id
stringlengths 36
40
| text
stringlengths 50
3.89M
| source
stringclasses 1
value | added
stringdate 2025-04-07 03:56:57
2025-04-07 04:04:00
| created
stringlengths 0
16
| metadata
dict |
|---|---|---|---|---|---|
001fcd9a-cd27-45e8-b23f-f447874e0974.91 | *2.2. BTEX Determination by HS-SPME-GC-MS Analysis*
| doab | 2025-04-07T03:56:58.097518 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 91
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.92 | 2.2.1. Standards and Reagents
The reference standard benzene (99.96%), toluene (99.93%), ethylbenzene (≥ 99.90%), p-xylene (99.90%), and benzene-d6 (99.99%), the latter used as internal standard (IS), were purchased from Sigma Aldrich. The reagents methanol and propylene glycol used for the preparation of standard/calibration solutions, as well as blank and samples solutions of a purity grade of more than 99%, were purchased from Sigma Aldrich.
| doab | 2025-04-07T03:56:58.097544 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 92
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.93 | 2.2.2. Standards and Calibration Solutions
For each compound, internal standard included, two standard stock solutions were preliminarily prepared. The first set of standard solutions (S1) was prepared diluting reference standards in methanol at a concentration of about 9 <sup>×</sup> 10<sup>7</sup> <sup>μ</sup>g/L. The second set of standard solutions (S2) was prepared diluting S1 solutions in methanol (1:100 dilution) to obtain a concentration of about 9 <sup>×</sup> 105 <sup>μ</sup>g/L.
**Table1.**E-liquidscompositionandinformation:manufacturingcountry,%ofthemaincomponents,characteristicflavorandnicotinecontent(expressedasmg/mL
*Atmosphere* **2020** , *11*, 374
\* mg/g; (-) means that information was not provided on product label.
as
Starting from S2 and with subsequent dilution with methanol, five solutions for each compound were prepared in the concentration range, approximately 20.0–450.0 μg/L (S3–S7). In order to simulate e-liquid basic composition, five matrix-matched calibration solutions for each compound were prepared by adding 100 μl of the corresponding S3–S7 solutions and 100 μl of benzene-d6 solution (S2 set) in a headspace (HS)-vial containing 1 ml of laboratory-made liquid (90% propylene glycol, 10% water). Similarly, a blank solution was also prepared by adding 100 μl of benzene-d6 solution (S2 set) and 100 μl of methanol in a HS-vial containing 1 ml of laboratory-made liquid (90% propylene glycol, 10% water). Both blank and matrix-matched calibration solutions were used for calibration, resulting in five concentration levels for each compound in the dynamic range between limit of quantification (LOQ) value and 45.0 μg/L.
| doab | 2025-04-07T03:56:58.097597 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 93
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.94 | 2.2.3. Sample Preparation
Sample preparation prior analysis required the dilution of an aliquot of refill liquid (1 ml) with 100 μl of methanol and 100 μl of IS solution. The dilution with a proper solvent is fundamental to avoid inhomogeneous samples due to the difficulty in sampling exact volumes of high viscosity fluids [35].
| doab | 2025-04-07T03:56:58.097702 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 94
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.95 | 2.2.4. HS-SPME-GC-MS Method Conditions and Performance Characteristics
The collection of BTEX in the volatile fraction of both calibration and sample solutions was carried out in 20-ml HS vials with magnetic screw caps provided with polytetrafluoroethylene (PTFE)/silicone septa (Agilent Technologies). BTEX were collected through adsorption onto the polydimethylsiloxane (PDMS) stationary phase-coated fused silica fiber (thickness 100 μm, length 1 cm) introduced into the sample vial. The PDMS fiber was left in the vial for 30 s at 50 ◦C. Mechanical stirring was performed for 5 s with a stirring speed of 500 rpm. Analyses were performed using a gas chromatograph (7890B Agilent Tecnologies, Santa Clara CA, USA) equipped with an automated sampler (Pal System, CTC Analytics AG, Zwingen, Switzerland), a split/splitless injector and a single-quad mass spectrometer (5977A Agilent Technologies, Santa Clara CA, USA). Once incubation was completed, the heated gas-tight syringe containing the fiber was automatically transferred into the GC injector via the automated sampler and BTEX were thermally desorbed at 250 ◦C for 300 s and injected into the GC column in split injection mode (split ratio 1:10). Separation was performed on capillary column semivolatiles, 30 m × 0.25 mm, i.d. 0.25 μm film thickness (Phenomenex). Helium (purity ≥ 99.999%) was applied as carrier gas at a constant flow rate of 1 ml/min. The GC oven temperature program used for optimal separation was: 40 ◦C for 2 min, ramped 8 ◦C/min up to 80 ◦C, then ramped 60 ◦C/min up to 250 ◦C. Transfer line and ion source temperatures were kept at 260 ◦C and 270 ◦C, respectively. The mass spectrometer was operated in electron impact (EI) ionization mode (70 eV). Identification of BTEX was based on comparison of the obtained mass spectra with those included in the National Institute of Standards and Technology (NIST) library (MassHunter software) and considered positive by library search match >800 for both forward and reverse matching. Further criteria for compounds identification were: (a) the matching of relative retention times (tR) with those of the authentic standards within the allowed deviation of ± 0.05 min; and (b) the matching of ion ratios collected with those of the authentic standards within a tolerance of ± 20%. Quantification was performed in a selected ion monitoring mode (SIM). One quantifier ion and two qualifier ions were selected for each compound on the basis of their selectivity and abundance: 79 m/z as quantifier ion and 51 and 39 m/z as qualifier ions for benzene; 91 m/z as quantifier ion and 65 and 39 m/z as qualifier ions for toluene; and 91 m/z as quantifier ion and 106 and 51 m/z as qualifier ions for ethylbenzene and xylenes. Five point matrix-matched calibration curves were constructed for quantification (r<sup>2</sup> > 0.995) reporting compound/benzene-d6 quantifier ion peak areas ratio vs amount ratio. Calibration curves were in the range 2.6–41.6 μg/L for benzene, 2.7–43.2 μg/L for toluene and xylenes isomers and 2.8–44.8 μg/L for ethylbenzene. The xylenes isomers were quantified on the basis of p-xylene response factor (e.g., p-xylene calibration curve) and reported as sum in Table 2. Chromatograms of a blank sample and a sample spiked with the BTEX standard solution (calibration level 3) were
compared in Figure S1 (Supplementary Material, Figure S1). The main performance characteristics of the HS-SPME-GC-MS method were also evaluated. Linearity was calculated on the basis of three sets of replicates for each calibration level on three different days. As for the results, all matrix-matched calibration curves were linear over the set concentration ranges: relative accuracy (%) for each point was within the <sup>±</sup> 5% of the expected concentrations, and all coefficients of determination (r2) were >0.995. Selectivity/specificity was assessed directly onto the chromatograms obtained from the blank and from spiked matrices. The occurrence of possible extra peaks was tested by monitoring in SIM mode qualifier and quantifier ions characteristic for each investigated compound onto the blank matrix chromatograms, within the retention time window expected for the analyte elution. Limit of detection (LOD) and LOQ values were assessed in the spiked matrix by determining the lowest concentration of the analytes that resulted in a signal-to-noise (S/N) ratio of ≥ 3 and ≥ 10, respectively. LOD values were 1.4 μg/L for benzene and toluene, 1.5 μg/L for xylenes, and 1.6 μg/L for ethylbenzene. LOQ values were 2.6 μg/L for benzene, 2.7 μg/L for toluene and xylenes and 2.8 μg/L for ethylbenzene. Repeatability expressed as intra-day coefficients of variation (CV%) was evaluated on a set of results (n = 6 replicates) obtained for each analyte at three validation levels (i.e., LOQ values; 10.4 μg/L for benzene, 10.8 μg/L for toluene and xylenes, 11.2 μg/L for ethylbenzene; 41.6 μg/L for benzene, 43.2 μg/L for toluene and xylenes and 44.8 μg/L for ethylbenzene). Intra-day CV% values were 1.2–4.5% for benzene, 1.2–9.9% for toluene, 3.2–10.9% for ethylbenzene and 2.8–11.4% for xylenes. Intermediate precision (expressed as inter-day CV%) and recovery were calculated by analyzing the series within the three different days (n = 18 replicates). Inter-day CV% values were 5.1–15.3% for benzene, 6.6–10.0% for toluene, 8.8–14.6% for ethylbenzene and 9.4–15.4% for xylenes. Finally, recoveries were in the range of 96.6–113.0%.
| doab | 2025-04-07T03:56:58.097745 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 95
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.96 | *2.3. Identification of Flavoring Additives by GC-MS-O Analysis*
GC-MS-O methodology was revealed to be a powerful approach for accurate identification of volatile odor-active compounds in high-level complexity matrices through coupling traditional chromatographic analysis with human sensory perception [36–38]. For this reason, GC-MS-O methodology was applied in the present study, allowing us to accurately identify, on a limited number of e-liquids, the odor-active compounds responsible for the overall flavor perceived or of specific flavor notes.
| doab | 2025-04-07T03:56:58.098085 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 96
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.97 | 2.3.1. Sample Selection and Preparation
The e-liquids subjected to the in-depth investigation were e-liquids with ID A 1-5 manufactured in China, with medium-high nicotine content and characterized by flavors covering different categories, from tobacco to fruits (Table 1). The aforementioned e-liquids were chosen for further study on the basis of collected data from BTEX investigation that highlighted high level of contamination. Moreover, during the preliminary survey and e-liquid selection made by the National Institute of Health, the brand A was already considered worthy of particular attention due to previous precautionary seizing actions made by Italian authorities and financial police. The preparation of the gaseous sample for GC-MS-O analysis starting from e-liquid formulation involved the use of the Adsorbent Tube Injector System device (ATIS™, Supelco). Before gaseous sample preparation, 250 μl of each e-liquid was preliminarily diluted, adding 250 μl of methanol, resulting in a solution with final volume of 500 μl. An aliquot (100 μL) of the obtained solution was injected by a syringe through the septum of the ATIS injection glassware and the volatile fraction was conveyed by ultrapure air flow (50 mL/min) into a collecting bag (Nalophan®), connected at the outlet of the injection glassware, resulting in a gaseous sample with a final volume of 2 L. The temperature, controlled by a thermometer inserted into the heating block, was set at 120◦C. As a result, only the volatile fraction was collected into the bag, avoiding the vaporization of the high-boiling point fraction composed by propylene glycol and glycerol that would have resulted in two broad chromatographic peaks in the GC chromatogram.
| doab | 2025-04-07T03:56:58.098135 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 97
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.98 | 2.3.2. GC-MS-O Analysis Conditions
The VOCs collected were analyzed using an air sampler-thermal desorber integrated system (UNITY 2™Markes International Ltd, Llantrisant, UK) connected to a gas chromatograph (7890 Agilent Technologies, Santa Clara CA, USA) equipped with an Olfactory Detection Port (ODP 3 Gerstel GmbH&Co, Mülheim an der Ruhr, Germany) and a single-quad mass spectrometer (5975 Agilent Technologies, Santa Clara CA, USA). The collection of VOCs onto the sorbent-pack focusing trap at −10◦C of the desorption system UNITY2™ was performed by connecting the Nalophan bag to the inlet port of the automated air sampling device. The cold trap was flash heated to 300 ◦C and the compounds were transferred via the heated transfer line (200 ◦C) to the GC column and to the ODP port. The chromatographic separation was performed on a HP5-MS capillary column (30m × 250μm × 0.25μm). Carrier gas (Helium) flow was controlled by constant pressure and equal to 1.7 ml/min. The GC oven temperature program was set as follows: from 37 ◦C up to 100 ◦C at 3.5 ◦C/min (ramp 1); and from 100 ◦C up to 250 ◦C at 15 ◦C/min (ramp 2). After the GC separation, the column flow was split into two parts (ratio 1:1), one part was connected to the MS detector and the other one to ODP. The transfer line connecting the GC column and MS detector was kept at 250 ◦C. The mass spectrometer was operated in electron impact (EI) ionization mode (70eV) in the mass range 20–250 m/z. The effluent from the capillary column was connected to the ODP port through an uncoated transfer line (deactivated silica capillaries), constantly heated to prevent compounds condensation. Two trained panelists, one male and one female (24 years old), were asked to sniff in the conical ODP simultaneously with the GC run, indicating exactly when they start and stop perceiving the odor and providing a qualitative description of the odor (using suitable descriptors) [36] and odor intensity based on an intensity scale from 0 (no odor perceived) to 4 (strong odor). Auxiliary air (make-up gas) was added to the GC effluent to prevent the assessors' nose mucous membranes drying, which may potentially cause discomfort, especially in extended analysis sessions. The panelists involved in the present study had previously been selected according to a standardized procedure used for the panel selection in Dynamic Olfactometry, the official methodology for odor emissions assessment standardized by a European technical law (EN 13725/2003) [39]. The standardized procedure provides for individuals with average olfactory perception sensitivity that constitute a representative sample of the human population. The screening was performed evaluating the response to the most used reference gas, 1-butanol. Only assessors who fulfilled predetermined repeatability and accuracy criteria were selected as panelists. The identification of flavoring additives and other VOCs in e-liquid formulation was performed by comparing the mass spectra obtained with those listed in the NIST library (Agilent Technologies). It was considered valid when the confidence rating of mass spectra comparison was superior or equal to 95%. The attribution was further confirmed using the retention times of authentic compounds. Before GC-MS-O sessions, panelists were asked to carry out preliminary sensory tests by sniffing and vaping the liquid formulations. This preliminary approach revealed to be useful in appreciating discrepancies between the flavors reported on e-liquid labels and the overall flavor perceived by panelists' noses and mouths (see Section 3.2 in results section).
| doab | 2025-04-07T03:56:58.098261 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 98
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.100 | *3.1. Quantitative Analysis: BTEX Contamination of the Investigated E-Liquids*
Single and total BTEX concentrations, expressed in μg/L, are reported in Table 2. As shown, most of the e-liquids investigated in the present study (85/97) were revealed to be affected, to a lesser or greater extent, by BTEX contamination. Only a few exceptions were observed with BTEX levels below the LOQ of the analytical methodology applied. Across all of the brands investigated (ID A-N, Table 1), concentration levels ranged from 2.7 μg/L to 30,200.0 μg/L for benzene, from 1.9 μg/L to 447.8 μg/L for ethylbenzene, from 1.9 μg/L to 1,648.4 μg/L for toluene and, finally, from 1.7 μg/L to 574.2 μg/L for m,p,o-xylenes. HS-SPME-GC-MS analysis of e-liquids with ID A (1-5), manufactured in China, highlighted a relevant contamination by BTEX with concentration levels up to four order of magnitude
higher than those determined in all the other investigated e-liquids, regardless of the manufacturing country and the chemical composition. More specifically, within brand A, benzene concentration levels ranged from 7,200.0 μg/L (sample 4-A) to 30,200.0 μg/L (sample 3-A), toluene concentration levels ranged from 764.4 μg/L (sample 1-A) to 1,648.4 μg/L (sample 4-A), ethylbenzene concentration levels ranged from 187.9 μg/L (sample 1-A) to 447.8 μg/L (sample 4-A) and, finally, m,p,o-xylenes concentration levels ranged from 201.8 μg/L (1-A) to 574.2 μg/L (sample 5-A). Moreover, making a comparison among samples ID A in terms of BTEX total concentration, it is possible to observe that 3-A shows the highest BTEX total concentration, equal to 32,151.1 μg/L. The comparison between samples ID A with all the other samples under investigation (ID B-N) revealed that benzene concentrations in 1-5 A samples were between one and four orders of magnitude higher than those determined in all the other e-liquids. Moreover, toluene concentrations in 1-5 A samples were up to three order of magnitude higher than those determined in all the other e-liquids, whilst ethylbenzene and m,p,o-xylenes were up to two order of magnitude higher. Benzene concentrations in 1–5 A samples were higher than toluene concentrations (from 4 to 22 times higher), a finding that was not observed for all the other samples characterized by toluene concentrations higher than benzene concentrations, with very few exceptions. To mention some examples, e-liquids with ID E and F manufactured in China showed toluene concentrations ranging from 20.7 μg/L to 96.2 μg/L and from 6.8 μg/L to 385.9 μg/L, respectively, in both cases one up to two order of magnitude higher than benzene concentrations. As already mentioned, some of the samples investigated were not affected by BTEX contamination. It is possible to observe that in most of the samples C (i.e., 1,2,3,8,10,12,14,15 and 17) and in samples 5D, G5, G6 the presence of BTEX was not detected at all with all concentration levels below the LOQ of the analytical methodology applied. Therefore the samples with ID C manufactured in Italy were revealed to be the highest quality e-liquids among all the tested samples. On the contrary, across samples with ID B-N, the highest BTEX total concentrations were associated with samples belonging to the batch with ID F (manufacture country China) with samples 12-F and 17-F showing the highest values, equal to 739.2 μg/L and 743.8 μg/L respectively. Therefore, it is possible to state that the highest BTEX contamination was observed in e-liquids belonging to two different brands (A and F), both of Chinese origin. Another important observation is that the highest BTEX total concentrations observed for most of the brands were associated with e-liquids characterized by mint flavor (brands B, F and L) and tobacco flavor (brands D, E, F and I).
**Table 2.** Benzene, toluene, ethylbenzene and xylenes (BTEX) concentration (expressed in μg/L) in the investigated e-liquids.
**Table 2.** *Cont.*
| doab | 2025-04-07T03:56:58.098510 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 100
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.101 | *3.2. GC-MS-O Qualitative Analysis: Identification of Flavoring Additives*
The sensory evaluation report by GC-MS-O analysis of e-liquids ID A 1–5 is shown in Table 3. Molecular formula, CAS number and retention time (TR), expressed in minutes, of identified odor-active compounds, as well as the intensity of the odor perceived and the associated qualitative description provided by both trained panelists, are reported. GC/MS-O analysis of the sample 1-A with labelled flavor Coca cola allowed to distinctly identify 4 odor-active compounds: ethoxyethane, 2-ethoxybutane, camphene, and γ-terpinene. In more detail, the integration of chromatographic data with sensory perception revealed that the first odorous stimulus perceived by both assessors with intensity 3 (clear odor) and qualitatively described with the descriptor 'sweet' was associated with ethoxyethane eluted at 2.8 min. The odor-active compounds 2-ethoxybutane and γ-terpinene, eluted at 4.8 and 18.1 min respectively, were associated with the characteristic flavor of coca cola beverage and related to the overall flavor perceived during the preliminary odor test with the refill liquid. More specifically, 2-ethoxybutane was perceived by both panelists with intensity 3 and described as coca cola-like flavor while γ-terpinene was perceived by both the panelists with intensity 2, described with the descriptor 'bitter' and referred to the bitter aftertaste of coca cola. Another odor-active compound detected at the olfactory port and chromatographically identified was camphene, perceived by both assessors with intensity 2 and associated with citrus and fresh notes. The odor-active compounds 2-ethoxybutane, camphene and γ-terpinene are all classified by FEMA as flavoring agents with a specific flavor profile. 2-ethoxybutane is associated with the flavor profile 'floral' while camphene and γ-terpinene to the flavor profile 'camphor/oil' and 'bitter/citrus' respectively. Other sources e.g., The Good Scents Company (TGSC) Information System reports a more detailed flavor profile of camphene including minty, fresh, woody and citrus notes depending on the concentration confirming, in part, the assessors' olfactory perception. Finally, as shown in Table 3, two odorous stimuli although distinctly perceived at the olfactory port approximately at 8.9 and 21.3 min were not identified due to chromatographic peaks not sufficiently intense to allow accurate identification. The lack of clear correspondence between sensory perception and chromatographic data highlights that, despite the potentialities of GC-MS-O technique, in certain cases the sensory perception of human nose is more sensitive than the analytical detection as reported by Plutowska et al., 2008 [40]. The GC-MS-O analysis of e-liquid 2-A with the characteristic kiwi flavor resulted in the identification of seven odor-active compounds. Most of the odorous stimuli were qualitatively described by assessors with the odor descriptors 'sweet' and 'fruity'. The odor-active compounds identified, in order of chromatographic elution, were: ethoxy ethane (sweet, 2.8 min), ethyl acetate (aromatic/alcoholic, 3.6 min), 2-ethoxybutane (sweet/fruity, 4.8 min), methyl butanoate (fruity, 5.3 min), ethyl butanoate (fruity, 7.6 min), ethyl 2-methyl butanoate (fruity, 9.3 min), and methyl hexanoate (fruity, 12.3 min). Two were in common with e-liquid 1-A, i.e., ethoxy ethane and 2-ethoxy butane perceived by both assessors with intensity 3 and 2, respectively. The esters methyl butanoate, ethyl butanoate, ethyl 2-methylbutanoate, and methyl hexanoate are odor-active compounds with fruity attributes and represent a characteristic portion of the volatile aroma profile of fruits. They are also classified by FEMA as flavoring agents and are primarily used to impart fruity flavor in foods and beverages. Ethyl acetate is also included in the FEMA list of flavoring agents (with specification as food additive, carrier solvent) but its flavor profile is based on aromatic, brandy, and grape odor notes. Among the 'sweet' and 'fruity' odorous stimuli, both the assessors clearly indicated the one associated with the characteristic kiwi flavor, with odor intensity equal to 3. Comparing GC-MS results with the sensory response provided by both the panelists, ethyl 2-methyl butanoate was identified as the odor-active compound responsible of the kiwi flavor of the refill. This specific ester has been already identified in previous investigations by GC-MS and GC-MS-O as the key contributor of the aroma profile of several fruits such as pineapples [41], strawberries [42], cranberries [43] and melons [44]. The preliminary sensory tests (e.g., sniffing and vaping) on e-liquids 3-A, 4-A and 5-A, performed by both the panelists before GC-MS-O analytical sessions, allowed to appreciate a significant discrepancy between the flavor reported on the label and the overall flavor perceived. E-liquids 3-A and 4-A labels 'Davidoff' and 'Green USA mix' referred to tobacco brands
whilst e-liquid 5-A label reported 'cigar' flavor. In all three cases, the overall flavor coming from e-liquids vaporization should have simulated the characteristic notes of the tobacco leaves aroma (i.e., woody, leather). Instead, the qualitative description provided by both assessors highlighted that the overall e-liquids flavors were dominated by sweet and caramel-like notes with the only exception of e-liquid 5-A that in addition was characterized by distinct woody notes. GC/MS-O analysis of sample 3-A ('Davidoff' flavor) allowed to confirm the role of ethoxyethane in giving the formulation a characteristic sweet and pleasant flavor. Moreover, the odor-active compound found to be the key contributor to the caramel notes of the overall flavor was 2,3-butanedione (or diacetyl), whose relevance as a flavoring additive will be deeply discussed in Section 4. Similarly to samples 1-A and 2-A, other odorous stimuli perceived approximately at 17.6 and 20.4 min and resembling tobacco flavor were associated with low intensity chromatographic peaks and, as a result, the tentative attribution was not allowed. At this regard, it has been already highlighted in Tierney et al., 2015 that the majority of tobacco flavored liquids were found to contain confectionary flavor chemicals instead of tobacco extracts therefore it is likely that the flavor chemicals pattern (i.e., benzyl alcohol, vanillin, ethylacetate, maltol) included in the formulations for resembling tobacco flavor is not necessarily what is expected to be found in a tobacco extract [45]. Considerations made for sample 3-A are relevant also for sample 4-A ('Green USA mix' flavor). Ethoxyethane and diacetyl were also detected in sample 4-A and associated, similarly with sample 3-A, to sweet and caramel-like flavor notes respectively. In addition, ethoxybutane was identified and associated with sweet flavor notes. The attribution for other odorous stimuli perceived during the GC/MS run, approximately at 8.7, 11.6 and 17.6 min (the latter similarly with sample 3-A), was not successful due to low intensity chromatographic peaks. More specifically, in addition to tobacco-like flavor, hearbaceous and grass/mint notes were perceived by assessors and this perception was considered reliable taking into account that, at least in principle, the formulation 'Green USA mix' should have simulated menthol-tobacco cigarettes and its characteristic menthol and herbaceous flavor notes. A comprehensive list of flavoring additives was obtained for sample 5-A ('cigar' flavor). Ethoxyethane and 2-ethoxybutane (both perceived with intensity 2) were confirmed as key contributors for sweet flavor notes while diacetyl (perceived with intensity 3) responsible for the caramel-like flavor. An interesting GC-MS-O outcome, allowing us to characterize the odor profile of the sample 5-A in a more distinctive way, was the identification of three odor-active compounds, perceived with odor intensity ranging from 1 to 2: α-terpinene (woody, 16.3 min), α-phellandrene (woody, 18.3 min), and α-terpinolene (woody/pine, 19.4 min). They all are classified as flavoring agents by FEMA: the associated flavor profile varies from woody, fresh, citrus, and spice notes in the case of α-phellandrene to pine flavor notes in the case of α-terpinolene. Their inclusion in the liquid formulation is therefore related to the intention of enriching the overall flavor profile of the product with woody and pine flavor notes with the purpose to simulate as closely as possible the cigar flavor. Finally, the integration of sensory perception and GC-MS chromatographic data failed in the identification of the odor-active compound perceived by both evaluators as responsible for the tobacco-like and burnt flavor, similar to what was previously observed for the sample 3-A.
| doab | 2025-04-07T03:56:58.098722 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 101
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.103 | *4.1. Discussion on BTEX Results*
HS-SPME-GC-MS analysis of 97 e-liquids highlighted BTEX contamination. Experimental data obtained suggest that, during the period 2013–2015, contaminated e-liquids were commercially available on the EU market, particularly e-liquids imported into EU member states and manufactured in China. Taking into account all of the data obtained, no correlation was found between BTEX contamination levels and nicotine content, nor nicotine presence. The variability observed in BTEX contamination levels from one brand to another one is therefore likely to be related to the variability in contamination level of the basic components (i.e., propylene glycol and glycerol) and/or the flavoring additives included. In addition, the variability in BTEX contamination levels observed within the same brand is likely to be related to the flavoring additives used, and in the specific case of samples 10, 11 and 12 C, given the same flavor and nicotine content, to the contamination of basic components used in the production process of different batches. According to Regulation (EC) No 1272/2008 on Classification, Labelling and Packaging of substances and mixtures (CLP), benzene, toluene, ethylbenzene and m,o,p-xylenes are included in Annex VI, Table 3. Benzene is classified as carcinogenic for humans (Carc. 1A, H350: May cause cancer by inhalation), mutagenic (Muta. 1B, H340: May cause genetic defects), and represents a hazard when inhaled (Asp. Tox 1, H304: May be fatal if swallowed and enters airways; STOT RE 1, H372: causes damage to organs through prolonged and repeated exposure) [31]. Toluene is classified as reprotoxic (Repr. 2, H361d: Suspected of damaging the unborn child) and represents a hazard when inhaled (Asp.Tox 1, H304: May be fatal if swallowed and enters airways). Ethylbenzene and xylenes are both classified as follows: Acute tox. 4, H332: harmful if inhaled. Given all the information on toxicity classification reported above, more attention has necessarily to be paid to benzene, a human mutagenic and genotoxic carcinogen, detected in some e-liquids at high concentration levels. Therefore, an in-depth analysis of potential health effects due to inhalation exposure to benzene is due. Epidemiological studies over the years have provided evidence of a causal relationship between chronic inhalation exposure to benzene and serious adverse health effects and diseases, from non-cancer health effects (i.e., hematologic diseases and/or functional aberrations of immune, nervous, endocrine systems) to cancer (i.e., myeloid leukemia, non-Hodgkins lymphoma) [46]. Numerous studies have demonstrated that benzene metabolites, especially p-benzoquinone, are involved in the progression from cytotoxicity to carcinogenicity, as they activate oxygenated radical species able to cause DNA damage [47]. It has been estimated that approximately 50% of the quantity of inhaled benzene is adsorbed into the human body. Once introduced into the human body through the respiratory apparatus, benzene is preferentially adsorbed in fat-rich tissues (i.e., fat and bone marrow), owing to its lipophilic nature. Great concern about potential health hazards has been historically linked to occupational exposure (where higher benzene concentrations than in general environments are likely to be encountered) but knowledge on the issue, acquired over the years, has led the scientists and epidemiologists to be more and more focused on health effects induced by long term exposure of the general population to low concentrations of benzene. Although benzene is recognized as a 'non-threshold carcinogen' on the basis of the assumption that any exposure may result in some increase of risk, in the present study the carcinogenic risk related to the inhalation exposure to benzene resulting from the consumption of e-liquids affected by the highest contamination (brand A) has been estimated.
As reported in the results section, across all 97 e-liquids tested, benzene concentration levels ranged from 2.7 μg/L (in samples 3-B and 6-D, both produced in Italy) to 30,200.0 μg/L (sample 3-A produced in China). This means that, if we consider the daily average consumption of e-liquids by a regular vaper approximately equal to 3 ml per day [48], the total amount of benzene potentially inhaled by the vaper within one day would have ranged from 0.0081 μg to 90.6 μg. For the most contaminated Chinese brand (brand A) the total amount of daily inhaled benzene with 3 ml e-liquid consumption would have varied in the range 21.6–90.6 μg. Taking into account a regular vaper represented by an adult person with an average body weight of 60 kg, the daily consumption of brand A e-liquids would
result in benzene exposure of 0.00036–0.00151 mg/kg/day. A carcinogenic risk assessment for benzene may be performed comparing the estimated exposure with derived minimal effect level (DMEL) value, representing the level of exposure expressed as mg/kg/day below which the risk level of cancer is considered tolerable/acceptable (indicative tolerable risk level is 10-5 associated with a life-time risk for cancer of 1 per 100000 exposed individuals). The DMEL value for benzene, derived from reference values reported on Integrated Risk Information System (IRIS) website of United States Environmental Protection Agency (USEPA), is 0.0000182 mg/kg/day [49]. The comparison exposure-DMEL allows to point out that the daily consumption of Chinese e-liquids belonging to brand A would have resulted in a serious inhalation exposure scenario for active users with a risk level of cancer that is not acceptable. These results are of particular concern, also in light of the World Health Organization (WHO) guidelines for indoor air quality, published in 2010, where it is clearly stated that 'no safe level of exposure to benzene can be recommended' and that 'from a practical standpoint, it is expedient to reduce exposure levels to as low as possible' reducing or eliminating activities and materials that may release it [50].
| doab | 2025-04-07T03:56:58.099327 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 103
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.104 | *4.2. Discussion of Flavoring Additives Results*
Among the flavoring additives identified, diacetyl is certainly worthy of an in-depth analysis. Diacetyl is a volatile α-diketone and is a natural constituent of many regularly consumed foods (i.e., dairy products, fruits, coffee). Due to its flavor characteristics, it is widely used in the food manufacturing industry as a flavoring additive. It is added to a wide selection of foods and beverages to mainly impart butter and caramel taste and smell, depending on the concentration used. Its use in the food manufacturing industry is approved by competent governmental bodies such as U.S. Food and Drug Administration (U.S. FDA) and the National Institute for Occupational Safety and Health (NIOSH) and is currently authorized in EU member states according to EU Regulation No 872/2012. The potential risks for consumers health associated with the dietary exposure have been deeply evaluated over the years. As a result of safety evaluations, diacetyl has been determined to be 'generally recognized as safe' (e.g., GRAS) by the FEMA Expert Panel, and has been included in the FEMA GRAS list of authorized flavoring substances [51]. The European Food Safety Authority was also asked to take a position on the issue and the final opinion was that, on the basis of the safety evaluations carried out so far, the use of diacetyl in food is of no safety concern for humans. In this regard, however, it is important to point out that toxicological evaluations used to approve and support diacetyl as a flavoring additive in foods are related to ingestion, and therefore do not provide assurance of safety when other routes of exposure are involved, such as inhalation. In the early 2000s, concerns were raised with respect to potential toxicity for humans associated with inhalation exposure to diacetyl following the reported cases of a severe obstructive lung disease in diacetyl-exposed workers at microwave popcorn manufacturing plants in USA [52]. Preliminary evidence of an association between the occupational exposure to diacetyl and adverse effects on human respiratory apparatus has been reported by Kreiss et al., from a decline in respiratory function to development of a rare irreversible lung disease characterized by fixed airflow obstruction, called bronchiolitis obliterans [52]. Extensive scientific research on diacetyl has been carried out from then both confirming preliminary hypothesis on exposure-occurrence of lung disease association and adding new relevant scientific data [53]. Recently published papers have highlighted both neurotoxicity and impairment of cilia function in human airway epithelium [54,55]. Therefore, in light of the knowledge progressively acquired, the inclusion of diacetyl as flavoring additive in the manufacturing process of liquid formulations for e-cigs has rapidly become a much-debated issue in the scientific community due to foreseeable toxicological implications from direct inhalation exposure. In reaction to this, a prompt response came from e-liquids manufacturers with the replacement of diacetyl with 2,3-pentanedione (acetylpropionyl), an α-diketone showing similar flavor properties, but this option was soon revealed to be unsuccessful when scientific data on acetylpropionyl toxicity started to be published [56]. Our findings, although related to a limited number of samples, are in line with the results obtained in previous investigations highlighting the presence of diacetyl in e-liquids commercially available in EU member states in the
pre-TPD implementation period and with characteristic flavors appealing to teenagers and young adults [19,20,30]. Farsalinos et al., 2015 analyzed both liquid and aerosol matrices of a total number of 159 samples purchased from 36 manufacturers and retailers in 7 different countries. Diacetyl was found in 74% of the samples investigated and in a large proportion of sweet-flavored e-liquids, with similar concentrations in both liquid and aerosol. The simultaneous presence of acetylpropionyl also suggested that, instead of being used as a replacement, acetylpropionyl is often used in conjunction with diacetyl. Further, the authors highlighted that, for 47% of diacetyl-containing e-liquids, the daily exposure level (μg/day) for vapers could be higher than NIOSH-defined safety limits for occupational exposure. Barhdadi et al. investigated 12 flavored e-liquids by applying the HS/GC-MS method, properly developed for the screening and quantification of diacetyl and acetylpropionyl in e-liquids. The samples were provided by the Belgium Federal Agency for Medicinal and Health Products and collected either upon inspections in vaping shops or through seizure activity by Belgian authorities in the period 2013–2015, similar to the present study. The authors reported that only two sweet-flavored e-liquids contained measurable amounts of diacetyl and the determined concentrations were 6.04 μg/g and 98.84 μg/g. Finally, 42 e-liquids selected from among the 14 most popular brands dominating both the USA and EU markets in 2013 were investigated by Varlet et al. in terms of chemical and biological constituents. Diacetyl was detected in three e-liquids, two of them characterized by tobacco flavors and one by candy flavor. Similarly to Farsalinos et al., comparison with the NIOSH safety limit was made, revealing that one tobacco flavored e-liquid that resulted diacetyl-positive could lead to exposure higher the recommended limit. Although approximate for estimating risk for e-cig users, the use of occupational exposure limits is affected by several limitations [19,57]. This approach has raised some resistance, mainly because occupational safety limits for toxicants, for instance for diacetyl, have been set for workers not for the general population and are related to inhalation exposure scenarios not applicable to e-cigs users. According to the authors' knowledge, other two studies carried out by Allen et al. in 2017 and Omayie et al. in 2019 have raised concerns about diacetyl, confirming its inclusion as flavoring additive in refill liquids for e-cigs (diacetyl detected in 39 of 51 tested refills and in 150 of 277 samples, respectively), but in both cases the investigated samples were considered dominating the current extra-EU market and therefore are not representative of the EU market before the implementation of TPD. To summarize, our findings on diacetyl, although related to a limited number of e-liquids manufactured in China and commercially available in the EU during the period 2013-2015, are in line with the results obtained in other investigations made on larger sets of samples representative of the EU market at that time. The only discrepancy on diacetyl presence detectable among the studies performed before the TPD implementation was reported by Girvalaki et al. in 2018. The authors evaluated the chemical composition of 122 e-liquids selected among the most commonly sold brands in 9 EU member states in mid-2016 before the TPD implementation. The result of this comprehensive investigation was a list of 177 compounds detected (e.g., flavoring additives and other VOCs), the majority with associated Globally Harmonized System of Classification and labeling of Chemicals (GHS) health hazard statements. Diacetyl, however, was not detected in the samples tested, and therefore not included in the list. This discrepancy between Girvalaki et al. and the other abovementioned studies may be related or to the different period of e-liquids selection (2013–2015 versus 2016), although both periods were before TP -implementation, when the first actions aimed to the progressive replacement/elimination of diacetyl started to be made on a voluntary basis by some EU manufacturers and importers, or it simply reflects the potential heterogeneity due to the multitude of samples commercially available on the EU market in the period of reference. To date, following the implementation of TPD in most EU member states in 2016, both manufacturers and importers are obliged to submit a notification to competent authorities reporting detailed information on refill liquids (Article 20) [13]. The notification must report the list of all the ingredients (including flavoring additives) contained in e-liquid formulations for e-cigs available on the market and indication of related quantities as well. It must be noted, however, that according to TPD, the use of diacetyl is neither explicitly prohibited nor subjected to restriction. In addition, due to difficulty in defining a
typical inhalation exposure scenario fitting all vapers habits (high variability in daily e-liquid amount consumed), there is no scientific consensus on the maximum allowed level of diacetyl in e-liquids. Therefore, to date, diacetyl use as a flavoring additive in e-liquids remains an open issue, suggesting not only that quality controls remain necessary, even in e-liquids labelled as diacetyl-free, but also that the potential solution at the EU level to ensure that e-liquids supplied to consumers are safe is to follow the direction of some EU member states that proposed the ban of diacetyl and other flavoring additives of concern [58].
| doab | 2025-04-07T03:56:58.099713 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 104
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.105 | **5. Conclusions**
In the present paper, results from a study on the chemical characterization of levels of BTEX in 97 e-liquids, representative of the EU market between 2013 and 2015 prior the implementation of TPD in most EU member states, are reported. To our knowledge, there have been very few studies focused on BTEX analysis in refill fluids and cartridges for e-cigs commercially available on the EU market in the pre-TPD implementation period. Therefore, although the e-liquids investigated may not be representative of the current EU market, our findings may represent a useful reference for the ongoing evaluation on the effectiveness of e-liquid safety and quality requirements under the current legislative framework. Most of the e-liquids investigated were revealed to be affected, to a lesser or greater extent, by BTEX contamination. Few exceptions were observed (12 of 97 samples). High variability in BTEX total concentration level was observed from one brand to another, ranging from 2.7 μg/L to 32,151.1 μg/L. The contamination is likely to be related to the contamination of propylene glycol and glycerol, and/or the flavoring additives used. No correlation was found between BTEX concentration levels and nicotine content/presence. Moreover, it was estimated that an inhalation exposure of very high concern would have occurred for active users vaping the most contaminated e-liquids (brand A), characterized by high concentration levels (7,200–30,200 μg/L) of benzene, a known human carcinogen. Our findings, therefore, point out that higher quality ingredients should have been used and that quality control on the formulations should have been applied prior their introduction on the EU market in 2013–2015 period. Further investigations carried out on a limited number of e-liquids aimed at the identification of flavoring additives through GC-MS-O application confirmed, in the reference period of the present study, the use of diacetyl, a flavoring additive approved for foods but associated with the onset of a severe lung disease when inhaled. This finding is in line with results obtained by other investigations made in the same period on a larger number of e-liquids sold in EU, highlighting the use of diacetyl in the e-liquid manufacturing industry due to poor awareness of the potential harm to humans. There are now sufficient toxicological data on the potential adverse effects of diacetyl and other flavoring chemicals when directly inhaled into the human airways, and therefore harmonized regulation at EU level on flavoring additives use in e-liquids, resulting in ban or restriction, should be fully addressed, in order to ensure health protection.
**Supplementary Materials:** The following are available online at http://www.mdpi.com/2073-4433/11/4/374/s1, Figure S1: Comparison of the Chromatograms of a blank sample and a sample spiked with the BTEX standard solution (calibration level 3).
**Author Contributions:** Conceptualization: J.P., R.D. and G.d.G.; methodology: J.P., C.A., M.F. and L.P.; investigation: J.P., C.A., M.F. and L.P.; data curation: J.P., C.A., M.F., A.D.G. and L.P.; writing—original draft preparation: J.P., A.D.G. and L.P.; supervision: R.D. and G.d.G. project administration: C.A., M.F., G.d.G. and R.D.; funding acquisition: G.d.G. and R.D. All authors have read and agreed to the published version of the manuscript.
**Funding:** The work has been carried with the coordination of the Italian National Institute of Health in the framework of the National project entitled 'New articles and new health risks: the electronic cigarette' (project code: CCM 2013, date of approval: 12 December 2013), financially supported by the Ministry of Health of Italy.
**Conflicts of Interest:** The authors declare no conflict of interest.
| doab | 2025-04-07T03:56:58.100289 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 105
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.107 | **The Relevance of Indoor Air Quality in Hospital Settings: From an Exclusively Biological Issue to a Global Approach in the Italian Context**
| doab | 2025-04-07T03:56:58.100534 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 107
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.108 | **Gaetano Settimo 1, Marco Gola 2,\* and Stefano Capolongo <sup>2</sup>**
Received: 2 March 2020; Accepted: 7 April 2020; Published: 8 April 2020
**Abstract:** In the context of the architectures for health, it is an utmost priority to operate a regular and continuous updating of quality, efficacy, and efficiency's processes. In fact, health promotion and prevention take place through a proper management and design of healing spaces, in particular with regard to the most sensitive users. In recent decades, there has been increasing attention to indoor air quality in healthcare facilities. Nowadays, this issue must involve the implementation of a series of appropriate interventions, with a global approach of prevention and reduction of risk factors on users' health, which allows, in addition to a correct management of hospital settings, the realization of concrete actions. To date, in Italy, despite the indoor air being taken in consideration in numerous activities and studies aimed at understanding both building hygiene and environmental aspects, the greatest difficulty is strongly related to the absence of an integrated national policy. The scope of the paper is to underline the relevance of indoor air quality in hospital settings, highlighting the need of procedures, protocols, and tools for strengthening and improving interventions for health prevention, protection, and promotion of users.
**Keywords:** indoor air quality; healthcare settings; chemical and biological pollution; quality improvement; Italian context
| doab | 2025-04-07T03:56:58.100569 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 108
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.109 | **1. The Relevance of Built Environment: The Case of Healing Spaces**
In a strategic field such as care and assistance, diagnostics, prevention, research, training, and safeguarding of public health by architectures for health (hospitals, community health centers, clinics and outpatient centers, etc.), both public and private ones, it is utmost a priority to operate a regular and continuous updating of quality's processes, efficacy, and efficiency of healthcare practices. The approach should apply to its entirety with prevention techniques, training, health education, and promotion activities, in relation to the needs for the health protection of users (both patients, visitors, and staff), with particular attention to the most sensitive and vulnerable groups in hospital settings [1–4].
In this scenario the healthcare facilities, affected by the requirement of promoting greater innovation and improving the quality of services and processes, have given rise to a considerable amount of concrete actions and interventions, such as the improvement of staff's training, exceeding and updating the level of organizational, management, and structural standards of healthcare [1,5,6]. In several ways they contribute not only to the efficiency of territorial assistance and care [2] but also to the dissemination of the value of individual and public health prevention, with a broad perspective of citizens' health status, increasing as a consequence the years of life [5,7,8].
In particular, in an Italian context, in order to correctly respond to the healthcare needs of the population, in the Health Pact for the years 2014–2016, signed by the Permanent Conference for the Relations between the State, the Regions, and the Autonomous Provinces, the state of health has been defined no longer as a source of cost, but as an economic and social investment, identifying a series of interventions to achieve and offer the best products for citizens' health and to promote the development of health and the competitiveness of the whole country [9]. This application has constituted a strategic and important opportunity to tackle some of the crucial and highly relevant issues of recent years with greater awareness, such as:
The methodologies for assessing the healthcare costs incurred by the various countries were developed by the Organization for Economic Cooperation and Development (OECD), in which 20% of the total health expenditure, quantified in the report "Tackling Wasteful Spending on Health", does not contribute to a real improvement in populations' health status [15]. For this reason, several authors highlight the importance to promote health through design actions in the built environment (urban health strategies, healthy indoor spaces, etc.) [16–19].
Moreover, in relation to the Italian case, with the Decree no. 50/2015—Regulation for hospital assistance, structural, technological, qualitative, and quantitative standards relating to healthcare are aimed at promoting the expansion of the areas, increasing hospitable features of the environments, safety and security, and real and adequate quality of care, which must be adopted to create the conditions to produce benefits and high quality of the entire National Health System (NHS) network [2,20].
| doab | 2025-04-07T03:56:58.100703 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 109
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.110 | **2. Design and Management Aspects that A**ff**ect Indoor Air in Hospital Settings**
In this evolutionary context, there has been growing attention to indoor air quality's issue in healthcare facilities, which, in order to satisfy primarily the requests of patients, healthcare users and workers, administrative and non-administrative staff, etc., have been affected to a series of new adjustments and design approaches (i.e., configuration and rationalization of spaces and flows, the use of specific products and materials, etc.) [21,22], structural and functional actions (i.e., requalification, restructuring, energy efficiency improvement, etc.) [23], engineering plants' system (i.e., optimizing the performance of the centralized heating and cooling systems, energy performances, etc.), [24] and management strategies (i.e., the correct daily management of the ventilations systems, the reduction of costs, accounting for consumption, etc.) [25], with the aim of expanding the services supplied, the quality of healthcare services, obtaining greater organizational and working flexibility, and attempting to reduce the economic costs of healthcare facilities [26]. Gola et al. have highlighted the factors that mostly affect a healing space, as Figure 1 synthetizes [40].
In all these healthcare environments for different needs, the healthcare and technical and administrative staffs, and the users (caregivers, elderly people, children, volunteers, students, visitors, outsourcing services' staffs, maintenance workers and suppliers, etc.)—some of them with reduced mobility, too—interact, stay, live, and work [27,28]. For this reason, specific prevention measures are necessary, considering the exposure of key actors (from the users to hospital staff), whose roles, knowledge and background, motivations, and individual relationships have changed and evolved, becoming increasingly an informed, active, and willing participation to collaborate for improving the environments' quality, services, and treatments. Their exposure takes on particular significance and importance both for the vulnerabilities of the users (i.e., patients with various pathologies, with an
acute health status, with different immune responses, people with disabilities elderly, etc.), and for the times of permanence in the hospital [29–33].
**Figure 1.** Factors that affect hospital environments.
In the specific case of the activities carried out in the healthcare facilities, it is essential to consider the close relationships between the behaviors and activities of medical e and technical-administrative staffs, and the different ones of patients, visitors, volunteers, students, professionals of external companies (i.e., cleaning, maintenance, suppliers, etc.), the quality of the spaces, and daily relationships with the organizational and management procedures of functional processes, that define the complex scenario of activities to be delivered [34,35]. The use of technological systems designed to perform and satisfy the various tasks in the best economic conditions, the technical furnishings, the level of use, the ordinary and extraordinary cleaning and sanitization activities (providing targeted actions according to the health status and the type of risk of patients, with different levels of contamination, and with microbiological monitoring), the maintenance, the procedures, and the organic management of the multiple routine prevention activities implemented and shared within the spaces, are all factors that contribute significantly to indoor air quality, and the health (this is even more concrete in view of the emergency period for SARS-CoV-2 virus that currently the population is experiencing) and satisfaction of all those users who attend the healing spaces [24,25,35].
In general, these interventions and initiatives have been adopted to address the significant change in healthcare needs, which affects the growth of requests for services and diagnostic treatments, as well as new fields of assistance and research, which require greater functionality of spaces, a reduction in the average length of hospitalization, the occupancy rate of beds, and inter-regional flows of healthcare mobility, overcoming social and territorial inequalities [36,37].
Specifically, on the operational level as regards the interventions carried out, it is necessary to highlight how often the choices of products and construction materials (i.e., paints, varnishes, etc.), finishing, (i.e., adhesives, silicones, etc.), furniture components (i.e., decors, curtains, etc.), products for cleaning and detergents for daily use, products for ordinary (methods and frequency that independently must always be adapted to the use of the area, to the flows of inpatients or medical staff, visitors, etc.) and extraordinary sanitization (i.e., use of more or less products concentrated, or not specific for cleaning surfaces, etc.), as well as engineering plant's management and maintenance activities (i.e., various air conditioning systems and centralized controlled mechanical ventilation systems), etc. were carried out in a disordered manner, without an adequate assessment of the emission behavior of pollutants from the materials and products used (i.e., VOCs—volatile organic compounds—and other substances emissions). In fact, the specificity and the protective value that the environments must respond to specific environmental conditions of use (i.e., temperature, relative humidity, air changes, etc.), the presence of patients, healthcare users, temporary visitors, volunteers, activities carried out by healthcare staff and not, and hygienic conditions of the environments depending on the health status, the type or risk of patients or, in general, of the daily flows (i.e., presence of microbial and fungal communities with a capacity for persistence, variability of concentration, and diversity in healthcare environments, which can generate an extension of the length of hospitalization stay, additional diagnostic and/or therapeutic interventions and additional costs, etc.) [38,39].
| doab | 2025-04-07T03:56:58.100986 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 110
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.111 | **3. Chemical and Biological Concentrations in Indoor Air in Healthcare Environments**
It should be underlined that, until a few years ago, in Italy, most of the activities and direct and indirect interventions of prevention and training were limited exclusively to select and identified healthcare environments with specific professional exposure to: chemical and biological agents (i.e., monitoring in the air of anesthetic gases in operating rooms, in laboratories dedicated to the preparation and administration of antiblastic drugs, in premises or areas of chemical sterilization, in histology and pathological anatomy departments for use of preservatives or disinfectants (i.e., formaldehyde, waste storage, and transport activities, etc.); ergonomic and physical factors (i.e., patient movement, sudden movements with efforts, critical or prolonged working posture, and in the administrative offices related to the workplace, etc.); video terminals (i.e., in administrative offices, call centers, back offices, departments, etc.); accidents (i.e., falls, etc.); psychosocial (i.e., excessive workload, stress and satisfaction levels, etc.); microclimatic factors such as temperature, relative humidity, air changes (both in health and administrative areas, etc.); implementation of programs of multidisciplinary hygiene surveillance and control such as those developed by the hospital infection committees for the control of infections, of the Supervisory Commissions, composed of a group of dedicated professional figures and with guidelines and protocols for the control of pollutants of biological origin (provided in compliance with ministerial acts), in order to prevent patient-related and non-healthcare staff and non-healthcare-related infections, which have always been a major concern for all hospitals [40–45].
For this reason, these aspects are increasingly integral components of the quality of services, therapies, healthcare services, activities and training, and information plans continuously provided, contributing to obtaining an effective and adequate indoor air quality, which responds to the main references elaborated for some time by the World Health Organization (WHO), and which currently constitute a valuable contribution worldwide.
In general, although the biological pollutants are constantly under analysis, they have already been studied and investigated by several research groups, and several countries have defined guidelines and very detailed protocols (that need to be improved more and more), such as Legionella, etc. [32,46].
Unlike the activities on biological compounds, investigations or monitoring activities of indoor air quality dedicated to the presence (or assessment) of the concentrations of chemical pollutants also to other environments have been carried out only recently and marginally, in some functional areas and environments of the hospital. Never before have such monitoring activities been brought to the attention of management by users, healthcare staff, etc., who complain of uncomfortable circumstances while living and working in the hospital settings or in carrying out their work activities that do not involve the use of chemical or biological agents [45,46]. Often at the operational level, these are requested that usually occur for complaints to situations related to an inadequate air exchange, the presence of new furnishings, the change of the room, during maintenance or renovation activities in specific areas and/or in punctual rooms, when the intended uses vary, when using cleaning and detergent products, or due to the inadequate or incorrect operation of the ventilation systems, etc. [45].
Therefore nowadays, this must entail the implementation of a series of appropriate and organized interventions (not limited to single and specific actions), with a global approach of prevention and reduction of risk factors on the health of all users, which allow, in addition to a correct management of the various environments of healthcare facilities, the realization of concrete actions on indoor air quality according to the priority principles and guidelines identified by WHO [47] and in part already listed as goals in various European and international programs of the prevention measures [12].
With particular attention to chemical pollutants, an examination of the current situation in the European Union (EU) shows that some Member States, such as France, Belgium, Finland, Portugal, Poland, and Lithuania have fully entered the quality of the indoor air in their national legislations with quantitative values (reference values, guidelines, etc.), and with practical guidelines which contain indications for the control, self-assessment sheets for identifying potential indoor sources (or close to the facilities), and the procedures for the development of indoor air monitoring, which are in many cases in line with the current WHO values published in 2009 and 2010 on the basis of the main scientific evidences [12].
In these countries, compliance with the legal requirements and the correct application of practical protocols remain one of the fundamental points for achieving good indoor air quality in the various healthcare environments [48]. In particular, France has foreseen a series of specific interventions including mandatory monitoring of indoor air quality in healthcare facilities as early as 2023 [49].
Until today, in Italy, despite being the quality of indoor air subjected to numerous activities and investigations aimed at understanding both the environmental and hygiene aspects, the greatest difficulty remains the absence of an integrated national policy about indoor air quality, with specific legislative references, which report the national references (i.e., guide values, references, etc.) and the rules for the data analysis of the results, and with documents that list the recommendations for an adequate management and evaluation of indoor air quality [50]. In the absence of national references, it is possible to use those present in the WHO documents related to indoor air quality or those in the legislation of other European countries or, by analogy, to other standards such as those relating to the ambient air for which specific legislative references have been issued on a limited number of pollutants, etc. [51].
There is no doubt that the current system of health prevention and protection laws has led to a confusion of language and knowledge that indeed has often confused and disoriented the practitioners, engaged in various capacities in the programs and evaluations in these environments and structures [37]. In this process of approach and strengthening of prevention actions, it is necessary to bring about a concrete harmonization, revision, innovation, updating and expansion on specific aspects, also to current standards [52]).
The aims and scope are to provide the procedures and tools necessary to strengthen, optimize, and improve interventions for the prevention, protection, and promotion of the health of users in healthcare environments that represent one of the priority objectives of the NHS's strategy in the prevention programs, with monitoring activities within the healing spaces [38,53].
Additionally, with regard to biological pollutants, although there are recommendations from international agencies and institutions, there are no legislative values or standards for the microbiological parameters of indoor air quality due to the difficulties encountered in correlating the data of the microbiological tests with those of the epidemiological investigations [46].
| doab | 2025-04-07T03:56:58.101542 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 111
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.112 | **4. Future Perspectives**
In conclusion, hospital facilities are complex constructions, with very different needs, users, and requirements compared to other building facilities, and they work 24/7, all year long. For this reason, every action should be assessed in relation to their performances and the aim to interrupt medical activities as little as possible.
It is clever that indoor air quality is a very broad topic in which any variable can affect the performances of air in indoor environments both in biological and chemical terms, as one of the goals of UN 2030—United Nations Sustainable Development. As several authors states, adequate design and management strategies, in relation to different procedures, can decrease or increase the quality performances of the healthcare environments.
The Scientific Community should continue to investigate the issue, define smart and efficient procedures, protocols for monitoring and tools, instrumentations for the investigations, etc. for strengthening and improving interventions, and guaranteeing protection and promotion of users. The new challenge should investigate the correlations between the chemical and biological pollutants and their effects in indoor air and the quality of the healthcare facility.
**Author Contributions:** Conceptualization, G.S. and M.G.; writing—original draft preparation, M.G. and G.S.; writing—review and editing, M.G.; supervision, S.C. All authors have read and agreed to the published version of the manuscript.
**Funding:** This research received no external funding.
**Conflicts of Interest:** The authors declare no conflict of interest.
| doab | 2025-04-07T03:56:58.102355 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 112
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.114 | *Article* **Indoor Comfort and Symptomatology in Non-University Educational Buildings: Occupants' Perception**
| doab | 2025-04-07T03:56:58.102488 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 114
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.115 | **Miguel Ángel Campano-Laborda, Samuel Domínguez-Amarillo, Jesica Fernández-Agüera \* and Ignacio Acosta**
Instituto Universitario de Arquitectura y Ciencias de la Construcción, Escuela Técnica Superior de Arquitectura, Universidad de Sevilla, 41012 Seville, Spain; [email protected] (M.Á.C.-L.); [email protected] (S.D.-A.); [email protected] (I.A.)
**\*** Correspondence: [email protected]; Tel.: +34-954-557-024
Received: 14 February 2020; Accepted: 2 April 2020; Published: 7 April 2020
**Abstract:** The indoor environment in non-university classrooms is one of the most analyzed problems in the thermal comfort and indoor air quality (IAQ) areas. Traditional schools in southern Europe are usually equipped with heating-only systems and naturally ventilated, but climate change processes are both progressively increasing average temperatures and lengthening the warm periods. In addition, air renewal is relayed in these buildings to uncontrolled infiltration and windows' operation, but urban environmental pollution is exacerbating allergies and respiratory conditions among the youth population. In this way, this exposure has a significant effect on both the academic performance and the general health of the users. Thus, the analysis of the occupants' noticed symptoms and their perception of the indoor environment is identified as a potential complementary tool to a more comprehensive indoor comfort assessment. The research presents an analysis based on environmental sensation votes, perception, and indoor-related symptoms described by students during lessons contrasted with physical and measured parameters and operational scenarios. This methodology is applied to 47 case studies in naturally ventilated classrooms in southern Europe. The main conclusions are related to the direct influence of windows' operation on symptoms like tiredness, as well as the low impact of CO2 concentration variance on symptomatology because they usually exceeded recommended levels. In addition, this work found a relationship between symptoms under study with temperature values and the environmental perception votes, and the special impact of the lack of suitable ventilation and air purifier systems together with the inadequacy of current thermal systems.
**Keywords:** educational buildings; schools; field measurements; ventilation; indoor air quality (IAQ); thermal comfort; thermal perception; health symptoms; CO2 concentration; air infiltration
| doab | 2025-04-07T03:56:58.102521 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 115
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.116 | **1. Introduction**
#### *1.1. State of the Art*
Non-university educational buildings are one of the most widespread building typologies, in which teenagers, a more sensitive population than adults and with specific different thermal preferences due to their different metabolic rate values [1–4], spend more than 25% of their day time during winter and midseasons. Thus, indoor environment in non-university classrooms is one of the most analyzed problems in the thermal comfort and indoor air quality (IAQ) areas [5], being widely studied for cold [6–10], mild [11–17], and warm climates [18].
Traditional schools in southern Europe solve thermal control basically by heating-only systems (without mechanical ventilation), relying on air renewal to uncontrolled infiltration and users' frequent windows' operation, much more than usually found in central and northern Europe. This develops a behavior that could be defined as hybrid or mixed mode, with thermal systems operated and with a significant part of the time with the windows open. In addition, climate change processes are progressively lengthening the warm periods with greater presence within the school season. In addition, urban environmental pollution and pollen are exacerbating allergies and respiratory conditions among the youth population [19,20], especially in the case of outdoor atmospheric particulate matter (PM) with a diameter of less than 2.5 micrometers (PM 2.5) [12,21,22]. This context generates a situation of specificity where further study is necessary, given the different exposure scenarios with a greater influx from the outside although varying over time.
Given that ventilation is one of the main variables which affects the degree of environmental comfort [23,24], the European ventilation standard EN 13779:2008 [25], through its Spanish transposition [26], establishes a minimum outdoor airflow to guarantee the adequate indoor air quality (IAQ) in non-residential buildings. Mainly, its focus is to control CO2 concentration, pollutants, and suspended particles [27] to avoid the development of symptomatology and respiratory health related to prolonged periods of exposure [28]. According to the national regulation, this ventilation must be mechanically controlled since 2007, also including an air filtering system, to ensure this IAQ, but given that the adaptation could entail a huge investment and a higher energy consumption, several public institutions in Spain are imposing natural ventilation as the only system for IAQ control, against standards.
In this way, previous studies in classrooms of southern Spain [16,17], Portugal [12], France [29], Italy [30], and other south European locations [31] have shown poor indoor conditions, both thermal and clean air, which can relate to the appearance of symptoms like dizziness, dry skin, headache, or tiredness. This environmental exposure has a significant effect on both the academic performance [32–34], the general health of the users and their psychological and social development [35], existing evidences of poor indoor air quality in schools with correlation with negative effects on the students' health, which potentially can lead to asthma or allergic diseases [36], which are two of the most prevalent diseases in children and young people [37], and can be mainly related to the high values found in classrooms for bacteria and PM, given their pro-inflammatory role [38].
In this way, previous studies in European schools analyzed the link between the IAQ conditions, obtained through measurements of CO2, PM, and volatile organic compounds (VOCs), with health questionnaires made by parents, spirometry, exhaled nitric oxide tests, and asthma tests with medical kits [29,38]. This approach required complex equipment and tests, and were not directly related to on-site symptomatology but to long-term symptom development, as it was gathered in housing studies [39]. Users' perception of environmentally related symptoms had a direct potential to draw an actual comfort situation, not only determined by room-physical conditions but to occupants' responses, as was shown in [40–43], also with the capacity to identify individual answers, such as those related to gender or emotional situation [44–46].
Thus, the analysis of the occupants' symptoms and their environmental perception was identified as a potentially affordable complementary tool to obtain a more accurate indoor comfort condition assessment with a high degree of widespread applicability, together with the widely accepted rational (RTC) [23,47,48] or adaptive (ATC) [49,50] thermal comfort indicators, especially those analyzed in the Mediterranean area including educational buildings [51–53] or in non-air conditioned buildings in warm climates [54].
#### *1.2. Objectives*
The first objective of this research was to present the physical and operational characterization of the indoor environment of a representative sample of multipurpose classrooms in a wide area of southern Spain, as well as the environmental perception votes, personal clothing, and symptoms expressed by the occupants (aged 12–17 years) exposed to this environment during the measurement campaigns.
The second objective of the study was to contrast environmental sensation votes, perception, and indoor-related symptoms described by students during lessons with physical and environmental
parameters and operational scenarios (focusing on windows' and doors' operation), in order to evaluate the impact and relationship between them.
| doab | 2025-04-07T03:56:58.102713 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 116
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.117 | **2. Methods and Materials**
The acquisition of both the physical measurement data and the occupants' sensation votes during a normal school day was developed through the following phases:
The data collection was performed both in winter and midseason in two sets per day: One in the early morning, at the beginning of the first lesson, and another previous to the midmorning break.
| doab | 2025-04-07T03:56:58.103068 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 117
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.118 | *2.1. Definition of the Study Sample*
The study sample was composed of 47 multipurpose classrooms (for ages 12–17) from 8 educational buildings, selected from the most representative climate zones of the region of Andalusia according to the Spanish energy performance zoning [55–57] (zones A4, B4, C3, and C4), which include temperate to cold zones in winter (types A, B, or C), as well as average to warm summers (3 or 4). These zones can also be classed in the Köppen climate scale [58] as cold semi-arid climate (Köppen BSk) and hot summer Mediterranean climate (Köppen CSa), as it can be seen in Table 1**.**
**Table 1.** Study samples by location and climate zone.
These multipurpose classrooms followed the design standards established by the regional educational agency (Andalusian Agency of Public Education) [59], with classrooms measuring approximately 50 m<sup>2</sup> and 3 meters high for accommodating up to 30 students with their teacher. This standard also defined the common access corridor with the adjoining classrooms and the distribution of the furniture, as well as the location of the windows to the left of the occupants and the two entrance doors in the partition to the corridor, as it can be seen in Figure 1.
The most common composition of the external vertical wall was a brick masonry cavity wall with some kind of thermal insulation with a simple hollow brick wall with plaster setting in the inner surface. The internal partitions were usually composed of a half-brick wall with plaster on either side.
The regional standards established hot water (HW) radiators as the main heating system of schools, with no provision for cooling systems [59]. In this way, all the classrooms under study were equipped with this heating system and, in addition, two of the schools had some add-on split-systems for cooling. Ventilation is traditionally performed in the Mediterranean area by user windows' operation and uncontrolled infiltrations. Despite the current Spanish standard on thermal installations in buildings (RITE) [26] that establishes the mechanical ventilation as the only option for new non-residential
buildings, these systems are not normally started up in order to save energy when the building is equipped with them.
**Figure 1.** Multipurpose classroom according to Andalusian design standards: (**a**) Plant with windows, doors, and furniture standard distribution. (**b**) A-A' vertical section.
| doab | 2025-04-07T03:56:58.103129 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 118
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.119 | *2.2. Characterization of the Airtightness of the Samples*
The assessment of the infiltration level of the classrooms under study was performed by a series of airtightness tests (doors and windows closed) in order to obtain their expected average infiltrations rates (Figure 2). These tests consisted of decreasing the room pressure by using a fan, which extracted air until the indoor-outdoor differential pressure was stabilized. It was achieved by balancing the extracted airflow with the entering airflow through the envelope cracks. Then, the depressurization was decreased in steps by lowering the fan speed, in order to obtain the regression curve of the pressure/extracted airflow relation, which showed the entering airflow when the indoor pressure was equal to the atmospheric one.
**Figure 2.** Protocols developed for the characterization of the airtightness of the classrooms.
These tests were performed by using enclosure pressurization-depressurization equipment or "blower door", as specified in the ISO standard 9972: 2015 [6], considering each classroom as a single zone to be analyzed. The specific model used was the Minneapolis Blower Door Model 4/230 V System, which was controlled by the TECTITE Express software.
The higher-pressure difference used to create this regression curve must be at least ± 50 Pa; in this study, it was reached until a ± 70 Pa differential pressure.
When the classroom had a single access point, the pressurization-depressurization test characterized the airflow that can pass through the envelope by sealing the corresponding door and the adjacent classrooms and common area. However, in most of the studied classrooms there were two access points, so it was necessary to perform three measurements in each classroom, changing the location of the blower door and sealing, or not, and the door in which the blower door was disposed. In this way, it was possible to determine the real airflow that entered the classroom during its normal operation. Adjacent classrooms and common area were sealed, too.
This protocol was designed for medium rooms with two access doors like the one under study, and required three different measurements (Figure 2):
where *V*<sup>50</sup> *<sup>P</sup>*<sup>1</sup> is the air leakage rate at 50 Pa in Protocol 1, *V*<sup>50</sup> *<sup>P</sup>*<sup>2</sup> is the air leakage rate at 50 Pa in Protocol 2, *V*<sup>50</sup> *<sup>P</sup>*<sup>3</sup> is the air leakage rate at 50 Pa in Protocol 3.
Infiltration values measured in each of these three ± 50 Pa depressurization test hypotheses, developed in each classroom, were obtained by the following expressions of the British Standard 5925 standard, obtained from a simplification of the "crack flow equation":
$$m\_{50,AT1} = \frac{V\_{50,DoverA} + V\_{50,env}}{V} \tag{1}$$
$$m\_{50,AT2} = \frac{V\_{50,DovB} + V\_{50,env}}{V} \tag{2}$$
$$m\_{50,AT3} = \frac{V\_{50,cw}}{V} \tag{3}$$
$$m\_{50,t} = \frac{V\_{50,D\text{var}A} + V\_{50,D\text{var}B} + V\_{50,c\text{inv}}}{V} \tag{4}$$
$$n\text{50}\_{\text{f}} = n\text{50}\_{\text{f}}\text{p1}\_{\text{i}} + n\text{50}\_{\text{i}}\text{p2}\_{\text{i}} - n\text{50}\_{\text{i}}\text{p3}\_{\text{i}}\tag{5}$$
where *n*50,*AT1* is the infiltration rate at 50 Pa in protocol 1, in h<sup>−</sup>1; *n*50,*AT2* is the infiltration rate at 50 Pa in protocol 2, in h<sup>−</sup>1; *n*50,*AT3* is the infiltration rate at 50 Pa in protocol 3, in h<sup>−</sup>1; *n*50,*<sup>t</sup>* is the infiltration rate at 50 Pa through the envelope and doors of the room, in h<sup>−</sup>1. *V*50,*DoorA* is the air leakage rate at 50 Pa which circulates through door A, in m3/h; *V*50,*DoorB* is the air leakage rate at 50 Pa which circulates through door B, in m3/h; *V*50,*env* is the air leakage rate at 50 Pa which circulates through the envelope, in m3/h; *V* is the internal volume of the room, in m3.
| doab | 2025-04-07T03:56:58.103329 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 119
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.120 | *2.3. Field Measurements*
The measurement campaign was developed in the selected classrooms according a data collection protocol [16,17,60] in which physical parameters relating to hygrothermal comfort, CO2 concentration, and illuminance were obtained in a spatial matrix [61] previously and throughout the survey distribution period (30 min, twice per day), taking the average of the values obtained, both outdoor and indoor, as can be seen in Figure 3 and Table 2:
**Figure 3.** The 3-D array of measurement points superimposed in a multipurpose classroom. Red dots show values at 0.6 m, dark red dots at 1.7 m, orange dots represent measures in the room's envelope, and blue dot is the window.
**Table 2.** Acquisition points per physical parameters measured.
During the measurements both the initial state and the operation of doors, windows, solar devices, heating systems, and electric lighting, as well as changes in the occupants' distribution, were collected. Performance and uncertainty of measurement instrumental are described in Table 3:
<sup>1</sup> The instruments for hygrothermal measurements listed comply with the requirements of ISO 7726 standard [62] for class C (comfort).
Mean radiant temperature (*tr*, in ◦C) was calculated using Equation (6) [63]:
$$\overline{t\_r} = \left[ \left( t\_\% + 273 \right)^4 + \frac{1.10 \cdot 10^8 \cdot \upsilon\_a^{0.6}}{\varepsilon \cdot D \lg^{0.4}} \cdot \left( t\_\% - T\_a \right) \right]^{\frac{1}{4}} - 273,\tag{6}$$
where *va* is air velocity (in m/s), Ta is dry bulb air temperature (◦C), *tg* is black globe temperature (◦C), ε is emissivity (dimensionless, 0.95 for black globe), and Dg is globe diameter (m).
The operative temperature (to, in ◦C) was obtained through Equations (7) and (8) [64]:
$$\mathbf{T\_o = A \times T\_a + (1 - A) \times \overline{t\_r}}\tag{7}$$
$$\mathbf{A} = \begin{cases} 0.5 & \text{if } \mathbf{V\_a} < 0.2 \text{ m/s} \\ 0.6 & \text{if } 0.2 \text{ m/s} \le \mathbf{v\_a} < 0.6 \text{ m/s} \\ 0.7 & \text{if } 0.6 \text{ m/s} \le \mathbf{v\_a} < 1.0 \text{ m/s} \end{cases} \tag{8}$$
where va is air velocity (in m/s), Ta is dry bulb air temperature (◦C), and tg is black globe temperature (◦C).
| doab | 2025-04-07T03:56:58.103574 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 120
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.121 | *2.4. Design and Distribution of Surveys*
The survey was designed to collect information and votes from occupants in order to comprehensively assess the environment in conjunction with the measurement of the physical parameters. The survey design was based on the experience of previous research [16,17,60,61,65] with the aim to collect data with an objective approach. The completion of the survey took around 20 minutes per classroom, being distributed during the measurement campaigns, and were performed twice per day both in a winter and a midseason day.
The survey distributed included questions about the following issues (the specific layout of the questionnaire is presented in Appendix A):
- Sensation: Thermal sensation vote (TSV) using the 7-points ASHRAE scale.
- Preference: Thermal preference vote (TPV) using the 7-points ASHRAE scale.
- Acceptance: Thermal environment rejection percentage (PDacc) from 0 (rejection) to 1 (acceptance).
- Level of comfort: Thermal comfort vote (TCV) from 4 (extremely uncomfortable) to 0 (comfortable).
- -Difficulty concentrating (DC).
- -Dry throat (DT).
- -Dizziness (D).
- -Itchiness (I).
- -Dry skin (DS).
- -Nausea (N).
- -Nasal congestion (NC).
- -Eye irritation (EI).
- -Headache (H).
- -Chest oppression (CO).
- -Tiredness (T).
The perception of the hygrothermal environment was formulated to the occupants according the protocol established in the Spanish version of Standard ISO 10551 [66].
The clothing insulation values worn by the occupants were obtained from the surveys and subsequently quantified according to EN ISO 9920 [67] and EN ISO 7730 [23], considering the corrections proposed by Havenith et al. [68] for seated occupants, with a thermal insulation of clothing (Icl) lower than 1.84 clo and air velocities under 0.15 m/s.
In addition, the protocol for the analysis of the surveys included a screening for the exclusion of the sample when the participant had previous and subsequent symptoms, related health problems, were developing some sickness, were taking medication for a long time, or when a strange answer was found for the multiple choices of a given question. In this way, during the measurements, a total number of 977 valid surveys was obtained (Table 4).
**Table 4.** Students participating in the survey campaign according to season and sex.
| doab | 2025-04-07T03:56:58.103809 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 121
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.122 | **3. Results**
The results of the present study, part of a PhD dissertation [69], can be grouped into five subsections:
These values were analyzed according to seasons (winter, W, and midseasons, MS) and windows' and doors' operation (open windows, OW, closed windows, CW, open door, OD). In this way, 26 classrooms (55% of the case studies) had the windows closed during the measurement period, with 23 of these during the winter session, and 21 had the windows open (45% of the case studies), with 11 of them during the winter period. No intervention by the researchers was made to modify classroom-state, allowing us to gather operational actual conditions.
#### *3.1. Mean Values of Physical Parameters*
The measured interior air temperature (Ta) ranged between 17.8 and 22.7 ◦C during winter season (Table 5), with the lowest mean temperature values obtained for the case studies with closed windows, especially when inner doors were open, with values of 20 ◦C. It can be related with the outdoor conditions, given that the lowest outdoor temperature values (Ta, outdoor) were measured for classrooms with closed windows and open inner doors. In addition, 8 of the case studies had the windows open during winter, which can mean that there was a bad regulation of the heating system and the heat excess had to be dissipated, or that the students considered that they had to ventilate the classroom due to a poor environment perception. Indoor air temperature in midseasons was oscillating around 22.4 ◦C, without a direct relation with window operation. Although winter time temperature expectations range between near 20–22 ◦C to 20.4–22.6 ◦C if windows are open (central quartile lower and upper values), this band nearly doubles in middle season, when temperatures from 20.6 to 24 ◦C may be expected (21.1 to 24.5 ◦C if windows are open). A quartile distribution plot for indoor thermal parameters, air temperature, and operative temperature is proposed in Figure 4. It is noteworthy to highlight that there was a statistical significance between seasons in a windows-state with independent behavior aspect that was verified through test of comparison of samples, F-test for the variance and a K-S (Kolmogórov-Smirnov) for the distributions of probability with *p*-values under 0.05 in all the cases.
**Table 5.** Mean values of environmental parameters obtained during the field measurements related to seasons and windows' and doors' operation.
W are measurements during winter, MS are measurements during midseasons, OW-W are measurements during winter with open windows, OW-MS are measurements during midseasons with open windows, CW-W are measurements during winter with closed windows, CW-MS are measurements during midseasons with closed windows, CW OD-W are measurements during winter with closed windows and open doors, SD are standard deviation.
Although average values of mean radiant temperature (t*r*) were within the recommended operating temperature ranges for classrooms according to ISO 7730 standard [23] (22.0 ± 2.0 ◦C for category B), there was a high dispersion of figures with a standard deviation (SD) between 1.7 ◦C in winter with closed windows and 3.7 ◦C in midseasons with open windows, which was due to the operation of HW radiator system, especially when windows were open, with t*<sup>r</sup>* values of 27.2–28.0 ◦C with radiators on and values of 17.5–19.0 ◦C when radiators were turned off. This caused operative temperature to swing usually between 20 and 25 (central quartiles) during middle season, with typical values of 20.6 to 22.5 ◦C during winter, with a very similar band of 20.4 to 22.6 ◦C if windows were open, highlighting the effect of surface thermal control performed by the radiator heating system.
Relative humidity (RH) in winter was always over 40%, with a maximum value of 64% in the case of one of the classrooms with windows closed and inner doors opened. In midseasons, relative humidity was lower but with a higher oscillation, with a minimum value of 29%.
Air velocity (Va) values were oscillating under 0.05 m/s, both in winter and midseasons, only exceeding the recommended design limit for comfort category B established by the ISO 7730 standard [23] of 0.16 m/s in one of the case studies with open windows, with a value of 0.18 m/s. In the case of closed windows, air velocity was always under 0.09 m/s. This showed poor air movement and limited air displacement potential.
Measurements of the CO2 concentration usually show figures well above typical thresholds (Figure 5). The World Health Organization (WHO) recommends a limit for healthy indoor spaces of 1000 ppm [70]. In this way, the probability distribution derived from the measures showed that more than 92% of the distribution for closed windows was above this limit, while this only decreased to 88% of the time when windows were open. In addition, 47.5% of classrooms with windows closed exceeded the 2000 ppm threshold. The greatest relative effect of window operation was seen in the winter, when CO2 concentration can be decreased by 25%, comparing median values. However, figures were above desirable levels, indicating the lack of capacity of the window operation to solve a suitable ventilation. In general, during the intermediate season, the operation of the windows did not provide a significant improvement of indoor air quality, which may be related to the lack of thermal differential between indoor and outdoor air, limiting the air exchange due to the absence of a thermodynamic effect (Figure 6).
**Figure 4.** Quartile distribution for indoor air temperature and operative temperature. (**a**) Quartile distribution for indoor air temperature; (To) winter (top) and mid-season (down) with windows closed (0) and open (1). (**b**) Quartile distribution for operative temperature (To); winter (top) and mid-season (down) with windows closed (0) and open (1).
**Figure 5.** CO2 concentration distribution. (**a**) General CO2 concentration density trace for winter (blue) and mid-season (red). (**b**) Detailed CO2 concentration density trace for winter (blue) and mid-season (red) with closed windows (continuous line) and windows open (dot line).
The median room mean illuminance (E) in the case studies oscillated between 461 and 560 lx (both cases with a SD of 222) according to the season, with an average lighting uniformity (Uo) of 0.48. However, although there seemed to be a greater illumination associated with the half-season period, it was not possible to rule out, without further measurements, the fact of being biased by the activities in execution during the measurements. The high SD was due both to the use of the projector and the
solar protection devices (as low as 15 lx) and the lack of use of a solar protection device with direct solar radiation (figures as high as 1710 lx) (Figure 7). The homogeneity of the lighting solutions in almost all buildings generated visual fields with very similar characteristics, mainly dominated by the behavior of their electric lighting. The correlated color temperature was similar in all cases, varying from 3500 to 5500 K; hence, it can be considered that both the amount of light and hue did not affect the thermal perception of the participants, as exposed by Bellia et al. [71] and Acosta et al. [72].
**Figure 6.** Indoor and outdoor air temperature values (Ta), mean radiant temperature values (t*r*), relative humidity (HR) values and indoor and outdoor CO2 concentration values related to seasons and windows' and doors' operation.
**Figure 7.** Room illuminance. (**a**) Hourly distribution for mean-room illuminance by season: winter (blue) and middle-season (red). (**b**) Quartile distribution of mean-room-illuminance by season (top) winter (down) mid-season.
#### *3.2. Mean Values of Airtightness of the Samples*
The values of airtightness of the classrooms under study with a difference of pressure between indoor and outdoor of 50 Pa (n50 range) varied from 2.6 h–1 to 10 h–1, with an average value of n50 of 6.97 h–1 and a SD of 2.06 h–1.
| doab | 2025-04-07T03:56:58.103974 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 122
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.123 | *3.3. Mean Values of Occupants' Votes*
The mean thermal sensation vote (TSV) of the students in both seasons was "slightly warm", with a value of +0.32 on the ASHRAE scale in winter and +0.38 in midseasons, having a SD of between 0.93 and 0.83, respectively. This can be identified as a common situation among poorly ventilated and crowded spaces. Even with open windows the actual air-removal capacity looked very limited as previously evaluated (Table 6). These thermal perceptions were higher (+0.10 points) when windows
were open, highlighting an excess of heat release of the heating system due to inefficient regulation and the wish of the users of dissipation. In this case, the occupants' thermal preference vote (TPV) expressed was softer and closer to neutrality than the TSV, not fitting at all with the perceived thermal sensations (R<sup>2</sup> <sup>=</sup> <sup>−</sup>0.47, moderate correlation), as showed by Teli et al. [14,53].
**Table 6.** Mean values of occupants' votes obtained during the field measurements related to seasons and windows' and doors' operation.
TSV is the thermal sensation vote, TPV is the thermal preference vote, PDacc is the thermal environment rejection percentage, TCV is the thermal comfort vote, and EPV is the environmental perception vote.
The average thermal environment rejection percentage (PDacc) expressed by students, based on a scale from 1 (acceptance) to 0 (rejection), was low and homogeneous in both seasons, with a mean value of 0.81 in winter conditions and 0.85 for midseasons. In addition, thermal acceptance was, in general, slightly better in classrooms with closed windows in both seasons, but in the case of closed windows and open doors. The thermal comfort vote (TCV) allowed us to qualify this acceptance-rejection PDacc index, given that less than 70% of students found "comfortable" the thermal environment in winter conditions compared to more than 80% in midseasons. This percentage increased to 96% for students with "comfortable" or "a bit uncomfortable" votes in winter conditions, but without reaching 92% in midseasons. By contrast, the number of users who, accepting a slight discomfort, considered the acceptable environment was superior in winter than in midseason, where the feeling of discomfort was slightly more marked, can be seen in Figure 8.
**Figure 8.** Accumulated frequency for the environmental perception and thermal comfort votes, related to seasons and windows' and doors' operation.
The mean environmental perception vote (EPV) showed during winter a 1.03 value (slightly bad odor), with low differences regarding windows' operation (less than 0.03 points); in midseasons, EPV was more favorable (0.61), with more than 0.10 points of difference regarding windows' operation. Figure 8 also shows the accumulated distribution of the EPV, in which less than 35% of students voted "without odor" in winter, while more than 50% voted it during midseasons. In addition, almost 30% of students perceived a slight odor or worse in winter in comparison to the midseasons, with 10%. Finally, around 7% of students voted "bad odor" or worse in winter, while there were no votes in this way during midseasons.
During the winter there is a more evident feeling of a poorly ventilated (not healthy) environment, in line with the measured CO2 values acting as a token of the indoor ambient renovation state. The operation of windows produced little to no effect on the improvement of the environmental quality, especially during the winter. Although it was found that the opening of windows in this period generated noticeable dilution of the interior atmosphere, it was still insufficient to guarantee pleasant environments.
During midseason, although the ventilation mechanism was less effective (by means of a lack of thermal differential), the capability of diluting the indoor environment to threshold levels was perceived by the users as somewhat better. The assessment of these user perception-thresholds was a key aspect of research, since it will allow the design of more adequate and well-accepted spaces.
| doab | 2025-04-07T03:56:58.104451 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 123
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.124 | *3.4. Mean Values of Occupants' Clothing Insulation*
The occupants' clothing insulation (Icl) showed two models of response linked to the season, as it can be seen in Table 7. Clothing distribution in winter was homogeneous, with a mean value of 0.90 clo and a SD of 0.19, common both for open and closed windows, and a minor divergence of 5% around 0.6–0.7 clo values related to windows' operation, showed in Figure 9. It should also be noted that the biggest slope of the insulation distribution was during winter with closed windows and inner doors open, which highlighted the smaller variation in clothing insulation of this group of case studies.
**Figure 9.** Accumulated frequency for the designed insulated level related to seasons and windows' and doors' operation.
**Table 7.** Mean values of occupants' clothing insulation obtained during the field measurements related to seasons and windows' and doors' operation.
In midseasons, the clothing insulation was lower and variable, with a SD of 0.23 and an asymmetrical distribution. There was a divergence of up to 25% in the frequency of the lowest levels of clothing insulation during midseasons regarding the windows' operation, coinciding both frequencies around the value of 0.90 clo (75–80% of the accumulated frequency).
#### *3.5. Symptoms and Related Health E*ff*ects*
The most commonly reported severe symptoms were headache and concentration difficulty (around 10%), followed by tiredness and a dry throat (under 10%), with a greater prevalence during wintertime and closed windows' operation. The action of the windows (Figure 10) was relatively weak, indicating the limited actual ventilation capacity of these spaces with only the opening of windows (reductions were around 25% less, in general). However, the perception of mild symptoms was very common in the classrooms, with tiredness, headache, and difficulty in concentration presenting a prevalence in the range of 40% to 50% for closed operation and slightly lower when windows were open (decreasing around 10–15 %), as shown Table 8.
**Figure 10.** Average probability of reported symptoms by seasons (W is winter, MS is midseasons) and windows' operation (0 is closed, 1 is open) (severe, left group; light perception, right group for each symptom).
This situation changed in midseason, where the symptom report was lower, even for the situation of closed windows. However, unlike winter, symptomatic perception increased when the windows were open for both perceptions, severe and mild, especially for dry throat, itchiness, nasal congestion, and headache, which are symptoms that can be linked to the penetration of external species (in many cases aerobiological such as pollen [73–75]).
Aiming to evaluate the overall impact of the different perceptions of symptoms, while assuming the variability component of the subjective responses and different individual sensitivity to the environments, unlike the evaluation of physical parameters, users were asked to assess the intensity of the perception of discomfort on a scale of 0 to 1 (0 none and 1 maximum intensity). Although this was not a standardized parameter (it may vary between different users) it had a great potential to represent the importance that each user assigned to the nuisance and, therefore, to assess the actual perception of the indoor conditions. Similar subjective ratings in conjunction with objective environmental measures were used in relevant studies, such as [76–79]. An overall indicator was collected through the addition of the specific scores or valuations generated by the users of each symptom or condition.
Icl is the clothing insulation level of the occupants.
This represented a global assessment of perceived impact, with a fundamentally qualitative character, since there was no univocal relationship but strong enough to highlight health discomfort ant to categorize best and worse indoor environments. The main values from the different classrooms are grouped by seasons and windows' situation in Table 9. This table contains the statistical summary for the data samples. Of particular interest are standardized bias and standardized kurtosis since, in all the cases (except the kurtosis of MS\_1) these statistics were outside the range of −2 to +2 standard deviation, thus indicating significant deviations from normal.
**Table 8.** Relative probability of occupants' relating symptoms and health effects, from N (not perceived), to L (lightly perceived), and H (severe perception), with closed windows (0) and open (1).
DC is difficulty concentrating, DT is dry throat, D is dizziness, DS is dry skin, IT is itchiness, N is nausea, NC is nasal congestion, EI is eye irritation, H is headache, CO is chest oppression, and T is tiredness.
The distribution of symptoms' samples for each scenario (Figure 11) was asymmetrical, not normal (Shapiro–Wilk test with *p*-value less than 0.05 in all cases, so it can be ruled out with 95% confidence) with bias. Median values located between 1.4 as the lower impact case in half a season (closed windows) up to 2.10 for winter (also with closed windows). Although values concentrated around 2.00, there was a significant dispersion, reaching values of up to 11, which meant a maximum vote in practically all the symptoms. (This specific case must be understood as outlier). This highlighted that even in the best scenario analyzed, there was a significant perception of ambient-related symptoms and problems by the users. By contrast, there was also a non-negligible presence of users that did not reflect any discomfort or effects, especially in the midseason scenario with closed windows, with percentiles that
stood at 39%, compared to lower values in the other states, where this group went from 6.1% to 16.4% (W0 to MS1). In this way, the low level of difference in the distribution according to windows' operation can also show that the ventilation airflow through windows was not enough to guarantee a noticeable reduction of the students' symptoms, although it can modify slightly the physical parameters of the interior environment. This aspect was of singular importance, since it indicated that the mere control of the usual environmental values did not guarantee satisfaction with the interior environment, at least with regard to the absence of bothersome symptoms. In the case of midseason, symptoms described with open windows can be due to the higher level of external aerobiological particles entering into the classrooms, such as pollen. That is why the appropriate ventilation to provide a perceptive reduction of the symptomatology should be done by means of fans with filter system.
**Table 9.** Statistics from symptoms' scores for individuals' response by season and windows' situation (MS, middle season; W, winter; 0, windows closed; and 1, open).
**Figure 11.** Probabilistic density trace distribution for individual symptoms' scores (winter, red; middle season, blue; windows closed, solid line; windows open, dashed line).
The probabilistic distributions of individual related symptoms' scores for the different scenarios showed some similarity in the global pattern response and central values, mainly for open windows, except for the MS\_0 (closed windows). A set nonparametric contrast through K–S test (Kolmogorov–Smirnov for the global parameter) was developed to evaluate the pertinence to a common distribution. In all four cases, comparisons for accumulated distances of the samples showed statistically significant differences at 95% significance between the distributions (with all the cases with a *p*-value < 0.05 and DN values over Dcrit.0.05), with DN around 0.122 to 0.148 for the samples with closest distribution (windows open winter vs. middle season and winter open vs. closed windows) and the greater DN value 0.380 for the furthest. So it can be established that there were different distributions for all the cases
| doab | 2025-04-07T03:56:58.104721 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 124
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.125 | *3.6. Airtightness*
The average value of the infiltration rate at 50 Pa (n50) was 6.97 h<sup>−</sup>1, with a standard deviation of 2.06 h<sup>−</sup>1. Models with the lowest n50 values were those in the C3 climate zone, where the lowest average temperature values are recorded in the winter. The values of n50 ranged from 10 h−<sup>1</sup> (maximum) to 2.6 h−<sup>1</sup> (minimum), both recorded in the B4 climate zone (Table 10).
**Table 10.** Average values and standard deviation of n50.
### **4. Discussion**
This section is focused on the analysis of the relationships between the symptoms described by occupants and the rest of the parameters under study (physical, building operation, and votes).
### *4.1. Relationship between Physical Parameters and Classroom Operation*
It could be assumed that manual opening of windows in naturally ventilated buildings should depend on outdoor conditions, as this is the main element of control. However, it was observed that, despite the fact that in midseason windows remain open longer than in winter, no clear linear trend can be observed. In winter, the need for ventilation or indoor air changes is considered more important than the need to control the entry of outdoor cold air. Analysis by categories of the opening of windows (Figure 12) showed this occurs mostly in mid temperatures, although it was also observed in cooler conditions when necessary. Furthermore, no progressive growth was observed with the increased temperature, as could be expected. In midseason, it is more common to open windows, although there was no clear correlation with temperature, some of which was similar to winter, where more windows are opened in comparison.
**Figure 12.** Cross-tabulation for windows' opening and outdoor air temperature. (**a**) Cross-tabulation for windows opening and outdoor air temperature in winter: windows closed (gray) and open (blue). (**b**) Cross-tabulation for windows opening and outdoor air temperature in mid-season: windows closed (gray) and open (blue).
It could be deduced that users are psychologically or culturally conditioned to some extent as to how and when they open windows. Although it would be preferable for the classroom windows to remain open, the act of opening was seen as a reaction to poorer indoor air quality, which was more noticeable for the same thermal conditions in spring. It, therefore, appears that there is an adaptation process.
As it can be seen in Figure 13, although there was statistical significance between the CO2 concentration and the outdoor-indoor air temperature differential (*p*-value < 0.05), the correlation was somehow weak and more clear in winter time (R<sup>2</sup> = 0.249) than in midseason (R*<sup>2</sup>* = 0.145), with a better fit to a y-reciprocal relation. However, the predictive mathematical model lacked enough accuracy to be of utility to forecast actual situations. Besides the wide dispersion on values, there was a trend in the worsening of the indoor environment as DT increased, as can be usually expected, due to the lack of a controlled ventilation system.
**Figure 13.** Fitted regression model plot for CO2 indoor concentration related to indoor-outdoor air temperature differential, closed windows (red), and windows open (blue).
Is noteworthy to highlight that moderate DT winter and midseason trends were very similar, which matched with the foreseen windows' operation patterns, when most of the apertures occurred around cold-mild external temperatures.
| doab | 2025-04-07T03:56:58.105241 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 125
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.126 | *4.2. Relationship between Physical Parameters and Symptoms Described*
Some symptoms were more predominant when outdoor temperature was lower, although no clear linear relationship could be established. These symptoms were more frequent in winter when the thermal differential is at its highest, usually linked to a lack of ventilation at the time, as supported by the high CO2 indices as a general air quality indicator.
Although the symptoms often appeared to be more evident when the windows were open, this should be seen as a consequence, not a cause, as user perception of the symptoms was generally clearer when opening the windows. This is interesting to note, as it could be due to a situation which exceeded the perception threshold. In the winter, it is more common to observe symptoms such as difficulty concentrating, dry throat, and tiredness. These are very closely linked to poor hygrothermal control, even with windows open, where temperature and relative humidity are far more important, especially with open windows, as well as increased indoor CO2 linked to poor ventilation. In contrast, itchiness and chest tightness were barely noticeable.
The situation changed in midseason and symptoms, such as difficulty concentrating, tiredness, and nasal congestion, were less widely reported. However, symptoms less connected with the absence of hygrothermal regulation increased, while there was a greater presence of symptoms that may be linked to outdoor exposure.
When both lighting parameters, illuminance (E) and illuminance uniformity (Uo), were analyzed and referred to symptomatology, no clear correlation was obtained, as other previous studies showed for educational buildings [80]. This may be because illuminance values in the classrooms under study were generally over 350–400 lx with a uniformity of 0.40–0.50, so they were values good enough to not influence students at a symptomatic level.
The infiltration rate (n50) and the symptoms related by occupants showed a very tenuous connection, with some weak trends in the case of tiredness, as well as difficulty on concentrating, dry throat, and headache. Given that the airtightness of the classrooms was, in general, adequate or even good, with an average value of 6.97 h−<sup>1</sup> with a maximum value of 10 h−1, its influence can be moderate due to its low impact on air renewal. It also indicates that other variables, like time spent inside the classroom or windows' and doors' operation, can have more importance than the airtightness of the room.
There was no clear linear correlation between the students' clothing insulation and the symptoms described during measurements. The possibility of freely varying the level of clothing insulation by the students, according to their individual thermal needs, may be a factor that influenced this lack of relationship between clothing and symptomatology, besides psychological factors linked to clothing.
When symptomatology was assessed as global, there were some trends that could be identified. If CO2 was assumed as an overall indicator of indoor air renovation (not as a contaminant itself), the worsening of indoor environment linked with the increase of symptoms related. It can be approximated to a logarithmic regression relation (Figure 14), although a wide spread of values must be assumed. Different patterns for winter and midseason were described due to adaptation of users and the influence of outdoor species. Although this model presents some uncertainty for its use as a prediction tool, it does have the capacity to act as a qualitative indicator.
**Figure 14.** General average related symptoms' scores relation with indoor CO2 grouped by measured classrooms (green for middle season and blue for winter).
A linear trend model was calculated for the average symptom score and given a record of the average indoor CO2 (logarithmic fit). The model was statistically significant at *p* < 0.05, having a high correlation coefficient (R<sup>2</sup> = 0.8833) and a mean square error (MSE) of 0.6160.
A somewhat weaker linear relationship (logarithmic fit also) was seen (R<sup>2</sup> = 0.509 for midseason and R<sup>2</sup> = 0.143 for winter) but with statistical signification (*p*-value < 0.05 in both cases) and an error of MSE 0.425. Although dispersion was high, it was also a useful qualitative trend indicator, and was found between the overall perception of symptoms and the indoor operative temperature (Figure 15). In this case, it can be established that the symptoms tended to be more frequent when indoor temperatures increased, also with specific patterns for winter and middle season.
**Figure 15.** General average related symptoms' scores relation with indoor operative temperature grouped by measured classrooms (green for middle season and blue for winter).
### **5. Conclusions**
A wide study sample of 47 naturally ventilated multipurpose classrooms of the most representative climate zones of southern Spain was characterized and analyzed through field measurements and surveys distributions, in order to contrast environmental sensation votes, perception, and indoor-related symptoms described by 977 students during lessons with physical and environmental parameters, as well as operational scenarios.
The main operational case to be analyzed, according to votes and symptoms, was the windows' operation. In this sense, the 61% of the case studies during winter season had the windows open, which can be related both to a bad regulation of the heating system (the slight heat excess had to be dissipated) as well as to a poor indoor environment perception. In this way, the case studies with open windows in winter had a higher mean indoor air temperature value (21.5 ◦C versus 21.0 ◦C) and higher standard deviation of the mean radiant temperature (2.6 ◦C versus 1.6 ◦C). The mean thermal perception of students in winter season with open windows reinforced this slight heat excess, given that it was in a comfort range but 0.15 points warmer than in the case of closed windows, also expressing a thermal preference of thermal neutrality-mild cold (−0.06 on the ASHRAE scale) with open windows in contrast to the preference for a warmer environment when the windows were closed (+0.13). The thermal assessment of the environment through the thermal comfort vote (TCV) also had a poorer value with open windows (−0.44 versus −0.35 from 0 to −4), also showing a higher deviation in the votes (0.75 versus 0.54) and a somewhat higher linear correlation with CO2 concentration. Therefore, the architectural design should take into account to guarantee the air quality of the venue, as well as a comfortable heating system, in order to lead students to not open the windows uncontrollably, which produces, as explained above, a noticeable energy consumption and distorts interior comfort control.
The operation of windows during winter helps to decrease the mean value of CO2 concentration, with 1537 ppm versus 2164 ppm with windows closed; but, in most of cases, this decrease was
insufficient both to be within the standard recommendations for healthy environments and to reach threshold values of perceptions of the users. Given that the mean CO2 concentration level was still high even when windows were open, the mean environmental perception of the students (EPV) was not strongly influenced by the opening of windows, with almost 30% of students expressing a certain level of annoying odor in both cases, but also having a moderate correlation between poor environment perception and CO2 concentration just when windows were closed. Therefore, it can be stated that there was not a high correlation between the CO2 value and the students' perceptions, mainly due to the olfactory adaptation phenomenon, irrespective of the need to provide a suitable air quality for healthiness purpose. In this way, when symptoms reported were added to this analysis, they presented a not-direct relationship with EPV, with the higher complaint values when windows were open. This odor perception was also somehow related with tiredness, difficulty on concentrating, eye irritation, headache, and dry throat.
In midseasons, windows' operation led to a greater variation of indoor thermal values, both air and radiant, also maintaining in general CO2 levels over the WHO recommendations (mean vale of 1537 ppm). In addition, students' TSVs were higher with open windows, close to the thermal comfort limit by warmth. Furthermore, the odor perception (EPV) was also poorer (0.63 value versus 0.52) when windows were open in midseasons, reinforcing the finding that windows alone are not able to provide an adequate renewal capacity for the indoor environment.
The study of the symptoms reported during measurements showed that they were largely expressed by students, both for windows open and closed, particularly in the case of difficulty of concentrating (52%), headache and tiredness (46%), followed by dry throat and nasal congestion (39%), which also were the symptoms most frequently combined with the other symptoms. According to the studied scenario, without a mechanically controlled ventilation system, complaints were more often found during winter, especially when windows were closed. In midseason conditions, symptoms were somewhat less common, but students expressed more acute symptomatology when windows were open, especially for dry throat, itchiness, nasal congestion, and headache, which are symptoms that can be related to hypersensitivity to external agents such as allergies and other respiratory conditions. This conclusion states the clear need to provide a ventilation system with a suitable filtering.
Regarding the relationship with indoor temperature, it can also be established that the symptoms tended to be more frequent when indoor temperatures increased, also with specific patterns for winter and middle season, also related to the occupants' thermal perceptions.
Other operation factors, like illuminance and illuminance uniformity, as well as students' clothing insulation, were analyzed referred to this symptomatology, but no clear correlation was obtained. In the case of lighting parameters, almost all the classrooms under study were generally over 350–400 lx with a uniformity of 0.40–0.50, so they were values good enough to not influence students at a symptomatic level. The correlated color temperature was similar in all cases, varying from 3500 to 5500 K; hence, it can be considered that both the amount of light and hue did not affect the thermal perception of the participants. On the other hand, students had the possibility of freely varying the level of clothing insulation, according to their individual thermal needs, so its impact on symptomatology was diminished.
In conclusion, the findings of this study show that effectively controlled ventilation systems are needed to assure an actual indoor ambient renovation and clean air supply. The special sensibility to external species make it advisable to incorporate filtering and cleaning systems for outdoor air beyond the impact on investment costs and energy use that this may entail. In addition, the study of symptomatology suggests that CO2 indicator should be complemented by other pollutants' measurements to assure a proper interpretation of data, given that they could not be correctly identified exclusively using this single CO2 control parameter. As explained above, CO2 levels have a fuzzy influence in the students' symptomatology; hence, the air quality should be complementarily assessed through other parameters, such as particle or VOCs' levels.
The following points can be established as key aspects:
The use of CO2 as a standalone indicator of environmental quality, especially for the management of ventilation systems or driving the windows' opening, may be insufficient and can derivate in situations of increased user discomfort, alongside thermal-ambient disturbance. Although there was evidence that there is a relationship with the indoor CO2 levels growing (assumed as general index) and the increase in reported global symptoms, this was not a direct link and tended to be asymptotic from certain threshold levels (around 2000 ppm).
In most cases, natural ventilation systems are not able to solve properly the removal of pollutants, generating situations with high rates of complaints even when windows are open, although they can mitigate the situations during indoor peak situations (such as produced in winter season). In many cases, windows' opening can be counterproductive, given that, although the classic indicators of the indoor environment valuation improve, the perception of the users was negative or, at least, worse than in situations with closed windows.
Assuming that indoor ambient is a complex and multifactor model, in the current state of the art of school buildings, the use of natural ventilation by itself (with the typical configuration of classrooms and enclosures of the buildings in the region) does not guarantee adequate control of the indoor environment, against popular assumption in the area, both by users and administrators. This aspect, although it was previously included in the text, has been emphasized.
This fact may be related to the need to review the classic indicators and parameters commonly used in the environmental management of these spaces. This research found situations of discomfort even within the ranges generally assumed as comfortable by the standards and design guides. Thus, it is necessary to develop complementary indicators based on the perception and the probability of developing symptoms that allow contributing to the correct valorization of the indoor environments from the users' points of view.
In this way, this analysis should also be complemented with corresponding measurements and surveys distributions in classrooms with mechanical ventilation systems in order to develop a comparison of results with adequate CO2 levels, so further research on this field is required.
**Author Contributions:** Conceptualization, M.Á.C.-L., S.D.-A., J.F.-A., and I.A.; methodology, M.Á.C.-L., S.D.-A., J.F.-A., and I.A.; software, M.Á.C.-L., S.D.-A., J.F.-A., and I.A.; validation, M.Á.C.-L., S.D.-A., J.F.-A., and I.A.; formal analysis, M.Á.C.-L., S.D.-A., J.F.-A., and I.A.; investigation, M.Á.C.-L., S.D.-A., J.F.-A., and I.A.; resources, M.Á.C.-L., S.D.-A., J.F.-A., and I.A.; data curation, M.Á.C.-L., S.D.-A., J.F.-A., and I.A.; writing—original draft preparation, M.Á.C.-L., S.D.-A., J.F.-A., and I.A.; writing—review and editing, M.Á.C.-L., S.D.-A., J.F.-A., and I.A.; visualization, M.Á.C.-L., S.D.-A., J.F.-A., and I.A.; supervision, M.Á.C.-L., S.D.-A., J.F.-A., and I.A.; project administration, M.Á.C.-L., S.D.-A., J.F.-A., and I.A.; funding acquisition, M.Á.C.-L., S.D.-A., J.F.-A., and I.A. All authors have read and agreed to the published version of the manuscript.
**Funding:** This research was partially funded by the PIF Program of the Universidad de Sevilla (V Plan Propio).
**Acknowledgments:** The authors wish to express their gratitude to Jaime Costa-Luque for reviewing the manuscript and helping with several of the graphics, to Blas Lezo for the encouragement of this article, as well as to the students, teachers, and management team of the secondary schools which were part of this study. Finally, to the Public Entity "Agencia Pública Andaluza de Educación" from the Regional Government of Andalucía.
**Conflicts of Interest:** The authors declare no conflicts of interest.
| doab | 2025-04-07T03:56:58.105484 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 126
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.128 | **Table A1.** Excerpt from the surveys distributed to students.
| doab | 2025-04-07T03:56:58.106543 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 128
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.130 | *Article* **Indoor Air Quality Improvement by Simple Ventilated Practice and** *Sansevieria Trifasciata*
| doab | 2025-04-07T03:56:58.106573 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 130
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.131 | **Kanittha Pamonpol 1,\*, Thanita Areerob <sup>2</sup> and Kritana Prueksakorn 2,3,\***
Received: 21 January 2020; Accepted: 3 March 2020; Published: 9 March 2020
**Abstract:** Optimum thermal comfort and good indoor air quality (IAQ) is important for occupants. In tropical region offices, an air conditioner is indispensable due to extreme high temperatures. However, the poor ventilation causes health issues. Therefore, the purpose of this study was to propose an improving IAQ method with low energy consumption. Temperature, relative humidity, and CO2 and CO concentration were monitored in a poorly ventilated office for one year to observe seasonal variation. The results showed that the maximum CO2 concentration was above the recommended level for comfort. Simple ventilated practices and placing a number of *Sansevieria trifasciata* (*S. trifasciata*) plants were applied to improve the IAQ with the focus on decreasing CO2 concentration as well as achieving energy saving. Reductions of 19.9%, 22.5%, and 58.2% of the CO2 concentration were achieved by ventilation through the door during lunchtime, morning, and full working period, respectively. Placing *S. trifasciata* in the office could reduce the CO2 concentration by 10.47%–19.29%. A computer simulation was created to observe the efficiency of simple practices to find the optimum conditions. An electricity cost saving of 24.3% was projected for the most feasible option with a consequent reduction in global warming potential, which also resulted in improved IAQ.
**Keywords:** computational fluid dynamics; CO2 concentration; indoor air quality; *Sansevieria trifasciata*; ventilation
| doab | 2025-04-07T03:56:58.106602 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 131
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.132 | **1. Introduction**
In tropical regions, where the mean annual temperature (T) exceed 30 ◦C, [1,2], air conditioning (AC) is often used in buildings to make the conditions more comfortable for occupants and it can be found in commercial buildings, government buildings, factories, universities, schools, and homes. AC is often used in closed rooms with low ventilation in order to prevent ambient air pollution [3], caused by high occupancy [4] as well as to maintain low T and save energy [5]. The desire to save energy and a lack of awareness regarding health and safety issues relating to indoor air quality (IAQ) have resulted in rooms being designed in tropical countries, including Thailand, which allow for all doors and windows to be closed, causing poor ventilation, particularly in hotels [6]. Gases that are generated in closed rooms, which cannot be effectively ventilated can increase to harmful levels, resulting in negative health effects [7], especially for office workers, who spend most of their working hours inside buildings [8].
Keeping windows open or using a ventilation fan can improve IAQ, but energy consumption is thereby increased, since AC systems have to work harder to maintain the indoor T within a comfortable range. Therefore, an important aspect of room design is how to sustain good IAQ and thermal comfort with low energy consumption. An additional benefit from low energy use is the reduction of greenhouse gas (GHG) emissions due to human activities, which are a cause of climate change. It is now generally accepted that GHG emissions have a considerable impact by raising ambient T [9], which results in higher energy use for air conditioning to make rooms comfortable for their occupants, thus creating a vicious circle.
However, poor indoor air quality is not only associated with closed rooms, but it can also result from errors in the design of ventilation systems, which can introduce pollutants from outside into indoor areas [10]. Major indoor air pollutants that have been studied include PM2.5, PM10, O3, CO2, CO, NO2, NH3, volatile organic compounds (VOCs), and aldehydes, which can be derived from a number of potential sources [11–15]. Moreover, in addition to physical pollutants, biological pollutants, such as bacteria, fungi, and mold, can be suspended in indoor air in the form of particles and they are considered to be indoor air pollutants [16,17], and people in buildings affected by indoor air pollutants are at risk of acquiring sick building syndrome (SBS). The symptoms of SBS are various and non-specific, but they include tiredness, feeling unwell, itching skin, high blood pressure, and heart rate, and even difficulty in concentrating. Sometimes, these effects are rapidly relieved after leaving the building [11,15,18], but this may not be an option for those affected.
In this study, the air quality and comfort parameters that were selected for study were relative humidity (RH), T and CO2 concentration, and how they affect the proper design of rooms [19]. Also included is the CO concentration as a representative outdoor air pollutant generated by incomplete combustion of fossil fuels [9], being mostly derived from automotive exhaust fumes from roads and parking areas around buildings.
A high RH content in the ambient air can result in the low evaporation of perspiration from the surface of human skin with a consequent reduction in the excretion of substances by evaporation [20]. Further, exposure to high T has been found to not only affect work performance, but also to result in symptoms, such as mental fatigue and changes in blood pressure [21]. In addition, inhaling excessive amounts of CO2 above 10,000 ppm can cause a condition that is known as acidosis (low pH of blood: <7.35), in which the body's defense mechanisms are stimulated, resulting in, e.g., an increase in breathing rate and volume and high blood pressure and heart rate [12]. Exposure to CO2 levels of approximately 50,000 ppm can lead to the failure of the central nervous system (brain and spinal cord), possibly causing death [22]. On the other hand, breathing high levels of CO can lead to death due to tissue hypoxia, as CO can bind to hemoglobin more effectively than oxygen [23].
The results of research that was conducted by NASA's environmental scientists into the improvement of IAQ while using plants was published in 1989, with a number of houseplants being tested as a means of treating indoor air pollution by removing trace organic pollutants from the air in closed environments in energy-efficient buildings. The organic chemicals tested consisted of benzene, trichloroethylene, and formaldehyde, and the scientific names of the plants investigated were *Chamaedorea seifritzii, Aglaonema modestum, Hedera helix, Ficus benjamina, Gerbera jemesonii, Deacaena deremensis, Deacaena marginata, Dracaena massangeana, Sansevieria laurentii, Spathiphyllum, Chrysanthemum morifollum,* and *Dracaena deremensis* [24]. Among the plants that are generally found in tropical regions is *Sansevieria laurentii* (mother-in-law's tongue), which is a size that is suitable for a small office and it was selected in this study to test its effect on IAQ improvement.
IAQ has been an important issue in Europe and America since the 18th century [4]. However, there has been relatively limited research on the topic in Association of South East Asean Naitons (ASEAN) countries [15,25] and few long-term studies have been conducted [26]. The main aim of this research was to improve the IAQ of an air-conditioned office in Thailand with a poor ventilation system by practicing simple ventilation operations and locating the mother-in-law's tongue plants in the office, with the second aim of achieving lower energy consumption. Another beneficial outcome of
this study was finding ways of reducing GHG emissions that are associated with the use of electricity. In this paper, alternative energy-saving scenarios are presented to demonstrate the effectiveness of simple operations in reducing GHG emissions and improving IAQ for office workers.
| doab | 2025-04-07T03:56:58.106749 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 132
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.134 | *2.1. Studied Site*
The study site was an air-conditioned office in Valaya Alongkorn Rajabhat University under the Royal Patronage (VRU), which is located in Pathum Thani province, a suburban area 15 km north of Bangkok, Thailand at 14◦8.004 north latitude and 100◦36.961 east longitude, with the site, on average, 5 m above sea level. The office was located on the second floor of a four-story building that was surrounded by a parking area. The room dimensions were: Length × Width × Height of around 4 m × 10 m × 3 m for six occupants. The office was air conditioned with the only means of ventilation being a door, leading to open corridor, which was generally closed to maintain a low T and save energy. The air conditioner was a wall mounted type with cooling capacity 36,000 Btu/hour (Daisenko International Co., LTD., Thailand). Figure 1 presents a diagram of the office.
**Figure 1.** Diagram of the office.
| doab | 2025-04-07T03:56:58.107201 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 134
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.135 | *2.2. Measurement of IAQ Parameters*
Data were collected in the air-conditioned office for approximately one year from May 2017 to May 2018 to indicate the quality of the indoor air by the recommended levels of the American Society of Heating, Refrigerating, and Air-conditioning Engineers (ASHRAE) [27]. RH, T, CO, and CO2 were monitored every minute using a FLUKE 975 AirMeter, a portable device for the measurement of IAQ. The specifications of the measurement device are, as follows; CO2: accuracy ± 2.75%, range 0 to 5000 ppm; CO: accuracy ± 5% or ± 3 ppm at 20 ◦C and 50% RH, range 0 to 500 ppm; T: accuracy ± 0.5 ◦C from 5 ◦C to 40 ◦C, range −20 ◦C to 50 ◦C; and, RH: accuracy ± 2%, range 10% to 90% RH [28]. The device was installed on the desk in the middle of the room, at the same height as the breathing zone during working hours.
| doab | 2025-04-07T03:56:58.107288 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 135
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.136 | *2.3. Simple Ventilation Practices for Improving IAQ*
After obtaining the results of the one-year observation, four different systems were implemented in June 2018 during working hours (9:00–17:00) to discover the simplest and most efficient means of removing stale air from inside the room, and introducing fresh air from outside. The experimental condition was conducted in real practice where people in the office were working and doing activities as usual. The four systems tested were as follows: Case 1: the AC was turned on all day (normal case) from 1–7 June, Case 2: the AC was turned off during the lunch hour (12:00–13:00) from 8–14 June, Case 3: the AC was turned off for half a day (9:00–13:00) from 15–21 June, and Case 4: the AC was turned off all day (9:00–17:00) from 22–28 June. The statistical analysis was analyzed by one way ANOVA to consider among four cases at confidential level 95% (p < 0.05) by Statistic 8 Software (Version 8, USA).
For Case 3, it was decided to turn off the AC in the morning, because the high T in the afternoon [2] had a negative effect on work efficiency. Turning off the AC all day (Case 4) could not realistically be applied, since it would probably result in problems, such as heat strain. This system was included to establish the maximum rate of full-day ventilation with the AC turned off, the door constantly opened, and an electric fan mounted on the ceiling turned on.
An assessment of the envelope air permeability of the room was obtained by the infiltration rate, which was calculated by the following equation:
$$Q = -\frac{V}{t} \times \ln\left[\frac{\mathbf{C}\_t - \mathbf{C}\_{ext}}{\mathbf{C}\_0 - \mathbf{C}\_{ext}}\right]$$
where, *Q* is infiltration rate of air entering the room, *V* is volume of air in the office (m3), *t* is time interval (s), *Ct* is indoor concentration of CO2 at time *t* (ppm), *Cext* is concentration of CO2 in the ambient air (ppm), and *C0* is indoor CO2 concentration at time 0 (ppm) [29,30].
The volume of air in the office (V) was calculated from the size of the room (4 m x 10 m x 3 m), Interval (t) was 3600 seconds from hourly average data, *Cext* was average monitoring outdoor CO2 concentration at 430 ppm, *Ct* was the monitored CO2 concentration at time t (i.e., 18:00), and *C0* was the monitored CO2 concentration at one hour before *t* (i.e., 17:00). The frequency of measurement was every one minute, so the raw data were calculated to hourly data for both indoor and outdoor CO2 concentration. The indoor CO2 concentration was obtained by monitoring inside the office at 1 m height or nose level while people were sitting. The condition in the room was no plant and no ventilation for a long period. The door was opened when people came in and went out in a short time, not over one minute. The outdoor CO2 concentration was monitored at 1.5 m above ground level in the ambient air.
#### *2.4. Sansevieria Trifasciata for IAQ Improvement*
Another option tested for improving IAQ was locating the mother-in-law's tongue plants in the office to reduce the CO2 concentration in the ambient air through their photosynthesis. In this experiment, the *S. trifasciata* was put in a pot that contained soil. The plant was watered twice a week. The experiments were conducted by monitoring the air quality for six conditions, as follows: 0, 2, 3, 4, 5, and 6 Mother-in-law's tongue plants with three replicates for each case. The plants were placed on the floor near the desks where people worked, as shown in Figure 1. The number of plants was limited by the space of the rooms in which they were located. The IAQ was monitored from March to April 2019 for 24 hours each day to observe the amount of CO2 that the plants consumed for photosynthesis during the daytime and to establish the amount that they released through respiration at night. RH, T, CO, and CO2 were monitored by indoor air meter (FLUKE 975 AirMeter, USA) every minute. We monitored CO2 in a real situation to represent working activities or business as usual in tropical areas. Only the room temperature was controlled by air conditioner to comfort people at 25-degree Celsius, which was the general setting temperature in tropical countries.
| doab | 2025-04-07T03:56:58.107373 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 136
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.137 | *2.5. Numerical Study*
In addition to the measurement of the IAQ, a simulation was performed to estimate the efficiency of the office ventilation while using the computational fluid dynamics (CFD) software system, ANSYS Airpak 3.0.16 (Fluent Inc., Lebanon, NH, USA). Airpak simulation software has been broadly applied
for numerical simulation of the indoor air alteration under conditions [31–34]. Based on the finite volume method, Airpak uses the FLUENT CFD solver engine for the thermal and fluid-flow calculations to solve equations for the conservation of mass, energy, and momentum of air. The two-equation K-epsilon turbulence model was chosen to solve turbulent flow equations. The number of cells in the domain was approximately 1.5 million, while using hexa-unstructured geometry to discretize. For this function, all of the element types were used to fit the mesh to the geometry. The simulation was iterated to a convergence level of 10−<sup>3</sup> until the solutions were stable. Additionally, a mesh refinement study was conducted for quantifying and minimizing the error due to discretization. Four different mesh systems, i.e., coarser, course, medium, and fine were generated, to perform the test.
The investigated parameter was the mean age of the air, indicating the average time taken for the air to pass through the room, with a shorter time denoting higher air freshness [35]. A three-dimensional (3D) simulation of an experimental room with the same dimensions as that shown in Figure 1 was constructed using the ANSYS Airpak software and it is illustrated in Figure 2. Section 3.4 presents the results of numerical study.
**Figure 2.** Model of the room in Airpak.
| doab | 2025-04-07T03:56:58.107678 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 137
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.138 | *2.6. Estimation of Mitigation of Electricity Use and GHG Emissions*
Electricity usage directly impacts the increases in GHG emissions. Reducing electricity consumption not only helps decrease GHG emissions, but also reduces the cost of electricity. Table 1 presents the possible options for electricity saving scenarios.
**Table 1.** Options for electricity saving scenario.
Notes: Superscripts represent assumptions as follows: <sup>a</sup> the spare computer near the door was not in used; <sup>b</sup> all computers turned off during lunchtime; <sup>c</sup> not used continuously for printing; used for approximately one hour per day; <sup>d</sup> turned on only when in use; <sup>e</sup> no food kept in the office refrigerator overnight; <sup>f</sup> turned on from 09.00–12.00 and 13.00–17.00 (turned off during lunchtime); <sup>g</sup> turned on from 13.00-17.00; <sup>h</sup> turned on instead of AC; <sup>i</sup> turned off during lunchtime.
The different options that are shown in Table 1 were designed in collaboration with the room occupants, who were interested in the effect on the cost of electricity and global warming potential (GWP) if they all agreed to try them. The duration of working without AC for Cases A, B, and C were aligned with Cases 1, 2, and 3 in Section 2.3, respectively. Case 4 (no AC) was not re-assessed, since its results could be estimated from the other cases and its application was, in any event, not realistic. Case A was a typical case, whereas Cases B and C for other appliances represented situations that were not convenient, but feasible in practice.
### **3. Results and Discussion**
| doab | 2025-04-07T03:56:58.107935 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 138
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.139 | *3.1. Results of Monthly Monitoring Data*
The data from 24-hour monitoring of IAQ at one-minute intervals were converted into monthly results, as presented in Figure 3. The average RH of the office was 67.50% ± 3.98%, which was a little over the range of the recommended standard of 65% or less [36]. Thailand is located in a tropical zone and Thais are accustomed to a hot-humid climate, so the ambient humidity can be higher than the recommended level for the USA [37]. However, a T of 26 ◦C and an RH of 50–60% were preferred, according to the results of a survey of thermal comfort for air conditioned buildings in Thailand [25]. During the period studied, the Bangkok's mean annual RH was 74% [38], which was consistent with the wide range of outdoor RH between 34 and 78% RH, detected at 17 locations referred by the study of Ongwandee et al. [25]. Therefore, opening the office door might only be helpful in reducing the RH at some times.
**Figure 3.** Monthly relative humidity (RH) data.
Figure 3 shows that the greatest variation in RH was apparent for the minimum values during September and October 2017, because of the influence of outdoor air that is caused by various groups visiting the office at that time. Reducing the RH by introducing ambient air is only practicable while taking that the RH value can vary diurnally or hourly into consideration.
The room T was maintained close to the comfortable standard (23 to 26 ◦C) [39], consistent with the preferred environmental conditions for Thailand established in the thermal comfort survey (26 ◦C at 50–60% RH) [25], through the use of the AC, as shown in Figure 4. The annual country average outside T was 27.61 ± 1.31 ◦C and high temperature was found to be higher than usual during August to November in 2017, because of rain, so the winter started late in December [40]. The outdoor T was usually higher, reaching more than 40 ◦C in the afternoon. The highest indoor T was detected during September and October 2017, which was consistent with the variation in the minimum RH data, and it was also possibly due to the number of visitors entering and leaving the office during that period.
**Figure 4.** Monthly Temperature data.
The CO concentrations were measured and the data were analyzed by converting from ppm at the local T, to the standard ppm at 25 ◦C. Figure 5 shows that the average CO concentration was 1.32 ± 0.29 ppm, while the maximum concentrations were 4.57 ppm, 4.91 ppm in September, and October 2017, respectively. The CO detected must have originated from the ambient air outside the room with the probability that this was associated with the parking outside the room with the probability that this was associated with the parking area surrounding the building, since there was no source of CO generation in the room and the minimum values were close to zero. Moreover, the findings are also consistent with the findings related to T and RH in September and October. Further, while fluctuations can be observed between the minimum, maximum, and average levels of CO, these three parameters were closest in March 2018, because there were no events scheduled in that month. Nevertheless, although the ambient air outside the office probably influenced the concentration of CO, the level was not a significant factor in the IAQ, because it was lower than the indoor air quality standard, (9 ppm for eight hours and 35 ppm for one hour) [37].
**Figure 5.** Monthly CO data.
Figure 6 shows that, from May 2017 to May 2018, the maximum, average (± standard deviation), and minimum indoor CO2 concentrations were 1456.79 ppm, 600.67 ± 42.80 ppm, and 387.32 ppm, respectively. The maximum level of CO2 was found to be above the comfortable level of 1000 ppm—as recommended by various standards [4,11,12] in every month, except December 2017 and April 2018 (maximum values, 987 ppm and 977 ppm, respectively). The maximum 24-hour CO2 concentration was found in January 2018, on a day when all staff members (six people) were in the office together
with an additional four people attending a long meeting. This emphasizes that human respiration was a significant source of indoor CO2 (generally two pounds of CO2 per day) [22]. There were no significant fluctuation in the level of CO2 detected during September and October 2017, which implied that the exchange of air between inside and outside the office only occurred in the area near the door, and the stale air inside was not effectively removed, due to the lack of a ventilation system to support the exchange process.
| doab | 2025-04-07T03:56:58.108064 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 139
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.140 | *3.2. Results of Simple Ventilated Practices for The Improvement of IAQ*
Based on the measurements for the entire year, CO2 was the parameter that most obviously exceeded the comfortable standard and there was no obvious method of solving this problem, which did not involve reconstruction and the installation of a ventilation system. Therefore, this was the chosen parameter for the experiment to test the performance of simple ventilation practices. The results for the IAQ improvement in June 2017 between Cases A, B, and C that are shown in Table 1 corresponding to Cases 2–4 in Section 2.3, respectively, and Case 1 (normal case) are presented in Table 2.
**Table 2.** Measurement results in the operational practices for the improvement of indoor air quality (IAQ).
Reductions of 19.9%, 22.5%, and 58.2% of the maximum CO2 concentrations were found by turning off the AC during lunchtime (12:00–13:00, Case 2); in the morning (9:00–13:00, Case 3); and. during the full working time (9:00–17:00, Case 4), respectively, as can be deduced from Table 2. In all cases, the concentrations of CO2 were reduced below the comfortable standard (1000 ppm) [4,11,12]. Cases 2 and 3 were able to reduce the maximum concentration of CO2 by almost the same amount. The concentration of CO2 was being reduced when the air conditioning was more time inactive because the air inside and outside the room could be exchanged due to the door leading to the opened corridor. More time opening the door resulted in more time for air to be exchanged. Higher CO2 in the room was released into the ambient air then plants could lower CO2 in the ambient air through photosynthesis process.
The measured values are in agreement with research that was conducted in 21 offices, in Taiwan in which the CO2 levels were measured while using a Q-TRAK indoor air quality tester (Model 7575, TSI Corporation, Bangkok, Thailand)) with an average level of 708.2 ± 190.5 ppm, a maximum level of 1193.6 ppm, and a minimum level 464.0 ppm [11]. The normal case results were consistent with the indoor CO2 level in Thai classrooms, which was measured by the similar method using Indoor Air Quality Meters (IAQ-CALCTM) Model 8760/876, a real time monitoring device that used a dual wavelength non-dispersive infrared sensor (NDIR) for CO2 and Electro chemical sensor for CO by frequency of five minutes. The average CO2 concentration in classroom 1 (carpet) was 711 ± 272 ppm and classroom 2 (wooden) was 1332 ± 609 ppm [41].
The statistical analysis results by One-Way ANOVA found that the result was significant at *p* = 0.002223 (*p* < 0.05) and the *f*-ratio value was 5.18726. This result means concentrations of CO2 were different among four cases.
The T and RH were also measured to check whether they aligned with the comfort standards. Figures 7 and 8 present the results of T and RH, respectively.
**Figure 8.** Diurnal relative humidity in the ambient air and in the room during experiments.
From Figure 7, case 2 would be the optimum choice if the occupants of the room preferred not to work under hot condition. The T in Case 3 exceeded the comfortable standard, but it did not reach 28 ◦C, the level with which 80% of Thai workers have been found to be comfortable [42]. Thus, the adoption of Case 3 would have to be based upon the agreement of all the occupants of the room. The values of T between Cases 3 and 4 are visibly different because the ambient T in the afternoon is higher than that in the morning. Certainly, the greatest saving of electricity can be achieved by not using AC at all during working hours, but it is not a realistic option with a working temperature of over 30 ◦C for 7–8 hours [35], which would be likely to affect work performance. From Figure 8, the RH was lower in the air conditioned room, so the skin would be dried when you stayed for a long time. If the door was opened in Case 2–4, the RH was increased in the room to be more comfortable. The monitored wind velocity in front of the room was 0.05 ± 0.04 m/s, which was calm wind that mostly blew from southwest and west direction. The door was in the west, so the mild wind entered the room to comfort people and exchanged the air between inside and outside the room to reduce CO2.
Energy saving from turning off air conditioner was calculated. The air conditioner size 36,000 BTU/Hour (international) was converted into 10.55056 kWh. For case 2, the air conditioner was turned off one hour for 238 working days (not including special holiday 24 days and weekend 104 days in Thailand 2020). The conversion factor of Thailand grid mixed electricity year 2016–2018 was 0.5986 kg CO2eq/kWh (LCIA method IPCC 2013 GWP 100a V1.03, Thai National LCI Database, TIIS-MTEC-NSTDA (with TGO, Electricity 2016–2018) updated in December 2019) [43]. Therefore, we can reduce CO2eq emission for 1503.10 CO2eq per year.
| doab | 2025-04-07T03:56:58.108380 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 140
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.141 | *3.3. Results of Using Sansevieria trifasciata for IAQ Improvement*
Table 3 presents the results of monitoring the parameters that are relevant to IAQ during the experimental placement of mother-in-law's tongue plants in the poorly ventilated room.
**Table 3.** Results of indoor air quality improvement using *Sansevieria trifasciata*.
Table 3 shows that the average concentration of CO2 was decreased by placing 2, 3, 4, 5, and 6 mother-in-law's tongue plants in the office as compared to there being no plants in the room, as can be visually observed in Figure 9. The concentration of CO2 in the room was not varied by the number of plants, but influenced by temperature of the room. This means if the door is opened, the temperature will be high and the CO2 concentration will be reduced. Closed chamber is required to control infiltration and ventilation to only consider the influence of plant on indoor CO2.
**Figure 9.** The trend of 24-hour CO2 concentration and number of *Sansevieria trifasciata*.
From Figure 9, it can be seen that there was an increasing trend in CO2 concentrations from 9:00 to 15:00, after which they declined to an ambient air concentration at 473.23 ± 8.66 ppm. The reductions in the percentage CO2 concentrations with 2, 3, 4, 5, and 6 of *S. trifasciata* when compared to no plants being placed in the office were 19.29%, 21.99%, 14.36%, 13.66%, and 10.47%, respectively. The overall average CO2 decreased by 15.95 ± 4.13%, which was slightly (±4%) lower than the reduction that was achieved in Case 2 by turning off the AC during lunchtime. The statistical analysis by One-Way ANOVA results was significant at *p* = 0.009132 (*p*<0.05) and *f*-ratio value is 3.53877. The results are different among treatments. However, the result of statistical analysis to compare between Case 2 (turned off AC at noon) and 3–6 indoor plants found *p* = 0.061, so it is not significant at the 95% confidence level. Therefore, reducing indoor CO2 could be done by turning off air conditioner and opening the door during lunch hour, or by planting *S. trifasciata* in offices because there was no difference from statistical analysis results.
These results were consistent with those of other studies, which have found that plants could reduce indoor CO2 concentrations [44,45]. Therefore, during the daytime, human respiration is clearly the key factor in increasing the CO2 concentration, while ventilation is the main factor in decreasing the level of CO2 in a room. Hence, the number of people in the room and their activities are the main drivers of the CO2 concentration. Thus, the concentration of CO2 is not directly related to the number of plants that are placed in the room. It was found that the respiration of plants during the night had no effect on the CO2 concentration when compared with no plants in the room with a declining trend in CO2 to the same level among different options from 0–6 mother-in-law's tongue plants being apparent.
The envelope air permeability of the room was the air passing through 2–3 millimeters around the door, which was caused by damage of the sealed material. The infiltration rate of the room was considered from the CO2 concentration when there was no plant in the room to avoid the effect from photosynthesis or respiration of plant and microorganism in soil. The maximum rate was found at around 17:00–18:00 when people went back after finish working. The average maximum value of ventilation rate was 0.0152±0.0006 m3/s. The maximum infiltration was found at 0.0162 m3/s.
| doab | 2025-04-07T03:56:58.108705 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 141
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.142 | *3.4. Results of The Simulation*
As noted above, the stale air inside the room might not be properly ventilated and this section presents the results of the computer simulation generated to explain that hypothesis, as shown in Figure 10.
**Figure 10.** Assessment of air ventilation: mean age of air, (**a**) fan mounted on the ceiling was turned off (**b**) fan mounted on the ceiling was turned on.
Figure 10 shows the comparative results for the mean age of air when the fan mounted on the ceiling is turned off (Figure 10a) and turned on (Figure 10b). The electric fan was only used when the AC was switched off so there was no effect from the AC on the movement of air in this simulation. The average mean age of the air in Figure 10a,b were 747 s and 164 s, respectively. Thus, the efficiency of mixing the air by a 39-watt ceiling fan, from which the volume of air blown is 30 m3/min., is improved by approximately four times. Moreover, the exchange of air between the inside and outside of the room only occurred in the area near the door, and the stale air (i.e., that with the highest concentration of CO2 from human exhalation) was retained inside of the room, and this is clearly illustrated in Figure 10a.
The fan needs to be used for mixing air and the door must be opened, as shown in both Figures 10a and 10b, since otherwise there will be no movement of air near the door, as presented in Figure 10a. Thus, the ventilation is undoubtedly much better when compared to when the AC is turned on when the door is closed. This finding should encourage room occupants to sometimes apply general ventilation to increase the IAQ.
A mesh refinement study was performed based on four grids, i.e., coarser, coarse, medium, and fine grid, with approximately 0.4–0.5 million, 0.8–0.9 million, 1.4–1.5 million, and 1.6–1.8 million cells, respectively, to minimize and ensure that the error was below the tolerance level. The average mass flow rate in the room was used for comparison of the meshes. The differences between fine grid and the other grids (coarser, coarse, and medium grids) are about 7.4%, 3.6%, and 1.7%, respectively. The performance of the medium grid, selected for the simulation of this numerical study, was not significantly different to the fine one, and it was concluded as the suitable option.
| doab | 2025-04-07T03:56:58.109288 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 142
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.143 | *3.5. Mitigation of GHG Emissions*
Table 4 summarizes the reduction in electricity costs and hence GWP for the three experimental cases.
**Table 4.** Estimation of mitigations of greenhouse gas (GHG) emissions based on the electricity saving scenarios.
Notes: Cases A, B, and C are detailed in Table 1. Superscripts: <sup>a</sup> electricity cost per unit varies from month to month. The cost was averaged from the annual bill based on a figure of 3.96 Baht per kWh; <sup>b</sup> The GWP is calculated based on IPCC guidelines [46]. The emission factor for electricity production in Thailand is 0.5821 kgCO2-eq [47].
When the feasible measures for the electrical appliances in the office were applied representing Case 2/B with the AC turned off from 12.00–13.00, or as in Case 3/C with the AC turned off from 9.00–13.00, the electricity cost and GWP can be reduced by 7.5% and 24.3%, respectively. AC is generally accepted as the largest consumer of electrical energy per unit and, thus, contributes most to the emission of GHGs. However, the total amount of electricity that is used by computers is higher due to the number of computers in the office (6 units). Overall therefore, since computers consume the greatest amount of energy, turning them off for only one hour per day during lunchtime can help to reduce the overall electricity consumption by almost 10%. However, this option was least convenient as compared to other choices, based on the opinion of the room occupants. The different practices relating to computers and the AC are the reason why the reductions that result from Cases B and C are so great.
Therefore, the methods that were investigated in this study to reduce electricity consumption and GWP would be feasible means of mitigating costs by reducing the usage of appliances in offices and could be adopted and implemented in energy saving plans in universities and other workplaces. However, it should be noted that the figures that are presented in Table 4 are only the estimated values, and based on the power ratings (wattage) of the various devices. The actual electricity consumption also depends on the settings actually used (such as standby mode for computers and T settings of the AC and refrigerator). In general, by observation, there is unlikely to be a significant effect from the adjustment of the T setting of the AC based on seasonal variation (i.e., the T setting is not varied during the year based on the outside temperature). More accurate data could be obtained if an electricity meter was installed in each room to monitor the actual electricity consumption in different scenarios.
| doab | 2025-04-07T03:56:58.109476 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 143
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.144 | **4. Conclusions**
This study conducted measurements of four IAQ parameters i.e., T, RH, CO, and CO2, in a small room with six occupants and found that the levels of some of the parameters exceeded the recommended levels, particularly the level of CO2, the source of which was human respiration. Therefore, the number of occupants in the room and its ventilation efficiency are the key factors for the CO2 concentration. Poor conditions (i.e., a CO2 concentration of over 1000 ppm) were not detected
after the simple mitigation practices were implemented. However, T is a significant factor that must be taken into consideration in the adoption of measures to reduce electricity consumption. Without reconstruction of the office space and the installation of a ventilation system, Case 3/C was the best option with good IAQ, and it would achieve a reduction of electricity consumption of 24.3% based on the situation without taking any mitigating action. Moreover, this system would be feasible with the agreement of all members of staff working in the office. Another feasible option for CO2 reduction was placing mother-in-law's tongue plants in the office. The average reduction in the CO2 level based on using between two and six plants was almost 16%, which rendered the CO2 concentration within the standard comfortable level of 1000 ppm. However, human activities are the key factor in the CO2 concentration in a room, not the number of plants. The data that were derived from measuring the actual IAQ parameters in various scenarios and the results of the computer simulation are helpful in identifying and promoting simple practices that aimed at achieving good IAQ while reducing electricity costs and, hence, GWP in office situations. Further research should be directed towards measuring other IAQ parameters, and the causes of SBS, which are possibly associated with the CO, particulate matter, and VOCs in car exhaust fumes.
**Author Contributions:** Conceptualization, K.P. (Kanittha Pamonpol); methodology, K.P. (Kanittha Pamonpol) and K.P. (Kritana Prueksakorn); software model K.P. (Kritana Prueksakorn); validation K.P. (Kanittha Pamonpol) and K.P. (Kritana Prueksakorn); formal analysis, K.P. (Kanittha Pamonpol) and K.P. (Kritana Prueksakorn); investigation, K.P. (Kanittha Pamonpol); resources, K.P. (Kanittha Pamonpol); data curation, K.P. (Kanittha Pamonpol) and T.A.; writing-original draft preparation, K.P. (Kanittha Pamonpol) and K.P. (Kritana Prueksakorn); writing-review and editing, T.A.; visualization, K.P. (Kanittha Pamonpol) and K.P. (Kritana Prueksakorn); supervision, K.P. (Kanittha Pamonpol); funding acquisition, K.P. (Kanittha Pamonpol), K.P. (Kritana Prueksakorn), and T.A.; project administration, T.A. All authors have read and agreed to the published version of the manuscript.
**Funding:** This research was performed with the financial support of the Research and Development Institute, Valaya Alongkorn Rajabhat University under the Royal Patronage and the Andaman Environment and Natural Disaster Research Center, Prince of Songkla University, Phuket Campus.
**Acknowledgments:** The authors would like to express our appreciation to the Research and Development Institute, Valaya Alongkorn Rajabhat University under the Royal Patronage, Prince of Songkla University, Phuket Campus for the financial support. The support by assistants, i.e., Tip Sophea and Hong Anh Thi Nguyen are also acknowledged. The authors would like to thank Robert William Larsen for kind advice on English writing.
**Conflicts of Interest:** The authors declare no conflict of interest.
| doab | 2025-04-07T03:56:58.109684 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 144
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.147 | **Indoor Particle Concentrations, Size Distributions, and Exposures in Middle Eastern Microenvironments**
**Tareq Hussein 1,2,\*, Ali Alameer 1, Omar Jaghbeir 1, Kolthoum Albeitshaweesh 1, Mazen Malkawi 3, Brandon E. Boor 4,5, Antti Joonas Koivisto 2, Jakob Löndahl 6, Osama Alrifai <sup>7</sup> and Afnan Al-Hunaiti <sup>8</sup>**
Received: 11 November 2019; Accepted: 25 December 2019; Published: 28 December 2019
**Abstract:** There is limited research on indoor air quality in theMiddle East. In this study, concentrations and size distributions of indoor particles were measured in eight Jordanian dwellings during the winter and summer. Supplemental measurements of selected gaseous pollutants were also conducted. Indoor cooking, heating via the combustion of natural gas and kerosene, and tobacco/shisha smoking were associated with significant increases in the concentrations of ultrafine, fine, and coarse particles. Particle number (PN) and particle mass (PM) size distributions varied with the different indoor emission sources and among the eight dwellings. Natural gas cooking and natural gas or kerosene heaters were associated with PN concentrations on the order of 100,000 to 400,000 cm−<sup>3</sup> and PM2.5 concentrations often in the range of 10 to 150 μg/m3. Tobacco and shisha (waterpipe or hookah) smoking, the latter of which is common in Jordan, were found to be strong emitters of indoor ultrafine and fine particles in the dwellings. Non-combustion cooking activities emitted comparably less PN and PM2.5. Indoor cooking and combustion processes were also found to increase concentrations of carbon monoxide, nitrogen dioxide, and volatile organic compounds. In general, concentrations of indoor particles were lower during the summer compared to the winter. In the absence of indoor activities, indoor PN and PM2.5 concentrations were generally below 10,000 cm−<sup>3</sup> and 30 μg/m3, respectively. Collectively, the results suggest that Jordanian indoor environments can be heavily polluted when compared to the surrounding outdoor atmosphere primarily due to the ubiquity of indoor combustion associated with cooking, heating, and smoking.
**Keywords:** indoor air quality; aerosols; particle size distributions; ultrafine particles; particulate matter (PM); smoking; combustion
| doab | 2025-04-07T03:56:58.109909 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 147
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.148 | **1. Introduction**
Indoor air pollution has a significant impact on human respiratory and cardiovascular health because people spend the majority of their time in indoor environments, including their homes, offices, and schools [1–9]. The World Health Organization (WHO) has recognized healthy indoor air as a fundamental human right [4]. Comprehensive indoor air quality measurements are needed in many regions of the world to provide reliable data for evaluation of human exposure to particulate and gaseous indoor air pollutants [10].
Indoor air pollutant concentrations depend on the dynamic relationship between pollutant source and loss processes within buildings. Source processes include the transport of outdoor air pollution, which can be high in urban areas [11–13], into the indoor environment via ventilation and infiltration, and indoor emission sources, which include solid fuel combustion, electronic appliances, cleaning, consumer products, occupants, pets, and volatilization of chemicals from building materials and furnishings, among others [10,14–28]. Loss processes include ventilation, exfiltration, deposition to indoor surfaces, filtration and air cleaning, and pollutant transformations in the air (i.e., coagulation, gas-phase reactions). Indoor emission sources can result in substantial increases in indoor air pollutant concentrations, exceeding contributions from the transport of outdoor air pollutants indoors. Air cleaning technologies, such as heating, ventilation, and air conditioning (HVAC) filters and portable air cleaners, can reduce concentrations of various indoor air pollutants.
Evaluation of indoor air pollution and concentrations of particulate and gaseous indoor air pollutants in Middle Eastern dwellings has been given limited attention in the literature. In Jordan, one study investigated the effects of indoor air pollutants on the health of Jordanian women [29] and three studies evaluated concentrations of indoor particles in Jordanian indoor environments [30–32]. These studies provided useful insights on the extent of air pollution in selected Jordanian indoor environments and the role of cultural practices on the nature of indoor emission sources. However, these studies did not provide detailed information on the composition of indoor air pollution, including indoor particle number and mass size distributions, concentrations of ultrafine particles (UFPs, diameter < 0.1 μm), and concentrations of various gaseous pollutants.
The objective of this study was to evaluate size-fractionated number and mass concentrations of indoor particles (aerosols) in selected Jordanian residential indoor environments and human inhalation exposures associated with a range of common indoor emission sources prevalent in Jordanian dwellings, such as combustion processes associated with cooking, heating, and smoking. The study was based upon a field campaign conducted over two seasons in which portable aerosol instrumentation covering different particle size ranges was used to measure particle number size distributions spanning 0.01–25 μm during different indoor activities.
| doab | 2025-04-07T03:56:58.110178 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 148
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.150 | *2.1. Residential Indoor Environment Study Sites in Jordan*
The residential indoor environments targeted in this study were houses and apartments covering a large geographical area within Amman, the capital city of Jordan (Figure 1). The selection was based upon two main criteria: (1) prevalence of smoking indoors and (2) heating type, such as kerosene heaters, natural gas heaters, and central heating systems. The selected residential indoor environments included two apartments (A), one duplex apartment (D), three ground floor apartments (GFA), and two houses (H). Table 1 lists the characteristics of each study site. All indoor environments were naturally ventilated. The occupants documented their activities and frequency of cooking, heating, and smoking during the measurement campaign.
**Figure 1.** A map showing the Amman metropolitan region with the locations of the selected indoor environment study sites. The type of dwelling is referred to as: (A) apartment, (H) house, (D) duplex apartment, and (GFA) ground floor apartment. Table 1 provides additional details for each dwelling.
**Table 1.** Characteristics of the selected residential indoor environments. The heating method refers to: kerosene heater (Ker.), natural gas heater (Gas), air conditioning system (AC), electric heaters (El.), and central heating system (Cen.). Cigarette smoking is denoted as (Cig.).
#### *2.2. Indoor Aerosol Measurements and Experimental Design*
| doab | 2025-04-07T03:56:58.110401 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 150
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.151 | 2.2.1. Measurement Campaign
Indoor aerosol measurements were performed during two seasons: winter and summer, as indicated in Table 2. The winter campaign occurred from 23 December 2018 to 12 January 2019. All eight study sites participated in the winter campaign. The summer campaign occurred from 16 May to 22 June 2019. Only GFA2, GFA3, and H2 participated in the summer campaign.
**Table 2.** Measurement periods and lengths of the two campaigns.
### 2.2.2. Aerosol Instrumentation
Aerosol instrumentation included portable devices to monitor size-fractionated particle concentrations. Supplemental measurements of selected gaseous pollutants were also conducted. The aerosol measurements included particle number and mass concentrations within standard size fractions: submicron particle number concentrations, micron particle number concentrations, PM10, and PM2.5. Table 3 provides an overview of the portable aerosol instrumentation deployed at each study site. The use of portable aerosol instruments has increased in recent years, with a number of studies evaluating their performance in the laboratory, the field, or through side-by-side comparisons with more advanced instruments [33–46]. The instruments were positioned to sample side-by-side without the use of inlet extensions. The instruments were situated on a table approximately 60 cm above the floor inside the living room of each dwelling. The sample time was set to 1 min for all instruments, either by default or through time-averaging of higher sample frequency data.
**Table 3.** List of the portable air quality instruments and the measured parameters.
Two condensation particle counters (CPCs) with different lower size cutoffs (TSI 3007-2: cutoff size 10 nm; TSI P-Trak 8525: cutoff size 20 nm) were used to measure total submicron particle number concentrations. The maximum detectable concentration (20% accuracy) was 105 cm−<sup>3</sup> and 5 <sup>×</sup> 105 cm−<sup>3</sup> for the CPC 3007 and the P-Trak, respectively. The sample flow rate for both CPCs was 0.1 lpm (inlet flow rate of 0.7 lpm). A handheld optical particle counter (AeroTrak 9306-V2, TSI, MI, USA) was used to monitor particle number concentrations within 6 channels (user-defined) in the diameter range of 0.3–25 μm. The cutoffs for these channels were defined as 0.3, 0.5, 1, 2.5, 10, and 25 μm. The sample flow rate was 2.83 lpm. A handheld laser photometer (DustTrak DRX 8534, TSI, MI, USA) monitored particle mass (PM) concentrations (PM1, PM2.5, respirable (PM4), PM10, and total) in the diameter range of 0.1–15 μm (maximum concentration of 150 mg/m3). The sample flow rate for the DustTrak was 3 lpm. A personal aerosol monitor (SidePak AM520, TSI, MI, USA) with a PM2.5 inlet was used for additional measurements of PM2.5 concentrations. The SidePak is a portable instrument with a small form factor equipped with a light-scattering laser photometer. The CPCs were calibrated in the laboratory [40], whereas the AeroTrak, DustTrak, and SidePak were factory calibrated. Additionally, a portable gas monitor (S500, AeroQual, New Zealand) estimated the concentrations of gaseous pollutants by installing factory calibrated plug-and-play gas sensor heads. The sensor heads included ozone (O3), formaldehyde (HCHO), carbon monoxide (CO), nitrogen dioxide (NO2), sulfur dioxide (SO2), and total volatile organic compounds (TVOCs).
Each instrument was started at different times during the campaigns; and thus, they did not record concentrations at the same time stamp. Therefore, we interpolated the concentrations of each instrument into a coherent time grid so that we evaluated the number of concentrations in each size fraction with the same time stamp. The built-in temperature and relative humidity sensors used in the aerosol instruments cannot be confirmed to be accurate for ambient observations because these sensors were installed inside the instruments and can be affected by instrument-specific conditions, such as heat dissipation from the pumps and electronics. Therefore, those observations were not considered here.
| doab | 2025-04-07T03:56:58.110511 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 151
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.152 | *2.3. Processing of Size-Fractionated Aerosol Concentration Data*
The utilization of portable aerosol instruments with different particle diameter ranges and cutoff diameters enables derivations of size-fractionated particle number and mass concentrations [47]: Super-micron (1–10 μm) particle number and mass concentrations, submicron (0.01–1 μm) particle number concentrations, PM2.5 mass concentrations, PM10 mass concentrations, and PM10–1 mass concentrations. Additionally, we derived the particle number size distribution *n*0 *<sup>N</sup>* <sup>=</sup> *dN dlog*(*Dp*) within eight diameter bins:
The particle mass size distribution was estimated from the particle number size distribution by assuming spherical particles:
$$m\_M^0 = \frac{dM}{d\log\left(D\_p\right)} = \frac{dN}{d\log\left(D\_p\right)}\frac{\pi}{6}D\_p^3\rho\_p = n\_N^0\frac{\pi}{6}D\_p^3\rho\_p\tag{1}$$
where *n*<sup>0</sup> *<sup>M</sup>* is the particle mass size distribution, *dM* is the particle mass concentration within a certain diameter bin normalized to the width of the diameter range *dlog Dp* of that diameter bin, *dN* is the particle number concentration within that diameter bin (also normalized with respect to *dlog Dp* to obtain the particle number size distribution, *n*<sup>0</sup> *<sup>N</sup>*), *Dp* is the particle diameter, and ρ*<sup>p</sup>* is the particle density, here assumed to be unit density (1 g cm−3). In practice, the particle density is size-dependent and variable for different aerosol populations (i.e., diesel soot vs. organic aerosol); therefore, size-resolved effective density functions should be used. However, there is limited empirical data on the effective densities of aerosols produced by indoor emission sources. Thus, the assumption of 1 g cm−<sup>3</sup> for the particle density will result in uncertainties (over- or underestimates, depending on the source) in the estimated mass concentrations.
The size-fractionated particle number concentration was calculated as:
$$PN\_{D\_{p2} - D\_{p1}} = \bigcap\_{D\_{p1}}^{D\_{p2}} n\_N^0(D\_P) \cdot d\log(D\_P) \tag{2}$$
where *PNDp*2−*Dp*<sup>1</sup> is the calculated size-fractionated particle number concentration within the particle diameter range *Dp*1*–Dp*2. Similarly, the size-fractionated particle mass concentration *PMDp*2−*Dp*<sup>1</sup> was calculated as:
$$P M\_{D\_{p2} - D\_{p1}} = \int\_{D\_{p1}}^{D\_{p2}} n\_M^0(D\_P) \cdot d\log(Dp) = \int\_{D\_{p1}}^{D\_{p2}} n\_N^0(D\_P) \frac{\pi}{6} D\_p^3 \rho\_{p'} d\log(Dp) \tag{3}$$
PM2.5 and PM10 can be also calculated by using Equation (3) and integrating over the particle diameter range starting from 10 nm (i.e., the lower cutoff diameter according to our instrument setup) and up to 2.5 μm (for PM2.5) or 10 μm (for PM10).
| doab | 2025-04-07T03:56:58.110779 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 152
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.154 | *3.1. Comparisons between Di*ff*erent Aerosol Instruments—Technical Notes*
The co-location of different aerosol instruments covering similar size ranges provides a basis to compare concentration outputs as measured through different techniques. First, the PM2.5 and PM10 concentrations reported by the DustTrak can be compared to evaluate the contribution of the submicron fraction to the total PM concentration in Jordanian indoor environments. According to the DustTrak measurements, it was observed that most of the PM was in the submicron fraction as the mean PM10/PM2.5 ratio was 1.03 ± 0.04 (Figure 2). This was somewhat expected as most of the tested indoor activities in this field study were combustion processes (smoking, heating, and cooking) that produce significant emissions in the fine particle range. However, more sophisticated aerosol instrumentation would be needed to verify this finding, such as an aerodynamic particle sizer (APS) and scanning mobility particle sizer (SMPS).
**Figure 2.** Comparison between the PM10 and PM2.5 concentrations measured with the DustTrak.
The DustTrak and SidePak both employ a light-scattering laser photometer to estimate PM concentrations. As such, their output can be compared for the same particle diameter range. In general, the PM2.5 concentrations measured with the DustTrak were lower than the corresponding values measured with the SidePak (Figure 3). This trend was consistent across the measured concentration range from approximately 10 to >1000 μg/m3. The mean SidePak/DustTrak PM2.5 concentration ratio was 2.15 ± 0.48. These differences can be attributed to technical matters related to the internal setup of the instruments and their factory calibrations. For example, the SidePak inlet has an impactor plate with a specific aerodynamic diameter cut point (here chosen as PM2.5), whereas the DustTrak differentiates the particle size based solely on the optical properties of particles.
Following the methodology outlined in Section 2.3, we converted the measured particle number size distributions (via CPC 3007, P-Trak, and AeroTrak) to particle mass size distributions assuming spherical particles of unit density. From integration of the latter, we calculated the PM2.5 and PM10 concentrations. The calculated PM2.5 and PM10 concentrations can be compared with those reported by the DustTrak. The calculated PM2.5 concentrations were found to be less than those reported by the DustTrak (Figure 4). More variability was observed for PM10, with the calculated PM10 both underand overestimating the DustTrak-derived values across the measured concentration range. The mean calculated-to-DustTrak PM2.5 ratio was 0.63 ± 0.58 and that for PM10 was 1.46 ± 1.27.
**Figure 3.** Comparison between the PM2.5 concentrations measured with the DustTrak and SidePak.
**Figure 4.** Comparison between the PM2.5 and PM10 concentrations measured with DustTrak and those calculated using the measured particle number size distributions, assuming spherical particles of unit density.
This brief comparative analysis of the PM concentrations measured by the DustTrak, SidePak, and calculated via measured particle number size distributions illustrates that portable aerosol instruments have limitations and their output is likely to be inconsistent. Relying on a single instrument output may not provide an accurate assessment of PM concentrations. The utilization of an array of portable aerosol instruments can provide lower and upper bounds on PM concentrations in different indoor environments. Calculating PM concentrations from measured particle number size distributions is uncertain in the absence of reliable data on size-resolved particle effective densities for different indoor emission sources.
| doab | 2025-04-07T03:56:58.110943 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 154
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.155 | *3.2. Overview of Indoor Particle Concentrations in Jordanian Dwellings*
| doab | 2025-04-07T03:56:58.111176 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 155
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.156 | 3.2.1. Indoor Particle Concentrations during the Winter Season
An overview of the indoor submicron particle number (PN) concentrations and PM2.5 and PM10 concentrations is presented Tables 4 and 5 (mean ± SD and 95%) and illustrated in Figure 5 for each of the eight Jordanian dwellings investigated in this study. Particle concentration time series are presented in the supplementary material (Figures S1–S8). Indoor particle concentrations (mean ± SD) were also evaluated during the nighttime, when there were no indoor activities reported in the dwellings and the concentrations were observed to be at their lowest levels (Table 6).
**Table 4.** Indoor particle number and mass concentrations (mean ± SD and 95%) during the winter campaign.
**Table 5.** Indoor particle number and mass concentrations (mean ± SD and 95%) during the summer campaign.
**Figure 5.** Overall mean indoor particle concentrations during the measurement period in each dwelling: (**a**) submicron particle number (PN) concentrations measured with the condensation particle counter (CPC 3007) and (**b**) PM2.5 concentrations measured with the DustTrak. The blue bars represent the winter campaign and the orange bars represent the summer campaign.
Submicron PN concentrations were the lowest in apartment A2, which was equipped with an air conditioning (AC) heating/cooling setting and nonsmoking occupants. For example, the overall mean submicron PN concentrations in A2 was approximately 1.6 <sup>×</sup> 104 cm−3. The second lowest PN concentrations were observed in the ground floor apartment GFA2, which was equipped with a central heating system (water radiators) and, periodically, electric heaters. Occupants in GFA2 were nonsmokers. The overall mean submicron PN concentration in GFA2 was approximately double that of A2 at 3.2 <sup>×</sup> 104 cm<sup>−</sup>3.
The highest submicron PN concentrations were measured in duplex apartment D1, with a mean of 1.3 <sup>×</sup> 105 cm−3. This apartment had a kerosene heater and one of the occupants smoked shisha (waterpipe or hookah) on a daily basis. The second highest submicron PN concentrations were observed in houses H1 and H2, with overall mean values of 1.2 <sup>×</sup> 105 cm−<sup>3</sup> and 9.7 <sup>×</sup> 10<sup>4</sup> cm−3, respectively. House H1 was heated by using a natural gas heater and smoking shisha was often conducted by more than one occupant. House H2 was heated with a kerosene heater and cooking activities occurred frequently.
The ground floor apartments, GFA3 and GFA1, showed intermediate submicron PN concentrations among the study sites, with mean concentrations of 6.3 <sup>×</sup> 104 cm−<sup>3</sup> and 5.4 <sup>×</sup> 104 cm−3, respectively. Although occupants in GFA3 heavily smoked tobacco and shisha, the concentrations were lower than those observed in D1 and H1, where shisha was also smoked. The building envelopes of D1 and H1 may be more tightly sealed, with lower infiltration rates compared to GFA3. Furthermore, GFA3 used a natural gas heater and cooking activities were not as frequent. As for GFA1, the heating was a combination of a kerosene heater and a natural gas heater. The cooking activities in GFA1 were minimal and not frequent. Occupants in apartment A1 were nonsmokers. Indoor emission source manipulations were conducted in A1, including various cooking activities and the use of three different types of heating (kerosene heater, natural gas heater, and AC). The overall mean submicron PN concentration in A1 was approximately 4.3 <sup>×</sup> 104 cm<sup>−</sup>3.
For PM2.5 concentrations, the lowest levels were observed not in A2 (highest submicron PN concentrations), but rather in GFA2, with a mean of approximately 29 μg/m3. GFA2 was heated by means of a central heating system and, periodically, with electric heaters. Ground floor apartment GFA1 and apartment A2 exhibited intermediate overall mean PM2.5 concentrations among the study sites, with mean values of 42 μg/m<sup>3</sup> and 44 μg/m3, respectively. As previously discussed, the occupants in GFA1 did not conduct frequent cooking activities and heated their dwelling by means of kerosene and natural gas heaters, whereas A2 was heated via an AC. GFA1 was built in the 1970s, whereas A2 was relatively new (less than 10 years old); therefore, A2 is expected to be a more tightly sealed indoor environment compared to GFA1. However, infiltration rate and air leakage (i.e., blower door) measurements were not conducted for the dwellings in this study.
Apartment A1, in which manipulations of various cooking activities and heating methods were conducted, showed an overall mean PM2.5 concentration of 91 μg/m3. The impact of shisha smoking on PM2.5 concentrations in D1 and H1 was clearly evident, with overall mean PM2.5 concentrations of 131 μg/m<sup>3</sup> and 138 μg/m3, respectively. The influence of a kerosene heater and intense cooking activities in H2 was also evident, with an overall mean PM2.5 concentration of 156 μg/m3. The highest PM2.5 concentrations were recorded in GFA3 (approximately 433 μg/m3), which reflects the frequent shisha and tobacco smoking in this dwelling.
In the absence of indoor activities (Table 6), the submicron PN concentrations were the lowest (approximately 6 <sup>×</sup> 103 cm<sup>−</sup>3) in A1 and A2 and the highest in D1 (approximately 1.3 <sup>×</sup> 104 cm<sup>−</sup>3) and GFA3 (approximately 1.5 <sup>×</sup> 104 cm<sup>−</sup>3). As for the PM2.5 concentrations measured with the DustTrak, the lowest concentrations (approximately 10 μg/m3) were observed in A2 and GFA2 and the highest concentrations were observed in GFA3 (approximately 67 μg/m3). It is important to note that the measured indoor particle concentrations were primarily the result of the transport of outdoor particles indoors via ventilation and infiltration. However, indoor-generated aerosols during the day may still have traces overnight. For example, the dwellings with combustion and smoking activities also had background concentrations higher than other dwellings. Furthermore, differences in background concentrations among dwellings can be due to the geographical location of the dwelling within the city; this might reflect the outdoor aerosol concentrations at a given location [16,48].
#### 3.2.2. Indoor Particle Concentrations: Summer Versus Winter
Indoor aerosol measurements were repeated for three apartments in the summer campaign. We selected a dwelling (H2) that was heated with a kerosene heater and had nonsmoking occupants, a dwelling (GFA2) that was not heated with combustion processes and had nonsmoking occupants, and a dwelling (GFA3) that was heated with a natural gas heater and the occupants were smokers. Although the number of selected indoor environments was fewer in the summer campaign, the measurement period in each dwelling was longer and more extensive than what was measured during the winter campaign.
In general, the observed concentrations during the summer campaign were lower than those observed during the winter campaign (Tables 4 and 5, Figure 5). The overall mean submicron PN concentration during the summer campaign in GFA2 was approximately 1.5 <sup>×</sup> 10<sup>4</sup> cm<sup>−</sup>3, which was about 40% of that during the winter campaign. As for the PM2.5 concentrations, the overall mean during the summer campaign was approximately 30 μg/m3, which was almost the same as that observed during the winter campaign.
The overall mean submicron PN concentrations in GFA3 and H2 were similar (approximately 1.6–1.9 <sup>×</sup> 10<sup>4</sup> cm−3), whereas the corresponding mean PM2.5 concentrations were higher in H2 (approximately 46 μg/m3) compared to GFA3 (approximately 31 μg/m3). The summer/winter ratio for submicron PN concentrations for GFA3 and H2 were 0.3 and 0.2, respectively. The corresponding PM2.5 ratios were approximately 0.1 and 0.3. The primary reason for higher particle concentrations during the winter was the use of fossil fuel combustion for heating (i.e., kerosene and natural gas
heaters). Furthermore, the dwellings during the summer were more likely to be better ventilated than during the winter, when the dwellings had to conserve energy during heating periods.
| doab | 2025-04-07T03:56:58.111203 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 156
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.157 | *3.3. Indoor Particle Number and Mass Size Distributions in Jordanian Dwellings*
| doab | 2025-04-07T03:56:58.111627 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 157
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.158 | 3.3.1. Indoor Particle Size Distributions in the Absence of Indoor Activities
The mean particle number and mass size distributions for each dwelling in the absence of indoor activities during the winter campaign are presented in Figure S9. Significant differences in the mean particle number and mass size distributions were observed among the eight dwellings. Based on the number size distributions, the submicron PN concentration was the lowest (approximately 6 × 10<sup>3</sup> cm−3, with a corresponding PM2.5 of 5 μg/m3) in dwellings A1 and A2 and the highest in GFA3 (approximately 1.5 <sup>×</sup> 104 cm<sup>−</sup>3, with a corresponding PM2.5 of 12 <sup>μ</sup>g/m3) and D1 (approximately 1.3 <sup>×</sup> 104 cm−3, with a corresponding PM2.5 of 8 <sup>μ</sup>g/m3). The mean submicron PN concentration was between 9 <sup>×</sup> 103 cm−<sup>3</sup> and 10<sup>4</sup> cm−<sup>3</sup> and the mean PM2.5 was 7–9 <sup>μ</sup>g/m3 in the remainder of the dwellings. It should be noted that GFA3 had the highest submicron PN concentration, whereas H2 had the highest PM2.5 concentration (approximately 13 μg/m3). Differences between the PN and PM concentrations among the eight dwellings is an indicator of variability in the shape and magnitude of the aerosol size distributions, as illustrated in Figure S9.
The coarse PN concentrations were the lowest in A1 (approximately 0.4 cm<sup>−</sup>3, with a corresponding PMcoarse of 0.9 μg/m3) and D1 (approximately 0.4 cm−3, with a corresponding PMcoarse of 1.3 μg/m3) and the highest was in H2 (approximately 5.2 cm<sup>−</sup>3, with a corresponding PMcoarse of 39.9 μg/m3) and the second highest was in H1 (approximately 2.5 cm<sup>−</sup>3, with a corresponding PMcoarse of 17.3 μg/m3). As for A2, GFA1, and GFA3, the coarse PN concentrations were approximately 0.9 cm−<sup>3</sup> for each of the dwellings, but the corresponding PMcoarse was about 6.3, 3.5, and 5.6 μg/m3, respectively. The similarity in the coarse PN concentrations, compared to the differences observed for the PMcoarse concentrations, in these dwellings is an indication of differences in the coarse size fraction of the indoor particle size distributions. This likely reflects differences in indoor emission sources of coarse particles among the dwellings. For example, H2 had the highest coarse PN and PM concentrations which could be explained by the existence of pets (more than two cats), in addition to the geographical location of this dwelling, which was close to an arid area in southeast Amman, where dust events and coarse particle resuspension are common.
| doab | 2025-04-07T03:56:58.111655 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 158
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.159 | 3.3.2. Overall Mean Indoor Particle Number and Mass Size Distributions
The overall mean particle number and mass size distributions were calculated for each dwelling for the entire winter measurement campaign (Figures 6 and 7). This includes periods with and without indoor activities. In the following section, we will present and discuss the characteristics of the indoor particle number and mass size distributions during different indoor activities. Each dwelling had a unique set of particle number and mass size distributions that reflected the indoor aerosol emission sources associated with the inhabitants' activities, heating processes, and dwelling conditions. For example, among all dwellings, the lowest UFP concentrations were observed in apartment A2 because combustion processes (i.e., cooking using a natural gas stove) were minimal and the indoor space was heated via AC units. GFA2 had the second lowest UFP concentrations because the heating was via water-based central heating and, occasionally, electric heaters. Furthermore, both A2 and GFA2 were nonsmoking dwellings.
**Figure 6.** Mean particle number size distributions calculated for the entirety of the winter measurement campaign at each dwelling: (**a**) apartment A1, (**b**) ground floor apartment GFA1, (**c**) duplex D1, (**d**) ground floor apartment GFA3, (**e**) house H1, (**f**) apartment A2, (**g**) house H2, and (**h**) ground floor apartment GFA2.
**Figure 7.** Mean particle mass size distributions calculated for the entirety of the winter measurement campaign at each dwelling: (**a**) apartment A1, (**b**) ground floor apartment GFA1, (**c**) duplex D1, (**d**) ground floor apartment GFA3, (**e**) house H1, (**f**) apartment A2, (**g**) house H2, and (**h**) ground floor apartment GFA2.
Indoor combustion processes had a pronounced impact on submicron particle concentrations, especially UFPs. For example, the impact of using kerosene heaters was evident in A1, D1, GFA1, and H2. Similarly, the impact of using natural gas heaters was evident in A1, GFA1, GFA3, and H1. Shisha smoking was reported in D1, GFA3, and H1, and the impact can be seen in the high concentrations of UFPs that were measured. D1 never obtained a stable background aerosol concentration during the nighttime likely due to traces of the kerosene heater and shisha smoking.
| doab | 2025-04-07T03:56:58.111799 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 159
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.160 | 3.3.3. The Impact of Indoor Activities on Indoor Particle Size Distributions and Concentrations
As listed in Table 1, the heating processes reported in this study included both combustion (natural gas heater and/or kerosene heater) and non-combustion (central heating, electric, and air conditioning). The cooking activities were reported on stoves using natural gas. The use of microwaves, coffee machines, and toasters were very rare. Table 7 presents a classification of selected activities and the mean PN and PM concentrations during these activities. The location (i.e., dwelling) and duration of the activities are listed in Table S1. Figures S9–S17 in the supplementary material present the mean particle number and mass size distributions during these activities. In this section, the reported PM concentrations were calculated from the particle mass size distributions by assuming spherical particles of unit density, as previously discussed.
**Table 7.** Classification of indoor activities and corresponding particle number and mass concentrations. Combustion heating is denoted as (Heat.) and the types are natural gas heater (NG) and kerosene heater (K). Cooking on a natural gas stove is denoted as (Stov.) and smoking cigarettes is denoted by (Cig.).
Cooking Activities without Combustion Processes
Cooking activities were the most commonly reported indoor emission source in all eight dwellings. Periodically, they were reported in the absence of combustion processes (such as a natural gas stove or heating). The non-combustion cooking activities included: microwave (GFA2, Figure S17), brewing coffee (A1, Figure S10), and toasting bread (A1, Figure S10). When compared to the background concentrations (i.e., in the absence of indoor activities), the concentrations during these activities had a minor impact on the indoor air quality in each dwelling.
Brewing coffee had the smallest impact on indoor aerosol concentrations, with a mean calculated PM2.5 concentration of approximately 7 <sup>μ</sup>g/m3 (submicron PN concentration of 1.1 <sup>×</sup> 10<sup>4</sup> cm−3) and mean calculated PM10 concentration of approximately 31 μg/m<sup>3</sup> (coarse PN concentration of 1 cm<sup>−</sup>3). Using the toaster doubled the PM2.5 concentration and increased the submicron PN concentration four-fold. However, it had a negligible impact on the coarse PN and PM concentrations. Using the microwave had a similar impact on concentrations of fine particles as that observed when using a toaster.
| doab | 2025-04-07T03:56:58.112015 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 160
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.161 | Cooking Activities in the Absence of Combustion Heating Processes
Cooking on a stove (natural gas) can be classified as light or intensive. Light cooking activities were reported in dwelling A1 as cooking soup and making chai latte (Figure S10). During these two activities, the mean calculated PM2.5 concentration was approximately 40 μg/m3. The mean submicron PN concentration was approximately 1.4 <sup>×</sup> 105 cm−<sup>3</sup> and 1.6 <sup>×</sup> 105 cm−<sup>3</sup> during cooking soup and making chai latte, respectively. The corresponding calculated PM10 concentrations were approximately 144 μg/m<sup>3</sup> and 160 μg/m3 and the coarse PN concentrations were approximately 3 cm−<sup>3</sup> and 1 cm<sup>−</sup>3, respectively. Here, the differences in the PM10 and coarse PN concentrations were unlikely due to the cooking processes, but rather driven by occupancy and occupant movement-induced particle resuspension near the instruments, which was more intense during cooking soup.
Light cooking activities (such as making tea and/or coffee) were also reported in GFA2, which had a central heating system. During the making of tea and coffee, the mean calculated PM2.5 concentrations were approximately 16 μg/m<sup>3</sup> and 31 μg/m3, respectively (Figure S17). The mean submicron PN concentrations were approximately 1.2 <sup>×</sup> 105 cm−<sup>3</sup> and 4.6 <sup>×</sup> 104 cm<sup>−</sup>3, respectively. The corresponding calculated PM10 concentrations were approximately 52 μg/m<sup>3</sup> and 42 μg/m3, respectively, and the coarse PN concentrations were about 1 cm<sup>−</sup>3. This indicates that similar activities might have different impacts on particle concentrations depending on the indoor conditions and the way in which the activity was conducted. For example, variability in dwelling ventilation may play a role, as well as the burning intensity of the natural gas stove.
Intensive cooking activities were reported in dwelling GFA2 (Figure S17, central heating) and A2 (Figure S15, AC heating). Indoor aerosol concentrations during these intensive cooking activities were higher than those observed during light cooking activities (in the absence of combustion heating processes). For example, the mean calculated PM2.5 concentrations were between 62 μg/m<sup>3</sup> and 88 <sup>μ</sup>g/m3. The mean submicron PN concentrations were between 7.4 <sup>×</sup> 104 cm−<sup>3</sup> and 2.1 <sup>×</sup> 105 cm<sup>−</sup>3. The corresponding mean calculated PM10 concentrations were between 112 μg/m3 and 201 μg/m<sup>3</sup> and the mean coarse PN concentrations were between 3 cm−<sup>3</sup> and 14 cm<sup>−</sup>3.
| doab | 2025-04-07T03:56:58.112166 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 161
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.162 | Concurrent Cooking Activities and Combustion Heating Processes
Periodically, the cooking activities occurred concurrently with a combustion heating process (natural gas or kerosene heaters). All of these cooking activities, aside from two, did not report the type of cooking; therefore, it was not possible to classify them as light or intensive cooking. One of the activities was very intensive cooking (grilling burger and sausages) and the other one was a birthday party (candles burning with more than 15 people in the living room). During cooking activities accompanied by a natural gas heater, the mean calculated PM2.5 concentrations were between <sup>9</sup> <sup>μ</sup>g/m3 and 70 <sup>μ</sup>g/m3 (submicron PN concentrations between 6.8 <sup>×</sup> <sup>10</sup><sup>4</sup> cm−<sup>3</sup> and 2.7 <sup>×</sup> <sup>10</sup><sup>5</sup> cm<sup>−</sup>3). The corresponding mean calculated PM10 concentrations were between 16 μg/m3 and 81 μg/m3.
Grilling had a significant impact on indoor aerosol concentrations: the mean calculated PM2.5 concentration was approximately 378 <sup>μ</sup>g/m<sup>3</sup> (submicron PN concentration of 3.8 <sup>×</sup> <sup>10</sup><sup>5</sup> cm<sup>−</sup>3) and the mean calculated PM10 concentration was approximately 2100 μg/m<sup>3</sup> (mean coarse PN concentration of 130 cm−3). The birthday party event had a clear impact on both submicron and micron aerosol concentrations: the mean calculated PM2.5 concentration was approximately 65 μg/m3 (submicron PN concentration of 1.7 <sup>×</sup> 10<sup>5</sup> cm−3) and mean calculated PM10 concentration was 374 <sup>μ</sup>g/m3. Using a kerosene heater instead of a natural gas heater further elevated the concentrations of indoor aerosols. During these activities, the mean calculated PM2.5 concentrations were between 43 μg/m3 and 130 μg/m3
(submicron PN concentration between 1.7 <sup>×</sup> 105 cm−<sup>3</sup> and 3.2 <sup>×</sup> 105 cm<sup>−</sup>3). The corresponding mean calculated PM10 concentrations were between 90 μg/m3 and 460 μg/m3.
| doab | 2025-04-07T03:56:58.112300 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 162
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.163 | Indoor Smoking of Shisha and Tobacco
Smoking indoors is prohibited in Jordan. However, this is often violated in many indoor environments in the country. In this study, shisha smoking and/or tobacco smoking was reported in three dwellings (GFA3, H1, and D1). It was not possible to separate the smoking events from the combustion processes used for heating or cooking. Therefore, the concentrations reported here were due to a combination of smoking and heating/cooking activities.
Tobacco smoking increased indoor aerosol concentrations as follows: the mean calculated PM2.5 concentrations were between 40 <sup>μ</sup>g/m<sup>3</sup> and 100 <sup>μ</sup>g/m3 (submicron PN concentrations between 9 <sup>×</sup> 104 cm−<sup>3</sup> and 1.5 <sup>×</sup> 105 cm−3). The corresponding mean calculated PM10 concentrations were between 160 μg/m<sup>3</sup> and 190 μg/m3 (mean coarse PN concentrations between 6 cm−<sup>3</sup> and 8 cm−3). Shisha smoking had a more pronounced impact on indoor aerosol concentrations compared to tobacco smoking. The mean calculated PM2.5 concentrations were between 60 μg/m<sup>3</sup> and 140 μg/m3 (submicron PN concentrations between 1.2 <sup>×</sup> 105 cm−<sup>3</sup> and 4 <sup>×</sup> 105 cm−3). The corresponding mean calculated PM10 concentrations were between 90 μg/m3 and 290 μg/m3 (mean coarse PN concentrations between 2 cm−<sup>3</sup> and 15 cm<sup>−</sup>3).
For shisha smoking, the tobacco is mixed with honey (or sweeteners), oil products (such as glycerin), and flavoring products. Charcoal is used as the source of heat to burn the shisha tobacco mixture. Usually, the charcoal is heated up indoors on the stove prior to the shisha smoking event. Shisha and cigarette smoking produces a vast range of pollutants in the form of primary and secondary particulate and gaseous pollution [49–58]. It was also reported that cigarette and shisha smoke may contain compounds of microbiological origin, in addition to hundreds of compounds of known carcinogenicity and inhalation toxicity [49].
| doab | 2025-04-07T03:56:58.112416 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 163
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.164 | *3.4. Concentrations of Selected Gaseous Pollutants in Jordanian Dwellings*
The indoor activities documented in the eight dwellings were associated with emissions of gaseous pollutants for which exceptionally high concentrations were observed (Figures S1–S8). For example, the shisha smoking and preceding preparation (i.e., charcoal combustion) were associated with CO concentrations that reached as high as 10 ppm in D1 and GFA3. The CO concentrations were further elevated in H1, with concentrations approaching 100 ppm. Emissions of SO2 were also recorded in D1 during charcoal combustion that accompanied shisha smoking. During shisha smoking, the CO concentrations exceeded the exposure level of 6 ppm due to smoking a single cigarette, as reported by Breland et al. [56], and 2.7 ppm as reported by Eissenberg and Shihadeh [52]. Previous studies have reported CO concentrations in the range of 24–32 ppm during shisha smoking events [51–53].
The eight dwellings exhibited variable concentrations of TVOCs, NO2, and HCHO. For instance, TVOC concentrations were in the range of 100–1000 ppm in A2 and H2, whereas they were in the range of 1000–10,000 ppm in all ground floor apartments (GFA1, GFA2, and GFA3). NO2 concentrations were in the range of 0.01–1 ppm in the duplex apartment (D1), ground floor apartments (GFA1, GFA2, and GFA3), and houses (H1 and H2). HCHO concentrations were in the range of 0.01–1 ppm in A2 and GFA1 and reached as high as 5 ppm in H2. O3 was not detected in any of the dwellings. It should be noted that the gaseous pollutant concentrations presented here are estimates and are likely uncertain due to technical limitations of the low-cost sensing module employed.
### *3.5. Indoor Versus Outdoor Particle Concentrations*
It is important to note that the indoor aerosol measurement periods at each dwelling were short during the winter campaign. Outdoor aerosol measurements were made on a few occasions at each dwelling; however, they were not of sufficient length to make meaningful conclusions about the aerosol indoor-to-outdoor relationship. However, comprehensive measurements of ambient aerosols have been made in the urban background in Amman [40,41,59–62], for which comparisons with the indoor measurements presented in this study can be made.
In the urban background atmosphere of Amman [62], outdoor PN concentrations were typically higher during the winter compared to the summer; the ratio can be 2–3 based on the daily means. Based on the hourly mean, the outdoor PN concentration had a clear diurnal and weekly pattern, with high concentrations during the workdays, especially during traffic rush hours. For example, the PN concentration diurnal pattern was characterized by two peaks: morning and afternoon. The afternoon peak (wintertime highest concentration range of 3 <sup>×</sup> 104–3.5 <sup>×</sup> 104 cm<sup>−</sup>3) was rather similar on all weekdays; however, the first peak was higher on workdays compared to weekends (wintertime highest concentration range of 4.5 <sup>×</sup> 104–6.5 <sup>×</sup> 104 cm−3). The lowest outdoor concentrations were typically observed between 3:00 to 6:00 in the morning, when they are as low as 1.8 <sup>×</sup> 104 cm−<sup>3</sup> during the wintertime.
When compared to the results reported in this study (Tables 4–7), the mean indoor PN concentrations were generally higher than those outdoors during the daytime, when indoor activities were taking place. For example, PN concentrations inside all dwellings were less than 1.5 <sup>×</sup> <sup>10</sup><sup>4</sup> cm−<sup>3</sup> between midnight and early morning; i.e., in the absence of indoor activities. However, the overall mean PN concentrations during the winter campaign inside the studied dwellings were in the range of 1.6 <sup>×</sup> 104–1.3 <sup>×</sup> 105 cm<sup>−</sup>3. Looking at the mean concentrations during the indoor activities, the PN concentrations were as high as 4.7 <sup>×</sup> 10<sup>4</sup> cm−<sup>3</sup> during non-combustion cooking activities. During cooking activities conducted on a natural gas stove, the PN concentrations were in the range of 4.6 × <sup>10</sup>4–3.8 <sup>×</sup> <sup>10</sup><sup>5</sup> cm<sup>−</sup>3. The combination of cooking activities and combustion processes (as the main source of heating) increased the PN concentrations to be in the range of 6.8 <sup>×</sup> 104–2.7 <sup>×</sup> 10<sup>5</sup> cm−3. Grilling sausages and burger indoors was associated with a substantial increase in mean PN concentrations, with levels reaching as high as 3.8 <sup>×</sup> 105 cm−<sup>3</sup> (PM2.5 = 378 <sup>μ</sup>g/m<sup>3</sup> and PM10 = 2094 <sup>μ</sup>g/m3). Both tobacco and shisha smoking were also associated with significant increases in PN concentrations, with levels reaching 9.1 <sup>×</sup> 104–4.0 <sup>×</sup> <sup>10</sup><sup>5</sup> cm<sup>−</sup>3.
It is very well documented in the literature that the temporal variation in indoor aerosol concentrations closely follows those outdoors in the absence of indoor activities [20,30,32,63–74]. As such, the aerosol indoor-to-outdoor relationship depends on the size-resolved particle penetration factor for the building envelope, the ventilation and infiltration rates, and the size-resolved deposition rate onto available indoor surfaces [20,30,64]. As can be seen here, and also reported in previous studies, indoor aerosol emission sources, which are closely connected to human activities indoors, produce aerosol concentrations that are usually several times higher than those found outdoors [17,75–77]. Indoor aerosol sources can thus have a significant adverse impact on human health given that people spend the majority of their time indoors [10,11,32].
| doab | 2025-04-07T03:56:58.112546 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 164
} |
001fcd9a-cd27-45e8-b23f-f447874e0974.165 | **4. Conclusions**
Indoor air quality has been given very little attention in the Middle East. Residential indoor environments in Jordan have unique characteristics with respect to size, ventilation modes, occupancy, activities, cooking styles, and heating processes. These factors vary between the winter and summer. In this study, we reported the results of one of the first comprehensive indoor aerosol measurement campaigns conducted in Jordanian indoor environments. Our methodology was based on the use of portable aerosol instruments covering different particle diameter ranges, from which we could investigate particle number and mass size distributions during different indoor activities. We focused on standard particle size fractions (submicron versus micron, fine versus coarse). The study provides valuable information regarding exposure levels to a wide range of pollutant sources that are commonly found in Jordanian dwellings.
In the absence of indoor activities, indoor PN concentrations varied among the dwellings and were in the range of 6 <sup>×</sup> 103–1.5 <sup>×</sup> 104 cm−<sup>3</sup> (corresponding PM2.5 of 5–12 <sup>μ</sup>g/m3). The coarse PN concentrations were in the range of 0.4–5.2 cm−<sup>3</sup> (corresponding PMcoarse of 0.9–39.9 μg/m3). Indoor activities significantly impacted indoor air quality by increasing exposure to particle concentrations that exceeded what could be observed outdoors. Non-combustion cooking activities (microwave, brewing coffee, and toasting bread) had the smallest impact on indoor aerosol concentrations. During such activities, the PN concentrations were in the range of 1.1 <sup>×</sup> 104–4.7 <sup>×</sup> 10<sup>4</sup> cm<sup>−</sup>3, PM2.5 concentrations were in the range of 7–25 μg/m3, micron PN concentrations were in the range of 1–9 cm−3, and PM10 concentrations were in the range of 44–181 μg/m3. Cooking on a natural gas stove had a more pronounced impact on indoor aerosol concentrations compared to non-combustion cooking, with measured PN concentrations in the range of 4.6 <sup>×</sup> 104–2.1 <sup>×</sup> 105 cm−3, PM2.5 concentrations in the range of 16–88 μg/m3, micron PN concentrations in the range of 1–14 cm−3, and PM10 concentrations in the range of 42–201 μg/m3.
The combination of cooking activities (varying in type and intensity) with heating via combustion of natural gas or kerosene had a significant impact on indoor air quality. PN concentrations were in the range of 6.8 <sup>×</sup> 104–2.7 <sup>×</sup> 105 cm−3, PM2.5 concentrations were in the range of 9–130 <sup>μ</sup>g/m3, micron PN concentrations were in the range of 1–27 cm<sup>−</sup>3, and PM10 concentrations were in the range of 16–458 μg/m3. Grilling sausages and burgers indoors was identified as an extreme event, with mean PN concentration reaching 3.8 <sup>×</sup> 105 cm<sup>−</sup>3, PM2.5 concentrations reaching 378 <sup>μ</sup>g/m3, micron PN concentrations reaching 131 cm<sup>−</sup>3, and PM10 concentrations reaching 2094 μg/m3.
Both tobacco and shisha smoking adversely impacted indoor air quality in Jordanian dwellings, with the latter being more severe. During tobacco smoking, the PN concentrations were in the range of 9.1 <sup>×</sup> 104–1.5 <sup>×</sup> 105 cm−3, PM2.5 concentrations were in the range of 40–98 <sup>μ</sup>g/m3, micron PN concentrations were in the range of 6–8 cm−3, and PM10 concentrations were in the range of 158–189 <sup>μ</sup>g/m3. During shisha smoking, the PN concentrations were in the range of 1.2 <sup>×</sup> 105–4.0 <sup>×</sup> 105 cm<sup>−</sup>3, PM2.5 concentrations were in the range of 61–173 μg/m3, micron PN concentrations were in the range of 2–36 cm<sup>−</sup>3, and PM10 concentrations were in the range of 92–424 μg/m3.
The above-mentioned concentration ranges were reported during the winter campaign, when the houses were tightly closed for heating purposes. Indoor aerosol concentrations during the summer campaign were generally lower. The overall mean PN concentrations during the summer campaign were less than 2 <sup>×</sup> 104 cm−<sup>3</sup> and PM2.5 concentrations were less than 50 <sup>μ</sup>g/m3. Some of the reported indoor activities were accompanied with high concentrations of gaseous pollutants. TVOC concentrations exceeded 100 ppm. NO2 concentrations were in the range of 0.01–1 ppm. HCHO concentrations were in the range of 0.01–5 ppm. During shisha smoking and preceding preparation (e.g., charcoal combustion), the mean CO concentrations reached as high as 100 ppm.
There are a number of limitations of the present study: (1) the measurement periods were short at each dwelling during the winter campaign, (2) the sample population was small (eight dwellings), and (3) outdoor measurements were only conducted on a few occasions for short periods. These limitations can be addressed in future indoor–outdoor measurement campaigns in Jordan. However, indoor aerosol concentrations were compared to long-term outdoor PN measurements conducted in past studies in Jordan.
The results of this study can offer several practical recommendations for improving indoor air quality in Jordanian indoor environments: source control by prohibiting the smoking of tobacco and shisha indoors, improved ventilation during the use of fossil fuel combustion for heating, and cooking with a natural gas stove under a kitchen hood.
**Supplementary Materials:** The following are available online at http://www.mdpi.com/2073-4433/11/1/41/s1. Table S1: Average particle mass and number concentrations (mean ± stdev) during selected indoor activities. Figure S1: Aerosol concentrations inside apartment A1 during the winter campaign (23–25 December 2018). Figure S2: Aerosol concentrations inside ground floor apartment GFA1 during the winter campaign (25–27 December 2018). Figure S3: Aerosol concentrations inside duplex apartment D1 during the winter campaign (28–30 December 2018). Figure S4: Aerosol concentrations inside ground floor apartment GFA3 during the winter campaign (31 December 2018–2 January 2019). Figure S5: Aerosol concentrations inside house H1 during the winter campaign (2–4 January 2019). Figure S6: Aerosol concentrations inside apartment A2 during the winter campaign (4–5 January 2019). Figure S7: Aerosol concentrations inside house H2 during the winter campaign (6–9 January 2019). Figure S8: Aerosol concentrations inside ground floor apartment GFA2 during the winter campaign (9–12 January
2019). Figure S9: Mean particle number size distributions and corresponding particle mass size distributions in the absence of indoor activities during the winter campaign at each study site. Figure S10: Mean particle number size distributions and particle mass size distributions during selected activities reported inside Apartment A1 during the winter campaign (23–25 December2018). Figure S11: Mean particle number size distributions and particle mass size distributions during selected activities reported inside ground floor apartment GFA1 during the winter campaign (25–27 December 2018). Figure S12: Mean particle number size distributions and particle mass size distributions during selected activities reported inside duplex D1 during the winter campaign (28–30 December 2018). Figure S13: Mean particle number size distributions and particle mass size distributions during selected activities reported inside ground floor apartment GFA3 during the winter campaign (31 December 2018–2 January 2019). Figure S14: Mean particle number size distributions and particle mass size distributions during selected activities reported inside house H1 during the winter campaign (2–4 January 2019). Figure S15: Mean particle number size distributions and particle mass size distributions during selected activities reported inside apartment A2 during the winter campaign (4–5 January 2019). Figure S16: Mean particle number size distributions and particle mass size distributions during selected activities reported inside house H2 during the winter campaign (6–9 January 2019). Figure S17: Mean particle number size distributions and particle mass size distributions during selected activities reported inside ground floor apartment GFA2 during the winter campaign (9–12 January 2019).
**Author Contributions:** Conceptualization, T.H., M.M., A.A.-H., and O.A.; methodology, T.H., O.J., K.A., A.A., and O.A.; validation, T.H.; formal analysis, T.H., O.J., and A.A.; investigation, T.H.; resources, T.H. and M.M.; data curation, T.H., O.J., K.A., and A.A.; writing—original draft preparation, T.H. and A.A.-H.; writing—review and editing, T.H., B.E.B., A.J.K., J.L., M.M., and A.A.-H.; visualization, T.H.; supervision, T.H. and A.A.-H.; project administration, T.H. and A.A.-H.; funding acquisition, T.H. and M.M. All authors have read and agreed to the published version of the manuscript.
**Funding:** This research was funded by the World Health Organization regional office in Amman. The research infrastructure utilized in this project was partly funded by the Deanship of Academic Research (DAR, project number 1516) at the University of Jordan and the Scientific Research Support Fund (SRF, project number BAS-1-2-2015) at the Jordanian Ministry of Higher Education. This research was part of a close collaboration between the University of Jordan and the Institute for Atmospheric and Earth System Research (INAR/Physics, University of Helsinki) via the Academy of Finland Center of Excellence (project No. 272041 and 1307537).
**Acknowledgments:** The first author would like to thank the occupants for allowing the indoor measurement campaigns to be conducted in their dwellings. Some of them also helped in follow-up aerosol measurements and reporting of indoor activities. This manuscript was written and completed during the sabbatical leave of the first author (T.H.) that was spent at the University of Helsinki and supported by the University of Helsinki during 2019. Open access funding was provided by the University of Helsinki.
**Conflicts of Interest:** The authors declare no conflict of interest.
| doab | 2025-04-07T03:56:58.112877 | 11-1-2022 14:33 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "001fcd9a-cd27-45e8-b23f-f447874e0974",
"url": "https://mdpi.com/books/pdfview/book/3938",
"author": "",
"title": "Indoor Air Quality: From Sampling to Risk Assessment in the Light of New Legislations",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036509464",
"section_idx": 165
} |
00210593-c9b9-4d6f-9f8b-238faf7dc64f | **Life Course Research and Social Policies 17**
# Sandra V. Constantin
# A Life Course Perspective on Chinese Youths
From the Transformation of Social Policies to the Individualization of the Transition to Adulthood
# **Life Course Research and Social Policies**
Volume 17
#### **Series Editors**
Flavia Fossati, Bâtiment IDHEAP, Quartier UNIL-Mouline, Lausanne, Switzerland Andreas Ihle, University of Geneva, Geneva, Switzerland Jean-Marie Le Goff, LIVES, Batiment Geopolis, University of Lausanne, Lausanne, Switzerland Núria Sánchez-Mira, Institute of Sociology, University of Neuchâtel, Neuchâtel, Switzerland Matthias Studer, NCCR Lives, University of Geneva, Genève, Geneve, Switzerland
The book series puts the spotlight on life course research. The series publishes monographs and edited volumes presenting theoretical, methodological, and empirical advances in the study of the life course, thereby elaborating on possible implications for society and social policies applications. Topics appropriate for the series include, among others, the following:
#### *Life Course Research and Social Policies*
Books commissioned for the series aim to encourage debates on life course research in various countries and regions across the world. Volumes in this series will be of interest to researchers, professionals and policy makers in social sciences and related felds.
The series is edited by a team of scholars affliated to the Swiss LIVES Centre: Flavia Fossati (UNIL), Andreas Ihle (UNIGE), Jean-Marie Le Goff (UNIL), Núria Sánchez-Mira (UNIL) and Matthias Studer (UNIGE).
If you are interested in flling a gap in coverage, providing a focus on a certain area, or contributing a new perspective or approach, we would be delighted to receive a book proposal from you. The book proposal should include a description of the proposed book, Table of Contents, unique or special features compared to competing titles, anticipated completion date, and CV with brief biography. The book proposal can be sent to the publisher, Evelien Bakker, at evelien.bakker@ springer.com.
Sandra V. Constantin
# A Life Course Perspective on Chinese Youths
From the Transformation of Social Policies to the Individualization of the Transition to Adulthood
Sandra V. Constantin Institute of Sociological Research (ISR) University of Geneva Geneva, Switzerland
ISSN 2211-7776 ISSN 2211-7784 (electronic) Life Course Research and Social Policies ISBN 978-3-031-57215-9 ISBN 978-3-031-57216-6 (eBook) https://doi.org/10.1007/978-3-031-57216-6
This work was supported by Schweizerischer Nationalfonds zur Förderung der Wissenschaftlichen Forschung
© The Editor(s) (if applicable) and The Author(s) 2024 . This book is an open access publication.
**Open Access** This book is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.
The images or other third party material in this book are included in the book's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the book's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specifc statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affliations.
This Springer imprint is published by the registered company Springer Nature Switzerland AG The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
If disposing of this product, please recycle the paper.
# **Preface**
This book explores whether and how trends in individualization have shaped the transition to adulthood of two cohorts of Chinese living in Beijing in the 2010s, the frst one born between 1950 and 1959, in the aftermath of the foundation of the PRC, and the second one constituted by their children, born between 1980 and 1989. It revisits individualization in China through the prism of the transition to adulthood. A major contribution of this book, which is based on a Ph.D. thesis in Sociology defended at the University of Geneva in 2017, is that it proposes new insights based on original data, both qualitative and quantitative, on what it might mean to make the transition to adulthood in China, a context that has experienced major social, political and economical shifts during the last 50 years, with potential consequences for life chances of distinct cohorts. This issue has sadly been largely neglected by previous research based on the life-course paradigm. The quantitative approach presented in the book uses an original convenience sample collected by the author as part of her doctoral feldwork. It is coupled with several qualitative data collections, such as in-depth interviews and analysis of television series, also part of the doctoral research.
The book provides a number of interesting fndings, some of which I wish to highlight in this preface. The frst thing that stands out in my view is the high degree of standardization in educational pathways and family formation in the two cohorts studied. In both cohorts, a small number of types of life course are suffcient to account for the variety of life courses available to individuals. Indeed, we fnd ample evidence in the book that the cohorts under consideration show a high degree of homogeneity. This standardization of actual life trajectories goes hand in hand with a strong standardization of biographisation, i.e. the way actors defne their life projects, with a massive focus on educational and occupational success and further upward social mobility. The importance of being seen by others as a 'quality person' is well described and explained in the book. Overall, individuals belonging to the two cohorts on which Dr Constantin's research focuses are anchored in highly standardized projects and life trajectories. The book provides a breadth of information and analysis about such individual standardized life courses that makes it relevant to scholars seeking to understand the shaping of life trajectories by individualization trends outside Europe and North America.
This standardization remains, the book reveals, under the crucial infuence of inherited social statuses, through the 'hukou' (household register, residence permit), and parental social background (as revealed by parental educational attainment). Far from freeing individuals from the grip of their family group or class destiny, changes in Chinese society over the past 50 years seem to have reinforced the impact of inherited rather than achieved social statuses. The book thus provides crucial evidence about the limits and peculiarities of individualization trends in urban China, showing the maintenance of classical logics of social reproduction in shaping life trajectories. According to Dr Constantin's research, such trajectories remain under massive normative control by the State through a variety of institutional instruments, such as TV series, the imposition of the notion of the 'quality person', the residence permit, access control to higher education and the Communist Party, laws on marriage and family and so on. Such institutions, the book shows, in most cases do not open up biographical possibilities but, on the contrary, contribute to the formation of an elite consisting mainly of university graduates who follow a standard life course.
These are important contributions to the understanding of China as a national context that is still largely neglected by life-course studies, as the book rightly points out. Moreover, they provide us with some clues as to how individualization processes may have been constructed in countries that have followed different historical trajectories than Europe or North America, the regions from which the bulk of life-course studies have been drawn.
These are important contributions to the understanding of a national context that is still largely neglected by life-course studies, as the book points out. They are also important because they give us a better understanding of how individualization processes may have been constructed in countries that have not followed the same historical courses as Europe and North America, the regions from which the bulk of life-course studies to date have been drawn.
Department of Sociology Eric D. Widmer University of Geneva Geneva, Switzerland
# **Acknowledgements**
This book is the fruit of many fascinating encounters, enriching collaborations and stimulating scientifc discussions.
I would like to express my deep gratitude to all the people in Beijing who agreed to share the intimacy of their life stories with me. I hope I have lived up to their expectations, providing as accurate a picture as possible of their paths towards adulthood. I would also like to thank Tong Xing for welcoming me so warmly and integrating me into the Department of Sociology at Peking University and for the theoretical discussions we had on the issue of individualization during my stay in Beijing. The discussions I was able to have with my PhD comrades and Master's students at Peking University greatly enriched my analyses. I would like to thank Zhou Lüjun for his insight, sound advice and availability, as well as all the interviewers who helped me collect quantitative data in the feld; without their enthusiasm, this research would not have been possible.
Numerous scientifc exchanges upstream and downstream of the feld research also enriched my research. In particular, I would like to thank my colleagues from the NCCR LIVES, in particular Jean-Marie Le Goff and Eric Widmer, for their insights on the life-course theory and the construction of biographical analysis models. I would like to thank Glen Elder for taking the time to write to me and call me while I was in Beijing to advise me on drawing up the data collection protocol. I also wish to thank Laura Bernardi for giving me the opportunity to publish my PhD thesis in the form of this book.
Many thanks to Evelien Bakker and Bernadette Deelen-Mans from Springer for their patience and encouragements all along the way. Synthesizing 700 pages of analysis into a single book was not an easy game!
I would like to thank Marylène Lieber, Isabelle Attané, Isabelle Thireau, Olivier Galland, Nicolas Zufferey, Michel Bonnin and Nadia Sartoretti for their comments on earlier versions of the manuscript.
I wish to thank the members of the Demography, gender and society unit at the French Institute for Demographic Studies (INED), the members of the Centre d'études sur la Chine moderne et contemporaine, Laura Downs from the European University Institute in Florence and Mary Daly from the Department of Social
Policy and Intervention at the University of Oxford for their warm welcome during my scientifc peregrinations. They all contributed in one way or another in the fnal outcome.
This book would not have been possible without funding from the University of Geneva, the Ernst and Lucie Schmidheiny Foundation, the Geneva-Asia Association, Peking University and the Swiss National Science Foundation.
Last but not least, my heartfelt thanks go to Yvan, who has supported me every step of the way, and to Maria who proofread this manuscript.
# **Contents**
# **Chapter 1 Introduction**
#### **Contents**
Bibliography 3
*"Like in a night, the members of the post-80 generation have become "old". First, they are nostalgic. They sing "the old boy", they nostalgically remember the summers when they wore striped T-shirts and leather sandals, read comics, and chased after girls. They sigh that they are old. Some who have become parents seem to still be children, facing their children younger than them they sigh "we have become old", "the tired heart, with the feeling of not being able to love again" ... ".*
The author of the article continues: *"What makes a young generation full of vigor and vitality become lethargic? [...]".*
This article, untitled "Do not let youth become apathetic" 1 , was published on May 14, 2013, in the "opinions" section of the People's daily (*Renmin ribao*). At the time it sparked strong reactions2 . Why these harsh judgments on this generation, which was the frst one since the decollectivization and the country's opening-up to market economy to grow up and move towards adulthood?
<sup>1</sup>People's Daily, May 14, 2013, *Don't let youth be tainted with the air of twilight (Mò ràng qīngchūn rǎn mùqì).* http://opinion.people.com.cn/n/2013/0514/c1003-21470995.html. Retrieved on May 15, 2013. "It seems that overnight, the post-80s generation has collectively become 'old'. First, there is nostalgia. They sing 'Old Boys', lamenting the summers of wearing sea soul shirts and leather sandals that have faded in memory, reminiscing about the comic books they have read, and the girls they chased together in those years. Then, there is the sigh of aging. A group of post-80s who are still children in their parents' eyes, sigh 'I'm old', 'My heart is so tired, I feel like I can't love anymore' in front of children younger than themselves... What is it that makes the young generation, who should be full of vitality, become so gloomy?".
<sup>2</sup>Other press articles qualify young adults born in the 1980s as the "tragic generation (杯具的一代 *, bei ju de yi dai*)". This last expression refers to a popular saying ("life is a cup, it's up to you to decide if it's a cup for drinking or a cup for brushing your teeth"). It is composed of several puns: In Chinese the word cup (杯子, *beizi* ) is a homophone for the word life (辈子, *beizi* ). The word cup for drinking (杯具 ,*beiju* ) is a homophone of the word tragedy (悲剧, *beiju* ). Finally, the word cup for brushing your teeth (洗具, *xiju* ) is a homophone of the word comedy (戏剧 ,*xiju* ).
S. V. Constantin, *A Life Course Perspective on Chinese Youths*, Life Course Research and Social Policies 17, https://doi.org/10.1007/978-3-031-57216-6\_1
In this new socioeconomic context, social risks are no longer assumed by the collective, as was the case for their parents who were socialized in Maoist China. These risks have become more and more individualized. For instance, young adults are increasingly enjoined to seek solutions by themselves to respond to new social risks such as: unemployment, precariousness, increased competition in the labor market, as well as the disintegration of the social security system and, at the same time, they are enjoined to fulfll the obligation – not only customary but also legal – to provide moral and/or fnancial support to their aging parents.
These socioeconomic and institutional transformations that have marked China over the past 30 years have been extensively studied by Chinese and Western researchers. However, very few have considered these developments as part of a process of individualization (Delman & Yin, 2008; Hansen & Pang, 2008, 2010; Hansen & Svarverud, 2010; Li, 2010; Mühlhahn, 2010; Rolandsen & H., 2008; Svarverud, 2010; Thogersen & Ni, 2008; Wedell-Wedellsborg, 2010; Yan, 2009). Among the scientifc literature that adopts this theoretical approach, the analysis presented focuses only on a specifc dimension of this process. Moreover, the fndings are based mostly on qualitative data collected either in urban or rural areas.
In this landscape, the present research, that aims to capture the dynamics of individualization in post-collectivist China, is unique because it is based on longitudinal data – qualitative and quantitative – collected among both urban and rural populations. It also has the interest of focusing on an aspect that has been left in the shadows by Chinese and Western researchers: the life-course of young adults. To my knowledge, no research mobilizes a life-course approach and, more specifcally, the study of the transition to adulthood to analyze this process. Shaped by various institutional regulations, the life-course is a privileged site of observation to grasp the different forms taken by the individualization process.
The analyses presented in this book focus on the life-courses of two emblematic birth cohorts at the crossroad between two eras: collectivist and post-collectivist China. It took the Chinese only 40 years to live through two such dissimilar eras. These ideological, economic, social and institutional transformations, condensed in time and without equivalent, are embodied in the life-course of two birth cohorts. On the one hand, there is the "generation of the new China (*xin zhongguo de yidai*)", born between 1950 and 1959, in the aftermath of the foundation of the PRC. This birth cohort grew up and experienced its transition to adulthood in Maoist China. On the other hand, their children, the "generation of reforms (*zhuanxing de yidai)* ", born between 1980 and 1989 and whose members – now adults – have lived their entire lives in the era of reforms initiated by Deng Xiaoping.
This book is structured around three parts. The frst part, entitled revisiting individualization in China through the prism of the transition to adulthood, consists of three chapters. The frst chapter traces the path followed by the individualization process in China from the pre-revolutionary period to today and highlights its characteristics. The second chapter questions this path in the light of two governability instruments: the discourse on the quality of the population (*renkou suzhi*) and the residence booklet system (*hukou*). This chapter considers television series as a privileged place of observation of the norms and values carried by the Party-State. The
third chapter details the theoretical framework and the methodological design elaborated to examine the impact of the dismantling of the collectivist social security system on the dynamics of (in)equality over the life-course. It explains why I have chosen to analyze the path to adulthood of two emblematic birth cohorts as a period of observation. It also details how quantitative and qualitative methods were articulated, and how the two birth cohorts' life-course were reconstructed and analyzed.
The second part of the book opens with a ffth chapter that examines the paths to adulthood. Along with the presentation of quantitative analysis, it reveals the meaning given by young adults to each transition to adulthood. It shows that the paths to adulthood tend to stretch over time due to the emergence of a new social time in the life-course of young adults: higher education. In this chapter, qualitative analyses also highlight that, unlike Western young people who emphasize individualistic criteria, the respondents underline the centrality of family roles and responsibilities in the meaning they give to the transition to adulthood. A sixth chapter sheds light on new social risks that result from decollectivization and the process of individualization of the social security system, as well as their impact on young adults' professional paths. Data not only show that young adults' work life is becoming more fragile and precarious, but also that they are unequal in facing the challenges they must overcome. The supports they can expect vary according to their anchoring into the social stratifcation.
The third part of the book questions the impact of decollectivization and the rise of uncertainties on the dynamics of family transitions. The three chapters around which this last part is articulated reveal the emergence of neo-familialism. It is showed in a seventh chapter that young adults tend to postpone the formation of a family, notably because of the precariousness of the labor market. The eighth chapter reveals that they also aspire to have more intimacy, in comparison to their parents' generation. The ninth and fnal chapter reveals that if the failures of the welfare state place a number of socio-economic risks on young adults and their families' shoulders, family solidarities are transforming: these solidarities tend to become elective and to concentrate on the nuclear family.
#### **Bibliography**
**Open Access** This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.
The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.
# **Part I Revisiting Individualization in China Through the Prism of the Transition to Adulthood**
Nowadays in China, neoliberal thought is highly infuential in the economic and social spheres, as well as in the formation of public policies. Neoliberalism encourages individuals to become autonomous and refexive. In other words, it incites people to become self-critical and to take responsibility for the successes and failures over their life-course, while inscribing their actions within the framework of existing norms (Cosmo, 2007). According to this meaning, neoliberal thought produces a form of individualization.
According to Elias, this social process results from the transition from "traditional societies" to "modern societies" (Elias, 1991). If in so-called "traditional" societies, the functions of control and protection of individuals were ensured by the community of origin (the clan, the village, the lord, the corporation or the class), in the societies described by Elias as "modern" because they are highly urbanized, it is state organizations and institutions that fulfll these functions. In the latter, individuals that become more mobile gradually break away from the limits of the community of origin. As a result, their dependence and identifcation with these groups tends to decrease, they tend to become autonomous and to differentiate themselves.
In this frst part, after having examined in a frst chapter the transformations of the relationship between Individual-Society-State from the pre-revolutionary period to the present day, I will present in a second chapter the discourse on the quality of the population. We will see that this discourse can be understood as an instrument mobilized by the Party-State to support the socio-economic development of the country while, at the same time, reinforcing the dynamics of individualization. A third chapter will present the methodological design developed to capture the Chinese characteristics of this process and its impact on the life-course of young adults. This chapter will also reveal the advantages of a life-course approach for analyzing, not only, the biographical structuring of the events and transitions experienced by individuals, but also, the resources mobilized, and the roles held by them over their life while capturing the role played by the sociohistorical context to shape their life-course. The approach developed in this book is innovative. It links three
sociological felds: sociology of China, sociology of youth and sociology of lifecourse that have, to my knowledge, never been held together in French-speaking research and, barely in English and Chinese-speaking research.
#### **Bibliography**
Cosmo, H. (2007). *Contested individualization. Debates about contemporary personhood*. Palgrave Macmillan.
Elias, N. (1991). *La Société Des Individus*. Fayard.
# **Chapter 2 Institutionalized Individualism in Post-collectivist China**
#### **Contents**
**Abstract** This second chapter explores the dynamics of individualization in China across time to unravel the complexities of the individualization process in a rapidly transforming society.
The pre-Maoist period, often mischaracterized as solely group-oriented, is reevaluated through the lens of Chinese sociologist Fei Xiaotong. It not only reveals a society rooted in individual autonomy within a network of personal relationships, but it also sheds light on how the transformative intellectual movements of the early twentieth century sought to redefne the "Chinese national character" to modernize the country, challenging traditional family structures and promoting individual autonomy. The May 4th movement marked a peak in this reformist wave, introducing elements of individualism that clashed with Confucian morals.
Furthermore, the chapter reveals that during the Maoist period, despite a shift towards collectivism, Maoist policies contributed to a partial disembedding of individuals from traditional solidarities, particularly for women.
In a last movement, the chapter shed light on the subsequent economic reforms from the late 1970s and how it marked a new phase. While the Party-State maintains strong political and economic control, it enjoins individuals to take responsibility for their life-course. It supports individualization in the labor market while rejuvenating family obligations and solidarities.
**Keywords** Individualization process · Complexity · Traditional solidarities · Mao · Collectivization · Reforms
#### **2.1 Varieties of Individualization**
Individualism is a polysemic concept, which took shape at the beginning of the nineteenth century to characterize the frst European modernity (Elias, 1991; Le Bart, 2008). Although authors differentiate several forms of individualism1 , in the collective imagination, this concept remains negatively connoted. It is often associated with economic neoliberalism and competition between individuals, which can lead to individuals becoming withdrawn, selfsh, and indifferent to others (Singly (de), 2011). However, individualism also refers notably to the dignity and sanctity of the person, to autonomy and the ability to act sovereignly, to respect for private intimacy and self-development (Corcuff, 2003; Giddens, 2015; Le Bart, 2008; Singly (de), 2011).
According to proponents of the theory of individualization, under the effect of globalization2 , traditional social anchors are crumbling and dissolving. Individuals are gradually detached from the family, tradition, and the collective, which prescribed their ways of acting and their behaviors. This disembedding3 of individual life paths from the traditional social fabric has the consequence of affecting the relationship that the individual maintains with the social. While individuals have gained control over their life-course by freeing themselves from some social determinisms, they have also become responsible for the successes and failures encountered during their biographical trajectories. Similarly, the individual autonomy gained in relation to the family and the collective, thanks to the development of the European welfare state, has the corollary of a greater dependence of individuals on institutions (schools, social insurances, etc.). These institutions contribute to the (re)structuring of life-course by imposing new constraints and requirements, sometimes contradicting each other. Individualization constitutes, in this sense, a complex process of socialization because it sometimes relies on contradictory injunctions (Bauman, 2001; Beck, 2008; Beck & Beck-Gernsheim, 2011; Giddens, 2007).
<sup>1</sup>Bourgeois individualism, capitalist individualism (Le Bart, 2008:91–92), market individualism (Aron, 2002; Le Bart, 2008:93), authoritarian individualism, possessive individualism, romantic individualism (Le Bart, 2008:66–97), citizen individualism (Singly (de), 2011), creative individualism, individualism of otherness or dissimilarity (Le Bart, 2008:100–104), anomic individualism (Lash, 2011) or even modern individualism (Le Bart, 2008:116).
<sup>2</sup>Beck prefers the notion of *cosmopolitanization* to that of *globalization*. The *cosmopolitanization* refers according to him to a concrete multidimensional process, characterized by interdependencies, cultural mixtures and a common destiny that binds humanity (Beck & Grande, 2010). For Beck, the *globalization* is an ambivalent notion, which is generally apprehended from an economic point of view.
<sup>3</sup>The notions of embedding and disembedding refer respectively to the dependence and independence that individuals maintain in relation to various aspects of the social world (institutions, politics, economy, etc.), which constitute for them both resources and constraints: "Embedding is a process of increasing dependencies and [disembedding] decoupling is a process of empowerment, of strengthening specifcity, of emergence" (Grossetti, 2015:8).
In Europe, as elsewhere in the world, this historical process of socialization does not result from the free choice of individuals but from the transformation of societies. European welfare states and the benefts delivered by these tend to punctuate and standardize the temporality of biographical trajectories (Kohli, 2007; Mayer & Schoepfin, 2009). However, since the 1970s, this "social compromise" is gradually being undermined. The rise of unemployment and the precariousness of working conditions compel workers to be fexible. Therefore, individuals are called upon to become refexive. This is the promotion of the "biographical model" or "biographization" (Beck, 2008; Martuccelli & de Singly, 2010). In other words, individuals must become the "entrepreneurs of their own lives" by making the necessary choices and sometimes reconversions in their careers, their hobbies or even their emotional lives (Ehrenberg, 1995). This form of freedom is "precarious" since individuals are both constrained to navigate in a context determined by socioeconomic and political structures and they have no certainty about the expected effects of their choices. Moreover, not everyone is equally armed to deal with the demand for refexivity (Furlong & Cartmel, 2007). Some are better endowed than others with objective resources (economic capital, cultural capital, and social capital) on which they can rely to make their decisions. Consequently, this contributes to maintaining, or even reinforcing, social inequalities which become less visible, as the process of individualization tends to give individual colorations to systemic social problems.
In China, the process of individualization is also present (Hansen & Svarverud, 2010; Yan, 2009). As will be developed in the following sections, contrary to the received idea that in pre-Maoist China the individual was subject to the group and according to which collective interests predominated over individual interests, there were already forms of individualization at this time, even if the interdependencies between individuals remained strong and exerted tight control over them4 . Similarly, in Maoist China, despite collectivist economic and social policies, some policies paradoxically contributed to individualization. These social felds that had been individualized would, to some extent, serve the post-collectivist project three decades later.
#### **2.2 The Individual and the Shadow of Ancestors in Pre-Maoist China**
The pre-Maoist period is understood in this book as a historical stage that extends from the end of the two Opium Wars, which preceded the fall of the Manchu dynasty in 1911 and the proclamation of the Republic of China on January 1st 1912, to the foundation of the People's Republic of China (PRC), on October 1st 1949.
<sup>4</sup>The social felds covered here: the family feld, the religious feld and the administrative feld.
According to the Chinese sociologist Fei Xiaotong, pre-Maoist China was never solely group-oriented. It was, he explains, traditionally centered on the individual and woven from networks created from personal relationships (*guanxi*) linking the individual to other individuals in multiple directions and placing in each relational context individual and specifc moral obligations (Fei, 1992). Individuals were interdependent and they retained a form of autonomy. Apart from the father-son relationship for which social relations were prescribed by a fxed status and responsibilities, individuals had the freedom to choose whether or not to enter a social relationship and to defne their own roles and those of others, as well as the boundaries of the groups around which they revolved.
In this societal organization, the individuals evolved in a network of social relations, where they were emotionally attached to the obligations defned by these relations. This embedding of social relations was intended to maintain and guarantee social stability (Fei, 1992). This organization of social ties is found in the thought of Confucius, who describes fve fundamental relationships: son/daughter, husband/ wife, father/mother, brother/sister, and friends. The individual is therefore led to defne himself more in terms of his relationships with others than by himself. If forms of individualism could exist at this time, they were strongly internalized and could not be directly expressed socially.
During the pre-Maoist period, large families were an ideal (Hsu, 1971). Therefore, nuclear-type families were not considered as independent units5 (Fei, 1992:83). Social stability was ensured by the institutionalization of interdependent social networks. In other words, social stability was safeguarded through legitimate patriarchal authority in the family sphere, the authority of the elders in the villages, and social control exercised by the notables. In the system of norms and values of the time, the relationship between father and son was fundamental for the perpetuation of the family lineage. It served as a link between the (paternal) ancestors and the descendants of the same family lineage. The organization of this system was fundamentally unequal, as it was legitimately admitted that husbands had a higher status than their wives and brothers than their sisters. Romantic love had very little place. Marriages were indeed generally arranged by parents. In these arrangements, The wife's role was to perpetuate the family lineage (on the paternal side) (Hsu, 1971:240–41).
It was in opposition to this organization of society that, from the 1910s, the notion of the "ideal man" emerged in Chinese intellectual thought. According to the intellectuals of this time, the modernization of the country had to go through fundamental changes in the "national character" (Cheng, 2009). The two Opium Wars (1842 and 1860), which marked the beginning of foreign presence in China, indeed highlighted the socio-economic backwardness of the Middle Kingdom compared to the newly industrialized Western powers (Bergère et al., 1989; Fairbank, 1989).
<sup>5</sup> "[w]e cannot say that the nuclear family household does not exist, but we should never think of it as an independent unit."
Following these wars, a feeling of humiliation, crisis, and national disintegration spread among the population.
Part of the intelligentsia then turned to Europe and Japan to fnd solutions to quickly modernize the country. The "Chinese national character", characterized by political indifference, resignation to destiny and power, a lack of individual initiative and resistance to social change, was denounced6 (Cheng, 2009:51). The writings of Liang Qichao marked the reformist movement of the time. He supported the idea that a new type of individual/citizen (*xinmin*) had to be born to allow the country to modernize (Cheng, 1997, 2009; Schell & Delury, 2013; Svarverud, 2010). This new Chinese citizen had to be characterized by "a strong sense of nationalism and patriotism, a spirit of adventure, awareness of personal rights and freedom, a sense of autonomy, self-esteem, the ability to form a cohesive community, persistence, a sense of responsibility, a militaristic mentality, a public ethic, and a private morality" (Huang, 1972:63). The "abolition" of the traditional family model and traditional social ties inherited from Confucian thought were then seen as one of the conditions for modernizing the country (King, 1996). In an editorial in the journal *The China Progress*, Liang notably claimed "gender equality, the opening of schools for women, and the abolition of foot binding" (Cheng, 1997:624).
This reformist movement gradually gained infuence and it posed in China "in an unusually acute and persistent way" the problem of the individual's autonomy in relation to the state (Fairbank & Goldman, 2010:376). A partial form of individualism then gradually crept into society. As an illustration, Chen Duxiu, the founder of the Chinese Communist Party (CCP), launched the magazine *New Youth* (*xin qingnian*) in 1915. Intellectuals began to gather in study groups, clubs, or societies in order to better disseminate their ideas within society. The most famous organization was the one founded by Mao Zedong and Cai Hesen in 1918, "the study group of the new Chinese Man" (*xinmin xuehui*) (Cheng, 2009).
The May 4th, 1919 movement (*siwu yundong*) brought this reformist movement to its climax. The leaders were mostly intellectuals trained in Japan, permeable to the ideas of Enlightenment thinkers and, for some, to Marxist ideology. Full of nationalism7 , they defended their ideas on science and democracy, while denouncing Confucian thought and the ties imposed by the traditional family system. They also defended "individual expression and even sexual freedom." According to historians Fairbank and Goldman, for the time, "[t]he romantic individualism and selfrevelation at work in some pioneers, the act of carrying a narrative in the frst person or expressing oneself in the style of a diary, all this was quite shocking in the face of strictly Confucian morals" (Fairbank & Goldman, 2010:390–91).
<sup>6</sup> "[a] political indifference, succumbing to fate with resignation, a slavish attitude toward authority, lack of individual initiative, and resistance to social change"
<sup>7</sup> In reaction to the Treaty of Versailles, the attribution of German concessions in China to Japan, and the latter's territorial claims in China.
#### **2.3 The Individual and the Collective in Maoist China**
Before the communists took power in China, Mao had formulated in three essays his idea of the revolutionary model and the ideal type of citizen that the PRC would need to build the new China (Mao, 2001).
He uses simple language and a vivid style that mobilizes the fgure of model characters8 . He thus describes the individuals that the CCP needs:
*"We must educate a lot of people – the kind of people who are the vanguard of the revolution, who have political farsightedness, who are prepared for battle and sacrifce, who are frank, loyal, positive and upright; the kind of people who seek no self-interest, but only national and social emancipation and forwardness in the face of hardships; the kind of people who are neither undisciplined nor fond of limelight, but practical, with their feet frmly on the ground. If China possesses many men like this, the tasks of the Chinese revolution can easily be fulflled." ((Chen, 1970) cited in (*Cheng*,* 2009*, p. 62)).*
He depicts in an equally vivid manner, from characters with "model" actions, the fgure of the cadre of the CCP, the soldier and the peasant that the country needs to recover.
The Maoist period (1949–1976), characterized by the state's monopoly at the political, economic and social levels, is even more marked by the instrumentalization of model fgures. These "reform of the thought" campaigns had two main objectives: on the one hand, to socialize individuals to Mao's political thought by erasing the ideological and cultural imprints of traditional China and replacing them with new ones. This political re-socialization was essential for the regime to control and mobilize the population in the implementation of new socio-economic policies. On the other hand, it was about promoting collectivism as a moral principle and as a guideline in the lives of individuals and uprooting individualism (Cheng, 2009).
The free labor market system was replaced by a state-controlled planned economic system, in which private companies were gradually eliminated. In urban areas, it became the state's responsibility to allocate everyone a job within a work unit, the *danwei*. The work unit regulated all aspects of individuals' lives: not only did it provide a lifetime job, housing, medical care, retirement allowance, but it also fnanced the education of its workers' children, monitored, and controlled the workers and their families through "the system of political fles" (Bray, 2005; Lü & Perry, 1997). The education of individuals to collectivism was thus used by the central government "to subdue individual preference in seeking a career, even choosing a residence, subjecting people to the economic needs of the Party and the government" (Bray, 2005:74–75). The aim of such campaigns was to promote
<sup>8</sup>This pictorial style is the one that was mobilized by classical thinkers such as Confucius.
collectivist and socialist values in opposition to individualism and to place the interests of the Nation above individual preferences.
Social relations were trying to be reinvented. They should no longer be centered on the individuals and their family group, but on the relationship between individuals and the Party-State. Individuals were called to reinvent themselves to defne themselves less by their family affliation than by their citizenship and class status. As of 1952, the Party-State began to structure the population into classes. *Jieji* means "class" and *chengfen* "component element", each class status (*jieji chengfen*) was associated with particular rights and prestige. There were more than 60 designations and "[e]very Chinese knew his own. In all his papers and in all the fles that concerned him, the mention of his class status was mandatory" (Billeter, 1987:143–44). Class statuses were heritable. They directly infuenced the individuals' place in society, their relationship with the Party, the ambitions they could "legitimately harbor in the political or professional feld [... and] the possibilities of social promotion of [their] children" (Billeter, 1987:144). This system, which prevailed in both urban and rural areas, had a deterministic power over individuals' life-course and all their social ties. By establishing the principle of the hereditary nature of class status, which was a determining factor in the allocation of political, social and economic resources, the Party-State nevertheless – contrary to its political ambition – contributed at the level of individual practices to maintain a link between the individual and his family group.
In rural areas, the dynamic was the same. People's communes were imposed in three stages to become *in fne* mandatory by the summer of 1958. A new collective social organization was gradually introduced: "many aspects of private life such as cooking and eating, rearing children, bathing, tailoring, and looking after the elderly were […] collectivized […]. The most conspicuous form of such collective life was the free meal in the public dining halls set up in every village" (Cheng, 2009:81). The goal of the Party-State was to substitute the family, traditional social organization, with the people's commune to eliminate the idea of private property in the minds of the rural population and to cultivate socialist and communist thought instead. Despite a strong political will, it is important to note that this project did not fully succeed in the sense that the family group remained an important social institution in the countryside.
During the Maoist period, the Party-State never stopped to mobilize and build different heroic fgures and models to shape the "new Chinese woman" and the "new Chinese man". Xiang Xiuli (Fig. 2.1) and Lei Feng (Fig. 2.2) embody, for example, revolutionary virtues. They were exemplary comrades in their daily actions. Propaganda highlights their spirit of mutual aid, camaraderie, their devotion to Mao's thought, or even the frugality of their lifestyles. Propaganda posters show them ready to sacrifce their lives for the community and the revolutionary cause.
Xiang Xiuli was born 1933 into a poor working-class family in Guangzhou. She worked in a pharmaceutical factory. In December 1958, fre broke out in her work unit. Because of the presence of infammable sodium, the blaze quickly spread. According to the Party narrative, she bravely fought the fames, suffering severe
**Fig. 2.1** Xiang Xiuli (向秀丽) braved the fre and sacrifced her life to save the machines and textile production of her work unit in Guangzhou. (Source: BG E15/775, Landsberger collection, https://chineseposters.net/themes/xiangxiuli)
burns, to save her comrades and the factory. She died 33 days later (Min et al., 2003, p.162).
The authorities not only used posters to disseminate these "model" behaviors, but they also organized group meetings and mass rallies at which citizens of the new China were expected to learn from the behavior of these models and try to live and work like them. To serve this purpose the Party State exposed, for instance, posters showing Lei Feng, a soldier and driver in the People's Liberation Army (PLA), doing good deeds and inviting the population to do the same. Lei Feng was born in 1940 in a poor rural family in Hunan province. According to the Party narrative, he became an orphan at the age of 7. In the late 1950s, he became a tractor driver, a bulldozer driver and then a truck driver in a factory. Lei Feng died in August 1962 in an accident while performing a transportation task. On March 5, 1963 China's national newspapers published images and Mao's dedication: "Learn from Comrade Lei Feng" (Fig. 2.2, top right). Later on other slogans appeared, such as: "Learn from the exemplary culture of Lei Feng. Serve the People with all your heart", or "The life of individuals is limited but serving the People has no limit. For the good of the People, study all aspects of Lei Feng's behavior". Lei Feng would have compared his loyalty to the Party to a mother's love for her child: "I am like a toddler, and the party is like my mother, who helps me, leads me, teaches me to walk…My beloved Party, my loving mother, I am always your loyal son" (Lei Feng cited in (Cheng, 2009, pp. 93–94)).
**Fig. 2.2** "Learn from Comrade Lei Feng". (Source: BG E13/428, Landsberger collection, https:// chineseposters.net/themes/leifeng-3)
Along the mobilization of heroic fgures, the Party-State created collective models. For example, the Dazhai production brigade, in Shanxi province, was promoted as a model of rural development. On one of the posters related to this brigade, we can read, written on the dam: "Farmers study [the model of] Dazhai" (Fig. 2.3).
**Fig. 2.3** "Learn to move mountains from the example set by Dazhai". (Source: BG D25/30, Landsberger collection, https://chineseposters.net/posters/d25-30)
It is written in red ink on top of the barrage "Learn to move mountains to change China" because the People's commune achieved large harvests "despite a barren mountainous landscape and the most primitive living and farming conditions" (Min et al., 2003, p.164).
In 1964, Mao Zedong called upon the Nation "In agriculture, to learn from Dazhai !". The poster below says: "Unite and fght to learn from Dazhai, work hard and change to win the harvest" (Fig. 2.4).
In Maoist China, the Party-State created an "organized dependence" of individuals on the collective. Individuals were socio-economically dependent on the work unit or commune to which they belonged, politically dependent on the State organization, and personally dependent on the cadres of the CCP. The control exercised by the collective over the individual resulted in standardized biographical trajectories among people born in the 1950s, in other words, those who grew up in Maoist China. The Party-State's plan to transform the society deeply infltrated into the family sphere. The "traditional" family organization, intergenerational solidarities, and Confucian ethics were condemned as feudal practices; while the freedom to choose a spouse in marriage, as well as gender equality, were advocated by the State (Yan, 2003a, b).
**Fig. 2.4** Workers and peasants strive hand in hand for a good harvest. (Source: BG E15/537, Landsberger collection, https://chineseposters.net/posters/e15-537)
In Maoist China, on the surface, individuals were embedded in the collectivist society: "the individual almost entirely had lost her/his freedom and autonomy as she/he could not even choose where to work or to reside, much less to which social or political group she/he would belong"; however, paradoxically, the socialist campaign of "creating a new Chinese man" allowed to some extent, at a deeper level, the individual to free himself from traditional social ties (the family, the family group, the community), as well as from traditional patriarchal and Confucian norms and values. The impact of this emancipation was particularly strong for women who were socially dominated in the organization that prevailed in pre-Maoist China.
Maoist China is characterized by a profound modifcation of economic, social, and institutional structures. This period is imprinted by the strengthening of the welfare state since biographical trajectories and risks are taken care of by either the work unit or the commune. Yet social norms under this period encourage a relative emancipation of individuals from "traditional" roles and constraints thanks to the economic and social supports provided by the institutions. The intellectual project of creating "a new Chinese man" to modernize the country did not start in the Maoist period. However, until the foundation of the PRC, this project remained at the idea stage. To realize it, the newly empowered Party-State did not hesitate to attack a central pillar of the social and economic organization of Chinese society: the family institution. By promulgating the marriage law in 1950, the CCP sought to eliminate past inequalities: "free women from the yoke of feudalism by establishing a new family regime on liberal bases. [...] the re-organization of the institution of marriage and the family attacks the foundations of traditional Chinese society, in which the all-powerful head of the family cell shields the Party's infuence on individuals: their emancipation requires that traditional structures be broken frst, and that mediating frameworks be eliminated" (Blayo, 1998:20). This law aims to free marriage from the family control by establishing the free choice of spouses, free consent to marriage, equal rights for men and women, and the right to divorce. Ultimately this law redefnes marriage "as a social contract, instead of a pact between two families" (Blayo, 1998:21).
The agrarian reform law enacted the same year and, a few years later, the policy of collectivization and the work unit system also contributed to emancipate individuals from traditional social organization forms and to empower various social felds9 . Interdependencies within the society remained nevertheless tight and exerted close control over the individual.
It should be noted that, while Maoist policies contributed to partially disembedding individuals from "traditional" solidarities, they simultaneously re-embedded them in new forms of social ties. Marriages, for example, freed from family control, were henceforth subject to the control of the Party, and the same goes for divorces, since Party offcials had decision-making power over family affairs.
Communes and work units, relying on Maoist institutions, somewhat replaced "traditional" community solidarities. Through a political project to build a "collectivist" society, the Party-State imposed roles, norms, status, and identities on individuals. Likewise, it produced generic fgures such as the worker or the student. *In fne* the bureaucracy borne of this political project was paradoxically both impersonal and individualizing.
#### **2.4 The Rise of Individualism in Post-Maoist China**
China's transition to a market economy has not only liberalized the economy, it has also gradually transformed – or even dissolved – certain social constraints that tied individuals to a specifc lifestyle. This period of transition and reforms has also allowed to broaden the range of choices for individuals.
From the late 1970s, the old members of the Chinese Communist Party, who had experienced the Long March, in particular Deng Xiaoping, returned to power with
<sup>9</sup>By social felds we mean the family feld, the political feld, the administrative feld, the economic feld, etc.
the aim of transforming the country at the socio-economic level. To this end, they undertook to shape again a "new Chinese man", who would still be subject to the Party-State politically but emancipated at the economic and social levels. They had two important assets to achieve their ambitions and undertake the economic and social reform of society: a literate and healthy male and female population (resulting from the investments of the Maoist period in these areas) and individuals eager to express their individuality. This latter dimension is apparent in the historical path of the reforms.
According to the offcial historiography, the reforms started from the Chinese countryside. During the winter of 1978, for fear of the famine that threatened, and while the memory of the disaster following the Great Leap Forward was still fresh in their minds, 18 households in the village of Xiaogang in Anhui province decided, with the support of local offcials, to freely dispose of their agricultural production, after having paid the tax due to the government and provided it with the legal grain quota. The following year, it seems that the agricultural production of this village increased, while elsewhere the trend was at best stagnation. Other villages in the county would have decided to follow the example (Kelliher, 1992; Oi, 1989; Roux, 2006). The central state would not only have turned a blind eye to such initiatives, but also, given their effectiveness at the local level, decided to apply them nationally in the 1980s by introducing the "household responsibility system". From then on, the Peasants no longer had to produce according to directives, they could freely decide on their agricultural production and sell the surplus on local markets or in cities. If decollectivization in the countryside began in the early 1980s, it was not until 1985 that the people's communes were completely dismantled and village life "deregulated" (*songbang*). These reforms allowed the countryside to gradually disembed individuals from the collective, in other words, from the people's commune. At the time, the primary sector represented the most dynamic sector of the economy. In the early 1980s, 80% of the Chinese population still lived in the countryside.
Deng Xiaoping and the reformers "crossed the river by feeling the stones"10 Pragmatic, they cautiously extended the reforms to urban areas. Initially, the government turned a blind eye to the development of an informal economy in the cities, which was the result of both peasants living in the urban periphery and coming to sell part of their harvest in the city, and young urbanites not wanting to return to the countryside where they were supposed to be re-educated by the peasant masses (Bonnin, 2004; Peng, 2009). The Party-State had an ambivalent attitude towards the re-emergence of these activities (street vendors): "on the one hand, they represented a real threat to the socialist planned economy, but on the other hand, they also helped the party-state solve the problem of mounting unemployment and declining economic growth" (Yan, 2010:496). After a decade of
<sup>10</sup>摸着石头过河, *mozhe shitou guohe*.
cultural revolution (1966–1976)11, the country was paralyzed both economically and politically. The discontent of the urban population manifested itself through strikes, absenteeism, and the grouping of young people into gangs (De Beer & Rocca, 1995). The central government, fearing strong opposition from the urban population, deemed it preferable to wait until 1984 before applying the policy of reforms and opening-up to urban areas.
In a pragmatic approach aimed at maintaining as much stability as possible, the central government initially opened "pockets" in the planned economy system, in which individuals could slip in and exploit opportunities to increase their income. In this dual system (*shuangguizhi*) where planned economy and "free" market coexist, public enterprises gradually produced a part of their production "outside the plan". "Individual industrial and commercial households" (*geti gongshang hu*) emerged and, from 1988 private enterprises developed as well12. The country's streets saw the rebirth of small local businesses, street vendors and small crafts, to which the central government recognizes "an irreplaceable utility for improving the living conditions of the people" (Resolution adopted by the third plenum of the XII e congress of the CCP in 1984, cited in Monteil, 2010:104).
The country's opening to the market economy was not done in a linear way and without encountering opposition within the government. De Beer and Rocca attribute an "uneven, sinusoidal" character to the reform policy, and Marie-Claire Bergère speaks of zigzag policy (Bergère et al., 1990; De Beer & Rocca, 1995). These reforms – institutional, economic, and social changes – initiated by the central government, which allowed the re-introduction of a partly free labor market in China, also enjoined individuals to dis-embed themselves from the Maoist dependency system.
Starting from 1992, the reforms were relaunched and deepened. The main objective was the socio-economic development of the Nation. Deng Xiaoping's formula symbolizes the spirit of this period: "It doesn't matter whether the cat is white or black, as long as it catches mice it's a good cat"13. This second phase of reforms (1992–2023) was characterized by the brutal retreat of the welfare state in China and the increased exposure of individuals to socio-economic risks. The Party-State undertook a diffcult task: the dismantling of state-owned enterprises. By the mid-1990s, 40% of workers were laid off (nearly 50 million people found
<sup>11</sup>There are two defnitions of the cultural revolution. In the strict sense, it is the period from May 1966 to April 1969 during which Mao, through the armed wing of the Red Guards, restored his power within the CCP. In the broad sense (imposed by the CCP and retained by Chinese historians), the period of the cultural revolution extends from 1966 to 1976, this latter date marking the death of Mao and the fall of the Gang of Four.
<sup>12</sup>See the work of Barry Naughton for a detailed explanation of this mechanism (Naughton, 1995, 2007).
<sup>13</sup>不管白猫黑猫 会捉老鼠就是好猫, *buguan bai mao hei mao, hui zhuo laoshu jiu shi hao mao.*
themselves without employment). The transition was overwhelming. These workers, then assured of a job for life (*tiefanwan*) and embedded in the system of protection and social relations of the work unit, were suddenly made responsible for their life-course. They were called upon to look for a job on their own, to change jobs if necessary, and to move to seek new professional opportunities. In 2003, 80% of the working population in China worked in the private sector, whereas 20 years earlier this sector was non-existent (Naughton, 2007).
This institutionalized disembedding of individuals, both from communes and work units, is part of a national project of "societization" which aims to give more autonomy to individuals. The reforms, by "emancipating" the economy and individuals from the tutelage of the socialist state, without a desire to re-embed individuals, had a unique impact on the dynamics of the individualization process in China. They created new opportunities and an increase in income for some, but a degradation of living conditions for others (Whyte, 2010). With the disappearance of popular communes and the dismantling of work units, the welfare state inherited from the Maoist period had been diluted. Individuals had no choice but to internalize the negative externalities of economic liberalization; namely, the liberalization of the labor market, the liberalization of prices in the health, education, housing and food sectors (sectors which were previously taken care of by the collective). Access to these resources has become the responsibility of individuals and highly competitive over the years.
Following these various social, economic, and institutional reforms, individuals have gradually been emancipated, at the social and economic levels but not at the political level, from the constraint of Maoist institutions that controlled and conditioned the stages of their life-course. Therefore, today, the process of individualization in China follows its own timeline and takes a particular form: it is "limited and controlled by the State" (Beck & Beck-Gernsheim, 2010). While maintaining strong control over the political and economic spheres, the Party-State urges individuals to become responsible for the successes and failures in their biographical trajectories. Through rhetorical efforts, individual actions and "do it yourself biographies" are promoted, which has the collateral effect of transferring risks responsibility onto individuals. In this book, individualization is understood as a complex process. Given that in China, the process remains partial, and because social policies and legal provisions produce different levels of individualization depending on the lifecourse's domains. Moreover, the discourse on the quality of the population (*suzhi renkou*) contributes to support individualization in the labor market, while also rejuvenating family obligations and solidarities rather than diluting them. It will be shown in the following chapters that these two dynamics concurrently unfold and interact together.
#### **Bibliography**
Aron, R. (2002). *Le Marxisme de Marx*. Editions de Fallois.
**Open Access** This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.
The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.
# **Chapter 3** *Suzhi* **Discourse as a Structural Component of Institutionalized Individualism in Post-Maoist China**
#### **Contents**
**Abstract** This third chapter sheds light on how the State's discourse on population "quality" (*suzhi*) represents a sophisticated means of governance from afar.
Initially emerging in the 1980s, the concept focused on population control. Eugenic slogans promoting "fewer births, better births" refected a belief that both genetic and environmental factors shape quality. This approach aimed to cultivate a vibrant, competitive youth capable of propelling China onto the global stage.
In the 1990s and 2000s, the discourse gained prominence and underwent a shift. It ultimately encompasses moral, intellectual, psychological, ideological and physical characteristics. Zhang Weiqing's 2007 speech highlights the Party-State's push for an ideal of quality that would produce competent workers and citizens.
The concept is infused through various organizations, including TV dramas. The chapter argues that the State Administration of Radio, Film, and Television controlling their content shown in, TV series act as powerful vehicles for norm dissemination. They actively participate in the socialization of the population. The chapter examine fve TV dramas and how they depict the transition to adulthood through the prism of this discourse.
**Keywords** *Suzhi* discourse · Population "quality" · TV dramas · TV series · Transition to adulthood · Narrative
#### **3.1 Engineering "Quality Citizens"**
At the beginning of the 1980s, the concept *suzhi* was somewhat unheard of in public discourse and often used interchangeably with *zhiliang* (质量). From 1986 and throughout the 1990s, *suzhi* began to impose itself in offcial rhetoric to assess the quality level of the population1 (Fig. 3.1). The concept refers to the intrinsic quality of a person as well as their conduct. It encompasses moral, intellectual, psychological, ideological, and physical qualities2 (Jacka, 2009; Kipnis, 2006; Murphy, 2004).
From then on, relying on propaganda posters, the Party-State has conveyed a message about the "quality" of the population, valuing people with a high level of quality. In other words, people who comply with the norms and values disseminated by the Party-State and the agencies directly and indirectly associated with it.
At the end of 1980s, Chinese offcials no longer only aimed at containing the country's demographic growth, but also at "producing" a population of "quality". Chinese scientists then judged the country's population too numerous, too rural and too uneducated compared to the populations of major world powers. The discourse on *suzhi* was therefore used to transform the country's citizens by acting on both their appearance and their way of thinking to allow the country to become competitive on the international stage (Greenhalgh, 2010). Slogans, such as "fewer births, better births, to invigorate the Chinese Nation" (*shaosheng yousheng zhenxing zhonghua*) appeared (Greenhalgh & Winckler, 2005). At this time, the discourse on the "quality" of the population was eugenic. The word " *yousheng* " (优生) mobilized in the slogans, which means "better births", referred to the idea that genetic characteristics was not the only factor to infuence the quality of human beings.
**Fig. 3.1** Occurrence of *renkou suzhi* and *renkou zhiliang* in Google books written in simplifed Chinese, 1979–2015. (Source: Google books Ngram Viewer)
<sup>1</sup>*Zhiliang* is henceforth used to designate the quality of a thing or an institution.
<sup>2</sup> Jie Sizhong, in a book dedicated to the analysis of the problem of "quality" in contemporary China, identifes eight different categories of " *suzhi* ": personality, spirit, morality, culture, science, health, profession, and Aestheticism (Jie, 1997, cited in Friedman, 2010:234).
The environment in which they grew up and the education they receive were also among the factors not to be overlooked. In other words, acting on these factors could shape individuals who met the needs of the Nation. Such slogans also encouraged healthy couples to give birth and raise a vibrant youth capable of competing with those who have grown up in countries that are among the great economic and political powers (Greenhalgh, 2010).
Then, during the 1990s and 2000s, following the drastic drop in fertility levels, the discourse on the "quality" of the population moved towards a more encompassing notion. The notion of "quantity" is henceforth secondary and the word " *suzhi* " (素质) replaces " *yousheng* " (Greenhalgh, 2011). The idea of "quality" covered by this concept is broad: it includes education, health, patriotism, skills, ethics, civics, and cosmopolitanism. As illustrated at the beginning of 2007 by the speech given by Zhang Weiqing, Minister of the National Commission for Population and Family Planning, all social actors were encouraged to strive towards this ideal that will produce quality workers and citizens (see box below).
*[…] The Decision is a programmatic document guiding population and family planning program in the new era. Its promulgation represents an important measure for implementing the concept of scientifc development and the strategic thinking of building a harmonious socialist society, and marks the entry of China's population and family planning program into a new stage of stabilizing the low fertility level, addressing population issues in a comprehensive way and promoting all-round human development.*
*The Decision adheres to the scientifc development concept as a general guiding principle, regards all-round human development as its central focus and stresses upon comprehensive solution to population issues as the main theme.*
*[…] The plan and the program, as a component of the six special programs for China's national economic and social development, have recently been approved by the State Council for implementation. The plan and the program state the national strategic thinking and objectives for population development and put forward major tasks for population development during the 11th Five-Year Plan Period. (1) Stabilize the current fertility policy, and implement socioeconomic development policies in an integrated way, so as to maintain the total fertility rate at around 1.8 and ensure realization of the quantitative population objective; (2) Upgrade general health of newborn population, comprehensively address unbalanced sex ratio at birth and proactively respond to population ageing; (3) Prioritize development of education and fully develop human resources; (4) Take coordinated development of urban and rural areas and of different geographic regions into overall consideration and guide orderly movement and rational distribution of population; (5) Develop the undertakings for public health, women and children, social welfare, and promote social harmony and equity.*
(continued)
Source: Excerpts from the speech of Zhang Weiqing, Minister of the National Commission for Population and Family Planning, January 23, 20073 .
Zhang's speech reveals that, in the eyes of the Party-State, in order to build a society of small prosperity, many contradictions still need to be resolved. This includes maintaining control of the number of births within households. However, unlike the strict fertility limitation practiced in the 1980s, since 2007 couples where both the husband and wife are only children can have a second child. This provision of the law implies that the problem of excessive fertility is concentrated on the rural population. It is also striking to note that this law, which legislates not only on birth control issues but also on various aspects related to the "quality" of the population, initially only allowed city dwellers to have two children. This law, imbued with the discourse on the "quality" of the population, emerges in the wake of neoliberal ideology. It shows that, in the eyes of the Party-State, it is young urbanites who are seen as the standard-bearers of the national project to produce "quality" individuals who are "self-entrepreneurs" (refexive).
At the end of the 1980s, while the use of this umbrella concept spread within society, its meaning remained vague in the minds of Chinese researchers. At the end of a conference convened to defne this concept, the only consensus found was this: *"suzhi [however defned] is for the most part higher in the city than in the countryside, higher in Han areas than in minority areas, higher in the economically advanced areas than in backward areas. And generally, the quality level of the Chinese population is too low"* (Li, 1988:60). According to Kipnis, the use of the concept *suzhi* at the beginning of the twentieth century stems from the desire of Chinese intellectuals to ensure that efforts to improve the quality of the population were not only aimed at eugenics, but served the national project more broadly by increasing both the level of education and its physical, psychological, and moral qualities (Kipnis, 2006).
Subsequently, the concept was subject to several reappropriations. During the launch of the fve-year plan in 1991, the country's economic backwardness was attributed to the low-quality level of the rural population (Anagnost, 2004). The concept was then used to get the population to adhere to socioeconomic reforms initiated since the mid-1980s, the dissemination of this concept and its reappropriation by the Party-State was probably facilitated by the fact that it existed in the thought of Confucius (Cheng, 1997:67–68).
Since 2010, the concept *suzhi* has been mobilized by a multitude of actors: governmental organization such as family planning, the CCP urging its members and citizens to improve their moral and political level, business leaders who complain about the low level of "quality" of their employees, parents who try to improve the quality level of their children by providing them with the best care, food and
<sup>3</sup>The full version of the speech can be consulted at the following address: http://china.org.cn/enews/news070123-1.htm . Retrieved on January 5, 2017.
education, the urban population who complain about the low level of quality of migrant workers, etc. (Fig. 3.1).
As the interviews I conducted illustrate, *suzhi* not only has a normative character, it also refers to a social value. The defnition proposed by the respondents refers to both the moral value of a person (*daode suzhi*), as well as their physical (*shenti suzhi*), psychological (*xinli suzhi*), or even intellectual qualities. A respondent named Han formulates it this way: *"if the rural population and the urban population do not have the same level of education and the same values, it is because they do not have the same quality level"* (born in 1989, rural, Jilin *hukou* at birth, Bachelor).
The use of this concept allows young people to distance themselves from certain people and assert their difference and their open-mindedness (*kaifang*), which is characteristic of cosmopolitan youth. In other words, its use allows to differentiate and hierarchize individuals: urban and rural population, rural and urban migrants, planned and unplanned person4 , etc. (Anagnost, 2004; Greenhalgh, 2003). The use of the discourse on "quality", mobilized by the Chinese government, while aiming to form autonomous young adults, capable of leading China towards its historical destiny, namely the construction of a rich, powerful and respected Nation on the international stage, tends, at the same time, to essentialize and legitimize the socioeconomic inequalities that are deepening in post-collectivist China5 (Sigley, 2009; Woronov, 2009; Yan, 2003a, b).
#### **3.2** *Suzhi* **Discourse and** *Hukou* **as Instruments of Governmentality**
The construction and mobilization of the discourse on the "quality" of the population by the State and its institutions can be analyzed as a means of governance from afar (Ong & Zhang, 2008). If the Party-State exercised frm control over the individual during the Maoist period, by the end of the twentieth century, its infuence had loosened and become less direct. The discourse on "quality" allowed the Party-State to defne and instill "indirectly" in the public new norms and values in health and education, via a multitude of relays within public administration (laws, public policies, school, etc.), businesses, associations, families and individuals themselves.
Under the Hu Jintao-Wen Jiabao administration, the problem of population growth was stated as a question of social, economic, and human development. The goal was to transition from an economy reliant on cheap labor to a knowledge economy benefting from skills of highly qualifed workers. However, in 2008, offcials
<sup>4</sup> In other words, people born within the legally imposed quotas by birth planning policies and people born outside these quotas.
<sup>5</sup>Between 1980 and 2010, the Gini coeffcient, which measures income disparities, went from 0.25 to 0.50. Chinese society thus went from being one of the least unequal to one of the most unequal in the world during this period (Jacka et al., 2013:220–21).
from the National Commission for Population and Family Planning estimated that the quality of the workforce at the time was detrimental to social development and harmony, effcient resource use, and the country's competitiveness. The solution found to solve this problem was to link birth control policies to the improvement of care and education systems in cities. In other words, in order to assert itself on the international stage, the country needed educated, well-bred, patriotic, highly skilled young adults who are ready to learn throughout their life-course (Greenhalgh & Winckler, 2005:244). State investments to develop human resources, combined with efforts to shape young people ready to become self-entrepreneurs is therefore, a cornerstone on which the Chinese dream rests: regaining the country's past greatness.
Young adults whose level of "quality" is deemed "high" are refexive, meaning they are able to choose their lifestyle, their identity and they are responsible for their life-course. Young adults belonging to the urban middle class6 symbolize the government's success in forming a politically docile youth, but eager to contribute to the ethical and moral elevation of their community, for instance in the name of consumers' rights, social stability, or virtue (Tomba, 2009). The continuous improvement of their individual "quality" level and that of their family becomes both a personal aspiration and a social injunction.
The residence booklet system (*hukou*) is another instrument of governance established by the Party-State. It institutionalizes forms of discrimination and exclusion between different population groups, making it more diffcult for people from the provinces or the countryside to conform to the ideal of "high quality". The *hukou* (*jumin hukou bu*) was introduced by the Party-State in the 1950s7 to control population fows within the country (Cai, 2011; Liu, 2005; Wang, 2005). To this day, it serves individuals to prove their identity, since it records information on the type of *hukou* (collective household8 or family household9 ) and on the different members of the household: the name, date and place of birth, the relationship with the head of household, sex, ethnicity, children under 16 are also mentioned along with the address, paternal grandfather's address, religion, identity card number, education level, marital status, height, blood group, profession and workplace. A copy of this document is kept by the public security offce. Two indications with signifcant
<sup>6</sup>The concept of "middle class" (中产阶级, *zhongchan jieji*) emerged in China in the late 1980s, but it was at the turn of the 2000s that it became established within the Chinese scientifc community. In the early research on the subject, researchers used either the term "middle stratum" (中 间层, *zhongjian ceng*), or "middle income stratum" (中间收入阶层, *zhongjian shouru jieceng*) or "middle income group" (中等收入群体, *zhongdeng shouru qunti*) (C. (李成) Li 2013).
<sup>7</sup> Initial measures were introduced in 1951, but freedom of movement was still in place. Moreover, in June 1955 a directive was signed to introduce a residence booklet system (户籍制度, *huji zhidu*). The law was promulgated in 1958 by the Standing Committee of the National People's Assembly (NPA) (Chan & Li, 1999).
<sup>8</sup>Collective household, *jiti hu.* People living in the dormitories of a company (collective or not), a school, a temple, etc.
<sup>9</sup>Family household, *jiating hu*. People living under the same roof or alone.
implications are also included in the residence booklet: the type of *hukou, hukou leibie* (agricultural, *nongye hukou* or non-agricultural, *feinongye hukou*) 10 and its place of registration, *hukou suozaidi* (Hukou, 1958).
The combination of these two criteria leads to a "complex categorization of individuals", dividing the population in two ways (Froissart, 2008). On the one hand, by distinguishing between rural and urban populations, and assigning them unequal rights; and on the other hand, by establishing a social hierarchy between the local population and the population coming from outside. Individuals are attached to a locality, whether it is rural or urban, and their rights and duties are dependent on the economic and social resources of this locality. The Maoist organization, which favored the economic and social development of major urban centers, resulted in a spatial hierarchy in which villages are located at the bottom of the scale, followed by county capitals, district capitals, municipalities, provincial capitals, autonomous municipalities and Beijing, country's capital at the top. As a result, people at the top of the pyramid receive and beneft comparatively more from social investments than residents of villages, towns, or even small and medium-sized cities.
#### **3.3 Chinese TV Drama as a Vector of Institutionalization**
The discourse on the "quality" of the population is widely present in the media and in TV drama, which are channels for disseminating the norms and values the Party-State considers important. They can be envisaged as a means to govern "from afar" (Ong & Zhang, 2008).
The content of TV drama is very interesting to study from a sociological point of view. Since in China it is the State administration of the press, publication, radio, flm and television (SARFT)11 that controls – from conception to broadcast – the content of television production, TV drama refect norms and values endorsed by the Party-State. Scripts have to be submitted for approval to the SARFT before they can be produced by a television unit registered with the administration. It should be noted, that the fnal script must go before the censorship commission of the SARFT, or its local affliated offce, to receive a license that offcially validates its distribution/production12 (Zhu, 2008; Zhu et al., 2008). The regulations issued by the SARFT produce a powerful instrument of self-censorship for scriptwriters.
<sup>10</sup>Also referred to as *hukou* rural (乡村户口, *xiangcun hukou*) and *hukou* urban (城市户口, *chengshi hukou*).) (Chan & Li, 1999:821–22).
<sup>11</sup>SARFT, State Administration of Press, Publication, Radio, Film and Television of the People's Republic of China, http://www.sarft.gov.cn. Retrieved on 24.03.2023.
<sup>12</sup>According to Zhu, retired offcials who are loyal to the CCP doctrine are sometimes called upon to carry out the political and ideological control of television series after viewing, because this task is very time-consuming.
Especially since broadcast authorization does not necessarily mean that the television series will be broadcast in its entirety. Through the SARFT, the Party-State can at any time decide to suspend the broadcast or even demand modifcations in the script. While television production in China is subject to political infuences, it also responds to market infuences such as fnancial proftability.
During the conduct of the interviews with the respondents, they frequently referred to TV drama to illustrate their accounts, so it felt necessary to me to "take television series seriously" (Laugier, 2022). Especially since they reach a very large audience given that the entire population in China has access to television (directly or indirectly through the Internet). Chinese viewers watch an average of 52 min of TV drama per day (Sartoretti, 2014:37). As producers of norms and values approved by the Party-State, television series can be considered as actively participating in the socialization of the population (Laugier, 2023; Widder, 2004). In this sense, they constitute a relevant window through which to observe the coexistence and mixing of norms and values borrowing from both classical cultural elements (pre-Maoist period), elements dating from the collectivist China (Maoist period) and others revealing a China open to the world (the contemporary period). The reconfguration of these past and present, national and transnational elements, as well as their reappropriation and reinterpretation by the viewers contribute ultimately to the production of shared representations, which spread in daily conversations. Located at the interface between public and private spheres, we will see that there is a very strong convergence between the paths to adulthood valued in the corpus of TV dramas I analyzed and the representations that young Chinese have of a successful path to life.
As part of this research, fve television series were examined in detail: "Beijing Youth" (*beijing qingnian*), "Ants' Struggle" (*yizu de fendou*), "Rules Before Divorce" (*lihun qian guize*), "The Era of Naked Marriage" (*luohun shidai*) and "Struggle" (*fendou*). All were produced in mainland China. Each episode lasts on average 45 min and is available for free online on the Internet in Chinese and subtitled in Chinese. I chose these fve television series in particular because they were often mentioned by the respondents during the conduct of the feld research. The frst three TV dramas had just been released on screens and were very popular at the time. The other two series are a bit older.
"Beijing Youth" was broadcast during the summer of 2012. Through its 36 episodes, the spectator can follow the transition to adulthood of four cousins (on the paternal side) from Beijing: Hedong, Hexi, Henan and Hebei whose social origin, character and aspirations differ. The series "Ants' Struggle", which was frst broadcast in 2012, consists of 33 episodes. It tells the hopes and disappointments of young University graduates who are not originally from Beijing. They studied in this city, but once they graduated, they only found low paid jobs that barely allow them to rent a tiny room together. Their strained fnancial situation makes it diffcult for them to fulfll their desire to get married. "Rules Before Divorce" was also frst broadcast in 2012. The TV drama takes place in Beijing. It features, in over 40 episodes, the marital trajectories of three young adults couples. There are, on the one hand, Wang Mingxuan and Jiang Xinyao who are freshly graduated from University and they defy Mingxuan's parents' opposition to their marriage. On the other hand, Zhao Yatong, a girl from a good family, marries Wen Hao, a young freelance photographer, in a rush. Finally, Li Xin, a second-generation rich young man, gets married with Zhang Xiaofan, whose parents are ordinary blue collars workers, to cover up the birth of their child out of wedlock. Once married, the young couples quarrel daily over trivial matters.
These frst three television series, which had just been released on the screens, were very popular at the time of the conduct of the feld research. The following two series are a bit older, but they were also recommended to me several times by the respondents. "The Era of Naked Marriage" was broadcast in 2011 and consists of 30 episodes. The two main characters have been dating for 10 years and have reached a point in their relationship where they want to get married. However, they do not own a home, they do not have a car, they have no savings, no means to buy an engagement ring or even afford a wedding ceremony. They decide to get married anyway because the young woman is pregnant. Without individual housing, they decide to cohabit under the same roof as the groom's parents. However, they eventually divorce because of the diffcult intergenerational cohabitation.
Last but not least, the TV drama "Struggle" was broadcast in 2007. It takes place in Beijing. The 32 episodes depict the aspirations of young university graduates, as well as their diffculties of living their love stories in a changing society that values personal achievement.
In the fve TV dramas, the transition to adulthood is characterized by four main dimensions: the end of studies, access to a frst stable job, the frst marriage, which is sometimes associated with the birth of a child, and access to individual housing. These dimensions are associated with another central value: cosmopolitanism, which is associated with a level of high "quality". By depicting several paths to adulthood, these TV dramas convey a dual message to young people: it is "normal" to desire and they show what is "normal" to desire in post-collectivist China. The same construction processes are used. Relying on counter-models and symbols, they display the same model of pathways. They function as a place where different norms and values coexist, while producing meaning through their reappropriation and reinterpretation by the audience (Rofel, 2007). They stage young people who share the same desire to become accomplished cosmopolitan urban adults on professional and emotional levels. We see them relentlessly seeking to brave the vicissitudes of life. Life is shown as diffcult, although the moral is always optimistic: with perseverance and patience the future is brighter. Young people ultimately manage to achieve their objectives in terms of professional or family trajectories provided they accept to take charge of their life-course and they are able to make choices for themselves. While this evokes characteristics of the individualization, the TV dramas also remind us that private initiatives, personal expression, the quest for material gain and personal fulfllment must be carried out within the political limits set by the Party-State. In these TV series, the transition to adulthood is completely depoliticized. They never show young people committed to a cause.
The fve TV dramas take place in a cosmopolitan city: Beijing. They stage the city and its infrastructures through a choice of scenic shots that value less its role as a capital, than the image of a cosmopolitan, clean city, open to the globalized world, technologically and economically developed.
The urban environment occupies an important place in each sequence. The city's cosmopolitan character is highlighted through references to the 2008 Olympic Games, which testify to the city's ties with major international powers, with references to the Sanlitun district where embassies, modern shopping centers, and a portion of the foreign population are concentrated, or by shooting scenes of the symbolic buildings of the capital, such as the tower of the national television network CCTV and the immense expressways on which dense traffc circulates and the new guarded residential districts with tall buildings.
Beijing embodies "the" Chinese city *par excellence* in the TV dramas. It portrays the life of middle-class Chinese. Its name appears directly in four of the series. It is only in "Struggle" (*fendou*) that Beijing is referred to indirectly. This narrative creates a myth of the cosmopolitan city, bustling with activities and with modern architecture (Sartoretti, 2014). Beijing embodies dreams, desires, diffculties, and fears of cosmopolitan emerging adults, but not all of them are equally successful on their path to adulthood. The image of the cosmopolitan city is also chosen to stage and promote lifestyles associated with the middle and upper classes. In creating a sense of belonging (Bourdieu, 2007), consumption is presented as a means to achieve social distinction and affrmation of a cosmopolitan identity. Tingting, a young Shanghainese, declared during a discussion in which I asked him to describe his impressions of contemporary China:
*To me, modernity is urbanization. It's also a way of life. Some foreigners think that we still dress in a Chinese way, but we dress in a Western way..*
By staging young adults born in the 1980s–1990s who have a Sino-Western lifestyle or who appreciate it, TV dramas act as a powerful means to convey and promote norms and values associated with a high level of *suzhi*. Cosmopolitanism thus appears as a central concept for understanding the process of individualization in China. The images broadcast show individual and State desires that converge in the formation of a same project: reassert China's power on the international stage by forming high-quality citizens. In other words, citizens with a high level of education, healthy, cosmopolitan, and faithful to ethical and civic values. The government project aims to create a "high quality" workforce, competitive and citizens able to navigate at ease in a globalized world. Therefore, individualization is the consequence of the development strategy carried out by the Party-State.
#### **Bibliography**
**Open Access** This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.
The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.
# **Chapter 4 A Life-Course Perspective on the Individualization Process in Post-collectivist China**
#### **Contents**
**Abstract** After an introduction to the life-course perspective, this fourth chapter gives an overview of the research design. It details the reasons behind the choice of the birth-cohorts under study: the "generation of new China" and the "generation of reforms" that encapsulate the profound transformations from Maoist collectivism to Deng Xiaoping's era of economic reforms.
Furthermore, the chapter reveals how the mixed methods research design was built and it gives insights into the sample composition. The qualitative strand of the research captures variations in the transition to adulthood experienced by 45 individuals born between 1978 and 1993. It unveils lived experiences, challenges, and authentic voices. The quantitative strand reveals the journey to adulthood of 615 young people born between 1980 and 1985, and of 301 people born between 1950 and 1959. The quantitative data are retrospective longitudinal data. They were collected using a life-course matrix (calendar) to provide complete information on individuals' biographical paths and their temporality.
In the initial phase of the research, qualitative and quantitative data were separately analyzed to address specifc dimensions of the research question. The chapter explains that subsequent step involved triangulation, facilitating the integration of insights from both data types.
**Keywords** Biographical analysis · Mixed methods · Life-course matrix · Young adults · China
#### **4.1 The Transition to Adulthood as a Period of Observation**
To understand the forms taken in China by the process of individualization and its impact on the life-course, I have chosen the transition to adulthood as a period of observation. The journey to adulthood is indeed conditioned by a multitude of opportunities and constraints infuenced both by social and institutional structures (legislation, public policies, etc.) and by the socio-economic and political contexts (educational systems, labor market, forms of citizenship, etc.), as well as by cultural systems (emergence of new values, new intergenerational relationships, etc.). The life-course perspective1 allows to examine over a long period of time to what extent structural changes shape the individual trajectories (Mills, 2006).
While Cain was the frst sociologist to propose a defnition of the concept (Cain Jr., 1964), it was not until Riley that the infuence of socioeconomic and political contexts on life-courses was conceptualized. The author also notes that the infuence played by socio-historical changes varies according to birth cohorts and that these cohorts in turn can be producers of social change (Riley, 1979). Neugarten, on the other hand, highlights the infuence of age in determining social behaviors, as each society has specifc expectations regarding the behavior that individuals should adopt according to their age (Neugarten et al., 1965). Following these works, Elder identifed four founding principles of the life-course theory (Elder, 1999; Giele & Elder, 1998): (1) Life-courses are embedded in a given socio-historical context (Location in time and place); (2) Individual lives are embedded in a network of social interdependencies, which can vary according to socio-historical transformations (Linked lives); (3) Individuals have a relative capacity to act on their lifecourse. It should be noted that the extent of their capacity for action depends on the opportunities and social, economic, institutional, and historical constraints to which they are subjected (Human agency); (4) The effects of events and transitions experienced by individuals during their lives vary according to their temporality, sequence, and their alignment with social expectations (Timing of lives). In more recent literature, a ffth principle, which was implicit in Elder's work, is mentioned: (5) Human development and aging are lifelong processes (Marshall & Mueller, 2003). Only longitudinal studies make it possible to analyze the transformation of the life-course over time. They enable us to capture "a pattern of socially defned age-graded events and roles which is subject to historical change in culture and social structure" (Elder, 1999:302). The ability to seize opportunities or face constraints indeed depends on individuals' personal history and their differentiated anchoring in the social structure (Bidart, 2006; Heinz, 2009).
<sup>1</sup>Excellent literature reviews on this concept have been carried out by French-speaking and English-speaking researchers: see notably (Elder et al., 2003; Giele & Elder, 1998; Heinz et al., 2009; Heinz & Marshall, 2003; Kohli, 2009; d'Epinay et al., 2005; Marshall & Mueller, 2003; Mayer, 2001, 2004; Sapin et al., 2007). Contributions from the feld of developmental psychology, which are part of a sociological perspective, are not presented in this book.
It should be mentioned that Chinese sociologists recognize the interest of the theoretical perspective of the life-course as well to explain major transformations undergone in recent decades by China and their impact on individuals' lives (Bao, 2005; Li et al., 1999). Glen Elder's seminal work was translated into Chinese (Elder, 2002) at the same time as research in this perspective began to spread (Bao, 2012; Guo & Chang, 2006; Lin, 2013; Meng & Gregory, 2002; Zhou, 2004; Zhou & Hou, 1999).
Analyzing the transition to adulthood in terms of pathway and transition enable to move away from an essentialist defnition of young people and adults in terms of age class or unifed group. By allowing young people to be conceived as both actors and subjects of their history, it offers a nuanced and dynamic vision of this stage of the life-course (Baudelot & Establet, 2000; Furstenberg et al., 2008; Settersten et al., 2008). These concepts also refer to the temporality and duration of the transitions experienced, which depend not only on individuals' place in the social stratifcation but also on their social characteristics. The path followed can be more or less long and winding. It can change over time, although its initial imprint strongly encourages individuals to follow it (Hogan & Astone, 1986). In studies that fall within the feld of sociology of youth and sociology of ages, fve thresholds are generally retained to analyze the transition to adulthood: the end of studies; leaving the parental home; access to stable employment; cohabitation; and the birth of a frst child (Furstenberg et al., 2008; Galland, 2009; Liefbroer & Toulemon, 2010; Shanahan, 2000). Based on research conducted in the United States, Arnett (2000, 2004) also identifed three individualistic characters, mentioned indiscriminately according to gender: accepting responsibility for oneself; making independent decisions; and becoming fnancially independent.
An increasing number of studies suggest that, since the 1980s, a new model of transition to adulthood has emerged in Western Europe and the United States. This transition has become individualized. The paths to adulthood are both longer and de-standardized2 (Billari & Liefbroer, 2010; Buchmann & Kriesi, 2011; Liefbroer & Toulemon, 2010; Shanahan, 2000). Young people stay in school longer. They delay marriage, childbearing, and have out-of-wedlock births (Brückner & Mayer, 2005; Cherlin, 2004). A weakening of the links between family, school, and professional trajectories has also been observed in the path to adulthood of Western youth. Entry into adult life is becoming more diverse, uncertain, gradual, complex, and less uniform. In other words, the timing and sequence of transitions are less predictable, more prolonged over time, more diversifed, and more disordered (Arnett & Tanner, 2006; Berlin et al., 2010; Booth et al., 1999; Furlong & Cartmel, 2007; Furstenberg, 2010; Galland, 2000; Modell et al., 1976; Settersten & Ray, 2010).
Longer periods of study, later entry to the labor market and more precarious career paths have on average led to an extension of the period of economic dependence of young people on their families or, in the case of Scandinavian countries for instance, on welfare states (Schoeni & Ross, 2008; Van de Velde, 2008).
<sup>2</sup>Standardization refers to the relative uniformity of the timing and sequence of transitions in a population, while destandardization refers to the relative heterogeneity of timing and sequence.
Young adults respond to the uncertainties of neoliberal societies with what researchers call "biographical tinkering", "yo-yo transitions", or "boomerang transitions" consisting of back-and-forths between periods of fnancial independence and periods of precariousness, which sometimes require a return to the parents' home for some time (Cavalli & Galland, 1993; Du Bois-Reymond & Blasco, 2003; Mitchell, 2006). There are multilayered raisons for this: the restructuring of labor markets, the demand from companies for skilled and fexible workers, and the loosening of welfare states are among them. These social and economic transformations affect the ability of Western youth to establish themselves as independent young adults (Furlong & Cartmel, 2007; Furstenberg et al., 2008; Heinz, 2009). While young adults can afford to live more diverse life experiences, this should not overshadow to the determinist role played by social markers, such as gender and social background, in shaping their pathways to adulthood.
Since the 1990s in Central and Eastern Europe, and East Asia as well, the transition to adulthood has also been delayed and become more uncertain (Lesthaeghe, 2010). Since the collapse of the Soviet Union in 1989, in Central and Eastern European countries, marriage and childbearing have been signifcantly postponed, and the proportion of out-of-wedlock births has soared (Kohler et al., 2002; Perelli-Harris, 2008; Thorton & Philipov, 2008). Since the 1990s in East Asia, marriage and childbearing have also been postponed (Jones et al., 2011; Jones & Yeung, 2014). In Asian societies, studies focusing on the transition to adulthood are surprisingly few given that 60% of the world's youth live in the Asia-Pacifc region (Fukuda, 2013; Huang, 2013; Ishida, 2013; Ji, 2013; Nahar et al., 2013; Park, 2013; Utomo et al., 2013; Xenos et al., 2007; Yeung & Alipio, 2013). Research on the transition to adulthood based on Chinese data is, to my knowledge, still relatively rare (Badger et al., 2006; Bao, 2012; Fulda et al., 2019; Hannum & Liu, 2005; Kane & Li, 2021; Li, 2013a, b; Lin, 2013; Nelson & Chen, 2007; Tian, 2016; Yeung & Alipio, 2013; Yeung & Hu, 2013; Zhang, 2004; Zhong & Arnett, 2014). Moreover, the different stages of the transition to adulthood are most often examined independently. While these studies focus on the transitions mobilized by Western researchers, namely: leaving school and entering the labor market, getting married and becoming a parent, they do not analyze the sequence of these events. Except for the work of Kane and Li (2021), leaving the parental home is generally not considered as a constitutive stage of the transition to adulthood.3 Hannum and Liu reveal, as has been observed elsewhere, a postponement of the milestones characterizes the transition to adulthood. Young people tend to complete their studies later, to enter the job market later, to get married later, and to give birth to a frst child later in life. Their work reveals however, that young adults from rural backgrounds tend to enter the labor market more swifty and complete the characteristic stages of the transition to adulthood earlier (Hannum & Liu, 2005). For Yeung and Hu (2013), the transition
<sup>3</sup>However, it should be noted that Chen Feinian, Zhang Qian Forrest, Chai Yanwei and Zhao Zhongwei and Chen Wei have studied separately and from a life-course perspective the evolution of residential trajectories since China's opening to the market economy (Chai, 2009; Chen, 2005; Chen and Korinek, 2010; Zhang, 2004; Zhao and Chen, 2008).
to adulthood in China refects the consequences of the complex interrelation between drastic political interventions, and socio-economic and ideological transformations on the family. Several studies highlight the importance of the family's place in the transition to adulthood. According to these surveys, family responsibilities (marriage, the birth of children within wedlock and flial piety) constitute structuring norms shaping young people's paths to adulthood (Fuligni & Zhang, 2004; Hwang, 1999; Nelson & Chen, 2007; Oyserman et al., 2002). The importance of role transitions given by young people in the defnition of adulthood is sometimes interpreted as indicative of the role played by family obligations in their lives. For instance, the completion of school and access to a frst job are envisioned as means to support the family. Marriage and the birth of a child are envisaged as means to ensure the posterity and well-being of the family group (Badger et al., 2006). Zhong and Arnett (2014) reveal that accepting responsibility for oneself, making independent decisions, and fnancial independence are less substantial attributes of the transition to adulthood for migrant women than family responsibilities towards their spouse, children, and parents. The current state of research, however, does not allow us to know whether these values, which refer to the Confucian value of flial piety and collectivism, are specifc to young migrants. Available scholarships barely examine the variations in the conception of adulthood according to social origin (urban/rural) and gender.
The present study will shed a bright light on this issue. It will also examine whether the transition to adulthood in post-socialist China tends to converge with the dynamics observed in Western societies, and to what extent the paths are specifc; a largely unexplored area in the literature. Because the country has undergone the fastest socio-economic transformations over the past four decades, and is home to the world's largest youth population, China is an important informative case for assessing and testing the regularities in the transformations observed elsewhere. To my knowledge, so far, no study on the subject, which mobilizes the theory of individualization as a theoretical framework, is based on quantitative data. All the research undertaken is based on qualitative data collected in the city or in the countryside. This book proposes to fll this gap. It will examine how individualization manifests itself in the paths to adulthood, using a research design that combines quantitative and qualitative methods.
#### **4.2 A Place and Two Birth Cohorts Emblematic of the People's Republic of China's History: Beijing, the Post-1950s and Post-1980s Generations**
I chose to conduct the feld research in Beijing because this metropolis is not only the political and administrative center of the country, but also its scientifc, cultural, and economic center4 . *Zhongnanhai*, located in the heart of Beijing, a stone's throw
<sup>4</sup>Website of the municipality of Beijing (http://www.beijing.gov.cn. Retrieved on 7.02.23).
from the Forbidden City, is the seat of Chinese power.5 The country's most prestigious universities are located in the capital, which also houses national research centers, such as the Chinese Academy of Social Sciences, or the National Bureau of Statistics. The Chinese capital is also the leading economic center in Northeast China, with the second-highest Gross Domestic Product (GDP) in China (behind Shanghai).6 The tertiary sector is dynamic (76.9%) thanks to fnancial activities, information and technology services, and sales and trading (BMBS, 2022).7 Beijing has the status of a municipality (*zhixiashi*), which places the city directly under the control of the central state (Cabestan, 2014).
The Chinese population is unevenly distributed across the territory. It is concentrated on the coast, where the country's most dynamic economic zones are located (BNS, 2019). Beijing is the third most populous municipality in the country after Chongqing and Shanghai (BNS, 2019). At the time of conducting the feld research, more than 19,000 million people had their residence booklets registered in Beijing for more than 6 months. More than 90% of the inhabitants then lived in urban districts (peripheral and downtown), and nearly 60% of them were established in the city center and hypercenter (BNS, 2012). For this reason, I decided to conduct this research in the eight urban districts that form the center of the city: Dongcheng (including Xuanwu), Xicheng (including Chongwen), Chaoyang, Haidian, Shijingshan, and Fengtai (see appendix). Despite the extent of the territory, this choice was made because these eight districts, which are located inside the fve peripheral roads that surround the metropolis, are very well served, both by car, taxi, and by public transports (subways and bus). It takes about an hour with the subway to reach Fengtai from Haidian. This mobility was further encouraged by the low cost of transport tickets: a bus ticket then cost 0.40 *renminbi* (0.06 CHF) or 1 yuan (0.15 CHF) for a person without a transport card, while a metro ticket cost 2 *renminbi* (CHF). The decision to conduct the research across all these districts, which form the city center, was also motivated by my desire to diversify the profle of the respondents as much as possible, beyond the stratifcation criteria defned for the constitution of the sample.
During my research in the Chinese capital, young adults born between 1978 and 1993 were distributed as follows: About 60% did not have a Beijing *hukou*, and among them, nearly 50% had an urban residence booklet. About 40% had a Beijing *hukou*, and among them, 9% had an agricultural residence booklet (BNS, 2012). Producing and asserting statuses into statistical registers refects the symbolic power held by the Party-State. General data indeed conceal a multitude of situations, which are the product of social class, gender and the residence booklet system. Statistical
<sup>5</sup>For a comprehensive presentation of the organization of the Chinese political system, see notably the works of J.P. Cabestan (Cabestan, 1994, 2014).
<sup>6</sup> http://www.china.org.cn/business/2023-02/16/content\_85110076.htm. Retrieved on 7.02.23.
<sup>7</sup>The composition of Beijing's GNP is as follows: primary sector, 0.3%; secondary sector, 18%; and tertiary sector, 81.7%, https://nj.tjj.beijing.gov.cn/nj/main/2022-tjnj/zk/indexeh.htm. Retrieved on 7.02.2023.
records distinguish four population groups8 : frstly, the population with an urban *hukou* from Beijing; secondly, the population with a rural *hukou* from Beijing; thirdly, the population with an urban *hukou* but who is not registered in Beijing; fourthly, the population with a rural *hukou* registered in a town or village outside the municipality of Beijing *.* Within these groups, gender and social class constitute invisible sub-demarcations. One should add a ffth population group to these four groups: those who migrate informally to Beijing to work and live, without having obtained a temporary or permanent residence permit.
Two birth cohorts perfectly embody all the economic, social, and institutional changes that have shaken China over the past seven decades. These are, on the one hand, the "generation of the new China (*xin zhongguo de yidai*)", born between 1950 and 1959, in the aftermath of the foundation of the PRC. This birth cohort grew up and experienced its transition to adult life in Maoist China. And on the other hand, their children, the "generation of reforms (*zhuanxing de yidai)* ", born between 1980 and 1989 and whose members – today's adults – have lived their entire lives in the era of reforms initiated by Deng Xiaoping.
During the last decades, the socio-economic transformations have drastically undermined the institutional organization set up under Mao Zedong (1949–1976). The country's opening to neoliberal economic ideology concurrent with the maintenance of a single political party in power has given rise to new institutional arrangements. These transformations have had a differentiated impact on the population's life-course depending on the position of individuals in social structures and their affliations (Bian & Logan, 1996; Wu & Xie, 2003; Xie & Hannum, 1996; Zhou, 2004; Zhou et al., 1997). Young adults have been confronted with new opportunities during their transition to adulthood. Self-realization, personal fulfllment, and the development of individual capacities, which were not encouraged during the collectivist period, are now valued. Since the 1980s, there are also more opportunities to pursue higher education and in terms of employment in various geographical locations (Connelly & Zheng, 2007). Young adults no longer need to wait for the State to assign them a job. They can directly apply for the desired job in the labor market.
Over half (57%) of China's population lives in the countryside, and less than 20% of Chinese attend university after high school (BNS, 2012). Young adults born in rural areas who, in the past, had almost no opportunity to settle in urban areas, can now migrate to cities with fewer constraints (Wang, 2005). The downside is that lifetime jobs have given way to precarious jobs (Constantin, 2016, 2018). Many employment-related benefts such as affordable housing, cheap medical care, generous retirement pensions, and childcare subsidies have disappeared (Lü & Perry, 1997). In today's China, young adults must carefully plan each transition over their life-course. For example, if housing is an essential prerequisite for family
<sup>8</sup>The statistical registers of the population censuses distinguish, in addition to the four aforementioned groups, people who have their *hukou* registered in Beijing but who have gone to study or work abroad for a defned duration, and *hukou* agricultural (农业户口, *nongye hukou*) from *hukou* non-agricultural (非农业户口, *fei nongye hukou*).
formation, the rising house prices do not always allow young adults and their families to own their own home (Constantin, 2021). Especially as social and economic inequalities have widened since the country opening-up to market economy (Zhou & Moen, 2001).
#### **4.3 A Mixed Methods Research Design to Reconstruct the Transition to Adulthood**
To capture the experience of the transition to adulthood objectively and subjectively in the face of the transition to market economy, I developed a survey design based on mixed methods. Qualitative data are used to interpret and contextualize quantitative fndings. Data collection was part of my doctoral thesis. The feld research took place in Beijing between 2012 and 2014. The interviews were conducted in Chinese (Mandarin).
The quantitative data are retrospective longitudinal data. They aim to reconstruct the respondents' paths to adulthood. They were collected using a life calendar, which I developed on a computer. This methodological tool not only provides complete information on individuals' biographical paths but also reconstructs their temporality (Belli et al., 2009; Lelièvre, 2005). The life calendar associates temporality, events, and domains of the life-course. They function as cues to help respondents structure their autobiographical memory and to support them in recalling their past. The interviews carried out for the collection of quantitative data were organized around a comprehensive discussion. This extensive exchange had the advantage of reducing the risk of obtaining responses intended to "satisfy" the researcher. During the interviews, conducted with the help of about 20 students, the life calendar was displayed on our computer screens. The respondents could thus visualize at the same time the questions concerning the different areas of their life-course (professional, educational, family, etc.) and the temporality (years and age). Calendar years were used as a unit. With life calendars I not only collected quantitative data (N = 916), but I also used them as a support for conducting semi-structured interviews (N = 45). The convenience sample was stratifed by age, sex, place of origin, *hukou* (urban/rural), and educational achievement. For each stratifcation criterion, based on the 2010 Beijing's population census, I computed quotas. The quantitative data reveal the journey to adulthood of 615 young people born between 1980 and 1985 (F = 290 and M = 325), and of 301 people born between 1950 and 1959 (F = 151 and M = 150).9 At the beginning of the data collection, I accompanied each of the students and, throughout the process, we held meetings every week to check the quality of the data collected. These sessions also served as a forum for exchanging best survey practices.
<sup>9</sup>For reasons of survey feasibility, for the cohort born between 1950 and 1959, interviews were conducted only among the urban population.
Regarding the qualitative part of the research, I collected the data alone. I tried to submerge myself as much as possible in the daily life as it is experienced by young adults by placing myself in a situation of prolonged interaction (close to the respondents and in their environment). I stayed for over a year in Beijing and lived for several months in cohabitation with some of the respondents. I also conducted indepth interviews to bring out (until theoretical saturation was reached) a variation of positions and points of view. The interviews, which were anonymized, were conducted in different neighborhoods of Beijing. I interviewed 45 urban and rural young people born between 1978 and 1993, and aged 19 to 36 years old (F = 25 and M = 20) at the time of the survey. They had been living in Beijing for at least 6 months at the time of the survey. Each interview began with a discussion about the transition to adulthood, with the life calendar as support. This frst part of the interview lasted on average 2 h. Then, we agreed on a second meeting to discuss a specifc topic defned in the interview canvas (becoming an adult, marriage, family, employment, values, social pressure, and consumption). This second part, depending on the respondent's interest in the research, could end after one meeting or continue, with regular meetings over a year. Twenty respondents agreed with the latter option. I found it more diffcult to establish contact with rural men. Therefore, the content of these interviews is more superfcial. The combination of quantitative and qualitative methods allowed me to grasp the individual and social representations attached to the thresholds of the transition to adulthood by testing the different categorizations.
#### *4.3.1 Quantitative Data Collection*
Thanks to the fnancial support of the Ernst and Lucie Schmidheiny Foundation and the Geneva-Asia Association, I was able to form a team of investigators to help me collect the quantitative data. Once in the feld, the frst challenge was to fnd the right balance in the number of investigators to recruit. It was necessary not only to consider the fact that a signifcant number of investigators would stop their mission during the research, but also to respond to the risk of interaction bias between the investigator and the respondent. Indeed, the responses of the latter can be biased by their perception of the expectations of the investigator or, unconsciously, the investigator can infuence to some extent the responses by the way he/she formulates the questions. Therefore, the more interviews conducted by an investigator, the greater the risk of bias. To reduce the likelihood of this risk, it appears preferable to recruit a large number of investigators and have them each conduct a small number of interviews. However, this poses the problem of recruiting investigators trained in quantitative data collection. The number of investigators I ultimately recruited resulted from the arbitration between these specifc constraints.
I posted, with the help of Chinese colleagues, a small ad on the Internet BBS sites of Peking University, People's University, and Beijing Normal University. In the ad, I did not mention the remuneration, as I wanted to make sure that I would recruit social science students who were motivated by the project. To motivate the investigators, I had planned a relatively high remuneration: 50 RMB (or 8 CHF) per valid interview.10 Indeed, the work of reconstructing biographical paths is diffcult: the interviews are long, lasting between 1 h and 1 h 30, and it is not easy to fnd respondents who meet the sampling criteria and have the time and motivation to participate in the survey. Initially, I made appointments with all the people who had applied. Among these, some did not show up. I then met a total of 35. In a second step, I invited 23 of them to participate in a two-day training. Again, some did not show up, while others after the frst half-day came to the conclusion that the work did not suit them and withdrew. At the end of the training, the team consisted of 5 male investigators and 15 female investigators including myself. Half of them were students from Peking University, the others were enrolled at People's University, only one person came from Beijing Normal University.
As pointed out in the scientifc literature, I very quickly encountered a problem of attrition in the team of interviewers. During the frst semester, two people dropped out and I had to ask two others to stop due to signifcant methodological problems in conducting the surveys that led me to invalidate and reject all their interviews. To strengthen the team, I decided to organize a second recruitment campaign after the Chinese New Year holidays. Following this campaign, I recruited four people (among the 15 people met). I again organized a two-day training for these three new female investigators and one new male investigator. At the end of the training, these people worked for 2 weeks in pairs with me. An investigator, who was among the best in the team, supported me by playing this role with a new investigator as well. But very quickly, 2 weeks after the training, one of the new female interviewers dropped out. Then, during the semester fve people gradually stopped. *In fne* 16 people collected the quantitative data during the spring. Then, as summer approached, we were 13 people to complete the data collection. Aware from the start of the risk of attrition, I continuously sought to motivate the team through encouragement during follow-up meetings that took place every 2 weeks, or by inviting them in the "best" campus restaurant on the eve of the Chinese New Year holidays and before everyone went back to their family for a month to celebrate.
To gain the trust of the respondents and encourage them to participate in the survey, I gave it a name: Beijing Youth (*beijing qingnian*). On the advice of the investigators, I printed a small brochure. It presented the research and was distributed to each respondent when we contacted them. At this occasion we also stated our identity and affliations. Then, we explained that we were conducting a sociological survey (*shehui diaocha*), which has a positive connotation in China. It is associated with the academic environment and State efforts to reform the country. Then, we briefy presented the objectives of the research as well as the type of interview we wished to conduct and its duration. If the person was not immediately available, we agreed on an appointment to meet them later in a quiet place, without
<sup>10</sup> in Beijing, the average wage was 6688 RMB per month in 2015.
any third party present who could negatively infer the quality of the information gathered during the interview.
Because the supervision of investigators and the control of data being two key variables for collecting reliable data, I developed a strict monitoring and control system. During the frst 2 weeks, I individually accompanied each investigator in the feld to help them get a good handle on the life calendar device. It was also an opportunity to advise them on their self-introduction to respondents, as well as to provide technical support. Following this, I organized weekly follow-up meetings. The investigators were invited to send me in advance all the life calendars collected, so that I could check them and discuss with them the diffculties encountered, as well as the recurring errors identifed during the verifcation of the matrices. These sessions also aimed to serve as a place for exchange and advice on best practices to adopt. At the end of each session, I individually met with the investigators to discuss the inconsistencies found in their life calendars. After 3 months of feldwork, I realized that all the investigators had mastered the matrix. Consequently, I changed the format of the meetings to one-on-one to discuss the life calendars individually. In situations where minor errors were identifed, we could correct them by getting back in touch with the respondents since we asked for their contact details in the frst part of the interview. However, in cases where we could not recontact the respondents to verify certain inconsistencies or complete missing data, the matrices were considered as invalid and, therefore, rejected (Appendix 7).
To track the progress of the feldwork and the work of each investigator, I developed two Excel fles: a general tracking fle and an investigator-specifc tracking fle. The frst fle allowed me to visualize using graphical representations, on the one hand, the overall progress of data collection according to the different stratifcation criteria of the sample and, on the other hand, to compare the number of interviews collected by the investigators by looking more closely at how many life calendars each one collected for each of the stratifcation criteria of the sample. The second tracking fle was designed to individually follow the investigators. For each meeting and for each person, I printed the graphical illustration of the summary of their work, as well as a sheet compiling all of my remarks on the matrices. This monitoring work allowed me to ensure the proper distribution of the sample during the data collection process, by continuously reevaluating the number of matrices to collect for each stratifcation criterion. Despite this methodological caution, the fnal sample shows a slight underrepresentation of respondents born between 1980 and 1985 with a middle school education level (*chuzhong*). They represent 27% of the sample while according to calculations based on the population census, this category should have reached 35%. Perhaps the reasons for this underrepresentation are to be found in the social distance between the young university investigators and the respondents. Or in the form of symbolic violence that any survey carries. Faced with students enrolled in the country's top universities, respondents may have felt a "symbolic loss", in the sense that they may have felt reminded of their own diffculties and disappointments. This explanation seems plausible since the fnal sample shows an overrepresentation of people with a master's degree or higher. Their share is 13%, while the quota to reach was established at 7%, according to the population
census. This is a recurring problem regardless of the respondents' gender, even though the breakdown of the sample according to sex is in line with the quotas calculated based on this census. Finally, while it was diffcult at the beginning to fnd within this birth cohort and in the eight districts that form the downtown of Beijing respondents with a Beijing but rural (agricultural) residence booklet, in Spring 2013 I organized a daytrip in the district of Miyun. This district, located in the outskirt of the city, is mainly rural. 65% of the population has a Beijing agricultural residence booklet. Thanks to this expedition, at the end of the day, we managed to meet our data collection targets. While the total ratio of women and men in the sample meet the quotas, it should be noted that regardless of sex, the sample shows an underrepresentation of two point of percentage in the number of men and women with an urban *hukou* registered in the capital. We indeed had a lot of diffculties in meeting native Beijingers. There is also a three-percentage point overrepresentation of the number of people with a rural residence booklet registered outside of Beijing.
Regarding the cohort born in the 1950s, it was diffcult to meet women with a primary school level of education. They constitute 14% of the respondents while they should have represented 18% of the women of this cohort. The sample also shows an underrepresentation of men with a middle school level of education (47% instead of 49%). For the high school level of education, there is an overrepresentation of women since 24.5% of the women of this birth cohort have this level of education in the sample, while according to the Beijing population census their share should have been 19%. Women with a specialized university level of education (short-cycle tertiary education) are also underrepresented by two percentage points compared to data from the national statistics bureau. While men who have reached this level of education are overrepresented by three points. Finally, the collected sample shows a slight underrepresentation of men who have completed a bachelor's degree, and an overrepresentation of men and women who have reached a level of education higher than bachelor. Apart from these points of vigilance, it should be emphasized that the distribution of the sample according to the residence booklet and sex is excellent. The sample is fairly representative of the two populations. Nevertheless, the inferences cannot be generalized to the entire population belonging to the two birth cohorts. The mode of recruitment of the respondents may have induced a bias in the constitution of the sample.
#### *4.3.2 Qualitative Data Collection*
Since the early 1980s, the People's Republic of China has widely opened its doors to social science researchers. Access to the Chinese feld has a signifcant entry cost, as it requires prior learning of the Chinese language and knowledge of Chinese culture. I negotiated my entry into the feld by seeking an institutional affliation within Peking University. Throughout the duration of the feld survey, I was a student at this University. From June to August 2012, I took intensive Chinese courses to refresh my oral language skills. Then, from September 2012 to August 2013, I was an invited doctoral researcher within the department of sociology. In addition to the interviews conducted with young adults, this prolonged exposure in the feld allowed me to grasp a relative authenticity in the behavior of young adults and thus to collect relatively spontaneous analysis material (Quivy & Van Campenhoudt, 1995). On campus, I tried to obtain authorization to live in the same building as the Chinese doctoral students in order to get as close as possible to the daily life of my peers and to interact constantly with them, but the housing offcials refused my request arguing that I was better housed in the building reserved for foreign students. While I had no diffculty being accepted by the students and professors, I felt a certain frustration being referred by the school authorities to my foreign status and segregated in terms of housing. The main problem encountered during participant observation was that of taking notes on the spot in my feld journal. During classes, I could discreetly complete it, but during daily interactions or during moments of confdence, note-taking was not desirable because it would have ended these privileged moments. It would have had the effect of changing my status from comrade to researcher and from then on, I would have taken the risk that my pairs, feeling observed, would change their behavior. The diffculty for me was to immerse myself in these conversations and then transcribe their substance in my notebook out of sight of my comrades.
In parallel with participant observation, I practiced indirect observation by seizing every opportunity to walk around the city or participate in activities (visits with friends, discovery of religious practices on invitation, etc.). I was thus able to easily expand my network of acquaintances, outside of the university community. Moreover, apart from those who worked in the service sector, it was more diffcult for me to socialize with young men of rural origin who resided in Beijing.
Regarding the conduct of the interviews, I mobilized two strategies to capture the representations and points of view of the respondents: semi-structured interviews and the constitution of a focus group. For the semi-structured interviews, I favored direct access to the respondents by approaching people and explaining my research to them. For the constitution of the focus group, I contacted several professors from the French department at Peking University via email offering their students free tandems. Only one professor responded to my offer. He invited me to his class to introduce me and present my project. Following this, several students contacted me, including fve who were diligent throughout the year. These are the people born in 1993 in the sample. Given that all the participants had a good level of French, the support courses could take the form of informal exchanges around the interview canvas themes to start the discussion. At the end of the session, I would go over grammar or vocabulary points with them. The group was relatively homogeneous socially, although more frequented by young men. The exchanges, which took place every 2 weeks, lasted on average 2 h 30.
I chose to conduct all the interviews in public, neutral spaces, such as cafes. Creating a convivial moment around a cup of coffee or tea helped to reduce the social distance between me and the respondents. All the interviews were recorded using the Audacity software. As a safeguard, I also took notes in a feld notebook. This precaution proved particularly useful because on two occasions the sound was
cut off during the recording. Thanks to the notes taken, when transcribing them into the MAXQDA software I was able to complete them. The public space chosen to conduct the interviews had the merit of approaching a situation of ordinary interaction, namely conversation, and thus it helped the respondents forget the presence of the recording.
While estimating an optimal sample size is possible for the collection of quantitative data, conversely, there are no strict sampling rules for qualitative research. At the beginning of the survey each new interview brings new information that contributes to the development of models, which become clearer and more stable as the research progresses. When saturation is reached, the latest information collected almost no longer teaches anything (Glaser & Strauss, 1967). As "no category of actors holds all the objective knowledge alone, but that each one's vision contains its share of truth" (Bertaux, 2013:27), I sought respondents with diverse positions and points of view to gradually build the sample and avoid the monographic compartmentalization of a certain type of population. To obtain a unit of analysis, I sought to avoid extreme cases in the construction of the sample. To balance the different sampling criteria and contrast individuals and situations as much as possible according to these variables, I chose to conduct the interviews in various districts of the city: in Haidian, which is located in the northwest; in Dongcheng, which is in the hypercenter; and in Chaoyang, which is located in the eastern part of the capital. The interviews were conducted in three languages according to the wishes of the respondents: in Chinese, French, and English-Chinese. Some interlocutors agreed to participate in the survey for a year by meeting me every 2 weeks, because they saw it as a way to practice a foreign language. Although I would have preferred to conduct all the interviews in Chinese, I was not able to impose this. The risk of losing the enthusiasm of the participants was too great. In situations where the respondents preferred to practice a foreign language, they made great efforts to answer the questions in detail.
I also felt that my status as a foreigner was an advantage, in the sense that some respondents felt that they could easily confde in me since I was outside their network of relationships, and they knew that eventually I would return to Switzerland. The guarantee of preserving their anonymity and our frequent meetings allowed me to collect real confdences from some of them; sometimes even asking me for advice. In the various interactions with the interviewees, belonging to the same birth cohort as I, I often found myself in the position of a sympathetic friend. This attitude was reinforced by the fact that the topics discussed were, from the respondents' point of view, real daily concerns.
#### *4.3.3 Data Analysis*
In a frst step, quantitative and qualitative data were analyzed separately to answer each of the dimensions of the research question to which they related. This separation between quantitative and qualitative analyses is obviously not "pure". Given that the different analyses were made by the same person, the results obtained on one side infuenced my reading of the results obtained on the other side. This process is referred to as "cross-talk" between the different phases of the analysis (Teddlie & Tashakkori, 2009). In a second step, the triangulation of the results allowed me to make inferences. The iterative process of triangulation between statistical and thematic analyses enabled me, on the one hand, to identify trends and recurrences in the quantitative and qualitative data, and on the other hand, to make links between the different levels of analysis.
Regarding the quantitative strand of the research, the analyses rely on biographical analysis. To examine the sequence of transitions over the life-course and their variations, I used the R software and the package TraMineR. These longitudinal analyses consider as a unit of analysis the entire sequences of successive states occupied by the individuals during a given period. It makes it possible to highlight continuities and changes over the life-course in one specifc trajectory, or in several at once. This latter method, which allows to analyze several trajectories simultaneously, is called "multichannel sequence analysis (MCSA)" (Gauthier et al., 2010). From these analyses, I sought to obtain a typology using Optimal Matching (Gabadinho et al., 2011). I also constructed logistic regression models on SPSS in order to measure the association between the occurrence of an event and the factors likely to explain it. Using a "generalized linear model" I also sought to estimate the risk of occurrence of a given event by evaluating the effect of explanatory variables on the transitions examined (Mills, 2011).11
In general, the quality of inferences is determined by the validity of the results produced. Statistical tests allow to verify the validity of the models, or if the relationships found between dependent and independent variables are reliable. The intensity of the conclusions produced is then proportional to the level of signifcance of the results found in the statistical analyses. Regarding the qualitative strand of the research, the quality of inferences depends on the position of the researcher and the interviewees at the time of conducting the interviews. The position of the researcher is not neutral. Hence extensive and detailed descriptions of the context and the conduct of the research enable the researcher to adopt a refexive stance in relation to the process of data collection, their analysis, and the production of fndings. It is a question of distancing oneself, of situating the respondents' answers in relation to the context, their social condition, and their specifc situation with regard to the issues addressed (Fielding, 2008; Schatzmann & Strauss, 1973). Regarding focus groups and observation, the relative positions of the participants need to be taken into account in order to identify potential biases in the respondents' answers and in the researcher's preliminary interpretations (Bertaux, 2013).
To analyze the corpus of qualitative data, I looked for a thematic coherence between the interviews by transcribing and coding the segments that referred to shared topics. I relied on thematic analysis to reveal the respondents' practices and representations. Based on the initial hypotheses and the data, I identifed the themes
<sup>11</sup> I warmly thank Jean-Marie Le Goff for guiding and advising me in the construction of this model.
and built the analytical framework. This is an iterative process between the hypotheses and the corpus (Blanchet & Gotman, 2007). The MAXQDA software enables to hierarchize the analysis grid into main themes and secondary themes to form units of meaning. This has the advantage of decomposing the information as much as possible, separating the factual elements and the elements of meaning, and thus minimizing uncontrolled interpretations. During this process, I made sure that the division into themes did not change the meaning of the isolated interview excerpts. Seeking to develop the knowledge as close as possible to reality, I supplemented this corpus with secondary data, such as statistics, reports, newspapers, literature, broadcasts, and TV series related to my subject of study. Although interviews are the only way of accessing people's experiences and point of view, secondary data can be used to contextualize them and relate them to facts. In particular, the analysis of TV dramas enabled me to bring out norms and values related to the transition to adulthood conveyed within society, with a particular attention paid to representations of femininity and masculinitiy. Therefore, the whole corpus of data (primary and secondary, qualitative, and quantitative) enabled me to look in different directions: from the inside, from the outside, from above, from below and from the side.
#### **Bibliography**
Baudelot, C., & Establet, R. (2000). *Avoir 30 ans en 1968 et en 1998*. Seuil.
BNS, (National Bureau of Statistics). (2019). *China statistical yearbook*. China Statistics Press.
W. McNeish, & A. Walther (Eds.), *Young people and contradictions of inclusion* (Towards integrated transition policies in Europe). Policy Press.
**Open Access** This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.
The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.
# **Part II Coming of Age in Uncertain Times**
By the end of the 1970s, it had become clear to China's leaders that acting on the quality of its population–by seeking to raise it—constituted a way for the country to regain its former greatness (Greenhalgh, 2010). This project involved the implementation of population policies. These public policies, located at the intersection between biology, medical and social sciences, are eminently political. They are used by the Party-State to administer and govern the population. Among other things, they constitute a means to control migration. For instance, through the residence booklet system which enables the State to control people's mobility within the country. Population policies are also useful to control demographic growth. Acting either, for example, on the mortality rate through investment in the health sector, or on the fertility rate by limiting the number of authorized births.
Since the mid-1950s, and especially since the late 1970s, the population has become a central object of governance. The Party-State then dream of transforming the country into a prosperous and modern Nation that enjoys a strong footing on the international stage. In this context, families who invest, both emotionally and materially, in the development of their children, are socially valued. This discourse about "quality" especially resonates with families where the closure of Universities during the Cultural Revolution and the sending-down movement to the countryside of urban youth1 meant that it was not possible to pursue higher education. Among them, some project their social ascent through their children, or rather that of the family group (Chicharro, 2010; Fong, 2004). This collective behavior is further reinforced through the publication of successful novels, such as "Harvard Girl – Liu Ying—Documentary on quality training" published in 2001 and reissued in 2009 and 2014 (Liu & Zhang, 2001), which illustrate how the ferce competition within the educational system and on the labour market leads individuals to maneuver to stand out from each other and to give their children the ability to stand out. Tingting's parents subscribed to this logic. From the time their only son was a child, they
<sup>1</sup> In the People's Republic of China (PRC), between 1966 and 1978, during the ultural revolution, about 17 million young Chinese who had completed their secondary education had to leave their hometowns to work in the countryside (Bonnin, 2004).
sought to maximize his chances of getting into one of the country's top universities. Tingting was raised in Shanghai in a family belonging to the upper-middle class (born in 1993, urban, Shanghai *hukou*, Bachelor). Tingting's journey is emblematic of the young elites. At the time I was conducting the interviews, his father was a sales manager and his mother a project manager. His parents saved money all their life so that their only son could go to a prestigious University. The family project was for Tingting to enroll in the best high school in the city in order to maximize his chances of obtaining excellent results in the University entrance exam (*gaokao*) and thus be accepted at the University of his choice. Once admitted to one of the best Universities in the country, his parents continued to invest in their son's educational path by fnancing his university fees (5000 RMB per year), his accommodation (1000 RMB per year) and his pocket money (between 1000 and 1500 RMB per month). In February 2013, during the Chinese New Year holidays and his third year at university (bachelor degree), he participated in the GIMUN (Geneva International Model United Nations) in New York. His parents fnanced the trip (40,000 RMB). They did it again in 2013–2014, when he went to study a semester at Sciences-Po Paris. Because of his excellent academic record, he was awarded a scholarship to study abroad. While it covered his university fees, his parents had to fnance his accommodation, as well as his daily expenses in Paris. Tingting, like all the young adults who participated in this research, believes that "*Nowadays to enjoy life, one must not only work hard but also study hard*". According to the respondents, there is a royal road to success ("*chenggong lu*") that goes through a good education. This is not surprising since according to the values disseminated the level of "quality" of a person is partly correlated to his academic level. Yet this rational conduct induces a collective irrationality. It encourages more and more youth to graduate from university, and that in turn leads to a progressive devaluation of the value of diplomas on the labor market (Liu, 2011).
Tingting is aware of the important role played by "interpersonal relationships" (*guanxi*) to achieve his goals and build a career. He admits that "*those around me are elites. I prefer to be part of elite groups"*. One of the strategies implemented by some young adults to increase their chances of professional success is joining the Chinese Communist Party (CCP): 20% of the respondents born between 1980 and 1985, regardless of gender, made this choice. Data shows that it is primarily young adults with a high level of education who seek to become members of the CCP. Among all the young adults met (data corpus combining both quantitative and qualitative), over 60% of the respondents who have obtained a doctorate or a master's degree are members of the CCP. Whereas less than 10% of the respondents with an education level equal to or lower than high school are members of the CCP. Beyond the numbers, it should be kept in mind that it is primarily the students from the country's best universities who become members of the CCP (Rosen, 2004). This trend refects the CCP's desire to focus its recruitment efforts on young intellectuals. This move towards the formation of a "technocratic elite" marks a turning point from the Maoist period, during which workers and peasants were favored by the Party-State (Andreas, 2009; Li & Walder, 2001). Tingting explains that being a member of the CCP: *"has advantages for fnding a civil service job or* *for working in a state-owned company. You have more chances of being promoted too. If you want to continue in the system, it's necessary"*. He thinks that "*it is out of pragmatism that students decide to join the CCP: to fnd a job, to be promoted, to be elected as a cadre*", because "*to secure a promising career it is better to be a member of the CCP* "(born in 1993, urban, Shanghai *hukou*, Bachelor). Moreover, from the employers' point of view, the fact that someone is a member of the CCP may indicate that this person is disciplined and will not seek to disturb public order (Hsu, 2007).
The respondents are very clear-sighted. Whether they are urban or rural, from Beijing or not, men or women, they all identify the *hukou* as a source of discrimination. Several have clearly stated that the *hukou* system "*makes no sense*", that it is "*not fair*" and that it "*should be abolished*". This chapter shows, on the one hand, how young adults from rural and urban backgrounds transition to adulthood in an uncertain society exposed to market values and competition. On the other hand, it sheds light on the strategies they deploy to cope with the many diffculties of integrating into a job market that is increasingly precarious, competitive, and characterized by discrimination, institutionalized among other things by the residence booklet system (*hukou*).
#### **Bibliography**
# **Chapter 5 Exploring Pathways to Adulthood**
#### **Contents**
**Abstract** Through examining the life-course of the cohorts born post-1950s and post-1980s, this ffth chapter reveals a substantial extension in the duration of education for the latter group, impacting subsequent life events.
The chapter also discusses the signifcance of family roles and responsibilities in the transition to adulthood in China, emphasizing the central role of marriage in this process. It explores how Confucian thought, government policies, and societal expectations contribute to a normative approach to marriage and parenthood. In particular, it sheds light on President Xi Jinping's discourses that promote family values as essential for national development and social harmony.
Moreover, the analyses delve into the portrayal of marriage and housing in Chinese TV series. Refecting societal norms and expectations, they reveal and discuss the pressure on women to be married at a certain age, by calling those, who are not, "leftover women". TV drama can be understood as a window on the tensions existing between traditional values and changing gender roles. Last but not least, the fndings highlight the challenges young adults face in achieving homeownership, and the meaning they give to this transition, as a symbol of stability and fnancial autonomy, in their pathway to adulthood.
**Keywords** Transition to adulthood · Longer · Roles · Responsibilities · Leftover women · Gender
#### **5.1 A Longer Transition to Adulthood in Post-collectivist China**
The inter-cohort analyses of the stages of the transition to adulthood reveal the emergence of a new social time – higher education – in the life-course of young adults (Fig. 6.1). Urban youth study on average 3 years longer than their elders: the post-1980s fnish their education on average around 21 years old, while the post-1950s fnish it on average around 18 years old. At 22 years old, 80% of respondents born between 1950 and 1959 had fnished their studies; whereas at the same age only 65% of urban respondents born between 1980 and 1985 have completed their schooling.
One should remember that most universities were closed during the cultural revolution (1966–1972). During this period many students and professors were sent-down to the countryside to either work on farms, monitor the borders, or teach (Bonnin, 2004). Among this birth cohort, urban young adults who were able to stay in the city were mainly assigned to a state or collective enterprise to work.
From 1972, universities gradually reopened their doors (Bernstein, 1977; Unger, 1982). The university entrance exam was reinstated in 1977. As part of the "four modernizations" championed by Deng Xiaoping, schooling and diploma regained a foremost place. In 1986, the Party-State declared 9 years of primary and secondary education compulsory for all citizens. At the same time, numerous technical schools opened (Zhou et al., 1998). These changes were followed in 1999 by a policy in favor of the development of higher education (Li, 2013a, 2013b). China's entry into the World Trade Organization (WTO) in 2001 further strengthened public policies in favor of equality in access to education. This movement is refected in the publication of laws and directives in favor of the expansion of higher education (Ren & Zhu, 2014).
The analyses carried out indicate that these policies have been effective (Fig. 5.1). However, while access to education has signifcantly improved for both men and women, the formers nevertheless tend to study longer than the latter. The analyses also reveal the structuring role played by the type of residence booklet (urban/rural) in the educational path of young adults. Urban dwellers tend to pursue longer studies than young people from rural areas.
The emergence of this new stage in the life-course of the post-1980 birth cohort impacts the age of occurrence of other events that are constitutive of the transition to adulthood. These specifc events not only tend to occur later in the life-course, but the fndings also reveal variations in the duration spent in these different states. The longer time spent at school has led to a relative delay in the mean age at frst employment and therefore to young adults' fnancial independence. While examining the average age at transitions for the two birth-cohorts, we observe among respondents born in the 1980s the almost perfect juxtaposition of the average age for school termination and the average age at frst employment. On average urban dwellers, regardless of their gender, fnish school just before 21-year-old and enter the labor market at 21-year-old. While young people from rural areas fnish their studies on average between 18 or 19-year-old and start working at 19-year-old.
5.1 A Longer Transition to Adulthood in Post-collectivist China
**Fig. 5.1** Intergenerational comparison of the mean age at transitions (1950–1959 and 1980–1985) (Agricultural employment is also considered a frst job)
The delayed age for school termination, the postponement of the age to frst job, and variations in the average duration spent in a professional activity for young people born in the 1980s, compared to the cohort born in the 1950s, are the result of public policies supporting the expansion of post-secondary education in the 1980s and the dismantling of work units in the 1990s. They also had the consequence of pushing back the age of young adults' fnancial autonomy. In post-collectivist China, it is no longer the political profle1 of candidates that matters to fnd a job, but their level of qualifcation. Indeed, young adults are no longer guaranteed a job for life at the end of their studies. They must fnd their place in an extremely competitive labor market. In this context, the level of education (diploma) turns out to be a comparative advantage for fnding a job that meets their life expectations. In this context, to support their children, families invest more and more in their education. Families' bigger investments in their children's education, and the trend towards a later access to a stable job also leads to a delay in the transition to marriage and
<sup>1</sup>During the Maoist period, the position held in a work unit, the possibility of continuing or not continuing studies, or even vocational training were evaluated based on the political profle of the candidates. The "blood theory" prevailed, according to which not only the "counter-revolutionaries" had to be punished, but also their children and parents in application of the principle: "Hero father, prodigal son; reactionary father, bastard son" (laozi yinxiong, er haohan, laozi fandong, er hundan) (Béja, 2011, p. 12).
parenthood. Whereas their paths to adulthood were disrupted by the Cultural revolution and the send-down movement to the countryside, 87% of the respondents born in the 1950s were married before the age of 28 and 70% gave birth to a child before this very age.
Life-course theorists who studied this birth cohort revealed that these sociohistorical changes had the effect of delaying the stages of their transition to adulthood. Some young people belonging to the sent-down generation resumed their studies after 1976. Other members of this birth cohort only acceded to a job (nonagricultural job) after the end of this State policy in 1976. In addition, interesting research pointed out that this generation tended to wait for their defnitive return to the city to get married and have a child (Chen, 1999; Lin, 2013; Meng & Gregory, 2002; Zhou & Hou, 1999).
The birth control campaign (*wanxishao*), introduced in 1973, may also explain the delay in the age of frst marriage and frst childbirth for members of this birth cohort. Since this demographic policy urged the population to late marriage and procreation, spacing between births and few births (Greenhalgh & Winckler, 2005).2 Out-of-wedlock births were not socially accepted at this time (and still largely today). Therefore, as revealed by the analyses carried out in this book, women's age at frst birth was *de facto* delayed by this campaign.
The end of the "*wanxishao*" campaign at the end of the 1970s, as well as the new marriage law implemented on January 1, 1981 – according to which the legal age of marriage is 20 years for women and 22 years for men (article 5 of the marriage law)3 – could have led to an earlier age at marriage and parenthood for the birth cohort born in the 1980s. However, the data reveal a delay in the average age towards these transitions. If we consider the paths of respondents up to the age of 27, we observe that the average age at frst marriage went from 23–24-year-old for urban women born in the 1950s to 25–26-year-old for city dwellers born in the 1980s (Fig. 6.1). The intercohort comparative analysis of transition rates to frst marriage and frst birth of young people born in the 1980s confrms a postponement in the age of entry into the roles of spouses and parents. At 27-year-old, only 42% of urban respondents born between 1980 and 1985 were married, and 18% had had a child. Among them, just over 50% of women and 34% of men were married at the age of 27. They were 23% to have become mothers and 14% to have become fathers. At the same age, 63% of rural respondents were married and 50% gave birth to a child. Rural women were nearly 70% to be married at 27-year-old, and 57% had become mothers. Rural men were 58% to be married, and 44% to had become fathers at this age.
<sup>2</sup> In cities, where the policy was strictest, women were encouraged not to marry before 25 and men before 28. Couples were also encouraged not to have more than two children, with a 3–4 year interval between births. In the countryside, the marriage age is 23 for girls and 25 for boys. Couples are encouraged not to have more than three children, with 3 or 4 years between births. National minorities (10% of the population) are not affected by these measures.
<sup>3</sup>According to the same article (article 5), late marriage and late births are strongly encouraged: the age of marriage should not be earlier than 22 years for boys and 20 years for girls. Late marriage and late births are strongly encouraged (Marriage Law, 1981).
If the fndings account for a lengthening of the journey towards adulthood in China, unlike what was observed in Germany by Brückner and Mayer (2005), the visible signs of a complexifcation of the life-course are less present in family trajectories than in work trajectories.
#### **5.2 Family Roles and Responsibilities as Central Values**
#### *5.2.1 Marriage, "an existential question" for Young Adults and their Families*
Family roles and responsibilities play a pivotal part in the meaning attributed by the respondents to the transition to adulthood, whereas in research conducted in Western countries, individualistic criteria such as fnancial autonomy, independent decision-making and a sense of responsibility are ranked frst (Arnett & Galambos, 2003). In Confucian thought, the family occupies a privileged position. It is seen as "an extension of the individual" (Cheng, 1997:79). In pre-Maoist China, marriage was a relationship between two family groups that served to perpetuate the family lineage. The marriage law has institutionalized the legal foundation of marriage (the marriage contract) as the basis for living together (Marriage Law, 1981). This normative prescription is relayed in offcial discourses. For instance, Xi Jinping declared in December 2016 that the family constitutes an "*important foundation for national development, progress and social harmony*"4 (CPCNews, 2016). Through this discourse, he explicitly urges the cadres of the Chinese Communist Party (CCP) to serve as an example to the population by promoting and valuing family values. In this speech, President Xi recalls that: "*families form the cells of society. When these are healthy, society prospers; but when family values are lacking, it is diffcult to avoid social unrest*" (CPCNews, 2016). Families are described by Xi as the primary agent of socialization: good family traditions infuence individual values and it is the responsibility of each individual to disseminate them (CPCNews, 2016).
Through this discourse, Xi Jinping openly places the family at the center of the country's project towards prosperity. As was already the case in the early 1950s with the marriage law, and then in the early 1980s with the institutionalization of birth control policies, the Party-State explicitly gives a social, political and economic role to the family by reminding that respect for family traditions benefts not only individuals and their family, but the society as a whole. Through this discourse, the Chinese President promotes a form of familialism, which gives the illusion, supported by the Party-State, that families have the power and resources to fnd solutions to systemic problems.
<sup>4</sup>Already in 2013, Xi Jinping had reminded, in front of the organization of women of China, the importance of promoting family values.
Marriage appears to be not only an important step in the transition to adulthood in China, but also a logical one. At no point in the interviews did the young adults question the validity of the normative model of the couple and the family. It is mentioned 75 times that marriage provides the necessary basis for the formation of a family and the birth of children. For male or female respondents, urban or rural, regardless of their level of education, marriage is not to be taken lightly. It is seen as "the most important thing in an individual's life" (Niu, born in 1982, urban, Hebei *hukou*, specialized college). According to Tingting, who is now an interpreter, marriage is a "normal" process: "*When you love each other, getting married to stabilize your life, I think it's quite natural*" (born in 1993, urban, Shanghai *hukou*, Bachelor). Han's words, who is an administrative employee, also go in this direction:
"*I fnd it very strange not to want to start a family, because it's a core that brings comfort. Getting married, having children, it's something perfectly normal, a duty*" (born in 1989, rural, Jilin *hukou*, Bachelor).
According to Qian, who grew up in a city in Jilin province, in northeast China, where he returned to work as a University professor once he obtained his PhD from Peking University:
*The family, especially a harmonious family, gives a person dignity. [...] that's why in China, marriage, is not just a matter of feeling, but also an existential issue for individuals. [...] A person who does not get married is single and childless. [...] Through marriage the family is formed. [...] Marriage is the frst step to having a family. No registered marriage, no child in China.* (Born in 1982, urban, Jilin *hukou*, PhD)
This normative approach to marriage and parenthood has diverse consequences for men and women I interviewed.
#### *5.2.2 Being Good Wives and Good Mothers (***Xianqi Liangmu***)*
The themes of love and marriage are also at the heart of the TV series studied in the corpus. Cohabitation appears as a possible prerequisite before marriage, but the latter remains the fnal goal of the protagonists. The centrality of marriage shown in TV series could aim to respond to the continuous increase in the number of divorces, which doubled between 1985 and 1995, then tripled in 2005 (Kong, 2008).
The TV series examined convey an ideal of femininity which, unlike the portrayed ideal of masculinity, is less linked to women's professional trajectories than to their family trajectories. In "Beijing Youth" The dialogue below between Quanzheng and her friend, a psychologist, is revealing:
Quanzheng: "*A woman who only has her career and no family (understood in the sense of having a husband and a child) is not a fulflled woman*."5
<sup>5</sup> "A woman who only has a career and no family is not a successful woman" (*zhi you shiye mei you jiating de nüren bu shi yige chenggong de nüren*), DVD 1, sequence 3:08:30–3:08:39.
Her friend responds: "*But a man who only has his career and no family is however highly desirable*."6
Then, the latter continues: "*According to custom, the expectations for men are professional success and it is expected that women be good wives and caring mothers. If, we, women, wish for professional success like men, we must make a double effort. We must juggle a career and take care of our family, otherwise we face social condemnation*."7
Young women who wish to pursue a career are therefore subject to a double constraint and if they fail to juggle both, they expose themselves to social condemnation for having violated the norm of the good wife and devoted mother.
Television series convey another normative constraint addressed to young women: to get married before 28 years old. The term *shengnü*, which means "leftover women", is used in Chinese to refer to young women who are not married after 27 years old. Since 2007 and the offcial use of this term by the All-China Women's Federation and the Ministry of Education, Chinese media have also widely disseminated it through articles, surveys, cartoons, or editorials stigmatizing in particular young women who have completed higher education and who were still single after 27 years (Hong Fincher, 2014). The creation of this new social category by State structures can be interpreted as a desire of the latter to encourage marriages, birth rates and to maintain social peace. Indeed, "[t]he State Council [has named since 2007] the sex-ratio imbalance as one of the population pressures because it « causes a threat to social stability »"(Hong Fincher, 2014:28). The encouragement of marriage and reproduction of educated young women (through the stigmatization of those who are not married and who do not have children) is a means used by the government to increase the quality of the population. This goal is indeed one of the key objectives of the government. The latter has mandated, in this sense, the All-China Women's Federation, propaganda services, the Public Security Bureau and civil affairs to help it implement the family planning policy (Hong Fincher, 2014:29).
The interviews refect the tensions produced by the normative prescriptions that concern young people in relation to marriage. Young women face a double contradictory injunction. A discourse on the "quality of the population ", which emerged when they were still children, values independence, refexivity and physical and intellectual performance. While in parallel, another discourse, also conveyed by the State and its institutions, values wives and mothers capable of taking care of their spouse and their children by ensuring the formation of harmonious families. For all
<sup>6</sup> "But a man who only has a career and no family is indeed a diamond bachelor", (*keshi guangyou shiye mei you jiating de nanren queshi zuanshi wanglaowu*), DVD 1, sequence 3:08:30–3:08:39.
<sup>7</sup> "The secular expectation for men is to have a successful career, for women it is to be a good wife and mother. If we women want to achieve the same success as men, we have to put in double the effort. At the same time, we must also take care of the family, otherwise we will be condemned by society." (*shisu dui nanren de qiwang shi shiye youcheng, dui nüren de qiwang shi xianqi liangmu. Ruguo women nüren xiang huode tamen nanren tongyang de chenggong, jiu dei fuchu shuangbei de nuli. Tongshi hai dei jiangu jiating, fouze jiu yao zaodao shehui de qianze*), DVD 1, sequence 3:08:44–3:08:59.
the female respondents, marriage plays a central role in the meaning they give to the transition to adulthood, more important than the transition to employment. Concepts such as "stability" (*wending*), "guarantee" (*baozhang*) and "feeling of security" (*anquangan*) are frequently mentioned in the respondents' discourses on marriage. Wei, who is a teacher in Beijing and comes from Heilongjiang province, in northeast China, explains:
*A cohabiting household cannot be considered a real family because the situation is not stable. [...] Such a situation would make me anxious because I have no guarantee. The government gives its approval for marriage, you receive a certifcate, a marriage certifcate. So, I consider marriage as legal (hefa) and cohabitation as illegal (fei hefa). If you cohabit, people might wonder: Why don't you get married? Why do you cohabit? I think that marriage is the frst step, then you can live with your spouse. Because in China if you don't get married and you live in cohabitation, people will not only think that the couple has a problem, but that the whole family has a problem. Everyone will wonder: But why don't they formalize their situation in front of the government by getting married? Living in cohabitation without getting married, it's like a marriage, but with the bad sides only for a woman. You would have to do the housework, cook, without feeling secure. So why not get married?* (Born in 1981, urban, Heilongjiang *hukou*, Master)
If the idealized conception of marriage refers to a stable union, which is meant to last a lifetime and legitimized by law, young women are down-to-earth. Like Suzhi, who raises her son alone and who works in a massage parlor in Beijing, her words clarify those of Wei about the idea of guarantee/insurance that marriage provides:
*Marriage is like life insurance for women. If a woman cohabits with her boyfriend in the long run, in case of separation she risks ending up without a house, without money and without work if she did not work. Not cohabiting is an argument to ensure that the relationship with the spouse ends with a marriage.* (Born in 1978, rural, Hebei *hukou*, college)
Linguistic changes also refect the weight of the injunction to marriage in the lives of young adults, and particularly young women. Sentences such as "I am part of the tribe of the unmarried" (*wo shi bu hunzu*) used by Shasha, or even that of " leftover women" (*shengnü*) evokes a dichotomy between married and unmarried people that leaves little (or no) room for other ways of "making family". Women who remain single beyond 27 years are stigmatized. Wei describes the situation in these terms:
*Society tells us that if you are not married by 25–27 years old, there is something wrong with you. In China our traditional culture tells us that from 25–27 years old you should marry a man. And if you don't do it others will think you are strange, or ugly, or that you have private things [...].* (Born in 1981, urban, Heilongjiang *hukou*, Master)
If the transition to adulthood is experienced as a journey that takes time, its duration is socially constrained, especially for young women, regardless of their social anchoring. This is illustrated by Wei's path. She complains about the social pressure she feels because she is not yet married at over 30 years old (born in 1981, urban, Heilongjiang *hukou*, master's degree). It is also exemplifed by Suzhi's life-course. She feels socially sanctioned because of her single mother status and because she gave birth to a child out of wedlock. The attitudes of other respondents illuminate the weight of these social injunctions on the shoulders of young women as well. To try to respect the social script, they accelerate the passage of role transitions characteristic of adulthood around the age of 27–28 years old.
According to Qian,
*marriage is a prerequisite for obtaining a [adult] status within society. [...] a woman or a man who would remain single would have a somewhat vague social status, as the Chinese generally refer to the family status of people. If an individual does not get married, she or he could be socially discriminated against, and be perceived as an incomplete person in the sense that she would not have fulflled her responsibilities.* (Born in 1982, urban, Jilin *hukou*, PhD).
Getting married and having children are transitions intimately linked in the minds of young adults. The birth of children is indeed an integral part of the normative expectations surrounding marriage. Like most of the respondents, Xiaocui, a teacher in a private school in the Chinese capital, admits to never having asked herself the question of having or not having a child and even less of its birth outside the legitimate framework of marriage:
"*To me it's not a question, I frst want to get married and then have a child*" (born in 1986, rural, Heilongjiang, *hukou*, Bachelor).
For Han, it is clear that this is what motivates many young women:
*In China many women are concerned about the issue of children. I think that many women only want to get married for this reason. [...] Really, it's to have a child that they want to get married before 30 years old, because that's when the body is in better condition to give birth to a child.* (Born in 1989, rural, Jilin *hukou*, Bachelor)
Contrary to what has been observed in the United States, marriage and parenthood are not disconnected in the minds of young people in China (Furstenberg, 2010). These transitions confer new social roles and personal attributes to young adults.
#### *5.2.3 "Housing [...] Is the Foundation of the Family"*
"Financial autonomy" (*duli shenghuo*) from parents, is a central aspect in the respondents' defnitions of adulthood. According to Qian, being an adult means no longer depending on fnancial aid from parents to live (born in 1982, urban, Jilin *hukou*, PhD). This aspiration is more present in the discourse of male respondents. They feel an obligation, once adult, to be the main provider for their family. In this sense, fnancial autonomy from parents is less about proving their individualism than it is about them acting as breadwinners and no longer being a burden on their parents. It is about moving from a logic of assistance to a logic of reciprocity based on flial ties. Once schooling is fnished, it is up to young people to support themselves fnancially, or even to help their parents fnancially. The respondents' comments denote a deep structuring of trajectories by a strong valorization of employment, especially in male paths, with employment being the main vector of fnancial autonomy.
Financial autonomy is never mentioned as the only element marking the transition to adulthood. Having one's own housing is another central element in the respondents' discourse. Renting a home is not a satisfactory solution for the respondents. The comments of Qian and Xiaocui are emblematic. The respondents consider housing as a solid and stable foundation for starting a family:
*There is no religion in China, but there is a belief: devotion to the family. Devotion is not for the Nation, for society or leaders. It is for the family. The family is the heart of the Nation. We all think that housing is not just a construction, a building, it is the foundation of the family [...] renting [a home] does not constitute this foundation.* (Qian, born in 1982, urban, Jilin *hukou*, PhD)
*Without a house, if marriage is still possible, it is however unthinkable to have a child.* (Xiaocui, born in 1986, rural, Heilongjiang *hukou*, Bachelor).
If data suggest that young adults achieve residential autonomy early, becoming a homeowner is however not easy for them. Since the country's opening-up to market economy and the privatization of housing, men remain the main providers of housing (Kane & Li, 2021). Access to property constitutes in the collective imagination an element of identifcation with the middle class (Zhang, 2008). Access to property, however, is not just a vector of social distinction. In a context where the State has largely withdrawn from its social and protective functions, property ownership is also seen as a form of economic security. This is one of the reasons why the purchase of a home plays such a central place in marriage negotiations. This subject came up nearly a hundred times in the interviews.
Although the article 3 of the Marriage law prohibits the transfer of money or gifts in connection with marriage, practices are moving away from this prohibition (Marriage Law, 1981). Even today, even if it is sanctioned by the Marriage law, the practice of the "bride price" persists both in rural and urban areas. Traditionally, once married most young women would go to live in their spouse's family. The "bride price" was used to thank the parents of the bride for having raised her until her marriage and to compensate them for the loss of a resource (Croll, 1981; Yan, 2009). Since the 1990s in cities, the standard of private three-room apartments with "a room of one's own" has replaced that of overcrowded apartments offered by work units (Davis, 2002). In recent decades, this new need for intimacy within the private sphere and the staggering rise in real estate prices make intergenerational cohabitation increasingly rare in urban areas.
While it is socially expected that men (or their families) provide housing at the time of marriage, the interviews revealed a more nuanced reality that depends on the fnancial situation of families. Very few young men were in the position of being able to offer housing to their fancée. The interviews reveal several confgurations: in the frst, the spouse's parents had already bought additional housing in anticipation of their son's marriage, so that the young married couple could settle there. These are urban families from Beijing and Shanghai. In the second confguration, the girl's parents also acquired additional housing, which they transferred to their only daughter so that she can earn a rental income in case she does not live there. However, as in the frst situation, her parents think it is the responsibility of their daughter's husband and his family to provide housing for the couple. In the third confguration, which was the most common, the spouse's parents fnance part or all of the down payment necessary for the purchase of a house, and the spouses together repay the mortgage. I observed that in this situation young women usually use their dowry to invest in the purchase of the house. In the last confguration, called by the respondents "naked marriage" (*luohun*), they pool together their fnancial resources in order to be able to take out a mortgage and repay it. The future spouses can set up a common strategy and unite, as Xiaocui and her spouse did, to negotiate the highest possible "bride price". This strategy aims to increase the amount of their savings and the part of the down payment they can invest in the property. Xiaocui thought that her in-laws would be able to fnance the entire amount of the down payment for their mortgage. After the frst discussions between her fancé and his parents, it quickly became apparent that this would not be the case. They therefore decided that he would negotiate directly with his parents. Xiaocui believed it is important that her name appears on the deed of purchase of the property so that she is not harmed in the event of divorce. Since not only has she invested the money she received for the "bride price" as well as her savings in the apartment, but she is also repaying their mortgage at the same level as her spouse (Xiaocui, born in 1986, rural, Heilongjiang *hukou*, Bachelor).
This example shows that access to property is seen by the respondents as a marker of the transition to adulthood. For the male respondents, it is a central element of their identity, as it is a criterion by which men are judged. It embodies the autonomy and ability that men have to earn their living and to provide for their family (Kane & Li, 2021). However, as the interviews illustrate, the reality is more complex, as women also often play their part to enable the couple to become homeowners.
Housing also occupies a central place in each of the TV series analyzed. The choice to stage this private space is heavy with meaning. Until the 1990s, in an attempt to erase private spaces from social imagination, these spaces were not shown in TV series. The series "Yearnings" (*kewang*), frst aired at the end of 1990, marks a turning point towards the individualization of housing. For the frst time, the narrative is set in the family space and focuses on family life (Zhu, 2008:81).
In each of the TV dramas examined, housing plays a key symbolic role in young people's transition to adulthood. It is depicted as a symbol of stability. In the TV drama "Ants' Struggle", the narrative revolves around the daily struggle of young adults from outside Beijing (*waidi ren*) to make a place for themselves in the city. The staging of housing is used to show the ideal of material comfort and social status they hope to achieve through hard work. The frst scene shows the dormitories in which the main characters live. Male characters live in a small room made up of several beds, without privacy, a bathroom and a communal kitchen. Women's dormitories are separated from those of men. The second scene shows Xiaoyan, a young migrant. She comes from the countryside and left school after high school. She is portrayed as preparing an apartment she rented especially for the arrival of her boyfriend Hu Yifan's mother. He, originally from a provincial city, has completed university studies in the capital. They both live in separate dormitories, in a residence located in the Tangjialing district. To convince her future mother-inlaw to agree to her marrying Hu Yifan, Xiaoyan rented for the duration of her stay (one month) an apartment. The place, located in a freshly built residential complex inhabited mainly by middle-class families, has two bedrooms, a living room, a kitchen and a bathroom. This staging, orchestrated by Xiaoyan, aims to symbolize the stability of the young couple, as well as their social ascent through material comfort and social insertion in Beijing.
The purchase of a home also constitutes an obstacle to the marriage of Tong Jiaqian and Liu Yiyang in the eyes of the young woman's parents, in the series "Naked Wedding Era". Liu Yiyang is thus led to prove, in a moving way, to Tong Jiaqian's mother his willingness to work hard to provide a home for his fancée:
*I know that in your eyes I am not good enough for your daughter, you think that I have no money, no savings, that I am neither a homeowner nor a car owner. But the fact that I don't have this now does not mean that I won't have it later. You older people, you say that one can say how much one loves each other, but that does not allow one to survive [in other words, one cannot live on love and fresh water alone]. But ayi, I think that Jiaqian and I are capable of it. […] Once I graduated from University when I was looking for a job, I was determined to fnd a job that pays a lot, that allows me to buy a house, a car and that allows me to propose to Tong Jiaqian in grand style. But despite all my efforts, I cannot keep up with the speed of real estate infation*. 8
If his efforts are not enough to achieve this for the moment, he believes they will be in the future. That's why once married the young couple plans to temporarily rent an apartment.9 Moreover, as the tirade below suggests, the narrative places the responsibility for acquiring a home, which the couple would own, on the young man's shoulders. Tong Jiaqian asks her fancé who agrees:
*This rental is only a transitional period. It is certain that you cannot let me live all my life in a rental, can you?*<sup>10</sup>
<sup>8</sup> "I know you've always looked down on me, thinking I have no money, no savings, no house, no car. But just because I don't have them now doesn't mean I won't have them in the future. You always say that our verbal expressions of love can't be eaten as food. But auntie, I think they can in my and Jiaqian's case. […] After graduating from university and looking for a job, I was desperate to fnd a job that paid enough for me to afford a house, a car, and to proudly marry Tong Jiaqian. But even if I risked my life, I couldn't keep up with the speed of the housing price increase." (*wo zhidao nin da xinyanr li jiu qiaobushang wo, juede wo mei qian, mei jixu, meifang, meiche. Danshi wo xianzai meiyou bu daibiao wo yihou meiyou a. ni lao shuo women guang zuishang ai lai ai qu de, bu neng dang fanchi. Danshi ayi, zai wo he jiaqian zher wo juede neng. […] houlai*).
<sup>9</sup> "Increase the proportion of the middle-income group, *kuoda zhongdeng shouru qunti de bizhong* ".
<sup>10</sup>The residence booklet (*hukou*), institutionalized in the 1950s and still in force today although softened several times, is a key administrative tool to understand social policies in China. It not only allows to control population fows within the country, but also to determine the place of perception of social benefts. This administrative document produces forms of discrimination and exclusion within the population by institutionalizing a division between the population holding an urban *hukou* and those holding a rural *hukou* (Wang, 2005).
He also says he is ready to work hard to ensure the material comfort of the couple.11 However, their mothers are opposed to the idea of a "naked marriage" and prefer instead multi-generational cohabitation. Again, in the TV drama "Rules before Divorce", the renting of a home as proposed by the young married couple, Mingxuan and Xinyao, is not feasible from the parents' point of view. After a tough negotiation between the two families, it is decided that the young married couple will buy, with the help of their parents (especially those of Mingxuan, the husband), a new apartment in a residential compound.12
In post-collectivist China, where 82.3% of the population owned their own home in 2007, the style of home purchased and inhabited became a vector of social distinction. The middle class turned away from socialist buildings. These three-storey buildings have basic equipment. They prefer to live in gated residential compound, which group several buildings over 30 foors high (Man, 2013; Tomba, 2013). These are the kinds of homes that are said in TV drama to typify "good taste". The decoration of middle-class homes no longer refects their allegiance to the Party-State, but the tastes, personalities, and lifestyles of their occupants.
The housing reform initiated by the State began in the early 1980s. The policy was then extended after 1988. In a frst step (1980–1998),13 the aim was to privatize public housing and housing belonging to work units. These homes were sold to employees at preferential rates.14 As a result, the benefts enjoyed by some workers during the Maoist period have been perpetuated in post-collectivist China.15 With the rise in property prices and the government interventionist policies,16 these lucky workers were able to use cheaply bought homes to invest in new property in the 1990s. These years marked the beginning of the construction of commercial housing, in a new type of buildings: gated communities (*shequ fengbi*), which have become the new architectural standard (Tomba, 2013:187).
<sup>11</sup>The work unit (单位, *danwei)* refers to state and collective enterprises, as well as administrative units in urban areas.
<sup>12</sup>These services notably cover assistance to orphans, to disabled people, to elderly people without family or with too modest incomes. If these services were already provided by the residents' committees during the Maoist period (居民委员会, *jumin weiyuanhui*), the scope of neighborhood community activities has expanded: they also provide fnancial support to people who have been dismissed from their posts during the restructuring of state enterprises (下岗, *xiagang*), to parents who have lost their only child and who can no longer have offspring to take care of them, they also relay the directives of family planning in terms of birth control, contraception and reproductive health.
<sup>13</sup>Article 33, chapter 2 of the Constitution of the People's Republic of China: http://www.npc.gov. cn/npc/xinwen/node\_505.htm#2
<sup>14</sup>On the concept of new social risks refer to the article by Giuliano Bonoli (Bonoli, 2005).
<sup>15</sup>The 1993 law on fscal decentralization has reinforced the heterogeneity of interests between administrative entities.
<sup>16</sup> In the early 1980s, the State still imposed on agricultural households production quotas that will gradually disappear as the reforms progress.
These changes in the housing policy have had the effect of creating a new segregation of residential spaces, which is illustrated in TV drama by opposing the two categories of housing. In the TV drama "The Era of Naked Marriage", the different lifestyles are used to indicate a difference in social status between two families from Beijing. Three generations of the Liu family live in a popular three-storey building. The fat they bought from the work unit is on the frst foor. It consists of a kitchen, a living room in which the parents – who are workers – live, and two independent bedrooms – one is for their son and the other is for the paternal grandmother. The apartment is rudimentary. The kitchen and the pots seem to date from the Maoist era and the decoration is made from cheap objects. For their part, the Tong family lives in a two-generation (parents-child) apartment located on the upper foors of a recently built and secure residence. The apartment has a modern kitchen, living room, bathroom and bedrooms. The parents, who are civil servants, and their daughter have separate bedrooms. The decoration of their apartment is cosmopolitan. It recalls the interiors presented in IKEA store catalogs.
In the TV drama "Rules Before Divorce" the same opposition technique is used to stage the homes and implicitly signal the social status of their occupants: the Zhang family and the Li family. The Zhangs live in a popular three-storey building, an apartment bought from their work unit. The decoration of the apartment has been worked on by its occupants, but it refects their modest fnancial means. Whereas the Li family live in a posh house located in an upscale suburb of the capital. The staging of the father's professional success is refected in the decoration of their interior with prestigious goods (solid wood traditional Chinese furniture, works of art, abundance of fruits and green plants, etc.) and the volume of spaces.
The presentation of the median and upper middle classes as homeowners of a home in new gated residential compounds or in a posh suburb is to be seen as part of the State discourse on the "quality" of the population and the construction of a society of "small prosperity" (*xiaokang shehui*). This latter concept refers to the Book of Rites, a classic text in which the concept frst appears (Tomba, 2013). This historical reference places the project of societization proposed by the Party-State in a millennial historical continuity. By skillfully marrying socialism with classical thought, it manages to resolve the ideological confict posed by the opening-up of the country to market economy without making a radical break with the past. The pursuit of private economic interests was recognized as legitimate and became the new standard. The rapid emergence of the consumer society not only increased individuals' choices and material comfort, but by granting them greater autonomy to some extent, it also destroyed the State's monopoly over individual lives and made citizens responsible for their own lives.
The places, the way they are decorated and the consumer practices that are highlighted by the *mise en scène* in television series are highly signifcant. Through these social practices and the appropriation of lifestyles associated with the middle classes, young adults seek to distance themselves from what they consider to be old-fashioned and to move closer to the feeling of belonging to the middle class. However, anyone familiar with Beijing will realize that the homes portrayed in TV series are not always in line with social reality. There is a distortion aimed at valorizing housing associated with high social statuses. Yet, the people who live there do
not always have a professional activity that would enable them to afford such luxury. The use of these representations contributes to valorizing the socioeconomic attributes of the middle class. Indeed, in 2002 during the People's National Assembly, the Party-State called for an expansion of the proportion of the "middle income group" within the society17 (Li, 2013a, b:11). The middle class is seen by the Party-State as an asset and a political ally for the socioeconomic development and social cohesion of the country, offering hope of upward social mobility for the most disadvantaged fringe of the population, as well as a model level of "quality" to be reached.
#### **Bibliography**
Bonoli, G. (2005). The politics of the new social policies: Providing coverage against new social risks in mature welfare states. *Policy & Politics, 33*(3), 431–439.
<sup>17</sup>*Migrant population* (迁移人口), *migratory population* (迁徙人口), *foreign population* (外来人 口), *infow population* (流入人口), *foreign mobile population* (外来流动人口), *foreign workers and business personnel* (外来务工经商人员), *temporary resident population* (暂住人口), *spontaneous migrant population* (自发迁移人口), *self-fowing population* (自流人口), *lodger population* (寄住人口), *foreign temporary resident population* (外来暂住人口), *short-term migrant population* (短期迁移人口), *temporary migrant population* (暂时性迁移人口), *informal migrant population* (非正式迁移人口), *non-registered migrant population* (非户口迁移人口), *migratory population* (流迁人口), *peasant workers* (民工), *laborer* (打工仔), *female laborer* (打工妹), *rural migrant workers* (农民工), *surge of migrant workers* (民工潮), *blind fow* (盲流), *population separated from household registration* (人户分离人口), *overbirth guerrillas* (超生游击队), *new citizens* (新市民), *blue-stamp household population* (蓝印户口人口), etc. (Duan et al., 2012; Hou, 2010; Shen, 2011; Sun & Li, 2015; Wang, 2010; Yan, 1999; Yang & Yang, 2009).
Croll, E. (1981). *The politics of marriage in contemporary China*. Cambridge University Press.
**Open Access** This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.
The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.
# **Chapter 6 The Rise of New Social Risks in Post-collectivist China**
#### **Contents**
**Abstract** This chapter highlights in a frst section the complex changes in China's Welfare system, marked by historical legacies, reforms, and the ongoing tension between collectivist and individualized approaches.
The chapter reveals that China's transition to market economy triggered shifts in employment dynamics. The dismantling of State and collective sectors in 1992 led to a surge in job insecurity, with most jobs moving to the private sector by 2005. Efforts to formalize labor contracts began in 1986 but faced challenges. The 1994 Labor Law standardized contracts, but informal jobs persisted.
The chapter uncovers that in this context of uncertainties, intergenerational solidarities tend to replace the collectivist provisions. Demographic shifts strain this new contract. Faced with the challenges of precarious employment, particularly in the informal sector, young adults exhibit varying vulnerabilities. Migrants face increased vulnerabilities, often lacking social security schemes. Many respondents, including those with degrees, experience frustration due to fxed-term contracts, impacting income and personal life. The risk of disaffliation is exacerbated by long working hours, hindering social interactions. Despite attempts with the 2008 labor law to address job precariousness, poor law enforcement persists, especially after the 2008 economic crisis.
**Keywords** Hukou · New social risks · Welfare State · Uncertainties · Employment, Intergenerational solidarities
#### **6.1 From Collective to Individualized Social Policies**
The new middle class, like other social groups, is not immune to new social risks that result from the decollectivization and the individualization of the social protection system. To understand the transformation of employment policies and social policies in China, it is important to keep in mind that the Chinese Communist Party (CCP) still exercises complete control over their development (Chan et al., 2008). In a political system where there is *de facto* a single political party, the CCP is not subject to the same political demands as in multi-party systems. The CCP is, of course, not a homogeneous and uniform block. It indeed covers several conficting visions - conservative, moderate and reformist - on social policies and, therefore, on solutions to social problems and demands from civil society.
During the collectivist period, the social protection system was already characterized by its duality. The distinction between urban and rural population, institutionalized through the residence booklet (*hukou*) 1 which operates a social hierarchy between local population and population coming from outside, allowed, and still allows, to assign unequal social rights to different categories of citizens. Individuals are attached to a community - rural or urban - and resources they receive from the community depend on its fnancial situation. This results in a social and spatial hierarchy in which township are at the bottom of the scale and autonomous municipalities, like Beijing, the country's capital, are at the top. Therefore, people coming from places with better resources (top of the scale) beneft from a higher level of social protection than others.
In the planned economy system (1949–1978), the structure of social policies, organized around the work unit or the people's commune, aimed to encourage the population to participate in the productive effort. Groups of population directly involved in the industrialization effort of the country (offcials, urban workers) were favored, at the expense of rural workers who were not attached to a work unit2 (Cook, 2000). While the rural population had to rely entirely on its own resources, from the early 1950s the urban population began to beneft, not only from lifetime employment, housing, education, and subsidized childcare, but also from a social security system (*shehui baozhang*). This included social insurances (*shehui baoxian*), social protections (*shehui fuli*), social assistance (*shehui jiuji*), and assistance to disabled or deceased revolutionaries and military personnel and their families (*shehui youfu*). These mechanisms, managed by government agencies, targeted different groups of the population at different stages of their life-course.
<sup>1</sup>According to experts, data on the number of migrants published in offcial sources are probably below reality.
<sup>2</sup> "Atypical" jobs are defned in opposition to the "typical" employment relationship, which according to the International Labour Organization involves "working continuously full-time, within a direct relationship between the employer and the employee", https://www.ilo.org/infostories/ fr-FR/Stories/Employment/Non-tandard-Employment#what-is-non-standard-employment
From the early 1980s, the collectivist social security system began to be reformed where it was least effcient, that is, in the countryside, with the dismantling of people's communes. Then, the reform of this system continued where its institutional anchoring was the strongest, that is, in the cities. From the early 1980s, and especially from the mid-1990s, the Party-State sought to transfer the responsibility and the obligation to fnance the social security system from work units or people's communes to State agencies, the community, and individuals (Cook, 2000; Gao et al., 2013). Various non-State actors (NGOs, associations, and communities) have emerged in the social landscape, albeit maintaining close ties with the Party-State. Alongside unemployment insurance, adopted as early as 1986 to address the growing problem of urban unemployment caused by the dismantling of state and collective enterprises, the Party-State implemented neighborhood communities (*shequ jianshe*) to take over part of the role played by the work units. These local organizations combine service to residents (*shequ fuwu*) 3 with a degree of social control over them (Monteil, 2010). By 2014, the unemployment insurance system had been revised four times since its creation. This insurance is administered by local governments and fnanced both by local, provincial governments, the employer (2% of all wages) and the employee (1% of his gross income). The systems of health and maternity insurances institutionalized in 1951 have been adjusted no less than 14 times since the late 1970s, with a State desire, since 2011, to unify the system at the national level. Again, benefts are partly conditioned by workers' contributions. Dating from 1951, the insurance system covering work accidents has been amended seven times since 1978. It is administered by local governments and fnanced by the employer (1% of all wages). China has also a pension system, disability and death insurances (in case of death of the spouse or parents). They have been reformed many times since 1951. Since 2011, the Chinese government has also expressed the desire to unify these social insurances at the national level (with the aspiration to cover the rural population). These insurances are fnanced both by the central government and by local governments, employers and employees in varying proportions depending on the localities (SSPTW, 2012; Wang & Ding, 2012). Citizens who hold a rural residence booklet remain largely excluded from these reforms, although *de jure* article 33 of the Constitution guarantees the same civic, social and political rights to the entire population.4
The process of decollectivization has led to the emergence of new social risks, particularly in the labor market and in terms of intergenerational solidarity.
<sup>3</sup>Precarity is understood as "more labile relationships to work that contrast with the stability and consistency" of permanent jobs (Castel, 2009, p. 163).
<sup>4</sup>While it is common to use the word *gaokao* to refer to this exam, it is actually the *putong gaodeng xuexiao zhaosheng kaoshi*, which can be literally translated as "general recruitment exam for higher schools". This exam is national, but its content varies according to provinces and municipalities. The number of points obtained in this test determines the quality of the universities that students can join.
#### **6.2 The Rise of New Social Risks5**
#### *6.2.1 The Rise of Work Uncertainties*
With the opening-up to the market economy, allocated jobs are no longer the norm. It has been supplanted by "chosen" employment. Before 1978, almost all urban jobs were in the State and collective sectors. Then, with the dismantling of State and collective enterprises undertaken in 1992, by 1995, 40% of urban workers had been dismissed from their posts. This represents 50 million unemployed people. In 2005, State and collective sectors accounted for only 27% of urban jobs. Since then, more than 70% of jobs in cities have been redirected to the private economy sector (Naughton, 2007:184; Park & Cai, 2011:17). The State has thus gradually withdrawn from the responsibility of guaranteeing a right "to" work and providing a solid safety net of social protections to the entire urban population of working age. As a result, candidates have been enjoined to learn to "sell themselves" and their skills through the writing of *curriculums vitae* and cover letters. Employers, for their part, have learnt to present an attractive image of their companies to attract the best candidates. Deprived of job security, workers are now in competition for job positions (Naughton, 2007; Park & Cai, 2011).
As early as 1986, the central government attempted to implement "Temporary Regulations on the Labor Contract System" (RTSCT, 1986). Despite this attempt, a number of companies have continued to recruit workers without signing a labor contract with them. These companies, relying on cheap labor, have continued to favor work fexibility at the expense of job security and social protection for workers (Gallagher & Dong, 2011). To attract foreign investments during the frst stage of reforms (1978–1992), local governments were inclined to turn a blind eye to the continuous increase in the number of informal jobs, that is to say precarious and non-guaranteed jobs, which do not result from the signing of an agreement or a labor contract and do not guarantee social protection for workers (Gallagher et al., 2011).
It was not until 1994 that labor law made its frst developments nationwide. For the frst time since the country's opening-up to market economy, the 1994 Labor law standardized labor law in a single text (LDF, 1995). Before this law, a multitude of laws, regulations, and directives coexisted. Their content varied depending on the type of company and the sector of activity. The 1994 law established several categories of labor contracts, each entitling to specifc levels of social protection. Only open-ended contracts (OEC) guarantee a right to social insurances based on a contribution shared between the employer and the employee. They encompass health, accident, pension, unemployment and maternity insurances. With fxed-term contracts (FTC) and fxed-term contracts for a specifc mission (FTCM), it is the workers who must fnance their social insurance on their own (LDF, 1995). Despite the
<sup>5</sup> Javornik prefers the term "State familialism" to analyze the transformations that have taken place in post-socialist Eastern European countries in the feld of social policies (Javornik, 2014).
1994 labor law, which requires companies to sign a labor contract with their employees, many have ignored the legislation. In response to the competition introduced by neoliberalism, they prefer to recruit fexible staff, or even staff without work contracts because these employees cost less, rather than employees with open-ended contracts (Lin, 2011; Peng, 2009; Swider, 2011; Zhang, 2011). In 2005, between 27% and 36% of jobs were in the informal economy sector (Park & Cai, 2011; Peng, 2009; Tong, 2009). Number discrepancies between studies in the literature can be explained by the use of different defnitions of informal employment according to authors.
Despite the safeguards provided to workers by the 1994 labor law, opening the door to fexible jobs and guaranteeing companies a relative autonomy from the State, the law has ultimately better served the interest of employers than those of workers. This law gave the *coup de grâce* to the socialist welfare system (the iron rice bowl). It institutionalized as a national standard the practice of individualized labor contracts, which provided differentiated social protections in place of permanent jobs and unifed social protection from cradle to grave. In this sense, the new legislation defnitively broke with the old socialist social pact. In 2008, the Party-State tried to remedy this situation by institutionalizing the new labor contract law (LDF, 2008). The aim of this law was to respond to the growing precariousness of work by requiring employers to sign a work contract with their employees. The law also requires employers to offer their employees an open-ended contract after two consecutive fxed-term contracts, in order to prevent the succession of fxed-term contracts. Moreover, to terminate a labor contract, it is expected that the employer motivates his reasons. The implementation of this legislation is part of the Hu-Wen administration's project to build a harmonious society.
As will be detailed in the following section, the 2008 labor law has not succeeded in stemming the presence of informal jobs and inequalities in access to employment. The analyses conducted in the context in this book indicate that young adults from Beijing are signifcantly more likely to fnd a job in public administration or a state-owned enterprise. These "in-the-system" jobs (*tizhi nei*) require a university degree and an urban residence booklet. The life-course being more uncertain than in the past, *tizhi nei* jobs are nowadays highly sought after. They offer not only economic security, but also social prestige. The relative ineffectiveness of the 2008 labor law can be explained in part by the attitude of provincial and local governments. As the latter play a major role in the funding6 and implementation of social policies, their costs, combined with the need to balance their budgets, sometimes lead them to resist the implementation of policies issued at the national level.
<sup>6</sup>The position held in a work unit, the possibility of continuing or not continuing studies, or even vocational training were evaluated based on the political profle of the candidates. During the Maoist period, the "blood theory" prevailed, according to which not only the "counterrevolutionaries" had to be punished, but also their children and parents in application of the principle: "Hero father, prodigal son; reactionary father, bastard son" (*laozi yinxiong, er haohan, laozi fandong, er hundan)*, (Béja, 2011).
#### *6.2.2 Increased Dependency on Intergenerational Solidarities*
During the Maoist period, Marxist ideology, the collectivization of the economy, and the almost general elimination of private property destabilized the foundations on which family solidarities rested in pre-Maoist China. The social protection system offered to the population and, in particular to workers in the work units, provided them with a pension, housing, and medical care once they reached retirement age. These social supports provided by the State supplemented traditional support from the family (Sheng & Settles, 2006). In this sense, they had an individualizing effect by allowing individuals to no longer depend on the "traditional" family organization and solidarity system. This also had the effect of making them dependent on the social protection system guaranteed by the socialist State.
Since the 1980s, with the policy of decollectivization and the crumbling of the social security system, the Party-State has gradually transferred the role once played by the socialist State (the people's communes and work units) to close family members: parents, children, grandparents and siblings. A new social contract is then born. It relies on intergenerational solidarities, as refected in the marriage law. Its article 29 stipulates that older brothers and sisters who have the means to do so have a duty to look after their minor brothers and sisters if their parents are unable to do so (HYF, 1981).
This process of re-familialization should be understood as the State's gradual withdrawal from social policies in favor of families. It can be observed both in rural and urban areas. In the countryside before the early 1980s, the People's communes constituted the basic political and economic unit to which individuals were attached. With their dismantling, the family responsibility system succeeded them. Farmers then gained a right of usufruct over their land. Households, which had once again become the unit of production, were allowed to manage their farm7 (Bianco, 2005). The loss of the social safety net previously provided by the People's communes, even if it was rudimentary, tends to strengthen family solidarity. In urban areas at the beginning of the reforms, private entrepreneurship, although encouraged by offcial rhetoric, remained limited. The urban population clung frmly to the work unit system and the socio-economic benefts it provided. In this system, urban families played a very weak role in social protection (Davis & Harrell, 1993), but the situation would suddenly change after 1992. Following the dismantling of the work unit system, and due to the lack of institutionalization of robust and inclusive social policies, some individuals had no choice but to remobilize family solidarities to compensate for the shortcomings in the new welfare system. Article 21 of the marriage law states that parents have the duty to raise and educate their children, and that children have a duty to support and assist their parents (HYF, 1981).
However, the combined effect of longer life expectancy and demographic policies aimed at containing the population growth are putting this new intergenerational contract under strain. Although since the 1980s the country has benefted
<sup>7</sup>Agricultural employment is also considered a frst job.
from a "demographic bonus", refected in a low dependency ratio of the inactive population to the active population, this ratio now is inexorably shrinking. A social protection system and a pay-as-you-go pension scheme have not yet been generalized to the entire population, so many among those benefting from a pension scheme fnd it diffcult to make ends meet. In the mid-2000s, half of the retirees could only survive with the help of a family member. This family member is often an only child (Attané, 2011), who has sometimes been compelled to migrate to meet the family's economic needs. Change in employment policies and the rise of uncertainty in the labor market have gone hand in hand with a loosening of the Party-State's control over internal migration fows. The absence of young adults makes it diffcult for them to meet the daily needs of their elders in terms of social, medical, or administrative support.
The 1980s saw the start of major waves of migration from the countryside to the cities. These migrants are referred to as "foating population" (*liudong renkou*) because if they work in the city, their *hukou* and entitlement to social benefts remain registered in their place of origin. The authors have identifed in the literature more than 20 concepts to designate the migrant population.8 Some refer to legal temporary work migrations from the countryside to the cities or interurban, while others refer to illegal migrations or even the type of activity carried out. It should be underlined that government statistical reports differ on the temporal indicator used to make the calculations. The 1987 and 1995 surveys, focusing on the migrant population (*qianyi renkou*) and carried out on a sample of 1% of the population, consider as migrants those who have left their place of registration of the *hukou* for more than 6 months (Wang, 2010:3); while in the population census carried out in 1990, those considered as migrants are people who have left their place of registration of the *hukou* for more than a year. The 2000 census also uses the duration of 6 months as an indicator. The plurality of defnitions associated with the concepts *qianyi renkou* and *liudong renkou* explains the differences in the fgures presented by the various statistical sources. The latest population census does not defne the migrant population in Beijing as *qianyi renkou*, but as *wailai renkou*. According to the defnition adopted by the National Bureau of Statistics (NBS) who conducted the census, the migrant population refers to the long-term resident population (*changzhu*) in possession of a *hukou* registered in another province or city and who have been away from their place of origin for at least 6 months. The National Bureau of Statistics also excludes foreign nationals or those from Hong Kong, Macao and Taiwan from this defnition (NBS, 2012). People who have come to work in the capital for less than 6 months are therefore not included in these data, nor are informal workers. The ambiguity of the various categories makes it diffcult for researchers to estimate the exact number of migrants in China in general and in Beijing in particular.
Today in China, these migratory fows are numerous as they involve more than 200 million people (Chan, 2010). Between 1978 and 2011, the urban population's
<sup>8</sup> In rural areas, the marriage age is 23 for girls and 25 for boys. Couples are encouraged not to have more than three children, with an interval of 3 or 4 years between births. National minorities (10% of the population) are not affected by these measures.
**Fig. 6.1** Distribution of the population born between 1978 and 1993, Beijing (percentage). (Source: Figure based on the data of the 2010 population census in Beijing)
share in the total population rose from 18% to over 51% (NBS, 2012). According to the population census, there were over 19 million inhabitants in the municipality of Beijing in 2010 (19,612,368 people). Among these people the majority have a *hukou* from Beijing (64%) and for half of them their residence booklet is urban (50%) (NBS, 2012:6–9 and 308–311). If we consider only young adults born between 1978 and 1993, 58% of them are migrants (*wailai renkou*) 9 as defned by the NBS (Fig. 6.1) (NBS, 2012:172–180 and 318–335). Slightly more men (30%) than women (28%), and among these migrants the large majority (85%) are of urban origin (*chengshi*).
In the eyes of the Party-State, precarious jobs and internal migrations constitute signifcant risks of social instability.
#### **6.3 Precarious Employment as New Normality**
The transition to market economy, along with the wage-earner model and the multiplication of labor contract status, have given rise to a new organization of labor relations. In cities, it is no longer the work unit that organizes the welfare of workers, but a system of social insurance fnanced by employers, workers, and the State. This is associated with a form of social insecurity, since protections against social risks vary according to the type of labor contract. In an environment that has become
<sup>9</sup>According to the same article, late marriage and late births are strongly encouraged: 第五条 结婚 年龄,男不得早于二十二周岁,女不得早于二十周岁。晚婚晚育应予鼓励 (*di wu tiao, jiehun nianling, nanbude zao yu ershier zhousui, nü bude zao yu ershi zhousui. Wan hun wan yu ying yu guli*, Article 5, the marriage age, should not be earlier than 22 years for boys, 20 years for girls. Late marriage and late births are strongly encouraged) (HYF, 1981).
competitive, job categories have multiplied and the number of atypical jobs10 has increased to the point of becoming the norm. The analysis of the career paths of young adults born in the 1980s reveals such a trend.
Four biographical model clusters emerge from the analysis. They are named according to their main characteristics. The fndings reveal that nearly 80% of respondents hold a precarious job.11 While almost 60% of respondents are on fxedterm contracts (groups 3 and 4), and 18% of the young adults we met do not have a work contract (group 2), only 20% of respondents are fortunate enough to have a permanent job (open-ended contract) (group 1).
The analyses show that, in post-collectivist China, a large proportion of jobs created take the form of atypical employment, "meaning that they escape the form of open-ended contract which provides time insurance and signifcant social coverage" (Castel, 1995:19). By institutionalizing different categories of labor contracts instead of lifetime employment, the 1994 labor law opened the door to fexible and less costly forms of labor (Gallagher & Dong, 2011). Unlike lifetime employment, individualized labor contracts not only served the interests of companies - whether private, public or collective - and of State services, but they also contributed to breaking with the old socialist social pact by depriving urban workers of their right "to" employment and the generous welfare provisions that went with it. Until the 1994 Labor law, the differentiation of labor contracts created inequalities in the access to social insurance, since under the 1994 law only permanent contracts provided employees protection on the basis of a contribution shared between the employer and the employee. According to the law, with other forms of labor contracts, social contributions to health, accident, pension, unemployment and maternity insurance are only paid by workers (LDF, 1995).
Employment contracts established by most companies for increasingly short durations create a form of permanent insecurity. The average duration of contracts, which was about 3–5 years in the mid-1990s, has given way since the early 2000s to contracts with an average duration of 1 year (Gallagher & Dong, 2011).
Young adults are not on an equal footing when it comes to coping with job insecurity and the widespread use of precarious contracts. The analyses below show that respondents who are not from Beijing (that is to say with a *hukou* not registered in Beijing) face a higher risk of having an atypical job during their lifetime. Alongside social origin, respondents' level of education also plays a decisive role in explaining their career paths (Fig. 6.3).
<sup>10</sup>Yeung and Hu observed in an article published in 2013 that young people born between 1976 and 1981 (without making a distinction according to the type of *hukou*) married and had a child on average earlier than people born between 1946 and 1955 (Yeung & Hu, 2013). It is probably because we are not working on exactly the same birth cohorts that the results between our two studies on the transition to adulthood differ on this point.
<sup>11</sup>After evaluating the different quality indicators (ASW, PBC, HC, HG) for different group numbers, testing several clustering algorithms and visually analyzing the clusters produced, we chose the PAM algorithm in four groups (Studer, 2012).
The less educated young adults are, the more vulnerable they are to job insecurity. Findings reveal that respondents with a university degree are signifcantly less likely to be in precarious employment than those with no qualifcation, but they are not completely immune against this risk, since about 33% of respondents in precarious employment have at least a bachelor's degree (group 4, Figs. 6.2 and 6.3). However, the latter beneft from medical insurance, which is not always the case for respondents who work in atypical jobs and have a baccalaureate or bac + 2 diploma (group 3, Fig. 6.2).
**Fig. 6.2** Typology of career paths of the respondents born in the 1980s and of their inclusion in the health and accident insurance system
Insecurity, which affects a relatively heterogeneous population of young adults, tends to become permanent (Fig. 6.4). The longitudinal representation of the typology of professional paths according to the type of labor contract shows that this situation is not transitory, in the sense that fxed-term contracts would eventually lead to open-ended contracts. Workers are placed in a "temporary" situation that tends to become permanent. Precarious work is fragilizing a signifcant proportion of young adults, especially since many young people in this situation do not have the means to fnance, through the system of private insurances, a level of social protection identical to that provided by open-ended labor contracts.
#### **6.4 Young Adults' Work Expectations: From Idealism to Realism**
Faced with the fragility of career paths and the rise of precarious employment, individuals are not equal when it comes to coping with the challenges they face. Young adults working in the informal sector are indeed the most vulnerable. The level of education of these young people, most of whom are migrants, tends not to exceed general or specialized/technical middle school. They often come to Beijing to work and are not covered by a health and accident insurance scheme (Figs. 6.2 and 6.3). Social and economic vulnerabilities produced by employment insecurity can lead to the disaffliation of the young adults, since these situations occur at the end of a twofold process of dropping out in relation to work and in relation to relational insertion (Castel, 1994, p. 13). Young people working in the informal economy are not in a situation of non-employment or even in a situation of social isolation, since they develop sociabilities linked to their professional activity. However, they fnd themselves in a "zone of vulnerability". In other words, in "a social space of instability and turbulence" (*Ibid.*, p. 16). Their relationship with work and their integration into relationships being fragile, they run the risk of falling into disaffliation (*Ibid.*).
This is the case for Suzhi, a 35-year-old single mother from a village in the Hebei province. She dropped out of middle school. After a few years spent in her village, she then migrated to the Chinese capital at the age of 25. After a brief training course lasting a few months, she was employed - without a labor contract - by a massage parlor, where she was still working at the time of the interview. She declared that she did not feel integrated in Beijing, or even despised by Beijingers (inhabitants with a Beijing *hukou*). Her only friends were her work colleagues, who were also migrants, but she did not consider them as her confdantes. Her mother plays this role. She was the only person aware of her out-of-wedlock pregnancy. She supported her from a distance to get through this "ordeal", because in China out-ofwedlock births remain stigmatized.
The vulnerability expressed in Suzhi's discourse is echoed by many of the respondents who came to Beijing to study and/or to work. These adults have not managed to fnd in Beijing a job that offers them the stability they were looking for.
If they have a labor contract, it is for a fxed term. This leaves a particularly bitter taste for Wei and Lixin, who are of urban origin. Wei was born in 1981 in Haerbin, in the north-eastern province of Heilongjiang. After graduating with a bachelor's degree from the University of Haerbin in her home province, she began working part-time on a fxed-term contract. For 4 years she worked as a teacher in a private school. At the age of 27, she decided to leave her hometown and her job to "go discover Beijing" and work there. She worked part-time for a private school for 4 years, without a labor contract. She took the opportunity to enroll in a master's program at a university in Beijing. At the age of 30, she graduated. Her degree enables her to teach Chinese to foreigners. Thanks to this diploma, she found a fxed-term job in a language school in Beijing, where she was still teaching at the time of the interview.
**Fig. 6.4** Longitudinal representation of the typology of professional trajectories of young adults born between 1980 and 1985 according to the type of labor contract
She was then very angry at the permanent insecurity that her professional situation brought:
*Relationships in Beijing are not fair* […) *I am here because my income is higher here than in Haerbin. I am here for work* […), *but despite my degree to teach Chinese to foreign students, my income does not allow me to rent an apartment in Beijing on my own*.
She therefore lives, not by choice, but out of necessity, in a shared apartment. This gives her a huge sense of frustration, as her income does not allow her to live the adult life she aspires to. Since she started University, she has also often felt alone. She thinks it is because she does not have many friends and because none of them live in Beijing. She is worried because she no longer wishes to confde her worries to her elderly parents, so as not to worry them. As she puts it, she fnds herself " *alone with her worries* " (*wo ziji chiku*).
As for Lixin, he was so disappointed by his professional experience in the capital that he went back to work in his home province, Zhejiang. This young man, born in 1983, is an only child because his mother's second pregnancy was forcibly terminated by the family planning. A brilliant student, as soon as he passed his high school diploma,12 he was admitted to a prestigious University in Beijing to study foreign languages. This degree, along with some work experience in the United States, were not enough to overcome the discrimination he experienced and his professional insecurity. For this reason, at the time of our meeting (2013–2014), he had decided to return to his home province to open a consulting agency.
The risk of relational disaffliation for young adults in atypical jobs is heightened by the long working hours, which leaves them little free time and few opportunities to socialize. The simultaneous longitudinal analysis of respondents' career paths and working hours indicates that the more unstable their professional situation is, the more likely they are to exceed the statutory working hours, which are set at 8 h a day and 44 h a week, with at least 1 day off a week (Constantin, 2016; LDF, 1995, art.36–42 §4).
Precarious employment makes young adults vulnerable not only economically, but also socially. Research based on data dating from before 2008 indicates that only 50% of companies (state and non-state) had signed a work contract with their employees in 2007; and among them, only 20% of private companies had done so. Of the contracts signed, 60–70% were fxed-term contracts of less than a year (Friedman & Lee, 2010:509). The 2008 labor law is an attempt to remedy the growing precariousness of jobs. The provision relating to the employment contract aims, in particular, to address the problem of informal work. One of the articles aims to curb situations of lasting precariousness by limiting the chaining of fxed-term contracts. For the same position within the same company, the law stipulates that after two consecutive fxed-term contracts, the employee must be offered a permanent contract. To terminate a contract of employment, the employer is now required to justify the reasons (LDF, 2008). This legislation, intended as a tool for social peace by President Hu Jintao and his Prime Minister Wen Jiabao, unfortunately arrived at the wrong time. It coincided with the 2008 economic crisis and it has so far been poorly enforced.
#### **Bibliography**
<sup>12</sup>After evaluating the different quality indicators (ASW, PBC, HC, HG) for different group numbers, testing several clustering algorithms and visually analyzing the clusters produced, we chose the Ward algorithm in three groups (Studer, 2012).
**Open Access** This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.
The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.
# **Part III The Rise of Neo-familialism**
In Confucian thought, the family holds a privileged position as the center of the solidarity system. It is considered as "an extension of the individual" (Cheng, 1997). The concept "family" (*jiating*) then referred to a group much larger than the parentchild relationship. It included all members of the family clan (*jiazu*) gathered in networks and at the center of which is the individual (Fei, 1992). In this type of vertical family organization, control was exercised by the patriarch over other members of the family.
Recent work by Chinese sociologists of the family highlight changes in meaning taken by the concept. They describe this concept as "fexible" (*shensuo*) because in post-collectivist China the family tends to follow the model of the nuclear family. The family unit, which was previously an economic unit, is becoming more elective (Ma, 1999; Shen, 2013). Ethnographic research conducted by Yan Yunxiang has shown that, in contemporary China, individuals no longer systematically place the interests of the family group above their individual interests (Yan, 2003a, b). As a result, researchers are observing a shift from vertical patriarchal family organizations (with the ancestors at the centre) towards more horizontal organizations (with conjugality at the centre). Thus, in China as elsewhere, under the effect of individualization, the meaning of the family is changing. The family is becoming relational (Beck & Beck-Gernsheim, 1995; Evans, 2008; Singly (de), 2012). Families are taking the form of units of feelings and affection, values that have partially dethroned those of obligation around which family solidarities were articulated (Friedman, 2006; Gonçalo & Harrell, 2017; Hansen & Svarverud, 2010; Yan, 2009).
In my doctoral thesis, I described these transformations as part of a form of neofamilialism (Constantin, 2017). I then refected on the concept of familialism developed in the feld of social policy. This concept makes it possible to understand the structure of social rights by revealing whether they are familialized or individualized (Lister, 1990). Building on this concept, the concepts of "familialization" and "defamilialization" emerged to account for State injunctions made to families, and more particularly to women, to take on the tasks of *care* (Hantrais, 2004; Leitner, 2003; Ostner, 2004; Saraceno, 2016) 1 . For example, the childcare schemes set up in China during the collectivist period refer to the concept of defamilialization. Childcare facilities were set up to relieve families of this workload to enable them to work within work units. Later, the dismantling of this social policy has participated to a form of "familialization" since families have been urged to fnd private solutions for looking after preschool children. These solutions may take the form of a diminution in the activity rate of one of the parents (usually the mother), the solicitation of grandparents (usually grandmothers), or the involvement of third parties. As this form of familialism is no longer based on the same principles as in the past, it seems to me that the concept neo-familialism allows to account for this change and for the new value system around which it is articulated, as it will be unfolded in this third and fnal part.
#### **Bibliography**
<sup>1</sup>After evaluating the different quality indicators (ASW, PBC, HC, HG) for different group numbers, testing several clustering algorithms and visually analyzing the clusters produced, we chose the PAM algorithm in three groups (Studer, 2012).
# **Chapter 7 The Postponement of Family Formation Due to Employment Instability**
#### **Contents**
**Abstract** This chapter reveals that the lengthening of the period of schooling, for the 1980s birth-cohort, delayed their entry into the workforce and fnancial autonomy. Job-seeking is likened to a "rat race", emphasizing the importance of prestigious University degrees to fnd a job. Young adults face more and more challenges fnding stable employment, with many experiencing job instabilities. Unemployment rates rise, particularly impacting those with rural *hukou* and lower educational achievements.
Prolonged schooling and delayed stable employment lead to postponed transitions to marriage and parenthood for the 1980s birth-cohort compared to the 1950s. These shifts often create tensions and misunderstandings with parents who grew up and who were socialized in a very different social context.
This chapter examines in-depth family trajectories and depicts the rise of independent lifestyles nestled between flial and conjugal ties. During this specifc time lapse, young adults have jobs but are not yet burdened by family responsibilities. Findings reveal that while the housing market leans towards buying, gaining homeownership proves more and more challenging, particularly for those in precarious employment.
**Keywords** Youth unemployment · Instability · Precarity · Housing · Middle class · Family
#### **7.1 A Lengthier Path Toward Family Formation**
As revealed in Chap. 5, the extension of the time spent at school in the life calendar of young people born in the 1980s has induced a delay in the average age at frst job, and therefore to fnancial autonomy (Fig. 5.1). In post-collectivist China, looking for a job can be compared to a rat race according to some respondents. There is a growing sense amongst them that if they do not work hard to get a diploma, they will be ousted by the very competitive job market. Degrees from prestigious Universities constitute a comparative advantage to fnd a job and a job that matches with their expectations. But more investments from families in the education of young people does not mean the latter will not have to struggle to fnd a stable employment. Some young adults coming from cities outside Beijing can count on their fngers how many times they had to change jobs (see Chap. 6). If there are many jobs available, a large number of them are precarious (short term contracts, part-time, low paid). As shown by earlier studies (Lian, 2009, 2010; Wang & Guo, 2012), the data collected indicate that the unemployment rate is rising. Unemployment affects categories of workers differently. All other things being equal, this phenomenon mainly impacts young people who have a rural *hukou* at the time the research was conducted. This social group tends to have a lower school achievement than urban youths.
Longer school trajectories and delay in the age at frst stable employment induce later transition to marriage and parenthood. The inter-cohort comparative analysis of transition rates to frst marriage and birth of a frst child confrms the delay in the age of entry into the roles of spouses and parents for young people born in the 1980s (Fig. 5.1). Whereas 87% of respondents born in the 1950s were married before the age of 28 and 70% had a child before this age (Figs. 7.1 and 7.2), at the same age
**Fig. 7.1** Percentage of respondents born between 1950 and 1959 married before the age of 28-year-old
**Fig. 7.2** Percentage of respondents born between 1950 and 1959 who became parents before the age of 28-year-old
only 42% of urban respondents born between 1980 and 1985 were married and 18% had had a child (Figs. 7.4 and 7.5).
It is worth recalling that the 1950s birth cohort's path to adulthood was disrupted by the Cultural revolution and the Sent-down movement of urban youth to the countryside. These sociohistorical changes had the effect of delaying the stages of their transition to adulthood. For instance, amongst the *zhiqing* (sent-down youth) many waited for the end of the movement and their defnitive return to the city to get married and have a child because they feared that getting married in the countryside, with a rural spouse, could mark a break with their urban origins and deter their chances of returning to the city (Chen, 1999; Lin, 2013; Meng & Gregory, 2002; Zhou & Hou, 1999; Bonnin, 2004).
The birth control campaign "*wanxishao* (晚稀少)" implemented in 1973 may also explain the delay in the age of frst marriage and frst childbirth of the members of this birth cohort, since it urged individuals to late marriage and procreation, spacing between births and few births. In cities, where the policy was the strictest, women were encouraged not to marry before 25 and men before 28. Couples were also enjoined not to have more than two children, with a three-to-four-year interval between births.1 Out-of-wedlock births were not socially accepted at this time in China (and still largely today).
<sup>1</sup> In the sample:
<sup>22.6%</sup> of respondents have had a car at least once in their life (21.72% of women and 23.38% of men).
Among all young adults of urban origin, 28.77% have had a car at least once in their life (28.84% of women and 28.7% of men).
Among all young adults of rural origin, 7.34% have had a car at least once in their life (1.33% of women and 11.76% of men).
The analyses reveal that 97% of female respondents were married at 27-year-old, while 78% of male respondents were married at the exact same age (Fig. 7.1). After 27-year-old, nearly 30% of men are married. If women born in the 1950s mainly get married between 22 and 26- year-old, men mainly take the step between 24 and 28- year-old.
Amongst this birth cohort, 82% of female respondents had become mothers at 27-year-old, and 57% of male had become fathers at this age (Fig. 7.3). It is mainly between 22 and 28- year-old that women reach motherhood and between 24 and 29-year-old for men. Women are signifcantly more likely than men to be married and have a child at/or before 27. This is especially true if their level of education does not exceed middle school (low level of education, Fig. 7.3).
The abandonment of the " *wanxishao* " campaign at the end of the 1970s and the marriage law of January frst 1981 according to which the legal age for marriage is 20 for women and 22 for men (article 5 of the marriage law)2 could have led to an advancement in the age of marriage and parenthood for the birth cohort born in the 1980s. However, the data reveals a delay in the average age towards these transitions. If we consider the paths of the respondents up to the age of 27, we observe that the average age at frst marriage has moved from 23–24-year-old for urban women born in the 1950s to 25–26-year-old for city dwellers born in the 1980s (Fig. 5.1). Urban women born in the 1950s gave birth to their frst child on average 1 year after their marriage. Whereas young women born in the 1980s tend to give birth to their frst child in the same year as their marriage.
Just over 50% of city women and 34% of city men are married at the age of 27 (Fig. 7.4). They are 23% to have become mothers and 14% to have become fathers (Fig. 7.5). At the same age, 63% of rural respondents were married and 50% had had a child. Rural women are nearly 70% to be married at 27 and 57% have had a child. Rural men are 58% to be married and 44% to have become fathers at this age (Figs. 7.4 and 7.5).
The logistic regression models indicate that amongst the post-1980 birth cohort, it is the women, the rural respondents, or those with a low level of education who have signifcantly more chance of being married or having a child before the age of 28 (Fig. 7.3). Because they married later, young adults born in the 1980s and of urban origin had been married for a shorter time than their peers from post-1950 at the same age.3 However, young rural people born in the 1980s, who married before this age, tended to do so slightly earlier than the cohort born in the 1950s (Fig. 5.1).
These results indicate that transitions to marriage and parenthood occured later over the life-course of young adults. If the timing of the family trajectories of young urban people differs from that of city dwellers born in the 1950s, that of young rural people tends to approach it. The delay in the age at frst marriage and frst childbirth
<sup>2</sup>Expression frequently used by respondents to express this idea.
<sup>3</sup>N = 571 because only those who answered the question are considered in the analyses.
**Fig. 7.3** Logistic regression models (*odds ratio*) on the risk for respondents to be married and have a child before 28-year-old (1950–1959 and 1980–1985)
cannot, however, be explained by the disaffection of the family institution. Indeed, Ji and Yeung's research shows that contrary to what has been observed in Western countries and in several Southeast Asian countries, the marriage rate in China is not declining (Jones & Yeung, 2014; Shanahan, 2000). Marriage and births within marriage remain a prevalent social phenomenon.
**Fig. 7.4** Percentage of respondents born between 1980 and 1985 married before the age of 28-year-old
**Fig. 7.5** Percentage of respondents born between 1980 and 1985 who became parents before the age of 28-year-old
The rise of uncertainties in Chinese society over the past 20 years partly explains the delay in the age of marriage and parenthood. Decollectivization, by "emancipating" the economy and individuals from the tutelage of the socialist state to some extent, has created new opportunities and an increase in income for some; and a degradation of living conditions for others. With the disintegration of the socialist welfare state, individuals had no choice but to internalize the negative externalities of economic liberalization; namely, the liberalization of the labor market, the liberalization of prices in the health, education, housing, and food sectors (sectors which were previously taken care of by the collective). Access to these resources has become highly competitive over the years and is now the responsibility of individuals. It is in this context that more and more young people born in the 1980s are extending their studies and consequently the stages of their transitions to adulthood. They are delaying certain transitions, such as transitions to the roles of spouses and parents. Moreover, for many respondents, the formation of the family is conditioned on the prior achievement of economic independence, a transition that also comes later in the life-course of young cohorts due to the average lengthening of study duration. The delay in the age of marriage and parenthood can be a source of misunderstandings and tensions within families. The parents of young adults, who became adults in Maoist society, were subject to different social expectations. Under pressure from their parents, some young people – who are facing new realities – try
to meet their parents' expectations, but many cannot and are forced to forge new paths to adulthood.
#### **7.2 From Filial to Conjugal Ties**
The young adults' family trajectories reveal the existence of independent and exploratory lifestyles slipping between flial and conjugal ties. This social phenomenon has been described by researchers as "the rice bowl of youth" because they have jobs and are not yet burdened by family responsibilities (Zhang, 2000). Young adults take advantage of this time to assert themselves as actors and actresses of Chinese "modernity". Young people from rural areas use this time as well to migrate to the city and "see the country" (Chang, 2008; Pun, 2005). Migrations, which are often not only motivated by economic reasons for this birth-cohort, offer young women an opportunity to escape family obligations, parental control and to acquire autonomy, leading to a delay in the age of marriage (Hershatter, 2007; Murphy, 2002).
Multichannel sequence analyses (MCSA) reveal that marriage still plays a central role in the lives of young adults, despite a relatively long single life (Fig. 7.6) 4 .
The fndings uncover that respondents who have completed short studies, not exceeding high school (groups 1 "Precarious employment and "early" marriage" and group 4 "Informal employment and "early" marriage" Figs. 7.6 and 7.7), tend to marry earlier than those who have completed higher education – Bac + 2 and beyond (groups 2 "Stable employment and "late" marriage" and group 3 "Precarious employment and "late" marriage" Figs. 7.6 and 7.7). If the extension of the duration of schooling plays a role in the delay of the age at frst marriage, the precarization of professional trajectories also seems to exert an infuence. It does not indeed appear easy for young urban and rural people from outside Beijing to reach this stage when they have atypical labor contracts.
Respondents with higher educational background and who have a precarious employment contract (Group 3 "Precarious employment and "late" marriage" Fig. 7.6) tend to marry later than young adults who have the same level of education, but who have a permanent employment contract (Group 2 "Stable employment and "late" marriage" Fig. 7.6). Xiaocui's situation, for instance, is emblematic of these trade-offs. She met her future husband in 2012 and explains that, for fnancial reasons, they have postponed the date of their wedding several times. They both have a bachelor's degree from a prestigious university in Beijing and have been working in the city for several years. As their parents were farmers, their income alone did not allow them to organize a wedding that met their "expectations" (born in 1986, rural, Heilongjiang *hukou*, Bachelor). As we shall see in the next chapter, these expectations are complex in the sense that they are at the crossroads of
<sup>4</sup>Comparing the education level of the partners' parents-in-law would also be useful to complete the analyses conducted. Unfortunately, this is not possible because I only have the education level of the respondents' parents.
**Fig. 7.6** Cross-sectional representation of the typology of professional and marital paths of young adults born between 1980 and 1985 (MCSA)
different norms stemming from social expectations dating from different eras (pre-Maoist, Maoist, post-collectivist). They wanted to celebrate their engagement in Beijing a year before the wedding, so that their parents could meet. As they each
**Fig. 7.7** Logistic regressions (*odds ratio*) – Professional and marital paths of young adults born between 1980 and 1985
lived in shared accommodation in cramped housing, it was impossible for them to accommodate them at home. Her future husband's parents agreed to pay for their trip from Shandong province. This was not the case with Xiaocui's parents, who live in Heilongjiang province. They announced that they "did not have time to come to Beijing". She was very saddened by this, as she wanted to oversee every stage of her wedding so that it would be "perfect". She sadly told me that, not having the means to offer her parents the trip, they decided to cancel their engagement and get married straight away. They were subsequently forced to postpone the date of their wedding again because they were not fnancially able to organize a reception that suited their wishes. They wanted to take souvenir photos with a photographer in Beijing, organize a frst reception in Xiaocui's village with family and neighbors, a second in her husband's village with family and neighbors, a third in Beijing with their friends and colleagues, and go on a honeymoon. They waited another year before getting married to save money. In the end, however, they had to forego the reception in Beijing (born in 1986, rural, Heilongjiang *hukou*, Bachelor).
#### **7.3 The Quest for a Place to Call Home**
A trend has been observed in European countries (especially in southern Europe) and North America for young adults to return to their parents' home despite their entry in the labor market. This is due to the development of fexible and precarious jobs that do not allow young adults to meet their needs without the support of their parents (Settersten & Ray, 2010; Van de Velde, 2008). Multichannel sequence analyses do not reveal such a trend in China. Whether young adults have a permanent contract, have no employment contract or have an atypical contract, they tend to live
in shared accommodation before living alone or as a couple (Fig. 7.8).5 Even in the third group "Precarious employment and cohabitation", including mainly Beijingers, the analyses do not show back-and-forth trajectories between the parental home and residential autonomy.
**Fig. 7.8** Longitudinal representation of the typology of professional and residential paths of young adults born between 1980 and 1985 (MCSA)
<sup>5</sup>During the interviews, I asked all the respondents what they think about cohabitation, but I did not ask them to specify their views on people who practice cohabitation.
Although the housing market is more oriented towards buying rather than renting, and young adults do not receive support from public institutions to help them become self-suffcient, there is no trend towards prolonging family cohabitation, even for young people who fnd themselves in the most precarious situations (Group 2 "Informal employment and cohabitation" and Group 3 "Precarious employment and cohabitation" Fig. 7.8). In our view, this does not mean that family solidarity no longer exists, but rather that it has evolved, changed and adapted to the new context of post-collectivist society.
Although young adults achieve residential autonomy at an early age, it is not easy for them to gain access to home ownership, which in the collective imagination is an essential element of identifcation with the middle class. The clusters obtained through the Optimal Matching (OMA) method6 (Fig. 7.8) and the transition rates to home ownership calculated for each group (Fig. 7.10) indicate that it is even more diffcult for precarious workers to become homeowners. By the age of 27, just over 15% of young adults working informally and slightly more than 20% of those with atypical labor contracts had become homeowners (Fig. 7.10). In addition, about 25% of permanent contract employees were able to buy a home at the same age, even though they entered the labor market later (Fig. 7.10). Higher education beyond high school protects against the risk of working in the informal economy, but it does not necessarily shield young adults from job insecurity7 (Fig. 7.2, group 3 "Precarious Employment" and Fig. 7.9).
**Fig. 7.9** Logistic regressions (*odds ratio*) – professional paths of young adults born between 1980 and 1985 according to the type of employment contract
<sup>6</sup>For example: http://sh.eastday.com/m/20141008/u1ai8378295.html; http://emotion.pclady.com. cn/133/1336199\_all.html; http://www.shuolianai.com/thread-10495-1-1.html
<sup>7</sup> "A couple shall go through marriage registration if it has not done so" (HYF, 1981).
**Fig. 7.10** Transition rates to homeownership of young adults born between 1980 and 1985 according to the typology above (Fig. 7.8)
The creation of a new lifestyle, in other words, practices and/or representations associated with a specifc social group, through the media, offcial discourse, and public policies creates a strong tension among young adults between their high expectations and the reality of their economic situation. Young people born in the 1980s were led to believe that they were lucky and privileged to have grown up in a China undergoing rapid economic growth. Now, as adults, they fnd themselves alone in facing ruthless competition and struggling to be part of the new middle class, symbol of the "society of small prosperity" sought by the Chinese government. Tired of this new "cult of performance," the young adults I met denounced the competition, the cult of money, corruption, the decline in values or morality in the society and the "illusion of a better life they were lulled into" (Chicharro, 2010; Ehrenberg, 2010). They do not, however, give up on wanting to be part of the middle class, whose membership is socially valued, as its members are represented as welleducated (*wenhua*), civilized (*wenming*) and of quality (*suzhi*) (Rocca, 2016:6). 45% of the young adults I met identify with the middle class.8 All other things being equal, logistic regressions indicate that, among all respondents, the urban population in general is more likely than the rural population to identify with this social group, and that women are more likely than men to declare that they belong to this
<sup>8</sup> "Marriage is a major event in life, it needs to be taken very seriously, *hunyin shi rensheng dashi, xuyao hen shenzhong"*.
**Fig. 7.11** Logistic regressions (*odds ratio*) on the feeling of social belonging of the respondents born between 1980 and 1985
group. This sense of belonging is also signifcantly correlated with the level of education of young adults: The higher their level of education is, the more likely they are to identify with the middle class (Fig. 7.11).
According to research on the Chinese middle class, owning an individual home in a secure compound and a car have become inseparable elements of the feeling of belonging to the middle class (Davis, 2000; Li, 2013). The analyses confrm the very strong link that exists in individuals' subjectivity between "owning one's home" and "identifcation with the middle class". Young adults who do not own their home signifcantly tend to have a feeling of belonging to the working class (Fig. 7.11). However, the data does not establish a signifcant correlation between "identifying with the middle class" and "owning a car" (Fig. 7.11). It should be noted, however, that among all young adults from urban areas, 28% have owned a car at least once in their lives, while only 7% of young adults from rural areas had done so.9
<sup>9</sup> "Getting married and having children is a matter of course, *jiehun shenghaizi shi tianjingdiyi de shiqing*".
#### **Bibliography**
**Open Access** This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.
The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.
# **Chapter 8 Young Adults' Aspiration for Intimacy in Post-collectivist China**
#### **Contents**
**Abstract** This chapter delves into the dynamics of marital choices in contemporary China. The chapter presents contrasting responses between men and women regarding parental infuence. While men often assert independence, women frequently seek parental advice in their marital decisions. Notably, several women employ strategic approaches to circumvent parental disagreement, refecting a delicate balance between individual autonomy and familial harmony. The analysis unveils a complex interplay of societal norms and aspirations, indicating that, despite legal emphasis on free choice, familial considerations continue to shape marital decisions.
Additionally, the chapter explores social homogamy and hypergamy trends, revealing gendered patterns in the importance placed on education, income, and family background in partner selection.
Furthermore, the fndings shed light on a growing form of intimacy in post-collectivist China: cohabitation before marriage. Encountering social resistance, particularly from older generations, a new concept surfaced: "trial marriage" (*shihun*). This strategic term emerged among young adults seeking to legitimize living together before formalizing the union. It bridges the gap between societal norms and evolving practices.
The fndings reveal a complex interplay between societal expectations, individual choices, and the enduring infuence of Confucian values in shaping attitudes towards cohabitation and marriage in contemporary China.
**Keywords** Intimacy · Marital choice · Parents infuence · Gender · Confucian values · Cohabitation · *Shihun*
#### **8.1 From Free Choice to Social Homogamy**
On May 1, 1950, the government of the People's Republic of China announced the abolition of all laws supporting arranged and arbitrary marriages by implementing the marriage contract based on individual free choice (Croll, 1981). This law was replaced on January 1, 1981 by a new law on marriage (HYF, 1981). This law stipulates that marriage is the foundation of communal life (art.8):
Article 8: Both the man and the woman desiring to contract a marriage shall register in person with the marriage registration offce. If the proposed marriage is found to conform with the provisions of this Law, the couple shall be allowed to register and issued marriage certifcates. The husband-and-wife relationship shall be established as soon as they obtain the marriage certifcates. A couple shall go through marriage registration if it has not done so.
The law also stipulates the right to free choice of marriage (art.5) and the prohibition of forced and arranged marriages (art.3). It mentions the equal place of spouses within the couple (art.2). It indicates that the spouses are free to choose to use their own surname (art.14) and that the children can bear either the surname of their mother or that of their father (art.22). Once the marriage has been registered, depending on the wishes of both parties, the wife or the husband can become a member of his or her in-laws (art.9). Through these provisions, marriage is no longer a matter between two families (as it was the case in pre-Maoist China). It becomes an affair between spouses, who have autonomous rights enabling them to take care of their marital affairs without interference from third parties. By giving spouses control of marriage negotiations, the marriage law has given marriage a new signifcance for the spouses and helped to spread new family values.
The free choice of a spouse enables women to become active players in the formation of their couple. 58% of respondents born between 1980 and 1985 declared that their parents could not infuence them in their choice of spouse. Among respondents born in the 1950s, i.e. after the enactment of the marriage law, 59% said they had not followed their parents' infuence in choosing the person with whom to build a family. It is particularly interesting to note that, in both birth cohorts, it is the men who are the least infuenced by their parents. 63% of men born in the 1980s and in the 1950s claim not to be infuenced by their parents in choosing a spouse. In contrast, 52% of women born in the 1980s and 55% of women born in the 1950s where able to overcome the infuence of their parents. Overall, men seem to be more emancipated from parental infuence than women. This is even more acute in the responses of young adults from rural backgrounds: 73% of men and 55% of women born between 1980 and 1985 say that they do not allow themselves to be infuenced by their parents in a matter as intimate as the formation of their couple. The redefnition of marriage by the Party-State and the direct intervention of the government in the family sphere has contributed to weakening parental infuence and opening a space for the rise of intimacy.
In addition, the radical change in the organization of society during the Maoist period also contributed to transforming the social organization of families. From the mid-1950s onwards, the loss by these families of their land, industrial and real estate properties, as well as the questioning of the cult of ancestors, directly weakened the basis on which patriarchal authority rested. Production was organized by collectives, the work unit in town and the commune in the countryside, which took the place of parents in assigning individuals a specifc job to do. There are three reasons that could explain why parental infuence is weaker in the countryside. Firstly, since the opening-up to the market economy and the importation of new production techniques, the knowledge of parents has become relatively outdated in the eyes of their children. Secondly, young adults, and especially sons, have been able to access a higher level of education than their parents. This has given them the legitimacy to participate in the social life of the village (Yan, 2009:76). Thirdly, the experience of migration to urban centers has enlarged opportunities for young people to socialize with people of the opposite sex. This has had the effect of opening up their range of options when it comes to choosing a partner. Enriched by these experiences, young rural men have gained in agency. This makes them feel more empowered to take charge of their own lives and question parental authority.
It is however signifcant to observe that just over half of the women born in the 1980s declare not wanting or not having followed the recommendations of their parents in terms of marital choice. Qualitative interviews reveal that, regardless of the level of education, it is important for the respondents to discuss the choice of their spouse with their parents. Yet, the analyses reveal that men are more likely to tell their parents about their choices (after they made up their decision on their own), while women are more likely to ask their parents for advice. The responses of Niu and Zhiqiang are emblematic of the corpus of interviews:
Niu was born in 1982 in a city in Hebei province. He is engaged. He declared:
*"My parents could not infuence my choice, because it's my business. My parents ask me questions about my private life, but they do not want, they do not seek, to infuence my choices on the choice of a girlfriend"* (born in 1982, urban, Hebei *hukou*, specialized college).
Zhiqiang was also born in 1982. He grew up in a village in Henan. He explains to us:
*"Generally, my parents do not infuence my choices. Maybe because they live far away. I inform them. I tell them: I intend to do this. I intend to do that"* (born in 1982, rural, Henan *hukou*, Bachelor).
On the other hand, the women's responses are much more nuanced. I have selected four emblematic answers:
Min was born in 1981. She is from a city in Shanxi province. She declared:
"*I discussed my choice of spouse with my parents. But they could not have infuenced my choice in the sense of forcing me to change my mind. [...] Especially for a girl, it is important to have the consent of her parents, because if the parents appreciate our spouse, it facilitates family relations. Lovers want to share their joy with their parents*" (born in 1981, urban, Shanxi *hukou*, Bachelor).
Lina, who was born in 1993 in a village in Jiangsu province, explained that "*parents may not support a marriage, but they cannot prevent it.*" She elaborates by recounting the strategy devised by her cousin, whose parents were opposed to her union with a young man whose fnancial means were very modest.
"*My cousin wanted to marry a man who had neither a house nor a good situation. They had a son who is now one year old. They got married in December 2013. My cousin's parents are not happy, but they could not change the situation by opposing the marriage as my cousin and her spouse already had a child. Her parents had to accept the reality*" (born in 1993, rural, Jiangsu *hukou*, Bachelor).
According to the marriage law, her cousin could have married without her parents' consent. However, this was so important to her to legitimize her union, that she chose to force their hand by having a child out of wedlock. This deviation from the norm being even more unbearable to her parents than letting their daughter marry a penniless young man, they fnally, reluctantly gave their consent to their union.
Without going so far, Shasha, who was born in 1986 in Tianjin, also developed a strategy to circumvent her parents' disagreement and to put all the chances on her side for them to accept her love relationship:
"*I know that my parents do not agree that I marry my boyfriend who is Russian and who is more than two years younger than me. So that they do not try to infuence my choice, I chose not to tell them that I have a boyfriend. I will only introduce him to them when he has a stable situation*" (born in 1986, urban, Tianjin *hukou*, Bachelor).
In other words, when he would have completed his master's in international relations at the People's University and found a job.
For Wei too, her parents' consent is very important. She was born in 1981 in the capital of Heilongjiang province. While her parents have never forced her to do anything, she thinks her parents could infuence her choice of spouse. She gives me the example of her last three serious love stories. Each time, she discussed her choices with her parents. During her last love relationship, it was they who suggested she end this relationship "*because the young man was not honest*". He wanted her to help him take advantage of naive migrants and make money for himself (born in 1981, urban, Heilongjiang *hukou*, Master).
All respondents are aware that their parents could not impose a spouse on them or legally prevent them from getting married. However, while the analyses show that decisions about who to marry are becoming more individualized, this remains limited. In post-collectivist China, "free" marital choices are still thought out in terms of family norms and aspirations, especially for young women who are looking for a compromise that will maintain family harmony.
The marriage law and its emphasis on the free choice of spouse could be a factor encouraging social heterogamy. However, in practice, marital choices remain dependent on the socio-economic status of the future spouses. If for all respondents physical and moral characteristics are important, young adults born before 1985, who were not yet married when I met them, also attached importance to the future partner's socio-economic situation. When I asked them what they consider important in the choice of a spouse, men were more likely to mention "family background". While women mentioned "income", "education level" and "job held". Men who had studied at university would not envisage building a family with a woman who would not be from the "same background" (*mendanghudui*).1 Tingting, who graduated from Beijing University, particularly thinks that "*to meet someone you love, you must have common interests*". He continues: "*I don't think I could meet someone who is poorly educated*". By "poorly educated", he implies a person who would not have attended an elite University. Like other respondents, his point of view is closer to that of his grandparents than to that of his parents:
"*If I decided to marry someone who does not belong to my social background, I think my grandparents could be an obstacle because they are more traditional. For them, the whole family must be well matched*" (born in 1993, urban, Shanghai *hukou*, Bachelor).
The quantitative data reveal a high degree of social homogamy. Young urban men with at least a bachelor's degree are all in a relationship or married to young women who have attended university. Urban men have also signifcantly stated that before their marriage, there was no economic difference between their family and that of their spouse (Fig. 8.1). Young adults' frequentation of places specifc to their social background increases the probability that they meet partners who belong to the same milieu as them. Indeed, romantic encounters are often the result of a social process through which individuals belonging to the same background are placed in a position to meet (Bourdieu, 2007; Girard, 2012).
The analyses show, alternately, a signifcant trend towards social hypergamy among young women and respondents of rural origin. Their spouse's family had a higher economic level than their family before their marriage (Fig. 8.1). Young women from rural backgrounds, who have an education that does not exceed high school level, as well as young women from urban backgrounds, who have an education level that does not exceed a two-year degree after baccalaureate, are more likely to marry or be in a relationship with someone who has a level of education higher than theirs. The level of education is also an important criterion for Wei when choosing a partner, because from her point of view:
*"a man who has not studied at university has a very low level of "quality". He knows little. He does not have the ability to understand me, to understand my choices". She adds: "income is a very, very important element. It is indeed linked to the level of education and*
<sup>1</sup>This practice is common in China as shown by Sun Peidong in his Chinese-language work "Who will come to ask for my daughter's hand in marriage? (谁来娶我的女儿?, *shei lai qu wo de nü er?*)" (Sun, 2013).
**Fig. 8.1** Logistic regressions (*odds ratio*) on the economic situation of the spouse's family compared to that of the respondent's family among the 1980–1985 birth cohort (Male child)
*the job held. Income is important to me because my income is not very high. So if we want a child, it is obvious that income is very important. If his income was identical to mine, it is obvious that I would not want to marry because in no way could we afford to build a family. The sector of activity is important to me because it is linked to income and the level of "quality". For example, if a man is a driver, it is certain that his level of education is not high"* (born in 1981, urban, Heilongjiang *hukou*, Master).
For Xiaocui, who has a university education but comes from a rural background, the level of education is also an important criterion in the choice of a spouse. During the various interviews, we discussed how she met her current husband. She explained:
*"My circle of relationships is narrow; they are mostly my colleagues who are almost all women. So, I chose to use a dating site on the Internet to fnd my future husband [...]. My selection criteria were: appearance (xiangmao), the level of education (xueli), income (shouru) and the level of quality (suzhi). [...] I was looking for a boyfriend who had at least the same social and economic status as me (mendanghudui)"* (born in 1986, rural, Heilongjiang *hukou*, Bachelor).
Like her, her spouse has a bachelor's degree. In other interviews, when we were discussing together about her spouse, Xiaocui indicated that the economic situation of the family of her partner was and still is better than that of her family. During the discussion, she declared with a smile: "*I am too pragmatic! (Wo tai xianshi*!)".
These examples illustrate the importance of considering not only the level of education of the respondents and their spouses but also the socio-economic situation of both families.2 While women are more likely than men to marry or be in a relationship with people who have a level of education higher than theirs, this pattern is not absent from men's trajectories. Given the imbalance in the sex ratio at birth in favor of men, it remains particularly diffcult for men with a very low level of education to fnd a spouse (Attané et al., 2013; Ji & Yeung, 2014).
#### **8.2** *Shihun:* **An Understatement for Premarital Cohabitation**
Cohabitation (*tongju*) before marriage is uncontestably a new form of intimacy that is gradually developing in post-collectivist China. However, for the time being it remains little accepted socially. Cohabitation before marriage, like almost all changes affecting the customary practice of marriage, initially provokes disapproval from many parents, as well as conficts between generations. At a later stage, it will probably become socially accepted. At the time of the feldwork, this was a very thorny issue for the respondents and in the news. Among the 660 young adults who participated in this research (quantitative and qualitative interviews), only a minority had lived in cohabitation with a partner without being married (in yellow on Fig. 8.2). When I tried to determine the profle of these respondents, it turned out that no explanatory criterion was statistically signifcant.
**Fig. 8.2** Cross-sectional representation of the marital trajectories of young adults born between 1980 and 1985
<sup>2</sup> "There are three unflial acts, the greatest of which is to have no offspring, *bu xiao you san, wu hou wei da*". http://baike.baidu.com/item/不孝有三,无后为大. Consulted on March 23, 2017.
While quantitative longitudinal data reveal the emergence of this new social norm, we need to go back to respondents' life stories to grasp its complexity. It is particularly interesting to observe that all young adults born in the 1990s see no problem in cohabiting before marriage, while those born in the 1980s are a little more reluctant.3 The latter agree to cohabiting before marriage under one condition: it must be a prerequisite to marriage and does not replace it. For this reason, they prefer to speak of "trial marriage" (*shihun*) rather than "cohabitation" (*tongju*). The expression "*shihun*" is a lexical invention of the post-1980s. Xiaocui mentioned several times during the interviews that she did not want to live with her boyfriend until they were married:
*"because of the infuence of tradition, [...] of what other people say. People say that you cannot live together before marriage, have sexual relations before marriage. So maybe in my mind all this is engraved [...]. But little by little my way of thinking has changed".*
Xiaocui wanted to get engaged, before getting married. But as they could not get engaged, they decided to practice "trial marriage" before getting married:
*"We moved in together as a trial marriage, to pretend we were married. For me, it's a form of engagement. Before, we lived separately on our own. I lived in a dormitory. Now, we're renting a small fat near Chaoyang Park. We plan to cohabit together (shihun) for a year, until we get married"* (born in 1986, rural, Heilongjiang *hukou*, Bachelor).
As this interview excerpt shows, Xiaocui appropriates the new concept of "*shihun*" to give new meaning and legitimacy to her practice of living together before marriage. For her, it is a form of engagement, a form of commitment between future spouses. On the occasion of one interview, she named another symbolic act to prove the seriousness of her commitment in the relationship with her boyfriend: for Chinese New Year, she was going to give her in-laws some of her own money as a New Year gift. Before the *shihun*, she did not want to do it and her parents did not want her to give gifts to her boyfriend's parents either. This is the second time she would spend New Year's Eve with them. The previous year, as they were not yet "engaged", her future in-laws gave her a *hongbao* as a gift*.* The following year (at the time of the interview) she received nothing from them, as the young lovers decided to give a joint *hongbao* and gifts to her partner's parents as a sign of their mutual commitment to marriage through the practice of *shihun*. Through this symbolic act, Xiaocui sends a strong message: She is not a "*bad girl*". With this expression, she means a girl of loose morals, who would have sexual relations easily with men or who would be with a man only for his money. She modestly uses the expression "*fasheng guanxi*", which literally means something that happens within a relationship, to refer to sexual relations. In her mind, *shihun* gives legitimacy to cohabiting with her partner and to the fact that they have sexual relations before marriage, because it only takes place for a determined period, that is to say, the time to organize their wedding. In this case, the practice of cohabitation becomes socially respectable for Xiaocui since the process ultimately ends with a marriage.
<sup>3</sup> "I think that now Chinese people do not necessarily need a son to take care of them in old age, *wo juede xianzai zhongguoren meiyou yiding yao erzi yanglao de shuofa"*.
The other respondents born in the 1980s also emphasize the risk that cohabitation lasts and does not lead to marriage. Gezi's and Han's responses are enlightening in this respect. Gezi declared:
*"Living together before marriage, I agree in principle but not entirely, because if it's too long it's likely that we'll never get married. I fnd it good because we get to know each other better when there is love between two people. But I'm afraid to say that I'm totally for cohabitation before marriage. It makes me feel nervous. I'm afraid I won't end up getting married. I'm afraid of having children out of wedlock..."* (born in 1988, rural, Henan *hukou*, College).
#### Han is also in favor of cohabitation because:
*"it allows two people to get to know each other and see if they can live together."*
#### She specifed:
*"However, I am not for cohabiting more than six months with someone. Because if it's too long the situation can become negative. We become an old couple. The young man may keep postponing the marriage, and the young girl becomes, in the meantime, older and older. It is therefore increasingly diffcult for her to separate from her partner. She becomes trapped in the situation, as her chances of fnding someone else decrease over time. [...] I have a friend who has been cohabiting with her boyfriend for three years, but he doesn't want to get married. In such a situation, I would have separated from him after six months, because for me this attitude is a sign that the young man is irresponsible. He doesn't want to take on his responsibilities"* (born in 1989, rural, Jilin *hukou* at birth, Bachelor).
This fear of ending up in a situation of permanent cohabitation is present only in the responses of young women. Men are less eloquent on this subject. In principle, they agree because they do not see any disadvantages. For them, it is a way of getting to know their partner better before legally formalizing their union and building a family. Whereas for young women, whose morals are more subject to social scrutiny, cohabitation is a very serious issue and constitutes a turning point that must be negotiated carefully. It is important for them that cohabitation does not become a substitute for marriage, as the institution of marriage still prevails as the dominant norm for conjugal life.
Risks linked with premarital cohabitation before marriage for young women give rise to intense discussions. This was, for example, the subject of the show "*Baogong lai le*" on March 18, 2013. The title of the show refers to the offcial and judge Bao Zheng (999–1062) who was renowned for his integrity and respect for the law. The show brought together several young adult couples and "experts" to debate their positions on cohabitation before marriage. This is also the subject of many discussions circulating on Chinese websites.4 Like in the interviews I conducted, these discussions emphasize the risks of cohabitation for young women. These discussions are very discreet when it comes to addressing the issue of sexuality. Indirect expressions or euphemisms are preferred when talking about sexual relations. The problem with cohabitation is that it publicly reveals the dissociation of sexual relations and marriage. This is still not socially accepted for young women if the
<sup>4</sup> http://www.u148.net/tale/6736.html. Retrieved on February 10, 2017.
relationship does not end in marriage. Women who cohabitated before marriage would be less desirable partners, because considered as frivolous. In addition, with cohabitation, the passionate love of the early days of marriage would become nonexistent. It would have time to fade during the period of cohabitation and the spouses would become an "old couple" (*laofu laoqi de zhuangtai*). Finally, in addition to the risk of out-of-wedlock birth, TV drama and internet dialogues mention a research conducted in the United States which suggests that the divorce rate is higher among couples who cohabited before marriage.
For these reasons and because it explicitly states that cohabitation is an intermediate stage between single life and the life of a young married woman, the expression *shihun* is preferred by some young women. They thus feel reassured, because the vocabulary used suggests that their behavior is not completely outside the dominant norm. In fact, they take a detour to then return to the path of marriage. Article 8 of the Marriage Law stipulates that couples have a legal obligation to formalize their union.5 As the practice of cohabitation is not yet fully accepted by families and society, *shihun* is a compromise that may develop towards social acceptance of cohabitation, and it may thus gradually lose its negative connotation. Consequently, the individualization of premarital cohabitation paths is limited since it does not tend to replace marriage, as it is the case in European and American countries.
Marriage in China, in other words, the legal union between two spouses of the opposite sex, remains the norm as well as a necessary and unavoidable prerequisite for the founding a family. As in other Southeast Asian societies, it remains the bedrock of family life. At no point in the interviews did young adults question the validity of the couple and family model. It is mentioned 75 times in the in-depth interviews that marriage provides the necessary foundation to build a family and give birth to children. For male or female respondents, "*marriage is not to be taken lightly, as it is the most important thing in an individual's life*".6
For Niu, marriage is a serious event for women, because "*after a divorce a woman's value decreases in China and it is particularly diffcult for her to build a new family*".
He continues: "*divorce also has an impact on our parents, because the neighbors speak negatively about their daughter's divorce*" (born in 1982, urban, Hebei *hukou*, specialized college).
Like those of other respondents who brought up the subject, Xiaolin's words abound in this sense. She explains:
"*Marriage in China is a very important event in a person's life [...]. For example, if a woman divorces and wants to remarry, it is terribly diffcult for her. Or if you have a child, it's very, very diffcult. You are under a lot of social pressure*" (born in 1981, rural, Sichuan *hukou*, College).
<sup>5</sup> In the sample, in 2013, before or after 27 years, 37% of respondents had completed all the transitions marking the journey towards adulthood.
<sup>6</sup>Cheng, Tiejun and Selden, Mark, 1994, *The origins and social consequences of China's Hukou system*, in *The China Quarterly*, p. 649.
The notions of "stability (*wending*)", guarantee (*baozhang*) and feeling of security (*anquangan*) frequently come up in respondents' conversations about marriage.
For Wei, "*a cohabiting household cannot be considered a real family, because the situation is not stable*".
She adds: *"Such a situation would make me anxious because I have no guarantee*" (born in 1981, urban, Heilongjiang *hukou*, Master).
As this notion of guarantee that marriage would provide came up many times in the discussions, I sought to understand what they meant. In the discourse of young men, the notion "Guarantee" refers to the relationship between the two spouses. To quote Tingting:
"*Marriage is something certain that guarantees the relationship between two people*" (born in 1993, urban, Shanghai *hukou*, Bachelor).
As stated earlier, according to Wei, the guarantee is provided by the legal status of marriage:
"*The government gives its approval for marriage, you receive a certifcate, a marriage certifcate. So I consider marriage as legal (hefa) and cohabitation as illegal (fei hefa). If you cohabit, people might wonder: Why don't you get married? Why do you cohabit? I think marriage is the frst step, then you can live with your spouse. Because in China if you don't get married and you live in cohabitation, people will not only think that the couple has a problem, but that the whole family has a problem. Everyone will wonder: But why don't they formalize their situation in front of the government by getting married? Living in cohabitation without getting married is like a marriage but with only the bad sides. You would have to do the housework, cook, without feeling safe. So why not get married?*" (born in 1981, urban, Heilongjiang *hukou*, Master).
Protected by legal provisions, the marriage is envisioned by Suzhi as life-insurance:
"*Marriage is like life insurance for women. If a woman cohabits with her boyfriend in the long term, in case of separation she risks ending up without a house, without money and without work if she did not work. Not cohabiting is an argument to ensure that the relationship with the spouse ends in marriage*" (born in 1978, rural, Hebei *hukou*, college).
In this sense, marriage provides a sense of security. Shasha believes "*that in China girls want to get married because of tradition, but also because they are looking for a sense of security*" (born in 1993, rural, Jiangsu *hukou*, Bachelor).
This commitment marks, in her opinion, the seriousness of the love relationship. In this respect, during a period of cohabitation before marriage, Xiaocui's boyfriend entrusted her with his bank card as a sign of commitment and to provide her with a sense of security. According to her, he wanted to prove to her that she could trust him (born in 1986, rural, Heilongjiang *hukou*, Bachelor).
The pressure to marry is real. It is not only the young adults, or even their parents who want them to marry and have children, it is also the Chinese Communist Party. In a recent speech at the Party's All-China Women's Federation, President Xi Jinping declared that women must play a critical role and must establish a new trend of family, as birth rate plummets and the Nation grapples with an ageing population (Xinhua, 2023). Infuenced by the mainstream narrative, Han fnds it "*very strange* *not to want to build a family, because it is a core that brings comfort*". She adds: "*Getting married, having children, it's perfectly normal, a duty*"7 (born in 1989, rural, Jilin *hukou* at birth *hukou*, Bachelor).
Language development also refects the importance of the marriage norm in the lives of young adults. The expression "*I am part of the tribe of the unmarried*" (*wo shi bu hunzu*) coined by Shasha, or even that of "leftover women" (*shengnü*), which the All-China Women's Federation has been promoting since 2007, evoke a dichotomy between married and unmarried people that leaves little (or no) room for other ways of "*faire famille*". Women who remained single beyond 27 years old are stigmatized and placed in a specifc category (Hong Fincher, 2014). The trend is exemplifed by a decision made in 2023 by a small town in Zhejiang province. The local government announced that they would "offer couples 1'000 yuan (130 euros) as a "reward" if the bride was 25 years or younger" (BBC 2023). As illustrated with this example, Chinese young women face a paradox: they are encouraged to become educated, cosmopolitan, and independent during their childhood and youth, but once they reach adulthood they are under strong pressure to marry before 27–30 years old, to give birth to children and to "return to family life" carrying forward "the traditional virtues of Chinese nation, establish a good family tradition, and create a new trend of family civilization" (Xinhua, 2023).
According to Hong Fincher, the promotion of marriage is instrumentalized by the Party-State to counter the risk of social instability that could be caused by the forced celibacy of a growing segment of the male population (resulting from the imbalance of the sex ratio at birth induced by demographic policies). Parents relay the Chinese government discourse that promotes "traditional" family values and make "pressure" on their children to marry and "give" them grandchildren. Respondents who were not yet married were particularly stressed upon returning from the Chinese New Year. On her return from Tianjin, Shasha explained that she is not completely happy to have returned to her family. In an upset tone, she declared:
"*It is likely that in the end I will succumb to the pressures of my parents for marriage and motherhood, because I am tired of hearing them repeat or insinuate this. Every time my mother calls me, the subject comes up. Before my departure for Beijing, she introduced me to people in Tianjin. Now that I live and work in Beijing, she tries to introduce me to people here! To give me a desire for marriage, she pushes me to attend friends' weddings. Her behavior is very heavy and burdensome. This year during the New Year holidays, I got angry at my mother's insistent and pressing attitude. Now, she acts more indirectly. For example, my parents talk about marriage in the living room while I watch television. They*
<sup>7</sup>They represented, in December 1954, 39% of rural households. See Roux, Alain, 2006, *China in the twentieth century*, Armand Colin, Paris, p. 89, and CNA. (1981). 农业集体化重要文件汇编 *1949–1957 (Nongye jitihua zhongyao wenjian huibian 1949–1957, Offcial documents on the period of agricultural collectivization 1949–1957)*. Beijing: 中共中央党校出版社*,* p. 360.
*know that I hear what they say". She concludes our discussion on the subject by saying: "It is impossible to control parents. For example, my mother discussed in parks with other mothers who asked for pictures of me and who were there to fnd a wife for their son*<sup>8</sup> " (born in 1986, urban, Tianjin *hukou*, Bachelor).
While there is no doubt that the government's campaigns overlook men's responsibilities and rather focus on women responsibility for carrying on marriage, childbearing and family values, men also feel pressure to marry from their families. Having noticed that Yan was spending much less money on his meals upon his return from Hebei, where he had gone to celebrate the New Year with his family, I asked him if everything was okay. In response, he explained to me that his parents had put pressure on him to save money so that he could buy a house and get married (born in 1984, rural, Hebei *hukou*, College).
The birth of children being an integral part of the expectations surrounding marriage, the marriage of the most educated segment of the population could serve the country by producing "high quality" children (Greenhalgh, 2010). For many young adults, the desire for a child is one of the reasons behind their decision to marry. Xiaocui indeed declares "*for me it's not a question, I frst want to get married and then have a child*" (born in 1986, rural, Heilongjiang *hukou*, Bachelor). For Han, this is what motivates many young women:
"*In China many women are concerned about the issue of children. I think that many women only want to get married for this reason. [...] Really, it's to have a child that they want to get married before thirty, because that's when the body is in the best condition to give birth to a child*" (born in 1989, rural, Jilin *hukou* at birth, Bachelor).
Out-of-wedlock births remain stigmatized. One of the respondents, Suzhi, whom I frst met in the summer of 2011 and with whom I maintained regular contact until 2015, found herself in this uncomfortable situation (born in 1978, rural, Hebei *hukou*, College*).* While the social sanctions faced were severe, the marriage law protects children born out of wedlock. Article 25 stipulates that children born out of wedlock should enjoy the same rights as children born within a marital union and that they should not suffer from discrimination. The natural parents of the child, even if one of the two does not live with the child, must cover the child's daily and school expenses until he or she is able to support themselves (HYF, 1981).
However, Suzhi explained that she had to pay a fne for their son to obtain a residence booklet and, thus, a legal existence. The amount varies depending on the locality. Despite the law providing some safeguards for children born out of wedlock, customary practices remain tenacious. For instance, Qian, a University professor graduated from a prestigious University, stated on the subject that:
<sup>8</sup>Li, Huaiyin, 2009, *Village China under socialism and reform. A micro-history, 1948–2008*, Stanford, Stanford University Press, p. 39.
*"without marriage, births are illegal in China. He has no hukou. A person who does not marry is single and without children. [...] Through marriage, the family is formed. [...] Marriage is the frst step to having a family. No registered marriage, no children in China"* (born in 1982, urban, Jilin *hukou*, PhD).
The strength of this custom is refected in the wedding celebration rites. The bridal bedroom where the newlyweds share their frst night after the wedding is the subject of meticulous preparations. At Xiaocui's wedding, the bedroom where they were going to spend their frst wedding night had been all decorated in red (the color of good luck) and with signs of double happiness. The family had pasted photographs of babies, especially baby-boys, on the walls of the room to encourage them to have a child soon, and preferably a son. On the red blanket that covered the bed, and whose set was offered to the newlyweds, a heart was drawn with jujubes (*zao*), peanuts (*huasheng*), longans (*guiyuan*) and seeds (*zi*). The choice of fruits and seeds is not left to chance. Their symbolic function is to encourage the newlyweds to give birth to a son as soon as possible. Indeed, the Chinese character for jujube is pronounced *zao*, like the word "early". Peanut is pronounced *huasheng* and, if we only keep the last part, "sheng" is written and pronounced like "birth". The frst part of the word longan *guiyuan* is pronounced like "precious" (*gui*). Finally, the word "seed" has the same pronunciation in Chinese as "child"9 (*zi*). The decoded message reads: "Give birth to a precious son quickly (*zao sheng guizi*)". This message is sometimes explicitly written on the bed of the newlyweds with the same fruits and seeds. As the interviews show, the decisions to marry and have children are intimately linked in the minds of young adults. The Confucian thinker Mencius himself said that "there are three ways of being unflial, and not having offspring (a son) is the most serious".10
#### **Bibliography**
Attané, I., Zhang, Q., Li, S., Yang, X., & Guilmoto, C. Z. (2013). Bachelorhood and sexuality in a context of female shortage: Evidence from a survey in rural Anhui, China. *The China Quarterly, 215*, 703–726.
Bourdieu, P. (2007). *La Distinction. Critique Sociale Du Jugement*. Les Editions de Minuit.
Croll, E. (1981). *The politics of marriage in contemporary China*. Cambridge University Press.
Fincher, H., & Leta. (2014). *Leftover women: The resurgence of gender inequality in China*. Zed Books.
Girard, A. (2012). *Le Choix Du Conjoint. Une Enquête Psychologique En France*. Armand Colin.
<sup>9</sup>Goodman, David, 2004, *China's campaign to « Open up the West »: National, Provincial, and Local Perspectives*, Cambridge, Cambridge University Press.
<sup>10</sup> http://english.gov.cn/offcial/2005-07/29/content\_18334.htm. Consulted on May 3, 2010.
**Open Access** This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.
The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.
# **Chapter 9 What Does the Individualization Process Do to Intergenerational Solidarities?**
#### **Contents**
**Abstract** This chapter argues that TV series disseminate values of intergenerational solidarity to cope with daily struggles. The themes of flial piety and family relationships are recurrent. They convey a moral code designed to counteract the negative effects of individualization and the shortcomings of the social security system.
The dynamics of family solidarity refect a nuanced evolution in flial piety practices. While the Confucian thought placed different obligations on sons and daughters, flial piety is transforming from an unconditional duty to a more reciprocal and negotiated form, infuenced by factors such as affects, economic and family situations.
The fndings also reveal the individualization of living spaces and simultaneously the maintaining of strong family ties. Financial and non-fnancial supports from parents remain a signifcant aspect of family solidarities. It is observed that the type of support changes over the life-course. For instance, once parents, 70% of respondents receive non-fnancial support from their families.
The chapter concludes highlighting the challenges faced by the "sandwich generation" caught between supporting elderly parents and raising their own children. The precariousness of working conditions and the absence of a robust Welfare system further contribute to the strain felt by young adults.
**Keywords** Filial piety · Intergenerational solidarity · Individualization · Reciprocity · Support · Welfare system
#### **9.1 "Better Not to Have a Child Than an Unflial Child"**
TV dramas are another vector through which the Party-State promotes a familialist ideology, in other words, conservative family values. In the corpus of television series analyzed, the family personifes the country. The Chinese character that designates the family (*jia*) makes up the word for country or the Nation (*guojia*). For centuries, the family or the family unit has played a role in Chinese thought. The family is thought as constitutive of the Nation and the State (Kong, 2008). In Maoist era, the family was political and subordinate to the State. During the process of decollectivization, the family reemerged as a semi-private sphere. The Party-State retained control over the family sphere, but at the same time it privatized the family by enjoining it to take over functions it had previously assumed.
In TV dramas, the central place given to the family and family relationships refects this logic. At a time when viewers are confronted with the problem of unemployment, economic and social pressures, and the rise of uncertainties in their daily lives, the representations of the family portrayed in TV series not only offer them emotional identifcation, but also provide benchmarks by disseminating values of solidarity between generations as a way of coping with everyday struggles. By tackling the theme of flial piety and family relationships, TV dramas convey a moral code that aims to counter the negative effects of the individualization. If they cannot manage to solve their problems on their own, individuals are less expected to turn to the State for answers, than to members of their family.
This discourse refers to the logics of family solidarities in force in pre-Maoist China. It is widely accepted by the respondents, who all declared that "it is better not to have a child than an unflial child" (*yanger buxiao, buru buyang*). Chinese New Year is their favorite holiday because it is a family celebration. It gives them the opportunity to show their flial affection for their parents by giving them gifts. It is about "*giving something back*" to their parents as Ruyue puts it, "*taking care of them*". She continues:
"*You want to express your love, your piety. It allows you not to feel guilty. [...]. By earning 10,000 yuan per month, I can give them between 2,000 and 5,000 yuan per year*" (born in 1990, urban, Guizhou *hukou*, Bachelor).
Wanheng thinks that "*most Chinese give a part of their frst salary to their parents to thank them*". He adds:
"*The amount is not important. It's mainly the action, the gesture that matters. When you have children, you have less time to spend with your parents. So you give money to express your feelings. Because you have less time to go see them, spend time with them*" (born in 1980, rural, Shanxi *hukou*, College).
The fnancial contribution is sometimes not suffcient or not necessary when the parents of young adults have enough resources. Wei realizes, for example, that her parents need her more to "help them physically and morally". This is also what Yule observes:
"*Parents in China need their children when they retire. They need someone to accompany them, talk to them, entertain them*" (born in 1986, rural, Heilongjiang *hukou*, Bachelor).
On July 3, 2013, the Chinese government decided to strengthen the 1996 law on the "protection of the rights and interests of the elderly", which institutionalizes the duty of assistance and support by children and their spouses to their parents and inlaws over 60 years old (LQBF, 2012). According to the law, children not only have a duty to provide fnancial support to their parents in need, but also to visit them regularly. These legal provisions transform flial piety from a moral duty into a legal obligation. As with Article 21 of the 1981 Marriage and Family Law, it is no longer only sons who have the duty to be flial, but also daughters. Historically, this duty fell to sons because the family inheritance was theirs. Even in the poorest families, it was socially expected that sons fulfll this obligation of flial piety. Breaking this norm and disrupting family harmony exposed them to severe social sanctions. The Chinese character *xiao*, which means flial piety, is signifcant in this respect. It is made up of two characters: the character *lao*, which means old, is on top, and the character *zi*, which can be translated as son, is located below. Symbolically, the ideogram shows that the elder patriarch is above his son and that the latter supports the former (Ikels, 2004).
The intergenerational contract was concluded in different terms for daughters. They were expected to contribute to the family economy until their marriage. Once married, they were expected to contribute to the family economy of their husband's family. Passing on all or part of the family inheritance to them would have been seen, according to the Indian proverb, as "watering the neighbor's garden". In other words, it would have been envisioned as a loss since the women would eventually leave their family of origin. In contrast, expenses made by families for their sons were appreciated as an investment since the sons would not only remain within the family lineage, but they would also perpetuate it (Ikels, 1993).
Number of studies carried out after the country's opening-up to the market economy point out a weakening of family solidarities. The growing individual mobility is said to have contributed to uprooting young adults from their communities of origin. Consequently, they fnd it easier to distance themselves from the social pressures enjoining them to comply with flial piety obligations (Ikels, 2004; Yan, 2003a, b). The most recent research nuances this thesis (Cockain, 2012; Fong, 2004; Liu, 2020). Like in this book, they identify a form of renegotiation of the intergenerational contract. The interviews I conducted indicate that once married it is not only sons who remain flial, but also daughters. Moreover, it appears that the practice of flial piety takes different forms depending on the economic and family situation of the respondents. Han shared with me:
"*If I had an older brother or a younger brother, it is certain that it would not be up to me to take care of my parents. But many families, like mine, having only a daughter, by force of circumstances, the situation evolves. Today, in my home village it is common that once married, daughters continue to support their parents. I think that today the Chinese do not have* *a determined practice that would only oblige sons to take care of their elderly parents.*<sup>1</sup> *Because today the situation is no longer the same as before. Many people do not have sons. Moreover, I think that more and more people are wondering why only sons should be responsible for the elderly. I think that it is quite a lot of pressure for them. It is possible that they sometimes do not feel capable and try to avoid it. There are also cases where sons do not fulfll their flial obligations towards their parents because they have been too spoiled by them. I think this is the case in families where the parents had several daughters before fnally being able to have a son. In this case, it is usually the daughters who take the place of their brother to help their parents. For me, flial piety is not only a duty, it is also a responsibility that I have. This is related to morality (daode). You cannot disregard the elderly*" (born in 1989, rural, Jilin *hukou* at birth, Bachelor).
Gezi, who is also of rural origin and has a sister, declared:
"*Once married, I will certainly continue to help my parents. It is certain that my spouse will not oppose it since I will have chosen him. A person's values (renpin) are very important*" (born in 1988, rural, Henan *hukou*, College).
Xiaocui, who is also of rural origin but has a sister and a brother, also decided to continue helping her parents once married. On Chinese New Year, they offer both families *hongbao*. She explains that her sister, who is married, still helps her parents. She often offers gifts to her parents in secret because she fears her husband's reaction if he fnds out. For the Chinese New Year, she and her husband give the same amount of money to both families. For Xiaocui, flial piety was an important criterion in the choice of her spouse because:
"*a person who does not show flial piety towards his parents, even though they have brought him up for many years, if he does not even support his parents, it is certain that he will not be generous with other people*".
As the discussion continued, Xiaocui explained that she considers it her obligation or duty (*yiwu*) to look after her parents (born in 1986, rural, Heilongjiang *hukou*, Bachelor).
These interview excerpts reveal the transformation of values and practices in terms of family solidarity in the countryside. Regardless of their level of education, the composition of their siblings and their fnancial situation, it is not conceivable, once married, for these young women not to support morally and fnancially their parents. The interviews conducted with these women who have migrated to the capital do not indicate the existence of a "crisis of flial piety" as observed by Yan Yunxiang in the late 1990s (Yan, 2003a, b:189). They testify to its evolution. Filial
<sup>1</sup> In 2003 the NRCMS is defned as: "[It] provides mutual help and beneft, mainly focusing on and curing severe diseases. It is organized, led and supported by the government and with voluntary participation of the farmers. The system is fnanced jointly by individuals, collectives and government," "*Guanyu jianli xinxing hezuo yiliao zhidu yijian*" ("Opinions about the introduction of NRCMS"), Guobanfa, No. 3 (16 January 2003), http://www.jswst.gov.cn/dfnewsdisplay. php?newsid=436, cited in Klotzbücher, Sascha; Lässig, Peter; Qin, Jiangmei; and Weigelin-Schwiedrzik, Susanne, 2010,"What is New in the "New Rural Co-operative Medical System"? An Assessment in One Kazak County of the Xinjiang Uyghur Autonomous Region", *The China Quarterly*, 201, March, p. 38.
piety is no longer unconditional. It is now based on a logic of reciprocity. The intensity of the relationship and the level of solidarity offered by the children to their parents depends on the generosity and attitude of the latter. For example, after getting married Xiaocui scaled down the level of solidarity she intended to offer to her parents. She did not express it directly this way, but when I analyzed the interviews, which ran from June 2012 to August 2014, I noticed a change in tone or rather a disappointment. She recounted:
"*At my brother's wedding, my parents had given him all the hongbao they had received. At my wedding, they kept them all. They did this because he is a boy, and I am a girl. My parents also helped my brother to buy a house, although they are not going to help me*".
When I asked her why, she answered with weariness: "*because I am a girl*". I insisted and asked her what she meant by that. She added: "*It's because my parents are hoping my brother's will look after them once they get old. They plan to go live with him*". Asking her if she fnds this situation unfair, she replied: "*I fnd it normal, I do not fnd it unfair, because once my parents are old, if my brother does not take care of them, they will go to a nursing home that they will pay for themselves, as they have given a lot of money to my brother and not to me, I think it's clear in my parents' minds*" (born in 1986, rural, Heilongjiang *hukou*, Bachelor).
Xiaocui did not directly blame her parents for their behavior, but her words indicate that she does not endorse it. Before her wedding, the support she considered giving to her parents was unconditional. Afterwards, her words refect a change in position. She no longer feels as much indebted as before to her parents since they have decided not to support her fnancially in planning her wedding and buying a home. These were two signifcant milestones in her transition to adulthood and, for the second, it will be very diffcult for her to achieve it without family support.
As for urban single daughters, it is inconceivable for them to favor their in-laws at the expense of their parents. Weiwei is aware that providing support and assistance to her parents is a legal obligation. However, she declares that this is not the reason why she does it, it's because she feels close to her parents. Like all the urban respondents, her parents receive a pension. Weiwei's parents also own their home. Therefore, she does not need to support them fnancially. However, she needs to provide them with physical and moral support when they need it (born in 1982, urban, Beijing *hukou*, Bachelor).
The letters that circulate on Chinese websites illustrate the tensions at the heart of the renegotiation of intergenerational solidarities.2 A number of young women write that they do not refuse to help their in-laws as long as they show gratitude and do not take it for granted. In this sense, they do not question the notion of flial piety and the idea of solidarity between generations. However, they believe that relationships must be reciprocal and demonstrate mutual respect. Since the reforms, research shows that emotional ties have become tighter in both rural and urban areas
<sup>2</sup>Op.cit., Li, Huaiyin, p. 325.
(Evans, 2010; Shen, 2013; Yan, 2023). As the interviews reveal, young women who grew up in such an environment fnd it diffcult to accept hierarchical relationships devoid of reciprocity and mutual respect once they get married. It is also unthinkable for them to break the strong ties that bind them to their parents.
#### **9.2 Elective Intergenerational Solidarities Revolving around the Family Nucleus**
The majority of respondents born in the 1980s do not to cohabit with their parents, even after the birth of a child (Fig. 9.1). The analyses reveal that at the age of 27, most urban and rural respondents are not living with their parents or in-laws after becoming parents (in orange, Fig. 9.1). All other things being equal and as expected, Beijingers are signifcantly more likely to cohabit with their parents or in-laws than other respondents.
The individualization of living arrangements stems from the desire of young adults to preserve their privacy and avoid conficts with their parents, but it does not necessarily lead to a decrease in family solidarity. The discourse and practices of the respondents express, on the contrary, a very strong sense of responsibility towards
**Fig. 9.1** Intergenerational cohabitation with parents and/or in-laws, respondents born between 1980 and 1985 with at least one child
**Fig. 9.2** Financial support for marriage received by respondents born between 1980 and 1985 – Logistic regression models (*odds ratio*)
their parents. The nuclear family plays a crucial role, as it is the main source of support that young adults and their parents can rely on to respond to social, economic, and personal uncertainties as well as the risks that comes along the process of individualization. For the young adults who took part in this research, their family – they mean by this the parent-child nucleus – forms a small collective unite around which family solidarities revolve. A third of the respondents said that they had received fnancial help from their parents a year before getting married or in the year of the marriage. Young adults from Beijing are signifcantly more likely to beneft from this type of support than other respondents. The analyses also suggest that the greater likelihood of receiving fnancial support is correlated with the respondents' level of education. The higher the level of education, the more likely they are to have received fnancial aid from their parents before getting married (Fig. 9.2).
Half of the respondents who have become homeowners received fnancial aid from their parents either the previous year, the same year, or the following year. Among the young adults I met, those of urban origin and Beijingers were signifcantly more likely to have received family support to become homeowners (Fig. 9.3). For instance, Han explained during the interview that her mother is going to give her money when she gets married so that she can contribute (along with her spouse) to
**Fig. 9.3** Family fnancial aid for the purchase of a home in the paths of respondents born between 1980 and 1985—Logistic regression models (*odds ratio*)
the purchase of a home. She points out that she will not need to pay back her mother because it will not be a loan, but rather a gift. These fndings corroborate sociologist Serge Paugam's thesis that family solidarities have the effect of reinforcing social inequalities rather than reducing them, since this type of support remains weak among the most deprived population groups (Paugam, 2008).
Family solidarities are not only a matter of fnancial support. Over 70% of respondents with children received non-fnancial support from their family. Other kind of family support might not depend on the family's socioeconomic background. However, logistic regressions did not reveal any signifcant explanatory variables. According to research conducted in the 1990s, relationships between grandparents and preschool-aged grandchildren were frequent. This was due to the proximity between generations. The mismatch between supply and demand in the urban housing stock had the effect to encourage multigenerational cohabitation. Moreover, a relatively large proportion of young children were entrusted to the care of their grandparents at the latter's home (Chen, 2014). This confguration refects the family network (Widmer & Jallinoja, 2008). Relationships between non-co-resident grandparents, their children and their grandchildren are not only frequent, but the adults share signifcant family responsibilities. While respondents would consider their parents and in-laws coming to live with them for a few months to help them look after their child(ren), they all insist on the limited duration of such arrangements. This is, for example, what Suzhi's mother did to help her after the birth of her son. Despite her modest fnancial means and social disapproval, she came from the countryside to support her daughter in her home, staying for several months or weeks at a time. At the time, her daughter lived as a single mother in an accommodation paid for them by her son's father, who was meanwhile living with his wife and child in the same city (born in 1978, rural, Hebei *hukou*, College).
Findings highlight that family solidarities tend to reverse over time. The older young adults get, the more their responsibilities towards their parents increase. In both rural and urban areas, they tend to provide fnancial and emotional support whenever they can. The fndings further reveal that by the age of 22-23, fnancial support becomes predominant compared with the other forms of help given to parents (in light blue on Fig. 9.4).
**Fig. 9.4** Support provided by young adults born between 1980 and 1985 to their parents (crosssectional representation)
The concentration of family solidarities around the nuclear family is a source of stress for young adults, particularly for those who are only children. Many of them, like Tingting (an only child), regret not having a brother or sister with whom to share this responsibility. Barely 20 years old, he laments:
142
"*With brothers and sisters, we could share the responsibility of taking care of our parents. But with the one-child policy, we really have a heavier burden than before. I have already thought about this problem. When I get my job, if my parents are ill, who is going to look after them? Parents in China need their children when they retire. They need someone to be with them, to talk to them, to distract them. And as only children, we have a greater responsibility than the generations before us*" (born in 1993, urban, Shanghai *hukou*, Bachelor).
According to Han, the mental and fnancial burdens borne by members of her generation are "gigantic" (born in 1989, rural, Jilin *hukou* at birth, Bachelor). While they are caught in the whirlwind of their professional and family life, which often require them to work long hours and migrate to other localities, they must, according to the law, fnd time to regularly visit their parents. The respondents often want to do so, but how to reconcile the irreconcilable? They lack time! That is why they tend to prefer fnancial support to their parents over emotional support or care. However, as it was often stated in interviews, in the event that their parents are dependent for a decent living on the fnancial support they provide, this puts young adults under "*heavy economic pressure*". Once they are married, they may have up to four parents to support in addition to other spendings, such as: bringing up their children and providing for them, mortgage or rent, etc. All of this, in a context where the cost of living keeps rising due to infation. If in urban areas only children fnd themselves alone in fulflling their flial obligations, their parents often beneft from better medical cover and pension benefts than in the countryside. Han uses the image of an hourglass (*shalou*) to describe the heavy pressure felt by young adults (born in 1989, rural Jilin *hukou* at birth, Bachelor). The Chinese media refer to this phenomenon as the "sandwich generation" (*jiaxin zu*).
Young adults born in the 1980s are the frst birth cohort to have to cope with longer life expectancy. As long as their parents are in good health, the situation is relatively manageable, but acute problems arise when the elderly lose their autonomy. The failures of the welfare state means that these risks are *de facto* shouldered by individuals and their families. In the absence of robust social insurance schemes for the entire population and a lack of subsidized specialized facilities to care for the elderly and preschool-aged children, the pressure on young adults is increasing as the years go by. The precariousness of working conditions makes this pressure even stronger. Against this backdrop, Yan Yunxiang has recently observed a tendency for parents of young adults to go out of their way to support their children and grandchildren, putting their own needs second. He describes this transformation of intergenerational solidarity dynamics as an inversion of the family hierarchy (Yan, 2021).
#### **Bibliography**
**Open Access** This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.
The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.
# **Chapter 10 Conclusion**
#### **Contents**
Bibliography 151
This book, which stems from my doctoral thesis defended in 2017, was born out of a questioning of the impact of the dismantling of collectivist-type social policies on the dynamics of inequalities formation over the life-course. In contemporary societies, it seems that the feld of possibilities for young adults has never been so open. Individual risk-taking, self-realization, personal fulfllment, capacity development, mobility and change are highly valued, while failures in life-course increasingly take the form of individual crises. Being assimilated to personal failures, systemic problems are deprived of their political dynamics. The diversifcation of the paths towards adulthood has the effect of making invisible the logics underlying the construction and reproduction, or even the intensifcation, of inequalities over the lifecourse. The young adults' paths tend to be interpreted as the result of personal choices. However, such an interpretation conceals the role played by institutions in the construction of social inequalities.
China constitutes a particularly interesting case study since the country has undergone an extremely rapid transition from a collectivist welfare system to an individualization of the social protection system. In the early 1950s, Marxist and Maoist ideologies supplanted the logic of market. It gave birth to a new social organization based on a collectivist model. As described in this book, barely thirty years later, without naming it, the Party-State rejected this form of social organization, which had only partially succeeded. The opening-up to market economy led the country towards profound changes. The allocation of jobs and means of subsistence to individuals by work units or people's communes, in return for their participation in the socialist model of societization, gave way to individualization in the economic, social and, to some extent, private spheres. From the 1990s onwards, individuals were enjoined to "liberate" (*jiefang*) their individual capacities, to "rely on themselves" (*kao ziji*) and no longer "depend on the State" (*kao guojia*) (Ong & Zhang, 2008). The pursuit of personal proft and self-enterprising were also valued.
The popular view is that young adults are now free to seize lives with both hands, as well as all the new opportunities offered by market economy. However, this process of individualization comes with constraints. The individualization is partial. It is not only regulated and structured by the State, but it also carries social inequalities.
The analyzes produced in this book are unique in that they examine the dynamics of individualization through the prism of the transition to adulthood of two birth cohorts that are emblematic of the reforms carried out in China: The frst made most of its journey to adulthood in collectivist China, while the second made its way to adulthood in post-collectivist China. The respondents had been living in Beijing for at least six months at the time the research was conducted. Among 615 respondents born between 1980 and 1985, three types of paths to adulthood stood out: The frst one is centered on relatively early fnancial independence (28% of respondents), the second one on a relatively long period of studies (43% of respondents) and the third one on relatively late decohabitation (29% of respondents) (Fig. 10.1).
Young adults who achieve fnancial autonomy relatively early (cluster 1 "Relatively early fnancial autonomy") tend to live with their parents until they start working, around the age of 17 (yellow). By the age of 19, almost all members of this group are working and no longer live with their families (pink). From the age of 21 onwards, they begin to live with a partner or get married (turquoise) and move towards parenthood (purple) while remaining professionally active.
The later entry into the labor market of some respondents can be explained by their educational background (cluster 2 "Relatively long period of study "). These young adults continued their training in higher education after high school (orange). They began to enter the labor market around the age of 20 (pink). By the age of 23,
**Fig. 10.1** Cross-sectional representation of the three types of transition to adulthood for young people born between 1980 and 1985
most young adults in this cluster are professionally active. They start living with a partner (turquoise) and a few years later some become parents (purple) (Fig. 10.1).
The path followed by young adults in the third cluster "Relatively late decohabitation", in other words, who leave the parental home relatively late, is very similar to that of the young adults in the frst group. The majority of them continue to live with their parents after graduating from high school. By the age of 22, almost all of them made their professional transition (green). At this age, they begin to live with a partner and have a child. For some, this means continuing to live under the same roof as their parents or parents-in-law (Fig. 10.1).
This cluster analysis reveals a limited diversifcation of the paths to adulthood followed by the young adults. The analyses highlight social determinisms and the normative relevance of education in young adults' life-course. School achievement is the main factor that signifcantly explains age differences in access to fnancial autonomy. Indeed, young adults who are not coming from Beijing and who did not go on to higher education have a greater propensity to leave the family home quickly after post-compulsory schooling to work in order to achieve fnancial autonomy. In this specifc situation, early family decohabitation is often explained by the constraints of the labor market that leads them to be mobile to fnd a job.
The third cluster is made up mainly of young adults with a Beijing *hukou*. With the same level of education as the previous group, these young adults tend to stay with their parents even though they have a job that provides them with fnancial autonomy.
Sequence analyses reveal a clear intra-cohort standardization of the pathways to adulthood since the temporality of state sequences is almost uniform for women and men who share the same type of *hukou*. The analyses also reveal that the school system has a strong normative infuence on the temporality of the transition to adulthood (Fig. 10.2). Although the residence booklet policy (*hukou*) has been slightly relaxed and access to higher education has been expanded in post-collectivist China, the comparison of the paths of young rural and urban people, originally from Beijing or elsewhere in China, of women and men with high and low education levels indicates that the *hukou* retains its deterministic effect on people's life paths.
The emergence of this new life stage in the young adults' life-course should be considered in the context of the offcial rhetoric on social and economic progress and the discourse on the "quality" of the population. In 2001, a reform of the primary school system was implemented. The objective of this reform was to move from "exam-oriented education" (*yingshi jiaoyu*) to "quality education" (*suzhi jiaoyu*) (Chicharro, 2010:183)*.* According to a member of the Ministry of Education, the reform aimed to "*prepare a new generation of "quality" citizens capable of serving China in its modernization*" (Chicharro, 2010:184). The ambition to shape a new type of citizen has always involved educational reform in China. Even after the Communist Party came to power, one of the slogans of this period was "*fght illiteracy, shape a new man*". Unlike the Maoist period, today the aim of the CCP is to no longer to encourage social homogeneity and political fervor. It is about shaping a type of person who corresponds to what post-collectivist China needs to achieve the country's dream of "national rejuvenation". The balance is diffcult to fnd between, on the one hand, encouraging the development of individualities,
**Fig. 10.2** Cross-sectional representation of intergenerational changes in the sequence of the transition to adulthood (1950–1959 and 1980–1985)
shaping autonomous young adults who are capable of taking personal initiatives but, on the other hand, who remain patriotic and do not destabilize the CCP. In this political project, the professional success of young people contributes to increase the country's economic power and international prestige.
The analyses carried out for this book also indicate that the journey of young people towards adulthood does not always coincide with the representation they have of the transition to adulthood. It is particularly interesting to observe that home ownership as a prerequisite for marriage or the birth of a child is a new norm reappropriated in the discourse of young people, but which in reality only applies to a privileged fringe of the population. Homeowners are mainly young Beijingers who have an urban *hukou* or who have attended University. All other things being equal, the discrete time logistic regression model (GLM) indicates that young people who have become homeowners are signifcantly more likely to have experienced all the stages constituting the transition to adulthood (Fig. 10.3).1 In this sense, access to home ownership, which constitutes a specifc element of the transition to adulthood in China, can make the process smoother or, on the contrary, sometimes generate stress.
<sup>1</sup>Chan, Kam Wing, and Buckingham, Will, 2008, "Is China "Abolishing the Hukou System?", *The China Quarterly,* 195, September, p. 602.
**Fig. 10.3** Discrete time logistic regression model
While, like elsewhere, the transition to adulthood in China has been lengthening over time (Fig. 10.2), one of the major contributions of this research is to reveal, through the respondents' discourse, the importance of the place of the family and family obligations in this stage of the life-course. Family responsibilities and marriage constitute a structuring norm in their pathways. Contrary to what has been observed in Western countries, marriage and parenthood, which confer on young adults new social roles and personal attributes, are not disconnected in the respondents' minds. For the respondents, there is not a unique experience that marks the passage towards adulthood. These are multiple events and the gradual accumulation of experiences that accompany the transitions that enable them to feel adults. They attach individualistic attributes to these transitions, such as accepting new responsibilities (*you zeren xin*), in other words, the ability to take responsibility for their actions. In particular, respondents mention professional responsibilities and family responsibilities towards their parents and their own family; or once married, towards their parents-in-laws. Autonomy is also mentioned. This refers to the ability to "make decisions by themselves" (*ziji zixing*) and to "solve problems independently" (*duli jiejue wenti*) without the help of parents. They identify fnancial autonomy, associated with employment and access to home ownership as a condition for their transition to marriage and parenthood. They attach to these transitions, which confer upon them new social roles and personal attributes, the blooming of a sense of responsibility towards others. They also envision the transition to adulthood as a process of psychological development towards maturity.
The fndings presented suggest that regardless of the social origin and level of education of the respondents, the family continues to play a signifcant role in the lives of young adults. This is particularly true in terms of intergenerational solidarity. Both parties are legally bound by intergenerational solidarity obligations, but it
is no longer only men who can count on their parents' support or who must be flial to them. Changes in family law are not the only factor driving this trend towards neo-familialization. The dismantling of the collectivist welfare state, which has yet to fnd a substitute, is another factor that explains this process. The social assistance system is still rudimentary and is designed for people who are physically unable to carry out a professional activity and/or who fnd themselves without descendants. Furthermore, since the end of 2016, President Xi Jinping has explicitly placed the family at the center of the national project of rejuvenation in a series of speeches. By emphasizing that respect for family traditions benefts not only individuals and their family, but the society as a whole, the Party-State promotes a process of neofamilialization which bounds individualization in family relations.
As highlighted in this book, individualization is a complex phenomenon. Social policies and changes in the legal framework support a high degree of individualization in the labor market but limit the depth of the process in the family sphere. Intergenerational solidarity obligations are frmly engraved in the law. These two forms of individualization take place concurrently and interact to address the problems of inequality and social anomy that result from China's rapid entry into the globalized market economy. As socioeconomic supports provided by the family depends on its resources, the resort to family solidarities contributes *in fne* to reinforcing social inequalities rather than reducing them. The same applies to gender inequalities, since women are still the majority of those who take on the tasks of *care*.
The recent emphasis in TV dramas, magazines and mainstream discourses on traditional Confucian culture and women's role as virtuous wives and mothers once married are characteristic of the differentialist ideology diffused in post-collectivist China. In contrast with communist slogans, such as "what men can do, women can do too" mobilized to favor gender equality and women's participation in productive activities, today's China is permeated by the idea that a difference in nature exists between women and men. As depicted in the TV drama "Ant's struggle", the main character, Chuchu, resigns from her job after becoming involved with Rongsheng. Once married, she devotes herself to their new home. She is portrayed as not hesitating to sacrifce fnancially and emotionally for her husband and child. These qualities, valued by the narrative and the scenography, refect those associated with the ideal of the virtuous woman in Confucian thought. In the corpus of TV series analyzed in this book, the articulation between work and family life is always pictured as a female dilemma. The latter dimension is systematically emphasized in the lifecourse of young women. Women are portrayed as belonging to the domestic sphere. The narratives encourage them to devote time and energy to their family responsibilities, rather than to pursue a fourishing career. This essentialist view of men and women' social roles reinforce gender inequality.
This rhetorical shift, that has been intensifying since 2007, should be seen against a backdrop of rising youth unemployment, which mainly affects young graduates from "ordinary" higher education institutions (*putong gaoxiao*). The Party-State discourse enjoining women to marry before 27 years old, to give birth and to "return to family life" carrying forward "the traditional virtues of Chinese nation, establish a good family tradition, and create a new trend of family civilization" echoes the
State discourse on "women returning home". This narrative emerged in the 1990s, in a context where unemployment rate was rising due to the dismantling of the *danwei* system. It was a strategic response from the State, inviting women to quit their job and devote themselves to care responsibilities, to contain the surge of unemployment. This insistence on women's domestic role is a departure from gender equality. With the dismantling of the collectivist Welfare State, nowadays public spendings for pre-school childcare facilities and raising children are not suffcient. Through this narrative the Party State is turning on women to assume the tasks of care and to offset the decline in family policy.
Young women may not be keen to go down this path without resistance. Despite the raise since May 2021 of the legal limit on the number of children per couple from two to three, young people appear to delay childbirth, as well as buying a home amid a faltering economy and rampant unemployment. This emerging social phenomenon is known as the "young refuseniks". They reject the traditional fourfold path to adulthood: fnding a mate, marriage, mortgages and raising a family.
#### **Bibliography**
Chicharro, G. (2010). *Le Fardeau Des Petits Empereurs*. Société d'ethnologie. Ong, A., & Zhang, L. (2008). *Privatizing China. Socialism from Afar*. Cornell University Press.
**Open Access** This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.
The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.
# **Appendices**
#### **Appendix 1: Timeline of the History of Contemporary China**
owners to landless peasants and those who lack it (this practice was already widespread in communist communities before 1949). Entry of Chinese troops into Tibet, over which China asserts its sovereignty. Sending troops to Korea. China's support for the Viet Minh in the Indochina War against France
<sup>1</sup> http://english.gov.cn/offcial/2005-07/29/content\_18346.htm. Accessed on May 3, 2010.
Mao Zedong gives a speech to accelerate the march towards collectivization and compares "the cadres who refuse to do so to women with bound feet who trot on the road of history seeing a tiger in front and a dragon behind"2 .
<sup>2</sup>People's Daily ( *Renmin ribao*), March 6, 2004.
<sup>3</sup>No. 1 central document issued: http://english.gov.cn/index.htm; and interviews with the peasants of the Danian commune.
<sup>4</sup> http://english.gov.cn/offcial/2006-03/14/content\_227248.htm. Retrieved on May 3, 2010.
<sup>5</sup>No. 1 central document issued: http://english.gov.cn/index.htm.
<sup>6</sup> "A ministry report said 12 places, including Hebei, Liaoning, Shandong provinces, the Guangxi Zhuang Autonomous Region and Chongqing Municipality, had launched pilot programs to experiment with a system that narrowed differentiation between rural and urban residents." Article appeared in the China Daily on March 31, 2007: http://www.chinadaily.com.cn/china/2007-03/31/ content\_840877.htm. Retrieved on May 5, 2010.
<sup>7</sup>Mun, Young Cho. (2010). On the Edge between "the People" and "the Population": Ethnographic Research on the Minimum Livelihood Guarantee. *The China Quarterly, 201* (March), 20-37.
<sup>8</sup>People's Daily ( *Renmin ribao* ) of April 15, 2010; *China Daily* of April 16-18, 2010.
Wen Jiabao becomes prime minister, start of the controversial construction of the Three Gorges Dam. State directive9 aimed at facilitating the employment of the rural population in the city and eliminating discriminatory policies or arbitrary taxes against them.
<sup>9</sup>China Daily of March 5, 2010.
<sup>10</sup>China Daily of March 15, 2010.
<sup>11</sup>Document No. 1 of the CPC Central Committee promulgated at the end of January 2010.
<sup>12</sup> http://english.gov.cn/offcial/2006-03/14/content\_227248.htm. Consulté le 3 mai 2010. Retrieved on May 3, 2010.
<sup>13</sup> http://english.gov.cn/index.htm: No. 1 central document issued.
<sup>14</sup> «A ministry report said 12 places, including Hebei, Liaoning, Shandong provinces, the Guangxi Zhuang Autonomous Region and Chongqing Municipality, had launched pilot programs to experiment with a system that narrowed differentiation between rural and urban residents.», China Daily,
March 31, 2007: http://www.chinadaily.com.cn/china/2007-03/31/content\_840877.htm. Retrieved on May 5, 2010.
<sup>15</sup>Mun, Young Cho. (2010). On the Edge between "the People" and "the Population": Ethnographic Research on the Minimum Livelihood Guarantee. The China Quarterly, 201(March), 20–37.
<sup>16</sup>人民日报 (Renmin ribao), April 15, 2010; China Daily, April 16-18, 2010.
<sup>17</sup>China Daily, March 5, 2010.
<sup>18</sup>China Daily, March 5, 2010.
<sup>19</sup>Document No. 1, CCP Central Committee, promulgated at the end of January 2010.
#### **Appendix 2: Administrative Division of the Municipality of Beijing**
#### **Appendix 4: Quantitative Convenience Sample Stratifed by Quota - Birth Cohort 1980–1985 (Targeted Number of People)**
The distribution of quota in each category was calculated according to the 2010 Beijing population census.
#### **Appendix 5: Quantitative Convenience Sample Stratifed by Quota - Birth Cohort 1950–1959 (Targeted Number of People)**
The distribution of quota in each category was calculated according to the 2010 Beijing population census.
#### **Appendix 6: Detailed Description of the Qualitative Sample (1978–1993)**
### **Appendix 7: Functional Diagram of the Matrix Control Procedure**
# **Bibliography**
Anagnost, A. (2004). The corporeal politics of quality (Suzhi). *Public Culture, 16*, 189–208.
© The Editor(s) (if applicable) and The Author(s) 2024 167
S. V. Constantin, *A Life Course Perspective on Chinese Youths*, Life Course
Research and Social Policies 17, https://doi.org/10.1007/978-3-031-57216-6
| doab | 2025-04-07T03:56:58.116903 | 14-6-2024 06:12 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "00210593-c9b9-4d6f-9f8b-238faf7dc64f",
"url": "https://library.oapen.org/bitstream/20.500.12657/90953/1/978-3-031-57216-6.pdf",
"author": "Constantin, Sandra V.",
"title": "A Life Course Perspective on Chinese Youths",
"publisher": "Springer Nature",
"isbn": "",
"section_idx": 0
} |
0024144a-a6ce-4490-ad8f-59c145166ff3 | # The Ik language
## Dictionary and grammar sketch
Terrill B. Schrock
African Language Grammars and Dictionaries 1
## African Language Grammars and Dictionaries
Chief Editor: Adams Bodomo Editors: Ken Hiraiwa, Firmin Ahoua
In this series:
# The Ik language
## Dictionary and grammar sketch
Terrill B. Schrock
Terrill B. Schrock. 2017. *The Ik language*: *Dictionary and grammar sketch* (African Language Grammars and Dictionaries 1). Berlin: Language Science Press. This title can be downloaded at: http://langsci-press.org/catalog/book/98 © 2017, Terrill B. Schrock Published under the Creative Commons Attribution 4.0 Licence (CC BY 4.0): http://creativecommons.org/licenses/by/4.0/ ISBN: 978-3-944675-95-4 (Digital) 978-3-944675-96-1 (Hardcover) 978-3-944675-68-8 (Softcover) 978-1-544669-06-9 (Softcover US) DOI:10.5281/zenodo.344792
Cover and concept of design: Ulrike Harbort Typesetting: Sebastian Nordhoff, Terrill B. Schrock Illustration: Monika Feinen Proofreading: Ahmet Bilal Özdemir, Andreas Hölzl, Bev Erasmus, Christian Döhler, Claudia Marzi, Don Killian, Eitan Grossman, Esther Yap, Greg Cooper, Kilu von Prince, Mykel Brinkerhoff, Rosey Billington, Steve Pepper, Tom Gardner Fonts: Linux Libertine, Arimo, DejaVu Sans Mono Typesetting software: XƎLATEX
Language Science Press Unter den Linden 6 10099 Berlin, Germany langsci-press.org Storage and cataloguing done by FU Berlin
Language Science Press has no responsibility for the persistence or accuracy of URLs for external or third-party Internet websites referred to in this publication, and does not guarantee that any content on such websites is, or will remain, accurate or appropriate.
For Amber Dawn
#### Contents
### Contents
## **Preface**
When I first heard about the Ik back in September 2005, I was thoroughly intrigued. Here was a people just emerging from the mists of time, from that now dark and shrouded realm of African prehistory. Judging by appearances, their journey had not been easy. Their story spoke of great suffering in the form of sickness, suppression, starvation, and slaughter. And yet, somehow, there they were, limping into the 21st century as survivors of conditions most of us cannot imagine. Having grown up in a safe and serene community in the American South, I thought the Ik seemed stranger than fiction. People like this actually exist out there? I found myself wanting to know more about them, wanting to know who they are. Subconsciously I sensed that anyone who could endure what they had endured could perhaps teach me something about being truly human.
My quest to know the Ik has led me down a winding path to the present. Over the years I have been frustrated by my inability to enter fully into their world, to see reality through their eyes. More than once I wished I were an anthropologist, so I could get a better grasp of their essence as a people. But time and time again, life steered me right back to the language – to *Icétôd*. I gradually learned to accept that their language is a doorway to their spirit, and that as a linguist, I could only open the door for others, and point the way to the Promised Land while I remain at the threshold. This book can act as a key to that door, a key that has been carefully shaped and smoothened by hands tired yet trembling with purpose.
Living in Ikland has taught me a lot about being human, but not in the way I expected. It was not by becoming 'one with the people' that I learned what it is like to survive subhuman conditions. And it was not physical starvation, or sickness, or slaughter that I was forced to endure. No, I was spared those things. Yet all the same, in Ikland I became acquainted with spiritual starvation, social sickness, and the wholesale slaughter of my cultural, religious, and intellectual idols. And just as the Ik have learned that life does not consist in 'bread alone', nor in health, nor in security – but can carry on living with dignity and humanity – I have learned that at the rock bottom of my soul, where my self ends and the world begins, there is where Life resides. That realization is my 'pearl of great price' for which I have sold everything else and would do it all over again.
## **Acknowledgements**
Compiling a dictionary such as this one is a massive undertaking, far more so than I ever imagined it would be. Although I myself spent many hours, days, and months working alone on this project, a whole host of people put me in a position to do so. And it is here that I wish to acknowledge and thank them all.
First, I want to express a heartfelt *Ɨlákásʉƙɔtíàà zùkᵘ* to all the Ik people of the Timu Forest area for welcoming us into their community and patiently putting up with the long process of a foreigner trying to learn their language. To the following Ik men and women, I give thanks for their participation in a word-collecting workshop that took place in October 2009, during which roughly 7,000 Ik words were amassed: Ariko Hillary, Kunume Cecilia, Lochul Jacob, Lokure Jacob, Longoli Philip, Losike Peter, Lotuk Vincent, Nakiru Rose, Nangoli Esther, †Ngiriko Fideli, Ngoya Joseph, Ochen Simon Peter, Sire Hillary, and Teko Gabriel.
A second group of Ik men are sincerely thanked for giving me a clearer view of the Ik sound system and for helping me edit several hundred words during an orthography workshop in April 2014: Amida Zachary, Dakae Sipriano, Lokauwa Simon, Lokwameri Sylvester, Lomeri John Mark, Longoli Philip, Longoli Simon, Lopeyok Simon, †Lopuwa Paul, and Lotuk Paul. One of those men, Longoli Philip, deserves special thanks for the years he spent as my main guide into the grammar and lexicon of his mother tongue. The number and quality of entries in this book are owed in large part to his skillful labors. Four other men – Lojere Philip, †Lochiyo Gabriel, Lokwang Hillary, and †Lopuwa Paul – also deserve my thanks for teaching me bits of the language at various points along the journey.
But it is another group of Ik men that I wish to give special honor. These are the ones who for an entire year went with me through every word in this dictionary to refine their spellings and define their meanings. They include the respectable elders Iuɗa Lokauwa, Locham Gabriel, and Lemu Simon, as well as our translators Kali Clement, Lotengan Emmanuel, and Lopeyok Simon. The three elders not only shared their intimate knowledge of the language with me but also befriended me with a grace and humility that can only come with age. Every moment I spent with them was a blessing I will never forget. As they said, if I ever come back, I should ask if those old men are still around. I pray they are.
#### Acknowledgements
Although teaching foreigners Ik-speak has usually been the domain of men, I wish to bring special attention to two Ik women who, through their resilient friendship and lively conversation, greatly enhanced my grasp of the language. These are the highly esteemed Nachem Esther and Nakiru 'Akóóro' Rose.
Next, I want to gratefully mention those in the long line of linguists who worked on the Ik language and – in person or publication – passed their knowledge down to me: Fr. J. P. Crazzolara who wrote the first recorded grammatical description of the language; A. N. Tucker whose series of articles on Ik expanded my knowledge considerably; Fritz Serzisko who penned several insightful articles and books on Ik and Kuliak; Bernd Heine who wrote numerous works on Ik and Kuliak and authored a grammar sketch and dictionary of the language (1999); Richard Hoffman who studied the grammar and lexicon, devised a practical orthography, and tirelessly supported language development efforts on behalf of the Ik; Christa König who wrote several articles and an entire book on the Ik case system; Ron Moe who helped me lead a word-collection workshop; Keith Snider who trained me in tone analysis; Kate Schell who collected dozens of hours of recorded Ik texts; and Dusty Hill who supervised me throughout the process.
Another group I wish to thank are our friends and family members whose generous and faithful donations have made it possible for us to live and work in Uganda since 2008. It has been a privilege to be financially supported in doing long-term work on the Ik language, and I do not take that for granted.
For all their hard work pushing this project through to completion, I gratefully acknowledge the series editors: Adams Bodomo, Ken Hiraiwa, and Firmin Ahoua. My sincere thanks also go the reviewers and proofreaders who helped me improve this manuscript, to Monika Feinen for drawing up a lovely map of Ikland (Figure 1), and lastly to Sebastian Nordhoff, whose patient help and technical expertise in manuscript preparation I could never have done without.
I also want to thank my dear family: my two adopted Ik daughters, Kaloyang Mercy and Lemu Immaculate, and my wife Amber Dawn. Their loving presence enabled me to carry out this long work in an otherwise isolated and often very lonely environment. The existence of this book is owed in large measure to Amber's innumerable sacrifices big and small. It came into being at great cost to her. For that and many other reasons, I thank her from the bottom of my heart.
Above all else, I want to praise the God whose Word became flesh – ὁ λόγος σὰρξ ἐγένετο – making a linguistic cosmos where my mind and the Ik language could collide and radiate bright rays of new knowledge out into the world.
## **Abbreviations**
Abbreviations
## **Part I Introduction**
## **1 The Ik language**
Ik is the native language of the Ik people who live on a narrow swath of land in the northeastern corner of Uganda, East Africa. The people call their language *Icétôd*, which means 'Ik-speech' or 'Ik-talk' and is pronounced *ee-CHAY-TOad* or in phonetic symbols, [ītʃétôd]. Ik belongs to a small cluster of languages called ̻ 'Kuliak', which also includes Nyang'ía of Lobalangit and Soo/Tepeth of Mounts Moroto, Napak, and Kadam – all in Uganda's magnificent Karamoja Region.
At the outset, let me state definitively that Ik is *not* a dialect of Karimojong, nor is it even Nilotic or 'Hamitic'. And it is certainly not Bantu (as some have asked me). Scholars disagree as to whether it is related to Karimojong at all, but if it is, it would be a distant relationship within the great Nilo-Saharan language family, much as English is related to Russian or Hindi within Indo-European.
One reason people assume Ik is a dialect of Karimojong is that the Ik have long been surrounded and dominated by the pastoralist Dodoth, Toposa, Turkana, and Jie. These groups, as well as the Karimojong proper, all speak mutually intelligible forms of a speech variety called 'Ateker', 'Teso-Turkana', or 'Tunga'. Another reason Ik seems similar to Karimojong is that it has borrowed many hundreds of words from Teso-Turkana speech varieties over the centuries. In addition to lexical borrowing, the close contact between the Ik and Teso-Turkana peoples has caused Ik grammar to become more like Teso-Turkana in various ways.
But despite the many superficial similarities one may see between Ik and Teso-Turkana, their grammatical systems are actually quite different. For instance, while their vowel inventories are similar, Ik has many more consonants than Teso-Turkana, including the ejectives /ƙ/ and /ts'/, which are found in no other Ugandan language. Ik also has an elaborate case system with eight cases all marked with suffixes, whereas Teso-Turkana languages mark only four cases, some using only tone to do so. And although both Ik and Teso-Turkana order their words as Verb-Subject-Object in main clauses, in subordinate clauses, Ik changes the order to Subject-Verb-Object. These are but a few examples among others that show the significant differences between Ik and Teso-Turkana.
## **2 The dictionary**
This book contains a bilingual Ik-English dictionary and an English-Ik reversal index. The dictionary section lists all the Ik words I have recorded up to now and offers English definitions for them. Including proper names, there are approximately 8,700 entries in the dictionary. While I have done all I could to collect as
many words as possible within the limits of time and resources, no doubt many hundreds of other words still lurk out there in the recesses of Ikian minds. It will not be until more texts are written in Ik that these missing words might be gently coaxed out onto the page and into more books like the present one.
Although the presumed purpose of a dictionary is to propound the current meanings of the words of a language, I fear that purpose is only partly achieved in this volume. The true meanings of words are lived meanings, intended by living beings in a living world. To capture them on a page is to encase them in black rock and white ice. A native speaker of Ik may recognize in my English definitions familiar traces of true meaning but never all of it. As a foreign, nonnative speaker of the language, my grasp of the living meanings of Ik words is severely limited. For the only way to learn living linguistic meanings is to experience life linguistically, *through* a language, through its words and phrases and tropes. Still, I have been fortunate enough to have had a few real-life experiences in Ik, for instance, when I learned the living meaning of the verb *ɨsɛɛs* 'to miss' by actually missing a bushpig boar as I tried to spear it when it charged toward me out of a thicket. The young Ik hunters never let me forget that miss, and as they retold the story with glee, they always used that particular verb. So when I hear it, I not only know what it means in terms of 'missing', but I also *feel* the living overtones that include shame, regret, loss of opportunity, diminution of manhood, and so on. *That* is how one learns the meanings of words.
Due to the exceptional nature of such experiences, most of the Ik words in this volume I have had to define extrinsically, from the outside. Unfortunately, as a foreign lexicographer, *I do not inhabit the words*. All I could really do was try to understand the words as best I could and render them in perspicacious English, marking out a felicitous meeting place between two very different modes of linguistic being-in-the-world. To the degree that I succeeded in this endeavor, this is what I hope to be a worthwhile first full-scale Ik-English lexicon.
The English definitions the reader will find are of various types. Some Ik words lend themselves easily to one-word, entirely accurate glosses, for example, *gʉɓɛ́rá-* as 'leopard'. Others require a short phrase in English, for instance, *ƙóré*as the 'back of the knee'. Still others, the ones that are conceptually more distant from English, call for longer descriptions, as when *makúlí-* is defined as a 'round grass beehive cover that goes over the end of a hollow beehive'.
As well as being a record of modern Ik to be used for modern purposes, this dictionary also provides much data for historical research. Because the Ik have left little in the way of archaeology over the ages, and because oral histories tend to be vague, inconsistent, and undated, language is one of the few lenses through
#### 2 The dictionary
which to investigate prehistory. Already the Ik lexicon gives some tantalizing hints as to the ancient northern East African origins of the Ik, for example in the link between words like *sɔkɔ́-* 'hoof' and Arabic *saaq* 'foot' and Gumuz *tʃagw* 'foot', or between *ƙídz-* 'bite' and Maltese Arabic *gidem* 'bite' and Uduk *k'ūcūr* 'suck'. Every Ik word is a cultural relic, a linguistic artifact sticking out of the red clays of time and memory. Each one has been molded by a million mouthings, much as grains of sand are ground down by wind and water. Each has its own history, an origin and a tortuous path of descent to its present form, the same path, we can assume, that its many speakers have taken. This is where the fields of etymology and historical linguistics (or 'paleolinguistics') can provide some evidence on which to build a grounded sense of identity and cultural history.
A deeply rooted sense of history and identity is important because it could help give the Ik a more sure footing as they transition into a nationally-minded Ugandan society and a globally-minded international society. If I imagine the future fate of the Ik language, I can see two possible developmental paths it could take. The first is that it could be lost by being totally assimilated by Karimojong, much like Nyang'ía already has and Soo/Tepeth is in danger of doing, or by succumbing to the dazzling promise of upward mobility that English seems to offer. If either of these forms of language death should take place, at least this book would remain as a monument to a once noble language-mediated world-view.
The second path the Ik language could take into the future is the one I have often daydreamed of. It is the one that would fulfill my scholarly strivings and confirm my greatest hopes for the Ik. In this path, Ik would go on to become the language of a highly literate populace who would use it skillfully to promote their own well-being. With explicit knowledge of their grammar and lexicon, educated Ik people would harness the expressive power of their native-born tongue and make it a vehicle of music, poetry, fiction, philosophy, theology, medicine, education, policy – the full gamut of human expression. This scrappy language that has barely scraped by countless threats to its existence yet somehow managed to pull through, this language that contains the linguistic genes of so many other languages from unrelated stocks, this small language of a small people in small place, could go on to become an enduring symbol of the Ikian spirit.
As portrayed in Figure 1, the Ik language area can be viewed imaginatively from an 'Ik-centric' perspective as a 'heart' of East Africa. There it lies, near the arterial convergence of four East African nations: Uganda, South Sudan, Ethiopia, and Kenya. Over the centuries the Ik have migrated through and throughout each of these four countries. While doing so, their language absorbed words and grammatical traits from the many languages spoken there. So, in a real sense, Ik
Figure 1: Ik language area in an 'Ik-centric' perspective (CC-BY Monika Feinen)
#### 3 Using the dictionary
embodies the linguistic heritage of northern East Africa. Thus, could it be that Ik is providentially situated to blossom into a language that can serve the full range of communicative needs of a modernized Ik society, and then extend its fruited boughs over the escarpment in four directions to become a blessing to the neighboring nations? In the end, only time will tell, and yet it is toward the fulfillment of that dream that this work on Ik has been lovingly consecrated.
## **3 Using the dictionary**
## **3.1 Writing system**
The Ik script used in this dictionary and grammar sketch is based on what is called the Linguistic Orthography (LingO) as described in Schrock (2015). The LingO is a compromise between the simpler Popular Orthography (PopO) and a more scientific writing system. The main reason for choosing the LingO over the PopO is that the LingO encodes three very important features of the Ik sound system: voiceless vowels, vowel harmony, and tone. Although these three features are difficult to remember and write, they are indispensable for the correct pronunciation of Ik. Therefore it was decided that for this book to be an accurate and reliable record of the language, the proper pronunciations would have to be reflected in the spellings. LingO writing can easily be converted to PopO, but the reverse is not true, since it requires greater linguistic awareness.
The alphabetical order of Ik letters is given below. Note that the vowel pairs E/Ɛ, I/Ɨ, O/Ɔ, and U/Ʉ – whose two members differ only in terms of a linguistic feature called Advanced Tongue Root [ATR] – are alphabetized as if they were the same letter. This is done to assist non-native speakers of Ik in finding words beginning with vowels they might not be able to distinguish at first. Also note that the letter (Ʒ) is in parentheses because even though it belongs to the alphabet, no recorded Ik word begins with it. For the pronunciation of these letters, the reader is referred ahead to §2.1 of the grammar sketch section.
• Ik alphabetical order: A B Ɓ C D Ɗ Dz E/Ɛ F G H Hy I/Ɨ J J' K Ƙ L M N Ɲ Ŋ O/Ɔ P R S T Ts Ts' U/Ʉ W X Y Z (Ʒ)
## **3.2 Structure of entries**
The Ik-English dictionary section contains entries of the following kinds of Ik words: nouns, pronouns, demonstratives, quantifiers, numerals, prepositions, verbs, adverbs, ideophones, interjections, nursery words, complementizers, and
connectives (or conjunctions). For a brief description of each word class, the reader is referred to §3 of the grammar sketch at the back of the book. The goal of the present section is to explain to the user the structure of lexicographical entries. To do this, an example of a noun entry and a verb entry are discussed.
A typical noun entry has several components. To identify them, match the numbered components in this explanatory paragraph with the superscript number in the model entry below. 1) The lexical headword is in bold typeface. It is the citation form of the noun, that is, the form of the noun spoken in isolation. In Ik, the citation form takes the nominative case (see §7.2). 2) The root or lexical form is in parentheses. It is hyphenated to show that it still needs a case ending, and it is the form on which to base all other case forms of this noun. This particular noun is also hyphenated in the middle to signify that it is a compound noun made of two parts (see §4.3 of the grammar). 3) This is an abbreviation for 'plural', indicating that the next item is the plural form of the headword. 4) This is the plural form of the singular headword *bàdìàm*. 5) This number (1) indicates that what follows is the first and primary sense or meaning of the headword. 6) This is an abbreviation of the grammatical category of the word, in this case *n.* for 'noun'. 7) After the primary sense, one or more other numbered senses of the word may be added. 8) After the senses, one or more notes may mention further information about the entry, for example cultural details or suggestions for synonyms or near-synonyms.
**<sup>1</sup>bàdìàm** <sup>2</sup> (bàdì-àmà-) <sup>3</sup>*pl.* <sup>4</sup>badiikᵃ <sup>5</sup> 1) <sup>6</sup>*n.* sorcerer, wizard <sup>7</sup> 2) anything spooky, weird, or uncanny | <sup>8</sup>The concept of *bàdìàm* includes nocturnal animals like bats, hyenas, and owls that have strange characteristics… tobacco is also called *bàdìàm* because its strong physiological effects are not attributable to human agency.
A typical verb entry has similar components but also some different ones: 1) Just as with nouns, the verbal headword is shown in bold typeface. This is the citation form of the verb, which in Ik appears in the infinitive form and nominative case (see §8.2 in the grammar). As an infinitive, the verb is acting as a noun at this point, much like 'to go' or 'going' in English. To use an Ik infinitive as a verb, simply remove the infinitive suffix (either *-ònì-* or *-ésí-*) and use the appropriate suffixes (see §8.7). 2) Then, the form in the parentheses is the lexical form of the infinitival headword, the one that is the base for all other case-inflected forms of the verb. 3) This number (1) indicates that what follows is the first and primary sense or meaning of the headword. 4) This is an abbreviation of the grammatical category of the headword, in this case *v.* for 'verb'. 5) After the
3 Using the dictionary
primary sense, one or more other senses of the headword may be added. 6) This short note directs the user to a synonym or near-synonym of the headword.
**<sup>1</sup>betsínón** <sup>2</sup> (betsínónì-) <sup>3</sup> 1) <sup>4</sup> *v.* to be awkward, gauche, inept <sup>5</sup> 2) to be left-handed, sinistral | <sup>6</sup>See also *ɨɓaŋɨɓáŋɔn*.
Over a hundred Ik verb roots end in /a/, /e/, or /ɛ/, meaning that when an infinitive suffix is added to the root, these root-final vowels are assimilated (see §2.4.4). For example, though the root for 'miss' is *ɨsá-*, the infinitive form is *ɨsɛɛs*, which obscures the root-final vowel. Lest the dictionary user hear a form of the root *ɨsá-* in speech and then fail to deduce that its infinitive is *ɨsɛɛs*, both root and infinitive have been listed in the dictionary. The entry for *ɨsá-* includes the notation (<ɨsɛɛs) which indicates that *ɨsɛɛs* is the entry the user should go to for the definition. Conversely, the entry for *ɨsɛɛs* 'to miss' includes both the lexical form of the infinitive and the bare root, as in: **ɨsɛɛs** (ɨsɛɛsí-/ɨsá-).
## **3.3 Tips for finding words**
Finally, because a good number of Ik words have more than one form, and because many of them can be reasonably spelled in multiple ways, let me offer the user the following tips for locating polymorphous words in the dictionary:
## **Part II**
## **Ik-English dictionary**
#### aaii
## **a**
**àdɛ̀nɛ̀s** (àdɛ̀nɛ̀sà-) *n*. bird species.
**àɗ<sup>e</sup>** (àɗè) *num*. three.
**aƙʉ́ƙʉ́rɔ̀n** (aƙʉ́ƙʉ́rɔ̀nì-) *v*. to creep.
**aƙóláánón** (aƙóláánónì-) *v*. to swing.
**Amérìkààm** (Amérìkà-àmà-) *pl.* **Amérikaik<sup>a</sup>** . *n*. American.
**amʉtsanés** (amʉtsanésí-) *v*. to collect on a debt.
**anás** (anásí-) *n*. male greater kudu. *Tragelaphus strepsiceros*.
**ànɛ̀** (ànɛ̀ɛ̀-) *n*. vine species whose tuberous roots are peeled and eaten raw or roasted and whose beanlike seeds are cooked and eaten. *Vigna frutescens*. Possibly the same vine as *málákʉ́r*.
**anɛ́sʉ́ƙɔt<sup>a</sup>** (anɛ́sʉ́ƙɔtí-) *v*. to recall, remember (often with regret).
**anɛtɛ́s** (anɛtɛ́sí-) *v*. to recall, recollect, remember. See also *tamɛtɛ́s*.
**aniesúƙot<sup>a</sup>** (aniesúƙotí-) *v*. to recall repeatedly (often with regret).
**aŋaras** (aŋarasá-) *n*. gravel.
**aŋarasááƙw<sup>a</sup>** (aŋarasá-áƙɔ̀-) *n*. gravelly area.
**Aŋatár** (Aŋatárì-) *n*. name of a hill or mountain.
**aŋaw<sup>a</sup>** (aŋaú-) *n*. yellowish tobacco leaves.
**aŋɨrɛs** (aŋɨrɛsí-) *1 v*. to turn, twist. *2 v*. to steer (a vehicle).
**aŋiriesón** (aŋiriesónì-) *v*. to swerve or veer repeatedly.
**Aŋolekók<sup>a</sup>** (Aŋolekókò-) *n*. name of a hill or mountain.
**Apáálokiɓúk<sup>a</sup>** (Apáálokiɓúkù-) *n*. personal ox-name of a colonial British District Commissioner.
**Apáálokúk<sup>a</sup>** (Apáálokúkú-) *n*. name of an Italian priest (Father Daniel) who founded the Kaabong Catholic mission and was killed by the Turkana.
**Apáálomúƙ<sup>a</sup>** (Apáálomúƙú-) *n*. a personal name.
**Apáálòŋìrò** (Apáálòŋìròò-) *n*. a personal name.
**Apáásiá** (Apáásiáà-) *n*. a personal name. **apápánà** (<apápánɔ̀ɔ̀n) *v*.
**Apʉs** (Apʉsí-) *n*. a personal name.
**aráɡwaníékw<sup>a</sup>** (aráɡwaní-ékù-) *n*. full moon. *Lit.* 'moon-eye'.
**Aramasán** (Aramasánì-) *n*. personal name of a Bokora man who settled in Ikland and married an Ik.
**Árápííʝí** (Árápííʝíì-) *n*. place named after rocket-propelled grenades (RPG).
**arasí** (arasíì-) *n*. councillor in the Local Council I (LCI), an administrative unit in the Ugandan government at the village level.
**arétón** (arétónì-) *v*. to cross (this direction, to this side).
**arí** (aríɛ́-) *pl.* **aríík<sup>a</sup>** . *n*. section of the small intestine.
16
**arír** (arírá-) *pl.* **arírík<sup>a</sup>** . *n*. flame mixed with smoke.
**aukes** (aukesí-) *v*. to fill (one's mouth) with drink before/without swallowing.
**aw<sup>a</sup>** (awá-) *pl.* **àwìk<sup>a</sup>** . *1 n*. abode, home, homestead, manyatta, village. *2 n*. place.
**awa ná zè** *1 n*. big home or village. *2 n*. capital city. *3 n*. Heaven.
**awa Ɲákuʝí** *n*. Heaven.
#### bàbà
## **b**
**bɛná** (<bɛnɔ́ɔ́n) *v*.
**bɛrɛtɛ́sá mɛná<sup>ɛ</sup>** *v*. to come to a consensus.
**bɨrá** (<bɨrɔ́ɔ́n) *v*.
**biy<sup>a</sup>** (biyá-) *n*. outside.
**biyáxán** (biyá-xánà-) *n*. outside.
**botibotosíám** (botibotosí-ámà-) *pl.* **botibotosíík<sup>a</sup>** . *n*. drifter, migrant, nomad.
**botitín** (botitíní-) *n*. baggage, cargo, luggage.
**bòtòn** (bòtònì-) *v*. to migrate, move.
**botonuƙot<sup>a</sup>** (botonuƙotí-) *v*. to migrate or move away.
**bótsón** (bótsónì-) *1 v*. to be clear, open, vacant. *2 v*. to be empty, hollow.
**bótsóna iká<sup>e</sup>** *v*. to be clear, sober (of one's mind).
**bóx** (bóxá-) *1 n*. nightjar. *2 n*. idiot, moron, stupid person.
**boxoƙorét<sup>a</sup>** (boxoƙorétí-) *pl.* **boxoƙorétík<sup>a</sup>** . *n*. tall softwood tree species whose bland, red berries are eaten by children and whose wood is carved into bowls and cups. *Cussonia arborea*.
**bú** (búá-) *n*. airborn dust, dust cloud.
**buanítésuƙot<sup>a</sup>** (buanítésuƙotí-) *v*. to lose, hide, make disappear, misplace.
**buanón** (buanónì-) *v*. to be lost, disappeared, misplaced.
**buanónuƙot<sup>a</sup>** (buanónuƙotí-) *v*. to disappear, fade, evaporate, get lost.
**búbuiem** (búbui-emé-) *n*. back part or underpart of an animal's leg, from the ankle to the thigh, which is the women's special cut of meat.
**búdès** (búdèsì-) *1 v*. to bury, inhume, inter, lay to rest. *2 v*. to conceal, hide. See also *muɗés* and *tʉnʉkɛs*.
**búdòs** (búdòsì-) *v*. to be concealed, covert, hidden, private, secret.
**bùɗàm** (bùɗàmà-) *n*. darkness.
**bùf** (bùfù) *ideo*. spongily.
**bufúdòn** (bufúdònì-) *v*. to be spongy.
# **ɓ**
**ɓɛkánón** (ɓɛkánónì-) *v*. to be incitive, inflammatory, provoking, rankling.
**ɓɛk<sup>ɛ</sup>** (ɓɛkɛ) *ideo*. snap! (sound of something thin snapping).
**ɓɛkɛ́s** (ɓɛkɛ́sí-) *v*. to perforate, puncture.
**ɓɛkɛtɛ́s** (ɓɛkɛtɛ́sí-) *1 v*. to perforate, puncture. *2 v*. to incite, provoke, rankle.
**ɓɛkɛ́tɔ́n** (ɓɛkɛ́tɔ́nì-) *v*. to hatch (of chicks). See also *ɨɓɛ́ɓɛ́ɛ̀tɔ̀n*.
**ɓɛkɨɓɛ́kɔ́n** (ɓɛkɨɓɛ́kɔ́nì-) *v*. to rustle.
**ɓɛƙɛ́s** (ɓɛƙɛ́sí-) *1 v*. to walk. *2 v*. to travel. *3 v*. to move.
**ɓɛƙɛ́sá buɗámík<sup>e</sup>** *v*. to move blindly.
**ɓɛƙɛ́sá kútúŋìk<sup>o</sup>** *v*. to walk on the knees.
**ɓɛƙɛ́sá kwɛ̀tìk<sup>ɔ</sup>** *v*. to walk on the hands.
**ɓɛƙɛ́sá turúùk<sup>e</sup>** *v*. to stumble ahead.
**ɓɛƙɛ́sá wɛwɛɛs** *v*. to walk leisurely.
**ɓɛƙɛ́sá ziál** *v*. to walk laboriously (like an obese or pregnant person).
**ɓɛƙɛ́síàm** (ɓɛƙɛ́sí-àmà-) *pl.* **ɓɛƙɛ́síik<sup>a</sup>** . *1 n*. pedestrian, walker. *2 n*. traveler, wayfarer.
**ɓɛƙɛ́síama mukú** *n*. one who walks at night (like a lover or wizard or merely someone who has not reached their destination by dark).
**ɓɛƙɛ́síkabáɗ<sup>a</sup>** (ɓɛƙɛ́sí-kabáɗá-) *pl.* **ɓɛƙɛ́ síkabáɗík<sup>a</sup>** . *n*. identity card, passport.
**ɓɛlɛ́ɓɛ́lánón** (ɓɛlɛ́ɓɛ́lánónì-) *v*. to be chapped, cracked, split open.
**ɓelémón** (ɓelémónì-) *1 v*. to crack or split open (like burnt skin). *2 v*. to break (of day), dawn.
**ɓɛlɛ́rɛ́mɔ̀n** (ɓɛlɛ́rɛ́mɔ̀nì-) *v*. to be bugeyed.
**ɓɛlɛ́s** (ɓɛlɛ́sí-) *v*. to crack, split.
**ɓɛtɛ́lɛ́mɔ̀n** (ɓɛtɛ́lɛ́mɔ̀nì-) *v*. to be flatly or shallowly concave. See also *fɛtɛ́lɛ́mɔ̀n*.
**ɓɔɗáʝʉ́m** (ɓɔɗá-ʝʉ́mʉ̀-) *n*. dirt mixed with threshed grain that is then sifted.
**ɓɔ́l** (ɓɔ́lá-) *pl.* **ɓɔ́lítín**. *n*. shin.
**ɓʉ̀nɔ̀n** (ɓʉ̀nɔ̀nì-) *v*. to move past, pass by.
**ɓúrukúkón** (ɓúrukúkónì-) *v*. to germinate, sprout.
**ɓútánés** (ɓútánésí-) *v*. to have sex repeatedly and often.
**ɓut<sup>u</sup>** (ɓutu) *ideo*. all, entirely.
**ɓutúrúmòn** (ɓutúrúmònì-) *v*. to be bulky, hulky.
**ɓuumón** (ɓuumónì-) *v*. to get dislocated, out-of-joint.
#### caál
## **c**
*3 n*. fuel: diesel, paraffin (kerosene), petrol (gas).
**cicianón** (cicianónì-) *v*. to reform, repent.
**cicídè** (cicídèà-) *n*. bird species.
**cíkóróìkànànès** (cíkóróìkànànèsì-) *n*. boundedness, having boundaries.
**cíkóroy<sup>a</sup>** (cíkóroí-) *pl.* **cíkóróìk<sup>a</sup>** . *n*. border, boundary, limit.
**cikw<sup>a</sup>** (cikó-) *pl.* **cikóík<sup>a</sup>** . *n*. male animal.
**cìɔ̀n** (cìɔ̀nì-) *v*. to be full, sated, satiated, satisfied.
**cɨɔnʉƙɔt<sup>a</sup>** (cɨɔnʉƙɔtí-) *v*. to become full, sated, satiated.
**Cɔ́ƙɔ́tɔ̀m** (Cɔ́ƙɔ́tɔ̀mɛ̀-) *n*. Dodoth people.
**Cɔ́ƙɔ́tɔ̀mɛ̀àm** (Cɔ́ƙɔ́tɔ̀mɛ̀-àmà-) *n*. Dodoth person .
**còòkààm** (còòkà-àmà-) *pl.* **cookaik<sup>a</sup>** . *1 n*. cowherd, shepherd. *2 n*. guard, watchman. *3 n*. defender, guardian, protector.
**cookaama zíkɛ́siicé** *n*. prison guard.
**cookaika ínó<sup>e</sup>** *n*. wildlife authorities. *Lit.* 'guardians of animals'.
**cookés** (cookésí-) *1 v*. to shepherd, tend (livestock). *2 v*. to defend, guard, protect.
**cookotós** (cookotósí-) *v*. to be guarded, protected, tended.
**coór** (coorí-) *pl.* **coorík<sup>a</sup>** . *n*. leg rattle tied below the knee.
**cooríɡwà** (coorí-ɡwàà-) *n*. bee-eater. *Lit.* 'rattle-bird'. *Merops sp*. See also *keseníɡwà*.
**cuáák<sup>a</sup>** (cuá-ákà-) *n*. permanent water source (like a spring or well). *Lit.* 'water-mouth'.
**cuanón** (cuanónì-) *v*. to be fluid, liquid.
**cuanónuƙot<sup>a</sup>** (cuanónuƙotí-) *v*. to become liquid, liquify, melt.
**cuc<sup>u</sup>** (cucu) *ideo*. very black.
**cucue** (cucué-) *n*. damp chill.
**cucuéétòn** (cucuéétònì-) *1 v*. to cool down/off (of pain or weather). *2 v*. to feel mercy, sympathize.
**Cùcùèìk<sup>a</sup>** (Cùcùè-ìkà-) *n*. name of a hill or mountain.
**cucuéítésuƙot<sup>a</sup>** (cucuéítésuƙotí-) *v*. to chill, cool down.
**cucuéítésuƙota ɡúró<sup>e</sup>** *v*. to calm or cool down one's heart.
**cucuéón** (cucuéónì-) *1 v*. to be chilly, cool. *2 v*. to be weak . *3 v*. to be bland, stale.
**cucuéónuƙot<sup>a</sup>** (cucuéónuƙotí-) *v*. to cool down/off (of pain or weather).
**cue** (cué-) *1 n*. liquid, water. *2 n*. baby girl. *3 n*. taboo of failing to give water to the elders first.
**cúédòm** (cúé-dòmà-) *pl.* **cúédomitín**. *n*. water pot used to keep clean water inside the hut.
**cueina mɛ́sɛ̀** *n*. beer leftover from a ceremony or group work-day. *Lit.* 'waters of beer'.
**cúémúcè** (cúé-múcèè-) *pl.* **cúémúcèìk<sup>a</sup>** . *n*. ditch, watercourse, waterway.
**cúénêb<sup>a</sup>** (cúé-nébù-) *pl.* **cúénébitín**. *n*. body of water.
**cúkúɗùm** (cúkúɗùmù-) *n*. male mountain reedbuck. *Redunca fulvorufula*.
**Curuk<sup>a</sup>** (Curukú-) *n*. name of a hill or mountain. *Lit.* 'bull'.
**cúrúk<sup>a</sup>** (cúrúkù-) *pl.* **cúrúkaikw<sup>a</sup>** . *1 n*. bull. *2 n*. male, sire, stud.
**cúrúkà mɛ̀sɛ̀** *n*. barm or yeast used in brewing beer. *Lit.* 'sire of beer'.
hand' that may have taken on a narrower meaning for some Ik speakers.
**cwɛtéém** (cwɛté-émè-) *n*. bicep and/or tricep. *Lit.* 'upper arm flesh'.
## **d**
**dà** (<dòòn) *v*.
**daás** (daasí-) *1 n*. beauty, handsomeness, loveliness, prettiness. *2 n*. generosity, magnanimity, philanthropy. *3 n*. agreeableness, niceness, pleasantness. *4 n*. glory, radiance, splendor. *5 n*. holiness, sanctity.
**dàbìʝ<sup>a</sup>** (dàbìʝà-) *n*. bird species.
**dàƙw<sup>a</sup>** (dàƙwà) *ideo*. weakly.
**dɛááƙw<sup>a</sup>** (dɛá-áƙɔ̀-) *n*. sole of the foot.
**dɛɨkatsɨrím** (dɛɨka-tsɨrímʉ́-) *pl.* **dɛɨkatsɨrímík<sup>a</sup>** . *n*. metal anklet.
**dɛ̀ìk<sup>ɔ</sup>** *n*. by foot, on foot.
**detés** (detésí-) *v*. to bring.
rainy weather ailments like body aches and pains. *4 n*. soldiers.
**dimés** (dimésí-) *v*. to deny, refuse, reject.
**dimésá bubue ŋɔɛsí** *n*. dyspepsia, indigestion. *Lit.* 'refusal of the stomach to grind'.
**dɔ́dɔrɔnʉƙɔt<sup>a</sup>** (dɔ́dɔrɔnʉƙɔtí-) *v*. to move away on one's buttocks.
**dòɗ<sup>a</sup>** (dòɗì-) *pl.* **doɗitín**. *n*. vagina.
**dòɗìdòɗìɡwà** (dòɗìdòɗì-ɡwàà-) *n*. bird species. *Lit.* 'vagina-vagina bird'.
**dimités** (dimitésí-) *v*. to forbid, prohibit, proscribe. See also *itáléés*.
**dónésìàm** (dónésì-àmà-) *pl.* **dónésiik<sup>a</sup>** . *n*. donor, philanthropist.
**dónésuƙot<sup>a</sup>** (dónésuƙotí-) *v*. to donate, give out, present.
**dosés** (dosésí-) *v*. to thatch.
**dʉbam** (dʉbamá-) *n*. dough. *Lit.* 'kneadable'.
**dús** (dúsé-) *pl.* **dúsítín**. *n*. grassland, plain, savannah.
#### ɗa
# **ɗ**
clothing; hunters often remove clothing before trying to penetrate its brush; found down in the Turkana plains, it is used for fencing and is eaten by livestock. *Acacia senegal*. See also *lofílitsí*.
a
down low (e.g. into a deep valley). *3 v*. to screw around on an errand.
**ɗɔ̀r** (ɗɔ̀rɔ̀) *ideo*. slickly, slipperily.
**ɗorôɡ<sup>a</sup>** (ɗoróɡè-) *pl.* **ɗoróɡìk<sup>a</sup>** . *n*. roan antelope. *Hippotragus equinus*.
**ɗòs** (ɗòsì-) *n*. gum, mucilage.
**ɗɔ̀s** (ɗɔ̀sɔ̀) *ideo*. gummily.
**ɗɔsɔ́** (ɗɔsɔ́ɔ̀-) *n*. vine species whose seeds are eaten raw or cooked and whose roots are eaten raw, cooked, or roasted. *Vigna sp*.
**ɗɔsɔ́dɔ̀n** (ɗɔsɔ́dɔ̀nì-) *v*. to gummy.
**ɗotíɗótòn** (ɗotíɗótònì-) *v*. to go restlessly from place to place.
**ɗɔtɔ́**(ɗɔtɔ́ɔ̀-) *1 n*. rubbery gum (like that of barat<sup>a</sup> ). *2 n*. rubber. *3 n*. chewing gum.
**ɗɔtsánón** (ɗɔtsánónì-) *1 v*. to be joined together. *2 v*. to cooperate. *3 v*. to be compacted. See also *kumutsánón*.
**ɗɔtsánónuƙot<sup>a</sup>** (ɗɔtsánónuƙotí-) *1 v*. to join or meet together. *2 v*. to cooperate.
**ɗɔtsɛ́s** (ɗɔtsɛ́sí-) *1 v*. to add. *2 v*. to add or mix in. *3 v*. to add together, combine, join.
**ɗɔtsɛ́sá ɦyekesí** *v*. to contribute resources. *Lit.* 'to add life'.
**ɗɔtsɛtɛ́sá así** *v*. to gather oneselves.
**ɗɔtsɛtɛ́sá tódà<sup>e</sup>** *v*. to negotiate, reach a consensus. *Lit.* 'to add up speech'.
**ɗɔtsɔ́s** (ɗɔtsɔ́sí-) *1 v*. to be added. *2 v*. to be attached, joined. *3 v*. to be compacted.
**ɗòwòn** (ɗòwònì-) *1 v*. to be pristine, unspoiled, untouched, virginal (like a forest). *2 v*. to be left alone, unused (like a path or old village).
**ɗòx** (ɗòxò) *ideo*. fitly, healthily.
**ɗɔ̀x** (ɗɔ̀xɔ̀) *ideo*. unsteadily.
**ɗɔ́xɛatá na ɓɛƙɛ́s** *n*. satellite. *Lit.* 'star that moves'.
**ɗɔ́xɛatá na tsúwà** *n*. meteor, shooting star. *Lit.* 'start that runs'.
**Ɗɔ́xɛatá tsòònì** *n*. morning star.
**Ɗɔ́xɛatá xìŋàtà<sup>e</sup>** *n*. evening star.
**ɗoxódòn** (ɗoxódònì-) *v*. to be fit, healthy, in shape.
**ɗɔxɔ́dɔ̀n** (ɗɔxɔ́dɔ̀nì-) *v*. to be rickety, wobbly (of people, while walking).
**ɗɔ̀xɔ̀ƙ a** (ɗɔ̀xɔ̀ƙɔ̀-) *pl.* **ɗɔxɔ́ƙík<sup>a</sup>** . *n*. gourd flesh, pulp.
**ɗuɗuanón** (ɗuɗuanónì-) *v*. to growl, grumble, gurgle (of digestion).
**ɗʉ̀ɗʉ̀ŋ** (ɗʉ̀ɗʉ̀ŋʉ̀) *ideo*. to the end. See also *tùtùr*.
**ɗués** (ɗuesí-) *v*. to deracinate, extirpate, pull up, uproot. See also *rués*.
#### ɗutúdòn
#### dzàbùl
## **dz**
dzôɡ
a
**dzoluɡánón** (dzoluɡánónì-) *v*. to have hidden motives.
**dzòn** (dzònì-) *pl.* **dzóníkw<sup>a</sup>** . *n*. hand-dug well.
#### eakwa
## **e/ɛ**
**eɡés** (eɡésí-) *v*. to place, put.
**eɡésá kwɛtá<sup>ɛ</sup>** *v*. to sign. *Lit.* 'to put the hand'.
**eɡésá ɲáʝálaák<sup>e</sup>** *v*. to jail, put in jail.
**eɡésá zíkɛ́sìk<sup>ɛ</sup>** *v*. to imprison, jail, lock up. *Lit.* 'to put in tying'.
**Ɛ́kìtɛ̀là** (Ɛ́kɨtɛlaá) *n*. a personal name.
**ekoɗit<sup>a</sup>** (ekoɗití-) *n*. tree species whose young roots are eaten raw and whose fruit is liked by children and birds; its bark is chewed as medicine for coughing, and its wood is used to carve stools and wooden containers. *Lannea schimperi*. See also *meleke*.
**èkòn** (èkònì-) *v*. to get out of the way, move aside. Also pronounced as *ècòn*.
**ɛkwítɛ́sʉƙɔt<sup>a</sup>** (ɛkwítɛ́sʉƙɔtí-) *v*. to put ahead or in front.
**ɛ̀kwɔ̀n** (ɛ̀kwɔ̀nì-) *1 v*. to be early or first. *2 v*. to be ahead or in front.
**eletiésuƙot<sup>a</sup>** (eletiesúƙotí-) *v*. to blow, squander, waste. See also *iɲékésuƙot<sup>a</sup>* .
**em** (emé-) *1 n*. flesh, meat. *2 n*. muscle.
**emetá** (emetáa-) *pl.* **emetátikw<sup>a</sup>** . *n*. parent in-law (of men).
**eminiés** (eminiesí-) *v*. to draw apart, pull apart, stretch.
**emitaakón** (emitaakónì-) *v*. to swell (of many).
**elánétòn** (elánétònì-) *v*. to accompany here, come after/with.
**érítòn** (érítònì-) *v*. to flitter, flutter (of termite alates underground just before emerging and taking flight).
**erutánón** (erutánónì-) *1 v*. to low, moo. *2 v*. to roar (e.g. lions).
# **f**
fá
**fitetés** (fitetésí-) *v*. to wash up.
**fitídòn** (fitídònì-) *v*. to be dull (of blades).
**fìtⁱ** (fìtì) *ideo*. dully (of blades).
**fɔfɔ́ʝ a** (fɔfɔ́ʝá-) *pl.* **fɔfɔ́ʝík<sup>a</sup>** . *n*. dried fruit of the *ts'ɔƙɔm* tree.
**fɔ́fɔ́tɛ́s** (fɔ́fɔ́tɛ́sí-) *v*. to drag, grate, scrape (along the ground). See also *ɨfɔɛs*.
**fóʝón** (fóʝónì-) *v*. to whistle.
**fɔ̀k ɔ** (fɔ̀kɔ̀) *ideo*. lightweightly.
**fɔkɔ́dɔ̀n** (fɔkɔ́dɔ̀nì-) *v*. to be lightweight. See also *ɔfɔ́dɔ̀n* and *olódòn*.
**folólómòn** (folólómònì-) *v*. to be wide open (like an open road).
**fòlòn** (fòlònì-) *v*. to exuviate, molt, shed (scales or skin).
**fɔrɔ́sìtà** (fɔrɔ́sìtàà-) *n*. forest.
# **g**
**ɡaanaakón** (ɡaanaakónì-) *v*. to be bad, evil, wicked (of many).
**ɡaánàs** (ɡaánàsì-) *1 n*. badness, evil, wickedness. *2 n*. anger, annoyance, fury, rage. *3 n*. danger, peril, risk.
**ɡàànìk<sup>e</sup>** *v*. badly, poorly.
**ɡaɗikam** (ɡaɗikamá-) *n*. foraging.
**ɡáƙón** (ɡáƙónì-) *v*. to leave early (before dawn). See also *isókón*.
**ɡaƙúrúmòn** (ɡaƙúrúmònì-) *v*. to be grouchy, grumpy. See also *ŋízìmɔ̀ɔ̀n*.
**Gàlàts<sup>a</sup>** (Gàlàtsì-) *n*. name of a hill or mountain.
**ɡamam** (ɡamamá-) *n*. firewood, kindling, tinder. *Lit.* 'kindlable'.
**ɡamés** (ɡamésí-) *v*. to kindle, light or start (a fire).
**ɡamésá ts'aɗí** *v*. to kindle, light, or start a fire.
**ɡasoŋwa** (ɡaso-ŋwaá-) *n*. warthog sow. **ɡàts<sup>a</sup>** (ɡàtsà) *ideo*. rockily, stonily.
**ɡìd<sup>a</sup>** (ɡìdà-) *pl.* **ɡíditín**. *n*. cloud.
**ɡirésá mɛná<sup>ɛ</sup>** *v*. to keep one's thoughts secret.
**ɡirésíàw<sup>a</sup>** (ɡirésí-àwà-) *pl.* **ɡirésíawík<sup>a</sup>** . *n*. storage place, store (cupboard, cabinet, cave, etc.).
**ɡirú** (ɡirúù-) *n*. locust.
**ɡɔ̀ɓ a** (ɡɔ̀ɓà-) *pl.* **ɡɔ́ɓítín**. *n*. knot in wood.
**ɡɔ̀ɡɔ̀r** (ɡɔ̀ɡɔ̀rɔ̀) *ideo*. decrepitly.
**ɡòɡòròʝ<sup>a</sup>** (ɡòɡòròʝì-) *pl.* **ɡóɡòròʝìk<sup>a</sup>** . *1 n*. spine. *2 n*. midrib. *3 n*. straight middle part. *4 n*. ridge.
**ɡɔ́lìɗ<sup>a</sup>** (ɡɔ́lìɗì-) *n*. gold.
**ɡɔ́lɔ́ɡɔlánón** (ɡɔ́lɔ́ɡɔlánónì-) *v*. to be crooked, twisted (like a river or stick).
```
ɡɔn (ɡɔnɛ́-) pl. ɡɔ́nítín. n. stump.
```
#### ɡubes
# **h**
**hoetésá ɲásáatí** *v*. to plan a time. *Lit.* 'to cut out an hour'.
**hoɡwarí** (ho-ɡwaríì-) *n*. roof.
to be confused with *hɔnɛ́s*.
**hɔ́tɔ̀** (hɔ́tɔ̀ɔ̀-) *n*. bustard.
## ɦyakwés
# **ɦy**
**ɦyakwés** (ɦyakwésí-) *v*. to hush, shush.
**ɦyeésúƙot<sup>a</sup>** (ɦyeésúƙotí-) *v*. to come to know, learn, learn how.
**ɦyeímós** (ɦyeímósí-) *v*. to be kin, related. *Lit.* 'to know each other'. Also pronounced as *ɦyeínós*.
**ɦyeínósá na ƙɔ́ɓà<sup>ɛ</sup>** *v*. kinship or relation by birth. *Lit.* 'relation of the umbilical cord'.
**ɦyeínósá na séà<sup>e</sup>** *v*. to be related by blood. *Lit.* 'relation of blood'.
**ɦyeínósá sits'ésú** *v*. to be related by marriage.
**ɦyeítésihò** (ɦyeítési-hòò-) *pl.* **ɦyeítésihoík<sup>a</sup>** . *n*. examination room.
**ɦyeitésúƙot<sup>a</sup>** (ɦyeitésúƙotí-) *v*. to inform, let know, tell.
**ɦyeitetés** (ɦyeitetésí-) *v*. to inform, let know, tell.
**ɦyekétón** (ɦyekétónì-) *v*. to come back to life, resurrect, revive.
**ɦyɛ̀nɔ̀n** (ɦyɛ̀nɔ̀nì-) *v*. to barf, hurl, puke, regurgitate, vomit, upchuck.
**ɦyɔ̀** (ɦyɔ̀ɔ̀-) *n*. cattle, cow(s). Can refer to one or more cattle/cows.
## ɨáíá
## **i/ɨ**
**ɨáíá** (<ɨáíɛ́ɛ́s) *v*.
**ɨatɔs** (ɨatɔsí-) *1 v*. to added, increased. *2 v*. to be expanded, widened.
**ibét<sup>a</sup>** (ibétí-) *pl.* **íbètìk<sup>a</sup>** . *n*. thorn-tree species whose wood is used for building, fencing, and carving tool handles; its twigs are used as toothbrushes, and its red seeds are worn as beads by the Turkana. *Commiphora africana*.
**íbìrìbìròn** (íbìrìbìrònì-) *v*. to babble, blather.
**íbɨtɛ́s** (íbɨtɛ́sí-) *v*. to plant, sow.
**íbànòn** (íbànònì-) *v*. to go in the late afternoon or evening. See also *irípón*.
**íbɔtsɛ́sá así** *v*. to be agitated, churn, roil.
**ɨɓáɓá** (<ɨɓáɓɛ́ɛ́s) *v*.
**ɨɓatíɓátɔ̀n** (ɨɓatíɓátɔ̀nì-) *v*. to be hampered, handicapped, hindered.
**iɓátísa** (<iɓátíseés) *v*.
**iɓátíseés** (iɓátíseesí-/iɓátísa-) *v*. to dunk, baptize.
**ɨɓɛ́ɓɛ́**(<ɨɓɛ́ɓɔ́ɔ̀n) *v*.
**ɨɓɛ́ɓɛ́ɛsʉƙɔt<sup>a</sup>** (ɨɓɛ́ɓɛ́ɛsʉƙɔtí-) *v*. to lay (eggs).
**ɨɓɛ́ɓɛ́ɛ̀tɔ̀n** (ɨɓɛ́ɓɛ́ɛ̀tɔ̀nì-) *v*. to hatch out (of chicks). See also *ɓɛkɛ́tɔ́n*.
**ɨɓɛ́ɓɛ́lɛ́s** (ɨɓɛ́ɓɛ́lɛ́sí-) *v*. to split into pieces.
**ɨɓɛ́ɓɔ́ɔ̀n** (ɨɓɛ́ɓɔ́ɔ̀nì-/ɨɓɛ́ɓɛ́-) *v*. to hatch (of chicks).
**ɨɓɛkíɓɛ́kɛ́s** (ɨɓɛkíɓɛ́kɛ́sí-) *v*. to break or snap off (e.g. dry sticks).
**ɨɓɛ́lɛ́**(<ɨɓɛ́lɔ́ɔ̀n) *v*.
**ɨɓɛ́lɛ́ánón** (ɨɓɛ́lɛ́ánónì-) *v*. to be impetuous, impulsive.
**iɓéléés** (iɓéléésí-) *1 v*. to overturn, roll or turn over. *2 v*. to alter, change, transform. See also *bukures* and *iɓélúkéés*.
**iɓéléésuƙota así** *v*. to change one's direction, turn oneself around.
**iɓéléetés** (iɓéléetésí-) *v*. to overturn, roll or turn over.
**iɓéléìmètòn** (iɓéléìmètònì-) *1 v*. to overturn, roll or turn over (on its own). *2 v*. to change, transform. See also *iɓékúkáìmètòn*.
**iɓelíɓélésa tódà<sup>e</sup>** *v*. to change statements or the story.
**ɨɓɛ́lɔ́ɔ̀n** (ɨɓɛ́lɔ́ɔ̀nì-/ɨɓɛ́lɛ́-) *v*. to be impetuous, impulsive.
**iɓélúká** (<iɓélúkéés) *v*.
**iɓélúkáìmètòn** (iɓélúkáìmètònì-) *v*. to overturn, turn over (on its own). See also *iɓéléìmètòn*.
**iɓélúkéés** (iɓélúkéésí-/iɓélúká-) *v*. to overturn, turn over, upset. See also *bukures* and *iɓéléés*.
**ɨɓɛ́rɔ́ánón** (ɨɓɛ́rɔ́ánónì-) *v*. to be energetic, hard-working, industrious.
**ɨɓɛsíɓɛ́sɛ́s** (ɨɓɛsíɓɛ́sɛ́sí-) *v*. to break up into small pieces (like sticks for kindling).
**iɓíléròn** (iɓílérònì-) *1 v*. to be forgotten, misplaced. *2 v*. to be baffled, confused, perplexed. See also *ɡwèlòn*.
**iɓíléronuƙot<sup>a</sup>** (iɓíléronuƙotí-) *v*. to become forgotten or misplaced.
**iɓilíɓílésá así** *v*. to keep turning oneself over (e.g. in bed at night).
**ɨɓíɔ́n** (ɨɓíɔ́nì-) *v*. to shart: defecate unintentionally while passing gas.
**ɨɓɨtsíɓítsɛ́s** (ɨɓɨtsíɓítsɛ́sí-) *v*. to struggle or wriggle into (an opening).
**iɓóɓólés** (iɓóɓólésí-) *v*. to chip or pull (bark) off in pieces (rather than strips).
**iɓóɓóŋètòn** (iɓóɓóŋètònì-) *v*. to come back, return, turn around this way. See also *itétón*.
**iɓóɓóŋòn** (iɓóɓóŋònì-) *v*. to go back, return, turn around.
**iɓóɓórés** (iɓóɓórésí-) *v*. to core or hollow out.
**iɓóɓórós** (iɓóɓórósí-) *v*. to be cored or hollowed out.
**iɓokes** (iɓokesí-) *v*. to jiggle, shake.
**iɓókésuƙot<sup>a</sup>** (iɓókésuƙotí-) *v*. to shake or throw off.
**iɓoletés** (iɓoletésí-) *v*. to covenant, promise.
**iɓolíɓólés** (iɓolíɓólésí-) *v*. to plunder, ransack, rifle through.
**iɓolíɓólésuƙot<sup>a</sup>** (iɓolíɓólésuƙotí-) *v*. to plunder, ransack, rifle through.
**Icéɛ́n** (Icé-ɛ́ní-) *n*. Ik culture and/or language.
**Icéhò** (Icé-hòò-) *pl.* **Icéhoík<sup>a</sup>** . *n*. traditional beehive-shaped grass hut once the primary dwelling of the Ik people. *Lit.* 'Ik-hut'.
**Icémóríɗókàk<sup>a</sup>** (Icé-móríɗó-kàkà-) *n*. cowpea leaves (which the Ik eat as a vegetable). *Lit.* 'Ik-bean leaves'.
**Icéódòw<sup>a</sup>** (Icé-ódòù-) *n*. Ik day: a general Ik gathering held every January in Kamion to celebrate Ik identity and discuss challenges.
**ídadamɔ́s** (ídadamɔ́sí-) *v*. to be groping (e.g. walking in darkness or on a slope).
**ídèm** (ídèmè-) *pl.* **ídèmìk<sup>a</sup>** . *1 n*. serpent, snake. *2 n*. intestinal worm.
**ídemecɛmɛ́r** (ídeme-cɛmɛ́rí-) *n*. small plant species with yellow flowers; its roots, when roasted and ground, are used to treat snakebites. *Lit.* 'snakeherb'. Also called *ídèmèdàkw<sup>a</sup>* .
**ídèmèkwàyw<sup>a</sup>** (ídèmè-kwàyò-) *pl.* **ídemekwaitín**. *n*. snake fang.
**ídòkàts<sup>a</sup>** (ídò-kàtsì-) *pl.* **ídokatsín**. *n*. nipple. *Lit.* 'breast-tip'.
rub vigorously (e.g. when one is hit on the body, or to soften up a fruit).
**ídwà nì ɓàr** *n*. sour milk.
**ɨɗá** (<ɨɗɛɛs) *v*.
**ɨɗakɛ́s** (ɨɗakɛ́sí-) *v*. to lack, miss.
are banging'. See also *Ɗiwamúce* and *Lósʉ́ɓán*.
**ɨɗɛ́ɗɛ́**(<ɨɗɛ́ɗɔ́ɔ̀n) *v*.
*v*. to be covered, scoured, searched over thoroughly.
**iɗílón** (iɗílónì-) *v*. to get a headstart (e.g. on the day by leaving early).
**ɨɗɨmɛ́s** (ɨɗɨmɛ́sí-) *1 v*. to fix, make, repair. *2 v*. to organize, prepare, ready.
**ɨɗɨmɛ́sá así** *v*. to sit decently (by arranging your legs and clothes modestly).
**ɨɗɨmɛ́sá buƙúⁱ** *v*. to arrange a marriage.
**ɨɗɨmɛ́sá ìʉ̀mà<sup>ɛ</sup>** *v*. to arrange a marital engagement.
**ɨɗɨmɛ́síàm** (ɨɗɨmɛ́sí-àmà-) *pl.* **ɨɗɨmɛ́ síik<sup>a</sup>** . *1 n*. fixer, mechanic. *2 n*. artist, creator, inventor.
**ɨɗɨmɛ́sɔ́n** (ɨɗɨmɛ́sɔ́nì-) *v*. to be prepared, ready.
**ɨɗɨmɛ́sʉ́ƙɔt<sup>a</sup>** (ɨɗɨmɛ́sʉ́ƙɔtí-) *1 v*. to fix, make, repair. *2 v*. to organize, prepare, ready.
**ɨɗɨmɛtɛ́s** (ɨɗɨmɛtɛ́sí-) *1 v*. to create, make. *2 v*. to fix, repair. See also *iroketes* and *tɔsʉɓɛs*.
**ɨɗɨmɛtɛ́síàm** (ɨɗɨmɛtɛ́sí-àmà-) *pl.* **ɨɗɨmɛtɛ́síik<sup>a</sup>** . *n*. creator, maker.
**iɗimiés** (iɗimiesí-) *v*. to organize, plan, prepare, ready.
**iɗimiesá así** *v*. to prepare or ready oneself.
**iɗimiesíàm** (iɗimiesí-àmà-) *pl.* **iɗimiesíik<sup>a</sup>** . *1 n*. organizer. *2 n*. usher.
**iɗimiesúƙot<sup>a</sup>** (iɗimiesúƙotí-) *v*. to organize, plan, prepare, ready.
**ɨɗímɔ́n** (ɨɗímɔ́nì-) *v*. to speak a foreign language, talk foreignly.
**ɨɗíŋɔ́n** (ɨɗíŋɔ́nì-) *1 v*. to be narrow. *2 v*. to be tight (of a space, e.g. inside a vehicle).
**ɨɗíɔ́n** (ɨɗíɔ́nì-) *v*. to drop down (of seeds from a dried out plant pod).
**ɨɗíɔ́na iká<sup>e</sup>** *v*. migraine headache.
**iɗipes** (iɗipesí-) *v*. to dip into (e.g. a container, fire, etc.).
**ɨɗírírɔ̀n** (ɨɗírírɔ̀nì-) *v*. to move straight ahead.
**ɨɗírítɛ́sʉƙɔt<sup>a</sup>** (ɨɗírítɛ́sʉƙɔtí-) *v*. to straighten, unbend.
**ɨɗírɔ́n** (ɨɗírɔ́nì-) *1 v*. to be or move in a straight line. *2 v*. to aim, take aim.
**ɨɗɨtsɛs** (ɨɗɨtsɛsí-) *v*. to cane, lash, whip.
**ɨɗɔ́bɛ̀s** (ɨɗɔ́bɛ̀sì-) *v*. to arrange, order. See also *ɨnábɛ̀s* and *itíbès*.
**ɨɗɔ́bɛtɛ́s** (ɨɗɔ́bɛtɛ́sí-) *v*. to arrange, order.
**ɨɗɔ́ɗɔ́ɛ́s** (ɨɗɔ́ɗɔ́ɛ́sí-) *v*. to cook up, rustle up (a light meal or snack).
**iɗóɗókánón** (iɗóɗókánónì-) *v*. to be heaped, piled or stacked up.
**iɗóɗókés** (iɗóɗókésí-) *v*. to heap, pile, or stack on top of.
**iɗoes** (iɗoesí-) *v*. to drop into (e.g. food in the mouth).
**iɗokes** (iɗokesí-) *v*. to add on top.
**iɗókóliés** (iɗókóliesí-) *v*. to choose, pick, select (e.g. fruit that is ripe).
**ɨɛtɛ́sá así** *v*. to save oneself.
**ɨfɔɛs** (ɨfɔɛsí-) *v*. to drag, grate, scrape (along the ground). See also *fɔ́fɔ́tɛ́s*.
**ɨfɔɛsa así** *v*. to drag oneself.
**ɨkáká** (<ɨkákɛ́ɛ́s) *v*.
**ɨkɛɗɛs** (ɨkɛɗɛsí-) *v*. to curse with a difficult labor and delivery.
**ɨkɛɗíkɛ́ɗɛ́s** (ɨkɛɗíkɛ́ɗɛ́sí-) *v*. to stimulate digitally, titillate, touch lightly. See also *ɨkwatíkwátɛ́s*.
**ikeimétòn** (ikeimétònì-) *1 v*. to be lifted, raised up. *2 v*. to be developed.
**ikékéɲòn** (ikékéɲònì-) *1 v*. to be stable, steady, sturdy. *2 v*. to be dependable, reliable (in work).
**ikoŋíkóŋés** (ikoŋíkóŋésí-) *v*. to knock or rap on (e.g. a door). See also *ɨɗɔŋíɗɔ́ŋɛ́s*.
**ikóóbés** (ikóóbésí-) *1 v*. to close, fold up or in half (e.g. a book or leather mat). *2 v*. to collect, gather (e.g. yard rubbish).
**ikóteré** (ikóteré) *1 prep*. because of, due to. *2 subordconn*. because, due to the fact that, for the reason that. *3 subordconn*. in order that, so that. A noun following this word takes the oblique case. See also *kóteré*.
**ikwá** (<ikwóón) *v*.
**íkwà** (íkwà) *adv*. apparently, seemingly, it seems. See also *ókò*.
**ikwáánitetés** (ikwáánitetésí-) *1 v*. to equalize, equate, treat equally. *2 v*. to model, simulate. See also *iríánitetés*.
**ikwáánòn** (ikwáánònì-) *1 v*. to be the same. *2 v*. to resemble, be similar. *3 v*. to be symmetrical.
**ɨkwákwárɛ́s** (ɨkwákwárɛ́sí-) *v*. to spread around .
**ɨkwalɛs** (ɨkwalɛsí-) *1 v*. to clean, clear (a surface). *2 v*. to even, plane, shave (e.g. tool handles or wooden poles). See also *ɨkʉlɛs*.
**ikwárétòn** (ikwárétònì-) *v*. to resuscitate, revive.
**ɨkwaríkwárɛ́s** (ɨkwaríkwárɛ́sí-) *1 v*. to spread around. *2 v*. to spend wildly.
**ɨkwaríkwarɔ́s** (ɨkwaríkwárɔ́sí-) *v*. to be spread around.
**ɨkwatíkwátɛ́s** (ɨkwatíkwátɛ́sí-) *v*. to stimulate digitally, titillate, touch lightly. See also *ɨkɛɗíkɛ́ɗɛ́s*.
**ɨkwɛ́rɛ́ɗɔ̀n** (ɨkwɛ́rɛ́ɗɔ̀nì-) *v*. to wrestle out, wriggle free (of a hold).
**ɨkwɛtíkwɛ́tɛ́s** (ɨkwɛtíkwɛ́tɛ́sí-) *v*. to corral, round up (e.g. animals, by lightly hitting them).
**ɨkwɨlíkwílɛ́s** (ɨkwɨlíkwílɛ́sí-) *v*. to tickle.
**ɨkwílílɔ̀n** (ɨkwílílɔ̀nì-) *v*. to scream, shriek.
**ɨkwɨɲíkwíɲɔ́n** (ɨkwɨɲíkwíɲɔ́nì-) *v*. to be active, energetic. See also *kwɨɲídɔ̀n*.
**ikwóón** (ikwóónì-/ikwá-) *v*. to crow.
**ɨƙáálɛ́s** (ɨƙáálɛ́sí-) *v*. to skim off. See also *iripetés*.
**ɨƙaíƙá** (<ɨƙaíƙɛ́ɛ́s) *v*.
**ɨƙaíƙɛ́ɛ́s** (ɨƙaíƙɛ́ɛ́sí-/ɨƙaíƙá-) *v*. to fight, oppose, resist.
**ɨƙáƙá** (<ɨƙáƙɛ́ɛ́s) *v*.
**ɨƙáƙɛ́ɛ́s** (ɨƙáƙɛ́ɛ́sí-/ɨƙáƙá-) *v*. to fight, oppose, resist.
**ɨƙáƙɛ́ɛtɛ́s** (ɨƙáƙɛ́ɛtɛ́sí-) *v*. to fight, oppose, resist.
**ɨƙalíƙálɛ́s** (ɨƙalíƙálɛ́sí-) *v*. to confine, hold back/in, restrain (e.g. animals in a pen or someone in a fight).
**ɨƙaŋɛs** (ɨƙaŋɛsí-) *v*. to hold or prop up, support.
**ɨƙárárɔ̀n** (ɨƙárárɔ̀nì-) *v*. to sit on a stool.
**ɨƙɛɓíƙɛ́ɓɛ́sa ts'aɗí** *v*. to slash a firebreak.
**ɨƙɛ́ƙɛ́ɛ́s** (ɨƙɛ́ƙɛ́ɛ́sí-) *1 v*. to crack open (e.g. the seeds of gourds or pumpkins). *2 v*. to pick out, select.
**ɨƙɛ́lɛ́mɛ́s** (ɨƙɛ́lɛ́mɛ́sí-) *v*. to castrate (by crushing the spermatic cords).
**ɨƙɛlɛs** (ɨƙɛlɛsí-) *v*. to choose, pick out, select.
**ɨƙɛlɛtɛ́s** (ɨƙɛlɛtɛ́sí-) *v*. to choose, pick out, select.
**iƙémíƙémés** (iƙémíƙémésí-) *v*. to cut or slice away, fillet (i.e. meat or skin).
**iƙenes** (iƙenesí-) *1 v*. to beseech, entreat, plead with. *2 v*. to forgive, have mercy on, pardon.
**ɨƙɛníƙɛ́nɛ́s** (ɨƙɛníƙɛ́nɛ́sí-) *v*. to cluck at (e.g. animals or a honeyguide).
**ɨƙɛrɛs** (ɨƙɛrɛsí-) *v*. to draw, mark, trace (non-linguistically).
**ɨƙɨɛs** (ɨƙɨɛsí-) *v*. to block, deflect, shield. See also *ɨɓatɛs*.
**ɨƙɨlíƙílɔ̀n** (ɨƙɨlíƙílɔ̀nì-) *v*. to drip. See also *ɨlímɔ́n*.
**ɨƙɨrɛs** (ɨƙɨrɛsí-) *v*. to draw, mark, write (linguistically).
ɨƙɨrɛs
**ɨlɔ́lɔ́rɛ́s** (ɨlɔ́lɔ́rɛ́sí-) *v*. to smear, smudge (e.g. food on a child's face to trick its parents into believing it has eaten).
**ɨlɔ́ɔ́n** (ɨlɔ́ɔ́nì-/ɨlá-) *v*. to go for a visit, take a trip, travel. See also *ipásóòn*.
**ɨlɔ́ɔ́nʉƙɔt<sup>a</sup>** (ɨlɔ́ɔ́nʉƙɔtí-/ɨláíƙɔt-) *v*. to go for a visit or on a trip, travel away.
**ɨlɔpɛs** (ɨlɔpɛsí-) *v*. to move, reassign, transfer.
**ilotsesa zɛƙɔ́ ɛ** *v*. to migrate, move, relocate one's home.
**ilotsímétòn** (ilotsímétònì-) *v*. to change, transform. See also *iɓéléìmètòn*.
**ilúƙúretés** (ilúƙúretésí-) *v*. to coil, curl, spiral, wind.
**ilúƙúrètòn** (ilúƙúrètònì-) *v*. to coil or curl up (as in the fetal position).
**ilúƙúròn** (ilúƙúrònì-) *v*. to be coiled or curled up (as in the fetal position).
**ɨmáráɗàɗòòn** (ɨmáráɗàɗòònì-/ɨmáráɗaɗa-) *v*. to be fancy, flashy, gaudy.
**iméérés** (iméérésí-) *v*. to shift, transfer (from one like thing to another, i.e. container, place, etc.).
**ɨmɛɛs** (ɨmɛɛsí-/ɨmá-) *v*. to warm (by fire or sunlight).
**ɨmɛlɛs** (ɨmɛlɛsí-) *v*. to flicker, flitter (e.g. the tongue).
**iméníkánón** (iméníkánónì-) *v*. to be untrue to one's word, untrustworthy.
**Ìmɛ̀r** (Ìmɛ̀rà-) *n*. name of a hill or mountain.
**imetsés** (imetsésí-) *1 v*. to fill in for, replace, substitute for, take over for. *2 v*. to inherit.
**imetsités** (imetsitesí-) *v*. to fill or put in, replace, substitute.
**ɨmɨɗímíɗɛ́s** (ɨmɨɗímíɗɛ́sí-) *v*. to dilate, enlarge (a small hole).
**ɨmɨɗímíɗɛ́sa mɛná<sup>ɛ</sup>** *1 v*. to elaborate or expand on the issues. *2 v*. to exaggerate or hyperbolize.
**ɨmíɗítsɛ́s** (ɨmíɗítsɛ́sí-) *v*. to fill, plug, stop up (a hole).
**ɨmíɗítsɛ́sa así** *v*. to commit, devote, or plug oneself in.
**ɨmíʝílɛ́s** (ɨmíʝílɛ́sí-) *v*. to wink.
**ɨmɨlɛtɛ́s** (ɨmɨlɛtɛ́sí-) *v*. to dribble, drip, drop.
**ɨmɨlímílɔ̀n** (ɨmɨlímílɔ̀nì-) *v*. to pool, puddle (in small amounts).
**ɨmímíʝɛ́s** (ɨmímíʝɛ́sí-) *1 v*. to convulse, twitch (the face for fun). *2 v*. to shrug.
**ɨmɔ́ɗɛ́sʉƙɔt<sup>a</sup>** (ɨmɔ́ɗɛ́sʉƙɔtí-) *v*. to cheat, heist, rip off.
species with narrow round stems containing milky sap toxic to the eyes; it is used widely as a protective hedge around homes. *Euphorbia tirucalli*.
**ɨŋáŋá** (<ɨŋáŋɛ́ɛ́s) *v*.
**iŋáyá** (<iŋáyéés) *v*.
**iŋóɗyáìmètòn** (iŋóɗyáìmètònì-) *v*. to become chaotic, descend into chaos.
**iŋókíánón** (iŋókíánónì-) *v*. to be as poor or wretched as a dog.
**iŋolíŋólós** (iŋolíŋólósí-) *v*. to be wary: looking this way and that.
**ɨŋɔ́písà** (<ɨŋɔ́písɔ̀ɔ̀n) *v*.
**ɨŋʉ́ŋʉ́nɔ̀n** (ɨŋʉ́ŋʉ́nɔ̀nì-) *v*. to whine.
**ɨɔk<sup>a</sup>** (ɨɔkɔ́-) *n*. nectar, pollen.
**ɨɔ́kɔ́n** (ɨɔ́kɔ́nì-) *1 v*. to bloom, blossom, flower. *2 v*. to be fertile, flourish.
**iona ɛɗá** *v*. to be alone, solitary.
**iona muceék<sup>e</sup>** *v*. to be en route, on the way.
**iona ńdà** *v*. to be with, have.
**iona ńda sea ni itsúr** *v*. to be active, energetic. *Lit.* 'to have blood that boils'.
**ɨpáɗáɲɔ̀n** (ɨpáɗáɲɔ̀nì-) *v*. to be flat, level.
**ɨpáʝɔ́n** (ɨpáʝɔ́nì-) *v*. to sit on the ground.
**ɨpakɛs** (ɨpakɛsí-) *v*. to swipe: make a sweeping motion with one's hand.
**ɨpakɛsa cué** *v*. to fling water.
**ɨpákɛ́sʉƙɔt<sup>a</sup>** (ɨpákɛ́sʉƙɔtí-) *v*. to fling, swipe away/off.
**ɨpáláƙɔ̀n** (ɨpáláƙɔ̀nì-) *1 v*. to be feeble, frail, weak. *2 v*. to be fickle at work, reliable only when the boss is around. See also *ʝuódòn*.
**iona ŋítsaník<sup>ɛ</sup>** *v*. to be in trouble, have problems.
**ipiipíyeés** (ipiipíyeesí-) *v*. to smoothen.
**ɨpííríánón** (ɨpííríánónì-) *v*. to have a great memory, remember clearly.
**ipúɲá** (<ipúɲéés) *v*.
**ipúɲéés** (ipúɲéésí-/ipúɲá-) *v*. to make a funeral goat sacrifice to prevent the deceased's ghost from disturbing the relatives (by stunting their growth and the yield of their crops). See also *sɛ́ɛ́s*.
**ipuŋes** (ipuŋesí-) *1 v*. to tilt. *2 v*. to tuck (into one's clothing).
**ipúrá** (<ipúréés) *v*.
**ipúrá** (<ipúróòn) *v*.
**ipúréés** (ipúréésí-/ipúrá-) *1 v*. to fumigate, smoke out. *2 v*. to ritually cover in smoke (by sacrificing a chicken). See also *iwaŋíwáŋés* and *ts'udités*.
**ipúréètòn** (ipúréètònì-) *v*. to begin to smoke, billow up, evaporate.
**ipúróòn** (ipúróònì-/ipúrá-) *v*. to billow, fume, smoke, waft.
**ɨpʉtɛs** (ɨpʉtɛsí-) *v*. to beat or knock down/off (e.g. dew from grass).
**ɨpʉ́tɛ́sʉƙɔta así** *v*. to rush or take off.
**ɨpʉ́tɛ́sʉƙɔta muceé** *v*. to blaze a trail.
**ɨpʉtsɛs** (ɨpʉtsɛsí-) *v*. to plaster.
**ipwáákés** (ipwáákésí-) *v*. to accuse or act falsely.
**ɨrá** (<ɨrɛɛs) *v*.
**ɨrábɛs** (ɨrábɛsí-) *v*. to harvest finger millet.
**Iraf** (Irafá-) *n*. name of a river that flows down from Kamion.
**ɨraírá** (<ɨraírɔ́ɔ̀n) *v*.
**ɨraírɔ́ɔ̀n** (ɨraírɔ́ɔ̀nì-/ɨraírá-) *v*. to glare, shine brightly.
**ɨrákáánás** (ɨrákáánásì-) *n*. envy, jealousy.
**ɨrákáánón** (ɨrákáánónì-) *v*. to be envious, jealous.
**ɨrakɛs** (ɨrakɛsí-) *v*. to daze, stun (e.g. by slapping very hard).
**ɨrakɛsa así** *v*. to make oneself envious or jealous.
**ɨrákɛ́sʉƙɔt<sup>a</sup>** (ɨrákɛ́sʉƙɔtí-) *v*. to daze, stun.
**ɨrákɛ́sʉƙɔta así** *1 v*. to get oneself high (on drugs). *2 v*. to be in ecstasy, have an orgasm.
**ɨrakiesúƙota así** *v*. to have seizures, seize.
**ɨrákímétòn** (ɨrákímétònì-) *1 v*. to have a seizure, seize. *2 v*. to feel sexual afterglow. *Lit.* 'to be stunned'.
**iram** (iramá-) *pl.* **irámík<sup>a</sup>** . *n*. thinlysliced dehydrated food (e.g. meat and pumpkin). *Lit.* 'sliceable'.
**ɨramírámɛ́s** (ɨramírámɛ́sí-) *v*. to hit repeatedly.
**ɨraɲ** (ɨraɲí-) *n*. small corncobs and pieces of cobs left over after the larger cobs have been tied together by the husks.
**ɨraŋɛs** (ɨraŋɛsí-) *v*. to ruin, spoil.
**ɨraŋɛtɛ́s** (ɨraŋɛtɛ́sí-) *v*. to ruin, spoil.
**ɨraŋímétòn** (ɨraŋímétònì-) *1 v*. to become ruined or spoiled. *2 v*. to become upset.
**ɨraŋɔs** (ɨraŋɔsí-) *v*. to be ruined, spoiled.
**ɨráŋʉ́nánón** (ɨráŋʉ́nánónì-) *1 v*. to be ruined, spoiled. *2 v*. to be corrupt, depraved. *3 v*. to be dirty, soiled.
**ɨrapɛs** (ɨrapɛsí-) *v*. to reclaim, recover.
**ɨrápɛ́sʉƙɔt<sup>a</sup>** (ɨrápɛ́sʉƙɔtí-) *v*. to reclaim, recover.
**ɨrapɛtɛ́s** (ɨrapɛtɛ́sí-) *v*. to reclaim, recover.
**ɨráráƙɛ́s** (ɨráráƙɛ́sí-) *v*. to crack into pieces .
**ɨrárátés** (ɨrárátésí-) *v*. to gather, glean, harvest, reap (from the ground).
**ɨrarɛs** (ɨrarɛsí-) *v*. to gather, glean, harvest, reap (from the ground). See also *ɨrárátés* and *tarares*.
**ɨratírátɛ́s** (ɨratírátɛ́sí-) *v*. to spatter, splatter (e.g. heavy rain on the ground). See also *irwates* and *irwaírwéés*.
**ɨrɛɓɛs** (ɨrɛɓɛsí-) *v*. to clip, snip.
**iríƙá** (<iríƙéés) *v*.
**irímá** (<iríméés) *v*.
**irukes** (irukesí-) *v*. to chase or run after. See also *ɨlɔŋɛs*.
**irúkésuƙot<sup>a</sup>** (irúkésuƙotí-) *v*. to chase or run after.
**iruketés** (iruketésí-) *v*. to heap or pile up. See also *kitsetés*.
**irúkón** (irúkónì-) *v*. to sing.
**ɨrʉ́mʉ́nɔ́s** (ɨrʉ́mʉ́nɔ́sí-) *v*. to embrace or hug each another.
**irúpá** (<irúpóòn) *v*.
**Irwátà** (Irwátàà-) *n*. a personal name.
**isólólòòn** (isólólòònì-) *v*. to be visually clear (eyes, sky).
**isómá** (<isóméés) *v*.
which are overcrowded, thereby no producing heads). *2 v*. to be overfull to the bursting point of the container (e.g. a granary or a sack).
**ísw<sup>a</sup>** (ísó-) *n*. flood, torrent.
**isyá** (<isyees) *v*.
**itáléánón** (itáléánónì-) *v*. to be forbidden, off limits, prohibited, taboo.
**ɨtátá** (<ɨtátɛ́ɛ́s) *v*.
**isyónónuƙot<sup>a</sup>** (isyónónuƙotí-) *v*. to become merciful, soften (emotionally).
**ɨtáƙálɛ́s** (ɨtáƙálɛ́sí-) *1 v*. to mislay, misplace. *2 v*. to do/make poorly .
**ɨtíɗíɗɛ́sá así** *v*. to slink,sneak, tiptoe.
**ɨtíílɛ́s** (ɨtíílɛ́sí-) *v*. to repel, repulse, turn away.
**itíírà** (<itííròòn) *v*.
**ɨtílɛ́sʉƙɔt<sup>a</sup>** (ɨtílɛ́sʉƙɔtí-) *v*. to push over. **ɨtɨlɛtɛ́s** (ɨtɨlɛtɛ́sí-) *v*. to pull over.
**ɨtíŋá** (<ɨtíŋɛ́ɛ́s) *v*.
**ɨtɨŋɛsa así** *v*. to force oneself.
**itiŋésíàw<sup>a</sup>** (itiŋésí-àwà-) *pl.* **itiŋésíawík<sup>a</sup>** . *n*. cooking place, kitchen.
**itíón** (itíónì-) *v*. to go back, return there. See also *itéón*.
**ɨtíɔ́n** (ɨtíɔ́nì-) *v*. to delay. See also *asínón*.
**itípá** (<itípéés) *v*.
**itípéés** (itípéésí-/itípá-) *v*. to deviate, divert (e.g. enemies by making a preventative sacrifice or by talking them down from their plan).
**Itírá** (Itíráà-) *n*. a personal name.
**itírákés** (itírákésí-) *v*. to recoil from, shrink back from (i.e. something dangerous or embarrassing).
**itíríƙà** (<itíríƙòòn) *v*.
**itírónòn** (itírónònì-) *v*. to be fast, quick, speedy. See also *wɛ́ɛ́nɔ̀n*.
**itítíketés** (itítíketésí-) *v*. to hold back, restrain, retain.
**itítíŋòn** (itítíŋònì-) *v*. to be brave, courageous, dauntless, fearless, intrepid.
**itítírés** (itítírésí-) *v*. to hinder, obstruct, prevent.
**ɨtɨwɛs** (ɨtɨwɛsí-) *v*. to filter, strain. See also *ɨʝɨwɛs*.
**itíyá** (<itíyéés) *v*.
**itoɓes** (itoɓesí-) *1 v*. to hole, make a hole in. *2 v*. to cut in, interrupt.
**ɨtɔkɔɗɛs** (ɨtɔkɔɗɛsí-) *v*. to clench, grip (e.g. one's side while running).
**itóƙóƙòòn** (itóƙóƙòònì-) *v*. to be bony, cadaverous, emaciated, skeletal. See also *iróƙóòn* and *kwédekwedánón*.
**ɨtɔ́ƙɔ́ɔ̀n** (ɨtɔ́ƙɔ́ɔ̀nì-) *v*. to gimp, hobble. See also *ɨsɛ́pɔ́n* and *itsúkúkòn*.
**ɨtɔ́mɔ́nìàm** (ɨtɔ́mɔ́nì-àmà-) *pl.* **ɨtɔ́mɔ́ niik<sup>a</sup>** . *n*. neighbor.
**ɨtɔ́mʉ́nɔ́s** (ɨtɔ́mʉ́nɔ́sí-) *v*. to be neighbors, next to each other.
**ítón** (ítónì-) *v*. to be (some size or amount).
**ítónà dìdìk<sup>e</sup>** *v*. to become big or many. *Lit.* 'to reach high'.
**itoŋes** (itoŋesí-) *v*. to twine, twist.
**ɨtsɛ́ɛ́rà** (<ɨtsɛ́ɛ́rɔ̀ɔ̀n) *v*.
**ɨtsɔ́ítɛ́sʉƙɔt<sup>a</sup>** (ɨtsɔ́ítɛ́sʉƙɔtí-) *v*. to satisfy (someone's) hunger for meat.
**itsók<sup>a</sup>** (itsókó-) *n*. sunbird.
**ɨtsɔkítsɔ́kɛ́s** (ɨtsɔkítsɔ́kɛ́sí-) *v*. to whip back and forth.
**itsól** (itsólá-) *n*. bird species.
**itsópé** (<itsópóòn) *v*.
**ɨtsʉlɛs** (ɨtsʉlɛsí-) *v*. to pay a fine for (impregnating a girl out of wedlock).
#### itúrútésìàm
**iúrá** (<iúréés) *v*.
**iúréés** (iúréésí-/iúrá-) *v*. to pick clean, scavenge.
**iwá** (<iwees) *v*.
**ɨwáíƙɔt<sup>a</sup>** (<ɨwɔ́ɔ́nʉƙɔt<sup>a</sup> ) *v*.
**Iwam** (Iwamá-) *n*. name of a river.
**iwanetés** (iwanetésí-) *v*. to enlarge, expand, make bigger. See also *zeites*.
**iwánétòn** (iwánétònì-) *v*. to become bigger, enlarge, expand. See also *zoonuƙot<sup>a</sup>* .
**ɨwáwɛ́ɛ́s** (ɨwáwɛ́ɛ́sí-/ɨwáwá-) *v*. to caress, fondle, stroke.
**ɨwɛ́ɛ́lɛ́sʉƙɔt<sup>a</sup>** (ɨwɛ́ɛ́lɛ́sʉƙɔtí-) *v*. to disperse, dissipate, scatter, spread out.
**ɨwɛ́ɛ́lɛtɛ́s** (ɨwɛ́ɛ́lɛtɛ́sí-) *v*. to disperse, dissipate, scatter, spread out.
**ɨwɨɗɛs** (ɨwɨɗɛsí-) *v*. to grind finely.
**iwies** (iwiesí-) *v*. to spread around (e.g. food in order to cool it). See also *ɓátsɛ́s*.
**iwííɲés** (iwííɲésí-) *v*. to singe.
**iwóŋón** (iwóŋónì-) *v*. to exult, shout triumphantly (e.g. having won a contest or killed a wild animal).
**ɨwɔ́ŋɔ́n** (ɨwɔ́ŋɔ́nì-) *v*. to intend or mean (to do), premeditate. Must be followed by an infinitive in the dative case.
**ɨwɔ́ɔ́n** (ɨwɔ́ɔ́nì-) *v*. to be flagging, slacking, slowing.
**ɨwɔ́ɔ́nʉƙɔt<sup>a</sup>** (ɨwɔ́ɔ́nʉƙɔtí-/ɨwáíƙɔt-) *v*. to ease up, flag, slack off, slow down.
**iwórón** (iwórónì-) *v*. to roam, stray, wander.
**iwósétòn** (iwósétònì-) *v*. to be dangerously steep, preciptious.
**iwótsóòn** (iwótsóònì-) *v*. to binge, glut, overeat, pig out.
**iwówá** (<iwówéés) *v*.
**iwówéés** (iwówéésí-/iwówá-) *v*. to swarm over, teem around.
**ɨwʉ́lɔ́n** (ɨwʉ́lɔ́nì-) *v*. to boast, brag.
**iwulúwúlés** (iwulúwúlésí-) *v*. to rub around (e.g. on the ground to make dirty).
**ɨxáxá** (<ɨxaxɛɛs) *v*.
**ɨyáyá** (<ɨyáyɛ́ɛ́s) *v*.
**ɨyáyɛ́ɛ́s** (ɨyáyɛ́ɛ́sí-/ɨyáyá-) *v*. to shout or yell 'ya yaǃ' at (e.g. to drive animals or control a crowd).
**iyééseetésá así** *v*. to lower oneself.
**iyoes** (iyoesí-) *v*. to aim, direct, train.
# **j**
#### **jèjè** (jèjèì-) *pl.* **jéjèìk<sup>a</sup>** . *1 n*. leather. *2 n*. leather mat, sleeping skin.
**jíjè** (jíjèì-) *pl.* **jíjèìk<sup>a</sup>** . *n*. opposite side of a ravine, river, or valley.
## **ʝ**
**ʝàm** (ʝàmù) *ideo*. silkily, smoothly.
#### Kaaɓɔ́ŋ
## **k**
**kaiɗeíbɔrɔƙɔ́ƙ a** (kaiɗeí-bɔrɔƙɔ́ƙɔ́-) *pl.* **kaiɗeíbɔrɔƙɔ́ƙík<sup>a</sup>** . *n*. piece of a cut up pumpkin. *Lit.* 'pumpkin-cone'.
**Kàrɛ̀ŋà** (Kàrɛ̀ŋàà-) *n*. name of a river with reddish soil.
**Kàrɛ̀ŋààm** (Kàrɛ̀ŋà-àmà-) *pl.* **Karɛŋaik<sup>a</sup>** . *n*. Napore person.
money). *2 n*. bank check, check. *Lit.* 'money-tree paper'.
**kêd<sup>a</sup>** (kédì-) *n*. degree, measure.
**kétsóibaráts<sup>a</sup>** (kétsói-barátsó-) *n*. day after tomorrow, tomorrow next.
**kétsóita ke** *n*. three days from now.
**kɛ́xána kɛ** *dem*. that direction/way.
**kɛ́xɛ́s** (kɛ́xɛ́sì-) *1 v*. to fry (with or without oil). *2 v*. to dry by cooking (when no oil is used).
**kíʝá na ɨɗɨmɔtɔ́s** *n*. creation.
**kíʝáìm** (kíʝá-ìmà-) *pl.* **kíʝáwik<sup>a</sup>** . *n*. fairy, imp, sprite: small humanoid that inhabits the wild places, has magical powers, and interacts with humans in mischievous and mysterious ways. *Lit.* 'earth-child'.
**kíʝák<sup>e</sup>** *n*. down.
**Kocí** (Kocíì-) *n*. a personal name.
**koisiés** (koisiesí-) *v*. to wait in vain.
**kɔkɛ́sʉ́ƙɔt<sup>a</sup>** (kɔkɛ́sʉ́ƙɔtí-) *v*. to shut out.
**kɔlɨl** (kɔlɨlí-) *n*. cucumber.
**kom** (komá-) *n*. lots, many, multitude.
**kòm** (kòmà) *quant*. many.
**kɔŋɛ́síàm** (kɔŋɛ́sí-àmà-) *pl.* **kɔŋɛ́síik<sup>a</sup>** . *n*. chef, cook.
**kɔŋɛ́sídàkw<sup>a</sup>** (kɔŋɛ́sí-dàkù-) *pl.* **kɔŋɛ́sídakwitín**. *n*. cooking stick (for stirring food).
**kɔrɨtɛtɛ́s** (kɔrɨtɛtɛ́sí-) *1 v*. to blacken, char, scorch. *2 v*. to beat, crush, or destroy (slang for defeating someone).
**Kɔrɔmɔt<sup>a</sup>** (Kɔrɔmɔtá-) *n*. Toposa.
**Kɔrɔmɔtáhó** (Kɔrɔmɔtá-hóò-) *pl.* **kɔrɔmɔtáhóík<sup>a</sup>** . *n*. conical hut made in the style of the Toposa people.
**kòrrr** (kòrrr) *ideo*. swish swish (sound of someone moving by).
**Kotorúbé** (Kotorúbéè-) *n*. name of a hill or mountain.
**kúbam** (kúbamá-) *n*. more. *Lit.* 'unseeable'.
**kʉ́bɛ̀l** (kʉ́bɛ̀là-) *pl.* **kʉ́bɛ̀lìk<sup>a</sup>** . *n*. push hoe.
**kʉ́bɛ̀lɛ̀mɔ̀n** (kʉ́bɛ̀lɛ̀mɔ̀nì-) *v*. to be precipitous, steep (on two sides).
**kúbòn** (kúbònì-) *v*. to be invisible, out of sight, unobserved, unseen.
**kúbonuƙot<sup>a</sup>** (kúbonuƙotí-) *v*. to disappear, go out of sight, vanish.
**kúbùr** (kúbùrà-) *pl.* **kúburaikw<sup>a</sup>** . *n*. big container (e.g. gourd, jerrycan, tank).
**kùɓ<sup>a</sup>** (kùɓà-) *pl.* **kúɓítín**. *n*. hill.
**Kùɓààw<sup>a</sup>** (Kùɓà-àwà-) *n*. name of a hill. *Lit.* 'hill-place'.
**kuɓaɡwarí** (kuɓa-ɡwaríì-) *n*. hilltop.
**kúɗón** (kúɗónì-) *v*. to be low, short.
**kʉkʉ́ʉ́kᶶ** (kʉkʉ́ʉ́kʉ̀) *ideo*. cockle-doodledoo! (sound of a rooster crowing).
**kʉláɓ<sup>a</sup>** (kʉlaɓá-) *n*. bushbuck. *Tragelaphus scriptus*.
**Kumet<sup>a</sup>** (Kumetí-) *n*. name of a river.
are pounded and soaked in water yielding a decoction that is drunk for chest problems; the residue can also be applied to alleviate back and chest pain. *Vepris glomerata*.
roasted and eaten; a root decoction is drunk for body pain.
**kʉ́tʉ́k a** (kʉ́tʉ́kʉ̀-) *n*. first portion of edible termites to be eaten. See also *wàxìdòm*.
**kutúŋón** (kutúŋónì-) *v*. to be kneeling.
**Kʉwám** (Kʉwámʉ̀-) *n*. a personal name.
**kùx** (kùxù) *ideo*. greasy, oily.
**kwàà** (kwàà) *nurs*. pee-pee: a nursery word for urine or urinating.
**kwààk<sup>e</sup>** *n*. ago, before, since. May also be spelled as *kɔ̀wà kè*.
**kwaake nák<sup>a</sup>** *n*. since earlier today. May also be spelled as *kɔwa ke nák<sup>a</sup>* .
**kwààkè nòk<sup>o</sup>** *n*. long since, since long ago. May also be spelled as *kɔ̀wà kè nòk<sup>o</sup>* .
**kwààkè sìn** *n*. since yesterday. May also be spelled as *kɔ̀wà kè sìn*.
**kwaár** (kwaárá-) *pl.* **kwaárík<sup>a</sup>** . *n*. troop of baboons.
**kwaídòn** (kwaídònì-) *v*. to be chewy, tough to chew. See also *kaŋádòn*.
**kwàìn** (kwàìnì-) *1 n*. edges, sides. *2 n*. vulval (genital) labia.
**kwalíkwálɔ̀n** (kwalíkwálɔ̀nì-) *v*. to quake, quiver, shake, shiver, tremble. See also *irikíríkòn*.
**kwan** (kwaní-) *pl.* **kwanɨtín**. *1 n*. penis, phallus. *2 n*. stinger.
**kwaníékw<sup>a</sup>** (kwaní-ékù-) *n*. penis hole, urethral meatus. *Lit.* 'penis-eye'.
**kwaníts'ɛ́**(kwaní-ts'ɛ́à-) *n*. foreskin. *Lit.* 'penis-skin'.
**kwaɲɛ́s** (kwaɲɛ́sí-) *v*. to foil, thwart.
**kwaɲɛ́sʉ́ƙɔt<sup>a</sup>** (kwaɲɛ́sʉ́ƙɔtí-) *v*. to foil, thwart.
**kwar** (kwará-) *pl.* **kwàrìk<sup>a</sup>** . *n*. mountain.
**kwarádɛ̀**(kwará-dɛ̀à-) *pl.* **kwarɨkadɛík<sup>a</sup>** . *n*. base or foot of a mountain.
**kwaráɡwarí** (kwará-ɡwaríì-) *n*. mountaintop, peak, summit.
**kwaréékw<sup>a</sup>** (kwaré-ékù-) *n*. mountain saddle. *Lit.* 'mountain-eye'.
**kwàrìkààm** (kwàrìkà-àmà-) *pl.* **kwarɨkaik<sup>a</sup>** . *n*. mountain dweller.
**Kwarikabubúík<sup>a</sup>** (Kwarika-bubúíkà-) *n*. name of a place. *Lit.* 'mountainbellies'.
**kwats<sup>a</sup>** (kwatsí-) *1 n*. pee, urine. *2 n*. offspring, progeny.
**kwatsíém** (kwatsí-émè-) *n*. soft flesh below the buttock. *Lit.* 'urine-meat'.
**kwátsíkaakón** (kwátsíkaakónì-) *1 v*. to be little, small (of many). *2 v*. to be young (of many).
**kwatsítésuƙot<sup>a</sup>** (kwatsítésuƙotí-) *v*. to decrease size, make smaller, shrink down.
**kwatsitésúƙota así** *v*. to humble oneself.
**kwátsón** (kwátsónì-) *1 v*. to be little, small. *2 v*. to be young.
**kwátsónuƙot<sup>a</sup>** (kwátsónuƙotí-) *v*. to become little or small, decrease in size, shrink down.
**kweeda ƙwázà<sup>e</sup>** *n*. hem.
### kwirídòn
**kwits'ídòn** (kwits'ídònì-) *v*. to be juicy. **kwits'íkwíts'ánón** (kwits'íkwíts'ánónì-) *v*. to be moody, tempermental. **kwìts'ⁱ** (kwìts'ì) *ideo*. juicily.
## **ƙ**
**ƙà** (<ƙòòn) *v*.
*v*. to bemoan, complain, lament. See also *topóɗón*.
**ƙʉ́ɗʉ́nɔ́s** (ƙʉ́ɗʉ́nɔ́sí-) *v*. to suck on each other (e.g. during foreplay).
**ƙʉʝʉ́dɔ̀n** (ƙʉʝʉ́dɔ̀nì-) *v*. to garble.
**ƙʉ́ƙ a** (ƙʉ́ƙá-) *pl.* **ƙʉ́ƙítín**. *n*. runt.
**ƙʉƙʉmánónuƙot<sup>a</sup>** (ƙʉƙʉmánónuƙotí-) *v*. to turn back to back.
**ƙwàɗ<sup>e</sup>** (ƙwàɗè) *quant*. few.
**ƙwaɗiƙwáɗón** (ƙwaɗiƙwáɗónì-) *v*. to be fewer.
**ƙwàɗòn** (ƙwàɗònì-) *v*. to be few, little.
**ƙwaɗonuƙot<sup>a</sup>** (ƙwaɗonuƙotí-) *v*. to become fewer, decrease in number.
**ƙwár** (ƙwárá-) *pl.* **ƙwárítín**. *1 n*. cicatrix, scar. *2 n*. bruise, contusion.
#### ƙwɨxídɔ̀n
#### ƙwàz
**ƙwìx** (ƙwìxì) *ideo*. greenly.
**ƙwɨxídɔ̀n** (ƙwɨxídɔ̀nì-) *v*. to be verdant, very green. See also *xídɔ̀n*.
#### Laatso
## **l**
**Laatso** (Laatsoó-) *n*. name of a hill or mountain.
**laɓ<sup>a</sup>** (laɓá-) *pl.* **láɓíkw<sup>a</sup>** . *n*. cache, stash (whose location may be forgotten).
**laɓáɲámòn** (laɓáɲámònì-) *v*. to be gaping, wide-mouthed, yawning. See also *lafárámòn*.
**làf** (làfʉ̀-) *pl.* **láfítín**. *n*. breast (of meat), pec, pectoral muscle.
**lafárámòn** (lafárámònì-) *v*. to be gaping, wide-mouthed, yawning. See also *laɓáɲámòn*.
**láɡalaɡetés** (láɡalaɡetésí-) *v*. to check or spy on/out.
**làʝ<sup>a</sup>** (làʝà) *ideo*. loosely.
**laʝádòn** (laʝádònì-) *v*. to be loosely tied down, unsecured. See also *haʝádòn* and *yaŋádòn*.
**laʝámétòn** (laʝámétònì-) *1 v*. to collapse, crumple, fall down. *2 v*. to wilt, wither. *3 v*. to dissolve, melt (of fat). See also *ɲalámétòn*.
**laʝetés** (laʝetésí-) *1 v*. to lay down/over loosely (e.g. the last layer of grass on a thatched roof). *2 v*. to take off (e.g. beads from one's neck).
**lakámétòn** (lakámétònì-) *v*. to descend, go down (out of sight).
**lakámón** (lakámónì-) *v*. to descend, go down (out of sight).
**lakates** (lakatesí-) *v*. to push into/over the side.
**lakatiés** (lakatiesí-) *1 v*. to push into/ over the side repeatedly. *2 v*. to down, gulp down, inhale (food). See also *itúlákáɲés*.
**láládziránón** (láládziránónì-) *v*. to be ripped, shredded, in shreds.
**lalatíɓón** (lalatíɓónì-) *pl.* **lalatíɓónìk<sup>a</sup>** . *n*. flat stone, stone slab (used to grind tobacco, carry rubbish, cover granaries, or cover a rock well to protect it from the befouling of baboons).
**làlòn** (làlònì-) *v*. to be hideous, ugly. See also *itópénòn*.
**lalʉ́ʝɔ́n** (lalʉ́ʝɔ́nì-) *v*. to be roomy, spacious. See also *ɨlɔ́lɔ́mɔ̀n*.
**láŋ** (láŋá-) *pl.* **láŋítín**. *n*. bogus, counterfeit, fake, phoney, pseudo-.
**laŋádòn** (laŋádònì-) *v*. to be stifling, sultry, unpleasantly warm.
**laŋírímòn** (laŋírímònì-) *v*. to be broad, stout (e.g. bodies, buildings).
**laŋírón** (laŋírónì-) *v*. to be broad, stout (e.g. bodies, buildings).
**làr** (làrà-) *pl.* **láríkw<sup>a</sup>** . *n*. tobacco pipe.
**laradakw<sup>a</sup>** (lara-dakú-) *pl.* **lárákódakwitín**. *n*. pipe-stem (often made from Carissa stems).
**látsó** (látsóò-) *pl.* **látsóìk<sup>a</sup>** . *n*. drop-off, edge of a cliff or rock, precipice.
**látsóìk<sup>a</sup>** (látsóìkà-) *n*. falls, waterfall. *Lit.* 'cliff edges'.
**leat<sup>a</sup>** (leatí-) *pl.* **leatíkw<sup>a</sup>** . *1 n*. his/her/its brother. *2 n*. his/her cousin (father's brother's son).
**leatíím** (leatí-ímà-) *pl.* **leatíwík<sup>a</sup>** . *n*. his/her niece or nephew (brother's child).
**leatínánès** (leatínánèsì-) *n*. brotherhood, brotherliness.
**lɛ̀ɓ a** (lɛ̀ɓà-) *n*. liquid honey.
**lèɓ<sup>u</sup>** (lèɓù) *ideo*. pudgily, puffily.
**léɗ<sup>a</sup>** (léɗá-) *n*. gecko species?
**lɛɛmɛ́tɔ̀n** (lɛɛmɛ́tɔ̀nì-) *v*. to stick out/up (like a snake from the grass).
**leɡé** (leɡéè-) *1 n*. craziness, insanity, madness, mental illness. *2 n*. demon possession. See also *lejé*.
**lejé** (lejéè-) *1 n*. craziness, insanity, madness, mental illness. *2 n*. demon possession. Also pronounced as *leɡé*.
**lɛ́ʝ ɛ** (lɛ́ʝɛ́) *ideo*. flash!: sound of bursting in flames.
**lɛ́ʝɛ́tɔ́n** (lɛ́ʝɛ́tɔ́nì-) *v*. to catch fire, erupt in flames, ignite.
**lɛkɛ́s** (lɛkɛ́sí-) *v*. to retrieve from storage (e.g. grain from a granary).
**lemúánètòn** (lemúánètònì-) *1 v*. to be hornless. *2 v*. to be bare, naked, nude.
**lɛŋ** (lɛŋá-) *n*. honey badger, ratel. *Mellivora capensis*.
**lɛŋɛ́rɛ́mɔ̀n** (lɛŋɛ́rɛ́mɔ̀nì-) *v*. to be circumcised.
**leŋúrúmòn** (leŋúrúmònì-) *v*. to be bare, naked, nude.
**léó** (léóò-) *pl.* **léóín**. *1 n*. your brother. *2 n*. your cousin (father's brother's son).
**léóím** (léó-ímá-) *pl.* **léówík<sup>a</sup>** . *n*. your niece or nephew (brother's child).
**lɛ̀r** (lɛ̀rà-) *pl.* **lɛ́rítín**. *n*. fever tree, or Naivasha thorn: tall acacia with powdery green-white bark, whose wood is used to carve stools. *Acacia xanthophloea*.
**Locóto** (Locótoó-) *n*. name of a place.
**lóɗíkór** (lóɗíkóró-) *n*. scorpion.
**lóɗíkórócɛmɛ́r** (lóɗíkóró-cɛmɛ́rí-) *n*. scorpion herb: succulent plant species whose latex is applied to scorpion stings and whose reddish stems may be worn by girls as a kind of wig. *Euphorbia prostrata*.
**loɗúwa** (loɗúwaá-) *n*. jet, jet plane.
**lɔɡɛ́m** (lɔɡɛ́mʉ̀-) *n*. game warden, game ranger, wildlife authorities. From English 'game' park.
**loɡeréɲo** (loɡeréɲoó-) *n*. green stink bug. *Pentatomidae*.
**Lɔɡyɛ́l** (Lɔɡyɛ́lì-) *n*. a personal name.
**loiɓóròk<sup>a</sup>** (loiɓóròkù-) *n*. tall grass species found in the forest which is used to rain-proof granaries.
**Loíkí** (Loíkíì-) *n*. a personal name.
**lɔ̀ʝʉ̀rʉ̀tà** (lɔ̀ʝʉ̀rʉ̀tàà-) *n*. slipknot.
**loʝúulú** (loʝúulúù-) *n*. sorghum variety with drooping seed-heads, red seeds; it is very bitter and used to make leaven.
**Lɔkaaƙɨlɨt<sup>a</sup>** (Lɔkaaƙɨlɨtí-) *n*. name of a hill or mountain.
**Lòkààpèlòt<sup>a</sup>** (Lòkààpèlòtò-) *n*. name of the river draining *Nàkòrìtààw<sup>a</sup>* .
**lɔkaapín** (lɔkaapíní-) *pl.* **lɔkaapíník<sup>a</sup>** . *n*. shoelace, shoe-strap.
**lɔkabʉ́ás** (lɔkabʉ́ásì-) *n*. crumbly rock.
#### lɔkáʝʉ́
rule in northeast Uganda and northwest Kenya.
**Lokwaŋ** (Lokwaŋá-) *n*. February: month of planting. See also *Ɓèts'òn*.
**Lokwaŋ** (Lokwaŋá-) *n*. a personal name.
**loƙeƙes** (loƙeƙesí-) *pl.* **loƙéƙésìk<sup>a</sup>** . *n*. glutton, gobbler, gourmand. See also *lòkòɗòŋìròàm*.
**loƙírot<sup>a</sup>** (loƙírotí-) *n*. robin-chat (whitebrowed and others?). *Cossypha sp*.
**loƙól** (loƙólé-) *n*. eagle.
**lóƙólíl** (lóƙólílá-) *pl.* **lóƙólílík<sup>a</sup>** . *n*. swing.
**lɔ́ƙɔ́ŋ** (lɔ́ƙɔ́ŋʉ̀-) *pl.* **lɔ́ƙɔ́ŋìk<sup>a</sup>** . *n*. sacred tree where ceremonies like *itówéés* are held. Also called *ɲɔ́ƙɔ́ŋ*.
**lɔ́ƙɔ́ŋʉ̀dɛ̀** (lɔ́ƙɔ́ŋʉ̀-dɛ̀à-) *pl.* **lɔ́ƙɔ́ŋɨkadɛík<sup>a</sup>** . *n*. base of the sacred tree.
**loƙózòmòn** (loƙózòmònì-) *v*. to be longnecked (of gourd or women who wear neck-beads).
**lɔlɛ́ɛʉ́** (lɔlɛ́ɛʉ́ʉ̀-) *n*. cattle disease that causes meat to become bitter.
**lóliit<sup>a</sup>** (lóliití-) *n*. false accuser or witness. See also *kɛ́rínɔ́síàm*.
**lolíts<sup>a</sup>** (lolítsí-) *n*. dense forest, jungle.
**Lolítsíàƙw<sup>a</sup>** (Lolítsí-àƙɔ̀-) *n*. name of a densely forested place. *Lit.* 'in the jungle'.
**lɔlɔanón** (lɔlɔanónì-) *v*. to be discontent, dissatisfied.
**Lɔlɔɓáy<sup>a</sup>** (Lɔlɔɓáí-) *n*. October. See also *Terés*.
**lolómónuƙot<sup>a</sup>** (lolómónuƙotí-) *1 v*. to shrivel up (e.g. seeds in the ground). *2 v*. to decay, dry out (e.g. bones, hair).
**Lóloy<sup>a</sup>** (Lóloí-) *n*. name of a river.
**Lɔmaaníkɔ** (Lɔmaaníkɔɔ́-) *n*. name of a river.
**Lomaruk<sup>a</sup>** (Lomarukú-) *n*. April: month of mushrooms. See also *Lɔmɔ́y a* .
**Lomataŋaáw<sup>a</sup>** (Lomataŋá-áwà-) *n*. name of a place where some Ik used to live.
**Lɔ́mɛ́ʝ a** (lɔ́mɛ́ʝà-) *n*. name of a mountain sloping westward off *Morúŋole*.
**lɔmɛ́ʝɛ́kɛlɛ́** (lɔmɛ́ʝɛ́kɛlɛ́ɛ̀-) *n*. cockroach, roach.
**lóméléw<sup>a</sup>** (lóméléwá-) *pl.* **lóméléwáikw<sup>a</sup>** . *n*. widow(er). See also *ɲepúrósit<sup>a</sup>* .
**loménio** (loménió-) *n*. swallow, swift.
**Lɔmɛ́r** (Lɔmɛ́rà-) *n*. a personal name.
**Loméríɗok<sup>a</sup>** (Loméríɗokó-) *n*. name of a hill or mountain.
**lomerúk<sup>a</sup>** (lomerúká-) *n*. sweetsmelling plant species growing underground, invisible until guinea-fowl uncover it; a decoction from its red roots is drunk as a medicine.
**Lɔmíʝ<sup>a</sup>** (Lɔmíʝí-) *n*. name of a mountain and surrounding area that was home to a fairy who told on wrongdoers.
**Lómìl** (Lómìlà-) *n*. name of a river.
**loŋórómòn** (loŋórómònì-) *v*. to be domical, hemispherical (like a round hut).
**lɔŋɔ́t a** (lɔŋɔ́tá-) *n*. enemies, foes.
**lɔŋɔ́tánànès** (lɔŋɔ́tánànèsì-) *n*. enmity, hostility.
**loŋɔ́tásìts'<sup>a</sup>** (loŋɔ́tá-sìts'à-) *n*. sacrifice for warding off enemies.
**Loocíkwa** (Loocíkwaá-) *n*. name of a hill.
**Lɔɔɗíŋ** (Lɔɔɗíŋì-) *n*. name of a mountain in Didingaland, South Sudan. Also called *Lotukéì*.
**Lòòɗòs** (Lòòɗòsì-) *n*. name of a hill or mountain.
**Looɗóy<sup>a</sup>** (Looɗóì-) *n*. name of a hill in Timu, the surrounding area, and associated human habitations.
**lɔɔmʉ́yá** (lɔɔmʉ́yáà-) *n*. blue-eared starling (greater or lesser). *Lamprotornis*.
**lɔɔrán** (lɔɔrání-) *n*. glow of a fire at night.
**lóórì** (lóórìì-) *pl.* **lóórììk<sup>a</sup>** . *n*. lorry, truck. See also *ɲolórì*.
**lɔɔrʉ́k a** (lɔɔrʉ́kʉ́-) *n*. whydah (Eastern paradise or pin-tailed). *Vidua sp*.
**Lɔɔsɔ́m** (Lɔɔsɔ́mɔ̀-) *n*. name of a river.
**lɔ́pɛ́ɗɛpɛ́ɗ a** (lɔ́pɛ́ɗɛpɛ́ɗɛ́-) *n*. bat.
**Lopéɗó** (Lopéɗóò-) *n*. name of a mountainous area south of Ikland.
**lopem** (lopemú-) *pl.* **lopémík<sup>a</sup>** . *1 n*. flat area. *2 n*. plateau, tableland. *3 n*. level, storey.
**lopemúím** (lopemú-ímà-) *pl.* **lopémíkawik<sup>a</sup>** . *n*. stair, step.
**lopéren** (lopérení-) *n*. ghost, ghoul, phantom, wraith (associated with rivers).
**Lɔpɛ́t a** (Lɔpɛ́tí-) *n*. name of a hill or mountain.
**lópey<sup>a</sup>** (lópeí-) *pl.* **lópèìk<sup>a</sup>** . *n*. pancreas.
**Lopéyók<sup>a</sup>** (Lopéyóko-) *n*. a personal name.
**Lopíar** (Lopíarí-) *n*. Year of *Lopíar* (1980), a year that brought disease and famine on the Ik, leading to the death and displacement of many.
**Lopíè** (Lopíè-) *n*. a personal name.
**lɔ́pírɨpír** (lɔ́pírɨpírá-) *n*. wood-boring insect (identified by the piles of sawdust it leaves below).
**lɔpɨtá** (lɔpɨtáí-) *pl.* **lɔpɨtáík<sup>a</sup>** . *1 n*. drying rack. *2 n*. platform, podium. *3 n*. altar.
**lɔpɨtáá na ƙófóikó<sup>e</sup>** *n*. dish-drying rack.
**Lopokók<sup>a</sup>** (Lopokókò-) *n*. name of a mountain to the east of Timu. Also called *Soƙoɡwáás*.
**Loporukɔlɔ́ŋ** (Loporukɔlɔ́ŋì-) *n*. name of a place.
**lɔpɔ́ts<sup>a</sup>** (lɔpɔ́tsá-) *n*. clear fluid found in an elephant's stomach.
**lɔ́pʉ́l** (lɔ́pʉ́lì-) *pl.* **lɔ́pʉ́lìk<sup>a</sup>** . *n*. small oblong edible gourd. *Lagenaria sp*.
**Lopúsór** (Lopúsórì-) *n*. a personal name.
**Lotyak<sup>a</sup>** (Lotyakí-) *n*. September. See also *Nakariɓ<sup>a</sup>* .
**Lotyaŋ** (Lotyaŋí-) *n*. a personal name.
**loúk<sup>a</sup>** (loukú-) *1 n*. carnivore, predator. *2 n*. greedyguts.
**Loukómor** (Loukómorú-) *n*. name of a mountain in Turkanaland, Kenya.
soaked in water, and drunk for stomachaches. *Pachycarpus schweinfurthii*.
**Lourien** (Louriení-) *n*. a personal name.
#### mà
## **m**
**màŋ** (màŋà) *ideo*. thickly.
**maráŋónìàm** (maráŋónì-àmà-) *pl.* **maráŋóniik<sup>a</sup>** . *n*. good, kind person.
**mídzatetés** (mídzatetésí-) *v*. to catch scent of, smell.
**míɡiriɡíránón** (míɡiriɡíránónì-) *v*. to be dusky, twilit (at dawn or dusk).
**míʝés** (míʝésì-) *v*. to decline, reject, scorn, turn down.
**mísì** (mísì) *subordconn*. if, whether.
**mísɨ … mísɨ …** *coordconn*. either … or ….
**moona ɡúró<sup>e</sup>** *v*. to be heartsick (from anger, guilt, sadness, etc.).
**moós** (moosí-) *v*. to be given.
buried as a way to prevent enemies or sickness).
**muts'utiesúƙot<sup>a</sup>** (muts'utiesúƙotí-) *v*. to keep closing or shutting up (e.g. termite holes where one doesn't want them exiting).
**mʉtʉ** (mʉtʉʉ́-) *pl.* **mʉtʉ́ík<sup>a</sup>** . *1 n*. thick needle/pin with a wooden handle (used for mending gourds and leather skins). *2 n*. firing pin.
**Mʉtʉ́nan** (Mʉtʉ́naní-) *n*. name of a river.
## **n**
**naíké** *dem*. here.
**Nakɔŋ** (Nakɔŋʉ́-) *n*. a personal name.
**Narúkyeɲ** (Narúkyeɲí-) *n*. name of a hill or mountain and surrounding area.
**nayé** *dem*. here.
**nayé kɔ̀nà** *dem*. right here.
**nayé na** *dem*. here.
**nayé ne** *dem*. just there, there.
**názɛ̀ƙwà** (názɛ̀ƙwà) *1 n*. while. *2 n*. now, soon. When used in the sense of 'while', this word is followed by a verb with the dummy pronoun attached to it.
**nàⁱ** (nàì) *ideo*. viscously.
**ńdà** (ńdà) *1coordconn*. and. *2 prep*. with. In a series of nouns linked by this word, every noun after the first takes the oblique case.
**ńda nébèè kɔ̀n** *n*. eleven.
**ndaicé** (ndaicé-) *n*. um, uh, whatca-macallit.
**ndaík<sup>e</sup>** *pro*. where?
**nday<sup>o</sup>** *n*. by what path? which way?
**ndayúk<sup>o</sup>** *n*. it is where?
**ndéé** *1 n*. from where? *2 interj*. whatever! (an expression of disagreement).
**ǹdò** *pl.* **ndoín**. *pro*. who?
**ndóó mìtìɛ̀** *v*. what if (it is …).
**ne** (=ne) *dem*. that (just there).
**ne** (ne) *interj*. here! here you go! (an expression of giving).
**nêb<sup>a</sup>** (nébù-) *pl.* **nébitín**. *1 n*. body. *2 n*. self.
**nébàdà** (nébàdì-) *n*. colossus, giant, monster of a, whale of a. Referring to an entity a slight distance away.
**nébèd<sup>a</sup>** (nébèdè-) *pl.* **nébìn**. *n*. himself, herself, itself: the very person/thing.
**nébùnànès** (nébùnànèsì-) *n*. bodiliness, embodiment.
**nébùsìts'<sup>a</sup>** (nébù-sìts'à-) *n*. body hair.
**nédà** (nédì-) *dem*. there (near).
**néda ne** *dem*. just there, there.
**nɛ́ɛ́** (nɛ́ɛ́) *subordconn*. when. The main verb that follows this word takes the simultaneous aspect.
**nɛ́ɛ́**(nɛ́ɛ́) *1 prep*. from. *2 prep*. through. A noun following this word takes the genitive case.
**nɛɛ́s** (nɛɛsí-/na-) *v*. to abide, bear, deal with, endure, tolerate.
**nɛɛsʉ́ƙɔt<sup>a</sup>** (nɛɛsʉ́ƙɔtí-/naɨƙɔt-) *v*. to abide, bear, deal with, endure, tolerate.
**néíta ne** *dem*. just there, there.
**nɛpɛ́ƙáàm** (nɛpɛ́ƙá-àmà-) *pl.* **nɛpɛ́ƙáik<sup>a</sup>** . *1 n*. arguer, argumentative person. *2 n*. atheist, unbeliever.
**nɛpɛƙanitetés** (nɛpɛƙanitetésí-) *v*. to challenge, contradict.
**nɛpɛƙánón** (nɛpɛƙánónì-) *v*. to argue, debate, disagree, protest.
**nɛ̀rɛ̀** (nɛ̀rɛ̀) *ideo*. teeteringly.
**nɛrɛ́dɔ̀n** (nɛrɛ́dɔ̀nì-) *v*. to be teetering, tottering.
**nɛ́rɨnɛ́rɔ́n** (nɛ́rɨnɛ́rɔ́nì-) *1 v*. to quiver, shudder. *2 v*. to lurch, stagger.
**nés** (nésé) *adv*. oh, I see; oh, you mean.
**nɛsɛƙánón** (nɛsɛƙánónì-) *v*. to be healthy, hygienic.
**nesés** (nesésí-) *v*. to hear. See also *nesíbès*.
**nesíbes** (nesíbesí-) *1 v*. to hear. *2 v*. to listen. *3 v*. to comprehend, understand. *4 v*. to heed, obey. See also *nesés*.
**nesíbesíám** (nesíbesí-ámà-) *pl.* **nesíbesíík<sup>a</sup>** . *n*. listener.
**nesíbiés** (nesíbiesí-) *v*. to obey habitually.
**nesíbos** (nesíbosí-) *v*. to be understood.
**nesíbunós** (nesíbunósí-) *1 v*. to understand each other. *2 v*. to be understood.
**nótsóò nòk<sup>o</sup>** *n*. day before yesterday.
**ńtá** (ńtá) *pro*. where?
**ntsúó ts'ɔɔ** *pro*. it's likely, probably.
(#7 in the historical line). *Lit.* 'Leopard-Folk'.
**nʉʉnʉ́**(nʉʉnʉ́) *nurs*. yum-yum! (a nursery word for breastfeeding!.
#### ɲaɓʉraídàkwa
#### ɲáaɲún
# **ɲ**
**ɲanʉ́pít<sup>a</sup>** (ɲanʉ́pítì-) *n*. belief, faith.
#### ɲasal
**ɲáwaawá** (ɲáwaawáà-) *pl.* **ɲáwaawáìk<sup>a</sup>** . *n*. large gunny sack. See also *lomóŋin*.
wife's sibling, my brother's wife's sibling, my sister's husband). *2 n*. my sister's husband's sibling).
**ɲécáy<sup>a</sup>** (ɲécáì-) *n*. tea.
**ɲelerum** (ɲelerumú-) *n*. argument, disputation, quarreling.
**ɲélúru** (ɲélúruú-) *n*. quail.
**ɲɛ́mɛlɛkʉ́dàkw<sup>a</sup>** (ɲɛ́mɛlɛkʉ́-dàkù-) *pl.* **ɲɛ́mɛlɛkʉ́ɨkadakwitín**. *n*. hoe handle.
**ɲɛmɛray<sup>a</sup>** (ɲɛmɛraí-) *n*. sorghum variety with tall, red seed-heads and sweet canes.
**ɲépiskóópì** (ɲépiskóópìì-) *pl.* **ɲépiskóópììk<sup>a</sup>** . *n*. bishop.
**ɲɛpɨtɛ** (ɲɛpɨtɛɛ́-) *pl.* **ɲɛpɨteicík<sup>a</sup>** . *1 n*. behavior, habit, manner. *2 n*. method, procedure, way.
berries for dying their hair. *Lit.* 'girls*dzôɡ*'.
**ɲétsúpa** (ɲétsúpaá-) *pl.* **ɲétsúpàìk<sup>a</sup>** . *n*. glass bottle.
**ɲétsúur** (ɲétsúurí-) *pl.* **ɲétsúùrìk<sup>a</sup>** . *n*. hole in a riverbed hollowed out by churning water.
are ground, cooked, and drunk for stomachache. *Indigofera arrecta*.
**ɲezeí** (ɲezeíì-) *pl.* **ɲézeíìk<sup>a</sup>** . *n*. inkpad, stamp pad.
**ɲícwéɲé** (ɲícwéɲéè-) *n*. sugar bush: hardwoood tree species whose wood is used for building but which nails cannot pierce. *Protea gaguedi*.
**ɲimanitésíàw<sup>a</sup>** (ɲimanitésí-àwà-) *pl.* **ɲimanitésíawík<sup>a</sup>** . *n*. edge, joint.
**ɲimánón** (ɲimánónì-) *v*. to encounter, meet.
**ɲɔ́ɓɔɔ́**(ɲɔ́ɓɔɔ́ɔ̀-) *n*. lentils.
**ɲorótónitíɔ́k a** (ɲorótónití-ɔ́kà-) *pl.* **ɲorótónitikɔɔkɨtín**. *n*. humerus: bone of the upper arm.
**ɲosoƙoloké** (ɲosoƙolokéè-) *pl.* **ɲosoƙolokéìk<sup>a</sup>** . *n*. shorts, pair of trunks.
**ɲɔt<sup>a</sup>** (ɲɔtɔ́-) *1 n*. men. *2 n*. husbands.
**ɲotánánès** (ɲotánánèsì-) *1 n*. friendliness, friendship (with non-Ik). *2 n*. to be related indirectly by marriage (e.g. to the parents of a child's or sibling's spouse).
**ɲótóts<sup>a</sup>** (ɲótótsì-) *pl.* **ɲótótsìk<sup>a</sup>** . *1 n*. flashlight, torch. *2 n*. lip herpes.
**ɲɔ́tsɔ́ɓɛ** (ɲɔ́tsɔ́ɓɛɛ́-) *pl.* **ɲɔ́tsɔ́ɓɛ̀ìk<sup>a</sup>** . *n*. cap made of giraffe-tail hairs.
**ɲɔ́tsɔ́ɓɨtsɔɓ<sup>a</sup>** (ɲɔ́tsɔ́ɓɨtsɔɓí-) *n*. hodgepodge, melange, mishmash.
**ɲótsorón** (ɲótsorónì- ) *pl.* **ɲótsorónìk<sup>a</sup>** . *n*. latrine, outhouse, toilet. See also *ets'íhò*.
**ɲóvakáɗò** (ɲóvakáɗòò-) *n*. avocado (tree and fruit). *Persea americana*.
**ɲówoɗí** (ɲówoɗíì-) *n*. seed butter, tahini (miture of peanut and sesame pastes).
**ɲʉmɛ́s** (ɲʉmɛ́sí-) *v*. to want, wish for.
#### ŋabér
# **ŋ**
**ŋábitetés** (ŋábitetésí-) *v*. to dress up, get dressed.
**ŋáɓɔ́ɔla** (ŋáɓɔ́ɔlaá-) *pl.* **ŋáɓɔ́ɔ̀làìk<sup>a</sup>** . *n*. cent, penny.
**ŋáɓutús** (ŋáɓutúsù-) *pl.* **ŋáɓutúsìk<sup>a</sup>** . *n*. boot.
**Ŋákiswahílìtòd<sup>a</sup>** (Ŋákiswahílì-tòdà-) *n*. Swahili language.
**ŋám** (ŋámá-) *n*. sorghum.
**ŋámá na buɗám** *n*. sorghum variety with black seeds.
**ŋámá nà ɓèts'<sup>a</sup>** *n*. sorghum variety with white seeds.
**ŋámá nà ɗìw<sup>a</sup>** *n*. sorghum variety with red seeds.
**ŋamarʉwáy<sup>a</sup>** (ŋamarʉwáì-) *n*. millet beer. See also *rébèmɛ̀s*.
**ŋamíá** (ŋamíáì-) *n*. hundred (100). No plural form.
**ŋamɨŋámɔ́n** (ŋamɨŋámɔ́nì-) *v*. to rush into things (eating, talking, etc.).
the zebra as its totem (#4 in the historical line). The Ik name is *Zɨnáík<sup>a</sup>* .
**ŋókítsùts<sup>a</sup>** (ŋókí-tsùtsà-) *n*. dogfly: species of fly associated with dogs.
**ŋoléánètòn** (ŋoléánètònì-) *1 v*. to be white-faced. *2 v*. to be bald. *3 v*. to be treeless.
**ŋɔr** (ŋɔrɛ́-) *pl.* **ŋɔrɨtín**. *1 n*. colored clay, ocher. *2 n*. color.
**ŋɔra na buɗám** *n*. black clay.
**ŋɔra na ɗíw<sup>a</sup>** *n*. red clay.
**ŋɔrɛ́ám** (ŋɔrɛ́-ámà-) *pl.* **ŋɔréík<sup>a</sup>** . *n*. Turkana person (who wears a colored clay headdress).
**ŋɔrɔ́ɲɔ́mɔ̀n** (ŋɔrɔ́ɲɔ́mɔ̀nì-) *v*. to be dirty, soiled, unclean (from dirt or food). See also *ɲɔŋɔ́rɔ́mɔ̀n*.
**ŋɔ́rɔ́rɔ̀n** (ŋɔ́rɔ́rɔ̀nì-) *1 v*. to saw logs, snore. *2 v*. to growl. *3 v*. to purr.
**ŋorótsánón** (ŋorótsánónì-) *v*. to be filthy, nasty, putrid (e.g. of water or wounds).
**ŋɔt<sup>a</sup>** (ŋɔtá-) *1 n*. cowdung, manure. See also *ɦyɔ̀èts'<sup>a</sup>* . *2 n*. millet heads when they turn brown and curl.
**ŋʉɗʉ́sʉ́mɔ̀n** (ŋʉɗʉ́sʉ́mɔ̀nì-) *1 v*. to be low, short. *2 v*. to be limbless.
ground is lain). *Lit.* 'mother of the stone'.
**ŋ́k a** (ɲ́cì-) *pro*. I/me/my.
**Ŋ́kaleesó** (Ŋ́kaleesóò-) *n*. Ateker name for the traditional Ik men's age-group with the ostrich as its totem (#9 in the historical line). The Ik name is *Leweɲiik<sup>a</sup>* .
**ŋ́kaŋók<sup>a</sup>** (ŋ́kaŋókì-) *n*. acne, pimples.
**ŋ́karakocóy<sup>a</sup>** (ŋ́karakocóì-) *n*. bottlecap game.
## **o/ɔ**
**ódzadidí** (ódza-didíì-) *n*. light rain at the beginning of dry season.
spears, and walking sticks and are used to build houses and granaries. *Grewia tenax*. See also *alárá*.
**ɔríyɔ́**(ɔríyɔ́ɔ̀-) *n*. bird species.
**Ɔrɔ́m** (Ɔrɔ́mʉ̀-) *n*. name of a mountain on the border of Karamoja and Acholiland.
**otésúƙot<sup>a</sup>** (otésúƙotí-) *v*. to pour into.
**otetés** (otetésí-) *1 v*. to pour out into. *2 v*. to miscarry.
**otí** (otí) *interj*. whoa! (an expression of awe or mystery).
**ɔ́zààk<sup>a</sup>** (ɔ́zà-àkà-) *n*. anus.
**ɔ́zàhò** (ɔ́zà-hòò-) *pl.* **ɔ́zɨtíníhoík<sup>a</sup>** . *n*. anal sphincter.
**ɔ́zàsìts'<sup>a</sup>** (ɔ́zà-sìts'à-) *n*. pubic hair. See also *didisísíts'<sup>a</sup>* .
#### pààɗòk o
# **p**
**pààɗòk<sup>o</sup>** (pààɗòkò) *ideo*. bang! kaboom! (sound of a gunshot).
**Páɗɛ̀rɛ̀hò** (Páɗɛrɛ-hoó-) *n*. name of a hill where the Italian priest *Apáálolúk* used to stop for the night. *Lit.* 'priest-hut'.
**paɗókómòn** (paɗókómònì-) *v*. to be caved in, collapsed (e.g. one's stomach).
**pàɗw<sup>a</sup>** (pàɗò-) *pl.* **páɗíkw<sup>a</sup>** . *n*. small cave (often used for secret storage).
**pákà** (pákà) *1 prep*. all the way to, until, up to (followed by a noun). *2 subordconn*. until (followed by a dependent clause). *3 adv*. forever, indefinitely. As a preposition, this word is followed by a noun in the oblique case.
**pakámón** (pakámónì-) *v*. to split in two.
**Palúùkùɓ<sup>a</sup>** (Palúù-kùɓà-) *n*. name of a hill in Timu. *Lit.* 'Palu-hill'.
**pànɛ̀ɛ̀s** (pànɛ̀ɛ̀sì-) *n*. teenage boys, young men. See also *ŋísɔ́rɔk<sup>a</sup>* .
**pápà** (pápàà-) *n*. pope: head of the Roman Catholic church.
**papaɗós** (papaɗósí-) *n*. small stash hidden from others (e.g. food).
**parɨpárɔ́n** (parɨpárɔ́nì-) *v*. to gleam, glisten. See also *piripírón*.
**pás** (pásì-) *n*. lousy, pathetic, or useless person or thing. See also *tsar*.
**pásìnànès** (pásìnànèsì-) *n*. patheticness, lousiness, uselessness. See also *tsarínánès*.
**pásìtà** (pásìtàà-) *pl.* **pásìtàìk<sup>a</sup>** . *n*. minister, pastor, rector.
**patsólómòn** (patsólómònì-) *v*. to be bare (of a patch or spot).
**páupáw<sup>a</sup>** (páupáù-) *n*. scout bee that scouts out sources of food and water.
**pɛɗɛpɛ́ɗɔ́n** (pɛɗɛpɛ́ɗɔ́nì-) *v*. to flutter (of hearts, wings).
**pɛɛ́ɲɛ́mɔ̀n** (pɛɛ́ɲɛ́mɔ̀nì-) *v*. to walk in a small-buttocked way.
**pɛ̀l** (pɛ̀lɛ̀) *ideo*. slickly, slipperily.
**pɛlɛ́dɔ̀n** (pɛlɛ́dɔ̀nì-) *v*. to be precariously slick or slippery.
**pɛ́lɛ́ɗɛ̀k a** (pɛ́lɛ́ɗɛ̀kɛ̀-) *n*. long-leaf tobacco. See also *loríónómor*.
**pɛlɛ́mɛ́tɔ̀n** (pɛlɛ́mɛ́tɔ̀nì-) *v*. to appear, come into view, emerge. See also *lɛlɛ́tɔ́n*.
**pɛlɛ́mɔ́na fetí** *v*. dawn, daybreak, sunrise. *Lit.* 'appearance of the sun'.
**Pelén** (Peléní-) *n*. a personal name.
**pelérémòn** (pelérémònì-) *v*. to be squinted, squinty.
**penitésìyà** (penitésìyàà-) *n*. Catholic sacrament of penance.
**pɛntɛkɔ́stɛ̀** (pɛntɛkɔ́stɛ̀ɛ̀-) *n*. Pentecost: seventh Sunday after Easter.
**pɛ̀s** (pɛ̀sɛ̀) *ideo*. boom! (sound or effect of something explosive).
**pɛsɛlam** (pɛsɛlamá-) *n*. chip, small piece.
**pɛ́sɛ́lamed<sup>a</sup>** (pɛ́sɛ́lamede-) *n*. chip, small piece (of sth. in particular).
**pɛsɛlɛs** (pɛsɛlɛsí-) *v*. to break a piece off, chip, knap.
**pɛsɛ́mɛ́tɔ̀n** (pɛsɛ́mɛ́tɔ̀nì-) *v*. to break, chip, or crumble off in small pieces.
**pɛsɛ́pɛ́sánón** (pɛsɛ́pɛ́sánónì-) *v*. to be brittle, crumbly (e.g. biscuits).
**Pétèrò** (Pétèròò-) *1 n*. Peter. *2 n*. Peter: short New Testament letters.
**pìc** (pìcì) *ideo*. very full.
**piɗés** (piɗésí-) *v*. to cut across, go through, traverse. See also *tɔkɛ́ɛ́rɛ́s*.
**pìɗ<sup>ɨ</sup>** (pìɗì) *ideo*. sleekly, slickly. **pɨɗídɔ̀n** (pɨɗídɔ̀nì-) *v*. to be sleek, slick. See also *ɨpɛlípɛ́lɔ̀n*.
**Píipí** (Píipíì-) *n*. a personal name.
**pikódòn** (pikódònì-) *v*. to be worn smooth (of ground, stones, etc.).
**pìl** (pìlɔ̀) *ideo*. smoothly.
**pílè** (pílè) *ideo*. all, completely, totally.
**Pɨlɛmɔ́nɛ̀** (Pɨlɛmɔ́nɛ̀ɛ̀-) *n*. Philemon: book in the New Testament.
**Pílíkìts<sup>a</sup>** (Pílíkìtsì-) *n*. name of a mountain near *Káákʉma* in Turkanaland, Kenya.
**Pilípoik<sup>a</sup>** (Pilípo-icé-) *n*. Philippians: book in the New Testament.
**pɨlírímɔ̀n** (pɨlírímɔ̀nì-) *v*. to have strabismal amblyopia: a lazy eye that looks in different directions.
**pilís** (pilísì-) *n*. game of tag.
**pɨlɔ́dɨtɛ́sʉ́ƙɔt<sup>a</sup>** (pɨlɔ́dɨtɛ́sʉ́ƙɔtí-) *v*. to make smooth, smoothen out.
**pɨlɔ́dɔ̀n** (pɨlɔ́dɔ̀nì-) *v*. to be smooth.
**pìɔ̀** (pìɔ̀) *ideo*. splat! spoot! (sound of diarrhea or vomit). The vowels in this word are pronounced silently.
**pír** (pírí) *ideo*. glittering white.
**Píré** (Píréè-) *n*. name of a slope on the north of *Morúŋole* where some Ik used to be congregated.
**pirídòn** (pirídònì-) *v*. to be gleaming, shiny (like a bald head). See also *ɓarídɔ̀n*.
**piripírón** (piripírónì-) *v*. to gleam, glisten. See also *parɨpárɔ́n*.
**pɨrɨs** (pɨrɨsɨ) *ideo*. pop! (sound of coming or popping out, as in pus from a wound or an animal from a thicket).
**pìrrr** (pìrrr) *ideo*. very hot (of the weather).
**pìrⁱ** (pìrì) *ideo*. appearing suddenly, out of nowhere.
**pìs** (pìsì) *ideo*. quish! (sound of flesh being punctured). See also *tùs*.
**pítⁱ** (pítí) *ideo*. all.
**poɗés** (poɗésí-) *v*. to husk, shuck.
**poɗetés** (poɗetésí-) *v*. to husk, shuck.
**pɔ̀ɗ ɔ** (pɔ̀ɗɔ̀) *ideo*. agilely, nimbly, spryly.
**pɔɗɔ́dɔ̀n** (pɔɗɔ́dɔ̀nì-) *v*. to be agile, nimble, spry.
**pokés** (pokésí-) *1 v*. to break off. *2 v*. to stick, wedge (e.g. what mud does to a vehicle). See also *wakés*.
**poketés** (poketésí-) *v*. to break off.
**pókíetésá asíɛ́ kédìè kɔ̀n** *v*. to break off in groups.
**pòk<sup>o</sup>** (pòkò) *ideo*. breakably, brittlely.
**pokódòn** (pokódònì-) *v*. to be breakable, brittle. See also *ɛɔmɔ́dɔ̀n* and *wɛts'ɛ́dɔ̀n*.
**pokómón** (pokómónì-) *v*. to break off.
**pólìs** (pólìsì-) *n*. police.
**pɔ̀t ɔ** (pɔ̀tɔ̀) *ideo*. slickly, slipperily.
**pɔtɔ́dɔ̀n** (pɔtɔ́dɔ̀nì-) *v*. to be slick, slippery (e.g. a slimy wet rock).
**pòx** (pòxò) *ideo*. chattily, talkatively.
**poxés** (poxésí-) *v*. to peel, skin.
**pukésúƙot<sup>a</sup>** (pukésúƙotí-) *v*. to overturn away, turn out away (solid substances).
**Pʉlʉkɔ́l** (Pʉlʉkɔ́lì-) *n*. a personal name.
**pulúmétòn** (pulúmétònì-) *v*. to come out, egress, emerge.
**pulúmítésúƙot<sup>a</sup>** (pulúmítésúƙotí-) *v*. to take out.
**pulúmón** (pulúmónì-) *v*. to exit, go out.
**pulutetés** (pulutetésí-) *v*. to bring out, issue, produce (e.g. newborn twins out of the hut).
**pulutiés** (pulutiesí-) *1 v*. to perforate, pierce, or puncture repeatedly. *2 v*. to bore, drill. See also *ɨpɨrípírɛ́s*.
**púrurú** (púrurúù-) *n*. measles, rubeola.
#### rààrààrà
## **r**
respond. *3 v*. to account or answer for. *4 v*. to profit from.
**rátsiés** (rátsiesí-) *v*. to mend, patch, or repair repeatedly (by sewing).
string of bark for carrying). *2 v*. to skewer, spit.
**Rɔ́ɡɛ̀hò** (Rɔ́ɡɛ̀-hòò-) *n*. name of a hill or mountain. *Lit.* 'reedbuck-hut'.
**rutet<sup>a</sup>** (rutetí-) *n*. hillside, mountainside. **rutet<sup>o</sup>** *n*. along the side of a slope.
**rʉtsɛ́s** (rʉtsɛ́sí-) *v*. to cram, jam, ram, stuff. See also *ɨsɨkɛs*.
## **s**
**sêd<sup>a</sup>** (sédà-) *pl.* **sédìk<sup>a</sup>** . *n*. garden.
**seekw<sup>a</sup>** (seekó-) *n*. bouillon, broth, soup.
**Sɛkɛt<sup>a</sup>** (Sɛkɛtɛ́-) *n*. name of a river.
**sèrèy<sup>a</sup>** (sèrèì-) *pl.* **seréík<sup>a</sup>** . *n*. gourd basin made from a large gourd cut in half.
be sinewy, wiry (e.g. when malnourished).
mingling sticks are carved. *Albizia grandibracteata*.
**sír** (sírì-) *n*. dropsy, edema.
**sotés** (sotésí-) *v*. to carve, sculpt.
**sotetés** (sotetésí-) *v*. to carve, sculpt.
**sʉpaicíká ni nesíbòs** *n*. voiced vowels.
## **t**
are eaten raw, whose branches are used as fighting switches, and whose bark fibers are used to string beads. *2 n*. allium species: plant growing from a bulb and that produces a single large red flower worn by children as a hat. *Allium sp*.
**tábàìàm** (tábàì-àmà-) *pl.* **tábaiik<sup>a</sup>** . *n*. westerner.
**tábaɨxan** (tábaɨ-xaná-) *dem*. westerly direction.
**taɓolos** (taɓolosí-) *v*. to be exultant, gleeful, gloating.
**taɓóɲómòn** (taɓóɲómònì-) *v*. to have flat buttocks.
**taɗá** (<taɗɛɛs) *v*.
**taɗaɗáŋón** (taɗaɗáŋónì-) *v*. to be slightly bitter but edible.
**taɗaŋes** (taɗaŋesí-) *v*. to abide, bear, endure, put up with, stand, tolerate.
**taɗapes** (taɗapesí-) *1 v*. to mend, patch, repair. *2 v*. to ambush, waylay. See also *ɨɗaarɛ́s*.
**taɗapetés** (taɗapetésí-) *1 v*. to mend or patch up, repair. *2 v*. to ambush, waylay.
**Takaniƙʉlɛ́**(Takani-ƙʉlɛ́ɛ̀-) *n*. name of a place where the Turkana found an old woman because her elbow was visible. *Lit.* 'appearing-elbow'.
**takanités** (takanitesí-) *v*. to detect, discover, find.
**takánón** (takánónì-) *1 v*. to be present, seen, visible. *2 v*. to be clear, evident, obvious.
**takár** (takárí-) *pl.* **takárík<sup>a</sup>** . *1 n*. forehead. *2 n*. face, visage.
**takárɛ̂d a** (takárɛ́dɛ̀-) *n*. face, front.
**takátákánón** (takátákánónì-) *v*. to be cracked, fissured, fractured (e.g. heated rock, dried out mud puddle). See also *médemedánón*.
**tákés** (tákésì-) *v*. to mean, mention, refer to.
**takiés** (takiesí-) *v*. to lift carefully (palm upward).
**takwés** (takwésí-) *v*. to step or tread on.
**takwésá ɲamɨɨlɨí** *v*. to pedal a bicycle.
**takwiesúƙota dáŋá<sup>e</sup>** *v*. to trample edible termites.
**takwitakwiés** (takwitakwiesí-) *v*. to step all over, trample.
**taƙáá na ŋáɲɔ́s** *n*. open-toed shoe.
**taƙámón** (taƙámónì-) *v*. to chance upon, come across, happen upon.
**tàƙàt<sup>a</sup>** (tàƙàtà-) *n*. call-and-response group prayer.
**taƙates** (taƙatesí-) *v*. to lead call-andresponse group prayers.
**taƙátésuƙot<sup>a</sup>** (taƙátésuƙotí-) *v*. to pray against/away.
**taƙáy<sup>a</sup>** (taƙáí-) *pl.* **taƙáík<sup>a</sup>** . *n*. shoe.
**talakes** (talakesí-) *1 v*. to free, let go, release. *2 v*. to allow, permit. See also *hoɗés*.
**tàlàlìdòm** (tàlàlìdòmò-) *n*. small animal species that steals food from homes, especially from pots of food (possibly a type of mongoose).
**talóón** (talóónì-) *v*. to be nauseated, queasy.
**tamá** (<tamɛɛs) *v*.
**tamáísánón** (tamáísánónì-) *v*. to smile.
**tamanɛs** (tamanɛsí-) *v*. to circumvent, encircle, go around, skirt.
**tamanɛtɛ́s** (tamanɛtɛ́sí-) *v*. to circumvent, encircle, go around, skirt.
**tamátámatés** (tamátámatésí-) *v*. to consider, contemplate, mull over, ponder, think on. See also *tamítámiés*.
**Tamateeɓon** (Tamateeɓonó-) *n*. name of a river.
**tamɛɛs** (tamɛɛsí-/tamá-) *1 v*. to adore, extol, laud, praise (e.g. one's spouse, a friend, or even an animal like a favorite ox; may involve complimentary words and affectionate physical touch). *2 v*. to give an admiring nickname to.
**tamɛɛsíêd<sup>a</sup>** (tamɛɛsí-édì-) *pl.* **tamɛɛsíéditín**. *n*. affectionate nickname.
**tamɛ́s** (tamɛ́sí-) *v*. to think.
**tamɛ́síàm** (tamɛ́sí-àmà-) *pl.* **tamɛ́síik<sup>a</sup>** . *n*. contemplative, thinker.
**tamɛ́sʉ́ƙɔt<sup>a</sup>** (tamɛ́sʉ́ƙɔtí-) *v*. to recall, recollect, remember, think back on.
**tamɛtɛ́s** (tamɛtɛ́sí-) *1 v*. to consider, imagine, ponder, think about. *2 v*. to recall, recollect, remember. See also *anɛtɛ́s*.
**támínɔ́s** (támínɔ́sí-) *1 v*. to think about each other. *2 v*. to compare each other.
**tamítámiés** (tamítámiesí-) *v*. to consider, contemplate, mull over, ponder, think on. See also *tamátámatés*.
**taná** (<tanɔ́ɔ́n) *v*.
**tànàŋ** (tànàŋà-) *n*. mud plaster.
**tapá** (<tapɛɛs) *v*.
**tapáínɔ́s** (tapáínɔ́sí-) *v*. to accuse or blame each other falsely.
**tatíón** (tatíónì-) *v*. to drip (of rain).
**tatiós** (tatiosí-) *v*. to be blessed.
**tàtòn** (tàtònì-) *v*. to spit.
**tatónón** (tatónónì-) *v*. to sit dejectedly with one's chin in one's hand or on one's knees (due to coldness, depression, sadness, etc.).
**tatsá** (<tatsɔ́ɔ́n) *v*.
**tawanes** (tawanesí-) *v*. to afflict, harm, hurt.
**tawanímétòn** (tawanímétònì-) *v*. to be afflicted, badly off, suffering.
**tawánítetés** (tawánítetésí-) *v*. to afflict, harm, hurt.
**tɛ́bɛ̀s** (tɛ́bɛ̀sì-) *v*. to scoop, take up.
**Teɓur** (Teɓurí-) *n*. Abim, Labwor.
**tɛ́ɛ́tɔ̀n** (tɛ́ɛ́tɔ̀nì-) *v*. to drop, fall.
**teɡeles** (teɡelesí-) *v*. to bar, barricade, block. See also *toƙólésuƙot<sup>a</sup>* .
building and fencing; a bark decoction is drunk to treat pain and snakebites. *Ziziphus mucronata*.
**tiŋátiŋá** (tiŋátiŋáà-) *n*. rat species.
**tírɨríŋɔ́n** (tírɨríŋɔ́nì-) *v*. to be fortunate, lucky.
**tiritirikwáy<sup>a</sup>** (tiritiri-kwáì-) *n*. vine species that is worn around the necks of age-group initiates' wives during the post-initiation ceremony of beer drinking and slaughtering a goat.
**tiróŋ** (tiróŋí-) *pl.* **tiróŋík<sup>a</sup>** . *n*. molar.
**titímáiƙot<sup>a</sup>** (<titímóonuƙot<sup>a</sup> ) *v*.
**titiretés** (titiretésí-) *1 v*. to hold up, prop up, support, undergird. *2 v*. to delay, hold up, prevent from doing.
**tɨts'ɛ́s** (tɨts'ɛ́sí-) *1 v*. to block, dam, plug. *2 v*. to conceal, cover (a hole, the truth).
**tɨts'ɛ́sʉ́ƙɔt<sup>a</sup>** (tɨts'ɛ́sʉ́ƙɔtí-) *1 v*. to block, dam, or plug up. *2 v*. to conceal, cover up (a hole, the truth).
**tódinós** (tódinósí-) *v*. to converse, speak or talk to each other, tell each other.
**tɔkɛ́tɛ́sʉƙɔt<sup>a</sup>** (tɔkɛ́tɛ́sʉƙɔtí-) *v*. to extract, pull off/out, remove.
**tɔkɛtɛtɛ́s** (tɔkɛtɛtɛ́sí-) *v*. to extract, pull out, remove. See also *toletés*.
**tɔkíɔ́n** (tɔkíɔ́nì-) *v*. to confess.
**Tɔ̀kɔ̀b a** (Tɔ̀kɔ̀bà-) *n*. a personal name.
**tɔ̀kɔ̀b a** (tɔ̀kɔ̀bà-) *v*. agriculture, cultivation, farming, plowing.
**tɔ̀kɔ̀bààm** (tɔ̀kɔ̀bà-àmà-) *pl.* **tɔkɔbaik<sup>a</sup>** . *n*. agriculturalist, cultivator, farmer, plower.
**tɔkɔbam** (tɔkɔbamá-) *n*. arable, cultivatable, farmable, tillable.
**tɔkɔbatsóy<sup>a</sup>** (tɔkɔba-tsóí-) *n*. farming season, plowing season.
**tɔkɔ́bɛs** (tɔkɔ́bɛsí-) *v*. to cultivate, farm, plow, till, work.
**tɔkɔ́bɨtɛtɛ́s** (tɔkɔ́bɨtɛtɛ́sí-) *v*. to make cultivate or plow (e.g. oxen).
**tɔkɔ́bɨtɔtɔ́s** (tɔkɔ́bɨtɔtɔ́sí-) *v*. to be cultivated, plowed, tilled (usually by oxen).
**tokódòn** (tokódònì-) *v*. to be tightly tied down so as to be immovable.
**tɔkɔɗɛs** (tɔkɔɗɛsí-) *v*. to grasp, hold by a handle.
**tɔkɔɗíkɔ́ɗɔ̀n** (tɔkɔɗíkɔ́ɗɔ̀nì-) *v*. to cramp (abdominally).
**tɔ̀kɔ̀n** (tɔ̀kɔ̀nì-) *v*. to be slender, slim. See also *kaɗótsómòn* and *sídɔ̀rɔ̀mɔ̀n*.
**tokopes** (tokopesí-) *v*. to grab, seize, snatch. See also *ɨkamɛs*, *ɨrɛɗɛs*, and *ŋusés*.
**tokópésuƙot<sup>a</sup>** (tokópésuƙotí-) *v*. to grab, seize, or snatch away. See also *taŋátésuƙot<sup>a</sup>* .
**tɔkɔrɛs** (tɔkɔrɛsí-) *1 v*. to dispense, disperse, distribute, divide out/up. *2 v*. to divide mathematically. See also *kisanes* and *kísés*.
**tɔkɔ́rɛ́sʉƙɔt<sup>a</sup>** (tɔkɔ́rɛ́sʉƙɔtí-) *v*. to dispense, disperse, distribute, divide out/up. See also *kisanes* and *kísés*.
**tɔkɔ́rʉ́kɔ́rɛ́sʉƙɔt<sup>a</sup>** (tɔkɔ́rʉ́kɔ́rɛ́sʉƙɔtí-) *v*. to dispense, disperse, distribute, divide out/up. See also *kisanes* and *kísés*.
**tokúétòn** (tokúétònì-) *v*. to react suddenly: crack, jerk, snap. See also *tokúréètòn*.
**tokúréètòn** (tokúréètònì-) *v*. to react suddenly: crack, jerk, snap. See also *tokúétòn*.
**tɔƙam** (tɔƙamá-) *n*. edible termites that are dewinged and dried.
**toƙírá** (<toƙíróòn) *v*.
**toƙíróòn** (toƙíróònì-/toƙírá-) *v*. to bear down, charge, react against, turn on (as if to attack).
**toƙízèètòn** (toƙízèètònì-) *v*. to move in, settle in (e.g. bad weather).
**toƙízòòn** (toƙízòònì-/toƙíza-) *v*. to hang around, stay a while, stick around (e.g. bad weather).
**toƙólésuƙot<sup>a</sup>** (toƙólésuƙotí-) *v*. to bar, barricade, block off. See also *teɡeles*.
**tɔƙɔtɔƙ<sup>a</sup>** (tɔƙɔtɔƙɔ́-) *n*. slug, snail.
**tɔƙɔtɔƙáhò** (tɔƙɔtɔƙá-hòò-) *pl.* **tɔƙɔtɔƙáhoík<sup>a</sup>** . *n*. snail shell.
**tɔƙʉmʉ́ƙʉ́mɛ́s** (tɔƙʉmʉ́ƙʉ́mɛ́sí-) *v*. to fire on, open fire on.
**topéɗésuƙot<sup>a</sup>** (topéɗésuƙotí-) *1 v*. to be able to, can, capable of. *2 v*. to have authority.
**topéɗésuƙotíám** (topéɗésuƙotí-ámà-) *pl.* **topéɗésuƙotíík<sup>a</sup>** . *n*. person in authority.
**toráƙádòs** (toráƙádòsì-) *n*. nongovernmental organization (NGO).
**towá** (<towóón) *v*.
**Tówotó** (Tówotóò-) *n*. nickname of an old woman named *Matsú*.
**towutses** (towutsesí-) *v*. to bulldoze, level, raze.
**towútsónìàm** (towútsónì-àmà-) *pl.* **towútsóniik<sup>a</sup>** . *n*. bulldozer.
**tʉɗʉtɛtɛ́s** (tʉɗʉtɛtɛ́sí-) *v*. to stiffen by stirring (e.g. meal mush).
**tʉɗʉtɔs** (tʉɗʉtɔsí-) *v*. to be stiff from being stirred (e.g. meal mush).
**tùf** (tùfà) *1 ideo*. bouncily, springily. *2 ideo*. bluntly, dully.
## **ts**
```
tsá (<tsóón) v.
```
thin stick is drilled to create enough friction to light a fire. *Lit.* 'firedrillwoman'.
**tsàm** (tsàmʉ̀) *adv*. just, I guess, I suppose.
**tsàpòn** (tsàpònì-) *v*. to be bored or drilled full of holes (e.g. wood bored by insects).
**tsè** (<tsòòn) *v*.
**tsɛrɛ́k a** (tsɛrɛ́kí-) *pl.* **tsɛrɛ́kík<sup>a</sup>** . *n*. shinbone, tibia.
person (who crouches instead of sits or moves around instead of settles).
**tsúts<sup>a</sup>** (tsútsá-) *n*. fly.
**tsʉ̀tᶶ** (tsʉ̀tʉ̀) *adv*. ever, never.
**tsʉ́ʉ́r** (tsʉ́ʉ́rà-) *pl.* **tsʉ́ʉ́rìk<sup>a</sup>** . *n*. white thorn acacia: a scrubby, plentiful tree species whose extremely hard wood is good for building and fencing; its resin is chewed and a bark decoction taken for bloating. *Acacia hockii*.
**tsùwà** (tsùwàà-) *1 v*. to run. *2 v*. to race.
#### ts'ábès
## **ts'**
**ts'aɗídàkw<sup>a</sup>** (ts'aɗí-dàkù-) *n*. firewood.
**ts'áɡòn** (ts'áɡònì-) *v*. to be cruddy, dirty, filthy, grimy. See also *itútsón*.
**ts'aɡús** (ts'aɡúsé) *num*. four.
**ts'aɡusátìk<sup>e</sup>** *v*. four-by-four.
**ts'aɡús<sup>o</sup>** *num*. four times.
**ts'aɡúsón** (ts'aɡúsónì-) *v*. to be four.
**ts'áɡwà** (<ts'áɡwòòn) *v*.
**ts'aletés** (ts'aletésí-) *v*. to remove, take out (as from a cooking pot).
**ts'é** (<ts'óón) *v*.
**ts'ɛ̀** (ts'ɛ̀à-) *pl.* **ts'ɛ́ítín**. *n*. hide, pelt, skin.
**ts'ɛa na kwaní** *n*. foreskin.
**ts'eiƙot<sup>a</sup>** (<ts'oonuƙot<sup>a</sup> ) *v*.
**ts'éíƙot<sup>a</sup>** (<ts'óónuƙot<sup>a</sup> ) *v*.
**ts'eites** (ts'eitesí-) *v*. to extinguish, put out, quench.
**ts'eítésuƙot<sup>a</sup>** (ts'eítésuƙotí-) *1 v*. to extinguish, put out, quench. *2 v*. to switch or turn off electrically. *3 v*. to snuff out (i.e. kill, murder).
**ts'ɛ̀tà kɔ̀nà** *n*. likewise, in the same way.
**ts'íínáá** (ts'íínáá) *n*. all the time.
**ts'íínúó** (ts'íínúó) *n*. everywhere.
**ts'ɨƙ<sup>a</sup>** (ts'ɨƙá-) *1 n*. bee, honeybee. *Apis mellifera*. *2 n*. honey.
**ts'ɨƙábòt<sup>a</sup>** (ts'ɨƙá-bòtà-) *pl.* **ts'ɨƙábotitín**. *n*. swarm of bees on the move.
**ts'ɨƙáhò** (ts'ɨƙá-hòò-) *pl.* **ts'ɨƙáhoík<sup>a</sup>** . *n*. honeycomb. *Lit.* 'bee-house'. See also *ɗàɗàhò*.
**ts'ìn** (ts'ìnɔ̀-) *n*. taboo against husbands of pregnant wives participating in any food-gathering activities such as hunting.
**ts'ìnɔ̀àm** (ts'ìnɔ̀-àmà-) *pl.* **ts'ɨnoik<sup>a</sup>** . *n*. husband prohibited from participating in food-gathering activities due to the pregnancy of his wife.
**ts'ìnɔ̀n** (ts'ìnɔ̀nì-) *v*. to be pregnant, with the result that one's husband is prohibited from participating in any foodgathering activities.
**ts'írítɔ̀n** (ts'írítɔ̀nì-) *v*. to spurt, squirt (e.g. blood or spit).
**ts'íts'ɛ́s** (ts'íts'ɛ́sì-) *v*. to track.
**ts'íts'ɛ́sà dɛ̀ìkà<sup>ɛ</sup>** *v*. to track footprints.
**ts'íts'ítɛ́sʉƙɔt<sup>a</sup>** (ts'íts'ítɛ́sʉƙɔtí-) *v*. to make sharp, sharpen.
**ts'íts'ítɛtɛ́s** (ts'íts'ítɛtɛ́sí-) *v*. to make sharp, sharpen.
**ts'ɔƙɔ́m** (ts'ɔƙɔ́má-) *n*. tree species whose yellow fruits are eaten raw, whose seeds are cracked, mashed, and fried to be eaten, and from whose wood stools and troughs are carved. *Sclerocarya birrea*.
**ts'òlòn** (ts'òlònì-) *v*. to dribble, drip, drizzle. See also *ɨɗɔ́nɔ́n*.
**ts'óónuƙot<sup>a</sup>** (ts'óónuƙotí-/ts'éíƙot-) *v*. to die off/out (of more than one). Not to be confused with *ts'oonuƙot<sup>a</sup>* .
**ts'ʉ́bʉlát<sup>a</sup>** (ts'ʉ́bʉlátí-) *pl.* **ts'ʉ́bʉlátíkw<sup>a</sup>** . *n*. plug or stopper made from grass or leaves. See also *ts'ʉ̂b a* .
**ts'ûd<sup>a</sup>** (ts'údè-) *1 n*. smoke. *2 n*. exhaust, fumes. *3 n*. smoke signal. *4 n*. tobacco.
#### ts'údemucé
**ts'ʉnɛ́s** (ts'ʉnɛ́sí-) *v*. to kiss.
**ts'ʉts'ʉ́áw<sup>a</sup>** (ts'ʉts'ʉ́-áwà-) *pl.* **ts'ʉts'ʉ́áwík<sup>a</sup>** . *n*. garbage dump, rubbish pile.
## **u/ʉ**
**útɔ̀** (útɔ̀ɔ̀-) *n*. seed oil (e.g. simsim or sunflower).
## **w**
**wat<sup>a</sup>** (waté-) *n*. rain.
**wàxɛ̀d a** (wàxɛ̀dɛ̀-) *n*. point of departure, starting point.
**wàxʉ̀** *1 n*. ahead, in front. *2 n*. before, earlier, first.
**wâz** (wázò-) *pl.* **wázìkw<sup>a</sup>** . *n*. young female (human or non-human).
**wɛ́dɔ̀n** (wɛ́dɔ̀nì-) *v*. to detour, change course, take a diverson. See also *ƙeƙérón*.
**wɛɗɨwɛ́ɗɔ́n** (wɛɗɨwɛ́ɗɔ́nì-) *v*. to flap, flutter.
**wɛ́ɛ́nɔ̀n** (wɛ́ɛ́nɔ̀nì-) *v*. to be fast, quick, speedy. See also *itírónòn*.
**weés** (weésí-/wa-) *v*. to harvest, reap.
**weesá dakwí** *v*. to collect firewood.
**weesa kíʝá<sup>e</sup>** *v*. to play the field (sexually), sleep around. *Lit.* 'reaping the land'.
Wús
## **x**
**xàɓ<sup>u</sup>** (xàɓù) *ideo*. softly.
**xakútsúmòn** (xakútsúmònì-) *v*. to be abyssal, bottomless, unfathomable.
**xaƙarés** (xaƙarésí-) *v*. to clear the throat, harrumph, hawk. See also *hákátòn*.
**xáƙátòn** (xáƙátònì-) *v*. to gag, heave, retch. See also *ʝaƙátós*.
**xáƙw<sup>a</sup>** (xáƙʉ́-) *pl.* **xáƙwítín**. *n*. dried flesh left on a skinned hide. Compare with *ɲɔpɔɗɛ*.
**xán** (xáná-) *pl.* **xáín**. *1 n*. side. *2 n*. direction, location.
**xàr** (xàrà-) *pl.* **xárítín**. *n*. bladder.
**xaramucé** (xara-mucéè-) *pl.* **xárítínímucéík<sup>a</sup>** . *n*. urethra.
**xawííts'<sup>ɨ</sup>** (xawííts'ì) *ideo*. burnt to ashes, incinerated.
**xɛɓás** (xɛɓásí-) *n*. cowardice, fear, shyness, timidity.
**xɛɓásíàm** (xɛɓásí-àmà-) *pl.* **xɛɓásíik<sup>a</sup>** . *n*. coward, shy or timid person.
**xɛ̀ɓ ɛ** (xɛ̀ɓɛ̀) *ideo*. tenderly (of plants).
**xɛɛsʉ́ƙɔt<sup>a</sup>** (xɛɛsʉ́ƙɔtí-) *1 v*. to dust, sprinkle. *2 v*. to return a bride (to her home by sprinkling water on the gate of her parents' home).
**xɛɛtɛ́s** (xɛɛtɛ́sí-) *v*. to dust, sprinkle.
**xèr** (xèrà-) *n*. belch, burp.
**xɨkɛ́s** (xɨkɛ́sí-) *v*. to hang up.
**xɛɓɛ́dɔ̀n** (xɛɓɛ́dɔ̀nì-) *v*. to be tender (of new plant growth). See also *rusúdòn*.
**xɛɓɨtɛs** (xɛɓɨtɛsí-) *v*. to frighten, intimidate, scare.
# **y**
**Yakóɓò** (Yakóɓòò-) *1 n*. Jacob, James. *2 n*. James: New Testament book.
**yàŋ** (yàŋà) *ideo*. sludgily.
**yáŋìɲòt<sup>a</sup>** (yáŋì-ɲòtà-) *n*. my mother-inlaw (sibling's spouse's mother).
**yáó** (yáóò-) *pl.* **yáóín**. *n*. your sister.
**yáóím** (yáó-ímá-) *pl.* **yáówík<sup>a</sup>** . *n*. your niece or nephew (sister's child).
**Yarán** (Yarání-) *n*. a personal name.
**yeá** (yeáà-) *pl.* **yeáín**. *n*. my sister.
**yeé'** (yeé') *interj*. huh! (an expression of derision or disbelief). The apostrophe at the end represents a glottal stop, an abrupt end to the word's voicing.
**Yésù** (Yésùù-) *n*. Jesus.
**yír** (yírí) *ideo*. zoom! (sound of a swiftly moving object). See also *wír*.
**yù** (yùù) *ideo*. softly (of soil).
**yʉɛ** (yʉɛ́-) *n*. falsehood, lie, prevarication, untruth.
**yʉɛ́ám** (yʉɛ́-ámà-) *pl.* **yuéík<sup>a</sup>** . *n*. liar.
**yùm** (yùmù) *ideo*. softly inside.
**yús** (yúsì-) *n*. young people, youth.
**yuúdòn** (yuúdònì-) *v*. to be soft (of soil).
**yɛ̀l** (yɛ̀lɛ̀) *ideo*. dryly.
## **z**
**zeiƙot<sup>a</sup>** (<zoonuƙot<sup>a</sup> ) *v*.
*v*. to develop, foster, grow, promote (e.g. plants or the economy).
age, mature. See also *iwánétòn*.
**zɔ́tɛ́tɔ̀n** (zɔ́tɛ́tɔ̀nì-) *v*. to join, link up.
## **Part III**
## **English-Ik reversal index**
**a** kɔ́níɛ́n *pro*. **a bit** kédììm *n*. **a few** kíníɛ́n *pro*. **a fifth (be)** kɔnɔna túdònù *v*. **a little bit** kédiima kwáts<sup>a</sup> *n*. **a little while** kédiima kwáts<sup>a</sup> *n*. **a long time** dèŋ *ideo*.; kéda zikîb<sup>a</sup> *n*.; ódowicíká nì kòm *n*. **a long while** kéda zikîb<sup>a</sup> *n*.; ódowicíká nì kòm *n*. **a lot** mbáyà *adv*.; pʉ́n *ideo*.; tàn *n*. **a short time** kédiima kwáts<sup>a</sup> *n*.; ódowicíká nì ƙwàɗ<sup>e</sup> *n*. **a short while** ódowicíká nì ƙwàɗ<sup>e</sup> *n*. **a time** kédììm *n*. **a while** kédììm *n*. **a while ago** nótsò *adv*. **a year ago** sakɛɨn *n*. **aahǃ** wóí *interj*. **aardvark** ɲoɗôd<sup>a</sup> *n*. **aardwolf** náƙírà *n*. **ab** ɲákwálɨkwal *n*. **abandon** ɡóózés *v*.; ɡóózesuƙot<sup>a</sup> *v*. **abandoned** ɡóózosuƙot<sup>a</sup> *v*. **abbattoir** hoesího *n*. **abdomen** bùbù *n*. **abdominal disease** ɡɛf *n*. **abdominal muscle** ɲákwálɨkwal *n*. **abduct** ɨɛpɛtɛ́s *v*.; toreɓes *v*. **abhor** tʉlʉŋɛs *v*. **abide** nɛɛ́s *v*.; nɛɛsʉ́ƙɔt<sup>a</sup> *v*.; taɗaŋes *v*. **ability** ɲapéɗór *n*. **Abim** Teɓur *n*. **ablactate** topétésuƙot<sup>a</sup> *v*. **able to** topéɗésuƙot<sup>a</sup> *v*. **able-bodied (of middle age)** toipánón *v*.
**able-bodied adult** toipánónìàm *n*. **abode** aw<sup>a</sup> *n*.; zɛƙɔ́áw<sup>a</sup> *n*. **abominate** tʉlʉŋɛs *v*. **abort** iɲétséetés *v*.; iɲétsóòn *v*.; iyétséetés *v*.; iyétsóòn *v*. **abort repeatedly** iyétséyeés *v*. **above** nɔ́ɔ́kwar<sup>ɔ</sup> *n*. **abruptly** ŋàm *ideo*.; ùrùƙùs *ideo*. **abscess** tún *n*. **absent-minded** ɓotsódòn *v*. **absorb (attention)** itúmúránitésúƙot<sup>a</sup> *v*. **absorb heat** róbiróbòn *v*. **absorbed** ts'ʉ́ʉ́tɔnʉƙɔt<sup>a</sup> *v*. **absorbed (attention)** itúmúránón *v*. **abstain (from food)** ɨkáɲɔ́n *v*. **abuse** ɨlarɛs *v*.; ɨlwarɛs *v*. **abuse sexually** tarates *v*. **abuse verbally** iyaŋes *v*. **Abutilon species** ídòkàk<sup>a</sup> *n*. **abyssal** xakútsúmòn *v*. **Abyssinia** Isópìà *n*.; Sópìà *n*. **acacia (Gerrard's)** ɓʉkʉ́lá *n*. **acacia (gum)** ɗerét<sup>a</sup> *n*.; lofílitsí *n*. **acacia (hook-thorn)** kásít<sup>a</sup> *n*. **acacia (Naivaisha thorn)** lɛ̀r *n*. **acacia (Nubian)** ɡɔmɔr *n*. **acacia (scented-pod)** kɨlɔ́rít<sup>a</sup> *n*. **acacia (wait-a-bit)** kʉ́rà *n*. **acacia (white thorn)** tsʉ́ʉ́r *n*. **Acacia brevispica** kʉ́rà *n*. **Acacia gerrardii** ɓʉkʉ́lá *n*. **Acacia hockii** tsʉ́ʉ́r *n*. **Acacia mellifera** kásít<sup>a</sup> *n*. **Acacia nilotica** kɨlɔ́rít<sup>a</sup> *n*. **Acacia nubica** ɡɔmɔr *n*. **Acacia senegal** ɗerét<sup>a</sup> *n*.; lofílitsí *n*. **Acacia species** ròr *n*.; tíír *n*.
**Acacia tortilis** sèɡ<sup>a</sup> *n*. **Acacia xanthophloea** lɛ̀r *n*. **Acalypha fruticosa** ʝɨʝîd<sup>a</sup> *n*. **accept** tsamɛtɛ́s *v*. **accept a curse** tsamɛ́tɛ́sa ìlàmà<sup>ɛ</sup> *v*. **accident (have an)** rúmánòn *v*. **accompany** iníámésuƙot<sup>a</sup> *v*. **accompany away** elánónuƙot<sup>a</sup> *v*. **accompany here** elánétòn *v*. **account** ɲáɗís *n*. **account for** raʝetés *v*. **accountable** wàsɔ̀n *v*. **accumulate** ɨɗaɲɛtɛ́s *v*.; torítéetés *v*. **accumulate one-by-one** ɨɗɛ́bɛtɛ́s *v*. **accuse** ɨsíítɛ́s *v*. **accuse each other falsely** kɛ́rínɔ́s *v*.; tapáínɔ́s *v*. **accuse falsely** ipwáákés *v*.; kɛrɛ́s *v*.; tapɛɛs *v*. **accuser** ɨsíítɛ́sìàm *n*. **accuser (false)** kɛ́rínɔ́síàm *n*.; lóliit<sup>a</sup> *n*. **accustom** ɨtalɛs *v*.; naínɛ́ɛtɛ́s *v*.; naítɛ́sʉƙɔt<sup>a</sup> *v*. **accustomed** ɨtalɔs *v*. **acerbic** ɓatsilárón *v*. **ache** áts'ɛ́s *v*.; dódòn *v*. **ache (begin to)** dódètòn *v*. **ache (of joints)** ɨtɨlítílɔ̀n *v*. **ache (of lymph nodes)** titisíánón *v*. **ache (of teeth)** isálílòn *v*. **achieve** enés *v*. **Acholi language** Ŋíkátsolítôd<sup>a</sup> *n*. **Acholi person** Ŋíkátsolíám *n*. **acid reflux** áts'ɛ́sa ɡúró<sup>e</sup> *n*. **acne** ŋ́kaŋók<sup>a</sup> *n*. **acolyte** túbèsìàm *n*.
**acquaint** naínɛ́ɛtɛ́s *v*.; naítɛ́sʉƙɔt<sup>a</sup> *v*. **acquainted (mutually)** náínɔ́s *v*. **acquiesce to** tsamɛtɛ́s *v*. **acquire** iryámétòn *v*. **acquit** ɨɛtɛ́sá ɨsíítɛ́sʉ́ *v*. **acrid** ɓatsilárón *v*. **act** iɲétsóòn *v*.; iyétsóòn *v*. **act falsely** ipwáákés *v*. **act in sync** ilíréètòn *v*. **act in unison** ilíréètòn *v*. **actionable** itíyéetam *n*. **active** ɨkwɨɲíkwíɲɔ́n *v*.; iona ńda sea ni itsúr *v*.; kʉrʉ́kʉ́rɔ́s *v*.; kwɨɲídɔ̀n *v*.; tsuwoós *v*. **actively** kwìɲ *ideo*. **Acts of the Apostles** Teréɡiicíká Dɛɨkaicé *n*. **actually** kárɨká *adv*.; rò *adv*. **add** ɗɔtsɛ́s *v*.; ɗɔtsɛ́sʉ́ƙɔt<sup>a</sup> *v*.; ɨatɛs *v*. **add in** ɗɔtsɛ́s *v*.; ɗɔtsɛ́sʉ́ƙɔt<sup>a</sup> *v*. **add on** tasaɓes *v*. **add on top** iɗokes *v*. **add repeatedly** ɨatiés *v*.; ɨatiésuƙot<sup>a</sup> *v*. **add together** ɗɔtsɛ́s *v*.; ɗɔtsɛ́sʉ́ƙɔt<sup>a</sup> *v*. **add up** ɗɔtsɛtɛ́s *v*. **add water to** ɨlálátɛ́s *v*.; ɨlatɛs *v*. **added** ɗɔtsɔ́s *v*.; ɨatímétòn *v*.; ɨatɔs *v*. **adder (dwarf)** narɛ́w<sup>a</sup> *n*. **adder (puff)** bɛf *n*. **Adenium obesum** dɛrɛ́ƙ <sup>a</sup> *n*.; ʝɔt<sup>a</sup> *n*. **adhere** ƙídzɔ̀n *v*.; nɔtsɔ́mɔ́n *v*. **adhere to** ɨnɔtsɛs *v*. **adhesive** nɔtsɔ́dɔ̀n *v*. **adhesively** nɔ̀ts<sup>ɔ</sup> *ideo*. **adipose tissue** ceím *n*.; ɛfás *n*. **adjacent** ɨɓákɔ́n *v*. **adjacent to (move)** ɨɓákɔ́nʉƙɔt<sup>a</sup> *v*.
**adjudicate** ŋʉrɛ́s *v*.; ŋurutiés *v*.; ŋurutiesúƙot<sup>a</sup> *v*. **adjudicated** ŋurutiós *v*. **administer** ɨmɔlɛs *v*. **administer in drops** ts'olites *v*.; ts'olítésuƙot<sup>a</sup> *v*. **administration** ɲápukán *n*.; ɲeryaŋ *n*. **administrator** ɨmɔlɛsíám *n*. **admire** iséméés *v*.; iséméetés *v*. **admirer** mínɛ́sìàm *n*. **adopt (a mascot)** totores *v*. **adore** mínɛ́s *v*.; tamɛɛs *v*. **adorn** daites *v*.; naƙwídɛtɛ́s *v*. **adornment** ɲewale *n*. **adult** ámáze *n*. **adulterer** ɓúƙónìàm *n*. **adulteress** ɓúƙónìàm *n*. **adulthood (of many)** roɓazeikánánès *n*. **adults** roɓazeik<sup>a</sup> *n*. **advertise (attributes)** ƙɔƙɔanón *v*. **advice each other** táŋínɔ́s *v*. **advise** táŋɛ́s *v*.; taŋɛtɛ́s *v*. **adze** lorokon *n*. **aerate** iwúlákés *v*. **affair (have an)** ríínós *v*. **affect** kʉpɛ́s *v*. **affix** mínɛ́s *v*. **afflict** sikwetés *v*.; tawanes *v*.; tawánítetés *v*. **afflicted** tawanímétòn *v*. **affluence** ídzànànès *n*. **affront** risés *v*.; tatés *v*. **aflutter** kìtòn *v*. **afraid** paupáwón *v*.; xɛ̀ɓɔ̀n *v*. **Africa** Buɗámóniicékíʝ<sup>a</sup> *n*. **African** buɗámónìàm *n*.
**after** ʝìrʉ̀ *n*.; térútsù *adv*.; tórútsù *adv*. **after all** ʝâb<sup>o</sup> *adv*. **afterlife** didiɡwarí *n*. **again** naɓó *adv*. **age** dunétón *v*.; zeís *n*.; zoonuƙot<sup>a</sup> *v*. **age-group** ɲanákɛ́t <sup>a</sup> *n*. **age-group (Blood-Strugglers)** Ŋímarɨɔkɔ́t <sup>a</sup> *n*.; Ŋuésíìkà sèà<sup>e</sup> *n*. **age-group (Buffalo)** Gasaraik<sup>a</sup> *n*.; Ŋíkósowa *n*. **age-group (Eland)** Basaúréik<sup>a</sup> *n*.; Ŋíwápɛtɔ *n*. **age-group (Elephant)** Ŋɨtɔ́mɛ́ *n*.; Oŋoriik<sup>a</sup> *n*. **age-group (Gazelle)** Kodowíík<sup>a</sup> *n*.; Ŋíŋóleɲaŋ *n*. **age-group (Giraffe)** Gwaíts'íik<sup>a</sup> *n*.; Ŋkóryó *n*. **age-group (Leopard)** Ŋisíráy<sup>a</sup> *n*.; Nʉsíík<sup>a</sup> *n*. **age-group (Lion)** Máóik<sup>a</sup> *n*.; Ŋíŋátuɲo *n*. **age-group (Ostrich)** Leweɲiik<sup>a</sup> *n*.; Ŋímérimoŋ *n*.; Ŋ́kaleesó *n*. **age-group (Umbrella Thorn)** Ŋítíɨra *n*.; Seɡaik<sup>a</sup> *n*. **age-group (Zebra)** Ŋítúkoy<sup>a</sup> *n*.; Zɨnáík<sup>a</sup> *n*. **age-set** ɲanákɛ́t <sup>a</sup> *n*. **aggravate** ɨwíwínɛ́s *v*.; rúbès *v*. **aggregate rock** ɲɛ́kɔ́kɔ́tɛ́*n*. **agile** pɔɗɔ́dɔ̀n *v*. **agilely** pɔ̀ɗ ɔ *ideo*. **agitate** íbɔbɔtsɛ́s *v*.; íbɔtsɛ́s *v*.; ɨlɔ́lɔ́ŋɛ́s *v*. **agitated** íbɔtsɛ́sá así *v*.; iƙúrúmós *v*.; walɨwálɔ́n *v*. **ago** kwààk<sup>e</sup> *n*. **agree to** tsamɛtɛ́s *v*. **agree with each other** tsámʉ́nɔtɔ́s *v*. **agreeableness** daás *n*.
#### agreeble
**agreeble** dòòn *v*. **agricultural course** áɡɨrɨkácà *n*. **agriculturalist** tɔ̀kɔ̀bààm *n*. **agriculture** tɔ̀kɔ̀b a *v*. **ahead** ɛ̀kwɔ̀n *v*.; wàx *n*.; wàxìk<sup>ɛ</sup> *n*.; wàxʉ̀ *n*. **aid** ɨŋaarɛ́s *v*. **aided** ɨŋaarímétòn *v*. **AIDS** lóɓúlukúɲ *n*.; sílím *n*. **aim** ɨɗírɔ́n *v*.; ɨtsírítɛtɛ́s *v*.; iyoes *v*.; tsírítɛtɛ́s *v*.; tɔɓɛɨtɛtɛ́s *v*. **aim for** ɨpɨmɛs *v*.; iyoesa así *v*. **air** ɡwa *n*.; suɡur *n*. **aircraft** iɗékè *n*. **airfield** ɲakwaanʝa *n*. **airplane** iɗékè *n*. **airport** ɲakwaanʝa *n*. **airstrip** ɲakwaanʝa *n*. **ajar** bɔrɔ́ɔ́n *v*.; ŋawíɔ́n *v*. **AK-47** ɲámakaɗá *n*. **alarm** ɲakalo *n*. **albizia (large-leaf)** síɔ̀ɔ̀t <sup>a</sup> *n*. **Albizia anthelmintica** óbìʝòɔ̀z *n*. **Albizia grandibracteata** síɔ̀ɔ̀t <sup>a</sup> *n*. **alcohol craving** mɛ̀sɛ̀ɲɛ̀ƙ <sup>a</sup> *n*. **alcoholic** ɛ́sáàm *n*. **alcoholism** ɛ́s *n*. **alert** ɡonés *v*.; itsópóòn *v*.; ŋízɛ̀s *v*.; ɲakalo *n*. **alien** ʝalánón *v*.; kíʝíkààm *n*. **alight** toɗóón *v*.; zɛƙwɛ́tɔ́n *v*. **alive** ɦyekes *v*. **all** ɓut<sup>u</sup> *ideo*.; boted<sup>o</sup> *n*.; ɗàŋìɗàŋ *quant*.; mùɲ *quant*.; mùɲùmùɲ *quant*.; pílè *ideo*.; pítⁱ *ideo*.; tsíɗ<sup>ɨ</sup> *quant*.; tsíɗɨtsíɗ<sup>ɨ</sup> *quant*.
**all day** ódàtù *n*. **all done** tɛ́zɛ̀tɔ̀n *v*. **all fours (get on)** tíɡàkètòn *v*. **all gone** tɛ́zɛ̀tɔ̀n *v*. **all night** tsoík<sup>o</sup> *n*.; tɛrɛƙɛs *ideo*. **all the time** àrìŋàs *adv*.; ʝíìkⁱ *adv*.; ts'íínáá *n*. **all the way to** ɡònè *prep*.; pákà *prep*. **allergic reaction (skin)** nabêz *n*. **allergy (skin)** nabêz *n*. **Allium species** tâb<sup>a</sup> *n*. **Allophylus species** àɗèŋèlìò *n*. **allot** ɨmɔlɛs *v*. **allow** talakes *v*. **allure** ɨmɔɗɛtɛ́s *v*. **almost** ɦyɔtɔ́ɡɔ̀n *v*. **almost do** bɛ́ɗɛ́s *v*. **aloe** tikorotót<sup>a</sup> *n*. **Aloe species** tikorotót<sup>a</sup> *n*. **alone** ɛɗá *adv*.; kɔ̀nɔ̀n *v*. **along the side** rutet<sup>o</sup> *n*. **along the way** tsìl *ideo*. **alsatian (dog)** ɲeryaŋíŋók<sup>a</sup> *n*. **also** ʝìk<sup>ɛ</sup> *adv*. **altar** lɔpɨtá *n*. **alter** iɓéléés *v*. **alternated** iɗómíòn *v*. **altitude** zikíbàs *n*. **aluminum** ŋkwáŋáy<sup>a</sup> *n*. **aluminum lip plug** ŋkwáŋáy<sup>a</sup> *n*. **always** àrìŋàs *adv*.; ʝíìkⁱ *adv*. **amass** ɨɗaɲɛtɛ́s *v*.; torítéetés *v*. **amass one-by-one** ɨɗɛ́bɛtɛ́s *v*. **ambitious** ɨkázànòòn *v*. **amble** ɨpɛ́ɛ́ɲɛ́sá así *v*.; tasɔ́ɔ́n *v*. **amblyopic** pɨlírímɔ̀n *v*. **ambulatory** ɓɛƙɛsɔs *v*.
**ambush** ɨɗaarɛ́s *v*.; taɗapes *v*.; taɗapetés *v*. **ambusher** tàɗàpààm *n*. **ameliorate** maraŋités *v*.; maraŋítésuƙot<sup>a</sup> *v*. **amenable** tsolólómòn *v*. **America** Amérìkà *n*.; Ɓets'oniicékíʝ<sup>a</sup> *n*. **American** Amérìkààm *n*.; ɓèts'ònìàm *n*. **amniotic fluid** baúcùè *n*. **amuse** fekitetés *v*.; ɨmʉ́mwárés *v*. **an** kɔ́níɛ́n *pro*. **anal sphincter** ɔ́zàhò *n*. **analyze** ŋʉrɛtɛ́s *v*. **anchor** ɲólóit<sup>a</sup> *n*. **ancient** kɔ̀wɔ̀n *v*. **and** ńdà *coordconn*. **and then** náàtì *coordconn*. **and yet (earlier today)** tenák<sup>a</sup> *adv*. **and yet (long ago)** tènòk<sup>o</sup> *adv*. **and yet (yesterday)** tèsìn *adv*. **Andropogon chinensis** ʝan *n*.; ʝɛn *n*. **anemic** dɛrɛ́dɔ̀n *v*. **anemically** dɛ̀r *ideo*. **angel** ŋímaláíkàn *n*. **anger** ɡaánàs *n*.; ɡaanítésuƙot<sup>a</sup> *v*.; ɲɛlɨl *n*. **Anglican** sɛ́mìs *n*. **angry** ɡaanón *v*.; ɨlílíɔ̀n *v*.; iŋóyáánón *v*. **angry (become)** ɡaanónuƙot<sup>a</sup> *v*.; ɨlílíɔnʉƙɔt<sup>a</sup> *v*. **angry at each other** ɨlílíɨnɔ́s *v*. **anguish** ɨtsanítsánɛ́s *v*. **animal (domestic)** ínwá na awá<sup>e</sup> *n*. **animal (wild)** ínwá na riʝááƙɔ̀ <sup>ɛ</sup> *n*. **animal bed** ɗípɔ̀ *n*.; nakús *n*. **animal doctor** ɗakɨtárɨama ínó<sup>e</sup> *n*.
**animal species** tàlàlìdòm *n*. **animal(s)** ínw<sup>a</sup> *n*. **animal-likeness** ínónànès *n*. **animateness** ínónànès *n*. **animism** ɲakuʝíícík<sup>a</sup> *n*. **animist** ɲakuʝíícíkáàm *n*. **ankle** dɛámórók<sup>a</sup> *n*.; kɔpɨkɔp<sup>a</sup> *n*. **anklet (coiled)** ɲɛ́kílɔɗa *n*. **anklet (gold)** ɲámaritóít<sup>a</sup> *n*. **anklet (metal)** dɛɨkatsɨrím *n*. **Ankole cow** Ŋíɲaŋkóléɦyɔ́*n*. **annihilate** ɨƙɔmɛs *v*.; kánɛ́s *v*. **annihilated** kanímétòn *v*. **announce** ɗoɗésúƙot<sup>a</sup> *v*.; síránòn *v*. **announcement (morning)** sír *n*. **announcer** síráàm *n*. **annoy** ɡaanítésuƙot<sup>a</sup> *v*.; ɨtsanɛs *v*. **annoyance** ɡaánàs *n*.; ɲɛlɨl *n*. **annoyed** ɡaanón *v*.; ɨlílíɔ̀n *v*.; iŋóyáánón *v*. **annoyed (become)** ɡaanónuƙot<sup>a</sup> *v*.; ɨlílíɔnʉƙɔt<sup>a</sup> *v*. **annoying** fìfòn *v*.; ɨtsánánòn *v*. **anoint** iɲóɲóés *v*.; kwírɛ́s *v*.; tsáŋés *v*. **anoint the sick** tsáŋésa mayaakóniicé *v*. **anointed** tsáŋós *v*. **another** kɔn *pro*. **another day** kónító ódòwì *n*. **answer** raʝés *v*.; raʝetés *v*.; taatses *v*.; taatsésuƙot<sup>a</sup> *v*.; taatsetés *v*. **answer for** raʝetés *v*. **ant (cocktail acacia)** tàbàrìbàr *n*. **ant (safari)** ƙúduƙûd<sup>a</sup> *n*. **ant (small worker)** sokomet<sup>a</sup> *n*. **ant (soldier)** lókók<sup>a</sup> *n*.; tɛƙɛram *n*. **ant species (black biting)** ɲémúkùɲ *n*. **ant species (black flying)** amózà *n*.; tɨɲíɲ *n*.
arithmetic
**ant species (black)** ɓáritson *n*.; íleɡûɡ<sup>a</sup> *n*.; sɨŋíl *n*. **ant species (blackish flying)** kwɛlɛ́ɗ <sup>a</sup> *n*. **ant species (red flying)** kútúŋùdàd<sup>a</sup> *n*. **ant species (sugar-eating)** ɗɔ́ɡɨɗɔ̂ɡ <sup>a</sup> *n*. **antbear** ɲoɗôd<sup>a</sup> *n*. **anteater** ɲoɗôd<sup>a</sup> *n*. **antelope (male)** sakalʉ́k <sup>a</sup> *n*. **antelope (roan)** ɗorôɡ<sup>a</sup> *n*. **antelope (whistling)** ɲétíli *n*. **antenna (insect)** ɛ̂b <sup>a</sup> *n*. **anthill** kutút<sup>a</sup> *n*. **anthill (holey)** lòkòsòs *n*. **anthill chamber** ɓarán *n*. **antler** ɛ̂b <sup>a</sup> *n*. **ants species (singing)** ʝɔrɔr *n*. **antsy** kakáánón *v*. **anus** kazɨt<sup>a</sup> *n*.; ɔ́zààk<sup>a</sup> *n*. **anvil (stone)** ityakesíɡwàs *n*. **anxious** alólóŋòn *v*. **any** mùɲ *quant*. **apocalyse** tasálétona kíʝá<sup>e</sup> *n*. **appall** ɨɓálɛ́tɔ̀n *v*.; lilétón *v*.; toɓules *v*. **appalling** ɨɓálɔ́n *v*.; toɓúlón *v*. **apparently** íkwà *adv*.; ókò *adv*.; tsábò *adv*. **apparition** kúrúkúr *n*. **appeal to** rɨmɛ́s *v*.; tɔmɨnɛs *v*. **appear** lɛlɛmánétòn *v*.; lɛlɛ́tɔ́n *v*.; pɛlɛ́mɛ́tɔ̀n *v*.; takánétòn *v*. **appear randomly** iɗotíɗótòn *v*. **appearing** lɛlɛmánón *v*. **appearing suddenly** pìrⁱ *ideo*. **appease** ɨkanɛ́s *v*.; ɨkaníkánɛ́s *v*. **appendage (arm-like)** kwɛt<sup>a</sup> *n*. **appendicitis** lɔkapɛt<sup>a</sup> *n*.
**appendix** lɔkapɛt<sup>a</sup> *n*. **appetite (for meat)** bisák<sup>a</sup> *n*. **appetite (have an)** ɨrɔ́rɔ́kánón *v*. **apply heat to** ɨmaɗɛs *v*.; ɨmáɗɛ́sʉƙɔt<sup>a</sup> *v*. **apportion** ɨmɔlɛs *v*. **appreciate** ɨlákásítɛ́sʉƙɔt<sup>a</sup> *v*. **appreciative** ɨlákásɔ́nʉƙɔt<sup>a</sup> *v*. **apprehend** ƙanetés *v*.; zíkɛ́s *v*.; zíkɛ́sʉƙɔt<sup>a</sup> *v*.; zíkɔ́s *v*. **apprehensive** ísánòn *v*. **apprehensive (become)** ísánonuƙot<sup>a</sup> *v*. **approach** ɦyɔtɔ́ɡɛ̀tɔ̀n *v*. **approach (going)** ɦyɔtɔ́ɡɔnʉƙɔt<sup>a</sup> *v*. **approach death** inunúmétòn *v*. **appropriate** itémón *v*. **approximate** ɦyɔtɔ́ɡɛ̀tɔ̀n *v*.; ɦyɔtɔ́ɡɔ̀n *v*. **April** Lomaruk<sup>a</sup> *n*.; Lɔmɔ́y <sup>a</sup> *n*. **Arab** Ŋímaraɓúìàm *n*.; Oríáé *n*. **Arabic language** Múɗùɡùrìtòd<sup>a</sup> *n*.; Oríáénítòd<sup>a</sup> *n*. **arable** tɔkɔbam *n*. **arbitrate** terés *v*. **arc** ilúkúɗòn *v*. **arch-backed** zaɗíɗímòn *v*. **arched** zaɗíɗímòn *v*. **area** ay<sup>a</sup> *n*.; bácík<sup>a</sup> *n*.; ɲɛ́tɛɛr *n*. **area (adjoining)** nabɨɗɨt<sup>a</sup> *n*. **argue** nɛpɛƙánón *v*. **argue (of many)** ilérúmùòn *v*. **argue with** déƙwítetés *v*. **arguer** nɛpɛ́ƙáàm *n*. **argument** dèƙw<sup>a</sup> *n*.; ɲelerum *n*. **argumentative** deƙwideƙos *v*. **argumentative person** nɛpɛ́ƙáàm *n*. **Arisaema ruwenzoricum** ɓóéɗ<sup>a</sup> *n*. **Aristida adoensis** ɲeɗuar *n*. **arithmetic** ɲámára *n*.
**arm** kwɛt<sup>a</sup> *n*. **arm (upper)** cwɛt<sup>a</sup> *n*.; ɲorótónit<sup>a</sup> *n*. **arm bone (upper)** ɲorótónitíɔ́k <sup>a</sup> *n*. **armored vehicle** ɡasoa na ɨlír *n*. **armpit** bàbà *n*. **armpit muscle** ɡuféém *n*. **arms** kúrúɓáa ni cɛmá<sup>ɛ</sup> *n*. **army** kéà *n*. **aromatic** tukukúɲón *v*. **arouse** ɨɓʉ́rɛ́tɔ̀n *v*. **aroused** iɓurímétòn *v*. **aroused sexually** kwídikwidós *v*. **arrange** ɨɗɔ́bɛ̀s *v*.; ɨɗɔ́bɛtɛ́s *v*.; ɨnábɛs *v*.; ɨnábɛ̀sʉ̀ƙɔ̀t a *v*.; ɨnábɛtɛ́s *v*.; itíbès *v*.; itíbesúƙot<sup>a</sup> *v*. **arrange engagement** ɨɗɨmɛ́sá ìʉ̀mà<sup>ɛ</sup> *v*. **arrange marriage** ɨɗɨmɛ́sá buƙúⁱ *v*. **arrest** wasɨtɛs *v*.; wasítɛ́sʉƙɔt<sup>a</sup> *v*.; zíkɛ́s *v*.; zíkɛ́sʉƙɔt<sup>a</sup> *v*. **arrested** zíkɔ́s *v*. **arrive (here)** ɨtɛ́tɔ́n *v*. **arrive in** kʉ̀tɔ̀n *v*. **arrogant** itúrón *v*.; ɨwɔ́ƙɔ́n *v*. **arrogant person** ɨwɔ́ƙɔ́nìàm *n*. **arrow** ɲámal *n*.; ɲɛ́cɨpɨtá *n*. **arrow hole** ɲámalíák<sup>a</sup> *n*. **arrow rings** ɨlíŋírɛ́síàw<sup>a</sup> *n*. **arrow shaft** ɲámalídàkw<sup>a</sup> *n*. **arrowhead** naƙaf *n*. **arrowhead base** ɓólóèd<sup>a</sup> *n*. **artery** tsòrìt<sup>a</sup> *n*. **arthritic** rɔʝɔ́dɔ̀n *v*. **arthritically** rɔ̀ʝ ɔ *ideo*. **articulation (anatomical)** ɲékel *n*. **artifacts** kúrúɓáà nùù kɔ̀w<sup>a</sup> *n*. **artificial** ɨɗɨmɔtɔ́sá ròɓ<sup>o</sup> *v*.
**artillery gun** komótsɛ́ɛ̀b <sup>a</sup> *n*. **artist** ɨɗɨmɛ́síàm *n*. **as** ɗítá *prep*. **as a whole** boted<sup>o</sup> *n*. **as well** ʝìk<sup>ɛ</sup> *adv*. **ascend** tóbìrìbìròn *v*.; totírón *v*. **Asclepiadaceae species** lócén *n*.; xoúxoú *n*. **ash(es)** káw<sup>a</sup> *n*. **ashamed** kweelémòn *v*. **ashamed (become)** ɨɛ́ɓɛ́tɔ̀n *v*.; iryámétona ŋiléétsìk<sup>e</sup> *v*. **ask** esetés *v*.; esetésúƙot<sup>a</sup> *v*.; esetetés *v*. **ask for** tɔɓɛ́ɲɛ́tɔ̀n *v*.; wáánɛtɛ́s *v*. **asleep (limbs)** isálílòn *v*. **Asparagus flagellaris** kòkòròts<sup>a</sup> *n*. **asphyxiate** tʉɓʉnɛ́s *v*.; tuɓunímétòn *v*. **ass** ɔ̂z *n*. **ass (donkey)** ɗìɗ<sup>a</sup> *n*. **assault** iríɓéés *v*. **assault sexually** itikiesúƙot<sup>a</sup> *v*. **assay** ɨkatɛs *v*. **assemble** ɨrírɛ́ɛ́sʉƙɔt<sup>a</sup> *v*.; ɨrírɛ́ɛtɛ́sá así *v*.; iryámíryámètòn *v*.; itóyéés *v*.; itóyéésa así *v*.; ɨtsʉ́nɛ́tɔ̀n *v*.; ɨʉɗɛs *v*.; ɨʉɗɛtɛ́s *v*. **assembly** kur *n*.; ɲatʉ́kɔ́t <sup>a</sup> *n*.; ɲékíìkò *n*.; ɲémítìŋ *n*. **assess** esetés *v*.; iniŋes *v*. **assist** ɨŋaarɛ́s *v*. **assistant** ɨŋaarɛ́síàm *n*.; karan *n*. **assisted** ɨŋaarímétòn *v*. **assort** ɨsíílɛ́s *v*. **assortment** ɲalíɲalí *n*. **astonish** ɨɓálɛ́tɔ̀n *v*.; lilétón *v*. **astonishing** ɨɓálɔ́n *v*.; toɓúlón *v*. **astringent** ƙɛ́rɨƙɛ́rɔ́n *v*.; tɛrɛrɛ́ɔ́n *v*. **at dawn** ɲaɓáít<sup>ɔ</sup> *n*. **at daytime** ódò<sup>o</sup> *n*.
#### at dusk
**at dusk** xɨŋat<sup>ɔ</sup> *n*. **at night** mukú *n*. **at once** ikéé kɔ̀n *n*.; ɲásáàtɔ̀ kɔ̀n *n*. **at one time** kédìè kɔ̀n *n*.; kédò kɔ̀n *n*. **at the same time** ɲásáàtɔ̀ kɔ̀n *n*. **Ateso language** Ŋítésótôd<sup>a</sup> *n*. **atheist** nɛpɛ́ƙáàm *n*. **atmosphere** didiɡwarí *n*. **atom** kiɗoɗots<sup>a</sup> *n*. **atone for** iɓutes *v*. **attach** ɗɔtsɛtɛ́s *v*.; ramɛtɛ́s *v*. **attached** ɗɔtsɔ́s *v*. **attack** iríɓéés *v*.; toɓés *v*.; toɓésúƙot<sup>a</sup> *v*. **attacker** toɓésíàm *n*. **attain** enés *v*. **attempt** ɨkatɛs *v*. **attempt repeatedly** ɨkatíkátɛ́s *v*. **attend to** ewanes *v*.; ewanetés *v*. **attend to (garden)** kɔɛ́s *v*. **attendant (garden)** kɔɛsíàm *n*. **attention span (have a short)** dʉmɛ́ɗɛ́ mɔ̀n *v*. **attentive** itsópóòn *v*. **attest to** itsáɗénés *v*. **attire (fine)** ɲɛ́nɨs *n*. **attract** ɨtʉ́lɛ́ɛtɛ́s *v*.; tɔmɨnɛs *v*. **attractive** dòòn *v*. **attractive person** ɨɓʉrɛtɛ́síàm *n*. **attribute (personal)** ɲɛpɨtɛa ámá<sup>e</sup> *n*. **auction** ókísèn *n*. **August** Ɗiwamúce *n*.; Iɗátáŋɛ́r *n*.; Lósʉ́ɓán *n*. **aunt (his/her father's brother's wife)** ŋwáát<sup>a</sup> *n*. **aunt (his/her father's sister)** tatat<sup>a</sup> *n*.
**aunt (his/her mother's brother's wife)** momotícék<sup>a</sup> *n*. **aunt (his/her mother's sister)** totot<sup>a</sup> *n*. **aunt (mother's brother's wife)** momócèk<sup>a</sup> *n*. **aunt (mother's sister)** totó *n*. **aunt (my father's brother's wife)** yáŋ *n*. **aunt (my father's sister)** tátà *n*. **aunt (your father's brother's wife)** ŋɔ́*n*. **aunt (your father's sister)** tátó *n*. **authority** zeís *n*. **authority (have)** topéɗésuƙot<sup>a</sup> *v*. **authority (person)** zeísíàm *n*. **automobile** ɲómotoká *n*. **avail** bírɛ́s *v*. **available** tɔɔsɛ́tɔ̀n *v*. **avenge** ɲaŋés *v*.; ɲaŋésúƙot<sup>a</sup> *v*. **average** ŋwanɨŋwánɔ́n *v*. **avert eyes** kurukúrón *v*. **avocado** ɲóvakáɗò *n*. **avoid eye contact** kurukúrón *v*. **avoid repeatedly** iwitíwítòn *v*. **avoidant** firifíránón *v*.; wíríwíránón *v*. **avow** ikóŋón *v*. **await** kɔɛ́s *v*.; kɔɛtɛ́s *v*. **awake** ɡonés *v*. **awake for sex** ɡòkòn *v*. **awaken** ɡonésétòn *v*. **award** tɔ́rɔ́bɛs *v*.; tɔ́rɔ́bɛsa na ílɔɛsí *n*. **awkward** betsínón *v*.; ɨɓaŋíɓáŋɔ̀n *v*.; pɔsɔ́kɔ́mɔ̀n *v*. **awl** ɓɔtsɔ́t <sup>a</sup> *n*. **axe** dzibér *n*. **axe (modern)** nanɨŋɨnɨŋ *n*.; naŋɨnɨŋɨn *n*. **axe (traditional)** kuɲukúdzibér *n*. **axe-blessing ceremony** dzíbèrìkàmɛ̀s *n*. **axehead (modern)** nanɨŋɨnɨŋ *n*.; naŋɨnɨŋɨn *n*.
## banditry
#### baa!
**baa!** bèrrr *ideo*.; mɛ́ɛ̀ɛ̀ *ideo*.; rɛ̀rrr *ideo*. **babble** íbìrìbìròn *v*.; imátôd<sup>a</sup> *n*. **baboon** tsɔ́r *n*. **baboon (alpha male)** òlìòt<sup>a</sup> *n*.; tìmùòz *n*. **baboon (female)** tsɔ́ráŋwa *n*. **baboon (lone male)** òrèɡèm *n*. **baboon troop** kwaár *n*. **baby** ɗiak<sup>a</sup> *n*.; im *n*. **baby carrier** ɗoɗôb<sup>a</sup> *n*. **baby hair** imásíts'<sup>a</sup> *n*. **baby primate** kíɗɔlɛ́*n*. **baby sling** ɗoɗôb<sup>a</sup> *n*. **baby talk** imátôd<sup>a</sup> *n*. **baby wipe** ŋííɗɛ́sìƙwàz *n*. **back** ʝìr *n*.; ʝìrìk<sup>ɛ</sup> *n*.; kanɛd<sup>a</sup> *n*.; ɔ́zɛ̀d <sup>a</sup> *n*. **back (lower)** ɲɛ́kɨpɛtɛ́t <sup>a</sup> *n*. **back (upper)** kan *n*. **back of hand** kwɛtákán *n*. **back of leg** búbuiem *n*. **back out** ɨsʉ́rʉ́mɔ̀n *v*.; raʝánón *v*. **back part** ʝírɛ̂d <sup>a</sup> *n*. **back side** namɛ́ɗɔ́ɛ̀d <sup>a</sup> *n*. **back then** ódowicíká kì nùù kì *n*.; ódowicíkó nùk<sup>u</sup> *n*. **back up** ikutúkútés *v*.; ikutúkútòn *v*. **backbone** ɡòɡòròʝòɔ̀k <sup>a</sup> *n*. **backside** ɔ́zɛ̀d <sup>a</sup> *n*.; ɔ̂z *n*. **backward** ʝìrìk<sup>ɛ</sup> *n*. **backyard** bɔlɔl *n*. **bad** ɡaanón *v*. **bad (make)** ɡaanítésuƙot<sup>a</sup> *v*. **bad (of many)** ɡaanaakón *v*. **bad eye (have a)** ɗooɲómòn *v*. **badger** ilúlúés *v*. **badly** ɡàànìk<sup>e</sup> *v*.
**badly off** tawanímétòn *v*. **badness** ɡaánàs *n*. **baffled** iɓíléròn *v*.; ɨcɔ́ŋáimetona iká<sup>e</sup> *v*. **bag** ɲɛ́ɓɛ́k <sup>a</sup> *n*.; ofur *n*. **bag (burlap)** ɲéɡuniyá *n*. **bag (cloth)** ɲáwáróófúr *n*. **bag (goat-leather)** riéófúr *n*. **bag (leather)** èw<sup>a</sup> *n*. **bag (plastic)** ɲápaalí *n*. **baggage** botitín *n*. **bagworm** mɔɗɔ́ɗ <sup>a</sup> *n*. **bait** ɨmɔɗɛtɛ́s *v*. **bait (bees)** sɨsɨɓɛs *v*. **bake** ʝʉɛ́s *v*. **baked** ʝʉɔ́s *v*. **balance** iríánitetés *v*.; ɨtátɛ́ɛ́s *v*. **Balanites aegyptiaca** tsʉm *n*. **Balanites pedicellaris** ɓòŋ *n*. **bald** ŋoléánètòn *v*. **bald on top** palórómòn *v*. **balefire** ɲáɡaaɗi *n*. **balk** kwɛ́rɛɗɛ́ɗɔ́n *v*.; wasɛ́tɔ́n *v*. **ball** ɲɛ́pɨɨrá *n*. **ball field** ɲakwaanʝa *n*. **ball up** imúnúkukúón *v*. **ball-shaped** ɨlʉ́lʉ́ŋɔ́s *v*. **ball-shaped (make)** ɨlʉ́lʉ́ŋɛ́s *v*. **ballot** kàbàɗ<sup>a</sup> *n*. **bamboo (mountain)** ɲéɡiróy<sup>a</sup> *n*. **banana** ɲómototó *n*. **bandage** ɨmakɛs *v*. **bandit** lotáɗá *n*. **bandit (bush)** ríʝíkààm *n*. **banditry** lotáɗánànès *n*.; ŋirúkóìnànès *n*.
**bang into** íbaɗɛ́s *v*. **bang into repeatedly** íbaɗiés *v*.; tanaŋínáŋesuƙot<sup>a</sup> *v*. **bang on** ɨɗatɛs *v*. **bang!** pààɗòk<sup>o</sup> *ideo*. **bank** ɓɔ́kɔ̀ɲ *n*.; kaɓéléɓelánón *v*.; ɲáɓáŋk<sup>a</sup> *n*. **bank check** kaúdzokabáɗ<sup>a</sup> *n*. **banquet** írésiŋƙáƙ<sup>a</sup> *n*. **bantam** puusúmòn *v*. **bao (game)** ɲékilelés *n*. **baptism** ɓatísimú *n*. **baptismal certificate** ɓatísimúkabáɗ<sup>a</sup> *n*. **baptismal name** ɓatísimúêd<sup>a</sup> *n*. **baptize** iɓátíseés *v*. **bar** ɲáɓá *n*.; teɡeles *v*.; toƙólésuƙot<sup>a</sup> *v*. **bar (wooden)** naƙólít<sup>a</sup> *n*. **barb** omén *n*. **barbeque** ɨtɔlɛs *v*. **barbeque spot** nakíríkɛ̀t <sup>a</sup> *n*. **barbequed** ɨtɔlɔs *v*. **barbet (ground)** loyeté *n*. **bare** ilérón *v*.; lemúánètòn *v*.; leŋúrúmòn *v*.; sɨlɔ́ʝɔ́mɔ̀n *v*.; tuɗúsúmòn *v*. **bare (of a patch)** patsólómòn *v*. **bare (of a tree)** sáɡwàràmòn *v*. **bare foot** ɡɔ̀rìɡɔ̀r *n*. **bare teeth** kwíɲíkɔ̀ɔ̀n *v*. **barefoot** ɡɔ́ríɡɔ̀rìk<sup>ɔ</sup> *n*. **barf** ɦyɛnɛ́tɔ́n *v*.; ɦyɛ̀nɔ̀n *v*. **bark** bɔɗɔ́k <sup>a</sup> *n*.; íɡòmòn *v*. **bark at** doƙofiés *v*. **bark soap** ʝaobɔɗɔ́k <sup>a</sup> *n*. **Barleria acanthoides** ɲólíkàf *n*. **barm** sîb<sup>a</sup> *n*.
**barrack** ɲáɓáràkìs *n*.; ɲeryaŋíɓór *n*. **barrage with words** ídʉrɛ́sá tódà<sup>e</sup> *v*. **barrel (gun)** morókêd<sup>a</sup> *n*. **barrel (large)** ɲépípa *n*. **barrel (plastic)** ɲékakúŋ́ɡù *n*. **barrel-shaped** ɗatáɲámòn *v*. **barren** ikólípánón *v*.; osorosánón *v*. **barren (animal or person)** ɲokólíp<sup>a</sup> *n*. **barren person** òsòròs *n*. **barricade** teɡeles *v*.; toƙólésuƙot<sup>a</sup> *v*. **barrier (thorny)** ɲéríwi *n*. **base** dɛ *n*.; dɛɛd<sup>a</sup> *n*.; dziŋ *n*. **base of a boulder** taɓádɛ̀ *n*. **base of a fence** marɨŋídɛ̀ *n*. **base of a mountain** kwarádɛ̀ *n*. **base of beehive tree** kànàxàdɛ̀ *n*. **base of ridge** tsɨɨr *n*. **base of sacred tree** lɔ́ƙɔ́ŋʉ̀dɛ̀ *n*. **Basella alba** lòɓòlìà *n*. **bash** inipes *v*. **basin** ɲáɓáf *n*.; ɲéɓésèn *n*. **basin (gourd)** sèrèy<sup>a</sup> *n*. **basket (winnowing)** zít<sup>a</sup> *n*. **bastard** ŋabɔ́bòìm *n*. **bat** lɔ́pɛ́ɗɛpɛ́ɗ <sup>a</sup> *n*. **bat species** watéɡwà *n*. **bateleur** oromén *n*. **bath** féíàw<sup>a</sup> *n*. **bathe** féítetés *v*.; féón *v*. **bathing** féy<sup>a</sup> *n*. **bathing stone** ɡwasa na féí *n*. **bathroom** féíàw<sup>a</sup> *n*. **bathtub** itúɓ<sup>a</sup> *n*. **batis (chin-spot)** iwótsíɡwà *n*.; nàŋʉ̀ràmɓɔ̀ *n*. **battalion** ɲéɓatál *n*. **batter** ipukúpúkés *v*.
**battery** ɡúr *n*.; ɡwas *n*.; ɲéɓeterí *n*. **bawl** xérón *v*. **be (make)** mɨtɨtɛs *v*. **be (not)** beníón *v*.; bɛnɔ́ɔ́n *v*. **be (some size)** ítón *v*. **be (somehow)** ìròn *v*. **be (someone or something)** mìtɔ̀n *v*. **be (somewhere)** ìòn *v*. **be about** tábès *v*. **be alone** ìònà ɗòk<sup>u</sup> *v*.; iona ɛɗá *v*. **be doing** cɛ̀mɔ̀n *v*. **be en route** iona muceék<sup>e</sup> *v*. **be in trouble** iona ŋítsaník<sup>ɛ</sup> *v*. **be neighbors** narúétinós *v*. **be not yet** sárón *v*. **be on the way** iona muceék<sup>e</sup> *v*. **be solitary** ìònà ɗòk<sup>u</sup> *v*.; iona ɛɗá *v*. **be the only** ìònà ɗòk<sup>u</sup> *v*. **be with** iona ńdà *v*. **bead** ɨɗɛrɛs *v*. **bead (big white)** lokoit<sup>a</sup> *n*. **beaded** ɨɗɛrɔs *v*. **beaded vest** ɲáɓol *n*. **beads** ŋabit<sup>a</sup> *n*. **beads (strung)** rɔam *n*. **beadwork** ŋabit<sup>a</sup> *n*. **beak** aka ɡwaá<sup>e</sup> *n*.; ɡwáák<sup>a</sup> *n*. **beam of light** bás *n*.; sʉ́w<sup>a</sup> *n*. **bean (red)** màràŋɡwà *n*. **bean variety** marɨŋímóríɗ<sup>a</sup> *n*. **bean(s)** mòrìɗ<sup>a</sup> *n*. **bear** nɛɛ́s *v*.; nɛɛsʉ́ƙɔt<sup>a</sup> *v*.; taɗaŋes *v*. **bear a child** ƙwaatetés *v*. **bear down** toƙíróòn *v*. **bear legs first** ɨʝʉlɛtɛ́s *v*.
**bear prematurely** ɨsɔɛtɛ́s *v*. **bear twins** ɨmʉ́ítɛtɛ́s *v*. **bear witness to** itsáɗénés *v*. **beard** ɲɛ́pɛ́nɛk<sup>a</sup> *n*.; tɛ̀mʉ̀r *n*. **beast (mythical)** ɲaŋu *n*. **beast(s)** ínw<sup>a</sup> *n*. **beat** ɨnɔmɛs *v*. **beat (defeat)** kurés *v*.; kurésúƙot<sup>a</sup> *v*. **beat (heart)** ƙádiƙádòn *v*. **beat (outdo)** ɨlɔɛs *v*.; ɨlɔɛtɛ́s *v*. **beat (pulsate)** dìkwòn *v*. **beat (rhythm)** ɨrɛɛsa dikwá<sup>e</sup> *v*. **beat (tired)** ɨlɔ́ɛ́tɔ̀n *v*.; ɨlɔ́yɔ́n *v*.; ziálámòn *v*.; zíkímétòn *v*.; ziláámòn *v*. **beat (win)** ɨrɛɛs *v*.; kɔrɨtɛtɛ́s *v*. **beat back** ɨɗáfɛ́sʉƙɔt<sup>a</sup> *v*.; ɨɗáfɛ́sʉƙɔta así *v*. **beat down** íbutsés *v*.; ɨɗáfɛ́sʉƙɔt<sup>a</sup> *v*.; ɨlɔ́ítɛ́sʉƙɔt<sup>a</sup> *v*. **beat each other** tábunós *v*. **beat off** ɨpʉtɛs *v*. **beat out** íbutsés *v*. **beautiful** dòòn *v*. **beautify** daites *v*.; íɡwɨɡwɨʝɛ́s *v*.; naƙwídɛtɛ́s *v*. **beautify oneself** daitetésá así *v*. **beauty** daás *n*. **because** ikóteré *subordconn*.; kóteré *subordconn*. **because (of)** ɗúó *pro*. **because of** ikóteré *prep*.; kóteré *prep*. **Becium species** ɔ́ʝítínícɛmɛ́r *n*. **become (some size)** ítónuƙot<sup>a</sup> *v*. **become (somehow)** ironuƙot<sup>a</sup> *v*. **become (someone/something)** mɨtɔnʉƙɔt<sup>a</sup> *v*. **become like** ƙámétòn *v*.; ƙámónuƙot<sup>a</sup> *v*. **bed** ɲɛ́kítaɗa *n*.
**bed (animal)** ɗípɔ̀ *n*.; nakús *n*. **bedbugs** ŋítʉ́mìk<sup>a</sup> *n*. **beddings** kúrúɓáa ni epwí *n*. **bedlam** ɲɔ́ŋɔtsán *n*. **bee** ts'ɨƙ<sup>a</sup> *n*. **bee (carpenter)** dʉrʉdʉr *n*. **bee (ground)** mɔ́ɗ <sup>a</sup> *n*. **bee (stingless)** lɔwɨɲ *n*. **bee (sweat)** lɔwɨɲ *n*. **bee (tree)** mʉ́ƙás *n*. **bee (worker)** naaseɲaŋ *n*. **bee eater** keseníɡwà *n*. **bee larva** sîd<sup>a</sup> *n*. **bee larva chamber** sídàhò *n*. **bee queen** lókílóróŋ *n*.; okílóŋór *n*. **bee scout** páupáw<sup>a</sup> *n*. **bee summoning** wówóʝ<sup>o</sup> *ideo*. **bee swarm (mobile)** ts'ɨƙábòt<sup>a</sup> *n*. **bee-eater** cooríɡwà *n*. **beehive** kànàxà *n*. **beehive (ground bee)** kùkùsèn *n*. **beehive (in rock)** wàts'w<sup>a</sup> *n*. **beehive (in tree)** hàb<sup>a</sup> *n*. **beehive cover** makúl *n*. **beehive hut** Icéhò *n*. **beehive shell** ɗòl *n*. **beer** mɛ̀s *n*. **beer (bottled)** ɲéɓía *n*. **beer (bottom layer)** ɔ́zɛ̀dà mɛ̀sɛ̀ *n*. **beer (breakfast)** lɔkátɔ́rɔ̀t <sup>a</sup> *n*. **beer (for birth ceremony)** mɛsa ƙwaaté *n*. **beer (for naming ceremony)** mɛsa édì *n*. **beer (for sale)** ɲɛ́kɨráɓ<sup>a</sup> *n*. **beer (honey)** sɨs *n*.; ts'ɔƙam *n*.
**beer (leftover)** cueina mɛ́sɛ̀ *n*.; mɛsɛcue *n*. **beer (millet)** ŋamarʉwáy<sup>a</sup> *n*.; rébèmɛ̀s *n*. **beer (stale)** ɗìɗèkwàts<sup>a</sup> *n*. **beer (wedding)** ɲalakʉts<sup>a</sup> *n*. **beer barm** cúrúkà mɛ̀sɛ̀ *n*. **beer dregs** ɗʉká *n*.; dàʝ<sup>a</sup> *n*. **beer dregs (in a pot)** dómóɔ̀zà mɛ̀sɛ̀ *n*. **beer head** ikeda mɛ́sɛ̀ *n*. **beer hunger** mɛ̀sɛ̀ɲɛ̀ƙ <sup>a</sup> *n*. **beer porridge** rùt<sup>a</sup> *n*. **beer pot** mɛ̀sɛ̀dòm *n*. **beer yeast** cúrúkà mɛ̀sɛ̀ *n*. **beeswax (black)** sɔs *n*. **beeswax (chewed)** sàsàr *n*. **beetle (brown jewel)** dùràts<sup>a</sup> *n*. **beetle (bruchid)** dʉmʉ́ná mòrìɗò<sup>e</sup> *n*. **beetle (burrowing ground)** lɔ́ŋízɨŋîz *n*. **beetle (dung)** dʉmʉ́n *n*. **beetle (water)** dʉɗɛ́r *n*. **beetle larva (tiger)** sikusába *n*. **before** ɗàmʉ̀s *subordconn*.; ɗɛ̀mʉ̀s *subordconn*.; kwààk<sup>e</sup> *n*.; wàx *n*.; wàxʉ̀ *n*. **before dawn** ɲaɓáít<sup>ɔ</sup> *n*. **beg** itsenes *v*.; wáán *v*. **beg from each other** wáánínɔ́s *v*. **beg relentlessly** itseniés *v*. **beg relentlessly (begin to)** itsenietés *v*. **beggar** wáánààm *n*. **beggar (persistent)** tikorotót<sup>a</sup> *n*. **begging** wáán *n*. **begin** iséétòn *v*.; isóón *v*.; itsyákétòn *v*.; toɗóón *v*. **beginning point** itsyákétònìàw<sup>a</sup> *n*. **behave badly** tarates *v*.; taratiés *v*. **behavior** ɲɛpɨtɛ *n*. **behind** ʝìr *n*.; ʝìrìk<sup>ɛ</sup> *n*.; ʝìrʉ̀ *n*.; kanɛd<sup>a</sup> *n*.; kànɛ̀dɛ̀k <sup>ɛ</sup> *n*.; kanɨtínʉ́ *n*.
**behind bars (jailed)** zíkɔ́s *v*. **being** ɦyekes *n*. **belabor** ɨɗɔtsɛs *v*. **belch** xèr *n*.; xerétón *v*.; xérón *v*. **belief** ɲanʉ́pít<sup>a</sup> *n*. **believable** tɔnʉpam *n*. **believe** tɔnʉpɛs *v*. **believer** tɔnʉpɛsíám *n*. **bellicose** cɛmɨcɛmɔs *v*. **bellow** ábʉ̀bʉ̀ƙɔ̀n *v*.; béúrètòn *v*.; ɨkílɔ́n *v*.; xérón *v*. **belly** bùbù *n*.; ɡwàʝ<sup>a</sup> *n*. **belly (of a pot)** bakútsêd<sup>a</sup> *n*. **belly button** ƙɔ̀ɓ <sup>a</sup> *n*. **bellyache** áts'ɛ́sà bùbùì *n*. **belongings** kúrúɓâd<sup>a</sup> *n*.; kúrúɓáicík<sup>a</sup> *n*. **below** kɔ́ɔ́kíʝ<sup>o</sup> *n*.; nɔ́ɔ́kíʝ<sup>o</sup> *n*. **belowground** ɗis *n*. **belt** ɲámisíp<sup>a</sup> *n*.; ɲémisíp<sup>a</sup> *n*. **belt (beaded)** ɲátóè *n*. **bemoan** ƙɔ̀ɗɔ̀n *v*.; topóɗón *v*. **bench** ɲɛ́ɓɛ́ɲc *n*.; ɲófóm *n*. **bend** itúkúɗètòn *v*.; itúkúɗòn *v*.; nɔƙɨnɔ́ƙɔ́n *v*.; tɔɓɨlɛs *v*.; tukuɗes *v*. **bend over** dʉ́ɡʉ̀mɛ̀tɔ̀n *v*.; rétítésúƙot<sup>a</sup> *v*. **bend sideways** kwɛ́dɔ̀n *v*. **bendily** lɛ̀ts'<sup>ɛ</sup> *ideo*.; nàƙw<sup>a</sup> *ideo*.; nà<sup>u</sup> *ideo*.; nɔ̀ƙ ɔ *ideo*. **bendy** lɛts'ɛ́dɔ̀n *v*.; naƙwádòn *v*.; naúdòn *v*.; nɔƙɔ́dɔ̀n *v*. **benefit** ɨkɛ́ítɛtɛ́s *v*. **bent** dɔ́ɡɔ̀lɔ̀mɔ̀n *v*.; tukúɗón *v*. **bent back** tɛ́zɛ̀ɗɔ̀n *v*. **bent forward** koɗósón *v*. **bent over** dʉ́ɡʉ̀mɔ̀n *v*.; ƙúzùmòn *v*.; rétón *v*.; tɔɓɨlɛsa así *v*.
**berate** dɔxɛ́s *v*.; dɔxɛ́sʉ́ƙɔt<sup>a</sup> *v*. **Berchemia discolor** ɗewen *n*. **beseech** iƙenes *v*. **beside** ɨɓákɔ́n *v*.; ŋabér<sup>o</sup> *n*. **beside (move)** ɨɓákɔ́nʉƙɔt<sup>a</sup> *v*. **besmear** iɲóɲóés *v*. **bestialist** tirésíama ínó<sup>e</sup> *n*. **betray** tolúétòn *v*.; tolúónuƙot<sup>a</sup> *v*. **betray each other** tolúúnós *v*. **betrayer** ŋítúrumúám *n*.; tolúónìàm *n*. **better** maraŋités *v*.; maraŋítésuƙot<sup>a</sup> *v*. **better (get)** doonuƙot<sup>a</sup> *v*.; iŋáléètòn *v*.; maráŋónuƙot<sup>a</sup> *v*.; toíónuƙot<sup>a</sup> *v*.; xɔ́dɔnʉƙɔt<sup>a</sup> *v*. **better (make a bit)** ɓarɨɓárɔ́nʉƙɔt<sup>a</sup> *v*. **better slightly** ŋwaníŋwánɨtɛ́s *v*. **between** sɨsɨkák<sup>ɛ</sup> *n*. **beverage** wetam *n*. **bewail** topóɗón *v*. **beware** tɔtsɔ́ɔ́n *v*. **bewildered** tɔmɛrímɛ́rɔ̀n *v*. **bewitch** ipeɗes *v*.; sʉɓɛ́s *v*. **bewitch by pointing** ɗóɗiés *v*. **bewitch by the evil eye** ɨɗɛmɛs *v*. **bewitcher** ìpèɗààm *n*.; sʉɓɛ́síàm *n*. **bhang** ɲábaŋɡí *n*. **Bible** Ɲáɓáíɓɔ̀l *n*. **bicep** cwɛtéém *n*. **bicep/tricep** ɲorótónitíém *n*. **bicker** iƙúmúnós *v*.; ɨɲʉ́ɲʉ́rɔ̀n *v*.; ŋʉ́zʉmánón *v*. **bicycle** ɲamɨɨlɨ *n*. **biddy** ɡwaŋwa *n*. **biddy (chick)** ɲɔ́kɔkɔróím *n*. **Bidens pilosa** tʉ̀fɛ̀rɛ̀ƙ <sup>a</sup> *n*. **big** zòòn *v*.
**big (become)** ítónà dìdìk<sup>e</sup> *v*. **big (of many)** zeikaakón *v*. **big day** ódowa ná zè *n*. **big man** ámáze *n*.; ámázeám *n*. **big people** roɓazeik<sup>a</sup> *n*. **big person** ámáze *n*. **bigger (become)** iwánétòn *v*.; zoonuƙot<sup>a</sup> *v*. **bigger (make a bit)** ɓarɨɓárítɛ́sʉƙɔt<sup>a</sup> *v*. **bigger (make)** iwanetés *v*.; zeites *v*.; zeitésuƙot<sup>a</sup> *v*. **bigness** zeís *n*. **bike** ɲamɨɨlɨ *n*. **bile** bìɗ<sup>a</sup> *n*. **bill** ɡwáák<sup>a</sup> *n*.; kaúdzokabáɗ<sup>a</sup> *n*. **bill (monetary)** ɲónót<sup>a</sup> *n*. **billfold** ɲɔ́pɔ́c *n*. **billion** ɗòlà kòn *n*.; ɲéɓílìòn *n*. **billow** ipúróòn *v*. **billow up** ipúréètòn *v*. **biltong** ŋátɔɔsa *n*. **bind** ɨmakɛs *v*.; ɨrɨɗɛtɛ́s *v*.; zíkɛ́s *v*. **bind the stomach** ɨrɨɗɛsá bùbùà<sup>e</sup> *v*. **bind up** zíkɛ́sʉƙɔt<sup>a</sup> *v*.; zɨkɛtɛ́s *v*. **bine** ɨrɨɗɛs *v*. **binge** iwótsóòn *v*. **binoculars** ɲáɓaraɓín *n*. **bird** ɡwa *n*. **bird (domestic)** ɡwaa na awá<sup>e</sup> *n*. **bird (female)** ɡwaŋwa *n*. **bird (male)** ɡwácúrúk<sup>a</sup> *n*. **bird (nocturnal)** mukúáɡwà *n*. **bird species** àdɛ̀nɛ̀s *n*.; bíro *n*.; cicídè *n*.; ɗɨɨt<sup>a</sup> *n*.; dàbìʝ<sup>a</sup> *n*.; dòɗìdòɗìɡwà *n*.; fetíɡwà *n*.; fírits'ár *n*.; itsól *n*.; ʝílífɨfí *n*.; kìlòlòɓ<sup>a</sup> *n*.; kiyér *n*.; kólór *n*.;
lorokoníɡwà *n*.; lotiwúót<sup>a</sup> *n*.; napóɗè *n*.; ɲíkwaamwíyá *n*.; ɲɔ́kɔɗɔɔŋɔ́r *n*.; ɔ̀fàɡwà *n*.; óƙírot<sup>a</sup> *n*.; ɔríyɔ́ *n*.; sʉ́ƙʉ́sʉƙáɡwa *n*.; tsóɗitsóno *n*. **bird tail** tsʉ́ɓ <sup>a</sup> *n*. **bird-scarer** kɔɛsíàm *n*. **birdhouse** ɡwáho *n*. **birth** ƙwaát<sup>a</sup> *n*. **birth relation** ɦyeínósá na ƙɔ́ɓà<sup>ɛ</sup> *v*. **birthing complications** ƙwaata ná ɡààn *n*. **biscuit (sweet)** ɲéɓisikót<sup>a</sup> *n*. **bisect** ɨwáwáɗɛ́s *v*.; pakés *v*. **bishop** ɲépiskóópì *n*. **bishop bird (black-winged red)** tsól *n*. **bishop bird (red/yellow)** loɓúrútùt<sup>a</sup> *n*. **bitch** ŋókíŋwa *n*. **bite** áts'ɛ́s *v*.; ƙídzɛ̀s *v*. **bite each other** ƙídzɨnɔ́s *v*. **bite hard** áts'ɛ́sʉƙɔt<sup>a</sup> *v*. **bite repeatedly** ƙídzatiés *v*. **bitter** fàɗòn *v*. **bitter (but edible)** taɗaɗáŋón *v*. **bitter (make)** faɗites *v*. **bitter (slightly)** ɓakɨɓákɔ́n *v*. **bitter thing** kèrèmìdz<sup>a</sup> *n*. **bivouac** napérít<sup>a</sup> *n*.; ɲákamɓí *n*. **blabber** ilemílémòn *v*. **blabbermouth** ɡáʒadɨŋwa *n*. **black** buɗámón *v*. **black (of many)** buɗamaakón *v*. **black (very)** cuc<sup>u</sup> *ideo*.; múɗèr *ideo*.; tíkⁱ *ideo*. **Black Jack** tʉ̀fɛ̀rɛ̀ƙ <sup>a</sup> *n*. **black person** buɗámónìàm *n*. **black with white cheeks** kɨɗapánɛ́tɔ̀n *v*. **black with white rump** kedíánètòn *v*. **black-and-white** kábìlànètòn *v*.
**blackboard** ɲáɓáo *n*. **blacken** kɔrɛtɛ́s *v*.; kɔrɨtɛtɛ́s *v*.; tɛwɛrɛs *v*. **blackened** kɔrɛ́tɔ́n *v*. **blackmail** ɨŋaalɛ́s *v*. **blacksmith** ìtyàkààm *n*. **bladder** xàr *n*. **bladder area** heʝú *n*. **blade** ɗàw<sup>a</sup> *n*. **blade (fan)** suɡuráɗáw<sup>a</sup> *n*. **blade (propeller)** suɡuráɗáw<sup>a</sup> *n*. **blame each other falsely** tapáínɔ́s *v*. **blame falsely** tapɛɛs *v*. **bland** cucuéón *v*.; ɗɛ̀ƙwɔ̀n *v*.; ʝɔ̀lɔ̀n *v*.; muʝálámòn *v*. **blanket** ɡubés *v*.; ɲéɓurankít<sup>a</sup> *n*. **blanket (cotton)** ɲákamariɗúk<sup>a</sup> *n*. **blanket (light)** tírìƙà *n*. **blanket (Maasai style)** lorwaneta *n*. **blanket over** ɡubésúƙot<sup>a</sup> *v*. **blaspheme** ébetiés *v*. **blather** íbìrìbìròn *v*. **blaze a trail** ɨpʉ́tɛ́sʉƙɔta muceé *v*. **bleb** ɔ́ʝáìm *n*. **bleed** tɔyɔ́ɔ́n *v*. **bleed for drinking** tsɔrɛ́s *v*. **bleed for healing** kɔ́ɛ́s *v*. **bleed from nose** ƙòlòn *v*. **blend** iɗyates *v*.; iɲales *v*.; ɨtsɔɓítsɔ́ɓɛ́s *v*.; itsulútsúlés *v*. **blend (grains)** ikáɗóés *v*. **blended** ɨtsɔɓítsɔ́ɓɔ̀n *v*.; ɨtsɔɓítsɔ́ɓɔ́s *v*. **bless** ɨmwaímwɛ́ɛ́s *v*.; tatiés *v*. **blessed** tatiós *v*. **blesser** tatiesíám *n*. **blind** múɗúkánón *v*.
**blink** íbɛ̀ɗìbɛ̀ɗɔ̀n *v*.; irwapírwápòn *v*. **blister** bubuxánónuƙot<sup>a</sup> *v*.; ileɓíléɓòn *v*.; ɔ́ʝáìm *n*. **blistered** bubuxánón *v*. **bllgh bllgh (boiling)** ƙwɛ̀ʝ ɛ *ideo*. **bloated** teɓúsúmòn *v*. **bloated from overeating** xɛxánón *v*. **block** ɗɨnɛ́s *v*.; ɨɓatɛs *v*.; ɨƙɨɛs *v*.; ɲéɓulók<sup>a</sup> *n*.; teɡeles *v*.; tɨts'ɛ́s *v*. **block off** toƙólésuƙot<sup>a</sup> *v*. **block repeatedly** ɨɓatíɓátɛ́s *v*. **block the road** teɡelesa ɲerukuɗeé *v*. **block up** tɨts'ɛ́sʉ́ƙɔt<sup>a</sup> *v*.; tits'ímétòn *v*. **blocked** ɗɨnɔ́s *v*.; tɨts'ɔ́s *v*. **blockheaded** lɛrɛ́dɔ̀n *v*. **blood** sè *n*. **blood (coagulated)** ŋazul *n*. **blood covenant** iɓólínósa na séà<sup>e</sup> *v*. **blood herb** seacɛmɛ́r *n*. **blood red** tsòn *ideo*. **blood relation** ɦyeínósá na séà<sup>e</sup> *v*. **blood vessel** seamucé *n*.; tsòrìt<sup>a</sup> *n*. **bloodthirsty person** sèààm *n*. **bloom** ɨɔ́kɔ́n *v*. **blossom** ɨɔ́kɔ́n *v*.; ŋátur *n*. **blow** fútés *v*.; fútón *v*. **blow (a projectile)** ɗɛ́tɛ́s *v*. **blow (of breeze)** ipiipíòn *v*. **blow (waste)** eletiésuƙot<sup>a</sup> *v*.; iɲekes *v*.; iɲékésuƙot<sup>a</sup> *v*.; iɲekíɲékés *v*. **blow away** fútésuƙot<sup>a</sup> *v*. **blow nose** rʉ́tɛ́s *v*. **blow off** fútésuƙot<sup>a</sup> *v*. **blow on gently** borotsiés *v*. **blow up** taɲɛ́ɔ́n *v*.; xuanón *v*.; xuxuanitetés *v*.; xuxuanón *v*. **blow up (explode)** toɗúón *v*. **blubber** ceím *n*.
**blue-gray** bósánòn *v*.; kábusubusánón *v*. **blunder (verbally)** eletiesá mɛná<sup>ɛ</sup> *v*. **blunt** duŋúlúmòn *v*.; líídòn *v*.; tufádòn *v*. **bluntly** lì *ideo*.; tùf *ideo*. **blurt news** eletiesá mɛná<sup>ɛ</sup> *v*. **blush** ɓʉnʉ́mɔ́nà sèà<sup>e</sup> *v*. **bluster** íɡòmòn *v*. **board** ɲáɓáo *n*.; otsés *v*. **boast** íɡòmòn *v*.; ɨwʉ́lɔ́n *v*. **boasting** itúrónìtòd<sup>a</sup> *n*. **boat** itúɓ<sup>a</sup> *n*. **bob** iupúúpòn *v*. **bodiliness** nébùnànès *n*. **body** nêb<sup>a</sup> *n*. **body hair** nébùsìts'<sup>a</sup> *n*. **body of water** cúénêb<sup>a</sup> *n*. **body part** ɲekiner *n*. **body snatcher** tukutesíáma ts'óóniicé *n*. **boggily** fɔ̀ts'<sup>ɔ</sup> *ideo*. **boggy** fɔts'ɔ́dɔ̀n *v*. **bogus** láŋ *n*. **boil** féés *v*.; íɡulaʝitetés *v*.; íɡùlàʝòn *v*.; itsúrítetés *v*.; itsúrón *v*.; ƙwɛʝɛ́dɔ̀n *v*.; ƙwɛʝíƙwɛ́ʝɔ̀n *v*.; tún *n*.; wádòn *v*. **boil (large)** ɲáɓús *n*. **boiling** ƙwɛʝɛ́dɔ̀n *v*.; ƙwɛʝíƙwɛ́ʝɔ̀n *v*. **boing!** tùd<sup>u</sup> *ideo*. **Bokora person** mɔƙɔrɔ́ám *n*.; Ŋíɓɔ́kɔráám *n*. **bolt carrier** bùbùàƙw<sup>a</sup> *n*. **boma** ɓór *n*. **bombard** ídʉrɛ́s *v*. **bombard with spears** bɛrɛ́s *v*. **bombard with words** ídʉrɛ́sá tódà<sup>e</sup> *v*. **bombinate** ilólúés *v*. **bond** nɔtsɔ́mɔ́n *v*.
**bone** ɔk<sup>a</sup> *n*. **bone (cancellous)** ɲɛ́ɲam *n*. **bone (costal)** ŋabéríɔ̀k <sup>a</sup> *n*. **bone (inner ear)** bòsìɔ̀k <sup>a</sup> *n*. **bone (nasal)** aƙatíɔ́k <sup>a</sup> *n*. **bone (occipital)** nàmɛ̀ɗɔ̀ *n*. **bone (pubic)** didisíɔ́k <sup>a</sup> *n*. **bone (shoulder)** sawatɔ́ɔ́k <sup>a</sup> *n*. **bone (spongy)** ɲɛ́ɲam *n*. **bone (supraorbital)** ekúɔ́k <sup>a</sup> *n*. **bone (temporal)** bòsìɔ̀k <sup>a</sup> *n*. **bone (zygomatic)** akáƙúm *n*. **bone below sternum** toroɓóɔ́k <sup>a</sup> *n*. **bone marrow** hɛ̀ɡ <sup>a</sup> *n*. **boneheaded** lɛrɛ́dɔ̀n *v*. **bonfire** ɲáɡaaɗi *n*. **bony** ɡɔ́ɡɔ̀rɔ̀mɔ̀n *v*.; iróƙóòn *v*.; itóƙóƙòòn *v*.; kwédekwedánón *v*. **book** ɲáɓúk<sup>a</sup> *n*. **book of prayers** ɲáɓúka wáánà<sup>e</sup> *n*. **boom** ɗukuɗúkón *v*. **boom (voice)** ɔ́bɛ̀s *v*. **boom!** pɛ̀s *ideo*. **booster tower** ɲéɓusitá *n*. **boot** ŋáɓutús *n*. **border** cíkóroy<sup>a</sup> *n*. **bore** ɨlɔ́ɛ́tɔ̀n *v*.; ɨpɨrípírɛ́s *v*.; pulutiés *v*. **bored** bɔ́rɔ́n *v*.; ɨlɔ́yɔ́n *v*. **bored (become)** bɔrɛ́tɔ́n *v*. **bored (drilled)** tsàpòn *v*. **borehole** ɲatsʉʉma *n*. **borehole casing** ɲatsʉʉmánêb<sup>a</sup> *n*. **borehole footing** ɲatsʉʉmádɛ̀ *n*. **borehole handle** ɲatsʉʉmákwɛ́t <sup>a</sup> *n*. **borehole pipe** ɲatsʉʉmáárí *n*. **borehole shaft** ɲatsʉʉmáhò *n*. **borehole water** ɲatsʉʉmácúé *n*.
## boring
**boring** itópénòn *v*.; ʝɔ̀lɔ̀n *v*. **born again (religiously)** hoɗetésá así *v*. **born handless** duŋúlúmòn *v*. **borrow** iɗenes *v*.; iɗenetés *v*.; wáánɛtɛ́s *v*. **borrow from each other** wáánínɔ́s *v*. **Boscia angustifolia** ɓàʝ<sup>a</sup> *n*. **Boscia coriacea** ɡɛbɛʝ<sup>a</sup> *n*. **Boscia salicifolia** ròr *n*. **boss** ámáze *n*.; ámázeám *n*.; ámázeáma teréɡì *n*. **botch** hamʉʝɛ́s *v*. **both** ɡáí *quant*. **bothersome** fìfòn *v*.; ɨtsánánòn *v*. **bottle (glass)** ɲétsúpa *n*. **bottle (plastic)** ɲɛ́rɔɓɨrɔ́ɓ <sup>a</sup> *n*. **bottlecap game** ŋ́karakocóy<sup>a</sup> *n*. **bottom** ɔ́zɛ̀d <sup>a</sup> *n*.; ɔ̂z *n*. **bottom layer of beer** ɔ́zɛ̀dà mɛ̀sɛ̀ *n*. **bottom of pot/pan** dómóɔ̀z *n*. **bottomless** xakútsúmòn *v*. **boubou (slate-colored)** kukuɗets<sup>a</sup> *n*. **bough** dakúkwɛ́t <sup>a</sup> *n*. **bouillon** seekw<sup>a</sup> *n*. **boulder** taɓ<sup>a</sup> *n*. **boulder (flat)** ɡɨzá *n*.; lolataɓ<sup>a</sup> *n*. **boulder base** taɓádɛ̀ *n*. **bounce** íbotitésúƙot<sup>a</sup> *v*.; íbotitetés *v*.; íbòtòn *v*. **bounce along** ɨŋɔ́písɔ̀ɔ̀n *v*. **bounce off** iɗótón *v*. **bouncily** tùf *ideo*.; tùs *ideo*. **bouncy** tufádòn *v*.; tusúdòn *v*. **bound** iɗótón *v*.; íɡɔ̀rɔ̀bɔ̀n *v*.; ɨrɨɗɔs *v*.; zíkɔ́s *v*. **bound up together** zíkízɨkánón *v*.
**boundaries (having)** cíkóróìkànànès *n*. **boundary** cíkóroy<sup>a</sup> *n*. **boundary (garden)** ɲókorimít<sup>a</sup> *n*. **boundedness** cíkóróìkànànès *n*. **bow (head)** turúnétòn *v*.; turúnón *v*. **bow (weapon)** ɲakaw<sup>a</sup> *n*. **bowed** dɔ́ɡɔ̀lɔ̀mɔ̀n *v*. **bowed over** rétón *v*. **bowel** bùbùàƙw<sup>a</sup> *n*. **bowl (gourd)** kàɓàɲ *n*. **bowl (of child)** imáƙɔ́fɔ́*n*. **bowl-shaped** sakánámòn *v*.; tsuƙúlúmòn *v*. **box** ɲásaɗukú *n*.; tanaŋes *v*. **box in** rʉ́ɛ́s *v*. **boy** ɲámal *n*.; sore *n*. **boy (baby)** ɓɨs *n*. **boy (little)** soréím *n*. **boy (teenage)** ŋísɔ́rɔkɔ́ám *n*. **boys (teenage)** ŋísɔ́rɔk<sup>a</sup> *n*.; pànɛ̀ɛ̀s *n*. **bozo** ɲɛ́ɛ́s *n*. **bra** ɲákɨláƙ<sup>a</sup> *n*. **brace** íbunutsés *v*.; ɨrʉ́rʉ́ɓɛ́s *n*.; ƙaƙates *n*. **bracelet (coiled)** ɲɛ́kílɔɗa *n*. **bracelet (gold)** ɲámaritóít<sup>a</sup> *n*. **bracelet (white leather)** ɲácáɗa *n*. **brag** íɡòmòn *v*.; ɨwʉ́lɔ́n *v*. **bragging** itúrónìtòd<sup>a</sup> *n*. **braid** bɛrɛ́s *v*.; sikwés *v*. **braid up** sikwetés *v*. **brain** ɗam *n*. **brake pedal** titiritésíàw<sup>a</sup> *n*. **branch** dakúkwɛ́t <sup>a</sup> *n*.; kwɛt<sup>a</sup> *n*.; taŋatsárón *v*.; tɛlɛ́tsɔ́n *v*.; toŋélón *v*. **branch pile (dry)** ràm *n*. **brand** ɨmátsárɛ́s *v*.; ɨtsʉŋɛs *v*.; ɲámátsar *n*.
**brassiere** ɲákɨláƙ<sup>a</sup> *n*. **brave** itítíŋòn *v*. **bray** werétsón *v*. **bread** tɔbɔŋ *n*. **bread (flat)** ɲécapatí *n*. **break** badonuƙot<sup>a</sup> *v*.; cɛɛ́s *v*.; cɛɛ́sʉ́ƙɔt<sup>a</sup> *v*.; ɗɛsɛ́mɔ́n *v*.; ɗusés *v*.; ɗusúmón *v*.; ɗusutes *v*.; ɨkɛ́ŋɛ́ɗɛ́s *v*.; ŋʉrɛ́s *v*.; ŋʉrʉ́mɔ́n *v*.; tɔŋɛɗɛs *v*. **break (make)** badítésuƙot<sup>a</sup> *v*. **break (of day)** ɓelémón *v*. **break apart** ɗɛsɛ́ɗɛ́sánón *v*.; ɨʝʉƙʉ́ʝʉ́ƙɛ́s *v*. **break away** tatsáɗón *v*.; tɔpɛ́ɔ́n *v*. **break away (and come)** tɔpɛ́ɛ́tɔ̀n *v*. **break away (and go)** tɔpɛ́ɔ́nʉƙɔt<sup>a</sup> *v*. **break due to weight** xuƙúmétòn *v*. **break into** itúwéés *v*. **break into pieces** ɓɨlíɓílɛ́s *v*.; ɨɓɛsíɓɛ́sɛ́s *v*.; ɨtɛítɛ́ɛ́s *v*.; ŋɨlíŋílánón *v*. **break into song** tofóróƙétòn *v*. **break new ground** taɗates *v*.; túburés *v*. **break off** ɗusésúƙot<sup>a</sup> *v*.; ɨɓɛkíɓɛ́kɛ́s *v*.; pɛsɛlɛs *v*.; pɛsɛ́mɛ́tɔ̀n *v*.; pokés *v*.; poketés *v*.; pokómón *v*.; wakés *v*.; wɛts'ɛ́mɔ́n *v*.; wɛts'ɛ́s *v*.; wɛts'ɛtɛ́s *v*. **break off in groups** ŋɨlíŋílánón *v*.; pókíetésá asíɛ́kédìè kɔ̀n *v*. **break off in pieces** ɨwɛts'íwɛ́ts'ɛ́s *v*.; tomuɲes *v*.; wakatiés *v*.; wakáwákatés *v*.; wets'etiés *v*. **break out (of skin)** iŋárúrètòn *v*.; morétón *v*. **break the law** ŋʉrɛ́sá ìtsìkà<sup>ɛ</sup> *v*. **break up** ɨʝʉƙʉ́ʝʉ́ƙɛ́s *v*.; ɨwɛ́ɛ́lɛ́s *v*.; tɔpwaɲípwáɲɛ́s *v*. **break up (a group)** toɓwaŋes *v*. **break wind** fenétón *v*.
**breakable** ɛɔmɔ́dɔ̀n *v*.; ŋʉɓʉ́dɔ̀n *v*.; pokódòn *v*.; tɛɛmɛ́mɔ̀n *v*.; wɛts'ɛ́dɔ̀n *v*. **breakably** ɛ̀ɔ̀m *ideo*.; ŋʉ̀ɓ<sup>ᶶ</sup> *ideo*.; pòk<sup>o</sup> *ideo*.; wɛ̀ts'<sup>ɛ</sup> *ideo*. **breakage point** ɗusutesíáw<sup>a</sup> *n*. **breaker** ŋʉrɛ́síàm *n*. **breakfast** ŋƙáƙá na baratsó<sup>e</sup> *n*. **breast** îdw<sup>a</sup> *n*. **breast (area)** bakuts<sup>a</sup> *n*. **breast (cut of meat)** làf *n*. **breast areola** ídoeɗ<sup>a</sup> *n*. **breast milk** ámáìdw<sup>a</sup> *n*. **breast twitch** lɔkɨsíná *n*. **breastbone** ɡɔɡɔm *n*.; toroɓ<sup>a</sup> *n*. **breastbone (of a chicken)** bɨbɨʝ<sup>a</sup> *n*. **breastfeed** naƙwɛ́s *v*.; naƙwɛ́sʉ́ƙɔt<sup>a</sup> *v*.; naƙwɨtɛs *v*. **breastmilk (have)** cɛ̀rɔ̀n *v*. **breath** sʉ̀p <sup>a</sup> *n*. **breathe** ɨɛ́ŋɔ́n *v*.; sʉ́pɔ́n *v*. **breathe fitfully** ɗíɗítɔ̀n *v*. **breathe heavily** fúútòn *v*. **breathe in** sʉpɛ́tɔ́n *v*. **breathe out** sʉ́pɔ́nʉƙɔt<sup>a</sup> *v*. **breeze** suɡur *n*. **brew** otés *v*.; tsapés *v*.; waatɛ́s *v*. **brew (mead)** ts'ɔƙɛ́s *v*. **brew beer** tsapésá mɛ̀sɛ̀*v*.; waatɛ́sá mɛ̀sɛ̀ *v*. **brewski** mɛ̀s *n*. **bribe** ɨlʉŋʉ́lʉ́ŋɛ́s *v*. **brick** ɲéɓulók<sup>a</sup> *n*. **bride** bɔƙátín *n*. **bride (return a)** xɛɛsʉ́ƙɔt<sup>a</sup> *v*. **bridegroom** bɔƙátíníèàkw<sup>a</sup> *n*. **brideprice** buƙ<sup>a</sup> *n*. **brideprice (extract)** buƙitetés *v*.
**brideprice gift** buƙotam *n*. **brideprice payer** buƙúám *n*. **bridge** àrònìàw<sup>a</sup> *n*.; ɲɔɗɔrɔcá *n*.; tɛ̀wɛ̀r *n*. **bridge of nose** sarɨsar *n*. **brigandry** ŋirúkóìnànès *n*. **bright** ɡwɛɲɛ́mɔ́n *v*.; ɨmɛ́ɗɔ́n *v*. **bright (of sky)** tatsɔ́ɔ́n *v*. **brilliant** ɡwɛɲɛ́mɔ́n *v*.; ɨmɛ́ɗɔ́n *v*. **bring** detés *v*.; ɨakɛtɛ́s *v*.; ɨʝʉkɛtɛ́s *v*. **bring all of** ɨsʉɲɛtɛ́s *v*. **bring alongside** ɨnapɛtɛ́s *v*.; napɛtɛ́s *v*. **bring back** raʝetés *v*. **bring beside** ɨnapɛtɛ́s *v*.; napɛtɛ́s *v*. **bring down** ruɓutetés *v*. **bring matters to a close** kɔkɛtɛ́sá mɛná<sup>ɛ</sup> *v*. **bring out** pulutetés *v*. **bring slowly** ɨɛmɛtɛ́s *v*. **bring to a boil** itsúrítetés *v*. **bring together** itóyéés *v*. **bring up** ilímítés *v*.; tasɛɛs *v*. **bristling** bʉlʉbʉlɔs *v*. **bristly** bʉlʉbʉlɔs *v*. **British colonial government** ɡɛrɛ́sà *n*. **brittle** bɔɲɔ́dɔ̀n *v*.; ɛɔmɔ́dɔ̀n *v*.; ɨwɛlɛ́wɛ́lánón *v*.; ŋʉɓʉ́dɔ̀n *v*.; pɛsɛ́pɛ́sánón *v*.; pokódòn *v*.; wɛts'ɛ́dɔ̀n *v*. **brittlely** bɔ̀ɲ *ideo*.; ɛ̀ɔ̀m *ideo*.; ŋʉ̀ɓᶶ *ideo*.; pòk<sup>o</sup> *ideo*.; wɛ̀ts'<sup>ɛ</sup> *ideo*. **broad** laŋírímòn *v*.; laŋírón *v*.; zòòn *v*. **broad (of many)** zeikaakón *v*. **broad-bladed** fakádòn *v*. **broad-bladedly** fàk<sup>a</sup> *ideo*. **broadcast** tɛwɛɛs *v*. **broaden** zoonuƙot<sup>a</sup> *v*.
**broke (go)** ŋʉrʉ́mɔ́n *v*. **broken** bàdòn *v*.; ŋʉrɔ́s *v*. **broken (get)** ŋʉrʉ́mɔ́n *v*. **broken beyond repair** tɛ̀s *ideo*. **broken out (of skin)** iŋárúròn *v*.; katúrúturánón *v*. **broken up** ŋʉrʉ́ŋʉ́ránón *v*. **brood** ƙʉ́ƙín *n*.; siŋírón *v*. **broom** ʝan *n*.; ʝɛn *n*. **broomgrass** ʝan *n*.; ʝɛn *n*. **broth** ɲɔ́ɓɔ́ka *n*.; seekw<sup>a</sup> *n*. **brother (his/her/its)** leat<sup>a</sup> *n*. **brother (my)** ɛdɛ́*n*. **brother (your)** léó *n*. **brother-in-law (brother's wife's brother)** uɡwam *n*. **brother-in-law (her husband's brother)** ntsínámúí *n*. **brother-in-law (his brother's wife's brother)** ntsúɡwám *n*. **brother-in-law (his wife's brother)** ntsúɡwám *n*. **brother-in-law (his/her child's spouse's father)** ntsíɲót<sup>a</sup> *n*. **brother-in-law (his/her sister's husband's brother)** ntsúɡwám *n*. **brother-in-law (husband's brother)** námúí *n*. **brother-in-law (my brother's wife's brother)** ɲ́cuɡwám *n*. **brother-in-law (my husband's brother)** ɲ́cinamúí *n*. **brother-in-law (my sister's husband)** ɲ́cuɡwám *n*. **brother-in-law (my wife's brother)** ɲ́cuɡwám *n*. **brother-in-law (sister's husband's brother)** uɡwam *n*.
**brother-in-law (sister's husband)** ntsúɡwám *n*.; uɡwam *n*. **brother-in-law (wife's brother)** uɡwam *n*. **brother-in-law (your brother's wife's brother)** buɡwám *n*. **brother-in-law (your child's spouse's father)** biɲót<sup>a</sup> *n*. **brother-in-law (your husband's brother)** binamúí *n*. **brother-in-law (your sister's husband's brother)** buɡwám *n*. **brother-in-law (your sister's husband)** buɡwám *n*. **brother-in-law (your wife's brother)** buɡwám *n*. **brotherhood** leatínánès *n*. **brotherliness** leatínánès *n*. **brow bone** ekúɔ́k <sup>a</sup> *n*. **brown** muƙíánètòn *v*.; ɔŋɔ́ránètòn *v*. **brown (dark)** kɨpʉ́ránètòn *v*.; ts'aráfón *v*. **brown-skinned** ɗìwòn *v*. **brown-skinned person** ɗìwònìàm *n*. **bruise** ƙwár *n*.; rùròn *v*. **brush** ɨƙwɛrɛs *v*.; séɓés *v*.; sʉ́ƙʉ́tɛ́s *v*.; sʉ́ʉ́tɛ́s *v*.; tsekís *n*.; tsekísíàƙw<sup>a</sup> *n*. **brush (scrub)** ɲecaaƙo *n*. **brush (thick)** môɡ<sup>a</sup> *n*. **brush aside** ɨpalípálɛ́s *v*. **brush away** séɓésuƙot<sup>a</sup> *v*. **brush off** itútúés *v*.; itútúésuƙot<sup>a</sup> *v*.; séɓésuƙot<sup>a</sup> *v*. **brush out** ɨƙwɛ́rɛ́sʉƙɔt<sup>a</sup> *v*. **brushed** ɨƙwɛrɔs *v*. **brutal** ɨsílíánón *v*. **brute(s)** ínw<sup>a</sup> *n*.
**bubble** íɡùlàʝòn *v*. **bubble (make)** íɡulaʝitetés *v*. **buck (dollar)** ɲɔɗɔ́la *n*. **buck-toothed** ɨŋísímɔ̀n *v*. **bucket** ɲáɓákɛ̀t <sup>a</sup> *n*.; ɲɛ́ɓákɛ̀t <sup>a</sup> *n*. **bucket (metal)** ɲépeelí *n*. **bud** tún *n*. **buffalo** ɡàsàr *n*. **buffalo bull** eûz *n*. **buffalo calf** ɡàsàràìm *n*. **buffalo cow** ɡasaraŋwa *n*. **buffalo thorn tree** tílàŋ *n*. **bug** ilúlúés *v*.; ƙʉts'<sup>a</sup> *n*. **bug-eyed** ɓɛlɛ́rɛ́mɔ̀n *v*. **build** bɛrɛ́s *v*.; toyeetés *v*. **build on the ground** ɨkamɛtɛ́s *v*. **build up (of a termite mound)** tabúón *v*. **builder** bɛrɛ́síàm *n*.; ŋífunɗíàm *n*. **building (modern)** ɲeryaŋíhò *n*. **bulbous** bulúƙúmòn *v*.; lɔrɔ́dɔ̀n *v*. **bulbously** lɔ̀r *ideo*. **bulbul (common)** òtsìɓìl *n*.; tsìɓìl *n*. **bulge** tɨbíɛ́tɔ̀n *v*. **bulging** bulúƙúmòn *v*.; lɔrɔ́dɔ̀n *v*.; tsʉrʉ́ɗʉ́mɔ̀n *v*.; tɨbíɔ́n *v*. **bulgingly** lɔ̀r *ideo*. **bulkily** bɛ̀f *ideo*. **bulky** ɓutúrúmòn *v*.; bɛfʉ́dɔ̀n *v*.; bɛfʉ́kʉ́mɔ̀n *v*. **bull** cúrúk<sup>a</sup> *n*. **bulldoze** towutses *v*. **bulldozer** towútsónìàm *n*. **bullet** bʉbʉn *n*.; eɗed<sup>a</sup> *n*.; ɲámal *n*. **bullet hole** ɲámalíák<sup>a</sup> *n*. **bullet wound** bʉbʉnɔ́ɔ́ʝ <sup>a</sup> *n*. **bullrush** ìsìk<sup>a</sup> *n*. **bum** lɛŋɛ́s *v*.; lɛŋɛ́síàm *n*.; ɲakárámɨt<sup>a</sup> *n*.
300
**bump** íbaɗɛ́s *v*.; ɨɓaɲɛs *v*.; ɨɛ́bɛ̀s *v*.; ɨɛ́bɛtɛ́s *v*.; toyeres *v*. **bump (skin)** síts'ádɛ̀ *n*. **bump off (kill)** ɨɗɛɛs *v*.; ɨɗɛ́ɛ́sʉƙɔt<sup>a</sup> *v*. **bump repeatedly** íbaɗiés *v*. **bumpy** ƙumúƙúmánón *v*. **bumpy (of skin)** katúrúturánón *v*. **bunch** ɓòtòŋ *n*.; tutukesíáw<sup>a</sup> *n*.; zɨkam *n*. **bunch (of bees)** ɲénéne *n*. **bunch up** tutuketés *v*. **bunched up** tutukánón *v*. **bundle** ɨɗɨlɛs *v*.; méy<sup>a</sup> *n*.; zɨkam *n*. **bundle (of crops)** ɲénéne *n*. **bungle** hamʉʝɛ́s *v*. **bunk** ɨɓááŋàsìtòd<sup>a</sup> *n*. **bunny** tulú *n*. **buoy** ilélébètòn *v*. **burble** ábʉ̀bʉ̀ƙɔ̀n *v*.; ábʉ̀bʉ̀ƙɔ̀n *v*. **burden** bot<sup>a</sup> *n*.; ɨnʉɛs *v*.; isites *v*. **burdensome** ìsòn *v*. **burglar** dzúám *n*. **burgle** dzuesés *v*.; dzuesetés *v*. **buried** tʉnʉkɔs *v*. **burlgary** dzú *n*. **burn** ɨtsʉŋɛs *v*.; kʉpɛ́s *v*.; kʉpɛ́sʉ́ƙɔt<sup>a</sup> *v*.; ts'aɗíɔ́ʝ <sup>a</sup> *n*. **burn (blistered)** ɲéleɓuléɓu *n*. **burn (of eyes)** ŋaɓɨŋáɓɔ́n *v*.; ŋàɓɔ̀n *v*. **burn (of pain)** ɓɛɨɓɛ́ɔ́n *v*. **burn a little** ɨrɔɗírɔ́ɗɛ́s *v*. **burn around** ɨɗɛɨɗɛ́ɛ́s *v*. **burn off (land)** iróróbes *v*. **burn poorly** ɨkáwílɔ̀n *v*. **burn to ashes** wuɗétón *v*. **burn up** ɨtsʉ́ŋɛ́sʉƙɔt<sup>a</sup> *v*.
**burnt to ashes** wùɗòn *v*.; xawííts'<sup>ɨ</sup> *ideo*. **burp** xèr *n*.; xerétón *v*.; xérón *v*. **burrow** ak<sup>a</sup> *n*. **burst** ɓilímón *v*.; toɗúón *v*. **bury** búdès *v*.; búdesuƙot<sup>a</sup> *v*.; muɗés *v*.; tʉnʉkɛs *v*. **bury (make)** tʉnʉkɨtɛtɛ́s *v*. **bury a bird** muɗésá ɡwaá<sup>e</sup> *v*. **bury a stink-bug** muɗésá loɡeréɲoé *v*. **bury life of one's children** muɗésá ɦyekesíé wicé *v*. **bury medicine** muɗésá cɛmɛ́ríkà<sup>ɛ</sup> *v*. **bus** ɲáɓás *n*. **bush** ríʝ<sup>a</sup> *n*. **bush barbet** kɔkíríkɔk<sup>a</sup> *n*. **bush country** ríʝíkaaʝík<sup>a</sup> *n*. **bush(es)** tsekís *n*.; tsekísíàƙw<sup>a</sup> *n*. **bushbaby** ɡwan *n*. **bushbuck** kʉláɓ<sup>a</sup> *n*. **bushbuck (female)** natsíɓɨlí *n*. **bushbuck (male)** òɗòmòr *n*. **bushbuck leaf** kʉláɓákàk<sup>a</sup> *n*. **bushcamp** napérít<sup>a</sup> *n*. **bushpig** bòròk<sup>a</sup> *n*. **bushpig boar** borokucúrúk<sup>a</sup> *n*. **bushpig piglet** bòròkùìm *n*. **bushpig sow** borokuŋwa *n*. **bushy** tsèkòn *v*. **business** dzîɡw<sup>a</sup> *n*. **bustard** hɔ́tɔ̀ *n*. **busy** íɡùʝùɡùʝòn *v*. **but** kòt<sup>o</sup> *coordconn*. **butcher** hoés *v*.; hoesíàm *n*.; tɔ̀ŋɔ̀lààm *n*.; tɔŋɔlɛs *v*. **butchery** hoesího *n*. **butt** ɔ̂z *n*.; tɔɗɔ́pɔ́n *v*.
**butt (of gun)** dɛɛd<sup>a</sup> *n*. **butt cheek** komos *n*. **butt in** íbʉbʉŋɛ́s *v*.; íbʉbʉŋɛ́sʉ́ƙɔt<sup>a</sup> *v*. **butter** íbɔtsam *n*. **butter (seed)** ɲówoɗí *n*. **butter flask** ɲéɓur *n*. **butterfly** béɗíbeɗú *n*.; bóɗíboɗú *n*. **buttock** komos *n*. **buttock underside** kwatsíém *n*. **buttocks (have flat)** taɓóɲómòn *v*. **button** ɲáʝaará *n*. **buy** dzíɡwès *v*.; dzíɡwetés *v*. **buy off** ɨlʉŋʉ́lʉ́ŋɛ́s *v*. **buyer** dzíɡwààm *n*.; dzíɡwèsìàm *n*. **buzz around** ilólúés *v*. **buzzard (augur)** alálá *n*. **by foot** dɛ̀ìk<sup>ɔ</sup> *n*. **by night** mukú *n*. **by what path?** nday<sup>o</sup> *n*. **bypass** ɨsʉkɛs *v*.; sʉ́kɛ́s *v*. **cabbage** ɲákáɓìc *n*. **cabinet** ɲákábàt<sup>a</sup> *n*. **cache** irwanes *v*.; laɓ<sup>a</sup> *n*. **cackle** ɨkɛ́kɛ́mɔ̀n *v*. **cactus species** ɲɛ́ɲɛwán *n*. **Cadaba farinosa** mét<sup>a</sup> *n*.; súr *n*. **cadaver** loukúéts'<sup>a</sup> *n*.; ɲɛ́lɛl *n*. **cadaverous** iróƙóòn *v*.; itóƙóƙòòn *v*.; kwédekwedánón *v*. **cadge** lɛŋɛ́s *v*. **cadger** lɛŋɛ́síàm *n*. **café** ɲéótèl *n*. **cage (trap)** ɲáɓáo *n*. **cajole** ɨmámɛ́ɛ́s *v*. **calabash** kàŋɛ̀r *n*.
**calendar** ɲákalɛ́nɗà *n*. **calf** kɔrɔ̂b <sup>a</sup> *n*. **Calisa edulis** turunet<sup>a</sup> *n*. **call** óés *v*. **call (of animals)** ƙɔ̀ɗɔ̀n *v*. **call (in alarm)** íbòfòn *v*. **call (name)** óés *v*. **call continuously** óísiés *v*. **call each other** óímós *v*. **call for rain** dʉbɛ́s *v*. **call here** oetés *v*. **call off** ŋííɗɛ́sʉ́ƙɔt<sup>a</sup> *v*.; tasálétòn *v*.; tasálón *v*. **call on the phone** iwésúƙot<sup>a</sup> *v*.; iwetés *v*. **call repeatedly** óésés *v*. **call sweetly** ɨmámɛ́ɛtɛ́s *v*. **callous** ɓulúrúmòn *v*. **calm** tisílón *v*.; toikíkón *v*. **calm down** ɨʝɛ́mɔ́nʉƙɔt<sup>a</sup> *v*.; itiketésá ɡúróe kíʝák<sup>e</sup> *v*.; toíésuƙot<sup>a</sup> *v*.; zɛƙwɛ́tɔ́n *v*.; zɛƙwɨtɛtɛ́s *v*. **calm down the heart** cucuéítésuƙota ɡúró<sup>e</sup> *v*. **calm dowwn** ɗipímón *v*. **calm person** ɡúróàm *n*. **camel** ɡwaíts'<sup>a</sup> *n*.; ɲákáal *n*. **camp** napérít<sup>a</sup> *n*.; ɲákamɓí *n*. **camp (garden)** ɲóɓóot<sup>a</sup> *n*. **camp kitchen** ŋƙáƙáhò *n*. **campaign** was *n*. **campaign ads** ɲáɓás *n*. **can** topéɗésuƙot<sup>a</sup> *v*. **can (large)** ɲéɗépe *n*. **can (metal)** ɲɛ́kɛ́n *n*. **cancel** ŋííɗɛ́sʉ́ƙɔt<sup>a</sup> *v*.; tasálétòn *v*.; tasálón *v*. **cancel (make)** ŋííɗítɛ́sʉƙɔt<sup>a</sup> *v*. **candidacy** was *n*.
cartilage
**candidate** wasɔ́ám *n*. **candidiasis** losúk<sup>a</sup> *n*. **candle wax** sɔs *n*. **candy** ɲátamɨtám *n*. **cane** ɨɗɨtsɛs *v*.; ɨnɔmɛs *v*.; kasír *n*.; sɛ̀w<sup>a</sup> *n*. **cane (hooked)** ɲótooɗó *n*. **cane rat** ŋʉr *n*. **canine tooth** bàdìàm *n*. **cannabis** lɔ́tɔ́ɓa ná zè *n*. **cannibal** áts'ɛ́sìàmà ròɓà<sup>e</sup> *n*. **cannon** komótsɛ́ɛ̀b <sup>a</sup> *n*. **canteen (gourd)** nasɛmɛ́*n*. **Canthium lactescens** kómoló *n*. **Canthium species** dodík<sup>a</sup> *n*.; kàrɛ̀ *n*.; milékw<sup>a</sup> *n*. **canvass (an area)** ɨkáyɛ́ɛ́s *v*. **cap** ɲákopiyá *n*. **cap (giraffe-tail)** ɲɔ́tsɔ́ɓɛ *n*. **cap (human hair)** ɲóɓókot<sup>a</sup> *n*. **cap (ignition)** ɔ́zɛ̀d <sup>a</sup> *n*. **capability** ɲapéɗór *n*. **capable of** topéɗésuƙot<sup>a</sup> *v*. **capillary** tsòrìt<sup>a</sup> *n*. **Capparaceae species** tsàl *n*. **Capparis tomentosa** lókúɗukuɗét<sup>a</sup> *n*. **capture** dʉ́bɛ̀s *v*.; ɨkamɛ́sʉ́ƙɔt<sup>a</sup> *v*. **car** kàèìm *n*.; ɲómotoká *n*. **car (small)** pɔ́ɗɛ̀ *n*. **caracal** naɨtakípʉ́rat<sup>a</sup> *n*. **carbon black** ɲémúɗets<sup>a</sup> *n*.; ɲémúɗuɗu *n*. **carcass** ɗòl *n*. **card (identity)** ɲákáɗ<sup>a</sup> *n*. **card (Kenyan ID)** ɲɛkɨpanɗɛ *n*. **card (playing)** ɲákáɗ<sup>a</sup> *n*.
**cardboard (thin)** ɲɛ́páìl *n*. **Cardiospermum corindum** tìl *n*. **cards (playing)** ɲákáɗìk<sup>a</sup> *n*. **care** ɨkatsɛs *v*.; ɨmɨsɛs *v*. **care for** bɔnɛ́s *v*.; ɨrɨtsɛ́s *v*. **care for (the sick)** maitetés *v*. **care for oneself** ɨrɨtsɛ́sá así *v*. **care-free person** ɨɓámɔ́nìàm *n*. **careful** tɔtsɔ́ɔ́n *v*. **carefully** hɨíʝ<sup>a</sup> *adv*.; hɨíʝ<sup>ɔ</sup> *adv*. **carelessly** càc<sup>ɨ</sup> *adv*.; fùts'àts'<sup>a</sup> *ideo*.; tsàr *ideo*. **caress** ɨɓɔníɓɔ́nɛ́s *v*.; iɓoníɓóniés *v*.; ɨwáwɛ́ɛ́s *v*. **caretaker** bɔnɛ́ám *n*.; ɨrɨtsɛ́síàm *n*.; itelesíám *n*. **caretaking** bɔn *n*. **cargo** bot<sup>a</sup> *n*.; botitín *n*. **Carica papaya** ɲápaɨpáy<sup>a</sup> *n*. **carinated** toŋórómòn *v*.; toróŋómòn *v*. **carissa shrub** turunet<sup>a</sup> *n*. **carnivore** loúk<sup>a</sup> *n*. **carpenter** ɲáɓáòìkààm *n*. **carrier** tsídzèsìàw<sup>a</sup> *n*. **carrion** itam *n*. **carrot** kárʉ̀ts'<sup>a</sup> *n*. **carry away on the head** tsídzesuƙot<sup>a</sup> *v*. **carry by hand** taɓakés *v*. **carry in arms** taɓakés *v*. **carry many of** rúdzès *v*. **carry on shoulder** tʉ́zʉŋɛ́s *v*. **carry on the back** édès *v*. **carry on the head** tsídzès *v*. **carry on the head (make)** tsídzitetés *v*. **carry together** ilélébés *v*. **carry under** ɨlʉkɛs *v*. **cartilage** ŋɔrɔɓɔɓ<sup>a</sup> *n*.
**cartilaginous** rɔɓɔ́dɔ̀n *v*. **cartridge** ɲéɓurocó *n*.; ɲɛsɛpɛɗɛ *n*. **carve** sotés *v*.; sotetés *v*. **case (legal)** ɲékés *n*. **cash** kaúdzà nì ɓèts'<sup>a</sup> *n*. **casing** hò *n*. **casing (borehole)** ɲatsʉʉmánêb<sup>a</sup> *n*. **casing (shell)** ɲéɓurocó *n*.; ɲɛsɛpɛɗɛ *n*. **cassava** ɲómoŋɡó *n*. **Cassia hildebrandtii** ɲasal *n*. **Cassia singueana** lɔ́kɛ́rʉ́ *n*. **cast** ɡóózés *v*. **cast (for divination)** ipés *v*. **cast away** ɡóózesuƙot<sup>a</sup> *v*.; hábatsésúƙot<sup>a</sup> *v*. **cast down** hábatsetés *v*. **cast sandals (in divination)** ipésá taƙáíkà<sup>ɛ</sup> *v*. **cast stones (in divination)** ipésá ɡwàsìkà<sup>e</sup> *v*. **cast this way** ɡóózetés *v*. **castor-oil plant** ɨmánán *n*. **castrate** ɨƙɛ́lɛ́mɛ́s *v*. **casual** faɗétón *v*. **cat** púùs *n*. **cat's-tail** ìsìk<sup>a</sup> *n*. **catapult** ɲapaaru *n*. **cataract** kɛ̀s *n*. **catch** dʉ́bɛ̀s *v*.; ɨkamɛ́sʉ́ƙɔt<sup>a</sup> *v*.; ɨkamɛtɛ́s *v*.; sáɡwès *v*. **catch (up with)** rítsɛ́s *v*. **catch a whiff** wetésá kɔíná<sup>ɛ</sup> *v*. **catch fire** aeétón *v*.; lɛ́ʝɛ́tɔ́n *v*. **catch off guard** bóɡès *v*.; itúúmés *v*. **catch one's breath** ɨɛ́ŋɔ́nʉƙɔt<sup>a</sup> *v*.
**catch sight of** enésúƙot<sup>a</sup> *v*.; tataɲes *v*. **catch the attention of** ɨtʉ́lɛ́ɛtɛ́s *v*. **catechism** ɲákatékísìmù *n*. **catechist** ŋíkatikisítààm *n*. **categorize** ɨsíílɛ́s *v*. **cater for** íɡɔɲɛ́s *v*. **Catha edulis** ɲémurúŋ́ɡù *n*. **Catholic church** Katólìkà *n*. **Catholic person** Katólìkààm *n*. **Catholic priest** páɗɛ̀r *n*. **Catholicism** Katólìkà *n*. **cattle** ɦyɔ̀ *n*. **cattle disease** lɔlɛ́ɛʉ́ *n*. **cattle herd** ɦyɔ̀bàr *n*. **Caucasian** ɓèts'ònìàm *n*.; ɲémúsukit<sup>a</sup> *n*. **cause abdominal pain** ɲimíɲímàtòn *v*. **cause pain** áts'ɛ́s *v*. **cause problems** ɓuƙetésá mɛná<sup>ɛ</sup> *v*. **cause problems for** ɨtsánítɛtɛ́s *v*. **cause torment for** ɨtsanítsánɨtɛtɛ́s *v*. **cauterize** itsues *v*. **cautious** toikíkón *v*. **cavalier** faɗétón *v*. **cave** pakw<sup>a</sup> *n*. **cave (small)** pàɗw<sup>a</sup> *n*. **cave (vertical)** ɲáɗúy<sup>a</sup> *n*. **cave entrance** pakóásák<sup>a</sup> *n*. **cave interior** pakóáƙw<sup>a</sup> *n*. **caved in** paɗókómòn *v*. **cavern (vertical)** ɲáɗúy<sup>a</sup> *n*. **cavernous** wòò *ideo*. **cavity (abdominal)** bùbùàƙw<sup>a</sup> *n*. **cavity (oral)** ak<sup>a</sup> *n*. **cease** bɔlɔnʉƙɔt<sup>a</sup> *v*. **cease (blowing or boiling)** tilímón *v*. **ceaseless** rítsírɨtsánón *v*. **cedar (African pencil)** asʉnán *n*.
**catch scent of** mídzatetés *v*.
charmer
**ceiling (upper)** lɔɓîz *n*. **celebrate** ɨnʉmʉ́nʉ́mɛ́s *v*.; ɨnʉ́nʉ́mɛ́s *v*.; iyóómètòn *v*. **celebration** ɲápáti *n*. **cellular network** suɡur *n*. **Celosia schweinfurthiana** sɔ́ɡɛ̀kàk<sup>a</sup> *n*. **cement** ɲésímìt<sup>a</sup> *n*. **census** ɲɛkɨmar *n*. **cent** ŋáɓɔ́ɔla *n*.; ŋásɛntáìèkw<sup>a</sup> *n*. **center** ekw<sup>a</sup> *n*.; sɨsɨk<sup>a</sup> *n*.; sɨsíkɛ̂d <sup>a</sup> *n*. **center of snare** siméékw<sup>a</sup> *n*. **center point** ekw<sup>a</sup> *n*. **centipede** ƙɔ́dɔxɔ́*n*. **central part** bakútsêd<sup>a</sup> *n*. **cents** ŋásɛntáy<sup>a</sup> *n*. **cereal** eɗ<sup>a</sup> *n*. **ceremonial meal** írésiŋƙáƙ<sup>a</sup> *n*. **ceremony (conduct a)** írés *n*. **certainly** kárɨká *adv*. **cervical vertebrae** ɦyʉƙʉmʉ́ɔ́k <sup>a</sup> *n*. **cervix** dòɗìèkw<sup>a</sup> *n*. **cessation** was *n*. **cestode** apéléle *n*. **chafed** ɲɛɗɛ́dɔ̀n *v*. **chafedly** ɲɛ̀ɗ ɛ *ideo*. **chaff dust** ŋawíl *n*. **chaffs** nakariɓ<sup>a</sup> *n*. **chain** zɔ̀t <sup>a</sup> *n*. **chained** zɔ́tɔ́n *v*. **chair** kàràts<sup>a</sup> *n*. **chairperson (of village)** ámázeáma awá<sup>e</sup> *n*. **chalk** ɲɛ́cɔ́ka *n*. **chalkboard** ɲáɓáo *n*. **chalky (dry)** pʉrákámòn *v*.; pʉráŋámòn *v*.; pʉsɛ́lɛ́mɔ̀n *v*.
**challenge (verbally)** nɛpɛƙanitetés *v*. **chamber (bee larva)** sídàhò *n*. **chameleon** híƙɔ́*n*. **chance** ŋawɨlɛs *v*. **chance upon** taƙámón *v*. **change** beníónuƙot<sup>a</sup> *v*.; iɓéléés *v*.; iɓéléìmètòn *v*.; icéɲʝeés *v*.; icéɲʝèìmètòn *v*.; ilotses *v*.; ilotsímétòn *v*. **change (money)** toɓwaŋes *v*. **change allegiences** tolúétòn *v*.; tolúónuƙot<sup>a</sup> *v*.; tolúútésuƙot<sup>a</sup> *v*. **change course** wɛ́dɔ̀n *v*. **change decisions** ilotsesa mɛná<sup>ɛ</sup> *v*. **change one's direction** iɓéléésuƙota así *v*. **change position** ɨsʉ́tɔ́n *v*. **change statements** iɓelíɓélésa tódà<sup>e</sup> *v*. **change the story** iɓelíɓélésa tódà<sup>e</sup> *v*. **change the subject** ɨʝʉlɛsa tódà<sup>e</sup> *v*. **chaotic (beocome)** iŋóɗyáìmètòn *v*. **chapati** ɲécapatí *n*. **chapeau** ɲákakar *n*. **chapel** ɲɛ́cápɔ̀l *n*. **chapped** ɓɛlɛ́ɓɛ́lánón *v*. **chapter** ɔ́dɔ̀k <sup>a</sup> *n*. **char** kɔrɛtɛ́s *v*.; kɔrɨtɛtɛ́s *v*. **charcoal** leûz *n*.; ɲámakáy<sup>a</sup> *n*. **charge (accuse)** ɨsíítɛ́s *v*. **charge (attack)** toƙíróòn *v*. **charge (electrically)** cɨɨtɛ́sʉƙɔt<sup>a</sup> *v*.; hábitésúƙot<sup>a</sup> *v*.; ŋƙɨtɛtɛ́s *v*.; wetitésuƙot<sup>a</sup> *v*. **charge (order)** ɨtsɨkɛs *v*. **Charismatic** Ŋímorokóléìàm *n*. **charitable** waŋádòn *v*. **charm** sʉ́bɛ̀s *v*.; sʉ́bɛsʉƙɔt<sup>a</sup> *v*. **charmer** sʉ́bɛ̀sìàm *n*.
#### charred
**charred** kɔrɛ́tɔ́n *v*. **chase** ɨfalífálɛ́s *v*.; ɨlɔŋɛs *v*. **chase after** ɨlɔ́ŋɛ́sʉƙɔt<sup>a</sup> *v*.; irukes *v*.; irúkésuƙot<sup>a</sup> *v*. **chase away** ikutses *v*.; ikútsésuƙot<sup>a</sup> *v*. **chase down (the throat)** itikes *v*. **chase off** ikutses *v*.; ikútsésuƙot<sup>a</sup> *v*. **chattily** ɗɛ̀m *ideo*.; pòx *ideo*. **chatty** ɗɛmɛ́dɔ̀n *v*.; ɗɛmɨɗɛ́mɔ́n *v*.; ɨkʉ́tʉ́kánón *v*.; poxódòn *v*. **chaw** ts'àf *n*. **cheap** batánón *v*. **cheat** ɨmɔɗɛs *v*.; ɨmɔ́ɗɛ́sʉƙɔt<sup>a</sup> *v*. **check (bank)** kaúdzokabáɗ<sup>a</sup> *n*. **check (mark)** totsetes *v*. **check on** ɨfátɛ́sʉƙɔt<sup>a</sup> *v*.; láɡalaɡetés *v*. **check out** ɡonés *v*.; ɦyeités *v*.; ɨfátɛ́sʉƙɔt<sup>a</sup> *v*.; iséméés *v*.; iséméetés *v*.; láɡalaɡetés *v*.; tɨrɨfɛs *v*.; tɨrɨfɛtɛ́s *v*. **check out (here)** ɡonetés *v*. **check out (there)** ɡonésúƙot<sup>a</sup> *v*. **check out thoroughly** tirifiés *v*. **checkers** ɲɛ́ɗʉráp<sup>a</sup> *n*. **cheek** ɔ̂b <sup>a</sup> *n*. **cheek (butt)** komos *n*. **cheekbone** akáƙúm *n*. **cheep cheep!** ƙíɛƙíɛƙíɛ *ideo*. **cheese** ɲéɗíol *n*. **cheetah** ɲárará *n*. **chef** kɔŋɛ́síàm *n*. **chemical** cɛ̀mɛ̀r *n*. **Chenopodium opulifolium** tsamʉya *n*. **cherish** mínɛ́s *v*. **chest** bakuts<sup>a</sup> *n*. **chest (inner)** ɡúr *n*. **chest (storage)** ɲásaɗukú *n*.
**chest disease** ɲeɗekea bákútsìkà<sup>e</sup> *n*. **chest hair** bakutsísíts'<sup>a</sup> *n*. **chew** áts'ɛ́s *v*.; ɨɲáɗʉ́tɛ́s *v*. **chew (tobacco)** ɨmátáŋɛ́s *v*.; mataŋɛs *v*. **chew extractively** ts'afɛ́s *v*. **chew on** ɲɛɓɛ́s *v*. **chew roughly** ɨkaŋíkáŋɛ́s *v*.; ɨŋaíŋɛ́ɛ́s *v*.; ɨŋáŋɛ́ɛ́s *v*. **chewily** kàŋ *ideo*.; kwàⁱ *ideo*. **chewy** dɨrɨɲíɲɔ́n *v*.; kaŋádòn *v*.; kwaídòn *v*. **chick** ɡwáím *n*.; ɲɔ́kɔkɔróím *n*. **chicken** ɲɔ́kɔkɔr *n*. **chicken backbone** ɲéturuƙúƙu *n*. **chicken coop** ɲɔ́kɔkɔrɔ́hò *n*. **chickenpox** ɲɛtʉnɛ *n*.; puurú *n*. **chide** dɔxɛ́s *v*.; dɔxɛ́sʉ́ƙɔt<sup>a</sup> *v*. **chief** ámáze *n*.; ámázeám *n*.; ɲéríósit<sup>a</sup> *n*. **chief (crew)** ámázeáma teréɡì *n*. **chief (parish)** ɲékúŋut<sup>a</sup> *n*. **chief (subcounty)** ɲɛʝákáɨt<sup>a</sup> *n*. **chief elder** ámázeáma awá<sup>e</sup> *n*.; diyoama ná zè *n*. **child** im *n*. **child (foreign)** ɦyòìm *n*. **child (his/her/its)** ntsíím *n*. **child (my)** ɲ́cììm *n*. **child (of someone)** ámáìm *n*. **child (your)** biím *n*. **child's bowl** imáƙɔ́fɔ́*n*. **childhood** imánánès *n*.; wicénánès *n*. **childishness** imánánès *n*.; wicénánès *n*. **childless** ikólípánón *v*.; osorosánón *v*. **childless person** ɲokólíp<sup>a</sup> *n*.; òsòròs *n*. **childlikeness** imánánès *n*.; wicénánès *n*. **children** wik<sup>a</sup> *n*. **children (young)** kómósikaa ɓets'aakátìk<sup>e</sup> *n*.
**chill** cucuéítésuƙot<sup>a</sup> *v*. **chill (damp)** cucue *n*. **chill out** zɛƙwɛ́tɔ́n *v*. **chilly** cucuéón *v*. **chin** tatʉ́n *n*. **chinches** ŋítʉ́mìk<sup>a</sup> *n*. **chink against** íɡatsɨɡatsɛ́s *v*. **chip** ŋɛlɛ́s *v*.; pɛsɛlam *n*.; pɛ́sɛ́lamed<sup>a</sup> *n*.; pɛsɛlɛs *v*.; tɛŋɛlɛs *v*. **chip (wood)** kíɓɛ́zam *n*. **chip in** tɔ́ƙɛ́s *v*. **chip off** ŋɛlɛ́mɛ́tɔ̀n *v*.; pɛsɛ́mɛ́tɔ̀n *v*.; wɛts'ɛ́mɔ́n *v*.; wɛts'ɛ́s *v*.; wɛts'ɛtɛ́s *v*. **chip off (bark)** iɓóɓólés *v*. **chip off in pieces** wets'etiés *v*. **chip repeatedly** ɨwɛts'íwɛ́ts'ɛ́s *v*. **chirp chirp!** ƙíɛƙíɛƙíɛ *ideo*. **choir** ɲákwáya *n*. **choke** ɨkɛtɛs *v*.; iketiés *v*. **cholera** kɔlɛ́rà *n*. **chomp** iŋomes *v*.; ƙídzɛ̀s *v*. **choose** ɗʉmɛtɛ́s *v*.; iɗókóliés *v*.; ɨƙɛlɛs *v*.; ɨƙɛlɛtɛ́s *v*.; ɲʉmɛtɛ́s *v*.; tɔsɛɛtɛ́s *v*.; xɔ́bɛtɛ́s *v*. **choosen** xɔ́bɔtɔ́s *v*. **chop in pieces** ŋurutiés *v*. **chop up in pieces** ŋurutiesúƙot<sup>a</sup> *v*. **chop!** pùà *ideo*. **chopped in pieces** ŋurutiós *v*. **chopper** naƙílɨƙíl *n*. **chorus** ɲákwáya *n*. **chosen** ƙanotós *v*. **Christian** Ŋíkiristóìàm *n*. **Christian name** ɓatísimúêd<sup>a</sup> *n*. **Christianity** ŋíkiristóìnànès *n*. **Christlikeness** ŋíkiristóìnànès *n*.
**Christmas service** nuélì *n*. **chronic** kɔ̀wɔ̀n *v*. **chronically ill** maimoos *v*. **chronically ill person** maimoosíám *n*. **chubbily** dɛ̀ʝ ɛ *ideo*. **chubby** dɛʝɛ́dɔ̀n *v*. **chuckle** ɡaɡaanón *v*. **chug** ɡéɡès *v*.; ɨʝíírɛ́sʉƙɔt<sup>a</sup> *v*. **chunk** ʝulam *n*. **chunk (small)** ʝulamáím *n*. **chunky** ŋʉmʉ́ŋʉ́mánón *v*. **church** ɲakuʝíhò *n*.; ɲéɡelesíà *n*.; wáánàhò *n*. **churn** íbɔbɔtsɛ́s *v*.; íbɔtsɛ́s *v*.; íbɔtsɛ́sá así *v*.; íɡùlàʝòn *v*. **chyme** ey<sup>a</sup> *n*.; ɡaɗár *n*.; ŋkʉ́ʝít<sup>a</sup> *n*. **cicada** tswíɨtswí *n*. **cicatrix** ƙwár *n*. **cigarette** ɲɔsɔkatá *n*. **cinder** bʉbʉn *n*. **cinema** ɲévíɗyòhò *n*. **circle** tɔlʉkɛs *v*.; tɔlʉ́kɛ́sʉƙɔt<sup>a</sup> *v*.; tɔlʉkɛtɛ́s *v*. **circles in (make)** tɔlʉkʉ́lʉ́kɛ́s *v*. **circular** ɗukúdòn *v*. **circularly** ɗùk<sup>u</sup> *ideo*. **circulate** ɨlɔmílɔ́mɔ̀n *v*.; irímétòn *v*.; irímítetés *v*.; irímón *v*. **circulate in** irimes *v*. **circumcise** ɨlíŋírɛ́s *v*.; ɨlɨrɛs *v*. **circumcised** lɛŋɛ́rɛ́mɔ̀n *v*. **circumvent** tamanɛs *v*.; tamanɛtɛ́s *v*. **circumventing** firifíránón *v*.; wíríwíránón *v*. **circumvolve** ɨlɨrɛs *v*. **circumvolve repeatedly** ɨlɨrílírɛ́s *v*. **Cissus rhodesiae** bɛfácɛ́mɛ́r *n*. **Citrullus species** nàdɛ̀kwɛ̀l *n*.
**citrus fruit** ɲámucúŋ́ƙà *n*. **city** zɛƙɔ́áwa ná zè *n*. **city (capital)** awa ná zè *n*. **civet (African)** mɨnít<sup>a</sup> *n*. **clack** ɨrɔʝírɔ́ʝɛ́s *v*. **clamber** ɨlɛ́pɔ́n *v*. **clamber down** ɨtsɔrɛtɛ́sá así *v*.; ɨtsɔ́rɔ́nʉƙɔt<sup>a</sup> *v*. **clamber up** ɨlɛ́pɛ́sʉƙɔt<sup>a</sup> *v*.; ɨtsɛ́tsɛ́ɛ́s *v*. **clamp** rɨɗɛ́s *v*.; wɨɗɨɗánón *v*.; wíɗíwɨɗánón *v*. **clamp shut** riɗímétòn *v*. **clan** bònìt<sup>a</sup> *n*. **clan (Gaɗukuɲ)** Gaɗukúɲ *n*. **clan (Iléŋ)** Iléŋ *n*. **clan (Komokua)** Kòmòkùà *n*. **clan (paternal)** ɔ́dɔ̀k <sup>a</sup> *n*.; àsàk<sup>a</sup> *n*. **clan (Siƙetia)** Sìƙètìà *n*. **clan (Telek)** Tɛ́lɛ́k <sup>a</sup> *n*. **clan (Uzet)** Úzɛ̀t <sup>a</sup> *n*. **clan (Ŋiɓoŋorona)** Ŋíɓɔ́ŋɔrɔna *n*. **clan (Ŋiɗotsa)** Ŋíɗɔ́tsa *n*. **clan (Ɲorobat)** Ɲɔrɔbat<sup>a</sup> *n*. **clan member (Gaɗukuɲ)** Gaɗukúɲùàm *n*. **clan member (Iléŋ)** Iléŋíàm *n*. **clan member (Komokua)** Kòmòkùààm *n*. **clan member (Siƙetia)** Sìƙètìààm *n*. **clan member (Telek)** Tɛ́lɛ́kìàm *n*. **clan member (Uzet)** Úzɛ̀tìàm *n*. **clan member (Ŋiɓoŋorona)** Ŋíɓɔ́ŋɔrɔnáám *n*. **clan member (Ŋiɗotsa)** Ŋíɗɔ́tsáám *n*. **clan member (Ɲorobat)** Ɲɔrɔbatíám *n*. **clap** ɨɗafɛs *v*.
**clap clap!** rààrààrà *ideo*. **clap hands** ɨɗafɛsa kwɛ́tìkà<sup>ɛ</sup> *v*. **clapping dance** ɲókorot<sup>a</sup> *n*. **clarify** enitetés *v*. **clash** iƙúmúnós *v*. **clasp** mʉkʉtɛs *v*. **class** ɲɛ́kɨlás *n*.; ɲókós *n*. **class (school)** hò *n*. **classify** ɨsíílɛ́s *v*. **classroom** ɲɛ́kɨlás *n*. **clavicle** ɲálaƙamááìtìɔ̀k <sup>a</sup> *n*. **claw** soƙóríties *v*.; tíbòlòkòɲ *n*. **claw up** ikúkúrés *v*.; tukutes *v*. **clay (black)** ŋɔra na buɗám *n*. **clay (colored)** ŋɔr *n*. **clay (red)** ŋɔra na ɗíw<sup>a</sup> *n*. **clay (smear with)** ŋɔrɨtɛtɛ́s *v*. **clay pot (blackened)** dómá na buɗám *n*. **clay pot (small)** ɲeƙulu *n*. **clean** ɓèts'òn *v*.; fítés *v*.; kɨlíwítánón *v*.; ŋííɗɛ́s *v*.; sʉ́ƙʉ́tɛ́s *v*.; sʉ́ʉ́tɛ́s *v*. **clean (a surface)** ɨkʉlɛs *v*.; ɨkwalɛs *v*.; ɨlɨwɛs *v*. **clean off** ŋííɗɛ́sʉ́ƙɔt<sup>a</sup> *v*. **clean off (a surface)** ɨkʉ́lɛ́sʉƙɔt<sup>a</sup> *v*.; ɨlɨwílíwɛtɛ́s *v*. **clean up** ŋííɗɛtɛ́s *v*. **clean up (a surface)** ɨkʉlɛtɛ́s *v*. **cleaned off** ɨlɨwílíwɔ́s *v*. **cleaning rod** sʉ́ƙʉ́tɛ́sítsɨrím *n*. **clear** ɓèts'òn *v*.; bótsón *v*.; fotólón *v*.; ɨɛtɛ́sá ɨsíítɛ́sʉ́ *v*.; kánɔ́n *v*.; kɨlíwítánón *v*.; takánón *v*. **clear (a surface)** ɨkʉlɛs *v*.; ɨkwalɛs *v*.; ɨlɨwɛs *v*. **clear (grass)** íɛ́s *v*.; ireɲes *v*. **clear (of a path)** kɔlánétòn *v*. **clear (of many)** ɓets'aakón *v*.
**clear (of mind)** bótsóna iká<sup>e</sup> *v*. **clear (of weather)** tatsɔ́ɔ́n *v*. **clear (the throat)** hákátòn *v*.; xaƙarés *v*. **clear (visually)** isólólòòn *v*. **clear a path in** ɨɗíɗíwɛ́s *v*. **clear a path through** utés *v*.; utésúƙot<sup>a</sup> *v*. **clear away (grass)** iréɲésuƙot<sup>a</sup> *v*. **clear off (a surface)** ɨkʉ́lɛ́sʉƙɔt<sup>a</sup> *v*.; ɨlɨwílíwɛtɛ́s *v*. **clear off (brush)** iɓúɓúés *v*. **clear off (visually)** isólólòètòn *v*. **clear off/up** tatsɛ́ɛ́tɔ̀n *n*. **clear out** bulútésuƙot<sup>a</sup> *v*. **clear out (brush)** iɓúɓúés *v*. **clear up (a surface)** ɨkʉlɛtɛ́s *v*. **clear up (of sickness)** ɨɛ́ɓɔ́nʉƙɔt<sup>a</sup> *v*. **clear up (visually)** isólólòètòn *v*. **clearable area** kawam *n*. **cleared (of grass)** íɔ́s *v*. **cleared forest** tsɛ̀f *n*. **cleared off** ɨlɨwílíwɔ́s *v*.; topwatímétòn *v*. **clearing** ɓɔɗ<sup>a</sup> *n*.; ɗípɔ̀ *n*. **cleft (rock)** tsarátán *n*. **cleft palate** akáts'ɛ́a na pakós *n*. **cleg** ɲɔ́pɔɗɔkʉ́ *n*. **cleg (black)** lɔŋɨzɛt<sup>a</sup> *n*. **clench** imúnúkukúón *v*.; ɨtɔkɔɗɛs *v*.; mʉkʉtɛs *v*.; wɨɗɨɗánón *v*.; wíɗíwɨɗánón *v*. **clench buttocks** iíɗón *v*. **clench teeth** ƙídzìkɔ̀ɔ̀n *v*. **clench up** mʉkʉtɛtɛ́s *v*. **clever** nɔɔsánón *v*. **clever person** nɔɔsáàm *n*.
**cleverness** nɔɔ́s *n*. **click disapprovingly** ɨtswɛtítswɛ́tɔ̀n *v*. **cliff chat (bird)** akɔ́ŋɨkɔŋ *n*. **clifflike** fuʝúlúmòn *v*. **cliffy** fuʝúlúmòn *v*. **climb** otsés *v*.; tóbìrìbìròn *v*.; totírón *v*. **climb down (that way)** kídzimonuƙot<sup>a</sup> *v*. **climb down (this way)** kídzìmètòn *v*.; tsíɡìmètòn *v*. **climb on** otsésúƙot<sup>a</sup> *v*. **climb up** otsésúƙot<sup>a</sup> *v*. **climb with gear** itséƙéés *v*.; itséƙóòn *v*. **clinch** rɨɗɛ́s *v*.; wɨɗɨɗánón *v*.; wíɗíwɨɗánón *v*. **cling** ɗɛlɛ́mɔ́n *v*.; ƙídzɔ̀n *v*. **cling to** ɨnɔtsɛs *v*.; ɨrʉmɛs *v*.; ŋɔtsɛ́s *v*. **clinic** ɗakɨtár *n*. **clink against** íɡatsɨɡatsɛ́s *v*. **clink!** ƙwíl *ideo*. **clip** ɨrɛɓɛs *v*. **clip (of a gun)** ɲókópo *n*. **clip off** ɨrɛ́ɓɛ́sʉƙɔt<sup>a</sup> *v*. **clique** kábùn *n*. **clitoris** sɔn *n*. **clitoris head** sɔníík<sup>a</sup> *n*. **clitoris tip** sɔníík<sup>a</sup> *n*. **cloak** ɲáléso *n*.; ɲáwáro *n*. **cloak (leather)** xɔŋɔŋ *n*. **clobber** ipukúpúkés *v*. **clock** fet<sup>a</sup> *n*. **clogged** firímón *v*. **close** ɦyɔtɔ́ɡɔ̀n *v*.; ikóóbés *v*.; ikóóbetés *v*.; kɔkɛ́s *v*.; kɔkɛtɛ́s *v*.; kokímétòn *v*.; mʉts'ʉtɛs *v*. **close (make)** kɔkɨtɛtɛ́s *v*. **close by** ɦyàtàk<sup>a</sup> *n*. **close the eyes** múɗúkánón *v*.
coiled up
**close to each other** ɦyɔtɔ́ɡɨmɔ́s *v*. **close up** mʉts'ʉ́tɛ́sʉƙɔt<sup>a</sup> *v*. **close up repeatedly** muts'utiesúƙot<sup>a</sup> *v*. **closing prayer** wáána na tɛ́zɛ̀tɔ̀nì *n*. **clot** iɗíkétòn *v*.; iɗikitetés *v*. **cloth (small)** ƙwàzàìm *n*. **cloth (waist-)** riɗiesíƙwàz *n*. **cloth(es)** ƙwàz *n*. **clothing** ƙwàz *n*. **clothing (leather)** ínóƙwàz *n*. **clothing (sheep-leather)** ɗóɗòƙwàz *n*. **clotted** iɗíkón *v*. **cloud** ɡìd<sup>a</sup> *n*.; ɲáɗís *n*. **cloud (dark)** ɡida ná buɗám *n*. **cloud (white)** ɡìdà nà ɓèts'<sup>a</sup> *n*. **cloud cover** kùp<sup>a</sup> *n*. **cloud over/up** ɡobétón *v*.; kupétón *v*.; kupukúpón *v*.; mɔƙɨmɔ́ƙɔ́n *v*. **cloudiness** kùp<sup>a</sup> *n*.; kùpààƙw<sup>a</sup> *n*. **cloudless** kánɔ́n *v*. **cloudy** kùpòn *v*. **club** sɛ̀w<sup>a</sup> *n*. **club (group)** ɲéɡurúf *n*. **clubfooted** ƙɔɔlɔ́mɔ̀nà dɛ̀ìkà<sup>ɛ</sup> *v*.; sɔƙɔ́ɲɔ́mɔ̀n *v*. **clubhanded** sɔƙɔ́ɲɔ́mɔ̀n *v*. **cluck at** ɨƙɛníƙɛ́nɛ́s *v*. **cluck disapprovingly** ɨtswɛtítswɛ́tɔ̀n *v*. **cluck!** ƙút<sup>u</sup> *ideo*. **clump** ɓòtòŋ *n*. **clumsy** ɨɓaŋíɓáŋɔ̀n *v*.; pɔsɔ́kɔ́mɔ̀n *v*. **clunky** pɔsɔ́kɔ́mɔ̀n *v*. **cluster** ɓòtòŋ *n*. **co- (my)** ɲ́citaŋá *n*. **co- (plural)** taŋáíkìn *pro*.
**co- (singular)** taŋɛ́ɛ̂d <sup>a</sup> *pro*. **co- (your)** bitáŋá *n*. **co-op** ɲéɡurúf *n*. **co-wife** ɛán *n*. **coagulate** iɗíkétòn *v*.; iɗikitetés *v*. **coagulated** iɗíkón *v*. **coal** bʉbʉn *n*. **coarse** ɡwɛrɛ́ʝɛ́ʝɔ̀n *v*. **coat** ɲókóti *n*. **coax** ɨmámɛ́ɛ́s *v*. **coax into coming** ɨmámɛ́ɛtɛ́s *v*. **cobra** loúpal *n*. **cobweb** abûb<sup>a</sup> *n*. **cock** ɡwácúrúk<sup>a</sup> *n*. **cock (weapon)** wakés *v*. **cocking lever** wakésíàw<sup>a</sup> *n*. **cockle-doodle-doo!** kʉkʉ́ʉ́kᶶ *ideo*. **cockroach** lɔmɛ́ʝɛ́kɛlɛ́*n*. **cocky** ɨwɔ́ƙɔ́n *v*. **cocky person** ɨwɔ́ƙɔ́nìàm *n*. **cocoon** mɔɗɔ́ɗ <sup>a</sup> *n*. **cocoon opening** mɔɗɔ́ɗɔ́èkw<sup>a</sup> *n*. **coerce** rɛ́ɛ́s *v*.; rɛɛtɛ́s *v*.; tɔrɛɛs *v*. **coerced** toreimétòn *v*. **coexistence (peaceful)** zɛƙwa ná dà *n*. **coffee** ɲákáwa *n*. **cohort** taŋɛ́ɛ̂d <sup>a</sup> *pro*. **cohort (your)** bitáŋá *n*. **cohorts** taŋáíkìn *pro*. **coil** ilúƙúretés *v*.; ɨnɔɛs *v*. **coil around** ɨnɔɛtɛ́s *v*. **coil up** ilúƙúrètòn *v*. **coiled** iyérón *v*. **coiled loosely** ŋarúxánòn *v*. **coiled up** ilúƙúròn *v*.
**coin** kaúdzèèkw<sup>a</sup> *n*.; ŋárɔpɨyéékw<sup>a</sup> *n*.; ɲésimón *n*.; ɲókóìn *n*. **coinhabit** ínínós *v*. **cold** ɨɛ́ɓɔ́n *v*. **cold (become)** ɨɛ́ɓɛ́tɔ̀n *v*.; ɨɛ́ɓɔ́nʉƙɔt<sup>a</sup> *v*. **cold (make)** ɨɛ́ɓítɛ́sʉƙɔt<sup>a</sup> *v*. **cold (virus)** ɲarʉ́kʉ́m *n*. **cold weather** ɨɛ́ɓɔ́na kíʝá<sup>e</sup> *v*. **collaborate on** ɨŋáŋárɛtɛ́s *v*. **collapse** badonuƙot<sup>a</sup> *v*.; laʝámétòn *v*.; ɲalámétòn *v*.; ɲalámónuƙot<sup>a</sup> *v*.; ruɓétón *v*.; ruɓonuƙot<sup>a</sup> *v*.; taɲáléetésá así *v*.; taɲálóòn *v*. **collapse due to weight** xuƙúmétòn *v*. **collapsed** paɗókómòn *v*. **collar** ɦyʉƙʉma ƙwázà<sup>e</sup> *n*.; ikeda ƙwázà<sup>e</sup> *n*. **collar (animal)** rɔɓ<sup>a</sup> *n*. **collarbone** ɲálaƙamááìtìɔ̀k <sup>a</sup> *n*. **colleague** taŋɛ́ɛ̂d <sup>a</sup> *pro*. **colleague (my)** ɲ́citaŋá *n*. **colleague (your)** bitáŋá *n*. **colleagues** taŋáíkìn *pro*. **collect** ikóóbés *v*.; ikóóbetés *v*.; ɨrírɛ́ɛ́s *v*.; ɨrírɛ́ɛ́sʉƙɔt<sup>a</sup> *v*.; ɨtsʉnɛs *v*.; ɨtsʉnɛtɛ́s *v*.; ɨʉɗɛs *v*.; ɨʉɗɛtɛ́s *v*. **collect (contributions)** bɔsɛtɛ́s *v*. **collect a debt** amʉtsanés *v*. **collect firewood** weesá dakwí *v*. **collect rubbish** ɲaɗaɗés *v*. **college** ɲésukúl *n*.; yunivásìtì *n*. **colloquium** ɲésémìnà *n*. **colonize** ínésuƙot<sup>a</sup> *v*. **color** ŋɔr *n*. **colored soil** ɲálámʉɲɛna *n*. **Colossians (biblical)** Ŋíkolosáik<sup>a</sup> *n*.
**colossus** kébàdà *n*.; nábàdà *n*.; nébàdà *n*. **colostrum** ɲóɗós *n*. **coma (be in a)** bàdòn *v*. **comatose** ifáfúkós *v*. **comb** ɨƙwɛrɛs *v*.; ɲɛkɛ́sɛ́t <sup>a</sup> *n*. **comb out** ɨƙwɛ́rɛ́sʉƙɔt<sup>a</sup> *v*. **combatant** cɛmáám *n*. **combative** cɛmɨcɛmɔs *v*. **combed** ɨƙwɛrɔs *v*. **combine** ɗɔtsɛ́s *v*.; ɗɔtsɛ́sʉ́ƙɔt<sup>a</sup> *v*.; ɗɔtsɛtɛ́s *v*.; iɗyates *v*.; iɲales *v*. **combine (grains)** ikáɗóés *v*. **Combretium species** tʉ̀tʉ̀f *n*. **come** àtsòn *v*.; ɨnapɛtɛ́sá así *v*. **come (get to)** ɨsʉ́ŋʉ́rɛ́s *v*. **come across** bunétón *v*.; bùnòn *v*.; ɨtsɔŋɛtɛ́s *v*.; ŋawɨlɛs *v*.; taƙámón *v*. **come after** elánétòn *v*. **come alongside** nápɔ́n *v*. **come and go in bunches** ts'ʉ̀wɔ̀n *v*. **come apart** ɗɛsɛ́ɗɛ́sánón *v*.; ɗusúmón *v*. **come around** ɨlɔ́ɗɛ́tɔ̀n *v*.; irímétòn *v*. **come back** iɓóɓóŋètòn *v*.; itétón *v*.; tɔrʉ́ɓɔ́n *v*. **come back around** ɨƙʉlʉ́ƙʉ́lɔ̀n *v*. **come back to life** ɦyekétón *v*. **come beside** nápɔ́n *v*. **come by** ɨɛ́bɛtɛ́s *v*. **come by way of** tɔmɛɛtɛ́s *v*. **come close** ɦyɔtɔ́ɡɛ̀tɔ̀n *v*. **come down** kídzìmètòn *v*.; tsíɡìmètòn *v*. **come early** isókétòn *v*. **come edging** itsoɗiétòn *v*. **come for a visit** ɨlɛ́ɛ́tɔ̀n *v*. **come free** hoɗómón *v*. **come in (of teeth)** morétón *v*.
concave
**come in convoy** ɨtílɛ́tɔ̀n *v*. **come in procession** ɨtílɛ́tɔ̀n *v*. **come inching** itsoɗiétòn *v*. **come into view** lɛlɛmánétòn *v*.; lɛlɛ́tɔ́n *v*.; pɛlɛ́mɛ́tɔ̀n *v*. **come late** íbànètòn *v*.; irípétòn *v*. **come of age** iríétòn *v*. **come off** hoɗómón *v*.; tolómétòn *v*. **come out** pulúmétòn *v*.; tolómétòn *v*. **come out (of stars)** ʝʉ́ɛ́tɔ̀n *v*. **come quickly** ikóméètòn *v*. **come successively** torópétòn *v*. **come to (wake)** tsídzètòn *v*. **come to a consensus** bɛrɛtɛ́sá mɛná<sup>ɛ</sup> *v*. **come to a stop** wasɔnʉƙɔt<sup>a</sup> *v*. **come together** ikóóbetésá así *v*. **come undone** hoɗómón *v*. **come uninvitedly** imíŋóòn *v*. **come via** tɔmɛɛs *v*.; tɔmɛɛtɛ́s *v*. **come with** elánétòn *v*. **coming into view** lɛlɛmánón *v*. **commander** ɨtsɨkɛsíám *n*. **commandment (divine)** ɨtsɨkɛsa Ɲákuʝí *v*. **Commelina species** ɡùʝ<sup>a</sup> *n*. **commence** iséétòn *v*.; isóón *v*.; itsyákétòn *v*. **commerce** dzîɡw<sup>a</sup> *n*. **commerce (do)** dzíɡwès *v*. **comminge** itsulútsúlés *v*. **commingle** ídulés *v*.; íduludulés *v*.; ɨtsɔɓítsɔ́ɓɛ́s *v*. **commingled** ɨtsɔɓítsɔ́ɓɔ̀n *v*.; ɨtsɔɓítsɔ́ɓɔ́s *v*. **Commiphora africana** ibét<sup>a</sup> *n*. **Commiphora campestris** lɔ́lɔwí *n*.
**Commiphora species** tʉlárɔ́y <sup>a</sup> *n*. **commit adultery** ɓúƙón *v*. **commit oneself** ɨmíɗítsɛ́sa así *v*. **commix** ídulés *v*.; íduludulés *v*. **commodities** kúrúɓáicík<sup>a</sup> *n*. **commodity** dzíɡwam *n*.; dzíɡwetam *n*.; dzííƙotam *n*. **common** tɔɔsɛ́tɔ̀n *v*. **Communion** Ŋƙáƙá Komúnió<sup>e</sup> *n*. **community** narúét<sup>a</sup> *n*. **compacted** ɗɔtsánón *v*.; ɗɔtsɔ́s *v*.; dirídòn *v*.; tɔrɔ́dɔ̀n *v*. **compactedly** dìr *ideo*.; tɔ̀r *ideo*. **companion** itumetésíàm *n*. **company** itumetésíàm *n*.; itúmétòn *v*. **company of girls** ɲèrààƙw<sup>a</sup> *n*. **compare** iríánitetés *v*.; ƙámítetés *v*. **compare each other** támínɔ́s *v*. **compel** ɨtɨŋɛs *v*. **compete** ɨkátínɔ́s *v*.; ɨlɔ́ímɔ́s *v*.; ɨlɔ́ínɔ́s *v*. **complain** ɨŋʉrʉ́ŋʉ́rɔ̀n *v*.; ƙɔ̀ɗɔ̀n *v*. **comple** ɨtíŋɛ́ɛ́s *v*. **complete** ŋábɛsʉƙɔt<sup>a</sup> *v*. **completed** ŋábɔnʉƙɔt<sup>a</sup> *v*.; nábɔnʉƙɔt<sup>a</sup> *v*. **completely** ʝɨkî *adv*.; kɔ́nítɨák<sup>e</sup> *pro*.; pílè *ideo*.; tsʉ́tɔ̀ *adv*. **compliant** tsolólómòn *v*. **compose (music)** iroketés *v*. **compound** awááƙw<sup>a</sup> *n*. **comprehend** nesíbes *v*. **compress** ɨrɨɗɛtɛ́s *v*.; rɔɲɛ́s *v*. **compressed** ɨrɨɗɔs *v*.; tɔrɔ́dɔ̀n *v*. **computer** ɲókompyútà *n*. **comrade** taŋɛ́ɛ̂d <sup>a</sup> *pro*. **comrades** taŋáíkìn *pro*. **con** ɨmɔɗɛs *v*. **concave** sakánámòn *v*.; tsuƙúlúmòn *v*.
consume
**concave (flatly)** ɓɛtɛ́lɛ́mɔ̀n *v*.; fɛtɛ́lɛ́mɔ̀n *v*. **conceal** búdès *v*.; búdesuƙot<sup>a</sup> *v*.; ɨɗɛɛs *v*.; ɨɗɛ́ɛ́sʉƙɔt<sup>a</sup> *v*.; ipáŋwéés *v*.; tɨts'ɛ́s *v*.; tɨts'ɛ́sʉ́ƙɔt<sup>a</sup> *v*. **conceal matters** kɔkɛ́sá mɛná<sup>ɛ</sup> *v*. **conceal oneself** budés *v*. **concealed** budésón *v*.; búdòs *v*. **concede** óɡoés *v*. **conceited** itúrón *v*. **concern (topic)** tábès *v*. **concerned** ísánòn *v*. **concerned (become)** ísánonuƙot<sup>a</sup> *v*. **conciliate** ɨsílítɛ́sʉƙɔt<sup>a</sup> *v*. **conclude matters** kɔkɛtɛ́sá mɛná<sup>ɛ</sup> *v*. **conclusion** ɲémíso *n*. **condense (of water)** ikópíòn *v*. **condense (water)** tsɨpɨtsípɔ́n *v*. **cone of tobacco** bɔrɔƙɔƙ<sup>a</sup> *n*.; lɔ́tɔ́ɓabɔrɔƙɔ́ƙ <sup>a</sup> *n*. **confess** tɔkíɔ́n *v*. **confine** ɨƙalíƙálɛ́s *v*.; ɨrɨɗɛs *v*.; ɨrɨɗɛtɛ́s *v*. **confirm** nɨnɛtɛ́s *v*. **confirm the case** enésá mɛná<sup>ɛ</sup> *v*. **confirmation (religious)** ɲékipeyés *n*. **Confirmation (Confirmation)** koferemáásìò *n*. **conflagration** kóméts'àɗ<sup>a</sup> *n*. **conflate** ɨtsɔɓítsɔ́ɓɛ́s *v*.; itsulútsúlés *v*. **conflated** ɨtsɔɓítsɔ́ɓɔ̀n *v*.; ɨtsɔɓítsɔ́ɓɔ́s *v*. **conflict** ɲéƙúruƙur *n*.; ɲɛ́píɗɨpɨɗ<sup>a</sup> *n*. **conflict (cause)** irémóòn *v*. **confound** iƙures *v*. **confounded** iƙúrúmétona iká<sup>e</sup> *v*. **confuse** iƙures *v*.; ilotses *v*.
**confused** iɓíléròn *v*.; ɨcɔ́ŋáimetona iká<sup>e</sup> *v*.; iƙúrúmétona iká<sup>e</sup> *v*.; lɔŋɔanón *v*. **confused (become)** lɔŋɔanónuƙot<sup>a</sup> *v*. **congeal** iɗíkétòn *v*.; iɗikitetés *v*.; tɔsɔ́ɗɔ́kɔ̀n *v*. **congeal when cooled** ɨpʉ́kákòn *v*. **congealed** iɗíkón *v*. **congested (nasally)** ɗɨnɔ́s *v*. **Congo** Kóŋɡò *n*. **congregate** ikóóbetésá así *v*.; ɨrírɛ́ɛtɛ́sá así *v*.; iryámíryámètòn *v*.; itóyéésa así *v*.; ɨtsʉ́nɛ́tɔ̀n *v*.; itukánón *v*. **congregation** ɲatʉ́kɔ́t <sup>a</sup> *n*. **connect** ɗɛsɛ́mɔ́n *v*.; imánétòn *v*.; imánónuƙot<sup>a</sup> *v*.; tɔŋɛtɛs *v*.; tɔŋɛ́tɔ́nʉƙɔt<sup>a</sup> *v*.; toropes *v*.; tɔrʉtsɛs *v*. **conserve** ɨmɨnímínɛ́s *v*. **consider** tamátámatés *v*.; tamɛtɛ́s *v*.; tamítámiés *v*. **consonant** ŋurutiesá tódà<sup>o</sup> *n*. **constant** rítsírɨtsánón *v*. **constellation (five-star)** Ɲerawik<sup>a</sup> *n*.; Tùtùk<sup>a</sup> *n*. **constellation (seven-star)** Lorokonídàkw<sup>a</sup> *n*. **constellation (six-star)** Torikaika Rié *n*. **constellation (Southern Cross)** Ɲémusaláɓà *n*. **constipated** firímón *v*.; tíbɨɗɛ́sɔ́n *v*.; tɨbíɛ́tɔ̀n *v*. **constrained** ɨrɨɗɔs *v*. **constrict** ɨrɨɗɛs *v*.; ɨrɨɗɛtɛ́s *v*.; riɗímétòn *v*. **constricted** ɨrɨɗɔs *v*.; rɔ́ƙɔ́rɔƙánón *v*. **construct** bɛrɛ́s *v*.; toyeetés *v*. **construct (a saying)** taɗápítetés *v*. **consume** ɗáɗítés *v*.; ɨɓalíɓálɛ́s *v*.; ŋƙáƙés *v*.; ŋƙɛ́s *v*.
cored out
**contain** ɨrɨtsɛ́s *v*. **container (big)** kúbùr *n*. **contemn** tsíítés *v*. **contemplate** ɲɛɓɛ́s *v*.; tamátámatés *v*.; tamítámiés *v*. **contemplative** tamɛ́síàm *n*. **contend** ɨlɔ́ímɔ́s *v*.; ɨlɔ́ínɔ́s *v*.; kóríètòn *v*. **contentious** deƙwideƙos *v*. **continue** itemes *v*.; itemetés *v*. **contort** imákóitetés *v*. **contorted** imákóòn *v*. **contorted (become)** imákwéètòn *v*. **contract** riɗímétòn *v*.; tɔ́ɗɔ́nʉƙɔt<sup>a</sup> *v*. **contract (cramp)** ɨtɨɓítíɓɔ̀n *v*. **contract (work)** kabaɗa na teréɡì *n*.; teréɡikabáɗ<sup>a</sup> *n*. **contracted** tɔ́ɗɔ́n *v*. **contradict** déƙwítetés *v*.; nɛpɛƙanitetés *v*. **contribute** tɔ́ƙɛ́s *v*. **contribute resources** ɗɔtsɛ́sá ɦyekesí *v*. **contributions (extract)** bɔsɨtɛtɛ́s *v*. **contuse** rùròn *v*. **contusion** ƙwár *n*. **convene** ikíkóanón *v*.; itukánón *v*. **converge** itóyéésa así *v*. **conversation** ɲékukwá *n*. **converse** ɨɛ́nɔ́n *v*.; tódinós *v*. **convex** tsʉrʉ́ɗʉ́mɔ̀n *v*. **convulse** ɨmímíʝɛ́s *v*. **Conyza species** ɲɛ́ʉrʉmɛmɛ́*n*. **coo (of infants)** ɨŋɨníŋínɔ̀n *v*. **cook** itiŋés *v*.; kɔŋɛ́síàm *n*. **cook by boiling** féés *v*. **cook by stirring** kɔŋɛ́s *v*. **cook for friends** féésa íditíní *v*.
**cook out (poison)** natsɛ́s *v*. **cook quickly** hataikánón *v*. **cook up** ɨɗɔ́ɗɔ́ɛ́s *v*. **cooked well** ɗʉmʉ́dɔ̀n *v*. **cooked medium** tsɛɓɛ́kɛ́mɔ̀n *v*. **cooked tough** haʉ́dɔ̀n *v*. **cooked very tough** hàᶶ *ideo*. **cooker** ɲɛsɨŋƙɨrɨ *n*. **cookie** ɲéɓisikót<sup>a</sup> *n*. **cooking hut** ɲéʝiɡón *n*. **cooking place** itiŋésíàw<sup>a</sup> *n*. **cooking stick** kɔŋɛ́sídàkw<sup>a</sup> *n*. **cooking stone** caál *n*. **cool** cucuéón *v*. **cool down** cucuéétòn *v*.; cucuéítésuƙot<sup>a</sup> *v*.; cucuéónuƙot<sup>a</sup> *v*.; ɗipímón *v*. **cool down (emotionally)** ɨɛ́ɓítɛ́sʉƙɔta ɡúró<sup>e</sup> *v*. **cool down (food)** ɨkáʉ́tɛ́s *v*. **cool down the heart** cucuéítésuƙota ɡúró<sup>e</sup> *v*. **cool off** cucuéétòn *v*.; cucuéónuƙot<sup>a</sup> *v*. **coop (chicken)** ɲɔ́kɔkɔrɔ́hò *n*. **cooperate** ɗɔtsánón *v*.; ɗɔtsánónuƙot<sup>a</sup> *v*. **cooperate on** ɨŋáŋárɛtɛ́s *v*. **coordinate** ilíráitetés *v*. **coordinated** ilíróòn *v*. **cope with** totseres *v*. **coppice (round)** ɲalʉ́kɛ́t <sup>a</sup> *n*. **copulate with** tirés *v*. **copy** iŋitiés *v*.; toputes *v*.; toputetés *v*. **cord** sim *n*. **Cordia sinensis** ɲóɗomé *n*. **cordon-bleu (red-cheeked)** ŋíkaɗiiɗí *n*. **core** ekwed<sup>a</sup> *n*. **core out** iɓóɓórés *v*. **cored out** iɓóɓórós *v*.
covenant
**Corinthians (biblical)** Ŋíkoríntoik<sup>a</sup> *n*. **corn** màlòr *n*.; màlòrìèɗ<sup>a</sup> *n*.; ɲaɓʉra *n*. **corncob** ɲaɓʉraídàkw<sup>a</sup> *n*. **corncobs (leftover)** ɨraɲ *n*. **corner** ɲɔkɔɔna *n*.; rʉ́ɛ́s *v*. **corner of the eye** ɗoɗékw<sup>a</sup> *n*. **corpse** loukúéts'<sup>a</sup> *n*.; ɲɛ́lɛl *n*. **corral** ɓór *n*.; ikoŋetés *v*.; ɨkwɛtíkwɛ́tɛ́s *v*.; ƙalíƙálɛ́s *v*.; sáɡwès *v*. **correct** ɨtsírítɛtɛ́s *v*.; ɨtsírɔ́n *v*.; iyóón *v*.; tsírítɛtɛ́s *v*.; tsírɔ́n *v*.; tɔɓɛɨtɛtɛ́s *v*.; tɔɓɛ́ɔ́n *v*. **correct (typcially)** toɓéíón *v*. **corridor** bácík<sup>a</sup> *n*. **corroborate** nɨnɛtɛ́s *v*. **corrode** iróróòn *v*.; sɨmírɔ́n *v*. **corroded** sɨmɨránón *v*. **corrupt** iƙurúƙúròn *v*.; ɨlʉŋʉ́lʉ́ŋɛ́s *v*.; ɨráŋʉ́nánón *v*. **corruption** ɲéƙúruƙur *n*. **cosmos** kíʝ<sup>a</sup> *n*. **cost** ɲéɓéy<sup>a</sup> *n*. **costlessness** tsàm *n*. **cot** ɲɛ́kítaɗa *n*. **coterie** kábùn *n*. **cotton** ɲápáma *n*. **coucal** múɗuɗú *n*. **cough** ɔ̀f *n*.; ɔ̀fɔ̀n *v*. **councillor** ŋurutiesíama mɛná<sup>ɛ</sup> *n*.; ɲákáńsɔ̀là *n*. **counsel** táŋɛ́s *v*.; taŋɛtɛ́s *v*. **counsel each other** táŋínɔ́s *v*. **count** ɨmaarɛ́s *v*. **counterfeit** láŋ *n*. **country** kíʝ<sup>a</sup> *n*. **county** ɲákáúńtì *n*.
**courageous** itítíŋòn *v*. **courgette** lomuƙe *n*. **courier** dɛáám *n*. **course** isépón *v*.; ɲókós *n*. **court** ɲókót<sup>a</sup> *n*.; sits'és *v*. **court each other** síts'ímós *v*. **court fee** kárátsìk<sup>a</sup> *n*. **courthouse** ɲókót<sup>a</sup> *n*. **cousin (his/her father's brother's child)** babatíím *n*. **cousin (his/her father's brother's son)** leat<sup>a</sup> *n*. **cousin (his/her father's sister's child)** tatatíím *n*. **cousin (his/her mother's brother's child)** momotíím *n*. **cousin (his/her mother's sister's child)** tototíím *n*. **cousin (mother's brother's child)** momóìm *n*. **cousin (mother's sister's child)** totóìm *n*. **cousin (my father's brother's child)** ŋɡóím *n*. **cousin (my father's brother's son)** ɛdɛ́ *n*. **cousin (my father's brother's wife's child)** yáŋììm *n*. **cousin (my father's sister's child)** tátàìm *n*. **cousin (your father's brother's child)** bábòìm *n*. **cousin (your father's brother's son)** léó *n*. **cousin (your father's sister's child)** tátóím *n*. **covenant** iɓoletés *v*.
**cover** ɡubés *v*.; kɔkɛ́s *v*.; kɔkɛtɛ́s *v*.; tɨts'ɛ́s *v*. **cover (a corpse)** sínɛ́s *v*. **cover (an area)** ɨɗɛŋɛs *v*.; ɨkáyɛ́ɛ́s *v*. **cover (an opening)** tʉɓʉnɛ́s *v*. **cover (flat)** ɲápár *n*. **cover (termites)** mɔkɛ́s *v*. **cover oneself** kɔkɛ́sá así *v*. **cover up** bukúrésuƙot<sup>a</sup> *v*.; ɡubésúƙot<sup>a</sup> *v*.; tɨts'ɛ́sʉ́ƙɔt<sup>a</sup> *v*. **cover up issues** kɔkɛ́sá mɛná<sup>ɛ</sup> *v*. **coverall (leather)** kɔ́lɔ́ts<sup>a</sup> *n*. **covered** iɗéŋímètòn *v*. **covered (get)** tuɓunímétòn *v*. **covered in sores** sómomóʝón *v*.; tɔmɔ́tɔ́mánón *v*. **covert** búdòs *v*. **covetous** ts'íts'ɔ́n *v*. **cow** ɦyɔŋwa *n*. **cow (Ankole)** Ŋíɲaŋkóléɦyɔ́*n*. **cow (elephant)** oŋoriŋwa *n*. **cow leather** ɦyɔjejé *n*. **cow milk** ɦyòìdw<sup>a</sup> *n*. **cow udder** ɦyòìdw<sup>a</sup> *n*. **cow urine** tsét<sup>a</sup> *n*. **cow(s)** ɦyɔ̀ *n*. **cow-leather shoe** ɦyɔtaƙáy<sup>a</sup> *n*. **coward** xɛɓásíàm *n*. **cowardice** xɛɓás *n*. **cowardly** ʝàƙw<sup>a</sup> *ideo*.; ʝaƙwádòn *v*.; wúrukukánón *v*.; xɛ̀ɓɔ̀n *v*. **cowbell** ɲákááɗoŋot<sup>a</sup> *n*. **cowdung** ɦyɔ̀èts'<sup>a</sup> *n*.; ŋɔt<sup>a</sup> *n*. **cowfly** lótsóts<sup>a</sup> *n*. **cowherd** còòkààm *n*. **cowhide** ɦyɔjejé *n*.
**cowpea leaves** Icémóríɗókàk<sup>a</sup> *n*. **cowpeas** Icémóríɗ<sup>a</sup> *n*. **cowskin** ɦyɔjejé *n*. **crack** ɓɛlɛ́s *v*.; ɓɛlɛtɛ́s *v*. **crack (react)** tokúétòn *v*.; tokúréètòn *v*. **crack (sound)** ɗɛɗɛanón *v*.; ɨrɔʝírɔ́ʝɛ́s *v*.; rɛɗɛɗánón *v*. **crack apart** itotoles *v*. **crack in pieces** ɨráráƙɛ́s *v*. **crack knuckles** ɨrɔʝírɔ́ʝɛ́sá kɔrɔ́kíkà<sup>ɛ</sup> *v*. **crack open** ɓelémón *v*.; ɨƙɛ́ƙɛ́ɛ́s *v*. **crack open (bones)** ikokes *v*. **crack slightly** beemón *v*. **crack!** ɦyòm *ideo*. **cracked** ɓaaɓánón *v*.; ɓɛlɛ́ɓɛ́lánón *v*.; ɓɛlɔ́s *v*.; médemedánón *v*.; takátákánón *v*. **cracked skin (on feet)** ɲaɓaɓa *n*. **cracker** ɲéɓisikót<sup>a</sup> *n*. **crackle** ɗɛɗɛanón *v*.; rɛɗɛɗánón *v*. **crackly** xà<sup>u</sup> *ideo*. **crackly (in sound)** xaúdòn *v*. **craft** bɛrɛtɛ́s *v*. **craft (a saying)** taɗápítetés *v*. **craftiness** nɔɔ́s *n*. **crafty** nɔɔsánón *v*. **crag** ɲɛ́ɛ́sɛ *n*. **Craibia laurentii** kaûdz<sup>a</sup> *n*. **cram** ɨsɨkɛs *v*.; rʉtsɛ́s *v*.; rʉtsɛ́sʉ́ƙɔt<sup>a</sup> *v*. **cramp** ɨtɨɓítíɓɔ̀n *v*. **cramp (abdominally)** tɔkɔɗíkɔ́ɗɔ̀n *v*. **cramp up** imúnúkukúón *v*. **cranium** ikóɔ́k <sup>a</sup> *n*.; ɔka iká<sup>e</sup> *n*. **crap** nts'áƙón *v*. **crash (sound)** tɔtɔanón *v*. **crash through brush** ɗáɗítésa ríʝá<sup>e</sup> *v*. **Crataeva adansonii** ɲéyoroeté *n*.
cross-eyed
**craving (have a)** ɨrɔ́rɔ́kánón *v*. **craw** ɡwà *n*. **crawl** akúkúròn *v*.; tolíón *v*. **crawl up** ikókórés *v*. **crawl up this way** ikókóretés *v*. **craziness** leɡé *n*.; lejé *n*.; lejéèd<sup>a</sup> *n*.; lejénánès *n*.; ŋkérép<sup>a</sup> *n*. **crazy** iworós *v*. **crazy (go)** doʝánónuƙot<sup>a</sup> *v*. **crazy person** lejéàm *n*. **cream off** ɨʝɨlɛtɛ́s *v*. **crease** zeket<sup>a</sup> *n*. **creased** rʉʝanón *v*.; rʉʝʉrʉ́ʝánón *v*.; turúʝón *v*.; zamʉʝánón *v*. **creasy** turuʝúrúʝánón *v*. **create** ɨɗɨmɛtɛ́s *v*.; iroketés *v*.; tɔsʉɓɛs *v*. **create peace** fítésa kíʝá<sup>e</sup> *v*.; ɨlɔ́tɛ́sʉƙɔta kíʝá<sup>e</sup> *v*. **created** ɨɗɨmɔtɔ́s *v*. **creation** kíʝá na ɨɗɨmɔtɔ́s *n*. **creator** ɨɗɨmɛ́síàm *n*.; ɨɗɨmɛtɛ́síàm *n*.; tɔsʉɓɛtɛ́síàm *n*. **creature** ɨɗɨmɔtɔ́s *n*. **credible** tɔnʉpam *n*. **credit** iɗenes *v*.; iɗenetés *v*.; kál *n*. **creek** ìàwìàw<sup>a</sup> *n*. **creep** aƙʉ́ƙʉ́rɔ̀n *v*.; ƙʉts'áám *n*.; totséɗón *v*. **creep up** íbɛ̀ɗìbɛ̀ɗɔ̀n *v*. **creep up on** tɔlɛ́lɛ́ɛtɛ́s *v*. **creeper** ɲɛ́lɔ́kɨlɔk<sup>a</sup> *n*. **crescent-shaped** toɗóánètòn *v*. **crest (bird)** ɲálem *n*. **crested** tsowírímòn *v*. **crevice (rock)** tsarátán *n*. **crew** ɲáʝore *n*.
**cricket** ʝòrìʝòr *n*. **cricket (armoured)** ɨkɔ́kɔ́yà *n*. **crier** weretsíám *n*. **crime** tɛ́ŋɛ́r *n*. **criminal** ɲɔ́mɔkɔsáàm *n*.; tɛ́ŋɛ́rìàm *n*. **crimp** zeket<sup>a</sup> *n*. **crimson** tsòn *ideo*. **cringe** ŋaxɛ́tɔ́n *v*. **crinkle** zeket<sup>a</sup> *n*. **crinkled** rʉʝanón *v*.; turúʝón *v*.; zamʉʝánón *v*. **crinkly** turuʝúrúʝánón *v*. **cripple** ŋwàxɔ̀nìàm *n*. **crippled** ŋwàxɔ̀n *v*. **crock (soot)** ɲémúɗets<sup>a</sup> *n*.; ɲémúɗuɗu *n*. **crocodile** ɲetíɲáŋ *n*. **crombec (woodland)** natúɗusé *n*. **crook** itúkúɗètòn *v*.; itúkúɗòn *v*.; tukuɗes *v*. **crook (cane)** ɲótooɗó *n*. **crook (climbing)** ƙɔ̀ɗɔ̀t <sup>a</sup> *n*. **crook (staff)** ɲɛ́sɛɛɓɔ́*n*. **crooked** ɡɔ́lɔ́ɡɔlánón *v*.; ríbiribánón *v*.; tukúɗón *v*. **crooked (corrupt)** iƙurúƙúròn *v*. **crooked neck (have a)** lɔkɔɗíkɔ́ɗɔ́n *v*. **crop** ɡwà *n*.; isésélés *v*. **crop up** tʉwɛ́tɔ́n *v*.; tʉ̀wɔ̀n *v*. **crops (nearly ripe)** porór *n*. **cross** kámáránón *v*.; kámáránónuƙot<sup>a</sup> *v*.; ŋʉrɛ́s *v*.; ɲémusaláɓà *n*.; tɔkɛ́ɛ́rɛ́s *v*. **cross over** ɡórés *v*.; íɡorés *v*.; íɡorésúƙot<sup>a</sup> *v*. **cross over a spear** ɡóriesá ɓɨsá<sup>ɛ</sup> *v*. **cross repeatedly** ɡóriés *v*.; íɡoriés *v*. **cross to that side** aronuƙot<sup>a</sup> *v*. **cross to this side** arétón *v*. **cross-eyed** kámáránón *v*.; ríbiribánón *v*.
**crossing** àrònìàw<sup>a</sup> *n*. **crossing (river)** ôd<sup>a</sup> *n*.; ódèèkw<sup>a</sup> *n*. **crossroad** bézèkètìkìn *n*. **Crotalaria lachnocarpoides** ɨɛƙɨɛƙ<sup>a</sup> *n*. **crotch** lɔkɔ́r *n*. **crotch of a tree** bɔ̀k <sup>a</sup> *n*.; bɔ̀kɛ̀d <sup>a</sup> *n*. **Croton dichogamus** ɓólìs *n*. **crouch** dɛ́ɡɛ̀mɔ̀n *v*.; rábʉ̀xɔ̀n *v*.; tsɔ́nɔ́n *v*. **crow** ikwóón *v*. **crow (pied)** kʉ́ràk<sup>a</sup> *n*. **crowbar** ɲotolim *n*. **crowd** itsuɗútsúɗés *v*.; ituɗútúɗés *v*.; ɲéɓúku *n*.; ɲerípírìp<sup>a</sup> *n*.; òdìòs *n*. **crowded together** lolotánón *v*.; tutukánón *v*. **crown (bird)** ɲálem *n*. **crown of head** ikáɡwarí *n*. **crud** ts'âɡ<sup>a</sup> *n*. **cruddy** itútsón *v*.; ts'áɡòn *v*. **crumble** ɗukúmétòn *v*.; ɗukúmón *v*.; ɲalámétòn *v*.; ɲalámónuƙot<sup>a</sup> *v*.; taɲáléetésá así *v*.; taɲálóòn *v*.; tɔpwaɲípwáɲɛ́s *v*. **crumble off** pɛsɛ́mɛ́tɔ̀n *v*. **crumbly** bùɲ *ideo*.; buɲádòn *v*.; ɨwɛlɛ́wɛ́lánón *v*.; pɛsɛ́pɛ́sánón *v*. **crumbly dry** tsaʉ́dɔ̀n *v*. **crumbly substance** búɲèn *n*. **crumbly-dryly** tsàᵘ *ideo*. **crumby** ŋʉmʉ́ŋʉ́mánón *v*. **crumple** laʝámétòn *v*.; tʉmʉɗʉŋɛs *v*. **crumple up** tʉmʉɗʉŋɛtɛ́s *v*. **crunch** ɨʉmʉ́íʉ́mɛ́s *v*. **crunch (food)** iruɓes *v*. **crunch (sound)** ɨrɔʝírɔ́ʝɛ́s *v*. **crunch crunch** ƙɛ̀ƙ ɛ *ideo*.; tsʉ̀bɛ̀ɗ ɛ *ideo*.
**crunchily** xà<sup>u</sup> *ideo*. **crunchy** karʉ́ts'ʉ́mɔ̀n *v*. **crunchy (in sound)** xaúdòn *v*. **crush** iɗoses *v*.; ɨlɛɗɛs *v*.; ɨtsakɛs *v*. **crush (win)** kɔrɨtɛtɛ́s *v*. **crush into powder** itsomes *v*. **crush up** ɨtsakítsákɛ́s *v*. **crust** ɔmɔ́x *n*. **crusted over** rɔ́bɔ̀ɗɔ̀mɔ̀n *v*. **crusty** rɔ́bɔ̀ɗɔ̀mɔ̀n *v*. **cry** ƙɔ̀ɗɔ̀n *v*.; werets<sup>a</sup> *n*. **cry (make)** ƙɔɗɨtɛs *v*. **cry breathily** xíƙwítós *v*. **cry easily** dzálón *v*.; ɨɲɛ́ɛ́mɔ̀n *v*. **cry out** bɔ́rɔ́rɔ̀n *v*.; werétsón *v*.; werétsónuƙot<sup>a</sup> *v*. **crystallize** ɲaʉ́dɔnʉƙɔt<sup>a</sup> *v*. **crystallized** ɲaʉ́dɔ̀n *v*. **crystallizedly** ɲàᶶ *ideo*. **cucumber** kɔlɨl *n*. **cucumber grass** kɔlɨlíkú *n*. **cucumber juice** kɔlɨlícúé *n*. **cucumber seed** kɔlɨlíékw<sup>a</sup> *n*. **Cucumis dipsaceus** tsɔ́ráɗoɗôb<sup>a</sup> *n*. **Cucumis figarei** loɗeɗ<sup>a</sup> *n*. **cultivatable** tɔkɔbam *n*. **cultivate** tɔkɔ́bɛs *v*. **cultivate (make)** tɔkɔ́bɨtɛtɛ́s *v*. **cultivate early** ɨɗɔ́rɛ́ʉƙɔt<sup>a</sup> *v*. **cultivated** tɔkɔ́bɨtɔtɔ́s *v*. **cultivation** tɔ̀kɔ̀b a *v*. **cultivation (early)** ìɗɔ̀r *n*. **cultivator** tɔ̀kɔ̀bààm *n*. **cumbersome** pɔsɔ́kɔ́mɔ̀n *v*. **cuneal** lɨkíɗímɔ̀n *v*. **cunning** nɔɔ́s *n*.; nɔɔsánón *v*. **cup** ɲókópo *n*.
#### cupboard
**cupboard** ɲákábàt<sup>a</sup> *n*. **Cupressus lusitanica** asʉnán *n*. **cure** maraŋítésuƙot<sup>a</sup> *v*.; mínɛ́s *v*. **curl** ilúƙúretés *v*. **curl around** imanímánés *v*. **curl up** ilúƙúrètòn *v*.; tusuketés *v*.; tusúkón *v*. **curl up (to rest)** touríánòn *v*. **curled up** ilúƙúròn *v*. **current issues** kíʝámɛ̀n *n*. **current of air** suɡur *n*. **curse** ébetiés *v*.; ìlàm *n*.; ɨlamɛs *v*.; itsenes *v*.; kɔ́tɛ́s *v*. **curse with a difficult birth** ɨkɛɗɛs *v*. **cursed** ɨlamɔs *v*. **cursed (of a birth)** ɨkɛɗɔs *v*. **curser of natural resources** tɛ́bɛsɨama kíʝá<sup>e</sup> *n*. **curses** ŋítsen *n*. **curtain (doorway)** ɲétiriƙá *n*. **curve** ilúkúɗòn *v*. **curve backward (of horns)** toryóŋón *v*. **curved forward** soƙolánón *v*. **curvy** tukúɗúkuɗánón *v*. **cushion** ɲápalís *n*.; ɲápalís *n*. **cuspid** bàdìàm *n*. **cuss out** ébetiés *v*.; kɔ́tɛ́s *v*. **Cussonia arborea** boxoƙorét<sup>a</sup> *n*. **custom** ɲatal *n*.; ɲeker *n*. **customer** dzíɡwààm *n*.; dzíɡwèsìàm *n*. **cut** ɗɛsɛ́mɔ́n *v*.; ɗusúmón *v*.; ɗusutes *v*.; hoés *v*.; ɨkɛ́ŋɛ́ɗɛ́s *v*.; ŋʉrɛ́s *v*.; ŋʉrɔ́s *v*.; tɔŋɛɗɛs *v*. **cut (of meat)** ɲekiner *n*. **cut (vegetation)** ɨɗɛtɛs *v*.; ɨrɛʝɛs *v*. **cut a circle** tɔlʉkɛtɛ́s *v*.
**cut a ring in** ɨlíŋírɛ́s *v*. **cut across** piɗés *v*. **cut around** ɨlíŋírɛ́s *v*.; ɨlɨrɛs *v*. **cut away** iƙémíƙémés *v*.; ŋʉrɛ́sʉ́ƙɔt<sup>a</sup> *v*. **cut bluntly** ifitífítés *v*. **cut down** ruɓutetés *v*. **cut dully** iŋulúŋúlés *v*. **cut in (verbally)** itoɓes *v*. **cut in chunks** ʝulés *v*. **cut in conversation** itoɓítóɓésa tódà<sup>e</sup> *v*. **cut in pieces** ŋurutiés *v*.; ŋurutiós *v*. **cut in slices** írés *v*.; irikíríkés *v*. **cut in strips** ɨɗɨɗɛs *v*. **cut in two** ɨwáwáɗɛ́s *v*. **cut into** hoetés *v*. **cut noisily** ɨwɔxíwɔ́xɛ́s *v*. **cut off** ŋʉrɛ́sʉ́ƙɔt<sup>a</sup> *v*.; ŋurúmétòn *v*. **cut off (branches)** iteɗes *v*. **cut off (verbally)** iƙofes *v*. **cut off completely** fùt<sup>u</sup> *ideo*. **cut out** ɓilés *v*.; hoetés *v*. **cut out (exlude)** tɔlʉ́kɛ́sʉƙɔt<sup>a</sup> *v*. **cut out respiratory organs** itórópés *v*. **cut short** ŋurúmétòn *v*. **cut through** ɓʉ́nɛ́s *v*.; ɓʉnɛtɛ́s *v*. **cut up** hoetés *v*.; ŋʉrɛtɛ́s *v*.; ŋʉrʉ́ŋʉ́ránón *v*. **cut up in pieces** ɨŋɨlíŋílánón *v*.; ɨŋɨlíŋílɛ́s *v*.; ŋurutiesúƙot<sup>a</sup> *v*. **cut with a blade** kawés *v*. **cuttable** ɓʉnɛtam *n*. **cutting of grass** sɨláxìŋ *n*. **cycle** ɨrɔmɛs *v*. **cylindrical** ɗatáɲámòn *v*. **Cynachium species** ɲákamɔ́ŋɔ *n*. **Cynodon dactylon** mʉrɔn *n*. **Cyperus alternifolius** bùsùbùs *n*.
**Cyperus distans** ɡòmòʝòʝ<sup>a</sup> *n*. **Cyphostemma junceum** ɲɛ́caal *n*. **daily life** zɛƙw<sup>a</sup> *n*. **dam** tábàr *n*.; tɨts'ɛ́s *v*. **dam up** itéƙélés *v*.; tɨts'ɛ́sʉ́ƙɔt<sup>a</sup> *v*.; tits'ímétòn *v*. **damage** kwɛts'ɛ́s *v*. **damaged (get)** kwɛts'ɛ́mɔ́n *v*. **dammed** tɨts'ɔ́s *v*. **damn** ɨlamɛs *v*. **damned** ɨlamɔs *v*. **damnǃ** ɗʉ̀rʉ̀ *n*. **damp** ɗɔ̀ƙɔ̀n *v*. **damp (become)** ɗɔƙɔnʉƙɔt<sup>a</sup> *v*. **dampen** ɗɔƙítɛ́sʉƙɔt<sup>a</sup> *v*. **dance** dikw<sup>a</sup> *n*.; dikwétón *v*. **dance (clapping)** ɲókorot<sup>a</sup> *n*. **dance (ecstatic)** ʝàkàlʉ̀kà *n*. **dance (like to)** dikwidikos *v*. **dance (stork-style)** dikwa na tsokóbè *n*. **dance (style of)** ɲɛ́lɛmá *n*. **dance and sing** óìdìkwòn *v*. **dance at danceground** dikwétóna ŋabɔ́bɔ̀ɔ̀ *v*. **dance hall** dikwáhò *n*. **dance toward** ilépón *v*. **dance with singing** wááka dikwitíní *n*. **dance with stomping** ìpàs *n*. **dance-walk** ɨpásɛ́tɔ̀n *v*. **danceground** ŋabɔ́bɔ̀ *n*. **dancing** ɲaɓolya *n*. **danger** ɡaánàs *n*.; ɲárém *n*. **dangerous** ɡaanón *v*. **dangle** alólóánitetés *v*.; alólóánón *v*.; alólóés *v*. **dapple** ɨtsɔɓítsɔ́ɓɛ́s *v*.
**dappled** ɨtsɔɓítsɔ́ɓɔ̀n *v*.; ɨtsɔɓítsɔ́ɓɔ́s *v*. **dark** buɗámón *v*. **dark (of many)** buɗamaakón *v*. **darken** kaikóón *v*.; kurukúrétòn *v*.; witsiwítsétòn *v*. **darkness** bùɗàm *n*. **darkness (pitch)** ɲákámus *n*. **dart** ɓɨsáím *n*.; ɨrʉtsɛsa así *v*. **dart away** ɨpʉ́mɔ́nʉƙɔt<sup>a</sup> *v*. **dart off** ɨpʉ́mɔ́nʉƙɔt<sup>a</sup> *v*. **dart out** ɨpʉ́mɛ́tɔ̀n *v*. **dash** dzɛ̀rɔ̀n *v*. **dash away** ɨpʉ́mɔ́nʉƙɔt<sup>a</sup> *v*. **dash off** ɨpʉ́mɔ́nʉƙɔt<sup>a</sup> *v*. **dash out** ɨpʉ́mɛ́tɔ̀n *v*. **dassie** kwɨnɨƙ<sup>a</sup> *n*. **date** sits'és *v*. **date (desert)** òwà *n*. **date each other** síts'ímós *v*. **Datura stramonium** lɔmɔ́y <sup>a</sup> *n*. **daub** ɨwarɛs *v*. **daubed** ɨwarɔs *v*. **daughter** ʝàɡw<sup>a</sup> *n*.; ɲàràm *n*. **daughter (his/her)** ʝáɡwèd<sup>a</sup> *n*. **daughter-in-law** bɔƙátín *n*.; imácék<sup>a</sup> *n*. **daughters** ɲèr *n*. **dauntless** itítíŋòn *v*. **dawn** ɓelémón *v*.; ítóna kíʝée ts'ɛɛ *v*.; pɛlɛ́mɔ́na fetí *v*.; tsòòn *v*.; tsoonuƙot<sup>a</sup> *v*.; walámón *v*. **dawn cooly** wéreƙéƙón *v*. **dawn red** ɨpɨrípírɛ̀tɔ̀n *v*. **day** ódòw<sup>a</sup> *n*. **day after tomorrow** kétsóibaráts<sup>a</sup> *n*. **day before yesterday** nótsóò nòk<sup>o</sup> *n*. **daybreak** pɛlɛ́mɔ́na fetí *v*. **daze** ɨrakɛs *v*.; ɨrákɛ́sʉƙɔt<sup>a</sup> *v*.
#### dazed
**dazed** ɨʝárɔ́n *v*.; ʝarámétòn *v*.; tɔmɛrímɛ́rɔ̀n *v*. **deacon** ɨrɨtsɛ́síàm *n*. **dead (almost)** inunúmónuƙot<sup>a</sup> *v*. **dead people** ts'óóniik<sup>a</sup> *n*. **dead person** bàdònìàm *n*. **deaf** ilios *v*.; mìɲɔ̀n *v*. **deaf person** ámá nà mìɲ *n*.; bositíníàm *n*. **deal out** ɨmɔlɛs *v*. **deal with** nɛɛ́s *v*.; nɛɛsʉ́ƙɔt<sup>a</sup> *v*.; totseres *v*. **deal with each other** totsérímós *v*. **death (natural)** badona ná jèjèⁱ *n*. **debate** nɛpɛƙánón *v*. **debile** bʉláʝámɔ̀n *v*.; daƙwádòn *v*. **debt** amʉ́ts<sup>a</sup> *n*.; kál *n*. **debtor** amʉ́tsáàm *n*.; amʉ́tsáàm *n*. **decay** ɗutúɗútánónuƙot<sup>a</sup> *v*.; lolómónuƙot<sup>a</sup> *v*.; masánétòn *v*.; mʉsánétòn *v*. **decayed** ɗatáɗátánón *v*.; ɗutúdòn *v*. **decayed (very)** ɗùt<sup>u</sup> *ideo*. **decaying** ɗutúɗútánón *v*. **decease** ɨríɗɛ́tɔ̀n *v*. **deceased** bàdònìàm *n*.; tás *n*. **deceive** ɨmɔɗɛs *v*. **decelerate** tosipetés *v*. **December** Ìɓùɓù *n*.; Raraan *n*. **decide on** ɲʉmɛtɛ́s *v*. **decked out** naƙwídɔ̀n *v*. **declare** síránòn *v*. **decline** míʝés *v*.; rárímòn *v*. **decompose** ɗutúɗútánónuƙot<sup>a</sup> *v*.; mʉsánétòn *v*. **decomposed** ɗatáɗátánón *v*.; ɗutúdòn *v*. **decomposing** ɗutúɗútánón *v*.
**decorate** isires *v*.; naƙwídɛtɛ́s *v*. **decorate with beads** ɨɗɛrɛs *v*. **decorated** isiros *v*. **decorated with beads** ɨɗɛrɔs *v*. **decoration** ŋásír *n*.; ɲewale *n*. **decrease** rárímetés *v*.; rárímòn *v*. **decrease in number** ƙwaɗonuƙot<sup>a</sup> *v*. **decrease in size** kwátsónuƙot<sup>a</sup> *v*. **decrease size** kwatsítésuƙot<sup>a</sup> *v*. **decreased** bɨráʉ́tɔ̀n *v*. **decrepit** ɡɔ́ɡɔ̀rɔ̀mɔ̀n *v*. **decrepitly** ɡɔ̀ɡɔ̀r *ideo*. **deep** ɓòɓòn *v*. **deep (very)** wòò *ideo*. **deep asleep** nʉsʉ́dɔ̀n *v*. **deepen** ɓoɓonuƙot<sup>a</sup> *v*. **deeper (become)** ɓoɓonuƙot<sup>a</sup> *v*. **deeply** dùù *ideo*. **defame** itúrúmés *v*. **defeat** ɨlɔɛs *v*.; ɨlɔɛtɛ́s *v*.; ipíyéésuƙot<sup>a</sup> *v*.; kurés *v*.; kurésúƙot<sup>a</sup> *v*. **defeated** iloimétòn *v*. **defecate** nts'áƙón *v*. **defecate often** iɓutúɓútòn *v*. **defecation spot** nts'áƙáàw<sup>a</sup> *n*. **defend** cookés *v*.; ɨɛtɛ́s *v*. **defender** còòkààm *n*.; ɨɛtɛ́síàm *n*. **deficient** ɨɗákɔ́n *v*.; taɗatsánón *v*. **deflate** ɓilímón *v*.; tɔ́ɗɔ́nʉƙɔt<sup>a</sup> *v*. **deflated** fɔrɔ́ts'ɔ́mɔ̀n *v*.; tɔ́ɗɔ́n *v*. **deflect** ɨɓatɛs *v*.; ɨƙɨɛs *v*. **deflect repeatedly** ɨɓatíɓátɛ́s *v*. **deformed** rétón *v*. **deformed (of an eye)** ɗooɲómòn *v*. **degree** kêd<sup>a</sup> *n*. **dejected** ɨsɔ́nɛ́sɔ̀n *v*.
**delay** asínón *v*.; inípónítésúƙot<sup>a</sup> *v*.; ɨtíɔ́n *v*.; itúmésuƙot<sup>a</sup> *v*.; titiretés *v*. **delegate** irotes *v*. **deliberate** iyótsóós *v*. **delicacy** ɡwadam *n*.; ɲɛmʉna *n*. **delicate** yɛmɛ́dɔ̀n *v*. **delicate (thin)** bɛɗɛ́dɔ̀n *v*. **delicately** yɛ̀m *ideo*. **delicious** ritídòn *v*. **deliciously** rìtⁱ *ideo*. **delight** ɨlákásítɛ́sʉƙɔt<sup>a</sup> *v*.; ɨmʉ́mʉ́ɨtɛtɛ́s *v*. **deliver a child** ƙwaatetés *v*. **demolish** ipáríés *v*.; ɨtɔ́tɔ́ɛ́s *v*. **demon** ɲɛkípyɛ́*n*. **demon possession** leɡé *n*.; lejé *n*.; lejéèd<sup>a</sup> *n*.; lejénánès *n*.; ŋkérép<sup>a</sup> *n*. **demon-possessed person** lejéàm *n*. **demons** ŋípyɛn *n*. **demonstrate** ɗoɗésúƙot<sup>a</sup> *v*.; itétémés *v*. **den** ak<sup>a</sup> *n*. **denied** rébìmètòn *v*. **denigrate** itúrúmés *v*. **denizen** zɛƙɔ́ám *n*. **denomination (religious)** ɲéɗíni *n*. **dense** rɔ́mɔ́n *v*. **dense (become)** moɡánétòn *v*. **dense (of a thicket)** moɡánón *v*. **dense (of undergrowth)** bòmòn *v*. **dent** luɗés *v*.; rábaɗamitésúƙot<sup>a</sup> *v*. **dented** luɗúmón *v*.; rábàɗàmòn *v*. **dented (get)** rábaɗamonuƙot<sup>a</sup> *v*. **dentifrice** ɲókólíƙèt<sup>a</sup> *n*. **dentist** tolésíàm *n*. **denuded** sɨlɔ́ʝɔ́mɔ̀n *v*. **deny** dimés *v*.; rébès *v*.
**departed (dead)** tás *n*. **depend on** ɨƙɔŋɛs *v*. **dependable** ikékéɲòn *v*. **dependant** ɦyekesíám *n*. **dependence** bɔnánés *n*. **dependent** bɔnán *n*. **depict** iwetés *v*. **depleted** ikarímétòn *v*. **deplore** topóɗón *v*. **deploy** eréɡes *v*. **depraved** ɨráŋʉ́nánón *v*. **depress** luɗés *v*. **depressed** ɨsɔ́nɛ́sɔ̀n *v*.; sìŋòn *v*. **deprive** rébès *v*. **deprived** rébìmètòn *v*. **deracinate** ɗués *v*.; ɗuetés *v*.; rués *v*. **deranged** iworós *v*. **derelict** kɔlɔlánón *v*. **deride with a sucking sound** ts'ʉ́ʉ́tɛ́s *v*. **descend** ɗipímón *v*.; kídzìmòn *v*. **descend (out of sight)** lakámétòn *v*.; lakámón *v*. **descend (that way)** kídzimonuƙot<sup>a</sup> *v*. **descend (this way)** kídzìmètòn *v*.; tsíɡìmètòn *v*. **descend into chaos** iŋóɗyáìmètòn *v*. **desert** ɡóózés *v*.; ɡóózesuƙot<sup>a</sup> *v*.; nalɔŋɨzat<sup>a</sup> *n*. **desert date** tsʉm *n*. **desert date (fruit)** òwà *n*. **desert dated (ripe)** kwílɨlí *n*. **desert rose** dɛrɛ́ƙ <sup>a</sup> *n*.; ʝɔt<sup>a</sup> *n*. **deserted** ɡóózosuƙot<sup>a</sup> *v*. **designs** ŋáƙɨran *n*. **desire** ƙanetés *v*.; wíránés *v*. **desist** bɔlɔnʉƙɔt<sup>a</sup> *v*. **desk** ɲéméza na íƙìrà<sup>ɛ</sup> *n*.
**despise** ɨlɛ́lɛ́ɛ́s *v*.; ts'ábès *v*.; tsíítés *v*.; takaɗes *v*. **despise each other** ɨlɛ́lɛ́ɨmɔ́s *v*.; ts'ábunós *v*. **despoil** taɓales *v*. **dessicated** seƙelánón *v*. **destitute** ikúrúfánón *v*.; iɓúlíánón *v*. **destitute person** iɓúlíánónìàm *n*.; ikúrúfánóníàm *n*. **destroy** ináƙúés *v*.; ináƙúetés *v*.; ipáríés *v*.; ɨtɔ́tɔ́ɛ́s *v*. **destroy violently** ɡaɗɛ́s *v*. **destroyed** ináƙúós *v*.; ináƙúotós *v*. **detachment (military)** ɲɛ́ɗɨtác *n*. **detain** itúmésuƙot<sup>a</sup> *v*. **detect** takanités *v*. **detective** tɨrɨfɛtɛ́síàm *n*. **detergent** ɲásaɓuní *n*. **detergent (laundry)** hómò *n*.; ɲéómò *n*. **determined** ɨkázànòòn *v*.; ɨmʉ́káánón *v*. **detest** ɨlɛ́lɛ́ɛ́s *v*.; ts'ábès *v*.; tʉlʉŋɛs *v*. **detest each other** ts'ábunós *v*. **detour** wɛ́dɔ̀n *v*. **detour widely** ƙeƙérón *v*. **devastate** ɨsílíánɨtɛtɛ́s *v*. **develop** bɛrɛ́s *v*.; zeites *v*.; zeitésuƙot<sup>a</sup> *v*. **developed** ikeimétòn *v*. **deviate** imámáɗós *v*.; itípéés *v*.; iwitités *v*. **devil** Siitán *n*. **devilry** badirét<sup>a</sup> *n*. **devise** iroketés *v*. **devote onself** ɨmíɗítsɛ́sa así *v*. **devotee** mínɛ́sìàm *n*.; túbèsìàm *n*. **devour** ɗáɗítés *v*.; ŋƙáƙésuƙot<sup>a</sup> *v*. **devout person** ɲakuʝíám *n*.
**dew** siƙ<sup>a</sup> *n*. **dewlap** ɓòlìɓòl *n*. **Dhaasanac people** Ŋíɗóŋiro *n*. **diagnose** ɨpɨmɛs *v*. **diaper** ets'íƙwâz *n*. **diaphragm** kàɓ<sup>a</sup> *n*. **diarrhate** harítɔ́n *v*. **diarrhea** hár *n*. **diarrhea (chunky)** hárá na ɡwɛrɛʝɛ́ʝ <sup>a</sup> *n*. **diarrhea (have explosive)** ifulúfúlòn *v*. **diarrhea (have liquidy)** harítɔ́na pɨɔ *v*. **diarrhea (have)** harítɔ́n *v*. **diarrhea (liquid)** hárá na tílɨw<sup>a</sup> *n*. **diarrhea (severe)** hárá ná zè *n*. **diarrheal** dulúmón *v*. **diarrheal mucus** háríɡaɗár *n*. **dice** ŋurutiés *v*. **dice up** ŋurutiesúƙot<sup>a</sup> *v*. **diced** ŋurutiós *v*. **Dichrostachys cinerea** ɡʉ̀r *n*. **dictionary** ɲéɗíkìxònàrì *n*. **did not** máa *adv*. **Didinga language** Ŋííɗɨŋátôd<sup>a</sup> *n*. **Didinga man** kɛ́xɛ́sìàm *n*. **Didinga person** lɔ́tɔ́ɓààm *n*.; Ŋííɗɨŋáám *n*.; ɲémurúŋ́ɡùàm *n*. **die** bàdòn *v*.; badonuƙot<sup>a</sup> *v*.; tɛ́zɛ̀tɔ̀n *v*. **die (make)** badítésuƙot<sup>a</sup> *v*. **die (of many)** ts'óón *v*. **die down** ɗipímón *v*. **die off** bulonuƙot<sup>a</sup> *v*.; raraanón *v*. **die off (of many)** ts'óónuƙot<sup>a</sup> *v*. **die out** bulonuƙot<sup>a</sup> *v*. **die out (of fire)** ts'oonuƙot<sup>a</sup> *v*. **die out (of many)** ts'óónuƙot<sup>a</sup> *v*. **die suddenly** ɨtʉ́lɛ́sʉƙɔta así *v*. **diesel** ceím *n*.
#### differ
**differ** ʝalánónuƙot<sup>a</sup> *v*. **different** ʝaláʝálánón *v*.; ʝalánón *v*. **different (become)** ʝalánónuƙot<sup>a</sup> *v*. **differentiate** ʝalanites *v*. **difficult** imákóòn *v*.; itíónòn *v*. **difficult (become)** imákwéètòn *v*.; kurósúƙot<sup>a</sup> *v*. **difficult (make)** imákóitetés *v*. **difficulty** ŋátíónis *n*. **dig** úɡès *v*. **dig (for water)** tawaɗes *v*. **dig by scratching** ikúkúrés *v*. **dig out** ikuɗúkúɗés *v*.; tukuretés *v*.; tukutetés *v*.; úɡetés *v*. **dig randomly** ɨtsɛɗítsɛ́ɗɛ́s *v*. **dig up** tukures *v*.; tukutes *v*.; úɡetés *v*. **digest** ŋɔ́ɛ́s *v*. **digested** ŋɔ́ɔ́s *v*. **digger** ɲɛ́tɛrɛƙɨtaa na kwɛtá<sup>ɛ</sup> *n*.; tɛ́bɛsɨama ʝʉmwí *n*. **digging stick** fen *n*. **digit** kɔrɔ́k <sup>a</sup> *n*. **digit (number)** ɲánamɓá *n*. **digress** hakonuƙot<sup>a</sup> *v*.; ɨɓátɛ́sʉƙɔta mɛná<sup>ɛ</sup> *v*. **dik-dik (Gunther's)** ɲól *n*. **dik-dik mushroom** ɲólíkɨnám *n*. **dik-dik thornbush** ɲólíkàf *n*. **dilapidated** kɔlɔlánón *v*. **dilate** ɨmɨɗímíɗɛ́s *v*. **diluted** sɨlaɓánón *v*. **dim** kaikóón *v*. **dim (in intellect)** buɗámón *v*. **dim (of eyesight)** buɗámón *v*. **dim (of many)** buɗamaakón *v*. **diminish** rárímetés *v*.; rárímòn *v*.
**ding** rábaɗamitésúƙot<sup>a</sup> *v*. **dinged** rábàɗàmòn *v*. **dinged (get)** rábaɗamonuƙot<sup>a</sup> *v*. **dining area** ŋƙáƙáày<sup>a</sup> *n*. **dinner** ŋƙáƙá na wídzò<sup>e</sup> *n*. **Dioscorea species** ewêd<sup>a</sup> *n*. **Diospyros scabra** ɡodiyw<sup>a</sup> *n*. **dip** ilumes *v*.; ilúmésuƙot<sup>a</sup> *v*.; iúpón *v*. **dip away** iúpónuƙot<sup>a</sup> *v*. **dip down** iúpétòn *v*. **dip into** iɗipes *v*. **dip out (liquid)** ábʉbʉƙɛ́s *v*. **dipper** ƙɔ́r *n*. **dipteran** dililits'<sup>a</sup> *n*. **direct** ɨtsírítɛtɛ́s *v*.; iyoes *v*.; tsírítɛtɛ́s *v*.; tɔɓɛɨtɛtɛ́s *v*.; tɔɓɛ́ɔ́n *v*. **direction** xán *n*. **dirt** ʝʉm *n*. **dirt (red)** boŋórén *n*. **dirt mixed with grain** ɓɔɗáʝʉ́m *n*. **dirt-poor** iɓúlíánón *v*. **dirt-poor person** iɓúlíánónìàm *n*. **dirtiness** ts'âɡ<sup>a</sup> *n*. **dirty** buɗámón *v*.; ɨráŋʉ́nánón *v*.; kaɓúrútsánón *v*.; ŋɔrɔ́ɲɔ́mɔ̀n *v*.; ɲɔŋɔ́rɔ́mɔ̀n *v*.; ts'áɡòn *v*. **dirty (of many)** buɗamaakón *v*. **disability** ŋwaxás *n*. **disabled** ŋwàxɔ̀n *v*. **disabled person** ámá nà ŋwàx *n*.; ŋwàxɔ̀nìàm *n*. **disagree** nɛpɛƙánón *v*. **disappear** buanónuƙot<sup>a</sup> *v*.; ɨíɗɔ́n *v*.; kúbonuƙot<sup>a</sup> *v*.; wɨɗímɔ́nʉƙɔt<sup>a</sup> *v*. **disappear (make)** buanítésuƙot<sup>a</sup> *v*.; iwítésuƙot<sup>a</sup> *v*. **disappeared** buanón *v*. **disc (aluminum)** ɲépéɗe *n*.
#### discard
**discard** hábatsésúƙot<sup>a</sup> *v*.; hábatsetés *v*. **discern** ɦyeités *v*. **discharge** ɓɔ́rítɔ̀n *v*.; ídzès *v*.; ídzesuƙot<sup>a</sup> *v*.; ídzòn *v*.; taŋasɛs *v*. **discharge (mucupurulent)** dɔ̀x *n*. **discharge continuously** ídziidziés *v*. **disciple** dɛáám *n*. **disclose** ilééránitetés *v*. **disco** dikwáhò *n*. **discontent** lɔlɔanón *v*. **discord** ɲéƙúruƙur *n*.; ɲɛ́píɗɨpɨɗ<sup>a</sup> *n*. **discourage** faɗites *v*.; ɨkarɛtɛ́s *v*.; ɨlɔ́ítɛ́sʉƙɔt<sup>a</sup> *v*. **discover** enés *v*.; takanités *v*. **discovery** nòìn *n*. **discriminate** ɨlɔ́ɗíŋɛ́s *v*. **discrimination** ɲoloɗiŋ *n*. **discriminatory** ɨlɔ́ɗíŋánón *v*. **discuss** ɨɛ́nɛ́tɔ̀n *v*.; ɨɛ́nɔ́n *v*. **discuss details** ituetésá tódà<sup>e</sup> *v*. **discuss matters (as a group)** itukanetésá mɛná<sup>ɛ</sup> *v*. **discussion** ɲékukwá *n*. **disdain** ɨlɛ́lɛ́ɛ́s *v*.; tsíítés *v*.; takaɗes *v*. **disease** màyw<sup>a</sup> *n*.; ɲeɗeke *n*. **disease (liver)** ɲeɗekea sakámá<sup>e</sup> *n*. **disease (mild)** ɲeɗekéím *n*. **disease of chest** ɲeɗekea bákútsìkà<sup>e</sup> *n*. **disembowel** ɓilésúƙot<sup>a</sup> *v*. **disgorge** xerétón *v*. **disgrace** iryámítetésá ŋiléétsìk<sup>e</sup> *v*.; ŋilééts<sup>a</sup> *n*. **disgraced (become)** iryámétona ŋiléétsìk<sup>e</sup> *v*. **disgracefulness** ŋiléétsìnànès *n*. **disgust** ɨlɛ́lɛ́ítɛtɛ́s *v*.
**dish out** ɡárés *v*.; ɨmɔlɛs *v*. **dish up** ɡárés *v*. **dish-rack** lɔpɨtáá na ƙófóikó<sup>e</sup> *n*. **dishevel** imóɲíkees *v*.; imóɲíkeetés *v*. **dishonest** yʉanón *v*. **disintegrate** ɗukúmétòn *v*.; ɗukúmón *v*. **disintegrating** ɨɲɨlíɲílánón *v*. **dislocate** toletés *v*. **dislocate (joint)** ƙwɨʝɛ́s *v*. **dislocated** ƙwɨʝímɔ́n *v*. **dislocated (become)** tolómétòn *v*. **dislocated (get)** ɓuumón *v*. **dislodge** iloílóés *v*.; toletés *v*. **dislodged (become)** tolómétòn *v*. **dismantle** ɨʝʉƙʉ́ʝʉ́ƙɛ́s *v*. **disobedience** ɲɛ́sɛ́ƙɔ *n*. **disobedient** ɨsɛ́ƙɔ́ánón *v*. **disobedient person** ɲɛ́sɛ́ƙɔ́ám *n*. **disorderly** lɔŋɔanón *v*. **disorderly (become)** lɔŋɔanónuƙot<sup>a</sup> *v*. **disorientation (topographical)** ŋíjokopí *n*. **dispel** itwares *v*. **dispense** kisanes *v*.; kísés *v*.; tɔkɔrɛs *v*.; tɔkɔ́rɛ́sʉƙɔt<sup>a</sup> *v*.; tɔkɔ́rʉ́kɔ́rɛ́sʉƙɔt<sup>a</sup> *v*. **disperse** alámááránón *v*.; ɓʉnʉ́mɔ́n *v*.; ɓʉnʉtɛ́s *v*.; ɨlámááránón *v*.; itwares *v*.; ɨwɛ́ɛ́lánón *v*.; ɨwɛ́ɛ́lɛ́s *v*.; ɨwɛ́ɛ́lɛ́sʉƙɔt<sup>a</sup> *v*.; ɨwɛ́ɛ́lɛtɛ́s *v*.; kisanes *v*.; kísés *v*.; toɓwaŋes *v*.; tɔkɔrɛs *v*.; tɔkɔ́rɛ́sʉƙɔt<sup>a</sup> *v*.; tɔkɔ́rʉ́kɔ́rɛ́sʉƙɔt<sup>a</sup> *v*. **dispersed** ɨwɛ́ɛ́lɔ́s *v*. **dispersion** kís *n*. **displace** dzuƙés *v*.; ilóʝésuƙot<sup>a</sup> *v*.; ɨlɔ́líɛ́s *v*. **displace away** dzuƙésúƙot<sup>a</sup> *v*. **displace this way** dzuƙetés *v*. **displease** íɡaɗɛ́s *v*.
## disprove
**disprove** ɨsalɛs *v*.; ɨsalɛtɛ́s *v*.; ɨsalɨtɛ́s *v*. **disproven** ɨsálímétòn *v*. **disputation** ɲelerum *n*. **dispute** dèƙw<sup>a</sup> *n*. **dispute (of many)** ilérúmùòn *v*. **disregard** bálábálatés *v*.; balɛ́s *v*.; balɛtɛ́s *v*. **disrespect** tatés *v*. **disrupt** íbʉbʉŋɛ́s *v*.; íbʉbʉŋɛ́sʉ́ƙɔt<sup>a</sup> *v*. **disruptive** rɛ́bɔ̀n *v*. **dissatisfied** lɔlɔanón *v*. **dissatisfy** íɡaɗɛ́s *v*. **dissect** pakés *v*. **dissimilar** ʝalánón *v*. **dissipate** alámááránón *v*.; ɨlámááránón *v*.; ɨwɛ́ɛ́lánón *v*.; ɨwɛ́ɛ́lɛ́sʉƙɔt<sup>a</sup> *v*.; ɨwɛ́ɛ́lɛtɛ́s *v*. **dissipated** ɨwɛ́ɛ́lɔ́s *v*. **dissolve** ɨlɔƙɛs *v*.; laʝámétòn *v*. **dissolve (in mouth)** ɨnʉƙʉ́nʉ́ƙwɛ́s *v*. **dissuade** táŋɛ́s *v*.; taŋɛtɛ́s *v*. **distance** ɨɛƙás *n*.; ɨɛƙítɛ́sʉƙɔt<sup>a</sup> *v*.; zikíbàs *n*. **distance (in miles)** máìrɔ̀ɛ̀d <sup>a</sup> *n*. **distant** ìɛ̀ƙɔ̀n *v*. **distant (slightly)** ɨɛƙíɛ́ƙɔ̀n *v*. **distended** teɓúsúmòn *v*. **distill** tsɨpɨtsípɔ́n *v*. **distinguish** ʝalanites *v*. **distort** imákóitetés *v*. **distort (truth)** isuɗes *v*.; isuɗetés *v*. **distorted** imákóòn *v*.; rétón *v*. **distorted (become)** imákwéètòn *v*. **distract** ɡwelítésuƙot<sup>a</sup> *v*.; hakaikitetés *v*.; itúmúránitésúƙot<sup>a</sup> *v*. **distracted** itúmúránón *v*.
**distractible** dʉmɛ́ɗɛ́mɔ̀n *v*. **distressed (become)** ɨlárímétòn *v*.; ɨlwárímétòn *v*. **distribute** kisanes *v*.; kísés *v*.; tɔkɔrɛs *v*.; tɔkɔ́rɛ́sʉƙɔt<sup>a</sup> *v*.; tɔkɔ́rʉ́kɔ́rɛ́sʉƙɔt<sup>a</sup> *v*. **distribution** kís *n*. **distributor** kísáàm *n*.; kisanesíám *n*. **district** ɲéɗísítùrìk<sup>a</sup> *n*. **District Commissioner** ɗiisí *n*. **disturb** ɨtsanɛs *v*. **disturbed** walɨwálɔ́n *v*. **disunited with each other** teretiinós *v*. **ditch** cúémúcè *n*.; hábatsésúƙot<sup>a</sup> *v*.; hábatsetés *v*. **diverse** ʝaláʝálánón *v*. **diversionary tactic** ɲépípa *n*. **divert** itípéés *v*.; iwitités *v*. **divide** pakés *v*.; taɲáléés *v*.; terés *v*. **divide in two** tɔŋɛ́ɛ́rɛ́s *v*. **divide mathematically** tɔkɔrɛs *v*. **divide out** kisanes *v*.; kísés *v*.; tɔkɔrɛs *v*.; tɔkɔ́rɛ́sʉƙɔt<sup>a</sup> *v*.; tɔkɔ́rʉ́kɔ́rɛ́sʉƙɔt<sup>a</sup> *v*. **divide up** hoetés *v*.; kisanes *v*.; kísés *v*.; taɲáléetés *v*.; terétéránitésúƙot<sup>a</sup> *v*.; tereties *v*.; tɔkɔrɛs *v*.; tɔkɔ́rɛ́sʉƙɔt<sup>a</sup> *v*.; tɔkɔ́rʉ́kɔ́rɛ́sʉƙɔt<sup>a</sup> *v*. **divide up an animal** teretiesá ínó<sup>e</sup> *v*. **divided amongst each other** teretiinós *v*. **divided in pieces** ŋurutiós *v*. **divided up** terétéránón *v*.; teretiós *v*. **divination** faɗás *n*. **divine** fàɗòn *v*. **diviner** fàɗònìàm *n*. **divinity** ɲakuʝínánès *n*. **division (military)** ɲéɗivíxìòn *n*. **division (space)** naƙʉ́lɛ́*n*.
**divorce** hoɗés *v*.; terémón *v*.; terémónuƙot<sup>a</sup> *v*. **divulge** kwɛts'ɛ́s *v*. **divulged** kwɛts'ɛ́mɔ́n *v*. **divvy up** hoetés *v*. **dizziness** taítayó *n*. **dizzy** imáúròn *v*. **do** itíyéés *v*.; itíyéetés *v*. **do again** iɓóŋón *v*.; iɲaƙes *v*.; iɲoƙes *v*.; iɲóƙésuƙot<sup>a</sup> *v*. **do aimlessly** ɨpɛípɛ́ɛ́s *v*. **do away with** ɡʉts'ʉrɛs *v*.; ɡuts'uriés *v*.; itsúrúés *v*. **do business** dzíɡwès *v*.; itsúrútseés *v*. **do faster** ɨmʉ́mʉ́rɛ́s *v*. **do in** ɨtsʉ́tɛ́sʉƙɔt<sup>a</sup> *v*. **do in patches** xʉ́rɛ́s *v*. **do like** iretes *v*. **do not** eʝá *adv*.; máa *adv*. **do partially** ɨlɨŋɛs *v*. **do patchily** xʉ́rɛ́s *v*. **do poorly** ɨtáƙálɛ́s *v*. **do properly** iyomes *v*. **do quickly** ɨɓʉrʉ́ɓʉ́rɔ̀n *v*. **do streakily** ɨlɨŋɛs *v*. **do wrongly** hamʉʝɛ́s *v*. **do(es) not** ńtá *adv*. **do-nothing** ɲakárámɨt<sup>a</sup> *n*. **doable** ikásíetam *v*.; itíyéetam *n*. **doctor** ɗakɨtárìàm *n*. **doctor (animal)** ɗakɨtárɨama ínó<sup>e</sup> *n*. **document** kàbàɗ<sup>a</sup> *n*.; ɲákaratás *n*. **doddering** kamudurudádòn *v*. **doddery** dúnésòn *v*.; itúléròn *v*. **doddery (become)** itúléronuƙot<sup>a</sup> *v*. **dodge** kɨɗɔnʉƙɔt<sup>a</sup> *v*.
**dodge repeatedly** iwitíwítòn *v*. **Dodoth County** Ɗàsòƙ<sup>a</sup> *n*. **Dodoth dialect** Gwáɡwaicétôd<sup>a</sup> *n*. **Dodoth people** Cɔ́ƙɔ́tɔ̀m *n*. **Dodoth person** Cɔ́ƙɔ́tɔ̀mɛ̀àm *n*.; Gwáɡwààm *n*. **doe (goat)** rieŋwa *n*. **dog** ŋók<sup>a</sup> *n*. **dog (female)** ŋókíŋwa *n*. **dog (male)** ŋókícikw<sup>a</sup> *n*. **dog (wild hunting)** tsoe *n*. **dog poop** ŋókíèts'<sup>a</sup> *n*. **dogfly** ŋókítsùts<sup>a</sup> *n*. **dole out** ɨɲíɲínɛ́s *v*. **Dolichos kilimandscharicus** túḿbàb<sup>a</sup> *n*. **Dolichos oliveri** dàlìs *n*. **dollar** ɲɔɗɔ́la *n*. **Dombeya goetzenii** xuxûb<sup>a</sup> *n*. **Dombeya quinqueseta** warɨwar *n*. **domestic animal** ínwá na awá<sup>e</sup> *n*. **domestic violence** ɡaɗ<sup>a</sup> *n*. **domesticate** bɔnɛ́s *v*. **domical** loŋórómòn *v*. **dominate** ɔ́bɛ̀s *v*. **donate** dónés *v*.; dónésuƙot<sup>a</sup> *v*. **donate sporadically** donitiésuƙot<sup>a</sup> *v*. **done** itíyáìmètòn *v*. **done (finished)** ŋábɔnʉƙɔt<sup>a</sup> *v*.; nábɔnʉƙɔt<sup>a</sup> *v*. **done (get)** itíyéetés *v*. **donkey** ɗìɗ<sup>a</sup> *n*. **donkey (female)** ɗiɗeŋwa *n*. **donkey (male)** ɗiɗecúrúk<sup>a</sup> *n*. **donkey (young)** ɗìɗèìm *n*. **donor** dónésìàm *n*. **doomed** ɨlamɔs *v*. **door** àsàk<sup>a</sup> *n*.
**door body** àsàkànèb<sup>a</sup> *n*. **doorframe** ɲɛ́fɨrɛ́m *n*. **doorstep** lòkìtòŋ *n*.; lòrìòŋòn *n*. **doorway** àsàk<sup>a</sup> *n*. **dormouse** ɲáaɲún *n*. **dot** bàsɔ̀n *v*.; iɗolíɗólés *v*. **dotted** iɗolíɗólòn *v*.; merixánón *v*. **double-crosser** tolúónìàm *n*. **doubt** iŋáyéés *v*.; itóŋóòn *v*. **doubt matters** itóŋóiesá mɛná<sup>ɛ</sup> *v*. **doubtful** iƙóƙós *v*. **dough** dʉbam *n*. **doughily** nìr *ideo*. **doughy** nɨrídɔ̀n *v*. **douse** ɨɛ́ɓítɛtɛ́s *v*. **dove** bîb<sup>a</sup> *n*. **dove (red-eyed)** túrúƙuƙú *n*. **down** ɡíɡìròk<sup>e</sup> *n*.; kíʝák<sup>e</sup> *n*.; kɔ́ɔ́ kíʝ<sup>o</sup> *n*.; nɔ́ɔ́kíʝ<sup>o</sup> *n*. **down (drink)** ɨʝíírɛ́sʉƙɔt<sup>a</sup> *v*. **down (food)** lakatiés *v*.; lukutiés *v*. **down (gulp̠ )** itúlákáɲés *v*. **down-striped** ɨƙwɛrɔs *v*. **downcast** iʝúrúròn *v*.; ɨsɔ́nɛ́sɔ̀n *v*. **downcast (of eyes)** iʝúrúʝúròn *v*. **downfall** rumet<sup>a</sup> *n*. **downgrade** rárímetés *v*. **downward** ɡíɡìròk<sup>e</sup> *n*.; iʝúrúròn *v*.; kɔ́ɔ́ kíʝ<sup>o</sup> *n*. **downward (gaze)** iʝúrúʝúròn *v*. **downward place** ɡíɡìr *n*. **doze off** ɨlʉ́zɛ̀tɔ̀n *v*. **draff** ɗʉká *n*.; dàʝ<sup>a</sup> *n*. **drag** fɔ́fɔ́tɛ́s *v*.; ɨfɔɛs *v*. **drag along** béberiés *v*.
**drag away** béberésúƙot<sup>a</sup> *v*. **drag off** béberésúƙot<sup>a</sup> *v*. **drag oneself** ɨfɔɛsa así *v*. **dragonfly** ɗàɗàɡwà *n*. **drain** ídzòn *v*. **drain blood** ts'olítésuƙota séà<sup>e</sup> *v*. **drainage area** ɲɛrɛ́t <sup>a</sup> *n*. **drama** wááka na támɔtɔ́s *n*. **drape** ɨlɔ́ɡɔtsɛ́s *v*. **draw** eminés *v*.; ɨƙɛrɛs *v*.; iwetés *v*. **draw (attract)** tɔmɨnɛs *v*. **draw (signs)** ɨƙɨrɛs *v*. **draw apart** eminiés *v*. **draw away** eminésúƙot<sup>a</sup> *v*. **draw back** ɨsʉ́rʉ́mɔ̀n *v*. **draw near** ɦyɔtɔ́ɡɛ̀tɔ̀n *v*. **draw off** eminésúƙot<sup>a</sup> *v*. **draw out** béberés *v*.; eminetés *v*. **draw out speech** ɨɛ́nítɛtɛ́s *v*.; tóítetés *v*. **draw saliva** ɨmʉʝʉ́mʉ́ʝɔ̀n *v*. **draw up** eminetés *v*. **DRC** Kóŋɡò *n*. **dread** paupáwón *v*. **dream** asínítòn *v*. **dreamer** asínítònìàm *n*. **dregs** ɗʉká *n*.; dàʝ<sup>a</sup> *n*. **drenched** ts'alídòn *v*. **drenchedly** ts'àl *ideo*. **dress** ŋábès *v*.; ɲékiteitéy<sup>a</sup> *n*. **dress up** ŋábitetés *v*.; naƙwídɛtɛ́s *v*. **dressed (get)** ŋábitetés *v*. **dressed up** naƙwídɔ̀n *v*. **dressed up (very)** nàƙw<sup>ɨ</sup> *ideo*. **dresser (fine)** ɲásírìàm *n*.
**dribble** ɨɗɔ́nɔ́n *v*.; ɨmɨlɛtɛ́s *v*.; ɨpɨnípínɔ̀n *v*.; ts'olites *v*.; ts'olítésuƙot<sup>a</sup> *v*.; ts'òlòn *v*.; tɔlɛ́lɛ́ɔ̀n *v*. **dried out** mɔ̀sɔ̀n *v*.; seƙelánón *v*. **dried up** kɔlɔlánón *v*. **drift** ɨlɔ́líɛ́sá así *v*. **drift away** ilélébonuƙot<sup>a</sup> *v*. **drift off** hakonuƙot<sup>a</sup> *v*. **drifter** botibotosíám *n*. **drill** ɨpɨrípírɛ́s *v*.; pulutiés *v*. **drilled** tsàpòn *v*. **driller** pulutiesíàm *n*. **Drimia altissima** bʉlʉbʉlát<sup>a</sup> *n*. **drink** wetam *n*.; wetés *v*. **drink (carbonated)** ɲɔ́sɔ́ɗa *n*. **drink (give)** wetités *v*.; wetitésuƙot<sup>a</sup> *v*. **drink (orange)** ɲɛ́kwɨɲcá *n*. **drink (strong)** kombót<sup>a</sup> *n*.; tule *n*. **drink a lot of** tolepetés *v*. **drink like a cow** bútés *v*. **drink slowly** ɨwɛtɛs *v*. **drink to last drop** ɨʝíírɛ́sʉƙɔt<sup>a</sup> *v*. **drink too much** wɔ̀ɔ̀n *v*. **drink with a straw** ɓíɓítɛ́s *v*. **drinkable** wetam *n*. **drinker** wetésíàm *n*. **drinking straw** ɲálamorú *n*. **drip** ɨƙɨlíƙílɔ̀n *v*.; ɨlɨmɛsa así *v*.; ɨlímɔ́n *v*.; ɨmɨlɛtɛ́s *v*.; itáɓóòn *v*.; ts'òlòn *v*. **drip (of rain)** tatíón *v*. **drip continuously** ɨɗɔ́nɔ́n *v*. **drive** ɨfalífálɛ́s *v*. **drive (a vehicle)** hɔnɛ́s *v*. **drive (animals)** hɔ́nɛ́s *v*. **drive away** hɔnɛ́sʉ́ƙɔt<sup>a</sup> *v*. **drive off** hɔnɛ́sʉ́ƙɔt<sup>a</sup> *v*.
**drive out (here)** hɔnɛtɛ́s *v*. **drive out animals** hɔnɛtɛ́sá ínó<sup>e</sup> *v*. **drive through** ututetés *v*.; xutés *v*.; xutésúƙot<sup>a</sup> *v*. **drive through repeatedly** ututiés *v*. **driven** ɨmʉ́káánón *v*. **driver** ŋíɗɛrɛpáìàm *n*. **driver's license** kabaɗa na hɔnɛ́síɛ̀ kàèè *n*. **driving permit** kabaɗa na hɔnɛ́síɛ̀ kàèè *n*. **drizzle** ɨlɨmílímɔ̀n *v*.; ɨsɛ́ísɛ́ɔ̀n *v*.; itáɓóòn *v*.; ɲɛ́límɨlɨm *n*.; ts'olites *v*.; ts'olítésuƙot<sup>a</sup> *v*.; ts'òlòn *v*. **drongo (fork-tailed)** bɛ̀ʉ̀r *n*.; mɛ̀ʉ̀r *n*. **drool** iʝókón *v*. **droop** ɨƙɔ́nɔ́nɔ̀ɔ̀n *v*. **droop (of eyelids)** irwápón *v*. **droopy** ratatáɲón *v*. **drop** ɗaɗátésuƙot<sup>a</sup> *v*.; ɨmɨlɛtɛ́s *v*.; tɛ́ɛ́tɔ̀n *v*.; tɛɨƙɔ́tɔ̀n *v*.; tɛɨtɛ́sʉƙɔt<sup>a</sup> *v*.; tɛɨtɛtɛ́s *v*.; tuɓutes *v*. **drop down** ɨɗíɔ́n *v*. **drop in numbers** tsakitésúƙot<sup>a</sup> *v*. **drop into** iɗoes *v*. **drop off** rárímòn *v*. **drop-off** látsó *n*. **droppings** ets'<sup>a</sup> *n*. **dropsy** sír *n*. **drought** ɲɔrɔn *n*.; tsóóna kíʝá<sup>e</sup> *v*. **drowsy** ɨlʉ́zɔ̀n *v*.; iyalíyálòn *v*. **drug** cɛ̀mɛ̀r *n*. **drum** iwóts<sup>a</sup> *n*.; ɲéɓur *n*. **drum (large)** ɲépípa *n*. **drum (plastic)** ɲékakúŋ́ɡù *n*. **drum (talking)** ɲɛ́ɗíɨt<sup>a</sup> *n*. **drunk** ɛ́sáàm *n*.; ɛsánón *v*. **drunkard** ɛ́sáàm *n*.
**drunken talk** ɛ́sátòd<sup>a</sup> *n*. **drunkenness** ɛ́s *n*. **dry** tsáítés *v*.; tsóón *v*. **dry (partially)** mɔsímɔ́sɔ̀n *v*. **dry and crumbly** tsaʉ́dɔ̀n *v*. **dry and dusty** puɗádòn *v*. **dry and thick (of grass)** sakátánòn *v*. **dry by cooking** kɛ́xɛ́s *v*. **dry out** lolómónuƙot<sup>a</sup> *v*.; mɔsɔnʉƙɔt<sup>a</sup> *v*.; tsáítésúƙot<sup>a</sup> *v*.; tsóónuƙot<sup>a</sup> *v*. **dry over fire** ɨlílíɛ́s *v*. **dry season** ôdz<sup>a</sup> *n*.; ódzatsóy<sup>a</sup> *n*. **dry season leaves** ódzàkàk<sup>a</sup> *n*. **dry season rain** ódzadidí *n*. **dry up** tsaiƙótòn *v*.; tsáítésúƙot<sup>a</sup> *v*.; tsóónuƙot<sup>a</sup> *v*. **drying rack** lɔɓaɓal *n*. **dryly** ɗàk<sup>a</sup> *ideo*.; wòròƙòƙ<sup>o</sup> *ideo*.; yɛ̀l *ideo*. **dryly as dust** pùɗ<sup>a</sup> *ideo*. **dubious** iƙóƙós *v*. **duck** iúpón *v*.; ɲáɓata *n*.; rʉʝɛ́tɔ́n *v*. **duck away** iúpónuƙot<sup>a</sup> *v*. **duck down** dɛ́ɡɛ̀mɔ̀n *v*.; iúpétòn *v*. **duck up and down** iupúúpòn *v*. **dude** ɲɛ́ɛ́s *n*. **due to** ikóteré *prep*.; kóteré *prep*. **due to (the fact that)** ɗúó *pro*. **due to the fact that** ikóteré *subordconn*.; kóteré *subordconn*. **duel** ɲèùrìà *n*.; ɲeuríétòn *v*. **duiker (common)** ŋamur *n*. **dull** duŋúlúmòn *v*.; iʝíŋáánón *v*.; líídòn *v*.; tufádòn *v*. **dull (boring)** itópénòn *v*.; ʝɔ̀lɔ̀n *v*. **dull (in intellect)** buɗámón *v*.
**dull (of blades)** fitídòn *v*. **dully** lì *ideo*.; tùf *ideo*. **dully (of blades)** fìtⁱ *ideo*. **dumb** ɨɓááŋɔ̀n *v*.; mɨɲɔna íkèdè *v*. **dump** ƙúdès *v*. **dump (garbage)** bɔlɔl *n*. **dump out** iɓutúɓútés *v*.; ixóxóƙés *v*.; ixuƙúxúƙés *v*.; ixúxúƙés *v*.; ƙúdesuƙot<sup>a</sup> *v*.; ƙúdetés *v*. **dump over** bukúrésuƙota así *v*. **dump truck** ƙúdèsìàm *n*.; ɲétípa *n*. **dung** ets'<sup>a</sup> *n*. **dung (cow)** ɦyɔ̀èts'<sup>a</sup> *n*.; ŋɔt<sup>a</sup> *n*. **dunk (baptize)** iɓátíseés *v*. **duplicate** toputes *v*.; toputetés *v*. **duramen** ɡúréda dakwí *n*. **during the day** ódò<sup>o</sup> *n*. **during the evening** wîdz<sup>o</sup> *n*. **during the night** mukú *n*. **during twilight** xɨŋat<sup>ɔ</sup> *n*. **dusk** xɨŋat<sup>a</sup> *n*.; xɨŋatétón *v*. **dusky** míɡiriɡíránón *v*. **dust** bur *n*.; ízuzués *v*.; kabas *n*.; kábàsìn *n*.; xɛɛ́s *v*.; xɛɛsʉ́ƙɔt<sup>a</sup> *v*.; xɛɛtɛ́s *v*. **dust (airborn)** bú *n*. **dust (of chaff)** ŋawíl *n*. **dust bath (of birds)** bùts<sup>a</sup> *n*. **dust cloud** bú *n*. **dust devil** lòtàbùsèn *n*. **dust off** itútúés *v*.; itútúésuƙot<sup>a</sup> *v*. **duster** ɲáɗasɨtá *n*. **dusty-dry** puɗádòn *v*. **duty (work)** ɲetits<sup>a</sup> *n*. **dwarfish** puusúmòn *v*. **dwelling** zɛƙɔ́áw<sup>a</sup> *n*. **dye black** tɛwɛrɛs *v*.
#### dying
**dying** inunúmétòn *v*.; inunúmónuƙot<sup>a</sup> *v*. **dying in numbers** rutukuɲ *ideo*. **dynamite** ɲáɓarút<sup>a</sup> *n*. **dyspepsia** dimésá bubue ŋɔɛsí *n*. **each one** kéŋán *n*.; kíníŋàn *n*.; ŋàn *pro*. **eager** olódòn *v*. **eagle** loƙól *n*. **eagle (long-crested)** kílíkɨlɨká *n*. **ear** bòs *n*. **ear hair** bòsìsìts'<sup>a</sup> *n*. **ear hole** bòsìèkw<sup>a</sup> *n*. **ear infection** ɲɛ́míɗɨmɨɗ<sup>a</sup> *n*. **earler** wàxʉ̀ *n*. **earlier** ŋɔ́rɔ́n *v*. **earlier time** wàx *n*. **earlier today** nák<sup>a</sup> *adv*. **earlobe scar** ɲɛ́tɛ́lɨtɛl *n*. **early** ɛ̀kwɔ̀n *v*. **earn a living** bɛ́ɗɛ́sa ɦyekesí *v*. **earring** ɲɛ́tílɨtɨl *n*. **earth** ʝʉm *n*.; kíʝ<sup>a</sup> *n*. **earth (scorched)** kàròƙ<sup>a</sup> *n*. **earthquake** ɲéríkirik<sup>a</sup> *n*. **earthworm** ídèmèìm *n*. **earwax** ŋókíèts'<sup>a</sup> *n*.; ɲéɗípor *n*. **earwig** natélétsìò *n*. **ease (in place)** ɨɛpɛs *v*. **ease down** ɨɛpɛtɛ́s *v*. **ease off** ɨɛ́pɛ́sʉƙɔta así *v*.; ɨɛ́pɔ́nʉƙɔt<sup>a</sup> *v*. **ease off with** ɨɛ́pɛ́sʉƙɔt<sup>a</sup> *v*. **ease over** ɨɛpɛsa así *v*. **ease up** iŋáléètòn *v*.; ɨwɔ́ɔ́nʉƙɔt<sup>a</sup> *v*. **ease up on someone** ɨɛpɛsa asíɛ́ámák<sup>e</sup> *v*. **east** fetíékw<sup>a</sup> *n*. **easterly direction** fetíékuxan *dem*.
**easterner** fetíékùàm *n*. **easy** batánón *v*.; ɔfɔ́dɔ̀n *v*. **easy (to seduce)** batánón *v*. **easy to manage** olódòn *v*. **eat** ŋƙáƙés *v*.; ŋƙɛ́s *v*. **eat a lot of** tolepetés *v*. **eat all** áts'ɛ́sʉƙɔt<sup>a</sup> *v*. **eat gingerly** ɨláláŋɛ́s *v*. **eat live termites** ƙídzɛsa dáŋá<sup>e</sup> *v*. **eat too much** wɔ̀ɔ̀n *v*. **eat up** áts'ɛ́sʉƙɔt<sup>a</sup> *v*.; ŋƙáƙésuƙot<sup>a</sup> *v*. **eatable** ŋƙam *n*. **eatery** ɲéótèl *n*. **eating** ŋƙáƙ<sup>a</sup> *n*. **eating place** ŋƙáƙáày<sup>a</sup> *n*. **eclipse (lunar)** badona aráɡwaní *n*. **eclipse (solar)** badona fetí *n*. **economize** ɨmɨnímínɛ́s *v*. **ecstasy (be in)** ɨrákɛ́sʉƙɔta así *v*. **ecstatic (become)** ɛfɔnʉƙɔt<sup>a</sup> *v*. **edema** sír *n*. **edge** itsóɗón *v*.; kwayw<sup>a</sup> *n*.; kweed<sup>a</sup> *n*.; ɲimanitésíàw<sup>a</sup> *n*. **edge (cliff)** látsó *n*. **edge over here** itsoɗiétòn *v*. **edges** kwàìn *n*. **edgy** rukurúkón *v*. **edible** ŋƙam *n*. **educate** isómáitetés *v*. **education** ɲósomá *n*. **effloresce** ɲaʉ́dɔnʉƙɔt<sup>a</sup> *v*. **egg** ɓìɓ<sup>a</sup> *n*. **egg on** ɨsʉ́sʉ́ɛ́s *v*.; itsótsóés *v*. **egg yolk** ɗukes *n*. **eggplant** tùlèl *n*. **eggsack** ɓiɓáhò *n*. **eggshell** ɔfɔrɔƙ<sup>a</sup> *n*.
**egocentric** reídòn *v*. **egress** pulúmétòn *v*. **egret** ɲɛɓɔŋ *n*. **Egyptian thorn** kɨlɔ́rít<sup>a</sup> *n*. **eight** tude ńdà kìɗì àɗ<sup>e</sup> *num*. **eight o'clock** ɲásáatɨkaa leɓetsátìk<sup>e</sup> *n*. **eighteen** toomíní ńda kiɗi túde ńdà kìɗì àɗ<sup>e</sup> *n*. **eighty** toomínékwa túde ńdà kìɗì àɗ<sup>e</sup> *n*. **either … or …** mísɨ … mísɨ … *coordconn*. **ejaculate** ƙúdetésá ɗíró<sup>e</sup> *v*. **elaborate** ɨmɨɗímíɗɛ́sa mɛná<sup>ɛ</sup> *v*. **elaborate on** ɨɲɛɛs *v*.; taɲɛɛs *v*. **eland** basaúr *n*. **eland (female)** dzàbùl *n*. **eland (male)** karám *n*. **eland visceral fat** tʉkán *n*. **elbow** ƙʉlɛ́*n*. **elder** dìyòàm *n*.; ʝákám *n*.; ŋurutiesíama tódà<sup>e</sup> *n*. **elder (chief)** ámázeáma awá<sup>e</sup> *n*.; diyoama ná zè *n*. **elder who judges** bɨlamʉ́síàm *n*. **elder's first food** tʉ́w<sup>a</sup> *n*. **elderliness** ʝákámànànès *n*. **elderly** dúnésòn *v*. **elderly (of many)** dunaakón *v*. **elderly (the)** dunaakóniik<sup>a</sup> *n*. **elderly person** dúnésìàm *n*. **elders** ʝák<sup>a</sup> *n*. **elders (obese)** ɨʝʉkɛsííka búbùìkà<sup>e</sup> *n*. **eldership** ʝákámànànès *n*. **eldest child** tíɡàràmàts<sup>a</sup> *n*. **elect** ƙanotós *v*.; tɔsɛɛtɛ́s *v*.; xɔ́bɛtɛ́s *v*. **elect government officials** xɔ́bɛtɛ́sá ɲápukání *v*.
**elected** xɔ́bɔtɔ́s *v*. **elephant** òŋòr *n*. **elephant (female)** oŋoriŋwa *n*. **elephant bull** kílíŋìt<sup>a</sup> *n*. **elephant grass** òŋòrìkù *n*. **elephant gun** ɲáturuɡéy<sup>a</sup> *n*. **elephant trunk** ɡìɡ<sup>a</sup> *n*.; òŋòrìkwɛ̀t <sup>a</sup> *n*. **elephant tusk** òŋòrìkwàyw<sup>a</sup> *n*. **Eleusine coracana** rêb<sup>a</sup> *n*. **elevate** ɓuƙés *v*.; íbokés *v*.; ɨkɛ́ɛ́sʉƙɔt<sup>a</sup> *v*.; ɨkɛɛtɛ́s *v*. **elevate (make)** ɨkɛ́ítɛtɛ́s *v*. **elevated** ɨkɔɔtɔ́s *v*. **eleven** ńda nébèè kɔ̀n *n*.; toomíní ńdà kɛ̀ɗì kɔ̀n *n*. **eleven o'clock** ɲásáatɨkaa tudátìk<sup>e</sup> *n*. **elicit** tɔɓɛ́ɲɛ́tɔ̀n *v*. **elicit speech** ɨɛ́nítɛtɛ́s *v*.; tóítetés *v*. **elliptical** semélémòn *v*. **elude** ɨɓʉtsɛs *v*.; ɨɓʉtsɛtɛ́sá así *v*. **emaciated** iróƙóòn *v*.; itóƙóƙòòn *v*.; lotímálèmòn *v*. **embankment** ɓɔ́kɔ̀ɲ *n*. **embarrass** ɓets'itetés *v*.; iryámítetésá ŋiléétsìk<sup>e</sup> *v*. **embarrassed** kweelémòn *v*. **embarrassed (become)** ɨɛ́ɓɛ́tɔ̀n *v*.; iryámétona ŋiléétsìk<sup>e</sup> *v*. **embellish** daites *v*.; íɡwɨɡwɨʝɛ́s *v*.; isires *v*. **embellished** isiros *v*. **ember** bʉbʉn *n*. **embitter** faɗites *v*. **emblems** ŋáƙɨran *n*. **embodiment** nébùnànès *n*. **embrace** ɨrʉmɛs *v*.; tɔtʉnɛs *v*. **embrace each other** ɨrʉ́mʉ́nɔ́s *v*.; tɔtʉ́nʉ́mɔ́s *v*.
**embrocate** kwírɛ́s *v*.; tsáŋés *v*. **embroider** isires *v*. **embroidered** isiros *v*. **embryo** eɗed<sup>a</sup> *n*. **emerge** lɛlɛmánétòn *v*.; lɛlɛ́tɔ́n *v*.; pɛlɛ́mɛ́tɔ̀n *v*.; pulúmétòn *v*. **emerge (of termites)** ƙídzɔ̀n *v*. **emerging** lɛlɛmánón *v*. **emissary** eréɡesíám *n*. **emit** ɓɔ́rítɔ̀n *v*.; ídzès *v*.; ídzesuƙot<sup>a</sup> *v*.; ídzòn *v*.; itsues *v*.; itsuetés *v*. **emit continuously** ídziidziés *v*. **emit nocturnally (semen)** ƙúdesa ɗíróe así *v*. **employ** ɓuƙítésuƙot<sup>a</sup> *v*.; eréɡes *v*.; ikásíitetés *v*.; teréɡanitetés *v*. **employee** ɲákásìàm *n*.; teréɡìàm *n*. **employer** ámázeáma teréɡì *n*.; ɓuƙítésuƙotíám *n*. **employment** ɲákási *n*.; terêɡ<sup>a</sup> *n*. **empty** bótsón *v*.; bùlòn *v*. **empty by shaking** ɨwɔkɛs *v*. **empty out** bulonuƙot<sup>a</sup> *v*.; bulútésuƙot<sup>a</sup> *v*. **empty sound** fùù *ideo*. **empty-handed** kwɛtíkín<sup>ɔ</sup> *n*.; seát<sup>o</sup> *n*. **empty-handedly** ɡwèlèʝ<sup>e</sup> *ideo*. **emulate** iŋitiés *v*. **emulsify** ɨlɔƙɛs *v*. **en route** múkò *n*. **encampment** ɲákamɓí *n*. **encircle** ɨrɨkɛs *v*.; ɨríkɛ́sʉƙɔt<sup>a</sup> *v*.; ɨrɨkɛtɛ́s *v*.; tamanɛs *v*.; tamanɛtɛ́s *v*. **encircle repeatedly** tamaniés *v*. **enclose** ɨpʉ́pʉ́ŋɛ́s *v*.; ɨpʉ́pʉ́ŋɛtɛ́s *v*. **enclose (termites)** mɔkɛ́s *v*.
**encounter** imánétòn *v*.; imánónuƙot<sup>a</sup> *v*.; ŋimánétòn *v*.; ɲimánétòn *v*.; ɲimánón *v*. **encourage** ɨmʉ́káitetés *v*. **encroach** sáítòn *v*. **encrusted** rɔ́bɔ̀ɗɔ̀mɔ̀n *v*. **encumber** ɨnʉɛs *v*.; isites *v*. **end** ŋurúmétòn *v*.; ɲémíso *n*. **end of the world** tasálétona kíʝá<sup>e</sup> *n*. **ending** ɲémíso *n*. **endure** ɨmʉ́kɔ́ɔ̀n *v*.; nɛɛ́s *v*.; nɛɛsʉ́ƙɔt<sup>a</sup> *v*.; taɗaŋes *v*.; topíánètòn *v*. **enemies** lɔŋɔ́t <sup>a</sup> *n*. **enemy** lɔŋɔ́tɔ́m *n*. **energetic** ɨɓɛ́rɔ́ánón *v*.; ɨkwɨɲíkwíɲɔ́n *v*.; iona ńda sea ni itsúr *v*.; kʉrʉ́kʉ́rɔ́s *v*.; kwɨɲídɔ̀n *v*. **energetically** kwìɲ *ideo*. **enfold** tʉmʉɗʉŋɛs *v*.; tʉmʉɗʉŋɛtɛ́s *v*. **engage** sits'és *v*. **engage each other** síts'ímós *v*. **engagement** ìʉ̀m *n*. **engine** ɡúr *n*. **English language** Ɓets'oniicétôd<sup>a</sup> *n*.; Ŋímusukúìtòd<sup>a</sup> *n*. **enhaloed** itówóòn *v*. **enhance** maraŋités *v*.; maraŋítésuƙot<sup>a</sup> *v*. **enjoy** iyóómètòn *v*.; tsamɛ́s *v*. **enlarge** ɨɲɛɛs *v*.; iwanetés *v*.; iwánétòn *v*.; taɲɛɛs *v*.; zeites *v*.; zeitésuƙot<sup>a</sup> *v*. **enlarge (a hole)** iƙúƙúrés *v*. **enlarge (a small hole)** ɨmɨɗímíɗɛ́s *v*. **enlarge slightly** ɓarɨɓárítɛ́sʉƙɔt<sup>a</sup> *v*.; ɓarɨɓárɔ́nʉƙɔt<sup>a</sup> *v*. **enlist** xɔ́bɛtɛ́sá así *v*. **enmity** lɔŋɔ́tánànès *n*.
**enough** itémón *v*.; ŋábɔnʉƙɔt<sup>a</sup> *v*.; nábɔnʉƙɔt<sup>a</sup> *v*. **enrich** barítɛ́sʉƙɔt<sup>a</sup> *v*. **enroll** xɔ́bɛtɛ́sá así *v*. **ensared (become)** kotsonuƙot<sup>a</sup> *v*. **enshroud** ikáburés *v*. **enslavement** ŋiléɓúìnànès *n*.; ŋɨpɔ́táìnànès *n*. **ensnare** kotsítésuƙot<sup>a</sup> *v*.; sáɡwès *v*. **ensnared** kòtsòn *v*. **entangle** sáɡwès *v*. **entangled** sáɡoanón *v*. **enter** ɓuƙétón *v*.; ɓuƙítésuƙot<sup>a</sup> *v*.; ɓùƙòn *v*.; ɓuƙonuƙot<sup>a</sup> *v*.; ilumetésá así *v*. **enter (make)** ɓuƙítésuƙot<sup>a</sup> *v*. **entertain** fekitetés *v*.; ɨmʉ́mwárés *v*. **entertaining** ɛ̀fɔ̀n *v*. **enthusiastic** olódòn *v*. **entice** ɨmɔɗɛtɛ́s *v*.; ɨsʉ́ŋʉ́rɛ́s *v*. **entire** ɗàŋìɗàŋ *quant*.; mùɲ *quant*.; mùɲùmùɲ *quant*.; tsíɗ<sup>ɨ</sup> *quant*.; tsíɗɨtsíɗ<sup>ɨ</sup> *quant*. **entirely** ɓut<sup>u</sup> *ideo*.; boted<sup>o</sup> *n*. **entities** kúrúɓâd<sup>a</sup> *n*. **entity** kɔ́rɔ́ɓâd<sup>a</sup> *n*. **entourage member** túbesiama ámázeámà<sup>e</sup> *n*. **entrails (inspect)** ɦyeitésá arííkà<sup>ɛ</sup> *v*. **entrance** ak<sup>a</sup> *n*. **entrap** kotsítésuƙot<sup>a</sup> *v*. **entrapped** kòtsòn *v*.; sáɡoanón *v*. **entrapped (become)** kotsonuƙot<sup>a</sup> *v*. **entreat** iƙenes *v*. **enumerate** ɨmaarɛ́s *v*. **envelope** ɲáɓaasá *n*. **envious** ɨrákáánón *v*.
**envious (make oneself)** ɨrakɛsa así *v*. **environment** kíʝ<sup>a</sup> *n*. **envision** asínítòn *v*. **envoy** eréɡesíám *n*. **envy** ɨrákáánás *n*. **enwrap** ikáburés *v*.; tʉmʉɗʉŋɛs *v*.; tʉmʉɗʉŋɛtɛ́s *v*. **Ephesians (biblical)** Ŋíepesóík<sup>a</sup> *n*. **epicardium** máxìŋ *n*. **epistle** ɲáɓáruwa *n*. **equal** iríánòn *v*. **equal (become)** iríánonuƙot<sup>a</sup> *v*. **equal (make)** iríánitetés *v*. **equalize** ikwáánitetés *v*.; iríánitetés *v*. **equate** ikwáánitetés *v*.; iríánitetés *v*.; iríánonuƙot<sup>a</sup> *v*. **eradicate** bulútésuƙot<sup>a</sup> *v*. **Eragrostis braunii** ɦyɔ̀ʝàn *n*. **Eragrostis superba** ɲamaɗaŋíkú *n*. **erase** iɲíɲíés *v*. **eraser** ɲáráɓa *n*. **eraser (chalkboard)** ɲáɗasɨtá *n*. **erect** ɨtsírítɛtɛ́s *v*.; ɨtsírɔ́n *v*.; tsírítɛtɛ́s *v*.; tsírɔ́n *v*.; wasɨtɛs *v*. **erection (have an)** kwídètòn *v*.; kwídòn *v*. **erode** iƙúƙúrés *v*. **erosion** bɔ̀rɔ̀ts<sup>a</sup> *n*.; dìdìàk<sup>a</sup> *n*. **err** hakonuƙot<sup>a</sup> *v*.; tɔsɛ́sɔ́n *v*. **errand-runner** eréɡesíám *n*. **error** ɲasécón *n*.; ɲɔ́mɔkɔsá *n*. **erudite** nɔɔsánón *v*. **erudition** nɔɔ́s *n*. **erupt** ɓilímón *v*.; dulúmón *v*.; toɗúón *v*.; tuƙúmétòn *v*. **erupt (in flames)** lɛ́ʝɛ́tɔ́n *v*. **Erythrina abyssinica** ɨtítí *n*. **escalate** iɗíkétòn *v*.; iɗikitetés *v*.
**escape** iɓutsúmétòn *v*.; iwálílòn *v*.; tɔpɛ́ɔ́n *v*.; ùtòn *v*.; utonuƙot<sup>a</sup> *v*. **escape (and come)** tɔpɛ́ɛ́tɔ̀n *v*. **escape (and go)** tɔpɛ́ɔ́nʉƙɔt<sup>a</sup> *v*. **escarpment** ɓo *n*. **escarpment edge** ɓóák<sup>a</sup> *n*. **escort** iníámésuƙot<sup>a</sup> *v*. **escort (government)** túbesiama ámázeámà<sup>e</sup> *n*. **esophagus** moróká na kwáts<sup>a</sup> *n*. **especially** zùk<sup>u</sup> *adv*. **essence** ekwed<sup>a</sup> *n*. **etchings** ŋáƙɨran *n*. **Ethiopia** Isópìà *n*.; Sópìà *n*. **ethnic group** dìyw<sup>a</sup> *n*.; ɲákaɓɨlá *n*. **Ethur language** Ŋítéɓurítôd<sup>a</sup> *n*. **eucalypt** ɲɛ́pɔrɛ́sìtà *n*. **Eucalyptus species** ɲɛ́pɔrɛ́sìtà *n*. **Euclea schimperi** èmùsìà *n*. **Euphorbia bussei** isókóy<sup>a</sup> *n*. **Euphorbia candelabrum** mʉ̀s *n*. **Euphorbia prostrata** lóɗíkórócɛmɛ́r *n*. **Euphorbia tirucalli** ɨnw<sup>a</sup> *n*. **Europe** Ɓets'oniicékíʝ<sup>a</sup> *n*.; Uláyà *n*. **European** ɓèts'ònìàm *n*.; ɲémúsukit<sup>a</sup> *n*. **European language** Ɓets'oniicétôd<sup>a</sup> *n*.; Ŋímusukúìtòd<sup>a</sup> *n*. **evade** ɨɓʉtsɛs *v*.; ɨɓʉtsɛtɛ́sá así *v*.; iɓutsúmétòn *v*. **evade repeatedly** iwitíwítòn *v*. **evaluate** iniŋes *v*. **evangelist** ɨtátámɛ́síàm *n*. **evangelize** ɨtátámɛ́s *v*. **evaporate** buanónuƙot<sup>a</sup> *v*.; ipúréètòn *v*.; tsaiƙótòn *v*.; tsóónuƙot<sup>a</sup> *v*. **evasive** firifíránón *v*.; wíríwíránón *v*.
**even** ɨkʉlɛs *v*.; ɨkwalɛs *v*.; nááƙwa *n*.; tònì *prep*. **even if** átà *subordconn*.; tònì *subordconn*. **even out** ɨkʉlɛtɛ́s *v*. **evening** wîdz<sup>a</sup> *n*. **evening near sunset** wídzèèkw<sup>a</sup> *n*. **evening time** wîdz<sup>o</sup> *n*. **ever** tsʉ̀tᶶ *adv*. **every person** kéám *n*.; níám *n*. **everywhere** ts'íínúó *n*. **evidence** ɲekísíɓìt<sup>a</sup> *n*. **evident** takánón *v*. **evidently** tsábò *adv*. **evil** ɡaánàs *n*.; ɡaanón *v*. **evil (of many)** ɡaanaakón *v*. **evil eye (having the)** ekúnánès *n*. **evil-eye gazer** ekúám *n*. **eviscerate** ɓilésúƙot<sup>a</sup> *v*. **evoke** tɔɓɛ́ɲɛ́tɔ̀n *v*. **Evolvus alsinoides** ídocɛmɛ́r *n*. **ewe** ɗóɗoŋwa *n*. **exacerbate** rúbès *v*. **exactly** dàn *adv*.; rò *adv*. **exaggerate** ɨmɨɗímíɗɛ́sa mɛná<sup>ɛ</sup> *v*.; tasaɓesa mɛná<sup>ɛ</sup> *v*. **exalt** itúrútés *v*. **examination** ɲɛ́tɛ́sìt<sup>a</sup> *n*. **examination room** ɦyeítésihò *n*. **exasperate** ɨwíwínɛ́s *v*. **excavate** úɡès *v*.; úɡetés *v*. **excavator** ɲɛ́tɛrɛƙɨtaa na kwɛtá<sup>ɛ</sup> *n*.; òŋòrìkwɛ̀t <sup>a</sup> *n*.; tɛ́bɛsɨama ʝʉmwí *n*. **exceed** ɨlɔɛs *v*.; ɨsʉkɛs *v*.; sʉ́kɛ́s *v*. **exchange** ilókótsés *v*.; ɨxɔtsɛs *v*.; xɔ́tsɛ́s *v*. **exchange (money)** toɓwaŋes *v*. **exchange words** bezekanitetésá tódà<sup>e</sup> *v*.; ilókótsésa mɛná<sup>ɛ</sup> *v*.
excise
**excise** ɓilés *v*. **excise (teeth)** tɛŋɛlɛs *v*. **exclude** tɔlʉ́kɛ́sʉƙɔt<sup>a</sup> *v*. **excrement** ets'<sup>a</sup> *n*. **excuse** óɡoés *v*. **excused** óɡoós *v*. **exempt** óɡoés *v*.; óɡoós *v*. **exercise** itétémés *v*. **exercise (the body)** ƙɔƙatés *v*. **exhalation** sʉ̀p <sup>a</sup> *n*. **exhale** sʉ́pɔ́nʉƙɔt<sup>a</sup> *v*. **exhaust** ts'ûd<sup>a</sup> *n*. **exhaust pipe** ts'údemucé *n*. **exhausted** bitsétón *v*.; ikarímétòn *v*.; ziálámòn *v*.; zíkímétòn *v*.; ziláámòn *v*. **exhausted (become)** ziláámètòn *v*. **exhaustion** baƙáík<sup>a</sup> *n*.; ɲɛ́kítɔ́wɔ́*n*. **exhibit** ɲekísíɓìt<sup>a</sup> *n*. **existence** ɦyekes *n*.; zɛƙw<sup>a</sup> *n*. **exit** pulúmón *v*. **exonerate** ɨɛtɛ́sá ɨsíítɛ́sʉ́ *v*. **exorbitant, high in price** itíónòn *v*. **exorcist** kɔ́tɛ́sìàm *n*. **exorcize** hɔnɛ́sʉ́ƙɔt<sup>a</sup> *v*.; kɔ́tɛ́s *v*.; tɔɗɛ́tɛ́sʉƙɔt<sup>a</sup> *v*. **exorcize spirits** kɔ́tɛ́sa súɡùrìkà<sup>e</sup> *v*. **expand** ɨɲɛɛs *v*.; iwanetés *v*.; iwánétòn *v*.; taɲɛɛs *v*.; taɲɛ́ɔ́n *v*.; tʉwɛ́tɔ́n *v*.; tʉ̀wɔ̀n *v*.; xuanón *v*.; xuxuanón *v*. **expand on the issues** ɨmɨɗímíɗɛ́sa mɛná<sup>ɛ</sup> *v*. **expanded** ɨatɔs *v*. **expect** kɔɛ́s *v*.; kɔɛtɛ́s *v*. **expedition** ɲásápari *n*. **expel** ídzès *v*.; ídzesuƙot<sup>a</sup> *v*.; ɨlɔ́líɛ́s *v*. **expel continuously** ídziidziés *v*.
**expend** isítíyeés *v*. **expense** ɲéɓéy<sup>a</sup> *n*. **expensive** itíónòn *v*.; ŋìxɔ̀n *v*. **experience (gain)** nɔɔsánétòn *v*. **expiate** iɓutes *v*. **expire** bitsétón *v*.; ɨríɗɛ́tɔ̀n *v*. **expire (die)** tɛ́zɛ̀tɔ̀n *v*. **explain** enitetés *v*. **explode** ɓilímón *v*.; toɗúón *v*. **exploit** ɨnɔmɛtɛ́s *v*. **expose** ɗoɗésúƙot<sup>a</sup> *v*.; ilééránitetés *v*.; ilététaitetés *v*. **exposed** ilééránón *v*.; ilététòòn *v*.; tuɗúsúmòn *v*. **expound on** ɨɲɛɛs *v*.; taɲɛɛs *v*. **exsufflate** rʉ́tɛ́s *v*. **extend** zikíbitésúƙot<sup>a</sup> *v*. **extend (hands)** taɓɛɛs *v*. **extend up** zikíbitésúƙot<sup>a</sup> *v*.; zikíbonuƙot<sup>a</sup> *v*. **extensor (muscle)** ɲépísikitíém *n*. **exterminate** kánɛ́s *v*. **extinct (go)** bulonuƙot<sup>a</sup> *v*.; kanímétòn *v*. **extinguish** ts'eites *v*.; ts'eítésuƙot<sup>a</sup> *v*. **extirpate** ɗués *v*.; ɗuetés *v*.; rués *v*. **extol** itúrútés *v*.; tamɛɛs *v*. **extort** rɛ́ɛ́s *v*.; rɛɛtɛ́s *v*.; tɔrɛɛs *v*. **extorted** toreimétòn *v*. **extra** ɨɓámɔ́n *v*.; ʝɛʝɛ́tɔ́n *v*. **extract** béberés *v*.; faɗetés *v*.; tɔkɛtɛs *v*.; tɔkɛ́tɛ́sʉƙɔt<sup>a</sup> *v*.; tɔkɛtɛtɛ́s *v*.; tolés *v*.; toletés *v*.; tuɓutetés *v*.; tuutes *v*.; tuutetés *v*. **extract (evil charms)** totues *v*. **extract brideprice** buƙitetés *v*. **extract repeatedly** tolotiés *v*. **extremity (arm-like)** kwɛt<sup>a</sup> *n*. **exude** tɔfɔ́ɗɔ́n *v*.
**exult** iwóŋón *v*.; taɓólón *v*. **exultant** taɓolos *v*. **exuviate** fòlòn *v*. **eye** ekw<sup>a</sup> *n*.; iséméés *v*.; iséméetés *v*. **eye disease** lokiyo *n*. **eye jaundice** xídɔna ekwitíní *v*. **eyeball** ɡɔ̀k <sup>a</sup> *n*. **eyebrow** ekúkɔ́ɔ́kwarósíts'<sup>a</sup> *n*.; ekúsíts'<sup>a</sup> *n*. **eyeglasses** ɲékiyóìk<sup>a</sup> *n*. **eyelash** ekúsíts'<sup>a</sup> *n*. **eyelid** ɡɔrɔ́x *n*. **eyesight (have poor)** múɗúkánón *v*.; ŋwaxɔna ekwitíní *v*. **fable (animal)** emuta ínó<sup>e</sup> *n*. **face** takár *n*.; takárɛ̂d <sup>a</sup> *n*. **face back to back** ƙʉƙʉmánón *v*. **face each other** maímós *v*. **face in one direction** ɨtɛ́ɛ́rɨtɛtɛ́s *v*. **face one direction** ɨtɛ́ɛ́rɔ̀n *v*. **facial incisions (make)** dzeretiés *v*. **fade** buanónuƙot<sup>a</sup> *v*. **fade (of color)** kitsonuƙot<sup>a</sup> *v*.; sɛkɔnʉƙɔt<sup>a</sup> *v*. **Fadigura species** fàɗìɡùr *n*. **fail** ɨlaʝíláʝɛ́s *v*.; iɲétsóòn *v*.; iyétsóòn *v*.; rúmánitésúƙot<sup>a</sup> *v*.; rúmánòn *v*.; totóánonuƙot<sup>a</sup> *v*. **fail (make)** rúmánitésúƙot<sup>a</sup> *v*. **fail to cross paths** bezekánón *v*. **fail to get** fàlòn *v*. **fail to meet** bezekánón *v*. **failure** rumet<sup>a</sup> *n*. **failure of a season** tasálétona kíʝá<sup>e</sup> *n*. **faint** ɨɲʉ́ɲʉ́ánón *v*.; rèŋòn *v*.
**faint-hearted** ʝaƙwádòn *v*.; wúrukukánón *v*. **fair** dòòn *v*. **fairy** kíʝáìm *n*. **faith** ɲanʉ́pít<sup>a</sup> *n*. **faith (religion)** ɲéɗíni *n*. **faith in (have)** tɔnʉpɛs *v*. **fake** ɨtsárʉ́ánón *v*.; láŋ *n*. **falcon** tsìts<sup>a</sup> *n*. **fall** ruɓétón *v*.; ruɓonuƙot<sup>a</sup> *v*.; rúmánòn *v*.; rumet<sup>a</sup> *n*.; tɛ́ɛ́tɔ̀n *v*.; tɛɨƙɔ́tɔ̀n *v*. **fall (make)** rúmánitésúƙot<sup>a</sup> *v*.; tɛɨtɛ́sʉƙɔt<sup>a</sup> *v*.; tɛɨtɛtɛ́s *v*. **fall apart** ɗukúmétòn *v*.; ɗukúmón *v*. **fall back** isíɗéètòn *v*.; maɗámón *v*. **fall behind** isíɗéètòn *v*. **fall continuously in numbers** tsákíètòn *v*. **fall down** laʝámétòn *v*. **fall down (crumble)** ɲalámétòn *v*.; ɲalámónuƙot<sup>a</sup> *v*.; taɲáléetésá así *v*.; taɲálóòn *v*. **fall down repeatedly** téíètòn *v*. **fall for (in love)** isémúreés *v*. **fall from up high** xɛ́mɛ́tɔ̀n *v*.; xɛmɔnʉƙɔt<sup>a</sup> *v*. **fall gently** raraanón *v*. **fall in love with** isémúreés *v*. **fall in numbers** tsakétón *v*. **fall off** tolómétòn *v*. **fall out** tolómétòn *v*. **fall sick** meetón *v*. **fall upon (attack)** toɗóón *v*. **falling apart** ɨɲɨlíɲílánón *v*. **falling sound** hyeaa *ideo*. **falls** látsóìk<sup>a</sup> *n*.
fearfully
**false information** isuɗam *n*. **falsehood** yʉɛ *n*. **falsely accuse each other** tɔrɛ́mʉ́nɔ́s *v*. **falsify** isuɗes *v*.; isuɗetés *v*. **familiarize** naínɛ́ɛtɛ́s *v*.; naítɛ́sʉƙɔt<sup>a</sup> *v*. **familiarize oneself with** ɦyeités *v*. **family** ts'aɗíékw<sup>a</sup> *n*. **famine** ɲɔrɔƙɔ *n*. **famished** ɲɛƙánón *v*. **famous** arútón *v*.; ɦyoós *v*. **fan** ipukes *v*. **fancy** ɨmáráɗàɗòòn *v*. **fang** ƙídzɛ̀sìkwàyw<sup>a</sup> *n*. **fang (snake)** ídèmèkwàyw<sup>a</sup> *n*. **far** ìɛ̀ƙɔ̀n *v*. **far (a bit)** ɦyàƙàtàk<sup>a</sup> *n*. **far (slightly)** ɨɛƙíɛ́ƙɔ̀n *v*. **far from each other** ɨɛ́ƙímɔ́s *v*. **faraway** ɦyàƙàtàk<sup>a</sup> *n*. **farm** tɔkɔ́bɛs *v*. **farmable** tɔkɔbam *n*. **farmer** tɔ̀kɔ̀bààm *n*. **farming** tɔ̀kɔ̀b a *v*. **farming season** tɔkɔbatsóy<sup>a</sup> *n*. **farmland** ɲámanikór *n*. **farness** ɨɛƙás *n*. **fart** fèn *n*.; fenétón *v*. **fashionable** titianón *v*. **fashionista** ɲásírìàm *n*. **fast** ɗàmʉ̀s *adv*.; ɗɛ̀mʉ̀s *adv*.; ɨkáɲɔ́n *v*.; itírónòn *v*.; torónón *v*.; wɛ́ɛ́nɔ̀n *v*. **fasten** iwetés *v*.; zíkɛ́s *v*. **fastened** zíkɔ́s *v*. **fastened tightly** ŋɔ̀tsɔ̀n *v*. **faster (be)** ɨmʉ́mʉ́rɔ̀n *v*.
**faster (do)** ɨmʉ́mʉ́rɛ́s *v*. **fat** ceím *n*.; ɛfás *n*.; zízòn *v*. **fat (congealed)** ɡwadam *n*. **fat (epicardial)** máxìŋ *n*. **fat (intra-abdominal)** sábà *n*. **fat (organ)** sábà *n*. **fat (visceral)** sábà *n*. **fat person** zízònìàm *n*. **fat slug** bɛ́bam *n*. **fat-ass** bɛ́bam *n*. **father (his/her/its)** babat<sup>a</sup> *n*. **father (my)** abáŋ *n*. **father (your)** bábò *n*. **father-in-law (her)** bɔbat<sup>a</sup> *n*. **father-in-law (his)** ntsíémetá *n*. **father-in-law (his/her sibling's spouse's father)** babatíɲót<sup>a</sup> *n*. **father-in-law (my sibling's spouse's father)** abáŋìɲòt<sup>a</sup> *n*. **father-in-law (my, of men)** ɲ́ciemetá *n*. **father-in-law (my, of women)** bɔbá *n*. **father-in-law (of men)** emetá *n*. **father-in-law (your sibling's spouse's father)** báboɲót<sup>a</sup> *n*. **father-in-law (your, of men)** biemetá *n*. **father-in-law (your, of women)** bɔ́bɔ̀ *n*. **fatherhood** babatínánès *n*. **fatherliness** babatínánès *n*. **fatigue (extreme)** ɲɛ́kítɔ́wɔ́*n*. **fatso** bɛ́bam *n*. **fatten** zízités *v*. **fatten up** tuɓútitésúƙot<sup>a</sup> *v*.; tuɓútónuƙot<sup>a</sup> *v*.; zízitésúƙot<sup>a</sup> *v*.; zízonuƙot<sup>a</sup> *v*. **fault** ɲɔ́mɔkɔsá *n*. **fear** mòròn *v*.; xɛɓás *n*. **fearful** ʝaƙwádòn *v*.; wúrukukánón *v*.; xɛ̀ɓɔ̀n *v*. **fearfully** ʝàƙw<sup>a</sup> *ideo*.
#### fearless
**fearless** itítíŋòn *v*. **feasible** ikásíetam *v*.; itíyéetam *n*. **feast** írésiŋƙáƙ<sup>a</sup> *n*. **feather** tùk<sup>a</sup> *n*. **February** Ɓèts'òn *n*.; Lokwaŋ *n*. **feces** ets'<sup>a</sup> *n*. **feces (loose)** eruxam *n*. **feces (Rhus natalensis)** mɨsáíèts'<sup>a</sup> *n*. **feckless** ɨkáláʝaránón *v*. **feeble** dɛrɛ́dɔ̀n *v*.; ɨpáláƙɔ̀n *v*.; ʝuódòn *v*. **feebly** dɛ̀r *ideo*.; ʝù<sup>o</sup> *ideo*. **feed** bɔnɛ́s *v*.; ŋƙáƙítetés *v*.; ŋƙɨtɛtɛ́s *v*. **feed (fire)** iríréés *v*.; iríréetés *v*. **feed nocturnally** aúɡòn *v*. **feeding** ŋƙáƙ<sup>a</sup> *n*. **feel** enés *v*.; ɨmɨnímínɛ́s *v*.; mɨnímínatés *v*.; tábès *v*. **feel around** ɨtapítápɔ̀n *v*. **feel bad (make)** ɓarites *v*.; ɓarítésuƙot<sup>a</sup> *v*. **feel faint** tikítíkona ɡúró<sup>e</sup> *v*. **feel mercy** cucuéétòn *v*. **feel weak** tikítíkona ɡúró<sup>e</sup> *v*. **feign** iɲétsóòn *v*.; iyétsóòn *v*. **fell** ruɓutetés *v*. **fell (trees)** toukes *v*.; touketés *v*. **fellow** ɛbam *n*. **felt (on the skin)** fiifíón *v*. **felted** kémúsánón *v*. **female (animal)** ŋwa *n*. **female (young)** wâz *n*. **femininity** cekínánès *n*. **femur** ɡubesíɔ́k <sup>a</sup> *n*. **fence** iriwes *v*.; mɛtsɛ́s *v*. **fence (outer thorny)** ɲéríwi *n*.
**fence (wooden)** marɨŋ *n*. **feral** ɨsílíánón *v*. **ferment** ɓaronuƙot<sup>a</sup> *v*. **ferocious** ɦyɛtɨɦyɛtɔs *v*.; ɦyɛ̀tɔ̀n *v*. **ferocity** ɦyɛtás *n*. **ferry** iríítés *v*. **fertile** bòmòn *v*.; ɨɔ́kɔ́n *v*. **fertile (of soil)** zízòn *v*. **fertile soil** ʝʉma na zîz *n*. **fertilize (plants)** zízités *v*.; zízitésúƙot<sup>a</sup> *v*. **festering** tatifíánón *v*. **fetch** tɛ́bɛtɛ́s *v*. **fetch (water)** kotsés *v*.; kotsetés *v*. **fetid** ɗɛtsɨɗɛ́tsɔ́n *v*. **fetidly** ɗɛ̀ts<sup>ɛ</sup> *ideo*. **fetus** ɗʉ́r *n*. **fever** hábona nébwì *n*.; suɡur *n*. **fever tree** lɛ̀r *n*. **feverish** titianón *v*. **few** ɨmaarɔ́s *v*.; ƙwàɗ<sup>e</sup> *quant*.; ƙwàɗòn *v*. **fewer** ƙwaɗiƙwáɗón *v*. **fewer (become)** ƙwaɗonuƙot<sup>a</sup> *v*. **fiber** ŋísɨl *n*. **fiber (twiste)** sim *n*. **fibrous** simánón *v*. **fickle at work** ɨpáláƙɔ̀n *v*. **Ficus ingens** namʉ́ɗɨt<sup>a</sup> *n*. **Ficus platyphylla** barat<sup>a</sup> *n*. **Ficus species** náínɛnɛ́ *n*.; ɲákárat<sup>a</sup> *n*.; ɲékerum *n*. **Ficus sycomorus** áts'<sup>a</sup> *n*. **fiddle** ɲakaw<sup>a</sup> *n*. **fiddle with** íɡuʝuɡuʝés *v*. **fidgety** kakáánón *v*.
**field** rɔw<sup>a</sup> *n*. **field (ball)** ɲakwaanʝa *n*. **field (large)** ɲámanikór *n*. **field glasses** ɲáɓaraɓín *n*. **fierce** ɦyɛtɨɦyɛtɔs *v*.; ɦyɛ̀tɔ̀n *v*.; ɨɲɛ́ɛ́mɔ̀n *v*. **fierceness** ɦyɛtás *n*. **fifteen** toomíní ńda kiɗi tûd<sup>e</sup> *n*. **fifth (one)** ɗa túdònì *pro*. **fifty** toomínékwa tûd<sup>e</sup> *n*. **fig tree base** baratídɛ̀ *n*. **fig tree species** áts'<sup>a</sup> *n*.; barat<sup>a</sup> *n*.; náínɛnɛ́ *n*.; namʉ́ɗɨt<sup>a</sup> *n*.; ɲákárat<sup>a</sup> *n*.; ɲékerum *n*. **fight** cɛ̀mɔ̀n *v*.; déƙwítetés *v*.; ɨƙaíƙɛ́ɛ́s *v*.; ɨƙáƙɛ́ɛ́s *v*.; ɨƙáƙɛ́ɛtɛ́s *v*.; iƙúmúnós *v*. **fighter** cɛmáám *n*. **fighter (be a)** cɛmɛkánón *v*. **figs (dried out)** kòsòw<sup>a</sup> *n*. **figure out** hoetés *v*. **filch** ɨtíɗíɗɛ́s *v*. **file** tɔɗʉ́pɔ́n *v*.; torópón *v*. **file in** torópétòn *v*. **filiariasis** ɲɛkwɨ *n*. **fill** cɨɨtɛ́sʉƙɔt<sup>a</sup> *v*.; éétòn *v*.; ɨmíɗítsɛ́s *v*. **fill completely** lʉʝʉlʉ́ʝɛ́s *v*. **fill in** imetsités *v*. **fill in for** imetsés *v*. **fill mouth** aukes *v*. **fill up** éítésuƙot<sup>a</sup> *v*.; éítetés *v*.; ilílíés *v*.; iluʝúlúʝés *v*.; ʝɨríʝírɛ̀tɔ̀n *v*.; ʝɨríʝírɔ̀n *v*. **fill with air** xuxuanitetés *v*. **fillet** iƙémíƙémés *v*. **film (movie)** kúrúkúríka ni ɓɛƙɛ́s *n*.; ɲévíɗyo *n*. **filter** ɨʝɨwɛs *v*.; ɨtɨwɛs *v*. **filth** ts'âɡ<sup>a</sup> *n*.
**filthiness** ts'âɡ<sup>a</sup> *n*. **filthy** itútsón *v*.; ŋorótsánón *v*.; ts'áɡòn *v*. **filthy person** ts'áɡààm *n*. **fin** taban *n*. **final to arrive** irúpóòn *v*. **final to come** irúpéètòn *v*. **finance** kaúdzòmɛ̀n *n*. **financial matters** kaúdzòmɛ̀n *n*. **find** ɨtɛ́tɔ́n *v*.; ɨtɔ́ɔ́n *v*.; nòìn *n*.; takanités *v*. **find a way into** útétòn *v*. **find a way through** utés *v*.; utésúƙot<sup>a</sup> *v*. **find an entrance to** útétòn *v*. **find each other missing** ítínós *v*. **find missing** ítés *v*. **find proof** enésá mɛná<sup>ɛ</sup> *v*. **find refuge in** rɨmɛ́s *v*. **find remains of** ítés *v*. **fine** dòòn *v*.; ɲáfaín *n*. **fine for impregnation** ɨtsʉ́lítɛtɛ́s *v*.; **fine (impregnation)** ɲɛkɨtsʉl *n*. **fine (marital)** ɲápaín *n*. **fine for sexual misconduct** kasurúɓé *n*. **fine!** maráŋ *interj*. **finger** ɨmɨnímínɛ́s *v*.; kɔrɔ́k <sup>a</sup> *n*.; mɨnímínatés *v*. **finger bone** kɔrɔ́kɔ́ɔ̀k <sup>a</sup> *n*. **finger out (food)** ts'álóbiés *v*. **finger sexually** ts'álóbiés *v*. **fingernail** tíbòlòkòɲ *n*. **fingers (extra)** ɲɛ́ɗɔ́nɨɗɔn *n*. **finish** ɨríkɛ́sʉƙɔt<sup>a</sup> *v*.; ŋábɛsʉƙɔt<sup>a</sup> *v*. **finish off** ɨmʉ́ɲɛ́sʉƙɔt<sup>a</sup> *v*.; ɨmʉɲɛtɛ́s *v*.; iɲóɗésuƙot<sup>a</sup> *v*.; ɨríkɛ́sʉƙɔt<sup>a</sup> *v*.; ɨtsʉtɛs *v*.; ɨtsʉ́tɛ́sʉƙɔt<sup>a</sup> *v*. **finish off (crops)** iróríkés *n*.
#### field
**finished** ŋábɔnʉƙɔt<sup>a</sup> *v*.; nábɔnʉƙɔt<sup>a</sup> *v*.; tɛ́zɛ̀tɔ̀n *v*. **finished (totally)** kwɛ̀rɛ̀t ɛ *ideo*. **finished off** imúɲúmétòn *v*.; itsútúmétòn *v*. **fire** hoɗésúƙot<sup>a</sup> *v*.; ídzès *v*.; ídzesuƙot<sup>a</sup> *v*.; ɨtsʉŋɛs *v*.; ts'aɗ<sup>a</sup> *n*. **fire (a weapon)** tɔɗɛ́tɛ́sʉƙɔt<sup>a</sup> *v*. **fire a weapon** tɔɗɛtɛsa ɛ́bà<sup>ɛ</sup> *v*. **fire herb** ts'aɗícɛ́mɛ̀r *n*. **fire lily** bʉlʉbʉlát<sup>a</sup> *n*. **fire on** ɗamatés *v*.; tɔƙʉmʉ́ƙʉ́mɛ́s *v*. **fire repeatedly** ídziidziés *v*. **firearm** ɛ̂b <sup>a</sup> *n*. **fireboard** tsakúdècèk<sup>a</sup> *n*. **firebreak (make a)** ɨɗɛɨɗɛ́ɛ́s *v*. **firedrill** tsakûd<sup>a</sup> *n*.; tsakúdèèàkw<sup>a</sup> *n*. **firefinch** lòtsòr *n*. **firefly** lɔ́ɗʉ́mɛ́l *n*. **fireplace** ts'aɗíékw<sup>a</sup> *n*. **firewood** dakw<sup>a</sup> *n*.; ɡamam *n*.; ts'aɗídàkw<sup>a</sup> *n*. **firing pin** mʉtʉ *n*. **firmament** lúl *n*. **first** ɛ̀kwɔ̀n *v*.; wàxʉ̀ *n*. **first (be the)** mɨtɔna ɗíɛ́wàxì *v*. **first (one)** ɗa kɔ́nɔ̀nì *pro*.; ɗa wáxì *pro*. **first person** wàxìàm *n*. **firstborn** tíɡàràmàts<sup>a</sup> *n*.; wàxìàm *n*. **fiscal** bílɔɔrɔ́*n*. **fish** ŋkɔ́líà *n*. **fish out** tukuretés *v*.; tukutetés *v*. **fissured** médemedánón *v*.; takátákánón *v*. **fist** ɨlʉlʉŋam *n*.; mʉkʉtam *n*. **fistful** ɨlʉlʉŋam *n*.
**fit** ɗoxódòn *v*. **fit (physically)** itsyátón *v*. **fitly** ɗòx *ideo*. **fitting** itémón *v*. **five** tùd<sup>e</sup> *num*.; tùdòn *v*. **five o'clock** ɲásáatɨkaa mɨtátie toomíní ńdà kɛ̀ɗì kɔ̀n *n*. **five times** tùd<sup>o</sup> *num*. **fix** ɨɗɨmɛ́s *v*.; ɨɗɨmɛ́sʉ́ƙɔt<sup>a</sup> *v*.; ɨɗɨmɛtɛ́s *v*.; ramɛtɛ́s *v*.; zɔ́bɛ̀s *v*. **fix a handle on** xɔ́ŋɛ́s *v*. **fix on** mínɛ́s *v*. **fixed** ɨɗɨmɔ́s *v*. **fixer** ɨɗɨmɛ́síàm *n*. **fixture (tool)** mínɛ́sìàw<sup>a</sup> *n*. **fizz** tabúón *v*. **fizz up** tabúétòn *v*. **flabbily** lùʝ<sup>u</sup> *ideo*. **flabby** ɡerúsúmòn *v*.; luʝúdòn *v*.; rexúkúmòn *v*. **flaccid** luʝúdòn *v*. **flaccid (become, of the penis)** ziláámètòn *v*. **flaccidly** lùʝ<sup>u</sup> *ideo*. **flag** ɨwɔ́ɔ́nʉƙɔt<sup>a</sup> *v*.; ɲébenɗéra *n*. **flagging** ɨwɔ́ɔ́n *v*. **flake (wood)** kíɓɛ́zam *n*. **flame** arír *n*.; ts'aɗíák<sup>a</sup> *n*. **flame tree** ɨtítí *n*. **flank** ŋábèrèd<sup>a</sup> *n*. **flap** ipukúpúkés *v*.; iwakúwákòn *v*.; wɛɗɨwɛ́ɗɔ́n *v*. **flapping** pɛɗɛ́pɛ́ɗánón *v*. **flare** ɨmɛ́ɗɔ́n *v*. **flare up** dulúmón *v*.; ɨmɛ́ɗɛ́tɔ̀n *v*.; tuƙúmétòn *v*. **flare up (of skin)** iŋárúrètòn *v*. **flared up (of skin)** iŋárúròn *v*.
*v*.
flash
**flash** ɨmɛ́ɗɛ́tɔ̀n *v*.; ɨmɛ́ɗɔ́n *v*.; itweɲítwéɲòn *v*. **flash!** lɛ́ʝ ɛ *ideo*. **flashlight** ɲótóts<sup>a</sup> *n*. **flashy** ɨmáráɗàɗòòn *v*. **flask (butter)** ɲéɓur *n*. **flask (gourd)** nasɛmɛ́*n*.; ɲekúrúm *n*. **flask (small gourd)** ɲékútàm *n*. **flat** ɨpáɗáɲɔ̀n *v*.; pʉ̀nʉ̀kᶶ *ideo*. **flat (deflated)** fɔrɔ́ts'ɔ́mɔ̀n *v*. **flat (of an area)** kalápátánón *v*. **flat (of land)** ɗàsòn *v*. **flat (of objects)** ɗapálámɔ̀n *v*. **flat area** lopem *n*. **flat buttocks (have)** taɓóɲómòn *v*. **flat-topped** tɔpɛ́tɔ́n *v*. **flatbed** tsídzèsìàw<sup>a</sup> *n*. **flatland** rɔw<sup>a</sup> *n*. **flatly** ɓɛlɛlɛts<sup>ɛ</sup> *ideo*.; ɗɛ̀ɲ *ideo*. **flatly concave** ɓɛtɛ́lɛ́mɔ̀n *v*.; fɛtɛ́lɛ́mɔ̀n *v*. **flatten** epitésúƙòtà ɗèɲ *v*.; iɲíkéésuƙot<sup>a</sup> *v*.; ɲaɗés *v*. **flatten out** kalápátánitetés *v*.; ɲaɗésúƙot<sup>a</sup> *v*. **flatten out (an area)** kalápátánónuƙot<sup>a</sup> *v*. **flatten repeatedly** ɲaɗiés *v*. **flatulence** fèn *n*. **flatulent** fɛnɛ́dɔ̀n *v*. **flatulently** fɛ̀n *ideo*. **flaunt** ƙɔƙɔanón *v*. **flavor** ɛfɨtɛs *v*.; íbutsurés *v*.; iwéwérés *v*. **flavorful** ɛ̀fɔ̀n *v*. **flavorful (become)** ɛfɔnʉƙɔt<sup>a</sup> *v*. **flavoring** ɲɛ́ɓɨsár *n*. **flavorless** ɗɛ̀ƙwɔ̀n *v*.; ʝɔ̀lɔ̀n *v*.; muʝálámòn *v*. **flea (chigoe)** túkútùk<sup>a</sup> *n*. **flea (sand)** túkútùk<sup>a</sup> *n*. **flea(s)** naɗɛ́p <sup>a</sup> *n*. **fleas** ŋíkaɗɛpíɗɛ́p <sup>a</sup> *n*. **fleck** bàsɔ̀n *v*. **flecked** ɨlɨmílímɔ̀n *v*. **flecked with fat** kábìlànètòn *v*. **flee** duƙésúƙota mòrà<sup>e</sup> *v*.; moronuƙot<sup>a</sup> **flee (of many)** iɗúzòn *v*. **fleece** ɗóɗòsìts'<sup>a</sup> *n*. **flesh** em *n*. **flesh dried on hide** xáƙw<sup>a</sup> *n*. **flesh left on hide** ɲɔpɔɗɛ *n*. **flex** itúkúɗètòn *v*.; itúkúɗòn *v*.; nɔƙɨnɔ́ƙɔ́n *v*. **flex (muscles)** xuxuanitetés *v*. **flexible** naƙwádòn *v*.; naúdòn *v*.; nɔƙɔ́dɔ̀n *v*. **flexibly** nàƙw<sup>a</sup> *ideo*.; nà<sup>u</sup> *ideo*.; nɔ̀ƙ ɔ *ideo*. **flexor (muscle)** ɲépísikitíém *n*. **flick** tɔɗɛtɛs *v*.; towates *v*.; towatetés *v*. **flick away** tɔɗɛ́tɛ́sʉƙɔt<sup>a</sup> *v*. **flick off** tɔɗɛ́tɛ́sʉƙɔt<sup>a</sup> *v*. **flicker** ɨmɛɗímɛ́ɗɔ̀n *v*.; ɨmɛlɛs *v*.; itweɲítwéɲòn *v*.; tɔɗɛtɛs *v*. **flicker (of light)** tɔɗɛtɛsa así *v*. **flinch** ŋaxɛ́tɔ́n *v*. **flinch (make)** iniŋíníŋés *v*. **fling** ɨpákɛ́sʉƙɔt<sup>a</sup> *v*.; ɨrʉtsɛs *v*.; towates *v*. **fling away** towátésúƙot<sup>a</sup> *v*. **fling off** towátésúƙot<sup>a</sup> *v*. **fling water** ɨpakɛsa cué *v*. **flip** ɨʝʉlɛs *v*.; ɨʝʉlɛtɛ́s *v*.; tɔɗɛtɛs *v*. **flip away** tɔɗɛ́tɛ́sʉƙɔt<sup>a</sup> *v*.
#### fold in half
#### flip off
**flip off** tɔɗɛ́tɛ́sʉƙɔt<sup>a</sup> *v*. **flip over** aɓúlúkánón *v*. **flipped** ɨʝʉlɔs *v*. **flit** ɨkaɨkɛ́ɛ́s *v*. **flit around** ɨpɛrípɛ́rɔ̀n *v*. **flitter** ɨmɛlɛs *v*.; ɨpɛrípɛ́rɔ̀n *v*.; iwolíwólòn *v*.; tɔɗɛtɛs *v*. **flitter (of termites)** érítòn *v*. **float** ilélébètòn *v*. **float away** ilélébonuƙot<sup>a</sup> *v*. **flock** bàr *n*. **flock (small)** bàròìm *n*. **flock of birds** ɡwábàr *n*. **flock of goats** riébàr *n*. **flock of sheep** ɗóɗòbàr *n*. **flocks** ŋíɓarɛn *n*. **flog** iɓúŋéés *v*. **flood** ísw<sup>a</sup> *n*. **flood rubbish** ɲérímama *n*. **floodplain** ɲéɓúruɓur *n*. **floor** dziŋ *n*.; kíʝ<sup>a</sup> *n*. **flop!** kùm *ideo*. **flour** kabas *n*. **flour (dry)** ŋápʉp<sup>a</sup> *n*. **flour (moist)** ŋámírɔ̀ *n*. **flourish** ɨɔ́kɔ́n *v*. **flourish (of plants)** ɡáruɓúɓón *v*.; karuɓúɓón *v*. **flow** isépón *v*. **flow (menses)** iona aráɡwaník<sup>ɛ</sup> *v*. **flow away** isépónuƙot<sup>a</sup> *v*. **flow in waves** ídulumona cué *v*. **flower** ɨɔ́kɔ́n *v*.; ŋátur *n*. **flower (sorghum)** ƙádetésá kadɨxá<sup>ɛ</sup> *v*. **flower bud** tún *n*.
**flu** ɲarʉ́kʉ́m *n*.; suɡur *n*. **fluctuate** iɲíkétòn *v*.; iɲikiétòn *v*.; iɲikíɲíkòn *v*.; iɲíkón *v*. **fluid** cuanón *v*. **flump down** rɛfɛ́kɛ́ɲɔ̀n *v*. **flunk** rúmánòn *v*. **flush** ɓʉnʉ́mɔ́nà sèà<sup>e</sup> *v*. **flush out** tsídzètòn *v*.; tsídzitetés *v*. **flushed** iyaŋíyáŋòn *v*. **flustered** iyaŋíyáŋòn *v*. **flute** ɲálamorú *n*. **flutter** ipukúpúkés *v*.; iwolíwólòn *v*.; pɛɗɛpɛ́ɗɔ́n *v*.; wɛɗɨwɛ́ɗɔ́n *v*. **flutter (of termites)** érítòn *v*. **fluttering** pɛɗɛ́pɛ́ɗánón *v*. **fly** bʉrɛ́tɔ́n *v*.; bʉ̀rɔ̀n *v*.; iwálílòn *v*.; tsúts<sup>a</sup> *n*. **fly (biting)** lɔŋɨzɛt<sup>a</sup> *n*.; lótsóts<sup>a</sup> *n*.; ɲɔ́pɔɗɔkʉ́ *n*. **fly (flee, of many)** iɗúzòn *v*. **fly (tsetse)** ɲɛ́ɗíɨt<sup>a</sup> *n*. **fly away** bʉrɔnʉƙɔt<sup>a</sup> *v*. **fly off** bʉrɔnʉƙɔt<sup>a</sup> *v*. **fly species (green)** tukéy<sup>a</sup> *n*. **flycatcher (paradise)** iboboy<sup>a</sup> *n*. **foal** ɗìɗèìm *n*. **foal (female)** ɗìɗèwàz *n*. **foam** ɡuf *n*.; tabúón *v*. **foam up** tabúétòn *v*. **fodder** wà *n*. **foe** lɔŋɔ́tɔ́m *n*. **foes** lɔŋɔ́t <sup>a</sup> *n*. **fog** ɡóʒòw<sup>a</sup> *n*. **foil** kwaɲɛ́s *v*.; kwaɲɛ́sʉ́ƙɔt<sup>a</sup> *v*. **foist** tɔnɛɛtɛ́s *v*. **foist oneself** tɔnɛɛtɛ́sá así *v*. **fold** iwoles *v*.; tɔɓɨlɛs *v*. **fold in half** ikóóbés *v*.; ikóóbetés *v*.
**fold over** iwoletés *v*. **fold up** ikóóbés *v*.; ikóóbetés *v*.; tɔɓɨlɛtɛ́s *v*.; tusuketés *v*.; tusúkón *v*. **folded over** tɔɓɨlɛsa así *v*. **folk** ròɓ<sup>a</sup> *n*. **folks!** òɓà *interj*.; ròɓà *interj*. **follicle (hair)** síts'ádɛ̀ *n*. **follow** iɗupes *v*.; tɔmɛɛs *v*.; tɔmɛ́ɛ́sʉƙɔt<sup>a</sup> *v*.; túbès *v*. **follow (in doing)** itáƙúòn *v*. **follow after** elánónuƙot<sup>a</sup> *v*. **follow away** túbesuƙot<sup>a</sup> *v*. **follow each other** túbunós *v*. **follow here** tɔmɛɛtɛ́s *v*. **follow in line** tɔtʉ́pɔ́n *v*. **follow off** túbesuƙot<sup>a</sup> *v*. **follower** dɛáám *n*.; túbèsìàm *n*. **folly** ɨɓááŋàs *n*. **fondle** ídadamɛ́s *v*.; ɨwáwɛ́ɛ́s *v*.; tárábes *v*. **fondle all over** tárábiés *v*. **fontanelle** baɗɨbaɗas *n*.; bɔɗɨbɔɗɔs *n*. **food** ŋƙáƙ<sup>a</sup> *n*. **food (for elders first)** tʉ́w<sup>a</sup> *n*. **food (give)** ŋƙáƙítetés *v*.; ŋƙɨtɛtɛ́s *v*. **food (gnawable)** ats'am *n*. **food (peelable)** ɨsɨmam *n*. **food (ready)** aeam *n*. **food (ripe)** aeam *n*. **food (sippable)** abutiam *n*. **food (slurpable)** ɨsɔrɔɓam *n*. **food residue** ɲéɗúruɗur *n*. **food slices (dried)** iram *n*. **fool!** ɗʉ́rʉɗɔ́ɔ̀ *interj*. **fool's gold** ɲésiɓalitútu *n*. **foolish** ɨɓááŋɔ̀n *v*.
**foolishness** ɨɓááŋàs *n*. **foot** dɛ *n*.; dɛɛd<sup>a</sup> *n*. **foot (non-animal)** dɛ *n*.; dɛɛd<sup>a</sup> *n*.; dziŋ *n*. **foot of a boulder** taɓádɛ̀ *n*. **foot of a fence** marɨŋídɛ̀ *n*. **foot of a mountain** kwarádɛ̀ *n*. **foot of a tree** dakúdɛ̀ *n*. **football** ɲɛ́pɨɨrá *n*. **footer** dɛ̀ɛ̀dà hò<sup>e</sup> *n*. **footfall** kímáts<sup>a</sup> *n*. **footing of a borehole** ɲatsʉʉmádɛ̀ *n*. **footman** dɛáám *n*. **footpath (fresh)** fʉ́fʉ́t <sup>a</sup> *n*. **footprint** dɛ *n*. **footstep** kímáts<sup>a</sup> *n*. **footsure** tsɛ́rɛkɛ́kɔ́n *v*. **for good** kìŋ *ideo*. **for nothing** ŋálàk<sup>a</sup> *ideo*. **for the reason that** ikóteré *subordconn*.; kóteré *subordconn*. **for what?** isiɛník<sup>ɛ</sup> *n*. **forager** ɡaɗikamáám *n*. **forager (of greens)** wààm *n*. **foraging** ɡaɗikam *n*. **foraging in the valleys** fátárààkànànès *n*. **forbid** dimités *v*.; dimitetés *v*.; itáléés *v*. **forbidden** itáléánón *v*.; itálóós *v*. **force** ɨtíŋɛ́ɛ́s *v*.; ɨtɨŋɛs *v*.; rɛ́ɛ́s *v*.; rɛɛtɛ́s *v*.; tɔrɛɛs *v*. **force oneself** ɨtɨŋɛsa así *v*. **force out** ɨlɔ́líɛ́s *v*. **force through** ututetés *v*.; xutés *v*.; xutésúƙot<sup>a</sup> *v*. **force through repeatedly** ututiés *v*. **force together** rɔɲɛ́s *v*. **forceps** ɲɔkɔ́ɲɛ́t <sup>a</sup> *n*.
**ford** àrònìàw<sup>a</sup> *n*.; ôd<sup>a</sup> *n*.; ódèèkw<sup>a</sup> *n*. **forearm** ƙɔ́dɔ̀l *n*.; ɲepísíkit<sup>a</sup> *n*. **forearm muscle** ɲépísikitíém *n*. **forego** bɔlɛ́sʉ́ƙɔt<sup>a</sup> *v*.; tasálétòn *v*.; tasálón *v*. **forehead** takár *n*. **foreign** ʝalánón *v*. **foreign child** ɦyòìm *n*. **foreign language** ɦyɔɛn *n*.; ɦyɔ̀tòd<sup>a</sup> *n*. **foreign woman** ɦyɔ̀cèk<sup>a</sup> *n*. **foreigner** ámá na biyá<sup>e</sup> *n*.; ɦyɔ̀àm *n*.; kíʝíkààm *n*.; ŋíɓúkúìàm *n*.; ɲeɓúkúit<sup>a</sup> *n*. **foreleg** ɲepísíkit<sup>a</sup> *n*. **foreleg muscle** ɲépísikitíém *n*. **foreman** ámázeáma teréɡì *n*. **foremilk** ɲóɗós *n*. **foreskin** kwaníts'ɛ́*n*.; ts'ɛa na kwaní *n*. **forest** dakúáƙw<sup>a</sup> *n*.; fɔrɔ́sìtà *n*.; ríʝ<sup>a</sup> *n*. **forest (cleared)** tsɛ̀f *n*. **forest (dense)** lolíts<sup>a</sup> *n*. **forest (planted)** ɲɛ́pɔrɛ́sìtà *n*. **forest dombeya** xuxûb<sup>a</sup> *n*. **forest entrance** ríʝáàk<sup>a</sup> *n*. **forestall** titikes *v*. **forever** ʝɨkî *adv*.; mʉ̀kà *adv*.; pákà *adv*. **forever and ever** kaíníka ńda kaíník<sup>a</sup> *n*. **forfeit** kuritésúƙota así *v*. **forge** ityakes *v*. **forger** ìtyàkààm *n*. **forget** hakaikés *v*. **forget (make)** ɡwelítésuƙot<sup>a</sup> *v*.; hakaikitetés *v*. **forgetful** imáɗíŋánón *v*.
**forging stone** ityakesíɡwàs *n*.
**forgive** iƙenes *v*.; óɡoés *v*.
**forgotten** ɡwèlòn *v*.; hakaikós *v*.; hakonuƙot<sup>a</sup> *v*.; iɓíléròn *v*. **forgotten (become)** iɓíléronuƙot<sup>a</sup> *v*. **fork** ɲácaƙwarát<sup>a</sup> *n*.; ɲɔ́fɔ́k <sup>a</sup> *n*.; taŋatsárón *v*.; tɛlɛ́tsɔ́n *v*.; toŋélón *v*. **fork (of a tree)** bɔ̀kɛ̀d <sup>a</sup> *n*. **fork of a tree** bɔ̀k <sup>a</sup> *n*. **form** bɛrɛtɛ́s *v*.; itues *v*.; ituetés *v*.; ituetésá así *v*. **form droplets** tsɨpɨtsípɔ́n *v*. **form rainclouds** mɔƙɨmɔ́ƙɔ́n *v*. **forsake** ɡóózés *v*.; ɡóózesuƙot<sup>a</sup> *v*. **forsake each other** ɡóózinósúƙot<sup>a</sup> *v*. **forsaken** ɡóózosuƙot<sup>a</sup> *v*. **fortify** ŋɨxítɛ́sʉƙɔt<sup>a</sup> *v*. **fortunate** tírɨríŋɔ́n *v*. **fortune** muce *n*.; ɲaɓáát<sup>a</sup> *n*. **fortune (decent)** mucea na ɓárɨɓár *n*. **fortune (good)** mucea na títìàn *n*.; ɲaréréŋ *n*. **fortune (ill)** mucea ná ʝɔ̀l *n*. **fortune (terrible)** mucea na ináƙúós *n*. **forty** toomínékwa ts'aɡús *n*. **forward** wàxìk<sup>ɛ</sup> *n*. **foster** zeites *v*.; zeitésuƙot<sup>a</sup> *v*. **foul** ɗɛtsɨɗɛ́tsɔ́n *v*.; ɨmʉ́sɔ́ɔ̀n *v*. **foul (become)** ɨmʉ́sɛ́ɛ̀tɔ̀n *v*. **foul-smelling** wízɨlílɔ́n *v*. **foully** ɗɛ̀ts<sup>ɛ</sup> *ideo*. **foundation** dɛ̀ɛ̀dà hò<sup>e</sup> *n*. **fountain** ɲɛɨtánɨt<sup>a</sup> *n*. **four** ts'aɡús *num*.; ts'aɡúsón *v*. **four o'clock** ɲásáatɨkaa mɨtátie toomín *n*. **four times** ts'aɡús<sup>o</sup> *num*.
**four years ago** kaɨnɔ nɔɔ kɛ *n*.; nɔkɛɨna kenóó ke *n*. **four-by-four** ts'aɡusátìk<sup>e</sup> *v*. **fourteen** toomíní ńda kiɗi ts'aɡús *n*. **fourth (be a)** kɔnɔna ts'aɡúsónù *v*. **fourth (one)** ɗa ts'aɡúsónì *pro*. **fowl** ɡwa *n*. **fox (bat-eared)** bùràts<sup>a</sup> *n*. **fractured** takátákánón *v*. **fragile** tɛɛmɛ́mɔ̀n *v*.; yɛmɛ́dɔ̀n *v*. **fragilely** yɛ̀m *ideo*. **fragment** ɡúɗúsam *n*. **fragrant** tukukúɲón *v*. **frail** dɛrɛ́dɔ̀n *v*.; ɨpáláƙɔ̀n *v*.; ʝuódòn *v*. **frailly** dɛ̀r *ideo*.; ʝù<sup>o</sup> *ideo*. **frame (door)** ɲɛ́fɨrɛ́m *n*. **francolin (crested)** bílíkɛrɛtɛ́*n*. **francolin (Jackson's)** kwítsíladidí *n*. **francolin (scaly)** tìkɔ̀rɔ̀ts<sup>a</sup> *n*. **frayed** ɲɛɗɛ́dɔ̀n *v*. **frayedly** ɲɛ̀ɗ ɛ *ideo*. **freak** ƙʉts'áám *n*. **freak out** doʝánónuƙot<sup>a</sup> *v*. **free** ɓàŋɔ̀n *v*.; bùlòn *v*.; hoɗés *v*.; hoɗésúƙot<sup>a</sup> *v*.; hoɗetés *v*.; ɨɓámɔ́n *v*.; talakes *v*. **free of charge** ɨɓámɔ́n *v*. **free oneself** hoɗetésá así *v*. **free to walk** zíbos *v*. **freeload** lɛŋɛ́s *v*. **freeloader** lɛŋɛ́síàm *n*. **freeloading** olíɓó *n*. **freer** hoɗetésíàm *n*. **fresh** erútsón *v*. **fret** alólóŋòn *v*. **friar** purutél *n*.
**Friday** Ɲákásíá tùdìk<sup>e</sup> *n*. **friend (agreer)** tsámʉ́nɔtɔ́síàm *n*. **friend (foreign)** ɲòt<sup>a</sup> *n*. **friend (his/her foreign)** ntsíɲót<sup>a</sup> *n*. **friend (in-group)** ɛbam *n*. **friend (my foreign)** ɲ́cìɲòt<sup>a</sup> *n*. **friend (my)** nádzàƙ<sup>a</sup> *n*. **friend (sharer)** tɔ̀mɔ̀rààm *n*. **friend (your foreign)** biɲót<sup>a</sup> *n*. **friendliness (in-group)** ɛbamánánès *n*. **friendliness (out-group)** ɲotánánès *n*.; ɲótíkónánès *n*. **friends (make foreign)** ɲotánónuƙot<sup>a</sup> *v*. **friendship (in-group)** ɛbamánánès *n*. **friendship (out-group)** ɲotánánès *n*.; ɲótíkónánès *n*. **frighten** kitítésuƙot<sup>a</sup> *v*.; ŋaxɨtɛtɛ́s *v*.; xɛɓɨtɛs *v*.; xɛɓɨtɛ́sʉ́ƙɔt<sup>a</sup> *v*. **frighten away** ɨrɛmɛs *v*. **frighten off** ɨrɛmɛs *v*. **frightened** ŋaxɛ́tɔ́n *v*. **frigid (sexually)** ɨɛ́ɓɔ́n *v*. **fringe** kwayw<sup>a</sup> *n*.; kweed<sup>a</sup> *n*. **frog** ƙwaát<sup>a</sup> *n*. **from** nàpèì *prep*.; nɛ́ɛ́*prep*.; ɲàpèì *prep*. **from the heart** ɡúróɛ́n <sup>ɔ</sup> *n*. **from there** ts'ɛ́dɔ́ɔ́kɔ̀nà *pro*. **from when** nàpèì *subordconn*.; ɲàpèì *subordconn*. **from where?** ndéé *n*. **front** takárɛ̂d <sup>a</sup> *n*.; wàx *n*. **frontal bone (upper)** matáŋíɡwarí *n*. **froofy** bʉlʉbʉlɔs *v*. **froth** ɡuf *n*.; tabúón *v*. **froth up** tabúétòn *v*. **frown** iɲíkón *v*. **frowning (begin)** iɲíkétòn *v*.
garden
**frozen (in fear)** dodimórón *v*. **frozen (still)** lɛrɛ́dɔ̀n *v*. **fruit of** *ts'ɔƙɔm* fɔfɔ́ʝ <sup>a</sup> *n*.; ŋalúɓ<sup>a</sup> *n*. **fruit(s)** eɗin *n*. **fruit-laden** bɔ́ŋɔ́n *v*. **fruitless (become)** ɨkárímétòn *v*. **frustrate** ɨkarɛtɛ́s *v*.; ɨlɔ́ítɛ́sʉƙɔt<sup>a</sup> *v*. **fry** kɛ́xɛ́s *v*. **fryer** kɛ́xɛ́sìàm *n*. **fuck** ɨtɛpɛs *v*. **fuel** ceím *n*. **fuel (fire)** iríréés *v*.; iríréetés *v*. **Fuerstia africana** kʉláɓákàk<sup>a</sup> *n*. **fulfill** ɨríkɛ́sʉƙɔt<sup>a</sup> *v*. **full** cìɔ̀n *v*.; eódòn *v*. **full (become)** cɨɔnʉƙɔt<sup>a</sup> *v*.; éétòn *v*. **full (completely)** tsɛ̀k ɛ *ideo*. **full (very)** pìc *ideo*.; tìr *ideo*. **full completely** lʉʝʉlʉ́ʝɔ́n *v*. **full of it (lying)** isuɗesa mɛná<sup>ɛ</sup> *v*. **fully** fìr *ideo*. **fumble around** ɨtapítápɔ̀n *v*. **fume** ipúróòn *v*. **fume (angrily)** siŋírón *v*. **fumes** ts'ûd<sup>a</sup> *n*. **fumigate** ipúréés *v*.; iwaŋíwáŋés *v*.; ts'udités *v*. **fun** ɛ̀fɔ̀n *v*. **fun (become)** ɛfɔnʉƙɔt<sup>a</sup> *v*. **fun (have)** iyóómètòn *v*. **fun (make)** ɛfɨtɛs *v*. **function** ɲákásìèd<sup>a</sup> *n*. **funerary goat** kɔ̀p <sup>a</sup> *n*.; ɲépúɲa *n*. **funerary-goat killer** ɲépúɲáám *n*. **funny** ɛ̀fɔ̀n *v*.
**funny (become)** ɛfɔnʉƙɔt<sup>a</sup> *v*. **funny (make)** ɛfɨtɛs *v*. **fur** ínósìts'<sup>a</sup> *n*. **furnace (brick)** ɲéripipí *n*. **furrow** ɲɛ́pɛ́lʉ *n*. **furry** tsèkòn *v*. **furthermore** naɓó *coordconn*.; toni naɓó *n*. **fury** ɡaánàs *n*.; ɲɛlɨl *n*. **fuse** iɲales *v*. **futile (become)** ɨkárímétòn *v*. **future** fàr *adv*. **future (distant)** tsò *adv*. **fuzzy** saúkúmòn *v*. **gabby** mɔɲɨmɔɲɔs *v*. **gag** ʝaƙátós *v*.; toukes *v*.; touketés *v*.; xáƙátòn *v*. **gain ground on** iɗiles *v*. **gain weight** tuɓútónuƙot<sup>a</sup> *v*.; zízonuƙot<sup>a</sup> *v*. **galago (lesser)** ɡwan *n*. **Galatians (biblical)** Ŋíɡalatíaik<sup>a</sup> *n*. **gall** bìɗ<sup>a</sup> *n*. **gallbladder** bìɗ<sup>a</sup> *n*.; bìɗàhò *n*. **game** ɲaɓolya *n*.; wáák<sup>a</sup> *n*. **game (Ludo)** ɲélúɗo *n*. **gang** ɲáʝore *n*. **gang up on** ɨŋáŋárɛtɛ́s *v*. **gangly** sawátsámòn *v*. **gap** ɨkálámɛ́s *v*.; tsàŋ *n*. **gaping** hádòlòmòn *v*.; laɓáɲámòn *v*.; lafárámòn *v*. **gappy (teeth)** ɲaŋálómòn *v*. **garbage** ts'ʉts'ʉ *n*. **garbage dump** ts'ʉts'ʉ́áw<sup>a</sup> *n*. **garble** ƙʉʝʉ́dɔ̀n *v*. **garble speech** iŋolíŋólésa tódà<sup>e</sup> *v*. **garden** sêd<sup>a</sup> *n*.
get away from
**garden (multi-year)** ŋwan *n*. **garden (newly cleared)** kɨwíl *n*. **garden (newly plowed)** túbùr *n*. **garden (tiny)** fɔ́ɗ <sup>a</sup> *n*. **garden (vegetable)** waicíkásèd<sup>a</sup> *n*. **garden boundary** ɲókorimít<sup>a</sup> *n*. **garden camp** ɲóɓóot<sup>a</sup> *n*. **garden edge** lɔkɨram *n*. **garden rain shelter** lɔtɔ́ƙ <sup>a</sup> *n*. **garment** ƙwàz *n*. **gas (intestinal)** fèn *n*. **gas (petrol)** ceím *n*. **gas (propane)** ɲáɡás *n*. **gash** dzɛrɛ́s *v*. **gasoline** ɲépetorón *n*. **gassily** fɛ̀n *ideo*. **gassy** fɛnɛ́dɔ̀n *v*. **gate** ɔ́dɔ̀k <sup>a</sup> *n*. **gateway** ɔ́dɔ̀k <sup>a</sup> *n*. **gather** ikóóbés *v*.; ɨrírɛ́ɛ́s *v*.; ɨtsʉnɛs *v*.; ɨtsʉnɛtɛ́s *v*.; ɨtsʉ́nɛ́tɔ̀n *v*.; itukánón *v*.; itukes *v*.; ɨʉɗɛs *v*.; ɨʉɗɛtɛ́s *v*. **gather (contributions)** bɔsɛtɛ́s *v*. **gather (glean)** ɨrárátés *v*.; ɨrarɛs *v*.; tarares *v*. **gather and move** ɲʉ́ɲɛ́s *v*. **gather and remove** ɲʉ́ɲɛ́sʉƙɔt<sup>a</sup> *v*. **gather oneselves** ɗɔtsɛtɛ́sá así *v*.; ɨrírɛ́ɛtɛ́sá así *v*. **gather saliva** ɨmʉʝʉ́mʉ́ʝɔ̀n *v*. **gather socially** ɨtsʉnɛtɛ́sá así *v*. **gather up** ikóóbetés *v*.; ɨrírɛ́ɛ́sʉƙɔt<sup>a</sup> *v*.; ituketés *v*. **gather up and move** ɲʉɲɛtɛ́s *v*. **gatherer (of greens)** wààm *n*. **gathering** kur *n*.; ɲatʉ́kɔ́t <sup>a</sup> *n*.
**gathering (big)** zɛƙɔ́áwa ná zè *n*. **gauche** betsínón *v*. **gaudy** ɨmáráɗàɗòòn *v*. **gaunt** ɨkárɔ́n *v*.; kɔrɔ́ɗɔ́mɔ̀n *v*.; lotímálèmòn *v*. **gaze at emptily** itelesa bàrìrrr *v*. **gazelle** kodow<sup>a</sup> *n*. **gearshift** cɛbɛn *n*.; ɲɛ́sɛɛɓɔ́*n*. **gearstick** cɛbɛn *n*.; ɲɛ́sɛɛɓɔ́*n*. **gecko** lókíɓoɓó *n*. **gecko (dwarf)** kíɓíɓìt<sup>a</sup> *n*. **gecko species?** léɗ<sup>a</sup> *n*.; ɲámɨlɨɔŋɔ́r *n*. **gel** tɔsɔ́ɗɔ́kɔ̀n *v*. **gelatinous** milílón *v*. **generosity** daás *n*. **generous** dòòn *v*.; waŋádòn *v*. **generous person** dòònìàm *n*. **genet** lɔ̀rìƙìlà *n*. **genitalia (female)** didis *n*. **gentle** batánón *v*. **genuflect** kutúŋétòn *v*. **germ** ƙʉts'<sup>a</sup> *n*. **German shepherd** ɲeryaŋíŋók<sup>a</sup> *n*. **germinate** ɓúrukúkón *v*.; morétón *v*.; rʉ́bɛ̀tɔ̀n *v*.; rʉ́bɔ̀n *v*.; tʉwɛ́tɔ́n *v*.; tʉ̀wɔ̀n *v*.; xúbètòn *v*.; xúbòn *v*. **germinate (of grain)** xokómón *v*. **germinate fully (of grain)** yɛlíyɛ́lɔ̀n *v*. **get** iryámétòn *v*.; ɨtɛ́tɔ́n *v*.; ɨtɔ́ɔ́n *v*.; ƙanetés *v*.; tɛ́bɛtɛ́s *v*. **get (cause to)** ƙanitetés *v*. **get (water)** kotsés *v*.; kotsetés *v*. **get a rise out of** ɨtɔ́ŋɔ́ɛ́s *v*. **get away** iɓutsúmétòn *v*.; iwálílòn *v*.; ùtòn *v*.; utonuƙot<sup>a</sup> *v*. **get away (from home)** ɨpɛ́ɛ́rɔ̀n *v*. **get away from** ɨɓʉtsɛs *v*.; ɨɓʉtsɛtɛ́sá así *v*.
give in
**get close to** rɔɲɛ́sá así *v*. **get cloudy** kupétón *v*.; kupukúpón *v*. **get dark** witsiwítsétòn *v*. **get down** rʉʝɛ́tɔ́n *v*. **get going** ɗóɗésa muceé *v*. **get here** ɨtaɨtɛtɛ́s *v*. **get light (of sun)** ítóna kíʝée ts'ɛɛ *v*. **get lost** hakonuƙot<sup>a</sup> *v*. **get on** otsés *v*.; otsésúƙot<sup>a</sup> *v*. **get out** tɔpɛ́ɔ́n *v*.; utonuƙot<sup>a</sup> *v*. **get out of the way** ècòn *v*.; èkòn *v*. **get payback** ɲaŋés *v*. **get ready to go** súɓánòn *v*.; suɓétón *v*. **get rid of** ɡʉts'ʉrɛs *v*.; ɡuts'uriés *v*.; itsúrúés *v*. **get rid of (kill)** ɨɗɛɛs *v*.; ɨɗɛ́ɛ́sʉƙɔt<sup>a</sup> *v*. **get saved** hoɗetésá así *v*. **get shady** kurukúrétòn *v*. **get sick** meetón *v*. **get there** ɨtaɨtɛ́s *v*. **get through** utonuƙot<sup>a</sup> *v*. **get to know** ɦyeités *v*. **get together** ɨtsʉnɛtɛ́sá así *v*. **get up** ŋkáítetés *v*.; ŋkéétòn *v*.; ŋkóón *v*. **get up (of many)** ŋkaíón *v*. **ghost** kúrúkúr *n*.; lopéren *n*.; tás *n*. **ghoul** lopéren *n*.; tás *n*. **giant** bàd<sup>a</sup> *n*.; kébàdà *n*.; nábàdà *n*.; nébàdà *n*. **gift** meetam *n*.; tɔ́rɔ́bɛs *v*. **gifts** meetésíicík<sup>a</sup> *n*. **giggle** ɡaɡaanón *v*. **giggly** fekifekos *v*. **gimp** ɨsɛ́pɔ́n *v*.; ɨtɔ́ƙɔ́ɔ̀n *v*.; itsúkúkòn *v*. **gimpily** itsúkúk<sup>u</sup> *ideo*.
**gimpy** ŋoɗólómòn *v*. **gin (Sunny)** ɲásáníjìn *n*. **gingiva** diriʝiʝ<sup>a</sup> *n*. **giraffe** ɡwaíts'<sup>a</sup> *n*. **giraffe bull** ɡwaíts'ícikw<sup>a</sup> *n*. **giraffe calf** ɡwaíts'íìm *n*. **giraffe cow** ɡwaíts'íŋwa *n*. **giraffe-tail cap** ɲɔ́tsɔ́ɓɛ *n*. **girl** ɲàràm *n*. **girl (baby)** cue *n*. **girl (little)** ɲàràmàìm *n*. **girls' company** ɲèrààƙw<sup>a</sup> *n*. **girl-crazy** iɲéráánón *v*. **girlfriend** ɲàràm *n*. **girls** ɲèr *n*. **girls (little)** ɲerawik<sup>a</sup> *n*. **girls' grass** ɲèràkù *n*. **give** bírɛ́s *v*.; iʝokes *v*.; iʝókésuƙot<sup>a</sup> *v*.; meés *v*.; meetés *v*. **give (chewing tobacco)** matáŋɨtɛtɛ́s *v*. **give a ride to** ɨɛ́bɛ̀s *v*.; ɨɛ́bɛsʉƙɔt<sup>a</sup> *v*.; ɨɛ́bɛtɛ́s *v*. **give across** íɡorésúƙot<sup>a</sup> *v*. **give away** maƙésúƙot<sup>a</sup> *v*. **give back** raʝésúƙot<sup>a</sup> *v*.; raʝetés *v*. **give birth** ƙwaatón *v*. **give birth (help)** ƙwaatítetés *v*. **give birth to** ƙwaatetés *v*. **give birth to prematurely** ɨsɔɛtɛ́s *v*. **give chace** irúkésuƙot<sup>a</sup> *v*. **give drink** wetités *v*. **give each other** maímós *v*. **give food relief** bɔnɛ́s *v*. **give gifts to** tɔ́rɔ́bɛs *v*. **give in** kuritésúƙota así *v*.
**give marching orders to** taŋasɛs *v*. **give medicine** wetitésá cɛmɛ́ríkà<sup>ɛ</sup> *v*. **give off** itsues *v*.; itsuetés *v*. **give oneself away** maƙésúƙota así *v*. **give out** dónés *v*.; dónésuƙot<sup>a</sup> *v*.; maƙésúƙot<sup>a</sup> *v*. **give over** íɡorésúƙot<sup>a</sup> *v*. **give rectally** ɨtʉrɛs *v*. **give suck to** naƙwɨtɛs *v*. **give up** bɔlɛ́sʉ́ƙɔt<sup>a</sup> *v*.; kuritésúƙota así *v*.; taʝales *v*.; taʝálésuƙot<sup>a</sup> *v*.; taʝaletés *v*. **given** moós *v*. **giver** meesíám *n*. **giving birth** ƙwaát<sup>a</sup> *n*. **gizzard** ŋìl *n*. **glad** ɨlákásɔ̀n *v*. **glad (become)** ɨlákásɔ́nʉƙɔt<sup>a</sup> *v*. **glade** ɓɔɗ<sup>a</sup> *n*.; ɗípɔ̀ *n*. **glance** ɨɛ́bɛ̀s *v*.; ɨɛ́bɛsʉƙɔt<sup>a</sup> *v*.; ɨɛ́bɛsʉƙɔt<sup>a</sup> *v*.; ɨɛ́bɛtɛ́s *v*. **glance (bump)** toyeres *v*. **glance off** iɗótón *v*. **glance sidelong at** ɨŋɔ́ɓɛ́lɛ́s *v*. **gland** ƙùts'àts'<sup>a</sup> *n*. **glandular swelling** lɔ́mílɨmíl *n*. **glans penis** eɗed<sup>a</sup> *n*.; kwaníéɗ<sup>a</sup> *n*. **glare** ɨraírɔ́ɔ̀n *v*. **glare at** ŋɔ́zɛ̀s *v*.; ŋóziés *v*. **glare at each other** ŋɔ́zɨnɔ́s *v*. **glarer** ŋɔ́zɛ̀sìàm *n*. **glasses** ɲékiyóìk<sup>a</sup> *n*. **gleam** parɨpárɔ́n *v*.; piripírón *v*. **gleam when wet** tsalɨtsálɔ́n *v*. **gleaming** ɓalídɔ̀n *v*.; pirídòn *v*. **glean** ɨrárátés *v*.; ɨrarɛs *v*.; tarares *v*. **gleeful** taɓolos *v*.
**glide** ɨfɛlɛsa así *v*.; ɨfɛ́lɔ́nʉƙɔt<sup>a</sup> *v*.; ɨɔ́ɔ́rɛ́s *v*.; ɨɔ́ɔ́rɔ̀n *v*. **glide through** ɨsɛ́lɛ́tɛ́sʉƙɔta así *v*. **glimmer** riɓiríɓón *v*. **glimmer when wet** tsalɨtsálɔ́n *v*. **glint** tɔɗɛtɛsa así *v*. **glisten** parɨpárɔ́n *v*.; piripírón *v*. **glistening** ɓalídɔ̀n *v*. **glitter** ɨmɛɗímɛ́ɗɔ̀n *v*. **glitterily** mìl *ideo*. **glittery** mɨlídɔ̀n *v*. **gloat** taɓólón *v*. **gloating** taɓolos *v*. **globe** kíʝ<sup>a</sup> *n*. **gloomy** sìŋòn *v*. **glorious** dòòn *v*. **glory** daás *n*. **glow (of fire)** lɔɔrán *n*. **glue** ɲáɡám *n*. **gluey** ɗɔmɔ́dɔ̀n *v*.; ɨríítánón *v*. **gluily** ɗɔ̀m *ideo*. **glut** iwótsóòn *v*. **glutton** lòkòɗòŋìròàm *n*.; loƙeƙes *n*. **gluttonousness** lokoɗoŋironánés *n*. **gluttony** aeásá bùbùì *n*.; lòkòɗòŋìrò *n*. **gnarled** lɛrɛ́kɛ́mɔ̀n *v*. **gnat** dililits'<sup>a</sup> *n*. **gnaw** áts'ɛ́s *v*. **Gnidia subcordata** mɨsíás *n*. **go** ƙòòn *v*. **go [a sound]** kʉ̀tɔ̀n *v*.; kʉtɔnʉƙɔt<sup>a</sup> *v*. **go across** ɡórés *v*.; íɡorés *v*.; íɡorésúƙot<sup>a</sup> *v*.; kámáránón *v*.; kámáránónuƙot<sup>a</sup> *v*. **go across repeatedly** ɡóriés *v*.; íɡoriés *v*. **go after** ɨlɔŋɛs *v*.; ɨlɔ́ŋɛ́sʉƙɔt<sup>a</sup> *v*.; túbesuƙot<sup>a</sup> *v*. **go after each other** ríínós *v*.
**go ahead** ƙoona wáxìk<sup>ɛ</sup> *v*. **go along the side** ƙoona rutet<sup>o</sup> *v*. **go around** ɨlɔ́ɗɔ́nʉƙɔt<sup>a</sup> *v*.; ɨsʉkɛs *v*.; sʉ́kɛ́s *v*.; tamanɛs *v*.; tamanɛtɛ́s *v*.; tamánɛ́tɔ̀n *v*. **go around in** irimes *v*. **go around repeatedly** tamaniés *v*. **go around restlessly** ɗotíɗótòn *v*. **go asleep (of limbs)** isálílètòn *v*. **go astray** hakonuƙot<sup>a</sup> *v*.; itwáŋón *v*. **go at dawn** ifúlón *v*. **go away forever** ƙòònà ʝìr *v*. **go back** iɓóɓóŋòn *v*.; itéón *v*.; itíón *v*.; ƙoona ʝírìk<sup>ɛ</sup> *v*. **go behind** ƙoona ʝírìk<sup>ɛ</sup> *v*. **go behind clouds** ƙooná ɡìdààƙɔ̀k ɛ *v*. **go bring** ƙáidetés *v*. **go broke** ŋʉrʉ́mɔ́n *v*. **go by** ɨɛ́bɛ̀s *v*.; ɨɛ́bɛsʉƙɔt<sup>a</sup> *v*.; ilúɲón *v*.; ilúɲónuƙot<sup>a</sup> *v*. **go by way of** tɔmɛ́ɛ́sʉƙɔt<sup>a</sup> *v*. **go crazy** doʝánónuƙot<sup>a</sup> *v*.; itwáŋón *v*. **go directly** ɨtsírɔ́n *v*.; iyoesa así *v*.; tsírɔ́n *v*. **go down** ƙoona ɡíɡìròk<sup>e</sup> *v*.; kídzimonuƙot<sup>a</sup> *v*.; raʝámón *v*.; raʝánón *v*. **go down (of sun in afternoon)** tɔɔsɔ́ŋɔ́n *v*. **go down (of sun)** itsólóŋòn *v*.; tɔɔnʉƙɔt<sup>a</sup> *v*. **go down (out of sight)** lakámétòn *v*.; lakámón *v*. **go down low** ɗipímón *v*. **go early** isókón *v*. **go extinct** bulonuƙot<sup>a</sup> *v*. **go for a visit** ɨlɔ́ɔ́n *v*.; ɨlɔ́ɔ́nʉƙɔt<sup>a</sup> *v*.; ipásóòn *v*.
**go for a walk** ɨɗámɔ́n *v*.; zíbòn *v*. **go get** ƙáidetés *v*. **go head-over-heels** tíbìɗìlɔ̀n *v*. **go horizontally** kámáránónuƙot<sup>a</sup> *v*. **go hungry** torónón *v*. **go in** ɓuƙétón *v*.; ɓùƙòn *v*.; ɓuƙonuƙot<sup>a</sup> *v*. **go in a rage** tʉlʉ́ŋɔ́n *v*. **go in convoy** ɨtílɔ́nʉƙɔt<sup>a</sup> *v*. **go in front** ƙoona wáxìk<sup>ɛ</sup> *v*. **go in procession** ɨtílɔ́nʉƙɔt<sup>a</sup> *v*. **go into exile** totóánonuƙot<sup>a</sup> *v*.; xàtsɔ̀n *v*. **go late** íbànòn *v*.; irípón *v*. **go near** ɦyɔtɔ́ɡɔnʉƙɔt<sup>a</sup> *v*. **go off (rot)** masánétòn *v*.; mʉsánétòn *v*. **go off course** kwɛ́dɔ̀n *v*. **go off in a rage** tʉlʉ́ŋɔ́nʉƙɔt<sup>a</sup> *v*. **go off one-by-one** kónionúƙot<sup>a</sup> *v*. **go off topic** hakonuƙot<sup>a</sup> *v*. **go off track** imámáɗós *v*. **go on** pórón *v*. **go on (to do)** itáƙúòn *v*. **go on a trip** ɨlɔ́ɔ́nʉƙɔt<sup>a</sup> *v*. **go out** pulúmón *v*. **go out (of fire)** ts'oonuƙot<sup>a</sup> *v*. **go out of sight** kúbonuƙot<sup>a</sup> *v*. **go over** ɡórés *v*.; íɡorés *v*.; íɡorésúƙot<sup>a</sup> *v*. **go over matters** bɛrɛ́sá mɛná<sup>ɛ</sup> *v*. **go past** ɓʉnɔnʉƙɔt<sup>a</sup> *v*. **go quickly** ikómóonuƙot<sup>a</sup> *v*. **go round and round** ɨlɔɗílɔ́ɗɔ̀n *v*. **go straight** isérérèòn *v*. **go through** ɨlámɔ́n *v*.; piɗés *v*.; pʉtʉ́mɔ́n *v*. **go to seed** eɡésá ekwí *v*.; tutufánón *v*. **go to sleep** eponuƙot<sup>a</sup> *v*. **go under (water)** ɓuƙonuƙot<sup>a</sup> *v*.
go up
**go up** ƙooná dìdìk<sup>e</sup> *v*.; tóbìrìbìròn *v*.; totírón *v*. **go via** tɔmɛɛs *v*.; tɔmɛ́ɛ́sʉƙɔt<sup>a</sup> *v*. **go with** elánónuƙot<sup>a</sup> *v*. **go-away bird (white-bellied)** ƙwáaƙwá *n*. **goad** ɨʝʉkʉ́ʝʉ́kɛ́s *v*. **goat (female)** rieŋwa *n*. **goat (funerary)** kɔ̀p <sup>a</sup> *n*.; ɲépúɲa *n*. **goat flock** riébàr *n*. **goat kid** riéím *n*. **goat(s)** ri *n*. **goat-leather bag** riéófúr *n*. **goatee** ɲɛ́pɛ́nɛk<sup>a</sup> *n*.; tɛ̀mʉ̀r *n*. **gobble down** ifáfúkés *v*.; ŋɔfɛ́s *v*. **gobbler (of food)** lòkòɗòŋìròàm *n*.; loƙeƙes *n*. **gobbling (of food)** lòkòɗòŋìrò *n*. **God** áméda kíʝá<sup>e</sup> *n*.; didiɡwarí *n*.; Ɲakuʝ<sup>a</sup> *n*. **god** Ɲakuʝ<sup>a</sup> *n*. **godhood** ɲakuʝínánès *n*. **godly person** ɲakuʝíám *n*. **gods** ɲakuʝíícík<sup>a</sup> *n*. **goer** ƙòònìàm *n*. **going for good** wìʉ̀ *ideo*. **goiter** ɓòlìɓòl *n*. **gold** ɡɔ́lìɗ<sup>a</sup> *n*. **gold dust** ɲɛ́ɲɨɲí *n*. **gold flecks** ŋkaɗɛɛɗɛ́y <sup>a</sup> *n*. **golden earring** ɲámaritóít<sup>a</sup> *n*. **goliath** bàd<sup>a</sup> *n*. **Gomphocarpus fruticosus** lɔsalát<sup>a</sup> *n*. **goo** ɡaɗár *n*. **good** maráŋón *v*. **good (become)** maráŋónuƙot<sup>a</sup> *v*.
**good (make)** maraŋités *v*.; maraŋítésuƙot<sup>a</sup> *v*. **good (of many)** dayaakón *v*.; maráŋaakón *v*. **good person** maráŋónìàm *n*. **good!** maráŋ *interj*. **good-looking** dòòn *v*. **goodness** maráŋás *n*. **goods** kúrúɓâd<sup>a</sup> *n*.; kúrúɓáicík<sup>a</sup> *n*. **gooey** ɗɔmɔ́dɔ̀n *v*.; nɨrídɔ̀n *v*.; xɔrɔ́dɔ̀n *v*. **gooily** ɗɔ̀m *ideo*.; nìr *ideo*.; xɔ̀r *ideo*. **goop** ɡaɗár *n*. **goopily** xɔ̀r *ideo*. **goopy** ɓɔrɔ́tɔ́mɔ̀n *v*.; bɔrɔ́tsɔ́mɔ̀n *v*.; xɔrɔ́dɔ̀n *v*. **goose** ɲáɓata *n*. **gorge** fòts<sup>a</sup> *n*.; ɲɔ́kɔ́pɛ̀ *n*. **gorged** itéɓúkòn *v*. **goshawk** tsìts<sup>a</sup> *n*. **gossamer** bɛɗɛ́dɔ̀n *v*. **gossip about** mɔ́ɲɛ́s *v*. **gossiper** ɡáʒadɨŋwa *n*.; sɛsɛanónìàm *n*. **gossipy** mɔɲɨmɔɲɔs *v*. **gourd** kàŋɛ̀r *n*. **gourd (big oblong)** ɲátúɗu *n*. **gourd (big round)** ɓoló *n*.; ɲápaɗɛr *n*. **gourd (bitter)** óbìʝòkwàts<sup>a</sup> *n*. **gourd (broken)** ƙwɛsɛ́*n*. **gourd (butter)** ɗɛ̀k <sup>a</sup> *n*. **gourd (cracked)** naturutur *n*. **gourd (dried)** ìƙòlòt<sup>a</sup> *n*. **gourd (edible)** lomuƙe *n*.; lɔ́pʉ́l *n*. **gourd (flask)** ɲekúrúm *n*. **gourd (funnel-stemmed)** lɔkʉtʉ́r *n*. **gourd (milking)** ɲelépít<sup>a</sup> *n*. **gourd (round)** ƙɔfɔ́*n*. **gourd (small round)** dúlúƙuƙú *n*.; ƙɔfóìm *n*.; túkulét<sup>a</sup> *n*.
grasping
**gourd (small-mouthed)** loƙú *n*. **gourd (wide-bottom)** ɗùt<sup>a</sup> *n*. **gourd (wide-mouthed)** ƙoƙó *n*.; rìƙòŋ *n*. **gourd basin** sèrèy<sup>a</sup> *n*. **gourd bowl** ƙɔfɔ́*n*. **gourd enema** bulukét<sup>a</sup> *n*. **gourd flask** bòrèn *n*.; nasɛmɛ́*n*. **gourd flask (small)** ɲékútàm *n*. **gourd flesh** ɗɔ̀xɔ̀ƙ <sup>a</sup> *n*. **gourd ladle (broken)** tebeleƙes *n*. **gourd leaves** ŋátuɓe *n*. **gourd plug** akatɛ́t <sup>a</sup> *n*. **gourd pulp** ɗɔ̀xɔ̀ƙ <sup>a</sup> *n*. **gourd rattle** ɲɛ́ɛ́ƙɨɛ́ƙ <sup>a</sup> *n*. **gourd scoop** ɓòlòkòts<sup>a</sup> *n*. **gourd seed** kàŋɛ̀rèèkw<sup>a</sup> *n*. **gourd spout** sɔ́k <sup>a</sup> *n*. **gourmand** lòkòɗòŋìròàm *n*.; loƙeƙes *n*. **govern** ipúkéés *v*. **government** ɲápukán *n*.; ɲeryaŋ *n*. **government (British colonial)** ɡɛrɛ́sà *n*. **government employee** ɲápukáníàm *n*. **government official** ámázeáma ɲápukaní *n*.; túbesiama ɲápukání *n*. **governor** ipúkéésíàm *n*.; tòtwàrààm *n*. **gown** ɲékiteitéy<sup>a</sup> *n*. **grab** ídadamɛ́s *v*.; ɨkamɛ́sʉ́ƙɔt<sup>a</sup> *v*.; ɨkamɛtɛ́s *v*.; ɨrɛɗɛs *v*.; ŋusés *v*.; ŋusésúƙot<sup>a</sup> *v*.; taŋates *v*.; tokopes *v*. **grab away** taŋátésuƙot<sup>a</sup> *v*.; tokópésuƙot<sup>a</sup> *v*. **grab repeatedly** ɨkamíkámɛ́s *v*. **grabbiness** dzɔɗátínànès *n*. **grabby** ɨtɔ́kɔ́ánòn *v*.; ts'íts'ɔ́n *v*. **grace (Catholic)** ɲɛ́ɡɨrasíà *n*. **grade (roads)** séɓés *v*.
**grade (school)** hò *n*. **grader (of roads)** séɓésìàm *n*. **grain** eɗ<sup>a</sup> *n*. **grain (fallen)** sɛkɛmán *n*. **grain (new)** eɗa ni erúts<sup>a</sup> *n*. **grain (regrown)** lɔ́ʝɔʝɔ́*n*.; ɲópoté *n*. **grain beer** eɗímɛ́s *n*. **grain disease** lɔɓɛlɛɲ *n*. **grain harvest (first)** ƙɔfɔ́èɗ<sup>a</sup> *n*. **grain thief** lokoɓél *n*. **grain-filled dirt** ɓɔɗáʝʉ́m *n*. **grainy** sɨƙɨsíƙánétòn *v*.; síƙísɨƙánón *v*. **Gramineae species** ʝay<sup>a</sup> *n*.; kɔlɨlíkú *n*.kwɨnɨƙíkú *n*.; ; lɔ̀tàfàr *n*.; máyákù *n*.; mɨlíár *n*.; ɲálaaʝáít<sup>a</sup> *n*.; ɲámá *n*.; ɲèràkù *n*.; ɲokoɗopey<sup>a</sup> *n*.; sòlìsòl *n*.; tɨɓ<sup>a</sup> *n*. **granary** loɗúrú *n*. **granary (small)** lɔkʉ́ɗ <sup>a</sup> *n*. **granary base** loɗúrúdɛ̀ *n*. **granary cover** ɲésóto *n*.; ɲósóto *n*. **granary reed** loɗúrúkɔ̀k <sup>a</sup> *n*. **granary reed (vertical)** ɲétémets<sup>a</sup> *n*. **grand** zòòn *v*. **grandchild** lobá *n*. **grandchild (his/her)** ntsílóbà *n*. **grandchild (my)** ɲ́cilobá *n*. **grandchild (your)** bilóbà *n*. **grandfather (his/her)** bɔbat<sup>a</sup> *n*. **grandfather (my)** bɔbá *n*. **grandfather (your)** bɔ́bɔ̀ *n*. **grandmother (his/her)** dádàt<sup>a</sup> *n*. **grandmother (my)** dadáŋ *n*. **grandmother (your)** dádò *n*. **granite** rikírík<sup>a</sup> *n*. **grapes** ɲéviiní *n*. **grasp** ŋɔtsɛ́s *v*.; tɔkɔɗɛs *v*. **grasping** ɨtɔ́kɔ́ánòn *v*.
grass
**grass** kù *n*. **grass (cow-broom)** ɦyɔ̀ʝàn *n*. **grass (matted)** kémús *n*. **grass (soft)** ûd<sup>a</sup> *n*. **grass (thick dry)** sakátán *n*. **grass hut** ɲérwám *n*. **grass patch** xʉram *n*. **grass shelter** ɲɛ́kɨsakát<sup>a</sup> *n*. **grass species** fàɗìɡùr *n*.; ʝay<sup>a</sup> *n*.; kaxɨt<sup>a</sup> *n*.; kwɨnɨƙíkú *n*.; loiɓóròk<sup>a</sup> *n*.; lɔ́mɔ́ɗaát<sup>a</sup> *n*.; lɔ̀tàfàr *n*.; lɔ́tílɨtíl *n*.; lɔtsɔ́ɡɔ̀m *n*.; máyákù *n*.; mèlèt<sup>a</sup> *n*.; mɨlíár *n*.; mʉrɔn *n*.; ɲálaaʝáít<sup>a</sup> *n*.; ɲámá *n*.; ɲamaɗaŋíkú *n*.; ɲaɲɛnɨʝɛ́n *n*.; ɲeɗuar *n*.; ɲèràkù *n*.; ɲéuɗuúɗu *n*.; ɲokoɗopey<sup>a</sup> *n*.; ɔlí *n*.; òŋòrìkù *n*.; sòlìsòl *n*.; tɨɓ<sup>a</sup> *n*. **grass stub** rumurúm *n*. **grasshopper** ƙɛƙɛ́r *n*. **grassland** dús *n*. **grate** fɔ́fɔ́tɛ́s *v*.; ɨfɔɛs *v*. **gratifyingly** wìɲ *ideo*. **gratifying (sensually)** wɨɲídɔ̀n *v*.; wɨɲɨwíɲɔ́n *v*. **gratis** tsàm *n*. **gratuitousness** tsàm *n*. **grave** rip<sup>a</sup> *n*.; tás *n*.; tásɛ̂d <sup>a</sup> *n*. **grave (pile of stones)** ɡwàsìkàkìts<sup>a</sup> *n*. **gravedigger** muɗésíàm *n*.; tʉnʉkɛsíám *n*. **gravedigger prayer** wáána na muɗésíàmà<sup>e</sup> *n*. **gravel** aŋaras *n*. **gravelly** ŋàr *ideo*.; ŋarʉ́dɔ̀n *v*.; rakákámòn *v*. **gravelly area** aŋarasááƙw<sup>a</sup> *n*. **graverobber** tukutesíáma ts'óóniicé *n*. **gravesite** tás *n*.; tásɛ̂d <sup>a</sup> *n*.
**gravitate** irídòn *v*. **gravy** ɲɔ́ɓɔ́ka *n*. **gray** ɔŋɔ́ránètòn *v*. **gray (of weather)** kùpòn *v*. **gray-brown** ɔŋɔ́ránètòn *v*. **gray-haired** kwɛrɛ́xɔ́n *v*. **graze** waitetés *v*. **grazing** wà *n*. **grease** ŋiites *v*. **grease up** ŋiitésúƙot<sup>a</sup> *v*. **greasy** kùx *ideo*. **great** zòòn *v*. **greatness** zeís *n*. **greediness** dzɔɗátínànès *n*. **greedy** ts'íts'ɔ́n *v*.; tírésa dzɔɗátí *v*. **greedy person** kìɓèɓèàm *n*. **greedyguts** loúk<sup>a</sup> *n*. **Greek language** Ŋíɡɨríkìtòd<sup>a</sup> *n*. **Greek person** Ŋíɡɨríkìàm *n*. **green** ɨlíɓɔ́n *v*. **green (of many)** ɨlɨɓaakón *v*. **green (turn)** ɨlíɓɔ́nʉƙɔt<sup>a</sup> *v*. **green (turn, of many)** ɨlɨɓaakónuƙot<sup>a</sup> *v*. **green (very)** ƙwɨxídɔ̀n *v*.; xídɔ̀n *v*. **greenly** ƙwìx *ideo*. **greens** wà *n*. **greet** ɨmáxánɛ́s *v*. **greet each other** ɨmáxánínɔ́s *v*. **Grewia bicolor** ʝàw<sup>a</sup> *n*. **Grewia tenax** alárá *n*.; ɔɡɔn *n*. **Grewia villosa** mɔ̀z *n*. **grieve** turúnón *v*. **grill** ɨtɔlɛs *v*. **grill (of vehicle)** sarísárìk<sup>a</sup> *n*. **grilled** ɨtɔlɔs *v*. **grimace** ɨɲɛ́ɓɛ́rɛ́s *v*. **grime** ts'âɡ<sup>a</sup> *n*.
#### grimy
**grimy** ts'áɡòn *v*. **grin** ɨmʉ́mʉ́ɔ̀n *v*. **grind** ŋɔ́ɛ́s *v*. **grind (make)** ŋɔítɛ́sʉƙɔt<sup>a</sup> *v*. **grind again** iŋáɓúkés *v*. **grind coarsely** ɨkaŋíkáŋɛ́s *v*.; ɨŋáámɛ́s *v*.; ɨŋaíŋɛ́ɛ́s *v*.; ɨŋáŋɛ́ɛ́s *v*. **grind finely** ɨwɨɗɛs *v*. **grind quickly** ɨpʉnɛs *v*. **grind well** hamʉʝɛ́s *v*. **grindable** ŋɔam *n*. **grinding mill** ŋɔɛsíɡwàs *n*.; ɲámasín *n*. **grinding stone** ɡwas *n*.; ŋɔɛsíɡwàs *n*. **grinding stone (hand-held)** imeda ɡwasá<sup>e</sup> *n*.; kɨnata *n*. **grinding stone (lower)** ŋwááteda ɡwasá<sup>e</sup> *n*. **grip** ɨkamɛs *v*.; ɨkamɛ́sʉ́ƙɔt<sup>a</sup> *v*.; ɨkamɛtɛ́s *v*.; ɨtɔkɔɗɛs *v*.; tírésìàw<sup>a</sup> *n*. **grip repeatedly** ɨkamíkámɛ́s *v*. **grist** kabas *n*. **grist (dry)** ŋápʉp<sup>a</sup> *n*. **grist (moist)** ŋámírɔ̀ *n*. **gristle** ŋɔrɔɓɔɓ<sup>a</sup> *n*. **gristlely** rɔ̀ɓ ɔ *ideo*. **gristly** rɔɓɔ́dɔ̀n *v*. **gritty** ɡwɛrɛ́ʝɛ́ʝɔ̀n *v*. **grizzly** kwɛrɛ́xɔ́n *v*. **groan** éɓútòn *v*.; ɛ́mítɔ̀n *v*.; émúròn *v*. **groggy** mususánón *v*. **groin** lɔkɔ́r *n*. **groom** bɔƙátíníèàkw<sup>a</sup> *n*. **grope** ídadamɛ́s *v*. **grope (sexually)** tárábes *v*. **grope all over** tárábiés *v*. **groping** ídadamɔ́s *v*.
**grouchy** ɡaƙúrúmòn *v*.; ŋízìmɔ̀ɔ̀n *v*. **ground** ʝʉm *n*.; ŋɔ́ɔ́s *v*. **ground (anchor)** ɲólóit<sup>a</sup> *n*. **ground (burnt)** kàròƙ<sup>a</sup> *n*. **ground (newly broken)** túbùr *n*. **ground bee** mɔ́ɗ <sup>a</sup> *n*. **ground bee species** ɲásaŋáɲo *n*.; ɲésíìt<sup>a</sup> *n*. **ground beehive** kùkùsèn *n*. **ground finely** ɨwɨɗɔs *v*. **ground level (be at)** ràtòn *v*. **groundnut(s)** ɲépulé *n*.; taráɗá *n*. **grounds** awááƙw<sup>a</sup> *n*. **groundwater** ʝʉmʉ́cúé *n*. **groundwork** dɛ̀ɛ̀dà hò<sup>e</sup> *n*. **group** ɲáʝore *n*.; ɲéɡurúf *n*. **group (social)** kábùn *n*. **group discussion** natúk<sup>a</sup> *n*. **grove** ɡwi *n*. **grow** ɨatímétòn *v*.; morétón *v*.; zeites *v*.; zeitésuƙot<sup>a</sup> *v*. **grow back** ʝɔɓɛ́tɔ́n *v*.; tɔrʉ́ɓɔ́n *v*. **grow bigger** zoonuƙot<sup>a</sup> *v*. **grow bushy** tsekétón *v*. **grow furry** tsekétón *v*. **grow hairy** tsekétón *v*. **grow high** zikíbètòn *v*. **grow in age** zoonuƙot<sup>a</sup> *v*. **grow limber** tsutsukúmétòn *v*. **grow long** zikíbètòn *v*. **grow old** dunétón *v*. **grow over** tsekétón *v*. **grow supple** tsutsukúmétòn *v*. **grow tall** tɔwʉ́tɔ́n *v*.; zikíbètòn *v*. **grow up** iríétòn *v*.; zeetón *v*. **grow up and back (of horns)** tʉ́zʉ̀ŋɔ̀n *v*. **growl** ŋɔ́rɔ́rɔ̀n *v*.
**growl (of stomach)** ɗuɗuanón *v*. **grown underground** ɨʝɛ́ɛ́lɔ̀n *v*. **grown up (of many)** zeikaakón *v*. **grown-up** ámáze *n*. **grub** ƙʉts'<sup>a</sup> *n*. **grub (food)** ŋƙáƙ<sup>a</sup> *n*. **grub (rhinocerus beetle)** lóɓúlukúɲ *n*. **gruel** ŋáítɔ̀ *n*.; ɲéúʝi *n*. **gruff (of voice)** ƙoƙórómòn *v*.; rɔ́ƙɔ́rɔƙánón *v*. **grumble** ɨŋʉrʉ́ŋʉ́rɔ̀n *v*. **grumble (of stomach)** ɗuɗuanón *v*. **grumble to oneself** ɲɛɓɛ́sá tódà<sup>e</sup> *v*. **grume** ŋazul *n*. **grumpy** ɡaƙúrúmòn *v*.; ŋízìmɔ̀ɔ̀n *v*. **grunt** ɨɗíɲɔ́n *v*. **grunt in pain** rúɓón *v*. **guard** còòkààm *n*.; cookés *v*.; ɨrɨtsɛ́s *v*.; ŋasíƙáárìàm *n*. **guard (local council)** ŋíɲampáràyàm *n*. **guard (prison)** cookaama zíkɛ́siicé *n*. **guarded** cookotós *v*. **guardian** còòkààm *n*. **guava** ɲóɡóva *n*. **guest** wáánààm *n*. **guest (long-term)** ɦyekesíám *n*. **guide** ɨtsírítɛtɛ́s *v*.; tsírítɛtɛ́s *v*.; tɔɓɛɨtɛtɛ́s *v*.; tòrìkààm *n*.; torikes *v*.; torikesíám *n*. **guide away** toríkésuƙot<sup>a</sup> *v*. **guide this way** toriketés *v*. **guile** nɔɔ́s *n*. **guileful** nɔɔsánón *v*. **guineafowl (helmeted)** ʝáɓúɡwà *n*. **guitar** ɲéɡitá *n*. **gullet** moróká na kwáts<sup>a</sup> *n*. **gully** ɲéƙúrumot<sup>a</sup> *n*.; urúr *n*.
**gulp** áɡʉʝɛ́s *v*.; ɡéɡès *v*.; itúlákáɲés *v*. **gulp down** lakatiés *v*.; lukutiés *v*. **gulp!** ɡʉ̀lʉ̀ʝᶶ *ideo*. **gum** ɲáɡám *n*. **gum (chewing)** ɗɔtɔ́*n*. **gum (food)** iŋulúŋúlés *v*. **gum (of mouth)** diriʝiʝ<sup>a</sup> *n*. **gum (of trees)** ɗòs *n*. **gum (rubbery)** ɗɔtɔ́*n*. **gummily** ɗɔ̀s *ideo*. **gummy** ɗɔsɔ́dɔ̀n *v*.; mɨníkímɔ̀n *v*. **gun** ɛ̂b <sup>a</sup> *n*. **gun (homemade)** ɲamatiɗa *n*. **gun (large-bore)** ɲáturuɡéy<sup>a</sup> *n*. **gun (long-barreled)** ɲɛ́pɛn *n*. **gun sight** ɲɛ́lɨmɨrá *n*. **gun type** ɲáɗúle *n*. **gunkily** xɔ̀r *ideo*. **gunky** ɓɔrɔ́tɔ́mɔ̀n *v*.; xɔrɔ́dɔ̀n *v*. **gunny sack** ɲɛ́ɗɛpɨɗɛ́p <sup>a</sup> *n*.; ɲéɡuniyá *n*. **gunny sack (large)** lomóŋin *n*.; ɲáwaawá *n*. **gunpowder** leúzìn *n*. **gunstock** ɛ́bàdɛ̀ *n*. **gurgle** ábʉ̀bʉ̀ƙɔ̀n *v*. **gurgle (of stomach)** ɗuɗuanón *v*. **gut** ɓilésúƙot<sup>a</sup> *v*.; bùbù *n*.; bùbùàƙw<sup>a</sup> *n*. **gutter** sɔ́k <sup>a</sup> *n*. **guy** ɲɛ́ɛ́s *n*. **guzzle** áɡʉʝɛ́s *v*.; ɡéɡès *v*.; íɡʉʝɛ́s *v*. **gwuf gwuf** tsèfètsèf *ideo*. **habit** ɲɛpɨtɛ *n*. **habitation** zɛƙɔ́áw<sup>a</sup> *n*. **habituate** ɨtalɛs *v*.; naínɛ́ɛtɛ́s *v*.; naítɛ́sʉƙɔt<sup>a</sup> *v*. **habituated** ɨtalɔs *v*. **hacksaw** ɲɛ́mʉsʉmɛ́n *n*.
**haggard** ɨkárɔ́n *v*.; kɔrɔ́ɗɔ́mɔ̀n *v*. **hail** tìkɔ̀r *n*. **hair** síts'<sup>a</sup> *n*. **hair (of a baby)** imásíts'<sup>a</sup> *n*. **hair (of head)** ikásíts'<sup>a</sup> *n*. **hair (pubic)** didisísíts'<sup>a</sup> *n*.; ɔ́zàsìts'<sup>a</sup> *n*.; tɛ̀mʉ̀r *n*. **hair (stiff)** ɲéɡets<sup>a</sup> *n*. **hair column** ɲotókósit<sup>a</sup> *n*. **hair follicle** síts'ádɛ̀ *n*. **hair jewelry** ɲéméle *n*. **hair ridge** sìɡìrìɡìr *n*. **hair-patch (styled)** nàɗìàk<sup>a</sup> *n*. **hairstyle** ɲékiɲés *n*. **hairy** saúkúmòn *v*.; tsèkòn *v*. **half** ɲɛ́nʉ́s *n*.; xɔnɔ́ɔ́kɔn *n*. **half (be a)** kɔnɔna leɓétsónù *v*. **half-asleep state** mɔɗɔ́ɗɔ́èkw<sup>a</sup> *n*. **half-awake state** mɔɗɔ́ɗɔ́èkw<sup>a</sup> *n*. **half-cooked** tsɛɓɛ́kɛ́mɔ̀n *v*. **half-striped** ŋurutiós *v*. **halfway** sɨsɨkák<sup>ɛ</sup> *n*. **halfway point** sɨsíkɛ̂d <sup>a</sup> *n*. **halt** wasɨtɛs *v*.; wasítɛ́sʉƙɔt<sup>a</sup> *v*.; wasɔnʉƙɔt<sup>a</sup> *v*. **halting** was *n*. **hammer** ityakes *v*.; ɲɛ́ɲɔnɗɔ́*n*.; tɔ́ts'ɛ́s *v*. **hammerstone (black)** sàbàɡwàs *n*. **hamper** ɨɓatɛs *v*. **hamper repeatedly** ɨɓatíɓátɛ́s *v*. **hampered** ɨɓatíɓátɔ̀n *v*. **hand** ɨkɔɓɛs *v*.; kwɛt<sup>a</sup> *n*. **hand (left)** betsínákwɛ̀t <sup>a</sup> *n*. **hand over here** ɨkɔɓɛtɛ́s *v*. **hand over there** ɨkɔ́ɓɛ́sʉƙɔt<sup>a</sup> *v*.
**hand-crafted** ɨɗɨmɔtɔ́sá kwɛ̀tìk<sup>ɔ</sup> *v*. **hand-held radio** dʉrʉdʉra na tímoí *n*. **handbag** ɲáníɓàk<sup>a</sup> *n*. **handful** ɨlʉlʉŋam *n*. **handgrip** ɲɔkɔ́ɗɛ́t <sup>a</sup> *n*.; tírésìàw<sup>a</sup> *n*. **handgun** ɲɛ́písítɔ̀l *n*. **handicap** ɨɓatɛs *v*.; ŋwaxás *n*. **handicap repeatedly** ɨɓatíɓátɛ́s *v*. **handicapped** ɨɓatíɓátɔ̀n *v*.; ŋwàxɔ̀n *v*. **handicapped person** ŋwàxɔ̀nìàm *n*. **handkerchief** ɲákitamɓára *n*. **handle** dɛɛd<sup>a</sup> *n*.; ɨkamɛs *v*.; ɲɔkɔ́ɗɛ́t <sup>a</sup> *n*.; tírés *v*.; tírésìàw<sup>a</sup> *n*. **handle (borehole)** ɲatsʉʉmákwɛ́t <sup>a</sup> *n*. **handle (hoe)** ɲɛ́mɛlɛkʉ́dàkw<sup>a</sup> *n*. **handle (manage)** totseres *v*. **handle carefully** ɨɓáɓɛ́ɛ́s *v*. **handmade** ɨɗɨmɔtɔ́sá kwɛ̀tìk<sup>ɔ</sup> *v*. **handmade object** kwɛtákɔ́rɔ́ɓád<sup>a</sup> *n*. **handsaw** ɲɛ́ƙɨrɨƙír *n*. **handsome** dòòn *v*. **handsomeness** daás *n*. **handyman** ŋífunɗíàm *n*. **hang (kill)** ɨkɛtɛs *v*. **hang around** toƙízòòn *v*.; tɔʉ́rʉ́mɔ̀n *v*. **hang back** isíɗóòn *v*. **hang by tucking** rʉ́bɛ̀s *v*. **hang freely** alólóánón *v*. **hang in there** taɗáŋón *v*. **hang low with weight** xuƙúmánòn *v*. **hang on to** ɨrɨtsɛ́s *v*. **hang oneself** ɨkɛtɛsa así *v*. **hang out** ɨtɛ́mɔ́ɔ̀n *v*.; itúmétòn *v*. **hang up** inénéés *v*.; inénéésuƙot<sup>a</sup> *v*.; xɨkɛ́s *v*.; xɨkɛ́sʉ́ƙɔt<sup>a</sup> *v*.
**hang up (in storage)** tɨkɨɛtɛ́s *v*. **hang up by tucking** rʉ́bɛsʉƙɔt<sup>a</sup> *v*. **hankering (have a)** ɨrɔ́rɔ́kánón *v*. **Haplocoelum foliolosum** ŋʉrʉ́sá *n*. **happen** ikásíìmètòn *v*.; itíyáìmètòn *v*. **happen again** tɔrʉ́ɓɔ́n *v*. **happen quickly** ɗoriɗórón *v*. **happen upon** bunétón *v*.; bùnòn *v*.; ŋawɨlɛs *v*.; taƙámón *v*. **happiness** ɲalakas *n*. **happy** eaŋanes *v*.; ɨlákásɔ̀n *v*. **happy (become)** ɨlákásɔ́nʉƙɔt<sup>a</sup> *v*. **happy (make)** ɨlákásítɛ́sʉƙɔt<sup>a</sup> *v*. **harass** isyees *v*.; tɛɲɛfɛs *v*. **harass each other** tɛɲɛ́fʉ́nɔ́s *v*. **hard** ŋìxɔ̀n *v*.; tɔrɔ́dɔ̀n *v*. **hard (difficult)** itíónòn *v*. **hard (filled)** dirídòn *v*. **hard (impenetrably)** ɡɔɓɔ́dɔ̀n *v*. **hard (make)** ŋɨxítɛ́sʉƙɔt<sup>a</sup> *v*. **hard (of wood)** dèwòn *v*. **hard (of a substance)** lɛrɛ́dɔ̀n *v*. **hard-of-hearing** ilios *v*. **hard-working** ɨɓɛ́rɔ́ánón *v*. **harden** ŋɨxítɛ́sʉƙɔt<sup>a</sup> *v*.; ŋɨxɔnʉƙɔt<sup>a</sup> *v*. **harder (become)** ŋɨxɔnʉƙɔt<sup>a</sup> *v*. **hardheadedness** ŋɨxɔna iká<sup>e</sup> *v*. **hardly** ɡɔ̀ɓ ɔ *ideo*.; lɛ̀r *ideo*. **hardship** ŋítsan *n*. **hare** tulú *n*. **hare nickname** bositíníàm *n*. **harm** tawanes *v*.; tawánítetés *v*. **harmfulness** ƙʉts'ánánès *n*. **harmonious** ɨsílɔ́n *v*. **harmonize** ɨsílítɛ́sʉƙɔt<sup>a</sup> *v*.
**Harrisonia abyssinica** kèlèrw<sup>a</sup> *n*. **harrumph** hákátòn *v*.; xaƙarés *v*. **harry** isyees *v*. **haruspicate** ɦyeitésá arííkà<sup>ɛ</sup> *v*. **harvest** ɨrárátés *v*.; ɨrarɛs *v*.; karɔŋ *n*.; tarares *v*.; watsóy<sup>a</sup> *n*.; weés *v*. **harvest (honey)** ɗusés *v*. **harvest (wild food)** hakwés *v*. **harvest bountifully** cɛɛtɛ́s *v*. **harvest millet** ɨrábɛs *v*. **harvest of new grain** eɗa ni erúts<sup>a</sup> *n*. **harvest termites** hakwésá dáŋá<sup>e</sup> *v*. **harvest time** karɔŋ *n*.; watsóy<sup>a</sup> *n*. **harvester** weésíàm *n*. **has not** máa *adv*. **hassle** tɛɲɛfɛs *v*. **hassle each other** tɛɲɛ́fʉ́nɔ́s *v*. **hasten** ɨɓʉrʉ́ɓʉ́rɔ̀n *v*.; ikómóòn *v*. **hasten here** ikóméètòn *v*. **hasten there** ikómóonuƙot<sup>a</sup> *v*. **hat** ɲákopiyá *n*. **hat (wide-brim)** ɲákakar *n*. **hatch (of chicks)** ɓɛkɛ́tɔ́n *v*.; ɨɓɛ́ɓɔ́ɔ̀n *v*. **hatch out (of chicks)** ɨɓɛ́ɓɛ́ɛ̀tɔ̀n *v*. **hatchet** dzibér *n*. **hatchet (traditional)** kuɲukúdzibér *n*. **hate** ts'ábès *v*.; takaɗes *v*. **hate each other** dɛŋʉ́ɲʉ́nɔ́s *v*.; ts'ábunós *v*. **have** ɡirés *v*.; iona ńdà *v*.; tírés *v*. **have been around** kɔ̀wɔ̀n *v*. **have circles** tɔlʉkʉ́lʉ́kɔ̀n *v*. **have contractions** wúrukukánón *v*. **have free time** ipásóòn *v*. **have fun** iyóómètòn *v*.
heave away
**have hidden motives** dzoluɡánón *v*. **have mercy on** iƙenes *v*. **have multiples** ramɛ́s *v*. **have not** máa *adv*. **have pain** toryáɓón *v*. **have problems** iona ŋítsaník<sup>ɛ</sup> *v*. **have sex** èpòn *v*. **have sex (frequently)** ɓútánés *v*. **have sex with** tirés *v*. **have sex with each other** tirímós *v*. **have time to walk around** zíbos *v*. **have unprotected sex** ts'íts'ɔ́n *v*. **hawk** hákátòn *v*.; tsìts<sup>a</sup> *n*.; xaƙarés *v*. **hawker** ŋímutsurúsìàm *n*. **hawker (being a)** ŋímutsurúsìnànès *n*. **haze (atmospheric)** súm *n*. **hazy** imítíròn *v*. **he** nts<sup>a</sup> *pro*. **head** ik<sup>a</sup> *n*.; iked<sup>a</sup> *n*. **head of beer** ikeda mɛ́sɛ̀ *n*. **head of trail** mucédɛ̀ *n*. **head honcho** ámáze *n*.; ámázeám *n*. **head-butt** tɔɗɔ́pɔ́n *v*. **head-pad** ɨkɨt<sup>a</sup> *n*. **headband (beaded)** ɲɛ́wakɔ́l *n*. **headband (men's)** ɲɛcáát<sup>a</sup> *n*. **heading** iked<sup>a</sup> *n*. **headlight** ɲómotokéèkw<sup>a</sup> *n*. **headmaster** ámázeáma ɲésukúluⁱ *n*. **headrest** dɨƙwam *n*.; kàràts<sup>a</sup> *n*. **headstart (get a)** iɗílón *v*. **headway on (make)** iɗiles *v*. **heal** maraŋítésuƙot<sup>a</sup> *v*.; maráŋónuƙot<sup>a</sup> *v*.; mínɛ́s *v*. **heal up** iŋáléètòn *v*.; toíónuƙot<sup>a</sup> *v*.
**healer (traditional)** cɛmɛ́ríkààm *n*.; irésíàm *n*.; ŋƙwa *n*.; wetitésíàm *n*. **health center** ɗakɨtár *n*. **healthily** ɗòx *ideo*. **healthy** ɗoxódòn *v*.; iŋáléòn *v*.; nɛsɛƙánón *v*.; zízòn *v*. **healthy (get)** zízonuƙot<sup>a</sup> *v*. **healthy (keep)** ɨrɨtsɛ́sá ɲeɗekéícíká<sup>o</sup> *v*. **healthy person** zízònìàm *n*. **heap** ɨnʉkʉ́nʉ́kɛ́s *v*.; itukes *v*.; kìts<sup>a</sup> *n*.; ɲatúkít<sup>a</sup> *n*.; tutukesíáw<sup>a</sup> *n*. **heap on** iɗóɗókés *v*. **heap up** iruketés *v*.; ituketés *v*.; kitsetés *v*.; tutuketés *v*. **heaped up** iɗóɗókánón *v*.; tutukánón *v*. **hear** nesés *v*.; nesíbes *v*. **heart (physical)** ɡúróèɗ<sup>a</sup> *n*. **heart (soul)** ɡúr *n*. **heart fat** máxìŋ *n*. **heart-rot (have)** ɓʉɓʉsánón *v*. **heartache** áts'ɛ́sa ɡúró<sup>e</sup> *n*. **heartburn** áts'ɛ́sa ɡúró<sup>e</sup> *n*.; kíɓɔ́ɔ̀z *n*. **hearth** ts'aɗíékw<sup>a</sup> *n*. **hearthstone** caál *n*. **heartsick** moona ɡúró<sup>e</sup> *v*. **heartwood** ɡúréda dakwí *n*. **heat** ɨmaɗɛs *v*. **heat up** hábètòn *v*.; hábitésúƙot<sup>a</sup> *v*.; hábonuƙot<sup>a</sup> *v*.; ɨmáɗɛ́sʉƙɔt<sup>a</sup> *v*.; ɨmɔ́lɔ́ŋɛtɛ́s *v*. **heated (angry)** ɨlílíɔ̀n *v*. **heated (become)** ɨlílíɔnʉƙɔt<sup>a</sup> *v*. **heathen** ŋíkafírìàm *n*. **heave** ʝaƙátós *v*.; ɲɛɲɛrɛs *v*.; toremes *v*.; xáƙátòn *v*. **heave away** ɨtsɔ́rɛ́sʉƙɔt<sup>a</sup> *v*.
#### Heaven
**Heaven** awa ná zè *n*.; awa Ɲákuʝí *n*.; didiɡwarí *n*.; ɡìdòɔ̀k <sup>a</sup> *n*. **heavens** lúl *n*. **heavily** lìr *ideo*. **heavy** ìsòn *v*. **heavy (make)** ɨnʉ́ítɛ́sʉ́ƙɔt<sup>a</sup> *v*.; isites *v*. **heavy-laded** ɨʉ́ƙɔ́n *v*. **Hebrew language** Ŋíyuɗáítòd<sup>a</sup> *n*. **Hebrew person** Ŋíyuɗáíàm *n*. **Hebrews (biblical)** Ŋíeɓuráiik<sup>a</sup> *n*. **hedge** itsóɗón *v*. **hedgehog (four-toed)** náabʉ́s *n*. **hedging in (of prey)** nakítsòɗ<sup>a</sup> *n*. **hee-haw** werétsón *v*. **heed** nesíbes *v*. **heedlessly** càc<sup>ɨ</sup> *adv*.; fùts'àts'<sup>a</sup> *ideo*.; tsàr *ideo*. **heel** tɨtíʝ<sup>a</sup> *n*. **heft** ɲɛɲɛrɛs *v*.; toremes *v*. **heftily** bɛ̀f *ideo*. **hefty** bɛfʉ́dɔ̀n *v*.; bɛfʉ́kʉ́mɔ̀n *v*. **height** zikíbàs *n*. **heir** ɨrɨtsɛ́síàm *n*. **heist** ɨmɔ́ɗɛ́sʉƙɔt<sup>a</sup> *v*. **Helianthus species** ɲɛ́kɨɗɛkɨɗɛ́ *n*.; ɲɛ́tɔɔkíɗɛ́*n*. **Helichrysum odoratissimum** ɲéúlam *n*. **helicopter** naƙílɨƙíl *n*. **help** bírɛ́s *v*.; ɨŋaarɛ́s *v*. **help each other** ɨŋáárínɔ́s *v*. **help give birth** ɨŋaarínɔ́sá ƙwaaté<sup>o</sup> *v*. **helped** ɨŋaarímétòn *v*. **helper** ɨŋaarɛ́síàm *n*. **hem** kweeda ƙwázà<sup>e</sup> *n*. **hem (bottom)** sɔka na ƙwázà<sup>ɛ</sup> *n*. **hemispherical** loŋórómòn *v*.
**hemp** ɲábaŋɡí *n*. **hen** ɡwaŋwa *n*. **henhouse** ɲɔ́kɔkɔrɔ́hò *n*. **her** nts<sup>a</sup> *pro*. **herald** síráàm *n*.; síránòn *v*. **herb (medicinal)** cɛ̀mɛ̀r *n*. **herbalist** cɛmɛ́ríkààm *n*.; wetitésíàm *n*. **herd** bàr *n*. **herd (small)** bàròìm *n*. **herd of cattle** ɦyɔ̀bàr *n*. **herds** ŋíɓarɛn *n*. **here** naíké *dem*.; náíta na *dem*.; nayá *n*.; nayé *dem*.; nayé na *dem*.; nɔ́ɔ́ *dem*. **here you go!** ne *interj*. **hereǃ** ne *interj*. **hermitic** iɓóótánón *v*. **herniate** ɨtsʉ́bʉ̀ɗɔ̀n *v*. **herniated** ɨtsʉ́bʉ̀ɗʉ̀mɔ̀n *v*. **herpes (of lips)** ɲótóts<sup>a</sup> *n*. **hers** ntsɛ́n *pro*. **herself** nébèd<sup>a</sup> *n*.; ntsínêb<sup>a</sup> *n*. **hesitate** ɨmɔ́mɛ́tɔ̀n *v*.; isíƙóòn *v*.; itóŋóòn *v*.; mɔ́mɛ́tɔ̀n *v*.; wasɛ́tɔ́n *v*. **hex** ipeɗes *v*.; sʉɓɛ́s *v*. **hexer** ìpèɗààm *n*.; sʉɓɛ́síàm *n*. **hey!** hèz *interj*. **hibernate** ɨtáƙálɛ́sá así *v*. **Hibiscus cannbinus** ɔ́bɛ̀r *n*. **Hibiscus esculentus** ɲɔlɔlɔt<sup>a</sup> *n*. **hiccough** xíƙón *v*.; xìƙw<sup>a</sup> *n*.; xíƙwítòn *v*. **hiccup** xíƙón *v*.; xìƙw<sup>a</sup> *n*.; xíƙwítòn *v*. **hidden** budésón *v*.; búdòs *v*. **hidden (become)** budésónuƙot<sup>a</sup> *v*. **hide** buanítésuƙot<sup>a</sup> *v*.; búdès *v*.; búdesuƙot<sup>a</sup> *v*.; ɨɗɛɛs *v*.; ipáŋwéés *v*.; ts'ɛ̀ *n*. **hide away** ɨɗɛ́ɛ́sʉƙɔt<sup>a</sup> *v*.; iwítésuƙot<sup>a</sup> *v*.
mouth
**hide oneself** budés *v*. **hide poorly** itárákáɲés *v*. **hide repeatedly** ɨɗaiyes *v*. **hideous** itópénòn *v*.; làlòn *v*. **high** zikíbòn *v*. **high (get oneself)** ɨrákɛ́sʉƙɔta así *v*. **high (grow)** zikíbonuƙot<sup>a</sup> *v*. **high (make)** zikíbitésúƙot<sup>a</sup> *v*. **high (of many)** zikíbaakón *v*. **high-pitched** ɓòɓòn *v*. **higher than** ileŋes *v*. **highway** ɲerukuɗe *n*. **hill** kùɓ<sup>a</sup> *n*. **hillside** rutet<sup>a</sup> *n*. **hillside (unseen)** kûb<sup>a</sup> *n*. **hilltop** kuɓaɡwarí *n*. **hilltop (flat)** kùɓààƙw<sup>a</sup> *n*. **hilly** ƙumúƙúmánón *v*. **him** nts<sup>a</sup> *pro*. **himself** nébèd<sup>a</sup> *n*.; ntsínêb<sup>a</sup> *n*. **hind-apron (leather)** dek<sup>a</sup> *n*. **hinder** ɨɓatɛs *v*.; itítírés *v*. **hinder repeatedly** ɨɓatíɓátɛ́s *v*. **hindered** ɨɓatíɓátɔ̀n *v*. **hindleg (right)** ɲálán *n*. **hinge** ɲɛ́pɛtá *n*. **hip** obólén *n*. **hip joint socket** loŋóléhò *n*. **hipbone (lower)** obólénìɔ̀k <sup>a</sup> *n*. **hipbone (upper)** róróìɔ̀k <sup>a</sup> *n*. **hippo** ɲépírìà *n*. **Hippocratea africana** míʒìʒ *n*. **hippopotamus** ɲépírìà *n*. **hire** ɓuƙítésuƙot<sup>a</sup> *v*.; ikásíitetés *v*.; ipáŋƙeés *v*.; teréɡanitetés *v*.
**hire temporarily** ɨlɛʝílɛ́ʝɨtɛtɛ́s *v*. **hired hand** teréɡìàm *n*. **hirer** ɓuƙítésuƙotíám *n*. **his** nts<sup>a</sup> *pro*.; ntsɛ́n *pro*. **his/her cousin** tatat<sup>a</sup> *n*. **hiss** fúútòn *v*. **history** emutíká nùù kɔ̀w<sup>a</sup> *n*. **hit** ɗálútés *v*.; íbaɗɛ́s *v*.; iwés *v*.; iwésúƙot<sup>a</sup> *v*. **hit (exploit)** ɨnɔmɛtɛ́s *v*. **hit from behind** kɔnɨtɛtɛ́s *v*. **hit repeatedly** íbaɗiés *v*.; ɨramírámɛ́s *v*. **hit the target** ɨtsírɔ́n *v*.; tsírɔ́n *v*. **HIV** lóɓúlukúɲ *n*.; sílím *n*. **hive** kànàxà *n*. **hoard** irwanes *v*. **hoarse** ƙoƙórómòn *v*.; rɔ́ƙɔ́rɔƙánón *v*. **hoary** kwɛrɛ́xɔ́n *v*. **hobble** ɨsɛ́pɔ́n *v*.; ɨtɔ́ƙɔ́ɔ̀n *v*.; itsúkúkòn *v*. **hobbling** ŋoɗólómòn *v*. **hobblingly** itsúkúk<sup>u</sup> *ideo*. **hodgepodge** ɲɔ́tsɔ́ɓɨtsɔɓ<sup>a</sup> *n*. **hoe** ɲákakurá *n*.; ɲɛ́mɛlɛkʉ́ *n*. **hoe (push type)** kʉ́bɛ̀l *n*. **hoe handle** ɲɛ́mɛlɛkʉ́dàkw<sup>a</sup> *n*. **hoe up (grass)** íɛ́s *v*.; ireɲes *v*. **hoed up (of grass)** íɔ́s *v*. **hog** ɲéɡuruwé *n*. **hog-tie** ɨtʉsɛtɛ́s *v*. **hogwash** ɨɓááŋàsìtòd<sup>a</sup> *n*. **hold** ɨkamɛs *v*.; tírés *v*.; tírésìàw<sup>a</sup> *n*. **hold back** ɨƙalíƙálɛ́s *v*.; isíƙéés *v*.; itítíkés *v*.; itítíketés *v*.; raʝés *v*.; titikes *v*. **hold by handle** tɔkɔɗɛs *v*. **hold by the mouth** taʝakes *v*.
**hold down** ɗaɗátésuƙot<sup>a</sup> *v*.; itikes *v*. **hold each other** tírínós *v*. **hold fast** itííròòn *v*. **hold hands** tírínósá kwɛ̀tìkà<sup>ɔ</sup> *v*. **hold in** ɨƙalíƙálɛ́s *v*. **hold inside** ipáŋwéés *v*. **hold off** raʝés *v*. **hold onto** ɨkamíkámɛ́s *v*.; ŋɔtsɛ́s *v*. **hold out (hands)** taɓɛɛs *v*. **hold religious service** wáán *v*. **hold up** alólóés *v*.; ɨƙaŋɛs *v*.; itúmésuƙot<sup>a</sup> *v*.; titiretés *v*. **hold with teeth** titikes *v*. **hole** ak<sup>a</sup> *n*.; iɓólóɲés *v*.; itoɓes *v*.; rip<sup>a</sup> *n*. **hole (in river)** ɲétsúur *n*. **hole (in tree)** kotím *n*. **hole repeatedly** itoɓítóɓés *v*. **hole up (hide)** ɨtáƙálɛ́sá así *v*. **holiday** ɲákarám *n*. **holiday (national)** ɲésukukú *n*.; ódowa ná zè *n*. **holiness** daás *n*. **holler** ɨkílɔ́n *v*.; iƙúétòn *v*.; iƙúón *v*.; iƙúónuƙot<sup>a</sup> *v*. **hollow** bótsón *v*.; rip<sup>a</sup> *n*. **hollow (in tree)** kotím *n*. **hollow out** iɓóɓórés *v*.; iróróƙés *v*. **hollowed out** iɓóɓórós *v*. **holy** ɓèts'òn *v*.; dòòn *v*. **Holy Communion** Ŋƙáƙá Komúnió<sup>e</sup> *n*. **holy ground** ɲɛkíwɔ́rìt<sup>a</sup> *n*. **Holy Spirit** Suɡura ná Dà *n*. **home** aw<sup>a</sup> *n*.; zɛƙɔ́áw<sup>a</sup> *n*. **home (big)** awa ná zè *n*. **home (clean, orderly)** zɛƙɔ́áwa na maráŋ *n*.
**home life** awááƙw<sup>a</sup> *n*. **homebody** awáám *n*. **homeless person** tsɔnɨtsɔnɔsíám *n*. **homestead** aw<sup>a</sup> *n*. **homestead (abandoned)** ɲóɓóot<sup>a</sup> *n*.; on *n*.; oníáw<sup>a</sup> *n*. **honest person** easíám *n*. **honesty** eas *n*. **honey** ɗàɗ<sup>a</sup> *n*.; ts'ɨƙ<sup>a</sup> *n*. **honey (crystallized)** sʉ̀ƙʉ̀tɛ̀là *n*. **honey (liquid)** lɛ̀ɓ <sup>a</sup> *n*. **honey badger** lɛŋ *n*. **honey bag** ɗàɗèèw<sup>a</sup> *n*.; èw<sup>a</sup> *n*. **honey beer** sɨs *n*.; ts'ɔƙam *n*. **honey hunter** lɛŋɛ́síàm *n*. **honeybee** ts'ɨƙ<sup>a</sup> *n*. **honeycomb** ɗàɗàhò *n*.; ts'ɨƙáhò *n*. **honeycomb (crystallized)** dòkìr *n*. **honeycomb (dry)** ɔfɔrɔƙ<sup>a</sup> *n*. **honeycomb (new)** lòmìl *n*. **honeycomb (old black)** bɔkɔ́k <sup>a</sup> *n*. **honeyguide** tsíts<sup>a</sup> *n*. **honor** tɔ́rɔ́bɛs *v*.; xɛ̀ɓɔ̀n *v*. **hood (of a snake)** ɓòlìɓòl *n*. **hoof** ɡɔ̀rìɡɔ̀r *n*.; sɔk<sup>a</sup> *n*. **hook (metal)** sʉ́ƙʉ́tɛ́sítsɨrím *n*. **hook-shaped** lɔkɔ́ɗɔ́n *v*. **hooked** lɔkɔ́ɗɔ́n *v*.; sokóɗómòn *v*. **hooklike** sokóɗómòn *v*. **hoopoe (green wood-)** ƙáraƙár *n*. **hoot (of owls)** rúɓón *v*. **hop** iɗótón *v*. **hop along** iɗotíɗótòn *v*. **hop on one leg** ɨtsɔ́ɗɔ́kɔ̀n *v*. **hope** ítánòn *v*. **hope for** ítánésuƙot<sup>a</sup> *v*. **hopscotch** cédicedí *n*.
#### horizontal
#### hunch over
**horizontal** kámáránón *v*. **horizontalize** kámáriés *v*.; kámáriésúƙot<sup>a</sup> *v*. **horizontalized** kámáriós *v*. **horn** ɛ̂b <sup>a</sup> *n*. **horn (musical)** ɛ̂b <sup>a</sup> *n*.; ɲɛ́rʉpɛpɛ́*n*. **horn (oryx)** tsarʉ́kʉ́ɛ̀b <sup>a</sup> *n*. **horn (voice-amplifying)** ɲárʉpɛpɛ́*n*. **hornbill (African grey)** tìlòkòts<sup>a</sup> *n*. **hornbill (red-billed)** ƙɔ̀ƙɔ̀t <sup>a</sup> *n*. **hornet** dɛrɛ́ƙ <sup>a</sup> *n*. **hornless** lemúánètòn *v*. **horny (sexually)** kwídikwidós *v*. **horrify** ɨɓálɛ́tɔ̀n *v*.; lilétón *v*.; toɓules *v*. **horrifying** ɨɓálɔ́n *v*.; toɓúlón *v*. **horse** ɲaŋólé *n*. **Hoslundia opposita** ɲɛ́rɨkɨrík<sup>a</sup> *n*. **hospital** ɗakɨtár *n*. **hospital ward** mayaakóniicéhò *n*. **hostel** epúáw<sup>a</sup> *n*.; ɲólóʝ<sup>a</sup> *n*. **hostility** lɔŋɔ́tánànès *n*. **hot** hábòn *v*. **hot (become blazing)** ririanétòn *v*. **hot (become)** hábètòn *v*.; hábonuƙot<sup>a</sup> *v*. **hot (blazing)** ririanón *v*. **hot (make)** hábitésúƙot<sup>a</sup> *v*. **hot (of the ground)** tábʉrʉbʉ́rɔ́n *v*. **hot (piping)** titianón *v*. **hot (very)** pìrrr *ideo*.; titianón *v*. **hot weather** hábona kíʝá<sup>e</sup> *n*. **hot-tempered** ɡúránòn *v*.; ɡúránòn *v*.; ɡúránós *v*.; ɡúránós *v*. **hot-tempered person** ɡúróàm *n*. **hotel** epúáw<sup>a</sup> *n*.; ɲólóʝ<sup>a</sup> *n*. **hothead** ɡúróàm *n*.
**hotheaded** ɡúránòn *v*.; ɡúránós *v*. **hour** ɲásáat<sup>a</sup> *n*. **house** hò *n*. **house help** teréɡiama awá<sup>e</sup> *n*. **house of prayer** wáánàhò *n*. **house of worship** itúrútésiho *n*.; ɲakuʝíhò *n*. **house rat** ɗérá na áwìkà<sup>e</sup> *n*. **housecat** púùs *n*. **housewarming (host a)** fáídomés *v*. **housing** hò *n*. **how (it is)** naítá *subordconn*. **how many?** tanɔ́ɔ́n *v*. **how?** ńtí *adv*. **howitzer** ɲómóta *n*. **howl** iƙúétòn *v*.; iƙúón *v*.; iƙúónuƙot<sup>a</sup> *v*.; irúrúmòòn *v*. **hsss!** rìrrr *ideo*. **hug** ɨrʉmɛs *v*.; tɔtʉnɛs *v*. **hug each other** ɨrʉ́mʉ́nɔ́s *v*.; tɔtʉ́nʉ́mɔ́s *v*. **huge** zòòn *v*. **huge (of many)** zeikaakón *v*. **huh!** yeé' *interj*. **hulky** ɓutúrúmòn *v*. **human being** buɗámónìàm *n*. **humanitarian** lɔʝɔkɔtáw<sup>a</sup> *n*. **humanitarians** roɓa ni ɡúrítínía dayaák<sup>a</sup> *n*. **humanness** ámánànès *n*. **humble** batánón *v*. **humble oneself** kwatsitésúƙota así *v*. **humerus** ɲorótónitíɔ́k <sup>a</sup> *n*. **humiliate** ɓets'itetés *v*. **humiliated** kweelémòn *v*. **humor** céŋ *n*. **hump (animal)** rʉk<sup>a</sup> *n*. **hunch over** dʉ́ɡʉ̀mɛ̀tɔ̀n *v*.
#### hunched
**hunched** mʉƙʉ́rʉ́mɔ̀n *v*. **hunched over** dʉ́ɡʉ̀mɔ̀n *v*.; ɡʉ́ɡʉ̀rɔ̀n *v*.; rʉ́ɡʉ̀ɗʉ̀mɔ̀n *v*. **hundred** ŋamíá *n*. **hunger** ɲɛ̀ƙ <sup>a</sup> *n*. **hunger for meal mush** tɔbɔŋɔ́ɲɛ́ƙ <sup>a</sup> *n*. **hunger for meat** bisák<sup>a</sup> *n*. **hungover** mususánón *v*. **hungriness** ɲɛ̀ƙ <sup>a</sup> *n*. **hungry (be quickly)** hɔɓɔ́mɔ́n *v*. **hunker** rábʉ̀xɔ̀n *v*. **hunt** ƙàƙ<sup>a</sup> *n*.; ƙaƙés *v*.; tɔɓɛlɛs *v*. **hunt (honey)** lɛŋɛ́s *v*. **hunter** ƙàƙààm *n*. **hunters' call for help** waín *n*. **hunting** ƙàƙ<sup>a</sup> *n*. **hunting ground** itsóɗónìàw<sup>a</sup> *n*. **hunting with ropes** ƙaƙa ŋúnítín<sup>o</sup> *n*. **hurdle** itúlúmòòn *v*. **hurl** ɨmasɛs *v*.; ɨrʉtsɛs *v*. **hurl (vomit)** ɦyɛnɛ́tɔ́n *v*.; ɦyɛ̀nɔ̀n *v*.; ɨlɔ́ɓɔ́tɛtɛ́s *v*. **hurl away** ɨmásɛ́sʉƙɔt<sup>a</sup> *v*. **hurl this way** ɨmasɛtɛ́s *v*. **hurry** ɨɓʉ́ŋɔ́n *v*.; ɨɓʉrʉ́ɓʉ́rɔ̀n *v*.; ɨkɨríkírɔ̀n *v*.; ikómóòn *v*. **hurry over there** ikómóonuƙot<sup>a</sup> *v*. **hurry this way** ikóméètòn *v*.; ŋusetésá así *v*. **hurt** áts'ɛ́s *v*.; dódòn *v*.; ɨríɗɔ́n *v*.; tawanes *v*.; tawánítetés *v*. **hurt (begin to)** dódètòn *v*. **hurt (in the chest)** iláfúkòn *v*. **hurt (of an injury)** wɨlímɔ́n *v*. **hurt (of teeth)** isálílòn *v*.; itóŋílés *v*. **hurt intermittently** áts'ietés *v*.
**husband** eakw<sup>a</sup> *n*. **husband (her)** ntsíéákw<sup>a</sup> *n*. **husband (my)** ɲ́cìèàkw<sup>a</sup> *n*. **husband (of my wife's sister)** kʉɓá *n*. **husband (tabooed)** ts'ìnɔ̀àm *n*. **husband (your)** biéákw<sup>a</sup> *n*. **husbands** ɲɔt<sup>a</sup> *n*. **hush** ɦyakwés *v*.; ɨʝɛ́mítɛ́sʉƙɔt<sup>a</sup> *v*. **hush up** ɨʝɛ́mɔ́nʉƙɔt<sup>a</sup> *v*. **hushed** ɨʝɛ́mɔ́n *v*. **hushedly** ʝìr *ideo*. **husk** bɔɗɔ́k <sup>a</sup> *n*.; poɗés *v*.; poɗetés *v*. **husks** nakariɓ<sup>a</sup> *n*. **husky (of voice)** ƙoƙórómòn *v*.; rɔ́ƙɔ́rɔƙánón *v*. **hussy** dekitíníàm *n*. **hut** hò *n*. **hut (grass)** ɲérwám *n*. **hut (Toposa)** Kɔrɔmɔtáhó *n*. **hyena** haú *n*. **hyena (spotted)** atɔŋ *n*.; natɨŋá *n*. **hyena (striped)** ɲetutu *n*. **hyena rider** otsésíama haúùⁱ *n*. **hyena species** oyóŋ *n*. **hygienic** nɛsɛƙánón *v*. **Hymenodictyon floribundum** sésèn *n*. **hyperbolize** ɨmɨɗímíɗɛ́sa mɛná<sup>ɛ</sup> *v*.; tasaɓesa mɛná<sup>ɛ</sup> *v*. **hypocritical** ɨtsárʉ́ánón *v*. **Hypoxis obtusa** ɲɛ́sʉ́tɛ̀ *n*. **hyrax** kwɨnɨƙ<sup>a</sup> *n*. **hyrax (bush)** ɔlír *n*. **hyrax (rock)** barís *n*. **hyrax grass** kwɨnɨƙíkú *n*. **hyrax half** irutumén *n*. **hyrax nickname** lotúɗuzé *n*. **I** ŋ́k <sup>a</sup> *pro*.
**I guess** tsàm *adv*. **I suppose** tsàm *adv*. **I.D.** ɓɛƙɛ́síkabáɗ<sup>a</sup> *n*. **i.e.** tàà *comp*. **ice** tìkɔ̀r *n*. **ichneumon** mútèts<sup>a</sup> *n*. **identification card (Kenyan)** ɲɛkɨpanɗɛ *n*. **identity car** ɓɛƙɛ́síkabáɗ<sup>a</sup> *n*. **idiocy** ɨɓááŋàs *n*. **idiot** bóx *n*.; ɨɓááŋàsìàm *n*. **idiot!** ɗʉ́rʉɗɔ́ɔ̀ *interj*. **idiotic** ɨɓááŋɔ̀n *v*. **idle** ɨlárɔ́n *v*.; ɨlwárɔ́n *v*. **idler** ɲakárámɨt<sup>a</sup> *n*. **idol** kúrúkúr *n*. **if** mísì *subordconn*.; na *subordconn*. **if … had (a while ago)** nànòò *subordconn*. **if … had (earlier)** nanáá *subordconn*. **if … had (yesterday)** nábèè *subordconn*. **if … would** naƙánàk<sup>a</sup> *subordconn*. **if … would have (a while ago)** naƙánòk<sup>o</sup> *subordconn*. **if … would have (earlier)** naƙánàk<sup>a</sup> *subordconn*. **if … would have (yesterday)** naƙásàm *subordconn*. **ignite** dulúmón *v*.; lɛ́ʝɛ́tɔ́n *v*.; tsapés *v*.; tuƙúmétòn *v*. **ignition** aeitetésíàw<sup>a</sup> *n*. **ignore** bálábálatés *v*.; balɛ́s *v*.; balɛtɛ́s *v*. **Ik country** Icékíʝ<sup>a</sup> *n*. **Ik county** Icékíʝ<sup>a</sup> *n*. **Ik dance** Icédìkw<sup>a</sup> *n*. **Ik day** Icéódòw<sup>a</sup> *n*.
**Ik language** Icéɛ́n *n*.; Icénáƙáf *n*.; Icétôd<sup>a</sup> *n*. **Ik people** Ik<sup>a</sup> *n*.; Tɛʉ́sɔ̀ *n*. **Ik person** Icéám *n*. **Ik tobacco** kwílɨlí *n*. **Ik tribe** Icédìyw<sup>a</sup> *n*. **Ikland** Icékíʝ<sup>a</sup> *n*. **Iklikeness** Icénánès *n*. **Ikness** Icénánès *n*. **ill** mòòn *v*. **ill (make)** ɨnʉɛs *v*. **ill (nauseated)** iláƙízòn *v*. **ill (of many)** mayaakón *v*. **ill-fitting** nalóʝón *v*. **illegal** toɗyakos *v*. **illegitimate child** ŋabɔ́bòìm *n*. **illiteracy** ɨɓááŋàs *n*. **illiterate** ɨɓááŋɔ̀n *v*. **illness** màyw<sup>a</sup> *n*.; ɲeɗeke *n*. **illness (mental)** lejénánès *n*. **illness (mild)** ɲeɗekéím *n*. **illusion of movement** lɔ́wírɨwír *n*. **illustrate** ɗoɗésúƙot<sup>a</sup> *v*.; itétémés *v*. **imagine** tamɛtɛ́s *v*. **imbecile** ɨɓááŋàsìàm *n*. **imbibe** béberetés *v*.; wetés *v*. **imbiber** wetésíàm *n*. **imitate** iŋitiés *v*. **immaculate** xɔ́dɔ̀n *v*.; xɔtánón *v*. **immediately** wɛrɛƙɛs *ideo*. **immerse** ilumes *v*.; ilúmésuƙot<sup>a</sup> *v*. **immerse oneself** ilumetésá así *v*. **immigrant** botáám *n*. **immovable** diriɓóón *v*. **immunize** ɨtsɨpítsípɛ́s *v*. **imp** kíʝáìm *n*. **impede** ɨɓatɛs *v*.
**impede repeatedly** ɨɓatíɓátɛ́s *v*. **impel** ɨfalífálɛ́s *v*. **impenetrable** rɔ́mɔ́n *v*. **impervious** itsyátón *v*. **impetuous** ɨɓɛ́lɛ́ánón *v*.; ɨɓɛ́lɔ́ɔ̀n *v*. **impliable** ɡɔkɔ́dɔ̀n *v*. **implore** wáán *v*. **import** zeísêd<sup>a</sup> *n*. **importance** zeís *n*. **important** itíónòn *v*. **impose** tɔnɛɛtɛ́s *v*. **impose oneself** tɔnɛɛtɛ́sá así *v*. **impoverishment** ŋókínànès *n*. **imprecate** ɨlamɛs *v*. **imprecation** ìlàm *n*. **impregnated (recently)** sɨbánón *v*. **imprison** eɡésá hòòk<sup>e</sup> *v*.; eɡésá zíkɛ́sìk<sup>ɛ</sup> *v*.; zíkɛ́s *v*.; zíkɛ́sʉƙɔt<sup>a</sup> *v*. **imprisoned** zíkɔ́s *v*. **improve** doonuƙot<sup>a</sup> *v*.; iŋáléètòn *v*.; maraŋités *v*.; maraŋítésuƙot<sup>a</sup> *v*.; maráŋónuƙot<sup>a</sup> *v*.; xɔ́dɔnʉƙɔt<sup>a</sup> *v*. **improve slightly** ŋwaníŋwánɨtɛ́s *v*. **impudent** ɲɔmɔránón *v*. **impulsive** ɨɓɛ́lɛ́ánón *v*.; ɨɓɛ́lɔ́ɔ̀n *v*. **in addition** ʝìk<sup>ɛ</sup> *adv*. **in charge** wàsɔ̀n *v*.; zeísíàm *n*.; zòònìàm *n*. **in charge of things** wasɔna kúrúɓádù *v*. **in charge of work** zoona teréɡù *v*. **in concert** ikéé kɔ̀n *n*. **in flux** iɲíkón *v*. **in front** ɛ̀kwɔ̀n *v*.; wàxìk<sup>ɛ</sup> *n*.; wàxʉ̀ *n*. **in labor** koríón *v*. **in labor (difficult)** imákóòn *v*.
**in large denominations (of cash)** ɗukúdòn *v*. **in order that** ikóteré *subordconn*.; kánì *subordconn*.; kánɨ náa táa *subordconn*.; kóteré *subordconn*. **in order that … not** kánɨ mookóo *subordconn*. **in pain** toryáɓón *v*. **in shape** ɗoxódòn *v*.; itsyátón *v*. **in shock** ɨʝárɔ́n *v*.; ʝarámétòn *v*. **in shreds** dzɛ́rɛ́dzɛránón *v*. **in that direction** kɔ́ɔ́kɛ *dem*. **in the back** ʝìrʉ̀ *n*. **in the center** sɨsɨkák<sup>ɛ</sup> *n*. **in the future** fàr *adv*. **in the hellǃ** ɲák<sup>a</sup> *adv*. **in the middle** sɨsɨkák<sup>ɛ</sup> *n*. **in the morning** barats<sup>o</sup> *n*. **in the rear** ʝìrʉ̀ *n*. **in the same way** ts'ɛ̀tà kɔ̀nà *n*. **in the worldǃ** ɲák<sup>a</sup> *adv*. **in this direction** nɔ́ɔ́ *dem*.; nɔ́ɔ́na *dem*. **in three years** nakaɨna far *n*. **in two years** nakaɨna tso *n*. **in twos** leɓetsíón *v*. **in unison** ilíróòn *v*. **in vain** ŋálàk<sup>a</sup> *ideo*. **in-law** ceŋetíám *n*. **in-law (my child's spouse's parent)** ɲ́cìɲòt<sup>a</sup> *n*. **in-law (parent)** emetá *n*. **in-law (removed)** ɲòt<sup>a</sup> *n*. **in-laws (being)** ceŋetínánès *n*. **inactive** ɨlárɔ́n *v*.; ɨlwárɔ́n *v*. **inadequate** luƙáámòn *v*. **inadequate (make)** luƙáámitésúƙot<sup>a</sup> *v*.
#### inattentive
influencer
**inattentive** ɓotsódòn *v*. **inaugurate** tsáŋés *v*. **inaugurate (new year)** itówéés *n*. **inbreeder** kɔ́nísìàm *n*. **inbreeding** kɔ́nís *n*. **incarcerate** zíkɛ́s *v*. **incarcerated** zíkɔ́s *v*. **incest** kɔ́nís *n*. **incestuous person** kɔ́nísìàm *n*. **inch** itsóɗón *v*. **inch over here** itsoɗiétòn *v*. **incinerate** wuɗétón *v*. **incinerated** wùɗòn *v*.; xawííts'<sup>ɨ</sup> *ideo*. **incisor** ƙídzɛ̀sìkwàyw<sup>a</sup> *n*. **incite** ɓɛkɛtɛ́s *v*.; ɨsʉ́sʉ́ɛ́s *v*.; itsótsóés *v*. **incite (verbally)** ɨtɔ́ŋɔ́ɛ́s *v*. **incite desire** ɨɓʉ́rɛ́tɔ̀n *v*. **incited** iƙúrúmós *v*. **incitement** ɓɛkam *n*. **incitive** ɓɛkánón *v*. **inclined** sémédedánón *v*. **inclose** íbʉbʉŋɛ́s *v*.; íbʉbʉŋɛ́sʉ́ƙɔt<sup>a</sup> *v*. **include** ɓuƙítésuƙot<sup>a</sup> *v*. **including** nááƙwa *n*. **incorrectly** kèɗè *adv*. **increase** bɨtɛ́tɔ́n *v*.; bɨtɨtam *n*.; bɨtɨtɛtɛ́s *v*.; ɨatɛs *v*.; ɨatímétòn *v*.; komítésuƙot<sup>a</sup> *v*.; tasaɓes *v*. **increase (size)** zeites *v*.; zeitésuƙot<sup>a</sup> *v*. **increased** ɨatɔs *v*. **incubate** tɔʉ́rʉ́mɔ̀n *v*. **indeed** rò *adv*. **indefinitely** pákà *adv*. **indent** luɗés *v*. **indented** luɗúmón *v*.
**indentureship** ŋiléɓúìnànès *n*. **Independence Day** Úrù *n*. **Indian** Ŋímiiɗíàm *n*. **Indian jujube** ɨláŋ *n*. **Indian jujube grove** ɨláŋíɡwì *n*. **indicate** ɗóɗés *v*.; ɗoɗésúƙot<sup>a</sup> *v*. **indigestion** dimésá bubue ŋɔɛsí *n*. **Indigofera arrecta** ɲɛʉrʉlats<sup>a</sup> *n*. **indistinct (visibly)** imítíròn *v*. **indolent** ɓʉɓʉsánón *v*. **induct** tasápánitetés *v*. **inducted** tasapánón *v*. **inducted (get)** tasápétòn *n*. **industrious** ɨɓɛ́rɔ́ánón *v*. **inebriated** ɛsánón *v*. **inebriation** ɛ́s *n*. **inept** betsínón *v*.; hádaadánón *v*.; ɨkáláʝaránón *v*. **infant** ɗiak<sup>a</sup> *n*. **infection (urinary)** lɔríɗ<sup>a</sup> *n*. **inferno** kóméts'àɗ<sup>a</sup> *n*. **infertile** ikólípánón *v*.; osorosánón *v*. **infertile (animal or person)** ɲokólíp<sup>a</sup> *n*. **infertile person** òsòròs *n*. **infiltrate** sáítòn *v*. **infirmary** ɗakɨtár *n*. **infix** otés *v*. **inflammatory** ɓɛkánón *v*. **inflate** ɨsɨkɛs *v*.; xuanón *v*.; xuxuanitetés *v*.; xuxuanón *v*. **inflated** teɓúsúmòn *v*. **inflexible** ɓotsódòn *v*.; kɛtɛ́rɛ́mɔ̀n *v*. **influence** ɲɛsʉp<sup>a</sup> *n*.; sʉ́bɛ̀s *v*.; sʉ́bɛsʉƙɔt<sup>a</sup> *v*. **influence each other** sʉ́bʉnɔ́s *v*. **influencer** sʉ́bɛ̀sìàm *n*.
#### inform
**inform** ɦyeitésúƙot<sup>a</sup> *v*.; ɦyeitetés *v*. **informant** tsíts<sup>a</sup> *n*. **information** emut<sup>a</sup> *n*.; emútík<sup>a</sup> *n*. **infrequent** búúbuanón *v*. **infuse** waatɛ́s *v*. **ingest** béberetés *v*. **ingratiate** rɔɲɛ́sá así *v*. **inhabit** ínés *v*.; ínésuƙot<sup>a</sup> *v*. **inhabitant** zɛƙɔ́ám *n*. **inhalation** sʉ̀p <sup>a</sup> *n*. **inhale** sʉpɛ́tɔ́n *v*. **inhale (food)** ɗáɗítés *v*.; lakatiés *v*.; lukutiés *v*. **inherit** imetsés *v*. **inheritor** ɨrɨtsɛ́síàm *n*. **inhume** búdès *v*.; muɗés *v*.; tʉnʉkɛs *v*. **initiate** tasápánitetés *v*. **initiate (a dance)** iwees *v*. **initiated** tasapánón *v*. **initiated (get)** tasápétòn *n*. **initiation** ɲékipeyés *n*. **inject** ilúkútsés *v*. **injuredly** wìl *ideo*.; wìlìwìl *ideo*. **injury** ɲárʉ́má *n*. **ink** ɲéwiinó *n*. **inkpad** ɲezeí *n*. **inn** epúáw<sup>a</sup> *n*.; ɲólóʝ<sup>a</sup> *n*. **inner ear bone** bòsìɔ̀k <sup>a</sup> *n*. **innocence** ɨɓááŋàs *n*. **innocent** ɨɓááŋɔ̀n *v*. **inquire** esetés *v*.; esetésúƙot<sup>a</sup> *v*.; esetetés *v*. **inquiry** esetés *n*. **insane talk** lejétòd<sup>a</sup> *n*. **insanity** leɡé *n*.; lejé *n*.; lejéèd<sup>a</sup> *n*.; lejénánès *n*.
**insect** ƙʉts'<sup>a</sup> *n*. **insect (cloth-eating)** kurukur *n*. **insect (wood-boring)** lɔ́pírɨpír *n*. **insect species** ts'anán *n*. **insecticide** ƙʉts'ácɛ́mɛ́r *n*. **insecure (of an area)** tsakátsákánón *v*. **insecurity** ɲárém *n*. **insecurity (create)** irémóòn *v*. **inseminate** ɨtɛpɛs *v*. **insensitive** ɨlɛtílɛ́tɔ̀n *v*.; ɨlɛ́tʉ́ránón *v*. **inseparable** upánón *v*. **insert** íbʉbʉŋɛ́s *v*.; íbʉbʉŋɛ́sʉ́ƙɔt<sup>a</sup> *v*.; iwoɗíwóɗés *v*.; otés *v*.; zɔ́bɛ̀s *v*. **inside** aƙw<sup>a</sup> *n*.; áƙwɛ̂d <sup>a</sup> *n*. **inside (a house)** hoaƙw<sup>a</sup> *n*. **insincere** ɨtsárʉ́ánón *v*. **insolent** dirídòn *v*.; ɲɔmɔránón *v*. **insomnia** ɡòk<sup>a</sup> *n*. **insomnia (have)** ɡòkòn *v*. **inspect** ɡonés *v*.; ɨpíʝíkɛ́s *v*.; iséméés *v*.; iséméetés *v*. **inspect (here)** ɡonetés *v*. **inspect (there)** ɡonésúƙot<sup>a</sup> *v*. **inspect entrails** ɦyeitésá arííkà<sup>ɛ</sup> *v*. **install (a beehive)** rɔ́ƙɛ́s *v*. **instantly** wɛrɛƙɛs *ideo*. **instead (of)** àkìlɔ̀ *prep*. **institute (educational)** ɲésukúl *n*. **instruct** ɨtátámɛ́s *v*.; ɨtsɨkɛs *v*.; nɔɔsanitetés *v*. **instructor** ɨtátámɛ́síàm *n*.; ŋímaalímùàm *n*. **instrument (stringed)** aɗúŋkú *n*. **insubordinate** ɨsɛ́ƙɔ́ánón *v*. **insubordination** ɲɛ́sɛ́ƙɔ *n*. **insufficient** ɡàɗɔ̀n *v*.; ɨɗákɔ́n *v*.; luƙáámòn *v*.; taɗatsánón *v*. **insufficient (make)** luƙáámitésúƙot<sup>a</sup> *v*.
#### insult
**insult** iyaŋes *v*.; risés *v*.; tatés *v*. **integrity** eas *n*. **intelligence** akílìk<sup>a</sup> *n*.; nɔɔ́s *n*. **intelligence officer** tsíts<sup>a</sup> *n*. **intelligent** nɔɔsánón *v*. **intelligent person** nɔɔsáàm *n*. **intend (to do)** bɛ́ɗɛ́s *v*.; ɨwɔ́ŋɔ́n *v*. **intend to do** kʉ̀tɔ̀n *v*. **intense** iɗíkón *v*. **intensify** iɗíkétòn *v*.; iɗikitetés *v*. **intentional** iyótsóós *v*. **inter** búdès *v*.; muɗés *v*.; tʉnʉkɛs *v*. **intercede** terés *v*. **intercourse (sexual)** ep<sup>a</sup> *n*. **interesting** ɛ̀fɔ̀n *v*. **interfere** íbʉbʉŋɛ́s *v*.; íbʉbʉŋɛ́sʉ́ƙɔt<sup>a</sup> *v*. **interior** aƙw<sup>a</sup> *n*.; áƙwɛ̂d <sup>a</sup> *n*. **interject** ilúkútsés *v*. **interlace** ɨlɔ́ƙɛ́rɛ́s *v*. **interlock** ɨlɔ́ƙɛ́rɛ́s *v*. **interpose** ilúkútsés *v*. **interpret** ŋʉrɛtɛ́s *v*. **interred** tʉnʉkɔs *v*. **interrogate** esetiés *v*.; ɨnɨnɛ́s *v*. **interrupt** iƙofes *v*.; itoɓes *v*. **interrupt conversation** itoɓítóɓésa tódà<sup>e</sup> *v*. **interruptive** rɛ́bɔ̀n *v*. **interspace** ilores *v*. **intersperse** iɗomes *v*. **interspersed** iɗómíòn *v*. **intertwine** ɨmɔ́ʝírɛ́s *v*. **intervals (do in)** iɗomes *v*. **interweave** ɨlɔ́ƙɛ́rɛ́s *v*. **intestine** nasoroɲ *n*.
**intestine (large)** bɔ̀ *n*.; ɲétenús *n*. **intestine (small)** arí *n*. **intimidate** ɨɔ́ɓɔ́rɛ́s *v*.; kitítésuƙot<sup>a</sup> *v*.; xɛɓɨtɛs *v*.; xɛɓɨtɛ́sʉ́ƙɔt<sup>a</sup> *v*. **intoxicated** ɛsánón *v*. **intravenous drip** ɲɛ́ɗʉríp<sup>a</sup> *n*. **intrepid** itítíŋòn *v*. **introduce** ɗoɗésúƙot<sup>a</sup> *v*. **inunct** kwírɛ́s *v*.; tsáŋés *v*. **invade** sáítòn *v*. **invent** iroketés *v*. **inventor** ɨɗɨmɛ́síàm *n*. **invert** ɨtsʉ́bʉ̀ɗɔ̀n *v*.; tuɗúlútés *v*. **inverted** ɨtsʉ́bʉ̀ɗʉ̀mɔ̀n *v*.; tuɗúlón *v*. **investigate** esetiés *v*.; ɨpíʝíkɛ́s *v*.; tɨrɨfɛs *v*.; tɨrɨfɛtɛ́s *v*.; tirifiés *v*.; tɨrɨfírífɛ́s *v*. **investigator** tɨrɨfɛtɛ́síàm *n*. **investigator (government)** tirifiesíáma ɲápukání *n*. **invisible** kúbòn *v*. **invite** óés *v*. **Ipomoea spathulata** tʉkʉtʉkán *n*. **Ipomoea wightii** kapʉrat<sup>a</sup> *n*. **irascible** ɡúránòn *v*.; ɡúránós *v*. **iron disulfide** ɲésiɓalitútu *n*. **iron sheets** kua ni ɲeryaŋí *n*. **ironstone** ŋaríám *n*. **irresponsible** ɨkáláʝaránón *v*. **irrigate** wetités *v*.; wetitésuƙot<sup>a</sup> *v*. **irritate** ɨtsanɛs *v*. **irritating** fìfòn *v*.; ɨtsánánòn *v*. **Islam** Ŋísɨlám *n*. **isolate** ɨlɔ́ɗíŋɛ́s *v*.; ɨpátsɛ́sʉƙɔt<sup>a</sup> *v*.; tɔlʉ́kɛ́sʉƙɔt<sup>a</sup> *v*. **isolate oneself** ɨpátsɛ́sʉƙɔta así *v*. **issue** pulutetés *v*. **issues** mɛn *n*. **it** nts<sup>a</sup> *pro*.
**it seems** íkwà *adv*.; ókò *adv*. **it's likely** ntsúó ts'ɔɔ *pro*. **itch** sʉ́ƙɔ́n *v*. **item** kɔ́rɔ́ɓâd<sup>a</sup> *n*. **items** kúrúɓâd<sup>a</sup> *n*. **its** nts<sup>a</sup> *pro*.; ntsɛ́n *pro*. **itself** nébèd<sup>a</sup> *n*.; ntsínêb<sup>a</sup> *n*. **ivory** òŋòrìkwàyw<sup>a</sup> *n*. **jab** ɡafarɛs *v*.; ɡɛfɛrɛs *v*.; ɨɓaɲɛs *v*.; iƙumes *v*. **jab repeatedly** ɡafariés *v*. **jabber** ilemílémòn *v*. **jack** ɗiɗecúrúk<sup>a</sup> *n*. **jackal** isér *n*. **jackal (golden)** ɲekiliriŋ *n*. **jacket** ɲɛ́ʝákɛ̀t <sup>a</sup> *n*. **Jacob** Yakóɓò *n*. **jagged** ríbiribánón *v*. **jail** eɡésá hòòk<sup>e</sup> *v*.; eɡésá ɲáʝálaák<sup>e</sup> *v*.; eɡésá zíkɛ́sìk<sup>ɛ</sup> *v*.; lɔʝála *n*.; ɲáʝála *n*.; zíkɛ́s *v*.; zíkɛ́sìàw<sup>a</sup> *n*.; zíkɛ́sʉƙɔt<sup>a</sup> *v*. **jailed** zíkɔ́s *v*. **jam** ɨɗɨlɛs *v*.; ɨɗɔ́tsɔ́n *v*.; rʉtsɛ́s *v*.; rʉtsɛ́sʉ́ƙɔt<sup>a</sup> *v*. **jam into** ipúkútsésuƙot<sup>a</sup> *v*. **James** Yakóɓò *n*. **James (biblical)** Yakóɓò *n*. **jammed together** lolotánón *v*. **January** Kùpòn *n*.; Lomuk<sup>a</sup> *n*. **jar** kurétón *v*. **jaundice** ɲaŋáánètòn *v*. **jaundice (of eyes)** xídɔna ekwitíní *v*. **jaw** ƙálíts'<sup>a</sup> *n*. **jawbone** ƙálíts'ìɔ̀k <sup>a</sup> *n*. **jawbone corner** ƙaƙúŋ *n*. **jealous** ɨrákáánón *v*.
**jealous (make oneself)** ɨrakɛsa así *v*. **jealousy** ɨrákáánás *n*. **jeer** tɔʝɛmɛs *v*. **jelly-like** milílón *v*. **jenny** ɗiɗeŋwa *n*. **jerk** iɓwates *v*.; ɨlɨkílíkɛ́s *v*.; ipoles *v*. **jerk (react)** tokúétòn *v*.; tokúréètòn *v*. **jerk out** ipoletés *v*. **jerk up** ipoletés *v*. **jerky** ŋátɔɔsa *n*. **jerrycan** ɲɛ́ʝɨrɨkán *n*. **jerrycan (1-liter)** túkulét<sup>a</sup> *n*. **jerrycan (half)** ɲéɓukuɓúk<sup>a</sup> *n*. **Jesus** Yésù *n*. **jet (plane)** loɗúwa *n*. **Jew** Ŋíyuɗáíàm *n*. **jewelry (hair)** ɲéméle *n*. **Jie dialect** Fetíícétôd<sup>a</sup> *n*. **Jie person** Fetíám *n*. **jigger** túkútùk<sup>a</sup> *n*. **jiggily** lòk<sup>o</sup> *ideo*. **jiggle** iɓokes *v*.; ɨlɨkílíkɛ́s *v*.; ɨyɔŋíyɔ́ŋɛ́s *v*. **jiggly** lokódòn *v*. **jinx** ipeɗes *v*.; sʉɓɛ́s *v*. **jinxed** ɨɲʉ́ɲʉ́ánón *v*. **jinxer** ìpèɗààm *n*. **jittery** rukurúkón *v*.; tsʉɗʉtsʉɗɔ́s *v*. **job** ɲákási *n*.; ɲetits<sup>a</sup> *n*.; terêɡ<sup>a</sup> *n*. **jobless** ɨlwárɔ́na teréɡù *v*. **jog** isipísípòn *v*.; ɨsɔƙísɔ́ƙɔ̀n *v*.; ɨsʉmʉ́sʉ́mɔ̀n *v*. **John** Yoánà *n*. **John (biblical)** Yoánà *n*. **join** ɗɛsɛ́mɔ́n *v*.; ɗɔtsɛ́s *v*.; ɗɔtsɛ́sʉ́ƙɔt<sup>a</sup> *v*.; ɗɔtsɛtɛ́s *v*.; ɲimanites *v*.; ramɛtɛ́s *v*.; tɔŋɛtɛs *v*.; toropes *v*.; tɔrʉtsɛs *v*.; xɔ́bɛtɛ́sá así *v*.
**join end-to-end** ƙídzɨtɛtɛ́s *v*. **join in** ɓuƙonuƙot<sup>a</sup> *v*. **join together** ɗɔtsánónuƙot<sup>a</sup> *v*. **join up** imánónuƙot<sup>a</sup> *v*.; tɔŋɛ́tɔ́nʉƙɔt<sup>a</sup> *v*.; zɔ́tɛ́tɔ̀n *v*. **joined** ɗɔtsɔ́s *v*. **joined together** ɗɔtsánón *v*.; kumutsánón *v*. **joint** ɲékel *n*.; ɲimanitésíàw<sup>a</sup> *n*. **joint (costovertebral)** ŋábèrìkàdɛ̀ *n*. **joke** céŋ *n*. **joke around** ceŋánón *v*. **joker** céŋáàm *n*. **jokester** céŋáàm *n*. **joking** céŋ *n*. **jolt** kurétón *v*. **jounce** íbotitésúƙot<sup>a</sup> *v*.; íbotitetés *v*. **journey** ɲásápari *n*. **jowl** bòkìbòk<sup>a</sup> *n*. **joyful** eaŋanes *v*. **Jude (biblical)** Yúdà *n*. **judge** ŋʉrɛ́s *v*.; ŋʉrɛtɛ́s *v*.; ŋʉrɛtɛsíàm *n*.; ŋurutiés *v*.; ŋurutiesíàm *n*.; ŋurutiesúƙot<sup>a</sup> *v*. **judge (pardoning)** óɡoesíám *n*. **judged** ŋurutiós *v*. **judging elder** bɨlamʉ́síàm *n*. **jug** ɲéʝá *n*. **jug (wooden)** ɲɛkʉlʉmɛ *n*. **juice** ɲéʝúùs *n*. **juicily** kwìts'ⁱ *ideo*. **juicy** kwits'ídòn *v*. **juicy (of meat)** duláts'ámòn *v*. **July** Ilíŋéèts'<sup>a</sup> *v*.; Lomoɗokoɡéc *n*. **jumble up** ɨmɔrímɔ́rɛ́s *v*. **jumbled** lɔŋɔanón *v*.
**jumbled up** ɨmɔrímɔ́rɔ́s *v*. **jumbled up (become)** lɔŋɔanónuƙot<sup>a</sup> *v*. **jump** íbòtòn *v*.; iɗótón *v*. **jump (attack)** tonyámónuƙot<sup>a</sup> *v*. **jump (make)** íbotitésúƙot<sup>a</sup> *v*.; íbotitetés *v*. **jump (startle)** tsídzitetés *v*. **jump down** apííròn *v*.; ipííròn *v*. **jump excitedly** ɗamɨɗámɔ́n *v*. **jump off** apííròn *v*.; ipííròn *v*. **jump rope** íɡoriesá simá<sup>e</sup> *v*. **jump to it** apííròn *v*.; ipííròn *v*. **jump up** ɨɗɛ́ɛ́tɔ̀n *v*.; ɨɗɛ́ɛ́tɔ̀n *v*. **jumping out** ɨɗɛ́ɔ́n *v*. **jumping up** ɨɗɛ́ɔ́n *v*. **jumpy** rukurúkón *v*.; tsʉɗʉtsʉɗɔ́s *v*. **junction** bézèkètìkìn *n*. **June** Nàmàƙàr *n*.; Yɛlíyɛ́l *n*. **jungle** lolíts<sup>a</sup> *n*. **jury member** kárátsìkààm *n*. **just** ʝá *adv*.; ɲák<sup>a</sup> *adv*.; tsàm *adv*. **just there** néda ne *dem*.; néíta ne *dem*. **Justicia species** ʉrʉ́sáy<sup>a</sup> *n*. **jut** rúɡèts<sup>a</sup> *n*. **Kaabong Road** Kaaɓɔ́ŋɨmucé *n*. **Kaabong town** Kaaɓɔ́ŋ *n*. **Kaakuma town** Káákuma *n*. **kabob** rɔam *n*. **kaboom!** pààɗòk<sup>o</sup> *ideo*. **kalah (game)** ɲékilelés *n*. **Kamion** Kámíón *n*. **Kamion Road** Kámíónomucé *n*. **Kathile town** Kàsìlè *n*. **keeled** toŋórómòn *v*.; toróŋómòn *v*. **keen (of eyesight)** tsɔ́tsɔ́n *v*. **keep** ɡirés *v*.; ɨrɨtsɛ́s *v*. **keep an eye on** ɨfátɛ́sʉƙɔt<sup>a</sup> *v*.
**keep aside** óɡoɗés *v*.; óɡoɗésúƙot<sup>a</sup> *v*.; oƙésúƙot<sup>a</sup> *v*. **keep away** tatsáɗésuƙot<sup>a</sup> *v*. **keep back** itúmésuƙot<sup>a</sup> *v*.; tatsáɗésuƙot<sup>a</sup> *v*. **keep calling** óísiés *v*. **keep company** itúmétòn *v*. **keep going** pórón *v*. **keep on** itemes *v*.; itemetés *v*.; taɗáŋón *v*. **keep one's mouth shut** iɓótóŋésá aká<sup>e</sup> *v*. **keep the beat** ɨrɛɛsa dikwá<sup>e</sup> *v*. **keep the law** ɡirésá ìtsìkà<sup>ɛ</sup> *v*. **keep thoughts secret** ɡirésá mɛná<sup>ɛ</sup> *v*. **keep to oneself** ɡirésá así *v*. **keep waiting** ɨlarɨtɛtɛ́s *v*.; ɨlwarɨtɛtɛ́s *v*. **keeper** ɨrɨtsɛ́síàm *n*. **keloid** ɲépóros *n*. **Kenya** Kéɲà *n*. **kernel** eɗed<sup>a</sup> *n*.; **kernel** ekwed<sup>a</sup> *n*. **kernels (hollowed)** dololoƙ<sup>a</sup> *n*.; dololots'<sup>a</sup> *n*. **kerosene** ceím *n*. **kerplunk!** rìtⁱ *ideo*. **key** ɲófuŋƙúwo *n*. **khat** ɲémurúŋ́ɡù *n*. **khat-chewer** ɲémurúŋ́ɡùàm *n*. **kick** tɛtsɛ́s *v*. **kick hides** tɛtsɛ́sá jéjèìkà<sup>e</sup> *v*. **kick off (a dance)** iwees *v*. **kick off (new year)** itówéés *n*. **kick repeatedly** tetsítétsiés *v*. **kid (child)** im *n*. **kid (goat)** riéím *n*. **kidnap** ɨɛpɛtɛ́s *v*. **kidney** ɲaŋalúr *n*.
**kids** wik<sup>a</sup> *n*. **Kigelia africana** sosóbòs *n*. **kill (many)** sáɓés *v*. **kill (singly)** cɛɛ́s *v*.; cɛɛ́sʉ́ƙɔt<sup>a</sup> *v*. **kill each other** sáɓúmós *v*. **kill for** ipéyéés *v*.; topues *v*. **kill for (make)** ipéyéitésúƙot<sup>a</sup> *v*. **kill more than one (make)** sáɓítetés *v*. **kill off** ɨɗɛ́ɛ́sʉƙɔt<sup>a</sup> *v*. **kill serially** ɨkɛɲíkɛ́ɲɛ́s *v*. **kill time** ipásóòn *v*. **killed (pay to have)** ɨlʉŋʉ́lʉ́ŋɛ́s *v*. **killer** sèààm *n*. **killer (of many)** sáɓésìàm *n*. **killer (serial)** ɨkɛɲíkɛ́ɲɛ́síàm *n*. **killer (singly)** cɛɛsíám *n*. **killing one-by-one** kɛ̀ɲ *ideo*. **kiln** ɲéripipí *n*. **kin** ɦyeímós *v*.; ɦyeínós *v*. **kind** batánón *v*.; bònìt<sup>a</sup> *n*.; maráŋón *v*.; ɲákaɓɨlá *n*. **kind person** maráŋónìàm *n*. **kindle** ɡamés *v*. **kindle a fire** ɡamésá ts'aɗí *v*. **kindling** ɡamam *n*.; lúulú *n*. **kindling (small)** ɗɛ̀rɛ̀ts<sup>a</sup> *n*. **king** ɲéríósit<sup>a</sup> *n*. **kingfisher** kílootór *n*. **kinship by birth** ɦyeínósá na ƙɔ́ɓà<sup>ɛ</sup> *v*. **kinship by blood** ɦyeínósá na séà<sup>e</sup> *v*. **kiosk** ɗʉkán *n*.; ɲáɗʉkán *n*. **kiss** ts'ʉnɛ́s *v*. **kiss each other** ts'ʉ́nʉ́nɔ́s *v*. **kitchen** itiŋésíàw<sup>a</sup> *n*.; ɲéʝiɡón *n*. **kitchen (camp)** ŋƙáƙáhò *n*. **kite (black)** ʝolíl *n*.
**Kleinia species** kìmɔ̀ɗɔ̀rɔ̀ts<sup>a</sup> *n*.; loɓóŋiɓóŋ *n*. **kleptomania** kɔrɔ́kíkànànès *n*. **kleptomaniac** kɔrɔ́kíkààm *n*. **klipspringer** tʉ́s *n*. **klipspringer rat** tʉ́síɗèr *n*. **klop klop** kùr *ideo*. **knap** pɛsɛlɛs *v*.; wɛts'ɛ́s *v*.; wɛts'ɛtɛ́s *v*. **knap repeatedly** ɨwɛts'íwɛ́ts'ɛ́s *v*. **knead** dʉbɛ́s *v*.; ɨɗaŋíɗáŋɛ́s *v*. **knee** kútúŋ *n*. **knee (posterior)** ƙór *n*. **kneecap** bùrùkùts<sup>a</sup> *n*. **kneel** kutúŋétòn *v*. **kneeling** kutúŋón *v*. **knife** ɗàw<sup>a</sup> *n*.; rutésúƙot<sup>a</sup> *v*. **knife (wrist)** ɲáɓaarát<sup>a</sup> *n*. **knitting needle** ɲɛ́sɨlɨɓá *n*. **knobby** lɛrɛ́kɛ́mɔ̀n *v*. **knock** íbaɗɛ́s *v*.; ɨɲatɛs *v*.; tanaŋes *v*. **knock around** tanaŋínáŋesuƙot<sup>a</sup> *v*. **knock back** ɨɓatɛs *v*. **knock back repeatedly** ɨɓatíɓátɛ́s *v*. **knock down** íbatɛ́s *v*.; íbatɛtɛ́s *v*.; ɨpʉtɛs *v*. **knock down repeatedly** íbatiés *v*. **knock off** ɨpʉtɛs *v*. **knock on** ɨɗɔŋíɗɔ́ŋɛ́s *v*.; ikoŋíkóŋés *v*. **knock over** íbatɛ́s *v*.; íbatɛtɛ́s *v*. **knock over repeatedly** íbatiés *v*. **knock repeatedly** íbaɗiés *v*.; ɨɲatiés *v*. **knot** simááƙát<sup>a</sup> *n*. **knot (in wood)** ɡɔ̀ɓ <sup>a</sup> *n*.; ɔ́ʝ <sup>a</sup> *n*. **know** enés *v*.; ɦyeés *v*.; íyeés *v*. **know (come to)** ɦyeésúƙot<sup>a</sup> *v*.
**knowledge** nɔɔ́s *n*. **knowledgeable** nɔɔsánón *v*. **known** ɦyoós *v*. **kob** máyá *n*. **kob grass** máyákù *n*. **kraal** ɓór *n*. **kudu (female greater)** kòtòb<sup>a</sup> *n*. **kudu (greater)** akín *n*. **kudu (lesser)** àbɛ̀t <sup>a</sup> *n*. **kudu (male greater)** anás *n*. **Kuliak people** Ŋkúlák<sup>a</sup> *n*. **label** ɨmátsárɛ́s *v*. **labia (vulval)** kwàìn *n*. **labor (be in)** koríón *v*. **labor (difficult)** ƙwaata ná ɡààn *n*. **laboriously** ziál *ideo*. **labret** ɡwalát<sup>a</sup> *n*. **labret combo** ɲétépes *n*. **Labwor** Teɓur *n*. **Labworian** Ŋítéɓuríám *n*. **lack** bɨrɔ́ɔ́nìmɛ̀n *n*.; ɨɗakɛ́s *v*. **lacking** bɨrɔ́ɔ́n *v*.; ɡàɗɔ̀n *v*.; ɨɗákɔ́n *v*.; luƙáámòn *v*.; taɗatsánón *v*. **lacking (make)** luƙáámitésúƙot<sup>a</sup> *v*. **lackluster** ʝɔ̀lɔ̀n *v*. **ladder** ɲáláɗa *n*. **ladle** ƙɔ́r *n*. **ladle (broken)** tebeleƙes *n*. **lag** isíɗéètòn *v*.; isíɗóòn *v*. **Lagenaria species** lomuƙe *n*.; lɔ́pʉ́l *n*. **Lagenaria sphaerica** óbiʝoets'<sup>a</sup> *n*. **laid flat** fatsámánòn *v*. **laid to rest** tʉnʉkɔs *v*. **lair** ak<sup>a</sup> *n*. **lake** ɲánam *n*.
#### lamb
**lamb** ɗóɗòìm *n*. **lame** ŋwàxɔ̀n *v*. **lame person** ŋwàxɔ̀nìàm *n*. **lameness** ŋwaxás *n*. **lament** itseniés *v*.; ƙɔ̀ɗɔ̀n *v*. **lamp (oil)** ɲɔ́tɔɗɔpá *n*. **land** kíʝ<sup>a</sup> *n*.; toɗóón *v*.; zɛƙwɛ́tɔ́n *v*. **land deed** ʝʉmʉ́kábaɗ<sup>a</sup> *n*. **land title** ʝʉmʉ́kábaɗ<sup>a</sup> *n*. **land transformation** beniitesa kíʝá<sup>e</sup> *n*. **landowner** áméda kíʝá<sup>e</sup> *n*. **landslide** bɔ̀rɔ̀ts<sup>a</sup> *n*.; dìdìàk<sup>a</sup> *n*. **language** naƙaf *n*.; tôd<sup>a</sup> *n*. **language (European)** Ɓets'oniicétôd<sup>a</sup> *n*.; Ŋímusukúìtòd<sup>a</sup> *n*. **language (foreign)** ɦyɔɛn *n*.; ɦyɔ̀tòd<sup>a</sup> *n*. **lanky** sawátsámòn *v*.; sʉrʉsʉ́rɔ́n *v*. **Lannea schimperi** ekoɗit<sup>a</sup> *n*.; meleke *n*. **Lantana trifolia** tíkòŋ *n*. **lantern** ɲátaayá *n*. **lap up** ɨʝaƙíʝáƙɛ́s *v*. **lappet** ɲɛ́tɛ́lɨtɛl *n*. **larceny** dzú *n*. **large** zòòn *v*. **large (of many)** zeikaakón *v*. **large numbers (in)** ɓʉlɛs *ideo*. **large tin can** ɲákaɓurúr *n*. **largeness** zeís *n*. **larva (bee)** sîd<sup>a</sup> *n*. **larva (tiger beetle)** sikusába *n*. **larynx** ɡɔ̀k <sup>a</sup> *n*. **lash** iɓúŋéés *v*.; ɨɗɨtsɛs *v*. **last** topíánètòn *v*. **last (be the very)** ɨts'íɗɛ́ɛ̀tɔ̀n *v*. **last (be the)** mɨtɔna ɗíɛ́ʝìrì *v*.
**last (one)** ɗa ʝírì *pro*. **last (unit of time)** bàts<sup>e</sup> *adv*. **last born (being the)** eŋúnúnànès *n*. **last person** ʝìrìàm *n*. **last to show up** irúpéètòn *v*.; irúpóòn *v*. **last year** kaɨnɔ sɨn *n*.; nótso kaɨnɔ sɨn *n*.; sakɛɨn *n*. **lastborn** eŋún *n*.; ʝìrìàm *n*. **late (dead)** tás *n*. **later** ʝìrʉ̀ *n*. **latest person** ʝìrìàm *n*. **latrine** ets'íhò *n*.; ɲótsorón *n*. **laud** tamɛɛs *v*. **laugh** fèkòn *v*. **laugh (make)** fekitetés *v*. **laugh a lot** fekifekos *v*. **laugh uproariously** ɗɛtɛ́ɗɛ́tánón *v*. **launch** ɨɗɛ́ɛ́tɔ̀n *v*.; itówéés *n*. **law** ɨtsɨkɛs *n*. **law of God** ɨtsɨkɛsa Ɲákuʝí *v*. **lawbreaker** tɛ́ŋɛ́rìàm *n*. **lawbreaking** tɛ́ŋɛ́r *n*. **lawful order** ŋíkísila *n*.; ɲɛkɨsɨl *n*. **laws (local)** ɨtsɨkɛsíícíká kíʝá<sup>e</sup> *v*. **laws of the land** ɨtsɨkɛsíícíká kíʝá<sup>e</sup> *v*. **lay (eggs)** ɨɓɛ́ɓɛ́ɛsʉƙɔt<sup>a</sup> *v*. **lay down** epítésuƙot<sup>a</sup> *v*. **lay flat** epitésúƙòtà ɗèɲ *v*. **lay it out (issues)** tɔlɔɛsa mɛná<sup>ɛ</sup> *v*. **lay loosely** laʝetés *v*. **lay out** ɨpɛ́pɛ́tɛ́s *v*.; tɔlɔɛs *v*. **lay over** ɨlɔ́ɡɔtsɛ́s *v*. **lay prostrate** bukites *v*. **lay to rest** búdès *v*.; muɗés *v*.; tʉnʉkɛs *v*. **lay waste to** ɨsílíánɨtɛtɛ́s *v*.
**laying of eggs** ƙwaata ɓíɓà<sup>e</sup> *n*. **lazy** ɨtátsámánón *v*.; karámóòn *v*.; wéésánón *v*. **lazy eye (have a)** pɨlírímɔ̀n *v*. **LCI** ámázeáma na kɔ́nɔ̀nì *n*.; arasí *n*. **lead** torikes *v*. **lead astray** hakítésuƙot<sup>a</sup> *v*.; itwáŋítésúƙot<sup>a</sup> *v*. **lead away** toríkésuƙot<sup>a</sup> *v*. **lead group prayers** taƙates *v*. **lead off** toríkésuƙot<sup>a</sup> *v*. **lead slowly** ɨɛmɛs *v*. **lead this way** toriketés *v*. **leader** tòrìkààm *n*.; torikesíám *n*. **leaders** roɓazeik<sup>a</sup> *n*. **leadership (of many)** roɓazeikánánès *n*. **leaf** kak<sup>a</sup> *n*. **leaflike** katálámòn *v*. **leak** ɨpɨnípínɔ̀n *v*.; tɔ̀wɔ̀n *v*. **lean** liƙés *v*. **lean against** ɨƙɔ́ŋítɛ́s *v*.; ɨƙɔ́ŋítɛ́sʉƙɔt<sup>a</sup> *v*.; tonokes *v*. **lean on** ɨƙɔŋɛs *v*.; tébinés *v*. **lean over** íboɗolés *v*. **leaned against** ɨƙɔ́ŋítɔ́s *v*. **leaned on** ɨƙɔ́ŋítɔ́s *v*. **leap** íbòtòn *v*.; iɗótón *v*.; íɡɔ̀rɔ̀bɔ̀n *v*. **leap into action** ipííròn *v*. **leapfrog** íɡoriés *v*. **learn** ɦyeésúƙot<sup>a</sup> *v*.; nɔɔsánétòn *v*. **lease** ipáŋƙeés *v*. **leash** rɔɓ<sup>a</sup> *n*. **leather** jèjè *n*. **leather (cow)** ɦyɔjejé *n*. **leather brassiere** ɲákɨláƙ<sup>a</sup> *n*. **leather cloak** xɔŋɔŋ *n*.
**leather clothing** ínóƙwàz *n*. **leather leggings** ŋapokóy<sup>a</sup> *n*. **leather mat** jèjè *n*. **leather shawl** xɔŋɔŋ *n*. **leather shoe** ŋáɓʉʉrá *n*. **leather strap** ƙíw<sup>a</sup> *n*. **leather strips** ŋamɔ́lɔ́l *n*. **leather tassel** ƙíítínísɔ̀k <sup>a</sup> *n*. **leatherily** tùɗ<sup>a</sup> *ideo*. **leathery** tuɗádòn *v*. **leave** óɡoés *v*. **leave (go)** ƙòòn *v*. **leave a mess** nts'áƙóna sèrèìk<sup>e</sup> *v*. **leave aside** ɨnápɛ́sʉƙɔt<sup>a</sup> *v*. **leave behind** ilaŋés *v*. **leave early** ɡáƙón *v*. **leave gaps in** ɨkálámɛ́s *v*. **leave home** xàtsɔ̀n *v*. **leave in a huff** ɡwaítón *v*.; íɡwìʝìrɔ̀n *v*. **leave in the open** ilététaitetés *v*. **leave one's memory** hakonuƙot<sup>a</sup> *v*. **leave open** iŋáɓólés *v*. **leave to** iʝokes *v*.; iʝókésuƙot<sup>a</sup> *v*. **leaven** sîb<sup>a</sup> *n*. **leaves (dry)** ódzàkàk<sup>a</sup> *n*. **leaves (Hibiscus cannabinus)** ɔ́bɛ̀ràkàk<sup>a</sup> *n*. **leech** kɔ́ɛ́s *v*. **left** óɡoós *v*. **left alone** ɗòwòn *v*. **left hand** betsínákwɛ̀t <sup>a</sup> *n*. **left in the open** ilététòòn *v*. **left rib** betsínáŋabér *n*. **left-handed** betsínón *n*. **leftover** ʝɛʝɛ́tɔ́n *v*.; ʝírɛ̂d <sup>a</sup> *n*.; óɡoɗesam *n*. **leftovers** ʝírín *n*.; ɲomokoʝo *n*. **leg** dɛ *n*.; dɛɛd<sup>a</sup> *n*.
**leg (of furniture)** sɔk<sup>a</sup> *n*. **leg-meat taboo** dɛ *n*. **leggings (leather)** ŋapokóy<sup>a</sup> *n*. **leggy** tsáƙólómòn *v*.; tsɔ́ɡɔ̀rɔ̀mòn *v*. **legible** isómáìmètòn *v*. **legion** miƙídòn *v*. **leisure** ɲɛ́tɛmá *n*. **leisure (be at)** ɨtɛ́mɔ́ɔ̀n *v*. **leisurely** wɛ̀wɛ̀ɛ̀s *ideo*. **lemon** ɲámucúŋ́ƙà *n*. **length** zikíbàs *n*. **lengthen** zikíbètòn *v*.; zikíbitésúƙot<sup>a</sup> *v*.; zikíbonuƙot<sup>a</sup> *v*. **lengthy** zikíbòn *v*. **lengthy (of many)** zikíbaakón *v*. **lentils** ɲɔ́ɓɔɔ́*n*. **Leonotis species** ɲétúlerú *n*. **leopard** ɡʉɓɛ́r *n*. **leopard (male)** nʉs *n*. **less** bɨráʉ́tɔ̀n *v*. **less than** ɨlɔɛs *v*. **less-than-full** kíón *v*. **lest** náa táà *subordconn*. **let** óɡoés *v*.; óɡoós *v*. **let down** iyolíyólés *v*. **let down (milk)** cɛ̀rɔ̀n *v*. **let go** talakes *v*. **let go of** taʝales *v*.; taʝálésuƙot<sup>a</sup> *v*.; taʝaletés *v*. **let know** ɦyeitésúƙot<sup>a</sup> *v*.; ɦyeitetés *v*. **let loose** itsues *v*.; itsuetés *v*. **let not** eʝá *adv*. **let oneself be known** ɗóɗítetésá así *v*. **let out** itsues *v*.; itsuetés *v*. **let out (secret)** kwɛts'ɛ́mɔ́n *v*.; kwɛts'ɛ́s *v*.
**let slide** ɨsɔɛs *v*.; ɨsɔɛtɛ́s *v*. **let slip** ɨsɔɛs *v*.; ɨsɔɛtɛ́s *v*. **letter** béɗíbeɗú *n*.; bóɗíboɗú *n*.; ɲáɓáruwa *n*. **letter (alphabetical)** ɲéɡutá *n*.; ɲéɲuɡutá *n*. **level** ɗàsòn *v*.; ɨpáɗáɲɔ̀n *v*.; lopem *n*.; ɲaɗés *v*.; ɲolókér *n*.; towutses *v*. **level (of an area)** kalápátánón *v*. **level (roads)** séɓés *v*. **level (school)** hò *n*. **level out** kalápátánitetés *v*.; ɲaɗésúƙot<sup>a</sup> *v*. **level out (an area)** kalápátánónuƙot<sup>a</sup> *v*. **level repeatedly** ɲaɗiés *v*. **levy** ɲéutsúr *n*. **lexicon** ɲéɗíkìxònàrì *n*. **liaise** terés *v*. **liar** yʉɛ́ám *n*. **liar (be a)** yʉanón *v*. **liberal** waŋádòn *v*. **liberate** hoɗés *v*.; hoɗésúƙot<sup>a</sup> *v*.; hoɗetés *v*. **lice** ts'an *n*. **lice egg(s)** ɨnak<sup>a</sup> *n*. **lick** kánɛ́s *v*. **lick (of flames)** ɨnɛpínɛ́pɔ̀n *v*. **lick clean** tɨmɨɗɛs *v*. **lick up** tɨmɨɗɛs *v*. **lid (flat)** ɲápár *n*. **lie** isuɗam *n*.; isuɗesa mɛná<sup>ɛ</sup> *v*.; itoŋetésá tódà<sup>e</sup> *v*.; yʉanitetés *v*.; yʉanón *v*.; yʉɛ *n*. **lie (legs straight)** ɨtɛ́ɛ́lɔ̀n *v*. **lie around flat** fatsifatsos *v*. **lie down** eponuƙot<sup>a</sup> *v*.; itsólóŋòn *v*. **lie down prostrate** bukonuƙot<sup>a</sup> *v*. **lie face-up** fátsón *v*.
listener
**lie in wait for** ɨɗaarɛ́s *v*. **lie on the back** fátsón *v*.; ɨɗɛ́ɗɔ́ɔ̀n *v*. **lie on the side** epona ŋabér<sup>o</sup> *v*. **lie prostrate** bùkòn *v*.; bukukánón *v*. **life** ɦyekes *n*. **life (daily)** zɛƙw<sup>a</sup> *n*. **lift** ɓuƙés *v*. **lift (make)** ɨkɛ́ítɛtɛ́s *v*. **lift (steal)** ɨtíɗíɗɛ́s *v*. **lift carefully** takiés *v*. **lift off** ɓuƙetés *v*. **lift together** ilélébés *v*. **lift up** ɨkɛ́ɛ́sʉƙɔt<sup>a</sup> *v*.; ɨkɛɛtɛ́s *v*. **lifted** ikeimétòn *v*. **lifted up** ɨkɔɔtɔ́s *v*. **ligament** kon *n*. **ligament (iliolumbar)** ɛkɛw<sup>a</sup> *n*. **light** aeitetés *v*. **light (fire)** ɡamés *v*.; tɔƙɛrɛtɛ́s *v*. **light in color** bɛɗɛ́dɔ̀n *v*.; ɓèts'òn *v*. **light (in color, of many)** ɓets'aakón *v*. **light (slightly)** ɓèts'ìɓèts'òn *v*. **light a fire** ɡamésá ts'aɗí *v*. **light up** aeétón *v*.; iléúrés *v*.; inwakes *v*. **lightbeam** bás *n*.; sʉ́w<sup>a</sup> *n*. **lightheaded** imáúròn *v*. **lightheadedness** taítayó *n*. **lightly (in color)** bɛ̀ɗ ɛ *ideo*. **lightning** ɨmɛ́ɗɔ́nà dìdìì *n*. **lightweight** fɔkɔ́dɔ̀n *v*.; ɔfɔ́dɔ̀n *v*.; olódòn *v*. **lightweightly** fɔ̀k ɔ *ideo*. **like** ɗítá *prep*.; ƙámón *v*.; tsamɛ́s *v*. **like (a lot)** ɡaanón *v*. **like (it is)** naítá *subordconn*.
**like (to do)** ɨtáŋátɔ̀n *v*. **like each other** tsámʉ́nɔ́s *v*. **like thatǃ** ńtí *interj*.; ńtíà ʝà *adv*.; ńtía ʝɨkî *adv*. **like this** ƙámónà ts'ɛ̀ɛ̀n *v*. **like this!** ńtí *adv*. **liken** ƙámítetés *v*. **likewise** ts'ɛ̀tà kɔ̀nà *n*. **lily (fire)** bʉlʉbʉlát<sup>a</sup> *n*. **limb (of a tree)** dakúkwɛ́t <sup>a</sup> *n*. **limber** ɡwidídòn *v*.; tsutsukes *v*. **limberly** ɡwìdⁱ *ideo*. **limbless** ŋʉɗʉ́sʉ́mɔ̀n *v*. **limit** cíkóroy<sup>a</sup> *n*.; ɨrɨɗɛs *v*.; ɨrɨɗɛtɛ́s *v*. **limited** ɨrɨɗɔs *v*. **limp** itsóɗón *v*. **line** ɲálaín *n*.; ɲériŋƙís *n*. **line (furrow)** ɲɛ́pɛ́lʉ *n*. **line (level)** ɲolókér *n*. **line (raised)** zeket<sup>a</sup> *n*. **linger** ɓaɓaránón *v*.; tɔʉ́rʉ́mɔ̀n *v*. **link** tɔŋɛtɛs *v*.; toropes *v*.; tɔrʉtsɛs *v*. **link up** tɔŋɛ́tɔ́nʉƙɔt<sup>a</sup> *v*.; zɔ́tɛ́tɔ̀n *v*. **linked** zɔ́tɔ́n *v*. **lion** máw<sup>a</sup> *n*. **lip** akákwáyw<sup>a</sup> *n*. **lip herpes** ɲótóts<sup>a</sup> *n*. **lip plug** ɡwalát<sup>a</sup> *n*. **lip plug (combo)** ɲétépes *n*. **liquid** cuanón *v*.; cue *n*.; tsɔlílɔ́n *v*. **liquid (become)** cuanónuƙot<sup>a</sup> *v*. **liquify** cuanónuƙot<sup>a</sup> *v*. **liquor** kombót<sup>a</sup> *n*.; tule *n*. **list** ɗɔtsɛtɛ́s *v*. **list of names** éditíníkabáɗ<sup>a</sup> *n*. **listen** nesíbes *v*. **listener** nesíbesíám *n*.
**lit** àèòn *v*. **literacy** nɔɔ́s *n*. **literate** nɔɔsánón *v*. **lithe** ɡwidídòn *v*. **lithely** ɡwìdⁱ *ideo*. **litter** ƙʉ́ƙín *n*. **little (become)** kwátsónuƙot<sup>a</sup> *v*. **little (of many)** kwátsíkaakón *v*. **little in amount** ƙwàɗòn *v*. **little in size** kwátsón *v*. **little in volume** tilóts'ómòn *v*. **live** ɦyekes *v*.; zɛ̀ƙwɔ̀n *v*. **live in** ínés *v*. **live long** ikoŋíkóŋòn *v*. **live on** ʝɛ̀ʝɔ̀n *v*. **live solitarily** iɓóótánón *v*. **live through** pʉ̀rɔ̀n *v*. **live together** ínínós *v*. **livelihood** ɦyekes *n*. **lively** tsuwoós *v*. **liver** sakám *n*. **liver disease** ɲeɗekea sakámá<sup>e</sup> *n*. **livestock** ŋíɓarɛn *n*. **livestock pen** ɓór *n*. **living** ɦyekes *n*. **lizard** pɔ̀pɔ̀s *n*. **lizard (Nile monitor)** náɡanâɡ<sup>a</sup> *n*. **lizard sp.** lúulú *n*. **load** bot<sup>a</sup> *n*. **load a load** ɓuƙésá botá<sup>e</sup> *v*. **loaded down** ɨʉ́ƙɔ́n *v*. **loader** ŋiɓóóìàm *n*. **loafer** ɲakárámɨt<sup>a</sup> *n*. **loan** kál *n*. **loathe** tʉlʉŋɛs *v*.
**local** áméda kíʝá<sup>e</sup> *n*. **Local Councillor I** ámázeáma na kɔ́nɔ̀nì *n*.; arasí *n*. **locality** ɲɛ́tɛɛr *n*. **location** ay<sup>a</sup> *n*.; ɡwarí *n*.; kíʝ<sup>a</sup> *n*.; xán *n*. **lock** iɓótóŋés *v*.; iwetés *v*.; ɲékifúl *n*. **lock up** eɡésá hòòk<sup>e</sup> *v*.; eɡésá zíkɛ́sìk<sup>ɛ</sup> *v*. **lock up (imprison)** zíkɛ́sʉƙɔt<sup>a</sup> *v*. **locked up (imprisoned)** zíkɔ́s *v*. **lockjaw** ɲeɗekea na ɨtɛnítʉ́ƙɔta ámák<sup>a</sup> *n*. **locust** ɡirú *n*. **locust (milkweed)** ɓɔlɔrɔts<sup>a</sup> *n*. **lodge** epúáw<sup>a</sup> *n*.; ɲólóʝ<sup>a</sup> *n*. **Lodwar town** Lóɗwàr *n*. **log** ƙúl *n*. **log (door-barring)** naƙólít<sup>a</sup> *n*. **loincloth** fɔ́ɗ <sup>a</sup> *n*. **loincloth (beaded)** lèt<sup>a</sup> *n*.; tìlàlèt<sup>a</sup> *n*. **loincloth (tâb-shrub)** tábòlèt<sup>a</sup> *n*. **loincloth (women's)** ɲákáɗeŋo *n*. **loiter** ɓaɓaránón *v*. **Lokichokio town** Lokicókio *n*. **lonely** tisílón *v*. **lonesome** tisílón *v*. **long** ítánòn *v*.; zikíbòn *v*. **long (of many)** zikíbaakón *v*. **long ago** kaíníkò nùk<sup>o</sup> *n*.; kɔ̀wɛ̀ nòk<sup>o</sup> *v*.; nòk<sup>o</sup> *adv*.; ódowicíká kì nùù kì *n*.; ódowicíkó nùk<sup>u</sup> *n*. **long for** ítánésuƙot<sup>a</sup> *v*.; wíránés *v*. **long since** kwààkè nòk<sup>o</sup> *n*. **long-legged** tsáƙólómòn *v*.; tsɔ́ɡɔ̀rɔ̀mòn *v*. **long-necked** loƙózòmòn *v*. **longer (become)** zikíbonuƙot<sup>a</sup> *v*. **look around** tɨrɨfírífɛ́s *v*. **look around cautiously** imwáŋón *v*. **look at** ɡonés *v*.
**look at (here)** ɡonetés *v*. **look at (there)** ɡonésúƙot<sup>a</sup> *v*. **look at each other** ɡónímós *v*. **look down** turúnétòn *v*.; turúnón *v*. **look for** bɛ́ɗɛ́s *v*.; bɛɗɛtɛ́s *v*.; ɨkʉʝɛs *v*. **look like** topútétòn *v*. **look out** tɔtsɔ́ɔ́n *v*. **look over** iséméés *v*.; iséméetés *v*. **looking great** naƙwídɔ̀n *v*. **looking very good** nàƙw<sup>ɨ</sup> *ideo*. **lookout** dìyw<sup>a</sup> *n*. **loop** ɨnɔɛs *v*. **loop around** ɨnɔɛtɛ́s *v*. **loose** ɓàŋɔ̀n *v*.; dolódòn *v*.; ɨɓámɔ́n *v*.; lokilókón *v*.; lokódòn *v*.; nalóʝón *v*.; roiróón *v*. **loose (of stool)** dulúmón *v*.; erúxón *v*. **loose (sexually)** ɨmáláánón *v*. **loosely** dòl *ideo*.; hàʝ<sup>a</sup> *ideo*.; làʝ<sup>a</sup> *ideo*.; lòk<sup>o</sup> *ideo*. **loosely tied** haʝádòn *v*. **loosely tied down** laʝádòn *v*.; yaŋádòn *v*. **loosen** hoɗómón *v*.; ɨlaʝíláʝɛ́s *v*.; iloílóés *v*.; ɨnɔɨnɔɛ́s *v*. **loosen (soil)** iwúlákés *v*. **looseness** ɓaŋás *n*. **loot** taɓales *v*. **lop** isésélés *v*. **lop off (branches)** iteɗes *v*. **lope** isipísípòn *v*.; ɨsɔƙísɔ́ƙɔ̀n *v*.; ɨsʉmʉ́sʉ́mɔ̀n *v*. **lord** ámáze *n*.; ámázeám *n*. **Lord's Prayer** Ɲakuʝíwáán *n*. **lorry** lóórì *n*.; ɲolórì *n*. **lose** buanítésuƙot<sup>a</sup> *v*.; ɡóózesuƙot<sup>a</sup> *v*.; iloimétòn *v*.; rúmánòn *v*.
**lose (drop)** tuɓutes *v*.; tuɓútésuƙot<sup>a</sup> *v*. **lose (sth. valuable)** iɲekes *v*.; iɲékésuƙot<sup>a</sup> *v*. **lose interest** bɔrɛ́tɔ́n *v*.; kitsonuƙot<sup>a</sup> *v*. **lose one's mind** itwáŋón *v*. **lose red-brown color** teɓúránétòn *v*. **lose the way** hakonuƙot<sup>a</sup> *v*.; itwáŋón *v*. **losing teeth** tolótólánón *v*. **loss (take a)** totóánonuƙot<sup>a</sup> *v*. **lost** buanón *v*.; ɡóózosuƙot<sup>a</sup> *v*.; totóánonuƙot<sup>a</sup> *v*. **lost (get)** buanónuƙot<sup>a</sup> *v*.; itwáŋón *v*.; totóánonuƙot<sup>a</sup> *v*. **lot** muce *n*.; ɲaɓáát<sup>a</sup> *n*. **lots** kom *n*. **loud** bɔrɔ́ɔ́n *v*.; ilélémùòn *v*. **loud person** nɔ̀sààm *n*. **lounge** iríƙímánón *v*. **louse** ts'an *n*. **lousiness** pásìnànès *n*. **lousy** pás *n*. **love** mínɛ́s *v*. **love (make)** èpòn *v*. **love child** ŋabɔ́bòìm *n*. **love each other** mínínɔ́s *v*. **love of neighbors** mínínɔ́sá na áwìkà<sup>e</sup> *v*. **love that is feigned** mínínɔ́sá na ɨɓám *v*. **loveliness** daás *n*. **lovely** dòòn *v*. **lovely (make)** daites *v*. **lover** epúám *n*.; mínɛ́sìàm *n*.; mínínɔ́síàm *n*. **lover of sleep** epúám *n*. **low** kúɗón *v*.; ŋʉɗʉ́sʉ́mɔ̀n *v*.; ŋʉsʉ́lʉ́mɔ̀n *v*. **low (of cows)** erutánón *v*. **low (of many)** kúɗaakón *v*.
#### lower
**lower** ɓuƙetés *v*.; iyolíyólés *v*. **lower oneself** iyééseetésá así *v*. **lowland** ɡíɡìr *n*. **lowland living** dziŋánànès *n*. **lowness** kuɗás *n*. **lubricant** ceím *n*. **lubricate** ŋiites *v*.; ŋiitésúƙot<sup>a</sup> *v*. **Lucifer** Siitán *n*. **luck** muce *n*.; ɲaɓáát<sup>a</sup> *n*. **luck (average)** mucea na ɓárɨɓár *n*. **luck (awful)** mucea na ináƙúós *n*. **luck (bad)** mucea ná ʝɔ̀l *n*. **luck (good)** mucea na títìàn *n*.; ɲaréréŋ *n*. **lucky** tírɨríŋɔ́n *v*. **Ludo (game)** ɲélúɗo *n*. **Luganda** Ŋímúɡanɗáétòd<sup>a</sup> *n*. **luggage** botitín *n*. **Luke** Lúkà *n*. **Luke (biblical)** Lúkà *n*. **lukewarm** muʝálámòn *v*. **lunar eclipse** badona aráɡwaní *n*. **lung** ɡàfìɡàf *n*. **lurch** ɡakímón *v*.; nɛ́rɨnɛ́rɔ́n *v*. **lurch (of heart)** mulúráŋòn *v*. **lure** ɨmɔɗɛtɛ́s *v*.; ɨsʉ́ŋʉ́rɛ́s *v*. **lure (bees)** sɨsɨɓɛs *v*. **luxate** ƙwɨʝɛ́s *v*. **luxated** ƙwɨʝímɔ́n *v*. **Lycopersicon esculentum** ɲɛ́ɲaaɲá *n*. **lying down** èpòn *v*. **lymph node** ƙùts'àts'<sup>a</sup> *n*. **machete** ɲápaŋká *n*. **machine** ɲámasín *n*. **machine gun (belt-fed)** zɔ̀tɛ̀ɛ̀b <sup>a</sup> *n*.
**mad person** lejéàm *n*. **made** ɨɗɨmɔ́s *v*.; ɨɗɨmɔtɔ́s *v*. **madness** leɡé *n*.; lejé *n*.; lejéèd<sup>a</sup> *n*.; ŋkérép<sup>a</sup> *n*. **Maerua angolensis** lóɗíwé *n*. **Maerua pseudopetalosa** ɡomóí *n*. **Maerua triphylla** iroroy<sup>a</sup> *n*. **magazine (of a gun)** ɲókópo *n*. **magnanimity** daás *n*. **magnanimous** dòòn *v*. **magnanimous person** dòònìàm *n*. **maiden** ɲàràm *n*. **maidens** ɲèr *n*. **maintain** ɨsáɓísɨŋɛɛ́s *v*. **maintain poorly** ɨsɔ́ɓɔ́lɛ́s *v*. **maize** màlòr *n*.; màlòrìèɗ<sup>a</sup> *n*.; ɲaɓʉra *n*. **maize (unripe)** íɡùm *n*.; káruɓú *n*.; ɲaŋárʉ́tɛ̀ *n*. **maize cob** ɲaɓʉraídàkw<sup>a</sup> *n*. **maize deformity** ŋƙwa *n*.; tsɔ́rákwɛ̀t <sup>a</sup> *n*. **maize kernels (milky)** îdw<sup>a</sup> *n*. **maize kernels (tough)** lúɡùm *n*. **maize variety** áɡɨrɨkácà *n*.; aráɡwànà kɔ̀n *n*.; ɲóɗòmòŋòlè *n*. **maize variety (black and white)** lɔkíríɗɨɗí *n*. **maize variety (multicolored)** katólìkà *n*. **maize variety (white)** katʉmán *n*.; nakatʉmán *n*. **make** bɛrɛ́s *v*.; ɨɗɨmɛ́s *v*.; ɨɗɨmɛ́sʉ́ƙɔt<sup>a</sup> *v*.; ɨɗɨmɛtɛ́s *v*.; itues *v*.; ituetés *v*.; tɔsʉɓɛs *v*. **make a** *sh-sh* **sound** wɔxɔ́dɔ̀n *v*. **make a hole in** iɓólóɲés *v*.; itoɓes *v*. **make a mistake** hakonuƙot<sup>a</sup> *v*.; tɔsɛ́sɔ́n *v*. **make a racket** nɔsátón *v*.
manyatta
**make a way through** utés *v*.; utésúƙot<sup>a</sup> *v*. **make blocks** iwésá ɲéɓulókìkà<sup>e</sup> *v*. **make bricks** iwésá ɲéɓulókìkà<sup>e</sup> *v*. **make excellently** iyomes *v*. **make fun of** tɔʝɛmɛs *v*. **make hole in** pulés *v*. **make holes in repeatedly** itoɓítóɓés *v*. **make ill** ɨnʉɛs *v*. **make into one** kɔnítɛ́sʉƙɔt<sup>a</sup> *v*. **make into porridge** iúʝietés *v*. **make like** iretes *v*. **make money** itsúrútseés *v*. **make noise** arútón *v*.; fútón *v*. **make oneself look good** daitetésá así *v*. **make peace** apápánɛ̀ɛ̀tɔ̀n *v*.; apápánɔ̀ɔ̀n *v*. **make plans** ipáŋƙeés *v*. **make poorly** ɨtáƙálɛ́s *v*. **make roadblock** teɡelesa ɲerukuɗeé *v*. **make to collect firewood** waitésuƙota dakwí *v*. **make trouble for** ɨtsánítɛtɛ́s *v*. **make up (invent)** iroketés *v*. **make up (lies)** isuɗes *v*.; isuɗetés *v*. **make use of** eréɡes *v*. **maker** ɨɗɨmɛtɛ́síàm *n*.; tɔsʉɓɛtɛ́síàm *n*. **maladroit** hádaadánón *v*. **malady** màyw<sup>a</sup> *n*. **malar** akáƙúm *n*. **malaria** suɡur *n*. **male** cúrúk<sup>a</sup> *n*. **male animal** cikw<sup>a</sup> *n*. **malediction** ìlàm *n*. **malicious person** ɲɛ́kɨsɨránìàm *n*. **maliciousness** ɲɛ́kɨsɨrán *n*.
**malleable** lumúdòn *v*. **malleably** lùm *ideo*. **malnourished child** dódòk<sup>a</sup> *n*. **malt** ìkɔ̀ŋ *n*. **mamba** míɔ̀k <sup>a</sup> *n*. **man** eakw<sup>a</sup> *n*. **man (old)** ʝákám *n*. **man (young)** ŋímɔ́kɔkáám *n*.; ŋísɔ́rɔkɔ́ám *n*. **manage** ɨrɨtsɛ́s *v*.; totseres *v*. **manage each other** totsérímós *v*. **manageable** ɔfɔ́dɔ̀n *v*.; olódòn *v*. **manager** ɨrɨtsɛ́síàm *n*. **mancala (game)** ɲékilelés *n*. **mandible** ƙálíts'<sup>a</sup> *n*. **mandibular angle** ƙaƙúŋ *n*. **mandibular bone** ƙálíts'ìɔ̀k <sup>a</sup> *n*. **mane** sìɡìrìɡìr *n*. **maneuver (manually)** ƙɔxɛ́s *v*. **mange** ɲɛ́kʉ́rara *n*. **manger** itúɓ<sup>a</sup> *n*. **Mangifera indica** ɲɛ́mɨɛ́ḿɓɛ̀ *n*. **mangle** ɨɛmíɛ́mɛ́s *v*. **mango** ɲɛ́mɨɛ́ḿɓɛ̀ *n*. **manhood** eakwánánès *n*. **Manihot species** ɲómoŋɡó *n*. **manioc** ɲómoŋɡó *n*. **manipulate (manually)** ƙɔxɛ́s *v*. **manliness** eakwánánès *n*. **manmade** ɨɗɨmɔtɔ́sá ròɓ<sup>o</sup> *v*. **manner** ɲɛpɨtɛ *n*. **manure** ɦyɔ̀èts'<sup>a</sup> *n*.; ŋɔt<sup>a</sup> *n*. **many** kom *n*.; kòm *quant*.; kòmòn *v*.; tʉ̀mɛ̀ɛ̀ *n*. **many (become)** ítónà dìdìk<sup>e</sup> *v*.; komonuƙot<sup>a</sup> *v*. **manyatta** aw<sup>a</sup> *n*.
**manyness** komás *n*. **map** ɲámáp<sup>a</sup> *n*. **maraud** toɓés *v*. **marauder** toɓésíàm *n*. **march** iríánònà dɛ̀ìkà<sup>ɛ</sup> *v*.; ɲéperét<sup>a</sup> *n*.; tɛtsɛ́sá ɲéperétì *v*. **March** Dáŋ *n*.; Lɔɗʉ́ŋɛ *n*. **marching orders** tàŋàs *n*. **marginalization** ɲoloɗiŋ *n*. **marginalize** ɨlɔ́ɗíŋɛ́s *v*. **marginalizing** ɨlɔ́ɗíŋánón *v*. **marijuana** lɔ́tɔ́ɓa ná zè *n*. **Mark** Máríkò *n*. **mark** ɨƙɛrɛs *v*.; ɨmátsárɛ́s *v*.; ɨsɛɓɛs *v*.; itwelítwélés *v*.; iwetés *v*.; ɲámátsar *n*.; totsetes *v*. **Mark (biblical)** Máríkò *n*. **mark (on skin)** tás *n*. **mark (signs)** ɨƙɨrɛs *v*. **marked** ɨsɛɓɔs *v*.; itwelítwélós *v*. **market** dzíɡwààw<sup>a</sup> *n*.; ɲámákɛ̀t <sup>a</sup> *n*. **market (Kenyan)** ɲɛ́fíl *n*. **market (open air)** ɗípɔ̀ *n*. **maroon** kɨpʉ́ránètòn *v*. **married (of a bride)** buƙós *v*. **marrow** hɛ̀ɡ <sup>a</sup> *n*. **marry by taking** ɨʉmɛs *v*. **marry by taking away** ɨʉ́mɛ́sʉƙɔt<sup>a</sup> *v*. **marry polygamously** ramɛ́s *v*. **marsh** ɲéʝem *n*. **marsh (seasonal)** ɲɛkípɔ́r *n*.; ɲotóbòr *n*. **marshily** fɔ̀ts'<sup>ɔ</sup> *ideo*. **marshy** fɔts'ɔ́dɔ̀n *v*. **masculinity** eakwánánès *n*. **mash (grist)** mʉrɛ́s *v*. **mash (sour)** ɓaram *n*.
**massacre** ɡɨʝɛtɛ́s *v*. **massage** ʝʉ́rɛ́s *v*. **massage out** ʝʉ́rɛ́sʉƙɔt<sup>a</sup> *v*.; ʝʉrɛtɛ́s *v*. **massager** ʝʉ́rɛ́sìàm *n*. **masseuse** ʝʉ́rɛ́sìàm *n*. **master** ámáze *n*.; ámázeám *n*. **masterpiece** iyomam *n*. **masticate** ɨɲáɗʉ́tɛ́s *v*. **masticate (tobacco)** ɨmátáŋɛ́s *v*.; mataŋɛs *v*. **mastitis** ídoɲeɗeké *n*. **masturbate** ɨʝɔƙíʝɔ́ƙɛ́sá kwaní *v*. **mat** ɲámát<sup>a</sup> *n*. **mat (leather)** jèjè *n*. **mat (small leather)** ɗɛ́f *n*. **mat (termite-drying)** uré *n*. **match** ɲékiɓirít<sup>a</sup> *n*.; topútétòn *v*. **matchstick** ɲékiɓirít<sup>a</sup> *n*. **mate with** tirés *v*. **mate with each other** tirímós *v*. **math** ɲámára *n*. **mathematics** ɲámára *n*. **matted** kémúsánón *v*. **matted hair** kémús *n*. **matters** mɛn *n*. **Matthew** Matéò *n*. **Matthew (biblical)** Matéò *n*. **mattress** ɲápalís *n*. **mature** iríétòn *v*.; kɔkɔsánón *v*.; zòòn *v*.; zoonuƙot<sup>a</sup> *v*. **mature (of many)** zeikaakón *v*. **mature sexually** teɓúránétòn *v*. **mature sexually (of boys only)** ɨɓʉyákòn *v*. **maturity** zeís *n*. **May** Kɨnám *n*.; Titímá *n*. **may …** kóʝ<sup>a</sup> *adv*.
… **maybe** ƙámá kiɗíé *v*.; ndóó ɦyè *n*. **Maytenus undata** múrotsíò *n*. **mbira** lokemú *n*. **me** ŋ́k <sup>a</sup> *pro*. **mead** sɨs *n*.; ts'ɔƙam *n*. **meadow (flat)** rɔw<sup>a</sup> *n*. **meal mush** ɨlɨram *n*.; tɔbɔŋ *n*.; tʉɗʉtam *n*. **meal mush (solid)** lúɡùm *n*. **meal mush (watery)** ɓɔtí *n*. **mealy** ɡwɛrɛ́ʝɛ́ʝɔ̀n *v*. **mean** ɦyɛtɨɦyɛtɔs *v*.; ɦyɛ̀tɔ̀n *v*.; tákés *v*. **mean to do** ɨwɔ́ŋɔ́n *v*. **mean talker** hákátònìàm *n*. **meander** ɨkɔɗíkɔ́ɗɔ̀n *v*.; iƙulúƙúlòn *v*.; **meander/weave** lúkúɗukuɗánón *v*. **meaning** zeísêd<sup>a</sup> *n*. **meaningful** zízòn *v*. **meaningless** buɗámón *v*. **meanness** ɦyɛtás *n*. **measles** púrurú *n*. **measure** ɨpɨmɛs *v*.; kêd<sup>a</sup> *n*. **measure words** íziɗesa tóda<sup>e</sup> *v*. **meat** em *n*. **meat (charred)** kɔr *n*. **meat (dried)** ŋátɔɔsa *n*. **meat (rib)** ŋábèrìkèèm *n*. **meat (skewered)** rɔam *n*. **meat dried on hide** xáƙw<sup>a</sup> *n*. **meat hunger** bisák<sup>a</sup> *n*. **meat hunger (satisfy)** ɨtsɔ́ítɛ́sʉƙɔt<sup>a</sup> *v*.; ɨtsɔ́ɔ́nʉƙɔt<sup>a</sup> *v*. **meat-carrying call** waín *n*. **mechanic** ɨɗɨmɛ́síàm *n*.; ŋífunɗíàm *n*. **medal** ɡwas *n*.
**mediate** terés *v*. **medicate** irés *v*. **medicine** cɛ̀mɛ̀r *n*. **mediocre** ŋwanɨŋwánɔ́n *v*. **medium-sized** ɓarɨɓárɔ́n *v*.; ʝɔ̀ƙɔ̀n *v*.; lerúkúmòn *v*. **medulla spinalis** lɔ́ɓírɨɓír *n*. **meek** ɨɛ́ɓɔ́n *v*. **meet** ikíkóanón *v*.; imánétòn *v*.; itóyéésa así *v*.; ɨtsʉnɛtɛ́sá así *v*.; ɨtsʉ́nɛ́tɔ̀n *v*.; itukánón *v*.; ŋimánétòn *v*.; ɲimánétòn *v*.; ɲimánón *v*. **meet together** ɗɔtsánónuƙot<sup>a</sup> *v*.; iryámíryámètòn *v*. **meet up** imánónuƙot<sup>a</sup> *v*. **meet while dancing** ilépón *v*. **meet with** iryámétòn *v*. **meeting** kur *n*.; ɲékíìkò *n*.; ɲémítìŋ *n*. **melange** ɲalíɲalí *n*.; ɲɔ́tsɔ́ɓɨtsɔɓ<sup>a</sup> *n*. **melon species** nàdɛ̀kwɛ̀l *n*. **melt** cuanónuƙot<sup>a</sup> *v*.; laʝámétòn *v*. **melt (in mouth)** ɨnʉƙʉ́nʉ́ƙwɛ́s *v*. **melt away** rìmòn *v*. **memorize** tamɨtɛtɛ́s *v*. **memory (have a good)** ɨpííríánón *v*. **men** ɲɔt<sup>a</sup> *n*. **men (old)** ʝák<sup>a</sup> *n*. **men (young)** karatsʉ́na *n*.; ŋímɔ́kɔka *n*.; ŋísɔ́rɔk<sup>a</sup> *n*.; pànɛ̀ɛ̀s *n*. **men-crazy** ɨɲɔ́táánón *v*. **menace** zízɛ̀s *v*. **mend** ɲimanites *v*.; rátsɛ́s *v*.; taɗapes *v*. **mend repeatedly** rátsiés *v*. **mend up** taɗapetés *v*. **mend with fire** ɨtsʉŋɛs *v*. **mend with mud** nutsés *v*.
**mended** taɗapos *v*. **mendicant** purutél *n*. **Mening language** Ŋímeniŋítôd<sup>a</sup> *n*. **Mening person** Ŋímeniŋíám *n*. **meningitis** ɲɛ́tɛrɛƙɛ́ƙɛ *n*.; tɛ́rɛƙɛ́ƙɛ *n*. **menopause** ŋʉrʉ́mɔ́na ƙwaaté<sup>o</sup> *v*. **menstruate** iona aráɡwaník<sup>ɛ</sup> *v*.; itáléés *v*. **mental illness** leɡé *n*.; lejé *n*.; lejéèd<sup>a</sup> *n*.; lejénánès *n*.; ŋkérép<sup>a</sup> *n*. **mention** ilímítés *v*.; tákés *v*. **mentum** tatʉ́n *n*. **merchandise** dzíɡwam *n*.; dzíɡwetam *n*.; dzííƙotam *n*. **merciful** isyónón *v*. **merciful (become)** isyónónuƙot<sup>a</sup> *v*. **mercy on (have)** isyones *v*. **mesh** ɨlɔ́ƙɛ́rɛ́s *v*.; ƙídzatiés *v*. **mess up** hamʉʝɛ́s *v*.; imóɲíkees *v*.; imóɲíkeetés *v*. **mess with (in fun)** wáákitetés *v*. **message** ɲéripót<sup>a</sup> *n*. **message (morning)** sír *n*. **message (send out a)** mɛnɔnʉƙɔt<sup>a</sup> *v*. **messenger** dɛáám *n*. **metabolism (have a high)** hɔɓɔ́mɔ́n *v*. **metal** tsɨrɨm *n*. **metal pot** tsɨrɨmʉ́dòm *n*. **metal ringlet** àɡìt<sup>a</sup> *n*. **metalworker** ìtyàkààm *n*. **mete out** ɨɲíɲínɛ́s *v*. **meteor** ɗɔ́xɛatá na tsúwà *n*. **meter** ɲémíta *n*.; ɔkɔ́ts<sup>a</sup> *n*. **method** muce *n*.; ɲɛpɨtɛ *n*. **metropolis** zɛƙɔ́áwa ná zè *n*. **mewl** ɨɲɨɨɲíɔ̀n *v*.; ɨɲíɲíɔ̀n *v*. **Meyna tetraphylla** lòŋìr *n*.
**microbe** ƙʉts'<sup>a</sup> *n*. **midday** ikáɡwaríìk<sup>e</sup> *n*.; ódoo bɨrɨr *n*. **middle** sɨsɨk<sup>a</sup> *n*.; sɨsíkɛ̂d <sup>a</sup> *n*. **middle child** sɨsɨkáám *n*. **middle of path** mucéák<sup>a</sup> *n*.; mucéékw<sup>a</sup> *n*. **middle part** bakútsêd<sup>a</sup> *n*. **midget** puusúmòn *v*. **midnight** mukúásísík<sup>a</sup> *n*. **midrib** ɡòɡòròʝ<sup>a</sup> *n*. **midriff** kàɓ<sup>a</sup> *n*. **midwife** ʝʉ́rɛ́sìàm *n*.; ƙwaatítetés *v*.; ƙwaatítetésíàm *n*. **might** zeís *n*. **migraine headache** ɨɗíɔ́na iká<sup>e</sup> *v*. **migrant** botáám *n*.; botibotosíám *n*. **migrate** bòtòn *v*.; ilotsesa zɛƙɔ́ ɛ *v*. **migrate away** botonuƙot<sup>a</sup> *v*. **migrate this way** botétón *v*. **migration** bot<sup>a</sup> *n*. **migratory** botibotos *v*. **mild** ɨɛ́ɓɔ́n *v*. **mildew** lóburuʝ<sup>a</sup> *n*.; ɲóróiroy<sup>a</sup> *n*. **mile** máìrɔ̀ *n*. **mileage** máìrɔ̀ɛ̀d <sup>a</sup> *n*. **milk** îdw<sup>a</sup> *n*.; ʝʉ́tɛ́s *v*. **milk (cow)** ɦyòìdw<sup>a</sup> *n*. **milk (fresh)** ŋalɛ́pán *n*. **milk (from breast)** ámáìdw<sup>a</sup> *n*. **milk (sour)** ídwà nì ɓàr *n*. **milk bush** ɨnw<sup>a</sup> *n*. **milk tea** tábarɨcue *n*. **milk tooth** ídòkwàyw<sup>a</sup> *n*. **milk-leaf** ídòkàk<sup>a</sup> *n*. **milking gourd** ɲelépít<sup>a</sup> *n*. **mill** ŋɔ́ɛ́s *v*.; ŋɔɛsíɡwàs *n*.; ɲámasín *n*. **mill (gastric)** ŋìl *n*. **milled** ŋɔ́ɔ́s *v*.
**millet (brown)** ŋɔt<sup>a</sup> *n*. **millet (finger)** rêb<sup>a</sup> *n*. **millet (harvest)** ɨrábɛs *v*. **millet beer** ŋamarʉwáy<sup>a</sup> *n*.; rébèmɛ̀s *n*. **millet left in field** kírérebú *n*. **million** dakwa kɔn *n*.; ɲémílìòn *n*. **millipede** iƙórú *n*. **Mimusops kummel** lokum *n*. **mind** akílìk<sup>a</sup> *n*.; ɨkatsɛs *v*.; ɨmɨsɛs *v*.; ŋátámɛta *n*. **mine** ɲjɛ́n *pro*. **mingle in** íburuburés *v*. **mingle until stiff** tʉɗʉtɛtɛ́s *v*. **mingled stiff** tʉɗʉtɔs *v*. **mingling stick** tʉɗʉtɛsídàkw<sup>a</sup> *n*. **mingling stick (pronged)** ŋɛ́r *n*. **miniature** tɔ́ɗɔ́n *v*. **miniscule** dununúòn *v*.; tɔ́ɗɔ́n *v*. **minister** pásìtà *n*. **ministry** terêɡ<sup>a</sup> *n*. **mira** ɲémurúŋ́ɡù *n*. **miracle (perform a)** ikúʝíánón *v*. **miracles** itíónàs *n*. **miracles (do)** itíónòn *v*. **mirage** ɲéríɓiriɓ<sup>a</sup> *n*. **mire** dɔ̂b <sup>a</sup> *n*. **mirror** ɲɛ́ɡɨlás *n*.; ɲɛ́rʉ́ɛ́t <sup>a</sup> *n*. **misbehave** imákwéètòn *v*. **misbehave (get to)** imákóitetés *v*. **misbehaving** imákóòn *v*. **miscarry** iɲétséetés *v*.; iyétséetés *v*.; ƙúdetés *v*.; otés *v*.; otetés *v*. **miscreant** ɲárásíám *n*. **miserable** tsʉ́kʉɗʉ́ɗɔ́n *v*. **mishap (have a)** rúmánòn *v*.
**mishmash** ɲɔ́tsɔ́ɓɨtsɔɓ<sup>a</sup> *n*. **mislay** ɨtáƙálɛ́s *v*. **mislead** hakítésuƙot<sup>a</sup> *v*.; itwáŋítésúƙot<sup>a</sup> *v*. **misplace** buanítésuƙot<sup>a</sup> *v*.; ɡóózesuƙot<sup>a</sup> *v*.; ɨtáƙálɛ́s *v*. **misplaced** buanón *v*.; ɡóózosuƙot<sup>a</sup> *v*.; iɓíléròn *v*. **misplaced (become)** iɓíléronuƙot<sup>a</sup> *v*. **miss** fàlòn *v*.; ɨɗakɛ́s *v*. **miss (a shot)** ɨsɛɛs *v*. **miss narrowly** iwitses *v*. **miss out on** fàlòn *v*. **miss repeatedly** ɨsaɨsáyées *v*. **miss the point** ɨsɛɛsa mɛná<sup>ɛ</sup> *v*. **misscary repeatedly** iyétséyeés *v*. **missing (of an eye)** ɗooɲómòn *v*. **mission** ɲémíxòn *n*. **missionary** lɔʝɔkɔtáw<sup>a</sup> *n*. **missive** béɗíbeɗú *n*.; bóɗíboɗú *n*.; ɲáɓáruwa *n*. **mist** ɡóʒòw<sup>a</sup> *n*. **mistake** ɲasécón *n*.; ɲɔ́mɔkɔsá *n*. **mistake (make a)** hakonuƙot<sup>a</sup> *v*. **mistake for** ilotses *v*. **mistakenly** kèɗè *adv*. **mister** ámázeám *n*. **mistletoe** lɛ̂z *n*. **misuse** ɨɓalíɓálɛ́s *v*.; ɨlarɛs *v*.; ɨlwarɛs *v*. **mix** iɗyates *v*.; iɲales *v*.; ɨtsɔɓítsɔ́ɓɛ́s *v*.; itsulútsúlés *v*. **mix (grains)** ikáɗóés *v*. **mix (honey and termites)** ƙɛ́ƙɛ́rɛ́s *v*. **mix (porridge and mash)** toremes *v*. **mix in** ɗɔtsɛ́s *v*.; ɗɔtsɛ́sʉ́ƙɔt<sup>a</sup> *v*.; ídulés *v*.; íduludulés *v*.
moron
**mix up** ɨmɔrímɔ́rɛ́s *v*. **mix up (confuse)** ilotses *v*. **mixed** ɨtsɔɓítsɔ́ɓɔ̀n *v*.; ɨtsɔɓítsɔ́ɓɔ́s *v*. **mixed up** ɨmɔrímɔ́rɔ́s *v*. **mixture** ɲalíɲalí *n*. **mixture of fat and meat (be a)** tsokótsókánón *v*. **moan** éɓútòn *v*.; ɛ́mítɔ̀n *v*.; émúròn *v*. **mob** ɲéɓúku *n*. **mobile** ɓɛƙɛsɔs *v*. **mobile phone** dʉrʉdʉr *n*.; ɲásím *n*. **mobilize** iríréetés *v*. **mobilizer** ìʉ̀ɗààm *n*. **mock** tɔʝɛmɛs *v*. **model** ikwáánitetés *v*. **modern building** ɲeryaŋíhò *n*. **modern society** ɲeryaŋ *n*. **modernity** ɲeryaŋ *n*. **moist** ɗɔ̀ƙɔ̀n *v*. **moist (become)** ɗɔƙɔnʉƙɔt<sup>a</sup> *v*. **moisten** ɗɔƙítɛ́sʉƙɔt<sup>a</sup> *v*.; ɨpápɛ́ɛ́s *v*. **moisture-resistant** pʉrákámòn *v*.; pʉráŋámòn *v*.; pʉsɛ́lɛ́mɔ̀n *v*. **moisturize (skin)** ɨwasɛs *v*. **molar** tiróŋ *n*. **mold** lóburuʝ<sup>a</sup> *n*.; ɲóróiroy<sup>a</sup> *n*. **mole** ɲɛ́nʉkʉnʉ́kʉ *n*. **molecule** kiɗoɗots<sup>a</sup> *n*. **molest** tarates *v*. **molt** fòlòn *v*. **molting** ɗaráɗáránón *v*. **Monday** Ɲáɓarásà *n*.; Ɲákásíá kɔ̀nìk<sup>ɛ</sup> *n*. **money** kaûdz<sup>a</sup> *n*.; ŋárɔpɨyá *n*. **money tree** kaûdz<sup>a</sup> *n*. **mongoose (dwarf)** múɗèr *n*.
**mongoose (Egyptian)** mútèts<sup>a</sup> *n*. **mongoose (gray)** mútèts<sup>a</sup> *n*. **mongoose (slender)** sílɔlɔ́ʝ <sup>a</sup> *n*. **mongoose (white-tailed)** lóɓíliwás *n*. **moniliasis** losúk<sup>a</sup> *n*. **monitor** ɨfátɛ́sʉƙɔt<sup>a</sup> *v*. **monkey (colobus)** ɲɛ́cʉma *n*. **monkey (female)** ɔɡɛraŋwa *n*. **monkey (male)** ɔ̀ɡɛ̀r *n*. **monkey (patas)** koliméw<sup>a</sup> *n*. **monkey (vervet)** kaɗokóy<sup>a</sup> *n*. **monocular vision (have)** ɗooɲómòn *v*. **monotonous** itópénòn *v*. **monster** ɲaŋu *n*. **monster (of a)** kébàdà *n*.; nábàdà *n*.; nébàdà *n*. **month** aráɡwan *n*. **month of bad honey** Lotséto *n*. **month of honey** Nakaɓinín *n*. **month of weeding** Lɔɓalɛl *n*. **moo** erutánón *v*. **moo!** buúù *ideo*. **mooch** lɛŋɛ́s *v*. **moocher** lɛŋɛ́síàm *n*. **mooching** olíɓó *n*. **moody** kwits'íkwíts'ánón *v*. **moon** aráɡwan *n*. **moon (full)** aráɡwaníékw<sup>a</sup> *n*. **moon (new)** aráɡwaníɛ́bɨtín *n*. **more** kúbam *n*.; sa *pro*. **more than** ɨlɔɛs *v*. **moreover** naɓó *coordconn*.; toni naɓó *n*. **morning** barats<sup>a</sup> *n*. **morning glory** lòɓòlìà *n*. **moron** bóx *n*.; ɨɓááŋàsìàm *n*.
move in
**morrow** barats<sup>a</sup> *n*. **mortar** iwóts<sup>a</sup> *n*.; ɲómóta *n*. **mortar bottom** iwótsíɔ̀z *n*. **mortar mouth** iwótsíàk<sup>a</sup> *n*. **mosquito** kímʉ́r *n*. **mosquito (small)** tsorokoní *n*. **most likely** kárɨká *adv*. **mote** símíɗiɗí *n*. **motel** epúáw<sup>a</sup> *n*. **mother (his/her/its)** ŋwáát<sup>a</sup> *n*. **mother (my)** yáŋ *n*. **mother (your)** ŋɔ́*n*. **mother-in-law (her)** dádàt<sup>a</sup> *n*. **mother-in-law (his)** ntsíémetá *n*. **mother-in-law (his/her sibling's spouse's mother)** ŋwáátìɲòt<sup>a</sup> *n*. **mother-in-law (my sibling's spouse's mother)** yáŋìɲòt<sup>a</sup> *n*. **mother-in-law (my, of men)** ɲ́ciemetá *n*. **mother-in-law (my, of women)** dadáŋ *n*. **mother-in-law (of men)** emetá *n*. **mother-in-law (your sibling's spouse's mother)** ŋɔ́ɲót<sup>a</sup> *n*. **mother-in-law (your, of men)** biemetá *n*. **mother-in-law (your, of women)** dádò *n*. **motherhood** ŋwáátìnànès *n*. **motherliness** ŋwáátìnànès *n*. **motor** ɡúr *n*. **mottle** ɨtsɔɓítsɔ́ɓɛ́s *v*. **mottled** ɨtsɔɓítsɔ́ɓɔ̀n *v*.; ɨtsɔɓítsɔ́ɓɔ́s *v*. **mould** bɛrɛ́s *v*. **mound** ɨnʉkʉ́nʉ́kɛ́s *v*.; kìts<sup>a</sup> *n*.; ɲatúkít<sup>a</sup> *n*.
**mount** otsés *v*. **mount (a beehive)** rɔ́ƙɛ́s *v*. **mount an offensive** iríɓéés *v*. **mount up** otsésúƙot<sup>a</sup> *v*. **mountain** kwar *n*. **mountain dweller** kwàrìkààm *n*. **mountain saddle** kwaréékw<sup>a</sup> *n*. **mountainside** rutet<sup>a</sup> *n*. **mountaintop** kwaráɡwarí *n*. **mourn** ilúrón *v*.; turúnón *v*. **mourner** turúnónìàm *n*. **mouse** ɗér *n*. **mouse species** naɓálámorú *n*. **mousebird (speckled)** tsówír *n*. **mouth** ak<sup>a</sup> *n*.; akɛd<sup>a</sup> *n*. **mouth cover** ɲákáparat<sup>a</sup> *n*. **move** ɓɛƙɛ́s *v*.; dzuƙés *v*.; ɨlɔpɛs *v*.; ilotsesa zɛƙɔ́ ɛ *v*.; ɨsʉ́tɔ́n *v*.; rités *v*. **move (an object)** ɨsʉtɛs *v*. **move (emotionally)** tábès *v*. **move (migrate)** bòtòn *v*. **move after dark** buɗamés *v*. **move around** ɨlɨrɛs *v*.; ɨlɔmílɔ́mɔ̀n *v*.; irímón *v*.; ɨwaríwárɛ́s *v*. **move around in** ɨlɔpílɔ́pɛ́s *v*. **move around repeatedly** ɨlɨrílírɛ́s *v*. **move aside** ècòn *v*.; èkòn *v*. **move away** botonuƙot<sup>a</sup> *v*.; dzuƙésúƙot<sup>a</sup> *v*.; ɨsʉ́tɛ́sʉƙɔt<sup>a</sup> *v*.; ritésúƙot<sup>a</sup> *v*. **move away on buttocks** dɔ́dɔrɔnʉƙɔt<sup>a</sup> *v*. **move back** raʝánón *v*. **move blindly** ɓɛƙɛ́sá buɗámík<sup>e</sup> *v*. **move down** kídzìmòn *v*. **move emotionally** kʉpɛ́s *v*. **move in** ínésuƙot<sup>a</sup> *v*.; toƙízeesá así *v*.; toƙízèètòn *v*.; zɛƙwɛ́tɔ́n *v*.
**move in single file** ɨʝíílɔ̀n *v*.; tɔɗʉ́pɔ́n *v*.; torópón *v*. **move off** ɨsʉ́tɛ́sʉƙɔt<sup>a</sup> *v*. **move on all fours** tolíón *v*. **move on buttocks** dɔ́dɔ̀rɔ̀n *v*. **move oneself rhythmically** ɨtɨnítínɛ́sá así *v*. **move out** ritetés *v*. **move past** ɓʉ̀nɔ̀n *v*. **move quickly** ikómóòn *v*. **move rhythmically** ɨtɨnítínɛ́s *v*. **move slowly** inípónòn *v*. **move straight** ɨɗírírɔ̀n *v*. **move this way** botétón *v*.; dzuƙetés *v*.; ɨsʉtɛtɛ́s *v*. **move to a point** irídòn *v*. **move together** toríkínós *v*. **move up and down alternately** iyopíyópòn *v*. **movement (migration)** bot<sup>a</sup> *n*. **movie** kúrúkúríka ni ɓɛƙɛ́s *n*.; ɲévíɗyo *n*. **mow** ɨɗɛtɛs *v*.; ɨrɛʝɛs *v*. **much** iruɓes *v*. **muchomo** kɔr *n*. **mucilage** ɗòs *n*. **muck** dɔ̂b <sup>a</sup> *n*. **muck things up** nts'áƙóna sèrèìk<sup>e</sup> *v*. **mucus** ɗɔ̀ƙɔ̀n *n*.; ɲarʉ́kʉ́m *n*. **mucus (cervical)** ɡaɗár *n*. **mucus (dried)** dɔ̀x *n*. **mud** dɔ̂b <sup>a</sup> *n*. **mud (plaster)** tanaŋes *v*. **muddled talk** dɔ́bàtòd<sup>a</sup> *n*. **mudflap** fɔ́ɗ <sup>a</sup> *n*. **mudslide** bɔ̀rɔ̀ts<sup>a</sup> *n*.; dìdìàk<sup>a</sup> *n*. **muffler** ts'údemucé *n*.
**mug** ɲámáƙ<sup>a</sup> *n*. **Muganda** Ŋímúɡanɗáéàm *n*. **mull over** ɲɛɓɛ́s *v*.; tamátámatés *v*.; tamítámiés *v*. **multi-patterned** mɛrɨmɛ́ránètòn *v*. **multiplicity** komás *n*. **multiply** bɨtɛ́tɔ́n *v*.; bɨtɨtɛtɛ́s *v*.; komítésuƙot<sup>a</sup> *v*. **multiply oneselves** komitésá así *v*. **multitude** kom *n*.; ɲerípírìp<sup>a</sup> *n*.; òdìòs *n*.; tʉ̀mɛ̀ɛ̀ *n*. **multitudinous** kòmòn *v*. **mumble (food)** iŋulúŋúlés *v*. **munch happily** ɨlʉ́mʉ́lʉ́mɛ́s *v*. **munchy** haʉ́dɔ̀n *v*. **murder** ɨɗɛɛs *v*. **murder (many)** sáɓés *v*. **murder (singly)** cɛɛ́s *v*.; cɛɛ́sʉ́ƙɔt<sup>a</sup> *v*. **murder each other** ɨɗáínɔ́s *v*. **murder repeatedly** ɨɗaiyes *v*. **murderer** sèààm *n*. **murderer (of many)** sáɓésìàm *n*. **murderer (singly)** cɛɛsíám *n*. **murky** kaɓúrútsánón *v*. **murmur** ídʉlɨdʉlɛ́sa tódà<sup>e</sup> *v*.; ɨŋʉrʉ́ŋʉ́rɔ̀n *v*. **Musa species** ɲómototó *n*. **muscle** em *n*. **muscle (abdominal)** ɲákwálɨkwal *n*. **muscle (cowl)** ɲálaƙamáít<sup>a</sup> *n*. **muscle (external oblique)** ɲɔpɔl *n*. **muscle (intercostal)** kileleɓú *n*. **muscle (mylohyoid)** mukét<sup>a</sup> *n*. **muscle (perineal)** ɲalamatsar *n*.; ɲekiɗoŋit<sup>a</sup> *n*. **muscle (plantaris)** ƙóróèm *n*. **muscle (rhomboid)** ɲɛ́sílɨsɨl *n*. **muscle (sacral)** ɲɛtsɨr *n*.
**muscle twitching** bàɗ<sup>a</sup> *n*.
**mush (meal)** ɨlɨram *n*.; tɔbɔŋ *n*.; tʉɗʉtam *n*.
**mushily** bùr *ideo*.; dàb<sup>u</sup> *ideo*.; dùl *ideo*.; dùx *ideo*.
**mushroom** kɨnám *n*.
**mushroom (dik-dik)** ɲólíkɨnám *n*.
**mushroom (elephant)** oŋorikɨnám *n*.
**mushroom species** lɔmɔ́y <sup>a</sup> *n*.; ŋíɓalɛl *n*.; ŋits'e *n*.
**mushy** burádòn *v*.; dabúdòn *v*.; dulúdòn *v*.; duxúdòn *v*.
**Muslim** tuɗúlónìàm *n*.
**must** ɨtámáánón *v*.
**must have … (earlier today)** nábàts<sup>e</sup> *adv*.
**must have … (long ago)** nánòk<sup>o</sup> *adv*.
**must have … (yesterday)** násàm *adv*.
**muster** iríréetés *v*.
**mute** mìɲɔ̀n *v*.
**mutilate** ɨɛmíɛ́mɛ́s *v*.
**mutter** ídʉlɨdʉlɛ́sa tódà<sup>e</sup> *v*.
**mutton** ɗóɗòèm *n*.
**muzzle (of weapon)** akɛd<sup>a</sup> *n*.
**my** ŋ́k <sup>a</sup> *pro*.
**my cousin child** totóìm *n*.
**my friend** nádzàƙ<sup>a</sup> *n*.
**myopic** mumúánón *v*.
**myself** ɲ́cìnèb<sup>a</sup> *n*.
**mythical beast** ɲaŋu *n*.
**nah** ńtóodó *interj*.; ńtóondó *interj*.
**nail** ɡóɡès *v*.; ɲɔ́sʉmár *n*.
**nail (finger)** tíbòlòkòɲ *n*.
**nailrod** ìsìk<sup>a</sup> *n*.
**naive** ɨɓááŋɔ̀n *v*.
**naivete** ɨɓááŋàs *n*.
**naked** ilérón *v*.; lemúánètòn *v*.; leŋúrúmòn *v*.; sɨlɔ́ʝɔ́mɔ̀n *v*.; tuɗúsúmòn *v*.
**naked (totally)** wɛ̀r *ideo*.
**name** êd<sup>a</sup> *n*.; kʉ̀tɔ̀n *v*.; óés *v*.
**name (hill/mountain)** Aŋatár *n*.; Aŋolekók<sup>a</sup> *n*.; Cùcùèìk<sup>a</sup> *n*.; Curuk<sup>a</sup> *n*.; Curukúdɛ̀ *n*.; Ɗʉmánámérìx *n*.; Dímánìàk<sup>a</sup> *n*.; Dúnémorók<sup>a</sup> *n*.; Gàlàts<sup>a</sup> *n*.; Gomóíàw<sup>a</sup> *n*.; Góʒòwìk<sup>a</sup> *n*.; Ìmɛ̀r *n*.; Iwar *n*.; Kaacikóy<sup>a</sup> *n*.; Kaakámár *n*.; Kaatíríám *n*.; Kaɓʉ́tákurí *n*.; Kádzàn *n*.; Kakaɗ<sup>a</sup> *n*.; Kàkʉ̀tà *n*.; Kàlèànàŋìrò *n*.; Kàlèwèr *n*.; Kalɔbɛɲɛ́ɲ *n*.; Kaloŋoléárɛ́ŋan *n*.; Kamɔ́rɔ́mɔrát<sup>a</sup> *n*.; Kanamútó *n*.; Kanatárúk<sup>a</sup> *n*.; Kàpɛ̀tà *n*.; Kapɛtapʉ́s *n*.; Karéɲaŋ *n*.; Karʉmɛmɛ́ *n*.; Kátárʉkɔ́t <sup>a</sup> *n*.; Katsakól *n*.; Katsolé *n*.; Kàxìɛ̀rà *n*.; Keepák<sup>a</sup> *n*.; Kɛ́tɛ́l *n*.; Kilóróŋ *n*.; Kɔ́cɔ́kɨɔ *n*.; Kocom *n*.; Kòfòè *n*.; Kókósowa *n*.; Kɔ̀pàkwàr *n*.; Kotorúbé *n*.; Kùɓààw<sup>a</sup> *n*.; Kʉráhò *n*.; Laatso *n*.; Lɔbɛɛ́l *n*.; Lɔcáráƙwat<sup>a</sup> *n*.; Locom *n*.; Lɔcɔ́ríàlɔ̀sìà *n*.; Lɔ́ɗɔ́wɔ̀n *n*.; Lɔɗʉ́r *n*.; Lɔkaaƙɨlɨt<sup>a</sup> *n*.; Lòkìlè *n*.; Lokinéne *n*.; Lokipáka *n*.; Lɔkɨtɔ́y <sup>a</sup> *n*.; Lɔkwakaramɔ́y <sup>a</sup> *n*.; Lɔ́mɛ́ʝ a *n*.; Loméríɗok<sup>a</sup> *n*.; Lɔmíʝ<sup>a</sup> *n*.; Lɔ́mɨl *n*.; Lomoɗóɲ *n*.; Lɔɲákw<sup>a</sup> *n*.; Lɔ́ŋʉ́sʉl *n*.; Loocíkwa *n*.; Lɔɔɗíŋ *n*.; Lòòɗòs *n*.; Looɗóy<sup>a</sup> *n*.; Lopéɗó *n*.; Lɔpɛ́t <sup>a</sup> *n*.; Lopokók<sup>a</sup> *n*.; Lopúwà *n*.; Lɔ̀sɛ̀rà *n*.; Lósíl *n*.; Lòsòlìà *n*.; Lotim *n*.; Lotíyá *n*.; Lòtsòròɓò *n*.; Loukómor *n*.; Lowákuʝ<sup>a</sup> *n*.; Mɔƙɔrɔ́ɡwàs *n*.; Morúaɲáo *n*.; Moruaŋákiné *n*.; Morúaŋápɨɔn *n*.; Moruaŋípi *n*.; Morúaŋítà *n*.; Morúápólón *n*.; Morúárɛ́ŋán *n*.; Morúatap<sup>a</sup> *n*.; Morúédikay<sup>a</sup> *n*.; Morúéris *n*.; Morukoyan *n*.; Morúlem *n*.; Morúɲaŋ *n*.; Morúŋole *n*.; Mùƙè *n*.; Ŋasɛp<sup>a</sup> *n*.; Ŋusuman *n*.; Náápoŋo *n*.; Naɡomocóm *n*.; Naiɗíɗ<sup>a</sup> *n*.; Náìtà *n*.;
Náɨtáy<sup>a</sup> *n*.; Nakaɗapaláít<sup>a</sup> *n*.; Nakalalé *n*.; Nakíríkɛ̀t <sup>a</sup> *n*.; Nàkòrìtààw<sup>a</sup> *n*.; Nakɔrɔɗɔ́ *n*.; Naŋóléɓok<sup>a</sup> *n*.; Napitiro *n*.; Napóroto *n*.; Narúkyeɲ *n*.; Nasurukéɲ *n*.; Natípem *n*.; Natsíátà *n*.; Naʉrat<sup>a</sup> *n*.; Nawáɗow<sup>a</sup> *n*.; Ɲakwác<sup>a</sup> *n*.; Ɲèràdzòɡ<sup>a</sup> *n*.; Ɲèràtàɓ<sup>a</sup> *n*.; Ɔrɔ́m *n*.; Ɔ́pʉs *n*.; Paalakán *n*.; Páɗɛ̀rɛ̀hò *n*.; Palúùkùɓ<sup>a</sup> *n*.; Pílíkìts<sup>a</sup> *n*.; Pútá *n*.; Rɔ́ɡɛ̀hò *n*.; Ròŋòt<sup>a</sup> *n*.; Sààŋìròàw<sup>a</sup> *n*.; Seɡeríkwár *n*.; Séíkwàr *n*.; Séítíníkokór *n*.; Sɛkɛɗíáw<sup>a</sup> *n*.; Soƙoɡwáás *n*.; Tsakɨrɨk<sup>a</sup> *n*.; Tsakúdèɓò *n*.; Tsíɡàk<sup>a</sup> *n*.; Tsɔŋɔ́rán *n*.; Tsɔ́ráàw<sup>a</sup> *n*.; Taɓákókór *n*.; Tòlòyà *n*.; Tòòrwààk<sup>a</sup> *n*.; Tutét<sup>a</sup> *n*.
**name (personal)** Acók<sup>a</sup> *n*.; Acúkwa *n*.; Àɗùpà *n*.; Aemun *n*.; Akaɗéérót<sup>a</sup> *n*.; Akɔl *n*.; Akóóro *n*.; Akúɗúkori *n*.; Apáálokiɓúk<sup>a</sup> *n*.; Apáálokúk<sup>a</sup> *n*.; Apáálomúƙ<sup>a</sup> *n*.; Apáálòŋìrò *n*.; Apáásiá *n*.; Apérít<sup>a</sup> *n*.; Apʉs *n*.; Aramasán *n*.; Aríkó *n*.; Áryánkòrì *n*.; Asiróy<sup>a</sup> *n*.; Ceɡem *n*.; Ɗɔan *n*.; Dakáy<sup>a</sup> *n*.; Ɛ́kìtɛ̀là *n*.; Erupe *n*.; Gutí *n*.; Íʝéekw<sup>a</sup> *n*.; Ilʉ́kɔ́l *n*.; Irwátà *n*.; Itírá *n*.; Ìʉ̀ɗà *n*.; Kali *n*.; Kalɨmapʉ́s *n*.; Kalɔyáŋ *n*.; Kawes *n*.; Kinimé *n*.; Kocí *n*.; Kɔɛ́s *n*.; Kɔkɔ́ *n*.; Kokóy<sup>a</sup> *n*.; Koríye *n*.; Koroɓé *n*.; Koryaŋ *n*.; Kɔsɔŋ *n*.; Kúrúlè *n*.; Kʉsɛ́m *n*.; Kʉwám *n*.; Lemú *n*.; Loɓúɓúwo *n*.; Lɔcám *n*.; Lɔcáp<sup>a</sup> *n*.; Locíyo *n*.; Locóm *n*.; Locómín *n*.; Lɔɡyɛ́l *n*.; Loíkí *n*.; Lóʝérè *n*.; Lokapel *n*.; Lɔ̀kàts<sup>a</sup> *n*.; Lokauwa *n*.; Lokéɲériɓɔ *n*.; Lɔkíʝʉká *n*.; Lókírù *n*.; Lɔkɔl *n*.; Lɔ̀kʉ̀ɗà *n*.; Lɔkʉwám *n*.; Lokwaŋ *n*.; Lolém *n*.; Lɔmɛ́r *n*.; Lomoŋin *n*.; Lɔmɔ́y <sup>a</sup> *n*.; Lɔmʉ́ɲɛ́n *n*.; Lómúrìà *n*.; Lomutsú *n*.; Loɲá *n*.; Loɲáŋálem *n*.; Lɔ̀ɲàŋàsʉ̀wà *n*.; Loŋóle *n*.; Lòŋòlè *n*.; Loŋólépalɔ́r *n*.; Loŋólì *n*.; Lopeleméri *n*.; Lopéyók<sup>a</sup> *n*.; Lopíè *n*.; Lopúsór *n*.; Lópúwà *n*.; Lorukuɗe *n*.; Losíke *n*.; Losíroy<sup>a</sup> *n*.; Lotsul *n*.; Lɔ̀tʉ̀ɗɔ̀ *n*.; Lotuk<sup>a</sup> *n*.; Lotukéy<sup>a</sup> *n*.; Lotyaŋ *n*.; Lourien *n*.; Loyaŋorok<sup>a</sup> *n*.; Lúkà *n*.; Maarʉk<sup>a</sup> *n*.; Mamʉkíria *n*.; Máríkò *n*.; Matéò *n*.; Matsú *n*.; Moɗiŋ *n*.; Moɗó *n*.; Ŋiriko *n*.; Ŋoya *n*.; Nacapíò *n*.; Naɗóóɲ *n*.; Nàɗù *n*.; Nakíŋa *n*.; Nákírù *n*.; Nakɔŋ *n*.; Nakyéɲ *n*.; Namɔ́y <sup>a</sup> *n*.; Naŋetéɓ<sup>a</sup> *n*.; Náŋólì *n*.; Nápíyò *n*.; Napoliso *n*.; Narót<sup>a</sup> *n*.; Nátɔmɛ́ *n*.; Natsapúó *n*.; Natsíámu *n*.; Nàwà *n*.; Nayaón *n*.; Ɲáɓáts<sup>a</sup> *n*.; Ɲáɓoliɡúr *n*.; Ɲakalees *n*.; Ɲákamʉ *n*.; Ɲákáy<sup>a</sup> *n*.; Ɲálem *n*.; Ɲaŋasir *n*.; Ɲaŋorok<sup>a</sup> *n*.; Ɲékuɗuɗ<sup>a</sup> *n*.; Ɲéléle *n*.; Ɲɛlɛtsa *n*.; Ɲémuƙ<sup>a</sup> *n*.; Ɲɛpʉlɔ *n*.; Ɲétayoŋ *n*.; Ɲókoɗós *n*.; Ɲoŋoleɓók<sup>a</sup> *n*.; Ɲɔ́rɔ́cɔm *n*.; Océn *n*.; Ɔŋɔr *n*.; Pelén *n*.; Píipí *n*.; Pʉlʉkɔ́l *n*.; Rúfa *n*.; Saŋaɲ *n*.; Sɛʉsɛ́w<sup>a</sup> *n*.; Silóy<sup>a</sup> *n*.; Síré *n*.; Suɡur *n*.; Tsɨlá *n*.; Tekó *n*.; Timatéw<sup>a</sup> *n*.; Títò *n*.; Tɔ̀kɔ̀b <sup>a</sup> *n*.; Topér *n*.; Tówotó *n*.; Yakóɓò *n*.; Yarán *n*.; Yoánà *n*.
**name (place)** Árápííʝí *n*.; Ɓèlèkw<sup>a</sup> *n*.; Ɓets'oniicékíʝ<sup>a</sup> *n*.; Bɔ̀rɔ̀tsààk<sup>a</sup> *n*.; Buɗámóniicékíʝ<sup>a</sup> *n*.; Burukáy<sup>a</sup> *n*.; Caalíím *n*.; Ɗàsòƙ<sup>a</sup> *n*.; Ɗìɗèàw<sup>a</sup> *n*.; Ɗómòk<sup>a</sup> *n*.; Dímán *n*.; Gàràʝìàw<sup>a</sup> *n*.; Icékíʝ<sup>a</sup> *n*.; Ilúúkori *n*.; Irikakokor *n*.; Isókóìàƙw<sup>a</sup> *n*.; Íwá *n*.; Íwɔlɔ́*n*.; J'àòàw<sup>a</sup> *n*.; Kaaɓɔ́ŋ *n*.; Káákuma *n*.; Kaaláɓè *n*.; Kaehíƙɔ́ *n*.; Kaɨkɛm *n*.; Kaɨkɔ́ɓà *n*.; Kámíón *n*.; Kanarɔ́ *n*.; Kapalú *n*.; Kapísima *n*.; Kàsìlè *n*.; Kawalakɔ́l *n*.; Kiɓíc<sup>a</sup> *n*.; Koror *n*.; Kʉ́ràìàƙw<sup>a</sup> *n*.; Kùrùmò *n*.; Kwarikabubúík<sup>a</sup> *n*.; Lɛ̀rààƙw<sup>a</sup> *n*.; Lɔɓʉrák<sup>a</sup> *n*.; Locóto *n*.; Lóɗwàr *n*.; Lòìtà *n*.; Lokicókio *n*.; Lɔkíŋɔ́l *n*.; Lɔkɨtɛlɛ́ɛ́lɔɓ<sup>a</sup> *n*.; Lɔ́kɔ̀l *n*.; Lòkòrìkìpì *n*.; Lɔ́kʉ́rʉ́k <sup>a</sup> *n*.; Loƙúm *n*.; Lɔlɛ́líà *n*.; Lolítsíàƙw<sup>a</sup> *n*.; Lɔmálɛ́r *n*.; Lomataŋaáw<sup>a</sup> *n*.; Lɔpɛlɨpɛl *n*.; Loporukɔlɔ́ŋ *n*.; Lɔrɛŋ *n*.; Loriɓóɓó *n*.; Losíroíáw<sup>a</sup> *n*.; Losor *n*.; Lotíɲam *n*.;
Lotirém *n*.; Lɔtɔ́ƙíkààw<sup>a</sup> *n*.; Lɔtɔlɛ́r *n*.; Loúsúnà *n*.; Loyóro *n*.; Morícoro *n*.; Móróɗ<sup>a</sup> *n*.; Ŋʉrak<sup>a</sup> *n*.; Náápoŋo *n*.; Nacákʉ́nɛ̀t <sup>a</sup> *n*.; Nakalelé *n*.; Naɔyakíŋɔ́l *n*.; Natɔ́rɔ́kɔkítɔ́ *n*.; Nayapan *n*.; Ɲálámʉɲɛna *n*.; Òŋòrìàw<sup>a</sup> *n*.; Òŋòrìpàkw<sup>a</sup> *n*.; Píré *n*.; Rɔ́ƙɔ́dɛ̀ *n*.; Sèɡààw<sup>a</sup> *n*.; Siƙák<sup>e</sup> *n*.; Ts'aɗíáw<sup>a</sup> *n*.; Tsùtsùkààw<sup>a</sup> *n*.; Takaniƙʉlɛ́ *n*.; Tasapetíáw<sup>a</sup> *n*.; Teɓur *n*.; Tɔrɔŋɔ́ *n*.; Tulútúl *n*.; Wús *n*.
**name (river)** Cakalatɔ́m *n*.; Cerûb<sup>a</sup> *n*.; Ɗóɗò *n*.; Dìdìàk<sup>a</sup> *n*.; Dímánìàk<sup>a</sup> *n*.; Dɔ́dɔ̀f *n*.; Dúlél *n*.; Gɔrɨs *n*.; Íbotokok<sup>a</sup> *n*.; Iraf *n*.; Iryɔ́kɔ́ *n*.; Ísɛ́ *n*.; Iwam *n*.; Ƙolomúsábá *n*.; Kàƙòlò *n*.; Kàlɔ̀ʝɔ̀kɛ̀ɛ̀s *n*.; Kalɔtʉ́kɔ́ *n*.; Kaloturum *n*.; Kalouwan *n*.; Kámíónòàk<sup>a</sup> *n*.; Kanaɗáp<sup>a</sup> *n*.; Kanákɛ́rɛt<sup>a</sup> *n*.; Kaɲíkààl *n*.; Kàrɛ̀ŋà *n*.; Katoposiɲaŋ *n*.; Kátɔ́rɔ̀sà *n*.; Katsakól *n*.; Kátsápeto *n*.; Kàwàlɛ̀ɛ̀s *n*.; Kéékoŋa *n*.; Kerûb<sup>a</sup> *n*.; Kiɗorinamót<sup>a</sup> *n*.; Kɔ́cɔ́kɨɔ *n*.; Kumet<sup>a</sup> *n*.; Lɔɓɔsɔɔŋɔ́r *n*.; Lɔcɔ́ríàlɔ̀sìà *n*.; Lɔɗʉ́r *n*.; Lɔɨsíká *n*.; Lɔɨtánɨt<sup>a</sup> *n*.; Lòkààpèlòt<sup>a</sup> *n*.; Lɔkasaŋatɛ́ *n*.; Lòkìlè *n*.; Lokipáka *n*.; Lɔkɨtɔ́y <sup>a</sup> *n*.; Lɔkʉ́ma *n*.; Lóloy<sup>a</sup> *n*.; Lɔmaaníkɔ *n*.; Lɔmacarɨwárɛt<sup>a</sup> *n*.; Lómìl *n*.; Lɔŋása *n*.; Lɔɔsɔ́m *n*.; Loteteleít<sup>a</sup> *n*.; Lòtsòròɓò *n*.; Meletisabá *n*.; Mɔƙɔ́rík<sup>a</sup> *n*.; Mukulit<sup>a</sup> *n*.; Mʉtʉ́ nan *n*.; Nakamemeot<sup>a</sup> *n*.; Nakoɗíle *n*.; Nàkwàŋà *n*.; Namerí *n*.; Namétúròn *n*.; Namórú *n*.; Naŋóléɓok<sup>a</sup> *n*.; Napitiro *n*.; Natɔkɔ́ɔ́ŋɔr *n*.; Natsuƙúl *n*.; Natʉrʉkan *n*.; Nòf *n*.; Ɲerasabá *n*.; Ɲɔrɔbat<sup>a</sup> *n*.; Oŋorisabá *n*.; Oŋórîz *n*.; Óríɓò *n*.; Óríɓosabá *n*.; Pakósábà *n*.; Palú *n*.; Popá *n*.; Puɗápúɗ<sup>a</sup> *n*.; Sabaa Damán *n*.; Saloloŋ *n*.; Saŋar *n*.; Sɛkɛɗíáw<sup>a</sup> *n*.; Sɛkɛt<sup>a</sup> *n*.; Síɔ̀ɔ̀t <sup>a</sup> *n*.; Sɔ́ɡɛsabá *n*.; Tsakúdèɓò *n*.; Tamateeɓon *n*.; Tíríkɔ̀l *n*.; Tòòrwààk<sup>a</sup> *n*.; Turakwareekw<sup>a</sup> *n*.;
Tùrùmàrààk<sup>a</sup> *n*. **name a newborn** óésa édie imá<sup>e</sup> *v*. **name of honor** éda na moranâd<sup>e</sup> *n*. **nanny (goat)** rieŋwa *n*. **nape (of neck)** fètìfèt<sup>a</sup> *n*. **nape of neck (fatty)** nìtsìnìts<sup>a</sup> *n*. **Napore person** Kàrɛ̀ŋààm *n*.; Ŋíkátapʉ́ám *n*.; Nàpɔ̀rɛ̀ààm *n*.; tɔbɔŋɔ́ám *n*. **nappy** ets'íƙwâz *n*. **narrate** isíséés *v*. **narrative** emut<sup>a</sup> *n*.; ɲáɗís *n*. **narrator** emútíkààm *n*.; isíséésíàm *n*. **narrow** ɨɗíŋɔ́n *v*.; ɨrɨɗɔs *v*.; riɗímétòn *v*.; rɔ́ƙɔ́rɔƙánón *v*. **narrow (of an opening)** tɨɨts'ímɔ̀n *v*. **nasal bridge** sarɨsar *n*. **nasty** itútsón *v*.; ŋorótsánón *v*. **nation** kíʝ<sup>a</sup> *n*. **native** áméda kíʝá<sup>e</sup> *n*. **nature** kíʝ<sup>a</sup> *n*. **naughty** tarates *v*. **naughty (habitually)** taratiés *v*. **nauseated** iláƙízòn *v*.; talóón *v*. **navel** ƙɔ̀ɓ <sup>a</sup> *n*. **navel hair** ƙɔ̀ɓàsìts'<sup>a</sup> *n*. **near** ɦyɔtɔ́ɡɔ̀n *v*.; ɨtɔ́mɔ́n *v*. **near death** inunúmétòn *v*. **near maturity (of grain)** titímóonuƙot<sup>a</sup> *v*. **near to each another** ɦyɔtɔ́ɡɨmɔ́s *v*. **nearby** ɦyàtàk<sup>a</sup> *n*. **nearly ripe** bɔ́ŋɔ́n *v*. **neb** loɓôz *n*. **necessary** ɨtámáánón *v*. **neck** ɦyʉƙʉm *n*. **neckband** ɦyʉƙʉma ƙwázà<sup>e</sup> *n*. **neckbone** bɔkɔ́s *n*.; ɦyʉƙʉmʉ́ɔ́k <sup>a</sup> *n*.
**neckrest** kàràts<sup>a</sup> *n*. **neckring (metal)** ɦyʉƙʉmʉ́tsírím *n*. **necktie** ɲátáy<sup>a</sup> *n*. **necktie (money-keeping)** ɲɛƙɨl *n*. **necrophilist** tirésíama ts'óóniicé *n*. **necrotize** iɗéɗéŋés *v*. **nectar** ɨɔk<sup>a</sup> *n*. **need** bɛ́ɗɛ́s *v*. **needle** mʉtʉ *n*.; ɲɛ́sɨnɗán *n*. **needle (knitting)** ɲɛ́sɨlɨɓá *n*. **needle-thin** tɨwídɔ̀n *v*. **needle-thinly** tìw *ideo*. **neglect** balɛ́s *v*.; balɛtɛ́s *v*.; hakaikés *v*.; ɨlaʝíláʝɛ́s *v*. **neglect (property)** ɨsɔ́ɓɔ́lɛ́s *v*. **neglect oneself** balɛ́sá así *v*. **neglected** hakaikós *v*. **negligent** ɨtátsámánón *v*. **negotiate** ɗɔtsɛtɛ́sá tódà<sup>e</sup> *v*. **neighbor** ɨtɔ́mɔ́nìàm *n*.; ŋƙáƙínósíàm *n*.; narúétìàm *n*. **neighbor (agreer)** tsámʉ́nɔtɔ́síàm *n*. **neighbor (close)** ɦyɔtɔ́ɡɔ̀nìàm *n*. **neighbor (sharer)** tɔ̀mɔ̀rààm *n*. **neighbor each other** narúétinós *v*. **neighborhood** narúét<sup>a</sup> *n*. **neighbors (be)** ɨtɔ́mʉ́nɔ́s *v*. **Neotonia wightii** simísímàt<sup>a</sup> *n*. **nephew (her husband's sibling's son)** ntsínámúíìm *n*. **nephew (his brother's son)** ntsíím *n*. **nephew (his/her brother's son)** leatíím *n*. **nephew (his/her sister's son)** yeatíím *n*. **nephew (my brother's son)** ɛdéìm *n*.; ɲ́cììm *n*.
**nephew (my husband's sibling's son)** ɲ́cinamúíìm *n*. **nephew (my sister's son)** yeáìm *n*. **nephew (sororal)** momó *n*. **nephew (your brother's son)** biím *n*.; léóím *n*. **nephew (your husband's sibling's son)** binamúíìm *n*. **nephew (your sister's son)** yáóím *n*. **nephew-in-law (his/her child's spouse's brother)** ntsíɲótàìm *n*. **nephew-in-law (my child's spouse's brother)** ɲ́cìɲòtàìm *n*. **nephew-in-law (your child's spouse's brother)** biɲótáìm *n*. **nerve** ŋísɨl *n*. **nervous** rukurúkón *v*. **nest** ɡwáho *n*. **net (trap)** sáɡòsìm *n*. **net-trapping** sâɡw<sup>a</sup> *n*. **network (cellular)** ɲénétìwàk<sup>a</sup> *n*.; suɡur *n*. **never** ʝɨkî *adv*.; mʉ̀kà *adv*.; tsʉ̀tᶶ *adv*. **never-ending** rítsírɨtsánón *v*. **new** erútsón *v*. **new (of foliage)** ɨlíɓɔ́n *v*. **new (plant growth)** lìtɔ̀n *v*. **new thing** ŋípyà *n*. **news** emut<sup>a</sup> *n*.; emútík<sup>a</sup> *n*.; ɲéripót<sup>a</sup> *n*. **next** tɔtʉ́pɔ́n *v*. **next (be the)** mɨtɔna ɗíɛ́tûb<sup>a</sup> *v*. **next (time)** táá *adv*. **next to** ɨɓákɔ́n *v*.; ɨtɔ́mɔ́n *v*. **next to (move)** ɨɓákɔ́nʉƙɔt<sup>a</sup> *v*. **next to each other** ɨɓákínɔ́s *v*.; ɨtɔ́mʉ́nɔ́s *v*. **next year** kaɨnɔ na táà *n*.; kɛɨnats<sup>a</sup> *n*.
**night-walker** ɓɛƙɛ́síama mukú *n*. **nightjar** bóx *n*. **nighttime** mukú *n*. **nighty-night!** bubú *nurs*. **nimble** pɔɗɔ́dɔ̀n *v*. **nimbly** pɔ̀ɗ ɔ *ideo*. **nine** tude ńda kiɗi ts'aɡús *num*. **nine o'clock** ɲásáatɨkaa aɗátìk<sup>e</sup> *n*. **nineteen** toomíní ńda kiɗi túde ńda kiɗi ts'aɡús *n*. **ninety** toomínékwa túde ńda kiɗi ts'aɡús *n*. **nipple** ídòkàts<sup>a</sup> *n*. **nit(s)** ɨnak<sup>a</sup> *n*. **no** ńtóodó *interj*.; ńtóondó *interj*. **no-no!** kɔkɔ́*nurs*. **nocturnal emission** ɗír *n*. **nod** itéƙítéƙés *v*. **nod off** ɨlʉ́zɛ̀tɔ̀n *v*. **noise** nɔ̀s *n*. **noise (make a)** arútónuƙot<sup>a</sup> *v*. **noise (make)** arútón *v*. **noise (make, of a vehicle)** fútón *v*. **noisy** ilélémùòn *v*. **nomad** botibotosíám *n*. **nomadic** botibotos *v*. **nominate for office** wasɨtɛs *v*.; wasítɛ́sʉƙɔt<sup>a</sup> *v*. **nomination** was *n*. **nominee** wasɔ́ám *n*. **non-governmental organization** lɔʝɔkɔtáw<sup>a</sup> *n*.; toráƙádòs *n*. **nonchalant** faɗétón *v*. **nonsense** dɔ́bàtòd<sup>a</sup> *n*.; ɨɓááŋàsìtòd<sup>a</sup> *n*. **noon** ikáɡwaríìk<sup>e</sup> *n*.; ódoo bɨrɨr *n*. **normal (return to)** xɔ́dɔnʉƙɔt<sup>a</sup> *v*. **north** kɔ́ɔ́kwar<sup>ɔ</sup> *n*.; nɔ́ɔ́kwar<sup>ɔ</sup> *n*.
obvious
**North America** Amérìkà *n*.; Ɓets'oniicékíʝ<sup>a</sup> *n*. **northerly direction** ɡwárixan *dem*. **northerner** kɔ́ɔ́kwarɔ́ám *n*. **northward** kɔ́ɔ́kwar<sup>ɔ</sup> *n*. **nose** aƙat<sup>a</sup> *n*. **nosebleed (have a)** ƙòlòn *v*. **nosebone** aƙatíɔ́k <sup>a</sup> *n*. **nostril** aƙat<sup>a</sup> *n*.; aƙatíékw<sup>a</sup> *n*. **not** eʝá *adv*.; máa *adv*.; mòò *adv*.; ńtá *adv*. **not be** beníón *v*.; bɛnɔ́ɔ́n *v*. **not be (somewhere)** bɨrɔ́ɔ́n *v*. **not enough** ɡàɗɔ̀n *v*. **not full** kíón *v*. **not make sense** ɗɛ̀ƙwɔ̀n *v*. **not sit well** ts'ábès *v*. **not there** bɨrɔ́ɔ́n *v*. **not yet be** sárón *v*. **notch** ɨsɛɓɛs *v*.; ɨtɛɓɛs *v*. **notch (ears)** tɔmʉɲɛs *v*.; topones *v*. **notch (jugular)** tɔ̀k <sup>a</sup> *n*. **notched** ɨsɛɓɔs *v*. **note (monetary)** kaúdzokabáɗ<sup>a</sup> *n*.; ɲónót<sup>a</sup> *n*. **notice** ewanes *v*.; ewanetés *v*. **novelty** ŋípyà *n*. **November** Kawés *n*.; Loipo *n*. **now** nápáka na *adv*.; názɛ̀ƙwà *n*.; ts'ɔ̀ɔ̀ *adv*. **now now!** tíɔ̀ *interj*.; tíɔ ʝɔ́ɔ̀ *interj*. **nowadays** ódowicíkó nì *n*.; ódowicíkó nì kɔ̀nà *n*. **nude** ilérón *v*.; lemúánètòn *v*.; leŋúrúmòn *v*.; tuɗúsúmòn *v*. **nugget** ɡwas *n*. **numb (of body parts)** ɨmʉ́nʉ́kʉ̀kʉ̀ɔ̀n *v*.
**number** ɨmaarɛ́s *v*.; ɲánamɓá *n*. **number (large)** tʉ̀mɛ̀ɛ̀ *n*. **numbered** ɨmaarɔ́s *v*. **numbness (of body)** ɲɛ́mʉnʉkʉ́ *n*. **numerous** miƙídòn *v*. **numerously** mìƙⁱ *ideo*. **nun** ɓíkìrà *n*. **nurse** ɗakɨtárìàm *n*. **nurse (the sick)** maitetés *v*. **nurture (newborn)** tɔʉ́rʉ́mɔ̀n *v*. **Nyangatom people** Ŋíóyatom *n*. **Nyangatom person** Ŋísaakɔ́lìàm *n*. **Nyang'ia language** Ŋíɲaŋíyátòd<sup>a</sup> *n*. **Nyang'ia person** Ŋíɲaŋíyáàm *n*. **nylon** náìlòn *n*. **obese** iɓutúɓútòn *v*. **obey** nesíbes *v*. **obey (habitually)** nesíbiés *v*. **object** kɔ́rɔ́ɓâd<sup>a</sup> *n*. **object (large)** òrìkìrìk<sup>a</sup> *n*. **object (surveiled)** rɔtam *n*. **objects** kúrúɓâd<sup>a</sup> *n*. **obligation** amʉ́ts<sup>a</sup> *n*. **obliquely** ŋabér<sup>o</sup> *n*. **obliterate** ɨƙɔmɛs *v*. **oblong** sʉlʉ́tʉ́mɔ̀n *v*. **obscure** ɨɲʉ́ɲʉ́ánón *v*. **observatory** dìyw<sup>a</sup> *n*. **observe** tɨrɨfɛs *v*.; tɨrɨfɛtɛ́s *v*.; tɨtɨmɛs *v*. **observe (ceremony)** ɨnʉmʉ́nʉ́mɛ́s *v*.; ɨnʉ́nʉ́mɛ́s *v*. **obsessed with girls** iɲéráánón *v*. **obsessed with men** ɨɲɔ́táánón *v*. **obstinate** ɨɗíkílɔ̀n *v*. **obstruct** ɗɨnɛ́s *v*.; itítírés *v*. **obstructed** ɗɨnɔ́s *v*. **obvious** takánón *v*.
#### obviously
**obviously** tsábò *adv*. **occupy** ɔ́bɛ̀s *v*. **occur** ikásíìmètòn *v*.; itíyáìmètòn *v*. **ocean** ɲánam *n*. **ocher** ŋɔr *n*. **Ochna species** ɲéleɓuléɓu *n*. **October** Lɔlɔɓáy<sup>a</sup> *n*.; Terés *n*. **odd jobs** ɲɛ́lɛ́ʝɨlɛʝ<sup>a</sup> *n*. **odd jobs (do)** ɨlɛʝílɛ́ʝɛ́s *v*. **odds with each other (be at)** ɗúlúnós *v*. **odor** ɔn *n*.; ɔnɛd<sup>a</sup> *n*. **Oenanthe palustris** ŋálómóyá *n*. **of age** zòòn *v*. **off (kill)** ɨɗɛɛs *v*. **off (rotten)** masánón *v*. **off limits** itáléánón *v*.; itálóós *v*. **offend** risés *v*. **offender** ɲɔ́mɔkɔsáàm *n*.; tɛ́ŋɛ́rìàm *n*. **offense** tɛ́ŋɛ́r *n*. **offering (animal)** ɲapʉɔ́t <sup>a</sup> *n*. **offerings** meetésíicík<sup>a</sup> *n*. **office** ɲápís *n*. **officer** ŋurutiesíama tódà<sup>e</sup> *n*. **official** ŋurutiesíama mɛná<sup>ɛ</sup> *n*. **official (government)** ámázeáma ɲápukaní *n*.; túbesiama ɲápukání *n*. **offload a load** ɓuƙetésá botá<sup>e</sup> *v*. **offspring** kwats<sup>a</sup> *n*. **oh my God!** Ɲakuʝ<sup>a</sup> *interj*. **oh my goodness** hóítá kwí *interj*. **oh my wordǃ** hóítá kwí *interj*. **oh!** ábaŋ *interj*.; té *interj*.; yáŋ *interj*. **oh, I see** nés *adv*. **oh, you mean** nés *adv*. **oil** ceím *n*.
**oil (seed)** útɔ̀ *n*. **oily** kùx *ideo*. **okay!** maráŋ *interj*. **okra** ɲɔlɔlɔt<sup>a</sup> *n*. **old** dúnésòn *v*.; kɔ̀wɔ̀n *v*.; zòòn *v*. **old (of many)** dunaakón *v*. **old man** ʝákám *n*. **old men** ʝák<sup>a</sup> *n*. **old people** dunaakóniik<sup>a</sup> *n*. **old person** dúnésìàm *n*. **old woman** dúnéìm *n*. **old-fashioned** kɔ̀wɔ̀n *v*. **oldness** zeís *n*. **Olea europaea (africana)** dèmìyw<sup>a</sup> *n*. **Olinia rochetiana** ɓets'akáw<sup>a</sup> *n*. **omasum** ɲémékweɲ *n*. **on** ɡwaríédek<sup>e</sup> *n*. **on all fours** tíɡàkòn *v*. **on empty stomach** kùk<sup>u</sup> *ideo*. **on foot** dɛ̀ìk<sup>ɔ</sup> *n*. **on that day** ódeedóó *n*. **on the feet** dɛ̀ìkà<sup>ɔ</sup> *n*. **on the legs** dɛ̀ìkà<sup>ɔ</sup> *n*. **on the move** ɓɛƙɛsɔs *v*. **on the way** múkò *n*. **on top** ɡwaríédek<sup>e</sup> *n*. **once** kɔn<sup>ɔ</sup> *num*. **once and for all** kɔ́nítɨák<sup>e</sup> *pro*. **once upon a time** kónító ódòwì *n*. **one** kɔ̀n *num*.; kɔ̀nɔ̀n *v*. **one (make into)** kɔnítɛ́sʉƙɔt<sup>a</sup> *v*. **one at a time** kóníátìk<sup>e</sup> *v*.; kóníón *v*. **one day** kónító ódòwì *n*.; na tsóíta kɔní *n*. **one o'clock** ɲásáatɨkaa tudátie ńda kiɗi léɓèts<sup>e</sup> *n*.
**one time** kɔn<sup>ɔ</sup> *num*. **one-by-one** kóníátìk<sup>e</sup> *v*.; kóníón *v*. **onion** ɲékuduŋƙúru *n*. **only** ɛɗá *adv*. **onward** wàxìk<sup>ɛ</sup> *n*. **ooze** tɔfɔ́ɗɔ́n *v*. **opaque (thick)** tìnòn *v*. **open** bɔrɔ́ɔ́n *v*.; bótsón *v*.; fotólón *v*.; ŋáɲámòn *v*.; ŋáɲɛ́s *v*.; ŋáɲɔ́s *v*.; ŋawíɔ́n *v*. **open (completely)** ùwòò *ideo*. **open fire on** ɗamatés *v*.; tɔƙʉmʉ́ƙʉ́mɛ́s *v*. **open up** tɛlɛɛs *v*. **open up (that way)** ŋáɲɛ́sʉƙɔt<sup>a</sup> *v*. **open up this way** ŋaɲɛtɛ́s *v*. **open wide** hádoletés *v*. **open-topped** lɛɓɛ́ɲɛ́mɔ̀n *v*. **opening** ak<sup>a</sup> *n*.; akɛd<sup>a</sup> *n*. **opening (center)** wɛ̀lèèkw<sup>a</sup> *n*. **opening (small)** wɛ̀l *n*. **operate on** hoés *v*.; hoetés *v*. **operating room** hoesího *n*. **operation (conduct an)** iríɓéés *v*. **operation (military)** ɲériɓá *n*. **operator** ŋíɗɛrɛpáìàm *n*. **oppose** ɨƙaíƙɛ́ɛ́s *v*.; ɨƙáƙɛ́ɛ́s *v*.; ɨƙáƙɛ́ɛtɛ́s *v*. **opposite side** jíjè *n*. **or** kèɗè *coordconn*.; kòrì *coordconn*. **OR (operating room)** hoesího *n*. **oracle** ɲakuʝíícíkáàm *n*. **orange** ɲámucúŋ́ƙà *n*. **orange drink** ɲɛ́kwɨɲcá *n*. **orbit** ɨríŋɔ́n *v*. **orchestrate** itukanitetés *v*.
**order** ɨɗɔ́bɛ̀s *v*.; ɨɗɔ́bɛtɛ́s *v*.; ɨnábɛs *v*.; ɨnábɛtɛ́s *v*.; itíbès *v*.; itíbesúƙot<sup>a</sup> *v*.; ɨtsɨkɛs *v*. **order out** taŋasɛs *v*. **orders (for marching)** tàŋàs *n*. **organ fat** sábà *n*. **organization** máɗíŋ *n*.; ɲéɡurúf *n*. **organize** ɨɗɨmɛ́s *v*.; ɨɗɨmɛ́sʉ́ƙɔt<sup>a</sup> *v*.; iɗimiés *v*.; iɗimiesúƙot<sup>a</sup> *v*.; ipáŋƙeés *v*.; itukanitetés *v*. **organized** ɨɗɨmɔ́s *v*. **organizer** iɗimiesíàm *n*.; ìʉ̀ɗààm *n*. **organs (ritual)** ɲorópúò *n*. **orgasm (have an)** ɨrákɛ́sʉƙɔta así *v*. **orgasmic (become)** ɛfɔnʉƙɔt<sup>a</sup> *v*. **oribi** kɔtɔ́r *n*. **origin** itsyákétònìàw<sup>a</sup> *n*. **Ormocarpum trichocarpum** mozokoɗ<sup>a</sup> *n*. **orphan** bɔnán *n*. **orphaned** ikókíánón *v*. **orphanhood** bɔnánés *n*. **oryx (Beisa)** tsarʉ́k <sup>a</sup> *n*. **oryx (male)** tènùs *n*. **oryx horn** tsarʉ́kʉ́ɛ̀b <sup>a</sup> *n*. **os temporale** bòsìɔ̀k <sup>a</sup> *n*. **oscillate** iŋolíŋólés *v*. **ostrich** lèwèɲ *n*. **Osyris abyssinica** tsereɗeɗí *n*. **other(s)** kiɗíása *pro*. **otherwise** náa táà *subordconn*. **ouch!** aaii *interj*.; áí *interj*. **ought** ɨtámáánón *v*. **our (exclusive)** ŋɡw<sup>a</sup> *pro*. **our (inclusive)** ɲjín *pro*. **ours (exclusive)** ŋɡóɛ́n *pro*. **ours (inclusive)** ɲjíníɛ̀n *pro*. **ourselves (exclusive)** ŋɡónébitín *n*.
**ourselves (inclusive)** ɲjínínebitín *n*. **oust** ɨlɔ́líɛ́s *v*. **out (finished)** tɛ́zɛ̀tɔ̀n *v*. **out (of stars)** ʝʉ̀ɔ̀n *v*. **out of sight** kúbòn *v*.; lwàŋ *ideo*. **out of work** ɨlwárɔ́na teréɡù *v*. **out-of-joint (get)** ɓuumón *v*. **outbuilding** ɲésitó *n*. **outcrop** rikírík<sup>a</sup> *n*. **outcropping** kúc<sup>a</sup> *n*. **outcry** werets<sup>a</sup> *n*. **outdo** ɨlɔɛs *v*.; ɨlɔɛtɛ́s *v*.; ipíyéésuƙot<sup>a</sup> *v*. **outer part** kanɛd<sup>a</sup> *n*. **outer stomach** kɨrarap<sup>a</sup> *n*. **outfit** ɲéyúnìfòm *n*. **outhouse** ets'íhò *n*.; ɲótsorón *n*. **outlaw** mɛnáám *n*. **outmoded** kɔ̀wɔ̀n *v*. **outside** biy<sup>a</sup> *n*.; biyáxán *n*.; kànɛ̀dɛ̀k <sup>ɛ</sup> *n*. **outsider** ámá na biyá<sup>e</sup> *n*.; ɦyɔ̀àm *n*.; kíʝíkààm *n*.; ŋíɓúkúìàm *n*.; ɲeɓúkúit<sup>a</sup> *n*. **oval** semélémòn *v*. **ovary** ɓiɓáhò *n*. **over here** nɔ́ɔ́na *dem*. **over there** kéda ke *dem*.; kéíta ke *dem*.; kíxána ke *n*.; kɔ́ɔ́kɛ *dem*. **overabundance (here)** níbàdà *n*. **overabundance (there)** kíbàdà *n*. **overall (skin)** kɔ́lɔ́ts<sup>a</sup> *n*. **overcast** kùpòn *v*. **overcast (become)** ɡobétón *v*. **overcast weather** kùpààƙw<sup>a</sup> *n*. **overcome** iloimétòn *v*.; ipíyéésuƙot<sup>a</sup> *v*. **overcrowd** ƙídzɛ̀s *v*. **overdo** taɗɛɛs *v*.
**overeat** iwótsóòn *v*.; ɲaɗésá ŋƙáƙá<sup>e</sup> *v*.; wɔ̀ɔ̀n *v*. **overflow** ɨlápɛ́tɔ̀n *v*. **overfull** ɨsʉ́wɔ́ɔ̀n *v*. **overgrown** sɔɓɔ́lɔ́mɔ̀n *v*.; tsèkòn *v*. **overlook** hakaikés *v*.; ɨlaʝíláʝɛ́s *v*. **overlooked** hakaikós *v*. **overpower** itikes *v*. **override** ilotsesa mɛná<sup>ɛ</sup> *v*. **overrun** ƙídzɛ̀s *v*. **oversee** ɨrɨtsɛ́s *v*. **oversleep** towúryánòn *v*. **overtake** ilaŋés *v*.; ɨsʉkɛs *v*.; rítsɛ́s *v*.; sʉ́kɛ́s *v*. **overthrown** rúmánònà kàràtsʉ̀ *v*. **overturn** bukures *v*.; bukúrésuƙot<sup>a</sup> *v*.; bukúrésuƙota así *v*.; buƙusítésuƙot<sup>a</sup> *v*.; iɓéléés *v*.; iɓéléetés *v*.; iɓéléìmètòn *v*.; iɓélúkáìmètòn *v*.; iɓélúkéés *v*.; pukés *v*.; puketés *v*. **overturn away** pukésúƙot<sup>a</sup> *v*. **overweight** iɓutúɓútòn *v*. **overwhelm** kurés *v*.; kurésúƙot<sup>a</sup> *v*. **overwhelming (become)** kurósúƙot<sup>a</sup> *v*. **ovum** ɓìɓ<sup>a</sup> *n*. **ow!** aaii *interj*.; áí *interj*. **oware (game)** ɲékilelés *n*. **owl (African scops-)** ƙórór *n*. **owl (eagle-)** lófúk<sup>a</sup> *n*. **owner** ámêd<sup>a</sup> *n*. **owner (of land)** áméda kíʝá<sup>e</sup> *n*. **ox** râɡw<sup>a</sup> *n*. **ox name** ráɡòèd<sup>a</sup> *n*. **ox plow** ɲɛ́mɛlɛkʉ́à nà ɦyɔ̀ɔ̀ <sup>ɛ</sup> *n*. **ox song** ráɡòdìkw<sup>a</sup> *n*. **oxpecker (red-billed)** dzàr *n*.
**Ozoroa insignis (reticulata)** mókol *n*. **pace off** ɨmaarɛ́sá dɛ̀ìkà<sup>ɛ</sup> *v*. **Pachycarpus schweinfurthii** lóúpè *n*. **pacify** ɨkanɛ́s *v*.; ɨkaníkánɛ́s *v*. **pack** ɨɗɨlɛs *v*.; ɨsɨkɛs *v*. **pack down** ɨɗɛŋɛs *v*. **pack down personally** ɨɗɛŋíɗɛ́ŋɛ́s *v*. **package** méy<sup>a</sup> *n*. **packed down** tɔrɔ́dɔ̀n *v*. **packed down (become)** iɗéŋímètòn *v*. **packet** méy<sup>a</sup> *n*. **paddle** iƙures *v*. **paddle (spank)** ipíkéés *v*. **padlock** ɲékifúl *n*. **pagan** ŋíkafírìàm *n*. **pail** ɲáɓákɛ̀t <sup>a</sup> *n*.; ɲɛ́ɓákɛ̀t <sup>a</sup> *n*. **pail (metal)** ɲépeelí *n*. **pain (cause sharp)** ɨrɛɓírɛ́ɓɛ́s *v*. **painful** dódòn *v*. **painfully** wìl *ideo*.; wìlìwìl *ideo*. **paint** ŋɔrɨtɛtɛ́s *v*.; ɲáraŋɡí *n*.; tsáŋés *v*. **paintbrush** tsɨtsín *n*. **painted** tsáŋós *v*. **palate** aƙár *n*. **palate (cleft)** akáts'ɛ́a na pakós *n*. **palm (African fan)** ŋíɗʉkan *n*. **palm (African wild date)** lɔ̀kàtàt<sup>a</sup> *n*. **palm (Borassus)** ŋíɗʉkan *n*. **palm (of hand)** kwɛtááƙw<sup>a</sup> *n*. **palm tree species** ɲétenɗé *n*. **palpebra** ɡɔrɔ́x *n*. **palpitate** dìkwòn *v*. **pan** dóm *n*.; ɨlaƙɛs *v*.; ɨlaƙíláƙɛ́s *v*.; ɨláláƙɛ́s *v*.; ɲákaláát<sup>a</sup> *n*.
**pan (metal)** ɲásipiryá *n*.; ɲésipiriyá *n*.
**pan (small)** dómáìm *n*. **pan bottom** dómóɔ̀z *n*. **pancreas** lópey<sup>a</sup> *n*. **pandemonium** ɲɔ́ŋɔtsán *n*. **pandemonium (go into)** doʝánónuƙot<sup>a</sup> *v*.; lɔŋɔanónuƙot<sup>a</sup> *v*. **panel (solar)** ɲɔ́sɔ́la *n*. **panga** ɲápaŋká *n*. **pangolin** mɛkɛmɛkán *n*. **panic** doʝánónuƙot<sup>a</sup> *v*.; lɔŋɔanónuƙot<sup>a</sup> *v*.; ɲɔ́ŋɔtsán *n*. **Panicum maximum** òŋòrìkù *n*. **panties (pair of)** ɲekúrúm *n*. **pants (pair of)** ɲétorós *n*.; ɲótorós *n*. **papaya** ɲápaɨpáy<sup>a</sup> *n*. **paper** kàbàɗ<sup>a</sup> *n*.; ɲákaratás *n*. **paper (file)** ɲɛ́páìl *n*. **Pappea capensis** dzôɡ<sup>a</sup> *n*. **parable** taɗápítotós *n*. **parade** ɲéperét<sup>a</sup> *n*.; tɛtsɛ́sá ɲéperétì *v*. **parade about** inésóòn *v*. **paraffin** ceím *n*. **paralyzed (from fear)** dodimórón *v*. **parasite** ƙʉts'<sup>a</sup> *n*. **parasitic plant** lɛ̂z *n*. **parasitic plant species** ɲáɓús *n*.; tìlàlɛ̀z *n*. **parcel** taɲáléés *v*. **parcel out** ɨɲíɲínɛ́s *v*.; taɲáléetés *v*. **parch** mɔsɔnʉƙɔt<sup>a</sup> *v*. **parched** mɔ̀sɔ̀n *v*.; paupáwón *v*. **parched (lightly)** mɔsímɔ́sɔ̀n *v*. **pardon** iƙenes *v*.; óɡoés *v*. **pare** ɨlɨmɛs *v*.; ɨpɛlɛs *v*. **pare down** ɨlɨmɛtɛ́s *v*. **parent** ƙwaatetésíàm *n*.
**parent-in-law** emetá *n*. **parent-in-law (his)** ntsíémetá *n*. **parent-in-law (my, of men)** ɲ́ciemetá *n*. **parent-in-law (your, of men)** biemetá *n*. **parenthesia (be in)** isálílòn *v*. **paresthesia (go into)** isálílètòn *v*. **parish** ɲápárìx *n*. **parish chief** ámázeáma ɲépárìxì *n*.; ɲékúŋut<sup>a</sup> *n*. **parish security officer** tsítsá na kwáts<sup>a</sup> *n*. **parking place** ɨnábɛ̀sìàw<sup>a</sup> *n*. **parrot** loki *n*. **parry** ɨɓatɛs *v*. **parry repeatedly** ɨɓatíɓátɛ́s *v*. **part** bácík<sup>a</sup> *n*.; xɔnɔ́ɔ́kɔn *n*. **part (back)** ʝírɛ̂d <sup>a</sup> *n*. **part (body)** ɲekiner *n*. **part (inner)** áƙwɛ̂d <sup>a</sup> *n*. **part (middle)** bakútsêd<sup>a</sup> *n*. **part (straight, middle)** ɡòɡòròʝ<sup>a</sup> *n*. **part (top)** ɡwaríêd<sup>a</sup> *n*.; iked<sup>a</sup> *n*. **part ways** terémétòn *v*.; terémón *v*.; terémónuƙot<sup>a</sup> *v*. **participate** ɓuƙonuƙot<sup>a</sup> *v*. **particle** kiɗoɗots<sup>a</sup> *n*.; símíɗiɗí *n*. **partition** naƙʉ́lɛ́*n*. **party** iyóómètòn *v*.; ɲápáti *n*. **pass** ɨkɔɓɛs *v*.; ɨlámɔ́n *v*.; ilaŋés *v*.; ilúɲón *v*.; ilúɲónuƙot<sup>a</sup> *v*. **pass (a test)** ɡórés *v*. **pass (time)** dzuƙés *v*. **pass a law** eɡésá ɨtsɨkɛsí *v*. **pass along here** ɨkɔɓɛtɛ́s *v*. **pass along there** ɨkɔ́ɓɛ́sʉƙɔt<sup>a</sup> *v*. **pass away** bitsétón *v*.
**pass by** ɓʉ̀nɔ̀n *v*.; ilúɲón *v*.; ilúɲónuƙot<sup>a</sup> *v*. **pass by going** ɓʉnɔnʉƙɔt<sup>a</sup> *v*. **pass gas** fenétón *v*. **pass here via** ɨɛ́bɛtɛ́s *v*. **pass off** iʝokes *v*.; iʝókésuƙot<sup>a</sup> *v*. **pass on** iʝokes *v*.; iʝókésuƙot<sup>a</sup> *v*. **pass on here** ɨkɔɓɛtɛ́s *v*. **pass on problems** iʝokesa mɛná<sup>ɛ</sup> *v*. **pass on there** ɨkɔ́ɓɛ́sʉƙɔt<sup>a</sup> *v*. **pass out** rèŋòn *v*. **pass over** ɡórés *v*.; íɡorés *v*.; íɡorésúƙot<sup>a</sup> *v*. **pass over a spear** ɡóriesá ɓɨsá<sup>ɛ</sup> *v*. **pass over repeatedly** ɡóriés *v*.; íɡoriés *v*. **pass round to each other** ɨkɔ́ɓínɔ́s *v*. **pass through** ɓuƙonuƙot<sup>a</sup> *v*.; pʉtʉ́mɔ́n *v*. **pass time** ɨtɛ́mɔ́ɔ̀n *v*. **pass via** ɨɛ́bɛ̀s *v*.; ɨɛ́bɛsʉƙɔt<sup>a</sup> *v*. **passageway** wɛ̀l *n*. **passageway (center)** wɛ̀lèèkw<sup>a</sup> *n*. **passport** ɓɛƙɛ́síkabáɗ<sup>a</sup> *n*.; kabaɗa na ɓɛƙɛ́sí *n*. **past (distant)** tsò *adv*. **paste** ilies *v*.; iliílíés *v*. **pasted** iliílíós *v*.; ilios *v*. **pastor** pásìtà *n*. **pasture** wà *n*.; waitetés *v*. **pat** iturútúrés *v*. **pat down** ɨnatsínátsɛ́s *v*.; tárábes *v*.; tárábiés *v*. **patch** rátsɛ́s *v*.; taɗapes *v*. **patch (bare)** ɲapatsole *n*. **patch (hard)** ɲapáyál *n*. **patch of cleared forest** tsɛ̀f *n*. **patch of grass** xʉram *n*. **patch repeatedly** rátsiés *v*. **patch up** taɗapetés *v*.
**patched** komolánón *v*.; koríánètòn *v*.; tábàsànètòn *v*.; taɗapos *v*. **pate** ikáɡwarí *n*. **patella** bùrùkùts<sup>a</sup> *n*. **path** muce *n*. **pathetic** pás *n*. **patheticness** pásìnànès *n*. **patients** mayaakóniik<sup>a</sup> *n*. **patriclan** àsàk<sup>a</sup> *n*.; ɔ́dɔ̀k <sup>a</sup> *n*. **paunch** ɡwàʝ<sup>a</sup> *n*. **paunchy** heɓúlúmòn *v*. **pause** ɨmɔ́mɛ́tɔ̀n *v*.; isíƙóòn *v*.; mɔ́mɛ́tɔ̀n *v*. **pawpaw** ɲápaɨpáy<sup>a</sup> *n*. **pay** taatses *v*. **pay (tax)** ƙúdès *v*. **pay a fine unjustly** taatsesa káwí *v*. **pay brideprice** buƙés *v*.; buƙetés *v*. **pay fine (for impregnation)** ɨtsʉlɛs *v*. **pay haphazardly** taatsesá bùɗàmàk<sup>e</sup> *v*. **pay in vain** taatsesa tsam *v*. **pay off** ɨlʉŋʉ́lʉ́ŋɛ́s *v*.; taatsésuƙot<sup>a</sup> *v*. **pay out brideprice** buƙésúƙot<sup>a</sup> *v*. **pay tax** taatsesa ɲéutsúrùⁱ *v*. **pay toward** tɔ́ƙɛ́s *v*. **payback (get)** ɲaŋésúƙot<sup>a</sup> *v*. **payment** tààts<sup>a</sup> *n*. **payment slip** taatsakabáɗ<sup>a</sup> *n*. **peace** ŋíkísila *n*.; ɲɛkɨsɨl *n*. **peace (make)** ɨsílítɛ́sʉƙɔt<sup>a</sup> *v*. **peaceful** ɨsílɔ́n *v*.; tisílón *v*. **peaceful (become)** ɨsílɔ́nʉƙɔt<sup>a</sup> *v*. **peaceful person** ɨsílɔ́nìàm *n*. **peak** kwaráɡwarí *n*. **peanut(s)** ɲépulé *n*.; taráɗá *n*.
**pebble** ɡwas *n*. **pec** làf *n*. **peck** itoɗítóɗés *v*.; ɨtɔ́tɔ́ŋɛ́s *v*. **pectoral muscle** làf *n*. **pectus** bakuts<sup>a</sup> *n*. **pedal a bicycle** takwésá ɲamɨɨlɨí *v*. **peddler** ŋímutsurúsìàm *n*. **peddler (being a)** ŋímutsurúsìnànès *n*. **pedestrian** ɓɛƙɛ́síàm *n*. **pee** kʉtsáƙón *v*.; kwats<sup>a</sup> *n*. **pee-pee** kwàà *nurs*. **peek at** ikórímés *v*.; ilóíkés *v*. **peek out** ts'ʉ̀fɔ̀n *v*. **peek through** tɛkɛɲɛs *v*. **peek through repeatedly** tekeɲiés *v*. **peel** ɨpɛlɛs *v*.; ɔmɔ́x *n*.; poxés *v*. **peel off** ɨwalɛtɛ́s *v*.; moxés *v*.; poxésúƙot<sup>a</sup> *v*. **peel with teeth** ɨsɨmɛs *v*. **peelable food** ɨsɨmam *n*. **peeling** ɨpɛlɛtam *n*. **peep at** ikórímés *v*.; ilóíkés *v*. **peep out** ts'ʉ̀fɔ̀n *v*. **peer at** tɨrɨfɛs *v*.; tɨrɨfɛtɛ́s *v*. **peer at over** ɨrɨnɛs *v*. **peer through** tɛkɛɲɛs *v*. **peer through repeatedly** tekeɲiés *v*. **peg** ɡóɡès *v*.; kìnòròt<sup>a</sup> *n*. **Pellaea adiantoidea** ts'aɗícɛ́mɛ̀r *n*. **pelt** ídʉrɛ́s *v*.; ínósìts'<sup>a</sup> *n*.; ts'ɛ̀ *n*. **pelvis** róróìɔ̀k <sup>a</sup> *n*. **pen** ɲákalám *n*. **penalty (financial)** ɲáfaín *n*. **penance (Catholic)** penitésìyà *n*. **pencil** ɲɛ́pínísìl *n*.
**penetrate (area)** utés *v*.; utésúƙot<sup>a</sup> *v*. **penile shaft** ɲɛ́sɛɛɓɔ́*n*. **penis** kwan *n*. **penis hole** kwaníékw<sup>a</sup> *n*. **penitentiary** zíkɛ́sìàw<sup>a</sup> *n*. **penny** ŋáɓɔ́ɔla *n*. **Pentarrhinum insipidum** urém *n*. **Pentecost** pɛntɛkɔ́stɛ̀ *n*. **Pentecostal** Ŋímorokóléìàm *n*. **people** ròɓ<sup>a</sup> *n*. **people (tribe)** dìyw<sup>a</sup> *n*.; ɲákaɓɨlá *n*. **people!** òɓà *interj*.; ròɓà *interj*. **pepper** ɲépilipíli *n*. **pepper (red)** ɲákamʉlára *n*. **perceive** enés *v*. **perch** itsélélèòn *v*. **perfect** xɔ́dɔ̀n *v*.; xɔtánón *v*. **perforate** ɓɛkɛ́s *v*.; ɓɛkɛtɛ́s *v*.; húbutés *v*.; pulés *v*.; ruɗés *v*. **perforate (with a tool)** ɡóɡès *v*. **perforate noisily** rɔɗɛ́s *v*. **perforate repeatedly** pulutiés *v*. **perform a miracle** ikúʝíánón *v*. **perfume (natural)** ɓʉ́ɓʉ́s *n*. **perhaps** ƙámá kiɗíé *v*.; ndóó ɦyè *n*. **peril** ɡaánàs *n*. **perilous** ɡaanón *v*.; ɨpáríŋánón *v*. **perineal muscle** ɲalamatsar *n*.; ɲekiɗoŋit<sup>a</sup> *n*. **peripheral vision** ɗoɗékw<sup>a</sup> *n*. **perish** bitsétón *v*.; ɨríɗɛ́tɔ̀n *v*. **perjure** isuɗes *v*.; isuɗetés *v*. **perjury** ɲɔ́pɔkɔca *n*. **permanently** kìŋ *ideo*. **permit** talakes *v*.
**perplexed** iɓíléròn *v*.; ɨcɔ́ŋáimetona iká<sup>e</sup> *v*. **Persea americana** ɲóvakáɗò *n*. **persevere** ɨɗaŋíɗáŋɔ̀n *v*.; ɨmʉ́kɔ́ɔ̀n *v*. **persist** ɨɗaŋíɗáŋɔ̀n *v*.; taɗáŋón *v*. **person** ám *n*. **person (indigenous)** áméda kíʝá<sup>e</sup> *n*. **person (surveiled)** rɔtam *n*. **person in authority** topéɗésuƙotíám *n*. **person who prays** wáánààm *n*. **personal item** ámákɔrɔ́ɓâd<sup>a</sup> *n*. **personal property** ámákɔrɔ́ɓâd<sup>a</sup> *n*. **personhood** ámánànès *n*. **perspiration** kirot<sup>a</sup> *n*. **perspire** kirotánón *v*. **persuade** sʉ́bɛ̀s *v*.; sʉ́bɛsʉƙɔt<sup>a</sup> *v*. **perturbed** walɨwálɔ́n *v*. **pervert** ɲárásíám *n*. **pester** ilúlúés *v*. **pestilence** koɗó *n*. **pestle** àʝ<sup>a</sup> *n*.; iwótsídàkw<sup>a</sup> *n*.; kuɲuk<sup>a</sup> *n*. **pet (have as a)** totores *v*. **Peter** Pétèrò *n*. **Peter (biblical)** Pétèrò *n*. **petite** tsaʉ́ɗímɔ̀n *v*. **petrol** ceím *n*.; ɲépetorón *n*. **phallus** kwan *n*. **phantom** lopéren *n*.; tás *n*. **philanthropic** dòòn *v*. **philanthropist** dónésìàm *n*.; lɔʝɔkɔtáw<sup>a</sup> *n*. **philanthropists** roɓa ni ɡúrítínía dayaák<sup>a</sup> *n*. **philanthropy** daás *n*. **Philemon (biblical)** Pɨlɛmɔ́nɛ̀ *n*. **Philippians (biblical)** Pilípoik<sup>a</sup> *n*.
**phlebotomize** kɔ́ɛ́s *v*. **phlegm** ɗɔ̀ƙɔ̀n *n*.; ɲarʉ́kʉ́m *n*. **phlegm (newborn)** kíɓɔ́ɔ̀z *n*. **Phoenix reclinata** lɔ̀kàtàt<sup>a</sup> *n*. **phone** dʉrʉdʉr *n*.; ɲásím *n*. **phone in** iwetés *v*. **phone out** iwésúƙot<sup>a</sup> *v*. **phony** ɨtsárʉ́ánón *v*.; láŋ *n*. **phoot!** pʉ̀ʉ̀tᶶ *ideo*.; rɛ̀s *ideo*. **photo(graph)** kúrúkúr *n*.; ɲɛ́pítsa *n*. **photograph** iwetés *v*. **Phymateus species** ɓɔlɔrɔts<sup>a</sup> *n*. **physical therapist** ʝʉ́rɛ́sìàm *n*. **physically fit** itsyátón *v*. **physician** ɗakɨtárìàm *n*. **pick** ɗʉmɛtɛ́s *v*.; iɗókóliés *v*.; ɲésurúr *n*. **pick (choose)** xɔ́bɛtɛ́s *v*. **pick (nibble)** ɲɛ́ɲɛ́s *v*. **pick (teeth)** mínɛ́s *v*. **pick at** itoɗítóɗés *v*. **pick categorially** ɨsíílɛtɛ́s *v*. **pick clean** iúréés *v*. **pick fight with** itoʝiés *v*. **pick off** ɨkáábɛs *v*.; ɨkákápɛ́s *v*.; moxés *v*. **pick out** ikuɗúkúɗés *v*.; ɨƙɛlɛs *v*.; ɨƙɛlɛtɛ́s *v*.; ƙɛ́lɛ́s *v*.; tɔsɛɛtɛ́s *v*. **pick up** ɗʉmɛ́s *v*.; ɗʉmɛtɛ́s *v*.; ɨɗɛpɛs *v*.; ɨɗɛpɛtɛ́s *v*. **pick up (multiply)** ɨɗɛpíɗɛ́pɛ́s *v*.; ɨɗɛpíɗɛ́pɛtɛ́s *v*. **pick up a scent** wetésá kɔíná<sup>ɛ</sup> *v*. **pick up and bring** ɨɛ́bɛtɛ́s *v*. **pick up and take** ɨɛ́bɛsʉƙɔt<sup>a</sup> *v*. **pick-pocket** ɗíɗítɛ́sʉƙɔt<sup>a</sup> *v*.; ɗíɗítɛtɛ́s *v*. **pickaxe** ɲésurúr *n*. **picked (chosen)** xɔ́bɔtɔ́s *v*.
**picture** kúrúkúr *n*.; ɲɛ́pítsa *n*. **piece** ʝulam *n*. **piece (small)** ʝulamáím *n*.; pɛsɛlam *n*.; pɛ́sɛ́lamed<sup>a</sup> *n*. **piece of junk** ƙwɛsɛ́*n*. **pierce** itsumés *v*.; pulés *v*. **pierce noisily** rɔɗɛ́s *v*. **pierce repeatedly** pulutiés *v*. **piercer** pulutiesíàm *n*. **pig** ɲéɡuruwé *n*. **pig out** iwótsóòn *v*.; ɲaɗésá ŋƙáƙá<sup>e</sup> *v*. **pigeon** bîb<sup>a</sup> *n*. **pigeon (green)** orómó *n*. **pigeon (olive)** lótúrum *n*. **pigeon (speckled)** rutúdùm *n*. **pigment** ɲáraŋɡí *n*. **pile** ɨnʉkʉ́nʉ́kɛ́s *v*.; itukes *v*.; kìts<sup>a</sup> *n*.; ɲatúkít<sup>a</sup> *n*.; tutukesíáw<sup>a</sup> *n*. **pile (of dry branches)** ràm *n*. **pile on** iɗóɗókés *v*. **pile on (add)** tasaɓes *v*. **pile up** iruketés *v*.; ituketés *v*.; kitsetés *v*.; tutuketés *v*. **piled up** iɗóɗókánón *v*.; tutukánón *v*. **pilfer** ɨtíɗíɗɛ́s *v*. **pill** cɛ̀mɛ̀rìèkw<sup>a</sup> *n*. **pillage** toɓésúƙot<sup>a</sup> *v*. **pillage and bring** toɓetés *v*. **pillow** dɨƙwam *n*. **pimples** ŋ́kaŋók<sup>a</sup> *n*. **pin** mʉtʉ *n*. **pin (safety)** ɲákwác *n*. **pinch** rɨɗɛ́s *v*.; tʉnɛ́s *v*. **pinch all over** tʉnʉ́tʉ́natés *v*. **pinch each other** tʉ́nítʉ́nímɔ́s *v*. **pinch flirtatiously** tunútúniés *v*. **pinch off** ƙɨtɛ́s *v*.; tɔɲɨmɛtɛ́s *v*.
**pinch up (granules)** tʉsɛ́s *v*. **pinkish-red** ɗiwiɗíwón *v*. **pinnacle (of hut)** lómoloró *n*. **pinpoint** ɗóɗiés *v*. **pinworms** lɔkɨtʉ́r *n*. **pipe (borehole)** ɲatsʉʉmáárí *n*. **pipe (tobacco)** làr *n*. **pipe-stem** laradakw<sup>a</sup> *n*. **piquant** ɓariɓárón *v*.; ɓárikíkón *v*. **PISO (parish intelligence and security officer)** tsítsá na kwáts<sup>a</sup> *n*. **pissed off** ɨlílíɔ̀n *v*. **pissed off (become)** ɨlílíɔnʉƙɔt<sup>a</sup> *v*. **pistol** ɲɛ́písítɔ̀l *n*. **pistol grip** ƙɔ̀ɓ <sup>a</sup> *n*. **pit (trapping)** ɲɔ́sɔ́ɔ́ƙat<sup>a</sup> *n*. **pitch** tɔrɛ́s *v*. **pitch (mead)** ts'ɔƙɛ́s *v*. **pitch (soccer)** ɲakwaanʝa *n*. **pitch away** tɔrɛ́sʉ́ƙɔt<sup>a</sup> *v*. **pitch this way** tɔrɛtɛ́s *v*. **pitcher** ɲewataʝá *n*. **pitcher (wooden)** ɲɛkʉlʉmɛ *n*. **pitcherful** ɲewataʝá *n*. **pitfall trap** ɲɔ́sɔ́ɔ́ƙat<sup>a</sup> *n*. **pitted** tsakátsákánón *v*. **Pittosporum viridiflorum** ɲékwaŋa *n*. **pity (have)** isyónón *v*. **pity on (take)** isyones *v*. **placate** ɨkanɛ́s *v*.; ɨkaníkánɛ́s *v*. **place** aw<sup>a</sup> *n*.; bácík<sup>a</sup> *n*.; eɡés *v*.; eɡetés *v*.; kíʝ<sup>a</sup> *n*. **place of honor** zɛƙɔ́áwa na maráŋ *n*. **placenta** ŋaxɔ̂b <sup>a</sup> *n*. **placid** tisílón *v*.
**plague** koɗó *n*. **plain** ɓàŋɔ̀n *v*.; dús *n*.; ɨɓámɔ́n *v*. **plainness** ɓaŋás *n*. **plait** bɛrɛ́s *v*.; sikwés *v*. **plait up** sikwetés *v*. **plan** iɗimiés *v*.; iɗimiesúƙot<sup>a</sup> *v*.; ipáŋƙeés *v*. **plan a time** hoetésá ɲásáatí *v*. **plane (air-)** iɗékè *n*. **plane (even)** ɨkʉlɛs *v*.; ɨkʉlɛtɛ́s *v*.; ɨkwalɛs *v*. **plane off** ɨkʉ́lɛ́sʉƙɔt<sup>a</sup> *v*. **plank** ɲáɓáo *n*. **plant** dakw<sup>a</sup> *n*.; íbɨtɛ́s *v*. **plant (unknown)** kɔ́rɔ́ɓáìdàkw<sup>a</sup> *n*.
**plant disease** xoúxoú *n*.
**plant species** bɛfácɛ́mɛ́r *n*.; bʉlʉbʉláta na sábàìkà<sup>e</sup> *n*.; bùsùbùs *n*.; dàlìs *n*.; dodík<sup>a</sup> *n*.; ɡàsàràkwàts<sup>a</sup> *n*.; ɡòmòʝòʝ<sup>a</sup> *n*.; ɡùʝ<sup>a</sup> *n*.; ídemecɛmɛ́r *n*.; ídèmèdàkw<sup>a</sup> *n*.; ídocɛmɛ́r *n*.; ɨɗak<sup>a</sup> *n*.; ʝáláts<sup>a</sup> *n*.; ʝɨʝîd<sup>a</sup> *n*.; ʝʉ̀ʝʉ̀ *n*.; kèlèrw<sup>a</sup> *n*.; kìmɔ̀ɗɔ̀rɔ̀ts<sup>a</sup> *n*.; kòkòròts<sup>a</sup> *n*.; kɔmɔ́m *n*.; komóts<sup>a</sup> *n*.; loɓóŋiɓóŋ *n*.; lomerúk<sup>a</sup> *n*.; lɔmɔ́y a *n*.; lɔsalát<sup>a</sup> *n*.; mɛ́rɛ́ɗɛɗɛ́ *n*.; múmùt<sup>a</sup> *n*.; ŋálómóyá *n*.; ŋímáarɔy<sup>a</sup> *n*.; ŋítɛ́sʉrɔ *n*.; ɲáɓáɓú *n*.; ɲálamorú *n*.; ɲápat<sup>a</sup> *n*.; ɲasal *n*.; ɲɛcaɓoy<sup>a</sup> *n*.; ɲɛ́ɛkɨmá *n*.; ɲékilitón *n*.; ɲesuƙuru *n*.; ɲɛ́sʉ́tɛ̀*n*.; ɲétúlerú *n*.; ɲéúɗe *n*.; ɲéúlam *n*.; ɲɛʉrʉlats<sup>a</sup> *n*.; ɲɛ́ʉrʉmɛmɛ́ *n*.; ɲoɗokole *n*.; ɔ́bɛ̀r *n*.; ɔ́ʝítínícɛmɛ́r *n*.; òŋòrìkwàts<sup>a</sup> *n*.; ráɡàn *n*.; sʉ́ƙʉ́sʉƙá *n*.; ts'aɗícɛ́mɛ̀r *n*.; tsákàts<sup>a</sup> *n*.; tsamʉya *n*.; tâb<sup>a</sup> *n*.; túḿbàb<sup>a</sup> *n*. **plantain** ɲómototó *n*.
**plaster** ɨlɔɓílɔ́ɓɛ́s *v*.; ɨpʉtsɛs *v*.; ɨwarɛs *v*.
**plaster (mud)** tànàŋ *n*.
**plaster (with mud)** tanaŋes *v*.
**plastered** ɨwarɔs *v*.
#### plate
**plate** ɲásaaní *n*. **plateau** lopem *n*. **platform** lɔpɨtá *n*. **platform (make a)** ɨpɛ́tɛ́ɛ́s *v*. **platoon** ɲépalatún *n*. **play** ɲaɓolya *n*.; wáák<sup>a</sup> *n*.; wáák<sup>a</sup> *v*. **play (dramatic)** wááka na támɔtɔ́s *n*. **play around with** wáákitetés *v*. **play the field (sexually)** weesa kíʝá<sup>e</sup> *v*. **play with** ɨmɨnímínɛ́s *v*.; mɨnímínatés *v*. **player** wáákààm *n*. **playful** wáákós *v*. **plead with** iƙenes *v*. **pleasant** dòòn *v*. **pleasantness** daás *n*. **please** ɨlákásítɛ́sʉƙɔt<sup>a</sup> *v*.; ɨmʉ́mʉ́ɨtɛtɛ́s *v*.; kóʝ<sup>a</sup> *adv*. **Plecthranthus species** ɡàsàràkwàts<sup>a</sup> *n*. **plentitude (here)** níbàdà *n*. **plentitude (there)** kíbàdà *n*. **plenty** ŋábɔnʉƙɔt<sup>a</sup> *v*.; nábɔnʉƙɔt<sup>a</sup> *v*.; tàn *n*. **pliable** lumúdòn *v*. **pliably** lùm *ideo*. **pliers** ɲɔkɔ́ɲɛ́t <sup>a</sup> *n*. **plop down** rɛfɛ́kɛ́ɲɔ̀n *v*. **plop!** pùs *ideo*. **plot (owned)** ɲeɗúkór *n*. **plot against** ɨmanɛs *v*. **plow** tɔkɔ́bɛs *v*. **plow (make to)** tɔkɔ́bɨtɛtɛ́s *v*. **plow (ox-)** ɲɛ́mɛlɛkʉ́à nà ɦyɔ̀ɔ̀ <sup>ɛ</sup> *n*. **plowed** tɔkɔ́bɨtɔtɔ́s *v*. **plower** tɔ̀kɔ̀bààm *n*. **plowing** tɔ̀kɔ̀b a *v*.
**plowing season** tɔkɔbatsóy<sup>a</sup> *n*. **pluck** sɔrɛ́s *v*.; tɔtsʉɗɛs *v*.; tʉtsʉɗɛs *v*. **pluck off** ɓotsetés *v*. **pluck off repeatedly** ɓotsotiés *v*. **plug** ɨmíɗítsɛ́s *v*.; ts'ʉ́bʉlát<sup>a</sup> *n*.; ts'ʉ̂b <sup>a</sup> *n*.; tɨts'ɛ́s *v*.; tʉ́zʉɗɛ́s *v*. **plug (chewable)** ts'àf *n*. **plug (lip)** ɡwalát<sup>a</sup> *n*. **plug oneself in** ɨmíɗítsɛ́sa así *v*. **plug up** tɨts'ɛ́sʉ́ƙɔt<sup>a</sup> *v*.; tits'ímétòn *v*.; tʉ́zʉɗɛ́sʉ́ƙɔt<sup>a</sup> *v*. **plugged** tɨts'ɔ́s *v*. **plume** tùk<sup>a</sup> *n*. **plumed** tsowírímòn *v*. **plump** dɛʝɛ́dɔ̀n *v*.; zízòn *v*. **plump person** zízònìàm *n*. **plumply** dɛ̀ʝ ɛ *ideo*. **plunder** iɓolíɓólés *v*.; iɓolíɓólésuƙot<sup>a</sup> *v*.; taɓales *v*.; toɓésúƙot<sup>a</sup> *v*. **plunder and bring** toɓetés *v*. **plural** kòmòn *v*. **plurality** komás *n*. **pneumonia** ɲeɗekea bákútsìkà<sup>e</sup> *n*. **pock** itwelítwélés *v*. **pocked** itwelítwélós *v*.; tsakátsákánón *v*. **pocket** ofur *n*. **pocket (back)** ofura na ʝírì *n*. **pocket (front)** ofura na wáxì *n*. **pockmarked** tsakátsákánón *v*. **podium** lɔpɨtá *n*. **point** ekw<sup>a</sup> *n*.; ekwed<sup>a</sup> *n*.; náƙáfɛ̀d <sup>a</sup> *n*. **point (topic)** mɛnéékw<sup>a</sup> *n*. **point (word)** tódèèkw<sup>a</sup> *n*. **point at** ɗóɗés *v*.; ɗoɗésúƙot<sup>a</sup> *v*. **point at sunset** ɗóɗiesá tsòònì *v*. **point backward** kámáránón *v*. **point downward (of horns)** ɨlʉ́kánètòn *v*.
**point of departure** wàxɛ̀d <sup>a</sup> *n*. **point out** ɗóɗés *v*.; ɗoɗésúƙot<sup>a</sup> *v*. **point to** ɗóɗés *v*.; ɗoɗésúƙot<sup>a</sup> *v*. **point to secretly** ɗóɗiés *v*. **pointed** ɨwítsɔ́n *v*.; ts'íts'ɔ́n *v*. **pointless** buɗámón *v*. **pointless talk** ɨɓámɔ́nìtòd<sup>a</sup> *n*. **points** áƙátìkìn *n*.; ekwin *n*. **pointy** tsuɓáánètòn *v*. **poison** cɛ̀mɛ̀r *n*.; ɨŋaalɛ́s *v*.; ɲekísórìt<sup>a</sup> *n*. **poisoner** ɨŋaalɛ́síàm *n*. **poke** iƙumes *v*.; itsemes *v*. **poke around on** itsemítsémés *v*. **Pokot person** Ŋúupéám *n*. **pole (forked)** titír *n*. **pole (horizontal)** rìkw<sup>a</sup> *n*. **pole (wooden)** dakw<sup>a</sup> *n*. **polecat** ɲewuruŋorok<sup>a</sup> *n*. **police** pólìs *n*. **police post** ɲɛ́pɔ́sìt<sup>a</sup> *n*. **polish** iríƙéés *v*. **politick** sʉ́bɛ̀s *n*. **pollen** ɗukes *n*.; ɨɔk<sup>a</sup> *n*. **polling station** ɡóózésíàw<sup>a</sup> *n*. **pollywog** ŋʉ́ɗʉ́ŋʉ́ɗ <sup>a</sup> *n*. **polydactyly** ɲɛ́ɗɔ́nɨɗɔn *n*. **pond** tábàr *n*. **pond water** tábarɨcue *n*. **ponder** ɲɛɓɛ́s *v*.; tamátámatés *v*.; tamɛtɛ́s *v*.; tamítámiés *v*. **poo-poo!** dí *nurs*. **pooched out** ɓotólómòn *v*. **poofy** bʉlʉbʉlɔs *v*. **pool** ɨmɨlímílɔ̀n *v*.; tábàr *n*. **pool (riverbed)** ɲéɓwál *n*.
**pool (rock)** sát<sup>a</sup> *n*. **poop** ets'<sup>a</sup> *n*.; nts'áƙón *v*. **poor** bùlòn *v*.; ikúrúfánón *v*. **poor as a dog** iŋókíánón *v*. **poor eyesight (have)** múɗúkánón *v*.; ŋwaxɔna ekwitíní *v*. **poor person** ikúrúfánóníàm *n*.; ŋók<sup>a</sup> *n*. **poorly** ɡàànìk<sup>e</sup> *v*. **pop (soda)** ɲɔ́sɔ́ɗa *n*. **pop (sound)** ɗɛɗɛanón *v*.; rɛɗɛɗánón *v*. **pop out** ɨpírísɛtɛ́s *v*. **pop!** pɨrɨs *ideo*. **pope** abáŋ *n*.; pápà *n*. **populate** ínésuƙot<sup>a</sup> *v*. **population** ɲɛkɨmar *n*. **porch** hodzíŋ *n*. **porcupine** tɔ̀rɔ̀mìɲ *n*. **porridge** ŋáítɔ̀ *n*.; ɲéúʝi *n*. **porridge (fermented)** rùt<sup>a</sup> *n*. **porridge (thick)** ízotam *n*. **portion** taɲáléés *v*.; xɔnɔ́ɔ́kɔn *n*. **portion (best)** ɲopol *n*. **portion (first)** ɲopol *n*. **portion (of meat)** ɲekiner *n*. **portion out** taɲáléetés *v*. **Portulaca quadrifida** ɨɗak<sup>a</sup> *n*. **posho** tɔbɔŋ *n*. **posho (solid)** ɨlɨram *n*.; lúɡùm *n*. **posho (stiff)** tʉɗʉtam *n*. **posho (watery)** ɓɔtí *n*. **position** was *n*. **position (social)** zeís *n*.; zeísínànès *n*. **possess** ɡirés *v*. **possession (demonic)** lejénánès *n*. **possessions** ɲámáli *n*. **possible** itíyéetam *n*. **post (police)** ɲɛ́pɔ́sìt<sup>a</sup> *n*.
**postpone** íbokés *v*. **postpone repeatedly** dzúƙudzuƙiés *v*.; irotírótés *v*. **posture** was *n*. **pot** dóm *n*. **pot (metal)** ɲásipiryá *n*.; ɲésipiriyá *n*.; tsɨrɨmʉ́dòm *n*. **pot (of beer)** mɛ̀sɛ̀dòm *n*. **pot (small clay)** ɲeƙulu *n*. **pot (small)** dómáìm *n*. **pot bottom** dómóɔ̀z *n*. **pot of edible termites (first)** wàxìdòm *n*. **pot-belly** ɡwàʝ<sup>a</sup> *n*. **pot-holed** ƙumúƙúmánón *v*. **potable** wetam *n*. **potato (wild)** keîdz<sup>a</sup> *n*. **potato(es)** ɲɛ́ɓɨás *n*. **potbellied** heɓúlúmòn *v*. **potsherd** tɔɓɔk<sup>a</sup> *n*. **potter** bɛrɛ́síama dómítíní *n*. **pottery (broken)** tɔɓɔk<sup>a</sup> *n*. **pouch** ofur *n*. **poultry** ɲɔ́kɔkɔr *n*. **pounce** tonyámónuƙot<sup>a</sup> *v*. **pound** ɗúlútés *v*.; ɨɗatɛs *v*.; iɗoses *v*.; tɔ́ ts'ɛ́s *v*. **pound (in a mortar)** iwotses *v*. **pound (with a pestle)** íɲɛ́s *v*. **pound repeatedly** itsomítsómés *v*. **pounded (with a pestle)** íɲɔ́s *v*. **pour** ƙídɨƙídɔ̀n *v*.; ƙúdès *v*. **pour down** iyééseetés *v*. **pour from small opening** ádʉdʉƙɛ́s *v*. **pour into** ɨtʉrɛs *v*.; otés *v*.; otésúƙot<sup>a</sup> *v*. **pour out** furúdòn *v*.; iyééseetés *v*.; ƙúdesuƙot<sup>a</sup> *v*.; ƙúdetés *v*.
**pour out (noisily)** ídʉlɨdʉlɛ́s *v*. **pour out into** otetés *v*. **pour out to last drop** ɨʝíírɛ́sʉƙɔt<sup>a</sup> *v*. **pour to last drop** ɨʝíírɛ́s *v*. **pout** ɨɓʉ́tʉ́ŋɔ̀n *v*.; imutúmútòn *v*. **pouty** ɨmʉtʉ́mʉ́tɔ́s *v*. **poverty** ŋókínànès *n*. **poverty-stricken** ikúrúfánón *v*. **poverty-stricken person** ikúrúfánóníàm *n*. **powder** kabas *n*.; kábàsìn *n*. **powderily** lyàm *ideo*. **powdery** ɨwɨɗɔs *v*.; lyamádòn *v*. **power** ŋɡúf *n*.; ŋɨxás *n*.; ɲapéɗór *n*.; zeís *n*. **powerful** ŋìxɔ̀n *v*. **practice** itétémés *v*. **praise** itúrútés *v*.; tamɛɛs *v*. **praise oneself** itúrútésá así *v*. **prattle** ilemílémòn *v*. **pray** wáán *v*. **pray (call-and-response)** taƙates *v*. **pray against** taƙátésuƙot<sup>a</sup> *v*. **pray away** taƙátésuƙot<sup>a</sup> *v*. **prayer** wáán *n*. **prayer (call-and-response)** tàƙàt<sup>a</sup> *n*. **prayer (closing)** wáána na tɛ́zɛ̀tɔ̀nì *n*. **prayer book** ɲáɓúka wáánà<sup>e</sup> *n*. **prayer for gravedigger** wáána na muɗésíàmà<sup>e</sup> *n*. **prayerful person** wáánààm *n*. **praying mantis** tʉ́w<sup>a</sup> *n*. **preach** ɨtátámɛ́s *v*. **preacher** ɨtátámɛ́síàm *n*. **precarious** ɨpáríŋánón *v*. **precipice** látsó *n*. **precipitation** dìdì *n*.
**precipitous** iwósétòn *v*.; kʉ́bɛ̀lɛ̀mɔ̀n *v*. **precisely** dàn *adv*. **predator** loúk<sup>a</sup> *n*. **predawn** eúzòn *v*.; ɲaɓáít<sup>a</sup> *n*. **preeminence** zeís *n*. **pregnancy taboo** ts'ìn *n*. **pregnant** taríɔ́n *v*. **pregnant (newly)** ƙeɗétón *v*. **pregnant (prohibitively)** ts'ìnɔ̀n *v*. **pregnant (recently̠ )** sɨbánón *v*. **premeditate** ɨwɔ́ŋɔ́n *v*. **preoccupied** íɡùʝùɡùʝòn *v*.; itúmúránón *v*.; wasɨtɛsa iká<sup>e</sup> *v*. **preoccupy** itúmúránitésúƙot<sup>a</sup> *v*. **preparation (for travel)** sùɓèt<sup>a</sup> *n*. **prepare** ɨɗɨmɛ́s *v*.; ɨɗɨmɛ́sʉ́ƙɔt<sup>a</sup> *v*.; iɗimiés *v*.; iɗimiesúƙot<sup>a</sup> *v*.; itemités *v*. **prepare (food)** itiŋés *v*. **prepare oneself** iɗimiesá así *v*. **prepare to go** súɓánòn *v*.; suɓétón *v*. **prepared** ɨɗɨmɛ́sɔ́n *v*. **presence** ɡwarí *n*. **present** dónés *v*.; dónésuƙot<sup>a</sup> *v*.; takánón *v*.; tɔ́rɔ́bɛs *v*. **presently** nápáka na *adv*. **press** bízès *v*.; ɨɗɔtsɛs *v*.; ɨsɨkares *v*. **press all over** bízibizatés *v*. **press for details** ɨnɨnɛ́s *v*. **press on** ɨɗaŋíɗáŋɔ̀n *v*.; ɨmʉ́kɔ́ɔ̀n *v*.; ʝʉ́rɛ́s *v*. **press out** bízetés *v*.; ʝʉ́rɛ́sʉƙɔt<sup>a</sup> *v*.; ʝʉrɛtɛ́s *v*. **press repeatedly** ɨɗaŋíɗáŋɛ́s *v*. **pressed** rɔ́ƙɔ́rɔƙánón *v*. **pressure** ɨsɨkares *v*.; rɛ́ɛ́s *v*.; rɛɛtɛ́s *v*.; tɔrɛɛs *v*.
**pressured** toreimétòn *v*. **pressurize** ɨsɨkares *v*. **pretend** iɲétsóòn *v*.; iyétsóòn *v*. **prettiness** daás *n*. **pretty** dòòn *v*. **prevail** taɗáŋón *v*. **prevaricate** isuɗesa mɛná<sup>ɛ</sup> *v*.; itoŋetésá tódà<sup>e</sup> *v*.; yʉanitetés *v*.; yʉanón *v*. **prevarication** yʉɛ *n*. **prevent** ɨrɛtɛs *v*.; isíƙéés *v*.; itítírés *v*. **prey on** tɔɓɛlɛs *v*. **price** dzîɡw<sup>a</sup> *n*.; dzíɡwèsèd<sup>a</sup> *n*.; ɲéɓéy<sup>a</sup> *n*. **prick** ts'ɔɗɨtɛs *v*. **priest (Catholic)** páɗɛ̀r *n*. **primate (female)** ɔɡɛraŋwa *n*. **primate (half-grown)** kukát<sup>a</sup> *n*. **primate (male)** ɔ̀ɡɛ̀r *n*. **primate infant** kíɗɔlɛ́*n*. **primer (ignition)** ɔ́zɛ̀d <sup>a</sup> *n*. **principal** ámázeáma ɲésukúluⁱ *n*. **print** ipíríntiŋeetés *v*. **print a book** iwetésá ɲáɓúkwì *v*. **prison** zíkɛ́sìàw<sup>a</sup> *n*. **prison guard** cookaama zíkɛ́siicé *n*. **prisoner** ŋímamɓʉ́sìàm *n*. **pristine** ɗòwòn *v*.; tɨlíwɔ́n *v*.; xɔ́dɔ̀n *v*.; xɔtánón *v*. **private** búdòs *v*. **probably** kárɨká *adv*.; ntsúó ts'ɔɔ *pro*. **probe** ɨnɨnɛ́s *v*. **problems** mɛn *n*.; ŋítsan *n*. **proboscis (elephant)** òŋòrìkwɛ̀t <sup>a</sup> *n*. **procedure** ɲɛpɨtɛ *n*. **proceed** pórón *v*. **proceed (to do)** itáƙúòn *v*.
**process issues** bɛrɛ́sá mɛná<sup>ɛ</sup> *v*. **procession** ɲɛɗʉpɛ *n*. **prod** iƙumes *v*.; itsemes *v*. **produce** ƙwaatítetés *v*.; pulutetés *v*. **produce a lot of** cɛɛtɛ́s *v*. **produce seeds** eɡésá ekwí *v*. **product** dzíɡwam *n*.; dzíɡwetam *n*.; dzííƙotam *n*. **profit** bɨtɨtam *n*.; ɨkɛ́ítɛtɛ́s *v*. **profit from** raʝetés *v*. **progeny** kwats<sup>a</sup> *n*. **prohibit** dimités *v*.; dimitetés *v*.; itáléés *v*. **prohibited** itáléánón *v*.; itálóós *v*. **prohibition** ɲatal *n*. **project** ɨrʉtsɛs *v*.; ɲɛ́prɔ́ʝɛ̀kìt<sup>a</sup> *n*. **project anorectum** doletésá ɔ́zà<sup>ɛ</sup> *v*. **prolapse** ɨtsʉ́bʉ̀ɗɔ̀n *v*. **prolapsed** ɨtsʉ́bʉ̀ɗʉ̀mɔ̀n *v*. **prolific** bòmòn *v*. **prolong** zikíbètòn *v*.; zikíbitésúƙot<sup>a</sup> *v*. **prolonged (become)** zikíbonuƙot<sup>a</sup> *v*. **promenade** tasɔ́ɔ́n *v*. **promiscuous (sexually)** furés *v*.; ɨmáláánón *v*. **promise** iɓoletés *v*. **promise each other** iɓólínós *v*. **promote** zeites *v*.; zeitésuƙot<sup>a</sup> *v*. **pronely** ɓɛlɛlɛts<sup>ɛ</sup> *ideo*. **pronounce** kʉtɔnʉƙɔt<sup>a</sup> *v*. **prop** ɨƙɔŋɛs *v*.; titirés *v*. **prop (the head)** díƙwɛ́s *v*. **prop against** ɨƙɔ́ŋítɛ́s *v*.; ɨƙɔ́ŋítɛ́sʉƙɔt<sup>a</sup> *v*. **prop on** ɨƙɔ́ŋítɛ́s *v*.; ɨƙɔ́ŋítɛ́sʉƙɔt<sup>a</sup> *v*. **prop up** ɨƙaŋɛs *v*.; titiretés *v*. **propane gas** ɲáɡás *n*.
**propeller blade** ɗàw<sup>a</sup> *n*.; suɡuráɗáw<sup>a</sup> *n*. **proper** itémón *v*. **proper (of many)** dayaakón *v*. **property** kíʝ<sup>a</sup> *n*.; ɲámáli *n*. **prophecy** faɗás *n*. **prophesy** fàɗòn *v*.; ikúʝíánón *v*. **prophet** fàɗònìàm *n*.; ɲakuʝíícíkáàm *n*. **propolis** sɔs *n*. **propped against** ɨƙɔ́ŋítɔ́s *v*. **propped on** ɨƙɔ́ŋítɔ́s *v*. **proprietor** ámêd<sup>a</sup> *n*. **proscribe** dimités *v*.; dimitetés *v*. **prosecutor** ɨsíítɛ́sìàm *n*. **prospect** ɦyeités *v*. **prosperity** ídzànànès *n*.; zɛƙwa ná dà *n*. **prostitute** ɲamáláɨt<sup>a</sup> *n*. **prostrate** bùkòn *v*.; bukukánón *v*. **Protea gaguedi** ɲícwéɲé *n*. **protect** cookés *v*. **protected** cookotós *v*. **protector** còòkààm *n*. **protest** nɛpɛƙánón *v*. **Protestant** sɛ́mìs *n*. **protract** zikíbitésúƙot<sup>a</sup> *v*. **protracted (become)** zikíbonuƙot<sup>a</sup> *v*. **protrude** tɨbíɛ́tɔ̀n *v*. **protrude (of ears)** kweelémòn *v*. **protruding** ɓotólómòn *v*.; tɨbíɔ́n *v*. **protrusion** rúɡèts<sup>a</sup> *n*. **protuberance** rúɡèts<sup>a</sup> *n*. **proud** itúrón *v*. **proverb** taɗápítotós *n*. **provide for** bɔnɛ́s *v*.; íɡɔɲɛ́s *v*. **provider** bɔnɛ́ám *n*. **provision** bɔn *n*. **provocation** ɓɛkam *n*. **provocativeness** ʝʉ́ránànès *n*.
#### provoke
**provoke** ɓɛkɛtɛ́s *v*.; ɨsʉ́sʉ́ɛ́s *v*.; itsemes *v*.; itsótsóés *v*. **provoke (verbally)** ɨtɔ́ŋɔ́ɛ́s *v*. **provoking** ɓɛkánón *v*.; ʝʉ́ránòn *v*. **prowl** tonyámón *v*.; totséɗón *v*. **prune** ɨƙwáƙwárɛ́s *v*.; isésélés *v*. **pry apart** ɓereɲiés *v*. **pry bar** ɲotolim *n*. **pry open** ɓereɲiés *v*. **pseudo-** láŋ *n*. **Pseudocedrela species** ɲókotit<sup>a</sup> *n*. **Psidium guajava** ɲóɡóva *n*. **psyche** ɡúr *n*. **ptooey!** tʉ̀ *ideo*. **puberty (enter)** teɓúránétòn *v*. **puberty (enter, of boys)** ɨɓʉyákòn *v*. **pubic area** didis *n*. **pubic bone** didisíɔ́k <sup>a</sup> *n*. **pubic hair** didisísíts'<sup>a</sup> *n*.; ɔ́zàsìts'<sup>a</sup> *n*.; tɛ̀mʉ̀r *n*. **pubis** didisíɔ́k <sup>a</sup> *n*. **public** ŋáɲɔ́s *v*. **publish a book** iwetésá ɲáɓúkwì *v*. **puddle** ɨmɨlímílɔ̀n *v*. **pudendum** didis *n*. **pudgily** lèɓ<sup>u</sup> *ideo*. **pudgy** ɡerúsúmòn *v*.; leɓúdòn *v*.; rexúkúmòn *v*. **puff adder** bɛf *n*. **puff up** xuanón *v*.; xuxuanitetés *v*.; xuxuanón *v*. **puffily** bɔ̀f *ideo*.; lèɓ<sup>u</sup> *ideo*. **puffy** bɔfɔ́dɔ̀n *v*.; bʉlʉbʉlɔs *v*.; dúduránón *v*.; leɓúdòn *v*. **puke** ɦyɛnɛ́tɔ́n *v*.; ɦyɛ̀nɔ̀n *v*.
**pull** béberés *v*.; eminés *v*.; iɓwates *v*.; ɨʝʉkɛtɛ́s *v*.; ipoles *v*.; ɨtsɔrɛtɛ́s *v*. **pull (make)** ɨʝʉ́kítɛtɛs *v*. **pull along** béberiés *v*. **pull apart** ɗusés *v*.; ɗusésúƙot<sup>a</sup> *v*.; ɗusutes *v*.; eminiés *v*.; ɨkɛ́ŋɛ́ɗɛ́s *v*.; tɔŋɛɗɛs *v*. **pull away** béberésúƙot<sup>a</sup> *v*.; eminésúƙot<sup>a</sup> *v*. **pull back** dolés *v*.; doletés *v*.; rʉʝɛ́s *v*. **pull back foreskin** doletésá kwaní *v*. **pull down** inietés *v*.; lɔkɔɗɛtɛ́s *v*. **pull forcefully** iɓwatetés *v*. **pull in** béberetés *v*. **pull off** ɓotsetés *v*.; eminésúƙot<sup>a</sup> *v*.; tɔkɛtɛs *v*.; tɔkɛ́tɛ́sʉƙɔt<sup>a</sup> *v*.; tolés *v*.; toletés *v*. **pull off (bark)** iɓóɓólés *v*. **pull off repeatedly** ɓotsotiés *v*.; tolotiés *v*. **pull on** ɗʉ́rɛ́s *v*.; ɗʉtɛ́s *v*. **pull oneself away** ƙɛlɛtɛ́sá así *v*. **pull oneself back** rʉʝɛtɛ́sá así *v*. **pull out** ɗʉrɛtɛ́s *v*.; ɗʉtɛtɛ́s *v*.; eminetés *v*.; faɗetés *v*.; ipoletés *v*.; ritetés *v*.; ruutésuƙot<sup>a</sup> *v*.; ruutetés *v*.; tɔkɛtɛs *v*.; tɔkɛ́tɛ́sʉƙɔt<sup>a</sup> *v*.; tɔkɛtɛtɛ́s *v*.; tolés *v*.; toletés *v*.; tɔtsʉɗɛs *v*.; tʉtsʉɗɛs *v*.; tuutes *v*.; tuutetés *v*. **pull out repeatedly** tolotiés *v*. **pull over** ɨtɨlɛtɛ́s *v*. **pull this way** béberetés *v*. **pull up** ɗués *v*.; ɗuetés *v*.; eminetés *v*.; ipoletés *v*.; rués *v*. **pulsate** dìkwòn *v*.; ƙádiƙádòn *v*. **pulse** ƙádiƙádòn *v*. **pulverize** itsomes *v*. **pummel** ɗúlútés *v*.
#### pump
**pump** íɲɛ́s *v*. **pump up** ɨsɨkɛs *v*. **pumpkin** kaiɗey<sup>a</sup> *n*. **pumpkin (oblong)** naperorwá *n*.; tsòkòlòr *n*. **pumpkin (small unripe)** ɗɔ̀l *n*. **pumpkin (unripe)** ŋíkalʉtʉ́rɔ *n*. **pumpkin juice** kaiɗeícúé *n*. **pumpkin piece** kaiɗeíbɔrɔƙɔ́ƙ <sup>a</sup> *n*. **pumpkin ring** ɨbɔt<sup>a</sup> *n*. **pumpkin seed** kaiɗeíékw<sup>a</sup> *n*. **pumpkin stem base** kaiɗeíáƙát<sup>a</sup> *n*. **punch** tanaŋes *v*. **punch (a hole)** húbutés *v*.; pulés *v*.; ruɗés *v*. **puncture** ɓɛkɛ́s *v*.; ɓɛkɛtɛ́s *v*.; itsumés *v*.; pulés *v*.; ruɗés *v*. **puncture repeatedly** pulutiés *v*. **punish** iɗoŋes *v*. **puny** ɡɔɗírímɔ̀n *v*. **pupil** isóméésíàm *n*.; ɲósomáám *n*. **pupil (of eye)** tiléŋ *n*. **puppy** ŋókíìm *n*. **pure** ɓèts'òn *v*.; tɨlíwɔ́n *v*.; xɔ́dɔ̀n *v*.; xɔtánón *v*. **purity (of food)** lɛtsɛ́kɛ́ɛd<sup>a</sup> *n*. **purplish-red** kɨpʉ́ránètòn *v*. **purpose** ɲákásìèd<sup>a</sup> *n*. **purr** ŋɔ́rɔ́rɔ̀n *v*. **purse** ɲáníɓàk<sup>a</sup> *n*. **pursue** ɨlɔŋɛs *v*.; ɨmítíŋɛɛ́s *v*. **pursue after** ɨlɔ́ŋɛ́sʉƙɔt<sup>a</sup> *v*. **pursue each other sexually** ríínós *v*. **pus** báts'<sup>a</sup> *n*. **push** ɨɗɔtsɛs *v*.; ɨʝʉkɛs *v*.; ɨpʉnɛs *v*.; rités *v*.
**push along** ɨʝʉkʉ́ʝʉ́kɛ́s *v*. **push aside repeatedly** iʝúkúmiés *v*. **push away** ɨʝʉ́kɛ́sʉƙɔt<sup>a</sup> *v*.; ritésúƙot<sup>a</sup> *v*. **push buttons (provoke)** itsemes *v*. **push down** ɗaɗátésuƙot<sup>a</sup> *v*. **push in and out** irúrúƙés *v*. **push into** ipúkútsésuƙot<sup>a</sup> *v*.; lakates *v*. **push into repeatedly** lakatiés *v*. **push near to** bɨɲɛ́s *v*. **push on** bízès *v*. **push over** ɨtílɛ́sʉƙɔt<sup>a</sup> *v*. **push over side** lakates *v*. **push over side repeatedly** lakatiés *v*. **pushing** ʝʉ́ránòn *v*. **put** eɡés *v*.; eɡetés *v*. **put ahead** ɛkwɨtɛs *v*.; ɛkwítɛ́sʉƙɔt<sup>a</sup> *v*. **put alongside** ɨnapɛs *v*. **put aside** ɨnápɛ́sʉƙɔt<sup>a</sup> *v*.; ɨŋáɗɛ́ɛ́s *v*.; ɨŋáɗɛ́ɛ́sʉƙɔt<sup>a</sup> *v*.; óɡoɗés *v*.; óɡoɗésúƙot<sup>a</sup> *v*.; oƙésúƙot<sup>a</sup> *v*. **put away** óɡoɗés *v*.; óɡoɗésúƙot<sup>a</sup> *v*.; oƙésúƙot<sup>a</sup> *v*. **put back** raʝésúƙot<sup>a</sup> *v*.; raʝetés *v*. **put beside** ɨnapɛs *v*. **put in** ɓuƙítésuƙot<sup>a</sup> *v*.; imetsités *v*. **put in a sling** íbatalɛ́s *v*. **put in front** ɛkwɨtɛs *v*.; ɛkwítɛ́sʉƙɔt<sup>a</sup> *v*. **put in jail** eɡésá ɲáʝálaák<sup>e</sup> *v*. **put in order** ɨnábɛ̀sʉ̀ƙɔ̀t a *v*.; itíbès *v*.; itíbesúƙot<sup>a</sup> *v*. **put in twos** leɓetsítésuƙot<sup>a</sup> *v*. **put inside** xutésúƙot<sup>a</sup> *v*. **put nearby** taraŋés *v*. **put off** íbokés *v*. **put off odor** mídzòn *v*. **put off repeatedly** dzúƙudzuƙiés *v*.; irotírótés *v*. **put on** iwales *v*.; ŋábès *v*.
**put on (beads)** otés *v*. **put on a feather** iwalesa túkà<sup>e</sup> *v*. **put out** ts'eites *v*.; ts'eítésuƙot<sup>a</sup> *v*. **put to sleep** epítésuƙot<sup>a</sup> *v*. **put to work** ikásíitetés *v*.; teréɡanitetés *v*. **put together** itóyéés *v*. **put two-by-two** leɓetsítésuƙot<sup>a</sup> *v*. **put up** íbokés *v*. **put up with** taɗaŋes *v*. **put upright** ɨtsírítɛtɛ́s *v*.; tsírítɛtɛ́s *v*. **put weight on** tuɓútitésúƙot<sup>a</sup> *v*. **put in** otés *v*. **putrefy** mʉ́mʉ́tɛ̀tɔ̀n *v*. **putrid** mʉ́mʉ́tɔ̀n *v*.; ŋorótsánón *v*. **pyrosis** kíɓɔ́ɔ̀z *n*. **python** ɲomórótòt<sup>a</sup> *n*. **python tail-tip** ɲéɡets<sup>a</sup> *n*. **quaff** itúlákáɲés *v*. **quail** ɲélúru *n*. **quake** irikíríkòn *v*.; kwalíkwálɔ̀n *v*.; ɲéríkirik<sup>a</sup> *n*. **quaking sound** ɓʉlʉɓʉl *ideo*. **qualms (have)** paupáwón *v*. **quarrel** dèƙw<sup>a</sup> *n*. **quarrel (of many)** ilérúmùòn *v*. **quarrel with (start a)** déƙwítetés *v*. **quarreler** deƙwideƙosíám *n*. **quarreling** ɲelerum *n*. **quarrelsome** deƙwideƙos *v*. **quarter (area)** nabɨɗɨt<sup>a</sup> *n*. **quartzite** séy<sup>a</sup> *n*. **queasy** iláƙízòn *v*.; itikítíkòn *v*.; talóón *v*. **queen bee** lókílóróŋ *n*.; okílóŋór *n*. **queen termite** dádata dáŋá<sup>e</sup> *n*.; dáŋádadát<sup>a</sup> *n*.; ŋwááta dáŋá<sup>e</sup> *n*.
**quelea (red-billed)** kimír *n*. **quench** ts'eites *v*.; ts'eítésuƙot<sup>a</sup> *v*. **question** esetés *n*.; esetiés *v*.; iŋáyéés *v*. **question things** itóŋóiesá mɛná<sup>ɛ</sup> *v*. **quibble** ɨɲʉ́ɲʉ́rɔ̀n *v*. **quick** itírónòn *v*.; wɛ́ɛ́nɔ̀n *v*. **quickly** ɗàmʉ̀s *adv*.; ɗɛ̀mʉ̀s *adv*. **quiet** ɨʝɛ́mɔ́n *v*.; líídòn *v*.; tisílón *v*. **quiet down** ɨʝɛ́mítɛ́sʉƙɔt<sup>a</sup> *v*.; ɨʝɛ́mɔ́nʉƙɔt<sup>a</sup> *v*. **quietly** ʝìr *ideo*.; lì *ideo*.; sokósíìk<sup>e</sup> *v*. **quish!** pìs *ideo*.; tùs *ideo*. **quit** kuritésúƙota así *v*. **quiver** kìtòn *v*.; kwalíkwálɔ̀n *v*.; nɛ́rɨnɛ́rɔ́n *v*. **quiver (begin to)** kitétón *v*. **quiver (make)** kitítésuƙot<sup>a</sup> *v*. **rabbit** tulú *n*. **rabbit nickname** bositíníàm *n*. **race** ɨrʉtsɛsa así *v*.; tsùwà *v*.; tsuwa na ɨɓákɔ́nì *n*. **rack (drying)** lɔpɨtá *n*. **racket** nɔ̀s *n*. **racket (make a)** ilélémùòn *v*. **radiance** daás *n*. **radiant** dòòn *v*. **radiator** ɡàfìɡàf *n*. **radio** dʉrʉdʉr *n*.; ɲéréɗi *n*. **rafter stick** tɨmɛ́l *n*. **rag** kàbàɗ<sup>a</sup> *n*. **rage** ɡaánàs *n*.; ɨlɛ́ɔ́n *v*.; ɲɛlɨl *n*. **ragged** ɨkárɔ́n *v*.; kɔrɔ́ɗɔ́mɔ̀n *v*.; rídziridzánón *v*.; rúɡuruɡánón *v*. **raid** toɓés *v*.; toɓésúƙot<sup>a</sup> *v*. **raid and bring** toɓetés *v*. **raider** toɓésíàm *n*. **rain** dìdì *n*.; wat<sup>a</sup> *n*.; wàtòn *v*.
**rain (drizzling)** ɲɛ́límɨlɨm *n*. **rain (dry season)** ódzadidí *n*. **rain (gentle)** ɗéródɛík<sup>a</sup> *n*. **rain (light)** kʉ́f *n*.; rɛ̀b <sup>a</sup> *n*. **rain elsewhere** ɨtsɛ́ɛ́rɔ̀ɔ̀n *v*. **rain from the west** tsóéàm *n*. **rain sickness** didiɲeɗeké *n*. **rain-stopper** tuɗúlónìàm *n*. **rainbow** nàtɔ̀lɔ̀kà *n*. **rains (eastern)** obólén *n*. **rains (intermittent)** ɲerupe *n*. **rainy season** diditsóy<sup>a</sup> *n*.; ɔtáy<sup>a</sup> *n*. **raise** ɓuƙés *v*.; ŋkáítetés *v*.; tasɛɛs *v*.; zikíbitésúƙot<sup>a</sup> *v*. **raise (make)** ɨkɛ́ítɛtɛ́s *v*. **raise buttocks** tsúdòn *v*. **raise the head** wasɨtɛsa iká<sup>e</sup> *v*. **raise to kick** dɛŋɛlɛs *v*. **raise up** ɨkɛ́ɛ́sʉƙɔt<sup>a</sup> *v*.; ɨkɛɛtɛ́s *v*. **raise up (develop)** bɛrɛ́s *v*. **raised up** ikeimétòn *v*.; ɨkɔɔtɔ́s *v*. **rake** ɨƙwɛrɛs *v*.; ŋírés *v*.; ɲaƙwárɛ́t <sup>a</sup> *n*. **rake (with nails)** soƙóríties *v*. **rally** iríréetés *v*.; sùtòn *v*. **ram** rʉtsɛ́s *v*.; rʉtsɛ́sʉ́ƙɔt<sup>a</sup> *v*. **ram (goat)** bɔfɔƙɔr *n*. **ram (sheep)** ɗóɗocurúk<sup>a</sup> *n*. **ram (young goat)** kɔl *n*. **ram into** ipúkútsésuƙot<sup>a</sup> *v*. **rampage** ɨlɛ́ɔ́n *v*. **ramshackle** kɔlɔlánón *v*. **ranger (game)** lɔɡɛ́m *n*. **rank** ɗɛtsɨɗɛ́tsɔ́n *v*.; ɨmʉ́sɔ́ɔ̀n *v*.; mʉ́mʉ́tɔ̀n *v*.; wízɨlílɔ́n *v*.; zeís *n*.; zeísínànès *n*. **rank (become)** ɨmʉ́sɛ́ɛ̀tɔ̀n *v*.
**rankle** ɓɛkɛtɛ́s *v*. **rankling** ɓɛkánón *v*. **ransack** iɓolíɓólés *v*.; iɓolíɓólésuƙot<sup>a</sup> *v*.; taɓales *v*. **rap on** ɨɗɔŋíɗɔ́ŋɛ́s *v*.; ikoŋíkóŋés *v*.; ɨtɔ́tɔ́ŋɛ́s *v*. **rap on repeatedly** ɨlɛrílɛ́rɛ́s *v*. **rap repeatedly** ɨɗɛɨɗɛ́ɛ́s *v*. **rapaciousness** lokoɗoŋironánés *n*. **rape** itikiesúƙot<sup>a</sup> *v*. **rapturous (become)** ɛfɔnʉƙɔt<sup>a</sup> *v*. **rare** búúbuanón *v*. **rashy** katúrúturánón *v*.; sómomóʝón *v*. **rat** ɗér *n*. **rat (giant Gambia)** lòlòt<sup>a</sup> *n*. **rat (house)** ɗérá na áwìkà<sup>e</sup> *n*. **rat (striped ground)** nàtsɛ̀r *n*. **rat poison** ɗérócɛmɛ́r *n*. **rat species** natélewá *n*.; tiŋátiŋá *n*.; tufúl *n*.; tʉ́síɗèr *n*. **ratel** lɛŋ *n*. **ration** ɨmɨnímínɛ́s *v*. **rattily** rɛ̀s *ideo*. **rattle (animal-hoof)** lots'ilots'<sup>a</sup> *n*. **rattle (gourd)** ɲɛ́ɛ́ƙɨɛ́ƙ <sup>a</sup> *n*. **rattle (leg)** coór *n*. **ratty** rɛsɛ́dɔ̀n *v*. **ravage** ɨsílíánɨtɛtɛ́s *v*. **raven (fan-tailed)** kʉ́ràk<sup>a</sup> *n*. **ravine** fòts<sup>a</sup> *n*.; ɲɔ́kɔ́pɛ̀ *n*. **ravine (river)** ɔrɔr *n*. **raw** ts'áɡwòòn *v*. **ray of light** bás *n*.; sʉ́w<sup>a</sup> *n*. **raze** ɡɨʝɛtɛ́s *v*.; ɲaɗésúƙot<sup>a</sup> *v*.; towutses *v*. **razor (handmade)** ɡìʝìt<sup>a</sup> *n*. **razorblade** ɲɔ́ŋɔmɓɛ́*n*.
**reach (a destination)** ɨtɔ́ɔ́n *v*. **reach (make)** ɨtaɨtɛ́s *v*. **reach a consensus** ɗɔtsɛtɛ́sá tódà<sup>e</sup> *v*. **reach and pull down** likiɗes *v*. **reach here (make)** ɨtaɨtɛtɛ́s *v*. **react against** toƙíróòn *v*. **react suddenly** tokúétòn *v*.; tokúréètòn *v*. **read** isóméés *v*. **read (teach to)** isómáitetés *v*. **readable** isómáìmètòn *v*. **reader** isóméésíàm *n*. **readings** ɲósomáicík<sup>a</sup> *n*. **ready** ɨɗɨmɛ́s *v*.; ɨɗɨmɛ́sɔ́n *v*.; ɨɗɨmɛ́sʉ́ƙɔt<sup>a</sup> *v*.; iɗimiés *v*.; iɗimiesúƙot<sup>a</sup> *v*.; itemités *v*. **ready (make)** itemités *v*. **ready (to eat)** àèòn *v*. **ready oneself** iɗimiesá así *v*. **ready to eat (become)** aeonuƙot<sup>a</sup> *v*. **ready to fight** iríríƙòn *v*. **ready to go (get)** súɓánòn *v*.; suɓétón *v*. **ready, set, go!** mérímeritsíò *interj*. **readying for harvest (of gardens)** aeonuƙota kíʝá<sup>e</sup> *n*. **real estate** kíʝ<sup>a</sup> *n*. **realize** walámón *v*. **really** easík<sup>e</sup> *n*.; kárɨká *adv*.; mʉ̀kà *adv*. **really (much)** pʉ́n *ideo*. **ream out** irúútés *v*. **reanimate** ɦyekitetés *v*. **reap** ɨrárátés *v*.; ɨrarɛs *v*.; tarares *v*.; weés *v*. **reaper** weésíàm *n*. **rear** ʝìr *n*.; kanɛd<sup>a</sup> *n*.; ɔ́zɛ̀d <sup>a</sup> *n*.; ɔ̂z *n*.; tasɛɛs *v*.
**rear end** ʝírɛ̂d <sup>a</sup> *n*.; ɔ́zɛ̀d <sup>a</sup> *n*. **reassign** ɨlɔpɛs *v*. **rebel** mɛnáám *n*.; ɲɛkɛsʉpan *n*.; terémón *v*. **rebels (Sudanese)** Ŋɨɲɛ́ɲɛ́y <sup>a</sup> *n*. **rebound** íbòtòn *v*.; iɗótón *v*. **rebuff** ɨmɛ́ɗɛ́lɛ́s *v*. **rebuke** dɔxɛ́s *v*.; dɔxɛ́sʉ́ƙɔt<sup>a</sup> *v*. **recall** anɛ́sʉ́ƙɔt<sup>a</sup> *v*.; anɛtɛ́s *v*.; tamɛ́sʉ́ƙɔt<sup>a</sup> *v*.; tamɛtɛ́s *v*. **recall repeatedly** aniesúƙot<sup>a</sup> *v*. **recce** irimírímés *v*. **recede** raʝámón *v*.; raʝánón *v*. **receipt** taatsakabáɗ<sup>a</sup> *n*. **receive** tɛ́bɛtɛ́s *v*. **recent** erútsón *v*. **recently** ts'ɔ̀ɔ̀ *adv*. **reclaim** ɨrapɛs *v*.; ɨrápɛ́sʉƙɔt<sup>a</sup> *v*. **recline** eponuƙot<sup>a</sup> *v*.; ɨɗɛ́ɗɔ́ɔ̀n *v*.; itsólóŋòn *v*. **recognize** ɦyeités *v*. **recoil from** itírákés *v*. **recollect** anɛtɛ́s *v*.; tamɛ́sʉ́ƙɔt<sup>a</sup> *v*.; tamɛtɛ́s *v*. **reconcile** apápánɛ̀ɛ̀tɔ̀n *v*.; apápánɔ̀ɔ̀n *v*.; ɨsílítɛ́sʉƙɔt<sup>a</sup> *v*. **reconnoiter** irimírímés *v*. **record** ɨrɛ́kɔ́ɗɨŋeés *v*.; tamɨtɛtɛ́s *v*. **record of attendance** éditíníkabáɗ<sup>a</sup> *n*. **recorded (on paper)** ɨƙɨrɔs *v*. **recount** isíséés *v*. **recover** ɨrapɛs *v*.; ɨrápɛ́sʉƙɔt<sup>a</sup> *v*.; ɨrapɛtɛ́s *v*.; ɨrapɛtɛ́s *v*. **rectify** ɨtsírítɛtɛ́s *v*.; tsírítɛtɛ́s *v*.; tɔɓɛɨtɛtɛ́s *v*. **rector** pásìtà *n*. **rectum** dzɔɗát<sup>a</sup> *n*. **recur** tɔrʉ́ɓɔ́n *v*.
remain
**recycle** ɨrɔmɛs *v*. **red** ɗìwòn *v*. **red (become)** ɗiwonuƙot<sup>a</sup> *v*. **red (make)** ɗiwítésuƙot<sup>a</sup> *v*. **red (of many)** ɗiwaakón *v*. **red (very)** tsòn *ideo*. **red-pod terminalia** ɡáʒàd<sup>a</sup> *n*. **redden** ɗiwítésuƙot<sup>a</sup> *v*.; ɗiwonuƙot<sup>a</sup> *v*. **reddish-brown** bɔɨbɔ́ɔ́n *v*. **redeemer** hoɗetésíàm *n*. **redo** iɲaƙes *v*.; iɲoƙes *v*.; iɲóƙésuƙot<sup>a</sup> *v*. **reduce** raʝámón *v*.; raʝánón *v*. **reed** kɛ̀ɗ <sup>a</sup> *n*. **reed (granary)** ɲétémets<sup>a</sup> *n*. **reed ring** nàtsìkw<sup>a</sup> *n*. **reed species** sɔ́ɡɛ̀kàk<sup>a</sup> *n*.; sɔ̂ɡ <sup>a</sup> *n*. **reed wreath** nàtsìkw<sup>a</sup> *n*. **reedbuck (Bohor)** ɲeɓuri *n*. **reedbuck (female mountain)** rɔ́ɡɛŋwa *n*. **reedbuck (male mountain)** cúkúɗùm *n*. **reedbuck (mountain)** rɔ̂ɡ <sup>a</sup> *n*. **reedmace** ìsìk<sup>a</sup> *n*. **reeds (small)** kɔ̀k <sup>a</sup> *n*. **reek** ilíánòn *v*.; mídzona ɗɛtsɨɗɛ́tsík<sup>ɛ</sup> *v*.; mídzònà ɗùk<sup>u</sup> *v*. **reel** ɡakímón *v*.; ɨtɛrítɛ́rɔ̀n *v*. **reenact** ɨŋɨtɛs *v*. **refer to** tákés *v*. **reflection** kúrúkúr *n*. **reform** cicianón *v*. **refuse** dimés *v*.; ts'ʉts'ʉ *n*.; wasɛ́tɔ́n *v*. **refuse treatment** béberésuƙota así *v*. **refute** ɨsalɛs *v*.; ɨsalɛtɛ́s *v*.; ɨsalɨtɛ́s *v*. **refuted** ɨsálímétòn *v*. **regale** ɨmʉ́mwárés *v*.
**regalia** ɲɛ́nɨs *n*. **regime** ɲápukán *n*. **region** ɲɛ́tɛɛr *n*. **regress** raʝánón *v*. **regret** anɛ́sʉ́ƙɔt<sup>a</sup> *v*. **regrind** iŋáɓúkés *v*. **regrow** ʝɔɓɛ́tɔ́n *v*.; tɔrʉ́ɓɔ́n *v*. **regrow (of hair)** ŋʉrʉrʉ́ɲɔ́n *v*. **regrowing** ʝɔ̀ɓɔ̀n *v*. **regurgitate** ɦyɛnɛ́tɔ́n *v*.; ɦyɛ̀nɔ̀n *v*.; xerétón *v*. **reign over** ipúkéés *v*. **reject** dimés *v*.; ɨmɛ́ɗɛ́lɛ́s *v*.; míʝés *v*. **rejoin** ɗɛsɛ́mɔ́n *v*.; tɔŋɛ́tɔ́nʉƙɔt<sup>a</sup> *v*. **rekindle (with breath)** fúts'iés *v*. **related** ɦyeímós *v*.; ɦyeínós *v*. **related by birth** ɦyeínósá ƙwaaté<sup>o</sup> *v*. **related by marriage** ɦyeínósá sits'ésú *v*.; ɲotánánès *n*. **relations (sexual)** ep<sup>a</sup> *n*. **relax** torwóónuƙot<sup>a</sup> *v*. **relaxed** torwóón *v*. **relay** ilotses *v*. **relay tower** ɲéɓusitá *n*. **release** hoɗés *v*.; hoɗésúƙot<sup>a</sup> *v*.; hoɗetés *v*.; itsues *v*.; itsuetés *v*.; talakes *v*. **releaser** hoɗetésíàm *n*. **reliable** ikékéɲòn *v*. **religion** ɲéɗíni *n*. **religious matters** ɲakuʝímɛ́n *n*. **relinquish** bɔlɛ́sʉ́ƙɔt<sup>a</sup> *v*.; taʝales *v*.; taʝálésuƙot<sup>a</sup> *v*.; taʝaletés *v*. **relocate** dzuƙés *v*.; ilóʝésuƙot<sup>a</sup> *v*. **relocate away** dzuƙésúƙot<sup>a</sup> *v*. **relocate one's home** ilotsesa zɛƙɔ́ ɛ *v*. **relocate this way** dzuƙetés *v*. **rely on** ɨƙɔŋɛs *v*. **remain** ʝɛʝɛ́tɔ́n *v*.; ʝɛ̀ʝɔ̀n *v*.
**remain behind** maɗámón *v*. **remainder** ʝírɛ̂d <sup>a</sup> *n*.; óɡoɗesam *n*. **remainder (of food)** ɲomokoʝo *n*. **remainders** ʝírín *n*. **remains (find)** ítés *v*. **remember** anɛ́sʉ́ƙɔt<sup>a</sup> *v*.; anɛtɛ́s *v*.; tamɛ́sʉ́ƙɔt<sup>a</sup> *v*.; tamɛtɛ́s *v*. **remember clearly** ɨpííríánón *v*. **remember often** aniesúƙot<sup>a</sup> *v*. **remind** tamɨtɛtɛ́s *v*. **remove** hoɗésúƙot<sup>a</sup> *v*.; hoɗetés *v*.; ƙanésúƙot<sup>a</sup> *v*.; ƙanetés *v*.; ts'álés *v*.; ts'aletés *v*.; tɔkɛ́tɛ́sʉƙɔt<sup>a</sup> *v*.; tɔkɛtɛtɛ́s *v*.; tuɓutes *v*.; tuɓútésuƙot<sup>a</sup> *v*.; tuɓutetés *v*.; tuutes *v*.; tuutetés *v*. **remove a bird** ƙanésúƙota ɡwaá<sup>e</sup> *v*. **remove gingerly** ɗítɛ́s *v*.; ɗítɛ́sʉƙɔt<sup>a</sup> *v*. **remove shoes** hoɗetésá taƙáíkà<sup>ɛ</sup> *v*. **remove the jaw of** taʝakes *v*. **removed from office** rúmánònà kàràtsʉ̀ *v*. **remunerate** taatses *v*. **remuneration** tààts<sup>a</sup> *n*. **rend** dzɛrɛ́s *v*. **render** ɨkɔɓɛs *v*. **rendezvous (sexual)** tirésíàw<sup>a</sup> *n*. **renounce claim over** óɡoés *v*. **rent** ipáŋƙeés *v*. **rent (torn)** dzɛrɔ́sɔ́n *v*. **repair** ɨɗɨmɛ́s *v*.; ɨɗɨmɛ́sʉ́ƙɔt<sup>a</sup> *v*.; ɨɗɨmɛtɛ́s *v*.; ɲimanites *v*.; rátsɛ́s *v*.; taɗapes *v*.; taɗapetés *v*. **repair repeatedly** rátsiés *v*. **repaired** ɨɗɨmɔ́s *v*.; taɗapos *v*. **repay** taatsésuƙot<sup>a</sup> *v*. **repeat** iɓóŋón *v*.; iɲaƙes *v*.; iɲoƙes *v*.; iɲóƙésuƙot<sup>a</sup> *v*.
**repeat endlessly** íɡuʝuɡuʝésa tódà<sup>e</sup> *v*. **repel** ɨɗáfɛ́sʉƙɔt<sup>a</sup> *v*.; ɨtíílɛ́s *v*. **repelled** ɨɗáfɛ́sʉƙɔta así *v*. **repent** cicianón *v*.; fítésuƙota ɡúró<sup>e</sup> *v*. **repent of sins** tʉlʉŋɛsa tɔsɛ́sɔ́nì *v*. **replace** imetsés *v*.; imetsités *v*. **replant (a garden)** iɓures *v*. **replicate** toputes *v*.; toputetés *v*. **reply** raʝés *v*.; raʝetés *v*.; taatses *v*.; taatsésuƙot<sup>a</sup> *v*.; taatsetés *v*. **report** ɗoɗésúƙot<sup>a</sup> *v*.; ɲéripót<sup>a</sup> *n*. **repose** ɨɗɛ́ɗɔ́ɔ̀n *v*. **representative** tódààm *n*. **reprimand** dɔxɛ́s *v*.; dɔxɛ́sʉ́ƙɔt<sup>a</sup> *v*. **reprobate** ɲárásíám *n*. **reproduce (copy)** toputes *v*.; toputetés *v*. **repulse** ɨlɛ́lɛ́ítɛtɛ́s *v*.; ɨtíílɛ́s *v*. **repulse each other** ɡʉts'ʉ́rínɔ́s *v*. **reputed** ɦyoós *v*. **request** wáánɛtɛ́s *v*. **rescue** ɨɛtɛ́s *v*. **rescuer** ɨɛtɛ́síàm *n*. **resect** ɓilés *v*. **resemble** ikwáánòn *v*.; ƙámón *v*.; topútétòn *v*. **reserved** toikíkón *v*. **residence** zɛƙɔ́áw<sup>a</sup> *n*. **resident** zɛƙɔ́ám *n*. **residue (beer)** ɗʉká *n*.; dàʝ<sup>a</sup> *n*. **residue (food)** ɲéɗúruɗur *n*. **resile** rʉʝɛ́s *v*. **resilient** itsyátón *v*. **resist** ɨƙaíƙɛ́ɛ́s *v*.; ɨƙáƙɛ́ɛ́s *v*.; ɨƙáƙɛ́ɛtɛ́s *v*.; ɨrɛtɛs *v*.; kwɛ́rɛɗɛ́ɗɔ́n *v*.; raʝés *v*. **resistant** nɨkwídɔ̀n *v*. **resistantly** nìkw<sup>ɨ</sup> *ideo*. **resistent** itsyátón *v*.
**resolve** itemités *v*. **resolve an issue** epitésúƙota tódà<sup>e</sup> *v*. **resound** arútón *v*. **respect** mòròn *v*.; xɛ̀ɓɔ̀n *v*. **respect each other** mórímós *v*. **respire** ɨɛ́ŋɔ́n *v*.; sʉ́pɔ́n *v*. **resplendent** dòòn *v*. **respond** raʝés *v*.; raʝetés *v*.; taatses *v*.; taatsésuƙot<sup>a</sup> *v*.; taatsetés *v*. **respond as a group** sùtòn *v*. **respond repeatedly** tébitebiés *v*. **responsible** ámêd<sup>a</sup> *n*.; wàsɔ̀n *v*. **responsible for things** wasɔna kúrúɓádù *v*. **resprout** ʝɔɓɛ́tɔ́n *v*. **resprouting** ʝɔ̀ɓɔ̀n *v*. **rest** ɨɛ́ŋɔ́n *v*. **rest (the head)** díƙwɛ́s *v*. **rest against** ɨƙɔŋɛs *v*.; ɨƙɔ́ŋítɛ́s *v*.; ɨƙɔ́ŋítɛ́sʉƙɔt<sup>a</sup> *v*.; tonokes *v*. **rest on** ɨƙɔŋɛs *v*.; ɨƙɔ́ŋítɛ́s *v*.; ɨƙɔ́ŋítɛ́sʉƙɔt<sup>a</sup> *v*. **rest up** ɨɛ́ŋɔ́nʉƙɔt<sup>a</sup> *v*. **restaurant** ŋƙáƙáhò *n*.; ɲéótèl *n*. **rested against** ɨƙɔ́ŋítɔ́s *v*. **rested on** ɨƙɔ́ŋítɔ́s *v*. **resting place** ìɛ̀ŋààw<sup>a</sup> *n*. **restless** ɗɛɲɨɗɛɲɔs *v*. **restless (unsettled)** tsɔnɨtsɔnɔ́s *v*. **restrain** ɨƙalíƙálɛ́s *v*.; isíƙéés *v*.; itítíkés *v*.; itítíketés *v*. **restrict** ɨrɨɗɛs *v*.; ɨrɨɗɛtɛ́s *v*. **restricted** ɨrɨɗɔs *v*. **resurrect** ɦyekétón *v*.; ɦyekitetés *v*. **resuscitate** fúts'iés *v*.; ikwárétòn *v*.
**retain** itítíkés *v*.; itítíketés *v*.; tatsáɗésuƙot<sup>a</sup> *v*. **retaliate** ɲaŋés *v*.; ɲaŋésúƙot<sup>a</sup> *v*. **retard** inípónítésúƙot<sup>a</sup> *v*. **retch** ʝaƙátós *v*.; toukes *v*.; touketés *v*.; xáƙátòn *v*. **retell** ɨŋɨtɛs *v*. **retire** rumétón *v*. **retort** taatses *v*.; taatsésuƙot<sup>a</sup> *v*.; taatsetés *v*. **retrace one's steps** ɨƙʉlʉ́ƙʉ́lɔ̀n *v*. **retract** dolés *v*.; doletés *v*.; rʉʝɛ́s *v*. **retract foreskin** doletésá kwaní *v*. **retract oneself** rʉʝɛtɛ́sá así *v*. **retreat** ɨpɛ́ɛ́rɔ̀n *v*.; raʝánón *v*.; rumétón *v*. **retrieve** tukuretés *v*.; tukutetés *v*. **retrieve (food)** lɛkɛ́s *v*. **retriever (of food)** lɛkɛ́síàm *n*. **return** iɓóɓóŋòn *v*.; raʝés *v*.; raʝésúƙot<sup>a</sup> *v*.; raʝetés *v*.; tɔrʉ́ɓɔ́n *v*. **return bride** xɛɛsʉ́ƙɔt<sup>a</sup> *v*. **return here** itétón *v*. **return there** itéón *v*.; itíón *v*. **return this way** iɓóɓóŋètòn *v*. **return to normal** xɔ́dɔnʉƙɔt<sup>a</sup> *v*. **reveal** ɗoɗésúƙot<sup>a</sup> *v*.; ɗóɗítetés *v*.; enitésúƙot<sup>a</sup> *v*.; enitetés *v*.; ilééránitetés *v*.; kwɛts'ɛ́s *v*. **revealed** kwɛts'ɛ́mɔ́n *v*. **Revelation (biblical)** Enitetés *n*. **revenge** ɲaŋés *v*.; ɲaŋésúƙot<sup>a</sup> *v*. **revere** itúrútés *v*.; mòròn *v*. **reverse** ikutúkútés *v*.; ikutúkútòn *v*. **revert** raʝánón *v*. **revive** fúts'iés *v*.; ɦyekétón *v*.; ɦyekitetés *v*.; ikwárétòn *v*.
**revolt** ɨlɛ́lɛ́ítɛtɛ́s *v*. **revolve** ɨríŋɔ́n *v*. **reward** tɔ́rɔ́bɛs *v*.; tɔ́rɔ́bɛsa na ílɔɛsí *n*. **rheum (dried)** dɔ̀x *n*. **rhinoceros (black)** óbìʝ<sup>a</sup> *n*. **rhomboid muscle** ɲɛ́sílɨsɨl *n*. **Rhus natalensis** mɨsá *n*. **Rhynchosia hirta** ɲéŋéso *n*. **rib** ŋabér *n*. **rib (lefthand)** betsínáŋabér *n*. **rib (lowest)** sʉ̀ɗ <sup>a</sup> *n*. **rib (meat)** kileleɓú *n*. **rib (righthand)** ŋƙáƙáŋabér *n*. **rib (upper)** tsɛ̀tsɛ̀kw<sup>a</sup> *n*. **rib bone** ŋabéríɔ̀k <sup>a</sup> *n*. **rib meat** ŋábèrìkèèm *n*. **rich** bàrɔ̀n *v*.; iʝákáánón *v*. **rich (get)** bárɛ́tɔ̀n *v*.; barɔnʉƙɔt<sup>a</sup> *v*. **rich (in taste)** wɨɲídɔ̀n *v*.; wɨɲɨwíɲɔ́n *v*. **rich (make)** barítɛ́sʉƙɔt<sup>a</sup> *v*. **rich person** bàrɔ̀àm *n*.; bàrɔ̀nìàm *n*. **riches** bàr *n*. **richly in taste** wìɲ *ideo*. **Ricinus communis** ɨmánán *n*. **rickety** ɗɔxɔ́dɔ̀n *v*.; ɡɔ́ɡɔ̀rɔ̀mɔ̀n *v*. **ricochet** iɗótón *v*. **rid oneself of** ɡʉts'ʉrɛs *v*.; itsúrúés *v*. **riddle** taɗápítotós *n*. **ride** otsés *v*. **ride (a bicycle/motorcycle)** hɔnɛ́s *v*. **ridge** ɡòɡòròʝ<sup>a</sup> *n*.; itóróɲés *v*.; kokór *n*.; zeket<sup>a</sup> *n*. **ridge (of hair)** sìɡìrìɡìr *n*. **ridge (vertical)** fátár *n*.; ɲɔɗɔ́kɛ́t <sup>a</sup> *n*. **ridge base** tsɨɨr *n*.
**ridged** toŋórómòn *v*.; toróŋómòn *v*. **ridges in (make)** itóróɲés *v*. **ridgetop (vertical)** fátárààk<sup>a</sup> *n*. **ridicule** tɔʝɛmɛs *v*. **rifle (bolt-action)** lomucir *n*. **rifle (short bolt-action)** ɲápaŋƙaláít<sup>a</sup> *n*. **rifle through** iɓolíɓólés *v*.; iɓolíɓólésuƙot<sup>a</sup> *v*. **rifled** iyérón *v*. **right** ɨtsírɔ́n *v*.; iyóón *v*.; tsírɔ́n *v*.; tɔɓɛ́ɔ́n *v*. **right (make)** ɨtsírítɛtɛ́s *v*.; tsírítɛtɛ́s *v*.; tɔɓɛɨtɛtɛ́s *v*. **right (typically)** toɓéíón *v*. **right away** ɗìr *adv*. **right hand** ŋƙáƙákwɛ̀t <sup>a</sup> *n*. **right here** nayé kɔ̀nà *dem*. **right hindleg** ɲálán *n*. **righthand rib** ŋƙáƙáŋabér *n*. **rigid** ɓotsódòn *v*.; kɛtɛ́rɛ́mɔ̀n *v*.; tsɛ́rɛkɛ́kɔ́n *v*. **rigidly** ɓòts<sup>o</sup> *ideo*. **rind** bɔɗɔ́k <sup>a</sup> *n*. **ring** ɨlíŋírɛ́s *v*.; ɨlɨrɛs *v*. **ring (a bell)** iwés *v*. **ring (finger)** ɲákaɓɔɓwáát<sup>a</sup> *n*. **ring (of ears)** iwákón *v*. **ring (stick)** ɲókokor *n*.; ɲɔkɔlɔɓɛr *n*. **ring hollow** ɗɛʉɗɛ́wɔ́n *v*. **ring of reeds** nàtsìkw<sup>a</sup> *n*.; ɲéleƙeré *n*. **ring-beam** ɲolóɗo *n*. **ring-tone** dikw<sup>a</sup> *n*. **ringed** itówóòn *v*. **ringlet (metal)** àɡìt<sup>a</sup> *n*. **ringworm** aɗáɗá *n*. **rinse** ɨlaɨlɛ́ɛ́s *v*.; ɨlɔ́lɔ́tsɛ́s *v*. **rinse (mouth)** íɡʉʝʉɡʉʝɛ́s *v*.; ɨmʉ́mʉ́ʝɛ́s *v*. **rip** dzɛrɛ́s *v*.
**rip off (cheat)** ɨmɔɗɛs *v*.; ɨmɔ́ɗɛ́sʉƙɔt<sup>a</sup> *v*. **ripe** àèòn *v*. **ripe (nearly)** ɨts'ɔ́ƙɔ́n *v*. **ripen** aeonuƙot<sup>a</sup> *v*.; ƙádòn *v*. **ripen quickly** hataikánón *v*. **ripen up** aeétón *v*. **ripening** kɔ̀ɓɔ̀n *v*. **ripped** dzɛrɔ́sɔ́n *v*.; láládziránón *v*. **rise** ŋkéétòn *v*.; ŋkóón *v*.; zikíbonuƙot<sup>a</sup> *v*. **rise (of sun)** tsòòn *v*.; tsoonuƙot<sup>a</sup> *v*. **rise (of voice)** ɔ́bɛ̀s *v*. **rise up (rebel)** terémón *v*. **risk** ɡaánàs *n*. **risky** ɡaanón *v*. **ritual (do a)** írés *n*. **ritual killing** síts'<sup>a</sup> *n*. **river** sàbà *n*. **river (small)** ɔrɔr *n*. **river basin** ɲɛrɛ́t <sup>a</sup> *n*. **river bottom** sàbààƙw<sup>a</sup> *n*. **riverbank (opposite)** ƙìròt<sup>a</sup> *n*. **riverbed** sàbààƙw<sup>a</sup> *n*. **riverbed pool** ɲéɓwál *n*. **roach** lɔmɛ́ʝɛ́kɛlɛ́*n*. **road** muce *n*.; ɲerukuɗe *n*. **road grader** séɓésìàm *n*. **roam** iwórón *v*.; tɛ́rɛ́s *v*. **roamer** ɓɛƙɛsɔsíám *n*.; irimesíám *n*.; iwórónìàm *n*. **roar** ábʉ̀bʉ̀ƙɔ̀n *v*.; béúrètòn *v*.; erutánón *v*.; irúrúmòòn *v*.; xérón *v*. **roast** ʝʉɛ́s *v*. **roast lightly** ɨɔ́ɓɔ́rɛ́s *v*. **roasted** ʝʉɔ́s *v*. **roasting ground** nakíríkɛ̀t <sup>a</sup> *n*.
**robber** lotáɗá *n*.; ŋuésíàm *n*. **robbery** lotáɗánànès *n*. **robe** ɲákaasó *n*. **robin-chat** loƙírot<sup>a</sup> *n*. **rock (crumbly)** lɔkabʉ́ás *n*. **rock (large)** taɓ<sup>a</sup> *n*. **rock (sedimentary)** sáɲamát<sup>a</sup> *n*. **rock (small)** ɡwas *n*. **rock (soft)** ɲékúkuse *n*.; sáɲamát<sup>a</sup> *n*. **rock (table)** ɡɨzá *n*. **rock back and forth** iikííkés *v*.; iukúúkés *v*. **rock crevice** tsarátán *n*. **rock pool** sát<sup>a</sup> *n*. **rock pool water** sátíkócue *n*. **rock well** mɔƙɔr *n*. **rock well water** mɔƙɔrɔ́cúé *n*. **rockily** ɡàts<sup>a</sup> *ideo*.; ŋàr *ideo*. **rocky** ɡatsádòn *v*.; ŋarʉ́dɔ̀n *v*.; rakákámòn *v*. **rocky outcrop** kúc<sup>a</sup> *n*.; rikírík<sup>a</sup> *n*. **rod** ʝʉrʉm *n*. **rod (cleaning)** sʉ́ƙʉ́tɛ́sítsɨrím *n*. **roil** íbɔtsɛ́sá así *v*.; íɡùlàʝòn *v*. **roll** kaɓéléɓelánón *v*.; tsitsikes *v*. **roll a root** tsitsikesa ʝɔtɛ́*v*. **roll around** aɓíɓílánón *v*. **roll around (in mouth)** ɨŋɔ́lɔ́ɓɔ́ɲɛ́s *v*. **roll away** tsitsíkésuƙota así *v*. **roll between hands** tsutsukes *v*. **roll over** iɓéléés *v*.; iɓéléìmètòn *v*. **roll repeatedly** tsitsikiés *v*. **roll this way** tsitsiketésá así *v*. **roll up** ɨpɔ́pírɛ́s *v*.; kakɨrɛ́s *v*.; tɔɓɨlɛtɛ́s *v*. **rolling sound** dɛ̀rɛ̀dɛ̀r *ideo*.; kùrùkùr *ideo*. **Romans (biblical)** Ŋírɔmánɔ́niik<sup>a</sup> *n*.
#### roof
**roof** hoɡwarí *n*. **roof (of mouth)** aƙár *n*. **roof ring (of reeds)** ɲéleƙeré *n*. **roof tip (woven)** tsùt<sup>a</sup> *n*. **roofing sheet** ɲámaamɓát<sup>a</sup> *n*. **rooftop (inner)** lɔɓîz *n*. **room** naƙʉ́lɛ́*n*. **room (space)** zɛƙɔ́áw<sup>a</sup> *n*. **roomy** ɨlɔ́lɔ́mɔ̀n *v*.; lalʉ́ʝɔ́n *v*. **roost** itsélélèòn *v*. **rooster** ɡwácúrúk<sup>a</sup> *n*. **root** dakúsɔ́k <sup>a</sup> *n*.; sɔk<sup>a</sup> *n*.; sɔkɛd<sup>a</sup> *n*. **root (of a tooth)** kwayɔ́ɔ́k <sup>a</sup> *n*. **rope** ŋún *n*.; ún *n*. **rope (braided)** natɨɓ<sup>a</sup> *n*. **rope (tree bark)** tɔ̀fɔ̀l *n*. **rot** ɗutúɗútánónuƙot<sup>a</sup> *v*.; masánétòn *v*.; mʉsánétòn *v*. **rotate** ɨlɨrɛs *v*.; irímétòn *v*.; irímítetés *v*.; irímón *v*. **rotate around in** irimes *v*. **rotate repeatedly** ɨlɨrílírɛ́s *v*. **rotted** ɗatáɗátánón *v*. **rotten** ɗutúdòn *v*.; masánón *v*. **rotten (very)** ɗùt<sup>u</sup> *ideo*. **rotten at core** ɓʉɓʉsánón *v*. **rotting** ɗutúɗútánón *v*. **rotund** pʉŋʉ́rʉ́mɔ̀n *v*. **rough** ƙumúƙúmánón *v*.; rúɡuruɡánón *v*. **rough (of a road)** ƙumúƙúmánón *v*. **rough (of a surface)** ŋaráɓámòn *v*. **roughen** cɛ́bɛ̀s *v*. **roughly** ɡwɛ̀ʝ ɛ *ideo*. **round** ɗukúditésúƙot<sup>a</sup> *v*.; ɗukúdòn *v*.; ɨlʉ́lʉ́ŋɔ́s *v*.; ɲásápari *n*.
**round (and thick)** baƙúlúmòn *v*. **round (make)** ɗukúditésúƙot<sup>a</sup> *v*.; ɨlʉ́lʉ́ŋɛ́s *v*. **round up** ikoŋetés *v*.; ɨkwɛtíkwɛ́tɛ́s *v*.; ƙalíƙálɛ́s *v*.; sáɡwès *v*. **roundly** ɗùk<sup>u</sup> *ideo*. **roused** iƙúrúmós *v*. **rove** tɛ́rɛ́s *v*. **rover** ɓɛƙɛsɔsíám *n*.; irimesíám *n*. **row** iƙures *v*. **rub** iríƙéés *v*.; ʝʉ́rɛ́s *v*.; ŋííɗɛ́s *v*. **rub around** iwulúwúlés *v*. **rub between fingers** sɨmɨɗímíɗɛ́s *v*. **rub down/out** ʝʉ́rɛ́sʉƙɔt<sup>a</sup> *v*.; ʝʉrɛtɛ́s *v*. **rub in hands** tsutsukes *v*. **rub off** iɲíɲíés *v*.; ŋííɗɛ́sʉ́ƙɔt<sup>a</sup> *v*. **rub vigorously** ídʉlɨdʉlɛ́s *v*. **rubber** ɗɔtɔ́*n*. **rubber (eraser)** ɲáráɓa *n*. **rubberily** rɔ̀ɓ ɔ *ideo*. **rubbery** rɔɓɔ́dɔ̀n *v*. **rubbish** ts'ʉts'ʉ *n*. **rubbish pile** ts'ʉts'ʉ́áw<sup>a</sup> *n*. **rubeola** púrurú *n*. **ruffle** dʉbɛ́s *v*. **rugged** rúɡuruɡánón *v*. **ruin** imóɲíkees *v*.; imóɲíkeetés *v*.; ináƙúés *v*.; ináƙúetés *v*.; ɨraŋɛs *v*.; ɨraŋɛtɛ́s *v*. **ruined** ináƙúós *v*.; ináƙúotós *v*.; ɨraŋɔs *v*.; ɨráŋʉ́nánón *v*. **ruined (become)** ɨraŋímétòn *v*. **ruinousness** ƙʉts'ánánès *n*. **rule** ipúkéés *v*.; ɨtsɨkɛs *n*. **ruler** ipúkéésíàm *n*.; tòtwàrààm *n*. **rumble** ɗukuɗúkón *v*.; ɨkílɔ́n *v*.; tɔtɔanón *v*. **rumble off** itíƙíròòn *v*.; itíríƙòòn *v*.
salivate
**rumen** ɲɛ́pʉnʉk<sup>a</sup> *n*. **ruminate** ɨɲáɗʉ́tɛ́s *v*.; ɲɛɓɛ́s *v*. **rumple** imóɲíkees *v*. **rumple up** imóɲíkeetés *v*. **run** tsùwà *v*. **run (a direction)** ŋàtɔ̀n *v*. **run (multiply)** ŋatíón *v*. **run after** irukes *v*.; irúkésuƙot<sup>a</sup> *v*. **run after each other** ríínós *v*. **run away** duƙésúƙota mòrà<sup>e</sup> *v*.; moronuƙot<sup>a</sup> *v*.; ŋatɔnʉƙɔt<sup>a</sup> *v*. **run away (of many)** iɗúzòn *v*. **run cold (of blood)** ɓʉnʉ́mɔ́nà sèà<sup>e</sup> *v*. **run hot (of blood)** ɓʉnʉ́mɔ́nà sèà<sup>e</sup> *v*. **run into** íbaɗɛ́s *v*.; imánónuƙot<sup>a</sup> *v*.; ɲimánétòn *v*. **run into (meet)** imánétòn *v*. **run into repeatedly** íbaɗiés *v*. **run off** ikutses *v*.; ikútsésuƙot<sup>a</sup> *v*.; ŋatɔnʉƙɔt<sup>a</sup> *v*. **run out** ídzòn *v*. **run this way** ŋatɛ́tɔ́n *v*. **run-down** kɔlɔlánón *v*. **running (send off)** ŋatítɛ́sʉƙɔt<sup>a</sup> *v*. **running jump (get a)** itseɗítséɗòn *v*. **running naked** lèdèr *ideo*. **running water** lɔkáʝʉ́ *n*. **runt** ƙʉ́ƙ <sup>a</sup> *n*. **runty** sɛ́rɛ́ƙɛƙánón *v*. **rush** ɨɓʉ́ŋɔ́n *v*.; ɨkɨríkírɔ̀n *v*. **rush into things** ŋamɨŋámɔ́n *v*. **rush off** ɨpʉ́tɛ́sʉƙɔta así *v*. **rush out** tsídzètòn *v*. **rust** iróróòn *v*.; sɨmírɔ́n *v*. **rustle** ɓɛkɨɓɛ́kɔ́n *v*. **rustle up (food)** ɨɗɔ́ɗɔ́ɛ́s *v*.
**rusty** sɨmɨránón *v*. **rusty (get)** iɗolíɗólòn *v*. **Saba comorensis** ɲamalil *n*. **sack** hoɗésúƙot<sup>a</sup> *v*. **sack (gunny)** ɲɛ́ɗɛpɨɗɛ́p <sup>a</sup> *n*.; ɲéɡuniyá *n*. **sack (huge leather)** tun *n*. **sack (large gunny)** lomóŋin *n*.; ɲáwaawá *n*. **sack (leather)** lokóoɗo *n*. **sack (nylon)** ɲɛ́kɨsɛsɛ́*n*. **sack (small plastic)** ɲápaalí *n*. **sacrament** ɲásakaraméntù *n*. **sacred place** ɲɛkíwɔ́rìt<sup>a</sup> *n*. **sacrifice** síts'<sup>a</sup> *n*. **sacrifice against enemies** loŋɔ́tásìts'<sup>a</sup> *n*. **sacrifice (funeral goat)** ipúɲéés *v*.; sɛ́ɛ́s *v*. **sacrifice (wedding)** ɲɛ́kílama *n*.; ɲɛkʉma *n*. **sacrifice a goat** cɛɛsá rié sàbàk<sup>e</sup> *v*. **sacrum** ɲɛtsɨr *n*.; ɲɛtsɨríɔ́k <sup>a</sup> *n*. **sad** itásónòn *v*.; sìŋòn *v*.; tasónón *v*. **saddle (donkey)** ɲásaaʝ<sup>a</sup> *n*. **saddle (of a mountain)** kwaréékw<sup>a</sup> *n*. **safari ant** ƙúduƙûd<sup>a</sup> *n*. **safe** ɲébeŋɡí *n*.; toikíkón *v*. **safe-box** ɲébeŋɡí *n*. **safety (gun)** bɔɗɔ́k <sup>a</sup> *n*.; ŋáɲɛ́sìàw<sup>a</sup> *n*. **safety pin** ɲákwác *n*. **sag** ɨƙɔ́nɔ́nɔ̀ɔ̀n *v*. **sag (of eyelids)** irwápón *v*. **saggy** ratatáɲón *v*. **sail** ɨɔ́ɔ́rɛ́s *v*.; ɨɔ́ɔ́rɔ̀n *v*. **saintly** dòòn *v*. **saliva** tat<sup>a</sup> *n*. **salivary glands** ƙuts'áts'íka ni tatí *n*. **salivate** mʉ́lʉƙʉ́ƙɔ́n *v*.
#### saloon
**saloon** ɲáɓá *n*. **salt** didiɡwarí *n*.; ɲémíli *n*. **Salvadora persica** ɓaláŋ *n*. **same** iríánòn *v*. **same (be the)** ikwáánòn *v*. **sample** kaites *v*. **sample from** ɨsɛkísɛ́kɛ́s *v*.; ɨsɛ́sɛ́kɛ́s *v*. **sample many** hamomos *v*. **sanctity** daás *n*. **sand** ʝʉmʉʝʉmás *n*.; séɓés *v*.; sʉ́ʉ́tɛ́s *v*. **sand spring** ɲakúʝá *n*. **sandal** ŋaɗɛ́tá *n*. **sandal (rubber)** ɲásánɗɔ̀l *n*. **Sansevieria robusta** màlòr *n*. **Sansevieria species** barat<sup>a</sup> *n*.; ʝɔ̂d <sup>a</sup> *n*. **sanza** lokemú *n*. **sapless** bʉláʝámɔ̀n *v*.; daƙwádòn *v*. **Satan** Siitán *n*. **sate** cɨɨtɛ́sʉƙɔt<sup>a</sup> *v*. **sated** cìɔ̀n *v*. **sated (become)** cɨɔnʉƙɔt<sup>a</sup> *v*.; topwatímétòn *v*. **satellite** ɗɔ́xɛatá na ɓɛƙɛ́s *n*. **satiate** cɨɨtɛ́sʉƙɔt<sup>a</sup> *v*. **satiated** cìɔ̀n *v*. **satiated (become)** cɨɔnʉƙɔt<sup>a</sup> *v*.; topwatímétòn *v*. **satisfied** cìɔ̀n *v*. **satisfy** cɨɨtɛ́sʉƙɔt<sup>a</sup> *v*. **satisfy hunger** topwátón *v*. **satisfy meat hunger** ɨtsɔ́ítɛ́sʉƙɔt<sup>a</sup> *v*.; ɨtsɔ́ɔ́n *v*.; ɨtsɔ́ɔ́nʉƙɔt<sup>a</sup> *v*. **saturated** ilébìlèbètòn *v*. **Saturday** Ɲárámɨram *n*. **Satureja species** òŋòrìkwàts<sup>a</sup> *n*.
**saucepan** dóm *n*.; ɲásipiryá *n*.; ɲésipiriyá *n*. **saucepan (small)** dómáìm *n*. **saucer** ɲásaaní *n*. **saunter** ɨpɛ́ɛ́ɲɛ́sá así *v*. **sausage tree** sosóbòs *n*. **savage** ɦyɛtɨɦyɛtɔs *v*.; ɦyɛ̀tɔ̀n *v*.; ɨɲɛ́ɛ́mɔ̀n *v*.; ɨsílíánón *v*. **savagery** ɦyɛtás *n*. **savannah** dús *n*. **save** ɡirés *v*.; ɨɛtɛ́s *v*. **save (spiritually)** hoɗetés *v*. **save oneself** ɨɛtɛ́sá así *v*. **saved (get)** hoɗetésá así *v*. **savior** hoɗetésíàm *n*.; ɨɛtɛ́síàm *n*. **saw** iriƙíríƙés *v*.; ɲɛ́ƙɨrɨƙír *n*. **saw away at** ifitífítés *v*. **say** kʉ̀tɔ̀n *v*.; kʉtɔnʉƙɔt<sup>a</sup> *v*.; tódètòn *v*. **say hello to** ɨmáxánɛ́s *v*. **saying** taɗápítotós *n*. **scab** ɔmɔ́x *n*. **scabby** sómomóʝón *v*.; tɔmɔ́tɔ́mánón *v*. **scabby person** tɔmɔ́tɔ́mánìàm *n*. **scabies** kɔ́ts<sup>a</sup> *n*. **scaffolding (make)** ɨpɛ́tɛ́ɛ́s *v*. **scald** kʉpɛ́s *v*.; kʉpɛ́sʉ́ƙɔt<sup>a</sup> *v*. **scale** fâd<sup>a</sup> *n*.; ikókórés *v*.; kɔmɔ́m *n*. **scale (weighing)** ɲératíl *n*. **scale this way** ikókóretés *v*. **scales** ɲératíl *n*. **scaly** saŋáŋóòn *v*. **scamper up** ɨtsɛ́tsɛ́ɛ́s *v*. **scapula** sawatɔ́ɔ́k <sup>a</sup> *n*. **scar** ƙwár *n*.; ɔ́ʝátàs *n*.; tás *n*.; tásɛ̂d <sup>a</sup> *n*. **scar (big)** ɲépóros *n*.
**scarce** búúbuanón *v*.
**scare** kitítésuƙot<sup>a</sup> *v*.; xɛɓɨtɛs *v*.; xɛɓɨtɛ́sʉ́ƙɔt<sup>a</sup> *v*.
**scare away** ɨrɛmɛs *v*.
**scare off** ɨrɛmɛs *v*.
**scarified** ɨsɛɓɔs *v*.
**scarify** ɓunutiés *v*.; ɨsɛɓɛs *v*.; ɨsɛɓísɛ́ɓɛ́s *v*.
**scarlet** tsòn *ideo*.
**scarred** ɓulúrúmòn *v*.
**scarred up** seɓuránón *v*.
**scat** ets'<sup>a</sup> *n*.
**scatter** ɓʉnʉ́mɔ́n *v*.; ɓʉnʉtɛ́s *v*.; ɨɗɛrɛs *v*.; itwares *v*.; ɨwɛ́ɛ́lánón *v*.; ɨwɛ́ɛ́lɛ́s *v*.; ɨwɛ́ɛ́lɛ́sʉƙɔt<sup>a</sup> *v*.; ɨwɛ́ɛ́lɛtɛ́s *v*.; toɓwaŋes *v*.
**scatter (seeds)** tɛwɛɛs *v*.
**scatter around** ɨɗɛríɗɛ́rɛ́s *v*. **scattered** ɨɗɛrɔs *v*.; ɨwɛ́ɛ́lɔ́s *v*.; kazaanón *v*. **scattered around** apɛ́tɛ́pɛ́tánón *v*.; ɨɗɛríɗɛ́rɔ́s *v*. **scavenge** furés *v*.; iúréés *v*. **scavenger** furésíàm *n*. **scent** kɔín *n*.
**schlip!** ʝʉ́rʉ́tᶶ *ideo*.; sɛ̀lɛ̀t ɛ *ideo*.
**school** ɲésukúl *n*.
**school (technical)** tɛ́kɛ̀nìkɔ̀l *n*.
**school (vocational)** tɛ́kɛ̀nìkɔ̀l *n*.
**schooling** ɲósomá *n*.
**schunk!** pùrùs *ideo*.
**science** ɲɛ́sɛ́ànìs *n*.
**scissors** ɲámakás *n*.
**Sclerocarya birrea** ts'ɔƙɔ́m *n*.
**scold** dɔxɛ́s *v*.; dɔxɛ́sʉ́ƙɔt<sup>a</sup> *v*.
**scoliosis (have)** toíɗón *v*.
**scoop** cɛbɛn *n*.; tɛ́bɛ̀s *v*. **scoop off** ɨlaɓɛtɛ́s *v*. **scoop out (water)** ɗalés *v*. **scoop out/up** tɛ́bɛtɛ́s *v*. **scoop up** cɛɓɛ́s *v*.; towoɗetés *v*. **scoop with fingers** ɡafariés *v*. **scoot** béberésuƙota así *v*. **scorch** kɔrɛtɛ́s *v*.; kɔrɨtɛtɛ́s *v*. **scorched** kɔrɛ́tɔ́n *v*. **score** ɨsɛɓɛs *v*. **score (a goal)** iƙólésuƙot<sup>a</sup> *v*. **scored** ɨsɛɓɔs *v*. **scorn** míʝés *v*.; tsíítés *v*. **scorpion** lóɗíkór *n*. **scorpion (water)** lòcòrò *n*. **scorpion herb** lóɗíkórócɛmɛ́r *n*. **scour (an area)** ɨɗɛŋɛs *v*. **scoured** iɗéŋímètòn *v*. **scourge** iɓúŋéés *v*. **scout** irimesíám *n*.; rɔtɛ́ám *n*. **scout bee** páupáw<sup>a</sup> *n*. **scout out** irimírímés *v*. **scowl** iɲíkón *v*. **scowling (begin)** iɲíkétòn *v*. **scraggy** ɡɔ́ɡɔ̀rɔ̀mɔ̀n *v*. **scramble** ɨfáfáɲɛ́s *v*.; ɨlɛ́pɔ́n *v*. **scramble down** kukuanón *v*. **scramble up** ɨfɛ́ɗɛ́lɛ́s *v*.; ɨlɛ́pɛ́sʉƙɔt<sup>a</sup> *v*.; ɨmɔrímɔ́rɛ́s *v*. **scrambled up** ɨmɔrímɔ́rɔ́s *v*. **scrap** ƙwɛsɛ́*n*. **scrap metal (piece of)** ɲɛpɛlɛrɛŋ *n*. **scrape** fɔ́fɔ́tɛ́s *v*.; ɨfɔɛs *v*.; wówóʝés *v*. **scrape clean** tikitetés *v*. **scrape off** bátsɛ́s *v*.; ɨwalɛtɛ́s *v*.; rɛ́kɛ́s *v*.; tʉkʉrɛs *v*.
**scratch** ɡweɡweritiés *v*.; sʉ́ƙʉ́tɛ́s *v*.; sʉ́ʉ́tɛ́s *v*. **scratch (with claws)** soƙóríties *v*. **scratch off** sɛkɛ́s *v*.; sɛkɛ́sʉ́ƙɔt<sup>a</sup> *v*. **scratch up** ikúkúrés *v*.; tukures *v*.; tukutes *v*. **scratch vigorously** koxésúƙot<sup>a</sup> *v*. **scratched** ɡweɡweritiós *v*. **scrawl** ɡweɡweritiés *v*.; wíziwizetés *v*. **scrawled** ɡweɡweritiós *v*. **scrawny** kalɛ́ɛ́tsɛránón *v*.; rɛƙɛ́ɲɛ́mɔ̀n *v*. **scream** ɨkwílílɔ̀n *v*. **screech** iyíyéés *v*. **screw around on errand** ɗipímón *v*. **screw up** hamʉʝɛ́s *v*. **scribble** ɡweɡweritiés *v*.; wíziwizetés *v*. **scribbled** ɡweɡweritiós *v*. **scripture (Christian)** Ɲáɓáíɓɔ̀l *n*. **scrotal swelling** ɲɛkwɨ *n*. **scrounge for** furés *v*. **scrub** ríʝ<sup>a</sup> *n*.; sʉ́ƙʉ́tɛ́s *v*.; sʉ́ʉ́tɛ́s *v*. **scrub brush** ɲecaaƙo *n*. **scrub off** sɛkɛ́s *v*.; sɛkɛ́sʉ́ƙɔt<sup>a</sup> *v*. **scrubby** kalɛ́ɛ́tsɛránón *v*. **scrubland** ɲáɓwa *n*.; ríʝíkaaʝík<sup>a</sup> *n*. **scruff** fètìfèt<sup>a</sup> *n*. **scruff (fat)** nìtsìnìts<sup>a</sup> *n*. **scrumptious** ɗɔkɔ́dɔ̀n *v*.; ritídòn *v*. **scrumptiously** ɗɔ̀k ɔ *ideo*. **scrunch** ɨʉmʉ́íʉ́mɛ́s *v*. **scrunch up** ɨtʉsɛtɛ́s *v*.; tusuketés *v*.; tusúkón *v*. **scrutinize** ɨpíʝíkɛ́s *v*. **sculpt** bɛrɛtɛ́s *v*.; sotés *v*.; sotetés *v*. **scum** kɨrarap<sup>a</sup> *n*. **scurf** kɔmɔ́m *n*.
**scurfy** saŋáŋóòn *v*. **scurry up** ɨfɛ́ɗɛ́lɛ́s *v*.; sekweres *v*. **sea** ɲánam *n*. **seal** ilies *v*.; iliílíés *v*. **sealed** iliílíós *v*.; ilios *v*. **seamless** iliílíós *v*. **seamster** tʉfɛ́síàm *n*. **seamstress** tʉfɛ́síàm *n*. **sear** iɓues *v*. **search (an area)** ɨɗɛŋɛs *v*. **search (pat down)** tárábes *v*. **search all over (pat down)** tárábiés *v*. **search for** bɛ́ɗɛ́s *v*.; bɛɗɛtɛ́s *v*.; ɨkʉʝɛs *v*. **search in vain** iróʝíés *v*. **searched over** iɗéŋímètòn *v*. **season** ɛfɨtɛs *v*.; íbutsurés *v*.; iwéwérés *v*.; tsóy<sup>a</sup> *n*. **season (dry)** ôdz<sup>a</sup> *n*.; ódzatsóy<sup>a</sup> *n*. **season (rainy)** diditsóy<sup>a</sup> *n*.; ɔtáy<sup>a</sup> *n*. **seasoning** ɲɛ́ɓɨsár *n*. **seat** kàràts<sup>a</sup> *n*.; zɛƙɔ́áw<sup>a</sup> *n*. **seated** zɛ̀ƙwɔ̀n *v*. **seated (of many)** ɡóƙón *v*. **sebum** îdw<sup>a</sup> *n*. **secede** tatsáɗón *v*. **seclude** ɨpátsɛ́sʉƙɔt<sup>a</sup> *v*. **seclude oneself** ɨpátsɛ́sʉƙɔta así *v*. **second (be the)** mɨtɔna ɗíɛ́leɓétsónì *v*. **second (one)** ɗa leɓétsónì *pro*. **secret** búdòs *v*. **secretary** karan *n*. **secrete** ɓɔ́rítɔ̀n *v*. **secretor** ƙùts'àts'<sup>a</sup> *n*. **section** bácík<sup>a</sup> *n*.; itiɓes *v*.; itiɓítíɓés *v*.; ʝulés *v*.; ƙɔ́dɔ̀l *n*. **section (area)** nabɨɗɨt<sup>a</sup> *n*. **section (military)** ɲésékíxìòn *n*.
**section (plant)** ɲékel *n*. **section (space)** naƙʉ́lɛ́*n*. **secure** toikíkón *v*. **Securinega virosa** ɲalakas *n*. **security** ŋíkísila *n*.; ɲɛkɨsɨl *n*. **security officer (government)** tirifiesíáma ɲápukání *n*. **seduce** sʉ́bɨtɛ́sʉ́ƙɔta así *v*. **see** enés *v*.; enésúƙot<sup>a</sup> *v*.; walámón *v*. **see stars** ɨmɛ́ɗɛ́tɔna ekwí *v*. **see-through** tsaórómòn *v*. **seed** eɗed<sup>a</sup> *n*.; eɡésá ekwí *v*.; ekw<sup>a</sup> *n*.; ekwed<sup>a</sup> *n*.; iŋárúrètòn *v*. **seed butter** ɲówoɗí *n*. **seed mixture (nuptial)** loŋazut<sup>a</sup> *n*. **seed oil** útɔ̀ *n*. **seed(s)** kiɲom *n*. **seedeater (yellow-rumped)** ʝɨlíw<sup>a</sup> *n*. **seeded** iŋárúròn *v*. **seeds** ekwin *n*. **seeds (have)** iŋárúròn *v*. **seeds of ʝàw<sup>a</sup>** ílekó *n*. **seeing as how** naítá *subordconn*. **seek** bɛ́ɗɛ́s *v*.; bɛɗɛtɛ́s *v*.; ɨkʉʝɛs *v*.; ɨmítíŋɛɛ́s *v*. **seemingly** íkwà *adv*.; ókò *adv*. **seen** lɛ́lɔ́n *v*.; takánón *v*. **seen (make)** kɛtɛ́lɨtɛtɛ́s *v*. **seen clearly** ilééránón *v*.; kɛtɛ́lɔ́n *v*. **seen dimly** misimísón *v*. **seen faintly** misimísón *v*. **seep** tɔfɔ́ɗɔ́n *v*. **seer** ɲakuʝíícíkáàm *n*. **seesaw** iyopíyópòn *v*. **seethe** tabúón *v*. **seethe over** tabúétòn *v*.
**segment** itiɓes *v*.; itiɓítíɓés *v*.; ʝulam *n*.; ʝulés *v*. **segment (plant)** ɲékel *n*. **segment (small)** ʝulamáím *n*. **segregate** ɨlɔ́ɗíŋɛ́s *v*.; tereties *v*. **segregated** teretiós *v*. **segregation** ɲoloɗiŋ *n*. **segregative** ɨlɔ́ɗíŋánón *v*. **seism** ɲéríkirik<sup>a</sup> *n*. **seize** ɛ́nɛ́sʉƙɔt<sup>a</sup> *v*.; ɨkamɛ́sʉ́ƙɔt<sup>a</sup> *v*.; ɨrakiesúƙota así *v*.; ɨrákímétòn *v*.; ɨrɛɗɛs *v*.; tokopes *v*.; tokópésuƙot<sup>a</sup> *v*. **seize frequently** reŋíónuƙot<sup>a</sup> *v*. **seizure (have a )** ɨrákímétòn *v*. **seizures (have)** ɨrakiesúƙota así *v*.; reŋíónuƙot<sup>a</sup> *v*. **Selaginella phillipsiana** múmùt<sup>a</sup> *n*. **select** ɗʉmɛtɛ́s *v*.; iɗókóliés *v*.; ɨƙɛ́ƙɛ́ɛ́s *v*.; ɨƙɛlɛs *v*.; ɨƙɛlɛtɛ́s *v*.; ƙɛ́lɛ́s *v*.; tɔsɛɛtɛ́s *v*.; xɔ́bɛtɛ́s *v*. **select categorially** ɨsíílɛtɛ́s *v*. **select iteratively** ƙélíetés *v*. **selected** xɔ́bɔtɔ́s *v*. **self** as *pro*.; nêb<sup>a</sup> *n*. **self-centered** reídòn *v*. **self-cleanse** fítésuƙota ɡúró<sup>e</sup> *v*. **self-controlled** ɨrɨtsɛ́sá así *v*.; toikíkón *v*. **self-important** ɨwɔ́ƙɔ́n *v*. **selfish** hábòn *v*. **selfish person** kìɓèɓèàm *n*. **selfishness** hábàs *n*. **sell** dzíɡwès *v*.; dzíɡwesuƙot<sup>a</sup> *v*. **seller** dzíɡwesuƙotíám *n*. **semen** ɗír *n*. **semester** ɲátám *n*. **seminar** ɲésémìnà *n*. **send** eréɡes *v*. **send back** raʝésúƙot<sup>a</sup> *v*.
**send early** isókítésuƙot<sup>a</sup> *v*. **send in a message** mɛnɛ́tɔ́n *v*. **send off running** ŋatítɛ́sʉƙɔt<sup>a</sup> *v*. **send out a message** mɛnɔnʉƙɔt<sup>a</sup> *v*. **send soaring** ɨɔ́ɔ́rɛ́s *v*. **send straight to** tɔɓɛɨtɛtɛ́s *v*. **senile** dúnésòn *v*.; itúléròn *v*.; kamudurudádòn *v*. **senile (become)** itúléronuƙot<sup>a</sup> *v*. **sense** ƙanetés *v*. **sensitive (to light)** tɔtsɔ́ɔ́n *v*. **sentry** itelesíám *n*.; itelesíáma kíʝá<sup>e</sup> *n*. **separate** ɗusés *v*.; ɗusésúƙot<sup>a</sup> *v*.; terémétòn *v*.; terémón *v*.; terémón *v*.; terémónuƙot<sup>a</sup> *v*.; terés *v*.; tereties *v*. **separate by shaking** ɨkákɛ́ɛ́s *v*. **separate oneself** terésúƙota así *v*. **separate out** terétéránitésúƙot<sup>a</sup> *v*. **separate out by shaking** ɨkákɛ́ɛ́sʉƙɔt<sup>a</sup> *v*. **separated** teretiós *v*. **separated out** terétéránón *v*. **September** Lotyak<sup>a</sup> *n*.; Nakariɓ<sup>a</sup> *n*. **sequester** ɨpátsɛ́sʉƙɔt<sup>a</sup> *v*. **sequester oneself** ɨpátsɛ́sʉƙɔta así *v*. **serial killer** ɨkɛɲíkɛ́ɲɛ́síàm *n*. **serious (be)** mɨtɔna sírìàs *v*. **seriously** ɓa<sup>u</sup> *ideo*. **serous fluid** tsét<sup>a</sup> *n*. **serpent** ídèm *n*. **serum** tsét<sup>a</sup> *n*. **serval** ɲálukutúʝu *n*. **servant** ŋípáƙásíàm *n*. **servant (domestic)** teréɡiama awá<sup>e</sup> *n*. **servant (indentured)** ŋiléɓúìàm *n*. **serve (food)** ɡárés *v*. **service** ɨsáɓísɨŋɛɛ́s *v*.; terêɡ<sup>a</sup> *n*.
**service (religious)** wáán *n*. **sesame** kaɲʉm *n*. **Sesamum indicum** kaɲʉm *n*. **set (joint)** raʝés *v*. **set (of sun)** itsólóŋòn *v*.; tɔɔnʉƙɔt<sup>a</sup> *v*. **set agape** hádoletés *v*. **set aside** ɨŋáɗɛ́ɛ́s *v*.; ɨŋáɗɛ́ɛ́sʉƙɔt<sup>a</sup> *v*. **set fire to** ɗamatés *v*. **set free** hoɗés *v*.; hoɗésúƙot<sup>a</sup> *v*.; hoɗetés *v*. **set loose** itsues *v*.; itsuetés *v*. **set nearby** taraŋés *v*. **set oneself apart** ƙɛ́lɛ́sʉƙɔta así *v*.; terésúƙota así *v*. **set out** ɗóɗésa muceé *v*. **set record straight** ɨtɛ́nɨtɛtɛ́sá tódà<sup>e</sup> *v*. **set straight** ɨtɛ́nɨtɛtɛ́s *v*. **set up** ɨnábɛtɛ́s *v*. **set up (a beehive)** rɔ́ƙɛ́s *v*. **set up (gel)** tɔsɔ́ɗɔ́kɔ̀n *v*. **set upright** ɨtsírítɛtɛ́s *v*.; tsírítɛtɛ́s *v*. **settle** ínésuƙot<sup>a</sup> *v*.; zɛƙwɛ́tɔ́n *v*. **settle a dispute** epitésúƙota tódà<sup>e</sup> *v*. **settle down** ɗipímón *v*.; epítésuƙot<sup>a</sup> *v*.; toíésuƙot<sup>a</sup> *v*.; zɛƙwɛ́tɔ́n *v*.; zɛƙwɨtɛtɛ́s *v*. **settle in** toƙízeesá así *v*.; toƙízèètòn *v*. **settle on** ɲʉmɛtɛ́s *v*. **seven** tude ńda kiɗi léɓèts<sup>e</sup> *num*. **seven o'clock** ɲásáatíá kɔ̀nìk<sup>ɛ</sup> *n*. **seventeen** toomíní ńda kiɗi túde ńda kiɗi léɓèts<sup>e</sup> *n*. **seventy** toomínékwa túde ńda kiɗi léɓèts<sup>e</sup> *n*. **sever** ɗusés *v*.; ɗusésúƙot<sup>a</sup> *v*.; ɗusutes *v*.; ɨkɛ́ŋɛ́ɗɛ́s *v*.; tɔŋɛɗɛs *v*. **several** ʝaláʝálánón *v*. **sew** tʉfɛ́s *v*. **sewing machine** ɲájarán *n*.
**sex** ep<sup>a</sup> *n*. **sex (have frequent)** ɓútánés *v*. **sex (have unprotected)** ts'íts'ɔ́n *v*. **sex (have)** èpòn *v*. **sexual afterglow (feel)** ɨrákímétòn *v*. **sexual intercourse** ep<sup>a</sup> *n*. **sexual relations** ep<sup>a</sup> *n*. **sexually insatiable (of a woman)** ɓòɓòn *v*. **sexually-transmitted disease** ɲamakaʝe *n*. **shade** kur *n*.; kúrúkúr *n*. **shadeless** sáɡwàràmòn *v*. **shades (glasses)** ɲékiyóika ni fetí *n*. **shadow** ikókótés *v*.; kúrúkúr *n*. **shaft** morók<sup>a</sup> *n*. **shaft (arrow)** ɲámalídàkw<sup>a</sup> *n*. **shaft (borehole)** ɲatsʉʉmáhò *n*. **shaft (penile)** ɲɛ́sɛɛɓɔ́*n*. **shaft of light** bás *n*.; sʉ́w<sup>a</sup> *n*. **shaggy** ɡaúsúmòn *v*. **shake** iɓokes *v*.; kìtòn *v*.; kwalíkwálɔ̀n *v*. **shake (begin to)** kitétón *v*. **shake (make)** kitítésuƙot<sup>a</sup> *v*. **shake back and forth** ɨlílíŋɛ́s *v*.; ɨlɨŋílíŋɛ́s *v*.; ilitsílítsés *v*. **shake in a pan** ɨlaƙɛs *v*.; ɨlaƙíláƙɛ́s *v*.; ɨláláƙɛ́s *v*. **shake off** iɓókésuƙot<sup>a</sup> *v*.; ilílítsés *v*.; iwatíwátés *v*. **shake out** iɓutúɓútés *v*.; ilílítsés *v*.; ixóxóƙés *v*.; ixuƙúxúƙés *v*.; ixúxúƙés *v*. **shake out noisily** ɨxaxɛɛs *v*. **shake side to side** ilitsílítsés *v*. **shake to separate** ɨkákɛ́ɛ́s *v*.; ɨkákɛ́ɛ́sʉƙɔt<sup>a</sup> *v*.
**shake up and down** itéƙítéƙés *v*. **shake vigorously** íbɔbɔtsɛ́s *v*.; íbɔtsɛ́s *v*. **shaking sound** ɓʉlʉɓʉl *ideo*. **shallow** tɛ́ƙɛ́dɛ̀mɔ̀n *v*.; tɛ́ƙɛ́zɛ̀mɔ̀n *v*.; tiƙódzòmòn *v*. **shallowly concave** ɓɛtɛ́lɛ́mɔ̀n *v*.; fɛtɛ́lɛ́mɔ̀n *v*. **shame** ɓets'itetés *v*.; iryámítetésá ŋiléétsìk<sup>e</sup> *v*.; ŋilééts<sup>a</sup> *n*. **shameful person** ŋiléétsìàm *n*. **shamefulness** ŋiléétsìnànès *n*. **shape** bɛrɛtɛ́s *v*.; itues *v*.; ituetés *v*. **shape (with a blade)** bɔ́tɛ́s *v*. **shard** ɡúɗúsam *n*. **share** tɔmɔram *n*.; tɔmɔrɛs *v*. **share with each other** tɔmɔ́rínɔ́s *v*. **shareable** tɔmɔram *n*. **sharp** ts'íts'ɔ́n *v*. **sharp (of eyesight)** tsɔ́tsɔ́n *v*. **sharp (make)** ts'íts'ítɛ́sʉƙɔt<sup>a</sup> *v*.; ts'íts'ítɛtɛ́s *v*. **sharp in taste** ɓariɓárón *v*.; ɓárikíkón *v*. **sharpen** banɛ́s *v*.; ts'íts'ítɛ́sʉƙɔt<sup>a</sup> *v*.; ts'íts'ítɛtɛ́s *v*. **shart** ɨɓíɔ́n *v*. **shatter** ɓɨlíɓílɛ́s *v*.; kwɛts'ɛ́s *v*. **shattered (get)** kwɛts'ɛ́mɔ́n *v*. **shave** bɔ́tɛ́s *v*.; ɨpɛlɛs *v*. **shave (even)** ɨkʉlɛs *v*.; ɨkʉlɛtɛ́s *v*.; ɨkwalɛs *v*. **shave (hair)** ɡíʝɛ́s *v*. **shave off** ɨƙwɛ́ƙwɛ́rɛ́s *v*. **shave off (hair)** ɡɨʝɛtɛ́s *v*. **shaving** ɨpɛlɛtam *n*. **shaving (wood)** bɔtɛtam *n*. **shawl** ɲáléso *n*.; ɲáwáro *n*. **shawl (cotton)** ɲákamarɨkán *n*. **shawl (leather)** xɔŋɔŋ *n*.
#### shortness
#### sheaf
**sheaf** zɨkam *n*. **sheaf (of crops)** ɲénéne *n*. **sheath** nakɨrɔ́r *n*.; ɲaɓʉ́rɛ́t <sup>a</sup> *n*. **shed** fòlòn *v*.; ɲésitó *n*.; tuɓutes *v*.; tuɓútésuƙot<sup>a</sup> *v*. **shed blood** tɔyɔ́ɔ́n *v*. **shedding** ɗaráɗáránón *v*. **sheep** ɗóɗò *n*. **sheep tail** ɗóɗotimóy<sup>a</sup> *n*. **sheep-leather clothing** ɗóɗòƙwàz *n*. **sheep-leather skirt** ɗóɗòƙwàz *n*. **sheet of paper** kàbàɗ<sup>a</sup> *n*. **sheet of roofing** ɲámaamɓát<sup>a</sup> *n*. **shell** fâd<sup>a</sup> *n*. **shell (casing)** ɲéɓurocó *n*.; ɲɛsɛpɛɗɛ *n*. **shell (of a beehive)** ɗòl *n*. **shell (snail)** irex *n*.; tɔƙɔtɔƙáhò *n*. **shell (tortoise)** ròɡìròɡ<sup>a</sup> *n*. **shelter** kur *n*.; rɨmɛ́s *v*. **shelter (grass)** ɲɛ́kɨsakát<sup>a</sup> *n*. **shelter from rain** rɨmɛ́sá dìdìù *v*. **shepherd** còòkààm *n*.; cookés *v*. **shield** ɨƙɨɛs *v*.; kesen *n*. **shift** iméérés *v*.; irotes *v*.; ɨsʉ́tɔ́n *v*. **shift (position)** ɨsɔ́sɔ́ŋɔ́s *v*. **shift repeatedly** irotírótés *v*. **shilling** kaúdzèèkw<sup>a</sup> *n*.; ŋásɛntáìèkw<sup>a</sup> *n*. **shillings** ŋásɛntáy<sup>a</sup> *n*. **shimmer** riɓiríɓón *v*. **shin** ɓɔ́l *n*. **shinbone** tsɛrɛ́k <sup>a</sup> *n*. **shindig** ɲápáti *n*. **shine** ɨwírɔ́n *v*. **shine (begin to)** ɨwírɛ́tɔ̀n *v*. **shine brightly** ɨraírɔ́ɔ̀n *v*.
**shine forth (of heavenly bodies)** ʝʉ́ɛ́tɔ̀n *v*. **shinny up** ɨfɛ́ɗɛ́lɛ́s *v*. **shiny** pirídòn *v*. **ship** irotes *v*.; tsídzès *v*. **ship away** tsídzesuƙot<sup>a</sup> *v*. **ship off** tsídzesuƙot<sup>a</sup> *v*. **ship repeatedly** irotírótés *v*. **shirt** ɲásáti *n*. **shish kebab** rɔam *n*. **shit** nts'áƙón *v*. **shitǃ** ɗʉ̀rʉ̀ *n*. **shiver** kìtòn *v*.; kwalíkwálɔ̀n *v*. **shiver (begin to)** kitétón *v*. **shiver (make)** kitítésuƙot<sup>a</sup> *v*. **shock** ɨɓálɛ́tɔ̀n *v*.; lilétón *v*.; toɓules *v*. **shocking** ɨɓálɔ́n *v*.; toɓúlón *v*. **shoe** taƙáy<sup>a</sup> *n*. **shoe (cow-leather)** ɦyɔtaƙáy<sup>a</sup> *n*. **shoe (elephant leather)** oŋoritaƙáy<sup>a</sup> *n*. **shoe (leather)** ŋáɓʉʉrá *n*. **shoe (open-toed)** taƙáá na ŋáɲɔ́s *n*. **shoe (tire rubber)** ɲómotokátáƙáy<sup>a</sup> *n*. **shoe-strap** lɔkaapín *n*. **shoelace** lɔkaapín *n*. **shoot** ídzès *v*.; ídzesuƙot<sup>a</sup> *v*.; ƙádès *v*.; ƙádesuƙot<sup>a</sup> *v*. **shoot across** ídzesa así *v*. **shoot over** ídzesa así *v*.; toɓésá así *v*. **shoot repeatedly** ídziidziés *v*.; ƙádiƙadiés *v*. **shop** ɗʉkán *n*.; ɲáɗʉkán *n*. **short** kúɗón *v*.; ŋʉɗʉ́sʉ́mɔ̀n *v*.; ŋʉsʉ́lʉ́mɔ̀n *v*. **short (make)** kuɗítésuƙot<sup>a</sup> *v*. **short (of many)** kúɗaakón *v*. **shorten** kuɗítésuƙot<sup>a</sup> *v*. **shortness** kuɗás *n*.
**shorts** ɲosoƙoloké *n*. **shorts (pair of)** ɲésiriwáli *n*. **shortsighted** mumúánón *v*. **should** ɨtámáánón *v*. **shoulder** sawat<sup>a</sup> *n*. **shoulder bone** sawatɔ́ɔ́k <sup>a</sup> *n*. **shout** bofétón *v*.; bófón *v*.; ɨkílɔ́n *v*.; iƙúétòn *v*.; iƙúón *v*.; iƙúónuƙot<sup>a</sup> *v*.; nɔsátón *v*. **shout at** ɨyáyɛ́ɛ́s *v*. **shout triumphantly** iwóŋón *v*. **shouter** nɔ̀sààm *n*. **shouting** nɔ̀s *n*. **shove** ɲékitiyó *n*.; ɲɛɲɛrɛs *v*.; toremes *v*. **shove away** iɓwátésuƙot<sup>a</sup> *v*.; ɨtsɔ́rɛ́sʉƙɔt<sup>a</sup> *v*. **shovel (power-)** ɲɛ́tɛrɛƙɨtaa na kwɛtá<sup>ɛ</sup> *n*. **show** ɗóɗés *v*.; ɗoɗésúƙot<sup>a</sup> *v*.; ɗóɗítetés *v*.; enitésúƙot<sup>a</sup> *v*.; enitetés *v*.; ilééránitetés *v*.; itétémés *v*. **show appreciation to caregiver(s)** taatsesa ɗoɗóbò<sup>e</sup> *v*. **show favoritism** tereties *v*. **show hospitality to** ewanes *v*.; ewanetés *v*. **show off** inésóòn *v*. **show oneself** ɗóɗítetésá así *v*. **show to be wrong** ɨsalɨtɛ́s *v*. **show up** takánétòn *v*. **show up unwelcomely** imíŋóòn *v*. **shower** féíàw<sup>a</sup> *n*.; féón *v*. **shower (bombard)** ídʉrɛ́s *v*. **showering** féy<sup>a</sup> *n*. **shred** dzɛrɛ́s *v*.; dzeretiés *v*.; dzeretiésuƙot<sup>a</sup> *v*.; kàbàɗ<sup>a</sup> *n*. **shredded** dzɛ́rɛ́dzɛránón *v*.; dzɛrɔ́sɔ́n *v*.; láládziránón *v*.
**shrew** ɗɔ́f *n*. **shrewd** nɔɔsánón *v*. **shrewdness** nɔɔ́s *n*. **shriek** ɨkwílílɔ̀n *v*.; iyíyéés *v*. **shrike** bílɔɔrɔ́*n*. **shrike (white-crested)** kíyɔɔrɔ́*n*. **shrill** ɓòɓòn *v*. **shrimpy** sɛ́rɛ́ƙɛƙánón *v*. **shrink** kɨɗɔnʉƙɔt<sup>a</sup> *v*.; tɔ́ɗɔ́nʉƙɔt<sup>a</sup> *v*. **shrink back** rìmòn *v*. **shrink back from** itírákés *v*. **shrink down** kwatsítésuƙot<sup>a</sup> *v*.; kwátsónuƙot<sup>a</sup> *v*. **shrink up** hɛ́ɗɔ́nʉƙɔt<sup>a</sup> *v*. **shrivel** kɨɗɔnʉƙɔt<sup>a</sup> *v*.; tɔ́ɗɔ́nʉƙɔt<sup>a</sup> *v*. **shrivel up** hɛ́ɗɔ́nʉƙɔt<sup>a</sup> *v*.; lolómónuƙot<sup>a</sup> *v*. **shriveled** bɔrɔ́ɗɔ́mɔ̀n *v*.; ƙɔ́rɔmɔmɔ́n *v*.; mɨtírímɔ̀n *v*.; tɔ́ɗɔ́n *v*. **shrub** dakw<sup>a</sup> *n*. **shrub species** alárá *n*.; ɓets'akáw<sup>a</sup> *n*.; ɓóéɗ<sup>a</sup> *n*.; dìdì *n*.; ɡɛbɛʝ<sup>a</sup> *n*.; ɡomóí *n*.; ikitínícɛmɛ́r *n*.; ʝàw<sup>a</sup> *n*.; lócén *n*.; loʝeméy<sup>a</sup> *n*.; lorít<sup>a</sup> *n*.; lóúpè *n*.; marʉ́kʉ́cɛmɛ́r *n*.; mét<sup>a</sup> *n*.; milékw<sup>a</sup> *n*.; mɨsá *n*.; mɨsíás *n*.; mɔ̀z *n*.; múrotsíò *n*.; ɲɛ́rɨkɨrík<sup>a</sup> *n*.; ɲónomokére *n*.; ɔɡɔn *n*.; suɡur *n*.; súr *n*.; ts'ʉ́ɡʉram *n*.; tâb<sup>a</sup> *n*.; tíkòŋ *n*.; turunet<sup>a</sup> *n*. **shrug** ɨmímíʝɛ́s *v*. **shrunk** tɔ́ɗɔ́n *v*. **shrunken** bɔrɔ́ɗɔ́mɔ̀n *v*. **shuck** poɗés *v*.; poɗetés *v*. **shudder** nɛ́rɨnɛ́rɔ́n *v*.; tsábatsabánón *v*. **shuffle** iyaŋíyáŋés *v*. **shush** ɦyakwés *v*. **shut** kɔkɛ́s *v*.; kɔkɛtɛ́s *v*.; kokímétòn *v*.; mʉts'ʉtɛs *v*. **shut (make)** kɔkɨtɛtɛ́s *v*.
**shut down** kɔkɛtɛ́s *v*. **shut oneself in** kɔkɛ́sá así *v*. **shut out** kɔkɛ́sʉ́ƙɔt<sup>a</sup> *v*. **shut up** iɓótóŋésá aká<sup>e</sup> *v*.; ɨʝɛ́mítɛ́sʉƙɔt<sup>a</sup> *v*.; ɨʝɛ́mɔ́nʉƙɔt<sup>a</sup> *v*.; mʉts'ʉ́tɛ́sʉƙɔt<sup>a</sup> *v*. **shut up (lock)** iɓótóŋés *v*. **shut up repeatedly** muts'utiesúƙot<sup>a</sup> *v*. **shy** xɛ̀ɓɔ̀n *v*. **shy person** xɛɓásíàm *n*. **shyness** xɛɓás *n*. **sibling** ŋɡóím *n*. **sick** mòòn *v*. **sick (of many)** mayaakón *v*. **sick (queasy)** itikítíkòn *v*. **sickle bush** ɡʉ̀r *n*. **sickness** màyw<sup>a</sup> *n*.; ɲeɗeke *n*. **sickness (kind of)** loɓáy<sup>a</sup> *n*. **sickness (mild)** ɲeɗekéím *n*. **sickness spirit** ɲeɗekéím *n*. **side** ay<sup>a</sup> *n*.; kwayw<sup>a</sup> *n*.; kweed<sup>a</sup> *n*.; xán *n*. **side (of hill or mountain)** rutet<sup>a</sup> *n*. **side of clothing** ŋabérá ƙwàzà<sup>ɛ</sup> *n*. **side part** ŋábèrèd<sup>a</sup> *n*. **side-striped** kámáriós *v*. **sidearm** ɲɛ́písítɔ̀l *n*. **sides** kwàìn *n*. **sidestep** kɨɗɔnʉƙɔt<sup>a</sup> *v*. **sideswipe** ɨɛ́bɛ̀s *v*. **sidetrack a discussion** ɨɓátɛ́sʉƙɔta mɛná<sup>ɛ</sup> *v*. **sideways** ŋabér<sup>o</sup> *n*. **sidle up to** rɔɲɛ́sá así *v*. **sieve** ɨsalɛs *v*.; ɨsalɛtɛ́s *v*.; ɲékeikéy<sup>a</sup> *n*.; rɔrɛ́s *v*. **sieve (make)** ɨsalɨtɛ́s *v*. **sift** ɨsalɛs *v*.; ɨsalɛtɛ́s *v*.; rɔrɛ́s *v*.
**sift (make)** ɨsalɨtɛ́s *v*. **sight (front, of a weapon)** ɲeteeɗe *n*. **sight (weapon)** ɲɛ́lɨmɨrá *n*. **sign** eɡésá kwɛtá<sup>ɛ</sup> *v*.; ɨsáánɨŋeés *v*.; iwetés *v*.; ɲámátsar *n*. **sign the cross** iwésá ɲémusaláɓà<sup>e</sup> *v*. **sign up for** xɔ́bɛtɛ́sá así *v*. **signal** ŋízɛ̀s *v*. **signal (smoke)** ts'ûd<sup>a</sup> *n*. **significance** zeísêd<sup>a</sup> *n*. **significant** itíónòn *v*.; zízòn *v*. **silence** ɨʝɛ́mítɛ́sʉƙɔt<sup>a</sup> *v*. **silent** ɨʝɛ́mɔ́n *v*.; líídòn *v*. **silently** ʝìr *ideo*.; lì *ideo*. **silk** ŋísɨl *n*. **silkily** ʝàm *ideo*. **silky-smooth** ʝamúdòn *v*. **similar** ikwáánòn *v*. **simmer** wádòn *v*. **simple** ɓàŋɔ̀n *v*.; batánón *v*.; ɨɓámɔ́n *v*. **simplicity** ɓaŋás *n*. **simsim** kaɲʉm *n*. **simulate** ikwáánitetés *v*. **simultaneously** ikéé kɔ̀n *n*. **sin** ɲasécón *n*.; tɔsɛ́sɔ́n *v*. **since** kwààk<sup>e</sup> *n*.; naítá *subordconn*.; nàpèì *subordconn*.; ɲàpèì *subordconn*. **since earlier today** kwaake nák<sup>a</sup> *n*. **since long ago** kwààkè nòk<sup>o</sup> *n*. **since yesterday** kwààkè sìn *n*. **sincerely** ɡúróɛ́n <sup>ɔ</sup> *n*. **sinew** kon *n*. **sinewy** simánón *v*. **sing** irúkón *v*. **sing and dance** óìdìkwòn *v*. **sing while walking** tofóróƙánón *v*. **singe** iwííɲés *v*.
singer
**singer** ìrùkààm *n*.; irukósíàm *n*. **singing** ìrùk<sup>a</sup> *n*. **singing ants** ʝɔrɔr *n*. **singing hall** ìrùkàhò *n*. **single file** ƙídzɨnɔ́s *v*. **single out** tereties *v*. **singled out** teretiós *v*. **sinistral** betsínón *n*. **sink** ɓuƙonuƙot<sup>a</sup> *v*. **sink (of heart)** mulúráŋòn *v*. **sink teeth into** titikes *v*. **sip** abʉtɛtɛ́s *v*.; ɨwɛtɛs *v*.; tsʉɓɛ́s *v*.; tsʉɓɛtɛ́s *v*. **sip continually** abutiés *v*. **sippable food** abutiam *n*. **sir** ámáze *n*.; ámázeám *n*. **sire** cúrúk<sup>a</sup> *n*. **sire (dog)** ŋókícikw<sup>a</sup> *n*. **sisal rope** baratísím *n*.; màlòr *n*. **sisal species** bàdònìsìm *n*.; barat<sup>a</sup> *n*.; ʝɔ̂d a *n*.; màlòr *n*. **sister (Catholic)** ɓíkìrà *n*. **sister (his/her/its)** yeát<sup>a</sup> *n*. **sister (my)** yeá *n*. **sister (your)** yáó *n*. **sister's husband's sibling (my)** ɲ́cuɡwám *n*. **sister-in-law (brother's wife's sister)** uɡwam *n*. **sister-in-law (brother's wife)** námúí *n*. **sister-in-law (her husband's sister)** ntsínámúí *n*. **sister-in-law (his brother's wife's sister)** ntsúɡwám *n*. **sister-in-law (his brother's wife)** ntsínámúí *n*.
**sister-in-law (his wife's sister)** ntsúɡwám *n*. **sister-in-law (his/her child's spouse's mother)** ntsíɲót<sup>a</sup> *n*. **sister-in-law (his/her sister's husband's sister)** ntsúɡwám *n*. **sister-in-law (husband's brother's wife)** ɛán *n*. **sister-in-law (husband's sister)** námúí *n*. **sister-in-law (my brother's wife's sister)** ɲ́cuɡwám *n*. **sister-in-law (my brother's wife)** ɲ́cinamúí *n*. **sister-in-law (my husband's sister)** ɲ́cinamúí *n*. **sister-in-law (my wife's sister)** ɲ́cuɡwám *n*. **sister-in-law (my)** ɛdɛ́cèk<sup>a</sup> *n*. **sister-in-law (sister's husband's sister)** uɡwam *n*. **sister-in-law (wife's sister)** uɡwam *n*. **sister-in-law (your brother's wife's sister)** buɡwám *n*. **sister-in-law (your brother's wife)** binamúí *n*. **sister-in-law (your child's spouse's mother)** biɲót<sup>a</sup> *n*. **sister-in-law (your husband's sister)** binamúí *n*. **sister-in-law (your sister's husband's sister)** buɡwám *n*. **sister-in-law (your wife's sister)** buɡwám *n*. **sisterhood** yeatínánès *n*. **sisterliness** yeatínánès *n*. **sit** zɛƙwɛ́tɔ́n *v*. **sit (legs straight)** ɨtɛ́ɛ́lɔ̀n *v*.
**sit (make)** zɛƙwɨtɛtɛ́s *v*. **sit alone (silently)** zɛ̀ƙwɔ̀nà lìòò *v*. **sit around** iríƙímánón *v*. **sit decently** ɨɗɨmɛ́sá así *v*. **sit dejectedly** tatónón *v*. **sit down** zɛƙwɛ́tɔ́n *v*.; zɛƙwɨtɛtɛ́s *v*. **sit down (of many)** ɡoƙaakétòn *v*. **sit indecently** iŋáúánón *v*.; tafakésá así *v*. **sit legs apart** iŋátsátsóòn *v*. **sit on a stool** ɨƙárárɔ̀n *v*. **sit on the ground** ɨpáʝɔ́n *v*. **sitting** zɛ̀ƙwɔ̀n *v*. **sitting (of many)** ɡóƙón *v*. **sitting as a group** ɡóƙ<sup>a</sup> *n*. **sitting place** dìyw<sup>a</sup> *n*.; ɡóƙáàw<sup>a</sup> *n*.; zɛƙɔ́áw<sup>a</sup> *n*. **sitty-sit!** dʉʉdʉ́ *nurs*. **six** tude ńdà kɛ̀ɗì kɔ̀n *num*. **six o'clock** ɲásáatɨkaa mɨtátie toomíní ńda kiɗi léɓèts<sup>e</sup> *n*. **sixteen** toomíní ńda kiɗi túde ńdà kɛ̀ɗì kɔ̀n *n*. **sixty** toomínékwa túde ńdà kɛ̀ɗì kɔ̀n *n*. **skeletal** iróƙóòn *v*.; itóƙóƙòòn *v*.; kwédekwedánón *v*.; lotímálèmòn *v*. **skeleton** ɔkɨtín *n*. **skewer** rɔ́ɛ́s *v*. **skewered** rɔ́ɔ́s *v*. **skid** béberésuƙota así *v*.; ɨfɛ́lɔ́nʉƙɔt<sup>a</sup> *v*. **skill** akílìk<sup>a</sup> *n*. **skim off** ɨkáábɛs *v*.; ɨkákápɛ́s *v*.; ɨƙáálɛ́s *v*.; ɨlaɓɛtɛ́s *v*.; iripetés *v*. **skin** hoés *v*.; poxés *v*.; ts'ɛ̀ *n*. **skin (cracked)** kɔmɔ́m *n*. **skin (plant)** ɔmɔ́x *n*.
**skin bump** síts'ádɛ̀ *n*. **skin off** poxésúƙot<sup>a</sup> *v*. **skin on milk** kɨrarap<sup>a</sup> *n*. **skink** lɔ́mɨlí *n*. **skip out** tɔpɛ́ɔ́n *v*. **skip out (and come)** tɔpɛ́ɛ́tɔ̀n *v*. **skip out (and go)** tɔpɛ́ɔ́nʉƙɔt<sup>a</sup> *v*. **skip rope** íɡoriesá simá<sup>e</sup> *v*. **skirmish** iƙúmúnós *v*. **skirt** itsóɗón *v*.; ɲɛ́mɨríńɗà *n*.; tamanɛs *v*.; tamanɛtɛ́s *v*. **skirt (sheep-leather)** ɗóɗòƙwàz *n*. **skirt repeatedly** tamaniés *v*. **skit** wááka na támɔtɔ́s *n*. **skitter up** sekweres *v*. **skull** ikóɔ́k <sup>a</sup> *n*.; ɔka iká<sup>e</sup> *n*. **skullcap (nylon)** nàɗìàk<sup>a</sup> *n*. **skunk-like animal** ɲɔ́kʉɗɔmʉ́tʉ̀ *n*. **sky** didiɡwarí *n*.; ɡwa *n*.; lúl *n*. **sky (clear)** dìdìɔ̀k <sup>a</sup> *n*.; ɡìdòɔ̀k <sup>a</sup> *n*. **slab (stone)** lalatíɓón *n*. **slack** ɓàŋɔ̀n *v*.; ɨtátsámánón *v*. **slack off** ɨwɔ́ɔ́nʉƙɔt<sup>a</sup> *v*. **slacker** ɲakárámɨt<sup>a</sup> *n*. **slacking** ɨwɔ́ɔ́n *v*. **slackness** ɓaŋás *n*. **slander** itúrúmés *v*. **slanderer** ŋítúrumúám *n*. **slant outward (horns)** tɔpɛ́tɔ́n *v*. **slap** ɨɗafɛs *v*.; ɨɗáfɛ́sʉƙɔt<sup>a</sup> *v*. **slap around** ɨɗafiés *v*. **slap the shoulders** ɨɗafesa sáwátìkà<sup>ɛ</sup> *v*. **slash** dzɛrɛ́s *v*. **slash (vegetation)** ɨɗɛtɛs *v*.; ɨrɛʝɛs *v*. **slash firebreak** ɨƙɛɓíƙɛ́ɓɛ́sa ts'aɗí *v*. **slasher** ɲɛ́sɨláx *n*. **slashing of grass** sɨláxìŋ *n*.
**slather** ɨmɔɗímɔ́ɗɛ́s *v*.; ɨmɔ́mɔ́ɗɛ́s *v*. **slaughter** hoés *v*.; tɔŋɔlɛs *v*. **slaughterer** tɔ̀ŋɔ̀lààm *n*. **slaughterhouse** hoesího *n*. **slave** ŋípáƙásíàm *n*.; ŋɨpɔ́táìàm *n*.; ɲɔpɔ́táy<sup>a</sup> *n*. **slavery** ŋiléɓúìnànès *n*.; ŋɨpɔ́táìnànès *n*. **slay** tɔŋɔlɛs *v*. **slay (many)** sáɓés *v*. **slay (singly)** cɛɛ́s *v*. **slayer (of many)** sáɓésìàm *n*. **sleek** ɨpɛlípɛ́lɔ̀n *v*.; mɨlɔ́dɔ̀n *v*.; pɨɗídɔ̀n *v*. **sleekly** mìl *ideo*.; pìɗ<sup>ɨ</sup> *ideo*. **sleep** ep<sup>a</sup> *n*.; èpòn *v*.; mɔɗɔ́ɗ <sup>a</sup> *n*. **sleep (of many)** baráʝónuƙot<sup>a</sup> *v*. **sleep (put to)** epítésuƙot<sup>a</sup> *v*. **sleep a lot** epopos *v*. **sleep around (sexually)** epopos *v*.; weesa kíʝá<sup>e</sup> *v*. **sleeper** epúám *n*. **sleeping** èpòn *v*. **sleeping deeply** nʉ̀s *ideo*.; nʉsʉ́dɔ̀n *v*. **sleeping place** epúáw<sup>a</sup> *n*. **sleeping skin** jèjè *n*. **sleepless** ɡòkòn *v*. **sleepnessness** ɡòk<sup>a</sup> *n*. **sleepy** ɨlʉ́zɔ̀n *v*.; iyalíyálòn *v*. **sleeve** kwɛt<sup>a</sup> *n*. **slender** kaɗótsómòn *v*.; kiɗíwítsánón *v*.; sídɔ̀rɔ̀mɔ̀n *v*.; tɔ̀kɔ̀n *v*. **slenderly** kaɗóts<sup>o</sup> *ideo*. **sleuth** rɔtɛ́ám *n*. **sleuth on** rɔ́tɛ́s *v*. **slice** hoés *v*.; írés *v*. **slice away** iƙémíƙémés *v*. **slice up** irikíríkés *v*.
**sliced food (dry)** iram *n*. **slick** ɗɔrɔ́dɔ̀n *v*.; ɨpɛlípɛ́lɔ̀n *v*.; ʝʉrʉ́tʉ́mɔ̀n *v*.; ʝʉrʉtʉ́tɔ́n *v*.; kwirídòn *v*.; ŋìòn *v*.; pɛlɛ́dɔ̀n *v*.; pɨɗídɔ̀n *v*.; pɔtɔ́dɔ̀n *v*. **slickly** ɗɔ̀r*ideo*.; kwìr*ideo*.; pɛ̀l *ideo*.; pìɗ<sup>ɨ</sup> *ideo*.; pɔ̀t ɔ *ideo*. **slide** ɨfɛlɛsa así *v*.; ɨfɛ́lɔ́nʉƙɔt<sup>a</sup> *v*.; ɨsɔɛs *v*.; ɨsɔɛtɛ́s *v*. **slide off** ɗarámɔ́n *v*. **slide oneself through** ɨsɔɛtɛ́sá así *v*. **slide out** sɛlɛ́tɛ́tɔ̀n *v*. **slide through** ɨsɛ́lɛ́tɛ́sʉƙɔta así *v*. **slight** kiɗíwítsánón *v*. **slight (with food)** itáósés *v*. **slightly numerous** komikómón *v*. **slim** kaɗótsómòn *v*.; kiɗíwítsánón *v*.; sídɔ̀rɔ̀mɔ̀n *v*.; tɔ̀kɔ̀n *v*. **slime** ɡaɗár *n*.; kɨrarap<sup>a</sup> *n*. **slimly** kaɗóts<sup>o</sup> *ideo*. **sling** íbatalɛ́s *v*.; ɲapaaru *n*. **sling over** toryoŋes *v*. **slingshot** ɲapaaru *n*.; ɲɛ́pɨɨrá *n*. **slink** ɨtíɗíɗɛ́sá así *v*. **slink away/off** nʉ́nʉ́tɔ̀n *v*. **slip** ɨfɛlɛsa así *v*.; ɨsɔɛs *v*.; ɨsɔɛtɛ́s *v*. **slip away** ɨfɛlɛsa así *v*. **slip in** sɛrɛpɛs *v*. **slip into** íbʉbʉŋɛ́s *v*.; íbʉbʉŋɛ́sʉ́ƙɔt<sup>a</sup> *v*.; sɛrɛ́pɛ́sʉƙɔt<sup>a</sup> *v*. **slip off** ɨfɛlɛsa así *v*. **slip oneself through** ɨsɔɛtɛ́sá así *v*. **slip out** sɛlɛ́tɛ́tɔ̀n *v*. **slip through** ɨsɛ́lɛ́tɛ́sʉƙɔta así *v*.; sɛrɛpɛs *v*.; sɛrɛ́pɛ́sʉƙɔt<sup>a</sup> *v*. **slipknot** lɔ̀ʝʉ̀rʉ̀tà *n*. **slipper** ŋaɗɛ́tá *n*. **slipperily** ɗɔ̀r *ideo*.; kwìr *ideo*.; pɛ̀l *ideo*.; pɔ̀t ɔ *ideo*.
**slippery** ɗɔrɔ́dɔ̀n *v*.; ʝʉrʉ́tʉ́mɔ̀n *v*.; ʝʉrʉtʉ́tɔ́n *v*.; kwirídòn *v*.; pɛlɛ́dɔ̀n *v*.; pɔtɔ́dɔ̀n *v*. **slippy** kwirídòn *v*.; sɛlɛ́tɛ́mɔ̀n *v*. **slither** lúkúɗukuɗánón *v*. **slithery** sɛlɛ́tɛ́mɔ̀n *v*. **slitted** mɨʝílímɔ̀n *v*. **slope** ɓɔ́kɔ̀ɲ *n*. **sloped** sémédedánón *v*. **slosh** ɨpɔkɛs *v*.; ts'álúbòn *v*. **slosh through water** íbuɗésá cué *v*. **slothful** ɓʉɓʉsánón *v*.; wéésánón *v*. **slouch** rʉ̀ʝɔ̀n *v*. **slough off** ɗarámɔ́n *v*. **slow** inípónòn *v*. **slow (mentally)** mɨɲɔna íkèdè *v*. **slow down** inípónítésúƙot<sup>a</sup> *v*.; ɨwɔ́ɔ́nʉƙɔt<sup>a</sup> *v*.; tosipetés *v*. **slowing** ɨwɔ́ɔ́n *v*. **slowly** hɨíʝ<sup>a</sup> *adv*.; hɨíʝ<sup>ɔ</sup> *adv*.; kédie kwáts<sup>a</sup> *n*.; wɛ̀wɛ̀ɛ̀s *ideo*. **slowly (very)** pààì *ideo*. **sludgily** yàŋ *ideo*. **sludgy** bɔrɔ́tsɔ́mɔ̀n *v*.; yaŋádòn *v*. **slug** tanaŋes *v*.; tɔƙɔtɔƙ<sup>a</sup> *n*. **slug (bullet)** bʉbʉn *n*. **sluggish** iʝíŋáánón *v*. **slumber** ep<sup>a</sup> *n*.; èpòn *v*.; mɔɗɔ́ɗ <sup>a</sup> *n*. **slump** rʉ̀ʝɔ̀n *v*. **slur** ɗáƙón *v*.; iŋáʝápánón *v*. **slurp** ɨsɔ́rɔ́ɓɛ́s *v*.; xáɓútés *v*. **slurp continually.** abutiés *v*. **slurpable food** ɨsɔrɔɓam *n*. **smack** ídirés *v*. **small** kwátsón *v*. **small (become)** kwátsónuƙot<sup>a</sup> *v*.
**small (of many)** kwátsíkaakón *v*. **small (opening)** tɨɨts'ímɔ̀n *v*. **small-bodied** tsaʉ́ɗímɔ̀n *v*. **smaller (make)** kwatsítésuƙot<sup>a</sup> *v*. **smash** ɨlɛɗɛs *v*.; ɨtsakɛs *v*. **smash up** ɨtsakítsákɛ́s *v*. **smear** ɨlɔ́lɔ́rɛ́s *v*.; ɨmɔɗímɔ́ɗɛ́s *v*.; ɨmɔ́mɔ́ɗɛ́s *v*. **smear (goat dung)** síɛ́s *v*. **smear (reputation)** itúrúmés *v*. **smear with clay** ŋɔrɨtɛtɛ́s *v*. **smell** mídzatés *v*.; mídzatetés *v*.; mídzòn *v*.; ɔn *n*.; ɔnɛd<sup>a</sup> *n*.; wetésá kɔíná<sup>ɛ</sup> *v*. **smell fetid** mídzona ɗɛtsɨɗɛ́tsík<sup>ɛ</sup> *v*. **smell rotten** mídzònà ɗùk<sup>u</sup> *v*. **smell to death** mídzatés *v*. **smelly** ɨmʉ́sɔ́ɔ̀n *v*. **smelly (become)** ɨmʉ́sɛ́ɛ̀tɔ̀n *v*. **smelly (make)** mídzitésúƙot<sup>a</sup> *v*. **smelly (very)** ɗùk<sup>u</sup> *ideo*. **smile** ɨmʉ́mʉ́ɔ̀n *v*.; tamáísánón *v*. **smile (make)** ɨmʉ́mʉ́ɨtɛtɛ́s *v*. **smock (leather)** kɔ́lɔ́ts<sup>a</sup> *n*. **smoke** ipúróòn *v*.; iwaŋíwáŋés *v*.; ts'ûd<sup>a</sup> *n*.; ts'udités *v*.; wetés *v*. **smoke (a cigarette)** ɨsɔ́kɔ́teés *v*. **smoke (begin to)** ipúréètòn *v*. **smoke (ritually)** ipúréés *v*. **smoke out** ipúréés *v*. **smoke signal** ts'ûd<sup>a</sup> *n*. **smolder** iɲipíɲípòn *v*. **smooth** lɨwídɔ̀n *v*.; pɨlɔ́dɔ̀n *v*. **smooth (make)** pɨlɔ́dɨtɛ́sʉ́ƙɔt<sup>a</sup> *v*. **smoothen** ipiipíyeés *v*. **smoothen (with water)** iláɓúés *v*. **smoothen out** pɨlɔ́dɨtɛ́sʉ́ƙɔt<sup>a</sup> *v*. **smoothly** ʝàm *ideo*.; lìw *ideo*.; pìl *ideo*.
**smother** tʉɓʉnɛ́s *v*. **smudge** ɨlɔ́lɔ́rɛ́s *v*. **smuggle** ɨɛpɛtɛ́s *v*. **smut fungus** lósínák<sup>a</sup> *n*. **snail** tɔƙɔtɔƙ<sup>a</sup> *n*. **snail shell** irex *n*.; tɔƙɔtɔƙáhò *n*. **snake** ídèm *n*. **snake (blind)** lokaliliŋ *n*. **snake (rufous beaked)** oŋerep<sup>a</sup> *n*. **snake (sand)** nakɔlɨták<sup>a</sup> *n*. **snake (small green)** ílebéɗ<sup>a</sup> *n*. **snake fang** ídèmèkwàyw<sup>a</sup> *n*. **snake venom** ídèmètàt<sup>a</sup> *n*. **snake-bite** ídemeƙɨdzɛ́s *n*. **snap** ɗusúmón *v*. **snap (react)** tokúétòn *v*.; tokúréètòn *v*. **snap (snarl)** ɨɲɛ́ɛ́mɔ̀n *v*. **snap a photo** iwetésá ɲɛ́pítsaá<sup>ɛ</sup> *v*. **snap a photo of** iwetés *v*. **snap off** ɨɓɛkíɓɛ́kɛ́s *v*.; wakés *v*. **snap off in pieces** wakatiés *v*.; wakáwákatés *v*. **snap!** ɓɛk<sup>ɛ</sup> *ideo*.; ɗì *ideo*.; tɛ̀ *ideo*. **snapshot** kúrúkúr *n*.; ɲɛ́pítsa *n*. **snare** kotsítésuƙot<sup>a</sup> *v*.; sáɡòsìm *n*.; sáɡwès *v*. **snare (neck)** ɲákol *n*. **snare (wire neck)** ɲáwáya *n*. **snare rope** lozikinet<sup>a</sup> *n*.; lozikit<sup>a</sup> *n*. **snare spring** kàsw<sup>a</sup> *n*. **snare stick** tɨmɛ́l *n*. **snare stick (bent)** tɔmɔƙɔrɛs *n*. **snare trigger** kwanɛd<sup>a</sup> *n*. **snared** kòtsòn *v*.; sáɡoanón *v*. **snared (become)** kotsonuƙot<sup>a</sup> *v*.
**snaring** sâɡw<sup>a</sup> *n*. **snarl** ɨɲɛ́ɛ́mɔ̀n *v*. **snatch** ɨrɛɗɛs *v*.; ŋusés *v*.; ŋusésúƙot<sup>a</sup> *v*.; taŋates *v*.; tokopes *v*.; toreɓes *v*. **snatch away** taŋátésuƙot<sup>a</sup> *v*.; tokópésuƙot<sup>a</sup> *v*. **sneak** ɗíɗítɛ́sʉƙɔt<sup>a</sup> *v*.; ɗíɗítɛtɛ́s *v*.; ɨsʉ́mɔ́n *v*.; ɨtíɗíɗɛ́s *v*.; ɨtíɗíɗɛ́sá así *v*.; totséɗón *v*. **sneak away** dzuesésúƙota así *v*.; ɨfɛlɛsa así *v*. **sneak off** dzuesésúƙota así *v*.; ɨfɛlɛsa así *v*.; ɨsʉ́mɔ́nʉƙɔt<sup>a</sup> *v*. **sneak up** íbɛ̀ɗìbɛ̀ɗɔ̀n *v*.; ɨsʉ́mɛ́tɔ̀n *v*. **sneak up on** tɔlɛ́lɛ́ɛtɛ́s *v*. **sneaky** iɗásón *v*. **sneeze** síƙón *v*.; sìƙw<sup>a</sup> *n*. **sniff** mídzatés *v*. **sniff (tobacco)** ʝʉ́rɛ́s *v*. **snip** iɲipes *v*.; ɨrɛɓɛs *v*. **snip off** ɨrɛ́ɓɛ́sʉƙɔt<sup>a</sup> *v*. **snitch** ɗíɗítɛ́sʉƙɔt<sup>a</sup> *v*.; ɗíɗítɛtɛ́s *v*. **snitch on** ilíítés *v*. **snoop around** tɨrɨfírífɛ́s *v*. **snore** ŋɔ́rɔ́rɔ̀n *v*. **snort** síƙón *v*.; sìƙw<sup>a</sup> *n*. **snort at** ifúƙúfuƙés *v*. **snot** ɗɔ̀ƙɔ̀n *n*.; ɲarʉ́kʉ́m *n*. **snout** loɓôz *n*. **snub** ɨmɛ́ɗɛ́lɛ́s *v*. **snub (with food)** itáósés *v*. **snuff** wetés *v*. **snuff (tobacco)** ʝʉ́rɛ́s *v*.; ʝʉ́rɛ́sʉƙɔt<sup>a</sup> *v*. **snuff container** ɲeɓuryaŋ *n*. **snuff out (life)** ts'eítésuƙot<sup>a</sup> *v*. **snuffle at** ifúƙúfuƙés *v*. **so** kòt<sup>o</sup> *coordconn*.
**so that** ikóteré *subordconn*.; kánì *subordconn*.; kánɨ náa táa *subordconn*.; kóteré *subordconn*. **so that … not** kánɨ mookóo *subordconn*. **so then** ɓàz *interj*.; kíná *coordconn*. **so there!** ɓàz *interj*. **so-and-so** tatanám *n*. **so-so** ŋwanɨŋwánɔ́n *v*. **soak** ɨɛ́ɓítɛtɛ́s *v*. **soak (grist)** mʉrɛ́s *v*. **soaked** ts'alídòn *v*. **soap** dàlìs *n*.; ɲásaɓuní *n*. **soap (laundry)** hómò *n*.; ɲéómò *n*. **soar** ɨɔ́ɔ́rɔ̀n *v*. **sober (not drunk)** bótsóna iká<sup>e</sup> *v*. **soccer** ɲɛ́pɨɨrá *n*. **sock** ɲósóƙis *n*. **soda** ɲɔ́sɔ́ɗa *n*. **soda ash** ɲaɓáláŋɨt<sup>a</sup> *n*.; ɲámakaɗí *n*. **sodium carbonate** ɲaɓáláŋɨt<sup>a</sup> *n*.; ɲámakaɗí *n*. **Sodom apple** tùlèl *n*. **soft** bubuxánón *v*.; buɗúdòn *v*.; burádòn *v*.; dabúdòn *v*.; heɓúdòn *v*.; ʝaulímòn *v*.; ɲipídòn *v*.; xaɓúdòn *v*. **soft (become)** bubuxánónuƙot<sup>a</sup> *v*. **soft (make)** buɗúditésúƙot<sup>a</sup> *v*. **soft (of metal)** lumúdòn *v*. **soft (of soil)** yuúdòn *v*. **soft (powdery)** ɲapíɗímòm *v*. **soft and tender** dabúdòn *v*. **soft inside** yumúdòn *v*. **soft spot** baɗɨbaɗas *n*.; bɔɗɨbɔɗɔs *n*. **soften** bubuxánónuƙot<sup>a</sup> *v*. **soften (emotionally)** isyónónuƙot<sup>a</sup> *v*. **soften up** buɗúditésúƙot<sup>a</sup> *v*.
**softly** bùɗ<sup>u</sup> *ideo*.; dàb<sup>u</sup> *ideo*.; hèɓ<sup>u</sup> *ideo*.; lùm *ideo*.; ɲìpⁱ *ideo*.; sokósíìk<sup>e</sup> *v*.; xàɓ<sup>u</sup> *ideo*. **softly (of soil)** yù *ideo*. **softly inside** yùm *ideo*. **soggy** fɔts'ɔ́dɔ̀n *v*. **soil** ʝʉm *n*. **soil (colored)** ɲálámʉɲɛna *n*. **soil (fertile)** ʝʉma na zîz *n*. **soil (red)** boŋórén *n*.; ɲapala *n*. **soiled** ɨráŋʉ́nánón *v*.; ŋɔrɔ́ɲɔ́mɔ̀n *v*.; ɲɔŋɔ́rɔ́mɔ̀n *v*. **Solanum incanum** tùlèl *n*. **solar eclipse** badona fetí *n*. **solar panel** ɲɔ́sɔ́la *n*. **soldier** ʝɔrɔrɔ́ám *n*.; kéààm *n*. **soldier ant** lókók<sup>a</sup> *n*. **soldier termite** lókók<sup>a</sup> *n*. **soldiers** dìdì *n*.; ʝɔrɔr *n*. **sole** ɗòk<sup>u</sup> *adv*. **sole (of foot)** dɛááƙw<sup>a</sup> *n*. **solely** ɛɗá *adv*. **solicit** tɔɓɛ́ɲɛ́tɔ̀n *v*.; wáán *v*. **soliciting** wáán *n*. **solidified** iɗíkón *v*. **solidify** iɗíkétòn *v*.; iɗikitetés *v*. **solitary** ɗòk<sup>u</sup> *adv*. **solvable problem** itémítuƙotam *n*. **solve** hoetés *v*.; itemités *v*.; ŋurutiés *v*.; ŋurutiesúƙot<sup>a</sup> *v*. **solved** ŋurutiós *v*. **Somali** Oríáé *n*. **Somali language** Ŋísʉmálìtòd<sup>a</sup> *n*.; Oríáénítòd<sup>a</sup> *n*. **Somali person** Ŋísʉmálìàm *n*. **Somalia** Somálìà *n*. **some (plural)** kíníɛ́n *pro*. **some (singular)** kɔ́níɛ́n *pro*.
southerner
**some more** sa *pro*. **some other** sa *pro*. **some other (sg.)** kɔn *pro*. **somebody** kɔ́nɛ́ɛ́ná ámá<sup>e</sup> *n*. **someone** kɔ́nɛ́ɛ́ná ámá<sup>e</sup> *n*.; kɔníám *pro*. **somersault** aɓúlúkánón *v*.; tíbìɗìlɔ̀n *v*. **something** kɔ́nɛ́ɛ́ná kɔ́rɔ́ɓádì *n*. **sometimes (hours)** sayó ɲásáàtìkà<sup>ɛ</sup> *n*. **somewhere else** kɔ́náy<sup>a</sup> *pro*. **son** dzàƙ<sup>a</sup> *n*.; sore *n*. **son (his/her)** dzàƙɛ̀d <sup>a</sup> *n*. **son (of my father)** abáŋídzàƙ<sup>a</sup> *n*. **son (young)** soréím *n*. **song** dikw<sup>a</sup> *n*.; ìrùk<sup>a</sup> *n*. **Soo language** Ŋítépesítôd<sup>a</sup> *n*. **Soo people** Ŋítépes *n*. **soon** názɛ̀ƙwà *n*.; ts'ɔ̀ɔ̀ *adv*. **soot** ɲémúɗets<sup>a</sup> *n*.; ɲémúɗuɗu *n*. **soot (tobacco)** ɲéɗípor *n*. **sooty** ɨmɔ́ɗɔ́rɔ̀n *v*. **sopping** ts'alídòn *v*. **soppingly** ts'àl *ideo*. **sorcerer** bàdìàm *n*. **sorcerer (who stops rain)** tuɗúlónìàm *n*. **sorcery** badirét<sup>a</sup> *n*.; badirétínànès *n*.; ƙʉts'ánánès *n*. **sore** ɔ́ʝ <sup>a</sup> *n*. **sore (small)** ɔ́ʝáìm *n*. **sorghum** ŋám *n*. **sorghum flowers** kadɨx *n*. **sorghum variety (black)** ŋámá na buɗám *n*. **sorghum variety (brownish-gray)** ɗìɗèŋàm *n*. **sorghum variety (droopy)** loʝúulú *n*. **sorghum variety (hairy)** ɲákaɓír *n*.
**sorghum variety (purplish)** serínà *n*. **sorghum variety (red)** ŋámá nà ɗìw<sup>a</sup> *n*.; ɲɛmɛray<sup>a</sup> *n*. **sorghum variety (round-headed)** nalíɨlí *n*. **sorghum variety (Toposa)** Kɔrɔmɔtáŋám *n*. **sorghum variety (Turkana)** ɲékimyét<sup>a</sup> *n*.; Pakóícéŋám *n*. **sorghum variety (white)** ŋámá nà ɓèts'<sup>a</sup> *n*. **sorghum variety (yellow)** ɗókóts<sup>a</sup> *n*.; natɛ́ɓ <sup>a</sup> *n*.; oɲaŋ *n*. **sorghum varity (tall)** walá *n*. **sorrowful** itásónòn *v*.; tasónón *v*. **sort** ɨsalɛs *v*.; ɨsalɛtɛ́s *v*.; rɔrɛ́s *v*. **sort (make)** ɨsalɨtɛ́s *v*. **soul** ɡúr *n*. **sound** arútón *v*. **sound an alarm** iwákón *v*. **sound empty** ɗɛʉɗɛ́wɔ́n *v*. **sound out** arútónuƙot<sup>a</sup> *v*. **sounding alarm** iwáákós *v*. **sounding like** *sh-sh* wɔ̀x *ideo*. **soup** seekw<sup>a</sup> *n*. **sour** ɓàròn *v*. **sour (become)** ɓaronuƙot<sup>a</sup> *v*. **sour (make)** ɓarites *v*.; ɓarítésuƙot<sup>a</sup> *v*. **sour (of malt)** mʉránón *v*. **sour mash** ɓaram *n*. **source** itsyákétònìàw<sup>a</sup> *n*. **source of water** cuáák<sup>a</sup> *n*. **souse** ɨɛ́ɓítɛtɛ́s *v*. **south** kɔ́ɔ́kíʝ<sup>o</sup> *n*.; nɔ́ɔ́kíʝ<sup>o</sup> *n*. **South Sudan** Sʉɗán *n*. **southerly direction** ɡíɡiroxan *dem*. **Southern Cross** Ɲémusaláɓà *n*. **southerner** kɔ́ɔ́kíʝóàm *n*.
**southward** kɔ́ɔ́kíʝ<sup>o</sup> *n*. **sovereign** ipúkéésíàm *n*.; tòtwàrààm *n*. **sow** íbɨtɛ́s *v*.; tɛwɛɛs *v*. **space** ɨlɔ́lɔ́kɛ́s *v*.; ilores *v*.; zɛƙɔ́áw<sup>a</sup> *n*. **space (outer)** didiɡwarí *n*. **space too closely** itsuɗútsúɗés *v*.; ituɗútúɗés *v*. **spacious** ɨlɔ́lɔ́mɔ̀n *v*.; lalʉ́ʝɔ́n *v*. **spade** ɲakáɓɛ́t <sup>a</sup> *n*.; ɲékitiyó *n*. **spade (wooden)** nakút<sup>a</sup> *n*. **spank** ipíkéés *v*. **spar** ɲèùrìà *n*.; ɲeuríétòn *v*. **spare** ɨɓámɔ́n *v*. **sparklely** mìl *ideo*. **sparkly** mɨlídɔ̀n *v*. **sparks** ŋkaɗɛɛɗɛ́y <sup>a</sup> *n*. **sparrow (parrot-billed)** mɨdɨƙ<sup>a</sup> *n*. **spatter** ɨratírátɛ́s *v*.; irwaírwéés *v*.; irwates *v*.; ɨwɛ́ɛ́lánón *v*.; tɔfɔ́ɗɔ́n *v*. **spatula (wooden)** cɛbɛn *n*.; ɲémiikó *n*. **speak** ɨɛ́nɔ́n *v*.; tódètòn *v*.; tódòn *v*. **speak (begin to)** tódonuƙot<sup>a</sup> *v*. **speak about** tódetés *v*. **speak eloquently** isiresa aká<sup>e</sup> *v*. **speak harshly** ɡuts'uriesá tódà<sup>e</sup> *v*. **speak harshly to** doƙofiés *v*. **speak indistinctly** ɗáƙón *v*.; iŋáʝápánón *v*. **speak meanly** ɡuts'uriesá tódà<sup>e</sup> *v*. **speak pointlessly** ɨpɛípɛ́ɛ́sá tódà<sup>e</sup> *v*. **speak slowly** ízìɗòn *v*. **speak to each other** tódinós *v*. **speak vaguely** ɨŋaɨŋɛ́ɛ́sa tóda<sup>e</sup> *v*. **speaker** taatsaama tódà<sup>e</sup> *n*.; tódààm *n*. **spear** ɓɨs *n*.; toɓés *v*. **spear (long-headed)** ɲátúm *n*.
**spear (of many)** bɛrɛ́s *v*. **spear (sharpened stick)** ʝìrɔ̀k <sup>a</sup> *n*. **spear a tree** toɓésá dakwí *v*. **spear bluntly** iŋulúŋúlés *v*. **spear from afar** ɨtsɛ́tsɛ́ɛ́s *v*. **spear repeatedly** toɓítóɓiés *v*. **spear shaft (long)** narwá *n*. **spear shaft (short)** erumén *n*. **spear through** ɲɛ́ŋɛ́s *v*. **spearhead** ɓɨsáák<sup>a</sup> *n*. **spearhead (long-necked)** ɲɛlɨrát<sup>a</sup> *n*. **spearhead (short-necked)** ɲɛ́ɓítɨ *n*. **spearhead neck** ɓɨsáɓóló *n*. **speartip (rear)** ɲɛ́rʉ́mats<sup>a</sup> *n*. **special** ƙanotós *v*. **specialty** ɲɛmʉna *n*. **species** bònìt<sup>a</sup> *n*.; ɲákaɓɨlá *n*. **speck** kiɗoɗots<sup>a</sup> *n*.; símíɗiɗí *n*. **speckle** iɗolíɗólés *v*.; itwelítwélés *v*. **speckled** iɗolíɗólòn *v*.; itwelítwélós *v*. **spectacles** ɲékiyóìk<sup>a</sup> *n*. **spectator** enésúƙotíám *n*. **specter** kúrúkúr *n*. **speech** tôd<sup>a</sup> *n*. **speech (careless)** múɗúkánónìtòd<sup>a</sup> *n*. **speech (muddled)** tóda ni buɗám *n*. **speed** ɨrʉtsɛsa así *v*.; ɲésipíɗ<sup>a</sup> *n*. **speedy** itírónòn *v*.; wɛ́ɛ́nɔ̀n *v*. **spell it out** tɔmɛɛtɛ́sá tódà<sup>e</sup> *v*. **spend the day** iríóonuƙot<sup>a</sup> *v*. **spend time** iríóòn *v*.; iríóonuƙot<sup>a</sup> *v*. **spend wildly** ɨkwaríkwárɛ́s *v*. **spent (tired)** ziálámòn *v*.; zíkímétòn *v*.; ziláámòn *v*. **sperm** ɗír *n*. **spew** ɨlɔ́ɓɔ́tɛtɛ́s *v*. **sphenoid bone** matáŋíɔ̀k <sup>a</sup> *n*.
#### spoil everything
#### spherical
**spherical** ɨlʉ́lʉ́ŋɔ́s *v*. **spherical (make)** ɨlʉ́lʉ́ŋɛ́s *v*. **sphincter (anal)** ɔ́zàhò *n*. **spice** ɛfɨtɛs *v*. **spice up** ɛfɨtɛs *v*. **spider** abûb<sup>a</sup> *n*. **spiderweb** abûb<sup>a</sup> *n*. **spike** omén *n*. **spill** ɗaɗátésuƙot<sup>a</sup> *v*.; ƙúdès *v*.; ƙúdesuƙot<sup>a</sup> *v*.; ƙúdetés *v*. **spill all over** ɨtsʉ́rʉ́tsʉ́rɛ́s *v*.; ɨtsʉ́rʉ́tsʉ́rɛ́sʉ́ƙɔt<sup>a</sup> *v*. **spill over** bukúrésuƙota así *v*.; ɨlápɛ́tɔ̀n *v*. **spin** irímítetés *v*.; irímón *v*.; ɨríŋítɛ́s *v*.; ɨríŋɔ́n *v*. **spinach** ɲásalátà *n*. **spinal cord** lɔ́ɓírɨɓír *n*. **spindly** sawátsámòn *v*. **spine** ɡòɡòròʝ<sup>a</sup> *n*.; ɡòɡòròʝòɔ̀k <sup>a</sup> *n*. **spiral** ilúƙúretés *v*.; iyérónuƙot<sup>a</sup> *v*. **spiraled** iyérón *v*. **spirit** suɡur *n*.; sʉ̀p <sup>a</sup> *n*. **spirit (earth)** ɲɛkípyɛ́*n*. **spirit (evil)** ɲɛkípyɛ́*n*. **spirit (sickness-causing)** ɲeɗekéím *n*. **spirit dance** ʝàkàlʉ̀kà *n*. **spirits (earth)** ŋípyɛn *n*. **spirits (evil)** ŋípyɛn *n*. **spit** tat<sup>a</sup> *n*.; tatés *v*.; tàtòn *v*. **spit (skewer)** rɔ́ɛ́s *v*. **spit far** tsɨrɨtɛs *v*. **spit on** ɨmwaímwɛ́ɛ́s *v*. **spit out** tatésúƙot<sup>a</sup> *v*. **spit out repeatedly** tatíésuƙot<sup>a</sup> *v*. **spit repeatedly** tatiés *v*.
**spit-inducer** tatitésuƙotíám *n*. **spit-pestle plant** tatíáʝ<sup>a</sup> *n*. **spiteful person** ɲɛ́kɨsɨránìàm *n*. **spitefulness** ɲɛ́kɨsɨrán *n*. **spitter** tatiesíám *n*. **spittle** tat<sup>a</sup> *n*. **splash** ts'álúbòn *v*. **splash!** bùlùƙ<sup>u</sup> *ideo*. **splat!** pìɔ̀ *ideo*. **splatter** ɨratírátɛ́s *v*.; ɨwɛ́ɛ́lánón *v*.; tɔfɔ́ɗɔ́n *v*. **spleen** máɗíŋ *n*. **splendid** dòòn *v*. **splendor** daás *n*. **splint** íbunutsés *v*. **splish-splash!** calúɓ<sup>u</sup> *ideo*. **split** ɓɛlɛ́s *v*.; ɓɛlɛtɛ́s *v*.; ɓɛlɔ́s *v*.; taŋatsárón *v*.; tɛlɛ́tsɔ́n *v*.; terés *v*.; tɔɓɛlɛs *v*.; toŋélón *v*. **split apart** ɓɛ́ɓɛ́lɛ́s *v*.; ɓɛ́ɓɛ́lɔ́s *v*.; itotoles *v*.; terémón *v*.; tɔɓɛ́lɛ́sʉƙɔta<sup>a</sup> *v*. **split apart multiply** ɓeletiés *v*. **split in pieces** ɨɓɛ́ɓɛ́lɛ́s *v*. **split in two** pakámón *v*.; pakés *v*. **split multiply** pakatiés *v*. **split open** ɓɛ́ɓɛ́lɛ́s *v*.; ɓɛ́ɓɛ́lɔ́s *v*.; ɓɛlɛ́ɓɛ́lánón *v*.; ɓelémón *v*. **split open (of pods)** kwɛ́dɔ̀n *v*. **split up** terémétòn *v*.; terémón *v*.; terémónuƙot<sup>a</sup> *v*.; terétéránitésúƙot<sup>a</sup> *v*.; terétéránón *v*.; tereties *v*.; teretiós *v*.; tɔɓɛ́lɛ́sʉƙɔta<sup>a</sup> *v*. **splitch!** rùt<sup>u</sup> *ideo*. **splosh** ɨpɔkɛs *v*. **spoil** imóɲíkees *v*.; imóɲíkeetés *v*.; ɨraŋɛs *v*.; ɨraŋɛtɛ́s *v*.; masánétòn *v*. **spoil everything** nts'áƙóna sèrèìk<sup>e</sup> *v*.
**spoiled** ɨraŋɔs *v*.; ɨráŋʉ́nánón *v*.; masánón *v*. **spoiled (become)** ɨraŋímétòn *v*. **spokesperson** taatsaama tódà<sup>e</sup> *n*.; tódààm *n*. **sponge** lɛŋɛ́s *v*.; lɛŋɛ́síàm *n*. **spongily** bùf *ideo*. **sponging** olíɓó *n*. **spongy** bufúdòn *v*. **spongy bone** ɲɛ́ɲam *n*. **spooky thing** bàdìàm *n*. **spoon (metal)** ɲékiʝikó *n*. **spoon (wooden)** ƙolom *n*. **spoot!** pìɔ̀ *ideo*. **sport** ɲaɓolya *n*.; wáák<sup>a</sup> *n*. **spot** bàsɔ̀n *v*.; ɨtsɔɓítsɔ́ɓɛ́s *v*. **spot (bare)** ɲapatsole *n*. **spot (claimed)** ɲeɗúkór *n*. **spot (hard)** ɲapáyál *n*. **spot (place)** bácík<sup>a</sup> *n*. **spotless** xɔ́dɔ̀n *v*.; xɔtánón *v*. **spotted** iɗolíɗólòn *v*.; ɨtsɔɓítsɔ́ɓɔ́s *v*.; komolánón *v*.; merixánón *v*.; tsɨpɨtsípɔ́n *v*.; tábàsànètòn *v*. **spotted (black-and-white)** ŋorokánón *v*.; ŋorókón *v*. **spotter** weretsíám *n*. **spotty** ɨtsɔɓítsɔ́ɓɔ̀n *v*. **sprawl** ɨpɛ́pɛ́tánón *v*. **sprawl out** ɨpɛ́pɛ́tánónuƙot<sup>a</sup> *v*. **spray** bítés *v*. **spray (bombard)** ídʉrɛ́s *v*. **spread** ɨmɔɗímɔ́ɗɛ́s *v*.; ɨmɔ́mɔ́ɗɛ́s *v*.; ɨwɛ́ɛ́lɛ́s *v*. **spread (legs) apart** dɛŋɛlɛs *v*. **spread about** ɨpɛ́pɛ́tɛ́s *v*.
**spread apart** tɛlɛɛs *v*. **spread around** ɓátsɛ́s *v*.; ɨkwákwárɛ́s *v*.; ɨkwaríkwárɛ́s *v*.; ɨkwaríkwarɔ́s *v*.; iríríʝés *v*.; iwies *v*. **spread circularly** ɨmalímálɛ́s *v*. **spread oneself open** ɓátsɛ́sa así *v*. **spread out** ɨwɛ́ɛ́lánón *v*.; ɨwɛ́ɛ́lɛ́sʉƙɔt<sup>a</sup> *v*.; ɨwɛ́ɛ́lɛtɛ́s *v*.; ɨwɛ́ɛ́lɔ́s *v*.; tɔlɔɛs *v*.; tɔpɛtɛs *v*.; tɔpɛ́tɛ́sʉƙɔt<sup>a</sup> *v*. **spread out under** tafakés *v*. **spread over (an area)** ɨkáyɛ́ɛ́s *v*. **spread soil** ɨwɛ́ɛ́lɛ́sá ʝʉmwí *v*. **spring** íbòtòn *v*.; ɨɗɛ́ɛ́tɔ̀n *v*.; iɗótón *v*.; íɡɔ̀rɔ̀bɔ̀n *v*. **spring (feather-holding)** ɲeteeɗe *n*. **spring (of a trap)** ɨɗálɛ́sʉƙɔta así *v*. **spring (in sand)** ɲakúʝá *n*. **spring (a trap)** ɨɗalɛs *v*. **spring (of water)** ɲɛɨtánɨt<sup>a</sup> *n*.; ɲɛlɛ́lyá *n*. **spring mechanism (of weapons)** zɔ̀t <sup>a</sup> *n*. **springily** tùf *ideo*.; tùs *ideo*. **springing** ɨɗɛ́ɔ́n *v*. **springy** tufádòn *v*.; tusúdòn *v*. **sprinkle** irwaírwéés *v*.; irwates *v*.; iyikes *v*.; towates *v*.; towatetés *v*.; xɛɛ́s *v*.; xɛɛsʉ́ƙɔt<sup>a</sup> *v*.; xɛɛtɛ́s *v*. **sprinkle (granulates)** ízuzués *v*. **sprinkle (rain)** kʉ́f *n*. **sprinkle ashes on paths** ƙúdesa káúe mucéíkàk<sup>e</sup> *v*. **sprite** kíʝáìm *n*. **spritz** iyikes *v*. **sprout** ɓúrukúkón *v*.; morétón *v*.; rʉ́bɔ̀n *v*.; tɔɓɔ́rɔ́kánón *v*.; tʉwɛ́tɔ́n *v*.; tʉ̀wɔ̀n *v*.; xúbètòn *v*.; xúbòn *v*. **sprout (of grain)** xokómón *v*. **sprout (of leaves)** ŋʉrʉrʉ́ɲɔ́n *v*.
**sprout (of maize cobs)** ɨsínákòn *v*. **sprout up** ŋʉrʉ́ɲʉ́ɲɛ̀tɔ̀n *v*.; rʉ́bɛ̀tɔ̀n *v*. **spry** pɔɗɔ́dɔ̀n *v*. **spryly** pɔ̀ɗ ɔ *ideo*. **spud** ɲɛ́ɓɨás *n*. **spur** ɨʝʉkʉ́ʝʉ́kɛ́s *v*.; ɲéɡets<sup>a</sup> *n*. **spur on** ɨmʉ́káitetés *v*. **spurfowl (yellow-necked)** kɔ̀dz<sup>a</sup> *n*. **spurn** ɨmɛ́ɗɛ́lɛ́s *v*. **spurt** ts'írítɔ̀n *v*. **sputum** ɗɔ̀ƙɔ̀n *n*.; ɲarʉ́kʉ́m *n*. **spy** irimesíám *n*.; rɔtɛ́ám *n*. **spy on** láɡalaɡetés *v*.; rɔ́tɛ́s *v*.; toreɓes *v*. **spy on from afar** towates *v*. **spy out** láɡalaɡetés *v*. **squabble** iƙúmúnós *v*.; ɨɲʉ́ɲʉ́rɔ̀n *v*.; ŋʉ́zʉmánón *v*. **squander** eletiésuƙot<sup>a</sup> *v*.; iɲekes *v*.; iɲékésuƙot<sup>a</sup> *v*.; iɲekíɲékés *v*. **squash** ɓirés *v*.; ɓírítésuƙot<sup>a</sup> *v*.; ɨlɛɗɛs *v*.; lomuƙe *n*.; lɔ́pʉ́l *n*.; rɛɗɛ́s *v*. **squashed (get)** ɓirímón *v*. **squashily** ɓìr *ideo*.; ɲàl *ideo*.; rɔ̀ʝ ɔ *ideo*. **squashy** ɓirídòn *v*.; ɲalídòn *v*.; rɔʝɔ́dɔ̀n *v*. **squat** tsɔ́nɔ́n *v*. **squeak squeak!** tswíítswí *ideo*. **squeaky voice (have a)** ɨsɨrísírɔ́n *v*. **squeezable** heɓúdòn *v*. **squeezably** hèɓ<sup>u</sup> *ideo*. **squeeze** bízès *v*.; ɨrɨɗɛs *v*.; ɨrɨɗɛtɛ́s *v*.; ʝʉ́tɛ́s *v*.; rɨɗɛ́s *v*.; tʉtsʉɛs *v*. **squeeze all over** bízibizatés *v*. **squeeze out** bízetés *v*.; ɨpírísɛtɛ́s *v*.; ʝʉ́tɛ́sʉƙɔt<sup>a</sup> *v*.; tʉtsʉ́ɛ́sʉƙɔt<sup>a</sup> *v*. **squeezy** heɓúdòn *v*. **squelchily** rɔ̀ʝ ɔ *ideo*.
**squelchy** rɔʝɔ́dɔ̀n *v*. **squiggle** ŋʉɗʉŋʉ́ɗɔ́n *v*. **squint** ɨmʉɗʉ́mʉ́ɗɔ̀n *v*.; wízɨlɛ́s *v*. **squint at** katsés *v*. **squinted** pelérémòn *v*. **squinty** pelérémòn *v*.; wízìlìmɔ̀n *v*. **squinty-eyed** wízìlìmɔ̀n *v*. **squirm away** ɡwìrɔ̀n *v*. **squirmy** wʉlʉ́kʉ́mɔ̀n *v*. **squirrel (striped ground)** taráɗá *n*. **squirrel (tree)** luk<sup>a</sup> *n*. **squirt** ts'írítɔ̀n *v*. **squish** ɓirés *v*.; ɓírítésuƙot<sup>a</sup> *v*.; rɛɗɛ́s *v*.; rɨɗɛ́s *v*. **squished (get)** ɓirímón *v*. **squishily** ɓìr *ideo*.; ɲàl *ideo*.; rɔ̀ʝ ɔ *ideo*. **squishy** ɓirídòn *v*.; dulúdòn *v*.; duxúdòn *v*.; ɲalídòn *v*.; rɔʝɔ́dɔ̀n *v*. **stab** ɡafarɛs *v*.; ɡɛfɛrɛs *v*. **stab repeatedly** ɡafariés *v*. **stabile** diriɓóón *v*. **stability** ŋíkísila *n*.; ɲɛkɨsɨl *n*. **stabilize** ɨrʉ́rʉ́ɓɛ́s *n*.; ɨsílɔ́nʉƙɔt<sup>a</sup> *v*.; ƙaƙates *n*. **stable** ikékéɲòn *v*. **stack up** ɨnábɛ̀sʉ̀ƙɔ̀t a *v*.; ɨnábɛtɛ́s *v*. **stack up on** iɗóɗókés *v*. **stacked up** iɗóɗókánón *v*. **staff** ʝʉrʉm *n*. **staff (hook-necked)** ɲɛ́sɛɛɓɔ́*n*. **stagger** ɡakímón *v*.; ɨtɛrítɛ́rɔ̀n *v*.; nɛ́rɨnɛ́rɔ́n *v*. **stagnant** wàsɔ̀n *v*. **stair** lopemúím *n*. **stake** kìnòròt<sup>a</sup> *n*. **stale** cucuéón *v*. **stalk** arʉ́rʉ́bɔ̀n *v*.; kasír *n*.; morók<sup>a</sup> *n*.; tonyámón *v*.
steadily
stall
**stall** titikes *v*. **stammer** ɗɔkɔ́lɔ́mɔ̀n *v*.; ɨƙʉʝʉ́kʉ́ʝɔ̀n *v*. **stammering** ɡaʝádòn *v*.; kaŋádòn *v*. **stammering speech** ɡaʝádònìtòd<sup>a</sup> *n*. **stammeringly** ɡàʝ<sup>a</sup> *ideo*. **stamp** itirítírés *v*. **stamp down** iɲíkéésuƙot<sup>a</sup> *v*. **stamp pad** ɲezeí *n*. **stance** was *n*. **stand** taɗaŋes *v*.; wàsɔ̀n *v*. **stand (for nomination)** bɛ́ɗɛ́sa wasɔ́ ɛ *v*. **stand (of many)** ɡwámón *v*. **stand (of trees)** ɡwi *n*. **stand apart** iwásíòn *v*. **stand around** síbɔ̀n *v*.; towóón *v*.; towóónuƙot<sup>a</sup> *v*. **stand around (make)** síbɨtɛ́sʉ́ƙɔt<sup>a</sup> *v*. **stand around as a group** síbiónuƙot<sup>a</sup> *v*. **stand by each other** tɔmɛ́ínɔ́s *v*. **stand firm** itííròòn *v*. **stand out** kɛtɛ́lɔ́n *v*. **stand still** wasɔna ts'ír *v*.; wasɔnʉƙɔt<sup>a</sup> *v*. **stand up** ŋkáítetés *v*.; ŋkéétòn *v*.; ŋkóón *v*.; wasɨtɛs *v*.; wasítɛ́sʉƙɔt<sup>a</sup> *v*. **stand up (of many)** ɡwamétón *v*. **stand upright** wasɔna ts'ír *v*. **standard of living (high)** zɛƙwa ná dà *n*. **star** ɗɔ́xɛát<sup>a</sup> *n*. **star (evening)** Ɗɔ́xɛatá xìŋàtà<sup>e</sup> *n*. **star (morning)** Ɗɔ́xɛatá na baratsó<sup>e</sup> *n*.; Ɗɔ́xɛatá tsòònì *n*. **star (shooting)** ɗɔ́xɛatá na tsúwà *n*. **stare at** ŋɔ́zɛ̀s *v*.; ŋóziés *v*. **stare at each other** ŋɔ́zɨnɔ́s *v*. **stare at emptily** itelesa bàrìrrr *v*. **starer** ŋɔ́zɛ̀sìàm *n*.
**starling (blue-eared)** lɔɔmʉ́yá *n*. **start** iséétòn *v*.; isóón *v*.; itsyákétòn *v*.; toɗóón *v*. **start (fire)** ɡamés *v*.; tsapés *v*. **start a fight with** itoʝiés *v*. **start a fire** ɡamésá ts'aɗí *v*. **start early** ɛkwɛ́tɔ́n *v*. **start first** ɛkwɛ́tɔ́n *v*. **start off (a dance)** iwees *v*. **start raining** tosípón *v*. **starting point** wàxɛ̀d <sup>a</sup> *n*. **startle** iniŋíníŋés *v*.; ŋaxɨtɛtɛ́s *v*. **startle awake** tsídzètòn *v*. **startled** ŋaxɛ́tɔ́n *v*.; toúmón *v*. **starvation** ɲɔrɔƙɔ *n*. **starving** ɲɛƙánón *v*. **stash** irwanes *v*.; laɓ<sup>a</sup> *n*. **stash (small)** papaɗós *n*. **state** kíʝ<sup>a</sup> *n*. **state the verdict** tódetés *v*. **Stathmostelma peduncalatum** ŋímáarɔy<sup>a</sup> *n*. **station (missionary)** ɲémíxòn *n*. **stationary** diriɓóón *v*.; wàsɔ̀n *v*. **status** zeís *n*.; zeísínànès *n*. **stay** ʝɛ̀ʝɔ̀n *v*.; zɛ̀ƙwɔ̀n *v*. **stay a while** toƙízeesá así *v*.; toƙízòòn *v*.; tɔʉ́rʉ́mɔ̀n *v*. **stay behind** isíɗóòn *v*. **stay in** ínés *v*. **stay lying** towúryánòn *v*. **stay on** ʝɛʝɛ́tɔ́n *v*. **stay put** iɗúkóós *v*. **stay together** ínínós *v*. **stay-at-home person** awáám *n*. **STD** ɲamakaʝe *n*. **steadily** ɓa<sup>u</sup> *ideo*.
#### steady
**steady** ikékéɲòn *v*. **steal** dzuesés *v*.; dzuesetés *v*. **stealer** dzúám *n*. **stealthy** iɗásón *v*. **steam** lɔkapʉ́r *n*. **steep** ɨɔ́lɔ́lɔ̀n *v*.; kʉ́bɛ̀lɛ̀mɔ̀n *v*.; waatɛ́s *v*. **steep (dangerously)** iwósétòn *v*. **steer (a vehicle)** aŋɨrɛs *v*. **steer clear** ƙeƙérón *v*. **steering clear** firifíránón *v*.; wíríwíránón *v*. **steering wheel** ɲókokor *n*. **Steganotaenia araliacea** seɡer *n*. **stench** ɔn *n*.; ɔnɛd<sup>a</sup> *n*. **step** kímáts<sup>a</sup> *n*.; lopemúím *n*.; ɔkɔ́ts<sup>a</sup> *n*. **step all over** takwitakwiés *v*. **step off (measure)** ɨmaarɛ́sá dɛ̀ìkà<sup>ɛ</sup> *v*. **step on** takwés *v*. **Sterculia stenocarpa** ɡàràʝ<sup>a</sup> *n*. **Stereospermum kuntianum** seínení *n*. **sterile** ikólípánón *v*.; ɨsʉ́wɔ́ɔ̀n *v*.; osorosánón *v*. **sterile (animal or person)** ɲokólíp<sup>a</sup> *n*. **sterile person** òsòròs *n*. **sternum** ɡɔɡɔm *n*.; toroɓ<sup>a</sup> *n*. **sternutate** síƙón *v*. **sternutation** sìƙw<sup>a</sup> *n*. **stich** tʉfɛ́s *v*. **stick** ɗɛlɛ́mɔ́n *v*.; dakw<sup>a</sup> *n*.; ɨɗɔ́tsɔ́n *v*.; ɨnábɛs *v*.; ʝʉrʉm *n*.; ƙídzɔ̀n *v*.; nɔtsɔ́mɔ́n *v*.; pokés *v*.; sɛ̀w<sup>a</sup> *n*.; ts'ɔɗɨtɛs *v*. **stick (climbing)** ƙɔ̀ɗɔ̀t <sup>a</sup> *n*. **stick (honey)** tsɨtsín *n*. **stick (hooked)** lɔkɔ́ɗ <sup>a</sup> *n*. **stick (long digging)** ɲéɗiŋ *n*. **stick (metal-tipped)** ɲákálɨrɨkɨt<sup>a</sup> *n*.
**stick (mingling)** tʉɗʉtɛsídàkw<sup>a</sup> *n*. **stick (net-holding)** naƙwín *n*. **stick (round-headed)** ɲéɓiró *n*. **stick (small hooked)** pòròt<sup>a</sup> *n*. **stick around** iɗúkóós *v*.; toƙízòòn *v*. **stick in and out** ɨʝɔƙíʝɔ́ƙɛ́s *v*. **stick insect** tʉ́w<sup>a</sup> *n*. **stick out** sábʉ̀rʉ̀rɔ̀n *v*.; tɨbíɛ́tɔ̀n *v*. **stick out (of ears)** kweelémòn *v*. **stick out of sight** ɨkíɗítsɛ́s *v*. **stick out/up** lɛɛmɛ́tɔ̀n *v*. **stick ring** ɲókokor *n*.; ɲɔkɔlɔɓɛr *n*. **stick to** iɗupes *v*.; ɨnɔtsɛs *v*. **stick to (keep on)** ɗɛlɛ́mɔ́n *v*. **stick with** tɔmɛɛs *v*. **stick with (keep on)** ɗɛlɛ́mɔ́n *v*. **sticker** kàf *n*. **stickily** nɔ̀ts<sup>ɔ</sup> *ideo*. **sticking out** tɨbíɔ́n *v*. **sticky** ɨríítánón *v*.; mɨníkímɔ̀n *v*.; nɔtsɔ́dɔ̀n *v*. **stiff** ɓotsódòn *v*.; ɡɔkɔ́dɔ̀n *v*.; kɛtɛ́rɛ́mɔ̀n *v*.; tsɛ́rɛkɛ́kɔ́n *v*. **stiffen by stirring** tʉɗʉtɛtɛ́s *v*. **stiffly** ɓòts<sup>o</sup> *ideo*.; ɡɔ̀k ɔ *ideo*. **stifling (weather)** laŋádòn *v*. **still** ɨʝɛ́mɔ́n *v*.; lɛrɛ́dɔ̀n *v*.; wàsɔ̀n *v*. **still (become)** ɨʝɛ́mɔ́nʉƙɔt<sup>a</sup> *v*. **still be** sárón *v*. **still if** tònì *subordconn*. **stimulate digitally** ɨkɛɗíkɛ́ɗɛ́s *v*.; ɨkwatíkwátɛ́s *v*. **sting** áts'ɛ́s *v*.; ƙídzɛ̀s *v*. **sting (of pain)** ɓɛɨɓɛ́ɔ́n *v*. **sting thoroughly** áts'ɛ́sʉƙɔt<sup>a</sup> *v*. **stinger** kwan *n*. **stinginess** hábàs *n*.
**stingy** hábòn *v*.; mɨníkímɔ̀n *v*. **stink** ilíánòn *v*.; mídzòn *v*.; mídzona ɗɛtsɨɗɛ́tsík<sup>ɛ</sup> *v*.; mídzònà ɗùk<sup>u</sup> *v*. **stink bug (green)** loɡeréɲo *n*. **stinking** wízɨlílɔ́n *v*. **stinky** ɨmʉ́sɔ́ɔ̀n *v*. **stinky (become)** ɨmʉ́sɛ́ɛ̀tɔ̀n *v*. **stinky (very)** ɗùk<sup>u</sup> *ideo*. **stipple** iɗolíɗólés *v*. **stir** iƙures *v*.; ɨlɔ́lɔ́ŋɛ́s *v*. **stir (emotionally)** tábès *v*. **stir (restlessly)** ɨsɔ́sɔ́ŋɔ́s *v*. **stir around** ɨfáfáɲɛ́s *v*. **stir in** íburuburés *v*. **stirred stiff** tʉɗʉtɔs *v*. **stirred up** iƙúrúmós *v*. **stock (gun)** ɛ́bàdɛ̀ *n*. **stock-still** lɛrɛ́dɔ̀n *v*. **stocky** kikímón *v*. **stomach** ɡwà *n*. **stomach (first, of ruminants)** ɲɛ́pʉnʉk<sup>a</sup> *n*. **stomach (third, of ruminants)** ɲémékweɲ *n*. **stomach ache** áts'ɛ́sà bùbùì *n*. **stomach contents** ey<sup>a</sup> *n*. **stomach fluid (elephant)** lɔpɔ́ts<sup>a</sup> *n*. **stomp** itirítírés *v*. **stone** ɡwas *n*.; zébès *v*. **stone (cooking)** caál *n*. **stone (flat)** lalatíɓón *n*. **stone (hard black)** lokítoɲí *n*. **stone (sharpening)** lósùàɲ *n*. **stone (supporting)** caál *n*.; titirésíɡwàs *n*. **stone repeatedly** turues *v*.; turuetés *v*.
**stone that way** zébesuƙot<sup>a</sup> *v*. **stone this way** zébetés *v*. **stone-deaf** ɗɨnɔ́s *v*. **stone-still** lɛrɛ́dɔ̀n *v*. **stonily** ɡàts<sup>a</sup> *ideo*. **stony** ɡatsádòn *v*. **stool** kàràts<sup>a</sup> *n*. **stool (fecal)** ets'<sup>a</sup> *n*. **stool (loose)** eruxam *n*. **stool (three-legged)** dɛ̀ìkà àɗ<sup>e</sup> *n*.; lɔ̀cɛ̀ɡɛ̀r *n*. **stool (two-legged)** ɗìɗèsɔ̀k <sup>a</sup> *n*.; ɲámakuk<sup>a</sup> *n*. **stool carver** kárátsìkààm *n*. **stoop over** rɔ́rɔ́tɔ̀n *v*. **stooped** mʉƙʉ́rʉ́mɔ̀n *v*. **stooped over** ɡʉ́ɡʉ̀rɔ̀n *v*.; rʉ́ɡʉ̀ɗʉ̀mɔ̀n *v*. **stop** bɔlɨtɛ́sʉ́ƙɔt<sup>a</sup> *v*.; ɨmɔ́mɛ́tɔ̀n *v*.; mɔ́mɛ́tɔ̀n *v*.; wasɨtɛs *v*.; wasítɛ́sʉƙɔt<sup>a</sup> *v*.; wasɔ́áw<sup>a</sup> *n*.; wasɔnʉƙɔt<sup>a</sup> *v*. **stop (blowing or boiling)** tilímón *v*. **stop (plug)** tʉ́zʉɗɛ́s *v*. **stop beating** toɗúón *v*. **stop doing** bɔlɔnʉƙɔt<sup>a</sup> *v*. **stop hurting** toíónuƙot<sup>a</sup> *v*. **stop swarming (of termites)** ɡwɛɛ́ ts'ɛ́mɔ̀n *v*. **stop up** ɗɨnɛ́s *v*.; ɨmíɗítsɛ́s *v*.; tʉ́zʉɗɛ́sʉ́ƙɔt<sup>a</sup> *v*. **stopover** wasɔ́áw<sup>a</sup> *n*. **stoppage** was *n*. **stopped up** ɗɨnɔ́s *v*. **stopper** ts'ʉ́bʉlát<sup>a</sup> *n*.; ts'ʉ̂b <sup>a</sup> *n*. **stopping** was *n*. **storage hole** kùkùsèn *n*. **storage place** ɡirésíàw<sup>a</sup> *n*. **store** ɗʉkán *n*.; ɡirés *v*.; ɡirésíàw<sup>a</sup> *n*.; ɲáɗʉkán *n*.; ɲésitó *n*.; óɡoɗés *v*.
stripe
**store away** óɡoɗésúƙot<sup>a</sup> *v*.; oƙésúƙot<sup>a</sup> *v*. **storehouse** loɗúrú *n*.; lótsúm *n*. **storeroom** ɲésitó *n*. **storey** lopem *n*. **stork (Abdim's)** tsokôb<sup>a</sup> *n*. **stork-style dance** dikwa na tsokóbè *n*. **storm** itúúmés *v*. **storm (attack)** bóɡès *v*. **storm off** ɡwaítón *v*.; íɡwìʝìrɔ̀n *v*.; tʉlʉ́ŋɔ́n *v*.; tʉlʉ́ŋɔ́nʉƙɔt<sup>a</sup> *v*. **story** emut<sup>a</sup> *n*.; ɲáɗís *n*. **storyteller** emútíkààm *n*.; isíséésíàm *n*. **storytelling** emútík<sup>a</sup> *n*. **stout** laŋírímòn *v*.; laŋírón *v*. **stove (cooking)** ɲɛsɨŋƙɨrɨ *n*. **strabismic** kámáránón *v*.; ríbiribánón *v*. **straddle** dɛŋɛlɛsá dɛá<sup>ɛ</sup> *v*. **straight** ɨɗírɔ́n *v*.; ɨtɛ́nɔ́n *v*.; sʉrʉsʉ́rɔ́n *v*.; tɔɓɛ́ɔ́n *v*. **straight (horizontally)** isérérèòn *v*. **straight (set)** ɨtɛ́nɨtɛtɛ́s *v*. **straight (vertically)** iséréròn *v*. **straight away** ɗìr *adv*. **straight part** ɡòɡòròʝ<sup>a</sup> *n*. **straighten** ɨɗírítɛ́sʉƙɔt<sup>a</sup> *v*.; ɨtɛ́nɨtɛtɛ́s *v*.; ƙɔɛ́s *v*.; ƙɔƙatés *v*. **strain** ɨɗíɲɔ́n *v*.; ɨʝɨwɛs *v*.; ɨtɨwɛs *v*. **strain (muscles)** ɗukés *v*. **strainer** ɲékeikéy<sup>a</sup> *n*. **strange talk** kínítòd<sup>a</sup> *n*. **stranger** kɔ́nɔ́m *n*. **strangers** kíníám *n*. **strangle** iketiés *v*. **strap** ƙíw<sup>a</sup> *n*. **strap across** ízokomés *v*. **strasbismal** pɨlírímɔ̀n *v*.
**straw (drinking)** ɲálamorú *n*.; ɲeɓune *n*. **stray** iwórón *v*.; iwórónìàm *n*. **stray off** hakonuƙot<sup>a</sup> *v*. **streaked** ɨlíŋánètòn *v*.; ɨlíŋɔ́n *v*. **stream** ƙídɨƙídɔ̀n *v*. **stream (large)** ɔrɔr *n*. **stream (small)** ìàwìàw<sup>a</sup> *n*. **stream out** furúdòn *v*. **streaming out** fùr *ideo*. **streamlined** mɨlɔ́dɔ̀n *v*. **strength** ŋɡúf *n*.; ŋɨxás *n*.; ɲaƙóƙóŋ *n*. **strengthen** ŋɨxítɛ́sʉƙɔt<sup>a</sup> *v*.; ŋɨxɔnʉƙɔt<sup>a</sup> *v*. **stressed out (become)** ɨlárímétòn *v*.; ɨlwárímétòn *v*. **stretch** eminiés *v*.; ƙɔɛ́s *v*.; ƙɔƙatés *v*.; ƙɔƙɔanón *v*. **stretch across** ɨkámárɛ́s *v*. **stretch out (to rest)** torwóónuƙot<sup>a</sup> *v*. **stretched out (resting)** torwóón *v*. **strew** ɨɗɛrɛs *v*. **strew about** ɨɗɛríɗɛ́rɛ́s *v*. **strewn** ɨɗɛrɔs *v*.; kazaanón *v*. **strewn about** apɛ́tɛ́pɛ́tánón *v*.; ɗɛtɛ́ɗɛ́tánón *v*.; ɨɗɛríɗɛ́rɔ́s *v*. **stride** dɛŋɛlɛsá dɛá<sup>ɛ</sup> *v*.; ɔkɔ́ts<sup>a</sup> *n*. **strife** ɲéƙúruƙur *n*.; ɲɛ́píɗɨpɨɗ<sup>a</sup> *n*. **strike** ɗálútés *v*.; iwés *v*.; iwésúƙot<sup>a</sup> *v*.; toɓés *v*. **strike (a match)** dzɛrɛ́s *v*. **striking** ɨɗɛ́ɔ́n *v*. **string** rɔ́ɛ́s *v*.; sim *n*. **string (nylon)** ɲákol *n*. **stringy** simánón *v*. **strip** dzeretiés *v*.; dzeretiésuƙot<sup>a</sup> *v*.; ɨtakɛs *v*. **strip off** ɨɓɔ́lɔ́tsɛ́s *v*.; ɨɓɔtɛs *v*.; ɨtákɛ́sʉƙɔt<sup>a</sup> *v*.; tɔɲílíɲílɛ́s *v*. **stripe** dzeretiés *v*.
**striped** ɡweɡweritiós *v*.; ɨƙɨrɔs *v*.; ɨlíŋánètòn *v*.; ɨlíŋɔ́n *v*.; wíziwizatós *v*. **striped down the spine** kɔlánétòn *v*. **stroke** ɨwáwɛ́ɛ́s *v*. **stroke affectionately** ɨɓɔníɓɔ́nɛ́s *v*.; iɓoníɓóniés *v*. **stroll** ɨtɛ́mɔ́ɔ̀n *v*.; tasɔ́ɔ́n *v*.; zíbòn *v*. **strong** ŋìxɔ̀n *v*. **stronger (become)** ŋɨxɔnʉƙɔt<sup>a</sup> *v*. **struggle** kóríètòn *v*.; kɔrɔanón *v*. **struggle against** cɛ̀mɔ̀n *v*. **struggle for** ŋués *v*. **struggle into** ɨɓɨtsíɓítsɛ́s *v*. **struggle over** ŋués *v*. **strung** rɔ́ɔ́s *v*. **strut** ɨƙɔ́ɔ́rɛ́sá así *v*.; ƙɔ̀rɔ̀n *v*. **stub (grass)** rumurúm *n*. **stubble (plant)** sʉ́s *n*. **stubborn** ɨɗíkílɔ̀n *v*. **stubby** ŋɨríɓímɔ̀n *v*.; poŋórómòn *v*. **stubby-toothed** ŋɨríɓímɔ̀n *v*. **stuck** ɨpɔ́kɔ́n *v*. **stuck (become)** bokímón *v*. **stud** cúrúk<sup>a</sup> *n*. **student** isóméésíàm *n*.; ɲósomáám *n*. **studies** ɲósomá *n*. **study** isóméés *v*.; tɨtɨmɛs *v*. **stuff** ɨsɨkɛs *v*.; kúrúɓáicík<sup>a</sup> *n*.; mɛnáícík<sup>a</sup> *n*.; rʉtsɛ́s *v*.; rʉtsɛ́sʉ́ƙɔt<sup>a</sup> *v*. **stuffed** itéɓúkòn *v*. **stumble** rúmánòn *v*. **stumble (make)** ɨlɛ́ƙwɛ́rɛ́s *v*.; ɨlɛ́ƙwɛ́rɛtɛ́s *v*. **stumble ahead** ɓɛƙɛ́sá turúùk<sup>e</sup> *v*. **stumble repeatedly** iɲatiesá kíʝá<sup>e</sup> *v*. **stump** ɡɔn *n*.
**stun** ɨrakɛs *v*.; ɨrákɛ́sʉƙɔt<sup>a</sup> *v*. **stunned** ɨʝárɔ́n *v*.; ʝarámétòn *v*. **stunted** rɛƙɛ́ɲɛ́mɔ̀n *v*.; sɛ́rɛ́ƙɛƙánón *v*. **stunted growth** lɔɓʉ́kɛʝɛ́n *n*. **stupefied** tɔmɛrímɛ́rɔ̀n *v*. **stupid** ɨɓááŋɔ̀n *v*. **stupid person** bóx *n*.; ɨɓááŋàsìàm *n*. **stupidity** ɨɓááŋàs *n*. **sturdy** ikékéɲòn *v*. **stutter** ɗɔkɔ́lɔ́mɔ̀n *v*.; ɨƙʉʝʉ́kʉ́ʝɔ̀n *v*. **stuttering** ɡaʝádòn *v*.; kaŋádòn *v*. **stutteringly** ɡàʝ<sup>a</sup> *ideo*. **stylish** titianón *v*. **stymie** ɨɓatɛs *v*. **stymie repeatedly** ɨɓatíɓátɛ́s *v*. **subcounty** ɲásáɓúkáúntì *n*. **subdue** ɨɗáfɛ́sʉƙɔt<sup>a</sup> *v*.; itikes *v*. **subdued** ɨɗáfɛ́sʉƙɔta así *v*. **submerge** ilumes *v*.; ilúmésuƙot<sup>a</sup> *v*. **subparish** ɲásáɓúpárìx *n*. **subsist** topíánètòn *v*. **substitute** imetsités *v*. **substitute for** imetsés *v*. **subtract** ƙanésúƙot<sup>a</sup> *v*. **subversion** ɲéƙúruƙur *n*. **subverter** ɲɛkɛsʉpan *n*. **succeed** ɨlámɔ́n *v*. **success** ídzànànès *n*. **succession** ɲɛɗʉpɛ *n*. **suck** ƙʉɗɛ́s *v*. **suck on** ƙʉɗɛ́s *v*. **suck on each other** ƙʉ́ɗʉ́nɔ́s *v*. **suck out** ƙʉɗɛtɛ́s *v*.; ts'ʉ́ts'ʉ́tɛ́s *v*.; ts'ʉ́ʉ́tɛ́s *v*. **suck up** ts'ʉ́ts'ʉ́tɛ́s *v*.; ts'ʉ́ʉ́tɛ́s *v*. **sucked dry** ts'ʉ́ʉ́tɔnʉƙɔt<sup>a</sup> *v*. **sucked up** ts'ʉ́ʉ́tɔnʉƙɔt<sup>a</sup> *v*.
**suckle** naƙwɛ́s *v*.; naƙwɛ́sʉ́ƙɔt<sup>a</sup> *v*.; naƙwɨtɛs *v*. **suckling** ɗiak<sup>a</sup> *n*. **Sudan (South)** Sʉɗán *n*. **Sudan gum arabic** ɗerét<sup>a</sup> *n*.; lofílitsí *n*. **Sudanese rebels** Ŋɨɲɛ́ɲɛ́y <sup>a</sup> *n*. **suddenly** ŋàm *ideo*.; ùrùƙùs *ideo*. **suffer** ɨríɗɔ́n *v*. **suffer internally** ɗuƙúkón *v*. **suffer quietly** ɗuƙúkón *v*. **suffering** tawanímétòn *v*. **sufficient** ŋábɔnʉƙɔt<sup>a</sup> *v*.; nábɔnʉƙɔt<sup>a</sup> *v*. **suffocate** tʉɓʉnɛ́s *v*.; tuɓunímétòn *v*. **sugar** ɲósukarí *n*. **sugar ant** ɗɔ́ɡɨɗɔ̂ɡ <sup>a</sup> *n*. **sugar bush** ɲícwéɲé *n*. **sugary** diridírón *v*. **suit (legal)** ɲékés *n*. **suitable** itémón *v*. **Suk person** Ŋúupéám *n*. **sulk** ɨɓʉ́tʉ́ŋɔ̀n *v*.; imutúmútòn *v*.; siŋírón *v*. **sulky** ɨmʉtʉ́mʉ́tɔ́s *v*. **sultry (weather)** laŋádòn *v*. **sum up** bɔsɛtɛ́s *v*.; ɗɔtsɛ́sʉ́ƙɔta mɛná<sup>ɛ</sup> *v*. **summarize** bɔsɛtɛ́s *v*.; ɗɔtsɛ́sʉ́ƙɔta mɛná<sup>ɛ</sup> *v*.; ɨtsʉnɛtɛ́s *v*. **summit** kwaráɡwarí *n*. **summon** iríréetés *v*.; óés *v*. **summon by whistling** iwéwérés *v*.; iwówórés *v*. **summon here** oetés *v*. **sun** fet<sup>a</sup> *n*. **sun range** fetíhò *n*. **sun watcher** itelesíáma fetí *n*. **sun-watching point** itelesíáwa fetí *n*.
**sunbeam** fetíbàs *n*. **sunbird** itsók<sup>a</sup> *n*. **Sunday** Ɲásaɓét<sup>a</sup> *n*. **sunflower** ɲɛ́kɨɗɛkɨɗɛ́*n*.; ɲɛ́tɔɔkíɗɛ́*n*. **sunglasses** ɲékiyóika ni fetí *n*. **sunray** fetíbàs *n*. **sunrise** pɛlɛ́mɔ́na fetí *v*. **superiority** zeís *n*. **supervise** ɨrɨtsɛ́s *v*. **supper** ŋƙáƙá na wídzò<sup>e</sup> *n*. **supple** ʝaulímòn *v*.; tsutsukes *v*. **supplication** wáán *n*. **supply** íɡɔɲɛ́s *v*. **support** ɨƙaŋɛs *v*.; titirés *v*.; titiretés *v*.; tɔmɛɛs *v*.; wasɨtɛs *v*. **support each other** tɔmɛ́ínɔ́s *v*. **suppress** ɨɗáfɛ́sʉƙɔt<sup>a</sup> *v*.; itikes *v*. **suppressed** ɨɗáfɛ́sʉƙɔta así *v*. **suppurating** tatifíánón *v*. **suppuration** báts'<sup>a</sup> *n*. **suprapubic area** heʝú *n*. **surefooted** tsɛ́rɛkɛ́kɔ́n *v*. **surface** takánétòn *v*. **surgeon** hoesíàm *n*. **surgery** hoesího *n*. **surgery (perform)** hoés *v*.; hoetés *v*. **surpass** ɨlɔɛs *v*.; ɨlɔɛtɛ́s *v*.; ɨsʉkɛs *v*.; sʉ́kɛ́s *v*. **surpass in height** ileŋes *v*. **surprise** bóɡès *v*.; itúúmés *v*. **surrender** taʝales *v*.; taʝálésuƙot<sup>a</sup> *v*.; taʝaletés *v*. **surround** ɨrɨkɛs *v*.; ɨríkɛ́sʉƙɔt<sup>a</sup> *v*.; ɨrɨkɛtɛ́s *v*.; itsóɗón *v*. **surrounding (of prey)** nakítsòɗ<sup>a</sup> *n*. **surveil** rɔ́tɛ́s *v*.; toreɓes *v*. **survey** ɦyeités *v*. **survival** ɦyekes *n*.
**survive** ʝɛ̀ʝɔ̀n *v*.; pʉ̀rɔ̀n *v*.; topíánètòn *v*. **survive (a mishap)** ɨsɛɛs *v*. **suspect each other** ɨpíʝíkɨmɔ́s *v*. **suspend in air** alólóánitetés *v*.; alólóés *v*. **swagger** ɨƙɔ́ɔ́rɛ́sá así *v*.; ƙɔ̀rɔ̀n *v*. **Swahili language** Ŋákiswahílìtòd<sup>a</sup> *n*. **swallow** luƙés *v*. **swallow (bird)** loménio *n*. **swamp** ɲéʝem *n*. **swamp (seasonal)** ɲɛkípɔ́r *n*.; ɲotóbòr *n*. **swampily** fɔ̀ts'<sup>ɔ</sup> *ideo*. **swampy** fɔts'ɔ́dɔ̀n *v*. **swap** ilókótsés *v*.; ɨxɔtsɛs *v*.; xɔ́tsɛ́s *v*. **swarm of bees (mobile)** ts'ɨƙábòt<sup>a</sup> *n*. **swarm over** iwówéés *v*. **sway** ɲɛsʉp<sup>a</sup> *n*.; sʉ́bɛ̀s *v*.; sʉ́bɛsʉƙɔt<sup>a</sup> *v*. **sway gently** ʝikiʝíkón *v*. **swear** ikóŋón *v*. **swear (make)** ikóŋítetés *v*. **swear an oath** tsamɛtɛ́sá ikóŋónì *v*. **sweat** kirot<sup>a</sup> *n*.; kirotánón *v*. **sweep** séɓés *v*. **sweep aside** ɨpalípálɛ́s *v*. **sweep away** séɓésuƙot<sup>a</sup> *v*. **sweep off** séɓésuƙot<sup>a</sup> *v*. **sweep up** séɓetés *v*. **sweeper** séɓésìàm *n*. **sweet** diridírón *v*.; ɡwéts'ón *v*.; ɲátamɨtám *n*. **sweet (slightly)** nɨƙwídɔ̀n *v*. **sweet potato** ɲakaíta *n*. **sweet potato leaves** ɲakaítákák<sup>a</sup> *n*. **sweet-and-sour** mɨtɨmítɔ́n *v*.; taasámòn *v*. **sweet-smelling** tukukúɲón *v*.
**sweet-talk** ɨmámɛ́ɛ́s *v*. **sweeten** ɛfɨtɛs *v*. **sweetly** nìƙw<sup>ɨ</sup> *ideo*. **swell** èmòn *v*.; itéɓúkòn *v*.; tʉ̀wɔ̀n *v*.; xuanón *v*.; xuxuanón *v*. **swell (make)** emites *v*. **swell (of many)** emitaakón *v*. **swell up** ʝɨríʝírɛ̀tɔ̀n *v*.; ʝɨríʝírɔ̀n *v*.; tʉwɛ́tɔ́n *v*. **sweltering** ririanón *v*. **sweltering (become)** ririanétòn *v*. **swerve** iwítón *v*. **swerve repeatedly** aŋiriesón *v*.; iwitíwítòn *v*. **swift (bird)** loménio *n*. **swift (white-rumped)** tsòrìàm *n*. **swim** ɨlʉ́lʉ́mʉ̀ɔ̀n *v*. **swindle** ɨmɔɗɛs *v*. **swine** ɲéɡuruwé *n*. **swing** aƙóláánón *v*.; lóƙólíl *n*. **swing by** ɨɛ́bɛ̀s *v*.; ɨɛ́bɛsʉƙɔt<sup>a</sup> *v*.; ɨɛ́bɛtɛ́s *v*. **swing side to side** iŋolíŋólés *v*. **swipe** ɨɛ́bɛsʉƙɔt<sup>a</sup> *v*.; ɨɛ́bɛtɛ́s *v*.; ɨpakɛs *v*.; toyeres *v*. **swipe away** ɨpákɛ́sʉƙɔt<sup>a</sup> *v*. **swipe clean** ɨƙʉ́ʉ́lɛ́s *v*. **swipe off** ɨpákɛ́sʉƙɔt<sup>a</sup> *v*. **swirl** ɨwaríwárɛ́s *v*. **swirl up** tɔpɨrípírɔ̀n *v*. **swish** ɨɗííɗɛ́s *v*. **swish (mouth)** íɡʉʝʉɡʉʝɛ́s *v*.; ɨmʉ́mʉ́ʝɛ́s *v*. **swish swish** kòrrr *ideo*. **swish!** swèèè *ideo*. **switch** aeitetésíàw<sup>a</sup> *n*.; ɨʝʉlɛs *v*.; ɨʝʉlɛtɛ́s *v*.; kíxw<sup>a</sup> *n*.
**switch (whip)** ízaɓɨzaɓɛ́s *v*.; zaɓatiés *v*. **switch off (electrically)** ts'eítésuƙot<sup>a</sup> *v*. **switch off (of electricity)** ts'oonuƙot<sup>a</sup> *v*. **switch on (electrically)** aeitetés *v*. **switch on (of electricity)** aeétón *v*. **switch!** ɗì *ideo*. **switched** ɨʝʉlɔs *v*. **swollen** dúduránón *v*.; itéɓúkòn *v*. **swoon** rèŋòn *v*. **swoosh** ɨɗííɗɛ́s *v*. **symmetrical** ikwáánòn *v*. **sympathize** cucuéétòn *v*. **synchronize** ilíráitetés *v*. **synchronized** ilíróòn *v*. **table** ɲéméza *n*. **tableland** lopem *n*. **tablet** cɛ̀mɛ̀rìèkw<sup>a</sup> *n*. **tabletop** ɲémézaɡwarí *n*. **taboo** itáléánón *v*.; itálóós *v*.; ɲatal *n*. **taboo (make)** itáléés *v*. **taboo of eating first** isóón *n*. **taboo of eating prematurely** ifófóés *n*. **taboo of leg meat** dɛ *n*. **taboo of sitting on elders' stools** zɛƙwɔna karatsɔɔ ʝáká<sup>e</sup> *n*. **taboo of watching mother-in-law** itelesa ceŋetíámà<sup>e</sup> *n*. **taboo of water** cue *n*. **tadpole** ŋʉ́ɗʉ́ŋʉ́ɗ <sup>a</sup> *n*. **tag** pilís *n*. **tahini** ɲówoɗí *n*. **tail** ikókótés *v*.; timóy<sup>a</sup> *n*. **tail (chicken)** bòrèn *n*. **tail (of a bird)** tsʉ́ɓ <sup>a</sup> *n*. **tail-hair** lɔ̀d <sup>a</sup> *n*.
**tail-tip (of a python)** ɲéɡets<sup>a</sup> *n*. **tailor** tʉfɛ́s *v*.; tʉfɛ́síàm *n*. **take (somewhere)** duƙésúƙot<sup>a</sup> *v*. **take (swallow)** béberetés *v*. **take a break** ɨɛ́ŋɔ́nʉƙɔt<sup>a</sup> *v*.; ɨmɔ́mɛ́tɔ̀n *v*.; mɔ́mɛ́tɔ̀n *v*. **take a diversion** wɛ́dɔ̀n *v*. **take a loss** totóánonuƙot<sup>a</sup> *v*. **take a picture** iwetésá ɲɛ́pítsaá<sup>ɛ</sup> *v*. **take a picture of** iwetés *v*. **take a seat** zɛƙwɛ́tɔ́n *v*. **take a shot at** ídzesuƙot<sup>a</sup> *v*. **take a sip** abʉtɛtɛ́s *v*.; tsʉɓɛtɛ́s *v*. **take a trip** ɨlɔ́ɔ́n *v*. **take advantage of** ɨnɔmɛtɛ́s *v*. **take aim** ɨɗírɔ́n *v*. **take all of** ɨsʉɲɛs *v*. **take an oath** ikóŋón *v*.; tsamɛtɛ́sá ikóŋónì *v*. **take apart** ɨʝʉƙʉ́ʝʉ́ƙɛ́s *v*. **take away** ɨákɛ́sʉƙɔt<sup>a</sup> *v*.; ƙanésúƙot<sup>a</sup> *v*. **take away all of** ɨsʉ́ɲɛ́sʉƙɔt<sup>a</sup> *v*. **take away gingerly** ɗítɛ́sʉƙɔt<sup>a</sup> *v*. **take back** raʝésúƙot<sup>a</sup> *v*. **take by force** ɛ́nɛ́sʉƙɔt<sup>a</sup> *v*.; toreɓes *v*. **take by surprise** bóɡès *v*.; itúúmés *v*. **take care of** ɨrɨtsɛ́s *v*. **take far away** ɨɛƙítɛ́sʉƙɔt<sup>a</sup> *v*. **take flight** bʉrɛ́tɔ́n *v*. **take for a walk** ɨláítɛ́sʉƙɔt<sup>a</sup> *v*.; ɨtátɛ́ɛ́s *v*. **take form** ituetésá así *v*. **take gingerly** ɗítɛ́s *v*. **take hold of** ɨkamɛ́sʉ́ƙɔt<sup>a</sup> *v*.; ɨkamɛtɛ́s *v*.; ƙanetés *v*.; ŋusés *v*.; ŋusésúƙot<sup>a</sup> *v*. **take hold of each other** ɨkámʉ́nɔ́sʉ́ƙɔt<sup>a</sup> *v*. **take in** ɓuƙítésuƙot<sup>a</sup> *v*. **take in hand** ƙanés *v*.
**take medicine** wetésá cɛmɛ́ríkà<sup>ɛ</sup> *v*. **take nearly all** kɔnɨtɛtɛ́s *v*. **take note of** ewanetés *v*. **take note off** ewanes *v*. **take off** ɗóɗésa muceé *v*.; hoɗésúƙot<sup>a</sup> *v*.; hoɗetés *v*.; laʝetés *v*.; tuɓutes *v*.; tuɓútésuƙot<sup>a</sup> *v*. **take off (running)** ɨpʉ́tɛ́sʉƙɔta así *v*.; tsídzonuƙot<sup>a</sup> *v*. **take off flying** bʉrɛ́tɔ́n *v*.; bʉrɔnʉƙɔt<sup>a</sup> *v*. **take off hopping** itseɗítséɗòn *v*. **take off shoes** hoɗetésá taƙáíkà<sup>ɛ</sup> *v*. **take office** itsyákétòn *v*. **take on credit** iɗenes *v*.; iɗenetés *v*. **take out** ƙanetés *v*.; pulúmítésúƙot<sup>a</sup> *v*.; ts'álés *v*.; ts'aletés *v*.; tuɓutetés *v*. **take over for** imetsés *v*. **take place** ikásíìmètòn *v*.; itíyáìmètòn *v*. **take shape** ituetésá así *v*. **take shelter in** rɨmɛ́s *v*. **take time** iríóòn *v*. **take to court** wasɨtɛs *v*.; wasítɛ́sʉƙɔt<sup>a</sup> *v*. **take to pasture** waitetés *v*. **take up** tɛ́bɛ̀s *v*. **taken off guard** toúmón *v*. **tale** emut<sup>a</sup> *n*. **talent** akílìk<sup>a</sup> *n*. **talk** ɨɛ́nɛ́tɔ̀n *v*.; ɨɛ́nɔ́n *v*.; tôd<sup>a</sup> *n*.; tódòn *v*. **talk (get to)** ɨɛ́nítɛtɛ́s *v*.; tóítetés *v*. **talk (reckless)** múɗúkánónìtòd<sup>a</sup> *n*. **talk about** mɔ́ɲɛ́s *v*.; tódetés *v*. **talk at once (crowds)** réɡiréɡòn *v*. **talk foreignly** ɨɗímɔ́n *v*. **talk straight to the issue** ɨtɛ́nɨtɛtɛ́sá tódà<sup>e</sup> *v*. **talk to each other** tódinós *v*.
**talkative** ɗɛmɛ́dɔ̀n *v*.; ɗɛmɨɗɛ́mɔ́n *v*.; ɨkʉ́tʉ́kánón *v*.; poxódòn *v*. **talkatively** ɗɛ̀m *ideo*.; pòx *ideo*. **talker** akáám *n*.; tódààm *n*. **tall** zikíbòn *v*. **tall (grow)** tɔwʉ́tɔ́n *v*.; zikíbonuƙot<sup>a</sup> *v*. **tall (make)** zikíbitésúƙot<sup>a</sup> *v*. **tall (of many)** zikíbaakón *v*. **tall anthill** kìts<sup>a</sup> *n*. **tallness** zikíbàs *n*. **tallow** ɲéɗíol *n*. **talon** tíbòlòkòɲ *n*. **tamarind** rɔ́ƙɔ́*n*. **tamarind seeds** ɗɛ̂ɡ <sup>a</sup> *n*. **Tamarindus indica** rɔ́ƙɔ́*n*. **tame** bɔnɛ́s *v*. **tamp** ɨɗɛŋɛs *v*.; itirítírés *v*. **tamp repeatedly** ɨɗɛŋíɗɛ́ŋɛ́s *v*. **tamped down (become)** iɗéŋímètòn *v*. **tan** ŋirotsánón *v*. **tangible** tabam *n*. **tangle** imóɲíkees *v*. **tangle up** imóɲíkeetés *v*. **tangled** sáɡoanón *v*. **tangy** mázɨmázɔ̀n *v*. **tank (military)** ɡaso *n*. **tap on** ɨtɔ́tɔ́ŋɛ́s *v*. **tap out** ɨtɔŋítɔ́ŋɛ́s *v*. **tap repeatedly** ɨɗɛɨɗɛ́ɛ́s *v*. **tapeworm** apéléle *n*. **Tarenna graveolens** tsɛ̀tsɛ̀kw<sup>a</sup> *n*.; tsɨkw<sup>a</sup> *n*. **target** ɨpɨmɛs *v*. **tarp** ɲéema *n*. **tarry** tɔʉ́rʉ́mɔ̀n *v*.
**tart** ɓariɓárón *v*.; ɓárikíkón *v*.; ɓàròn *v*.; mázɨmázɔ̀n *v*. **task** ɲákási *n*.; ɲetits<sup>a</sup> *n*.; terêɡ<sup>a</sup> *n*. **tassel (animal-tail)** lɔ̀d <sup>a</sup> *n*. **tassel (giraffe-tail)** ɡwaíts'ílɔ̀d <sup>a</sup> *n*. **tassel (of maize)** kâʒw<sup>a</sup> *n*. **taste** kaites *v*. **tasteless** ɗɛ̀ƙwɔ̀n *v*.; ʝɔ̀lɔ̀n *v*.; muʝálámòn *v*. **tasty** ɛ̀fɔ̀n *v*.; ɡwéts'ón *v*. **tasty (become)** ɛfɔnʉƙɔt<sup>a</sup> *v*. **tattered** rídziridzánón *v*. **tattily** rɛ̀s *ideo*. **tattle on** ilíítés *v*. **tattoo** ɨtsɨpítsípɛ́s *v*. **tatty** rɛsɛ́dɔ̀n *v*. **tawny** ŋirotsánón *v*. **tax** ɲéutsúr *n*. **tax collector** ɲéútsuríám *n*. **tchagra** kíɗɔ̀ *n*. **tea** ɲécáy<sup>a</sup> *n*. **tea (African)** tábarɨcue *n*. **tea (black)** kotímácùè *n*. **tea (milk)** tábarɨcue *n*. **teach** ɨtátámɛ́s *v*.; nɔɔsanitetés *v*. **teach to read** isómáitetés *v*. **teacher** ɨtátámɛ́síàm *n*.; ŋímaalímùàm *n*. **teacher (head)** ámázeáma ɲésukúluⁱ *n*. **team** ɲétím *n*. **tear** ɗusés *v*.; ɗusúmón *v*.; ɗusutes *v*.; dzɛrɛ́s *v*.; ɨkɛ́ŋɛ́ɗɛ́s *v*.; tɔŋɛɗɛs *v*. **tear off** ɗusésúƙot<sup>a</sup> *v*.; dzeretiés *v*.; dzeretiésuƙot<sup>a</sup> *v*.; dzɛ̀rɔ̀n *v*. **tear off (running)** tsídzonuƙot<sup>a</sup> *v*. **tear off in strips** tɔɲílíɲílɛ́s *v*. **tear out** ruutésuƙot<sup>a</sup> *v*.; ruutetés *v*.
**teardrop rain** ekúcúédidí *n*. **tears** ekúcé *n*. **tease** ceŋánón *v*. **teaser** céŋáàm *n*. **teat** îdw<sup>a</sup> *n*. **technical school** tɛ́kɛ̀nìkɔ̀l *n*. **teclea (small fruited)** kɛ́láy<sup>a</sup> *n*.; ɲɛmaɨlɔŋ *n*. **Teclea nobilis** kɛ́láy<sup>a</sup> *n*.; ɲɛmaɨlɔŋ *n*. **teem** ƙídɨƙídɔ̀n *v*. **teem around** iwówéés *v*. **teetering** nɛrɛ́dɔ̀n *v*. **teeteringly** nɛ̀rɛ̀ *ideo*. **telephone** ɲásím *n*. **television** kúrúkúríka ni ɓɛƙɛ́s *n*.; ɲévíɗyo *n*. **tell** ɦyeitésúƙot<sup>a</sup> *v*.; ɦyeitetés *v*.; isíséés *v*.; tódètòn *v*.; tódòn *v*. **tell apart** ɦyeités *v*. **tell each other** tódinós *v*. **tell on** ilíítés *v*. **tell the time** ɗoɗésúƙota ɲásáatí *v*.; itelesa fetí *v*. **tell the truth** ɨtsírɔ́na tódàk<sup>e</sup> *v*. **teller** tódààm *n*. **telling** tôd<sup>a</sup> *n*. **temperature (high)** hábona nébwì *n*. **tempermental** kwits'íkwíts'ánón *v*. **temple area (of head)** matáŋ *n*. **temple area (upper)** matáŋíɡwarí *n*. **temple bone** matáŋíɔ̀k <sup>a</sup> *n*. **tempt** sʉ́bɛ̀s *v*.; sʉ́bɛsʉƙɔt<sup>a</sup> *v*.; sʉ́bɨtɛ́sʉ́ƙɔta así *v*. **tempter** sʉ́bɛ̀sìàm *n*. **ten** toomín *n*. **ten o'clock** ɲásáatɨkaa ts'aɡúsátìk<sup>e</sup> *n*. **tenacious** ɨkázànòòn *v*. **tend (garden)** kɔɛ́s *v*.
**tend (livestock)** cookés *v*. **tend (to do)** ɨtáŋátɔ̀n *v*. **tended** cookotós *v*. **tender** rɛɓɛ́dɔ̀n *v*.; rɛɗɛ́dɔ̀n *v*.; rusúdòn *v*. **tender (of plants)** xɛɓɛ́dɔ̀n *v*. **tenderly** dàb<sup>u</sup> *ideo*.; rɛ̀ɗ ɛ *ideo*.; rùs *ideo*. **tenderly (of plants)** xɛ̀ɓ ɛ *ideo*. **tendon** kon *n*. **tendon (Achilles)** tɨtíʝíkòn *n*. **tent** ɲéema *n*. **tenth (be a)** kɔnɔna toomínú *v*. **Tepeth language** Ŋítépesítôd<sup>a</sup> *n*. **Tepeth people** Ŋítépes *n*. **tercel** ɡwácúrúk<sup>a</sup> *n*. **teres major** ɡuféém *n*. **term (school)** ɲátám *n*. **term of endearment** mínɛ́sìèd<sup>a</sup> *n*. **Terminalia brownii** ɡáʒàd<sup>a</sup> *n*. **terminate** hoɗésúƙot<sup>a</sup> *v*.; iɲétséetés *v*.; iyétséetés *v*. **termite (early-flying)** erún *n*. **termite (edible noctural)** mukúádàŋ *n*. **termite (edible)** dáŋ *n*. **termite (first eaten portion)** kʉ́tʉ́k <sup>a</sup> *n*. **termite (small worker)** sokomet<sup>a</sup> *n*. **termite (soldier)** lókók<sup>a</sup> *n*.; tɛƙɛram *n*. **termite (tiny)** léts<sup>a</sup> *n*. **termite colony (active)** abér *n*. **termite colony (inactive)** wàrɔ̀t <sup>a</sup> *n*. **termite column** dáŋámorók<sup>a</sup> *n*. **termite dirt** dáŋáʝʉ̀m *n*. **termite housing** dáŋáhò *n*. **termite mound** kutút<sup>a</sup> *n*. **termite mound (holey)** lòkòsòs *n*. **termite mound (old)** dáŋákìts<sup>a</sup> *n*.
**termite mound base** dáŋádɛ̀ *n*. **termite mound chamber** ɓarán *n*. **termite opening** dáŋáàk<sup>a</sup> *n*. **termite outlet** dáŋéèkw<sup>a</sup> *n*. **termite paste (edible)** másálúk<sup>a</sup> *n*. **termite queen** dádata dáŋá<sup>e</sup> *n*.; dáŋádadát<sup>a</sup> *n*.; ŋwááta dáŋá<sup>e</sup> *n*. **termite rain** dáŋádidí *n*. **termite season** dáŋátsóy<sup>a</sup> *n*. **termite soil** kerets'<sup>a</sup> *n*. **termite trap** akarér *n*. **termite wings** síts'<sup>a</sup> *n*. **termite worker** nateɓú *n*. **termite(s)** ɛs *n*. **termite-and-honey dish** ƙɛƙɛram *n*. **termite-drying mat** uré *n*. **termites (dried, wingless)** tɔƙam *n*. **termites (first pot eaten)** wàxìdòm *n*. **termites (late-flying)** ɓɛʝɛ́kw<sup>a</sup> *n*. **termites (pounded)** iwótsíɔ̀z *n*. **terrapin** sídilé *n*. **Teso person** Ŋítésóàm *n*. **test** esetés *v*.; iniŋes *v*.; ɨpɨmɛs *v*.; kaites *v*.; ɲɛ́tɛ́sìt<sup>a</sup> *n*. **testicle** míts'<sup>a</sup> *n*. **testify to** itsáɗénés *v*. **testimony (false)** ɲɔ́pɔkɔca *n*. **testis** míts'<sup>a</sup> *n*. **tetanus** ɲeɗekea na ɨtɛnítʉ́ƙɔta ámák<sup>a</sup> *n*. **Teuso** Tɛʉ́sɔ̀ *n*. **textured** sɨƙɨsíƙánétòn *v*.; síƙísɨƙánón *v*. **thank** ɨlákásítɛ́sʉƙɔt<sup>a</sup> *v*. **thank (with grain)** otés *v*. **thankful** ɨlákásɔ́nʉƙɔt<sup>a</sup> *v*. **that** tòìmɛ̀n *n*.
**that (a while ago)** nótsò *dem*.; nótsò *rel*. **that (a while ago, pl.)** nútsù *rel*. **that (already known)** déé *dem*. **that (earlier)** nák<sup>a</sup> *dem*.; nák<sup>a</sup> *rel*. **that (earlier, pl.)** níkⁱ *rel*. **that (is)** tàà *comp*. **that (just there)** ne *dem*. **that (long ago)** nòk<sup>o</sup> *dem*.; nòk<sup>o</sup> *rel*. **that (long ago, pl.)** nùk<sup>u</sup> *rel*. **that (over there)** ke *dem*. **that (plural)** ni *rel*. **that (singular)** na *rel*. **that (yesterday)** sìn *dem*.; sìn *rel*. **that (yesterday, pl.)** sìn *rel*. **that direction** kɛ́xána kɛ *dem*. **that is (to say)** tòìmɛ̀n *n*. **that one** ɗa ne *pro*. **that one (just there)** kɛɗá *pro*. **that one (over there)** kɛɗa *pro*. **that way** kɛ́xána kɛ *dem*.; kíxána ke *n*. **that way!** ńtía ʝɨkî *adv*. **that wayǃ** ńtíà ʝà *adv*. **thatch** dosés *v*. **thatching (first layer)** ɲáɓarasán *n*. **thatching layer** kerêb<sup>a</sup> *n*. **the coming year** kɛɨnats<sup>a</sup> *n*. **the good life** zɛƙwa ná dà *n*. **the others** kiɗíása *pro*. **the very person/thing** nébèd<sup>a</sup> *n*. **the whole day** ódàtù *n*. **the whole night** tsoík<sup>o</sup> *n*.; tɛrɛƙɛs *ideo*. **theater (movie)** ɲévíɗyòhò *n*. **theater (surgery)** hoesího *n*. **theft** dzú *n*. **their** ńt<sup>a</sup> *pro*.
**theirs** ńtíɛ̀n *pro*. **them** ńt<sup>a</sup> *pro*. **themselves** ńtínebitín *n*. **then** ɓàz *interj*.; ʝá *adv*.; kíná *coordconn*.; kòt<sup>o</sup> *coordconn*.; ts'ɛ́dɔ́ɔ́kɔ̀nà *pro*. **theology** ɲakuʝímɛ́n *n*. **there** ƙɛ́daikén *dem*.; kɔ́ɔ́ *dem*.; ts'ɛ́ daikén *dem*. **there (already known)** ts'ɛ́dɛ́ɛ́ *dem*.; tʉmɛdɛ́ɛ́*dem*. **there (far)** kéda ke *dem*.; kéíta ke *dem*.; kɔ́ɔ́kɛ *dem*. **there (near)** nayé ne *dem*.; nédà *dem*.; néda ne *dem*.; néíta ne *dem*. **there there!** tíɔ̀ *interj*.; tíɔ ʝɔ́ɔ̀ *interj*. **therefore** kòt<sup>o</sup> *coordconn*. **thermometer** ɲátamóómìtà *n*. **these** ni *dem*. **these areas/places** niyá ni *dem*. **these days** ódowicíkó nì *n*. **these guys, I tell you!** ɲɔto ni *interj*. **these kids, I tell you!** wice ni *interj*. **these ones** ɗa *pro*.; ɗa ni *pro*.; niɗa ni *pro*. **these very days** ódowicíkó nì kɔ̀nà *n*. **Thessalonians (biblical)** Ŋítesalóníkaik<sup>a</sup> *n*. **they** ńt<sup>a</sup> *pro*. **thick** rɔ́mɔ́n *v*.; tetíŋón *v*. **thick (and round)** baƙúlúmòn *v*. **thick (become)** moɡánétòn *v*. **thick (flat)** maŋídòn *v*. **thick (mentally)** mɨɲɔna íkèdè *v*. **thick (of brush)** moɡánón *v*. **thick (of undergrowth)** bòmòn *v*. **thick (opaque)** tìnòn *v*. **thick (optimally)** lɔɓɔ́dɔ̀n *v*. **thick (sludgy)** yaŋádòn *v*. **thick (undesirably)** maŋádòn *v*.
#### thicken
**thicken** iɗíkétòn *v*.; iɗikitetés *v*. **thicken up (optimally)** lɔɓɔ́dɨtɛtɛ́s *v*. **thickened** iɗíkón *v*. **thicket** tsekís *n*.; tsekísíàƙw<sup>a</sup> *n*. **thicket (dense)** môɡ<sup>a</sup> *n*. **thicket (round)** ɲalʉ́kɛ́t <sup>a</sup> *n*. **thickly** lɔ̀ɓ ɔ *ideo*.; màŋ *ideo*. **thickset** kikímón *v*. **thief** dzúám *n*. **thief (of grain)** lokoɓél *n*. **thieve** dzuesés *v*.; dzuesetés *v*. **thievery** dzú *n*.; dzúnánès *n*. **thieving** dzúnánès *n*. **thigh** ɡubes *n*. **thigh meat** ɲámoɗ<sup>a</sup> *n*. **thighbone** ɡubesíɔ́k <sup>a</sup> *n*. **thin** ɨkárɔ́n *v*.; kɔrɔ́ɗɔ́mɔ̀n *v*. **thin (delicately)** bɛɗɛ́dɔ̀n *v*. **thin (needle-)** tɨwídɔ̀n *v*. **thin (of a surface)** kwɛxɛ́dɔ̀n *v*. **thin (too)** ɗɛpɛ́dɔ̀n *v*. **thin out** ɨlɔ́lɔ́kɛ́s *v*. **thing** kɔ́rɔ́ɓâd<sup>a</sup> *n*. **things** kúrúɓâd<sup>a</sup> *n*.; mɛnáícík<sup>a</sup> *n*. **things (newly discovered)** kúrúɓáìnòìn *n*. **think** tamɛ́s *v*. **think about** tamɛtɛ́s *v*. **think about each other** támínɔ́s *v*. **think back on** tamɛ́sʉ́ƙɔt<sup>a</sup> *v*. **think on** ɲɛɓɛ́s *v*.; tamátámatés *v*.; tamítámiés *v*. **thinker** tamɛ́síàm *n*.; turúnónìàm *n*. **thinly** bɛ̀ɗ ɛ *ideo*.; ɗɛ̀p ɛ *ideo*.; kwɛ̀x *ideo*. **thinned out** sɨlaɓánón *v*.
**third (be a)** kɔnɔna áɗònù *v*. **third (be the)** mɨtɔna ɗíɛ́áɗònì *v*. **third (one)** ɗa áɗònì *pro*. **third time** aɗoniɛn *n*.; àɗònìk<sup>e</sup> *n*. **thirst** fet<sup>a</sup> *n*. **thirsty** paupáwón *v*. **thirteen** toomíní ńdà kìɗi àɗ<sup>e</sup> *n*. **thirty** toomínékwà àɗ<sup>e</sup> *n*. **this** na *dem*. **this direction** náxána na *dem*. **this kid, I tell you!** ima na *interj*. **this one** ɗa *pro*.; ɗa na *pro*.; naɗa na *pro*. **this way** náxána na *dem*. **this year** kaɨnɔ na *n*.; nakaɨn *n*. **thorax** bakuts<sup>a</sup> *n*. **thorn** kàf *n*. **thornbush (dik-dik)** ɲólíkàf *n*. **those (a while ago)** nútsù *dem*. **those (already known)** díí *dem*. **those (earlier)** níkⁱ *dem*. **those (long ago)** nùk<sup>u</sup> *dem*. **those (over there)** ki *dem*. **those (yesterday, pl.)** sìn *dem*. **those areas** kiyá ki *dem*. **those days** ódowicíkó nùk<sup>u</sup> *n*. **those ones (just there)** kiɗá *pro*. **those ones (over there)** kiɗa *pro*. **those places** kiyá ki *dem*. **thoughts** ŋátámɛta *n*. **thousand** álìf *n*. **thrash** ipés *v*. **thread** iléƙwéries *v*.; ɲéúsi *n*.; rɔ́ɛ́s *v*. **threaded** rɔ́ɔ́s *v*. **threaten** ɨŋaalɛ́s *v*.; kitítésuƙot<sup>a</sup> *v*.; zízɛ̀s *v*.
**threaten to displace** ɨlɔ́líɛtɛ́s *v*. **three** àɗ<sup>e</sup> *num*.; àɗòn *v*. **three days from now** kétsóita ke *n*. **three o'clock** ɲásáatɨkaa tudátie ńda kiɗi ts'aɡús *n*. **three times** àɗ<sup>o</sup> *num*. **three years ago** kaɨnɔ nɔk<sup>ɔ</sup> *n*.; nɔkɛɨna ke *n*. **three years from now** kaɨnɔ na far *n*.; nakaɨna far *n*. **thresh** ipés *v*. **threshing floor** ɓɔɗ<sup>a</sup> *n*.; ɗípɔ̀ *n*. **threshold** lòkìtòŋ *n*.; lòrìòŋòn *n*. **threshold consciousness** mɔɗɔ́ɗɔ́èkw<sup>a</sup> *n*. **thrice** àɗ<sup>o</sup> *num*. **thrifty** toikíkón *v*. **thrive (of plants)** ɡáruɓúɓón *v*.; karuɓúɓón *v*. **throat** morók<sup>a</sup> *n*. **throat infection** tòmàlàɗò *n*. **throb** dìkwòn *v*. **throng** ɲéɓúku *n*.; ɲerípírìp<sup>a</sup> *n*. **throttle** iketiés *v*. **through** nɛ́ɛ́*prep*. **throw** ɡóózés *v*.; ɨmasɛs *v*. **throw a spear** toɓésúƙota ɓɨsá<sup>ɛ</sup> *v*. **throw a stone** zébès *v*. **throw a stone that way** zébesuƙot<sup>a</sup> *v*. **throw a stone this way** zébetés *v*. **throw away** ɡóózesuƙot<sup>a</sup> *v*.; ɨmásɛ́sʉƙɔt<sup>a</sup> *v*. **throw down** ɡwarés *v*. **throw down carelessly** futs'áts'ésuƙot<sup>a</sup> *v*. **throw off** iɓókésuƙot<sup>a</sup> *v*.
**throw stones** turues *v*.; turuetés *v*. **throw this way** ɡóózetés *v*.; ɨmasɛtɛ́s *v*. **thrower (of meat)** ɡóózésíàm *n*. **thrush (rock-)** nàlèmùdzòɗà *n*. **thrust** ututetés *v*.; xutés *v*.; xutésúƙot<sup>a</sup> *v*. **thrust (a knife)** rutésúƙot<sup>a</sup> *v*. **thrust repeatedly** ututiés *v*. **thud** ƙádiƙádès *v*. **thud!** ɗùl *ideo*. **thumb** kɔrɔ́ká ná zè *n*. **thumb piano** lokemú *n*. **thump** dìkwòn *v*.; ɨɲatɛs *v*.; ƙádiƙádès *v*. **thump a tree** iwésá dakwí *v*. **thump repeatedly** ɨɲatiés *v*. **thump thump** kímáts<sup>a</sup> *ideo*.; kùk<sup>u</sup> *ideo*. **Thunbergia alata** ɲápat<sup>a</sup> *n*. **thunder** ɗukuɗúkón *v*.; ɨkílɔ́n *v*.; irúrúmòòn *v*.; ƙìròn *v*.; ƙironuƙot<sup>a</sup> *v*.; tɔtɔanón *v*. **thunder off** itíƙíròòn *v*.; itíríƙòòn *v*.; ƙironuƙot<sup>a</sup> *v*. **Thur person** Ŋítéɓuríám *n*. **Thursday** Ɲákásíá ts'aɡúsík<sup>e</sup> *n*. **thus** kòt<sup>o</sup> *coordconn*. **thwart** kwaɲɛ́s *v*.; kwaɲɛ́sʉ́ƙɔt<sup>a</sup> *v*. **tibia** tsɛrɛ́k <sup>a</sup> *n*. **tick** ɲamaɗaŋ *n*. **tick (mark)** totsetes *v*. **tick grass** ɲamaɗaŋíkú *n*. **tickbird** dzàr *n*. **tickle** ɨkwɨlíkwílɛ́s *v*. **tie** ɲátáy<sup>a</sup> *n*.; zíkɛ́s *v*. **tie around** kɛkɛrɛs *v*. **tie off** ɨlɨɗɛ́s *v*.; ɨlɨɗɛtɛ́s *v*. **tie off umbilical cord** zíkɛ́sà ƙɔ̀ɓà<sup>ɛ</sup> *v*. **tie tightly** ɨrɨʝɛs *v*. **tie up** inénéés *v*.; inénéésuƙot<sup>a</sup> *v*.; zíkɛ́sʉƙɔt<sup>a</sup> *v*.; zɨkɛtɛ́s *v*.
**tied** zíkɔ́s *v*. **tied down loosely** yaŋádòn *v*. **tied down tightly** tokódòn *v*. **tied off** ɨlɨɗɔ́s *v*. **tied tightly** ɨrɨʝɔs *v*.; ŋɔ̀tsɔ̀n *v*. **tied together** zíkízɨkánón *v*. **tight** ɨɗíŋɔ́n *v*.; ɨpɨtɔs *v*.; ɨrɨɗɔs *v*. **tighten hard** ɨpɨtɛs *v*. **tightened hard** ɨpɨtɔs *v*. **tightly** tìɓ<sup>ɨ</sup> *ideo*. **till** tɔkɔ́bɛs *v*. **till morning** tsoík<sup>o</sup> *n*.; tɛrɛƙɛs *ideo*. **tillable** tɔkɔbam *n*. **tilled** tɔkɔ́bɨtɔtɔ́s *v*. **tiller (hand)** ɲɛ́tɛrɛƙɨtaa na kwɛ́tìkà<sup>ɛ</sup> *n*. **tilt** ipuŋes *v*.; liƙés *v*. **tilt over** ipuŋetés *v*. **time** ɲásáat<sup>a</sup> *n*.; ɲásápari *n*. **time off** ɲákarám *n*. **timid** xɛ̀ɓɔ̀n *v*. **timid person** xɛɓásíàm *n*. **timidity** xɛɓás *n*. **Timothy** Timatéw<sup>a</sup> *n*. **Timothy (biblical)** Timatéw<sup>a</sup> *n*. **Timu Road** Tímumucé *n*. **tinder** ɡamam *n*.; lúulú *n*. **tinder (small)** ɗɛ̀rɛ̀ts<sup>a</sup> *n*. **tinkerbird (red-fronted)** kɔkíríkɔk<sup>a</sup> *n*. **tiny** dununúòn *v*.; ɡɔɗírímɔ̀n *v*.; tsaʉ́ɗímɔ̀n *v*.; tɔ́ɗɔ́n *v*. **tiny (opening)** mɨríɗímɔ̀n *v*. **tip** eɗ<sup>a</sup> *n*.; eɗ<sup>a</sup> *n*.; iɲipes *v*.; kàts<sup>a</sup> *n*.; kátsɛ̂d <sup>a</sup> *n*. **tip over** íboɗolés *v*. **tipped** ɨwítsɔ́n *v*.
**tipper** ƙúdèsìàm *n*.; ɲétípa *n*. **tiptoe** ɨtíɗíɗɛ́sá así *v*.; itseɗítséɗòn *v*. **tiptop (of hut)** lómoloró *n*. **tire** bɔrɛ́tɔ́n *v*.; dɛ *n*. **tire shoe** kaetaƙáy<sup>a</sup> *n*. **tire track** dɛ *n*. **tired** bɔ́rɔ́n *v*.; ɨlɔ́ɛ́tɔ̀n *v*.; ɨlɔ́yɔ́n *v*. **tired (become)** bɔrɛ́tɔ́n *v*. **tissue (osseous)** ɔk<sup>a</sup> *n*. **titillate** ɨkɛɗíkɛ́ɗɛ́s *v*.; ɨkwatíkwátɛ́s *v*. **title** éda na moranâd<sup>e</sup> *n*.; iked<sup>a</sup> *n*. **Titus** Títò *n*. **Titus (biblical)** Títò *n*. **to the end** ɗʉ̀ɗʉ̀ŋ *ideo*.; tùtùr *ideo*. **to the rear** ʝìrìk<sup>ɛ</sup> *n*. **toad** ƙwaát<sup>a</sup> *n*.; ƙwaatá na áwìkà<sup>e</sup> *n*. **toast** ɨɔ́ɓɔ́rɛ́s *v*. **tobacco** lɔ́tɔ́ɓ <sup>a</sup> *n*.; ts'ûd<sup>a</sup> *n*. **tobacco (long-leaf)** loríónómor *n*.; pɛ́lɛ́ɗɛ̀k <sup>a</sup> *n*. **tobacco (pounded)** lɔʉtsʉ́r *n*. **tobacco cone** bɔrɔƙɔƙ<sup>a</sup> *n*.; lɔ́tɔ́ɓabɔrɔƙɔ́ƙ a *n*. **tobacco garden** lɔ́tɔ́ɓàsèd<sup>a</sup> *n*. **tobacco garden (grassy)** mʉrɔn *n*. **tobacco grinding stone** lɔ́tɔ́ɓàɡwàs *n*. **tobacco horn** ɲeɓuryaŋ *n*. **tobacco hunger** lɔ́tɔ́ɓàɲɛ̀ƙ <sup>a</sup> *n*. **tobacco leaves** aŋaw<sup>a</sup> *n*. **tobacco pipe** làr *n*. **tobacco user** lɔ́tɔ́ɓààm *n*. **today** nóódwáá *n*. **Toddalia asiatica** lókóɗém *n*. **toe** dɛákɔ́rɔ́k <sup>a</sup> *n*.; kɔrɔ́k <sup>a</sup> *n*. **toe (big)** kɔrɔ́ká ná zè *n*.
#### tied
**toe bone** kɔrɔ́kɔ́ɔ̀k <sup>a</sup> *n*. **toe cut** lɔ́ŋízɨŋîz *n*. **toenail** tíbòlòkòɲ *n*. **toes (extra)** ɲɛ́ɗɔ́nɨɗɔn *n*. **together** ikéé kɔ̀n *n*.; kédìè kɔ̀n *n*.; kédò kɔ̀n *n*. **toilet** ets'íhò *n*.; ɲótsorón *n*. **tolerate** nɛɛ́s *v*.; nɛɛsʉ́ƙɔt<sup>a</sup> *v*.; taɗaŋes *v*. **tomato** ɲɛ́ɲaaɲá *n*. **tomb** rip<sup>a</sup> *n*.; tás *n*.; tásɛ̂d <sup>a</sup> *n*. **tomorrow (morning)** barats<sup>o</sup> *n*.; táábarats<sup>a</sup> *n*.; táábarats<sup>o</sup> *n*. **tomorrow next** kétsóibaráts<sup>a</sup> *n*. **tongs** ɲɔkɔ́ɲɛ́t <sup>a</sup> *n*. **tongue** naƙaf *n*. **too** ʝìk<sup>ɛ</sup> *adv*. **tool (handle-less)** lolemukán *n*. **tool (hooked stick)** pòròt<sup>a</sup> *n*. **tooth** kwayw<sup>a</sup> *n*. **tooth gap (have a)** ɲaŋálómòn *v*. **toothbrush** ɲáɓʉrás *n*.; sʉ́ƙʉ́tɛ́sídàkw<sup>a</sup> *n*.; tsɨtsín *n*. **toothbrush tree** ɓaláŋ *n*. **toothless** ŋalólómòn *v*. **toothless gums** ŋalúɓ<sup>a</sup> *n*. **toothpaste** ɲókólíƙèt<sup>a</sup> *n*. **toothpick (grass)** kua mínɛ́sɨɛ kwaɨtíní *n*. **toothy** ɨŋísímɔ̀n *v*. **top** ɡwarí *n*.; ɨsʉkɛs *v*.; kàts<sup>a</sup> *n*.; kátsɛ̂d a *n*.; sʉ́kɛ́s *v*. **top of a gorge** fòtsàìk<sup>a</sup> *n*. **top of foot** dɛáɡwarí *n*. **top of head** ikáɡwarí *n*. **top part** ɡwaríêd<sup>a</sup> *n*.; iked<sup>a</sup> *n*. **topi** ɲémúƙet<sup>a</sup> *n*.
**topic** mɛnéékw<sup>a</sup> *n*.; tódèèkw<sup>a</sup> *n*. **topics** áƙátìkìn *n*. **Toposa** Kɔrɔmɔt<sup>a</sup> *n*. **Toposa dialect** Kɔrɔmɔtátôd<sup>a</sup> *n*. **Toposa person** dzònìàm *n*.; tɔ́ɓɔ́kìkààm *n*. **topple** íbatɛ́s *v*.; íbatɛtɛ́s *v*. **topple repeatedly** íbatiés *v*. **topsy-turvy** lɔŋɔanón *v*. **torch** kâʒw<sup>a</sup> *n*.; ɲótóts<sup>a</sup> *n*. **torment** ɨtsanítsánɛ́s *v*. **torn** dzɛrɔ́sɔ́n *v*. **torpid** iʝíŋáánón *v*. **torrent** ísw<sup>a</sup> *n*. **tortoise** kàè *n*. **tortoise hatchling** kàèìm *n*. **tortoise shell** kàèƙwàz *n*.; ròɡìròɡ<sup>a</sup> *n*. **torture** ɨtsanítsánɛ́s *v*. **toss** ɡóózés *v*.; ɨmasɛs *v*.; tɔrɛ́s *v*.; towates *v*. **toss (for divination)** ipés *v*. **toss aside** hábatsésúƙot<sup>a</sup> *v*.; hábatsetés *v*. **toss away** ɡóózesuƙot<sup>a</sup> *v*.; ɨmásɛ́sʉƙɔt<sup>a</sup> *v*.; tɔrɛ́sʉ́ƙɔt<sup>a</sup> *v*.; towátésúƙot<sup>a</sup> *v*. **toss in mouth** iɗómóés *v*. **toss off** towátésúƙot<sup>a</sup> *v*. **toss out of sight** ɨsɔmɛs *v*. **toss this way** ɡóózetés *v*.; ɨmasɛtɛ́s *v*.; tɔrɛtɛ́s *v*. **totally** ʝɨkî *adv*.; kɔ́nítɨák<sup>e</sup> *pro*.; mʉ̀kà *adv*.; pílè *ideo*.; tsʉ́tɔ̀ *adv*. **tottering** nɛrɛ́dɔ̀n *v*. **touch** tábès *v*. **touch (make)** tábitetés *v*. **touch all over** tábodiés *v*. **touch down** toɗóón *v*. **touch each other** tábunós *v*.
**touch lightly** ɨkɛɗíkɛ́ɗɛ́s *v*.; ɨkwatíkwátɛ́s *v*. **touch on (topic)** tábès *v*. **touchable** tabam *n*. **touchwood** lúulú *n*. **tough** dɨrɨɲíɲɔ́n *v*.; itsyátón *v*.; nɨkwídɔ̀n *v*. **tough (leathery)** tuɗádòn *v*. **tough (to chew)** kaŋádòn *v*.; kwaídòn *v*. **tough when cooked** haʉ́dɔ̀n *v*. **toughen** ŋɨxítɛ́sʉƙɔt<sup>a</sup> *v*. **toughly** nìkw<sup>ɨ</sup> *ideo*.; tùɗ<sup>a</sup> *ideo*. **tousle** dʉbɛ́s *v*. **tow** béberés *v*. **towel** ɲátauló *n*. **tower (celluar/radio)** ɲéɓusitá *n*. **town** ɲálaín *n*.; táùn *n*. **town-dweller** ɲáláínìkààm *n*. **toxin** ɲekísórìt<sup>a</sup> *n*. **trace** iɗupes *v*.; ɨƙɛrɛs *v*. **trachea** moróká ná zè *n*. **track** ts'íts'ɛ́s *v*. **track footprints** ts'íts'ɛ́sà dɛ̀ìkà<sup>ɛ</sup> *v*. **tractor** ɲɛ́tɛrɛƙɨta *n*. **tractor (hand)** ɲɛ́tɛrɛƙɨtaa na kwɛ́tìkà<sup>ɛ</sup> *n*. **trade** dzîɡw<sup>a</sup> *n*.; dzíɡwès *v*.; ilókótsés *v*.; ɨxɔtsɛs *v*.; xɔ́tsɛ́s *v*. **trade with each other** xɔ́tsínɔ́s *v*. **trader** ŋímutsurúsìàm *n*. **trader (being a)** ŋímutsurúsìnànès *n*. **trading center** ɲálaín *n*.; táùn *n*. **tradition** ɲatal *n*.; ɲeker *n*. **traditional healer** irésíàm *n*. **Tragia insuavis** sʉ́ƙʉ́sʉƙá *n*. **trail** ikókótés *v*.; muce *n*. **trail (fresh)** fʉ́fʉ́t <sup>a</sup> *n*.
**trailer** iƙórú *n*.; ɲɛtʉrɛ́ɛ́là *n*. **trailhead** mucédɛ̀ *n*. **train** ɨtátámɛ́s *v*.; iyoes *v*.; nɔɔsanitetés *v*. **trainer** ɨtátámɛ́síàm *n*.; ŋímaalímùàm *n*. **training** ɲókós *n*. **trait (personality)** ɲɛpɨtɛa ámá<sup>e</sup> *n*. **traitor** tolúónìàm *n*. **trample** íbuɗés *v*.; iɲíkéésuƙot<sup>a</sup> *v*.; takwitakwiés *v*. **trample termites** takwiesúƙota dáŋá<sup>e</sup> *v*. **trample to pieces** fíríts'és *v*. **transfer** iʝokes *v*.; iʝókésuƙot<sup>a</sup> *v*.; ɨkɔɓɛs *v*.; ilóʝésuƙot<sup>a</sup> *v*.; ɨlɔpɛs *v*.; iméérés *v*.; iríítés *v*.; irotes *v*. **transfer here** ɨkɔɓɛtɛ́s *v*. **transfer repeatedly** irotírótés *v*. **transfer there** ɨkɔ́ɓɛ́sʉƙɔt<sup>a</sup> *v*. **transform** beníónuƙot<sup>a</sup> *v*.; iɓéléés *v*.; iɓéléìmètòn *v*.; ilotses *v*.; ilotsímétòn *v*. **transformation of land** beniitesa kíʝá<sup>e</sup> *n*. **transgender person** ɲéliwolíwo *n*. **translate** ɨkɔɓɛs *v*.; ɨkɔɓɛtɛ́s *v*.; ilotses *v*. **translate back and forth** ɨkɔ́ɓínɔ́s *v*. **transmit** iʝokes *v*.; iʝókésuƙot<sup>a</sup> *v*. **transmit trouble** iʝokesa mɛná<sup>ɛ</sup> *v*. **transparency** eas *n*. **transparent** ɓèts'òn *v*.; tsaórómòn *v*. **transparent (of many)** ɓets'aakón *v*. **transpire** ikásíìmètòn *v*.; itíyáìmètòn *v*. **transplant** irotes *v*. **transport** iríítés *v*.; irotes *v*.; tsídzès *v*. **transport away** tsídzesuƙot<sup>a</sup> *v*. **transvestite** ɲéliwolíwo *n*. **trap** kotsítésuƙot<sup>a</sup> *v*.; rʉ́ɛ́s *v*.; tɔlɔkɛ́s *v*. **trap (cage)** ɲáɓáo *n*. **trap (large animal)** ɲéritá *n*. **trap (metal)** ɲétéƙe *n*.
**trap (net)** sáɡòsìm *n*. **trap (small-animal)** lɔwɨɗ<sup>a</sup> *n*. **trap (spike)** ɲátats<sup>a</sup> *n*. **trap (termite)** akarér *n*. **trap (termites)** kokoes *v*. **trap with net** sáɡwès *v*. **trapezius** ɲálaƙamáít<sup>a</sup> *n*. **trapped** kòtsòn *v*. **trapped (become)** kotsonuƙot<sup>a</sup> *v*. **trapping** tɔ̀lɔ̀k <sup>a</sup> *n*. **trapping pit** ɲɔ́sɔ́ɔ́ƙat<sup>a</sup> *n*. **trapping with snares** sâɡw<sup>a</sup> *n*. **trash** ts'ʉts'ʉ *n*. **trash (flashflood)** ɲérímama *n*. **travel** ɓɛƙɛ́s *v*.; ɨlɔ́ɔ́n *v*. **travel away** ɨlɔ́ɔ́nʉƙɔt<sup>a</sup> *v*. **travel here** ɨlɛ́ɛ́tɔ̀n *v*. **travel preparation** sùɓèt<sup>a</sup> *n*. **travel together** ɓɛƙɛ́sínɔ́s *v*. **traveler** ɓɛƙɛ́síàm *n*.; ƙòònìàm *n*. **traveler (preparing)** súɓánònìàm *n*. **traverse** piɗés *v*.; tɔkɛ́ɛ́rɛ́s *v*. **treacherous** ɨmaɗímáɗɔ̀n *v*. **tread on** takwés *v*. **treat** ɡwadam *n*.; ɨmʉ́mwárés *v*.; irés *v*.; ɲɛmʉna *n*. **treat (medicinally)** wetitésá cɛmɛ́ríkà<sup>ɛ</sup> **treat a wound** ɨmaɗɛsa ɔ́ʝá<sup>ɛ</sup> *v*. **treat equally** ikwáánitetés *v*. **treat gently** ɨɓáɓɛ́ɛ́s *v*. **treat respectfully** iríméés *v*. **treatment** cɛ̀mɛ̀r *n*. **tree** dakw<sup>a</sup> *n*. **tree (sacred)** lɔ́ƙɔ́ŋ *n*.; ɲɔ́ƙɔ́ŋ *n*. **tree (unknown)** kɔ́rɔ́ɓáìdàkw<sup>a</sup> *n*.
*v*.
**tree species** àɗèŋèlìò *n*.; ɓàʝ<sup>a</sup> *n*.; ɓólìs *n*.; ɓòŋ *n*.; basaúréèkw<sup>a</sup> *n*.; boxoƙorét<sup>a</sup> *n*.; ɗewen *n*.; dzôɡ<sup>a</sup> *n*.; ekoɗit<sup>a</sup> *n*.; èmùsìà *n*.; fàìdw<sup>a</sup> *n*.; ɡàràʝ<sup>a</sup> *n*.; ɡodiyw<sup>a</sup> *n*.; ibét<sup>a</sup> *n*.; iroroy<sup>a</sup> *n*.; isókóy<sup>a</sup> *n*.; ƙɔ́ɓʉƙɔ́ɓ <sup>a</sup> *n*.; kàrɛ̀ *n*.; kɛ́láy<sup>a</sup> *n*.; kómoló *n*.; kunét<sup>a</sup> *n*.; kùr *n*.; lóɗíwé *n*.; lɔ́kɛ́rʉ́ *n*.; lokum *n*.; lɔ́lɔwí *n*.; lòŋìr *n*.; meleke *n*.; mókol *n*.; mozokoɗ<sup>a</sup> *n*.; mʉ̀s *n*.; ŋʉrʉ́sá *n*.; naarákɨlɛ *n*.; ɲáɓata *n*.; ɲákaɓurúr *n*.; ɲákátɨríɓa *n*.; ɲamalil *n*.; ɲécaal *n*.; ɲɛ́caal *n*.; ɲɛ́kɨsí *n*.; ɲékwaŋa *n*.; ɲéleɓuléɓu *n*.; ɲɛmaɨlɔŋ *n*.; ɲéŋéso *n*.; ɲépípa *n*.; ɲéyoroeté *n*.; ɲóɗomé *n*.; ɲókotit<sup>a</sup> *n*.; ɲóƙoloƙolét<sup>a</sup> *n*.; óbìʝòɔ̀z *n*.; rirís *n*.; ròr *n*.; rukûdz<sup>a</sup> *n*.; seɡer *n*.; seínení *n*.; sésèn *n*.; ts'ɔƙɔ́m *n*.; tsàl *n*.; tsereɗeɗí *n*.; tsɛ̀tsɛ̀kw<sup>a</sup> *n*.; tsɨkw<sup>a</sup> *n*.; tɛɛtɛ́ *n*.; tʉlárɔ́y <sup>a</sup> *n*.; tùr *n*.; tʉ̀tʉ̀f *n*.; ʉrʉ́sáy<sup>a</sup> *n*.; warɨwar *n*.; xuxûb<sup>a</sup> *n*. **tree trunk** dakúɓɔ́l *n*. **treeless** ŋoléánètòn *v*. **tremble** kìtòn *v*.; kwalíkwálɔ̀n *v*.; tsábatsabánón *v*. **tremble (begin to)** kitétón *v*. **tremble (make)** kitítésuƙot<sup>a</sup> *v*. **tremor** irikíríkòn *v*. **trench** ɲéƙúrumot<sup>a</sup> *n*.; urúr *n*. **tri-colored** ɓokóánètòn *v*.; eséánètòn *v*. **trial (legal)** ɲékés *n*. **tribe** dìyw<sup>a</sup> *n*.; ɲákaɓɨlá *n*. **Tribulus cistoides** ɲesuƙuru *n*. **tribunal** ɲókót<sup>a</sup> *n*. **tribute** meetésíicík<sup>a</sup> *n*. **tricep** cwɛtéém *n*. **tricep (lower)** ƙʉlɛ́èm *n*. **trick** ɨmɔɗɛs *v*.; itwáŋítésúƙot<sup>a</sup> *v*. **trickle** mɨrɨmírɔ́n *v*.; tɔlɛ́lɛ́ɔ̀n *v*. **tricky** ɨmaɗímáɗɔ̀n *v*.
### trigger
**trigger** ɲɛɗɛ́sɛ̂d <sup>a</sup> *n*.; ɲétíƙa *n*. **trigger (trap)** ɨɗalɛs *v*. **trim** ɨƙwáƙwárɛ́s *v*.; ɨlɨmɛs *v*. **trim back** ɨlɨmɛtɛ́s *v*.; isésélés *v*. **trip** ɨlɛ́ƙwɛ́rɛ́s *v*.; ɲásápari *n*.; rúmánitésúƙot<sup>a</sup> *v*.; rúmánòn *v*. **trip (of a trap)** ɨɗálɛ́sʉƙɔta así *v*. **trip (trap)** ɨɗalɛs *v*. **trip repeatedly** iɲatiesá kíʝá<sup>e</sup> *v*. **trip up** ɨlɛ́ƙwɛ́rɛtɛ́s *v*. **tripod** lèwèɲìdɛ̀ *n*. **Triumfetta annua** lɔ́mɔ́ɗaát<sup>a</sup> *n*. **trochanter (greater)** obólénìɔ̀k <sup>a</sup> *n*. **tromp** íbuɗés *v*. **troop (of baboons)** kwaár *n*. **trophy** tɔ́rɔ́bɛsa na ílɔɛsí *n*. **trot** isipísípòn *v*.; ɨsɔƙísɔ́ƙɔ̀n *v*.; ɨsʉmʉ́sʉ́mɔ̀n *v*. **troublemaker** mɛnáám *n*. **troubles** mɛn *n*.; ŋítsan *n*. **trough** itúɓ<sup>a</sup> *n*.; sɔ́k <sup>a</sup> *n*. **trounce** ipés *v*.; ɨrɛɛs *v*. **trousers (pair of)** ɡwan *n*.; ɲétorós *n*.; ɲótorós *n*. **trowel** ɲɛ́tʉráwɛ̀l *n*. **truck** lóórì *n*.; ɲolórì *n*. **truck (small)** kàè *n*. **true** ɨtsírɔ́n *v*.; tsírɔ́n *v*.; tɔɓɛ́ɔ́n *v*. **true (typically)** toɓéíón *v*. **truly** easík<sup>e</sup> *n*. **trumpet** ɨkílɔ́n *v*.; ɲɛ́rʉpɛpɛ́*n*. **trunk (elephant)** ɡìɡ<sup>a</sup> *n*.; komóts<sup>a</sup> *n*.; òŋòrìkwɛ̀t <sup>a</sup> *n*. **trunk (tree)** dakúɓɔ́l *n*. **trunks (pair of)** ɲésiriwáli *n*.; ɲosoƙoloké *n*.
**trust** tɔnʉpɛs *v*. **truster** tɔnʉpɛsíám *n*. **trustworthy** tɔnʉpam *n*. **truth** eas *n*. **truth (be the)** mɨtɔna eas *v*. **truthful person** easíám *n*. **try** ɨkatɛs *v*.; kaites *v*. **try in court** iniŋes *v*. **try repeatedly** ɨkatíkátɛ́s *v*. **tryst** tirésíàw<sup>a</sup> *n*. **tsetse fly** ɲɛ́ɗíɨt<sup>a</sup> *n*. **tub** itúɓ<sup>a</sup> *n*. **tubby** ɡerúsúmòn *v*.; poŋórómòn *v*.; rexúkúmòn *v*. **tuberculosis (pulmonary)** ɡafíɡáfikaɲeɗeké *n*.; ɲeɗekea bákútsìkà<sup>e</sup> *n*. **tuck** ipuŋes *v*. **tuck away** ɨkíɗítsɛ́s *v*. **tuck into** íbʉbʉŋɛ́s *v*.; íbʉbʉŋɛ́sʉ́ƙɔt<sup>a</sup> *v*. **tuck up** rʉ́bɛ̀s *v*.; rʉ́bɛsʉƙɔt<sup>a</sup> *v*. **tuckered** ziálámòn *v*.; zíkímétòn *v*.; ziláámòn *v*. **Tuesday** Ɲákásíá lèɓètsìk<sup>e</sup> *n*. **tuft** tsulát<sup>a</sup> *n*. **tug** ɗʉ́rɛ́s *v*.; ɗʉtɛ́s *v*. **tug back and forth** ɨlɨkílíkɛ́s *v*. **tumble** íbatɛ́s *v*.; íbatɛtɛ́s *v*.; ruɓétón *v*.; ruɓonuƙot<sup>a</sup> *v*. **tumble down** íbatɛsa así *v*. **tumble repeatedly** íbatiés *v*. **tumefy** emites *v*.; èmòn *v*. **tumid** bɔfɔ́dɔ̀n *v*. **tune out** bálábálatés *v*. **tuner** ɲéréɗi *n*. **tunnel** wɛ̀l *n*. **tunnel (center)** wɛ̀lèèkw<sup>a</sup> *n*. **turaco** fúluƙurú *n*. **turbid** kaɓúrútsánón *v*.
**turgid** bɔfɔ́dɔ̀n *v*. **Turkana language** Pakóícétôd<sup>a</sup> *n*. **Turkana person** ŋɔrɛ́ám *n*.; Pakóám *n*.; rɔwáám *n*. **Turkanaland** Burukáy<sup>a</sup> *n*. **turkey** ɲékulukúl *n*. **turn** aŋɨrɛs *v*.; ɨʝʉlɛs *v*.; ɨríŋítɛ́s *v*.; ɨríŋɔ́n *v*.; iwoles *v*. **turn against** tolúétòn *v*.; tolúónuƙot<sup>a</sup> *v*.; tolúútésuƙot<sup>a</sup> *v*. **turn against each other** tolúúnós *v*. **turn around** iɓóɓóŋòn *v*.; ɨʝʉlɛtɛ́s *v*.; ɨríŋítɛ́sʉƙɔt<sup>a</sup> *v*.; ɨríŋɔ́nʉƙɔt<sup>a</sup> *v*. **turn away** ɨríŋítɛ́sʉƙɔt<sup>a</sup> *v*.; ɨríŋɔ́nʉƙɔt<sup>a</sup> *v*.; ɨtíílɛ́s *v*. **turn back this way** iɓóɓóŋètòn *v*. **turn back to back** ƙʉƙʉmánítésuƙot<sup>a</sup> *v*.; ƙʉƙʉmánónuƙot<sup>a</sup> *v*. **turn down** ɨmɛ́ɗɛ́lɛ́s *v*.; míʝés *v*. **turn off** kɔkɛtɛ́s *v*. **turn off (electrically)** ts'eítésuƙot<sup>a</sup> *v*. **turn off (of electricity)** ts'oonuƙot<sup>a</sup> *v*. **turn on (attack)** toƙíróòn *v*. **turn on (betray)** tolúétòn *v*.; **turn on** tolúónuƙot<sup>a</sup> *v*. **turn on (electrically)** aeitetés *v*. **turn on (of electricity)** aeétón *v*. **turn on each other** tolúúnós *v*. **turn one's back to** ƙʉƙʉmanés *v*. **turn oneself around** iɓéléésuƙota así *v*. **turn oneself over repeatedly** iɓilíɓílésá así *v*. **turn out** pukés *v*.; puketés *v*. **turn out away** pukésúƙot<sup>a</sup> *v*. **turn over** bukures *v*.; bukúrésuƙot<sup>a</sup> *v*.; buƙusítésuƙot<sup>a</sup> *v*.; iɓéléés *v*.; iɓéléetés
*v*.; iɓéléìmètòn *v*.; iɓélúkáìmètòn *v*.; iɓélúkéés *v*.; ɨʝʉlɛtɛ́s *v*.; iwoletés *v*. **turn over (soil)** iwúlákés *v*. **turn this way** ɨríŋɛ́tɔ̀n *v*. **turn up** takánétòn *v*. **turn upside-down** tuɗúlútés *v*. **turned** ɨʝʉlɔs *v*. **turned on (sexually)** iɓurímétòn *v*.; kwídikwidós *v*. **turtle** sídilé *n*. **tusk (elephant)** òŋòrìkwàyw<sup>a</sup> *n*. **tusker** òŋòr *n*. **tussock** tsulát<sup>a</sup> *n*. **tweak** tʉnɛ́s *v*. **tweak off** tɔɲɨmɛtɛ́s *v*. **twelve** toomíní ńda kiɗi léɓèts<sup>e</sup> *n*. **twelve o'clock** ɲásáatɨkaa tudátie ńdà kɛ̀ɗì kɔ̀n *n*. **twenty** toomínékwa léɓèts<sup>e</sup> *n*. **twice** lèɓèts<sup>o</sup> *num*. **twiddle** íɡuʝuɡuʝés *v*. **twilight** xɨŋat<sup>a</sup> *n*.; xɨŋatétón *v*. **twilit** míɡiriɡíránón *v*. **twin (be a)** ɨmʉ́ɔ́n *v*. **twine** ɨmɔ́ʝírɛ́s *v*.; itoŋes *v*.; natɨɓ<sup>a</sup> *n*. **twinkle** itweɲítwéɲòn *v*. **twins** ŋímúí *n*. **twirl** tɔpɨrípírɛ́s *v*. **twirl between hands** tsapés *v*. **twirlable** tsapetam *n*. **twist** aŋɨrɛs *v*.; ɨmɔ́ʝírɛ́s *v*.; itoŋes *v*.; tɔpɨrípírɛ́s *v*. **twist round** ɨtʉ́tʉ́rɛ́s *v*. **twist the truth** itoŋetésá tódà<sup>e</sup> *v*. **twist up** imákóitetés *v*.; ɨpɔ́pírɛ́s *v*.; kakɨrɛ́s *v*.
#### twisted
**twisted** ɡɔ́lɔ́ɡɔlánón *v*. **twisted round** ɨtʉ́tʉ́rɔ́s *v*. **twisted up** imákóòn *v*. **twisted up (become)** imákwéètòn *v*. **twitch** ɨmímíʝɛ́s *v*.; irikíríkòn *v*. **two** lèɓèts<sup>e</sup> *num*.; leɓétsón *v*. **two (make)** leɓetsítésuƙot<sup>a</sup> *v*. **two o'clock** ɲásáatɨkaa tudátie ńdà kìɗì àɗ<sup>e</sup> *n*. **two times** lèɓèts<sup>o</sup> *num*. **two years ago** kaɨnɔ nótso *n*.; nɔkɛɨn *n*. **two years from now** kaɨnɔ na tso *n*.; nakaɨna tso *n*. **two-by-two** leɓetsíón *v*. **type** bònìt<sup>a</sup> *n*.; ɲákaɓɨlá *n*. **Typha species** bʉlʉbʉláta na sábàìkà<sup>e</sup> *n*. **udder** ídoho *n*. **udder (cow)** ɦyòìdw<sup>a</sup> *n*. **Uganda** Uɡánɗà *n*. **ugly** itópénòn *v*.; làlòn *v*. **uh …** ndaicé *n*. **uh-huh, sure!** yóói *interj*. **ulcer** ɗɔ̀l *n*. **ulcer (stomach)** bùbùɔ̀ʝ <sup>a</sup> *n*. **ululate** iyíyéés *v*. **um …** át<sup>a</sup> *n*.; ndaicé *n*. **umber** ts'aráfón *v*. **umbilical cord** ƙɔɓasim *n*. **umbilical hernia** ƙɔɓa na zikîb<sup>a</sup> *n*. **umbrella thorn tree** sèɡ<sup>a</sup> *n*. **unadorned** ɨɓámɔ́n *v*.; sɨlɔ́ʝɔ́mɔ̀n *v*. **unaffixed** dolódòn *v*.; roiróón *v*. **unattractive** itópénòn *v*. **unavailable** bɨrɔ́ɔ́n *v*. **unbeliever** nɛpɛ́ƙáàm *n*.
**unbend** ɨɗírítɛ́sʉƙɔt<sup>a</sup> *v*. **unbending** kɛtɛ́rɛ́mɔ̀n *v*. **unbreakable** lɛrɛ́dɔ̀n *v*.; mɛkɛlɛ́lɔ́n *v*. **unbreakably** lɛ̀r *ideo*. **unburdened** bùlòn *v*. **uncertain** ɨʉ́ƙɔ́n *v*. **unchewable** mɛkɛlɛ́lɔ́n *v*.; ts'afʉ́dɔ̀n *v*. **unchewably** ts'àf *ideo*. **uncle (his/her father's brother)** babat<sup>a</sup> *n*. **uncle (his/her father's sister's husband)** tatatíéákw<sup>a</sup> *n*. **uncle (his/her mother's brother)** momot<sup>a</sup> *n*. **uncle (his/her mother's sister's husband)** tototíéákw<sup>a</sup> *n*. **uncle (mother's brother)** momó *n*. **uncle (mother's sister's husband)** totóèàkw<sup>a</sup> *n*. **uncle (my father's brother)** abáŋ *n*. **uncle (my father's sister's husband)** tátàèàkw<sup>a</sup> *n*. **uncle (your father's brother)** bábò *n*. **uncle (your father's sister's husband)** tátóéákw<sup>a</sup> *n*. **unclean** ŋɔrɔ́ɲɔ́mɔ̀n *v*.; ɲɔŋɔ́rɔ́mɔ̀n *v*. **unclear (information)** kìtsòn *v*. **unclouded** kánɔ́n *v*. **uncomfortable** ɗɛɲɨɗɛɲɔs *v*. **uncomplicated** ɓàŋɔ̀n *v*. **unconscious** bàdòn *v*.; ifáfúkós *v*. **unconscious (go)** badonuƙot<sup>a</sup> *v*.; rèŋòn *v*. **uncooked** ts'áɡwòòn *v*. **uncooperative with each other** ɗúlúnós *v*.
unseen
**uncoordinated** hádaadánón *v*.; ɨɓaŋíɓáŋɔ̀n *v*. **uncover** ŋáɲɛ́sʉƙɔt<sup>a</sup> *v*. **undependable** iléʝíánón *v*.; imáɗíŋánón *v*. **underarm** bàbà *n*. **underbelly** búbùèd<sup>a</sup> *n*. **underclothes (pair of)** ɲekúrúm *n*. **undercook** kitsonuƙot<sup>a</sup> *v*. **undercooked** ɨmʉ́ránón *v*.; kìtsòn *v*. **underfoot** dɛ̀ìkà<sup>ɔ</sup> *n*. **undergird** titirés *v*.; titiretés *v*. **underground** ɗis *n*.; ʝʉmááƙw<sup>a</sup> *n*. **underminer** ɲɛkɛsʉpan *n*. **underside** búbùèd<sup>a</sup> *n*. **understand** enés *v*.; nesíbes *v*.; walámón *v*. **understand each other** nesíbunós *v*. **understood** nesíbos *v*.; nesíbunós *v*. **underwear (pair of)** ɲekúrúm *n*. **undeveloped** ikúrúfánón *v*. **undigestible** ts'afʉ́dɔ̀n *v*. **undigestibly** ts'àf *ideo*. **undivided (of cash)** ɗukúdòn *v*. **undress** hoɗésúƙot<sup>a</sup> *v*.; hoɗetés *v*. **unearth** úɡès *v*.; úɡetés *v*. **uneducated** ɨɓááŋɔ̀n *v*. **unemployed** ɨlwárɔ́na teréɡù *v*. **uneven** ƙumúƙúmánón *v*. **unfastened** dolódòn *v*.; roiróón *v*. **unfathomable** xakútsúmòn *v*. **unfixed** ɨɓámɔ́n *v*. **unfurl** tɔpɛtɛs *v*.; tɔpɛ́tɛ́sʉƙɔt<sup>a</sup> *v*. **uniform** ɲéyúnìfòm *n*. **unintelligent** ɨɓááŋɔ̀n *v*. **uninterested** bɔ́rɔ́n *v*.
**uninteresting** itópénòn *v*.; ʝɔ̀lɔ̀n *v*. **unique** beníón *v*.; bɛnɔ́ɔ́n *v*. **unison (act in)** ilíréètòn *v*. **unite** kɔnítɛ́sʉƙɔt<sup>a</sup> *v*. **universe** kíʝ<sup>a</sup> *n*. **university** yunivásìtì *n*. **unkempt** sɔɓɔ́lɔ́mɔ̀n *v*. **unknowing** ɨɓááŋɔ̀n *v*. **unlawful** toɗyakos *v*. **unless** ɗàmʉ̀s *subordconn*.; ɗɛ̀mʉ̀s *subordconn*. **unlidded** lɛɓɛ́ɲɛ́mɔ̀n *v*. **unliftably** ɓa *ideo*. **unload** ɓuƙetés *v*. **unload a load** ɓuƙetésá botá<sup>e</sup> *v*. **unmanageable** imákóòn *v*. **unmanageable (become)** imákwéètòn *v*. **unmoving** wàsɔ̀n *v*. **unobserved** kúbòn *v*. **unobstructed** fotólón *v*. **unoccupied** bùlòn *v*.; ɨɓámɔ́n *v*.; ipásóòn *v*. **unoccupied (become)** bulonuƙot<sup>a</sup> *v*. **unpierced (ears)** mʉɗáŋámɔ̀n *v*. **unproductive** ɨsʉ́wɔ́ɔ̀n *v*. **unprotected** ilététòòn *v*. **unrelated** ʝalánón *v*. **unreliable** iléʝíánón *v*.; imáɗíŋánón *v*. **unresponsive** ɨlɛtílɛ́tɔ̀n *v*.; ɨlɛ́tʉ́ránón *v*. **unrest** ɲárém *n*. **unrewarded** kwɛtíkín<sup>ɔ</sup> *n*.; seát<sup>o</sup> *n*. **unripe** ts'áɡwòòn *v*. **unroll** tɔpɛtɛs *v*.; tɔpɛ́tɛ́sʉƙɔt<sup>a</sup> *v*. **unruly (of hair)** ɡaúsúmòn *v*. **unsafe (of an area)** tsakátsákánón *v*. **unsecured** haʝádòn *v*.; laʝádòn *v*. **unseen** kúbòn *v*.
**unsettled** ɡokirós *v*. **unsettled (homeless)** tsɔnɨtsɔnɔ́s *v*. **unsighted** múɗúkánón *v*. **unspoiled** ɗòwòn *v*. **unstable** ɗatólóɲòn *v*.; ikáɓóɓánón *v*. **unsteadily** ɗɔ̀x *ideo*.; ɡwèlèʝ<sup>e</sup> *ideo*. **unsteady** ikáɓóɓánón *v*. **unstick** ɨtakɛs *v*.; ɨtákɛ́sʉƙɔt<sup>a</sup> *v*. **unsturdy** ɗatólóɲòn *v*. **unsupple** ɡɔkɔ́dɔ̀n *v*. **unsure** ɨʉ́ƙɔ́n *v*. **until** akání *prep*.; ɗàmʉ̀s *subordconn*.; ɗɛ̀mʉ̀s *subordconn*.; ɡònè *prep*.; pákà *prep*.; pákà *subordconn*. **untouched** ɗòwòn *v*. **untrue to one's word** iméníkánón *v*. **untrustworthy** iméníkánón *v*. **untruth** yʉɛ *n*. **unused** ɗòwòn *v*.; ɨɓámɔ́n *v*.; ɨlárɔ́n *v*.; ɨlwárɔ́n *v*. **unusual way** kɔ́náxàn *n*. **unutilized** ɨlárɔ́n *v*.; ɨlwárɔ́n *v*. **unwanted** ɨtsárʉ́ánón *v*. **unwise** ɨɓááŋɔ̀n *v*. **unyielding** nɨkwídɔ̀n *v*. **up** dìdìk<sup>e</sup> *n*.; kɔ́ɔ́kwar<sup>ɔ</sup> *n*.; nɔ́ɔ́kwar<sup>ɔ</sup> *n*. **up to** akání *prep*.; ɡònè *prep*.; pákà *prep*. **up-up!** kukú *nurs*. **upchuck** ɨlɔ́ɓɔ́tɛtɛ́s *v*. **upper end** iked<sup>a</sup> *n*. **upright** iséréròn *v*.; ɨtsírɔ́n *v*.; tsírɔ́n *v*. **uproot** ɗués *v*.; ɗuetés *v*.; rués *v*.; ruutésuƙot<sup>a</sup> *v*.; ruutetés *v*. **upset** bukures *v*.; bukúrésuƙot<sup>a</sup> *v*.; buƙusítésuƙot<sup>a</sup> *v*.; ɡaanítésuƙot<sup>a</sup> *v*.; ɡaanón *v*.; iɓélúkéés *v*.; iŋóyáánón *v*.
**upset (become)** ɡaanónuƙot<sup>a</sup> *v*.; ɨraŋímétòn *v*. **upset (emotionally)** ɓarites *v*.; ɓarítésuƙot<sup>a</sup> *v*. **upside-down** tuɗúlón *v*. **upward** dìdìk<sup>e</sup> *n*.; kɔ́ɔ́kwar<sup>ɔ</sup> *n*. **urban center** zɛƙɔ́áwa ná zè *n*. **urbanite** ɲáláínìkààm *n*. **urethra** xaramucé *n*. **urethral meatus** kwaníékw<sup>a</sup> *n*. **urge** ɨmʉ́káitetés *v*. **urge (nicotine)** lɔ́tɔ́ɓàɲɛ̀ƙ <sup>a</sup> *n*. **urge on** ɨsʉ́sʉ́ɛ́s *v*.; itsótsóés *v*. **urinate** kʉtsáƙón *v*. **urinating spot** kʉtsáƙáàw<sup>a</sup> *n*. **urine** kwats<sup>a</sup> *n*. **urine (cow)** tsét<sup>a</sup> *n*. **us (exclusive)** ŋɡw<sup>a</sup> *pro*. **us (inclusive)** ɲjín *pro*. **use** eréɡes *v*.; isítíyeés *v*.; ɲákásìèd<sup>a</sup> *n*. **useable** eréɡam *n*. **used to** ɨtalɔs *v*. **used to (make)** ɨtalɛs *v*. **used to each other** náínɔ́s *v*. **used up** bitsétón *v*.; tɛ́zɛ̀tɔ̀n *v*. **useful** eréɡam *n*. **useless** ɨtsárʉ́ánón *v*.; pás *n*. **useless thing** tsar *n*. **uselessness** pásìnànès *n*.; tsarínánès *n*. **usher** iɗimiesíàm *n*. **usher in (new year)** itówéés *n*. **uterus** epúáw<sup>a</sup> *n*.; ɲapéryɛ́t <sup>a</sup> *n*. **uterus (prolapsed)** kìɓèɓè *n*.; kìtʉ̀lɛ̀ *n*. **utilize** isítíyeés *v*. **utter** kʉtɔnʉƙɔt<sup>a</sup> *v*.
**uvula** ɲoɗokole *n*. **vacant** bótsón *v*.; bùlòn *v*. **vacant (become)** bulonuƙot<sup>a</sup> *v*. **vacate** bulútésuƙot<sup>a</sup> *v*. **vacation** ɲákarám *n*. **vaccinate** ɨtsɨpítsípɛ́s *v*. **vacillate** iɲikiétòn *v*.; iɲikíɲíkòn *v*. **vagabond** xikóám *n*. **vagabond (bush)** ríʝíkààm *n*. **vagabondage** xikw<sup>a</sup> *n*. **vagabondage (in the bush)** ríʝíkànànès *n*. **vagina** dòɗ<sup>a</sup> *n*. **vagrancy** xikw<sup>a</sup> *n*. **vagrancy (in the bush)** ríʝíkànànès *n*. **vagrant** xikóám *n*. **vague** ɨɲʉ́ɲʉ́ánón *v*.; kìtsòn *v*. **vague (visibly)** imítíròn *v*. **vain** itúrón *v*. **vain (become)** ɨkárímétòn *v*. **valley (wide)** ɲéɓúruɓur *n*. **vanish** ɨíɗɔ́n *v*.; kúbonuƙot<sup>a</sup> *v*.; wɨɗímɔ́nʉƙɔt<sup>a</sup> *v*. **vanishingly** wìɗ<sup>ɨ</sup> *ideo*. **vanquish** ɨrɛɛs *v*. **vapid** ɗɛ̀ƙwɔ̀n *v*.; ʝɔ̀lɔ̀n *v*.; muʝálámòn *v*. **vapor** lɔkapʉ́r *n*. **varicella** ɲɛtʉnɛ *n*.; puurú *n*. **variety** bònìt<sup>a</sup> *n*.; ɲákaɓɨlá *n*.; ɲalíɲalí *n*. **various** ʝaláʝálánón *v*. **vary in height** iyópón *v*. **vault** itúlúmòòn *v*. **veer** iwítón *v*.; kwɛ́dɔ̀n *v*. **veer repeatedly** aŋiriesón *v*.; iwitíwítòn *v*.
**vegetable garden** waicíkásèd<sup>a</sup> *n*.
**vegetables** wà *n*. **vegetation (thick)** tsekís *n*.; tsekísíàƙw<sup>a</sup> *n*. **vehicle** kàè *n*.; ɲómotoká *n*. **vehicle (small)** kàèìm *n*. **vein** seamucé *n*.; tsòrìt<sup>a</sup> *n*. **vellicate** irikíríkòn *v*. **velocity** ɲésipíɗ<sup>a</sup> *n*. **velvety** ʝamúdòn *v*. **vendue** ókísèn *n*. **venerate** mòròn *v*. **venom** ɲekísórìt<sup>a</sup> *n*. **vent (volcanic)** ɲáɗúy<sup>a</sup> *n*. **ventriculus** ŋìl *n*. **Vepris glomerata** kùr *n*. **veranda** hodzíŋ *n*. **verdant** ƙwɨxídɔ̀n *v*.; xídɔ̀n *v*. **verity** eas *n*. **Vernonia cinerascens** mɛ́rɛ́ɗɛɗɛ́*n*. **verrucose** tɔmɔtsɔkánón *v*. **verse** àsàk<sup>a</sup> *n*. **vertebrae** ɡòɡòròʝòɔ̀k <sup>a</sup> *n*. **vertex** ikáɡwarí *n*. **vertibrae (upper cervical)** bɔkɔ́s *n*. **vertigo** taítayó *n*. **very** ʝɨkî *adv*.; zùk<sup>u</sup> *adv*. **very much** mbáyà *adv*. **vesicate** bubuxánónuƙot<sup>a</sup> *v*.; ileɓíléɓòn *v*. **vesicated** bubuxánón *v*. **vespid** dɛrɛ́ƙ <sup>a</sup> *n*. **vessel (blood)** seamucé *n*.; tsòrìt<sup>a</sup> *n*. **vessel (water)** itúɓ<sup>a</sup> *n*. **vest** ɲábʉlán *n*.; ɲɛ́bʉlán *n*. **vest (beaded)** ɲáɓol *n*. **veterinarian** ɗakɨtárɨama ínó<sup>e</sup> *n*. **veto** ilotsesa mɛná<sup>ɛ</sup> *v*.
#### viable
**viable** ikásíetam *v*. **vicinity (in the)** ɦyàtàk<sup>a</sup> *n*. **vicious** ɨɲɛ́ɛ́mɔ̀n *v*.; ɨsílíánón *v*. **victual** ŋƙam *n*. **video** kúrúkúríka ni ɓɛƙɛ́s *n*.; ɲévíɗyo *n*. **view** ɡonés *v*.; iléúrés *v*.; inwakes *v*. **view (here)** ɡonetés *v*. **view (there)** ɡonésúƙot<sup>a</sup> *v*. **Vigna frutescens** ànɛ̀ *n*. **Vigna oblongifolia** kìtsàɗɔ̀s *n*. **Vigna species** ɗɔsɔ́*n*.; málákʉ́r *n*. **village** aw<sup>a</sup> *n*. **village (abandoned)** on *n*.; oníáw<sup>a</sup> *n*. **village (big)** awa ná zè *n*. **vine** ɲɛ́lɔ́kɨlɔk<sup>a</sup> *n*. **vine species** àdàbì *n*.; ànɛ̀ *n*.; ɗilat<sup>a</sup> *n*.; ɗɔsɔ́ *n*.; ewêd<sup>a</sup> *n*.; ɨɛƙɨɛƙ<sup>a</sup> *n*.; ɨnw<sup>a</sup> *n*.; ʝooʝo *n*.; kapʉrat<sup>a</sup> *n*.; kìtsàɗɔ̀s *n*.; kutsúbàè *n*.; loɗeɗ<sup>a</sup> *n*.; lókóɗém *n*.; lókúɗukuɗét<sup>a</sup> *n*.; lótórobét<sup>a</sup> *n*.; málákʉ́r *n*.; míʒìʒ *n*.; ɲákamɔ́ŋɔ *n*.; ɲákamʉ́ka *n*.; ɲalakas *n*.; ɲéŋeɗo *n*.; óbiʝoets'<sup>a</sup> *n*.; simísímàt<sup>a</sup> *n*.; tsɔ́ráɗoɗôb<sup>a</sup> *n*.; tìl *n*.; tiritirikwáy<sup>a</sup> *n*.; tɔt<sup>a</sup> *n*.; tʉkʉtʉkán *n*.; urém *n*.; wɛ́ƙɛ́ƙ a *n*.; xoúxoú *n*. **vinery** ɲéviinísèd<sup>a</sup> *n*. **vineyard** ɲéviinísèd<sup>a</sup> *n*. **violate sexually** itikiesúƙot<sup>a</sup> *v*. **violence (domestic)** ɡaɗ<sup>a</sup> *n*. **violent** ifulúfúlòn *v*.; iréɲíánón *v*. **violent person** ɡaɗɛ́ám *n*. **violin** ɲakaw<sup>a</sup> *n*. **viper** bɛf *n*. **viper (Gaboon)** bɛfa na ɡóɡòròʝìkà<sup>e</sup> *n*. **virgin** ɲarama na ɓéts'<sup>a</sup> *n*.; ɲarama na tílɨw<sup>a</sup> *n*.
**virginal** ɗòwòn *v*.; tɨlíwɔ́n *v*.; xɔ́dɔ̀n *v*.; xɔtánón *v*. **visage** takár *n*. **viscous** naídòn *v*. **viscously** nàⁱ *ideo*. **visible** ilééránón *v*.; kɛtɛ́lɔ́n *v*.; lɛ́lɔ́n *v*.; takánón *v*. **visible (completely)** ùwòò *ideo*. **visible (make)** kɛtɛ́lɨtɛtɛ́s *v*. **visit** énímós *v*. **visitor** wáánààm *n*. **vivacious person** ɨɓʉrɛtɛ́síàm *n*. **vocalist** ìrùkààm *n*.; irukósíàm *n*. **vocals** ìrùk<sup>a</sup> *n*. **vocational school** tɛ́kɛ̀nìkɔ̀l *n*. **voice** morók<sup>a</sup> *n*. **voice (loud)** moróká ná zè *n*. **voice (soft)** moróká na kwáts<sup>a</sup> *n*. **voicebox** ɡɔ̀k <sup>a</sup> *n*. **void** bùlòn *v*. **volcano** ɲáɗúy<sup>a</sup> *n*. **vomit** ɦyɛnɛ́tɔ́n *v*.; ɦyɛ̀nɔ̀n *v*.; ɨlɔ́ɓɔ́tɛtɛ́s *v*. **vomit liquid** ɦyɛnɛ́tɔ́na pɨɔ *v*. **vote for** ɡóózés *v*. **voter** ɡóózésíàm *n*. **voucher** taatsakabáɗ<sup>a</sup> *n*. **vow to harm** ɨmanɛs *v*. **vowel** sʉ̀p <sup>a</sup> *n*. **vowels** sʉpaicíká tódà<sup>o</sup> *n*. **vowels (heavy)** sʉpaicíká ni isaák<sup>a</sup> *n*. **vowels (light)** sʉpaicíká ni ɔlɔ́daák<sup>a</sup> *n*. **vowels (voiced)** sʉpaicíká ni nesíbòs *n*. **vowels (voiceless)** sʉpaicíká ni líîd<sup>a</sup> *n*. **vreevreeew!** tíítɨɨtíì *ideo*. **vroom!** zììì *ideo*.
#### vulture
warthog boar
**vulture** kɔ̀p <sup>a</sup> *n*. **vulture (African white-backed)** kɔtɔl *n*. **vulture (lappet-faced)** náʉ́mɔ *n*. **wa-wa!** kó *nurs*. **wad** ts'àf *n*. **waffle** wasɛ́tɔ́n *v*. **waft** ipúróòn *v*. **wag** iwítsíwítsés *v*. **wag (tail)** ɨwɨɗíwíɗɛ́s *v*. **waggle** iwítsíwítsés *v*. **wail** iƙúétòn *v*.; iƙúón *v*.; iƙúónuƙot<sup>a</sup> *v*.; iwákón *v*.; ƙɔ̀ɗɔ̀n *v*. **wail (make)** ƙɔɗɨtɛs *v*. **wail for** itseniés *v*. **wailing** iwáákós *v*. **waist** róróy<sup>a</sup> *n*. **waist (of clothing)** ikeda ƙwázà<sup>e</sup> *n*. **waist-cloth** riɗiesíƙwàz *n*. **waistline** róróy<sup>a</sup> *n*. **wait (for/on)** kɔɛ́s *v*.; kɔɛtɛ́s *v*. **wait (make)** ɨlarɨtɛtɛ́s *v*.; ɨlwarɨtɛtɛ́s *v*. **wait in vain** koisiés *v*. **wake suddenly** bʉrɛ́tɔ́n *v*.; tsídzètòn *v*. **wake up** ɡonésétòn *v*. **wakefulness** ɡòk<sup>a</sup> *n*. **walk** ɓɛƙɛ́s *v*.; ɨláítɛ́sʉƙɔt<sup>a</sup> *v*.; ɨtátɛ́ɛ́s *v*. **walk (leisurely)** ɲɛ́tɛmá *n*. **walk crunchily** ƙɛƙɛanón *v*. **walk feebly** ɨsɔ́wɔ́ɔ̀n *v*. **walk hesitantly** tsìkòn *v*. **walk laboriously** ɓɛƙɛ́sá ziál *v*. **walk leisurely** ɓɛƙɛ́sá wɛwɛɛs *v*.; ɨtɛ́mɔ́ɔ̀n *v*. **walk on hands** ɓɛƙɛ́sá kwɛ̀tìk<sup>ɔ</sup> *v*. **walk on knees** ɓɛƙɛ́sá kútúŋìk<sup>o</sup> *v*.
**walk on tippytoes** itseɗítséɗòn *v*. **walk slowly** ɨɛ́mɔ́n *v*. **walk small-buttocked-ly** pɛɛ́ɲɛ́mɔ̀n *v*. **walk springily** ɨŋɔ́písɔ̀ɔ̀n *v*. **walk together** ɓɛƙɛ́sínɔ́s *v*. **walk with cane** itséƙóòn *v*. **walker** ɓɛƙɛ́síàm *n*. **walkie-talkie** dʉrʉdʉra na tímoí *n*. **walking stick (insect)** tʉ́w<sup>a</sup> *n*. **walkway** bácík<sup>a</sup> *n*. **wall** ɲarátát<sup>a</sup> *n*. **wall (back interior)** ɦyʉƙʉ́n *n*. **wallet** ɲɔ́pɔ́c *n*. **wallop** inipes *v*. **wander** ɨlɔ́líɛ́sá así *v*.; iwórón *v*.; tɛ́rɛ́s *v*. **wander aimlessly** ɨpɛípɛ́ɛ́sá kíʝá<sup>e</sup> *v*. **wander off** imámáɗós *v*. **wanderer** ɓɛƙɛsɔsíám *n*.; iwórónìàm *n*. **want** bɛ́ɗɛ́s *v*.; ɲʉmɛ́s *v*. **war** cɛma kíʝíkà<sup>e</sup> *n*. **warbler (willow)** dɛ̀dɛ̀s *n*. **ward** ɲáwáɗ<sup>a</sup> *n*. **ward (hospital)** mayaakóniicéhò *n*. **warden (game)** lɔɡɛ́m *n*. **ware** dzíɡwam *n*.; dzíɡwetam *n*.; dzííƙotam *n*. **warm** ɨɓʉ́rɔ́n *v*.; ɨmɛɛs *v*. **warm (make)** ɨɓʉ́rítɛ́sʉƙɔt<sup>a</sup> *v*. **warm (unpleasantly)** laŋádòn *v*. **warm up** ɨɓʉ́rɛ́tɔ̀n *v*.; iɓurímétòn *v*.; ɨɓʉ́rítɛ́sʉƙɔt<sup>a</sup> *v*.; ikues *v*.; ikuetés *v*.; ɨmɔ́lɔ́ŋɛtɛ́s *v*.; iwáŋón *v*. **warming** ìwàŋ *n*. **warn** zízɛ̀s *v*. **warthog** ɡaso *n*. **warthog boar** bèkw<sup>a</sup> *n*.
**warthog piglet** ɡasóím *n*. **warthog sow** ɡasoŋwa *n*. **warty** tɔmɔtsɔkánón *v*. **wary** iŋolíŋólós *v*. **wash** fítés *v*.; ɨlɔ́tɛ́sʉƙɔt<sup>a</sup> *v*. **wash away** fítésuƙot<sup>a</sup> *v*. **wash hands** fítésuƙota kwɛ́tìkà<sup>ɛ</sup> *v*. **wash up** fítésuƙot<sup>a</sup> *v*.; fitetés *v*. **wasp** dɛrɛ́ƙ <sup>a</sup> *n*. **wasp (large)** oŋoridɛrɛ́ƙ <sup>a</sup> *n*. **wasp (small)** ɲólídɛrɛ́ƙ <sup>a</sup> *n*. **waste** eletiésuƙot<sup>a</sup> *v*.; ɨɓalíɓálɛ́s *v*.; ɨlɛkílɛ́kɛ́s *v*.; iɲekes *v*.; iɲékésuƙot<sup>a</sup> *v*.; iɲekíɲékés *v*. **waste (time)** dzuƙés *v*. **waste time of** ɨlarɨtɛtɛ́s *v*.; ɨlwarɨtɛtɛ́s *v*. **wasted away** kɔlɔlánón *v*. **watch** fet<sup>a</sup> *n*.; iteles *v*.; ɲásáat<sup>a</sup> *n*. **watch (here)** iteletés *v*. **watch (spy)** toreɓes *v*. **watch (there)** itélésuƙot<sup>a</sup> *v*. **watch each other** itélínós *v*. **watch the sun** itelesa fetí *v*. **watchful** itsópóòn *v*. **watchman** còòkààm *n*.; itelesíám *n*.; itelesíáma kíʝá<sup>e</sup> *n*. **water** cue *n*.; wetités *v*.; wetitésuƙot<sup>a</sup> *v*. **water (borehole)** ɲatsʉʉmácúé *n*. **water (pond)** tábarɨcue *n*. **water (rock pool)** sátíkócue *n*. **water (rock well)** mɔƙɔrɔ́cúé *n*. **water (tree hollow)** kotímácùè *n*. **water pot** cúédòm *n*. **water source** cuáák<sup>a</sup> *n*. **water table** ʝʉmʉ́cúé *n*. **water-logged** ilébìlèbètòn *v*.
**water-resistant** pʉrákámòn *v*.; pʉráŋámòn *v*.; pʉsɛ́lɛ́mɔ̀n *v*. **waterbuck (Defassa's)** ɲéɓéɓut<sup>a</sup> *n*. **watercourse** cúémúcè *n*. **watercraft** itúɓ<sup>a</sup> *n*. **waterfall** látsóìk<sup>a</sup> *n*. **waterhole** tábàr *n*. **waterily** tsàk<sup>a</sup> *ideo*. **watermelon** nàdɛ̀kwɛ̀l *n*. **waters (amniotic)** baúcùè *n*. **watershed (area)** ɲɛrɛ́t <sup>a</sup> *n*. **watershed (ridge)** murut<sup>a</sup> *n*. **watershed centerpoint** murutéékw<sup>a</sup> *n*. **waterway** cúémúcè *n*. **waterworm (red)** bìɲ *n*. **watery** tsakádòn *v*. **wattle** ɲɛ́tɛ́lɨtɛl *n*. **wave** ɨmáxánɛ́s *v*.; ipukes *v*. **wave (of migration)** bot<sup>a</sup> *n*. **wave around** iwítsíwítsés *v*. **wave in eyes** iwitsíwítsésá ekwitíní *v*. **wave wildly** apɛ́tɛ́pɛ́tánón *v*. **waver** isíƙóòn *v*.; itóŋóòn *v*. **wax (candle)** sɔs *n*. **wax eloquent** isiresa aká<sup>e</sup> *v*. **way** muce *n*. **way (it is)** naítá *subordconn*. **way (method)** ɲɛpɨtɛ *n*. **way (of doing)** muce *n*. **wayfarer** ɓɛƙɛ́síàm *n*. **waylay** ɨɗaarɛ́s *v*.; taɗapes *v*.; taɗapetés *v*. **we (exclusive)** ŋɡw<sup>a</sup> *pro*. **we (inclusive)** ɲjín *pro*. **weak** bʉláʝámɔ̀n *v*.; cucuéón *v*.; daƙwádòn *v*.; ɨɛ́ɓɔ́n *v*.; ɨpáláƙɔ̀n *v*.; ʝuódòn *v*.
#### weakly
**weakly** dàƙw<sup>a</sup> *ideo*.; ʝù<sup>o</sup> *ideo*. **wealth** bàr *n*.; ɲámáli *n*. **wealthy** bàrɔ̀n *v*.; iʝákáánón *v*. **wealthy (get)** bárɛ́tɔ̀n *v*.; barɔnʉƙɔt<sup>a</sup> *v*. **wealthy person** bàrɔ̀àm *n*.; bàrɔ̀nìàm *n*. **wean** topétésuƙot<sup>a</sup> *v*. **weapon** ɛ̂b <sup>a</sup> *n*. **weapons** kúrúɓáa ni cɛmá<sup>ɛ</sup> *n*. **wear** iwales *v*.; ŋábès *v*. **wear (beads)** otés *v*. **wear a feather** iwalesa túkà<sup>e</sup> *v*. **wear across** ízokomés *v*. **wear down** ɨlɔ́ítɛ́sʉƙɔt<sup>a</sup> *v*. **wear out** ats'ímétòn *v*. **weary** ɨlɔ́yɔ́n *v*. **weather** dìdì *n*. **weather (cold)** ɨɛ́ɓɔ́na kíʝá<sup>e</sup> *v*. **weather (hot)** hábona kíʝá<sup>e</sup> *n*. **weave** bɛrɛ́s *v*.; ɨkɔɗíkɔ́ɗɔ̀n *v*.; iƙulúƙúlòn *v*.; tʉtʉkɛs *v*. **weave around** iléƙwéries *v*. **weave woof** kámáriés *v*. **weaverbird** tsario *n*. **web** abûb<sup>a</sup> *n*. **wedding beer** ɲalakʉts<sup>a</sup> *n*. **wedding sacrifice** ɲɛ́kílama *n*.; ɲɛkʉma *n*. **wedge** pokés *v*.; rɨɗɛ́s *v*. **wedge-shaped** lɨkíɗímɔ̀n *v*. **Wednesday** Ɲákásíá àɗìk<sup>e</sup> *n*. **weed** ɗɔanés *v*. **weed species** karimésém *n*.; ɲɛ́ʉrɛrɛ́*n*. **weed(s)** ɗɔan *n*. **weeding** ɗɔan *n*. **weedy** tsèkòn *v*.
**week** ɲásaɓét<sup>a</sup> *n*. **weeny** ɡɔɗírímɔ̀n *v*. **weep** ƙɔ̀ɗɔ̀n *v*. **weep (make)** ƙɔɗɨtɛs *v*. **weevil (maize)** lɔkaʉɗ<sup>a</sup> *n*. **weigh** ɨpɨmɛs *v*. **weigh down** isites *v*. **weigh words** íziɗesa tóda<sup>e</sup> *v*. **weighed down** ɨʉ́ƙɔ́n *v*. **weight (gain)** tuɓútónuƙot<sup>a</sup> *v*. **weight dwon** ɨnʉɛs *v*. **weighty** ìsòn *v*. **weird thing** bàdìàm *n*. **weirdo** ƙʉts'áám *n*. **welcome** tɛ́bɛtɛ́s *v*. **welcome warmly** ewanes *v*.; ewanetés *v*. **welcome! (plural)** tɛ́bɛtaná bìt<sup>a</sup> *v*. **welcome! (singular)** tɛ́bɛtaná bì *v*. **well** iŋáléòn *v*.; maráŋík<sup>e</sup> *v*. **well (get)** iŋáléètòn *v*. **well (hand-dug)** dzòn *n*. **well (in rocks)** mɔƙɔr *n*. **well (natural)** ɲɛlɛ́lyá *n*. **well up** ʝɨríʝírɛ̀tɔ̀n *v*.; ʝɨríʝírɔ̀n *v*. **well worn (of paths)** ɗìwòn *v*. **well-cooked (very)** ɗʉ̀m *ideo*. **well-done** ɗʉmʉ́dɔ̀n *v*. **well-fed** zízòn *v*. **well-known** arútón *v*.; ɦyoós *v*. **well-off** bàrɔ̀n *v*.; iʝákáánón *v*. **well-prepared** toikíkón *v*. **well-seasoned** ɛ̀fɔ̀n *v*. **west** tábày<sup>a</sup> *n*. **westerly direction** tábaɨxan *dem*. **western rain** tsóéàm *n*. **westerner** tábàìàm *n*. **wet** ɗɔ̀ƙɔ̀n *v*.
468
**wet (become)** ɗɔƙɔnʉƙɔt<sup>a</sup> *v*. **wet (make)** ɗɔƙítɛ́sʉƙɔt<sup>a</sup> *v*. **whack** inipes *v*.; ɨɲatɛs *v*. **whack repeatedly** ɨɲatiés *v*. **whack!** pùk<sup>u</sup> *ideo*.; zɨɗát<sup>a</sup> *ideo*. **whale (of a)** kébàdà *n*.; nábàdà *n*.; nébàdà *n*. **wham!** ɡwàʝ<sup>a</sup> *ideo*.; kùm *ideo*. **what about (when) …?** ndóó *subordconn*. **what about …?** ndóó *prep*. **what color?** kɨtɔ́ɔ́sɔ̀n *v*. **what exactly …?** ín *adv*. **what if (it is)** ndóó mìtìɛ̀ *v*. **what shape?** kɨtɔ́ɔ́sɔ̀n *v*. **what texture?** kɨtɔ́ɔ́sɔ̀n *v*. **what?** ìs *pro*. **whatcha-ma-callit** ndaicé *n*. **whatever!** hà *interj*.; ndéé *interj*. **whats-their-name** tatanám *n*. **whatsoever** mùɲ *quant*. **wheedle** ɨmámɛ́ɛ́s *v*. **wheel** dɛ *n*. **wheel (steering)** ɲókokor *n*. **wheelbarrow** ɲáɡaɗiɡáɗ<sup>a</sup> *n*.; ɲaƙaari *n*. **wheelchair** ɲamɨɨlɨa ŋwáxɔ̀nìàmà<sup>e</sup> *n*. **wheeze** ɛ́mítɔ̀n *v*.; xíƙwítós *v*. **whelp** ŋókíìm *n*. **when** náà *subordconn*.; náa táà *subordconn*.; nɛ́ɛ́*subordconn*. **when (a while ago)** nótsò *subordconn*. **when (earlier today)** náà *subordconn*. **when (hypothetically)** na *subordconn*. **when (long ago)** nòò *subordconn*. **when (yesterday)** sìnà *subordconn*. **when already** térútsù *adv*.; tórútsù *adv*.
**when … had (a while ago)** nànòò *subordconn*. **when … had (earlier)** nanáà *subordconn*. **when … had (yesterday)** nàsàmʉ̀ *subordconn*. **when?** ńtóódò *n*. **where (it is)?** ndayúk<sup>o</sup> *n*. **where?** ndaík<sup>e</sup> *pro*.; ńtá *pro*. **whet** banɛ́s *v*. **whether** mísì *subordconn*. **whetstone** lósùàɲ *n*. **which (a while ago)** nótsò *rel*. **which (a while ago, pl.)** nútsù *rel*. **which (earlier)** nák<sup>a</sup> *rel*. **which (earlier, pl.)** níkⁱ *rel*. **which (long ago)** nòk<sup>o</sup> *rel*. **which (long ago, pl.)** nùk<sup>u</sup> *rel*. **which (one)?** ńtɛ́ɛ́n *pro*. **which (ones)?** ńtíɛ́n *pro*. **which (plural)** ni *rel*. **which (singular)** na *rel*. **which (yesterday)** sìn *rel*. **which (yesterday, pl.)** sìn *rel*. **which way?** nday<sup>o</sup> *n*. **whiff** kɔín *n*. **while** názɛ̀ƙwà *n*. **while (earlier today)** tenák<sup>a</sup> *adv*. **while (long ago)** tènòk<sup>o</sup> *adv*. **while (not yet)** káɗìò *subordconn*. **while (yesterday)** tèsìn *adv*. **while hungry** ɲɛ́ƙín<sup>ɔ</sup> *n*. **whimper** ɨɲɨɨɲíɔ̀n *v*.; ɨɲíɲíɔ̀n *v*. **whine** ɨŋʉ́ŋʉ́nɔ̀n *v*. **whip** ɨɗɨtsɛs *v*.; ɨnɔmɛs *v*. **whip (leather)** ɲánɨnɔ́*n*. **whip all over** iléƙwéries *v*.
**whip back and forth** ɨtsɔkítsɔ́kɛ́s *v*.; nɨƙwíníƙɔ̀n *v*. **whip lightly** irwatesa kíx<sup>o</sup> *v*.; ízaɓɨzaɓɛ́s *v*.; zaɓatiés *v*. **whippily** lɛ̀ts'<sup>ɛ</sup> *ideo*. **whippy** lɛts'ɛ́dɔ̀n *v*. **whirl around** tɔpɨrípírɔ̀n *v*. **whirlwind** lòtàbùsèn *n*. **whisper** sɛsɛanón *v*. **whisperer** sɛsɛanónìàm *n*. **whistle** fóʝón *v*.; síƙón *v*.; sìƙw<sup>a</sup> *n*. **whistle (metallic)** ɲákápɨrɨt<sup>a</sup> *n*. **whistle (wooden)** ɲétúle *n*. **whistle for** iwéwérés *v*.; iwówórés *v*. **white** ɓèts'òn *v*. **white (glittering)** pír *ideo*. **white (make)** ɓets'itetés *v*. **white (of many)** ɓets'aakón *v*. **white (slightly)** ɓèts'ìɓèts'òn *v*. **white (very)** lìà *ideo*.; pàkⁱ *ideo*. **white person** ɓèts'ònìàm *n*.; ɲémúsukit<sup>a</sup> *n*. **white with black eye patches** tulíánètòn *v*. **white-eye (yellow)** baratíɡwà *n*. **white-faced** ŋoléánètòn *v*. **whiten** ɓets'itetés *v*. **whitish** ɓèts'ìɓèts'òn *v*.; ɓɛts'ɨɗɔ́ɗɔ́n *v*.; xóuxówòn *v*. **whiz by** ídzesa así *v*. **who knows?** ndóó ɦyè *n*. **who?** ǹdò *pro*. **whoa!** otí *interj*. **whole** ɗàŋìɗàŋ *quant*.; mùɲ *quant*.; mùɲùmùɲ *quant*.; tsíɗ<sup>ɨ</sup> *quant*.; tsíɗɨtsíɗ<sup>ɨ</sup> *quant*.
**whoosh** wààà *ideo*. **whoosh!** wùòò *ideo*. **whooshǃ** fùùt<sup>u</sup> *ideo*. **whorl** iyérónuƙot<sup>a</sup> *v*. **whorled** iyérón *v*. **why … of course!** ɲák<sup>a</sup> *adv*. **why?** isiɛník<sup>ɛ</sup> *n*. **whydah** lɔɔrʉ́k <sup>a</sup> *n*. **wicked (of many)** ɡaanaakón *v*. **wickedness** ɡaánàs *n*. **wide** zòòn *v*. **wide (of many)** zeikaakón *v*. **wide awake** ɡwɛɲɛ́mɔ́n *v*. **wide open** folólómòn *v*. **wide-eyed** ɡonésá kom<sup>o</sup> *v*.; ŋɔɓɔ́dɔ̀n *v*.; ŋɔ̀ɓɔ̀n *v*. **wide-eyedly** ŋɔ̀ɓ ɔ *ideo*. **wide-legged** ƙaƙótsómòn *v*. **wide-mouthed** ɓolóɲómòn *v*.; laɓáɲámòn *v*.; lafárámòn *v*. **widen** zoonuƙot<sup>a</sup> *v*. **widened** ɨatɔs *v*. **widow(er)** lóméléw<sup>a</sup> *n*.; ɲepúrósit<sup>a</sup> *n*. **wife** cek<sup>a</sup> *n*. **wife (co-)** ɛán *n*. **wife (his)** ntsícék<sup>a</sup> *n*. **wife (last)** ʝìrìàm *n*.; kárátsìkààm *n*. **wife (my)** ɲ́cìcèk<sup>a</sup> *n*. **wife (of someone)** ámácèk<sup>a</sup> *n*. **wife (your)** bicék<sup>a</sup> *n*. **wiggle in** ɨnɨƙwíníƙwɛ́s *v*. **wiggly** lokilókón *v*. **wild (area)** ɲáraƙɔ́áƙw<sup>a</sup> *n*.; ɲáraƙw<sup>a</sup> *n*. **wild animal** ínwá na riʝááƙɔ̀ <sup>ɛ</sup> *n*. **wild hunting dog** tsoe *n*.
woman (old)
**wild olive tree** dèmìyw<sup>a</sup> *n*. **wilderness** ɲáraƙɔ́áƙw<sup>a</sup> *n*.; ɲáraƙw<sup>a</sup> *n*.; ɲɛ́kítɛla *n*. **wildfire** kóméts'àɗ<sup>a</sup> *n*. **wildlife authorities** cookaika ínó<sup>e</sup> *n*.; lɔɡɛ́m *n*. **will not** ńtá *adv*. **willing** tsolólómòn *v*. **wilt** ɨtɔ́ɗɔ́n *v*.; laʝámétòn *v*. **wimpy** kalɛ́ɛ́tsɛránón *v*.; sikwárámòn *v*. **win** ɨlámɔ́n *v*.; ɨsʉkɛs *v*.; sʉ́kɛ́s *v*. **win the support of** sʉ́bɛ̀s *n*. **wind** ilúƙúretés *v*.; ɨnɔɛs *v*.; lúkúɗukuɗánón *v*.; suɡur *n*. **wind around** iƙulúƙúlòn *v*.; ɨlɔkílɔ́kɛ́s *v*.; ɨlɔkílɔ́kɛtɛ́s *v*.; imanímánés *v*.; ɨnɔɛtɛ́s *v*.; kɛkɛrɛs *v*.; tamánɛ́tɔ̀n *v*. **wind up** ɨmakímákɛ́s *v*. **winding** tukúɗúkuɗánón *v*. **window (of a house)** hòwɛ̀l *n*. **windpipe** moróká ná zè *n*. **wine** ɲéviiní *n*. **wine (Rhus natalensis)** mɨsáícùè *n*. **wing** taban *n*. **wink** íbɛ̀ɗìbɛ̀ɗɔ̀n *v*.; ɨmíʝílɛ́s *v*.; irwapírwápòn *v*. **winnow** ilélébés *v*. **winnow (by pouring)** síkɔ́ɔ́rɛ́s *v*. **winnow (by tossing)** fɔ́tɛ́s *v*. **wipe** ŋííɗɛ́s *v*. **wipe (rear end)** ɨtɔ́tɔ́rɔ̀n *v*. **wipe clean** ɨƙʉ́ʉ́lɛ́s *v*.; ɨtsɨɗɛs *v*. **wipe off** kánɛ́s *v*.; ŋííɗɛ́sʉ́ƙɔt<sup>a</sup> *v*.; ɲimirés *v*. **wipe out** bulútésuƙot<sup>a</sup> *v*.; ɨƙɔmɛs *v*.; ɨmʉ́ɲɛ́sʉƙɔt<sup>a</sup> *v*.; ɨmʉɲɛtɛ́s *v*.; ɨtsʉtɛs *v*.; ɨtsʉ́tɛ́sʉƙɔt<sup>a</sup> *v*.; kánɛ́s *v*.
**wipe up** kánɛ́s *v*.; ŋííɗɛtɛ́s *v*. **wiped out** ikarímétòn *v*.; kanímétòn *v*.; ziálámòn *v*.; zíkímétòn *v*.; ziláámòn *v*. **wire** ɲáwáya *n*. **wiry** simánón *v*. **wisdom** nɔɔ́s *n*. **wise** nɔɔsánón *v*. **wise person** nɔɔsáàm *n*. **wiser (grow)** nɔɔsánétòn *v*. **wish for** ƙanetés *v*.; ɲʉmɛ́s *v*.; wíránés *v*. **witchdoctor** ŋƙwa *n*. **with** ńdà *prep*. **with hunger** ɲɛ́ƙín<sup>ɔ</sup> *n*. **Withania somnifera** ikitínícɛmɛ́r *n*.; ɲónomokére *n*. **withdraw** ɨpɛ́ɛ́rɔ̀n *v*.; ɨsʉ́rʉ́mɔ̀n *v*. **wither** ɨtɔ́ɗɔ́n *v*.; laʝámétòn *v*. **wither up** mɔsɔnʉƙɔt<sup>a</sup> *v*. **withered** ƙɔ́rɔmɔmɔ́n *v*.; mɨtírímɔ̀n *v*.; mɔ̀sɔ̀n *v*. **withheld from** rébìmètòn *v*. **withhold from** rébès *v*. **without eating** kùk<sup>u</sup> *ideo*. **witness** enésúƙotíám *n*.; itelesíám *n*.; ŋítsaɗénìàm *n*. **witness (false)** kɛ́rínɔ́síàm *n*.; lóliit<sup>a</sup> *n*. **wives** cɨkám *n*. **wizard** bàdìàm *n*. **wizard (hyena-riding)** otsésíama haúùⁱ *n*. **wizardry** badirét<sup>a</sup> *n*.; badirétínànès *n*.; ƙʉts'ánánès *n*. **wobbly** ɗɔxɔ́dɔ̀n *v*.; ɡwèlèʝ<sup>e</sup> *ideo*. **wocked** ɡaanón *v*. **wolf down (food)** ifáfúkés *v*.; ŋɔfɛ́s *v*. **woman** cek<sup>a</sup> *n*. **woman (foreign)** ɦyɔ̀cèk<sup>a</sup> *n*. **woman (old)** dúnéìm *n*.; fɔ́ɗítíníàm *n*.
**woman (unmarried)** dekitíníàm *n*.; ɲàràm *n*. **womanhood** cekínánès *n*. **womanliness** cekínánès *n*. **womb** epúáw<sup>a</sup> *n*.; ɲapéryɛ́t <sup>a</sup> *n*. **women** cɨkám *n*. **women (young unmarried)** ɲèr *n*. **wonders** itíónàs *n*. **wonders (perform)** itíónòn *v*. **woo** sits'és *v*. **wood** dakw<sup>a</sup> *n*. **wood (piece of)** dakw<sup>a</sup> *n*. **woodland** dakúáƙw<sup>a</sup> *n*.; ríʝíkaaʝík<sup>a</sup> *n*. **woodpecker** cɛŋ *n*. **woods** ríʝ<sup>a</sup> *n*. **woodworker** ɲáɓáòìkààm *n*. **woof** íɡòmòn *v*. **wool** ɗóɗòsìts'<sup>a</sup> *n*. **woolly** saúkúmòn *v*. **woozy** imáúròn *v*.; itikítíkòn *v*. **word** mɛnéékw<sup>a</sup> *n*.; tódèèkw<sup>a</sup> *n*. **work** ikásíés *v*.; ikásíitetés *v*.; ɲákási *n*.; ɲetits<sup>a</sup> *n*.; terêɡ<sup>a</sup> *n*.; tereɡanés *v*.; teréɡanitetés *v*. **work (knead)** dʉbɛ́s *v*. **work (temporary)** ɲɛ́lɛ́ʝɨlɛʝ<sup>a</sup> *n*. **work (the soil)** tɔkɔ́bɛs *v*. **work contract** teréɡikabáɗ<sup>a</sup> *n*. **work for pay** teréɡa na kaúdzò<sup>e</sup> *n*. **work in (insert)** ɨnɨƙwíníƙwɛ́s *v*. **work into** iwoɗíwóɗés *v*. **work of art** iyomam *n*. **work on (beat)** iɗiles *v*. **work over (beat)** iɗiles *v*. **work project** ɲɛ́prɔ́ʝɛ̀kìt<sup>a</sup> *n*. **work temporarily** ɨlɛʝílɛ́ʝɛ́s *v*.
**work with long tool** iƙoríƙórés *v*. **workable** ikásíetam *v*. **worked up (sexually)** iɓurímétòn *v*. **worker** ɲákásìàm *n*.; teréɡìàm *n*. **worker termite** nateɓú *n*. **working for government** ɲeryaŋínánès *n*. **workshop** ɲókós *n*. **world** kíʝ<sup>a</sup> *n*. **World Vision** Loúnoy<sup>a</sup> *n*. **world's end** tasálétona kíʝá<sup>e</sup> *n*. **worm** ƙʉts'<sup>a</sup> *n*. **worm (bee-eating)** ɡɔɗɔ́ɛ̀ *n*. **worm (biting)** hoƙʉts'<sup>a</sup> *n*. **worm (intestinal)** ídèm *n*. **worn out** ɨlɔ́ɛ́tɔ̀n *v*.; rɛsɛ́dɔ̀n *v*.; ziálámòn *v*.; zíkímétòn *v*.; ziláámòn *v*. **worn out (become)** ziláámètòn *v*. **worn smooth** pikódòn *v*. **worried** ísánòn *v*.; tsʉ́kʉɗʉ́ɗɔ́n *v*. **worried (become)** ísánonuƙot<sup>a</sup> *v*. **worry** alólóŋòn *v*. **worse (become)** ɡaanónuƙot<sup>a</sup> *v*. **worsen** ɡaanítésuƙot<sup>a</sup> *v*.; ɡaanónuƙot<sup>a</sup> *v*.; rúbès *v*. **worship** itúrútés *v*.; wáán *n*.; wáán *v*. **worship leader** wáánɨtɛtɛ́síàm *n*. **worshipper** itúrútésìàm *n*. **wort (fermenting)** ɲɛ́wíɲɨwɨɲ *n*. **worthless thing** tsar *n*. **worthlessness** tsarínánès *n*. **would have … (earlier today)** ƙánàk<sup>a</sup> *adv*. **would have … (long ago)** ƙánòk<sup>o</sup> *adv*. **would have … (yesterday)** ƙásàm *adv*. **wound** ɔ́ʝ <sup>a</sup> *n*. **wound (bullet)** bʉbʉnɔ́ɔ́ʝ <sup>a</sup> *n*.
#### wow!
**wow!** ín *adv*.; wúlù *interj*.; yáŋ *interj*. **wowǃ** ábaŋ *interj*. **wraith** kúrúkúr *n*.; lopéren *n*.; tás *n*. **wrap** ɨpʉ́pʉ́ŋɛ́s *v*. **wrap (with clothing)** ikáburés *v*. **wrap around** ɨlɔkílɔ́kɛ́s *v*.; ɨlɔkílɔ́kɛtɛ́s *v*. **wrap up** ɡubésúƙot<sup>a</sup> *v*.; ɨpʉ́pʉ́ŋɛtɛ́s *v*. **wreath of reeds** nàtsìkw<sup>a</sup> *n*. **wreck** ináƙúés *v*.; ináƙúetés *v*.; ipáríés *v*. **wrecked** ináƙúós *v*.; ináƙúotós *v*. **wrest** ŋusés *v*. **wrestle** kóríètòn *v*.; kɔrɔanón *v*. **wrestle out** ɨkwɛ́rɛ́ɗɔ̀n *v*. **wretched** tsʉ́kʉɗʉ́ɗɔ́n *v*. **wretched as a dog** iŋókíánón *v*. **wriggle** ŋʉɗʉŋʉ́ɗɔ́n *v*. **wriggle around** akwɛ́tɛ́kwɛ́tánón *v*. **wriggle free** ɨkwɛ́rɛ́ɗɔ̀n *v*. **wriggle in** ɨnɨƙwíníƙwɛ́s *v*. **wriggle into** ɨɓɨtsíɓítsɛ́s *v*. **wriggle out** ɡwìrɔ̀n *v*. **wriggly** wʉlʉ́kʉ́mɔ̀n *v*. **wring** ʝʉ́tɛ́s *v*.; tʉtsʉɛs *v*. **wring out** ʝʉ́tɛ́sʉƙɔt<sup>a</sup> *v*.; tʉtsʉ́ɛ́sʉƙɔt<sup>a</sup> *v*. **wrinkled** rʉʝanón *v*.; rʉʝʉrʉ́ʝánón *v*.; turúʝón *v*.; zamʉʝánón *v*. **wrinkly** turuʝúrúʝánón *v*. **wrist** kwɛtámórók<sup>a</sup> *n*. **wrist knife** ɨbɔt<sup>a</sup> *n*.; ɲáɓaarát<sup>a</sup> *n*. **wristwatch** ɲásáat<sup>a</sup> *n*. **write** ɨƙɨrɛs *v*. **writhe around** akwɛ́tɛ́kwɛ́tánón *v*. **writing desk** ɲéméza na íƙìrà<sup>ɛ</sup> *n*. **written** ɨƙɨrɔs *v*. **wrong thing** róŋ *n*.
**wrongdoer** ɲɔ́mɔkɔsáàm *n*. **wussy** kalɛ́ɛ́tsɛránón *v*.; sikwárámòn *v*. **Ximenia americana** kunét<sup>a</sup> *n*. **xiphoid process** toroɓóɔ́k <sup>a</sup> *n*. **yank** ɗʉ́rɛ́s *v*.; ɗʉtɛ́s *v*.; iɓwates *v*.; ipoles *v*. **yank out** ɗʉrɛtɛ́s *v*.; ɗʉtɛtɛ́s *v*.; ipoletés *v*. **yank over** iɓwatetés *v*. **yank this way** ɨtsɔrɛtɛ́s *v*. **yank up** ipoletés *v*. **yard** awááƙw<sup>a</sup> *n*.; ɔkɔ́ts<sup>a</sup> *n*. **yawn** áƙáfòn *v*. **yawning** hádòlòmòn *v*.; laɓáɲámòn *v*.; lafárámòn *v*. **yeah** ee *interj*. **yeah right!** héé' *interj*. **yeah!** ńtí *interj*. **year** kaɨn *n*. **year after next** kaɨnɔ na tso *n*. **year before last** kaɨnɔ nótso *n*.; nɔkɛɨn *n*. **Year of Lopíar** Lopíar *n*. **Year of Lotííra** Lotííra *n*. **Year of Lɔkʉlɨt** Lɔkʉlɨt<sup>a</sup> *n*. **Year of Nawólóʝam** Nawólóʝam *n*. **yearn** ítánòn *v*. **yearn for** ítánésuƙot<sup>a</sup> *v*. **years ago** kaíníkò nùk<sup>o</sup> *n*. **yeast** sîb<sup>a</sup> *n*. **yell** bofétón *v*.; bófón *v*.; ɨkílɔ́n *v*.; iƙúétòn *v*.; iƙúón *v*.; iƙúónuƙot<sup>a</sup> *v*.; nɔsátón *v*. **yell at** ɨyáyɛ́ɛ́s *v*. **yeller** nɔ̀sààm *n*. **yellow** ɲaŋáánètòn *v*. **yellow color** ɗukes *n*. **yellowish color (gazelle)** kodow<sup>a</sup> *n*.
zygomatic area
**yelp** iƙwéón *v*. **yes** ee *interj*. **yester-** bàts<sup>e</sup> *adv*. **yesterday** sáásò sìn *n*. **yesteryear** kaɨnɔ sɨn *n*. **yikes!** wúlù *interj*. **yip** iƙwéón *v*. **yogurt** ŋakɨɓʉk<sup>a</sup> *n*. **yolk (egg)** ɗukes *n*. **you (plural)** bìt<sup>a</sup> *pro*. **you (singular)** bì *pro*. **you dog!** ŋók<sup>a</sup> *n*. **young** kwátsón *v*. **young (of many)** kwátsíkaakón *v*. **young children** kómósikaa ɓets'aakátìk<sup>e</sup> *n*. **young female** wâz *n*. **young man** karatsʉ́náám *n*. **young monkeys** lɔ́tɔ́ɓàɡwàs *n*. **young people** yús *n*. **young tortoise** kàèìm *n*. **youngster** im *n*. **your (plural)** bìt<sup>a</sup> *pro*. **your (singular)** bì *pro*. **yours (plural)** bitiɛn *pro*. **yours (singular)** biɛ́n *pro*. **yourself (singular)** binêb<sup>a</sup> *n*. **yourselves (plural)** bitinebitín *n*. **youth** yús *n*. **youth (be a)** ɨsɔ́rɔ́kánón *v*. **youth (male)** karatsʉ́na *n*.; ŋímɔ́kɔka *n*.; ŋímɔ́kɔkáám *n*. **youthful (of middle-age)** toipánón *v*. **youthful adult** toipánónìàm *n*.
**yum-yum!** ɓá *nurs*.; mamá *nurs*.
**yum-yum! (for milk)** nʉʉnʉ́ *nurs*. **yummily** ɗɔ̀k ɔ *ideo*. **yummy** ɗɔkɔ́dɔ̀n *v*.; ɡwéts'ón *v*. **Zanthoxylum chalybeum** rukûdz<sup>a</sup> *n*. **zebra** zɨn *n*. **Zehneria scabra** lótórobét<sup>a</sup> *n*. **zigzag** ɨkɔɗíkɔ́ɗɔ̀n *v*.; lúkúɗukuɗánón *v*. **zigzagging** tukúɗúkuɗánón *v*. **zing!** fiuu *ideo*.; líùù *ideo*.; pìùù *ideo*. **zip over** toɓésá así *v*. **zipper** ɲɛ́ʝíp<sup>a</sup> *n*. **Ziziphus mauritiana** ɨláŋ *n*. **Ziziphus mucronata** tílàŋ *n*. **zlop!** pùk<sup>u</sup> *ideo*. **zone** ɲɛ́tɛɛr *n*. **zoom!** wír *ideo*.; yír *ideo*. **zorilla** ɲewuruŋorok<sup>a</sup> *n*. **zucchini** lomuƙe *n*. **zygomatic area** matáŋ *n*.
#### yelp
## **Part IV Grammar sketch**
1 Introduction
## **1 Introduction**
Although the bulk of this book is devoted to the dictionary and reversal index, the following section offers an overview sketch of Ik grammar that covers most important features of the total grammatical system. Those who wish to dig deeper are encouraged to consult the fuller treatment published as *A grammar of Ik (Icétód): Northeast Uganda's last thriving Kuliak language* (Schrock 2014), which is available for free downloading from several websites on the internet.
Linguistic concepts are most easily defined with linguistic terminology. Thus, due to limitations of time and space, this sketch of Ik grammar is geared in style toward the general linguist. And yet a primary aim has been to clearly define some of the key terms used and to describe the grammatical structures in simple, straightforward language. Unfortunately, some of the discussion may still remain opaque to non-linguist readers. If such persons wish to know more, I am very willing to clarify or explain in layman's terms any point raised in this grammar sketch. Feel free to contact me any time at: [email protected].
The grammar sketch begins with the language's sound system (phonology) and then proceeds to words and word-building strategies (morphology). It ends with a shallow dip into syntax. Because of its length and technical nature, the grammar sketch is probably most useful as a reference tool. However, should the reader have the opportunity, it may prove beneficial to read the sketch from front to back in order to gain a bird's-eye view of the whole system.
Learning any language from printed sources alone is rarely ideal. Rather, every learner would ideally have the chance to soak up language naturally as children do. Sadly, most adult learners do not have that luxury. Because of that, I recommend creatively mixing language-learning approaches to suit one's personality, learning style, schedule, and responsibilities. Studying grammar from a book like this one will not appeal to everyone, yet all learners will occasionally get stuck on points of grammar during the course of their learning. Just as the foregoing dictionary can help you fill in gaps where specific words need to be, this grammar sketch can help fill in holes in your understanding of how Ik works.
## **2 Phonology**
## **2.1 Consonants and vowels**
Ik has an array of thirty consonants and nine vowels, which are presented in Table 1. In the table's first column are shown the alphabetical letters used to represent these sounds. The second column shows the phonetic symbol for the
#### Grammar sketch
sound used by the International Phonetic Alphabet (IPA). Then in the third column, an approximate English equivalent is given in bold typeface, or else an explanation of how the sound is made if there is no English approximation.
Those sounds in Table 1 that have a small square under the IPA symbol are pronounced with the tip of the tongue a bit farther forward than in English. Especially [d], [n ̻ ̻], and [t] are affected; sometimes they are fronted so much that ̻ they touch the back of the front teeth. It is important not to pronounce [d] ex- ̻ actly like an English 'd' as this sounds more like the Ik sound [ɗ] which contrasts with [d]. The sounds [ɓ, ɗ, ɠ, ʝ] are called ̻ implosives because they are made by 'imploding' or sucking air into the mouth rather than expelling air from the lungs. The sounds [k'] and [ts'] are called ejectives because they are made by ejecting air from the throat cavity instead of from the lungs. Lastly, the sound [ɦʲ], unlike an [h], is made with the vocal chords vibrating, giving it a raspy, throaty sound. It only occurs at the beginning of words. The nine Ik vowels – [a, e, ɛ, i, ɨ, ɔ, o, ʉ, u] – operate in a vowel harmony system, which is discussed in §2.5.
## **2.2 Consonant devoicing**
At the end of an Ik word, if silence immediately follows, voiced consonants are devoiced. In other words, they sound more like unvoiced consonants in that environment. This is similar to German, for instance, where the word *Tag* 'day' is pronounced as [tak]. Consonant devoicing most noticeably affects /d/ and /g/ in Ik, as when *êd* 'name' sounds like [êt] or when *hɛ̀g* 'marrow' sounds like [hɛ̀k].
## **2.3 Vowel devoicing**
Ik vowels are also devoiced before a pause. This is important because every word in every grammatical context – without exception – ends in a vowel. If that final vowel is not immediately followed by another sound, then it is whispered or even left totally inaudible (for example, after the consonants /f, m, n, ɲ, ŋ, r, s, z, ʒ/). It has become a tradition in scholarly writing on Ik to write whispered vowels with the following raised (superscript) symbols: <<sup>i</sup> ,ᶤ,ᵉ,ᵋ,ᵃ,ᵓ,ᵒ,ᶶ,ᵘ>.
## **2.4 Morphophonology**
## **2.4.1 Deaffrication**
The affricates /c/ and /j/ are occasionally deaffricated or 'hardened' into their nonaffricate counterparts /k/ and /g/, respectively. This is not a general phonological
Table 1: Ik sound inventory
#### Grammar sketch
tendency in the language but is, rather, limited to a small handful of words. Moreover, the principle is applied in different ways to different words. For instance, in the word *muceé-* 'path, way', the /c/ is hardened to /k/ when the word is used in the instrumental case (see §7.7): *muko* 'on the way'. Secondly, as an instance of idiolectal variation, the plural inclusive pronoun *ɲjíní-* 'we all (including addressees)' is pronounced idiosyncratically as *ŋgíní-* by a minority of speakers. Thirdly, when the words *Icé-* 'Ik people' and *wicé-* 'children' are declined for the nominative or instrumental cases, their /c/ hardens to /k/. This type of deaffrication can be clearly seen in a case declension, like the one in Table 2. Note that, as explained later in §2.4.3, all cases have non-final and final forms:
Table 2: Case declension of *Icé-* 'Ik' and *wicé-* 'children'
#### **2.4.2 Haplology**
In Ik, when a consonant in one morpheme is made at the same place of articulation as a consonant in the next morpheme, haplology may occur – the deletion of the first of the two similar consonants. One example of this involves the venitive suffix {-ét-} and the andative suffix {-uƙot-}, both of which end in /t/. If another suffix containing /t/, /d/, or /s/ is attached to either of these, their final /t/ may be omitted. To illustrate this, Table 3 presents a conjugation of the verb *ŋatɛ́tɔ́n* 'to run this way'. Notice how the /t/ in {-ét-} disappears from the suffix in the forms for 2sg ('you'), 1pl.inc ('we all'), and 2pl ('you all'). The 3pl form ('they') is an exception as it does not drop its final /t/ in the same environment.
A second example of haplology occurs when a verb root ending in /g/, /k/, or /ƙ/ is followed directly by the andative suffix {-uƙot-}. When this happens, the
#### 2 Phonology
Table 3: Haplology in *ŋatɛ́tɔ́n* 'to run this way'
Table 4: Haplology in verbs ending in a velar consonant
final velar consonant of the verb root gets omitted in anticipation of the velar /ƙ/ in {-uƙot-}. Table 4 illustrates this by listing a few verbs ending in /g/, /k/, or /ƙ/, which disappear when the next morpheme is the andative suffix {-uƙot-}.
#### **2.4.3 Non-final consonant deletion**
Ik makes a clear distinction between non-final and final forms of all morphemes and words. Presumably this is to delineate syntactic boundaries, often with stylistic overtones. Non-final forms are those that occur within a string of speech, with at least one element immediately following them. Final forms, by contrast, are those that occur at the end of a string of speech, before a pause, with nothing immediately following. This basic distinction was already shown to affect the voicing of vowels in §2.3. In the case of a small number of morphemes, it also affects consonants. Table 5 presents a few of these morphemes whose final forms contain consonants that are omitted in their non-final forms. The first column of the table shows the underlying form (uf) of the morpheme in question. This is followed in the next two columns by the non-final (nf) and
#### Grammar sketch
final (ff) forms that actually occur in speech. Notice how the non-final forms are missing one consonant that is fully present in the uf and the ff.
Table 5: Consonant deletion in non-final forms
#### **2.4.4 Vowel assimilation**
In addition to consonants, Ik vowels also undergo phonological changes at morpheme boundaries. For instance, when two dissimilar vowels come in contact with each other as a result of two morphemes joining together, there is a powerful urge for them to become more like each other. This vowel assimilation was already seen at work in Table 4, as when putting the root *torík-* 'lead' and affix *-uƙot-* 'away' together led to *torííƙot-* instead of \**toríúƙot-*. It is also seen in Table 6 where the 'yester-' adverb *bàtsè* becomes *bèè* in its non-final form instead of \**bàè*. Ik vowel assimilation only takes place between morphemes and not inside morphemes. Inside morphemes, many combinations of dissimilar vowels are allowed, for example in *kaɨn* 'year', *mɛ̀ʉ̀r* 'drongo', and *kɔín* 'scent'.
Ik vowel assimilation can be clearly seen throughout the lexicon, as when the transitive infinitive suffix {-és} and the intransitive infinitive suffix {-òn} are affixed to verb roots. If the verb root that these suffixes attach to ends in /a/ or /e/, the vowel of the suffix fully assimilates it. Table 6 offers a few examples of this kind of vowel assimilation in verbal infinitives.
Another environment illustrating Ik vowel assimilation is the case declension of nouns. Since all Ik nouns end in a vowel, and since seven of the eight case suffixes consist of or contain a vowel, case suffixation creates a fertile ground for
#### 2 Phonology
#### Table 6: Vowel assimilation in verbal infinitives
Table 7: Vowel assimilation in the declension of *ŋókí-* 'dog'
vowel assimilation. For example, as Table 7 illustrates, in the declension of the noun root *ŋókí-* 'dog', the /o/ in the ablative case suffix {-o} and the copulative case suffix {-ko} partially assimilate the final /i/ of *ŋókí-* to /u/.
Other vowel assimilation effects are shown in the case declension of a noun like *ŋʉrá-* 'cane rat', as in Table 8, where the final /a/ of *ŋʉrá-* is assimilated by the dative, genitive, ablative, and copulative case suffixes in their non-final forms.
#### Grammar sketch
Table 8: Vowel assimilation in the declension of *ŋʉrá-* 'cane rat'
Ik vowel assimilation may be partial, as when the word *ŋókí-kᵒ* 'It is a dog' is rendered as *ŋókú-kᵒ*. There, the /i/ at the end of *ŋókí-* 'dog' only moves back in the mouth to become /u/; it does not fully assimilate to become identical to the /o/ in the suffix. But vowel assimilation can also be total, as when *ŋʉrá-ɛ́*'of the cane rat' becomes *ŋʉrɛ́-ɛ́*. In that instance, the /a/ at the end of *ŋʉrá*- becomes fully identical to the vowel in the suffix.
Ik vowel harmony can be regressive as in both prior examples, where a vowel exerts pressure on a preceding one. But it can also be progressive, as in the example of *torí-úƙot-* becoming *torí-íƙot*-, where the /i/ acts ahead on the /u/.
#### **2.4.5 Vowel desyllabification**
When the back-of-the-mouth vowels /ɔ/, /o/, /ʉ/ or /u/ wind up next to another vowel across a morpheme boundary, they may lose their status as the nucleus of a syllable and become the semi-vowel /w/ instead. When such vowel desyllabification occurs, the syllabic 'weight' of the vowel gets transferred to the following vowel in a process called compensatory lengthening. This phonological change is evident in the transitive infinitives of verbs ending in a back vowel. Table 9 depicts how the back vowel at the end of the verb root changes to /w/ and then lengthens the vowel in the transitive suffix {-és}.
Vowel desyllabification also takes place in the case declensions of nouns. Any noun root that ends in a back vowel can have that vowel desyllabified to /w/, with the result that the following case suffix is lengthened. As Table 10 demonstrates, this happens with a noun like *dakú-* 'plant, tree' which ends with the back vowel /u/. In five of the eight cases – accusative, dative, genitive, ablative, copulative –
Table 9: Vowel desyllabification in verbs
the final /u/ of *dakú-* changes to /w/ and then lengthens the case suffix. Note that in the nominative case, the /u/ of *dakú-* is desyllabified but does not lengthen the nominative suffix {-a}. This irregularity is a peculiarity of the nominative case only and is seen in many other noun declensions.
Table 10: Vowel desyllabification in nouns
## **2.5 Vowel harmony**
Ik vowels participate in a phonological system called vowel harmony. This means that the language's sound system seeks vocalic 'harmony' by ensuring that all vowels in a single word belong to the same vowel class. The vowel classes involved are the following: 1) the [+ATR] or 'heavy' vowels /i, e, o, u/ that are made with a larger cavity in the throat, giving them a 'heavier', more resonant sound, and 2) the [-ATR] or 'light' vowels /ɨ, ɛ, ɔ, ʉ/ that are made with a smaller cavity in the throat, giving them a 'lighter', less resonant sound. Where the ninth vowel /a/ fits in with these two classes is a theoretical question that has not been conclusively resolved. However, what is clear is that in Ik, /a/ sometimes behaves
#### Grammar sketch
as a [+ATR] vowel and other times as a [-ATR] vowel. And it certainly is found together with vowels from both classes within a single word. The Ik vowel classes anchored by the low vowel /a/ are depicted in Table 11:
Table 11: Ik vowel classes
Because of vowel harmony, all the vowels in a single word will generally belong to one of the vowel classes shown in Table 11. This is clearly evident in the lexicon where verbs consisting of multiple syllables and morphemes contain either [+ATR] or [-ATR] vowels, but not both. Table 12 shows an opposing set of such verbs. Notice how all the vowels in each word belong to one vowel class.
In some situations though, /a/ blocks vowel harmony from spreading to all the morphemes in a word. For example, when the stative suffix {-án-} falls between a verb with [-ATR] vowels and the intransitive suffix {-òn-}, the /a/ in {-án-} prevents the spread of harmony to the whole word. Table 13 gives a few examples of the harmony-blocking behavior of /a/. Notice how [-ATR] vowels are found to the left of {-án-} (in bold), while the [+ATR] /o/ in {-òn-} comes after it.
Ik has three suffixes which are said to be dominant in that they always spread their [+ATR] value as far as they can within a word. These include the pluractional suffix {-í-}, the middle suffix {-ím-}, and the plurative suffix {-íkó-}, all of which contain the vowel /i/. Unless an /a/ blocks the way, these three suffixes will cause all the vowels in the word they are found in to harmonize to [+ATR]. This dominant behavior is illustrated in Table 14. Notice how the [-ATR] vowels in the first column all become [+ATR] in the third column as a result of the dominance of the suffixes (in bold typeface).
Two other issues surrounding vowel harmony deserve mention. First, when two nouns are joined together to form a compound word (§4.3), vowel harmony does not occur between them. For example, the noun roots *rébè-* 'millet' and *mɛ̀sɛ̀-* 'beer' can be joined into the compound *rébèmɛ̀sɛ̀-* 'millet beer', in which, notice, the vowels belong to two different [ATR] vowel classes. An exception to this rule is when the second noun in the compound begins with the vowel /i/, in which case /i/ harmonizes the last vowel of the first noun, as when *ɲɔ́kɔkɔrɔímà-* 'chick' becomes *ɲɔ́kɔkɔró-ímà-* (where the first noun's /ɔ/ is harmonized to
### Table 12: Vowel harmony in the lexicon
Table 13: Vowel harmony blocking behavior of /a/
Table 14: Ik dominant suffixes
#### Grammar sketch
/o/). Second, many of Ik's clitics take on the [ATR] value of their host word, for example when the anaphoric pronoun *déé* becomes *dɛ́ɛ́*in the phrase *mɔƙɔrɔ́ɛ́=dɛ́ɛ́* 'in that rock pool'. Again, the exception is when the clitic contains /i/, in which case it becomes dominant, harmonizing its host, as when *bárítínʉ́ɔ=díí* 'from those corrals' becomes *bárítínúo=díí* (where the vowels /ʉ́ɔ/ become /úo/).
## **2.6 Tone**
### **2.6.1 Tone inventory**
Ik is a tonal language. In terms of acoustics, this means that every vowel is identified not only by where it is formed in the vocal chamber but also by the pitch with which it is uttered. This further entails that every syllable, morpheme, word, and phrase exhibits a specific and indispensable tone pattern. At a phonological (or psychological) level, Ik has just two tones: high (H) and low (L). All other tones that one hears can be traced back to these two. However, for practical applications like orthography and language learning, four sub-tones must be recognized. These include: high, high-falling, mid, and low. High tone is pronounced with a level, relatively high pitch. High-falling tone falls quickly from relatively high to relatively low pitch, often in the presence of a depressor consonant (see §2.6.4). Mid tone is a level, relatively medium-height pitch, while low tone is either relatively low and flat or tapering off before a pause. Table 15 presents the Ik tones with their names in the first column, pitch profiles in the second, and the orthographic diacritics for writing them in the third (the same diacritics employed throughout the foregoing dictionary sections):
Table 15: Ik tones
### **2.6.2 Lexical tone**
As mentioned above, every word in the Ik lexicon has a tone pattern or 'melody'. That is, Ik words are not identified solely on the basis of consonants and vowels
#### 2 Phonology
(as in non-tonal languages like English) but also on their tone pattern, which must be learned. Since every vowel and therefore every syllable bears a tone, the combination of many syllables in words produces a large inventory of tone patterns. And since the tone pattern of a word is totally unpredictable, language learners must resort to memorizing the pattern with the word. Table 16 gives a sample of the lexical tone patterns found on some short words in Ik:
Table 16: Ik lexical tone patterns
#### **2.6.3 Grammatical tone**
Ik does not have grammatical tone, whereby tone alone can carry out a grammatical function. But tone often accompanies other grammatical signals, thereby reinforcing them. Thus, in that regard, it could be said that Ik has 'semi-grammatical' tone. For example, when the suffix {-íkó-} is used to pluralize a singular noun, the tone over the singular root usually changes, as when *kɔl* 'ram' becomes *kólíkwᵃ*. Similarly, when the venitive suffix {-ét-} is added to a verb stem, it often changes the overall tone pattern, as when *bɛ́ɗɛ́s* 'to want' becomes *bɛɗɛtɛ́s* 'to look for', whereby the tone of the root *bɛ́ɗ-* goes from high to mid. Indeed, many of the suffixes of the language are associated with significant tone changes to the stem. So even if one learns the tonal melodies of nouns and verbs on their own, these melodies may change in particular grammatical contexts. This type of tone changeability is one of the system's more difficult aspects.
The Ik tone system is challenging for foreigners and is not yet fully understood from an analytical point of view. Still, the good news is that with lots of practice, language learners can reasonably expect to develop a certain degree of communicative competency. For the most complete description of the tone system to
#### Grammar sketch
date, the reader is invited to consult §3.2 in *A grammar of Ik (Icé-tód)* (Schrock 2014). That section expands on what has been presented here and includes more detailed discussions of other features of the Ik tone system.
## **2.6.4 Depressor consonants**
In Ik, the voiced consonants /b, d, dz, g, ɦy, j, z, ʒ/ plus /h/ act as depressor consonants. Depressor consonants are so-called because they 'depress' or pull down the pitch of neighboring vowels. In doing so, they act almost as if they had a very low tone of their own. The effect of Ik depressors is so strong that, over time, it led to the creation of a whole new set of lexical tone patterns. For instance, all Ik verbs with a HL pattern in their roots have a depressor as the first consonant after the initial high tone: *dɛ́gɛ̀m-* 'crouch', *gʉ́gʉ̀r-* 'hunched', *íbòt-* 'jump', *kídzìm-* 'descend', and *ts'ágwà-* 'be raw'. This is because, in anticipation of the extra-low pitch of the depressor, the language compensated by putting a high tone before it where there used to be none. As another example, all nouns with the root tone pattern HL have a depressor as the only consonant between two vowels, as in: *dɔ́bà-* 'mud', *ɛ́bà-* 'horn', *édì-* 'name', *nébù-* 'body', and *wídzò-* 'evening'. And when these types of nouns lose their final vowel due to vowel devoicing, that is when the high-falling contour tone comes into play, as in *dɔ̂bᵃ* 'mud', *ɛ̂bᵃ* 'horn', *êdᵃ* 'name', *nêbᵃ* 'body', and *wîdzᵃ* 'evening'.
## **3 Morphology**
## **3.1 Overview**
morphology is the system by which a language grammar makes words. While the preceding chapter introduced meaningful sound units (phonemes), the present chapter describes larger meaningful units called morphemes. Ik exhibits three types of morpheme: word, affix, and clitic. A word is defined as a free morpheme that can meaningfully stand alone. An affix is a bound morpheme that must attach to a word to maintain its integrity. Affixes are indicated in this grammar by a hyphen before (and sometimes after) them, as in {-án-}, the stative adjectival suffix. A clitic is a hybrid: in some constructions it acts like a word standing alone, while in other constructions, it attaches to a word like an affix. Clitics may be marked in this grammar by an equals sign, as in {=kì} 'those'.
Traditionally, languages are described as having word classes, that is, categories of morphemes that have certain characteristics. These classes include the familiar major ones like 'nouns' and 'verbs' but often several others as well. For
the purposes of this grammar sketch, free-standing words and clitics are considered 'words', while affixes are not. In Ik, thirteen word classes are recognized and include the following: nouns, pronouns, demonstratives, quantifiers, numerals, prepositions, verbs, adverbs, ideophones, interjections, nursery words, complementizers, and connectives (or conjunctions). Each of these word classes is briefly introduced in the following subsections, while a full list of Ik affixes can be found later in Appendix A.
## **3.2 Nouns**
Nouns and verbs make up the language's only two open word classes, meaning that they may have new members added to them. Nouns make up roughly 47% of the total Ik lexicon. Noun roots can be short, like *eí-* 'stomach contents', or long like *ɲákaɓɔɓwáátá-* 'finger ring', but they all have at least two syllables in their root form. This structural condition is necessary because some case suffixes delete the last vowel of the noun root when they affix to it. All Ik nouns, without a single exception, end in a vowel in their root forms. Noun roots are represented throughout this book with hyphenated forms, indicating that in actual Ik speech, any noun must have at least a case suffix. In addition to case, nouns may take singulative or plurative suffixes and may be joined with other nouns to make compound nouns. §4 is devoted to expounding on Ik nouns.
## **3.3 Pronouns**
Pronouns form a closed word class, incapable of admitting new members. They 'stand in' for nouns whose specific names need not always be mentioned or repeated. Pronouns make up less than 1% of the Ik lexicon and yet have great grammatical importance. Most Ik pronouns are free, capable of standing on their own, while others are inextricably bound to verbs. They may be personal, capable of specifying grammatical person, or impersonal. Other categories of pronoun include: indefinite, interrogative, demonstrative, relative, and reflexive. §5 is devoted to describing the various kinds of pronouns in Ik.
## **3.4 Demonstratives**
Demonstratives form another closed word class, admitting no new members. They 'demonstrate' nouns by 'pointing them out', referring to them spatially, temporally, or discursively. They too make up less than 1% of the lexicon. Many Ik demonstratives have been analyzed as clitics: They seem sometimes to act
#### Grammar sketch
like separate words, and yet in terms of vowel harmony, they act like suffixes. As clitics, they may be written connected to words in linguistic writing (with =), whereas in non-linguistic writing, they are written separately. For example, the phrase 'these trees' would be written as *dakwítína=ni* in linguistic publications and as *dakwítína ni* elsewhere. Ik has four kinds of demonstrative: spatial, temporal, anaphoric, and locative adverbial – all of which are discussed in §6.
## **3.5 Quantifiers**
As their name implies, qantifiers'quantify' the nouns that precede them. That is, they are separate words that follow nouns and convey the general quantity of the noun in terms of allness, bothness, fewness, or manyness. Specific, numeric quantity is expressed by the numerals, which are the topic of the next subsection. Ik quantifiers sometimes act more like numerals by directly following the noun they modify without an intervening relative pronoun, as in *wika ƙwaɗᵉ* 'few children'. But other times they act more like adjectival verbs by taking a relative pronoun between them and the noun they modify, for example, *wika ni ƙwaɗᵉ* 'children that (are) few'. In the former function as numerals, they have a distinct, perhaps more ancient root, as in *ƙwàɗè*, whereas in their function as adjectival verbs, they have a truncated root in a verbal infinitive, in this case *ƙwàɗ-òn* 'to be few'. The eight known Ik quantifiers are given in Table 17:
Table 17: Ik quantifiers
#### 3 Morphology
## **3.6 Numerals**
Numerals convey the specific number of the noun they modify. Ik has a quinary or 'base-5' counting system, meaning that it has individual words for the numbers 1-5 and then builds numbers 6-9 by adding the appropriate number to 5, as in *tude ńda kiɗi ts'agús* 'five and those four', which is 9. The number 10 is technically not a numeral, but rather, a noun: *toomíní-*. Ik numerals directly follow the noun they modify, without an intervening relative pronoun. Just as the quantifiers *ƙwàɗè* 'few' and *kòmà* 'many' can function as verbs, the numerals 1-5 can also function as verbs. Table 18 presents Ik numerals 1-9:
Table 18: Ik numerals
To form numbers 11-19, Ik builds off the noun *toomíní-* 'ten' and then repeats the quinary system shown in Table 18. For example, the number 17 is expressed as *toomín ńda kiɗi túde ńda kiɗi léɓètsᵉ* 'ten and those five and those two'. Then, after 19, the numbers 20, 30, 40, etc. are based on the compound *toomín-ékù-* 'teneye', as in *toomínékwa léɓètsᵉ* 'ten-eye two', which is 20. The numbers for 100 (*ŋamíáì-*) and 1,000 (*álìfù-*) have both been borrowed from Swahili.
## **3.7 Prepositions**
Prepositions are usually small particles 'pre-posed', that is, put in front of a noun to indicate what its relationship is to another noun or to the wider sentence in which it occurs. Many of the functions that prepositions fulfill in other languages are handled by cases in Ik (see §7). However, Ik still has a very small, closed group of prepositions that somehow have survived the hegemony of case.
#### Grammar sketch
Nonetheless, they interact closely with case as each preposition selects the case that its noun head (or host) must take. Table 19 presents all the known Ik prepositions with their meanings and the cases they require on nouns:
Table 19: Ik prepositions
The following example sentences (1)-(8) offer an opportunity to see the prepositions from Table 19 in a variety of natural language contexts:
## **3.8 Verbs**
Verbs comprise the second of Ik's two large open word classes. Like nouns, Ik verbs make up approximately 48% of the lexicon. Verb roots can be short like *ó-* 'call', long like *gwɛrɛʝɛ́ʝ-* 'be coarse', or reduplicated like *diridír-* 'be sugary' and *ɨpɨrípír-* 'drill'. Verb roots are represented throughout this book with hyphenated forms, indicating that in actual Ik speech, any verb must have at least one suffix. That minimal suffix may be a subject-agreement suffix or a tense-aspectmood (TAM) suffix like an imperative or optative. Ik verb stems can stand alone as an independent, self-contained clause and can have many suffixes strung together, as in *soƙórítiísínàkᵃ* 'we all have clawed' and *zeikááƙotinîdᵉ* 'and they all grew large there'. Among the many suffixes that can derive nouns from verbs or inflect verbs for different meanings, there are: deverbatives, subject-agreement markers, directionals, the dummy pronominal, modals, aspectuals, voice and valency changers, and adjectivals. All these verb-related topics (and others) are treated more fully later on in §8.
## **3.9 Adverbs**
Adverbs make up a catch-all category of words that modify verbs or whole clauses. The sixty-or-so Ik adverbs make up less than 1% of the total lexicon. They include 'manner' adverbs like *hɨíʝɔ́* 'slowly' and *zùkù* 'very', epistemic adverbs like *tsábò* 'apparently' and *tsamʉ* 'of course', and general adverbs like *ɛɗá* 'only'
#### Grammar sketch
and *naɓó* 'again'. Other important categories of adverbs are the tense-marking adverbs, certainty and contingency markers, and the conditional-hypothetical adverbs. All these types of Ik adverbs are discussed later in §9.
## **3.10 Ideophones**
Ideophones form a word class that is characterized by highly expressive words that denote physical phenomena like color, motion, sound, shape, volume, etc. They are often 'sound-symbolic' or onomatopoeic. That means that their very sound as they are pronounced evokes the physical perception they signify. For example, the ideophone *bùlùƙᵘ* means 'the sound something makes when dropping into water', like 'splashǃ' or 'kersplunkǃ' in English. At present, one hundred forty Ik ideophones (1.6% of total) have been recorded, but there are certainly many more in the language. And they are probably continually created. Table 20 offers a sample of the colorful variety of Ik ideophones on record:
Table 20: Ik ideophones
3 Morphology
## **3.11 Interjections**
Like adverbs, interjections form a bit of a catch-all word class. Interjections include any word that expresses emotions or mental states of any kind, usually outside the grammar of a sentence. The roughly thirty Ik interjections that have been recorded make up less than 1% of the total lexicon. Ik interjections may consist of a single word like *aaii* 'ouchǃ' or *wúlù* 'yikesǃ' or a short phrase like *wika ni* 'these kids (I tell you)ǃ' or *tíɔ ʝɔ́ɔ̀* 'there, there (it's okay)ǃ'. Several of the other interjections on record are provided in Table 21:
## **3.12 Nursery words**
Nursery words make up a small class of one-word expressions that act as commands or encouragements to babies or toddlers to do something. The ten Ik nursery words on record are lain out in Table 22 with English glosses.
## **3.13 Complementizers**
Complementizers are words that introduce reported speech or thought. For example, in the English sentence 'She said that she agrees', the word *that* is the complementizer that introduces that reported statement *she agrees*. Ik has only two complementizers. One of them, *tòìmɛ̀nà-* 'that', is technically a noun and thus belongs in the noun word class. But because of its function, it is dealt with here. The word *tòìmɛ̀nà-*, a compound of the verb *tód-* 'speak' and *mɛná-* 'words',
#### Grammar sketch
is used with a variety of speaking and thinking verbs. The second Ik complementizer, *tàà*, is a probably a derivative of the verb *kʉta* '(s)he says' that has been reduced over time. Even now it is usually used after the verb *kʉ̀t-* 'say'. Example (9) shows how *tòìmɛ̀nà-* is used in a sentence to introduce the clause *mɨtída bɔnán* 'you are an orphan'. And example (10) shows the complementizer*tàà* introducing the clause *iya ɲjíníkiʝa kɔ́ɔ́kɛ* 'our land is over there':
## **3.14 Connectives**
Connectives or 'conjunctions' are words whose function is to join together other words, phrases, or clauses. If they are coordinating connectives like *ńdà* 'and', then they join grammatical units of equal status, like a word to a word, or an independent clause to another independent clause. Whereas if they are subordinating connectives like *na* 'if', then they join grammatical units of unequal status, usually a dependent clause to an independent one. Even though their role is to link grammatical units, not all of them come between the units they link. Many come before both, often as the first word in the sentence. Ik has roughly
#### 3 Morphology
eight coordinating connectives and thirty subordinating ones – making up less than 1% of the lexicon. The coordinating connectives are presented in Table 23, while Table 24 offers a sampling of the subordinating connectives:
Table 23: Ik coordinating connectives
The following natural-language examples illustrate three of the more commonly used coordinating connectives: *kèɗè*, *kòtò*, and *ńdà*. In example (11), the connective *kèɗè* 'or' joins two equal constituents, the nouns *Tábayɔɔ* and *Fetíékù*. In (12), the connective *kòtò* 'and, but, then,' links two independent but semantically related clauses, and in (13), the connective *ńdà* 'and' connects two passive clauses:
In contrast to the coordinating connectives shown in Table 23 and examples (11)- (13), *sub*ordinating connectives join units of unequal status, usually a subordinate (dependent) clause to a main one. Table 24 provides a sample of the thirty Ik subordinating connectives, while examples (14)-(16) below illustrate the function of some of these connectives in a few natural-language environments:
Grammar sketch
Table 24: Ik subordinating connectives
In example (14) below, the subordinating connective *ɗɛ̀mʉ̀sʉ̀* 'before, unless, until' introduces a dependent clause that connects semantically to the following independent one. The same grammatical structure is also evident in (15) and (16), where the connectives *mísì* 'if, whether' and *na* 'if, when' set off short dependent clauses that logically lead into main clauses that follow them:
- b. isio what:cop noo pst3 ŋábìàn? wear:plur:ips what was typically worn?'
#### 4 Nouns
## **4 Nouns**
## **4.1 Overview**
Single Ik nouns in a speaker's mental lexicon consist minimally of a root. Roots are words that cannot be analyzed into smaller parts from the perspective of modern Ik. (Historical research may reveal how roots were put together over time, but that is the domain of etymology.) When plucked from the lexicon and put into actual Ik speech, every noun root must receive at least one suffix, which must be a case suffix. Every noun root ends in a vowel, and case suffixes either delete or attach to this final vowel. In addition to case suffixes, an Ik noun may take on a number suffix or may be joined with one or two other nouns to form a compound. Case suffixes are fully explained later in §7, while number suffixes and compounds are covered in the rest of this chapter.
Ik number suffixes include pluratives and singulatives. Many noun roots can be pluralized if they are inherently singular in number. A few others can be singularized because they are inherently plural. In addition to these standard number-markers, Ik also has special possessive number suffixes that combine the notions of number and possession into one suffix. And yet other nouns are mass nouns, naming entities in the world perceived as inherently plural unities (like dust or water). These take no suffixes but are treated grammatically as plurals. Finally, some nouns are transnumeral, construed as singular or plural and given the appropriate singular or plural modifiers, if needed.
Compounding is the primary way Ik acquires or makes new nouns – besides borrowing them from other languages. Compounds in Ik are made by putting two or three nouns together into a new composite word with special emergent characteristics. The first noun describes or specifies the second noun to make an aggregate meaning that is often different than that of the two separate nouns. Compounding and types of compounds are discussed below in §4.3.
Ik nominal suffixes differ individually in how they affix to noun roots. With the exception of five case suffixes, all nominal suffixes first delete the final vowel of the noun to which they attach. This is known as subtractive morphology. The case suffixes that preserve the final vowel are the accusative, dative, genitive, ablative, and oblique. For more on how case suffixes attach to nouns, see §7.
#### Grammar sketch
## **4.2 Number**
### **4.2.1 Pluratives (plur)**
Ik has four ways to show that a noun is plural: three plurative suffixes and suppletive plurals. The three plurative suffixes are: 1) {-íkó-}, 2) {-ítíní-}, and 3) {-ìkà-}. The first plurative suffix, {-íkó-}, is dominant in terms of vowel harmony, meaning it changes the vowels of a [-ATR] noun to [+ATR] unless /a/ intervenes and blocks it. For example, in some instances, the vowel /a/ spontaneously appears between the singular root and the suffix {-íkó-}. (This /a/ is a relic of an ancient singulative suffix \**-at-* that is no longer in use in current Ik.)
The use of {-íkó-} is limited to a small number of nouns (roughly 100); it is not applied to newly borrowed nouns. Table 25 presents several examples of nouns pluralized with this suffix. Note how the suffix harmonizes the vowels of the singular root except where the vowel /a/ blocks the leftward spread of harmony. Notice also that in some cases the suffix alters the tone of the singular root.
The second plurative, {-ítíní-}, is used to pluralize nouns that have only two syllables in their root. Table 26 gives a sample of disyllabic nouns pluralized with {-ítíní-}. Notice that if the singular noun has [-ATR] vowels, then the plurative suffix harmonizes to {-ítíní-}. Unlike the suffix {-íkó-}, {-ítíní-} never alters the tone of the root, though its own tone may conform to the tone of the root.
The third plurative, {-ìkà-}, is used primarily to pluralize nouns with three or more syllables in their lexical root. Table 27 provides a sample of polysyllabic nouns pluralized with {-ìkà-}. Notice that if the singular noun has [-ATR] vowels, then the plurative suffix harmonizes to {-ìkà-}. Like {-íkó-}, {-ìkà-} sometimes alters the tone of the singular noun as well as having its own tone altered.
Secondarily, the plurative {-ìkà-} is used to pluralize a few nouns that have only two syllables in their lexical root. Why these few nouns do not take {-ítíní-} as a plurative instead is not known. A bit of speculation on this point might invoke the notion of mora or the unit of syllable weight. Among the seven examples shown in Table 28, three of them contain the semi-vowel /w/ which may be thought to contain its own mora, as a vowel would. Likewise, two of the examples (*hòò-* and *sédà-*) contain depressor consonants which may also count for one mora. Perhaps in the remaining two (*kíʝá-* and *ríʝá-*), the voiced stop /ʝ/ used to be a depressor consonant. Regardless of the historical explanation, Table 28 presents a few examples of {-ìkà-} being used to pluralize disyllabic nouns.
Table 25: The plurative suffix {-íkó-}
Table 26: The plurative suffix {-ítíní-}
Table 27: The plurative suffix {-ìkà-} with polysyllabic nouns
#### Grammar sketch
Table 28: The plurative suffix {-ìkà-} with disyllabic nouns
### **4.2.2 Suppletive plurals**
Ik also has a handful of singular nouns that cannot be pluralized in a productive way with any of the three suffixes discussed above. Three of these nouns on record are truly suppletive in that their singular and plural forms bear absolutely no resemblance to each other. These are the first three in Table 29. The last three examples in Table 29 represent nouns that are semi-suppletive; even though one can discern a similarity between the singular and plural forms, the way the two forms are derived from each other is not productive in the language:
Table 29: Ik suppletive plurals
## **4.2.3 Singulatives (sing)**
In contrast to pluratives, singulatives convert an inherently plural noun root to a derived singular. Ik has one such suffix that may be considered a true singulative in the contemporary grammar of the modern language, and that is {-àmà-}
#### 4 Nouns
or {-ɔ̀mà-}. Since this singulative is only used with personal entities, it seems likely that it is related etymologically to the word *ámá-* 'person'. Table 30 gives the only four unambiguous examples of when this singulative is used. Note that its tone pattern may be altered by the tone of the plural root:
Table 30: The Ik singulative {-àmà-}
### **4.2.4 Possessive number suffixes (poss)**
In addition to standard pluratives and a singulative, Ik has what may be called possessive number suffixes. These possessive suffixes – {-èdè-} in the singular and {-ìnì-} in the plural – each fuse the notions of number and possession into one morpheme. When they are affixed to a noun, they specify a) the grammatical number of the noun and b) its association with another entity (hence the 'possession'). They do not specify the number of the possessor(s). For example, the word *akedᵃ*, a stem consisting of *aká-* 'den' and {-èdè-} (nominative case) can mean both 'its den' or 'their den'. And the word *akɨn*, consisting of *aká-* 'den' and {-ìnì-} (nominative case), can mean either 'its dens' or 'their dens'.
Within the broad notion of 'possession', the possessive number suffixes {-èdè-} and {-ìnì-} can signify more specific semantic relationships like part-whole, kinship, and association. Table 31 gives some examples of {-èdè-} expressing a partwhole relationship with the unnamed entity. Note how the meanings of the noun roots are extended metaphorically to denote structural parts of things. Note also that the tone of the root may be altered in the presence of {-èdè-}.
The plural possessive suffix {-ìnì-} has two special applications with human possessors. In the first, it is used to pluralize kinship terms, where a kinship association is explicit. In the second, it refers to people associated with a certain person in general terms. Table 32 illustrates both of these nuances, showing the singular root in the first column, and in the second, the root plus {-ìnì-}.
#### Grammar sketch
Table 31: The Ik singular possessive {-èdè-}
Table 32: The Ik plural possessive {-ìnì-}
#### **4.2.5 Mass nouns**
A small group of Ik noun roots are classified as non-count mass nouns. These nouns are inherently, lexically plural. As such, they require plural demonstratives and relative pronouns. This group includes words for powders, liquids, and gases various particulate substances. Table 33 presents seven examples of mass nouns. The roots are in the table's first column, followed in the third column by the noun in a phrase with the plural demonstrative *ni* 'those'. Note that in the English gloss, the equivalent is provided but with a singular interpretation:
4 Nouns
Table 33: Ik non-countable mass nouns
### **4.2.6 Transnumeral nouns**
Another small group of Ik noun roots are inherently transnumeral, meaning that they can be singular or plural depending on the speaker's intention. Whatever number is imputed to them must be reflected in the grammar of the rest of the sentence, for example in subject-agreement on the verb or in any demonstratives or relative pronouns used to modify them. Ik transnumeral nouns cannot be pluralized in any of the ways discussed up to this point. But with the bound nominal morpheme *-icíká-* (see §4.3.4), they can be given a sense of distributiveness or variation. Table 34 presents three examples of Ik transnumeral nouns with their singular, plural, and distributive interpretations:
Table 34: Ik transnumeral nouns
## **4.3 Compounds**
For word-building purposes, Ik relies heavily on compounding, joining two or more nouns together into a new composite word. The first noun (or pronoun) in a compound retains its lexical root form (that is hyphenated throughout this book), including its lexical tone. The last noun in a compound takes whichever case ending the syntactic context calls for. For example, in the compound *riéwíkᵃ* 'goat kids', the first root *rié-* 'goat' keeps its lexical form, while the second, *wicé-* 'children', has been modified by the nominative case suffix {-ᵃ}. If compounding changes the tone of its constituent parts, it will be the first noun that affects the others. In the rare compound with three constituent nouns, the first two stay in their lexical form (not counting tone), while the third is inflected for case, for example in *Icémóríɗókàkà-* 'cowpea leaves', a compound of *Icé-* 'Ik', *mòrìɗò-* 'beans', and *kaká-* 'leaves'. In *Icé-móríɗó-kàkà-*, note that while the last two elements retain their lexical segments, their tone patterns have changed dramatically due to the influence of *Icé-* in spreading H tone throughout the word.
Ik compounds create two kinds of new meaning: 1) a narrower, more specific meaning in which the first noun specifies the second, or 2) a completely novel, unpredictable meaning. An example of the first type would be *bʉbʉnɔ́ɔ́ʝà-* 'ember-wound' or 'bullet wound' where the first noun *bʉbʉná-* 'ember' narrows down the possible references of *ɔ́ʝá-* 'wound' to a wound caused by a bullet. And an example of the second type of compounded meaning would be *óbiʝoets'í-* , a compound that literally means 'rhino urine' but is actually the name of a species of vine (that nonetheless was apparently the favorite urination spot of rhinos). Through both types of meaning-making, Ik compounds add a considerable amount of expressiveness and color to the language's vocabulary.
In addition to the two broader semantic categories of compounds discussed above, five other categories of Ik compounds are recognized. These include the agentive, diminutive, internal, variative, and relational, all discussed in the sections to follow.
## **4.3.1 Agentive (agt)**
1 Ik forms agentive compounds by using the root *ámá-* 'person' (for singular) or *icé-* (for plural) as the last element in a compound. Although the root *Icé-* simply means 'Ik people' when standing on its own, in the agentive construction it denotes plural agents. Here 'agent' is understood broadly as any person or thing that does or is whatever is characterized by the first element in the com-
#### 4 Nouns
pound. The first element may be a noun, as in *dɛá-ámà-* 'messenger', literally 'foot-person', or a verb as in *ŋwàxɔ̀nì-àmà-* 'lame person', literally 'to be lameperson'. Note, however, that even though *ŋwàxɔ̀n* is a verb semantically, it has been deverbalized into a noun by the infinitive suffix {-òn}. Ik agentive compounds can be translated into English in various ways, depending on what is appropriate. Table 35 presents several examples of agentive compounds:
Table 35: Ik agentive compounds
### **4.3.2 Diminutive (dim)**
Ik forms diminutive compounds by using the root *imá-* 'child' (for singular) and *wicé-* 'children' (for plural) as the second element in a compound. In the more literal interpretation, the first element is the animate being (animal or human) of which the second element is the 'child' or 'children', as in *ɗóɗò-ìmà-* 'lamb' or *ɗóɗo-wicé-* 'lambs'. But when the first element is inanimate, the diminutive construction conveys a sense of 'a small X' or 'small Xs', for example *ƙɔfó-ìmà-* 'a small gourd bowl' and *ƙɔfó-wicé-* 'small gourd bowls'. Lastly, the two interpretations can also get blurred, as when an animate being is perceived as smaller than normal but not as the child of anything. This can be seen, for instance, in the compound *ídèmè-ìmà-* 'earthworm', literally 'snake-child'. Table 36 offers several more examples of the diminutive compound. Notice that when the whole construction is pluralized, both elements may get pluralized, as when *ámá-ìmà-* 'someone's child' becomes *roɓa-wicé-* 'someone's (pl.) children':
#### Grammar sketch
Table 36: Ik diminutive compounds
### **4.3.3 Internal (int)**
So-called internal compounds are made with the bound nominal root *aʝíká-* 'among/inside'. When appended to plural noun, this nominal conveys a sense of interiority or internality to the noun. The internal compound, which occurs relatively rarely, is exemplified in Table 37:
Table 37: Ik internal compounds
### **4.3.4 Variative (var)**
So-called variative compounds are made with the bound nominal root *icíká-* 'various (kinds of)'. When appended to a noun – singular or plural – this nominal morpheme communicates a sense of variety or the multiplicity of a type. As a kind of pluralizer itself, *icíká-* is may be called upon to pluralize five kinds of nouns: 1) transnumeral nouns, 2) nouns not usually pluralizeable in the usual sense, 3) inherently plural nouns, 4) already pluralized nouns, and 5) verb infinitives. Table 38 presents one example for each of these five kinds of nouns that the variative bound nominal *icíká-* can be used to pluralize:
4 Nouns
Table 38: Ik variative compounds
#### **4.3.5 Relational**
Ik compounding is also used to create relational nounsthat express the spatial or structural relationship one thing has to another. In this way, Ik metaphorically extends body-part terms to other non-bodily relationships. Table 39 presents some of the Ik body-part terms used metaphorically:
Table 39: Ik body-part terms with extended meanings
So then, in an Ik relational compound, terms like those in Table 39 form the second element in the compound, a position in which they denote the 'part' in a 'whole-part' semantic relationship. Accordingly, the first element in the relational compound represents the 'whole' in the structural relationship. Table 40 displays a handful of such 'whole-part' compounds:
#### Grammar sketch
Table 40: Ik relational compounds
## **5 Pronouns**
## **5.1 Overview**
Pronouns 'stand in' for nouns that are not explicitly mentioned. Most Ik pronouns are free-standing words, but the subject-agreement pronominals and the dummy pronominal are suffixes that are bound to verbs (and so are treated in §8 on verbs). In a sentence, free pronouns are handled just like nouns in that they take case and modifiers. The free pronouns discussed in this section fall into the following nine categories: personal, impersonal possessum, indefinite, interrogative, demonstrative, relative, reflexive, distributive, and cohortative.
## **5.2 Personal pronouns**
Ik personal pronouns represent the various grammatical persons that can be referred to in a sentence. The name is slightly misleading in that the pronouns can also denote nonpersonal, inanimate entities expressed by 'it' and 'they' (when referring to things). The Ik personal pronoun system operates along three axes: person (1, 2, 3), number (sg, pl), and clusivity (exc, inc). The 'first person' refers to 'I' and 'we', the second to 'you', and the third to 'she', 'he', 'it', and 'they'. 'Number' (singular or plural) obviously has to do with whether the entity is one or more than one. And 'clusivity' (exclusive or inclusive) indicates whether the addressee of the speech is *ex*cluded from or *in*cluded in the reference of 'we'. Table 41 presents the seven Ik personal pronouns in their lexical root forms, while Table 42 presents the same but in their full case declension:
Table 41: Ik personal pronouns
Table 42: Case declension of Ik personal pronouns
#### Grammar sketch
## **5.3 Impersonal possessum pronoun (pssm)**
Ik also has a special pronoun whose only function is to represent a possessum, that is, generally, an entity associated with another entity (a possessor). This pronoun has the form *ɛní-* and must be bound to another noun or pronoun as the last element in a compound construction. It is impersonal in that it communicates nothing about the possessor or the possessum except for the relationship of possession itself. The impersonal possessum pronoun can be used in a compound construction with personal pronouns or other nouns. Table 43 shows *ɛní-* in conjunction with all seven personal pronouns. It can also be used with full nouns (including deverbalized verbal infinitives) as the compound's first element. This type of possessive construction is illustrated in Table 44:
Table 43: Ik impersonal possessum with pronouns
Table 44: Ik impersonal possessum with nouns
## **5.4 Indefinite pronouns**
Pronouns that are indefinite stand for other entities but with a certain degree of indefiniteness or vagueness. All but one of the Ik indefinite pronouns are based
#### 5 Pronouns
on the root *kɔní-* 'one' or its plural counterpart *kíní-* 'more than one'. The other one that is not based on these roots is *saí-* 'some more/other', a root that may not actually belong with this set but is included on the basis of its English translation. Table 45 provides a rundown of these Ik indefinite pronouns:
Table 45: Ik indefinite pronouns
## **5.5 Interrogative pronouns**
The role of interrogative pronouns is to query the identity of the entity they represent. As a result, they are used to form questions. All but one of the Ik interrogative pronouns incorporate the ancient northeastern African interrogative particle *\*nd-/nt-*, and the one that does not has the form *ìsì-* 'what'. The seven Ik interrogative pronouns are presented in Table 46. Note that in the table's first column, forms are hyphenated when there is a hypothesis as to their internal morphological composition, which is reflected in the second column:
Table 46: Ik interrogative pronouns
#### Grammar sketch
In the formation of a question, Ik interrogative pronouns fill the same slot as the nouns they are representing. It is common for the interrogative pronoun to be 'fronted': moved for emphasis to the first place in the sentence. When this happens, the interrogative pronoun takes the copulative case (see §7.8), as exemplified in sentences (17)-(18). Both demonstrated word orders are perfectly acceptable. For more on how questions are formed in Ik, please see §10.4.3.
- b. **Isio** what:cop bɛ́ɗîdᵃ? want:2sg 'What do you want?
- b. **Ndaíó** where:cop iâdᵉ? be:3sg:dp 'Where is it?'
## **5.6 Demonstrative pronouns**
Ik also has a set of demonstrative pronouns that referentially 'demonstrate' or point to an entity. They are all based on either the singular form *ɗɨ-* 'this (one)' or the plural form *ɗi-* 'these (ones)' that differ formally only in regard to their vowel (/ɨ/ versus /i/). The Ik demonstrative pronoun system is divided in three categories based on spatial distance from the speaker: 1) proximal, meaning near the speaker, 2) medial, meaning a relatively medium distance from the speaker, and 3) distal, meaning relatively far from the speaker. The medial and distal forms, for both singular and plural, consist of the root *ɗɨ-/ɗi-* preceded by the cliticized distal demonstratives *kɨ* 'that' (derived from *ke*) for singular and *ki* 'those' for plural. The only difference between the medial and distal pronouns is the tone pattern whereby the medial form has a high tone on the last syllable, while the distal form does not. Table 47 presents these pronouns in their six lexical forms, while Table 48 gives their full case declensions. Note that the medial and distal forms are indistinguishable except in the nom, ins, and obl cases:
#### 5 Pronouns
Table 47: Ik demonstrative pronouns
Table 48: Case declensions of the demonstrative pronouns
## **5.7 Relative pronouns (rel)**
The role of relative pronouns is to introduce a relative clause: a clause embedded in a main clause to specify the reference of an entity in the main clause. One of the most fascinating features of the Ik relative pronoun system is that it is tensed. That is, it is able to encode the time period at which the statement contained in the relative clause holds or held true. The five time periods covered by these pronouns are 1) non-past, 2) recent past (earlier today), 3) removed past (yester-, last), 4) remote past (a while ago), and 5) remotest past (long ago).
The Ik relative pronouns are all enclitics based on the proto-demonstratives *na* 'this' and *ni* 'these' (see §6.2 below). Those proto-forms are identical to the non-past relative pronouns *na* 'that/which' and *ni* 'that/which (pl)' shown in Table 49. The remaining tensed relative pronouns are built from the proto-forms with a variety of ancient prefixes and suffixes such as *sɨ-/si-* and *-tso/-tsu*.
As shown in examples (19)-(20), no matter where an Ik relative clause (rc) appears in a sentence, the relative pronoun will introduce it as the first element in
#### Grammar sketch
the clause. The entity in the main clause that the relative clause is modifying – called the common argument – must be the last word before the relative clause. As a clitic, the relative pronoun attaches to the common argument. To learn more about the syntax of relative clauses, please see §10.3.2.
## **5.8 Reflexive pronoun**
Ik has a reflexive pronoun that 'reflects' the impact of a verb back onto the subject of the verb. In other words, with the reflexive, the subject and object of an action are the same entity. The Ik reflexive pronoun has the form *así-* in the singular and *ásíkà-* in the plural, translated as '-self' and '-selves', respectively.
These reflexive pronouns are used extensively to make semi-transitive verbs: verbs falling between transitive and intransitive. For example, while the verb *ídzòn* 'to discharge, emit' is intransitive and the verb *ídzès* 'to discharge, emit, shoot' is transitive, the verb *ídzesa así* 'to shoot across (literally 'to shoot -self')' is 'semi-transitive' because the subject and object of the shooting are the same entity. The full case declensions of the singular and plural reflexive pronouns are given below in Table 50, and example sentences (21)-(22) illustrate both the reflexive and the semi-transitive usages of these special pronouns:
#### 6 Demonstratives
Table 50: Case declensions of the reflexive pronouns
## **6 Demonstratives**
## **6.1 Overview**
Ik's demonstratives grammatically point to a referent. In the case of nominal demonstratives, the referent is an entity named by a noun, whereas adverbial demonstratives point to a scene or situation of some sort, encoded by a whole clause. The Ik nominal demonstratives are all enclitics that come just after their host (the referent), as in *ámá=nà* 'this person'. Because the locative adverbial demonstratives function as adverbs, they tend to come at the end of the clause they are modifying. Unlike demonstrative pronouns (see §5.6), spatial and temporal demonstratives are not nouns and never take case endings.
#### Grammar sketch
## **6.2 Spatial demonstratives (dem)**
Ik's spatial demonstratives locate their referent in physical space in degrees of distance from the speaker. For singular referents, there are three degrees of distance: proximal (near), medial (relatively near/far), and distal (more distant). For plural referents, the language inexplicably only distinguishes between proximal and distal. The singular demonstratives are usually translated into English as 'this' and 'that' and the plural ones as 'these' or 'those'. Table 51 below presents the whole set of spatial nominal demonstratives. Notice that in their final forms (ff), their final vowels may be whispered or omitted altogether:
Table 51: Ik spatial demonstratives
Spatial demonstratives usually directly follow their referent, as in (23)-(24):
- cut:sps then forest=dem.sg.dist
'And then that forest over there was cut down.'
## **6.3 Temporal demonstratives (dem.pst)**
The temporal demonstratives, by contrast, locate their referent in five periods of time: non-past (present and future), recent past (earlier today), removed past (yester-, last), remote past (a while ago before yesterday), and remotest past (long ago). Ik has both singular and plural temporal nominal demonstratives, and these are listed below in Table 52. These temporal demonstratives are usually translated into English as 'this' and 'that' in the singular, and 'these' and 'those' in the plural, but with a sense of time rather than physical location. Recall from Table 49 that Ik's relative pronouns are identical in form to the temporal
#### 6 Demonstratives
demonstratives in Table 52, except that because relative pronouns never occur before a pause, they lack the final forms (ff) of those in Table 52:
Table 52: Ik temporal demonstratives
Just like spatial demonstratives, temporal demonstratives directly follow the noun they refer to, as example sentences (25)-(26) illustrate:
## **6.4 Anaphoric demonstratives (anaph)**
The anaphoric demonstratives locate their referent not in space or time *per se* but rather in *shared communicative context*. In other words, they point back to a referent that has either been mentioned already in the same discourse or is already known by both speaker and hearer by some other means. Ik has a singular and a plural anaphoric demonstrative which are enclitics that have the same form in both non-final and final environments (i.e., their final vowels are not omitted). These invariant anaphoric demonstratives, translated into English as 'that' in the singular and 'those' in the plural, are presented in Table 53:
Table 53: Ik anaphoric demonstratives
#### Grammar sketch
Ik anaphoric demonstratives also directly follow their referents, as in (27)-(28):
## **6.5 Adverbial demonstratives**
### **6.5.1 Overview**
Besides the three types of nominal demonstratives described above, Ik also has a set of adverbial demonstratives that involve both locative and anaphoric locative reference. Unlike the nominal demonstratives, the adverbial demonstratives are technically nouns themselves in that they are marked for case and can take their own nominal demonstratives. Their function, however, is adverbial.
## **6.5.2 Locative adverbial demonstratives**
The first type of adverbial demonstrative, the locative adverbial demonstrative, locates the state or event expressed in a clause in physical space. Ik has three sets of such demonstratives. As shown in Table 54, Sets 1 and 2 are built on degree of distance, while Set 3, in addition to degree of distance, is also split into singular and plural. These demonstratives are usually translated into English as 'here', 'there', and 'over there', depending on relative distance:
Examples (29)-(30) illustrate the locative adverbial demonstratives:
#### 6 Demonstratives
Table 54: Ik locative adverbial demonstratives
#### **6.5.3 Anaphoric locative demonstratives**
The second type of Ik adverbial demonstrative are called the anaphoric locatives, which are nouns with a demonstrative function. Like the locative nominal demonstratives, these demonstratives point to a specific place – or metaphorically, a specific time – while also signifying anaphorically that that place or time is already known, either from earlier in the discourse or for some other reason. Ik has two such demonstratives with roughly the same meaning: *ts'ɛ́dɛ́-* and *tʉmɛdɛ́-*, both of which are typically translated as 'there' or more rarely 'then'. Because these words are technically nouns, Table 55 presents them in a full case declension, while examples (31)-(32) illustrate them in sentences.
Table 55: Case declension of anaphoric locative demonstratives
#### Grammar sketch
## **7 Case**
## **7.1 Overview**
Ik has a case system. This means that every noun has a special marking to show what role it has in the sentence. The language marks this role by means of a set of case suffixes (endings). Four of the cases are marked with suffixes consisting of a single vowel, while for three others, the suffix consists of /k/ plus a vowel. An eighth case, the oblique, is marked by the absence of any suffix. In the following examples, (33)-(40), notice how the word *ŋókí-* 'dog' at the end of each sentence has a different ending depending on the case for which it is marked:
7 Case
Eight examples are given above because Ik has eight cases: nominative, accusative, dative, genitive, ablative, instrumental, copulative, and oblique. Table 56 presents the non-final and final forms of the suffixes that mark all eight of these cases. Keep in mind that the null symbol <Ø> signifies either 1) that the case suffix is inaudible or, for the oblique case, 2) that there is no case suffix:
Table 56: Ik case suffixes
From Table 56, there may appear to be significant ambiguity in the Ik case system. For instance, the non-final forms of the nominative and accusative suffixes, the dative and genitive suffixes, and the ablative, instrumental, and copulative suffixes all look the same, respectively. In most cases, the key to disambiguating the suffixes is called 'subtractive' morphology. Two of the Ik case suffixes (namely nominative and instrumental) are subtractive in that they subtract or delete the final vowel of the noun to which they attach. So, for example, while the non-final forms of the nominative and accusative are identical, their morphological behavior is not: the nominative {-a} subtracts the noun's final vowel,
#### Grammar sketch
as when *ŋókí-* 'dog' becomes *ŋók-á* 'dog:nom'; by contrast, the accusative suffix is non-subtractive, as in *ŋókí-à* 'dog:acc'. Other case ambiguities like genitive versus dative and ablative versus copulative, in their non-final forms, can be resolved in the context of the sentence. Different verbs require different cases.
Since every Ik noun ends in a vowel, and since that vowel can be any of the nine (/i, ɨ, e, ɛ, a, ɔ, o, ʉ, u/), the collision of nouns and case suffixes gives rise to all kinds of vowel assimilation (see §2.4.4). The next two tables present declensions of two nouns illustrating vowel assimilation. Table 57 shows the noun *fetí-* 'sun' declined for all eight cases. In particular, notice how the vowel /o/ in the ablative and copulative suffixes partially assimilate the /i/ in *fetí-* to become /u/:
Table 57: Case declension of *fetí-* 'sun'
While Table 57 shows partial vowel assimilation caused by case suffixation, Table 58 reveals an instance of total assimilation. In this table, the noun *kíʝá-* 'land' is declined for all eight cases. Note how the final /a/ of *kíʝá-* becomes totally assimilated by the non-final dative, genitive, ablative, and copulative suffixes.
## **7.2 Nominative (nom)**
The nominative case, marked by the suffix {-a}, is the 'naming' case, whose role is to: 1) mark the subject of main clauses, 2) mark the subject of sequential clauses (see §8.10.7), and 3) mark the direct object of clauses with 1st and 2nd person subjects ('I', 'we', 'you'). Three examples ((41)-(43)) are provided below, each one illustrating one of the three grammatical roles of the nominative case. The third example contains seven sentences that show how Ik object marking is split: objects after 3-person subjects ((s)he/it, they) take the accusative case, while 1- or 2-person subjects (you, we) take objects in the nominative case:
Table 58: Case declension of *kíʝá-* 'land'
Subject of a main clause
(41) Atsáá come:prf lɔŋɔ́t-**ᵃ**! enemies-nom 'The enemies have come!'
Subject of a sequential clause
(42) Toɓuo spear:seq ƙaƙaam-**a** hunter-nom kʉláɓákᵃ. bushbuck:acc 'And the hunter speared the bushbuck.'
Object of a clause with a 1/2-person subject
- b. Ŋƙída eat:2sg tɔbɔŋ-**a**=na. mush-nom=this 'You eat this meal mush.'
- c. Ŋƙa eat:3sg tɔbɔŋɔ́-á=na. mush-acc=this 'She eats this meal mush.'
- d. Ŋƙímá eat:1pl.exc tɔbɔŋ-**a**=na mush-nom=this 'We eat this meal mush.'
#### Grammar sketch
## **7.3 Accusative case (acc)**
The accusative case, marked by the suffix {-ka}, is also split with regard to its basic function. One of its basic functions, that for which it is named, is to mark the direct object of any clause with a 3-person subject ((s)he/it, they). Its other common function is to mark the subject *and* any object of several kinds of subordinate clauses (including relative and temporal clauses). Each of these functions is exemplified by one of the sentences in examples (44)-(47). In the first example, a sentence with a 1-person subject is also given to show the contrast:
Direct object of a clause with a 3-person subject
Subject and object of a subordinate clause
#### 7 Case
## **7.4 Dative (dat)**
The dative case, marked by the suffix {-ke}, is the 'to' or 'in' case, whose role is to mark indirect objects (also called 'extended' or 'secondary' objects). These indirect objects may encode semantic notions like destination, location, recipient, experiencer, possession, and purpose. These are illustrated in examples (48)-(53):
Destination
(48) Ƙeesíá go:fut:1sg awá-**kᵉ**. home-dat 'I'm going home.'
Location
(49) Ia be:3sg sédà-**kᵉ**. garden-dat 'She's in the garden.'
Recipient
(50) Tɔkɔráta divide:3pl kabasáá flour:acc ròɓà-**kᵉ**. people-dat 'They are dividing out flour to people.'
#### Experiencer
(51) Ɨɓálá appall:3sg ɲcì-**è** I-dat zùkᵘ. very 'It really appalls me.' (Lit: 'It is very appalling to me.')
### Possession
(52) Ia be:3sg ɦyɔa cattle:nom ntsí-**kᵉ**. he-dat 'He has cattle.' (Lit: 'There are cattle to him.')
#### Grammar sketch
### Purpose
(53) Ƙaa go:3sg ɲera girls:nom dakúáƙɔ̀-**kᵋ**. wood:inside-dat 'The girls go for firewood.'
## **7.5 Genitive (gen)**
The genitive case, marked by the suffix {-e}, is the 'of' case, whose role is to encode a possessive or associative relationship a noun has with another noun (or, in rare cases, with a verb). Within the broad notions of possession and association are finer nuances such as: ownership, part-whole relationship, kinship, and attribution. These nuances are illustrated in examples (54)-(57):
#### Ownership
(54) Hɔ́nɨnɨ drive:seq ɦyɔa cattle:acc ńtí-**e** they-gen ɓórékᵉ. corral:dat 'And they drove their cattle to the corral.'
Part-whole relationship
(55) Wasá stand:3sg dɛɛdɛɛ foot:dat kwará-**ᵉ**. mountain-gen 'He's standing at the foot of the mountain.'
#### Kinship
(56) Míná love:3sg cekíá wife:acc ntsí-**é** he-gen zùkᵘ. very 'He loves his wife very much.'
#### Attribution
(57) Maráŋá good:3sg muceá way:nom bì-Ø. you-gen 'Your luck is good.' (lit: Your way is good.)
The genitive case has two further roles. One is the nominalization of clauses, that is, the process by which a whole clause is changed into a noun phrase that can be used as a subject or object in another clause. For example, the clause *Cɛɨƙɔta náa eakwa ídèmèkᵃ* 'The man killed the snake' can be compressed into
the nominalized *cɛɛ́sʉ́ƙɔta eakwéé ídèmè* 'the killing of the man of the snake' or 'the man's killing of the snake'. The other secondary role of the genitive has to do with verb *ƙámón* 'to be like'. For unknown historical reasons, this particular verb requires genitive case marking on its complement, as in *Ƙámá ròɓèè mùɲ* 'He's like all people', where *ròɓè-è* is analyzed as 'people-gen' or 'of people'.
## **7.6 Ablative (abl)**
The ablative case, marked by the suffix {-o}, is the 'from' case (or in some situations 'at' or 'in'), whose function is to mark objects with the following semantic roles: origin/source, cause, stimulus, source of judgment, location of activity (versus static location, which is covered by the dative case). Each of these semantic roles of the ablative are illustrated among example sentences (58)-(62):
Origin/source
(58) Atsía come:1sg awá-**ᵒ**. home-abl 'I come from home.'
Cause
(59) Baduƙota=noo die:3sg=pst ɲɛ́ƙɛ̀-**ᵓ**. hunger-abl 'He died from hunger.'
Stimulus
(60) Xɛɓa fear:3sg ɲérà-**ᵒ**. girls-abl 'He's shy of girls.'
Source of judgment
(61) Daa nice:3sg ɲ́cù-<sup>Ø</sup>. I-abl 'It's nice to me.'
Location of activity
(62) Cɛmáta fight:3pl sédìkà-**ᵒ**. gardens-abl 'They are fighting in the gardens.'
#### 7 Case
#### Grammar sketch
## **7.7 Instrumental (ins)**
The instrumental case, marked by the suffix {-o}, is the 'by' or 'with' case. Unlike the ablative suffix {-o}, the instrumental suffix is subtractive, meaning that it first deletes the noun's final vowel. The function of the instrumental case is to mark secondary objects with such semantic roles as instrument/means, pathway, accompaniment, manner, time, and occupation. Each of these nuances are illustrated by one sentence each in example sentences (63)-(68):
Instrument/means
(63) Toɓíá=noo spear:1sg=pst gasoa warthog:nom ɓɨs-**ᵓ**. spear-ins 'I speared a warthog with a spear.'
Pathway
(64) Ƙaini go:3pl fots-**o** ravine-ins gígìròkᵉ. downside:dat 'And they went down by way of the ravine.'
#### Accompaniment
(65) Atsímá=naa come:1pl=pst kúrúɓád-**o** things-ins ŋgóᵉ. we:gen 'We came with our things.'
Manner
(66) Ráʝétuo answer:3sg ɲcie I:dat gáánàs-**ᵓ**. badness-ins 'And he answered me with hostility.'
Time
(67) Bɨraa lack:3sg ɲɛƙa hunger:nom ódoicik-**ó**=ni. days-ins=these 'There is no hunger these days.'
Occupation
(68) Cɛma fight:3sg fítés-**o** washing-ins ƙwázìkàᵉ. clothes:gen 'She's washing clothes.' (lit: 'She is fighting with the washing of clothes.')
7 Case
## **7.8 Copulative (cop)**
The copulative case, marked by the suffix {-ko}, is the 'is' or 'coupling' case, whose function is to link one noun to another in a relationship of exact identity. In this function, the copulative marks three kinds of nouns: 1) a focused (fronted) noun, 2) the complement of a verbless copula (linking verb) clause, and 3) the complement of a negative copula of identity clause. These different uses of the copulative are illustrated in examples sentences (69)-(73):
Fronted subject
(69) Ŋgó-**ó**=naa we-cop=pst wetím. drink:1pl.exc 'It was we (who) drank (it).'
Fronted object
(70) Emó-**ó** meat-cop bɛ́ɗí. want:1sg 'It is meat (that) I want.'
Fronted secondary object
(71) Ɲɛƙɔ-**ɔ** hunger-cop ƙaiátèè go:plur:3pl ƙàƙààƙɔ̀kᵋ. hunt:inside:dat 'It is (due to) hunger (that) they keep going hunting.'
Verbless copula complement
- b. Ámá-**kᵒ**. person-cop 'It's a person.'
Negative copula complement
(73) Bɛna=náá not.be:3sg=pst ɲ́cù-**kᵒ**. I-cop 'It was not me!'
#### Grammar sketch
## **7.9 Oblique (obl)**
The obliqe case, marked by the absence of any suffix, is the 'leftover' case. As such, it is employed to mark nouns in a variety of disparate grammatical roles and functions. Among these are the following: 1) The subject and/or object of an imperative clause, 2) the subject and/or object of an optative clause, 3) the object of a preposition, and 4) a vocative noun (used when calling someone). Each of these uses of the oblique case are demonstrated in examples (74)-(78):
Subject and/or object of an imperative clause
(74) Deté bring:imp bi you:obl cue=dííǃ water:obl=those 'You bring that water!'
Subject and/or object of an optative clause
(75) Ɲ́ci I:obl nesíbine listen:1sg:opt emuti story:obl ntsí. he:gen 'Let me listen to her story.'
Object of a preposition
### Vocative
(78) Éé hey wice, children:obl atsúǃ come:imp 'Hey children, come!'
## **8 Verbs**
## **8.1 Overview**
Ik verbs consist of a verbal root (written in this book with a hyphen, as in *wèt-* 'drink') and at least one of a variety of available derivational and inflectional
#### 8 Verbs
suffixes. The language has no prefixes except those borrowed centuries ago that no longer have any active function, for example the /a/ in *ábʉ̀bʉ̀ƙ-* 'bubble' or the /i/ in *iɓóɓór-* 'hollow out'. Reduplicating a verb root, partially or totally, has long been a strategy for creating a sense of continuousness or repetitiveness, as when *ɨtsán-* 'disturb' becomes *ɨtsanítsán-* 'torment relentlessly'.
Ik employs a large number of suffixes to create longer verb stems. Among these are the infinitive and other deverbalizing suffixes that change a verb into a noun that can take case endings, demonstratives, relative clauses, etc. One very key verb-building strategy of Ik is the directional suffixes that signify the direction of the verb's movement to or away from the speaker. These two directionals have also been extended metaphorically to express the beginning or completion of actions or processes. Another set of verbal suffixes deal with voice and valency, that is, the number of objects the verb requires. Among these are the passive, impersonal passive, middle, causative, and reciprocal.
Once a verb is taken from the mental lexicon and used in speech, it often requires subject-agreement marking, which Ik accomplishes through pronominal suffixes. Ik also has a special verbal suffix, the dummy pronoun, that goes on the verb whenever a peripheral argument, like a place or time designation, has been moved to the front of the clause or removed entirely.
The Ik verbal system has a variety of verbal paradigms based on mood and aspect. The basic distinction in mood is between realis and irrealis, or things that have happened and things that have not. Other modal distinctions include the optative, subjunctive, imperative, and negative. As for aspect, the specification of the internal structure of a verb – complete or incomplete – Ik has suffixes that mark present perfect, intentional-imperfective, pluractional, seqential, and simultaneous. Lastly, Ik exhibits a special set of adjectival suffixes to cover the language's need to express adjectival concepts.
## **8.2 Infinitives (inf)**
#### **8.2.1 Intransitive**
intransitive verbs allow only a subject and possibly an indirect object – a direct object does not figure into their semantic schema. The Ik intransitive infinitive suffix is {-ònì-}. It converts an intransitive verb to a morphological noun that can be used as a noun in a noun phrase. The infinitive is the citation form of a verb, the form one cites in a dictionary or in isolation from other words. Table 59 gives a few examples of intransitive infinitives from the lexicon:
#### Grammar sketch
Table 59: Ik intransitive infinitives
Because the infinitive is a noun morphologically, it can be fully declined for case as all nouns can. Table 60 gives the case declension of the verb *wàtònì-* 'to rain', which shows some vowel assimilation effects on [+ATR] vowels, as when /io/ becomes /uo/ in the ablative and copulative cases. Table 61 does the same for the [-ATR] verb *wɛ́dɔ̀nì-* 'to detour'. Note /ɨɔ/ becoming /ʉɔ/ there as well:
Table 60: Case declension of *wàtònì-* 'to rain'
#### **8.2.2 Transitive**
transitive verbs are those that admit a subject *and* a direct object into their schematic of an active event. The Ik transitive infinitive suffix is {-ésí-}. It converts a transitive verb to a morphological noun that can be used as a noun in a noun phrase. Table 62 presents a few examples of transitive infinitives:
Table 61: Case declension of *wɛ́dɔ̀nì-* 'to detour'
Table 62: Ik transitive infinitives
Table 63 gives the case declension of the deverbalized noun *wetésí-* 'to drink', which shows vowel assimilation effects on [+ATR] vowels. Table 64 does the same for the [-ATR] verb for the [-ATR] verb *wɛts'ɛ́sí-* 'to knap'.
#### **8.2.3 Semi-transitive**
semi-transitive verbs fall between transitive and intransitive in that they take an object, but the object is the reflexive pronoun *así-* '-self', referring to the subject. This means that semi-transitive verbs are morphologically transitive but almost intransitive semantically. Another name for this is 'middle' (although see another Ik middle verb in §8.6.3). Table 65 provides a sample of semi-transitive verbs. No case declension is given for these because they decline the same way as the transitive infinitives shown in Table 63 and Table 64:
Table 63: Case declension of *wetésí-* 'to drink'
Table 64: Case declension of *wɛts'ɛ́sí-* 'to knap'
Table 65: Ik semi-transitive infinitives
#### 8 Verbs
## **8.3 Deverbalizers**
### **8.3.1 Abstractive (abst)**
The abstractive suffix {-ásí-} can be used to replace the intransitive suffix {-ònì-} for converting an intransitive verb to an abstract noun, for example, when *hábòn* 'to be hot' becomes *hábàs* 'heat'. Table 66 gives several examples of abstract nouns derived from intransitive verbs:
Because verbs deverbalized by the abstractive suffix are morphological nouns, they are fully declined for case. Table 67 gives one such case declension of the abstract noun *kuɗásí-* 'shortness':
Table 67: Case declension of *kuɗásí-* 'shortness'
#### Grammar sketch
### **8.3.2 Behaviorative (bhvr)**
The behaviorative suffix {-nànèsì-} is a complex suffix possibly consisting of the stative suffix {-án-} from §8.11.4 and the transitive suffix (§8.2.2) or the abstractive suffix (§8.3.1). Regardless of its composition, the suffix as a whole creates abstract concepts based on simple nouns, like *ámánànès* 'personhood' or 'personality' from *ámá-* 'person'. Table 68 provides a few more examples:
Table 68: Ik behaviorative abstract nouns
Because behavioratives are nouns morphologically, they are declined for case. Table 69 gives the case declension for the word *eakwánánèsì*- 'manhood':
8 Verbs
#### **8.3.3 Patientive (pat)**
The patientive suffix {-amá-} converts a verb to a noun that is characterized by the meaning of the verb. It is called 'patientive' because the derived noun usually has the meaning of 'patient' or object of the original verb, as when *meetés* 'to give' produces *meetam* 'gift'. Table 70 gives some examples of patientive nouns:
Table 70: Ik patientive nouns
Because patientives are nouns morphologically, they are fully declined for case. Table 71 gives the full declension of the noun *wetamá-* 'drink(able)':
#### Grammar sketch
## **8.4 Directionals**
## **8.4.1 Venitive (ven)**
The venitive suffix {-ét-} denotes a direction *toward* a deictic center, usually (but not always) the speaker. It can be translated variously as 'here', 'this way', 'out', or 'up', but it is the Middle English word 'hither' that captures its essence nicely. The venitive suffix comes between the verb root and the infinitive suffix, whether intransitive or transitive. It can be used to augment any verb whose meaning includes motion or movement of any kind. Table 72 gives a few examples:
Venitive infinitives are morphological nouns and thus are declined for case. See §8.2.1 and §8.2.2 for case declensions that show the relevant endings.
### **8.4.2 Andative (and)**
The andative suffix {-uƙot(í)-} denotes direction *away from* a deictic center, usually the speaker (but not always). It can be translated variously as 'away', 'off', 'out', 'that way', or 'there', but it is the Middle English word 'thither' that captures its essence nicely. Unlike the venitive suffix, the andative comes after both the verbal root and the infinitive suffix (in an infinitival construction). It can be used to augment any verb whose meaning includes motion or movement of any kind. Table 73 provides a few examples of andative verbs.
Because the andative suffix comes after infinitive suffixes, whenever an andative infinitive is declined for case, it is the andative suffix that takes case endings. Table 74 gives a declension of the [+ATR] andative verb *séɓésuƙotí-* 'to sweep off', while Table 75 does the same for the [-ATR] verb *sɛkɛ́sʉ́ƙɔtí-* 'to scrub off':
Table 73: Ik andative verbs
Table 74: Case declension of *séɓésuƙotí-* 'to sweep off'
Table 75: Case declension of *sɛkɛ́sʉ́ƙɔtí-* 'to scrub off'
#### Grammar sketch
## **8.5 Aspectuals**
## **8.5.1 Inchoative (inch)**
The inchoative suffix {-ét-} is identical to the venitive suffix described in §8.4.1, and this is because its meaning is a metaphorical extension of the meaning of the venitive. That is, the venitive meaning of 'hither' was extended to mean the beginning of a state or activity (for intransitives) or the starting up of some action or process (for transitives). The inchoative behaves morphologically (including case declensions) exactly the same as the venitive. Table 76 gives a few examples of intransitive and transitive verbs in the inchoative aspect:
### **8.5.2 Completive (comp)**
The completive suffix {-uƙot(í)-} is identical to the andative suffix described in §8.4.2, and this is because its meaning is a metaphorical extension of the meaning of the andative. That is, the andative meaning of 'thither' was extended to mean the completion of a change of state or activity (for intransitives) or the fulfillment of some action or process (for transitives). The completive behaves morphologically (including case declensions) exactly the same as the andative. Table 77 gives a few examples of lexical verbs in the completive aspect:
#### **8.5.3 Pluractional (plur)**
The pluractional suffix {-í-} denotes an action or state that is construed as inherently *plural* in its realization. This notion of plurality can mean any of the following: 1) an intransitive action done more than once or done by more than
8 Verbs
Table 77: Ik completive verbs
one subject, 2) a state attributed more than once or of more than one subject, 3) a transitive action done more than once, done by more than one subject, or done to more than one object. In short, the pluractional suffix conveys the idea that the application of the verb is multiple. The pluractional suffix comes just before the infinitive suffix and is a dominant [+ATR] suffix that harmonizes [-ATR] vowels. Table 78 gives a few examples of intransitive and transitive pluractional verbs:
Table 78: Ik pluractional verbs
## **8.6 Voice and valence**
### **8.6.1 Passive (pass)**
The Ik passive suffix {-ósí-} has the unusual distinction of being able to modify both intransitive and transitive verbs. With intransitive verbs, it adds the nuance of characteristicness to the meaning of the verb, often with the help of root reduplication. With transitive verbs, it has the usual function of a passive, which is
#### Grammar sketch
to convert the object of a transitive verb into the subject of an intransitive verb. Table 79 gives examples of both intransitive and transitive passives:
Table 79: Ik passives
Another quirky feature of the Ik passive {-ósí-} is that it can function both as a passive infinitive suffix (taking case) and as a regular inflectional suffix followed by subject-agreement pronouns. When it is declined for case, it declines just like the transitive suffix {-ésí-} in §8.2.2. Example (79) below illustrates this in a sentence where the passive infinitive *búdòsì-* 'to be hidden' gets the accusative case. Then, example (80) shows the same passive acting as a verb proper, taking the 3pl subject-agreement pronominal suffix {-át-}:
## **8.6.2 Impersonal passive (ips)**
The impersonal passive suffix {-àn-} behaves like a typical passive in that it eliminates the agent of a transitive verb and promotes the object to subject. However, unlike the passive {-ósí-} described above, the impersonal passive cannot be specified for the person or number of its subject. Instead, it remains marked for 3sg regardless of who or what the subject may be. Another strange property of {-àn-} is that it can be used with intransitive verbs as well (just like the passive). When
8 Verbs
used with intransitive verbs, it has the function of downplaying the identity of the subject. For this reason, it can often be translated as 'People …' or 'One …', as in *Tódian* 'People say (it)'. The impersonal passive is a grammatical morpheme not listed in the lexicon, and so it must be illustrated in examples like (81)-(82):
## **8.6.3 Middle (mid)**
Ik has two middle suffixes: {-m-} and {-ím-}. Like the semi-transitive construction discussed in §8.2.3, the middle suffixes convert simple transitive verbs into something in the 'middle' of transitive and intransitive. That is, the Ik middle verbs convey that idea that if an action is done to an entity, it is the entity itself – if anything – doing it to itself alone, apart from any other explicit agent. The middles eliminate the agent and promote the patient to subject.
The middle suffix {-m-} always has a vowel between it and the preceding verb root. This vowel is usually a copy of the root vowel, as when *ɗusés* 'cut' becomes *ɗusúmón* 'to cut (alone/on its own)', but it can also have a non-copy vowel as in *bokímón* 'to get caught'. For its part, the middle suffix {-ím-} – a dominant [+ATR] suffix – is always paired with the inchoative suffix {-ét-}, thereby forming the complex morpheme {-ímét-}. Table 80 below gives some examples of these two suffixes converting transitive verbs to middle verbs.
#### **8.6.4 Reciprocal (recip)**
The reciprocal suffix {-ínósí-} denotes a reciprocal relationship that a verb's subject has with itself. That is, the reciprocal collapses the subject and direct object of a transitive verb, or the subject and a secondary object of an intransitive verb, into just the subject of a reciprocal verb. In this regard, it is similar to the semi-transitive verbs from §8.2.3 that use the reflexive pronoun *así-* '-self'. Table 81 provides a few examples of reciprocals derived from other verbs:
#### Grammar sketch
Table 80: Ik middle verbs
Like the passive {-ósí-} discussed in §8.6.1, the reciprocal suffix can take either case endings (as a morphological noun) or subject-agreement endings (as a morphological verb). A case declension of *ínínósí-* 'to cohabitate' is shown in Table 82, and in example (83) below, the reciprocal verb *ɨɓákínɔ́sí-* 'to be next to each other' gets the accusative case. Then, example (84) shows the same verb acting as a verb proper, with the 3pl subject-agreement marker {-át-}:
Table 82: Case declension of *ínínósí-* 'to cohabitate'
### **8.6.5 Causative (caus)**
Ik expresses causativity with a morphological causative, the causative suffix {-ìt- }. When this suffix is added to a verb with meaning X, it changes the meaning of the verb to 'cause/make (to) X'. This suffix can be used to causativize intransitive and transitive verbs and comes right after the verb root, before the infinitive marker (if present) and any other suffixes like an inchoative or pluractional. If the last vowel of the verb root is /u/, the causative may be assimilated to the form {-ùt-}. Table 83 gives several examples of causativized verbs.
## **8.7 Subject-agreement**
Whenever Ik grammar requires verbs to agree with their subjects, one of the seven pronominal suffixes in Table 84 are used. Note that if the verb contains [- ATR] vowels, these suffixes will also be harmonized to [-ATR]. Just like the free pronouns described back in §5.2, these bound pronominal suffixes are organized along three axes: 1) person (1/2/3), 2) number (singular/plural), and 3) clusivity
#### Grammar sketch
Table 83: Ik causative verbs
(exclusive/inclusive). The form these pronominals ultimately take depends on the grammatical mood of the verb to which they attach. If the verb is in the irrealis mood (see §8.9.1), the suffixes appear with their underlying forms. Whereas if they are in the realis mood (see §8.9.2), the realis suffix {-a} first subtracts or deletes their final vowel. The difference in the two mood-based paradigms is depicted in Table 84. To see instances of the Ik subject-agreement suffixes in actual language use, you may refer back to example (43) in §7.2.
#### 8 Verbs
## **8.8 Dummy pronoun (dp)**
Ik has a special verbal affix called the dummy pronoun because it represents a secondary (indirect) object that has been (re)moved. That is, the dummy pronoun is a form of object-marking on the verb, but not of direct object marking. For example, if an indirect object expressing location or time or means is moved to the front of a clause for emphasis, it leaves a trace on the verb in the form of the dummy pronoun. Seen from another perspective, the dummy pronoun is always a clue that there is a missing syntactic constituent in the clause.
The dummy pronoun has the form {-ˊdè} and is very volatile in terms of allomorphy, dramatically changing its form in different morpho-phonological environments. Once the /d/ is lost in non-final forms, vowel assimilation and vowel harmony so distort the dummy pronoun as to make it almost unrecognizable at times. Table 85 below is given to illustrate its diverse allomorphy:
Examples (85)-(86) illustrate the dummy pronoun in two different morphological forms: final and non-final. Note that the tones associated with the pronoun in these examples do not match what is shown in Table 85; this is because of local tonal interference. In terms of function, the dummy pronoun in (85) indicates that an indirect object – the destination of the verb *ƙáátà* 'they go (went)' – has been displaced from its usual spot after the verb to a place of focus at the beginning of the sentence (*Ntsúó*). Then in (86), the dummy pronoun marks an indirect object – the location of staying – that is missing from the clause entirely. Since this sentence was taken out of context from a story, most likely the missing object had been already mentioned earlier in the discourse:
#### Grammar sketch
## **8.9 Mood**
## **8.9.1 Irrealis (irr)**
A basic distinction in grammatical mood cleaves Ik verbal aspects and modalities right down the center, and this distinction is between irrealis and realis. As it applies specifically to Ik, the irrealis mood includes states and events whose *actuality* or *reality* are not expressly encoded in the grammar. Another way of saying this is that irrealis verbs in Ik can convey anything *but* whether a state or event has happened, is happening, or will happen. The morphological manifestation of the irrealis is that the final suffix of an irrealis verb – a subject-agreement pronoun – surfaces with its underlying form (see Table 84).
The verbal aspects and modalities that fall under the irrealis mood include the optative, subjunctive, imperative, negative, seqential, and simultaneous.
#### **8.9.2 Realis (real)**
In contrast to irrealis, the realis mood includes states and events whose actuality or reality *are* encoded in the grammar. That is to say, realis verbs in Ik include in their meaning the fact that something has taken place, is taking place, or will take place in the real world. The morphological manifestation of the realis mood is seen in the realis suffix {-a} that subtracts or deletes the final vowel of the subject-agreement suffix to which it attaches (again, see Table 84). In terms of verb types, the realis mood includes declarative statements in the past or nonpast, questions about the past or non-past, and, rather paradoxically, negative imperatives (which one might expect to fall under irrealis).
#### 8 Verbs
## **8.10 Verb paradigms**
### **8.10.1 Intentional-imperfective (int/ipfv)**
The intentional-imperfective aspect suffix {-és-} has two functions, hence its hyphenated title. One is to denote either an intention on the part of animate subjects or an imminence on the part of inanimate subjects. It is in this role that it finds use as the usual translation for the English future tense. It is also the answer to the question, "How do you express future tense in Ik?" A second function is to denote grammatical imperfectivity: a sense that a state or event is ongoing or incomplete. The two concepts collapse into one when intention/imminence is viewed as the incomplete coming-to-be of a future state or event. And even though intention or imperfectivity may seem to fall under an irrealis mood, {-és-} can actually be used with verbs in either the realis or irrealis mood. In Table 86, {-és-} is illustrated with the verb *àts-* 'come' in its imperfective sense with a recent past tense marker (*nákᵃ*) and then in its intentional (English 'future') sense:
Table 86: Ik intentional-imperfective aspect
#### Grammar sketch
#### **8.10.2 Present perfect (prf)**
The Ik present perfect suffix {-ˊka} denotes a state or event recently completed ('perfected') but still relevant in the present. The suffix has a 'floating' high tone that shows up on the preceding syllable of 3sg verbs, for example in *Nabʉƙɔták ᵃ* 'It is finished'. The /k/ in {-ˊka} disappears in non-final environments, making {-ˊa} an allomorph. Table 87 presents the paradigm of the present perfect with the verb *àts-* 'come' in both non-final and final environments:
Table 87: Ik present perfect aspect
#### **8.10.3 Optative (opt)**
The Ik optative mood is used to express wishes, even ironic ones like 'Let the enemies comeǃ'. Optative verbs are often introduced with imperative verbs like *Ógoe* or *Taláké*, both of which mean 'Let …'. And all Ik optative verbs are translated into English with a sentence beginning with 'Let …' or 'May …'.
Morphologically, the optative is marked by a combination of tone and special irregular suffixes. All optative verbs except 3pl show a kind of high-tone 'leveling' in the subject-agreement suffixes. The leveled high tone is pushed out to the end, creating a floating high tone. This high tone is not seen except in the fact that the last syllable of the subject-agreement suffixes remains at mid-tone level. Besides tone, special irregular suffixes mark the optative in 1sg, 1pl.exc, and 1pl.inc verbs, while standard irrealis suffixes are used for the other paradigm members. Note that the 1pl.inc may also be called the 'hortative'. Another peculiarity of the Ik optative is that there is no difference between its non-final and final forms. Table 88 presents the optative on the verb *àts-* 'come':
Table 88: Ik optative mood
### **8.10.4 Subjunctive (subj)**
The Ik subjunctive mood is used to encode statements that are somehow contingent or temporally unrealized. In that regard, it is an essentially irrealis verb form because it captures states or events that have not yet happened. It is also essentially irrealis in that it is marked simply by the absence of any marking. In other words, the subject-agreement suffixes surface with their underlying forms in the subjunctive mood, just as they appear in Table 84. The subjunctive is usually introduced either by *ɗɛmʉsʉ* 'unless, until' or *damu (kóʝa)* 'may'. Table 89 gives the full subjunctive paradigm with the verb *àts-* 'come':
Table 89: Ik subjunctive mood
#### **8.10.5 Imperative (imp)**
The imperative mood is used to issue commands or instructions. If the recipient of the command is singular, then the suffix used is {-eˊ}, and if the recipient is
#### Grammar sketch
plural, the suffix is {-úó}. The singular {-eˊ} has a floating high tone that raises any preceding low tones to mid. Both imperative suffixes are appended to the end of the verb stem, and no subject-agreement markers are needed. Both imperative suffixes are subject to vowel devoicing before a pause, as shown in Table 90:
#### **8.10.6 Negative**
Ik negates clauses by means of verblike particles that come first in the negative clause. If the negated clause has a realis verb, then the negator particle used is *ńtá* 'not'. If the negated clause has an irrealis verb, then the negator particle is *mòò* or *nòò*. Lastly, if the negated clause is past tense realis or present perfect realis, then the negator particle used is *máà* or *náà*. In the negated clause, the negator particle comes first, followed by the subject, and then the verb. Any negated verb takes the irrealis mood with the appropriate form of subject-agreement suffixes (see Table 84). To make all this more concrete, Table 91 gives example of the different negator particles used with different types of clauses.
### **8.10.7 Sequential (seq)**
The Ik seqential aspect expresses states or events that happen in sequence. Usually a sequence of verbs starts with an anchoring non-sequential verb and/or time expression, and then a clause chain begins in the sequential aspect. For example, when someone tells a story, they may start with one or two past tense realis verbs to set the stage and then continue the narrative with sequential verbs. Or if someone is giving a set of instructions, they may start with one or two imperative verbs followed by a chain of sequential verbs. Because of its versatility, the Ik sequential aspect is the language's most frequently used verb form.
Table 91: Ik negative mood
Morphologically, Ik sequential verbs are recognized by a combination of tone, irregular subject-agreement suffixes, and the sequential aspect suffix {-ko}. Specifically, all 1 and 2-person sequential verbs exhibit high-tone leveling in their subject-agreement suffixes, which pushes a high tone out to the right of the verb. This floating high raises the preceding low tones to mid. These tone effects, plus the irregular suffixes, and the sequential marker {-ko} are shown in Table 92. Note that the sequential paradigm also has an impersonal passive marked with the suffix {-ese}. Its function is identical to that of the impersonal passive described back in §8.6.2. For more on how the sequential aspect works in actual language contexts, skip ahead to the discussion of clause-chaining in §10.8.2.
### **8.10.8 Simultaneous (sim)**
The Ik simultaneous aspect is used to express states or events that are happening simultaneously to another state or event. In contrast to the sequential, the simultaneous aspect can only be used in subordinate clauses. That is to say, simultaneous clauses usually cannot stand alone without a main clause (with some exceptions). Because of its role of supporting sequential clauses, the simultaneous aspect is also commonly found in narratives and other longer discourses. It can be given a perfective interpretation as in 'when I came' or an imperfective one as in 'while I was coming'. Morphologically, the simultaneous aspect is
#### Grammar sketch
Table 92: Ik sequential aspect
marked by the suffix {-ke}, which is affixed to the subject-agreement suffixes in their irrealis forms. Table 93 presents the simultaneous paradigm of *àts-* 'come':
Table 93: Ik simultaneous aspect
## **8.11 Adjectival verbs**
### **8.11.1 Overview**
Since Ik does not have a separate word class of adjectives, it conveys adjectival concepts with adjectival verbs. These verbs have adjectival meanings but otherwise mostly behave like intransitive verbs. One way they do differ from normal intransitive verbs, though, is in the specific adjectival suffixes they can take. The next four subsections briefly describe these special adjectival suffixes.
#### 8 Verbs
#### **8.11.2 Physical property I (phys1)**
The physical property i adjectival suffix {-ˊd-} is found on adjectival verbs that express physical properties like appearance, size, shape, consistency, texture, and other tangible attributes. As a result, physical property I verbs are some of the language's most colorful adjectivals. Physical property I verbs all contain two syllables with LH tone pattern, and in the infinitive, they take the intransitive suffix {-ònì-}. Table 94 gives a sample of these colorful descriptive terms:
#### **8.11.3 Physical property II (phys2)**
The physical property ii adjectival suffix {-m-} is found in adjectival verbs that also express physical properties like appearance, color, consistency, posture, shape, and texture. It can also express less tangible attributes like strength, weakness, quality, or personality traits. Physical property II verbs usually contain two syllables with a LH tone pattern or three syllables with a LHH tone pattern (without the infinitive suffix), and in the infinitive, they take the intransitive suffix {-ònì}. Table 95 gives a sample of these descriptive adjectival verbs in two groupings.
#### **8.11.4 Stative (stat)**
The stative adjectival suffix {-án-} forms adjectival verbs that express an ongoing state characterized by the meaning of a noun or a transitive verb. Because {-án-} contains the vowel /a/, it prevents vowel harmony from spreading between the verbal root and any suffixes that follow the stative suffix (for example, infinitive or subject-agreement suffixes). Table 96 and Table 97 present a few examples of stative adjectival verbs derived from nouns and verbs, respectively:
Table 95: Ik physical property II adjectival verbs
#### Table 96: Ik stative verbs derived from nouns
Table 97: Ik stative verbs derived from transitive verbs
9 Adverbs
### **8.11.5 Distributive (distr)**
Ik has two distributive adjectival suffixes: {-aák-} and {-ìk-}. These suffixes have the function of distributing the meaning of an adjectival verb to more than one subject. The suffix {-aák-} can be used with all kinds of adjectival verbs, including the physical property and stative varieties, while the suffix {-ìk-} has been found only with the two verbs of size, *kwáts-* 'small' and *zè-* 'large'. Moreover, it commonly occurs together with {-aák-}, as in *kwátsíkaakón* 'to be small (of many)' and *zeikaakón* 'to be large (of many)'. Table 98 gives a sampling of adjectival verbs with the distributive suffix:
Table 98: Ik distributive adjectival verbs
## **9 Adverbs**
## **9.1 Overview**
The word class called adverbs is a catch-all category that includes words and clitics of various sorts that say something descriptive about a whole clause, for example, 'how' or 'when' it takes place, or how the speaker feels about the certainty or contingency of the clause. Accordingly, Ik adverbs can be divide up into manner adverbs, temporal adverbs, and epistemic adverbs. The following subsections take up each of these adverbial categories in a brief discussion.
## **9.2 Manner adverbs**
manner adverbs modify whole clauses by commenting on, for example, the manner in which a state comes across or in which an action is done. Manner adverbs usually come near or at the end of the clause they modify, as shown in example sentences (87)-(88) below. Table 99 presents a sampling of these adverbs:
Table 99: Ik manner adverbs
## **9.3 Temporal adverbs**
## **9.3.1 Overview**
The Ik temporal adverbs situate their clause somewhere in the course of time. Ik has sets of temporal adverbs that deal with past tense, past perfect tense, and non-past (including future) tense. The past and past perfect tense adverbs are enclitics that come directly after the verb they modify. The future tense adverbs are free adverbs that come near the end or at the end of the clause.
## **9.3.2 Past tense adverbs (pst)**
Ik divides past tense into four time periods and marks them with adverbial enclitics. They are: 1) recent past that covers the current day and is marked with =*nákà*, 2) removed past that covers yesterday (or any last or 'yester-' time period like 'yesterday' or 'yesteryear') and is marked with =*bàtsè*, 3) remote past that covers a few days or weeks before yesterday and is marked with =*nótsò*, and finally, 4) remotest past that covers everything before the remote past and is marked with =*nòkò*. Each of these tense enclitics comes directly after ther verb
#### 9 Adverbs
and has a non-final and final form. Table 100 illustrates the Ik tense markers in all their forms, and examples (89)-(90) illustrate their typical post-verbal position in a sentence:
Table 100: Ik past tense markers
#### **9.3.3 Past perfect tense adverbs (pst.prf)**
The past tense can be combined with a perfect aspect to yield the past perfect tense. Unlike the simple past tense adverbs, Ik past perfect tense adverbs operate along only three periods of time: recent (earlier today), removed (yester-), and remote (before yester-). Table 101 presents the Ik past perfect tense adverbs, and example sentences (91)-(92) illustrate their use in natural contexts:
Table 101: Ik past perfect tense markers
#### Grammar sketch
### **9.3.4 Non-past tense adverbs**
Ik divides the non-past tense into three rather vaguely defined time periods suggested by three adverbs. They are: 1) the distended present that includes just before and just after the present and is expressed by the adverb *ts'ɔ̀ɔ̀*, 2) the removed future that includes the *next* future time period (next hour, next day, next year) and is expressed by the adverb *táá*, and 3) the remote future expressed by the adverb *fàrà* (occasionally *fàrò*). Table 102 arranges these adverbs in a paradigm, while (93)-(94) below illustrates them in natural sentences:
Table 102: Ik non-past tense markers
- b. Atsésíà come:int:1sg **ts'ɔ̀ɔ̀**. soon 'I will come soon.'
9 Adverbs
## **9.4 Epistemic adverbs**
### **9.4.1 Overview**
The Ik epistemic adverbs express how the speaker feels or thinks about the certainty or contingency of the clause. Accordingly, this set of adverbs can be divided into the categories of inferential, confirmational, and conditionalhypothetical. All of the epistemic adverbs are enclitics that follow the verb in normal main clauses, but some of them can also be moved in front of the verb.
### **9.4.2 Inferential adverbs (infr)**
Ik can communicate a degree of uncertainty about a situation by means of a set of inferential tense-based adverbs. This sense of making a tentative inference based on an observation can be translated into English with such turns of phrase as 'Apparently …', 'Maybe …', 'It seems that …', 'must have', etc. Two of these inferential particles consist of the proclitic *ná* plus a past-tense particle, while the third combines *ná* with the adverb *tsamʉ*. Table 103 presents the three inferential adverbial particles in their final and non-final forms. Note that compared to the past-tense markers above in Table 100, the inferential time-scale is moved up one notch more recent. Examples (95)-(96) show the Ik inferential adverbs in context. Note that they can be placed before or after the main verb.
(95) a. Baduƙota=**nábàtsᵉ**. die:comp:3sg=infr 'It died, apparently.'
> b. **Nábee** infr baduƙotᵃ. die:comp:3sg 'Apparently, it died.'
Grammar sketch
(96) **Nánoo** infr teremátᵃ. separate:3pl 'It looks like they separated.'
## **9.4.3 Confirmational adverbs (conf)**
Ik can also issue a confirmation of a state or event by means of a set of confirmational adverbs that are derived from the tensed relative pronouns described back in §5.7. When used, these adverbs are placed before the verb, and the verb surfaces in its non-final form, almost like a question rendered in English 'Why yes, did X *not* happen?' – meaning that, of course, it *did* happen. These suffixes are first presented in Table 104 and then demonstrated in (97)-(98):
Table 104: Ik confirmational markers
- b. **Sɨna** conf ŋƙáƙótíà. eat:comp:1sg 'Yes, of course I did.'
- b. **Nòò** conf dètà. bring:3sg 'Yes, of course she did.'
## **9.4.4 Conditional-hypothetical adverbs (cond/hypo)**
If a state or event has not taken place but*could* or *would* take place, Ik can express that contingency with its conditional-hypothetical adverbs. There are three
10 Basic syntax
of these adverbs, but they are used to cover four periods of time. The first adverb covers non-past and recent past, the second removed past, and third remote past. These conditional-hypothetical adverbs are presented in Table 105:
Table 105: Ik conditional-hypothetical adverbs
The conditional-hypothetical adverbs come after the main verb:
## **10 Basic syntax**
## **10.1 Noun phrases**
The Ik noun phrase consists first and foremost of a noun 'head', either a lexical noun or a nominalized lexical verb. As a head-initial language, Ik places its noun phrase head first in the phrase. Any subordinate, supporting elements follow the head. These optional elements may include anaphoric demonstratives, possessive markers, relative pronouns/temporal demonstratives, number markers, and spatial demonstratives. The Ik noun phrase structure can be formalized as follows, where elements in parentheses are optional:
#### Grammar sketch
(101) Ik NP structure: head (anaph)(poss)(num)(rel/temp) (dem)
The syntactical structure of noun phrases formalized in (101) is fleshed out among the real Ik noun phrases presented below in examples (102)-(110):
(102) head wikᵃ 'children' (103) head wika anaph díí 'those (specific) children' (104) head wika poss ɲ́cì 'my children' (105) head wika anaph díí poss ɲ́cì 'those (specific) children of mine' (106) head wika anaph díí poss ɲ́cìè num lèɓètsè 'those two (specific) children of mine' (107) head wika anaph díí poss ɲ́cie rel [ni leɓetse]rel 'those (specific) children of mine, two in number' (108) head wika anaph díí poss ɲ́cie num leɓetse rel [ní dà] rel 'those two nice (specific) children of mine' (109) head wika anaph díí poss ɲ́cie num leɓetse rel [ní daa]rel dem ni 'those two nice (specific) children of mine, these' (110) head wika anaph díí poss ɲ́cie num leɓetse temp níi dem ni 'those two (specific) children of mine from earlier, these'
10 Basic syntax
## **10.2 Clause structure**
### **10.2.1 Intransitive**
Ik intransitive clauses consist minimally of a verb (v) and a subject (s) in a vs constituent order. The subject may be explicit, in which case it follows the verb, or it may be implicit, in which case it is merely marked on the verb. Basic intransitive clause structure is illustrated in example (111):
(111) Epa<sup>v</sup> sleep:3sg ŋókᵃ<sup>s</sup> . dog:nom 'The dog sleeps.'
When a tense adverb is needed, it comes directly after the verb and before any explicit subject. And any other adverbial elements like extended objects (e) or adverbs, in that order, come after the subject. This word order is shown in (112):
(112) Epáv=beetense sleep:3sg=yesterŋóká<sup>s</sup> dog:nom kurúe. shade:abl 'The dog slept in the shade yesterday.'
## **10.2.2 Transitive**
Ik transitive clauses consist minimally of a transitive verb (v), an agent (a), and an object (o) in a vao constituent order. The subject may be explicit, in which case it comes between the verb and object, or it may merely be marked on the verb with a suffix. The object may also be dropped, in which case it is inferred from the context. Example (113) illustrates basic transitive clause structure:
(113) Áts'á<sup>v</sup> gnaw:3sg ŋóká<sup>a</sup> dog:nom ɔkákᵃo. bone:acc 'The dog gnaws the bone.'
When a tense adverb is needed, it comes directly after the verb and before any explicit subject. And any other adverbial elements like extended objects (e) or adverbs, in that order, come after the subject. This syntax is shown in (114):
(114) Áts'áv=beetense gnaw:3sg=yesterŋóká<sup>a</sup> dog:nom ɔkáá<sup>o</sup> bone:acc ódàtù<sup>e</sup> day:ins 'The dog gnawed the bone all day yesterday.'
#### Grammar sketch
### **10.2.3 Ditransitive**
Ik ditransitive clauses consist minimally of a ditransitive verb (v), an agent (a), an object (o), and an extended object (e) in a vaoe constituent order. If the agent is not mentioned explicitly, then it will still be marked with a suffix on the verb. The object and extended object may be left implicit but will be understood from context. The basic ditransitive clause structure is illustrated in (115):
(115) Maa<sup>v</sup> give:3sg ƙaƙaama<sup>a</sup> hunter:nom ɔkáá<sup>o</sup> bone:acc ŋókíkᵉe. dog:dat 'The hunter gives a bone to the dog.'
### **10.2.4 Causative**
By adding an extra element in the form of a causing agent, Ik causative verbs change the structure of a clause. If the original clause was a vs intransitive one, then the causative changes it to a transitive vao. If the original clause was a transitive vao, then the causative changes it to a ditransitive vaoe. The following two examples, (116)-(119), show causative verbs making these structural changes:
Intransitive vs → Causative vao
Transitive vao → Causative vaoe
10 Basic syntax
#### **10.2.5 Auxiliary**
Ik has both true auxiliary verbs and pseudo-auxiliary verbs. Both types modify sentence syntax. The true auxiliaries, shown in Table 106, function as the syntactic main verb in a clause, while the *semantic* main verb follows the subject (s/a) in a morphologically defective form that consists of the bare verb stem plus a suffix {-a} (which may be the realis marker from §8.9.2). This means the constituent order of clauses with true auxiliary verbs is auxSV for intransitives, auxAVO for transitives, and auxAVOE with extended objects. Again, in all these constructions, the aux acts as the main verb from a syntactic perspective, while the defective verb carries the main meaning of the verbal schema. Another way to analyze this construction would be to say that the auxiliary verb and the defective verb *together* fill the single verb slot of the clausal syntax.
The true auxiliaries have both lexical and aspectual meanings, which are nevertheless practically identical in their semantics. However, in their lexical function, the verbs in Table 106 do not require a second, morphologically defective verb to augment them; in their strictly lexical usage, they stand alone:
Table 106: Ik true auxiliary verbs
Example (120) illustrates the use of the recentive aspectual auxiliary verb *erúts*in an intransitive clause with the structure auxSVE:
(120) ErútsímaAuxS recent:1pl.exc atsa<sup>v</sup> come sédàᵒe. garden:abl
'We just came from the garden.'
Example (121), on the other hand, shows the use of the anticipative verb *ŋɔ́r-* in a transitive clause with the structure auxAVOE:
(121) Ŋɔ́ráAuxA=naa anticip:3sg=pst1 cɛa<sup>v</sup> kill riáá<sup>o</sup> goat:acc baratsoe=nákᵃ. morn:ins=dem.pst1 'He already killed the goat earlier this morning.'
#### Grammar sketch
Lastly, sentence (122) exemplifies the durative aspectual verb *sár-* in a simple transitive clause working with the defective verb *ts'ágwa-*:
(122) SáráAux dur:3sg séda<sup>s</sup> garden:nom ts'ágwàv. unripe 'The garden is still unripe.'
In contrast to the above examples, the pseudo-auxiliary verbs only mimic true auxiliaries in that they are fully lexical verbs yet ones with potentially aspectual meanings, including the completive, inchoative, and occupative. However, because they are not *syntactically* auxiliary, they take complements as any lexical verb would (direct objects for the transitive ones and extended objects for the intransitive one). The pseudo-auxiliaries are presented in Table 107 with their lexical and aspectual meanings and the cases required in their complements:
Table 107: Ik pseudo-auxiliary verbs
Each of the aspectual meanings listed in Table 107 are given one example in the following sentences. The brackets in example (123) signify that the bracketed noun phrase as a whole is the object of the verb:
Completive
(123) Nábʉƙɔtíáava finish:1sg:prf [isóméésá to.read:nom ɲáɓúkwi]o. book:gen 'I have finished reading the book.'
Inchoative
(124) Itsyaketátaava begin:3pl:prf wáánàkᵃo. praying:acc 'They have begun praying.'
10 Basic syntax
Occupative
(125) Cɛma<sup>v</sup> fight:3 wika<sup>s</sup> children:nom wáákᵒe. playing:ins 'The children are busy playing.'
### **10.2.6 Copular**
Ik copular clauses have relational rather than referential meanings. They link a copular subject (cs) to a copular complement (cc) which represents an entity or attribute, depending on the specific copular verb involved. The constituent order of copular clauses is therefore v-cs-cc. Ik has three distinct copular or 'be' verbs that can express five copular relationships between them. These copular verbs are presented in Table 108 below, along with the case markings their subjects and complements are obligated to have:
Table 108: Ik copular verbs
The three copular verbs in Table 108 and their five potential meaning are each exemplified briefly in the example sentences (126)-(130):
Existence
(126) Ia<sup>v</sup> be:3sg didigwarícs. rain.top:nom 'Heaven [i.e. God] is (there).'
Location
(127) Ia<sup>v</sup> be:3 lɔŋɔ́tács enemies:nom muceékᵉcc. way:dat 'Enemies are on the way.'
#### Grammar sketch
## Attribution
(128) Iravcs be:3sg tíyéadv. like.this 'It is like this.'
Identity
(129) Mɨtíá<sup>v</sup> be:1sg ŋkacs I:nom bábòcc. father.your:obl 'I am your father.'
### Possession
(130) Mɨta<sup>v</sup> be:3sg [awa=na]cs home:nom=this ŋgóᵉcc. we:gen 'This house is ours.'
## **10.2.7 Fronted**
Ik can put special emphasis on any core nominal element by moving it to the front of the clause, before the verb, subject, and other constituents. Doing so obviously disrupts the usual syntactic structure of main clauses. Two kinds of fronting are observed in the language: 1) a cleft construction and 2) left-dislocation. In a cleft construction, the emphasized noun is moved to the front and given the copulative case. This puts it in an identifying relationship with the original clause out of which it just came. As a result, the newly arranged clause can be viewed as a kind of copular clause where the fronted element is the copular subject and the original clause the copular complement. This can in turn be formulized as: [NP:cop]cs [clause]cc. To make this more concrete, the next examples show the cleft construction with a simple transitive clause in (131) whose object (*mɛ̀s*) gets fronted and marked with the copulative case in (132):
Cleft construction
10 Basic syntax
Whereas the cleft construction involves removing a clausal element from a clause and building a new clause, left-dislocation simply relocates the element to the front of the clause, but still within the same clause. In this fronted position it is given the nominative case. This type of fronting can be formulized as: [NP:nom ‖ clause]clause, where the double vertical line symbolize a short pause. This type of left-dislocation is illustrated between example sentences (133)-(134):
#### Left-dislocation
## **10.3 Subordinate clauses**
#### **10.3.1 Overview**
The constituent order of Ik subordinate clauses differs from that of main clauses. Ik subordinate clauses exhibit an sv order with intransitive verbs, an av order with transitives, and an ave order with ditransitives – in short 'sv' instead of the usual 'vs'. Case marking in subordinate clauses is also different: The fronted subject/agent and *every* direct object take the accusative case.
The next two subsections deal with two key kinds of Ik subordinate clause, the relative clause (§10.3.2) and the adverbial clause (§10.3.3).
### **10.3.2 Relative clauses**
relative clauses are subordinate clauses that modify a noun within a main clause. Ik relative clauses are restrictive, meaning they can only narrow the reference of their head noun rather than merely adding extra details about it. Relative clauses are introduced by the tensed relative pronouns discussed back in (§5.7), which, within the relative clause, stand in for a noun in the main clause called the common argument (ca). As such, the common argument is a full verbal argument in the main clause, while in the relative clause, the relative pronoun fills its syntactic slot.
#### Grammar sketch
As a subordinate clause, an Ik relative clause exhibits a different constituent order than typical main clauses. Specifically, an intransitive relative clause has the ordersv (instead of vs), and a transitive relative clause has the order oav (instead of vao). In the former (intransitive), the subject slot (s) is filled by the relative pronoun, and in the latter (transitive), it is the object (o) that is represented by the relative pronoun. Furthermore, apart for the relative pronouns themselves, all subjects and direct objects in relative clauses are marked with the accusative case – another sign of grammatical subordination in Ik.
These attributes of Ik relative clauses are illustrated in examples (135)-(136). In (135), the common argument in the main clause is *emuta* 'story', which is modified by the relative clause *nɛ ɛ́f* 'that is funny'. Note how the subject slot of the relative clause is filled by the relative pronoun *nɛ* (*na* with its vowel assimilated). Then, in (136), the common argument of the main clause is *ima* 'child', modified by the relative clause *náa ɲcia tákí* 'that I mentioned'. Since the verb of the relative clause is transitive (*tákés* 'to mean, mention'), it requires an object, which in this case is fulfilled by the relative pronoun *náa* representing the noun *ima*:
Intransitive (sv)
(135) Nesíbimaa hear:1pl.exc:prf emutaca=[nɛ<sup>s</sup> story:nom=rel ɛ́fv]rel. sweet:3sg 'We've heard a story that is funny.'
Transitive (oav)
(136) Atsáá come:3sg:prf imaca=[náa<sup>o</sup> child=rel ɲcia<sup>a</sup> I:acc tákív]rel. mention:1sg 'The child I mentioned earlier has come.'
### **10.3.3 Adverbial clauses**
The category of adverbial clausesis rather broad as it includes any subordinate clause that modifies a main clause adverbially. Adverbial clause are subordinate or 'dependent' precisely because they cannot stand alone but must be linked to an independent main clause. As subordinate clauses, adverbial clauses exhibit a constituent order that differs from both main clauses and relative clauses. Specifically, intransitive adverbial clauses have the order sv, while transitive adverbial clauses have the order avo. Another correlate of subordination seen in most adverbial clauses – except for the conditional and hypothetical ones – is accusative case-marking on all core constituents (s/a/o) if they are explicitly mentioned.
10 Basic syntax
Among the main kinds of adverbial clause in Ik are the following: temporal, simultaneous, conditional, hypothetical, manner, reason/cause, and concessive. Most types of adverbial clause – except for manner – have their own dedicated connective (or 'conjunction') or set of connectives, many of which are listed back in Table 24 under §3.14. Without exception, the subordinating connectives come first in the adverbial clause. Lastly, in terms of position, Ik adverbial clauses may come before or after the main clause they modify. Each of these types of adverbial clause is given one example apiece in (137)-(143):
Temporal
(137) [Noo when ntsíá he:3sg baduƙotâdᵉ]temp, die:3sg:dp ƙɔ́ɗɨakᵒ. cry:1sg:seq 'When he died, I cried.'
Simultaneous
(138) [Náa as ntsíá he:3sg badúƙótìkᵉ]simul, die:3sg:sim ƙɔ́ɗɛ́sɨakᵒ. cry:ipfv:1sg:seq 'As he was dying, I was crying.'
Conditional
(139) [Na if ntsa he:nom badúƙótùkᵒ]cond, die:3sg:seq ƙɔ́ɗɨakᵒ. cry:1sg:seq 'If he dies, I'll cry.'
Hypothetical
- b. ƙɔ́ɗɨaa cry:1sg:seq ƙánòkᵒ. would've I would've cried.'
### Manner
(141) Badúƙótuo die:3sg:seq [(ntsíá) (he:acc) tisílíkᵉ]manner. peaceful:3sg:sim 'And he died peacefully (lit. 'he being peaceful').'
#### Grammar sketch
Reason/cause
(142) Baduƙotáá die:3sg:prf [ɗúó because ídzanâdᵉ]reason. shoot:ips:3sg:dp 'He has died because he was shot.'
Concessive
(143) [Áta even ntsíá he:acc badúƙótìkᵉ]concess, die:3sg:sim ńtá not ƙɔ́ɗí. cry:1sg 'Even if he dies, I will not cry.'
## **10.4 Questions**
#### **10.4.1 Overview**
Questions in Ik can be formed in two mutually exclusive ways: 1) by leaving the final word in the question in its non-final form (along with a questioning intonation) or 2) by using interrogative pronouns and often rearranging the syntax of the sentence. The first method is employed with what is called polar or yes/no questions: those whose answer is either 'yes' or 'no'. The second method is used for content or wh-questions: those whose answer is a substantive response to such interrogative pronouns as *who?*, *what?*, *when?*, *where?*, etc. These two types of question are briefly described in the following two subsections.
#### **10.4.2 Polar questions**
Polar questions are those that elicit a 'yes' or 'no' in response. In Ik, they are formed by leaving the last word or particle of the question in its non-final form (revisit §2.3 and §2.4.3 for a review). This open-endedness of form is a fascinating way the grammar reflects the open-endedness of a question – open to a response. Besides the non-final form of the last word, polar questions are identified by a change in intonation. This interrogative intonation is enacted by what is called a boundary low tone: a low tone that attaches to the final syllable. If the final syllable already has a low tone, then the boundary tone is not audible. But if the final syllable has a high tone, the boundary tone manifests as a high-low glide.
Examples (144)-(145) illustrate these features of polar questions. Note in the first part of (144) how the present perfect suffix {-ˊka} shows up in its non-final form (*-ˊà*), while in the second part, the final form is used (*-ˊkᵃ*). Then, (145) shows the interrogative boundary low tone attaching to the high tone on the final syllable of *cekúó* 'is a woman', creating a high-low down-glide (*cekúô*):
10 Basic syntax
- b. Ee, yes nábʉƙɔtákᵃ. finish:comp:3sg:prf[ff] 'Yes, it is finished.'
(145) a. Cekúô? woman:cop[nf]
'Is it a woman?'
b. Ee, yes cekúó woman:cop ntsaᵃ. she:nom 'Yes, it's a woman.'
## **10.4.3 Content questions**
In contrast to polar questions, content questions cannot logically take 'yes' or 'no' for an answer. Rather, answers to content questions – as their name implies – must contain content relevant to the specific interrogative pronoun used to make the inquiry (Ik interrogative pronouns are listed in Table 46). So if the question contains the pronoun *ǹdò-* 'who?', the answer must include a person. Or if the question contains the pronoun *ndaí-* 'where?', the response must refer to a specific location, and so on. Ik forms content questions by placing an interrogative pronoun in the syntactic slot of the unknown entity being queried (i.e. a person, place, time, manner, etc.). For example, in (146), the interrogative pronoun *ndaí-* 'where?' is filling the normal place where an object encoding the destination of *ƙà-* 'go' would go. A similar thing occurs in (147), where the pronoun *ìsì-* 'what?' fills the direct object slot required by the verb *bɛ́ɗ-* 'want':
However, what is more common is for the interrogative pronoun to be fronted for emphasis. As in other instances of fronting in Ik (see §10.2.7), the fronted
#### Grammar sketch
element takes the copulative case marker {-ko}. In (148)-(149), examples (146)- (147) are repeated in their fronted (focused) forms, and two other interrogative pronouns are used in (150)-(151) to illustrate content questions:
## **10.5 Quotations**
Quotations involve reporting someone's speech (or thought) – the speaker's own or someone else's – directly or indirectly. Ik fulfills this communicative need through the use of the verb *kʉ̀t-* 'say' followed by the actual quotation treated as an add-on clause. That is, unlike complements described below in §10.6, a quoted sentence in Ik is technically *not* an object of the verb *kʉ̀t-*. Instead, it is tacked on 'extra-syntactically' and given the oblique case (the 'leftover' case). This is proven by the fact that when the pronoun *ìsì-* 'what?' appears to be the object of *kʉ̀t-* with a 3sg or 3pl subject, *ìsì-* takes the oblique case instead of the accusative case as one would expect otherwise from case grammar (§7.3).
Many languages, English included, distinguish between direct and indirect quotative formulas, for example the direct "I said, 'I will come'" versus the indirect "I said I will come". By contrast, Ik does not distinguish the two grammatically. Instead, the proper sense has to be discerned from the context (and possibly from intonation). So the statement *Kʉtíá naa atsésí* could mean either "I said, 'I will come'" or "I said I will come", depending on factors other than syntax.
In Ik quotative sentences, if there is an addressee of the quotation, they will appear in the dative case. And the quotative particle *tàà* 'that' is often inserted
#### 10 Basic syntax
just before the quotation, though by all appearances it is optional. The example sentences (152)-(153) provide a demonstration of the quotative construction:
## **10.6 Complements**
Complements are individual clauses that function as an 'argument' of the verb – as either subject or object. In other words, they are clauses within clauses. Unlike subordinate clauses which are added *onto* main clauses, complement clauses are added *into* other clauses. The main type of Ik complement clause is introduced by the complementizer *tòìmɛ̀nà-* 'that', which is combination of a form of the verb *tód-* 'speak' and the noun *mɛná-* 'issues, words'. This compound word gives some evidence that Ik complement clauses (of this particular type) evolved from quotative clauses like those described above in §10.5.
Because a complement clause fits within the clausal grammar, it must somehow be declined for case (because all arguments of a verb in Ik take case, without exception). To meet this requirement, the complementizer *tòìmɛ̀nà-* bears the burden of case on behalf of the whole complement clause it is introducing. So technically, it is the complementizer – not the complement clause alone – that is the verbal argument. But because *tòìmɛ̀nà-* plus the complement is a frozen quotative formula, the whole construction can be analyzed as an argument.
To illustrate this, (154) presents a simple complement clause governed by the cognitive verb *èn-* 'see'. The {curly brackets} indicate the boundaries of the main clause from the point of view of the syntax, in which the verb *èn-* 'see' selects its object *tòìmɛ̀nà-* 'that' for the accusative case. The [square brackets] mark the boundary of the complement clause seen from the point of view of semantics, for the actual content of 'seeing' is the clause *that we have become very rich*:
(154) {Enáta see:3pl [toimɛnaa}obj that:acc barʉƙɔtímáà rich:comp:1pl.exc:prf zùkᵘ]compl very 'They see that we have become very rich.'
#### Grammar sketch
In addition to a direct object, an Ik complement clause can also function as an indirect object or even the 'complement' of a copular clause. For instance, in (155) below, *tòìmɛ̀nà-* and by extension the whole complement clause is acting as the indirect object of the verb *xɛ̀ɓ-* 'be afraid of, fear', which requires the ablative case. Then, in (156), the verb is the copular verb *mìt-* 'be', which requires its nominal compliment to be in the oblique case, as is seen with *tòìmɛ̀nà-*:
## **10.7 Comparatives**
Comparatives are grammatical constructions that allow the comparison of two entities on the basis of some shared characteristic. Ik has two strategies for doing this: 1) the mono-clausal, which involves one simple clause, and 2) the bi-clausal, which involves a complex clause. Mono-clausal comparatives place the comparee (entity being compared) in the nominative case and the standard (entity the comparee is being compared to) in the ablative case. Since most comparable attributes are expressed as intransitive verbs in Ik, the parameter (attribute) of the comparison is also an adjectival verb in such constructions. For example, in (157)-(158) below, the intransitive verbs *zè-* 'big' and *dà-* 'nice' are acting as the parameters, while their subjects are the comparees in the nominative case and their extended objects the standards in the ablative case:
Bi-clausal comparatives, on the other hand, combine a main clause with a subordinate or 'co-subordinate' clause (§10.8.2). Both types are introduced by the verb
#### 10 Basic syntax
*ɨlɔ́-* 'exceed, surpass', which acts as the index of the comparison (the gauge of the degree of difference between compared entities). If the indexical verb introduces a subordinate clause, it takes the simultaneous aspect, while if it introduces a co-subordinate clause, it takes the sequential aspect. In such bi-clausal comparatives, the comparee is still the subject of the main clause, while the standard is the object of the dependent clause. The parameter remains with the main clause verb (as in mono-clausal comparatives). But unlike mono-clausals, bi-clausal comparatives can have intransitive or transitive parametric verbs. In other words, actions as well as attributes can be compared in this type of construction.
In (159), the parameter lies with the verb *tɔkɔ́b-* 'cultivate', and 'he' (marked as 3sg on the verb) is being compared with 'us' (*ŋgó-*). The index of the comparison is the verb *ɨlɔ́íɛ* 'he surpassing', which reveals the inequality of the compared actions of the two entities. Example (160) follows the exact same logic, only that the indexical verb *ɨlɔ́ɨnɨ* is in the sequential aspect instead of the simultaneous:
## **10.8 Clause combining**
### **10.8.1 Clause coordination**
Two or more clauses can be linked in Ik through clause coordination. This can result in clause addition ('and'), which joins two independent clauses of equal status. It can result in contrast ('but'), which joins clauses of equal syntactic status, the second of which is a counterexpectation to the first. And thirdly, clause coordination can result in disjunction ('or'), in which two clauses of equal status are presented as different possible options.
Clause addition is achieved in two ways: 1) simply adjoining the clauses with a pause in between (represented by a period or comma in writing) or 2) linking the clauses with a coordinating connective like *kòtò* 'and, but, then' or *ńdà* 'and'. These first two methods are illustrated in (161)-(162). A third way to add one clause to another is to nominalize it – change all its main parts to nouns, put them in a noun phrase, and link it up to the other clause with *ńdà*. Note from
#### Grammar sketch
(163) that with this third method, because the word *ńdà* 'and' is acting as a sort preposition, it requires its head noun(s) to be in the oblique case. Its head nouns in (163) are the subject (*ŋgo*) and infinitive (*ŋƙɛ́sí*) – both in the oblique case:
- b. ńda and ƙaíána=noo go:plur:ips:real=pst waa pick:nom waicíkée greens:gen ńtí? how and how did they used to go picking greens?'
- b. ńda and ŋgo we:obl ŋƙɛ́sí to.eat:obl tɔbɔŋɔ́ᵉ. mush:gen and we ate mealmush.'
Contrast between two clauses in Ik can be expressed in two primary ways: 1) by simply adjoining the two clauses with a brief pause in between (marked with by a comma or period in writing) or 2) by linking the two clauses with the contrastive connective *kòtò*, which can mean 'but' as well as 'and, then, therefore, etc.'. These two types are demonstrated in examples (164)-(165), respectively:
- b. koto but máa=naa not=pst1 ŋunetí. find:1sg but I did not find (it).'
10 Basic syntax
Lastly, the idea of disjunction is expressed in Ik through the use of the connectives *kèɗè* 'or' or *kòrì* 'or', as illustrated in example sentences (166)-(167):
- b. keɗe or ńtá not tɔkɔ́bɛsîdᶤ? farm:ipfv:2sg or are you not farming (it)?'
- b. kori or maráŋaakátìkᵉ? good:distr:3pl:sim or as being good all around?
## **10.8.2 Clause chaining**
But in fact, the most common way Ik links independent clauses is through clause 'co-subordination' or clause chaining. To create a chain of clauses, the grammar starts with an anchoring phrase or clause to set the stage modally or temporally, and then it puts all the following mainline verbs in the sequential aspect (see §8.10.7), creating a chain of two or more clauses. When clause chaining is used in a story, the temporal 'anchor' can be a simple time expression like *kaíníkò nùkᵘ* 'in those years' or a tensed statement like *Atsa noo ámá ntanée taa Apáálɔrɛ́ŋ* 'There came a man named Apaaloreng'. In (168), the clause chain is anchored by the initial adverbial phrase *Na kónító ódoue baratsoó* 'One day, in the morning', which puts the whole sentence in a temporal frame. Thenceforth, the clause chain proceeds clause by clause, each marked as seq1, seq2, etc:
- b. [ipu**o** cast:3sg:seq taƙáíkakᵃ] shoes:acc seq1 he cast (his) shoes (in divination),
#### Grammar sketch
Although the sequential aspect and clause chains are common in narratives, they are also used extensively for other types of discourse, for example, exposition and instruction. The following expository clause chain in (169) details some of the steps taken in the process of grinding tobacco leaves. Note that there are two anchoring adverbial clauses, one at the beginning and one in the third line. After each one, there is a string of one or more verbs set in the sequential aspect:
- b. [ŋɔɛ́**ɛsɛ** grind:inch:sps ɲaɓáláŋɨtᵃ]seq1 soda.ash:nom soda ash is ground up.
- c. [náa when ɲaɓáláŋɨtíá soda.ash:acc iwíɗímètìkᵉ]adv2 pulverize:mid:sim When the soda ash is ground to powder,
- d. [páka until ɲapúɗúmùƙòtù**kᵒ**]seq4 powdery:comp:seq until it becomes fine powder.'
Finally, the sequential aspect and clause chaining is often found operating in a set of commands or instructions. Such a clause chain may begin with one or more imperative verbs, followed by the sequential verbs in a chain of further commands or instructions. This type of clause chain is shown in (170):
- b. [kawete cut:imp titíríkᵃ, pole:pl kɛɗɨtín, reed:pl ńda and sim]imp2 fiber Cut poles, reeds, and fiber,
- c. [iréɲuƙoidu**o** clear:comp:2sg:seq bácíkᵃ]seq1 area:nom clear away the area,
- d. [úgidu**o** dig:2sg:seq ripitín]seq2 hole:pl:nom dig holes,
- e. [otídu**kó**é pour:2sg:seq:dp titíríkᵃ]seq3 pole:pl:nom and put the poles into them.'
## **Appendix A: Ik affixes**
All of the affixes discussed in the preceding grammar sketch are listed in the table below for easy reference. When looking for an affix in the list, keep in mind that if it has two forms (for example the {-e} and {-ɛ} of the genitive case), both forms are given their own separate entry. Affixes that cannot be the terminal morpheme in a word have no final form, while those that can, have non-final and final forms.
Table 1: Full list of Ik affixes
Appendix A: Ik affixes
Appendix A: Ik affixes
## **References**
Heine, Bernd. 1999. *Ik dictionary*. Köln: Rüdiger Köppe Verlag. Schrock, Terrill. 2014. *A grammar of Ik (Icé-tód): Northeast Uganda's last thriving*
## **Subject index**
ablative case, 483, 582 accusative case, 482, 526, 546, 548, 575, 576, 580, 581 adjectival, 492, 535, 558–561, 582, 589– 591 adverb, 482, 564, 565, 567, 569, 573 adverbial clause, 575, 577 agentive, 508, 509 allomorphy, 551 anaphoric, 488, 492, 521–523, 567 andative, 542–544 aspect, 535, 544, 553, 556, 557, 563, 589–592 ATR, 7, 485–488, 502, 536, 537, 542, 545, 547, 549 auxiliary verb, 571 boundary tone, 578 causative, 549, 550, 570 certainty, 496, 561, 565 clause chaining, 585, 586 clause structure, 569, 570 cleft construction, 574, 575 clitic, 488, 490, 518 clusivity, 512, 549 common argument, 518, 575, 576 complement, 531, 533, 574, 581, 582 complementizer, 497, 498, 581 completive aspect, 544 constituent order, 569–571, 573, 575, 576
copula, 533 copular, 573, 574, 582 copulative case, 482, 483, 516, 574, 580 dative, 480–484, 501, 525, 526, 531, 542, 580, 592 definiteness, 514 depressor consonant, 488, 502 devoiced, 478 devoicing, 478, 490, 556 dialect, 3 diminutive, 508–510 directional, 590, 592 disjunction, 585 distributive adjectival, 561 dummy pronoun, 482, 551 genitive case, 530, 531, 589 haplology, 480 ideophone, 496 imperative, 495, 534, 554, 556, 586 imperfective, 553, 557 inchoative, 544, 547, 549, 572 infinitive, 8, 9, 482, 492, 509, 535, 536, 542, 545, 546, 549, 559, 584, 590–592 instrumental case, 480, 532 intentional, 553 interrogative, 491, 512, 515, 516, 578– 580
#### Subject index
intonation, 578, 580 intransitive, 482, 486, 518, 535–537, 539, 542, 544–547, 549, 558, 559, 569–572, 575, 576, 582, 583 left-dislocation, 575 main clause, 517, 518, 527, 557, 575– 577, 581–583 modality, 589, 590 mora, 502 negative copula, 533 nominative case, 8, 485, 505, 508, 526, 575, 582 noun phrase, 530, 535, 536, 567, 572, 583 oblique case, 525, 534, 580, 582, 584 orthography, xi, xii, 488 particle, 515, 556, 565, 578, 580 passive, 499, 535, 545–548, 557, 589, 590 patient, xii, 541, 547 plurative, 486, 491, 502–504 possession, 501, 505, 514, 529, 530 possessor, 505, 514 preposition, 494, 534, 584 proclitic, 565 quotative, 580, 581 recipient, 529, 555 reciprocal, 547, 548 reduplication, 545 reflexive, 491, 512, 518, 519, 537, 547 relative clause, 517, 518, 575, 576 relative pronoun, 492, 493, 517, 518, 575, 576
reported speech, 497 semi-vowel, 484, 502 sequential aspect, 556, 557, 583, 585, 586 simultaneous aspect, 557, 583 singulative, 491, 502, 504, 505 stative adjectival, 490, 559 subjunctive mood, 555 subordinate clause, 528, 575, 576, 583 syllable, 484, 488, 489, 502, 516, 554, 578 tense, 553, 556, 562–565, 569 time expression, 556, 585 vowel assimilation, 482–484, 526, 536, 537, 551 vowel harmony, 7, 478, 484, 486, 492, 502, 551, 559
# Did you like this book?
This book was brought to you for free
Please help us in providing free access to linguistic research worldwide. Visit http://www.langsci-press.org/donate to provide financial support or register as a community proofreader or typesetter at http://www.langsci-press.org/register.
## The Ik language
This book is a dictionary and grammar sketch of Ik, one of the three Kuliak (Rub) languages spoken in the beautiful Karamoja region of northeastern Uganda. It is the lexicographic sequel to *A grammar of Ik (Icé-tód): Northeast Uganda's last thriving Kuliak language* (Schrock 2014). The present volume includes an Ik-English dictionary with roughly 8,700 entries, followed by a reversed English-Ik index. These two main sections are then supplemented with an outline of Ik grammar that is comprehensive in its coverage of topics and written in a simple style, using standard linguistic terminology in a way that is accessible to interested non-linguists as well. This book may prove useful for language preservation and development among the Ik people, as a reference tool for non-Ik learners of the language, and as a source of data, not only for the comparative study of Kuliak but also the wider Afroasiatic and Nilo-Saharan language families. | doab | 2025-04-07T03:56:58.157539 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "0024144a-a6ce-4490-ad8f-59c145166ff3",
"url": "https://library.oapen.org/bitstream/20.500.12657/31322/1/631280.pdf||https://library.oapen.org/bitstream/20.500.12657/31322/1/631280.pdf||https://library.oapen.org/bitstream/20.500.12657/31322/1/631280.pdf",
"author": "Schrock, Terrill",
"title": "The Ik language: Dictionary and grammar sketch",
"publisher": "Language Science Press",
"isbn": "",
"section_idx": 0
} |
|
0025fd58-dc3d-4c1b-9075-3b20d5ae0a98 | # MITOCHONDRIA AND ENDOPLASMIC RETICULUM DYSFUNCTION IN PARKINSON'S DISEASE
EDITED BY : Sandeep Kumar Barodia, Krishnan Prabhakaran, Smitha Karunakaran, Vikas Mishra and Victor Tapias, PUBLISHED IN : Frontiers in Neuroscience
#### Frontiers eBook Copyright Statement
The copyright in the text of individual articles in this eBook is the property of their respective authors or their respective institutions or funders. The copyright in graphics and images within each article may be subject to copyright of other parties. In both cases this is subject to a license granted to Frontiers. The compilation of articles constituting this eBook is the property of Frontiers.
Each article within this eBook, and the eBook itself, are published under the most recent version of the Creative Commons CC-BY licence. The version current at the date of publication of this eBook is CC-BY 4.0. If the CC-BY licence is updated, the licence granted by Frontiers is automatically updated to the new version.
When exercising any right under the CC-BY licence, Frontiers must be attributed as the original publisher of the article or eBook, as applicable.
Authors have the responsibility of ensuring that any graphics or other materials which are the property of others may be included in the CC-BY licence, but this should be checked before relying on the CC-BY licence to reproduce those materials. Any copyright notices relating to those materials must be complied with.
Copyright and source acknowledgement notices may not be removed and must be displayed in any copy, derivative work or partial copy which includes the elements in question.
All copyright, and all rights therein, are protected by national and international copyright laws. The above represents a summary only. For further information please read Frontiers' Conditions for Website Use and Copyright Statement, and the applicable CC-BY licence.
ISSN 1664-8714 ISBN 978-2-88963-334-0 DOI 10.3389/978-2-88963-334-0
#### About Frontiers
Frontiers is more than just an open-access publisher of scholarly articles: it is a pioneering approach to the world of academia, radically improving the way scholarly research is managed. The grand vision of Frontiers is a world where all people have an equal opportunity to seek, share and generate knowledge. Frontiers provides immediate and permanent online open access to all its publications, but this alone is not enough to realize our grand goals.
#### Frontiers Journal Series
The Frontiers Journal Series is a multi-tier and interdisciplinary set of open-access, online journals, promising a paradigm shift from the current review, selection and dissemination processes in academic publishing. All Frontiers journals are driven by researchers for researchers; therefore, they constitute a service to the scholarly community. At the same time, the Frontiers Journal Series operates on a revolutionary invention, the tiered publishing system, initially addressing specific communities of scholars, and gradually climbing up to broader public understanding, thus serving the interests of the lay society, too.
#### Dedication to Quality
Each Frontiers article is a landmark of the highest quality, thanks to genuinely collaborative interactions between authors and review editors, who include some of the world's best academicians. Research must be certified by peers before entering a stream of knowledge that may eventually reach the public - and shape society; therefore, Frontiers only applies the most rigorous and unbiased reviews.
Frontiers revolutionizes research publishing by freely delivering the most outstanding research, evaluated with no bias from both the academic and social point of view. By applying the most advanced information technologies, Frontiers is catapulting scholarly publishing into a new generation.
#### What are Frontiers Research Topics?
Frontiers Research Topics are very popular trademarks of the Frontiers Journals Series: they are collections of at least ten articles, all centered on a particular subject. With their unique mix of varied contributions from Original Research to Review Articles, Frontiers Research Topics unify the most influential researchers, the latest key findings and historical advances in a hot research area! Find out more on how to host your own Frontiers Research Topic or contribute to one as an author by contacting the Frontiers Editorial Office: [email protected]
# MITOCHONDRIA AND ENDOPLASMIC RETICULUM DYSFUNCTION IN PARKINSON'S DISEASE
Topic Editors:
Sandeep Kumar Barodia, University of Alabama at Birmingham, United States Krishnan Prabhakaran, Norfolk State University, United States Smitha Karunakaran, Centre for Brain Research, Indian Institute of Science, India Vikas Mishra, Babasaheb Bhimrao Ambedkar University, India Victor Tapias, Feil Family Brain & Mind Research Institute, Weill Cornell Medicine, United States
Several pathogenic mechanisms are involved in the pathogenesis of Parkinson's Disease (PD), a neurodegenerative disease characterized by the loss of substantial nigra (SN) dopamine (DA) neurons. Alterations in calcium (Ca2+) homeostasis, cellular proteostasis, axonal transport, mitochondrial function, and neuroinflammation are linked to PD. However, research involving inter-organelle communication and their significance as precise mechanisms underlying neuronal death in PD remain to be elucidated.
Evidence showed that perturbations in the mitochondria-endoplasmic reticulum (ER) network play an important role in the pathogenesis of PD. Alterations in the mitochondria-ER interface have been reported in PARK2 knockout mice and patients harboring PARK2 mutations. Enhanced parkin levels maintain mitochondria-ER cross-talk and assure regulated Ca2+ transfer to sustain cell bioenergetics. Several familial PD-related proteins, including Parkin and PINK1, may lead to modifications in the mitochondria-ER signaling. Interestingly, mitochondria-ER tethering suppresses mitophagy and parkin/PINK1-dependent mechanism regulates the destruction of mitochondria-ER contact sites by catalyzing a rapid burst of Mfn2 phospho-ubiquitination to trigger p97-dependent disassembly of Mfn2 complexes from the outer mitochondrial membrane. Mitofusin-mediated ER stress elicited neurodegeneration in Pink1/Parkin models of PD. α-Synuclein, a presynaptic protein, can bind to the ER-mitochondria tethering protein vesicle-associated membrane protein-associated protein B (VAPB) to disrupt Ca2+ homeostasis and mitochondrial ATP production. It has been reported that ER stress and mitochondrial cell death pathways might mediate A53T mutant α-synuclein-induced toxicity.
Mitochondria-ER signaling mechanism is poorly characterized in neurons and its association in neuronal pathophysiology remains uncertain. The presence of mitochondria-ER contacts in neurons, preferentially at synapses, suggests a potential role in regulating synaptic activity. Alterations in mitochondria-ER associations are expected to be potentially detrimental to neurons, especially to SN DA neurons. Compounds from an unbiased chemical screen reverse both ER-to-Golgi trafficking defects and associated mitochondrial dysfunction in different PD models. In addition, a dibenzoylmethane derivative protects DA neurons against ER stress. Thus, mitochondria-ER signaling may represent a possible upstream drug target as potential therapeutic strategy for PD.
In this Research Topic, we bring together knowledge that emphasizes the importance of mitochondria-ER communication and its impact to further dissect the pathogenic mechanisms in PD.
Citation: Barodia, S. K., Prabhakaran, K., Karunakaran, S., Mishra, V., Tapias, V., eds. (2020). Mitochondria and Endoplasmic Reticulum Dysfunction in Parkinson's Disease. Lausanne: Frontiers Media SA. doi: 10.3389/978-2-88963-334-0
# Table of Contents
*05 Editorial: Mitochondria and Endoplasmic Reticulum Dysfunction in Parkinson's Disease*
Sandeep Kumar Barodia, Krishnan Prabhakaran, Smitha Karunakaran, Vikas Mishra and Victor Tapias
#### *62 PERK-Mediated Unfolded Protein Response Activation and Oxidative Stress in PARK20 Fibroblasts*
Giuseppina Amodio, Ornella Moltedo, Dominga Fasano, Lucrezia Zerillo, Marco Oliveti, Paola Di Pietro, Raffaella Faraonio, Paolo Barone, Maria Teresa Pellecchia, Anna De Rosa, Giuseppe De Michele, Elena Polishchuk, Roman Polishchuk, Vincenzo Bonifati, Lucio Nitsch, Giovanna Maria Pierantoni, Maurizio Renna, Chiara Criscuolo, Simona Paladino and Paolo Remondelli
Ingrid González-Casacuberta, Diana Luz Juárez-Flores, Constanza Morén and Gloria Garrabou
*108 Intracellular and Intercellular Mitochondrial Dynamics in Parkinson's Disease*
Dario Valdinocci, Rui F. Simões, Jaromira Kovarova, Teresa Cunha-Oliveira, Jiri Neuzil and Dean L. Pountney
*116 The Impairments of* a*-Synuclein and Mechanistic Target of Rapamycin in Rotenone-Induced SH-SY5Y Cells and Mice Model of Parkinson's Disease* Mahesh Ramalingam, Yu-Jin Huh and Yun-Il Lee
# Editorial: Mitochondria and Endoplasmic Reticulum Dysfunction in Parkinson's Disease
Sandeep Kumar Barodia<sup>1</sup> \*, Krishnan Prabhakaran<sup>2</sup> , Smitha Karunakaran<sup>3</sup> , Vikas Mishra<sup>4</sup> and Victor Tapias <sup>5</sup> \*
*<sup>1</sup> Center for Neurodegeneration and Experimental Therapeutics, Birmingham, AL, United States, <sup>2</sup> Department of Biology, Norfolk State University, Norfolk, VA, United States, <sup>3</sup> Centre for Brain Research, Indian Institute of Science, Bangalore, India, <sup>4</sup> Department of Pharmaceutical Sciences, Basanaheb Bhirao Ambedkar University, Lucknow, India, <sup>5</sup> Feil Family Brain and Mind Research Institute, Weill Cornell Medicine, New York, NY, United States*
Keywords: Parkinson's disease, mitochondria, endoplasmic reticulum, α-synuclein, PINK1, parkin, oxidative stress, MAMs
**Editorial on the Research Topic**
#### **Mitochondria and Endoplasmic Reticulum Dysfunction in Parkinson's Disease**
Endoplasmic reticulum (ER) and mitochondria are distributed in close communication via a dynamic ER-calcium (Ca2+) mitochondria interconnection and regulate a plethora of vital cellular functions, including Ca2<sup>+</sup> homeostasis, mitochondrial transport and dynamics, bioenergetics, ER stress, apoptotic signaling, and inflammation (Erpapazoglou et al., 2017). Alteration in the ER-mitochondria communication adversely affects overall physiology of the cell (Gómez-Suaga et al., 2018). ER-mitochondria communication is also involved in lipid transport, suggesting that lipidomic approach may be useful to study the potential mechanisms leading to impaired neuropeptidergic signaling (Valadas et al., 2018). Mitochondria-associated membranes (MAMs) are defined as specialized subdomains connecting ER and mitochondria in order to regulate physiological functions, maintain Ca2<sup>+</sup> signaling and other vital cellular processes (Rodríguez-Arribas et al., 2017). Neurons are highly dependent on MAMs to exchange metabolites and signaling molecules between ER and mitochondria, suggesting that altered function of MAMs due to toxin insults such as rotenone and manganese could play a crucial role in the pathogenesis of neurodegenerative diseases, including Parkinson's disease (PD) (Krols et al., 2016; Harischandra et al.; Ramalingam et al.; Valdinocci et al.). Modifications in the communication between ER and mitochondria cause a reduction in mitochondrial Ca2<sup>+</sup> homeostasis in several animal models of neurodegeneration, such as PD, an age-dependent neurodegenerative disorder characterized by the progressive loss of dopamine (DA)-producing neurons in the substantia nigra (Paillusson et al., 2016; Lee et al., 2018). Several cellular mechanisms have been identified to be involved in the DAergic neuronal death, including mitochondrial dysfunction, impaired bioenergetics, oxidative stress, autophagy and impaired intracellular Ca2<sup>+</sup> homeostasis in patientderived cell models of PD (González-Casacuberta et al.; Segura-Aguilar). However, mechanisms underlying how organelle crosstalk (especially between mitochondria and ER) could affect the progression of pathogenesis in PD still remain unknown. ER stress activates unfolded protein response through the upregulation of the ER chaperone GRP78 and caspases as well as evokes Ca2<sup>+</sup> flux that induces mitochondrial dysfunction and associated loss of DA neurons (Arduíno et al., 2009; Baek et al.). Interestingly, increased ROS production through PERK/eIF2α/ATF4/CHOP pathway of UPR and concomitant alteration of the mitochondrial network morphology have been reported in PARK20 fibroblasts (Amodio et al.). Emerging evidence supporting significance of altered ER–mitochondria communication suggests that damaged ER–mitochondria signaling could be a potential therapeutic strategy to treat neurodegenerative diseases.
#### Edited and reviewed by:
*Wendy Noble, King's College London, United Kingdom*
#### \*Correspondence:
*Sandeep Kumar Barodia [email protected] Victor Tapias [email protected]*
#### Specialty section:
*This article was submitted to Neurodegeneration, a section of the journal Frontiers in Neuroscience*
Received: *05 September 2019* Accepted: *16 October 2019* Published: *08 November 2019*
#### Citation:
*Barodia SK, Prabhakaran K, Karunakaran S, Mishra V and Tapias V (2019) Editorial: Mitochondria and Endoplasmic Reticulum Dysfunction in Parkinson's Disease. Front. Neurosci. 13:1171. doi: 10.3389/fnins.2019.01171*
The present Research Topic is an effort to showcase the significance of MAM in PD pathogenesis. Here, we discuss the recent findings in PD research with main focus on molecular and cellular mechanisms involving mitochondria and ER. Pathophysiological significance of ER-mitochondria interaction has been demonstrated in the case of PD-related genes, such as α-synuclein (α-syn) (Guardia-Laguarta et al., 2014), DJ-1 (Ottolini et al., 2013), PINK1 (Celardo et al., 2016; Gelmetti et al., 2017), and Parkin (Van Laar et al., 2015; Celardo et al., 2016; Gautier et al., 2016; Gelmetti et al., 2017; Zheng et al., 2017). Several clinical cases with diagnosed PD show a welldefined Lewy body pathology (Cookson et al., 2008), which are composed of α-syn. Protein aggregation and imbalanced cellular proteostasis are key factors leading to accumulation of misfolded α-syn (Lehtonen et al.). Within neurons, α-syn is diversely localized to cytosolic and membrane compartments including synaptic vesicles, mitochondria and the ER (Guardia-Laguarta et al., 2015; Colla). Membrane localization of α-syn is well-targeted to lipid rafts (detergent-resistant membranes) that are enriched in cholesterol and acidic phospholipids (Fortin et al., 2004). Interestingly, a subpopulation of α-syn is shown to be enriched in MAM fraction in immortalized cell lines and in the mouse and human brain (Poston et al., 2013; Guardia-Laguarta et al., 2014; Paillusson et al., 2016). Certainly, identification of the A53T mutation in the gene encoding for α-syn (SNCA) provides us better understanding of both the genetics and the neuropathology of PD (Polymeropoulos et al., 1997). It has been demonstrated that A53T mutant showed a decreased association with MAM and an elevated mitochondrial fragmentation, as compared to wild-type α-syn (Guardia-Laguarta et al., 2014). Moreover, overexpression of either wild-type or mutant α-syn decreases ER–mitochondria contacts (Paillusson et al., 2016). Thus, substantial accumulation of α-syn aggregates could be linked to the loss of function of this protein at the MAMs. Interestingly, subcellular localization of α-syn to MAM could be related to both normal and pathological states (Guardia-Laguarta et al., 2014, 2015). A recent study demonstrated that α-syn binds to VAPB (an ER-mitochondria tethering protein) to disrupt Ca2<sup>+</sup> homeostasis and mitochondrial ATP production (Paillusson et al., 2016).
PINK1/Parkin-mediated mitophagy could be an underlying mechanism of nigral DA neuron death in PD (Thomas et al., 2011; Kane et al., 2014; Barodia et al., 2017). ER-mitochondria contact sites were shown to constitute the initiation sites for this process (Yang and Yang, 2013). During mitophagy, PINK1 and BECN1 re-localize at MAM, which induces ER-mitochondria tethering and autophagosome formation (Gelmetti et al., 2017).
#### REFERENCES
Parkin expression was significantly increased in the MAM fraction of neurons following glutamate excitotoxicity (Van Laar et al., 2015), which also ubiquitylated several proteins of the ER-mitochondria interface including Mfn2, VDACs and Miro (Sarraf et al., 2013; Pickrell and Youle, 2015). Parkin may regulate ER-mitochondria communication via Mfn2 (Basso et al., 2018). Mitochondrial and ER stress results in an upregulation of Parkin levels via ATF4 (Bouman et al., 2011). ER-mitochondria communication was reported to be increased in fibroblasts from patients with PARK2 or PARK6 mutations compared to control group (Celardo et al., 2016; Gautier et al., 2016). This alteration was associated with higher mitochondrial Ca2<sup>+</sup> absorption, upon IP3R stimulation. Similar structural changes were observed in MEFs from PARK2 knock-out mice (Gautier et al., 2016). Parkin has recently been reported to co-regulate ERmitochondria communication together with the transcription factor peroxisome proliferator activated receptor g coactivator 1a (PGC-1α), a key modulator of mitochondrial biogenesis (Zheng et al., 2017). ER–mitochondria associations have also been linked to the formation of the inflammasome. Cellular stress in neurodegenerative diseases are detected by the innate immune system through pattern recognition receptors (Paillusson et al., 2016). Reactive oxygen species (ROS) from mitochondria are one signal for activation of the NLRP3 inflammasome (Abais et al., 2015). Elevated ROS generation led to NLRP relocation to MAM, which may provide a mechanism whereby NLRP senses damage mitochondria to activate the inflammasome (Zhou et al., 2011). Due to the importance of MAMs in understanding the fundamental mechanisms of PD pathogenesis and their potential use as a therapeutic approach, further research is needed to investigate on the communications between the ER and mitochondria.
#### AUTHOR CONTRIBUTIONS
SB collected the relevant references and wrote the manuscript. VT, KP, SK, and VM edited the manuscript and provided thorough reviews on the manuscript.
#### ACKNOWLEDGMENTS
We would like to thank you the authors who have contributed to this Research Topic and the dedicated reviewers who helped us reach the highest quality standards. We gratefully acknowledge the valuable input of the Frontiers editorial team members for their support in editing and publishing the scientific content.
Neurochem. Int. 55, 341–348. doi: 10.1016/j.neuint.2009. 04.004
**Conflict of Interest:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Copyright © 2019 Barodia, Prabhakaran, Karunakaran, Mishra and Tapias. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
# On the Role of Aminochrome in Mitochondrial Dysfunction and Endoplasmic Reticulum Stress in Parkinson's Disease
Juan Segura-Aguilar\*
Molecular and Clinical Pharmacology, Faculty of Medicine, ICBM, University of Chile, Santiago, Chile
Keywords: mitochondrial dysfunction, dopamine, aminochrome, endoplasmic reticulum stress, Parkinson's disease, glutathione-S-transferase (GST), DT-diaphorase neurodegeneration
#### Edited by:
Victor Tapias, Feil Family Brain & Mind Research Institute, Weill Cornell Medicine, United States
#### Reviewed by:
Éva M. SzegÅ, Eötvös Loránd University, Hungary Catherine Brenner, INSERM U1180 Signalisation et Physiopathologie Cardiovasculaire, France
\*Correspondence:
Juan Segura-Aguilar [email protected]
#### Specialty section:
This article was submitted to Neurodegeneration, a section of the journal Frontiers in Neuroscience
Received: 21 November 2018 Accepted: 07 March 2019 Published: 29 March 2019
#### Citation:
Segura-Aguilar J (2019) On the Role of Aminochrome in Mitochondrial Dysfunction and Endoplasmic Reticulum Stress in Parkinson's Disease. Front. Neurosci. 13:271. doi: 10.3389/fnins.2019.00271 The identity of what triggers the loss of dopaminergic neurons containing neuromelanin in Parkinson's disease (PD) is still unknown. Fifty years since its introduction in PD therapy, L-dopa is still the gold-standard drug despite severe side effects observed after 4 to 6 years of being treated with it. There are no new therapies that can halt or slow down the progression of the disease and much of the research efforts in this context have been destined to treat L-dopa-induced dyskinesia. There is huge concern about the difficulties that have been observed in the translation of successful preclinical results into clinical studies and new therapies in PD. The discovery of genes associated with familiar forms of PD has made an enormous input into basic research, which seeks to understand the degenerative process resulting in the loss of dopaminergic neurons in the nigrostriatal system. Several mechanisms have been suggested to be involved in the degeneration of nigrostriatal neurons in PD, including mitochondrial dysfunction, endoplasmic reticulum stress, lysosomal and proteasomal protein degradation dysfunction, the formation of neurotoxic alpha-synuclein (SNCA) oligomers, neuroinflammation, and oxidative stress.
#### MITOCHONDRIAL DYSFUNCTION
The brain is completely dependent on chemical energy (ATP) in order to perform the release of neurotransmitters such as dopamine. Therefore, the existence of functional mitochondria is essential to the performed role of a dopaminergic neuron, i.e., to release dopamine. Postmortem brains with PD presented a deficiency in Complex I activity (Shapira et al., 1990; Esteves et al., 2011). Reduced Complex I activity in platelet mitochondria, purified from patients with idiopathic PD, has been observed (Esteves et al., 2011). CHCHD2 mutation in PD patient fibroblasts reduces oxidative phosphorylation in Complexes I and IV and induces fragmentation of the mitochondrial reticular morphology (Lee et al., 2018). A meta-analysis supports the deficit in Complexes I and IV in the case of peripheral blood, the frontal cortex, the cerebellum and the substantia nigra in PD (Holper et al., 2018). Analysis of mitochondria morphology in PD samples compared to controls revealed a significant decrease in the number of healthy mitochondria per cell. Several genes associated with familial forms of PD (PINK-1, DJ-1, Parkin, HTRA2) are linked to mitochondrial impairment (Larsen et al., 2018). Parkinson's disease, associated with vacuolar protein sorting 35 mutation, affects Complex I activity (Zhou et al., 2017). PINK1 and DJ-1 mutation induce energetic inefficiency (Lopez-Fabuel et al., 2017). SNCA induces mitochondrial dysfunction (Devi et al., 2008; Chinta et al., 2010; Nakamura et al., 2011; Martínez et al., 2018).
#### ENDOPLASMIC RETICULUM STRESS
Endoplasmic reticulum is involved in secretory protein translocation and the quality control of secretory protein folding. Misfolded or unfolded proteins in the lumen accumulate under endoplasmic reticulum stress, which causing an integrated adaptive response identified as the unfolded protein response (UPR), which seeks to restore proteostasis within the secretory pathway (Cabral-Miranda and Hetz, 2018).
The UPR activation markers, phosphorylated eukaryotic initiation factor 2alpha and phosphorylated pancreatic endoplasmic reticulum kinase, were detected in dopaminergic neurons containing neuromelanin in the substantia nigra of PD patients. Interestingly, phosphorylated pancreatic endoplasmic reticulum kinase was colocalized with an increased level of SNCA (Hoozemans et al., 2007). Neuropathological analysis of PD postmortem brain tissue revealed that pIRE1α is expressed within neurons containing elevated levels of α-synuclein or Lewy bodies (Heman-Ackah et al., 2017). SNCA triggers endoplasmic reticulum stress via the protein kinase RNA-like endoplasmic reticulum kinase/eukaryotic translation initiation factor 2α signaling pathway (Liu et al., 2018). N370S mutation and β-glucocerebrosidase-1 retention within the endoplasmic reticulum induce endoplasmic reticulum stress activation, triggering UPR and Golgi apparatus fragmentation (García-Sanz et al., 2017). It has been reported that endoplasmic reticulum stress activates the chaperone-mediated autophagy pathway via an EIF2AK3/PERK-MAP2K4/MKK4-MAPK14/p38-dependent manner (Li et al., 2018).
#### DOPAMINE OXIDATION AND PARKINSON'S DISEASE
One of the most characteristic features of the pathology of PD, which results in the onset of motor symptoms, is the massive loss of dopaminergic neurons containing neuromelanin in the nigrostriatal system. As mentioned before, several mechanisms, including mitochondrial dysfunction and endoplasmic reticulum stress, have been proposed as being involved in the degeneration of the nigrostriatal neurons in PD, but the question concerns what triggers these mechanisms in dopaminergic neurons containing neuromelanin. Many times, it has been suggested that the involvement of exogenous neurotoxins triggers these mechanisms, but the severe Parkinsonism induced by MPTP in just 3 days in drug addicts who used synthetic drugs contaminated with this compound undermines this idea (Williams, 1986). The rate of the degenerative process in PD takes years (Braak et al., 2004). The extremely slow degeneration of the nigrostriatal neurons and slow progression of the disease challenge the possible role of exogenous neurotoxins in the loss of dopaminergic neurons containing neuromelanin, suggesting that some endogenous neurotoxin must trigger these mechanisms. A neurotoxic event, triggered by an endogenous neurotoxin, will affect a single neuron without propagative effects, which explains the extremely slow rate of this degenerative process in PD. Among possible endogenous neurotoxins are the neurotoxic SNCA oligomers. However, the prion-like hypothesis of SNCA in PD pathogenesis is based on the propagation (neuronto-neuron transfer) of neurotoxic SNCA oligomers (Brundin and Melki, 2017). According to this prion-like hypothesis, a relatively rapid process is expected, in contrasting with what happens in PD, which takes years. In addition, what triggers the formation of neurotoxic SNCA oligomers inside the dopaminergic neurons containing neuromelanin? Braak stage hypothesis use the intraneuronal inclusion bodies to follow the development of Parkinson's disease where SNCA is one of the aggregated proteins (Braak et al., 2004). What induces SNCA aggregation in other brain region involved in non-motor symptoms remains unclear. A possible explanation is that an endogenous neurotoxin is formed inside dopaminergic neurons containing neuromelanin during dopamine oxidation. The formation of the pigment called neuromelanin in these neurons is the result of dopamine oxidation into ortho(o)-quinones, which is a pathway that involves the formation of three o-quinones in a sequential manner (dopamine −→dopamine o-quinone −→ aminochrome−→ 5,6- indolequinone−→ neuromelanin).
Dopamine o-quinone is able to form adducts with proteins, such as ubiquitin carboxy-terminal hydrolase L1 (UCHL-1) and Parkinsonism-associated deglycase (DJ-1, PARK7), as well as ubiquinol-cytochrome c reductase core protein 1, glucoseregulated protein 75/mitochondrial HSP70/mortalin, mitofilin, mitochondrial creatine kinase and glutathione peroxidase-4, and a human dopamine transporter (Whitehead et al., 2001; Van Laar et al., 2009; Hauser et al., 2013). Incubation of purified tyrosine hydroxylase with dopamine and tyrosinase also forms adducts with dopamine (Xu et al., 1998). Dopamine o-quinone induces mitochondrial dysfunction (Berman and Hastings, 1999). Exposure of cells to dopamine induced the formation of dopamine adducts with parkin (LaVoie et al., 2005), but the identity of the o-quinone involved in this reaction (dopamine o-quinone or aminochrome) is not clear. Dopamine o-quinone is completely unstable at physiological pH and cyclizes immediately into aminochrome; thus, the question concerns whether dopamine o-quinone has the opportunity to form adducts with parkin in the cell cytosol overcrowded with other proteins, molecules and organelles.
Aminochrome has been reported to be neurotoxic on account of inducing mitochondrial dysfunction, endoplasmic reticulum stress, autophagy dysfunction, proteasomal dysfunction, oxidative stress, neuroinflammation, the disruption of the cytoskeleton architecture and the formation of neurotoxic SNCA oligomers (Arriagada et al., 2004; Zafar et al., 2006; Fuentes et al., 2007; Zhou and Lim, 2009; Paris et al., 2010, 2011; Aguirre et al., 2012; Muñoz et al., 2012, 2015; Huenchuguala et al., 2014, 2017; Xiong et al., 2014; Briceño et al., 2016; Santos et al., 2017; de Araújo et al., 2018; Segura-Aguilar and Huenchuguala, 2018) (**Figure 1**).
5,6-Indolequinone, the precursor of neuromelanin, is able to form adducts with SNCA (Bisaglia et al., 2010). Dopaminochrome has also been reported to form adducts with SNCA (Norris et al., 2005) and to be neurotoxic in cell cultures (Linsenbardt et al., 2009, 2012). The unilateral injection of dopaminochrome induced degeneration of the
dopaminergic neurons within the substantia nigra (Touchette et al., 2015). However, the structure of dopaminochrome has not been determined by NMR; nor do we know the nature of this structure. The dopaminochrome structure is different to the aminochrome structure because dopaminochrome has an absorption maximum of 303 and 479 nm (Ochs et al., 2005), while aminochrome has an absorption maximum of 280 and 475 nm and its structure has been confirmed by NMR (Paris et al., 2010).
#### AMINOCHROME AND PARKINSON'S DISEASE
Dopamine oxidation into neuromelanin is a normal and harmless pathway because neuromelanin accumulates with age, with dopaminergic neurons containing neuromelanin remaining intact in the substantia nigra of healthy seniors (Zecca et al., 2002). Aminochrome is the most stable and studied oquinone formed during dopamine oxidation into neuromelanin. Paradoxically, aminochrome under certain conditions can be neurotoxic as a result of inducing mitochondrial dysfunction (Arriagada et al., 2004; Paris et al., 2011; Aguirre et al., 2012; Huenchuguala et al., 2017; Segura-Aguilar and Huenchuguala, 2018), endoplasmic reticulum stress (Xiong et al., 2014), the formation of neurotoxic SNCA oligomers (Muñoz et al., 2015; Muñoz and Segura-Aguilar, 2017), proteasome dysfunction (Zafar et al., 2006; Zhou and Lim, 2009), autophagy dysfunction (Muñoz et al., 2012; Huenchuguala et al., 2014), lysosome dysfunction (Meléndez et al., 2018), neuroinflammation (Santos et al., 2017; de Araújo et al., 2018), cytoskeleton architecture disruption (Paris et al., 2010; Briceño et al., 2016) and oxidative stress (Arriagada et al., 2004). Aminochrome in vivo induces neuronal dysfunction as a consequence of mitochondrial dysfunction, decreased axonal transport resulting in a significant decrease in the number of synaptic monoaminergic vesicles, reduced dopamine release accompanied by an increase in GABA levels, and a dramatic change in the neurons' morphology characterized as cell shrinkage (Herrera et al., 2016). The explanation as to why dopamine oxidation into neuromelanin is not a harmful pathway, despite the formation of potential neurotoxic o-quinones, is because the existence of two enzymes [DT-diaphorase and glutathione transferase M2-2 (GSTM2)], which are able to prevent aminochrome neurotoxicity. DT-diaphorase is expressed in dopaminergic neurons and astrocytes and catalyzes the twoelectron reduction of aminochrome into leukoaminochrome, preventing aminochrome one-electron reduction into the leukoaminochrome o-semiquinone radical, catalyzed by flavoenzymes that transfer one electron and use NADH or NADPH. DT-diaphorase prevents aminochrome-induced cell death (Lozano et al., 2010), mitochondrial dysfunction (Arriagada et al., 2004; Paris et al., 2011; Muñoz et al., 2012), cytoskeleton architecture disruption (Paris et al., 2010), lysosomal dysfunction (Meléndez et al., 2018), the formation of neurotoxic SNCA oligomers (Muñoz et al., 2015; Muñoz and Segura-Aguilar, 2017), oxidative stress (Arriagada et al., 2004); dopaminergic neurons' degeneration in vivo (Herrera-Soto et al., 2017) and astrocytes dell death (Huenchuguala et al., 2016). GSTM2 catalyzes the GSH conjugation of aminochrome into 4-S-glutathionyl-5,6-dihydroxyindoline, which is resistant to biological oxidizing agents such as oxygen, hydrogen peroxide, and superoxide (Segura-Aguilar et al., 1997). GSTM2 also catalyzes the GSH conjugation of dopamine o-quinone into 5-glutathionyl-dopamine (Dagnino-Subiabre et al., 2000), which degrades into 5-cysteinyl-dopamine. Interestingly, 5-cysteinyldopamine is a stable metabolite that can be eliminated from the cells. 5-Cysteinyl-dopamine has been found in substantia nigra, caudate nucleus, putamen, globus pallidus, neuromelanin, and the cerebrospinal fluid of PD patients (Rosengren et al., 1985; Carstam et al., 1991; Cheng et al., 1996). GSTM2 prevents aminochrome-induced cell death, mitochondrial dysfunction, autophagy, and lysosome dysfunction (Huenchuguala et al., 2014; Segura-Aguilar, 2017a; Segura-Aguilar and Huenchuguala, 2018). The GSH conjugation of aminochrome prevents the formation of neurotoxic SNCA oligomers by generating nontoxic SNCA oligomers (Huenchuguala et al., 2018). GSTM2 is expressed in human astrocytes and it has been reported
#### REFERENCES
that astrocytes secrete GSTM2, while dopaminergic neurons are able to internalize this enzyme into the cytosol, protecting these neurons against aminochrome-induced neurotoxicity (Cuevas et al., 2015; Segura-Aguilar, 2015, 2017b).
Mitochondrial dysfunction and endoplasmic reticulum stress are two very important mechanisms involved in the loss of dopaminergic neurons containing neuromelanin in the nigrostriatal neurons in idiopathic PD. However, the question concerns the common denominator in these mechanisms: i.e., what triggers these mechanisms in dopaminergic neurons containing neuromelanin in the nigrostriatal system? We propose that aminochrome is the endogenous neurotoxin that triggers mitochondrial dysfunction and endoplasmic reticulum stress because aminochrome is formed inside dopaminergic neurons of the nigrostriatal system. In addition, aminochrome also triggers other mechanisms involved in the loss of dopaminergic neurons in the nigrostriatal system, such as the formation of neurotoxic SNCA oligomers, oxidative stress, neuroinflammation, and proteasomal and lysosomal protein degradation dysfunction.
#### AUTHOR CONTRIBUTIONS
The author confirms being the sole contributor of this work and has approved it for publication.
### FUNDING
FONDECYT 1170033.
of alpha-synuclein impair complex I in human dopaminergic neuronal cultures and Parkinson disease brain. J. Biol. Chem. 283, 9089–9100. doi: 10.1074/jbc.M710012200
in the mesencephalic cell line, MN9D. J. Neurochem. 122, 175–184. doi: 10.1111/j.1471-4159.2012.07756.x
**Conflict of Interest Statement:** The author declares that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Copyright © 2019 Segura-Aguilar. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
# Dysfunction of Cellular Proteostasis in Parkinson's Disease
Šárka Lehtonen1,2, Tuuli-Maria Sonninen<sup>1</sup> , Sara Wojciechowski<sup>1</sup> , Gundars Goldsteins<sup>1</sup> and Jari Koistinaho1,2 \*
<sup>1</sup> A.I. Virtanen Institute for Molecular Sciences, University of Eastern Finland, Kuopio, Finland, <sup>2</sup> Neuroscience Center, Helsinki Institute of Life Science, University of Helsinki, Helsinki, Finland
Despite decades of research, current therapeutic interventions for Parkinson's disease (PD) are insufficient as they fail to modify disease progression by ameliorating the underlying pathology. Cellular proteostasis (protein homeostasis) is an essential factor in maintaining a persistent environment for neuronal activity. Proteostasis is ensured by mechanisms including regulation of protein translation, chaperone-assisted protein folding and protein degradation pathways. It is generally accepted that deficits in proteostasis are linked to various neurodegenerative diseases including PD. While the proteasome fails to degrade large protein aggregates, particularly alpha-synuclein (α-SYN) in PD, drug-induced activation of autophagy can efficiently remove aggregates and prevent degeneration of dopaminergic (DA) neurons. Therefore, maintenance of these mechanisms is essential to preserve all cellular functions relying on a correctly folded proteome. The correlations between endoplasmic reticulum (ER) stress and the unfolded protein response (UPR) that aims to restore proteostasis within the secretory pathway are well-established. However, while mild insults increase the activity of chaperones, prolonged cell stress, or insufficient adaptive response causes cell death. Modulating the activity of molecular chaperones, such as protein disulfide isomerase which assists refolding and contributes to the removal of unfolded proteins, and their associated pathways may offer a new approach for disease-modifying treatment. Here, we summarize some of the key concepts and emerging ideas on the relation of protein aggregation and imbalanced proteostasis with an emphasis on PD as our area of main expertise. Furthermore, we discuss recent insights into the strategies for reducing the toxic effects of protein unfolding in PD by targeting the ER UPR pathway.
Keywords: proteostasis, alpha-synuclein, refolding, ER stress, UPR, protein disulfide isomerase
### INTRODUCTION
In Parkinson's disease (PD), the loss of dopaminergic (DA) neurons in the substantia nigra pars compacta (SNpc) and subsequent loss of dopamine in the striatum leads to typical motor impairments in PD, such as bradykinesia, rigidity, rest tremor, and postural instability. There are various non-motor symptoms also associated with PD including anosmia, gastrointestinal motility issues, sleep disturbances, sympathetic denervation, anxiety, and depression. These non-motor symptoms generally precede the motor impairments by years (Kalia and Lang, 2015). The presence of Lewy bodies (LBs) with an accumulation of the protein alpha-synuclein (α-SYN) is one of the
#### Edited by:
Sandeep Kumar Barodia, University of Alabama at Birmingham, United States
#### Reviewed by:
Kalle Gehring, McGill University, Canada Mahesh Narayan, The University of Texas at El Paso, United States
> \*Correspondence: Jari Koistinaho [email protected]
#### Specialty section:
This article was submitted to Neurodegeneration, a section of the journal Frontiers in Neuroscience
Received: 01 March 2019 Accepted: 23 April 2019 Published: 10 May 2019
#### Citation:
Lehtonen Š, Sonninen T-M, Wojciechowski S, Goldsteins G and Koistinaho J (2019) Dysfunction of Cellular Proteostasis in Parkinson's Disease. Front. Neurosci. 13:457. doi: 10.3389/fnins.2019.00457
**14**
pathological hallmarks in PD (Kalia and Lang, 2015; Sveinbjornsdottir, 2016). There is not yet a cure, although, treatments are available to relieve symptoms. Approximately 20 PD-associated genes have been identified to date even though most cases are late onset and sporadic with no evidence for inheritance or genetic cause (Klein and Westenberger, 2012). The phenotypes of both the sporadic and familial forms are essentially indistinguishable, implying that they might share common underlying mechanisms. Moreover, many similarities including protein misfolding and aggregation are also commonly seen in other neurodegenerative diseases. While the exact role of protein aggregation in disease pathology is still under debate, discovering these similarities offers hope for therapeutic advances that could affect many diseases simultaneously. In this review, we summarize recent progress in the studies on the mechanism of endoplasmic reticulum (ER) stressinduced unfolded protein response (UPR) in PD, how protein aggregation relates to imbalanced proteostasis and how to remedy the toxic effects of protein unfolding in PD by targeting the ER UPR pathway.
### DESCRIPTION OF CELLULAR PROTEOSTASIS DEFICITS IN PD
### Physiological Role of α-SYN and Aggregation
α-SYN is a small (14 kDa) protein that is highly expressed in neurons but can also be found in peripheral tissues and blood (Witt, 2013; Malek et al., 2014). A recent report also demonstrated its expression in astrocytes (di Domenico et al., 2019). The physiological function of α-SYN remains mostly undefined (Devine et al., 2011; Liu et al., 2012; Kalia and Kalia, 2015), nevertheless, the involvement in synaptic maintenance, mitochondrial homeostasis, dopamine metabolism, and chaperone activity has been studied. Typically, α-SYN is a monomer with three structural regions (Villar-Piqué et al., 2016). The N-terminal domain (1–60) contains a multirepeated consensus sequence (KTKEGV) and is responsible for the membrane-binding capacity. The central domain (61–95) is known as the non-amyloid-beta component and contains a highly hydrophobic motif which is involved in α-SYN aggregation. The C-terminal domain's (96–140) proline residues have been found to be acidic. The exact native structure of α-SYN is not completely established, but several studies have described it as a soluble protein with a disordered monomeric structure (Binolfi et al., 2012; Fauvet et al., 2012; Waudby et al., 2013). In addition, soluble tetramers have been identified (Bartels et al., 2011), but the physiologically relevant structure of α-SYN may differ depending on the cellular location and environment. The non-amyloid-beta domain of α-SYN is prone to aggregate, but in its native structure, it appears to be protected by the N- and C-termini (Bertoncini et al., 2005). Changes in environment, mutations and/or post-translational modifications (PTMs) may disrupt the native conformation of α-SYN and induce misfolding and aggregation.
Initially, α-SYN was identified in the nucleus, but this is still in dispute (Huang et al., 2011). It has been proposed that the nuclear protein TRIM28 regulates its translocation into the nucleus and α-SYN may play a role in transcription regulation and histone acetylation (Kontopoulos et al., 2006; Rousseaux et al., 2016). Several studies have shown that PD associated mutations, PTMs and oxidative stress can increase the nuclear localization of α-SYN (Kontopoulos et al., 2006; Xu et al., 2006; Schell et al., 2009; Gonçalves and Outeiro, 2013; Fares et al., 2014). In addition, animal and cellular models and patient studies have shown altered activation of transcription factors upon α-SYN translocation. These include decreased activation of the mitochondrial biogenesis factor PGC-1α, reduced activation of the autophagy-lysosomal pathway (ALP) transcription factor EB (TFEB), and increased activation of calcineurin and subsequent nuclear translocation of nuclear factor of activated T cells (Decressac et al., 2013; Ryan et al., 2013; Luo et al., 2014; Eschbach et al., 2015).
α-SYN is associated with several neurodegenerative disorders, collectively known as synucleinopathies (Wong and Krainc, 2017). α-SYN fibrils are the main component found in LB and Lewy neurites (LNs) in PD and dementia with Lewy bodies (DLBs). LBs are spherical aggregates of α-SYN found in neuronal cell bodies, while LNs are aggregate structures found in neuronal dendrites and axons. Structurally, LBs are made up of insoluble eosinophilic amyloid that is surrounded by fibrils of α-SYN which are typically ubiquitinated (Beyer et al., 2009). In sporadic PD, α-SYN accumulates in neuronal cell bodies and processes resulting in LBs and LNs, respectively. Duplication of SNCA results in late-onset autosomal dominant forms of PD and triplication results in early-onset PD (Singleton et al., 2003). This demonstrates that α-SYN levels correlate with the onset of PD. In addition, other mutations causing familial PD, like mutations in leucine-rich repeat kinase 2 (LRRK2), can develop LB pathology (Zimprich et al., 2004). When comparing the pathology of DLB, there are some similarities with PD, but the clinical symptoms are closer to Alzheimer's disease (Spillantini et al., 1998). In PD, the substantia nigra (SN) is affected, while in DLB the pathology is seen in the cortex. In addition to PD and DLB, α-SYN accumulation is present in multiple system atrophy (MSA) and pure autonomic failure (PAF). In MSA the α-SYN inclusions are present in the cytosol of oligodendrocytes. Mutations in α-SYN can cause both PD and MSA symptoms (Fanciulli and Wenning, 2015). LBs and LNs are found in the sympathetic nervous system in PAF (Arai et al., 2000). In addition to synucleinopathies, α-SYN toxicity has been associated with lysosomal storage disorders such as Gaucher's disease, a rare genetic disorder characterized by the deposition of glucocerebroside in cells of the macrophage-monocyte system (Blanz and Saftig, 2016). Mutations in GBA1, which encodes glucocerebrosidase (GCase) and causes Gaucher's disease, are the most common risk factors for PD. Some patients carrying these mutations may develop parkinsonism, a clinical syndrome characterized by movement disorders commonly seen in PD, with LB pathology.
Oligomers and fibrils are considered to be the toxic species of α-SYN, but there remains some disagreement regarding their
toxicity. Several studies have suggested that soluble oligomers are more toxic than fibrils or aggregates. For example, increased levels of soluble oligomers have been identified in α-SYN transgenic mice and in PD and DLB patient brains. Oligomeric α-SYN caused more severe DA neuron loss than fibrils in rats (Sharon et al., 2003; Winner et al., 2011). By contrast, some studies have shown fibrils to be more toxic compared to the oligomers and caused increased motor impairment, DA cell loss and synaptic impairment (Peelaerts et al., 2015).
In neurons, α-SYN is known to localize in presynaptic terminals and regulate synaptic transmission. The release of neurotransmitters requires cycles of soluble N-ethylmaleimidesensitive factor attachment protein receptor (SNARE)-complex assembly and disassembly. α-SYN has been shown to bind to SNARE protein synaptobrevin-2/vesicle-associated membrane protein 2 (VAMP2) and promote SNARE-complex assembly (Burré et al., 2010). The same study also demonstrated that triple knockout mice developed neurological impairments and had decreased SNARE assembly. Subsequently, it was described that α-SYN promotes vesicle-clustering activity, which is dependent on the interaction of α-SYN with synaptobrevin-2/VAMP2 and anionic lipids (Diao et al., 2013). These studies suggest that the major cellular function of α-SYN are interactions of α-SYN with cell membranes, and that the cytosolic state may be transient.
While α-SYN is normally localized in presynaptic terminals, the oligomers and aggregates can be found in cell bodies and neurites, as well as in other cell types, including astrocytes which indicates a widespread toxic action. The pathological effects of α-SYN can affect the function of several different organelles, including synaptic vesicles, mitochondria, lipid bilayers, cell's cytoskeleton, ER, Golgi, proteasomes, lysosomes, and nucleus. α-SYN oligomers can disrupt the SNARE complex formation, dopamine release and synaptic-vesicle motility (Choi et al., 2013; Wang et al., 2014). Increased levels of α-SYN can also decrease the synaptic-vesicle recycling-pool size and mobility leading to a disrupted neurotransmitter release (Nemani et al., 2010; Scott and Roy, 2012). It was discovered that dopamine neurotransmission can be disrupted by high levels of α-SYN. Transgenic mice overexpressing α-SYN showed a DA terminal loss, deficient release and altered synaptic-vesicle distribution (Masliah et al., 2000; Janezic et al., 2013). Moreover, the reduction in dopamine reuptake and defective dopamine transporter function has been linked to increased levels of α-SYN (Lundblad et al., 2012).
The homeostasis of mitochondria can be disrupted by α-SYN toxicity. Mice with an A53T α-SYN mutation have increased mitochondrial DNA damage and upregulated mitophagy (Martin et al., 2006; Choubey et al., 2011; Chen et al., 2015). In contrast, a recent study showed delayed mitophagy in PD patient neurons caused by abnormal accumulation of Miro protein (Shaltouki et al., 2018). α-SYN oligomers also reduced axonal mitochondria transport in induced pluripotent stem cell (iPSC)- derived neurons (Prots et al., 2018). Recent studies have also shown that α-SYN translocated to the mitochondrial matrix and caused impairment of complex I leading to decreased ATP synthesis and increased reactive oxygen species (ROS) production (Martínez et al., 2018). These results suggest that α-SYN can disrupt the mitochondrial homeostasis in several ways.
α-SYN oligomers can interact with and permeabilize lipid membranes causing structural alterations of the intracellular and plasma membranes, increase of intracellular calcium levels, and activation of calpain (van Rooijen et al., 2010; Melachroinou et al., 2013; Ronzitti et al., 2014). Additionally, α-SYN oligomers can inhibit tubulin polymerization and impair neurite network morphology and overexpression in cultured cells and cause microtubule destabilization and neurite degeneration (Lee et al., 2006; Chen et al., 2007; Prots et al., 2013). α-SYN fibrils have also been shown to impair axonal transport of autophagosomes and endosomes but the fibrils didn't affect the transport of synaptophysin or mitochondria (Volpicelli-Daley et al., 2014). However, a recent study found that α-SYN oligomers disrupted anterograde axonal transport of mitochondria and caused subcellular changes in transport-regulating proteins in iPSCderived neurons (Prots et al., 2018).
## Major Pathways of Alpha-Synuclein Clearence in PD
The protein degradation system is part of a protein quality control machinery which clears non-essential misfolded, or damaged proteins. The two major protein degradation systems are the ubiquitin-proteasome pathway (UPP) and ALP. Both are affected in synucleinopathies. These pathways have been shown to be responsible for degrading α-SYN, and failure in one or both can lead to accumulation. The progressive accumulation of α-SYN typical in PD can be linked to the disruption of the UPP by aggregation (Lindersson et al., 2004) as well as different types of autophagy (Winslow et al., 2010; Malkus and Ischiropoulos, 2012). It has been shown that aggregated α-SYN can bind to the membrane proteins of lysosomes and block their function (Malkus and Ischiropoulos, 2012) as well as inhibit certain enzymatic activity domains of proteasomes (Lindersson et al., 2004). α-SYN also inhibits the expression of proteins relevant to autophagosome assembly leading to inefficient removal of aggregated proteins due to impairment in macroautophagy (Winslow et al., 2010).
#### The Ubiquitin-Proteasome Pathway
In the UPP, short-living proteins that are coupled with ubiquitin molecules are degraded by proteasomes (Pickart, 2001; Glickman and Ciechanover, 2002) (**Figure 1A**). The first evidence of UPP failure in PD came from post-mortem studies that used enzymatic assays to evaluate proteasome activity in brain tissues. These studies showed a significant decrease in the chymotrypsinlike and trypsin-like proteasome activity in the SN of PD patients in comparison to age-matched controls. No evidence of defective proteasome activity was seen in other brain regions but rather increased activity was observed in the unaffected areas (Furukawa et al., 2002). These studies suggested that reduced proteasome activity is specific for certain brain regions, like SN. In contrast, it is probable that the decreased activity of proteasomes could be a consequence of the neurodegeneration in this region.
Reduced levels of proteasome subunits have been observed in PD patients. Several genes that code for proteasome subunits
were downregulated in the SN of PD patients and were linked to reduced levels of 20S proteasome core, α-subunit and 19S regulatory caps (Furukawa et al., 2002; McNaught et al., 2002a, 2003; Grünblatt et al., 2004; Chu et al., 2009; Bukhatwa et al., 2010).
Some studies have demonstrated altered proteasome function in peripheral blood cells of PD patients, but the results were significant only in patients treated with L-DOPA and dopamine agonists (Blandini et al., 2006; Ullrich et al., 2010). This indicates that dopamine levels can alter the proteasome function and is supported by animal and in vitro models (Yoshimoto et al., 2005; Berthet et al., 2012).
Studies with disease models have implicated the dysfunction of UPP in PD. Treatment with proteasome inhibitor lactacystin leads to dose-dependent neurodegeneration and formation of ubiquitin and α-SYN positive inclusions in α-SYN-eGFP transfected mouse primary neurons, rat ventral mesencephalic primary neurons, and cultured PC12 cells (McLean et al., 2001; Rideout et al., 2001; McNaught et al., 2002b). McNaught and colleagues used a systemic application of proteasome inhibitor in rats which led to motor deficits and main pathological features typical for PD (McNaught et al., 2004). Also, neurons were found to contain α-SYN and ubiquitin-positive inclusion bodies. Since then, this model has been challenged due to several laboratories inability to replicate the model (more information, see review, Bentea et al., 2017). Besides these studies, toxinbased animal and cellular models have implicated a link between sporadic PD and UPP failure. The toxin 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) appears to target specifically those neurons that are involved in PD. MPTP can easily cross the blood–brain barrier and then is metabolized by astrocytes to become 1-methyl-4-phenylpyridinium (MPP++) ion which is also toxic. MPP+ is released from astrocytes and taken up by DA neurons. MPTP administration causes nigral cell loss, striatal dopamine loss and behavioral deficits (Meredith and Rademacher, 2011). Several in vitro experiments have shown decreased proteasome activity after exposure of pesticides and environmental toxins linked to PD (Wang et al., 2006; Caneda-Ferrón et al., 2008; Chou et al., 2010). Consistent with these findings, in vivo studies showed reduced proteasome activity after rotenone and MPTP administration (Fornai et al., 2005; Betarbet et al., 2006). The UPP impairment caused by MPTP was alleviated in mice lacking α-SYN suggesting that it increases the detrimental effects of MPTP on the UPP (Fornai et al., 2005).
Numerous in vitro studies with purified proteins or cell culture systems, have demonstrated that mutant or wildtype α-SYN can inhibit 20S or 26S proteasome activity, especially in the case of oligomer or fibril formation. In PC12 cells expressing human mutant A53T α-SYN, cells
exhibited accumulation of cytoplasmic ubiquitinated aggregates corresponding with decreased proteasomal chymotrypsin-like activity measured from cell lysates (Stefanis et al., 2001). The finding was confirmed by another study using the same cell line which showed that chymotrypsin-like, trypsinlike and caspase-like activities of the proteasome were all decreased (Tanaka et al., 2001). Also, mutant α-SYN increased cell death in the presence of a proteasome inhibitor. In M17 neuroblastoma cells, mutant α-SYN (A30P or A53T) increased sensitivity to proteasome inhibitors by decreasing proteasome function measured by a GFP reporter system (Petrucelli et al., 2002). These studies indicated reduced proteasome function related to mutant α-SYN, but not to wild-type. However, Snyder et al. (2003) showed inhibition of the proteasome with overexpressing wild-type α-SYN in neuroblastoma M17 cells. A study with yeasts revealed impairment of proteasome-mediated protein degradation in cells expressing wild-type and mutant (A30P) α-SYN (Chen et al., 2005). The cells expressing α-SYN also exhibited a decrease in chymotrypsin-like activity, but no other proteolytic activity of the proteasome was altered.
Contrary to these studies, overexpression of wild-type or mutant (A30P, A53T) α-SYN in PC12 cells or transgenic mice did not result in dysfunction of the UPP (Martìn-Clemente et al., 2004). In a more recent study, Zondler et al. (2017) demonstrated that the impairment of proteasome activity by α-SYN is dependent upon the cellular background. In this study, recombinant α-SYN oligomers and fibrils in vitro or transient expression of wild-type or mutant (A30P, A53T) α-SYN in U2OS ps 2042 [Ubi(G76V)-GFP] cells did not affect 20S proteasome function. In contrast, in DA SH-SY5Y and PC12 cells, stable expression of both wild-type and mutant α-SYN resulted in impairment of the chymotrypsin-like 20S/26S proteasomal protein cleavage.
The proposed mechanisms of how α-SYN inhibits the activity of the proteasome may be by direct binding to the S6<sup>0</sup> or the Rpt5 subunit of the 19S proteasome, or to the β5 subunit of the 20S proteasome (Ghee et al., 2000; Snyder et al., 2003; Lindersson et al., 2004). In addition, different α-SYN species have been implicated in UPP dysfunction. PC12 cells expressing wild-type or mutant α-SYN produce soluble intermediate sized oligomers that associate with the 26S proteasome and increase in amount after treatment with proteasomal inhibitor, indicating specific degradation of the 26S proteasome (Emmanouilidou et al., 2010). In fact, the expression of α-SYN leads to inhibition of all proteasome activities. This study suggested that only a subset of soluble cellderived α-SYN oligomers are targeted to the 26S proteasome for degradation. Simultaneously, these species can inhibit the proteasome function.
#### The Autophagy-Lysosomal Pathway
The ALP is responsible for degrading long-lived proteins, cellular components and organelles through the lysosomal compartment (Parzych and Klionsky, 2014). The ALP has two main purposes: to clear deleterious intracellular components and recycle macromolecules from organelles and proteins to guarantee proteome renewal. Depending on the delivery method, ALP can be divided into three pathways: macroautophagy, chaperonemediated autophagy (CMA) and microautophagy (**Figure 1B**). Because of the lack of evidence linking microautophagy to α-SYN, the focus here is on macroautophagy and CMA.
Macroautophagy is an evolutionary and highly conserved process and the best known of the three autophagic mechanisms (Parzych and Klionsky, 2014; Bento et al., 2016). After the discovery of autophagy-related genes (Atg), the molecular pathway of macroautophagy has been well-characterized. Macroautophagy involves the formation, elongation, and nucleation of double-membrane organelles called autophagosomes that sequester the substrate before fusion with lysosomes. CMA is a particular system based on the recognition of a specific amino acid sequence (KFERQ) (Dice, 1990). The cytosolic chaperone heat shock cognate protein of 70 kDa (hsc70) recognizes the specific motif and translocates the substrate into the lysosome membrane where it interacts with the lysosome-associated membrane protein type 2A (LAMP2A) (Cuervo and Dice, 1996). The final step of the translocation requires the presence of lysosome-associated hsc70 (lys-hsc70) which disassembles LAMP2A into monomers and initiates a new cycle of substrate uptake and degradation (Agarraberes et al., 1997).
Several genetic factors related to PD are involved or interact with ALP (Gan-Or et al., 2015). Mutations in GBA1 that encode the lysosomal hydrolase GCase can lead to lysosomal dysfunction and disruption of autophagy. For example, SCARB2, which encodes the lysosomal integral membrane protein type 2 and interacts with GCase, has been associated with a reduced risk for PD (Gan-Or et al., 2015). Mutations in the ATP13A2 (PARK9) gene, encoding a lysosomal ATPase, causes Kufor-Rakeb syndrome, a rare form of atypical, juvenile-onset autosomal recessive parkinsonism with pyramidal neurodegeneration and dementia.
Along with mutations in genes coding for lysosomal components, other PD-related mutations have been implicated in the process of autophagy. Mutations in the gene encoding vacuolar protein sorting-associated protein 35 (VPS35) cause a rare form of autosomal dominant PD (Gan-Or et al., 2015). VPS35 is involved in endosomal-lysosomal trafficking which is associated with autophagy. Several autosomal recessive PD genes, like parkin (PARK2), PINK1 (PARK6), DJ-1 (PARK7), and FBXO7 (PARK15) have been linked to mitophagy, the process of degradation of dysfunctional mitochondria by autophagy (Burchell et al., 2013; Gan-Or et al., 2015; Pickrell and Youle, 2015). Mutations in LRRK2 are the most common known genetic causes for PD. LRRK2 can be degraded by macroautophagy and CMA, but the most common mutation, G2019S, is poorly degraded by this pathway (Orenstein et al., 2013). In addition, mutated LRRK2 impaired CMA leading to an accumulation of other CMA substrates, including α-SYN. Moreover, mutant LRRK2 caused an increase of autophagic vacuoles in a neuronal cellular model, proposing a more general role of LRRK2 in autophagy (Plowey et al., 2008). In a recent study, Ho et al. (2018) showed that LRRK2 mediates phosphorylation of leucyl-tRNA synthetase leading to impairment of autophagy.
After the initial finding of accumulation of autophagic vacuoles in the SN of PD patients (Anglade et al., 1997), several pathological studies suggested that macroautophagy and CMA are deregulated along with several key proteins related to macroautophagy. Increased levels of beclin-1, which is responsible for the formations and maturation of autophagosomes, and increased levels of autophagosome marker LC3II have been found in the SN of PD patients (Dehay et al., 2010; Miki et al., 2016). Likewise, decreased levels of lysosomal-associated membrane protein type 1 (LAMP1) were evident in nigral neurons in PD patients (Chu et al., 2009; Dehay et al., 2010). The impairment of CMA is associated with the pathogenesis of PD since chaperone Hsc70 and LAMP2 were less expressed in several structures of PD brains (Alvarez-Erviti et al., 2010; Murphy et al., 2015). Decreased levels of several lysosomal markers have been shown in the SN of PD patients. These include the structural protein LAMP1 (Dehay et al., 2010), the lysosomal P-type ATPase ATP13A2 (Dehay et al., 2012), GCase (Gegg et al., 2012; Murphy et al., 2014) and heat shock protein 73 (Chu et al., 2009). In addition, altered activities of lysosomal enzymes, like GCase, Cathepsin A and D have been detected in PD brains (Chu et al., 2009; Gegg et al., 2012; Murphy et al., 2014; Chiasserini et al., 2015).
Transcriptome studies have revealed deregulation of the ALP in PD brains and alterations of several autophagy-related processes, including mTOR, PI3K/AKT, and 14-3-3 protein signalings (Elstner et al., 2011; Mutez et al., 2014; Dijkstra et al., 2015). Increased levels of mTOR protein expression were found in the temporal cortex of patients with DLB, in particular in neurons displaying α-SYN accumulation (Crews et al., 2010). The alteration of other upstream autophagy-related proteins has also been demonstrated, including the immunoreactivity of UNC-51 like kinase 1 (ULK-1), ULK-2, VPS35 and autophagy/Beclin-1 regulator 1 (AMBRA1) within mature LBs, increased levels of Beclin-1, and changed subcellular localization of transcription factor EB (TFEB) (Decressac and Bjorklund, 2013; Miki et al., 2016). Moreover, in a recent study a downregulation of 6 core autophagy genes (ULK 3, Atg2A, Atg4B, Atg5, Atg16L1, and histone deacetylase 6), and increased protein levels of ULK1, Beclin1, and AMBRA1 were detected in peripheral blood mononuclear cells (PBMCs) of PD patients (Miki et al., 2018). These protein levels correlated with increased α-SYN levels in PBMCs. These results suggest a decrease in autophagy properties in PD patients.
#### **The autophagy-lysosomal pathway and** α**-SYN**
While α-SYN can be cleared by UPP, the main pathway for its degradation appears to be lysosomal (Webb et al., 2003; Vogiatzi et al., 2008). α-SYN can be degraded by both macroautophagy and CMA, but the structure and mutations may change the final path of degradation. Small soluble forms of α-SYN are more likely to be degraded by CMA but in the pathological condition the burden shifts to macroautophagy even though both pathways can compensate for each other.
Induction of autophagy with rapamycin leads to the clearance of overexpressed wild-type and mutant α-SYN in cell cultures (Webb et al., 2003). This study established that inhibition of macroautophagy with 3-methyladenine causes the accumulation of mutant – but not wild-type – α-SYN. In contrast, a study with PC12 cells showed an increase in both endogenous and overexpressed wild-type α-SYN when macroautophagy was inhibited with 3-methyladenine (Vogiatzi et al., 2008). Other studies with neuronal cells or transgenic mice overexpressing wild-type α-SYN showed accumulation of α-SYN only upon general lysosomal inhibition, and not in suppressed macroautophagy (Lee et al., 2004; Klucken et al., 2012). However, another study has revealed an increase of A53T α-SYN oligomers after pharmacological or molecular inhibition of macroautophagy (Yu et al., 2009). Conditional depletion of Atg7 in DA neurons caused age-related neuronal loss, the formation of ubiquitinated protein aggregates and increase in monomeric α-SYN (Ahmed et al., 2012). In addition, α-SYN aggregates were detected in striatal axonal swellings of 20-monthold mice after depletion of Atg7 (Friedman et al., 2012). Both studies suggested a role of macroautophagy in α-SYN turnover in vivo, since macroautophagy impairment caused modest alterations in endogenous α-SYN. Overall, these studies indicate that degradation of α-SYN by macroautophagy may depend on the conformation of α-SYN. It is likely that small amounts of wild-type α-SYN are degraded by CMA, but in cases of overexpression or mutations, macroautophagy becomes a more important pathway. The macroautophagic degradation of α-SYN could also depend on PTMs. Phosphorylation and SUMOylation [Small Ubiquitin-like Modifier (SUMO)] have been reported to increase α-SYN degradation by macroautophagy in yeast and PD models (Oueslati et al., 2013; Shahpasandzadeh et al., 2014; Tenreiro et al., 2014). Inside the lysosome, α-SYN is mainly degraded by Cathepsin D, and overexpression of the mutant form of this protease leads to increased levels of α-SYN (Crabtree et al., 2014). In addition, Cathepsin D knockout mice exhibited an accumulation of higher molecular weight α-SYN species (Cullen et al., 2009). Overexpression of mutant A53T α-SYN has been reported to enhance autophagic flux which causes increase in autophagic vacuoles and macroautophagic degradation (Cuervo et al., 2004; Xilouri et al., 2009; Choubey et al., 2011). Similar effects have been reported with wild-type α-SYN, although to a lesser extent. α-SYN can also inhibit macroautophagy via interaction with Rab proteins leading to Atg9 mislocalization (Winslow et al., 2010). Furthermore, α-SYN has been shown to enhance mitophagy. In a transgenic mouse model expressing A53T specifically in DA neurons, the induction of mitophagy was detected (Chen et al., 2015). In the cell culture model, α-SYN overexpression caused increased mitophagy leading to neuronal death (Choubey et al., 2011). However, elevated macroautophagic flux was evident in primary midbrain neurons overexpressing wild-type and A53T α-SYN, without significant alterations in mitophagy (Koch et al., 2015).
Besides neuronal cells, the relation of α-SYN and autophagy has also been demonstrated in other cell types. DJ-1 knockdown microglia exhibited an impaired uptake of α-SYN and had lower autophagy-dependent degradation of p62 and LC3 proteins (Nash et al., 2017). In immortalized astrocyte cell lines overexpressing wild-type, A30P and A53T mutant α-SYN showed decreased LC3-II and increased p62 protein levels, suggesting the
inhibition of autophagy (Erustes et al., 2018). In addition, iPSCderived astrocytes with LRRK2 G2019S mutation accumulated α-SYN and had impaired macroautophagy and dysfunctional CMA (di Domenico et al., 2019). PD astrocytes displayed LAMP2A positive vesicles all around the cell body, whereas in control lines the vacuoles were in the perinuclear area. In addition, α-SYN co-localized with LAMP2A receptor in PD astrocytes. LAMP1 -positive vesicles were also found throughout the cell in PD astrocytes, and there was an increase in autophagic vacuoles. Furthermore, higher basal levels of LC3-II, p62 and impaired autophagic flux were detected from PD astrocytes.
The link between α-SYN and CMA was initially established in purified lysosomes demonstrating that α-SYN could be actively degraded by CMA (Cuervo et al., 2004). Interestingly, A30P and A53T α-SYN mutations had higher affinity to LAMP2A and blocked and totally impaired the CMA pathway. Since then, the higher affinity of the mutant α-SYN to LAMP2A was confirmed in neuronal cultures and other cell culture models (Vogiatzi et al., 2008; Alvarez-Erviti et al., 2010). In the neuronal systems, the inhibition of CMA leads to the formation of high molecular weight or detergent-insoluble oligomeric α-SYN conformations (Vogiatzi et al., 2008). Also, in mice where α-SYN expression was enhanced with paraquat or transgenic overexpression, the intralysosomal content of α-SYN was increased as well (Mak et al., 2010). The overexpression of α-SYN in mice also led to upregulation of LAMP2A and hsc70. Another study with mice with VPS35 deficiency or expression of PD-linked mutation D620N showed accumulation of α-SYN in DA neurons and DA degeneration (Tang et al., 2015). This was accomplished by an impaired endosome-to-Golgi retrieval of LAMP2A leading to decreased levels of LAMP2A and a reduced α-SYN clearance. In Drosophila, which lacks CMA, neuronal expression of human LAMP2A protected against starvation and oxidative stress and delayed the locomotor decline in aging flies (Issa et al., 2018). LAMP2A also alleviated the progressive locomotor and oxidative defects induced by neuronal expression of PDassociated human A30P α-SYN. LAMP2A stimulated autophagy in adult Drosophila, and neuronal expression of LAMP2A upregulated levels of Atg5.
PTMs can affect the degradation of α-SYN through CMA. Oxidation and nitration of α-SYN slightly inhibited the CMA, whereas phosphorylation and exposure to dopamine almost completely block the CMA degradation system. However, only dopamine-modified α-SYN blocks the degradation of other substrates (Martinez-Vicente et al., 2008). The same study reported that CMA could degrade only monomeric or dimeric α-SYN, but not oligomers. Blocking the CMA by aberrant forms of α-SYN can also have a toxic effect and impact other degradation pathways. PD-linked mutations like A30P and A53T or dopamine-modified wild-type α-SYN can inhibit the function of CMA leading to activation of macroautophagy and increased toxicity in cells (Martinez-Vicente et al., 2008; Xilouri et al., 2009).
Recently, micro-RNAs (miRNAs) have been implicated in CMA function and α-SYN clearance. Several miRNAs have been described to target LAMP2A and Hsc70 and decrease α-SYN degradation (Alvarez-Erviti et al., 2013; Li et al., 2014). The initial study found four miRNAs that reduce LAMP2A levels, and three that decreased Hsc70 levels. This was accompanied by increased accumulation of α-SYN in SH-SY5Y neuroblastoma cells. These miRNAs were also found up-regulated in brains of PD patients and correlated with decreased protein levels of CMA (Alvarez-Erviti et al., 2013).
#### **Autophagy enhancing agents as a potential therapeutic strategy for PD**
Because ALP is an important pathway in α-SYN degradation, the opportunity to use autophagy enhancement as a strategy against α-SYN aggregation in PD has raised considerable interest. Pioneering studies with rapamycin and other macroautophagy enhancing agents have demonstrated an increased α-SYN clearance in several PD models. However, the selectivity of these early autophagy enhancers is limited. Selective targeting of ALP components, like TFEB, lysosomes, and CMA, may provide more potential for development of new therapies for PD. The main findings are listed in **Table 1**, and for more details see the literature (Moors et al., 2017).
The most studied and used macroautophagy-enhancer is rapamycin which inhibits mTORC1 signaling (Bové et al., 2011). Rapamycin has been shown to reduce α-SYN accumulation in wild-type, A30P, or A53T α-SYN expressing PC12 cells and in mice (Crews et al., 2010) and rats (Decressac and Bjorklund, 2013) with overexpressed α-SYN. Rapamycin also improved the motor function in mice with overexpressed A53T α-SYN (Bai et al., 2015). The drawback of mTORC1 inhibition is the interference with numerous other pathways. Prolonged treatment with rapamycin can inhibit mTORC2 and stimulate other cellular pathways, including cell survival mechanisms (Bové et al., 2011). The activation of macroautophagy can be achieved by activating AMPK, leading to downstream inhibition of mTORC1. Several agents which act through this pathway have been described, such as metformin, 5-aminoimidazole-4 carboxamide ribonucleotide (AICAR) and resveratrol (Curry et al., 2018). Metformin, commonly used to treat Diabetes Mellitus, showed neuroprotective effects in in vitro and in vivo models of PD (Ng et al., 2013; Dulovic et al., 2014; Patil et al., 2014). Metformin also decreased phosphorylated levels of α-SYN in SH-SY5Y cells and MPTP-treated mice (Pérez-Revuelta et al., 2014; Katila et al., 2017). Another agent that affects the AMPK signaling is trehalose, which inhibits members of the SLC2A (GLUT) family of glucose transporters leading to AMPK-dependent increase of macroautophagy (DeBosch et al., 2016). Trehalose-induced autophagy has shown to increase cell survival and α-SYN clearance in cell lines and in multiple in vivo models (Sarkar et al., 2007, 2014; Rodríguez-Navarro et al., 2010; Lan et al., 2012; Tanji et al., 2015; He et al., 2016). However, a recent study did not find improvement in neuronal survival after exposure to α-SYN pre-formed fibrils (Redmann et al., 2017). Although increasing autophagy by AMPK pathway has shown beneficial effects, AMPK is involved in several other cellular functions, and its modulation is likely to induce unwanted effects.
Recently, several other agents acting through an mTORdependent pathway have been studied in cell cultures and PD animal models. Sheng et al. (2017) showed that uric acid treatment increased autophagy in PC12 cell in dose- and
TABLE 1 | Commonly used autophagy enhancing agents.
mTORC1, mammalian target of rapamycin complex 1 or mechanistic target of rapamycin complex 1; PREP, prolyl oligopeptidase; 2-HPβCD, 2-hydroxypropylβ-cyclodextrin; α-SYN, alpha-synuclein; DA, dopaminergic; ROS, reactive oxidative stress; WT, wild-type; 6-OHDA, 6-hydroxydopamine; MPTP, 1-methyl-4-fenyl-1,2,3,6 tetrahydropyridine; Gcase, glucocerebrosidase; TFEB, transcription factor EB; SLC2A, (GLUT) family of glucose.
time-dependent manners. Moreover, uric acid reduced α-SYN accumulation in PC12 cells overexpressing wild-type or A53T mutant α-SYN. In vivo, uric acid modulated autophagy markers increased the autophagosome/autolysosome formation and reduced α-SYN accumulation in the midbrain of SNCA A53T transgenic mice. Suresh et al. (2017) showed that a novel autophagy modulator 6-Bio alleviated α-SYN toxicity. In yeast and mammalian cell lines, 6-Bio induced autophagy and enhanced autolysosome formation which resulted in α-SYN degradation and clearance. In vivo studies with a MPTP mouse model demonstrated that 6-Bio has a neuroprotective activity, enhances autophagy and clearance of toxic protein aggregates and ameliorates MPTP-induced behavioral deficits. The results demonstrated that 6-Bio modulates autophagy in a GSK3B-dependent manner and the induction of autophagy in mammalian cells appears to be mTOR dependent.
Instead of general activation of macroautophagy, targeting selective ALP components including Beclin-1, TFEB, and lysosomes have been tried. Activation of Beclin-1 induces autophagosome formation and initiation of autophagy. Overexpression of Beclin-1 has been shown to reduce accumulation of α-SYN in PC12 cells and mice with overexpressed α-SYN (Spencer et al., 2009; Wang et al., 2016). In addition, the drug-induced activation of Beclin-1 has been demonstrated to increase autophagy and promotes α-SYN clearance in neuronal cell lines and PD animal models (Lu et al.,
2012; Savolainen et al., 2014). One attractive target to stimulate macroautophagy downstream of mTORC1 is modulating transcriptional levels of TFEB. TFEB regulates macroautophagy and lysosomes and acts as a link between upstream signaling pathways (Settembre et al., 2011). Overexpression of TFEB eliminated α-SYN oligomers and rescued midbrain DA neurons from α-SYN toxicity in overexpressing rats (Decressac and Bjorklund, 2013). Another strategy for stimulating ALP in PD is a direct modulation of lysosomes. The potential of targeting the lysosome system has been demonstrated with acidic nanoparticles which were able to stimulate lysosomal degradation and revert the lysosomal dysfunction in genetic PD models (Baltazar et al., 2012; Bourdenx et al., 2016). Ambroxol, AT2101 (isofagomine) and histone deacetylase inhibitors can correct the folding of GCase and therefore increase the GCase and lysosome function (Blanz and Saftig, 2016). The small-molecule chaperones have been demonstrated to enhance GCase activity, improve lysosomal function and enhance α-SYN clearance in preclinical models of PD (Steet et al., 2006; Khanna et al., 2010; Sun et al., 2012; Yang et al., 2013; McNeill et al., 2014; Richter et al., 2014; Ambrosi et al., 2015). Especially ambroxol is widely studied presently. In mice overexpressing α-SYN or heterozygous L444P mutation in CBA1, ambroxol treatment increased the GCase activity while decreasing phosphorylated and endogenous levels of α-SYN (Migdalska-Richards et al., 2016, 2017b). In non-human primates, ambroxol increased brain GCase activity (Migdalska-Richards et al., 2017a). In patients with Gaucher disease, ambroxol was able to cross the blood–brain barrier and high-dose oral administration was safe and well-tolerated (Narita et al., 2016).
Currently, ambroxol is in phase II clinical trials tested for treatment of PD and PD with dementia (ClinicalTrials.gov Identifier: NCT02941822 and NCT02914366, respectively) (Silveira et al., 2019).
Downstream targeting of the CMA components presents an alternative approach to develop new strategies for PD. Induced overexpression of LAMP2A in human SH-SY5Y cells, rat primary cortical neurons in vitro and nigral DA neurons in vivo decreased α-SYN accumulation and protected α-SYNinduced DA degeneration (Xilouri et al., 2013). Retinoic acid alpha receptors have been identified as CMA inhibitors, and synthetic derivatives of all-trans-retinoic acid were developed to reverse this effect (Anguiano et al., 2013). These derivates specifically stimulated CMA and LAMP2A was identified as one of the targets.
### ROLE OF ER STRESS IN PD AS A RESULT OF DYSFUNCTIONAL CELLULAR PROTEOSTASIS
Endoplasmic reticulum is the first component of the secretory pathway mainly responsible for protein synthesis, posttranslational processing and folding of newly synthesized proteins. The proteins are then transported to their final destinations in membrane-bound vesicles. Disturbance in any of these functions including proper folding capacity and disposal of misfolded proteins leads to ER stress and activation of intracellular signal transduction pathway that is essentially intended to re-establish ER homeostasis. These biological processes are collectively called the UPR. Inability to restore ER functions induces cell death via apoptosis. Growing evidence from studies in human PD post-mortem brain, additionally to genetic and neurotoxin models, suggests that ER stress is a common feature in PD that contributes to PD pathology. Recently, the generation of neuronal cultures from iPSCs derived from PD patients indicated that ER stress leads to the accumulation of ER-associated degradation (ERAD) substrates and placed this ER dysfunction as an early component of PD pathogenesis (Chung et al., 2013; Heman-Ackah et al., 2017).
## Causes of ER Stress in PD
The ER is crucial for protein folding, trafficking to the Golgi, UPR, and calcium buffering. The imbalance between the load on ER functions and ER capacity leads to ER stress. In PD, the mechanisms leading to ER stress and the actual role of the UPR in degeneration of the DA neurons are the object of intensive research. Oligomeric α-SYN has been shown to accumulate in the ER in animal models and PD patient brains (Colla et al., 2012). The aggregation of α-SYN induces ER stress, which eventually results in inflammation and neurodegeneration.
A number of studies have shown that α-SYN affects Rab1, a protein involved in trafficking components from the ER to the Golgi. Over-expression of Rab1 in animal models of PD reduced stress levels and protected DA neurons against degeneration (Coune et al., 2012). Further, α-SYN directly interacts with nascent activating transcription factor 6 (ATF6), effectively preventing its association with COPII vesicles that generally transfer proteins to the Golgi. This has specific implications: (1) interfering with the Rab1 protein could lead to accumulation of unfolded proteins in the ER, and (2) inhibition of ATF6 would generally stop the ERAD, triggering the cell to signal apoptosis (Credle et al., 2015). Other studies are linking PD genes with alteration of the secretory pathway, including LRRK2, Parkin, DJ-1, ATP13A2 (Mercado et al., 2013), and VPS35 (Zimprich et al., 2011), which may result in pathological levels of ER stress contributing to the etiology of the disease. Furthermore, increase in cytoplasmic Ca2<sup>+</sup> levels induced by 6-hydroxydopamine (6-OHDA) (a toxin capable to generate, in vitro and in vivo, some features of PDassociated neurodegeneration) was detected in 6-OHDA-treated rats. Pretreatment with ryanodine or ER stress inhibitor 4-PBA inhibited RyR receptor Ca2<sup>+</sup> channels and protective midbrain DA neurons from degeneration (Huang et al., 2017).
### UPR Response in PD
The primary function of the UPR is a maintenance of ER protein homeostasis. When cells undergo constant ER stress, the UPR is responsible for the elimination of damaged cells through apoptotic mechanisms, some of which appear to be specific to ER stress and others that are included in general apoptotic pathways (Xu et al., 2005). The activation of UPR depends on three ER stress sensors proteins, protein kinase RNA-like endoplasmic reticulum kinase (PERK) receptor, inositol-requiring enzyme 1
(IRE1), and ATF6 (Schröder and Kaufman, 2005) (**Figure 2**). Under normal physiological conditions, all three effectors are negatively regulated by the ER chaperone glucose-regulated protein 78/ binding immunoglobulin protein (GRP78/BiP), which suppresses their activity by binding to their luminal ends (Bertolotti et al., 2000).
Under conditions of ER stress and increase in unfolded proteins, BiP dissociates from UPR sensors inducing their activation. Activation of the ER pathways helps to fight the cellular stress through the combined actions of suppressing the translation of new proteins, inducing ER chaperones that promote protein refolding and activating the proteasome to degrade misfolded/unfolded proteins. There is direct evidence that GRP78/BiP levels are increased in cell as well as animal models of PD forming a complex with α-SYN (Bellucci et al., 2011). Moreover, it has been demonstrated that aging leads to a significant decline in GRP78/BiP expression (up to 40%; Naidoo, 2009). However, under chronic ER stress, UPR sensors shift their signaling toward induction of cell death by apoptosis (Urra et al., 2013).
#### PERK Signaling
PERK is a type I ER transmembrane protein kinase with a luminal domain and a cytoplasmic domain that has kinase activity (Liu et al., 2002). Upon ER stress, BiP releases the luminal domain of PERK, which then dimerizes and autophosphorylates to become active (Harding et al., 1999). Following transautophosphorylation, this kinase phosphorylates the alpha subunit of eukaryotic initiation factor-2 (eIF2), inactivating it by Ser-51 phosphorylation and attenuating protein translation. This inhibitory effect of translation helps to alleviate ER stress by decreasing the overload of misfolded proteins and thereby protecting the cells under conditions where proteins cannot achieve proper folding (Fels and Koumenis, 2006). This event leads to activating transcription factor 4 (ATF4). The UPRrelated transcriptional factor ATF4 upregulates a subset of genes that control oxidative stress, metabolism, protein folding, and glutathione biosynthesis (Harding et al., 2000). Important targets of ATF4 include NRF2, which regulates the functions of a variety of antioxidant genes (He et al., 2001), and CHOP, which conversely is a key in the activation of apoptotic pathways and cell death (Han et al., 2013). In PD patients, the activation of PERK/ATF4 signaling is observed in different brain areas. α-SYN has been shown to accumulate within the ER of nigral DA cells, directly activating the PERK/eIF2α signaling and increasing the expression of ATF4 (Bellucci et al., 2011). The activation of this pathway overlapped with pro-apoptotic changes.
Further evidence supporting this pathway comes from PDassociated gene studies. By inducing the A53T α-SYN mutation to PC12 cells, UPR activates CHOP and GRP78/BiP by increasing their expression and increases the phosphorylation of eIF2α (Smith et al., 2005). Interventions to block ER stress and caspase activity using inhibitors, and to knock down the expression of caspase-12 using siRNA, protected against A53T α-SYN induced cell death (Smith et al., 2005). In PINK1- and Parkinassociated PD models, mitofusins cause enhanced ER stress
signaling, by interconnecting damaged mitochondria to the ER membranes (Celardo et al., 2016). PERK signaling inhibition, either pharmacological or genetic suppression, was beneficial in these experimental models of PD (Celardo et al., 2016). LRRK2 mutations also cause familial PD by accumulation and aggregation of α-SYN and ubiquitinated proteins over time mainly due to the impairment of protein degradation pathways (Tong et al., 2010). This is likely to result in the buildup of unfolded proteins leading to ER stress, although there is little evidence of UPR activation yet. On the other hand, studies with GBA1 gene mutations revealed that treatment with chemical chaperones (ambroxol and isofagomine) can combat GBA-mediated ER stress by increasing GBA levels and activity in fly models and in fibroblasts from PD patients (Sanchez-Martinez et al., 2016). In rodent models of PD, ATF4 upregulation in DA neurons of SN resulted in severe nigrostriatal degeneration caused by activating caspase 3/7 dependent pathway (Gully et al., 2016). On the other hand, Grp78/BiP over-expression exerted neuroprotective effects in a rat model of PD (Gorbatyuk et al., 2012).
#### IRE1 Signaling
IRE1 is a type I ER transmembrane sensor and cell fate executor. IRE1 gets activated in response to the accumulation of misfolded proteins by autophosphorylation. The activation induces RNase activity that is consequently needed for unconventional splicing of X-box binding protein 1 (XBP1) (Calfon et al., 2002). Spliced XBP1 translocates to the nucleus where it commands transcription of genes responsible for quality control, protein folding, lipid synthesis and ERAD pathway (Sriburi et al., 2004). In PD, XBP1 controls the survival of DA neurons (Valdés et al., 2014). The developmental ablation of XBP1 preconditions DA neurons against the effect of the neurotoxin, 6-OHDA (Mollereau et al., 2014, 2016). The effect is specific to SNpc, as it is not seen in other brain regions. Contrary, reduction of XBP1 in DA neurons of adult mice caused ER stress with CHOP induction leading to degeneration (Valdés et al., 2014), highlighting the critical role of XBP1 depending on the development stage. In addition, a gene therapy approach using adeno-associated viral vectors to deliver XBP1 active form to the SNpc confers protection of DA neurons against 6-OHDA-mediated toxicity (Valdés et al., 2014). Moreover, XBP1 transgene delivered to mouse striatum using recombinant adenoviral vectors protected DA neurons against MPTP-induced degeneration (Sado et al., 2009). XBP1 is also protective when it is delivered in neural stem cells transfected with this transcription factor resulting in increased survival and improved behavior in a rotenone-induced rat model of PD (Si et al., 2012). Among other functions, XBP1 and ATF6 mediate the transcription of BiP. While the overexpression of this chaperone protects DA neurons and increases motor performance in a rat model of PD, the age-related decline in BiP expression as well as siRNA-mediated downregulation, increases DA neuron vulnerability to α-SYN in the same PD model (Salganik et al., 2015).
macroautophagy or CMA. Alternatively, misfolded α-SYN undergoes refolding in the ER. However, excessive refolding upregulates PDI reduction. Re-oxidation of PDI is linked with an increase in hydrogen peroxide generation causing dysregulation of IP3R permeability and an increase in cytosolic calcium. Calcium release from the ER may activate calpain and eventually lead to apoptosis. The pharmacological inhibition of PDI by bacitracin or cystamine prevents ER redox imbalance and downstream proapoptotic events. The inhibition of the ERO1 catalyzed re-oxidation of PDI by EN460 results in a protective effect similar to PDI inhibitor.
#### ATF6 Signaling
fnins-13-00457 May 10, 2019 Time: 15:3 # 12
ATF6 is a type II ER transmembrane protein. Upon the accumulation of misfolded proteins in the ER, ATF6 moves to the Golgi apparatus where it is cleaved by two proteases (Haze et al., 1999). The cytosolic domain of ATF6 is translocated to the nucleus where it activates the transcription of ER chaperones (Gotoh et al., 2002). Contrarily, reduced levels of ATF6 in the nucleus of cells due to the impairment in ATF6 trafficking to the Golgi are likely to trigger apoptosis (Credle et al., 2015). In PD, α-SYN directly targets ATF6 and inhibits ATF6 processing leading to an impaired up-regulation of ERAD genes, which sensitizes cells to apoptosis (Credle et al., 2015). In ATF6α knockout animals, the accumulation of ubiquitin-positive inclusions and enhanced loss of DA neurons induced by MPTP, a PD-triggering neurotoxin, was detected (Egawa et al., 2011). This suggests that activation of the UPR has an important adaptive function to maintain protein homeostasis in this model. ATF6 mainly controls the levels of BiP and ERAD elements rather than development and survival of DA neurons in mice under resting conditions (Egawa et al., 2011).
#### SUPPRESSION OF EXCESSIVE PROTEIN OXIDATIVE FOLDING AS AN ALTERNATIVE SOLUTION FOR LOWERING ER STRESS
As indicated above, ER stress is increasingly implicated in PD, and emerging evidence highlights the complexity of the UPR, with both protective and detrimental components being described. Mild insults increase the activity of chaperones, such as protein disulfide isomerase (PDI) that is responsible for the oxidative folding through formation of disulfide bonds in proteins (Rao and Bredesen, 2004). To promote correct disulfide bond formation in unfolded/misfolded proteins, the redox environment in the ER is oxidatively maintained (Hwang et al., 1992). In neurons, the increased activity of PDI represents an adaptive response that is induced to protect the cells (Haynes et al., 2004). In contrast, recent studies have revealed alternative roles for PDI in neurodegenerative diseases (Hoffstrom et al., 2010; Lehtonen et al., 2016).
In PD, ER homeostasis is disrupted in DA neurons in SNpc and PDI co-localizes with α-SYN in LBs (Conn et al., 2004). We have recently demonstrated that treatment with 1 methyl-4-phenylpyridinium (MPP+), a neurotoxin associated with PD, upregulated the expression of α-SYN and PDI in human neuroblastoma SH-SY5Y cells and that α-SYN co-localized with PDI. The α-SYN accumulation not only activated PDI but resulted in the accumulation of a reduced form of PDI due to an increasingly reduced ER redox environment (**Figure 3**).
#### REFERENCES
Agarraberes, F. A., Terlecky, S. R., and Dice, J. F. (1997). An intralysosomal hsp70 is required for a selective pathway of lysosomal protein degradation. J. Cell Biol. 137, 825–834. doi: 10.1083/jcb.137.4.825
Protein disulfide isomerase inhibitors bacitracin and cystamine prevented the accumulation of α-SYN and MPP+ induced reductive shift in the ER by hindering PDI excessive reduction (Lehtonen et al., 2016). Moreover, the data suggest that redox misbalance and hydrogen peroxide production due to PDI re-oxidation in the ER are the outcome of a severe toxic insult caused by α-SYN accumulation. Calpain is a Ca2<sup>+</sup> -sensitive non-lysosomal protease reported to be disruptive in SN of PD patients as well as in experimental PD models (Crocker et al., 2003). In the model described by Lehtonen et al. (2016), the release of Ca2<sup>+</sup> was succesfully blocked not only by 2-aminoethoxydiphenyl borate (2- APB), an inositiol-3-phosphate receptor (IP3R) inhibitor, but also by bacitracin, a PDI inhibitor, and it promoted neuroblastoma cell survival. Additionally, ALLN (N-acetylleu-leu-norleual, N-acetyl-L-leucyl-L-leucyl-L-norleucinal), a calpain I inhibitor, protected these cells from MPP+ toxicity. Overall, these results are in line with a previously published study using a PC12 cell model of Huntington's disease showing the potential of PDI inhibitors to suppress the cells' death induced by misfolded proteins (Hoffstrom et al., 2010). Importantly, a beneficial effect of PDI inhibition is coupled with consecutive enhancement of autophagy that is turned on to support cell survival. Furthermore, PDI inhibition also protected against MPP+ -induced DA neurodegeneration in Caenorhabditis elegans.
Collectivelly, excessive protein refolding taking place in the ER leads to an increase in the reduced form of PDI and to the activation of the PDI-Ero1 cycle, causing overproduction of hydrogen peroxide and promoting generation of ROS. These events lead to the dysregulation of IP3R that causes an increase in cytosolic calcium followed by calpain activationinduced apoptosis. In contrast, PDI inhibition prevents ER redox imbalance and enhances the autophagic clearance pathway. Therefore, when considering therapeutic approaches, it is necessary to take into account the balance between ER-linked refolding or/and alternative protein clearance by autophagy.
#### AUTHOR CONTRIBUTIONS
All authors listed have made a substantial, direct and intellectual contribution to the work, and approved it for publication.
#### FUNDING
This study was supported by Academy of Finland, University of Eastern Finland, University of Helsinki and the Finnish Parkinson Foundation.
Ahmed, I., Liang, Y., Schools, S., Dawson, V. L., Dawson, T. M., and Savitt, J. M. (2012). Development and characterization of a new Parkinson's disease model resulting from impaired autophagy. J. Neurosci. Off. J. Soc. Neurosci. 32, 16503–16509. doi: 10.1523/JNEUROSCI.0209-12. 2012
disease and Alzheimer disease. Neurology 66, 529–534. doi: 10.1212/01.wnl. 0000198511.09968.b3
alpha-synuclein in neuronal cells via inducing autophagy. Autophagy 8, 98–108. doi: 10.4161/auto.8.1.18313
mutation-linked Parkinson disease cells. Brain J. Neurol. 137, 1481–1495. doi: 10.1093/brain/awu020
pleomorphic pathology. Neuron 44, 601–607. doi: 10.1016/j.neuron.2004. 11.005
Zondler, L., Kostka, M., Garidel, P., Heinzelmann, U., Hengerer, B., Mayer, B., et al. (2017). Proteasome impairment by α-synuclein. PLoS One 12:e0184040. doi: 10.1371/journal.pone.0184040
**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Copyright © 2019 Lehtonen, Sonninen, Wojciechowski, Goldsteins and Koistinaho. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
# Linking the Endoplasmic Reticulum to Parkinson's Disease and Alpha-Synucleinopathy
#### Emanuela Colla\*
Bio@SNS Laboratory, Scuola Normale Superiore, Pisa, Italy
Accumulation of misfolded proteins is a central paradigm in neurodegeneration. Because of the key role of the endoplasmic reticulum (ER) in regulating protein homeostasis, in the last decade multiple reports implicated this organelle in the progression of Parkinson's Disease (PD) and other neurodegenerative illnesses. In PD, dopaminergic neuron loss or more broadly neurodegeneration has been improved by overexpression of genes involved in the ER stress response. In addition, toxic alpha-synuclein (αS), the main constituent of proteinaceous aggregates found in tissue samples of PD patients, has been shown to cause ER stress by altering intracellular protein traffic, synaptic vesicles transport, and Ca2<sup>+</sup> homeostasis. In this review, we will be summarizing evidence correlating impaired ER functionality to PD pathogenesis, focusing our attention on how toxic, aggregated αS can promote ER stress and cell death.
#### Edited by:
Sandeep Kumar Barodia, The University of Alabama at Birmingham, United States
#### Reviewed by:
Arthi Kanthasamy, Iowa State University, United States Scott Oakes, University of California, Oakland, United States
> \*Correspondence: Emanuela Colla [email protected]
#### Specialty section:
This article was submitted to Neurodegeneration, a section of the journal Frontiers in Neuroscience
Received: 24 March 2019 Accepted: 15 May 2019 Published: 29 May 2019
#### Citation:
Colla E (2019) Linking the Endoplasmic Reticulum to Parkinson's Disease and Alpha-Synucleinopathy. Front. Neurosci. 13:560. doi: 10.3389/fnins.2019.00560 Keywords: alpha-synuclein, ER stress, UPR, misfolded proteins, Parkinson's disease, alpha-synucleinopathy, alpha-synuclein aggregates
### INTRODUCTION
Misfolded proteins are a common cellular abnormality that is shared among neurodegenerative diseases. Parkinson's Disease (PD), a multifactorial neurodegenerative disorder that affects the motor, cognitive and peripheral system, is characterized by the accumulation of misfolded, aggregated alpha-synuclein (αS) fibrils into proteinaceous intracellular inclusions in neuronal soma or neurites, named Lewy bodies (LB) or Lewy neurites (LN) (Goedert et al., 2013). The aggregation process of αS is a nucleation-type of reaction where αS monomer converts into a β-sheet conformation that elongates into filamentous structures called protofibrils and fibrils, becoming progressively insoluble (Lashuel et al., 2002; Cremades et al., 2012; Tuttle et al., 2016). The presence of αS inclusions is associated with neuronal damage and has been found in other types of neurodegenerative diseases besides PD that are collectively referred as α-synucleinpathies. In PD one of the most affected neuronal populations, but surely not the only one, is represented by the dopaminergic neurons of the substantia nigra pars compacta. Here widespread dopaminergic neuronal death causes depletion of striatal dopamine, whose reduction is responsible for motor and cognitive dysfunction experienced by PD patients. Since its discovery in 1997, many observations have pointed to the aggregation of αS as one of the culprits of neuronal demise (as examples Feany and Bender, 2000; Masliah, 2000; Lee et al., 2002; Lakso et al., 2003). In addition, the prion-like ability of the protein to spread and propagate its toxic template has shown how αS pathology can disseminate between different anatomically connected areas, from the peripheral nervous system to the brain (Braak et al., 2003; Luk et al., 2012; Masuda-Suzukake et al., 2013; Rey et al., 2013; Holmqvist et al., 2014; Recasens et al., 2014; Sacino et al., 2014). At a neuronal level, αS toxicity has been associated with impairment in numerous cellular functional aspects, including mitochondrial, proteasomal and lysosomal abnormalities, axonal transport
deficits and alteration in synaptic transmission (Bendor et al., 2013). More recent evidence has emerged supporting the endoplasmic reticulum (ER) stress, a condition of altered ER functionality, as a mediator of αS toxicity. In this review, we will summarize the link between PD and ER stress, focusing our attention on how pathological αS impairs ER functionality, induces ER stress and ultimately contributes to neuronal death.
#### ER STRESS AND UNFOLDED PROTEIN RESPONSE
Folding of secreted and transmembrane proteins is one of the main functions of the ER. Membrane and extracellularly targeted proteins are translated on ribosomes localized on the cytosolic surface of the rough ER and promptly inserted into the ER membrane or lumen (Görlach et al., 2006). In the ER, proteins achieve a specific folded conformation, acquire post-translational modifications such as glycosylation and formation of disulfide bonds, and are selectively targeted for secretion or destined for the plasma membrane or other cellular compartments. The ER is also responsible for biosynthesis of lipids and steroid hormones and is a primary site for Ca2<sup>+</sup> storage. Proteins that failed to fold properly are retro-translocated into the cytosol by the ERassociated degradation (ERAD) pathway and degraded by the proteasomes (Smith et al., 2011). To sustain extensive protein folding capability, cells must promptly maintain an adequate level of ER folding machinery and ERAD proteins. Because of the high concentration of proteins in the ER (estimated at 100 mg/mL), the ER quality control is a fundamental mechanism that maintains and preserves cell metabolism and normal functions. Perturbations of this balance lead to accumulation of aberrant proteins in the ER, a condition called ER stress, that if left unchecked, can have deleterious consequences and can lead to the collapse of the whole secretory pathway and cellular homeostasis. In addition to defects in the protein folding machinery, other conditions culminating directly or indirectly in the accumulation of misfolded proteins (including starvation, infections, changes in ER Ca2<sup>+</sup> concentration and dysregulation in the redox potential of the ER) are able to trigger ER stress. Because of this fundamental role in protein homeostasis, eukaryotic organisms have developed a concerted and coordinated multi-signaling pathway, named unfolded protein response (UPR), that aims to restore ER functionality through the increase of cellular folding capacity and the transient reduction of the flux of proteins entering the ER (Walter and Ron, 2011). To achieve such a status, a massive transcriptional upregulation of ER chaperons, foldases, glycosylases, ERAD proteins, lipid biosynthesis to facilitate ER membrane expansion and, at the same time, degradation of selective mRNA messengers with attenuation of general protein translation, must be well coordinated in order to protect cells from ER stress and recover ER protein quality control (Harding et al., 1999; Hollien and Weissman, 2006; Hollien et al., 2009). However, when the ER stress is too severe and there is a persistent build-up of misfolded proteins that cannot be efficiently eliminated, the UPR can become cytotoxic and can directly initiate programmed cell death through both caspase-dependent and independent pathways (Lin et al., 2007).
In eukaryotes, the UPR is highly conserved and comprises three parallel branches, each of them initiated by a specific ER stress sensor. Such sensor is represented by an ER resident protein, which is sensitive to ER perturbations and signals this information to the cytosol and the nucleus. There are three ER stress sensors: (1) the inositol-requiring enzyme 1 (IRE1); (2) the double-stranded RNA-activated protein kinase (PKR)-like ER kinase (PERK); (3) the activating transcription factor-6 (ATF6) (**Figure 1**).
PERK is an ER-resident type I transmembrane protein with a cytosolic kinase domain. Upon ER stress, PERK phosphorylates the α subunit of the eukaryotic translational initiation factor 2α (eIF2α) at residue Ser51 (Shi et al., 1998). Phosphorylation inactivates eIF2, disrupting the formation of GTP·eIF2α·MettRNAi ternary complex required for mRNA translation leading to a reduction in general protein synthesis and consequently to a decrease of the protein influx into the ER lumen (Harding et al., 2000a; Scheuner et al., 2001). However, in conditions of limited availability of eIF2, specific mRNAs, that contain inhibitory upstream open reading frame sequences in their 5<sup>0</sup> -untranslated region, are preferentially translated (a process called attenuation). One of these transcripts encodes for the activating transcription factor 4 (ATF4), that selectively upregulates transcription of genes involved in restoring ER functionality such as enzymes for amino acid biosynthesis and transport, protein folding and antioxidant response (Vattem and Wek, 2004; Starck et al., 2016). Within ATF4<sup>0</sup> s known targets is CHOP, a C/EBP homologous transcription factor that controls the upregulation of components involved in apoptosis (Harding et al., 2000a; Ma et al., 2002). Additionally, CHOP binds and promotes transcription of growth arrest and DNA damage-inducible protein 34 (GADD34), an inducible regulatory subunit of the protein phosphatase PP1C. PP1C dephosphorylates eIF2α, providing a feedback mechanism for tightly regulating the phosphorylation status of eIF2α and, in turn, for controlling inhibition of protein synthesis (Connor et al., 2001; Novoa et al., 2001). Another target of PERK kinase activity is NRF2, a transcription factor that induces the translation of antioxidant proteins and detoxifying enzymes (Cullinan and Diehl, 2004, 2006; Marciniak, 2004).
Similar to PERK, IRE1 is a bifunctional ER type I transmembrane protein, highly conserved through evolution and with a carboxy-terminal cytoplasmic kinase and RNase domains (Wang et al., 1998). Mammalian IRE1 has two homologs, IRE1α, an ubiquitous protein and IRE1β which is expressed specifically in the gastrointestinal and respiratory tracts (Bertolotti et al., 2001; Tsuru et al., 2013). In the presence of ER stress, both IRE1 homologs form homo-oligomers through the association of their ER luminal domain. The oligomer then cleaves a 26-base intron from the mRNA encoding the X-box binding protein-1 (XBP1) (Yoshida et al., 2001). Spliced Xbp1, sXBP1, activates downstream a wide set of genes encoding proteins involved in ER membrane biogenesis, protein folding and ERAD (Lee et al., 2003; Acosta-Alvear et al., 2007). In addition to XBP1 cleavage, IRE1 is also implicated in the degradation of specific mRNAs of membrane and secreted proteins, 28S ribosomal RNA and microRNAs as a part of a Regulated IRE1-Dependent Decay
(RIDD) pathway (Hollien and Weissman, 2006). RIDD may indirectly contribute to the reduction of protein influx in the ER. However, exhaustive RIDD activity, such as in conditions of chronic ER stress, can indiscreetly degrade messengers of protein involved in protein folding, exacerbating the overload of misfolded polypeptides and indirectly contributing to induction of cell death (Han et al., 2009; Upton et al., 2012).
Unlike PERK and IRE1, ATF6 is an ER-associated type 2 transmembrane protein with (carboxy-terminal luminal domain stress-sensing) a basic leucine zipper domain functioning as transcription factor (Haze et al., 1999). In unstressed conditions, ATF6 is an oligomer that upon activation dissociates into a monomeric form and translocates in the Golgi where it is sequentially cleaved by serine protease site-1 (S1P) and metalloprotease site-2 (S2P) to release an active cytosolic form, ATF6 (N) (Schindler and Schekman, 2009). After migration in the nucleus, ATF6 (N) binds to ER stress response element (ERSE) and activates the transcription of genes involved in ERAD, ER homeostasis and folding machinery such as the ER chaperons BiP/grp78 and Grp94 (Yoshida et al., 2000). Cross-talk between the different UPR pathways has been shown for sXBP1 mRNA whose expression can be unregulated also by PERK and ATF6 (Yoshida et al., 2001; Tsuru et al., 2016) and in the case of CHOP, whose expression appears to be stimulated also by ATF6 (N) (Yoshida et al., 2000; Tsuru et al., 2016).
How can PERK, IRE1, and ATF6 sense misfolded proteins? It is thought that each UPR sensor is maintained in an inactive or quiescent state through binding with the ER chaperon BiP/grp78 (Bertolotti et al., 2000). BiP/grp78 is part of the ER translocon pore and is involved in numerous ER-related functions, such as translocation of nascent polypeptides, protein folding, targeting of misfolded proteins to ERAD machinery and ER calcium homeostasis (Hendershot, 2004). During ER stress BiP/grp78 senses and binds misfolded proteins, dissociating from the binding to the luminal domain of UPR sensors, with concomitant activation of these signaling proteins and initiation of the three different UPR cascades. More recent evidence has suggested that IRE1 can directly sense and bind misfolded polypeptides, without the mediation of BiP/grp78, becoming activated and inducing UPR (Credle et al., 2005; Kimata et al., 2007; Gardner and Walter, 2011). It is not clear whether a similar mechanism applies also to PERK or ATF6 although selective deletion of BiP/grp78 binding site on ATF6 does not result in the constitutive activation of ATF6-dependent branch of the UPR in unstressed conditions (Shen et al., 2005).
Besides its protective function, the UPR has been recently implicated in memory and synaptic plasticity as it has been shown that the PERK-eIF2α branch or XBP1 can regulate gene expression of proteins implicated in long-term potentiation, memoryformation and synapsis remodeling (Trinh and Klann, 2013; Martínez et al., 2016). Thus it appears that the UPR is not only a rescue mechanism in case of stressful conditions in the ER but also a way to modulate normal cellular function and homeostasis.
#### ER Stress-Induced Apoptosis
When the initial cellular response fails to restore ER homeostasis and misfolded proteins overload cannot be efficiently removed, the UPR switches from an adaptive response to induce cell death. Although the mechanism and key players have not been entirely identified, it appears that the same UPR branches involved in
the initial prosurvival response have the capacity to induce apoptosis in the case of severe ER stress. Activation of CHOP by PERK/ATF4 or ATF6 or XBP1 (Harding et al., 2000a,b; Scheuner et al., 2001) appears to be central for the induction of ER stress-driven apoptotic signal. Pro-apoptotic activity of CHOP is mediated by both the upregulation of Bim, a protein that belongs to the BH3-only family (Puthalakath et al., 2007) and by the suppression of the anti-apoptotic protein Bcl-2. The BH-3 only family is comprised of proteins able to induce formation of the mitochondrial pore and, consequently, to induce the release of cytochrome c. Instead, the Bcl-2 protein family inhibits the formation of the mitochondrial outer membrane pore. For CHOP to activate the apoptotic cascade, both factors, Bim and Bcl-2, have to be regulated, although in the opposite direction. Release of cytochrome c then activates caspase-9 in the apoptosome with consecutive cleavage of caspase 3 and initiation of the apoptotic process. More recently, CHOP activity has been proposed to be mediated by PUMA, a p53-upregulated modulator of apoptosis (Galehdar et al., 2010; Ghosh et al., 2012).
In addition to CHOP, upon extensive ER stress, IRE1 can also promote apoptosis through binding to the TNF-α receptorassociated factor 2 (TRAF2) and stimulation of apoptosis signal-regulating kinase-1 (ASK1) (Nishitoh et al., 2002). ASK1 in turn, activates JNK that phosphorylates and inhibits antiapoptotic factors Bcl-2 and Bcl-XL. Interestingly, Bax and Bak, which belong to the BH-3 only family, can modulate directly IRE1 activity by relocating to the ER membrane under ER stress conditions (Zong et al., 2003; Hetz et al., 2006; Klee et al., 2009). Moreover, cells lacking Bax and Bak are resistant to apoptosis after treatment with different ER stress stimuli (Scorrano et al., 2003; Buytaert et al., 2006). Thus, exhaustive ER stress can induce cell death through a tight and well controlled cross-talk with the mitochondria.
However, besides induction of apoptosis through the mitochondria, other pathways have been proposed to take part in cellular demise upon UPR activation. Initial observations had suggested how the initiation of the UPR-dependent cell death cascade could be mediated directly by the ER through activation of the ER-resident caspase, caspase 12 (Nakagawa et al., 2000; Yoneda et al., 2001). However, inhibition of caspase 12 expression in MEF cells does not make cells more vulnerable to ER stressors indicating that caspase 12 is not specifically activated in conditions of ER stress (Obeng and Boise, 2005; Saleh et al., 2006). Also caspase 12 in humans does not appear to be functional whereas the pro-inflammatory caspase 4 seems to now be a more suitable candidate in mediating the ER stress-induced apoptosis (Lu et al., 2014). Nevertheless, the PERK–ATF4–CHOP pathway has been shown to promote ER stress-dependent apoptosis bypassing the mitochondria by recruiting cell death receptors such as TRAIL-R1/DR4 and TRAIL-R2/DR5 and their death ligands (Martín-Pérez et al., 2012; Li et al., 2015). Here, activation of such receptors leads to cleavage of pro-caspase 8 that, in turn, can selectively cleave caspase 3. Moreover, activation of the same PERK branch can also lead to autophagy, where transcriptional upregulation of autophagy-related genes such as Atg5, Atg3, Atg7, Atg10, Atg12, Atg16l1, Becn1, p62, and Nbr, is downstream of ATF4 expression (B'chir et al., 2013). In summary, in case of chronic ER dysfunction, it appears that multiple signal pathways involving the ER, mitochondria and the cytosol can contribute to the ER stress-induced cell death.
#### ER Stress and Inflammation
The UPR is also actively involved in inducing inflammation through the stimulation of the transcriptional activity of NFκB and JNK, key mediators of the proinflammatory response. Bacterial infections can induce all three branches of the UPR and activation of the UPR is necessary for production of proinflammatory cytokines (Smith et al., 2013). Stimulation of Toll-like receptors (TLRs), innate immune receptors known to sense pathogen invasion, such as TLR4 and TLR2, specifically activate the IRE-1/XBP-1 branch for production of cytokines such as IL-1β, IL-6, TNF-α and interferon in macrophages (Martinon et al., 2010; Shenderov et al., 2014). TLR4 signaling appears mediated by MyD88 and TRAF6 that interact and activate IRE-1 through ubiquitination and by blocking its inactivation by PP2A phosphatase activity (Qiu et al., 2013). More recently the activity of two other immune receptors, NOD1 and NOD2, known to sense bacterial cell wall molecules, was shown to be mediated by IRE1α activation after Brucella abortus or Chlamydia muridarum infections (Keestra-Gounder et al., 2016). In addition to IRE1, also the PERK/eIF2/CHOP pathway can mediate TLR4 signaling during inflammation (Afrazi et al., 2014). In conditions of ER stress, attenuation of global mRNA translation, mediated by the PERK/eIF2α phosphorylation, reduces the protein level of IκB, an inhibitory protein that sequesters NF-κB in a quiescent state through binding. Without IκB, NF-κB can migrate into the nucleus and can transcriptionally activate the upregulation of proinflammatory genes (Deng et al., 2004). In addition to PERK, IRE-1α can also stimulate NF-κB activity, through the recruitment of TRAF2 and consequent binding and activation of IκB kinase (IKK) (Hu et al., 2006). Phosphorylation of IκB by IKK signals selective degradation of IκB through the proteasome and promotes activation of NF-κB. BesidesNF-κB, the IRE-1α-TRAF2 complex can also induce inflammation by direct recruitment and activation of the JNK signaling and consecutive recruitment of AP-1 and transcription of proinflammatory genes (Urano et al., 2000). In addition, other mechanisms, such as the production of reactive oxygen species (ROS) in the ER, the level of glutathione and the release of intracellular Ca2<sup>+</sup> can activate NF-κB signaling inducing inflammation. Production of ROS, in the form of oxygen peroxide, occurs normally in the ER during the catalysis of disulfide bonds formation and it is mediated by two ER-resident proteins PDI and ERO1 (Görlach et al., 2015). Similarly, oxidative stress in the ER is also the result of increased consumption of glutathione, employed as reducing agent of improperly formed disulfide bonds. Thus, an increase in the ER protein load may lead to an overproduction of ROS and, in turn, may initiate an inflammatory response. To control the level of oxidative stress the PERK pathway, through NRF2 and ATF4, induces transcription of antioxidant and oxidantdetoxifying enzymes, including genes involved in regulating cellular level of glutathione (Cullinan and Diehl, 2004). Thus, ER stress through activation of the IRE1 and PERK branches can directly initiate neuronal inflammation, a key process in the
pathogenesis of neurodegenerative diseases, providing a direct link between accumulation of misfolded/aggregated protein and pro-inflammatory conditions.
#### ER STRESS AND PD PATHOGENESIS
Several reports support the link between ER stress and PD pathogenesis. One of the first of these was obtained in pharmacological neurotoxic models of PD where acute treatment with MPTP, 6-hydroxydopamine (6-OHDA) or rotenone, in cell cultures induced, although at different extent, activation of the UPR genes (Ryu et al., 2002; Holtz and O'Malley, 2003). Moreover ablation of CHOP in mice protected dopaminergic neurons against 6-OHDA, indicating that the ER stress response contributes directly to neurodegeneration in vivo (Silva et al., 2005). Specific sensitivity of the dopaminergic system to ER stress was also confirmed by more recent evidence and could partly explain how this population is particularly vulnerable to protein misfolding. For instance, inhibition of XBP1 protein expression in the substantia nigra of adult mice triggered chronic ER stress and specific neurodegeneration of dopaminergic neurons, whereas local recovery of XBP1 level through gene therapy increased neuronal survival and reduced striatal denervation after 6-OHDA treatment (Valdes et al., 2014). Similar results were obtained in mice after MPTP administration or in neuroblastoma cell lines treated with MPTP or proteasome inhibitors (Sado et al., 2009). In both cases, overexpression of XBP1 rescued neuronal cells from dying, indicating that the UPR plays a pivotal role in dopaminergic neuronal survival. In the same way knocking down ATF6 expression in mice exacerbated neurotoxicity after MPTP insult (Egawa et al., 2011). Interestingly, treatment with MPTP has been shown to induce UPR by affecting ER Ca2<sup>+</sup> homeostasis through inhibition of store-operated calcium entry (SOCE), whose activity is fundamental for maintaining ER Ca2<sup>+</sup> level (Selvaraj et al., 2012). In this context, MPTP would inhibit the expression of transient receptor potential channel 1 (TRPC1), a regulator of SOCE, decreasing Ca2<sup>+</sup> entry into the cells. Overexpression of TRPC1 protected against MPTP-induced loss of SOCE and UPR, while knocking down the gene in mice increased UPR and cell death of dopaminergic neurons. Thus, at least for MPTP, induction of UPR appears to be directly linked to Ca2<sup>+</sup> imbalance.
Activation of the ER stress response was also reported in human PD brain. Accumulation of ER chaperons was found in LBs (Conn et al., 2004) while increased PERK/p-eIF2α signaling was demonstrated in dopaminergic neurons of the substantia nigra in post-mortem tissue from PD cases, confirming that PD pathology is intimately associated with activation of ER stress in vivo (Hoozemans et al., 2007). Interestingly, at least two protective mechanisms against ER stress have been shown to involve modulation of genes such as Parkin and LRRK2, whose mutated forms have been associated with familiar cases of PD. Parkin, an E3 ubiquitin ligase implicated in the regulation of mitophagy, was found increased after treatment with ER or mitochondria stressors and this increase was mediated directly by ATF4 binding to the parkin promoter (Bouman et al., 2011; Sun et al., 2013). Overexpression of parkin protected cells from ER stress by promoting splicing of XBP-1 and the induction of the UPR prosurvival response (Duplan et al., 2013). Also mutations causing loss of function of LRRK2, a protein involved in maintaining neuronal cellular stability, have shown to abrogate upregulation of BiP/grp78 level after 6OHDA treatment or overexpression of αS, enhancing neuronal death in vitro and in vivo (Yuan et al., 2011). Thus, additional protective mechanisms may be important in preserving cellular environment from detrimental effects of ER stress whereas alteration in such pathways may contribute to PD progression.
#### ER Stress and α-Synucleinopathy
In genetic models of PD obtained by overexpression of αS, the association between ER stress and α-synucleinopathy has been studied extensively. In mammalian cell cultures, mice and yeast, toxicity due to overexpression of human wild-type, A53T mutant or C-terminal truncated αS was correlated with ER stress and activation of the UPR (Smith et al., 2005; Cooper et al., 2006; Bellucci et al., 2011; Colla et al., 2012a; Chung et al., 2013; Heman-Ackah et al., 2017). Interestingly, αS-induced dysfunctional ER and ER-stress activated cell death were both attenuated by treatment with L-DOPA, a dopamine analog and the only known treatment for PD at the moment (Song et al., 2017). In pc12 cells, overexpression of mutant αS induced cellular stress in a time-dependent matter that initiated with oxidative and proteasome damage and later culminated with ER stress and activation of ER-stress dependent cell death program (Smith et al., 2005). Blockage of caspases activity with inhibitors, siRNA or treatment with the ER stress inhibitor salubrinal, protected against A53T αS-induced cell death, indicating that the ER stress mediates αS toxicity and contributes to cellular death (Boyce et al., 2005). In addition, since proteasome and mitochondria deficits appeared before UPR activation, this suggested that onset of the ER stress response was the final protective pathway to contain αS damage before apoptosis had to be initiated.
Because αS was not known to be a resident protein of the ER, questions on how αS induces ER stress remained open until multiple observations placed αS in close proximity or within the ER and showed its direct interaction with ER and vesicular traffic components (**Figure 2**). In yeast, overexpression of mutant A53T αS caused toxicity through inhibition ER-Golgi vesicular transport that was completely abrogated by the overexpression of Rab1, a member of the Rab/GTPase family important for intracellular protein traffic modulation (Cooper et al., 2006). Interestingly, Rab1 overexpression was able to rescue dopaminergic neurons from toxicity induced by αS overexpression in other PD animal models, such as Drosophila, Caenorhabditis elegans and rat primary cultures. In yeast, vesicular transport deficit was consistent with the inhibition of docking and fusion of vesicles to the Golgi membrane and could be also rescued by overexpression of other members of the Rab family, such as Rab3A and Rab8A (Gitler et al., 2008). Rab3A and Rab8A are responsible for promoting vesicles' transport at other sites such as the presynaptic button and the post-Golgi element. This suggested that αS overexpression caused traffic defects not only at the ER-Golgi step, but also at multiple sites
in the secretion pathway, including at the plasma membrane, an observation that fits well with the physiological role of αS in promoting neurotransmission. As a matter of fact, at the synapse, αS has been described to interact with the SNARE complex and to promote vesicles docking and fusion to the membrane in the presynaptic button (Burre et al., 2010; Diao et al., 2013; Wang et al., 2014). In vivo, overexpression of wildtype αS in mice significantly inhibited neurotransmission by delaying vesicles recycling and reclustering after endocytosis at the synaptic terminal (Nemani et al., 2010). In Drosophila, overexpression of αS induced accumulation of synaptic vesicles with a larger size at the neuromuscular junction, a defect that was rescued by Rab11 (Breda et al., 2015). Interestingly, in the yeast model overexpressing αS, the observed traffic defect due to the build-up of clustered vesicles unable to fuse, initiated at the membrane level and later expanded in a retrograde manner to the Golgi and the ER. Other reports pointed out that αS could impair traffic of specific vesicles cargo, such as COPII vesicles loaded with ATF6 protein or vesicles containing lysosomal-targeted hydrolases moving from the ER to the Cis-Golgi (Credle et al., 2015; Mazzulli et al., 2016).
Additionally, other reports suggested how protein traffic deficit was not due to αS expression per se but related to an acquired toxic function. For instance, formation of aggregates in axonal terminals of primary cultures after exogenous administration of αS pre-formed fibrils did not cause initially a generalized defect in axonal transport but impaired primarily Rab7 and TrkB receptor–containing endosomes and autophagosomes cycling (Volpicelli-Daley et al., 2014). Consistent with this observation, we demonstrated in A53T αS transgenic mice how α-synucleinopathy, intended as accumulation of αS toxic species along the ER and the secretory pathway, but not the overexpression of the protein, was associated with the induction of UPR and ER stress-induced cell death in vivo (Colla et al., 2012a). Very importantly, appearance of ER-associated αS oligomers preceded α-synucleinopathy and ER stress whereas treatment with salubrinal, delayed α-synucleinopathy onset in transgenic mice and in the AA2V-A53T αS rat model and reduced the level of αS oligomers and aggregates associated with the ER, but not the total amount of αS, suggesting that accumulation of ER-associated αS species results in ER stress (Colla et al., 2012b). Additionally, αS species that accumulate along the secretory pathway appear to have specific distinct biochemical properties compared to non-membranous associated αS aggregates and can be extremely neurotoxic (Colla et al., 2018). For instance, while mature microsomes-αS aggregates isolated from diseased A53T mice exogenously added to mouse primary neurons induced endogenous aggregation and cell death, same species but isolated from presymptomatic mice, without overt α-synucleinopathy, were still cytotoxic but with lesser extent and were found unable to propagate. Such differences in behavior suggest some sort of toxic maturation of αS species right at the ER/microsomal membrane, pushing forward the hypothesis that the ER, Golgi and synaptic vesicles membranes may be a key site for αS aggregation and toxicity.
In addition, we and others have shown that αS interacts with Bip/Grp78 in physiological conditions (Bellucci et al., 2011; Colla et al., 2012a) suggesting that in case of αS aggregation, accumulation of αS aggregates along the ER membrane might directly signal distress to the ER through its interaction with BiP/grp78. Interestingly overexpression of BiP/Grp78 in rats or XBP-1 in C. elegans has been shown to alleviate ER stress and protect dopaminergic neurons from αS neurotoxicity (Gorbatyuk et al., 2012; Ray et al., 2014). On the other hand, a study in mammalian cells and αS transgenic mice reported that mutant
A53T αS induced cell death and UPR by destabilizing ER Ca2<sup>+</sup> homeostasis. Overexpression of homocysteine-inducible ER stress protein (Herp), a protein that plays a role in maintaining ER Ca2<sup>+</sup> balance, markedly reduced A53T αS-induced toxicity in mice, whereas knockdown of Herp exacerbated ER stress leading to a significant rise in toxicity (Belal et al., 2012). Similarly, αS aggregates but not the monomer have been shown to bind to and activate SERCA, an ER Ca2<sup>+</sup> pump, inducing Ca2<sup>+</sup> release in the cytosol (Betzer et al., 2018). Treatment with CPA, a SERCA inhibitor, normalized Ca2<sup>+</sup> level and was neuroprotective against αS aggregates toxicity in C. elegans. Ultimately, because ER Ca2<sup>+</sup> level is particularly sensitive to an increase in ROS, including radical species derived from dysfunctional mitochondria, toxic αS could indirectly contribute and exacerbate ER stress by impairing mitochondria metabolism and the respiratory chain (Görlach et al., 2015).
Thus, because of its preference to bind biological membranes that puts αS in direct contact with the ER/Golgi membrane and synaptic vesicles, toxic, aggregated αS is able to promote ER stress by destabilizing Ca2<sup>+</sup> homeostasis and inhibiting intracellular protein trafficking and vesicles release, affecting the whole secretory pathway and contributing to the build up of misfolded proteins in the ER with consequent impairment in ER functionality.
#### CONCLUSION
In recent years because of its importance in regulating protein homeostasis, the ER has emerged as a central organelle in
#### REFERENCES
the pathogenesis of neurodegenerative diseases. Accumulating evidence support a key role of the UPR cascade in PD progression, correlated specifically to dopaminergic neuronal death and αS toxicity. Specifically, it appears that because of αS's cellular function and its ability to interact with membranes of organelles and vesicles of the secretory pathway and intracellular protein trafficking, toxicity of αS mediated by aggregation might directly or indirectly affect ER functionality and induce ER stress. In addition because ER stress contributes to promote neuroinflammation, a central process in neurodegeneration, regulating the UPR cascade through therapy may become an efficient cellular target that can lower misfolded protein overload as well as improve inflammation. Thus, future genetic or pharmacology-based approaches tackling the UPR in order to recover ER functionality could potentially lead to disease-modifying therapy in PD and possibly other neurodegenerative diseases.
#### AUTHOR CONTRIBUTIONS
EC conceived the idea, wrote, edited, and reviewed the manuscript before submission.
#### FUNDING
This research was funded by Italian Ministry of University and Research (MIUR), Career Reintegration Grant scheme (RLM Program for Young Researchers).
protein trafficking in human midbrain synucleinopathy models. Proc. Natl. Acad. Sci. U.S.A. 113, 1931–1936. doi: 10.1073/pnas.1520335113
fnins-13-00560 May 29, 2019 Time: 9:2 # 9
produce mature IL-1 in response to TLR4 stimulation through a caspase-8- and TRIF-dependent pathway. J. Immunol. 192, 2029–2033. doi: 10.4049/jimmunol. 1302549
by transmembrane protein kinase IRE1. Science 287, 664–666. doi: 10.1126/ science.287.5453.664
**Conflict of Interest Statement:** The author declares that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Copyright © 2019 Colla. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
fnins-13-00560 May 29, 2019 Time: 9:2 # 10
# Manganese-Induced Neurotoxicity: New Insights Into the Triad of Protein Misfolding, Mitochondrial Impairment, and Neuroinflammation
Dilshan S. Harischandra† , Shivani Ghaisas† , Gary Zenitsky, Huajun Jin, Arthi Kanthasamy, Vellareddy Anantharam and Anumantha G. Kanthasamy\*
#### Edited by:
Krishnan Prabhakaran, Norfolk State University, United States
#### Reviewed by:
Aaron B. Bowman, Purdue University, United States Maria Xilouri, Biomedical Research Foundation of the Academy of Athens, Greece
> \*Correspondence: Anumantha G. Kanthasamy [email protected]
#### †Present address:
Dilshan S. Harischandra, Covance Greenfield Laboratories, Greenfield, IN, United States Shivani Ghaisas, Department of Cancer Biology, University of Pennsylvania, Philadelphia, PA, United States
#### Specialty section:
This article was submitted to Neurodegeneration, a section of the journal Frontiers in Neuroscience
Received: 10 April 2019 Accepted: 06 June 2019 Published: 26 June 2019
#### Citation:
Harischandra DS, Ghaisas S, Zenitsky G, Jin H, Kanthasamy A, Anantharam V and Kanthasamy AG (2019) Manganese-Induced Neurotoxicity: New Insights Into the Triad of Protein Misfolding, Mitochondrial Impairment, and Neuroinflammation. Front. Neurosci. 13:654. doi: 10.3389/fnins.2019.00654 Department of Biomedical Sciences, Parkinson's Disorder Research Laboratory, Iowa State University, Ames, IA, United States
Occupational or environmental exposure to manganese (Mn) can lead to the development of "Manganism," a neurological condition showing certain motor symptoms similar to Parkinson's disease (PD). Like PD, Mn toxicity is seen in the central nervous system mainly affecting nigrostriatal neuronal circuitry and subsequent behavioral and motor impairments. Since the first report of Mn-induced toxicity in 1837, various experimental and epidemiological studies have been conducted to understand this disorder. While early investigations focused on the impact of high concentrations of Mn on the mitochondria and subsequent oxidative stress, current studies have attempted to elucidate the cellular and molecular pathways involved in Mn toxicity. In fact, recent reports suggest the involvement of Mn in the misfolding of proteins such as α-synuclein and amyloid, thus providing credence to the theory that environmental exposure to toxicants can either initiate or propagate neurodegenerative processes by interfering with disease-specific proteins. Besides manganism and PD, Mn has also been implicated in other neurological diseases such as Huntington's and prion diseases. While many reviews have focused on Mn homeostasis, the aim of this review is to concisely synthesize what we know about its effect primarily on the nervous system with respect to its role in protein misfolding, mitochondrial dysfunction, and consequently, neuroinflammation and neurodegeneration. Based on the current evidence, we propose a 'Mn Mechanistic Neurotoxic Triad' comprising (1) mitochondrial dysfunction and oxidative stress, (2) protein trafficking and misfolding, and (3) neuroinflammation.
Keywords: manganese neurotoxicity, Parkinson's disease, protein aggregation, exosome, cell-to-cell transmission and neuroinflammation
#### METALS IN BIOLOGY
At least 13 metals have been identified as essential for life, and four of these (sodium, potassium, magnesium, and calcium) occur in large amounts. The remaining nine trace metals (manganese, iron, cobalt, vanadium, chromium, molybdenum, nickel, copper, and zinc) assume vital roles in building organic biomolecules as well as in regulating biological functions. In the last couple of decades, the importance of metal ions in protein biology has been an increasingly attractive
**43**
research subject given their association with many human diseases, for which metals have been identified as a causative or stimulatory agent. Metals are essential because of their integral role in enzymes that catalyze the basic metabolic or biochemical processes shared by all forms of life on earth. About onethird of all proteins depend on metal ions to carry out their biological functions (Holm et al., 1996). When considering all six classes of enzymes – oxidoreductases, transferases, hydrolases, lyases, isomerases, and ligases – over 40% of all enzymes contain metals (Andreini et al., 2008). Moreover, the chemistry of metals allows for a broader set of protein-metal reactions. For instance, redox-active metal ions are often interchangeable depending on the metal concentration and their affinities to protein. Protein affinities for trace metals are substantially determined by universal series, which for divalent metals is the Irving-Williams series (Mn2<sup>+</sup> < Fe2<sup>+</sup> < Co2<sup>+</sup> < Ni2<sup>+</sup> < Cu2<sup>+</sup> > Zn2+), wherein Cu2<sup>+</sup> is highly competitive and can replace lower order metals (Tottey et al., 2008).
These "metalloproteins" are involved in many key biological processes, such as gas transport, cell respiration, antioxidant defense, photosynthesis, and many other vital redox reactions driven by their interaction with metals. Well-characterized examples for redox-active metalloprotein systems are bluecopper proteins, heme-binding proteins and iron-sulfur-cluster proteins. Moreover, recent advances in synthetic chemistry have focused on the study of metal sites in metalloproteins and metalloenzymes to influence biological processes in the battle against many daunting human diseases. Advanced medicinal chemistry approaches have given us new, innovative medicinal applications of metal complexes and organometallic agents. Prime examples for such uses of metals include platinumcontaining anticancer drugs (e.g., Cisplatin), lithium-containing depression drugs (e.g., Camcolit), and manganese (Mn) containing anticancer drugs (e.g., SOD mimics) (Farrell, 2003).
Presumably, all metalloproteins would bind to their desired metal ligands, and this binding can regulate their folding. However, despite the wealth of structural information, the coupled protein-folding, metal-binding pathways for metalloproteins remain largely unknown (Wittung-Stafshede, 2002). Proper protein folding is critical to the conformational integrity and function of proteins. However, metal ligand binding can also induce undesirable structural transitions in proteins that eventually lead to the formation of pathological protein aggregates. Indeed, the pathologies of Alzheimer's disease (AD), PD, and prion diseases are linked to abnormal misfolding of otherwise harmless neural proteins. For example, in AD, increased levels of metals, such as Cu2<sup>+</sup> and Zn2+, are linked to the aggregation of Aβ protein in vitro (Kenche and Barnham, 2011). The theory of metal-induced aggregation is supported by numerous studies tying metal concentrations in the brain with AD, PD, and amyotrophic lateral sclerosis (ALS) in in vivo and in vitro studies employing recombinant proteins (Brown et al., 2005; Brown, 2011).
In this review, we will focus on α-synuclein, one of the major proteins implicated in PD, and its interactions with metals, specifically, its interaction with Mn in oxidative stress, protein aggregation and neurodegeneration.
### PARKINSON'S DISEASE
Parkinson's disease is recognized as the second most prevalent neurodegenerative disorder after AD, affecting roughly 1% of the population over the age of 65. It is also the most common movement disorder in the elderly, resulting in bradykinesia, resting tremor, and rigidity (Lotharius and Brundin, 2002). Several non-motor symptoms involving the autonomic nervous system have also been gaining attention (Pfeiffer, 2009; Schapira et al., 2017). PD is characterized histopathologically by the degeneration of dopaminergic neurons in the substantia nigra pars compacta (SNpc), leading to the progressive loss of the neurotransmitter dopamine and hence the above-mentioned cardinal motor deficits. Even though PD is also often associated with the abnormal accumulation of misfolded proteins, primarily α-synuclein, in cytoplasmic inclusions called Lewy bodies (LB) and Lewy neurites, the pathophysiological association between Lewy pathology and disease pathogenesis is not well understood. Similar neuropathological lesions involving the deposition of abnormal proteins also characterize other neurological disorders (Ross and Poirier, 2004), including AD (Kotzbauer et al., 2001; Uchikado et al., 2006), Lewy body dementia (LBD) (McKeith et al., 2004), Huntington's disease (HD) (Davis et al., 2014), multiple system atrophy (MSA) (Shoji et al., 2000), and some prion diseases (Aguzzi and Calella, 2009; Aguzzi and O'Connor, 2010).
Although aging remains the greatest risk factor for idiopathic PD, a small fraction of patients were identified with familial PD, which is caused by mutations in several genes associated with protein metabolism, ion transport and mitochondrial function. Genes associated with early-onset PD include α-synuclein (PARK1), parkin (PARK-2), PINK1 (PARK6), DJ-1 (PARK7) and ATP13A2 (PARK9), while those linked with late-onset PD include LRRK2 (PARK8) and VPS35 (PARK-17) (Dawson et al., 2010; Roth, 2014). A growing number of epidemiological and clinical studies have identified environmental risk factors for PD, including repeated head trauma, heavy metal toxicity, pesticide toxicity, obesity, and some surrogate measures such as rural living, contaminated well water, substance abuse, and farming (Priyadarshi et al., 2001; Dick et al., 2007). Interestingly, some of these environmental triggers and toxins induce pathophysiological features that mimic PD when they are administered in experimental animal settings. One such toxin is MPTP (methyl-4-phenyl-1,2,3,6-tetrahydropyridine), a compound produced as an impurity during synthesis of the illicit narcotic desmethylprodine. MPTP causes chronic and severe Parkinsonism by selectively damaging the SN, resulting in PD-related motor deficits (Langston et al., 1983; Ballard et al., 1985; Appendino et al., 2014). Other compounds widely used in experimental models to study the etiopathogenesis of PD include the narcotic methamphetamine, the dopamine derivative 6-hydroxydopamine, and pesticides such as rotenone, paraquat, and dieldrin. These neurotoxins cause nigrostriatal cell death by interfering with mitochondrial function, inducing oxidative stress, protein aggregation, and modifying proteasomal function (Kanthasamy et al., 2008; Latchoumycandane et al., 2011; Ghosh et al., 2013; Jin et al., 2015b). In addition, exposure to
heavy metals (e.g., iron, lead, mercury, cadmium, arsenic, and Mn) and metal-based nanoparticles increases the risk of PD through the neurotoxic accumulation of metals in the SNpc and by increasing oxidative stress-induced apoptosis (Afeseh Ngwa et al., 2009; Milatovic et al., 2009; Afeseh Ngwa et al., 2011; Kanthasamy et al., 2012; Aboud et al., 2014; Harischandra et al., 2015a).
#### MANGANESE
Manganese is considered to be a key inhaled environmental pollutant as well as a putative risk factor for environmentally linked PD and related neurodegenerative disorders. Being the 12th most abundant element and composing approximately 0.1% of the earth's crust, Mn is ubiquitously present in the environment (Martinez-Finley et al., 2012). Besides the earth's crust, other Mn sources include direct atmospheric deposition, wash-off from plant and other surfaces, leaching from plant tissues, ocean spray, and volcanic activity. Mn occurs in trace amounts in all body tissues as it is essential for many ubiquitous enzymatic reactions, including the synthesis of amino acids (AA), lipids, proteins, and carbohydrates. It also plays a key nutritional role in bone growth, fat and carbohydrate metabolism, blood sugar regulation, and calcium absorption (Bowman et al., 2011). Being present in whole grains, rice, nuts, tea, leafy green vegetables, and Mn-containing nutritional supplements, the primary route of Mn exposure in humans is through dietary intake. The abundance of Mn-enriched food in the typical daily diet makes it relatively easy to accrue the daily reference intake (DRI) of 2.3 mg/day for men and 1.8 mg/day for women (Aschner and Aschner, 2005), thereby minimizing the risk of Mn deficiency-related birth defects, impaired fertility, osteoporosis, and enhanced susceptibility to seizures (Dendle, 2001; Aschner and Aschner, 2005; Sarban et al., 2007).
Despite its nutritional benefits, prenatal and postnatal overexposure to Mn affects infant neurodevelopment, exemplifying its role as both an essential nutrient and a toxicant (Zota et al., 2009; Claus Henn et al., 2010). High Mn exposure in early life is associated with poor cognitive performance, especially in the verbal domain of children (Menezes-Filho et al., 2011). In older cohorts, chronic excessive exposure to occupational or environmental sources of Mn causes manganism, which is characterized by a severe neurological deficit that often resembles the involuntary extrapyramidal symptoms associated with PD (Kwakye et al., 2015). Couper (1837), at the University of Glasgow, reported the first case of Mn-induced neurotoxicity, which was discovered in employees of Charles Tennant and Co., a manufacturer of bleaching powder. Later, public awareness of Mn neurotoxicity arose as more clinical studies identified a PD-like syndrome in workers employed at a Mn ore-crushing plant and a ferromanganese factory (Cook et al., 1974; Huang et al., 1989). In addition, Rodier (1955) detailed clinical features of manganese neurotoxicity in Moroccan miners. Since then, the commercial applications for Mn have broadened considerably so that now Mn exposure also occurs through its use as an additive in gasoline (methylcyclopentadienyl manganese tricarbonyl, MMT) and fertilizers, and as manganese violet in paint and cosmetics (Martinez-Finley et al., 2012). Mn neurotoxicity occurs often in agricultural workers exposed to organic Mn-containing pesticides, such as manganese ethylene-bis-dithiocarbamate (Maneb) and in chronic abusers of the street drug 'Bazooka', a cocaine-based drug contaminated with manganese carbonate (Ensing, 1985). The other major anthropogenic sources of environmental Mn include municipal wastewater discharge, welding, mining and mineral processing, metal (alloy, steel, and iron) manufacturing emissions, fossil fuel combustion, and drycell manufacturing. Although the precise mechanisms through which Mn is absorbed into the body are not fully understood, it is known to accumulate predominantly in the brain's basal ganglia region. Beyond the many commonalities shared between manganism and PD, it is also worth pointing out their differences. Behaviorally, manganism is mainly characterized by milder and less frequent resting tremor that tends to be postural or actional, a propensity to fall backward, excessive salivation, and frequent dystonia consisting of facial grimacing, hand dystonia, and/or plantar flexion (Calne et al., 1994). Manganism patients were also reported to have symptoms of irritability, emotional lability, and hallucinations and psychoses referred to as "manganese madness" (Huang, 2007). Pathologically, Mn neurotoxicity affects primarily neurons in both the globus pallidus and striatum, whereas PD predominantly affects dopaminergic neurons in the SNpc (Roth, 2014). Therefore, in fact, the PD-like behavior deficits in manganism result from Mn's capability to suppress dopamine release from the striatum, thus generating behavioral dysfunctions common to both PD and manganism (Kim et al., 2002; Racette et al., 2005; Fitsanakis et al., 2006; Roth et al., 2013).
#### MANGANESE HOMEOSTASIS
The homeostasis of Mn and other metal ions is maintained through tightly regulated mechanisms of uptake, storage, and secretion that strictly limit their abundance in the cellular compartment. The distribution and neurotoxicity of Mn is governed largely by the routes of exposure, which are primarily ingestion and inhalation. In humans, the primary route of exposure is through Mn-enriched food and well water. However, the molecular mechanisms of oral Mn absorption are not well understood. Roughly 3–5% of the Mn ingested gets absorbed into the body from the gastrointestinal tract (GIT) (Finley et al., 1994). Under homeostatic conditions, Mn enters the portal circulation through either passive diffusion (Bell et al., 1989) or active transport via divalent metal transporter 1 (DMT1) (Erikson and Aschner, 2006; Fitsanakis et al., 2007), which was the first mammalian transmembrane iron transporter to be identified. Formerly known as Nramp2 or DCT1, DMT1 is a 12-transmembrane domain protein responsible for the uptake of various divalent metals including Fe2+, Mn2+, Zn2+, Co2+, and Ni2+, and it transfers iron across the apical surface of intestinal cells and out via transferrin (Tf)-cycle endosomes (Andrews, 1999). Besides using a mechanism similar to that for iron, there are no known metal transporters specific for transporting Mn
into cells. In plasma, approximately 80% of Mn2<sup>+</sup> is bound to α-macroglobulin or albumin, while only a small fraction (<1%) of Mn3<sup>+</sup> is bound to Tf. It has been proposed that, like iron, Mn in plasma is oxidized from Mn2<sup>+</sup> to the Mn3<sup>+</sup> valence state by the ferroxidase enzyme ceruloplasmin and loaded onto plasma Tf for circulating into tissues (Davidsson et al., 1989). Circulating Mn diffuses throughout the body, including bone, kidney, pancreas, liver, and brain (Martinez-Finley et al., 2012).
Once in the brain, Mn3<sup>+</sup> entry into neurons occurs by the Tf-Mn3<sup>+</sup> complex binding to the transferring receptor (TfR) and it becomes localized in endosomes. Subsequent recruitment of v-ATPases acidifies endosomes and dissociates Mn3<sup>+</sup> from the Tf/TfR complex, reducing it to Mn2+, which is quite stable at physiological pH, and thereafter, neuronal transport occurs via DMT1 independent of the Tf pathway. In the brain, DMT1 is highly expressed in the DA-rich basal ganglia, putamen, cortex, and SN (Huang et al., 2004; Salazar et al., 2008), which may account for Mn's pattern of accumulation and neurotoxicity. Other primary transport mechanisms for Mn are through capillary endothelial cells of the blood-brainbarrier (BBB) (Crossgrove et al., 2003) or through the CSF via the choroid plexus (Murphy et al., 1991). Since Mn neurotoxicity primarily occurs through occupational exposure, such as inhalation of Mn fumes or dust in welding, dry-cell battery manufacturing, and the smelting industry, the nasal passage through the olfactory epithelium to the olfactory nerve is another major Mn transport mechanism into the brain (Tjalve et al., 1996). In fact, DMT1 is highly expressed in the olfactory epithelium and is required for Mn transport across the olfactory epithelium, as has been shown in the rat (Thompson et al., 2007). Evidence also exists for Mn transport into the central nervous system (CNS) through store-operated calcium channels (Crossgrove and Yokel, 2005), ionotropic glutamate receptor calcium channels (Kannurpatti et al., 2000), and Mn citrate transporters (Crossgrove et al., 2003).
Another mechanism regulating Mn homeostasis in the brain involves Mn being transferred with high affinity into cells by the Zinc transporters ZIP-8 and ZIP-14, which are Zrt-/Irt-related protein (ZIP) family metal transporters encoded by SLC39A8 and SLC39A14, respectively. These transporters are highly expressed in the liver, duodenum, kidney, and testis, and are localized on apical surfaces of brain capillaries (Girijashanker et al., 2008; Wang et al., 2012). Taking advantage of its particular magnetic properties, Aoki et al. (2004) employed magnetic resonance imaging (MRI) to show that Mn uptake also occurs through the choroid plexus. One day after they systemically administered Mn<sup>2</sup> <sup>+</sup> to rats, the distribution of Mn in the brain extended to the olfactory bulb, cortex, basal forebrain, and basal ganglia, overlapping specific brain structures vulnerable to Mn-induced neurotoxicity (Aoki et al., 2004). In cells such as neurons and astrocytes, toxic accumulations of Mn are found primarily in the mitochondria, heterochromatin, and nucleoli (Lai et al., 1999; Morello et al., 2008).
Mn also shares the Ca2<sup>+</sup> uniporter mechanism and the rapid mode (RaM) of Ca2<sup>+</sup> uptake of mitochondrial calcium influx, resulting in Mn sequestration in mitochondria, which gets removed only very slowly from the brain (Gavin et al., 1990). This Mn accumulation inhibits the efflux of calcium, decreases MAO activity, and inhibits the respiratory chain and ATP production (Zhang et al., 2003), which may partly explain the role of mitochondrial dysfunction in Mn neurotoxicity. Previously, Mn detoxification and efflux from cells was thought to be primarily regulated by ferroportin (Fpn), also known as HFE4, MTP1, and IREG1, which are proteins encoded by the SLC40A1 gene. Although Fpn was initially identified as the iron exporter, more recent findings suggest that Fpn also interacts with Mn, zinc, and cobalt to export them from the cell (Troadec et al., 2010; Yin et al., 2010; Madejczyk and Ballatori, 2012). Furthermore, Mn exposure increases Fpn mRNA levels in mouse bone marrow macrophages (Troadec et al., 2010) and it significantly increases Fpn protein levels in HEK293T cells (Yin et al., 2010). Increasing Fpn levels were linked to reduced Mn accumulation in both the cerebellum and cortex of mice treated with Mn (Yin et al., 2010), further confirming that Fpn removes Mn and reduces Mn-induced neurotoxicity.
Recently, the secretory pathway of the Ca2+/Mn2<sup>+</sup> ATPases SPCA1 and SPCA2, which are localized at the Golgi, was suggested as an alternative way of cytosolic Mn detoxification by sequestering into the Golgi lumen (Sepulveda et al., 2009). Overexpressing SPCA1 in HEK293T cells conferred tolerance of manganese (Mn2+) toxicity by facilitating Mn2<sup>+</sup> accumulation in the Golgi, thereby increasing cell viability (Leitch et al., 2011). However, the degree to which SPCA1 and SPCA2 regulate Mn homeostasis has yet to be determined. Another mode for Mn egress through Golgi has been attributed to SLC30A10 in humans (Tuschl et al., 2012). Recently, SLC30A10 was shown to be localized on the cell surface where it acted as a Mn efflux transporter to reduce cellular Mn levels and protect against Mn-induced toxicity (Leyva-Illades et al., 2014). Mutations in the SLC30A10 gene have been associated with hepatic cirrhosis, dystonia, polycythemia, Parkinsonian-like gait disturbances, and hypermanganesemia in cases unrelated to environmental Mn exposure (Tuschl et al., 2012). A genome-wide association study mapping genes involved in regulating Mn homeostasis mapped serum Mn levels to SLC30A10 (Ng et al., 2015). Along with its expression in the liver and CNS, SLC30A10 is also expressed in the GIT. Interestingly, this transporter is present mainly on the apical surface of enterocytes that line the GIT and presumably help transport Mn to the lumen. In fact, it is the liver and GIT that are primarily responsible for maintaining Mn homeostasis in the body as indicated by whole body as well as endoderm-specific SLC30A10 knockouts (KOs) resulting in hypermanganesemia, while pan neuronal/glial SLC30A10 KOs produce normal levels of Mn in the CNS (Taylor et al., 2019). The authors also found that a lack of SLC30A10 in the CNS led to an increased accumulation of Mn in the basal ganglia and thalamus when these mice were exposed to elevated Mn levels. Importantly, these recent discoveries involving SLC30A10 and its mutations reinforce its crucial role in humans as a Mn transporter, broadening our understanding of familial Parkinsonism as a result of SLC30A10 mutations.
The p-type transmembrane ATPase protein ATP13A2 (or PARK9) located at the lysosome also protects cells from Mn-induced toxicity (Tan et al., 2011). Although ATP13A2's
function in mammalian cells remains elusive, loss-of-function mutations in ATP13A2 cause Kufor-Rakeb Syndrome (KRS), an autosomal recessive form of early-onset Parkinsonism with pyramidal degeneration and dementia (Ramirez et al., 2006). Overexpression of wild-type (WT) ATP13A2, but not KRS pathogenic ATP13A2 mutants, protects mammalian cell lines and primary rat neuronal cultures from Mn2+-induced cell death by reducing intracellular Mn concentrations and cytochrome c release, suggesting a role of ATP13A2 in Mn detoxification and homeostasis (Tan et al., 2011). A summary of the abovementioned receptors and channels involved in cellular Mn homeostasis appears in **Figure 1**.
#### MANGANESE AND α-SYNUCLEIN PROTEIN MISFOLDING
Belonging to a family that includes β- and γ-synuclein, α-synuclein (αSyn) is a small 140-AA, highly conserved vertebrate protein encoded by a single 7-exon gene located on chromosome 4. It is predominantly a neuronal protein expressed in presynaptic terminals throughout the mammalian brain and CSF where it is estimated to account for as much as 1% of total protein in soluble cytosolic brain fractions. Functionally, αSyn remains poorly understood, but emerging evidence points to roles in membrane trafficking, dopamine regulation, and synaptic plasticity. The link between αSyn and PD pathogenesis is based on case studies of familial and sporadic PD patients presenting with misfolded αSyn-rich Lewy pathology during autopsy (Poulopoulos et al., 2012). Also, compelling evidence demonstrates that mutations in the gene encoding αSyn are directly linked to the onset of PD (Liu et al., 2012). Furthermore, rare familial forms of PD also have been linked to the overexpression of αSyn due to SNCA gene duplication and triplication.
The aggregation and fibrillation of αSyn, forming intracellular proteinaceous aggregates, have been implicated in several other neurodegenerative disorders besides PD, including LBD, Lewy body variant of AD, MSA, and Hallervorden–Spatz disease. The idea that extracellular αSyn species can accelerate the spread of PD pathology throughout the brain gained much consideration with the findings of host-to-graft propagation of αSyn-positive Lewy pathology in fetal ventral mesencephalic and embryonic nigral neurons transplanted in human PD patients (Kordower et al., 2008; Li et al., 2008) and misfolded αSyn species in human CSF and plasma (El-Agnaf et al., 2003; Kordower et al., 2008). Although multiple studies have hypothesized the intercellular transmission of pathological αSyn species in PD (Lee et al., 2008; Desplats et al., 2009; Dunning et al., 2013), its exact mechanistic role in disease pathogenesis and related synucleinopathies largely remains unknown. Available in vitro evidence thus far postulates that extracellular αSyn induces pathogenic actions by multiple mechanisms including, but not limited to, the triggering of neuroinflammatory responses and mitochondrial dysfunction leading to neurodegenerative processes (Su et al., 2008; Emmanouilidou et al., 2010).
As a member of the family of intrinsically unstructured proteins, αSyn is natively unfolded and lacks a defined secondary protein structure. However, upon interaction with lipid membranes, it adopts an α-helical conformational change, and under conditions that trigger aggregation, αSyn undertakes the characteristic crossed β-conformation and self-aggregates into soluble oligomers, which gradually form insoluble amyloidlike fibrils. The αSyn protein comprises three main structural domains (**Figure 2**): (1) an N-terminal amphipathic region, (2) an amyloid-binding central domain (NAC), and (3) a C-terminal acidic tail. The N-terminus (residues 1-60) contains four series of 11-AA repeats containing the highly conserved consensus sequence KTKEGV, which also is important for α-helix conformation upon binding to phospholipid membranes. The core NAC region (residues 61–91) is important in protein aggregation and it also contains two additional KTKEGV repeats. Within the NAC, a hydrophobic GAV peptide motif (residues 66–74), consisting of Ala, Val, and Gly AA residues, has been identified as the required core for human αSyn protein fibrillization and cytotoxicity (Du et al., 2006). Finally, the proline-rich C-terminus (residues 91–140) is highly acidic and accounts for the intrinsically disordered properties of αSyn (Harischandra et al., 2015a). The N-terminal and NAC regions form αSyn's membrane binding domain, whereas the C-terminal region is believed to contain protein–protein and protein–small molecule interaction sites.
Importantly, αSyn wields its metalloprotein properties through its three metal-binding sites: two each at the N-terminus and one at the C-terminus. A systematic analysis of mono-, di-, and trivalent metal ligands (Li+, K+, Na+, Cs+, Ca2+, Co2+, Cd2+, Cu2+, Fe2+, Mg2+, Mn2+, Zn2+, Co3+, Al3+, and Fe3+) revealed that metal binding induces conformational changes that cause normally benign αSyn protein to aggregate (Uversky et al., 2001). Of the 15 metal cations studied, Uversky et al. (2001) determined Al3<sup>+</sup> to be the most effective stimulator of protein fibril formation followed by Cu2+, Fe2+, Co3+, and Mn2+, with each causing conformational changes detectable by intrinsic protein fluorescence and far UVcircular dichroism. Furthermore, Uversky's team also showed that Mn3<sup>+</sup> induced immediate di-tyrosine formation, suggesting that Mn is responsible for the metal-induced oxidation of αSyn. Among the three metal-binding sites, those located at the N-terminus, specifically the <sup>1</sup>MDVFMKGLS<sup>9</sup> and <sup>48</sup>VAHGV<sup>52</sup> regions, demonstrated high-affinity binding for Cu2<sup>+</sup> (K<sup>d</sup> ∼ 0.1 µM) (Rasia et al., 2005), whereas metal-interaction sites near residues 49–52 and residues 110–140 are known to bind with divalent metals like Mn (Uversky et al., 2001; Binolfi et al., 2006, 2008). In a detailed study, the metal ions Mn2+, Fe2+, Co2+, and Ni2<sup>+</sup> bound preferentially to the <sup>119</sup>DPDNEA<sup>124</sup> motif, in which Asp121 acted as the main anchoring site with low affinity (mM) to metal ligands (Binolfi et al., 2006). These discoveries on the structural components of αSyn's interaction with metals strengthen the link between metal neurotoxicity and PD, further suggesting that metal dyshomeostasis plays an even more important role in the development of neurodegenerative disorders than previously acknowledged (Binolfi et al., 2006).
Our in vitro studies show that physiological levels of human WT αSyn attenuate acute Mn-induced dopaminergic neuronal degeneration. However, this neuroprotective effect is diminished by chronic exposure to Mn toxicity, which accelerates αSyn misfolding (Harischandra et al., 2015a). Furthermore, using a genetically modified Caenorhabditis elegans model system, Bornhorst et al. (2014) reported enhanced Mn accumulation and oxidative stress in pdr1 and djr1.1 mutants, which were reduced by αSyn expression. This protective role of αSyn in Mn-induced neurotoxicity was further validated using αSyn transgenic animals (Yan et al., 2019). By treating αSyn KO (αSyn−/−) and WT (αSyn+/+) mice with different Mn concentrations, this study demonstrated that the presence of αSyn ameliorates high-dose Mn-induced neurotoxicity. Taken together, these findings point to a novel, neuroprotective role of WT αSyn in attenuating acute Mn
toxicity, an effect which may stem directly from its metal-binding capability (**Figure 3**).
Although the physiological role of αSyn with respect to Mn toxicity still needs to be fully validated, the effects of Mn on αSyn expression, aggregation, and subsequent cytotoxicity have been studied in in vitro, in vivo, and ex vivo models of PD (Gitler et al., 2009; Verina et al., 2013; Xu et al., 2013). Studies conducted with neuronal cell culture models show that Mn treatment upregulates cellular αSyn levels and leads to αSyn aggregation (Cai et al., 2010). In contrast, knocking down αSyn using antisense αSyn treatment (Li et al., 2010) or siRNA (Cai et al., 2010) can reverse Mn-induced cytotoxicity. In parallel studies, overexpressing αSyn in rat mesencephalic cells (MES 23.5) not only enhanced their susceptibility to Mn exposure (Prabhakaran et al., 2011), but also attenuated Mn release from Mn-treated cells without significantly attenuating the major Mn transporter proteins DMT1, VGCC, and Fpn1 (Ducic et al., 2015). Thus, these studies further suggest that αSyn's metal-binding capacity serves as an intracellular Mn store that helps to regulate free-roaming Mn cations.
### MANGANESE AND ENDOSOMAL TRAFFICKING
Accumulating evidence indicates that secretion and cell-tocell trafficking of pathological forms of αSyn may explain the typical progression of PD. In particular, vesicular trafficking has attracted considerable attention as an initiator or enhancer of the neurodegenerative process underlying PD. Dysfunction of the cellular trafficking pathway can compromise synaptic function and lead to the accumulation of misfolded αSyn. Similarly, changes in endosomal sorting and degradation greatly influence the intracellular trafficking of misfolded proteins, thereby enabling the cell-to-cell transmission of toxic αSyn species in a prion-like manner. Recent genetic studies also suggest that defects of endolysosomal function could disrupt αSyn homeostasis and mitochondrial function, causing neurotoxicity through unknown mechanisms (Kett and Dauer, 2016). Indeed, several PD-linked gene mutations or polymorphisms (DNAJC13/RME-8, VPS35, ATP13A2, ATP6AP2, RAB7L1, GBA, GAK, LRRK2) interrupt protein trafficking and degradation via the endosomal pathway (Perrett et al., 2015), highlighting the importance of the endosomal pathway in the progression of neurodegenerative disease.
It has been shown that αSyn overexpression blocks endoplasmic reticulum (ER)-to-Golgi vesicular trafficking (Cooper et al., 2006) and that αSyn is functionally associated with endocytic vesicular trafficking, retromer complex proteins, phosphatases, and Rab GTPases (Breda et al., 2015; Chung et al., 2017). In this regard, recent attempts to identify molecular regulators of αSyn oligomerization have identified several Rab proteins, including Rab8b, Rab11a, Rab13, Slp5
overexposure leads to progressive protein misfolding in the neurons and induces inflammation and finally neurodegeneration.
(Goncalves et al., 2016), and Rab1, which (Cooper et al., 2006) promote the clearance of αSyn inclusions and attenuate αSyninduced toxicity. Furthermore, Rab11a and Rab13 expression enhanced the endocytic recycling and secretion in cells accumulating αSyn inclusions (Goncalves et al., 2016). In contrast, Rab11 regulates the recycling of extracellular αSyn (Liu et al., 2009) and modulates αSyn-mediated defects in synaptic transmission and locomotor behavior in experimental PD models (Breda et al., 2015). This is particularly interesting as Rab11 has been identified as a major regulator of endosomal recycling (Grant and Donaldson, 2009) and controls the secretion of smaller αSyn oligomers by exosomes (Poehler et al., 2014).
Exosomes are nano-sized vesicles (50–150 nm) that are released from cells into the extracellular space (Thery et al., 2002). Exosomes circulate throughout the body and readily cross the blood–brain and other barriers. Great interest in exosomes is emerging because of their potential role in disease progression as well as their possible use in early biomarker discovery (Sarko and McKinney, 2017) and drug delivery (Luan et al., 2017). Toxicology researchers are building upon the discovery that environmental toxicants change the exosome signature of human health conditions such as cancer and neurodegenerative diseases (Harischandra et al., 2017; Munson et al., 2018; Ngalame et al., 2018). In this regard, the impact that Mn exposure has on the neuronal exosome signature and its subsequent effect on neuroinflammation and neurodegeneration have been studied in great detail in our laboratory (Harischandra et al., 2015a,b, 2017, 2018, 2019). We have shown that Mn exposure significantly upregulates the small GTPase Rab27a, which mediates the membrane fusion of multivesicular bodies (Pfeffer, 2010) that subsequently release exosomes into the extracellular environment (Harischandra et al., 2018). Furthermore, our miRNA profiling analysis of Mn-induced neuronal exosomes indicates increased expression of certain miRNAs (e.g., miR-210, miR-325, miR-125b, miR-450b) known to control key biological mechanisms, including inflammation, autophagy, protein aggregation, and hypoxia (Harischandra et al., 2018). In subsequent studies, we show how Mn exposure promotes the exosomal secretion of aggregated αSyn into the extracellular medium. These exosomes were endocytosed through caveolaemediated endocytosis, thereby inducing neuroinflammation that subsequently evoked neurodegenerative processes in both cell culture and animal models (Harischandra et al., 2019). Interestingly, serum exosome samples collected from welders chronically exposed to Mn-containing welding fumes show increased misfolded αSyn in their exosomes, further implicating environmental Mn exposure in developing Parkinsonism (Harischandra et al., 2019). In parallel studies, we revealed Mn's role in inflammasome activation in microglial cells. We found that Mn acts as signal 2 for NLRP3 inflammasome activation in LPS-primed microglial cells, triggering the exosomal release of ACS "prionoids," resulting in inflammasome propagation (Sarkar et al., 2019). Together, our results highlight Mn's role in modulating endosomal trafficking through the exosomal release of cargo capable of triggering neuroinflammation and progressive neurodegeneration.
### MANGANESE AND NEUROINFLAMMATION
In addition to the importance of oxidative stress in the Mn-induced dysfunction of dopaminergic neurons, glial cell activation also plays an important role in potentiating Mn neurotoxicity by inducing the release of non-neuronal-derived ROS and inflammatory mediators such as proinflammatory
cytokines. The state of glial activation is defined by its morphology and by the proliferation, migration and expression of immune modulatory molecules. The two major types of glial cells in the CNS are astrocytes and microglia, with the latter constituting about 10% of all glial cells in the CNS.
It is now well documented that glial activation is prominent in the brains of humans exposed to Mn, as well as in non-human primate and rodent models of Mn neurotoxicity (Erikson and Aschner, 2006; Huang, 2007; Perl and Olanow, 2007; Cordova et al., 2013). Neuroinflammation is regarded as a key mediator in mechanisms underlying the loss of dopaminergic neurons in PD. The activation of microglia plays a major role in the response to environmental stress and immunological challenges by scavenging excess neurotoxins, removing dying cells and cellular debris, and releasing proinflammatory cytokines (Carson et al., 2007; Tansey et al., 2008). Inducible nitric oxide synthase (iNOS), which produces large amounts of nitric oxide (NO), is released by microglia in response to inflammatory mediators such as LPS and cytokines. The levels of NO are reported to be elevated in the CNS of human PD cases and in animal models of PD (Mogi et al., 1994). Consistent with this finding, iNOS KO animals are resistant to MPTP-induced dopaminergic neuronal loss in the SN (Przedborski and Vila, 2003). The transcription factor NF-κB, required for transcribing proinflammatory molecules, is also activated in the SN of PD patients and MPTP-treated mice (Ghosh et al., 2007). In contrast to microglia, astrocytes do not attack any pathological targets, but instead produce factors that mediate inflammatory reactions seen in the SN of PD brains (Miklossy et al., 2006). Activated astroglial cells were recently found in human PD brains and in the MPTP mouse model of PD (Ghosh et al., 2007; Ghosh et al., 2009).
Astrocytes play a major role in Mn-induced neuroinflammation as they represent a "hub" for brain Mn homeostasis (Wedler and Denman, 1984). The transferrin receptors found on astrocytes readily bind to Tf-Mn3+, so it is not surprising to find more Mn in astrocytes than in any other neural cell types. Indeed, astrocytes can exhibit Mn concentrations 10- to 50-fold greater than those measured in neurons, making them more susceptible to Mn toxicity than other cell types. During glutamate-induced excitotoxicity, excess glutamate abruptly increases intracellular Ca2<sup>+</sup> to levels that block Mn2<sup>+</sup> uptake, prompting a release of mitochondrial Mn2<sup>+</sup> into the cytosol. High levels of cytosolic Mn2<sup>+</sup> in astrocytes activate glutamine synthetase, which removes excess glutamate (Wedler et al., 1994). However, excessive extracellular Mn2<sup>+</sup> can disrupt intracellular Ca2<sup>+</sup> signaling in astrocytes by competitively occupying Ca2+-binding sites, thus interfering with mitochondrial Ca2<sup>+</sup> homeostasis (Farina et al., 2013), which triggers astrogliosis. In addition, Mn3<sup>+</sup> causes astrocyte swelling via oxidative/nitrosative pathways (Rama Rao et al., 2007). Increased Mn levels in astrocytes elevate the expression of proinflammatory signals such as iNOS and IL-6 (Moreno et al., 2008). In vitro studies show that Mn-treated astrocytes use larger amounts of L-arginine, which is a substrate for NO (Hazell and Norenberg, 1998). While timely expression of these signals is necessary in response to neuronal stress or cellular damage, excessive production is counter-productive, often exacerbating the toxic insult. Microarray gene expression profiling of primary human astrocytes exposed to Mn reveals an upregulation of genes encoding proinflammatory cytokines with a concurrent downregulation of genes involved in cell cycle regulation and DNA replication and repair (Sengupta et al., 2007).
The glutamate-GABA cycle (GGC) is important especially in the context of astrocyte-neuron metabolism. The AA glutamine is a precursor for the production of both glutamate and GABA (Bak et al., 2006). Deamidation of neuronal glutamine to glutamate produces ammonia, which is then transferred to astrocytes and utilized in the amidation of glutamate. Glutamine released by astrocytes is taken up by glutamatergic and GABAergic neurons that incidentally show projections in the basal ganglia and help regulate voluntary movements (Sidoryk-Wegrzynowicz and Aschner, 2013). However, in response to excessive Mn in the brain, Mn rapidly enters astrocytic mitochondria. As mentioned in the previous section, high levels of mitochondrial Mn impair cellular respiration and prevent the production and activation of glutathione peroxidase (GPx). Taken together, astrocytes appear to be particularly affected by a disruption of Mn homeostasis in the brain. This in turn could negatively affect GABAergic and glutamatergic projections in the basal ganglia, leading to the motor deficits characterizing Mn neurotoxicity.
### MANGANESE IN OXIDATIVE STRESS AND NEURODEGENERATION
Although the mechanisms of Mn-induced nigrostriatal cell death are not well characterized, Mn neurotoxicity appears to be regulated by multiple factors, including oxidative injury, mitochondrial dysfunction, protein misfolding, and neuroinflammation.
Mn is a redox-active metal whose high reduction potential aids the removal of harmful byproducts of oxygen metabolism, such as superoxide (O<sup>2</sup> .−) and hydrogen peroxide (H2O2), when as a cofactor it forms manganese superoxide dismutase (MnSOD). However, when allowed to accumulate, Mn exacerbates oxidative damage. At just 2% of body weight while consuming 20% of the total oxygen and calories, the brain is highly metabolically active and hence highly susceptible to oxidative damage. Since Mn is known to accumulate in the globus pallidus and striatum, these regions are especially vulnerable to oxidative injury because of their intense oxygen consumption, significant dopamine content, and their high content of non-heme iron. A recent study evaluating the effect of Mn on dopamine transporter (DAT)-transfected and non-transfected HEK cells shows that Mn prevents dopamine reuptake in transfected cells and also mobilizes DAT receptors from the cell surface to intracellular compartments. Consequently, dopamine-induced cell toxicity is observed (Roth et al., 2013). Our laboratory systematically characterized the cell signaling mechanisms underlying Mn-induced oxidative stress. We showed that Mn treatment in rat-derived mesencephalic dopaminergic neuronal (N27) cells increases reactive oxygen species (ROS) production (Harischandra et al., 2015a) that can sequentially activate proapoptotic processes like mitochondrial cytochrome
c release, caspase-3 activation, and DNA fragmentation. This mitochondria-dependent apoptotic cascade did not involve caspase-8 activation, but was triggered by the Mn treatment (**Figure 4**) (Latchoumycandane et al., 2005). Moreover, the redox-sensitive protein kinase C delta (PKCδ), involved in neurodegenerative disorders such as AD, prion disease, and PD (Kanthasamy et al., 2006; Ciccocioppo et al., 2008; Jin et al., 2011; Harischandra et al., 2014), is reported to be a key mediator in Mn-induced apoptosis (Anantharam et al., 2002; Latchoumycandane et al., 2005). Later studies in differentiated N27 cells also demonstrate that chronic low-dose Mn exposure impairs tyrosine hydroxylase (TH), the rate-limiting enzyme in dopamine synthesis, through activation of PKCδ and protein phosphatase-2A (PP2A) activity (Zhang et al., 2011). Notably, in vitro and in vivo administration of the hydrophilic antioxidant vitamin E analog trolox (6-hydroxy-2,5,7,8-tetramethylchroman-2-carboxylic acid) reverses Mn-induced neurotoxicity and rescues dysfunctional dopaminergic transmission and Mninduced motor coordination deficits (Milatovic et al., 2011; Cordova et al., 2013), further emphasizing the relationship between oxidative stress and Mn-related neurodegeneration.
The neurotransmitter dopamine belongs to the catecholamine and phenethylamine families. The chemical structure of catecholamines predisposes them to oxidation, and their well-characterized metabolic routes can yield quinones and free radicals, suggesting that dopamine may also serve as a neurotoxin contributing to the neurodegenerative process through oxidative metabolism. By promoting dopamine auto-oxidation, Mn potentiates dopamine toxicity in high Mn-accumulating areas of the brain (e.g., globus pallidus and striatum). Under homeostatic conditions, monoamine oxidases (MAO) enzymatically oxidize dopamine to produce dihydroxyphenylacetic acid (DOPAC), which catechol-O-methyltransferase (COMT) methylates to homovanillic acid (HVA). Alternatively, COMT can convert dopamine to 3-methoxytyramine (3-MT), which MAO then oxidizes to HVA. H2O<sup>2</sup> is another byproduct of this dopamine turnover or deamination, generating inherent oxidative stress conditions in the nigrostriatal system. Dopamine can also be non-enzymatically oxidized by molecular oxygen, yielding H2O<sup>2</sup> and quinones. These quinones also undergo intramolecular cyclization and oxidative reactions to produce neuromelanin (Graham, 1978; Hermida-Ameijeiras et al., 2004). In dopaminergic SN neurons, neuromelanin augments dopamine's vulnerability to auto-oxidation through quinone modification (Graham, 1978). Therefore, the degradation of dopamine, either enzymatically or non-enzymatically, produces H2O2. Two prominent Mn valence states, Mn2<sup>+</sup> and Mn3+, are found in biological systems. In the presence of high levels of divalent Mn2+, H2O<sup>2</sup> can convert to highly toxic hydroxyl radicals (·OH) via the Fenton reaction. But because of its higher oxidative state, Ali et al. (1995) found Mn3<sup>+</sup> to be an order of magnitude more cytotoxic than Mn2<sup>+</sup> in Mn-dosed rats. In fact, Mn3+-induced dopamine oxidation, generating quinones and H2O2, appears to be independent of oxygen and far more rapid than that mediated by Mn2<sup>+</sup> (Archibald and Tyree, 1987). Since Mn2<sup>+</sup> can readily oxidize to Mn3<sup>+</sup> in the human brain via superoxides, the auto-oxidation of catecholamines can only further potentiate oxidative stress.
Impairment of the cellular antioxidant machinery, causing an imbalance between ROS generation and its elimination, plays a major role in the development of certain neurodegenerative processes. The antioxidant glutathione (GSH), present in both neurons and astrocytes, provides the first line of cellular defense against ROS. GSH actively disposes of exogenous peroxides by acting as a co-substrate in reactions catalyzed by GPx, thus playing important functional roles in the CNS. Altered striatal concentrations of GSH, glutathione disulfide (GSSG), ascorbic acid, malondialdehyde (MDA), and the activities of glutathione reductase (GR) and GPx have been previously reported with Mn neurotoxicity, suggesting that an impaired neuronal antioxidant system renders the brain susceptible to Mninduced neurotoxicity (Chen and Liao, 2002; Dukhande et al., 2006; Maddirala et al., 2015). Moreover, inhibiting GSH synthesis potentiates the Mn-induced increase in inosine, hypoxanthine, xanthine, and uric acid levels in the striatum and brainstem of aged rats (Desole et al., 2000), indicating that Mn-induced cytotoxicity is mediated through mitochondrial dysfunction. Therefore, the specific vulnerability of dopamine neurons to Mn plays a pivotal role in impairing cellular antioxidant defenses, wherein breakdown of the mitochondrial oxidative energy metabolism cascade leads to dopaminergic cell death. Excess ROS fuels the oxidation of membrane polyunsaturated fatty acids (PUFA), yielding numerous arachidonic acid (ARA) peroxidation products, including reactive aldehydes such as 4 hydroxy-trans-2-nonenal (4-HNE), 4-oxo-trans-2-nonenal (4- ONE), MDA, acrolein, F2-isoprostanes (F2-IsoPs), and isofurans (Esterbauer et al., 1991; Aluru et al., 2015). The lipid ARA had been released from neural membrane glycerophospholipids through the activation of cytosolic phospholipases A<sup>2</sup> (cPLA2), which are enzymes coupled to NMDA receptors (Farooqui and Horrocks, 2007; Farooqui and Farooqui, 2011). Since most biological membranes of cells and organelles are composed of PUFA, lipid peroxidation is the main molecular mechanism involved in the oxidative damage to cell structures and in toxicitymediated cell death. Consistent with these observations, primary rat cortical neurons exposed to a very high Mn dose (500 µM) for 6 h show structural damage to neurons and a roughly 50% increase in F2-IsoPs levels compared to controls (Milatovic and Aschner, 2009). Likewise, in primary astrocyte cultures exposed to the same experimental conditions, F2-IsoPs levels increased 51% compared to control cultures (Milatovic et al., 2007). However, the direct role of Mn in CNS toxicity associated with lipid peroxidation remains debatable as some investigators argue that in vivo administration of Mn alters cellular Ferrous (Fe2+), which plays a permissive role in increasing lipid peroxidation and augmenting neuronal vulnerability (Shukla and Chandra, 1981; Chen et al., 2000, 2006).
Moreover, dopamine-derived quinones are known to bind and modify several PD-related proteins such as αSyn, DJ-1, and parkin (Conway et al., 2001; LaVoie et al., 2005; Girotto et al., 2012). However, of all the cellular macromolecules prone to oxidative damage, damaged nucleic acids are particularly hazardous due to the elevated risk of potentially irreparable
genetic base mutations. Among the five nucleobases, guanine is the most susceptible to hydroxyl radical-mediated oxidation (Cooke et al., 2003; Cerchiaro et al., 2009), which produces the well-studied oxidized DNA product 8-hydroxyguanosine (8-OHG). Interestingly, elevated 8-OHG as well as reduced 8-hydroxyl-2-deoxyguanosine (8-OHdG) have been observed in the SN and cerebrospinal fluid (CSF) of PD patients (Zhang et al., 1999; Isobe et al., 2010). In contrast, in vitro
the release of cytochrome c, activating the apoptosis initiator caspase-9, which in turn cleaves caspase-3. The cleaved fragment of caspase-3 interacts with protein
kinase C delta (PKCδ), a pro-apoptotic protein. Caspase-3-mediated proteolytic cleavage of PKCδ leads to DNA fragmentation and apoptosis.
studies of Mn toxicity reported increased 8-oxo-7,8-dihydro-2<sup>0</sup> deoxyguanosine (8-oxodG) content in the DNA of dopaminetreated PC12 cells (Oikawa et al., 2006). Stephenson et al. (2013) have also shown that Mn catalyzes the auto-oxidation of catecholamines in SH-SY5Y cells with the ensuing oxidative damage to thymine and guanine DNA bases, further indicating the damaging effect of Mn-induced semi-quinone radical ions and ROS production on DNA.
Mn preferentially accumulates in mitochondria, through the mitochondrial Ca2<sup>+</sup> uniporter, where it is mainly bound to mitochondrial membrane or matrix proteins (Gavin et al., 1999). Succinate, malate, and glutamate are important substrates for mitochondrial respiration, but at high concentrations, Mn2<sup>+</sup> binds to these substrates effectively inhibiting mitochondrial respiration (Gavin et al., 1999). Interference in oxidative phosphorylation triggers the downstream release of inflammatory signals, leading ultimately to apoptosis. Recent evidence sheds light on Mn-induced ER stress and ER-mediated cellular apoptosis. Rats given three different doses of Mn for 4 weeks showed a dose-dependent increase in apoptotic cells in the striatum, as evidenced by chromatin condensation, as well as up-regulation of markers of mitochondrial and ER stress-mediated apoptosis (Wang et al., 2015). Furthermore, Mn induces the transcriptional and translational up-regulation of αSyn (Li et al., 2010), promoting susceptibility to Mn-induced neurotoxicity through ERK1/2 MAPK activation, NF-κB nuclear translocation, and activation of apoptotic signaling cascades leading to dopaminergic cell death (Li et al., 2010; Prabhakaran et al., 2011).
Mn affects not only cellular viability, but also various factors involved in neurotransmitter regulation. Acetylcholine esterase (AChE) is an enzyme that hydrolyses acetylcholine (ACh), thus regulating its availability in the synaptic cleft (Whittaker, 1990; Pohanka, 2012). Chronic exposure to high levels of Mn can inhibit the activity of AChE, thereby allowing ACh to accumulate in the synaptic cleft and subsequently overstimulating muscarinic and nicotinic ACh receptors. While the precise mechanism has not been determined, inhibiting AChE increases levels of ROS and RNS (Milatovic et al., 2006; Santos et al., 2012), which further leads to lipid peroxidation as well as production of citrulline, a marker of RNS activity. Ali et al. (1983) reported that Mn overexposure in rats on a low-protein diet reduces the level of GABA in the brain while increasing the animals' susceptibility to seizures. However, the effect depended on the treatment regime and age of rats. For instance, low-dose Mn given thrice weekly for 5 weeks increased GABA levels (Takagi et al., 1990). Additional mechanistic studies are needed to better understand Mn's role in GABA dysregulation. In the case of glutamate, high levels of Mn in the brain may trigger constitutive NMDA activation leading to excitotoxic-related neuronal death. Once released into the synaptic cleft, most glutamate is removed by astrocytes via the glutamate-aspartate transporter (GLAST). However, high levels of extracellular Mn2<sup>+</sup> decrease the expression of GLAST and induce astrocyte apoptosis (Erikson et al., 2002). Chronic exposure to Mn can also increase the amplitude of excitatory postsynaptic potentials (EPSPs) in striatal neurons. With respect to the neurotransmitter dopamine, Ingersoll et al. (1999) demonstrated Mn transport to dopaminergic neurons via DAT. Another study done on DAT−/<sup>−</sup> mice receiving high doses of Mn reported a lower amount of striatal Mn compared to WT mice given the same dose. Interestingly, only the normally DAT-rich region of the striatum showed this contrasting pattern, which was not seen in areas not expressing DAT (Erikson et al., 2005). Young non-human primates exposed to a low dose of Mn twice weekly for about 9 weeks show retracted microglial processes even while dopaminergic neurons remained unchanged (Verina et al., 2011). More information is needed on the effect of this microglial disturbance on nigrostriatal neurons.
To conclude, Mn influx and efflux are tightly controlled in the body by various receptors and ion channels. However, overexposure to Mn can lead to the toxic accumulation of Mn in the brain, especially in the basal ganglia, causing hyperactivity of cortico-striatal neurons. While contradictory evidence arises from different dose regimens, in general Mn also impairs the regulation of neurotransmitters, such as dopamine, glutamate, and GABA by inhibiting the enzyme activity that regulates optimum neurotransmitter levels. High levels of glutamate and/or acetylcholine in the synaptic cleft overstimulate NMDA receptors leading to excitotoxic neuronal death. Mn may get transported into dopaminergic neurons via DAT. Excess cellular Mn2<sup>+</sup> disrupts Ca2<sup>+</sup> homeostasis in cells, leading to decreased dopamine production and neuronal death. Mn also causes ER and mitochondrial stress leading to neuronal apoptosis and/or gliosis. In light of the mounting evidence pointing to the deleterious effects of Mn on neurons and glia, researchers are examining the use of metal chelators and antioxidants as therapeutic interventions against manganism.
#### MANGANESE IN OTHER DISEASES
Until the last decade, Mn neurotoxicity was mainly associated with Parkinsonism, and very little attention had been given to its potential role in other neurodegenerative diseases. However,
with growing interest in the neurobiology of heavy metals, Mn has now been linked to other major neurodegenerative diseases such as HD and prion diseases (Choi et al., 2010; Martin et al., 2011; Kumar et al., 2015). Furthermore, gene expression in the frontal cortex of cynomolgus macaques exposed to various Mn doses indicates that the amyloidβ (Aβ) precursor-like protein 1 (APLP1) of the amyloid precursor family was highly up-regulated, thereby linking Mn exposure to AD (Guilarte et al., 2008). Along with this gene array analysis, immunochemistry revealed the presence of Aβ plaques and αSyn aggregates, which have been linked to PD as well as AD, and which have also been seen in the gray and white matter of Mn-exposed animals (Guilarte, 2010).
In contrast to Mn-induced Parkinsonism, the pathogenesis of HD, an autosomal dominant disorder characterized by the neurodegeneration of medium spiny neurons in the striatum, appears to involve a Mn transport deficiency (Kumar et al., 2015). Recent experiments carried out with immortalized mutant HD cell lines (SThdhQ7/Q7 and SThdhQ111/Q111) show reduced TfR levels and substantial deficits in Mn uptake (Williams et al., 2010). In follow-up studies, YAC128 HD transgenic mice accumulated less Mn in their basal ganglia, including their striata, which are focal regions for both HD neuropathology and Mn accumulation (Madison et al., 2012). Furthermore, transition metal analysis of HD patients has shown significantly increased iron together with significantly decreased cortical copper and Mn (Rosas et al., 2012), further supporting the role of Mn in HD.
Prion protein (PrP) is widely known for its association with transmissible spongiform encephalopathies (TSE), a class of neurodegenerative diseases caused by the accumulation of an abnormal isoform of the prion protein known as PrPSc (Jin et al., 2015a). The cellular prion protein PrP<sup>C</sup> has a high-binding affinity for divalent metals. In fact, above-normal Mn content has been detected in the blood and brains of humans infected with Creutzfeldt–Jakob disease (CJD), in scrapie-infected mice, and in bovines infected with bovine spongiform encephalopathy (BSE) (Wong et al., 2001; Hesketh et al., 2007; Hesketh et al., 2008). The binding of Mn to prion protein mitigates Mn's neurotoxicity during the early acute phase of Mn exposure (Choi et al., 2007). However, prolonged Mn exposure alters the stability of prion proteins without changing gene transcription (Choi et al., 2010), suggesting that Mn contributes to prion protein misfolding and prion disease pathogenesis. Interestingly, prion proteins survive significantly longer in a Mn-enriched soil matrix (Davies and Brown, 2009), a finding with important implications for the environmental transmissibility of PrPSc. The role of Mn in TSE was further validated by our lab's discovery that it enhances the ability of the pathogenic PrPSc isoform to regulate Mn homeostasis (Martin et al., 2011) and by Davies and Brown (2009) reporting that Mn increases the infectivity of scrapie-infected cells. Therefore, deepening our understanding of how metals interact with disease-specific proteins will provide further insight into the pathogenesis and potential treatment of neurodegenerative diseases.
## FUTURE DIRECTIONS
Our review of existing literature related to Mn overexposure and associated health issues has revealed genetic, sex- and age-related susceptibilities, signaling cascades involved in Mn neurotoxicity, and comparisons between PD and other motor disorders. Yet, many aspects of Mn overexposure and homeostasis remain largely understudied. For example, early childhood exposure to drinking water containing elevated Mn levels has been conclusively shown to compromise certain aspects of memory and learning; however, how absorption of excessive levels of Mn via the GIT leads to cognitive deficits (a CNS component) is still largely unknown. Secondly, given that the nasal tract and GIT are two well-known microbial environments, does overexposure to Mn via inhalation or ingestion alter the community composition of nasal or gut microbes or otherwise cause dysbiosis? Are changes in microbial populations offset by other lines of host defense against Mn toxicity or do these changes exacerbate the neuropathology? Thirdly, MRI and positron emission tomography (PET) on Mnexposed individuals have shown changes in brain Mn accumulation and dopamine and GABA neurotransmitter levels. Yet despite the strong links between manganism and Parkinsonism, no case control study has examined the extent of elevated Mn accumulation in the brain and its associated neuropathology, including the presence of Lewy bodies/neurites and elevated phospho-α-synuclein expression. Additionally, longitudinal studies examining the immediate, intermediate and long-term effects of elevated Mn exposure on children and adults are needed to identify age- and sexspecific susceptibilities, and potential biological, psychological and cognitive biomarkers. Lastly, studies systematically identifying Mn exposure limits based on age, sex, and exposure duration as well as changes in signaling cascades associated with metal homeostasis and protein aggregation need to be conducted.
### CONCLUSION
Chronic exposure to excessive Mn induces various neurological and psychiatric symptoms. While the body can efficiently remove excess Mn, primarily through the gut and liver, the brain cannot. Because of direct passage via the nasal neuroepithelium, inhaling large doses of Mn can lead to its accumulation in the brain's basal ganglia. Astrocytes are particularly sensitive to Mn toxicity and may compound neuroinflammation by releasing proinflammatory cytokines in response to excess Mn. Mn can also bind to substrates of oxidative phosphorylation, thus inhibiting mitochondrial respiration and thereby augmenting oxidative stress. Chronic exposure to Mn causes benign α-synuclein monomers, present in all neurons, to undergo a conformational change to the oligomeric structures that are toxic to neurons. Additionally, Mn dysregulates key protein degradative and trafficking pathways including proteasomes, autophagy, and endosomal trafficking. Taken together, we entertain the notion
that a 'neurotoxic triad,' comprising mitochondrial dysfunction and oxidative stress, protein misfolding and trafficking, and neuroinflammation, plays a major pathogenic role in Mn neurotoxicity (**Figure 5**). Beyond its effects on the CNS, excess Mn also interferes with the body's iron metabolism and can cause kidney failure. Early detection and chelation therapy can effectively reverse the harmful effects caused by this metal; however, if it progresses untreated, it can cause severe neurological and physiological defects. As with any metal, the bioaccumulation and teratogenic effects of Mn remain a risk, yet this aspect has not been studied in detail. Similarly, an in-depth study of Mn's role in protein misfolding and the upregulation of genetic markers for various neurological diseases in humans must be conducted. By combining the results of epidemiological surveys with human case studies as well as mechanistic studies done in in vitro and in animal models of Mn toxicity, we will eventually decipher the causes and symptoms of neurodegeneration caused by Mn toxicity well enough to
#### REFERENCES
develop effective therapeutic strategies that can be readily used against environmentally linked PD and related chronic neurodegenerative diseases.
#### AUTHOR CONTRIBUTIONS
DSH and SG conceived and wrote the article. GZ, HJ, AK, VA, and AGK provided intellectual input for review content and edited the manuscript. All authors read and approved the manuscript.
#### FUNDING
This work was supported by National Institutes of Health R01 grants ES026892, ES019267, and ES025991 to AGK and NS088206 to AK. The W. Eugene and Linda Lloyd Endowed Chair and Armbrust Endowment to AGK are also acknowledged.
implications for oxidative stress-induced neuronal cell death. J. Biol. Chem. 286, 19840–19859. doi: 10.1074/jbc.M110.203687
trafficking and efflux activity. J. Neurosci. 34, 14079–14095. doi: 10.1523/ JNEUROSCI.2329-14.2014
**Conflict of Interest Statement:** AGK and VA have an equity interest in PK Biosciences Corporation located in Ames, IA, United States. The terms of this arrangement have been reviewed and approved by ISU in accordance with its conflict of interest policies.
The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Copyright © 2019 Harischandra, Ghaisas, Zenitsky, Jin, Kanthasamy, Anantharam and Kanthasamy. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
# PERK-Mediated Unfolded Protein Response Activation and Oxidative Stress in PARK20 Fibroblasts
Giuseppina Amodio<sup>1</sup>† , Ornella Moltedo<sup>2</sup>† , Dominga Fasano<sup>3</sup> , Lucrezia Zerillo<sup>3</sup> , Marco Oliveti<sup>1</sup> , Paola Di Pietro<sup>1</sup> , Raffaella Faraonio<sup>3</sup> , Paolo Barone<sup>4</sup> , Maria Teresa Pellecchia<sup>4</sup> , Anna De Rosa<sup>5</sup> , Giuseppe De Michele<sup>5</sup> , Elena Polishchuk<sup>6</sup> , Roman Polishchuk<sup>6</sup> , Vincenzo Bonifati<sup>7</sup> , Lucio Nitsch<sup>3</sup> , Giovanna Maria Pierantoni<sup>3</sup> , Maurizio Renna<sup>3</sup> , Chiara Criscuolo<sup>5</sup>‡ , Simona Paladino<sup>3</sup> \* and Paolo Remondelli<sup>1</sup> \* ‡
#### Edited by:
Victor Tapias, Weill Cornell Medicine, United States
#### Reviewed by:
Wensheng Lin, University of Minnesota Twin Cities, United States Licio A. Velloso, State University of Campinas, Brazil
\*Correspondence:
Simona Paladino [email protected] Paolo Remondelli [email protected] †These authors have contributed equally to this work
‡Co-last authors
#### Specialty section:
This article was submitted to Neurodegeneration, a section of the journal Frontiers in Neuroscience
Received: 21 March 2019 Accepted: 12 June 2019 Published: 27 June 2019
#### Citation:
Amodio G, Moltedo O, Fasano D, Zerillo L, Oliveti M, Di Pietro P, Faraonio R, Barone P, Pellecchia MT, De Rosa A, De Michele G, Polishchuk E, Polishchuk R, Bonifati V, Nitsch L, Pierantoni GM, Renna M, Criscuolo C, Paladino S and Remondelli P (2019) PERK-Mediated Unfolded Protein Response Activation and Oxidative Stress in PARK20 Fibroblasts. Front. Neurosci. 13:673. doi: 10.3389/fnins.2019.00673 <sup>1</sup> Department of Medicine, Surgery and Dentistry "Scuola Medica Salernitana", University of Salerno, Salerno, Italy, <sup>2</sup> Department of Pharmacy, University of Salerno, Salerno, Italy, <sup>3</sup> Department of Molecular Medicine and Medical Biotechnology, University of Naples Federico II, Naples, Italy, <sup>4</sup> Section of Neuroscience, Department of Medicine, Surgery and Dentistry, University of Salerno, Salerno, Italy, <sup>5</sup> Department of Neuroscience, Reproductive, and Odontostomatological Sciences, University of Naples Federico II, Naples, Italy, <sup>6</sup> Telethon Institute of Genetics and Medicine, Pozzuoli, Italy, <sup>7</sup> Department of Clinical Genetics, Erasmus MC, Rotterdam, Netherlands
PARK20, an early onset autosomal recessive parkinsonism is due to mutations in the phosphatidylinositol-phosphatase Synaptojanin 1 (Synj1). We have recently shown that the early endosomal compartments are profoundly altered in PARK20 fibroblasts as well as the endosomal trafficking. Here, we report that PARK20 fibroblasts also display a drastic alteration of the architecture and function of the early secretory compartments. Our results show that the exit machinery from the Endoplasmic Reticulum (ER) and the ER-to-Golgi trafficking are markedly compromised in patient cells. As a consequence, PARK20 fibroblasts accumulate large amounts of cargo proteins within the ER, leading to the induction of ER stress. Interestingly, this stressful state is coupled to the activation of the PERK/eIF2α/ATF4/CHOP pathway of the Unfolded Protein Response (UPR). In addition, PARK20 fibroblasts reveal upregulation of oxidative stress markers and total ROS production with concomitant alteration of the morphology of the mitochondrial network. Interestingly, treatment of PARK20 cells with GSK2606414 (GSK), a specific inhibitor of PERK activity, restores the level of ROS, signaling a direct correlation between ER stress and the induction of oxidative stress in the PARK20 cells. All together, these findings suggest that dysfunction of early secretory pathway might contribute to the pathogenesis of the disease.
Keywords: PARK20, PERK (PKR-like endoplasmic reticulum kinase), oxydative stress, ER stress, Synaptojanin 1, membrane trafficking, mitochondrial dysfunction, Parkinson's disease
#### INTRODUCTION
Parkinson's disease (PD) is the second most common neurodegenerative disorder, characterized by the progressive loss of dopaminergic neurons in the substantia nigra pars compacta (Gao and Hong, 2011; Cannon and Greenamyre, 2013; Beitz, 2014; Feng et al., 2015). A combination of environmental and genetic factors has been considered to concur to the neuronal death. However, the exact molecular mechanisms are still unknown. Notwithstanding, the alteration of
**62**
mitochondrial function (Winklhofer and Haass, 2010; Pilsl and Winklhofer, 2012), of reactive oxygen species (ROS) homeostasis (Gaki and Papavassiliou, 2014; Al Shahrani et al., 2017; Guo et al., 2018; Paladino et al., 2018) as well as the dysregulation of protein folding control and/or autophagic flux (McNaught and Olanow, 2006; Malkus et al., 2009; Karabiyik et al., 2017; Remondelli and Renna, 2017) have been implicated in PD pathogenesis.
Among genetic PD, PARK20 is a rare autosomal recessive juvenile Parkinson's form due to mutations in Synaptojanin1 (Synj1), a phosphatidylinositol phosphatase (PtdInsPP) (Krebs et al., 2013; Quadri et al., 2013; Olgiati et al., 2014). The homozygous R258Q mutation was almost simultaneously reported in three unrelated families from Iran and Italy (Krebs et al., 2013; Quadri et al., 2013; Olgiati et al., 2014). Subsequently, the p.R459P mutation was found in an Indian family (Kirola et al., 2016); and, more recently, another Iranian kindred has been described with the p.R839C mutation (Taghavi et al., 2018). Finally, a frameshift mutation (p.S552Ffs<sup>∗</sup> 5) in heterozygous state with the benign p.T1236M missense variant has been identified in one late onset PD patient from Moroccan consanguineous parents (Bouhouche et al., 2017), correlating Synj1 lesions to the risk of PD development.
Synj1 is a highly conserved PtdInsPP existing in two isoforms: the 145-kDa neuronal isoform and the ubiquitous 170-kDa isoform (Ramjaun and McPherson, 1996). Synj1 function relies on two sequential PtdInsPP domains: the N-terminal Sac1 and the central 5-phosphatase domains (5<sup>0</sup> -PP) (Di Paolo and De Camilli, 2006). The Sac1 domain of Synj1 acts on PtdIns 3- and 4-monophosphate, which are enriched in the endosomal and Golgi membranes respectively (Guo et al., 1999). Instead, the 5<sup>0</sup> -PP domain of Synj1 dephosphorylates phosphatidylinositol 4,5-bisphosphate [PtdIns(4,5)P2] located in the plasma membranes (McPherson et al., 1996; Cremona et al., 1999). Additionally, the Synj1 protein also contains a COOHterminal proline rich region that retains the ability to interact with SH3 domains of a variety of proteins that regulate its subcellular localization and function (McPherson et al., 1996; Dittman and Ryan, 2009).
Thanks to its double enzymatic activity, Synj1 exerts multiple roles in dependence on the cell context. In nerve terminals, Synj1 participates to the control of synaptic vesicles retrieval (McPherson et al., 1996; Song and Zinsmaier, 2003; Mani et al., 2007) and cooperates with DNAJC6, another PD-causative gene (PARK19), in the process of clathrin disassembly from synaptic vesicles during endocytosis (Chang-Ileto et al., 2011; Edvardson et al., 2012). Proper Synj1 activity is essential to control homeostasis and function of early endocytic pathways in different cell types, including neuronal cells (Fasano et al., 2018). Consistently, early endosomes of PARK20 fibroblasts resulted enlarged and the recycling trafficking impaired (Fasano et al., 2018). On the other hand, unbalanced Synj1 expression is significantly involved in a number of neurological and psychiatric disorders, such as: Bipolar Disorder (Saito et al., 2001; Stopkova et al., 2004), Down's Syndrome and Alzheimer's Disease (Berman et al., 2008; Voronov et al., 2008; Chang and Min, 2009; Cossec et al., 2012; Martin et al., 2014), unraveling a critical role in neurons.
The p.R258Q mutation into the Synj1 Sac1 domain was shown to abolish either the 3- or 4- phosphatase activity, while it does not affect the 5-phosphatase activity (Krebs et al., 2013). Therefore, the loss of Sac1 function could alter the rate of PtdIns3P and PtdIns4P, two crucial PtdInsPs for the control of structure and function of endosomal (Efe et al., 2005; Di Paolo and De Camilli, 2006) and ER or Golgi complex membranes (De Matteis and D'Angelo, 2007; D'Angelo et al., 2008), respectively. Moreover, as we have recently shown, Sac1 domain is necessary for proper endosomal trafficking and at least 50% of its activity is required to ensure correct functionality (Fasano et al., 2018).
Here, we investigated whether p.R258Q mutation in the Sac1 domain of Synj1 could also influence vesicular trafficking at the early stages of the secretory pathway. Our experiments show that the ER exit machinery and the ER-to-Golgi trafficking are markedly compromised in p.R258Q mutated cells. As a consequence, PARK20 fibroblasts accumulate larger amounts of cargo proteins within the ER. This condition, referred to as ER stress, activates the PERK/eIF2α/ATF4/CHOP pathway of the Unfolded Protein Response (UPR) and induces oxidative stress and mitochondrial damage.
## MATERIALS AND METHODS
#### Cell Cultures
Fibroblasts were derived directly from the skin punch biopsies of the two Italian patients carrying the p.R258Q mutation at homozygous state (Quadri et al., 2013; Olgiati et al., 2014). A written informed consent was obtained from each patient. As control cells, primary adult Human Dermal Fibroblasts (HDF) were purchased from Sigma-Aldrich. PARK20 fibroblasts and HDF were grown in one ready-to-use Fibroblast Growth Medium (FGM from Sigma-Aldrich) at 37◦C and 5% CO<sup>2</sup> in humidified atmosphere. Experiments were performed on both cell lines at similar culture passages (P5-P6). When indicated, cells were starved in Fibroblasts Basal Medium (FBM from Sigma-Aldrich), which does not contain FBS and growth factors supplement. Drug treatments were performed with 1 µM GSK2606414 (Calbiochem) or 500 nM Thapsigargin (Sigma-Aldrich) for the indicated time.
#### Immunofluorescence
Cells seeded on glass cover slips were washed in phosphatebuffered saline (PBS), fixed in PBS-4 % paraformaldehyde and permeabilized 30 min in PBS containing 0.5% BSA, 0.005% saponin and 50 mM NH4Cl. Cells were immunostained with the following primary antibodies: rabbit polyclonal anti-ERGIC-53 (α-CT) (Spatuzza et al., 2004), mouse monoclonal anti-GM130 (BD Biosciences), rabbit polyclonal anti-Giantin (Abcam), rabbit plyclonal anti-KIAA0310 (Bethyl Laboratories), rabbit polyclonal anti-Sec31a, rabbit plyclonal anti-Sar1 (Millipore), rabbit polyclonal anti-collagen IV (Rockland immunochemicals), mouse monoclonal anti-KDEL (StressGen). Primary antibodies were detected with Alexa 488- and Cy3-conjugated antibodies (Jackson Immuno Research Laboratories).
For mitochondria staining, cells were incubated for 30 min at 37◦C with 200 nM Mitotracker Red CMXRos (Invitrogen-Molecular Probes) in FBM before fixing in cold acetone for 5 min on ice. Images were acquired on a laser scanning confocal microscope (TCS SP5; Leica MicroSystems or LSM 510 Meta; Zeiss MicroSystems) equipped with a plan Apo 63X, NA 1.4 oil immersion objective lens. Quantitative analysis was performed on a minimum of 30 cells by setting the same threshold of fluorescence intensity in all the samples analyzed.
Co-localization analyses and the mean intensity fluorescence quantification were carried out by using either the Leica SP5 or Zeiss software or the ImageJ program as previously described (Paladino et al., 2008; Gorrasi et al., 2014; Iorio et al., 2018; Ranieri et al., 2018). Briefly, the number of co-localized pixels was normalized for the total fluorescent pixels in the image. The degree of colocalization was assessed by calculating the Pearson's correlation coefficient. Mean fluorescence intensity was measured in Region of Interest (ROI) of equal area in control and PARK20 samples. The number and size of SEC31a and SEC16a fluorescent spot was measured by using the ImageJ program. The distance from the nucleus of ERGIC-53 fluorescent spots was measured by using the scale bar drawing tool of Leica SP5 software.
### Electron Microscopy
Cells were fixed in 1% glutaraldehyde dissolved in 0.2 M HEPES buffer (pH 7.4) for 30 min at room temperature and then postfixed with a mixture of 2% OsO4 and 100 mM phosphate buffer (pH 6.8) (1 part 2% OsO4 plus 1 part 100 mM phosphate buffer) for 25–30 min on ice. Then, the cells were washed three times with water and incubated with 1% thiocarbohydrizide diluted in H2O for 5 min, incubated in a mixture of 2% OsO 4 and 3% potassium ferrocyanide (1 part 2% OsO4 plus 1 part 3% potassium ferrocyanide) for 25 min on ice and overnight at 4 ◦C in 0.5% uranyl acetate diluted in H2O. After dehydration in graded series of ethanol, the cells were embedded in epoxy resin and polymerized at 60◦C for 72 hr. Thin 60 nm sections were cut at the Leica EM UC7 microtome. EM images were acquired from thin sections using a FEI Tecnai-12 electron microscope equipped with a VELETTA CCD digital camera (FEI, Eindhoven, Netherlands).
### Western Blotting
Actively growing cells seeded on 60 mm dishes were starved in FBM for 18 h prior to be subjected to the indicated treatments. Cells were then harvested in lysis buffer (10 mM Tris-HCl pH7.4, 150 mM NaCl, 1 mM EDTA pH 8.0, 1% Triton X-100) supplemented with protease and phosphatase inhibitor cocktail (Roche). Equal amounts of protein extracts were analyzed by 8 or 10% SDS-PAGE and transferred on Protran nitro-cellulose membranes (Schleicher and Schuell). Membranes were blocked either in PBS containing 10% non-fat dry milk and 0.1% Tween-20, or in TBS containing 5% BSA and 0.1% Tween-20, depending on the antibody used.
Membranes were cut in stripes according to the molecular weight expected for the single proteins analyzed, incubated with the primary followed by secondary antibodies and then visualized by ECL reaction (Amersham International) (see **Supplementary Figures S3**, **S4**). The following primary antibodies were used: rabbit monoclonal anti-PERK (Cell Signalling Technology), rabbit polyclonal anti-eIF2α and anti-phospho-eIF2α (Cell Signalling Technology), rabbit monoclonal anti-ATF4 (Abcam), mouse monoclonal anti-GADD153 (Santa Cruz Biotechnology), mouse monoclonal anti-HO1 (Santa Cruz Biotechnology) and mouse monoclonal anti-α Tubulin (Santa Cruz Biotechnology). HRP-conjugated IgG (Jackson Immuno Research Laboratories) were used as secondary antibodies. Filters were exposed to ChemiDoc MP System (Bio-Rad Laboratories Inc.) and the densitometry analysis of autoradiographs was performed by the ImageJ program on three independent experiments.
### Oxidative Stress Assays
10<sup>6</sup> cells for each treatment were disposed in a well of BD Falcon 96-well black plates and starved in FBM for 18 h prior to be subjected to the indicated treatments. Cytosolic ROS were quantified by a fluorescence microplate reader [Tecan Infinite 200 Pro] using dihydrorhodamine 123 (DHR 123) probe (Santa Cruz Biotechnology), a cell-permeable non-fluorescent substance that undergoes intracellular oxidation in the presence of ROS. In detail, cells were incubated for 1 h with 50 µM of DHR123/HBSS and then washed two times with freshly prepared Hank's balanced salt solution. Subsequently, formation of DHR 123 has been monitored by fluorescence spectroscopy using excitation and emission respectively of 500 L nm and 536 L nm. In some experiments, cells were pre-incubated with 1 µM GSK2606414 for 2 h, before measurements. Fluorescence signals have been recorded using Tecan i-control software and expressed as arbitrary units.
In another approach, we measured the ROS on singlecells. To this purpose, cells grown on glass bottom dishes were incubated with 2<sup>0</sup> ,70 -dichlorodihydrofluorescein diacetate (DCFH-DA, 10 µM) for 10 min at 37◦C in culture medium without serum and, then, imaged in vivo in CO<sup>2</sup> independent medium as previously described (Piccoli et al., 2013). Images were collected by a Zeiss confocal LSM510 using Ar–Kr laser beam (λex 488 nm); same laser power and same settings were used for control and patient fibroblasts in all experimental conditions. Data are expressed as arbitrary units of fluorescence and reported as mean ± SD from three independent experimental conditions.
For NADPH oxidase activity measurement, the lucigeninenhanced chemioluminescence assay was used to determine NADPH oxidase-mediated superoxide radical (O<sup>2</sup> <sup>−</sup>) production as previously described (Carrizzo et al., 2017; Schiattarella et al., 2018). Cells, cultured in 100 mm dishes, were detached using 0.25% trypsin/EDTA (1 mmol/l), washed with PBS, and resuspended in modified HEPES buffer containing (mmol/l) NaCl 140, KCl 5, MgCl2 0.8, CaCl2 1.8, Na2HPO4 1, HEPES 25 and 1% glucose, pH 7. Subsequently, cells were homogenated using VWR pellet mixer [#431-0100] and 100 µg of extract were distributed on a 96-well microplate. The reaction was started by the addition of NADPH (0.1 mmol/l) to each well (250 µl final volume) and lucigenin (5 µmol/l). The luminescence was measured using Tecan Infinite M200 multimode microplate fluorometer at 37◦C every 10 s for 60 min. Each experiment
was performed in triplicate. In some experiments, cells were preincubated with 1 µM GSK2606414 for 2 h, before measurement of luminescence.
### RT-PCR and XBPI Splicing Assay
One microgram of DNAse-treated total RNA was retrotranscribed with the Easy-script plus cDNA synthesis Kit (abm) according to manufacturer instructions. Semi-quantitative PCR was performed on 3 µl of cDNA with the following primers Bip/Grp78-forward: 5<sup>0</sup> -CTG GGT ACA TTT GAT CTG ACT GG-3<sup>0</sup> ; Bip/Grp78-reverse: 5<sup>0</sup> -GCA TCC TGG TGG CTT TCC AGCCAT TC-3<sup>0</sup> ; GAPDH-forward: 5<sup>0</sup> -GAA GGT GAA GGT CGGAGT C-3<sup>0</sup> ; GAPDH-reverse: 5<sup>0</sup> -GAA GATGGT GAT GGG ATTTC-3<sup>0</sup> (Amodio et al., 2011). XBPI splicing assay was performed as previously described (Eletto et al., 2016) by using the following primers: 5<sup>0</sup> -A AAC AGA GTA GCA GCT CAG ACT GC-3<sup>0</sup> and 5<sup>0</sup> -C CTT CTG GGT AGA CCT CTG GGA G-3<sup>0</sup> . The resulted un-spliced and spliced XBP1 mRNA were separated by gel electrophoresis on 3% agarose gel. Ethidium bromide-stained amplicons were quantified by densitometry with ImageJ software.
#### Statistical Analysis
Data are expresses as mean ± SD. All statistical analyses using Student's t-test and histograms were completed with Prism statistical software (Graphpad, La Jolla, CA, United States) and differences were considered statistically significant when P < 0.05.
## RESULTS
### PARK20 Fibroblasts Show Unbalanced ER-to-Golgi Trafficking and Abnormal Structure of Golgi Membranes
To test whether membrane trafficking from the ER to the Golgi complex was affected by the p.R258Q mutation, we analyzed fibroblasts derived from homozygous R258Q/R258Q PARK20 patients and from healthy individuals.
Dynamics of membrane trafficking at the early steps of the secretory pathway were analyzed by looking at the intracellular distribution of vesicles carrying the cargo receptor ERGIC-53 (Appenzeller et al., 1999). Normally, the ERGIC-53 protein cycles between the ER and the Golgi complex (Appenzeller et al., 1999) and ERGIC-53 containing vesicles show their typical punctuate distribution depicted by higher concentration in the region closed to the cis-Golgi membranes, which in turn are labeled with the resident protein GM130 in wild-type cells (**Figure 1A**, wt). Instead, in patient fibroblasts ERGIC-53 vesicles were reduced both in size and fluorescence intensity (**Figure 1A**, PARK20). In addition, they are delocalized throughout the cytoplasm at higher distance from the perinuclear region with a mean value of 12.3 ± 3.4 µm in the PARK20 cells vs. 31.8 ± 4.3 µm in the control cells (**Figure 1C**). Interestingly, ERGIC53 vesicles redistribution pattern in the PARK20 cells overlapped with the membrane network of the ER, as shown by the fluorescence detected by the anti-KDEL antibody, which label ER resident proteins bearing the KDEL retrieval sequence (**Figure 1B**). In addition, we also detected dramatic changes in the organization of Golgi membranes of PARK20 fibroblasts (**Figures 1A,D**). Both the cis-Golgi membranes labeled by the resident protein GM130 (**Figure 1A**) and the overall Golgi architecture revealed by the structural Golgi protein giantin (**Figure 1D**) were more dispersed and relocated in tubular structures extending from the nucleus to the cell edge in the PARK20 cells with respect to control (**Figures 1A,D**). The ultrastructural analysis further showed that the Golgi complex is scattered throughout the cell in PARK20 fibroblasts (**Figure 1E**, arrows). Moreover, while GM130 colocalized almost completely with giantin in control cells, they resulted partially co-distributed in patient cells (**Figure 1D**).
Since both GM130 and giantin are involved in the ER-to-Golgi trafficking (Alvarez et al., 2001), these results further suggest that PARK20 cells undergo unbalanced trafficking at the early steps of the secretory pathway.
### PARK20 Fibroblasts Show Reduced Formation of COPII Carrier Vesicles
The results described above prompted us to test whether the abnormal organization of post-ER compartments observed in the PARK20 cells was the result of reduced flow of carrier vesicles budding from the ER. Formation of transport vesicles from the ER membranes requires the assembly of the vesicular coat (COPII) at specific ER membrane domains defined ER exit Sites (ERESs), recognized by the presence of the endogenous Sec16 protein (isoform A). As a rule, Sec16 recruits COPII components for their assembly at the ERESs (Watson et al., 2006). As expected, these latter are visible as punctuate structures dispersed throughout the cytoplasm (**Figure 2A**, Sec16/wt). Instead, in PARK20 cells the number of puncta of Sec16 fluorescence are reduced (**Figures 2A,B**; Sec16/PARK20), indicating that the number of ERESs is considerably decreased in PARK20 cells. As a consequence, the number of COPII vesicles, revealed by antibody recognizing Sec31 (**Figure 2A**, Sec31/wt), a component of the outer layer of the COPII vesicles, was also reduced (**Figures 2A,B**; Sec31/PARK20). Moreover, Sec16 co-localized with Sec31 at the same extent as in control cells (**Figure 2A**, merge), suggesting that Sec16 still organizes COPII assembly at ERESs, but with less efficiency.
Thus, our results strongly indicate that the Synj1 activity localized in the Sac1 domain of the protein is essential for the proper function of COPII. In particular, the reduced number of ERES found in the PARK20 cells indicates that altered phosphatase activity of the Synj1 Sac1 domain reduces the amount of ER exit sites, thus biasing the assembling and/or the stability of COPII vesicles.
### The Secretion of Collagen IV Is Impaired in PARK20 Fibroblasts
To test whether reduced formation of ERESs could have an effect on the rate of cargo proteins transport from the ER to the Golgi complex, we examined the level of distribution of endogenous collagen IV (COLIV) along the secretory pathway in PARK20
cells vs. control cells (**Figure 3A**). Typically, in normal fibroblasts COLIV is secreted from the cell and accumulates in structures located in the interstitial space out of the cell (**Figure 3A**, COLIV/wt) (Amodio et al., 2016). In PARK20 fibroblasts, COLIV was mainly found within intracellular membranes resembling the ER network stained by the KDEL antibody (**Figure 3A**: compare PARK20 pro-COLIV to KDEL), indicating that COLIV secretion is reduced. In PARK20 fibroblasts, COLIV did not accumulate
FIGURE 2 | The formation of COPII-coated vesicles are reduced in PARK20 fibroblasts. (A) HDF (WT) and PARK20 cells seeded on glass coverslips were fixed, stained with the indicated antibodies and analyzed by confocal immunofluorescence. Scale bars: 10 µm. (B) Histogram shows particle count (mean ± SD) of Sec31 (green) and Sec16 (red) fluorescent spots analyzed by Image J. N ≥ 30. <sup>∗</sup>p ≤ 0.05, ∗∗p ≤ 0.01, Student's t-test.
outside the cell (**Figure 3A**, PARK20), but it was mainly found within intracellular membranes, in structures that resemble the ER network stained by the KDEL antibody (**Figure 3A**: compare PARK20 pro-COLIV to KDEL). Consistently, by analyzing the rate of colocalization with ER resident proteins labeled by the anti-KDEL antibody, we found an increase in procollagen IV (pro-COLIV) colocalization with ER membranes: 87.1 ± 7.7% in the Synj1 mutated cells compared to 56.6 ± 13.3% in control cells (**Figure 3B**). In line with the immunofluorescence assays, we found that secretion of COLIV by PARK20 fibroblasts into culture medium was strongly reduced in comparison to control fibroblasts (**Supplementary Figure S1A**). All these data indicate that secretion of COLIV in the PARK20 cells was almost completely inhibited presumably as a consequence of the reduced function of the ER exit machinery. Moreover, a slight but appreciable reduction of total protein secretion is observed in PARK20 fibroblasts as well as a reduction of total delivery to the surface (**Supplementary Figure S1**), all suggesting an impairment of secretory trafficking.
#### The PERK/eIF2a/ATF4 Branch of the Unfolded Protein Response Is Activated in PARK20 Fibroblasts
The presence of higher amounts of proteins retained into the ER, prompted our investigation into whether such accumulation
could activate the ER stress and, as a consequence, the UPR (Walter and Ron, 2011). Therefore, we analyzed the activation state of components of the UPR signaling, such as PERK and IRE1, and the expression level of typical marker of ER stress (Franceschelli et al., 2011; Hiramatsu et al., 2011). To determine PERK activation, we analyzed by western blotting the phosphorylated PERK form (p-PERK) and eIF2α form (p-eIF2a) expressed in the cell extracts obtained from control and PARK20 fibroblasts (**Figure 4**). In particular, PERK phosphorylation was recognized throughout immunoblots (**Figure 4A**) showing the band-shift of p-PERK in western blot analyses as a consequence of the higher molecular weight acquired by the auto-phosphorylation (Harding et al., 1999; Eletto et al., 2016). As expected, in control cells we did not detect p-PERK form in basal conditions, but only after treatment with the UPR inducer thapsigargin (TG) (**Figure 4A**). Strikingly, p-PERK was highly detectable in uninduced PARK20 cells, suggesting that the PERK branch of the UPR was constitutively turned on in PARK20 fibroblasts (**Figure 4A**). As in control fibroblasts, cell exposure to the PERK inhibitor GSK abolished kinase activity of the PERK protein (**Figure 4A**), confirming that this pathway is activated in Synj1 mutated cells.
Then, we analyzed the phosphorylation status of elF2α by using a p-eIF2α antibody, which specifically detects the phosphorylated form of the protein (**Figure 4A**) and we found higher levels of p-eIF2α in PARK20 fibroblasts with respect control cells (**Figures 4A,B**). Moreover, PERK inhibitor GSK was able to reduce p-eIF2α amount in the Synj1 mutated cells (**Figures 4A,B**).
Finally, since increased eIF2α phosphorylation induces the ATF4/CHOP pathway of the ER stress, we analyzed the expression of ATF4 and CHOP in the PARK20 cells and found that both proteins enhanced with respect to control cells (**Figures 4A,B**). Moreover, after PERK inhibitor incubation ATF4 significantly reduced. Conversely, CHOP was not influenced by GSK, suggesting the involvement of positive feedback loops activated downstream to PERK induction (**Figures 4A,B**).
We also tested whether in the Synj1 mutated cells, the IRE1 arm of the UPR and/or the expression of genes under the transcriptional control of the ATF6 pathway of the UPR were equally activated. We found that either the IRE1 endonuclease activity or expression levels of ATF6-controlled genes in the PARK20 cells were similar to those detected in control cells (**Supplementary Figure S2**).
In summary, our results reveal that in the PARK20 cells the PERK/eIF2α/ATF4 pathway of the UPR is constantly activated, presumably as a result of the persistent activation of ER stress induced by the overload of cargo protein within the ER.
### Persistent Activation of the PERK Pathway of the UPR Induces Oxidative Stress in PARK20 Cells
The alteration of ER proteostasis and the consequent accumulation of misfolded proteins within the ER is associated with the increment of ER protein folding that strongly induces ROS production (Tu and Weissman, 2004; Santos et al., 2009). Since NADPH oxidase is one of the key sources of cytosolic ROS (Lambeth, 2004), we measured the activity of the NADPH oxidase (NOX) through a quantitative lucigenin-based luminescence assay. Higher NOX activity was observed in PARK20 cells compared to control cells at each time point (**Figure 5A**), demonstrating that PARK20 fibroblasts exhibited pronounced oxidative stress.
It is well documented that the UPR could modulate the oxidative state of the cell, in particular through the PERK/eIF2α/ATF4 pathway (Amodio et al., 2018). Therefore, given the activation of the PERK pathway found in PARK20 fibroblasts, we measured the levels of cytosolic Reactive Oxygen Species (ROS) by dihydroethidium (DHE) fluorescent probe, in absence or in presence of the PERK inhibitor GSK (**Figure 5B**). Upon basal conditions, a significant higher amount of ROS production was detected in PARK20 cells with respect to control cells (24000 vs. 38000 a. u.; 1.6 Fold) (**Figure 5B**). Interestingly, GSK treatment reduced drastically ROS-derived DHE fluorescence, retrieving it to the values found in the control cells (**Figure 5B**). Alternatively, ROS levels were assessed by confocal microscopy imaging cells with the redox-sensitive fluorescent probe 2<sup>0</sup> ,70
dichlorodihydrofluorescein diacetate (DCFH-DA) obtaining similar results (**Figure 5C**).
All these data indicate that PERK inhibition significantly reduces cytosolic ROS generation, thus providing evidence that the activation of the PERK pathway of UPR is responsible for the induction of oxidative stress in PARK20 fibroblasts.
Because the PERK pathway of the UPR can also activate antioxidant factors, such as the neuroprotective haemeoxygenase-1 (HO-1) enzyme (Alam et al., 1999; Cullinan et al., 2003; Kensler et al., 2007), we analyzed the expression level of the HO-1 protein in control and PARK20 fibroblasts (**Figure 5D**). Our results clearly show a consistent up-regulation and PERK-dependant expression of the HO-1 protein in PARK20 cells, suggesting that PERK activation could also induce an antioxidant response to oxidative stress in PARK20 cells (**Figure 5D**).
Furthermore, because mitochondria represent the major source of intracellular ROS and their dysfunction is widely reported in PD (Balaban et al., 2005; Lin and Beal, 2006), we investigated whether the observed cytosolic ROS overproduction is associated with mitochondria impairment. Because mitochondrial function is linked to the overall organization of the mitochondrial network (Chan, 2012), we examined the mitochondrial morphology by using the Mitotracker Red CMXRos probe (**Figure 6**). The morphology of mitochondrial network is profoundly altered in PARK20 fibroblasts compared to control cells (**Figure 6**). As evident by 3D reconstructions in the patient cells, mitochondrial network lost the typical interconnected tubular structure and exhibited thinner and shorter tubules with many fewer branch points (**Figure 6B**). Moreover, we also observed a reduced Mitotracker staining in PARK20 compared to control cells (**Figure 6A**). Considering that Mitotracker probe accumulates in mitochondria depending on its membrane potential (MMP), the reduced staining suggests a loss in MPP.
All these data indicate that, in addition to the increase in cytosolic ROS, in the PARK20 cells mitochondrial alteration was dependent on PERK activation and quite reversed by the PERK inhibitor GSK2606414 (**Figure 6B**).
### DISCUSSION
In the present work we investigated further into the molecular events underlying the biogenesis of the juvenile parkinsonism PARK20, highlighting a role of Synj1 in regulating the early steps of secretory pathway.
Previous studies have well established that Synj1 is a crucial player for synaptic vesicle endocytosis and renewal at the nerve terminals, thanks to its 5-phospatase activity (Cremona et al., 1999; Kim et al., 2002; Mani et al., 2007; Cao et al., 2017). Recently, it has been demonstrated that Synj1 controls endosomal trafficking in different cell types including neuronal cells, presumably by regulating the of PtdIns3P levels within the endocytic pathways (Efe et al., 2005; Di Paolo and De Camilli, 2006; Krebs et al., 2013; Fasano et al., 2018).
Our results reveal a novel role for Synj1, consisting in the regulation of membrane trafficking at the ER-to-Golgi boundaries. It is likely that this regulatory function might be associated to its PtdIns4P phosphatase activity since ER and Golgi membranes are enriched in this phosphoinositide and the PtdIns4P phosphatase activity of Synj1 is affected by R258Q mutation (Krebs et al., 2013). Consistently, similar effect can be observed when persistent decrease of PtdIns4P hydrolysis is generated by deletion of the PtdIns4P phosphatase Sac1, the major lipid 4-phosphatase in yeast and mammals (Foti et al., 2001; Tahirovic et al., 2005; Liu et al., 2008).
Interestingly, by the use of a newly developed highresolution imaging technology Sac1, which has been shown to shuttle between ER and Golgi in response to different stimuli (Blagoveshchenskaya and Mayinger, 2009), has been recently found to reside at the ER-Trans Golgi Network (TGN) contacts
sites, where its phosphatase activity controls either ER or Golgi PtdIns4P membrane content (Venditti et al., 2019a,b). ER can establish membrane contacts with TGN as well as with endosomal membranes, referred to as ER-endo-lysosomal contact sites (EELCS) (Eden, 2016; Henne, 2017).
Given the interconnection of Synj1 with the endolysosomal pathway, an interesting hypothesis, to be tested in future work, would be to establish, whether Synj1 resides at EELCS and how EELCS helps Synj1/Sac1 activity to maintain PtdIns4P homeostasis.
Another finding supporting a role of Synj1 in controlling the early steps of secretory pathway is the altered distribution of GM130 and giantin, two Golgi factors also involved in the ER-to-Golgi trafficking (Alvarez et al., 2001). Furthermore, PARK20 cells display dramatic changes in the organization of Golgi membranes. The loss of intact Golgi populations with concomitant increase in vesiculated and dispersed Golgi membranes might be due to the unbalanced trafficking or defective membrane tethering and fusion events, consequent to the loss of Synj1 PtdIns4P phosphatase activity (Foti et al., 2001; Tahirovic et al., 2005; D'Agostino et al., 2014, 2018).
Central finding of our work is the discovery that formation of carrier vesicles at the endoplasmic reticulum exit sites (ERESs) is largely inhibited. The consequent traffic jam of secretory proteins within the ER membranes generates the activation of the PERK pathway of the ER stress/UPR, which in turn induces oxidative stress and mitochondrial damage. Therefore, the primary event causing the ER stress activation might rely on the altered control of the PtdIns4P content at ER membranes with consequent impairment of carrier vesicles formation. Indeed, the dynamic control of PtdIns4P level is necessary to coordinate the progression of both ERES formation and COPII assembly (Nagaya et al., 2002; Pathre et al., 2003; Blumental-Perry et al., 2006; Farhan et al., 2008). In these events, a specific role is played by p125A that, upon PtdIns4P recognition, promotes the recruitment of Sec16 at the ERESs, which in turn favors COPII assembly and cargo export from the ER (Shimoi et al., 2005; Iinuma et al., 2007; Ong et al., 2010).
Our results indicate that loss of Sac1/Synj1 activity in the R258Q/R258Q PARK20 cells could profoundly alter these dynamic events, leading to defective export of secretory proteins from the ER.
In PARK20 cells the PERK/eIF2α/ATF4 pathway of the UPR is hyperactive in response to the persistent state of ER stress induced by the ER overload of cargo proteins. This finding opens a novel perspective in the understanding of the molecular events leading to the PARK20 phenotype. The activation of the PERK pathway of the UPR is a common hallmark of various neurodegenerative diseases. In particular, as shown from postmortem analyses in PD patients as well as in animal models of PD, the activation of the PERK pathway represents a common cause of death of dopaminergic neuron (Scheper and Hoozemans, 2015; Gully et al., 2016). In our case, we show that, in PARK20 cells, prolonged PERK activation generates a pronounced production of cytosolic ROS, whereas GSKmediated inhibition of PERK drastically reduces ROS production.
In this regard, it is worth noting that UPR, through multiple different pathways, can give rise to either pro-oxidant or antioxidant response (Malhotra and Kaufman, 2007; Amodio et al., 2018). In particular, during ER stress the increment of ER protein folding demand strongly induces ROS production (Tu and Weissman, 2004; Santos et al., 2009). In this context, PERK can play the role of a double-edged sword. In fact, in first instance, the PERK-eIF2α-ATF4 axis operates to restore ER proteostasis by reducing ER protein load and by inducing antioxidant pathways through the activation of the transcription factor nuclear factor erythroid 2-related factor 2 (NRF2) (Alam et al., 1999; Cullinan et al., 2003; Kensler et al., 2007). On this line, we found a significant up-regulation of HO-1 expression in PARK20 cells. On the other hand, the persistent activation of PERK in condition of unsolved protein misfolding can boost up ROS production and induce factors of the pro-oxidant signaling pathway. Among these, the transcription factor CHOP, activated
downstream to the PERK/eIF2α/ATF4/CHOP pathway, plays an important role. CHOP promotes the expression of both Ero1 and NOX that are responsible for ROS production during the oxidative protein folding and ER stress (Li et al., 2010; Anelli et al., 2012). Accordingly, we found a significant increased expression of CHOP in PARK20 cells besides the reported increase of cytosolic ROS. In this regard, it is worth remarking that normal CHOP expression is not recovered after GSK treatment. This finding is not surprising, since after its PERK-dependent activation, CHOP activates other downstream pathways that positively feedback on CHOP expression. Additionally, it is important to consider that CHOP triggers the ERo1-IP3R1- Ca2+/calmodulin-dependent protein kinase II (CaMKII)-NOX2 cascade, in which NOX2 finally induces CHOP expression in a manner independent by PERK (Li et al., 2010; Anelli et al., 2012). All in all, our data support co-existence of a PERKdependent pro-oxidant and anti-oxidant response. However, we do not know whether one of them may prevail in the etiopathological onset of PARK20 and further investigation is needed in this direction.
Mitochondrial dysfunction is a common hallmark of both sporadic and genetic PD and is often associated with neuronal cell death in a number of neurodegenerative diseases (Subramaniam and Chesselet, 2013). Mitochondria are strictly connected to the ER via the mitochondrial-associated ER membranes (MAMs) through which Ca2+, lipids and ROS are transmitted from the ER to mitochondria (Rainbolt et al., 2014). Our findings support that the activation of PERK-CHOP pathway during chronic ER stress can even potentiate MAMs altering the mitochondrial function. Accordingly, in PARK20 cells we observed that mitochondrial alteration of MMP and morphology is dependent on PERK activation and reversed by the PERK inhibitor GSK2606414. Similar results were obtained in pink1/parkin PD models, where mitofusin contacts with damaged mitochondria sustain PERK signaling, while suppression of PERK signaling by using GSK2606414 or by genetic inhibition has a neuroprotective effect (Celardo et al., 2016), suggesting common molecular features between PARK20 and other PD types.
#### CONCLUSION
In the present work we show that PARK20 fibroblasts display alteration of the early secretory compartments and impairment of the ER-to Golgi trafficking leading to PERK activation, OS induction and mitochondrial dysfunction. Thus, these results indicate that, beside the role of endosomal system previously shown (Fasano et al., 2018), defects of early secretory pathway could contribute to the PARK20 pathogenesis. Together with our previous findings, our data emphasize the link between membrane trafficking defects and PD. Moreover, although the correlation between mitochondrial dysfunction, OS and PERK activation in PARK20 cells needs to be further investigated, our current findings open a new lead in studying PD and phosphoinositide metabolism, providing possible novel biomarkers that can be used as diagnostic and prognostic tools for the disease.
## DATA AVAILABILITY
The datasets generated for this study are available on request to the corresponding author.
### AUTHOR CONTRIBUTIONS
GA, OM, DF, and LZ performed and quantified the immunofluorescence data. MO, PDP, and RF carried out the biochemical quantitative analyses of oxidative stress. PB, MP, GDM, VB, CC, and ADR identified the patients, conducted the biopsies and provided primary cultures of fibroblasts. GA, OM, and MR analyzed UPR signaling. GA, OM, LN, and GP carried out the mitochondrial imaging analyses. EP performed electron microscopy experiments. EP and RP carried out the EM analyses. SP and PR contributed to the conception and design of the work, and wrote sections of the manuscript.
### FUNDING
This work was supported by the grant FARO 2012 from San Paolo bank and Polo delle Scienze e delle Tecnologie per la Vita, University of Naples Federico II to SP.
### SUPPLEMENTARY MATERIAL
The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fnins. 2019.00673/full#supplementary-material
FIGURE S1 | The exocytic trafficking is impaired in PARK20 fibroblasts. (A) Wild-type and PARK20 fibroblasts, grown on 60 mm dishes, were incubated in culture medium containing 0.75% serum for 7 h. Then, equal volumes of corresponding culture mediums were collected, TCA-precipitated, separated by 7.5% SDS-PAGE and revealed by western blotting with collagen IV (COL IV) antibody (right panel). 1/20 of total was analyzed by 10% SDS-PAGE and total protein secretion was revealed by Coomassie-Blue staining (left panel). (B) Wild-type and PARK20 fibroblasts, grown on 60 mm dishes, were biotinylated to label selectively surface proteins and cell lysates were revealed by western blotting using HRP-conjugated streptavidin.
FIGURE S2 | Activation of UPR-dependent signaling pathways. (A) Semi-quantitative RT-PCR of Bip/Grp78 mRNA in HDF (WT) and PARK20 fibroblasts either untreated (0) or treated with 2 µg/ml Tunicamycin (TM) or 500 nM Thapsigargin (TG) for the indicated times. GAPDH mRNA was used as reference. One out of three independent experiment is shown. (C) Histogram shows the relative fold expression of Bip/Grp78 mRNA amplified as in (A) and calculated by densitometry analysis with ImageJ software. Values are expressed as mean ± SD. Controls (C) refers to untreated samples. N = 3. (B) XbpI splicing assay performed on samples treated as in (A). GAPDH mRNA was used as reference. Amplicons derived from unspliced-XbpI (u) and spliced-XbpI (s) are shown. Numbers refers to the percent of spliced-XbpI to total XbpI (mean values), quantified by densitometry analysis with Image J from three independent experiments.
FIGURE S3 | The figure represents the uncut or partially cut filters used to mount Figure 4A. Predicted MW of proteins are reported on the left of the panels. # Indicates filters of the same experiment stripped and re-probed with antibodies as indicated. <sup>∗</sup> Indicates the phosphorylated form of PERK.
FIGURE S4 | The figure represents the uncut filter used to mount Figure 5D. Predicted MW of the HO-1 protein is reported on the left of the panel.
### REFERENCES
fnins-13-00673 June 27, 2019 Time: 15:17 # 12
**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Copyright © 2019 Amodio, Moltedo, Fasano, Zerillo, Oliveti, Di Pietro, Faraonio, Barone, Pellecchia, De Rosa, De Michele, Polishchuk, Polishchuk, Bonifati, Nitsch, Pierantoni, Renna, Criscuolo, Paladino and Remondelli. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
# GRP78 Level Is Altered in the Brain, but Not in Plasma or Cerebrospinal Fluid in Parkinson's Disease Patients
Jean-Ha Baek\*, Dejan Mamula, Beata Tingstam, Marcela Pereira, Yachao He and Per Svenningsson\*
Department of Clinical Neuroscience, Karolinska Institute, Stockholm, Sweden
Accumulation of misfolded proteins results in cellular stress, and is detected by specific sensors in the endoplasmic reticulum, collectively known as the unfolded protein response (UPR). It has been prominently proposed that the UPR is involved in the pathophysiology of Parkinson's disease (PD). In the present study, the levels of the UPR proteins and mRNA transcripts were quantified in post mortem brain tissue from PD patients and matched controls. The level of a key mediator of the UPR pathway, glucose-regulated protein 78 (GRP78), was significantly decreased in temporal cortex and cingulate gyrus, whereas there were no significant changes in the caudate nucleus, prefrontal, or parietal cortex regions. On the other hand, GRP78 mRNA level was significantly increased in caudate nucleus, cingulate gyrus, prefrontal, and parietal cortex regions. GRP78 protein level was also measured in plasma and cerebrospinal fluid, but there were no differences in these levels between PD patients and control subjects. Furthermore, immunofluorescence labeling of the CD4<sup>+</sup> T cells from PD patients showed that GRP78 protein is found in the cytoplasm. However, GRP78 level in PD patients was not significantly different from control subjects. Unlike the previous Lewy body dementia study, the present investigation reports reduced cortical protein, but increased transcript levels of GPR78 in PD. In summary, these data provide further evidence that GRP78 regulation is dysfunctional in the brains of PD patients.
Keywords: Parkinson's disease, unfolded protein response, glucose-regulated protein 78, endoplasmic reticulum stress, neurodegenerative diseases
#### INTRODUCTION
Parkinson's disease (PD) is one of the most common neurodegenerative disease affecting 1–2% of the population over 60 years of age (Dorsey and Bloem, 2018). PD is diagnosed based on the presence of bradykinesia, resting tremor, and postural rigidity. PD is characterized by a progressive degeneration of dopaminergic neurons in the substantia nigra pars compacta and deposits of intracellular protein inclusions called Lewy bodies, where aggregates of misfolded α-synuclein (α-syn) are the major components (Spillantini et al., 1997). Although these causative factors for PD have been known for many years and extensive research have been done to halt the disease progression, at present, there are no disease modifying therapies available for PD. Current treatments only restore dopamine neurotransmission and reduce symptoms, but do not stop or slow down the disease progression.
#### Edited by:
Victor Tapias, Weill Cornell Medicine, United States
#### Reviewed by:
Paula Garcia-Esparcia, Bellvitge Biomedical Research Institute, Spain Paolo Remondelli, University of Salerno, Italy
#### \*Correspondence:
Jean-Ha Baek [email protected] Per Svenningsson [email protected]
#### Specialty section:
This article was submitted to Neurodegeneration, a section of the journal Frontiers in Neuroscience
Received: 18 February 2019 Accepted: 19 June 2019 Published: 05 July 2019
#### Citation:
Baek J-H, Mamula D, Tingstam B, Pereira M, He Y and Svenningsson P (2019) GRP78 Level Is Altered in the Brain, but Not in Plasma or Cerebrospinal Fluid in Parkinson's Disease Patients. Front. Neurosci. 13:697. doi: 10.3389/fnins.2019.00697
**76**
Accumulation of specific misfolded proteins is a salient feature of many neurodegenerative diseases, including PD. The buildup of misfolded proteins gives rise to cellular stress, and is detected by specific sensors in the endoplasmic reticulum (ER). To overcome ER stress, mammalian cells activate a specific signaling pathway in the ER called the unfolded protein response (UPR), which is initiated by the binding of the glucose-regulated protein 78 (GRP78, also known as binding immunoglobulin protein, BiP), an ER chaperone, to misfolded proteins (Harding et al., 1999; Rutkowski and Kaufman, 2004). The UPR consists of three pathways, in which includes ER-resident transmembrane proteins, known as protein kinase RNA-like ER kinase (PERK), inositol-requiring enzyme 1 (IRE1), and activating transcription factor 6 (ATF6). UPR serves as a protective mechanism against the buildup of toxic misfolded proteins, in which first inhibits the protein synthesis, then up-regulate genes that are involved in protein folding or disposal in order to stabilize the disturbed ER homeostasis (Walter and Ron, 2011; Hetz, 2012). The early response of the UPR pathway is activated by PERK (Nakka et al., 2016), which leads to the phosphorylation of the eukaryotic translation initiator factor, eIF2α. This event inhibits general protein synthesis in order to reduce the load on the ER, having an important pro-survival role (Harding et al., 2000; Fernandez et al., 2002). However, under prolonged or chronic stress, phosphorylation of eIF2α increases translation of activating of transcription-4 (ATF4) mRNA, which encodes a transcription factor that induces the expression of pro-apoptotic genes such as the C/EBP-homologous protein (CHOP; also known as DDIT3/GADD153) (Zinszner et al., 1998). Although the initial intention of the UPR activation is to overcome ER stress, if a cell fails to reach proteostasis due to chronic or irreversible ER stress, then the UPR signals to cell death by apoptosis (Tabas and Ron, 2011; Urra et al., 2013).
It has been shown that the UPR is activated in post mortem human brains of PD patients (Hoozemans et al., 2007; Selvaraj et al., 2012) as well as in animal and cell models of parkinsonism (Mercado et al., 2016), implying that the neurons are prone to ER stress, and that the UPR may have a role in the degeneration of dopaminergic neuron. Numerous studies have demonstrated that the pathological aggregation/accumulation of α-syn activates the UPR pathway, consequently inducing proapoptotic changes (Cooper et al., 2006; Sugeno et al., 2008; Bellucci et al., 2011). Despite the fact that there is a general acceptance of the UPR activation in PD, previous human post mortem studies have only focused on demonstrating the presence/existence of UPR activation in PD through semiquantitative immunohistochemical approach. Therefore, the primary aim of the current study was to accurately quantify the changes in the level of UPR proteins and mRNA transcripts in PD patients in various brain regions by using western blot and quantitative real-time PCR.
The cerebrospinal fluid (CSF) has been extensively investigated as a source of robust biomarkers for neurodegenerative diseases, particularly for Alzheimer's disease (AD), but also for atypical parkinsonian disorders (Lleo et al., 2015). CSF is the biological fluid closest to the brain as it is not separated from the brain by the blood brain barrier, unlike plasma. However, at the moment, there is no CSF biomarker available to diagnose PD in clinics. The UPR proteins have never been investigated as a potential biomarker for PD. Therefore, the second aim of the present study was to investigate the possibility of the UPR proteins, specifically GRP78 protein, as a novel biomarker for PD.
### MATERIALS AND METHODS
#### Post Mortem Human Brain Tissues
Post mortem brain tissue was obtained from the MRC London Neurodegenerative Diseases Brain Bank, United Kingdom. All participants gave informed consent for their tissue to be used in research and the study had ethics approval from the UK National Research Ethics Service (08/H1010/4 and KI IRB) and from the Regional Ethics Review Board of Stockholm (2014/1366-31). The demographic details of the patients and control subjects are shown in **Table 1**. It is worth to declare that authors did not have any control over the sample collection, and hence, long post mortem delays of these brain samples could not be avoided. Biochemical analyses were undertaken on five different brain regions: caudate nucleus (n = 36), prefrontal cortex (n = 40), temporal cortex (n = 41), anterior cingulate gyrus (n = 38), and parietal cortex (n = 39). Temporal cortex tissues were not available for RNA analysis due to technical reasons. Caudate nucleus was selected for its involvement in motor function in PD; prefrontal cortex was selected for its proposed role in executive function and cognition; cingulate gyrus was selected for the early development of pathology encountered in this region, while parietal cortex was selected because of its pathological predominance in AD as opposed to PD; temporal cortex was chosen due to its suggested role in auditory processing and language.
#### Participants for CSF and Plasma Collection
Consents from participants were collected according to the Declaration of Helsinki, which was approved by the regional ethical committees. CSF and plasma samples were collected as described previously (Bjorkhem et al., 2013). All participants fulfilled the clinical diagnostic criteria for PD (Gelb et al., 1999), and PD severity was scored with the Unified Parkinson's disease rating scale (UPDRS) and Hoehn and Yahr scale. The Montreal Cognitive Assessment (MoCA) scores were also obtained. Control subjects were healthy volunteers or had mild symptoms without any severe neurological diagnosis (e.g., temporary tension headache or sensory symptoms). Control subjects were age- and gender-matched to PD patients (**Table 1**).
#### CSF and Plasma Collection
The standardized lumbar puncture procedure was performed according to the Alzheimer's Disease Neuroimaging Initiative recommended protocol. CSF was collected into sterile polypropylene tubes, in which the first 2 mL was discarded, and approximately 10–12 mL of CSF from the first portion
#### TABLE 1 | Demographical and clinical characteristics of subjects in this study.
PD, Parkinson's disease; F, female; M, male; n/a, not applicable; H&Y, Hoehn and Yahr; UPDRS, unified Parkinson's disease rating scale; MoCA, Montreal Cognitive Assessment; PBMCs, peripheral blood mononuclear cells. All measures are presented as mean + SEM, except H&Y score is presented as median ± SEM.
was collected and gently mixed in order to minimize the gradient influence. Cell counts were measured and samples were centrifuged in the original tube at 1800 × g for 10 min at 4◦C. Blood was collected in ethylenediaminetetraacetic acid (EDTA) tubes and centrifuged at 800 × g for 20 min. Plasma was collected from the top phase of the gradient. Both CSF and plasma were aliquoted in polypropylene tubes, frozen on dry ice and stored at −80◦C until use. The maximum time interval from the sample collection until freezing was 30 min.
### Preparation of Tissue Samples for Western Blotting
Western blot samples were prepared as previously described (Baek et al., 2016). Briefly, 100 mg of frozen tissue was taken from each brain region, which was then homogenized in 1 mL of ice cold buffer (pH 7.4) containing 50 mM Tris–HCL, 5 mM ethylene glycol-bis(β-aminoethyl ether)-N,N,N<sup>0</sup> ,N0 -tetraacetic acid (EGTA), 10 mM EDTA, "complete protease inhibitor cocktail tablet" (Sigma), phosphatase inhibitor (PhosStop, Sigma), and 2 µg/mL pepstatin A dissolved in ethanol:dimethyl sulfoxide (DMSO) 2:1 (Sigma). Homogenization was performed using disposable pestles (Cat# BELAF199230001, VWR) until the liquid appeared homogenous. Protein concentration of each sample was measured by using BCA Protein Assay Kit (Thermo Fisher Scientific).
#### Western Blotting
Twenty micrograms of each sample was loaded on 8% SDS– polyacrylamide gel for protein separation then transferred to nitrocellulose membrane (Immobilon-F, Millipore). After blocking for nonspecific binding, the membranes were incubated with anti-GRP78 (rabbit polyclonal, 1:1000, Cat# ab21685, Abcam), anti-eIF2α (rabbit polyclonal, 1:1000, Cat# 9722, Cell signaling), anti-phosphorylated eIF2α (rabbit polyclonal, 1:1000, Cat# 9721, Cell signaling) primary antibodies followed by IRDye 800CW goat anti-rabbit secondary antibody (1:20,000, Cat#
926-32211, Li-Cor). Bands were detected using an Odyssey infrared fluorescent scanner, and the integral of intensity was quantified using Odyssey infrared imaging system application software version 2.1. β-Actin was chosen as a "house-keeping" protein in order to control for any inconsistency in loading samples. Each membrane was therefore probed for actin (mouse monoclonal, 1:10,000, Cat# A5441-100UL, Sigma–Aldrich) to normalize the level of immunolabeling of the protein-of-interest to actin, so that any potential variations in protein loading could be eliminated.
### RNA Extraction and Quantitative Real-Time PCR (qRT-PCR)
Total RNA was extracted from 30 mg of frozen human brain tissues using RNeasy Plus Mini Kit (Qiagen) according to manufacturer's protocol. The samples were then measured and evaluated for concentration and purity (260/280 nm ratio) using a Nanodrop (Marshall Scientific). RNA samples were stored at -80◦C until use. cDNA was synthesized from 30 ng of total RNA using QuantiTect Reverse Transcription Kit (Qiagen). Levels of human GRP78 (Assay ID: Hs00946084\_g1), eIF2α (Assay ID: Hs00187953\_m1), and CHOP (Assay ID: Hs00358796\_g1) transcripts were measured by qRT-PCR. Briefly, qRT-PCR reactions were prepared in duplicate for each sample with TaqMan assay (Thermo Fisher Scientific) and performed on a CFX96 Real-Time System (BioRad). All reactions were run at 55◦C as an annealing temperature and for 40 s for elongation time. Transcript levels were determined by the comparative cycle threshold method and glyceraldehyde-3 phosphate dehydrogenase (GAPDH) (Assay ID Hs02758991\_g1) was used as an internal control for normalization.
### Quantification of Circulating GRP78 Protein by ELISA
The levels of GRP78 protein were measured in CSF (neat, i.e., no dilution) and in plasma (1:20 dilution) using the human GRP78 ELISA kit (Enzo Life Sciences) with a detection range of 1.4–4500 ng/mL. The assay was performed according to manufacturer's instruction.
### Immunofluorescence Labeling of CD4<sup>+</sup> T Cells
Preparation of peripheral blood mononuclear cells (PBMCs) was performed as previously described (Green et al., 2017). CD4<sup>+</sup> T cells were isolated by using CD4<sup>+</sup> T Cell Isolation Kit (Cat# 130-096-533, Miltenyi Biotec) according to manufacturer's instructions. Purified CD4<sup>+</sup> T cells were resuspended in proliferation medium (RPMI 1640, glutamine, Pen/Strep, heat inactivated FCS, β-mercaptoethanol, non-essential amino acids, sodium pyruvate) containing Il-2 and Il-7 cytokines. Cells were then seeded in 96-well plates, pre-coated with 5 µg/mL anti-CD3 (clone 2C11), and 1 µg/mL anti-CD28 (clone 37.51) antibodies. The plate was then incubated in a humidified incubator with 5% CO<sup>2</sup> at 37◦C for 7 days. After 7 days of incubation, cells were transferred to 10 µg/mL human ICAM-1-coated µ-slides (Ibidi), and incubated for 45 min at 37◦C for adhesion and migration. Cells were then fixed and permeabilized for 20 min using Cytofix/Cytoperm solution (BD Biosciences). Cells were incubated with anti-GRP78 primary antibody (1:50, Cat# ab21685, Abcam), anti-α-syn primary antibody (1:50, Cat# 610787, BD Transduction), anti-calreticulin primary antibody (1:50, Cat# ab22683, Abcam) for overnight at 4◦C, followed by goat anti-mouse Alexa FluorTM 488 secondary antibody (1:500, Thermo Fisher Scientific), goat anti-rabbit Alexa FluorTM 568 secondary antibody (1:500, Thermo Fisher Scientific) for 1 h at room temperature. All cells were counterstained with DAPI nuclear stain (300 nM, Sigma–Aldrich).
### Confocal Microscopy and Image Analysis
Immunofluorescent images were acquired by ZEISS LSM 880 Airyscan confocal laser scanning microscopy equipped with ZEN2.1 software, using Plan-Apochromat 63×/1.4 Oil DIC M27 63× oil objective. The quantification of the GRP78 protein expression in CD4<sup>+</sup> T cells was pooled from five independent experiments, in which more than 20 cells were analyzed from each group (control or PD) for every experiment. The analyses were done using ImageJ. The outline of a cell was defined by image threshold, and the total immunofluorescence was measured by maximum projections of Z stack images after background subtraction. The mean fluorescent intensity of PD group was normalized to that of control group.
#### Statistical Analysis
Statistical analysis was carried out using GraphPad Prism 5. All descriptive statistics for the variables in the study were reported as means ± standard error of means (SEM), unless otherwise stated. Normality tests were run to assess data distribution. One-way ANOVA or parametric unpaired t-test was used for variables with normal distribution, whereas Mann–Whitney non-parametric analysis was used for the distorted distribution. Differences were considered statistically significant with P < 0.05.
### RESULTS
### Changes in the Level of GRP78 Protein and the Ratio Between p-eIF2α and Total eIF2α Proteins in Various Regions of the Brain of PD Patients
There was a significant decrease in the level of GRP78 protein in PD patients compared to control subjects in temporal cortex (P = 0.0007) and cingulate gyrus (P = 0.001, **Figure 1A**). Similar pattern was observed in the prefrontal cortex, in which the difference was very close to statistical significance (P = 0.0663, **Figure 1A**). In caudate and parietal cortex regions, there was an increasing trend in the level of GRP78 protein in PD patients compared to control subjects, though statistically insignificant (**Figure 1A**).
The ratio between phosphorylated-eIF2α (p-eIF2α) and total eIF2α proteins was significantly decreased in PD patients compared to control subjects in prefrontal cortex (P = 0.0115,
FIGURE 1 | Changes in the level of GRP78 protein and the ratio between p-eIF2α and total eIF2α proteins in various regions of the brain from PD patients and control subjects. (A) The level of GRP78 protein was significantly decreased in temporal cortex and cingulate gyrus regions (P = 0.0007; P = 0.001, respectively). (B) The ratio between p-eIF2α and total eIF2α was significantly decreased in prefrontal cortex (P = 0.0115). There were no changes in the other regions. (C) A representative western blot image showing GRP78, total eIF2α, and p-eIF2α protein expressions in cingulate gyrus region in control and PD subjects. (D) The level of GRP78 protein in cingulate gyrus was significantly higher in PD with dementia (PDD) and dementia with Lewy bodies (DLB) patients compared to both control subjects and PD patients (∗∗P = 0.0023; ∗∗∗∗P < 0.0001 compared to control; #P < 0.0001 compared to PD patients). (E) Representative western blot image illustrating the expression of GRP78 protein in cingulate gyrus in control, PD, PDD, and DLB subjects.
**Figure 1B**). This decrease was due to a significant increase in the level of total eIF2α protein (P = 0.002) in PD patients, while the p-eIF2α protein level remained unchanged (**Table 2**). In the temporal cortex, p-eIF2α and total eIF2α ratio was decreased almost to a significant level in PD patients compared to control subjects (P = 0.0514, **Figure 1B**). There were no changes in p-eIF2α and total eIF2α ratio in caudate, cingulate gyrus, or parietal cortex (**Figure 1B**). In order to validate this finding of a general decrease in the levels of GRP78 protein in PD patients, the western blot experiment from Baek et al. (2016) was repeated (**Figures 1D,E**), in which the authors showed that the levels of GRP78 protein in patients with Parkinson's disease with dementia (PDD) and dementia with Lewy bodies (DLB) were significantly higher compared to control subjects and AD patients in the cingulate gyrus. As expected, the results were consistent with the results of Baek et al. (2016), in which the level of GRP78 protein in cingulate gyrus was significantly higher in PDD and DLB patients compared to control subjects (∗∗∗∗P < 0.0001 for both PDD and DLB) and also to PD patients (#P < 0.0001 for both PDD and DLB) (**Figures 1D,E**).
### Changes in the mRNA Level of Grp78, eif2α, and Chop in Various Regions of the Brain of PD Patients
There were significant increase in the levels of Grp78 mRNA transcripts in all regions of the brain in PD patients compared to control subjects (caudate, P = 0.0015; prefrontal cortex, P = 0.0025; cingulate gyrus, P = 0.0007; parietal cortex, P = 0.0047; **Figure 2A** and **Table 3**). However, eif2α and Chop mRNA levels in PD patients were not significantly different to control subjects in any of the brain regions (**Figures 2B–D** and **Table 3**).
### Circulating GRP78 Protein in Plasma and CSF of PD Patients
There was a high concentration of GRP78 protein in the plasma of PD patients as well as in control subjects. Although there was a slight decrease in GRP78 protein level in PD patient compared to control subjects, it was not statistically significant (**Figure 3A**). In contrast to plasma level, the concentrations of GRP78 protein in CSF of control and PD patients were negligible, and there were no differences between PD patients and control subjects (**Figure 3B**).
## GRP78 Protein Expression in CD4<sup>+</sup> T Cells of PD Patients
Double immunofluorescence staining data with α-syn and calreticulin showed cytoplasmic localization of GRP78 in CD4<sup>+</sup> T cells (**Figures 4A,B**). The GRP78 protein was expressed in the cytoplasm of CD4<sup>+</sup> T cells in both control subjects and in PD patients (**Figure 4C**). However, the level of expression in PD patients was not significantly different from that of control subjects (**Figure 4D**).
## DISCUSSION
Growing experimental evidence suggests that UPR is involved in PD; and this is not surprising since the accumulation of misfolded α-syn is a central pathogenic process in PD. Indeed, changes in the expression of GRP78 protein and other UPR activation markers (p-PERK and p-eIF2α) have been observed in the brain of PD patients (Hoozemans et al., 2007; Selvaraj et al., 2012; Baek et al., 2016; Mercado et al., 2018). Nevertheless, there are only a handful of studies examining UPR activation in PD using post mortem human brain tissues, and furthermore, almost all of these human studies employ semi-quantitative immunohistochemical method. In the present study, changes in the levels of UPR proteins in various regions of the brain from PD patients were measured by means of quantitative western blot. Surprisingly, among five regions analyzed, there were significant decreases in the level of GRP78 protein in temporal cortex and cingulate gyrus of PD patients compared to control subjects, while there were no changes in caudate, prefrontal, or parietal cortical regions (**Figure 1A**). Although the observed decrease in the level of GRP78 protein in PD patients is in contrast to previous post mortem human studies, this decrease has been observed in other neurodegenerative diseases and during normal aging. For example, in the brain of AD patients, the level of GRP78 protein has been shown to be reduced compared to control subjects
TABLE 2 | Summary of changes in the UPR proteins in PD patients compared to control subjects.
Upward arrow: An increase in expression. Downward arrow: A decrease in expression. ns: Not significant.
(Katayama et al., 1999; Baek et al., 2016) or remained unchanged (Sato et al., 2000). In a mouse model of over-expressing mutated human presenilin-1 gene, the most prevalent mutation found in cases of familial AD, the expression of GRP78 was also found to be decreased (Katayama et al., 1999). Furthermore, Salganik et al. (2015) have shown the loss of GRP78 during normal aging, in
TABLE 3 | Summary of changes in the UPR mRNA in PD patients compared to control subjects.
Upward arrow: An increase in expression. Downward arrow: A decrease in expression. ns: Not significant.
which the old rats showed significantly lower levels of GRP78 protein in the nigrostriatal system compared to young animals. In the same study, it was shown that knockdown of GRP78 by specific small interfering RNAs in a rat model of over-expressing α-syn in the substantia nigra aggravated α-syn neurotoxicity in nigral dopamine neurons, which then lead to significantly greater neuronal loss and reduction of striatal dopamine. Moreover, the degree of GRP78 decline was correlated to the severity of neurodegeneration (Salganik et al., 2015).
In the present study, it was shown that changes in the level of GRP78 protein in a particular brain region did not directly correlate with the changes in the levels of downstream proteins such as eIF2α or p-eIF2α (**Figure 1B**). For instance, in the cingulate gyrus, there was a significant decrease in GRP78 protein level in PD patients compared to control subjects (**Figure 1A**), but there was no change in the ratio between p-eIF2α and eIF2α levels (**Figure 1B**). Similar results were found in a rodent study, in which neuronal GRP78 induction in α-syn over-expressing transgenic mice was not accompanied by an increase of p-eIF2α level, suggesting that α-synucleinopathy is linked to abnormal UPR which in turn could trigger cell death (Colla et al., 2012a). Taken together, absence of coherent changes in the level of proteins in the UPR pathway observed in the current study provide additional support for the above hypothesis that PD may be associated with impaired UPR.
Unlike GRP78 protein, the levels of GRP78 mRNA were significantly increased in PD in all areas of the brain analyzed (**Figure 2A**). However, mRNA levels of eIF2α or CHOP, a pro-apoptotic transcription factor, did not show any significant changes in PD patients compared to control subjects (**Figures 2B,C**). To explore the possibility whether the increased GRP78 mRNA expression is a compensatory response to the decreased protein level, individual GRP78 mRNA level was correlated to its matching protein level. Nevertheless, no correlation was found (data not shown). The mismatch between the levels of GRP78 mRNA and protein in PD further indicates that the UPR signaling may be dysfunctional, and that PDrelated pathology that causes ER stress, likely the accumulation of misfolded α-syn, in some way impair the induction of GRP78 protein, which may also indicate an increase of the vulnerability of neurons to ER stress. Indeed, it was shown that α-syn inhibited processing of ATF6 directly via physical interactions and indirectly by inhibiting ER to Golgi transport of COPII vesicles (Credle et al., 2015). Moreover, the disease-causing mutant α-syn also reduced ER to Golgi trafficking and aggravated ER stress (Thayanidhi et al., 2010; Colla et al., 2012a,b). The phenomenon that decrease in UPR signaling leading to a possible increase in sensitivity to ER stress was demonstrated by Katayama et al. (1999). They showed that inhibition of endogenous IRE1 significantly increased vulnerability to ER stress, and increase in sensitivity to ER stress caused by treatment of an ER stress inducer was reversed by the expression of recombinant GRP78. The results from Katayama et al. (1999), together with the current study, suggest that activation of UPR signaling is important for protective effects against the ER stress, and that the reduction of GRP78 protein level may cause vulnerability to ER stress in PD, which may then potentiate disease progression.
It has been hypothesized that with aging and/or disease progression, soluble α-syn monomers that are present in the ER form insoluble α-syn oligomers/aggregates and attribute to chronic ER stress and neurodegeneration (Colla et al., 2012a). Recently, a new ER stress rat model using intranigral injection of a well-known ER stress inducer, tunicamycin, was employed to investigate whether ER stress is able to induce PD features (Coppola-Segovia et al., 2017). It was shown that ER stress not only induced locomotor impairment and dopaminergic neuronal loss, but also substantial α-syn oligomerization in substantia nigra pars compacta, astroglial activation, and increased expression of ER stress markers (Coppola-Segovia et al., 2017). These results reinforce the notion that ER stress, hence the UPR, could be an important contributor to the pathophysiology of PD. Nevertheless, there are still not enough evidence to decipher the fact whether the activation of the UPR is a cause or consequence of neurodegeneration observed in PD. Furthermore, how and what controls the changeover switch of the UPR between neuroprotection and neurotoxicity remains largely obscure.
While the present study confirms that there is a turbulence in the UPR in different areas of the brain of PD patients, due to technical limitations with reagents, it was difficult to efficiently
FIGURE 4 | GRP78 protein cellular localization and expression in CD4<sup>+</sup> T cells derived from control subjects and PD patients. (A) Double immunofluorescence staining of GRP78 (red) and α-synuclein (green) in CD4<sup>+</sup> T cells derived from control subjects. (B) Double immunofluorescence staining of GRP78 (red) and calreticulin (green) in CD4<sup>+</sup> T cells derived from control subjects. (C) There was a moderate level of GRP78 protein expression (red) in both control subjects and PD patients. GRP78 was localized in the cytoplasm. Scale bar represents 10 mm. (D) The level of GRP78 protein expression in PD patients was not significantly different from control subjects.
evaluate the state of other UPR reporters such as XBP1 and PERK. Since induction of the ER chaperones and the XBP1 cleavage can occur independent of UPR mechanisms (Kim et al., 2008), lack of GRP78 and/or p-eIF2α inductions in PD patients observed in the present study may reflect activation of processes other than UPR. Moreover, to further validate the UPR changes observed in the present study, changes in the level of other ER resident proteins, such as GRP94, calnexin, and protein disulfide isomerase, should also be examined in the future.
Given that the UPR is a homeostatic stress response, it means that it is greatly controlled by positive and negative feedback loops. The interaction between the three signaling pathways of the UPR proposes that the alteration of one pathway will affect signaling of the other two pathways also. For example, inhibition of one pathway may in fact increase signaling through one of the other pathways. This phenomenon was demonstrated by Harding et al. (2001) where deletion of PERK resulted in an increased activity of IRE1α. Therefore, additional studies investigating the changes in the proteins of the other two arms of the UPR pathway, that is, IRE1α and ATF6, are needed to fully understand the relationship between different arms of the UPR pathway in PD, and how each arm of the UPR pathway is involved in the neurodegenerative process in PD.
Diagnosis of PD remains tricky and misdiagnosis rate with vascular or atypical parkinsonian disorders reaches up to 20–30% (Rajput and Rajput, 2014). At present, the assessment of clinical motor symptoms underlies the diagnosis of PD. The absence of a reliable biomarker with high sensitivity and specificity has significantly hindered the validation of potential therapies. One of the novel results of this study was the measurement of the GRP78 protein in plasma and CSF samples of PD patients as to discover a potential biomarker for PD. Although GRP78 protein has been detected in the plasma of endometrial cancer and obese patients (Ciortea et al., 2016; Khadir et al., 2016), it has never been measured in PD patients. In the plasma from control subjects and PD patients, high levels of GRP78 protein were detected, though its level in PD patients was not significantly different to control subjects (**Figure 3A**). To the best of authors' knowledge, the present study is first to evaluate the presence of GRP78 protein in CSF of PD patients. Nevertheless, unlike plasma, GRP78 protein in CSF was almost undetectable in both control subjects and PD patients (**Figure 3B**). A possible reason for the undetectable level of GRP78 in CSF could be that "whole" GRP78 protein is too large to be secreted in CSF. Therefore, it would be interesting to investigate whether fragments of GRP78 protein could be detected in CSF. Although the current study did not provide a clear evidence that GRP78 could serve as a possible biomarker for PD, further studies are required in order to investigate whether other UPR proteins and/or ER stress-related proteins have potential to be novel biomarkers for PD. A relatively high level of GRP78 protein in plasma from both control subjects and PD patients (**Figure 3A**) led us to further understand the implication(s) of this result in the disease state. Therefore, we investigated the level of GRP78 protein expression in PBMCs, specifically in CD4<sup>+</sup> T cells derived from control subjects and PD patients. However, when we compared the expression level of GRP78 in PD patients vs. control subjects, we found no difference (**Figures 4C,D**). Delpino and Castelli (2002) showed that extracellular GRP78 are mostly derived from an active release from living cells and are not solely due to the protein leakage from dead cells. Recent studies have also demonstrated that GRP78 release is increased in cancer, obesity, or upon ER stress (Khadir et al., 2016; Steiner et al., 2017). Taken together, it may be hypothesized that the high level of GRP78 protein observed in plasma of PD patients could be due to circulating PBMCs releasing GRP78 into the extracellular domain. However, since the level of GRP78 protein in plasma in PD patients was not different from control subjects, further investigation is inevitable to determine the difference in the state of GRP78 protein between control subjects and PD patients.
### CONCLUSION
In conclusion, the present study showed that there are changes in the level of UPR proteins and mRNAs, particularly GRP78, in various regions of the brain of PD patients (**Figures 1**, **2**). However, while there were central changes, there were no peripheral changes, as observed in the levels of GRP78 protein in CSF, plasma, and in immune cells (**Figures 3**, **4**). Based on these results, one can cautiously postulate that UPR changes may be limited to the site of neurodegeneration, and not influenced elsewhere. In other words, it may be that the UPR in PD is quite a specific response in terms of location of action rather than a generic reaction to ER stress or progression of PD. This highlights an attractive opportunity to explore the UPR as a novel therapeutic target for PD with negligible peripheral side effects.
### DATA AVAILABILITY
All datasets generated for this study are included in the manuscript and/or the Supplementary Files.
#### ETHICS STATEMENT
This study was carried out in accordance with the recommendations of the Alzheimer's Disease Neuroimaging Initiative recommended protocol with written informed consent from all subjects. All subjects gave written informed consent in accordance with the Declaration of Helsinki. The protocol was approved by the UK National Research Ethics Service and by the Regional Ethics Review Board of Stockholm.
#### AUTHOR CONTRIBUTIONS
fnins-13-00697 July 4, 2019 Time: 16:10 # 11
J-HB and PS conceived and designed the experiments. J-HB, BT, MP, DM, and YH performed the experiments. J-HB, BT, MP, and
#### REFERENCES
DM analyzed the data. J-HB wrote the manuscript. All authors have read and approved the final manuscript.
#### FUNDING
This work was financially supported by grants from the Swedish Foundation for Strategic Research (SBE 13-0115; #RIF14-0078) and ALF. PS is a Wallenberg Clinical Scholar.
**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Copyright © 2019 Baek, Mamula, Tingstam, Pereira, He and Svenningsson. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
# Bioenergetics and Autophagic Imbalance in Patients-Derived Cell Models of Parkinson Disease Supports Systemic Dysfunction in Neurodegeneration
Ingrid González-Casacuberta1,2, Diana Luz Juárez-Flores1,2, Constanza Morén1,2 and Gloria Garrabou1,2 \*
<sup>1</sup> Muscle Research and Mitochondrial Function Laboratory, Cellex-IDIBAPS, Faculty of Medicine and Health Sciences-University of Barcelona, Internal Medicine Service-Hospital Clínic of Barcelona, Barcelona, Spain, <sup>2</sup> CIBERER-U722, Madrid, Spain
#### Edited by:
Victor Tapias, Weill Cornell Medicine, United States
#### Reviewed by:
Renato Xavier Coelho dos Santos, University of Aberdeen, United Kingdom Diana F. F. Silva, University of Coimbra, Portugal
> \*Correspondence: Gloria Garrabou [email protected]
#### Specialty section:
This article was submitted to Neurodegeneration, a section of the journal Frontiers in Neuroscience
Received: 12 April 2019 Accepted: 09 August 2019 Published: 10 September 2019
#### Citation:
González-Casacuberta I, Juárez-Flores DL, Morén C and Garrabou G (2019) Bioenergetics and Autophagic Imbalance in Patients-Derived Cell Models of Parkinson Disease Supports Systemic Dysfunction in Neurodegeneration. Front. Neurosci. 13:894. doi: 10.3389/fnins.2019.00894 Parkinson's disease (PD) is the second most prevalent neurodegenerative disorder worldwide affecting 2–3% of the population over 65 years. This prevalence is expected to rise as life expectancy increases and diagnostic and therapeutic protocols improve. PD encompasses a multitude of clinical, genetic, and molecular forms of the disease. Even though the mechanistic of the events leading to neurodegeneration remain largely unknown, some molecular hallmarks have been repeatedly reported in most patients and models of the disease. Neuroinflammation, protein misfolding, disrupted endoplasmic reticulum-mitochondria crosstalk, mitochondrial dysfunction and consequent bioenergetic failure, oxidative stress and autophagy deregulation, are amongst the most commonly described. Supporting these findings, numerous familial forms of PD are caused by mutations in genes that are crucial for mitochondrial and autophagy proper functioning. For instance, late and early onset PD associated to mutations in Leucine-rich repeat kinase 2 (LRRK2) and Parkin (PRKN) genes, responsible for the most frequent dominant and recessive inherited forms of PD, respectively, have emerged as promising examples of disease due to their established role in commanding bioenergetic and autophagic balance. Concomitantly, the development of animal and cell models to investigate the etiology of the disease, potential biomarkers and therapeutic approaches are being explored. One of the emerging approaches in this context is the use of patient's derived cells models, such as skin-derived fibroblasts that preserve the genetic background and some environmental cues of the patients. An increasing number of reports in these PD cell models postulate that deficient mitochondrial function and impaired autophagic flux may be determinant in PD accelerated nigral cell death in terms of limitation of cell energy supply and accumulation of obsolete and/or unfolded proteins or dysfunctional organelles. The reliance of neurons on mitochondrial oxidative metabolism and their post-mitotic nature, may explain their increased vulnerability to undergo degeneration upon mitochondrial challenges or autophagic insults. In this scenario, proper mitochondrial function and
**88**
turnover through mitophagy, are gaining in strength as protective targets to prevent neurodegeneration, together with the use of patient-derived fibroblasts to further explore these events. These findings point out the presence of molecular damage beyond the central nervous system (CNS) and proffer patient-derived cell platforms to the clinical and scientific community, which enable the study of disease etiopathogenesis and therapeutic approaches focused on modifying the natural history of PD through, among others, the enhancement of mitochondrial function and autophagy.
Keywords: neurodegeneration, mitochondria, autophagy, Parkin, LRRK2, fibroblasts
### PARKINSON'S DISEASE
Parkinson's disease (PD) is the most common movement disorder and the second most frequent neurodegenerative disease affecting more than 6.5 million people worldwide (Teves et al., 2017), representing 2–3% of the population over 65 years (Connolly and Lang, 2014; Poewe et al., 2017). As the global life expectancy increases, the number of people with PD is expected to rise by more than 50% in 2030, constituting an important burden for public health (Kalia and Lang, 2015).
Although it was already known in ancient India under the name of "Kampavata," PD was first described in 1817 by James Parkinson in the Essay on the Shaking Palsy and later on refined and expanded by Jean-Martin Charcot who named the disorder "malaldie de Parkinson"(Kempster et al., 2007; Corti et al., 2011; Goetz, 2011). Despite PD was described more than two centuries ago, the conceptualization of the disease continues to evolve and it is now recognized as a systemic disease with multiple layers of complexity.
The cardinal symptoms of PD described by James Parkinson in 1817 and then refined by Jean-Martin Charcot, include bradykinesia, muscular rigidity, rest tremor and postural and gait impairment (Goetz, 2011) (**Figure 1**).
Pathologically, PD is a complex neurodegenerative disorder characterized by the prominent death of dopaminergic neurons (DAn) in the substantia nigra (SN) pars compacta (SNpc) located in the mesencephalon and the consequent striatal dopamine (DA) deficit that leads to the classical motor symptoms of the disease (Kalia and Lang, 2015). In addition to the loss of DAn, another hallmark of PD is the presence of intraneuronal inclusions in the soma of the remaining DAn. These inclusions, named Lewy bodies (LB) as well as Lewy neurites (LN), are collectively referred as Lewy pathology (LP). LB are round eosinophilic inclusions mainly formed by insoluble α-synuclein aggregates as well as ubiquitin and other proteins (Shults, 2006). The aggregation of these misfolded proteins has been shown to be common to PD, dementia with LB, and multiple system atrophy (Goedert et al., 2013).
#### Clinical Features
The dramatic loss of DAn in the SNpc, even in early stages of the disease, suggests that the degeneration in this region starts long before the motor symptoms appear (Poewe et al., 2017). In this context, PD is considered to occur in three stages: preclinical PD, when the neurodegeneration has started but no clinical signs or symptoms are present; premotor or prodromal PD, when clinical signs and/or symptoms are present but are insufficient to establish a diagnosis of PD; and clinical PD, when the diagnostic criteria are met (Kalia and Lang, 2015, 2016). Non-motor features are frequently present in the prodromal phase of the disease, which can last for 20 years or more, and involve a multitude of non-motor features including rapid eye movement (REM) sleep behavior disorder (RBD), constipation and hyposmia, as well as depression and anxiety (Poewe, 2008).
In this scenario, it is now widely accepted that PD is not a movement disorder simply induced by the loss of the DAn in the SNpc. The SN is not the only damaged region in PD, nor the first affected one. Brain sites other than the SN, such as the cerebral cortex and the limbic system, have also been reported to be impaired in patients during the presymptomatic phase (Dickson, 2018). In fact, several studies have shown that the degenerative process in PD is much more extensive and affects not only the central nervous system (CNS) but also the peripheral autonomic nervous system (PNS) and the organs outside the brain that the latter innervates (Braak et al., 2004). PNS dysfunction underlies the presence of some of the specific non-motor features that appear in the prodromal phase of PD and remain present over the course of the disease (Jain, 2011). In line with this, LP is not restricted to the brain but has also been encountered in the spinal cord and PNS including the vagus nerve, sympathetic ganglia, cardiac plexus, enteric nervous system (ENS), salivary glands, adrenal medulla, cutaneous and sciatic nerves (Tolosa and Vilas, 2015).
### Risk Factors
Parkinson's disease was thought to be primarily caused by environmental factors, but research reveals that the disease develops from a complicated interplay of ageing, genetics and environment. In fact, the vast majority of cases occur sporadically and genetic forms of the disease account for about 10% of patients (Klein and Westenberger, 2012).
In general, the average age of onset of PD is the late fifties, with a broad range from <40 to >80 years of age depending, among others, on its pattern of inheritance. Young-onset PD is commonly defined by an age of onset <45 years and >10% of these cases have a genetic basis; the proportion of genetically defined cases rises to >40% of those with disease onset before 30 years of age (Alcalay et al., 2010; Marder et al., 2010).
genes. RBD, rapid eye movement sleep behavior disorder; PNS, peripheral autonomic nervous system; CMA, chaperon-mediated autophagy; MAMs, mitochondrial associated membranes; NSAIs: non-steroidal antinflammatory drugs; Ca2+, calcium.
#### Non-genetic Risk Factors
The greatest risk factor for the development of neurodegenerative diseases, including PD, is ageing (Kalia and Lang, 2015). Incidence increases nearly exponentially from the sixth to the ninth decade of life by 5–10 fold (Kalia and Lang, 2015; Poewe et al., 2017). Total global prevalence is 0.3% and rises with age up to 3% in those >80 years of age (Poewe et al., 2017) (**Figure 1**). Many lines of evidence suggest that some molecular pathways including mitochondrial dysfunction, oxidative stress and abnormal cell wasting clearance (autophagy) have a central role in both, physiological aging and age-related neurodegenerative diseases, such as PD (Lin and Beal, 2006). This may be especially relevant in neurons, due to its postmitotic nature and scarce replacement, that prone them to store defects as they age. Accelerated or healthy aging and all factors responsible of modulating brain fragility play a major role in PD development, together with genetic and environmental factors.
Gender has also been reported to be a risk factor for PD, with approximately a 3:2 male-to-female ratio. Sex hormones have been proposed to play a neuroprotective role in the disease. In case of female hormones, the antioxidant capacity of estradiol, for instance, has been proposed to prevent neurodegeneration (Aguirre-Vidal et al., 2017). Interestingly, estradiol has also been demonstrated to activate metabolic signaling by regulating mitochondrial function, emerging as protective hormone in case of bioenergetic deficits (Pozdniakova et al., 2018). The neuroprotecting role of progesterone is also being evaluated (Bourque et al., 2019). Alternatively, gender associated differences could also be associated to sex-associated genetic mechanisms, to specific differences in exposure to environmental cues or to the contribution of inequality in health care (Kalia and Lang, 2015). Interestingly, in a few populations, including one study from Japan, no differences in gender, and even increased prevalence in females, was observed (Kusumi et al., 1996). The explanation for equal gender PD prevalence in these populations remains elusive. Dietetic, sociocultural, economic or even hormonal or molecular characteristics of Japanese population (as particular mitochondrial DNA haplogroups) may justify gender equivalence in PD development.
Other risk factors for PD are directly associated with environmental features are pesticide exposure, rural living, agricultural occupation, well-water drinking, prior head injury and β-blocker use. A special mention must be done for exposure to 1-methyl-4-phenyl-1,2,3,6-tetrahydro pyridine (MPTP), with
similar structure of some herbicides that increase the risk of PD, due to its historical importance. Since its incidental discovery in drug abusers after inadvertent self-administration, MPTP has been widely used to induce PD in animal models (Zhang et al., 2019). MPTP is lipid-soluble molecule that penetrates the blood–brain barrier and, once converted to its oxidized product (MPP+), interferes with mitochondrial respiration. MPP+ acts specifically in mitochondrial complex I, such as other toxic PD inducers (including rotenone and pesticides). Blockade of mitochondrial respiration has three main toxic consequences to cells: the inhibition of ATP generation and associated bioenergetic failure, the derived elevation of intracellular Ca2<sup>+</sup> that promotes cell death and the promotion of oxidative stress responsible of cell damage, all hallmarks of PD pathogenesis.
In contrast, tobacco smoking, coffee drinking, non-steroidal anti-inflammatory drug use, calcium channel blocker intake, and alcohol consumption have been associated with a decreased risk of PD development (Kalia and Lang, 2015).
Additionally, the incidence of PD seems to vary within different ethnicities. The prevalence of PD is high in Ashkenazi Jews of Israel, Inuit, Alaska native and native American populations (de Lau and Breteler, 2006; Poewe et al., 2017) and only one study reports that PD is more common in Hispanic and non-Hispanic whites compared to African Americans and Asian in United States (Van Den Eeden et al., 2003). The higher incidence of PD in these populations has been classically attributed to specific genetic burden.
#### Genetic Risk Factors
There are two different classes of genetic contributions to PD: gene mutations directly associated to genetically inherited forms of PD and genetic variations, indirectly accounting for risk factors of disease.
With respect to the first, the existence of heritable forms of PD was originally established through the discovery in 1997 that mutations in SNCA, the gene encoding for the α-synuclein protein, caused PD and the demonstration that α-synuclein was the major component of LB (Corti et al., 2011). Since then, the list of mutations causing monogenic types of PD continues to grow associated either to dominant or recessive inherited forms of PD (**Table 1** and **Figure 1**). Currently 8–10% of cases are familial (result from a genetic alteration leading to PD). They are caused by a subset of locus usually encoded by the prefix PARK and a number referring to their order of discovery. Those mutations affecting genes with an autosomal recessive pattern of inheritance usually result in early onset cases of PD, while mutations affecting genes with dominant autosomal inheritance usually cause forms of PD that resemble late-onset idiopathic PD.
The discovery of the molecular pathways orchestrated by proteins encoded by these genes associated with monogenic forms PD, have reinforced the notion that impaired mitochondrial and autophagy homeostasis are key events in disease etiology (Park et al., 2018). In fact, impaired mitochondrial function and autophagy have been directly linked to mutations of at least eleven of the genes associated to familial PD (**Table 1**). Among them, Leucine-rich repeat kinase 2 (LRRK2) and Parkin (PRKN) genes emerge amongst the most frequent forms of autosomal dominant and recessive forms of PD, respectively.
With respect to the second, genetic risk factors of PD account for the rest of 90–92% of non-inherited cases of PD, so-called idiopathic. They are caused by the complex interplay of an array of unknown factors, a part from the numerous genetic risk factors of PD (see **Table 1**): modifying effects by susceptibility alleles, environmental exposures and gene-environmental interactions that may condition gene expression. Some of these genetic risk factors that may conditionate the development of PD are in common to lysosomal storage diseases, mitochondrial pathologies or genes governing autophagic processes.
The definition of genetic and environmental cues in the development of PD is one of the novel areas of study in which growing and coming efforts should be focused and, probably, thanks to the development of new generation sequencing tools, the discovery of novel gens responsible of PD will arise, together with new putative genetic risk factors, thus reducing the number of idiopathic PD cases.
**Table 1** includes mutations directly associated to genetically inherited forms of PD (either of dominant or recessive inheritance) and genetic variations indirectly accounting for risk factors of disease. Gene name, locus and symbol have been described, together with the protein encoded by the gene and the function it exerts (the cell pathway governated), as well as the kind of PD associated to the genetic mutation or variation (responsible of early, late, sporadic or unknown onset). Among these entities, LRRK2 and PRKN genes emerge amongst the most frequent forms of autosomal dominant and recessive inheritedforms of PD, respectively.
### MODELS TO STUDY PARKINSON'S DISEASE
A major challenge to study PD is the inaccessibility of the target tissue of the disease (DAn), which is currently only available postmortem. In addition, by the time that clinical symptoms manifest, most of the cells targeted by PD have already been lost (Grosch et al., 2016). Thus, finding models that faithfully recapitulate the events in PD is essential to understand the impaired molecular processes that underlie the disease etiopathogenesis and its progression. In this regard, different experimental in vitro and in vivo models of study have been consistently used (Blesa et al., 2012; Falkenburger et al., 2016).
### Animal Models
Animal models have allowed the study of PD in vivo, partially reproducing the specific pathogenic events and behavioral outcomes of the disease. In fact, after the study of brain necropsies from PD patients, much of the current understanding of the etiology and the pathogenesis of PD has been obtained from the study of neurotoxin-based animal models (Bezard et al., 2013), only recently complemented by experimental approaches targeting genes responsible of the disease. Typical cases of toxic exposure to induce PD in animal models is the use rotenone and MPTP exposition, which reinforced mitochondrial implication
TABLE 1 | Classification of genes associated with familial forms of Parkinson's disease.
Modified from Del Rey et al. (2018).
in PD. However, toxin-based models rely on acute insults to the nervous system and do not model the slow neural degeneration and development of clinical manifestations characteristic of PD (Westerlund et al., 2010) or the presence of LB, hallmark of the disease, thus raising concerns on the recapitulation of PD pathology.
Later, with the identification of PD-related genes, transgenic models including yeast, Drosophila melanogaster, Caenorhabditis elegans and murine models have been developed (Blesa et al., 2012) as an alternative to the classical toxin-based ones. These models enabled to gain insights in the molecular events underlying the disease, such as mitochondrial and autophagic deregulation, widely demonstrated in the target tissue of PD. These models have shed light into PD pathogenic processes, but have fallen short in replicating the phenotype and pathology of human disease (Dawson et al., 2010). One of the main drawbacks in the use of animal models to study PD is life span difference between species, which may not allow reproducing age-related events that are relevant to disease pathogenesis. On the other hand, there is an important risk when specifically using invertebrates to study PD as relevant pathogenic factors are vertebrate-specific and may be absent in these models. Finally, most of them do not recapitulate the key clinical and neuropathological features of the disease (as trembling and neuromelanin withdrawal in DAn). Maybe this explains why biomarkers of disease previously verified in murine models and treatments that have shown positive outcomes in these models, have not later been predictive of therapeutic success in humans (Vandamme, 2014). The biological differences between mice and humans may be accountable for this fact, and it is an issue that researchers must be aware of and carefully account for when using these animal models.
In this regard, and probably due to a closer similarity, nonhuman primates (NHP) have been used to generate the most robust and clinically useful models of PD. The current gold standard animal model of PD is a toxin-based NHP induced
model, which shows stable, bilateral clinical features that closely resemble idiopathic PD (Johnston and Fox, 2014) and may even exhibit some features of RBD (Verhave et al., 2011). However, this model does neither recapitulate the major pathophysiological hallmark of idiopathic PD, LB, and its utility for the study of prodromal PD has still not been validated (Barraud et al., 2009).
The development of experimental models to elucidate disease etiology, find novel diagnostic/prognostic biomarkers and assay new therapeutic strategies remains as one of the most challenging gaps in PD research.
The use of toxic or genetic animal models of disease has strengths as reproducing the complex interplay between different neural and non-neural brain cells directly in the target tissue of the disease, and the assay of different therapeutic approaches in physiologic context. Unfortunately, in parallel, animal models fail to recapitulate important hallmarks of PD as clinical and anatomopathological features. Other weaknesses, a part of ethical concerns, high economic and facility costs, are that most of them fail to reproduce the influence of aging, epigenetic, and genetic modifying factors characteristic of PD patients. Novel cell models overcome part of these limitations.
### Cell Models
Immortalized cell lines of neural origin, either animal or human, that can also be subjected to toxin exposition or gene editing, have been widely used to model PD (Obinata, 2007). These cell models have provided consistent and reproducible results. Among their main strengths there is their identical genetic background that confers them a large homogeneity (Carter and Shieh, 2015). Moreover, they represent wide platforms for disease modeling due to low cost maintenance and editing easiness. One example is the human-derived neuroblastoma cell line SHSY5Y that can be used undifferentiated or differentiated to DAn to model PD (Lopes et al., 2010; Alvarez-Erviti et al., 2011).
However, immortalized cell lines are also associated to important weaknesses including the presence of genetic instability or the high rate of glycolytic metabolism which pushes toward the use of patient-derived cell models (Frattini et al., 2015).
The fact that the greatest proportion of PD is of unknown cause and the urgent need to find novel biomarkers at the prodromal phase of the disease has encouraged the development of specific patient-derived models. The advantages of these patient-derived cell models is that they recapitulate PD pathogenicity at an individual basis, thus partially circumventing the drawbacks of animal and established cell line models (Teves et al., 2017). In this context, the use of patient-derived cells that conserve patient-specific features has constituted a substantial progress in the study of PD, considering the great complexity and individual variability of the disease that encompasses unknown genetic and epigenetic factors, including aging, as well as environmental insults, has been recapitulated.
The use of patient-derived neural stem cell models has stirred up the field of PD research. Some works on PD have been done directly studying these stem cells (Sanberg, 2007) and some others by differentiating them into neural precursors or mature neurons (Le Grand et al., 2015; Yang et al., 2017). For instance, neural precursors as neurosphere models (free-floating clusters of neural stem cells) are widely used for the study of neuronal differentiation and neuronal disease (Matigian et al., 2010). For the study of mature-derived neurons, the development of induced pluripotent stem cells (iPSCs) has spawned a new approach to model PD allowing researchers to generate diseasespecific DAn in vitro by reprograming somatic cells from patients with the disease (Fernandez-Santiago and Ezquerra, 2016). It is expected that the access to iPSCs-derived neurons from PD patients will shed light into mechanistic insights of PD pathogenesis and serves as a platform for drug screening and early diagnosis.
One of the latest applications of iPSC is the generation of different brain cell lineages to create brain organoids that resemble neuronal architecture, self-organization and cell to cell interaction from the physiologic brain. They are threedimensional (3D) in vitro culture systems that recapitulate the developmental processes and organization of the developing human brain. These "mini-brains" provide a physiologically relevant model for the study of neurological development and disease processes that are unique to the human nervous system, together with other 2D and 3D models including neurospheres, neural aggregates, neural rosettes, and cortical spheroids. They all are emerging and promising models for the study of brain fragility and neurodegenerative diseases that will bring light into PD field in the next coming years but that are currently handicapped because of their novelty and setting up troubleshooting (Schwamborn, 2018).
Additionally, and despite being an exciting prospect for PD research, stem cell-derived neural lines and iPSC technologies have some other important inconveniences, including a considerable phenotypic variability unrelated to their genotype and their high maintenance costs and time-inefficiency (Jacobs, 2014). Additionally, stem cell or iPSC differentiation into neurons generally leads to low yields of DAn generation (Jacobs, 2014) thus obtaining heterogeneous cell pools where undifferentiated and DAn-derived cell types coexist (Fernandez-Santiago et al., 2015). Of note, whilst stem cells and iPSCs reprogramed from somatic cells highly rely on glycolytic metabolism, neurons are mainly energetically sustained by mitochondrial oxidative metabolism. In this scenario, one of the limitations of these cell models is the analysis of certain cell processes such as bioenergetics, oxidative stress or autophagy may become biased due to this confounding factor. Additionally, the genetic manipulation required to generate iPSC-derived DAn is frequently associated to genetic aberrations (e.g., copy number variations, somatic coding mutations, and chromosomal defects). Thus, the development of alternative patient-derived cell models is gaining in strength.
As previously mentioned, accumulating evidences suggest that PD is a multisystem disorder rather than a solely dopaminergic motor syndrome that encompasses central and peripheral clinical features (Djaldetti et al., 2009; Cersosimo and Benarroch, 2012). In line with this, PD pathological and molecular changes are also not confined in the CNS but are also present in the PNS and the organs that the latter innervates (Djaldetti et al., 2009). For instance, a great number of studies have reported α-synuclein
deposits in many different peripheral tissues derived from PD patients (Tolosa and Vilas, 2015). On the other hand, many other alterations at molecular level including transcriptional changes, mitochondrial dysfunction and associated increased oxidative stress as well as autophagy deregulation have been described in PD-derived peripheral tissues such as muscle (Cardellach et al., 1993), blood cells including platelets and leukocytes (Haas et al., 1995; Muftuoglu et al., 2004; Mutez et al., 2014) and fibroblasts (Mortiboys et al., 2010; González-Casacuberta et al., 2018; Juarez-Flores et al., 2018), supporting the use of novel peripheral approaches.
The use of patient skin-derived fibroblasts has been widely utilized to model numerous diseases of metabolic, neurodegenerative, and lysosomal origin (Solini et al., 2004; McNeill et al., 2014; Alvarez-Mora et al., 2017; Konrad et al., 2017). For mitochondrial diseases, fibroblasts constitute the model of choice to diagnose and often to support research of these entities (Cameron et al., 2004; Soiferman and Saada, 2015; Ferrer-Cortès et al., 2016). Several studies in PD have used human skin-derived fibroblasts to investigate the molecular mechanisms underlying disease etiopathogenesis (Auburger et al., 2012). Skin-derived fibroblasts offer considerable advantages (**Table 2**): they constitute a patient-specific cellular system that retains the genetic (mutations, polymorphisms, polygenic risk factors, etc.) and epigenetic background of the patients while potentially preserving the specific environmental, toxic and cumulative age history. They show relevant expression of most PD genes which make them also suitable for the study of monogenic forms of PD (Auburger et al., 2012; Ivanov et al., 2016). Since they are accessible peripheral cells, they can be obtained from PD patients and healthy controls through an easy and minimally invasive procedure. Furthermore, fibroblasts can be propagated in culture, frozen, and stored for long periods of time, and transformed in cell types that exhibit molecular characteristics of
TABLE 2 | Advantages and disadvantages of using skin-derived fibroblasts as a cell model to study PD.
the target tissue of the disease. It should be noted that fibroblasts make dynamic cell-to-cell contacts when cultured, which is similar to neuronal cells (Konrad et al., 2017).
Studies in skin-derived fibroblasts from patients with sPD have thrown relevant and consistent information regarding molecular pathways altered in this type of neurodegeneration. A recent study showed that fibroblasts from sPD patients have higher growth rates, altered morphology, increased mitochondrial susceptibility to UV-exerted stress and autophagic alterations (Teves et al., 2017).
Other works focused on the study of molecular alterations in monogenic-PD patient fibroblasts have reported diseaserelevant changes further supporting the adequacy of this model to study these forms of the disease. These changes include altered transcript (González-Casacuberta et al., 2018) and protein expression (Lippolis et al., 2015; Azkona et al., 2016), SNCA gene expression up-regulation (Hoepken et al., 2008), altered GCase enzyme activity (McNeill et al., 2014), microtubule destabilization (Cartelli et al., 2012), impaired autophagy (Dehay et al., 2012; Rakovic et al., 2013; Juarez-Flores et al., 2018), increased sensitivity to neurotoxins (Yakhine-Diop et al., 2014), bioenergetic deficits (Papkovskaia et al., 2012; Ambrosi et al., 2014; Juarez-Flores et al., 2018), mitochondrial alterations (Mortiboys et al., 2008, 2010), and enhanced apoptosis (Klinkenberg et al., 2010; Romani-Aumedes et al., 2014).
In summary, skin-derived fibroblasts show certain disadvantages (see **Table 2**) but constitute, at the same time, a patient-specific cellular system that, without genetic manipulation, can potentially recapitulate the main features of the disease (Auburger et al., 2012). In fact, many of the molecular hallmarks occurring in nigral DAn have been reported in fibroblasts from patients with sporadic and monogenic forms of the disease (Hoepken et al., 2008; Mortiboys et al., 2010; Ambrosi et al., 2014; Haylett et al., 2016).
### MOLECULAR MECHANISMS UNDERLYING PD PATHOGENESIS: MITOCHONDRIAL DYSFUNCTION, OXIDATIVE STRESS AND AUTOPHAGY IMPAIRMENT
Most of accumulated evidences derived from studying brains of PD patients, animal or cell models of disease stand for common molecular mechanisms underlying PD pathogenesis. Among them, neuroinflammation, apoptosis, proteasomal dysfunction, and especially mitochondrial impairment, reactive oxygen species production (ROS) and autophagic failure, emerge as key players in PD development. However, there are different schools of thought regarding the triggers of the disease. Two main hypotheses raised: according to the "protein depot cascade hypothesis," alpha-synuclein and other misfolded proteins stored as protein depots in neurons are the cause of PD. Other stand for the "mitochondrial cascade hypothesis," that foresees the origin of PD in a defect in the oxidative phosphorylation (OXPHOS) system. Interestingly, protein deposition and bioenergetics
Modified from Auburger et al. (2012).
appear to be closely related. Thus, alpha-synuclein can reduce OXPHOS function and OXPHOS deficiency can increase alphasynuclein production (Cardoso, 2011; Haelterman et al., 2014).
In line with the "mitochondrial cascade hypothesis," some primary mitochondrial diseases caused by monogenic mutations in mitochondrial-related genes that usually translate to OXPHOS deficiencies, clinically manifests as brain disease such as neurodegeneration and parkinsonism (Diez et al., 2017; Suomalainen and Battersby, 2017). In fact, mitochondria contribute to neurodegeneration through deficiencies in the mitochondrial respiratory chain (MRC) or OXPHOS and associated overproduction of ROS, as well as the accumulation of mitochondrial DNA (mtDNA) mutations or defects in its quality (deletions) or quantity (depletion) (Suomalainen and Battersby, 2017). Many evidences point out that mitochondrial dysfunction-derived oxidative stress is mainly centralized at the level of the MRC complexes I and III and play a central role in brain damage of PD patients (Poewe et al., 2017) (**Figure 1**).
The study of mitochondrial contribution to PD and other neurodegenerative diseases has been widely validated by the use of cybrids (Ghosh et al., 1999; Llobet et al., 2013). They allow the exploitation of mutation-independent, mitochondrialderived impairments (age-associated, mostly). Arduíno et al. described the generation of cytoplasmic hybrid cells (or cybrids) as a promising cellular model for the study of sPD. This approach consists on the fusion of platelets harboring mtDNA from sPD patients with cells in which the endogenous mtDNA has been depleted (Rho0 cells). This allows the comparison of different mtDNAs in the same nuclear context. The sPD cybrid model has been successful in recapitulating most of the hallmarks of sPD (including CI dysfunction, ROS generation, loss of calcium homeostasis, changes in mitochondrial morphology, increased proton leak and decreased maximal respiratory capacity, as well as protein aggregation in the form of Lewy body-like inclusions) (Swerdlow et al., 1996; Cassarino et al., 1997; Sheehan et al., 1997; Ghosh et al., 1999; Trimmer et al., 2004; Esteves et al., 2009, 2010a,b; Arduíno et al., 2012, 2013; Llobet et al., 2013), thus constituting a validated model for addressing the link between mitochondrial dysfunction and sPD pathology.
Mitochondrial function is essential in almost all cells of the organism, but specially in neurons. There are different reasons to explain their dependence on proper mitochondrial function, controlled oxidative stress production and adequate mitochondrial replacement through autophagy.
Neural metabolism is highly oxidative. Glucose is the obligatory energy substrate of the adult brain. However, under certain circumstances the brain has the capacity to use alternative blood-derived energy substrates, such as ketone bodies (during starvation or development) (Magistretti, 2006) and lactate (during periods of intense physical activity) (van Hall et al., 2009). Once inside the cell, glucose can be metabolized through glycolysis (leading to lactate production or mitochondrial metabolism) or through the pentose phosphate pathway (as glycogenesis can only be performed in astrocytes). Both mitochondrial metabolism and pentose phosphate pathway are the proper bioenergetic pathways enhanced in physiologic conditions to provide of ATP and antioxidant power to the cell, respectively. Contrarily, the metabolic activation of lactate production through the anaerobic glycolysis may be detrimental for long-term neuronal function and should be only sustained in certain punctual circumstances (Falkowska et al., 2015).
Metabolic oxidative activity of neurons explains they dependence on mitochondria to obtain energy. Noticeably, brain is one of the highest energy-demanding organs of the body due to the its intrinsic physiological activity. It is constituted by postmitotic cells (neurons) with less capacity for cellular regeneration compared to other organs. Thus, it is believed to be particularly vulnerable to mitochondrial dysfunction, ROS damaging effects and detrimental autophagic renewal of cell components, that may explain why is prone to manifest clinical evidences of mitochondrial, oxidative or autophagic alterations (Picard and McEwen, 2014). This is especially relevant in the case of DAn that synthesize the most pro-oxidant neurotransmitter of the CNS (DA), thus becoming especially vulnerable to oxidative environments, mitochondrial failure or autophagic imbalance.
In this scenario, mitochondrial dysfunction, associated oxidative stress and autophagic development becomes critical for neuronal survival. Several lines of evidence have implicated mitochondrial dysfunction as a key element in PD pathogenesis. Reduction of mitochondrial CI activity has been reported in several tissues isolated from PD patients including the CNS (Schapira and Gegg, 2011). In addition, the target genes of the mitochondrial master transcriptional regulator PGC1α have been reported to be under expressed in PD, together with strong evidences of increased oxidative stress and reduced autophagic function that may be finally responsible for defective mitochondrial and protein depot (Schapira and Gegg, 2011; Nixon, 2013; Siddiqui et al., 2015).
Unfortunately, when postmortem brain tissue of PD patients is studied, most of DAn have disappeared (Teves et al., 2017), thus difficulting the establishment of any potential etiologic causal link. In consequence, the study of experimental models of PD conveys the opportunity to foresee the triggers of neuronal degeneration.
As previously mentioned, initial models of PD were developed by using MPTP, a mitochondrial neurotoxin that specifically targeted DAn in primates and mice which was discovered to produce parkinsonism in humans (Langston, 2017). MPTP-based models and many other models using mitochondrial neurotoxins such as 6-hydroxydopamine (6-OHDA), rotenone or paraquat, have been used over the past years in the PD research field to replicate features of disease neuropathology (Jagmag et al., 2015). On the other hand, the depletion of mitochondrial proteins in mice that are essential for mtDNA maintenance (TFAM and En1) leads to neuronal degeneration of DAn in the SN which accounts for the development of several important features of PD neuropathology (Pickrell et al., 2013).
In PD patients, reduction of mitochondrial CI activity has been reported in several isolated tissues including the SN and peripheral tissues (Schapira and Gegg, 2011). In addition, the downregulation of the target genes of the master transcriptional regulator involved in mitochondrial biogenesis (peroxisome proliferator-activated receptor gamma coactivator 1-alpha or PGC-1α) have been reported in PD (Siddiqui et al., 2015). Low
levels of α-synuclein have been encountered in mitochondria in physiologic conditions and in vitro abnormal accumulation of this protein has proven to lead to mitochondrial CI dysfunction and associated oxidative stress, importantly linking these two events that have been repeatedly reported in PD (Poewe et al., 2017).
Different levels of ROS damage have been reported within the target brain region that undergoes selective neurodegeneration in PD. Specifically, lipid peroxidation markers such as 4 hydroxynonenal and malondialdehyde, have been identified in the SN of PD patients (Dias et al., 2013). However, it remains elusive whether oxidative stress occurs early in the disease or later during the decease of neurons and thus, as a consequence of cell degeneration.
Mitochondria are not isolated or static entities but instead are highly dynamic organelles that are transported on cytoskeletal proteins responsible for mitochondrial trafficking and are continuously subjected to fusion and fission processes in order to maintain their homeostasis (Johri and Beal, 2012). Mitochondrial dynamics have also been reported to be altered in PD (Van Laar and Berman, 2009). Similarly, mitochondrial turnover and quality control through selective autophagy (mitophagy), or mitochondrial relationship with other organelles as endoplasmic reticulum and its associated membranes (MAMs) have also been associated to PD (Hattori et al., 2017). Mitochondrial dysfunction and oxidative stress are associated with the impairment of the autophagy process through the accumulation of damaged mitochondria due to their defective turnover and through depletion of lysosomes, evidencing that different molecular pathways involved in PD pathogenesis are intimately related (Poewe et al., 2017).
Although our understanding of the regulatory pathways that control autophagy is still limited, recent advances have shed light on the importance of autophagy in a panoply of physiological processes and human diseases including neurodegeneration (Jing and Lim, 2012). The post-mitotic status of neurons prone them to be strictly dependent on optimal regulation of autophagy for the removal of dysfunctional or obsolete cell constituents, especially as the brain ages. In fact, the most prevalent pathological feature of many neurodegenerative diseases is the aggregation of misfolded proteins and the loss of certain neuronal populations (Guo et al., 2018), closely related to lysosomal function. This becomes more evident with the observation that the brain is often the most severely affected organ in primary lysosomal disorders and that mutations in genes involved in autophagy are causatively linked mechanisms to neurodegenerative disorders with exceptional frequency. This link between lysosomal performance and neurodegenerative diseases explains the high prevalence of genetic lysosomal variants that are found in PD by genome wide association studies (Levine and Kroemer, 2019).
On the other hand, it has been reported that in neurodegenerative disorders, such as Alzheimer's disease, amyotrophic lateral sclerosis and familial PD, defects arise at different stages of the autophagy pathway and have different implications for pathogenesis and therapy (Nixon, 2013). Recent research in the field of PD points out to alterations in specific steps of autophagic processes that could be of high relevance in the etiopathogenesis of the disease (Jing and Lim, 2012; Nixon, 2013; Guo et al., 2018) (**Figure 1**). For instance, an upregulation of macroautophagy following an overwhelmed chaperone-mediated autophagy (CMA) system as a result of overexpression of misfolded aggregates of α-synuclein has been shown in mice and in vitro PD model systems (Spencer et al., 2009; Yu et al., 2009; Ebrahimi-Fakhari et al., 2011). Moreover, an accumulation of autophagosomes (AP) and an early decrease in lysosome content as a result of lysosomal membrane destabilization and cytosolic release of cathepsins has been reported in DAn of a neurotoxin-based mouse model (Dehay et al., 2010). In addition, a decrease in lysosomal acidification and consequent decline of lysosomal protein turnover has been reported in mice overexpressing α-synuclein (Stefanis et al., 2001; Cuervo et al., 2004).
In post mortem brain samples of PD patients dysfunctional lysosomes and accumulation of AP were observed in neurons, indicating a pathogenic role of autophagy in PD (Guo et al., 2018).
#### Mitochondrial Dysfunction, Oxidative Stress, and Autophagy Disruption in Familial PD
Amongst all familial forms of PD, late and early onset PD associated to mutations in LRRK2 and PRKN genes, respectively, are responsible for the most frequent dominant and recessive inherited forms of PD. These genes have emerged as promising examples of disease due to their established role in commanding bioenergetic and autophagic balance (**Figure 2**).
Numerous genes responsible of inherited PD cause the impairment of essential functions for mitochondrial homeostasis: OXPHOS function, mitochondrial trafficking, oxidative stress, calcium imbalance, mitochondrial biogenesis, mitochondria dynamics and mitochondrial autophagy.
#### LRRK2-Associated PD (LRRK2-PD)
The LRRK2 gene (at the PARK8 locus) is located in chromosome 12, contains 51 exons and spans a genomic distance of 144 kb that includes 7500 nucleotides of coding sequence. Up to date, about 80 "probably pathogenic" and seven pathogenic LRRK2 mutations have been described, being the G2019S (G2019S-LRRK2) the most frequent pathogenic mutation (Corti et al., 2011). Most of LRRK2 mutations correspond to missense variants, which, along with the dominant inheritance, are consistent with a gain-of-function pathogenic mechanism (Corti et al., 2011). A number of non-pathologic variants are also known and some others variants that may act as PD risk factors have been reported using genome wide association studies (Islam and Moore, 2017).
LRRK2 encodes a 2527 amino-acid multi-domain protein (LRRK2), which is also known as dardarin, from the Basque word "dardara" which means trembling. LRRK2 has the particular feature of encoding a leucine-rich repeat (LRR), a ROC-COR GTPase, a mitogen-activated protein kinase, and WD40 domains in the same protein (Singh et al., 2019) (**Figure 3**). Of note,
pathogenic mutations appear to be located in functionally relevant regions of the protein such as in the specific case of the G2019S-LRRK2 mutations that affect the kinase domain making it more active (Corti et al., 2011) (**Figure 3**). LRRK2 is expressed in most organs including brain, heart, liver, and circulating immune cells.
Mutations in LRRK2 account for ∼10% of familial PD and for a significant fraction of sPD cases (Kalinderi et al., 2007). LRRK2 mutation frequencies vary between ethnic groups, being North African Arabs and Ashkenazi Jews the most affected populations (Correia Guedes et al., 2010). The G2019S-LRRK2 mutation is responsible for 1% of apparent sPD and 4% of familial PD worldwide (Healy et al., 2008). Importantly, even when proven pathogenic mutations are present, penetrance is age dependent and estimated between 30 and 74% (Ozelius et al., 2006). This fact makes the LRRK2-mutation carriers a very interesting target of study, as they present a subclinical stage with molecular alterations potentially determinant for disease progression.
Clinically, LRRK2-associated PD (LRRK2-PD) presents with a PD-typical phenotype with no sex association. A large systematic review of LRRK2-PD case reports that age of onset is around 57 years, with a mean disease duration of 10 years. The cardinal PD symptoms were reported with the following frequency: bradykinesia in 99%, rigidity in 99%, tremor in 88% and postural instability in 65%, while atypical signs of PD have been reported only anecdotally. In addition, autopsies of such patients showed prominent loss of melanized DAn in the SNpc (Trinh et al., 2018). Disease progression is slow and response to treatment is as good as in sPD (Corti et al., 2011). LRRK2-PD has demonstrated an unprecedentedly significant role of LRRK2 in PD pathogenesis as most of the clinical and pathological features are indistinguishable from those of sPD (Gosal et al., 2005). Thus, since its discovery, great efforts have been focused on the study of this form of the disease.
Within cells, LRRK2 associates with various intracellular membranes and vesicular structures including the endosomes, the lysosomes, the multivesicular bodies, the outer mitochondrial membrane (OMM), lipid rafts, microtubule associates vesicles, Golgi complex, and the endoplasmic reticulum (Cookson, 2012), thus it is highly associated to MAMs. Accordingly to its multi-domain nature, LRRK2 has been implied in many cellular functions such as cytoskeleton remodeling, vesicle trafficking and movement, protein translation, autophagy and mitochondrial function homeostasis (Cookson, 2010; Liu et al., 2012; Taymans et al., 2015; Roosen and Cookson, 2016; Juarez-Flores et al., 2018; Price et al., 2018). Specifically, studies consisting of modifications in the expression of LRRK2 in different neuralcell based models have reported an altered synaptic vesicle
trafficking and endocytosis (Shin et al., 2008; Piccoli et al., 2011). LRRK2 dysfunction has also been demonstrated to play a role in different pathologic scenarios, such as ∝-synuclein phosphorylation, microtubule dynamics, alterations in the uncoupling protein system (UPS) as well as in neurite growth and branching regulation that may trigger neurodegeneration (Dächsel et al., 2010). Silencing of LRRK2 reduced the inflammatory response in different human cell-derived and animal models (Lopez de Maturana et al., 2014).
A growing body of evidence supports a role for LRRK2 in mitochondrial dynamics and function. In fact, it interacts with a number of crucial proteins that regulate mitochondrial dynamics such as dynamin-related protein1 (DRP1), Mitofusin (MFN1) and 2 and optic atrophy1 (OPA1) (Wang et al., 2012; Stafa et al., 2014; Park et al., 2018). Thus, LRRK2 might directly affect mitochondrial homeostasis while indirectly regulating it through autophagy and cytoskeletal dynamics (Singh et al., 2019). This hypothesis is also supported by many studies reporting mitochondrial dysfunction in various animal models of G2019S-LRRK2 PD, in postmortem human tissues from LRRK2-PD patients (Mortiboys et al., 2010; Cooper et al., 2012; Sanders et al., 2014; Yue et al., 2015) and in different patient-derived cell models (Niu et al., 2012; Wang et al., 2012; Cherra et al., 2013; Su and Qi, 2013). Such studies reported mtDNA damage, decreased mitochondrial membrane potential (MMP) and ATP production, as well as altered mitochondrial dynamics and mitophagy (Singh et al., 2019). In addition, a protective role against oxidative stress has been reported for wild-type LRRK2, which seems to be lost in mutant forms of the protein (Liou et al., 2008).
#### Studies in Fibroblasts From LRRK2-Associated PD
Mitochondrial phenotypes have been characterized in LRRK2 fibroblasts at baseline and under conditions of pharmacological stress. The most common pharmacological approaches used to date include mitochondrial toxins such as MPTP, valinomycin, oligomycin, CCCP, and rotenone. As these approaches may mimic mitochondrial toxicity, they are far from dissecting the mitochondrial pathways affected under physiological conditions (Trentadue et al., 2012). Smith et al. (2015) demonstrated an increased sensitivity to valinomycin in a subset, but not all, of fibroblasts derived from PD patients, pinpointing again the great interindividual variability of the disease and, outstandingly, that the molecular characteristics of patientderived cell models do not always correlate with the clinical presentation of the disease.
Mortiboys et al. (2010) reported for the first time mitochondrial alterations in fibroblasts of human G2019S-LRRK2 mutation carriers, consisting of reduced MMP, reduced intracellular ATP levels, mitochondrial elongation and increased mitochondrial interconnectivity. These findings were further confirmed by Papkovskaia et al. (2012), who described decreased MMP and ATP levels as well as increased proton leakage and ROS levels with the associated increase in uncoupling protein 2 (UCP2) in fibroblasts from G2019S-LRRK2 PD patients. Several studies have repeatedly observed alterations in mitochondrial dynamics such as increased mitochondrial fragmentation in fibroblasts from LRRK2-PD patients (Su and Qi, 2013; Grünewald et al., 2014; Smith et al., 2015; Falkenburger et al., 2016).
A recent study compared mitochondrial function and autophagy in fibroblasts of G2019S-LRRK2-mutation carriers without clinical symptoms (so called non-manifesting carriers or NMC), with patients harboring G2019S-LRRK2-mutation and clinical manifested PD. Interestingly, fibroblasts of NMC showed an enhanced mitochondrial performance upon forcing mitochondrial oxidative metabolism with galactose and upregulation of autophagy (Juarez-Flores et al., 2018). These findings suggested that the exhaustion of the bioenergetic and autophagy reserve might contribute to the onset of clinical PD symptoms. Other authors have reported heightened autophagic flux and higher expression of autophagy markers as well as an increased mitophagy in G2019S-LRRK2-mutation carriers with clinical diagnosed PD (Smith et al., 2015; Su et al., 2015). The reduction in mitophagy and increased ROS production has been associated to defective histone acetyltransferase and deacetylase activities contributing to cell death also in LRRK2-fibroblasts (Yakhine-Diop et al., 2019). Additionally, the novel role of key regulators of autophagy (as TMEM230) interacting with Rab proteins as Raba or Rab32been described in LRRK2-patients fibroblasts is emerging as a promising new target in disease (Waschbüsch et al., 2014; Kim et al., 2017).
#### PRKN-Associated PD (PRKN-PD)
Other forms of PD have also been genetically associated to mitochondrial and autophagic imbalance, in this case through a recessive inheritance. This is the case of PRKN.
The locus of PRKN is mapped to the telomeric region of the long-arm of chromosome 6. More than 170 different mutations have been identified throughout the sequence of this particularly large gene (1.35 Mb) ranging from point and missense mutations to large deletions or multiplications and small deletions/insertions (Bruggemann and Klein, 1993; Klein and Westenberger, 2012) (**Figure 4**). Rare deletions extending in the neighboring PRKN coregulated gene (PACRG) result in the same early onset parkinsonism phenotype (Corti et al., 2011).
PRKN is a 465 amino acid protein that contains an NH2 terminal homologous to a ubiquitin-like domain (UBL) followed by three really interesting new gene (RING) finger domains (RING 0–2) separated by an In-Between-RING (IBR) domain in the COOH-terminal part, each of which bind two Zn2<sup>+</sup> (Zhang et al., 2015) (**Figure 4**). Functionally, PRKN is a member of a family of E3 ubiquitin protein-ligases responsible for the labeling of selected cargos, such as obsolete proteins and organelles, which need to be degraded through the ubiquitination process. This process comprises the transfer of activated ubiquitin molecules to the lysine residues of specific substrate proteins. Depending on the site and type of ubiquitination (mono, poly or multiubiquitination), certain cell signaling processes are activated, including proteosomal degradation but also non-degradative signaling roles (Dawson and Dawson, 2010).
Along with the original discovery of the PRKN function as an E3 ubiquitin ligase in PD-associated PRKN mutations, the hypothesis that loss of PRKN function would lead to the toxic
accumulation of one or several of its substrates raised. To date, no less than 25 PRKN putative substrates have been reported and new substrates continue to emerge periodically, especially those related to mitochondria (Zhang et al., 2015). In addition, many dynamically regulated ubiquitination sites in dozens of proteins have been identified, with strong enrichment for OMM proteins, indicating that PRKN dramatically alters the ubiquitination status of the mitochondrial proteome (Sarraf et al., 2013).
Nigral cell loss in PRKN-PD patients appears to be caused by a loss of function of the protein due to biallelic homozygous or compound heterozygous mutations in the PRKN gene. However, there is an ongoing debate with regard to whether heterozygous PRKN mutations may confer increased susceptibility to PD as heterozygous PRKN pathogenic variants have been detected in a large number of individuals with PD (Bruggemann and Klein, 1993; Mortiboys et al., 2008).
Although the population-based prevalence of PRKN-PD is largely unknown (Bruggemann and Klein, 1993), it is thought that PRKN mutations account for up to 50% of recessive familial forms and 80% in those patients with a PD onset before the age of 20 years (Corti et al., 2011). Women and men are equally affected, with an age at onset usually <40– 50 years (Mizuno et al., 2001) although some individuals may not develop PD until age 60 or 70 years (Klein et al., 2000; Lohmann et al., 2003). In addition to an earlier age at onset, PRKN-PD patients show a clinical phenotype similar to that of sPD being bradykinesia and tremor amongst the most common signs, but also a number of specific clinical features. PRKN-PD is also characterized by a relatively benign course with slow progression, remarkable and maintained response to low levodopa doses but with frequent severe treatment-related motor complications such as early motor fluctuations and the development of dyskinesias (Cheon et al., 2012). Pyramidal signs, cerebellar features, and psychiatric disorders have been reported, but dementia or dysautonomia seem to be rare (Corti et al., 2011; Johansen et al., 2018).
In the limited neurophathologic studies, PRKN mutations are associated with selective DAn loss in the SN and some cases reported a moderate decrease of noradrenergic neurons in the Locus coeruleus with gliosis and without LB. However, a few cases of LP have been reported in PRKN-PD, especially those associated to a later onset of the disease (Bruggemann and Klein, 1993; Johansen et al., 2018).
One of the best characterized functions of PRKN is its role in the process of mitophagy, which is the selective targeting of a damaged mitochondrion for autophagy. Compelling evidence suggests that PRKN acts together with and downstream of PINK1 in a common mitochondrial quality control pathway responsible for the detection and clearance of damaged mitochondria through mitophagy (Eiyama and Okamoto, 2015). In healthy mitochondria, PINK1 is constitutively imported into the OMM and inner mitochondrial membranes (IMM), cleaved by several proteases and subsequently degraded. Loss of MMP impedes the import of PINK1 in the IMM, thereby stabilizing PINK1 on the OMM and consequently recruiting PRKN from the cytosol. In its native state, PRKN is auto-inhibited by its N-terminal UBL domain, which blocks the binding site for any incoming E2 ubiquitin conjugate, required for PRKN ubiquitination activity. Upon mitochondrial depolarization, PINK1 phosphorylates cytoplasmic PRKN in its UBL domain, relieving PRKN autoinhibition (Eiyama and Okamoto, 2015). Activated PRKN ubiquitinates many OMM proteins including VDAC1, mitofusins and the translocase of the OMM 20 (TOMM20) (Kondapalli et al., 2012; Ivankovic et al., 2016). Together, PINK1 and phosphorylated PRKN extensively modify the OMM with phosphorylated ubiquitin (pUb) chains. pUb chains serve as a mitochondrial receptor for further allosteric activation and recruitment of PRKN to the OMM, resulting in a self-amplifying feed-forward loop. Ubiquitination of these substrates primes mitochondria for recruitment to phagophores that then mature to AP and fuse with lysosomes resulting in the degradation of dysfunctional mitochondria (Ichimura et al., 2013). In addition, recent evidence suggests that PRKN is also involved in the aggresome-autophagy pathway in which PRKN promotes the sequestration of misfolded proteins into aggresomes and its subsequent clearance by autophagy (Olzmann and Chin, 2008; Yung et al., 2016).
On the other hand, PRKN has been implicated in mitochondrial biogenesis specifically, through the regulation by ubiquitination of the protein levels of one of its substrates named PARIS (ZNF746) (Shin et al., 2011). PARIS represses the expression of the transcriptional coactivator PGC-1α, which is considered a master regulator of mitochondrial biogenesis. In this line, PARIS has been reported to accumulate in models of PRKN inactivation and in human PD brain (Shin et al., 2011). Thus, PRKN potentially acts as an intermediary between
mitochondrial biogenesis and autophagy, by both blocking mitochondrial biogenesis and mitochondrial turn-over, thus resulting in mitochondrial aging.
#### Studies in Fibroblasts From PRKN-PD Patients
Amongst all the studies using skin-derived fibroblasts as a cell model for PRKN-PD, it is worth stressing that many of them have focused on studying mitochondrial function leading to controversial outcomes.
Alterations in the enzymatic activities of the MRC have been previously reported in PRKN-PD fibroblasts (Mortiboys et al., 2008; Grunewald et al., 2010; Pacelli et al., 2011). Mortiboys et al. (2008) and Pacelli et al. (2011) described CI enzymatic deficiency in PRKN-PD fibroblasts, while Grunewald et al. (2010) observed preserved enzymatic activities in isolated mitochondria from a larger cohort (Mortiboys et al., 2008; Grunewald et al., 2010; Pacelli et al., 2011). Mitochondrial complex IV deficiency has only been described by Pacelli et al. (2011) in two PRKN-PD fibroblasts lines while others have reported unaltered enzymatic activity of this complex (Mortiboys et al., 2008; Grunewald et al., 2010). Mitochondrial respiration is frequently measured to assess how MRC enzymatic activities translate to global mitochondrial function. Haylett et al. (2016) and Zanellati et al. (2015) consistently observed increased basal mitochondrial respiration in PRKN-PD fibroblasts but reported opposite outcomes in ATPlinked respiration. In contrast to these findings, a previous study described an overall decrease in all respiratory parameters of PRKN-PD fibroblasts (Pacelli et al., 2011).
Mitochondrial membrane potential has also been widely explored as a general marker of mitochondrial integrity in PRKN-PD fibroblasts. While two authors were not able to demonstrate alterations in this parameter (Grunewald et al., 2010; Haylett et al., 2016), others reported decreased MMP (Zanellati et al., 2015; Koentjoro et al., 2017), especially when exposing cells to mitochondrial-challenging conditions (Mortiboys et al., 2008; Grunewald et al., 2010).
Many evidences point to the involvement of PRKN in the entire process of mitochondrial dynamics, including organelle biogenesis, fusion/fission, and mitochondrial clearance via mitophagy (Lim et al., 2012). In this context, mitochondrial network morphology has been of interest in PRKN-PD fibroblasts studies but with controversial results (Mortiboys et al., 2008; Grunewald et al., 2010; Pacelli et al., 2011; van der Merwe et al., 2014; Zanellati et al., 2015; Haylett et al., 2016). Although some studies seem to agree that mitochondrial length is conserved in these cells (Mortiboys et al., 2008; van der Merwe et al., 2014; Haylett et al., 2016), others have observed a fragmented mitochondrial network (Pacelli et al., 2011; Zanellati et al., 2015). While most of studies did not show significant alterations in mitochondrial branching (Grunewald et al., 2010; van der Merwe et al., 2014; Zanellati et al., 2015), Haylett et al. (2016) observed decreased levels and Mortiboys et al. (2008) reported increased rates. In line with this, mitochondrial content has been assessed in several works but only Grunewald et al. (2010) reported a significant increase in this feature in PRKN-PD fibroblasts whereas others have observed conserved (Mortiboys et al., 2008; van der Merwe et al., 2014) or decreased levels (Pacelli et al., 2011).
As previously discussed, oxidative stress is a hallmark of mitochondrial dysfunction that has often been related with neurodegeneration, specifically in PD (Guzman et al., 2010; Surmeier et al., 2011; Poewe et al., 2017). In accordance, previous authors demonstrated increased protein and lipid oxidation in small cohorts (Grunewald et al., 2010; Pacelli et al., 2011).
Surprisingly, to our knowledge, studies assessing mitophagy or autophagy in PRKN-PD fibroblasts are scarce. Only recently, Koentjoro et al. (2017) elegantly demonstrated that a PRKN-PD patient fibroblast cell line failed in initiating mitophagy upon induction of mitochondrial depolarization. Interestingly, they also examined an unusual homozygous PRKN mutation carrier who did not develop clinical PD by her eight decade and found preserved mitochondrial function due to the induction of a PINK1/Parkin-independent mitophagy mediated by Nix, which is a selective autophagic receptor located on the OMM (Koentjoro et al., 2017).
Other studies in PRKN-PD fibroblasts have reported alterations in alternative important cell processes which represent promising targets of disease pathogenesis to be further explored. For instance the regulation of endoplasmic reticulum-to-mitochondrial contacts by Parkin via Mfn-2 (Basso et al., 2018). Also, Pacelli et al. (2019) reported altered severe damping of the bioenergetic oscillatory patterns associated to circadian rhythms and molecular clockworks in fibroblasts from PRKN-PD patients that may conditioning mitochondrial quality control and mitophagy. One study performing a whole-genome expression analysis by RNA-sequencing found that different PRKN mutations were associated with a large number of gene expression changes at the transcriptome level (González-Casacuberta et al., 2018). Specifically, authors reported the upregulation of 1C-dependent anabolic biosynthetic pathways, which has been related with the activation of the mitochondrial integrated stress response (ISRmt) in front of mitochondrial dysfunction (Bao et al., 2016; Celardo et al., 2017; Suomalainen and Battersby, 2017). Additional studies in PRKN-PD fibroblasts have reported alterations in the protein expression and lipidome profiles (Lippolis et al., 2015; Lobasso et al., 2017) as well as cytoskeleton alterations such as microtubule destabilization (Cartelli et al., 2012; Vergara et al., 2015).
The characterization of fibroblasts of PD patients point out disrupted pathways to be targeted and therapeutic platforms, but some concerns and controversies arise. The low reproducibility of mitochondrial function analysis presented in most studies performed up to date could be attributed to the small sample sizes tested. Also differences in methodological approaches, protocols and experimental conditions (e.g., site of skin biopsy, passage number of cells, etc.) may partially account for the large variation obtained. For instance, the use of different high-resolution respirometry approaches in which oxygen consumption is measured from seeding fibroblasts or from cells in suspension. Similarly, assessing MRC enzymatic activities in intact cells or in mitochondrial enriched fractions may contribute to outcome disparities. Moreover, all the studies were performed in glycolytic conditions that may partially unveil mitochondrial deficits and contribute to controversy (Mortiboys et al., 2008; Grunewald et al., 2010; Pacelli et al., 2011; van der Merwe et al., 2014; Zanellati et al., 2015; Haylett et al., 2016). In this sense, the
use of alternative sources of energy, as galactose, serves for two purposes: to force and challenge oxidative metabolism (usage widely extended for the diagnosis of mitochondrial disorders) and mimic neuronal metabolism (mainly based in OXPHOS function). Novel studies focused in exploring mitochondrial or autophagic function in galactose may be useful to unveil pathogenic mechanisms of disease.
In summary, the particular area of research focused on the study of mitochondrial function in PRKN-PD fibroblasts has proved to be contentious, with several groups either describing similar defects or no apparent abnormalities. It would be of great importance that researchers join efforts on homogenization of protocols and analyzing a more significant number of PRKN-PD patient-derived cells in order to unveil if mitochondrial and autophagic dysfunction is a crucial event in PRKN-PD pathogenesis.
#### DISEASE MODIFYING THERAPIES
Since the disseverment of PD, different therapeutic options have been developed to ameliorate the symptoms of PD. The first one to be developed was levodopa, in 1960, a precursor that is transformed to DA in the brain, supplying the amount of DA that degenerated neurons are not able to produce. Other medications include DA agonists and monoamine oxidase-B (MAO-B) or Catechol-O-methyltransferase inhibitors (COMT) inhibitors, selegiline, and rasagiline, that decrease the activity of MAO-B, enzyme responsible of degrading DA. There are also other options to severe cases who do not respond to DA based on the use of apomorphine and duodopa administered by pumps, surgical interventions or deep brain stimulation (Rizek et al., 2016).
However, these treatments are supportive, only control the symptoms of the disease, the neurodegeneration is not stopped or reversed, consequently, there are no curative treatments for PD.
To unveil the pathophysiology of the diseases and develop new therapeutic strategies to reduce the impact of PD and find a cure is essential to develop novel models of disease.
The experimental models herein discussed hold potential for the development of PD modifying therapies. The complementary assay of any potential candidate in different experimental models confers strength to the potential therapeutic efficacy before translation into the clinical settings. Patient-derived cell models offer usefulness either as platforms for testing novel therapeutic approaches or for prompting the discovery of novel targets from disrupted pathways reported in these models.
The use of these experimental models in PD has permitted the discovery of different therapeutic candidates, with different degree of evidence on their potential therapeutic activity and security concerns. This is the case of FGF20, echinacoside, rosmarinic acid or autophagic modulators, as Threalose or Torin 1, among many others, targeting different disrupted cell pathways in disease (Liang et al., 2019; Lv et al., 2019; Wang et al., 2019). Depending on the subtype of PD, specific treatments have been proposed. This is the case of LRRK2-PD carriers, where the use of LRRK2 inhibitors has been proposed. They are currently being tested in clinical I trials. Unfortunately, their systemic action may unveil secondary effects, somehow bypassed by the targeting of specific neural effectors (as PAK6 or Rab GTPases) to modulate neural disrupted protein and organelle trafficking in PD (Kiral et al., 2018). Similarly, for PRKN-patients, selective mitochondrial drugs have been proposed. Experimental data supports the use of fusion or fission inhibitors (as MDIVI-1), that still rank in experimental settings (Manczak et al., 2019). Antioxidant and mitochondrial principles (as coenzyme Q, that failed in a phase III assay) and peroxisome proliferator-activated receptor-γ agonists that reduces proinflammatory cytokines and modulate mitochondrial biogenesis (as Pioglitazone) were also tested in clinical trials, but failed to demonstrate further efficacy.
Apart from symptomatic treatments (such as levodopa or surgical interventions), disease modification through neuroprotection remains as the main milestone in PD research. Neuroprotection tested by pramipexole (CALM-PD), ropinirole (REAL-PD), and pramipexol (PROUD-PD) failed to establish disease modification (Bartus et al., 2013; Obeso et al., 2017). In this sense, calcium channel blockers aimed to prevent calcium influx on nigral neurons are being tested (isradipine is being evaluated in phase I and II clinical trials), together with compounds able to increase urate antioxidant protection (inosine is undergoing phase II studies) (Obeso et al., 2017).
Additionally, aiming to support neuroprotection through the enhancement of neuronal viability, trophic factors are also being evaluated in PD, showing moderate or null therapeutic success. Of them, glial family ligands as glial derived neurotrophic factor (GDNF) and neurturin in preclinical studies demonstrated strong neuroprotection in multiple animal models. However, multiple clinical trials, including 2 phase II trials, failed to demonstrate their efficacy or showed significant side effects (Kordower et al., 2000; Obeso et al., 2017).
Alpha-synuclein has become lately the major target for PD therapeutics. Initial preclinical efforts concentrated on synuclein-lowering treatments such as siRNAs directed against alpha-synuclein, that resulted toxic in animals. Novel attempts focused to disaggregate aggregated synuclein, facilitate its clearance by augmenting autophagy pathways, or using antibodies to prevent its propagation from the periphery to the brain and once in the brain across the neural axis. Vaccines against alpha-synuclein as both active and passive immunization approaches have been attempted. Active immunotherapy attempts to stimulate the immune system against specific antigens (Bergström et al., 2016). Passive immunotherapy uses monoclonal antibodies against alphasynuclein molecule. Initial phase 1 safety trials are currently underway and show promising results in terms of safety and tolerability profiles. The enhancement of glucocerebrosidase (GBA) lysosomal activity to reduce alpha-synuclein levels is also being tested through small molecule chaperones in clinical trials (McNeill et al., 2014).
Probably the next coming years will open future perspectives for the development of new supportive and curative therapies in PD, where personalized medicine, mainly based on genetic and molecular counseling, will help to direct specific PD patients to a wide panoply of therapies.
The development of novel therapeutic options will depend of the efficacy of candidate compounds previously tested in preclinical settings and experimental models of disease, as the herein described, and the target of disrupted pathways, as the herein explained.
#### CONCLUSION
Parkinson's disease encompasses a wide panoply of genetic and molecular etiologies leading to common clinical manifestations. Different schools of thought differ in considering either mitochondria or protein deposition-cascade as the triggers of PD, but all they convey that PD pathogenesis is associated to the deregulation of both mitochondrial and autophagic clearance pathways, supporting its role in the disease. Mitochondrial dysfunction and their turnover through autophagy directly targets some types of PD (as those carrying mitochondrial or autophagic mutations) but also stand at the base of the rest of PD by providing the overdose of energy needed to support alternative deregulated pathways while maintaining oxidative stress levels within control ranges. Thus, proper mitochondrial and autophagic function protects against PD and exhaustion of mitochondrial and autophagy contributes to PD development, independently of the genetic base. These molecular alterations have been consistently reported in skin-derived fibroblasts from PD patients carrying mutations in LRRK2 and PRKN genes. These findings demonstrate the presence of molecular damage characteristic of the PD target tissue beyond the CNS and the usefulness of these patient-derived cells to model PD, models that can be metabolically upgraded to resemble neuron behavior and challenge mitochondrial and autophagic function by the use of galactose.
Current research gaps in PD research stand for the development of novel therapeutic candidates aimed to promote healthy brain aging and avoid or even cure PD, probably
#### REFERENCES
based in personalized-medicine guided by genetic and molecular counseling. New generation sequencing will increase the number of genes responsible of familial PD and the number of genetic risk factors accounting for sporadic PD, thus unveiling molecular imbalances underlying PD. Novel compounds against these targets will be discovered in experimental settings and disease models to set the path for further clinical trial testing. Complementary models of disease will be needed to dissect the disrupted pathways in PD and design specific therapeutic targets, but the use of patient-derived cells such as fibroblasts is gaining in strength because they constitute platforms to model disease etiopathogenesis and try new therapeutic approaches in the genetic and epigenetic background of the patient.
New challenges and potential developments in the field of PD entail the validation of these novel therapeutic candidates focused on modifying the course of PD through, among others, promoting mitochondrial and autophagic performance.
#### AUTHOR CONTRIBUTIONS
All authors listed have made a substantial, direct and intellectual contribution to the work, and approved it for publication.
### FUNDING
This work was supported by the Fondo de Investigación Sanitaria (FIS PI11/01199, PI18/00451, and PI18/00498) and the CIBERER (an initiative of ISCIII) granted by the Instituto de Salud Carlos III and co-funded by the Fondo Europeo de Desarrollo Regional de la Unión Europea "Una manera de hacer Europa", Suports a Grups de Recerca (2017/SGR) and the CERCA Program from the Generalitat de Catalunya, CONACyt, Fundació La Marató de TV3 [87/C/2015], and Fundació Cellex.
of glycogen metabolism. Int. J. Mol. Sci. 16, 25959–25981. doi: 10.3390/ ijms161125939
LRRK2, VPS35: MDSGene systematic review. Mov. Disord. 33, 1857–1870. doi: 10.1002/mds.27527
neurotoxin 1-methyl-4-phenylpyridinium dependent of autophagy. Toxicology 324, 1–9. doi: 10.1016/j.tox.2014.07.001
**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Copyright © 2019 González-Casacuberta, Juárez-Flores, Morén and Garrabou. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
# Intracellular and Intercellular Mitochondrial Dynamics in Parkinson's Disease
Dario Valdinocci<sup>1</sup> , Rui F. Simões<sup>2</sup> , Jaromira Kovarova<sup>3</sup> , Teresa Cunha-Oliveira<sup>2</sup> , Jiri Neuzil1,3 and Dean L. Pountney<sup>1</sup> \*
<sup>1</sup> School of Medical Science, Griffith University, Southport, QLD, Australia, <sup>2</sup> CNC - Center for Neuroscience and Cell Biology, University of Coimbra, Cantanhede, Portugal, <sup>3</sup> Institute of Biotechnology, Czech Academy of Sciences, Prague-West, Czechia
The appearance of alpha-synuclein-positive inclusion bodies (Lewy bodies) and the
loss of catecholaminergic neurons are the primary pathological hallmarks of Parkinson's disease (PD). However, the dysfunction of mitochondria has long been recognized as a key component in the progression of the disease. Dysfunctional mitochondria can in turn lead to dysregulation of calcium homeostasis and, especially in dopaminergic neurons, raised mean intracellular calcium concentration. As calcium binding to alphasynuclein is one of the important triggers of alpha-synuclein aggregation, mitochondrial dysfunction will promote inclusion body formation and disease progression. Increased reactive oxygen species (ROS) resulting from inefficiencies in the electron transport chain also contribute to the formation of alpha-synuclein aggregates and neuronal loss. Recent studies have also highlighted defects in mitochondrial clearance that lead to the accumulation of depolarized mitochondria. Transaxonal and intracytoplasmic translocation of mitochondria along the microtubule cytoskeleton may also be affected in diseased neurons. Furthermore, nanotube-mediated intercellular transfer of mitochondria has recently been reported between different cell types and may have relevance to the spread of PD pathology between adjacent brain regions. In the current review, the contributions of both intracellular and intercellular mitochondrial dynamics to the etiology of PD will be discussed.
#### Edited by:
Sandeep Kumar Barodia, The University of Alabama at Birmingham, United States
#### Reviewed by:
Karina Ckless, State University of New York Plattsburgh, United States Dilshan Shanaka Harischandra, Covance, United States Tatiana Rosado Rosenstock, Faculty of Medical Sciences of Santa Casa de São Paulo, Brazil
> \*Correspondence: Dean L. Pountney [email protected]
#### Specialty section:
This article was submitted to Neurodegeneration, a section of the journal Frontiers in Neuroscience
Received: 01 March 2019 Accepted: 19 August 2019 Published: 18 September 2019
#### Citation:
Valdinocci D, Simões RF, Kovarova J, Cunha-Oliveira T, Neuzil J and Pountney DL (2019) Intracellular and Intercellular Mitochondrial Dynamics in Parkinson's Disease. Front. Neurosci. 13:930. doi: 10.3389/fnins.2019.00930 Keywords: alpha-synuclein, tunneling nanotube, Parkinson's, mitophagy, mitochondria
## INTRODUCTION: PD, α-SYNUCLEIN AND MITOCHONDRIA
The principal histopathological marker of Parkinson's disease (PD) is the presence in neurons of α-synuclein (α-syn) protein aggregates that occur in inclusion bodies known as Lewy bodies (Shults, 2006; McCann et al., 2014). α-Syn is primarily expressed pre-synaptically and evidence exists of α-syn transfer from neurons to neuronal and non-neuronal cells in vitro, indicating that α-syn pathology spreads between anatomically adjacent brain regions by a cell-to-cell transfer mechanism (Valdinocci et al., 2017). α-Syn is a small (14 kDa), acidic protein expressed in the brain, peripheral nervous system and circulating erythrocytes (Thakur et al., 2019). Its pre-synaptic localization and high abundance implicate an important role in synaptic transmission (Burre et al., 2010) with specific functions implicated in synaptic vesicle recycling and regulating soluble NSF
attachment protein receptor (SNARE) interactions and dopamine biosynthesis (Theillet et al., 2016; Sulzer and Edwards, 2019). In vitro α-syn is a dynamically unfolded protein, although in vivo the membrane-associated tetrameric form is proposed to be α-helical (Bartels et al., 2011). Various factors, such as raised copper or calcium concentration, oxidative stress and post-translational modifications can trigger intracellular α-syn aggregation (Fujiwara et al., 2002; El-Agnaf et al., 2006; Reynolds et al., 2008; Rcom-H'cheo-Gauthier et al., 2016). Whilst a definitive link between mitochondrial dysfunction and initiation of PD still does not exist, it is clear that dysfunctional mitochondria are omnipresent in PD (Chen et al., 2019). Moreover, α-Syn can be located at mitochondrial membranes, especially under stress conditions (Cole et al., 2008; Devi et al., 2008), and its aggregation can be linked to mitochondrial dysfunction in PD (Perfeito et al., 2013; Celardo et al., 2014). α-Syn aggregates may in turn cause deleterious alterations in mitochondrial function, including intracellular dynamics. This review focusses on α-syn interactions with mitochondria in PD.
#### INTRACELLULAR MITOCHONDRIAL DYNAMICS IN PD
Mitochondria are highly dynamic and interconnected entities, with the ability to change their morphology in addition to mobilizing within the cell to help power critical functions (Burte et al., 2015). Mitochondrial dynamic processes include fusion/fission, transport and clearance, and are so interconnected and interdependent that they have been proposed to form an interactome that ultimately controls mitochondrial quality, quantity and metabolism (Dorn and Kitsis, 2015; Shirihai et al., 2015). Under normal conditions, mitochondria constantly undergo cycles of fusion and fission that affect their morphology and shape, processes that may be perturbed in PD. α-Syn has been shown to influence mitochondrial size both independently and dependent on fusion/fission proteins, with recent reports detailing these interactions as an attribute of the pathological variants such as oligomers and fibrils, ruling out any negative effects on mitochondrial dynamics as a normal function of the monomer (Wang et al., 2019). Factors such as the GTPases Mitofusin1 and 2 (Mfn1/2), and optic atrophy protein 1 (OPA1) are involved in fusing the outer mitochondrial membrane (OMM) and inner mitochondrial membrane (IMM), respectively, forming elongated structures that are more efficient in ATP generation (Chan, 2012; Tilokani et al., 2018). As illustrated in **Figure 1**, oligomeric α-syn can bind to lipids in the OMM and distress the membrane curvature, leading to a decrease in mitochondrial fusion rate (Pozo Devoto and Falzone, 2017). In addition, overexpression of α-Syn in transgenic mice led to reductions in Mfn1/2 protein levels, correlating with a decrease in mitochondrial fusion and smaller mitochondria (Xie and Chung, 2012). α-Syn knockdown was shown to trigger mitochondrial elongation (Kamp et al., 2010).
Conversely, mitochondrial fission is dependent on dynaminrelated protein 1 (Drp1), mitochondrial fission factor (Mff), mitochondrial fission protein 1 (Fis1) and mitochondrial dynamics proteins of 49 and 51 kDa (MiD49/51) for shrinkage of structures typically promoted during cellular replication (Korobova et al., 2013; Pagliuso et al., 2018). Depending on the isoform, post-translational modification of Drp1 by the small ubiquitin-like modifier (SUMO) drastically alters the effect on mitochondrial fission. SUMO-2/3 for instance has been shown to inhibit Drp1-Mff interaction preventing mitochondrial fragmentation, however, SUMO-1 stabilizes Drp1 leading to enhanced fragmentation; with evidence establishing a link between SUMO and PD, this further implicates mitochondrial dynamics in the pathogenesis of the disease (Guo et al., 2013, 2017; Fu et al., 2014; Vijayakumaran et al., 2015; Henley et al., 2018; Vijayakumaran and Pountney, 2018). Mitochondrial complex I inhibitors, Rotenone and MPP+, inducers of parkinsonian phenotypes, were shown to promote mitochondrial fission (Barsoum et al., 2006; Thomas et al., 2011) and Drp1 inactivation prevented the fission phenotype (Wang et al., 2011a). In the substantia nigra of sporadic PD patients, the short form of OPA1 (OPA1-S) was decreased in the absence of changes in Mfn1, further suggesting mitochondrial fusion deficiency (Zilocchi et al., 2018). Thus, a net increase in mitochondrial fission over fusion may ultimately lead to a fragmented mitochondrial network negatively impacting the efficiency of neuronal signaling in PD. Worth noting are reductions in Drp1 in the later stages of degeneration in transgenic mice (Xie and Chung, 2012). Mitochondria are also a major source of intracellular calcium which is released upon electron transport chain dysfunction, such as that caused by rotenone-mediated inhibition, and interacts with α-syn to promote aggregation. Indeed, induction of the endogenous neuronal calcium buffering protein, calbindin-D28k (CB), was able to block rotenone-induced α-syn aggregation and neurons expressing high levels of CB excluded α-syn inclusion bodies in both human and mouse model tissues (Rcom-H'cheo-Gauthier et al., 2016, 2017; McLeary et al., 2019). Moreover, the Miro1 protein acts as a calcium sensor at the mitochondrial outer membrane, reacting to high calcium by augmenting mitophagy (Nemani et al., 2018).
Intracellular mitochondrial dynamics involves the transport of mitochondrial units from one area of the cell to another, and is of key importance in neuronal cells that are polarized with long axons and dendrites (Lin and Sheng, 2015). Mitochondria need to be transported to synaptic terminals, active growth cones and axonal branches, where they maintain energy and Ca2<sup>+</sup> homeostasis (Sheng, 2017). Trafficking of mitochondria is microtubule (MT)-based, relying on ATP and motor proteins. MTs are polar α/β-tubulin polymers with the minus end within the cell body and the plus end to the cell extremity (Tas and Kapitein, 2018), and neurons can move mitochondria in both directions using independent motor proteins. Kinesins drive transport toward the cell extremities (anterograde transport) whereas dynein motors are responsible for retrograde transport (Hirokawa et al., 2010). Hydrolysis of ATP is essential to fuel mitochondrial movement in both directions (Zala et al., 2013). Bridging between mitochondria and MT-bound motor proteins are motor adaptor proteins such as Trafficking Kinesin Proteins 1 and 2 (TRAK1 and 2), which connect the OMM protein Mitochondrial Rho GTPase 1 (Miro1) to
kinesin (Melkov and Abdu, 2018). Indeed, Miro1 also mediates Drp1/Fis1-independent mitochondrial shape transition (Mist), needed for mitophagy (Nemani et al., 2018). Syntabulin can also link mitochondria and kinesins whilst syntaphilin acts as an anchor stopping mitochondrial movement (Cai et al., 2005; Chen and Sheng, 2013).
An increase in α-syn concentration was shown to result in mitochondrial traffic arrest even before axonal degeneration, affecting both anterograde and retrograde transport (O'Donnell et al., 2014; Pozo Devoto and Falzone, 2017). Nigral dopaminergic neurons especially display a "dying back" pattern, wherein anterograde trafficking of mitochondria becomes disrupted in early stage sporadic PD followed by retrograde transport at late stage (Chu et al., 2012). Loss of this trafficking severely reduces nigral neuron ability to regulate the conditions necessary for axonal signaling and neurotransmitter release at terminals. Regarding anterograde trafficking, α-syn oligomers disrupt through direct binding interactions between kinesin and the MT in addition to increasing expression of tau, a MT structure disruptor (Prots et al., 2013, 2018). α-Syn appears to induce MT fragmentation directly as well, hindering mitochondrial movement from distal cell areas (Melo et al., 2017). Interestingly, this finding produced the opposite outcome to Chu et al. (2012) whereby retrograde transport was disrupted first. This may be due to differences in α-syn species interaction with trafficking complexes disrupting mitochondrial transport as Chu et al. (2012) brought about "dying back" through viral overexpression of α-Syn whilst Melo et al. (2017) utilized an A53T transgenic model. Moreover, the PD-linked protein leucine rich repeat kinase-2 (LRRK2) seems to alter MT polymerization/depolymerization cycles, affecting mitochondrial trafficking (Gillardon, 2009; Godena et al., 2014). Impairment of mitochondrial transport is also induced by the parkinsonian toxin MPP+, which inhibits kinesin-1-mediated anterograde transport leading to an increase in dynein-dependent retrograde transport (Morfini et al., 2007). LRRK2 and PTEN-induced kinase 1 (PINK1) mutant drosophila also exhibit disturbed mitochondrial calcium homeostasis with functional involvement of Miro1 (Lee et al., 2018).
#### ROLE OF MITOPHAGY IN PD
Mitochondrial dynamics also includes mitochondrial clearance by mitophagy, a mitochondrion-specific autophagic process that drives dysfunctional mitochondria to degradation in
autophagosomes (Rodolfo et al., 2018). Several mitophagy mechanisms have been described, which can be dependent on or independent of mitochondrial receptors (Martinez-Vicente, 2017), including B-cell lymphoma 2 nineteen kilodalton interacting protein 3 (BNIP3), Nix, Bcl-2-like protein 13 (Bcl2-L-13) and Fun14 domain-containing protein 1 (FUNDC1), which interact with microtubule-associated proteins 1A/1B light chain 3B (LC3), recruiting the autophagosomal machinery (Chu, 2018). Cardiolipin may also mediate mitophagy, driving mitochondrial degradation when this phospholipid (that has a LC3-binding motif) moves from the IMM to the OMM (Chu et al., 2013). Receptor independent mitophagy involves the priming of mitochondria after PINK1 translocation from the cytosol to the mitochondria. The loss of mitochondrial membrane potential, and inhibition of PINK1-degrading proteases, leads to PINK1 accumulation in mitochondria, where it recruits the E3 ubiquitin ligase Parkin that initiates mitophagy by ubiquitinating OMM proteins, such as Mfn1 and Mfn2, Miro1, translocase of outer mitochondrial membrane 20 (TOM20), and voltage-dependent anion channel (VDAC) (Meissner et al., 2015). Poly-ubiquitinated OMM proteins act as mediators that ultimately cause mitochondrial engulfment by autophagosomes (Yoshii and Mizushima, 2015).
Evidence for a role of mitophagy in PD includes the observation that α-syn overexpression decreases the level of LC3 positive vesicles in human neuroblastoma cells (Winslow et al., 2010). Interestingly, mutations in PINK1 and Parkin genes cause autosomal recessive forms of familial early onset PD (Valente et al., 2001; Mata et al., 2004; Kumar et al., 2017), implicating a role of mitophagy in the aetiopathogenesis of PD (Dawson and Dawson, 2010). Mitochondrial morphology aberrations were found in a PINK1-mutant Drosophila model, and overexpression of Parkin was shown to rescue the phenotype (Clark et al., 2006). The role of Parkin mutations in mitophagy impairment was confirmed using iPSC-derived dopaminergic neurons with mutations in the Parkin gene (Suzuki et al., 2017). Additionally, Parkin has been demonstrated to be highly insoluble in PD, compromising autophagic systems (Lonskaya et al., 2013). Increased pathological α-syn leads to increased cytosolic Ca2<sup>+</sup> and Miro1 upregulation, however, the adaptor function of Miro1 between mitochondria and motor transport complexes is abrogated at high Ca2<sup>+</sup> concentration (Saotome et al., 2008; MacAskill et al., 2009; Wang and Schwarz, 2009). In addition to transport, increased Miro1 protects mitochondria from mitophagy. Thus, in PD it is likely that Miro1 upregulation in combination with PINK1 reduction serves to delay degradation allowing for unregulated ROS generation (Shaltouki et al., 2018). Interestingly, an increase in Miro1-dependent anterograde transport of mitochondria was found in a PINK1-knockout model (Liu et al., 2012). It is perhaps possible that in the early stages of PD, whilst cytosolic Ca2<sup>+</sup> is low, that the transport function of Miro1 is unaffected, then changing to act as a protector especially in later stages. Phosphorylation of Miro1 by PINK1 is required for Miro1 degradation, however, mutations in PD likely affect this function (Wang et al., 2011b; Shlevkov et al., 2016). In combination with the protective capabilities of Miro1, this may explain why there is a frequent accumulation of dysfunctional mitochondria in distal axonal areas in PD models whilst still ubiquitinated by Parkin. Recent work by Grassi et al. (2018) suggests the possibility of differing α-syn variants affecting different cellular systems, with a non-fibrillar phosphorylated α-syn species described to induce mitophagy. However, understanding of mitophagy in PD is relatively limited due to the lack of amenable in vivo experimental approaches.
### INTERCELLULAR TRANSFER OF MITOCHONDRIA AND α-SYNUCLEIN
Genes move from progenitors to progeny, i.e., in a vertical manner, while horizontal gene transfer (HGT) is rare among eukaryotes (Keeling and Palmer, 2008; Davis et al., 2014; Davis and Xi, 2015). Recently, HGT has been reported for mitochondrial genes, via horizontal transfer of mitochondria between cells in vitro (Rustom et al., 2004; Spees et al., 2006; Rogers and Bhatacharya, 2013; Ahmad et al., 2014; Wang and Gerdes, 2015; Rustom, 2016; Sinha et al., 2016). Wang and Gerdes (2015) showed that organelles, including mitochondria, move between cells via so-called tunnelling nanotubes (TNTs), narrow inter-cellular bridges with actin as a structural protein and with tubulin fibers as "tracks" for movement of subcellular structures between cells (Rustom et al., 2004). Co-culture studies showed that transfer of mitochondria from mesenchymal stem cells (MSCs) into cancer cells with defects in mitochondrial DNA (mtDNA) resulted in recovery of mitochondrial respiration in cancer cells (Spees et al., 2006) and mitochondria, moving from healthy cells via TNTs, rescued cancer cells exposed to mitochondrial insults during early stages of apoptosis (Wang and Gerdes, 2015). Mice with experimental lung disease were grafted with allogenic MSCs with labeled mitochondria resulting in movement of the mitochondria to the diseased cells, recovering their respiration and alleviating the pathology (Islam et al., 2012). Inter-cellular transfer of mitochondria maintains balanced heteroplasmy of mtDNA in outbred individuals (Jayaprakash et al., 2015), pointing to mitochondrial HGT as a more frequent event than previously considered (Berridge et al., 2015, 2016).
In relation to disease, transfer of damaged mtDNA has been observed in various mouse models of engrafted tumor cells with other works such as that by Dong et al. (2018) revealing that damaged mtDNA is transported within the mitochondria (Tan et al., 2015). Such a process allows for the damaged mtDNA to affect the acceptor cell through the generation and leakage of ROS, especially so if mitophagy or systems regulating the prevention of damaged mitochondrial spread are faulty. Regarding PD, knowledge on mtDNA and mitochondrial spread in general is still unclear. mtDNA mutations are observed in neuronal cells in early stages of PD (Braak Stage 3 onward) due to damage attained via oxidative stress (Lin et al., 2012). This is not observed in late stage PD. However, this could potentially be due to the increased chances of neurons that previously hosted the damaged mtDNA of being destroyed, in addition to a lower population of neurons with damaged mtDNA prior to death. Various interactions
of α-syn with mitochondria lead to the generation of ROS and thus increase risk of damaged mtDNA, including certain pathological α-syn species shown to bind with high affinity to the Tom20 mitochondrial outer membrane protein, thereby inhibiting mitochondrial protein uptake and promoting ROS generation (Di Maio et al., 2016; Grassi et al., 2018). Recent work by Wang et al. (2019) further reveals α-syn bound mitochondria only occurs with pathological α-syn aggregate species and not physiological monomers. This finding did not pertain only to PD but other synucleinopathies such as Dementia with Lewy Bodies and Multiple System Atrophy. In addition α-syn is also shown to inhibit Complex I directly, further compromising the energy production of the mitochondrion and increasing the generation of ROS (Reeve et al., 2015). No evidence has emerged to suggest that α-syn-affected mtDNA or mitochondria can spread to healthy cells, however, further investigation is required.
Mitochondrial transport is driven by motor complexes via binding through specific adaptor proteins, such as Miro1, suggesting that a potential mechanism of mitochondrial transfer between cells is mediated via the motor systems using TNTs (Ahmad et al., 2014; Sinha et al., 2016). This is based on experimental data (Ahmad et al., 2014) as well as analogy with the movement of mitochondria along axons in neuronal cells, where the kinesin and dynein motor systems include the motor protein plus two adaptor proteins Milton and Miro1, the latter with high affinity for mitochondria (Hase et al., 2009; Wang and Schwarz, 2009; MacAskill and Kittler, 2010; Kimura et al., 2012; Schiller et al., 2013). Indeed, mutations in the Miro1 gene, RHOT1, have recently been linked to PD cases, wherein the mutations led to fewer ER-mitochondria contact sites, calcium dyshomeostasis and exacerbation of calcium-dependent mitochondrial fragmentation and increased mitochondrial clearance (Grossmann et al., 2019). How the movement of mitochondria between cells is triggered and regulated is still not fully understood. Recent work has shown α-syn utilizing TNTs for spread. Thus, Dieriks et al. (2017) demonstrated the establishment of TNTs and transport of α-syn from neurons and pericytes in culture. Interestingly, Rostami et al. (2017) revealed TNT formation and α-syn transfer between astrocytic cells, transport of healthy mitochondria to rescue stressed mitochondria damaged by α-syn and transport of α-syn from α-syn affected astrocytes to healthy astrocytes. There were no reports of α-syn utilizing mitochondria for TNT-mediated migration, however, mitochondria may also represent an efficient
#### REFERENCES
carrier of α-syn between certain cell types. Furthermore, prion proteins, which have some similarities in mode of propagation to α-syn aggregates, have been established to hijack TNTs to mediate cell-to-cell transfer of the infectious protein (Gousset et al., 2009; Dorban et al., 2010; Prusiner et al., 2015; Steiner et al., 2018). Although there are many potential mechanisms of α-syn spread (Valdinocci et al., 2017), in view of findings that α-syn can migrate within TNTs bound to organelles such as lysosomal vesicles (Abounit et al., 2016) it is tempting to speculate that α-syn may transfer between neighboring cells bound to mitochondria that are translocated actively along TNTs (**Figure 1**).
### POTENTIAL FOR MITOCHONDRIALLY TARGETED PD THERAPEUTICS
The influences of pathological α-syn on mitochondrial dynamics in PD are potentially wide-ranging. Mitochondria-targeted drugs, such as Mito-Q and Mito-Apocynin showed therapeutic potential in experimental PD models whilst others such as Metformin are variable likely attributed to model utilized (Patil et al., 2014; Bayliss et al., 2016; Ismaiel et al., 2016; Lu et al., 2016; Langley et al., 2017; Xi et al., 2018). Although therapies that target calcium mobilization or oxidative stress tackle the effects of mitochondrial dysfunction, future innovative approaches could ameliorate mislocation of mitochondria or intercellular mitochondrial transfer. Further work is clearly needed to articulate the full significance of changed mitochondrial dynamics in PD etiology.
### AUTHOR CONTRIBUTIONS
All authors have contributed to the preparation of the manuscript, with DP taking responsibility for editing the final version of the manuscript.
### FUNDING
DV and DP were funded by Griffith University. JK and JN were funded by the Czech Academy of Sciences. RS and TC-O were funded by the ERDF through COMPETE 2020/FCT under research grants PD/BD/128254/2016 (RS), POCI-01-0145- FEDER-029297 (TC-O), and UID/NEU/04539/2019.
dynamin-related GTPases in neurons. EMBO J. 25, 3900–3911. doi: 10.1038/ sj.emboj.7601253
Shults, C. W. (2006). Lewy bodies. Proc. Natl. Acad. Sci. U.S.A. 103, 1661–1668.
**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Copyright © 2019 Valdinocci, Simões, Kovarova, Cunha-Oliveira, Neuzil and Pountney. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
# The Impairments of α-Synuclein and Mechanistic Target of Rapamycin in Rotenone-Induced SH-SY5Y Cells and Mice Model of Parkinson's Disease
#### Mahesh Ramalingam, Yu-Jin Huh and Yun-Il Lee\*
Well Aging Research Center, DGIST, Daegu, South Korea
#### Edited by:
Krishnan Prabhakaran, Norfolk State University, United States
#### Reviewed by: Senthil Selvaraj,
Sidra Medicine, Qatar Tito Cali', University of Padova, Italy Shreesh K. Ojha, United Arab Emirates University, United Arab Emirates
> \*Correspondence: Yun-Il Lee [email protected]
#### Specialty section:
This article was submitted to Neurodegeneration, a section of the journal Frontiers in Neuroscience
Received: 30 March 2019 Accepted: 10 September 2019 Published: 24 September 2019
#### Citation:
Ramalingam M, Huh Y-J and Lee Y-I (2019) The Impairments of α-Synuclein and Mechanistic Target of Rapamycin in Rotenone-Induced SH-SY5Y Cells and Mice Model of Parkinson's Disease. Front. Neurosci. 13:1028. doi: 10.3389/fnins.2019.01028 Parkinson's disease (PD) is characterized by selective degeneration of dopaminergic (DAergic) neurons in the substantia nigra pars compacta (SNpc). α-synuclein (α-syn) is known to regulate mitochondrial function and both PINK1 and Parkin have been shown to eliminate damaged mitochondria in PD. Mechanistic target of rapamycin (mTOR) is expressed in several distinct subcellular compartments and mediates the effects of nutrients, growth factors, and stress on cell growth. However, the contributions of these various regulators to DAergic cell death have been demonstrated mainly in culture with serum, which is known to dramatically influence endogenous growth rate and toxin susceptibility through nutrient and growth factor signaling. Therefore, we compared neurotoxicity induced by the mitochondrial inhibitor rotenone (ROT, 5 or 10 µM for 24 h) in SH-SY5Y cells cultured with 10% fetal bovine serum (FBS), 1% FBS, or 1% bovine serum albumin (BSA, serum-free). In addition, C57BL/6J mice were injected with 12 µg ROT into the right striatum, and brains examined by histology and Western blotting 2 weeks later for evidence of DAergic cell death and the underlying signaling mechanisms. ROT dose-dependently reduced SH-SY5Y cell viability in all serum groups without a significant effect of serum concentration. ROT injection also significantly reduced immunoreactivity for the DAergic cell marker tyrosine hydroxylase (TH) in both the mouse striatum and SNpc. Western blotting revealed that ROT inhibited TH and Parkin expression while increasing α-syn and PINK1 expression in both SH-SY5Y cells and injected mice, consistent with disruption of mitochondrial function. Moreover, expression levels of the mTOR signaling pathway components mTORC, AMP-activated protein kinase (AMPK), ULK1, and ATG13 were altered in ROT-induced PD. Further, serum level influenced mTOR signaling in the absence of ROT and the changes in response to ROT. Signs of endoplasmic reticulum (ER) stress and altered expression of tethering proteins mediating mitochondria-associated ER contacts (MAMs) were also altered concomitant with ROT-induced neurodegeneration. Taken together, this study demonstrates that complex mechanism involving mitochondrial dysfunction, altered mTOR nutrient-sensing pathways, ER stress, and disrupted MAM protein dynamics are involved in DAergic neurodegeneration in response to ROT.
Keywords: rotenone, SH-SY5Y, α-synuclein, mitofusin, stereotaxic, mTOR
## INTRODUCTION
fnins-13-01028 September 21, 2019 Time: 16:13 # 2
Parkinson's disease (PD) is an age-related neurodegenerative disorder (NDD) characterized by progressive loss of dopaminergic (DAergic) neurons in the substantia nigra (SN) pars compacta (SNpc) along with intracellular aggregation of α-synuclein (α-syn) in structures known as Lewy bodies (LBs) (Lee and Trojanowski, 2006). Neurons use multiple feedback controls to regulate metabolism in response to nutrients and other signals. Neurons depend on oxidative phosphorylation (OXPHOS) for most energy needs, a process that consumes oxygen and glucose to generate energy-storing ATP (Hsu and Sabatini, 2008). This process relies on electron flow via electron transport chain (ETC) components in the inner mitochondrial membrane, culminating in the reduction of oxygen in the matrix, and the generation of a membrane potential across the inner membrane that is exploited to convert ADP to ATP (Whitworth and Pallanck, 2017).
Rotenone (ROT) is a naturally occurring insecticide, pesticide, and piscicide extracted from the roots of plants of the genera Lonchocarpus and Derris. It is highly lipophilic and therefore easily crosses all biological membranes including blood-brain barrier (Martinez and Greenamyre, 2012). The mitochondrial toxin ROT is widely used to induced a PDlike pathology in culture cells and experimental animals. ROT impairs OXPHOS by inhibiting mitochondrial ETC complex I (reduced nicotinamide adenine dinucleotideubiquinone reductase), leading to reduced ATP production and the formation of reactive oxygen species (ROS) that can induce oxidative stress (Duty and Jenner, 2011; Martinez and Greenamyre, 2012). The major advantages of ROT treatment for PD modeling are its ability to induce α-syn-positive cytoplasmic inclusions in nigral neurons resembling LBs (Betarbet et al., 2000) and progressive neurodegeneration accompanied by PD-like motor and non-motor symptoms (Johnson and Bobrovskaya, 2015).
In addition to external stressors, nutrient signals, growth factors, and cell energy balance (a product of mitochondrial function) are major regulators of cellular growth, proliferation, and survival in health and disease (Yang et al., 2019). However, the signaling pathways underlying nutrient effects in PD are largely unknown. Mechanistic/mammalian target of rapamycin (mTOR) is a serine-threonine kinase that controls several important aspects of mammalian cellular function through nutrient signal transduction (Saxton and Sabatini, 2017). It exists in two distinct multiprotein complexes, mTORC1 and mTORC2 (Switon et al., 2017), each with its own unique subunit composition and functions. The mTORC1 complex comprised of mTOR, mLST-8, FKBP38, Deptor, PRAS40, and rapamycin-sensitive adaptor protein of mTOR (Raptor) regulates cell growth, proliferation, and metabolism (Bai and Jiang, 2010), whereas mTORC2 comprised of mTOR, mSIN1, mLST8, and the rapamycin-insensitive subunit Rictor (Laplante and Sabatini, 2009) controls cell survival and cell-cycle dependent cytoskeleton assembly. Each complex also utilizes distinct cofactors and substrates for regulation of these processes according to nutrient and energy status. Collectively, these two complexes regulate multiple physiologic processes (Morrison Joly et al., 2016), such as axonal growth, neuronal development and survival, and synaptic plasticity, thereby contributing to learning and memory (Jaworski and Sheng, 2006; Swiech et al., 2008). The AMP-activated protein kinase (AMPK) is another widely recognized energy-sensing serine/threonine kinase (Mihaylova and Shaw, 2011). AMPK responds to oxidative stress and critically involved in NDDs (Jiang et al., 2013). In addition, AMPK-driven mTOR downregulation serves as a turn-off switch of the cellular anabolic program (Swiech et al., 2008). AMPK directly regulates the autophagy-associated kinase ULK1 through phosphorylation under nutrient signaling. Moreover, mTORC1 also phosphorylates ULK1 at Ser757 and affect the interaction between ULK1 and AMPK (Kim et al., 2011; Shang et al., 2011). Therefore, these mutually interacting protein signaling pathways may sense and integrate countless stimuli from nutrients and growth factors to direct normal cellular processes and pathogenic processes in NDDs like PD (Linke et al., 2017).
Endoplasmic reticulum (ER) stress also contributes to multiple pathophysiological processes in NDDs (Lavoie et al., 2011). Autophagy, Ca2<sup>+</sup> homeostasis, lipid metabolism, mitochondrial ATP production, mitochondrial transport and biogenesis, ER stress, and the unfolded protein response (UPR) are fundamental cellular processes dependent on direct communication between the ER and mitochondria (Paillusson et al., 2017). Approximately 5–20% of the mitochondrial surface is closely apposed (within ∼10–30 nm) to ER membranes, forming specialized regions termed mitochondria-associated ER membranes (MAMs) (Paillusson et al., 2017). The MAM contains chaperones, oxidoreductases, calcium channels, calcium buffering proteins, and regulators of lipid metabolism. Thus, this subcellular compartment is likely involved in metabolic regulation by orchestrating protein folding, lipid synthesis, calcium buffering (Patergnani et al., 2011), and oxidation/reduction (Vance, 2014). MAM formation is dynamically regulated by tethering proteins between these organelles, such as glucose-regulated protein 75 (GRP75), mitofusin 1 (Mfn1), and Mfn2 (Ma et al., 2017). α-syn disrupts the MAM, which affects cellular exchange between the two organelles. Thus, MAM dysfunction may be a potential molecular mechanism linking α-syn to PD (Paillusson et al., 2017).
Despite extensive research efforts, it is still largely unclear how nutrients regulate the protein signaling pathways relevant to NDDs such as PD. Fetal bovine serum (FBS) has been used for mammalian cell culture to promote growth, differentiation, and survival (Piletz et al., 2018). However, serum is not a physiological fluid in vivo, and has been shown to induce aberrant cell growth characteristics, alter phenotype, and suppress neurotoxicity in vitro (Pirkmajer and Chibalin, 2011; Tekkatte et al., 2011). As many of these investigations have employed cells cultured in serum, the pathogenic pathways underlying the ROT-induced PD phenotype may differ substantially from PD pathogenesis in vivo. Therefore, it is critical to examine the effects of serumcontaining cell culture media on cellular models of PD. Here we examined ROT toxicity in SH-SY5Y neuroblastoma cells under three different cell culture conditions: (1) 10% FBS, (2) low (1%) FBS, and (3) serum-free medium (0% FBS) containing 1% bovine serum albumin (BSA) [used as synthetic serum as previously reported (Ramalingam and Kim, 2016)]. We also compared results to stereotaxic ROT injection in C57BL/6J mice. This study aimed to elucidate the molecular mechanism underlying ROT-induced toxicity, specifically the distinct contributions of mTOR/AMPK, and ER-mitochondrial tethering pathways.
## MATERIALS AND METHODS
fnins-13-01028 September 21, 2019 Time: 16:13 # 3
### Chemicals, Reagents and Antibodies
Dulbecco's modified Eagle's medium (DMEM), penicillin streptomycin (Pen Strep), trypsin-EDTA, and FBS were purchased from Welgene (South Korea). Rotenone (R8875), dimethyl sulfoxide (DMSO; D2650), Avertin (2,2,2- Tribromoethanol; T48402) were purchased from Sigma-Aldrich (St. Louis, MO, United States). All other chemicals and reagents were from commercial suppliers and of the highest purity available. Plastic materials were purchased from SPL Life Science (SPL, Seoul, South Korea). The primary and secondary antibodies used in this study were tabled in **Supplementary Table 1**.
### Cell Culture and Treatment
The human neuroblastoma cell line SH-SY5Y (CRL-2266) was obtained from ATCC (Manassas, VA, United States) and maintained in DMEM supplemented with 10% FBS, Pen Strep (100 U/ml; 100 mg/ml), and 2mM L-glutamine, at 37◦C in a humidified atmosphere containing 5% CO2/95% air. Confluent cultures were washed with phosphate-buffered saline (PBS), detached with 0.25% trypsin-EDTA solution, reseeded as 1 × 10<sup>5</sup> cells/ml of DMEM containing 10% FBS or 1% FBS or 1% BSA and used for experiments after overnight incubation. SH-SY5Y cells were incubated with the absence or presence of ROT for 24 h. Combining floating cells in the medium and adherent cells detached by trypsinization and subjected to cell counting and Western blotting.
### Cell Counting and Cell Morphology
After treated with ROT or solvent control (DMSO) at the indicated concentrations for 24 h, phase contrast images were taken using microscope Olympus CKX41 equipped with a camera. Damaged and deplated floating cells in the medium and adherent cells detached by trypsinization were combined and subjected to trypan blue cell counting method. Surviving cells, which cannot be stained with trypan blue dye, were counted under microscope. The cell count assay was performed in triplicates and expressed as a percentage (%) of control.
### Preparation of Total Cell Lysates and Immunoblotting
After treated with indicated concentrations of ROT or DMSO for 24 h, cells were harvested by scraping with media, pelleted and washed twice with PBS. Then, exposed to RIPA buffer (25 mM Tris–HCl (pH 7.6), 150 mM NaCl, 1% Nonidet P-40, 0.251% sodium deoxycholate, 1% sodium dodecyl sulfate (SDS): Thermo Fisher Scientific, United States) supplemented with protease and phosphatase inhibitors cocktail (Thermo Fisher Scientific, United States) and incubated for 30 min in ice. Lysates were centrifuged at 13,000 rpm for 20 min at 4◦C and the supernatants were collected as total cell lysate. Protein concentrations were determined by BCA method (Kit). Proteins (30 µg) were separated on 6–12% SDS-polyacrylamide gels and transferred to PVDF membranes (Millipore, Bellerica, MA, United States). The membranes were washed with Tris buffered saline (TBS; 10 mM Tris–HCl, 150 mM NaCl, pH 7.5) containing 0.5% (v/v) Tween 20 (TBST) followed by blocking with 5% (v/v) non-fat dried milk solution prepared in TBST and then incubated overnight with primary antibodies at 4◦C. The antibodies used are listed in **Supplementary Table 1**. After this, membranes were exposed to secondary antibodies conjugated to horseradish peroxidase for 2∼3 h at room temperature and further washed thrice with TBST. The immunoreactivity was detected by the luminolbased chemiluminescence (ECL) system. Equal protein loading was assessed by the expression level of β-actin. Densitometric analysis was performed using ImageJ (National Institute of Health, Bethesda, MD, United States) software.
### Triton-X-100-Soluble and -Insoluble Fractionation
Following ROT toxicity for 24 h, SH-SY5Y cells were lysed on ice in RIPA buffer containing protease and phosphatase inhibitors with 1% Triton-X-100 for 30 min. Lysates were centrifuged at 12,000 rpm for 20 min at 4◦C and the supernatants were collected as Triton-X-100-souble fraction. The cell pellets were washed with PBS then dissolved in the RIPA buffer containing protease and phosphatase inhibitors with 1% Triton-X-100 and 2% SDS and sonicated for 10 s and used as Triton-X-100-insoluble fraction. Protein samples were immunoblotted as described above.
### Animals and Stereotaxic Surgery
Five-week-old C57BL/6J male mice were purchased from DBL (South Korea) were housed at room temperature under 12 h ligh/dark cycle. Food and water were provided ad libitum for 1 week before intrastriatal surgery. All animal experiments were approved by the Ethical Committee of Animal Research of DGIST, Daegu, South Korea accordance with international guidelines (DGIST-IACUC-18010204-01). Mice underwent unilateral stereotaxic surgery under injectable 2.5% Avertin anesthesia. A hole was drilled in the skull and a cannula inserted at following stereotaxic coordinates at AP + 1.0, ML −2.5 from bregma and DV −3.0 below dura in the right striatum, and 12 µg of freshly prepared ROT (dissolved in 2 µl DMSO) was infused (0.2 µl/min for 10 min for infusion with 5 min for diffusion). Control animals were injected with vehicle DMSO. Fourteen days after surgery, mice were anesthetized with avertin and perfused.
#### Immunohistochemistry
Mice were sacrificed by terminal anesthesia and transcardially perfused with 50 ml PBS followed by 50 ml 4% paraformaldehyde (PFA). Brains were rapidly removed, post-fixed in 4% PFA for 24 h and stored in a 30% sucrose solution for 48 h or
more. Serial coronal sections (35 µm) were cut using a freezing sledge microtome and a 1:4 series of sections was used for all quantitative immunohistochemistry. For immunohistochemical analyses, blocking of non-specific secondary antibody binding (using 3% normal horse serum in PBS with 0.2% Triton X-100 at room temperature for 1 h), sections were incubated overnight at room temperature with the primary antibody of TH (1:1000) diluted in PBS with 0.2% Triton X-100. Sections were then incubated in a biotinylated secondary antibody for 1 h [horse anti-mouse (1:2000, Vector, United Kingdom)] followed by 1 h incubation in streptavidin-biotin-horseradish peroxidase solution (Vector, United Kingdom). Sections were developed in 0.5% solution of diaminobenzidine (DAB) tetrahydrochloride (Sigma, Ireland) and mounted on microscope slides coverslipped using DPX mountant (BDH chemicals, United Kingdom). Immunostained sections were photographed and the total number of TH positive neurons in the SNpc was determined using the Optical fractionator probe in Stereo Investigator software (MicroBrightfield, Williston, VT, United States). All stereological counting was performed in a blinded manner to mice treatments.
### Cell Lysates Preparation and Immunoblotting
After fourteen days of surgery, mice were anesthetized with avertin and perfused with PBS. Mice brain subregions of midbrain (MB) and striatum (ST) were located following procedures described previously (Jackson-Lewis and Przedborski, 2007). The lysates were prepared and Western blotting done as mentioned above.
#### Statistical Analysis
Data are expressed as mean ± standard error mean (SEM). The significance level of treatment effects was determined using one-way analysis of variance (ANOVA) followed by Tukey's multiple comparison test (in vitro; three or more groups) or an paired/unpaired two-tailed Student's t-test (in vivo; two-group comparisons). A probability of <5% (p < 0.05) was considered to be statistically significant. GraphPad Prism 5.0 software (La Jolla, CA) was used for data analyses and preparation of all graphs.
### RESULTS
#### Rotenone-Induced Death in SH-SY5Y Cells
SH-SY5Y cells were cultured in the same medium containing different serum concentrations and then treated with ROT (0, 0.5, 1, 2.5, 5, and 10 µM) for 24 h. Microscopic examination revealed that the majority of cells were damaged and deplated following 5 and 10 µM ROT treatment in 1% FBS and 1% BSA culture medium (**Supplementary Figure 1**). Trypan blue cell viability assays combining both floating and adherent cells detached by trypsinization revealed little proliferation in 1% FBS or 1% BSA (p < 0.001) compared to control 10% FBS culture media. ROT dose-dependently increased cell death in all groups after 24 h (all p < 0.001; **Figure 1A**). However, cell death was substantially greater in the low-serum and no-serum groups.
FIGURE 1 | SH-SY5Y cells were seeded as 1 × 10<sup>5</sup> cells/ml of DMEM containing 10% FBS or 1% FBS or 1% BSA and used for experiments after overnight incubation. Cells were incubated with the absence or presence of different concentrations of ROT (0, 0.5, 1, 2.5, 5, and 10 µM) for 24 h and assessed for trypan blue assay (A) or TH, PINK1, Parkin, α-syn and β-actin by Western blotting (B). In addition, cells were fractionated into 1% Triton X-100 soluble and insoluble fractions and analyzed for α-syn by Western blotting (C). Each picture is a representative of three independent experiments. Data are mean ± SEM of three independent experiments and analyzed by one-way of variance (ANOVA) followed by Tukey's post hoc test. Statistical significance: <sup>a</sup>compared with 10% FBS control; <sup>b</sup>compared with 1% FBS control; <sup>c</sup>compared with 1% BSA control; <sup>∗</sup>p < 0.05, ∗∗p < 0.01, and ∗∗∗p < 0.001.
### Rotenone Alters TH, Parkin, PINK1, and α-Syn Expression in SH-SY5Y Cells
To further evaluate ROT toxicity in SH-SY5Y cells, we measured expression of tyrosine hydroxylase (TH), the rate-limiting enzyme for dopamine (DA) synthesis. Indeed, ROT significantly reduced TH protein expression in all serum groups (p < 0.05; **Figure 1B**). Moreover, ROT treatment (10 µM for 24 h) significantly increased expression of PINK1 (p < 0.001 in 10% FBS, p < 0.01 in 1% FBS, p < 0.001 in 1% BSA) and decreased expression of Parkin (p < 0.05 in 10% and 1% FBS, p < 0.01 in 1% BSA) as evidenced by Western blotting (**Figure 1B**), suggesting effects on mitochondrial function, quality control, and mitophagy.
Parkinson's disease is characterized by the presence of abnormal intracellular α-syn inclusions. ROT treatment increased α-syn in total cell lysates from all three serum concentrations groups (p < 0.001; **Figure 1B**), suggesting that ROT reduces SH-SY5Y DAergic neuron viability by promoting α-syn accumulation. Separate Western blot analyzes of the Triton-X100-soluble and insoluble lysate fractions, which are thought to include bioavailable and aggregated α-syn proteins, respectively revealed an increase of the oligomeric form in the Triton-X100-insoluble fraction and an increase of the monomeric form in the Triton-X100-soluble fraction. Thus, ROT may induce SH-SY5Y cell death through enhanced α-syn production and ensuing aggregation (**Figure 1C**).
### Rotenone Alters mTORC and AMPK Expression Levels in SH-SY5Y Cells
The activities of mTORC1 and mTORC2 were assessed by measuring expression levels of phosphorylated (p)-mTORC1 (Ser2448) and p-mTORC2 (Ser2481), respectively. Control cells cultured in 1% FBS or 1% BSA exhibited slightly increased p-mTORC1 and significantly decreased p-mTORC2 (p < 0.001 and p < 0.01, respectively) (**Figure 2A**) compared to control cells cultured in 10% FBS. ROT (10 µM) dramatically increased p-mTORC1 (p < 0.05) and decreased p-mTORC2 (p < 0.001) expression in cells cultured with 10% FBS. Conversely, ROT decreased p-mTORC1 (p < 0.05 and p < 0.001, respectively) and increased p-mTORC2 (both p < 0.001) in cells cultured with 1% FBS or 1% BSA. These data suggest that FBS levels and ROT both alter mTOR signaling pathways in SH-SY5Y cells and that the effects of ROT differ depending on serum levels (nutrient availability). The overall effect of mTOR signaling depends on the specific complex activated, mTORC1 or mTORC2, which are distinguished by Raptor in mTORC1 and Rictor in mTORC2. Thus, we measured expression levels of p-Raptor (Ser792) and p-Rictor (Thr1135). Control cells cultured in 1% FBS or 1% BSA demonstrated no difference in p-Raptor expression but increased p-Rictor expression (p < 0.01 and p < 0.001, respectively) compared to control SH-SY5Y cells cultured in 10% FBS (**Figure 2B**), suggesting mTORC2 signaling predominance under low nutrient conditions. ROT treatment (10 µM for 24 h) decreased the expression levels of both p-Raptor (both p < 0.001) and p-Rictor (p < 0.05 in 10% and 1% FBS; p < 0.01 in BSA) in all three culture conditions, while total mTOR, Raptor and Rictor expression levels were unaffected (**Figures 2A,B**).
AMP-activated protein kinase is a major metabolic energy sensor that contributes to mTOR signaling through interactions with ULK1 and ATG13. Control cells cultured in 1% FBS or 1% BSA showed increased expression levels of p-AMPK (Thr172) (both p < 0.001), p-ULK1 (Ser757) (p < 0.001 and p < 0.05), and ATG13 (both p < 0.001) compared to control cells cultured in 10% FBS (**Figure 2C**). Treatment with ROT (10 µM for 24 h) significantly enhanced p-AMPK (Thr172), p-ULK1 (Ser757), and ATG13 expression by SH-SY5Y cells cultured in 10% FBS (all p < 0.001) but decreased expression levels of all three phosphorylated proteins in SH-SY5Y cells cultured with 1% FBS and 1% BSA, again indicating that serum influences mTOR signaling independently of ROT, and alters the mTOR signaling change in response to ROT.
### Rotenone Induces ER Stress and Disrupts MAM in SH-SY5Y Cells
The endoplasmic reticulum is the central organelle responsible for protein folding, and there is compelling evidence that ER stress and protein misfolding are involved in ROT-induced PD-like toxicity. To examine ER stress in SH-SY5Y cells, we measured changes in the expression levels of protein kinase RNA (PKR)-like ER kinase (PERK) and inositol-requiring enzyme 1 α (IRE-1α). ROT (10 µM) increased expression of p-PERK at Thr981 (p < 0.001) and IRE-1α (p < 0.01) (**Figure 3**), suggesting that ER stress is involved in ROT-induced neuronal dysfunction. We further investigated the protein expression levels of MAM tethering proteins GRP75, Mfn1, and Mfn2. Control cells cultured with 1% FBS or 1% BSA showed significantly increased expression levels of GRP75 (both p < 0.001), Mfn1 (p < 0.01 in 1% FBS and p < 0.001 in 1% BSA), and Mfn2 (both p < 0.05) compared to control cells in 10% FBS (**Figure 3**). In addition, ROT (10 µM for 24 h) further increased the protein expression levels of GRP75 (p < 0.001) in all culture conditions. However, Mfn1 and Mfn2 were significantly increased by ROT (10 µM for 24 h) in cells cultured with 10% FBS (p < 0.01 and p < 0.05, respectively) but decreased in cells cultured in 1% FBS (both p < 0.05) or 1% BSA (p < 0.001 for Mfn1 and p < 0.01 for Mfn2).
### Intrastriatal Rotenone Injection Alters TH, PINK1, Parkin, and α-Syn Expression Levels in C57BL/6J Mice
To examine the effects of ROT on mTOR signaling, ER stress, mitochondrial function, MAM function, and cell viability in vivo, we conducted single unilateral intrastriatal ROT infusions (12 µg) in C57BL/6J mice. Instrastriatal injection induced, significant depletion of TH immunoreactivity in the striatum and SN (**Figure 4A**), and significantly reduced TH-positive cell numbers in the ipsilateral (injection-side) SNpc after 14 days (p < 0.01). The expression levels of TH and Parkin were decreased while PINK1 and α-syn expression levels were increased in midbrain (**Figure 4B**) and striatum (**Figure 4C**) of ROT-injected mice compared to vehicle (DMSO)-injected mice
as measured by Western blotting, consistent with the effects of SH-SY5Y cells.
#### Intrastriatal Rotenone Alters mTOR and AMPK Signaling in C57BL/6J Mice
We then investigated the involvement of mTOR pathways in ROT toxicity. ROT injection increased p-mTORC1 (Ser2448) expression (p < 0.01) but reduced p-mTORC2 (Ser2481) expression (p < 0.05) in mouse midbrain (**Figure 5A**). Alternatively, ROT dramatically decreased both p-mTORC1 and p-mTORC2 in mouse striatum (p < 0.001; **Figure 5C**). In addition, p-Raptor and p-Rictor expression levels were reduced in the midbrain of ROT-injected mice (p < 0.01 and p < 0.001, respectively) (**Figure 5A**). Collectively, these findings suggest that ROT has region-specific effects on mTOR signaling pathways. Moreover, the protein expression levels of p-AMPK, p-ULK1, and ATG13 were decreased in ROT-injected mice (all p < 0.01), while t-AMPK and t-ULK1 expression levels remained unchanged (**Figure 5B**). In the striatum of ROT-injected mice, p-AMPK was increased while p-ULK1 and ATG13 expression levels were reduced (**Figure 5D**).
### Intrastriatal Rotenone Alters ER Stress and Disrupts MAM in C57BL/6J Mice
To examine possible ROT-induced ER stress in the brain, we measured expression of the ER stress-associated proteins PERK and IRE-1α in midbrain (**Figure 6A**) and striatum (**Figure 6B**) lysates. Expression levels of p-PERK (Thr981) and IRE-1α were markedly increased by ROT in both midbrain and striatum (p-PERK: p < 0.001 in both regions; IRE-1α: p < 0.001 in midbrain, p < 0.01 in striatum) implicating ER stress in ROT induced neurodegeneration. Moreover, ER malfunction involved MAM dysfunction as the MAM tethering proteins GRP75, Mfn1, and Mfn2 were downregulated in midbrain (p < 0.05 for GRP75 and Mfn1; p < 0.01 for Mfn2) and striatum (p < 0.05 for GRP75; p < 0.001 for Mfn1 and Mfn2) of ROT-injected mice.
#### DISCUSSION
In recent years, many studies have been conducted to elucidate the molecular pathophysiology of PD progression. ROT, a fish poison that inhibits mitochondrial complex I, induces PD-like changes in cultured neurons, and rodent
FIGURE 4 | Photomicrograph of ROT lesioned C57BL/6J mice striatum and substantia nigra pars compacta (SNpc) immunostained for TH (25×). The quantitative image analysis of TH positive cells in SNpc expressed as percentage (%). Each picture is a representative of three independent experiments. Data are mean ± SEM (n = 3) and analyzed by paired two-tailed Student's t-test. ∗∗p < 0.01 (A). Effects of ROT on TH, PINK1, Parkin, α-syn, and β-actin in the midbrain (B) and striatum (C) of stereotaxic C57BL/6J mice were analyzed by Western blotting. Data are mean ± SEM (n = 4 for control; n = 5 for ROT) and analyzed by unpaired two-tailed Student's t-test. <sup>∗</sup>p < 0.05, ∗∗p < 0.01, and ∗∗∗p < 0.001.
brain (Sherer et al., 2003). SH-SY5Y is a DA-producing human catecholaminergic neuroblastoma cell line widely used as an in vitro dopaminergic cell model (Martins et al., 2013). ROT treatment for 24 h dose-dependently reduced SH-SY5Y cell number. Moreover, serum starvation by culture in 1% FBS or 1% BSA medium inhibited cell growth and proliferation in the absence of ROT. Cells under serum starvation alone or with ROT detached from the culture surface and lost typical neuronal morphology within 24 h consistent with previous studies showing that serum starvation induces cell death (Chou and Yung, 1997) and that ROT-induced impairment of mitochondrial complex I activity leads to apoptosis, likely via excess ROS formation (Imamura et al., 2006).
Tyrosine hydroxylase, the rate-limiting enzyme in DA synthesis, is obviously critical for phenotypic expression (Nagatsu et al., 1964). ROT treatment for 24 h dramatically reduced TH expression in SH-SY5Y cells. Moreover, ROT infusion in mouse striatum reduced TH immunoreactivity and reduced THpositive cell count in the SNpc after 14 days, consistent with the dopaminergic neuronal degeneration along the nigrostriatal pathway that parallels the symptoms of PD (Carriere et al., 2016). In addition, surviving cells accumulate damaged mitochondria, leading to metabolic deficits, oxidative stress, mitophagy, and greater susceptibility to other pathogenic processes such as protein aggregation (Shaltouki et al., 2015).
Parkin (PARK2), a cytosolic ubiquitin ligase, and PTENinduced kinase 1 (PINK1; PARK6), a mitochondria-targeted kinase, act important mediators of mitochondrial quality control (Whitworth and Pallanck, 2017) by removing damaged or dysfunctional mitochondria and preserving a healthy mitochondrial population (Oh et al., 2017). Dysfunctional mitochondria can trigger cell degeneration via mitophagy in PD (Li et al., 2015). In the present study, ROT inhibited Parkin expression and enhanced PINK1 expression in SH-SY5Y cells as well as in mouse midbrain and striatum, suggesting that loss of coordinated Parkin/PINK1 function contributes to ROT-induced mitochondrial impairment, oxidative stress, and cell death. Loss of Parkin results in an
initial accumulation of damaged mitochondria while PINK1 accumulation may cause proteosomal dysfunction (Wang et al., 2005; Shaltouki et al., 2015), which reduces Parkin solubility in toxin-induced PD models. Moreover, reduced Parkin in turn leads to the formation of protein aggregates resembling LBs in PD (Um et al., 2009, 2010).
Protein aggregation and filament formation are histopathological hallmarks of NDDs. For instance, PD is characterized by abnormal intracellular protein inclusions (LB and Lewy neurites or LN) mainly composed of aggregated α-syn fibrils (Faustini et al., 2017). The aggregation of α-syn causes proteasomal dysfunction, which may lead directly to neurodegeneration (Tanaka et al., 2001). We observed α-syn accumulation in SH-SY5Y cells and mouse midbrain and striatum following ROT treatment, consistent with previous studies demonstrating that ROT can trigger α-syn accumulation both in vitro and in vivo (Yuan et al., 2015). Aggregation of α-syn can be spread via a prion-like mechanism to neighboring neurons (Bae et al., 2012), resulting in the functional decline and death of dopaminergic neurons throughout the SNpc (Li et al., 2008; Lee et al., 2010). Moreover, α-syn can bind to TH, so accumulation of α-syn can reduce DAergic transmission by surviving cells (Perez and Hastings, 2004).
Under normal conditions, α-syn is cleared by proteolytic degradation in the extracellular space or lysosomal degradation in neighboring cells (Sung et al., 2005; Lee et al., 2008). However, impairment of these degradative mechanisms or an incapacity to clear proteins that have already aggregated is proposed the as primary defect leading to accumulation of insoluble α-syn (Cantuti-Castelvetri et al., 2005) within the LBs characteristic of PD (Myohanen et al., 2012), dementia with LB, and multiple system atrophy (Iwata et al., 2003). Previous in vitro studies have reported that the Triton-X100-soluble fraction represents bioavailable α-syn, whereas the Triton-X100-insoluble component may represent α-syn aggregated or sequestered in oligomeric forms (Klucken et al., 2003; Alves da Costa et al., 2006). In our present study, the oligomeric form of α-syn was detected only in the Triton-X100-insoluble fraction, while the monomeric form of α-syn was detected only in the Triton-X100-soluble fraction. ROT increased both Triton-X100-insoluble α-syn oligomers and Triton-X100-soluble monomer, consistent with previous studies reporting that ROT interacts with α-syn to drive accumulation of insoluble forms in SH-SY5Y cells (Sherer et al., 2002; Lee et al., 2004).
Growth factors and nutrients enhance protein synthesis and suppress protein degradation (Zhao et al., 2015). mTORs integrate signals from nutrients and growth factors with current energy status to regulate many neuronal processes, including autophagy, ribosome biogenesis, and growth (Sarbassov et al., 2005), as well as synaptic plasticity, learning and memory, and food uptake in adult brain (Zhou et al., 2015). Our present study suggests that both mTORC1 and mTORC2 regulate the responses to FBS and ROT in SH-SY5Y cells. ROT increased p-mTORC1 but decreased p-mTORC2 protein expression in the presence of 10% FBS, consistent with changes in the midbrain following ROT injection. Therefore, mTORC1 appears to be activated by ROT in the presence of sufficient nutrients and growth factors. The exact mechanism by which mTORC1 is activated and mTORC2 inhibited by ROT under this nutrient-rich condition requires additional studies. Control cells treated with 1% BSA also exhibited enhanced p-mTORC1 and reduced p-mTORC2 consistent with a previous study (Peruchetti et al., 2014). Conversely, ROT decreased p-mTORC1 and increased p-mTORC2 expression in the presence of 1% FBS or 1% BSA, changes shown to directly inhibit cell growth, and mitochondrial proteins during nutrient starvation. Our results are in line with another study reporting decreased mTORC1 in PC12 cells following ROT treatment for 24 h (Laplante and Sabatini, 2009). These results suggest that ROT treatment under nutrient shortage (serum starvation) inhibits mTORC1, resulting in lower mitochondrial membrane potential, oxygen consumption, and cellular ATP levels (Schieke et al., 2006).
Studies also suggest that mTORC1 acts as a negative regulator of mTORC2 (Dibble et al., 2009; Xie and Proud, 2014). Another study found that Rheb activated mTORC1 but inhibited mTORC2, while TSC1/2 inhibited Rheb/mTORC1 but activated mTORC2, most likely by overcoming the negative feedback loop (Yang et al., 2006). To further explore the role of mTOR complex signaling pathways in ROT toxicity, we examined expression of Raptor, a required cofactor for rapamycin-sensitive mTORC1 signaling, and Rictor, a regulatory subunit of mTORC2 (Li et al., 2007). ROT treatment decreased the phosphorylation levels of Raptor and Rictor both in vitro and in vivo, inconsistent with the aforementioned changes in mTORC1 and mTORC2 activity and suggesting that ROT can differentially stimulate or inhibit mTOR complex activities under different nutrient conditions; however, further investigations using specific substrates of these mTOR complexes necessary to examine the underlying mechanisms.
mTORC1 is the major transducer of nutrient signaling for cell growth (Efeyan et al., 2012). Similarly, AMPK senses energy deficiency in the form of an increased AMP/ATP ratio to regulate a myriad of cellular processes (Poels et al., 2009). The substrates ULK1 (ATG1) and ATG13 function downstream of mTORs and AMPK (Hosokawa et al., 2009; Egan et al., 2011). In this study and others (Wu et al., 2011), ROT-induced AMPK phosphorylation at Thr172 in SH-SY5Y cells cultured with 10% FBS medium. Increased p-AMPK was also reported in ROT treated HepG2 cells (Hou et al., 2018). Furthermore, activated ULK1 was shown to directly phosphorylate AMPK (Loffler et al., 2011). In contrast, Sciarretta et al. (2018) reported that ROT decreased mTORC1 and inhibited AMPK-ULK1 under serum starvation. Under starvation, AMPK translocates to the lysosome and both lysosomal AMPK and mTORC1 contribute to autophagy via ULK1/ATG13/FIP200 regulation (Ha et al., 2015). In this study, ROT inhibited mTORC1, AMPK/ULK/ATG13 in cells cultured with 1% FBS or 1% BSA (nutrient starvation). In ROT-injected mouse midbrain, the activation of mTORC1 by ROT inhibited the AMPK/ULK1/ATG13 pathway after 14 days. From the above results, we speculate that the nutrient-sensing molecules mTORC1 and AMPK/ULK1/ATG13 are differently regulated by ROT toxicity in vitro and in vivo.
The endoplasmic reticulum is involved in protein folding, maintenance of Ca2<sup>+</sup> homeostasis and cholesterol synthesis, and ER dysfunction is implicated in the pathogenesis of α-syn mediated NDDs (Paillusson et al., 2017). The ER depends on ATP to correct misfolded protein errors, so ROT-induced ATP reduction can lead to ER stress, which in turn initiates the UPR through activation of PERK, and IRE-1α (Jiang et al., 2016). In the present study, ROT enhanced p-PERK Thr981 and IRE-1α expression levels, indicating ROT-induced ER stress, in both SH-SY5Y cells and mouse midbrain and striatum. This enhanced activation of PERK signaling results in a sustained reduction of global protein synthesis, leading to neuronal loss (Liu et al., 2015). In addition, PERK over-activation has been observed in postmortem brain and spinal cord tissues of NDD patients (Smith and Mallucci, 2016). Taken together, the present study suggests that ROT-induced ER stress may trigger the death of SH-SY5Y cells and DAergic neurons in mice.
Mitochondria-associated ER membrane abnormalities have been described in cellular models of a number of NDDs, including PD (Cali et al., 2013). GRP75 is essential for maintaining physical contact between the ER and mitochondria, thereby facilitating Ca2<sup>+</sup> exchange and transfer through ERbound IP3R and mitochondrial VDAC1 (Honrath et al., 2017). Expression of GRP75 was increased in SH-SY5Y cells by ROT treatment but decreased in midbrain and striatum of C57BL/6J mice. As previously reported, GRP75 overexpression in human DAergic cells enhanced vulnerability to ROT-induced cytotoxicity (Jin et al., 2006), but in vivo GRP75 overexpression reduced infarct size and protected against mitochondrial damage in a rat middle cerebral artery occlusion model of stroke (Xu et al., 2009) and a rat model of intracerebral hemorrhage (ICH) (Lv et al., 2017). GRP75 expression was also decreased in the mitochondrial fraction isolated from the SNpc of PD patients compared to controls (Jin et al., 2006). The above results suggest that GRP75 may be either beneficial or harmful in different pathogenic contexts, although the exact mechanisms are still unknown (Shi et al., 2008). One possible explanation is provided by reports that GRP75 interacted with MAM-associated α-syn in a DAergic cell line with strong membrane attachment affinity due to a high lipid:protein ratio (Zabrocki et al., 2008; Fantini and Yahi, 2011).
The MAM tethering proteins mitofusions Mfn1 and Mfn2 are dynamin-related GTPases responsible for membrane fusion via a large cytosolic GTPase domain embedded in the outer mitochondrial membrane (Rojo et al., 2002; Suarez-Rivero et al., 2016). Mfn2 heterodimerizes with Mfn1 to link ER and mitochondria to regulate organelle tethering (de Brito and Scorrano, 2008; Chen et al., 2012). Our study suggests that α-syn can affect mitochondrial morphology and modulate mitochondrial dynamics through reduction of Mfn1- and Mfn2 dependent tethering in SH-SY5Y cells and mouse brain (Xie and Chung, 2012). Moreover, these findings identify MAM interface and inter-organelle contact disruption as novel mechanisms of ROT-induced α-syn toxicity.
Many studies rely on artificially overexpressed or recombinantly tagged proteins to investigate the underlying signaling mechanisms in PD. However, in the present study, we compared ROT-induced toxicity among cultures under different nutrient availability conditions and in vivo to further elucidate the potential mechanisms of α-syn neurotoxicity. As expected, ROT decreased TH and concomitantly increased α-syn accumulation, indicating degeneration of dopaminergic neurons both in vitro and in vivo. The nutrient-sensing mTOR and AMPK pathways appear to form a negative feedback loop that regulates the extent of ROT toxicity. More importantly, the present study implies that ER and MAM tethering proteins are also intimately involved in ROT-induced neurodegeneration. From these findings, we conclude that ROTtreated cell culture systems and mouse models are limited for recapitulating the clinical and pathological phenotypes of PD. We also conclude that serum concentration in culture medium can greatly influence the effects of ROT on cell growth, oxidative stress, and the various underlying signaling pathways. Further research in vitro and in vivo is necessary to establish stronger links between ROT-induced pathogenic mechanisms and human PD.
#### DATA AVAILABILITY STATEMENT
All datasets generated for this study are included in the manuscript/**Supplementary Files**.
### ETHICS STATEMENT
All animal experiments were approved by the Ethical Committee of Animal Research of the DGIST, Daegu, South Korea accordance with international guidelines.
### AUTHOR CONTRIBUTIONS
MR and Y-IL conceived and designed the study, and wrote the manuscript. MR performed the major experiments. Y-JH helped to perform the animal experiments. MR analyzed the data. All authors read and approved the manuscript for publication.
### FUNDING
This work was supported by the DGIST R&D Program of the Ministry of Science, ICT and Future Planning (18-LC-01 and 18-BT-01).
#### SUPPLEMENTARY MATERIAL
The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fnins. 2019.01028/full#supplementary-material
### REFERENCES
fnins-13-01028 September 21, 2019 Time: 16:13 # 13
independently of AMPK activation. J. Cell Mol. Med. 22, 1316–1328. doi: 10. 1111/jcmm.13432
**Conflict of Interest:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Copyright © 2019 Ramalingam, Huh and Lee. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
digital media
of impactful research
article's readership | doab | 2025-04-07T03:56:58.432134 | 18-11-2021 17:18 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "0025fd58-dc3d-4c1b-9075-3b20d5ae0a98",
"url": "https://www.frontiersin.org/research-topics/8659/mitochondria-and-endoplasmic-reticulum-dysfunction-in-parkinsons-disease",
"author": "",
"title": "Mitochondria and Endoplasmic Reticulum Dysfunction in Parkinson’s Disease",
"publisher": "Frontiers Media SA",
"isbn": "9782889633340",
"section_idx": 0
} |
00274f8d-5b4d-46b7-be49-66627b48ea8d | # New technologies to improve the ex situ conservation of plant genetic resources
Fiona R. Hay, Aarhus University, Denmark; and Sershen, University of the Western Cape & Institute of Natural Resources, South Africa
# **New technologies to improve the ex situ conservation of plant genetic resources**
*Fiona R. Hay, Aarhus University, Denmark; and Sershen, University of the Western Cape & Institute of Natural Resources, South Africa*
1 Introduction
New technologies to improve the ex situ conservation of
plant genetic resources
# **1 Introduction**
Plant genetic resources, which include traditional and modern varieties, crop wild relatives, genetic stocks, breeding lines and weedy species, form the genetic basis for the improvement and selection of crops through breeding (IPGRI, 1994). While the ex situ conservation of crop genetic resources will contribute towards addressing the challenge of food insecurity, as set out in the United Nations Sustainable Development Goals (www.sustainable development.un.org/), the ex situ conservation of wild species can facilitate strategies (e.g. habitat restoration and species reintroduction) that improve ecosystem resilience under various climate change scenarios (Frankel, 1990; McFerson, 1998; Abbas and Qaiser, 2011). Biotechnology, that is, the use of living organisms or parts thereof to manufacture or modify a product, develop microorganisms for specific uses, or improve plants or animals (Uyoh et al., 2003), has also made plant genetic resource conservation an industry priority in many developed parts of the world.
In this regard, seed storage is the most efficient and cost-effective means of ex situ plant germplasm conservation, and has been used to conserve a sizeable amount of plant biodiversity worldwide for decades (IBPGR, 1976; Engels and Engelmann, 1998). The science of seed genebanking – of storing orthodox seeds under dry, cold conditions (orthodox seeds being those that tolerate such conditions) – is still relatively basic and, given the lengths of time that the seeds
http://dx.doi.org/10.19103/AS.2020.0085.14
New technologies to improve the ex situ conservation of
plant genetic resources
© The Authors 2021. This is an open access chapter distributed under a Creative Commons Attribution 4.0 License (CC BY)
are expected to remain alive compared with the age of many well-documented collections, could still be considered somewhat experimental. Nonetheless, over recent years, we have gained greater insight into the effectiveness of the basic processes that most genebanks follow in managing their collections of orthodox seeds and, since budgets are often limited, advanced technologies are being introduced, with the tag line of improving efficiency.
However, seed storage cannot be used to conserve genetic variation of all species, due to fundamental differences in innate reproductive characteristics (specifically the ability of seeds to survive long periods of dry storage) or the need to fix and multiply favourable genetic combinations. In such cases, seed-derived embryonic tissues, somatic embryos or vegetative tissues (e.g. shoot tips) are the materials of choice for storage. Collections housed in both the developed and developing world (Thailand, Australia, United States of America, Italy, Belgium, India and China, among others) include germplasm of vegetatively propagated fruit, commercial and speciality crops, ornamentals, nuts, vegetables and wild relatives, from temperate, subtropical and tropical zones (Jenderek and Reed, 2017). Where these are clonal collections the plant material is maintained in an actively growing state using a variety of techniques for one or more of the following purposes: (1) to preserve the selected genotypes; (2) to maintain their sterile nature; (3) to produce seed unamenable to storage. While the development of cryopreservation techniques for clonal crops began as far back as the 1970s, involving controlled-rate cooling and subsequent plunging in liquid nitrogen (LN) of shoot tips (Grout et al., 1978) and buds (Sakai and Nishiyama, 1978), the last two decades has seen the development of a range of cryopreservation techniques that have been applied across numerous genera and species (Benson, 1999; Reed, 2008; González-Arnao et al., 2014), including both orthodox-seeded (desiccation tolerant) and recalcitrant-seeded (desiccation sensitive) species (Berjak et al.*,* 2011a; Jenderek and Reed, 2017).
In this chapter, we describe some of the recent applications of science and technology to improve the management of genebank collections of orthodox seeds, not least through the introduction of automation. In addition, we look at how routine cryopreservation procedures are now possible for many species, and how storage procedures for recalcitrant-seeded and vegetatively propagated species continue to evolve, in light of the challenges they present in terms of long-term ex situ germplasm conservation.
### **2 Improving the management of orthodox seeds**
# *2.1 Routine operations*
Differences in seed post-harvest physiology are reflected in the categorisation of seeds as orthodox, recalcitrant (Roberts, 1973) and intermediate (Ellis et al., 1990), based on their responses to loss of moisture and to storage at different temperatures. Orthodox seeds can tolerate drying without damage to lowmoisture contents (2–6%; Roberts, 1973; Chin and Roberts, 1980; Hong and Ellis, 1996) and their longevity, spanning perhaps many decades, is increased as the moisture content and temperature at which they are stored (within limits) are reduced (Ellis and Roberts, 1981). The core operations of a genebank with the mandate to conserve seeds of orthodox species comprise seed production, drying, cleaning, viability and health testing, packing, storage and distribution; all of these activities are documented and driven by the genebank information system (Fig. 1).
Most genebanks store seeds of every accession in both an active collection and a base collection. The active collection is generally stored at 2–4°C (medium-term storage; MTS) and it is from this collection that samples are taken for distribution. This is therefore where the bulk of the seeds are stored. The base collection is stored at a lower temperature, typically −20°C (long-term storage; LTS). Samples in the base collection are usually much smaller than the samples in the active collection, and the purpose is longterm conservation. The point of having all the accessions in two environments is to reduce costs. In theory, the cost of storing a sufficient quantity of seeds for distribution at the lower temperature is too high. This is particularly true for agricultural genebanks with high rates of distribution. In contrast, at, for example, the Millennium Seed Bank (MSB) of the Royal Botanic Gardens Kew, whose main purpose is long-term conservation of species and which has relatively few sample requests, all seeds are under LTS conditions. Other genebanks may also maintain only one storage environment, based on operational needs and costs. The United States Department of Agriculture has a network of genebanks holding the active collections for different crops, with one main base collection serving all the regional genebanks (National Research Council, 1991). Some genebanks additionally cryopreserve samples of orthodox seeds, for example, of orthodox seeds which are expected to be very short-lived, even under 'conventional' genebank storage conditions of −20°C (Davies et al., 2018; Ballesteros et al., 2021).
Testing the viability of seeds before they are placed into storage, and at regular intervals during storage, is a key 'quality control' activity that genebanks are expected to do as a matter of routine (FAO, 2014). For genebanks with accessions in both an active and a base collection, if the sample in both comes from the same seed lot, it is not necessary to check the viability of the seeds in both collections. Since the longevity of orthodox seeds improves in a predictable manner as the temperature is reduced, the seeds in the base collection should always have higher viability than the seeds from the same seed lot stored in the active collection after the same period of time. In other words, the viability of the seeds in the active collection
**Figure 1** Overview of operations in a typical seed genebank. When a sample first arrives at the genebank, it will be assessed for uniqueness and initially grown under quarantine conditions to ensure there is no disease. If there are enough seeds, a small sample (e.g. 20–30 seeds) of the incoming seeds will also be added to the seed file, for future reference and variety verification. The accession will then be multiplied to produce sufficient seeds for storage. Characterisation of the accession for morpho-agronomic traits according to defined crop descriptors may be carried out on plants used for seed multiplication, or may be an independent operation. Characterisation also includes collection of data on features of the seeds. After harvest, seeds will be cleaned and dried. They may also undergo a pest control treatment. Samples will be taken for phytosanitary (PS) testing and for an initial viability test (VT). After final equilibrium drying, seeds will
#### **Figure 1** (*Continued*)
be packed for storage in the long-term (base) and medium-term (active) collections (if appropriate). Some genebanks may additionally cryopreserve a sample of seeds. Safety duplicate samples will be sent for long-term conservation, without any management interventions, at another genebank and/or in the Svalbard Global Seed Vault. Seeds in the active collection are used for distribution (Dist.n). They are also regularly tested for viability (viability test; VT). Once the viability falls below the threshold level or the quantity of seeds remaining gets low, the accession will be regenerated and the seeds produced are processed for storage in the active collection again. A few cycles of regeneration using seeds from the active collection may occur, before seeds from the base collection should be used. After replacing the seeds in the active collection, meaning that the seeds in the two collections are not from the same seed lot, viability testing of seeds in the base collection should commence. When the viability of the seeds in the base collection falls below the threshold level (or, but theoretically unusually, the quantity of seeds remaining gets low), a new sample should be produced for the base collection. A sample of seeds from this seed lot should also be sent as a safety duplicate sample, because it is expected that the samples in the base collection and in long-term storage in another location would lose viability at a similar rate. Seeds from this same seed lot may also be used to replenish the active collection and the 'cycle' begins again. Across all the operations, data is collected and entered into the genebank information system. This data includes information on the processes themselves, such as number of seeds sown, number of plants harvested, viability test date and so on. The data examples shown represent only a fraction of all the data that would, in reality, be recorded.
will fall faster than the seeds stored in the base collection – unless there has been any divergence in the handling that compromises the initial quality or moisture content of the seeds intended for storage in the base collection. It is only necessary to start testing the viability of seeds in the base collection when the seeds in the active collection are replaced by a new seed lot (Whitehouse et al., 2020). Similarly, if seed samples are taken from the same seed lot and prepared in the same way for storage in the base collection and for sending to another genebank as a safety duplicate, it is not necessary to monitor the viability of that safety duplicate sample. Safety duplicate samples are only intended to be returned to the originating genebank, and only if that genebank cannot recover the accessions from their own samples. When the viability of the seeds in the base collection declines, the safety duplicate sample may be replaced Until now, this has been happening at low rates, although it is likely to increase, particularly for species with relatively shortlived seeds, as the viability of more samples in the base collection falls and the samples are replaced. For many genebanks, in addition to having every accession in the two collections, a very small sample of the 'most original seeds' of each accession is also placed in the seed file. The seed file is the reference sample against which comparisons can be made, for example, of regenerated material or of new material that is submitted to the genebank. Ideally, the most original sample would comprise some of the seeds originally donated or arriving at the genebank, but in some cases it will be seeds from the first round of multiplication.
Given the interdependency of operations, a robust information system is crucial to record all the data collected for each process and to ensure timely and efficient sequencing of operations. Unfortunately, data management may be overlooked as key 'infrastructure', particularly in national genebanks, where resources may be limited. Genebank information systems also store passport data and, increasingly, characterisation and evaluation data, and links to genomic data, where available. More genebanks are genotyping accessions or even, whole collections, and/or are handling genetic stocks such as mapping populations, mutant populations and recombinant inbred lines. Genebank information systems such as GRIN-Global (www.grin-global.org) also allow users to search collections and order accessions online.
# *2.2 Changing procedures*
Recommendations on how to handle seeds in genebanks were first published in 1975 (Cromarty et al., 1990) and, while there has been some evolution, the essence of those recommendations has barely changed. For example, Cromarty et al. (1990) recommended drying seeds in a drying room at 'about 15°C and 10–15% relative humidity with good air circulation'; in the IPGRI/ FAO genebank standards published in 1994, the recommendations were to dry at 10–25°C and 10–15% relative humidity (IPGRI/FAO, 1994); and in the latest version of the standards, the recommended drying environment is 5–20°C and 10–25% relative humidity (FAO, 2014). Similarly, storage temperatures and monitoring intervals remain largely unchanged, even after decades of storing seeds in genebanks. Unfortunately, because some of those original recommendations are now relatively old, it may be difficult to know how they came about; old data sets might not be archived and knowledge of the hows and wherefores not passed on to younger genebank managers/scientists. Nonetheless, we should still be considering whether genebank standards are optimal and in particular, given that the standards are intended to cover such a broad range of species, whether they are optimal for particular species.
# *2.2.1 Seed drying*
The conditions under which orthodox seeds are dried are intended to result in seeds being at a moisture content where seed longevity is optimal (regardless of whether the seeds are subsequently stored at 2–4°C or −20°C). Various combinations of temperature and humidity would allow seeds to reach the same moisture content, however, the drying treatment itself should not be detrimental to the quality of the seeds. In theory, if very wet seeds are placed at a high temperature there could be a high rate of seed ageing. Thus, the genebank standards recommend drying at a relatively cool temperature, in contrast with the higher temperatures used for drying grain. Many genebanks have a drying room that runs at 15°C and 15% relative humidity. In the case of rice, however, it has been found that drying at 15°C is not optimal for the subsequent longevity of the seeds (Crisostomo et al., 2011; Whitehouse et al., 2015, 2018a). If the rice seeds, due to the typically humid environment of rice fields, are not able to dry in situ and have a high moisture content at the time of harvest (> 16.5%), initial drying at 40–45°C followed by final equilibrium drying at 15°C and 15% relative humidity improves the subsequent longevity of the seeds significantly compared with only drying at 15°C and 15% relative humidity (Whitehouse et al., 2018a). This response is perhaps not surprising for rice: in the tropics, ambient temperature is, of course, relatively high, and rice (and other grains) is often dried on roads or flat cement areas, where temperatures in the middle of the day under bright sunlight can get quite high. A similar response was also seen for seeds of four wild rice species (Timple and Hay, 2018). As a consequence of this research, the rice genebank at the International Rice Research Institute now routinely uses two-stage drying for all the collection. Freshly harvested seeds are initially dried for 3 days in a drying room at 40°C and 30% relative humidity, after which, they are transferred for final equilibrium drying in the drying room at 15°C and 15% relative humidity. There is some evidence that drying seeds of some other species at a higher temperature than 15°C may similarly be better for the subsequent longevity of the seeds (Whitehouse et al., 2018b).
# *2.2.2 Revising monitoring intervals*
As indicated above, many genebanks still adopt the default viability testing intervals of 5 or 10 years, depending on the expected longevity in storage (FAO, 2014). Otherwise, '*Viability monitoring test intervals should be set at onethird of the time predicted for viability to fall to 85% of initial viability or lower depending on the species or specific accessions, but no longer than 40 years*' (FAO, 2014). The recommended way to predict the viability period is to use the Ellis and Roberts (1980) viability equation, using parameters determined from experiments in which seeds are stored at a range of temperatures and moisture contents (e.g. Hay et al., 2003). This can be done using the seed viability constants menu in the Seed Information Database (SID; Royal Botanic Gardens Kew, 2020), but is only possible for a small number of species. Furthermore, as discussed by Hay and Whitehouse (2017), the time for viability to fall to 85% of the initial viability is perhaps not an ideal standard, since it depends on an estimate of initial viability, and because it means that the absolute threshold value could potentially vary among seed lots (i.e. of seed lots of the same or different accessions). For example, if the initial viability ranges between 85% and 100%, the monitoring intervals would be calculated based on the time for viability to fall anywhere between 72% and 85%. Further, it is known, based on the typical shape of the seed survival curve (the curve describing the loss of viability during storage), that once the viability of a seed lot has fallen to 72%, the rate of percentage viability loss is nearing its peak. In practice, most genebanks still use set-standard monitoring intervals, at the species level at least, and have a fixed viability threshold (e.g. 85%), rather than setting seed lot-specific test intervals and thresholds. Nonetheless, in future, there could be a more dynamic approach for setting viability-monitoring intervals (Whitehouse et al., 2020).
A more dynamic, flexible approach will probably only emerge as we understand more about the real performance of seed lots in genebanks. Some genebanks have already analysed their historical viability monitoring data, using a variety of approaches to model the longevity of seed lots, within species or species groups and in some cases, to formulate more efficient monitoring intervals (e.g. Walters et al., 2005; van Treuren et al., 2013; Hay et al., 2015; Ellis et al., 2018, 2019; Yamasaki et al., 2020). In some cases, the conclusions drawn from such studies, while contributing to scientific knowledge, are only really relevant to the genebank where the data was collected, due to crop focus or more likely, regeneration and processing protocols, and storage conditions that are perhaps unique to that genebank, even if they are more-or-less consistent with the Genebank Standards (which cannot be species-prescriptive because they are meant to cover so many species). In future, there may be more metaanalyses, using viability data for the same crop from different genebanks. In particular, this could help verify the applicability of the Ellis-Roberts viability equations, particularly at low temperatures for which data is still lacking (Pritchard and Dickie, 2003).
There have also been some advances in terms of understanding the molecular basis of seed longevity for genebank accessions, which raises the possibility of using genotype data to predict the relative longevity of seeds in genebank storage and hence set appropriate monitoring intervals (Lee et al., 2019). Screening seed lots for longevity under experimental conditions to get an initial measure of relative longevity could also help identify which seed lots to test first (Davies et al., 2016; Hay and Whitehouse, 2017), perhaps in particular if used in conjunction with genotype data. Taking this idea further, it has been suggested that a more efficient strategy for genebanks would be to preclude the need to do viability monitoring tests, by taking other measurements (ideally simple, cheap and non-destructive tests) that are predictive of the extent to which a seed lot has aged (Fu et al., 2015). However, this still seems to be somewhat futuristic in practice, and may only ever be feasible for large agricultural genebanks with many accessions of a single or few crops.
**Figure 2** Barcodes are used to track samples through genebank operations: seed drying (a) and transplanting (b) at the International Center for Tropical Agriculture (CIAT) in Colombia; and germination testing and scoring (c, d) at the International Institute of Tropical Agriculture (IITA) in Nigeria. All photos taken by the first author (Fiona R. Hay).
# *2.2.3 Electronic data collection*
The use of barcodes to label the packets of seeds going into genebank storage has been a common practice for some years now (Rao et al., 2006). Indeed, barcodes are now used to track samples through many, if not all, genebank operations, from the field to storage, including during characterisation/ evaluation activities and distribution (Fig. 2). This has also made it possible to collect data electronically, even in the field: the barcode can be scanned and then the relevant data entered. To further minimise the risk of wrong data entry, genebank staff may use a portable tablet or similar device with software that is customised to show, for example, value options with pictures for a particular characterisation trait. In theory, not only is this more accurate, but also improves efficiency since staff do not have to manually enter data from paper score sheets after data collection.
# *2.2.4 Improving viability testing*
Modifying and improving germination procedures used in viability monitoring should be a constant process in genebanks, particularly those conserving wild species diversity. Such species, compared with major crop species, are likely to have some sort of dormancy mechanism which prevents germination when moisture, temperature and light requirements are met. It is also more likely that dormancy-breaking procedures have not been documented and/or widely validated for such species. Despite the need for reliability in viability testing and to provide germplasm users with seed germination advice, research in this area has been under-resourced, particularly at crop genebanks where wild species have historically perhaps received less focus in the context of managing germplasm. Seed banks entirely focussed on conserving wild species, in contrast, have invested more resources in determining optimum germination protocols, as documented, for example, in the Seed Information Database (Royal Botanic Gardens Kew, 2020). There are other comprehensive resources that would be helpful to any genebank needing to improve their germination protocols (e.g. Baskin and Baskin, 2014; ISTA, 2020).
The preferred germination-test sample size in official testing is 400 seeds, sown as four replicates of 100 (ISTA, 2020). This is far too many seeds than most genebanks would like to use in a viability monitoring test, though the actual number tested varies among genebanks depending on species, quantity of seeds available, facilities and other logistical constraints. As also discussed elsewhere (Hay and Whitehouse, 2017), there are other ways in which the number of seeds used overall for viability monitoring might be reduced, not just by having more knowledge of longevity in storage, but also, for example, by following a sequential monitoring scheme (Ellis et al., 1985). A recent study across three international genebanks, involving 111 accessions from 11 species, confirmed that sequential sampling is an efficient alternative to the fixed size sampling method for making decisions based on viability for tropical forage species, although for three *Leucaena* species, there was no advantage in using sequential sampling to save seeds in viability monitoring (A. Sartie, L. Santos and Z. Kinyanjui, *pers. comm*.).
While 'viability' is most often assessed using a germination test, there has long been an interest in developing a non-destructive test to predict the viability of a seed lot. Research in this area has considered, for example, the nature and quantity of volatiles released by seeds during storage (Colville et al., 2012) and molecular markers (Boniecka et al., 2019; Fleming et al., 2019). However, while these studies certainly cast light on the process of seed deterioration in storage, there is still a lot of research needed before we get close to a practical nondestructive predictive test of seed viability during genebank storage (reviewed by Fu et al., 2015).
#### *2.3 Automating processes*
Some genebanks, notably those of the international genebanks of the CGIAR and a few national genebanks, have very large collections. For example, the genebank at the International Rice Research Institute (IRRI) in the Philippines is the largest single-crop collection with more than 132 000 accessions (https:// www.genesys-pgr.org/). Staff at such genebanks may work as if they were
**Figure 3** Images of the robotic storage and retrieval system in the long-term storage room of the National Agrobiodiversity Center, Rural Development Administration, Republic of Korea. Images provided by the National Agrobiodiversity Center.
working in a factory line, with groups of staff focussing on one particular activity at any one time and often for quite long periods of time. In such genebanks, that process thousands of seed samples of one or just a few crops each year, automating some of those processes is more likely to be feasible, compared with a genebank that is perhaps processing one species for one week, and a completely unrelated species, the next. The genebank storage rooms can thence be considered the warehouses for the factory and also be managed in a similar way.
# *2.3.1 Robotic storage and retrieval*
A few genebanks around the world, for example, the Genetic Resources Center of Japan's National Agriculture and Food Research Organization and the National Agrobiodiversity Center in the Republic of Korea (Fig. 3), have incorporated warehouse management-type technology to manage their collections. All the placing and retrieval of samples is handled by a robot, so staff rarely enter the stores themselves. There are a number of reasons why this could be considered advantageous. The long-term storage environment, at −20°C, is not a pleasant working environment for staff and could even present a health hazard to some people, even if they wear insulated clothing. Therefore, restricting access and limiting the length of time anyone is in that room, might reduce any risk associated with working at such a low temperature. On the other hand, it should not be necessary to enter this storage room very often, as its intended use is long-term conservation. The active collection is stored at a higher temperature (2–4°C), which is a more tolerable temperature, nonetheless, a robotic system will increase security and ensure correct placement of samples. It may also mean that there is less exchange of air so it is easier to maintain the correct temperature and energy is not being wasted. This is perhaps particularly relevant in hot and humid environments. Of course, genebanks that do not install a robot system, will still make sure that the location of accessions in both stores (e.g. shelf number, tray number) is recorded in the genebank information system.
# *2.3.2 Seed phenotyping*
Analysis of images captured by different types of cameras has many potential applications in seed testing (Dell'Aquila, 2009; Boelt et al., 2018). One way in
**Figure 4** (a) Samples from the genebank's physical seed file at the International Rice Research Institute (IRRI); (b) selecting the training-set seeds from the bulk prior to automated sorting (image used with permission of N. Ruaraidh Sackville Hamilton); (c) the custom seed sorter at IRRI. (a) and (c) were taken by the first author (Fiona R. Hay).
which image analysis could be applied to improve genebank operations is by using it as a virtual seed file (Edberg Hansen et al., 2016). Most agricultural genebanks keep a small sample of seeds of perhaps the material originally received at the genebank, or otherwise, of the first cycle of seed multiplication, for use as a reference against which new harvests can be compared. This is called the seed file (Fig. 4A). This seed file does not have a physical backup, and if any seed file samples are lost, they would have to be replaced by the next available 'most original sample' (i.e. from the first regeneration cycle, if available, rather than the second or third). The seed file is not just used to verify that harvested seeds appear to be of the same accession, but they are also used, for example, to check incoming material to see whether the sample is a potential varietal mixture with different seed phenotypes and/or to determine whether it might be a variety that is already represented in the genebank. However, use of the seed file relies on having trained personnel who are able to detect subtle differences between seeds. Some genebanks already capture images of the seeds of different accessions (as well as other parts of the plants), to use internally and/or to make available to germplasm browsers on their online ordering portal, and there is a lot of potential to analyse such images and make the validation/checking process more objective. Edberg Hansen et al. (2016) proposed the use of multispectral imaging since it offers the possibility of identifying and measuring some of the more subtle characteristics of seeds, which might not be apparent in images from an ordinary digital RGB camera.
# *2.3.3 Seed sorting*
Most seed lots intended for genebank storage will go through a cleaning process, though the extent and method of cleaning may vary. There are various types of seed cleaning equipment developed for the seed industry, including seed blowers, sieves, brushes and colour sorters, which can be used to remove non-seed material such as plant stems or dispersal structures. The choice of method will very much depend on the seeds being cleaned and the quantity of seeds to be cleaned. For very small seed lots, the amount of time it takes to clean the equipment between seed lots may mean that it is not practical to use such machines. Some genebanks do a considerable amount of hand sorting. For many years, all the seeds intended for storage in the rice genebank at the International Rice Research Institute (IRRI) were entirely sorted by hand, removing off-types, immature (green), damaged or diseased seeds. This used to be a year-round activity and as a consequence, for some seed lots, the length of time between harvest and final storage was many months. Although the seeds were kept in the drying room until they could be cleaned, this could nonetheless impact the initial viability and storage potential of the seeds when put into the genebank stores. The IRRI genebank now has a unique, purpose-built seed sorter that selects seeds by individual-seed image analysis, making sure the colour and shape parameters of each seed matches that of a training set (Fig. 4B, C). This automated seed sorting is considered a first-step in the cleaning process, and the 'selected' seeds are still manually sorted, but introducing this automation has increased the throughput of the seed cleaning process.
# *2.3.4 Germination scoring*
Automated scoring of the germination process has attracted quite a lot of interest in the seed testing community over the last decade or so, to increase the number of samples that can be tested for germination in a short period of time and/or to get more information on the rate of germination as a measure of seed lot vigour (Joosen et al., 2010; Harper and Long, 2011; Demilly et al., 2015). Some genebanks test the germination of thousands of samples every year, as part of the routine viability monitoring. For genebanks handling seeds of many different species, germination protocols (pretreatment, germination method and germination temperature) are likely to vary from one species to the next, and there is little opportunity to automate the scoring process. However, for large crop genebanks, testing one or a few species following the same protocols for each accession within a crop, automated scoring of germination may be feasible. Various systems are available, from fully automated robotic systems in a controlled-environment room with the capacity for hundreds of tests at any one time, to semi-automated systems in which tests are manually moved from the germination environment to the camera. The software for these systems use image analysis to determine how many seeds have germinated at each observation time and, once the test is finished, calculates various parameters, for example, final germination percentage and mean germination time. From the genebank perspective, the parameter of most interest is the final germination percentage at the end of the germination test, but other parameters may provide insight into the seed-ageing process (since vigour is expected to decline before there is loss of viability) and/or may be of interest to breeders who are trying to enhance the seed vigour of varieties, so that they can, for example, still give good crop establishment in marginal environments.
# **3 Improving the management of recalcitrant-seeded and vegetatively propagated species**
# *3.1 Routine operations: recalcitrant seeds and vegetatively propagated species*
Recalcitrant seeds are sensitive to desiccation (Roberts, 1973), freezing (reviewed by Walters et al., 2008) and very often chilling, in some cases at temperatures as high as 15°C (Berjak et al., 1995). This precludes their maintenance under conventional orthodox seed storage conditions (e.g. King and Roberts, 1980; Farrant et al., 1989). Short- to medium-term storage methods and long-term storage via cryopreservation of seed-derived explants represent some of the storage options employed for such species over the last few decades. Similarly, alternative germplasm conservation strategies, such as storage of actively growing cultures, minimal-growth storage and cryopreservation, have been adopted for species that do not produce seeds (e.g. banana), root and tuber vegetables, clonally propagated crops, and where the unique true-to-type genomic constitution of a cultivar needs to be maintained.
For recalcitrant seeds, short- to medium-term storage has involved maintenance at water contents as close to that at shedding, and at ambient or slightly lower temperatures, and this is defined as 'hydrated-storage' (e.g. Berjak et al., 1989). However, under these conditions, recalcitrant seeds still initiate germination-associated events, culminating in seed death (Berjak et al., 1989; Pammenter et al., 1994; Chandra et al., 2019). Storage at temperatures lower than ambient has been shown to postpone the onset of germination in recalcitrant seeds by slowing down the metabolic rate (e.g. Pritchard et al., 1995), provided the seeds are not chilling-sensitive. However, even then, storage longevity generally ranges from a few weeks to months e.g. for *Scadoxus membranaceus* and *Landolphia kirkii* (Farrant et al., 1989), various amaryllid species (Sershen et al., 2008) and *Madhuca latifolia* (Chandra et al., 2019). The lifespan of recalcitrant seeds in hydrated storage is often further curtailed by the proliferation of a spectrum of fungi (Mycock and Berjak, 1990; Sutherland et al., 2002), even when seeds are treated with fungicidal agents (Mycock and Berjak, 1995; Moothoo-Padayachie et al., 2018).
In vitro storage, that is, the use of tissue culture, by which cells, tissues or organs are excised from parent plants, decontaminated and then transferred to artificial growth media in vitro (Krøgstrup et al., 1992; George, 1993; Mandal et al., 2000), represents the major short- to medium-term germplasm storage option for vegetatively propagated species. A wide range of explants have been used (e.g. buds, cuttings, seeds, shoot apices and leaves) based on the fact that media can be manipulated to produce different cultures such as unorganised, undifferentiated callus or organised tissues and organs that can be converted into plantlets in the regeneration phase. Variations to impose minimal growth include reduction in nutrient (e.g. Schnapp and Preece, 1986) and/or sucrose concentration (e.g. Kartha and Engelmann, 1994) in the growth medium, alterations to culture medium osmotic potential (using osmotica such as sucrose, sorbitol, ribose and mannitol (e.g. Zandvoort et al., 1994)), the addition of growth retardants such as abscisic acid (ABA) (e.g. Jarret and Gawel, 1991; Taylor et al., 1996), and placing explants in a controlled atmosphere or beneath a liquid medium or mineral oil (Paunescu, 2009). Lowering the partial pressure of oxygen with temperature (between 0°C and 10°C (e.g. Blakesley et al., 1996)) or light (e.g. Grout, 1995) below optimum in the culture environment can also limit in vitro growth. However, irrespective of whether germplasm is stored in vitro as actively growing cultures (e.g. Krøgstrup et al., 1992) or as minimal-growth cultures (e.g. Schnapp and Preece, 1986), such storage will at some stage require transfer of material onto new media (Krøgstrup et al., 1992; Mandal et al., 2000; Mycock et al., 2004), introducing the risk of contamination. Surviving, uncontaminated material can potentially be rapidly micropropagated to bulk up reserves (Mandal et al., 2000) but such clonal propagation, apart from being labour intensive and expensive, limits biological diversity within the collection and can impose selection pressures and environmental stresses, resulting in plants with genetic modifications (Staritsky, 1997). This variation arises from somaclonal variation during culture (Panis and Lambardi, 2006), but the use of organised systems such as embryos, meristems and shoot tips can reduce this risk (Engelmann, 1997; Mandal et al., 2000).
Long-term germplasm storage of recalcitrant-seeded and vegetatively propagated species is achieved via cryopreservation which involves the cooling of biological material to, and subsequent storage at, cryogenic temperatures, typically in liquid nitrogen (LN; −196°C), or its vapour (at approximately -150°C). Less ideally, material can be stored at some other temperature below −80°C (Finkel and Ulrich, 1983; Berjak et al., 1999a). Cryopreservation is regarded as the ultimate long-term storage approach since it is believed to arrest metabolic activity and deterioration, thus minimising, if not precluding, genetic changes (Krøgstrup et al., 1992; reviewed by Harding, 2004). With the exception of mature orthodox seeds (Pritchard and Nadarajan, 2008) and certain varieties of pollen (Ganeshan et al., 2008) and spores (Ingram and Bartels, 1996), biological tissues almost always contain considerable cellular water. This implies that successful cryopreservation of hydrated explants from recalcitrant seeds and vegetatively propagated species is best achieved when lethal intracellular ice-crystal formation is mitigated, as this can cause irreversible intracellular damage (Wesley-Smith et al., 1992). While cryopreservation is generally considered in terms of water's liquid and solid (ice) phases, it is also possible to cryopreserve plant material by inducing the process of vitrification, that is, the transition from liquid to glass phase without ice-crystal formation (Sakai, 2004). The phenomenon of vitrification has allowed for the development of 'ice-free' cryopreservation, which has been extensively applied to plant tissues (Sakai et al., 1992) of a number of vegetatively propagated species. However, freezing of plant tissues that inherently have high water content will inevitably involve the conversion of at least some of the water to ice and this is what precludes this approach in recalcitrant seed-derived explants of a number of species (Wesley-Smith et al., 1992).
Even though cooling and dehydration are the greatest sources of failure, and under other circumstances, the greatest contributors to postcryo preservation survival, the success or failure of any plant cryopreservation protocol is a consequence of optimisation of all the manipulations involved in the preparation of the tissue for cooling, and all the steps involved in the recovery of that tissue after cooling (Berjak et al., 1999b; Pammenter and Berjak, 2014). These manipulations generally include variable combinations of the following (Funnekotter et al., 2017; Kaczmarczyk et al., 2012):
Samples that have been used for cryopreservation over the last three decades or so include buds, shoots, meristems, cell cultures, protoplast cultures, anthers, pollen, somatic and zygotic embryos, embryonic axes, callus, and whole seeds, if they are sufficiently small (Benson, 2008a). Success has been achieved in cryopreserving whole seeds (e.g. Hor et al., 1990; Kioko et al., 2003), shoot tips (Varghese et al., 2009), embryonic axes and zygotic embryos (Berjak et al., 2011b; Sershen et al., 2012) and somatic embryos of recalcitrant-seeded species (e.g. Mycock and Berjak, 1993). However, the number of successful protocols developed for embryonic axes and zygotic embryos from recalcitrant seeds is extremely low in the context of the number of species within this seed category (Subbiah et al., 2019).
A number of protocols for vegetative explants incorporate cold acclimation to pre-condition samples for exposure to cryogenic temperatures (reviewed by Reed, 1996, 2008; Benson, 2008b). Cold acclimation is generally induced in the laboratory by exposing explants to low in vitro growth temperatures (1–6°C) generally in combination with shortened day-length and/or high-sugar pretreatments (Reed, 1996, 2008). Most of these protocols have also been based on what are termed classical methods (reviewed by Engelmann, 1997) which generally involve explant chemical cryoprotection, followed by slow cooling (0.5–2.0°C min−1) down to −30°C to −40°C, or even −60°C (Krøgstrup et al., 1992). This controlled slow cooling (step 1) is said to encourage the formation of extracellular ice, progressively dehydrating the cells, as intracellular water is lost to exterior ice nucleation sites. This step is usually followed by immersion in LN (step 2). When freeze-induced dehydration during step 1 is too intense, various damaging events associated with the concentration of intracellular salts and changes in cellular membranes are possible (Mazur, 1990; Pritchard et al., 1995a). Also, some of the cells may fail to reach the optimum intracellular concentration and upon super-cooling undergo lethal intracellular ice-crystal formation (Mazur, 1990). Thus, while slow cooling may retain the integrity of individual cells, it has been shown to be less efficient at maintaining the tissue integrity required for survival of complex tissues, that is, meristems and embryos (Panis and Lambardi, 2006). Nevertheless, this 'twostep' cooling method, regarded as the first standard protocol developed for hydrated plant tissue (e.g. Withers and King, 1980), is still used for the cooling of undifferentiated culture systems (e.g. cell suspensions and calli (Withers and Engelmann, 1997)), and even differentiated structures such as the shoot apices of cold-tolerant species (e.g. Reed and Chang, 1997) and in isolated cases, excised embryonic axes/zygotic embryos from recalcitrant seeds (Mroginski et al., 2008).
# *3.2 Changing procedures: recalcitrant seeds and vegetatively propagated species*
While minimal-growth storage has been successfully applied to explants from recalcitrant seeds (e.g. Chin, 1996; Sershen et al., 2008), recent reports of its use for the storage of such germplasm are scarce. Its continued use for the storage of somatic embryos (Hassan, 2017) and vegetative tissues is, however, still prevalent in the literature (Chauhan et al., 2019). In this regard, temperature reduction appears to be the most widely applied procedure in slow growth preservation, but it is now clear that temperature requirements vary from species to species, possibly based on the agroclimatic conditions in which a particular species occurs (Thakur et al., 2015). A more recent development in the design of minimal growth storage protocols involves the use of artificial seeds, produced by encapsulating propagules such as shoot buds or somatic embryos in a synthetic matrix, and this approach has been used successfully for the medium-term storage of a variety of plant species and propagule types (Cruz-Cruz et al., 2013).
There is also an increasing emphasis being placed on monitoring explants during storage, for example, changes in sugar content (total and reducing and non-reducing) due to sucrose-imposed stress in order to understand the effect of sucrose concentrations during slow growth conditions on survival and regeneration rate (El-Dawayati et al., 2018). Accommodating the interactive effects of light and ABA, and assessing relative nuclear DNA content in regenerants to ensure retention of ploidy level have also become important considerations (El-Dawayati et al., 2018).
The short- to medium-term storage of recalcitrant seeds via hydrated storage is no longer a common practice, but there are recent reports (Moothoo-Padayachie et al., 2016; Chandra et al., 2019) that the exogenous application of reactive oxygen species blocking agents (e.g. diphenyleneiodonium (DPI) and dimethyl thiourea (DMTU)) can be used to extend the storage lifespan of recalcitrant seeds (by days to weeks). In terms of long-term storage of recalcitrant seeds, in recent years it has become increasingly apparent that the success of cryopreservation protocols for their zygotic germplasm depends on the optimisation of cooling rates in parallel with explant moisture content, to eliminate or at least minimise intracellular ice crystal formation (Wesley-Smith et al., 2004). Success in this regard has, however, been hampered by (a) lethal freezing damage occurring when hydrated seeds/embryos/axes are exposed to LN (Wesley-Smith et al., 1992; Berjak et al., 1999b); and (b) drying to water contents precluding ice formation to the extent of desiccation damage, which generally culminates in loss of viability (Walters et al., 2008). In contrast, pollen (reviewed by Ganeshan et al., 2008) and seeds and somatic embryos of most desiccation-tolerant species (reviewed by Pritchard and Nadarajan, 2008) appear to be highly amenable to cryopreservation, shifting the focus for successful cryopreservation of recalcitrant seed germplasm from freezing tolerance to dehydration tolerance (Panis and Lambardi, 2006). The large-scale, routine application of cryopreservation to recalcitrant-seeded species is very limited globally, with many protocols remaining unpublished, but a recent study has validated the use of embryonic axes for cryopreservation of this group of species by showing that viability can be retained for decades (Ballesteros and Pence, 2019). For some recalcitrant-seeded species, such as cocoa and *Avicennia marina*, however, seed-derived germplasm will never be suitable for cryopreservation based on high levels of microbial contamination and/or the absence of suitably small zygotic explants. In these cases storage of alternate explants (e.g. shoot tips and meristems) generated via tissue culture may be more suitable.
While classical cryopreservation protocols used freeze-induced dehydration, modern protocols are predominantly vitrification-based (e.g. Fu et al., 1990; reviewed by Engelmann, 1999). In the latter case, cell dehydration to increase cytoplasmic viscosity precedes cooling, promotes the formation of glasses and avoids the factors that encourage ice-crystal formation – before exposure to the cryogen (e.g. Stanwood, 1985; Wesley-Smith et al., 1992; reviewed by Walters et al., 2008). Cooling rates typically used to cool embryonic axes/zygotic embryos in these modern protocols, range from 10°C min−1 (e.g. Vertucci, 1989; Sershen et al., 2007) to hundreds of °C s−1 (e.g. Wesley-Smith et al., 2004; Walters et al., 2002). Additionally, most recent plant cryopreservation studies involve one of the following vitrification-based procedures: pre-growth, dehydration, pre-growth-dehydration, encapsulationdehydration, vitrification, encapsulation-vitrification, droplet-vitrification (González-Arnao et al., 2008) and, over the last decade or so, the novel cryoplate procedure (Yamamoto et al., 2011).
Pre-growth involves culturing explants on cryoprotectants, followed by rapid, direct immersion in LN (Engelmann, 2004), while for dehydration explants are usually partially dehydrated in a laminar flow or above a bed of silica gel before immersion in LN. Pre-growth-dehydration involves the combination of both these procedures. These methods are now widely applied for cryopreserving meristems and polyembryonic cultures, small seeds and seed zygotic embryos or embryonic axes (González-Arnao et al., 2008). However, reports on their successful application to recalcitrant seed embryonic explants are rare. The encapsulation-dehydration technique is based on the methods used for artificial seed production, and involves encapsulating explants in alginate beads, followed by pre-growth in sucroseenriched liquid medium and then partial desiccation in a laminar flow or using silica gel. The explants, often shoot apices (González-Arnao and Engelmann, 2006), are usually dried to a water content around 20% (fresh mass basis), the effects of which are minimised by being encapsulated, and they are then rapidly immersed in LN.
Vitrification techniques have remained relatively unchanged over the last few decades and involve the immersion of explants (most often somatic embryos, apices and cell suspensions) in cryoprotective solution (loading), chemical dehydration in highly concentrated plant vitrification solutions (PVS), followed by rapid cooling and rewarming, and finally removal (unloading) of cryoprotectants before in vitro recovery (Sakai and Engelmann, 2007; Sakai et al., 2008). Encapsulation-dehydration and vitrification procedures are sometimes combined (encapsulation-vitrification) and involve encapsulating explants before cryoprotection and cooling. The droplet vitrification technique (Kartha et al.*,* 1982) is also based on the vitrification procedure and is still in use; samples are loaded, dehydrated with vitrification solutions and then placed within small droplets of these solutions on aluminium foil, which are then immersed with the samples in LN.
The cryo-plate procedure (Yamamoto et al., 2011) is one of the most revolutionary technical developments in plant cryopreservation over the last decade and combines the droplet-vitrification and encapsulation-dehydration techniques. The methods involve using a thin layer of calcium alginate to attach shoot tips to an aluminium cryo-plate. The explants are then loaded, dehydrated with PVS and then cooled in LN by direct immersion of the cryo-plates (Yamamoto et al., 2011, 2012). It is evident that the commonalities between both these relatively recent cryopreservation techniques, droplet-vitrification and cryo-plate, are the rapid cooling and warming rates (achieved by immersion in a sucrose-enriched medium at ambient temperature) compared with other vitrification-based procedures. These rapid cooling and warming rates are a consequence of the fact that the aluminium contact surfaces have a very high thermal conductivity and this increases the probability of vitrification during cooling and the avoidance of devitrification during subsequent warming. This may explain why vitrification-based protocols have been shown to improve post-cryo preservation survival in complex organs (e.g. shoot-tips) of species that responded poorly to classical protocols (Panis, 1995; González-Arnao et al., 2003).
# **4 Conclusion and future trends**
Over the last 50–60 years since many of the first national and international genebanks were established, the science of conventional seed banking has barely changed and it seems unlikely to change drastically in the near future, not least since many genebanks have, to a seemingly large extent, effectively conserved and distributed seed germplasm. Despite that apparent success, it is difficult to know the extent of genetic erosion that might have occurred since the original sample arrived at the genebank, through cycles of regeneration or loss of alleles (due to seed death) during storage (Fu, 2017). There has been concern that seeds of many species, even though desiccation tolerant, are perhaps 'minimally orthodox' in that their lifespan in conventional genebank storage may not extend to many decades. Colville and Pritchard (2019) recently published a meta-analysis of seed longevity data that suggested that many more species have relatively short seed lifespans, compared with the extreme longevity reported for few species.
The number of seed accessions in genebank storage continues to increase, perhaps in response to the identification of gaps in existing collections (as discussed elsewhere in this book), but also due to changing priorities and advances in plant science. Many genebanks now conserve genetic stocks such as recombinant inbred lines, MAGIC populations and similar, and provide direct or indirect access to genomic information. A number of genebanks have also started or plan to genotype all or subsets of their collection (as discussed by McCouch et al., 2012). Genetic markers may also be used to, for example, identify duplicates within and among collections or to track accessions through storage and cycles of regeneration, although the cost and additional logistics required for such tracking is probably unworkable for most genebanks currently (van Treuren and van Hintum, 2014).
Large genebanks, particularly those conserving only one or two crops, process, perhaps, many thousands of new seed lots every year. As such, by necessity, they often adopt a factory-like approach to manage operations, with staff trained for a particular activity and the seed lot then moved to the next position/task. In contrast, smaller genebanks, or seed banks conserving diverse species, particularly wild species, may take a different approach and have the same member(s) of staff follow the whole cycle for a particular seed lot. Whichever strategy is in place, it is beneficial for the genebank and for the wider seed conservation community, for there to be regular evaluation of procedures and active innovation where improvements are needed or offer advantages, for example, in terms of accuracy, consistency and/or efficiency. Related to this, the documentation of standard operating procedures (SOPs) is expected to be revised as procedures are adapted, as part of the continuous improvement principle of a genebank quality management system (as discussed elsewhere in this book).
Some genebanks have a research team responsible for developing new procedures, technologies and understanding, although in general, this is not something that has been prioritised at most genebanks. Nonetheless, it is helpful for genebanks to follow the seed science/testing and plant phenotyping communities to see whether emerging technologies could be applied in the genebank context (Whitehouse et al., 2020). As with many industries, more genebanks may have robots in the future. However, there will always be a need for people with an understanding of the value and uniqueness of germplasm samples, not least since it is for that very uniqueness and the underlying genes, that we conserve our species and agrobiodiversity for future generations.
The last two to three decades have seen significant progress being made in terms of conserving non-orthodox seeded and vegetatively propagated plant species from temperate, tropical and subtropical zones. This has been aided by the development of in vitro seed germination, zygotic embryo and callus culture, somatic embryogenesis and micropropagation systems for many of these species (e.g. a collection of Brazilian species (Pilatti et al., 2011)). The development of these in vitro systems has enabled short- and medium-term conservation for a number of these species, but successful cryopreservation protocols are still elusive for the vast majority. Nevertheless, there are presently collections in cryobanks in close to 15 countries, based on a list published by Cruz-Cruz et al. (2013) and our consultation with various specialists in the field. These collections, which include callus, pollen, shoot tips, dormant buds, seeds and embryogenic cell lines for vegetatively propagated species (Cruz-Cruz et al., 2013), have achieved significant species coverage, with cryopreservation protocols being established for root and tubers, ornamentals, crops and fruit trees of temperate and tropical origin (Engelmann, 2000; Benelli et al., 2013). As these collections expand and the production of clones obtained from elite genotypes, unique/important cell lines and genetically transformed material increases, genotyping collections in cryobanks to avoid duplication is going to become increasingly important. A recent review by Wang et al. (2020), for example, reports on the use of stem disc-bearing adventitious buds, small leaf square-bearing adventitious buds, rhizome buds and microtubers as novel propagule types and hence, possible new alternative explants for cryopreservation. Cryobanks of the future are therefore likely to house much more diverse germplasm than at present.
While the latest two cryopreservation techniques, namely, dropletvitrification and cryo-plate, have proven to be more beneficial for cryopreserving complex organs (e.g. shoot-tips) than classical protocols, of tropical species in particular, successful cryopreservation of explants from recalcitrant seeds remains challenging for the vast majority of species of tropical and temperate origin (Ballesteros et al., 2021). Large-scale, routine application of cryopreservation is therefore still very limited in comparison with conventional low-temperature seed storage. However, the benefits of cryotherapy in eliminating viruses in explants from vegetatively propagated species (e.g. sugarcane (González-Arnao et al., 2020)) increased production of artificial seeds in breeding programmes for non-orthodox or non-seed producing plants (Ravi and Anand, 2012), and more examples of orthodox species with poor longevity in conventional seed/genebank storage (e.g. Ali et al., 2007; Mondoni et al., 2011; Davies et al., 2018) may encourage practitioners to integrate cryopreservation into existing plant biodiversity conservation procedures, whether or not these facilities focus on vegetatively propagated, orthodox- or recalcitrant-seeded species.
# **5 Where to look for further information**
# **6 References**
Abbas, H. A. I. D. E. R. and Qaiser, M. (2011). *Ruellia linearibracteolata*: conservation assessment and strategies to avoid extirpation, *Pak. J. Bot.* 43(5), 2351–2357.
Ali, N., Probert, R., Hay, F., Davies, H. and Stuppy, W. (2007). Post-dispersal embryo growth and acquisition of desiccation tolerance in *Anemone nemorosa* L. seeds, *Seed Sci. Res.* 17(3), 155–163.
and molecular markers of deterioration advancement in seeds of oilseed rape (*Brassica napus* L.), *Ind. Crop. Prod.* 130, 478–490.
Published by Burleigh Dodds Science Publishing Limited, 2021 | doab | 2025-04-07T03:56:58.462862 | 3-3-2023 05:26 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "00274f8d-5b4d-46b7-be49-66627b48ea8d",
"url": "https://library.oapen.org/bitstream/20.500.12657/61476/1/9781786768186_web.pdf||https://library.oapen.org/bitstream/20.500.12657/61476/1/9781786768186_web.pdf||https://library.oapen.org/bitstream/20.500.12657/61476/1/9781786768186_web.pdf",
"author": "Hay, Fiona R.||Sershen, Dr",
"title": "New technologies to improve the ex situ conservation of plant genetic resources",
"publisher": "Burleigh Dodds Science Publishing",
"isbn": "",
"section_idx": 0
} |
00315389-ac15-4185-b49d-fa76bff78a40.0 | # **Target and Non-Target Approaches for Food Authenticity and Traceability**
Edited by Joana S. Amaral Printed Edition of the Special Issue Published in *Foods*
www.mdpi.com/journal/foods
## **Target and Non-Target Approaches for Food Authenticity and Traceability**
## **Target and Non-Target Approaches for Food Authenticity and Traceability**
Editor
**Joana S. Amaral**
MDPI ' Basel ' Beijing ' Wuhan ' Barcelona ' Belgrade ' Manchester ' Tokyo ' Cluj ' Tianjin
*Editor* Joana S. Amaral Chemical and Biological Technology Polytechnic Institute of Braganc¸a Braganc¸a Portugal
*Editorial Office* MDPI St. Alban-Anlage 66 4052 Basel, Switzerland
This is a reprint of articles from the Special Issue published online in the open access journal *Foods* (ISSN 2304-8158) (available at: www.mdpi.com/journal/foods/special issues/Food Authenticity Traceability).
For citation purposes, cite each article independently as indicated on the article page online and as indicated below:
LastName, A.A.; LastName, B.B.; LastName, C.C. Article Title. *Journal Name* **Year**, *Volume Number*, Page Range.
**ISBN 978-3-0365-5458-7 (Hbk) ISBN 978-3-0365-5457-0 (PDF)**
© 2022 by the authors. Articles in this book are Open Access and distributed under the Creative Commons Attribution (CC BY) license, which allows users to download, copy and build upon published articles, as long as the author and publisher are properly credited, which ensures maximum dissemination and a wider impact of our publications.
The book as a whole is distributed by MDPI under the terms and conditions of the Creative Commons license CC BY-NC-ND.
## **Contents**
Reprinted from: *Foods* **2019**, *8*, 537, doi:10.3390/foods8110537 . . . . . . . . . . . . . . . . . . . . . **117**
## **About the Editor**
#### **Joana S. Amaral**
Joana S. Amaral achieved her degree in Pharmaceutical Sciences at the University of Porto, Portugal, and completed her Ph.D. in Food Chemistry and Nutrition in 2006 at the same institution. She is a Professor at the Polytechnic Institute of Braganc¸a and research member of the Mountain Research Centre (CIMO). She was President of the Food Chemistry Division of the Portuguese Chemical Society (2009–2012) and is currently the Chair of the Food Chemistry Division of the European Chemical Society (EuChemS). Throughout her scientific career, she has been the principal investigator of 6 national and international projects, a member of 26 projects and a COST action, and published over 100 peer-reviewed papers and book chapters (H-index 33). Since 2020, Prof. Amaral is Editor-in-Chief of the section Food Physics and (Bio)Chemistry of the journal *Foods*. Her main line of research focuses on food authentication using DNA-based approaches.
## *Editorial* **Target and Non-Target Approaches for Food Authenticity and Traceability**
**Joana S. Amaral 1,2**
In the last decade, consumers have become increasingly aware of and concerned about the quality and safety of food, in part due to several scandals that were widely disseminated by the media. Currently, consumers are requesting more information about the food they buy, not only from a nutritional point of view but also regarding origin, safety, traceability, and authenticity. In addition, concerns about environmental and ethical issues are on the rise, with more attention being given to topics such as biodiversity protection, production mode, and food authenticity. The growing demand for higher quality foods, the desire for new experiences associated with delicacy products or foods having particular organoleptic characteristics, together with the increasing willingness to pay more money for such products, provides an overall incentive for the adulteration of premium foods. Moreover, several factors such as international trade, market globalization, long and complex food supply chains, and the booming of e-commerce, further create opportunities for food fraud. While in several cases food adulteration poses no major risk for consumers' health (e.g., mislabeling of geographical origin), in others it can result in health hazards due to toxic or allergenic substances. However, even when health is not jeopardized, food fraud leads to unfair market competition and consumers being deceived. For all these reasons, the issue of food authenticity and food fraud has been receiving increased attention from several stakeholders, including government agencies and policymakers, control labs, producers, industry, and the research community, and different attempts have been made aiming for the definition of these concepts. According to the CEN Workshop Agreement 17369:2019, an authentic food product is "*a food product where there is a match between the actual food product characteristics and the corresponding food product claims; when the food product actually is what the claim says that is*" [1,2]. In the discussion paper on food integrity and food authenticity of the working group of the Codex Alimentarius Commission [3], food fraud is described as "*any deliberate action of businesses or individuals to deceive others in regards to the integrity of food to gain undue advantage*". Moreover, four key elements are identified, namely deliberate intent, deception, financial gain and misrepresentation, which are in line with the European Commission's key criteria to refer to when establishing if a case should be considered as fraud or as non-compliance, namely (i) violation of one or more rules of the European Union agri-food chain legislation as referred to in Article 1(2) of Regulation (EU) 2017/625, (ii) customer deception, (iii) economic gain, (iv) intention [2,4]. Furthermore, different types of food fraud have been described, including substitution, dilution, mislabeling, concealment, and unapproved enhancement, among others [2]. In order to identify, tackle and/or deter fraudulent practices in the agri-food sector, complementary approaches are needed to address this complex issue, including analytical testing and broader strategies such as implementing early warning systems, vulnerability assessments, and intelligence gathering, among which the development of new, fast and advanced analytical methods for checking food authenticity is a central aspect. Thus, several works have been published on the subject with respect to different
**Citation:** Amaral, J.S. Target and Non-Target Approaches for Food Authenticity and Traceability. *Foods* **2021**, *10*, 172. https://doi.org/ 10.3390/foods10010172
Received: 11 January 2021 Accepted: 12 January 2021 Published: 16 January 2021
**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.
**Copyright:** © 2021 by the author. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).
food matrices, putting in evidence a variety of analytical techniques that can be used for food authentication [2,5–10]. So far, the majority are targeted methods, which look for a pre-defined characteristic or adulterant, thus being focused on the detection of a few selected analytes [11–13]. However, in the last few years, non-targeted methods have increasingly come into focus. These methods do not rely on the analyses of selected individual analytes since the molecules to be detected are not known a priori, but instead aim at studying a global fingerprint that should be as comprehensive as possible [11–13]. This approach can be advantageous when no information about possible adulterants is yet known and/or when unconventional adulterants are added, which would be unlikely to be detected by conventional targeted approaches. Moreover, contrary to targeted methods that frequently need complex and expensive extraction processes, in non-targeted approaches a simple sample preparation is generally performed to get as many matrix components as possible [12]. Despite the many challenges that still need to be overcome, non-targeted methods are becoming increasingly used and their contribution to deterring food fraud, together with targeted methods, is expected to grow in the coming years.
In this regard, this Special Issue aimed at gathering original research and review papers focusing on the development and application of both targeted and non-targeted methodologies to verify food authenticity and traceability. This Special Issue includes eighteen notable contributions, comprising one review paper and seventeen original research papers, these last dealing with the authentication of different foods, including some considered as highly prone to food fraud such as olive oil [14,15], honey [16,17], fish [18–20] and meat [21–24].
Several research articles in this Special Issue reported the application of different analytical techniques including chromatography, spectrometry, and spectroscopy aiming for food authentication. Grazina et al. [18] used a targeted approach to determine nineteen fatty acids by gas-chromatography with flame ionization detection (GC-FID), which were used together with advanced chemometrics to discriminate wild from farmed salmon. Based on seventeen features obtained from the chemical analysis, all the tested approaches, namely principal components analysis (PCA), *t*-distributed stochastic neighbor embedding (*t*-SNE), and seven machine learning classifiers, allowed them to differentiate the two groups (wild vs. farmed). Moreover, five classifiers allowed distinguishing between groups of farmed salmon from different geographical origins. Detecting mislabeling of geographical origin is an issue that has been receiving increasing attention in the last few years, since certified products or those produced in certain regions are frequently associated with a higher price due to their quality and specific characteristics. Analytical testing for identifying the geographical origin of foods is generally considered of high complexity since specifications for agri-food products with geographic indication are frequently based on subjective characteristics such as organoleptic properties [25]. Kim et al. [26] reported the use of hydrophilic and lipophilic metabolite profiling by gas chromatography-mass spectrometry (GC-MS) coupled with orthogonal partial least squares discriminant analysis (OPLS-DA) to differentiate perilla and sesame seeds originating from China and Korea. Furthermore, the authors noticed that glycolic acid was a notable metabolite for discriminating between perilla seeds grown in China and Korea and proposed this compound as being a potential biomarker for such discrimination. Likewise, proline and glycine could be considered potential biomarkers to determine the geographical origin of sesame seeds. The importance of tracing the geographical origin was also addressed in the study of Vukašinovi´c-Peši´c et al. [16] on multifloral honeys from different regions of Montenegro. The mineral content determined by inductively coupled plasma-optical emission spectrometry (ICP-OES) and linear discriminant analysis allowed the researchers to distinguish honeys that originated from areas exposed to industrial pollution. A different approach was proposed by Lippoli et al. [17] aiming for the fast authentication of honey's geographical origin. The authors describe the development of a non-targeted method using direct analysis in real time and high resolution mass spectrometry (DART-HRMS) combined with multivariate statistical analysis to discriminate chestnut honey from Portugal and Italy and
acacia honey from Italy and China. A non-targeted method coupled with chemometrics was also the approach selected by Barbieri et al. [14] towards the authentication of virgin olive oils. In this study, a classification model was developed based on the raw data from the volatile fraction fingerprint obtained by flash gas chromatography and partial least squares-discriminant analysis (PLS-DA) to predict the commercial category of olive oils (extra virgin, virgin and lampante). The proposed classification model was shown to be robust since it included a high number of diversified samples classified by sensorial analysis (*n* = 331); it was also shown to have good performance, since it was able to correctly classify a high percentage of samples in both cross and external validation. Thus, the proposed approach represents a valid alternative for supporting official sensory panels and increasing the efficiency and fastness of controls, since it could be used as a screening tool allowing for a fast pre-classification of olive oil quality grade, thus supporting the panels by prioritizing the samples or even reducing the number of samples requiring sensory analysis. The comparison of targeted and non-targeted approaches for detecting the adulteration of fresh turkey meat by the fraudulent addition of protein hydrolysates was reported by Wagner et al. [21]. Turkey breast muscles were treated with plant or animal protein hydrolysates (those being produced by enzymatic and acidic hydrolysis and presenting different hydrolyzation degrees—partial or total) and analyzed by traditional high-performance liquid chromatography with ultraviolet-visible detection (HPLC-UV/VIS) targeting ten proteinogenic amino acids and by GC–MS and nuclear magnetic resonance (NMR) spectroscopy as two non-targeted metabolite profiling methodologies. While free amino acids analysis allowed the detection of the injection with fully hydrolyzed proteins, it was not suitable for the detection of food fraud in the case of partial hydrolysates. It was concluded that for lower hydrolyzation degrees, additional compounds originating from protein (such as sugars and the by-products released during hydrolysis) play an important role in the differentiation of nontreated samples and hydrolysate treated ones. Thus, the combination of amino acid profiling and additional compounds can provide stronger evidence for detecting and classifying this kind of adulteration.
The feasibility of using spectroscopic techniques as non-targeted approaches for food authentication was also demonstrated in this Special Issue. Truffles are very expensive mushrooms whose price depends mainly on their species but also on their origin, with the white Piedmont truffle (*Tuber magnatum*) and the black Périgord truffle (*Tuber melanosporum*) being the most valued species. In the paper of Segelke et al. [27] Fourier transform near-infrared (FT-NIR) spectroscopy combined with chemometrics is used to differentiate these truffle species from other species that are less valued but morphologically very similar. Various data pre-processing techniques were evaluated to avoid overfitting and the results compared using several classification models. The results showed the ability to differentiate the expensive white truffle *T. magnatum* from *Tuber borchii* with 100% accuracy, and *T. melanosporum* from *Tuber aestivum* and some species of Chinese black truffles with an accuracy of 99%. Moreover, Piedmont truffles could be differentiated from non-Italian *T. magnatum* truffles with an accuracy of 83%. Therefore, this work demonstrates the potential of FT-NIR spectroscopy as a fast and low-cost authentication tool, not requiring special training for sample preparation and equipment handling, thus being very suited for the industrial screening of samples.
In addition to chemical approaches, several works have been conducted so far describing the development and application of molecular biology techniques for food authentication purposes. These techniques are highly specific and sensitive and are frequently considered as the most suitable tools for the identification of species. Various research papers on the use of DNA-based approaches are also included in this Special Issue, from the comparison of different DNA extraction methods [28] to the use of multiplex polymerase chain reaction (PCR) [23], real-time PCR [19,22,24,29,30], or more advanced techniques such as Digital PCR [31]. Kim et al. [23] proposed the use of a simple qualitative assay based on the use of multiplex PCR to identify three deer species, namely red deer (*Cervus elaphus*), roe deer (*Capreolus capreolus*), and water deer (*Hydropotes inermis*). Three sets of
species-specific primers were developed, generating amplicons of different sizes for each species that were then visualized by capillary electrophoresis to increase resolution and accuracy for the detection of the multiple targets. In other works, the specific identification of species was achieved by using real-time PCR. Kim et al. [24] designed new species-specific primers and probe targeting the *cytb* region of donkey (*Equus asinus*) allowing the detection of as low as 0.001% donkey meat in raw and processed meat mixtures made with beef. Velasco et al. [29] reported the development of a real-time PCR based on the use of specific primers and a minor groove binding TaqMan probe targeting the COI (*Cytochrome Oxidase I*) region for the specific authentication of common cuttlefish (*Sepia officinalis*) in seafood products. Commercial samples were also analyzed by FINS (forensically informative nucleotide sequencing) in order to test the reliability of the developed method and guarantee the correctness of the level of mislabeling found in this work (25%). This low-cost method proved to be reliable in the differentiation of this species from other cephalopods and can be very useful for food control authorities, since species from the genus *Sepia* are frequently similar and very difficult to identify after processing because the characteristics for morphological identification are eliminated. Kyriakopoulou and Kalogianni [15] described the development of a new allele-specific real-time PCR to specifically differentiate olive oil from the valuable wild-type *Olea europaea* var *Sylvestris* from the commonly cultivated type *Olea europaea* L. var *Europaea*. Besides being used for species-specific identification, real-time PCR is also reported for quantification purposes [22,29,30]. While Oh et al. [29] estimate the percentage of corn (*Zea mays*) as an added adulterant in turmeric powder (*Curcuma longa*) by using the fluorescent dye SYBRGreen, others propose the use of specific probes [22,30]. Dolch et al. [22] developed two multiplex real-time PCR assays using specific primers and probes, one for the detection and quantification of chicken (*Gallus gallus*), guinea fowl (*Numida meleagris*) and pheasant (*Phasianus colchicus*), and other for quail (*Coturnix japonica*) and turkey (*Meleagris gallopavo*). For each system, three different quantification methods were compared for estimating the relative meat content of these poultry species in meat mixtures. According to the authors, each method had its pros and cons, although the matrix-specific multiplication factors method was the one presenting more accepted recovery rates. By the contrary, in the work of Grazina et al. [30] the ∆Ct method was chosen to estimate the percentage of *Ginkgo biloba* in commercial herbal infusions. The proposed normalized real-time PCR system, which required the amplification of the specific target (*G. biloba* ITS1 region) using the novel primer set and TaqMan probe and a reference endogenous gene (nuclear 18S rRNA), exhibited high performance parameters and was successfully validated using blind mixtures. To assess the occurrence of fraud in the swordfish supply chain, Ferrito et al. [20] suggested the use of a different molecular strategy encompassing the PCR amplification of the frequently used barcode COI gene combined with the restriction fragment length polymorphism (RFLP) technique. The COIBar-RFLP procedure was applied on several authenticated reference samples of swordfish (*Xiphias gladius*) and four different shark species to generate species-specific restriction enzyme patterns. Those were further used for the authentication of fresh and frozen commercial swordfish slices, allowing the detection of *Prionace glauca*, *Mustelus mustelus* and *Oxynotus centrina* in slices labeled as *Xiphias gladius*. A different technology, namely digital PCR, is reported in the work of Morcia et al. [31] to identify economically motivated adulteration in the pasta industry by the substitution of *Triticum durum* with cheaper common wheat (*Triticum aestivum*). Moreover, the proposed assay allowed the researchers to track the adulterant down to 3%, which is the critical value established in the legislation as a limit for accidental contamination.
Finally, closing this Special Issue, the review paper by Hassoun et al. [32] discusses the use of different analytical methods for detecting frauds in food products of animal origin, with particular attention being paid to non-targeted spectroscopic detection methods. The advantages, opportunities and challenges associated with the use of spectroscopic techniques are discussed and several application examples are given, covering relevant and recently published works.
Overall, the papers included in the Special Issue "Target and Non-Target Approaches for Food Authenticity and Traceability" put in evidence the global relevance of the topic and the importance of developing different approaches that can be used by control laboratories and governmental agencies to verify and guarantee food authenticity and traceability, allowing agencies to detect and expose eventual food fraud scenarios, and therefore protecting producers and industry from unfair competition as well as increasing consumers' confidence in purchased foods.
**Funding:** The author acknowledges the Foundation for Science and Technology (FCT, Portugal) for financial support by national funds FCT/MCTES to CIMO (UIDB/00690/2020).
**Conflicts of Interest:** The author declares no conflict of interest.
#### **References**
*Review*
## **Fraud in Animal Origin Food Products: Advances in Emerging Spectroscopic Detection Methods over the Past Five Years**
**Abdo Hassoun 1,\* , Ingrid Måge <sup>1</sup> , Walter F. Schmidt <sup>2</sup> , Havva Tümay Temiz <sup>3</sup> , Li Li <sup>4</sup> , Hae-Yeong Kim <sup>5</sup> , Heidi Nilsen <sup>1</sup> , Alessandra Biancolillo <sup>6</sup> , Abderrahmane Aït-Kaddour <sup>7</sup> , Marek Sikorski <sup>8</sup> , Ewa Sikorska <sup>9</sup> , Silvia Grassi <sup>10</sup> and Daniel Cozzolino <sup>11</sup>**
Received: 3 July 2020; Accepted: 1 August 2020; Published: 6 August 2020
**Abstract:** Animal origin food products, including fish and seafood, meat and poultry, milk and dairy foods, and other related products play significant roles in human nutrition. However, fraud in this food sector frequently occurs, leading to negative economic impacts on consumers and potential risks to public health and the environment. Therefore, the development of analytical techniques that can rapidly detect fraud and verify the authenticity of such products is of paramount importance. Traditionally, a wide variety of targeted approaches, such as chemical, chromatographic, molecular, and protein-based techniques, among others, have been frequently used to identify animal species, production methods, provenance, and processing of food products. Although these conventional methods are accurate and reliable, they are destructive, time-consuming, and can only be employed at the laboratory scale. On the contrary, alternative methods based mainly on spectroscopy have emerged in recent years as invaluable tools to overcome most of the limitations associated with traditional measurements. The number of scientific studies reporting on various authenticity issues investigated by vibrational spectroscopy, nuclear magnetic resonance, and fluorescence spectroscopy has increased substantially over the past few years, indicating the tremendous potential of these techniques in the fight against food fraud. It is the aim of the present manuscript to review the state-of-the-art research advances since 2015 regarding the use of analytical methods applied to detect fraud in food products of animal origin, with particular attention paid to spectroscopic measurements
coupled with chemometric analysis. The opportunities and challenges surrounding the use of spectroscopic techniques and possible future directions will also be discussed.
**Keywords:** authentication; authenticity; chemometric; fish; origin; honey; meat; milk; spectroscopy; species
#### **1. Introduction**
In recent years, consumers have become more concerned about the quality and safety of food products and have become keenly interested in knowing more about food authenticity and food fraud. In other words, consumers demand more complete information about their food, including what they are really buying, where the food comes from, and when and how it was produced. Although fraud and adulteration have been practiced since ancient times, it is only in recent years that food authenticity issues have been more exposed, and public attention has been intensively paid to the magnitude of this problem and the serious consequences of food fraud [1,2]. Furthermore, during the current pandemic period with coronavirus raging around the world, affecting every aspect of life, including food choices and nutrition habits, consumers have become even more concerned about safety, accessibility, affordability, and the origin of food products than any time before. This increased interest in food authenticity may also be explained by the numerous food scandals over the last few years (e.g., horsemeat scandal in 2013 and rotten meat from Brazil in 2017) and the increased consumer awareness about the impacts of food fraud in terms of illegal economic gain, as well as negative effects on the public health and the environment. Nonetheless, several recent studies have indicated that fraud or mislabeling is still a widespread practice, especially in food products of animal origin, which are often considered among the most frequently adulterated foods [3–6]. Market globalization and increases in international trade, driven by fewer obstacles to the export and import of food, a complex food production chain, and the complex nature of food products of animal origin, the huge variety of these products, as well as the emergence of tricky and more sophisticated forms of fraud are some of the reasons that could explain this rise in food fraud and why detection and prevention are challenging tasks [7–10].
Fraud in animal origin products can take many forms, including mislabeling of the provenance (geographical or botanical origin), species substitution, discrepancies in the production method and farming or breading technique, addition of non-declared substances, as well as fraudulent treatments and non-declaration of processes, such as previous freezing, irradiation, and microwave heating (Figure 1). To support this review and obtain the research published in the last few years on the authenticity of food products of animal origin, Scopus database was queried in May 2020, using the keyword "authenticity" or "authentication" and the different categories of animal origin food products. It can be noticed that a huge amount of studies dealing with authenticity and detection of fraud in fish, meat, milk, honey, and eggs has been published in recent years; the number of published works increased from 530 between 2010 and 2014 to 1000 between 2015 and 2019 (Figure 2a).
Fraud in fish and other seafood is a widespread issue, and seafood products are often ranked among the top food product categories that are susceptible to fraud. Substitution of a high-value fish species with a cheaper alternative and mislabeling of the geographical origin are among the most common fraudulent activities practiced in the fish and seafood sector. Determining whether fish is wild or farmed, tracing farming systems, and differentiating between fresh and frozen–thawed seafoods are among the seafood authenticity topics that have been widely investigated [8]. According to our literature review, meat and meat products are the most studied animal origin foods with regards to authenticity (Figure 2b). Meat authenticity has similar issues to those of fish. To address authentication issues related to muscle foods (fish and meat), a wide range of protein- and DNA-based techniques, chromatography, elemental profiling, and isotopic analysis, among many other measurements, have been frequently applied to this problem [10–12]. Similar techniques have also been established in
routine analysis for detecting fraud that occurs in other foods of animal origin (e.g., milk and dairy products, honey, and eggs). measurements, have been frequently applied to this problem [10–12]. Similar techniques have also been established in routine analysis for detecting fraud that occurs in other foods of animal origin (e.g., milk and dairy products, honey, and eggs). measurements, have been frequently applied to this problem [10–12]. Similar techniques have also been established in routine analysis for detecting fraud that occurs in other foods of animal origin
*Foods* **2020**, *9*, x FOR PEER REVIEW 3 of 44
**Figure 1.** Most reported authenticity issues in food products of animal origin. **Figure 1.** Most reported authenticity issues in food products of animal origin. **Figure 1.** Most reported authenticity issues in food products of animal origin.
**Figure 2.** *Cont.*
3
3
*Foods* **2020**, *9*, x FOR PEER REVIEW 4 of 44
**Figure 2.** Temporal evolution of published work on the authenticity of different categories of food products of animal origin during the last decade (**a**) and publications distributed between the different food categories (**b**). **Figure 2.** Temporal evolution of published work on the authenticity of different categories of food products of animal origin during the last decade (**a**) and publications distributed between the different food categories (**b**).
However, most of the aforementioned analytical methods are associated with several drawbacks, mostly related to the destructive nature of the measurements and the time required to perform the analysis. Therefore, there is still great interest in the development of non-destructive, rapid, accurate, robust, and high-throughput analytical methods for on-site and real-time food authentication. Spectroscopic techniques have gained much importance during the last few years, and spectroscopy has been a popular "buzz word" in the context of fighting fraud and verifying the authenticity of food products. The considerable interest in these non-targeted fingerprinting techniques may be due to the advancements in the analytical instruments and the increasing awareness in the food industry and research on the advantageous aspects of applying such techniques [13]. The number of scientific works regarding the use of spectroscopy for food authenticity increased from 134 papers during 2010–2014 to 369 papers during 2015–2019 (Figure 3a), while the number of total citations (Figure 3b) doubled during the last five years (20,784 citations between 2015 and 2019 versus 9666 citations between 2010 and 2014). Some examples of recent applications of spectroscopic techniques for authentication of food products of animal origin include detection of adulteration in meat [14,15] identification of milk species [16,17], detection of thawed muscle foods [18,19] identification of muscle foods species [20–22], and determination of the botanical However, most of the aforementioned analytical methods are associated with several drawbacks, mostly related to the destructive nature of the measurements and the time required to perform the analysis. Therefore, there is still great interest in the development of non-destructive, rapid, accurate, robust, and high-throughput analytical methods for on-site and real-time food authentication. Spectroscopic techniques have gained much importance during the last few years, and spectroscopy has been a popular "buzz word" in the context of fighting fraud and verifying the authenticity of food products. The considerable interest in these non-targeted fingerprinting techniques may be due to the advancements in the analytical instruments and the increasing awareness in the food industry and research on the advantageous aspects of applying such techniques [13]. The number of scientific works regarding the use of spectroscopy for food authenticity increased from 134 papers during 2010–2014 to 369 papers during 2015–2019 (Figure 3a), while the number of total citations (Figure 3b) doubled during the last five years (20,784 citations between 2015 and 2019 versus 9666 citations between 2010 and 2014). Some examples of recent applications of spectroscopic techniques for authentication of food products of animal origin include detection of adulteration in meat [14,15] identification of milk species [16,17], detection of thawed muscle foods [18,19] identification of muscle foods species [20–22], and determination of the botanical origin of honey [23,24], among many others.
origin of honey [23,24], among many others. Over the last few years, several review papers have been published focusing on either one of the authenticity issues, such as the geographical origin [25,26] or species [27]; or one category of food products of animal origin, such as fish [7,8], meat [28,29], or honey [30]. Other papers have reviewed one specific type of analytical method, such as multielement and stable isotype techniques [11], volatilomics [31], DNA-based methods [32,33], or infrared spectroscopy [34]. The current review will cover the most recent studies that shed light on the various authenticity-related issues (i.e., geographical or botanical origin, species, production method, farming or breeding technique, and processing method) for all food products of animal origin (fish, meat, milk, honey, and egg), highlighting a wide range of both traditional and emerging techniques. This review will first Over the last few years, several review papers have been published focusing on either one of the authenticity issues, such as the geographical origin [25,26] or species [27]; or one category of food products of animal origin, such as fish [7,8], meat [28,29], or honey [30]. Other papers have reviewed one specific type of analytical method, such as multielement and stable isotype techniques [11], volatilomics [31], DNA-based methods [32,33], or infrared spectroscopy [34]. The current review will cover the most recent studies that shed light on the various authenticity-related issues (i.e., geographical or botanical origin, species, production method, farming or breeding technique, and processing method) for all food products of animal origin (fish, meat, milk, honey, and egg), highlighting a wide range of both traditional and emerging techniques. This review will first introduce a brief description
4
of the common multivariate data analysis and analytical techniques related to detecting fraud in food products of animal origin. Several examples of applications of conventional and spectroscopic techniques will be then presented, covering the most relevant works published during the last five years. Finally, some difficulties and challenges, as well as future trends in applications of these techniques, will be discussed. To the best of our knowledge, this review paper is the first to combine results from recent studies on a wide range of analytical methods applied to authenticate fish, meat, milk, honey, and egg, as well as their products. introduce a brief description of the common multivariate data analysis and analytical techniques related to detecting fraud in food products of animal origin. Several examples of applications of conventional and spectroscopic techniques will be then presented, covering the most relevant works published during the last five years. Finally, some difficulties and challenges, as well as future trends in applications of these techniques, will be discussed. To the best of our knowledge, this review paper is the first to combine results from recent studies on a wide range of analytical methods applied to authenticate fish, meat, milk, honey, and egg, as well as their products.
*Foods* **2020**, *9*, x FOR PEER REVIEW 5 of 44
**Figure 3.** Numbers of published works related to food authenticity (blue bars) and use of spectroscopic techniques in relation to food authenticity (red line) (**a**). Numbers of citations including the words authenticity or authentication and spectroscopy (**b**) since 2010. Data obtained from Scopus database on May 25, 2020. **Figure 3.** Numbers of published works related to food authenticity (blue bars) and use of spectroscopic techniques in relation to food authenticity (red line) (**a**). Numbers of citations including the words authenticity or authentication and spectroscopy (**b**) since 2010. Data obtained from Scopus database on 25 May 2020.
#### **2. Multivariate Data Analysis 2. Multivariate Data Analysis**
Traditional chemometric methods are based on linear projections onto a lower dimensional latent variable space, and these powerful and simple methods still dominate the field. However, more flexible and data-intensive machine learning methods have gained traction lately. These methods have the ability to model complex, non-linear relationships; however, the curve fitting procedures, interpretation, and validation are often more complicated. In general, the choice of data analysis strategy depends on the research question, as well as the type and size of the available data. The data analysis pipeline consists of preprocessing, data exploration, modeling, and validation. Traditional chemometric methods are based on linear projections onto a lower dimensional latent variable space, and these powerful and simple methods still dominate the field. However, more flexible and data-intensive machine learning methods have gained traction lately. These methods have the ability to model complex, non-linear relationships; however, the curve fitting procedures, interpretation, and validation are often more complicated. In general, the choice of data analysis strategy depends on the research question, as well as the type and size of the available data.
The following sections give a brief description of each of these steps, with the main emphasis on recent trends and developments. For detailed overviews of data analysis in food authenticity, please refer to [35–38]. *2.1. Data Preprocessing* The data analysis pipeline consists of preprocessing, data exploration, modeling, and validation. The following sections give a brief description of each of these steps, with the main emphasis on recent trends and developments. For detailed overviews of data analysis in food authenticity, please refer to [35–38].
#### The aim of preprocessing is to reduce non-relevant variations in the signal stemming from *2.1. Data Preprocessing*
instrumental artifacts, surrounding effects, or sample preparation. The most used methods include standard normal variate (SNV), (extended) multiplicative signal correction ((E)MSC), derivatives, smoothing, baseline corrections, and peak alignments, which are often used in combination. The choice of preprocessing method is critical for the subsequent modeling and interpretation [39,40], and should be chosen based on knowledge of the samples and the measurement platform. Recent research suggests various strategies for making the modeling less sensitive to preprocessing, for instance by using a boosting approach [41], through Tikhonov regularization [42], or by using convolutional neural networks [43–45]. *2.2. Data Exploration* The aim of preprocessing is to reduce non-relevant variations in the signal stemming from instrumental artifacts, surrounding effects, or sample preparation. The most used methods include standard normal variate (SNV), (extended) multiplicative signal correction ((E)MSC), derivatives, smoothing, baseline corrections, and peak alignments, which are often used in combination. The choice of preprocessing method is critical for the subsequent modeling and interpretation [39,40], and should be chosen based on knowledge of the samples and the measurement platform. Recent research suggests various strategies for making the modeling less sensitive to preprocessing, for instance by using a boosting approach [41], through Tikhonov regularization [42], or by using convolutional neural networks [43–45].
#### overview of the data, deal with outliers, evaluate the effects of preprocessing, and get a first *2.2. Data Exploration*
5 Data exploration is an important step prior to the actual modeling. The aim is to gain an overview of the data, deal with outliers, evaluate the effects of preprocessing, and get a first impression of the
Data exploration is an important step prior to the actual modeling. The aim is to gain an
potential for discriminating between samples. Principal component analysis (PCA) is the most used tool for data exploration, providing a linear transformation of the original data by maximizing the explained variance. Cluster analysis is another group of exploratory methods based on a certain distance or similarity measure between samples. These methods can be more flexible than PCA, depending on the chosen similarity metric, and may be useful for very large sample sizes.
#### *2.3. Modeling*
Authentication tasks mainly aim to determine which category a food item belongs to, i.e., classification. There are two main approaches to classification: class modeling and class discrimination [46–48]. While class modeling focuses on modeling the similarities among samples from the same category, class discrimination focuses on finding the differences between a set of predefined categories. The most used methods in the scientific literature are the soft independent modeling of class analogies (SIMCA) and partial least squares discriminant analysis (PLS-DA) classical chemometric methods for class modeling and discrimination, respectively; however, methods such as support vector machines (SVM), random forests (RF), k-nearest neighbor (k-NN), and different types of neural networks (NN) are also frequently applied. Quantitative prediction models are also relevant in some cases, for instance when the objective is to quantify the amounts of specific adulterants. An overview of alternative methods for class modeling, discriminant analysis, and quantitative prediction can be found in [35–38].
*Data Fusion*: Data or sensor fusion is an emerging topic within food authentication. A combination of several instrumental techniques can lead to more accurate results, either by providing complementary information or by reducing uncertainty [49–53]. Data fusion is also an active research area in fields other than authenticity, and new methods for explorative analysis, classification, and prediction are presented frequently. In principle, all multivariate methods can be used for data fusion by (1) combining all the measured variables directly, called low-level data fusion; (2) combining extracted features such as principal components, called mid- or feature-level data fusion; or (3) combining predictions or classifications from different techniques through voting, called high- or decision-level fusion. There are also several methods that are tailored for data fusion problems. Examples of newly developed explorative techniques include methods that separate common and distinctive variations in multiple data blocks [54,55], whereas sequentially orthogonalized PLS (SO-PLS) [56,57] is a common example of multiblock regression methods.
*From Small to Big Data*: In general, the traditional chemometric methods, such as PCA, SIMCA, and PLS-DA/PLSR, are suited for small feasibility studies, while larger studies allow for use of more data-intensive methods, such as SVM, RF, and NN. In industrial applications, however, databases with hundreds of thousands of samples are often available. Such huge data sets call for completely different data analysis strategies. There has so far been little focus on authentication models based on large databases in the scientific literature, mainly because these databases are not open. There are, however, a few exceptions showing that local modeling is a promising strategy [58,59]. In local modeling, a new model is fitted for each new sample to be predicted, using only a subset of spectrally similar samples as a calibration set. More research is needed on the use of local modeling for classification and on the analysis of large databases in general.
#### *2.4. Validation*
One of the main barriers for the successful implementation of fingerprinting techniques in food authenticity is the lack of proper validation schemes [2,60–62]. A full validation scheme consists of four phases: (1) optimization of the analytical procedure, (2) statistical model selection and parameter optimization, (3) testing of the model performance, and (4) stability testing by system challenges [60]. Most published feasibility studies stop at phase two or three, while phase four is essential for successful implementation.
Phase one is specific for the analytical technique and will not be covered here. The aim of phase two is to select an optimal modeling strategy and model parameters. This is usually done by resampling methods, such as cross-validation or bootstrapping. Phase three involves testing the model performance using an independent test set, while phase four tests the extrapolation of the model, e.g., overtime or for different instruments and locations. Thorough reviews of both numerical and conceptual aspects of validation are given in [63,64].
#### **3. Overview of Fraud Detection Techniques**
#### *3.1. Spectroscopic Techniques*
#### 3.1.1. Vibrational Spectroscopy
Innovation pathways in vibrational spectroscopy during this past half decade are preludes to potential impacts and further practical achievements in the next half decade. Vibrational spectroscopy techniques, including infrared spectroscopy in the near (NIR)- and mid (MIR)-infrared spectral ranges, as well as Raman spectroscopy, enable a fingerprinting chemical analysis of an intact food sample in situ for adulteration in real time. The sample remains intact for confirmatory analysis using other techniques. Spectroscopic technologies require high levels of rigor in the evidence for authentication of both the food or food product and of the adulterant or contaminant.
Variance in the spectral signature of the food always can complicate the capacity to distinguish the amount and composition of an adulterant or contaminant. Recent state-of-the-art authentication of milk products has been reported [65,66]. The authentication of raw milk involves a different process—knowing its fingerprint identity enables detecting adulteration [67]. Products made from milk have also been authenticated. Desi ghee made from buffalo and from cow milk can be differentiated [68], while butter containing lard [69] and cream and yogurt [70] can be distinguished with chemometrics.
Authentication in meats is required for foods that are labeled as individual meats [71]. Horsemeat in minced beef [72], beef and mutton in pork [73], and rainbow trout in Atlantic salmon [74] each require sufficient data specific to substances to be labeled to assure the meat contaminant is properly characterized in order to identify markers characteristic of each additional component. Spectral data on the primary meat preferentially needs to be oversampled relative to that of a contaminant, or of minor or occasional components that could be misinterpreted as unrelated to the original meat. Factors such as diet can alter vibrational fingerprints. Eggs from poultry fed omega-3 fatty acids contain an intentional adulterant that can be detected in the spectral signature of the eggs [75]. Work involving fish fillet authentication using vibrational spectroscopy has also been published [21,76].
#### 3.1.2. Nuclear Magnetic Resonance
Nuclear magnetic resonance (NMR) spectroscopy, despite being a very well-established methodology in food analysis, has had limited new publications over the last five years. The major difficulties are that foods are inherently mixtures of components and adulterants may or may not be mixtures. Thus, identifying NMR chemical shifts that do not belong in a particular food first requires authentication of the fact that a particular set of peaks may not arise occasionally (i.e., more rarely) on its own. This is an innately complicated process because one needs to ascertain which chemical shifts are correlated. A major advantage of NMR is that modern NMR techniques can trace the fingerprints from finger to finger and ascertain one part of a fingerprint belongs to another hand. Publishing the results of such an effort is often difficult because someone else may have found the same compound in this (or another) food product. Further, if the compound found is of little apparent biological or food property relevance, journal reviewers can deem such research as having a correspondingly low relevance.
The specificity of NMR complicates the authentication of the composition of an adulterant. A unique and specific NMR peak at best detects only a single component of an adulterant. If an
adulterant happens to be a mixture of components, NMR is useful only for detecting chemical components one at a time. Thus, if in minced meat labelled beef porcine fat can be detected as an adulterant, NMR can only identify a chemical shift, which identifies a site on a specific unsaturated lipid as foreign to beef. It cannot likely identify which animal (or plant for that matter) was the source of the product contamination. Once markers have been authenticated properly to a specific chemical structure, this fingerprint is treated as a positive result awaiting verification by some other technique. Verifying the food commodity that has been used for adulteration requires significantly more spectroscopic data. Each and every spectral identification result that can be detected in a specific matrix can be a significant challenge and are important to know. Publishing such a finding is a more complicated endeavor.
Three recent NMR manuscripts involved detection of adulteration in milk, powdered milk, or butter [77–79]. Two publications involving edible lipids, including milk, additionally used more complicated NMR experiments (time-domain NMR and <sup>13</sup>C inept NMR) [80,81]. The more complicated NMR techniques enhance the resolution and quality of data collected. A similar enhancement using improved technologies and methodologies in milk was reported using Raman chemical imaging techniques [82]. One manuscript reported on the authentication of krill oil using NMR techniques [83].
The focus on NMR research in authenticating lipid compositions in foods is because specific lipids in mixtures of lipids appear to be characteristic of their origin. The high resolution of NMR enables deconvolution of the specificity of the lipid composition at the molecular level. Solvent effects, however, appear to complicate spectral assignments. Two publications verified that NMR techniques can fully distinguish omega-3 from omega-6 fatty acids in mixtures [84] and among three omega-3 fatty acid structural analogs, each in an intact lipid environment [85]. Authenticating the fingerprints of lipids is an essential component of and prerequisite for verifying adulteration correctly.
#### 3.1.3. Fluorescence Spectroscopy
Fluorescence spectroscopy is based on measurement of the spectral distribution of the intensity of the light emitted by electronically excited molecules. Fluorescence coupled with chemometrics has been widely used in food studies, including for products of animal origin [86–91]. The main advantages of fluorescence as compared to other spectroscopic techniques are its higher sensitivity and selectivity. Due to these features, fluorescence is particularly useful for studying minor and trace components in complex food matrices [87,91]. Characterization of real multifluorophoric food samples requires more advanced measurement techniques than conventional emission or excitation spectra. The advanced fluorescence techniques have often been used in food studies, including excitation–emission matrix (EEM) fluorescence spectroscopy, synchronous fluorescence spectroscopy (SFS), and total synchronous fluorescence spectroscopy (TSFS) [87,92].
Fluorescence patterns of food products are usually complex. Fluorophores in food include natural food components, process-derived compounds, food additives, and contaminants [89]. Autofluorescence of meat and fish originates mainly from collagen, adipose tissues, proteins, and oxidation products [89]. Milk and dairy products contain several intrinsic fluorophores, including free aromatic amino acids, nucleic acids, aromatic amino acids in proteins, vitamins A and B2, nicotinamide adenine dinucleotide (NAD), chlorophyll, and oxidation and Maillard reaction products [86,90]. Fluorescence in honey is ascribed to proteins, polyphenolic compounds, and Maillard reaction products [23,93,94]. The unique fluorescence patterns of food products have been successfully utilized in authentication studies of food of animal origin, including meat [95,96], fish [97,98], milk [16,17] dairy products [86,90], and honey [23,88,99–102].
#### 3.1.4. Other Spectroscopic Techniques
The number of studies on the potential use of novel spectroscopic techniques to detect fraudulent practices encountered in the food chain has gradually increased in recent years. In this section, information is given on the applications of laser-induced breakdown spectroscopy (LIBS), terahertz (THz) spectroscopy, and hyperspectral imaging (I) in food adulteration analysis.
LIBS has been presented as a potential alternative to the existing analytical atomic spectrometry techniques used to determine the elemental composition of food. Most of the samples need a minimum or no sample preparation to be analyzed by using LIBS. The simultaneous analysis of multiple elements can be achieved. It is highly applicable to at-, on-, and in-line measurements and remote sensing, enhancing its potential as an analytical technology process [103,104]. LIBS coupled with several chemometric methods has been widely used for species discrimination [105], determination of adulteration [106], and spatial mapping of the sample surfaces in meat, milk, and other products [107,108]. Recently, some studies utilized LIBS for analysis of honey adulteration [109,110] and determination of its geographical origin [111,112]. Although there is a significant amount of research in the literature reporting the high potential of LIBS as an at-line monitoring tool for the industry, there is still a need for further improvements in system components and configurations. Besides, more research is required to recommend alternatives to reduce the matrix effect and minimize sample preparation procedures in order to improve the predictive accuracy. Peng et al. have described the significant challenges and possible solutions to these in order to speed up the use of LIBS as an in situ monitoring tool [113,114].
Terahertz spectroscopy (THz) is another technique that provides an excellent alternative to X-rays in order to obtain high-resolution images from the interiors of solid objects. Frequency-domain and time-domain measurements are performed for both imaging and spectroscopy with THz waves [115]. There are a limited number of studies on the use of THz spectroscopy for the determination of food adulteration, which were previously compiled by Afsah-Hejri [115] and He [116]. Adulteration of milk with a fat powder [117], discrimination of honey samples [118], and determination of honey adulteration [119] were some of the recent study topics.
Hyperspectral imaging (HSI) is another relatively new technology, which has explicit potential to satisfy the needs for remote and real-time monitoring techniques. Being rapid, non-invasive, and providing spectral and spatial features simultaneously are some of its significant advantages. Numerous articles describe the pros and cons of HSI-based methods for food authenticity and adulteration analyses [14,15]. Nowadays, low-cost, rapid, and simple multispectral imaging systems are being designed for the determination of particular adulterations [120]. Efforts are being made to offer alternative methods for the interpretation of HSI data. The transition from the use of linear classifiers to machine learning and deep learning solutions offers a great variety of opportunities [121]. Another trend is to employ miniature devices called single shot or snapshot hyperspectral sensors, which are ultra-portable and able to acquire data at video rates [122]. The enormous potential of the HSI technique to detect many aspects of food adulterations has been shown in the literature. Even so, enhancement of the spectral and spatial resolution and presentation of alternative technologies for advanced data analysis would be positive contributions to the accuracy and cost-effectiveness of the developed methods.
#### *3.2. Other Analytical Methods*
#### 3.2.1. DNA-Based Techniques
To date, many DNA-based detection methods have been developed to determine animal species in food products. In particular, DNA-based methods have been used to detect target species in processed foods, because DNA is stable at high temperatures and pressures. Sequencing-based techniques (such as DNA barcoding and minibarcoding), polymerase chain reaction (PCR) coupled with restriction fragment length polymorphism (PCR-RFLP), real-time PCR, multiplex PCR, and species-specific PCR are among the most used techniques [32,123,124].
Identification of short DNA sequences, called DNA barcodes, has been widely exploited for species discrimination. DNA barcoding and minibarcoding were used to authenticate animal-derived food products sold in the Chinese market [125] and to identify selected brands of locally-produced canned and dried sardines in the Philippines [126].
In PCR-RFLP, the PCR products are cleaved with restriction enzymes, followed by gel electrophoresis and blotting [32,123]. The technique was successfully applied to differentiate four commercial shrimp types in India, and the developed PCR-RFLP protocol was validated by analyzing 52 commercial shrimp products [127].
Real-time PCR and multiplex PCR methods are the most common detection technologies in meat and meat products, fish and seafood, and other food categories that are known to have a high incidence of adulteration [124,128–130]. There are numerous reports in the literature demonstrating that real-time PCR is a powerful method that can be used as a reliable and sensitive technique for meat identification. For example, in one of the recent studies, a real-time PCR assay was developed for the detection of raw donkey meat and different processed meat mixtures [131]. Fang and Zhang used real-time PCR and TaqMan-specific probes for the detection of murine components in mutton meat products [129]. The results showed that the limit of detection was lower than 1 pg of DNA per reaction and 0.1% murine contamination in meat mixtures.
Many researchers have applied multiplex PCR methods for identification of meat species for simultaneous and rapid detection of multiple species in a single reaction. For example, two direct-triplex real-time PCR systems based on intercalating dyes were applied as a robust and precise quantitative PCR assay for meat species identification [124]. No DNA extraction was required and 92.5% of market samples of six commonly eaten meat species were successfully amplified. The multiplex PCR method was also applied to detect chicken, duck, and goose in beef, mutton, pork, or quail meat samples [132]. In a similar study, a multiplex PCR assay was used to identify lamb, beef, and duck in a meat mixture before and after heat treatment [133]. Similar approaches were developed to monitor commercial jerky products [134]; to detect chicken and pigeon in raw and heat-treated meats [135]; and to detect chicken, turkey, and duck in processed meat products [130]. Recently, a fast multiplex real-time PCR with TaqMan probes was performed to simultaneously detect pork, chicken, and beef in processed meat samples [136].
The species-specific PCR method has been used to a great extent for meat species identification in foods because of its high specificity and rapidity. For instance, El-Razik and co-authors used a species-specific PCR test to differentiate donkey and horse tissue in cooked beef meat products in Egypt [137]. In another study, a species-specific PCR was developed for the identification of beef in India [138].
In addition, more advanced high-throughput DNA sequencing methods, such as next-generation sequencing (NGS) [139,140], have emerged in recent years as valuable techniques for carrying out untargeted screening of complex samples.
#### 3.2.2. Protein-Based Techniques and Related Methods
Chromatographic, electrophoretic, and immunological methods have been widely used for different authenticity issues for food products of animal origin [29,123,141,142]. Different mass spectrometry (MS) techniques have emerged in recent years, and along with chromatographic and NMR techniques have become some of the most commonly applied approaches for metabolomic fingerprinting [142,143]. Traditionally, MS methods are coupled with chromatographic separation techniques, such as liquid chromatography mass spectrometry (LC-MS) [142]. More recently, direct MS analysis approaches, such as matrix-assisted laser desorption/ionization-time of flight (MALDI-TOF), real-time techniques (e.g., direct analysis in real-time (DART) technique), and high-resolution mass spectrometry (HR-MS), among others, have been developed and applied to many authentication problems [123,144–147]. For example, a DART HR-MS method was developed to discriminate Canadian wild salmon from the farmed fish produced in Canada, Chile, and Norway [144]. The results showed that PCA applied to the 30 most abundant signals generated from fatty acids after the DART HR-MS analysis of fillet lipid extracts enabled a rapid discrimination between farmed and wild fish, whereas
discriminant analysis (DA) gave a correct classification rate of 100%. In another study, the differences between rainbow trout and king and Atlantic salmons were studied using a lipidomical method based on hydrophilic interaction chromatography MS [147]. PCA was applied to recognize the variance among these fish species, which was attributed to the genetic origin, living environment, and feed ingredients, among others. A novel method based on quadrupole time-of-flight (Q-TOF) MS coupled with a surgical diathermy device was recently developed to distinguish cod from oilfish in real time [145]. PCA demonstrated that the clusters of oilfish were well separated from those of cod, while the application of discriminant analysis models showed that the fish tissue can be authenticated with 96–100% accuracy. Another recent study investigated the potential of ultra-performance liquid chromatography–triple time-of-flight–tandem mass spectrometry (UPLC−triple TOF−MS/MS) to determine lipid composition in the muscle tissue of four popularly consumed shrimp species [146]. About 600 lipid compounds from 14 classes were characterized and quantified, and PCA results of lipid profiles allowed the different species to be distinguished. In a similar investigation, the use of LC-TOF−MS allowed the detection of commercially available, highly processed mixed-meat products, including duck, goose, and chicken, along with pork and beef [148].
Besides the chromatographic and mass spectrometry techniques, enzyme-linked immunosorbent assay (ELISA) is one of the most widely used methods for meat identification, because it is cheap and easy to perform [141,149].
Although the aforementioned techniques have several advantages, such as stability during thermal processing and high sensitivity and selectivity, most of these measurements are time-consuming because several steps are required for sample preparation, protein extraction, and lipid extraction. In addition, the technical difficulty with MS and PCR in food adulteration is that they are useful mainly and sometimes only after the rest of the chemistry and spectroscopy work has been completed. Such techniques are especially valuable for verifying adulterations detected in situ by other technologies.
#### 3.2.3. Isotopic Technique
As the isotopic compositions of the plants or animals reflect the condition of natural environment where they grew up, the light stable isotopes <sup>13</sup>C/ <sup>12</sup>C, <sup>18</sup>O/ <sup>16</sup>O, <sup>2</sup>H/ <sup>1</sup>H, and <sup>15</sup>N/ <sup>14</sup>N, <sup>34</sup>S/ <sup>32</sup>S; and the heavy isotopes <sup>11</sup>B/ <sup>10</sup>B and <sup>87</sup>Sr/ <sup>86</sup>Sr are commonly used in food authentication [11]. Preliminary studies have demonstrated the usefulness of stable isotope analysis in determining the origins of animal origin products [150–154]. However, the animal origin products had more complicated life cycles than the plant origin products. The stable isotopes such as δ <sup>2</sup>H and δ <sup>18</sup>O were more likely to be affected by the ambient environment [151,155]. Camin and co-authors [151] reported the H/O ratios of Italian rainbow trout fillets were positively interrelated with the O ratio of tank water. However, the other stable isotopes <sup>13</sup>C/ <sup>12</sup>C, <sup>15</sup>N/ <sup>14</sup>N, and <sup>34</sup>S/ <sup>32</sup>S were reported to be affected by diet [11,156,157]. Taking shrimp as an example, the δ <sup>13</sup>C and δ <sup>15</sup>N values in shrimp are significantly related to the food sources [158]. During shrimp culture, the farmers may use several brands of commercial feeds with different ingredients and isotopic signatures. Li et al. [156] reported that the δ <sup>13</sup>C and δ <sup>15</sup>N values in 16 commercial feeds used in shrimp culture in China ranged from −23.03 to −24.75‰, and from 2.1 to 8.18‰, respectively. The dietary shifts could influence the stable isotope signature of shrimp. The effects of diet on the stable isotope signature of animal origin products should be considered when using traceability methods. Moreover, animals can only be sampled for traceability purposes when they are in isotopic equilibrium with their diet. In a recent study, Li and others suggested the sampling of shrimp that have been consistently fed with the same feed for more than twenty days [158].
The stable isotopes of animal origin products could also be affected by other environmental factors, such as culture seasons and salinity [159,160]. Compared with the marine ecosystem, the freshwater ecosystem generally has low δ <sup>13</sup>C and δ <sup>15</sup>N values [159,161]. Previous studies reported different δ <sup>13</sup>C and δ <sup>15</sup>N values in shrimp and fish cultured in freshwater and seawater [156,159]. Hence, all of these factors should be compared when using isotopic traceability methods to allow for deter animal origin product fraud.
#### 3.2.4. Elemental Technique
The isotopic technique is usually combined with an elemental profiling technique to increase the accuracy of the traceability methods [157,159,162–165]. Elemental profiling techniques rely on digestion of samples into ions, then concentration of the ions is followed by spectroscopic analysis, including atomic absorption spectroscopy (AAS), inductively coupled plasma–optical emission spectroscopy (ICP-OES), and ICP–mass spectrometry (ICP–MS) [11]. The analyzed elements include K, Ca, Na, Mg, Cu, Fe, Mn, Al, Zn, As, Cd, Cr, Mn, Ni, Zn, Ba, Sr, Li, Se, Co, Ti, and V. Those elements include bulk structural elements (P, S, Si, etc.), macroelements (K, Ca, Na, Mg, etc.), trace elements (Cu, Fe, Zn etc.), and ultratrace elements (As, Cd, Cr, Mn, Ni, Co, V, etc.). Both non-metal elements (P, S, As, etc.) and metals (Mn, Fe, Cu, etc.) have been used in analysis [166]. In recent years, the rare earth elements (REEs), including Y, Ce, Nd, Pr, Sm, Er, and Eu, have also been used in traceability methods [159,167]. Databases generated by chemical analysis are subjected to multivariate analysis for data exploration.
Elemental profiling was used in geographic traceability testing of plant origin products, because element compositions of the specimen were believed to be a distorted reflection of the elemental profiling of the soil environment in which they grew [11]. This fact is more complicated for animals, who derive their elements not only from the environment but also the food they consume. Hence, feed is a factor that needs to be seriously considered in the traceability of animal origin food products. Mineral concentrations of feed, such as fish feed, vary greatly due to differences in raw ingredients, addition of specific macro or trace mineral premixes, or contamination [11]. The culture environment of animals is also more complicated than plants and the elemental profiling of animals can be affected by factors such as the culture season, size of the animal, species, and water quality [160,168,169]. For example, Han and others [168] reported that the element compositions of salmonid obtained from the reservoir were vulnerable to seasonal changes. Although studies have demonstrated the usefulness of elemental profiling in tracing the origin of animal origin products, all factors should be considered in future studies to strengthen the accuracy of the method.
#### **4. Examples of Recent Use of Spectroscopic and Traditional Methods to Detect Fraud**
#### *4.1. Fish and Seafood Products*
*Identification of Geographical Origin*: Provenance or geographical origin has become one of the most important authenticity issues for fish and seafood due to the increasing awareness among consumers of the impacts of their purchasing choice of seafood on the marine environment. Many consumers are becoming more worried about fraud, which occurs when fraudsters conceal the geographical origin or hide an illegally harvested protected species or a species from a protected area. Thus, reporting of the country of origin or place of provenance of seafood is essential in the fight to preserve sustainable fisheries, for better management of fish stocks, and to prevent unreported and unregulated fishing. This is why a requirement with respect to a clear indication of the geographical origin of seafood products has been implemented in many countries, such as the European Union and the USA [123,170].
Several analytical methods have been developed in order to identify the origin of seafood. Trace elements fingerprinting, stable isotope analysis, and DNA-based methods are among the most used approaches for this purpose. While these techniques show promise for definitively identifying the geographical origin of fish and other seafood [32,171–173], they have certain drawbacks, especially in terms of the required time and the destructive nature of measurements.
Recently, some studies have demonstrated the usefulness of spectroscopic techniques for monitoring the geographical origin of seafood [174–176] (Table 1). In one of these studies, NIR spectroscopy was applied to classify tilapia fillets according to their 4 geographical origins, namely Guangdong, Hainan, Guangxi, and Fujian in China [174]. SIMCA performed on the spectra showed a classification ability ranging from 75% for the Guangxi provenance to more than 80% for
the other origins. In another study, a better classification efficiency of sea cucumber originating from nine Chinese locations was obtained using NIR spectroscopy combined with PCA [175]. More recently, a similar technique was used to trace the geographical origin of European sea bass collected from the Western, Central, or Eastern Mediterranean Sea [176]. Results showed correct classification rates of 100% 88%, and 85% for the fish originating from the Eastern, Central, and Western Mediterranean Seas, respectively, with lipid absorption bands being the major contributor to the discrimination ability of the spectra.
In the literature, there are few studies regarding the use of NMR or fluorescence spectroscopy for monitoring of the geographical origin of seafood. In one of these scarce studies [177], <sup>1</sup>H NMR spectroscopy combined with SIMCA and PLS-DA was successfully applied to discriminate caviar cans originating from producers in the Aquitaine region in France from other producers. Therefore, more spectroscopic studies should be conducted on this topic in order to draw valid conclusions about the potential of these techniques for determining the geographical origin of fish and other seafood.
*Tracing Wild and Farmed Seafood and Farming Methods*: During the last few years, there has been a rapid expansion of aquaculture as a result of overfishing and decreasing wild fish stocks. Consumers generally prefer wild fish over farmed fish, and when it comes to farming, organically farmed fish is usually believed to be healthier and of higher quality in terms of animal welfare and environmental perspectives compared to conventionally farmed fish. This is why labeling farmed fish as wild or conventionally raised fish as organic is considered a fraudulent practice.
Various approaches have been proposed over the years to trace production methods and farming systems. Elemental profiling, stable isotopes, fatty acid analysis, or combinations of these methods have been extensively applied [144,173,178,179]. For example, a technique based on stable isotope analysis allowed differentiation of organically farmed from conventionally farmed salmon and brown trout, independent of the type of processing, i.e., raw, smoked, or graved [180]. In another study, the combination of stable isotope ratio analysis with multielement analysis gave a correct classification of 100% of shrimp samples according to their geographical origin and production method (i.e., wild or farmed), while 93.5% of the samples were correctly classified according to species [163]. A more recent study confirmed the positive effects of combining the stable isotopes and elemental profiling techniques to determine the origin and production method of Asian sea bass collected from Australian and Asian sources [160].
Only a few studies regarding the use of spectroscopic techniques for distinguishing between wild and farmed fish or between different farming regimes have been published so far. Xu and co-authors studied the possibility of discriminating wild and farmed salmon with different geographical origins and farming systems using HSI operating in two spectral ranges (spectral set I: 400–1000 nm; spectral set II: 897–1753 nm) coupled with different chemometric tools [181]. The best results were obtained with SVM applied to spectral set I, giving a correct classification rate of 98.2%. In a more recent study, NIR spectroscopy in the range of 1100–2500 nm was applied to authenticate European sea bass [176]. Slight separation was observed between fish groups when PCA was applied. However, PLS-DA allowed a clear discrimination between wild and farmed fish with a correct classification rate of 100% being achieved. Moreover, the different farming systems, including extensive, semi-intensive, and intensive farming, were discriminated from each other with correct classification rates of 67%, 80%, and 100%, respectively. In this study, the absorption bands of proteins were reported to be the greatest contributors to the discrimination ability of the spectra.
*Detection of Species Fraud*: Substitution of valuable marine species with less desirable or cheaper ones is the most common type of fraud in fish and other seafood. Detection of this type of fraud is difficult, especially if the fish is in the form of a fillet without skin or if the seafood product has been processed [182,183]. Given the widespread practice of species fraud and the serious consequences it can have, it is no wonder that a wide variety of conventional methods and spectral fingerprinting techniques have been investigated in order to aid in addressing this issue. DNA analysis and MS methods are among the most commonly used techniques in this regard [126,145,146,184,185].
**Table 1.** Examples of applications of spectroscopic techniques with respect to various authenticity issues for fish and other seafood.
PCA, Principal Component Analysis; PCR, Principle Component Regression; LDA, Linear Discriminant Analysis; DA, Discriminant Analysis; RF, Random Forest; SIMCA, Soft Independent Modeling of Class Analogy; PLS-DA, Partial Least Squares Discriminant Analysis; PLSR, Partial Least Squares Regression; LS-SVM, Least Squares Support Vector Machines; PNN, Probabilistic Neural Network; VIS/NIR; Visible–Near-Infrared Spectroscopy; his, Hyper Spectral Imaging; LF-NMR, Low-Field Nuclear Magnetic Resonance; MRI, Magnetic Resonance Imaging; FT-IR, Fourier-Transform Infrared Spectroscopy; ELM, Extreme Learning Machine; HCA, Hierarchical Cluster Analysis.
Several spectroscopic techniques in conjunction with chemometric tools have been used to identify fish species and detect fraud. Alamprese and Casiraghi used FT-NIR and FT-MIR data coupled with two classification techniques (i.e., SIMCA and linear discriminant analyses (LDA)) in order to discriminate valuable fish species (i.e., red mullet and plaice) substituted with cheaper ones, namely Atlantic mullet and flounder [76]. The best results were obtained by the LDA model, giving a 100% correct classification rate for red mullet and Atlantic mullet, regardless of the used spectroscopic techniques. Regarding discrimination between plaice and flounder species, the best results were obtained using FT-IR, with more than 83% prediction ability and 100% specificity being achieved.
The progress in miniaturization accompanied by software development has led to the emergence of several handheld and portable devices based on spectroscopy for many applications in the food industry [198,199]. In this respect, an investigation based on a handheld NIR device and FT-NIR benchtop spectrometer was carried out in order to discriminate Atlantic cod from haddock fillets and patties [200]. The results obtained by applying LDA and SIMCA models to the spectra using the portable device demonstrated an equivalent discrimination power to those obtained by the stationary benchtop instrument.
Besides NIR spectroscopy, other vibrational spectroscopic techniques have been widely employed to detect fraud in seafood species. For instance, MIR spectroscopy was applied to detect fraud involving substituting Atlantic salmon with rainbow trout in mini-burgers [201]. Using PCA, the authors succeeded in discriminating 11 formulations with different percentages of these two species, and the percentage of the fraud in the mixture was successfully predicted using PLSR. The same authenticity issue (i.e., species identification) was later studied in a similar investigation, but with a different vibrational spectroscopic technique, namely Raman spectroscopy [74].
Again, few or no studies have been found in the literature regarding the application of NMR or fluorescence spectroscopy. A recent study investigated the use of HSI in 4 different spectroscopic modes, including reflectance in the VIS/NIR region, fluorescence, reflectance in the short-wave infrared region, and Raman spectroscopy for discriminating between 6 fish species and differentiating between fresh and frozen–thawed fish [21]. By testing several machine learning classifiers, the authors obtained the best results when using the VIS/NIR and the short-wave infrared techniques for the identification of fish species and detection of thawed fish, respectively.
*Checking of Processing Treatments*: Fish and other seafood products are highly perishable foods that must be processed or preserved properly and rapidly after catch or harvest in order to extend their shelf life and maintain quality. Freezing has been widely applied as one of the most common ways of achieving this purpose. However, fresh products are often considered by consumers to be of superior quality and are usually sold at higher prices then frozen food. Therefore, discrimination between fresh and frozen products is one of the most important authenticity issues. Enzymatic, electrophoretic, and histological methods have been commonly used to detect thawed fish and seafood [202–205].
Vibrational spectroscopy, NMR, and fluorescence spectroscopy have shown considerable potential as interesting alternatives to traditional measurements used to differentiate fresh from frozen–thawed seafood. For example, differentiation of fresh and frozen–thawed Atlantic mullet fillets was successfully reported with the use of SIMCA applied to FT-IR, with values of more than 98%, 88%, and 95% being obtained for classification ability, prediction ability, and specificity, respectively [76]. Similar results were reported by using PLS-DA on VIS/NIR spectra obtained for fresh and frozen–thawed tuna, and high sensitivity, specificity, and accuracy of the model were achieved [206].
Unlike the other vibrational spectroscopy, very little work has been devoted to examining the potential of Raman spectroscopy to differentiate fresh and frozen–thawed fish. Fat extracted from six fish species, namely horse mackerel, European anchovy, red mullet, bluefish, Atlantic salmon, and flying gurnard, was analyzed using Raman spectroscopy in order to discriminate between fresh, once-frozen–thawed, and twice-frozen–thawed fish [186]. PCA models were developed and displayed a clear discrimination between the 3 states of each fish species, indicating a strong ability of this technique to rapidly detect changes in the lipid structures of fish species compared to gas chromatography, which is usually used in classical analysis.
Although NMR has been widely used to monitor changes in fish occurring during freezing and frozen storage [207], little work has been done regarding the use of this technique to differentiate between fresh and frozen–thawed fish. Recently, NMR was used to deal with freshness authentication of Atlantic salmon by analyzing metabolic changes that occur during the thawing process [19]. A PCA score plot showed distinct fresh and frozen–thawed groupings, while the discrimination ability was attributed to the formation of aspartate in the thawed salmon.
Few studies on fluorescence spectroscopy have been reported in the scientific literature, showing the possibility of the application of this technique to study different authenticity issues in seafood. For instance, the potential of front-face fluorescence spectroscopy was investigated to discriminate between fresh and frozen–thawed sea bass [208]. In this study, four fluorophores were examined, including NADH (excitation at 340 nm), tryptophan (excitation at 290 nm) riboflavin (excitation at 380 nm), and vitamin A (emission set at 410 nm). The results showed that this technique coupled with some appropriate chemometric tools was able to discriminate not only between fresh and frozen–thawed fish, but also between frozen fish of differing quality before freezing and storage.
Many studies have demonstrated the potential use of HSI for various authentication purposes [209]. Discrimination between fresh and frozen–thawed cod fillets was investigated by using VIS/NIR HSI adapted for online measurements of fish fillets moving on a conveyor belt at a speed of 40 cm/s, a rate that meets the industrial production requirements [210]. The results showed that the technique was able to differentiate between both fresh and frozen–thawed cod fillets and between the fillets according to different freezing and thawing protocols as a function of sample freeze–thaw history. In this study, the discrimination ability was attributed to variations in the visible region of the spectrum induced by oxidation of hemoglobin and myoglobin and to scattering changes caused by protein denaturation and other structural modifications during the freezing–thawing processes.
In light of the herein reviewed results, it can be noticed that various spectroscopic methods have tremendous potential for the detection of fraud and verification of several authentication issues in fish and other seafood. Our literature review revealed that the detection of species fraud and thawed fish are the most studied topics, while vibrational spectroscopic techniques, particularly NIR spectroscopy, are the most investigated techniques. Our literature review shows that few spectroscopic studies have been conducted with respect to the determination of geographical origins and detection of the modality of production (capture or aquaculture) of fish and other seafood. The low number of studies regarding authenticity issues, such as geographical origin, may be due to the difficulty associated with modeling variability in the spectra due to challenges related to many factors affecting measurements, such as biological variability, water temperature, and salinity [8,176]. Surprisingly, only a few applications of fluorescence spectroscopy have been reported, although the high sensitivity and specificity of this technique compared to the other spectroscopic techniques is well known. Therefore, fluorescence spectroscopic techniques should be investigated more extensively in future works.
#### *4.2. Meat and Meat Products*
*Meat Species Adulteration*: Meat and meat products can have a wide range of market values, depending on several factors. Among other factors, the biological origin is one of the most relevant. In fact, some animals are considered of greater value because of their renowned organoleptic characteristics; consequently, they have a higher selling price. One of the most common adulterations in meat products is the addition of the flesh of a different animal of a lower market value.
In recent years, a lot of effort has been put into developing non-destructive approaches for detecting meat adulterations. In this regard, the choice has often been spectroscopy, especially infrared spectroscopy, which limits or completely avoids any loss of sample material [27] (Table 2). Among the different flesh used as an adulterant, pork, which undesirable for several reasons [29], is probably one of the most investigated and reported materials in the literature. For instance, Kuswandi and collaborators [211] very successfully exploited FT-IR spectroscopy (equipped with attenuated total reflection cell) to detect porcine meat in beef jerky. In order to achieve this goal, the authors exploited three different classifiers, namely LDA, SIMCA, and SVM, and the best results were provided by LDA, giving a total classification rate of 100%. Beside FT-IR, NIR spectroscopy has also been widely exploited in this regard. For instance, Kuswandi et al. used NIR coupled with PLS-DA to detect pork adulteration in beef meatballs [212]. This approach provided extremely satisfying results, since the optimal classification model detected all the adulterated samples. In a similar study proposed by Rady and Adedeji [213], pork adulteration in minced beef was evaluated by NIR spectroscopy combined with PLS-DA. This research provided slightly lower but very promising results.
After pork, another common adulterant in beef meat is poultry. Several studies have used spectroscopy to detect this kind of adulteration. One example is the work from Deniz and collaborators [214], who demonstrated the possibility of using a fast and non-destructive spectroscopic technique to detect chicken or turkey in beef minced meat. In more detail, adulterated samples of different proportions (5%, 10%, 20%, 40%, and 100%) were prepared and analyzed by FT-IR combined with hierarchical cluster analysis (HCA) and PCA. The data obtained by HCA gave less information than those obtained by PCA, while different spectral bands, especially those of lipids, exhibited noticeable differences between the different meat products (beef, chicken, turkey). A similar study was proposed by Alamprese and collaborators in 2016 [215], who also investigated beef adulteration with turkey, however they inspected fresh, thawed, and cooked meat samples using NIR spectroscopy. Eventually, they used PLS-DA to identify the adulterant and were able to distinguish between samples presenting a low level of adulteration (<20%) and highly adulterated ones (≥20%).
HSI has been widely used and has shown promise in overcoming the challenges related to measurements of heterogeneous food matrices, such as muscle foods (meat, fish). For instance, Kamruzzaman et al. used this technique coupled with PCA to detect pork [216] and chicken [217] adulteration in beef. Similarly, HSI was applied to detect fraud in minced beef [218]. The data
were preprocessed by MSC and SNV, and the performance of two classification models (SVM and RF) was compared. The best results were obtained using the optimized RF model developed on selected wavelengths, achieving an accuracy of 96.87%.
One of the main advantages of HSI is the possibility to generate a distribution map, allowing the visualization of adulteration levels [14,20]. On the other hand, the data generated from HSI are extremely vast, requiring complex data handling. Multispectral imaging (MSI), however, uses a lower number of spectral bands, thus the acquisition time and complexity of MSI are comparably lower than that of HSI. MSI was successfully used recently in order to detect minced beef adulteration with horsemeat [219]. In this study, the performance of three classification models, namely PLS-DA, RF, and SVM, was explored, and the best results were obtained by the SVM model, giving a correct classification rate of more than 95%.
Beside spectroscopic methods, the traditional ones (e.g., PCR) are still widely used in this field of quality control. For example, Hou et al. used a PCR method to detect different adulterants (duck, chicken, and goose) in pork, beef, and mutton [132]. Similarly, Kim et al. used it to detect undesired donkey meat in mixtures [131]. Several similar studies have been conducted recently for the same purpose [220–222]. Very recently, Yin and co-workers proposed a novel and highly sensitive molecular assay (PCR-based) for the fast revelation of pork components at a concentration of 0.01% in adulterated meat [223]. A relatively novel technique, which is widely used to detect adulterated meat, is DNA barcoding. As an example, Xing et al. successfully exploited DNA barcoding and DNA mini-barcoding to detect mislabeling of several products on the Chinese market [125]. In addition to the previously mentioned approaches, ELISA is another common tool used for species identification in food authentication. For example, it has been used to detect pork-adulterated beef by Mandli and collaborators [141], whereas Perestam et al. compared the performance of the ELISA and of PCR for detecting beef and pork—both approaches have advantages and disadvantages for this purpose [149].
**Table 2.** Examples of applications of spectroscopic techniques with respect to various authenticity issues in meat and meat products.
PCA, Principal Component Analysis; PCR, Principle Component Regression; LDA, Linear Discriminant Analysis; DA, Discriminant Analysis; QDA, Quadratic Discriminant Analysis; RF, Random Forest; SIMCA, Soft Independent Modeling of Class Analogy; PLS-DA, Partial Least Squares Discriminant Analysis; PLSR, Partial Least Squares Regression; LS-SVM, Least Squares Support Vector Machines; VIS/NIR, Visible–Near-Infrared Spectroscopy; HSI, Hyper Spectral Imaging; FT-IR; Fourier-Transform Infrared Spectroscopy; (D)CNN, (Deep) Convolution Neural Networks.
*Distinction Between Fresh and Thawed Meat*: Beside adulteration with undesired meats, scams concerning meat freshness are unfortunately common, and consequently in the literature it is possible to find different studies aiming to detect this kind of fraudulent action. It is not always
easy to discern the freshness of meat by sight, and mislabeling can occur accidentally or intentionally to make illicit profits by selling thawed meat as fresh. Regardless of the reason, it is important to possess suitable tools for the authentication of fresh meat. Once again, in recent years, spectroscopy has played a key role in the detection of this kind of fraud.
One of the meats investigated the most in this context is chicken, mainly because of the few visual differences that differentiate fresh and thawed products. Nevertheless, Grunert and collaborators have suggested that discrimination can be achieved by FT-IR spectroscopy; in fact, in their study they showed the possibility of using this technique coupled with artificial neural networks (ANN) to discern fresh and thawed samples (frozen and stored for time periods from 2 up to 85 days) [237]. The results were extremely satisfying, since twenty samples (of the twenty-one investigated) were correctly classified. A similar study was proposed by Parastar and collaborators, where fresh and thawed chicken samples were analyzed using a portable NIR instrument and then classified by different methods (random subspace discriminant ensemble (RSDE), PLS-DA, ANN, and SVM); the best results were obtained by using RSDE, providing extremely satisfying results with a classification accuracy higher than 95% [18].
*Detection of the Geographical Origin and Production Method*: The traceability of meat and meat products is relevant from different standpoints; for this reason, several approaches have been proposed to assess the origins of meat samples [238]. Traditionally, meat and meat products are traced by means of protein- and DNA-based methods [239]. An example is a recently published paper by Muñoz and collaborators, who focused on Iberian pork meat, which is used to prepare a Spanish typical cured meat product [240]. The authors proposed a single nucleotide variant genotyping panel suitable for recognizing purebreds (Duroc and Iberian) or crossbreds. Interesting solutions for the origin assessment of edible meats were also provided by means of stable isotope ratio analysis. For instance, Erasmus and co-workers showed that δ <sup>15</sup>N and δ <sup>13</sup>C can be used to discriminate South-African lamb breeds in diverse regions [241]. These authors related the isotope abundancies to the pedo-climatic conditions of the different areas. A similar study on a diverse animal species was conducted by Monahan et al., who investigated the possibility of using stable isotope ratio analysis to recognize Irish chickens [242]. Further applications can be found in [243].
Despite the tools mentioned above providing noteworthy outcomes, they are time-consuming, destructive, relatively expensive, and require complex sample preparation. During the first decade of this century (2000–2010), a lot of effort has been put into developing fast and non-destructive spectroscopy-based approaches to achieve the same purpose. However, during the last five years, not many novel strategies have been proposed. For example, recently Zhang and co-authors demonstrated that FT-IR spectroscopy integrated with second derivative infrared spectroscopy (SD-IR) and two-dimensional correlation infrared spectroscopy (2DCOS-IR) coupled with computer vision methodologies represent suitable choices for discrimination of different hams produced in three different locations [244].
There are few studies on the potential of spectroscopic techniques for the determination of the production method (dietary background) of meat. One example is a study conducted by Huang and co-authors [245], who applied reflectance spectroscopy in two spectral ranges (400–700 nm and 400–2500 nm) coupled with PLS-DA to discriminate carcasses of lambs reared with 3 feeding regimes, involving perirenal fat from pasture-fed, concentrate-fed, and concentrate-finished after pasture feeding diets. The results demonstrated that the 3 feeding regimes could be distinguished with overall correct classification rates of 95.1% and 99% for the 400–700 nm and 400–2500 nm spectral ranges, respectively.
*Other Common Adulterants or Contaminants in Meat*: A number of foreign ingredients can be introduced (voluntarily or accidentally) in meat and meat products. Some contaminants can be unintentional, while others are conceived to alter the characteristics of the treated food in order to make it more palatable to the consumers. For instance, the addition of food dyes in meat products is allowed by law, but the types of colorants are strictly regulated; consequently, the possible presence of forbidden dyes has to be checked [243]. Other forms of fraud in meat may involve unwanted or forbidden physical pretreatments, as is the case with irradiation. This practice, which is generally used to extend the shelf-life of food products, is allowed for some foods (for instance dry aromatic herbs) but it is banned for meat. As a consequence, different research studies have been conducted with the aim of developing analytical approaches suitable for the detection of this illicit practice, as in the case discussed by Varrà and co-authors, where irradiated and non-irradiated sausages were discriminated by NIR spectroscopy coupled with orthogonal partial least square–discriminant analysis (OPLS-DA) [246].
One further illegal practice is fraudulent mislabeling, consisting of substituting a high-value cut meat with a cheaper alternative, as in the case reported by Sanz and his group [247]. In their study, the authors investigated four different types of lamb muscles using HSI and discriminated the four diverse categories using seven classifiers. The most accurate outcome was achieved using linear least mean squares, which led to a total correct classification rate of 96.67%.
Only limited research has been found in the literature about the use of fluorescence spectroscopy for studying authenticity issues in meat and meat products. In one of the scarce studies, FFFS combined with chemometric tools (PLS and PLS-DA) was successfully applied to classify three different beef muscles, namely *the semitendinosus*, *rectus abdominis*, and *infraspinatus* muscles [248]. These results were confirmed recently in a similar study [95]; in this study, FFFS achieved better accuracy in discrimination of beef muscles than synchronous fluorescence spectroscopy.
#### *4.3. Milk and Dairy Products*
Thanks to its enhanced nutritional value provided by the presence of high-quality protein and minerals, milk is an essential food for people of all ages, from infants to elderly people [249]. Adulteration of milk by the addition of undeclared substances is a widely encountered problem in the dairy industry. Whey, melamine, starch, water, chlorine, formalin, and hydrogen peroxide are the most frequently used adulterants for this type of practice. Mixing milk from different species, replacement of milk fat with non-milk fats or oils, labelling a conventional product as an organic farming product, and false declaration of the processing technology and geographical origin are the other primary fraudulent practices. Several physicochemical methods, liquid and gas chromatography, isotope ratio analysis, and DNA-based techniques have been used for these issues, which involve drawbacks such as having a high cost and being labor-intensive. Spectroscopic techniques (Table 3), being rapid, easy to operate, and applicable to on-line and at-line measurements, as well as providing a high amount of data, are alternatives that can be used to overcome the disadvantages of existing methods [250].
*Addition of Non-Declared Substances*: Urea, melamine, dicyandiamide, sodium bicarbonate, ammonium sulfate, and sucrose are the most frequently used adulteration agents for milk and dairy products [251,252]. Infrared spectroscopy, FT-MIR, and MIRS have been widely applied to determine raw milk and milk powder adulteration by using waste whey [253,254]. In a comprehensive study by Coitinho et al. [67], the FT-IR MilkoScan FT1 device was calibrated and validated using a large number of raw milk samples. Then, the sensitivity (80–90%) and specificity (80–100%) of the method were designated for adulteration of raw milk with different adulterants. Several NIR spectroscopic methods have been utilized to detect milk and milk powder adulteration [255]. In a recent study, a non-targeted method employing benchtop FT-NIR and portable NIR devices coupled with SIMCA was developed to determine eleven potential adulterants in milk powder. The portable device provided lower sensitivity and specificity due to its lower spectral resolution and narrower spectral range [256].
**Table 3.** Examples of applications of spectroscopic techniques with respect to various authenticity issues in milk and dairy products.
PCA, Principal Component Analysis; LDA; Linear Discriminant Analysis; DA, Discriminant Analysis; SIMCA, Soft Independent Modeling of Class Analogy; PLS-DA, Partial Least Squares Discriminant Analysis; PLSR, Partial Least Squares Regression; <sup>1</sup>H NMR, High-Field Nuclear Magnetic Resonance; 2D-NMR, Two-Dimensional Nuclear Magnetic Resonance; FT-IR, Fourier-Transform Infrared Spectroscopy; HCA, Hierarchical Cluster Analysis; (D)CNN, (Deep) Convolution Neural Networks; k-NN, k-Nearest Neighbors; Q-control, Control Chart Q; GA-LDA, Genetic Algorithm Linear Discriminant Analysis; 2DCOS-SFS, Synchronous Fluorescence Spectroscopy coupled with Two-Dimensional Correlation Spectroscopy.
Raman spectroscopy is another vibrational spectroscopic technique that has been widely investigated for adulteration purposes. For example, a portable Raman spectrometer was employed to detect melamine, dicyandiamide, urea, ammonium sulfate, and sucrose adulteration of milk. The standard error of prediction and relative standard deviation values were 39 to 72 ppm and 8% for nitrogen-rich compounds, and 1400 ppm and 10% for sucrose, respectively. The selectivity and efficiency values were 100% for the PLS-DA model in discriminating pure milk samples from adulterated ones [267]. The obtained results were found to be comparable with those of a previous study of the same group, in which a Raman microprobe system was employed [268]. Considering the high-throughput Raman chemical-imaging-based method, it was possible to visualize the spatial distributions of melamine and urea in milk powder and quantify these at the 50 ppm level [82]. Moreover, vegetable oils that were fraudulently added to dairy cream and yogurt were detected by Raman spectroscopy [70,260]. Finding alternative sample preparation procedures is an essential point to be highlighted for efficient Raman spectroscopic analysis in milk and dairy products. Nedeljkovi´c et al. [269] performed a preheating process to butter and margarine samples before Raman measurements. In a recent study, the successful use of a portable Raman spectrometer to assess lard adulteration in butter was reported. Samples were melted and mixed thoroughly prior to the Raman measurements [69]. Lohumi et al. developed a line scan spatially offset Raman spectroscopy (SORS) technique that can collect data from packaged butter and margarine samples [270].
*Detection of Species Fraud*: Successful discrimination and quantification of milk from undeclared species have been carried out using infrared spectroscopy [271]. Equivalent promising results were reported with Raman spectroscopy [272]. Nonetheless, it is important to emphasize the fluorescence interference problem during Raman spectroscopy measurements, especially with 532 nm lasers. Studies employing lasers with different wavelengths (e.g., 785 and 1064 nm) have extended the use of this technique for milk and dairy product analyses.
There have been very few studies in the literature reporting the use of NMR for the determination of adulteration. Nonetheless, one study succeeded in discriminating soymilk, bovine milk, goat milk, and their adulterants after coupling chemometrics and metabolite analysis using 1D- and 2D-NMR,
with limit of quantification values ranging between 2% and 5% [273]. Some other studies highlighted the changing sensitivity and specificity of the <sup>1</sup>H time-domain NMR (TD-NMR) method, depending on the used adulterant [81,257].
The identification of milk species by employing different measurement techniques involving fluorescence spectroscopy has been studied by several authors [16,274]. Boukria et al. [261] highlighted that cow milk adulteration in camel milk could be detected through the application of the two-dimensional correlation spectroscopy method on SFS spectra. Inclusion of a higher number of samples in the calibration model and scanning of a more comprehensive wavelength range were emphasized as determinant factors in obtaining satisfying discrimination results.
The successful use of several DNA-based analytical methods has been reported for milk authentication and traceability in the dairy sector [275]. In recent applications, entirely satisfactory limit of detection values were achieved [276,277]. Efforts have been made to develop low-cost and user-friendly PCR devices with accuracy and stability comparable to commercial alternatives [278]. Commercial PCR-based assays designed for the detection and quantitative authentication of animal species in a specific dairy product are also available in the market [279,280].
*Identification of Geographical Origin and Production Method*: Over the last five years, various studies have been reported regarding the authentication of Mozzarella di Bufala Campana Protected Designation of Origin (MBC-PDO) cheese. For example, to combat fraud, Bontempo et al. [281] have successfully proposed the use of the stable isotope method combined with elemental analysis to differentiate both milk and cheese products produced in the PDO area from other products produced outside the PDO area. In another study, Salzano et al. [282] demonstrated that it was possible to distinguish MBC-PDO milk and cheese from non-MBC-PDO products using an advanced GC-MS method and metabolite identification.
Concerning spectroscopic techniques, most of the reported studies were performed in the infrared wavelength range. In more detail, Caredda et al. [264] showed that MIR correctly identified 99% of the ewe's milk from different geographical regions. In another study, Liu et al. [283] conducted a study to assess the interest in a portable micro-NIR spectrometer to discriminate organic milk from pasture and conventional milk. It was shown that the micro-NIRS device could distinguish between organic and conventional milk as efficiently as the FT-NIRS device (i.e., laboratory device).
The abovementioned studies prove how frequently spectroscopic techniques are used to detect adulteration of milk and dairy products. Nonetheless, there is an imbalance in use between the different available spectroscopic techniques. Vibrational spectroscopy has been clearly the most preferable applied method used to detect and identify the most common adulterants in milk. However, more studies comparing the performance of NIR, MIR, and Raman spectroscopy for detecting adulteration of milk samples are necessary. Based on the existing literature, it can be noticed that Raman spectroscopy has particular potential for use for routine analysis of milk and dairy products. However, there is still a need for further studies investigating the simultaneous use of adulterants and extending the scope by developing novel untargeted approaches. Regarding the identification or authentication of milk and dairy products based on their geographical origin and processing treatments, surprisingly only a few studies were conducted during the last five years using spectroscopic techniques. This conclusion is similar to that discussed above for fish and meat products. Thus, the use of spectroscopic techniques for differentiation of fresh and frozen–thawed milk and dairy products and investigation of the effects of the applied processes (milk preparation, cheese processing, etc.) or storage conditions that are important for compliance with specifications (such as PDO, protected geographical indication, etc.) are some of the issues that need to be further studied.
#### *4.4. Honey and Other Products of Animal Origin*
Honey is a natural sweet product made by bees from the nectar of plants or plant excretions combined with bees' own specific substances and maturated in the honeycomb. The characteristic flavor, nutritional value, and health benefits of honey depend on its origin and
production methods. As a high-quality food product with a high price, honey is often subjected to fraudulent practices, which include mislabeling and adulteration. Development of methods for assessing honey authenticity is of interest to consumers, the honey industry, and food law agencies. Several papers have reviewed the methods used for honey analysis [30,284–288].
*Botanical Origin*: The price of honey strictly depends on its botanical origin. According to botanical origin, honey is classified as unifloral, multifloral (polyfloral), and honeydew [30]. The monofloral honeys are often more expensive than multifloral honeys and are subject to mislabeling or adulteration with cheaper honeys [289].
The most used conventional method for determining honey quality related to its origin is melissopalynological analysis based on the identification and quantification of pollen grains in honey sediment [30]. The physicochemical (profiles) parameters, such as sugars, moisture, proline, and hydroxymethylfurfural (HMF) contents; acidity; electrical conductivity; diastase; and invertase activity are used to establish the origin of a honey. Analytical techniques including gas and liquid chromatography are often used to measure markers of honey origin, such as sugar, phenolic compounds, and flavor compounds. The profiling techniques, stable isotope ratio, and trace element analysis can provide an indication of the geographical origin of honey. The identification of plant species and varieties of honey by DNA fingerprinting is also utilized to assess honey origin.
Spectroscopic techniques have shown considerable potential as rapid and often non-destructive methods used to study the authenticity of honey. In recent years, several studies have demonstrated the potential use of various spectroscopic techniques for evaluation of the botanical origin of honeys (Table 4). For example, NIR spectroscopy and chemometrics were applied to palynological and mineral characteristics of honey collected from Northwestern Spain [290]. Prediction models using a modified PLSR for the main pollen types (Castanea, Eucalyptus, Rubus, and Erica) in honeys and their mineral compositions were established. The ratio of performance to deviation exhibited a good prediction capacity for Rubus pollen and for Castanea pollen, whereas these ratios were excellent for minerals, Eucalyptus pollen, and Erica pollen.
The benefit of data fusion obtained using different analytical techniques was demonstrated for classification tasks of honey according to the botanical origin. The honey samples from three different botanical origins were analyzed by attenuated total reflection IR spectroscopy (ATR/FT-IR) and headspace gas chromatography–ion mobility spectrometry (HS-GC-IMS) [291]. The obtained datasets were combined in a low-level data fusion approach with subsequent multivariate classification by principal component analysis–linear discriminant analysis (PCA-LDA) or PLS-DA. The results showed that data fusion is an effective strategy for improving the classification performance.
Raman spectroscopy techniques complement information obtained from infrared spectral data and can be used in honey authenticity assessment [287]. Raman spectroscopy, performed using fiber optics, was successfully used to distinguish the botanical origin of unifloral (chestnut, citrus, and acacia) honeys produced in the Italian region of Calabria [292]. Moreover, predictive models were built to quantify important marker indicators in nutraceuticals, such as the main sugars, potassium, and selected sensory properties.
A promising quick, automatic, and non-invasive approach for honey botanical origin classification was developed using a combination of VIS/NIR hyperspectral imaging and machine learning, namely SVM and k-NN [24]. The developed techniques include noisy band elimination, spectral normalization, and hierarchical classification. The proposed model showed promising results under several classification scenarios, achieving high classification performances.
The blending of expensive (pure and rare) honey with a cheaper (pure and plentiful) one is another form of honey adulteration. NMR spectroscopy allows the rapid detection of adulterants in honey, as well as the simultaneous quantification of various chemical compounds from a spectrum [287]. For example, <sup>1</sup>H NMR spectroscopy combined with chemometric techniques was applied to detect and quantify adulteration of acacia honey with cheaper rape honey [293]. The highest prediction accuracy for rape honey addition of −89.7% was obtained using canonical discriminant analysis (CDA), determined from compounds located in the spectral range corresponding to the aliphatic compounds and carbohydrates (3.00–6.00 ppm). Orthogonal projection to latent structure discriminant analysis (OPLS-DA) was used to further discriminate samples of pure acacia honey adulterated with different amounts of rape honey. A PLSR model established a linear fit between the actual and predicted adulterant concentrations, with an R<sup>2</sup> value of up to 0.9996.
The fluorescence of honey originates from several groups of compounds, such as amino acids, proteins, phenolic acids, vitamins, fluorescent Maillard reaction products, and other bioactive molecules [23,102,294]. Few studies have demonstrated the potential of fluorescence for authenticity assessment. Fluorescence spectroscopy in EEM mode coupled with parallel factor analysis (PARAFAC) and PLS-DA was applied for classification of honey samples of different botanical origin, including acacia, sunflower, linden, meadow, and fake honey [100]. The classes of honey of different botanical origin were differentiated mainly by emissions from phenolic compounds and Maillard reaction products. PLS-DA constructed from the PARAFAC model provided detection of fake honey samples with 100% sensitivity and specificity. Moreover, PLS-DA classification results gave errors of only 0.5% for linden, 10% for acacia, and about 20% for both sunflower and meadow mixes.
**Table 4.** Examples of applications of spectroscopic techniques with respect to various authenticity issues of honey.
PCA, Principal Component Analysis; LDA, Linear Discriminant Analysis; SIMCA, Soft Independent Modeling of Class Analogy; PLS-DA, Partial Least Squares Discriminant Analysis; PLSR, Partial Least Squares Regression; SVM, Support Vector Machines; VIS/NIR, Visible–Near-Infrared Spectroscopy; NMR, Nuclear Magnetic Resonance; FT-IR, Fourier-Transform Infrared Spectroscopy; HCA, Hierarchical Cluster Analysis; CDA, Canonical Discriminant Analysis; OPLS-DA, Orthogonal Projection to Latent Structure Discriminant Analysis; iPLS, Interval Partial Least Squares; HPLC-DAD, High-Performance Liquid Chromatography with Diode Array Detection.
*Adulteration Detection*: Honey is a natural product for which the addition of any other substance is prohibited by international regulations. However, due to its high economic value, it is often subject to adulteration. The most common adulterants in honey are sugars from high-fructose corn syrup, corn sugar syrup, inverted sugar syrup, and cane sugar syrup [287]. Adulteration of honey is not limited to direct addition of sugars into natural honey. A common fraudulent practice is overfeeding of bees with concentrated sugar solutions during the main nectar flowing season [30]. Among analytical methods, spectroscopic techniques have become popular for detecting the adulterants in honey [287].
FT-IR and PLSR were utilized for the determination of sucrose syrup adulteration of Turkish honeys [301]. The results indicated that the predicted sucrose concentration of honey samples by the spectroscopic method ranged between 4.52 and 15.16%, and that the obtained results were confirmed by chromatography. Several studies reported successful applications of NIR or VIS/NIR spectroscopy for evaluation of honey adulteration. For example, NIR spectra (1300–1800 nm) recorded with a fiber optic immersion probe were used for the detection of high-fructose corn syrup in four artisanal Robinia honeys [302]. The PLSR models developed using the spectral region containing absorption bands related to both water and carbohydrates allowed accurate (root mean squared error of cross-validation; RMSECV = 1.48; R<sup>2</sup> CV = 0.987) detection of the adulterant concentration. Recently, NIR and MIR spectroscopy coupled with SVM and data fusion were utilized to detect adulteration of 20 common honey types from 10 provinces in China [303]. Both pure honey and adulterated samples with different percentages of syrup were analyzed. Compared to low-level data fusion, intermediate-level data fusion significantly improved the detection model, achieving 100% accuracy, sensitivity, and specificity.
Fluorescence excitation–emission spectroscopy was effectively used for the non-destructive and fast detection of fake honey samples obtained during winter feeding of bee colonies with a sucrose solution [99]. Natural honey samples (acacias, lindens, sunflowers, and meadow mixes) were perfectly discriminated from fake honey samples using the developed LDA model. Natural and adulterated honey samples differed significantly in five spectral regions corresponding to aromatic amino acids, phenolic compounds, furosine, and Maillard reaction products.
Eggs are consumed worldwide and are well known as a source of vitamins, minerals, phospholipids, and high-quality proteins. EU regulation classifies egg production into four hen housing systems, including 0 for organic production, 1 for free range, 2 for barns, and 3 for cages. Consumers are willing to pay higher prices for eggs produced in a way that considers animal welfare [304], and chicken eggs are often a subject of food fraud. Therefore, there is a need for analytical methods that are suitable for classifying eggs and for detecting the fraudulent mislabeling of eggs obtained from different production systems.
Various procedures are used to discriminate eggs, including carotenoid profiling, fatty acid composition, and mineral content procedures. Eggs from various systems (1-, 2-, and 3-coded eggs) may be discriminated through fluorescent patterns on egg surfaces or stable nitrogen isotope compositions. Stable isotopes methods were used to develop authentication criteria of eggs laid under cage, barn, free range, and organic farming regimens [305]. Recently, discrimination of selected chicken eggs in China's retail market based on multielement and lipidomic analyses was reported [306].
UV-VIS/NIR spectroscopy and chemometrics were utilized for a complete detection of the housing systems declared on the eggs' label [307]. Eggs were perfectly classified into the four housing systems by applying quadratic discriminant analysis for UV-VIS/NIR spectra of the yolk lipid extracts. NMR spectroscopy was successfully utilized as a tool to screen eggs according to the different systems of husbandry [304]. In this study, <sup>1</sup>H NMR of freeze-dried egg yolk samples were analyzed using PCA followed by a linear discriminant analysis (PCA-LDA). The prediction model allowed for the correct classification of about 93% of the organic eggs, barn eggs, and free range eggs.
#### **5. Challenges and Future Trends**
Even though extensive research regarding the authenticity and detection of fraud by on-site and real-time approaches has been carried out in recent years, several key challenges still remain concerning both technique-related issues and the model validation framework.
Regardless of the non-destructive approach considered, the correct sampling procedure is pivotal to provide valuable information, and thus to embrace the complexity of modern food authentication [308]. Indeed, non-destructive approaches include non-targeted methods (i.e., fingerprint techniques) with the ability to detect multiple small modifications in the considered food product and to extract these modifications as relevant information using the proper multivariate statistic approach. However, the database used to address the authentication issue should consider the main sampling-related criteria, such as the definition of the sample unit, number of samples, sample variability, handling procedure, representativeness, and so on. The most important considerations that must be addressed when creating
a food authenticity database are discussed in the position paper by Donarski and co-authors [309]. These issues are highly relevant, as the database is used to define an "authentication rule", which is applied to compare the unknown sample fingerprint with those of authentic reference samples [308]. Even though the creation of the foodstuff-specific database was done considering the perfect sampling procedure and can quickly cover the variability expected from test samples, continuous maintenance of the database is needed to ensure long-term ability to return reproducible results, and most of the scientific publications do not meet this requirement.
Once the authentication issue has been defined and the database creation has been designed accordingly, consideration needs to be given to the definition of a standard operating procedure (SOP) from the sample preparation to the analytical protocol.
DNA-based methods, protein-based methods, and isotopic techniques require specific consideration when defining SOP. Indeed, in these cases, the required analytical steps for sample preparation highly influence the results and their interpretation [2]. As for any analytical technique, different experimental factors can influence the obtained results, introducing an analytical deviation that is not related to the authentication issue under study. These deviations should be reduced to the lowest terms and controlled to ensure that they do not introduce confounding results in the analysis [309]. The influence of experimental factors cannot be avoided, even in spectroscopic technologies (e.g., vibrational spectroscopy, NMR, and fluorescence spectroscopy), despite being reproducible and barely influenced by changes in sensitivity over time. Indeed, they do not generally require any sample preparation, guaranteeing long-term stability and online or in-line application along the production chain. This is particularly true for liquid "homogeneous" samples, whereas solid heterogeneous products, such as meat, fish, and dairy products, may require moderate sample preparation or multiple point measurements. Moreover, the choice of the proper acquisition mode is fundamental to obtain reliable spectroscopic results according to the nature of the food product, including the type of radiation (NIR, IR, NMR, or fluorescence spectroscopy), sample presentation (transmission, absorbance, reflectance, excitation or emission fluorescence, synchronous fluorescence, EEM), type of sample holder (cuvette, fiber probe, attenuated reflectance holder, integration sphere), and working temperature, among others.
Actually, HSI technologies are a valid alternative to point spectral scanning, whereby the spatial distribution of components in heterogeneous products can be distinguished using site-to-site spectroscopic fingerprint specificity. Food quality and authenticity, especially referring to meat products, have been widely investigated by HSI technologies associated with NIR radiation. However, most of the reported works are feasibility studies at the laboratory scale, whereas there is a lack of studies proving the model's robustness at the processing plant level. Furthermore, the huge disadvantages of HSI technology are related to the large amount of produced data for each single measure and the relatively long processing times for these data. However, simplified instruments (multispectral imaging systems) developed for specific applications could reduce the spectral range to be scanned to a few selected wavelengths, thus minimizing both the acquisition time and generated data, which could be managed quickly with the proper ad hoc chemometric method [310]. Simplified, miniaturized, and portable instruments have been developed for the whole spectroscopic field, which are oriented toward food authentication [311]. Certainly, the performance of these instruments in terms of the electromagnetic range covered, resolution, signal-to-noise ratio, specificity, and sensitivity is lower if compared to the results obtained by benchtop instruments [198]. However, their use for ad hoc authentication purposes and their combination with robust chemometric algorithms for classification applications are expected to be major trends in the coming years.
As described in Section 2, multivariate data analysis is the fundamental step taken to produce a model able to classify samples as authentic or non-authentic from any emerging detection method result. No matter the algorithm used to solve an authentication issue, robust validation of the model is mandatory to guarantee reliable and reproducible results and to favor the acceptance of these methodologies in legislation. This theme is quite contentious, and it is one of the major reasons for the refusal of emerging detection methods, along with the standardization procedures [2,60]. Although several attempts have been made to meet the need for common and reliable validation protocols, there is still a lack of validation programs for method developers, which is also reflected in the scientific literature. In the paper by Oliveri [46], a detailed analysis of the key aspects of model evaluation is discussed. This paper could be a landmark when defining a global workflow to solve an authentication issue using spectroscopic techniques.
Thus, it is undeniable that spectroscopic techniques have enormous advantages over the targeted approaches when addressing a food authentication issue; however, their wide application outside of laboratories remains challenging. Meeting these challenges will align emerging spectroscopic methods with the needs of food fraud risk management systems, paving the way for their use for food integrity assurance, such as with the EU-wide Rapid Alert System for Food and Feed (RASFF).
#### **6. Concluding Remarks**
This paper has reviewed and discussed papers published in the last 5 years on the use of different analytical methods used to target issues related to fraud in both food and products of animal origin. The available literature in the field has shown an increase in the number of applications combining rapid analytical methods (e.g., DNA analysis, vibrational spectroscopy) with modern data analytics (e.g., multivariate data analysis). The body of research as a whole presents indisputable evidence that these methods and techniques have enormous advantages over other approaches when addressing food authentication. However, several challenges still exist related to the wide application and implementation of these technologies in both research and commercial laboratories. This calls for the need for a continuous exchange between the food authentication stakeholders, together with the growth of a new generation of scientists able to work in both academic and industrial environments and who are skilled in facing all aspects of food authentication using non-targeted techniques.
**Author Contributions:** Conceptualization, methodology, writing—original draft preparation, A.H.; writing original draft preparation, I.M.; writing—original draft preparation, revision, W.F.S.; writing—original draft preparation, H.T.T.; writing—original draft preparation, L.L.; writing—original draft preparation, H.-Y.K.; project administration, supervision, manuscript revision, H.N.; writing—original draft preparation, A.B.; writing—original draft preparation, A.A.-K.; writing—original draft preparation, M.S.; writing—original draft preparation, E.S.; writing—original draft preparation, S.G.; writing—original draft preparation, D.C. All authors have read and agreed to the published version of the manuscript.
**Funding:** This research received no external funding.
**Acknowledgments:** This work was supported by the Norwegian Institute of Food, Fisheries, and Aquaculture Research (Nofima) through a Strategic Research Initiative (Spectec Project): Rapid and Non-Destructive Measurements to Enable Process Optimization.
**Conflicts of Interest:** The authors declare no conflict of interest.
#### **References**
© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
*Article*
## **Flash Gas Chromatography in Tandem with Chemometrics: A Rapid Screening Tool for Quality**
#### **Sara Barbieri <sup>1</sup> , Chiara Cevoli <sup>1</sup> , Alessandra Bendini 1,\* , Beatriz Quintanilla-Casas 2,3 , Diego Luis García-González <sup>4</sup> and Tullia Gallina Toschi <sup>1</sup>**
**Grades of Virgin Olive Oils**
**\*** Correspondence: [email protected]; Tel.: +39-0547-338121
Received: 11 June 2020; Accepted: 29 June 2020; Published: 2 July 2020
**Abstract:** This research aims to develop a classification model based on untargeted elaboration of volatile fraction fingerprints of virgin olive oils (*n* = 331) analyzed by flash gas chromatography to predict the commercial category of samples (extra virgin olive oil, EVOO; virgin olive oil, VOO; lampante olive oil, LOO). The raw data related to volatile profiles were considered as independent variables, while the quality grades provided by sensory assessment were defined as a reference parameter. This data matrix was elaborated using the linear technique partial least squares-discriminant analysis (PLS-DA), applying, in sequence, two sequential classification models with two categories (EVOO vs. no-EVOO followed by VOO vs. LOO and LOO vs. no-LOO followed by VOO vs. EVOO). The results from this large set of samples provide satisfactory percentages of correctly classified samples, ranging from 72% to 85%, in external validation. This confirms the reliability of this approach in rapid screening of quality grades and that it represents a valid solution for supporting sensory panels, increasing the efficiency of the controls, and also applicable to the industrial sector.
**Keywords:** virgin olive oil; quality; volatile compounds; sensory analysis; chemometrics
### **1. Introduction**
The official methodology for sensory evaluation of virgin olive oils (VOOs), known as a panel test, is a fundamental tool to assess the quality of products that cannot be replaced by instrumental methods, considering that the overall and complex perceptual attributes (e.g., fruity and defects) are the indicators of the quality of VOOs. Despite its proven effectiveness in evaluating the quality grades of samples, tested in EU countries since 1991 [1,2], the scientific community has highlighted some drawbacks on its application that are mainly related to the following: (i) the reproducibility of results among different panels; (ii) critical attribution of the category when, e.g., a defect is borderline; (iii) costs, assessor fatigue and other limitations associated with a method working with humans.
Specifically, according to decisions taken at International Olive Council (IOC) level, the Reg. (EU) 1348/2013 [3] recommends the number of oils to be assessed by the sensory panels, fixing a maximum number of four samples at each session. Moreover, a maximum of three sessions per day is specified, to leave enough time between a session and another, thus avoiding the contrast effect that could be produced by immediately tasting sequences of samples. These specifications strongly limit the number of samples that can be assessed by one panel per day. On the other hand, to enhance panel skills in recognizing, identifying, and quantifying sensory attributes, the introduction of new artificial reference materials (obtained by chemical or biotechnological approaches), could improve the proficiency of the individual panels and their global alignment by overcoming some limitations associated with a natural matrix (e.g., limited amounts available, difficultly obtaining, low homogeneity year by year) and offering advantages such as preparation in each laboratory, reproducibility over time, possibility of purchase, and therefore their availability for the market.
In this context, the development of an instrumental method for rapid screening of quality grades of samples (extra virgin olive oil, EVOO; virgin olive oil, VOO; lampante olive oil, LOO) could represent a solution to support sensory panels (particularly for large private industries), decreasing their daily work by reducing the samples that need to be assessed (e.g., by excluding those definitely compliant), with a consequent increase in the efficiency of quality controls and reducing the number of samples that need to be controlled.
In this way, improvement of the activity of sensory panels, whose work remains central to ensuring the quality of the product, would be achieved by focusing sensory analysis only on uncertain samples (i.e., borderline oils between two product categories that can be the object of disagreement among panels).
It is well known that volatile compounds are crucial to determine VOO quality and that they are responsible for the different VOO sensory profiles [4–6]; their determination in a rapid way (e.g., screening method) could support sensory analysis and represents one of the current challenges in the olive oil sector where fast, accurate, and easy-to-use approaches providing real-time results are required.
Recently, different analytical techniques combined with chemometric statistical approaches have been proposed to predict sensory information [7–9].
Alongside the traditional techniques (targeted) in which specific and selected molecular markers are monitored during the analysis to assess the presence or absence of compounds and their quantification, untargeted analyses, based on a holistic approach and able to provide information such as a spectral fingerprint, giving a simplified and overall picture of the food under analysis, have gained an increasing relevance over the last years [10].
Among the latter, different analytical methods for determination of volatile compounds combined with multivariate chemometric techniques for VOOs quality testing have been described in the literature and proposed to the industrial sector as fast and high throughput screening techniques [9,11–18].
In particular, as an alternative to headspace gas chromatography-mass spectrometry (HS-GC/MS), which is the most widely used technique to quantify and characterize the profiles of volatile compounds of VOOs thanks to its high sensitivity and selectivity, the application of the HS-GC ion mobility spectrometry (HS-GC-IMS) has been proposed. This technique combines high selectivity and sensitivity with high robustness and cost-efficiency, and has given promising results in discriminating VOOs according to quality grades [9,11,12,14,18] or geographical origin [13,15].
The need to support organoleptic analysis was also reported in a specific call of the Horizon 2020 EU program (H2020-SFS-14a-2014) and is one of the main objectives of the OLEUM project (Horizon 2020, Grant Agreement No. 635690). In the framework of this project, two analytical instrumental techniques, headspace-solid phase micro extraction–gas-chromatography/mass spectrometry (HS-SPME-GC/MS) and flash gas chromatography (FGC) based on the determination of volatile compounds, have been proposed as the most promising rapid screening methods that can support sensory panels in the determination of quality grades.
In a recent work by Quintanilla-Casas and co-authors (2020) [17], the results obtained with HS-SPME-GC/MS with a fingerprinting approach to classify VOO categories has been demonstrated. Herein, a classification model based on minor fraction fingerprints that is able to predict the commercial
category of olive oil samples (EVOOs, VOOs, LOOs) obtained by FGC is presented. The FGC is an innovative analytical approach for analysis of volatile compounds of VOOs based on the FGC separation: the headspace of VOOs, previously conditioned, is sampled by a syringe, the volatile organic compounds are adsorbed on a Tenax trap and subsequently desorbed by rapid heating, and, finally, transferred to a FGC step. The elution of analytes runs in parallel using two metal capillary columns with different polarity of the stationary phase. This gives rise to slight differences in the separation capability of molecules that are detected by a flame ionization detector (FID) located at the end of each column.
The main advantage of the FGC technique is its short analysis time (total separation time is 100 s); moreover, its application associated with sensory analysis for calibration and chemometric tools is promising to support the work of panel tests in discriminating samples of different product categories. A classification model, once built, could be easily applied in any laboratory or industry.
The effectiveness of this technique is already demonstrated by previous works aimed to differentiate VOOs according to their geographical origin declared by labels such as "100% Italian" and "non-100% Italian" oils [19] or "EU" and "extra-EU" [16].
The aim of this study was to classify VOOs according to quality grade, combining FGC data with the multivariate classification technique partial least squares discriminant analysis (PLS-DA). To provide robustness to our model, a set of 331 oils belonging to the three different commercial categories (EVOO, VOO, LOO) involving two harvesting/production years was analyzed. The adopted validation protocol (repeatability and reproducibility tests) and related performance are also shown.
#### **2. Materials and Methods**
#### *2.1. Olive Oil Samples*
An initial set of 334 EVOOs, VOOs, and LOOs oils representative of the most common olive cultivars, geographical origin, sensory positive attributes, and sensory defects were sampled. Specifically, in addition to a first set of 180 oils collected during the first year of the OLEUM project (2016–2017 olive season), another set of 154 samples (2017–2018 olive season) was collected and analyzed during the second year (Tables S1–S4 in the Supplementary Materials).
The panel test method was carried out by six panels involved in the OLEUM project as described by Barbieri et al. 2020 [20] and sensory data were expressed as mean of medians. The procedure deals with possible disagreement between panels with a decision tree in order to have definitive classification of samples in which definitive agreement is reached. In agreement with the sensory results reported in Tables S1 and S2 (Supplementary Materials), in the first year of the project 178 of 180 samples were immediately classified by panels (54 EVOO, 78 VOO, and 48 LOO). Classification was not possible for only two samples (UN\_10, UP\_14), as agreement among panels was not reached on the category (V/L). The sensory evaluation of oils from the second sampling allowed classification of 153 oils (69 EVOO, 51 VOO and 33 LOO); 1 sample was not classified due to an anomalous lemon smell (ZRS\_1) and was therefore excluded from the set [20]. For these reasons, the classification model was built on 331 samples.
The oils collected were representative of possible commercial samples and borderline samples that can be the object of disagreement between panels in terms of sensory characteristics. Different aliquots of the samples, stored in the lab at 10–12 ◦C (for sensory analysis) and at −18 ◦C (for instrumental analysis), were reconditioned at room temperature before analysis.
#### *2.2. Analytical Conditions*
The FGC system (FGC-E-nose Heracles II, AlphaMos, Toulouse, France) is based on the technology of ultra-fast gas-chromatography.
The FGC is equipped with two columns working in parallel: a non-polar column (MXT5: 5% diphenyl, 95% methylpolysiloxane, 10 m length and 180 µm diameter) and a polar column (MXT-1701: 14% cyanopropylphenyl/86% dimethyl polysiloxane, 10 m length, 180 µm diameter). At the end of each column, a FID detector is placed and the acquired signal is digitalized every 0.01 s.
The analytical conditions applied were the same described by Melucci et al. 2016 [19]. The only difference was related to the temperature of the conditioning step of the samples before injection: the vial is placed in the auto-sampler (HS 100, CTC Analytics), which moves it in a shaker oven where it remains for 20 min at 40 ◦C, shaken at 500 rpm.
#### *2.3. Validation Protocol*
To confirm that the analytical procedure employed has performance capabilities consistent with the required application, a validation strategy for non-targeted approaches was performed.
A QC (quality control) sample, representative of the qualitative and quantitative VOO volatile composition (presence of volatile compounds along the entire interval of the chromatogram), was used. In this study, the QC sample was obtained by pooling the same volume of three case-control samples (1 EVOO, 1 VOO with median of 1.9 for fusty-muddy defect, and 1 VOO with a median of 2.5 for rancid defect) and seven replicates were taken into consideration.
The quality of the instrumental performance intended for fingerprinting analysis was checked by the calculation of the relative standard deviation (RSD) as proposed by the Food and Drug Administration [21]. Specifically, the repeatability (intra-day repeatability and inter-day repeatability performed according to EC 657/2002) [22] of the chromatographic signal evaluated in terms of RSD% of each chromatogram data point, with intensities above noise signal of the replicates of the same QC samples, was considered [23,24].
Prior to RSD calculation, data were aligned using the COW algorithm (correlation optimized warping) [25] and autoscaled (mean-centering followed by division of variable by the standard deviation of that column) to correct shifts in retention time and possible differences in the signal amplification of the instrument. All elaborations were made using PLS Toolbox for Matlab (MatlabR2018a®) (Natick, MA, USA).). For calculation of RSD% for each chromatogram data point, the evaluation and exclusion of noise signal is carried out to avoid considering non-relevant RSD%.
For precision, the FDA recommends a RSD not higher than 15% regarding the analytical variability for target analysis, except for concentrations close to the detection limit where a RSD of 20% is acceptable (FDA Bioanalytical Method Validation-Guidance for Industry, 2018). This, in agreement with the trend described by the Horwitz equation for targeted methods [26], demonstrates that the repeatability is strongly correlated with the intensity of the variables.
Although fingerprinting represents a different analytical approach and more variation is expected when doing untargeted analysis, these guidelines are used as a benchmark towards repeatability evaluation. Specifically, for intra-day repeatability, the acceptance criteria were as follows: more than 90% of signals with RSD < 15%; more than 95% of signals with RSD < 20% and distribution of RSD% vs. signal intensity in accordance with the Horwitz equation. For inter-day repeatability or within-lab reproducibility, the acceptance criteria were as follows: more than 85% of signals with RSD < 15%, more than 90% of signals with RSD < 20% and distribution of RSD% vs. signal intensity in accordance with Horwitz's equation.
In addition, the examination of system performance by checking the signal to noise ratio in standard solutions (instead of the evaluation of representative VOO profiles) to facilitate the assessment and comparison of method sensitivity for other laboratories was proposed. The sensitivity of the analytical system was evaluated by analyzing 2 g of each standard solution in refined olive oil (ethanol 0.05 mg·kg−<sup>1</sup> , CAS Number 64-17-5; assay <sup>≤</sup> 97.2%; density 0.789 g/mL at 25 ◦C; hexanal, 0.1 mg·kg−<sup>1</sup> CAS Number 66-25-1; assay <sup>≥</sup> 95% (GC); density 0.815 g/mL at 25 ◦C; (*E*)-2-hexenal, 0.75 mg·kg−<sup>1</sup> CAS Number 6728-26-3; assay ≥ 97.0% (GC); density 0.846 g/mL at 25 ◦C). The S/N (S = intensity of the peak of the compound; N = mean intensity of the noise measured considering the baseline of the chromatographic zone between 43 and 50 s) for the selected analytes in the chromatograms should be >3 (acceptance criteria).
#### *2.4. Classification Models*
In order to predict the assignment of samples to a specific quality grade, full chromatograms were used to develop classification models. The raw data of each chromatogram, for a total of 19,900 points, were aligned by the COW algorithm and autoscaled using PLS Toolbox for Matlab (MatlabR2018a®). Subsequently, the noise was excluded and 8401 points were consecutively selected from first to last peak observed in the chromatogram.
Subsequently, PLS-DA (partial least square discriminant analysis) models [27] were built by using the intensity values of the points as variables X (matrix X), while the commercial categories (EVOO, VOO, LOO) were considered as variable Y. In particular, classification models with 2 categories were developed in sequence: EVOO vs. no-EVOO followed by VOO vs. LOO and LOO vs. no-LOO followed by EVOO vs. VOO, as proposed by Quintanilla-Casas et al. 2020 [17].
The sample dataset was split in calibration (venetian blinds cross validation, including 75% of the samples) and external validation set (25% of the samples) by using the Kennard–Stone method [28]. The dataset was deposited for possible consultation in an on-line repository [29].
The threshold value able to identify the belonging category of each sample into one of the groups was defined by using a probabilistic approach based on Bayes's rule [30]. Finally, to assess the goodness of the method, the receiver operating characteristic (ROC) curves were evaluated.
#### **3. Results and Discussion**
#### *3.1. Performance of FGC*
Most of the procedures proposed in the literature for validation of non-targeted methods focus on post-analytical data treatment and validation of statistical models. Nevertheless, a few studies have investigated control procedures as well as performance criteria and requirements to ensure the consistence of the analytical signal (fingerprint) [24,31].
Conventional performance criteria adopted for targeted methods are not applicable as such to fingerprinting methods. Fingerprinting methods intended for sample classification are not aimed at identification and quantification of analytes, but on finding distinctive patterns that are specific for a given food category (i.e., VOO commercial category) in raw analytical signals (i.e., chromatograms). Therefore, the main constraint of the fingerprinting analytical method is to provide a repeatable and reproducible signal with sufficient sensitivity to collect the information from samples for the final purpose of the method, i.e., quality classification.
For evaluation of intra-day repeatability, the pooled QC sample was analyzed by the same operator with the same equipment and in the same instrument operative conditions within the same day. For each variable (data points), mean value, SD, and RSD% were calculated considering the seven replicates. More than 97.5% of signals presented RSD < 10%, while it achieves 99.8% in correspondence of RSD < 20% (Table 1). To analyze the variability as related to the magnitude of the variables, RSD% was plotted versus signal intensity (data not shown). As expected, data points with RSD > 10% are characterized by low values of intensity. This is in agreement with the trend described by the Horwitz equation for targeted methods [26].
In the case of the inter-day repeatability (within-lab reproducibility), seven replicates of the pooled QC sample were analyzed by the same operator with the same equipment but on different days, consequently involving different environmental conditions, and the mean value, SD, and RSD% were calculated. More than 91% and 99.4% of the signals presented RSD < 10% and RSD < 20%, respectively (Table 1). A relation between intensity and RDS% was also observed in this study, similarly to that previously observed in the intra-day repeatability test.
As the fingerprinting approach intended for sample classification is not aimed in determining the concentration of single analytes, limits of detection or quantification cannot be calculated for the analytical outcome. However, the analytical method needs to be sufficiently sensitive to allow detection of minor constituents to avoid missing any valuable information.
**Table 1.** Frequency of each relative standard deviation percentage (RSD%) class obtained for intra-day and inter-day repeatability evaluated on the quality control (QC) sample.
On this basis, the method's sensitivity needs to be set as a reference parameter to be evaluated in the validation process. A target-type strategy applied to standard solutions was proposed.
Standard solution compounds were chosen as most representative of the qualitative and quantitative volatile composition of VOOs, especially regarding the presence of volatile compounds over the entire interval of the chromatogram considered in fingerprinting analysis. Differences between the concentrations used for each compound are related to their different amounts generally present in a VOO sample. Results of the S/N are reported in Table 2.
**Table 2.** Concentration (mg·kg−<sup>1</sup> ) of each compound included in the standard solution used for method's sensitivity evaluation and related S/N. The standard mix were prepared by spiking refined olive oil with each compound and analysed by flash gas chromatography (FGC). S = intensity of the peak of the compound; N = mean intensity of the noise measured considering the baseline of the chromatographic zone between 43 and 50 s.
#### *3.2. Classification Models*
A fingerprinting approach involving chemometric elaboration of the entire profiles in volatile molecules without identification and quantification was applied.
Two different classification strategies were taken into account: (i) a classification model able to discriminate EVOO and no-EVOO samples, followed by a model to classify VOO vs. LOO samples; (ii) a classification model able to discriminate LOO and no-LOO samples, followed by a model to classify VOO vs. EVOO samples.
The results, in terms of percentage and number of correctly classified samples, are reported in Table 3 for cross and external validation, respectively. Regarding the first classification strategy, the percentages of correctly classified samples ranged from 72 to 89% and from 72 to 85%, for cross and external validation, respectively. In particular, the best results were obtained during the second step useful to discriminate VOO vs. LOO. For the second strategy, conceptually more correct in terms of sequence because it first discriminates LOO which are not edible if not refined, the percentage ranged from 78 to 92% and from 73 to 85%, for cross and external validation, respectively. In this case, the highest percentages were reached using the first PLS-DA model (LOO vs. no-LOO). Furthermore, this latter model was the best of all PLS-DA models developed.
oil, LOO = lampante olive oil.
In general, the percentages are in the same range as those obtained by other authors who proposed chemometric models to discriminate VOO quality grades according to their volatile profile analyzed by different instrumental techniques [9,17]. In general, the percentages are in the same range as those obtained by other authors who proposed chemometric models to discriminate VOO quality grades according to their volatile profile analyzed by different instrumental techniques [9,17].
TOTAL: 129/156 = 83% TOTAL: 41/51 = 78% TOTAL: 158/189 = 84% TOTAL: 47/60 = 78%
*Foods* **2020**, *9*, x FOR PEER REVIEW 8 of 12
**Table 3.** Results in terms of percentage and number of samples correctly classified in cross and external validation of the two classification strategies applied based on the partial least squaresdiscriminant analysis (PLS-DA) sequential model. EVOO = extra virgin olive oil; VOO = virgin olive
**1st CLASSIFICATION STRATEGY 2nd CLASSIFICATION STRATEGY 1st Step: EVOO vs. no-EVOO 1st Step: LOO vs. no-LOO** Cross validation External validation Cross validation External validation EVOO: 70/90 (78%) EVOO: 26/32 (81%) LOO: 50/61 (81%) LOO: 17/20 (85%) No-EVOO: 132/164 (81%) No-EVOO: 37/48 (77%) No-LOO: 172/188 (92%) No-LOO: 55/65 (85%) TOTAL: 202/254 = 80% TOTAL: 63/80 = 79% TOTAL: 222/249 = 89% TOTAL: 72/85 = 85% **2nd Step: VOO vs. LOO 2nd Step: VOO vs. EVOO** Cross validation External validation Cross validation External validation VOO: 88/99 (89%) VOO: 22/26 (85%) VOO: 84/95 (88%) VOO: 23/27 (85%)
The ROC curves (Figure 1) evaluated the sensitivity (number of samples predicted as in the class divided by number actually in the class) and the specificity (number of samples predicted as not in the class divided by actual number not in the class) of all PLS-DA models (external validation) [16]. In particular, the area under the curve (AUC) identifies the degree of discrimination (ranged 0.8148 to 0.8899) and suggests that all the models are characterized by a good degree of discrimination. The ROC curves (Figure 1) evaluated the sensitivity (number of samples predicted as in the class divided by number actually in the class) and the specificity (number of samples predicted as not in the class divided by actual number not in the class) of all PLS-DA models (external validation) [16]. In particular, the area under the curve (AUC) identifies the degree of discrimination (ranged 0.8148 to 0.8899) and suggests that all the models are characterized by a good degree of discrimination.
**Figure 1.** Receiver operating characteristic (ROC) curves of all developed PLS-DA models used to discriminate samples according to quality grade; the red circles identify the sensitivity (number of samples predicted as in the class divided by number actually in the class) and the specificity (number of samples predicted as not in the class divided by actual number not in the class) of the models. EVOO = extra virgin olive oil; VOO = virgin olive oil, LOO = lampante Olive Oil. **Figure 1.** Receiver operating characteristic (ROC) curves of all developed PLS-DA models used to discriminate samples according to quality grade; the red circles identify the sensitivity (number of samples predicted as in the class divided by number actually in the class) and the specificity (number of samples predicted as not in the class divided by actual number not in the class) of the models. EVOO = extra virgin olive oil; VOO = virgin olive oil, LOO = lampante Olive Oil.
The results of all the models (cross and external validation), in term of probability of belonging to the correct class, are shown in Figure 2. The threshold value was fixed at 0.5, corresponding to a probability of 50%: a sample classified with a probability lower than this is considered as not correctly grouped [32].
The definition of a probability level, ranging from 50% to 100%, could be a means of identifying uncertain samples that need to be checked by sensory evaluation. In other words, the samples classified with a probability lower than the selected probability level should be submitted to panel test. These procedures would reduce the amount of the samples analyzed by the panel, but at the same time, it would insure the accuracy of the classification.
same time, it would insure the accuracy of the classification.
grouped [32].
The results of all the models (cross and external validation), in term of probability of belonging to the correct class, are shown in Figure 2. The threshold value was fixed at 0.5, corresponding to a probability of 50%: a sample classified with a probability lower than this is considered as not correctly
The definition of a probability level, ranging from 50% to 100%, could be a means of identifying uncertain samples that need to be checked by sensory evaluation. In other words, the samples classified with a probability lower than the selected probability level should be submitted to panel
**Figure 2.** Class prediction probability of all samples used to develop the models, in cross and external validation (grey area). Step 1—EVOO (green star) vs. no-EVOO (blue square); step 2—VOO (yellow diamond) vs. LOO (red circle); step 1—LOO (red circle) vs. no-LOO (yellow square); step 2—VOO (yellow diamond) vs. EVOO (green star). EVOO = extra virgin olive oil; VOO = virgin olive oil, LOO = lampante olive oil. **Figure 2.** Class prediction probability of all samples used to develop the models, in cross and external validation (grey area). Step 1—EVOO (green star) vs. no-EVOO (blue square); step 2—VOO (yellow diamond) vs. LOO (red circle); step 1—LOO (red circle) vs. no-LOO (yellow square); step 2—VOO (yellow diamond) vs. EVOO (green star). EVOO = extra virgin olive oil; VOO = virgin olive oil, LOO = lampante olive oil.
**4. Conclusions** Despite the undisputed validity of the panel test, its application is time consuming and expensive. Accordingly, companies and private and public quality control labs could benefit from **Table 3.** Results in terms of percentage and number of samples correctly classified in cross and external validation of the two classification strategies applied based on the partial least squares-discriminant analysis (PLS-DA) sequential model. EVOO = extra virgin olive oil; VOO = virgin olive oil, LOO = lampante olive oil.
#### **4. Conclusions**
Despite the undisputed validity of the panel test, its application is time consuming and expensive. Accordingly, companies and private and public quality control labs could benefit from robust instrumental pre-classifications, which would reduce the number of samples that have to be assessed by panels, or at least prioritize their assessment.
For this reason, the development of rapid screening methods to support the official panel test, to analyze olive oils and differentiate their quality grades, is one of the challenges in the olive oil sector, as reported in the EU framework program Horizon 2020.
In this work, FGC combined with the multivariate statistical technique was applied to discriminate samples according to different quality grades (EVOO, VOO and LOO; examples of GC traces for EVOOs and LOOs are shown in Figure S1 of the Supplementary Materials). The analytical technique proposed herein for fingerprinting olive oils combined with chemometrics was effective in reducing data complexity and time to obtain a response; this rapid screening tool could be adopted for a quick pre-classification of the quality grades, e.g., by control laboratories in companies of the OO sector, before buying or blending EVOOs.
In order to propose a robust chemometric model, a large set of samples (*n* = 331) involving two different harvesting/production years, the most common olive cultivars, geographical origin, sensory positive attributes, and sensory defects, was analyzed. In addition, a validation protocol was adopted for evaluate the reliability of the results.
The proposed analytical fingerprinting method provided repeatable and reproducible signals with sufficient sensitivity to collect valuable information about samples.
FGC associated with the two-category sequential classification model is promising to support sensory analysis in discriminating samples of different product categories. Among the proposed classification strategy, the second (1st step: LOO vs. no-LOO; 2nd step: VOO vs. EVOO) was the best of all PLS-DA models developed with percentages of correctly classified samples ranging from 78 to 92% and from 73 to 85%, for cross and external validation, respectively.
This analytical approach is very fast, and, in fact, only around 200 s are needed to analyze a single sample. The classification model, built by using a high number of robust samples classified by sensorial analysis and representative of the commercial variability (here we used a decision tree and six panels to ensure their classification) is easily applicable in any laboratory or industry.
Future studies could be addressed to the implementation of this methodology, even in relation to an increasing interest of the food sector towards volatile compounds and more widespread use of instruments such as FGC, which are less common in quality control laboratories. An even wider sampling phase including other variables among oils, since they are natural products, could lead to a better control of classifications and would lead to implementation of this technique to a broader extent. Lastly, the use of other statistical approaches, such as nonlinear techniques, could be investigated in order to improve the results of classification.
**Supplementary Materials:** The following are available online at http://www.mdpi.com/2304-8158/9/7/862/s1, Table S1: Sensory results of samples from the first year. Table S2: Sensory results of samples from the second year. Table S3: available information on samples collected and evaluated during the first year of the Oleum project. Table S4: available information on samples collected and evaluated during the second year of the Oleum project. Figure S1: overlapping of the GC traces of extra virgin (EVOO) and lampante (LOO) samples.
**Author Contributions:** Conceptualization, S.B., A.B. and T.G.T.; Formal analysis, S.B. and C.C.; Data curation, S.B. and C.C.; Writing—original draft preparation, S.B.; Writing—review and editing, S.B., C.C., A.B., B.Q.-C.; Supervision, T.G.T. and D.L.G.-G.; Funding acquisition, T.G.T. All authors have read and agree to the published version of the manuscript.
**Funding:** This work is supported by the Horizon 2020 European Research project OLEUM "Advanced solutions for assuring the authenticity and quality of olive oil at a global scale", which received funding from the European Commission within the Horizon 2020 Programme (2014–2020), grant agreement No. 635690.
**Acknowledgments:** The information expressed in this article reflects the authors' views; the European Commission is not liable for the information contained herein. We are grateful to all producers who provided us with VOOs for this study as well as the panel members who performed sensory analysis of VOOs from each institution involved: Eurofins Analytik GmbH, Hamburg, Germany; Institute of Agriculture and Tourism, Porˇec, Croatia; Institut des Corps Gras, Pessac, France; Alma Mater Studiorum-Università di Bologna; Science and Research Centre Koper, Slovenia and Ulusal Zeytin ve Zeytinyăgı Konseyi, Izmir, Turkey.
**Conflicts of Interest:** The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.
## **References**
© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
## **Genetic Identification of the Wild Form of Olive (***Olea europaea var. sylvestris***) Using Allele-Specific Real-Time PCR**
### **Christina I. Kyriakopoulou and Despina P. Kalogianni \***
Department of Chemistry, University of Patras, 26504 Rio Patras, Greece; [email protected] **\*** Correspondence: [email protected]
Received: 28 February 2020; Accepted: 7 April 2020; Published: 9 April 2020
**Abstract:** The wild-type of olive tree, *Olea europaea var Sylvestris* or oleaster, is the ancestor of the cultivated olive tree. Wild-type olive oil is considered to be more nutritious with increased antioxidant activity compared to the common cultivated type (*Olea europaea* L. *var Europaea*). This has led to the wild-type of olive oil having a much higher financial value. Thus, wild olive oil is one of the most susceptible agricultural food products to adulteration with other olive oils of lower nutritional and economical value. As cultivated and wild-type olives have similar phenotypes, there is a need to establish analytical methods to distinguish the two plant species. In this work, a new method has been developed which is able to distinguish *Olea europaea var Sylvestris* (wild-type olive) from *Olea europaea* L. *var Europaea* (cultivated olive). The method is based, for the first time, on the genotyping, by allele-specific, real-time PCR, of a single nucleotide polymorphism (SNP) present in the two olives' chloroplastic genomes. With the proposed method, we were able to detect as little as 1% content of the wild-type olive in binary DNA mixtures of the two olive species.
**Keywords:** *Olea europaea var Sylvestris*; oleaster; olive; olive oil; real-time PCR; adulteration; SNP; DNA
### **1. Introduction**
The wild form of the olive tree, formally named *Olea europaea var Sylvestris* or oleaster, is considered to be one of the oldest trees worldwide; it is found mainly in the Mediterranean Basin. Genetic pattering studies have shown that cultivated olive trees, i.e., *Olea europaea* L. *var Europaea*, are more similar to oleaster species, providing evidence to support the concept that oleasters are the ancestors of cultivated trees [1]. Both wild and cultivated olive oil have beneficial properties for human health, giving them high economic and nutritional value; however, this has made olive oil one of the most vulnerable agricultural products to fraud and fakery. Wild-type olive oil has higher antioxidant activity, as well as phenolic, tocopherolic and orthodiphenolic contents equal to or higher those in extra virgin cultivated olive oil [2]. Moreover, wild-type olive is a valuable natural resource due to its resistance to certain environmental and climatic conditions and diseases [3]. For the above reasons, its genetic characteristics have to be evaluated, and reliable molecular tools have to be developed for olive oil origin traceability (genetically and geographically) and wild-type olive oil identification. On the other hand, producers need accurate analytical tools for the genetic identification of their wild-type olive-related products to ensure their high added value [4].
Genetic variations between the two plant species have not been extensively explored by the research community. The analytical techniques used so far for the genetic identification of the wild form of olive tree include randomly amplified polymorphic DNA (RAPD), amplified fragment length polymorphisms (AFLPs) and intersimple and simple sequence repeats (ISSRs and SSRs), based on the chloroplastic and mitochondrial plant DNA [1]. Early research compared the genome of *Olea* *europaea* L. *var Europaea* to that of the wild-type olive, derived from many countries and two areas in Italy, using AFLP analysis as designed by Angiolillo et al., 1999, and Baldoni et al., 2006 [5,6]. RAPD analysis was used to distinguish oleasters from *Olea europaea* L. *var Europaea* trees on the Mediterranean islands of Corsica and Sardinia, as well as in Turkey [7,8]. Besnard et al. used RAPD markers and restriction fragment length polymorphism (RFLP) analysis based on mitochondrial and cytoplasmatic DNA to investigate the relationships among olive species and subspecies in the Mediterranean Basin and other countries in Asia and Africa. This research led to the discovery that there was a large degree of diversity among olive cultivated trees, but that they were more or less related to the local oleasters [9,10]. Moreover, ISSR and SSR markers have been utilized by many researchers to investigate the relation and differentiation of cultivated olives from wild-type olives [3,11–16]. Genome size estimation based on double-stranded DNA staining followed by flow cytometric analysis was also used for screening purposes between *Olea europaea var Sylvestris* and *Olea europaea* L. *var Europaea* species [17], while flow cytometry in combination with SSR profiles was used for the taxonomy of four olive subspecies, namely *Olea europaea ssp. cerasiformis*, *Olea europaea ssp. guanchica*, *Olea europaea var Sylvestris* and *Olea europaea* L. *var Europaea* [18].
Moreover, the wild olive has also been used for nonedible purposes in pharmacology and cosmetics to create products with specific valuable characteristics. Researches have also studied the antimicrobial activity of the wild olive against certain human bacterial pathogens [19]. Several plants, including the olive and its wild form, have also been used for the production of various food supplements [20]. Finally, phenolic extracts from wild olive leaves have been investigated for use in foodstuffs, food additives and functional food materials, due to their high antioxidant activity [21,22].
In 2017, the complete genome sequence of *Olea europaea var Sylvestris* was published by Unver et al. [23]. This will be useful, in the future, for the localization of specific genetic variations in the genome of oleasters compared to other olive subspecies.
For the first time, in this work, a single nucleotide polymorphism (SNP)-based method was developed for the detection and identification of the wild form of olive in order to distinguish it from the cultivated olive. Different olive cultivars contain different SNPs in their genome that are responsible for their unique phenotyping characteristics [24,25]. The method was based on an allele-specific, real-time PCR. The proposed method is able to detect wild-type olive DNA at levels as low as 1% in DNA derived from the cultivated olive.
#### **2. Materials and Methods**
#### *2.1. Materials and Instrumentation*
The Vent (exo-) DNA polymerase was purchased by New England Biolabs (Beverly, MA, USA). Deoxynucleoside triphosphates (dNTPs) were obtained from Kapa Biosystems (Wilmington, MA, USA). The fluorescent dye SYBR Green I 10<sup>4</sup> <sup>×</sup> concentrated was from Molecular Probes (Eugene, OR, USA). The primers used were from Eurofins Scientific (Brussels, Belgium) and are listed in Table 1. The size of the PCR products was 136 bp. An extra virgin olive oil sample (*Olea europaea* L. *var Europaea*) was purchased from a local market, while a certified wild-type olive oil sample (*Olea europaea var Sylvestris*) was kindly by local producer, Alexandros Karakikes, from the Olea Sylvestris estate (Agrielaio, Volos, Greece) [26].
Real-time PCR was performed using the Mini Opticon Real-Time PCR System from Biorad (Hercules, CA, USA), while the results were analyzed using the Bio–Rad CFX Manager 3.0 software.
**Table 1.** The primers used in the allele-specific, real-time PCR, two species-specific upstream primers and a common downstream primer, along with their melting temperatures (Tm).
\* according to Eurofins Scientific (Brussels, Belgium).
#### *2.2. DNA Isolation Procedure*
DNA was isolated from olive oil samples using the NucleoSpin Tissue kit from Macherey-Nagel (Düren, Germany) according to the manufacturer's instructions. The quantity and purity of the isolated DNA were determined using the Nanodrop UV/VIS Nanophotometer by Implen GmbH (Münich, Germany).
#### *2.3. Design of the Primers*
The primers used for the amplification of *Olea europaea var Sylvestris* (wild-type olive) and *var Europaea* (cultivated olive) were designed using the free online Oligo Analyzer software for primer evaluation (created by Dr. Teemu Kuulasmaa), based on the *Olea europaea var. sylvestris* NADH dehydrogenase subunit F gene, chloroplastic sequence (Accession Number: AY172114) and the *Olea europaea* L. NADH dehydrogenase subunit F (ndhF) gene chloroplastic sequence (Accession Number: DQ673278) [23].
#### *2.4. Allele-Specific, Real-Time PCR*
The allele-specific, real-time PCR reactions were conducted in a final volume of 50 µL and contained 1 <sup>×</sup> Thermopol Buffer (20 mM Tris-HCl, 10 mM (NH4)2SO4, 10 mM KCl, 0.1% Triton® X-100 at pH 8.8), 0.5 µM of each of the upstream and downstream primers, 0.2 mM of each of the four dNTPs, 0.5 mM MgCl2, 2 × SYBR Green I, one unit of Vent (exo-) DNA polymerase and 150 ng of isolated DNA. The reaction conditions involved a 95 ◦C incubation step for three min, followed by 45 cycles at 95 ◦C for 30 s, 62 ◦C for 30 s, 72 ◦C for 30 s and a final extension step at 72 ◦C for 10 min.
#### **3. Results and Discussion**
A new analytical method was developed for the detection and identification of *Olea europaea var Sylvestris* that refers to the wild form of the olive tree. The method was based on the detection of a specific Single Nucleotide Polymorphism (SNP) that is different in the genome of the wild olive plant. The method involves the following steps: (i) DNA isolation from olive oil samples and (ii) allele-specific, real-time PCR using an upstream primer specific to *Olea europaea var Sylvestris* or *var Europaea* species and a common downstream primer. The species-specific primers have the same 22-base sequence but differ only at the base at the 30 end that contains the SNP of interest. The DNA sequences were amplified using a DNA polymerase that lacked the 30 to 50 exonuclease activity, so only the primer that was perfectly complementary to the DNA target was extended by the enzyme. The amplicons were finally detected using the DNA intercalating fluorescent dye SYBR Green I. The principle of the proposed method is illustrated in Figure 1. SYBR Green I was chosen here instead of Taqman probes in order to develop a new analytical method that could be easily transferred, with few modifications, for the detection of other SNPs that will be found in the wild olive genome in the future.
*Foods* **2020**, *9*, x FOR PEER REVIEW 4 of 9
**Figure 1. (Upper panel)** Schematic illustration of the principle of the method that includes DNA isolation and purification from olive oil samples using spin cleanup columns, including the following steps: cell lysis of an olive oil sample, capture of DNA to the cleanup columns and elution of the DNA from the columns. **(Lower panel)** The allele-specific, real-time PCR. Two allele-specific upstream primers that contain the SNP of interest at their 3΄ ends and one common downstream primer were used in the amplification reaction. Only the perfectly complementary upstream primer to the target was extended by the DNA polymerase, while the amplicons were detected by the DNA intercalating dye, SYBR Green I. *3.1. DNA Isolation* **Figure 1.** (**Upper panel**) Schematic illustration of the principle of the method that includes DNA isolation and purification from olive oil samples using spin cleanup columns, including the following steps: cell lysis of an olive oil sample, capture of DNA to the cleanup columns and elution of the DNA from the columns. (**Lower panel**) The allele-specific, real-time PCR. Two allele-specific upstream primers that contain the SNP of interest at their 30 ends and one common downstream primer were used in the amplification reaction. Only the perfectly complementary upstream primer to the target was extended by the DNA polymerase, while the amplicons were detected by the DNA intercalating dye, SYBR Green I.
#### First, DNA was isolated from olive oil samples and its concentration was determined using a *3.1. DNA Isolation*
UV/VIS nanophotometer. It was found that the isolation procedure did not result in a constant DNA amount for all samples, with the DNA concentrations ranging from 8.4 to 142 ng/μL. To avoid fluctuation in the PCR yield due to different initial DNA concentrations, we decided to use the same amount (ng) of isolated DNA for all samples into the real-time PCR mixture. After amplification, the amplicons had a size of 136 bp. The quality of the isolated DNA was also determined by UV measurements; the ratios A260/A280 were from 1,174 to 1,739. DNA was considered to be of high quality when the ratio A260/A280 was above 1.8. *3.2. Optimization of the PCR Conditions* The real-time PCR conditions were initially optimized. The parameters studied were the amount First, DNA was isolated from olive oil samples and its concentration was determined using a UV/VIS nanophotometer. It was found that the isolation procedure did not result in a constant DNA amount for all samples, with the DNA concentrations ranging from 8.4 to 142 ng/µL. To avoid fluctuation in the PCR yield due to different initial DNA concentrations, we decided to use the same amount (ng) of isolated DNA for all samples into the real-time PCR mixture. After amplification, the amplicons had a size of 136 bp. The quality of the isolated DNA was also determined by UV measurements; the ratios A260/A<sup>280</sup> were from 1174 to 1739. DNA was considered to be of high quality when the ratio A260/A<sup>280</sup> was above 1.8.
#### of the isolated DNA, the concentration of the primers, the number of PCR cycles and the temperature of the annealing step of the reaction. At low DNA and primer concentrations, low temperature (55- *3.2. Optimization of the PCR Conditions*
60 °C) and number of cycles < 45, the PCR was not sufficiently efficient. The yield of the reaction also decreased when a high amount of initial DNA target was used. This may be attributed to the fact that the DNA isolated from olive samples has reduced quality, as it contains high amounts of PCR inhibitors that may inhibit the activity of the DNA polymerase [27]. We also observed that the highest reaction yield and specificity were obtained at an annealing temperature of 62 °C. The real-time PCR conditions were initially optimized. The parameters studied were the amount of the isolated DNA, the concentration of the primers, the number of PCR cycles and the temperature of the annealing step of the reaction. At low DNA and primer concentrations, low temperature (55–60 ◦C) and number of cycles < 45, the PCR was not sufficiently efficient. The yield of the reaction also decreased when a high amount of initial DNA target was used. This may be attributed to the fact that the DNA isolated from olive samples has reduced quality, as it contains high amounts of PCR inhibitors that may inhibit the activity of the DNA polymerase [27]. We also observed that the highest reaction yield and specificity were obtained at an annealing temperature of 62 ◦C.
#### *3.3. Specificity of the Allele-Specific Primers Foods* **2020**, *9*, x FOR PEER REVIEW 5 of 9
The specificity of the two species-dependent upstream primers was then studied as follows: both DNA targets, *Olea europaea var Sylvestris* (wild-type olive) and *var Europaea* (cultivated olive) were subjected to two separate amplification reactions using either the upstream primer specific to the wild-type olive or the cultivated olive-specific upstream primer. As shown, in Figure 2, each primer amplified only its fully complementary DNA sequence, proving the superior specificity of the primers. To ensure that the fluorescence signals were attributed only to the specific amplicons, a melting curve analysis was also performed after each amplification reaction. The melting curve analysis revealed only one peak for each PCR product, the melting temperature (Tm) of which was 77 ◦C for *Olea europaea var Sylvestris* (wild-type olive) and 78 ◦C for *var Europaea* (cultivated olive), allowing us to distinguish between the two allele-specific DNA sequences. *3.3. Specificity of the Allele-Specific Primers* The specificity of the two species-dependent upstream primers was then studied as follows: both DNA targets, *Olea europaea var Sylvestris* (wild-type olive) and *var Europaea* (cultivated olive) were subjected to two separate amplification reactions using either the upstream primer specific to the wild-type olive or the cultivated olive-specific upstream primer. As shown, in Figure 2, each primer amplified only its fully complementary DNA sequence, proving the superior specificity of the primers. To ensure that the fluorescence signals were attributed only to the specific amplicons, a melting curve analysis was also performed after each amplification reaction. The melting curve analysis revealed only one peak for each PCR product, the melting temperature (Tm) of which was 77 °C for *Olea europaea var Sylvestris* (wild-type olive) and 78 °C for *var Europaea* (cultivated olive),
allowing us to distinguish between the two allele-specific DNA sequences.
**Figure 2.** The real-time PCR curves, along with the corresponding melting curve analysis, obtained during the specificity study of the two-allele specific upstream primers with both DNA targets: *Olea europaea var Sylvestris* (wild-type of olive) **(a)** and *Olea europaea* L*. var Europaea* (cultivated olive) **(b)**. Each specific primer strictly amplifies the fully complementary DNA sequence. Tm: melting temperature, RFU: Relative Fluorescence Units. **Figure 2.** The real-time PCR curves, along with the corresponding melting curve analysis, obtained during the specificity study of the two-allele specific upstream primers with both DNA targets: *Olea europaea var Sylvestris* (wild-type of olive) (**a**) and *Olea europaea* L. *var Europaea* (cultivated olive) (**b**). Each specific primer strictly amplifies the fully complementary DNA sequence. Tm: melting temperature, RFU: Relative Fluorescence Units.
#### *3.4. Detectability of the Method in Binary DNA Mixtures*
Subsequently, the detectability of the method in olive DNA binary mixtures was evaluated. DNA mixtures that contained different proportions (1–50%) of DNA from *Olea europaea var Sylvestris* in DNA from *var Europaea* were prepared. An amount of 150 ng of each DNA mixture was then subjected to two separate allele-specific, real-time PCR reactions using each of the species-specific upstream primers along with the common downstream primer, respectively. A high amount of total DNA was used in the PCR in order to detect the low amount of wild olive DNA in the mixtures, e.g., for 150 ng of total DNA in the 1% mixture, only the 1.5 ng was the wild olive DNA. The results are presented in Figure 3. We were able to detect as little as 1% of DNA specific to *Olea europaea var Sylvestris* in the presence of DNA from *Olea europaea* L. *var Europaea*. The allelic ratios of the analyzed SNP for the above DNA mixtures were also calculated based on the fluorescence value at the 45th cycle of the reaction, and are
presented in the same Figure. The allelic ratios for all DNA mixtures were close to the value of 0.5, as expected for a heterozygote sample. the analyzed SNP for the above DNA mixtures were also calculated based on the fluorescence value at the 45th cycle of the reaction, and are presented in the same Figure. The allelic ratios for all DNA mixtures were close to the value of 0.5, as expected for a heterozygote sample.
*europaea var Sylvestris* in the presence of DNA from *Olea europaea* L. *var Europaea*. The allelic ratios of
*Foods* **2020**, *9*, x FOR PEER REVIEW 6 of 9
Subsequently, the detectability of the method in olive DNA binary mixtures was evaluated. DNA mixtures that contained different proportions (1%–50%) of DNA from *Olea europaea var Sylvestris* in DNA from *var Europaea* were prepared. An amount of 150 ng of each DNA mixture was then subjected to two separate allele-specific, real-time PCR reactions using each of the speciesspecific upstream primers along with the common downstream primer, respectively. A high amount of total DNA was used in the PCR in order to detect the low amount of wild olive DNA in the mixtures, e.g., for 150 ng of total DNA in the 1% mixture, only the 1.5 ng was the wild olive DNA.
*3.4 Detectability of the Method in Binary DNA Mixtures*
**Figure 3. (a)** The real-time PCR curves obtained for different % DNA content (0%–50%) of *Olea europaea var Sylvestris* (wild-type olive) DNA in binary mixtures with *Olea europaea* L*. var Europaea* (cultivated olive) DNA. **(b)** The allelic ratios of the binary DNA mixtures calculated as the ratio of the fluorescence intensity obtained with the upstream primer specific to *Olea europaea var Sylvestris* target versus the sum of the fluorescence intensity obtained by both allele-specific primers for *Olea europaea var Sylvestris* and *var Europaea* targets. All allelic ratios were close to the value of 0.5, which corresponds to a heterozygote sample. RFU: Relative Fluorescence Units. **Figure 3.** (**a**) The real-time PCR curves obtained for different % DNA content (0–50%) of *Olea europaea var Sylvestris* (wild-type olive) DNA in binary mixtures with *Olea europaea* L. *var Europaea* (cultivated olive) DNA. (**b**) The allelic ratios of the binary DNA mixtures calculated as the ratio of the fluorescence intensity obtained with the upstream primer specific to *Olea europaea var Sylvestris* target versus the sum of the fluorescence intensity obtained by both allele-specific primers for *Olea europaea var Sylvestris* and *var Europaea* targets. All allelic ratios were close to the value of 0.5, which corresponds to a heterozygote sample. RFU: Relative Fluorescence Units.
#### *3.5 Reproducibility of the Method 3.5. Reproducibility of the Method*
Finally, the reproducibility of the method was determined. Two different proportions, 1% and 10%, of the above DNA mixtures, were subjected, in triplicate, to real-time PCR. The % coefficients of variation (CV) were calculated based on the obtained Cq values for all samples. The CV for the 1%-content was 10.5% and for the 10%-content was 7.5%, demonstrating the reproducibility of the method. Finally, the reproducibility of the method was determined. Two different proportions, 1% and 10%, of the above DNA mixtures, were subjected, in triplicate, to real-time PCR. The % coefficients of variation (CV) were calculated based on the obtained Cq values for all samples. The CV for the 1%-content was 10.5% and for the 10%-content was 7.5%, demonstrating the reproducibility of the method.
#### **4. Conclusions 4. Conclusions**
A new allele-specific, real-time PCR-based analytical method was developed for the detection and identification of wild-type olive oil (*Olea europaea var Sylvestris*), compared to cultivated olive oil (*Olea europaea L. var Europaea*). The discrimination of the two similar plant species was based on genotyping a single nucleotide polymorphism (SNP) that is differently present in the genome of the two plant species. The detection of this SNP was carried out by an allele-specific, real-time PCR that was performed using two different species-specific upstream primers that contained the analyzed SNP and a common downstream primer. Each specific primer amplified only its fully complementary DNA sequence, leading to species identification. The detection of the amplicons was accomplished A new allele-specific, real-time PCR-based analytical method was developed for the detection and identification of wild-type olive oil (*Olea europaea var Sylvestris*), compared to cultivated olive oil (*Olea europaea* L. *var Europaea*). The discrimination of the two similar plant species was based on genotyping a single nucleotide polymorphism (SNP) that is differently present in the genome of the two plant species. The detection of this SNP was carried out by an allele-specific, real-time PCR that was performed using two different species-specific upstream primers that contained the analyzed SNP and a common downstream primer. Each specific primer amplified only its fully complementary DNA sequence, leading to species identification. The detection of the amplicons was accomplished using the DNA intercalating dye, SYBR Green I. With the proposed method, we were able to sucessfully distinguish between the two plant species in olive oil samples. Also, as little as 1% wild-type olive species was detected in binary DNA mixtures of the two analyzed plant species. In conclusion, the method is easy, rapid, has good detectability, is reproducible and can easily distinguish between species. The proposed method also contributes to the ability to add the higher financial value to wild-type olive-based products. In the future, the determination of different SNPs in the wild-type olive genome compared to all the known cultivated olive trees could lead to more accurate discrimination of the wild-type olive among other olive-based subspecies. The proposed method could also be applied, with some modifications, for the detection of wild olive-based ingredients in food supplements and cosmetic products. The global increase in food supplements has led to the mislabeling of these products and fraudulent practices. In both cases, the purity of the extracted DNA is more important than the
PCR yield itself, because several food additives and other ingredients may be present in the extracts, inhibiting the PCR amplification. Also, the amount of the extracted DNA may be extremely low. Thus, the DNA isolation protocols have to be properly justified to remove these inhibitors and increase the DNA recovery and the PCR yield. In some studies, however, the inability to extract DNA from some food supplements has been reported. Finally, in some products, DNA degradation may also occur due to thermal or chemical treatment, but the use of short-length amplicons can overcome this issue [28–31].
**Author Contributions:** Conceptualization, D.P.K.; methodology, C.I.K. and D.P.K.; software, C.I.K.; investigation, C.I.K. and D.P.K.; resources, C.I.K. and D.P.K.; data curation, C.I.K. and D.P.K.; writing—original draft preparation, C.I.K. and D.P.K.; writing—review and editing, D.P.K.; supervision, D.P.K.; project administration, D.P.K.; funding acquisition, D.P.K. All authors have read and agreed to the published version of the manuscript.
**Funding:** This work was funded by the project "Research Infrastructure on Food Bioprocessing Development and Innovation Exploitation—Food Innovation RI" (MIS 5027222), which is implemented under the Action "Reinforcement of the Research and Innovation Infrastructure", funded by the Operational Programme "Competitiveness, Entrepreneurship and Innovation" (NSRF 2014–2020) and co-financed by Greece and the European Union (European Regional Development Fund).
**Acknowledgments:** We acknowledge Alexandros Karakikes (Olea Sylvestris estate) for kindly providing us with the wild-type olive oil sample.
**Conflicts of Interest:** The authors declare no conflicts of interest.
#### **References**
© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
## *Article*
## **Using Chemometric Analyses for Tracing the Regional Origin of Multifloral Honeys of Montenegro**
**Vesna Vukašinovi´c-Peši´c 1,\*, Nada Blagojevi´c <sup>1</sup> , Snežana Brašanac-Vukanovi´c <sup>1</sup> , Ana Savi´c <sup>2</sup> and Vladimir Peši´c <sup>3</sup>**
Received: 19 January 2020; Accepted: 14 February 2020; Published: 18 February 2020
**Abstract:** This is the first study of mineral content and basic physicochemical parameters of honeys of Montenegro. We examined honey samples from eight different micro-regions of Montenegro, and the results confirm that, with the exception of cadmium in samples from two regions exposed to industrial pollution, none of the 12 elements analyzed exceeded the maximum allowable level. The samples from areas exposed to industrial pollution were clearly distinguished from samples from other regions of Montenegro in the detectable contents of Pb, Cd, and Sr. This study showed that chemometric techniques might enhance the classification of Montenegrin honeys according to their micro-regional origin using the mineral content. Linear discriminant analysis revealed that the classification rate was 79.2% using the cross-validation method.
**Keywords:** honey; regional origin; chemometric analysis; mineral content; Montenegro
### **1. Introduction**
Honey is a complex natural product, whose characteristics depend on the flower nectar from which it is obtained, but also on other factors such as geographical origin, bee species, season, type of processing and storage [1]. It is known that pollution and a number of different pollutants present in its foraging areas have an impact on honeybees [2] but also on nectar-providing plant species. Therefore, it is necessary to assure geographical traceability and determine the botanical origin of the foraging area of the beehive.
As stated by Karabagias and Karabournioti [3], the authentication of honey is gaining in importance and includes a number of contending parties from producers and sellers to consumers and control labs. A number of papers have shown that specific physicochemical parameters and mineral contents in combination with chemometric analyses can be a useful tool in discovering botanical and/or geographical origin of honeys that may enter the market [1,3,4].
Tracing the geographical origin of honeys can provide important information about the potential contamination of the area from which the honey production material comes. Therefore, ensuring high standards in terms of product safety leads to the need to examine the contents of essential and toxic elements in honey. Due to its bioaccumulation ability, honey can be used as an indicator of metal pollution, especially of toxic pollutants such as Pb, Cd, and As [5–7].
Due to its geographical position, climate conditions and richness of the nectar-providing plants, Montenegro provides favorable natural conditions for more intensive development of beekeeping. According to the data for 2011, the population in Montenegro was 625,266, while the honey production
for that year was 394 t, and the average annual consumption 1.2 kg per person, meaning that a large part of honey consumption in Montenegro is imported [8]. Data for the last few years show an increase in honey production (627 t for 2016) but also an increase in the average annual consumption of honey per person (2.76 kg) [9].
The majority of honey on the market in Montenegro are multifloral (derived from a large number of nectar-providing plant species in the honeybees foraging area). Most of these honey types are recognizable by their local or regional origin (e.g., Katunski med (= honey), Pivski med, Piperski med.). It is worth mentioning that Montenegro and its regions are known to harbor a high number of regional floral endemics [10] that likely affect the composition and properties of honey.
There is a lack of information on the mineral content and basic physicochemical parameters of honey from the territory of Montenegro. Moreover, there is no continuity in monitoring the quality of honey, especially in areas that are exposed to the effects of potential pollution sources. Due to the high consumption of local honey in the diet, the need and obligation for its systematic characterization are highly required.
This study is aimed to investigate the mineral content and the basic physicochemical parameters of honeys from different micro-regions of Montenegro. We evaluated the usefulness of chemometric analyses for the classification of honeys according to its regional origin.
#### **2. Materials and Methods**
Twenty-four honey samples as indicated in Figure 1 were collected from eight micro-regions of Montenegro, i.e., (1) Piva, (2) Zbljevo, (3) Potrlica, (4) Mijakovi´ci, (5) Piperi, (6) Martini´ci, (7) Katunska, and (8) Zeta. The Piva, Zbljevo, Potrlica and Mijakovi´ci micro-regions are situated in the continental part of Montenegro (Alpine biogeographical region, see Figure 1) while the four other micro-regions are situated in the sub-Mediterranean part of the country belonging to the Mediterranean biogeographical region [10]. The climate in the latter region is mainly Mediterranean-Adriatic with relatively dry and warm summers (the average air temperature of the warmest month > 20 ◦C), but humid and mild winters (the average air temperatures varies from 6 to 9 ◦C), while the Alpine region has a "continental" type climate, with relatively cool and humid summers and long and harsh winters [10].
Samples were taken from individual beekeepers during the harvesting season 2015. All samples were multifloral as confirmed by the suppliers. The samples were stored in glass flasks at room temperature before analysis. Physicochemical parameters (pH, electrical conductivity (EC), free acidity (FA) and moisture) were analyzed using the Harmonized Methods of the International Honey Commission [11].
The mineral composition of honey was analyzed by inductively coupled plasma-optical emission spectrometry (ICP-OES). About 1 g of each honey sample was digested with 14 mL 65% HNO<sup>3</sup> and 2 mL 35% H2O<sup>2</sup> on a hot plate to near dryness. The sample containing a volumetric flask was cooled at room temperature before the addition of deionized water to the mark on the flask. All samples were prepared in triplicate and their average value was assessed.
The concentration of twelve elements (Pb, Cd, Cu, Zn, Fe, Cr, Sr, Ba, Ca, Na, K, Mg) were determined by ICP-OES according to the iCAP 6000 spectrometer method.
All statistical analyses were performed using SPSS 17.0 (SPSS Statistics for Windows, Version 17.0. SPSS Inc., Chicago, IL, USA). Data were expressed as mean ± standard deviation. A Kolmogorov–Smirnov test showed that all analyzed physicochemical parameters were normally distributed, while the content of Pb, Cd, Sr and Ba in some regions exhibited significant differences from the normal distribution. The one-way analysis of variance (ANOVA) was performed on physicochemical parameters in order to determine if there any significant differences between studied micro-regions at the confidence level 0.05. The Kruskal–Wallis test was used to investigate whether the mineral contents varied significantly between the investigated micro-regions. The relationship between the mineral content and physicochemical parameters were analyzed using the Spearman's correlation analysis. For checking similarities between samples of honey of different geographical
origin we used two chemometric analyses: PCA and LDA. Statistical methods based on principal component analysis (PCA) and linear discriminant analysis (LDA) have been used. The LDA was performed using R.3.5.3, while the PCA was made by using MVSP version 3.21. *Foods* **2020**, *9*, x FOR PEER REVIEW 3 of 9
**Figure 1.** Map of Montenegro with marked locations of honey sampling from eight micro‐regions (in parentheses are given sampling location numbers): Piva (1–3), Zbljevo (4–6), Potrlica (7–9), Mijakovići (10–12), Piperi (13–15), Martinići (16–18), Katunska (19–21), and Zeta (22–24). **Figure 1.** Map of Montenegro with marked locations of honey sampling from eight micro-regions (in parentheses are given sampling location numbers): Piva (1–3), Zbljevo (4–6), Potrlica (7–9), Mijakovi´ci (10–12), Piperi (13–15), Martini´ci (16–18), Katunska (19–21), and Zeta (22–24).
#### All statistical analyses were performed using SPSS 17.0 (SPSS Statistics for Windows, Version **3. Results**
**3. Results**
Zbljevo.
17.0. SPSS Inc., Chicago, IL, USA). Data were expressed as mean ± standard deviation. A Kolmogorov–Smirnov test showed that all analyzed physicochemical parameters were normally distributed, while the content of Pb, Cd, Sr and Ba in some regions exhibited significant differences from the normal distribution. The one‐way analysis of variance (ANOVA) was performed on physicochemical parameters in order to determine if there any significant differences between studied micro‐regions at the confidence level 0.05. The Kruskal–Wallis test was used to investigate whether the mineral contents varied significantly between the investigated micro‐regions. The relationship between the mineral content and physicochemical parameters were analyzed using the The mineral content of honey samples from different geographical areas of Montenegro is presented in Table 1. The value presented for each element is the average concentration observed. A significant difference has been observed in the concentrations of Pb, Cd and Sr (*p* = 0.002) between studied micro-regions. In most analyzed samples the concentrations of above-listed elements were below the limit of detection except in the samples from Potrlica, Zbljevo and Mijakovi´ci. The highest Cd concentration was observed in samples from Potrlica (0.08 ± 0.01 mg/kg). The highest concentration of Pb (0.21 ± 0.06 mg/kg) and Sr (0.12 ± 0.00 mg/kg) were recorded in samples from Zbljevo.
Spearman's correlation analysis. For checking similarities between samples of honey of different geographical origin we used two chemometric analyses: PCA and LDA. Statistical methods based on principal component analysis (PCA) and linear discriminant analysis (LDA) have been used. The
LDA was performed using R.3.5.3, while the PCA was made by using MVSP version 3.21.
The mineral content of honey samples from different geographical areas of Montenegro is presented in Table 1. The value presented for each element is the average concentration observed. A significant difference has been observed in the concentrations of Pb, Cd and Sr (*p* = 0.002) between studied micro‐regions. In most analyzed samples the concentrations of above‐listed elements were
*Foods* **2020**, *9*, 210
**Table 1.** Mineral content and physicochemical parameters of honey samples from studied micro-regions of Montenegro.
72
The concentrations of examined physicochemical parameters of honey samples are given in Table 1. The moisture level had similar values across studied regions and ranged from 14.92 ± 0.78% (Piperi) to 16.22 ± 0.36% (Zeta). The pH of studied honey samples varies between 3.87 and 4.49 and was lowest in samples from the Piva region (3.87 ± 0.36) and highest in honey samples from Mijakovi´ci (4.49 ± 0.14). A significant difference has been observed in pH according to the honey regional origin (*p* = 0.048). The electrical conductivity varied from 0.39 to 0.93 mS/cm and was lowest in samples from Katunska (0.39 ± 0.08 mS/cm) and highest in honey samples from Zeta (0.93 ± 0.15 mS/cm). A significant difference has been observed in electrical conductivity according to the honey geographical origin (*p* = 0.013). Free acidity varied from 25.00 to 41.67 meq/kg and was lowest in samples from Katunska (25.00 ± 7.21 meq/kg) and highest in honey samples from the Piva region (41.67 ± 12.10 meq/kg).
Correlation analysis revealed significant correlation between contents of K (R = 0.800, significance < 0.001) and Mg (R = 0.758, significance < 0.001) from one side and pH from the other one (Table 2).
**Table 2.** Results of the correlation analysis between the mineral content and physicochemical parameters of analyzed honey samples from Montenegro.
\*\* significance < 0.01.
The first principal component explains 42.53% of the total variability and is mostly determined by Cd (R = 0.416), Pb (R = 0.398), and Cu (R = 0.352). The PC2 explains 17.97% and is mostly determined by Mg (R = −0.577), K (R = −0.54), and Na (R = −0.425). Mutual projections of factor scores and their loadings for the first two PCs are presented in Figure 2. As can be seen from the projection plot the separation of the analyzed honey samples is much clearer along the *X*-axis. On the one side, there are localities Piva, Piperi, Katunska, Zeta and Martini´ci in whose honey samples Cd, Pb, and Sr were not detected. On the other side, there are Mijakovi´ci, Potrlica and especially Zbljevo, whose honey samples concentrations of Cd, Pb and Sr were detected.
73
*Foods* **2020**, *9*, x FOR PEER REVIEW 6 of 9
**Figure 2.** Principal component analysis (PCA) of the mineral content scores of analyzed Montenegrin honey samples. **Figure 2.** Principal component analysis (PCA) of the mineral content scores of analyzed Montenegrin honey samples.
LDA performed on the geographical origin revealed that the cross‐validation classification was correct for 79.17% of samples (Table 3). The smallest percent of good classification was achieved in the case of honey samples from Katunska, while the highest in the case of honeys from Piperi, Zeta, Mijakovići, and Zbljevo. LDA performed on the geographical origin revealed that the cross-validation classification was correct for 79.17% of samples (Table 3). The smallest percent of good classification was achieved in the case of honey samples from Katunska, while the highest in the case of honeys from Piperi, Zeta, Mijakovi´ci, and Zbljevo.
**Table 3.** Classification of honey according to their regional origin using the linear discriminate **Table 3.** Classification of honey according to their regional origin using the linear discriminate analysis.
Zbljevo 0.00% 0.00% 33.33% 0.00% 0.00% 0.00% 33.33% 100.00% The bidimensional plot (Figure 3) of the first two functions show four distinct clusters, three of them corresponding to Mijakovići, Potrlica, and Zbljevo regions, while all other regions were clustered together. The first discriminant function explains 94.2% of the total variance and it is The bidimensional plot (Figure 3) of the first two functions show four distinct clusters, three of them corresponding to Mijakovi´ci, Potrlica, and Zbljevo regions, while all other regions were clustered together. The first discriminant function explains 94.2% of the total variance and it is dominated by Cd content (R = 0.95). The second discriminant function explains 4.4% of the total variance and is dominated by Pb (R = −0.55) and Sr (R = −0.59) contents.
dominated by Cd content (R = 0.95). The second discriminant function explains 4.4% of the total
variance and is dominated by Pb (R = −0.55) and Sr (R = −0.59) contents.
Potrlica 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 66.67% 0.00%
*Foods* **2020**, *9*, x FOR PEER REVIEW 7 of 9
**Figure 3.** Linear discriminant score plot of analyzed honey samples. **Figure 3.** Linear discriminant score plot of analyzed honey samples.
#### **4. Discussion 4. Discussion**
The values of the mineral contents have been compared with those established by the EU regulations [12]. With the exception of the concentrations of cadmium in samples from Zbljevo and Potrlica, none of the 12 elements analyzed exceeded the maximum allowable level established by the EU regulations. Our results revealed significant differences in the concentrations of Pb, Cd, and Sr between the studied geographical areas of Montenegro. The latter elements were detected only in the samples from Mijakovići, Potrlica and Zbljevo regions, which are likely under the influence of the Pljevlja Thermal Power Plant (Zbljevo, and in less extent Mijakovići) and the Pljevlja coalmine (Potrlica). The values of the mineral contents have been compared with those established by the EU regulations [12]. With the exception of the concentrations of cadmium in samples from Zbljevo and Potrlica, none of the 12 elements analyzed exceeded the maximum allowable level established by the EU regulations. Our results revealed significant differences in the concentrations of Pb, Cd, and Sr between the studied geographical areas of Montenegro. The latter elements were detected only in the samples from Mijakovi´ci, Potrlica and Zbljevo regions, which are likely under the influence of the Pljevlja Thermal Power Plant (Zbljevo, and in less extent Mijakovi´ci) and the Pljevlja coalmine (Potrlica).
The Pb content in the analyzed honey samples varied from 80–210 μg/kg. These values were lower in comparison with the honey from Serbia (290 μg/kg [13]) and Italy (289 μg/kg [14]) but higher in comparison with those from Croatia (5.43–11.3 μg/kg [15]) and Bosnia and Hercegovina (13.4 μg/kg [16]). All these values are below the maximum allowable level established by the EU regulations (0.5 mg/kg) [12]. The Pb content in the analyzed honey samples varied from 80–210 µg/kg. These values were lower in comparison with the honey from Serbia (290 µg/kg [13]) and Italy (289 µg/kg [14]) but higher in comparison with those from Croatia (5.43–11.3 µg/kg [15]) and Bosnia and Hercegovina (13.4 µg/kg [16]). All these values are below the maximum allowable level established by the EU regulations (0.5 mg/kg) [12].
The values reported for Cd in this study (20–80 μg/kg) were higher in comparison with the honey from Croatia (0.69–12.8 μg/kg [15]), Bosnia and Hercegovina (0.013–22.9 μg/kg [16]), Romania (0.5–11.60 μg/kg [17]), Italy (8–18 μg/kg [14]), Spain (0.7–50 μg/kg [17]) and Serbia (0.59–30 μg/kg [1,13]). The values of Cd content in the samples from Potrlica and Zbljevo exceeds the maximum allowable level established by the EU legislation (0.05 mg/kg) [12]. As the main sources of Cd are recognized as the presence in sewage sludge and smelting from the nearby Pljevlja Thermal Power Plant (Zbljevo), or mining from the Pljevlja coalmine (Potrlica). The values reported for Cd in this study (20–80 µg/kg) were higher in comparison with the honey from Croatia (0.69–12.8 µg/kg [15]), Bosnia and Hercegovina (0.013–22.9 µg/kg [16]), Romania (0.5–11.60 µg/kg [17]), Italy (8–18 µg/kg [14]), Spain (0.7–50 µg/kg [17]) and Serbia (0.59–30 µg/kg [1,13]). The values of Cd content in the samples from Potrlica and Zbljevo exceeds the maximum allowable level established by the EU legislation (0.05 mg/kg) [12]. As the main sources of Cd are recognized as the presence in sewage sludge and smelting from the nearby Pljevlja Thermal Power Plant (Zbljevo), or mining from the Pljevlja coalmine (Potrlica).
The Sr content in honey samples from our study varied from 0.07–0.12 μg/kg and was in the same range as those from Serbia (0.09–0.19 μg/kg [13]). The Sr content in honey samples from our study varied from 0.07–0.12 µg/kg and was in the same range as those from Serbia (0.09–0.19 µg/kg [13]).
The most abundant element in honey samples was kalium, followed by Ca, Mg, Na, and Fe. In our study, we found that the content of kalium and magnesium correlated with pH. The average
The most abundant element in honey samples was kalium, followed by Ca, Mg, Na, and Fe. In our study, we found that the content of kalium and magnesium correlated with pH. The average levels of K content ranged from 713–2589.33 mg/kg and was in the same range with those from Croatia (304.7–2824.4 mg/kg [15]), but lower in comparison with the maximum values established for honeys from Bosnia and Herzegovina (14.81–4895.73 µg/kg [16]). On the other hand, the range of concentrations of kalium in the honey from Serbia (400–1755 mg/kg [1,13]) and Slovenia (1090–1220 mg/kg [18]) were lower. The Mg content in honey samples from our study varied from 29.52 to 76.33 mg/kg. In neighboring countries the Mg content in honey varied in a similar range, as: 28.83 to 101.50 mg/kg [13] in Serbia, 2.18 to 166.04 mg/kg [16] in Bosnia and Herzegovina and from 8.02 to 59.1 mg/kg [16,19] in Croatia.
In our study, we used two chemometric analyses, PCA and LDA, respectively to test similarities between honey samples of honey of different geographical origins. Both applied methods separated the regions exposed to industrial pollution (Mijakovi´ci, Potrlica, and Zbljevo) which are characterized by detectable content of Cd, Pb and Sr in their honey samples.
Using LDA it's possible to evaluate the capacity to correctly predict the group to which the unknown samples belong. In our study LDA analysis performed on the geographical origin revealed that the cross-validation classification was correct for 79.17% of the samples. The obtained values are in the range for those from Serbia (Zlatibor: 94.73%, Vojvodina: 70.58% [1]). On the other hand, our value was greater than those reported in the case of Romania honeys where only 21.2% were correctly classified according to their geographical origin [4].
The smallest percentage of good classification was achieved in the case of honeys from Katunska. Of the three samples from the latter region, only one was correctly classified, while the other two being misclassified as Piperi and Zbljevo, respectively. It is known that large numbers of beekeepers (especially from Katunska) in a part of the year (most often in summertime) move their bee colonies to geographically distant areas. On the other hand, the highest percentage of good classification was achieved in the case of honeys from Piperi, Zeta, Mijakovi´ci, and Zbljevo. One cause may be that most of these sites (i.e., Zeta, Mijakovi´ci, and Zbljevo) are more exposed to industrial pollution, resulting in increased concentration of heavy metals (Pb, Cd, and Sr showing significant difference (*p* < 0.05) between studied regions) in their honeys, which, in turn, increase the success rate of the classification of honey according to their geographical origin.
**Author Contributions:** Investigation, V.V.-P., S.B.-V. and N.B.; statistical analysis, A.S.; writing—original draft preparation, V.P.; writing—review and editing, V.V.-P. and V.P. All authors have read and agreed to the published version of the manuscript.
**Funding:** This research received no external funding.
**Conflicts of Interest:** The authors declare no conflict of interest.
#### **References**
© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
### *Article*
## **Geographical Origin Discrimination of Monofloral Honeys by Direct Analysis in Real Time Ionization-High Resolution Mass Spectrometry (DART-HRMS)**
#### **Vincenzo Lippolis 1,\* , Elisabetta De Angelis <sup>1</sup> , Giuseppina Maria Fiorino <sup>1</sup> , Annalisa Di Gioia <sup>1</sup> , Marco Arlorio <sup>2</sup> , Antonio Francesco Logrieco <sup>1</sup> and Linda Monaci <sup>1</sup>**
### Received: 27 July 2020; Accepted: 28 August 2020; Published: 1 September 2020
**Abstract:** An untargeted method using direct analysis in real time and high resolution mass spectrometry (DART-HRMS) combined to multivariate statistical analysis was developed for the discrimination of two monofloral (chestnut and acacia) honeys for their geographical origins—i.e., Italy and Portugal for chestnut honey and Italy and China for acacia honey. Principal Component Analysis, used as an unsupervised approach, showed samples of clusterization for chestnut honey samples, while overlapping regions were observed for acacia honeys. Three supervised statistical approaches, such as Principal Components—Linear Discriminant Analysis, Partial Least Squares—Discriminant Analysis and k-nearest neighbors, were tested on the dataset gathered and relevant performances were compared. All tested statistical approaches provided comparable prediction abilities in cross-validation and external validation with mean values falling between 89.2–98.4% for chestnut and between 85.8–95.0% for acacia honey. The results obtained herein indicate the feasibility of the DART-HRMS approach in combination with chemometrics for the rapid authentication of honey's geographical origin.
**Keywords:** monofloral honey; direct analysis in real time (DART); high resolution mass spectrometry (HRMS); geographical origin; chemometrics
#### **1. Introduction**
Honey is a complex and high-quality natural product containing a wide range of nutritional and therapeutic properties but with a limited production and high commercial prices. Honey is defined by European Union legislation as the natural sweet substance produced by bees of *Apis mellifera* species from nectar or sugary secretions of plants, as well as from excretions of plant-sucking insects on the living parts of plants [1]. Both the European Union and Codex Alimentarius laws establish that the geographical origin, in terms of country of production, must be indicated on the label, also supplemented by specific reference to the floral or vegetable origin. Moreover, in the case of blends of honey, their origin should be declared as a "blend of EC honeys", "blend of non-EC honeys" or "blend of EC and non-EC honeys" [1,2].
Geographical and botanical origins of honey account for the peculiar chemical composition and organoleptic characteristics of the final product [3,4]. Monofloral honeys, mostly deriving from a single plant species (at least 45% of pollen grains), may considerably differ in their sensory properties with highly prominent flavor and aroma. Acacia (*Robinia pseudoacacia*) honey is one of the most consumed monofloral honeys in Europe, being appreciated for its permanently liquid state, light color, floral aroma and sweet and delicate taste [5]. Similarly, chestnut (*Castanea sativa*) honey is considered one of the most delicious and high-quality honeys, being a very good source for nectar and pollen [6,7]. For these reasons, monofloral honeys, and in particular those derived from acacia and chestnut, have recently gained consumer preferences, with an increased demand and commercial value [4]. Due to their increased commercial value, monofloral honeys are highly susceptible to fraudulent practices through mislabeling and mixing with cheaper and lower-quality honeys or with various sugar syrups.
Honey is produced in different areas of the world, with more than 2.3 million tonnes produced worldwide in 2018, with China and Turkey as main producers [8]. China is also the largest exporter of honeys in the world, while in Europe, Portugal is the country bearing the highest number of geographical protected labels on honey [9]. Honey composition is quite variable and strictly linked to its floral source and geographical origin, but external factors, including processing, packaging and storage conditions, could play an important role. Although Italy is one of the EU countries with the highest honey production [10], the market demand for honey is higher than domestic production, therefore a substantial amount of honey is imported from elsewhere in Europe and from third-world countries, in which production does not always meet the high food safety standards required. This can lead to honey mislabeled with regard to its geographical and/or botanical origin [11].
The traceability certifying the geographical origin of food products is of primary importance for traders and producers, as well as to reinforce consumer trust. The complex task of the determination of food origin is commonly applied to control products in both customs control and self-control programs of the food industry.
Melissopalynological analysis of pollen is the most used approach for the botanical and geographical origin classification of honey, as the pollen spectrum is strictly related to the environment where the nectar is collected [12]. This analysis is often complemented by other analytical methodologies, mainly based on chromatographic techniques, to assure the honey authenticity [13]. Often, the use of conventional and targeted methods is time-consuming and not sufficient to guarantee the evaluation of complex matrices, including honeys. For this reason, the development of rapid and reliable non-targeted analytical approaches, such as fingerprinting and profiling methods, is highly demanded. Indeed, these methods combined to chemometric tools allow for the detection of a high number of metabolites, leading to samples based on their pattern.
Several analytical techniques, mainly based on nuclear magnetic resonance [5,14,15], Raman and infrared spectroscopy [16,17], mass spectrometry [18–20], electronic tongue [21,22] and electronic nose [23,24], in combination with chemometrics, have been applied to discriminate the geographical origin of honey.
The use of ambient mass spectrometry (AMS) is continuously increasing in the field of metabolomic fingerprinting as a high-throughput alternative to more traditional hyphenated methods for authentication issues [25]. Among AMS techniques, direct analysis in real-time mass spectrometry (DART-MS), being simple and requiring a very limited sample preparation, has been shown to be the most promising and versatile technique, proving to be a rapid tool in the assessment of food authenticity and food quality, also thanks to the use of fast and streamlined protocols [25–27]. Such an approach offers several advantages over the conventional techniques, including direct sample analysis in open atmosphere, high sample throughput and minimal or no sample preparation requirements, the soft ionization of a wide range of both polar and apolar compounds. Several papers have been recently published demonstrating the applicability of DART-MS to assess food authenticity and detect food adulterations of olive oil [28], beer [29], wine [30], animal fat [27], milk [31] and salmon [32]. Only one paper reported the applicability of DART-MS to the discrimination of geographical origin of food—i.e., garlic produced in Czech Republic, Spain and China [33]. Regarding honey products, DART-HRMS was used as alternative approach for the determination of 5-hydroxymethylfurfural [34,35]. To the best of our knowledge, no studies based on DART-MS have been performed to date for the assessment of geographical origin of honeys.
In this context, the aim of this study was to demonstrate the feasibility of the DART-HRMS technique for the discrimination of the geographical origin of honeys. Specifically, a rapid and suitable non-targeted DART-HRMS method in combination with multivariate statistical analysis was developed and validated to discriminate two monofloral honeys varieties (i.e., chestnut and acacia) for their geographical origin (i.e., Italy and Portugal and Italy and China, respectively). Different statistical classification models were investigated and applied to the analysis of honey samples and performance results were compared.
#### **2. Materials and Methods**
#### *2.1. Chemicals and Reagents*
Methanol (HPLC grade) was purchased from Sigma-Aldrich (Milan, Italy). Ultrapure water was produced by a Milli-Q® Direct system (Merck KGaA, Darmstadt, Germany). Helium (99.9995% purity) was provided by Sapio S.r.l. (Bari, Italy). Regenerate cellulose (RC) syringe filters with 0.2 µm of porosity were purchased by from VWR International (Milan, Italy). OpenSpot (OS) Sample Cards were purchased by Ion Sense Inc. (Saugus, MA, USA).
#### *2.2. Honey Samples*
A total of 234 monofloral honey samples commonly found in marketplaces and collected in different countries with certified origins were selected for this study. Specifically, 117 chestnut honey samples were collected from Italy (39) and Portugal (78), while 117 acacia honey samples were collected from Italy (78) and China (39). The authenticity of the monofloral honeys was assessed by internal certified protocols performed by Coop Italia Soc. Cooperativa (Casalecchio di Reno, Italy) which provided samples. Only honey samples produced in seasons 2017–2018 were taken into account.
#### *2.3. Sample Preparation*
Sampling and homogenization of honey samples were performed according to AOAC 920.180 protocol [36]. For sample preparation, a rapid protocol aimed at retaining as many honey metabolites as possible—thus to obtain most comprehensive spectra applicable for discriminating between different geographical origin—was optimized for the DART-HRMS analysis. In particular, an aliquot (1 g) of homogenized honey was added to a mixture of MeOH/H2O (1:1, *v*/*v*), (50 mL) and the sample was vortexed for 3 min. After filtration using 0.2 µm RC syringe filter, the filtered extract was directly analyzed by DART-HRMS.
#### *2.4. DART-HRMS Analysis*
DART-HRMS analyses were carried out by using a DART ionization source SI-140-GIST (DART Thermo Ion Max Vapur Interface, Ion Sense Inc., Saugus, MA, USA) coupled to an Exactive™ monostage Orbitrap™ High Resolution mass spectrometer (Thermo Fisher Scientific, San Jose, CA, USA). An aliquot (2 µL) of the honey extract was placed onto the metallic grid of the OpenSpot® sample cards and kept at 60 ◦C for 5 min to facilitate solvent evaporation before its introduction into the DART source holder. The operating conditions of the DART source were: positive ion mode; helium flow of 3.2 L/min for 1 min and heated at 250 ◦C; discharge needle voltage kept at −6 kV; grid electrode voltage set to 250 V; distance between DART exit and MS inlet set at 5 mm. The operating conditions of DART source were set by DART-SVP controller (v. 4.0.x). The main settings of the Exactive™ mass spectrometer were the following: mass scan range of 100–600 m/z; resolution set at 25,000 (FWHM at m/z 200); microscan number of 4; Automatic Gain Control (AGC) Target of 3 <sup>×</sup> <sup>10</sup>−<sup>6</sup> ; maximum injection time (IT) of 250 ms; capillary voltage set to 30 V; tube lens voltage set to 65 V; capillary temperature kept
at 250 ◦C. Calibrations of the MS system were periodically performed by the direct infusion ESI-MS approach of the positive ion calibrating solution, provided by the manufacturer, in order to obtain a mass accuracy lower than 5 ppm. The MS system was controlled by using the Xcalibur™ v. 2.1 software (Thermo Fisher Scientific, San Jose, CA, USA).
To carry out the subtraction of the spectral background, a blank open spot card was acquired before analyzing each sample by DART-HRMS acquiring the relevant spectrum for 30 s.
#### *2.5. Data Processing and Statistical Analysis*
In the first step of data processing, DART-HRMS spectra acquired in the time range of 30 s were averaged and then subtracted of spectral background by using the Xcalibur™ software. Successively, for each honey sample, the full list of accurate m/z ratios and peak intensities obtained was exported and processed by MetaboAnalyst 3.0 (http://www.metaboanalyst.ca/) [37,38] for peak matching and alignment with mass tolerance of 0.25, imputation of missing values (replacing missing elements by using the half of the lowest measured peak intensity) and data filtering (by using Interquartile Range approach). Successively, after pre-processing obtained by data centering, the dataset was submitted to multivariate statistical analyses performed by V-Parvus software (release 2010, http://www.parvus.unige.it, Genova, Italy).
Principal Component Analysis (PCA) was used as an unsupervised technique to evaluate the presence of outliers. Specifically, PCA was applied to each single group of monofloral honey samples of different geographical origin, observing the relevant influence plots and excluding samples identified as extreme outliers. To establish the exact number of Principal Components (PCs) to be used to build PCA models, the Non-linear Iterative Partial Least Squares (NIPALS) algorithm was applied using V-fold of 10 (cross validation process, CV = 10). PCA was also used as exploratory technique to visualize the presence of natural sample clustering between monofloral honey samples in relation to their geographical origin [39].
Afterwards, three supervised pattern recognition techniques—i.e., Linear Discriminant Analysis (LDA), Partial Least Squares Discriminant Analysis (PLS-DA) and k-nearest neighbors (k-NN) [40], were exploited to classify monofloral honey samples on the basis of their geographical origin. For this purpose, the two data matrices were randomly split in two subsets: a modelling set (containing 60 samples) and a test set (containing 57 samples). Specifically, for each monofloral honey, a modelling set, composed by 30 samples for each geographical origin, was used to build the three different statistical models. Test sets, consisting of 9 Italian and 48 Portuguese chestnut honey samples and 48 Italian and 9 Chinese acacia honey samples, were used for the validation process.
The chemometric models of PCA-LDA was built by firstly performing PCA test to reduce the number of variables that exceeded the number of objects, thus preventing model overfitting; then the selected scores were used as classification variables for LDA [41,42]. Indeed, the number of variables should not exceed (n-g)/3, where n is the number of objects and g is the number of categories. Considering that modelling sets were composed by 60 objects (number of samples) and 2 categories (number of geographical origins) the maximum number of variables should be approximately 19.
The appropriate numbers of principal components, latent variables and k values, respectively, for PCA-LDA, PLS-DA and k-NN models were established by evaluating those determining the lowest prediction error rate in cross-validation (cross-validation segments, V = 10). This parameter guarantees to improve feature variables and, at the same time, to avoid model overfitting. Model performances for PCA-LDA, PLS-DA and k-NN, expressed as percentages, were compared with reference to their recognition ability—i.e., the ability to correctly classify samples of the modelling set—prediction ability in cross-validation (CV)—i.e., the ability to correctly classify samples of a test set generated in a V-fold cross validation—and prediction ability in external validation—i.e., the ability to correctly classify samples of the test set.
#### **3. Results and Discussion**
In the present study, the real-time mass spectrometry (DART-MS) combined with chemometric analysis, was used for the first time to the discrimination of two kind of monofloral honey samples, namely chestnut and acacia, based on their geographical origin. As for chestnut, Italian and Portuguese honey samples were compared to each other, while for acacia, Italian honeys were compared with samples from China.
Figure 1 reports four representative DART-HRMS average spectra, after blank subtraction, obtained for the chestnut honey extracts of Italian (Figure 1a) and Portuguese (Figure 1b) samples and acacia honey extracts of Italian (Figure 1c) and Chinese (Figure 1c) samples. *Foods* **2020**, *9*, x FOR PEER REVIEW 5 of 12 Figure 1
**Figure 1.** *Cont.*
**Figure 1.** Representative DART-HRMS positive ion spectra acquired for the sample extracts of chestnut honeys from Italy (**a**), chestnut honeys from Portugal (**b**), acacia honeys from Italy (**c**) and acacia honeys from China (**d**). NL: Normalization level. **Figure 1.** Representative DART-HRMS positive ion spectra acquired for the sample extracts of chestnut honeys from Italy (**a**), chestnut honeys from Portugal (**b**), acacia honeys from Italy (**c**) and acacia honeys from China (**d**). NL: Normalization level.
At first, a preliminary PCA was performed on pre-processed spectra of chestnut and acacia honey samples in order to explore the presence of outlier samples. PCA score plots highlighted that, in the case of chestnut samples, seven PCs described 96.3% of total variance for samples from Italy while nine PCs described 93.0% of total variance for samples from Portugal. In the case of acacia honey samples, PCA models showed that eight PCs described 93.7% of total variance for samples from Italy while nine PCs described 91.0% of total variance for samples from China. The absence of outliers in all classes was demonstrated using influence plots where the Mahalanobis distance was plotted versus sample residual. Subsequently, an explorative PCA was performed using the entire data set to obtain an overview At first, a preliminary PCA was performed on pre-processed spectra of chestnut and acacia honey samples in order to explore the presence of outlier samples. PCA score plots highlighted that, in the case of chestnut samples, seven PCs described 96.3% of total variance for samples from Italy while nine PCs described 93.0% of total variance for samples from Portugal. In the case of acacia honey samples, PCA models showed that eight PCs described 93.7% of total variance for samples from Italy while nine PCs described 91.0% of total variance for samples from China. The absence of outliers in all classes was demonstrated using influence plots where the Mahalanobis distance was plotted versus sample residual.
of the data distribution for each monofloral honey. Figure 2 shows the PCA score plot (PC1 vs. PC2) obtained for chestnut honey samples (Figure 2a) and for acacia honey samples (Figure 2b). A discrete visual clustering of the objects on the basis of their geographical origin was observed for chestnut honeys (PC1 and PC2 explained 88.6% and 10.2% of the total variance, respectively), while overlapping regions were observed for acacia honeys with a modest clustering for their geographical origin (with 88.4% and 9.7% of the total variance explained by PC1 and PC2, respectively). Additionally, by analyzing the score plots of the remaining PCs no visual clusterization was observed. Subsequently, an explorative PCA was performed using the entire data set to obtain an overview of the data distribution for each monofloral honey. Figure 2 shows the PCA score plot (PC1 vs. PC2) obtained for chestnut honey samples (Figure 2a) and for acacia honey samples (Figure 2b). A discrete visual clustering of the objects on the basis of their geographical origin was observed for chestnut honeys (PC1 and PC2 explained 88.6% and 10.2% of the total variance, respectively), while overlapping regions were observed for acacia honeys with a modest clustering for their geographical origin (with 88.4% and 9.7% of the total variance explained by PC1 and PC2, respectively). Additionally, by analyzing the score plots of the remaining PCs no visual clusterization was observed.
**Figure 2.** *Cont.*
**Figure 2.** PC1 vs. PC2 scatter plots for monofloral chestnut (**a**) and acacia (**b**) honey samples. Geographical origins: Italy (black filled circle), Portugal (grey filled triangle), China (grey filled rhombus).
These results were confirmed by analyzing the Fisher weight (FW) values of the principal components, which measure the between-class variance/within-class variance ratio. Indeed, FW values resulted to be 2.64 for the PC1 of chestnut honeys samples and lower than 1 for all the remaining PCs of both data sets (data not shown). These results indicated that the PCA was not able to discriminate honey samples on the basis of their geographical origin; therefore, it was necessary to treat data with three different supervised discriminant techniques—i.e., PCA-LDA, PLS-DA and k-NN. These classification techniques were tested on both chestnut and acacia honey samples previously split into two subsets: a modeling set and a test set. Overall, results are indicated in Tables 1 and 2, for chestnut and acacia honeys, respectively.
**Table 1.** Model performances in terms of recognition, cross validation (CV) prediction abilities and external prediction to classify chestnut honeys based on their geographical origin.
a : Italy; <sup>b</sup> : Portugal; <sup>c</sup> : Cross Validation; <sup>d</sup> : Principal Components—Linear Discriminant Analysis; <sup>e</sup> : Partial Least Squares—Discriminant Analysis; <sup>f</sup> : k-nearest neighbors.
As for LDA, PCA was used as strategy for variable reduction and to avoid model overfitting. The number of PCs (seven and nine for chestnut and acacia honeys, respectively) to be used to build the PCA-LDA models was selected on the basis of the error in prediction cross validation that has to be the lowest (CV procedure with V = 10). The PCA-LDA models provided mean value of recognition ability of 98.4% for chestnut honeys (Table 1) in both classification and CV prediction and 95.0 and 93.4% for acacia honeys (Table 2) in classification and CV prediction, respectively. The model applicability was also tested by using the test set providing mean prediction abilities of 90.3 and 89.2%, for chestnut and acacia honeys, respectively (Tables 1 and 2).
**Table 2.** Model performances in terms of recognition, cross validation (CV) prediction abilities and external prediction for all models built to classify acacia honeys based on their geographical origin.
a : Italy; <sup>b</sup> : Portugal; <sup>c</sup> : Cross Validation; <sup>d</sup> : Principal Components—Linear Discriminant Analysis; <sup>e</sup> : Partial Least Squares—Discriminant Analysis; <sup>f</sup> : k-nearest neighbors.
PLS-DA was applied as an alternative multivariate statistical approach of classification offering the advantage to avoid variables reduction processes. Specifically, by applying a 10-fold cross-validation, 10 and 12 latent variables (LVs) were found to produce the optimal model complexity for chestnut and acacia honey data sets, respectively. In these conditions, mean recognition rates were higher than 96.7% in both cases (Tables 1 and 2). Specifically, all Italian samples was correctly classified, while one Portuguese and two Chinese samples were not correctly assigned. The mean CV prediction rates were 96.7 and 95.0%, for chestnut and acacia honeys, respectively. In addition, mean prediction abilities of 89.2% and 85.8% for chestnut and acacia honeys samples, respectively, were obtained for the external validation procedure (Tables 1 and 2).
In the case of k-NN, the prediction error rate in cross-validation (V = 10) was calculated for each different k value. The smallest k value determining the lowest error was 3 for both data sets and therefore it was selected as the optimal value. The k-NN models provided mean recognition abilities in the range between 95.0–98.4%, while CV predictions were of 98.4 and 91.7%, for chestnut and acacia honeys samples, respectively. Finally, mean prediction abilities of 91.4% and 90.3% were obtained in the external validation for chestnut and acacia honeys, respectively.
The results herein obtained were in accordance with a similar study focused on the geographical authentication of Italian honey based on an NMR-metabolomic approach [5]. The authors developed a PLS2-DA model able to correctly discriminate 100% of Italian honeys from Eastern European ones. In another study, MIR analysis in combination with a PCA-LDA model were found able to distinguish geographical origins of monofloral honeys from Switzerland, Germany, and France, with prediction abilities ranged from 76 to 100% [17] although only a limited number of samples was used for the analysis.
In the current study, the DART-HRMS untargeted approach coupled with three supervised techniques, such as PCA-LDA, PLS-DA and k-NN, were investigated for discriminating Italian chestnut and acacia honey from Portuguese and Chinese samples. The results showed that all developed models provided acceptable and comparable prediction abilities, highlighting the robustness of the entire method, its applicability being unaffected by the statistical approach used to assess the authenticity of unknown samples. Moreover, these results demonstrated that DART-HRMS technique provides informative experimental data useful to build up appropriate models for the discrimination of monofloral honey samples on the basis of their geographical origin.
#### **4. Conclusions**
In this study, a rapid, easy-to-perform and low-cost method based on DART-HRMS analysis combined to multivariate statistical analysis was successfully developed and applied to classify monofloral honeys for their geographical origins, such as Italy and Portugal for chestnut samples and Italy and China for acacia samples. Specifically, three supervised approaches—i.e., PCA-LDA, PLS-DA and k-NN were evaluated. All tested models provided high and comparable recognition and prediction abilities in cross-validation and external validation, with mean values ranging from
89.2% and 98.4%. The performances of the proposed DART-HRMS method makes it an effective tool to assess the authenticity of honeys, for both industries of sector against unfair advantages of competitors and control bodies to fight food frauds. Future efforts will be directed to improve the current predictive models in order to discriminate honey samples from different production seasons and identify potential markers useful for developing a DART-HRMS target method aimed at honey authentication. Moreover, the use of the DART-HRMS approach, generating huge information in a single run, would be a useful tool for discriminating honey samples with similar organoleptic characteristics but different quality levels.
**Author Contributions:** Conceptualization: V.L., M.A., A.F.L. and L.M.; methodology, V.L. and L.M.; validation, V.L. and E.D.A.; formal analysis, G.M.F. and A.D.G.; writing—original draft preparation, V.L.; writing—review and editing, V.L., E.D.A., G.M.F., A.D.G., A.F.L., M.A. and L.M.; supervision, A.F.L. and L.M.; funding acquisition, M.A. All authors have read and agreed to the published version of the manuscript.
**Funding:** The present research has received funding from the European Union's Seventh Framework Programme for research, technological development and demonstration, under grant agreement No. 613688 2 "Food Integrity". The equipment used in this work was funded by project BioNet – PTP – "Biodiversità per la valorizzazione e sicurezza delle produzioni alimentari tipiche pugliesi (codice n. 73, PO Regione Puglia FESR 377 2000-2006)".
**Acknowledgments:** The authors thank Salvatore Cervellieri for his support in the statistical analysis and Fernando Gottardi from Coop Italia (Casalecchio di Reno, Italy) for providing the honey samples analyzed in the present study.
**Conflicts of Interest:** The authors declare no conflict of interest.
### **References**
© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
## **Machine Learning Approaches Applied to GC-FID Fatty Acid Profiles to Discriminate Wild from Farmed Salmon**
**Liliana Grazina <sup>1</sup> , P. J. Rodrigues <sup>2</sup> , Getúlio Igrejas <sup>2</sup> , Maria A. Nunes <sup>1</sup> , Isabel Mafra 1,\* , Marco Arlorio <sup>3</sup> , M. Beatriz P. P. Oliveira <sup>1</sup> and Joana S. Amaral 4,\***
Received: 23 September 2020; Accepted: 4 November 2020; Published: 7 November 2020
**Abstract:** In the last decade, there has been an increasing demand for wild-captured fish, which attains higher prices compared to farmed species, thus being prone to mislabeling practices. In this work, fatty acid composition coupled to advanced chemometrics was used to discriminate wild from farmed salmon. The lipids extracted from salmon muscles of different production methods and origins (26 wild from Canada, 25 farmed from Canada, 24 farmed from Chile and 25 farmed from Norway) were analyzed by gas chromatography with flame ionization detector (GC-FID). All the tested chemometric approaches, namely principal components analysis (PCA), t-distributed stochastic neighbor embedding (t-SNE) and seven machine learning classifiers, namely k-nearest neighbors (kNN), decision tree, support vector machine (SVM), random forest, artificial neural networks (ANN), naïve Bayes and AdaBoost, allowed for differentiation between farmed and wild salmons using the 17 features obtained from chemical analysis. PCA did not allow clear distinguishing between salmon geographical origin since farmed samples from Canada and Chile overlapped. Nevertheless, using the 17 features in the models, six out of the seven tested machine learning classifiers allowed a classification accuracy of ≥99%, with ANN, naïve Bayes, random forest, SVM and kNN presenting 100% accuracy on the test dataset. The classification models were also assayed using only the best features selected by a reduction algorithm and the best input features mapped by t-SNE. The classifier kNN provided the best discrimination results because it correctly classified all samples according to production method and origin, ultimately using only the three most important features (16:0, 18:2n6c and 20:3n3 + 20:4n6). In general, the classifiers presented good generalization with the herein proposed approach being simple and presenting the advantage of requiring only common equipment existing in most labs.
**Keywords:** authenticity; fish; *Salmo salar* L.; fatty acids; mislabeling; chemometrics; machine learning
#### **1. Introduction**
In recent decades, the consumption of fish has been increasingly recommended due to its health benefits, mainly related to the prevention of cardiovascular diseases [1]. In particular, fatty fishes from cold waters, such as salmon, are frequently rich in polyunsaturated fatty acids (PUFA), including the essential fatty acids linoleic (18:2n6) and α-linolenic (18:3n3), but also in several omega-3 PUFA such as eicosapentaenoic (EPA, C20:5n3) and docosahexaenoic (DHA, 22:6n3) acids. Besides being components of cell membranes, omega-3 PUFA are involved in the biosynthesis of eicosanoids and have been shown to influence health by affecting cell signaling cascades and gene expression, resulting in decreased expression of inflammatory and atherogenesis-related pathways [2,3]. Moreover, different studies showed that omega-3 PUFA play an important role in altering blood lipid profiles and associate their consumption with improved cardiovascular function and decreased risk of atherosclerosis and peripheral arterial disease [2,3].
In addition to these benefits, fish is largely consumed for its nutritional value and sensory aspects, making it one of the most traded food commodities. In this sense and considering that the world's wild fish stocks are limited, the production of farmed fish has been steadily increasing in recent last years. In fact, according to the Food and Agriculture Organization (FAO) Globefish Highlights, world fisheries capture was 92.5 million tonnes in 2017 with this figure expected to decrease to 91.3 million tonnes in 2019, by the contrary, fish capture arising from aquaculture is expected to grow from 80.1 to 86.5 million tonnes in the same period [4]. Concerning salmon, from 2000 to 2014, a much stronger increase was verified for aquaculture production (from 898,800 to 2,326,300 tonnes) compared to that of the world's capture of wild salmon (from 728,000 to 879,000 tonnes) [5]. Aquaculture allows wider consumer access to fish generally at more affordable costs, though it is known that fatty acid composition can significantly vary according to its production method (wild vs. aquaculture). Particularly for salmon, it has been reported that wild salmon generally present higher contents of valuable omega-3 PUFA [6–8]. This aspect, together with particular organoleptic characteristics, has driven several consumers to prefer wild salmon. Considering the limited availability of this type of salmon and its growing demand, prices have been increasing significantly, resulting in this product being prone to adulteration by origin mislabelling or even substitution with other lower-cost fish [9–11]. Whereas fish species authentication can be performed using well established and straightforward DNA-based methods [12], different approaches have been proposed so far to assess the origin of fish with respect to production method. These include, mainly, the use of nuclear magnetic resonance (NMR) [13,14], isotope ratio analysis [15,16], lipidic profile [17,18] or a combination of these [11,19–21]. Excellent discrimination (100%) between wild and farmed Atlantic salmon was reported by Aursand et al. [13] by applying support vector machines (SVM) to data obtained by 13C NMR. In another study of the same group, the lipid extract was analyzed by 13C NMR and by gas chromatography with flame ionization detector GC-FID for fatty acid composition to discriminate between wild and farmed Atlantic salmon and assign the origin of the aquaculture samples to the farms included in the study [19]. The application of chemometrics to the reference farmed fish showed very good results for both approaches, but, surprisingly, slightly better for GC-FID data. The use of stable isotope analysis based on isotope ratio mass spectrometry (IRMS) is also a promising approach, especially when combined with chemical composition analysis, notably fatty acids [11,15,21]. Yet, previous works have demonstrated that lipidic profile is sufficient to establish the production method of salmon samples, particularly when combined with chemometric analysis [8,19,20]. Recently, Fiorino et al. [8] analyzed the lipid extracts obtained from a total of 100 samples of farmed and wild salmon by direct analysis in real time (DART) coupled to high resolution mass spectrometry (HRMS). The proposed methodology showed to be fast and allowed a good discrimination between the two groups (wild vs. farmed), though without differentiating the geographical origin of the farmed fish. Moreover, the referred approach requires advanced and expensive equipment, which is not available in most control quality/analytical laboratories. In the present study, the fatty acid composition of the same samples of wild and farmed salmon used in the work of Fiorino et al. [8] was analyzed by GC-FID, an affordable equipment commonly available
in most laboratories. Subsequently, the obtained data were submitted to advanced chemometric analysis to establish the most suitable classifier able to discriminate the origin of salmon samples (wild vs. farmed, and the geographical origin among farmed samples) with the minimum possible computational effort.
#### **2. Materials and Methods**
#### *2.1. Samples*
In this study, a total of 100 authentic salmon samples obtained in the framework of the EU-funded project FOODINTEGRITY (Working Package 18) were analyzed. The samples included 26 wild salmon captured in Canada, and 74 farmed salmon samples from aquaculture farms in Canada (25), Norway (25) and Chile (24). No information was available about the gender of each specimen, neither of the diet or farming conditions used. The samples (entire fish) were transported frozen to the laboratory (Meriex Nutriscience, Chicago, IL, USA), allowed to defrost overnight at refrigerated temperature, and filleted in a cold room (4 ◦C). After removing the bones and skin, the muscles were grinded and distributed in labelled glass jars containing approximately 200 g each. The jars were immediately frozen and then shipped under freezing conditions (−20 ◦C) to the participating laboratories in different countries. After arriving, the samples were kept at −20 ◦C and submitted to lipid extraction as soon as possible.
#### *2.2. Lipid Extraction*
Lipids were extracted based on the Bligh and Dyer protocol [22] with some modifications. Briefly, about 13 g of each minced fillet were added with 13 mL of NaCl (1%) and 100 µL of butylated hydroxytoluene (BHT) (0.01% in n-hexane) to avoid oxidation, and homogenized for 1 min using an Ultra-Turrax at 13,500 rpm, keeping a low temperature by immersing the tube with the sample on ice. After that, 2.5 mL of the homogenate was transferred to a new tube and added with 2.5 mL of chloroform and 5 mL of methanol, both refrigerated. The solution was mixed vigorously by vortexing for 2 min. After centrifuging (4000 rpm, 15 min at 4 ◦C) the upper layer was discarded, and an additional 2.5 mL of refrigerated chloroform was added. After vortexing for 30 s and centrifuging under the same conditions, the chloroformic phase was transferred into a new tube and centrifuged (4000 rpm, 5 min at 4 ◦C). Finally, the chloroformic phase was collected into a previously weighted vial, flushed with a nitrogen stream and stored at −20 ◦C until further analysis. Each sample was submitted to independent extractions (*n* = 3).
#### *2.3. Fatty Acids Analysis by GC-FID*
Fatty acids were methylated using acid-catalysed trans-methylation with BF<sup>3</sup> [23]. Firstly, the lipidic chloroformic extracts, previously stored at −20 ◦C, were dried under nitrogen and the tubes weighted to calculate the extraction yield. After dissolving the obtained lipids in 1 mL of n-hexane, for each sample, the volume containing 12.5 mg of lipids was transferred for a glass tube and dried under nitrogen. After adding 100 µL of BHT (0.01% in n-hexane) to prevent oxidation phenomena, fatty acid methyl esters were prepared. For that purpose, 1.25 mL of KOH (0.5 M) in methanol were added and the mixture was heated for 10 min at 100 ◦C after vortex-mixing vigorously. After cooling down the tubes, 1.0 mL of 14% boron-trifluoride in methanol (≥99.0% purity) (Sigma-Aldrich, Steinheim, Germany) was added to the solution, which was homogenized by vortexing, and the tubes heated again for 30 min at 100 ◦C. After completely cooling down the tubes in ice, 2.0 mL of n-hexane high performance liquid chromatography HPLC grade (Merck, Darmstadt, Germany) was added and the solution was vortex-mixed. Then, 1.0 mL of a saturated NaCl solution was added, followed by vigorous mixing and then by centrifuging for 5 min at 3000 rpm to obtain a clear upper phase. After that, 1.5 mL of supernatant was transferred to a new vial, added with anhydrous Na2SO<sup>4</sup> and approximately 1.0 mL of FAME solution was transferred to an injection vial.
GC-FID analysis was carried out in a Shimadzu GC-2010 Plus gas chromatograph equipped with a Shimadzu AOC-20i auto-injector and a flame ionization detector (Shimadzu, Japan). FAME separation was achieved on a CP-Sil 88 silica capillary column (50 × 0.25 mm i.d., 0.20 µm, Varian, Middelburg, The Netherlands). The injector and detector temperatures were 250 ◦C and 270 ◦C, respectively. The oven parameters were set as follows: an initial temperature of 150 ◦C was increased at 3 ◦C/min to 160 ◦C and held for 2.0 min, then it was increased at 3 ◦C/min to 220 ◦C and held for 10 min. Helium was used as the carrier gas at a flow rate of 1 mL/min, and 1 µL of sample was injected using a split ratio of 1:50. Identification of compounds was performed by comparison of their retention times with those of authentic standards mixtures, namely 37 component FAME mix (certified reference material CRM47885) and PUFA nº.1 Marine source (standard 47,033) both from Supelco (Bellefonte, PA, USA). In addition, the fatty acid cis-11-octadecenoate (C18:1n7) was identified with and individual standard also purchased from Supelco. The results were expressed as the relative percentage of each fatty acid, calculated based on the chromatographic peak area. Each lipid extract was injected in duplicate.
#### *2.4. Chemometric Analysis*
#### 2.4.1. Dataset
The data used for chemometrics resulted from the chemical analyzes, totalizing 596 instances (4 chromatograms were excluded due to injection/chromatographic system problems) that were organized into four reasonably well balanced groups, each corresponding to a class of salmon: Norway Farmed (25 salmons), Chile Farmed (24 salmons), Canada Farmed (25 salmons), Canada Wild (26 salmons). Each salmon sample was represented by a block of 6 chromatograms. The number of independent features considered was 17, corresponding to the identified fatty acids.
#### 2.4.2. Statistical Analysis by One-Way ANOVA
The differences between groups were analysed using a one-way analysis of variance (ANOVA) followed by Tukey's honest significant difference post hoc test with *p* = 0.05. The analysis was carried out using the SPSS v. 23.0 program (SPSS v. 23.0; IBM Corp., Armonk, NY, USA).
#### 2.4.3. Data Modelling Tools
The data modelling tools used in this work are based on the Orange 3.24 software, which, in turn, uses libraries from the Scikit-learn, Numpy and Scipy written in Python. The graphical user interface uses the cross-platform Qt framework.
#### Data Visualization by PCA and t-SNE
As a first approach, the possibility of separating the data by classical and linear statistical methods was evaluated. For that purpose, principal component analysis (PCA) was used to check the possibility of obtaining a separation by linear composition in a subspace of principal components based on the PCA projections. When the PCA shows data superposition among groups, it means that the possibility of separating groups in the original dimension space cannot be performed, since the mapping from the original dimension space to the principal component space is always linear [24]. A manner to overcome this issue involves using the t-distributed stochastic neighbor embedding (t-SNE) method, which is able to replicate non-linear mappings in the original data space to the lower dimension [25]. Thus, a non-linear approach by t-SNE was used to observe separations in higher dimensions when they are projected in a two-dimensional space.
#### Machine Learning Classifiers
Several well-known classification models were evaluated, namely k-nearest neighbors (kNN), decision tree, support vector machine (SVM), random forest, artificial neural networks, naïve Bayes and AdaBoost, whose main characteristics are described as follows:
All these models are mappers with non-linear capabilities, each having different methods of statistical induction of knowledge. Thus, some may perform better for certain classification problems than others. For this reason, in this study we used a test bench formed by several models.
All these classifiers were developed/trained, in a first phase, using the 17 features present in the dataset and the obtained classification results used for the assessment of each model.
#### Reduction in Features
Aiming at decreasing the cost of analyses and complexity of data, accelerating the whole process, it is frequently important to reduce the number of features in chemometric analysis. On the other hand, the reduction in the number of features can enhance the generalization capability of the classifier. Therefore, after having the models parameterized using all the features, strategies were developed to explore the reduction in features. The selection of the best features was made using a ranking process that is based on the measurement of information entropy. In this case, the well-known information gain ratio criterion was applied [27]. This criterion measures the uncertainty in how the data are separated based on a specific feature; the value of the information gain ratio is calculated for each feature, representing its separation power in the dataset. The sorting of features, according to these values, establishes their ranking. This criterion is normalized regarding the number of data partitions that the usage of a given feature causes. This mechanism makes it possible to obtain a numerical criterion, independent from the classifying bias (overfitting), prompted by numerous potential partitions of information groups. Thus, in the next step, the minimum number of the best features, in that ranking, was determined, ensuring that the classification model still classifies the data accurately.
Aiming to evaluate model overfitting, assertiveness and generalization assessments of the classification models were made using both external and full-cross validation. For external validation, the test dataset was obtained by splitting the data into 20:80. For the cross-validation scheme a mechanism of leave-one-sample-out (each sample corresponding to a block of 6 chromatograms) was used. Moreover, that scheme allowed us to parameterize the models by observing the validation performance given by the average of all six-chromatogram groups. The used performance indicators were the accuracy (CA) and F1. F1 is a more revealing measure of the practical performance that a classification model presents, being more sensitive to poorly classified instances. Moreover, an assertiveness analysis was made by using confusion matrices.
During the process of feature reduction, the performance of the classifiers was tested using a successive bisection approach. Starting from a set of classification models that normally provide high assertiveness rates and using all the features sorted by the information gain ratio, the following method (Algorithm 1) was developed and applied. This algorithm allows for the optimization of the search for the minimum number of features to classify the samples, with an arbitrary minimum of 99% of accuracy.
#### **Algorithm 1** Searching the optimal number of features
Given a set of features *F* of *n* elements with gain ratio values *F*<sup>0</sup> , *F*<sup>1</sup> , *F*<sup>2</sup> , . . . , *Fn*−*<sup>1</sup>* sorted such that *F*<sup>0</sup> > *F*<sup>1</sup> > *F*<sup>2</sup> . . . > *Fn*−*<sup>1</sup>* , and the *accuracym* being the correctness classifying the dataset using the first *m* features. The following algorithm is based on the binary search to find the index *m* in *F* that corresponds to the minimum index to classify the dataset properly.
The algorithm is repeated independently for each of the classifiers under analysis. The minimum number of features is selected to further actions when the accuracy is ≥ 99% for at least one model. For each machine learning classifier, a trial-and-error approach was used to find the best parametrization, with classifiers being tuned at two stages. At the first stage, Algorithm 1 was applied to all the models using the maximum number of features (seventeen). This tuning aims at obtaining a good classification, concerning the dataset, for each classifier. The adjustment was done manually, in a trial-and-error fashion, changing the hyperparameters associated to each model. In this phase, to get a good functional response (selection) from Algorithm 1, it is not necessary to have a perfect tuning of the classifiers.
After this, the minimum number of features required to produce good classifications (accuracy of 99%) on a classifier are known. Thus, at the second stage, eventually, one could make new adjustments to the classifier models to improve functional performance subjected to the new subset of features selected after applying Algorithm 1. The details of the final parameters used for the best models are shown in Table 1. Figure 1 schematically describes the chemometric approaches and main process pipeline used in this work. *Foods* **2020**, *9*, x FOR PEER REVIEW 7 of 15 After this, the minimum number of features required to produce good classifications (accuracy of 99%) on a classifier are known. Thus, at the second stage, eventually, one could make new adjustments to the classifier models to improve functional performance subjected to the new subset of features selected after applying Algorithm 1. The details of the final parameters used for the best
**Table 1.** Details of the parametrization used to tune each of the final classification models. models are shown in Table 1. Figure 1 schematically describes the chemometric approaches and main process pipeline used in this work.
> kNN: k-nearest neighbors; SVM: support vector machine; ANN: artificial neural networks. SAMME.R; Regression loss function: Square.
kNN: k-nearest neighbors; SVM: support vector machine; ANN: artificial neural networks.
**Figure 1.** Main process pipeline. PCA: principal components analysis; t-SNE: t-distributed stochastic neighbor embedding. neighbor embedding.
#### **3. Results and Discussion 3. Results and Discussion**
#### *3.1. Fatty Acids Composition 3.1. Fatty Acids Composition*
Figure 2 shows representative chromatograms of fatty acid analysis obtained from wild and farmed salmon samples and Table 2 presents their relative contents for the four salmon groups under evaluation, namely, wild from Canada and farmed from Canada, Chile and Norway. **Figure 1.** Main process pipeline. PCA: principal components analysis; t-SNE: t-distributed stochastic Figure 2 shows representative chromatograms of fatty acid analysis obtained from wild and farmed salmon samples and Table 2 presents their relative contents for the four salmon groups under evaluation, namely, wild from Canada and farmed from Canada, Chile and Norway.
(**B**) salmon samples from Canada.
*Foods* **2020**, *9*, x FOR PEER REVIEW 8 of 15
**Figure 2.** Chromatograms of fatty acid profiles obtained by GC-FID analysis of wild (**A**) and farmed **Figure 2.** Chromatograms of fatty acid profiles obtained by GC-FID analysis of wild (**A**) and farmed (**B**) salmon samples from Canada.
**Table 2.** Fatty acid composition (relative% of the identified FAME) obtained by GC-FID analysis of lipids from the wild and farmed salmon samples of different origins. Results are given as mean ± SD **Table 2.** Fatty acid composition (relative% of the identified FAME) obtained by GC-FID analysis of lipids from the wild and farmed salmon samples of different origins. Results are given as mean ± SD of the total specimens analyzed for each group.
Σ SFA 23.25 ± 1.93 d 17.18 ± 1.42 b 17.98 ± 1.13 c 13.6 ± 1.98 a Σ MUFA 32.73 ± 3.22 a 52.89 ± 1.76 c 52.02 ± 1.63 b 52.41 ± 7.55 c Σ PUFA 43.95 ± 2.77 c 29.93 ± 0.62 a 30.00 ± 1.62 a 31.99 ± 4.62 b SFA: saturated fatty acids; MUFA: monounsaturated fatty acids; PUFA: polyunsaturated fatty acids. Different letters indicate significant differences (*p* < 0.05) between groups in the statistical analysis by one-way analysis of variance (ANOVA).
n3/n6 16.75 ± 1.61 b 0.90 ± 0.24 a 0.66 ± 0.04 a 0.82 ± 0.02 a
Striking differences can be observed between wild and farmed salmons, namely in terms of the sum of MUFA and PUFA, ratio between omega-3 and omega-6 fatty acids, and also regarding several individual fatty acids. For the same amount of derivatized lipids, and when compared to wild, farmed salmon presented a significantly higher (*p* < 0.05) content of oleic and linoleic acids and lower contents of EPA, DHA and C22:1 isomers. In general, the obtained results are in good agreement with previous knowledge since farmed salmons are frequently described as having higher amounts of C18:1, C18:2 and C18:3 fatty acids, while wild are richer in long chain omega-3 PUFA as well as saturated fatty acids (SFA) [6,7]. Nevertheless, in the present study, similar contents of α-linolenic acid (C18:3n3) were found between the wild and farmed groups and only a slightly higher amount was verified in terms of SFA. The obtained data confirm that the consumption of wild salmon can be associated with greater health benefits due to their favorable ratio omega-3/omega-6 fatty acids. As discussed in previous papers, the differences observed are most probably related with differences in the diets of fish from the wild and in aquaculture conditions [6,17,21].
Compared to the results previously reported for the analysis of the same samples (as part of the EU-funded project FOODINTEGRITY) using a different technique, namely DART-MS, some quantitative differences can be pointed out. Namely, the content reported by Fiorino et al. [8] for 16:0 was higher in both farmed and wild groups, while the present GC-FID results show higher contents for 18:3, 18:1 (mainly for the farmed group) and 22:6 (mainly for the wild group). These dissimilarities can be due to the different techniques used, one based on mass spectrometry and normalized abundances, and the other relying on flame ionization detection and relative peak areas.
#### *3.2. Chemometric Analysis of the Generated Data*
#### 3.2.1. Features Selection
The importance of each feature regarding the group separation was evaluated by applying the information gain ratio criterion, as described in the materials and methods section. Table 3 presents the ranking of features obtained based on that measurement. Subsequently, the developed algorithm (Algorithm 1) was used to determine the minimum number of features required for classifying the four groups accurately. That number was found to be six, corresponding to the following features: 16:0, 18:2n6c, 20:3n3 + 20:4n6, 14:0, 18:1n9 and 22:6n3.
**Table 3.** Features sorted by applying the information gain ratio criterion.
#### 3.2.2. Data Visualization by PCA and t-SNE 3.2.2. Data Visualization by PCA and t-SNE
As a first approach, PCA was applied to the dataset as a linear and unsupervised statistical method. This method is one of the most widespread exploratory data analysis tools, providing a fast data overview by projecting each data point onto a small number of principal components, thus reducing data dimensionality, while maintaining their variation as much as possible [24]. Moreover, this approach was used previously regarding the analysis of the same salmon samples by a distinct methodology, namely DART-MS analysis [8]. Figure 3A presents the data distribution on two principal components when all the 17 data features are used. As it can be observed, PC1 and PC2 accounted for 87.8% of the total variance and showed a clear separation between the wild samples and the farmed ones, similarly to the results reported by Fiorino et al. [8]. Although it was not possible to clearly distinguish the farmed samples according to their geographical origin, mainly due to overlapping of samples from farmed Canada and Chile groups, a better separation was achieved when compared to the results of Fiorino et al. [8] using DART-MS analyses. Interestingly, in their work, five out of the six fatty acids, exhibiting the most relevant differences between wild and farmed salmons, were in common with the ones selected by Algorithm 1. Linolenic acid (C18:3) was an exception because in the present work it ranked as the 15th position with a low information gain ratio value, thus not being relevant to distinguish the four groups using the GC-FID fatty acid profiles. Subsequently, PCA was also applied to the whole dataset, but using only the selected best six features (Figure 3B), evidencing results similar to the ones obtained with all the 17 features. As a first approach, PCA was applied to the dataset as a linear and unsupervised statistical method. This method is one of the most widespread exploratory data analysis tools, providing a fast data overview by projecting each data point onto a small number of principal components, thus reducing data dimensionality, while maintaining their variation as much as possible [24]. Moreover, this approach was used previously regarding the analysis of the same salmon samples by a distinct methodology, namely DART-MS analysis [8]. Figure 3A presents the data distribution on two principal components when all the 17 data features are used. As it can be observed, PC1 and PC2 accounted for 87.8% of the total variance and showed a clear separation between the wild samples and the farmed ones, similarly to the results reported by Fiorino et al. [8]. Although it was not possible to clearly distinguish the farmed samples according to their geographical origin, mainly due to overlapping of samples from farmed Canada and Chile groups, a better separation was achieved when compared to the results of Fiorino et al. [8] using DART-MS analyses. Interestingly, in their work, five out of the six fatty acids, exhibiting the most relevant differences between wild and farmed salmons, were in common with the ones selected by Algorithm 1. Linolenic acid (C18:3) was an exception because in the present work it ranked as the 15th position with a low information gain ratio value, thus not being relevant to distinguish the four groups using the GC-FID fatty acid profiles. Subsequently, PCA was also applied to the whole dataset, but using only the selected best six features (Figure 3B), evidencing results similar to the ones obtained with all the 17 features.
*Foods* **2020**, *9*, x FOR PEER REVIEW 10 of 15
24:1n9 0.505 22:5n3 0.464 18:1n7 0.461 18:4n3 0.446 20:5n3 0.423 16:1 0.402 18:3n3 0.378 20:1n9 0.366 18:0 0.353
**Figure 3.** Scatterplot obtained for the first two principal components after applying PCA to the whole dataset using (**A**): all the 17 features, (**B**): the 6 best features (16:0, 18:2n6c, 20:3n3 + 20:4n6, 14:0, 18:1n9 and 22:6n3); 0—Norway farmed, 1—Chile farmed, 2—Canada farmed, 3—Canada wild. **Figure 3.** Scatterplot obtained for the first two principal components after applying PCA to the whole dataset using (**A**): all the 17 features, (**B**): the 6 best features (16:0, 18:2n6c, 20:3n3 + 20:4n6, 14:0, 18:1n9 and 22:6n3); 0—Norway farmed, 1—Chile farmed, 2—Canada farmed, 3—Canada wild.
The interpretation of Figure 3A,B allows drawing two conclusions: (1) most of the data are strongly explained by the first principal component regardless of the number of used features, namely all the 17 or only the best six, which confirms that most of the features are not important for the correct classification; (2) some samples of Chile farmed and Canada farmed groups are not linearly separable with data projected on a 2D subspace, thus suggesting the need for non-linear classification models. Therefore, t-SNE was applied to the dataset, first using all the 17 features, and The interpretation of Figure 3A,B allows drawing two conclusions: (1) most of the data are strongly explained by the first principal component regardless of the number of used features, namely all the 17 or only the best six, which confirms that most of the features are not important for the correct classification; (2) some samples of Chile farmed and Canada farmed groups are not linearly separable with data projected on a 2D subspace, thus suggesting the need for non-linear classification models. Therefore, t-SNE was applied to the dataset, first using all the 17 features, and then only the selected best six (Figure 4A,B). This method allows the projection of the original dimension on two dimensions without losing the non-linear relations presented at the high dimensional space. Thus, it is a suitable tool to perceive the separability of groups at the original dimension. As shown in Figure 4, there is no data superposition and, in general, the groups are well separated according to this method. This information suggests a good data separability when the classification models can handle non-linearities in a high dimension space.
then only the selected best six (Figure 4A,B). This method allows the projection of the original dimension on two dimensions without losing the non-linear relations presented at the high dimensional space. Thus, it is a suitable tool to perceive the separability of groups at the original dimension. As shown in Figure 4, there is no data superposition and, in general, the groups are well
classification models can handle non-linearities in a high dimension space.
**Figure 4.** Scatterplot obtained after applying t-SNE to the whole dataset using (**A**): all the 17 features (**B**): only the 6 best features (16:0, 18:2n6c, 20:3n3 + 20:4n6, 14:0, 18:1n9 and 22:6n3); 0—Norway, Figure 1. Chile farmed, 2—Canada farmed, 3—Canada wild. **Figure 4.** Scatterplot obtained after applying t-SNE to the whole dataset using (**A**): all the 17 features (**B**): only the 6 best features (16:0, 18:2n6c, 20:3n3 + 20:4n6, 14:0, 18:1n9 and 22:6n3); 0—Norway, Figure 1. Chile farmed, 2—Canada farmed, 3—Canada wild.
A good separability among groups was also observed when the number of employed input features was only six (Figure 4B). This suggests that, in the high dimension original space, the separability is achieved based on only a few features. Normally, this is an advantage for subsequently used classifiers because it promotes generalization and tends to avoid overfitting, thus strongly suggesting that new samples will be properly classified based on such classifiers. A good separability among groups was also observed when the number of employed input features was only six (Figure 4B). This suggests that, in the high dimension original space, the separability is achieved based on only a few features. Normally, this is an advantage for subsequently used classifiers because it promotes generalization and tends to avoid overfitting, thus strongly suggesting that new samples will be properly classified based on such classifiers.
#### 3.2.3. Machine Learning Classifiers 3.2.3. Machine Learning Classifiers
In this work, a total of seven different classifiers were tested considering performance (classification accuracy) and required computational effort (evaluated as test time). Similarly, as was done for PCA and t-SNE, each classifier was first assayed using all the 17 features as inputs to the classifiers. The obtained performance is shown in Table 4, evidencing that ANN, random forest, SVM, naïve Bayes and kNN were the best models as they showed a maximum performance, allowing classifying, without error, for all of the test dataset. Nevertheless, they are closely followed by the remaining classifiers, with decision tree being the one that performed worst. In terms of performance time (test time), among the classifiers that allowed 100% accuracy (CA), naïve Bayes was the best one. This can be explained by two factors: first, one must consider that in this case the number of features exceeds the needs, thus, according to Occam's razor principle, the simpler model can achieve a good performance; second, as the model is very simple to implement, the number of required computational calculation steps is small, thus corresponding to a shorter time of performance. In this work, a total of seven different classifiers were tested considering performance (classification accuracy) and required computational effort (evaluated as test time). Similarly, as was done for PCA and t-SNE, each classifier was first assayed using all the 17 features as inputs to the classifiers. The obtained performance is shown in Table 4, evidencing that ANN, random forest, SVM, naïve Bayes and kNN were the best models as they showed a maximum performance, allowing classifying, without error, for all of the test dataset. Nevertheless, they are closely followed by the remaining classifiers, with decision tree being the one that performed worst. In terms of performance time (test time), among the classifiers that allowed 100% accuracy (CA), naïve Bayes was the best one. This can be explained by two factors: first, one must consider that in this case the number of features exceeds the needs, thus, according to Occam's razor principle, the simpler model can achieve a good performance; second, as the model is very simple to implement, the number of required computational calculation steps is small, thus corresponding to a shorter time of performance.
**Table 4.** Classifiers performance, in the test dataset, using all the 17 input features. **Table 4.** Classifiers performance, in the test dataset, using all the 17 input features.
Decision Tree 0.001 0.908 0.908 CA: accuracy; F1 score: harmonic mean of the precision and recall.
CA: accuracy; F1 score: harmonic mean of the precision and recall.
Next, the performance of classifiers was assayed with only the six best features as their inputs. As can be observed in Table 5, in this case, the ANN, SVM and kNN classifiers allowed 100% correct classification, as measured by accuracy and F1 indicators. It is possible that the elements that were not correctly classified by the remaining models do not have statistical significance to change the
parameters present on the learning mechanism to the rest of classifiers. Among the best classifiers the one that presented the best computational performance was the SVM.
**Table 5.** Classifier performance, in the test dataset, using the selected best 6 input features.
Overall, the remaining classifiers were very close to the performance of ANN, SVM and kNN, despite the reduced number of features used. For this reason, it was decided to further observe the classification performance when the features are remapped by the t-SNE method as inputs for the classifiers, keeping the same parametrization for all models, as in the previous scheme. By applying Algorithm 1 and extending the processing pipeline with the t-SNE block, namely by placing that block between the features used and the classifiers, it was possible to conclude that the classification can still be performed successfully by relying on only three features, namely 16:0, 18:2n6c and the sum of 20:3n3 + 20:4n6. The obtained results are presented in Table 6, evidencing 100% accuracy of sample classification using the kNN, with only three compounds being required in this model. In this scenario, the decision tree classifier showed the worst performance, being the only one presenting an accuracy < 95%.
**Table 6.** Classifiers performance, in the test dataset, using the selected best 3 input features mapped by t-SNE.
Figure 5 shows the confusion matrices, evidencing sample classification, for the best (kNN) and worst (decision tree) models, using only the three best features, as processed by t-SNE. While the confusion matrix for the kNN model presents all samples as being correctly classified, the confusion matrix for the decision tree evidences some errors because six samples from group zero (Norway farmed) were misclassified as being from group one (Chile farmed). This shows that the inductive learning mechanism present in the decision tree was not able to classify those samples correctly, as probably happens with the remaining classifiers, except for kNN that is not based on inductive learning.
**Figure 5.** Confusion matrix (showing proportion of actual) for the decision tree model (left) and confusion matrix for the kNN model (right), both processing only three features. 0—Norway farmed, 1—Chile farmed, 2—Canada farmed, 3—Canada wild. **Figure 5.** Confusion matrix (showing proportion of actual) for the decision tree model (**left**) and confusion matrix for the kNN model (**right**), both processing only three features. 0—Norway farmed, 1—Chile farmed, 2—Canada farmed, 3—Canada wild.
#### **4. Conclusions 4. Conclusions**
In general, the four evaluated groups of salmon (wild from Canada and farmed from Canada, Chile and Norway) showed different fatty acid profiles, with wild specimens presenting significantly higher contents of health beneficial omega-3 fatty acids, in particular DHA and EPA, while farmed salmon presented significantly higher (*p* < 0.05) amounts of oleic and linoleic acids. Among the three groups of farmed salmon with different geographical origins, specimens from Chile and Canada were more similar, with the ones from Norway being more distinct mainly due to their lower levels of SFA and higher levels of α-linolenic acid. The differences among farmed groups are most probably related to different types of feed used in each farm. However, information about relevant factors such as farming diet and conditions, which are known to affect the lipidic composition of fish, was not available. In this work, we demonstrated the possibility of discriminating between wild and farmed salmons, as well as differentiating the origin within farmed ones, based on the use of machine learning models applied to fatty acid composition obtained by GC-FID. Thus, compared to a previous approach reported for the same samples, namely the use of PCA applied to normalized intensities of the most abundant signals generated by DART-HRMS analysis of the lipid extracts, this method showed a higher discrimination power. Moreover, this method proved to be simple and it only requires the use of affordable equipment, commonly found in most laboratories. Nevertheless, this approach has the disadvantage of requiring a longer analysis time compared to DART-HRMS. The developed algorithm combined with the information gain ratio criterion allowed us to establish the number of optimal features, so the classification tasks can still attain a very good performance. The feature reduction offers a computational speedup during the classification process. Among the seven tested machine learning models, the best results were obtained with the k-nearest neighbors (kNN) classifier, allowing for the correct classification of all tested samples. Moreover, it was shown that using t-SNE in the processing pipeline boosts the reduction in features, while still maintaining 100% accuracy in data classification. The performance difference between the test dataset and the leaveone-sample-out cross-validation was residual, meaning a good generalization figure. In general, the four evaluated groups of salmon (wild from Canada and farmed from Canada, Chile and Norway) showed different fatty acid profiles, with wild specimens presenting significantly higher contents of health beneficial omega-3 fatty acids, in particular DHA and EPA, while farmed salmon presented significantly higher (*p* < 0.05) amounts of oleic and linoleic acids. Among the three groups of farmed salmon with different geographical origins, specimens from Chile and Canada were more similar, with the ones from Norway being more distinct mainly due to their lower levels of SFA and higher levels of α-linolenic acid. The differences among farmed groups are most probably related to different types of feed used in each farm. However, information about relevant factors such as farming diet and conditions, which are known to affect the lipidic composition of fish, was not available. In this work, we demonstrated the possibility of discriminating between wild and farmed salmons, as well as differentiating the origin within farmed ones, based on the use of machine learning models applied to fatty acid composition obtained by GC-FID. Thus, compared to a previous approach reported for the same samples, namely the use of PCA applied to normalized intensities of the most abundant signals generated by DART-HRMS analysis of the lipid extracts, this method showed a higher discrimination power. Moreover, this method proved to be simple and it only requires the use of affordable equipment, commonly found in most laboratories. Nevertheless, this approach has the disadvantage of requiring a longer analysis time compared to DART-HRMS. The developed algorithm combined with the information gain ratio criterion allowed us to establish the number of optimal features, so the classification tasks can still attain a very good performance. The feature reduction offers a computational speedup during the classification process. Among the seven tested machine learning models, the best results were obtained with the k-nearest neighbors (kNN) classifier, allowing for the correct classification of all tested samples. Moreover, it was shown that using t-SNE in the processing pipeline boosts the reduction in features, while still maintaining 100% accuracy in data classification. The performance difference between the test dataset and the leave-one-sample-out cross-validation was residual, meaning a good generalization figure.
**Author Contributions:** Conceptualization, J.S.A., I.M., P.J.R. and G.I.; methodology, L.G., G.I. and P.J.R.; chemical analyses, L.G. and M.A.N.; writing—original draft preparation, P.J.R., G. I. and J.S.A.; writing—review and editing, I.M. and J.S.A.; supervision, I.M., M.B.P.P.O. and J.S.A.; project administration, M.A., I.M., J.S.A. **Author Contributions:** Conceptualization, J.S.A., I.M., P.J.R. and G.I.; methodology, L.G., G.I. and P.J.R.; chemical analyses, L.G. and M.A.N.; writing—original draft preparation, P.J.R., G.I. and J.S.A.; writing—review and editing, I.M. and J.S.A.; supervision, I.M., M.B.P.P.O. and J.S.A.; project administration, M.A., I.M., J.S.A. All authors have read and agreed to the published version of the manuscript.
**Funding:** This work was supported by the European project FOODINTEGRITY (FP7-KBBE-2013-single-stage, under grant agreement No 613688) and FCT (Fundação para a Ciência e Tecnologia, Portugal) under the Partnership Agreements UIDB 50006/2020, UIDB 00690/2020 (CIMO) and UIDB/5757/2020 (CeDRI). L. Grazina and M.A. Nunes acknowledge the FCT grant SFRH/BD/132462/2017 and SFRH/BD/130131/2017 financed by **Funding:** This work was supported by the European project FOODINTEGRITY (FP7-KBBE-2013-single-stage, under grant agreement No 613688) and FCT (Fundação para a Ciência e Tecnologia, Portugal) under the Partnership Agreements UIDB 50006/2020, UIDB 00690/2020 (CIMO) and UIDB/5757/2020 (CeDRI). L. Grazina and M.A. Nunes acknowledge the FCT grant SFRH/BD/132462/2017 and SFRH/BD/130131/2017 financed by POPH-QREN (subsidised by FSE and MCTES).
POPH-QREN (subsidised by FSE and MCTES). **Acknowledgments:** The authors are thankful to Emiliano De Dominicis from Mérieux-Nutrisciences for **Acknowledgments:** The authors are thankful to Emiliano De Dominicis from Mérieux-Nutrisciences for providing the salmon samples analysed in the present project.
providing the salmon samples analysed in the present project. **Conflicts of Interest:** The authors declare no conflict of interest. **Conflicts of Interest:** The authors declare no conflict of interest.
## **References**
**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.
© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
## *Article* **A Real-Time PCR Method for the Authentication of Common Cuttlefish (***Sepia o*ffi*cinalis***) in Food Products**
### **Amaya Velasco \*, Graciela Ramilo-Fernández and Carmen G. Sotelo**
Instituto de Investigaciones Marinas (IIM-CSIC), Eduardo Cabello 6, 36208 Vigo (Pontevedra), Spain; [email protected] (G.R.-F.); [email protected] (C.G.S.)
**\*** Correspondence: [email protected]; Tel.: +34-986-23-19-30
Received: 6 February 2020; Accepted: 28 February 2020; Published: 4 March 2020
**Abstract:** Cephalopods are very relevant food resources. The common cuttlefish (*Sepia o*ffi*cinalis*) is highly appreciated by consumers and there is a lack of rapid methods for its authentication in food products. We introduce a new minor groove binding (MGB) TaqMan real-time PCR (Polymerase Chain Reaction) method for the authentication of *S. o*ffi*cinalis* in food products to amplify a 122 base pairs (bp) fragment of the mitochondrial COI (Cytochrome Oxidase I) region. Reference and commercial samples of *S. o*ffi*cinalis* showed a threshold cycle (Ct) mean of 14.40, while the rest of the species examined did not amplify, or showed a significantly different Ct (*p* < 0.001). The calculated efficiency of the system was 101%, and the minimum DNA quantity detected was 10−<sup>4</sup> ng. No cross-reactivity was detected with any other species, thus, the designed method differentiates *S. o*ffi*cinalis* from other species of the genus *Sepia* and other cephalopod species and works for fresh, frozen, grilled, cooked and canned samples of *Sepia* spp. The method has proved to be reliable and rapid, and it may prove to be a useful tool for the control of fraud in cuttlefish products.
**Keywords:** Sepia; common cuttlefish; *Sepia o*ffi*cinalis*; real-time PCR (Polymerase Chain Reaction); species identification; food authentication; COI (Cytochrome Oxidase I)
### **1. Introduction**
Cephalopods are a very diverse group of mollusks and include 28 families and more than 600 species, many of which are commercially important. As a sign of their relevance, captures of cephalopods in 2017 reached 3,772,565 t, with an estimated value of almost 8000 million dollars [1].
The common cuttlefish (*S. o*ffi*cinalis*) is highly appreciated by consumers around the world, and it is traded with different presentations particularly in Japan, the Republic of Korea, Italy and Spain. In the last decade, the world catches attributed to this species have registered numbers between 20,000 and 30,000 tons every year [2]. It is the species of cuttlefish with the highest commercial value.
European regulations regarding the labeling of fishery products [3,4] establish that these products must show the information about the species, with the commercial and/or scientific name depending on the type of product. Illicit substitution of one species for another may constitute economic fraud and/or misbranding violations. Furthermore, species substitution may cause potential food safety hazards to be overlooked by processors or end-users [5]. Species substitution is relatively frequent in seafood products [6], and particularly in products containing cephalopods, where several cases of species substitution have been reported [7,8].
Species belonging to the genus *Sepia* can look very similar to a non-trained consumer, especially when they are processed for the market (e.g., peeled, canned), making the visual differentiation almost impossible and increasing the possibilities of fraud. Thus, the reported cases of mislabeling in products
containing *Sepia* spp. have usually been substitutions between species belonging to the same genus [9,10]. These cases can be attributed to economic fraud (e.g., substitutions between species with different commercial value) or unintentional substitutions, which can be due to similar geographic distribution of species (e.g., *S. o*ffi*cinalis*/*S. orbignyanya*/*S. elegans*) and/or similar morphological characteristics (e.g., juveniles of *S. o*ffi*cinalis*/*S. elegans*) which can lead to misidentification at any level of the value chain (fisheries, processors and consumers). In order to control these substitutions, a variety of genetic methods have been published for the identification of several cephalopod species. The majority of these are labor-intensive and time-consuming, such as forensically informative nucleotide sequencing (FINS), barcoding [10–14] and RFLP [8,15]. Some rapid DNA-based methods have also been published for the authentication of some cephalopod species [7,16–18], but to date, there is not any rapid technique available for the genetic identification of *S. o*ffi*cinalis*.
This work presents a rapid and reliable method for the authentication of *S. o*ffi*cinalis* in different food matrices, including processed products. Therefore, it can be a useful tool for control authorities at different levels of the value chain.
#### **2. Materials and Methods**
#### *2.1. Sampling and DNA Extraction*
In this work, 14 samples of *S. o*ffi*cinalis* from different locations of Spanish and Portuguese waters were used as a reference. Also, 29 individuals from 20 other cephalopod species of 11 genera from the Instituto de Investigaciones Marinas (IIM-CSIC) own tissue collection were included for the specificity assay (Table 1). All reference individuals had a known origin and were identified visually prior to the FINS identification. Additionally, 16 commercial samples were collected from supermarkets and restaurants in Galicia region (Spain) for the application to commercial products (Table 2). All tissue samples were stored at −20 ◦C until analysis.
**Table 1.** Reference samples used in this study and threshold cycle (Ct) results.
**Table 1.** *Cont.*
FAO: Food and Agriculture Organization. SD: Standard Deviation.
**Table 2.** Commercial samples used for validation. The mislabeled samples are highlighted in red.
FINS: Forensically Informative Nucleotide Sequencing.
A portion of 0.3 g of muscle tissue from each sample was digested at 56 ◦C in a thermo shaker with 860 µL of lysis buffer (1% Sodium Dodecyl Sulfate (SDS), 150 mM NaCl, 2 mM Ethylenediaminetetraacetic acid (EDTA) and 10 mM Tris-HCl at pH 8), 100 µL of guanidinium thiocyanate 5 M and 40 µL of proteinase K (20 mg/mL). After 3 h, 40 µL of extra proteinase K was added and left overnight. DNA was isolated with the Wizard DNA Clean-up System kit (Promega, Madison, WI, USA) following the manufacturer's protocol. Double-stranded DNA obtained was quantified with Qubit dsDNA BR Assay Kit (Life Technologies, Carlsbad, CA, USA) and Qubit 3.0 fluorometer (Invitrogen, Carlsbad, CA, USA). Purified DNA was stored at −20 ◦C until further analysis.
#### *2.2. FINS Identification of Samples*
Reference and commercial samples were authenticated by FINS (forensically informative nucleotide sequencing) in order to test the reliability of the method developed. PCR reactions were carried out in a Verity 96 wells Thermal cycler (Applied Biosystems, Foster City, CA, USA) with Illustra PuReTaq Ready-To-Go PCR Beads (GE Healthcare, Chicago, IL, USA), 1 µL of each primer (10 µM) and 100 ng of template DNA in a final volume of 25 µL. Primers designed by Folmer [19] LCO1490-50GGTCAACAAATCATAAAGATATTGG30 and HCO2198-50TAAACTTCAGGGTGACCAA AAAATCA30 were used to amplify a 750 base pairs (bp) fragment of the mitochondrial COI region, with the following thermal protocol: a preheating step of 3 min at 95 ◦C, followed by 35 cycles of 1 min at 95 ◦C, 1 min at 40 ◦C and 1.5 min at 72 ◦C, with a final extension step at 72 ◦C for 7 min. When amplification of COI fragment failed, the 16SVAR primers described by Chapela [11] 16SVAR-F-5 0CAAATTACGCTGTTATCCCTATGG30 and 16SVAR-R- 50GACGAGAAGACCCTAATGAGCTTT30 were used to amplify a 210 bp fragment of the mitochondrial 16S rDNA, with the thermal protocol as follows: a preheating step of 3 min at 95 ◦C, followed by 35 cycles of 40 s at 94 ◦C, 40 s at 50 ◦C and 40 s
at 72 ◦C, with a final extension step at 72 ◦C for 7 min. Negative and positive controls were included in all PCR sets.
Primers designed in this study for the minor groove binding (MGB)-TaqMan assay were also used for FINS identification in 3 cases of processed commercial samples of *Sepia* spp. (cooked and canned), where both COI and 16S sets of primers failed to amplify, with the following thermal protocol: a preheating step of 3 min at 95 ◦C, followed by 35 cycles of 40 s at 95 ◦C, 40 s at 40 ◦C and 40 s at 72 ◦C with a final extension step at 72 ◦C for 7 min. PCR amplicons were visualized on a 2% agarose gel, using UV transillumination (BioRad, Hercules, CA, USA).
PCR products were purified with Illustra ExoProStar (GE Healthcare, Chicago, IL, USA) and sequencing reactions were performed with BigDye Terminator 1.1 (Applied Biosystems, Foster City, CA, USA), following the manufacturer's instructions. The automatic sequencing was carried out in an ABI PRISM 3130 (Applied Biosystems, Foster City, CA, USA). After automatic sequencing, F and R files were edited with Chromas and aligned with Bioedit [20] to obtain the complete sequence of the fragment. Bioedit software was also used to align the resulting sequence with reference ones from the NCBI and the IIM-CSIC sequence database, which consists of more than 2000 sequences from fish and mollusks specimens that have been collected during 30 years; most of these specimens were morphologically identified and also genetically authenticated. This alignment was imported with MEGA [21] for phylogenetic analysis. The phylogenetic model used for constructing the neighbor-joining tree was Tamura–Nei, with 1000 bootstrap replicates. The results were also authenticated with BLAST [22]. The multiple alignments and the BLAST tool were also used to check the quality and coverage of the resulting sequences.
The COI sequences obtained for reference and commercial samples of this study were uploaded to Genbank [23] (accession numbers: MN977128 to MN977135, MN977138, MN977143, MN977144, MN977146, MN977147, MN977149, MN977152, MN977154 to MN977156, MN977158, MN977159, MN977161 to MN977171, MN977173 to MN977177, MN977179 to MN977191).
#### *2.3. RT-PCR Design*
In order to find a suitable fragment to design a short and specific system, a large number of nuclear and mitochondrial cephalopod sequences from public and IIM-CSIC databases were aligned and analyzed. A fragment of the COI region was suitable for the design of an MGB-Taq-Man Primers and Probe set, complying with the requirements of showing low intraspecific variability and high interspecific variability and allowing the amplification of a short fragment (122 bp, primers included). The sequences of primers (F and R) and Probe (P) are the following (see Figure 1):
SOFI\_F: 50CTTCTCCTTACATTTAGCWGGRGTCT30 SOFI\_R: FAM-50TACCGAYCAAGCAAATAAAGGTAGG30 -MGB SOFI\_P: 50AGCGATTAACTTCATCA30
#### *2.4. Real-Time PCR Conditions and Data Treatment*
Concentrations of 50, 300 and 900 nM of each primer and 25, 50, 75, 100, 125, 150, 175, 200 and 225 nM of the probe were tested in order to select the optimal reaction conditions. The combination that gave the lowest threshold cycle (Ct) value and the highest final fluorescence was selected for the subsequent assays. The selected concentrations were 300 nM of SOFI\_F primer, 900 nM of SOFI\_R primer and 150 nM of SOFI\_P probe.
Thus, each 20 µL reaction contained 10 µL of TaqMan Fast Universal Master Mix (2X), No AmpErase UNG (Applied Biosystems, Foster City, CA, USA), 1 µL of Primer SOFI\_F (6 µM), 1 µL of Primer SOFI\_R (18 µM), 1 µL of Probe SOFI\_P (3 µM) and 100 ng of template DNA. Reactions were amplified in a 7500 fast real-time PCR System (Applied Biosystems, Foster City, CA, USA), with the fast ramp speed protocol: 95 ◦C for 20 s, followed by 40 cycles of 95 ◦C for 3 s and 60 ◦C for 30 s. Samples were analyzed in triplicate, and Ct mean and standard deviation of each individual were registered.
*Foods* **2020**, *9*, x FOR PEER REVIEW 6 of 10
**Figure 1.** Multiple sequence alignment of the mitochondrial COI (Cytochrome Oxidase I) fragment, showing the position of the primers and probe designed. **Figure 1.** Multiple sequence alignment of the mitochondrial COI (Cytochrome Oxidase I) fragment, showing the position of the primers and probe designed.
#### *2.4. Real-Time PCR Conditions and Data Treatment* **3. Results**
#### Concentrations of 50, 300 and 900 nM of each primer and 25, 50, 75, 100, 125, 150, 175, 200 and *3.1. E*ffi*ciency and Detection Limit*
225 nM of the probe were tested in order to select the optimal reaction conditions. The combination that gave the lowest threshold cycle (Ct) value and the highest final fluorescence was selected for the subsequent assays. The selected concentrations were 300 nM of SOFI\_F primer, 900 nM of SOFI\_R primer and 150 nM of SOFI\_P probe. Thus, each 20 µL reaction contained 10 µL of TaqMan Fast Universal Master Mix (2X), No AmpErase UNG (Applied Biosystems, Foster City, CA, USA), 1 µL of Primer SOFI\_F (6 µM), 1 µL of Primer SOFI\_R (18 µM), 1 µL of Probe SOFI\_P (3 µM) and 100 ng of template DNA. Reactions were Different quantities of template DNA of *S. o*ffi*cinalis*, from 10−<sup>5</sup> ng to 100 ng were used for the efficiency assay. Over this range of dilutions, the response was linear with a slope of −3.13, an *R* <sup>2</sup> of 0.999 and an efficiency of 101%, following the equation: E <sup>=</sup> <sup>10</sup>−1/<sup>b</sup> <sup>−</sup> 1 [24]. The acceptable efficiency values range from 90% to 110%, therefore, 101% can be considered ideally optimal. The minimum quantity of DNA detected was 10−<sup>4</sup> ng. The automatic threshold generated in this assay was 0.02, the value used in the subsequent analyses.
#### amplified in a 7500 fast real-time PCR System (Applied Biosystems, Foster City, CA, USA), with the *3.2. Inclusivity and Specificity*
tested.
*3.3. Application to Commercial Products*
fast ramp speed protocol: 95 °C for 20 s, followed by 40 cycles of 95 °C for 3 s and 60 °C for 30 s. Samples were analyzed in triplicate, and Ct mean and standard deviation of each individual were registered. A total of 14 samples of *S. o*ffi*cinalis* from different locations and dates of capture were tested (Table 1), obtaining Ct data between 12.59 and 16.16, with a Ct mean of 14.04 (Figure 2A). *Foods* **2020**, *9*, x FOR PEER REVIEW 7 of 10
*3.2. Inclusivity and Specificity* **Figure 2.** Amplification plots of the 10X dilution series of *Sepia officinalis* DNA (**A**): logarithmic, (**B**): linear. **Figure 2.** Amplification plots of the 10X dilution series of *Sepia o*ffi*cinalis* DNA (**A**): logarithmic, (**B**): linear.
A total of 14 samples of *S. officinalis* from different locations and dates of capture were tested (Table 1), obtaining Ct data between 12.59 and 16.16, with a Ct mean of 14.04 (Figure 2A). In the other 19 species tested (Table 1), none of them presented any fluorescence signal with the exception of one specimen of *Loligo vulgaris*, which showed a late amplification signal (Figure 3B). In view of these results, another specificity assay was carried out with seven additional individuals of *L. vulgaris*, obtaining a Ct mean of 34.0, a result that is significantly different from the Ct of *S. officinalis* when a mean comparison test (one way ANOVA) was run (*p* < 0.001). In the other 19 species tested (Table 1), none of them presented any fluorescence signal with the exception of one specimen of *Loligo vulgaris*, which showed a late amplification signal (Figure 3B). In view of these results, another specificity assay was carried out with seven additional individuals of *L. vulgaris*, obtaining a Ct mean of 34.0, a result that is significantly different from the Ct of *S. o*ffi*cinalis* when a mean comparison test (one way ANOVA) was run (*p* < 0.001).
**Figure 3.** (**A**) Inclusivity test: amplification pattern of reference samples of *Sepia officinalis*. (**B**) Specificity test: amplification pattern of reference samples of *Sepia officinalis* and the rest of the species
According to the Spanish regulations for the labeling of fresh, frozen and refrigerated fishery products, the commercial name "Sepia", "Choco" or "Jibia" is only accepted for products containing *S. officinalis*, while the commercial name "Sepias" can be used for all species of the genus *Sepia* [25]. In the same way, the commercial name "Jibia" or "Sepia" can be only applied to canned products containing the species *S. officinalis* [26]. Therefore, the system was also tested with 16 commercial samples labeled as "Sepia", "Choco" or "Sepias", from supermarkets and restaurants of Galicia (Spain), with different degrees of processing such as thawed, frozen, grilled, cooked and canned.
Following the above-mentioned criteria, the FINS identification results of this study revealed four cases of mislabeling regarding species (Table 2), all being substitutions between different species of the genus *Sepia*, constituting a mislabeling rate of 25%. The substitute species found were *Sepia pharaonis*, *Sepia aculeata*, *Sepia bertheloti* and a non-identified species. In four cases, it was not possible to reach the species level with the FINS identification, due to the lack of reference sequences in public databases, but authors could determine that these samples did not belong to *S. officinalis* species by analyzing the results of the neighbor-joining tree and the BLAST tool. The MGB TaqMan real-time PCR system worked in fresh and processed samples of *S. officinalis*, and the method was able to differentiate between products containing *S. officinalis* (Ct mean 15.23) and products containing other species of the Sepiidae family (Ct mean 33.82), with statistical significance (*p* < 0.001). The type of processing did not affect the Ct values, and a good differentiation was obtained both in fresh and
frozen products as well as in highly processed samples, such as canned.
linear.
when a mean comparison test (one way ANOVA) was run (*p* < 0.001).
**Figure 2.** Amplification plots of the 10X dilution series of *Sepia officinalis* DNA (**A**): logarithmic, (**B**):
In the other 19 species tested (Table 1), none of them presented any fluorescence signal with the exception of one specimen of *Loligo vulgaris*, which showed a late amplification signal (Figure 3B). In view of these results, another specificity assay was carried out with seven additional individuals of
**Figure 3.** (**A**) Inclusivity test: amplification pattern of reference samples of *Sepia officinalis*. (**B**) Specificity test: amplification pattern of reference samples of *Sepia officinalis* and the rest of the species **Figure 3.** (**A**) Inclusivity test: amplification pattern of reference samples of *Sepia o*ffi*cinalis*. (**B**) Specificity test: amplification pattern of reference samples of *Sepia o*ffi*cinalis* and the rest of the species tested.
#### tested. *3.3. Application to Commercial Products*
*3.3. Application to Commercial Products* According to the Spanish regulations for the labeling of fresh, frozen and refrigerated fishery products, the commercial name "Sepia", "Choco" or "Jibia" is only accepted for products containing *S. officinalis*, while the commercial name "Sepias" can be used for all species of the genus *Sepia* [25]. In the same way, the commercial name "Jibia" or "Sepia" can be only applied to canned products containing the species *S. officinalis* [26]. Therefore, the system was also tested with 16 commercial samples labeled as "Sepia", "Choco" or "Sepias", from supermarkets and restaurants of Galicia According to the Spanish regulations for the labeling of fresh, frozen and refrigerated fishery products, the commercial name "Sepia", "Choco" or "Jibia" is only accepted for products containing *S. o*ffi*cinalis*, while the commercial name "Sepias" can be used for all species of the genus *Sepia* [25]. In the same way, the commercial name "Jibia" or "Sepia" can be only applied to canned products containing the species *S. o*ffi*cinalis* [26]. Therefore, the system was also tested with 16 commercial samples labeled as "Sepia", "Choco" or "Sepias", from supermarkets and restaurants of Galicia (Spain), with different degrees of processing such as thawed, frozen, grilled, cooked and canned.
(Spain), with different degrees of processing such as thawed, frozen, grilled, cooked and canned. Following the above-mentioned criteria, the FINS identification results of this study revealed four cases of mislabeling regarding species (Table 2), all being substitutions between different species of the genus *Sepia*, constituting a mislabeling rate of 25%. The substitute species found were *Sepia pharaonis*, *Sepia aculeata*, *Sepia bertheloti* and a non-identified species. In four cases, it was not possible to reach the species level with the FINS identification, due to the lack of reference sequences in public databases, but authors could determine that these samples did not belong to *S. officinalis* species by analyzing the results of the neighbor-joining tree and the BLAST tool. The MGB TaqMan real-time PCR system worked in fresh and processed samples of *S. officinalis*, and the method was able to differentiate between products containing *S. officinalis* (Ct mean 15.23) and products containing other species of the Sepiidae family (Ct mean 33.82), with statistical significance (*p* < 0.001). The type of processing did not affect the Ct values, and a good differentiation was obtained both in fresh and Following the above-mentioned criteria, the FINS identification results of this study revealed four cases of mislabeling regarding species (Table 2), all being substitutions between different species of the genus *Sepia*, constituting a mislabeling rate of 25%. The substitute species found were *Sepia pharaonis*, *Sepia aculeata*, *Sepia bertheloti* and a non-identified species. In four cases, it was not possible to reach the species level with the FINS identification, due to the lack of reference sequences in public databases, but authors could determine that these samples did not belong to *S. o*ffi*cinalis* species by analyzing the results of the neighbor-joining tree and the BLAST tool. The MGB TaqMan real-time PCR system worked in fresh and processed samples of *S. o*ffi*cinalis*, and the method was able to differentiate between products containing *S. o*ffi*cinalis* (Ct mean 15.23) and products containing other species of the Sepiidae family (Ct mean 33.82), with statistical significance (*p* < 0.001). The type of processing did not affect the Ct values, and a good differentiation was obtained both in fresh and frozen products as well as in highly processed samples, such as canned.
frozen products as well as in highly processed samples, such as canned. The Ct results obtained for both reference and commercial samples containing *S. o*ffi*cinalis* ranged from 12.59 to 17.88, with a Ct mean of 14.40, while the rest of species remained undetected or showed late amplification, with Ct values of 23.62 and higher and a mean Ct of 33.40 and this Ct mean resulted significantly different from the Ct of samples containing *S. o*ffi*cinalis* (*p* < 0.001).
#### **4. Discussion**
Results confirm TaqMan real-time PCR technique as a powerful tool for species authentication, due to its characteristics of specificity, increased with Minor Groove Binding technology (MGB probes) [27], and its sensitivity, allowing the detection of very low quantities of target DNA. Real-time PCR also allows the detection and quantification of target DNA in one step, eliminating post-PCR steps and saving labor time. The method described in this work includes these characteristics of specificity, sensitivity and fastness, since the real-time PCR analysis takes around 40 min, which means that, depending on the tissue digestion protocol, the complete analysis from the tissue sample can be carried out in 3–4 h. This feature and the reduced equipment needed, opens the possibility of the optimization of the method for analyses on-site at the different levels of the value chain, including the point of sale. The cost of the analysis (less than 5 euros per sample) is also much lower than sequencing-based methods, which makes it affordable for low-resources control units.
*S. o*ffi*cinalis* is marketed under several types of processing, including those that eliminate the characteristics for visual identification, such as peeling, cutting, cooking and canning. This makes these products vulnerable to species substitution, intentional or not. Results also show that the design of the primer set allows the amplification and authentication of the species even in samples where processing may lead to DNA degradation and/or fragmentation, such as canning. The lack of rapid methods for this task makes the control of this market laborious and costly, and this technique emerges as the only available alternative at the moment.
The sampling at supermarkets has also revealed that Spanish legislation of the commercial names in canned products needs to be updated since it has not been reviewed since 1986. Taking into account the current legislation, canned products are not obliged to show the scientific name on the labels, i.e., commercial names such as "Sepia" can be found, which correspond to several species. The authors consider that this system is no longer suitable for the current market, where the amount of cephalopod species in the market has greatly increased while different species may achieve significant differences in market price.
The Ct values obtained in this study for the target species are at the same level or lower than other recent works using the TaqMan real-time PCR technique for species identification [28,29]. The significant differences found between the data corresponding to *S. o*ffi*cinalis* and the other species prove that Ct values can be used to determine whether a sample contains *S. o*ffi*cinalis* or another cephalopod species. Results also prove the high specificity of the system, which works for the differentiation of *S. o*ffi*cinalis* from the other species of the genus *Sepia* with commercial importance, demonstrating the utility of the method in food control, since the reported cases of mislabeling in the family Sepiidae show substitutions between species belonging to the same genus, as shown in previous publications [9,10] and confirmed in this study. Although the system has not been tested with all the species of the genus *Sepia*, this study included those with relevance to the market. Nonetheless, further analysis could be carried out to confirm the specificity of the method with other species of the genus *Sepia* which might have some commercial relevance in certain countries. The level of mislabeling found in this work (25%) is slightly lower than those found in the aforementioned articles, but still in the range of significant mislabeling. However, the different sampling procedures do not allow an adequate comparison, therefore, authors cannot affirm that there has been a decrease in the mislabeling rates. Nevertheless, these results highlight the need for an effective tool for the control of this type of product.
#### **5. Conclusions**
As a conclusion, this work presents a rapid, non-expensive and reliable method, able to differentiate *S. o*ffi*cinalis* from other species of the genus *Sepia* and other cephalopod species in food samples with different levels of processing, making it useful for food control authorities in the whole food value chain.
This study also found a moderate level of mislabeling in Sepia products, which highlights the need for more efficient control of the authenticity of this type of product.
**Author Contributions:** Conceptualization, A.V. and C.G.S.; methodology, A.V. and G.R.-F.; software, A.V.; supervision, C.G.S.; writing—original draft, A.V.; writing—review and editing, G.R.F. and C.G.S. All authors have read and agreed to the published version of the manuscript.
**Funding:** This study is part of the SEATRACES project (www.seatraces.eu), funded by the EU Interreg Atlantic Area Programme (project number EAPA\_87/2016).
**Acknowledgments:** We acknowledge the Border Control Post of Vigo (BCP Vigo), Rogério Mendes, Patricia Ramos and Marta Pérez for providing tissue samples.
**Conflicts of Interest:** The authors declare no conflict of interest.
## **References**
© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
## *Article* **Swordfish or Shark Slice? A Rapid Response by COIBar–RFLP**
## **Venera Ferrito, Alessandra Ra**ff**a, Luana Rossitto, Concetta Federico , Salvatore Saccone and Anna Maria Pappalardo \***
Department of Biological, Geological and Environmental Sciences, Section of Animal Biology "M. La Greca", University of Catania, Via Androne 81, 95124 Catania, Italy; [email protected] (V.F.); [email protected] (A.R.); [email protected] (L.R.); [email protected] (C.F.); [email protected] (S.S.)
**\*** Correspondence: [email protected]; Tel.: +39-0957-306051
Received: 17 September 2019; Accepted: 28 October 2019; Published: 1 November 2019
**Abstract:** Market transparency is in strong demand by consumers, and the authentication of species is an important step for seafood traceability. In this study, a simple molecular strategy, COIBar–RFLP (cytochrome oxidase I barcode–restriction fragment length polymorphism), is proposed to unveil commercial fraud based on the practice of species substitution in the swordfish trade. In particular, COI barcoding allowed the identification of the species *Prionace glauca, Mustelus mustelus*, and *Oxynotus centrina* in slices labeled as *Xiphias gladius*. Furthermore, the enzymatic digestion of COI amplicons using the *Mbo*I restriction endonuclease allowed the simultaneous discrimination of the four species. Interestingly, an intraspecific differential *Mbo*I pattern was obtained for the swordfish samples. This pattern was useful to differentiate the two different clades revealed in this species by phylogenetic analyses using several molecular markers. These results indicate the need to strengthen regulations and define molecular tools for combating the occurrence of fraud along the seafood supply chain and show that COIBar–RFLP could become a standardized molecular tool to assess seafood authenticity.
**Keywords:** COIBar–RFLP (cytochrome oxidase I barcode–restriction fragment length polymorphism); seafood; fraud; DNA barcoding; food authenticity
#### **1. Introduction**
Swordfish fishery is one of the most important fishing activities in the Mediterranean Sea, in particular in South Italy. Quotas have been established to combat overfishing, and fisheries have been closed over several months to protect juveniles. According to recent data from the International Commission for the Conservation of Atlantic Tuna, Italy ranks the highest in terms of swordfish catches, which amount to 45% of the total allowable in the period 2003–2016 [1]. The highest demand for fish products in general, and swordfish in particular, occurs during summer, especially in restaurants [2] but also in local markets. As a result of the high demand, the price of these large pelagic fishes is on average higher than that of small fishes [3].
With the increase in demand and price, alimentary fraud potentially increases too. This can include food mislabeling, substitution, counterfeiting, misbranding, dilution, and adulteration [4]. The mislabeling of seafood can be harmful for health, in terms of economic loss, as well as for the loss of biodiversity it may cause in the case of illegal trade of threatened species. For these reasons, European regulations have focused on traceability and, in particular, on the mandatory declaration of the species present in a product on the product label [5–10].
Despite these adequate legislative tools, the number of cases of food fraud perpetrated in the fish trade in Europe and worldwide is increasing. The results of recent investigations on this phenomenon have shown that the percentage of mislabeling was around 30% of the total samples collected [11–15]. To address this problem, researchers have increasingly asserted the importance of using molecular tools based on DNA sequencing for detecting food fraud. The most common mitochondrial (mt) genes used for this purpose have been cytochrome b, 16S rRNA, and cytochrome oxidase I (COI). Other mtDNA targets, such as the mtDNA control region (CR) [16,17], which has been the most popular molecular marker used for genetic population structure studies [18–24], have seen limited use in fish and seafood species identification. In recent years, COI has been standardized as a barcode gene for species identification in several animal taxa [25–33] including fishes [34–40]. More specifically, the high number of COI-barcode fish sequences available in the large public gene sequence databases (BOLD and GenBank) [41,42] have made this gene the most highly used gene to clearly identify fish species and cases of mislabeling of seafood products [38,43–54]. However, in the context of seafood traceability, the main goal for the implementation of these analyses is to reduce the time it takes from sampling to obtaining gene sequencing results, as well as the costs of processing.
An already well-proven technique for the identification of species is polymerase chain reaction (PCR)–restriction fragment length polymorphism (RFLP), by which the PCR product of an amplified gene is cut with different restriction endonucleases to obtain a species-specific RFLP [55–57], useful for species authentication. In this regard, the combination of DNA barcoding of COI and the consolidated method of RFLP analysis (COIBar–RFLP, cytochrome oxidase I barcode–restriction fragment length polymorphism) has been successfully used to discriminate several fish species belonging to the Engraulidae, Merluccidae, Soleidae, and Acipenseridae families in processed seafood products [49,52–54,58]. It should be noted that the time and cost of execution of the COIBar–RFLP are lower than those of DNA sequencing (about 7 h and 10 euros per sample *vs* 24 h and 17 euros per sample, respectively).
Focusing on swordfish adulteration problems, the most commonly used species for fraudulent substitution are elasmobranches, including some species of shark. It should be noted that the market of shark meat is very wide for both fresh and frozen foods also in Italy, and cases of mislabeling have been frequently recorded for these products imported from all over the world [59–61].
Therefore, food fraud occurs due to an economic return when using shark meat. However, the substitution of a more valuable fish, such as swordfish, with shark meat leads to an even more serious fraud in economic terms. In the last decades, several studies have been carried out to detect the rate of mislabeling of different seafood products, and in some cases, shortfin mako (*Isurus oxyrinchus*) and blue shark (*Prionace glauca*) have been found to be sold as swordfish [36,37,62,63].
On the basis of the considerations above, the aim of this work is to extend the use of COIBar–RFLP to investigate the identity of swordfish products in the south of Italy and to discriminate swordfish (*Xiphias gladius*) from other fish species to detect fraudulent actions, such as species substitution, which represent the most common fraud in seafood. First, we sequenced the conventional COI barcode in a large number of samples collected in local fish markets and supermarkets, labeled as swordfish slices. Subsequently, the COIBar–RFLP procedure was applied on reference samples of the COI-barcoded species to obtain a species-specific restriction enzyme pattern. Finally, this pattern was used for swordfish slice authentication.
#### **2. Materials and Methods**
#### *2.1. Sampling*
Fresh and frozen slices of swordfish were acquired in 2010 and 2018 from local fish markets and supermarkets of south Italy for a total of 35 samples. Another 10 samples from the local harbor were collected and identified on the basis of morphological traits [64,65] and used to construct a reference COI-barcode library. The samples collected in 2010 had already been processed [37] and were used in this study only for the application of COIBar–RFLP. The remaining samples, preserved at room temperature in 1.5 mL labeled tubes filled with 95% ethanol, were processed for DNA barcoding and COIBar–RFLP (Table 1). DNA samples were deposited as vouchers at the Department of Biological, Geological, and Environmental Science, Section of Animal Biology, in Catania, Italy.
**Table 1.** Samples examined in this study.
*Foods* **2019**, *8*, 537
DBGES (Department of Biological, Geological and Environmental Sciences). BLAST (Basic Local Alignment Search Tool). \* = species of the samples identified by using morphological features.
*Foods* **2019**, *8*, 537
#### *2.2. DNA Barcoding*
Genomic DNA was extracted from 25 mg of tissue using a commercial kit based on silica purification (DNeasy tissue kit, Qiagen, Hilden, Germany) following the manufacturer's guidelines. All samples were analyzed by amplifying a portion of about 650 bases of the COI gene in a 20 µL reaction mixture also containing the M13 tailed primers (VF2\_t1 and FishR2\_t1) described in Ivanova et al. [66] to improve the sequencing quality of the PCR products and following the PCR conditions reported by Pappalardo et al. [54]. All PCR products were checked by 0.8% agarose gel electrophoresis, visualized with SYBR® Safe (Thermo Fisher, Waltham, MA USA), displayed through a Safe Imager TM 2.0 Blue Light Transilluminator (Thermo Fisher, Waltham, MA USA), and then purified with the QIAquick PCR purification kit (Qiagen, Hilden, Germany). Sanger sequencing, using M13 primers, was subsequently conducted by Genechron in both forward and reverse directions to generate the DNA barcodes [67].
The sequence chromatograms were checked visually and assembled. Multiple-sequence alignment was carried out by the online version of MAFFT v.7 [68]. Ambiguous sequences were trimmed, and primer sequences were cut. The sequences were carefully checked for the presence of nuclear mitochondrial pseudogenes or NUMTs (nuclear mitochondrial DNA sequences), which could be easily coamplified with orthologous mtDNA sequences [69]. The EMBOSS Transeq tool [70] was used to translate the nucleotide sequences to amino acids to check for premature stop codons and to verify that the open reading frames were maintained in the protein-coding locus. To confirm the identity of the amplified sequences, we conducted BLAST (Basic Local Alignment Search) searches in GenBank with default parameters [71]. All sequences obtained from the present study were published in the National Center for Biotechnology Information database (NCBI), and their GenBank accession numbers are reported in Table 1. After the BLAST search, six shark species sequences (HM909857, JF493927, KF899461, KI709900, JF493694, JN641217) downloaded from GenBank were added to our dataset to construct a phylogenetic tree. We used jModelTest v 2.1.10 [72] to select the best-fitting substitution model for our sequences according to the corrected Akaike information criterion. A maximum likelihood (ML) tree by using a GTR + I + G model was implemented in MEGA v 6.0 (Biodesign Institute, Arizona, MA, USA) [73]. The evaluation of the statistical confidence of nodes was based on 1000 non-parametric bootstrap replicates [74].
#### *2.3. COIBar–RFLP*
The selection of the most suitable restriction enzymes to discriminate swordfish from other shark species (*Mustelus mustelus,* L., 1758, *Oxinotus centrina* (L., 1758), *P. glauca* (L., 1758)*, Scyliorhinus canicula* L., 1758) was performed through "Remap" [75]. The in silico analysis was preliminarily carried out using a total of 10 COI barcode sequences (of about 650 bases) of the examined species, downloaded from public databases (GenBank and BOLD) [41,42]. Five different restriction enzymes were tested to scan all validated sequences and to detect the expected size of the digested products: *Hpa*II (C\*CGG), *Hinf*I (G\*ANTC), *Mbo*I (\*GATC), *Rsa*I (GT\*AC), and *Hind*III (A\*AGCTT). Finally, a total of 49 COI sequences were analyzed by Remap to test for evidence of intraspecific variation at the recognition site of the restriction endonuclease suitable for simultaneous discrimination of the examined species (Figure 1, Table 2).
*Foods* **2019**, *8*, 537 7 of 15
**Figure 1.** Flow chart of COIBar–RFLP (cytochrome oxidase I barcode–restriction fragment length polymorphism) or species discrimination. DNA barcoding steps: DNA isolation from swordfish slices and barcode region PCR amplification. In silico analysis steps: search for an appropriate restriction enzyme. RFLP steps: incubation of barcode amplicons with *Mbo*I to obtain the COIBar–RFLP pattern. ML, maximum likelihood; nBLAST, nucleotide Basic Local Alignment Search Tool. **Figure 1.** Flow chart of COIBar–RFLP (cytochrome oxidase I barcode–restriction fragment length polymorphism) or species discrimination. DNA barcoding steps: DNA isolation from swordfish slices and barcode region PCR amplification. In silico analysis steps: search for an appropriate restriction enzyme. RFLP steps: incubation of barcode amplicons with *Mbo*I to obtain the COIBar–RFLP pattern. ML, maximum likelihood; nBLAST, nucleotide Basic Local Alignment Search Tool.
KT307360 648 ≈ 510 - 95 KT307361 620 ≈ 480 - 95 KT307362 648 ≈ 510 - 95 KT307363 648 ≈ 510 - 95 KT307364 648 ≈ 510 - 95 JF834320 672 ≈ 505 - 100 KY176547 642 ≈ 495 - 105
*Oxynotus centrina* 9
**Table 2.** *Cont*.
Afterwards, the COI-barcode PCR products obtained from *X. gladius* and shark samples were digested with the selected restriction enzymes. For each endonuclease, a 15 µL reaction volume containing 13 µL of unpurified PCR product, 1 µL of digestion buffer (1X), and 1 µL of each endonuclease (10 U each) was prepared. The reaction mixtures were incubated at an optimum temperature of 37 ◦C for 1 h. The digested amplicons were then separated on a 3% agarose gel using Trackit TM 100 bp DNA ladder (Invitrogen) as a size standard. The restriction pattern obtained from the validated samples was exploited to unequivocally identify the unknown commercial slices.
#### **3. Results**
#### *3.1. DNA Barcoding*
The length range of the obtained COI sequences was between 669 bases and 681 bases. Each of them was a functional mitochondrial sequence without stop codons. NUMTs generally smaller than 600 bases were not sequenced [71]. Five species were identified in all examined samples: *X. gladius* (Xiphiidae), *P. glauca* (Charcarinidae), *M. mustelus* (Triakidae), *S. canicula* (Scyliorinidae), and *O. centrina* (Oxynotidae). The sequences obtained from morphologically validated species were compared with the sequences retrieved from GenBank through a BLAST search. The identity percentage between the COI query sequences and their top-match sequences ranged from 98.07% to 100% (Table 1). The ML tree (Figure 2) showed the relationship between the sequences of several unidentified samples and the reference barcode sequences. High bootstrap values (>60%) supported the nodes connecting the sequences of the same species in the tree. The samples of *X. gladius* clustered into two main clades (named clade I and II), as already found by Pappalardo et al. [36,37]. Only one case of mislabeling
(1 out of 15) was found in the samples examined in 2010 (6.7%), while 15% (3 out of 20) of mislabeling was found in the samples collected during 2018 (Table 1). Swordfish was substituted with *P. glauca* (2 products), *M. mustelus* (1 product), and *O. centrina* (1 product). *Foods* **2019**, *8*, 537 9 of 15
**Figure 2.** Maximum likelihood (ML) tree showing the relationships of unknown samples sequences (X and Y) to validated reference barcode sequences. The numbers above the nodes represent bootstrap analyses after 1000 replicates. Bootstrap values greater than 60% are shown. The red square indicates swordfish mislabeled samples. Scale bar refers to a distance of 0.05 nucleotide substitutions per site. **Figure 2.** Maximum likelihood (ML) tree showing the relationships of unknown samples sequences (X and Y) to validated reference barcode sequences. The numbers above the nodes represent bootstrap analyses after 1000 replicates. Bootstrap values greater than 60% are shown. The red square indicates swordfish mislabeled samples. Scale bar refers to a distance of 0.05 nucleotide substitutions per site.
#### *3.2. COIBar–RFLP 3.2. COIBar–RFLP*
The preliminary in silico analysis using "Remap" showed that the *Mbo*I enzyme produced a species-specific pattern useful to discriminate simultaneously all examined species. No intraspecific variation of the *Mbo*I recognition sites was detected for any species tested by "Remap", with the exception of the *X. gladius* digestion pattern (Table 2). Figure 3 highlights both the size of the undigested COI amplicon, of about 750 bp, and the *Mbo*I differential restriction pattern obtained for The preliminary in silico analysis using "Remap" showed that the *Mbo*I enzyme produced a species-specific pattern useful to discriminate simultaneously all examined species. No intraspecific variation of the *Mbo*I recognition sites was detected for any species tested by "Remap", with the exception of the *X. gladius* digestion pattern (Table 2). Figure 3 highlights both the size of the undigested COI amplicon, of about 750 bp, and the *Mbo*I differential restriction pattern obtained for each species:
each species: one fragment of 510 bp was obtained for *O. centrina*; two fragments of 110 and 400 bp and of 150 and 400 bp were obtained, respectively, for *P. glauca* and *S. canicula*; finally, three one fragment of 510 bp was obtained for *O. centrina*; two fragments of 110 and 400 bp and of 150 and 400 bp were obtained, respectively, for *P. glauca* and *S. canicula*; finally, three fragments of 120, 180, and 390 bp were obtained for *M. mustelus*. The negative control is not shown in the figure. The enzymatic digestion of *X. gladius* amplicons produced two different patterns (Figure 4) corresponding to clades I and II, already described in this species. In particular, three fragments of 170, 220, and 240 bp were detected for clade I and three fragments of 170, 220, and 280 bp were found for clade II. On the basis of this intraspecific pattern, the swordfish sample shown in Figure 3 belongs to clade I. *Foods* **2019**, *8*, 537 10 of 15 fragments of 120, 180, and 390 bp were obtained for *M. mustelus*. The negative control is not shown in the figure. The enzymatic digestion of *X. gladius* amplicons produced two different patterns (Figure 4) corresponding to clades I and II, already described in this species. In particular, three fragments of 170, 220, and 240 bp were detected for clade I and three fragments of 170, 220, and 280 bp were found for clade II. On the basis of this intraspecific pattern, the swordfish sample shown in Figure 3 belongs to clade I. *Foods* **2019**, *8*, 537 10 of 15 fragments of 120, 180, and 390 bp were obtained for *M. mustelus*. The negative control is not shown in the figure. The enzymatic digestion of *X. gladius* amplicons produced two different patterns (Figure 4) corresponding to clades I and II, already described in this species. In particular, three fragments of 170, 220, and 240 bp were detected for clade I and three fragments of 170, 220, and 280 bp were found for clade II. On the basis of this intraspecific pattern, the swordfish sample shown in Figure 3 belongs to clade I.
**Figure 3.** Example of COIBar–RFLP identification of swordfish and shark species on a 3% agarose gel by restriction by *Mbo*I of the cytochrome oxidase I amplicons. Bands smaller than 100 bp were not considered. The 5ND and 5D bands differ in intensity because they were obtained from two different PCR amplifications. ND = not digested, D = digested. M = molecular weight marker (100 bp DNA ladder, biotechrabbit GmbH, Berlin, Germany). **Figure 3.** Example of COIBar–RFLP identification of swordfish and shark species on a 3% agarose gel by restriction by *Mbo*I of the cytochrome oxidase I amplicons. Bands smaller than 100 bp were not considered. The 5ND and 5D bands differ in intensity because they were obtained from two different PCR amplifications. ND = not digested, D = digested. M = molecular weight marker (100 bp DNA ladder, biotechrabbit GmbH, Berlin, Germany). **Figure 3.** Example of COIBar–RFLP identification of swordfish and shark species on a 3% agarose gel by restriction by *Mbo*I of the cytochrome oxidase I amplicons. Bands smaller than 100 bp were not considered. The 5ND and 5D bands differ in intensity because they were obtained from two different PCR amplifications. ND = not digested, D = digested. M = molecular weight marker (100 bp DNA ladder, biotechrabbit GmbH, Berlin, Germany).
#### **4. Discussion**
The results obtained in this study once again confirm the efficacy of COIBar–RFLP in discriminating fish species in commercial products and also highlight the fraudulent practice of species substitutions in seafood products, consisting in the use of less valuable shark species in place of swordfish. The *Mbo*I endonuclease restriction enzyme produced species-specific restriction patterns of the COI amplicons useful to differentiate *X. gladius* from shark species. Another interesting result proving the sensitivity of this methodology is the intraspecific differential *Mbo*I pattern obtained for the swordfish samples. This pattern was useful to discriminate the two different clades revealed in this species by phylogenetic analyses using several molecular markers [36,37,76–78]. COI DNA-barcoding showed that 15% of the swordfish samples purchased in local fish markets during 2018 was mislabeled, with an evident economic loss for the consumers. This percentage was at least two times higher than that recorded in 2010, demonstrating that despite the current European legislation focused on consumer protection against fraud, fraud remains frequent and widespread. In this context, there is no doubt that molecular tools are very useful and effective to fight commercial fraud and that DNA-based methods have become increasingly important for seafood authentication. However, while the practice of commercial fraud in the seafood market is a global concern, to date there is no standardized global methodology to expose this practice. Firstly, all states have not yet incorporated into their legislation the use of molecular methods to combat commercial fraud; this is true for Italy, for example. Secondly, significant differences among countries have been found in methods used by accredited laboratories for food authenticity [79]. Thirdly, together with the classic methods (protein- and DNA sequence-based methods), new and sophisticated methods are being developed to identify seafood species [80]. It is evident that the first two issues can be solved only by adopting a common global policy to fight food fraud. The European legislation, for example, could require, rather than only suggest, the application of DNA analysis in the context of seafood traceability [81], also indicating the most useful methodology to be used across European laboratories. In this regard, the features that molecular methods should have for a rapid authentication of species in seafood products can be debated. To be effective for routine activities carried out by local food safety and quality authorities, from the traceability of the catch to the labeling of the products, effectiveness in terms of cost and time-saving and correctness of species identification should be a priority. Among the classic methods, the protein-based methods, such as isoelectric focusing of sarcoplasmic proteins, are still used as official methods for fish species identification [82], but the DNA-sequencing methods, and the DNA-barcoding methodology in particular, have become more common in laboratories specialized in food authentication ([3] and literature therein). Increasingly, new methodologies are emerging for species identification, such as qPCR, DNA microarrays, high-resolution melting analysis, mass spectrometry, high-throughput sequencing, and the recently developed handheld testing devices [80], all of them suitable and effective in terms of cost and time consumption.
However, these new methodologies require, in some cases, extensive technical equipment and specific skills by the operators and need to be standardized for use as official methods. Furthermore, the application of these methods is limited to a few cases of species authentication, while wide databases of reference samples are needed for their validation as official methods. The methodological approach we propose, COIBar–RFLP, although it cannot substitute DNA sequencing in general, takes advantage of large databases of reference DNA sequences of fish species and of the positive results from several study cases for species of relevant commercial interest under various food matrices [49,52–54,58]. COIBar–RFLP successfully and simultaneously discriminated the fish species analyzed in these studies, through the banding pattern obtained after digestion with only one endonuclease restriction enzyme. This simple, robust, easy-to-perform, and cost-effective strategy can potentially cover a wide range of species and provide a versatile tool to monitor the mislabeling of fish products. However, it should be noted that poor enzyme storage, as well as the processing conditions, could compromise the advantages of the methodology in terms of expected time of processing and misleading results. In a recent investigation on the methodological approach performed in 45 European laboratories, Griffiths et al. [79] revealed that PCR–RFLP was used in 40% of the laboratories involved in seafood authentication; this result suggests that this method could become a standardized molecular tool to assess seafood authenticity.
#### **5. Conclusions**
The efficacy of COIBar–RFLP was tested for species authentication on slices labeled as swordfish. The illegal practice of species substitution was observed, with the species *P. glauca*, *M. mustelus*, and *O. centrina* being sold in place of swordfish. These results indicate the need to strengthen regulations and to define molecular tools to fight the occurrence of fraud along the seafood supply chain, from the traceability of the catch to the labeling of the products, and to achieve market transparency, which is highly demanded by the consumers. Finally, the future perspectives of COIBar–RFLP rest on the need to build a database of COI restriction patterns to be used for unequivocal species identifications.
**Author Contributions:** A.M.P. and V.F. conceived and designed the experiments; A.M.P., A.R., and L.R. performed the experiments; A.M.P., C.F., and S.S. analyzed the data; A.M.P. and V.F. wrote the paper.
**Funding:** This work was supported by the Annual Research Plan 2016-2018 of the Department of Biological, Geological and Environmental Sciences, University of Catania (Grants #22722132134).
**Acknowledgments:** We thanks the University of Catania for the economic support.
**Conflicts of Interest:** The authors declare no conflict of interest.
#### **References**
© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
*Article*
## **Comparison of Targeted (HPLC) and Nontargeted (GC-MS and NMR) Approaches for the Detection of Undeclared Addition of Protein Hydrolysates in Turkey Breast Muscle**
**Liane Wagner <sup>1</sup> , Manuela Peukert <sup>1</sup> , Bertolt Kranz <sup>1</sup> , Natalie Gerhardt <sup>2</sup> , Sabine Andrée 1 , Ulrich Busch <sup>2</sup> and Dagmar Adeline Brüggemann 1,\***
Received: 29 June 2020; Accepted: 6 August 2020; Published: 8 August 2020
**Abstract:** The adulteration of fresh turkey meat by the undeclared addition of protein hydrolysates is of interest for fraudsters due to the increase of the economic gain by substituting meat with low cost ingredients. The aim of this study was to compare the suitability of three different analytical techniques such as GC-MS and <sup>1</sup>H-NMR with HPLC-UV/VIS as a targeted method, for the detection of with protein hydrolysates adulterated turkey meat. For this, turkey breast muscles were treated with different plant- (e.g., wheat) and animal-based (e.g., gelatin, casein) protein hydrolysates with different hydrolyzation degrees (15–53%: partial; 100%: total), which were produced by enzymatic and acidic hydrolysis. A water- and a nontreated sample (REF) served as controls. The data analyses revealed that the hydrolysate-treated samples had significantly higher levels of amino acids (e.g., leucine, phenylalanine, lysine) compared with REF observed with all three techniques concordantly. Furthermore, the nontargeted metabolic profiling (GC-MS and NMR) showed that sugars (glucose, maltose) and/or by-products (build and released during acidic hydrolyses, e.g., levulinic acid) could be used for the differentiation between control and hydrolysates (type, degrees). The combination of amino acid profiling and additional compounds gives stronger evidence for the detection and classification of adulteration in turkey breast meat.
**Keywords:** <sup>1</sup>H-NMR; GC-MS; HPLC-UV/VIS; metabolomics; food fraud; protein hydrolysate; free amino acid contents; ProHydrAdd
#### **1. Introduction**
Meat is an important supplier of high-quality nutrients such as proteins, minerals and vitamins. Since it is sold on the market for a low price, which does not cover the increasing production cost, it is in focus for food fraud. Adulterators look for opportunities to increase the economic gain. Possibilities would be to misrepresent, use illegal supply chains and/or manipulate the food product, e.g., replace/substitute some, or all, premium quality materials with lower-grade, cheaper cuts of meat, meat from other species or nonmeat components (e.g., water, additives) [1–3]. The fraud can influence the consumers' satisfaction and confidence (religious, moral, cultural), but worse it can be extremely dangerous for human health, e.g., causing illness, provoke allergies or even causing
death (e.g., melamine scandal) [1,4]. Therefore, it is necessary to have reliable analytical methods to detect adulteration.
The water binding capacity of meat is strongly related to the rate of early postmortem metabolism and the ultimate pH value. It is lowest at a pH of 4.9–5.4, but will increase with increasing or decreasing of the pH [5]. The ultimate pH of poultry breast muscle is 5.67–5.69 and therefore, the water uptake capacity is high [6]. A simple exposure to just water can lead to weight gain of the meat and consequently increase the economic gain. For a fair market competition, the detection of such fraudulent practices is required. The traditional method to determine extraneous water in meat is to analyze the water/protein ratio [7]. In an untreated sample is the water/protein ratio of chicken as well as turkey breast muscle <3.40, of chicken legs between 4.05–4.30 and of turkey legs between 3.80–4.05 (Commission regulation No 543/2008) [8]. If water is added, the water/protein ratio will be higher.
For more than a decade an undeclared addition of protein hydrolysates to poultry meat or meat products could be observed. This way, the analytical protein content rises, masking the water addition [3]. Protein hydrolysates consist mainly of amino acids and possibly peptides. They are cheaper, better soluble and harder to detect than protein additions. Besides amino acids, hydrolysates contain additional compounds like carbohydrates, fatty acids and/or side products, which are formed during the hydrolyzation process. A suitable method is the detection of the free amino acid contents using high performance liquid chromatography with ultraviolet-visible detection (HPLC-UV/VIS). This technique is well established in laboratories and is used as an alternative to the official method in Germany (§64 LFGB: Determination of free amino acids in meat using gas chromatography with flame ionization detection (GC-FID)) [9].
Alternative analytical approaches such as nontargeted metabolomics provide an entire profile (chromatogram, spectrum, fingerprints, etc.) of a suspicious sample [10,11]. These methods have gained more and more in importance in recent years due to strong technical improvements. With these promising and valuable high-throughput tools such as mass spectrometry (MS) based techniques or nuclear magnetic resonance (NMR) spectroscopy, it is possible to identify and quantify small organic molecules with molecular weights of less than 1.5 kDa, including carbohydrates, peptides, nucleotides, lipids and amino acids [12]. In order to analyze such a broad spectrum of metabolites with diverse properties and concentrations (over several magnitudes) in just one single sample, it is important to have techniques, which are robust and sensitive. The major techniques are MS coupled to a chromatographic separation and NMR. The high sensitivity and selectivity make MS a powerful tool to detect molecular masses and fragmentation patterns for chemical structure identification. It is possible to profile myriads of metabolites due to the different combinations of separation, ionization and detection technique. On the other hand, NMR spectroscopy provides characteristic information on the metabolic profile by analyzing small amounts of one sample in a nondestructive, quantitative and short-time period way with convenient sample preparation [13]. However, independent of the used approach (targeted or nontargeted), large reference datasets are required to account for the natural variation in different products due to multiple influencing factors such as feeding or storage conditions (duration and temperature) [3,14].
Chemometrics is a powerful multivariate data analysis tool that reduces a huge amount of generated data by (1) grouping or ordering unknown samples with similar characteristics (qualitative) and (2) ascertaining adulterant analytes in sample (quantitative) or (3) for assessing their quality or authenticity [14]. Chemometrics, and for this purpose used clustering, principal component analyses (PCA) as well as regression analyses, is a routine complement for MS- and NMR-based metabolomics [15].
The objective of this study was to compare the suitability of three different analytical methods such as GC-MS and <sup>1</sup>H-NMR (nontargeted approaches) and HPLC-UV/VIS as a targeted method, for the detection of adulterated turkey breast muscle with protein hydrolysates. HPLC-UV/VIS was selected due to the fact that it requires inexpensive equipment frequently available in every control laboratory. The first part focuses on the identification of specific markers such as amino acids, which are detectable
with all three methods. The second part (nontargeted approaches) focuses on the identification of additional markers as well as possible classification of the added hydrolysates.
#### **2. Materials and Methods**
### *2.1. Chemicals*
Sodium hydroxide, 5-sulfo salicylic acid, sodium chloride (NaCl, 99%), hydrochloric acid (HCl), sodium acetate, ninhydrin, hydrindantin-dihydrat, methanol and chloroform were obtained from Merck (Darmstadt, Germany). Ethylenediaminetetraacetic acid (EDTA, 99%) was purchased from AppliChem (Darmstadt, Germany) and tris(hydroxymethyl)aminomethane (TRIS Pufferan, Tris) from Carl Roth (Karlsruhe, Germany). The reagents (eluent buffers A, B, C, D, E, F; cleaning solution W, derivatization reagent R, sampling dilution buffer and autosampler solution) used for the amino acid (AA) determination were bought from membraPure (Henningsdorf, Germany). The AA standards (AA standards physiological: acidics–neutrals–basics), the amino acid mix solution (certified) and l-asparagine, l-glutamine, s-(2-aminoethyl)-l-cysteine hydrochloride (thialysine), l-norleucine, isopropanol, gelatin hydrolysate enzymatic, HyPep® 4601 protein hydrolysate from wheat gluten, protein hydrolysate N-Z-amine® AS, casein from bovine milk, gelatin from porcine skin and gluten from wheat were bought from Sigma-Aldrich (Steinheim, Germany). 2-Methoxyethanol was obtained from Riedel-deHaen (Seelze, Germany). For GC-MS derivatization methoxyamine hydrochloride (MAH) and pyridine were purchased from Sigma-Aldrich (Steinheim, Germany), and *N*-methyl-*N*-(trimethylsilyl)trifluoroacetamide plus 1% chlorotrimethylsilane (MSTFA + 1% TMCS) were purchased from Thermo Fisher (Dreieich, Germany). For NMR analyses reagents such as deuterium oxide (D2O) containing 0.05 wt% TSP (sodium-3-(trimethylsilyl)-2,2,3,3-tetradeuteriopropionate) and maleic acid were obtained from Sigma-Aldrich (Steinheim, Germany) and D2O (99.96%) was obtained from VWR (Ismaning, Germany).
#### *2.2. Sampling and Adulteration of Turkey Breast*
**Meat.** Three female turkeys (BUT Big 6, *Meleagris gallopavo*) with an average weight of 10.3 kg and in average 112 days old were collected at 4 ◦C from a slaughterhouse in Germany directly after slaughter. The *Musculus pectoralis superficialis* was taken after dissection [16]. After sampling, all meat pieces were immediately frozen in liquid nitrogen and stored at −20 ◦C until further use.
**Protein hydrolyzation.** The protein powders (1.0 g of casein from bovine milk, gelatin from porcine skin or gluten from wheat) were hydrolyzed at 150 ◦C in 8 mL of 6 M aqueous HCl solution for 1 h (total hydrolyzation: TH, degree of hydrolyzation: 100%). After that, neutralization with solid 1.0 M NaOH was performed. The final concentration of amino acids was 55.6 g/L (~0.5 M).
The protein hydrolysate powders (gelatin hydrolysate enzymatic, HyPep® 4601 protein hydrolysate from wheat gluten, protein hydrolysate N-Z-amine® AS (casein)) were used without further hydrolyzation (partial hydrolyzation: PH). All three bought protein hydrolysates were peptones (enzymatic hydrolysis with pepsin (E.C.3.4.4.1) or acidic hydrolysis). The degree of hydrolyzation was analyzed photometrically as the following describes. From each sample, 250 µL (aqueous AA solution) were added to 250 µL of 4 M acetate buffer (pH 5.5) and mixed. After that, 75 µL of 1 M NaOH and 250 µL ninhydrin solution (174 mg Ninhydrin + 28 mg hydrindantin-dihydrat in 10 mL 2-methoxyethanol) were added. The reaction took place at 95 ◦C for 20 min and was stopped by cooling the mixture in an ice bath (0 ◦C) for 20 min. 300 µL of each mixture was diluted with 1000 µL isopropanol solution (50% *v*/*v* in water) and measured at room temperature at 570 nm (photometer DU 640, Beckman Coulter GmbH, Krefeld, Germany). The degree of hydrolyzation (DH) was calculated by comparing the free amino groups of the samples with a solution of the same not hydrolyzed protein (0% DH) and total hydrolyzed protein (100% DH). That implies a hydrolyzation degree for gelatin, wheat and casein of 15% ± 3%, 16% ± 2% and 53% ± 2%, respectively.
**Hydrolysate injection.** About 1 mL solution (0.5 M) hydrolyzed protein or commercially available protein hydrolysate or water was injected per g turkey breast meat across and along the muscle fiber. The samples were frozen at <sup>−</sup><sup>80</sup> ◦C and lyophilized for GC-MS and <sup>1</sup>H-NMR analyses. All samples were stored at −80 ◦C until further use.
**Sample code.** The sample codes were chosen as followed: reference sample without injection is called REF, an additional control sample injected with water is called water. The different protein hydrolysates (gelatin (G), wheat (W) and casein (C)) with different hydrolyzation degree (partial (PH) or total (TH)) are indicated with the following codes: GPH, WPH and CPH for partial, respectively, and GTH, WTH and CTH for total hydrolyzation, respectively. This means, for example, that the sample GPH contained protein hydrolysate gelatin and was partially hydrolyzed.
#### *2.3. Amino Acid Analysis Using HPLC-UV*/*VIS*
#### 2.3.1. Sample Preparation for Amino Acid Analysis
The frozen turkey samples (2 g each) were homogenized with an Ultra-Turrax T10 (12,000 rpm, IKA Werke GmbH und CO KG, Staufen, Germany) in 5 mL 0.025 M EDTA/0.100 M Tris buffer (pH 8.0) at −20 ◦C. As internal standards, l-norleucine and l-thialysine were used with a final concentration of 133 µM and 88 µM, respectively. The proteins and longer peptides were precipitated with 30% *v*/*v* of 15% 5-sulfosalicylic acid over 30 min at 4 ◦C at a resulting pH of about 2.2 (pH less than 2.5 is recommended for cation exchange chromatography). The samples were centrifuged (15 min, 4 ◦C, 6827× *g*), filtered (45 µm) and then frozen at −20 ◦C until further use. The centrifugation and filtration were repeated directly before the analysis.
#### 2.3.2. Amino Acid Content Determination
The samples were analyzed in duplicates with internal (l-norleucine- and l-thialysine-solution, see Section 2.3.1) and external AA standards. The external standards were worked up similar to the meat samples. A regularly calibration of the amino acid analyzer was done with a certified AA standard (amino acid mix solution [17]). For the cation exchange chromatography, a column with 3 µm beads with a separation over a pH-range from 2.9 to 10.4 and ninhydrin post column derivatization was performed. An injection volume of 20 µL (pH 2.2) and a flow rate of 180 µL/min were applied. For the spectrophotometric analysis two photometers with wavelengths of 440 nm and 570 nm were used for detection of the free amino acids. The limit of detection (LOD) and quantification (LOQ) were defined as three and ten times the signal to noise ratio of the external standard solution, respectively. All measured contents of the free amino acids (FAA) were above the LOQ (0.13 mg/100 g–0.33 mg/100 g, depending on the different AA).
#### *2.4. GC-MS-Based Metabolomics Analyses*
#### 2.4.1. Sample Preparation for Metabolomics Study of White Breast Meat
The sample set consisted of study samples, mix samples and blanks. The mix samples were prepared by combining an aliquot from each muscle powder and served for normalization. All sample types were extracted in the same manner. Study samples were extracted in triplicates. In a first step 20 mg of the dried, homogenous meat powder was extracted with 600 µL ice-cold 80% methanol containing 10 internal standards using a bead mill homogenizer (Minilys, Bertin Technologies SAS, Montigny-le-Bretonneux, France) for 2 times 30 s and an ultrasonic bath for 2 min. The raw extract was centrifuged at 15,000× *g* for 20 min at 4 ◦C. In a second step the pellet was re-extracted with 600 µL ice-cold methanol:chloroform (2:1 *v*/*v*) according to the first extraction step. Both supernatants were combined, mixed and centrifuged at 15,000× *g* for 20 min at 4 ◦C. 100 µL of the supernatant were transferred into 2 mL glass vials containing a 200 µL glass insert and evaporated in a vacuum
centrifuge (Christ Speedvac RVC 2¨C18 CD plus, Germany). The dried samples were stored under protective argon atmosphere at −80 ◦C until analysis.
#### 2.4.2. GC-MS Measurements and Data Processing
Prior to measurement samples were derivatized by methoximation and trimethylsilylation. For methoximation samples were shaken in 30 µL of a 20 mg/mL solution of MAH in pyridine at 50 ◦C for 1 h. Subsequently, 70 µL MSTFA+1%TMCS were added and samples shook at 70 ◦C for 1 h. GC-MS analysis was performed on a Shimdazu GCMS QP2010 instrument (Shimadzu, Duisburg, Germany) equipped with an OPTIC-4 injector (GL Sciences, Eindhoven, The Netherlands). A 1 µL aliquot of each sample was injected in a 1:7 split ratio. A series of n-alkanes (C7-C30) was used as a retention time standard. Analytes were separated on a 30 m Rxi-5SIL MS column containing a 10 m Integra-Guard column (Restek, 0.25.mm i.d., 0.25 µm film thickness), and with a linear temperature gradient starting from 80 ◦C to 300 ◦C with 5 ◦C/min and a final 5 min hold at 320 ◦C. Masses between 60 and 600 *m*/*z* were scanned.
For data analysis only those features were selected that were not present in blank samples. The annotation of compounds was performed using the NIST 14 library database implemented in GCMSsolution (Shimadzu, Duisburg, Germany). The retention times and selected masses used for relative quantification are listed in Supplementary Table S1. Samples were normalized according to the calculated means of the mix samples for each feature to reduce the impact of device maintenance (instrument tuning, liner exchange and septum exchange during the measurement batch). Briefly, signal intensities of the mix-samples within one measurement period (between device maintenance) were averaged as well as for the whole measurement batch. Using these means a correction factor was determined between the total means and the means for each measurement period between device maintenances. These correction factors were then applied to the corresponding features of study samples and calibration samples.
#### 2.4.3. Amino Acid Quantification
For AA quantification a calibration curve consisting of a reference standard mixture and 8 final concentrations in the range 1.96–250 pmol/µL was applied. These standard mixes were spiked into aliquots from pooled REF samples (50 µL of standard mixture into 50 µL of REF mixture) to reduce for the impact of the sample matrix. The calibration samples were prepared in duplicates. Calibration samples were evaporated in a vacuum centrifuge and prepared for GC-MS measurement according to Section 2.4.2. For quantification, the calibration samples were at first normalized and then the mean values (ion counts) of selected quantitative ions from calibration-reference samples were subtracted from calibration samples containing the standard mixtures. Calibration coefficients were R<sup>2</sup> > 0.99 with acceptance of alanine which had an R<sup>2</sup> = 0.97). The retention times and selected masses used for quantification are listed in Supplementary Table S2.
### *2.5. <sup>1</sup>H NMR-Based Metabolomics Analyses*
#### 2.5.1. Sample Preparation for Metabolomics Study of White Breast Meat
The same dried, homogenous breast muscle samples as used for GC-MS were prepared for NMR metabolomics as described previously by Wagner et al. [18] with slight modifications. In brief, 20 mg of lyophilized, grinded, homogeneous muscle powder was extracted using first ice-cold methanol, then ice-cold chloroform and finally ice-cold water (400 µL of each solvent). The samples were vortexed for 1 min and stored on ice for 10 min between each step, and then stored at 4 ◦C overnight for separation and finally after centrifugation (2000× *g*, 4 ◦C, 30 min) the aqueous phase was collected in a new tube. The collected samples (750 µL) were dried using a vacuum centrifuge. The dried samples were redissolved with 550 µL D2O, 25 µL MilliQ water and 25 µL D2O containing 0.05 wt% TSP as internal standard for quantification and chemical shift reference.
### 2.5.2. <sup>1</sup>H NMR Spectroscopy, Data Processing and Identification of the Signals
All samples were analyzed with a Bruker 400 MHz spectrometer (Bruker BioSpin GmbH, Rheinstetten, Germany). For the aqueous white breast muscle, a noesygppr1d pulse program at 25 ◦C with 64 scans, a spectral width of 8224 Hz collected into 65,536 data point and acquisition time of 3.98 s and an interscan relaxation delay of 4 s was used. <sup>1</sup>H-1H COSY, <sup>1</sup>H-1H TOCSY and <sup>1</sup>H-13C HSQC were obtained on one representative muscle sample for metabolite identification purposes.
All data were processed using Bruker Topspin 3.6.0 software (Bruker), Fourier-transformed after multiplication by line broadening of 0.30 Hz and subsequently referenced to standard peak TSP at 0.00 ppm. After spectral phase and baseline were corrected, each NMR spectrum was integrated using Matlab R2017b (Mathworks, Natick, MA, USA) into 0.01 ppm integral regions (buckets) between 8.60 ppm and 0.80 ppm (area between 4.75 ppm and 4.80 ppm corresponding to the water signal was excluded). Each muscle spectrum region was scaled to the intensity of internal standard (TSP) for quantitative measurements. Afterwards, the signals were identified using ChenomX NMR Suite 8.4 library (ChenomX Inc., Edmonton, AB, Canada), the Human Metabolome Database (www.hmdb.ca) and previous literature [18–20] and confirmed with 2D-NMR in case of multiplicity. For quantification (profiling approach), 86 metabolites were identified by overlapping with standard spectra (Supplementary Table S3) and their concentrations (µmol/mg) were calculated using ChenomX NMR Suite 8.4 library after accounting for overlapping signals. The absolute concentrations were presented as mg/100 g wet weight (Supplementary Table S4).
#### *2.6. Statistical Analysis*
Multivariate data analyses were performed for the GC-MS data, the NMR spectral data (buckets) and the absolute concentrations of the metabolites (profiling approach) using the Simca-P+ software (version 13.0; Umetrics, Umeå, Sweden). All variables were centered and "pareto-scaled" (Par) (GC-MS data and NMR spectral data) or "unit variance" (UV)-scaled (NMR data, absolute concentrations). Principal component analysis (PCA) was used to screen the data and search for outliers. Outliers were determined using PCA-Hotelling T<sup>2</sup> Ellipse (95% confidential interval (CI)).
All statistical calculations such as one-way analysis of variance (ANOVA) and Dunnett's test were done in JMP (13.1.0, SAS Institute Inc., Cary, NC, USA).
The amino acid contents (alanine, leucine, methionine, phenylalanine, proline, serine, tyrosine, histidine, lysine and glutamate) determined by HPLC, GC-MS and NMR are presented as mg/100 g wet weight, henceforth referred to as mg/100 g. All data presented are mean ± standard deviation and differences were considered significant when *p* < 0.01.
#### **3. Results and Discussion**
The aim of this study was to compare different analytical techniques in respect to their performance for the detection of undeclared protein hydrolysates in fresh turkey breast. For this purpose, a traditional HPLC-UV/VIS approach focusing on the detection of free proteinogenic amino acids was compared with two nontargeted metabolic profiling techniques, GC-MS and <sup>1</sup>H-NMR. Additionally, both nontargeted approaches were compared for their suitability in the detection of the adulterated turkey breast muscle with protein hydrolysates.
#### *3.1. Capability of Amino Acid Profiling for the Detection of Added Protein Hydrolysates*
The contents of the ten free amino acids (FAA) alanine, leucine, methionine, phenylalanine, proline, serine, tyrosine, histidine, lysine and glutamate out of the 20 proteogenic AA were analyzed by all methods. Those ten AA were selected as exemplary free amino acids with different properties, e.g., aliphatic (alanine, leucine, proline), aromatic (phenylalanine, tyrosine), acidic (glutamate), basic (lysine, histidine), hydroxylic (serine) and sulfur-containing (methionine). The results of five groups (REF
and addition of water, gelatin-, wheat- and casein-hydrolysates) were compared depending on the hydrolyzation degree (PH or TH).
The contents of FAA for the addition of partial and total protein hydrolysates are shown in Figures 1 and 2, respectively. All these data are also summarized in the Supplementary Table S5. The reference samples (REF) were not modified and variations were therefore only occurring through the analytical error and natural variations of the FAA contents. The natural FAA contents depend on several conditions, e.g., the gender of the birds [21] or special feed additives [22].
It is obvious that the amount of FAA contents determined by HPLC-UV/VIS (in mg/100 g) and GC-MS as well as <sup>1</sup>H-NMR (in mg/100 g) are quite different. The FAA contents determined by GC-MS and <sup>1</sup>H-NMR are on average 3.8 times (range: 0.6-fold to 7.0-fold) and 3.3 times (range: 0.1-fold to 10.2-fold) higher, respectively, compared with the contents determined by HPLC-UV/VIS. The differences can be arising from the different sample preparations. One possibility could be that the homogenization method (HPLC-UV/VIS) was not able to dissolve all FAA. The homogenization method for HPLC did not contain any specific extraction step with solvents, whereas for GC-MS and <sup>1</sup>H-NMR the samples were extracted using mixtures of water, methanol and chloroform. Moreover, it is possible that the extraction method (GC-MS and <sup>1</sup>H-NMR) led to protein hydrolysis. However, the different quantification procedures could also have caused these differences. Comparative experiments (e.g., with protease inhibitors) clearly demonstrated that no protein hydrolysis occurred by using the homogenization method (manuscript in preparation). This method was also used to determinate 18 of the 20 proteinogenic FAA contents of chicken breast meat (manuscript in preparation) and the contents were in agreement with Rikimaru and Takahashi [23].
The addition of water to the turkey breast muscle resulted in tendentially lower contents of FAA, but statistical significance was not reached by using the Dunnett's test (Figures 1 and 2 and Supplementary Table S5). The reduced mean values in water treated samples could be explained as dilution or even wash-out effect. When the amount of water injected to the sample exceeds its water binding capacity, some endogenous compounds (e.g., FAA) might be washed out.
#### 3.1.1. Comparison of Partial Hydrolyzed Wheat-, Gelatin- and Casein-Hydrolysates
Partial enzymatic hydrolysates from gelatin (GPH, hydrolyzation degree 15% ± 3%), wheat (WPH, hydrolyzation degree: 16% ± 2%) and casein (CPH, hydrolyzation degree 53% ± 2%) were added to the meat samples, respectively (Figure 1). WPH and GPH were only slightly hydrolyzed and therefore most of the protein was converted to peptides. Therefore, the amount of FAA in these two hydrolysates was lower compared to CPH, which was hydrolyzed to a higher hydrolyzation degree.
Hence, nearly no significant differences were found for the FAA contents of GPH and WPH related to the REF. Only the content of free lysine (1H-NMR, from 5.26 mg/100 g <sup>±</sup> 0.45 mg/100 g (REF) to 26.9 mg/100 g ± 3.37 mg/100 g) for GPH, as well as the free methionine (HPLC-UV/VIS) and free leucine contents (1H-NMR) for WPH showed significant differences (Supplementary Table S5).
Contrary to this, the CPH showed clearly significant different FAA contents compared to the REF. As determined with HPLC-UV/VIS-method, the FAA contents of leucine, methionine, phenylalanine and histidine were highly significant different (*p* < 0.001) and for serine significant different (*p* < 0.01). For example, the FFA content increased for leucine from 1.97 mg/100 g ± 0.33 mg/100 g (REF) to 19.9 mg/100 g ± 2.39 mg/100 g (CPH). The FAA contents analyzed with GC-MS showed significant differences for five of the ten listed FAA. Leucine was also highly increased (6.97 mg/100 g ± 2.78 mg/100 g for REF to 111 mg/100 g ± 26.4 mg/100 g (CPH), *p* < 0.001), as well as methionine, phenylalanine and lysine. The histidine content was increased significantly (*p* < 0.01). The analysis with <sup>1</sup>H-NMR showed seven increased FAA contents: leucine, methionine, phenylalanine, proline and lysine increased highly (*p* < 0.001), whereas alanine and serine increased significantly (*p* < 0.01). For example, leucine increased from 4.26 mg/100 g ± 1.25 mg/100 g (REF) to 78.6 mg/100 g ± 19.6 mg/100 g (CPH).
**Figure 1.** Free amino acids contents (mean ± standard deviation) of turkey breast meat samples treated with and without the addition of partial protein hydrolysates or water and analyzed via: (**a**) HPLC-UV/VIS (**b**) GC-MS (**c**) 1H-NMR. Sample codes: REF: Reference, Water: injected with water, GPH: partial hydrolysate gelatin, WPH: partial hydrolysate wheat; CPH: partial hydrolysate casein. **Figure 1.** Free amino acids contents (mean ± standard deviation) of turkey breast meat samples treated with and without the addition of partial protein hydrolysates or water and analyzed via: (**a**) HPLC-UV/VIS (**b**) GC-MS (**c**) <sup>1</sup>H-NMR. Sample codes: REF: Reference, Water: injected with water, GPH: partial hydrolysate gelatin, WPH: partial hydrolysate wheat; CPH: partial hydrolysate casein.
**Figure 2.** Free amino acids contents (mean ± standard deviation) of turkey breast meat samples treated with and without the addition of total protein hydrolysates or water and analyzed via: (**a**) HPLC-UV/VIS (**b**) GC-MS (**c**) 1H-NMR. Sample codes: REF: Reference, Water: injected with water, GTH: total hydrolysate gelatin, WTH: total hydrolysate wheat; CTH: total hydrolysate casein. **Figure 2.** Free amino acids contents (mean ± standard deviation) of turkey breast meat samples treated with and without the addition of total protein hydrolysates or water and analyzed via: (**a**) HPLC-UV/VIS (**b**) GC-MS (**c**) <sup>1</sup>H-NMR. Sample codes: REF: Reference, Water: injected with water, GTH: total hydrolysate gelatin, WTH: total hydrolysate wheat; CTH: total hydrolysate casein.
3.1.1. Comparison of Partial Hydrolyzed Wheat-, Gelatin- and Casein-Hydrolysates Partial enzymatic hydrolysates from gelatin (GPH, hydrolyzation degree 15% ± 3%), wheat (WPH, hydrolyzation degree: 16% ± 2%) and casein (CPH, hydrolyzation degree 53% ± 2%) were added to the meat samples, respectively (Figure 1). WPH and GPH were only slightly hydrolyzed and therefore most For a clear proof, several FAA contents should differ significantly from the reference sample. In this way other reasons for different FAA contents (e.g., feed supplementation with AA) can be excluded with higher probability. It can be concluded in this study, that in case of partial hydrolysate treatment the detection of fraud could not be ensured by using only these ten FAA contents.
lower compared to CPH, which was hydrolyzed to a higher hydrolyzation degree.
of the protein was converted to peptides. Therefore, the amount of FAA in these two hydrolysates was
#### 3.1.2. Comparison of Total Hydrolyzed Wheat-, Gelatin- and Casein-Hydrolysates
All the total hydrolyzed proteins (GTH, WTH, CTH) showed a hydrolyzation degree of 100% and were therefore only composed of AA. As expected, the FAA contents of all hydrolysate-treated samples were increased dramatically compared to the REF (Figure 2). The analysis with HPLC-UV/VIS revealed two significant and seven highly significant increases of FAA contents for GTH, eight highly significant increased FAA contents for WTH and nine rises of the amounts of FAA for CTH (*p* < 0.001). For example, the content of proline changed from 2.56 mg/100 g ± 0.42 mg/100 g (REF) to 52.5 mg/100 g ± 20.4 mg/100 g (GTH), 44.9 mg/100 g ± 16.8 mg/100 g (WTH) and 39.2 mg/100 g ± 13.8 mg/100 g (CTH), respectively. With GC-MS, only highly significant changes were found: for GTH seven, for WTH nine and for CTH all ten FAA contents were increased. As for HPLC-UV/VIS, the proline content was raised obviously: from 10.0 mg/100 g ± 5.81 mg/100 g to 134 mg/100 g ± 71.4 mg/100 g (GTH), 161 mg/100 g ± 31.8 mg/100 g (WTH) and 160 mg/100 g ± 30.5 mg/100 g (CTH), respectively. The same results were also found for the <sup>1</sup>H-NMR analysis: one significant and eight highly significant increases for GTH (proline: from 5.06 mg/100 g ± 0.78 mg/100 g (REF) to 77.5 mg/100 g ± 16.3 mg/100 g), and nine highly significant differences for WTH (proline: 72.8 mg/100 g ± 12.1 mg/100 g) and CTH (proline: 53.8 mg/100 g ± 16.1 mg/100 g). Therefore, only tyrosine (GTH), alanine and lysine (WTH) as well as alanine (CTH) showed no significant differences determined by HPLC-UV/VIS. For the GC-MS-method, only methionine, tyrosine, histidine (GTH) and lysine (WTH) were not significantly different. The analysis by <sup>1</sup>H-NMR revealed also nearly exclusive significant differences with only few exceptions (tyrosine, histidine for GTH, histidine for WTH and CTH).
Depending on the hydrolysate type, different AA were more affected. The addition of GTH resulted in higher levels of alanine, whereas WTH showed higher levels of glutamate and CTH higher levels of leucine, methionine, tyrosine and lysine (Figure 2 and Supplementary Table S5). The latter AA might be used as an indicator for animal-based protein origins whereas glutamate could indicate plant-based protein origins.
It was shown in this study that in case of total hydrolysate treatment a general detection of fraud is possible.
#### 3.1.3. General Aspects Regarding the Detection of Free Amino Acids in Treated Breast Muscles
All three methods used (HPLC-UV/VIS, GC-MS, <sup>1</sup>H-NMR) showed comparable results. Although the FAA contents of the first method were about three- to fourfold lower compared to the other two methods, the validity was given. This is due to the fact that all samples were compared to the corresponding REF, determined with the same method. Hence, the method of sample preparation plays an important role for absolute quantities, whereas regarding the differentiation between REF and hydrolyzed-treated samples (rations are kept independently of the sample preparation) the method has no impact.
It can be concluded that the differentiation of hydrolysate addition depends on the degree of hydrolyzation. If breast muscles were treated with low degree hydrolysates, the additional injected FAA might not induce a significant increase over the range of natural variation. It was shown that a high hydrolyzation degree significantly increased the free AA content of several AA independently of which analytical technique (HPLC-UV/VIS, GC-MS or <sup>1</sup>H-NMR) was used.
Specific FAA profiles might be used for a tentative classification of the origin of the hydrolysate type (e.g., plant-based vs. animal-based protein origins). Nevertheless, a clear classification and identification of the protein used for hydrolyzation was not possible. Thus, it is of interest whether further information about additional compounds might be helpful for the detection and classification of the hydrolysates. For this, the following hypotheses for section two were postulated: (1) Original protein sources are not clean and contain additional compounds, which can be introduced into the breast meat. (2) Acidic hydrolysis leads to formation of byproducts and these compounds are also possible to be found in the breast meat. (3) Additional metabolites can be washed out from the breast meat and (4) therefore, information from metabolite profiling might be of interest and was included in the analysis.
### *3.2. Metabolomics Approaches to Obtain Additional Information Regarding Hydrolysate-Treated Samples Independently of the Hydrolyzation Degree*
The detection of hydrolysate treatment in turkey breast muscle by amino acid profiling largely depends on the hydrolyzation degree. Our results clearly indicated that an addition of total hydrolyzation increases the free amino acid content tremendously (Figure 2) so that a detection with all three presented methods was possible. However, the lower the hydrolyzation degree the more uncertain is the validity of amino acid profiles between the natural variation and the differentiation due to hydrolysate treatment. Therefore, we applied two nontargeted metabolite profiling approaches (GC-MS and <sup>1</sup>H-NMR) to test for their suitability in the detection of hydrolysate treatment in turkey breast muscle. Both approaches allow to detect additionally several metabolites besides the proteogenic amino acids like carbohydrates, organic acids, lipids, et cetera. For both techniques PCA was used to check for the variation of metabolite profiles between controls and treatments and between the different types of hydrolysates (Figure 3). *Foods* **2020**, *9*, x FOR PEER REVIEW 12 of 16 *Foods* **2020**, *9*, x FOR PEER REVIEW 12 of 16 *Foods* **2020**, *9*, x FOR PEER REVIEW 12 of 16 *Foods* **2020**, *9*, x FOR PEER REVIEW 12 of 16 *Foods* **2020**, *9*, x FOR PEER REVIEW 12 of 16 *Foods* **2020**, *9*, x FOR PEER REVIEW 12 of 16 *Foods* **2020**, *9*, x FOR PEER REVIEW 12 of 16 *Foods* **2020**, *9*, x FOR PEER REVIEW 12 of 16 *Foods* **2020**, *9*, x FOR PEER REVIEW 12 of 16
**Figure 3.** The score (**a**,**b**) and the loading (**c**,**d**) plots of all data of the principal component analysis (PCA) based on the metabolic fingerprint from adulated turkey breast meat with different protein hydrolysates (type and degrees) analyzed by GC-MS (**a**,**c**) and 1H-NMR (**b**,**d**). REF ( ), Water ( ), GPH ( ), WPH ( ), CPH ( ), GTH (), WTH ( ) and CTH ( ). GC-MS: The first component is explained by 38.0% and the second component by 15.7% of the variation (model parameters: R2X = 98.4%, Q2 = 72.8%, 17 components). 1H-NMR: The first and second components explained 40.8% and 21.9% of variation, **Figure 3.** The score (**a**,**b**) and the loading (**c**,**d**) plots of all data of the principal component analysis (PCA) based on the metabolic fingerprint from adulated turkey breast meat with different protein hydrolysates (type and degrees) analyzed by GC-MS (**a**,**c**) and <sup>1</sup>H-NMR (**b**,**d**). REF ( **Figure 3.** The score (**a**,**b**) and the loading (**c**,**d**) plots of all data of the principal component analysis (PCA) based on the metabolic fingerprint from adulated turkey breast meat with different protein hydrolysates (type and degrees) analyzed by GC-MS (**a**,**c**) and 1H-NMR (**b**,**d**). REF ( ), Water ( ), GPH ( ), WPH ( ), CPH ( ), GTH (), WTH ( ) and CTH ( ). GC-MS: The first component is explained by 38.0% and the second component by 15.7% of the variation (model parameters: R2X = 98.4%, Q2 = 72.8%, 17 components). 1H-NMR: The first and second components explained 40.8% and 21.9% of variation, respectively (model parameters: R2X = 98.5%, Q2 = 93.4%, 16 components). ), Water ( **Figure 3.** The score (**a**,**b**) and the loading (**c**,**d**) plots of all data of the principal component analysis (PCA) based on the metabolic fingerprint from adulated turkey breast meat with different protein hydrolysates (type and degrees) analyzed by GC-MS (**a**,**c**) and 1H-NMR (**b**,**d**). REF ( ), Water ( ), GPH ( ), WPH ( ), CPH ( ), GTH (), WTH ( ) and CTH ( ). GC-MS: The first component is explained by 38.0% and the second component by 15.7% of the variation (model parameters: R2X = 98.4%, Q2 = 72.8%, 17 components). 1H-NMR: The first and second components explained 40.8% and 21.9% of variation, respectively (model parameters: R2X = 98.5%, Q2 = 93.4%, 16 components). ), GPH ( **Figure 3.** The score (**a**,**b**) and the loading (**c**,**d**) plots of all data of the principal component analysis (PCA) based on the metabolic fingerprint from adulated turkey breast meat with different protein hydrolysates (type and degrees) analyzed by GC-MS (**a**,**c**) and 1H-NMR (**b**,**d**). REF ( ), Water ( ), GPH ( ), WPH ( ), CPH ( ), GTH (), WTH ( ) and CTH ( ). GC-MS: The first component is explained by 38.0% and the second component by 15.7% of the variation (model parameters: R2X = 98.4%, Q2 = 72.8%, 17 components). 1H-NMR: The first and second components explained 40.8% and 21.9% of variation, respectively (model parameters: R2X = 98.5%, Q2 = 93.4%, 16 components). ), WPH ( **Figure 3.** The score (**a**,**b**) and the loading (**c**,**d**) plots of all data of the principal component analysis (PCA) based on the metabolic fingerprint from adulated turkey breast meat with different protein hydrolysates (type and degrees) analyzed by GC-MS (**a**,**c**) and 1H-NMR (**b**,**d**). REF ( ), Water ( ), GPH ( ), WPH ( ), CPH ( ), GTH (), WTH ( ) and CTH ( ). GC-MS: The first component is explained by 38.0% and the second component by 15.7% of the variation (model parameters: R2X = 98.4%, Q2 = 72.8%, 17 components). 1H-NMR: The first and second components explained 40.8% and 21.9% of variation, respectively (model parameters: R2X = 98.5%, Q2 = 93.4%, 16 components). ), CPH ( **Figure 3.** The score (**a**,**b**) and the loading (**c**,**d**) plots of all data of the principal component analysis (PCA) based on the metabolic fingerprint from adulated turkey breast meat with different protein hydrolysates (type and degrees) analyzed by GC-MS (**a**,**c**) and 1H-NMR (**b**,**d**). REF ( ), Water ( ), GPH ( ), WPH ( ), CPH ( ), GTH (), WTH ( ) and CTH ( ). GC-MS: The first component is explained by 38.0% and the second component by 15.7% of the variation (model parameters: R2X = 98.4%, Q2 = 72.8%, 17 components). 1H-NMR: The first and second components explained 40.8% and 21.9% of variation, respectively (model parameters: R2X = 98.5%, Q2 = 93.4%, 16 components). ), GTH ( **Figure 3.** The score (**a,b**) and the loading (**c,d**) plots of all data of the principal component analysis (PCA) based on the metabolic fingerprint from adulated turkey breast meat with different protein hydrolysates (type and degrees) analyzed by GC-MS (**a,c**) and 1H-NMR (**b,d**). REF ( ), Water ( ), GPH ( ), WPH ( ), CPH ( ), GTH ( ), WTH ( ) and CTH ( ). GC-MS: The first component is explained by 38.0% and the second component by 15.7% of the variation (model parameters: R2X = 98.4%, Q2 = 72.8%, 17 components). 1H-NMR: The first and second components explained 40.8% and 21.9% of variation, respectively (model parameters: R2X = 98.5%, Q2 = 93.4%, 16 components). ), WTH ( **Figure 3.** The score (**a**,**b**) and the loading (**c**,**d**) plots of all data of the principal component analysis (PCA) based on the metabolic fingerprint from adulated turkey breast meat with different protein hydrolysates (type and degrees) analyzed by GC-MS (**a**,**c**) and 1H-NMR (**b**,**d**). REF ( ), Water ( ), GPH ( ), WPH ( ), CPH ( ), GTH (), WTH ( ) and CTH ( ). GC-MS: The first component is explained by 38.0% and the second component by 15.7% of the variation (model parameters: R2X = 98.4%, Q2 = 72.8%, 17 components). 1H-NMR: The first and second components explained 40.8% and 21.9% of variation, respectively (model parameters: R2X = 98.5%, Q2 = 93.4%, 16 components). ) and CTH ( **Figure 3.** The score (**a**,**b**) and the loading (**c**,**d**) plots of all data of the principal component analysis (PCA) based on the metabolic fingerprint from adulated turkey breast meat with different protein hydrolysates (type and degrees) analyzed by GC-MS (**a**,**c**) and 1H-NMR (**b**,**d**). REF ( ), Water ( ), GPH ( ), WPH ( ), CPH ( ), GTH (), WTH ( ) and CTH ( ). GC-MS: The first component is explained by 38.0% and the second component by 15.7% of the variation (model parameters: R2X = 98.4%, Q2 = 72.8%, 17 components). 1H-NMR: The first and second components explained 40.8% and 21.9% of variation, respectively (model parameters: R2X = 98.5%, Q2 = 93.4%, 16 components). ). GC-MS: The first component is explained by 38.0% and the second component by 15.7% of the variation (model parameters: R <sup>2</sup>X = 98.4%, Q<sup>2</sup> = 72.8%, 17 components). <sup>1</sup>H-NMR: The first and second components explained 40.8% and 21.9% of variation, respectively (model parameters: R2X = 98.5%, Q<sup>2</sup> = 93.4%, 16 components).
respectively (model parameters: R2X = 98.5%, Q2 = 93.4%, 16 components). The total hydrolysate treated samples clearly separated from the controls in PC1 observed with both techniques. Interestingly, with GC-MS analysis the wheat and casein origins showed higher similarities to each other compared to gelatin (GTH). Whereas for NMR analyses, higher similarities were observed between GTH and WTH. Obviously, independently of the analytical technique, there is a clear separation between gelatin (GTH) and casein (CTH). From the loading plot (Figure 3c,d), it can be deduced that proteinogenic amino acids particularly contribute to the differentiation of the total The total hydrolysate treated samples clearly separated from the controls in PC1 observed with both techniques. Interestingly, with GC-MS analysis the wheat and casein origins showed higher similarities to each other compared to gelatin (GTH). Whereas for NMR analyses, higher similarities were observed between GTH and WTH. Obviously, independently of the analytical technique, there is a clear separation between gelatin (GTH) and casein (CTH). From the loading plot (Figure 3c,d), it can be deduced that proteinogenic amino acids particularly contribute to the differentiation of the total hydrolysate treated samples and the controls (in PC1), which is in accordance to the results presented The total hydrolysate treated samples clearly separated from the controls in PC1 observed with both techniques. Interestingly, with GC-MS analysis the wheat and casein origins showed higher similarities to each other compared to gelatin (GTH). Whereas for NMR analyses, higher similarities were observed between GTH and WTH. Obviously, independently of the analytical technique, there is a clear separation between gelatin (GTH) and casein (CTH). From the loading plot (Figure 3c,d), it can be deduced that proteinogenic amino acids particularly contribute to the differentiation of the total hydrolysate treated samples and the controls (in PC1), which is in accordance to the results presented The total hydrolysate treated samples clearly separated from the controls in PC1 observed with both techniques. Interestingly, with GC-MS analysis the wheat and casein origins showed higher similarities to each other compared to gelatin (GTH). Whereas for NMR analyses, higher similarities were observed between GTH and WTH. Obviously, independently of the analytical technique, there is a clear separation between gelatin (GTH) and casein (CTH). From the loading plot (Figure 3c,d), it can be deduced that proteinogenic amino acids particularly contribute to the differentiation of the total hydrolysate treated samples and the controls (in PC1), which is in accordance to the results presented The total hydrolysate treated samples clearly separated from the controls in PC1 observed with both techniques. Interestingly, with GC-MS analysis the wheat and casein origins showed higher similarities to each other compared to gelatin (GTH). Whereas for NMR analyses, higher similarities were observed between GTH and WTH. Obviously, independently of the analytical technique, there is a clear separation between gelatin (GTH) and casein (CTH). From the loading plot (Figure 3c,d), it can be deduced that proteinogenic amino acids particularly contribute to the differentiation of the total hydrolysate treated samples and the controls (in PC1), which is in accordance to the results presented The total hydrolysate treated samples clearly separated from the controls in PC1 observed with both techniques. Interestingly, with GC-MS analysis the wheat and casein origins showed higher similarities to each other compared to gelatin (GTH). Whereas for NMR analyses, higher similarities were observed between GTH and WTH. Obviously, independently of the analytical technique, there is a clear separation between gelatin (GTH) and casein (CTH). From the loading plot (Figure 3c,d), it can be deduced that proteinogenic amino acids particularly contribute to the differentiation of the total hydrolysate treated samples and the controls (in PC1), which is in accordance to the results presented The total hydrolysate treated samples clearly separated from the controls in PC1 observed with both techniques. Interestingly, with GC-MS analysis the wheat and casein origins showed higher similarities to each other compared to gelatin (GTH). Whereas for NMR analyses, higher similarities were observed between GTH and WTH. Obviously, independently of the analytical technique, there is a clear separation between gelatin (GTH) and casein (CTH). From the loading plot (Figure 3c,d), it can be deduced that proteinogenic amino acids particularly contribute to the differentiation of the total hydrolysate treated samples and the controls (in PC1), which is in accordance to the results presented The total hydrolysate treated samples clearly separated from the controls in PC1 observed with both techniques. Interestingly, with GC-MS analysis the wheat and casein origins showed higher similarities to each other compared to gelatin (GTH). Whereas for NMR analyses, higher similarities were observed between GTH and WTH. Obviously, independently of the analytical technique, there is a clear separation between gelatin (GTH) and casein (CTH). From the loading plot (Figure 3c,d), it can be deduced that proteinogenic amino acids particularly contribute to the differentiation of the total hydrolysate treated samples and the controls (in PC1), which is in accordance to the results presented The total hydrolysate treated samples clearly separated from the controls in PC1 observed with both techniques. Interestingly, with GC-MS analysis the wheat and casein origins showed higher similarities to each other compared to gelatin (GTH). Whereas for NMR analyses, higher similarities were observed between GTH and WTH. Obviously, independently of the analytical technique, there is a clear separation between gelatin (GTH) and casein (CTH). From the loading plot (Figure 3c,d), it can be deduced that proteinogenic amino acids particularly contribute to the differentiation of the total hydrolysate treated samples and the controls (in PC1), which is in accordance to the results presented With GC-MS a total of 129 features were considered for PCA. The first (horizontal) and second (vertical) components explained 38.0% and 15.7% of the variation, respectively, with R2X = 98.4%, Q<sup>2</sup> = 72.8% (Figure 3a). A clear separation between the controls (REF and water-treated control) and five of the hydrolysate-treated sample groups was observed. GPH did not segregate from the controls. WPH varied only in PC2 direction whereas CPH, with 53% hydrolyzation degree, stronger differentiated in PC1 direction. The total hydrolysate treated sample groups showed most variation in PC1 direction.
were 5-hydroxylysine, 3-MCPD or aminomalonic acid among several nonidentified molecular features. These additional compounds represent characteristics of the protein origin or are byproducts formed during the acidic hydrolyzation process. In addition to the amino acid profiles these byproducts might
The wheat protein source contained higher amounts of sugars, which was also observed for the WPH treated breast samples (see below). During the acidic hydrolysis of the protein source, the sugars contained therein such as maltose, saccharose, glucose or fructose are converted to levulinic acid in presence of hydrochloride and under high temperature [24,25]. Thus, the high levels of levulinic acid detected in our analyses could be used to differentiate plant-based hydrolysates and as a marker for acidic hydrolyzation treatment. Nevertheless, other plant-based hydrolysates need to be tested for their
The wheat protein source contained higher amounts of sugars, which was also observed for the WPH treated breast samples (see below). During the acidic hydrolysis of the protein source, the sugars contained therein such as maltose, saccharose, glucose or fructose are converted to levulinic acid in presence of hydrochloride and under high temperature [24,25]. Thus, the high levels of levulinic acid detected in our analyses could be used to differentiate plant-based hydrolysates and as a marker for acidic hydrolyzation treatment. Nevertheless, other plant-based hydrolysates need to be tested for their
The wheat protein source contained higher amounts of sugars, which was also observed for the WPH treated breast samples (see below). During the acidic hydrolysis of the protein source, the sugars contained therein such as maltose, saccharose, glucose or fructose are converted to levulinic acid in presence of hydrochloride and under high temperature [24,25]. Thus, the high levels of levulinic acid detected in our analyses could be used to differentiate plant-based hydrolysates and as a marker for acidic hydrolyzation treatment. Nevertheless, other plant-based hydrolysates need to be tested for their
The wheat protein source contained higher amounts of sugars, which was also observed for the WPH treated breast samples (see below). During the acidic hydrolysis of the protein source, the sugars contained therein such as maltose, saccharose, glucose or fructose are converted to levulinic acid in presence of hydrochloride and under high temperature [24,25]. Thus, the high levels of levulinic acid detected in our analyses could be used to differentiate plant-based hydrolysates and as a marker for acidic hydrolyzation treatment. Nevertheless, other plant-based hydrolysates need to be tested for their
The wheat protein source contained higher amounts of sugars, which was also observed for the WPH treated breast samples (see below). During the acidic hydrolysis of the protein source, the sugars contained therein such as maltose, saccharose, glucose or fructose are converted to levulinic acid in presence of hydrochloride and under high temperature [24,25]. Thus, the high levels of levulinic acid detected in our analyses could be used to differentiate plant-based hydrolysates and as a marker for acidic hydrolyzation treatment. Nevertheless, other plant-based hydrolysates need to be tested for their
The wheat protein source contained higher amounts of sugars, which was also observed for the WPH treated breast samples (see below). During the acidic hydrolysis of the protein source, the sugars contained therein such as maltose, saccharose, glucose or fructose are converted to levulinic acid in presence of hydrochloride and under high temperature [24,25]. Thus, the high levels of levulinic acid detected in our analyses could be used to differentiate plant-based hydrolysates and as a marker for acidic hydrolyzation treatment. Nevertheless, other plant-based hydrolysates need to be tested for their
The wheat protein source contained higher amounts of sugars, which was also observed for the WPH treated breast samples (see below). During the acidic hydrolysis of the protein source, the sugars contained therein such as maltose, saccharose, glucose or fructose are converted to levulinic acid in presence of hydrochloride and under high temperature [24,25]. Thus, the high levels of levulinic acid detected in our analyses could be used to differentiate plant-based hydrolysates and as a marker for acidic hydrolyzation treatment. Nevertheless, other plant-based hydrolysates need to be tested for their
The wheat protein source contained higher amounts of sugars, which was also observed for the WPH treated breast samples (see below). During the acidic hydrolysis of the protein source, the sugars contained therein such as maltose, saccharose, glucose or fructose are converted to levulinic acid in presence of hydrochloride and under high temperature [24,25]. Thus, the high levels of levulinic acid detected in our analyses could be used to differentiate plant-based hydrolysates and as a marker for acidic hydrolyzation treatment. Nevertheless, other plant-based hydrolysates need to be tested for their
These additional compounds represent characteristics of the protein origin or are byproducts formed during the acidic hydrolyzation process. In addition to the amino acid profiles these byproducts might
These additional compounds represent characteristics of the protein origin or are byproducts formed during the acidic hydrolyzation process. In addition to the amino acid profiles these byproducts might
These additional compounds represent characteristics of the protein origin or are byproducts formed during the acidic hydrolyzation process. In addition to the amino acid profiles these byproducts might
These additional compounds represent characteristics of the protein origin or are byproducts formed during the acidic hydrolyzation process. In addition to the amino acid profiles these byproducts might
The wheat protein source contained higher amounts of sugars, which was also observed for the WPH treated breast samples (see below). During the acidic hydrolysis of the protein source, the sugars contained therein such as maltose, saccharose, glucose or fructose are converted to levulinic acid in presence of hydrochloride and under high temperature [24,25]. Thus, the high levels of levulinic acid detected in our analyses could be used to differentiate plant-based hydrolysates and as a marker for acidic hydrolyzation treatment. Nevertheless, other plant-based hydrolysates need to be tested for their
The total protein hydrolysate from gelatin contained higher amounts of AA derivates such as hydroxyproline and hydroxylysine. Gelatin is the denatured form of collagen, which is one of the most abundant proteins in meat, ranging between 2 and 4 mg/g in chicken breast meat [26]. Most abundant amino acids of collagen are glycine, proline, glutamate and hydroxyproline [27]. Hydroxyproline is
The total protein hydrolysate from gelatin contained higher amounts of AA derivates such as hydroxyproline and hydroxylysine. Gelatin is the denatured form of collagen, which is one of the most abundant proteins in meat, ranging between 2 and 4 mg/g in chicken breast meat [26]. Most abundant amino acids of collagen are glycine, proline, glutamate and hydroxyproline [27]. Hydroxyproline is
The total protein hydrolysate from gelatin contained higher amounts of AA derivates such as hydroxyproline and hydroxylysine. Gelatin is the denatured form of collagen, which is one of the most abundant proteins in meat, ranging between 2 and 4 mg/g in chicken breast meat [26]. Most abundant amino acids of collagen are glycine, proline, glutamate and hydroxyproline [27]. Hydroxyproline is
The total protein hydrolysate from gelatin contained higher amounts of AA derivates such as hydroxyproline and hydroxylysine. Gelatin is the denatured form of collagen, which is one of the most abundant proteins in meat, ranging between 2 and 4 mg/g in chicken breast meat [26]. Most abundant amino acids of collagen are glycine, proline, glutamate and hydroxyproline [27]. Hydroxyproline is
The total protein hydrolysate from gelatin contained higher amounts of AA derivates such as hydroxyproline and hydroxylysine. Gelatin is the denatured form of collagen, which is one of the most abundant proteins in meat, ranging between 2 and 4 mg/g in chicken breast meat [26]. Most abundant amino acids of collagen are glycine, proline, glutamate and hydroxyproline [27]. Hydroxyproline is
The total protein hydrolysate from gelatin contained higher amounts of AA derivates such as hydroxyproline and hydroxylysine. Gelatin is the denatured form of collagen, which is one of the most abundant proteins in meat, ranging between 2 and 4 mg/g in chicken breast meat [26]. Most abundant amino acids of collagen are glycine, proline, glutamate and hydroxyproline [27]. Hydroxyproline is
The total protein hydrolysate from gelatin contained higher amounts of AA derivates such as hydroxyproline and hydroxylysine. Gelatin is the denatured form of collagen, which is one of the most abundant proteins in meat, ranging between 2 and 4 mg/g in chicken breast meat [26]. Most abundant amino acids of collagen are glycine, proline, glutamate and hydroxyproline [27]. Hydroxyproline is
The total protein hydrolysate from gelatin contained higher amounts of AA derivates such as hydroxyproline and hydroxylysine. Gelatin is the denatured form of collagen, which is one of the most abundant proteins in meat, ranging between 2 and 4 mg/g in chicken breast meat [26]. Most abundant amino acids of collagen are glycine, proline, glutamate and hydroxyproline [27]. Hydroxyproline is
The total protein hydrolysate from gelatin contained higher amounts of AA derivates such as hydroxyproline and hydroxylysine. Gelatin is the denatured form of collagen, which is one of the most abundant proteins in meat, ranging between 2 and 4 mg/g in chicken breast meat [26]. Most abundant amino acids of collagen are glycine, proline, glutamate and hydroxyproline [27]. Hydroxyproline is
contribute to a better classification of the protein sources.
contribute to a better classification of the protein sources.
contribute to a better classification of the protein sources.
contribute to a better classification of the protein sources.
contribute to a better classification of the protein sources.
contribute to a better classification of the protein sources.
contribute to a better classification of the protein sources.
contribute to a better classification of the protein sources.
contribute to a better classification of the protein sources.
carbohydrate content in contrast to animal-based protein sources.
carbohydrate content in contrast to animal-based protein sources.
carbohydrate content in contrast to animal-based protein sources.
carbohydrate content in contrast to animal-based protein sources.
carbohydrate content in contrast to animal-based protein sources.
carbohydrate content in contrast to animal-based protein sources.
carbohydrate content in contrast to animal-based protein sources.
carbohydrate content in contrast to animal-based protein sources.
carbohydrate content in contrast to animal-based protein sources.
hydrolysate treated samples and the controls (in PC1), which is in accordance to the results presented in 3.1. Nevertheless, besides the amino acids other compounds could be identified which play an additional role for the variation in PC1 such as hydroxyproline, levulinic acid, ornithine or glycerol (the
in 3.1. Nevertheless, besides the amino acids other compounds could be identified which play an additional role for the variation in PC1 such as hydroxyproline, levulinic acid, ornithine or glycerol (the complete feature tables are presented in Supplementary Tables S3 and S4). Compounds, which were
in 3.1. Nevertheless, besides the amino acids other compounds could be identified which play an additional role for the variation in PC1 such as hydroxyproline, levulinic acid, ornithine or glycerol (the complete feature tables are presented in Supplementary Tables S3 and S4). Compounds, which were
in 3.1. Nevertheless, besides the amino acids other compounds could be identified which play an additional role for the variation in PC1 such as hydroxyproline, levulinic acid, ornithine or glycerol (the complete feature tables are presented in Supplementary Tables S3 and S4). Compounds, which were
in 3.1. Nevertheless, besides the amino acids other compounds could be identified which play an additional role for the variation in PC1 such as hydroxyproline, levulinic acid, ornithine or glycerol (the complete feature tables are presented in Supplementary Tables S3 and S4). Compounds, which were
in 3.1. Nevertheless, besides the amino acids other compounds could be identified which play an additional role for the variation in PC1 such as hydroxyproline, levulinic acid, ornithine or glycerol (the complete feature tables are presented in Supplementary Tables S3 and S4). Compounds, which were
in 3.1. Nevertheless, besides the amino acids other compounds could be identified which play an additional role for the variation in PC1 such as hydroxyproline, levulinic acid, ornithine or glycerol (the complete feature tables are presented in Supplementary Tables S3 and S4). Compounds, which were
in 3.1. Nevertheless, besides the amino acids other compounds could be identified which play an additional role for the variation in PC1 such as hydroxyproline, levulinic acid, ornithine or glycerol (the complete feature tables are presented in Supplementary Tables S3 and S4). Compounds, which were
in 3.1. Nevertheless, besides the amino acids other compounds could be identified which play an additional role for the variation in PC1 such as hydroxyproline, levulinic acid, ornithine or glycerol (the complete feature tables are presented in Supplementary Tables S3 and S4). Compounds, which were
were 5-hydroxylysine, 3-MCPD or aminomalonic acid among several nonidentified molecular features. These additional compounds represent characteristics of the protein origin or are byproducts formed during the acidic hydrolyzation process. In addition to the amino acid profiles these byproducts might
were 5-hydroxylysine, 3-MCPD or aminomalonic acid among several nonidentified molecular features. These additional compounds represent characteristics of the protein origin or are byproducts formed during the acidic hydrolyzation process. In addition to the amino acid profiles these byproducts might
were 5-hydroxylysine, 3-MCPD or aminomalonic acid among several nonidentified molecular features. These additional compounds represent characteristics of the protein origin or are byproducts formed during the acidic hydrolyzation process. In addition to the amino acid profiles these byproducts might
These additional compounds represent characteristics of the protein origin or are byproducts formed during the acidic hydrolyzation process. In addition to the amino acid profiles these byproducts might
In addition, the <sup>1</sup>H-NMR spectra obtained were compared by PCA (PC1 vs. PC2) (Figure 3b). The first component (horizontal) which is explained by 40.8% of spectral variation clearly separates controls (REF and water-treated control, left) with total hydrolyzed-treated samples (right) and partial hydrolyzed-treated samples (middle). The second component explained 21.9% of variation and separates the controls and total hydrolyzed-treated samples (top) from partial hydrolyzed (bottom) samples. The model parameters were the following: R2X = 98.5%, Q<sup>2</sup> = 93.4%, 16 components. In order to identify metabolic changes, the absolute concentrations of 86 metabolites were quantified through a profiling approach from <sup>1</sup>H-NMR spectra.
The total hydrolysate treated samples clearly separated from the controls in PC1 observed with both techniques. Interestingly, with GC-MS analysis the wheat and casein origins showed higher similarities to each other compared to gelatin (GTH). Whereas for NMR analyses, higher similarities were observed between GTH and WTH. Obviously, independently of the analytical technique, there is a clear separation between gelatin (GTH) and casein (CTH). From the loading plot (Figure 3c,d), it can be deduced that proteinogenic amino acids particularly contribute to the differentiation of the total hydrolysate treated samples and the controls (in PC1), which is in accordance to the results presented in Section 3.1. Nevertheless, besides the amino acids other compounds could be identified which play an additional role for the variation in PC1 such as hydroxyproline, levulinic acid, ornithine or glycerol (the complete feature tables are presented in Supplementary Tables S3 and S4). Compounds, which were additionally detected by <sup>1</sup>H-NMR were pyruvate and acetate. Further compounds detected by GC-MS were 5-hydroxylysine, 3-MCPD or aminomalonic acid among several nonidentified molecular features. These additional compounds represent characteristics of the protein origin or are byproducts formed during the acidic hydrolyzation process. In addition to the amino acid profiles these byproducts might contribute to a better classification of the protein sources.
The wheat protein source contained higher amounts of sugars, which was also observed for the WPH treated breast samples (see below). During the acidic hydrolysis of the protein source, the sugars contained therein such as maltose, saccharose, glucose or fructose are converted to levulinic acid in presence of hydrochloride and under high temperature [24,25]. Thus, the high levels of levulinic acid detected in our analyses could be used to differentiate plant-based hydrolysates and as a marker for acidic hydrolyzation treatment. Nevertheless, other plant-based hydrolysates need to be tested for their carbohydrate content in contrast to animal-based protein sources.
The total protein hydrolysate from gelatin contained higher amounts of AA derivates such as hydroxyproline and hydroxylysine. Gelatin is the denatured form of collagen, which is one of the most abundant proteins in meat, ranging between 2 and 4 mg/g in chicken breast meat [26]. Most abundant amino acids of collagen are glycine, proline, glutamate and hydroxyproline [27]. Hydroxyproline is specific to collagen and its concentration in collagen is rather constant with ~12% [28]. Therefore, hydroxyproline is used to estimate the connective tissue content [29,30]. Regarding the treatment of turkey breast meat with protein hydrolysates, the hydroxyproline content can serve as a marker for animal-based protein sources such as gelatin. Hydroxylysine is another modified amino acid, which is unique to collagens. Similar to hydroxyproline, this amino acid becomes posttranslational hydroxylated and subsequently glycosylated forming the α-helical structure of collagens [31]. Therefore, 5-hydroxylysine might serve as an additional indicator of gelatin hydrolysate treatment. Aminomalonic acid that was most abundant in GTH followed by CTH treated samples represents an amino acid derivative, whose origin is suspected to be related to protein oxidation processes [32] and to play a role in the serine-glycine interconversion [33]. According to our results, the acidic hydrolysis process might increase the formation of aminomalonic acid in dependence of the protein source, namely the glycine-rich gelatin.
For casein, the second tested animal derived hydrolysate, it was not that a particular molecule was strongly increased, but the combination of several molecular features could hint towards this treatment. In addition to the amino acid profile, the casein treated samples had higher levels of 3-MCPD and a number of not-identified molecular features (Supplementary Table S1).
out.
**4. Conclusions**
were easily proofed within this study.
Interestingly, the acidic treatment of the different protein sources led to the formation of 3-MCPD. This compound can be found in numerous foodstuffs and is described to be present in acidic hydrolysates of proteins [34]. Depending on the remaining lipids in the original protein sources, different amounts of 3-MCPD and 3-MCPD fatty acid esters might be formed and injected into the breast meat. Thus, 3-MCPD represents an additional marker for acidic hydrolysis, similar to levulinic acid.
From the score plots of the partial hydrolysate treated breast muscles, we observed a clear separation of all three sample groups using <sup>1</sup>H-NMR technique, whereas by GC-MS analysis the GPH group largely overlapped with the controls. The WPH treated samples were in a medium distance to the controls, and the highest variation to the controls was observed for CPH treated samples. Those observations are in accordance with the different degrees of hydrolyzation in the partial hydrolysates with GPH having a hydrolyzation degree of 15% and CPH having a hydrolyzation degree of 53%. Even though WPH has a hydrolyzation degree of only 16%, the better separation compared to GPH might be explained by the plant-based origin and the present additional metabolites.
A closer look at the loading plots from the PCA models for control samples and partial hydrolysate treated samples indicated that proteinogenic amino acids play a minor role for the variation between controls (and GPH) and WPH, both having a low hydrolyzation degree (15% and 16%, respectively). The variation of CPH from controls, which was to 53% hydrolyzed, was already dominated by proteinogenic amino acids. Especially the plant-based hydrolysate contained additional sugars such as maltose (Figure 4) and hexoses like glucose, detected with both technical approaches. In addition, higher levels of glycerol were detected in WPH. As mentioned above, future studies will have to elucidate to which extent different sugars are present in plant-based protein extracts used for hydrolysis. With the animal-based protein hydrolysates, GPH and CPH, the contents of ornithine (Figure 4) were increased as detected with <sup>1</sup>H-NMR and GC-MS. A low level of levulinic acid and 3-MCPD was observed in CPH treated breast muscle, which might be related to the CPH production process (CPH was commercially obtained). Using <sup>1</sup>H-NMR technique, higher amounts of acetate (GPH, WPH, CPH), butyrate (CPH), carnitine (WPH, CPH), citrate (WPH, CPH), glutathione (WPH, CPH), pantothenate (CPH) and putrescine (GPH, WPH, CPH), myo-inositol (GPH, WPH) were additionally detected (Figure 4 and Supplementary Table S4), whereas with GC-MS analysis we obtained increased levels of oxoproline (CPH), urea (CPH), hydroxylysine (GPH) and malic acid (GPH, WPH, CPH) among a few nonidentified compounds (Figure 4 and Supplementary Table S1). It can be concluded that the lower the hydrolyzation degree the more important are the additional compounds from the protein origins for the differentiation of nontreated samples and hydrolysate treated samples. *Foods* **2020**, *9*, x FOR PEER REVIEW 14 of 16
**Figure 4.** Selected signals that were present in partial hydrolysates and show significant differences between references and hydrolysate treated samples obtained via GC-MS (**a**) and 1H-NMR (**b**). **Figure 4.** Selected signals that were present in partial hydrolysates and show significant differences between references and hydrolysate treated samples obtained via GC-MS (**a**) and <sup>1</sup>H-NMR (**b**).
reduced levels (not significant) of myo-inositol and the peptides glutathione and anserine were detected by 1H-NMR. With GC-MS profiling we detected reduced levels of 4-hydroxybutanoic acid, myoinositol, inosine and uracil among several nonidentified molecular features (Supplementary Table S1). In our approach, sample preparation was performed using ~2 g fresh turkey breast meat to minimize the effect of natural variation when comparing different hydrolysate types. Whether the observed washout effect can also be detected by using whole breast muscles has to be validated by further studies. It can be suspected that the natural variation has a greater impact than the detected small levels of a wash-
This study aimed at a comparison between different analytical methods and their possibility to detect adulteration of turkey breast meat with different hydrolysates. It showed that FAA profiling allows for the detection of protein hydrolysate treatments only above a certain threshold which is mainly related to the degree of hydrolyzation. The samples naturally strongly differ in their free amino acid contents as a result of feeding, genotype and meat age. Therefore, the FAA analyses under these conditions (e.g., determination of only ten FAA contents) were not suitable for the detection of food fraud in the case of partial hydrolysates. To overcome this limitation, the contents of more than ten FAA of the 20 proteinogenic AA should be analyzed. Further on, a much higher quantity of samples ought to be measured. The evaluation of these datasets enables the reduction of the variations and therefore more significant differences. The additions of hydrolysates with high amounts of AA to breast meat
The different profiling techniques revealed that protein sources contain different metabolites, which can be used as biomarkers for the detection of partial hydrolysates. Furthermore, byproducts formed during acidic hydrolysis provide additional evidence for the treatment of breast meat with protein hydrolysates. Therefore, a combination of FAA and metabolite (by-products) profiling makes it possible to identify and classify the addition of nondeclared hydrolysates to turkey breast meat. In addition, an advantage of this comprehensive analysis is that it might be possible to proof the addition of animal-based proteins or animal-based hydrolysates to vegetarian or vegan products. According to the advantages of NMR in terms of sample throughput and direct quantification of the identified compounds, 1H-NMR is used in further/detailed studies analyzing food fraud of turkey breast meat.
In accordance with the reduced AA content in water treated samples, we observed for several
In accordance with the reduced AA content in water treated samples, we observed for several endogenous metabolites of turkey breast muscle a similar reduction when samples were injected with the different kinds of hydrolysates. This effect was particularly obvious for highly water-soluble compounds such as creatinine and lactate, which were detected by GC-MS and <sup>1</sup>H-NMR. Additionally, reduced levels (not significant) of myo-inositol and the peptides glutathione and anserine were detected by <sup>1</sup>H-NMR. With GC-MS profiling we detected reduced levels of 4-hydroxybutanoic acid, myo-inositol, inosine and uracil among several nonidentified molecular features (Supplementary Table S1). In our approach, sample preparation was performed using ~2 g fresh turkey breast meat to minimize the effect of natural variation when comparing different hydrolysate types. Whether the observed wash-out effect can also be detected by using whole breast muscles has to be validated by further studies. It can be suspected that the natural variation has a greater impact than the detected small levels of a wash-out.
#### **4. Conclusions**
This study aimed at a comparison between different analytical methods and their possibility to detect adulteration of turkey breast meat with different hydrolysates. It showed that FAA profiling allows for the detection of protein hydrolysate treatments only above a certain threshold which is mainly related to the degree of hydrolyzation. The samples naturally strongly differ in their free amino acid contents as a result of feeding, genotype and meat age. Therefore, the FAA analyses under these conditions (e.g., determination of only ten FAA contents) were not suitable for the detection of food fraud in the case of partial hydrolysates. To overcome this limitation, the contents of more than ten FAA of the 20 proteinogenic AA should be analyzed. Further on, a much higher quantity of samples ought to be measured. The evaluation of these datasets enables the reduction of the variations and therefore more significant differences. The additions of hydrolysates with high amounts of AA to breast meat were easily proofed within this study.
The different profiling techniques revealed that protein sources contain different metabolites, which can be used as biomarkers for the detection of partial hydrolysates. Furthermore, byproducts formed during acidic hydrolysis provide additional evidence for the treatment of breast meat with protein hydrolysates. Therefore, a combination of FAA and metabolite (by-products) profiling makes it possible to identify and classify the addition of nondeclared hydrolysates to turkey breast meat. In addition, an advantage of this comprehensive analysis is that it might be possible to proof the addition of animal-based proteins or animal-based hydrolysates to vegetarian or vegan products. According to the advantages of NMR in terms of sample throughput and direct quantification of the identified compounds, <sup>1</sup>H-NMR is used in further/detailed studies analyzing food fraud of turkey breast meat.
**Supplementary Materials:** The following are available online at http://www.mdpi.com/2304-8158/9/8/1084/s1, Table S1: Molecular features used for GC-MS profiling. Annotation was performed with the NIST 14 library implemented in the GCMSsolution software (Shimadzu, Duisburg, Germany). The signal intensities of the quantifier ions were used for the relative quantitation and comparison between sample groups. According to the presented FAA contents the Dunnett's test was used for a comparison between control (REF) and the different treatments.; Table S2: Quantification of amino acids via GC-MS was performed using the following parameters.; Table S3: Assignment of <sup>1</sup>H-NMR signals which could be identified via the software ChenomX NMR Suite 8.4 library. 86 metabolites were identified and exemplary one NMR signal (ppm) was chosen which was the obvious signal for identification.; Table S4: Significantly different absolute concentrations of metabolites in turkey breast meat treated with different hydrolysates and analyzed via <sup>1</sup>H-NMR (mg/100 g).; Table S5: Significantly different absolute concentrations (mg/100 g) of amino acids in turkey breast meat treated with different hydrolysates and analyzed via HPLC-UV/VIS, GC-MS and <sup>1</sup>H-NMR.
**Author Contributions:** Conceptualization, L.W., M.P., B.K., N.G, S.A., D.A.B.; methodology, validation, data curation, formal analysis, investigation, B.K. (HPLC), M.P. (GC-MS), L.W. (NMR); investigation, N.G. (NMR); resources, U.B. (NMR), D.A.B.; writing original draft, B.K. (HPLC), M.P. (GC-MS), L.W. (NMR); writing- review and editing, all authors; L.W. coordinated the editing of the manuscript; visualization, B.K. (HPLC), M.P. (GC-MS), L.W. (NMR); supervision, D.A.B.; project administration, all authors. All authors have read and agreed to the published version of the manuscript.
**Funding:** This project was performed within the framework of the research project "Fremdeiweiß" (ProHydAdd), delegated by the Federal Ministry of Food and Agriculture, Germany.
**Acknowledgments:** The authors thank the technical staff Elke Gardill, Gabriele Schüßler and Katrin Weiß for their assistance in the laboratories.
**Conflicts of Interest:** The authors declare no conflict of interest.
### **References**
© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
## **Comparison of Real-Time PCR Quantification Methods in the Identification of Poultry Species in Meat Products**
## **Kerstin Dolch, Sabine Andrée \* and Fredi Schwägele**
Department of Safety and Quality of Meat, Max Rubner-Institute, E.-C.-Baumann-Str. 20, 95326 Kulmbach, Germany; [email protected] (K.D.); [email protected] (F.S.)
**\*** Correspondence: [email protected]
Received: 26 June 2020; Accepted: 31 July 2020; Published: 3 August 2020
**Abstract:** Poultry meat is consumed worldwide and is prone to food fraud because of large price differences among meat from different poultry species. Precise and sensitive analytical methods are necessary to control poultry meat products. We chose species–specific sequences of the *cytochrome b* gene to develop two multiplex real-time polymerase chain reaction (real-time PCR) systems: one for chicken (*Gallus gallus*), guinea fowl (*Numida meleagris*), and pheasant (*Phasianus colchicus*), and one for quail (*Coturnix japonica*) and turkey (*Meleagris gallopavo*). For each species, added meat could be detected down to 0.5% *w*/*w*. No cross reactions were seen. For these two real-time PCR systems, we applied three different quantification methods: (A) with relative standard curves, (B) with matrix-specific multiplication factors, and (C) with an internal DNA reference sequence to normalize and to control inhibition. All three quantification methods had reasonable recovery rates from 43% to 173%. Method B had more accepted recovery rates, i.e., in the range 70–130%, namely 83% compared to 75% for method A or C.
**Keywords:** real-time PCR; quantification; chicken; guinea fowl; pheasant; quail; turkey
### **1. Introduction**
Consumer awareness for food is growing. On one side, this may be due to health, religious, or ideological issues. On the other side, consumers are sensitized due to food fraud incidences like the horsemeat scandal [1]. Therefore, they want to know what they are getting for their money. For processed food products, the easiest information source is the label of ingredients. In the EU, regulation (EU) No 1169/2011 defines specifically what the label should contain and in which order [2]. For their control, if these regulations are complied, affordable and practical analytical methods are necessary. Hence, one of the main focuses of food authenticity testing is to have the right analytical methods in place and, if necessary, to develop new methods or to improve existing ones.
A change in detailedness of analytic results is one point where improvement is needed. Qualitative results are sufficient when many samples are screened to obtain a rough idea with respect to the contamination rate, and to find out suspected cases. However, in processed food with several ingredients, it is not always sufficient to detect a specific ingredient. Quite often, it is more important to know if the content of one ingredient is higher than another one [3], or if the concentration of an ingredient exceeds a certain threshold [4].
To check the correct declaration of ingredients of animal or plant origin, one strategy is to detect a specific sequence of the deoxyribonucleic acid (DNA) of the corresponding ingredient. This is possible as each species has a unique genome. One widely established method for this is the real-time polymerase chain reaction (real-time PCR). It has been used for a long time for different food products, and constantly new methods are published for meat, seafood, milk, and dairy products, as well as fruit juices [5]. The advantages of real-time PCR analysis are its easy handling and affordable laboratory equipment. The biggest issue, however, is to obtain valid results for legal purposes [6]. In literature, there exist several options.
The most straightforward idea is absolute quantification, where a serial dilution of the target DNA sequence is used as the standard curve [7]. Thereby, the template can be either directly isolated DNA of the pure target [8], a plasmid containing the cloned DNA sequence [9], or synthetically synthesized DNA [10]. This quantification method is well suited for raw food samples. However, for processed food, it is not practicable, as all three DNA sources for the standard curve have in common that they were not treated like the unknown sample. The measurement of DNA is an indirect quantification method for the added amount of animal or plant tissue. The detected concentration of DNA deriving from the target species should correlate directly with the amount of tissue added. But heat treatment may change the amount of DNA detectable due to heat degradation [11,12], which would lead to an underestimation of the actual amount of target tissue added.
The next possibility (method A) is to use a relative standard curve [13]. This step allows to perform quantification of processed and unprocessed food by co-analyzing the DNA of unknown samples together with the DNA of reference material. Therefore, reference material is produced under the same production conditions as the unknown sample. Before production, the target is added in quantities that comply with the measurement range, and DNA is isolated from these DNA standard samples as a reference. For each real-time PCR run, these DNA standard samples have to be applied and measured together with the DNA of the unknown samples [14].
Another possibility (method B) is to determine the matrix-specific multiplication factor of each species under each production condition. DNA is isolated from the reference material and from raw meat. These DNA samples are analyzed together to obtain the matrix-specific multiplication factor. In all further quantification experiments, the DNA of the reference material is not needed anymore. Instead, the DNA samples of the raw material are measured together with the DNA of the unknown samples, and corrected with the matrix-specific multiplication factors obtained earlier [3,15,16]. As the matrix-specific multiplication factors vary between laboratories [17], each laboratory has to determine their own multiplication factors for each animal species and each processing condition [17]. To overcome this time- and labor-consuming step, an internal DNA reference sequence is necessary to quantify via a normalized standard curve (method C). This can be a common DNA sequence like *myostatin* or a ribosome subunit [18], which detects the whole amount of eukaryotic DNA. In the subsequent analytical process, either the ratio is determined between target and reference sequence [19,20], or the difference between the detected amount of target and internal reference sequence (∆Cq) [21,22].
This leads to one of the most important decisions of real-time PCR: the choice of the target DNA sequence. While the usage of single-copy DNA sequences is preferred because of the more stable copy number per cell, the application of multi-copy DNA sequences is in favor of lower detection limits [23].
This study focused on the quantification of the relative meat content for five poultry species in meat products as poultry is the most consumed meat, and its consumption rate is still growing [24]. In addition, poultry products are ranked in the top-ten list of most susceptible product categories [25]. The main species for poultry meat are chicken (*Gallus gallus*) and turkey (*Meleagris gallopavo*), while guinea fowl (*Numida melegaris*), quail (*Coturnix japonica*), and pheasant (*Phasianus colchicus*) are less consumed in Germany [26]. For each bird species, a DNA sequence of the mitochondrial *cytochrome b* gene was chosen, which is often used for identifying animal species [27,28]. For the quantification method C, we chose a sequence of the *12S rRNA* gene as it is a mitochondrial DNA sequence as well. To determine if the processing temperature affects the possibility to detect meat from the five poultry species, sausages were prepared under two different temperatures and analyzed with the three different quantification methods.
### **2. Material and Methods**
#### *2.1. Material*
#### 2.1.1. Chemical Material
The following chemicals were used: Proteinase K (Machery-Nagel, Düren, Germany), hydrogen chloride, isopropanol, sodium chloride, and tris(hydroxymethyl)aminomethane (Merck, Darmstadt, Germany), dodecyl sulfate sodium salt (Serva, Heidelberg, Germany), ethylenediaminetetraacetic acid disodium salt (Riedel-de Haën, Seelze, Germany), guanidine hydrochloride, DNA-free water, and RNAse A (Sigma, St. Louis, USA), and ethanol (Th. Geyer, Renningen, Germany). The Wizard® Plus Minipreps DNA Purification System was from Promega (Mannheim, Germany), the QuantiTect Multiplex PCR NoRox from Qiagen (Hilden, Germany), and real-time PCR tubes from LTF-Labortechnik (Wasserburg, Germany).
Primers and probes were synthesized by Eurofins Genomics (Ebersberg, Germany).
#### 2.1.2. Sample Material
All meat samples were pectoral muscle meat. Chicken and turkey meat were obtained from C + C, Kulmbach, Germany, and guinea fowl, pheasant, and quail as whole carcasses from a breeder in Bad Wörishofen, Germany (Josef Maier). All other meat and plant samples were bought in local stores.
Emulsified type sausages were produced twice as two independent batches A and B on separate days. If necessary, the carcasses were dissected, and the meat was minced in a Bizerba Ladenwolf (Baling, Germany). The basic formulation consisted of 50% meat, 25% sunflower oil, 23% ice, 1.7% nitrite salting mix, and 0.3% phosphate. All % values are (*w*/*w*) per sausage filling. For each species, the ingredients were added to a meat grinder (Food Machines Saarbrücken MK13, Germany) and mixed at 2600 rpm. The ground meat from all five poultry species were combined in various percentages (Table 1), and then the additional ingredients were added. Sausages were filled into cans (type 99/36 mm, Dosen-Zentrale Züchner GmbH, Cologne, Germany) and cooked at low (75 ◦C; batch A for 30 min, batch B for 4 min) or high temperatures (117 ◦C; with final F values of 5.68 and 6.01 for batches A and B, respectively, cf. [29]).
**Table 1.** Composition of standard (S1-5) and unknown emulsified type sausages (U1-5).
#### *2.2. Methods*
#### 2.2.1. Bioinformatics
A DNA sequence of the mitochondrial genome was obtained from NCBI GenBank for chicken (NC\_040970.1), guinea fowl (NC\_034374.1), pheasant (NC\_015526.1), quail (NC\_003408.1), and turkey (NC\_034374.1). These sequences were aligned with the software Molecular Evolutionary Genetics Analysis (MEGA) [30]. Regions with high similarity were chosen for primer and probe binding sites in the area coding for the *12S rRNA* gene.
The theoretical specificity of all primers was checked with the Primer-BLAST software (Basic Local Alignment Search Tool, NCBI) with the same parameters as described in [14].
#### 2.2.2. DNA Isolation
The Wizard DNA isolation kit from Promega was used as the standard DNA isolation method [31]. All samples were prepared as duplicates according to the corresponding instruction. DNA-free water was used as negative control.
All DNA samples were quantified and qualified by measuring at 260, 280, and 340 nm with a spectrophotometer DU 7400 (Beckman Coulter, Brea, CA, USA).
#### 2.2.3. Real-Time PCR
#### Reaction Set-Up
All real-time PCR assays were performed on a RotorGene 6000 (Qiagen, Hilden, Germany) according to the QuantiTect Multiplex PCR handbook (Qiagen, Hilden, Germany). The reaction was set up in 25 µL with primer and probe concentrations according to Table 2. The following cycler regime was used: 15 min at 95 ◦C, 35 cycles of 15 s at 95 ◦C, and 1 min at 60 ◦C, collecting the fluorescence signal at the end of each cycle.
#### Templates
After DNA isolation of the poultry sausages, the duplicates were combined. They were either adjusted to a DNA concentration of 20 ng/µL or diluted 1:10 with elution buffer. The other animal and plant samples were adjusted to a DNA concentration of 2 ng/µL. For the determination of efficiency, R 2 , and limit of detection (LOD) values, the DNA samples of the poultry sausage and of the pure meat samples were diluted ten-fold with elution buffer.
All DNA samples were analyzed in triplicates (standard samples) or duplicates (unknown samples), or in sextets (influence of chicken DNA on detection of pheasant DNA and the LOD). Positive controls and no-template controls (water) were measured once.
#### 2.2.4. Calculation
#### Method A: Quantification with Reference Material
All DNA samples were diluted 1:10 with elution buffer. The DNA samples from S1–5 (Table 1) were used as standard material, and the corresponding Cq values were plotted against the logarithmic starting quantity. This standard curve was used to quantify the amount of meat from each poultry species in the unknown samples U1–5 (Table 1) for both production temperatures.
#### Method B: Quantification with Matrix-Specific Multiplication Factors
For establishing the matrix-specific multiplication factors for each species, DNA was isolated from poultry meat and from the standard emulsified type sausages S1–5 (Table 1) which were adjusted to a DNA-concentration of 20 ng/µL with elution buffer. The DNA from the poultry meat was used as standard material to obtain standard curves. For the detection of chicken and quail meat, this was obtained by using 0.01, 0.1, 1.0, 10, and 100 ng DNA per real-time PCR reaction. For the detection of guinea fowl, pheasant, and turkey meat, 0.1, 1.0, 5.0, 10, and 100 ng DNA per real-time PCR reaction were used. The DNA from the emulsified type sausages S1–5 were used for calculating the respective multiplication factors according to Köppel et al. [13,15] for each meat from poultry species, separately for cooking at low or high temperatures. These multiplication factors were used to calculate the amount of meat from poultry species added in the unknown emulsified type sausages U1-5 (Table 1).
#### Method C: Quantification with Internal Reference Sequence
This method was performed according to Soares et al. [34] with Equation (1):
$$
\Delta \mathbf{C} \mathbf{q} = \mathbf{C} \mathbf{q}\_{\text{target}} - \mathbf{C} \mathbf{q}\_{\text{reference}} \tag{1}
$$
The poultry-species-specific real-time PCR systems were used as the target, and the eukaryotic real-time PCR system was used as the reference.
All DNA samples were diluted 1:10 with elution buffer, and the ∆Cq values from S1–5 (Table 1) were plotted against the logarithmic starting quantity for the standard curve. With this standard curve, we calculated the amount of DNA in the unknown samples U1–5 (Table 1). This was performed for the meat from each species and under both processing temperatures.
#### 2.2.5. Statistical Analysis
Calculations were performed either with the Rotor-Gene Q Series Software (Qiagen, Hilden, Germany), Excel (Microsoft Office 2019, Redmond, WA, USA), or with JMP (SAS, Heidelberg, Germany). All factors were analyzed by multiple logistic regression, and the chi-squared values were recorded. The level of significance was set at 5%. Standard box plots were used to visualize the data. The box plots show the median, quantiles as boxes, and whiskers extend to 1.5 times the interquartile distance at most. Outliers were not omitted from the analysis.
#### **3. Results**
#### *3.1. Bioinformatics*
All primer pairs were checked theoretically for specificity against the ten most commonly eatable bird species (chicken, duck, emu, goose, guinea fowl, ostrich, partridge, pheasant, quail, and turkey). Additionally, the theoretical cross reactivity of the primers was checked for the triplex real-time PCR system (C-G-P) and for the duplex real-time PCR system (Q-T) against all eukaryotes. All false positive matches had several mismatches: the amplicons were either too short or too long, and/or the species were irrelevant as food. Consequently, there were no relevant false positive matches.
The primer pair for detecting all five species was checked theoretically against all entries for animal organisms in the NCBI GenBank database, and amplicons were obtained with a length of 143–146 bp. No mismatches were found for the five poultry species investigated.
#### *3.2. Development of One Triplex and One Duplex Real-Time PCR System*
A pentaplex real-time PCR system was proposed in a former publication. However, this system had a lack in precision and accuracy [32]. To overcome this problem, the pentaplex real-time PCR system was split into one triplex real-time PCR system for detecting meat of chicken, guinea fowl, or pheasant (C-G-P), and one duplex real-time PCR system for detecting meat of quail or turkey (Q-T).
DNA isolated from raw meat had concentrations of 333 ng/µL for chicken, 245 ng/µL for guinea fowl, 228 ng/µL for pheasant, 593 ng/µL for quail, and 209 ng/µL for turkey. Ten-fold dilution series of 10−1–10−<sup>7</sup> gave standard curves for detecting the meat of each species. For the triplex real-time PCR system, efficiency and *R* <sup>2</sup> values were 102% and 0.983 for chicken, 91% and 0.995 for guinea fowl, and 95% and 0.985 for pheasant; for the duplex system, the values were 94% and 0.998 for quail and 93% and 0.993 for turkey, respectively.
No signal was received with either real-time PCR system when DNA of the following animal species was used: bison, buffalo, camel, chamois, elk, fallow deer, goat, horse, llama, mouflon, pig, reindeer, roe deer, sheep, tuna, wild hare, zebra, or zebu, or DNA from the following plant species: bean, beetroot, black mustard, broccoli, Brussels sprouts, bunching onion, caraway, cardamom, carrot, cauliflower, celery, chili, Chinese cabbage, coriander, cress, cucumber, fennel, garden leek, garden radish, garlic, ginger, green cabbage, horseradish, Indian mustard, kohlrabi, lemon, marjoram, onion, parsley, pepper, pistachio, potato, pumpkin, radish, red cabbage, rutabaga, salsify, savoy cabbage, tomato, white mushroom, white mustard, white pepper, wood garlic, or zucchini.
However, signals were obtained for DNA of chicken, guinea fowl, pheasant, quail, or turkey, each with the respective real-time PCR system (Cq = 15–18) (Table 3). False positive signals were obtained for a few DNA samples with the earliest Cq value of 29. Additionally, all five real-time PCR systems had in common that the blank values gave signals with Cq values between 29 and 33. Therefore, a cut-off was set at Cq ≥ 29.
#### *3.3. Quantification*
The unknown emulsified type sausages were analyzed with three quantification methods. For each method, the predicted means were calculated for the unknown samples, together with standard deviations, coefficients of variation (CV), and bias. The CV represents the relative standard deviation of results obtained under repeatability conditions, and was accepted with CV ≤ 25 [13].
Bias was accepted in a range of ±25% relative to the mean. Additionally, recovery rates were calculated, and a range of ±30% was accepted, i.e., recovery rates of 70–130% [13].
#### 3.3.1. Method A: Quantification with Reference Material
For this method, DNA of the reference material was used to obtain a standard curve. The reference material with known concentrations of meat from the five poultry species was produced under the same conditions as the unknown emulsified type sausages. Most of the CV and bias values were within the accepted range. Some of the values were out of range, especially when detecting small concentrations of poultry meat or a high concentration of quail meat (Table 4).
**Table 3.** Cq values for various animal and plant species tested with the triplex real-time polymerase chain reaction (real-time PCR) system (C-G-P) and the duplex real-time PCR system (Q-T) systems.
All samples were measured in duplicates. - no Cq values were obtained until cycle 35.
**Table 4.** Predicted concentrations of meat from five poultry species in unknown emulsified type sausages under two temperature conditions, quantified with reference material.
<sup>a</sup> Values are the means of replicate assays (*n* = 12); <sup>b</sup> SD—standard deviation; <sup>c</sup> CV—coefficient of variation; <sup>d</sup> Bias <sup>=</sup> 100 \* ((mean value <sup>−</sup> actual value)/actual value).
Most recovery rates were within the accepted range of 70–130% (Figure 1). Lower recovery rates (<70%) were obtained for detecting small concentrations (0.5% meat) and higher recovery rates (>130%) for detecting 57.5% quail meat. *Foods* **2020**, *9*, 1049 8 of 18
**Figure 1.** Recovery rates of meat from five poultry species in emulsified type sausages (with 0.5–57.5% meat) quantified with reference material from standard emulsified type sausages (with 0–69% meat). All concentration levels were cooked at low (L) or high temperature (H). DNA was isolated in duplicate from each sausage from both batches, and three independent real-time PCRs were performed, i.e., box plots are from twelve measurements. The grey areas represent the accepted range **Figure 1.** Recovery rates of meat from five poultry species in emulsified type sausages (with 0.5–57.5% meat) quantified with reference material from standard emulsified type sausages (with 0–69% meat). All concentration levels were cooked at low (L) or high temperature (H). DNA was isolated in duplicate from each sausage from both batches, and three independent real-time PCRs were performed, i.e., box plots are from twelve measurements. The grey areas represent the accepted range of 70–130%.
of 70–130%. 3.3.2. Method B: Quantification with Matrix-Specific Multiplication Factors
3.3.2. Method B: Quantification with Matrix-Specific Multiplication Factors The DNA of the standard emulsified type sausages were used to calculate the matrix-specific multiplication factors, separately for low or high cooking temperatures, with the DNA of raw meat as standard material. The multiplication factors ranged from 0.90 (for pheasant meat) to 3.82 (for quail meat) at low cooking temperature, and from 0.09 (for pheasant meat) to 0.52 (for turkey meat) at high The DNA of the standard emulsified type sausages were used to calculate the matrix-specific multiplication factors, separately for low or high cooking temperatures, with the DNA of raw meat as standard material. The multiplication factors ranged from 0.90 (for pheasant meat) to 3.82 (for quail meat) at low cooking temperature, and from 0.09 (for pheasant meat) to 0.52 (for turkey meat) at high cooking temperature (Table 5).
cooking temperature (Table 5). **Table 5.** Matrix-specific multiplication factors to predict the concentration of meat from five poultry species in unknown emulsified type sausages under two temperature conditions. All of the CV values were within the accepted range, only the CV value was slightly higher for detecting 0.5% of pheasant meat at low temperature (Table 6). Most of the bias values were as well within the given range, only for detecting chicken, pheasant, and turkey meat some values were out of the range.
> **Temperature Batch Chicken Guinea Fowl Pheasant Quail Turkey** Low A 1.16 1.19 0.68 3.02 1.19 B 1.44 1.49 1.12 4.61 1.17
> > **Mean 1.30 1.34 0.90 3.82 1.18**
**Mean 0.24 0.11 0.09 0.30 0.52**
All of the CV values were within the accepted range, only the CV value was slightly higher for detecting 0.5% of pheasant meat at low temperature (Table 6). Most of the bias values were as well
High A 0.24 0.12 0.09 0.31 0.61 B 0.23 0.09 0.10 0.29 0.42
**Table 5.** Matrix-specific multiplication factors to predict the concentration of meat from five poultry species in unknown emulsified type sausages under two temperature conditions.
**Table 6.** Predicted concentrations of meat from five poultry species in unknown emulsified type sausages under two temperature conditions, quantified with matrix-specific multiplication factors.
<sup>a</sup> Values are the means of replicate assay (*n* = 12); <sup>b</sup> SD—standard deviation; <sup>c</sup> CV—coefficient of variation; <sup>d</sup> Bias = 100 \* ((mean value—actual value)/actual value).
Most recovery rates were within the accepted range of 70–130% (Figure 2). Lower recovery rates were obtained for detecting small concentrations of chicken and higher recovery rates were obtained for detecting small concentrations of pheasant meat.
#### 3.3.3. Method C: Quantification with an Internal Reference Sequence
A mitochondrial reference sequence was chosen because the specific target sequences were mitochondrial. This additional step did not only normalize the results, it also worked well as an amplification and PCR inhibition control, which is recommended for processed food products [13]. For detecting the meat of guinea fowl, quail, and turkey, most of these values are either close to the limit of the range or above (Table 7). On the contrary, the CV and bias values are mostly within the range for detecting chicken or pheasant meat.
*Foods* **2020**, *9*, 1049 10 of 18
**Figure 2.** Recovery rates of meat from five poultry species in emulsified type sausages (with 0.5–57.5% meat) quantified with pre-defined multiplication factors. All concentration levels were cooked at low (L) or high temperature (H). DNA was isolated in duplicate from each sausage from both batches, and three independent real-time PCRs were performed, i.e., box plots are from twelve measurements. The grey areas represent the accepted range of 70–130%. 3.3.3. Method C: Quantification with an Internal Reference Sequence **Figure 2.** Recovery rates of meat from five poultry species in emulsified type sausages (with 0.5–57.5% meat) quantified with pre-defined multiplication factors. All concentration levels were cooked at low (L) or high temperature (H). DNA was isolated in duplicate from each sausage from both batches, and three independent real-time PCRs were performed, i.e., box plots are from twelve measurements. The grey areas represent the accepted range of 70–130%.
mitochondrial. This additional step did not only normalize the results, it also worked well as an amplification and PCR inhibition control, which is recommended for processed food products [13]. For detecting the meat of guinea fowl, quail, and turkey, most of these values are either close to the **Table 7.** Predicted concentrations of meat from five poultry species in unknown emulsified type sausages under two temperature conditions, quantified with an internal reference sequence.
A mitochondrial reference sequence was chosen because the specific target sequences were
<sup>a</sup> Values are the means of replicate assay (*n* = 12); <sup>b</sup> SD—standard deviation; <sup>c</sup> CV—coefficient of variation; <sup>d</sup> Bias = 100 \* ((mean value—actual value)/actual value).
*Foods* **2020**, *9*, 1049
The median of most of the recovery rates were within the accepted range of 70–130% (Figure 3). However, the scattering of the values for the recovery rates were wide for detecting the five poultry meat species. Lower recovery rates were obtained for detecting small concentrations of guinea fowl meat (0.5% meat). *Foods* **2020**, *9*, 1049 12 of 18
**Figure 3.** Recovery rates of meat from five poultry species in emulsified type sausages (with 0.5–57.5% meat) quantified with an internal reference sequence. All concentration levels were cooked at low (L) or high temperature (H). DNA was isolated in duplicate from each sausage from both batches, and three independent real-time PCRs were performed, i.e., box plots are from twelve measurements. The grey areas represent the accepted range of 70–130%. **Figure 3.** Recovery rates of meat from five poultry species in emulsified type sausages (with 0.5–57.5% meat) quantified with an internal reference sequence. All concentration levels were cooked at low (L) or high temperature (H). DNA was isolated in duplicate from each sausage from both batches, and three independent real-time PCRs were performed, i.e., box plots are from twelve measurements. The grey areas represent the accepted range of 70–130%.
#### 3.3.4. Comparison 3.3.4. Comparison
Repeatability (CV) and bias were used to compare the three quantification methods. With low cooking temperature, the CV values differed (χ2 = 0.0079). For method A, 88% of the CV values were within the limits, 96% for method B, and 67% for method C. For the bias, no obvious difference was seen between method A, B, or C (χ2 = 0.3679). With high cooking temperature, the percentage of accepted CV values differed between the three methods (χ2 = 0.0395). For method A, 84% of the CV values were within the limits, 100% for method B, and 76% for method C. No differences were seen for the bias values (χ2 = 0.4000) (data not shown). Repeatability (CV) and bias were used to compare the three quantification methods. With low cooking temperature, the CV values differed (χ <sup>2</sup> = 0.0079). For method A, 88% of the CV values were within the limits, 96% for method B, and 67% for method C. For the bias, no obvious difference was seen between method A, B, or C (χ <sup>2</sup> = 0.3679). With high cooking temperature, the percentage of accepted CV values differed between the three methods (χ <sup>2</sup> = 0.0395). For method A, 84% of the CV values were within the limits, 100% for method B, and 76% for method C. No differences were seen for the bias values (χ <sup>2</sup> = 0.4000) (data not shown).
Another criterion is to compare the recovery rates, where the limits for acceptance were set to ±30%. A multiple logistic regression was performed, and all predictors which did not significantly contribute to the whole model (*p*-value > 5%) were removed from analysis. Thus, cooking temperature and batches were omitted from the model. The three quantification methods differed in the percentage of accepted recovery rates (*p* = 0.0110), with 75% for method A, 83% for method B, and 75% for method C. There was no obvious pattern for under- or for overestimation. Recovery rates Another criterion is to compare the recovery rates, where the limits for acceptance were set to ±30%. A multiple logistic regression was performed, and all predictors which did not significantly contribute to the whole model (*p*-value > 5%) were removed from analysis. Thus, cooking temperature and batches were omitted from the model. The three quantification methods differed in the percentage of accepted recovery rates (*p* = 0.0110), with 75% for method A, 83% for method B, and 75% for method C. There was no obvious pattern for under- or for overestimation. Recovery rates varied between
and well suited. For detecting turkey meat, method A showed the highest accepted recovery rates.
varied between poultry species (*p* = 0.0129) as well as between concentration levels of poultry meat
poultry species (*p* = 0.0129) as well as between concentration levels of poultry meat (*p* < 0.0001). For detecting chicken meat, all three quantification methods had low accepted recovery rates (Table 8). For detecting guinea fowl meat, the quantification method B showed high accepted recovery rates. For detecting pheasant and quail meat, all three quantification methods are similar and well suited. For detecting turkey meat, method A showed the highest accepted recovery rates.
**Table 8.** Overview of the applied quantification methods, with percentages of accepted bias, coefficients of variation (CV), and recovery rate.
a Isolation of DNA from reference material and unknown sample; one real-time PCR for calculation of unknown sample. <sup>b</sup> Isolation of DNA from unknown sample, reference material, and raw meat; one real-time PCR for calculation of multiplication factor and another one for calculation of unknown sample. <sup>c</sup> Isolation of DNA from reference material and unknown sample; two real-time PCR assays (one for target and one for reference sequence) for calculation of unknown sample. <sup>d</sup> for explanation see Section 3.3: Quantification.
#### **4. Discussion**
In this study, we compared three different methods to quantify the amount of meat from chicken, guinea fowl, pheasant, quail, or turkey in meat products, cooked at low or high temperatures [24–26].
For the detection of the two main poultry meat species—chicken and turkey—there is a large variety of real-time PCR systems. Most of them are single real-time PCR systems to detect a mitochondrial gene like *cytochrome b* [4,23,27,28,35,36]. For the detection of chicken meat, a few chromosomal genes are used like *interleukin-2* gene [37] or β*-actin* gene [38]. Fewer real-time PCR systems have been published for the detection of meat from guinea fowl [39], pheasant [35,39], or quail [35,39]. There is only one multiplex real-time PCR system for the combination of chicken and turkey meat [16]. To our knowledge, less prominent poultry meat species like guinea fowl, pheasant, or quail have not been considered so far. However, these species are also relevant as they are a delicacy and high-priced. Therefore, the focus was set on the combined detection of the two main poultry meat species, chicken and turkey, together with the high-priced poultry meat species guinea fowl, pheasant, and quail.
The bioinformatic testing of the primers resulted in no false positive matches as single systems, as well as within their combinations (C-G-P and Q-T). However, as the DNA databases are not complete, there is always the chance to miss a species [14]. Therefore, different animal and plant DNAs were tested with both multiplex systems. False positive signals were obtained with a few of these DNA samples. However, each of these Cq values appeared later than the Cq values of the blank samples, and each was below the cut-off value. Furthermore, no influence of chicken DNA on the Cq value of detecting pheasant DNA was shown and it was not possible to establish values for the LOD as all dilutions were to 100% detectable until the cut-off. This, together with high efficiency and R<sup>2</sup> values, indicates that both multiplex real-time PCR systems are precise, specific, sensitive, and suitable to differentiate meat from these five poultry species in meat products.
The two real-time PCR systems were established with DNA that was isolated from 300 mg fresh meat from each species. The DNA content for quail meat was almost twice as high than for each of the other four species. One explanation for this observation is the small size of quails, which is the smallest of the five poultry species investigated. A positive correlation between cell size and body mass among birds [40] implies higher DNA content per body weight in smaller than in larger bird species.
In the literature, there is a large number of methods to quantify material of animal or plant origin in food products [13,41]. Some of these methods are not suited for processed food products, and were therefore not considered in this study. All other methods have in common that standard reference material is required which should be prepared under identical conditions, with similar content, and in similar concentrations [42,43]. Therefore, standard and unknown emulsified type sausages were prepared under comparable conditions.
The three quantification methods compared in this study are in wide use. Quantification method A used DNA from reference material to establish a standard curve that was applied to quantify the amount of meat of each bird species in the unknown samples. At low cooking temperature, the recovery rates were between 70% (chicken or turkey meat) and 90% (pheasant meat) within the accepted limits. At high cooking temperature, the recovery rates were lower than at low temperature for most species, but 100% for turkey meat. Combined with the high bias values for the detection of a low concentration of 0.5% pheasant meat, and 57.5% of quail meat, it can be concluded that quantification with reference material at high cooking conditions is not suited for the whole concentration range for all poultry species. Overall, the idea of this quantification method is quite straight forward, but the main problem is to have the right standard material in stock. For research purposes, this is feasible, and this method was successfully applied to our unknown emulsified type sausages using the standard emulsified type sausages.
Quantification method B applied multiplication factors. This method was first published by Köppel and colleagues in 2011 to detect cow, pig, horse, and sheep [15]. It has been applied to many different animal species since [3,17]. In this study, the multiplication factors were established separately for each bird species and each cooking temperature. The multiplication factors were smaller for high than for low cooking temperatures. This might be due to the higher degradation rate of the DNA due to the higher processing temperature [12]. At low cooking temperature, the percentage of recovery rate values within the limits of ±30% reached from 73% for chicken or pheasant meat to 97% for guinea fowl or quail meat. At high cooking temperature, the percentage of recovery rate values within the limits ranged from 67% (chicken or turkey meat) to 97% (quail meat). Only for the detection of a concentration of 0.5% of pheasant meat, the CV and the bias values were out of range for both cooking temperatures. This implies that, with such a low concentration of pheasant meat, the quantification is not accurate. Overall, the quantification via multiplication factors was effectively applied to the unknown emulsified type sausages. However, as this quantification method normalizes the concentration determined for each species, this method is only practicable when all species added are both known and analyzed together. Therefore, both real-time PCR systems should be expanded if e.g., pork or beef meat were additional ingredients.
For quantification method C, an additional real-time PCR system was necessary. This system should amplify a specific sequence from all eukaryote species. Therefore, this system is a way to measure the total amount of eukaryotic DNA in a sample. In the literature, many different universal systems have been published. The most common system for quantification of mammal or poultry DNA is the *myostatin* gene. There are several systems which differ in their amplicon length [44–46]. However, none of these systems are suited for both of our multiplex real-time PCR systems which detect sequences of the multicopy and mitochondrial *cytochrome b* gene, while *myostatin* is single-copy and nuclear. Another gene which is used quite often is the *18S rRNA* sequence [34,47–49]. This gene is multicopy, however, it is also nuclear. Therefore, it was necessary to develop a new real-time PCR system for eukaryotes which amplifies a multicopy and mitochondrial sequence: the *12S rRNA* gene. Because the calculation of the ∆Cq is widely used [13], this method was applied in our study. At low cooking temperature, the percentage of the recovery rate within the limits ranged from 50% (guinea fowl meat) to 80% (pheasant meat), and at high cooking temperature, the values were between 73% (guinea fowl meat) and 87% (quail meat). Under both conditions, the CV and bias values were especially large for the lower concentrations. This method allowed the detection of meat from all five poultry species. As an advantage of this quantification system, the amplification of a reference sequence serves also as an inhibition control. However, the addition of another real-time PCR system duplicates the number of samples necessary, and consequently also the costs. Moreover, this quantification is not always precise.
In summary, each quantification method was successfully applied to detect meat from the five species in poultry meat products. While method A had a simple and easy line of action (just a standard curve from standard emulsified sausages), the other two methods were more labor-intensive. For method B, the multiplication factors had to be determined additionally, and for method C, an additional real-time PCR system had to be established and performed. For highly processed food products, an inhibition control is recommended and already included in method C. If the detection system is to be used more often, quantification method B was the easiest to operate: in future experiments, the standard emulsified type sausages are not needed anymore, and the DNA from raw meat can be used for preparing standard curves. In addition, with some minor exceptions, the percentage of acceptable values for CV and recovery rate were the highest for method B.
#### **5. Conclusions**
Overall, splitting the pentaplex real-time PCR system into one triplex and one duplex real-time PCR system led to a stable, precise, and specific detection method to identify chicken, guinea fowl, pheasant, quail, and turkey meat. All three quantification methods were successfully applied, although mitochondrial gene sequences were chosen. While each quantification method had its pros and cons, a final choice of the quantification method depends on the purpose of its application and the expected concentration of poultry meat species in the meat product.
**Author Contributions:** Conceptualization, K.D., F.S., and S.A.; methodology, K.D., F.S., and S.A.; validation, K.D., F.S., and S.A.; formal analysis, K.D.; investigation, K.D., F.S., and S.A.; resources, F.S.; writing—original draft preparation, K.D.; writing—review and editing, F.S. and S.A.; visualization, K.D.; supervision, F.S.; project administration, K.D., F.S., and S.A. All authors have read and agreed to the published version of the manuscript.
**Funding:** This research received no external funding.
**Acknowledgments:** We thank K. Fischer and E. Müller for their technical assistance in the lab, E. Schlimp, J. Haida, M. Spindler, and M. Zäh for producing the emulsified type sausages, and M. Judas for proof-reading.
**Conflicts of Interest:** The authors declare no conflict of interest.
#### **References**
© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
*Communication*
## **Species Identification of Red Deer (***Cervus elaphus***), Roe Deer (***Capreolus capreolus***), and Water Deer (***Hydropotes inermis***) Using Capillary Electrophoresis-Based Multiplex PCR**
## **Mi-Ju Kim, Yu-Min Lee, Seung-Man Suh and Hae-Yeong Kim \***
Institute of Life Sciences & Resources and Department of Food Science & Biotechnology, Kyung Hee University, Yongin 17104, Korea; [email protected] (M.-J.K.); [email protected] (Y.-M.L.); [email protected] (S.-M.S.)
**\*** Correspondence: [email protected]
Received: 26 June 2020; Accepted: 22 July 2020; Published: 23 July 2020
**Abstract:** To provide consumers correct information on meat species, specific and sensitive detection methods are needed. Thus, we developed a capillary electrophoresis-based multiplex PCR assay to simultaneously detect red deer (*Cervus elaphus*), roe deer (*Capreolus capreolus*), and water deer (*Hydropotes inermis*). Specific primer sets for these three species were newly designed. Each primer set only amplified target species without any reactivity against non-target species. To identify multiple targets in a single reaction, multiplex PCR was optimized and combined with capillary electrophoresis to increase resolution and accuracy for the detection of multiple targets. The detection levels of this assay were 0.1 pg for red deer and roe deer and 1 pg for water deer. In addition, its applicability was demonstrated using various concentrations of meat DNA mixtures. Consequently, as low as 0.1% of the target species was detectable using the developed method. This capillary electrophoresis-based multiplex PCR assay for simultaneous detection of three types of deer meat could authenticate deer species labeled on products, thus protecting consumers from meat adulteration.
**Keywords:** red deer; roe deer; water deer; multiplex PCR; capillary electrophoresis
### **1. Introduction**
The inaccurate information on the meat species in meat products has been globally concerned by consumers and regulatory agencies [1,2]. Since it is illegal to substitute meat species undeclared on the label of meat products, food manufactures must authenticate correct ingredients declared on their products [3,4]. In the meat industry, game meat consumed commercially is more expensive than meat from domesticated animals. This is because game meat has high nutritional value, such as higher protein and lower fat levels. In addition, it does not contain residues of antibiotics or growth hormones [3,5,6]. Accordingly, replacing game meat with relatively cheaper domesticated meat has taken place for the economic benefit [5]. For game meat products containing deer species, red deer (*Cervus elaphus*) and roe deer (*Capreolus capreolus*) are commonly used, meaning that these species are particularly susceptible to fraudulent labeling [7,8]. Several European countries traditionally permit game hunting [7]. Meanwhile, in Korea, wild animals, such as water deer (*Hydropotes inermis*), that damage crops can be temporarily hunted. However, their distribution and sale are limited, according to the Ministry of Environment guideline. In addition, water deer cannot be used as raw meat or processed food in Korea. To prevent food adulteration, an authentication method for differentiating red deer, roe deer, and water deer is essential.
Methods for detecting meat species have been developed based on DNA molecules and proteins [1,9]. Protein-based methods for deer species authentication have been used by enzyme-linked immunosorbent assay (ELISA), high-performance liquid chromatography (HPLC), and liquid chromatography-mass spectrometry (LC-MS) [10–12]. However, the thermal stability of nucleic acids compared to proteins can increase the amplification efficiency of target species in processed foods [13,14]. PCR, a representative DNA-based detection method, has been utilized for species identification in various fields [15–18]. For deer species, PCR-based detection methods, such as conventional PCR and real-time PCR, have been developed [3,8,19]. To differentiate closely related animal species, the development of specific primers for a target species is very crucial. Mitochondrial DNAs, such as cytochrome b, 12 S rRNA, and D-loop, are commonly used as target genes due to their sequence variations [2,20–22]. Furthermore, to increase the sensitivity of the DNA-based detection method in processed foods, a short fragment of PCR amplification is required because of DNA degradation during the manufacturing process [22,23]. Meanwhile, a multiplex PCR can simultaneously detect several species in a single reaction tube, resulting in effective detection [15,24,25]. Recently, to clearly separate similar sizes of amplicons of short PCR products, multiplex PCR methods combined with capillary electrophoresis have been developed and applied to simultaneously identify various target species [15,26].
The aim of this study was to develop a capillary electrophoresis-based multiplex PCR (CE-mPCR) method to verify the presence of wild animal species, such as red deer, roe deer, and water deer, in processed foods. The developed assay not only saves time and labor because it can simultaneously detect three target species but also can be utilized as a specific and sensitive method for a clear separation of these three species.
#### **2. Materials and Methods**
#### *2.1. Sample Preparation*
Raw tissue samples of 10 animal species (red deer: *Cervus elaphus*, water deer: *Hydropotes inermis*, roe deer: *Capreolus capreolus*, beef: *Bos taurus*, pork: *Sus scrofa domestica*, lamb: *Ovis aries*, goat: *Capra hircus*, horse: *Equus caballus*, chicken: *Gallus gallus*, and duck: *Anas platyrhynchos*) were collected from the Conservation Genome Resource Bank (CGRB, Seoul, Korea) or purchased from online and local markets of Korea. All samples were cut into small pieces and immediately stored at −20 ◦C until analysis.
#### *2.2. DNA Extraction*
DNAs were extracted from meat samples of animal species and processed products using a DNeasy Blood and Tissue Kit (Qiagen, Hilden, Germany), according to the manufacturer's instructions with slight modifications. For good quality of DNA, 25 mg of meat was ground, and all buffers for extraction were used at double quantity. The purity and concentration of extracted DNAs were measured with a Maestro spectrophotometer (Maestro, Las Vegas, NV, USA). DNAs with a 260/280 nm ratio between 1.8 and 2.0 were used as templates for PCR.
#### *2.3. Primer Design*
To select species-specific regions for red deer, roe deer, and water deer, nucleotide sequences of target genes of 19 various animals were downloaded from the GenBank database (Table S1) and aligned using Clustal Omega program (http://www.ebi.ac.uk/Tools/msa/clustalo/) (Figure 1). Species-specific primer sets were newly designed using Primer Designer, version 3.0 (Scientific and Educational Software, Durham, NC, USA). Primers used in this study are listed in Table 1. They were synthesized by Bionics (Seoul, Korea).
*Foods* **2020**, *9*, x FOR PEER REVIEW 3 of 12
**Figure 1.** The sequence alignment of red deer (**A**), roe deer (**B**), and water deer (**C**) specific primers in the mitochondrial cytochrome b, 12 S rRNA, and D-loop **Figure 1.** regions against various animal species. The sequence alignment of red deer (**A**), roe deer (**B**), and water deer (**C**) specific primers in the mitochondrial cytochrome b, 12 S rRNA, and D-loop regions against various animal species.
**Table 1.** Primers used in this study.
#### *2.4. Single and Multiplex PCR Conditions*
Single PCR was performed in a 25 µL final volume containing 10 × Buffer (Bioneer, Daejeon, Korea), 10 mM of dNTPs (Bioneer), 5 units of Hot Start *Taq* DNA polymerase (Bioneer), 0.4 µM of each primer, and 10 ng of DNA template. PCR reaction was carried out in a thermal cycler (Model PC 808, ASTEC, Fukuoka, Japan) as follows: pre-denaturation at 95 ◦C for 5 min, followed by 35 cycles of 95 ◦C for 30 s, 60 ◦C for 30 s, and 72 ◦C for 30 s, with a final extension step at 72 ◦C for 5 min.
PCR mixture for multiplex PCR was similar to single PCR except that it used 10 units of Hot Start *Taq* DNA polymerase (Bioneer) and optimized concentrations of primers. Annealing temperature concentrations of primers were optimized, considering specificity between three deer species. The annealing temperatures were estimated at 58, 59, 60, and 61 ◦C, and the red deer/roe deer/water deer primers combinations were 0.2/0.4/0.4, 0.2/0.4/0.5, and 0.4/0.4/0.4 µM. Finally, 0.2 µM of primers for red deer and 0.4 µM of primers for roe deer and water deer were used for multiplex PCR. Multiplex PCR reactions were carried out under the same conditions as single PCR. All PCR amplicons were electrophoresed on 3% agarose gels stained with ethidium bromide at 150 V for 25 min and confirmed by capillary electrophoresis using an Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA) with DNA 1000 Lab Chip kit (Agilent Technologies). Briefly, 1 µL of PCR product and 5 µL of markers were loaded into each of the 12 wells and applied with a gel-dye mix in the chip, which was run in the bioanalyzer.
#### *2.5. Specificity and Sensitivity of Multiplex PCR*
The specificity of each primer set was performed using DNAs (10 ng each) isolated from 10 animal samples, including red deer, roe deer, and water deer. The specificity of the developed multiplex PCR was conducted using DNAs of the three target species to determine whether there was any cross-reactivity between closely related species.
The sensitivity of multiplex PCR was estimated using serially diluted DNAs (from 10 ng to 0.01 pg per reaction) of the three target species. Detection limits were tested using meat DNA mixtures. The ratio of DNA used in the mixture is shown in Table 2. This test was validated independently using different PCR instruments by different operators. All PCR reactions included a positive control (target DNA) and negative control (no-template).
**Table 2.** The ratio of meat DNA mixtures used in this study.
### **3. Results and Discussion**
#### *3.1. The Specificity of Newly Designed Species-Specific Primers*
In this study, the species-specific primer sets targeting mitochondrial genes of cytochrome b, 12 S rRNA, and D-loop for red deer, roe deer, and water, respectively, were newly designed. As shown in Figure 1, sequences of each target species were compared with two closely related species and 16 other animal species. Considering the intraspecific variation of target species, each primer was selected to have specific sequences of target species (Figure 1). Primer design is very important in the development of multiplex PCR because the primer has to selectively amplify the target in a single reaction containing several primer sets [12]. For multiplex PCR, the sizes of PCR products amplified by each primer set were different for the three target species (79, 126, and 160 bp for red deer, roe deer, and water deer, respectively, Table 1). Each set of species-specific primers amplified only the target species without showing cross-reactivity with nine other species (Figure 2), demonstrating high primer specificity for the target species.
### *3.2. Specificity and Sensitivity of Capillary Electrophoresis-Based Multiplex PCR*
Using these newly designed primers for the identification of red deer, roe deer, and water deer, a multiplex PCR was first optimized by adjusting the concentration of each primer and annealing temperature of PCR condition. The specificity of this optimized assay was then evaluated using DNAs isolated from 10 animal species. As shown in Figure 3, each primer set for red deer, roe deer, and water deer in CE-mPCR specifically amplified target species, showing a high resolution between target species. These results indicated that these red deer-, roe deer-, and water deer-specific primers were sufficient to differentiate these closely related three species by multiplex PCR without causing any cross-amplification against non-target species.
The sensitivity of this CE-mPCR developed in this study was evaluated using DNA at different amounts ranging from 10 ng to 0.01 pg. The results are shown in Figure 4. In lane 6 of Figure 4, peaks of electropherogram were detected for red deer and water deer, but the peak of roe deer was not detected in lane 6 and shown in lane 5. Therefore, sensitivities for red deer, roe deer, and water
**3. Results and Discussion**
*3.1. The Specificity of Newly Designed Species-Specific Primers*
deer were 0.1, 1, and 0.1 pg, respectively. Such high sensitivity of this assay might lead to accurate and reliable detection and differentiation of meat from three target deer species. target species without showing cross-reactivity with nine other species (Figure 2), demonstrating high primer specificity for the target species.
deer, and water deer, respectively, Table 1). Each set of species-specific primers amplified only the
*Foods* **2020**, *9*, x FOR PEER REVIEW 5 of 12
In this study, the species-specific primer sets targeting mitochondrial genes of cytochrome b, 12 S rRNA, and D-loop for red deer, roe deer, and water, respectively, were newly designed. As shown in Figure 1, sequences of each target species were compared with two closely related species and 16 other animal species. Considering the intraspecific variation of target species, each primer was selected to have specific sequences of target species (Figure 1). Primer design is very important in the development of multiplex PCR because the primer has to selectively amplify the target in a single
**Figure 2.** Electropherograms of specificity results of the single PCRs using newly designed primer sets for red deer (**A**), roe deer (**B**), and water deer (**C**). FU: fluorescence, M: alignment marker, lane 1: red deer, 2: roe deer, 3: water deer, 4: beef, 5: pork, 6: lamb, 7: goat, 8: horse, 9: chicken, 10: duck, and N: non-template. **Figure 2.** Electropherograms of specificity results of the single PCRs using newly designed primer sets for red deer (**A**), roe deer (**B**), and water deer (**C**). FU: fluorescence, M: alignment marker, lane 1: red deer, 2: roe deer, 3: water deer, 4: beef, 5: pork, 6: lamb, 7: goat, 8: horse, 9: chicken, 10: duck, and N: non-template.
*Foods* **2020**, *9*, x FOR PEER REVIEW 7 of 12
**Figure 3.** Electropherograms of specificity result of the multiplex PCR assay for red deer, roe deer, and water deer. FU: fluorescence, M: alignment marker, lane P: positive control (red deer, roe deer, and water deer), lane 1: red deer, 2: roe deer, 3: water deer, 4: beef, 5: pork, 6: lamb, 7: goat, 8: horse, 9: chicken, 10: duck, and N: non-template. **Figure 3.** Electropherograms of specificity result of the multiplex PCR assay for red deer, roe deer, and water deer. FU: fluorescence, M: alignment marker, lane P: positive control (red deer, roe deer, and water deer), lane 1: red deer, 2: roe deer, 3: water deer, 4: beef, 5: pork, 6: lamb, 7: goat, 8: horse, 9: chicken, 10: duck, and N: non-template.
*Foods* **2020**, *9*, x FOR PEER REVIEW 8 of 12
**Figure 4.** Electropherograms of sensitivity results of the multiplex PCR assay. FU: fluorescence, M: alignment marker, lanes 1–7: 1.0 × 101, 100, 10−1, 10−2, 10−3, 10−4, and 10<sup>−</sup>5 ng of three target species, and lane N: non-template. **Figure 4.** Electropherograms of sensitivity results of the multiplex PCR assay. FU: fluorescence, M: alignment marker, lanes 1–7: 1.0 <sup>×</sup> <sup>10</sup><sup>1</sup> , 10<sup>0</sup> , 10−<sup>1</sup> , 10−<sup>2</sup> , 10−<sup>3</sup> , 10−<sup>4</sup> , and 10−<sup>5</sup> ng of three target species, and lane N: non-template.
#### *3.3. Application and Validation of Capillary Electrophoresis-Based Multiplex PCR Using Meat DNA Mixtures*
To determine detection limits of CE-mPCR and confirm its applicability to a real sample, two sets of meat DNA mixtures were prepared as follows: (1) roe deer and red deer commonly used as game meat were added in water deer to authenticate game meat species present in commercial deer meats, and (2) red deer was contaminated with roe deer and water deer to detect wild animal species not permitted commercially in several countries. As shown in Figure 5, the detection limit of this assay was at least 0.1% for roe deer and red deer in meat DNA mixtures. In another meat DNA mixture, as low as 0.1% of roe deer and water deer could be detected (Figure 6). Microchip-based capillary electrophoresis technology used in this study is known to provide better accuracy and resolution in multiple target detection [15]. In the present study, at a low concentration of 0.1% for water deer (lane 6 in Figure 6), the result obtained by capillary electrophoresis was clearer than a PCR band visualized on agarose gel (Figure S1). This can help overcome the drawback, such as a false-negative result. In addition, the detection limit was validated independently in duplicate. All results obtained through intra-laboratory validation analysis were similar. The 0.1% of roe deer and red deer mixed in water deer and 0.1% of roe deer and water deer mixed in red deer were detected in two independent PCR reactions using the developed primer sets. Thus, this CE-mPCR assay developed in this study was able to simultaneously detect up to 0.1% of red deer, roe deer, and water deer in meat DNA mixtures. Compared to the limit of detection of 0.1% for roe deer and red deer [4] and 0.5% for red deer [3], our method showed higher or similar sensitivity. Meanwhile, since this was the first study to apply a detection method for water deer to meat DNA mixtures, the detection limit of 0.1% of our developed method could not be compared to previous reports. However, this method might be sufficient to be utilized as a specific and sensitive molecular tool for monitoring these three types of deer meat.
**Figure 5.** Detection limits of red deer and roe deer in water deer by the multiplex PCR assay. Gel image (**A**) and electropherograms (**B**). FU: fluorescence, M: alignment marker, lane L: 100 bp DNA ladder, lane 1: positive control (10 ng of DNA from target species), lanes 2–6: 10, 5, 1, 0.5, and 0.1% red deer and roe deer in water deer, and lane N: non-template. a, b, and c indicate red deer, roe deer, and water deer, respectively. *Foods* **2020**, *9*, x FOR PEER REVIEW 10 of 12
**Figure 6.** Detection limits of roe deer and water deer in red deer by the multiplex PCR assay. Gel image (**A**) and electropherograms (**B**). FU: fluorescence, M: alignment marker, lane L: 100 bp DNA ladder, lane 1: positive control (10 ng of each DNA from target species), lanes 2–6: 10, 5, 1, 0.5, and 0.1% roe deer and water deer in red deer, and lane N: non-template. a, b, and c indicate red deer, roe deer, and water deer, respectively. **Figure 6.** Detection limits of roe deer and water deer in red deer by the multiplex PCR assay. Gel image (**A**) and electropherograms (**B**). FU: fluorescence, M: alignment marker, lane L: 100 bp DNA ladder, lane 1: positive control (10 ng of each DNA from target species), lanes 2–6: 10, 5, 1, 0.5, and 0.1% roe deer and water deer in red deer, and lane N: non-template. a, b, and c indicate red deer, roe deer, and water deer, respectively.
#### **4. Conclusions 4. Conclusions**
Drug Safety in Korea.
**Conflicts of Interest:** The authors declare no conflict of interest.
The CE-mPCR assay developed in this study could successfully detect three types of deer meat. Its applicability for authentication of meat species was verified using various ratios of meat DNA mixtures. This method is simple and user-friendly. It has high specificity and sensitivity for the simultaneous detection of red deer, roe deer, and water deer. However, despite several advantages of this method developed, since it is utilized for only qualitative detection, further study is required to the application of real-time PCR to quantify meat of target deer species in processed game meat. **Supplementary Materials:** The following are available online at www.mdpi.com/xxx/s1. Figure S1: Detection limits of the multiplex PCR assay. Lane M: 100 bp DNA ladder, lanes 1: positive control (10 ng of each DNA The CE-mPCR assay developed in this study could successfully detect three types of deer meat. Its applicability for authentication of meat species was verified using various ratios of meat DNA mixtures. This method is simple and user-friendly. It has high specificity and sensitivity for the simultaneous detection of red deer, roe deer, and water deer. However, despite several advantages of this method developed, since it is utilized for only qualitative detection, further study is required to the application of real-time PCR to quantify meat of target deer species in processed game meat.
from target species), lanes 2-6: 10, 5, 1, 0.5, and 0.1% roe deer and water deer in red deer, and lane N: nontemplate. Table S1: Mitochondrial gene sequences of various animals used for the sequence alignment. **Author Contributions:** Formal analysis, M.-J.K., Y.-M.L., and S.-M.S.; Funding acquisition, H.-Y.K.; Methodology, M.-J.K.; Project administration, H.-Y.K.; Supervision, H.-Y.K.; Validation, Y.-M.L.; Writingoriginal draft, M.-J.K.; Writing-review and editing, M.-J.K., Y.-M.L., S.-M.S., and H.-Y.K. **Supplementary Materials:** The following are available online at http://www.mdpi.com/2304-8158/9/8/982/s1. Figure S1: Detection limits of the multiplex PCR assay. Lane M: 100 bp DNA ladder, lanes 1: positive control (10 ng of each DNA from target species), lanes 2-6: 10, 5, 1, 0.5, and 0.1% roe deer and water deer in red deer, and lane N: non-template. Table S1: Mitochondrial gene sequences of various animals used for the sequence alignment.
**Funding:** This research was funded by the Ministry of Food and Drug Safety in Korea, grant number 17162MFDS065. **Acknowledgments:** This research was supported by a grant (17162MFDS065) from the Ministry of Food and **Author Contributions:** Formal analysis, M.-J.K., Y.-M.L., and S.-M.S.; Funding acquisition, H.-Y.K.; Methodology, M.-J.K.; Project administration, H.-Y.K.; Supervision, H.-Y.K.; Validation, Y.-M.L.; Writing-original draft, M.-J.K.; Writing-review and editing, M.-J.K., Y.-M.L., S.-M.S., and H.-Y.K. All authors have read and agreed to the published version of the manuscript.
**Funding:** This research was funded by the Ministry of Food and Drug Safety in Korea, grant number 17162MFDS065.
**Acknowledgments:** This research was supported by a grant (17162MFDS065) from the Ministry of Food and Drug Safety in Korea.
**Conflicts of Interest:** The authors declare no conflict of interest.
### **References**
© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
## **Development of a Real-Time PCR Assay for the Detection of Donkey (***Equus asinus***) Meat in Meat Mixtures Treated under Di**ff**erent Processing Conditions**
#### **Mi-Ju Kim** † **, Seung-Man Suh** † **, Sung-Yeon Kim, Pei Qin, Hong-Rae Kim and Hae-Yeong Kim \***
Department of Food Science and Biotechnology, Institute of Life Sciences & Resources, Kyung Hee University, Yongin 17104, Korea; [email protected] (M.-J.K.); [email protected] (S.-M.S.);
[email protected] (S.-Y.K.); [email protected] (P.Q.); [email protected] (H.-R.K.)
**\*** Correspondence: [email protected]; Tel.: +82-31-201-2660; Fax: +82-31-204-8116
† These authors contributed equally to the work.
Received: 25 December 2019; Accepted: 23 January 2020; Published: 26 January 2020
**Abstract:** In this study, a donkey-specific primer pair and probe were designed from mitochondrial *cytochrome b* gene for the detection of raw donkey meat and different processed meat mixtures. The PCR product size for donkey DNA was 99 bp, and primer specificity was verified using 20 animal species. The limit of detection (LOD) was examined by serially diluting donkey DNA. Using real-time PCR, 0.001 ng of donkey DNA could be detected. In addition, binary meat mixtures with various percentages of donkey meat (0.001%, 0.01%, 0.1%, 1%, 10%, and 100%) in beef were analyzed to determine the sensitivity of this real-time PCR assay. At least 0.001% of donkey meat was detected in raw, boiled, roasted, dried, grinded, fried, and autoclaved meat mixtures. The developed real-time PCR method showed sufficient specificity and sensitivity in identification of donkey meat and could be a useful tool for the identification of donkey meat in processed products.
**Keywords:** food adulteration; food fraud; donkey; *cytochrome b*; real-time PCR; meat products
### **1. Introduction**
Identification of animal species in meat products is important for preventing food adulteration and providing accurate information regarding meat species to consumers. Donkey meat products are highly nutritious; moreover, in many countries, including Korea, it is considerably more expensive than other meats owing to its low supply [1]. In Islamic countries, donkey meat consumption is prohibited on religious grounds [2]. Due to donkey meat being expensive, it is likely to be mixed with other cheaper meats for economic benefits, and there is a need to avoid donkey meat entering the food chain in Islamic countries. Therefore, it is necessary to develop reliable and specific detection methods that can accurately identify animal species from meat products to prevent cases of disguising meat from one species as another [3,4].
To date, many protein- and DNA-based detection methods have been developed to determine animal species in food products. In particular, DNA-based methods have been used to detect target species in processed foods because DNA is stable at high temperatures and pressures [5–7]. Real-time PCR is an effective tool that accurately amplifies target DNA. Several real-time PCR methods, particularly based on detection via TaqMan probes, have been developed with high sensitivity and accuracy to distinguish common meat species such as pork, lamb, and beef [8–12].
Mitochondrial DNA (mtDNA) has been mainly used to detect target species in meat products, and mtDNA sequences from related species have been phylogenetically studied [13]. In addition, because mtDNA evolves faster than nuclear DNA, mtDNA has been used to discriminate target species from similar species. Further, mitochondria are present in high copy numbers in cells. Thus, real-time PCR based on specific mtDNA sequences can amplify target DNA degraded by food processing or mixed with other species [14,15]. In many studies, mitochondrial *cytochrome b* has been used to develop species-specific real-time PCR detection methods [12,16,17].
Here, we designed a donkey-specific primer and probe from mitochondrial *cytochrome b* and developed a real-time PCR method to accurately identify donkey meat. Although there have been several previous studies for donkey meat detection [1,9], no study has applied donkey meats treated under a variety of processing conditions. Thus, in this study, we evaluated the applicability of the developed method for the detection of donkey meat using raw, boiled, roasted, dried, grinded, fried, and autoclaved meat mixtures.
#### **2. Materials and Methods**
#### *2.1. Preparation of Samples and Binary Meat Mixtures*
A total of 20 raw meat samples were obtained from the Conservation Genome Resource Bank for Korean Wildlife (CGRB), the National Institute of Animal Science (NIAS), and local markets in South Korea. All samples were homogenized in small pieces and stored at −20 ◦C until analysis.
Binary meat mixtures were prepared to determine the detection limit of donkey-specific real-time PCR assay. For binary raw meat mixtures, 10 g of each of donkey and beef was lyophilized for 24 h using a freeze dryer (Ilsin Biobase, Dongduchon, Korea) to remove moisture of raw meats without DNA degradation, and then ground. In addition, to evaluate the applicability of the developed method in processed meat products, two meats were treated under six different processing conditions as follows: (1) boiled at 100 ◦C for 15 min in water bath (MONO-TECH, Daegu, Korea), (2) roasted at 200 ◦C for 5 min in hot plate (Corning Co., New York, NY, USA), (3) dried at 65 ◦C for 12 h in dry oven (HANKUK S&I, Hwaseong, Korea), (4) grinded for 5 min in commercial grinder (Buwon Electronics, Daegu, Korea), (5) fried at 180 ◦C for 5 min in cooking oil, and (6) autoclaved at 121 ◦C 150 kPa for 30 min. Each meat for the six different mixtures was prepared in triplicate, which was made on different days and from meats of different origins. After treatments, each of binary meat mixtures containing six different percentages (0.001%, 0.01%, 0.1%, 1%, 10%, and 100% (*w*/*w*)) of donkey meats in beef was prepared. Samples (final weight, 100 mg) of various meat mixtures were used for analysis.
#### *2.2. DNA Extraction*
Genomic DNA was extracted from raw and autoclaved meat mixtures using the DNeasy Blood and Tissue kit (Qiagen, Hilden, Germany), following the manufacturer's instructions with minor modifications. Briefly, 100 mg of each sample was lysed with 3600 µL of ATL buffer and 400 µL of proteinase K (20 mg/mL) in a water bath at 56 ◦C for 1 h. After adding 40 µL of RNase A (100 mg/mL), the mixture was incubated at room temperature for 2 min. AL buffer (4000 µL) and 100% ethanol (4000 µL) were mixed with the DNA mixture, and the sample was centrifuged through a spin column. After washing with AW1 and AW2 buffers, the column-bound DNA was eluted with purified water. The purity and concentration of the extracted DNA were confirmed using a Maestro Nano spectrophotometer (Maestrogen, Las Vegas, NV, USA).
#### *2.3. Primer and Probe Design*
A donkey-specific primer pair and probe for the detection of donkey were designed to amplify the specific target DNA (Table 1). To design a donkey-specific primer pair, mitochondrial *cytochrome b* sequences from 20 animal species, including donkey, beef, and horse (Accession No.: FJ428510.1, D34635.1, and MH594485.1, respectively) were obtained from GenBank. All sequences were aligned using Clustal Omega program (http://www.ebi.ac.uk/Tools/msa/clustalo/). The primer pair and probe
were designed using the Primer Designer program version 3.0 (Scientific and Education Software, Durham, NC, USA) and synthesized by Bionics (Seoul, Korea) and Bioneer (Daejeon, Korea).
### *2.4. Conventional PCR Reaction*
Conventional PCR was performed using a thermal cycler (PC808, ASTEC, Kyoto, Japan) under the following conditions: pre-incubation at 94 ◦C for 5 min, 30 cycles of denaturation for 30 s at 94 ◦C, annealing for 30 s at 60 ◦C, extension for 30 s at 72 ◦C, and final extension for 5 min at 72 ◦C. The PCR reaction mixture comprised 400 nM of each primer, 0.5 U of Ampli-Gold Taq polymerase (Applied Biosystems, Foster City, CA, USA), 10× PCR buffer (Applied Biosystems), 2.5 mM of each dNTP (Applied Biosystems), 1.5 mM of MgCl<sup>2</sup> (Applied Biosystems), and 10 ng of DNA template isolated from each of animal species for the specificity test in a total reaction volume of 25 µL. All PCR products were electrophoresed on a 2% agarose gel and then visualized under UV irradiation.
### *2.5. Real-Time PCR Reaction*
Real-time PCR amplification was performed using an ABI 7500 Real-time PCR instrument (Applied Biosystems). The PCR reaction was performed in a final volume of 25 µL, containing 2× TaqMan Universal Master mix (Applied Biosystems), 400 nM of primer pairs, 200 nM of the probe, and 10 ng of the DNA template. Real-time PCR was performed with a holding stage at 95 ◦C for 10 min, followed by 40 cycles at 95 ◦C for 15 s and 60 ◦C for 1 min. All real-time PCR reactions were performed in triplicates; no-template control (NTC) was used as a negative control. Data were analyzed using 7500 Software V.2.3 (Applied Biosystems).
#### *2.6. Specificity and Sensitivity of Real-Time PCR*
The specificity of the donkey-specific primer pair and probe was tested using 10 ng of genomic DNA extracted from 20 animal species. To confirm the presence of DNA, endogenous primer pair and probe targeting the 18S rRNA gene were also used [18]. The sensitivity of the real-time PCR was measured using 10-fold serially diluted DNA (from 10 to 0.001 ng) extracted from donkey. The detection limit of real-time PCR in 6 processed binary mixtures containing donkey meat (ranging in concentration from 10% to 0.001%) mixed with beef meat was used.
#### **3. Results and Discussions**
### *3.1. Specificity*
The donkey-specific primer and probe were designed to get a small product size of 99 bp from mitochondrial *cytochrome b*. The specificity of donkey primer set was confirmed using the DNA from 20 animal species as templates for this assay. Only the DNA fragment specific for donkey was amplified by conventional PCR, and there was no amplification in 19 nontarget species (Table 2). The PCR product amplified by the donkey-specific primer was sequenced to verify the donkey species (Figure S1). The specificity of the real-time PCR method was additionally verified, and the donkey DNA was
specifically amplified without any cross-reactivity against the 19 other animal species tested (Table 2). To confirm the presence of DNA, eukaryotic PCR targeting the 18S rRNA gene was performed. As shown in Table 2, positive signals were observed in all PCR reactions. Thus, our results proved that the donkey-specific primer and probe were accurately amplified the target DNA.
**Table 2.** Specificity results using conventional and real-time PCR assays.
### *3.2. Sensitivity of the Donkey-Specific Real-Time PCR Assay*
The sensitivity of the donkey-specific real-time PCR targeting *cytochrome b* gene was determined using 10-fold serially diluted donkey DNA from 10 to 0.001 ng. Ct values were plotted against logarithmic DNA concentrations to construct the standard curve for the donkey DNA. The slope and correlation coefficient (R<sup>2</sup> ) of the standard curve were −3.79 and 0.997, respectively. PCR efficiency was calculated using the equation "E <sup>=</sup> (10(−1/slope) <sup>−</sup> 1)" and was determined to be 83.47% (Figure 1). Each PCR reaction was performed thrice, and 0.001 ng of the donkey DNA was detected in all the reactions. The absolute detection limit of the donkey-specific real-time PCR assay was as low as 0.001 ng. These results demonstrated that the real-time PCR method developed in this study has good linearity and sensitivity.
*3.2. Sensitivity of the Donkey-specific Real-time PCR Assay*
The sensitivity of the donkey-specific real-time PCR targeting *cytochrome b* gene was determined using 10-fold serially diluted donkey DNA from 10 to 0.001 ng. Ct values were plotted against logarithmic DNA concentrations to construct the standard curve for the donkey DNA. The slope and correlation coefficient (R2) of the standard curve were −3.79 and 0.997, respectively. PCR efficiency was calculated using the equation "E = (10(−1/slope) – 1)" and was determined to be 83.47% (Figure 1). Each PCR reaction was performed thrice, and 0.001 ng of the donkey DNA was detected in all the reactions. The absolute detection limit of the donkey-specific real-time PCR assay was as low as 0.001
**Figure 1.** Amplification plot (**A**) and standard curve (**B**) for the detection of donkey DNA using 10 fold serial dilutions (from 10 to 0.001 ng). **Figure 1.** Amplification plot (**A**) and standard curve (**B**) for the detection of donkey DNA using 10-fold serial dilutions (from 10 to 0.001 ng).
#### *3.3. Application of the Real-time PCR Assay to Meat Mixtures Treated under Different Processing 3.3. Application of the Real-Time PCR Assay to Meat Mixtures Treated under Di*ff*erent Processing Conditions*
*Conditions* The meat mixtures treated under six conditions were used to confirm the applicability of the developed method using highly processed meat products as well as raw meat, and various concentrations of binary meat mixtures were prepared for the limit of detection (LOD) test of this method. As shown in Table 3, 0.001% of donkey meat was successfully detected in all processed meat mixtures, despite high heat and pressure treatments of donkey meat. The average Ct values of three replicates using donkey DNA were 18.45 ± 0.7, 20.24 ± 0.97, 18.74 ± 0.06, 18.59 ± 0.31, 19.17 ± 0.6, 21.17 ± 0.55, and 20.86 ± 0.26 for raw, boiled, roasted, dried, ground, fried, and autoclaved meats, respectively. Ct values of the target species in the boiled, fried, and autoclaved meat mixtures were relatively higher than in other meat mixtures; this may be attributable to the fact that the DNA was The meat mixtures treated under six conditions were used to confirm the applicability of the developed method using highly processed meat products as well as raw meat, and various concentrations of binary meat mixtures were prepared for the limit of detection (LOD) test of this method. As shown in Table 3, 0.001% of donkey meat was successfully detected in all processed meat mixtures, despite high heat and pressure treatments of donkey meat. The average Ct values of three replicates using donkey DNA were 18.45 ± 0.7, 20.24 ± 0.97, 18.74 ± 0.06, 18.59 ± 0.31, 19.17 ± 0.6, 21.17 ± 0.55, and 20.86 ± 0.26 for raw, boiled, roasted, dried, ground, fried, and autoclaved meats, respectively. Ct values of the target species in the boiled, fried, and autoclaved meat mixtures were relatively higher than in other meat mixtures; this may be attributable to the fact that the DNA was degraded under the high pressure and temperature treatments [7,19].
degraded under the high pressure and temperature treatments [7,19]. The lowest percentage of donkey meat adulteration that could be detected by the real-time PCR method developed in this study was 0.001%, which was lower than 1% of detection limit reported by Chen et al. [1] and same or lower than 0.001% and 0.01% of detection limits reported by Kesmen et al. [19]. Therefore, this real-time PCR method can help to confirm the presence of donkey meat in highly processed meat products and provide accurate information on target meat species. For a more efficient detection method tool, a further study can be performed the development of multiplex real-The lowest percentage of donkey meat adulteration that could be detected by the real-time PCR method developed in this study was 0.001%, which was lower than 1% of detection limit reported by Chen et al. [1] and same or lower than 0.001% and 0.01% of detection limits reported by Kesmen et al. [19]. Therefore, this real-time PCR method can help to confirm the presence of donkey meat in highly processed meat products and provide accurate information on target meat species. For a more efficient detection method tool, a further study can be performed the development of multiplex real-time PCR for the detection of two genes, including the endogenous 18S rRNA gene.
time PCR for the detection of two genes, including the endogenous 18S rRNA gene.
Average Ct value ± standard deviation obtained from triplicate reactions.
### **4. Conclusions**
This study described the development of a real-time PCR method to identify donkey DNA. By targeting a 99 bp fragment of mitochondrial *cytochrome b*, the designed primer pair and probe specifically amplified the donkey DNA. The standard curve of the developed real-time PCR method has good linearity and sensitivity, which is adequate to successfully amplify the target DNA. Raw and highly processed meat mixtures were analyzed with a sensitivity of 0.001% to demonstrate the applicability of the method developed in the present study for detecting donkey meat in processed meat products. The applicability of this method was verified with six processing conditions that can be used for meat processing, and the applicability was confirmed under all processing conditions. Therefore, the real-time PCR method developed in this study could be a useful tool for the detection of donkey and determination of intentional adulterations or food fraud in highly processed meat products.
**Supplementary Materials:** The following are available online at http://www.mdpi.com/2304-8158/9/2/130/s1, Figure S1: The identity result of sequences of the PCR products for donkey-specific primer sets.
**Author Contributions:** H.-Y.K. conceived of the overall study. M.-J.K., S.-Y.K., S.-M.S., P.Q., and H.-R.K. carried out the experiment. M.-J.K. wrote the first draft of the manuscript. SMS reviewed the manuscripts. All authors have read and agreed to the published version of the manuscript.
**Funding:** This research was supported by grant 17162MFDS065 from the Ministry of Food and Drug Safety in Korea.
**Conflicts of Interest:** The authors declare no conflict of interest.
#### **References**
© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
*Article*
## **Metabolite Profiling and Chemometric Study for the Discrimination Analyses of Geographic Origin of Perilla (***Perilla frutescens***) and Sesame (***Sesamum indicum***) Seeds**
**Tae Jin Kim 1,**† **, Jeong Gon Park 1,**† **, Hyun Young Kim <sup>2</sup> , Sun-Hwa Ha <sup>3</sup> , Bumkyu Lee <sup>4</sup> , Sang Un Park <sup>5</sup> , Woo Duck Seo 2,\* and Jae Kwang Kim 1,\***
Received: 26 June 2020; Accepted: 21 July 2020; Published: 24 July 2020
**Abstract:** Perilla and sesame are traditional sources of edible oils in Asian and African countries. In addition, perilla and sesame seeds are rich sources of health-promoting compounds, such as fatty acids, tocopherols, phytosterols and policosanols. Thus, developing a method to determine the geographic origin of these seeds is important for ensuring authenticity, safety and traceability and to prevent cheating. We aimed to develop a discriminatory predictive model for determining the geographic origin of perilla and sesame seeds using comprehensive metabolite profiling coupled with chemometrics. The orthogonal partial least squares-discriminant analysis models were well established with good validation values (*Q*<sup>2</sup> = 0.761 to 0.799). Perilla and sesame seed samples used in this study showed a clear separation between Korea and China as geographic origins in our predictive models. We found that glycolic acid could be a potential biomarker for perilla seeds and proline and glycine for sesame seeds. Our findings provide a comprehensive quality assessment of perilla and sesame seeds. We believe that our models can be used for regional authentication of perilla and sesame seeds cultivated in diverse geographic regions.
**Keywords:** perilla; sesame; geographic origin; metabolomics; multivariate analysis; metabolite profiling
### **1. Introduction**
Perilla (*Perilla frutescens*) seed is a rich source of health-promoting compounds, such as tocopherols, phytosterols, policosanols and fatty acids, which have various bioactivities [1]. Tocopherols have an antioxidant effect and are known as vitamin E. Phytosterols show reduction of total cholesterols in the serum. They increase high-density lipoprotein cholesterol levels and reduce low-density lipoprotein cholesterol levels in the blood. Policosanols also have a serum lipid- and cholesterol-lowering effect and other beneficial effects, such as cytoprotection, antiaging, liver protection, antioxidant and anti-parkinsonian effects [2]. In addition, perilla seeds contain high levels of octacosanol (C28-ol) [1,2]. The fatty acid α-linolenic acid (C18:3n3) is found in high levels in perilla seeds, which is essential to human health; moreover, perilla seeds contain omega-3 fatty acid, which lowers inflammation and risk of cancer and cardiovascular and atopic diseases [3]. Sesame (*Sesamum indicum* L.) seeds also contain the abovementioned health-promoting compounds, and they are a good source of proteins rich in sulfur-containing amino acids [4,5]. Linoleic acid (C18:2n6), which is an essential fatty acid for humans, is the main fatty acid found in sesame seeds; in addition, oleic acid (C18:1n9) is the second most abundant fatty acid in sesame seeds [6]. In addition, γ-tocopherol is the main tocopherol in sesame seeds [6,7]. Sesame seeds reportedly contain high levels of phytosterols [5]. Although the composition and contents of various health beneficial compounds in the perilla and sesame seeds have been reported, to the best of our knowledge, a comprehensive comparative-analysis involving hydrophilic and lipophilic compounds has not been reported.
Metabolomics has been widely used to distinguish food products on the basis of differences in their chemical composition and metabolite contents [8,9]. Food metabolomics comprises analytical techniques and multivariate discriminant analysis (MVDA) techniques used for food substances. The analytical techniques usually used in food metabolomics are mass spectrometry (MS) coupled with separation techniques such as liquid chromatography (LC) and gas chromatography (GC) and nuclear magnetic resonance (NMR) [10]. For MVDA, the most commonly used methods are principal component analysis (PCA), partial least squares-discriminant analysis (PLS-DA) and orthogonal partial least squares discriminant analysis (OPLS-DA), which are useful tools for describing correlations and diagnosing differences among the studied samples and their metabolites. Therefore, food metabolomics strategies are suitable for analyzing food safety, authenticity, traceability and quality assessment and these strategies have been used to assess various foods and beverages, such as adzuki bean, olive oil, cabbage, wine, rice, coffee and tomato [11–16].
Perilla and sesame seeds are traditionally used as sources of edible oils in Korea, China, India and other Asian countries. Perilla is cultivated in Korea, China, Japan, India, Nepal and Thailand [17,18]. In Korea, the production of perilla seeds was average 40,448 tons per year over the last decade, and approximately 24,411 tons were imported per year [19]. Out of the imported perilla seeds, almost of 99% are Chinese perilla seeds [20]. Sesame is mainly produced in China, Myanmar, India and African countries such as Sudan, Nigeria and Tanzania. In Korea, the average production of sesame was 12,168 tons over the last decade, whereas approximately 76,812 tons were imported; the self-sufficiency rate in sesame production was 14% [19,20]. In particular, more than 90% of sesame seeds were imported from China (50%) and India (40%) [19]. The price of perilla and sesame seeds is influenced by their places of origin; therefore, identification of the geographic origin of these seeds is important [21]. Forging or mislabeling domestic seeds as imported seeds to gain economic benefits has increasingly become a crucial issue for both producers and consumers, and it affects food quality assurance and safety [22]. To prevent this problem, developing a precise and accurate method to identify the geographic origin of perilla and sesame seeds is needed. Recently, genomic and analytical approaches have been developed for such identification [4,6,15,23–25]. The genomics method is considerably accurate; however, it cannot determine the geographic origins of the same plant variety [14]. On the contrary, the analytical methods can accurately determine the different geographic origins of the same variety based on the differences in chemical composition. Previous studies have used multivariate analysis for discriminating between geographic origins of perilla and sesame seeds using genomics and analytical methods [4,22]. In the case of perilla, however, genomic methods have been reported to determine geographic origin, but analytical methods have not been developed [23].
We aimed to develop a method to discriminate the geographic origin of perilla and sesame seeds and to assess their nutritional quality. To discriminate the geographic origin, MVDA was performed with targeted metabolite profiling using gas chromatography-mass spectrometry (GC-MS). The hydrophilic and lipophilic metabolite profiling (including amino acids, organic acids, sugars, sugar alcohols, tocopherols, sterols, policosanols and fatty acids) of perilla and sesame seeds originated in the Korea and China was performed. Using this, a discrimination model was established for
the determination of geographic origins of perilla and sesame seeds. This is the first attempt to construct a discrimination model for perilla seeds using metabolomics. Further, potential biomarkers for distinguishing the geographic origins of perilla and sesame seeds were proposed. A comprehensive food quality assessment was also performed. Our findings can offer reliable information about food authenticity and traceability of perilla and sesame seeds.
#### **2. Materials and Methods**
#### *2.1. Sample and Chemicals*
Korean perilla and sesame cultivars were grown at the National Institute of Crop Science, Rural Development Administration, Wanju-gun, Korea, during the 2018 growing season (June to November). Chinese perilla and sesame samples were procured from a local market in Xinzhou and JiangXia district (Wuhan city), China. The Chinese samples including perilla and sesame were from the recent harvests of November 2017 and 2016, respectively. Three biologic replicates were prepared for each sample. 5α-Cholestane, ribitol, pentadecanoic acid, fatty acid methyl ester (FAME) mixture, *N*-methyl-*N*-trimethylsilyl trifluoroacetamide (MSTFA) and pyridine were purchased from Sigma-Aldrich (St. Louis, Mo, USA). All other chemicals used in this study were reagent grade unless stated otherwise.
### *2.2. Extraction and Analysis of Hydrophilic Compounds*
The extraction and analysis of hydrophilic compounds was performed as described previously [26]. A finely ground sample (10 mg) was mixed with 1 mL of a mixture of methanol, water and chloroform in the ratio 2.5:1:1 (*v*/*v*/v). Sixty microliters of ribitol (200 µg/mL) was added to the mixture as an internal standard (IS) and the mixture was incubated using a Thermomixer Comfort (model 5355, Eppendorf AG, Hamburg, Germany) at 37 ◦C for 30 min at a mixing frequency of 1200 rpm. The mixture was centrifuged at 16,000× *g* for 3 min. The upper layer (methanol/water phase) of 800 µL was pipetted into a fresh tube and mixed with 400 µL of water. The methanol/water fraction was centrifuged at 16,000× *g* for 3 min and 900 µL of the supernatant was collected into a fresh tube. The aliquots were evaporated for 2 h in a centrifugal concentrator (CC-105; TOMY, Tokyo, Japan) and freeze-dried for over 16 h. For derivatization, 80 µL of 2% methoxyamine hydrochloride (MOX) in pyridine (*w*/*v*) was added in freeze-dried samples and the mixture was incubated at 30 ◦C and 1200 rpm for 90 min using a Thermomixer Comfort (Eppendorf AG). Subsequently, 80 µL of MSTFA was added and the mixture was further incubated at 37 ◦C and 1200 rpm for 30 min. The hydrophilic compounds were separated on the GCMS-QP2010 Ultra system equipped with autosampler AOC-20i (Shimadzu, Kyoto, Japan) and a DB-5 column (30 m length, 0.25-mm diameter and 1.00 µm thickness). The temperatures for injection, interface and ion source were set at 280, 280 and 200 ◦C, respectively. The carrier gas was helium and the column flow rate was 1.1 mL/min. The temperature was held for 4 min at 100 ◦C, after which it was increased at a rate of 10 ◦C/min up to 320 ◦C and held for 11 min. The runtime was 4.00 to 37.00 min and the scan mode was used with a mass range of 45 to 600 *m*/*z*. The compounds were confirmed using standards and the Wiley9, NIST11 and OA TMS DB5 (Shimadzu) libraries (Table S1). For relative quantification, we used ribitol as an IS and the calculated the integrated peak area of all the analyte ratios relative to the peak area of the IS.
#### *2.3. Extraction and Analysis of Lipophilic Compounds*
Extraction and analysis of lipophilic compounds (policosanols, phytosterols, tocopherols and other terpenoids) was performed as described previously [27]. Finely ground samples weighing 10 mg were collected in 15-mL conical tubes, and 3 mL of ethanol containing 0.1% ascorbic acid (*w*/*v*) was added to the tubes. Fifty microliters of 5α-cholestane (10 µg/mL) was added to the mixture as an IS. Next, the samples were vortexed for 20 s and placed in a water bath at 85 ◦C for 5 min. Subsequently, 120 µL of potassium hydroxide (80%, *w*/*v*) was added for saponification, and the mixture was vortexed
for 20 s. The mixture was returned to the water bath at 85 ◦C for 10 min. The samples were then cooled on ice for 5 min, and 1.5 mL each of deionized water and hexane was added to each sample and vortexed for 20 s. The mixture was centrifuged at 1200× *g* for 5 min at 4 ◦C and the upper layer was pipetted into afresh tube. In order to re-extract the remaining compounds, 1.5 mL of hexane was added again into the remaining pellets. The hexane fraction was collected in fresh tubes and evaporated under a stream of N<sup>2</sup> gas in a centrifugal concentrator (TOMY). For the derivatization step, 30 µL of MSTFA and 30 µL of pyridine were added and incubated at 60 ◦C and 1200 rpm for 30 min using a Thermomixer Comfort (model 5355, Eppendorf AG, Hamburg, Germany). The GCMS-QP2010 Ultra system, equipped with the autosampler AOC-20i (Shimadzu), was installed with a Rtx-5MS column (30 m length, 0.25-mm-diameter and 0.25-µm-thickness) and used for the separation of lipophilic compounds. In total 1.0 µL of each sample was injected with split mode (10:1 ratio) and the injection temperature was set at 290 ◦C. Helium was used as a carrier gas and the column flow rate was 1.0 mL/min. The oven temperature was held for 2 min at 150 ◦C, increased at the rate of 15 ◦C/min up to 320 ◦C and finally held for 10 min. The chromatography runtime was 2.00–23.33 min. The MS interface and ion source temperatures were 280 and 230 ◦C, respectively. The Labsolutions GCMSsolution software version 4.20 (Shimadzu Kyoto, Japan) was used for the analysis of chromatograms and mass spectra. The calibration curve range of each lipophilic compound was 0.025–5.00 µg, and a fixed concentration (0.50 µg each) of the internal standard was used. Qualitative and quantitative analyses were conducted using standards (Table S2).
Extraction of fatty acids was performed according to a method described previously, but with slight modifications [28,29]. Briefly, 10-mg of sample was mixed with 2.5 mL of chloroform/methanol (2:1, *v*/*v*) and 10 µL of pentadecanoic acid (100 µg/mL) as an IS. The mixture was sonicated for 15 min. Next, 2.5 mL of 0.58% (*w*/*v*) sodium chloride (NaCl) in water was added to separate the extract into two phases (methanol-water and chloroform) and to remove proteinaceous matter from the chloroform fraction. The mixture was briefly vortexed and then centrifuged at 13,000× *g* for 5 min at 4 ◦C. Thereafter, the chloroform phase (bottom layer) was pipetted into a new tube and evaporated using a centrifugal concentrator (TOMY). Toluene (100 µL), 5 M sodium hydroxide (NaOH, 20 µL) and methanol (180 µL) were added to the dried sample, and the tube was incubated at 85 ◦C for 5 min. Next, 300 µL of 14% (*w*/*v*) boron trifluoride (BF3) in methanol was added for methylation, and the reaction was performed at 85 ◦C for 5 min. Afterward, 800 µL of pentane and 400 µL of distilled water were added to the tube, and the tube was centrifuged at 750 ×*g* for 15 min at 4 ◦C. The supernatant was collected into a new 2-mL tube and concentrated using the centrifugal concentrator. The concentrated sample was finally dissolved in 300 µL of hexane, filtered through a 0.5-µm syringe filter and analyzed by gas chromatography–quadrupole mass spectrometry (GC-qMS) (Shimadzu). The methylated fatty acids (1 µL) were separated in a DB-5 column (30 m × 0.25 mm × 1.00 µm; Agilent, Palo Alto, CA, USA) using a GCMS-QP2010 Ultra system with autosampler AOC-20i (Shimadzu). Injection volume of the samples was 1.0 µL and split mode was set at 10:1 ratio. Injection, ion source and interface temperatures were set at 280 ◦C, 200 ◦C and 280 ◦C, respectively. The column temperature conditions were as follows. The initial temperature was maintained at 40 ◦C for 2 min and raised to 320 ◦C at a rate of 6 ◦C/min. Helium was used as a carrier gas at a flow rate of 1.42 mL/min. Runtime was 2.86 to 49.00 min and scan mode was used with a mass range of 45 to 500 *m*/*z*. Qualitative and quantitative analyses of fatty acids were conducted using standards and a FAME Mix (C8–C24) (Table S3).
### *2.4. Statistical Analysis*
All analyses were performed no fewer than three times. Data obtained from GC-qMS were analyzed using PCA and OPLS-DA (SIMCA-P version 13.0; Umetrics, Umea, Sweden) to discriminate the geographic origin of perilla and sesame seeds. To determine the optimal OPLS-DA model, all the data were normalized with unit variance (UV)-scaling and pareto-scaling. PCA and OPLS–DA were based on the calculated eigenvectors and eigenvalues. The external validation test, permutation test and analysis of variance of the cross-validated residuals (CV-ANOVA) were conducted using SIMCA-P version 13.0 (Umetrics). The receiver operating characteristic (ROC) analysis and student's *t*-test were performed using MetaboAnalyst 4.0 (https://www.metaboanalyst.ca).
#### **3. Results**
### *3.1. Metabolite Profiling of Perilla and Sesame Seeds*
To discriminate the geographic origin of perilla and sesame seeds, we analyzed hydrophilic and lipophilic compounds using GC-qMS. We detected 35 hydrophilic compounds in 19 samples of perilla seeds and 31 hydrophilic compounds in 25 samples of sesame seeds (Tables S4 and S5). The lipophilic compounds, such as fatty acids, sterols, policosanols and tocopherols, were detected and quantified in all seed samples (Tables S6–S11). In total, 28 lipophilic compounds, including 11 fatty acids, 9 policosanols, 3 tocopherols, 3 sterols and 2 amyrins, were identified in perilla seeds (Tables S6, S8 and S10). In addition, 23 lipophilic compounds, including 10 fatty acids, 9 policosanols, 1 tocopherol and 3 sterols were detected in sesame seeds (Tables S7, S9 and S11). Unlike perilla seeds, α- and β-tocopherols, α- and β-amyrins and C18:3n3 were not detected in sesame seeds.
#### *3.2. PCA and OPLS-DA for Geographic Discrimination of Perilla and Sesame Seeds*
To discriminate the geographic origins of perilla and sesame seeds, the metabolite profiling data were processed using multivariate statistical analysis (PCA and OPLS-DA), which is an important tool for identifying the features of samples in complex data matrices. PCA uses an orthogonal linear transformation to transform the original data into a new set of variables, the principal component (PC). The scores and loading of PCs are represented in a bi-dimensional plot, which can formulate a dataset pattern from the raw data. The data were normalized with UV-scaling. In the PCA score plots, the two seeds did not show any variance according to geographic origins (Figures S1 and S2).
To improve the geographic discrimination of perilla and sesame seeds, we used OPLS-DA to determine the differences in metabolites arising due to differences in the geographic origin. OPLS-DA is a supervised classification method that features (X variables: metabolites) divides into two parts to separate the systematic variation: one that models the correlation between X and Y (prediction) and another that models the orthogonal components [30]. Thus, OPLS-DA has maximum separation by geographic origins based on their metabolites. The geographic origins (Y-variables) were set to 0 for Korea and 1 for China. Internal validation method was used to validate the model. The quality of the predictive model was measured by *R* <sup>2</sup> and *Q*<sup>2</sup> values of the validation results. The *R* <sup>2</sup> value indicates how much the proportion of variation in the data is explained by the model and the goodness of fit. The *Q*<sup>2</sup> value indicates how much proportion of variation in the data is predictable by the model and the goodness of prediction. The parameters *R* <sup>2</sup> and *Q*<sup>2</sup> were calculated minimum zero to maximum one; the *R <sup>2</sup>* value closer to 1 indicates a good value, *Q*<sup>2</sup> > 0.5 is regarded as a good prediction model and *Q*<sup>2</sup> > 0.9 is regarded as excellent prediction model. To develop a better discrimination model, the data were normalized by UV and pareto scaling. The optimal OPLS-DA model was established using UV-scaling, which showed higher *R* <sup>2</sup>Y (perilla; 0.822, sesame; 0.844) and *Q*<sup>2</sup> (perilla; 0.761, sesame; 0.799) values than pareto-scaling (*R* <sup>2</sup>Y: perilla; 0.575, sesame; 0.744/ *Q*<sup>2</sup> : perilla; 0.480, sesame; 0.715) (Table 1). The OPLS-DA models of both perilla and sesame seeds showed the *Q*<sup>2</sup> values to be above 0.5, indicating a good prediction model.
**Table 1.** Model validation results from orthogonal partial least squares discriminant analysis (OPLS–DA) with various scaling methods for discriminating the geographic origin of perilla and sesame seeds.
UV—unit variance; Par—pareto.
The OPLS-DA analysis was performed with UV-scaling data. The OPLS score plot of perilla seeds showed good separation on the basis of geographic origins (Korea and China) (Figure 1A). To identify the potential biomarkers for the geographic discrimination of perilla seeds, variable importance in projection (VIP) plots were used to explain the contribution of metabolites to the prediction models wherein VIP values greater than 1.00 indicate the significant influence on the model. In total, 29 metabolites had greater than 1.00 VIP values (Table S12). Glycolic acid, α-tocopherol and C20:0 were top-ranked metabolites in the VIP plots. The OPLS score plot of sesame seeds also showed good separation by region (Korea and China) (Figure 1B). In total, 26 metabolites showed a VIP cut off value of over 1.00 (Table S13). Proline, glycine and alanine were top-ranked in VIP plots.
The established OPLS-DA model for the discrimination of perilla and sesame seeds on the basis of geographic origin was subjected to an external validation test to determine its accuracy. In the case of perilla seeds, 57 samples were divided into 49 training samples and 8 test samples. The Y-variables were set to 0 for Korea and 1 for China. The OPLS projection model was established using 49 training samples, and then the 8 test samples were projected on the established OPLS projection model. The results of external validation test showed good discrimination of geographic origin of perilla seeds in the OPLS prediction model with *R* <sup>2</sup>X = 0.298, *R* <sup>2</sup>Y = 0.788 and *Q*<sup>2</sup> = 0.674. In addition, this OPLS model showed a root mean square error of prediction (RMSEP) = 0.229, which indicates the accuracy of prediction. The RMSEP value, being close to zero, indicated a good value. Furthermore, perilla seeds cultivated in Korea and China did not fall on the borderline of 0.5, which was a threshold level in the external validation test. Additionally, a permutation test and CV-ANOVA were conducted to test the risk of over-fitting the OPLS model. The permutation test was performed with 200 permuted models, which was constructed using randomized Y-variables. The reference distribution of the *Q*<sup>2</sup> value for random data from permuted models was compared with the *Q<sup>2</sup>* value of the real (unpermuted) OPLS model. When the *Q<sup>2</sup>* value from the permuted model is smaller than the *Q<sup>2</sup>* value of the original OPLS model, the model is considered as a predictable model. The results of the permutation test showed the *<sup>Q</sup>*<sup>2</sup> value of <sup>−</sup>0.496, which was lower than the *<sup>Q</sup>*<sup>2</sup> value of the original OPLS model (Figure 2A). The CV-ANOVA test was performed to testify the validity of the model. When the *p*-value was smaller than 0.05, the model was regarded as a validated model. The *p*-value of perilla seeds from the CV-ANOVA test was 3.05 <sup>×</sup> <sup>10</sup>−<sup>10</sup> .
To perform the external validation test for the OPLS-DA model of sesame seeds, the 78 samples were divided into 68 training samples and 10 test samples. The 68 training samples were used for the construction of the OPLS prediction model, and the 10 test samples were projected on the OPLS model. The external validation test results displayed good separation of sesame seeds samples on the basis of geographic origin in the OPLS projection model, which showed validation values with *R* <sup>2</sup>X = 0.320, *R* <sup>2</sup>Y = 0.812, *Q*<sup>2</sup> = 0.754 and RMSEP = 0.208. The results of the permutation test for the OPLS predictive model for sesame seeds showed the *<sup>Q</sup>*<sup>2</sup> value of <sup>−</sup>0.383, which was smaller than the *Q*<sup>2</sup> value of the real OPLS model. The CV-ANOVA test results of sesame seeds showed the *p*-value of 1.61 <sup>×</sup> <sup>10</sup>−18. Therefore, the OPLS-DA model for geographic discrimination of both of perilla and sesame seeds were successfully established and validated.
**Figure 1.** OPLS–DA score plots and VIP (variable importance in the projection) plots of (**A**) perilla and (**B**) sesame seeds from Korea and China. C20-ol—eicosanol; C21-ol—heneicosanol; C22-ol—docosanol; C23-ol—tricosanol; C24-ol—tetracosanol; C26-ol—hexacosanol; C27-ol—heptacosanol; C28-ol—octacosanol; C30-ol—triacontanol; C12:0—lauric acid; C14:0—myristic acid; C16:1n7—palmitoleic acid; C16:0—palmitic acid; C18:2n6—linoleic acid; C18:3n3—α-linolenic acid; C18:1n9—oleic acid; C18:0—stearic acid; C20:0—arachidic acid; C22:0—behenic acid; C24:0—lignoceric acid.
*Foods* **2020**, *9*, 989
**Figure 2.** External validation test and permutation test by OPLS-DA for discriminating the geographic origin of (**A**) perilla and (**B**) sesame seeds from Korea and China. The number of permutations for the permutation test was 200. (A: *R* 2X = 0.298, *R* 2Y = 0.788, *Q*2 = 0.674, RMSEP = 0.229; B: *R* 2X = 0.320, *R* 2Y = 0.812, *Q*2 = 0.754, RMSEP = 0.208).
### *3.3. Potential Biomarkers for the Discrimination of Perilla and Sesame Seeds Based on Their Geographic Origins*
The OPLS-biplot displayed a combination of observations (samples), X-variables (metabolites) and Y-variables (geographic origin) in a bi-dimensional space. This could easily explain the correlation of variables and the clustering of samples. The three ellipses—inner (0.50), middle (0.75) and outer (1.00)—indicate that the explained variances are 50%, 75% and 100%, respectively. If the variables are located close to the observations, the sample group has high levels of metabolites, whereas if they are opposite, the levels of metabolites are low. If the variables are closer to the outer circle (1.00) of the OPLS-biplot, the metabolites have more significantly contributed to the model.
In the OPLS-biplot of perilla seeds, glycolic acid, α-tocopherol and C20:0 were significant contributors, which were notably positioned the closest to the outer (1.00) circle and Y-variables (Figure 3A). In particular, only glycolic acid was located within middle (0.75) and outer (1.00) circles among these metabolites. In addition, these metabolites had top-ranked VIP values (glycolic acid, 1.82; α-tocopherol, 1.70; and C20:0, 1.48) in VIP plot. Therefore, to evaluate the predictive performance of these metabolites as potential biomarkers, ROC analysis was conducted. When the area under curve (AUC) values, which were a result of the ROC analysis, are to be closer to 1.00, the outcome is desirable [4]. Glycolic acid showed the AUC value of 1.000, indicating the excellent accuracy of discriminating Korean and Chinese perilla seeds (Figure 4A). In addition, α-tocopherol (AUC: 0.900) and C20:0 (AUC: 0.856) showed good accuracy to be considered as potential biomarkers. Therefore, glycolic acid was proposed as a potential biomarker for Chinese perilla seeds.
As shown in Figure 4B, proline, glycine and alanine, which were top-ranked (proline, 1.82; glycine, 1.57; and alanine, 1.49) in the VIP plot of sesame seeds, were located the closest to the outer circle and Y-variables. These metabolites showed AUC values in the range of 0.915–0.944, indicating their excellent accuracy as potential biomarkers for discriminating Korean and Chinese sesame seeds. Thus, proline, glycine and alanine were proposed as potential biomarkers for discriminating sesame seeds on the basis of geographic origin.
**Figure 3.** The OPLS-biplot for discriminating the geographic origin of (**A**) perilla and (**B**) sesame seeds using metabolite profiling data. The OPLS-biplot showed correlation of all metabolites (X-variables), sample clusters (observations) and geographic origins (Y-variables). C20:0; arachidic acid.
**Figure 4.** Receiver operating characteristic (ROC) curves for discriminating the geographic origins of (**A**) perilla and (**B**) sesame seeds using metabolite profiling data. ROC curves for (a) glycolic acid, (b) α-tocopherol and (c) C20:0 (arachidic acid) on discriminating (**A**) perilla seeds from Korea and China. ROC curves for (d) proline, (e) alanine and (f) glycine on discriminating (**B**) sesame seeds from Korea and China.
#### **4. Discussion**
The quality of perilla and sesame seeds and oils based on various health-related compounds such as fatty acids, tocopherols and sterols has been assessed previously [1,5]. However, to the best of our knowledge, a comprehensive metabolite profiling, which combines primary and secondary metabolites, has not been reported for perilla and sesame seeds. Therefore, we analyzed the primary metabolites and health-promoting compounds, which are abundantly found in perilla and sesame seeds, using GC-qMS. Perilla and sesame seeds are important oil crops, and they contain high levels of lipophilic compounds. In our analysis, perilla seeds showed high levels of α-linolenic acid (C18:3n3) and linoleic acid (C18:2n6), which are essential omega-3 and -6 fatty acids, respectively (Tables S10 and S11). On the contrary, α-linolenic acid (C18:3n3) was not detected in sesame seeds. However, linoleic acid (C18:2n6) and oleic acid (C18:1n9) were detected in higher levels in sesame seeds than in perilla seeds. Among tocopherols, γ-tocopherol was found in the highest amount in both perilla and sesame
seeds; however, α- and β-tocopherols were not detected in sesame seeds. Phytosterols were found in high amounts in perilla and sesame seeds (Tables S8 and S9). The levels of phytosterols in sesame seeds were approximately three times higher than those in perilla seeds. The above results were consistent with those of the previous studies [1]. Perilla seeds showed high levels of policosanols (Table S6). In particular, C28-ol was found in the highest level among policosanols in perilla seeds. However, sesame seeds showed low levels of policosanols (Table S7). These results agreed with those of the previous studies, which showed that perilla seeds and oils contain the highest levels of policosanols among other oil crops, while sesame seeds and oils contain negligible amounts of policosanols [31,32]. The hydrophilic metabolites, such as amino acids, organic acids and sugars, were detected in both perilla and sesame seeds (Tables S4 and S5). Almost all amino acids were found at higher levels in sesame seeds than in perilla seeds, except methionine and β-alanine. Sesame seeds are known as a good source of proteins rich in high sulfur-containing amino acids [4,5]. Therefore, sesame seeds may be consumed methionine for generating protein, which including high sulfur-containing amino acids. For the synthesis of high amount methionine, aspartic acid metabolism is activated. As a result, aspartic acid levels were higher in sesame seeds than in perilla seeds. In addition, sesame seeds have high levels of phenylalanine. Sesame seeds are also known to contain high amounts of lignans such as sesamin, sesamolin and sesamol [6,7]. Therefore, sesame seeds may have an activated phenylpropanoid pathway for the synthesis of lignans, resulting in the upregulated levels of phenylalanine.
To compare the compositional differences in seeds according to their origins, student's *t*-test was performed with metabolite profile data of perilla and sesame seeds. The *t*-test results of perilla seeds showed that 22 metabolites were considered statistically significant (0.05 ≥ *p*-value) between Korean and Chinese perilla seeds. In addition, these metabolites were shown to have compositional differences with geographic origins of perilla seeds. In the OPLS-DA loading plots of perilla seeds, the Korean perilla seeds had higher amounts of five terpenoids (α-, γ-tocopherols, β-sitosterol and α-, β-amyrin), five fatty acids (C14:0, C16:0, C18:0, C20:0 and C22:0) and methionine than Chinese seeds (Figure S3B). On the other hand, four policosanols (C20-ol, C22-ol, C24-ol and C26-ol), five organic acids (glycolic acid, phosphoric acid, nicotinic acid, lactic acid, glyceric acid), 4-aminobutyric acid and sucrose were shown to be present in higher levels in Chinese perilla seeds. In the case of sesame seeds, 25 metabolites were considered statistically significant between Korean and Chinese seeds. In the OPLS-DA loading plots of sesame seeds, three fatty acids (C14:0, C18:1n-9 and C24:0), four organic acids (citric acid, isocitric acid, malic acid and threonic acid), threonine and C22-ol were higher in concentration in Korean sesame seeds than in Chinese sesame seeds (Figure S4B). Whereas, the Chinese sesame seeds contained higher amounts of four amino acids (glycine, alanine, phenylalanine and 4-aminobutyric acid), two organic acids (succinic acid and glyceric acid), four policosanols (C24-ol, C28-ol, C26-ol and C30-ol), γ-tocopherol, glycerol, phosphoric acid, inositol and fructose than the Korean sesame seeds.
We determined and predicted the geographic origins of perilla and sesame seeds cultivated in China and Korea using OPLS-DA (Figure 1). The score plot of OPLS-DA showed good separation of both perilla and sesame seeds using appropriate data pretreatment. The optimal data preprocessing method for the OPLS-DA model was the UV-scaling method with the highest *Q*<sup>2</sup> and *R* <sup>2</sup>Y values in both of perilla and sesame seeds (Table 1). The selection of normalization methods is particularly important to reduce the unwanted instrumental errors of peak intensity measurements for relevant biologic differences. Thus, data normalization and scaling strategies should be chosen in such a way that the model shows optimal predictive ability of MVDA and retains meaningful biologic information [33].
The OPLS-biplots and VIP plots were generated to identify the biomarkers for discriminating perilla and sesame seeds on the basis of their geographic origins. Glycolic acid, α-tocopherol and C20:0 were identified as potential biomarkers for perilla seeds discrimination. Furthermore, proline, alanine and glycine were found to be potential biomarkers for sesame seeds discrimination. These potential biomarkers were further validated using ROC curve analysis. All AUC values of potential biomarkers were higher than 0.85, indicating that these metabolites significantly contribute to discriminating the seeds on the basis of their geographic origins. Kim et al. have reported that the VIP values of proline and glycine derived from the OPLS-DA model for discriminating the geographic origin of sesame seeds were higher than 1.0, indicating that these metabolites can be potential biomarkers for determining the regional origins of sesame seeds [4]. Thus, our results were consistent with those of a previous study. Glycolic acid is generated during photorespiration. Under low atmospheric CO<sup>2</sup> condition, C3 photosynthetic metabolism fixes the competing substrate O<sup>2</sup> instead of CO2. The oxygen fixation generates one molecule of 3-phosphoglycerate (3-PGA) and one molecule of 2-phosphoglycolate (2-PG) instead of two molecules of 3-PGA. Glycolic acid is generated from the dephosphorylation of 2-PG, and it can inhibit the rate of photosynthesis in the chloroplast. As a result, photorespiration under current atmospheric CO<sup>2</sup> concentrations reduces the efficiency of C3 photosynthesis by ~15% to 50%, depending upon the temperature in the growing season at that particular geographic location [34]. Therefore, this study suggests that glycolic acid could be a potential biomarker for geographic discrimination of perilla seeds and proline and glycine could be the same for sesame seeds.
Outlier detection is an important issue in chemometrics analysis. The outliers are observations that are extreme or that do not fit the PCA model. Furthermore, outliers can be both serious and interesting observations in the data. To discover the outliers in the PCA model, we used the Hotelling's *T* 2 . The Hotelling's *T* 2 is a multivariate generalization of student's *t*-test and provides a check for observations adhering to multivariate normality. In the PCA score plots, the ellipse of Hotelling's *T* 2 indicates 95% confidence. When observations fall outside the confidence ellipse, they are termed as strong outliers. Observations suggested as outliers were removed from the entire data set. This process was repeated until no outliers were displayed on the PCA score plot. Figures S5 and S6 show the outlier removal process. A total of 11 samples were identified as outliers, and 46 samples remained in the data set of perilla seeds. In the OPLS-DA score plot of perilla seeds (Figure 1), Chinese perilla seeds were more dispersed than Korean perilla seeds because the outliers were clustered in the upper right of the score plot (Figure S3A). In addition, the data set of sesame seeds retained 69 samples and eliminated 9 samples. These pretreated data sets of perilla and sesame seeds were subjected to OPLS-DA. Figure S7 shows OPLS-DA scores and VIP plots of the outlier removal data sets. The OPLS-DA model was established using UV-scaling, which showed higher *R* <sup>2</sup>Y (perilla; 0.928, sesame; 0.876) and *Q*<sup>2</sup> (perilla; 0.874, sesame; 0.842) values than the original data set *R* <sup>2</sup>Y (perilla; 0.822, sesame; 0.844) and *Q*<sup>2</sup> (perilla; 0.761, sesame; 0.799) values. The OPLS-DA score plots for the outlier removal data sets showed good separation of both perilla and sesame seeds. In particular, the OPLS-DA score plots of the outlier removal data set of perilla seeds showed clearer clustering of the Chinese samples than that of the original data set. Furthermore, the VIP plots of the outlier removal data sets of perilla and sesame seeds showed results that were almost same as those of the original data sets. Although the number of samples was reduced by more than 10% due to the outlier removal, the potential biomarker candidates were the same as those from the original data sets. These results demonstrated that the established OPLS-DA discrimination models for perilla and sesame seeds were reliable predictive models.
In conclusion, we performed comprehensive metabolite profiling, which included primary metabolites and health-promoting secondary metabolites, for perilla and sesame seeds cultivated in Korea and China. In addition, we established the OPLS-DA discriminative model for perilla and sesame seeds and validated it with good test results. The OPLS-DA results showed a clear separation of perilla and sesame seeds sourced from Korea and China on the basis of their geographic origins. The OPLS-biplot and VIP plot showed that glycolic acid was a notable metabolite for discrimination of perilla seeds based on geographic origin; therefore, we propose it as a potential biomarker for such discrimination. Furthermore, proline and glycine most significantly contributed for determining the geographic origins of sesame seed, and thus, they could be potential biomarkers for discrimination of sesame seeds based on the geographic origin. This study provides a reliable discriminatory predictive model to determine the geographic origins of perilla and sesame seeds cultivated in Korea and China. In addition, to the best of our knowledge, this is the first attempt to construct a discrimination model for perilla seeds using metabolomics. We believe that this model will be helpful in dealing with issues of selling domestic perilla and sesame seeds in the name of imported ones. In this study, the number
of samples and their source countries was limited. A future work should involve a larger sample size from more cultivated regions in various countries and evaluate the predictive ability of this model.
**Supplementary Materials:** The following are available online at http://www.mdpi.com/2304-8158/9/8/989/s1, Figure S1: PCA score (A) and loading (B) plots of perilla (*Perilla frutescens*) seeds from Korea and China, Figure S2: PCA score (A) and loading (B) plots of sesame (*Sesamum indicum*) seeds from Korea and China, Figure S3: OPLS-DA score (A) and loading (B) plots of perilla (*Perilla frutescens*) seeds from Korea and China, Figure S4: OPLS-DA score (A) and loading (B) plots of sesame (*Sesamum indicum*) seeds from Korea and China, Figure S5: PCA score plots and Hotelling's *T* 2 range column plots of perilla (*Perilla frutescens*) seeds from Korea and China for outlier removal process, Figure S6: PCA score plots and Hotelling's *T* 2 range column plots of sesame (*Sesamum indicum*) seeds from Korea and China for outlier removal process, Figure S7: OPLS–DA score plots and VIP (variable importance in the projection) plots of perilla (A) and sesame (B) seeds from Korea and China outlier removal data sets, Table S1: Relative retention times (RRT) and mass spectral data of hydrophilic compounds as trimethylsilyl derivatives, Table S2: Relative retention times (RRT) and mass spectral data of lipophilic compounds as trimethylsilyl derivatives, Table S3: Relative retention times (RRT) and concentration of fatty acid methyl esters (FAME) mixture and fatty acids, Table S4: Composition and content (ratio/g) of hydrophilic compounds in perilla (*Perilla frutescens*) cultivars, Table S5: Composition and content (ratio/g) of hydrophilic compounds in sesame (*Sesamum indicum*) cultivars, Table S6: Composition and content (µg/g) of policosanol compounds in perilla (*Perilla frutescens*) cultivars, Table S7: Composition and content (µg/g) of policosanol compounds in sesame (*Sesamum indicum*) cultivars, Table S8: Composition and content (µg/g) of sterol and terpenoid compounds in perilla (*Perilla frutescens*) cultivars, Table S9: Composition and content (µg/g) of sterol and terpenoid compounds in sesame (*Sesamum indicum*) cultivars, Table S10: Composition and content (mg/g) of fatty acids in perilla (*Perilla frutescens*) cultivars, Table S11: Composition and content (mg/g) of fatty acids in sesame (*Sesamum indicum*) cultivars, Table S12: OPLS-DA loading plots and VIP values of variables of perilla (*Perilla frutescens*) cultivars, Table S13: OPLS-DA loading plots and VIP values of variables of sesame (*Sesamum indicum*) cultivars.
**Author Contributions:** Conceptualization, methodology: J.K.K., W.D.S. and T.J.K.; formal analysis: J.G.P.; resources: H.Y.K. and W.D.S.; data curation: J.G.P. and T.J.K.; writing—original draft preparation: T.J.K. and J.G.P.; writing—review and editing: S.-H.H., B.L. and J.K.K.; project administration: S.U.P., J.K.K. and W.D.S. All authors have read and agreed to the published version of the manuscript.
**Funding:** This research received no external funding.
**Acknowledgments:** This work was supported "Cooperative Research Program for Agriculture Science & Technology Development (Project No. PJ013483042020)" funded by the Rural Development Administration (RDA), Republic of Korea and by Research Assistance Program (2019) in the Incheon National University, Republic of Korea.
**Conflicts of Interest:** The authors declare no conflict of interest.
#### **References**
© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
## **Food Authentication: Tru**ffl**e (***Tuber* **spp.) Species Di**ff**erentiation by FT-NIR and Chemometrics**
#### **Torben Segelke** † **, Stefanie Schelm** † **, Christian Ahlers and Markus Fischer \***
Hamburg School of Food Science—Institute of Food Chemistry, University of Hamburg, Grindelallee 117, 20146 Hamburg, Germany; [email protected] (T.S.);
[email protected] (S.S.); [email protected] (C.A.)
**\*** Correspondence: [email protected]; Tel.: +49-4042-838-43-57
† Authors contributed equally to this work.
Received: 16 June 2020; Accepted: 10 July 2020; Published: 13 July 2020
**Abstract:** Truffles are certainly the most expensive mushrooms; the price depends primarily on the species and secondly on the origin. Because of the price differences for the truffle species, food fraud is likely to occur, and the visual differentiation is difficult within the group of white and within the group of black truffles. Thus, the aim of this study was to develop a reliable method for the authentication of five commercially relevant truffle species via Fourier transform near-infrared (FT-NIR) spectroscopy as an easy to handle approach combined with chemometrics. NIR-data from 75 freeze-dried fruiting bodies were recorded. Various spectra pre-processing techniques and classification methods were compared and validated using nested cross-validation. For the white truffle species, the most expensive *Tuber magnatum* could be differentiated with an accuracy of 100% from *Tuber borchii*. Regarding the black truffle species, the relatively expensive *Tuber melanosporum* could be distinguished from *Tuber aestivum* and the Chinese truffles with an accuracy of 99%. Since the most expensive Italian *Tuber magnatum* is highly prone to fraud, the origin was investigated and Italian *T. magnatum* truffles could be differentiated from non-Italian *T. magnatum* truffles by 83%. Our results demonstrate the potential of FT-NIR spectroscopy for the authentication of truffle species.
**Keywords:** truffle; *Tuber* spp.; food authentication; species differentiation; near-infrared spectroscopy; chemometrics
### **1. Introduction**
Today's globalization leads to an increase of known cases of food fraud [1]. At the same time, consumer demand is moving towards food products of higher quality [2]. Many cases of food fraud pose a risk to health if toxic or allergenic substances get into the products through adulteration. However, even in cases of food fraud, which in many cases do not lead to a health hazard, it must be ensured that the consumer is not economically harmed, i.e., that no unjustifiably high prices are charged for inferior goods.
The increasing interest of the consumer in higher quality food [3], and also the willingness to pay more money for it, provides the incentive for criminally motivated actors to stretch high-end products with cheaper ingredients. Since many falsifications cannot be detected immediately by laymen or even by trained personnel in companies, it is becoming increasingly important to have appropriate instrumental detection methods for possible food adulteration at hand [4].
Because of the unique aroma and taste emitted from the fruiting bodies, truffles (*Tuber* spp.) are considered as delicacies. The underground growing ascomycetes represent the most expensive of all edible fungi, whereby the white Piedmont Truffle (*Tuber magnatum*) and the black Périgord
Truffle (*T. melanosporum*) are the most valuable species: prices do range between 3000–5000 €/kg and 700–1200 €/kg, respectively [5–7].
Because of their high price, truffles are often subject to fraud, especially when the species are very similar in their morphological appearance: *T. borchii* (syn. *Tuber albidum* Pico) is a truffle morphologically and biochemically similar to *T. magnatum*, both are classified as white truffles. The latter is the most expensive truffle species of all, so it is obvious that it is the subject of an intended counterfeit [8,9]. However, even unintentional cases of fraud are reported when other truffles, such as *T. borchii* are harvested, although the roots have initially been colonized by *T. magnatum* [10,11].
Amongst black truffles, the species *T. melanosporum* is the most expensive and highly valued for its organoleptic properties [12]. The Asian black truffles (e.g., *Tuber indicum*, *Tuber himalayense,* and *Tuber sinense*) form fruiting bodies morphologically very similar to *T. melanosporum* [13]. In view of the higher price of *T. melanosporum*, there is also a risk of fraud, especially since Asian black truffles are imported into Europe from China [14–16].
Due to the above-mentioned potential fraud cases, analytical authentication techniques are necessary, which must also be time-efficient due to the short-term storage of the industry.
In 2006, Zhao et al. compared five Chinese truffle fruiting bodies using Fourier transform infrared (FT-IR) spectroscopy [17] and successfully differentiated *T. magnatum*, *T. indicum,* and *Tuber excavatum* from each other. More recently, El Karkouri et al. proposed a matrix-assisted laser desorption/ionisation time of flight mass spectrometry (MALDI-TOF-MS) strategy, analysing proteins and applying database search algorithms [5]. In 2020, Krauß et al. analysed different tuber species regarding their geographical origin and species authentication via stable isotope ratio analysis showing that a differentiation with this method is possible [18]. However, these techniques still require costly instrumentation, maintenance and sophisticated handling. Instead, our practical approach is, to our knowledge, the first Fourier transform near-infrared (FT-NIR) spectroscopy study addressing the authentication of truffles with a relatively large number of samples.
FT-NIR spectroscopy is a simple and cost-effective approach, nowadays widely used for the monitoring as well as for the controlling of product quality and safety [19] alike the evaluation of the freshness [20] or of pesticide residues of fruits and vegetables [21]. FT-NIR spectroscopy is widely used for the authentication of foodstuffs [9,22–24] or for controlling the intentionally or unintentionally adulteration of exogenous substances or process by-products [25–27] and was recently used to monitor the post-harvest ripening of white truffles [28].
Data pre-processing of the obtained data is a crucial step in spectroscopic analysis. Therefore, pre-processing techniques, such as scatter correction, smoothing, or detrending steps are used in order to reduce the variability between samples due to scattering caused e.g., by heterogeneous sample size of powdery samples. Furthermore, additive and multiplicative effects in the spectra are removed and a subsequent exploratory analysis, a bi-linear calibration model or a classification model is improved [29]. It is essential to carefully compare and select the data pre-processing techniques to avoid misleading results and overfitting [29–31]. The decision on the classification model is crucial as well, and therefore, similarly to the evaluation of different data pre-treatment steps, we have examined and compared various classification models.
The aim of this study was to develop a reliable, easy-to-handle and low-cost method using the FT-NIR technology coupled to chemometric tools for the differentiation and authentication of five economically relevant truffle species. In this regard, we concentrated on the real truffles of the genus *Tuber* defined in the German Guidelines for mushrooms and mushroom products [32] and used in foodstuffs: the expensive species *T. melanosporum* and *T. magnatum,* as well as the less expensive species *T. aestivum*, *T. borchii,* and *T. indicum*. In this study, 75 truffle samples from three years of harvest and eleven growing countries were analysed. Different common pre-processing techniques were applied to the raw spectra and the results were compared using various classification models.
#### **2. Materials and Methods**
#### *2.1. Sample Acquisition*
In total, 75 truffle samples of relevant, market available white and black truffle species (harvest years 2017–2020) from 11 different countries were analysed in this study.
More precisely, the sample set consisted of two white species *T. magnatum* (20 samples) and *T. borchii* (5 samples) and three black species *T. melanosporum* (10 samples), *T. aestivum* (synonym *T. uncinatum* [33], 29 samples), and *T. indicum* (11 samples).
Regarding the *T. aestivum* species, molecular biological analyses have shown that *T. aestivum* and *T. uncinatum* are one species. Both terms should therefore be regarded as synonymous. Since *T. aestivum* was described before *T. uncinatum*, the species should be named *T. aestivum* [33]. Based on these molecular biological findings, *T. aestivum* and *T. uncinatum* were subsumed and named *T. aestivum* in this study.
An overview of the collected samples is given in Table S1. Some samples were commercially purchased and, therefore, considered as non-origin-authentic, so the origin is stated as 'unknown' in Table S1. Still, information regarding the truffle species were secured for all samples either by personal participation in harvest or by DNA analysis carried out within the Hamburg School of Food Science [34]. On arrival, all samples were frozen in liquid nitrogen and stored at −80 ◦C until further treatment.
#### *2.2. Sample Preparation*
Per sample, several fruiting bodies, at least 75 g, were cleaned with pure water obtained by a Direct-Q purifying system (Merck Millipore, Burlington, MA, USA) for removing remaining soil. Subsequently, the fruiting bodies were milled using a knife mill (Grindomix GM 300, Retsch, Haan, Germany) with dry ice at a ratio of 1:1 (*w*/*w*) and freeze-dried for 72 h [24]. The truffles were freeze-dried because of two reasons, which are more discussed in Section 3.1: (i) FT-NIR spectra of fresh truffles showed unspecific spectra with large water bands. (ii) It was known from the literature that such a freeze-drying step can enhance the accuracy of the classification models [35]. Freeze-dried material was crushed using a mortar and a pestle to obtain a fine homogeneous powder.
#### *2.3. Spectra Acquisition*
For the acquisition of near-infrared spectra, a TANGO FT-NIR spectrometer with an integrating sphere (Bruker Optics, Bremen, Germany) was used. The signals were recorded between 11550–3950 cm−<sup>1</sup> , collecting 50 scans at a resolution of 4 cm−<sup>1</sup> . All spectra were acquired at room temperature of 22 ± 2 ◦C. Samples of 300 mg, weighed in a glass vial (52.0 × 22 mm × 1.2 mm, Nipro Diagnostics Germany GmbH, Ratingen, Germany), were analysed in triplicate, in-between individual spectra recordings the lyophilisate was shaken in the glass vial.
#### *2.4. Spectra Pre-Processing*
FT-NIR spectra were pre-processed using MATLAB R2019a (The MathWorks Inc., Natick, MA, USA). After having omitted a specific range of higher wavenumbers (see Table 1 and discussion below), different pre-processing techniques or combinations of them were applied and compared (see Table 1) [36].
Multiplicative scatter correction (MSC) using the average of all spectra as the reference spectra was performed to eliminate scatter effects for all approaches i–vi. First order derivate (approach ii) was calculated to eliminate offset, baseline drifts and additive scattering effects, and second order derivate (approach iii) was calculated to remove multiplicative scattering effects in beyond. Detrending (polynomial order = 1) was applied for approach iv and vii. The effect of smoothing (moving average, span = 5) before MSC was investigated for approach v–vii.
**Table 1.** Pre-processing steps to the raw spectra in the order 1–2–3. For all approaches, a binning was added as a last step. MSC, multiplicative scatter correction.
After the pre-processing methods stated in Table 1, a binning by averaging 10 adjacent features was carried out with all spectra. Lastly, the triplicate spectra were averaged [24,25,36,37]. For certain issues (e.g., only black or white truffles or origin determination of *T. magnatum* samples), the MSC correction was only applied to the selected spectra.
#### *2.5. Multivariate Data Analysis*
For data investigation and visualization, principal component analysis (PCA) and line plots were calculated using MATLAB R2019a after applying spectra pre-treatments and mean centring the data.
For the different pre-processing approaches i–vii (see Table 1) it was each evaluated which classification model achieved the best prediction accuracy using MATLAB R2019a. The classification models examined in this context are stated in Table 2.
**Table 2.** Overview of the classification models examined in this study.
For optimising the model parameters and for obtaining an unbiased estimate of the model's performance, stratified nested cross-validation was used [44,45]. Therefore, the whole data set was split into four parts whereby the samples were stratified by the species to ensure a representative and balanced training set (three fourths) and test set (one fourth). For the training set, 10-fold cross validation was applied to select the optimal model parameters, referred to as inner cross-validation. The performance of the calculated model was then evaluated by predicting the test set. This process was repeated for all four folds, so every part of the four-fold outer cross validation was once used as the test set.
Finally, since the results by a single nested cross validation can vary, the entire nested cross-validation and the prediction of the test set were repeated 100 times, of which the mean accuracy and the standard deviation are reported. *Foods* **2020**, *9*, x FOR PEER REVIEW 5 of 16
#### **3. Results and Discussion 3. Results and Discussion**
#### *3.1. Spectra Investigation 3.1. Spectra Investigation*
Figure 1A shows all untreated spectra of the raw data, coloured in accordance to the different truffle species. As anticipated and seen from Figure 1A, the absorbance rises towards lower wavenumbers because of the transition probability which is higher for the first transition than for higher overtones [46]. Figure 1A shows all untreated spectra of the raw data, coloured in accordance to the different truffle species. As anticipated and seen from Figure 1A, the absorbance rises towards lower wavenumbers because of the transition probability which is higher for the first transition than for higher overtones [46].
**Figure 1.** (**A**) Raw Fourier transform near-infrared (FT-NIR) spectra, triplicate measurements from all 75 samples, coloured by truffle species. (**B**) Mean FT-NIR spectra for each truffle species after omitting the >9000 cm<sup>−</sup>1 range, MSC and binning. **Figure 1.** (**A**) Raw Fourier transform near-infrared (FT-NIR) spectra, triplicate measurements from all 75 samples, coloured by truffle species. (**B**) Mean FT-NIR spectra for each truffle species after omitting the >9000 cm−<sup>1</sup> range, MSC and binning.
However, in the range from 11,550−9000 cm−1 some spectra show strong absorbance. Calculating the corresponding wavelength, this region from 11,000–9000 cm−1 relates to the region from 1111−909 nm, which is close to the visual region. Here, the 4th overtone of the –OH bond occurs, and the colour of the truffle lyophilisate itself might cause an offset, which could have increased the absorbance [47]. Since the spectra vary in a strong way for this region, chemometric analyses, such as PCA, would excessively focus on this region and would neglect the information that is present in the spectra for smaller wavenumbers, so we excluded the >9000 cm−1 region. In fact, the range >9000 cm−1 is often excluded in various FT-NIR studies—also because this region is prone to noise when performing data pre-processing methods, such as first or second derivative [37,43]. Regarding the exclusion of some regions in the FT-NIR spectra, special care has to be taken to bands, which can be affected by the water content. Particularly in the region around 5312 cm−1 (O−H However, in the range from 11,550−9000 cm−<sup>1</sup> some spectra show strong absorbance. Calculating the corresponding wavelength, this region from 11,000–9000 cm−<sup>1</sup> relates to the region from <sup>1111</sup>−909 nm, which is close to the visual region. Here, the 4th overtone of the –OH bond occurs, and the colour of the truffle lyophilisate itself might cause an offset, which could have increased the absorbance [47]. Since the spectra vary in a strong way for this region, chemometric analyses, such as PCA, would excessively focus on this region and would neglect the information that is present in the spectra for smaller wavenumbers, so we excluded the >9000 cm−<sup>1</sup> region. In fact, the range >9000 cm−<sup>1</sup> is often excluded in various FT-NIR studies—also because this region is prone to noise when performing data pre-processing methods, such as first or second derivative [37,43].
stretching, first overtone) and around 7142 cm−1 (O−H deformation, second overtone), water can affect the absorbance of protein or carbohydrate specific bands [43]. The analysis of fresh truffle samples has shown that a drying step is necessary, as otherwise large water bands and unspecific spectra are obtained which superimpose the information beneath. Thus, the truffle samples were freeze-dried because such a sample preparation can enhance the accuracy of the classification models [35]. Due to the freeze-drying step, the water content in the samples can be seen as negligibly small and in the same range, so it should have no impact on the differentiation with chemometric models in the following steps. In addition, in the region 6500−5300 cm−1, not only water molecules absorb electromagnetic radiation, but also C–H vibrations do, which could be a useful parameter for the differentiation of the truffle species. In order to avoid the loss of useful information, we have not excluded other regions for this non-targeted approach, as several other research groups do in practice [24,37]. Regarding the exclusion of some regions in the FT-NIR spectra, special care has to be taken to bands, which can be affected by the water content. Particularly in the region around 5312 cm−<sup>1</sup> (O−H stretching, first overtone) and around 7142 cm−<sup>1</sup> (O−H deformation, second overtone), water can affect the absorbance of protein or carbohydrate specific bands [43]. The analysis of fresh truffle samples has shown that a drying step is necessary, as otherwise large water bands and unspecific spectra are obtained which superimpose the information beneath. Thus, the truffle samples were freeze-dried because such a sample preparation can enhance the accuracy of the classification models [35]. Due to the freeze-drying step, the water content in the samples can be seen as negligibly small and in the same range, so it should have no impact on the differentiation with chemometric models in the following steps. In addition, in the region 6500−5300 cm−<sup>1</sup> , not only water molecules absorb electromagnetic radiation, but also C–H vibrations do, which could be a useful parameter for the differentiation of the truffle species. In order to avoid the loss of useful information, we have not excluded other regions for this non-targeted approach, as several other research groups do in practice [24,37].
For powdered samples, multiplicative scatter effects occur due to differences in the materials' particle size, and have to be corrected for a reasonable data interpretation. To overcome such scattering effects, two approaches are commonly used: MSC and standard normal variate (SNV). According to Dhanoa et al., both pre-processing steps are two alternative approaches, which lead to similar results [48]. In the present study, MSC was chosen to correct for scattering effects. It should be noted that the sequence of the various pre-processing steps is always decisive. In Figure S1, the effect of applying MSC on the raw data, after having omitted the >9000 cm−<sup>1</sup> region, is shown. On the contrary, applying MSC first and omitting the >9000 cm−<sup>1</sup> region afterwards will have misleading results, as shown in Figure S1B on the right: the unwanted variance in the >9000 cm−<sup>1</sup> region is not excluded, but persists in the spectra as an error propagation. By applying pre-processing steps, it is therefore important to examine and to compare the impact of different orders, noted in the same way by Gerretzen et al. [49]. Any further pre-processing steps will be investigated and discussed more deeply in Section 2.4.
#### *3.2. Spectra Interpretation and Assignment of Bands*
The FT-NIR spectra reflect the major constituents of the truffles. Naturally low in fat, lyophilised truffle samples are rich in dietary fibre and proteins [50]. These components can be recognised in the spectrum by their characteristic bands; however, it should be noted that an exact assignment of bands for complex samples is difficult due to overlapping effects. For the sake of clarity, the mean spectra have been calculated for each truffle species, and the resulting representation is shown in Figure 1B. At around 6667 cm−<sup>1</sup> a vast band can be located induced by N−H stretching (first overtone) that can be attributed to proteins and amino acids. Furthermore, N−H combinations are also present around 4687 cm−<sup>1</sup> and the bands at 4859 cm−<sup>1</sup> and 4600 cm−<sup>1</sup> are caused by amide groups [24,38,47].
Regarding the carbohydrates, the double peak at 4338 cm−<sup>1</sup> and 4257 cm−<sup>1</sup> can be assigned to −CH<sup>2</sup> asymmetric stretching and symmetric stretching, respectively [51]. In addition, C−H stretching (first overtone) and <sup>−</sup>CH<sup>2</sup> vibration lead to peaks at 5760 cm−<sup>1</sup> and 5742 cm−<sup>1</sup> , respectively [24,37,47].
In order to put these observations into context, the work of Saltarelli et al. with an analysis of the protein and carbohydrate content of *T. magnatum*, *T. borchii*, *T. aestivum,* and *T. melanosporum* is of great importance. Although their work did not emphasise the species differentiation but storage effects, they have already noticed differences in the major constituents for the truffle species [52]. This can be illustrated well e.g., by the protein fraction: In ascending order, *T. melanosporum*, *T. aestivum*, *T. borchii,* and *T. magnatum* have a soluble protein content of 8.7, 11, 13, and 24%, respectively [52]. Such an order can be found at the wavenumber 6318 cm−<sup>1</sup> : *T. melanosporum* showing the lowest absorbance for this protein-specific region and *T. magnatum* the highest, so the above-mentioned study and our FT-NIR analysis is therefore consistent. Admittedly, this order is not properly given over the entire protein-specific range, especially *T. magnatum* shows an individual curve, but it should be noted that FT-NIR analysis is not capable of specifically measuring soluble proteins, as Saltarelli et al. (2008) did in their approach. Instead, it returns a general parameter, so the amount of scleroprotein and non-soluble protein fractions could cause the discrepancy. Consequently, it should be possible to distinguish species by—albeit very costly—quantitation of soluble protein and carbohydrate content. FT-NIR analysis, on the other hand, enables the indirect and rapid identification of these major constituents.
#### *3.3. Principal Component Analysis*
PCA is widely used for visualising high dimensional data by transforming them into a low dimensional space. As an unsupervised approach, it is useful for the qualitative data exploration, checking for potential outliers and rechecking the research hypothesis before using supervised classification models [53,54].
Figure 2A shows the score-plot for all 75 truffle samples. Tendencies of cluster formations according to the truffle species can be identified: the *T. magnatum* samples are located in the lower-left, whilst the *T. melanosporum* samples are located to the right and the *T. aestivum* samples are in the centre of the plot. *T. borchii* und *T. indicum* samples scatter across the plot. These intermediate results give reason to assume that a classification of truffle species is possible. However, with a differentiation of all five species we do not address real issues in the incoming goods inspection: the truffle's colour can be checked visually; thus, it only makes sense to consider the white and black truffles separately especially because falsification occurs within the white and within the black truffle, and these are not adulterated with each other. Therefore, PCA was calculated only for white and black truffle species and the score-plots are shown in Figure 2B–D, respectively. The trends from the score-plot in Figure 2A are also noticeable here, and FT-NIR analysis appears to be an appropriate method for differentiating truffle species. For the *T. indicum* samples in Figure 2C, some samples are spread over the entire score-plot, but tend to higher PC2 values in the PC4 vs. PC2 plane, already indicating the need for multivariate, non-linear analysis tools hereinafter. Moreover, the fact that there is still cluster formation shows that the important information for the species differentiation is not only contained in the >9000 cm−<sup>1</sup> region, which was omitted, but is present over the whole spectra. *Foods* **2020**, *9*, x FOR PEER REVIEW 7 of 16 give reason to assume that a classification of truffle species is possible. However, with a differentiation of all five species we do not address real issues in the incoming goods inspection: the truffle's colour can be checked visually; thus, it only makes sense to consider the white and black truffles separately especially because falsification occurs within the white and within the black truffle, and these are not adulterated with each other. Therefore, PCA was calculated only for white and black truffle species and the score-plots are shown in Figure 2B–D, respectively. The trends from the score-plot in Figure 2A are also noticeable here, and FT-NIR analysis appears to be an appropriate method for differentiating truffle species. For the *T. indicum* samples in Figure 2C, some samples are spread over the entire score-plot, but tend to higher PC2 values in the PC4 vs. PC2 plane, already indicating the need for multivariate, non-linear analysis tools hereinafter. Moreover, the fact that there is still cluster formation shows that the important information for the species differentiation is not only contained in the >9000 cm−1 region, which was omitted, but is present over the whole spectra.
**Figure 2.** Principal component analysis (PCA) score-plots with their respective loadings plots after pre-processing approach No. i of (**A**) all five truffle species, (**B**) only white truffle species, and only black truffle species in the (**C**) PC2 vs. PC1 plane and (**D**) in the PC4 vs. PC2 plane. **Figure 2.** Principal component analysis (PCA) score-plots with their respective loadings plots after pre-processing approach No. i of (**A**) all five truffle species, (**B**) only white truffle species, and only black truffle species in the (**C**) PC2 vs. PC1 plane and (**D**) in the PC4 vs. PC2 plane.
#### *3.4. Evaluation of Pre-Processing and the Suitability for the Species Classification Foods* **2020**, *9*, x FOR PEER REVIEW 8 of 16
Whereas applying MSC or SNV correction is necessary without question and is common practice in FT-NIR studies, the need and the impact of any further pre-processing steps should be investigated experimentally [55]. For evaluating the quality of such steps, only visual comparison of 'before-and-after' PCA plots is unlikely to find the most suitable pre-processing strategy and may mislead to an approach, which is not appropriate for a supervised model, so we calculated classification models and compared the prediction accuracy [36,49]. *3.4. Evaluation of Pre-Processing and the Suitability for the Species Classification* Whereas applying MSC or SNV correction is necessary without question and is common practice in FT-NIR studies, the need and the impact of any further pre-processing steps should be investigated experimentally [55]. For evaluating the quality of such steps, only visual comparison of 'before-andafter' PCA plots is unlikely to find the most suitable pre-processing strategy and may mislead to an approach, which is not appropriate for a supervised model, so we calculated classification models
Spectra comparison of different pre-processing approaches examined are shown in Figure 3. The effect of smoothing is not recognisable visually. In addition, it turned out that neighbouring wave numbers show almost identical absorbance values. In order to avoid redundant data and overfitting, a binning was carried out by calculating the mean value of the absorbance for 10 adjacent wavenumbers and combining the measuring points into 248 variables. and compared the prediction accuracy [36,49]. Spectra comparison of different pre-processing approaches examined are shown in Figure 3. The effect of smoothing is not recognisable visually. In addition, it turned out that neighbouring wave numbers show almost identical absorbance values. In order to avoid redundant data and overfitting, a binning was carried out by calculating the mean value of the absorbance for 10 adjacent wavenumbers and combining the measuring points into 248 variables.
**Figure 3.** Spectra comparison of different pre-processing approaches, also refer to Table 1**.** First row: one-step pre-processing: (**i**) MSC. Second row: two-step pre-processing: (**ii**) MSC, 1st derivative. (**iii**) MSC, 2nd derivative. (**iv**) MSC, detrending. Third row: three-step pre-processing: (**v**) smoothing, MSC, 1st derivative. (**vi**) Smoothing, MSC, 2nd derivative. (**vii**) Smoothing, MSC, detrending. **Figure 3.** Spectra comparison of different pre-processing approaches, also refer to Table 1**.** First row: one-step pre-processing: (**i**) MSC. Second row: two-step pre-processing: (**ii**) MSC, 1st derivative. (**iii**) MSC, 2nd derivative. (**iv**) MSC, detrending. Third row: three-step pre-processing: (**v**) smoothing, MSC, 1st derivative. (**vi**) Smoothing, MSC, 2nd derivative. (**vii**) Smoothing, MSC, detrending.
For every pre-processing approaches, all five classification models stated in Table 2 were calculated and validated using stratified nested cross-validation. As the main result parameter for comparing the approaches, we used the mean accuracy instead of the overall accuracy to account for the different size of the groups. The classification accuracies and precision for the test set for the differentiation of white and black truffles are given in Tables 3 and 4, respectively. For the training For every pre-processing approaches, all five classification models stated in Table 2 were calculated and validated using stratified nested cross-validation. As the main result parameter for comparing the approaches, we used the mean accuracy instead of the overall accuracy to account for the different size of the groups. The classification accuracies and precision for the test set for the differentiation of white and black truffles are given in Tables 3 and 4, respectively. For the training set used for validation, the classification accuracies and precisions are given in Tables S2 and S3, respectively.
**Table 4.** Mean accuracy and precision of the prediction of the external test set for different pre-treatments and classification models for the differentiation of the black truffle species (29 *T. aestivum* samples, 10 *T. melanosporum* samples, and 11 *T. indicum* samples, all values in %).
**actual species**
As can be seen in Table 3, all classification models provide good accuracy (>90%). Only the second derivation leads to significantly worse results. A pre-treatment of MSC with first derivation with both a linear and a quadratic SVM lead to an error-free classification of 100% (the most appropriate results are marked bold in the corresponding tables). Accordingly, any falsification of the expensive *T. magnatum* with the cheaper *T. borchii* can be detected. Because of the clear result based on the available and analysed truffle samples, the confusion matrix is not needed here, but can be seen in the supplement in Table S4.
A similar trend can be seen for the black truffles: Here too, high accuracies are generally achieved (>90%), only the second derivative without previous smoothing performs worse and a linear model does not seem to be sufficient for this ternary issue. Although the results overlap when the standard deviation is considered, the best accuracies of 99.1 ± 1.2 % are obtained when using MSC and the first derivative with the SSD model. A previous smoothing does not yield a significant improvement. Since every data pre-treatment is also a manipulation of the data, the model with the fewest steps should be preferred. The corresponding confusion matrix is shown in Table 5. In particular, fraud is common with *T. indicum*, which is counterfeited as the high-priced *T. melanosporum* because the two species are morphologically very similar and collected at the same harvesting times. Therefore, it is pleasing that the specificity for *T. melanosporum* is 97.5%—the error rate of mistakes is only 2.5%.
*T. indicum* 1073 1 26 97.5 *T. aestivum* 3 2897 0 99.9 *T. melanosporum* 1 0 999 99.9
precisions are given in Table S6. The corresponding confusion matrix is shown in Table S7.
DNA analysis is often used to authenticate species and varieties, while FT-NIR analysis is widely established in industrial incoming goods inspection. FT-NIR analysis does not require any specialised training for handling and any special, eventually hazardous chemicals for sample preparation and measurement, therefore the FT-NIR analysis is a "green method" [56]. Additionally, possible contamination due to exponential amplification by PCR quickly leads to false positive results. In order to keep this danger to a minimum, separate laboratories for sample preparation and DNA analysis are necessary, whereas NIR does not have such requirements. Optionally, it would be conceivable to use FT-NIR measurement for sample screening and to countercheck any conspicuous results by DNA analysis.
Regarding the determination of the geographical origin, however, DNA analysis cannot provide reasonable answers since the origin rather affects the phenotype. Here, FT-NIR analysis can be a tool for differentiating the origin [35] and the possibilities for the truffle differentiation by origin are examined in the following chapter.
#### *3.5. Influence of Harvest Year and Geographical Origin*
As shown in the PCA plot (Figure 2A), the truffle species has the dominant influence on the NIR spectrum, since the scores cluster according to their species in this unsupervised model. This can be demonstrated on the *T. magnatum* samples, which, although dominant from Italy, originate from Bulgaria, Croatia, and Romania, and are clustering together in the unsupervised PCA. This effect is similar for the *T. aestivum* samples originating mainly from Romania, but also from Bulgaria, France, Iran, Italy, Moldovia, and Slovenia. Thus, the species itself seems to have a much greater influence on the metabolome to be measured by FT-NIR spectroscopy than the origin.
One model for the origins of all truffle samples is not advisable for this reason, since most Italian samples are white truffles and most Romanian samples are *T. aestivum* truffles what is linked to their natural areas of origin. Such a model might, therefore, correlate on a false causality. However, the price depends primarily on the species whilst the origin is a second factor in the purchase decision. Accordingly, for the incoming goods inspection it is important especially for the most expensive *T. magnatum* truffle whether it comes from Italy or not, according to the consumer's expectations. For this Italy vs. non-Italy issue, all pre-processing was compared with classification models, analogous to the previous investigations when targeting the species. The results of the test set are shown in Table 6, and for the training set used for validation, the classification accuracies and precisions are given in Table S8. Best classification results of 88.4 % are reached after MSC and 2nd derivative in combination with a Random Forest (RF) classification model. However, we have decided not to pursue this pre-processing strategy because the spectra line plots in Figure 3iii have shown that a lot of noise occurs in the range of wavenumbers above 6000 cm−<sup>1</sup> and a smoothing an omitting this range is preferable. This alternative approach leads to a slightly worse accuracy of 82.8 ± 8.1% and the corresponding confusion matrix is shown in Table 7. The accuracy results provided by the LDA classification only differ by a few percentage points, and are even better in some cases. However, we chose the RF model since the PCA plots have arouse the impression that non-linear classification models might be more suitable for this issue.
**Table 6.** Mean accuracy and precision of the prediction of the external test set for different pre-treatments and classification models for the differentiation of Italian vs. non-Italian *T. magnatum* truffles (all values in %).
**Table 7.** Confusion matrix for classification for the differentiation of Italian vs. non-Italian T. magnatum truffles with the build RF model after smoothing. MSC and 2nd derivative; resulting in 82.8 ± 8.1% mean sensitivity. The predictions of 100 repetitions of the test set were accumulated.
Additionally, the PCA-plots for the *T. magnatum* samples were calculated and are shown in Figure 4, indicating and confirming that a non-linear classification model, such as RF, is more suited for this issue. Still, there are two aspects to consider: first, the standard deviation is remarkably high and second, the PCA plots show that the variance within the Italian samples is at least as large as the variance of the other origins. An origin model with acceptable accuracy is chemometrically possible, but should be checked with additional samples. *Foods* **2020**, *9*, x FOR PEER REVIEW 12 of 16
**Figure 4.** PCA score-plots with their respective loadings plots after pre-processing approach No. vi of **Figure 4.** PCA score-plots with their respective loadings plots after pre-processing approach No. vi of the *T. magnatum* samples from Italy and other countries (**A**) PC2 vs. PC1, (**B**) PC5 vs. PC1.
the *T. magnatum* samples from Italy and other countries (**A**) PC2 vs. PC1, (**B**) PC5 vs. PC1.
As the results show, FT-NIR can be used for the differentiation of black and white truffles, and Italian and non-Italian truffles of the species *T. magnatum*. Since FT-NIR is a simple and cheap method, it is suitable for industrial applications, for example, for the incoming goods inspection or authenticity checks on truffles. The process of authentication using FT-NIR is shown schematically in Figure 5.
and chemometrics.
**Figure 5.** Authentication protocol for the stepwise authentication assessment of truffles with FT-NIR and chemometrics. **Figure 5.** Authentication protocol for the stepwise authentication assessment of truffles with FT-NIR and chemometrics.
#### **4. Conclusions**
FT-NIR spectroscopy was combined with chemometrics to distinguish within the white truffles *T. borchii* and *T. magnatum* and the black truffles *T. aestivum*, *T. indicum,* and *T. melanosporum*. Different techniques for pre-processing in combination with various classification models and their effect on the accuracy of the model were compared. Classification accuracies >99% showed that the analysis of truffle samples by FT-NIR spectroscopy is a very suitable tool for species differentiation without sophisticated sample preparation or instruments. When differentiating between Italian and non-Italian *T. magnatum* samples, an accuracy of 83% was achieved. FT-NIR analysis requires no special training for handling and no special, possibly hazardous chemicals for sample preparation and measurement. In addition, most quality assurance laboratories already have FT-NIR instruments. Due to its simple, cost-effective application, FT-NIR analysis is very well suited for industrial screening samples during incoming goods inspection. Considering the number of 75 truffle samples used, we intend to extend the results of our study by analysing further samples, including a research on the potential effects of the harvest year.
**Supplementary Materials:** The following figures and tables are available online at http://www.mdpi.com/2304- 8158/9/7/922/s1, Figure S1: Influence of the order of pre-processing steps. (A) Raw data. (B) MSC and omitting the > 9000 cm−<sup>1</sup> range. (C) Omitting the >9000 cm−<sup>1</sup> range first and MSC; Table S1: Overview of the analysed truffle samples with number of samples, harvest year and country; Table S2: Mean accuracy and precision of the training set used for validation for different pre-treatment and classification models for the differentiation of the white truffle species (20 *T. magnatum* samples, 5 *T. borchii* samples, all values in %); Table S3: Mean accuracy and precision of the training set used for validation for different pre-treatment and classification models for the differentiation of the black truffle species (29 *T. aestivum* samples, 10 *T. melanosporum* samples and 11 *T. indicum* samples, all values in %); Table S4: Confusion matrix for classification of the white truffle species with the build linear SVM model after MSC and 1st derivative; resulting in 100% mean sensitivity. The predictions of 100 repetitions of the test set were accumulated; Table S5: Mean accuracy with standard deviation for different pre-treatment and classification models for the prediction of the test set for the differentiation of five truffle species (20 *T. magnatum* samples, 5 *T. borchii* samples, 29 *T. aestivum* samples, 10 *T. melanosporum* samples and 11 *T. indicum* samples, all values in %); Table S6: Mean accuracy and precision of the training set for different pre-treatment and classification models for the differentiation of the five truffle species (20 *T. magnatum* samples, 5 *T. borchii* samples, 29 *T. aestivum* samples, 10 *T. melanosporum* samples and 11 *T. indicum* samples, all values in %); Table S7: Confusion matrix for classification of five truffle species with the build subspace discriminant model after MSC and 1st derivative; resulting in 99.3 ± 0.9% mean sensitivity. The predictions of 100 repetitions of the test set were accumulated; Table S8: Mean accuracy and precision of the training set for different pre-treatment and classification models for the differentiation of Italian vs. non-Italian *T. magnatum* truffles (all values in %), MATLAB function for the creation of stratified parts for the nested cross validation.
**Author Contributions:** Conceptualization, T.S. and S.S.; methodology, T.S. and S.S.; validation, T.S. and C.A.; formal analysis, T.S. and S.S.; investigation, T.S., S.S. and C.A.; resources, M.F.; data curation, T.S. and S.S., writing—original draft preparation, T.S. and S.S.; writing—review and editing, T.S., S.S., C.A., and M.F.; visualization, T.S.; supervision, M.F.; project administration, M.F.; funding acquisition, M.F. All authors have read and agreed to the published version of the manuscript.
**Funding:** This study was performed within the project "Food Profiling—Development of analytical tools for the experimental verification of the origin and identity of food". This Project (Funding reference number: 2816500914) is supported by means of the Federal Ministry of Food and Agriculture (BMEL) by a decision of the German Bundestag (parliament). Project support is provided by the Federal Institute for Agriculture and Food (BLE) within the scope of the program for promoting innovation.
**Acknowledgments:** The authors gratefully thank the project partners "LA BILANCIA Trüffelhandels GmbH" and "Trüffelkontor e.K." for providing sample material. We would like to thank Maike Arndt and Bernadette Richter for their helpful discussion on the manuscript.
**Conflicts of Interest:** The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.
### **References**
© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
*Article*
## **Procedures for DNA Extraction from Opium Poppy (***Papaver somniferum* **L.) and Poppy Seed-Containing Products**
#### **Šarlota Ka ˇnuková 1 , Michaela Mrkvová 1 , Daniel Mihálik 1,2 and Ján Kraic 1,2,\***
Received: 4 September 2020; Accepted: 2 October 2020; Published: 9 October 2020
**Abstract:** Several commonly used extraction procedures and commercial kits were compared for extraction of DNA from opium poppy (*Papaver somniferum* L.) seeds, ground seeds, pollen grains, poppy seed filling from a bakery product, and poppy oil. The newly developed extraction protocol was much simpler, reduced the cost and time required for DNA extraction from the native and ground seeds, and pollen grains. The quality of extracted DNA by newly developed protocol was better or comparable to the most efficient ones. After being extended by a simple purification step on a silica membrane column, the newly developed protocol was also very effective in extracting of poppy DNA from poppy seed filling. DNA extracted from this poppy matrix was amplifiable by PCR analysis. DNA extracted from cold-pressed poppy oil and suitable for amplifications was obtained only by methods developed previously for olive oil. Extracted poppy DNA from all tested matrices was analysed by PCR using primers flanking a microsatellite locus (156 bp) and two different fragments of the reference tubulin gene (553 bp and 96 bp). The long fragment of the reference gene was amplified in DNA extracted from native seeds, ground seeds, and pollen grains. Poppy DNA extracted from the filling of bakery product was confirmed only by amplification of short fragments (96 bp and 156 bp). DNA extracted from cold-pressed poppy oil was determined also only by amplification of these two short fragments.
**Keywords:** DNA extraction; opium poppy; seed; pollen grains; bakery product; oil; PCR
## **1. Introduction**
Extraction of nucleic acids from various matrices is the first and crucial step in analysis of biological materials generally. Methods of DNA extraction have evolved over time [1], but still contain several basic and necessary steps such as cell disruption, removal of undesirable molecules (lipids, proteins, polyphenols, and others), and purification. In addition to the cell wall disruption, the chemical diversity of metabolites contained in the plant cell is a major complication in the DNA isolation process. There is no available universal protocol for extraction of DNA which would be applicable independently of plant species, plant tissues, and plant matrix [2,3]. Generally, extraction of DNA from young, fast-growing, and healthy tissues is much easier. However, it is often necessary to extract DNA from plant tissues rich in polysaccharides, lipids, secondary metabolites, or even from very complex matrices (processed seeds, oils, foods, feeds). This is also the case of oilseeds where extraction of DNA is considered more demanding than from vegetative plant tissues (e.g., young leaves). Lipids usually prevent the action of solvents during removal of polysaccharides and phenolic
compounds. Secondary metabolites can bind and precipitate with DNA and reduce efficiency of isolation procedure. Nevertheless, extraction of DNA from mature seeds may be often preferred over extraction from foliar tissues. Moreover, processing of plant seeds into foods is associated with determination of authenticity and traceability of foods what have recently become very important for various reasons [4–6]. The quantity and quality of DNA extracted from foods and oils tends to decrease to the extent in which the food/oil is processed [7,8]. Processing affects the DNA and may lead to degradation or removal of DNA from sample due to its hydrolysis, oxidation, and deamination [9]. Considering the DNA degradation and the presence of PCR inhibitors, DNA extraction from processed matrices is often a compromise between high yield and high purity [9–11]. The most appropriate extraction method should be chosen case by case. Extracted DNA is used for authentication of foods and feeds and detection of falsifications (e.g., blending of low-quality oil into high-quality oil) [12–14].
Oilseed crop with an interesting position in the world agriculture is the opium poppy (*Papaver somniferum* L.) grown under control only in some countries [15]. In addition to the production of alkaloids extracted from poppy straw, edible seeds are in great demand in cuisine. However, trading with poppy seeds, products (cake fillings, spreads), and oils suffers sometimes from adulteration practices [16]. Sometimes, high quality poppy seeds with a blue colour and a sweet taste are adulterated with technical poppy seed (grey-black colour, no taste). In addition to quality, they differ significantly in price. The consumer may be deceived in both quality and price. Such practices are then transferred to the food industry (poppy bakery products). Falsification is also a serious problem in the production of vegetable oils, especially the more expensive ones, including poppy seed oil. Chemical analyses of oils are used to determine the species origin of oil [17], but DNA analyses are appropriate to determine species origin and also the cultivar origin [18–20].
Poppy seeds with a high content of lipids and secondary metabolites are not a simple object for DNA extraction. This is even more complicated with ground seeds, poppy seed fillings from bakery products, and pressed oil. The number of relevant scientific reports in poppy is very limited and DNA extraction procedures have been published only from defatted seeds [21] and heroin samples [22]. Three commercial kits were tested for DNA extraction from seeds [23]. Very useful would be efficient, simple, and universal protocol for extraction of DNA from poppy seeds, grains, and products containing or made from poppy seeds. Therefore, the aim of this study was to test several methods of DNA extraction and try to design a new, effective procedure from different poppy seed matrices (native and ground seeds, pollen grains, poppy filling of the bakery product, poppy oil) with respect to DNA quality and suitability for amplification analyses.
#### **2. Materials and Methods**
#### *2.1. Plant and Food Material*
Mature seeds and pollen grains of opium poppy (*Papaver somniferum* L.) were collected from registered cultivar Major, cultivated at the Research and Breeding Station in Malý Šariš (Slovakia). They were stored at 4 ◦C before DNA extraction. Seeds and pollen grains were homogenized by pestle and mortar before the extraction. Seeds were also ground with the poppy seed mill. The poppy seeds roll (Tastino, Slovakia) and cold-pressed poppy seed oil (Juvamed Ltd., Tastino, Slovakia) were purchased in food store and stored at 4 ◦C before the DNA extraction.
#### *2.2. DNA Extraction from Seeds*
Genomic DNA from seeds was extracted from 0.2–0.5 g of seeds by six methods: Dellaporta et al. [24] with and without CTAB; Bayer BioScience N.V. [25]; Monsanto Company [26]; Murray and Thompson [27]; Sagwan et al. [21] and using four commercial kits: DNeasy® Plant Maxi Kit, QIAamp DNA Stool Mini Kit, PowerSoil DNA Isolation Kit (all from QIAGEN N.V., Hilden, Germany) and Plant DNAzol® Reagent (Thermo Fisher Scientific, Waltham, MA, USA).
Another extraction protocol was newly developed protocol designed on the basis of the Bayer BioScience N.V. procedure [25], but containing several modifications. The content of this protocol is as follows. The sample (200 mg) of seeds was ground to a fine powder with mortar and pestle and extracted with 2.7 mL of extraction buffer (50 mM EDTA, 100 mM Tris-HCl, pH 8.0, 500 mM NaCl), 190 µL of 20% SDS, and 10 µL of 2-mercaptoethanol. The mixture was vortexed and incubated at 65 ◦C for 30 min. During the incubation, the samples were mixed every 10 min. After incubation, 2.3 mL of mixture phenol:chloroform:isoamyl alcohol (25:24:1) was added, the mixture being shaken for 1 min and centrifuged for 20 min at 5500× *g*. The upper aqueous phase was transferred to a new tube, mixed with 2 mL of isopropanol and precipitated 30 min at −20 ◦C. Precipitated nucleic acids were transferred to Eppendorf tube and washed with 70% and 96% ethanol. Pellet after drying was dissolved in TE buffer (10 mM Tris-HCl, 1 mM EDTA, pH 8.0) and treated with 10 µL of RNase A (10 mg/mL) for 30 min at 37 ◦C. After the incubation, 800 µL of mixture chloroform:isoamyl alcohol (24:1) was added, shaken vigorously and centrifuged for 10 min in a microcentrifuge at maximum speed. The upper aqueous phase was transferred to a new Eppendorf tube, 600 µL of isopropanol was added and after vortexing was incubated for 20 min at −20 ◦C. Precipitate DNA was again washed with 70 and 96% ethanol, dried, dissolved in TE buffer, and stored at −20 ◦C.
#### *2.3. DNA Extraction from Ground Seeds*
Six different methods used for isolation DNA from 0.2 g of ground seeds were: Bayer BioScience N.V. [25], Monsanto Company [26], two commercial kits (DNeasy® Plant Maxi Kit, QIAamp DNA Stool Mini Kit), and newly developed protocol (described above).
#### *2.4. DNA Extraction from Pollen Grains*
Three methods used for isolation of DNA from 0.1 g of pollen grains included DNeasy® Plant Maxi Kit, QIAamp DNA Stool Mini Kit. The third was the newly developed protocol (described above). An efficient mechanical homogenization of pollen grains was particularly important.
#### *2.5. DNA Extraction from Poppy Seed Filling*
DNA was extracted from 0.5–2.0 g of filling of the bakery product using methods: Bayer BioScience N.V. [25], Monsanto Company [26], QIAamp DNA Stool Mini Kit, and newly developed protocol. Extracted DNA was purified through the silica membrane spin-columns [28].
#### *2.6. DNA Extraction from Poppy Oil*
DNA was extracted from 0.2–15 mL of oil according to Doveri et al. [29]; Monsanto Company [26]; Bayer BioScience N.V. [25]; Consolandi et al. [30]; Giménez et al. [31]; Raieta et al. [4], newly developed protocol, and commercial kit (QIAamp DNA Stool Mini Kit).
### *2.7. Qualitative and Quantitative Analysis of Extracted DNA*
Integrity of the extracted DNA from different poppy matrices was assessed by agarose gel electrophoresis. Parameters of extracted DNA were tested by UV spectrophotometry (NanoDrop ND-1000 spectrophotometer, Thermo Fisher Scientific, Waltham, MA, USA) as well as by electrophoresis in 0.8% agarose gel stained with ethidium bromide.
### *2.8. PCR Amplification*
Extracted DNA were amplified by PCR using primers for microsatellite locus psSSR69 [32]. Two pairs of primers for reference gene encoding tubulin beta-7 chain (Table 1) was designed from coding sequence (XM\_026557633.1, GenBank®, http://www.ncbi.nlm.nih.gov) [33] using the Primer3 Input software (Whitehead Institute for Biomedical Research, USA).
**Table 1.** Primer pairs used for amplification of opium poppy DNA. Tm: Melting temperature.
PCR reactions were carried out in 15 µL reaction containing 11.7 µL ddH2O, 1.5 µL 10× PCR buffer, 0.3 µL of both primer (0.20 µM), 0.3 µL each of dNTP (200 µM), 0.2 µL Taq-polymerase (1U/µL), and 1 µL DNA (25 ng/µL). Parameters of PCR for the psSSR69 locus were: 94 ◦C for 3 min, 45 cycles of 45 s at 94 ◦C, 1 min at 54 ◦C, 1 min at 72 ◦C, and additional 1 cycle at 72 ◦C for 10 min. The reference gene for tubulin beta-7 chain was amplified using the program: 94 ◦C for 5 min, 35 cycles of 45 s at 94 ◦C, 1 min at 59 ◦C, 1 min at 72 ◦C, and additional 1 cycle at 72 ◦C for 5 min. Amplicons were analysed in 2% agarose gels in TBE buffer and stained with ethidium bromide.
#### **3. Results and Discussion**
#### *3.1. DNA from Mature Seeds*
Poppy seeds are specific commercial commodity used in the food industry particularly in some regions of the world. However, the food quality and related price of seeds vary considerably for different *P. somniferum* L. cultivars. Unfortunately, it is likely that premium quality seeds (sweet taste, blue colour) of some poppy cultivars are intentionally handled while trading. They are usually exchanged with low-quality seeds or mixed with them, whether intentionally or not. Therefore, different protocols for extraction of total DNA from poppy seeds and poppy containing products were tested. DNA analysis should be used to determine poppy seed cultivar origin. In addition to the six extraction protocols and four commercial kits tested (Table 2), the modified extraction procedure ("newly developed protocol") was proposed in this study. It is based on the results and experiences obtained during testing of ten extraction procedures.
Spectrophotometric analysis as well as gel electrophoresis of DNA from seeds revealed significant differences between used extraction protocols, both in quantity and quality of obtained DNA. The qualitative parameters of DNA were primarily important (Table 2). The protocol of Dellaporta et al. [24] and its modification by incorporation of CTAB showed that mechanical homogenization of seeds directly in the extraction buffer, even without the use of liquid nitrogen, did not lead to deterioration in quality or amount of DNA (Table 2, Figure 1a). It may be concluded that the need to use liquid nitrogen during mechanical homogenization of poppy seeds is not necessary for prevention of degradation of extracted DNA [34–36]. Quality of extracted DNA varied according to the extraction procedure. Procedures according to Sangwan et al. [21], Bayer BioScience N.V. [25], Monsanto Company [26], Murray and Thompson [27], QIAamp DNA Stool Mini Kit, DNeasy® Plant Maxi Kit, as well as the newly developed protocol, provided poppy DNA with A260/<sup>280</sup> values in range 1.77–2.11. Procedures Dellaporta et al. [24], PowerSoil DNA Isolation Kit, and Plant DNazol® Reagent had the A260/<sup>280</sup> ratios in range 1.54–1.75 (Table 2).
extracted from 0.2 g of seeds extracted in volume of extraction buffer for 0.5 g of seeds.
*Foods* **2020**, *9*, x FOR PEER REVIEW 6 of 15
**Figure 1.** Genomic DNA extracted from opium poppy seeds (**a**) by: Dellaporta et al. [24] (lanes 1–6, where 1–3 homogenization with liquid nitrogen, 4–6 homogenization without liquid nitrogen), Dellaporta et al. [24] with CTAB (lanes 7–12, where 7–9 homogenization with liquid nitrogen, 10–12 homogenization without liquid nitrogen), Bayer BioScience N.V. [25] (lanes 13–15), Murray, Thompson [27] (lanes 16–18), Monsanto Company [26] (lanes 19–21), Sangwan et al. [21] (lanes 22–24), DNeasy® Plant Maxi Kit (lanes 25–28), Plant DNAzol® Reagent (lanes 29–31), QIAamp DNA Stool Mini Kit (lanes 32–35), PowerSoil DNA Isolation Kit (lanes 36–39), newly developed protocol (lanes 40–47). Lane M—λ-phage DNA. (**b**) Ground poppy seeds: DNA extracted by: Bayer BioScience N.V. [25] (lanes 1–7, lines 1–4, 0.2 g of seeds with extraction buffer volume for 0.2 g; lanes 5–7, 0.2 g of seed with extraction buffer volume for 0.5 g of seeds), Monsanto Company (26) (lanes 8–11), newly developed protocol (lanes 12–15), DNeasy® Plant Maxi Kit (lanes 16–19), QIAamp DNA Stool Mini Kit (lanes 20–23). Lane M—λ-phage DNA. **Figure 1.** Genomic DNA extracted from opium poppy seeds (**a**) by: Dellaporta et al. [24] (lanes 1–6, where 1–3 homogenization with liquid nitrogen, 4–6 homogenization without liquid nitrogen), Dellaporta et al. [24] with CTAB (lanes 7–12, where 7–9 homogenization with liquid nitrogen, 10–12 homogenization without liquid nitrogen), Bayer BioScience N.V. [25] (lanes 13–15), Murray, Thompson [27] (lanes 16–18), Monsanto Company [26] (lanes 19–21), Sangwan et al. [21] (lanes 22–24), DNeasy® Plant Maxi Kit (lanes 25–28), Plant DNAzol® Reagent (lanes 29–31), QIAamp DNA Stool Mini Kit (lanes 32–35), PowerSoil DNA Isolation Kit (lanes 36–39), newly developed protocol (lanes 40–47). Lane M—λ-phage DNA. (**b**) Ground poppy seeds: DNA extracted by: Bayer BioScience N.V. [25] (lanes 1–7, lines 1–4, 0.2 g of seeds with extraction buffer volume for 0.2 g; lanes 5–7, 0.2 g of seed withextraction buffer volume for 0.5 g of seeds), Monsanto Company (26) (lanes 8–11), newly developed protocol (lanes 12–15), DNeasy® Plant Maxi Kit (lanes 16–19), QIAamp DNA Stool Mini Kit (lanes 20–23). Lane M—λ-phage DNA.
Amplifications were successful from DNA extracted from mature native seeds by almost all of used protocols and the relevant fragments were generated (Figure 2). The only exception was DNA extracted by the Plant DNAzol® Reagent. DNA extracted by PowerSoil DNA Isolation Kit was probably highly degraded considering that 553 bp fragment of reference tubulin gene was not amplified, but a 156 bp length microsatellite marker was generated (Figure 2). The only one currently available protocol developed for extraction of DNA from poppy seeds [21] did not provide high quality of DNA within this study (Table 2, Figure 1a). The newly developed protocol has been proven as effective. Compared to the original protocol [25], extraction However, the success in amplification of extracted DNA is not guaranteed only by purity, but also by concentration and structural integrity of DNA [37,38]. Although values A<sup>260</sup> of DNA extracted by protocols Murray and Thomson [27], Sangwan et al. [21], Plant DNAzol® Reagent, and PowerSoil DNA Isolation Kit were high, DNA was not observed in agarose gel (Figure 1a, Table 2). There were probably only limited amounts of poppy DNA and absorbance values have been increased by the presence of RNA and other contaminants. DNA extracted by these protocols also had very low quality. Significant RNA contamination was reported only for the original CTAB method [27] and the QIAamp DNA Stool Mini Kit due to absence of RNase A treatment (Figure 1a).
steps were rearranged, time intervals between steps were changed, and some chemicals/enzymes were eliminated. Both absorbance parameters (A260/280 andA260/230) as well as electrophoretic profile of DNA predicted very good quality and quantity (Table 2, Figure 1a) that should be suitable for amplification by PCR (Figure 2). Amplifications were successful from DNA extracted from mature native seeds by almost all of used protocols and the relevant fragments were generated (Figure 2). The only exception was DNA extracted by the Plant DNAzol® Reagent. DNA extracted by PowerSoil DNA Isolation Kit was probably highly degraded considering that 553 bp fragment of reference tubulin gene was not amplified, but a 156 bp length microsatellite marker was generated (Figure 2).
*3.2. DNA from Ground Seeds* Ground poppy seeds are commonly available in food stores. The sensory values (especially taste and smell) and related varietal origin of high-quality seeds may be easily masked in ground seeds by various additives, mainly by sugar. Analytical testing and confirmation of the poppy seeds varietal origin is necessary in such cases. DNA from ground poppy seeds was extracted by two protocols [25,26], two commercial kits, as well as newly developed protocol. The QIAamp DNA The only one currently available protocol developed for extraction of DNA from poppy seeds [21] did not provide high quality of DNA within this study (Table 2, Figure 1a). The newly developed protocol has been proven as effective. Compared to the original protocol [25], extraction steps were rearranged, time intervals between steps were changed, and some chemicals/enzymes were eliminated. Both absorbance parameters (A260/<sup>280</sup> and A260/230) as well as electrophoretic profile of DNA predicted very good quality and quantity (Table 2, Figure 1a) that should be suitable for amplification by PCR (Figure 2).
#### Stool Mini Kit and DNeasy® Plant Maxi Kit produced DNA with the A260/280 and A260/230 ratios furthest *3.2. DNA from Ground Seeds*
from optimal values (Table 2). Both spectrophotometric ratios of DNA extracted by Monsanto Company [26] protocol indicated high contamination of DNA with proteins, organic solvents, and secondary metabolites, and also very low concentration (Table 2). The yield of DNA was significantly different between tested protocols, but at the same amount of loaded DNA (25 ng/μL) Ground poppy seeds are commonly available in food stores. The sensory values (especially taste and smell) and related varietal origin of high-quality seeds may be easily masked in ground seeds by various additives, mainly by sugar. Analytical testing and confirmation of the poppy seeds varietal
the electrophoretic profiles of all DNA samples were appropriate (Figure 1b). The highest quality
origin is necessary in such cases. DNA from ground poppy seeds was extracted by two protocols [25,26], two commercial kits, as well as newly developed protocol. The QIAamp DNA Stool Mini Kit and DNeasy® Plant Maxi Kit produced DNA with the A260/<sup>280</sup> and A260/<sup>230</sup> ratios furthest from optimal values (Table 2). Both spectrophotometric ratios of DNA extracted by Monsanto Company [26] protocol indicated high contamination of DNA with proteins, organic solvents, and secondary metabolites, and also very low concentration (Table 2). The yield of DNA was significantly different between tested protocols, but at the same amount of loaded DNA (25 ng/µL) the electrophoretic profiles of all DNA samples were appropriate (Figure 1b). The highest quality and concentration of DNA has been extracted by protocols Bayer Biocience N.V. [25] with changed ratio of sample–extraction buffer (w/v) and the newly developed protocol (Table 2, Figure 1b). Both protocols contained SDS in extraction buffer. It is suggested that SDS-based DNA extractions could be more appropriate for oily plant matrices like ground poppy seeds. The SDS-containing method modified for ground raw soybean seeds had the highest yield of DNA in comparison with the CTAB method and two commercial kits [39]. A lower amount of DNA yielded the CTAB method also from soybean flour [40].
**Figure 2.** Amplification of 156 bp microsatellite psSSR69 (**a**) and 553 bp fragment of gene for tubulin beta-7 chain (**b**) in DNA extracted from poppy seeds by: Dellaporta et al. [24] with and without liquid N<sup>2</sup> (lanes 1 and 2), Dellaporta et al. [24] with CTAB with and without liquid N<sup>2</sup> (lanes 3 and 4), Bayer BioScience N.V. [25] (5), Murray, Thompson [27] (6), Monsanto Company [26] (7), Sangwan et al. [21] (8), DNeasy® Plant Maxi Kit (9), Plant DNAzol® Reagent (10), QIAamp DNA Stool Mini Kit (11), PowerSoil DNA Isolation Kit (12), newly developed protocol (13–14), NC—negative control, PC—positive control, M1—25 bp ladder (Invitrogen), M2—100 bp DNA ladder (Solis BioDyne).
Amplifications of DNA from ground poppy seeds using primers flanking microsatellite marker psSSR69 and longer fragment of gene for tubulin beta-7 chain resulted in production of both the 156 and 553 bp amplicons in DNA extracted by all used protocols (Figure 3).
#### *3.3. DNA from Poppy Pollen Grains*
DNA was extracted by two commercial kits and by newly developed protocol (Table 2). Homogenization by pestle and mortar in liquid nitrogen was efficient for disruption of pollen exine with high structural integrity. Both ratios A260/<sup>280</sup> and A260/<sup>230</sup> confirmed that the best quality had DNA extracted by newly developed protocol (Table 2). This simple protocol produced also very high amount of DNA. On the opposite, the QIAamp DNA Stool Mini Kit and DNeasy® Plant Mini Kit extracted the least amount of DNA (Figure 4a). Amplifications of DNA from poppy pollen grains were basically without any complications. All primer pairs were able to amplify relevant amplicons (Figure 4). The genomic DNA is well protected inside the pollen grain therefore, a large fragment of the reference gene (553 bp) was simply amplified (Figure 4b). Amplifications of both shorter, the 156 bp microsatellite marker and 96 bp fragment of reference gene were also easily feasible (Figure 4c,d).
soybean flour [40].
**Figure 2.** Amplification of 156 bp microsatellite psSSR69 (**a**) and 553 bp fragment of gene for tubulin beta-7 chain (**b**) in DNA extracted from poppy seeds by: Dellaporta et al. [24] with and without liquid N2 (lanes 1 and 2), Dellaporta et al. [24] with CTAB with and without liquid N2 (lanes 3 and 4), Bayer BioScience N.V. [25] (5), Murray, Thompson [27] (6), Monsanto Company [26] (7), Sangwan et al. [21] (8), DNeasy® Plant Maxi Kit (9), Plant DNAzol® Reagent (10), QIAamp DNA Stool Mini Kit
PC—positive control, M1—25 bp ladder (Invitrogen), M2—100 bp DNA ladder (Solis BioDyne).
ratio of sample–extraction buffer (w/v) and the newly developed protocol (Table 2, Figure 1b). Both protocols contained SDS in extraction buffer. It is suggested that SDS-based DNA extractions could be more appropriate for oily plant matrices like ground poppy seeds. The SDS-containing method modified for ground raw soybean seeds had the highest yield of DNA in comparison with the CTAB method and two commercial kits [39]. A lower amount of DNA yielded the CTAB method also from
Amplifications of DNA from ground poppy seeds using primers flanking microsatellite marker psSSR69 and longer fragment of gene for tubulin beta-7 chain resulted in production of both the 156
and 553 bp amplicons in DNA extracted by all used protocols (Figure 3).
**Figure 3.** Amplification of 156 bp microsatellite psSSR69 (**a**) and 553 bp fragment of reference tubulin gene (**b**) in DNA extracted from ground seeds extracted by Bayer BioScience N.V. [25] (lanes 1–2, lane 1), Monsanto Company [26] (3), DNeasy® Plant Maxi Kit (4), QIAamp DNA Stool Mini Kit (5), newly developed protocol (6), NC—negative control, PC—positive control, M1—25 bp DNA ladder (Invitrogen), M2—100 bp DNA ladder (Solis BioDyne). **Figure 3.** Amplification of 156 bp microsatellite psSSR69 (**a**) and 553 bp fragment of reference tubulin gene (**b**) in DNA extracted from ground seeds extracted by Bayer BioScience N.V. [25] (lanes 1–2, lane 1), Monsanto Company [26] (3), DNeasy® Plant Maxi Kit (4), QIAamp DNA Stool Mini Kit (5), newly developed protocol (6), NC—negative control, PC—positive control, M1—25 bp DNA ladder (Invitrogen), M2—100 bp DNA ladder (Solis BioDyne).
**Figure 4.** Genomic DNA extracted from opium poppy pollen grains (**a**). Amplification of 553 bp (**b**) and 96 bp (**d**) fragments of reference tubulin gene, and 156 bp (**c**) microsatellite, respectively. (1)—newly developed protocol, (2)—DNeasy® Plant Maxi Kit, (3)—QIAamp DNA Stool Mini Kit, NC—negative control, PC—positive control, M—100 bp DNA ladder (**b**) (Invitrogen) and 25 bp DNA ladder (**c**,**d**) (Solis BioDyne).
Extraction of DNA from pollen grains is needed in different applications including monitoring of pollen grains transfer from transgenic opium poppy plants to the environment [41], detection of pollen species in food (e.g., in honey) for the prevention of allergens [42], forensic palynology [43] and others.
#### *3.4. DNA from Poppy Seed Filling*
DNA was extracted by two procedures, one commercial kit, and the newly developed protocol (Table 3). The purification step using the silica membrane spin-columns [28] was added to protocols Monsanto Company [26] and newly developed one. Both ratios A260/<sup>280</sup> and A260/<sup>230</sup> confirmed that DNA extracted using almost all extraction protocols had these values out of the optimal range (Table 3). Undamaged high molecular weight DNA extracted from poppy seed filling from the bakery product was not visualizable in agarose gel (data not shown). This reflects fragmentation of poppy DNA to very short fragments due to high degradation during baking. This is common for DNA extracted from a matrix that has undergone processing by high temperature [29] and a combination of grinding, mechanical manipulation, and thermal treatment [44]. However, the objective quality and usability of DNA extracted can only be revealed by its amplification.
Note: \*—subsequent purification of extracted DNA through silica membrane spin-columns [28], aq/oil—DNA isolated from the water (aq) or oily (o) phase.
Complex food matrices contain a variety of PCR inhibitors [45]. Other effects of the matrix include degradation, fragmentation, and restricted extractability of DNA, as well as presence of DNA from different organisms [46]. Baking temperature around 200 ◦C used in processing of bakery goods containing poppy seed filling substantially reduces the size of extracted DNA. Moreover, higher moisture content inside the product, in this case in poppy filling, contributes to greater degradation of DNA [9]. Amplifications of poppy DNA extracted from filling of the baked product were more difficult. As expected, primer pair designed for amplification of 553 bp fragment of reference gene was not able to generate amplicon (data not shown). The Bayer BioScience N.V. method [25] and the QIAamp DNA Stool Mini Kit provided DNA with quality allowing amplification of the 156 bp microsatellite and short (96 bp) fragment of reference gene (Figure 5). Both these methods were effective also without the need of purification in columns. DNA extracted by the Monsanto Company method [26] and newly developed protocol was amplifiable only if the purification step in the silica membrane column [28] was added (Figure 5). Columns were able to bind impurities and inhibitors of polymerase chain reaction from primary DNA extracts. Complex food matrices contain a variety of PCR inhibitors [45]. Other effects of the matrix include degradation, fragmentation, and restricted extractability of DNA, as well as presence of DNA from different organisms [46]. Baking temperature around 200 °C used in processing of bakery goods containing poppy seed filling substantially reduces the size of extracted DNA. Moreover, higher moisture content inside the product, in this case in poppy filling, contributes to greater degradation of DNA [9]. Amplifications of poppy DNA extracted from filling of the baked product were more difficult. As expected, primer pair designed for amplification of 553 bp fragment of reference gene was not able to generate amplicon (data not shown). The Bayer BioScience N.V. method [25] and the QIAamp DNA Stool Mini Kit provided DNA with quality allowing amplification of the 156 bp microsatellite and short (96 bp) fragment of reference gene (Figure 5). Both these methods were effective also without the need of purification in columns. DNA extracted by the Monsanto Company method [26] and newly developed protocol was amplifiable only if the purification step in the silica membrane column [28] was added (Figure 5). Columns were able to bind impurities and inhibitors of polymerase chain reaction from primary DNA extracts.
**Figure 5.** Amplification of 156 bp microsatellite psSSR69 in DNA extracted from the poppy seed filling (**a**) using: Bayer BioScience N.V. [25] (lane 1), QIAamp DNA Stool Mini Kit (2), newly developed protocol (3, 5, 7), Monsanto Company [26] (4, 6) NC—negative control, PC—positive control. Lanes 1–4 represent samples without, lanes 5–7 with purification through silica membrane columns. Amplification of 96 bp fragment of the reference tubulin gene (**b**) using: Bayer BioScience N.V. [25] (lanes 1, 2, 9, 10), QIAamp DNA Stool Mini Kit (lanes 3, 4, 11, 12), newly modified protocol (lanes 5–7, 13–16), Monsanto Company [26] (lane 8, 17) NC-negative control, PC-positive control. Lanes 1–8 represent samples without, lanes 9-17 samples with purification through columns. M1—25 **Figure 5.** Amplification of 156 bp microsatellite psSSR69 in DNA extracted from the poppy seed filling (**a**) using: Bayer BioScience N.V. [25] (lane 1), QIAamp DNA Stool Mini Kit (2), newly developed protocol (3, 5, 7), Monsanto Company [26] (4, 6) NC—negative control, PC—positive control. Lanes 1–4 represent samples without, lanes 5–7 with purification through silica membrane columns. Amplification of 96 bp fragment of the reference tubulin gene (**b**) using: Bayer BioScience N.V. [25] (lanes 1, 2, 9, 10), QIAamp DNA Stool Mini Kit (lanes 3, 4, 11, 12), newly modified protocol (lanes 5–7, 13–16), Monsanto Company [26] (lane 8, 17) NC-negative control, PC-positive control. Lanes 1–8 represent samples without, lanes 9-17 samples with purification through columns. M1—25 bp DNA ladder (Invitrogen).
#### bp DNA ladder (Invitrogen). *3.5. DNA from Poppy Oil*
*3.5. DNA from Poppy Oil* Oil from poppy seeds is mainly used for culinary and pharmaceutical purposes, but also for production of cosmetics, paints and varnishes. Cold-pressed oil is quite expensive, so it can sometimes be adulterated by much cheaper vegetable oils (e.g., from rapeseed, sunflower, oil palm). Techniques of analytical chemistry are developing for distinguishing between cheaper oils (e.g., sunflower, oilseed rape) and poppy oil [17]. However, chemical analysis may not be unambiguous [31] due to variation in chemical composition of vegetable oils among growing areas and seasons. Alternative approaches are based on the DNA analysis and require extraction of DNA from oil. Such protocols were developed mainly for olive oil. Four of such methods [4,29–31], the QIAamp DNA Stool Mini Kit as well as Bayer BioScience N.V. [25], Monsanto Company [26], newly developed protocols were tested for different volumes of poppy seed oil. Bayer BioScience N.V. [25], Monsanto Company [26] and newly developed protocol were unable to extract detectable and usable DNA (data not shown). DNA extracted by other protocols had also both absorbance parameters (A260/280, Oil from poppy seeds is mainly used for culinary and pharmaceutical purposes, but also for production of cosmetics, paints and varnishes. Cold-pressed oil is quite expensive, so it can sometimes be adulterated by much cheaper vegetable oils (e.g., from rapeseed, sunflower, oil palm). Techniques of analytical chemistry are developing for distinguishing between cheaper oils (e.g., sunflower, oilseed rape) and poppy oil [17]. However, chemical analysis may not be unambiguous [31] due to variation in chemical composition of vegetable oils among growing areas and seasons. Alternative approaches are based on the DNA analysis and require extraction of DNA from oil. Such protocols were developed mainly for olive oil. Four of such methods [4,29–31], the QIAamp DNA Stool Mini Kit as well as Bayer BioScience N.V. [25], Monsanto Company [26], newly developed protocols were tested for different volumes of poppy seed oil. Bayer BioScience N.V. [25], Monsanto Company [26] and newly developed protocol were unable to extract detectable and usable DNA (data not shown). DNA extracted by other protocols had also both absorbance parameters (A260/280, A260/230) far from the optimal values (Table 3); however, DNA was amplifiable by PCR (Figure 6). DNA in cold-pressed
A260/230) far from the optimal values (Table 3); however, DNA was amplifiable by PCR (Figure 6).
vegetable oil has undergone a process of significant degradation, caused by DNA nucleases released during crushing and malaxation of oily plant material. This will certainly happen when pressing oil from poppy seeds as well. If enzymatic mixtures of proteases are applied during this process, the DNA is prevented to damage and could be extracted with high integrity and concentration, similarly as from vegetative tissues [47]. However, this cannot be ensured in the already pressed oil. Another significant complication in the extraction of DNA is the time since pressing and conditions of the oil storage before the DNA extraction. After a relatively short time interval, a significant decreasing of quality of extracted DNA was observed due to oxidation damage [48]. Following the assumed high degradation, DNA has not even been electrophoretically controlled and only its amplifications revealed the potential utility of the extracted DNA. Statistical analysis did not reveal relationship between concentration, A260/A<sup>280</sup> ratio, and the ability to undergo amplification by PCR [49]. analysis did not reveal relationship between concentration, A260/A280 ratio, and the ability to undergo amplification by PCR [49]. Four extraction protocols [4,29–31] and the QIAamp DNA Stool Mini Kit provided different results (Figure 6). In addition, DNA extraction was also tested from different starting volumes of poppy seed oil. Extraction protocol developed for authentication of olive oils [30] was efficient either from 3 mL or 6 mL samples of poppy oil. Poppy DNA obtained by this protocol, from both the oily and water phases were amplifiable and provided templates for relevant amplicons. Other used DNA extraction protocols were also developed for olive oil, but based on the CTAB in extraction buffer [4,31]. The resulting poppy DNA behaved unreliably in the PCR reaction. Convincing and reliable amplifications were obtained from DNA extracted by another protocol, modified for olive oil [29] containing guanidine thiocyanate in extraction buffer. The capability of tested QIAamp DNA Stool Mini Kit for DNA extraction from poppy oil has been demonstrated in low oil volumes (0.2–1 mL).
controlled and only its amplifications revealed the potential utility of the extracted DNA. Statistical
*Foods* **2020**, *9*, x FOR PEER REVIEW 11 of 15
DNA in cold-pressed vegetable oil has undergone a process of significant degradation, caused by DNA nucleases released during crushing and malaxation of oily plant material. This will certainly happen when pressing oil from poppy seeds as well. If enzymatic mixtures of proteases are applied during this process, the DNA is prevented to damage and could be extracted with high integrity and concentration, similarly as from vegetative tissues [47]. However, this cannot be ensured in the already pressed oil. Another significant complication in the extraction of DNA is the time since pressing and conditions of the oil storage before the DNA extraction. After a relatively short time interval, a significant decreasing of quality of extracted DNA was observed due to oxidation damage
**Figure 6.** Agarose gel electrophoresis of PCR products obtained by amplification of 156 bp microsatellite psSSR69 (**a**) and 96 bp fragment of reference tubulin gene (**b**). M1—25 bp DNA ladder (Invitrogen). DNA extracted by Consolandi et al. [30] (lines 1–6) from 3 mL (1–3) or 6 mL (4–6) of oil, Raieta et al. [4] (lanes 7–10) from 3 mL (lanes 7,8) or 1 mL (lanes 9–10) oil, Doveri et al. [29] (line 11) extracted from 1 mL of oil, Giménez et al. [31] (lines 12–13) extracted from 0.5 mL (lane 12) or 3 mL (lane 13) of oil, QIAamp DNA Stool Mini Kit (lanes 14–16) extracted from 0.2 mL (lane 14b), 1 mL (lane 15b) or 15 mL (lanes 14a and 16b) of oil, NC—negative control, PC—positive control. aq/o—DNA from water (aq) or oily (o) phase. **Figure 6.** Agarose gel electrophoresis of PCR products obtained by amplification of 156 bp microsatellite psSSR69 (**a**) and 96 bp fragment of reference tubulin gene (**b**). M1—25 bp DNA ladder (Invitrogen). DNA extracted by Consolandi et al. [30] (lines 1–6) from 3 mL (1–3) or 6 mL (4–6) of oil, Raieta et al. [4] (lanes 7–10) from 3 mL (lanes 7,8) or 1 mL (lanes 9–10) oil, Doveri et al. [29] (line 11) extracted from 1 mL of oil, Giménez et al. [31] (lines 12–13) extracted from 0.5 mL (lane 12) or 3 mL (lane 13) of oil, QIAamp DNA Stool Mini Kit (lanes 14–16) extracted from 0.2 mL (lane 14b), 1 mL (lane 15b) or 15 mL (lanes 14a and 16b) of oil, NC—negative control, PC—positive control. aq/o—DNA from water (aq) or oily (<sup>o</sup> ) phase.
The quality and quantity of DNA extracted from native or processed poppy seeds strongly depended on the character of poppy matrix entering the extraction procedure as well as level of its processing. Amplifications of obtained DNA were also influenced by many factors, especially by the presence of contaminants and inhibitors. Positioning of used primers for PCR analysis considered the expected length of extracted DNA fragments depended on the expected disruption of DNA during processing (baking, pressing) of poppy seed matrix. DNA extracted from different poppy seed matrices by different extraction protocols was amplified using primer pairs flanking the 553-, 156-, and 96 bp fragments, respectively (Table 1). The presence of the longest 553 bp fragment was detected by PCR in poppy DNA extracted from native seeds and ground seeds, but not from processed poppy seed matrices (filling of the bakery product, oil). Both types of poppy seed Four extraction protocols [4,29–31] and the QIAamp DNA Stool Mini Kit provided different results (Figure 6). In addition, DNA extraction was also tested from different starting volumes of poppy seed oil. Extraction protocol developed for authentication of olive oils [30] was efficient either from 3 mL or 6 mL samples of poppy oil. Poppy DNA obtained by this protocol, from both the oily and water phases were amplifiable and provided templates for relevant amplicons. Other used DNA extraction protocols were also developed for olive oil, but based on the CTAB in extraction buffer [4,31]. The resulting poppy DNA behaved unreliably in the PCR reaction. Convincing and reliable amplifications were obtained from DNA extracted by another protocol, modified for olive oil [29] containing guanidine thiocyanate in extraction buffer. The capability of tested QIAamp DNA Stool Mini Kit for DNA extraction from poppy oil has been demonstrated in low oil volumes (0.2–1 mL).
processing (baking, pressing) reduced the effective concentration of poppy DNA fragments capable The quality and quantity of DNA extracted from native or processed poppy seeds strongly depended on the character of poppy matrix entering the extraction procedure as well as level of its processing. Amplifications of obtained DNA were also influenced by many factors, especially by the presence of contaminants and inhibitors. Positioning of used primers for PCR analysis considered the expected length of extracted DNA fragments depended on the expected disruption of DNA during processing (baking, pressing) of poppy seed matrix. DNA extracted from different poppy seed matrices by different extraction protocols was amplified using primer pairs flanking the 553-, 156-, and 96 bp fragments, respectively (Table 1). The presence of the longest 553 bp fragment was detected by PCR in poppy DNA extracted from native seeds and ground seeds, but not from processed poppy seed
matrices (filling of the bakery product, oil). Both types of poppy seed processing (baking, pressing) reduced the effective concentration of poppy DNA fragments capable of amplification of fragments longer than 100 bp, as was detected in maize cornmeal [50]. DNA from heat-processed and other highly degraded plant matrices should be amplified only in short DNA sequences. This is the strategy also in analysis of DNA from genetically modified organisms in processed foods [9,51]. Analysis of highly degraded DNA by PCR is more advantageous in DNA regions higher in GC content because their stability during heat treatment of the analysed matrices is higher [51].
Specific morphological characteristics, extreme heterogeneity and variation in chemical composition of plant cells cause many problems in DNA extraction. Although numerous protocols for plant DNA extraction have been published, none is found to be universally applicable [52]. Newly developed DNA extraction protocols are usually modifications of already existing protocols. The extraction protocol developed in our study demonstrated a relatively high degree of universality, with respect to poppy matrices. Compared to other DNA extraction protocols, it was quite universal. In comparison with the Bayer BioScience N.V. [25] protocol, from which the most steps were taken, it was approximately one third shorter in time. A significant reduction in time was achieved by adjusting the centrifugation steps. 2-mercaptoethanol ME was added to the first extraction buffer. Some steps during the extraction procedure were eliminated. Along with purification on silica membrane columns, the newly developed extraction protocol was highly efficient and represents a simple and inexpensive alternative to commercial DNA extraction kits. Extraction of DNA from oil required specific extraction protocols that were developed specifically for this type of matrix only.
#### **4. Conclusions**
Protocols tested for extraction of DNA from native and ground poppy seeds, pollen grains, poppy seed filling from the bakery product, and poppy oil have been differently effective and suitable depending on individual poppy seed matrices or products. DNA from seeds, ground seeds and pollen grains extracted by almost all extraction procedures had quantity and quality sufficient for PCR analysis of short microsatellite marker (156 bp) and also long fragment of the reference gene (553 bp). The best of these protocols have been tested for DNA extraction from the poppy seed filling from the bakery product. It has been very useful to use silica membrane columns for purification of the extracted DNA. Purified DNA was then amplifiable. Poppy DNA extracted from thermally processed poppy seed filling from the baking product did not amplify long fragment (553 bp) of the reference gene. However, primers designed for amplification of shorter fragment of the reference gene (96 bp) as well as for the microsatellite marker (156 bp) provided the appropriate amplicons. The new extraction protocol developed within this study has proven to be universally applicable to poppy seeds, pollen, and poppy seed containing products. It can be used for various control purposes in poppy breeding programmes, production and distribution of elite poppy seeds for crop production, control of poppy seeds identity as an interesting market commodity, control of products containing poppy seeds during food production. Protocols tested for extraction of poppy DNA from cold-pressed poppy oil were originally developed or modified for olive oil. The most of them [29,30] were effective, and extracted DNA was amplified using primers for the microsatellite marker and the short fragment of the reference gene.
**Author Contributions:** Conceptualization, J.K.; methodology, J.K., Š.K. and D.M.; investigation, Š.K., M.M. and D.M.; writing—original draft preparation, Š.K.; writing—review and editing, J.K.; visualization, Š.K. and M.M.; supervision, J.K. All authors have read and agreed to the published version of the manuscript.
**Funding:** This research was funded by the Slovak Research and Development Agency, projects no. APVV-16-0026, APVV-18-0005, APVV-16-0051, and the Operational Programme Research and Development co-financed from the European Regional Development Fund, project No ITMS 26210120039.
**Conflicts of Interest:** The authors declare no conflict of interest.
## **References**
© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
## **Development and Validation of a Real-Time PCR Based Assay to Detect Adulteration with Corn in Commercial Turmeric Powder Products**
## **Su Hong Oh and Cheol Seong Jang \***
Plant Genomics Laboratory, Department of Bioresource Sciences, Kangwon National University, Chuncheon 24341, Korea; [email protected]
**\*** Correspondence: [email protected]; Tel.: +82-33-250-6416; Fax: +82-33-259-5558
### Received: 9 June 2020; Accepted: 3 July 2020; Published: 5 July 2020
**Abstract:** Turmeric, or *Curcuma longa*, is commonly consumed in the South East Asian countries as a medical product and as food due to its therapeutic properties. However, with increasing demand for turmeric powder, adulterated turmeric powders mixed with other cheap starch powders, such as from corn or cassava, are being distributed by food suppliers for economic benefit. Here, we developed molecular markers using quantitative real-time PCR to identify adulteration in commercial turmeric powder products. Chloroplast genes, such as *matK, atpF*, and *ycf2*, were used to design species-specific primers for *C. longa* and *Zea mays*. Of the six primer pairs designed and tested, the correlation coefficients (R<sup>2</sup> ) were higher than 0.99 and slopes were −3.136 to −3.498. The efficiency of the primers was between 93.14 and 108.4%. The specificity of the primers was confirmed with ten other species, which could be intentionally added to *C. longa* powders or used as ingredients in complex turmeric foods. In total, 20 blind samples and 10 commercial *C. longa* food products were tested with the designed primer sets to demonstrate the effectiveness of this approach to detect the addition of *Z. mays* products in turmeric powders. Taken together, the real-time PCR assay developed here has the potential to contribute to food safety and the protection of consumer's rights.
**Keywords:** anti food fraud; *Curcuma longa*; DNA markers; species identification; SYBR-GREEN real-time PCR; *Zea mays*
### **1. Introduction**
Turmeric (*Curcuma longa*) belongs to the ginger family, Zingiberaceae, and is native to Southern Asia and India. Turmeric rhizomes, which have brown skin and a unique flavor, are commonly used as a coloring and flavoring agent in Asian cuisines. Due to its fragrant aroma and slightly bitter taste, turmeric is a common culinary spice in Indian cuisines, especially curry. Additionally, beyond food products, turmeric is commonly consumed as a medical product in South East Asian countries due to its therapeutic properties [1]. The market size of curcumin was valued at USD 58.4 million in 2019 and is expected to experience a CAGR (compound annual growth rate) of 12.7% from 2020 to 2027 [2]. Globally, the demand for turmeric has grown due to its therapeutic functions and low toxicity. Curcumin, (1,7-bis(4-hydroxy-3-methoxyphenyl)-1,6-heptadiene-3,5-dione), also known as diferuloylmethane, is the main natural polyphenol found in rhizomes of *C. longa* (turmeric) and in other *Curcuma* spp. [3]. It has been shown to target multiple signaling molecules while also demonstrating activity at the cellular level, which has helped support its multiple health benefits [4]. It has beneficial effects in inflammatory conditions [3], metabolic syndrome [5], and pain [6], as well as helps in the management of inflammatory and degenerative eye conditions [7]. While there appear to be countless therapeutic benefits of curcumin supplementation, most of them may be due to its antioxidant and anti-inflammatory effects [3].
Reports on the medicinal value of turmeric in treating a variety of ailments have further increased the global demand for turmeric [8]. In the United States, the largest market for turmeric supplements, turmeric was the top-selling herbal supplement, with sales exceeding US \$47.6 million in 2016 [8,9]. In addition, turmeric-based dietary supplements, which also include standardized extracts with high concentrations of curcumin, have seen a steady increase in popularity in the United States and elsewhere [10,11]. However, with the increasing demand for turmeric powder, adulterated turmeric powders mixed with other cheap starch powders, such as from corn or cassava, have been distributed by food supplies for economic benefit [12]. According to the United States Grocery Manufacturers Association, food fraud costs \$10–15 billion annually in the global food industry and affects approximately 10% of all commercial foods sold [13].
To detect fraudulent ingredients in complicated mixed foods, various technologies, such as sensory-, physicochemical-, chromatographic-, spectroscopic-, and DNA-based assays have been developed. DNA is generally believed to be stable enough to withstand various chemical treatments and high temperatures, and small quantities of DNA can be detected with specific primers using PCR-based methods [14]. DNA-based methods, such as quantitative real-time PCR (real-time PCR), multiplex PCR, and PCR-RFLP have been successfully applied to detect food fraud and adulteration due to their economical and time-saving advantages over other approaches [15,16].
Specifically, real-time PCR (real-time PCR) assay presents with high specificity and sensitivity, capable of detecting very small amounts of target DNA in complex foods. General types of real-time PCR approaches, probe-based real-time PCR (TaqMan assay), and DNA intercalating dye-based real-time PCR (SYBR Green I assay) have been employed for the detection and identification of DNA [17]. Probe-based real-time PCR detects the target sequence with specificity using probes designed to be complementary to a target sequence [18]; however, this approach requires many SNPs or indels to differentiate species, and it is difficult to design probes and optimize real-time PCR conditions [19,20]. Alternatively, SYBR Green I, an intercalating dye that binds to double-stranded DNA in a sequence independent manner, can provide a more flexible, convenient, and inexpensive method over probe-based methods [19].
It is generally believed that the nuclear genome of a cell has a single copy of a particular gene along with a few sequences in low copy numbers; hence, it is difficult to obtain high uniformity in PCR amplification. Especially, DNA extracted from processed commercial foods is of low quality, possibly because of degradation caused by the processes of drying, heating, fermentation, and addition of ingredients. Therefore, markers designed on the extracted nuclear DNA from processed foods exhibit a low ability to discriminate between species because of the low quality of nuclear DNA that has either a single gene or low-copies of genes [21,22]. The chloroplast genome size varies among species, ranging from 107 to 208 kb and consisting of a single circular molecule of DNA that is generally present in hundreds of copies per cell [23]. Chloroplasts are composed of two layers of membranes that enable chloroplasts to persist through decomposition during food processing [24]. The chloroplast genome is generally believed to contain 120–130 genes [22]. Some genes, such as *matK*, *ndhF*, *ycf* 2, and *ccsA*, exhibit higher frequencies of single-nucleotide polymorphisms (SNPs) and insertion/deletions (indels) than other chloroplast genes [23]. A variety of chloroplast markers, including *atpF*-*atpH* spacer, *matK* gene, *rbcL* gene, *rpoB* gene, *rpoC1* gene, *psbK-psbl* spacer, and *trnH-psbA* spacer, have been employed for species identification [25,26].
As described above, cheap corn powder with a similar color to turmeric has been wildly used in adulterated turmeric powders by food suppliers for illegal economic benefit. In this study, we developed SYBR Green-based quantitative real-time PCR assay to identify adulteration in commercial turmeric powder products using turmeric and corn species-specific primer sets. The real-time PCR methodology was optimized for both species-specific primers to correctly identify target species in complex powder products. Subsequently, the designed primers were applied to commercial turmeric products.
### **2. Materials and Methods**
#### *2.1. Plant and Food Sample Preparation*
Turmeric (*Curcuma longa*) rhizomes and corn (*Zea mays*) seeds were kindly provided by Gangwondo Agriculture Research and Extension Services (Chuncheon, Korea). Both plants were grown in a stable temperature greenhouse for four weeks with horticulture soil. Samples for DNA isolation were extracted from the leaves of each plant. All *C. longa* commercial products used for the analysis of food complexes were purchased from local markets and stored at room temperature.
#### 2.1.1. Reference Binary Mixtures
To generate a quantitative reference binary mixture model, binary mixtures containing different amounts (2 mg, 0.1%; 20 mg, 1%; 200 mg, 10%; and 2 g, 100%) of turmeric rhizome powders were mixed to prepare a final mixture of 2 g with corn powder, wheat flour, or rice flour purchased from a local market. Additionally, different amounts of corn powder were mixed (2 mg, 0.1%; 20 mg, 1%; 200 mg, 10%; and 2 g, 100%) to prepare final mixtures of 2 g with turmeric rhizome powders. Turmeric rhizomes and corn seeds were dried in a 55 ◦C dry oven for 48 hours and then ground with a mixing machine.
#### 2.1.2. Blind Samples
Blind powder samples (*n* = 20) were provided by the National Institute of Food and Drug Safety Evaluation of the Ministry of Food and Drug Safety (Cheongju, Korea). The blind samples consisted of different percentages of corn and turmeric rhizome powders. The corn powders were added to turmeric rhizome powders at concentrations of 0–10% *w*/*w*, to prepare final mixtures of 150 mg.
#### *2.2. DNA Extraction*
For the efficiency of the designed primer sets, genomic DNA used for standard curves was extracted from *C. longa* and *Z. mays* leaves using the Dneasy Plant Pro Kit (QIAGEN, Hilden, Germany) according to the manufacturer's protocol. Genomic DNA used to plot standard curves of reference binary mixtures was isolated from the binary mixture samples (2 g each) using a large scale CTAB-based genomic DNA isolation method [27]. Genomic DNA from the commercial turmeric products was extracted using the Dneasy Plant Pro Kit according to the manufacturer's protocol. To obtain high quality genomic DNA, DNA extracted with the large scale CTAB method was purified using the Wizard DNA Clean-Up system (Promega, Madison, USA). DNA quantity and purity were measured using a SPECTROstar Nano reader (BMG Labtech, Ortenberg, Germany). Purity of the DNA extracts was in the range of 1.7–2.
#### *2.3. Sequence Analysis and Primer Design*
Sequences of target chloroplast genes such as *matK*, *atpF*, and *ycf2* of two species (*C. longa* for NC\_042886.1 and *Z. mays* for NC\_001666.2) were downloaded from the National Center for Biotechnology Information (NCBI) and used to design target-specific primers. The nucleotide sequences of the both species were aligned using ClustalW2 (EMBL-EBI, Hinxton, Cambridgeshire, UK) and BioEdit 7.2 (Ibis Biosciences, Carlsbad, CA, USA). Species-specific primer sets were designed based on the variable region between *C. longa* and *Z. mays* using Beacon DesignerTM (PRIMER Biosoft, Palo Alto, CA, USA). Species-specific primers were commercially synthesized (Macrogen, Seoul, Korea).
#### *2.4. Quantitative Real-Time PCR*
Real-time PCR was performed in a final volume of 20 <sup>µ</sup>L using AccuPower® <sup>2</sup><sup>×</sup> GreenStar™ real-time PCR Master Mix with SYBR Green (Bioneer, Daejeon, Korea). The real-time PCR reaction mixture consisted of 10 µL 2× GreenStar Master Mix, 0.5 µL 10 pmol each primer, 1 µL of 10 ng/µL genomic DNA, and 0.25 µL ROX Dye. A QuantStudio 3 Real-Time PCR System (Applied Biosystems, Foster City, CA, USA) was used for real-time PCR amplification. The real-time PCR conditions were as follows: pre-denaturation (10 min at 95 ◦C), followed by 40 cycles of denaturation for 30 s at 95 ◦C, annealing for 20 s at 55–60 ◦C (depending on each targeting primer sequence), and extension for 30 s at 72 ◦C. All real-time PCRs were performed in technical triplicates for three biological replicates.
#### *2.5. Cloning of PCR Amplicons and Sequencing*
Conventional PCR was carried out using TaKaRa Ex TaqTM DNA polymerase (TaKaRa Bio Company, Kusatsu, Shiga, Japan) mixture with 10 ng DNA and 10 pmol each primer using a C1000 Thermal Cycler (BIO-RAD, California, USA). PCR conditions were as follows: pre-denaturation for 5 min at 95 ◦C, followed by 35 cycles of annealing and denaturation for 30 s at 95 ◦C, annealing for 20 s at 55–60 ◦C (depending on primer sequences), and extension 30 s at 72 ◦C, and final extension for 5 min at 72 ◦C. PCR products were amplified using target specific primers (CL\_matK, CL\_atpF, CL\_ycf2, ZM\_matK, ZM\_atpF, and ZM\_ycf2) and cloned using the RBC T&A Cloning Vector (Real Biotech Corporation, Taipei, Taiwan). Plasmid DNA was extracted from recombinant plasmids using the DokDo-Prep Plasmid Mini-Kit (ELPISBIOTECH, DaeJeon, South Korea) and sequenced by a commercial service (Macrogen, Seoul, Korea).
#### *2.6. Standard Curve Construction and Data Analysis*
The efficiency of the designed primer sets was evaluated using two approaches. First, species-specific PCR products were cloned into the RBC T&A Cloning Vector (Real Biotech Corporation, Taipei, Taiwan), and recombinant clones were then diluted serially (10<sup>7</sup> , 10<sup>6</sup> , 10<sup>5</sup> , 10<sup>4</sup> , and 10<sup>3</sup> copies) and used to quantify and confirm the efficiency of equivalent amplification [28,29]. Second, real-time PCR assays were applied to genomic DNA using target and non-target gDNA diluted ten-fold into five series (10 ng to 1 pg).
Each binary mixture with genomic DNAs extracted from the leaves or powder products of each species was diluted to a final concentration of 10 ng/µL. A baseline and a threshold were set for further analysis. The cycle number at the threshold level of log-based fluorescence was defined as the Ct (cycle threshold) number, which was the observed value in the conventional real-time PCR experiments [30]. Correlations between diluted DNAs and cycle threshold (Ct) standard curves were evaluated using a default parameter. The standard curve was calculated as y = −ax + b (a refers to the standard curve slope and b refers to the y-intercept). The efficiency of the reaction (E) was calculated as E = (10−1/<sup>a</sup> ), and the percent efficiency was evaluated as (E − 1) × 100% [29,30]. For all analyses, three technical replicates of each biological replicate were performed.
To evaluate amplification efficiency and sensitivity, two criteria were used to define an acceptable real-time PCR assay based on previous reports [28,29]: linear dynamic range and amplification efficiency. The linear dynamic range should ideally extend over four log<sup>10</sup> concentrations, with the coefficient of determination (R<sup>2</sup> ) being greater than 0.98, and the amplification efficiency should be in the range of 110-90%, corresponding to a slope between −3.1 and −3.6 [29].
To validate the specificity and sensitivity of the designed target-specific primers, interlaboratory validation was performed in two independent laboratories. Validation was performed in two laboratories using the same PCR conditions and with either an Applied Biosystems 7500 Fast Real-Time PCR Instrument System (Applied Biosystems, Foster City, CA, USA) or a CFX Connect Real-Time PCR Detection System (Bio-Rad, Hercules, CA, USA).
#### **3. Results and Discussion**
#### *3.1. Design of Species-Specific Primers*
To verify authenticity of *C. longa* commercial food products, we designed species-specific primer pairs for *C. longa* and *Z. mays*. Chloroplast genes, such as *matK, atpF,* and *ycf2*, with high frequencies of SNPs and indels between the two species [25] were targeted to design the species-specific primer sets. For designing species-specific primers, chloroplast genes of both species, as well as those of other starch crops (*Oryza sativa* and *Triticum aestivum*), were aligned using a software program with ClustalW2 (EMBL-EBI, Hinxton, Cambridgeshire, UK) and BioEdit 7.2 (Ibis Biosciences, Carlsbad, CA, USA; Supplementary Figures S1 and S2). We identified a variety of SNPs within three chloroplast genes among four species (Supplementary Figure S1). Food processing, such as heating, drying, and mixing, is known to damage and degrade DNA [31]. If the length of PCR amplicons is long, real-time PCR would be decreased in various food products. Therefore, based on species-specific SNPs, target-specific primers were designed to amplify short products ranging from 80 to 194 bp (Table 1).
**Table 1.** Primer sets designed for species-specific targeting.
#### *3.2. Amplification E*ffi*ciency of the Designed Primer Sets*
Amplification efficiency of the six primer sets (CL\_matK, CL\_atpF, CL\_ycf2, ZM\_matK, ZM\_atpF, and ZM\_ycf2) was evaluated by constructing standard curves using 10-fold serial dilutions (10<sup>7</sup> to 10<sup>3</sup> ) of each recombinant plasmid DNA, and regression analyses were performed (Figure 1, Supplementary Figure S3). The correlation coefficients (R<sup>2</sup> ) of the six primer pairs were higher than 0.99, and slopes ranged from −3.14 to −3.50. The efficiency of the primers was between 93.14 and 108.40% (Supplementary Table S1). All values fit the ENGL (European Network of GMO Laboratories) guidelines, with the coefficient of determination (R<sup>2</sup> ) being greater 0.98 and the amplification efficiency ranging from 110 to 90%, which corresponds to a slope between −3.1 and −3.6 [29]. Subsequently, we evaluated the efficiency of the primers using the 10-fold serially diluted genomic DNAs (from 10 ng to 1 pg) extracted from plant samples (Figure 2, Supplementary Figure S4). Similarly to the results of recombinant plasmid DNAs, standard curves in the gDNA samples also ranged from −3.42 to −3.54, exhibited *R* <sup>2</sup>> 0.99, and efficiency values of 91.78–95.92%, which also conformed to the ENGL guidelines (Supplementary Table S1) [29].
In addition, to evaluate the adaptability of the primes across machines, amplification efficiency was evaluated by two independent laboratories. As a result, the primer sets were found to meet the ENGL criteria (R<sup>2</sup> > 0.98 and efficiency ranges of 91.78–108.40; Supplementary Table S2). Based on the evaluation of amplification efficiency of the designed primer sets through three approaches, with recombinant plasmids, genomic DNA, and interlaboratory evaluation, the designed primer sets could be suitable to detect the target species.
*Foods* **<sup>2020</sup>**, *<sup>9</sup>*, 882 *Foods* **2020**, *9*, x FOR PEER REVIEW 6 of 12
*Foods* **2020**, *9*, x FOR PEER REVIEW 6 of 12
**Figure 1.** Standard curve of cycle threshold (Ct) values were obtained on the basis of efficiency and correlation of coefficient (*R*2) in serial dilution series recombinant plasmids (C. longa and Z. mays) using species-specific primer sets. The x-axis represents log number of plasmids and the y-axis represents means of Ct value ± SD. (**A**) C. longa targeting primer sets (CL\_matK, CL\_atpF, and CL\_ycf2). Green dots represent serial dilution series of recombinant plasmids (107–103) containing C. longa specific target gene (matK, atpF and ycf2) sequences; (**B**) Z. mays targeting primer sets (ZM\_matK, ZM\_atpF, and ZM\_ycf2). Blue dots represent serial dilution series of recombinant plasmids (107–103) containing Z. mays specific target gene (matK, atpF and ycf2) sequences. The real-**Figure 1.** Standard curve of cycle threshold (Ct) values were obtained on the basis of efficiency and correlation of coefficient (*R* 2 ) in serial dilution series recombinant plasmids (*C. longa* and *Z. mays*) using species-specific primer sets. The *x*-axis represents log number of plasmids and the *y*-axis represents means of Ct value ± SD. (**A**) *C. longa* targeting primer sets (CL\_matK, CL\_atpF, and CL\_ycf2). Green dots represent serial dilution series of recombinant plasmids (107–103) containing *C. longa* specific target gene (matK, atpF and ycf2) sequences; (**B**) *Z. mays* targeting primer sets (ZM\_matK, ZM\_atpF, and ZM\_ycf2). Blue dots represent serial dilution series of recombinant plasmids (107–103) containing *Z. mays* specific target gene (matK, atpF and ycf2) sequences. The real-time PCRs were carried out in triplicate (*n* = 3). correlation of coefficient (*R*2) in serial dilution series recombinant plasmids (C. longa and Z. mays) using species-specific primer sets. The x-axis represents log number of plasmids and the y-axis represents means of Ct value ± SD. (**A**) C. longa targeting primer sets (CL\_matK, CL\_atpF, and CL\_ycf2). Green dots represent serial dilution series of recombinant plasmids (107–103) containing C. longa specific target gene (matK, atpF and ycf2) sequences; (**B**) Z. mays targeting primer sets (ZM\_matK, ZM\_atpF, and ZM\_ycf2). Blue dots represent serial dilution series of recombinant plasmids (107–103) containing Z. mays specific target gene (matK, atpF and ycf2) sequences. The realtime PCRs were carried out in triplicate (*n* = 3).
**Figure 2.** Standard curve of cycle threshold (Ct) values were obtained on the basis of efficiency and correlation of coefficient (*R*2) in serial dilution series genomic DNA (C. longa and Z. mays) using species-specific primer sets. The x-axis represents log DNA concentration (ng) and the y-axis represents means of Ct value ± SD. (**A**) C. longa targeting primer sets (CL\_matK, CL\_atpF, and CL\_ycf2). Green dots represent serial dilution series of genomic DNA in C. longa leaves (10ng–1pg) and blue dots represent genomic DNA of Z. mays (10ng); (**B**) Z. mays targeting primer sets **Figure 2.** Standard curve of cycle threshold (Ct) values were obtained on the basis of efficiency and correlation of coefficient (*R*2) in serial dilution series genomic DNA (C. longa and Z. mays) using species-specific primer sets. The x-axis represents log DNA concentration (ng) and the y-axis represents means of Ct value ± SD. (**A**) C. longa targeting primer sets (CL\_matK, CL\_atpF, and CL\_ycf2). Green dots represent serial dilution series of genomic DNA in C. longa leaves (10ng–1pg) and blue dots represent genomic DNA of Z. mays (10ng); (**B**) Z. mays targeting primer sets (ZM\_matK, ZM\_atpF, and ZM\_ycf2). Blue dots represent serial dilution series of genomic DNA in Z. mays leaves (10ng–1pg) and green dot represents genomic DNA of C. longa (10ng). The real-time **Figure 2.** Standard curve of cycle threshold (Ct) values were obtained on the basis of efficiency and correlation of coefficient (*R* 2 ) in serial dilution series genomic DNA (*C. longa* and *Z. mays*) using species-specific primer sets. The *x*-axis represents log DNA concentration (ng) and the *y*-axis represents means of Ct value ± SD. (**A**) *C. longa* targeting primer sets (CL\_matK, CL\_atpF, and CL\_ycf2). Green dots represent serial dilution series of genomic DNA in *C. longa* leaves (10ng–1pg) and blue dots represent genomic DNA of *Z. mays* (10ng); (**B**) *Z. mays* targeting primer sets (ZM\_matK, ZM\_atpF, and ZM\_ycf2). Blue dots represent serial dilution series of genomic DNA in *Z. mays* leaves (10ng–1pg) and green dot represents genomic DNA of *C. longa* (10ng). The real-time PCRs were carried out in triplicate (*n* = 3).
#### (ZM\_matK, ZM\_atpF, and ZM\_ycf2). Blue dots represent serial dilution series of genomic DNA in Z. mays leaves (10ng–1pg) and green dot represents genomic DNA of C. longa (10ng). The real-time PCRs were carried out in triplicate (*n* = 3). *3.3. Sensitivity and Specificity of the Assay*
PCRs were carried out in triplicate (*n* = 3). In addition, to evaluate the adaptability of the primes across machines, amplification efficiency was evaluated by two independent laboratories. As a result, the primer sets were found to meet the Globally, most *C. longa*-containing foods are prepared with rhizomes and dry powders. Therefore, we tested the sensitivity and specificity of the designed *C. longa* primer sets with binary mixtures
In addition, to evaluate the adaptability of the primes across machines, amplification efficiency
(0.1–100% (*w*/*w*)) of *C. longa* dry rhizome powders containing each of three starch crops, including corn, rice, and wheat (Figure 3A–C). All three *C. longa* primer sets with slopes ranging from −3.35 to −3.550 exhibited *R* <sup>2</sup> > 0.99 and efficiency values of 91.29–98.84% when used on mixed powders of *C. longa* and each starch crop, supporting the high sensitivity of the primer sets for verifying the presence of *C. longa* in mixtures sets (Supplementary Table S3). Subsequently, sensitivity of the three *Z. mays* primer sets was tested with binary mixtures of *Z. mays* and *C. longa* (0.1–100% (*w*/*w*)). Similarly, the three *Z. mays* primer sets with slope ranging from −3.12 to −3.44 exhibited *R* <sup>2</sup> > 0.99 and efficiency values of 95.30–109.18% when used on mixed powders of *C. longa* and *Z. mays*, supporting the high sensitivity of the primer sets for verifying the presence of *Z*. *mays* in mixtures. Next, we determined the cut-off of Ct values based on the binary mixture standard collinearity equation of each primer set (Supplementary Table S3) to identify intended additions of cheap starch ingredients, such as *Z. mays*, in the *C. longa* powders. Ct values of 0.1% target species were determined as cut-off values for each primer set because additions of less than 0.1% of non-target species were not considered to be intended for illegal economic profit. The cut-off Ct values (0.1% target species in binary mixtures) were established to verify the presence of the target species from the calibration curves (Figure 3). The cut-off Ct values ranged from 26.82 to 29.59 cycles for each primer set targeting *C. longa* and 27.58 to 29.68 cycles for those targeting *Z. mays* (Supplementary Table S3).
Subsequently, we conducted a specificity test using the species-specific primer sets. A total of 10 species of cereals and vegetables were examined to assess cross-reactivity (Table 2). The cheap starch crops such as barley, wheat, oats, rice, sweat potato, and cassava, which are likely to be intentionally mixed as ingredients in complex turmeric foods for illegal economic profits, were included for the specificity test. In addition, one vegetable crop such as cabbage and one oilseed crop such as peanuts were used for the specificity test as out groups. 18S plant rRNA primer sets were used as a positive control [32], which exhibited lower Ct values than the cut-off. As shown in Table 2, Cl\_matK, CL\_atpF, and CL\_ycf2 exhibited *C. longa*-specific amplification but did not amplify the DNA of other species. Similarly, ZM\_matK, ZM\_atpF, and ZM\_ycf2 exhibited *Z. mays* specific amplification did not amplify the DNA of other species. The specificity test demonstrates that the primer sets could be useful for detecting the target species in unknown-ingredient powders and in complex food products.
**Table 2.** Results of the specificity test with other plants.
<sup>a</sup> Cycles, conventional PCR cycles based on cut-off (Ct) values of each specific primer sets (Ct values of 0.1% binary mixture ); <sup>b</sup> +, detected at less than Ct values of primers; <sup>c</sup> -, not detected before the primers0 Ct values.
could be suitable to detect the target species.
*3.3. Sensitivity and Specificity of the Assay*
*C. longa* and 27.58 to 29.68 cycles for those targeting *Z. mays* (Supplementary Table S3).
ENGL criteria (R2 > 0.98 and efficiency ranges of 91.78–108.40; Supplementary Table S2). Based on the evaluation of amplification efficiency of the designed primer sets through three approaches, with recombinant plasmids, genomic DNA, and interlaboratory evaluation, the designed primer sets
Globally, most *C. longa*-containing foods are prepared with rhizomes and dry powders. Therefore, we tested the sensitivity and specificity of the designed *C. longa* primer sets with binary mixtures (0.1–100% (*w*/*w*)) of *C. longa* dry rhizome powders containing each of three starch crops, including corn, rice, and wheat (Figure 3A–C). All three *C. longa* primer sets with slopes ranging from −3.35 to −3.550 exhibited *R*2 > 0.99 and efficiency values of 91.29–98.84% when used on mixed powders of *C. longa* and each starch crop, supporting the high sensitivity of the primer sets for verifying the presence of *C. longa* in mixtures sets (Supplementary Table S3). Subsequently, sensitivity of the three *Z. mays* primer sets was tested with binary mixtures of *Z. mays* and *C. longa* (0.1–100% (*w*/*w*)). Similarly, the three *Z. mays* primer sets with slope ranging from −3.12 to −3.44 exhibited *R*2 > 0.99 and efficiency values of 95.30–109.18% when used on mixed powders of *C. longa* and *Z. mays*, supporting the high sensitivity of the primer sets for verifying the presence of *Z*. *mays* in mixtures. Next, we determined the cut-off of Ct values based on the binary mixture standard collinearity equation of each primer set (Supplementary Table S3) to identify intended additions of cheap starch ingredients, such as *Z. mays*, in the *C. longa* powders. Ct values of 0.1% target species were determined as cut-off values for each primer set because additions of less than 0.1% of non-target species were not considered to be intended for illegal economic profit. The cut-off Ct values (0.1% target species in binary mixtures) were established to verify the presence of the target species from the calibration
**Figure 3.** Standard curve of cycle threshold (Ct) values obtained on the basis of efficiency and correlation of coefficient (*R*2) by reference binary mixtures. The x-axis represents log percentage of the target species (%) and the y-axis represents means of Ct value ± SD. plotted against the logarithm of the target species concentration (100, 10, 1, and 0.1%). (**A**–**C**); each C. longa rhizome powders were mixed with three different plant powders (Z. mays, O. sativa, and T. aestivum) by ten-fold dilutions (0.1, 1, 10 and 100%, final mass of 2g) and the each mixture gDNA(10ng/ul) was amplified using the C. longa targeting primer sets (CL\_matK, CL\_atpF, and CL\_ycf2). The green dotted line means the 0.1% binary mixture Cts amplified using the C' longa targeting primer sets, CL\_matK, CL\_atpF and CL\_ycf2) (**A**) binary mixture of C. longa and Z. mays; (**B**) binary mixture of C. longa and O. sativa; **Figure 3.** Standard curve of cycle threshold (Ct) values obtained on the basis of efficiency and correlation of coefficient (*R* 2 ) by reference binary mixtures. The *x*-axis represents log percentage of the target species (%) and the *y*-axis represents means of Ct value ± SD. plotted against the logarithm of the target species concentration (100, 10, 1, and 0.1%). (**A**–**C**); each *C. longa* rhizome powders were mixed with three different plant powders (*Z. mays*, *O. sativa*, and *T. aestivum*) by ten-fold dilutions (0.1, 1, 10 and 100%, final mass of 2g) and the each mixture gDNA(10 ng/uL) was amplified using the *C. longa* targeting primer sets (CL\_matK, CL\_atpF, and CL\_ycf2). The green dotted line means the 0.1% binary mixture Cts amplified using the *C. longa* targeting primer sets, CL\_matK, CL\_atpF and CL\_ycf2) (**A**) binary mixture of *C. longa* and *Z. mays*; (**B**) binary mixture of *C. longa* and *O. sativa*; (**C**) binary mixture of *C. longa* and *T. aestivum*. (**D**) *Z. mays* powders were mixed with *C. longa* rhizome powders by ten-fold dilutions (0.1, 1, 10 and 100%, final mass of 2g) and each mixture gDNA(10 ng/uL) was amplified using the *Z. mays* targeting primer sets (ZM\_matK, ZM\_atpF, and ZM\_ycf2). The blue dotted line means the 0.1% binary mixture Cts amplified using the *Z. mays* targeting primer sets, ZM\_matK, ZM\_atpF and ZM\_ycf2). The real-time PCRs were carried out in triplicate (*n* = 3).
#### *3.4. Application of the Developed Real-Time PCR Assay to Blind Samples*
A blind test was conducted to estimate the reliability of the developed real-time PCR assays. Twenty unknown powder samples of *C. longa* and *Z. mays* were mixed randomly by an independent research group. The 18S rRNA plant primer sets were used as positive amplification controls [32], which exhibited low Cts (13.27–16.21; Table 3).
Next, we determined whether *Z. mays* powder was present in the samples based on the cut-off Ct values of the designed primer sets (0.1% *Z. mays* in binary mixtures). As a result, we identified four samples (sample 3, 9, 12, and 19) with Ct values exceeding the cut-off Ct values, indicating that the samples did not contain *Z. mays* powder mixed in *C. longa* powder. The other 16 samples were exhibited lower Cts than the cut-off Ct values, indicating that those samples contained *Z. mays* powder. In addition, the ratio of *Z. mays* powder mixed into the 16 samples was predicted using the developed binary mixture assay of the three primer sets (ZM\_matK, ZM\_atpF, and ZM\_ycf2). The predicted percentage of *Z. mays* in each blind sample was extrapolated by inserting the Cts into the standard collinearity equation of each primer set (ZM\_matK, ZM\_atpF, and ZM\_ycf2). As a result, the predicted percentages of *Z. mays* present in each sample were consistent with those of the mixed samples (Table 3). Therefore, the real-time PCR methodologies developed in this study demonstrated high accuracy for detecting the addition of *Z. mays* in *C. longa* rhizome powders.
**Table 3.** Results of the blind mixture test for evaluating the reliability of the developed primer sets.
<sup>a</sup> Positive amplification control (18s rRNA); b, expected ratio of the *Z. mays*; C, not detected; <sup>d</sup> accordance.
#### *3.5. Application of the Developed Assay in Commercial Products*
To verify adulteration with corn powder of *C. longa* food products, we performed the developed real-time PCR assays on 10 *C. longa* commercial food products (Supplementary Table S4, Table 4). First, the quality of genomic DNA isolated from the food products was evaluated using a spectrometer. As depicted in Table 4, the 18S rRNA primer sets exhibited low Cts (14.01–19.82), indicating that the gDNA from all the commercial products was sufficient to provide amplifiable gDNA. We found that all *C. longa* commercial food products (samples 1–10) were amplified with lower Ct values (from 14.1 to 21.971 cycles) using the *C. longa* species-specific primers (CL\_matK, CL\_atpF, and CL\_ycf2) than the cut-off Ct values (Ct values of 0.1% *C. longa-*specific primer set in binary mixtures) for each primer set (CT values of CL\_matK, CL\_atpF, and CL\_ycf2 were 28.65, 28.60, and 29.59 cycles, respectively; Figure 3, Supplementary Table S3). Additionally, all samples were amplified with higher Ct values (from 30.23 cycles to not detected before 40 cycles) with *Z. mays* targeting primers (ZM\_matK, ZM\_atpF, and ZM\_ycf2) than the cut-off Ct values (Ct values of 0.1% *Z. mays-*specific primer sets in binary mixtures) for each primer set (28.41, 29.68, and 27.58 cycles, respectively; Supplementary Table S2). As a result, the commercial products purchased from local markets did not contain *Z. mays*, suggesting that the developed real-time PCR assays could be successfully applied to detect the presence of *Z. mays* in commercial complex *C. longa* products.
**Table 4.** Result of the real-time PCR assay using 10 commercial products.
<sup>a</sup> ND indicates not detected at less than 40 cycles.
#### **4. Conclusions**
A real-time PCR assay is a highly sensitive, rapid, and specific method to detect target-species in processed food complexes. We designed three chloroplast gene targeted primer sets for both *C. longa* and *Z. mays*. To assess the quantities of the target-species present, standard curves were constructed using recombinant plasmid DNA and binary DNA mixtures. The specificities of the designed primers were confirmed with ten other species. Blind sample analysis and the application to commercial *C. longa* food products supported the effectiveness of the real-time PCR assays to detect *Z. mays* products added for illegal economic profits. Therefore, the developed real-time PCR assay could contribute to food safety and the protection of consumer's rights.
**Supplementary Materials:** The following are available online at http://www.mdpi.com/2304-8158/9/7/882/s1, Figure S1: Alignment of the target chloroplast gene (*matK*, *atpF* and *ycf2*) nucleotide sequences of *C. longa*, *Z. mays* and starch crops (*O. sativa* and *T. aestivum*) mainly eating as powders amplified by *C. longa* specific primer sets (CL\_matK CL\_atpF and CL\_ycf 2), Figure S2: Alignment of the target chloroplast gene(*matK*, *atpF* and *ycf2*)nucleotide sequences of *Z. mays*, *C. longa* and starch crops(*O. sativa*, and *T. aestivum*) mainly eating as powders, amplified by *Z. mays* s pecific primer sets (ZM\_matK, ZM\_atpF and ZM\_ycf2), Figure S3: Real time PCR with SYBR Green and DNA melting curve analyses.(A) Serial dilution series recombinant plasmids (107–10<sup>3</sup> ) containing *C. longa* specific gene (*matK*, *atpF* and *ycf 2*) sequence were amplified using *C. longa* specific primer sets. (**B**) Serial dilution series recombinant plasmids (107–10<sup>3</sup> ) containing *Z. mays* specific gene (*matK*, *atpF* and *ycf 2*) sequence were amplified using *Z. mays* specific primer sets. The real time PCRs were performed on a QuantStudio 3 Real Time PCR System (Applied Biosystems, Foster City, CA, USA) and carried out in triplicate (*n* = 3), Figure S4: Real time PCR with SYBR Green and DNA melting curve analyses green lanes mean the *C. longa* blue lanes mean *Z. mays* and pink lanes mean NTC. (**A**) Serial dilution series of C. *longa* genomic DNA (10 ng–1 pg) was amplified using *C*. *longa* specific primer sets. (**B**) Serial dilution series of *Z*. *mays* (10 ng–1 pg) was amplified using *Z. mays* specific primer sets. The real time PCRs were performed on a QuantStudio 3 Real Time PCR System (Applied Biosystems, Foster City, CA, USA) and carried out in triplicate (*n* = 3), Table S1: Evaluation of slope, *R* 2 , and efficiency using the developed primer sets, Table S2: Result of the real-time PCR assay in an interlaboratory experiment, Table S3: Evaluation of the slope, *R* 2 , and efficiency using binary mixtures containing three different intentionally added powders, Table S4: Information on the commercial food products.
**Author Contributions:** C.S.J. conceived of the overall study; S.H.O. carried out the experiment; S.H.O and C.S.J. wrote the manuscript. All authors have read and agreed to the published version of the manuscript.
**Funding:** This research received no external funding.
**Acknowledgments:** This research was supported by a grant (17162MFDS065) from the Ministry of Food and Drug Safety.
**Conflicts of Interest:** The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
## **References**
© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
## **Authentication of** *Ginkgo biloba* **Herbal Products by a Novel Quantitative Real-Time PCR Approach**
#### **Liliana Grazina <sup>1</sup> , Joana S. Amaral <sup>2</sup> , Joana Costa <sup>1</sup> and Isabel Mafra 1,\***
Received: 7 August 2020; Accepted: 29 August 2020; Published: 4 September 2020
**Abstract:** *Ginkgo biloba* is a widely used medicinal plant. Due to its potential therapeutic effects, it is an ingredient in several herbal products, such as plant infusions and plant food supplements (PFS). Currently, ginkgo is one of the most popular botanicals used in PFS. Due to their popularity and high cost, ginkgo-containing products are prone to be fraudulently substituted by other plant species. Therefore, this work aimed at developing a method for *G. biloba* detection and quantification. A new internal transcribe spacer (ITS) marker was identified, allowing the development of a ginkgo-specific real-time polymerase chain reaction (PCR) assay targeting the ITS region, with high specificity and sensitivity, down to 0.02 pg of DNA. Additionally, a normalized real-time PCR approach using the delta cycle quantification (∆Cq) method was proposed for the effective quantification of ginkgo in plant mixtures. The method exhibited high performance parameters, namely PCR efficiency, coefficient of correlation and covered dynamic range (50–0.01%), achieving limits of detection and quantification of 0.01% (*w*/*w*) of ginkgo in tea plant (*Camellia sinensis*). The quantitative approach was successfully validated with blind mixtures and further applied to commercial ginkgo-containing herbal infusions. The estimated ginkgo contents of plant mixture samples suggest adulterations due to reduction or almost elimination of ginkgo. In this work, useful and robust tools were proposed to detect/quantify ginkgo in herbal products, which suggests the need for a more effective and stricter control of such products.
**Keywords:** adulteration; authenticity; *Ginkgo biloba*; plant infusions; real-time polymerase chain reaction
### **1. Introduction**
Ginkgo (*Ginkgo biloba* L.) is a millenary Chinese tree that belongs to the Ginkgoaceae family whose leaves are widely used for medicinal purposes [1]. Owing to its composition in pharmacologically active compounds, such as flavonol glycosides and terpene trilactones (bilobalides and ginkgolides) [2,3], ginkgo is used for its capacity to improve cognitive impairment in the elderly and quality of life in mild dementia. It is also known for its therapeutic action in peripheral circulatory illnesses, improving blood circulation and preventing clot formation [1,2,4–6]. Currently, different herbal products that have ginkgo as an ingredient are readily available in the global market, including in plant food supplements (PFS) and herbal infusions. According to recent surveys, ginkgo was the most popular botanical in PFS and is used in six European Union countries [7], while in the United States it ranked among the top 10 dietary supplements in the category of herbal/botanicals [8]. Moreover, the global market of *G. biloba* extracts, mainly intended for pharmaceutical and food supplement industries, was estimated to be US \$1590.5 million in 2018 and projected to reach US \$2379.2 million by 2028 [9]. The high demand of ginkgo in the global market and the increased value of ginkgo products make them potential targets for economically motivated adulteration. Frauds can be performed by the total or partial replacement of ginkgo with other plant species or by adding pure flavonols/flavonol glycosides or extracts (rich in flavonol glycosides) from other plant species, such as *Styphnolobium* japonicum (syn: *Sophora japonica*) and *Fagopyrum esculentum* Moench, belonging to the Fabaceae and Polygonaceae families, respectively [6].
Both pharmaceuticals and traditional herbal medicinal products (THMP) (either final products or the extracts used for their production) must comply with the Pharmacopeia standards, established for ginkgo leaves or extracts, to ensure the product's quality [3,10]. However, in the case of other ginkgo-containing products, such as herbal infusions and PFS that are legally considered as foods, they do not have to comply with those standards. Moreover, in these type of products, previous studies have reported adulterations associated with the partial or complete replacement of ginkgo with other plants [1,3]. Thus, it is crucial to provide analytical tools that allow the identification and quantification of *G. biloba* in herbal products classified as foods, making possible the verification of compliance with label statements.
Several analytical methodologies have been proposed for authenticity assessment of ginkgo-containing herbal products based on liquid chromatography coupled to mass spectrometry (LC-MS), high performance thin layer chromatography (HPTLC), HPTLC coupled with nuclear magnetic resonance and spectroscopy [1,3,11–14]. Those methodologies rely on the identification of bioactive compounds and/or chemical profile, which can be affected by several external factors, such as the plant part/tissue, plant age, environmental conditions, geographical location, and storage conditions, among others. Furthermore, chemical approaches can be less adequate when the formulation includes several plant species. On the contrary, DNA-based methodologies have been shown to be suitable tools for the identification/discrimination of species due to their high specificity and sensitivity, with different works reporting successful applications in the authentication of herbal products, namely food supplements or herbal infusions [15–17]. In this regard, different approaches including species-specific polymerase chain reaction (PCR), multiplex PCR, real-time PCR, high resolution melting (HRM) analysis, sequence characterization of amplified regions (SCAR), DNA barcoding, and next generation sequencing (NGS), among others, have been proposed to authenticate medicinal plants in herbal products [18]. Among them, real-time PCR offers the advantage of providing quantitative information, being a very sensitive, specific, and fast tool.
So far, only a few works regard the identification of *G. biloba* in herbal products and PFS using DNA-based approaches. Little [19] proposed the use of DNA barcoding targeting a short region of *matK* gene to identify gingko in PFS. Despite using a DNA mini-barcode (166 bp), 3 out of 40 samples were not successfully amplified. Besides, it should be noticed that this approach is not adequate for samples containing mixtures of ingredients/medicinal plants. Liu et al. [20] developed a rapid identification method to detect both gingko and a possible adulterant (*Sophora japonica*) in herbal products using a recombinase polymerase amplification (RPA) approach, which relied on the use of species-specific primers and a probe with high specificity, though with limited cross-reactivity testing. More recently, Dhivya et al. [21] developed a real-time PCR assay using a species-specific hydrolysis probe to identify *G. biloba* in natural health products. The method allowed the specific and sensitive detection of *G. biloba*, but without any quantitative analysis that should rely on the development of an adequate calibration model. Besides, the authors did not demonstrate its applicability in the analysis of processed/complex products. Therefore, the present work aimed at filling this gap by providing a specific, sensitive, high-throughput and cost-effective real-time PCR method that, besides establishing the unequivocal identification of *G. biloba* in herbal products, enables its quantification in plant mixtures. For this purpose, a normalized quantitative method was proposed, which was further validated and applied to assess the authenticity of ginkgo-containing commercial herbal infusions and to verify their labelling compliance.
### **2. Materials and Methods**
#### *2.1. Plant Species and Commercial Samples*
Leaves from *G. biloba* were kindly provided by the Botanical Garden of University of Porto, Botanical Garden of Bern, Serralves Garden and Botanical Garden of Madeira (Table S1, Supplementary Material). Leaves or seeds of 73 plant species corresponding to medicinal plants, fruits and spices were used for cross-reactivity testing (Table S1, Supplementary Material). A total of 20 herbal infusions were bought at local stores, including specialized herbalists, and from the internet (Table 1).
For method development, model mixtures with known amounts of dried leaves of *G. biloba* in *Camellia sinensis* were prepared to contain 50%, 10%, 5%, 1%, 0.5%, 0.1%, 0.05% and 0.01% (*w*/*w*). Firstly, a reference mixture with 50% of *G. biloba* was prepared by adding 10 g of ground *G. biloba* leaves to 10 g of ground plant material of *C. sinensis*. All the subsequent mixtures were prepared by sequential additions of *C. sinensis* plant material up to the level of 0.01% (*w*/*w*). For method validation, blind mixtures were independently prepared as described for the reference mixtures, with the proportion of 20%, 8%, 2%, and 0.2% (*w*/*w*) of *G. biloba* in *C. sinensis* plant material and were further analyzed as unknown samples.
Seeds were ground with a mortar, while the leaves and herbal infusions were ground in a laboratory mill Grindomix GM200 (Retsch, Haan, Germany).
#### *2.2. DNA Extraction*
The NucleoSpin Plant II kit (Macherey-Nagel, Düren, Germany) was chosen to perform the DNA extraction from 50 mg of each sample, according to the manufacturer's instructions with slight modifications, as described by Costa et al. [22]
#### *2.3. DNA Quality and Purity*
The yield and purity of DNA extracts were assessed by UV spectrophotometry using a Take3 micro-volume plate accessory, on a Synergy HT multi-mode microplate reader (BioTek Instruments, Inc., Winooski, VT, USA). The nucleic acid protocol was set for double-strand DNA in the Gen5 data analysis software version 2.01 (BioTek Instruments, Inc., Winooski, VT, USA), which was applied to absorbance data measured at 260 and 280 nm.
The quality of DNA extracts was further assessed by electrophoresis with 1% of agarose gel as previously described [22].
#### *2.4. Target Gene Selection, Oligonucleotide Primers and Probes*
A set of primers (Gkb2-F/Gkb2-R) and a specific probe (Gkb2-P) labelled with fluorescein (FAM) as a fluorescent reporter and black hole quencher 1 BHQ-1 as quencher, were designed to target the Internal Transcribed Space (ITS) region of *G. biloba* (GenBank: Y16892.1) (Table 2). In silico analysis of sequences and primers was performed using the BLAST and Primer-BLAST tools to verify fragment and primer specificity, respectively. OligoCalc software was used to check primer properties and ensure the absence of primer hairpins and self-hybridization.
(+) Positive amplification; (−) Negative amplification; +/− Doubtful amplification. b Mean cycle of quantification (Cq) values ± standard deviation (SD) (*n* = 4). c Mean percentage (%) values ± SD (*n* = 4). d NA—not applicable. e NL—not labelled.
**Table 2.** Data of primers used, targeting the ITS1 region of Ginkgo biloba and a conserved eukaryotic region.
To ensure the presence of amplifiable DNA, a universal eukaryotic primer pair (EG-F/EG-R), targeting a conserved 18S rRNA nuclear region, was used [23]. The same primer pair together with a probe (EG-P) was used as an endogenous control gene for developing the normalized real-time PCR system [24]. The primers and probes were synthesized by Eurofins MWG Operon (Ebersberg, Germany).
#### *2.5. Qualitative PCR*
PCR amplification was performed using a total reaction volume of 25 µL which contained 20 ng of DNA, buffer (67 mM Tris-HCl, pH 8.8, 16 mM (NH4)2SO4, 0.01% Tween 20), 3 mM of MgCl2, 1.0 U of SuperHot Taq DNA Polymerase (Genaxxon Bioscience GmbH, Ulm, Germany), 280 nM of each primer and 200 µM of dNTP (Grisp, Porto, Portugal) (Table 2). The reactions were carried out in a MJ Mini™ Gradient Thermal Cycler (Bio-Rad Laboratories, Hercules, CA, USA), using the following optimized programs: initial denaturation at 95 ◦C for 5 min; 35 or 40 cycles (for EG-F/EG-R or Gkb2-F/Gkb2-R primers, respectively) of amplification at 95 ◦C for 30 s, 63 ◦C or 62 ◦C (for EG-F/EG-R or Gkb2-F/Gkb2-R primers, respectively) for 30 s and extension at 72 ◦C for 30 s; and a final extension at 72 ◦C for 5 min.
PCR products were further analyzed by electrophoresis in a 1.5% agarose gel stained with 1× Gel Red (Biotium, Hayward, CA, USA) and running in 1× SGTB buffer (GRISP, Porto, Portugal) for 20–25 min at 200 V. Each extract was amplified in at least two independent assays.
#### *2.6. Real-Time PCR*
The reactions were performed using 20 µL of total reaction volume, containing 2 µL of DNA (20 ng), 1× SsoFast Probes Supermix (Bio-Rad Laboratories, Hercules, CA, USA), 300 nM or 400 nM of each primer set (EG-F/EG-R or Gkb2-F/Gkb2-R, respectively) and 200 nM of each probe (EG-P or Gkb2-P, for eukaryotic and *G. biloba* genes, respectively). A fluorometric thermal cycler CFX96 Real-time PCR Detection System (Bio-Rad Laboratories, Hercules, CA, USA) was used to amplify, simultaneously and in parallel reactions, each target sequence, under the following conditions: 95 ◦C for 5 min, 45 cycles at 95 ◦C for 15 s and 65 ◦C for 45 s, and the fluorescence signal was collected at the end of each cycle. The data evaluation, from each real-time PCR assay, was made using the software Bio-Rad CFX Manager 3.1 (Bio-Rad Laboratories, Hercules, CA, USA). Real-time PCR assays were performed, at least, in two independent runs using *n* = 3 or *n* = 4 replicates in each one.
For the construction of a calibration curve and for the determination of the absolute limits of detection (LOD) and quantification (LOQ), 10-fold serially diluted ginkgo DNA extracts (20 ng–0.002 pg) were amplified by real-time PCR. Additionally, a normalized calibration model was constructed based on the parallel amplification of the ITS1 region of *G. biloba* (target sequence) and the 18S rRNA gene (reference for eukaryotes) using the model mixtures (0.01–50%) of *G. biloba* in *C. sinensis*. The acceptance criteria established for real-time PCR assays were the PCR efficiency between 90–110%, the slope within −3.6 and −3.1 and the correlation coefficient (*R 2* ) above 0.98 [25,26]. The lowest amplified level for 95% of the replicates was considered as the LOD and the LOQ was set as the lowest amplified level within the linear dynamic range of the calibration curve, which should cover a minimum of 4 orders of magnitude and should extend to ideally 5 or 6 log<sup>10</sup> concentrations [25,26].
### **3. Results and Discussion**
#### *3.1. DNA Quality and Selection of Target Region*
In general, DNA extracts from the leaves, seeds and commercial samples showed adequate yields and purities, being in the range of 17.6–270.8 ng/µL and 1.4–2.1, respectively. Before the *G. biloba* specific amplification of target region, all extracts were tested by PCR targeting a universal eukaryotic region (EG-F/EG-R) to check the capacity of DNA amplification and avoid false negatives [23]. All DNA extracts used for reactivity testing were amplified (Table S1, Supplementary Material).
So far, different regions have been assessed, either as a single locus or in combination, for their adequacy as barcode markers in plant species, which include *matk*, *rbcL*, ITS and ITS2, among others [27]. In this work, the non-coding ITS region of nuclear ribosomal DNA was selected due to its high power of species discrimination over plastid regions, allowing the differentiation of closely related species [27–29]. This region has been previously proposed for the development of PCR assays using species-specific primers aiming at identifying medicinal plant species, with high specificity and sensitivity [30,31]. The specificity of the newly designed primers (Gkb2-F/Gkb2-R) was initially in silico verified and subsequently assayed experimentally against different DNA extracts from several plant species (*n* = 73). As expected, the primers proved to be specific since only the DNA extracts from *G. biloba* were amplified (Table S1, Supplementary Material). Afterwards, the optimized species-specific PCR assay, using a 10-fold serially diluted *G. biloba* DNA extract (20 ng), showed a sensitivity down to 0.002 ng (Figure S1, Supplementary Material) and was further applied in the analysis of the commercial samples (Table 1). The achieved sensitivity was much higher than that obtained by the RPA-lateral flow strip device reported by Liu et al., which was approximately 1 ng of purified DNA. Moreover, only a few plant species were used for cross-reactivity testing (*Crataegus pinnatifida*, *Epimedium brevicornu*, *Selaginella tamariscina* and *Arisaema heterophyllum*) by those authors. In the same work, a species-specific PCR assay targeting *G. biloba* DNA was also developed, but again with very limited specificity testing (only against *S. japonica*).
#### *3.2. Quantitative Real-Time PCR*
#### 3.2.1. Method Development
Following the demonstrated suitability of the proposed primers for *G. biloba* specific detection, a real-time PCR method was developed using a newly designed hydrolysis probe (Gkb2-P), increasing the sensitivity and specificity of the assay. Figure 1 presents the real-time PCR amplification curves and respective calibration curve using a 10-fold serially diluted ginkgo DNA extract. The average parameters of PCR efficiency (101.4%), slope (−3.284) and *R* 2 (0.988) were all within the acceptance criteria (Figure 1B), suggesting a high performance of the assay [25,26]. The dynamic range covered six orders of magnitude of the target analyte (20 ng to 0.02 pg of ginkgo DNA) and the absolute LOD of the real-time PCR assay was established as 0.02 pg of *G. biloba* DNA, corresponding to 0.285 genomic DNA copies (using the mean value of the Plant DNA C-value database [32]) and considering the amplification of all replicates (*n* = 6 from two independent assays). Since the LOD value was within the linear dynamic range of the calibration curve, the LOQ value was set at the same value (0.02 pg) [25,26].
**Figure 1.** Amplification curves (**A**) and respective calibration curve (**B**) of a real-time PCR assay with a hydrolysis probe targeting ITS1 region of *G. biloba*. The amplified extracts correspond to 10-fold serially diluted ginkgo DNA from 20 ng to 0.002 pg (*n* = 3 replicates). Cq (cycle of quantification, also known as Ct, cycle threshold). **Figure 1.** Amplification curves (**A**) and respective calibration curve (**B**) of a real-time PCR assay with a hydrolysis probe targeting ITS1 region of *G. biloba*. The amplified extracts correspond to 10-fold serially diluted ginkgo DNA from 20 ng to 0.002 pg (*n* = 3 replicates). Cq (cycle of quantification, also known as Ct, cycle threshold).
For establishing a quantitative model of ginkgo in herbal material, a normalized real-time PCR assay using the ∆Ct method was developed. This approach accounts with amplification variations due to inconsistent DNA recovery and quality/degradation among extracts as a result of processing [24,33–35]. It relies on the construction of a normalized calibration curve using the cycle of quantitation (Cq) values from the target region (ITS1) and a reference endogenous gene (nuclear 18S rRNA) by applying the expression ΔCq = Cq (ginkgo)—Cq (universal gene). The normalized calibration curve was obtained by plotting the calculated ΔCq values versus the logarithm of the For establishing a quantitative model of ginkgo in herbal material, a normalized real-time PCR assay using the ∆Ct method was developed. This approach accounts with amplification variations due to inconsistent DNA recovery and quality/degradation among extracts as a result of processing [24,33–35]. It relies on the construction of a normalized calibration curve using the cycle of quantitation (Cq) values from the target region (ITS1) and a reference endogenous gene (nuclear 18S rRNA) by applying the expression ∆Cq = Cq (ginkgo)−Cq (universal gene). The normalized calibration curve was obtained by plotting the calculated ∆Cq values versus the logarithm of the gingko concentration, using the
gingko concentration, using the binary mixtures with known quantities of *G. biloba* in *C. sinensis*
binary mixtures with known quantities of *G. biloba* in *C. sinensis* (50.0%, 10.0%, 5.0%, 1.0%, 0.5%, 0.1%, 0.05%, and 0.01%, *w*/*w*) (Figure 2). The choice of *C. sinensis*, also commonly known as the "tea plant", to prepare the reference mixtures was based on the high frequency of its use in mixed herbal infusions. The developed normalized real-time PCR approach exhibited high performance, as inferred from the obtained parameters of PCR efficiency (96.2%), *R 2* (0.982) and slope (−3.417) (mean values from 6 independent assays), covering 7 magnitude orders, which were all within the acceptable criteria. The approach enabled an LOD and LOQ down to 0.01% (*w*/*w*) (*n* = 12 from 3 independent assays), corresponding to 0.1 g of *G. biloba* per 1 kg of *C. sinensis*. (50.0%, 10.0%, 5.0%, 1.0%, 0.5%, 0.1%, 0.05%, and 0.01%, *w*/*w*) (Figure 2). The choice of *C. sinensis*, also commonly known as the "tea plant", to prepare the reference mixtures was based on the high frequency of its use in mixed herbal infusions. The developed normalized real-time PCR approach exhibited high performance, as inferred from the obtained parameters of PCR efficiency (96.2%), *R2* (0.982) and slope (−3.417) (mean values from 6 independent assays), covering 7 magnitude orders, which were all within the acceptable criteria. The approach enabled an LOD and LOQ down to 0.01% (*w*/*w*) (*n* = 12 from 3 independent assays), corresponding to 0.1 g of *G. biloba* per 1 kg of *C. sinensis*.
*Foods* **2020**, *9*, x FOR PEER REVIEW 8 of 12
**Figure 2.** Normalized calibration curves obtained by real-time PCR, targeting the ITS1 region of ginkgo, using the binary mixtures of *G. biloba* in *C. sinensis* (50%, 10%, 5%, 1%, 0.5%, 0.1%, 0.05% and 0.01% (*w*/*w*)). The normalized ΔCq method was performed by the parallel amplification of a **Figure 2.** Normalized calibration curves obtained by real-time PCR, targeting the ITS1 region of ginkgo, using the binary mixtures of *G. biloba* in *C. sinensis* (50%, 10%, 5%, 1%, 0.5%, 0.1%, 0.05% and 0.01% (*w*/*w*)). The normalized ∆Cq method was performed by the parallel amplification of a eukaryotic sequence (18S rRNA) as reference (mean values of six independent assays with *n* = 3 replicates).
eukaryotic sequence (18S rRNA) as reference (mean values of six independent assays with *n* = 3
replicates). Compared with the recent report of Dhivya et al. [21], describing a species-specific real-time PCR with a hydrolysis probe targeting the *matk* gene, the present approach achieved similar performance parameters in terms of PCR efficiency and *R*<sup>2</sup> using serially diluted leaf DNA of ginkgo. However, the proposed real-time PCR method provides a much wider dynamic range (seven orders of magnitude) and a higher sensitivity (0.02 pg of ginkgo DNA) than that obtained by Dhivya et al. [21] (five orders of magnitude and 10 pg of ginkgo DNA). Regarding specificity, the proposed primers and probe targeting the ITS region do not provide any cross-reactivity with any of the known potential adulterants (*Sophora japonica* and *Fagopyrum esculentum* Moench) (Figure S2), while the method of Dhivya et al. [21] was reactive with *S. japonica* at late amplification cycles, which compromised its sensitivity, and the potential reactivity with *F. esculentum* Moench was not verified by the referred authors. Therefore, the proposed method demonstrated full specificity and high Compared with the recent report of Dhivya et al. [21], describing a species-specific real-time PCR with a hydrolysis probe targeting the *matk* gene, the present approach achieved similar performance parameters in terms of PCR efficiency and *R* <sup>2</sup> using serially diluted leaf DNA of ginkgo. However, the proposed real-time PCR method provides a much wider dynamic range (seven orders of magnitude) and a higher sensitivity (0.02 pg of ginkgo DNA) than that obtained by Dhivya et al. [21] (five orders of magnitude and 10 pg of ginkgo DNA). Regarding specificity, the proposed primers and probe targeting the ITS region do not provide any cross-reactivity with any of the known potential adulterants (*Sophora japonica* and *Fagopyrum esculentum* Moench) (Figure S2), while the method of Dhivya et al. [21] was reactive with *S. japonica* at late amplification cycles, which compromised its sensitivity, and the potential reactivity with *F. esculentum* Moench was not verified by the referred authors. Therefore, the proposed method demonstrated full specificity and high sensitivity for gingko detection, with the important achievement of providing, for the first time, a normalized quantitative real-time PCR approach to enable a determination of the proportion of ginkgo in herbal products.
#### sensitivity for gingko detection, with the important achievement of providing, for the first time, a normalized quantitative real-time PCR approach to enable a determination of the proportion of 3.2.2. Method Validation
ginkgo in herbal products. 3.2.2. Method Validation To proceed with the validation of the method, the precision and accuracy should also be evaluated [25,26]. Therefore, blind mixtures containing 20.0%, 8.0%, 2.0%, and 0.2% (*w*/*w*) of *G. biloba* in *C. sinensis* were used. The results regarding the estimated values (%) of ginkgo and the comparative To proceed with the validation of the method, the precision and accuracy should also be evaluated [25,26]. Therefore, blind mixtures containing 20.0%, 8.0%, 2.0%, and 0.2% (*w*/*w*) of *G. biloba* in *C. sinensis* were used. The results regarding the estimated values (%) of ginkgo and the comparative analysis with the real values are presented in Table 3. The obtained values exhibited adequate coefficients of variation (CV), which were between 5.6–17.9% and, therefore, lower than the maximum acceptable (25%), demonstrating the high precision of the method over the considered dynamic range.
analysis with the real values are presented in Table 3. The obtained values exhibited adequate coefficients of variation (CV), which were between 5.6–17.9% and, therefore, lower than the Regarding the accuracy, three out of the four blind mixtures presented bias values in the range of 5.6–17.9%, being within the recommended range (±25%) [26]. Although the mixture with 0.2% (*w*/*w*) presented a slightly higher error (−27.4%), this is the lowest tested level, not likely to occur due to adulteration, but rather from contamination. Besides, according to Kang [36], bias within 25–30% have been considered as acceptable in real-time PCR methods for food analysis.
**Table 3.** Results of the validation assays using the normalized quantitative PCR system applied to blind mixtures of *G. biloba* in *C. sinensis.*
<sup>a</sup> Mean values <sup>±</sup> standard deviation (SD) (*<sup>n</sup>* <sup>=</sup> 4) of three independent assays. <sup>b</sup> Coefficient of variation (CV). <sup>c</sup> Error <sup>=</sup> ((mean estimated value—real value)/real value) <sup>×</sup> 100.
#### 3.2.3. Analysis of Commercial Herbal Infusions
For assessing the applicability of the method, the normalized real-time PCR system was used to analyze and further verify the authenticity and labelling compliance of several commercial herbal products (herbal infusions). The analyzed herbal infusions were all labelled as containing ginkgo, wholly or partially (Table 1). All the samples produced amplifiable DNA extracts, which were positive for the ginkgo-specific PCR assay. The samples of mixed herbal species were further assayed by quantitative real-time PCR to assess their ginkgo content. The quantitative results demonstrated that, out of five samples of herbal mixtures with labelled ginkgo contents, four samples (#4, #10, #11 and #16) declared 15% of ginkgo, but the obtained contents were within 0.01–2.98%. In particular, sample #10 had only trace amounts (0.01%) of gingko, suggesting its complete substitution with other plant(s). Sample #12 declared 30% of ginkgo, but the obtained content was 9.95%. Consequently, the results of samples #4, #11, #12, and #16 suggest the partial substitution of ginkgo with other plant(s). The other two mixed herbal samples (#14, #15) did not provide any quantitative information regarding gingko, having low estimated amounts (<3%), suggesting again its reduced use. Therefore, the results of mixed herbal products strongly suggest the practice of adulterations, probably due to the high market price of *G. biloba* and its increasing demand, with the industries using less quantity than they declared to raise their profits.
### **4. Conclusions**
In the herein presented work, a new molecular marker of the ITS region was identified for the species-specific detection of *G. biloba* by both qualitative PCR and real-time PCR with a TaqMan probe, providing high specificity and sensitivity, down to 0.02 pg of DNA (0.285 genomic DNA copies). For the effective quantification of ginkgo in herbal products, a novel normalized real-time PCR system based on the ∆Cq method was successfully developed using reference herbal mixtures. The method exhibited high performance parameters, namely PCR efficiency, coefficient of correlation and covered dynamic range (50–0.01%), achieving a LOD and LOQ of 0.01% (*w*/*w*) of ginkgo in tea plant. The quantitative approach was further validated with blind mixtures, demonstrating accuracy, repeatability, and trueness within the range of 20–2%. The applicability of the PCR approaches was demonstrated using a set of commercial ginkgo-containing herbal infusions (*n* = 20), confirming the presence of ginkgo in all the products. However, the obtained quantitative results regarding the estimated ginkgo content of seven herbal mixture samples suggest adulterations due to reduction or almost elimination of ginkgo. The proposed system was demonstrated to be a powerful and robust tool for control laboratories and regulatory authorities to ensure labelling compliance of ginkgo-containing
herbal products. Since it was demonstrated that the developed method has a high specificity and sensitivity, it can potentially be useful for further detecting *G. biloba* in other processed herbal products or foods.
**Supplementary Materials:** The following are available online at http://www.mdpi.com/2304-8158/9/9/1233/s1, Table S1: Results of cross-reactivity testing of ITS1 primers are presented. Figure S1: Sensitivity by qualitative PCR. Figure S2: Analytical specificity by real-time PCR.
**Author Contributions:** Conceptualization, I.M. and J.S.A.; methodology, L.G. and J.C.; validation, L.G. and J.C.; formal analysis, L.G. and J.C.; investigation, I.M. and J.S.A.; writing—original draft preparation, L.G.; writing—review and editing, I.M. and J.S.A.; supervision, I.M. and J.S.A.; project administration, I.M.; funding acquisition, I.M. All authors have read and agreed to the published version of the manuscript.
**Funding:** This work was supported by FCT (Fundação para a Ciência e Tecnologia) under the Partnership Agreements UIDB 50006/2020 and UIDB 00690/2020. L. Grazina is grateful to FCT grant (SFRH/BD/132462/2017) financed by POPH-QREN (subsidised by FSE and MCTES).
**Acknowledgments:** The authors are grateful for the supply of leaves from the Botanical Garden of University of Porto (Porto, Portugal), Botanical Garden of Bern (Bern, Switzerland), Serralves Garden (Porto, Portugal), Botanical Garden of Madeira and Botanical garden of UTAD (Vila Real, Portugal), as well as to the voucher seeds from the USDA Grin by the University of Arizona Herbarium (Tucson, AZ, USA) and the RBG (Kew, Ardingly, West Sussex, UK).
**Conflicts of Interest:** The authors declare no conflict of interest.
### **References**
© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
## **A Chip Digital PCR Assay for Quantification of Common Wheat Contamination in Pasta Production Chain**
**Caterina Morcia 1,**† **, Ra**ff**aella Bergami 2,**† **, Sonia Scaramagli <sup>2</sup> , Roberta Ghizzoni <sup>1</sup> , Paola Carnevali <sup>3</sup> and Valeria Terzi 1,\***
Received: 27 May 2020; Accepted: 7 July 2020; Published: 10 July 2020
**Abstract:** Pasta, the Italian product par excellence, is made of pure durum wheat. The use of *Triticum durum* derived semolina is in fact mandatory for Italian pasta, in which *Triticum aestivum* species is considered a contamination that must not exceed the 3% maximum level. Over the last 50 years, various electrophoretic, chemical, and immuno-chemical methods have been proposed aimed to track the possible presence of common wheat in semolina and pasta. More recently, a new generation of methods, based on DNA (DeoxyriboNucleic Acid) analysis, has been developed to this aim. Species traceability can be now enforced by a new technology, namely digital Polymerase Chain Reaction (dPCR) which quantify the number of target sequence present in a sample, using limiting dilutions, PCR, and Poisson statistics. In our work we have developed a duplex chip digital PCR (cdPCR) assay able to quantify common wheat presence along pasta production chain, from raw materials to final products. The assay was verified on reference samples at known level of common wheat contamination and applied to commercial pastas sampled in the Italian market.
**Keywords:** pasta; *Triticum aestivum*; *Triticum durum*; genetic traceability; digital PCR; semolina; species
#### **1. Introduction**
Pasta production is a strategic chain in the Italian agri-food sector, covering around the 6% of total industrial output [1]. Italy is at the same time the world's leading pasta producer, with an annual production around 3.2 million tons and, in the same time, is the largest consumer of pasta (26 kg per capita). A pillar of Italian pasta production chain is the grain identity: The use of *Triticum durum* derived semolina is in fact mandatory for Italian pasta, in which *Triticum aestivum* species is considered a contamination that must not exceed the 3% maximum level, as indicated by Law n.580 of 1967 [2] and by subsequent Decreto del Presidente della Repubblica (D.P.R.) 187, 9 February 2001 [3] and D.P.R. 41, 5 March 2013 [4]. Traditional Italian pasta, according to such regulations, is therefore the result of the extrusion, rolling and drying of dough made exclusively from durum wheat and water. The choice of *Triticum durum* is based on its peculiarities, among others the hardiness of the caryopsis, the intense yellow color due to carotenoids, the gluten composition. Thanks to such specific properties, starch is not lost during cooking, avoiding sticking and ensuring a unique and authentic taste to pasta.
Beyond fraudulent behavior, dictated by the lower price of common wheat compared to durum, the purity of the semolina can also be compromised during the various processing stages of the supply chain, which range from harvesting in the field to storing the grains. Analytical methods have been proposed aiming at the detection and quantification of the possible presence of common wheat in semolina and pasta. In this perspective, over the last 50 years, various electrophoretic, chemical and immuno-chemical methods have been proposed aimed at detecting the purity of the semolina [5–9]. Such methods are based on the identification and quantification of specific protein, which, however, can be degraded by the high temperatures nowadays used to dry pasta. To overcome this gap and taking advantage of the remarkable thermic stability of DNA (DeoxyriboNucleic Acid), a new generation of methods, based on DNA analysis, has been developed during the last two decades. PCR (Polymerase Chain Reaction) based assays to identify common wheat by distinguishing it from durum one has been developed by Bryan et al. [10], by Arlorio et al. [11] and by Sonnante et al. [12], using respectively *Dgas44* gene sequence, puroindoline B and SSR (Simple Sequence Repeats) related sequences. Untargeted DNA fingerprinting through tubulin-based polymorphism (TBP) have been optimized by Casazza et al. [13] and by Silletti et al. [14] for the authentication of cereal species, including wheat and farro. qPCR assays for the quantification of *Triticum aestivum* species have been proposed by Alary et al. [15], Terzi et al. [16], Matsuoka et al. [17], and by Imai et al. [18]. These two last assays have been in-house verified and compared by Paterno' et al. [19], with the aim to select a taxon-specific assay useful for unauthorized GM (Genetically Modified) wheat detection in wheat samples. An inter-laboratory validation in collaboration with public and private laboratories has been even reported by Morcia et al. [20] to determine the performance parameters of a qPCR assay based on the primers designed on puroindoline-b gene by Alary et al. [15] and on low molecular weight glutenin encoding sequence by Terzi et al. [16].
Species traceability can be now enforced by a new technology, namely digital PCR (dPCR) which quantify the number of target sequence present in a sample, using limiting dilutions, PCR and Poisson statistics [21]. The PCR mix is compartmentalized across a large number of partitions or droplets containing zero, one or more copies of the target sequence. After endpoint PCR amplification, a partition can be positive ("10 ', the presence of PCR product) or negative ("00 ', the absence of PCR product). The absolute number of target nucleic acid molecules contained in the original sample before partitioning can be calculated directly from the ratio of the number of positive to total partitions, obtained using Poisson statistics. It is an absolute quantification strategy because there is not the need to have a standard curve as reference for quantification. In the past several years, dPCR has achieved progress in in agri-food sector, especially for GMO (Genetically Modified Organism) testing [21,22] and for pathogen diagnostics and, at more limited extent, to the detection of animal- and plant-derived ingredients in food adulteration control [23].
The aim of this work has been to develop a chip digital PCR (cdPCR) assay able to quantify common wheat presence along pasta production chain, from raw materials to final products. The assay was verified on reference samples at known level of common wheat contamination and applied to commercial pastas sampled in the Italian market.
#### **2. Materials and Methods**
#### *2.1. Mono-Species Flour Samples Preparation and DNA Extraction*
Certified *Triticum durum* (Claudio variety) and *Triticum aestivum* (Eureka variety) seeds were obtained from CREA DC (Tavazzano, Italy). Such first-reproduction seeds are controlled and certified both at species and variety levels. In major details, at species purity level, the maximum admitted contamination is of 7 seeds belonging to different cereal species/500 g of certified seeds, according to the Italian D.P.R. n. 1065, 8 October 1973. The seeds were milled using a Cyclotec (FOSS Italia S.r.l., Padova, Italy) at 0.2 mm grid diameter, avoiding any contamination between samples. Samples of
100% durum wheat semolina and 100% common wheat flour were separately stored at controlled temperature and humidity conditions until further use.
DNA were extracted from three biological replicates of milled *Triticum aestivum* and *Triticum durum* seeds using the DNeasy mericon Food Kit (Qiagen, Milan, Italy), that is based on an improved cetyltrimethylammonium bromide (CTAB) extraction of total cellular nucleic acids. The flour samples (2 g) were extracted according to manifacturer's instructions. The evaluation of quality and quantity of extracted DNA was done using Qubit™ fluorometer in combination with the Qubit™ dsDNA BR Assay kit (Invitrogen by Thermo Fisher Scientific, Monza, Italy).
#### *2.2. Mixed Species DNA Samples Preparation*
*Triticum aestivum* and *Triticum durum* DNA, extracted from the mono-species flours described in point 2.1, were mixed to obtain the following samples:
#### *2.3. Mixed Species Flour Samples Preparation and DNA Extraction*
Common wheat flour was used to contaminate durum wheat semolina with the aim to produce durum wheat samples containing 0.3, 1.5, 3, 4.5, and 30% of common wheat. After weighing the common and durum wheat flour, samples containing different percentages of the two species were homogenized for 10 min. DNA were extracted from flours (2 g) with the DNeasy mericon Food Kit (Qiagen, Milan, Italy), as previously described. The evaluation of quality and quantity of extracted DNA was done using Qubit™ fluorometer in combination with the Qubit™ dsDNA BR Assay kit (Invitrogen by Thermo Fisher Scientific, Monza, Italy).
#### *2.4. Reference and Commercial Pasta Samples and DNA Extraction*
Four reference pasta samples were prepared by mixing tap water and wheat flours containing the following common wheat percentages: 1.5%, 3%, 4.5%, 10%. The samples were dried in oven at 80 ◦C for 1 hour, followed by 3 hours at decreasing temperature. Such desiccation thermal profile is those commonly used for commercial pasta preparation. DNA were extracted from two biological replicates of reference pasta using the DNeasy mericon Food Kit (Qiagen, Milan, Italy), Twenty commercial pasta samples of different brands were purchased from the market. The pasta samples were milled with M20 Universal Mill (IKA). Samples (2 g) were extracted in single replicate with the DNeasy mericon Food Kit (Qiagen, Milan, Italy), as previously described. The DNA obtained was measured using Qubit™ fluorometer in combination with the Qubit™ dsDNA BR Assay kit (Invitrogen by Thermo Fisher Scientific, Monza, Italy).
#### *2.5. Primers and Probes*
Primers and probes (Table 1) were designed using Primer Express 3.0.1 Software (Life Technologies Corporation). Each primer was checked for absence of self-complementarity and primer dimer formation with other primer pairs using the online tool Multiple Primer Analyzer (Thermo Fisher Scientific, Monza, Italy). Primer specificity was checked by blasting in EnsemblPlants (https://plants. ensembl.org/index.html) against the *Triticum aestivum* database.
**Table 1.** Primers and probes.
#### *2.6. Real-Time PCR*
The reaction mixture was prepared in a final volume of 25 µL consisting of 12.5 µL of SYBR Green PCR, 2× GoTaq qPCR Master Mix (Promega Italia, Milan, Italy), 0.25 µl of 100× Reference Dye (Promega Italia, Milan, Italy), 0.5 µL of each primer at 10 µM (final concentration 200 nmol), 4 µL of DNA template serial dilution (10, 5, 2.5, 0.5, 0.25 and 0.025 ng/µL) and water to 25 µL. Three technical real-time PCR replicates were done for each sample and control. The PCR mixture was activated at 95 ◦C for 10 min. Forty amplification cycles were carried out at 95 ◦C for 15 s followed by 60 ◦C for 1 min. A melting curve analysis was included in each run.
#### *2.7. Chip Digital PCR*
Chip digital PCR was performed using QuantStudioTM 3D Digital PCR System (Applied Biosystems by Life Technologies, Monza. Italy). The reaction mixture was prepared in a final volume of 16 µLconsisting of 8 µL QuantStudioTM 3D Digital PCR 2X Master Mix, 0.72 µL of each primer at 20 µM (final concentration 900 nmol), 0.32 µL of FAM and VIC-MGB probes at 10 µM (final concentration 200 nmol), 2 µl of DNA (40 ng/µL) and nuclease free-water. Also, a negative control with nuclease free-water as template was added. A total volume of 15 µL reaction mixture was loaded onto the QuantStudioTM 3D Digital PCR chips using QuantStudioTM 3D Digital chip loader, according to manufacturer protocol. Amplifications were performed in ProFlexTM 2Xflat PCR System Thermocycler (Applied Biosystems by Life Technologies, Monza, Italy) under the following conditions: 96 ◦C for 10 min, 45 cycle of 55 ◦C annealing for 2 min and 98 ◦C denaturation for 30 s, followed by 60 ◦C for 2 min and 10 ◦C. End-point fluorescence data were collected in QuantStudioTM 3D Digital PCR Instrument and files generated were analyzed using cloud-based platform QuantStudioTM 3D AnalysisSuite dPCR software, version 3.1.6. Each sample was analyzed in triplicate.
#### *2.8. Triticum aestivum Percentage Calculation*
For the common wheat percentage calculation, we start from the absolute copies/µL yielded by the QuantStudioTM 3D Analysis Suite dPCR software. In our assay the *T. aestivum* target sequence is marked with FAM, whereas the taxon target sequence is marked in VIC. Equation 1 was used to calculate the percentage of common wheat copies in the sample, in which FAM stands for the number of FAM copies/µL and VIC for the number of VIC copies/µL:
$$\frac{FAM}{\frac{VIC - 3 \star FAM}{2} + FAM} \ast 100\tag{1}$$
#### **3. Results**
#### *3.1. Reference Samples*
Several factors are important for accurate quantification of multiplexed assays, including target linkage, probe specificity and differential PCR efficiencies.
3. Results
The absence of linking between the two targets has been evaluated through literature and bioinformatic analysis. Nemoto et al. [24] demonstrated, through Southern blot analysis, that the *Triticum TaHd1* gene is present in single copy on each A, B and D genomes of wheat and maps on long arm of chromosome 6. *Pinb-D1*gene maps in D sub-genome and is located on chromosome 5 at the Hardness (Ha) locus. The two targets are therefore not linked. linkage, probe specificity and differential PCR efficiencies. The absence of linking between the two targets has been evaluated through literature and bioinformatic analysis. Nemoto et al. [24] demonstrated, through Southern blot analysis, that the Triticum TaHd1 gene is present in single copy on each A, B and D genomes of wheat and maps on long arm of chromosome 6. Pinb-D1gene maps in D sub-genome and is located on chromosome 5 at the Hardness (Ha) locus. The two targets are therefore not linked.
Foods 2020, 9, x 5 of 11
Several factors are important for accurate quantification of multiplexed assays, including target
Primers/probes specificity have been preliminarily evaluated in qPCR, finding that TritA\_APX assay gives a signal only in hexaploid wheat, whereas GranoCO2 assay gives a signal both in hexaploid and tetraploid wheats (including farro dicoccum and Kamut). Primers/probes specificity have been preliminarily evaluated in qPCR, finding that TritA\_APX assay gives a signal only in hexaploid wheat, whereas GranoCO2 assay gives a signal both in hexaploid and tetraploid wheats (including farro dicoccum and Kamut).
Amplification efficiency and reproducibility for each primer set were examined through a standard curve qPCR assay, using bread and durum wheat DNA dilutions (Figure 1). Efficiency of reactions were calculated from the slope using the formula E = 10−1/slope. The slope values obtained were of −3.44 for GranoCO2 primers, and of −3.17 was obtained for TritAPX primers. Amplification efficiencies were of 99.6 and 104%, respectively. Amplification efficiency and reproducibility for each primer set were examined through a standard curve qPCR assay, using bread and durum wheat DNA dilutions (Figure 1). Efficiency of reactions were calculated from the slope using the formula E = 10−1/slope. The slope values obtained were of −3.44 for GranoCO2 primers, and of −3.17 was obtained for TritAPX primers. Amplification efficiencies were of 99.6 and 104%, respectively.
Figure 1. qPCR standard curves obtained after amplification of the DNA dilutions reported in the graph with GranoCO2 primers (A) and with TriAPX primers (B). **Figure 1.** qPCR standard curves obtained after amplification of the DNA dilutions reported in the graph with GranoCO2 primers (**A**) and with TriAPX primers (**B**).
The duplex method was then optimized in cdPCR system for specificity on the reference samples described in Materials and Methods. The concentrations of primers and probes were optimized at 900 nmol and 200 nmol respectively and the annealing temperature was fixed at 55 °C. The resolution of the clusters (Figure 2) was obtained in absence of restriction digestion of the samples, therefore this time-consuming procedure was omitted from the protocol. The duplex method was then optimized in cdPCR system for specificity on the reference samples described in Materials and Methods. The concentrations of primers and probes were optimized at 900 nmol and 200 nmol respectively and the annealing temperature was fixed at 55 ◦C. The resolution of the clusters (Figure 2) was obtained in absence of restriction digestion of the samples, therefore this time-consuming procedure was omitted from the protocol.
The mean common wheat percentages experimentally determined in "mixed flour" and "mixed DNA" samples in comparison with actual percentages are reported in Table 2. The SD values reported in the same table express the precision of the method, i.e., the closeness of agreement between replicate measurements. At 3% level, the SD values are <35% for all the samples and therefore the precision is acceptable, according to Codex Alimentarius Commission/Guidelines 74–2010 [25]. In Table 2 are even reported some values informative about the precision and the accuracy of the method, such as the coefficient of variation (CV), the absolute error and the relative error.
The trueness of the method is usually defined as the degree of agreement of the expected value with the true value or accepted reference value. In GMO testing the trueness must be within 25% of the accepted reference value [25]. The trueness of our method fits the purpose: The estimated concentrations over the dynamic range tested were within the ± 25% acceptable bias as recommended by GMO analytical guidelines [26]. In particular, at 3% level the experimentally determined percentages are very close to the true one. In the evaluated dynamic range, the LOD (Limit of Detection) of
the method has been found at 0.3% common wheat contamination, whereas the LOQ (Limit of Quantification) at 1.5% level. Foods 2020, 9, x 6 of 11
Figure 2. Two-dimensional scatter graphs generated by chip digital PCR (cdPCR) analysis of eight different samples. NTC (No Template Control) is a blank sample without DNA; The other samples are made of durum wheat (dw) DNA or common wheat (bw) DNA or a mix of the two, as indicated in the figure; In this graph a partition can fall into one of four possible clusters: negative partition that contain no amplified targets (yellow), single positive partition for Triticum genus (red), single positive partition for common wheat (blue) and positive partitions that contain a positive signal for both targets (green, double-positive partitions). **Figure 2.** Two-dimensional scatter graphs generated by chip digital PCR (cdPCR) analysis of eight different samples. NTC (No Template Control) is a blank sample without DNA; The other samples are made of durum wheat (dw) DNA or common wheat (bw) DNA or a mix of the two, as indicated in the figure; In this graph a partition can fall into one of four possible clusters: negative partition that contain no amplified targets (yellow), single positive partition for *Triticum* genus (red), single positive partition for common wheat (blue) and positive partitions that contain a positive signal for both targets (green, double-positive partitions).
DNA" samples in comparison with actual percentages are reported in Table 2. The SD values reported in the same table express the precision of the method, i.e., the closeness of agreement between replicate measurements. At 3% level, the SD values are < 35% for all the samples and therefore the precision is acceptable, according to Codex Alimentarius Commission/Guidelines 74– **Table 2.** Actual common wheat percentages in comparison with those experimentally determined in two different classes of samples. "Mixed DNA" samples were obtained by mixing DNA extracted from pure common and durum wheat species. "Mixed flour" samples were obtained by extracting DNA from of common and durum wheat flours mixed at different percentages). CV: Coefficient of variation.
The mean common wheat percentages experimentally determined in "mixed flour" and "mixed
100 105.00 7.00 0.07 5.00 0.05 94.40 6.85 0.07 5.60 0.06
0.3 0.43 0.05 0.12 0.13 0.43 0.37 0.12 0.34 0.07 0.23 1.5 1.37 0.07 0.05 0.13 0.09 1.43 0.28 0.19 0.07 0.05 3 3.06 0.05 0.01 0.06 0.02 2.86 0.32 0.11 0.14 0.05 4.5 4.50 0.04 0.01 0.00 0.00 3.93 0.51 0.13 0.57 0.13 30 25.90 0.46 0.02 4.10 0.14 24.90 1.68 0.07 5.10 0.17 The trueness of the method is usually defined as the degree of agreement of the expected value with the true value or accepted reference value. In GMO testing the trueness must be within 25% of the accepted reference value [25]. The trueness of our method fits the purpose: The estimated concentrations over the dynamic range tested were within the ± 25% acceptable bias as recommended The Pearson's r between the expected and calculated common wheat percentages were determined in mixed DNA samples and in mixed flour samples. The correlation values found are respectively of 0.9985 and of 0.9993. Extracting DNA from mixed flours and their subsequent amplification is much more realistic model of real foods, rather than mixing DNA from different species/samples. However, the preparation of mixtures of flours can be potentially affected by weighting errors and by heterogeneity problems, due, for example, to variation in granulometry, in mixing and blending. On the other hand, DNA mixtures can be affected by errors in DNA quantification and mixing. Therefore, with the intent to minimize the inaccuracy of the reference materials we decided to prepare two series
by GMO analytical guidelines [26]. In particular, at 3% level the experimentally determined percentages are very close to the true one. In the evaluated dynamic range, the LOD (Limit of of blends using the two different options. After analyses, the two classes of reference materials gave the same results. No statistically significant differences were found among mean common wheat % values determined from mixed DNA samples and from mixed flours. It is therefore possible to conclude that the two classes of reference materials prepared worked in agreement. the two classes of reference materials gave the same results. No statistically significant differences were found among mean common wheat % values determined from mixed DNA samples and from mixed flours. It is therefore possible to conclude that the two classes of reference materials prepared worked in agreement.
Foods 2020, 9, x 7 of 11
Detection) of the method has been found at 0.3% common wheat contamination, whereas the LOQ
materials we decided to prepare two series of blends using the two different options. After analyses,
The Pearson's r between the expected and calculated common wheat percentages were determined in mixed DNA samples and in mixed flour samples. The correlation values found are respectively of 0.9985 and of 0.9993. Extracting DNA from mixed flours and their subsequent amplification is much more realistic model of real foods, rather than mixing DNA from different species/samples. However, the preparation of mixtures of flours can be potentially affected by weighting errors and by heterogeneity problems, due, for example, to variation in granulometry, in
Since 3% common wheat threshold is in percentage of mass ratio (% *m*/*m*) and since the analytical output is in number of common wheat and taxon target copies, a conversion factor is needed. This conversion factor, CF, mainly depends on the zygosity, but even on differences linked to DNA extraction and varieties. CF for GMO detection is available for each CRM (Certified Reference Material) [21]. Since 3% common wheat threshold is in percentage of mass ratio (% m/m) and since the analytical output is in number of common wheat and taxon target copies, a conversion factor is needed. This conversion factor, CF, mainly depends on the zygosity, but even on differences linked to DNA extraction and varieties. CF for GMO detection is available for each CRM (Certified Reference Material) [21].
For our homozygous samples, for which certified reference materials are not available, a conversion from % (copy/copy) to % (m/m) can be hypothesized. This same approach has been used in the study of Dong et al. [23] aimed to quantify kidney bean in lotus seed paste. For our homozygous samples, for which certified reference materials are not available, a conversion from % (copy/copy) to % (m/m) can be hypothesized. This same approach has been used in the study of Dong et al. [23] aimed to quantify kidney bean in lotus seed paste.
In 3% common wheat reference samples, a mean percent recovery of 100.44 has been obtained, that fully fits with the acceptable range for major components in low complexity matrices (95–105%). In 3% common wheat reference samples, a mean percent recovery of 100.44 has been obtained, that fully fits with the acceptable range for major components in low complexity matrices (95–105%).
#### *3.2. Reference and Commercial Pasta Samples* 3.2. Reference and Commercial Pasta Samples
(Limit of Quantification) at 1.5% level.
The applicability of duplex dPCR assay to pasta was evaluated in two different groups of samples: 4 reference pasta samples prepared in our laboratory and contaminated with 1.5%, 3%, 4.5%, and 10% common wheat and on 20 pasta samples of different brands commercialized in Italy. The applicability of duplex dPCR assay to pasta was evaluated in two different groups of samples: 4 reference pasta samples prepared in our laboratory and contaminated with 1.5%, 3%, 4.5%, and 10% common wheat and on 20 pasta samples of different brands commercialized in Italy.
The results are reported in Figure 3, from which it can be observed that the duplex dPCR assay performs well on reference pasta, with a correlation value of 0.99 among actual and measured percentages and a mean relative error of 0.07. The results are reported in Figure 3, from which it can be observed that the duplex dPCR assay performs well on reference pasta, with a correlation value of 0.99 among actual and measured percentages and a mean relative error of 0.07.
Figure 3. Common wheat percentages determined in 4 reference pasta (A) and in 20 commercial pasta samples (B) with duplex digital PCR (dPCR) assay. In (A) the percentages values before the word "pasta" indicate the common wheat contaminations. In (B) the red horizontal line indicates the maximum level of common wheat contamination allowed by law.) **Figure 3.** Common wheat percentages determined in 4 reference pasta (**A**) and in 20 commercial pasta samples (**B**) with duplex digital PCR (dPCR) assay. In (A) the percentages values before the word "pasta" indicate the common wheat contaminations. In (B) the red horizontal line indicates the maximum level of common wheat contamination allowed by law.)
As previously introduced, a body of Italian laws and regulations rule the product named "pasta" [2–4]. The denomination "pasta" strictly defines a product obtained after drawing, rolling and As previously introduced, a body of Italian laws and regulations rule the product named "pasta" [2–4]. The denomination "pasta" strictly defines a product obtained after drawing, rolling and subsequent drying of a dough exclusively made from durum wheat (flour or semolina or whole semolina) and water. In the final product the humidity must not exceed 12.50%. The production of pasta with common wheat flour is forbidden, but a maximum level of 3% common wheat flour is tolerated as result of accidental contamination during the production chain. The inclusion of ingredients different from durum wheat and water is reserved to "special pasta". The special pastas must be offered for sale in Italy with the name durum wheat semolina pasta supplemented by the mention of the ingredient used and, in the case of several ingredients, of that or the characterizing ones. Anyway, even in special pasta, common wheat is a contaminants. The special pasta represents a minor sector of Italian pasta production and consumption. Therefore, as representative of the market, pasta of different brands has been considered in this study. The analyzed samples were all labelled as "pasta" and all reported, as ingredients, durum wheat and water. According to Italian laws, a maximum 3% common wheat
presence is expected. All the commercial samples have been found below the 3% common wheat contamination threshold. The analytical data confirm that all the samples comply with the Italian laws.
#### **4. Discussion**
We have developed a duplex chip digital PCR analytical protocol to identify and quantify common wheat contamination in pasta production chain. The reason for developing such new assay is related to dPCR particularities. In comparison with conventional end-point PCR and qPCR, this technique has been reported to have many advantages (reviewed by Demeke et al. [27]), the major the absolute quantification of a target without reference to a standard/calibration curve. This fact reduces the errors deriving from the comparison of different matrices, i.e., the calibrant and the test sample. Moreover, because of the high-level sample partitioning, dPCR is less sensitive to PCR inhibitors and the results obtained are potentially very precise and accurate [27,28]. Thanks to the high resilience to inhibitors, the efficiency and the reproducibility on different platforms, dPCR is candidate as higher-order reference measurement methods and as the method for value assignment of reference materials [28]. On the other hand, a limitation of such approach is that it is more expensive than qPCR, but the use of multiplex approaches moves the scales in favor of dPCR [27]. From a technology transfer point of view, both the pasta industry and the large consumer cooperative, between the other involved in this work, expressed interest in developing and applying a dPCR strategy for control pasta chain. The key control points are in the passage of the grains from stackers to the mills, of semolina batches from the mills to the pasta factory and in the final product, the pasta. The pasta chain stakeholders interested in such analytical tool are therefore the farmer associations, the stackers, the mills, the pasta industry, the consumer associations and the public and private control bodies. All the stakeholders have the interest to share a method for common wheat contamination control in grains, semolina, and pasta. Several assays has been developed and validated for such purpose, but are all dependent on a calibration curve and suffer from the loss of certified reference materials for the construction of such curves. DigitalPCR, that works without the need of calibrants, can fill this gap. It can in fact be proposed as method for the validation of reference materials to be used for qPCR standard curves and as higher order reference measurement method. This hypothesis to apply dPCR technology to prepare reference materials has been advanced by other authors, e.g., Mehle et al. [29] in plant pathogen detection, by Dong et al. [30] in environmental microbiology and by Pavšiˇc et al. [31] in microbial diagnostics. The potential for synergy of qPCR and dPCR has been underlined by Debski et al. [32]. in the field of medical diagnostics. In conclusion, the opportunity to complement and strengthen the cheaper qPCR analyses justify the higher cost of dPCR assays.
Our cdPCR assay is based on duplex non-competing reactions: two amplicons are generated from two primer sets and the signal generated from a probe specific for each amplicon enable to distinguish the two targets within a single reaction. Such concurrent amplifications reduce technical errors, reagent and time needed. One of the target is a D-genome specific genic sequence and the other a *Triticum* specific genic sequence present in A, B and D genomes. This taxon-specific assay was designed on *TaHd1* gene sequence. Such gene, involved in the photoperiodic flowering pathway, has been demonstrated to be present in single copy in each of the A, B and D *Triticum* genomes [20]. The bread wheat specific assay was designed on *Pinb-D1*, a single-copy gene encoding for puroindoline b protein [15,33]. This gene belongs to the Ha locus, occurring only on chromosome 5D in common wheat [26]. Accordingly, we have developed the formula reported in Materials and Methods for the common wheat % calculation. In the formula we have considered:
The *Pinb-D1* gene sequence has been used to target common wheat in cqPCR assays previously developed, whereas the *TaHd1* gene sequence has never been used in pasta authenticity assessment.
As verified on reference samples, the proposed protocol highly performs to track 3% common wheat contamination, that is the critical value fixed by law as limit between accidental contamination and fraud. Its applicability has been evaluated on reference and commercial pasta samples. In conclusion, a cdPCR duplex assay has been developed to control pasta production chain from an economically motivated adulteration, that is the use of cheaper ingredient (i.e., common wheat) instead of durum wheat for pasta manufacturing. It is possible to quantify the mass of common wheat directly in flours and in highly processed food, such as pasta. The inter-laboratory validation of the method can be proposed as further step.
**Author Contributions:** Conceptualization, S.S., P.C. and V.T.; Data curation, R.B.; Funding acquisition, V.T.; Methodology, C.M. and R.B.; Validation, R.G.; Writing—original draft, C.M. and V.T.; Writing—review and editing, S.S. All authors have read and agreed to the published version of the manuscript.
**Funding:** This research was partially funded by Horizon 2020 INVITE project, grant number 817970 and by METROFOOD project, Horizon 2020 grant number 739568.
**Conflicts of Interest:** The authors declare no conflict of interest.
### **References**
© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
MDPI St. Alban-Anlage 66 4052 Basel Switzerland Tel. +41 61 683 77 34 Fax +41 61 302 89 18 www.mdpi.com
*Foods* Editorial Office E-mail: [email protected] www.mdpi.com/journal/foods
MDPI St. Alban-Anlage 66 4052 Basel Switzerland
Tel: +41 61 683 77 34
www.mdpi.com
ISBN 978-3-0365-5457-0 | doab | 2025-04-07T03:56:58.474611 | 25-10-2022 11:03 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "00315389-ac15-4185-b49d-fa76bff78a40",
"url": "https://mdpi.com/books/pdfview/book/6149",
"author": "",
"title": "Target and Non-Target Approaches for Food Authenticity and Traceability",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036554587",
"section_idx": 0
} |
00356cdb-56f1-45a1-b97a-312d2e8aacab.0 | # **New Developments in Geometric Function Theory**
Edited by Georgia Irina Oros Printed Edition of the Special Issue Published in *Axioms*
www.mdpi.com/journal/axioms
## **New Developments in Geometric Function Theory**
## **New Developments in Geometric Function Theory**
Editor
**Georgia Irina Oros**
MDPI ' Basel ' Beijing ' Wuhan ' Barcelona ' Belgrade ' Manchester ' Tokyo ' Cluj ' Tianjin
*Editor* Georgia Irina Oros Department of Mathematics and Computer Science University of Oradea Oradea Romania
*Editorial Office* MDPI St. Alban-Anlage 66 4052 Basel, Switzerland
This is a reprint of articles from the Special Issue published online in the open access journal *Axioms* (ISSN 2075-1680) (available at: www.mdpi.com/journal/axioms/special issues/Geometric Function Theory).
For citation purposes, cite each article independently as indicated on the article page online and as indicated below:
LastName, A.A.; LastName, B.B.; LastName, C.C. Article Title. *Journal Name* **Year**, *Volume Number*, Page Range.
**ISBN 978-3-0365-6345-9 (Hbk) ISBN 978-3-0365-6344-2 (PDF)**
© 2023 by the authors. Articles in this book are Open Access and distributed under the Creative Commons Attribution (CC BY) license, which allows users to download, copy and build upon published articles, as long as the author and publisher are properly credited, which ensures maximum dissemination and a wider impact of our publications.
The book as a whole is distributed by MDPI under the terms and conditions of the Creative Commons license CC BY-NC-ND.
## **Contents**
## **About the Editor**
#### **Georgia Irina Oros**
Georgia Irina Oros teaches at University of Oradea, Romania, since 2004. She is Associate Professor at Faculty of Informatics and Sciences, Department of Mathematics and Computer Science since 2013. She obtained her Ph.D. in 2006 in Geometric Function Theory at Babes,-Bolyai University, Cluj-Napoca, Romania under the supervision of Prof. dr. Grigore S, tefan Sal˘ agean. Habilitation ˘ Thesis defended in 2018 at Babes,-Bolyai University, Cluj-Napoca, Romania. She has over 100 papers published in the field of Complex Analysis and Geometric Function Theory.
#### *Editorial* **New Developments in Geometric Function Theory**
**Georgia Irina Oros**
Department of Mathematics and Computer Science, Faculty of Informatics and Sciences, University of Oradea, RO-410087 Oradea, Romania; georgia\_oros\[email protected]
#### **1. Introduction**
This Special Issue aims to highlight the latest developments in the research concerning complex-valued functions from the perspective of geometric function theory. Contributions were sought regarding any aspect of subordination and superordination, different types of operators specific to the research in this field, and special functions involved in univalent function theory with the hope that new approaches would emerge regarding the introduction and study of special classes of univalent functions using operators and the classical theories of differential subordination and superordination, as well as the newer adapted theories of strong differential subordination and superordination and fuzzy differential subordination and superordination. Authors were invited to submit their latest results related to analytic functions in all their variety and also related to their applications in other fields of research. Quantum calculus and its applications related to geometric function theory were also expected to provide interesting outcomes. The presentation of the results obtained by using any other technique that can be applied in the field of complex analysis and its applications was also encouraged.
This Special Issue is devoted especially to complex analysis and was proposed as a means to find new approaches using geometric function theory, to inspire further development in this field.
#### **2. Overview of the Published Papers**
The present Special Issue contains 14 papers accepted for publication after a rigorous reviewing process.
In the study [1], Richard D. Carmichael considers vector-valued analytic functions and distributions with values in Banach or Hilbert space. It is proved that certain vector-valued measurable functions generate the analytic functions using the Fourier–Laplace transform, and conversely, measurable functions are generated from the analytic functions, and it is shown that the analytic functions are representable through the generated measurable functions. Certain specific properties are obtained for the analytic functions and measurable functions, and it is proved that, under specified conditions, the analytic functions considered are in fact vector-valued Hardy functions, which immediately result in Cauchy and Poisson integral representations. The existence of boundary values of the analytic functions on the topological boundary is investigated, and problems to consider in future research are suggested. Notably, the author is convinced that future studies can focus on the integral representation, boundary values, and applications of the functions defined in this paper.
In another study [2], Hatun Özlem Güney, Georgia Irina Oros, and Shigeyoshi Owa provide an application of the well-known Sălăgean differential operator for defining a new operator, through which a new class of functions is defined, which has the classes of starlike and convex functions of order α as special cases. The renowned Jack–Miller– Mocanu lemma is applied for obtaining interesting properties for the newly defined class of functions. The new operator defined in this paper can be used to introduce other specific
**Citation:** Oros, G.I. New
Developments in Geometric Function Theory. *Axioms* **2023**, *12*, 59. https:// doi.org/10.3390/axioms12010059
Received: 28 December 2022 Accepted: 1 January 2023 Published: 4 January 2023
**Copyright:** © 2023 by the author. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).
subclasses of analytic functions, and quantum calculus can be also investigated in future studies.
The research of Gangadharan Murugusundaramoorthy and Teodor Bulboacă presented in reference [3] involves the new subclasses of bi-univalent functions defined in the open-unit disk, which are associated with the Gegenbauer polynomials and satisfy subordination conditions. Coefficient estimates are established for the defined classes, and the remarkable Fekete–Szeg˝o problem is also considered. For particular values of the parameters involved in the definition of the classes, the results obtained in this paper provide new insights into the Yamakawa family of bi-starlike functions associated with the Chebyshev and Legendre polynomials, which are left as an exercise to interested readers.
The authors of reference [4], Alaa H. El-Qadeem and Ibrahim S. Elshazly, study the Hadamard product features of certain subclasses of *p*-valent meromorphic functions defined in the punctured open-unit disc using the q-difference operator. Convolution properties and coefficient estimates are also established regarding the new subclasses defined in this study. The authors suggest that future researchers focus on the use of these subclasses in studies involving the theories of differential subordination and superordination and also the investigation of the Fekete–Szeg˝o problem.
In the research presented in reference [5], Georgia Irina Oros, Gheorghe Oros, and Ancut,a Maria Rus use the confluent hypergeometric function embedded in the theory of strong differential superordinations. The form of the confluent hypergeometric function and that of the previously defined Kummer–Bernardi and Kummer–Libera operators are adopted by considering certain classes of analytic functions depending on an extra parameter previously introduced related to the theory of strong differential subordination and superordination. Strong differential superordinations are investigated, and the best subordinates are given. The applications of the established theoretical results are illustrated through two examples. As potential future studies, the authors suggest the use of the dual notion of strong differential subordination for investigations concerning the confluent hypergeometric function and the two operators used in the present study, which could yield sandwich-type results if combined with the results contained in this paper.
The topic of introducing new subclasses of bi-starlike and bi-convex functions of a complex order associated with the Erdély–Kober-type integral operator in the open-unit disc is considered by Alhanouf Alburaikan, Gangadharan Murugusundaramoorthy, and Sheza M. El-Deeb [6]. The estimates of initial coefficients are given, and Fekete–Szeg˝o inequalities are investigated for the functions in those classes. Several consequences of the results are also highlighted as examples.
For the study presented in reference [7], Feras Yousef, Ala Amourah, Basem Aref Frasin, and Teodor Bulboacă again consider certain new subclasses of bi-univalent functions by exploiting the zero-truncated Poisson distribution probabilities and involving Gegenbauer polynomials and the concept of subordination. Coefficient-related problems are investigated, and the Fekete–Szeg˝o functional problem is solved for those classes. The authors suggest that the results offered in this paper would lead to other different new results involving Legendre and Chebyshev polynomials.
Considering the importance of the logarithmic coefficients, in reference [8], Sevtap Sümer Eker, Bilal ¸Seker, Bilal Çekiç, and Mugur Acu obtain the sharp bounds for the second Hankel determinant concerning the logarithmic coefficients of strongly starlike functions and strongly convex functions. The results presented here could inspire further studies that focus on other subclasses of univalent functions and obtain the boundaries for higher-order Hankel determinants.
New results on the radius of uniform convexity of two kinds of normalization of the Bessel function *J<sup>ν</sup>* in the case of *ν* ∈ (−2, −1) are presented by Lumini¸ta-Ioana Cotîrlă, Pál Aurel Kupán, and Róbert Szász in reference [9]. This study provides alternative proof regarding the radius of convexity of order alpha. The authors also provide alternative proof regarding the radius of convexity of order alpha and derive an interesting correlation between convexity and uniform convexity.
The research presented by Richard D. Carmichael in reference [10] is connected to the results obtained in reference [1]. A boundary value result concerning vector-valued tempered distributions as the boundary values of vector-valued analytic functions is given under the general norm growth on the analytic function, which is equivalent to the growth of Tillmann. The second goal of this paper was to obtain a Cauchy integral representation of the analytic functions by using the generally known structure of the spectral function and the structure of the tempered distributional boundary value. The analytic function used to obtain the boundary value was equated to the product of a polynomial and the constructed Cauchy integral. This paper concerns theoretical mathematics; however, the considered topics find applications in mathematical physics and the field of mathematics involving physical problems.
New results are obtained concerning fuzzy differential subordination theory and are highlighted by Alina Alb Lupas, [11]. A previously introduced operator defined by applying the Riemann–Liouville fractional integral to the convex combination of well-known Ruscheweyh and Sălăgean differential operators is used for defining a new fuzzy subclass. The convex property of this class is proved, and certain fuzzy differential subordinations involving the functions from this class and the operator mentioned earlier are obtained. The best fuzzy dominants are given for the considered fuzzy differential subordinations in theorems, and interesting corollaries emerge when specific functions with remarkable geometric properties are used as the best fuzzy dominants. Inspired by the research presented here, researchers can apply the operator used in this paper in future studies for the introduction of other subclasses of analytic functions. The dual theory of fuzzy differential superordination can also be used for obtaining similar results involving the operator and the class defined in this paper.
Using beta-negative binomial distribution series and Laguerre polynomials, Isra Al-Shbeil, Abbas Kareem Wanas, Afis Saliu, and Adriana Căta¸s [12] investigate a new family of normalized holomorphic and bi-univalent functions associated with Ozaki close-toconvex functions. They provide estimates on the initial Taylor–Maclaurin coefficients and discuss Fekete–Szeg˝o type inequality for the functions in this family in the special case of generalized Laguerre polynomials.
A symmetric–convex differential formula of normalized Airy functions in the openunit disk is developed by Samir B. Hadid and Rabha W. Ibrahim in reference [13]. The equation is taken into account as a differential operator in the development of a class of normalized analytic functions. Two-dimensional wave propagation in the earth–ionosphere wave path using *k*-symbol Airy functions is used for the investigation. It is shown that the standard wave-mode working formula may be determined by orthogonality considerations without the use of intricate justifications of the complex plane.
The applications of fractional differential operators in the field of geometric function theory are obtained by Mohammad Faisal Khan, Shahid Khan, Saqib Hussain, Maslina Darus, and Khaled Matarneh in reference [14]. The fractional differential operator and the Mittag–Leffler functions are combined to formulate and arrange a new operator of fractional calculus. A new class of normalized analytic functions is introduced using the newly defined fractional operator, and some of its interesting geometric properties are discussed in the open-unit disk. The authors suggest that the operator introduced here can be utilized to define other classes of analytic functions or to generalize other types of differential operators.
#### **3. Conclusions**
The 14 papers published as part of this Special Issue entitled "New Developments in Geometric Function Theory" concern a broad range of subjects. Researchers interested in different aspects of geometric function theory and its related topics would find interesting insights and inspiring results, leading to increased reference to these contributions and the propagation of this Special Issue to a large audience.
**Acknowledgments:** The guest editor of this Special Issue would like to thank all the authors who decided to submit their works and have contributed to the success of this Special Issue, as well as all the reviewers for their time, constructive remarks, and help in maintaining high standards for the published materials. Special thanks are also given to the editors of *Axioms* and especially to the Managing Editor of this Special Issue, Alex Zhang.
**Conflicts of Interest:** The author declares no conflict of interest.
#### **References**
**Disclaimer/Publisher's Note:** The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.
#### *Article* **Generalized Vector-Valued Hardy Functions**
**Richard D. Carmichael**
Department of Mathematics and Statistics, Wake Forest University, Winston-Salem, NC 27109, USA; [email protected]
**Abstract:** We consider analytic functions in tubes <sup>R</sup>*<sup>n</sup>* <sup>+</sup> *iB* <sup>⊂</sup> <sup>C</sup>*<sup>n</sup>* with values in Banach space or Hilbert space. The base of the tube *B* will be a proper open connected subset of R*<sup>n</sup>* , an open connected cone in R*<sup>n</sup>* , an open convex cone in R*<sup>n</sup>* , and a regular cone in R*<sup>n</sup>* , with this latter cone being an open convex cone which does not contain any entire straight lines. The analytic functions satisfy several different growth conditions in *L <sup>p</sup>* norm, and all of the resulting spaces of analytic functions generalize the vector valued Hardy space *H<sup>p</sup>* in C*<sup>n</sup>* . The analytic functions are represented as the Fourier–Laplace transform of certain vector valued *L p* functions which are characterized in the analysis. We give a characterization of the spaces of analytic functions in which the spaces are in fact subsets of the Hardy functions *H<sup>p</sup>* . We obtain boundary value results on the distinguished boundary <sup>R</sup>*<sup>n</sup>* <sup>+</sup> *<sup>i</sup>*{0} and on the topological boundary <sup>R</sup>*<sup>n</sup>* <sup>+</sup> *<sup>i</sup>∂<sup>B</sup>* of the tube for the analytic functions in the *L <sup>p</sup>* and vector valued tempered distribution topologies. Suggestions for associated future research are given.
**Keywords:** analytic functions; vector valued Hardy functions; boundary values
**MSC:** 32A26; 32A35; 32A40; 42B30
#### **1. Introduction**
In [1] and related work, we defined and analyzed vector-valued Hardy *H<sup>p</sup>* (*T B* , X ) functions on tubes *T <sup>B</sup>* <sup>=</sup> <sup>R</sup>*<sup>n</sup>* <sup>+</sup> *iB* <sup>⊂</sup> <sup>C</sup>*<sup>n</sup>* with values in Banach space <sup>X</sup> . We showed that any Banach space X vector-valued analytic function on *T <sup>B</sup>* which obtained a <sup>X</sup> vector-valued distributional boundary value was a *H<sup>p</sup>* (*T B* , X ), 1 ≤ *p* ≤ ∞, function with values in Banach space X if the X vector-valued boundary value was a *L p* (R*<sup>n</sup>* , X ), 1 ≤ *p* ≤ ∞, function. We showed that the *H<sup>p</sup>* (*T B* , X ), 1 ≤ *p* ≤ ∞, functions admitted a representation by the Poisson integral of *L p* (R*<sup>n</sup>* , X ), 1 ≤ *p* ≤ ∞, functions if the values of the analytic functions were in a certain type of Banach space and then obtained a pointwise growth estimate for the *H<sup>p</sup>* (*T B* , X ) functions for this Banach space. In additional analysis, we have obtained many general results concerning *H<sup>p</sup>* (*T B* , X ) functions with values in Banach space including representations as Fourier–Laplace, Cauchy, and Poisson integrals and the existence of boundary values.
Previously, we defined generalizations of *H<sup>p</sup>* (*T B* ) functions in the scalar-valued case by using several more general growth conditions on the *L <sup>p</sup>* norm of the analytic functions. Some of these scalar-valued results are contained in [2] (Chapter 5); other such results in the scalar-valued case are contained in papers listed under the author's name in the references in [1,2]. In this paper, we build upon these scalar-valued generalizations of *H<sup>p</sup>* (*T B* ) functions by considering the vector-valued case of functions and distributions with values in Banach or Hilbert space. The generalizations of the vector-valued analytic functions in *H<sup>p</sup>* (*T B* , X ), X being a Banach space, which we consider here are defined in Section 4 of this paper. Our results are obtained for the base *B* of the tube *T B* successively being a proper open connected subset of R*<sup>n</sup>* , an open connected cone in R*<sup>n</sup>* , an open convex cone in R*<sup>n</sup>* , and a regular cone in R*<sup>n</sup>* , with this latter cone being an open convex cone which does not contain any entire straight lines; as the base *B* of the tube *T B* is
**Citation:** Carmichael, R.D. Generalized Vector-Valued Hardy Functions. *Axioms* **2022**, *11*, 39. https://doi.org/10.3390/ axioms11020039
Academic Editor: Georgia Irina Oros
Received: 18 December 2021 Accepted: 17 January 2022 Published: 20 January 2022
**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.
**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).
specialized, increasingly precise results are obtained in the analysis. For *B* being a proper open connected subset of R*<sup>n</sup>* we show, for example, that the growth condition that defines the functions which generalize the Hardy functions can, in certain circumstances, be extended to the boundary of the base *B* of the tube *T B* . At the open convex cone stage in our analysis we are able to show the equivalence of two types of vector-valued functions which generate *H<sup>p</sup>* (*T B* , X ) functions. In the cone setting for base *B* we show that certain elements of the defined analytic functions are in fact *H<sup>p</sup>* (*T B* ) functions which leads to the representation of these functions as Fourier–Laplace, Cauchy, and Poisson integrals. In the case that *B* is a regular cone we study the boundary values on the topological boundary of the tube defined by the cone as points in *B* approach a point on its boundary through circular bands within *B*. In general, our goal in this paper is to obtain results for the functions defined in Section 4 treated as generalizations of *H<sup>p</sup>* (*T B* , X ) functions and as generalizations of the scalar-valued functions noted in [2] (Chapter 5) and in some of our papers referenced in [2] and hence to generalize results concerning *H<sup>p</sup>* (*T B* , X ) spaces and concerning the scalar-valued functions noted in [2] (Chapter 5) and in certain references of [2] to these new spaces of analytic functions. Additionally, our goal is to obtain additional new results for the analytic functions of Section 4 which we accomplish.
As noted above, the vector-valued analytic functions considered in this paper are defined in Section 4. In Section 5, we show that certain vector-valued measurable functions generate the analytic functions by the Fourier–Laplace transform; conversely, in Section 6, we generate the measurable functions from the analytic functions and show that the analytic functions are representable through the generated measurable functions. As the base *B* of the tube *T B* is made more specific the analytic functions and measurable functions obtain more specific properties. In Section 7, we show that under specified conditions the analytic functions considered are in fact vector-valued Hardy *H*<sup>2</sup> functions which immediately results in Cauchy and Poisson integral representations. Section 8 concerns the existence of boundary values of the analytic functions in vector-valued *L <sup>p</sup>* and in vector-valued <sup>S</sup> 0 topologies on both the distinguished boundary and the topological boundary of the tube. Problems for future research are considered in Section 9, and conclusions are provided in Section 10.
#### **2. Definitions and Notation**
Throughout, X will denote a Banach space, H will denote a Hilbert space, N will denote the norm of the specified Banach or Hilbert space, and Θ will denote the zero vector of the specified Banach or Hilbert space. We reference Dunford and Schwartz [3] for integration of vector-valued functions and for vector-valued analytic functions. For foundational information concerning vector-valued distributions we refer to Schwartz ([4,5]).
The n-dimensional notation used in this paper will be the same as that in [1,2]. The information concerning cones in R*<sup>n</sup>* needed here is contained in [2] (Chapter 1). We recall some very important notation and concepts of cones here that are necessary for this paper. *<sup>C</sup>* <sup>⊂</sup> <sup>R</sup>*<sup>n</sup>* is a cone (with vertex at <sup>0</sup> = (0, 0, ..., 0) <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* ) if *y* ∈ *C* implies *λy* ∈ *C* for all positive scalars *λ*. The intersection of *C* with the unit sphere |*y*| = 1 is called the projection of *C* and is denoted *pr*(*C*). A cone *C* 0 such that *pr*(*C*0) ⊂ *pr*(*C*) is a compact subcone of *C* which we will denote as *C* <sup>0</sup> ⊂⊂ *C*. The function
$$\mu\_{\mathbb{C}}(t) = \sup\_{y \in pr(\mathbb{C})} (-\langle t, y \rangle)\_{\prime} \text{ } t \in \mathbb{R}^n\text{ } t$$
is the indicatrix of *C*. The dual cone *C* ∗ of *C* is defined as
$$\mathcal{C}^\* = \{ t \in \mathbb{R}^n : \langle t, y \rangle \ge 0 \text{ for all } y \in \mathcal{C} \}$$
and satisfies *C* <sup>∗</sup> <sup>=</sup> {*<sup>t</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* : *uC*(*t*) ≤ 0}. An open convex cone which does not contain any entire straight lines will be called a regular cone. See [2] (Section 1.2) for examples of cones in R*<sup>n</sup>* . In this paper, we will be concerned with the distance from a point in a cone to the boundary of the cone; for *C* being an open connected cone in R*<sup>n</sup>* , the distance from *y* ∈ *C* to the topological boundary *∂C* of *C* is
$$d(y) = \inf\{|y - y\_1| : y\_1 \in \partial \mathcal{C}\}.$$
For an open connected cone *<sup>C</sup>* <sup>⊂</sup> <sup>R</sup>*<sup>n</sup>* , we know from [2] (p. 6, (1.14)) that
$$d(y) = \inf\_{t \in pr(\mathbb{C}^\*)} \langle t, y \rangle\_{\prime} \ y \in \mathbb{C}\_{\prime}$$
and 0 < *d*(*y*) ≤ |*y*|, *y* ∈ *C*. Additionally, *d*(*λy*) = *λd*(*y*), *λ* > 0.
The *L p* (R*<sup>n</sup>* , X ) functions, 1 ≤ *p* ≤ ∞, with values in X and their norm |**h**|*<sup>p</sup>* are noted in [3] (Chapter III). The Fourier transform on *L* 1 (R*<sup>n</sup>* ) or *L* 1 (R*<sup>n</sup>* , X ) is given in [2] (p. 3). All Fourier (inverse Fourier) transforms on scalar or vector-valued functions will be denoted *<sup>φ</sup>*<sup>b</sup> <sup>=</sup> <sup>F</sup>[*φ*(*t*); *<sup>x</sup>*] (<sup>F</sup> <sup>−</sup><sup>1</sup> [*φ*(*t*); *x*]). As stated in [6] the Plancherel theory is not true for vectorvalued functions except when X = H, a Hilbert space. The Plancherel theory is complete in the *L* 2 (R*<sup>n</sup>* , H) setting in that the inverse Fourier transform is the inverse mapping of the Fourier transform with <sup>F</sup> <sup>−</sup>1<sup>F</sup> <sup>=</sup> *<sup>I</sup>* <sup>=</sup> F F <sup>−</sup><sup>1</sup> with *<sup>I</sup>* being the identity mapping.
As usual, we denote <sup>S</sup>(R*<sup>n</sup>* ) as the tempered functions with associated distributions being S 0 (R*<sup>n</sup>* ) or associated vector-valued distributions being S 0 (R*<sup>n</sup>* , X ). The Fourier transform on S 0 (R*<sup>n</sup>* ) and on S 0 (R*<sup>n</sup>* , X ) is the usual such definition and is given in [4] (p. 73).
Let *<sup>B</sup>* be an open subset of <sup>R</sup>*<sup>n</sup>* and <sup>X</sup> be a Banach space. The Hardy space *<sup>H</sup><sup>p</sup>* (*T B* , X ), 0 < *p* < ∞, consists of those analytic functions **f**(*z*) on the tube *T <sup>B</sup>* <sup>=</sup> <sup>R</sup>*<sup>n</sup>* <sup>+</sup> *iB* <sup>⊂</sup> <sup>C</sup>*<sup>n</sup>* with values in the Banach space X such that for some constant *M* > 0 and every *y* ∈ *B*
$$\int\_{\mathbb{R}^n} (\mathcal{N}(\mathbf{f}(\mathbf{x} + iy)))^p d\mathbf{x} \le M;$$
the usual modification is made for the case *p* = ∞.
#### **3. Cauchy and Poisson Kernels and Integrals**
Let *C* be a regular cone in R*<sup>n</sup>* . *C* ∗ is the dual cone of *C*. The Cauchy kernel corresponding to *T <sup>C</sup>* = R*<sup>n</sup>* + *iC* is
$$K(z - t) = \int\_{\mathbb{C}^\*} e^{2\pi i \langle z - t, \eta \rangle} d\eta, \ t \in \mathbb{R}^n, \ z \in T^{\mathbb{C}}.$$
The Poisson kernel corresponding to *T <sup>C</sup>* is
$$Q(z;t) = \frac{K(z-t)\overline{K(z-t)}}{K(2iy)} = \frac{|K(z-t)|^2}{K(2iy)}, \ t \in \mathbb{R}^n, \ z \in T^\mathbb{C}.$$
Referring to [2] (Chapters 1 and 4) for details, we know for *z* ∈ *T <sup>C</sup>* that *<sup>K</sup>*(*<sup>z</sup>* − ·) <sup>∈</sup> D(∗, *L p* ) ⊂ D*<sup>L</sup> <sup>p</sup>* , 1 < *p* ≤ ∞; and *Q*(*z*; ·) ∈ D(∗, *L p* ) ⊂ D*<sup>L</sup> <sup>p</sup>* , 1 ≤ *p* ≤ ∞, where ∗ is Beurling (*Mp*) or Roumieu {*Mp*}. These ultradifferentiable functions are contained in the Schwartz space D*<sup>L</sup> <sup>p</sup>* = D(*L p* , R*<sup>n</sup>* ). Because of the combined properties of the Cauchy and Poisson kernels from [2], we know that the Cauchy and Poisson integrals
$$\int\_{\mathbb{R}^n} \mathbf{h}(t)K(z-t)dt,\ z \in T^{\mathbb{C}}\_{\omega}$$
and <sup>Z</sup>
$$\int\_{\mathbb{R}^n} \mathbf{h}(t) Q(z; t) dt, \; z \in T^{\mathbb{C}}\mu$$
are well defined for **h** ∈ *L p* (R*<sup>n</sup>* , X ), 1 ≤ *p* < ∞, and **h** ∈ *L p* (R*<sup>n</sup>* , X ), 1 ≤ *p* ≤ ∞, respectively, for X being a Banach space.
We conclude this section with a boundary value calculation concerning the integral which defines the Cauchy kernel. Our calculations here provide motivation and guidance for boundary value results concerning the analytic functions considered in this paper which we obtain subsequently. Let *C* be a regular cone and put
$$K(z) = \int\_{\mathbb{C}^\*} e^{2\pi i \langle z, t \rangle} dt, \; z \in T^{\mathbb{C}}.$$
We know that *K*(*z*) is analytic in *T <sup>C</sup>* and is a bounded function of *<sup>x</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* for *y* ∈ *C*. Thus, *K*(*x* + *iy*) ∈ S<sup>0</sup> (R*<sup>n</sup>* ) as a function of *<sup>x</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* for *y* ∈ *C*. Let *IC*<sup>∗</sup> (*t*) denote the characteristic function of *C* ∗ . We have the following result concerning points on the boundary of *C*, *∂C*.
**Theorem 1.** *Let y<sup>o</sup>* ∈ *∂C. We have*
$$\lim\_{y \to y\_{\circ}, y \in \mathbb{C}} \mathsf{K}(\mathfrak{x} + iy) = \mathcal{F}[I\_{\mathbb{C}^\*}(t)e^{-2\pi \langle y\_{\circ}, t \rangle}]$$
*in the strong topology of* S 0 (R*<sup>n</sup>* )*.*
**Proof.** For *y<sup>o</sup>* ∈ *∂C*, choose a sequence of points {*ym*}, *m* = 1, 2, ..., in *C* which converges to *yo*. We have
$$\langle y\_{o\prime}t\rangle = \lim\_{y\_m \to y\_o} \langle y\_m\, t\rangle \ge 0, \,\, t \in \mathbb{C}^\*.$$
Thus, *e* −2*π*h*yo*,*t*i *IC*<sup>∗</sup> (*t*) ∈ S<sup>0</sup> (R*<sup>n</sup>* ) and F[*e* −2*π*h*yo*,*t*i *IC*<sup>∗</sup> (*t*)] ∈ S<sup>0</sup> (R*<sup>n</sup>* ). Let *<sup>φ</sup>* ∈ S(R*<sup>n</sup>* ) and *y* ∈ *C*.
$$
\begin{split} & \langle K(\mathfrak{x} + i\mathfrak{y}) - \mathcal{F}[e^{-2\pi \langle y\_o, t \rangle} I\_{\mathbb{C}^\*}(t)], \phi(\mathfrak{x}) \rangle \\ & = \langle \mathcal{F}[(e^{-2\pi \langle y, t \rangle} - e^{-2\pi \langle y\_o, t \rangle}) I\_{\mathbb{C}^\*}(t)], \phi(\mathfrak{x}) \rangle \\ & = \langle (e^{-2\pi \langle y, t \rangle} - e^{-2\pi \langle y\_o, t \rangle}) I\_{\mathbb{C}^\*}(t), \widehat{\phi}(t) \rangle. \end{split} $$
Now
$$|(e^{-2\pi \langle y,t\rangle} - e^{-2\pi \langle y,t\rangle})I\_{\mathbb{C}^\*}(t)\widehat{\phi}(t)| \le (e^{-2\pi \langle y,t\rangle} + e^{-2\pi \langle y\_0,t\rangle})I\_{\mathbb{C}^\*}(t)|\widehat{\phi}(t)| \le 2|\widehat{\phi}(t)|.$$
By the Lebesgue dominated convergence theorem, we have
$$\lim\_{y \to y\_{\circ}, y \in \mathbb{C}} K(\mathfrak{x} + iy) = \mathcal{F}[I\_{\mathbb{C}^\*}(t)e^{-2\pi \langle y\_{\circ}, t \rangle}],$$
in the weak topology of S 0 (R*<sup>n</sup>* ). Since <sup>S</sup>(R*<sup>n</sup>* ) is a Montel space we have this convergence in the strong topology of S 0 (R*<sup>n</sup>* ) also.
In Theorem 1, notice that 0 is on the boundary of *C*. Thus, for *y<sup>o</sup>* = 0,
$$\lim\_{y \to \overline{0}, y \in \mathbb{C}} K(\mathfrak{x} + i\mathfrak{y}) = \mathcal{F}[I\_{\mathbb{C}^\*}(t)],$$
in the strong topology of S 0 (R*<sup>n</sup>* ) in the conclusion of Theorem 1.
#### **4. The Analytic Functions**
As previously noted, we have studied vector-valued Hardy spaces in [1]; previous to this analysis we had generalized scalar-valued Hardy spaces by placing a more general bound on the *L <sup>p</sup>* norm of the scalar-valued analytic function. These main scalar-valued generalizations are contained in [2] with other related work referenced in [2]. In the scalar-valued generalizations, we obtained Fourier–Laplace transform representation of the analytic functions and characterized the measurable function which generated this representation along with related results.
Given our recent work in studying vector-valued Hardy spaces, we now desire to study vector-valued generalizations of vector-valued Hardy spaces.
In this section, we introduce and define the vector-valued analytic functions that we study here. Throughout this section, *B* will denote a proper open connected subset of R*<sup>n</sup>* unless stated otherwise; and, as previously stated, X will denote a Banach space with norm N .
**Definition 1.** *H p A* (*T B* , X ), 1 ≤ *p* < ∞*, is the set of analytic functions f*(*z*) *on T <sup>B</sup> with values in* X *such that*
$$|f(\mathbf{x} + i\mathbf{y})|\_p = \left(\int\_{\mathbb{R}^n} (\mathcal{N}(f(\mathbf{x} + i\mathbf{y})))^p d\mathbf{x}\right)^{1/p} \le M(1 + (d(\mathbf{y}))^{-r})^s e^{2\pi A|\mathbf{y}|}, \ y \in \mathbb{B}\_n$$
*where r* ≥ 0,*s* ≥ 0, *A* ≥ 0*, and M* = *M*(*f*, *p*, *A*,*r*,*s*) > 0*.*
**Definition 2.** *R p A* (*T B* , X ), 1 ≤ *p* < ∞*, is the set of analytic functions f*(*z*) *on T <sup>B</sup> with values in* X *such that*
$$|f(x+iy)|\_p \le M(1+|y|^{-r})^s e^{2\pi A|y|}, \ y \in B\_{\epsilon}$$
*where r* ≥ 0,*s* ≥ 0, *A* ≥ 0*, and M* = *M*(*f*, *p*, *A*,*r*,*s*) > 0*.*
**Definition 3.** *V p A* (*T B* , X ), 1 ≤ *p* < ∞*, is the set of analytic functions f*(*z*) *on T <sup>B</sup> with values in* X *such that*
$$|f(x+iy)|\_p \le Me^{2\pi A|y|}, \ y \in B\_\nu$$
*were A* ≥ 0 *and M* = *M*(*f*, *p*, *A*) > 0*.*
We consider situations and examples which help emphasize containment of these spaces although the definitions of these sets of functions show the containment in many cases. If *B* is an open connected cone we know from Section 2 that *d*(*y*) ≤ |*y*|, *y* ∈ *B*; thus, *R p A* (*T B* , X ) ⊆ *H p A* (*T B* , X ) in general in this case. For specific examples which help show proper containment let us just consider scalar-valued analytic functions in half planes in C1 . Let *B* = (0, ∞); thus, *T* (0,∞) = R<sup>1</sup> + *i*(0, ∞). We have
$$f(z) = \frac{e^{-2\pi iz}}{z(i+z)} \in \mathbb{R}\_1^2(T^{(0,\infty)}, \mathbb{C}^1) \cap H\_1^2(T^{(0,\infty)}, \mathbb{C}^1), \ y = Im(z) \in (0,\infty)\_\*$$
as
$$||f(\mathfrak{x} + iy)|\_{L^2(\mathbb{R}^1)} \le \pi^{1/2} (1 + y^{-1}) e^{2\pi y}, \ y = \operatorname{Im}(z) > 0;$$
but this *f*(*z*) is not in *V* 2 1 (*T* (0,∞) , C<sup>1</sup> ). We have
$$f(z) = \frac{e^{-2\pi iz}}{i+z} \in V\_1^2(T^{(0,\infty)}, \mathbb{C}^1)$$
but is not in *H*<sup>2</sup> (*T* (0,∞) , C<sup>1</sup> ). Of course *<sup>f</sup>*(*z*) = 1/(*<sup>i</sup>* <sup>+</sup> *<sup>z</sup>*) <sup>∈</sup> *<sup>H</sup>*<sup>2</sup> (*T* (0,∞) , C<sup>1</sup> ) and hence is in all of *V* 2 1 (*T* (0,∞) , C<sup>1</sup> ), *R* 2 1 (*T* (0,∞) , C<sup>1</sup> ), and *H*<sup>2</sup> 1 (*T* (0,∞) , C<sup>1</sup> ). These examples help to see the containment of the defined spaces and the Hardy functions for most of the specified conditions on the base *B* of the tube *T B* in our analysis in this paper.
For our next set of analytic functions, we must remember properties of sequences *Mp*, *p* = 0, 1, 2, ..., with which ultradifferentiable functions and ultradistributions are defined. These sequences and properties are discussed in [2] (Section 2.1). In this paper, we are principally concerned with the properties (*M*.1) and (*M*.30 ) and with the associated function
$$M^\*(\rho) = \sup\_p \log(\rho^p p! M\_0/M\_p), \ 0 < \rho < \infty.$$
With these facts in mind we define additional vector-valued analytic functions.
**Definition 4.** *For B, being a proper open connected subset of* R*<sup>n</sup> which does not contain* 0*, H p* <sup>∗</sup> (*T B* , <sup>X</sup> ), 1 <sup>≤</sup> *<sup>p</sup>* <sup>&</sup>lt; <sup>∞</sup>*, is the set of analytic functions <sup>f</sup>*(*z*) *on T<sup>B</sup> with values in* <sup>X</sup> *such that*
$$|f(x+iy)|\_p \le K(1+(d(y))^{-r})^s e^{M^\*(w/|y|)}, \ y \in \mathbb{B}\_r$$
*where r* ≥ 0,*s* ≥ 0, *w* > 0*, and K* = *K*(*f*, *p*,*r*,*s*, *w*) > 0*.*
With Definition 4 in place, we can now state definitions for *R p* <sup>∗</sup>(*T B* , X ) and *V p* <sup>∗</sup> (*T B* , X ) from Definition 4 similarly as we did for *R p A* (*T B* , X ) and *V p A* (*T B* , X ) from Definition 1. In the scalar-valued case, we have proved that the Cauchy integral of ultradistributions *U* ∈ D<sup>0</sup> (∗, *L p* ), where ∗ is Beurling (*Mp*) or Roumieu {*Mp*}, is analytic in *T <sup>C</sup>* and satisfies the growth of Definition 4 where *C* is a regular cone in R*<sup>n</sup>* ; see [2] (Section 4.2). Additionally, we have obtained boundary value results for scalar-valued functions of the type in Definition 4 in [2] (Chapter 5).
Throughout this paper, results concerning *H p* <sup>∗</sup> (*T B* , X ) and its subsets and associated norm growth bounds are obtained under the assumption that the sequence of positive numbers *Mp*, *p* = 0, 1, 2, ..., from which the associated function *M*∗ (*ρ*) is defined, will always be assumed to satisfy properties (*M*.1) and (*M*.30 ) in [2] (p. 13).
#### **5. Measurable Functions Generating Analytic Functions**
The results which we will prove in this paper are obtained for functions in *H p A* (*T B* , X ) of Definition 1 and for functions in *H p* <sup>∗</sup> (*T B* , X ) of Definition 4 by very similar methods. The results corresponding to *H p A* (*T B* , X ) however are somewhat more general in nature than the corresponding ones for *H p* <sup>∗</sup> (*T B* , X ). Thus, we will concentrate our proofs on the results corresponding to *H p A* (*T B* , X ) and subsequently state the corresponding results for *H p* <sup>∗</sup> (*T B* , X ) which will be denoted by a \* next to the result number.
We begin by obtaining properties on measurable functions which we will use to generate analytic functions in *H p A* (*T B* , <sup>X</sup> ). Let *<sup>B</sup>* be a proper open connected subset of <sup>R</sup>*<sup>n</sup>* and let X be a Banach space. Let 1 ≤ *p* < ∞ and **g**(*t*) be a X valued measurable function on R*<sup>n</sup>* such that
$$|e^{-2\pi \langle y, t \rangle} \mathbf{g}(t)|\_p \le M(1 + (d(y))^{-r})^s e^{2\pi A|y|}, \ y \in \mathcal{B},\tag{1}$$
where *r* ≥ 0,*s* ≥ 0, *A* ≥ 0, and *M* = *M*(**g**, *p*, *A*,*r*,*s*) > 0 do not depend on *y* ∈ *B*.
**Theorem 2.** *For B, being a proper open connected subset of* R*<sup>n</sup> and* X *being a Banach space let <sup>g</sup>*(*t*) *be a* <sup>X</sup> *valued measurable function on* <sup>R</sup>*<sup>n</sup> such that* (1) *holds for y* ∈ *B and for* 1 ≤ *p* < ∞*. We have*
$$f(z) = \int\_{\mathbb{R}^n} \mathbf{g}(t) e^{2\pi i \langle z, t \rangle} dt, \; z = x + iy \in T^B,\tag{2}$$
*is a* X *valued analytic function of z* ∈ *T B .*
**Proof.** Let *y<sup>o</sup>* ∈ *B*. Choose an open neighborhood *N*(*yo*;*r*), *r* > 0, and a compact subset *<sup>S</sup>* <sup>⊂</sup> *<sup>B</sup>* such that *<sup>y</sup><sup>o</sup>* <sup>∈</sup> *<sup>N</sup>*(*yo*;*r*) <sup>⊂</sup> *<sup>S</sup>* <sup>⊂</sup> *<sup>B</sup>*. Decompose <sup>R</sup>*<sup>n</sup>* into a union of a finite number of non-overlapping cones *C*1, *C*2, ..., *C<sup>k</sup>* each having vertex at 0 and such that whenever two points *y*<sup>1</sup> and *y*<sup>2</sup> belong to one of these cones the angle between the rays from 0 to *y*<sup>1</sup> and from 0 to *y*<sup>2</sup> is less than *π*/4 radians; and hence h*y*1, *y*2i = |*y*1||*y*2|*cos*(*θ*) > |*y*1||*y*2|2 1/2/2 where *θ* is the angle between the two rays. There is a *δ* > 0 such that 0 < *δ* < *r* and {*y* : |*y* − *yo*| = *δ*} ⊂ *N*(*yo*;*r*). Put *e* = 2*πpδ*/2 1/2 > 0. For each *j* = 1, 2, ..., *k* choose a fixed *y<sup>j</sup>* such that *y<sup>o</sup>* − *y<sup>j</sup>* ∈ *C<sup>j</sup>* and |*y<sup>j</sup>* − *yo*| = *δ*. For each *j* = 1, 2, ..., *k* let *t* ∈ *C<sup>j</sup>* ; we have
$$|\langle y\_o - y\_{\mathbf{j}\prime}t \rangle| \ge |y\_o - y\_{\mathbf{j}}| |t| / 2^{1/2}, \ t \in \mathbb{C}\_{\mathbf{j}\prime} \ j = 1, 2, \dots, k.$$
Thus, for *t* ∈ *C<sup>j</sup>* , *j* = 1, 2, ..., *k*,,
$$\varepsilon|t| = (2\pi p \delta / 2^{1/2})|t| = 2\pi p |y\_o - y\_j||t|/2^{1/2} \le 2\pi p \langle y\_o - y\_{j\_l}t \rangle = -2\pi p \langle y\_j - y\_{o\cdot}t \rangle.$$
Hence, for each *j* = 1, 2, ..., *k*, using (1) we have
$$\begin{aligned} &\int\_{\mathbb{C}\_{j}} e^{-2\pi p \langle y\_{o}, t \rangle} e^{\varepsilon |t|} (\mathcal{N}(\mathbf{g}(t)))^{p} dt \\ &\leq \int\_{\mathbb{C}\_{j}} e^{-2\pi p \langle y\_{o}, t \rangle} e^{-2\pi p \langle y\_{j} - y\_{o}t \rangle} (\mathcal{N}(\mathbf{g}(t)))^{p} dt = \int\_{\mathbb{C}\_{j}} e^{-2\pi p \langle y\_{j}t \rangle} (\mathcal{N}(\mathbf{g}(t)))^{p} dt \\ &\leq \int\_{\mathbb{R}^{n}} (\mathcal{N}(e^{-2\pi \langle y\_{j}t \rangle} \mathbf{g}(t)))^{p} dt \leq M^{p} (1 + (d(y\_{j}))^{-r})^{sp} e^{2\pi p A |y\_{j}|} \end{aligned}$$
and
$$\int\_{\mathbb{R}^n} e^{-2\pi p \langle y\_0 t \rangle} e^{\varepsilon |t|} (\mathcal{N}(\mathbf{g}(t)))^p dt = \sum\_{j=1}^k \int\_{\mathbb{C}\_j} e^{-2\pi p \langle y\_0 t \rangle} e^{\varepsilon |t|} (\mathcal{N}(\mathbf{g}(t)))^p dt$$
$$0 \le M^p \sum\_{j=1}^k (1 + (d(y\_j))^{-r})^{sp} e^{2\pi p A |y\_j|} \tag{3}$$
for arbitrary *<sup>y</sup><sup>o</sup>* <sup>∈</sup> *<sup>B</sup>*. For *<sup>p</sup>* <sup>=</sup> <sup>1</sup> and the fact that (*e*|*t*|/2) <sup>≤</sup> *<sup>e</sup>*|*t*|, *<sup>t</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* , we have from (3) that
$$\int\_{\mathbb{R}^n} e^{-2\pi \langle y\_0, t \rangle} e^{\varepsilon |t|/2} \mathcal{N}(\mathbf{g}(t)) dt \le M \sum\_{j=1}^k (1 + (d(y\_j))^{-r})^s e^{2\pi A |y\_j|}. \tag{4}$$
For 1 < *p* < ∞, H*o*¨lder's inequality, the identity *e <sup>e</sup>*|*t*|/2*<sup>p</sup>* = *e e*|*t*|/*p e* <sup>−</sup>*e*|*t*|/2*<sup>p</sup>* and (3) yield
$$\int\_{\mathbb{R}^n} e^{-2\pi \langle y\_o, t \rangle} e^{\varepsilon |t|/2p} \mathcal{N}(\mathbf{g}(t)) dt \le ||e^{-\varepsilon |t|/2p}||\_{L^q(\mathbb{R}^n)} |e^{-2\pi \langle y\_o, t \rangle} e^{\varepsilon |t|/p} \mathbf{g}(t)|\_p$$
$$\le M ||e^{-\varepsilon |t|/2p}||\_{L^q(\mathbb{R}^n)} \left(\sum\_{j=1}^k (1 + (d(y\_j))^{-r})^{sp} e^{2\pi p A |y\_j|}\right)^{1/p} \tag{5}$$
where 1/*p* + 1/*q* = 1. If |*y* − *yo*| < *e*/4*πp*, *y* = Im(*z*), 1 ≤ *p* < ∞, then for *z* = *x* + *iy*
$$\begin{split} \mathcal{N}(\mathbf{g}(t)e^{2\pi i \langle z, t \rangle}) &= e^{-2\pi \langle y, t \rangle} \mathcal{N}(\mathbf{g}(t)) = e^{-2\pi \langle y - y\_o, t \rangle} e^{-2\pi \langle y\_o, t \rangle} \mathcal{N}(\mathbf{g}(t)) \\ &\leq e^{2\pi |y - y\_o||t|} e^{-2\pi \langle y\_o, t \rangle} \mathcal{N}(\mathbf{g}(t)) \leq e^{-2\pi \langle y\_o, t \rangle} e^{\varepsilon |t|/2p} \mathcal{N}(\mathbf{g}(t)) \end{split} \tag{6}$$
for all *<sup>t</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* . (4) and (5) now show that the right side of (6) is a *L* 1 (R*<sup>n</sup>* ) function which is independent of *y* = Im(*z*) such that |*y* − *yo*| < *e*/4*πp* for all cases 1 ≤ *p* < ∞. Since *y<sup>o</sup>* ∈ *B* is arbitrary we conclude from (6) that *f*(*z*) defined by (2) is a X valued analytic function of *z* ∈ *T B* . Further, (6) proves that *e* <sup>−</sup>2*π*h*y*,*t*i**g**(*t*) <sup>∈</sup> *<sup>L</sup>* 1 (R*<sup>n</sup>* , X ), *y* ∈ *B*, for all cases 1 ≤ *p* < ∞ in addition to the fact that *e* <sup>−</sup>2*π*h*y*,*t*i**g**(*t*) <sup>∈</sup> *<sup>L</sup> p* (R*<sup>n</sup>* , X ), *y* ∈ *B*, for each of the specific cases for *p*, 1 ≤ *p* < ∞, because of the assumption (1). The proof is complete.
The exact same method of proof used for Theorem 2 yields the following result corresponding to the growth for *H p* <sup>∗</sup> (*T B* , X ).
**Theorem 3.** *Let <sup>B</sup> be a proper open connected subset of* <sup>R</sup>*<sup>n</sup> which does not contain* <sup>0</sup> <sup>∈</sup> <sup>R</sup>*<sup>n</sup> , and let* <sup>X</sup> *be a Banach space. Let* <sup>1</sup> <sup>≤</sup> *<sup>p</sup>* <sup>&</sup>lt; <sup>∞</sup> *and <sup>g</sup>*(*t*) *be a* <sup>X</sup> *valued measurable function on* <sup>R</sup>*<sup>n</sup> such that*
$$|e^{-2\pi \langle y, t \rangle} \mathbf{g}(t)|\_p \le M(1 + (d(y))^{-r})^s e^{M^\*(w/|y|)}, \ y \in B\_{\nu}$$
*where r* ≥ 0,*s* ≥ 0, *w* > 0*, and M* = *M*(*g*, *p*,*r*,*s*, *w*) > 0 *are independent of y* ∈ *B. We have*
$$f(z) = \int\_{\mathbb{R}^n} \mathbf{g}(t) e^{2\pi i \langle z, t \rangle} dt, \ z \in T^B.$$
*is a* X *valued analytic function of z* ∈ *T B .*
The Fourier transform of vector-valued functions *L p* (R*<sup>n</sup>* , X ) with the Plancherel theory and Parseval identity holding occurs only if *p* = 2 and X = H, a Hilbert space. For *p* = 2 in order to have an isomorphism of the Fourier transform of *L* 2 (R*<sup>n</sup>* , X ) onto itself with the Parseval identity holding it is necessary and sufficient that X = H, a Hilbert space [6] (pp. 45, 61). We use the Fourier transform considerably in this paper, and its use is the reason we sometimes restrict the result to *p* = 2 and X = H. We obtain a corollary to Theorem 2.
**Corollary 1.** *Let B be a proper open connected subset of* R*<sup>n</sup> and* H *be a Hilbert space. Let g*(*t*) *be a* <sup>H</sup> *valued measurable function on* <sup>R</sup>*<sup>n</sup> such that (1) holds for <sup>p</sup>* <sup>=</sup> <sup>2</sup>*. We have <sup>f</sup>*(*z*) <sup>∈</sup> *<sup>H</sup>*<sup>2</sup> *A* (*T B* , H) *for f*(*z*) *defined in (2).*
**Proof. f**(*z*) is analytic in *T <sup>B</sup>* by Theorem 2. By the assumption (1) for *p* = 2 and the proof of Theorem 2, *e* <sup>−</sup>2*π*h*y*,*t*i**g**(*t*) <sup>∈</sup> *<sup>L</sup>* 1 (R*<sup>n</sup>* , H) ∩ *L* 2 (R*<sup>n</sup>* , H) for *y* ∈ *B*. Thus, **f**(*x* + *iy*) = F[*e* <sup>−</sup>2*π*h*y*,*t*i**g**(*t*); *<sup>x</sup>*], *<sup>y</sup>* <sup>∈</sup> *<sup>B</sup>*, with the Fourier transform being in the *<sup>L</sup>* 1 (R*<sup>n</sup>* , H) and the *L* 2 (R*<sup>n</sup>* , H) cases. By the Parseval equality |**f**(*x* + *iy*)|<sup>2</sup> = |*e* <sup>−</sup>2*π*h*y*,*t*i**g**(*t*)|<sup>2</sup> for *<sup>y</sup>* <sup>∈</sup> *<sup>B</sup>*. From (1) the desired growth on **<sup>f</sup>**(*<sup>x</sup>* <sup>+</sup> *iy*) of Definition <sup>1</sup> is obtained, and **<sup>f</sup>**(*z*) <sup>∈</sup> *<sup>H</sup>*<sup>2</sup> *A* (*T B* , H).
Under certain circumstances, the growth on the *L* 2 (R*<sup>n</sup>* , H) function *e* <sup>−</sup>2*π*h*y*,*t*i**g**(*t*), *<sup>y</sup>* <sup>∈</sup> *B*, in Corollary 1 can be extended to hold for *y* ∈ *B*.
**Corollary 2.** *Assume the hypotheses of Corollary 1 with the addition that (1) holds for p* = 2 *with r* = 0 *or s* = 0*. We have f*(*z*) ∈ *V* 2 *A* (*T B* , H) *for f*(*z*) *defined in (2) and*
$$|e^{-2\pi \langle y, t \rangle} \mathbf{g}(t)|\_2 \le Me^{2\pi A|y|}, \ y \in \overline{B}.$$
*Further if* 0 ∈ *∂B then g* ∈ *L* 2 (R*<sup>n</sup>* , H)*.*
**Proof.** From the proof of Corollary 1 and Definition 3, we have **f**(*z*) ∈ *V* 2 *A* (*T B* , H) for *r* = 0 or *s* = 0 in (1). Let *y<sup>o</sup>* ∈ *∂B* and let {*ym*} be a sequence of points in *B* which converges to *yo*. By Fatou's lemma we have
$$\begin{aligned} &\int\_{\mathbb{R}^n} e^{-4\pi \langle y\_o, t \rangle} (\mathcal{N}(\mathbf{g}(t)))^2 dt \le \limsup\_{m \to \infty} \int\_{\mathbb{R}^n} e^{-4\pi \langle y\_m, t \rangle} (\mathcal{N}(\mathbf{g}(t)))^2 dt \\ &\le \limsup\_{m \to \infty} M^2 e^{4\pi A \langle y\_m \rangle} = M^2 e^{4\pi A \langle y\_o \rangle} \end{aligned}$$
and
$$|e^{-2\pi \langle y\_{\circ}, t \rangle} \mathbf{g}(t)|\_{2} \le M e^{2\pi A |y\_{\circ}|\_{\cdot}}$$
Thus, (1) holds with *r* = 0 or *s* = 0 for *y* ∈ *B*. If 0 ∈ *∂B* then **g**(*t*) ∈ *L* 2 (R*<sup>n</sup>* , H) from the above inequality for *y<sup>o</sup>* = 0.
For *<sup>B</sup>* being a proper open connected subset of <sup>R</sup>*<sup>n</sup>* and <sup>X</sup> being a Banach space, assume (1) holds for 1 ≤ *p* < ∞ with *r* = 0 or *s* = 0 and for **g** having values in X . The proof of Corollary 2 shows that (1) will hold for *y* ∈ *B* in this situation.
We study the extension of **f**(*z*) or *e* <sup>−</sup>2*π*h*y*,*t*i**g**(*t*), *<sup>y</sup>* <sup>∈</sup> *<sup>B</sup>*, in norm to the *<sup>∂</sup><sup>B</sup>* in greater detail later in this paper in section 8.
The proof of the following result is the same as that of Corollary 1 using Theorem 3.
**Corollary 3.** *Let <sup>B</sup> be a proper open connected subset of* <sup>R</sup>*<sup>n</sup> which does not contain* <sup>0</sup> <sup>∈</sup> <sup>R</sup>*<sup>n</sup> , and let* <sup>H</sup> *be a Hilbert space such that the growth of Theorem <sup>3</sup> holds for <sup>p</sup>* <sup>=</sup> <sup>2</sup>*. We have <sup>f</sup>*(*z*) <sup>∈</sup> *<sup>H</sup>*<sup>2</sup> ∗ (*T B* , H) *for f*(*z*) *defined in (2).*
In several following results, we restrict the base *B* of the tube *T B* to cones and obtain additional properties of the function **g**(*t*) in the results. Throughout supp(**g**) denotes the support of **g**.
**Theorem 4.** *Let C be an open connected cone in* R*<sup>n</sup> and* 1 ≤ *p* < ∞*. Let g*(*t*) *be a Banach space* <sup>X</sup> *valued measurable function on* <sup>R</sup>*<sup>n</sup> such that (1) holds for <sup>y</sup>* <sup>∈</sup> *C. We have supp(g*) ⊆ {*<sup>t</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* : *uC*(*t*) ≤ *A*} *almost everywhere (a.e.).*
**Proof.** Assume **<sup>g</sup>**(*t*) <sup>6</sup><sup>=</sup> <sup>Θ</sup> on a set of positive measure in {*<sup>t</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* : *uC*(*t*) > *A*}; there is a point *<sup>t</sup><sup>o</sup>* ∈ {*<sup>t</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* : *uC*(*t*) > *A*} such that **g**(*t*) 6= Θ on a set of positive measure in the neighborhoods *<sup>N</sup>*(*to*, *<sup>η</sup>*) = {*<sup>t</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* : |*t* − *to*| < *η*} for arbitrary *η* > 0. Since *<sup>t</sup><sup>o</sup>* ∈ {*<sup>t</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* : *uC*(*t*) > *A*} there is a point *y<sup>o</sup>* ∈ *pr*(*C*) ⊂ *C* such that (−h*to*, *yo*i) > *A* ≥ 0. Using the continuity of (−h*t*, *yo*i) at *t<sup>o</sup>* as a function of *t*, there is a fixed *σ* > 0 and a fixed neighborhood *N*(*to*; *η* 0 ) such that −h*t*, *yo*i) > *A* + *σ* > 0 for all *t* ∈ *N*(*to*; *η* 0 ). Choose *η* above to be *η* 0 . For any *λ* > 0 we have
$$-\langle \lambda y\_o, t \rangle = -\lambda \langle y\_o, t \rangle > \lambda A + \lambda \sigma > 0, \ t \in N(t\_o; \eta'), \ \lambda > 0. \tag{7}$$
*y<sup>o</sup>* ∈ *pr*(*C*) ⊂ *C* and *C* being a cone imply *λy<sup>o</sup>* ∈ *C*, *λ* > 0. From (7) and (1) with *y* = *λy<sup>o</sup>* we have for all *λ* > 0 that
$$e^{2\pi p(\lambda A + \lambda \sigma)} \int\_{N(t\_0; \mathfrak{y}')} (\mathcal{N}(\mathbf{g}(t)))^p dt \le \int\_{N(t\_0; \mathfrak{y}')} e^{-2\pi \langle \lambda y\_o, t \rangle} (\mathcal{N}(\mathbf{g}(t)))^p dt$$
$$\le \int\_{\mathbb{R}^n} e^{-2\pi p \langle \lambda y\_o, t \rangle} (\mathcal{N}(\mathbf{g}(t)))^p dt \le M^p (1 + (d(\lambda y\_o))^{-r})^{sp} e^{2\pi p A |\lambda y\_o|}$$
$$= M^p (1 + \lambda^{-r} (d(y\_o))^{-r})^{sp} e^{2\pi p \lambda A} \tag{8}$$
since *y<sup>o</sup>* ∈ *pr*(*C*) and *d*(*λyo*) = *λd*(*yo*). The integral on the left of (8) is finite. From (8) we have
$$(1 + \lambda^{-r} (d(y\_o))^{-r})^{-sp} e^{2\pi p \lambda \sigma} \int\_{N(t\_0, \eta')} (N(\mathbf{g}(t)))^p dt \le M^p \tag{9}$$
for all *λ* > 0 with *σ* > 0 being fixed and independent of *λ*. Recall that *y<sup>o</sup>* depends only on *to*. The constants *d*(*yo*),*r*,*s*, *p*, *σ*, *η* 0 , and *M* are all independent of *λ* > 0. We have (1 + *λ* −*r* (*d*(*yo*))−*<sup>r</sup>* ) <sup>−</sup>*sp* = 1 if *r* = 0 or *s* = 0, and (1 + *λ* −*r* (*d*(*yo*))−*<sup>r</sup>* ) <sup>−</sup>*sp* <sup>→</sup> <sup>1</sup> as *<sup>λ</sup>* <sup>→</sup> <sup>∞</sup> if *r* > 0 and *s* > 0. We let *λ* → ∞ in (9) and conclude that **g**(*t*) = Θ almost everywhere in *N*(*to*; *η* 0 ) which contradicts the fact that **g**(*t*) 6= Θ on a set of positive measure in *N*(*to*, *η* 0 ). Thus, **<sup>g</sup>**(*t*) = <sup>Θ</sup> a.e. in {*<sup>t</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* : *<sup>u</sup>C*(*t*) <sup>&</sup>gt; *<sup>A</sup>*}, and supp(**g**) ⊆ {*<sup>t</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* : *uC*(*t*) ≤ *A*} a.e. since {*<sup>t</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* : *<sup>u</sup>C*(*t*) <sup>≤</sup> *<sup>A</sup>*} is a closed set in <sup>R</sup>*<sup>n</sup>* .
The proof of the corresponding result for the growth of Theorem 3 can be obtained by similar techniques as in Theorem 4.
**Theorem 5.** *Let C be an open connected cone in* R*<sup>n</sup> and* 1 ≤ *p* < ∞*. Let g*(*t*) *be a Banach space* <sup>X</sup> *valued measurable function on* <sup>R</sup>*<sup>n</sup> such that the growth of Theorem 3 holds for y* ∈ *C. We have supp*(*g*) ⊆ *C* ∗ *a.e.*
In [7,8], Vladimirov introduced a space of measurable functions on R*<sup>n</sup>* , denoted S 0 0 , which when multiplied by a polynomial raised to a suitable negative power become *L* 2 (R*<sup>n</sup>* ) functions. Analysis concerning the space S 0 0 can also be found in [9,10]. We now extend this space to the vector-valued case and for *p* such that 1 ≤ *p* < ∞. We then show that these new spaces of functions become equivalent to the measurable functions **g** of the preceding results in this section for each *p* and for the base of the tube being open convex cones in R*<sup>n</sup>* . **Definition 5.** *Let* X *be a Banach space.* S 0 *p* (R*<sup>n</sup>* , X ), 1 ≤ *p* < ∞*, is the set of all measurable functions <sup>g</sup>*(*t*), *<sup>t</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup> , with values in* X *such that there exists a real number m* ≥ 0 *for which* (1 + |*t*| *p* ) <sup>−</sup>*mg*(*t*) <sup>∈</sup> *<sup>L</sup> p* (R*<sup>n</sup>* , X )*.*
First note that S 0 *p* (R*<sup>n</sup>* , X ) ⊂ S<sup>0</sup> (R*<sup>n</sup>* , X ), 1 ≤ *p* < ∞. In our first result concerning the spaces S 0 *p* (R*<sup>n</sup>* , X ) the base of the tube *T <sup>C</sup>* will be an open connected cone.
**Theorem 6.** *Let C be an open connected cone in* R*<sup>n</sup> and* 1 ≤ *p* < ∞*. Let g*(*t*) *be a measurable function on* <sup>R</sup>*<sup>n</sup> with values in a Banach space* <sup>X</sup> *such that (1) holds for <sup>y</sup>* <sup>∈</sup> *C. We have <sup>g</sup>* ∈ S<sup>0</sup> *p* (R*<sup>n</sup>* , X ) *and supp*(*g*) ⊆ {*<sup>t</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* : *uC*(*t*) ≤ *A*} *a.e.*
**Proof.** The support property of **g** has been proved in Theorem 4. We now prove that **g** ∈ S<sup>0</sup> *p* (R*<sup>n</sup>* , X ). Choose a fixed point *y<sup>o</sup>* ∈ *pr*(*C*) and put *Y* = {*y* : *y* = *λyo*, 0 < *λ* ≤ 1} ⊂ *C*; choose a fixed compact subcone *C* <sup>0</sup> ⊂⊂ *C* such that *y<sup>o</sup>* ∈ *C* 0 . We have *Y* ⊂ *C* <sup>0</sup> ⊂⊂ *C*. Let *y* ∈ *Y* be arbitrary; using (1) we have
$$\int\_{\mathbb{R}^n} e^{-2\pi p \langle y, t \rangle} (\mathcal{N}(\mathbf{g}(t)))^p dt \le M^p (1 + (d(y))^{-r})^{sp} e^{2\pi p A |y|}, \ y \in \mathbb{C}\_r$$
and hence
$$(d(y))^{rsp} \int\_{\mathbb{R}^n} e^{-2\pi p \langle y, t \rangle} (\mathcal{N}(\mathbf{g}(t)))^p dt \le M^p (1 + (d(y))^r)^{sp} e^{2\pi p A |y|}, \ y \in \mathbb{C}. \tag{10}$$
(10) holds in particular for *y* ∈ *Y* for which |*y*| = *λ*|*yo*| = *λ*, 0 < *λ* ≤ 1, since *y<sup>o</sup>* ∈ *pr*(*C*); and <sup>h</sup>*y*, *<sup>t</sup>*i ≤ |*y*||*t*|, *<sup>t</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* , implies (−|*y*||*t*|) ≤ −h*y*, *<sup>t</sup>*i, *<sup>t</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* . Corresponding to *C* <sup>0</sup> ⊂⊂ *C* we use [7] (p. 6, (1.14)) and obtain *δ* = *δ*(*C* 0 ) > 0 depending only on *C* <sup>0</sup> and not on *y* ∈ *C* 0 such that
$$0 < \delta |y| \le d(y) \le |y|, \ y \in \mathcal{C}' \subset\subset \mathcal{C}.\tag{11}$$
Using (11) and (10), we have
$$(\delta\lambda)^{rsp} \int\_{\mathbb{R}^n} e^{-2\pi p\lambda|t|} (\mathcal{N}(\mathbf{g}(t)))^p dt \le (d(y))^{rsp} \int\_{\mathbb{R}^n} e^{-2\pi p\langle y, t\rangle} (\mathcal{N}(\mathbf{g}(t)))^p dt$$
$$\le M^p (1 + (d(y)))^r)^{sp} e^{2\pi p A|y|} \le M^p (1 + \lambda^r) e^{2\pi p\lambda A} \tag{12}$$
for *y* = *λy<sup>o</sup>* ∈ *Y* ⊂ *C* <sup>0</sup> ⊂⊂ *C*, 0 < *λ* ≤ 1, with *δ* being independent of *C* 0 and hence independent of *y* ∈ *Y* and independent of *λ*, 0 < *λ* ≤ 1. Let *e* > 1 be fixed. Multiply both sides of (12) by *λ* <sup>−</sup>1+*<sup>e</sup>* and integrate the result from (12) over <sup>0</sup> <sup>&</sup>lt; *<sup>λ</sup>* <sup>≤</sup> <sup>1</sup> with respect to *<sup>λ</sup>* to obtain
$$\int\_0^1 \lambda^{-1+\varepsilon} (\delta \lambda)^{\text{rsp}} \int\_{\mathbb{R}^n} e^{-2\pi p \lambda |t|} (\mathcal{N}(\mathbf{g}(t)))^p dt d\lambda \le M^p \int\_0^1 \lambda^{-1+\varepsilon} (1+\lambda^r)^{\text{sp}} e^{2\pi p \lambda A} d\lambda.$$
Now multiply this inequality by *δ* <sup>−</sup>*rsp* and use Fubini's theorem on the left to obtain
$$\int\_{\mathbb{R}^n} (N(\mathbf{g}(t)))^p \int\_0^1 \lambda^{rsp - 1 + \epsilon} e^{-2\pi p\lambda |t|} d\lambda dt \le M^p \delta^{-rsp} \int\_0^1 \lambda^{-1 + \epsilon} (1 + \lambda^r)^{ps} e^{2\pi p\lambda A} d\lambda. \tag{13}$$
We note that all constants *M*, *δ*,*r*,*s*, *p*, *e*, and *A* are independent of *y* = *λy<sup>o</sup>* ∈ *Y* and hence independent of *λ*, 0 < *λ* ≤ 1. Using the change of variable *u* = 2*πpλ*|*t*| in the inner integral on the left of (13) and considering the cases 0 < |*t*| ≤ 1/2*πp* and |*t*| > 1/2*πp* we obtain
$$\int\_0^1 \lambda^{rsp - 1 + \epsilon} e^{-2\pi p\lambda|t|} d\lambda = (2\pi p|t|)^{-rsp - \epsilon} \int\_0^{2\pi p|t|} u^{rsp - 1 + \epsilon} e^{-u} du \ge \tag{14}$$
$$\left\{ \begin{array}{ll} (\varepsilon rp + \varepsilon e)^{-1} (1 + |t|^p)^{-rs - \varepsilon/p} & \text{for} \quad 0 < |t| \le 1/2\pi p\\ (2\pi p)^{-rsp - \varepsilon} \int\_0^1 u^{rsp - 1 + \varepsilon} e^{-u} du (1 + |t|^p)^{-rs - \varepsilon/p} & \text{for} \quad |t| > 1/2\pi p \end{array} \right\}$$
Put
$$K = \min\left\{ (ersp + \epsilon e)^{-1}, (2\pi p)^{-rsp - \epsilon} \int\_0^1 u^{rsp - 1 + \epsilon} e^{-u} du \right\} > 0.$$
From (14), we have
$$\int\_0^1 \lambda^{rsp - 1 + \varepsilon} e^{-2\pi p\lambda |t|} d\lambda \ge \mathcal{K} (1 + |t|^p)^{-rs - \varepsilon/p} \,\_\prime |t| > 0,\tag{15}$$
with this inequality holding also at *t* = 0 by adjusting the constant *K* if needed. Putting (15), which holds for all *<sup>t</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* now, into (13) and recalling *<sup>e</sup>* <sup>&</sup>gt; 1, we have
$$\begin{aligned} &K \int\_{\mathbb{R}^n} (1+|t|^p)^{-r\mathbf{s}-\varepsilon/p} (\mathcal{N}(\mathbf{g}(t)))^p dt \le M^p \delta^{-rsp} \int\_0^1 \lambda^{-1+\varepsilon} (1+\lambda^r)^{ps} e^{2\pi p\lambda A} d\lambda \\ &\le M^p \delta^{-rsp} \mathfrak{D}^{ps} e^{2\pi p\Lambda} \end{aligned}$$
with the right side being a fixed constant. Thus, (1 + |*t*| *p* ) −*rs*/*p*−*e*/*p* 2 **g**(*t*) ∈ *L p* (R*<sup>n</sup>* , X ), and **g** ∈ S<sup>0</sup> *p* (R*<sup>n</sup>* , X ) since (*rs*/*p* + *e*/*p* 2 ) ≥ 0.
We similarly obtain the following result from Theorem 5.
**Theorem 7.** *Let C be an open connected cone in* R*<sup>n</sup> and* 1 ≤ *p* < ∞*. Let g*(*t*) *be a Banach space* <sup>X</sup> *valued measurable function on* <sup>R</sup>*<sup>n</sup> such that the growth of Theorem 3 holds for y* ∈ *C. We have g* ∈ S<sup>0</sup> *p* (R*<sup>n</sup>* , X ) *and supp(g)* ⊆ *C* ∗ *a.e.*
In order for the converse implication of Theorem 6 to hold we need the cone *C* to be convex as well as open.
**Theorem 8.** *Let C be an open convex cone in* R*<sup>n</sup>* , 1 ≤ *p* < ∞*, and A* ≥ 0*. Let g* ∈ S<sup>0</sup> *p* (R*<sup>n</sup>* , X ) *with supp(g*) ⊆ {*<sup>t</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* : *uC*(*t*) ≤ *A*} *a.e. where* X *is a Banach space. We have g is a measurable function with values in* X *such that (1) holds for all y* ∈ *C.*
**Proof.** From [9] (p. 74, Lemma 3), {*<sup>t</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* : *uC*(*t*) ≤ *A*} = *C* <sup>∗</sup> + *N*(0; *A*), *N*(0; *A*) = {*t* ∈ R*n* : <sup>|</sup>*t*<sup>|</sup> <sup>&</sup>lt; *<sup>A</sup>*}, since the cone *<sup>C</sup>* is open and convex here. Thus, *<sup>t</sup>* ∈ {*<sup>t</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* : *uC*(*t*) ≤ *A*} yields *t* = *t*<sup>1</sup> + *t*2, *t*<sup>1</sup> ∈ *C* ∗ , *t*<sup>2</sup> ∈ *N*(0; *A*). Since **g** ∈ S<sup>0</sup> *p* (R*<sup>n</sup>* , <sup>X</sup> ), **<sup>g</sup>** is measurable on <sup>R</sup>*<sup>n</sup>* and (1 + |*t*| *p* ) <sup>−</sup>*m***g**(*t*) <sup>∈</sup> *<sup>L</sup> p* (R*<sup>n</sup>* , X ) for some *m* ≥ 0; thus
$$\int\_{\mathbb{R}^n} (1+|t|^p)^{-mp} (\mathcal{N}(\mathbf{g}(t)))^p dt \le K < \infty$$
for a constant *K* > 0. Let *y* ∈ *C* be arbitrary. We have
$$\begin{split} &\int\_{\mathbb{R}^n} e^{-2\pi p \langle y, t \rangle} (\mathcal{N}(\mathbf{g}(t)))^p dt \\ &= \int\_{\mathbb{C}^n + \overline{N(\overline{0}; A)}} e^{-2\pi p \langle y, t \rangle} (1 + |t|^p)^{mp} (1 + |t|^p)^{-mp} (\mathcal{N}(\mathbf{g}(t))))^p dt \\ &\leq \sup\_{t \in \mathbb{C}^n + \overline{N(\overline{0}; A)}} ((1 + |t|^p)^{mp} e^{-2\pi p \langle y, t \rangle}) \int\_{\mathbb{R}^n} (1 + |t|^p)^{-mp} (\mathcal{N}(\mathbf{g}(t)))^p dt \\ &\leq K \sup\_{t \in \mathbb{C}^n + \overline{N(\overline{0}; A)}} (1 + |t|^p)^{mp} e^{-2\pi p \langle y, t \rangle} \\ &\leq K \sup\_{t\_1 \in \mathbb{C}^n, t\_2 \in \overline{N(\overline{0}; A)}} (1 + (|t\_1| + |t\_2|)^p)^{mp} e^{-2\pi p \langle y, t\_1 + t\_2 \rangle}. \end{split} \tag{16}$$
For *t*<sup>2</sup> ∈ *N*(0; *A*), we have |*t*2| ≤ *A* and
$$e^{-2\pi p\langle y, t\_2\rangle} \le e^{2\pi p|t\_2||y|} \le e^{2\pi p A|y|}, \ t\_2 \in \overline{N(\overline{0}; A)}, \ y \in \mathbb{C}.\tag{17}$$
For *t*<sup>1</sup> ∈ *C* ∗ , we have *t*<sup>1</sup> = *λ*1*t* ∗ <sup>1</sup> where *λ*<sup>1</sup> ≥ 0 and *t* ∗ <sup>1</sup> ∈ *pr*(*C* ∗ ). From Section 2 we have
$$d(y) = \inf\_{u \in pr(\mathbb{C}^\*)} \langle u, y \rangle = -\sup\_{u \in pr(\mathbb{C}^\*)} (-\langle u, y \rangle), \ y \in \mathbb{C}. \tag{18}$$
For *y* ∈ *C*, using (17) and (18) we continue (16) as
$$\begin{split} &\int\_{\mathbb{R}^{n}} e^{-2\pi p \langle y,t\rangle} (\mathcal{N}(\mathbf{g}(t)))^{p} dt \\ &\leq \mathcal{K} e^{2\pi p A |y|} \sup\_{\lambda\_{1} \geq 0, t\_{1}^{n} \in pr(\mathbb{C}^{\*})} \left( (1 + (\lambda\_{1} + A)^{p})^{mp} e^{-2\pi p \lambda\_{1} \langle t\_{1}^{\*} y \rangle} \right) \\ &\leq \mathcal{K} e^{2\pi p A |y|} \sup\_{\lambda\_{1} \geq 0} \left( (1 + (\lambda\_{1} + A)^{p})^{mp} e^{-2\pi p \lambda\_{1} d(y)} \right) \\ &\leq K (1 + (1 + A)^{p})^{mp} e^{2\pi p A |y|} \sup\_{\lambda\_{1} \geq 0} \left( (1 + \lambda\_{1}^{p})^{mp} e^{-2\pi p \lambda\_{1} d(y)} \right) \\ &\leq K (1 + (1 + A)^{p})^{mp} e^{2\pi p A |y|} \sup\_{\lambda\_{1} \geq 0} \left( (1 + \lambda\_{1})^{mp} e^{-2\pi p \lambda\_{1} d(y)} \right). \end{split} \tag{19}$$
The supremum in the last line of (19) is a maximum which can be obtained using the first derivative test. If (*mp*<sup>2</sup> <sup>−</sup> <sup>2</sup>*πpd*(*y*)) <sup>&</sup>gt; <sup>0</sup> then *<sup>m</sup>* <sup>&</sup>gt; <sup>0</sup> and the supremum occurs at *<sup>λ</sup>*<sup>1</sup> = (*mp*<sup>2</sup> <sup>−</sup> <sup>2</sup>*πpd*(*y*))/2*πpd*(*y*), and in this case
$$\begin{split} \sup\_{\lambda\_1 \ge 0} &((1+\lambda\_1)^{mp^2} e^{-2\pi p\lambda\_1 d(y)}) \le \left(1 + \frac{mp^2 - 2\pi pd(y)}{2\pi pd(y)}\right)^{mp^2} \\ &\le \left(1 + \frac{mp^2}{2\pi pd(y)}\right)^{mp^2} = \left(\frac{mp}{2\pi}\right)^{mp^2} \left(\frac{2\pi}{mp} + \frac{1}{d(y)}\right)^{mp^2} \\ &\le \max\{1, \left(\frac{mp}{2\pi}\right)^{mp^2}\} (1 + (d(y))^{-1})^{mp^2}. \end{split}$$
If (*mp*<sup>2</sup> <sup>−</sup> <sup>2</sup>*πpd*(*y*)) <sup>≤</sup> 0, the supremum in the last line of (19) occurs at *<sup>λ</sup>*<sup>1</sup> <sup>=</sup> 0 and
$$\sup\_{\lambda\_1 \ge 0} ((1 + \lambda\_1)^{mp^2} e^{-2\pi p \lambda\_1 d(y)}) = 1 \le (1 + (d(y))^{-1})^{mp^2}.$$
Combining (19) with the above two estimates on the supremum over *λ*<sup>1</sup> ≥ 0 we have for *y* ∈ *C*
$$\begin{aligned} &\int\_{\mathbb{R}^n} e^{-2\pi p \langle y, t \rangle} (\mathcal{N}(\mathbf{g}(t)))^p dt \\ &\leq K (1 + (1 + A)^p)^{mp} \max\{1, \left(\frac{mp}{2\pi}\right)^{mp^2}\} (1 + (d(y))^{-1})^{mp^2} e^{2\pi p A |y|}. \end{aligned}$$
Taking the *p*th root of this inequality, we obtain (1) holding for all *y* ∈ *C* with *r* = 1 and *s* = *mp*.
For *C* being an open convex cone in R*<sup>n</sup>* Theorems 6 and 8 show that **g** being a Banach space X valued measurable function with (1) holding for *y* ∈ *C*, *A* ≥ 0, and 1 ≤ *p* < ∞ is an equivalent statement to **g** ∈ S<sup>0</sup> *p* (R*<sup>n</sup>* , <sup>X</sup> ) with supp(**g**) ⊆ {*<sup>t</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* : *uC*(*t*) ≤ *A*} a.e. for *A* ≥ 0 and 1 ≤ *p* < ∞. Thus, for any future result concerning open convex cones *C*, these two statements are interchangeable in hypotheses.
If *<sup>A</sup>* <sup>=</sup> 0, {*<sup>t</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* : *uC*(*t*) ≤ 0} = *C* ∗ . In this case we have the following corollary to Theorem 8.
**Corollary 4.** *Let C be an open convex cone in* R*<sup>n</sup> and* 1 ≤ *p* < ∞*. Let g* ∈ S<sup>0</sup> *p* (R*<sup>n</sup>* , X ) *for* X *being a Banach space and supp*(*g*) ⊆ *C* ∗ *a.e. We have*
$$|e^{-2\pi \langle y, t \rangle} \mathbf{g}(t)|\_p \le M(1 + (d(y))^{-1})^{mp}, \ y \in \mathbb{C},$$
*for constants M* > 0 *and m* ≥ 0*.*
#### **6. Analytic Functions Generating Measurable Functions**
In this section, we consider generalized vector-valued Hardy functions and construct measurable functions which yield Fourier–Laplace transform representations. This material is followed in Section 7 by representing the analytic functions, in particular cases, by Cauchy and Poisson integrals.
We use the Fourier transform on *L* 2 (R*<sup>n</sup>* , H) considerably in this section and in Section 7. This causes us to restrict the results to *p* = 2 and functions having values in Hilbert space H as previously discussed in Section 2 in relation to the function Fourier transform.
To prove the Fourier–Laplace representation of functions in *H*<sup>2</sup> *A* (*T B* , H) in terms of a constructed measurable function we first need the following lemma.
**Lemma 1.** *Let B be a proper open connected subset of* R*<sup>n</sup> . Let <sup>f</sup>*(*z*) <sup>∈</sup> *<sup>H</sup>*<sup>2</sup> *A* (*T B* , H))*,where* H *is Hilbert space, and be bounded for <sup>x</sup>* <sup>=</sup> *Re*(*z*) <sup>∈</sup> <sup>R</sup>*<sup>n</sup> and y* = *Im*(*z*) *in any compact subset of B. Let e* > 0*. Put*
$$\mathbf{g}\_{\varepsilon,y}(t) = \int\_{\mathbb{R}^n} e^{-\varepsilon \sum\_{j=1}^n z\_j^2} f(\mathbf{x} + iy) e^{-2\pi i \langle \mathbf{x} + iy, t \rangle} d\mathbf{x}, \ y \in \mathcal{B}\_\prime \tag{20}$$
*and*
$$\mathbf{g}\_y(t) = \mathcal{F}^{-1}[e^{2\pi \langle y, t \rangle} f(x + iy); t], \ y \in \mathcal{B}, \ t \in \mathbb{R}^n,\tag{21}$$
*in L*<sup>2</sup> (R*<sup>n</sup>* , H)*. We have ge*,*<sup>y</sup>* (*t*) *is independent of y* ∈ *B for any e* > 0*;*
$$\lim\_{\varepsilon \to 0+} |\mathbf{g}\_{\varepsilon,y}(t) - \mathbf{g}\_y(t)|\_2 = 0, \ y \in B; \tag{22}$$
*and g<sup>y</sup>* (*t*) *is independent of y* ∈ *B.*
**Proof.** For *<sup>y</sup>* <sup>∈</sup> *<sup>B</sup>* and *<sup>t</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* , **f**(*x* + *iy*) ∈ *L* 2 (R*<sup>n</sup>* , H) and *e* 2*π*h*y*,*t*i **f**(*x* + *iy*) ∈ *L* 2 (R*<sup>n</sup>* , H) as functions of *<sup>x</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* . Further, (*e* 2*π*h*y*,*t*i *e* −*e* ∑ *n j*=1 *z* 2 *<sup>j</sup>* **f**(*x* + *iy*)) ∈ *L* 1 (R*<sup>n</sup>* , H) ∩ *L* 2 (R*<sup>n</sup>* , H) for *<sup>y</sup>* <sup>∈</sup> *<sup>B</sup>* and *<sup>t</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* . Thus, both **g***e*,*<sup>y</sup>* (*t*) and **g***<sup>y</sup>* (*t*) are well defined for *y* ∈ *B* and both are in *L* 2 (R*<sup>n</sup>* , H). We assume here that 0 < *e* ≤ 1 since we are letting *e* → 0+ in (22). We have for *y* ∈ *B*
$$|\mathbf{g}\_{\varepsilon,y}(t) - \mathbf{g}\_y(t)|\_2 = |\mathcal{F}^{-1}[e^{2\pi \langle y, t \rangle}(e^{-\varepsilon \sum\_{j=1}^y z\_j^2} - 1)\mathbf{f}(x + iy); t]|\_2$$
$$t = |e^{2\pi \langle y, t \rangle}(e^{-\varepsilon \sum\_{j=1}^y z\_j^2} - 1)\mathbf{f}(x + iy)|\_2. \tag{23}$$
For 0 < *e* ≤ 1
$$\begin{aligned} &\left(\mathcal{N}(e^{2\pi \langle y,t\rangle}(e^{-\epsilon \sum\_{j=1}^n z\_j^2} - 1)\mathbf{f}(x+iy))\right)^2 \\ &= |e^{-\epsilon \sum\_{j=1}^n z\_j^2} - 1|^2 e^{4\pi \langle y,t\rangle} (\mathcal{N}(\mathbf{f}(x+iy)))^2 \\ &\le (|e^{-\epsilon z\_1^2}| \dots |e^{-\epsilon z\_n^2}| + 1)^2 e^{4\pi \langle y,t\rangle} (\mathcal{N}(\mathbf{f}(x+iy)))^2 \\ &\le (e^{|y|^2} + 1)^2 e^{4\pi \langle y,t\rangle} (\mathcal{N}(\mathbf{f}(x+iy)))^2, \end{aligned}$$
and the right side of this inequality is independent of 0 < *e* ≤ 1 and is integrable as a function of *<sup>x</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* . By the Lebesgue dominated convergence theorem (22) follows from (23).
To show that **g***e*,*<sup>y</sup>* (*t*) is independent of *y* ∈ *B* let *S* be any compact subset of *B*, and let *y* ∈ *S* ⊂ *B*. We have
$$|e^{-\epsilon \sum\_{j=1}^{n} z\_j^2}| \le e^{\epsilon n \sigma^2} e^{-\epsilon |x|^2}, \ x \in \mathbb{R}^n, \ y \in \mathbb{S}\_{\epsilon}$$
where *<sup>a</sup>* <sup>=</sup> max*y*∈*S*{|*y*1|, <sup>|</sup>*y*2|, ..., <sup>|</sup>*yn*|}. For *<sup>y</sup>* <sup>∈</sup> *<sup>S</sup>* <sup>⊂</sup> *<sup>B</sup>* and *<sup>t</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>*
$$\int\_{\mathcal{S}} \mathcal{N}(e^{-\varepsilon \sum\_{j=1}^{n} z\_j^2} \mathbf{f}(x+iy)e^{-2\pi i \langle x+iy,t\rangle} dy$$
$$=\int\_{\mathcal{S}} |e^{-\varepsilon \sum\_{j=1}^{n} z\_j^2}| |e^{-2\pi i \langle x+iy,t\rangle}| \mathcal{N}(\mathbf{f}(x+iy)) dy$$
$$\leq A\_{\mathcal{S}} e^{\varepsilon n a^2} e^{-\varepsilon |\mathbf{x}|^2} \int\_{\mathcal{S}} e^{2\pi |y||t|} dy\tag{24}$$
where *<sup>A</sup><sup>S</sup>* is a bound on <sup>N</sup> (**f**(*<sup>x</sup>* <sup>+</sup> *iy*)) for *<sup>x</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* and *<sup>y</sup>* <sup>∈</sup> *<sup>S</sup>*; and the right side of (24) approaches 0 as |*x*| → ∞. An application of the Caucyh-Poincare theorem yields **g***e*,*<sup>y</sup>* is independent of *y* ∈ *S* for any *e* > 0 and hence independent of *y* ∈ *B* for any *e* > 0 since *S* is any arbitrary compact subset of *B*. In the future we refer to **g***e*,*<sup>y</sup>* , *y* ∈ *B*, as **g***<sup>e</sup>* since this function is independent of *y* ∈ *B* for any *e* > 0.
Now to prove that **g***<sup>y</sup>* (*t*) ∈ *L* 2 (R*<sup>n</sup>* , H) is independent of *y* ∈ *B* let *y*<sup>1</sup> and *y*<sup>2</sup> both be points of *B*. Since **g***<sup>e</sup>* = **g***e*,*<sup>y</sup>* is independent of *y* ∈ *B*, for any *e* > 0 we have
$$\begin{aligned} |\mathbf{g}\_{y\_1}(t) - \mathbf{g}\_{y\_2}(t)|\_2 &= |\mathbf{g}\_{y\_1}(t) - \mathbf{g}\_{\varepsilon, y\_1}(t) + \mathbf{g}\_{\varepsilon, y\_2}(t) - \mathbf{g}\_{y\_2}(t)|\_2 \\ &\le |\mathbf{g}\_{y\_1}(t) - \mathbf{g}\_{\varepsilon, y\_1}(t)|\_2 + |\mathbf{g}\_{y\_2}(t) - \mathbf{g}\_{\varepsilon, y\_2}(t)|\_2. \end{aligned} \tag{25}$$
Letting *e* → 0+ in (25) and using (22), the right side of (25) approaches 0 while the left side is independent of *e* > 0. Thus, **g***y*<sup>1</sup> (*t*) = **g***y*<sup>2</sup> (*t*) a.e., *<sup>t</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* , and **g***<sup>y</sup>* (*t*) defined in (21) is independent of *y* ∈ *B*. We write **g***<sup>y</sup>* (*t*) defined in (21) as **<sup>g</sup>**(*t*), *<sup>y</sup>* <sup>∈</sup> *<sup>B</sup>*, *<sup>t</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* , in the future; and recall that **g**(*t*) ∈ *L* 2 (R*<sup>n</sup>* , H).
We obtain a Fourier–Laplace representation of elements in *H*<sup>2</sup> *A* (*T B* , H) now.
**Theorem 9.** *Let B be a proper open connected subset of* R*<sup>n</sup> . Let <sup>f</sup>*(*z*) <sup>∈</sup> *<sup>H</sup>*<sup>2</sup> *A* (*T B* , H)*, where* H *is Hilbert space, and be bounded for <sup>x</sup>* <sup>=</sup> *Re*(*z*) <sup>∈</sup> <sup>R</sup>*<sup>n</sup> and y* = *Im*(*z*) *in any compact subset of B. There is a measurable function g*(*t*) ∈ *L* 2 (R*<sup>n</sup>* , H) *for which*
$$|e^{-2\pi \langle y, t \rangle} \mathbf{g}(t)|\_2 \le M(1 + (d(y))^{-r})^s e^{2\pi A|y|}, \ y \in \mathcal{B},\tag{26}$$
*where r* ≥ 0, *s* ≥ 0, *A* ≥ 0*, and M* = *M*(*g*,*r*,*s*, *A*) > 0 *are independent of y* ∈ *B; and*
$$f(z) = \int\_{\mathbb{R}^n} \mathbf{g}(t) e^{2\pi i \langle z, t \rangle} dt, \; z \in T^B. \tag{27}$$
**Proof.** From Lemma 1 the function **g**(*t*) = **g***<sup>y</sup>* (*t*) defined in (21) is independent of *y* ∈ *B* and is in *L* 2 (R*<sup>n</sup>* , H). From (21)
$$e^{-2\pi \langle y, t \rangle} \mathbf{g}(t) = \mathcal{F}^{-1}[\mathbf{f}(x + iy); t], \ y \in \mathcal{B},\tag{28}$$
and by the Parseval equality
$$|e^{-2\pi \langle y, t \rangle} \mathbf{g}(t)|\_2 = |\mathbf{f}(\mathfrak{x} + iy)|\_{2\prime} \ y \in \mathcal{B}\_{\prime}$$
where *e* <sup>−</sup>2*π*h*y*,*t*i**g**(*t*) <sup>∈</sup> *<sup>L</sup>* 2 (R*<sup>n</sup>* , H), *y* ∈ *B*. Thus, (26) holds from the norm growth on **<sup>f</sup>**(*z*) <sup>∈</sup> *<sup>H</sup>*<sup>2</sup> *A* (*T B* , H). Using the now obtained Equation (26), by the proof of Theorem 2 for *p* = 2 we have *e* <sup>−</sup>2*π*h*y*,*t*i**g**(*t*) <sup>∈</sup> *<sup>L</sup>* 1 (R*<sup>n</sup>* , H) ∩ *L* 2 (R*<sup>n</sup>* , H), *y* ∈ *B*, and
$$\int\_{\mathbb{R}^n} \mathbf{g}(t) e^{2\pi i \langle z, t \rangle} dt = \mathcal{F}[e^{-2\pi \langle y, t \rangle} \mathbf{g}(t); x], \; z = x + iy \in T^B\_{\lambda}$$
is analytic in *T <sup>B</sup>* with the Fourier transform being the *L* 1 (R*<sup>n</sup>* , H) transform. Thus, from (28),
$$\mathbf{f}(z) = \mathcal{F}[e^{-2\pi \langle y, t \rangle} \mathbf{g}(t); \mathbf{x}] = \int\_{\mathbb{R}^n} \mathbf{g}(t) e^{2\pi i \langle z, t \rangle} dt, \ z = \mathbf{x} + i\mathbf{y} \in T^{\mathbb{B}},$$
with the Fourier transform being in both the *L* 1 (R*<sup>n</sup>* , H) and *L* 2 (R*<sup>n</sup>* , H) sense, and (27) is obtained.
The structure of the proofs of Lemma 1 and Theorem 9 can be used to prove a result like Theorem 9 for functions in *H*<sup>2</sup> ∗ (*T B* , H); we state this result now.
**Theorem 10.** *Let <sup>B</sup> be an open connected subset of* <sup>R</sup>*<sup>n</sup> which does not contain* <sup>0</sup> <sup>∈</sup> <sup>R</sup>*<sup>n</sup> . Let <sup>f</sup>*(*z*) <sup>∈</sup> *<sup>H</sup>*<sup>2</sup> ∗ (*T B* , <sup>H</sup>)*, where* <sup>H</sup> *is Hilbert space, and be bounded for <sup>x</sup>* <sup>=</sup> *Re*(*z*) <sup>∈</sup> <sup>R</sup>*<sup>n</sup> and y* = *Im*(*z*) *in any compact subset of B. There is a measurable function g*(*t*) ∈ *L* 2 (R*<sup>n</sup>* , H) *for which*
$$|e^{-2\pi \langle y, t \rangle} \mathbf{g}(t)|\_2 \le M(1 + (d(y))^{-r})^s e^{M^\*(w/|y|)}, \ y \in \mathcal{B}\_r$$
*where r* ≥ 0,*s* ≥ 0, *w* > 0*, andM* = *M*(*g*,*r*,*s*, *w*) > 0 *are independent of y* ∈ *B; and*
$$f(z) = \int\_{\mathbb{R}^n} \mathbf{g}(t) e^{2\pi i \langle z, t \rangle} dt, \ z \in T^B.$$
By restricting the base *B* in Theorem 9, further information is obtained.
**Corollary 5.** *Let C be an open connected cone in* R*<sup>n</sup> . Let <sup>f</sup>*(*z*) <sup>∈</sup> *<sup>H</sup>*<sup>2</sup> *A* (*T <sup>C</sup>*, <sup>H</sup>)*, where* <sup>H</sup> *is Hilbert space, and be bounded for <sup>x</sup>* <sup>=</sup> *Re*(*z*) <sup>∈</sup> <sup>R</sup>*<sup>n</sup> and y* = *Im*(*z*) *in any compact subset of C. There is a measurable function g* ∈ *L* 2 (R*<sup>n</sup>* , H) ∩ *S* 0 2 (R*<sup>n</sup>* , <sup>H</sup>) *with supp*(*g*) ⊆ {*<sup>t</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* : *uC*(*t*) ≤ *A*} *a.e. such that (26) and (27) hold. Further, if C is an open convex cone in* R*<sup>n</sup> we have*
$$\lim\_{y \to \overline{0}, y \in \mathbb{C}} |f(x + iy) - \mathcal{F}[\mathbf{g}(t); \mathbf{x}]|\_2 = 0,\tag{29}$$
*and*
$$\lim\_{y \to \overline{0}, y \in \mathbb{C}} f(x + iy) = \mathcal{F}[\mathbf{g}(t); x] \tag{30}$$
*in the strong topology of* S 0 (R*<sup>n</sup>* , H)*.*
**Proof.** The existence of **g** ∈ *L* 2 (R*<sup>n</sup>* , H) such that (26) and (27) hold follow from Theorem 9. The facts that **g** ∈ S<sup>0</sup> 2 (R*<sup>n</sup>* , <sup>H</sup>) with supp(**g**) ⊆ {*<sup>t</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* : *uC*(*t*) ≤ *A*} a.e. now follow by Theorem 6. Let us further assume that the cone *C* is open and convex. From the proof of Theorem <sup>8</sup> we know that {*<sup>t</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* : *uC*(*t*) ≤ *A*} = *C* ∗ + *N*(0; *A*) where *C* ∗ is the dual cone of *<sup>C</sup>* and *<sup>N</sup>*(0; *<sup>A</sup>*) = {*<sup>t</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* : |*t*| < *A*} since *C* is assumed to be convex now. Thus, *<sup>t</sup>* ∈ {*<sup>t</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* : *uC*(*t*) ≤ *A*} yields *t* = *t*<sup>1</sup> + *t*2, *t*<sup>1</sup> ∈ *C* ∗ , *t*<sup>2</sup> ∈ *N*(0; *A*) as in the proof of Theorem 8. Returning to the proof of Theorem 9 we have for *y* ∈ *C*
$$|\mathbf{f}(\mathbf{x} + i\mathbf{y}) - \mathcal{F}[\mathbf{g}(t); \mathbf{x}]|\_2 = |\mathcal{F}[e^{-2\pi\langle y, t\rangle}\mathbf{g}(t); \mathbf{x}] - \mathcal{F}[\mathbf{g}(t); \mathbf{x}]|\_2$$
$$= |\mathcal{F}[(e^{-2\pi\langle y, t\rangle} - 1)\mathbf{g}(t); \mathbf{x}]|\_2 = |(e^{-2\pi\langle y, t\rangle} - 1)\mathbf{g}(t)|\_2. \tag{31}$$
In (29) and (30), we prove limit properties as *y* → 0, *y* ∈ *C*; so we assume that |*y*| ≤ 1, *y* ∈ *C*, in the remainder of this proof. For *t* = *t*<sup>1</sup> + *t*<sup>2</sup> ∈ *C* ∗ + *N*(0; *A*) we have
$$\begin{aligned} &(\mathcal{N}((e^{-2\pi\langle y,t\rangle}-1)\mathbf{g}(t)))^2 = |e^{-2\pi\langle y,t\rangle} - 1|^2 (\mathcal{N}(\mathbf{g}(t)))^2 \\ &\leq (e^{-2\pi\langle y,t\rangle} + 1)^2 (\mathcal{N}(\mathbf{g}(t)))^2 = (e^{-2\pi\langle y,t\_1\rangle} e^{-2\pi\langle y,t\_2\rangle} + 1)^2 (\mathcal{N}(\mathbf{g}(t)))^2 \\ &\leq (1 + e^{2\pi A})^2 (\mathcal{N}(\mathbf{g}(t)))^2 \end{aligned} \tag{32}$$
for |*y*| ≤ 1, *y* ∈ *C*, where h*y*, *t*1i ≥ 0, *y* ∈ *C* and *t*<sup>1</sup> ∈ *C* ∗ , and |*t*2| ≤ *A* for *t*<sup>2</sup> ∈ *N*(0; *A*). Since **g** ∈ *L* 2 (R*<sup>n</sup>* , H) and supp(**g**) ⊆ *C* ∗ + *N*(0; *A*), (32) and the Lebesgue dominated convergence theorem combined with (31) prove (29). For (30), let *<sup>φ</sup>* ∈ S(R*<sup>n</sup>* ). Using the H*o*¨lder inequality we have
$$\begin{aligned} &\mathcal{N}(\langle \mathbf{f}(\mathbf{x}+i\mathbf{y}), \boldsymbol{\phi}(\mathbf{x}) \rangle - \langle \mathcal{F}[\mathbf{g}(t); \mathbf{x}], \boldsymbol{\phi}(\mathbf{x}) \rangle \\ &\leq \int\_{\mathbb{R}^n} \mathcal{N}((\mathbf{f}(\mathbf{x}+i\mathbf{y}) - \mathcal{F}[\mathbf{g}(t); \mathbf{x}]) \boldsymbol{\phi}(\mathbf{x})) d\mathbf{x} \\ &\leq |\mathbf{f}(\mathbf{x}+i\mathbf{y}) - \mathcal{F}[\mathbf{g}(t); \mathbf{x}]|\_{2} ||\boldsymbol{\phi}||\_{L^{2}(\mathbb{R}^n)'} \end{aligned}$$
and the use of (29) now shows (30) in the weak topology of S 0 (R*<sup>n</sup>* , <sup>H</sup>). But <sup>S</sup>(R*<sup>n</sup>* ) is a Montel space; thus, (30) also holds in the strong topology of S 0 (R*<sup>n</sup>* , H).
We now desire a converse result to Corollary 5 in the setting of tubes *T <sup>C</sup>* where *C* is an open connected cone in R*<sup>n</sup>* .
**Corollary 6.** *Let C be an open connected cone in* R*<sup>n</sup> and* H *be a Hilbert space. Let g*(*t*) *be <sup>a</sup>* <sup>H</sup> *valued measurable function on* <sup>R</sup>*<sup>n</sup> such that* (26) *holds. We have g* ∈ S<sup>0</sup> 2 (R*<sup>n</sup>* , H) *with supp*(*g*) ⊆ {*<sup>t</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* : *<sup>u</sup>C*(*t*) <sup>≤</sup> *<sup>A</sup>*} *a.e., and <sup>f</sup>*(*z*) <sup>∈</sup> *<sup>H</sup>*<sup>2</sup> *A* (*T <sup>C</sup>*, <sup>H</sup>) *for <sup>f</sup>*(*z*)*defined as in* (27) *for z* ∈ *T <sup>C</sup>. Further, if C is an open convex cone in* R*<sup>n</sup> we have* (30) *holding in the strong topology of* S 0 (R*<sup>n</sup>* , H)*.*
**Proof.** We apply Theorem 6 and Corollary 1 to obtain **g** ∈ S<sup>0</sup> 2 (R*<sup>n</sup>* , H) with supp(**g**) ⊆ {*<sup>t</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* : *uC*(*t*) ≤ *A*} a.e. and to obtain that **f**(*z*) defined as in (27) for *z* ∈ *T <sup>C</sup>* is an element of *H*<sup>2</sup> *A* (*T <sup>C</sup>*, <sup>H</sup>). Now assume that *<sup>C</sup>* is an open convex cone in the remainder of this proof to obtain (30) here. Since **g** ∈ S<sup>0</sup> 2 (R*<sup>n</sup>* , H) ⊂ S<sup>0</sup> (R*<sup>n</sup>* , H), the Fourier transform F[**g**] is well defined in S 0 (R*<sup>n</sup>* , H). From the proof of Corollary 1 we have *e* <sup>−</sup>2*π*h*y*,*t*i**g**(*t*) <sup>∈</sup> *L* 1 (R*<sup>n</sup>* , H) ∩ *L* 2 (**R** *n* , H) for *y* ∈ *C*. Thus, **f**(*x* + *iy*) = F[*e* <sup>−</sup>2*π*h*y*,*t*i**g**(*t*); *<sup>x</sup>*], *<sup>y</sup>* <sup>∈</sup> *<sup>C</sup>*, with the Fourier transform being in the *L* 1 (R*<sup>n</sup>* , H), the *L* 2 (R*<sup>n</sup>* , H), and the S 0 (R*<sup>n</sup>* , H) cases. Recalling that supp(**g**) ⊆ {*t* ∈ **R** *n* : *uC*(*t*) ≤ *A*} a.e. and referring to [9] (p. 119), we choose a function *λ*(*t*) ∈ *C* <sup>∞</sup>, *<sup>t</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* , such that for any n-tuple *α* of nonnegative integers <sup>|</sup>*Dαλ*(*t*)| ≤ *<sup>M</sup>α*, *<sup>t</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* , where *M<sup>α</sup>* is a constant which depends only on *α*; and for *<sup>e</sup>* <sup>&</sup>gt; 0, *<sup>λ</sup>*(*t*) = <sup>1</sup> for *<sup>t</sup>* on an *<sup>e</sup>* neighborhood of {*<sup>t</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* : *uC*(*t*) ≤ *A*}, and *λ*(*t*) = 0 for *<sup>t</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* but not on a <sup>2</sup>*<sup>e</sup>* neighborhood of {*<sup>t</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* : *<sup>u</sup>C*(*t*) <sup>≤</sup> *<sup>A</sup>*}. For *<sup>φ</sup>* ∈ S(R*<sup>n</sup>* ) we have for *y* ∈ *C*
$$
\langle \langle \mathbf{f}(\mathbf{x} + iy), \boldsymbol{\phi}(\mathbf{x}) \rangle \rangle = \langle \mathcal{F}[e^{-2\pi \langle y, t \rangle} \mathbf{g}(t); \mathbf{x}], \boldsymbol{\phi}(\mathbf{x}) \rangle \\
= \langle \boldsymbol{\lambda}(t) e^{-2\pi \langle y, t \rangle} \mathbf{g}(t), \mathcal{F}[\boldsymbol{\phi}(\mathbf{x}); t] \rangle.
$$
For *C* being convex we apply [9] (p. 74, Lemma 3) as in our proof of Theorem 8 to obtin {*<sup>t</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* : *uC*(*t*) ≤ *A*} = *C* ∗ + *N*(0; *A*). The result (30) in this corollary now follows from the above equality, *<sup>φ</sup>* ∈ S(R*<sup>n</sup>* ), by the same analysis in [9] (p. 119, lines 2–22) in the weak topology of S 0 (R*<sup>n</sup>* , H) as *y* → 0, *y* ∈ *C*; and the weak topology implies the strong topology of S 0 (R*<sup>n</sup>* , H) as in the proof of (30) in Corollary 5. The proof is complete.
Note that we can not say that **g** ∈ *L* 2 (R*<sup>n</sup>* , H) in Corollary 6 and hence can not obtain the convergence (29) in this converse of Corollary 5.
For *<sup>B</sup>* being a proper open connected subset of <sup>R</sup>*<sup>n</sup>* and <sup>X</sup> being a Banach space, the spaces *V p A* (*T B* , X ) follow as subspaces of *H p A* (*T B* , X ) (or appropriately of *R p A* (*T B* , X )) by letting either *r* = 0 or *s* = 0 in the norm growth defining these other spaces. Thus, Theorem 9 holds for **f**(*z*) ∈ *V* 2 *A* (*T B* , H); and by the proof of Theorem 9, (26) will hold for the obtained function **g** in the form
$$|e^{-2\pi \langle y, t \rangle} \mathbf{g}(t)|\_2 \le e^{2\pi A|y|}, \ y \in B.$$
Using the same proof as in Corollary 2 we then can extend the norm growth on *e* <sup>−</sup>2*π*h*y*,*t*i**g**(*t*) to hold for *<sup>y</sup>* <sup>∈</sup> *<sup>B</sup>*. This is stated in the following corollary to Theorem 9.
**Corollary 7.** *Let B be a proper open connected subset of* R*<sup>n</sup> . Let f*(*z*) ∈ *V* 2 *A* (*T B* , H)*, where* H *is Hilbert space, and be bounded for <sup>x</sup>* <sup>=</sup> *Re*(*z*) <sup>∈</sup> <sup>R</sup>*<sup>n</sup> and y* = *Im*(*z*) *in any compact subset of B. There is a measurable function g*(*t*) ∈ *L* 2 (R*<sup>n</sup>* , H) *for which*
$$|e^{-2\pi \langle y, t \rangle} \mathbf{g}(t)|\_2 \le Me^{2\pi A|y|}, \ y \in \overline{B}\_\tau$$
*where A* ≥ 0 *and M* = *M*(*g*, *A*) > 0 *are independent of y* ∈ *B; and*
$$f(z) = \int\_{\mathbb{R}^n} \mathbf{g}(t) e^{2\pi i \langle z, t \rangle} dt, \ z \in T^B.$$
For the base of the tube being an open connected cone in R*<sup>n</sup>* we have the following corollary of Theorem 10 by combining Theorems 7 and 10. The limit properties in the following corollary will hold for *C* being an open connected cone in R*<sup>n</sup>* by similar techniques as in the proof of Corollary 5; *C* does not need to be convex here for these limit properties to hold because the support of **g** is in *C* ∗ .
**Corollary 8.** *Let C be an open connected cone in* R*<sup>n</sup> . Let <sup>f</sup>*(*z*) <sup>∈</sup> *<sup>H</sup>*<sup>2</sup> ∗ (*T <sup>C</sup>*, <sup>H</sup>)*, where* <sup>H</sup> *is Hilbert space, and be bounded for <sup>x</sup>* <sup>=</sup> *Re*(*z*) <sup>∈</sup> <sup>R</sup>*<sup>n</sup> and y* = *Im*(*z*) *in any compact subset of C. There is a measurable function g*(*t*) ∈ *L* 2 (R*<sup>n</sup>* , H) ∩ S<sup>0</sup> 2 (R*<sup>n</sup>* , H) *with supp*(*g*) ⊆ *C* ∗ *a.e. such that the norm inequality for e* <sup>−</sup>2*π*h*y*,*t*i*g*(*t*) *and the representation of f*(*z*) *hold as in the conclusions of Theorem 10. Further we have*
$$\lim\_{y \to \overline{0}, y \in \mathbb{C}} |f(x+iy) - \mathcal{F}[\mathbf{g}(t); \mathbf{x}]|\_2 = 0$$
*and*
$$\lim\_{y \to \overline{0}, y \in \mathbb{C}} f(x + iy) = \mathcal{F}[\mathbf{g}(t); x]$$
*in the strong topology of* S 0 (R*<sup>n</sup>* , H)*.*
#### **7. Subsets of** *H***<sup>2</sup>** (*T <sup>C</sup>***,** <sup>H</sup>)
Let *C* be an open connected cone in R*<sup>n</sup>* , and 1 ≤ *p* < ∞. Let **g**(*t*) be a measurable function on <sup>R</sup>*<sup>n</sup>* with values in a Banach space <sup>X</sup> such that
$$|e^{-2\pi \langle y, t \rangle} \mathbf{g}(t)|\_p \le M(1 + (d(y))^{-r})^s e^{2\pi A|y|}, \ y \in \mathbb{C},\tag{33}$$
where *A* ≥ 0, *r* ≥ 0, *s* ≥ 0, and *M* = *M*(**g**, *p*,*r*,*s*, *A*) > 0, or
$$|e^{-2\pi \langle y, t \rangle} \mathbf{g}(t)|\_p \le M(1 + (d(y))^{-r})^s e^{M^\*(w/|y|)}, \ y \in \mathbb{C}.\tag{34}$$
where *w* > 0, *r* ≥ 0, *s* ≥ 0, and *M* = *M*(**g**, *p*, *w*,*r*,*s*) > 0 with all constants being independent of *<sup>y</sup>* <sup>∈</sup> *<sup>C</sup>*. We have from Theorems <sup>4</sup> and <sup>5</sup> that supp(**g**) ⊆ {*<sup>t</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* : *uC*(*t*) ≤ *A*} a.e. and supp(**g**) ⊆ *C* <sup>∗</sup> a.e. respectively. Restricting to *p* = 2 and letting X = H, a Hilbert space, now we have from Corollarys 1 and 3 that the function
$$\mathbf{f}(z) = \int\_{\mathbb{R}^n} \mathbf{g}(t) e^{2\pi i \langle z, t \rangle} dt, \ z \in T^{\mathbb{C}},$$
is an element of *H*<sup>2</sup> *A* (*T <sup>C</sup>*, <sup>H</sup>) or *<sup>H</sup>*<sup>2</sup> ∗ (*T <sup>C</sup>*, <sup>H</sup>), respectively. Conversely, we have proved in Corollary <sup>5</sup> or Corollary <sup>8</sup> that if **<sup>f</sup>**(*z*) <sup>∈</sup> *<sup>H</sup>*<sup>2</sup> *A* (*T <sup>C</sup>*, <sup>H</sup>) or **<sup>f</sup>**(*z*) <sup>∈</sup> *<sup>H</sup>*<sup>2</sup> ∗ (*T <sup>C</sup>*, <sup>H</sup>) and in each case **f**(*z*) is bounded for *x* = Re(*z*) and *y* = Im(*z*) in any compact subset of *C* then in each case there exists a measurable function **g** ∈ *L* 2 (R*<sup>n</sup>* , H) ∩ S<sup>0</sup> 2 (R*<sup>n</sup>* , <sup>H</sup>) with supp(**g**) ⊆ {*<sup>t</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* : *uC*(*t*) ≤ *A*} a.e. and (33) holds for *p* = 2 or supp(**g**) ⊆ *C* ∗ a.e. and (34) holds for *p* = 2 with
$$\mathbf{f}(z) = \int\_{\mathbb{R}^n} \mathbf{g}(t) e^{2\pi i \langle z, t \rangle} dt, \ z \in T^{\mathbb{C}}.$$
in each case.
We will now show from these results that both spaces *H*<sup>2</sup> 0 (*T <sup>C</sup>*, <sup>H</sup>), *<sup>A</sup>* <sup>=</sup> 0, and *H*<sup>2</sup> ∗ (*T <sup>C</sup>*, <sup>H</sup>) are subsets of the Hardy space *<sup>H</sup>*<sup>2</sup> (*T <sup>C</sup>*, <sup>H</sup>) and obtain immediate results from these subset properties.
**Theorem 11.** *Let C be an open connected cone in* R*<sup>n</sup> and* H *be a Hilbert space. Let f*(*z*) ∈ *H*<sup>2</sup> 0 (*T <sup>C</sup>*, <sup>H</sup>) *or <sup>f</sup>*(*z*) <sup>∈</sup> *<sup>H</sup>*<sup>2</sup> ∗ (*T <sup>C</sup>*, <sup>H</sup>) *and in either case be bounded for <sup>x</sup>* <sup>=</sup> *Re*(*z*) <sup>∈</sup> <sup>R</sup>*<sup>n</sup> and y* = *Im*(*z*) *in any compact subset of C. In either case there is a measurable function g*(*t*) ∈ *L* 2 (R*<sup>n</sup>* , H) ∩ S 0 2 (R*<sup>n</sup>* , H) *with supp*(*g*) ⊆ *C* ∗ *a.e. such that*
$$f(z) = \int\_{\mathbb{R}^n} \mathbf{g}(t) e^{2\pi i \langle z, t \rangle} dt, \ z \in T^{\mathbb{C}};$$
$$\sup\_{y \in \mathbb{C}} |f(\mathbf{x} + iy)|\_2 = \sup\_{y \in \mathbb{C}} |e^{-2\pi \langle y, t \rangle} \mathbf{g}(t)|\_2 = |\mathbf{g}|\_2 \mathbf{x}$$
*and <sup>f</sup>*(*z*) <sup>∈</sup> *<sup>H</sup>*<sup>2</sup> (*T <sup>C</sup>*, <sup>H</sup>)*.*
**Proof.** As noted previously in this section a function **g** ∈ *L* 2 (R*<sup>n</sup>* , H)∩ S<sup>0</sup> 2 (R*<sup>n</sup>* , H) is obtained from previous results such that
$$\mathbf{f}(z) = \int\_{\mathbb{R}^n} \mathbf{g}(t) e^{2\pi i \langle z, t \rangle} dt, \ z \in T^{\mathbb{C}}.$$
Further from the analysis leading to Corollarys 5 and 8 we know *e* <sup>−</sup>2*π*h*y*,*t*i**g**(*t*) <sup>∈</sup> *L* 1 (R*<sup>n</sup>* , H) ∩ *L* 2 (R*<sup>n</sup>* , <sup>H</sup>), *<sup>y</sup>* <sup>∈</sup> *<sup>C</sup>*, in both cases. If *<sup>A</sup>* <sup>=</sup> 0, {*<sup>t</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* : *uC*(*t*) ≤ 0} = *C* ∗ ; thus, in both cases supp(**g**) ⊆ *C* ∗ a.e. In both cases we have
$$|\mathbf{f}(\mathbf{x} + iy)|\_2 = |e^{-2\pi \langle y, t \rangle} \mathbf{g}(t)|\_{2\prime} \ y \in \mathbb{C}.$$
In both cases
$$\int\_{\mathbb{R}^n} (\mathcal{N}(e^{-2\pi \langle y, t \rangle} \mathbf{g}(t)))^2 dt = \int\_{\mathbb{C}^\*} (e^{-4\pi \langle y, t \rangle} (\mathcal{N}(\mathbf{g}(t))))^2 dt \le \int\_{\mathbb{C}^\*} (\mathcal{N}(\mathbf{g}(t)))^2 dt = |\mathbf{g}|\_2^2$$
for all *y* ∈ *C*. We thus have for all *y* ∈ *C*
$$|\mathbf{f}(\mathbf{x} + iy)|\_2 = |e^{-2\pi \langle y, t \rangle} \mathbf{g}(t)|\_2 \le |\mathbf{g}|\_{2\prime} \ y \in \mathbb{C}\_{\prime}$$
which yields **<sup>f</sup>**(*<sup>x</sup>* <sup>+</sup> *iy*) <sup>∈</sup> *<sup>H</sup>*<sup>2</sup> (*T <sup>C</sup>*, <sup>H</sup>). Further,
$$\sup\_{y \in \mathbb{C}} |\mathbf{f}(\mathbf{x} + iy)|\_2 = \sup\_{y \in \mathbb{C}} |e^{-2\pi \langle y, t \rangle} \mathbf{g}(t)|\_2 \le \left( \int\_{\mathbb{C}^\*} (\mathcal{N}(\mathbf{g}(t)))^2 dt \right)^{1/2} = |\mathbf{g}|\_2. \tag{35}$$
But 0 ∈ *C* <sup>∗</sup> <sup>=</sup> {*<sup>t</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* : h*t*, *y*i ≥ 0 for all *y* ∈ *C*}. Hence, the inequality in (35) is an equality.
Because of this result we have immediate consequences for **f**(*x* + *iy*) in either space in Theorem 11 from previously proven results. If *C* is an open convex cone in R*<sup>n</sup>* which contains an entire straight line then **f**(*z*) = Θ, *z* ∈ *T <sup>C</sup>*, for both cases of **f**(*z*) in Theorem 11. If *C* is a regular cone in R*<sup>n</sup>* then
$$\mathbf{f}(z) = \int\_{\mathbb{R}^n} \mathcal{F}[\mathbf{g}(u); t] \mathbf{K}(z - t) dt = \int\_{\mathbb{R}^n} \mathcal{F}[\mathbf{g}(u); t] \mathbf{Q}(z; t) dt, \; z \in \mathbf{T}^{\mathbb{C}},$$
for the function **g**(*t*) in Theorem 11 and for both cases of **f**(*z*) in Theorem 11. Further, we note that Vindas has proved using functional analysis techniques in [1] that for *C* being a regular cone in <sup>R</sup>*<sup>n</sup>* and <sup>X</sup> being a dual Banach space having the Radon-Nikodým property, any **<sup>f</sup>**(*z*) <sup>∈</sup> *<sup>H</sup><sup>p</sup>* (*T <sup>C</sup>*, <sup>X</sup> ), 1 <sup>≤</sup> *<sup>p</sup>* <sup>≤</sup> <sup>∞</sup>, is the Poisson integral of some **<sup>h</sup>** <sup>∈</sup> *<sup>L</sup> p* (R*<sup>n</sup>* , X ), 1 ≤ *p* ≤ ∞. We say more about the use of functional analysis techniques in obtaining results corresponding to those of this paper and those of [1] in Section 9 below.
#### **8. Boundary Values on the Topological Boundary**
In Corollary 5 we obtained boundary value properties of *H*<sup>2</sup> *A* (*T <sup>C</sup>*, <sup>H</sup>) functions on the distinguished boundary of the tube *T <sup>C</sup>* where *C* is an open convex cone in R*<sup>n</sup>* . The boundary values were obtained in the *L* 2 (R*<sup>n</sup>* , H) and S 0 (R*<sup>n</sup>* , H) topologies. We now investigate boundary value properties of a subset of *H*<sup>2</sup> *A* (*T <sup>C</sup>*, <sup>H</sup>) on the topological boundary of the tube.
Our basic result in this section depends on the cone *C* being regular. We consider the subset *R* 2 *A* (*T <sup>C</sup>*, <sup>H</sup>) of *<sup>H</sup>*<sup>2</sup> *A* (*T <sup>C</sup>*, <sup>H</sup>) consisting of analytic functions **<sup>f</sup>**(*z*) in *<sup>T</sup> <sup>C</sup>* with values in H such that
$$|\mathbf{f}(\mathbf{x} + iy)|\_2 \le M(1 + |y|^{-r})^s e^{2\pi A|y|}, \ y \in \mathbb{C},\tag{36}$$
where *A* ≥ 0, *r* ≥ 0, *s* ≥ 0, and *M* = *M*(**f**, *A*,*r*,*s*) > 0 are all independent of *y* ∈ *C*. We prove that *R* 2 *A* (*T <sup>C</sup>*, <sup>H</sup>) functions have boundary values on the topological boundary of *T <sup>C</sup>* again in the *L* 2 (R*<sup>n</sup>* , H) and S 0 (R*<sup>n</sup>* , H) topologies. We have *R* 2 *A* (*T <sup>C</sup>*, <sup>H</sup>) <sup>⊆</sup> *<sup>H</sup>*<sup>2</sup> *A* (*T <sup>C</sup>*, <sup>H</sup>) since <sup>0</sup> <sup>&</sup>lt; *<sup>d</sup>*(*y*) ≤ |*y*<sup>|</sup> for *<sup>y</sup>* in any open connected cone in <sup>R</sup>*<sup>n</sup>* from [2] (p. 6, (1.14)); recall Section 2 above.
Before proving our main result in this section we focus on the growth bound as in (36). If we had used this growth bound of (36) in the inequality (1) for *e* <sup>−</sup>2*π*h*y*,*t*i**g**(*t*) and in the inequality for |**f**(*x* + *iy*)|*<sup>p</sup>* which defines *H p A* (*T B* , X ), that is if we replace *d*(*y*) by |*y*| in the growth bound, then the results, proofs, and conclusions from Theorem 2 through Theorem 11 in Sections 5–7 will all hold as before. In any conclusion in these results that contains the growth bound, the growth bound in the conclusion will be that of (36). We state this to emphasize the content of our proofs in this section which deal with *R* 2 *A* (*T <sup>C</sup>*, <sup>H</sup>) instead of *H*<sup>2</sup> *A* (*T <sup>C</sup>*, <sup>H</sup>).
**Theorem 12.** *Let C be a regular cone in* R*<sup>n</sup> . Let f*(*z*) ∈ *R* 2 *A* (*T <sup>C</sup>*, <sup>H</sup>) *and be bounded for <sup>x</sup>* <sup>=</sup> *Re*(*z*) <sup>∈</sup> <sup>R</sup>*<sup>n</sup> and y* = *Im*(*z*) *in any compact subset of C. Let y<sup>o</sup>* ∈ *∂C*, *y<sup>o</sup>* 6= 0*. There exists a function F*(*x* + *iyo*) ∈ *L* 2 (R*<sup>n</sup>* , H) *such that*
$$\lim\_{y \to y\_o} |f(\mathbf{x} + i\mathbf{y}) - \mathbf{F}(\mathbf{x} + iy\_o)|\_2 = 0 \tag{37}$$
*for y* ∈ {*y* ∈ *C* : 0 < *a* < |*y*| < *b*} *where a and b are any constants such that* 0 < *a* < |*yo*| < *b; and*
$$\lim\_{y \to y\_o} f(x + iy) = F(x + iy\_o) \tag{38}$$
*in the strong topology of* S 0 (R*<sup>n</sup>* , H) *with y* ∈ {*y* ∈ *C* : 0 < *a* < |*y*| < *b*} *again where a and b are any constants such that* 0 < *a* < |*yo*| < *b.*
**Proof.** As noted previously the growth (36) for *R* 2 *A* (*T <sup>C</sup>*, <sup>H</sup>) functions is a special case of the growth for *H*<sup>2</sup> *A* (*T <sup>C</sup>*, <sup>H</sup>) functions since <sup>0</sup> <sup>&</sup>lt; *<sup>d</sup>*(*y*) ≤ |*y*|, *<sup>y</sup>* <sup>∈</sup> *<sup>C</sup>*. Thus, **<sup>f</sup>**(*z*), *<sup>z</sup>* <sup>∈</sup> *<sup>T</sup> <sup>C</sup>*, in this theorem satisfies the hypotheses of Corollary 5; and the conclusions of Corollary 5 follow for the **f**(*z*), *z* ∈ *T <sup>C</sup>*, here. In fact the construction of proofs above leading to Corollary 5 for the growth bound of type
$$M(1 + (d(y))^{-r})^s e^{2\pi A|y|} \text{ } y \in \mathbb{C} \text{ } \mathbb{C}$$
would be the same for the growth of type (36) with *d*(*y*) replaced by |*y*| in the analysis of the proofs as noted before. Thus, there is a measurable function **g** ∈ *L* 2 (R*<sup>n</sup>* , H) ∩ S<sup>0</sup> 2 (R*<sup>n</sup>* , H) with supp(**g**) ⊆ {*<sup>t</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* : *uC*(*t*) ≤ *A*} a.e. such that (26) and (27) hold with *d*(*y*) replaced by |*y*| in (26), and *z* = *x* + *iy* ∈ *T <sup>C</sup>*. From the construction of **g** in Lemma 1 and the proof of Theorem 2, *e* <sup>−</sup>2*π*h*y*,*t*i**g**(*t*) <sup>∈</sup> *<sup>L</sup>* 1 (R*<sup>n</sup>* , H) ∩ *L* 2 (R*<sup>n</sup>* , H), *y* ∈ *C*. Let *y<sup>o</sup>* ∈ *∂C*, the boundary of C, *y<sup>o</sup>* 6= 0. Since |*yo*| > 0 choose constants *a* and *b* such that 0 < *a* < |*yo*| < *b* and consider the band {*y* ∈ *C* : 0 < *a* < |*y*| < *b*} ⊂ *C*. Let {*ym*}, *m* = 1, 2, ..., be a sequence of points in this band which converges to *yo*. For each *ym*, *m* = 1, 2, ..., in this band
$$\int\_{\mathbb{R}^n} (\mathcal{N}(e^{-2\pi \langle y\_m t \rangle} \mathbf{g}(t)))^2 dt \le M^2 (1 + |y\_m|^{-r})^{2s} e^{4\pi A |y\_m|} \le M^2 (1 + a^{-r})^{2s} e^{4\pi bA}.$$
Using Fatou's lemma we have
$$\begin{aligned} &\int\_{\mathbb{R}^n} (\mathcal{N}(e^{-2\pi \langle y\_o, t \rangle} \mathbf{g}(t)))^2 dt \le \limsup\_{y\_m \to y\_o} \int\_{\mathbb{R}^n} (\mathcal{N}(e^{-2\pi \langle y\_m, t \rangle} \mathbf{g}(t)))^2 dt \\ &\le M^2 (1 + a^{-r})^{2s} e^{4\pi bA}; \end{aligned}$$
and *e* <sup>−</sup>2*π*h*yo*,*t*i**g**(*t*) <sup>∈</sup> *<sup>L</sup>* 2 (R*<sup>n</sup>* , H) for *y<sup>o</sup>* ∈ *∂C*; further *e* <sup>−</sup>2*π*h*yo*,*t*i**g**(*t*) <sup>∈</sup> *<sup>L</sup>* 2 (R*<sup>n</sup>* , H) even if *y<sup>o</sup>* = 0 since **g** ∈ *L* 2 (R*<sup>n</sup>* , H). Recall **g** ∈ *L* 2 (R*<sup>n</sup>* , H) ∩ S<sup>0</sup> 2 (R*<sup>n</sup>* , H) and *e* <sup>−</sup>2*π*h*y*,*t*i**g**(*t*) <sup>∈</sup> *<sup>L</sup>* 1 (R*<sup>n</sup>* , H) ∩ *L* 2 (R*<sup>n</sup>* , H), *y* ∈ *C*. Form
$$\mathbf{F}(\mathfrak{x} + i y\_o) = \mathcal{F}[e^{-2\pi \langle y\_o, t \rangle} \mathbf{g}(t); \mathfrak{x}], \ y\_o \in \partial \mathbb{C}, \ y\_o \neq \overline{0}; \overline{\mathfrak{x}}$$
thus, **F**(*x* + *iyo*) ∈ *L* 2 (R*<sup>n</sup>* , H), *y<sup>o</sup>* ∈ *∂C*, *y<sup>o</sup>* 6= 0. From the definition of **F**(*x* + *iyo*) and Corollary 5 we have
$$\begin{aligned} |\mathbf{f}(\mathbf{x} + i\mathbf{y}) - \mathbf{F}(\mathbf{x} + i\mathbf{y}\_o)|\_2 &= |\mathcal{F}[(e^{-2\pi\langle y, t\rangle} - e^{-2\pi\langle y\_o, t\rangle})\mathbf{g}(t); \mathbf{x}]|\_2 \\ = |(e^{-2\pi\langle y, t\rangle} - e^{-2\pi\langle y\_o, t\rangle})\mathbf{g}(t)|\_2 \end{aligned} \tag{39}$$
for *y* ∈ *C* and *y<sup>o</sup>* ∈ *∂C*, *y<sup>o</sup>* 6= 0. We consider
$$\int\_{\mathbb{R}^n} (\mathcal{N}((e^{-2\pi \langle y, t \rangle} - e^{-2\pi \langle y\_o, t \rangle}) \mathbf{g}(t)))^2 dt$$
and want to show that this integral approaches 0 as *y* → *yo*, *y* ∈ {*y* ∈ *C* : 0 < *a* < |*y*| < *b*}. We have supp(**g**) ⊆ {*<sup>t</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* : *uC*(*t*) ≤ *A*} = *C* ∗ + *N*(0; *A*) since *C* is open and convex as noted before in the proof of Theorem 8; thus, *<sup>t</sup>* ∈ {*<sup>t</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* : *uC*(*t*) ≤ *A*} implies *t* = *t*<sup>1</sup> + *t*<sup>2</sup> where *t*<sup>1</sup> ∈ *C* <sup>∗</sup> and *t*<sup>2</sup> ∈ *N*(0, *A*). For *y* ∈ {*y* ∈ *C* : 0 < *a* < |*y*| < *b*} with 0 < *a* < |*yo*| < *b* by definition of *<sup>a</sup>* and *<sup>b</sup>* we have for almost all *<sup>t</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>*
$$\left(\mathcal{N}((e^{-2\pi \langle y,t\rangle} - e^{-2\pi \langle y\_o,t\rangle})\mathbf{g}(t))\right)^2 = |e^{-2\pi \langle y,t\_1+t\_2\rangle} - e^{-2\pi \langle y\_o,t\_1+t\_2\rangle}|^2 (\mathcal{N}(\mathbf{g}(t)))^2 \dots$$
Since *t*<sup>1</sup> ∈ *C* ∗ ,h*y*, *t*1i ≥ 0 for all *y* ∈ *C* which implies h*yo*, *t*1i ≥ 0 also. Continuing the preceding inequality we have for *t*<sup>1</sup> ∈ *C* ∗ , *t*<sup>2</sup> ∈ *N*(0, *A*), and all *y* ∈ {*y* ∈ *C* : 0 < *a* < |*y*| < *b*}
$$\begin{aligned} &(\mathcal{N}((e^{-2\pi\langle y,t\rangle}-e^{-2\pi\langle y\_o,t\rangle})\mathbf{g}(t)))^2 \leq (e^{-2\pi\langle y,t\_2\rangle}+e^{-2\pi\langle y\_o,t\_2\rangle})^2(\mathcal{N}(\mathbf{g}(t)))^2\\ &\leq (e^{2\pi\langle y||t\_2\rangle}+e^{2\pi\langle y\_o||t\_2\rangle})^2(\mathcal{N}(\mathbf{g}(t)))^2 \leq 4e^{4\pi bA}(\mathcal{N}(\mathbf{g}(t)))^2 \end{aligned}$$
with the bound being independent of *y* ∈ {*y* ∈ *C* : 0 < *a* < |*y*| < *b*} and being in *L* 1 (R*<sup>n</sup>* ) since **g** ∈ *L* 2 (R*<sup>n</sup>* , H). Since (*e* <sup>−</sup>2*π*h*y*,*t*<sup>i</sup> <sup>−</sup> *<sup>e</sup>* −2*π*h*yo*,*t*i )**g**(*t*) → Θ as *y* → *yo*, *y* ∈ {*y* ∈ *C* : 0 <
*a* < |*y*| < *b*} with 0 < *a* < |*yo*| < *b*, the Lebesgue dominated convergence theorem and (39) yield (37).
To prove (38) let *<sup>φ</sup>* ∈ S(R*<sup>n</sup>* ) and *y<sup>o</sup>* ∈ *∂C*, *y<sup>o</sup>* 6= 0. As before choose constants *a* and *b* such that 0 < *a* < |*yo*| < *b*. For *y* ∈ {*y* ∈ *C* : 0 < *a* < |*y*| < *b*} we have
$$\begin{aligned} &\mathcal{N}(\langle\mathbf{f}(\mathbf{x}+i\mathbf{y}),\boldsymbol{\phi}(\mathbf{x})\rangle - \langle\mathbf{F}(\mathbf{x}+i\mathbf{y}\_o),\boldsymbol{\phi}(\mathbf{x})\rangle) \\ &\leq \int\_{\mathbb{R}^n} \mathcal{N}((\mathbf{f}(\mathbf{x}+i\mathbf{y}) - \mathbf{F}(\mathbf{x}+i\mathbf{y}\_o))\boldsymbol{\phi}(\mathbf{x}))d\mathbf{x} \\ &\leq |\mathbf{f}(\mathbf{x}+i\mathbf{y}) - \mathbf{F}(\mathbf{x}+i\mathbf{y}\_o)|\_2 ||\boldsymbol{\phi}||\_{L^2(\mathbb{R}^n)}. \end{aligned}$$
Using (37) we obtain (38) in the weak topology of S 0 (R*<sup>n</sup>* , H) as *y* → *yo*, *y* ∈ {*y* ∈ *C* : 0 < *a* < |*y*| < *b*} with 0 < *a* < |*yo*| < *b*. Now (38) is obtained in the strong topology of S 0 (R*<sup>n</sup>* , <sup>H</sup>) since <sup>S</sup>(R*<sup>n</sup>* ) is a Montel space. The proof is complete.
Since both *R* 2 *A* (*T <sup>C</sup>*, <sup>H</sup>) and *<sup>V</sup>* 2 *A* (*T <sup>C</sup>*, <sup>H</sup>) are subsets of *<sup>H</sup>*<sup>2</sup> *A* (*T <sup>C</sup>*, <sup>H</sup>), functions in both of these subset spaces satisfy (29) and (30) on the distinguished boundary of *T <sup>C</sup>* with *C* being a regular cone. Also *V* 2 *A* (*T <sup>C</sup>*, <sup>H</sup>) functions will have the results of Theorem <sup>12</sup> since *V* 2 *A* (*T <sup>C</sup>*, <sup>H</sup>) <sup>⊆</sup> *<sup>R</sup>* 2 *A* (*T <sup>C</sup>*, <sup>H</sup>).
Boundary value results for the analytic functions on the topological boundary of the tube may be able to be obtained for various types of base sets *C* of the tube *T <sup>C</sup>*. For example one could consider *C* to be an open polyhedron in R*<sup>n</sup>* as defined in [11] and [12] (p. 97). One could follow this situation by considering an open convex subset *B* of R*<sup>n</sup>* with *y<sup>o</sup>* being a point on its boundary; consideration could be given then to constructing an open polyhedron in *B* with *y<sup>o</sup>* as boundary point and approaching *y<sup>o</sup>* within the open polyhedron as Stein and Weiss have done in [12] (p. 98) for functions in *H*<sup>2</sup> (*T B* ). Clearly the types of boundary values available will depend on the specifics of the analytic functions and on the base of the tube if boundary values exist at all. More will be stated in Section 9 concerning boundary values.
We have previously obtained boundary value results on the distinguished boundary of the tube for functions of type *V p* <sup>∗</sup> (*T <sup>C</sup>*), 1 <sup>&</sup>lt; *<sup>p</sup>* <sup>≤</sup> 2, in the scalar-valued ultradistribution sense where *C* is a regular cone in R*<sup>n</sup>* . That is, the norm growth on the analytic functions on *T <sup>C</sup>* is
$$||\mathbf{f}(\mathfrak{x} + i\mathfrak{y})||\_{L^p(\mathbb{R}^n)} \le K e^{M^\*(w/|y|)}, \ y \in \mathbb{C}\_{\prime}$$
where *w* > 0 and *K* = *K*(**f**, *p*, *w*) are independent of *y* ∈ *C*. We have proved that such functions obtain a boundary value at 0 in the ultradistribution space D<sup>0</sup> ((*Mp*), *L* 1 (R*<sup>n</sup>* )). We refer to [2] (p. 106, Theorem 5.2.1) and the preceding analysis in [2] (Section 5.2).
#### **9. Suggested Research**
In this section, we suggest problems to consider in future research which are associated with the analysis of this paper.
Let *B* be an open connected subset of R*<sup>n</sup>* . Stein and Weiss use a bound condition on *H<sup>p</sup>* (*T B* ) obtained in [12] (p. 99, Lemma 2.12) to prove [12] (p. 93, Theorem 2.3), the representation theorem for functions in *H*<sup>2</sup> (*T B* ). The bound condition holds for *z* in a tube whose base is restricted uniformly away from the complement of *B*. We have used a similarly needed growth condition, obtained in [2] (p. 87, Lemma 5.1.3), on the analytic functions studied in [2] (Chapter 5) in relation to boundary values in ultradistribution spaces.
Starting with Lemma 1 in Section 6 of this paper we have used the following assumption on **<sup>f</sup>**(*z*) <sup>∈</sup> *<sup>H</sup>*<sup>2</sup> *A* (*T B* , H) to obtain several results; the assumption on **f**(*z*) is that it "be bounded for *<sup>x</sup>* <sup>=</sup> Re(*z*) <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* and *<sup>y</sup>* <sup>=</sup>Im(*z*) in any compact subset of *<sup>B</sup>*". We conjecture that a bound condition like [12] (p. 99, Lemma 2.12) holds for **f**(*z*) ∈ *H p A* (*T B* , X ); such a result will allow us to delete the above quoted assumption used in Sections 6–8.
Additionally we suggest research to obtain a bound condition like [12] (p. 99, Lemma 2.12) for functions in *H<sup>p</sup>* (*T B* , X ).
Throughout this paper we have obtained boundary value results both on the distinguished boundary of the tube and on the topological boundary of the tube. In every case a question that had to be considered was the method to approach a point on the boundary by points in the base in order to obtain a desired result. Our results before Section 8 concerned tubes with base being a regular cone, an open connected cone in R*<sup>n</sup>* , or a proper open connected subset of R*<sup>n</sup>* . In these cases we could approach a considered boundary point *y<sup>o</sup>* on the boundary of the base by a sequence of points within the base. Because of the nature of the analytic functions considered in Section 8 we needed to approach any boundary point *yo*, *y<sup>o</sup>* 6= 0, on the boundary of the base, a regular cone, by a sequence of points inside a band contained in the cone in order to obtain the desired result. Indications of other boundary point approaches for consideration were stated at the end of Section 8.
Stein and Weiss [12] (pp. 94–98) discuss situations in which boundary values on the boundary of tubes can not be obtained as points within the base arbitrarily approach the point *y<sup>o</sup>* on the boundary of the base. In the first case a specific type of analytic function was constructed in order to show the non-existence of a boundary value for arbitrary approach to a point on the boundary by points within the base. In the second case a *H*<sup>2</sup> (*T B* ) function was constructed for which no limit in the *L* <sup>2</sup> norm existed for arbitrary approach to 0 within *B*; but if the base *B* was suitably restricted, any function in *H*<sup>2</sup> (*T B* ) for the restricted base *B* was shown to have a boundary value at any point on *∂B*. Considerations of the approach to the boundary by points within bases *B* of other types than those of this paper could be made concerning the types of analytic functions defined in this paper. Are there base sets *B* in which an analytic function will not have a boundary value at a specified point *y<sup>o</sup>* ∈ *∂B* or such that there could be a boundary value if the base *B* is specialized?
The basic results of Section 5, Theorems 2, 4, 6 and 8, have all been proved for the most general appropriate situation. *B* was an open connected subset of R*<sup>n</sup>* or open (or convex) connected cone in R*<sup>n</sup>* ; values were in Banach space X ; results held for all *p*, 1 ≤ *p* < ∞, in Section 5. In Sections 6–8, the base *B* of the tube remained an open connected subset of R*<sup>n</sup>* or a cone in R*<sup>n</sup>* as appropriate; but all of the main results of these sections were proved for values in Hilbert space H with *p* = 2.
Of course the reason for the restrictions in these sections to *p* = 2 and values in H is that the primary tool in our proofs was the Fourier transform which, as previously noted, is available in its desired completeness to the specific cases of *p* = 2 and values in H. We desire to extend the results of Sections 6–8 to 1 ≤ *p* < ∞ and values in Banach space X as appropriate by using different techniques. This has been done by Vindas in [1] where functional analysis techniques have been used to extend the Poisson integral representation of functions in *H<sup>p</sup>* (*T <sup>C</sup>*, <sup>H</sup>) from *<sup>p</sup>* <sup>=</sup> <sup>2</sup> with values in <sup>H</sup> to <sup>1</sup> <sup>≤</sup> *<sup>p</sup>* <sup>≤</sup> <sup>∞</sup> with values in X . See [1] (Theorem 2); similarly see also [1] (Theorem 1). Use of functional analysis techniques and accumulated knowledge related to vector-valued fuctions to obtain the desired extensions of the results noted in this paragraph should be considered. Extensions of results from *p* = 2 to 1 ≤ *p* < ∞ could possibly also be obtained here for Hilbert space H by applying limit processes using the *p* = 2 case. We believe that the basic results of Sections 6–8 can be extended to 1 ≤ *p* < ∞ and values in Banach space X as appropriate. We suggest consideration of this extension in future research.
For *p* = 2 we have proved in previous work that the S 0 (R*<sup>n</sup>* ) Fourier transform maps the distribution space D<sup>0</sup> *L* <sup>2</sup>(R*n*) one-one and onto S 0 2 ; further we have proved that the S 0 (R*<sup>n</sup>* ) Fourier transform maps D<sup>0</sup> *L p* (R*n*) , 1 ≤ *p* < 2, one-one and into S 0 *q* , (1/*p*) + (1/*q*) = 1. The proofs are obtained using the characterization results for the form of elements in D0 *L p* (R*n*) , 1 ≤ *p* ≤ 2. With knowledge of a characterization of elements in the vector-valued distribution space equivalent to D<sup>0</sup> *L* <sup>2</sup>(R*n*) we conjecture that the S 0 (R*<sup>n</sup>* , H) Fourier transform maps this vector-valued distribution space one-one and onto S 0 2 (R*<sup>n</sup>* , H). Of course the values of the vector-valued distributions would need to be in Hilbert space H because of the probable use of the function Fourier transform on *L* 2 (R*<sup>n</sup>* , H) functions.
Results similar to those of this paper may be in order concerning the functions defined as *H*(*C*) in [7]. We leave this for future research.
#### **10. Conclusions**
As stated in Section 1 our goal in this paper was to obtain results for the analytic functions defined in Section 4 treated as generalizations of *H<sup>p</sup>* (*T B* , X ) functions and as generalizations of the scalar-valued functions noted in [2] (Chapter 5) and in some of our papers referenced in [2] and hence to generalize results concerning *H<sup>p</sup>* (*T B* , X ) spaces and concerning the functions of [2] (Chapter 5) to these new spaces of analytic functions. Additionally, we stated that our goal also was to obtain additional new results for the analytic fuctions of Section 4.
We were successful in our goals in Section 5 for all of the results there that had as assumption that **g**(*t*) was a X valued measurable function for which the growth (1) held and for all of the results that had as assumption that **g** ∈ S<sup>0</sup> *p* (R*<sup>n</sup>* , X ); these results held for X being a Banach space and for all *p*, 1 ≤ *p* < ∞.
We were partially successful in our goals in Section 6 where the results depended on hypotheses on the analytic function concerning X and p. Because our proofs of these results depended on the Fourier transform we had to restrict X to H, a Hilbert space, and *p* = 2 as described previously. But under these restrictions in Section 6 we were able to obtain Fourier–Laplace integral representation and boundary value results on the distinguished boundary of the tube for the analytic functions. In Section 7, we were able to prove containment of certain analytic functions from Definitions 1–4 in the Hardy space *H*<sup>2</sup> (*T <sup>C</sup>*, <sup>H</sup>). In Section 8, we were able to obtain boundary value results on the topological boundary of the tube domain for the functions considered there. We desire to have the results of Sections 6–8 holding as well for X being a Banach space and for 1 ≤ *p* < ∞.
In our previous work concerning scalar-valued generalizations of *H<sup>p</sup>* (*T B* ) functions we have been able to obtain results under the assumption on the analytic functions of the type in Sections 6–8 for all *p*, 1 ≤ *p* < ∞. That is we have obtained Fourier–Laplace integral representation and boundary value results for all *p*, 1 ≤ *p* < ∞, on the assumed scalar-valued analytic function. Additionally, we have obtained Cauchy and Poisson integral representations as appropriate. Because of the existence of these results for all *p* in the scalar-valued case we have emphasized in Section 9 our belief that the basic results of Sections 6–8 can be extended to 1 ≤ *p* < ∞ and to values in Banach space X under assumption on the analytic function in the results. We believe that new techniques apart from the Fourier transform will be used to obtained these desired results as described in Section 9. We pursue the analysis of these topics for the generalized setting in the future.
The author believes that there is considerable additional interesting analysis in the generalized format of the results in this paper that can be obtained in regards to integral representation, boundary values, and applications for the functions of Definitions 1–4.
**Funding:** This research received no external funding.
**Conflicts of Interest:** The author declares no conflict of interest.
#### **References**
## *Article* **An Application of S˘al˘agean Operator Concerning Starlike Functions †**
**Hatun Özlem Güney <sup>1</sup> , Georgia Irina Oros <sup>2</sup> and Shigeyoshi Owa 3,\***
**Abstract:** As an application of the well-known S˘al˘agean differential operator, a new operator is introduced and, using this, a new class of functions *Sn*(*α*) is defined, which has the classes of starlike and convex functions of order *α* as special cases. Original results related to the newly defined class are obtained using the renowned Jack–Miller–Mocanu lemma. A relevant example is given regarding the applications of a new proven result concerning interesting properties of class *Sn*(*α*).
**Keywords:** analytic function; starlike function of order *α*; convex function of order *α*; S˘al˘agean differential operator; Alexander integral operator
#### **1. Introduction and Preliminaries**
Many operators have been used since the beginning of the study of analytic functions. The most interesting of these are the differential and integral operators. Since the beginning of the 20th century, many mathematicians, especially J.W. Alexander [1], S.D. Bernardi [2] and R.J. Libera [3], have worked on integral operators. It has become easier to introduce new classes of univalent functions with the use of operators. In his article, published in 1983, S˘al˘agean introduced differential and integral operators, which bear his name. Those operators were very inspiring and many mathematicians have obtained new, interesting results using these operators. In particular, researchers have introduced many new operators, examined their properties, and further used the newly defined operators to introduce classes of univalent functions with remarkable properties. At the same time, some mathematicians obtained interesting results in different lines of research by combining differential and integral operators, where S˘al˘agean differential operator was involved, as is seen, for example, in very recent papers [4–6]. The topic of strong differential subordination was also approached recently using S˘al˘agean differential operator in [7], and new operators were introduced using a fractional integral of S˘al˘agean and Ruscheweyh operators in [8]. The operators introduced using the S˘al˘agean differential operator were also recently used to obtain results related to the celebrated Fekete–Szegö inequality [9].
In this work, we introduce a new class as an application of the S˘al˘agean operator and discuss some interesting problems with this class.
Let *A* be the class of functions *f* of the form
$$f(z) = z + \sum\_{k=2}^{\infty} a\_k z^k \tag{1}$$
**Citation:** Güney, H.Ö.; Oros, G.I.; Owa, S. An Application of S˘al˘agean Operator Concerning Starlike Functions. *Axioms* **2022**, *11*, 50. https://doi.org/10.3390/ axioms11020050
Academic Editor: Kurt Bernardo Wolf
Received: 15 December 2021 Accepted: 25 January 2022 Published: 27 January 2022
**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.
**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).
which are analytic in the open unit disc U = {*z* ∈ C : |*z*| < 1} and S be the subclass of *A* consisting of univalent functions. Also,
$$\mathcal{S}^\*(\alpha) = \left\{ f \in A : \text{Re}\left(\frac{zf'(z)}{f(z)}\right) > \alpha, z \in \mathbb{U}, 0 \le \alpha < 1 \right\}$$
is the class of starlike functions of order *α* and
$$K(\mathfrak{a}) = \left\{ f \in A : \text{Re}\left(1 + \frac{zf''(z)}{f'(z)}\right) > \mathfrak{a}, z \in \mathbb{U}, 0 \le \mathfrak{a} < 1 \right\}.$$
is the class of convex functions of order *α*.
Let us start by recalling the well-known definitions for the S˘al˘agean differential and integral operators.
**Definition 1** (S˘al˘agean [10])**.** *For <sup>f</sup>* <sup>∈</sup> *<sup>A</sup>*, *the S˘al˘agean differential operator <sup>D</sup><sup>n</sup> is defined by D<sup>n</sup>* : *A* → *A*,
$$D^0 f(z) = f(z) = z + \sum\_{k=2}^{\infty} a\_k z^k \, \tag{2}$$
$$D^1 f(z) = Df(z) = zf'(z) = z + \sum\_{k=2}^{\infty} k a\_k z^k,\tag{3}$$
$$D^n f(z) = D(D^{n-1} f(z)) = z + \sum\_{k=2}^{\infty} k^n a\_k z^k \quad (n = 1, 2, 3, \dotsb), \tag{4}$$
*and S˘al˘agean integral operator D*−*<sup>n</sup> is defined by*
$$D^{-1}f(z) = \int\_0^z \frac{f(t)}{t} dt = z + \sum\_{k=2}^\infty \frac{1}{k} a\_k z^k \tag{5}$$
*and*
$$D^{-n}f(z) = D^{-1}(D^{-n+1}f(z)) = z + \sum\_{k=2}^{\infty} \frac{1}{k^n} a\_k z^k \quad (n = 1, 2, 3, \cdots). \tag{6}$$
In view of Definition 1, the following new operator is introduced:
**Definition 2.** *For f* ∈ *A*
$$D^j f(z) = z + \sum\_{k=2}^{\infty} k^j a\_k z^k \ (j = \dots , -2, -1, 0, 1, 2, \dots) \,. \tag{7}$$
With the above operator *D<sup>j</sup> f* , we introduce the subclass *Sn*(*α*).
**Definition 3.** *The subclass Sn*(*α*) *of A consists of functions f , which satisfy*
$$\operatorname{Re}\left(\frac{D^{n+1}f(z)}{D^nf(z)}\right) > \mathfrak{a} \quad (n = \cdots, -2, -1, 0, 1, 2, \cdots) \tag{8}$$
*for z* ∈ U, *where* 0 ≤ *α* < 1.
**Remark 1.** *Since D*<sup>0</sup> *f*(*z*) = *f*(*z*), *D*<sup>1</sup> *f*(*z*) = *z f* 0 (*z*) *and D*<sup>2</sup> *f*(*z*) = *z f* 0 (*z*) + *z* 2 *f* <sup>00</sup>(*z*), *f* ∈ *S*0(*α*) *satisfies*
$$\operatorname{Re}\left(\frac{zf'(z)}{f(z)}\right) > a \quad (z \in \mathbb{U}),\tag{9}$$
*and f* ∈ *S*1(*α*) *satisfies*
$$\operatorname{Re}\left(1+\frac{zf''(z)}{f'(z)}\right)>\mathfrak{a} \quad (z\in\mathbb{U}).\tag{10}$$
*Therefore, f* ∈ *S*0(*α*) = *S* ∗ (*α*) *is starlike of order α in* U, *and f* ∈ *S*1(*α*) = *K*(*α*) *is convex of order α in* U *(cf. Robertson [11]). Since D*−<sup>1</sup> *f is Alexander integral operator, D*−*<sup>n</sup> f* (*n* = 1, 2, 3, · · ·) *is the generalization for Alexander integral operator (cf. Alexander [1]).*
For a function *f* ∈ *A*, we introduce
$$M\_p(r,f) = \begin{cases} \left(\frac{1}{2\pi} \int\_0^{2\pi} |f(re^{i\theta})|^p d\theta\right)^{\frac{1}{p}} & , (0 < p < \infty) \\\\ \max\_{|z| \le r} |f(z)| & , \qquad (p = \infty). \end{cases} \tag{11}$$
For the above *Mp*(*r*, *f*), we define
$$\mathcal{H}^p = \left\{ f \in A : \|f\|\_p = \lim\_{r \to 1^-} M\_p(r, f) < \infty \right\}.\tag{12}$$
To discuss our problems, we have to introduce the following lemmas.
**Lemma 1** (Wilken and Feng [12])**.** *If f* ∈ *S*1(*α*), *then f* ∈ *S*0(*β*), *where*
$$\beta = \beta(a) = \begin{cases} \frac{2a - 1}{2(1 - 2^{1 - 2a})} & \text{, } (a \neq \frac{1}{2}) \\\\ \frac{1}{2\log 2} = 0.7213\dots & \text{, } (a = \frac{1}{2}) \end{cases} \tag{13}$$
*The result is sharp.*
**Lemma 2** (Eenigenburg and Keogh [13])**.** *If f* ∈ *S*0(*α*) *and*
$$f(z) \neq \frac{z}{(1 - ze^{i\theta})^2} \tag{14}$$
*then there exists δ* = *δ*(*f*) > 0 *such that <sup>f</sup>*(*z*) *<sup>z</sup>* ∈ H*δ*<sup>+</sup> <sup>1</sup> 2(1−*α*) .
**Lemma 3** (Nunokawa [14])**.** *Let a function p be analytic in* U *with p*(0) = 1. *If p satisfies*
$$\operatorname{Re}(p(z) + zp'(z)) > \frac{1 - 2\log 2}{2(1 - \log 2)} = -0.629\dots \quad (z \in \mathbb{U})\tag{15}$$
*then Rep*(*z*) > 0 (*z* ∈ U).
**Lemma 4** (Duren [15])**.** *If a function <sup>p</sup> is analytic in* <sup>U</sup> *and Rep*(*z*) <sup>&</sup>gt; <sup>0</sup> (*<sup>z</sup>* <sup>∈</sup> <sup>U</sup>), *then <sup>p</sup>* ∈ H*<sup>p</sup>* (0 < *p* < 1).
**Lemma 5** (Kim, Lee and Srivastava [16])**.** *If f* ∈ *A satisfies z γ <sup>f</sup>*(*z*) ∈ H*<sup>p</sup>* (0 < *p* < ∞) *for some real <sup>γ</sup>*, *then f* ∈ H*<sup>p</sup>* (0 < *p* < ∞).
**Lemma 6** (Duren [15])**.** *If f* <sup>∈</sup> *A satisfies f* <sup>0</sup> ∈ H*<sup>p</sup>* (0 < *p* < 1), *then f* ∈ H *p* 1−*p* .
Discussing our problems for S˘al˘agean operator, we need to introduce the following lemma due to Miller and Mocanu [17,18] (also, by Jack [19]).
**Lemma 7** (Miller and Mocanu [17,18])**.** *Let the function w given by*
$$w(z) = b\_n z^n + b\_{n+1} z^{n+1} + b\_{n+2} z^{n+2} + \dots \quad , \ n \in \mathbb{N} \tag{16}$$
*be analytic in* U *with w*(0) = 0. *If* |*w*(*z*)| *attains its maximum value on the circle* |*z*| = *r at a point z*<sup>0</sup> ∈ U, *then a real number k* ≥ *n exists, such that*
$$\frac{z\_0 w'(z\_0)}{w(z\_0)} = k \tag{17}$$
*and*
$$\operatorname{Re}\left(1+\frac{z\_0 w''(z\_0)}{w'(z\_0)}\right)\geq k.\tag{18}$$
The original results obtained by the authors and presented in this paper are contained in the next section. A new operator is introduced with S˘al˘agean differential operator as the inspiration. Using this newly introduced operator, a new class of functions denoted by *Sn*(*α*) is defined, with known classes as particular cases. Certain properties involving the applications of S˘al˘agean differential operator related to class *Sn*(*α*) are discussed in the theorems and corollaries. Examples are also included to prove the applications of the proved results.
#### **2. Main Results**
Now, we derive the following result.
**Theorem 1.** *If f* ∈ *Sn*(*α*), *then f* ∈ *Sn*−*j*(*αj*), *where n* > *j* ≥ 0 *and*
$$\mathfrak{a}\_{j} = \begin{cases} \frac{2a\_{j-1} - 1}{2(1 - 2^{1 - 2a\_{j-1}})} & \text{ / } (\mathfrak{a}\_{j} \neq \frac{1}{2}) \\\\ \frac{1}{2\log 2} = 0.7213\dots & \text{ / } (\mathfrak{a}\_{j} = \frac{1}{2}). \end{cases} \tag{19}$$
*Further, if*
$$D^{n-j}f(z) \neq \frac{z}{(1 - ze^{i\theta})^{2(1 - a\_j)}},\tag{20}$$
*then there exists δ* > 0*, such that Dn*−*<sup>j</sup> f* ∈ H *δ*+ <sup>1</sup> 2(1−*αj* ) .
**Proof.** We note that if *f* ∈ *Sn*(*α*), then
$$\operatorname{Re}\left(\frac{D^{n+1}f(z)}{D^nf(z)}\right) > a\_0 \quad (z \in \mathbb{U}),\tag{21}$$
where *α*<sup>0</sup> = *α*. Since
$$D^{n+1}f(z) = z(D^nf(z))' = z(D^{n-1}f(z))' + z^2(D^{n-1}f(z))''\tag{22}$$
and
$$D^n f(z) = z(D^{n-1}f(z))',\tag{23}$$
we see that
$$\operatorname{Re}\left(\frac{D^{n+1}f(z)}{D^nf(z)}\right) = \operatorname{Re}\left(1 + \frac{z(D^{n-1}f(z))''}{(D^{n-1}f(z))'}\right) > a\_0 \quad (z \in \mathbb{U}).\tag{24}$$
Applying Lemma 1, we say that
$$\begin{aligned} f \in \mathcal{S}\_n(\mathfrak{a}\_0) &\Leftrightarrow D^{n-1}f \in \mathcal{S}\_1(\mathfrak{a}\_0) \\ &\Rightarrow D^{n-1}f \in \mathcal{S}\_0(\mathfrak{a}\_1) \\ &\Leftrightarrow D^{n-2}f \in \mathcal{S}\_1(\mathfrak{a}\_1) \\ &\Rightarrow D^{n-2}f \in \mathcal{S}\_0(\mathfrak{a}\_2) \\ &\Leftrightarrow D^{n-3}f \in \mathcal{S}\_1(\mathfrak{a}\_2) \\ &\vdots \\ &\Leftrightarrow D^{n-j}f \in \mathcal{S}\_0(\mathfrak{a}\_{j-1}) \\ &\Rightarrow D^{n-j}f \in \mathcal{S}\_1(\mathfrak{a}\_j). \end{aligned} \tag{25}$$
This implies that
$$\operatorname{Re}\left(\frac{z(D^{n-j}f(z))'}{D^{n-j}f(z)}\right) = \operatorname{Re}\left(\frac{D^{n-j+1}f(z)}{D^{n-j}f(z)}\right) > a\_j \quad (z \in \mathbb{U}),\tag{26}$$
that is, that *f* ∈ *Sn*−*j*(*αj*). Further, applying Lemma 2, we see that if
$$D^{n-j}f(z) \neq \frac{z}{(1 - ze^{i\theta})^{2(1 - a\_j)}},\tag{27}$$
then there exists *δ* > 0, such that *Dn*−*<sup>j</sup> f* ∈ H *δ*+ <sup>1</sup> 2(1−*αj* ) .
**Example 1.** *Let us consider a function f belonging to the class S*3(*α*). *Then f* ∈ *S*2(*α*1) *with (19), where*
$$\mathfrak{a}\_1 = \begin{cases} \frac{2\mathfrak{a} - 1}{2(1 - 2^{1 - 2\mathfrak{a}})} & \text{, } (\mathfrak{a} \neq \frac{1}{2}) \\\\ \frac{1}{2\log 2} = 0.7213\dots & \text{, } (\mathfrak{a} = \frac{1}{2}) . \end{cases} \tag{28}$$
*Further, f* ∈ *S*1(*α*2), *where*
$$\mathfrak{a}\_2 = \begin{cases} \frac{2\mathfrak{a}\_1 - 1}{2(1 - 2^{1 - 2\mathfrak{a}\_1})} & \text{ / } (\mathfrak{a}\_1 \neq \frac{1}{2}) \\\\ \frac{1}{2\log 2} = 0.7213\dots & \text{ / } (\mathfrak{a}\_1 = \frac{1}{2}) . \end{cases} \tag{29}$$
*Also, f* ∈ *S*0(*α*3), *where*
$$\mathfrak{a}\_3 = \begin{cases} \frac{2a\_2 - 1}{2(1 - 2^{1 - 2a\_2})} & \text{ ( $a\_2 \neq \frac{1}{2}$ )}\\\\ \frac{1}{2\log 2} = 0.7213\dots & \text{ ( $a\_2 = \frac{1}{2}$ )} \end{cases} \tag{30}$$
*If we consider the case of α* = <sup>1</sup> 4 , *then we have*
$$\alpha\_1 = \frac{1}{4(\sqrt{2} - 1)} \doteq 0.60355,\tag{31}$$
$$u\_2 = \frac{3 - 2\sqrt{2}}{4(\sqrt{2} - 1)(1 - 2^{\frac{1 - \sqrt{2}}{2}})} \doteq 0.77436,\tag{32}$$
*and*
$$
\alpha\_3 \doteq 0.8672.\tag{33}
$$
*Further, if we consider the case of α* = <sup>1</sup> 8 , *then*
$$\alpha\_1 = \frac{3}{8(\sqrt[4]{8}-1)} \doteq 0.55002,\tag{34}$$
*and*
$$\alpha\_2 = \frac{7 - 4\sqrt[4]{8}}{8(\sqrt[4]{8} - 1)(1 - 2^{\frac{4\sqrt[4]{8} - 7}{2^{4(\sqrt[4]{8} - 1)}}})} \doteq 0.60607. \tag{35}$$
**Remark 2.** *For some positive integer j*, *we know that*
$$\alpha\_{j+1} = \frac{2\alpha\_j - 1}{2(1 - 2^{1 - 2\alpha\_j})} \quad \text{ } (\alpha\_j \neq \frac{1}{2}). \tag{36}$$
*If we consider*
$$g(\mathfrak{a}\_{j}) = \mathfrak{a}\_{j+1} - \mathfrak{a}\_{j} = \frac{2\mathfrak{a}\_{j} - 1}{2(1 - 2^{1 - 2\mathfrak{a}\_{j}})} - \mathfrak{a}\_{j} \quad , \ (\mathfrak{a}\_{j} \neq \frac{1}{2}), \tag{37}$$
*g*(0) = <sup>1</sup> 2 *and g*(1) = 0. *From this fact, we know that α<sup>j</sup>* < *αj*+<sup>1</sup> *for* 0 ≤ *α<sup>j</sup>* < 1. *This implies that*
$$0 \le \mathfrak{a} < \mathfrak{a}\_1 < \mathfrak{a}\_2 < \dots < \mathfrak{a}\_j < \dots < 1. \tag{38}$$
Letting *j* = *n* in Theorem 1, we see
**Corollary 1.** *If f* ∈ *Sj*(*α*), *then f* ∈ *S*0(*αj*). *If*
$$f(z) \neq \frac{z}{(1 - ze^{i\theta})^{2(1 - a\_j)}},\tag{39}$$
*then there exists δ* > 0*, such that f* ∈ H *δ*+ <sup>1</sup> 2(1−*αj* ) .
Next we have
**Theorem 2.** *If f* ∈ *A satisfies*
$$\operatorname{Re}\left(\frac{D^{n+1}f(z)}{z}\right) = \frac{1 - 2\log 2}{2(1 - \log 2)} = -0.629\dots \quad (z \in \mathbb{U})\tag{40}$$
*for some n* ∈ N, *then there exists p<sup>j</sup> , such that Dn*−*j*+<sup>1</sup> *<sup>f</sup>* ∈ H*p<sup>j</sup>* , *where*
$$p\_j > \frac{1}{j - k + 1} \quad (k = 1, 2, 3, \dots, j) \tag{41}$$
*and j* ≤ *n* + 1.
**Proof.** If we define *p* by
$$p(z) = \frac{D^n f(z)}{z},\tag{42}$$
then *p* is analytic in U with *p*(0) = 1. Since
$$p(z) + zp'(z) = \frac{D^{n+1}f(z)}{z},\tag{43}$$
we see that
$$\operatorname{Re}\left(\frac{D^{n+1}f(z)}{z}\right) = \operatorname{Re}\left(p(z) + zp'(z)\right) > \frac{1 - 2\log 2}{2(1 - \log 2)} \quad (z \in \mathbb{U}).\tag{44}$$
Applying Lemma 3, we have that
$$\operatorname{Re} p(z) = \operatorname{Re} \left( \frac{D^n f(z)}{z} \right) > 0 \quad (z \in \mathbb{U}).\tag{45}$$
Using Lemma 4, we know that
$$\frac{D^n f(z)}{z} \in \mathcal{H}^{p\_1} \quad (0 < p\_1 < \frac{1}{j}),\tag{46}$$
that is, that (*Dn*−<sup>1</sup> *<sup>f</sup>*(*z*))<sup>0</sup> ∈ H*p*<sup>1</sup> . By Lemma 6, we have that
$$D^{\eta-1}f \in \mathcal{H}^{p\_2} \quad (0 < p\_2 = \frac{p\_1}{1 - p\_1} < \frac{1}{j - 1}).\tag{47}$$
Noting that
$$D^{n-1}f(z) = z(D^{n-2}f(z))',\tag{48}$$
we obtain that
$$D^{n-2}f \in \mathcal{H}^{p\_3} \quad (0 < p\_3 = \frac{p\_2}{1 - p\_2} < \frac{1}{j - 2}).\tag{49}$$
Repeating the above, we have that
$$D^{n-j+2}f \in \mathcal{H}^{p\_{j-1}} \quad (0 < p\_{j-1} < \frac{1}{2}).\tag{50}$$
Finally, we get
$$D^{n-j+1}f \in \mathcal{H}^{p\_j} \quad (0 < p\_j < 1). \tag{51}$$
Making *j* = *n* + 1 in Theorem 2, we have
**Corollary 2.** *If f* ∈ *A satisfies*
$$\operatorname{Re}\left(\frac{D^{n+1}f(z)}{z}\right) > \frac{1 - 2\log 2}{2(1 - \log 2)} = -0.629\dots \quad (z \in \mathbb{U})\,\tag{52}$$
*then, there exists pn*+<sup>1</sup> *such that f* ∈ H*pn*+<sup>1</sup> (<sup>0</sup> <sup>&</sup>lt; *<sup>p</sup>n*+<sup>1</sup> <sup>&</sup>lt; <sup>1</sup>).
Next, we derive
**Theorem 3.** *If f* ∈ *A satisfies*
$$\left| \frac{D^{n+2}f(z)}{D^{n+1}f(z)} - 1 \right| < \frac{5\alpha - 2\alpha^2 - 1}{2\alpha} \quad (z \in \mathbb{U}), (n \in \mathbb{N}) \tag{53}$$
*for some real α* ( 1 <sup>3</sup> <sup>≤</sup> *<sup>α</sup>* <sup>≤</sup> <sup>1</sup> 2 ), *or*
$$\left| \frac{D^{n+2}f(z)}{D^{n+1}f(z)} - 1 \right| < \frac{n - 2a^2 + 1}{2a} \quad (z \in \mathbb{U}), (n \in \mathbb{N}) \tag{54}$$
*for some real α* ( 1 <sup>2</sup> <sup>≤</sup> *<sup>α</sup>* <sup>&</sup>lt; <sup>1</sup>), *then <sup>D</sup><sup>n</sup> <sup>f</sup>* <sup>∈</sup> *<sup>S</sup>*0(*α*), *that is, <sup>D</sup><sup>n</sup> f is starlike of order α in* U. *Further, if*
$$D^{n-j}f(z) \neq \frac{z}{(1 - ze^{i\theta})^{2(1 - a\_j)}},\tag{55}$$
*then, there exists δ* > 0 *such that Dn*−*<sup>j</sup> f* ∈ H *δ*+ <sup>1</sup> 2(1−*αj* ) , *where*
$$\mathfrak{a}\_{j} = \begin{cases} \frac{2a\_{j-1} - 1}{2(1 - 2^{1 - 2a\_{j-1}})} & \text{, } (\mathfrak{a}\_{j-1} \neq \frac{1}{2}) \\\\ \frac{1}{2\log 2} = 0.7213\dots & \text{, } (\mathfrak{a}\_{j-1} = \frac{1}{2}) \end{cases} \tag{56}$$
*and j* ≤ *n*.
**Proof.** Define a function *w* by
$$\frac{D^{n+1}f(z)}{D^n f(z)} = \frac{1 + (1 - 2\alpha)w(z)}{1 - w(z)} \quad (w(z) \neq 1). \tag{57}$$
It follows from the above that
$$\frac{D^{n+2}f(z)}{D^{n+1}f(z)} - \frac{D^{n+1}f(z)}{D^nf(z)} = \frac{(1-2\alpha)zw'(z)}{1+(1-2\alpha)w(z)} + \frac{zw'(z)}{1-w(z)}.\tag{58}$$
Therefore, we have that
$$\frac{D^{n+2}f(z)}{D^{n+1}f(z)} - 1 = \left(\frac{w(z)}{1 - w(z)}\right) \left\{ 2(1 - a) + \frac{zw'(z)}{w(z)} \left(1 + \frac{(1 - 2a)(1 - w(z))}{1 + (1 - 2a)w(z)}\right) \right\}.\tag{59}$$
Suppose that there exists a point *z*<sup>0</sup> ∈ U, such that
$$\max\_{|z| \le |z\_0|} |w(z)| = |w(z\_0)| = 1 \quad (w(z\_0) \ne 1). \tag{60}$$
Then, Lemma 7 say that *w*(*z*0) = *e <sup>i</sup><sup>θ</sup>* and *z*0*w* 0 (*z*0) = *kw*(*z*0) (*k* ≥ 1). This implies that
$$\begin{split} \left| \frac{D^{n+2}f(z\_0)}{D^{n+1}f(z\_0)} - 1 \right| &= \left| \frac{e^{i\theta}}{1 - e^{i\theta}} \right| \left| 2(1 - a) + k \left( 1 + \frac{(1 - 2a)(1 - e^{i\theta})}{1 + (1 - 2a)e^{i\theta}} \right) \right| \\ &\geq \frac{2(1 - a\_0) + k}{\left| 1 - e^{i\theta} \right|} - \frac{k|1 - 2a|}{\left| 1 + (1 - 2a)e^{i\theta} \right|} \\ &\geq \frac{2(1 - a\_0) + k}{2} - \frac{k|1 - 2a|}{2a} . \end{split} \tag{61}$$
If <sup>1</sup> <sup>3</sup> <sup>≤</sup> *<sup>α</sup>* <sup>&</sup>lt; <sup>1</sup> 2 , then
$$\left| \frac{D^{n+2}f(z\_0)}{D^{n+1}f(z\_0)} - 1 \right| \ge \frac{5\alpha - 2\alpha^2 - 1}{2\alpha} \tag{62}$$
and if <sup>1</sup> <sup>2</sup> ≤ *α* < 1, then
$$\left| \frac{D^{n+2}f(z\_0)}{D^{n+1}f(z\_0)} - 1 \right| \ge \frac{\mathfrak{a} - 2\mathfrak{a}^2 + 1}{2\mathfrak{a}}.\tag{63}$$
This contradicts our condition of the theorem. Thus we say that |*w*(*z*)| < 1 for all *z* ∈ U. From the definition (57) for *w*, we obtain that
$$\operatorname{Re}\left(\frac{D^{n+1}f(z)}{D^nf(z)}\right) > \mathfrak{a} \quad (z \in \mathbb{U}).\tag{64}$$
This means that *D<sup>n</sup> <sup>f</sup>* <sup>∈</sup> *<sup>S</sup>*0(*α*). Letting *<sup>α</sup>* <sup>=</sup> *<sup>α</sup>*<sup>0</sup> and using Lemma 1, we obtain *<sup>D</sup>n*−*<sup>j</sup> f* ∈ *S*0(*αj*), where *α<sup>j</sup>* is given by (56). Applying Lemma 2, we know that if
$$D^{n-j}f(z) \neq \frac{z}{(1 - ze^{i\theta})^{2(1 - a\_j)}},\tag{65}$$
then, there exists *δ* > 0 such that *Dn*−*<sup>j</sup> f* ∈ H *δ*+ <sup>1</sup> 2(1−*αj* ) .
Making *j* = *n* in Theorem 3, we have
**Corollary 3.** *If f* ∈ *A satisfies*
$$\left| \frac{D^{n+2}f(z)}{D^{n+1}f(z)} - 1 \right| < \frac{5\alpha - 2a^2 - 1}{2\alpha} \quad (z \in \mathbb{U}), \tag{66}$$
*for some real α* ( 1 <sup>3</sup> <sup>≤</sup> *<sup>α</sup>* <sup>≤</sup> <sup>1</sup> 2 ), *or*
$$\left| \frac{D^{n+2}f(z)}{D^{n+1}f(z)} - 1 \right| < \frac{\alpha - 2a^2 + 1}{2a} \quad (z \in \mathbb{U}), \tag{67}$$
*for some real α* ( 1 <sup>2</sup> <sup>≤</sup> *<sup>α</sup>* <sup>&</sup>lt; <sup>1</sup>), *then D<sup>n</sup> f* ∈ *S*0(*α*). *If*
$$f(z) \neq \frac{z}{(1 - ze^{i\theta})^{2(1 - \alpha\_n)}},\tag{68}$$
*then, there exists <sup>δ</sup>* <sup>&</sup>gt; <sup>0</sup>*, such that f* ∈ H*δ*<sup>+</sup> <sup>1</sup> 2(1−*αn*)
#### **3. Conclusions**
Inspired by the classic and well-known S˘al˘agean differential operator, a new operator is introduced in Definition 2. By applying this operator, a new class of functions is defined, denoted by *Sn*(*α*). It is shown that classes of starlike and convex functions of the order *α* are obtained for specific values of *n*. Some interesting problems concerning the class *Sn*(*α*) are discussed in the theorems and corollaries. One example is given as an application for special cases of *n* for the class *Sn*(*α*). The new operator defined in this paper can be used to introduce other certain subclasses of analytic functions. Quantum calculus can be also associated for future studies, as can be seen in paper [20] regarding the S˘al˘agean differential operator and involving symmetric S˘al˘agean differential operator in paper [21]. Symmetry properties can be investigated for this operator, taking the symmetric S˘al˘agean derivative investigated in [22] as inspiration.
.
**Author Contributions:** Conceptualization, S.O., H.Ö.G. and G.I.O.; Investigation, S.O., H.Ö.G. and G.I.O.; Methodology, S.O.; Writing—original draft, S.O.; Writing—review and editing, H.Ö.G. and G.I.O. All authors have read and agreed to the published version of the manuscript.
**Funding:** This research received no external funding.
**Institutional Review Board Statement:** Not applicable.
**Informed Consent Statement:** Not applicable.
**Data Availability Statement:** Not applicable.
**Conflicts of Interest:** The authors declare no conflict of interest.
#### **References**
### *Article* **Subclasses of Yamakawa-Type Bi-Starlike Functions Associated with Gegenbauer Polynomials**
**Gangadharan Murugusundaramoorthy 1,† and Teodor Bulboac˘a 2,\* ,†**
**Abstract:** In this paper, we introduce and investigate new subclasses (Yamakawa-type bi-starlike functions and another class of Lashin, both mentioned in the reference list) of bi-univalent functions defined in the open unit disk, which are associated with the Gegenbauer polynomials and satisfy subordination conditions. Furthermore, we find estimates for the Taylor–Maclaurin coefficients |*a*2| and |*a*3| for functions in these new subclasses. Several known or new consequences of the results are also pointed out.
**Keywords:** starlike and convex functions; hadamard product; subordination; bi-univalent functions; Fekete–Szeg˝o problem; Gegenbauer polynomials; Yamakawa-type bi-starlike functions
**MSC:** 30C45; 30C50
## **1. Introduction and Preliminaries**
In geometric function theory, there have been numerous interesting and fruitful usages of a wide variety of special functions, *q*-calculus and special polynomials; for example, the Fibonacci polynomials, the Faber polynomials, the Lucas polynomials, the Pell polynomials, the Pell–Lucas polynomials, and the Chebyshev polynomials of the second kind. The Horadam polynomials are potentially important in a variety of disciplines in the mathematical, physical, statistical, and engineering sciences. Gegenbauer polynomials or ultra spherical polynomials G*<sup>λ</sup> n* can be obtained using the Gram–Schmidt orthogonalization process for polynomials in the domain (−1, 1) with the weight factor 1 − ` 2 *λ*<sup>−</sup> <sup>1</sup> 2 , *λ* > − 1 2 . Also, G<sup>0</sup> *n* (`) is defined as lim *λ*→0 G*<sup>λ</sup> n* (`) *λ* , and for *λ* 6= 0 the resulting polynomial *Rn*(`) is multiplied by a number which makes the value at ` = 1 equal to (2*λ*)*n*/*n*! = 2*λ*(2*λ* + 1)(2*λ* + 2). . .(2*λ* + *n* − 1)/*n*!. For *λ* = 0 and *n* 6= 0, the value at ` <sup>=</sup> 1 is <sup>2</sup> *n* , while G<sup>0</sup> 0 (`) = 1.
The Gegenbauer polynomials (for details, see Kim et al. [1] and references cited therein) are given in terms of the Jacobi polynomials *P* (*ν*,*υ*) *<sup>n</sup>* , with *ν* = *υ* = *λ* − 1 2 , *λ* > − 1 2 , *λ* 6= 0 , defined by
$$\begin{split} \mathfrak{G}\_{n}^{\lambda}(\ell) &= \frac{\Gamma\left(\lambda + \frac{1}{2}\right) \Gamma(n + 2\lambda)}{\Gamma(2\lambda) \Gamma\left(n + \lambda + \frac{1}{2}\right)} P\_{n}^{\left(\lambda - \frac{1}{2}, \lambda - \frac{1}{2}\right)}(\ell) \\ &= \binom{n + 2\lambda - 1}{n} \sum\_{k=0}^{n} \frac{\binom{n}{k} (2\lambda + n)\_{k}}{\left(\lambda + \frac{1}{2}\right)\_{k}} \left(\frac{\ell - 1}{2}\right)^{k}, \end{split} \tag{1}$$
**Citation:** Murugusundaramoorthy, G.; Bulboac˘a, T. Subclasses of Yamakawa-Type Bi-Starlike Functions Associated with Gegenbauer Polynomials. *Axioms* **2022**, *11*, 92. https://doi.org/ 10.3390/axioms11030092
Academic Editors: Georgia Irina Oros and Kurt Bernardo Wolf
Received: 30 January 2022 Accepted: 22 February 2022 Published: 24 February 2022
**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.
**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).
where (*a*)*<sup>n</sup>* := *a*(*a* + 1)(*a* + 2). . .(*a* + *n* − 1), and (*a*)<sup>0</sup> := 1.
From (1), it follows that G*<sup>λ</sup> n* (`) is a polynomial of degree *n* with real coefficients, and G*<sup>λ</sup> n* (1) = *n* + 2*λ* − 1 *n* , while the leading coefficient of G*<sup>λ</sup> n* (`) is 2 *n n* + *λ* − 1 *n* . By the theory of Jacobi polynomials, for *µ* = *υ* = *λ* − 1 2 , with *λ* > − 1 2 , and *λ* 6= 0, we get
$$
\mathfrak{G}\_n^{\lambda}(-\ell) = (-1)^n \mathfrak{G}\_n^{\lambda}(\ell).
$$
It is easy to show that G*<sup>λ</sup> n* (`) is a solution of the Gegenbauer differential equation
$$(1 - \ell^2)y'' - (2\lambda)\ell y' + n(n + 2\lambda)y = 0\_\lambda$$
with ` = 0 an ordinary point; this means that we can express the solution in the form of a power series *y* = ∞ ∑ *n*=0 *an*` *n* , and the Rodrigues formula for the Gegenbauer polynomials is (see [2,3]) as follows:
$$\left(1 - \ell^2\right)^{\lambda - \frac{1}{2}} \mathfrak{G}\_n^{\lambda}(\ell) = \frac{(-2)^n (\lambda)\_n}{n! (n + 2\lambda)\_n} \left(\frac{d}{d\ell}\right)^n \left(1 - \ell^2\right)^{n + \lambda - \frac{1}{2}}$$
and the above relation can be easily derived from the properties of Jacobi polynomials. The generating function of Gegenbauer polynomials is given by (see [1,4])
$$\frac{2^{\lambda-\frac{1}{2}}}{(1-2\ell t+t^2)^{\frac{1}{2}}\left(1-\ell t+\sqrt{1-2\ell t+t^2}\right)^{\lambda-\frac{1}{2}}} = \frac{\left(\lambda-\frac{1}{2}\right)\_n}{(2\lambda)\_n} \mathfrak{G}\_n^{\lambda}(\ell)t^n. \tag{2}$$
and this equality can be derived from the generating function of Jacobi polynomials. From the above relation (2), we note that
$$\frac{1}{\left(1-2\ell t+t^{2}\right)^{\lambda}} = \sum\_{n=0}^{\infty} \mathfrak{G}\_{n}^{\lambda}(\ell)t^{n}, \; t \in \mathbb{C}, \; |t| < 1, \; \ell \in [-1,1], \; \lambda \in \left(-\frac{1}{2}, +\infty\right) \backslash \{0\}, \tag{3}$$
and the proof is given in [4] and Kim et al. [1] (also, see [5]) where the authors extensively studied many results from different perspectives. For *λ* = 1, the relation (3) gives the ordinary generating function for the Chebyshev polynomials, and for *λ* = 1 2 , we obtain the ordinary generating function for the Legendre polynomials (see also [6]).
In 1935, Robertson [7] proved an integral representation for the typically real-valued function class *T<sup>R</sup>* having the form
$$f(z) = z + \sum\_{n=2}^{\infty} a\_n z^n, \; z \in \Delta := \{ z \in \mathbb{C} : |z| < 1 \}, \tag{4}$$
which is holomorphic in the open unit disc ∆, real for *z* ∈ (−1, 1), and satisfies the condition
$$
\operatorname{Im} f(z) \operatorname{Im} z > 0, \; z \in \Delta \backslash (-1, 1).
$$
Namely, *f* ∈ *T<sup>R</sup>* if and only if it has the representation
$$f(z) = \int\_{-1}^{1} \frac{z}{1 - 2\ell z + z^2} \, d\mu\_{\prime} \, z \in \Delta\_{\prime}$$
where *µ* is a probability measure on [−1, 1]. The class *T<sup>R</sup>* has been extended in [8] to the class *TR*(*λ*), *λ* > 0, which was defined by
$$f(z) = \int\_{-1}^{1} \Phi\_{\ell}^{\lambda}(z) \, d\mu(\ell), \; z \in \Delta\_{\prime} \; -1 \le \ell \le 1,\tag{5}$$
where
$$\Phi\_{\ell}^{\lambda}(z) := \frac{z}{\left(1 - 2\ell z + z^2\right)^{\lambda}}, \ z \in \Delta, \ -1 \le \ell \le 1,\tag{6}$$
and *<sup>µ</sup>* is a probability measure on [−1, 1]. The function <sup>Φ</sup>*<sup>λ</sup>* ` (*z*) has the following Taylor– Maclaurin series expansion:
$$\Phi\_{\ell}^{\lambda}(z) = z + \mathfrak{G}\_1^{\lambda}(\ell)z^2 + \mathfrak{G}\_2^{\lambda}(\ell)z^3 + \mathfrak{G}\_3^{\lambda}(\ell)z^4 + \dots + \mathfrak{G}\_{n-1}^{\lambda}(\ell)z^n + \dots,\tag{7}$$
where G*<sup>λ</sup> n* (`) denotes the Gegenbauer (or ultra spherical) polynomials of order *λ* and degree *n* in `, which are generated by
$$\Phi\_\ell^\lambda(z) = \sum\_{n=0}^\infty \mathfrak{G}\_n^\lambda(\ell) z^n = z \left( 1 - 2\ell z + z^2 \right)^{-\lambda}.$$
In particular,
$$\mathfrak{G}\_0^{\lambda}(\ell) = 1, \quad \mathfrak{G}\_1^{\lambda}(\ell) = 2\lambda \ell, \quad \mathfrak{G}\_2^{\lambda}(\ell) = 2\lambda(\lambda + 1)\ell^2 - \lambda = 2(\lambda)\_2 \ell^2 - \lambda. \tag{8}$$
Of course, we have *TR*(1) ≡ *TR*, and if *f* given by (5) is written in the power expansion series (4), then we have
$$a\_n = \int\_{-1}^{1} \mathfrak{G}\_{n-1}^{\lambda}(\ell) \, d\mu(\ell).$$
One can easily see that the class *TR*(*λ*), *λ* > 0, is a compact and convex set in the linear space of holomorphic functions *f*(*z*) = *z* + ∞ ∑ *n*=2 *a<sup>n</sup> z <sup>n</sup>* which are holomorphic in ∆, endowed with the topology of local uniform convergence on compact subsets of ∆. The importance of the class *TR*(*λ*), *λ* > 0, follows as well from the paper of Hallenbeck [9], who studied the extreme points of some families of univalent functions and proved that
$$\cos \mathcal{S}\_R^\*(1-\lambda) = T\_R(\lambda), \quad \text{and} \quad \text{ext} \cos \mathcal{S}\_R^\*(1-\lambda) = \left\{ \frac{z}{\left(1-2\ell z + z^2\right)^{\lambda}} \colon \ell \in [-1;1] \right\},$$
where "co *A*" denotes the closed convex hull of *A*, "ext *A*" represents the set of the extremal points of *A*, while S ∗ *R* (*ϑ*) denotes the class of holomorphic functions given by (5), which are univalent and starlike of order *ϑ*, *ϑ* ∈ [0, 1), in ∆, and have real coefficients.
Let A represents the class of functions whose members are of the form
$$f(z) = z + \sum\_{n=2}^{\infty} a\_n z^n, \; z \in \Delta,\tag{9}$$
which are analytic in ∆, and let S be the subclass of A whose members are univalent in ∆. The Koebe one quarter theorem [10] ensures that the image of ∆ under every univalent function *<sup>f</sup>* ∈ A contains a disk of radius <sup>1</sup> 4 . Thus every univalent function *f* has an inverse *f* −1 satisfying
$$f^{-1}(f(z)) = z,\ (z \in \Delta) \quad \text{and} \quad f\left(f^{-1}(w)\right) = w,\ \left(|w| < r\_0(f),\ r\_0(f) \ge \frac{1}{4}\right).$$
A function *f* ∈ A is said to be bi-univalent in ∆ if both *f* and *f* <sup>−</sup><sup>1</sup> are univalent in ∆, and let Σ denote the class of bi-univalent functions defined in the unit disk ∆. Since *f* ∈ Σ
has the Maclaurin series given by (9), a computation shows that its inverse *g* = *f* <sup>−</sup><sup>1</sup> has the expansion
$$g(w) = f^{-1}(w) = w - a\_2 w^2 + \left(2a\_2^2 - a\_3\right) w^3 + \dots \tag{10}$$
We notice that the class Σ is not empty. For instance, the functions
$$f\_1(z) = \frac{z}{1 - z'} \quad f\_2(z) = \frac{1}{2} \log \frac{1 + z}{1 - z'} \quad f\_3(z) = -\log(1 - z)$$
with their corresponding inverses
$$f\_1^{-1}(w) = \frac{w}{1+w},\quad f\_2^{-1}(w) = \frac{e^{2w}-1}{e^{2w}+1},\quad f\_3^{-1}(w) = \frac{e^w-1}{e^w}$$
are elements of Σ. However, the Koebe function is not a member of Σ. Lately, Srivastava et al. [11] have essentially revived the study of analytic and bi-univalent functions; this was followed by such works as those of [12–17]. Several authors have introduced and examined subclasses of bi-univalent functions and obtained bounds for the initial coefficients (see [11–13,15]), bi-close-to-convex functions [18,19], and bi-prestarlike functions by Jahangiri and Hamidi [20].
Orthogonal polynomials have been broadly considered in recent years from various perceptions due to their importance in mathematical physics, mathematical statistics, engineering, and probability theory. Orthogonal polynomials that appear most often in applications are the classical orthogonal polynomials (Hermite polynomials, Laguerre polynomials, and Jacobi polynomials). The previously mentioned Fibonacci polynomials, Faber polynomials, the Lucas polynomials, the Pell polynomials, the Pell–Lucas polynomials, the Chebyshev polynomials of the second kind, and Horadam polynomials have been studied in several papers from a theoretical point of view and recently in the case of bi-univalent functions (see [21–28] also the references cited therein).
Here, in this article, we associate certain bi-univalent functions with Gegenbauer polynomials and then explore some properties of the class of bi-starlike functions based on earlier work of Srivastava et al. (also, see [11]). In addition, motivated by recent works by Murugusundaramoorthy et al. [29], Wannas [30] and Amourah et al. [31], we introduce a new subclass of the Yamakawa-type bi-starlike function class (see [32]) associated with Gegenbauer polynomials, obtain upper bounds of the initial Taylor coefficients |*a*2| and |*a*3| for the functions *f* ∈ GY<sup>Σ</sup> Φ*<sup>λ</sup>* ` defined by subordination, and consider the remarkable Fekete–Szeg˝o problem. We also provide relevant connections of our results with those of some earlier investigations.
First, we define a new subclass Yamakawa-type bi-starlike in the open unit disk, associated with Gegenbauer polynomials as below.
Unless otherwise stated, we let 0 ≤ *ϑ* ≤ 1, *λ* > 1 2 and ` ∈ 1 2 , 1 .
**Definition 1.** *For* 0 ≤ *ϑ* ≤ 1 *and* ` ∈ 1 2 , 1 *, a function f* ∈ Σ *of the form* (9) *is said to be in the class* GY<sup>Σ</sup> *ϑ*, Φ*<sup>λ</sup>* ` *if the following subordinations hold:*
$$\frac{f(z)}{(1-\theta)z+\theta z f'(z)} \prec \Phi\_\ell^\lambda(z),\tag{11}$$
*and*
$$\frac{g(w)}{(1-\theta)w + \theta w \emptyset'(w)} \prec \Phi\_\ell^\lambda(w) \tag{12}$$
*where z*, *<sup>w</sup>* <sup>∈</sup> <sup>∆</sup>*,* <sup>Φ</sup>*<sup>λ</sup>* ` *is given by* (6)*, and g* = *f* −1 *is given by* (10)*.*
By specializing the parameter *ϑ*, we state a new subclass of Yamakawa-type bi-starlike in the open unit disk, associated with Gegenbauer polynomials as below:
**Remark 1.** *For ϑ* = 1*, we get* YS<sup>∗</sup> Σ Φ*<sup>λ</sup>* ` := GY<sup>Σ</sup> 1, Φ*<sup>λ</sup>* ` *, thus f* ∈ YS<sup>∗</sup> Σ Φ*<sup>λ</sup>* ` *if f* ∈ Σ *and the following subordinations hold:*
$$\frac{f(z)}{zf'(z)} \prec \Phi\_\ell^\lambda(z) \quad and \quad \frac{g(w)}{wg'(w)} \prec \Phi\_\ell^\lambda(w)$$
*where z*, *w* ∈ ∆*, and g* = *f* −1 *is given by* (10)*.*
**Remark 2.** *For ϑ* = 0*, we get* N<sup>Σ</sup> Φ*<sup>λ</sup>* ` := GY<sup>Σ</sup> 0, Φ*<sup>λ</sup>* ` *, thus f* ∈ N<sup>Σ</sup> Φ*<sup>λ</sup>* ` *if f* ∈ Σ *and the following subordinations hold:*
$$\frac{f(z)}{z} \prec \Phi\_\ell^\lambda(z) \quad and \quad \frac{g'(w)}{w} \prec \Phi\_\ell^\lambda(w)$$
*where z*, *w* ∈ ∆ *and g* = *f* −1 *is given by* (10)*.*
Note that if in the above Remarks 1 and 2, we choose *λ* = 1 or *λ* = 1 2 , then we can state the new subclasses of YS<sup>∗</sup> Σ Φ*<sup>λ</sup>* ` and N<sup>Σ</sup> Φ*<sup>λ</sup>* ` related with Chebyshev polynomials and Legendre polynomials, respectively.
#### **2. Initial Taylor Coefficients Estimates for the Functions of** GY**<sup>Σ</sup>** *ϑ***, Φ***<sup>λ</sup>* `
To obtain our first results, we need the following lemma:
**Lemma 1** ([33], p. 172)**.** *Assume that ω*(*z*) = ∞ ∑ *n*=1 *ωnz n , z* ∈ U*, is an analytic function in* U *such that* |*ω*(*z*)| < 1 *for all z* ∈ U*. Then,*
$$|\omega\_1| \le 1, \quad |\omega\_n| \le 1 - |\omega\_1|^2, \ n = 2, 3, \dots$$
In the next result, we obtain the upper bounds for the modules of the first two coefficients for the functions that belong to the class GY<sup>Σ</sup> *ϑ*, Φ*<sup>λ</sup>* ` .
**Theorem 1.** *Let f given by* (9) *be in the class* GY<sup>Σ</sup> *ϑ*, Φ*<sup>λ</sup>* ` *. Then,*
$$|a\_2| \le \frac{2\lambda\ell\sqrt{2\lambda\ell}}{\sqrt{\left| (1-6\theta+6\theta^2)4\lambda^2\ell^2 - 2\left(2(\lambda)\_2\ell^2 - \lambda\right)(1-2\theta)^2 \right|}},\tag{13}$$
*and*
$$|a\_3| \le \frac{2(\lambda\ell)^2 \left(1 - 2\theta - 2\theta^2\right)}{|(1 - 3\theta)(1 - 2\theta)^2|} + \frac{2\lambda\ell}{|1 - 3\theta|},\tag{14}$$
*where ϑ* 6= 1 3
*.*
**Proof.** Let *f* ∈ GY<sup>Σ</sup> *ϑ*, Φ*<sup>λ</sup>* ` and *g* = *f* −1 . From the definition in Formulas (11) and (12), we have
$$\frac{f(z)}{(1-\theta)z+\theta z f'(z)} = \Phi\_\ell^\lambda(\mu(z))\tag{15}$$
and
$$\frac{g(w)}{(1-\theta)w + \theta w g'(w)} = \Phi\_\ell^\lambda(v(w)),\tag{16}$$
where the functions *u* and *v* are of the form
$$\mu(z) = c\_1 z + c\_2 z^2 + \dots,\tag{17}$$
and
$$w(w) = d\_1 w + d\_2 w^2 + \dots,\tag{18}$$
are analytic in ∆ with *u*(0) = 0 = *v*(0), and |*u*(*z*)| < 1, |*v*(*w*)| < 1, for all *z*, *w* ∈ ∆. From Lemma 1 it follows that
$$|c\_{j}| \le 1 \quad \text{and} \quad |d\_{j}| \le 1, \text{ for all } j \in \mathbb{N}. \tag{19}$$
Replacing (17) and (18) in (15) and (16), respectively, we have
$$\frac{f(z)}{(1-\theta)z + \theta z f'(z)} = 1 + \mathfrak{G}\_1^{\lambda}(\ell)u(z) + \mathfrak{G}\_2^{\lambda}(\ell)u^2(z) + \dots \tag{20}$$
and
$$\frac{\mathfrak{g}(w)}{(1-\theta)w + \theta w \mathfrak{g}'(w)} = 1 + \mathfrak{G}\_1^{\lambda}(\ell)v(w) + \mathfrak{G}\_2^{\lambda}(\ell)v^2(w) + \dots \tag{21}$$
In view of (9) and (10), from (20) and (21), we obtain
$$\begin{aligned} 1 + (1 - 2\theta)a\_2 z + \left[ (1 - 3\theta)a\_3 - 2\theta(1 - 2\theta)a\_2^2 \right] z^2 + \dots \\ &= 1 + \mathfrak{G}\_1^\lambda(\ell) c\_1 z + \left[ \mathfrak{G}\_1^\lambda(\ell) c\_2 + \mathfrak{G}\_2^\lambda(\ell) c\_1^2 \right] z^2 + \dots, \end{aligned}$$
and
$$\begin{aligned} 1 - (1 - 2\theta)(\alpha)a\_2w + \left\{ \left(1 - 4\theta + 2\theta^2\right)a\_2^2 - (1 - 3\lambda)a\_3 \right\} w^2 + \dots \\ &= 1 + \mathfrak{G}\_1^{\lambda}(\ell)d\_1w + \left[\mathfrak{G}\_1^{\lambda}(\ell)d\_2 + \mathfrak{G}\_2^{\lambda}(\ell)d\_1^2\right]w^2 + \dots \end{aligned}$$
which yields the following relations:
$$(1 - 2\theta)a\_2 = \mathfrak{G}\_1^{\lambda}(\ell)c\_1\tag{22}$$
$$(1 - 3\theta)a\_3 - 2\theta(1 - 2\theta)a\_2^2 = \mathfrak{G}\_1^\lambda(\ell)c\_2 + \mathfrak{G}\_2^\lambda(\ell)c\_1^2. \tag{23}$$
and
$$-(1 - 2\theta)a\_2 = \mathfrak{G}\_1^{\lambda}(\ell)d\_{1\prime} \tag{24}$$
$$-(1 - 3\theta)a\_3 + \left(1 - 4\theta + 2\theta^2\right)a\_2^2 = \mathfrak{G}\_1^\lambda(\ell)d\_2 + \mathfrak{G}\_2^\lambda(\ell)d\_1^2. \tag{25}$$
From (22) and (24), it follows that
$$c\_1 = -d\_{1\prime} \tag{26}$$
and
$$2(1 - 2\theta)^2 a\_2^2 = [\mathfrak{G}\_1^{\lambda}(\ell)]^2 (c\_1^2 + d\_1^2),$$
$$a\_2^2 = \frac{[\mathfrak{G}\_1^{\lambda}(\ell)]^2}{2(1 - 2\theta)^2} (c\_1^2 + d\_1^2) \tag{27}$$
Adding (23) and (25), using (27), we obtain
$$a\_2^2 = \frac{[\mathfrak{G}\_1^\lambda(\ell)]^3 (c\_2 + d\_2)}{(1 - 6\theta + 6\theta^2)[\mathfrak{G}\_1^\lambda(\ell)]^2 - 2(1 - 2\theta)^2 \mathfrak{G}\_2^\lambda(\ell)}.\tag{28}$$
Applying (19) for the coefficients *c*<sup>2</sup> and *d*<sup>2</sup> and using (8), we obtain the Inequality (13).
By subtracting (25) from (23), using (26) and (27), we get
$$\begin{split} a\_{3} &= \frac{\mathfrak{G}\_{1}^{\lambda}(\ell)(c\_{2}-d\_{2})}{2(1-3\vartheta)} + \frac{\left(1-2\vartheta-2\vartheta^{2}\right)[\mathfrak{G}\_{1}^{\lambda}(\ell)]^{2}}{2(1-3\vartheta)} a\_{2}^{2} \\ &= \frac{\left(1-2\vartheta-2\vartheta^{2}\right)[\mathfrak{G}\_{1}^{\lambda}(\ell)]^{2}(c\_{1}^{2}+d\_{1}^{2})}{4(1-3\vartheta)(1-2\vartheta)^{2}} + \frac{\mathfrak{G}\_{1}^{\lambda}(\ell)(c\_{2}-d\_{2})}{2(1-3\vartheta)}. \end{split} \tag{29}$$
Using (8) and once again applying (19) for the coefficients *c*1, *c*2, *d*1, and *d*2, we deduce the required Inequality (14).
By taking *ϑ* = 0 or *ϑ* = 1 and ` ∈ (0, 1), one can easily state the upper bounds for |*a*2| and |*a*3| for the function classes GYΣ(0, Φ) =: N<sup>Σ</sup> Φ*<sup>λ</sup>* ` and GYΣ(1, Φ) =: YS<sup>∗</sup> Σ Φ*<sup>λ</sup>* ` , respectively, as follows:
**Remark 3.** *Let f given by* (9) *be in the class* N<sup>Σ</sup> Φ*<sup>λ</sup>* ` *. Then,*
$$|a\_2| \le \frac{2\lambda\ell\sqrt{2\lambda\ell}}{\sqrt{\left|4\lambda^2\ell^2 - 2\left(2(\lambda)\_2\ell^2 - \lambda\right)\right|}}\sqrt{2}$$
*and*
$$|a\_3| \le \mathbf{2}(\lambda \ell)^2 + 2\lambda \ell.$$
**Remark 4.** *Let f given by* (9) *be in the class* YS<sup>∗</sup> Σ Φ*<sup>λ</sup>* ` *. Then,*
$$|a\_2| \le \frac{2\lambda\ell\sqrt{2\lambda\ell}}{\sqrt{\left|4\lambda^2\ell^2 - 2\left(2(\lambda)\_2\ell^2 - \lambda\right)\right|}}\sqrt{2}$$
*and*
$$|a\_3| \le 3(\lambda \ell)^2 + \lambda \ell.$$
**Remark 5.** *Let f given by* (9) *be in the class* GY<sup>∗</sup> Σ *ϑ*, Φ<sup>1</sup> ` *. Then,*
$$|a\_2| \le \frac{2\ell\sqrt{2\ell}}{\sqrt{\left| (1 - 6\theta + 6\theta^2) 4\ell^2 - 2(4\ell^2 - 1)(1 - 2\theta)^2 \right|}}\sqrt{$$
*and*
$$|a\_3| \le \frac{2\ell^2 \left(1 - 2\vartheta - 2\vartheta^2\right)}{|(1 - 3\vartheta)(1 - 2\vartheta)^2|} + \frac{2\ell}{|1 - 3\vartheta|}\sqrt{2}$$
*where ϑ* 6= 1 3 *.*
**Remark 6.** *Let f given by* (9) *be in the class* GY<sup>∗</sup> Σ *ϑ*, Φ 1/2 ` *. Then, for* ` 6= 1 √ 2 *,*
$$|a\_2| \le \frac{\ell \sqrt{\ell}}{\sqrt{\left| (1 - 6\vartheta + 6\vartheta^2) \ell^2 - (3\ell^2 - 1)(1 - 2\vartheta)^2 \right|}} \le \frac{\ell \sqrt{\ell}}{\sqrt{\ell}}$$
*and*
$$|a\_3| \le \frac{\ell^2 \left(1 - 2\vartheta - 2\vartheta^2\right)}{2|(1 - 3\vartheta)(1 - 2\vartheta)^2|} + \frac{\ell}{|1 - 3\vartheta|^{\prime}}$$
*where ϑ* 6= 1 3 *.*
In the above Remarks 3 and 4, by fixing *λ* = 1 and *λ* = 1 2 , we obtain the new estimates of |*a*2| and |*a*3| for the function classes YS<sup>∗</sup> Σ Φ*<sup>λ</sup>* ` and N<sup>Σ</sup> Φ*<sup>λ</sup>* ` related with Chebyshev polynomials and Legendre polynomials, respectively.
#### **3. Fekete–Szeg˝o Inequality for the Function Class** GY**<sup>Σ</sup>** *ϑ***, Φ***<sup>λ</sup>* `
Due to the result of Zaprawa [34], in this section, we obtain the Fekete–Szeg˝o inequality for the function classes GY<sup>Σ</sup> *ϑ*, Φ*<sup>λ</sup>* ` .
**Theorem 2.** *Let f given by* (9) *be in the class* GY<sup>Σ</sup> *ϑ*, Φ*<sup>λ</sup>* ` *, and µ* ∈ R*. Then, we have*
$$|a\_3 - \mu a\_2^2| \le \begin{cases} \frac{2\lambda\ell}{|1 - 3\mathcal{O}|} \prime & \text{if} \quad |h(\mu)| \le \frac{1}{2|1 - 3\mathcal{O}|}\\ 4\lambda\ell |h(\mu)| \prime & \text{if} \quad |h(\mu)| \ge \frac{1}{2|1 - 3\mathcal{O}|} \end{cases}$$
*where*
$$h(\mu) := \frac{2\lambda\ell^2 \left[2\lambda^2 \ell^2 \left(1 - 2\theta - 2\theta^2\right) - \mu(1 - 3\theta)\right]}{(1 - 3\theta)\left\{2\lambda\ell^2 (1 - 6\theta + 6\theta^2) - (1 - 2\theta)^2 \left[2(\lambda + 1)\ell^2 - 1\right]\right\}'}$$
*and ϑ* 6= 1 3 *.*
**Proof.** If *f* ∈ GY<sup>Σ</sup> *ϑ*, Φ*<sup>λ</sup>* ` is given by (9), from (28) and (29), we have
$$\begin{split} a\_3 - \mu a\_2^2 &= \frac{\mathfrak{G}\_1^\lambda(\ell)(c\_2 - d\_2)}{2(1 - 3\theta)} + \left( \frac{(1 - 2\theta - 2\theta^2)[\mathfrak{G}\_1^\lambda(\ell)]^2}{2(1 - 3\theta)} - \mu \right) a\_2^2 \\ &= \frac{\mathfrak{G}\_1^\lambda(\ell)(c\_2 - d\_2)}{2(1 - 3\theta)} + \left( \frac{(1 - 2\theta - 2\theta^2)[\mathfrak{G}\_1^\lambda(\ell)]^2}{2(1 - 3\theta)} - \mu \right) \\ &\times \frac{[\mathfrak{G}\_1^\lambda(\ell)]^3 (c\_2 + d\_2)}{(1 - 6\theta + 6\theta^2)[\mathfrak{G}\_1^\lambda(\ell)]^2 - 2(1 - 2\theta)^2 \mathfrak{G}\_2^\lambda(\ell)} \\ &= \mathfrak{G}\_1^\lambda(\ell) \left[ \left( h(\mu) + \frac{1}{2(1 - 3\theta)} \right) c\_2 + \left( h(\mu) - \frac{1}{2(1 - 3\theta)} \right) d\_2 \right] \end{split}$$
where
$$h(\mu) = \frac{(\left(1 - 2\theta - 2\theta^2\right)[\mathfrak{G}\_1^{\lambda}(\ell)]^2 - 2\mu(1 - 3\theta))[\mathfrak{G}\_1^{\lambda}(\ell)]^3}{2(1 - 3\theta)\{(1 - 6\theta + 6\theta^2)[\mathfrak{G}\_1^{\lambda}(\ell)]^2 - 2(1 - 2\theta)^2\mathfrak{G}\_2^{\lambda}(\ell)\}}.$$
Now, by using (8)
$$a\_3 - \mu a\_2^2 = 2\lambda \ell \left[ \left( h(\mu) + \frac{1}{2(1 - 3\vartheta)} \right) c\_2 + \left( h(\mu) - \frac{1}{2(1 - 3\vartheta)} \right) d\_2 \right]^2$$
where
$$\begin{split} h(\mu) &= \frac{2\lambda^2 \ell^2 \left[ 2\lambda^2 \ell^2 \left( 1 - 2\theta - 2\theta^2 \right) - \mu (1 - 3\theta) \right]}{(1 - 3\theta) \left\{ 2\lambda^2 \ell^2 (1 - 2\theta + 2\theta^2) - \lambda (1 - 2\theta)^2 [2(\lambda + 1)\ell^2 - 1] \right\}} \\ &= \frac{2\lambda \ell^2 \left[ 2\lambda^2 \ell^2 (1 - 2\theta - 2\theta^2) - \mu (1 - 3\theta) \right]}{(1 - 3\theta) \left\{ 2\lambda \ell^2 (1 - 6\theta + 6\theta^2) - (1 - 2\theta)^2 [2(\lambda + 1)\ell^2 - 1] \right\}} \end{split}$$
Therefore, in view of (8) and (19), we conclude that the required inequality holds.
#### **4. The Subclass** M**<sup>Σ</sup>** *τ***, Φ***<sup>λ</sup>* ` **of Bi-Univalent Functions**
In [35] Obradovi´c et al. gave some criteria for univalence expressed by Re *f* 0 (*z*) > 0 for the linear combination
$$
\tau \left( 1 + \frac{z f''(z)}{f'(z)} \right) + (1 - \tau) \frac{1}{f'(z)}, \ \tau \ge 1, \ z \in \Delta.
$$
Based on the above definitions, recently, Lashin [36] introduced and studied new subclasses of the bi-univalent function. In our further discussions, unless otherwise stated, we let *τ* ≥ 1, *λ* > 1 2 , and ` ∈ 1 2 , 1 .
**Definition 2.** *A function f* ∈ Σ *given by* (9) *is said to be in the class* M<sup>Σ</sup> *τ*, Φ*<sup>λ</sup>* ` *if it satisfies the conditions*
$$\left(\pi \left(1 + \frac{z f''(z)}{f'(z)}\right) + (1 - \pi) \frac{1}{f'(z)} \prec \Phi\_\ell^\lambda(z)\right) \tag{30}$$
*and*
$$\left(\tau \left(1 + \frac{w g''(w)}{g'(w)}\right) + (1 - \tau) \frac{1}{g'(w)} \prec \Phi\_\ell^\lambda(w) \right. \tag{31}$$
*where <sup>τ</sup>* <sup>≥</sup> <sup>1</sup>*, z*, *<sup>w</sup>* <sup>∈</sup> <sup>∆</sup>*,* <sup>Φ</sup>*<sup>λ</sup>* ` *is given by* (6)*, and the function g* = *f* −1 *is given by* (10)*.*
**Remark 7.** *For the particular case τ* = 1*, a function f* ∈ Σ *given by* (9) *is said to be in the class* M<sup>Σ</sup> Φ*<sup>λ</sup>* ` =: K<sup>Σ</sup> Φ*<sup>λ</sup>* ` *if it satisfies the subordination relations*
$$1 + \frac{z f''''(z)}{f'(z)} \prec \Phi\_\ell^\lambda(z) \quad \text{and} \quad 1 + \frac{w g''''(w)}{g'(w)} \prec \Phi\_\ell^\lambda(w),$$
*<sup>z</sup>*, *<sup>w</sup>* <sup>∈</sup> <sup>∆</sup>*,* <sup>Φ</sup>*<sup>λ</sup>* ` *is given by* (6)*, and g* = *f* −1 *is given by* (10)*.*
**Theorem 3.** *Let f be given by* (9) *and f* ∈ M<sup>Σ</sup> *τ*, Φ*<sup>λ</sup>* ` *, with τ* ≥ 1*. Then,*
$$|a\_2| \le \min\left\{ \frac{\lambda\ell}{2(2\tau - 1)}; \frac{\lambda\ell\sqrt{2\lambda\ell}}{2\sqrt{| (1+\tau)\lambda^2\ell^2 - 4(2\tau - 1)^2[2\ell^2(\lambda)\_2 - \lambda]|}} \right\}.\tag{32}$$
*and*
$$|a\_3| \le \min\left\{\frac{2\lambda\ell}{3(3\tau-1)} + \frac{\lambda^2\ell^2}{4(2\tau-1)^2};$$
$$\frac{2\lambda\ell}{3(3\tau-1)} + \frac{2\lambda^3\ell^3}{\left| (1+\tau)\lambda^2\ell \right|\_1^2 - (2\tau-1)^2 \left[2\ell^2(\lambda)\_2 - \lambda \right] \right|} \}.$$
**Proof.** *f* ∈ M<sup>Σ</sup> *τ*, Φ*<sup>λ</sup>* ` , from (30) and (31) it follows that
$$\tau \left( 1 + \frac{z f''(z)}{f'(z)} \right) + (1 - \tau) \frac{1}{f'(z)} = \Phi\_\ell^\lambda(\mu(z)),\tag{33}$$
and
$$
\tau \left( 1 + \frac{w g''(w)}{g'(w)} \right) + (1 - \tau) \frac{1}{g'(w)} = \Phi\_\ell^\lambda(v(w)),
\tag{34}
$$
where the functions *u* and *v* are analytic in ∆ with *u*(0) = 0 = *v*(0), such that |*u*(*z*)| < 1, |*v*(*w*)| < 1, for all *z*, *w* ∈ ∆, and are of the form (17) and (18), respectively.
From (33) and (34), we have
$$\begin{aligned} 1 + 2(2\tau - 1)a\_2 z + \left[ 3(3\tau - 1)a\_3 + 4(1 - 2\tau)\_2 a\_2^2 \right] z^2 + \dots \\ &= 1 + \mathfrak{G}\_1^{\lambda}(\ell) c\_1 z + \left[ \mathfrak{G}\_1^{\lambda}(\ell) c\_2 + \mathfrak{G}\_2^{\lambda}(\ell) c\_1^2 \right] z^2 + \dots \ge 1 \end{aligned}$$
and
$$\begin{aligned} 1 - 2(2\tau - 1)a\_2 w + \left[2(5\tau - 1)a\_2^2 - 3(3\tau - 1)a\_3\right] w^2 - \dots \\ &= 1 + \mathfrak{G}\_1^\lambda(\ell) d\_1 w + \left[\mathfrak{G}\_1^\lambda(\ell) d\_2 + \mathfrak{G}\_2^\lambda(\ell) d\_1^2\right] w^2 + \dots \dots \end{aligned}$$
and equating the coefficients of the above two relations, we get
$$2(2\pi - 1)a\_2 = \mathfrak{G}\_1^{\lambda}(\ell)c\_{1\prime} \tag{35}$$
$$2\mathfrak{F}(3\tau - 1)a\_3 + 4(1 - 2\tau)a\_2^2 = \mathfrak{G}\_1^\lambda(\ell)c\_2 + \mathfrak{G}\_2^\lambda(\ell)c\_{1'}^2\tag{36}$$
and
$$-2(2\tau - 1)a\_2 = \mathfrak{G}\_1^{\lambda}(\ell)d\_{1\prime} \tag{37}$$
$$2(5\tau - 1)a\_2^2 - 3(3\tau - 1)a\_3 = \mathfrak{G}\_1^\lambda(\ell)d\_2 + \mathfrak{G}\_2^\lambda(\ell)d\_1^2. \tag{38}$$
From (35) and (37), we get
$$p\_1 = -q\_1 \tag{39}$$
From (35), by using the Inequality (19) for the coefficients *c<sup>j</sup>* and *d<sup>j</sup>* , from (8), we have
$$|a\_2| \le \frac{\mathfrak{G}\_1^{\lambda}(\ell)}{\mathfrak{Z}(2\tau - 1)} = \frac{\lambda \ell}{(2\tau - 1)}.$$
Furthermore,
$$(8(2\tau -1)^2 a\_2^2 = \left(\mathfrak{G}\_1^{\lambda}(\ell)\right)^2 \left(c\_1^2 + d\_1^2\right)\lambda$$
that is,
$$a\_2^2 = \frac{\left(\mathfrak{G}\_1^{\lambda}(\ell)\right)^2 \left(c\_1^2 + d\_1^2\right)}{8(2\pi - 1)^2}. \tag{40}$$
Thus, from the Inequality (19) and using (8), we obtain
$$|a\_2| \le \frac{\mathfrak{G}\_1^{\lambda}(\ell)}{4(2\tau - 1)} = \frac{\lambda \ell}{2(2\tau - 1)}.\tag{41}$$
Now, from (36), (38) and using (40), we get
$$\left[2(1+\tau)\left(\mathfrak{G}\_1^{\lambda}(\ell)\right)^2 - 8(2\tau - 1)^2 \mathfrak{G}\_2^{\lambda}(\ell)\right] a\_2^2 = \left(\mathfrak{G}\_1^{\lambda}(\ell)\right)^3 (c\_2 + d\_2). \tag{42}$$
Thus, according to (42), we obtain
$$a\_2^2 = \frac{\left(\mathfrak{G}\_1^\lambda(\ell)\right)^3 (c\_2 + d\_2)}{2(1+\tau)\left(\mathfrak{G}\_1^\lambda(\ell)\right)^2 - 8(2\tau - 1)^2 \mathfrak{G}\_2^\lambda(\ell)}\sqrt{$$
hence,
$$|a\_2| \le \frac{\lambda \ell \sqrt{2\lambda \ell}}{2\sqrt{\left| (1+\tau)\lambda^2 \ell^2 - 4(2\tau - 1)^2 [2\ell^2(\lambda)\_2 - \lambda] \right|}},\tag{43}$$
and the Inequality (32) is proved.
From (36), (38) and using (39), we get
$$a\_3 = \frac{\mathfrak{G}\_1^{\lambda}(\ell)(c\_2 - d\_2)}{6(3\tau - 1)} + a\_{2\prime}^2 \tag{44}$$
which implies
$$|a\_3| \le \frac{2\lambda\ell}{3(3\tau - 1)} + |a\_2^2|. \tag{45}$$
From this inequality, using (41), we obtain
$$|a\_3| \le \frac{2\lambda\ell}{3(3\tau - 1)} + \frac{\lambda^2 \ell^2}{4(2\tau - 1)^2}.$$
Combining (45) and (43), it follows that
$$|a\_3| \le \frac{2\lambda\ell}{\Im(3\tau - 1)} + \frac{2\lambda^3\ell^3}{\left| (1+\tau)\lambda^2\ell \right|\_1^2 - (2\tau - 1)^2 [2\ell^2(\lambda)\_2 - \lambda] |}.$$
Motivated by the result of Zaprawa [34], we discuss the Fekete–Szeg˝o inequality [37] for the functions *f* ∈ M<sup>Σ</sup> *τ*, Φ*<sup>λ</sup>* ` .
**Theorem 4.** *For ν* ∈ R*, let f* ∈ M<sup>Σ</sup> *τ*, Φ*<sup>λ</sup>* ` *be given by* (9)*. Then,*
$$\left| a\_3 - \nu a\_2^2 \right| \le \begin{cases} \frac{2\lambda \ell}{3(3\pi - 1)}, & \text{if} \quad |h(\nu)| \le \frac{1}{6(3\pi - 1)},\\ 4|h(\nu)|, & \text{if} \quad |h(\nu)| \ge \frac{1}{6(3\pi - 1)}, \end{cases}$$
*where*
$$h(\nu) = \frac{(1 - \nu)\lambda\ell^2}{4\left\{(1 + \tau)\lambda\ell^2 - (2\tau - 1)^2[2\ell^2(\lambda + 1) - 1]\right\}}.\tag{46}$$
**Proof.** If *f* ∈ M<sup>Σ</sup> *τ*, Φ*<sup>λ</sup>* ` be given by (9), from (44) we have
$$a\_3 - \nu a\_2^2 = \frac{\mathfrak{G}\_1^\lambda(\ell)(c\_2 - d\_2)}{6(3\tau - 1)} + (1 - \nu)a\_2^2. \tag{47}$$
.
By substituting (42) in (47), we obtain
$$\begin{split} a\_3 - \nu a\_2^2 &= \frac{\mathfrak{G}\_1^\lambda(\ell)(c\_2 - d\_2)}{6(3\tau - 1)} + \frac{(1 - \nu)\left(\mathfrak{G}\_1^\lambda(\ell)\right)^3(c\_2 + d\_2)}{2(1 + \tau)\left(\mathfrak{G}\_1^\lambda(\ell)\right)^2 - 8(2\tau - 1)^2\mathfrak{G}\_2^\lambda(\ell)} \\ &= \mathfrak{G}\_1^\lambda(\ell) \left[ \left( h(\nu) + \frac{1}{6(3\tau - 1)} \right) c\_2 + \left( h(\nu) - \frac{1}{6(3\tau - 1)} \right) d\_2 \right] .\end{split}$$
where
$$h(\nu) = \frac{(1-\nu)\left(\mathfrak{G}\_1^{\lambda}(\ell)\right)^2}{2(1+\tau)\left(\mathfrak{G}\_1^{\lambda}(\ell)\right)^2 - 8(2\tau - 1)^2\mathfrak{G}\_2^{\lambda}(\ell)}$$
From (8), it follows
$$a\_3 - \nu a\_2^2 = 2\lambda \ell \left[ \left( h(\nu) + \frac{1}{6(3\tau - 1)} \right) c\_2 + \left( h(\nu) - \frac{1}{6(3\tau - 1)} \right) d\_2 \right],\tag{48}$$
where the function *h* is given by (46). Hence, by using the triangle inequality for the modulus of (48) together with (19), we get our result.
For *ν* = 1 the above theorem reduces to the following special case:
**Remark 8.** *If f* ∈ M<sup>Σ</sup> *τ*, Φ*<sup>λ</sup>* ` *is given by* (9)*, then*
$$\left| a\_3 - a\_2^2 \right| \le \frac{2\lambda\ell}{3(3\tau - 1)}.$$
#### **5. Conclusions**
Yamakawa-type bi-starlike functions related with the Gegenbauer polynomials are defined for the first time, and initial Taylor coefficients and Fekete–Szeg˝o inequality are obtained. Further, by fixing *λ* = 1 or *λ* = 1 2 , the Gegenbauer polynomials lead to the Chebyshev polynomials and the Legendre polynomials, respectively. Hence, our results represent a new study of the Yamakawa family of bi-starlike functions associated with Chebyshev and Legendre polynomials, which are also not considered in the literature. We have left this as an exercise to interested readers.
**Author Contributions:** Conceptualization, T.B. and G.M.; methodology, T.B. and G.M.; validation, T.B. and G.M.; formal analysis, T.B. and G.M.; investigation, T.B. and G.M.; resources, T.B. and G.M.; writing—original draft preparation, T.B. and G.M.; writing—review and editing, T.B. and G.M.; supervision, T.B. and G.M.; project administration, T.B. and G.M. All authors have read and agreed to the published version of the manuscript.
**Funding:** This research received no external funding.
**Institutional Review Board Statement:** Not applicable.
**Informed Consent Statement:** Not applicable.
**Data Availability Statement:** Not applicable.
**Acknowledgments:** The authors are grateful to the reviewers of this article who gave valuable comments and advice that allowed us to revise and improve the content of the paper.
**Conflicts of Interest:** The authors declare no conflict of interest.
#### **References**
### *Article* **Hadamard Product Properties for Certain Subclasses of** *p***-Valent Meromorphic Functions**
**Alaa H. El-Qadeem 1,\* and Ibrahim S. Elshazly <sup>2</sup>**
**Abstract:** We study the Hadamard product features of certain subclasses of *p*-valent meromorphic functions defined in the punctured open-unit disc using the q-difference operator. For functions belonging to these subclasses, we obtained certain coefficient estimates and inclusion characteristics. Furthermore, linkages between the results given here and those found in previous publications are highlighted.
**Keywords:** analytic function; univalent function; starlike function; convex function; meromorphic function; q-difference operator
**MSC:** 30C45; 30D30
#### **1. Introduction**
Let M*<sup>p</sup>* stand for the class of functions of the form:
*f*(*z*) = *z* <sup>−</sup>*<sup>p</sup>* + ∞ ∑ *k*=−*p*+1 *akz k* , (1)
which are analytic in the perforated unit disc *U*<sup>∗</sup> = *U*\{0} = {*z* : *z* ∈ C : 0 < |*z*| < 1}. The class M*<sup>p</sup>* refers to the a class of *p*-valent meromorphic functions. It is worth noting that M<sup>1</sup> = M, which is the class of univalent meromorphic functions. If the function *g* ∈ M*<sup>p</sup>* is given by
$$g(z) = z^{-p} + \sum\_{k=-p+1}^{\infty} b\_k z^k \rho$$
then the Hadamard product (or convolution) of *f* and *g* is provided by
$$(f\*g)(z) = z^{-p} + \sum\_{k=-p+1}^{\infty} a\_k b\_k z^k = (g\*f)(z).$$
Interesting traits such as coefficient estimates, subordination relations and univalence features related some subclasses of *p*-valent functions were obtained in [1–3] (see also, [4]). With the help of the q-differential operator, a new subclass of meromorphic multivalent functions in the Janowski domain were introduced by Bakhtiar et al. in [5] (see also, [6]). Moreover, new subclasses of meromorphically *p*-valent functions were defined using q-derivative operator and investigations related to geometric properties of the class are conducted in [7–9].
If *f* and *g* are analytic in the open unit disc *U*, we say that *f* is subordinate to *g*, written as *f* ≺ *g* in *U* or *f*(*z*) ≺ *g*(*z*)(*z* ∈ *U*), if there exists a Schwarz function *w*(*z*), which (by
**Citation:** El-Qadeem, A.H.; Elshazly, I.S. Hadamard Product Properties for Certain Subclasses of *p*-Valent Meromorphic Functions. *Axioms* **2022**, *11*, 172. https://doi.org/ 10.3390/axioms11040172
Academic Editor: Georgia Irina Oros
Received: 21 March 2022 Accepted: 11 April 2022 Published: 13 April 2022
**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.
**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).
definition) is analytic in *U* with *w*(0) = 0 and |*w*(*z*)| < 1, (*z* ∈ *U*) such that *f*(*z*) = *g*(*w*(*z*)) (*z* ∈ *U*) [10].
For 0 < *q* < 1, the q-difference operator, which was introduced by Jackson [11], is characterised with
$$\partial\_{\emptyset} f(z) = \begin{cases} \frac{f(qz) - f(z)}{(q-1)z}, & z \neq 0, \\\\ f'(0), & z = 0. \end{cases}$$
The Jackson q-difference operator is another name for the q-difference operator. Additionally, for *f* given by (1), one can write
$$
\partial\_q f(z) = -q^{-p} [p]\_q z^{-p-1} + \sum\_{k=-p+1}^{\infty} [k]\_q a\_k z^{k-1} (z \in \mathcal{U}^\*), \tag{2}
$$
where [*k*] *<sup>q</sup>* = 1 − *q k* /(<sup>1</sup> − *<sup>q</sup>*) is the well-known q-bracket, lim*q*→1<sup>−</sup> [*k*] *<sup>q</sup>* = *k* and lim*q*→1<sup>−</sup> *<sup>∂</sup><sup>q</sup> <sup>f</sup>*(*z*) = *<sup>f</sup>* 0 (*z*).
Now, for *<sup>n</sup>* <sup>∈</sup> <sup>N</sup><sup>0</sup> <sup>=</sup> <sup>N</sup> ∪ {0}, we define the operator <sup>D</sup>*<sup>n</sup> p*,*q* : M*<sup>p</sup>* −→ M*<sup>p</sup>* with the help of the q-difference operator, as follows:
$$\begin{aligned} \mathfrak{D}\_{p,q}^{0}f(z) &= f(z),\\ \mathfrak{D}\_{p,q}^{1}f(z) &= z^{-p}\partial\_{q}\left(z^{p+1}f(z)\right),\\ \mathfrak{D}\_{p,q}^{n}f(z) &= z^{-p}\partial\_{q}\left(z^{p+1}\mathfrak{D}\_{p,q}^{n-1}f(z)\right)(n\in\mathbb{N}).\end{aligned}$$
then
$$\mathfrak{D}\_{p,q}^{\mathfrak{n}}f(z) = z^{-p} + \sum\_{k=-p+1}^{\infty} [k+p+1]\_q^{\mathfrak{n}} a\_k z^k \ (n \in \mathbb{N}\_0) \,. \tag{3}$$
which satisfies the following recurrence relation:
$$q^{p+1}z\partial\_{\mathfrak{q}}\left(\mathfrak{D}\_{p,\mathfrak{q}}^{\mathfrak{n}}f(z)\right) = \mathfrak{D}\_{p,\mathfrak{q}}^{\mathfrak{n}+1}f(z) - [p+1]\_{\mathfrak{q}}\mathfrak{D}\_{p,\mathfrak{q}}^{\mathfrak{n}}f(z). \tag{4}$$
**Definition 1.** *Utilising the q-derivative ∂<sup>q</sup> f*(*z*)*, the subclasses*MS<sup>∗</sup> *p*,*q* (*A*, *B*) *and* MK*p*,*q*(*A*, *B*) *are introduced as follows:*
$$\mathcal{MS}\_{p,q}^\*(A,B) = \left\{ f \in \mathcal{M}\_p \colon \frac{-q^p z \partial\_q f(z)}{[p]\_q f(z)} \prec \frac{1+Az}{1+Bz} \right\},\tag{5}$$
$$(0 < q < 1; -1 \le B < A \le 1; z \in \mathcal{U}),$$
*and*
$$\mathcal{M}\mathcal{K}\_{p,q}(A,B) = \left\{ f \in \mathcal{M}\_p \colon \frac{-q^p \mathfrak{d}\_q(z\mathfrak{d}\_q f(z))}{[p]\_q \mathfrak{d}\_q f(z)} \prec \frac{1+Az}{1+Bz}, z \in \mathcal{U} \right\},\tag{6}$$
$$(0 < q < 1; -1 \le B < A \le 1; z \in \mathcal{U}).$$
Using (5) and (6), we have the following equivalence relation:
$$f(z) \in \mathcal{M} \mathcal{K}\_{p,q}(A,B) \Longleftrightarrow -\frac{q^p z \partial\_q f(z)}{[p]\_q} \in \mathcal{M} \mathcal{S}^\*\_{p,q}(A,B). \tag{7}$$
**Remark 1.** *We list the following subclasses by specialising the parameters p, q, A and B: (i)* MS<sup>∗</sup> *p*,*q* (1 − 2*α*, −1) = MS<sup>∗</sup> *p*,*q* (*α*) = { *<sup>f</sup>* ∈ M*<sup>p</sup>* : Re − *q p z∂q f*(*z*) [*p*] *q f*(*z*) > *α*; 0 ≤ *α* < 1, *z* ∈ *U*} *the subclass of p-valent meromorphic q-starlike functions, and*MK*p*,*q*(1−2*α*, −1) = MK*p*,*q*(*α*) =
{ *<sup>f</sup>* ∈ M*<sup>p</sup>* : Re − *q <sup>p</sup>∂<sup>q</sup>* (*z∂<sup>q</sup> f*(*z*)) [*p*] *q ∂q f*(*z*) > *α*; 0 ≤ *α* < 1, *z* ∈ *U*} *the subclass of p-valent meromorphic q-convex functions; (ii)* MS<sup>∗</sup> 1,*q* (1 − 2*α*, −1) = MS<sup>∗</sup> *q* (*α*) = { *<sup>f</sup>* ∈ M: Re − *qz∂q f*(*z*) *f*(*z*) > *α*; 0 ≤ *α* < 1, *z* ∈ *U*} *the subclass of meromorphic q-starlike functions, and* MK1,*q*(1 − 2*α*, −1) = MK*q*(*α*) = { *f* ∈ M: Re − *q∂<sup>q</sup>* (*z∂<sup>q</sup> f*(*z*)) *∂q f*(*z*) > *α*; 0 ≤ *α* < 1, *z* ∈ *U*} *the subclass of meromorphic q-convex functions; (iii)* lim*q*→1<sup>−</sup> MS<sup>∗</sup> *p*,*q* (*A*, *B*) = MS<sup>∗</sup> *p* (*A*, *B*) = { *f* ∈ M*<sup>p</sup>* : − *z f* 0 (*z*) *p f*(*z*) <sup>≺</sup> <sup>1</sup>+*Az* <sup>1</sup>+*Bz* ; −1 ≤ *B* < *A* ≤ 1, *<sup>z</sup>* <sup>∈</sup> *<sup>U</sup>*}*, and* lim*q*→1<sup>−</sup> MK*p*,*q*(*A*, *<sup>B</sup>*) = MK*p*(*A*, *<sup>B</sup>*) = { *<sup>f</sup>* ∈ M*<sup>p</sup>* : <sup>−</sup> <sup>1</sup> *p* 1 + *z f* 00(*z*) *f* 0(*z*) ≺ 1+*Az* <sup>1</sup>+*Bz* ; −1 ≤ *B* < *A* ≤ 1, *z* ∈ *U*}, *were introduced and studied by Ali and Ravichandran [12]; (iv)* lim*q*→1<sup>−</sup> MS<sup>∗</sup> 1,*q* (1 − 2*α*, −1) = MS<sup>∗</sup> (*α*) = { *<sup>f</sup>* ∈ M: Re − *z f* 0 (*z*) *f*(*z*) > *α*; 0 ≤ *α* < 1, *<sup>z</sup>* <sup>∈</sup> *<sup>U</sup>*}*, and* lim*q*→1<sup>−</sup> MK*p*,*q*(<sup>1</sup> <sup>−</sup> <sup>2</sup>*α*, <sup>−</sup>1) = MK(*α*) = { *<sup>f</sup>* ∈ M: Re −1 − *z f* 00(*z*) *f* 0(*z*) > *α*; 0 ≤ *α* < 1, *z* ∈ *U*}*, were introduced and studied by Kaczmarski [13]; (v)* lim*q*→1<sup>−</sup> MS<sup>∗</sup> 1,*q* (1, −1) = MS<sup>∗</sup> , *and* lim*q*→1<sup>−</sup> MK1,*q*(1, −1) = MK*, which are wellknown function classes of meromorphic starlike and meromorphic convex functions, respectively; see*
**Definition 2.** *For n* ∈ N<sup>0</sup> *and* 0 < *q* < 1*, we define the following subclasses:*
*Pommerenke [14], Clunie [15] and Miller [16] for more details.*
$$\mathcal{MS}\_{p,q}^{\*}(\mathfrak{n}; A, \mathcal{B}) = \left\{ f \in \mathcal{M}\_{p} : \mathfrak{D}\_{p,q}^{n} f(z) \in \mathcal{MS}\_{p,q}^{\*}(A, \mathcal{B}) \right\},\tag{8}$$
$$(\mathfrak{n} \in \mathbb{N}\_{0}; 0 < q < 1; -1 \le B < A \le 1; z \in \mathcal{U}),$$
and
$$\mathcal{M}\mathcal{K}\_{p,q}(n;A,B) = \left\{ f \in \mathcal{M}\_p : \mathfrak{D}\_{p,q}^n f(z) \in \mathcal{M}\mathcal{K}\_{p,q}(A,B) \right\},\tag{9}$$
$$(n \in \mathbb{N}\_0; 0 < q < 1; -1 \le B < A \le 1; z \in \mathcal{U}).$$
It is easy to show that
$$f(z) \in \mathcal{M} \mathcal{K}\_{p,q}(\mathfrak{n}; A, \mathcal{B}) \iff -\frac{q^p z \partial\_q f(z)}{[p]\_q} \in \mathcal{M} \mathcal{S}^\*\_{p,q}(\mathfrak{n}; A, \mathcal{B}).\tag{10}$$
There is extensive literature dealing with convolution properties of different families of analytic and meromorphic functions; for details, see [17–23]. More recently, the quantum derivative was utilised by Seoudy and Aouf [24] (see also [25]) to introduce the convolution features for certain classes of analytic functions. Here, we use the quantum derivative to obtain some convolution properties of the meromorphic functions. For this purpose, we defined the new classes MS<sup>∗</sup> *p*,*q* (*A*, *B*) and MK*p*,*q*(*A*, *B*). The convolution results are followed by some consequences such as necessary and sufficient conditions, the estimates of coefficients and inclusion characteristics of the subclasses MS<sup>∗</sup> *p*,*q* (*n*; *A*, *B*) and MK*p*,*q*(*n*; *A*, *B*).
#### **2. Convolution Properties**
**Theorem 1.** *The function f given by* (1) *is in the class* MS<sup>∗</sup> *p*,*q* (*A*, *B*)*, if and only if*
$$z^p \left[ f(z) \* \frac{1 + (\mathcal{C} - q)z}{z^p (1 - z)(1 - qz)} \right] \neq 0 \ (z \in \mathcal{U}), \tag{11}$$
*for all*
$$\mathcal{C} = \frac{B + e^{-i\theta}}{A - [p]\_q B - q[p-1]\_q e^{-i\theta}} ; \theta \in [0, 2\pi) ,\tag{12}$$
*and also for C* = 0*.*
**Proof.** It is simple to check the following two equalities
$$f(z) \* \frac{1}{z^p(1-z)} = f(z) \tag{13}$$
and
$$f(z) \* \left(\frac{1}{qz^p(1-z)(1-qz)} - \frac{[1+p]\_q}{qz^p(1-z)}\right) = q^p z \partial\_q f(z) \tag{14}$$
In view of (5), *f* ∈ MS<sup>∗</sup> *p*,*q* (*A*, *B*), if and only if (1.4) holds. Since the function <sup>1</sup>+*Az* 1+*Bz* is analytic function on *U*, it follows that *f*(*z*) 6= 0, *z* ∈ *U*<sup>∗</sup> ; that is *z p f*(*z*) 6= 0, *z* ∈ *U*, and using the first identity of (13). That is the same as saying that the relation (11) is satisfied for *C* = 0. According to the concept of subordination of two functions in (14), there exists an analytic function *w*(*z*) in *U* with *w*(0) = 0, |*w*(*z*)| < 1 in such a way that
$$\frac{-q^p z \partial\_q f(z)}{[p]\_q f(z)} = \frac{1 + Aw(z)}{1 + Bw(z)} \ (z \in \mathcal{U})\_{\prime\prime}$$
which leads to
$$\frac{1 - q^p z \partial\_q f(z)}{[p]\_q f(z)} \neq \frac{1 + Ae^{i\theta}}{1 + Be^{i\theta}} \left( f(z) \neq 0, z \in \mathcal{U}; 0 \le \theta < 2\pi \right).$$
or
$$z^p \left[ \left( q^p z \partial\_{\theta} f(z) \right) \left( 1 + B e^{i\theta} \right) + [p]\_q f(z) \left( 1 + A e^{i\theta} \right) \right] \neq 0 \tag{15}$$
We may now deduce the following from (13)–(15):
$$z^p \left[ \left( f(z) \ast \frac{1 - [1+p]\_q (1-qz)}{qz^p (1-z)(1-qz)} \right) \left( 1 + Be^{i\theta} \right) + \left( 1 + Ae^{i\theta} \right) \left( f(z) \ast \frac{1}{z^p (1-z)} \right) \right] \neq 0,$$
$$z^p \left[ f(z) \ast \left( \frac{\left( 1 - [1+p]\_q + q[1+p]\_q z \right) \left( 1 + Be^{i\theta} \right) + q(1-qz) \left( 1 + Ae^{i\theta} \right)}{qz^p (1-z)(1-qz)} \right) \right] \neq 0,$$
but 1 − [1 + *p*] *<sup>q</sup>* = −*q*[*p*] *q* ; then, the condition became
$$z^p \left[ f(z) \ast \left( \frac{q\left( [1+p]\_q z - [p]\_q \right) \left( 1 + Be^{i\theta} \right) + q(1-qz)\left( 1 + Ae^{i\theta} \right)}{qz^p (1-z)(1-qz)} \right) \right] \neq 0,$$
or,
$$z^p \left[ f(z) \ast \left( \frac{\left( [1+p]\_q z - [p]\_q \right) \left( 1 + B e^{i\theta} \right) + (1-qz) \left( 1 + A e^{i\theta} \right)}{z^p (1-z)(1-qz)} \right) \right] \neq 0,$$
or, equivalent to
$$z^p \left[ f(z) \ast \left( \frac{1 - [p]\_q + \left( A - [p]\_q \mathcal{B} \right) e^{i\theta} + \left( [1 + p]\_q - q + \left( [1 + p]\_q \mathcal{B} - qA \right) e^{i\theta} \right) z}{z^p (1 - z)(1 - qz)} \right) \right] \neq 0,$$
or,
$$z^p \left[ f(z) \ast \left( \frac{-q[p-1]\_q + \left( A - [p]\_q B \right) e^{i\theta} + \left( [1+p]\_q - q + \left( [1+p]\_q B - qA \right) e^{i\theta} \right) z}{z^p (1-z)(1-qz)} \right) \right] \neq 0,$$
or,
$$z^p \left[ f(z) \ast \left( \frac{1 + \frac{\left( [1+p]\_q - q + \left( [1+p]\_q B - qA \right) e^{i\theta} \right) z}{-q[p-1]\_q + \left( A - [p]\_q B \right) e^{i\theta}}}{z^p (1-z)(1-qz)} \left( \left( A - [p]\_q B \right) e^{i\theta} - q[p-1]\_q \right) \right) \right] \neq 0,$$
by dividing both sides by the non-zero quantity *A* − [*p*] *q B e <sup>i</sup><sup>θ</sup>* <sup>−</sup> *<sup>q</sup>*[*<sup>p</sup>* <sup>−</sup> <sup>1</sup>] *q* , then we have
$$z^p \left[ f(z) \ast \left( \frac{1 + \frac{\left( [1+p]\_q - q + \left( [1+p]\_q B - qA \right) e^{i\theta} \right) z}{-q[p-1]\_q + \left( A - [p]\_q B \right) e^{i\theta}}}{z^p (1-z)(1-qz)} \right) \right] \neq 0,$$
which is the same as
$$z^p \left[ f(z) \ast \left( \frac{1 + \left( \frac{[1+p]\_q - q + \left( [1+p]\_q B - qA \right) \epsilon^{i\theta} + q \left( -q[p-1]\_q + \left( A - [p]\_q B \right) \epsilon^{i\theta} \right) }{-q[p-1]\_q + \left( A - [p]\_q B \right) \epsilon^{i\theta}} - q \right) z} - q \right) z \right) \right] \neq 0,$$
or,
$$z^p \left[ f(z) \* \left( \frac{1 + \left( \frac{[1+p]\_q - q - q^2[p-1]\_q + \left( [1+p]\_q - q[p]\_q \right) B e^{i\theta}}{-q[p-1]\_q + \left( A - [p]\_q B \right) e^{i\theta}} - q \right) z}{z^p (1-z)(1-qz)} \right) \right] \neq 0,$$
$$\gamma\_1 = \gamma\_2, \quad \gamma\_3 = \gamma\_4, \quad \gamma\_5 = \gamma\_6, \quad \dots, \quad \gamma\_8 = \gamma\_9$$
but [1 + *p*] *<sup>q</sup>* − *q* − *q* 2 [*p* − 1] *<sup>q</sup>* = [1 + *p*] *<sup>q</sup>* − *q*[*p*] *<sup>q</sup>* = 1, then the convolution condition became
$$z^p \left[ f(z) \ast \left( \frac{1 + \left( \frac{e^{-i\theta} + B}{A - [p]\_q B - q[p-1]\_q e^{-i\theta}} - q \right) z}{z^p (1 - z)(1 - qz)} \right) \right] \neq 0,$$
This leads to (11), proving the first part of Theorem 1.
In contrast, because (11) holds for *C* = 0, it follows that *z p f*(*z*) 6= 0 for all *z* ∈ *U*, and hence the function.
$$\wp(z) = \frac{-q^p z \partial\_q f(z)}{[p]\_q f(z)} \rho$$
is analytic in *U* (i.e., it is regular at *z*<sup>0</sup> = 0, with *ϕ*(0) = 1). We obtain that because the assumption (11) is equivalent to (15), as shown in the first section of the proof.
$$\frac{1 - q^p z \partial\_{\boldsymbol{q}} f(z)}{[p]\_q f(z)} \neq \frac{1 + Ae^{i\theta}}{1 + Be^{i\theta}} \ (\theta \in [0, 2\pi), f(z) \neq 0, z \in \mathsf{U}),\tag{16}$$
if we denote
$$
\psi(z) = \frac{1 + Az}{1 + Bz'} \tag{17}
$$
therefore *ϕ*(*U*) ∩ *ψ*(*∂U*) = *φ*, with the help of the relation (16). Thus, the simply connected domain *ϕ*(*U*) is included in a connected component of C\*ψ*(*∂U*). As a result, a connected component of C\*ψ*(*∂U*) includes the simply connected domain *ϕ*(*U*). The fact that *ϕ*(0) = *ψ*(0) and the univalence of the function *ψ* lead to the conclusion that
*ϕ*(*z*) ≺ *ψ*(*z*). This completes the proof of the second item of Theorem 1 by representing the subordination (5), i.e., *f* ∈ MS<sup>∗</sup> *p*,*q* (*A*, *B*).
**Remark 2.** *(i) We obtain the results obtained in the paper of Aouf et al. in [17] (Theorem 4, with λ* = 0 *and b* = 1*) by putting p* = 1 *and q* → 1 − *in Theorem 1. See also, Bulboac˘a et al. [20] (Theorem 1, with b* = 1*) and El-Ashwah [21] (Theorem 1, with p* = 1*);*
*(ii) Putting p* = 1*, q* → 1 <sup>−</sup>*, A* = 1 *and B* = −1 *in Theorem 1, we obtain the result of Aouf et al. [18] (Theorem 1, with b* = *m* = 1*).*
In Theorem 1, we have the following corollary if *A* = 1 − 2*α* and *B* = −1.
**Corollary 1.** *The function f defined by* (1) *is in the class* MS<sup>∗</sup> *p*,*q* (*α*)*, if and only if*
$$z^p \left[ f(z) \ast \frac{1 + \left( \frac{\left(1 + q^2 \langle p - 1 \rangle\_q \right) \varepsilon^{-i\theta} - q \left(1 - 2a + \langle p \rangle\_q \right)}{1 - 2a + \langle p \rangle\_q - q \langle p - 1 \rangle\_q \varepsilon^{-i\theta}} \right) z}{z^p (1 - z)(1 - qz)} \right] \neq 0 \ (z \in \mathcal{U}),$$
*Taking q* → 1 <sup>−</sup>*, A* = 1 − 2*α and B* = −1 *in Theorem 1, we obtain the following corollary.*
**Corollary 2.** *The function f expressed in* (1) *belongs to* MS<sup>∗</sup> *p* (*α*)*, if and only if*
$$z^p \left[ f(z) \* \frac{1 + \left[ \frac{2(1 - a) + p\left(e^{-i\theta} - 1\right)}{1 - 2a + p - (p - 1)e^{-i\theta}} \right] z}{z^p \left(1 - z\right)^2} \right] \neq 0 \ (z \in \mathcal{U})\_p$$
**Theorem 2.** *The function f of the form* (1) *is a member of the class* MK*p*,*q*(*A*, *B*)*, if and only if*
$$z^p \left[ f(z) \ast \frac{1 - \frac{\left(1 - q^{p+2}\right) - q \left(1 - q^{p-1}\right) \left(\mathbb{C} - q\right)}{1 - q^p} z - \frac{q \left(1 - q^{p+1}\right) \left(\mathbb{C} - q\right)}{1 - q^p} z^2}{z^p (1 - z) (1 - qz) (1 - q^2 z)} \right] \neq 0 \ (z \in \mathcal{U}), \tag{18}$$
*for all C defined by* (12)*, and also for C* = 0*.*
**Proof.** If
$$g(z) = \frac{1 + (\mathcal{C} - q)z}{z^p (1 - z)(1 - qz)},\tag{19}$$
then
$$-\frac{q^p z \partial\_q g(z)}{[p]\_q} = \frac{-q^p z}{[p]\_q} \left[ \frac{1}{(q-1)z} (g(qz) - g(z)) \right]\_q$$
which leads to
$$1 - \frac{q^p z \partial\_q g(z)}{[p]\_q} = \frac{1 - \left(\frac{\left(1 - q^{p+2}\right) - q\left(1 - q^{p-1}\right)(\mathbb{C} - q)}{1 - q^p}\right)z - \left(\frac{q\left(1 - q^{p+1}\right)(\mathbb{C} - q)}{1 - q^p}\right)z^2}{z^p (1 - z)(1 - qz)(1 - q^2 z)}\tag{20}$$
The following identity remains true for two functions, *f* and *g*, which belong to M*p*.
$$\left(-\frac{q^p z \partial\_q f(z)}{[p]\_q}\right) \* g(z) = f(z) \* \left(-\frac{q^p z \partial\_q g(z)}{[p]\_q}\right). \tag{21}$$
Now, by using equivalence relation (7) and Theorem 1, the proof can be achieved by applying (20) and (21).
**Remark 3.** *(i) Putting p* = 1 *and q* → 1 − *in Theorem 2, we arrive at the results of Aouf et al. [17] (Theorem 6, with λ* = 0 *and b* = 1*) and Bulboac˘a et al. [20] (Theorem 2, with b* = 1*), and El-Ashwah [21] (Theorem 2, with p* = 1*);*
*(ii) Putting p* = 1*, q* → 1 <sup>−</sup>*, A* = 1 *and B* = −1 *in Theorem 2, we reach the conclusion of Aouf et al. [18] (Theorem 3, with b* = *m* = 1*).*
As a result, we have the following corollary by taking *A* = 1 − 2*α* and *B* = −1 in Theorem 2.
**Corollary 3.** *The function f* ∈ MK*p*,*q*(*α*)*, if and only if*
$$z^p \left[ f(z) \* \frac{1 - Dz - Ez^2}{z^p (1 - z)(1 - qz)(1 - q^2 z)} \right] \neq 0 \ (z \in \mathcal{U}),$$
*where*
$$D = \frac{\left(1 - q^{p+2}\right) - q\left(1 - q^{p-1}\right) \left(\frac{\left(1 + q^2[p-1]\_q\right)e^{-i\theta} - q\left(1 - 2a + [p]\_q\right)}{1 - 2a + [p]\_q - q[p-1]\_q e^{-i\theta}}\right)}{1 - q^p}$$
*and*
$$E = \frac{q\left(1 - q^{p+1}\right)\left(\left(1 + q^2[p-1]\_q\right)e^{-i\theta} - q\left(1 - 2\alpha + [p]\_q\right)\right)}{\left(1 - q^p\right)\left(1 - 2\alpha + [p]\_q - q[p-1]\_q e^{-i\theta}\right)}.$$
As a result, we have the following corollary by taking *q* → 1 <sup>−</sup>, *A* = 1 − 2*α* and *B* = −1 in Theorem 2.
**Corollary 4.** *The function f* ∈ MK*p*(*α*)*, if and only if*
$$z^p \left[ f(z) \ast \frac{1 - \frac{2p(1 - 2a + p) - \left(2p^2 - p - 1\right)e^{-i\theta}}{p(1 - 2a + p) - p(p - 1)e^{-i\theta}} z - \frac{(p + 2)\left(pe^{-i\theta} - (1 - 2a + p)\right) - 1}{p(1 - 2a + p) - p(p - 1)e^{-i\theta}} z^2}{z^p (1 - z)^3} \right] \neq 0 \text{ } (z \in \mathcal{U}).$$
**Theorem 3.** *The following are necessary and sufficient requirements for the function f* ∈ M*<sup>p</sup> to be in the class* MS<sup>∗</sup> *p*,*q* (*n*; *A*, *B*)*:*
$$1 + \sum\_{k=-p+1}^{\infty} [k+p+1]\_q^n a\_k z^{k+p} \neq 0 \ (z \in \mathcal{U}), \tag{22}$$
,
*or*
$$1 + \sum\_{k=-p+1}^{\infty} \left( [k+p]\_q \mathbb{C} + 1 \right) [k+p+1]\_q^n a\_k z^{k+p} \neq 0 \ (z \in \mathcal{U}), \tag{23}$$
*where C is defined by* (12)*.*
**Proof.** Let *f* ∈ M*p*, then, by using Theorem 1 and (8) we have *f* ∈ MS<sup>∗</sup> *p*,*q* (*n*; *A*, *B*), if and only if
$$z^p \left[ \left( \mathfrak{D}\_q^n f \right)(z) \* \frac{1 + (\mathbb{C} - q)z}{z^p (1 - z)(1 - qz)} \right] \neq 0 \ (z \in \mathsf{U}), \tag{24}$$
*for all C* = *<sup>B</sup>*+*<sup>e</sup>* −*iθ A*−[*p*] *q B*−*q*[*p*−1] *q e*−*i<sup>θ</sup>* ; *θ* ∈ [0, 2*π*), *and also for C* = 0. Since
$$\frac{1 + (0 - q)z}{z^p (1 - z)(1 - qz)} = z^{-p} + \sum\_{k = -p + 1}^{\infty} z^k. \tag{25}$$
by using (3) and (25) in (24) in case of *C* = 0, then we can obtain (22).
Similarly, it can be shown that
$$\frac{1 + (\mathcal{C} - q)z}{z^p (1 - z)(1 - qz)} = z^{-p} + \sum\_{k = -p + 1}^{\infty} \left( [k + p]\_q \mathcal{C} + 1 \right) z^k \tag{26}$$
then using (3) and (26) in (24), we can obtain (23). The proof is complete.
The next theorem can be established using the same method, and the proof is eliminated.
**Theorem 4.** *The following are necessary and sufficient requirements for the function f* ∈ M*<sup>p</sup> to be in the class* MK*p*,*q*(*n*; *A*, *B*)*:*
$$1 - \sum\_{k=-p+1}^{\infty} q[k]\_q [k+p+1]\_q^n a\_k z^{k+p} \neq 0 \ (z \in \mathcal{U}), \tag{27}$$
*or*
$$(1 - \sum\_{k=-p+1}^{\infty} q[k]\_q \left( [k+p]\_q \mathbb{C} + 1 \right) [k+p+1]\_q^n a\_k z^{k+p} \neq 0 \ (z \in \mathcal{U}).\tag{28}$$
#### **3. Estimates of Coefficients and Inclusion Characteristics**
In this section, as an application of Theorems 3 and 4, we introduce some estimates of the coefficients *a<sup>k</sup>* (*k* ≥ −*p* + 1) of functions of the form (1) which belong to the two main classes MS<sup>∗</sup> *p*,*q* (*n*; *A*, *B*) and MK*p*,*q*(*n*; *A*, *B*), respectively. Moreover, we give the inclusion relationships of the two classes.
**Theorem 5.** *If the function f* ∈ M*<sup>p</sup> fulfills the inequalities*
$$\sum\_{k=-p+1}^{\infty} [k+p+1]\_q^n |a\_k| < 1,\tag{29}$$
*and*
$$\sum\_{k=-p+1}^{\infty} \left( [k+p]\_q |\mathbb{C}| + 1 \right) [k+p+1]\_q^n |a\_k| < 1,\tag{30}$$
*then f* ∈ MS<sup>∗</sup> *p*,*q* (*n*; *A*, *B*).
#### **Proof.** According to (29), a simple calculation shows that
$$\begin{aligned} \left| 1 + \sum\_{k=-p+1}^{\infty} [k+p+1]\_q^n a\_k z^{k+p} \right| &\geq 1 - \left| \sum\_{k=-p+1}^{\infty} [k+p+1]\_q^n a\_k z^{k+p} \right| \\ &\geq 1 - \sum\_{k=-p+1}^{\infty} [k+p+1]\_q^n |a\_k| |z|^{k+p} \\ &> 1 - \sum\_{k=-p+1}^{\infty} [k+p+1]\_q^n |a\_k| > 0 \end{aligned}$$
which leads to satisfaction of (22), then *f* ∈ MS<sup>∗</sup> *p*,*q* (*n*; *A*, *B*). Similarly, using the assumption (30), we conclude that
$$\begin{aligned} & \left| 1 + \sum\_{k=-p+1}^{\infty} \left( [k+p]\_q \mathbb{C} + 1 \right) [k+p+1]\_q^n a\_k z^{k+p} \right| \\ & \ge 1 - \left| \sum\_{k=-p+1}^{\infty} \left( [k+p]\_q \mathbb{C} + 1 \right) [k+p+1]\_q^n a\_k z^{k+p} \right| \\ & \ge 1 - \sum\_{k=-p+1}^{\infty} \left( [k+p]\_q |\mathbb{C}| + 1 \right) [k+p+1]\_q^n |a\_k| |z|^{k+p} \\ & > 1 - \sum\_{k=-p+1}^{\infty} \left( [k+p]\_q |\mathbb{C}| + 1 \right) [k+p+1]\_q^n |a\_k| > 0, \end{aligned}$$
which shows that (23) holds true and *f* ∈ MS<sup>∗</sup> *p*,*q* (*n*; *A*, *B*); the proof is finished.
Similarly, results regarding MK*p*,*q*(*n*; *A*, *B*) can be introduced as follows:
**Theorem 6.** *If the function f* ∈ M*<sup>p</sup> fulfills the inequalities*
$$\sum\_{k=-p+1}^{\infty} q[k]\_q [k+p+1]\_q^n |a\_k| < 1,\tag{31}$$
*and*
$$\sum\_{k=-p+1}^{\infty} q[k]\_q \left( [k+p]\_q |\mathbb{C}| + 1 \right) [k+p+1]\_q^n |a\_k| < 1,\tag{32}$$
*then f* ∈ MK*p*,*q*(*n*; *A*, *B*)*.*
Now, using the appropriate technique due to Ahuja [26], we introduce the inclusion relationships of MS<sup>∗</sup> *p*,*q* (*n*; *A*, *B*) and MK*p*,*q*(*n*; *A*, *B*), respectively.
**Theorem 7.** *If n* ∈ N*o, then*
$$
\mathcal{MS}\_{p,q}^\*(n+1;A,B) \subset \mathcal{MS}\_{p,q}^\*(n;A,B). \tag{33}
$$
**Proof.** If *f* ∈ MS<sup>∗</sup> *p*,*q* (*n* + 1; *A*, *B*), then using Theorem 3, we can write
$$1 + \sum\_{k=-p+1}^{\infty} [k+p+1]\_q^{n+1} a\_k z^{k+p} \neq 0 \ (z \in \mathcal{U}), \tag{34}$$
or
$$1 + \sum\_{k=-p+1}^{\infty} \left( [k+p]\_q \mathbb{C} + 1 \right) [k+p+1]\_q^{n+1} a\_k z^{k+p} \neq 0 \ (z \in \mathcal{U}), \tag{35}$$
but (34) and (35) can be written as follows:
$$\left(1+\sum\_{k=-p+1}^{\infty} [k+p+1]\_q z^{k+p}\right) \ast \left(1+\sum\_{k=-p+1}^{\infty} [k+p+1]\_q^n a\_k z^{k+p}\right) \neq 0,\tag{36}$$
and
$$\left(1+\sum\_{k=-p+1}^{\infty} [k+p+1]\_q z^{k+p}\right) \ast \left(1+\sum\_{k=-p+1}^{\infty} \left([k+p]\_q \mathcal{C}+1\right) [k+p+1]\_q^n a\_k z^{k+p}\right) \neq 0. \tag{37}$$
Let us really define the function
$$h\_1(z) = 1 + \sum\_{k=-p+1}^{\infty} [k+p+1]\_q z^{k+p}.\tag{38}$$
We note that the assumption that *h*1(*z*) = 0 leads to |*z*| > 1, Thus, we deduce that *h*1(*z*) 6= 0. Using the property that if *h*<sup>1</sup> ∗ *g* 6= 0 and *h*<sup>1</sup> 6= 0, then *g* 6= 0. Thus from (36) and (37) and using the function *h*1(*z*) 6= 0, we obtain
$$1 + \sum\_{k=-p+1}^{\infty} [k+p+1]\_q^n a\_k z^{k+p} \neq 0,\tag{39}$$
and
$$\left(1+\sum\_{k=-p+1}^{\infty} \left( [k+p]\_q \mathbb{C} + 1 \right) [k+p+1]\_q^n a\_k z^{k+p} \neq 0,\tag{40}$$
then Theorem 3 tells us that *f* ∈ MS<sup>∗</sup> *p*,*q* (*n*; *A*, *B*).
The following theorem gives the inclusion relationship regarding MK*p*,*q*(*n*; *A*, *B*).
**Theorem 8.** *For n* ∈ N0, *we have*
$$
\mathcal{M}\mathcal{K}\_{p,\emptyset}(n+1;A,B) \subset \mathcal{M}\mathcal{K}\_{p,\emptyset}(n;A,B).\tag{41}
$$
*Our results in Theorems 7 and 8 above can be utilised to introduce the following consequences.*
**Corollary 5.** *Suppose that m* = *n* + 1, *n* + 2, . . .(*n* ∈ N0)*. Then*
$$f \in \mathcal{MS}\_{p,q}^\*(m; A, B) \Longrightarrow f \in \mathcal{MS}\_{p,q}^\*(n; A, B).$$
*Equivalently, if*
*then*
$$\text{The first-order coupling between the two-dimensional } \mathcal{N} \text{-matrices is the only possible } \mathcal{N} \text{-matrices with } \mathcal{N} = \{0, 1, 2, \dots, N\} \text{ and } \mathcal{N} = \{0, 1, 2, \dots, N\}.$$
*p*,*q*
(*A*, *B*),
*f* ∈ MS<sup>∗</sup> *p*,*q* (*n*; *A*, *B*).
*f*(*z*) ∈ MS<sup>∗</sup>
**Corollary 6.** *Suppose that m* = *n* + 1, *n* + 2, . . .(*n* ∈ N0)*. Then*
D*<sup>m</sup> q*
D*<sup>m</sup> q*
$$f \in \mathcal{M} \mathcal{K}\_{p,q}(\mathfrak{m}; A, B) \Longrightarrow f \in \mathcal{M} \mathcal{K}\_{p,q}(\mathfrak{n}; A, B).$$
*Equivalently, if*
*then*
$$f \in \mathcal{M} \mathcal{K}\_{p,q}(n; A, B).$$
#### **4. Conclusions**
We have defined a new operator on the set of meromorphically multivalent functions. With the help of this operator, we introduced the new subclasses MK*p*,*q*(*n*; *A*, *B*) and MS<sup>∗</sup> *p*,*q* (*n*; *A*, *B*). The study was concentrated on convolution conditions. Our suggestions for future studies on these subclasses is to use them in studies involving the theories of differential subordination and superordination. Additionally, one can define the results concerning the calculation of the bounds of coefficients of the bi-univalent functions, also obtaining the Fekete–Szegö functionals.
**Author Contributions:** Formal analysis and methodology, A.H.E.-Q.; resources, I.S.E. All authors have read and agreed to the published version of the manuscript.
$$\text{The first-order coupling between the two-dimensional } \mathcal{N} \text{-matrices is the only possible } \mathcal{N} \text{-matrices with } \mathcal{N} = \{0, 1, 2, \dots, N\} \text{ and } \mathcal{N} = \{0, 1, 2, \dots, N\}.$$
*f*(*z*) ∈ MK*p*,*q*(*A*, *B*),
$$\dots \dots \dots \dots \dots \dots \dots \dots \dots$$
**Funding:** The authors would like to thank the Common First Year Research Unit at King Saud University for giving us the funds for this article.
**Institutional Review Board Statement:** Not applicable.
**Informed Consent Statement:** Not applicable.
**Data Availability Statement:** Not applicable.
**Acknowledgments:** The authors would like to give thanks for the help of HM Abbas.
**Conflicts of Interest:** The authors confirm no competing interests.
#### **References**
## *Article* **Applications of Confluent Hypergeometric Function in Strong Superordination Theory**
**Georgia Irina Oros <sup>1</sup> , Gheorghe Oros <sup>1</sup> and Ancut,a Maria Rus 2,\***
**\*** Correspondence: [email protected]
**Abstract:** In the research presented in this paper, confluent hypergeometric function is embedded in the theory of strong differential superordinations. In order to proceed with the study, the form of the confluent hypergeometric function is adapted taking into consideration certain classes of analytic functions depending on an extra parameter previously introduced related to the theory of strong differential subordination and superordination. Operators previously defined using confluent hypergeometric function, namely Kummer–Bernardi and Kummer–Libera integral operators, are also adapted to those classes and strong differential superordinations are obtained for which they are the best subordinants. Similar results are obtained regarding the derivatives of the operators. The examples presented at the end of the study are proof of the applicability of the original results.
**Keywords:** analytic function; starlike function; convex function; strong differential superordination; best subordinant; confluent (Kummer) hypergeometric function
#### **1. Introduction**
The theory of strong differential subordination was initiated by Antonino and Romaguera [1] as a generalization of the classical concept of differential subordination introduced by Miller and Mocanu [2,3]. The results obtained by Antonino and Romaguera for the case of strong Briot–Bouquet differential subordinations inspired the development of the general theory related to strong differential subordination as seen for the classical case of differential subordination which is synthetized in [4]. The main aspects of strong differential subordination theory were established in a paper published in 2009 [5] by stating the three problems on which the theory is based on and by defining the notions of solution of a strong differential subordination and dominant of the solutions of the strong differential subordination. The class of admissible functions, a basic tool in the study of strong differential subordinations, was also introduced in this paper. The theory developed rapidly especially through studies associated to different operators like Liu–Srivastava operator [6], a generalized operator [7], multiplier transformation [8,9], Komatu integral operator [10], Sălăgean operator and Ruscheweyh derivative [11] or a certain differential operator [12]. The topic is still interesting for researchers as it is obvious from the numerous publications in the last two years when multiplier transformation and Ruscheweyh derivative [13] or integral operators [14] were used for obtaining new strong subordination results. We can refer to [15,16] for applications of differential operators in the analyses of phenomena from mathematical biology.
The dual notion of strong differential superordination was introduced also in 2009 [17] following the pattern set by Miller and Mocanu for the classical notion of differential superordination [18]. The special case of first order strong differential superordinations was next investigated [19]. Strong differential superodinations were applied to a general equation [20] and they were also related to different operators such as generalized Sălăgean
**Citation:** Oros, G.I.; Oros, G.; Rus, A.M. Applications of Confluent Hypergeometric Function in Strong Superordination Theory. *Axioms* **2022**, *11*, 209. https://doi.org/10.3390/ axioms11050209
Academic Editor: Sidney A. Morris
Received: 18 April 2022 Accepted: 21 April 2022 Published: 29 April 2022
**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.
**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).
and Ruscheweyh operators [21], new generalized derivative operator [22], or certain general operators [23]. This notion is still popular as it can be proved by listing a few more papers than already shown, published recently [24–26]. *Axioms* **2022**, *11*, x FOR PEER REVIEW 3 of 12
> In 2012 [27], some interesting new classes were introduced related to the theory of strong differential subordination and superordination. They are intensely used for obtaining new results ever since they were connected to the studies. **Definition 3.** *[27] Let be a set in* ℂ*,* (∙, ) ∈ , *and a positive integer. The class of admissible functions* ൣ, (∙, )൧ *consists of those functions* : ℂ<sup>ଷ</sup> ××ഥ → ℂ *that satisfy the*
The study presented in this paper uses those classes which we list as follows: *admissibility condition*
For *U* = {*z* ∈ C : |*z*| < 1} the unit disc of the complex plane, there are some notations used: *U* = {*z* ∈ *C* : |*z*| ≤ 1} and *∂U* = {*z* ∈ *C* : |*z*| = 1}. *H*(*U*) denotes the class of holomorphic functions in the unit disc. (, , ; , ) ∈ (A) *whenever* = (, ), = ௭ <sup>ᇲ</sup> (௭,) ୯మ '' (,) (,) + 1൨ , ∈ , ∈ ഥ\൫(∙, )൯
Let *H U* × *U* denote the class of analytic functions in *U* × *U*. , ቀ<sup>௧</sup> <sup>௦</sup> + 1ቁ ≤ <sup>ଵ</sup> ୯ '
The following subclasses of *H U* × *U* are defined in [27]: *and* ≥ ≥ 1. *When* =1 *we write* Φଵൣ, (∙, )൧ *as* Φൣ, (∙, )൧. *In the special case when* ℎ(∙, ) *is an analytic mapping of* ×ഥ *onto* ≠ ℂ *we denote the class*
$$H\_{\tilde{\varsigma}}[a, n] = \left\{ f \in H(\mathcal{U} \times \overline{\mathcal{U}}) : f(z, \zeta) = a + a\_n(\zeta)z^n + a\_{n+1}(\zeta)z^{n+1} + \dots, z \in \mathcal{U}, \zeta \in \overline{\mathcal{U}} \right\}$$
with *ak*(*ζ*) holomorphic functions in *U*, *k* ≥ *n*, *a* ∈ C, *n* ∈ N. ቆ(, ), ௭ <sup>ᇱ</sup> (, )
$$H\!\!\!\zeta\_{\mathcal{U}}(\mathcal{U}) = \left\{ f \in H\_{\zeta}[a,n] : f(\cdot,\zeta) \text{ univalent in } \mathcal{U} \text{ for all } \zeta \in \boxdot \mathcal{U} \right\}$$
$$A\mathcal{Z}\_{\mathbb{N}} = \left\{ f \in H\left(\mathcal{U} \times \overline{\mathcal{U}}\right) : f(z, \zeta) = z + a\_{n+1}(\zeta)z^{n+1} + \dots, z \in \mathcal{U}, \zeta \in \overline{\mathcal{U}} \right\}, \text{ with } A\zeta\_1 = A\zeta$$
$$A \subsetneq (\mathcal{T})\_{1 \times 1}, \dots, A \subsetneq \mathcal{U}, \zeta \in \mathcal{U}, \zeta \in \overline{\mathcal{U}}$$
and *ak*(*ζ*) holomorphic functions in *U*, *k* ≥ *n* + 1, *n* ∈ N. (, ) as follows:
$$\mathcal{S}^\*\mathcal{J} = \left\{ f \in A\mathcal{J} : \text{Re } \frac{zf\_z'(z,\zeta)}{f(z,\zeta)} > 0, z \in \mathcal{U}, \zeta \in \overline{\mathcal{U}} \right\}$$
denotes the class of starlike functions in *U* × *U*. *dinate to* (, ), *then there exist points* = ఏబ ∈ *and* ∈ \൫(, )൯ *and an* ≥ ≥1 *for which* ( × ഥబ) ⊂ ( × ഥ) *and*
$$K\mathcal{J} = \left\{ f \in A\mathcal{J} : \text{Re}\left(\frac{zf\_{z^2}''(z,\zeta)}{f\_z'(z,\zeta)} + 1\right) > 0, z \in \mathcal{U}, \zeta \in \overline{\mathcal{U}} \right\}$$
denotes the class of convex functions in *U* × *U*. () ቆ௭మ ௭ <sup>ᇱ</sup> (, ) + 1ቇ ≥ ቆ௭మ ௭ <sup>ᇱ</sup> (, ) + 1ቇ.
() ௭
*and*
*then*
For obtaining the original results of this paper, the following definitions and notations introduced in [27] are necessary: This lemma will be used in the next section for proving the theorems which contain
the original results. Another helpful result which will be used is the next lemma proved
**Definition 1 ([27]).** *Let h*(*z*, *ζ*)*and f*(*z*, *ζ*) *be analytic functions in U* × *U*. *The function f*(*z*, *ζ*) *is said to be strongly subordinate to h*(*z*, *ζ*)*, or h*(*z*, *ζ*) *is said to be strongly superordinate to f*(*z*, *ζ*) *if there exists a function w analytic in U with w*(0) = 0, |*w*(*z*)| < 1 *such that f*(*z*, *ζ*) = *h*(*w*(*z*), *ζ*), *for all ζ* ∈ *U*, *z* ∈ *U*. *In such a case, we write* in [28]. **Lemma 2.** *[28] Let* ℎ(, ) *be convex in for all* ∈ ഥ *with* ℎ(0, ) = , ≠ 0, > 0 *and* ∈ሾ, 1ሿ ∩ . *If* (, ) <sup>+</sup> ௭ <sup>ᇲ</sup> (௭,) ఊ *is univalent in for all* ∈ ഥ,
$$f(z, \zeta) \ll h(z, \zeta), \; z \in \mathcal{U}, \; \zeta \in \overline{\mathcal{U}}.$$
ᇱ
(, ) <sup>=</sup> ఊ න ℎ(, )ఊିଵ ௭ , **Remark 1 ([27]).** *(a) If f*(*z*, *ζ*) *is analytic in U* × *U and univalent in U for ζ* ∈ *U, then Definition 1 is equivalent to:*
$$f(0, \zeta) = h(0, \zeta), \\
\text{for all } \zeta \in \overline{\mathcal{U}} \text{ and } f(\mathcal{U} \times \overline{\mathcal{U}}) \subset h(\mathcal{U} \times \overline{\mathcal{U}}).$$
(, ) ⪻ (, ), ∈ , ∈ ഥ. *(b) If f*(*z*, *ζ*) = *f*(*z*), *h*(*z*, *ζ*) = *h*(*z*)*, then the strong superordination becomes the usual superordination.*
*The function is convex and is the best subordinant.* The connection between univalent function theory and hypergeometric functions **Definition 2 ([27]).** *We denote by Q<sup>ζ</sup> the set of functions q*(·, *ζ*) *that are analytic and injective, as function of z, on U*\*E*(*q*(*z*, *ζ*)) *where*
$$E(q(z, \zeta)) = \left\{ \xi \in \partial \mathcal{U} \, : \, \lim\_{z \to \zeta} q(z, \zeta) = \infty \right\}$$
function and gave conditions for its univalence was published in 1990 [30]. Ever since then, aspects of its univalence were further investigated [31,32], it was considered in connection with other important functions [33–37] and it was used in the definition of new
many investigations. One of the first papers which investigated confluent hypergeometric *and are such that q*0 *z* (*ξ*, *ζ*) 6= 0 *for ξ* ∈ *∂U*\*E*(*q*(*z*, *ζ*)), *ζ* ∈ *U*.
*The subclass of Q<sup>ζ</sup> for which q*(0, *ζ*) = *a is denoted by Q<sup>ζ</sup>* (*a*).
*Axioms* **2022**, *11*, x FOR PEER REVIEW 3 of 12
**Definition 3 ([27]).** *Let Ω<sup>ζ</sup> be a set in* C*, q*(·, *ζ*) ∈ *Ω<sup>ζ</sup>* , *and n a positive integer. The class of admissible functions Φ<sup>n</sup>* - *Ω<sup>ζ</sup>* , *q*(·, *ζ*) *consists of those functions <sup>ϕ</sup>* : <sup>C</sup><sup>3</sup> <sup>×</sup> *<sup>U</sup>* <sup>×</sup> *<sup>U</sup>* <sup>→</sup> <sup>C</sup> *that satisfy the admissibility condition* **Definition 3.** *[27] Let be a set in* ℂ*,* (∙, ) ∈ , *and a positive integer. The class of admissible functions* ൣ, (∙, )൧ *consists of those functions* : ℂ<sup>ଷ</sup> ××ഥ → ℂ *that satisfy the admissibility condition*
$$
\varphi(r, s, t; \mathfrak{f}, \mathfrak{f}) \in \Omega\_{\mathfrak{f}} \tag{A}
$$
*whenever r* = *q*(*z*, *ζ*), *s* = *zq*0 *z* (*z*, *ζ*) *m* , *Re t <sup>s</sup>* + 1 <sup>≤</sup> <sup>1</sup> *m Re zq* 00 *z* 2 (*z*,*ζ*) *q* 0 *<sup>z</sup>* (*z*,*ζ*) + 1 , *z* ∈ *U*, *ξ* ∈ *U*\*E*(*q*(·, *ζ*)) *whenever* = (, ), = ௭ <sup>ᇲ</sup> (௭,) , ቀ<sup>௧</sup> <sup>௦</sup> + 1ቁ ≤ <sup>ଵ</sup> ୯మ '' (,) ୯ ' (,) + 1൨ , ∈ , ∈ ഥ\൫(∙, )൯ *Axioms* **2022**, *11*, x FOR PEER REVIEW 3 of 12
*and m* ≥ *n* ≥ 1*. When n* = 1 *we write* Φ<sup>1</sup> -*Ω<sup>ζ</sup>* , *q*(·, *ζ*) *as* Φ -*Ω<sup>ζ</sup>* , *q*(·, *ζ*) . *In the special case when h*(·, *ζ*) *is an analytic mapping of U* × *U onto Ω<sup>ζ</sup>* 6= C *we denote the class* Φ*<sup>n</sup>* - *h* , *q*(*z*, *ζ*) *by* Φ*n*[*h*(*z*, *ζ*) , *q*(*z*, *ζ*)]. *and* ≥ ≥ 1. *When* =1 *we write* Φଵൣ, (∙, )൧ *as* Φൣ, (∙, )൧. *In the special case when* ℎ(∙, ) *is an analytic mapping of* ×ഥ *onto* ≠ ℂ *we denote the class* Φሾℎ( × ഥ) , (, )ሿ *by* Φሾℎ(, ) , (, )ሿ. **Definition 3.** *[27] Let be a set in* ℂ*,* (∙, ) ∈ , *and a positive integer. The class of ad-*
*U* × *U If <sup>ϕ</sup>* : <sup>C</sup><sup>2</sup> <sup>×</sup> *<sup>U</sup>* <sup>×</sup> *<sup>U</sup>* <sup>→</sup> <sup>C</sup>*, then the admissibility condition (A) reduces to If* : ℂ<sup>ଶ</sup> ××ഥ → ℂ*, then the admissibility condition (A) reduces to missible functions* ൣ, (∙, )൧ *consists of those functions* : ℂ<sup>ଷ</sup> ××ഥ → ℂ *that satisfy the*
$$\left(\varrho\left(q(z,\zeta),\frac{zq\_z'(z,\zeta)}{m};\xi,\zeta\right)\in\Omega\_{\tilde{\xi}'}\right)\tag{A'}$$
*where z* ∈ *U*, *ζ* ∈ *U*, *ξ* ∈ *U*\*E*(*q*(·, *ζ*)) *and m* ≥ *n* ≥ 1*. where* ∈ , ∈ ഥ, ∈ഥ\൫(∙, )൯ *and* ≥ ≥ 1. *whenever* = (, ), = ௭ <sup>ᇲ</sup> (௭,) , ቀ<sup>௧</sup> <sup>௦</sup> + 1ቁ ≤ <sup>ଵ</sup> ୯మ (,) ୯ '(,) + 1൨ , ∈ , ∈ ഥ\൫(∙, )൯
Miller—Mocanu lemma given in [18] was rewritten in [27] for functions *p*(*z*, *ζ*) and *q*(*z*, *ζ*) as follows: Miller—Mocanu lemma given in [18] was rewritten in [27] for functions (, ) and (, ) as follows: *and* ≥ ≥ 1. *When* =1 *we write* Φଵൣ, (∙, )൧ *as* Φൣ, (∙, )൧. *In the special case when* ℎ(∙, ) *is an analytic mapping of* ×ഥ *onto* ≠ ℂ *we denote the class* Φሾℎ( × ഥ) , (, )ሿ *by* Φሾℎ(, ) , (, )ሿ.
''
**Lemma 1 ([17,27]).** *Let p*(*z*, *ζ*) ∈ *Q*(*a*) *and let q*(*z*, *ζ*) = *a* + *an*(*ζ*)*z <sup>n</sup>* + *an*+1(*ζ*)*z <sup>n</sup>*+<sup>1</sup> + . . . *with ak*(*ζ*) *holomorphic functions in U, k* ≥ *n*, *q*(*z*, *ζ*) 6≡ *a and n* ≥ 1*. If q*(*z*, *ζ*) *is not subordinate to p*(*z*, *ζ*), *then there exist points z*<sup>0</sup> = *r*0*e <sup>i</sup>θ*<sup>0</sup> <sup>∈</sup> *<sup>U</sup> and <sup>ξ</sup>*<sup>0</sup> <sup>∈</sup> *<sup>∂</sup>U*\*E*(*p*(*z*, *<sup>ζ</sup>*)) *and an <sup>m</sup>* <sup>≥</sup> *<sup>n</sup>* <sup>≥</sup> <sup>1</sup> *for which q U* × *Ur*<sup>0</sup> ⊂ *p U* × *U and* (*i*) *q*(*z*0, *ζ*) = *p*(*ξ*0, *ζ*), (*ii*) *z*0*q* 0 (*z*0, *ζ*) = *mξ*<sup>0</sup> *p* 0 (*ξ*0, *ζ*) *and* **Lemma 1.** *([17],[27]) Let* (, ) ∈ () *and let* (, ) =+() + ାଵ()ାଵ + ⋯ *with* () *holomorphic functions in* ഥ*,* ≥ , (, ) ≢ *and* ≥ 1. *If* (, ) *is not subordinate to* (, ), *then there exist points* = ఏబ ∈ *and* ∈ \൫(, )൯ *and an* ≥ ≥1 *for which* ( × ഥబ) ⊂ ( × ഥ) *and* () (, ) = (, ), () ௭ <sup>ᇱ</sup> (, ) = ௭ <sup>ᇱ</sup> (, ) *and If* : ℂ<sup>ଶ</sup> ××ഥ → ℂ*, then the admissibility condition (A) reduces to* ቆ(, ), ௭ <sup>ᇱ</sup> (, ) ; , ቇ ∈ , (A') *where* ∈ , ∈ ഥ, ∈ഥ\൫(∙, )൯ *and* ≥ ≥ 1.
*z z* (*iii*) *Re z*0*q* 00 *z* 2 (*z*0, *ζ*) *q* 0 *<sup>z</sup>* (*z*0, *ζ*) + 1 <sup>≥</sup> *mRe ξ*<sup>0</sup> *p* 00 *z* 2 (*ξ*0, *ζ*) *p* 0 *<sup>z</sup>* (*ξ*0, *ζ*) + 1 . () ቆ௭మ ᇱᇱ (, ) ௭ <sup>ᇱ</sup> (, ) + 1ቇ ≥ ቆ௭మ ᇱᇱ (, ) ௭ <sup>ᇱ</sup> (, ) + 1ቇ. Miller—Mocanu lemma given in [18] was rewritten in [27] for functions (, ) and (, ) as follows:
ఊ
This lemma will be used in the next section for proving the theorems which contain the original results. Another helpful result which will be used is the next lemma proved in [28]. This lemma will be used in the next section for proving the theorems which contain the original results. Another helpful result which will be used is the next lemma proved in [28]. **Lemma 1.** *([17],[27]) Let* (, ) ∈ () *and let* (, ) =+() + ାଵ()ାଵ + ⋯ *with* () *holomorphic functions in* ഥ*,* ≥ , (, ) ≢ *and* ≥ 1. *If* (, ) *is not subordinate to* (, ), *then there exist points* = ఏబ ∈ *and* ∈ \൫(, )൯ *and an* ≥
**Lemma 2 ([28]).** *Let h*(*z*, *ζ*) *be convex in U for all ζ* ∈ *U with h*(0, *ζ*) = *a*, *γ* 6= 0, *Re γ* > 0 *and p* ∈ *H<sup>ζ</sup>* [*a*, 1] ∩ *Q*. *If p*(*z*, *ζ*) + *zp*0 *z* (*z*, *ζ*) *γ is univalent in U for all ζ* ∈ *U*, **Lemma 2.** *[28] Let* ℎ(, ) *be convex in for all* ∈ ഥ *with* ℎ(0, ) = , ≠ 0, > 0 *and* ∈ሾ, 1ሿ ∩ . *If* (, ) <sup>+</sup> ௭ <sup>ᇲ</sup> (௭,) *is univalent in for all* ∈ ഥ, ≥1 *for which* ( × ഥబ) ⊂ ( × ഥ) *and* () (, ) = (, ), () ௭ <sup>ᇱ</sup> (, ) = ௭ <sup>ᇱ</sup> (, ) *and* () ቆ௭మ ᇱᇱ (, ) ᇱᇱ (, )
$$h(z,\zeta)\ll p(z,\zeta)+\frac{zp\_z'(z,\zeta)}{\gamma}$$
௭
*and and* This lemma will be used in the next section for proving the theorems which contain
௭
*admissibility condition*
$$q(z, \zeta) = \frac{\gamma}{z^{\gamma}} \int\_0^z h(t, \zeta) t^{\gamma - 1} dt,$$
then
*is univalent in for all* ∈ ഥ,
ᇱ
*then* ∈ሾ, 1ሿ ∩ . *If* (, ) <sup>+</sup> ௭
*then*
in [28].
$$q(z,\zeta) \preccurlyeq p(z,\zeta), \; z \in \mathsf{U}, \mathsf{f} \in \; \overline{\mathsf{U}}.$$
was established in 1985 when de Branges used the generalized hypergeometric function
nection with other important functions [33–37] and it was used in the definition of new
for proving Bieberbach's conjecture [29]. Once hypergeometric functions were considered in studies regarding univalent functions, confluent hypergeometric function was used in many investigations. One of the first papers which investigated confluent hypergeometric function and gave conditions for its univalence was published in 1990 [30]. Ever since then, aspects of its univalence were further investigated [31,32], it was considered in connection with other important functions [33–37] and it was used in the definition of new
The connection between univalent function theory and hypergeometric functions
The connection between univalent function theory and hypergeometric functions *and* ௭ *The function q is convex and is the best subordinant.*
ఊ
for proving Bieberbach's conjecture [29]. Once hypergeometric functions were considered in studies regarding univalent functions, confluent hypergeometric function was used in many investigations. One of the first papers which investigated confluent hypergeometric function and gave conditions for its univalence was published in 1990 [30]. Ever since then, aspects of its univalence were further investigated [31,32], it was considered in con-(, ) <sup>=</sup> ఊ න ℎ(, )ఊିଵ , *then* (, ) ⪻ (, ), ∈ , ∈ ഥ. *The function is convex and is the best subordinant.* The connection between univalent function theory and hypergeometric functions was established in 1985 when de Branges used the generalized hypergeometric function for proving Bieberbach's conjecture [29]. Once hypergeometric functions were considered in studies regarding univalent functions, confluent hypergeometric function was used in many investigations. One of the first papers which investigated confluent hypergeometric function and gave conditions for its univalence was published in 1990 [30]. Ever since then, aspects of its univalence were further investigated [31,32], it was considered in connection with other important functions [33–37] and it was used in the definition of new operators [38]. This prolific function is used in the present paper for obtaining results related to another topic, strong differential superordinations. The function is considered as follows:
**Definition 4 ([30]).** *Let a and c be complex numbers with c* 6= 0, −1, −2, . . . *and consider*
$$\phi(a,c;z) = 1 + \frac{a}{c} \cdot \frac{z}{1!} + \frac{a(a+1)}{c(c+1)} \cdot \frac{z^2}{2!} + \dots, \; z \in \mathcal{U} \tag{1}$$
*This function is called confluent (Kummer) hypergeometric function, is analytic in* C, *and satisfies Kummer's differential equation:*
$$z \cdot w''(z) + [c - z] \cdot w'(z) - a \cdot w(z) = 0.$$
*If we let*
$$(d)\_k = \frac{\Gamma(d+k)}{\Gamma(d)} = d(d+1)(d+2)\dots(d+k-1) \text{ and } (d)\_0 = 1,$$
*then (1) can be written in the form*
$$\phi(a,c;z) = \sum\_{k=0}^{\infty} \frac{(a)\_k}{(c)\_k} \cdot \frac{z^k}{k!} = \frac{\Gamma(c)}{\Gamma(a)} \cdot \sum\_{k=0}^{\infty} \frac{\Gamma(a+k)}{\Gamma(c+k)} \cdot \frac{z^k}{k!} \tag{2}$$
In the study conducted for obtaining the original results presented in the next section of this paper, the operators introduced in [38] are adapted to the subclasses of *H U* × *U* defined in [27] as follows:
**Definition 5 ([38]).** *Let φ*(*a*, *c*; *z*) *be given by (1) and let γ* > 0*. The integral operator B* : *H<sup>ζ</sup>* [1, 1] → *H<sup>ζ</sup>* [1, 1],
$$B\left[\phi\left(a(\zeta),c(\zeta);z,\zeta\right)\right] = B\left(a(\zeta),c(\zeta);z,\zeta\right) = \frac{\gamma}{2^{\gamma}}\int\_{0}^{z}\phi\left(a(\zeta),c(\zeta);t,\zeta\right)t^{\gamma-1}dt\tag{3}$$
*z* ∈ *U*, *ζ* ∈ *U*, *is called Kummer–Bernardi integral operator. For γ* = 1 *the integral operator L* : *H<sup>ζ</sup>* [1, 1] → *H<sup>ζ</sup>* [1, 1] *is defined as*
$$L[\phi(a(\zeta), c(\zeta); z, \zeta)] = L(a(\zeta), c(\zeta); z, \zeta) = \frac{1}{z} \int\_0^z \phi(a(\zeta), c(\zeta); t, \zeta)dt,\tag{4}$$
*z* ∈ *U*, *ζ* ∈ *U* , *which is called Kummer–Libera integral operator.*
The form of the confluent hypergeometric function adapted to the new classes depending on the extra parameter *ζ* needed in the studies related to strong differential superordination theory is given in the next section. Strong differential superordinations are proved in the theorems for which the operators given by (3) and (4) and their derivatives with respect to *z* are the best subordinants considering *γ* in relation (3) both a real number, *γ* > 0, and a complex number with *Re γ* > 0. Examples are constructed as proof of the applicability of the new results.
#### **2. Main Results** *In the special case when* ℎ(∙, ) *is an analytic mapping of* ×ഥ *onto* ≠ ℂ *we denote the class*
*admissibility condition*
*whenever* = (, ), = ௭
*admissibility condition*
Considering confluent hypergeometric function defined by (1) or (2), if coefficients *a* and *c* complex numbers are replaced by holomorphic functions *a*(*ζ*), *c*(*ζ*) depending on the parameter *ζ* ∈ *U*, the function changes its form into the following: Φሾℎ( × ഥ) , (, )ሿ *by* Φሾℎ(, ) , (, )ሿ. *If* : ℂ<sup>ଶ</sup> ××ഥ → ℂ*, then the admissibility condition (A) reduces to* ቆ(, ), ௭ <sup>ᇱ</sup> (, ) *whenever* = (, ), = ௭ <sup>ᇲ</sup> (௭,) , ቀ<sup>௧</sup> <sup>௦</sup> + 1ቁ ≤ <sup>ଵ</sup> ୯మ ୯ ' (,) + 1൨ , ∈ , ∈ ഥ\൫(∙, )൯ *and* ≥ ≥ 1. *When* =1 *we write* Φଵൣ, (∙, )൧ *as* Φൣ, (∙, )൧. *In the special case when* ℎ(∙, ) *is an analytic mapping of* ×ഥ *onto* ≠ ℂ *we denote the class*
<sup>௦</sup> + 1ቁ ≤ <sup>ଵ</sup>
**Definition 3.** *[27] Let be a set in* ℂ*,* (∙, ) ∈ , *and a positive integer. The class of admissible functions* ൣ, (∙, )൧ *consists of those functions* : ℂ<sup>ଷ</sup> ××ഥ → ℂ *that satisfy the*
**Definition 3.** *[27] Let be a set in* ℂ*,* (∙, ) ∈ , *and a positive integer. The class of admissible functions* ൣ, (∙, )൧ *consists of those functions* : ℂ<sup>ଷ</sup> ××ഥ → ℂ *that satisfy the*
*Axioms* **2022**, *11*, x FOR PEER REVIEW 3 of 12
*Axioms* **2022**, *11*, x FOR PEER REVIEW 3 of 12
<sup>ᇲ</sup> (௭,) , ቀ<sup>௧</sup>
*and* ≥ ≥ 1. *When* =1 *we write* Φଵൣ, (∙, )൧ *as* Φൣ, (∙, )൧.
$$\phi(a(\zeta), c(\zeta); z, \zeta) = 1 + \frac{a(\zeta)}{c(\zeta)} \cdot \frac{z}{1!} + \frac{a(\zeta)[a(\zeta) + 1]}{c(\zeta)[c(\zeta) + 1]} \cdot \frac{z^2}{2!} + \dots, z \in \mathsf{U},\tag{5}$$
୯మ '' (,) ୯ '
> ''(,)
(, , ; , ) ∈ (A)
(, , ; , ) ∈ (A)
(,) + 1൨ , ∈ , ∈ ഥ\൫(∙, )൯
; , ቇ ∈ , (A')
୯ '
where (*ζ*) 6= 0, *c*(*ζ*) 6= 0, −1, −2, . . .. Miller—Mocanu lemma given in [18] was rewritten in [27] for functions (, ) and ቆ(, ), *Axioms* **2022**, *11*, x FOR PEER REVIEW 3 of 12
In [32], Corollary 4 the convexity in the unit disc of the function *φ*(*a*, *c*; *z*) given by (1) was proved. This property extends to the new form of the function (*a*(*ζ*), *c*(*ζ*); *z*, *ζ*), as seen in (5). (, ) as follows: **Lemma 1.** *([17],[27]) Let* (, ) ∈ () *and let* (, ) =+() + ାଵ()ାଵ + ⋯ *where* ∈ , ∈ ഥ, ∈ഥ\൫(∙, )൯ *and* ≥ ≥ 1. **Definition 3.** *[27] Let be a set in* ℂ*,* (∙, ) ∈ , *and a positive integer. The class of ad-*
The first original theorem presented in this paper uses the convexity of the function *φ*(*a*(*ζ*), *c*(*ζ*); *z*, *ζ*) and the methods related to strong differential superordination theory in order to find necessary conditions for Kummer–Bernardi integral operator presented in Definition 5 to be the best subordinant of a certain strong differential superordination involving confluent hypergeometric function *φ*(*a*(*ζ*), *c*(*ζ*); *z*, *ζ*). *with* () *holomorphic functions in* ഥ*,* ≥ , (, ) ≢ *and* ≥ 1. *If* (, ) *is not subordinate to* (, ), *then there exist points* = ఏబ ∈ *and* ∈ \൫(, )൯ *and an* ≥ ≥1 *for which* ( × ഥబ) ⊂ ( × ഥ) *and* () (, ) = (, ), () ௭ <sup>ᇱ</sup> (, ) = ௭ <sup>ᇱ</sup> (, ) *and* Miller—Mocanu lemma given in [18] was rewritten in [27] for functions (, ) and (, ) as follows: **Lemma 1.** *([17],[27]) Let* (, ) ∈ () *and let* (, ) =+() + ାଵ()ାଵ + ⋯ *with* () *holomorphic functions in* ഥ*,* ≥ , (, ) ≢ *and* ≥ 1. *If* (, ) *is not subordinate to* (, ), *then there exist points* = ఏబ ∈ *and* ∈ \൫(, )൯ *and an* ≥ *missible functions* ൣ, (∙, )൧ *consists of those functions* : ℂ<sup>ଷ</sup> ××ഥ → ℂ *that satisfy the admissibility condition* (, , ; , ) ∈ (A) *whenever* = (, ), = ௭ <sup>ᇲ</sup> (௭,) , ቀ<sup>௧</sup> <sup>௦</sup> + 1ቁ ≤ <sup>ଵ</sup> ୯మ '' (,) (,) + 1൨ , ∈ , ∈ ഥ\൫(∙, )൯
**Theorem 1.** *Consider the confluent hypergeometric function φ*(*a*(*ζ*), *c*(*ζ*); *z*, *ζ*) *defined by (5) and Kummer–Bernardi integral operator <sup>B</sup>*(*a*(*ζ*), *<sup>c</sup>*(*ζ*); *<sup>z</sup>*, *<sup>ζ</sup>*) *given by (3). Let <sup>ϕ</sup>* : <sup>C</sup><sup>2</sup> <sup>×</sup> *<sup>U</sup>* <sup>×</sup> *<sup>U</sup>* <sup>→</sup> <sup>C</sup> *be an admissible function with the properties seen in Definition 3. Suppose that φ*(*a*(*ζ*), *c*(*ζ*); *z*, *ζ*) *is a univalent solution of the equation* () ቆ௭మ ᇱᇱ (, ) ௭ <sup>ᇱ</sup> (, ) + 1ቇ ≥ ቆ௭మ ᇱᇱ (, ) ௭ <sup>ᇱ</sup> (, ) + 1ቇ. This lemma will be used in the next section for proving the theorems which contain the original results. Another helpful result which will be used is the next lemma proved ≥1 *for which* ( × ഥబ) ⊂ ( × ഥ) *and* () (, ) = (, ), () ௭ <sup>ᇱ</sup> (, ) = ௭ <sup>ᇱ</sup> (, ) *and* () ቆ௭మ ᇱᇱ (, ) <sup>ᇱ</sup> (, ) + 1ቇ ≥ ቆ௭మ ᇱᇱ (, ) <sup>ᇱ</sup> (, ) + 1ቇ. *and* ≥ ≥ 1. *When* =1 *we write* Φଵൣ, (∙, )൧ *as* Φൣ, (∙, )൧. *In the special case when* ℎ(∙, ) *is an analytic mapping of* ×ഥ *onto* ≠ ℂ *we denote the class* Φሾℎ( × ഥ) , (, )ሿ *by* Φሾℎ(, ) , (, )ሿ. *If* : ℂ<sup>ଶ</sup> ××ഥ → ℂ*, then the admissibility condition (A) reduces to*
௭
$$\phi(a(\zeta), c(\zeta); z, \zeta) = \underbrace{\phi\left(B(a(\zeta), c(\zeta); z, \zeta), z \cdot \mathbb{B}\_z'(a(\zeta), c(\zeta); z, \zeta); z, \zeta\right)}\_{:}.\tag{6}$$
*is univalent in for all* ∈ ഥ,
ఊ න ℎ(, )ఊିଵ
*If ϕ* ∈ *Φ<sup>n</sup>* - *h U* × *U* , *q*(*z*, *ζ*) , *p*(*z*, *ζ*) ∈ *Q<sup>ζ</sup>* (1) *and ϕ*(*p*(*z*, *ζ*), *z*·*p* 0 *z* (*z*, *ζ*); *z*, *ζ*) *are univalent in U for all ζ* ∈ *U, then strong superordination* **Lemma 2.** *[28] Let* ℎ(, ) *be convex in for all* ∈ ഥ *with* ℎ(0, ) = , ≠ 0, > 0 *and* ∈ሾ, 1ሿ ∩ . *If* (, ) <sup>+</sup> ௭ <sup>ᇲ</sup> (௭,) ఊ *is univalent in for all* ∈ ഥ, This lemma will be used in the next section for proving the theorems which contain the original results. Another helpful result which will be used is the next lemma proved in [28]. *where* ∈ , ∈ ഥ, ∈ഥ\൫(∙, )൯ *and* ≥ ≥ 1.
$$
\phi(a(\zeta), c(\zeta); z, \zeta) \rightsquigarrow \phi(p(z, \zeta), z \cdot p'\_z(z, \zeta); z, \zeta) \tag{7}
$$
ᇱ
,
*and implies* ∈ሾ, 1ሿ ∩ . *If* (, ) <sup>+</sup> ௭
*then*
in [28].
௭
$$B(a(\zeta), c(\zeta); z, \zeta) \preccurlyeq p(z, \zeta), \; z \in \mathsf{U}, \; \zeta \in \; \overline{\mathsf{U}}.$$
<sup>ᇱ</sup> (, ) *and*
௭
*then and* ௭ *The function q*(*z*, *ζ*) = *B*(*a*(*ζ*), *c*(*ζ*); *z*, *ζ*) *is the best subordinant. dinate to* (, ), *then there exist points* = ఏబ ∈ *and* ∈ \൫(, )൯ *and an* ≥ ≥1 *for which* ( × ഥబ) ⊂ ( × ഥ) *and*
<sup>ᇲ</sup> (௭,) ఊ
(, ) ⪻ (, ), ∈ , ∈ ഥ. (, ) <sup>=</sup> **Proof.** Using relation (3) we obtain () (, ) = (, ), () ௭ <sup>ᇱ</sup> (, ) = ௭
*then*
$$z^{\gamma} \cdot \mathcal{B}(a(\zeta), c(\zeta); z, \zeta) = \gamma \int\_0^z \phi(a(\zeta), c(\zeta); t, \zeta) t^{\gamma - 1} dt. \tag{8}$$
was established in 1985 when de Branges used the generalized hypergeometric function for proving Bieberbach's conjecture [29]. Once hypergeometric functions were considered in studies regarding univalent functions, confluent hypergeometric function was used in *The function is convex and is the best subordinant.* The connection between univalent function theory and hypergeometric functions Differentiating (8) with respect to *z*, following a simple calculation, the next equation is obtained: This lemma will be used in the next section for proving the theorems which contain the original results. Another helpful result which will be used is the next lemma proved
$$B(a(\zeta), c(\zeta); z, \zeta) + \frac{1}{\gamma} z \cdot \mathcal{B}'\_z(a(\zeta), c(\zeta); z, \zeta); z, \zeta) = \phi(a(\zeta), c(\zeta); z, \zeta). \tag{9}$$
$$\text{v.v.: } \quad \colon \quad \colon \cdot \colon \quad (0) \quad \qquad \qquad \qquad \qquad \qquad \qquad \qquad \qquad \colon \quad \div \quad \mathcal{F}' \tag{9}$$
nection with other important functions [33–37] and it was used in the definition of new in studies regarding univalent functions, confluent hypergeometric function was used in Using relation (9), strong superordination (7) becomes: ∈ሾ, 1ሿ ∩ . *If* (, ) <sup>+</sup> ௭ <sup>ᇲ</sup> (௭,) ఊ *is univalent in for all* ∈ ഥ,
$$\mathcal{B}(a(\zeta), c(\zeta); z, \zeta) + \frac{1}{\gamma} z \cdot \mathcal{B}'\_z(a(\zeta), c(\zeta); z, \zeta); z, \zeta) \preccurlyeq \varphi \left( p(z, \zeta), z \cdot p'\_z(z, \zeta); z, \zeta \right). \tag{10}$$
nection with other important functions [33–37] and it was used in the definition of new *and* (, ) <sup>=</sup> ఊ න ℎ(, )ఊିଵ ௭ , Let *<sup>ϕ</sup>* : <sup>C</sup><sup>2</sup> <sup>×</sup> *<sup>U</sup>* <sup>×</sup> *<sup>U</sup>* <sup>→</sup> <sup>C</sup> be an admissible function, *<sup>ϕ</sup>*(*r*,*s*; *<sup>z</sup>*, *<sup>ζ</sup>*) <sup>∈</sup> <sup>Φ</sup>*<sup>n</sup>* - *h U* × *U* , *q*(*z*, *ζ*) , defined by:
$$
\varphi(r, s; z, \zeta) = r + \frac{1}{\gamma} s, \text{ } r, s \in \mathbb{C}, \text{ } \gamma > 0. \tag{11}
$$
for proving Bieberbach's conjecture [29]. Once hypergeometric functions were considered in studies regarding univalent functions, confluent hypergeometric function was used in many investigations. One of the first papers which investigated confluent hypergeometric function and gave conditions for its univalence was published in 1990 [30]. Ever since then, aspects of its univalence were further investigated [31,32], it was considered in connection with other important functions [33–37] and it was used in the definition of new
*The function is convex and is the best subordinant.* Taking *r* = *B*(*a*(*ζ*), *c*(*ζ*); *z*, *ζ*), *s* = *z*·*B* 0 *z* (*a*(*ζ*), *c*(*ζ*); *z*, *ζ*); *z*, *ζ*) relation (11) becomes:
$$\begin{split} \rho(B(a(\zeta), c(\zeta); z, \zeta), z \cdot \mathbb{B}'\_z(a(\zeta), c(\zeta); z, \zeta); z, \zeta) \\ = \mathcal{B}(a(\zeta), c(\zeta); z, \zeta) + \frac{1}{\gamma} z \cdot \mathbb{B}'\_z(a(\zeta), c(\zeta); z, \zeta); z, \zeta). \end{split} \tag{12}$$
௭
Using relation (12) in (10) we get: ∈ሾ, 1ሿ ∩ . *If* (, ) <sup>+</sup> ௭
*admissibility condition*
(, ) as follows:
() (, ) = (, ),
௭
<sup>ᇱ</sup> (, ) = ௭
ᇱᇱ (, )
() ௭
in [28].
() ቆ௭మ
*whenever* = (, ), = ௭
$$\begin{array}{l} \text{rng} & \text{return (12) in (10) we get:}\\ \text{\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflim}}}}}}}}}}\\\$\reflectleft}}\text{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflim}}}}}}}}}\text{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflim}}}}}}}}}\\ \referleft}}\text{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflim}}}}}}}}}\text{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflectbox{\$\reflim}}}}}}}}\\ \referleft}}\text{.}\$\referleft} \text{.} \referleft} \end{.}} \end{array}} \end{array} .} \text{)\referleft} \text{.} \referleft} \text{.} \referleft} \text{.} \referleft} \text{.} \referleft} \text{.} \referleft} \text{.} \refer$$
*Axioms* **2022**, *11*, x FOR PEER REVIEW 3 of 12
<sup>ᇲ</sup> (௭,) , ቀ<sup>௧</sup>
Φሾℎ( × ഥ) , (, )ሿ *by* Φሾℎ(, ) , (, )ሿ.
≥1 *for which* ( × ഥబ) ⊂ ( × ഥ) *and*
<sup>ᇱ</sup> (, ) *and*
<sup>ᇱ</sup> (, ) + 1ቇ ≥ ቆ௭మ
*where* ∈ , ∈ ഥ, ∈ഥ\൫(∙, )൯ *and* ≥ ≥ 1.
*and* ≥ ≥ 1. *When* =1 *we write* Φଵൣ, (∙, )൧ *as* Φൣ, (∙, )൧.
ቆ(, ),
*If* : ℂ<sup>ଶ</sup> ××ഥ → ℂ*, then the admissibility condition (A) reduces to*
**Definition 3.** *[27] Let be a set in* ℂ*,* (∙, ) ∈ , *and a positive integer. The class of admissible functions* ൣ, (∙, )൧ *consists of those functions* : ℂ<sup>ଷ</sup> ××ഥ → ℂ *that satisfy the*
<sup>௦</sup> + 1ቁ ≤ <sup>ଵ</sup>
*In the special case when* ℎ(∙, ) *is an analytic mapping of* ×ഥ *onto* ≠ ℂ *we denote the class*
௭ <sup>ᇱ</sup> (, )
Miller—Mocanu lemma given in [18] was rewritten in [27] for functions (, ) and
**Lemma 1.** *([17],[27]) Let* (, ) ∈ () *and let* (, ) =+() + ାଵ()ାଵ + ⋯ *with* () *holomorphic functions in* ഥ*,* ≥ , (, ) ≢ *and* ≥ 1. *If* (, ) *is not subordinate to* (, ), *then there exist points* = ఏబ ∈ *and* ∈ \൫(, )൯ *and an* ≥
ᇱᇱ (, )
<sup>ᇱ</sup> (, ) + 1ቇ.
This lemma will be used in the next section for proving the theorems which contain the original results. Another helpful result which will be used is the next lemma proved
*is univalent in for all* ∈ ഥ,
ᇱ
୯మ '' (,) ୯ '
(, , ; , ) ∈ (A)
(,) + 1൨ , ∈ , ∈ ഥ\൫(∙, )൯
; , ቇ ∈ , (A')
*and* (, ) <sup>=</sup> ఊ න ℎ(, )ఊିଵ ௭ , Using Definition 1 and Remark 1, a), considering strong differential subordination (7) we get:
$$\phi(a(\zeta), c(\zeta); 0, \zeta) = \varphi(p(0, \zeta), 0; 0, \zeta)$$
and
$$
\mathfrak{q}\left(\mathcal{U}\times\overline{\mathcal{U}}\right)\subset\mathfrak{q}\left(\mathcal{U}\times\overline{\mathcal{U}}\right).\tag{13}
$$
in studies regarding univalent functions, confluent hypergeometric function was used in
(, , ; , ) ∈ (A)
*The function is convex and is the best subordinant.* Interpreting relation (13) we conclude that
*Axioms* **2022**, *11*, x FOR PEER REVIEW 3 of 12
$$
\varphi\left(p(\xi,\zeta),\xi\cdot p'\_z(\xi,\zeta);\xi,\zeta\right) \notin \phi\left(\mathcal{U}\times\overline{\mathcal{U}}\right), \xi\in \partial\mathcal{U}, \zeta\in\overline{\mathcal{U}}.\tag{14}
$$
for proving Bieberbach's conjecture [29]. Once hypergeometric functions were considered For *ξ* = *ξ*<sup>0</sup> ∈ *∂U*, relation (14) becomes: **Definition 3.** *[27] Let be a set in* ℂ*,* (∙, ) ∈ , *and a positive integer. The class of ad-*
$$
\varphi\left(p(\xi\_0, \zeta), \xi\_0 \cdot p'\_z(\xi\_0, \zeta); \xi\_0, \zeta\right) \notin \phi\left(\mathcal{U} \times \overline{\mathcal{U}}\right), \zeta \in \overline{\mathcal{U}}.\tag{15}
$$
୯ '
function and gave conditions for its univalence was published in 1990 [30]. Ever since then, aspects of its univalence were further investigated [31,32], it was considered in con-Using relation (6) we get:
*then*
$$\varrho\left(B(a(\zeta), c(\zeta); z, \zeta), z \cdot B\_z'(a(\zeta), c(\zeta); z, \zeta); z, \zeta\right) \in \phi\left(\mathcal{U} \times \overline{\mathcal{U}}\right), z \in \mathcal{U}, \zeta \in \overline{\mathcal{U}}.\tag{16}$$
$\text{Гал } z - z \subset \mathcal{U} \text{ (}\mathcal{U}\text{-) è } \omega\text{-simún } \omega.$
For *z* = *z*<sup>0</sup> ∈ *U*, (16) is written as: , ቀ<sup>௧</sup> *and* ≥ ≥ 1. *When* =1 *we write* Φଵൣ, (∙, )൧ *as* Φൣ, (∙, )൧.
$$\mathfrak{g}\left(\mathcal{B}(a(\zeta),c(\zeta);z\_0,\zeta),z\_0\cdot\mathcal{B}\_z'(a(\zeta),c(\zeta);z\_0,\zeta);z\_0,\zeta\right) \in \mathfrak{g}\left(\mathcal{U}\times\overline{\mathcal{U}}\right), z\_0 \in \mathcal{U}, \zeta \in \overline{\mathcal{U}}.\tag{17}$$
In order to finalize the proof, Lemma 1 and admissibility condition (*A* 0 ) will be applied. *If* : ℂ<sup>ଶ</sup> ××ഥ → ℂ*, then the admissibility condition (A) reduces to*
Suppose that *q*(*z*, *ζ*) = *B*(*a*(*ζ*), *c*(*ζ*); *z*, *ζ*) is not subordinate to *p*(*z*, *ζ*) for *z* ∈ *U*, *ζ* ∈ *U*. Then, using Lemma 1, we know that there are points *z*<sup>0</sup> = *r*0*e <sup>i</sup>θ*<sup>0</sup> <sup>∈</sup> *<sup>U</sup>* and *<sup>ξ</sup>*<sup>0</sup> <sup>∈</sup> *<sup>∂</sup>U*\*E*(*p*(*z*, *<sup>ζ</sup>*)) and an *m* ≥ *n* ≥ 1such that ቆ(, ), ௭ <sup>ᇱ</sup> (, ) ; , ቇ ∈ , (A')
$$(z\_0, \zeta) = \mathcal{B}(a(\zeta), c(\zeta); z\_0, \zeta) = p(\zeta\_0, \zeta) \text{ and}$$
$$z\_0 \cdot q\_z'(z\_0, \zeta) = z\_0 \cdot \mathcal{B}\_z'(a(\zeta), c(\zeta); z\_0, \zeta) = m\xi\_0 p\_z'(\zeta\_0, \zeta).$$
Using those conditions with *r* = *q*(*z*0, *ζ*) and *s* = *z*0·*q* 0 *z* (*z*0,*ζ*) *m* for *ξ* = *ξ*<sup>0</sup> in Definition 3 and taking into consideration the admissibility condition (*A* 0 ), we obtain: **Lemma 1.** *([17],[27]) Let* (, ) ∈ () *and let* (, ) =+() + ାଵ()ାଵ + ⋯ *with* () *holomorphic functions in* ഥ*,* ≥ , (, ) ≢ *and* ≥ 1. *If* (, ) *is not subor-*
$$\begin{array}{c} \varphi(p(\xi\_0,\xi),\xi\_0\mathbf{p}'\_x(\xi\_0,\xi);\xi\_0,\xi) = \varphi\Big(B(a(\zeta),c(\zeta);z\_0,\zeta),\frac{z\_0\cdot B'\_z(a(\zeta),c(\zeta);z\_0,\zeta)}{m};z\_0,\zeta\Big) \\ \in \phi(\mathcal{U}\times\overline{\mathcal{U}}).\end{array}$$
Using *m* = 1 in the previous relation, we get () ௭ <sup>ᇱ</sup> (, ) = ௭ <sup>ᇱ</sup> (, ) *and* ᇱᇱ (, ) ᇱᇱ (, )
*ϕ*(*p*(*ξ*0, *ζ*), *ξ*<sup>0</sup> *p* 0 *z* (*ξ*0, *ζ*); *ξ*0, *ζ*) = *ϕ*(*B*(*a*(*ζ*), *c*(*ζ*); *z*0, *ζ*), *z*0·*B* 0 *z* (*a*(*ζ*), *c*(*ζ*); *z*0, *ζ*); *z*0, *ζ*) ∈ *φ U* × *U* () ቆ௭మ ௭ <sup>ᇱ</sup> (, )+ 1ቇ ≥ ቆ௭మ ௭ <sup>ᇱ</sup> (, ) + 1ቇ.
and using (17) we write This lemma will be used in the next section for proving the theorems which contain the original results. Another helpful result which will be used is the next lemma proved
in [28].
*and*
*then*
(, ) as follows:
$$\phi\left(p(\mathfrak{f}\_{0\prime}\zeta), \mathfrak{f}\_{0}\mathfrak{p}'\_{\mathbf{z}}(\mathfrak{f}\_{0\prime}\zeta); \mathfrak{f}\_{0\prime}\zeta\right) \in \phi\left(\mathcal{U} \times \overline{\mathcal{U}}\right), z \in \mathcal{U}, \zeta \in \overline{\mathcal{U}},$$
which contradicts the result obtained in relation (15). Hence, the assumption made is false and we must have: **Lemma 2.** *[28] Let* ℎ(, ) *be convex in for all* ∈ ഥ *with* ℎ(0, ) = , ≠ 0, > 0 *and* ∈ሾ, 1ሿ ∩ . *If* (, ) <sup>+</sup> ௭ <sup>ᇲ</sup> (௭,) ఊ *is univalent in for all* ∈ ഥ,
$$B(a(\zeta), c(\zeta); z, \zeta) \preccurlyeq p(z, \zeta) \text{ for } z \in \mathcal{U}, \zeta \in \overline{\mathcal{U}}.$$
௭
(, ) ⪻ (, ), ∈ , ∈ ഥ.
The connection between univalent function theory and hypergeometric functions was established in 1985 when de Branges used the generalized hypergeometric function for proving Bieberbach's conjecture [29]. Once hypergeometric functions were considered in studies regarding univalent functions, confluent hypergeometric function was used in many investigations. One of the first papers which investigated confluent hypergeometric function and gave conditions for its univalence was published in 1990 [30]. Ever since then, aspects of its univalence were further investigated [31,32], it was considered in connection with other important functions [33–37] and it was used in the definition of new
(, ) <sup>=</sup>
ᇱ
,
ఊ න ℎ(, )ఊିଵ
*The function is convex and is the best subordinant.*
Since *q*(*z*, *ζ*) = *B*(*a*(*ζ*), *c*(*ζ*); *z*, *ζ*) satisfies the differential Equation (6), we conclude that *q*(*z*, *ζ*) = *B*(*a*(*ζ*), *c*(*ζ*); *z*, *ζ*) is the best subordinant. **Lemma 1.** *([17],[27]) Let* (, ) ∈ () *and let* (, ) =+() + ାଵ()ାଵ + ⋯ *with* () *holomorphic functions in* ഥ*,* ≥ , (, ) ≢ *and* ≥ 1. *If* (, ) *is not subordinate to* (, ), *then there exist points* = ఏబ ∈ *and* ∈ \൫(, )൯ *and an* ≥ Miller—Mocanu lemma given in [18] was rewritten in [27] for functions (, ) and (, ) as follows: (, , ; , ) ∈ (A) **Definition 3.** *[27] Let be a set in* ℂ*,* (∙, ) ∈ , *and a positive integer. The class of admissible functions* ൣ, (∙, )൧ *consists of those functions* : ℂ<sup>ଷ</sup> ××ഥ → ℂ *that satisfy the*
''(,)
Miller—Mocanu lemma given in [18] was rewritten in [27] for functions (, ) and
**Definition 3.** *[27] Let be a set in* ℂ*,* (∙, ) ∈ , *and a positive integer. The class of admissible functions* ൣ, (∙, )൧ *consists of those functions* : ℂ<sup>ଷ</sup> ××ഥ → ℂ *that satisfy the*
**Definition 3.** *[27] Let be a set in* ℂ*,* (∙, ) ∈ , *and a positive integer. The class of admissible functions* ൣ, (∙, )൧ *consists of those functions* : ℂ<sup>ଷ</sup> ××ഥ → ℂ *that satisfy the*
<sup>௦</sup> + 1ቁ ≤ <sup>ଵ</sup>
*In the special case when* ℎ(∙, ) *is an analytic mapping of* ×ഥ *onto* ≠ ℂ *we denote the class*
௭ <sup>ᇱ</sup> (, )
> ௭ <sup>ᇱ</sup> (, )
**Definition 3.** *[27] Let be a set in* ℂ*,* (∙, ) ∈ , *and a positive integer. The class of admissible functions* ൣ, (∙, )൧ *consists of those functions* : ℂ<sup>ଷ</sup> ××ഥ → ℂ *that satisfy the*
*In the special case when* ℎ(∙, ) *is an analytic mapping of* ×ഥ *onto* ≠ ℂ *we denote the class*
<sup>௦</sup> + 1ቁ ≤ <sup>ଵ</sup>
୯మ '' (,) ୯ '
୯మ '' (,) ୯ '
(, , ; , ) ∈ (A)
(, , ; , ) ∈ (A)
(,) + 1൨ , ∈ , ∈ ഥ\൫(∙, )൯
(,) + 1൨ , ∈ , ∈ ഥ\൫(∙, )൯
; , ቇ ∈ , (A')
; , ቇ ∈ , (A')
**Remark 2.** *For γ* = 1, *instead of Kummer–Bernardi integral operator, Kummer–Libera integral operator defined in (4) is used in Theorem 1 and the following corollary can be written:* ≥1 *for which* ( × ഥబ) ⊂ ( × ഥ) *and* () (, ) = (, ), () ௭ <sup>ᇱ</sup> (, ) = ௭ <sup>ᇱ</sup> (, ) *and* **Lemma 1.** *([17],[27]) Let* (, ) ∈ () *and let* (, ) =+() + ାଵ()ାଵ + ⋯ *with* () *holomorphic functions in* ഥ*,* ≥ , (, ) ≢ *and* ≥ 1. *If* (, ) *is not subordinate to* (, ), *then there exist points* = ఏబ ∈ *and* ∈ \൫(, )൯ *and an* ≥ *whenever* = (, ), = ௭ , ቀ<sup>௧</sup> <sup>௦</sup> + 1ቁ ≤ <sup>ଵ</sup> ୯మ ୯ ' (,) + 1൨ , ∈ , ∈ ഥ\൫(∙, )൯ *and* ≥ ≥ 1. *When* =1 *we write* Φଵൣ, (∙, )൧ *as* Φൣ, (∙, )൧. *admissibility condition* (, , ; , ) ∈ (A)
''
**Corollary 1.** *Consider the confluent hypergeometric function φ*(*a*(*ζ*), *c*(*ζ*); *z*, *ζ*) *defined by (5) and Kummer–Libera integral operator <sup>L</sup>*(*a*(*ζ*), *<sup>c</sup>*(*ζ*); *<sup>z</sup>*, *<sup>ζ</sup>*) *given by (4). Let <sup>ϕ</sup>* : <sup>C</sup><sup>2</sup> <sup>×</sup> *<sup>U</sup>* <sup>×</sup> *<sup>U</sup>* <sup>→</sup> <sup>C</sup> *be an admissible function with the properties seen in Definition 3. Suppose that φ*(*a*(*ζ*), *c*(*ζ*); *z*, *ζ*) *is a univalent solution of the equation* () ቆ௭మ ᇱᇱ (, ) ௭ <sup>ᇱ</sup> (, ) + 1ቇ ≥ ቆ௭మ ᇱᇱ (, ) ௭ <sup>ᇱ</sup> (, ) + 1ቇ. This lemma will be used in the next section for proving the theorems which contain the original results. Another helpful result which will be used is the next lemma proved ≥1 *for which* ( × ഥబ) ⊂ ( × ഥ) *and* () (, ) = (, ), () ௭ <sup>ᇱ</sup> (, ) = ௭ <sup>ᇱ</sup> (, ) *and* () ቆ௭మ ᇱᇱ (, ) <sup>ᇱ</sup> (, ) + 1ቇ ≥ ቆ௭మ ᇱᇱ (, ) <sup>ᇱ</sup> (, ) + 1ቇ. *In the special case when* ℎ(∙, ) *is an analytic mapping of* ×ഥ *onto* ≠ ℂ *we denote the class* Φሾℎ( × ഥ) , (, )ሿ *by* Φሾℎ(, ) , (, )ሿ. *If* : ℂ<sup>ଶ</sup> ××ഥ → ℂ*, then the admissibility condition (A) reduces to* ቆ(, ), ௭ <sup>ᇱ</sup> (, ) ; , ቇ ∈ , (A') *whenever* = (, ), = ௭ <sup>ᇲ</sup> (௭,) , ቀ<sup>௧</sup> <sup>௦</sup> + 1ቁ ≤ <sup>ଵ</sup> ୯మ (,) ୯ ' (,) + 1൨ , ∈ , ∈ ഥ\൫(∙, )൯ *and* ≥ ≥ 1. *When* =1 *we write* Φଵൣ, (∙, )൧ *as* Φൣ, (∙, )൧. *In the special case when* ℎ(∙, ) *is an analytic mapping of* ×ഥ *onto* ≠ ℂ *we denote the class* Φሾℎ( × ഥ) , (, )ሿ *by* Φሾℎ(, ) , (, )ሿ. *Axioms* **2022**, *11*, x FOR PEER REVIEW 3 of 12
$$\phi(a(\zeta), c(\zeta); z, \zeta) = \phi\left(L(a(\zeta), c(\zeta); z, \zeta), z \cdot \mathcal{L}\_z'(a(\zeta), c(\zeta); z, \zeta); z, \zeta\right).$$
*If ϕ* ∈ *Φ<sup>n</sup>* - *h U* × *U* , *q*(*z*, *ζ*) , *p*(*z*, *ζ*) ∈ *Q<sup>ζ</sup>* (1) *and ϕ*(*p*(*z*, *ζ*), *z*·*p* 0 *z* (*z*, *ζ*); *z*, *ζ*) *are univalent in U for allζ* ∈ *U*, *then strong superordination* **Lemma 2.** *[28] Let* ℎ(, ) *be convex in for all* ∈ ഥ *with* ℎ(0, ) = , ≠ 0, > 0 *and* ∈ሾ, 1ሿ ∩ . *If* (, ) <sup>+</sup> ௭ <sup>ᇲ</sup> (௭,) ఊ *is univalent in for all* ∈ ഥ, This lemma will be used in the next section for proving the theorems which contain the original results. Another helpful result which will be used is the next lemma proved in [28]. Miller—Mocanu lemma given in [18] was rewritten in [27] for functions (, ) and ቆ(, ), ௭ <sup>ᇱ</sup> (, ) ; , ቇ ∈ , (A') *admissibility condition* (, , ; , ) ∈ (A)
ᇱ
ᇱ
*is univalent in for all* ∈ ഥ,
$$
\phi(a(\zeta), c(\zeta); z, \zeta) \nleftrightarrows \rho \left(p(z, \zeta), z \cdot p'\_z(z, \zeta); z, \zeta\right)
$$
*and implies* ∈ሾ, 1ሿ ∩ . *If* (, ) <sup>+</sup> ௭ *with* () *holomorphic functions in* ഥ*,* ≥ , (, ) ≢ *and* ≥ 1. *If* (, ) *is not subor-*Miller—Mocanu lemma given in [18] was rewritten in [27] for functions (, ) and *and* ≥ ≥ 1. *When* =1 *we write* Φଵൣ, (∙, )൧ *as* Φൣ, (∙, )൧.
(, ) as follows:
*whenever* = (, ), = ௭
(, ) as follows:
<sup>ᇲ</sup> (௭,) , ቀ<sup>௧</sup>
in [28].
$$L(a(\zeta), c(\zeta); z, \zeta) \preccurlyeq p(z, \zeta), z \in \mathsf{U}, \,\zeta \in \overline{\mathsf{U}}.$$
௭
*then and* ௭ *The function q*(*z*, *ζ*) = *L*(*a*(*ζ*), *c*(*ζ*); *z*, *ζ*) *is the best subordinant.* () (, ) = (, ), () ௭ <sup>ᇱ</sup> (, ) = ௭ <sup>ᇱ</sup> (, ) *and* **Lemma 1.** *([17],[27]) Let* (, ) ∈ () *and let* (, ) =+() + ାଵ()ାଵ + ⋯ *with* () *holomorphic functions in* ഥ*,* ≥ , (, ) ≢ *and* ≥ 1. *If* (, ) *is not subor-If* : ℂ<sup>ଶ</sup> ××ഥ → ℂ*, then the admissibility condition (A) reduces to*
> <sup>ᇲ</sup> (௭,) ఊ
௭
ఊ
*dinate to* (, ), *then there exist points* = ఏబ ∈ *and* ∈ \൫(, )൯ *and an* ≥
<sup>ᇲ</sup> (௭,) ఊ
୯ '
*Axioms* **2022**, *11*, x FOR PEER REVIEW 3 of 12
*Axioms* **2022**, *11*, x FOR PEER REVIEW 3 of 12
<sup>ᇲ</sup> (௭,) , ቀ<sup>௧</sup>
Φሾℎ( × ഥ) , (, )ሿ *by* Φሾℎ(, ) , (, )ሿ.
Φሾℎ( × ഥ) , (, )ሿ *by* Φሾℎ(, ) , (, )ሿ.
*where* ∈ , ∈ ഥ, ∈ഥ\൫(∙, )൯ *and* ≥ ≥ 1.
*Axioms* **2022**, *11*, x FOR PEER REVIEW 3 of 12
*where* ∈ , ∈ ഥ, ∈ഥ\൫(∙, )൯ *and* ≥ ≥ 1.
<sup>ᇲ</sup> (௭,)
*and* ≥ ≥ 1. *When* =1 *we write* Φଵൣ, (∙, )൧ *as* Φൣ, (∙, )൧.
ቆ(, ),
ቆ(, ),
*and* ≥ ≥ 1. *When* =1 *we write* Φଵൣ, (∙, )൧ *as* Φൣ, (∙, )൧.
*If* : ℂ<sup>ଶ</sup> ××ഥ → ℂ*, then the admissibility condition (A) reduces to*
*If* : ℂ<sup>ଶ</sup> ××ഥ → ℂ*, then the admissibility condition (A) reduces to*
<sup>ᇲ</sup> (௭,) , ቀ<sup>௧</sup>
*admissibility condition*
(, ) as follows:
*admissibility condition*
*whenever* = (, ), = ௭
*whenever* = (, ), = ௭
*admissibility condition*
(, ) ⪻ (, ), ∈ , ∈ ഥ. *The function is convex and is the best subordinant.* The connection between univalent function theory and hypergeometric functions (, ) <sup>=</sup> ఊ න ℎ(, )ఊିଵ , *then* (, ) ⪻ (, ), ∈ , ∈ ഥ. **Theorem 2.** *Let q*(*z*, *ζ*) *be a convex function in the unit disc for all ζ* ∈ *U*, *consider the confluent hypergeometric function φ*(*a*(*ζ*), *c*(*ζ*); *z*, *ζ*) *defined by (5) and Kummer–Bernardi integral operator <sup>B</sup>*(*a*(*ζ*), *<sup>c</sup>*(*ζ*); *<sup>z</sup>*, *<sup>ζ</sup>*) *given by (3). Let <sup>ϕ</sup>* : <sup>C</sup><sup>2</sup> <sup>×</sup> *<sup>U</sup>* <sup>×</sup> *<sup>U</sup>* <sup>→</sup> <sup>C</sup> *be an admissible function with the properties seen in Definition 3 and define the analytic function* () ቆ௭మ ᇱᇱ (, ) ௭ <sup>ᇱ</sup> (, ) + 1ቇ ≥ ቆ௭మ ᇱᇱ (, ) ௭ <sup>ᇱ</sup> (, ) + 1ቇ. This lemma will be used in the next section for proving the theorems which contain *dinate to* (, ), *then there exist points* = ఏబ ∈ *and* ∈ \൫(, )൯ *and an* ≥ ≥1 *for which* ( × ഥబ) ⊂ ( × ഥ) *and* () (, ) = (, ), <sup>ᇱ</sup> (, ) = ௭ <sup>ᇱ</sup> (, ) *and* () ቆ௭మ ᇱᇱ (, ) ᇱᇱ (, ) ቆ(, ), ௭ <sup>ᇱ</sup> (, ) ; , ቇ ∈ , (A') *where* ∈ , ∈ ഥ, ∈ഥ\൫(∙, )൯ *and* ≥ ≥ 1.
$$\text{proplex uses seen in Definition } \mathcal{I} \text{ and } \text{uejue} \text{ the } \text{univariate} \text{ } \text{[main]}$$
$$h(z, \zeta) = \left(1 + \frac{1}{\gamma}\right) q(z, \zeta) + \frac{1}{\gamma} z \cdot q\_z'(z, \zeta), z \in \mathcal{U}, \zeta \in \overline{\mathcal{U}}.$$
many investigations. One of the first papers which investigated confluent hypergeometric function and gave conditions for its univalence was published in 1990 [30]. Ever since The connection between univalent function theory and hypergeometric functions was established in 1985 when de Branges used the generalized hypergeometric function for proving Bieberbach's conjecture [29]. Once hypergeometric functions were considered *If φ* 0 *z* (*a*(*ζ*), *c*(*ζ*); *z*, *ζ*) *and B* 0 *z* (*a*(*ζ*), *c*(*ζ*); *z*, *ζ*) ∈ *H<sup>ζ</sup>* [1, 1] ∩ *Q<sup>ζ</sup>* (1) *are univalent functions in U for all ζ* ∈ *U, then strong differential superordination* **Lemma 2.** *[28] Let* ℎ(, ) *be convex in for all* ∈ ഥ *with* ℎ(0, ) = , ≠ 0, > 0 *and* ∈ሾ, 1ሿ ∩ . *If* (, ) <sup>+</sup> ௭ <sup>ᇲ</sup> (௭,) ఊ *is univalent in for all* ∈ ഥ, This lemma will be used in the next section for proving the theorems which contain the original results. Another helpful result which will be used is the next lemma proved **Lemma 1.** *([17],[27]) Let* (, ) ∈ () *and let* (, ) =+() + ାଵ()ାଵ + ⋯ *with* () *holomorphic functions in* ഥ*,* ≥ , (, ) ≢ *and* ≥ 1. *If* (, ) *is not subor-*
$$h(z, \zeta) \ll \phi\_z'(a(\zeta), c(\zeta); z, \zeta) \tag{18}$$
ᇱ
function and gave conditions for its univalence was published in 1990 [30]. Ever since *and implies* ∈ሾ, 1ሿ ∩ . *If* (, ) <sup>+</sup> ௭ ≥1 *for which* ( × ഥబ) ⊂ ( × ഥ) *and* () (, ) = (, ), <sup>ᇱ</sup> (, ) = ௭ <sup>ᇱ</sup> (, ) *and*
<sup>ᇱ</sup> (, ) + 1ቇ ≥ ቆ௭మ
in [28].
௭
() ௭
(, ) as follows:
in [28].
ᇱᇱ (, )
*and*
௭
() ௭
*and*
*then*
() ቆ௭మ
$$q(z,\zeta) \prec \mathcal{B}'\_z(a(\zeta), c(\zeta); z, \zeta), \; z \in \mathcal{U}, \; \zeta \in \overline{\mathcal{U}}.$$
*is univalent in for all* ∈ ഥ,
*then* (, ) ⪻ (, ), ∈ , ∈ ഥ. (, ) <sup>=</sup> ఊ න ℎ(, )ఊିଵ ௭ , **Proof.** Using relation (9) from the proof of Theorem 1 and differentiating it with respect to *z*, we obtain: This lemma will be used in the next section for proving the theorems which contain the original results. Another helpful result which will be used is the next lemma proved
$$\phi\_z'(a(\zeta), c(\zeta); z, \zeta) = \left(1 + \frac{1}{\gamma}\right) \mathcal{B}\_z'(a(\zeta), c(\zeta); z, \zeta) + \frac{1}{\gamma} z \cdot \mathcal{B}\_{z2}''(a(\zeta), c(\zeta); z, \zeta), z \in \mathcal{U}, \zeta \in \overline{\mathcal{U}}.\tag{19}$$
was established in 1985 when de Branges used the generalized hypergeometric function *The function is convex and is the best subordinant.* Using (19), strong differential superordination (18) becomes: ∈ሾ, 1ሿ ∩ . *If* (, ) <sup>+</sup> ௭ <sup>ᇲ</sup> (௭,) *is univalent in for all* ∈ ഥ,
$$h(z, \zeta) \rightsquigarrow \left(1 + \frac{1}{\gamma}\right) \mathcal{B}'\_z(a(\zeta), c(\zeta); z, \zeta) + \frac{1}{\gamma} z \cdot \mathcal{B}''\_{z^2}(a(\zeta), c(\zeta); z, \zeta) . \tag{20}$$
function and gave conditions for its univalence was published in 1990 [30]. Ever since then, aspects of its univalence were further investigated [31,32], it was considered in conwas established in 1985 when de Branges used the generalized hypergeometric function for proving Bieberbach's conjecture [29]. Once hypergeometric functions were considered (, ) <sup>=</sup> ௭ , For the proof of this theorem to be complete, Lemma 1 and the admissibility condition (A0 ) will be applied.
nection with other important functions [33–37] and it was used in the definition of new in studies regarding univalent functions, confluent hypergeometric function was used in many investigations. One of the first papers which investigated confluent hypergeometric function and gave conditions for its univalence was published in 1990 [30]. Ever since ఊ න ℎ(, )ఊିଵ In order to do that, we define the admissible function *<sup>ϕ</sup>* : <sup>C</sup><sup>2</sup> <sup>×</sup> *<sup>U</sup>* <sup>×</sup> *<sup>U</sup>* <sup>→</sup> <sup>C</sup>, *ϕ*(*r*,*s*; *z*, *ζ*) ∈ Φ*<sup>n</sup>* - *h U* × *U* , *q*(*z*, *ζ*) , given by:
$$
\varphi(r, s; z, \zeta) = \left(1 + \frac{1}{\gamma}\right)r + \frac{1}{\gamma}s, \text{ } r, s \in \mathbb{C}, \ \gamma > 0. \tag{21}
$$
The connection between univalent function theory and hypergeometric functions
for proving Bieberbach's conjecture [29]. Once hypergeometric functions were considered in studies regarding univalent functions, confluent hypergeometric function was used in many investigations. One of the first papers which investigated confluent hypergeometric function and gave conditions for its univalence was published in 1990 [30]. Ever since then, aspects of its univalence were further investigated [31,32], it was considered in connection with other important functions [33–37] and it was used in the definition of new
Taking *r* = *B* 0 *z* (*a*(*ζ*), *c*(*ζ*); *z*, *ζ*), *s* = *z*·*B* 00 z 2 (*a*(*ζ*), *c*(*ζ*); *z*, *ζ*) relation (21) becomes: This lemma will be used in the next section for proving the theorems which contain
$$\begin{split} \rho \left( B\_z'(a(\boldsymbol{\zeta}), c(\boldsymbol{\zeta}); z, \boldsymbol{\zeta}), z \cdot B\_{\boldsymbol{z}^2}''(a(\boldsymbol{\zeta}), c(\boldsymbol{\zeta}); z, \boldsymbol{\zeta}); z, \boldsymbol{\zeta} \right) \\ = \left( 1 + \frac{1}{\gamma} \right) B\_z'(a(\boldsymbol{\zeta}), c(\boldsymbol{\zeta}); z, \boldsymbol{\zeta}) + \frac{1}{\gamma} z \cdot B\_{z^2}''(a(\boldsymbol{\zeta}), c(\boldsymbol{\zeta}); z, \boldsymbol{\zeta}); z, \boldsymbol{\zeta} \right). \end{split} \tag{22}$$
(, , ; , ) ∈ (A)
(,) + 1൨ , ∈ , ∈ ഥ\൫(∙, )൯
; , ቇ ∈ , (A')
Using relation (22) in (20) we get: ∈ሾ, 1ሿ ∩ . *If* (, ) <sup>+</sup> ௭ <sup>ᇲ</sup> (௭,) ఊ *is univalent in for all* ∈ ഥ, *Axioms* **2022**, *11*, x FOR PEER REVIEW 3 of 12
<sup>ᇲ</sup> (௭,)
௭
*Axioms* **2022**, *11*, x FOR PEER REVIEW 3 of 12
<sup>ᇲ</sup> (௭,) , ቀ<sup>௧</sup>
Φሾℎ( × ഥ) , (, )ሿ *by* Φሾℎ(, ) , (, )ሿ.
≥1 *for which* ( × ഥబ) ⊂ ( × ഥ) *and*
<sup>ᇱ</sup> (, ) *and*
<sup>ᇱ</sup> (, ) + 1ቇ ≥ ቆ௭మ
*where* ∈ , ∈ ഥ, ∈ഥ\൫(∙, )൯ *and* ≥ ≥ 1.
*and* ≥ ≥ 1. *When* =1 *we write* Φଵൣ, (∙, )൧ *as* Φൣ, (∙, )൧.
ቆ(, ),
*If* : ℂ<sup>ଶ</sup> ××ഥ → ℂ*, then the admissibility condition (A) reduces to*
*admissibility condition*
(, ) as follows:
() (, ) = (, ),
௭
in [28].
*and*
*then*
<sup>ᇱ</sup> (, ) = ௭
ᇱᇱ (, )
*whenever* = (, ), = ௭
**Definition 3.** *[27] Let be a set in* ℂ*,* (∙, ) ∈ , *and a positive integer. The class of admissible functions* ൣ, (∙, )൧ *consists of those functions* : ℂ<sup>ଷ</sup> ××ഥ → ℂ *that satisfy the*
<sup>௦</sup> + 1ቁ ≤ <sup>ଵ</sup>
*In the special case when* ℎ(∙, ) *is an analytic mapping of* ×ഥ *onto* ≠ ℂ *we denote the class*
௭ <sup>ᇱ</sup> (, )
Miller—Mocanu lemma given in [18] was rewritten in [27] for functions (, ) and
**Lemma 1.** *([17],[27]) Let* (, ) ∈ () *and let* (, ) =+() + ାଵ()ାଵ + ⋯ *with* () *holomorphic functions in* ഥ*,* ≥ , (, ) ≢ *and* ≥ 1. *If* (, ) *is not subordinate to* (, ), *then there exist points* = ఏబ ∈ *and* ∈ \൫(, )൯ *and an* ≥
ᇱᇱ (, )
<sup>ᇱ</sup> (, ) + 1ቇ.
୯మ '' (,) ୯ '
$$\ln(z,\zeta) \prec q\left(\mathcal{B}'\_z(a(\zeta),c(\zeta);z,\zeta),z\mathcal{B}''\_{\mathbf{z}^2}(a(\zeta),c(\zeta);z,\zeta);z,\zeta\right).$$
(, ) <sup>=</sup> ௭ Using Definition 1 and Remark 1, a) for this strong differential superordination, we get: **Definition 3.** *[27] Let be a set in* ℂ*,* (∙, ) ∈ , *and a positive integer. The class of admissible functions* ൣ, (∙, )൧ *consists of those functions* : ℂ<sup>ଷ</sup> ××ഥ → ℂ *that satisfy the*
$$h(0,\zeta) = \wp\left(B\_z'(a(\zeta), c(\zeta); 0, \zeta), 0; 0, \zeta\right)$$
and
*admissibility condition*
(, ) as follows:
() ௭
in [28].
*and*
*then*
() ቆ௭మ
௭
$$h\left(\mathcal{U}\times\overline{\mathcal{U}}\right)\subset\mathcal{q}\left(\mathcal{U}\times\overline{\mathcal{U}}\right).\tag{23}$$
*The function is convex and is the best subordinant.* Interpreting relation (23) we conclude that *whenever* = (, ), = ௭ , ቀ<sup>௧</sup> <sup>௦</sup> + 1ቁ ≤ <sup>ଵ</sup> ୯మ ୯ '(,) + 1൨ , ∈ , ∈ ഥ\൫(∙, )൯
$$\varphi\left(B\_{z}'(a(\zeta),c(\zeta);\xi,\zeta),\xi;B\_{z^{2}}'(a(\zeta),c(\zeta);\xi,\zeta);\xi,\zeta\right) \notin h\left(\mathcal{U}\times\overline{\mathcal{U}}\right),\ \xi\in\partial\mathcal{U},\zeta\in\overline{\mathcal{U}}.\tag{24}$$
was established in 1985 when de Branges used the generalized hypergeometric function for proving Bieberbach's conjecture [29]. Once hypergeometric functions were considered For *ξ* = *ξ*<sup>0</sup> ∈ *∂U*, relation (24) becomes: *If* : ℂ<sup>ଶ</sup> ××ഥ → ℂ*, then the admissibility condition (A) reduces to*
$$\varphi\left(\mathcal{B}'\_z(a(\boldsymbol{\zeta}), c(\boldsymbol{\zeta}); \boldsymbol{\xi}\_0, \boldsymbol{\zeta}), \boldsymbol{\xi}\_0; \mathcal{B}'\_{z^2}(a(\boldsymbol{\zeta}), c(\boldsymbol{\zeta}); \boldsymbol{\xi}\_0, \boldsymbol{\zeta}); \boldsymbol{\xi}\_0, \boldsymbol{\zeta}\right) \notin h\left(\mathcal{U} \times \overline{\mathcal{U}}\right), \boldsymbol{\zeta} \in \overline{\mathcal{U}}.\tag{25}$$
function and gave conditions for its univalence was published in 1990 [30]. Ever since then, aspects of its univalence were further investigated [31,32], it was considered in connection with other important functions [33–37] and it was used in the definition of new Suppose that *q*(*z*, *ζ*)is not subordinate to *B* 0 *z* (*a*(*ζ*), *c*(*ζ*); *z*, *ζ*)for *z* ∈ *U*, *ζ* ∈ *U*. Then, using Lemma 1, we know that there are points *z*<sup>0</sup> = *r*0*e <sup>i</sup>θ*<sup>0</sup> <sup>∈</sup> *<sup>U</sup>* and *<sup>ξ</sup>*<sup>0</sup> <sup>∈</sup> *<sup>∂</sup>U*\*E*(*<sup>B</sup>* 0 *z* (*a*(*ζ*), *c*(*ζ*); *z*, *ζ*)) and an *m* ≥ *n* ≥ 1 such that *where* ∈ , ∈ ഥ, ∈ഥ\൫(∙, )൯ *and* ≥ ≥ 1. Miller—Mocanu lemma given in [18] was rewritten in [27] for functions (, ) and
$$q(z\_0, \zeta) = B\_z'(a(\zeta), c(\zeta); z\_0, \zeta) = p(\xi\_0, \zeta) \text{ and}$$
$$z\_0 q\_z'(z\_0, \zeta) = m\xi\_0' B\_{z^2}'(a(\zeta), c(\zeta); z\_0, \zeta) = m\xi\_0' p\_z'(\xi\_0, \zeta).$$
Using those conditions with *r* = *B* 0 *z* (*a*(*ζ*), *c*(*ζ*); *z*0, *ζ*) and *s* = *ξ*0*B* 00 *z* 2 (*a*(*ζ*), *c*(*ζ*); *z*0, *ζ*) for *ξ* = *ξ*<sup>0</sup> in Definition 3 and taking into consideration the admissibility condition (*A* 0 ), we obtain: *dinate to* (, ), *then there exist points* = ఏబ ∈ *and* ∈ \൫(, )൯ *and an* ≥ ≥1 *for which* ( × ഥబ) ⊂ ( × ഥ) *and* () (, ) = (, ), <sup>ᇱ</sup> (, ) = ௭ <sup>ᇱ</sup> (, ) *and*
$$\begin{split} \varphi(q(z\_0,\zeta),z\_0q'\_z(z\_0,\zeta);z\_0,\zeta) &= \varphi\left(B'\_z(a(\zeta),c(\zeta);z\_0,\zeta), \frac{\xi\_0\theta'\_{\frac{\nu}{2}}(a(\zeta),c(\zeta);\xi\_0\zeta)}{m};z\_0,\zeta\right) \\ &\in h(\mathcal{U}\times\overline{\mathcal{U}}). \end{split}$$
Using *m* = 1 in the previous relation, we get This lemma will be used in the next section for proving the theorems which contain the original results. Another helpful result which will be used is the next lemma proved
$$\left(\rho\left(B\_z'(a(\zeta), c(\zeta); z\_0, \zeta), \xi\_0 B\_{z^2}''(a(\zeta), c(\zeta); z\_0, \zeta); z\_0, \zeta\right) \in h(\mathcal{U} \times \overline{\mathcal{U}}), \zeta \in \overline{\mathcal{U}},\right)$$
which contradicts the result obtained in relation (25). Hence, the assumption made is false and we must have: **Lemma 2.** *[28] Let* ℎ(, ) *be convex in for all* ∈ ഥ *with* ℎ(0, ) = , ≠ 0, > 0 *and* ∈ሾ, 1ሿ ∩ . *If* (, ) <sup>+</sup> ௭ <sup>ᇲ</sup> (௭,) ఊ *is univalent in for all* ∈ ഥ,
$$q(z,\zeta) \rightsquigarrow B\_z'(a(\zeta), c(\zeta); z, \zeta) \text{ for } z \in \mathsf{U}, \zeta \in \overline{\mathsf{U}}.$$
(, ) <sup>=</sup> ఊ න ℎ(, )ఊିଵ , **Remark 3.** *For γ* = 1, *instead of Kummer–Bernardi integral operator, Kummer–Libera integral operator defined in (4) is used in Theorem 2 and the following corollary can be written:*
௭
The connection between univalent function theory and hypergeometric functions was established in 1985 when de Branges used the generalized hypergeometric function
many investigations. One of the first papers which investigated confluent hypergeometric function and gave conditions for its univalence was published in 1990 [30]. Ever since then, aspects of its univalence were further investigated [31,32], it was considered in connection with other important functions [33–37] and it was used in the definition of new
(, ) ⪻ (, ), ∈ , ∈ ഥ. *The function is convex and is the best subordinant.* **Corollary 2.** *Let q*(*z*, *ζ*) *be a convex function in the unit disc for all ζ* ∈ *U*, *consider the confluent hypergeometric function φ*(*a*(*ζ*), *c*(*ζ*); *z*, *ζ*) *defined by (5) and Kummer–Libera integral operator*
*<sup>L</sup>*(*a*(*ζ*), *<sup>c</sup>*(*ζ*); *<sup>z</sup>*, *<sup>ζ</sup>*) *given by (4). Let <sup>ϕ</sup>* : <sup>C</sup><sup>2</sup> <sup>×</sup> *<sup>U</sup>* <sup>×</sup> *<sup>U</sup>* <sup>→</sup> <sup>C</sup> *be an admissible function with the properties seen in Definition 3 and define the analytic function:* ௭ <sup>ᇱ</sup> (, ) + 1ቇ ≥ ቆ௭మ ௭ <sup>ᇱ</sup> (, ) + 1ቇ. This lemma will be used in the next section for proving the theorems which contain () (, ) = (, ), <sup>ᇱ</sup> (, ) = ௭ <sup>ᇱ</sup> (, ) *and* () ቆ௭మ ᇱᇱ (, ) <sup>ᇱ</sup> (, ) + 1ቇ ≥ ቆ௭మ ᇱᇱ (, ) Φሾℎ( × ഥ) , (, )ሿ *by* Φሾℎ(, ) , (, )ሿ. *If* : ℂ<sup>ଶ</sup> ××ഥ → ℂ*, then the admissibility condition (A) reduces to* ௭ <sup>ᇱ</sup> (, ) *whenever* = (, ), = ௭ , ቀ<sup>௧</sup> <sup>௦</sup> + 1ቁ ≤ <sup>ଵ</sup> ୯మ ୯ ' (,) + 1൨ , ∈ , ∈ ഥ\൫(∙, )൯ *and* ≥ ≥ 1. *When* =1 *we write* Φଵൣ, (∙, )൧ *as* Φൣ, (∙, )൧.
*and* ≥ ≥ 1. *When* =1 *we write* Φଵൣ, (∙, )൧ *as* Φൣ, (∙, )൧.
**Definition 3.** *[27] Let be a set in* ℂ*,* (∙, ) ∈ , *and a positive integer. The class of admissible functions* ൣ, (∙, )൧ *consists of those functions* : ℂ<sup>ଷ</sup> ××ഥ → ℂ *that satisfy the*
*Axioms* **2022**, *11*, x FOR PEER REVIEW 3 of 12
<sup>ᇲ</sup> (௭,) , ቀ<sup>௧</sup>
Φሾℎ( × ഥ) , (, )ሿ *by* Φሾℎ(, ) , (, )ሿ.
<sup>ᇲ</sup> (௭,) , ቀ<sup>௧</sup>
*Axioms* **2022**, *11*, x FOR PEER REVIEW 3 of 12
≥1 *for which* ( × ഥబ) ⊂ ( × ഥ) *and*
*admissibility condition*
<sup>ᇱ</sup> (, ) *and*
*whenever* = (, ), = ௭
<sup>ᇲ</sup> (௭,) ఊ
*where* ∈ , ∈ ഥ, ∈ഥ\൫(∙, )൯ *and* ≥ ≥ 1.
ቆ(, ),
*admissibility condition*
*If* : ℂ<sup>ଶ</sup> ××ഥ → ℂ*, then the admissibility condition (A) reduces to*
*and* ≥ ≥ 1. *When* =1 *we write* Φଵൣ, (∙, )൧ *as* Φൣ, (∙, )൧.
*and* ≥ ≥ 1. *When* =1 *we write* Φଵൣ, (∙, )൧ *as* Φൣ, (∙, )൧.
<sup>௦</sup> + 1ቁ ≤ <sup>ଵ</sup>
**Definition 3.** *[27] Let be a set in* ℂ*,* (∙, ) ∈ , *and a positive integer. The class of admissible functions* ൣ, (∙, )൧ *consists of those functions* : ℂ<sup>ଷ</sup> ××ഥ → ℂ *that satisfy the*
ቆ(, ),
௭ <sup>ᇱ</sup> (, )
*In the special case when* ℎ(∙, ) *is an analytic mapping of* ×ഥ *onto* ≠ ℂ *we denote the class*
*If* : ℂ<sup>ଶ</sup> ××ഥ → ℂ*, then the admissibility condition (A) reduces to*
*admissibility condition*
*admissibility condition*
*whenever* = (, ), = ௭
(, ) as follows:
() (, ) = (, ),
<sup>ᇱ</sup> (, ) = ௭
≥1 *for which* ( × ഥబ) ⊂ ( × ഥ) *and*
ᇱᇱ (, )
*where* ∈ , ∈ ഥ, ∈ഥ\൫(∙, )൯ *and* ≥ ≥ 1.
() ௭
(, ) as follows:
in [28].
௭
() ௭
in [28].
*and*
*then*
() ቆ௭మ
*whenever* = (, ), = ௭
Φሾℎ( × ഥ) , (, )ሿ *by* Φሾℎ(, ) , (, )ሿ.
**Definition 3.** *[27] Let be a set in* ℂ*,* (∙, ) ∈ , *and a positive integer. The class of admissible functions* ൣ, (∙, )൧ *consists of those functions* : ℂ<sup>ଷ</sup> ××ഥ → ℂ *that satisfy the*
<sup>௦</sup> + 1ቁ ≤ <sup>ଵ</sup>
*In the special case when* ℎ(∙, ) *is an analytic mapping of* ×ഥ *onto* ≠ ℂ *we denote the class*
௭ <sup>ᇱ</sup> (, )
୯మ '' (,) ୯ '
Miller—Mocanu lemma given in [18] was rewritten in [27] for functions (, ) and
*Axioms* **2022**, *11*, x FOR PEER REVIEW 3 of 12
**Lemma 1.** *([17],[27]) Let* (, ) ∈ () *and let* (, ) =+() + ାଵ()ାଵ + ⋯ *with* () *holomorphic functions in* ഥ*,* ≥ , (, ) ≢ *and* ≥ 1. *If* (, ) *is not subordinate to* (, ), *then there exist points* = ఏబ ∈ *and* ∈ \൫(, )൯ *and an* ≥
> <sup>ᇲ</sup> (௭,) , ቀ<sup>௧</sup>
ᇱᇱ (, )
Miller—Mocanu lemma given in [18] was rewritten in [27] for functions (, ) and
*Axioms* **2022**, *11*, x FOR PEER REVIEW 3 of 12
**Lemma 1.** *([17],[27]) Let* (, ) ∈ () *and let* (, ) =+() + ାଵ()ାଵ + ⋯ *with* () *holomorphic functions in* ഥ*,* ≥ , (, ) ≢ *and* ≥ 1. *If* (, ) *is not subordinate to* (, ), *then there exist points* = ఏబ ∈ *and* ∈ \൫(, )൯ *and an* ≥
<sup>ᇲ</sup> (௭,)
୯మ '' (,) ୯ '
(, , ; , ) ∈ (A)
(, , ; , ) ∈ (A)
(,) + 1൨ , ∈ , ∈ ഥ\൫(∙, )൯
; , ቇ ∈ , (A')
<sup>௦</sup> + 1ቁ ≤ <sup>ଵ</sup>
*In the special case when* ℎ(∙, ) *is an analytic mapping of* ×ഥ *onto* ≠ ℂ *we denote the class*
''(,)
**Lemma 1.** *([17],[27]) Let* (, ) ∈ () *and let* (, ) =+() + ାଵ()ାଵ + ⋯ *with* () *holomorphic functions in* ഥ*,* ≥ , (, ) ≢ *and* ≥ 1. *If* (, ) *is not subor-*
**Definition 3.** *[27] Let be a set in* ℂ*,* (∙, ) ∈ , *and a positive integer. The class of admissible functions* ൣ, (∙, )൧ *consists of those functions* : ℂ<sup>ଷ</sup> ××ഥ → ℂ *that satisfy the*
(,) + 1൨ , ∈ , ∈ ഥ\൫(∙, )൯
; , ቇ ∈ , (A')
୯మ '' (,) ୯ '
(, , ; , ) ∈ (A)
(, , ; , ) ∈ (A)
(,) + 1൨ , ∈ , ∈ ഥ\൫(∙, )൯
$$h(z, \zeta) = \left(1 + \frac{1}{\gamma}\right) q(z, \zeta) + \frac{1}{\gamma} z \cdot q'\_z(z, \zeta), z \in \mathcal{U}, \zeta \in \overline{\mathcal{U}}.$$
*If φ* 0 *z* (*a*(*ζ*), *c*(*ζ*); *z*, *ζ*) *and L* 0 *z* (*a*(*ζ*), *c*(*ζ*); *z*, *ζ*) ∈ *H<sup>ζ</sup>* [1, 1] ∩ *Q<sup>ζ</sup>* (1) *are univalent functions in U for all ζ* ∈ *U, then strong differential superordination* **Lemma 2.** *[28] Let* ℎ(, ) *be convex in for all* ∈ ഥ *with* ℎ(0, ) = , ≠ 0, > 0 *and* ∈ሾ, 1ሿ ∩ . *If* (, ) <sup>+</sup> ௭ <sup>ᇲ</sup> (௭,) ఊ *is univalent in for all* ∈ ഥ, This lemma will be used in the next section for proving the theorems which contain the original results. Another helpful result which will be used is the next lemma proved Miller—Mocanu lemma given in [18] was rewritten in [27] for functions (, ) and ቆ(, ), ௭ <sup>ᇱ</sup> (, ) ; , ቇ ∈ , (A') *Axioms* **2022**, *11*, x FOR PEER REVIEW 3 of 12
$$h(z,\zeta) \rightsquigarrow \phi'\_z(a(\zeta), c(\zeta); z, \zeta)$$
*is univalent in for all* ∈ ഥ,
ᇱ
Miller—Mocanu lemma given in [18] was rewritten in [27] for functions (, ) and
*and implies* ∈ሾ, 1ሿ ∩ . *If* (, ) <sup>+</sup> ௭ *missible functions* ൣ, (∙, )൧ *consists of those functions* : ℂ<sup>ଷ</sup> ××ഥ → ℂ *that satisfy the admissibility condition Axioms* **2022**, *11*, x FOR PEER REVIEW 3 of 12
$$q(z,\zeta) \prec \mathcal{L}'\_z(a(\zeta), c(\zeta); z, \zeta), \; z \in \mathcal{U}, \; \zeta \in \overline{\mathcal{U}}.$$
*then* (, ) ⪻ (, ), ∈ , ∈ ഥ. *The function is convex and is the best subordinant.* (, ) <sup>=</sup> ఊ න ℎ(, )ఊିଵ ௭ , In Theorems 1 and 2, parameter *γ* is a real number, *γ* > 0. In the next theorem, a necessary and sufficient condition is determined such that Kummer–Bernardi integral operator is the best subordinant for a certain strong differential superordination considering *γ* a complex number with *Re γ* > 0. ≥1 *for which* ( × ഥబ) ⊂ ( × ഥ) *and* () (, ) = (, ), () ௭ <sup>ᇱ</sup> (, ) = ௭ <sup>ᇱ</sup> (, ) *and* () ቆ௭మ ᇱᇱ (, ) ௭ <sup>ᇱ</sup> (, ) + 1ቇ ≥ ቆ௭మ ᇱᇱ (, ) ௭ <sup>ᇱ</sup> (, ) + 1ቇ. **Lemma 1.** *([17],[27]) Let* (, ) ∈ () *and let* (, ) =+() + ାଵ()ାଵ + ⋯ *with* () *holomorphic functions in* ഥ*,* ≥ , (, ) ≢ *and* ≥ 1. *If* (, ) *is not subordinate to* (, ), *then there exist points* = ఏబ ∈ *and* ∈ \൫(, )൯ *and an* ≥ ≥1 *for which* ( × ഥబ) ⊂ ( × ഥ) *and whenever* = (, ), = ௭ <sup>ᇲ</sup> (௭,) , ቀ<sup>௧</sup> <sup>௦</sup> + 1ቁ ≤ <sup>ଵ</sup> ୯మ '' (,) ୯ ' (,) + 1൨ , ∈ , ∈ ഥ\൫(∙, )൯ *and* ≥ ≥ 1. *When* =1 *we write* Φଵൣ, (∙, )൧ *as* Φൣ, (∙, )൧. *In the special case when* ℎ(∙, ) *is an analytic mapping of* ×ഥ *onto* ≠ ℂ *we denote the class* **Definition 3.** *[27] Let be a set in* ℂ*,* (∙, ) ∈ , *and a positive integer. The class of admissible functions* ൣ, (∙, )൧ *consists of those functions* : ℂ<sup>ଷ</sup> ××ഥ → ℂ *that satisfy the admissibility condition*
The connection between univalent function theory and hypergeometric functions was established in 1985 when de Branges used the generalized hypergeometric function for proving Bieberbach's conjecture [29]. Once hypergeometric functions were considered in studies regarding univalent functions, confluent hypergeometric function was used in (, ) ⪻ (, ), ∈ , ∈ ഥ. *The function is convex and is the best subordinant.* **Theorem 3.** *Let h*(*z*, *ζ*) *with h*(0, *ζ*) = *a be a convex function in the unit disc for all ζ* ∈ *U and let γ be a complex number with Re γ* > 0*. Consider the confluent hypergeometric function φ*(*a*(*ζ*), *c*(*ζ*); *z*, *ζ*) *defined by (5) and Kummer–Bernardi integral operator B*(*a*(*ζ*), *c*(*ζ*); *z*, *ζ*) *given by (3). Let p*(*z*, *ζ*) ∈ *H<sup>ζ</sup>* [*a*, 1] ∩ *Q<sup>ζ</sup>* (*a*). This lemma will be used in the next section for proving the theorems which contain the original results. Another helpful result which will be used is the next lemma proved in [28]. () (, ) = (, ), () ௭ <sup>ᇱ</sup> (, ) = ௭ <sup>ᇱ</sup> (, ) *and* () ቆ௭మ ᇱᇱ (, ) ௭ <sup>ᇱ</sup> (, ) + 1ቇ ≥ ቆ௭మ ᇱᇱ (, ) ௭ <sup>ᇱ</sup> (, ) + 1ቇ. Φሾℎ( × ഥ) , (, )ሿ *by* Φሾℎ(, ) , (, )ሿ. *If* : ℂ<sup>ଶ</sup> ××ഥ → ℂ*, then the admissibility condition (A) reduces to* ቆ(, ), ௭ <sup>ᇱ</sup> (, ) ; , ቇ ∈ , (A') *whenever* = (, ), = ௭ <sup>ᇲ</sup> (௭,) , ቀ<sup>௧</sup> <sup>௦</sup> + 1ቁ ≤ <sup>ଵ</sup> ୯మ '' (,) ୯ ' (,) + 1൨ , ∈ , ∈ ഥ\൫(∙, )൯ *and* ≥ ≥ 1. *When* =1 *we write* Φଵൣ, (∙, )൧ *as* Φൣ, (∙, )൧. *In the special case when* ℎ(∙, ) *is an analytic mapping of* ×ഥ *onto* ≠ ℂ *we denote the class*
many investigations. One of the first papers which investigated confluent hypergeometric function and gave conditions for its univalence was published in 1990 [30]. Ever since The connection between univalent function theory and hypergeometric functions was established in 1985 when de Branges used the generalized hypergeometric function for proving Bieberbach's conjecture [29]. Once hypergeometric functions were considered *If p*(*z*, *ζ*) + *z*·*p* 0 *z* (*z*,*ζ*) *γ is univalent in U for all* ζ ∈ *U and the following strong differential superordination is satisfied* **Lemma 2.** *[28] Let* ℎ(, ) *be convex in for all* ∈ ഥ *with* ℎ(0, ) = , ≠ 0, > 0 *and* ∈ሾ, 1ሿ ∩ . *If* (, ) <sup>+</sup> ௭ <sup>ᇲ</sup> (௭,) ఊ *is univalent in for all* ∈ ഥ, This lemma will be used in the next section for proving the theorems which contain the original results. Another helpful result which will be used is the next lemma proved *where* ∈ , ∈ ഥ, ∈ഥ\൫(∙, )൯ *and* ≥ ≥ 1. Φሾℎ( × ഥ) , (, )ሿ *by* Φሾℎ(, ) , (, )ሿ. *If* : ℂ<sup>ଶ</sup> ××ഥ → ℂ*, then the admissibility condition (A) reduces to*
$$B(a(\zeta), c(\zeta); z, \zeta) + \frac{z \cdot B\_z'(a(\zeta), c(\zeta); z, \zeta)}{\gamma} \prec p(z, \zeta) + \frac{z \cdot p\_z'(z, \zeta)}{\gamma},\tag{26}$$
௭
ᇱ
ᇱ
*is univalent in for all* ∈ ഥ,
,
(, , ; , ) ∈ (A)
then, aspects of its univalence were further investigated [31,32], it was considered in con*then* **Lemma 1.** *([17],[27]) Let* (, ) ∈ () *and let* (, ) =+() + ାଵ()ାଵ + ⋯ *with* () *holomorphic functions in* ഥ*,* ≥ , (, ) ≢ *and* ≥ 1. *If* (, ) *is not subor-*
(, ) as follows:
() ௭
in [28].
*and*
*then*
*then*
in [28].
$$q(z,\zeta) = \mathcal{B}(a(\zeta), c(\zeta); z, \zeta) \prec q(z, \zeta), z \in \mathcal{U}, \zeta \in \overline{\mathcal{U}}.$$
*then and* ௭ *Function q*(*z*, *ζ*) = *B*(*a*(*ζ*), *c*(*ζ*); *z*, *ζ*) *is convex and is the best subordinant.* ≥1 *for which* ( × ഥబ) ⊂ ( × ഥ) *and* () (, ) = (, ), (, ) as follows:
*where* ∈ , ∈ ഥ, ∈ഥ\൫(∙, )൯ *and* ≥ ≥ 1.
<sup>ᇲ</sup> (௭,) ఊ
(, ) ⪻ (, ), ∈ , ∈ ഥ. *The function is convex and is the best subordinant.* (, ) <sup>=</sup> ఊ න ℎ(, )ఊିଵ , **Proof.** Lemma 2 will be used for the proof of this theorem. Using the definition of Kummer– Bernardi operator given by (3) and differentiating this relation with respect to *z*, we obtain: <sup>ᇱ</sup> (, ) = ௭ <sup>ᇱ</sup> (, ) *and* () ቆ௭మ ᇱᇱ (, ) ௭ <sup>ᇱ</sup> (, ) + 1ቇ ≥ ቆ௭మ ᇱᇱ (, ) ௭ <sup>ᇱ</sup> (, ) + 1ቇ. **Lemma 1.** *([17],[27]) Let* (, ) ∈ () *and let* (, ) =+() + ାଵ()ାଵ + ⋯ *with* () *holomorphic functions in* ഥ*,* ≥ , (, ) ≢ *and* ≥ 1. *If* (, ) *is not subordinate to* (, ), *then there exist points* = ఏబ ∈ *and* ∈ \൫(, )൯ *and an* ≥
$$\gamma \cdot z^{\gamma - 1} \cdot B(a(\zeta), c(\zeta); z, \zeta) + z^{\gamma} \cdot B\_2'(a(\zeta), c(\zeta); z, \zeta) = \gamma \cdot h(z, \zeta) \cdot z^{\gamma - 1}, \; z \in \mathsf{U}, \; \zeta \in \overline{\mathsf{U}}.$$
was established in 1985 when de Branges used the generalized hypergeometric function for proving Bieberbach's conjecture [29]. Once hypergeometric functions were considered *The function is convex and is the best subordinant.* After a simple calculation, we get: the original results. Another helpful result which will be used is the next lemma proved () (, ) = (, ), () ௭ <sup>ᇱ</sup> (, ) = ௭ <sup>ᇱ</sup> (, ) *and*
ఊ
∈ሾ, 1ሿ ∩ . *If* (, ) <sup>+</sup> ௭
$$B(a(\zeta), c(\zeta); z; \zeta) + \frac{z \cdot B\_z'(a(\zeta), c(\zeta); z; \zeta)}{\gamma} = h(z, \zeta), z \in \mathcal{U}, \zeta \in \overline{\mathcal{U}}.\tag{27}$$
function and gave conditions for its univalence was published in 1990 [30]. Ever since then, aspects of its univalence were further investigated [31,32], it was considered in conwas established in 1985 when de Branges used the generalized hypergeometric function for proving Bieberbach's conjecture [29]. Once hypergeometric functions were considered Using (27), the strong differential subordination (26) becomes ∈ሾ, 1ሿ ∩ . *If* (, ) <sup>+</sup> ௭ <sup>ᇲ</sup> (௭,) *is univalent in for all* ∈ ഥ, This lemma will be used in the next section for proving the theorems which contain
was established in 1985 when de Branges used the generalized hypergeometric function for proving Bieberbach's conjecture [29]. Once hypergeometric functions were considered
function and gave conditions for its univalence was published in 1990 [30]. Ever since then, aspects of its univalence were further investigated [31,32], it was considered in connection with other important functions [33–37] and it was used in the definition of new
$$h(z, \zeta) \ll p(z, \zeta) + \frac{z \cdot p\_z'(z, \zeta)}{\gamma}, z \in \mathsf{U}, \zeta \in \overline{\mathsf{U}}.$$
then, aspects of its univalence were further investigated [31,32], it was considered in connection with other important functions [33–37] and it was used in the definition of new (, ) <sup>=</sup> ఊ න ℎ(, )ఊିଵ ௭ , Since *h*(*z*, *ζ*) is a convex function and *p*(*z*, *ζ*) + *z*·*p* 0 *z* (*z*,*ζ*) *γ* is univalent in *U* for all *ζ* ∈ *U*, by applying Lemma 2 we obtain: **Lemma 2.** *[28] Let* ℎ(, ) *be convex in for all* ∈ ഥ *with* ℎ(0, ) = , ≠ 0, > 0 *and* ∈ሾ, 1ሿ ∩ . *If* (, ) <sup>+</sup> ௭ <sup>ᇲ</sup> (௭,) ఊ *is univalent in for all* ∈ ഥ,
$$q(z,\zeta) = B(a(\zeta), c(\zeta); z, \zeta) \prec p(z, \zeta), z \in \mathcal{U}, \zeta \in \overline{\mathcal{U}}.$$
(, ) ⪻ (, ), ∈ , ∈ ഥ. *The function is convex and is the best subordinant. and* (, ) <sup>=</sup> ௭ , Since function *q*(*z*, *ζ*) = *B*(*a*(*ζ*), *c*(*ζ*); *z*, *ζ*) satisfies Equation (27) and is analytic in *U* for all *ζ* ∈ *U*, we conclude that *q*(*z*, *ζ*) = *B*(*a*(*ζ*), *c*(*ζ*); *z*, *ζ*) is the best subordinant.
$$\begin{array}{ccccc}\hline\\\text{Example 1.} & \text{Let } a = -1, \ c = \frac{i}{\mathbb{Z}\_{\succ}^{\times}} & \frac{i}{\mathbb{Z}\_{\succ}^{\times}} \neq 0, -1, -2, \dots, \zeta \neq 0, \ \gamma \in \mathbb{C}, \text{ Re } \gamma > 0. \text{ We evaluate:}\\\hline\\\hline\\\end{array}$$
(, ) ⪻ (, ), ∈ , ∈ ഥ.
The connection between univalent function theory and hypergeometric functions was established in 1985 when de Branges used the generalized hypergeometric function for proving Bieberbach's conjecture [29]. Once hypergeometric functions were considered in studies regarding univalent functions, confluent hypergeometric function was used in many investigations. One of the first papers which investigated confluent hypergeometric function and gave conditions for its univalence was published in 1990 [30]. Ever since then, aspects of its univalence were further investigated [31,32], it was considered in connection with other important functions [33–37] and it was used in the definition of new
$$\phi\left(-1, \frac{i}{2\zeta}; z, \zeta\right) = 1 + \frac{-1}{\frac{i}{2\zeta}} \cdot \frac{z}{1!} = 1 - \frac{2\zeta}{i} \frac{z}{i} = 1 + 2i\zeta z.$$
*Further, we use this expression to obtain Kummer–Bernardi integral operator's expression:* () (, ) = (, ), () ௭ <sup>ᇱ</sup> (, ) = ௭ <sup>ᇱ</sup> (, ) *and* **Lemma 1.** *([17],[27]) Let* (, ) ∈ () *and let* (, ) =+() + ାଵ()ାଵ + ⋯ *with* () *holomorphic functions in* ഥ*,* ≥ , (, ) ≢ *and* ≥ 1. *If* (, ) *is not subor-*
*Axioms* **2022**, *11*, x FOR PEER REVIEW 3 of 12
<sup>ᇲ</sup> (௭,) , ቀ<sup>௧</sup>
Φሾℎ( × ഥ) , (, )ሿ *by* Φሾℎ(, ) , (, )ሿ.
<sup>ᇲ</sup> (௭,) , ቀ<sup>௧</sup>
*Axioms* **2022**, *11*, x FOR PEER REVIEW 3 of 12
*where* ∈ , ∈ ഥ, ∈ഥ\൫(∙, )൯ *and* ≥ ≥ 1.
ቆ(, ),
*If* : ℂ<sup>ଶ</sup> ××ഥ → ℂ*, then the admissibility condition (A) reduces to*
*and* ≥ ≥ 1. *When* =1 *we write* Φଵൣ, (∙, )൧ *as* Φൣ, (∙, )൧.
*and* ≥ ≥ 1. *When* =1 *we write* Φଵൣ, (∙, )൧ *as* Φൣ, (∙, )൧.
<sup>௦</sup> + 1ቁ ≤ <sup>ଵ</sup>
**Definition 3.** *[27] Let be a set in* ℂ*,* (∙, ) ∈ , *and a positive integer. The class of admissible functions* ൣ, (∙, )൧ *consists of those functions* : ℂ<sup>ଷ</sup> ××ഥ → ℂ *that satisfy the*
ቆ(, ),
௭ <sup>ᇱ</sup> (, )
*In the special case when* ℎ(∙, ) *is an analytic mapping of* ×ഥ *onto* ≠ ℂ *we denote the class*
*If* : ℂ<sup>ଶ</sup> ××ഥ → ℂ*, then the admissibility condition (A) reduces to*
**Definition 3.** *[27] Let be a set in* ℂ*,* (∙, ) ∈ , *and a positive integer. The class of admissible functions* ൣ, (∙, )൧ *consists of those functions* : ℂ<sup>ଷ</sup> ××ഥ → ℂ *that satisfy the*
<sup>௦</sup> + 1ቁ ≤ <sup>ଵ</sup>
*In the special case when* ℎ(∙, ) *is an analytic mapping of* ×ഥ *onto* ≠ ℂ *we denote the class*
௭ <sup>ᇱ</sup> (, )
୯మ '' (,) ୯ '
Miller—Mocanu lemma given in [18] was rewritten in [27] for functions (, ) and
**Lemma 1.** *([17],[27]) Let* (, ) ∈ () *and let* (, ) =+() + ାଵ()ାଵ + ⋯
୯మ '' (,) ୯ '
(, , ; , ) ∈ (A)
(, , ; , ) ∈ (A)
(,) + 1൨ , ∈ , ∈ ഥ\൫(∙, )൯
; , ቇ ∈ , (A')
(,) + 1൨ , ∈ , ∈ ഥ\൫(∙, )൯
; , ቇ ∈ , (A')
$$\begin{split} B\left(\boldsymbol{\phi}\left(-1,\frac{i}{\Delta\_{\boldsymbol{\zeta}}^{\boldsymbol{\gamma}}};z,\boldsymbol{\zeta}\right)\right) &= \frac{\gamma}{\boldsymbol{z}^{\gamma}}\int\_{0}^{z}\boldsymbol{\phi}\left(-1,\frac{i}{\Delta\_{\boldsymbol{\zeta}}^{\boldsymbol{\gamma}}};t,\boldsymbol{\zeta}\right)t^{\gamma-1}dt = \frac{\gamma}{\boldsymbol{z}^{\gamma}}\int\_{0}^{z}(1+2i\boldsymbol{\zeta}t)t^{\gamma-1}dt\\ &= \frac{\gamma}{\boldsymbol{z}^{\gamma}}\left(\frac{z^{\gamma}}{\gamma}+2i\underline{\zeta}z^{\frac{\gamma+1}{\gamma+1}}\right) = 1+2i\underline{\zeta}\frac{\gamma}{\gamma+1}z. \end{split}$$
*Functions p*(*z*, *ζ*) = 1 + *zζ and p*(*z*, *ζ*) + *z*·*p* 0 *z* (*z*,*ζ*) *<sup>γ</sup>* = 1 + *z ζ* + *ζ γ are univalent in U for all ζ* ∈ *U*. the original results. Another helpful result which will be used is the next lemma proved in [28]. () ቆ௭మ ᇱᇱ (, ) ௭ <sup>ᇱ</sup> (, ) + 1ቇ ≥ ቆ௭మ ᇱᇱ (, ) ௭ <sup>ᇱ</sup> (, ) + 1ቇ. (, , ; , ) ∈ (A) *whenever* = (, ), = ௭ <sup>ᇲ</sup> (௭,) , ቀ<sup>௧</sup> <sup>௦</sup> + 1ቁ ≤ <sup>ଵ</sup> ୯మ '' (,) (,) + 1൨ , ∈ , ∈ ഥ\൫(∙, )൯ **Definition 3.** *[27] Let be a set in* ℂ*,* (∙, ) ∈ , *and a positive integer. The class of admissible functions* ൣ, (∙, )൧ *consists of those functions* : ℂ<sup>ଷ</sup> ××ഥ → ℂ *that satisfy the*
*Using Theorem 3, we get:* **Lemma 2.** *[28] Let* ℎ(, ) *be convex in for all* ∈ ഥ *with* ℎ(0, ) = , ≠ 0, > 0 *and* This lemma will be used in the next section for proving the theorems which contain *admissibility condition*
(, ) as follows:
*admissibility condition*
*admissibility condition*
*whenever* = (, ), = ௭
*whenever* = (, ), = ௭
Φሾℎ( × ഥ) , (, )ሿ *by* Φሾℎ(, ) , (, )ሿ.
*where* ∈ , ∈ ഥ, ∈ഥ\൫(∙, )൯ *and* ≥ ≥ 1.
*If the following strong differential superordination is satisfied* ∈ሾ, 1ሿ ∩ . *If* (, ) <sup>+</sup> ௭ <sup>ᇲ</sup> (௭,) *is univalent in for all* ∈ ഥ, the original results. Another helpful result which will be used is the next lemma proved *and* ≥ ≥ 1. *When* =1 *we write* Φଵൣ, (∙, )൧ *as* Φൣ, (∙, )൧. *In the special case when* ℎ(∙, ) *is an analytic mapping of* ×ഥ *onto* ≠ ℂ *we denote the class* (, , ; , ) ∈ (A)
> <sup>ᇲ</sup> (௭,) ఊ
$$1 + 2i\zeta \frac{\gamma}{\gamma + 1} \cdot z + \frac{2i\zeta \cdot z}{\gamma + 1} \preccurlyeq 1 + z\left(\zeta + \frac{\zeta}{\gamma}\right),$$
...
௭
୯ '
*is univalent in for all* ∈ ഥ,
,
ఊ
*then* ∈ሾ, 1ሿ ∩ . *If* (, ) <sup>+</sup> ௭ *In the special case when* ℎ(∙, ) *is an analytic mapping of* ×ഥ *onto* ≠ ℂ *we denote the class*
in [28].
() ௭
*and*
$$1 + 2i\zeta \frac{\gamma}{\gamma + 1} \cdot z \preccurlyeq 1 + z\zeta, z \in \mathsf{U}, \; \zeta \in \mathsf{U}.$$
௭ <sup>ᇱ</sup> (, )
*then* (, ) ⪻ (, ), ∈ , ∈ ഥ. *and* (, ) <sup>=</sup> ௭ *Function q*(*z*, *ζ*) = 1 + 2*iζ γ γ*+1 ·*z is convex and is the best subordinant.* Miller—Mocanu lemma given in [18] was rewritten in [27] for functions (, ) and ቆ(, ), ௭ <sup>ᇱ</sup> (, ) ; , ቇ ∈ , (A')
*The function is convex and is the best subordinant.* ఊ න ℎ(, )ఊିଵ , *then* **Example 2.** *Let <sup>a</sup>* <sup>=</sup> <sup>−</sup>1, *<sup>c</sup>* <sup>=</sup> *<sup>i</sup>* 2*ζ* , *i* 2*ζ* 6= 0, −1, −2, . . . , *ζ* 6= 0, *γ* = 1 + *i* ∈ C, *Re γ* = 1 > 0*. We evaluate:* (, ) as follows: **Lemma 1.** *([17],[27]) Let* (, ) ∈ () *and let* (, ) =+() + ାଵ()ାଵ + ⋯ *where* ∈ , ∈ ഥ, ∈ഥ\൫(∙, )൯ *and* ≥ ≥ 1.
$$\phi\left(-1, \frac{i}{2\zeta}; z, \zeta\right) = 1 + \frac{-1}{\frac{i}{2\overline{\zeta}}} \cdot \frac{z}{1!} = 1 - \frac{2\zeta \cdot z}{i} = 1 + 2i\zeta z.$$
in studies regarding univalent functions, confluent hypergeometric function was used in The connection between univalent function theory and hypergeometric functions *Further, we use this expression to obtain Kummer–Bernardi integral operator's expression:* ≥1 *for which* ( × ഥబ) ⊂ ( × ഥ) *and* () (, ) = (, ), **Lemma 1.** *([17],[27]) Let* (, ) ∈ () *and let* (, ) =+() + ାଵ()ାଵ + ⋯
$$\begin{split} B\left(\boldsymbol{\phi}\left(-1,\frac{i}{2^{\gamma}};z,\zeta\right)\right) &= \frac{\gamma}{z^{\gamma}} \int\_{0}^{z} \boldsymbol{\phi}\left(-1,\frac{i}{2^{\gamma}};t,\zeta\right) t^{\gamma-1} dt = \frac{1+i}{z^{1+i}} \begin{subarray}{c} \begin{subarray}{c} \boldsymbol{z} \\ \end{subarray}}{\boldsymbol{0}} (1+2i\zeta t) t^{\gamma-1} dt \\ &= \frac{1+i}{z^{1+i}} \left(\frac{z^{1+i}}{1+i}+2i\zeta \frac{z^{1+i+1}}{1+i+1}\right) = 1+2i\zeta \frac{z(i+1)}{1+2} = 1+\frac{2}{5}(-1+3i)z\zeta. \end{subarray} \end{split}$$
function and gave conditions for its univalence was published in 1990 [30]. Ever since then, aspects of its univalence were further investigated [31,32], it was considered in connection with other important functions [33–37] and it was used in the definition of new *Functions p*(*z*, *ζ*) = 1 + *zζ and p*(*z*, *ζ*) + *z*·*p* 0 *z* (*z*,*ζ*) <sup>1</sup>+*<sup>i</sup>* <sup>=</sup> <sup>1</sup> <sup>+</sup> <sup>3</sup> 2 *zζ*(3 − *i*) *are univalent in U for all ζ* ∈ *U*. This lemma will be used in the next section for proving the theorems which contain the original results. Another helpful result which will be used is the next lemma proved in [28]. () ቆ௭మ ᇱᇱ (, ) ௭ <sup>ᇱ</sup> (, ) + 1ቇ ≥ ቆ௭మ ᇱᇱ (, ) ௭ <sup>ᇱ</sup> (, ) + 1ቇ.
*Using Theorem 3, we get:*
*If* 1 + <sup>3</sup> 2 *zζ*(3 − *i*) *is univalent in U for all ζ* ∈ *U and the following strong differential superordination is satisfied* **Lemma 2.** *[28] Let* ℎ(, ) *be convex in for all* ∈ ഥ *with* ℎ(0, ) = , ≠ 0, > 0 *and* ∈ሾ, 1ሿ ∩ . *If* (, ) <sup>+</sup> ௭ <sup>ᇲ</sup> (௭,) *is univalent in for all* ∈ ഥ, This lemma will be used in the next section for proving the theorems which contain the original results. Another helpful result which will be used is the next lemma proved
$$1 + 2i\tilde{\zeta}t \ll 1 + \frac{3}{2}z\tilde{\zeta}(3-i),$$
௭
*is univalent in for all* ∈ ഥ,
ఊ න ℎ(, )ఊିଵ
,
(, ) ⪻ (, ), ∈ , ∈ ഥ.
*and then* ∈ሾ, 1ሿ ∩ . *If* (, ) <sup>+</sup> ௭
in [28].
$$1 + \frac{2}{5}(-1 + 3i)z\zeta \ll 1 + z\zeta, z \in \mathcal{U}, \zeta \in \overline{\mathcal{U}}.$$
*then and* ௭ *Function q*(*z*, *ζ*) = 1 + <sup>2</sup> 5 (−1 + 3*i*)*zζ is convex and is the best subordinant.*
(, ) <sup>=</sup>
<sup>ᇲ</sup> (௭,) ఊ
ఊ
#### *The function is convex and is the best subordinant. then* **3. Discussion**
The connection between univalent function theory and hypergeometric functions was established in 1985 when de Branges used the generalized hypergeometric function for proving Bieberbach's conjecture [29]. Once hypergeometric functions were considered (, ) ⪻ (, ), ∈ , ∈ ഥ. *The function is convex and is the best subordinant.* The study presented in this paper is inspired by the nice results published which involve confluent hypergeometric function and certain operators defined by using this interesting function. For this research, the environment of the theory of strong differential
in studies regarding univalent functions, confluent hypergeometric function was used in many investigations. One of the first papers which investigated confluent hypergeometric
The connection between univalent function theory and hypergeometric functions
nection with other important functions [33–37] and it was used in the definition of new
in studies regarding univalent functions, confluent hypergeometric function was used in many investigations. One of the first papers which investigated confluent hypergeometric function and gave conditions for its univalence was published in 1990 [30]. Ever since then, aspects of its univalence were further investigated [31,32], it was considered in connection with other important functions [33–37] and it was used in the definition of new superordination is considered. Confluent hypergeometric function and Kummer–Bernardi and Kummer–Libera operators defined in [38] are used in order to obtain certain strong differential superordinations. Their best subordinants are given in the three theorems proved in the main results part. Theorems 1 and 2 use the convexity of confluent hypergeometric function *φ*(*a*(*ζ*), *c*(*ζ*); *z*, *ζ*) given in (5) where it is adapted to certain classes of analytic functions specific for the theory of strong differential superordination. The methods related to strong differential superordination theory are applied in order to find necessary conditions for Kummer–Bernardi integral operator presented in Definition 5, relation (3), to be the best subordinant of a certain strong differential superordination involving confluent hypergeometric function *φ*(*a*(*ζ*), *c*(*ζ*); *z*, *ζ*). As corollary, the similar result is given for Kummer–Libera operator. For those two theorems, the parameter *γ* is a real number, *γ* > 0. In Theorem 3, *γ* ∈ C, with *Re γ* > 0 is considered and a necessary and sufficient condition is determined such that Kummer–Bernardi integral operator to be the best subordinant for a certain strong differential superordination. Two examples are constructed for the case when *γ* ∈ C, with *Re γ* > 0.
#### **4. Conclusions**
In this paper, new strong differential superordinations are investigated using a special form of confluent hypergeometric function given in (5) and two operators previously introduced in [38]. In the three theorems proved as a result of the study, the two operators called Kummer–Bernardi and Kummer–Libera integral operators are the best subordinants of the strong differential superordinations.
The novelty of the study resides in the forms of the confluent hypergeometric function and of the two operators considered by adaptation to the new classes depending on the extra parameter *ζ* introduced in the theory of strong differential subordination in [27].
As future studies, the dual notion of strong differential subordination can be considered for investigations concerning confluent hypergeometric function and the two operators used in the present study. Sandwich-type results could be obtained as seen in recent papers [13,39,40].
New subclasses of univalent functions could be introduced in the context of strong differential subordination and superordination theories using the operators presented in this paper as seen in [41].
It might also be interesting to consider other hypergeometric functions and operators defined with them following the ideas presented in this paper.
**Author Contributions:** Conceptualization, G.I.O. and G.O.; methodology, G.I.O., G.O. and A.M.R.; software, G.I.O.; validation, G.I.O., G.O. and A.M.R.; formal analysis, G.I.O. and G.O.; investigation, G.I.O., G.O. and A.M.R.; resources, G.I.O. and G.O.; data curation, G.I.O. and G.O.; writing—original draft preparation, G.O.; writing—review and editing, G.I.O. and A.M.R.; visualization, G.I.O. and A.M.R.; supervision, G.O.; project administration, G.I.O.; funding acquisition, G.I.O. and A.M.R. All authors have read and agreed to the published version of the manuscript.
**Funding:** This research received no external funding.
**Data Availability Statement:** Not applicable.
**Conflicts of Interest:** The authors declare no conflict of interest.
#### **References**
### *Article* **Certain Subclasses of Bi-Starlike Function of Complex Order Defined by Erdély–Kober-Type Integral Operator**
**Alhanouf Alburaikan 1,† , Gangadharan Murugusundaramoorthy 2,† and Sheza M. El-Deeb 1,3,\* ,†**
**Abstract:** In the present paper, we introduce new subclasses of bi-starlike and bi-convex functions of complex order associated with Erdély–Kober-type integral operator in the open unit disc and find the estimates of initial coefficients in these classes. Moreover, we obtain Fekete-Szeg˝o inequalities for functions in these classes. Some of the significances of our results are pointed out as corollaries.
**Keywords:** univalent functions; analytic functions; bi-univalent functions; coefficient bounds; bi-starlike and bi-convex functions of complex order; fractional calculus; Erdély–Kober-type integral operator
**MSC:** 30C45; 30C50; 30C55
#### **1. Introduction and Preliminaries**
Let A signify the class of functions of the following form:
$$f(\xi) = \xi + \sum\_{n=2}^{\infty} a\_n \xi^n \tag{1}$$
which are analytic in the open unit disc U = {*ξ* : |*ξ*| < 1} and normalized as *f*(0) = 0 and *f* 0 (0) = 1. Furthermore, let S represent the class of all functions in A that are univalent in U. Some of the imperative and well-investigated subclasses of the univalent function class S include (for example) the class S∗ (*δ*) of starlike functions of order *δ* in U and the class K(*δ*) of convex functions of order *δ* (0 ≤ *δ* < 1) in U. It is known that if *f* ∈ S, then there exists inverse function *f* <sup>−</sup><sup>1</sup> because normalization is defined in some neighborhood of the origin. In some cases, *f* −1 can be defined in the entire U. Clearly, *f* −1 is also univalent. For this reason, class Σ is defined as follows.
It is well known that every function *f* ∈ S has an inverse *f* <sup>−</sup><sup>1</sup> defined by the following:
$$\begin{aligned} f^{-1}(f(\xi)) &= \xi \ (\xi \in \mathfrak{U})\\ \text{and} \quad f(f^{-1}(w)) &= w \ (|w| < r\_0(f); r\_0(f) \ge 1/4) \end{aligned}$$
where the following is the case.
$$f^{-1}(w) = g(w) = w - a\_2 w^2 + (2a\_2^2 - a\_3)w^3 - (5a\_2^3 - 5a\_2 a\_3 + a\_4)w^4 + \cdots \ . \tag{2}$$
**Citation:** Alburaikan, A.; Murugusundaramoorthy, G.; El-Deeb, S.M. Certain Subclasses of Bi-Starlike Function of Complex Order Defined by Erdély–Kober-Type Integral Operator. *Axioms* **2022**, *11*, 237. https://doi.org/10.3390/axioms 11050237
Academic Editors: Georgia Irina Oros and Federico G. Infusino
Received: 24 April 2022 Accepted: 16 May 2022 Published: 20 May 2022
**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.
**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).
A function *f*(*ξ*) ∈ A is said to be bi-univalent in U if both *f*(*ξ*) and *f* −1 (*ξ*) are univalent in U. Let Σ denote the class of bi-univalent functions in U given by (1). Note that the following functions:
$$f\_1(\xi) = \frac{\xi}{1 - \tilde{\xi}'} \qquad f\_2(\xi) = \frac{1}{2} \log \frac{1 + \tilde{\xi}}{1 - \tilde{\xi}'} \qquad f\_3(\xi) = -\log(1 - \tilde{\xi})$$
with their corresponding inverses
$$f\_1^{-1}(w) = \frac{w}{1+w'} \qquad f\_2^{-1}(w) = \frac{e^{2w}-1}{e^{2w}+1} , \qquad f\_3^{-1}(w) = \frac{e^w-1}{e^w} $$
are elements of Σ (see [1–3]). Certain subclasses of Σ are explicitly bi-starlike functions of order *δ*(0 < *δ* ≤ 1) denoted by S<sup>∗</sup> Σ (*δ*) and bi-convex function of order *δ* designated by KΣ(*δ*) familiarized by Brannan and Taha [1]. For each *f* ∈ S<sup>∗</sup> Σ (*δ*) and *f* ∈ KΣ(*δ*), non-sharp estimates on the first two Taylor–Maclaurin coefficients |*a*2| and |*a*3| were established [1,2], but the problem to find the general coefficient bounds on the following Taylor–Maclaurin coefficients:
$$|a\_n| \qquad (n \in \mathbb{N} \backslash \{1, 2\}; \ N := \{1, 2, 3, \dots\})$$
is still an open problem (see [1–5]). Several researchers (see [6–11]) have introduced and explored some inspiring subclasses Σ and they have initiated non-sharp estimates |*a*2| and |*a*3|. For two functions *f*<sup>1</sup> and *f*<sup>2</sup> ∈ A, we say that function *f*<sup>1</sup> is subordinate to *f*<sup>2</sup> if there exists a Schwarz function *ω* that is holomorphic in U with property *w*(0) = 0; |*ω*(*ξ*)| < 1 and satisfying *f*1(*ξ*) = *f*2(*w*(*ξ*)) This subordination is symbolically written as *f*1(*ξ*) ≺ *f*2(*ξ*). Lately, Ma and Minda [12]-unified subclasses of starlike and convex functions are subordinate to a general superordinate function. For this purpose, they considered an analytic function W with positive real parts in the unit disk U,W(0) = 1,W0 (0) > 0, and W maps U onto a region starlike with respect to 1 and is symmetric with respect to the real axis. In the consequence, it is assumed that W is an analytic function with positive real part in the unit disk U, with W(0) = 1,W0 (0) > 0, and W(U) is symmetric with respect to the real axis. Such functions are of the following form.
$$\mathfrak{W}(\mathfrak{f}) = 1 + \mathfrak{m}\_1 \mathfrak{f} + \mathfrak{m}\_2 \mathfrak{f}^2 + \mathfrak{m}\_3 \mathfrak{f}^3 + \cdots, \quad (\mathfrak{m}\_1 > 0). \tag{3}$$
The study of operators plays a central role in geometric function theory and its correlated fields. In the recent years, there has been an collective importance in problems concerning the evaluations of various differential and integral operators. For our study, we recall the Erdély–Kober type ([13] Ch. 5; also see [14–17]) for the integral operator definition, which shall be used throughout the paper as stated below.
#### *Erdély–Kober Fractional-Order Derivative*
Let *κ* > 0, *ς*, *τ* ∈ C be such that R(*τ* − *ς*) ≥ 0, an Erdély–Kober type integral operator:
$$\mathfrak{I}\_{\kappa}^{\xi,\tau}: \mathfrak{A} \to \mathfrak{A}$$
be defined for R(*τ* − *ς*) > 0 and R(*ς*) > −*κ* by the following.
$$\mathfrak{D}\_{\kappa}^{\varsigma,\tau}f(\xi) = \frac{\Gamma(\tau+\kappa)}{\Gamma(\varsigma+\kappa)} \frac{1}{\Gamma(\tau-\varsigma)} \int\_{0}^{1} (1-t)^{\tau-\varsigma-1} t^{\varsigma-1} f(\xi t^{\kappa}) dt, \kappa > 0. \tag{4}$$
For *κ* > 0, R(*τ* − *ς*) ≥ 0, R(*ϑ*) > −*κ* and *f* ∈ A of the form (1), we have the following:
$$\mathfrak{I}\_{\mathbb{K}}^{\xi,\mathsf{T}}f(\boldsymbol{\xi}) \quad = \; \mathsf{f} + \sum\_{n=2}^{\infty} \frac{\Gamma(\tau + \kappa)\Gamma(\boldsymbol{\xi} + n\kappa)}{\Gamma(\boldsymbol{\xi} + \kappa)\Gamma(\tau + n\kappa)} a\_{n} \boldsymbol{\xi}^{n} \quad (\boldsymbol{\xi} \in \mathsf{U}) \tag{5}$$
$$\mathfrak{l} = \begin{array}{c} \mathfrak{f} + \sum\_{n=2}^{\infty} \mathbf{Y}\_{\mathbb{K}}^{\xi, \mathsf{T}}(n) a\_n \mathfrak{f}^n \quad (\xi \in \mathfrak{U}) \end{array} \tag{6}$$
where the following is the case.
$$\Upsilon\_{\kappa}^{\xi,\tau}(n) = \frac{\Gamma(\tau+\kappa)\Gamma(\xi+n\kappa)}{\Gamma(\xi+\kappa)\Gamma(\tau+n\kappa)}\tag{7}$$
and Γ(*n* + 1) = *n*!.
Note that the following is the case.
$$\mathfrak{I}\_{\mathbb{K}}^{\xi,\mathfrak{c}}f(\mathfrak{f}) = f(\mathfrak{f})$$
**Remark 1.** *By fixing the parameters ς*, *τ*, *ϑ as mentioned below, the operator* I *ς*,*τ <sup>κ</sup> includes various operators studied in the literature as cited below:*
The motivation of our present investigation stems from (by Silverman and Silvia [27] (also see [28])) the seminal paper on bi-univalent functions by Srivastava et al. [8] and by the recent works by many authors (for example Deniz [7], Huo Tang et al. [6], EI-Deeb et al. [29–31], and Murugusundaramoorthy and Janani [32]). In the present paper, we introduce two new subclasses of the function class Σ of complex order *ϑ* ∈ C\{0}, involving the linear operator I *ς*,*τ <sup>κ</sup>* given in Definition 1. We find estimates on the coefficients |*a*2| and |*a*3| for functions *f* ∈ S *ς*,*τ* <sup>Σ</sup>,W(*ϑ*, `). Several related classes are also considered, and connections to earlier known results are provided. Moreover we obtain the Fekete-Szeg˝o inequalities for *f* ∈ S *ς*,*τ* <sup>Σ</sup>,W(*ϑ*, `) and *f* ∈ K *ς*,*τ* <sup>Σ</sup>,W(*ϑ*, `).
**Definition 1.** *Let f* ∈ Σ *be assumed by (1) and f* ∈ S *ς*,*τ* <sup>Σ</sup>,W(*ϑ*, `), *if the subsequent conditions holds:*
$$1 + \frac{1}{\theta} \left( \frac{\mathfrak{f}(\mathfrak{I}\_{\mathbb{K}}^{\xi, \mathsf{T}} f(\mathfrak{f}))'}{\mathfrak{I}\_{\mathbb{K}}^{\xi, \mathsf{T}} f(\mathfrak{f})} + \left( \frac{1 + \mathfrak{e}^{i\ell}}{2} \right) \frac{\mathfrak{f}^{2}(\mathfrak{I}\_{\mathbb{K}}^{\xi, \mathsf{T}} f(\mathfrak{f}))''}{\mathfrak{I}\_{\mathbb{K}}^{\xi, \mathsf{T}} f(\mathfrak{f})} - 1 \right) \prec \mathfrak{W}(\xi) \tag{8}$$
*and*
$$1 + \frac{1}{\theta} \left( \frac{w(\mathfrak{I}\_{\mathbf{x}}^{\xi, \mathsf{T}} g(w))'}{\mathfrak{I}\_{\mathbf{x}}^{\xi, \mathsf{T}} g(w)} + \left( \frac{1 + e^{i\ell}}{2} \right) \frac{w^2 (\mathfrak{I}\_{\mathbf{x}}^{\xi, \mathsf{T}} g(w))''}{\mathfrak{I}\_{\mathbf{x}}^{\xi, \mathsf{T}} g(w)} - 1 \right) \prec \mathfrak{W}(w),\tag{9}$$
*where ϑ* ∈ C\{0}; ` ∈ (−*π*, *π*]; *ξ*, *w* ∈ U *and g is given by (2).*
**Definition 2.** *Let f* ∈ Σ *be assumed by (1) and f* ∈ K *ς*,*τ* <sup>Σ</sup>,W(*ϑ*, `), *if the subsequent conditions are satisfied:*
$$1 + \frac{1}{\theta} \left( \frac{[\mathfrak{T}(\mathfrak{T}\_{\mathbf{x}}^{\xi, \mathsf{T}} f(\boldsymbol{\xi}))' + \left(\frac{1 + \epsilon^{i\ell}}{2}\right) \mathfrak{T}^2 (\mathfrak{T}\_{\mathbf{x}}^{\xi, \mathsf{T}} f(\boldsymbol{\xi}))' ]'}{(\mathfrak{T}\_{\mathbf{x}}^{\xi, \mathsf{T}} f(\boldsymbol{\xi}))'} - 1 \right) \prec \mathfrak{W}(\boldsymbol{\xi}) \tag{10}$$
*and*
$$1 + \frac{1}{\theta} \left( \frac{[w(\mathfrak{I}\_{\mathbf{x}}^{\xi, \mathbf{r}} \mathcal{g}(w))' + \left(\frac{1 + \epsilon^{i\ell}}{2}\right) w^2 (\mathfrak{I}\_{\mathbf{x}}^{\xi, \mathbf{r}} \mathcal{g}(w))'']'}{(\mathfrak{I}\_{\mathbf{x}}^{\xi, \mathbf{r}} \mathcal{g}(w))'} - 1 \right) \prec \mathfrak{W}(w),\tag{11}$$
*where ϑ* ∈ C\{0}; ` ∈ (−*π*, *π*]; *ξ*, *w* ∈ U *and g is given by (2).*
**Remark 2.** *For a function f*(*ξ*) ∈ Σ *specified by (1) and for* ` = *π, interpret that* S *ς*,*τ* <sup>Σ</sup>,W(*ϑ*, `) ≡ S *ς*,*τ* <sup>Σ</sup>,W(*ϑ*) *satisfies the ensuing conditions :*
1 + 1 *ϑ ξ*(I *ς*,*τ <sup>κ</sup> f*(*ξ*))<sup>0</sup> I *ς*,*τ <sup>κ</sup> f*(*ξ*) − 1 <sup>≺</sup> <sup>W</sup>(*ξ*) and 1 + 1 *ϑ w*(I *ς*,*τ <sup>κ</sup> g*(*w*))<sup>0</sup> I *ς*,*τ <sup>κ</sup> g*(*w*) − 1 ≺ W(*w*)
*where ϑ* ∈ C\{0}; *ξ*, *w* ∈ U *and g is given by (2).*
**Remark 3.** *A function f*(*ξ*) ∈ Σ *specified by (1) and for* ` = *π, we interpret that* K *ς*,*τ* <sup>Σ</sup>,W(*ϑ*, `) ≡ K *ς*,*τ* <sup>Σ</sup>,W(*ϑ*) *satisfies the ensuing conditions correspondingly:*
$$\left[1+\frac{1}{\theta}\left(\frac{\mathfrak{T}(\mathfrak{I}\_{\mathbf{x}}^{\xi,\mathsf{T}}f(\mathfrak{f}))''}{(\mathfrak{I}\_{\mathbf{x}}^{\xi,\mathsf{T}}f(\mathfrak{f}))'}\right)\right] \prec \mathfrak{W}(\mathfrak{f}) \text{ and } \left[1+\frac{1}{\theta}\left(\frac{w(\mathfrak{I}\_{\mathbf{x}}^{\xi,\mathsf{T}}g(w))''}{(\mathfrak{I}\_{\mathbf{x}}^{\xi,\mathsf{T}}g(w))'}\right)\right] \prec \mathfrak{W}(w).$$
*where ϑ* ∈ C\{0}; *ξ*, *w* ∈ U *and g is given by (2).*
**Remark 4.** *For a function f*(*ξ*) ∈ Σ *given by (1) and for ϑ* = 1, *we note that* S *ς*,*τ* <sup>Σ</sup>,W(*ϑ*, `) ≡ S *ς*,*τ* <sup>Σ</sup>,W(`) *and satisfies the following conditions, respectively:*
$$\left(\frac{\mathfrak{T}(\mathfrak{I}\_{\mathbb{K}}^{\xi,\mathsf{T}}f(\xi))'}{\mathfrak{I}\_{\mathbb{K}}^{\xi,\mathsf{T}}f(\xi)} + \left(\frac{1+\varepsilon^{i\ell}}{2}\right) \frac{\mathfrak{I}^2(\mathfrak{I}\_{\mathbb{K}}^{\xi,\mathsf{T}}f(\xi))''}{\mathfrak{I}\_{\mathbb{K}}^{\xi,\mathsf{T}}f(\xi)}\right) \prec \mathfrak{W}(\xi)$$
*and the following is the case.*
$$\left(\frac{w(\mathfrak{I}\_{\mathbf{x}}^{\xi,\mathsf{T}}g(w))'}{\mathfrak{I}\_{\mathbf{x}}^{\xi,\mathsf{T}}g(w)} + \left(\frac{1+\varepsilon^{i\ell}}{2}\right) \frac{w^2(\mathfrak{I}\_{\mathbf{x}}^{\xi,\mathsf{T}}g(w))''}{\mathfrak{I}\_{\mathbf{x}}^{\xi,\mathsf{T}}g(w)}\right) \prec \mathfrak{W}(w).$$
*Moreover,* K *ς*,*τ* <sup>Σ</sup>,W(*ϑ*, `) ≡ K *ς*,*τ* <sup>Σ</sup>,W(`) *and it satisfies the following conditions:*
$$\left( \frac{[\mathfrak{f}(\mathfrak{I}\_{\mathbb{K}}^{\xi,\mathsf{T}}f(\mathfrak{f}))' + \left(\frac{1+\epsilon^{i\ell}}{2}\right)\mathfrak{f}^{2}(\mathfrak{I}\_{\mathbb{K}}^{\xi,\mathsf{T}}f(\mathfrak{f}))'']'}{(\mathfrak{I}\_{\mathbb{K}}^{\xi,\mathsf{T}}f(\mathfrak{f}))'} \right) \prec \mathfrak{W}(\mathfrak{f})' $$
*and the following is the case:*
$$\left( \frac{[w(\mathfrak{I}\_{\mathbf{k}}^{\xi,\mathsf{T}}g(w))' + \left(\frac{1+\epsilon^{i\ell}}{2}\right)w^2(\mathfrak{I}\_{\mathbf{k}}^{\xi,\mathsf{T}}g(w))'']'}{(\mathfrak{I}\_{\mathbf{k}}^{\xi,\mathsf{T}}g(w))'} \right) \prec \mathfrak{W}(w)\_{\mathsf{T}}$$
*where* ` ∈ (−*π*, *π*]; *ξ*, *w* ∈ U *and g is given by (2).*
**2. Coefficient Estimates for** *f* ∈ S *ς***,***τ* **<sup>Σ</sup>,**W(*ϑ***,** `) **and** *f* ∈ K *ς***,***τ* **<sup>Σ</sup>,**W(*ϑ***,** `)
For notational simplicity, in the sequel we let the following be the case:
$$
\kappa > 0, \Re(\tau - \varsigma) \ge 0, \quad \Re(\varsigma) > -\kappa \quad \text{and} \quad \Im^{\varsigma, \tau}\_{\kappa} f(\overline{\varsigma}).
$$
and it is provided by (5):
$$\mathbf{Y}\_{2} = \mathbf{Y}\_{\kappa}^{\xi, \tau}(\mathbf{2}) = \frac{\Gamma(\tau + \kappa)\Gamma(\xi + 2\kappa)}{\Gamma(\xi + \kappa)\Gamma(\tau + 2\kappa)},\tag{12}$$
$$\mathbf{Y}\_3 = \mathbf{Y}\_\mathbf{k}^{\xi, \tau}(\mathfrak{Z}) = \frac{\Gamma(\tau + \kappa)\Gamma(\xi + 3\kappa)}{\Gamma(\xi + \kappa)\Gamma(\tau + 3\kappa)}\tag{13}$$
and the following.
$$
\ell \in ( -\pi, \pi ].
$$
For deriving our main results, we need the following lemma.
**Lemma 1.** *Ref. [33] states that if h* ∈ P*, then* |*c<sup>k</sup>* | ≤ 2 *for each k, where* P *is the family of all functions h analytic in* U *for which* <(*h*(*ξ*)) > 0 *and the following is the case.*
$$h(\mathfrak{f}) = 1 + c\_1 \mathfrak{f} + c\_2 \mathfrak{f}^2 + \dots \colon for \, \mathfrak{f} \in \mathfrak{U}.$$
Define the functions *p*(*ξ*) and *q*(*ξ*) by the following:
$$p(\xi) := \frac{1 + \mu(\xi)}{1 - \mu(\xi)} = 1 + \wp\_1 \mathfrak{f} + \wp\_2 \mathfrak{f}^2 + \dotsb$$
and the following.
$$q(w) := \frac{1 + v(w)}{1 - v(w)} = 1 + \mathfrak{q}\_1 w + \mathfrak{q}\_2 w^2 + \dots \dots$$
It follows that the following is the case:
$$\mu(\mathfrak{f}) := \frac{p(\mathfrak{f}) - 1}{p(\mathfrak{f}) + 1} = \frac{1}{2} \left[ \wp\_1 \mathfrak{f} + \left( \wp\_2 - \frac{\wp\_1^2}{2} \right) \mathfrak{f}^2 + \dotsb \right]$$
and
$$w(w) := \frac{q(w) - 1}{q(w) + 1} = \frac{1}{2} \left[ \mathfrak{q}\_1 w + \left( \mathfrak{q}\_2 - \frac{\mathfrak{q}\_1^2}{2} \right) w^2 + \dotsb \right].$$
Then, *p*(*ξ*) and *q*(*w*) are analytic in U with *p*(0) = 1 = *q*(0).
Since *u*, *v* : U → U, the functions *p*(*ξ*) and *q*(*w*) have a positive real part in U, and |℘*i* | ≤ 2 and |q*<sup>i</sup>* | ≤ 2 for each *i*.
**Theorem 1.** *Let f given by (1) be in the class* S *ς*,*τ* <sup>Σ</sup>,W(*ϑ*, `)*, ϑ* ∈ C\{0} *and* ` ∈ (−*π*, *π*]*. Then, we have the following:*
$$|a\_2| \le \frac{|\vartheta| \mathbf{m}\_1 \sqrt{\mathbf{m}\_1}}{\sqrt{|\vartheta| (5 + 3e^{i\ell}) \mathbf{Y}\_3 - (2 + e^{i\ell}) \mathbf{Y}\_2^2 |\mathbf{m}\_1^2 + (2 + e^{i\ell})^2 (\mathbf{m}\_1 - \mathbf{m}\_2) \mathbf{Y}\_2^2|}}\tag{14}$$
*and the following.*
$$|a\_3| \le \frac{|\boldsymbol{\theta}|^2 \mathbf{m}\_1^2}{|2 + e^{i\boldsymbol{\ell}}|^2 \mathbf{Y}\_2^2} + \frac{|\boldsymbol{\theta}| \mathbf{m}\_1}{|5 + 3e^{i\boldsymbol{\ell}}| \mathbf{Y}\_3}.\tag{15}$$
**Proof.** It follows from (8) and (9) that we have the following:
$$1 + \frac{1}{\theta} \left( \frac{\xi(\Im\_{\mathsf{X}}^{\xi,\mathsf{T}} f(\xi))'}{\Im\_{\mathsf{X}}^{\xi,\mathsf{T}} f(\xi)} + \left( \frac{1 + e^{i\ell}}{2} \right) \frac{\xi^2 (\Im\_{\mathsf{X}}^{\xi,\mathsf{T}} f(\xi))''}{\Im\_{\mathsf{X}}^{\xi,\mathsf{T}} f(\xi)} - 1 \right) = \mathfrak{W}(\mathsf{u}(\xi)) \tag{16}$$
and
$$1 + \frac{1}{\theta} \left( \frac{w(\mathfrak{I}\_{\mathbb{K}}^{\xi, \mathsf{T}} \mathcal{g}(w))'}{\mathfrak{I}\_{\mathbb{K}}^{\xi, \mathsf{T}} \mathcal{g}(w)} + \left( \frac{1 + e^{i\ell}}{2} \right) \frac{w^2 (\mathfrak{I}\_{\mathbb{K}}^{\xi, \mathsf{T}} \mathcal{g}(w))''}{\mathfrak{I}\_{\mathbb{K}}^{\xi, \mathsf{T}} \mathcal{g}(w)} - 1 \right) = \mathfrak{W}(v(w)),\tag{17}$$
where
$$\mathfrak{W}(\mathfrak{u}(\xi)) = \frac{1}{2}\mathfrak{m}\_1\wp\_1\mathfrak{f} + \left(\frac{1}{2}\mathfrak{m}\_1(\wp\_2 - \frac{\wp\_1^2}{2}) + \frac{1}{4}\mathfrak{m}\_2\wp\_1^2\right)\mathfrak{f}^2 + \dotsb \,. \tag{18}$$
and
$$\mathfrak{W}(v(w)) = \frac{1}{2}\mathfrak{m}\_1\mathfrak{q}\_1w + \left(\frac{1}{2}\mathfrak{m}\_1(\mathfrak{q}\_2 - \frac{\mathfrak{q}\_1^2}{2}) + \frac{1}{4}\mathfrak{m}\_2\mathfrak{q}\_1^2\right)w^2 + \dotsb \,. \tag{19}$$
For a given *f*(*z*) of form (1), a computation shows the following:
$$\frac{zf'(z)}{f(z)} = 1 + a\_2 \mathbf{Y}\_2 z + (2\mathbf{Y}\_3 a\_3 - a\_2^2 \mathbf{Y}\_2^2) z^2 + (3a\_4 \mathbf{Y}\_4 + a\_2^3 \mathbf{Y}\_2^3 - 3a\_3 a\_2 \mathbf{Y}\_2 \mathbf{Y}\_3) z^3 + \dotsb$$
and
$$\frac{z f''(z)}{f'(z)} = 2a\_2 \mathbf{Y}\_2^2 z + (6a\_3 \mathbf{Y}\_3 - 4a\_2^2 \mathbf{Y}\_2^2) z^2 + \dotsb \ .$$
Using these in the left hand side of (16) and (17), a simple computation produces the following:
$$\begin{split} 1 + \frac{1}{\theta} \left( \frac{\mathfrak{f}(\mathfrak{I}\_{\mathbf{x}}^{\xi,\mathsf{T}}f(\mathfrak{f}))'}{\mathfrak{I}\_{\mathbf{x}}^{\xi,\mathsf{T}}f(\mathfrak{f})} + \left( \frac{1 + e^{i\ell}}{2} \right) \frac{\mathfrak{f}^{2}(\mathfrak{I}\_{\mathbf{x}}^{\xi,\mathsf{T}}f(\mathfrak{f}))''}{\mathfrak{I}\_{\mathbf{x}}^{\xi,\mathsf{T}}f(\mathfrak{f})} - 1 \right) &= 1 + \frac{1}{\theta} (2 + e^{i\ell}) \mathfrak{Y}\_{2} a\_{2} \mathfrak{g} \\ &+ \frac{1}{\theta} \Big[ (\mathfrak{f} + 3e^{i\ell}) \mathfrak{Y}\_{3} a\_{3} - (2 + e^{i\ell}) \mathfrak{Y}\_{2}^{2} a\_{2}^{2} \Big] \mathfrak{g}^{2} + \cdots \end{split}$$
and
$$\begin{split} 1 + \frac{1}{\theta} \left( \frac{w(\mathfrak{I}\_{\mathbf{x}}^{\xi,\mathsf{T}} \operatorname{g}(w))'}{\mathfrak{I}\_{\mathbf{x}}^{\xi,\mathsf{T}} \operatorname{g}(w)} + \left( \frac{1 + e^{i\ell}}{2} \right) \frac{w^2 (\mathfrak{I}\_{\mathbf{x}}^{\xi,\mathsf{T}} \operatorname{g}(w))''}{\mathfrak{I}\_{\mathbf{x}}^{\xi,\mathsf{T}} \operatorname{g}(w)} - 1 \right) &= 1 - \frac{1}{\theta} (2 + e^{i\ell}) \operatorname{Y}\_{2} a\_{2} w \\ + \frac{1}{\theta} \Big( [2(5 + 3e^{i\ell}) \operatorname{Y}\_{3} - (2 + e^{i\ell}) \operatorname{Y}\_{2}^{2}] a\_{2}^{2} - (5 + 3e^{i\ell}) \operatorname{Y}\_{3} a\_{3} \Big) w^{2} = \cdot \cdot \cdot \,\_{1} \end{split}$$
Thus, by equating the coefficients of *ς* and *ς* 2 in (16) and (17), we obtain the following:
$$\frac{1}{2}(2+e^{i\ell})\mathbf{Y}\_2a\_2 = \frac{1}{2}\mathbf{m}\_1\wp\_{1\prime} \tag{20}$$
$$\frac{1}{\theta} \left[ (\mathbf{5} + 3\mathbf{e}^{i\ell}) \mathbf{Y}\_3 a\_3 - (\mathbf{2} + \mathbf{e}^{i\ell}) \mathbf{Y}\_2^2 a\_2^2 \right] = \frac{1}{2} \mathbf{m}\_1 (\wp\_2 - \frac{\wp\_1^2}{2}) + \frac{1}{4} \mathbf{m}\_2 \wp\_1^2 \tag{21}$$
$$-\frac{1}{\theta}(\mathbf{2} + e^{i\ell})\mathbf{Y}\_2 a\_2 = \frac{1}{2}\mathbf{m}\_1 \mathbf{q}\_{1\prime} \tag{22}$$
and
$$\frac{1}{\theta} \Big( [2(5+3e^{i\ell})\mathbf{Y}\_3 - (2+e^{i\ell})\mathbf{Y}\_2^2]a\_2^2 - (5+3e^{i\ell})\mathbf{Y}\_3a\_3 \Big) = \frac{1}{2}\mathbf{m}\_1(\mathbf{q}\_2 - \frac{\mathbf{q}\_1^2}{2}) + \frac{1}{4}\mathbf{m}\_2\mathbf{q}\_1^2. \tag{23}$$
From (20) and (22), we obtain the following:
$$
\wp\_1 = -\mathfrak{q}\_1 \tag{24}
$$
and
$$\begin{array}{rcl}8(2+e^{i\ell})^2\mathbf{Y}\_2^2a\_2^2 &=& \vartheta^2\mathbf{m}\_1^2(\wp\_1^2+\mathfrak{q}\_1^2) \\ a\_2^2 &=& \frac{\vartheta^2\mathbf{m}\_1^2(\wp\_1^2+\mathfrak{q}\_1^2)}{8(2+e^{i\ell})^2\mathbf{Y}\_2^2}. \end{array} \tag{25}$$
Now, by adding (21) and (23) and then using (25), we obtain the following.
$$a\_2^2 = \frac{\theta^2 \mathbf{m}\_1^3 (\wp\_2 + \mathfrak{q}\_2)}{\left(4\{\theta[(\mathfrak{F} + 3\mathfrak{e}^{i\ell})\mathbf{Y}\_3 - (\mathfrak{Z} + \mathfrak{e}^{i\ell})\mathbf{Y}\_2^2] \mathbf{m}\_1^2 + (2 + \mathfrak{e}^{i\ell})^2 (\mathfrak{m}\_1 - \mathfrak{m}\_2)\mathbf{Y}\_2^2\right)}. \tag{26}$$
Applying Lemma (1) to the coefficients ℘<sup>2</sup> and q2, we have the following.
$$|a\_2| \le \frac{|\vartheta| \mathbf{m}\_1 \sqrt{\mathbf{m}\_1}}{\sqrt{\left|\vartheta[(\mathfrak{F} + \mathfrak{B}\dot{\epsilon}^{i\ell})\mathbf{Y}\_3 - (2 + \dot{\epsilon}^{i\ell})\mathbf{Y}\_2^2\right| \mathbf{m}\_1^2 + (2 + \dot{\epsilon}^{i\ell})^2 (\mathbf{m}\_1 - \mathbf{m}\_2)\mathbf{Y}\_2^2}}.$$
Next, in order to find the bound on |*a*3|, by subtracting (21) from (23) and using (24), we obtain the following.
$$\frac{4}{\theta} \frac{(\mathbf{5} + 3e^{i\ell})}{2} \mathbf{Y}\_3 (a\_3 - a\_2^2) = \frac{\mathbf{m}\_1}{2} (\wp\_2 - \mathfrak{q}\_2)$$
$$a\_3 = a\_2^2 + \frac{\theta \mathbf{m}\_1 (\wp\_2 - \mathfrak{q}\_2)}{4(\mathbf{5} + 3e^{i\ell}) \mathbf{Y}\_3}. \tag{27}$$
Substituting the value of *a* 2 2 given by (25), we obtain the following.
$$a\_3 = \frac{\theta^2 \mathbf{m}\_1^2 (\wp\_1^2 + \mathfrak{q}\_1^2)}{8(2 + e^{i\ell})^2 \mathbf{Y}\_2^2} + \frac{\theta \mathbf{m}\_1 (\wp\_2 - \mathfrak{q}\_2)}{4(5 + 3e^{i\ell}) \mathbf{Y}\_3}.$$
Applying Lemma 1 once again to the coefficients ℘1, ℘2, q<sup>1</sup> and q2, we obtain the following.
$$|a\_3| \le \frac{|\vartheta|^2 \mathbf{m}\_1^2}{|2 + e^{i\ell}|^2 \mathbf{Y}\_2^2} + \frac{|\vartheta| \mathbf{m}\_1}{|\mathbf{5} + \mathbf{3}e^{i\ell}| \mathbf{Y}\_3}.$$
**Theorem 2.** *Let f given by (1) be in the following class:* K *ς*,*τ* <sup>Σ</sup>,W(*ϑ*, `)*, ϑ* ∈ C\{0} *and* ` ∈ (−*π*, *π*]*. Then, we have the following:*
$$|a\_2| \le \frac{|\vartheta| \mathbf{m}\_1 \sqrt{\mathbf{m}\_1}}{\sqrt{|\vartheta| \mathbf{3} (\mathbf{5} + 3\varepsilon^{i\ell}) \mathbf{Y}\_3 - 4 (\mathbf{2} + \varepsilon^{i\ell}) \mathbf{Y}\_2^2| \mathbf{m}\_1^2 + 4 (\mathbf{2} + \varepsilon^{i\ell})^2 (\mathbf{m}\_1 - \mathbf{m}\_2) \mathbf{Y}\_2^2|}}\tag{28}$$
*and*
$$|a\_3| \le \frac{|\vartheta|^2 \mathbf{m}\_1^2}{4|2 + e^{i\ell}|^2 \mathbf{Y}\_2^2} + \frac{|\vartheta| \mathbf{m}\_1}{3|5 + 3e^{i\ell}| \mathbf{Y}\_3}.\tag{29}$$
**Proof.** By Definition 2,the argument inequalities in (10) and (11) can be equivalently written as follows:
$$1 + \frac{1}{\theta} \left( \frac{[\xi(\mathfrak{I}\_{\mathbf{x}}^{\xi, \mathsf{T}} f(\xi))' + \left(\frac{1 + \epsilon^{j\ell}}{2}\right) \xi^2 (\mathfrak{I}\_{\mathbf{x}}^{\xi, \mathsf{T}} f(\xi))'']'}{(\mathfrak{I}\_{\mathbf{x}}^{\xi, \mathsf{T}} f(\xi))'} - 1 \right) = \mathfrak{W}(u(\xi)) \tag{30}$$
and
$$1 + \frac{1}{\theta} \left( \frac{[w(\mathfrak{I}\_{\mathbf{x}}^{\xi, \pi} \mathfrak{g}(w))' + \left(\frac{1 + \epsilon^{i\ell}}{2}\right) w^2 (\mathfrak{I}\_{\mathbf{x}}^{\xi, \pi} \mathfrak{g}(w))'']'}{(\mathfrak{I}\_{\mathbf{x}}^{\xi, \pi} \mathfrak{g}(w))'} - 1\right) = \mathfrak{W}(v(w)),\tag{31}$$
and proceeding as in the proof of Theorem 1, we can arrive at the following relations:
$$\begin{split} 1 + \frac{1}{\theta} \left( \frac{[\xi(\mathfrak{I}\_{\mathbb{K}}^{\xi,\mathbb{T}}f(\xi))' + \left(\frac{1+e^{i\ell}}{2}\right)\xi^{2}(\mathfrak{I}\_{\mathbb{K}}^{\xi,\mathbb{T}}f(\xi))'']'}{(\mathfrak{I}\_{\mathbb{K}}^{\xi,\mathbb{T}}f(\xi))'} - 1 \right) &= 1 + \frac{2}{\theta}(2+e^{i\ell})\mathsf{Y}\_{2}a\_{2}\xi \\ + \frac{1}{\theta}[3(5+3e^{i\ell})\mathsf{Y}\_{3}a\_{3} - 4(2+e^{i\ell})\mathsf{Y}\_{2}^{2}a\_{2}^{2}]\xi^{2} + \cdots \end{split}$$
and
$$\begin{split} 1 + \frac{1}{\theta} \left( \frac{[w(\mathfrak{I}\_{\mathbf{x}}^{\xi,\tau}g(w))' + \left(\frac{1+e^{i\ell}}{2}\right)w^2(\mathfrak{I}\_{\mathbf{x}}^{\xi,\tau}g(w))']'}{(\mathfrak{I}\_{\mathbf{x}}^{\xi,\tau}g(w))'} - 1 \right) &= 1 - \frac{2}{\theta}(2+e^{i\ell})\mathsf{Y}\_2a\_2w \\ &+ \frac{1}{\theta}[3(5+3e^{i\ell})(2a\_2^2 - a\_3)\mathsf{Y}\_3 - 4(2+e^{i\ell})\mathsf{Y}\_2^2a\_2^2]w^2 + \cdots \ . \end{split}$$
From (30) and (31), equating the coefficients of *ς* and *ς* 2 , we obtain the following:
$$\frac{2}{\theta}(2+e^{i\ell})\mathbf{Y}\_2a\_2 = \frac{1}{2}\mathbf{m}\_1\wp\_{1\prime} \tag{32}$$
$$\frac{1}{2}[3(5+3e^{i\ell})\mathbf{Y}\_3a\_3 - 4(2+e^{i\ell})\mathbf{Y}\_2^2a\_2^2] = \frac{1}{2}\mathbf{m}\_1(\wp\_2 - \frac{\wp\_1^2}{2}) + \frac{1}{4}\mathbf{m}\_2\wp\_1^2\tag{33}$$
and
$$-\frac{2}{\theta}(2+e^{i\ell})\mathbf{Y}\_2a\_2 = \frac{1}{2}\mathbf{m}\_1\mathbf{q}\_{1\prime}\tag{34}$$
$$\frac{1}{\theta}[3(5+3e^{i\ell})(2a\_2^2 - a\_3)\mathbf{Y}\_3 - 4(2+e^{i\ell})\mathbf{Y}\_2^2 a\_2^2] = \frac{1}{2}\mathbf{m}\_1(\mathbf{q}\_2 - \frac{\mathbf{q}\_1^2}{2}) + \frac{1}{4}\mathbf{m}\_2\mathbf{q}\_1^2.\tag{35}$$
From (32) and (34), we obtain the following:
$$
\wp\_1 = -\mathfrak{q}\_1 \tag{36}
$$
and
$$32(2+e^{i\ell})^2 \mathbf{Y}\_2^2 a\_2^2 = \theta^2 \mathbf{m}\_1^2 (\wp\_1^2 + \mathfrak{q}\_1^2). \tag{37}$$
If we add (33) and (35) and substitute value ℘ 2 <sup>1</sup> + q 2 1 , we obtain the following.
$$a\_2^2 = \frac{\theta^2 \mathbf{m}\_1^3 (\wp\_2 + \mathfrak{q}\_2)}{4[\theta[3(5+3e^{i\ell})\varGamma\_3 - 4(2+e^{i\ell})\varGamma\_2^2] \mathbf{m}\_1^2 + 4(2+e^{i\ell})^2 (\mathbf{m}\_1 - \mathbf{m}\_2)\varGamma\_2^2]}.\tag{38}$$
Applying Lemma 1 to the coefficients ℘<sup>2</sup> and q2, we have the desired inequality given in (28).
Next, if we subtract (33) from (35), we easily observe the following.
$$\begin{aligned} \frac{12}{\vartheta} \frac{(\mathfrak{F} + 3e^{i\ell})}{2} (a\_3 - a\_2^2) \mathbf{Y}\_3 &=& \frac{\mathbf{m}\_1}{2} (\wp\_2 - \mathfrak{q}\_2) \\ a\_3 &=& \frac{\vartheta \mathbf{m}\_1 (\wp\_2 - \mathfrak{q}\_2)}{12 (\mathfrak{F} + 3e^{i\ell}) \mathbf{Y}\_3} + a\_2^2 \end{aligned}$$
Upon relieving the value of *a* 2 2 given in (37), the above equation leads to the following.
$$a\_3 = \frac{\theta \mathbf{m}\_1 (\wp\_2 - \mathfrak{q}\_2)}{12 (\mathbf{5} + 3e^{i\ell}) \mathbf{Y}\_3} + \frac{\theta^2 \mathbf{m}\_1^2 (\wp\_1^2 + \mathfrak{q}\_1^2)}{32 (\mathbf{2} + e^{i\ell})^2 \mathbf{Y}\_2^2}.$$
Applying Lemma (1) once again to the coefficients ℘1, ℘2, q1, and q2, we obtain the preferred coefficient provided in (29).
Fixing ` = *π* in Theorems (1) and (2), we can state the coefficient estimates for the functions in subclasses S *ς*,*τ* <sup>Σ</sup>,W(*ϑ*) and K *ς*,*τ* <sup>Σ</sup>,W(*ϑ*), defined in Remark (2).
**Corollary 1.** *Let f assumed as (1) be in the class* S *ς*,*τ* <sup>Σ</sup>,W(*ϑ*). *Then, the following is the case.*
$$|a\_2| \le \frac{|\boldsymbol{\theta}|\mathbf{m}\_1\sqrt{\mathbf{m}\_1}}{\sqrt{|\boldsymbol{\theta}|(2\mathbf{Y}\_3 - \mathbf{Y}\_2^2)\mathbf{m}\_1^2 + (\mathbf{m}\_1 - \mathbf{m}\_2)\mathbf{Y}\_2^2}} \quad \text{and} \quad |a\_3| \le \frac{|\boldsymbol{\theta}|^2\mathbf{m}\_1^2}{\mathbf{Y}\_2^2} + \frac{|\boldsymbol{\theta}|\mathbf{m}\_1}{2\mathbf{Y}\_3}.$$
**Corollary 2.** *Let f assumed as (1) be in class* K *ς*,*τ* <sup>Σ</sup>,W(*ϑ*). *Then, we have the following.*
$$|a\_2| \le \frac{|\boldsymbol{\theta}| \mathbf{m}\_1 \sqrt{\mathbf{m}\_1}}{\sqrt{2|\boldsymbol{\theta}|(3\mathbf{Y}\_3 - 2\mathbf{Y}\_2^2)\mathbf{m}\_1^2 + 4(\mathbf{m}\_1 - \mathbf{m}\_2)\mathbf{Y}\_2^2}} \quad \text{and} \quad |a\_3| \le \frac{|\boldsymbol{\theta}|^2 \mathbf{m}\_1^2}{4\mathbf{Y}\_2^2} + \frac{|\boldsymbol{\theta}| \mathbf{m}\_1}{6\mathbf{Y}\_3}.$$
Fixing *ϑ* = 1 in Theorems (1) and (2), we can state the coefficient estimates for the functions in the subclasses S *ς*,*τ* <sup>Σ</sup>,W(`) and K *ς*,*τ* <sup>Σ</sup>,W(`) defined in Remark (4).
**Corollary 3.** *Let f supposed by (1) be in class* S *ς*,*τ* <sup>Σ</sup>,W(`). *Then, we have the following:*
$$|a\_2| \le \frac{\mathbf{m}\_1 \sqrt{\mathbf{m}\_1}}{\sqrt{|[(\mathbf{5} + 3e^{i\ell})\mathbf{Y}\_3 - (\mathbf{2} + e^{i\ell})\mathbf{Y}\_2^2] \mathbf{m}\_1^2 + (2 + e^{i\ell})^2 (\mathbf{m}\_1 - \mathbf{m}\_2)\mathbf{Y}\_2^2|}}$$
*and the following is the case.*
$$|a\_3| \le \frac{\mathbf{m}\_1^2}{|2 + e^{i\ell}|^2 \mathbf{Y}\_2^2} + \frac{\mathbf{m}\_1}{|\mathbf{5} + \mathbf{3}e^{i\ell}| \mathbf{Y}\_3}.$$
**Corollary 4.** *Let f supposed by (1) be in class* K *ς*,*τ* <sup>Σ</sup>,W(`). *Then, we have the following:*
$$|a\_2| \le \frac{\mathbf{m}\_1 \sqrt{\mathbf{m}\_1}}{\sqrt{| [3(5 + 3e^{i\ell}) \mathbf{Y}\_3 - 4(2 + e^{i\ell}) \mathbf{Y}\_2^2] \mathbf{m}\_1^2 + 4(2 + e^{i\ell})^2 (\mathbf{m}\_1 - \mathbf{m}\_2) \mathbf{Y}\_2^2|}}$$
*and*
$$|a\_3| \le \frac{\mathbf{m}\_1^2}{4|2 + e^{i\ell}|^2 \mathbf{Y}\_2^2} + \frac{\mathbf{m}\_1}{3|5 + 3e^{i\ell}| \mathbf{Y}\_3}.$$
#### **3. Fekete-Szeg˝o Inequality**
In this section, we discuss the Fekete-Szeg˝o results [34] due toZaprawa [35] for functions *f* ∈ S *ς*,*τ* <sup>Σ</sup>,W(*ϑ*, `) and *f* ∈ K *ς*,*τ* <sup>Σ</sup>,W(*ϑ*, `).
**Theorem 3.** *Let f assumed by (1) be in class* S *ς*,*τ* <sup>Σ</sup>,W(*ϑ*, `) *and \$* ∈ R. *Then, we have the following:*
$$|\,|\,a\_3-\emptyset a\_2^2\vert\leq\begin{cases}\frac{\theta \mathbf{m}\_1}{|\mathbf{5}+3\epsilon^{i\ell}|\mathbf{Y}\_3|}\;\prime & 0\leq|\,\phi(\varrho)\mid\leq\frac{\theta \mathbf{m}\_1}{4|\mathbf{5}+3\epsilon^{i\ell}|\mathbf{Y}\_3|}\\4|\phi(\varrho)\vert\prime & |\phi(\varrho)\vert\geq\frac{\theta \mathbf{m}\_1}{4|\mathbf{5}+3\epsilon^{i\ell}|\mathbf{Y}\_3|}.\end{cases}$$
*where the following is obtained.*
$$\phi(\varrho) = \frac{(1-\varrho)\vartheta^2 \mathbf{m}\_1^3}{4\{\vartheta[(5+3e^{i\ell})\mathbf{Y}\_3 - (2+e^{i\ell})\mathbf{Y}\_2^2] \mathbf{m}\_1^2 + (2+e^{i\ell})^2 (\mathbf{m}\_1 - \mathbf{m}\_2)\mathbf{Y}\_2^2\}}.$$
**Proof.** From (26) and (27), we have the following:
$$\begin{split} \rho\_3 - \rho \mathbf{z}\_2^2 &= \frac{(1-\varrho)\theta^2 \mathbf{m}\_1^3 (\wp\_2 + \mathfrak{q}\_2)}{\left(4\{\boldsymbol{\theta}[(\mathfrak{F} + 3\boldsymbol{\epsilon}^{i\ell})\mathbf{Y}\_3 - (2 + \boldsymbol{\epsilon}^{i\ell})\mathbf{Y}\_2^2] \mathbf{m}\_1^2 + (2 + \boldsymbol{\epsilon}^{i\ell})^2 (\mathbf{m}\_1 - \mathbf{m}\_2) \mathbf{Y}\_2^2\right)} + \frac{\theta \mathbf{m}\_1 (\wp\_2 - \mathfrak{q}\_2)}{4(\mathfrak{F} + 3\boldsymbol{\epsilon}^{i\ell}) \mathbf{Y}\_3} \\ &= \quad \left[\boldsymbol{\phi}(\boldsymbol{\varrho}) + \frac{\theta \mathbf{m}\_1}{4(\mathfrak{F} + 3\boldsymbol{\epsilon}^{i\ell}) \mathbf{Y}\_3}\right] \boldsymbol{\wp}\_2 + \left[\boldsymbol{\phi}(\boldsymbol{\varrho}) - \frac{\theta \mathbf{m}\_1}{4(\mathfrak{F} + 3\boldsymbol{\epsilon}^{i\ell}) \mathbf{Y}\_3}\right] \boldsymbol{\mathfrak{q}}\_2 \end{split}$$
where the following is the case.
$$\phi(\varrho) = \frac{(1-\varrho)\vartheta^2 \mathbf{m}\_1^3}{4\{\vartheta[(5+3e^{i\ell})\mathbf{Y}\_3 - (2+e^{i\ell})\mathbf{Y}\_2^2]\mathbf{m}\_1^2 + (2+e^{i\ell})^2(\mathbf{m}\_1 - \mathbf{m}\_2)\mathbf{Y}\_2^2\}}$$
Thus, by applying Lemma 1, we obtain the following.
$$|\,|\,a\_3 - \varrho a\_2^2| \le \begin{cases} \frac{\theta \mathbf{m}\_1}{|\mathbf{5} + 3\mathbf{c}^{i\ell}| \mathbf{Y}\_3|} & 0 \le |\,\phi(\varrho)| \le \frac{\theta \mathbf{m}\_1}{4|\mathbf{5} + 3\mathbf{c}^{i\ell}| \mathbf{Y}\_3|}\\ 4|\phi(\varrho)| \, & |\phi(\varrho)| \ge \frac{\theta \mathbf{m}\_1}{4|\mathbf{5} + 3\mathbf{c}^{i\ell}| \mathbf{Y}\_3|}. \end{cases}$$
In particular, by fixing *\$* = 1, we obtain the following.
$$|\,a\_3 - a\_2^2 \mid \le \frac{\theta \mathbf{m}\_1}{|\mathbf{5} + \mathbf{3}e^{i\ell}| \mathbf{Y}\_3|}$$
.
**Theorem 4.** *Let f given by (1) be in class* K *ς*,*τ* <sup>Σ</sup>,W(*ϑ*, `) *and* ℵ ∈ R*. Then, we have the following:*
$$|\,|\,a\_3 - \aleph a\_2^2| \le \begin{cases} \frac{\theta \mathbf{m}\_1}{3|5 + 3\epsilon^{i\ell}|\mathbf{Y}\_3|} & 0 \le |\,\phi(\aleph)| \le \frac{\theta \mathbf{m}\_1}{12|5 + 3\epsilon^{i\ell}|\mathbf{Y}\_3|}\\ 4|\phi(\aleph)| , & |\phi(\aleph)| \ge \frac{\theta \mathbf{m}\_1}{12|5 + 3\epsilon^{i\ell}|\mathbf{Y}\_3|}. \end{cases}$$
*where*
$$\phi(\aleph) = \frac{(1-\aleph)\theta^2 \mathbf{m}\_1^3}{4[\theta[3(5+3e^{i\ell})\aleph\_3 - 4(2+e^{i\ell})\varUpsilon\_2^2]\mathbf{m}\_1^2 + 4(2+e^{i\ell})^2(\mathbf{m}\_1 - \mathbf{m}\_2)\varUpsilon\_2^2]}.$$
**Proof.** From (27) and (38), we have the following.
$$\begin{split} \mathfrak{a}\_{3} - \aleph\_{2}^{2} &= \frac{(1-\aleph)\mathfrak{a}^{2}\mathbf{m}\_{1}^{3}(\wp\_{2}+\mathfrak{q}\_{2})}{4[\varPhi]3(5+3\epsilon^{\acute{e}\ell})\mathbf{Y}\_{3} - 4(2+\epsilon^{\acute{e}\ell})\mathbf{Y}\_{2}^{2}]\mathbf{m}\_{1}^{2} + 4(2+\epsilon^{\acute{e}\ell})^{2}(\mathbf{m}\_{1}-\mathbf{m}\_{2})\mathbf{Y}\_{2}^{2}} + \frac{\theta\mathbf{m}\_{1}(\wp\_{2}-\mathfrak{q}\_{2})}{12(5+3\epsilon^{\acute{e}\ell})\mathbf{Y}\_{3}} \\ &= \quad \left[\varPhi(\aleph) + \frac{\theta\mathbf{m}\_{1}}{12(5+3\epsilon^{\acute{e}\ell})\mathbf{Y}\_{3}}\right] \wp\_{2} + \left[\varPhi(\aleph) - \frac{\theta\mathbf{m}\_{1}}{12(5+3\epsilon^{\acute{e}\ell})\mathbf{Y}\_{3}}\right] \mathfrak{q}\_{2} \end{split}$$
where the following is the case.
$$\phi(\aleph) = \frac{(1-\aleph)\theta^2 \mathbf{m}\_1^3}{4[\theta[\clubsuit(\sf S + 3e^{i\ell})\varUpsilon\_3 - 4(2+e^{i\ell})\varUpsilon\_2^2] \mathbf{m}\_1^2 + 4(2+e^{i\ell})^2(\mathbf{m}\_1 - \mathbf{m}\_2)\varUpsilon\_2^2]}.$$
Thus, by Lemma 1, we obtain the following.
$$|\,|\,a\_3 - \aleph a\_2^2| \le \begin{cases} \frac{\theta \mathbf{m}\_1}{3|\mathbf{5} + 3\mathbf{e}^{i\ell}| \mathbf{Y}\_3|} & 0 \le |\,\phi(\aleph)| \le \frac{\theta \mathbf{m}\_1}{12|\mathbf{5} + 3\mathbf{e}^{i\ell}| \mathbf{Y}\_3|}\\4|\phi(\aleph)| & |\phi(\aleph)| \ge \frac{\theta \mathbf{m}\_1}{12|\mathbf{5} + 3\mathbf{e}^{i\ell}| \mathbf{Y}\_3|}. \end{cases}$$
In particular, by taking ℵ = 1, we obtain the following.
$$|\,a\_3 - a\_2^2 \mid \le \frac{\theta \mathbf{m}\_1}{3|\mathbf{5} + 3e^{i\ell}|\mathbf{Y}\_3|}.$$
#### **4. Conclusions**
By fixing W(*ξ*) as listed below, one can determine new results as in Theorems 1–4 for the subclasses introduced in this paper by suitably fixing **m**<sup>1</sup> and **m**2:
In the current paper, we mainly obtain the upper bounds of the initial Taylors coefficients of bi-starlike and bi-convex functions of complex order involving Erdély–Kober-type integral operators in the open unit. Furthermore, we find the Fekete-Szeg˝o inequalities for the function in these classes. Several consequences of the results are also pointed out as examples. Moreover, we note that by assuming W with some particular functions as illustrated above, one can determine new results for the subclasses introduced in this paper. Moreover, by fixing ` = 0 and ` = *π* in the above Theorems, we can easily state the results for various subclasses of Σ illustrated in Remarks 2–4. By appropriately fixing the parameters in Theorems 3 and 4, we can deduce the Fekete-Szeg˝o functional for these function classes. Moreover, motivating further research on the subject-matter of this, we have chosen to draw the attention of the concerned readers toward a significantly large number of interrelated publications(see [49–52]) and developments in the area of Geometric Function Theory of Complex Analysis. In conclusion, we choose to reiterate an important observation, which was offered in the recently published survey-cum-expository article by Srivastava ([49], p. 340), who pointed out the fact that the results for the above-mentioned or new *q*− analogues can easily (and possibly or unimportantly) be interpreted into the equivalent results for the so-called (*p*; *q*)− analogues (with 0 < |*q*| < *p* ≤ 1) by smearing some recognizable parametric and argument variations with the additional parameter *p* being redundant.
**Author Contributions:** Conceptualization, A.A., G.M. and S.M.E.-D.; methodology, A.A., G.M. and S.M.E.-D.; validation, A.A., G.M. and S.M.E.-D.; formal analysis, A.A., G.M. and S.M.E.-D.; investigation, A.A., G.M. and S.M.E.-D.; resources, A.A., G.M. and S.M.E.-D.; writing—original draft preparation, A.A., G.M. and S.M.E.-D.; writing—review and editing, A.A., G.M. and S.M.E.-D.; supervision, A.A., G.M. and S.M.E.-D.; project administration, A.A., G.M. and S.M.E.-D. All authors have read and agreed to the published version of the manuscript.
**Funding:** This research received no external funding.
**Institutional Review Board Statement:** Not applicable.
**Informed Consent Statement:** Not applicable.
**Data Availability Statement:** Not applicable.
**Acknowledgments:** The researchers would like to thank the Deanship of Scientific Research, Qassim University, for funding the publication of this project. The authors are grateful to the referees of this article who provided valuable comments and advice that allowed us to revise and improve the content of the paper.
**Conflicts of Interest:** The authors declare no conflict of interest.
#### **References**
- 26. Livingston, A.E. On the radius of univalence of certain analytic functions. *Proc. Am. Math. Soc.* **1966**, *17*, 352–357. [CrossRef]
### *Article* **An Avant-Garde Construction for Subclasses of Analytic Bi-Univalent Functions**
**Feras Yousef 1,\* , Ala Amourah <sup>2</sup> , Basem Aref Frasin <sup>3</sup> and Teodor Bulboac˘a <sup>4</sup>**
**Abstract:** The zero-truncated Poisson distribution is an important and appropriate model for many real-world applications. Here, we exploit the zero-truncated Poisson distribution probabilities to construct a new subclass of analytic bi-univalent functions involving Gegenbauer polynomials. For functions in the constructed class, we explore estimates of Taylor–Maclaurin coefficients |*a*2| and |*a*3|, and next, we solve the Fekete–Szeg˝o functional problem. A number of new interesting results are presented to follow upon specializing the parameters involved in our main results.
**Keywords:** analytic bi-univalent functions; zero-truncated Poisson distribution; Gegenbauer polynomials; Fekete–Szeg˝o functional problem
**MSC:** 30C45; 33C45; 60E05
#### **1. Introduction**
In discrete probability distributions, the Poisson distribution has found an extensive and varied application in formulating probability models for a wide variety of real-life phenomena dealing with counts of rare events, such as reliability theory, queueing systems, epidemiology, medicine, industry, and many others. In some practical situations, only positive counts would be available and the zero count is ignored or is impossible to be observed at all. For instance: the length of stay in a hospital is recorded as a minimum of at least one day, the number of journal articles published in different disciplines, the number of occupants in passenger cars, etc. An appropriate Poisson distribution that applies to such a case is called a zero-truncated Poisson distribution.
The probability density function of a discrete random variable *X* that follows a zerotruncated Poisson distribution can be written as
$$P\_m(X=s) = \frac{m^s}{(e^m - 1)s!}, \text{ s } = 1, 2, 3, \dots, s$$
where the parameter mean *m* > 0.
Now, we introduce a novel power series whose coefficients are probabilities of the zero-truncated Poisson distribution
$$\mathbb{P}(m, z) := z + \sum\_{n=2}^{\infty} \frac{m^{n-1}}{(e^m - 1)(n - 1)!} z^n, \; z \in \mathbb{U}\_{\prime}$$
where *m* > 0 and U := {*z* ∈ C : |*z*| < 1} is the *open unit disk*. By ratio test, it is clear that the radius of convergence of the above series is infinity.
**Citation:** Yousef, F.; Amourah, A.; Frasin, B.A.; Bulboac˘a, T. An Avant-Garde Construction for Subclasses of Analytic Bi-Univalent Functions. *Axioms* **2022**, *11*, 267. https://doi.org/10.3390/ axioms11060267
Academic Editor: Georgia Irina Oros
Received: 15 May 2022 Accepted: 30 May 2022 Published: 1 June 2022
**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.
**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).
Orthogonal polynomials have been extensively studied in recent years from various perspectives due to their importance in mathematical statistics, probability theory, mathematical physics, approximation theory, and engineering. From a mathematical point of view, orthogonal polynomials often arise from solutions of ordinary differential equations under certain conditions imposed by certain model. Orthogonal polynomials that appear most commonly in applications are the classical orthogonal polynomials (Hermite polynomials, Laguerre polynomials, and Jacobi polynomials). The general subclass of Jacobi polynomials is the set of Gegenbauer polynomials, this class includes Legendre polynomials and Chebyshev polynomials as subclasses. To study the basic definitions and the most important properties of the classical orthogonal polynomials, we refer the reader to [1–4]. For a recent connection between the classical orthogonal polynomials and geometric function theory, we mention [5–10].
Gegenbauer polynomials *C α n* (*x*) for *<sup>n</sup>* <sup>=</sup> 2, 3, . . . , and *<sup>α</sup>* <sup>&</sup>gt; <sup>−</sup><sup>1</sup> 2 are defined by the following three-term recurrence formula
$$\begin{aligned} \mathbf{C}\_{0}^{\mathfrak{a}}(\mathbf{x}) &= 1; \\ \mathbf{C}\_{1}^{\mathfrak{a}}(\mathbf{x}) &= 2\alpha \mathbf{x}; \\ \mathbf{C}\_{n}^{\mathfrak{a}}(\mathbf{x}) &= \frac{1}{n} \Big[ 2\mathbf{x}(n+\mathfrak{a}-1)\mathbf{C}\_{n-1}^{\mathfrak{a}}(\mathbf{x}) - (n+2\mathfrak{a}-2)\mathbf{C}\_{n-2}^{\mathfrak{a}}(\mathbf{x}) \Big]. \end{aligned} \tag{1}$$
It is worth mentioning that by setting *α* = <sup>1</sup> 2 and *α* = 1 in Equation (1), we immediately obtain Legendre polynomials *Pn*(*x*) = *C* 1 2 *<sup>n</sup>* (*x*) and Chebyshev polynomials of the second kind *Un*(*x*) = *C* 1 *n* (*x*), respectively.
The generating function of Gegenbauer polynomials is given as
$$H\_{\mathfrak{A}}(\mathfrak{x}, \mathfrak{z}) = \frac{1}{(1 - 2\mathfrak{x}\mathfrak{z} + \mathfrak{z}^2)^{\mathfrak{A}'}}$$
where *x* ∈ [−1, 1] and *z* ∈ U. For fixed *x*, the function *H<sup>α</sup>* is analytic in U, so it can be expanded in a Taylor–Maclaurin series, as follows:
$$H\_{\mathfrak{a}}(\mathfrak{x}, \mathfrak{z}) = \sum\_{n=0}^{\infty} \mathbb{C}\_{\mathfrak{n}}^{\mathfrak{a}}(\mathfrak{x}) z^{n}, \; z \in \mathbb{U}. \tag{2}$$
#### **2. Preliminaries and Definitions**
Let A denote the class of all normalized analytic functions *f* written as
$$f(z) = z + \sum\_{n=2}^{\infty} a\_n z^n \text{ , } z \in \mathbb{U}. \tag{3}$$
Differential subordination of analytic functions provides excellent tools for study in geometric function theory. The earliest problem in differential subordination was introduced by Miller and Mocanu [11], see also [12]. The book of Miller and Mocanu [13] sums up most of the advancement in the field and the references to the date of its publication.
**Definition 1.** *Let f and g be two analytic functions in* U*. The function f is said to be subordinate to g, written as f*(*z*) ≺ *g*(*z*)*, if there is an analytic function ω in* U *with the properties*
$$|\omega(0) = 0 \quad \text{and} \quad |\omega(z)| < 1, \ z \in \mathbb{U}\_2$$
*such that*
$$f(z) = \operatorname{g}(\omega(z)), \ z \in \mathbb{U}.$$
**Definition 2.** *A single-valued one-to-one function f defined in a simply connected domain is said to be a univalent function.*
Let S denote the class of all functions *f* ∈ A, given by (3), that are univalent in U. Hence, every function *f* ∈ S has an inverse given by
$$f^{-1}(w) = w - a\_2 w^2 + \left(2a\_2^2 - a\_3\right) w^3 - \left(5a\_2^3 - 5a\_2 a\_3 + a\_4\right) w^4 + \dots \tag{4}$$
**Definition 3.** *A univalent function f is said to be bi-univalent in* U *if its inverse function f* −1 (*w*) *has an analytic univalent extension in* U*.*
Let Σ denote the class of all functions *f* ∈ A that are bi-univalent in U given by (3). For interesting subclasses of functions in the class Σ, see [14–24].
The coefficient functional
$$
\Delta\_{\eta}(f) = a\_3 - \eta a\_2^2 = \frac{1}{6} \left( f''''(0) - \frac{3\eta}{2} \left( f''(0) \right)^2 \right) \tag{5}
$$
of the analytic function *f* given by (3) is very important in the theory of analytic and univalent functions. Thus, it is quite natural to ask about inequalities for ∆*η*(*f*) corresponding to subclasses of bi-univalent functions in the open unit disk U. The problem of maximizing the absolute value of the functional ∆*η*(*f*) is called the Fekete–Szegö problem [25]. There are now several results of this type in the literature, each of them dealing with |*a*<sup>3</sup> − *ηa* 2 2 | for various classes of functions defined in terms of subordination (see, e.g., [26–31]).
Now, let us define the linear operator
$$\chi: \mathcal{A} \to \mathcal{A}$$
by
$$\chi\_{\mathfrak{m}}f(z) := \mathbb{P}(m, z) \* f(z) = z + \sum\_{n=2}^{\infty} \frac{m^{n-1}}{(e^m - 1)(n-1)!} a\_n z^n, \; z \in \mathbb{U}\_+$$
where the symbol "∗" denotes the Hadamard product of the two series.
To obtain our results we need the following lemma:
**Lemma 1** ([32], p. 172)**.** *Assume that ω*(*z*) = ∞ ∑ *n*=1 *ωnz n , z* ∈ U*, is an analytic function in* U *such that* |*ω*(*z*)| < 1 *for all z* ∈ U*. Then,* |*ω*1| ≤ 1, |*ωn*| ≤ 1 − |*ω*1| 2 , *n* = 2, 3, . . . *.*
Motivated essentially by the earlier work of Amourah et al. [33], we construct, in the next section, a new subclass of bi-univalent functions governed by the zero-truncated Poisson distribution series and Gegenbauer polynomials. Then, we investigate the optimal bounds for the Taylor–Maclaurin coefficients |*a*2| and |*a*3| and solve the Fekete–Szeg˝o functional problem for functions in our new subclass.
#### **3. The Class** *ζ***Σ**(*x***,** *α***,** *δ***,** *µ*)
Consider the function *f* ∈ Σ given by (3), the function *g* = *f* <sup>−</sup><sup>1</sup> given by (4), and *H<sup>α</sup>* is the generating function of Gegenbauer polynomials given by (2). Now, we are ready to define our new subclass of bi-univalent functions *ζ*Σ(*x*, *α*, *δ*, *µ*) as follows.
**Definition 4.** *A function f is said to be in the class ζ*Σ(*x*, *α*, *δ*, *µ*)*, if the following subordinations are fulfilled:*
$$(1 - \mu)\frac{\chi\_{\mathfrak{m}}f(z)}{z} + \mu(\chi\_{\mathfrak{m}}f(z))' + \delta z(\chi\_{\mathfrak{m}}f(z))'' \prec H\_{\mathfrak{a}}(\mathfrak{x}, z)\_{\mathsf{m}}$$
*and*
$$(1 - \mu) \frac{\chi\_m \text{g}(w)}{w} + \mu (\chi\_m \text{g}(w))' + \delta w (\chi\_m \text{g}(w))'' \prec H\_a(\chi, w)\mu$$
*where α* > 0*, µ*, *δ* ≥ 0*, and x* ∈ 1 2 , 1i
*.*
Upon allocating the parameters *µ* and *δ*, one can obtain several new subclasses of Σ, as illustrated in the following two examples.
**Example 1.** *A function f is said to be in the class ζ*Σ(*x*, *α*, *µ*) := *ζ*Σ(*x*, *α*, 0, *µ*)*, if the following subordinations are fulfilled:*
$$(1 - \mu) \frac{\chi\_m f(z)}{z} + \mu (\chi\_m f(z))' \prec H\_a(\chi, z),$$
*and*
$$(1 - \mu) \frac{\chi\_m g(w)}{w} + \mu (\chi\_m g(w))' \prec H\_a(\chi, w).$$
*where α* > 0*, µ* ≥ 0*, and x* ∈ 1 2 , 1i *.*
**Example 2.** *A function f is said to be in the class ζ*Σ(*x*, *α*) := *ζ*Σ(*x*, *α*, 0, 1)*, if the following subordinations are fulfilled:*
$$(\chi\_m f(z))' \prec H\_a(\mathfrak{x}, z)\_\star$$
*and*
$$(\chi\_m g(w))' \prec H\_\mathfrak{a}(\mathfrak{x}, w),$$
*where α* > 0 *and x* ∈ 1 2 , 1i *.*
#### **4. Main Results**
**Theorem 1.** *If the function f belongs to the class ζ*Σ(*x*, *α*, *δ*, *µ*)*, then*
$$|a\_2| \le \frac{2ax(e^m - 1)\sqrt{2x}}{m\sqrt{\left|\left[2a(1+2\mu+6\delta)(e^m - 1) - 2(1+a)(1+\mu+2\delta)^2\right]x^2 + (1+\mu+2\delta)^2\right|}},\tag{6}$$
*and*
$$|a\_3| \le \frac{4a^2\varkappa^2(e^m - 1)^2}{m^2(1+\mu+2\delta)^2} + \frac{4a\varkappa(e^m - 1)}{m^2(1+2\mu+6\delta)}$$
**Proof.** If *f* ∈ *ζ*Σ(*x*, *α*, *δ*, *µ*), from the Definition 4 there exist two analytic functions in U that are *w* and *v*, such that *w*(0) = *v*(0) = 0 and |*ω*(*z*)| < 1, |*v*(*w*)| < 1 for all *z*, *w* ∈ U, and
$$\left( (1 - \mu) \frac{\chi\_{\mathfrak{m}} f(z)}{z} + \mu (\chi\_{\mathfrak{m}} f(z))' + \delta z (\chi\_{\mathfrak{m}} f(z))'' \right)' = H\_{\mathfrak{a}}(\mathfrak{x}, \omega(z)), \; z \in \mathbb{U}, \tag{7}$$
.
and
$$(1 - \mu) \frac{\chi\_{\text{mg}}(w)}{w} + \mu (\chi\_{\text{mg}}(w))' + \delta w (\chi\_{\text{mg}}g(w))'' = H\_{\mathfrak{a}}(\mathfrak{x}, v(w)), \ w \in \mathbb{U},\tag{8}$$
From the equalities (7) and (8), we obtain
$$(1 - \mu) \frac{\chi\_m f(z)}{z} + \mu \left(\chi\_m f(z)\right)' + \delta z \left(\chi\_m f(z)\right)''$$
$$= 1 + \mathbb{C}\_1^a(\mathbf{x}) c\_1 z + \left[\mathbb{C}\_1^a(\mathbf{x}) c\_2 + \mathbb{C}\_2^a(\mathbf{x}) c\_1^2\right] z^2 + \dots, z \in \mathbb{U},\tag{9}$$
and
$$(1 - \mu) \frac{\chi\_{\text{wg}}(w)}{w} + \mu (\chi\_{\text{wg}}(w))' + \delta w (\chi\_{\text{wg}}(w))''$$
$$= 1 + \mathbb{C}\_1^u(\mathbf{x}) d\_1 w + \left[ \mathbb{C}\_1^u(\mathbf{x}) d\_2 + \mathbb{C}\_2^u(\mathbf{x}) d\_1^2 \right] w^2 + \dots, \ w \in \mathbb{U},\tag{10}$$
where
$$\omega(z) = \sum\_{j=1}^{\infty} c\_j z^j, \; z \in \mathbb{U}, \quad \text{and} \quad v(w) = \sum\_{j=1}^{\infty} d\_j w^j, \; w \in \mathbb{U}. \tag{11}$$
According to Lemma 1, if the above function *ω* and *v* has the form (11), then
Thus, upon comparing and equating the corresponding coefficients in (9) and (10), we have
$$\frac{(1+\mu+2\delta)m}{e^m-1}a\_2 = \mathbf{C}\_1^{\mathfrak{a}}(\mathfrak{x})c\_{1\prime} \tag{13}$$
$$\frac{(1+2\mu+6\delta)m^2}{2(e^m-1)}a\_3 = \mathcal{C}\_1^{\mathfrak{a}}(\mathfrak{x})c\_2 + \mathcal{C}\_2^{\mathfrak{a}}(\mathfrak{x})c\_1^2. \tag{14}$$
$$-\frac{(1+\mu+2\delta)m}{e^m-1}a\_2 = \mathcal{C}\_1^a(\mathfrak{x})d\_1. \tag{15}$$
and
$$\frac{(1+2\mu+6\delta)m^2}{2(e^m-1)}\left[2a\_2^2-a\_3\right]=\mathcal{C}\_1^a(\mathbf{x})d\_2+\mathcal{C}\_2^a(\mathbf{x})d\_1^2.\tag{16}$$
It follows from (13) and (15) that
$$c\_1 = -d\_{1\prime} \tag{17}$$
and
$$\frac{2(1+\mu+2\delta)^2m^2}{\left(e^m-1\right)^2}a\_2^2 = \left[\mathbb{C}\_1^a(\infty)\right]^2\left(c\_1^2+d\_1^2\right).\tag{18}$$
If we add (14) and (16), we get
$$\frac{(1+2\mu+6\delta)m^2}{(e^m-1)}a\_2^2 = \mathcal{C}\_1^a(\mathfrak{x})(c\_2+d\_2) + \mathcal{C}\_2^a(\mathfrak{x})\left(c\_1^2+d\_1^2\right). \tag{19}$$
Substituting the value of *c* 2 <sup>1</sup> + *d* 2 1 from (18) in the right hand side of (19), we deduce that
$$\left[ (1 + 2\mu + 6\delta) - \frac{2(1 + \mu + 2\delta)^2}{\left(\varepsilon^m - 1\right)} \frac{\mathbb{C}\_2^a(\mathbf{x})}{\left[\mathbb{C}\_1^a(\mathbf{x})\right]^2} \right] \frac{m^2}{\left(\varepsilon^m - 1\right)} a\_2^2 = \mathbb{C}\_1^a(\mathbf{x}) (c\_2 + d\_2). \tag{20}$$
Now, using (1), (12) and (20), we find that (6) holds. Moreover, if we subtract (16) from (14), we obtain
$$\frac{(1+2\mu+6\delta)m^2}{(e^m-1)}\left(a\_3-a\_2^2\right)=\mathcal{C}\_1^a(\infty)(c\_2-d\_2)+\mathcal{C}\_2^a(\infty)\left(c\_1^2-d\_1^2\right).\tag{21}$$
Then, in view of (17) and (18), Equation (21) becomes
$$a\_3 = \frac{(e^m - 1)^2 \left[\mathbb{C}\_1^{\alpha}(\mathfrak{x})\right]^2}{2m^2(1 + \mu + 2\delta)^2} \left(c\_1^2 + d\_1^2\right) + \frac{(e^m - 1)\mathbb{C}\_1^{\alpha}(\mathfrak{x})}{m^2(1 + 2\mu + 6\delta)} (c\_2 - d\_2).$$
Thus, applying (1), we conclude that
$$|a\_3| \le \frac{4\alpha^2 \varkappa^2 (e^m - 1)^2}{m^2 (1 + \mu + 2\delta)^2} + \frac{4\alpha \varkappa (e^m - 1)}{m^2 (1 + 2\mu + 6\delta)'},$$
and the proof of the theorem is complete.
The following result addresses the Fekete–Szeg˝o functional problem for functions in the class *ζ*Σ(*x*, *α*, *δ*, *µ*).
**Theorem 2.** *If the function f belongs to the class ζ*Σ(*x*, *α*, *δ*, *µ*)*, then*
$$\left| a\_3 - \eta a\_2^2 \right| \le \begin{cases} \frac{4a x (\epsilon^m - 1)}{m^2 (1 + 2\mu + 6\delta)}, & \text{if} \quad |\eta - 1| \le M, \\\\ \frac{8a^2 x (\epsilon^m - 1)^2 |1 - \eta|}{m^2 \left\{ \left[ 2a (1 + 2\mu + 6\delta) (\epsilon^m - 1) - 2(1 + a) (1 + \mu + 2\delta)^2 \right] x^2 + (1 + \mu + 2\delta)^2 \right\} \end{cases}, & \text{if} \quad |\eta - 1| \ge M, \end{cases}$$
*where*
$$M := \left| 1 - \frac{(1 + \mu + 2\delta)^2 \left[ 2(1 + \alpha)x^2 - 1 \right]}{2\alpha x^2 (e^m - 1)(1 + 2\mu + 6\delta)} \right|.$$
**Proof.** If *f* ∈ *ζ*Σ(*x*, *α*, *δ*, *µ*), from (20) and (21) we get
$$\begin{split} a\_3 - \eta a\_2^2 &= (1 - \eta) \frac{(e^m - 1)^2 \left[\mathcal{C}\_1^a(\mathbf{x})\right]^3 (c\_2 + d\_2)}{m^2 \left[\left(e^m - 1\right)\left(1 + 2\mu + 6\delta\right) \left[\mathcal{C}\_1^a(\mathbf{x})\right]^2 - 2\left(1 + \mu + 2\delta\right)^2 \mathcal{C}\_2^a(\mathbf{x})\right]} \\ &\quad + \frac{(e^m - 1)\mathcal{C}\_1^a(\mathbf{x})}{m^2 \left(1 + 2\mu + 6\delta\right)} (c\_2 - d\_2) \\ &= \mathcal{C}\_1^a(\mathbf{x}) \left[h(\eta) + \frac{(e^m - 1)}{m^2 \left(1 + 2\mu + 6\delta\right)}\right] c\_2 + \left[h(\eta) - \frac{(e^m - 1)}{m^2 \left(1 + 2\mu + 6\delta\right)}\right] d\_2 \end{split}$$
where
$$h(\eta) = \frac{(e^m - 1)^2 \left[\mathbb{C}\_1^a(\boldsymbol{x})\right]^2 (1 - \eta)}{m^2 \left[(e^m - 1)(1 + 2\mu + 6\delta) \left[\mathbb{C}\_1^a(\boldsymbol{x})\right]^2 - 2(1 + \mu + 2\delta)^2 \mathbb{C}\_2^a(\boldsymbol{x})\right]}.$$
Then, in view of (1), we conclude that
$$\left| a\_3 - \eta a\_2^2 \right| \le \begin{cases} \frac{4\alpha x (e^{\eta} - 1)}{m^2 (1 + 2\mu + 6\delta)'} & \text{if} \quad 0 \le |h(\eta)| \le \frac{(e^{\eta} - 1)}{m^2 (1 + 2\mu + 6\delta)'} \\\\ 4\alpha x |h(\eta)|\_{\prime} & \text{if} \quad |h(\eta)| \ge \frac{(e^{\eta} - 1)}{m^2 (1 + 2\mu + 6\delta)'} \end{cases}$$
,
which completes the proof of Theorem 2.
#### **5. Corollaries and Consequences**
Corresponding essentially to the Example 1 (setting *δ* = 0) and Example 2 (setting *δ* = 0 and *µ* = 1), from Theorems 1 and 2 we get the following consequences, respectively.
**Corollary 1.** *If the function f belongs to the class ζ*Σ(*x*, *α*, *µ*)*, then*
$$|a\_2| \le \frac{2\alpha \mathbf{x} (e^m - 1)\sqrt{2\mathbf{x}}}{m\sqrt{\left| \left[ 2\alpha (1 + 2\mu) (e^m - 1) - 2(1 + \mu)(1 + \mu)^2 \right] \mathbf{x}^2 + (1 + \mu)^2 \right|}},$$
$$|a\_3| \le \frac{4\alpha^2 \mathbf{x}^2 (e^m - 1)^2}{m^2 (1 + \mu)^2} + \frac{4\alpha \mathbf{x} (e^m - 1)}{m^2 (1 + 2\mu)},$$
*and*
$$\left| a\_3 - \eta a\_2^2 \right| \le \begin{cases} \frac{4\alpha x (\epsilon^m - 1)}{m^2 (1 + 2\mu)}, & \text{if} \quad |\eta - 1| \le N, \\\\ \frac{8\alpha^2 x^3 (\epsilon^m - 1)^2 |1 - \eta|}{\left| m^2 \left\{ \left[ 2\alpha (1 + 2\mu) (\epsilon^m - 1) - 2(1 + \mu)(1 + \mu)^2 \right] \mathbf{x}^2 + (1 + \mu)^2 \right\} \right|}, & \text{if} \quad |\eta - 1| \ge N\mu, \end{cases}$$
*where*
$$N := \left| 1 - \frac{(1+\mu)^2 \left[ 2(1+\alpha)x^2 - 1 \right]}{2\alpha x^2 (e^m - 1)(1+2\mu)} \right|.$$
$$\begin{aligned} |a\_2| &\le \frac{2\alpha \mathbf{x} (e^m - 1)\sqrt{2\mathbf{x}}}{m\sqrt{|\left[6\alpha (e^m - 1) - 8(1 + \alpha)\right] \mathbf{x}^2 + 4|}},\\ |a\_3| &\le \frac{\alpha^2 \mathbf{x}^2 (e^m - 1)^2}{m^2} + \frac{4\alpha \mathbf{x} (e^m - 1)}{3m^2},\end{aligned}$$
*and*
$$\left| a\_3 - \eta a\_2^2 \right| \le \begin{cases} \frac{4\alpha x (\epsilon^m - 1)}{3m^2}, & \text{if } \quad |\eta - 1| \le L\_\prime \\\\ \frac{8\alpha^2 x^3 (\epsilon^m - 1)^2 |1 - \eta|}{\left| m^2 \left\{ \left[ 6\alpha (\epsilon^m - 1) - 8(1 + \alpha) \right] \chi^2 + 4 \right\} \right|}, & \text{if } \quad |\eta - 1| \ge L\_\prime \end{cases}$$
*where*
$$L := \left| 1 - \frac{2\left[2(1+\alpha)x^2 - 1\right]}{3\alpha x^2(e^m - 1)} \right|.$$
#### **6. Concluding Remarks**
In the present work we have constructed a new subclass *ζ*Σ(*x*, *α*, *δ*, *µ*) of normalized analytic and bi-univalent functions governed with the zero-truncated Poisson distribution series and Gegenbauer polynomials. For functions belonging to this class, we have made estimates of Taylor–Maclaurin coefficients, |*a*2| and |*a*3|, and solved the Fekete–Szeg˝o functional problem. Furthermore, by suitably specializing the parameters *δ* and *µ*, one can deduce the results for the subclasses *ζ*Σ(*x*, *α*, *µ*) and *ζ*Σ(*x*, *α*) which are defined, respectively, in Examples 1 and 2.
The results offered in this paper would lead to other different new results for the classes *ζ*Σ(*x*, 1/2, *δ*, *µ*) for Legendre polynomials and *ζ*Σ(*x*, 1, *δ*, *µ*) for Chebyshev polynomials.
It remains an open problem to derive estimates on the bounds of |*an*| for *n* ≥ 4, *n* ∈ N, for the subclasses that have been introduced here.
**Author Contributions:** Conceptualization, F.Y. and A.A.; methodology, A.A.; validation, F.Y., A.A., B.A.F. and T.B.; formal analysis, A.A.; investigation, F.Y., B.A.F. and T.B.; writing—original draft preparation, F.Y. and A.A.; writing—review and editing, F.Y., T.B.; supervision, B.A.F. All authors have read and agreed to the published version of the manuscript.
**Funding:** This research received no external funding.
**Institutional Review Board Statement:** Not applicable.
**Informed Consent Statement:** Not applicable.
**Data Availability Statement:** No data were used to support this study.
**Conflicts of Interest:** The authors declare no conflict of interest.
#### **References**
## *Article* **Sharp Bounds for the Second Hankel Determinant of Logarithmic Coefficients for Strongly Starlike and Strongly Convex Functions**
**Sevtap Sümer Eker 1,\* , Bilal ¸Seker <sup>1</sup> , Bilal Çekiç <sup>1</sup> and Mugur Acu <sup>2</sup>**
**Abstract:** The logarithmic coefficients are very essential in the problems of univalent functions theory. The importance of the logarithmic coefficients is due to the fact that the bounds on logarithmic coefficients of *f* can transfer to the Taylor coefficients of univalent functions themselves or to their powers, via the Lebedev–Milin inequalities; therefore, it is interesting to investigate the Hankel determinant whose entries are logarithmic coefficients. The main purpose of this paper is to obtain the sharp bounds for the second Hankel determinant of logarithmic coefficients of strongly starlike functions and strongly convex functions.
**Keywords:** logarithmic coefficient; Hankel determinant; strongly starlike; strongly convex
**MSC:** 30C45; 30C50
#### **1. Introduction**
Let A stand for the standard class of analytic functions of the form
$$f(z) = z + \sum\_{k=2}^{\infty} a\_k z^k, \qquad z \in \mathbb{U} = \{ z \in \mathbb{C} : |z| < 1 \}, \tag{1}$$
and let S be the class of functions in A, which are univalent in U. A function *f* of the form (1) is said to be *starlike of order α* in U if
$$\Re\left\{\frac{zf'(z)}{f(z)}\right\} > a \qquad (z \in \mathbb{U}).$$
The set of all such functions is denoted by S ∗ (*α*).
Next, by K(*α*), we denote the class of *convex functions of order α* in U that satisfy the following inequality:
$$\Re\left\{1+\frac{zf''(z)}{f'(z)}\right\} > a \qquad (z \in \mathbb{U}).$$
A function *f* of the form (1) is said to be *strongly starlike of order α*, (0 < *α* ≤ 1), in U if
$$\left| \arg \frac{zf'(z)}{f(z)} \right| < \frac{\pi \mathfrak{a}}{2} \qquad (z \in \mathbb{U}).\tag{2}$$
**Citation:** Sümer Eker, S.; ¸Seker, B.; Çekiç, B.; Acu, M. Sharp Bounds for the Second Hankel Determinant of Logarithmic Coefficients for Strongly Starlike and Strongly Convex Functions. *Axioms* **2022**, *11*, 369. https://doi.org/10.3390/ axioms11080369
Academic Editor: Georgia Irina Oros
Received: 17 May 2022 Accepted: 7 June 2022 Published: 28 July 2022
**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.
**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).
The set of all such functions is denoted by S ∗ *s* (*α*). Moreover, a function *f* of the form (1) is said to be *strongly convex of order α*, (0 < *α* ≤ 1), in U if
$$\left| \arg \left( 1 + \frac{z f''(z)}{f'(z)} \right) \right| < \frac{\pi a}{2} \qquad (z \in \mathbb{U}).\tag{3}$$
The set of all such functions is denoted by K*c*(*α*).
The class S ∗ *s* (*α*) was independently introduced by Brannan and Kirwan [1] and Stankiewicz [2] (see also [3]). Clearly, S ∗ *s* (1) = S ∗ is the class of starlike functions and K∗ *c* (1) = K is the class of convex functions in U. We should observe that as *α* increases the sets S ∗ (*α*) and K(*α*) become smaller; however as *α* increases the sets S ∗ *s* (*α*) and K*c*(*α*) become larger. Furthermore, although the sharp coefficient bounds of the functions in the classes S ∗ (*α*) and K(*α*) are known, sharp coefficient bounds for the functions in the sets S ∗ *s* (*α*) and K*c*(*α*) are much harder to obtain, and only partial results are known [1,4].
Let P denote the class of analytic functions *p*(*z*) in U satisfying *p*(0) = 1 and < *p*(*z*) > 0. Thus, if *p* ∈ P, then have the following form:
$$p(z) = 1 + \sum\_{k=1}^{\infty} c\_k z^k, \qquad z \in \mathbb{U}. \tag{4}$$
Functions in P are called *Carathedory functions*.
Associated with each *f* ∈ S, is a well-defined logarithmic function
$$\mathcal{F}\_f := \log \frac{f(z)}{z} = 2 \sum\_{k=1}^{\infty} \gamma\_k z^k, \qquad z \in \mathbb{U}. \tag{5}$$
The numbers *γ<sup>k</sup>* are called the *logarithmic coefficients of f* . The logarithmic coefficients are very essential in the problems of univalent functions coefficients. The importance of the logarithmic coefficients is due to the fact that the bounds on logarithmic coefficients of *f* can transfer to the Taylor coefficients of univalent functions themselves or to their powers, via the Lebedev–Milin inequalities.
Relatively little exact information is known about the logarithmic coefficients of *f* when *f* ∈ S. The logarithmic coefficients of the Koebe function K(*z*) = *z*(1 − *z*) <sup>−</sup><sup>2</sup> are *γ<sup>k</sup>* = 1/*k*. Because of the extremal properties of the Koebe function, one could expect that *γ<sup>k</sup>* ≤ 1/*k*, for each *f* ∈ S; however, this conjecture is false even in the case *k* = 2. For the whole class S, the sharp estimates of single logarithmic coefficients are known only for
$$|\gamma\_1| \le 1 \qquad \text{and} \qquad |\gamma\_2| \le \frac{1}{2} + \frac{1}{\varepsilon^2} = 0.6353\dots$$
and are unknown for *k* ≥ 3. Recently, logarithmic coefficients have been studied by various authors and upper bounds of logarithmic coefficients of functions in some important subclasses of S have been obtained (e.g., [5–10]). For a summary of some of the significant results concerning the logarithmic coefficients for univalent functions, we refer to [11].
For *q*, *n* ∈ N, the Hankel determinant *Hq*,*n*(*f*) of *f* ∈ A of form (1) is defined as
$$H\_{q,n}(f) = \begin{vmatrix} a\_n & a\_{n+1} & \cdots & a\_{n+q-1} \\ a\_{n+1} & a\_{n+2} & \cdots & a\_{n+q} \\ \vdots & \vdots & & \vdots \\ a\_{n+q-1} & a\_{n+q} & \cdots & a\_{n+2(q-1)} \end{vmatrix}.$$
The Hankel determinant *H*2,1(*f*) = *a*<sup>3</sup> − *a* 2 2 is the well-known Fekete–Szegö functional. The second Hankel determinant *H*2,2(*f*) is given by *H*2,2(*f*) = *a*2*a*<sup>4</sup> − *a* 2 3 .
The problem of computing the upper bound of *Hq*,*<sup>n</sup>* over various subfamilies of A is interesting and widely studied in the literature on the geometric function theory of complex analysis. The upper bounds of *H*2,2, *H*3,1 and higher-order Hankel determinants for subclasses of analytic functions were obtained by various authors [12–24].
Very recently, Kowalczyk and Lecko [25] introduced the Hankel determinant *Hq*,*n*(*F<sup>f</sup>* /2), which are logarithmic coefficients of *f* , i.e.,
$$H\_{q,n}(\mathbb{F}\_f/2) = \begin{vmatrix} \gamma\_n & \gamma\_{n+1} & \cdots & \gamma\_{n+q-1} \\ \gamma\_{n+1} & \gamma\_{n+2} & \cdots & \gamma\_{n+q} \\ \vdots & \vdots & & \vdots \\ \gamma\_{n+q-1} & \gamma\_{n+q} & \cdots & \gamma\_{n+2(q-1)} \end{vmatrix}\_{\mathbb{F}}$$
For a function *f* ∈ S given in (1), by differentiating (5) one can obtain the following:
$$
\gamma\_1 = \frac{1}{2} a\_2, \ \gamma\_2 = \frac{1}{2} (a\_3 - \frac{1}{2} a\_2^2), \ \gamma\_3 = \frac{1}{2} (a\_4 - a\_2 a\_3 + \frac{1}{3} a\_2^3). \tag{6}
$$
.
Therefore, the second Hankel determinant of *F<sup>f</sup>* /2 can be obtained by
$$H\_{2,1}(\mathcal{F}\_f/\mathcal{Z}) = \gamma\_1 \gamma\_3 - \gamma\_2^2 = \frac{1}{4} \left( a\_2 a\_4 - a\_3^2 + \frac{1}{12} a\_2^4 \right). \tag{7}$$
Furthermore, if *f* ∈ S, then for
$$f\_{\theta}(z) = e^{-i\theta} f(e^{i\theta} z) \qquad (\theta \in \mathbb{R})\_{\prime}$$
we find that (see [26])
$$H\_{2,1}\left(\frac{F\_{f\_\theta}}{2}\right) = e^{4i\theta} H\_{2,1}\left(\frac{F\_f}{2}\right).$$
Kowalczyk and Lecko [26] obtained sharp bounds for *H*2,1(*F<sup>f</sup>* /2) for the classes of starlike and convex functions of order *α*. The problem of computing the sharp bounds of *H*2,1(*F<sup>f</sup>* /2) for starlike and convex functions with respect to symmetric points in the open unit disk has been considered by Allu and Arora [27].
In this paper, we calculate the sharp bounds for *H*2,1(*F<sup>f</sup>* /2) = *γ*1*γ*<sup>3</sup> − *γ* 2 2 for the classes S ∗ *s* (*α*) and K*c*(*α*).
To establish our main results, we will require the following Lemmas:
**Lemma 1** ([28] (see also [26]))**.** *If p* ∈ P *is of the form (4) with c*<sup>1</sup> ≥ 0*, then*
$$\begin{aligned} c\_1 &= 2d\_{1\prime} \\ c\_2 &= 2d\_1^2 + 2(1 - d\_1^2)d\_2 \\ c\_3 &= 2d\_1^3 + 4(1 - d\_1^2)d\_1d\_2 - 2(1 - d\_1^2)d\_1d\_2^2 + 2(1 - d\_1^2)(1 - |d\_2|^2)d\_3 \end{aligned} \tag{8}$$
*for some d*<sup>1</sup> ∈ [0, 1] *and d*2, *d*<sup>3</sup> ∈ U = {*z* ∈ C : |*z*| ≤ 1}*.*
*For d*<sup>1</sup> ∈ U *and d*<sup>2</sup> ∈ *∂*U = {*z* ∈ C : |*z*| = 1}*, there is a unique function p* ∈ P *with c*<sup>1</sup> *and c*<sup>2</sup> *as in (8), namely*
$$p(z) = \frac{1 + (\overline{d\_1}d\_2 + d\_1)z + d\_2z^2}{1 + (\overline{d\_1}d\_2 - d\_1)z - d\_2z^2}, \qquad z \in \mathbb{U}.$$
**Lemma 2** ([29])**.** *Given real numbers A, B, C, let*
$$Y(A, B, \mathbb{C}) = \max\left\{ \left| A + Bz + \mathbb{C}z^2 \right| + 1 - |z|^2 \, : \, z \in \mathbb{U} \right\}.$$
*I. If AC* ≥ 0*, then*
$$Y(A,B,\mathbb{C}) = \begin{cases} |A| + |B| + |\mathbb{C}| , & |B| \ge 2(1 - |\mathbb{C}|), \\\\ 1 + |A| + \frac{B^2}{4(1 - |\mathbb{C}|)^{\prime}} , & |B| < 2(1 - |\mathbb{C}|). \end{cases}$$
*II. If AC* < 0*, then*
$$Y(A,B,\mathbb{C}) = \begin{cases} 1 - |A| + \frac{B^2}{4(1 - |\mathbb{C}|)'} & -4AC(\mathbb{C}^{-2} - 1) \le B^2 \land |B| < 2(1 - |\mathbb{C}|), \\\\ 1 + |A| + \frac{B^2}{4(1 + |\mathbb{C}|)'} & B^2 < \min\{4(1 + |\mathbb{C}|)^2, \ -4AC(\mathbb{C}^{-2} - 1)\}, \\\\ R(A,B,\mathbb{C}), & \text{otherwise.} \end{cases}$$
*where*
$$\mathcal{R}(A,\mathcal{B},\mathbb{C}) = \begin{cases} |A| + |\mathcal{B}| - |\mathbb{C}|, & |\mathbb{C}|(|\mathcal{B}| + 4|A|) \le |A\mathcal{B}|, \\\\ -|A| + |\mathcal{B}| + |\mathbb{C}|, & |AB| \le |\mathbb{C}|(|\mathcal{B}| - 4|A|), \\\\ (|A| + |\mathbb{C}|) \sqrt{1 - \frac{B^2}{4AC}}, & \text{otherwise.} \end{cases}$$
**2. Second Hankel Determinant of Logarithmic Coefficients for the Class** S ∗ *s* (*α*) **Theorem 1.** *Let α* ∈ (0, 1]*. If f* ∈ S<sup>∗</sup> *s* (*α*)*, then*
$$\left|\gamma\_1\gamma\_3 - \gamma\_2^2\right| \le \frac{\mathfrak{a}^2}{4}.\tag{9}$$
*This inequality is sharp. Equality holds for the function*
$$f(z) = z \exp\int\_0^z \frac{(1 - u^2)^{-2\alpha} - 1}{u} \, \mathrm{d}u, \qquad z \in \mathbb{U}. \tag{10}$$
**Proof.** Let *α* ∈ (0, 1] and *f* ∈ S<sup>∗</sup> *s* (*α*) be of the form (1). Then by (2) we have
$$\frac{zf'(z)}{f(z)} = \left(p(z)\right)^a, \qquad z \in \mathbb{U},\tag{11}$$
for some function *p* ∈ P of the form (4). Since the class P and the functional |*H*2,1(*F<sup>f</sup>* /2)| are rotationally invariant, we may assume that *c*<sup>1</sup> ∈ [0, 2] (i.e., in view of (8) that *d*<sup>1</sup> ∈ [0, 1]). Equating the coefficients, we obtain
$$\begin{aligned} a\_2 &= \alpha c\_1\\ a\_3 &= \frac{a}{2} \left( c\_2 - \frac{1 - 3\alpha}{2} c\_1^2 \right) \\ a\_4 &= \frac{a}{3} \left( c\_3 + \frac{5\alpha - 2}{2} c\_1 c\_2 + \frac{17\alpha^2 - 15\alpha + 4}{12} c\_1^3 \right) .\end{aligned} \tag{12}$$
Hence by using (6)–(8) we obtain
$$\begin{split} \gamma\_1 \gamma\_3 - \gamma\_2^2 &= \frac{1}{4} \Big( a\_2 a\_4 - a\_3^2 + \frac{1}{12} a\_2^4 \Big) \\ &= \frac{a^2}{576} \Big[ (7+a)(1-a)c\_1^4 - 12(1-a)c\_1^2 c\_2 + 48c\_1 c\_3 - 36c\_2^2 \Big] \\ &= \frac{a^2}{36} \Big[ (4-a^2)d\_1^4 + 6a(1-d\_1^2)d\_1^2 d\_2 - (1-d\_1^2) \Big[ 12d\_1^2 + 9(1-d\_1^2) \Big] d\_2^2 \\ &\quad + 12(1-d\_1^2)(1-|d\_2|^2)d\_1 d\_3 \Big]. \end{split} \tag{13}$$
Now, we may have the following cases on *d*1: **Case 1.** Suppose that *d*<sup>1</sup> = 1. Then by (13) we obtain
$$\left| \gamma\_1 \gamma\_3 - \gamma\_2^2 \right| = \frac{\alpha^2}{36} (4 - \alpha^2)^2$$
**Case 2.** Suppose that *d*<sup>1</sup> = 0. Then by (13) we obtain
$$\left|\gamma\_1\gamma\_3 - \gamma\_2^2\right| = \frac{\alpha^2}{4} |d\_2|^2 \le \frac{\alpha^2}{4}.$$
**Case 3.** Suppose that *d*<sup>1</sup> ∈ (0, 1). By the fact that |*d*3| ≤ 1, applying the triangle inequality to (13) we can write
$$\begin{aligned} \left| \gamma\_1 \gamma\_3 - \gamma\_2^2 \right| &= \left| \frac{a^2 (1 - d\_1^2)}{3} \left[ \frac{4 - a^2}{12 (1 - d\_1^2)} d\_1^4 + \frac{a}{2} d\_1^2 d\_2 - \frac{12d\_1^2 + 9(1 - d\_1^2)}{12} d\_2^2 + (1 - |d\_2|^2) d\_1 d\_3 \right] \right| \\ &\le \frac{a^2 d\_1 (1 - d\_1^2)}{3} \left[ \left| \frac{4 - a^2}{12 (1 - d\_1^2)} d\_1^3 + \frac{a}{2} d\_1 d\_2 - \frac{12d\_1^2 + 9(1 - d\_1^2)}{12 d\_1} d\_2^2 \right| + 1 - |d\_2|^2 \right] \\ &= \frac{a^2 d\_1 (1 - d\_1^2)}{3} \left[ \left| A + B d\_2 + C d\_2^2 \right| + 1 - |d\_2|^2 \right] \end{aligned} \tag{14}$$
where
$$A = \frac{4 - \alpha^2}{12(1 - d\_1^2)} d\_1^3 \qquad B = \frac{\alpha}{2} d\_1 \qquad \mathbb{C} = -\frac{d\_1^2 + 3}{4d\_1}$$
.
Since *AC* < 0, we apply Lemma 2 only for the case II.
We consider the following sub-cases.
3 (a) Since
$$-4AC\left(\frac{1}{C^2} - 1\right) - B^2 = \frac{(4 - a^2)d\_1^2(d\_1^2 + 3)}{12(1 - d\_1^2)} \left(\frac{16d\_1^2}{(d\_1^2 + 3)^2} - 1\right) - \frac{a^2d\_1^2}{4} \le 0$$
equivalent to (1 − *α* 2 )*d* 2 <sup>1</sup> ≤ 9, which evidently holds for *d*<sup>1</sup> ∈ (0, 1). Further, the inequality |*B*| < 2(1 − |*C*|) is equivalent to 3 + (1 + *α*)*d* 2 <sup>1</sup> − 4*d*<sup>1</sup> < 0 which is false for *d*<sup>1</sup> ∈ (0, 1).
3 (b) Since
$$4(1+|\mathcal{C}|)^2 = \frac{(d\_1^2 + 4d\_1 + 3)^2}{4d\_1^2} > 0$$
and
$$-4AC\left(\frac{1}{C^2} - 1\right) = \frac{(4 - \alpha^2)d\_1^2(d\_1^2 - 9)}{12(d\_1^2 + 3)} < 0\_\lambda$$
we see that the inequality
$$\frac{\alpha^2 d\_1^2}{4} < \min \left\{ 4(1+|\mathcal{C}|)^2, \ -4AC(\frac{1}{\mathcal{C}^2}-1) \right\}.$$
is false for *d*<sup>1</sup> ∈ (0, 1).
3 (c) The inequality
$$|\mathbb{C}| \left( |\mathcal{B}| + 4|A| \right) - |AB| = \frac{(d\_1^2 + 3)}{4d\_1} \left( \frac{ad\_1}{2} + \frac{(4 - a^2)d\_1^3}{3(1 - d\_1^2)} \right) - \frac{a(4 - a^2)d\_1^4}{24(1 - d\_1^2)} \le 0,$$
is equivalent to
$$d^4(8+\alpha^3-2\alpha^2-7\alpha)+d^2(24-6\alpha^2-6\alpha)+9\alpha \le 0.5$$
It is easy to verify that
$$\begin{aligned} d^4(8+\mathfrak{a}^3-2\mathfrak{a}^2-7\mathfrak{a}) + d^2(24-6\mathfrak{a}^2-6\mathfrak{a}) + 9\mathfrak{a} \\ &> d^4(32+\mathfrak{a}^3-8\mathfrak{a}^2-13\mathfrak{a}) + 9\mathfrak{a} > 0. \end{aligned}$$
for *d*<sup>1</sup> ∈ (0, 1). Thus, the inequality |*C*| |*B*| + 4|*A*| ≤ |*AB*| does not hold for *α* ∈ (0, 1] and *d*<sup>1</sup> ∈ (0, 1).
3 (d) We can write
$$|AB| - |C| \left( |B| - 4|A| \right) = \frac{a(4 - a^2)d\_1^4}{24(1 - d\_1^2)} - \frac{(d\_1^2 + 3)}{4d\_1} \left( \frac{ad\_1}{2} - \frac{(4 - a^2)d\_1^3}{3(1 - d\_1^2)} \right)$$
$$= \frac{1}{24(1 - t)} \left( K\_1 t^2 + L\_1 t + M\_1 \right)$$
where *t* = *d* 2 <sup>1</sup> ∈ (0, 1) and
$$\begin{aligned} K\_1 &= -\alpha^3 - 2\alpha^2 + 7\alpha + 8 \\ L\_1 &= 6(4 + \alpha - \alpha^2) \\ M\_1 &= -9\alpha. \end{aligned}$$
It is easy to see that *K*<sup>1</sup> > 0, *L*<sup>1</sup> > 0 and *M*<sup>1</sup> < 0, for *α* ∈ (0, 1].
For the equation *K*1*t* <sup>2</sup> <sup>+</sup> *<sup>L</sup>*1*<sup>t</sup>* <sup>+</sup> *<sup>M</sup>*1, we have <sup>∆</sup> <sup>=</sup> <sup>144</sup>(<sup>4</sup> <sup>+</sup> <sup>4</sup>*<sup>α</sup>* <sup>−</sup> *<sup>α</sup>* 3 ) > 0. Since *K*<sup>1</sup> > 0, *M*<sup>1</sup> *K*1 < 0 and *K*<sup>1</sup> + *L*<sup>1</sup> + *M*<sup>1</sup> = 32 − *α* <sup>3</sup> <sup>−</sup> <sup>8</sup>*<sup>α</sup>* <sup>2</sup> <sup>+</sup> <sup>4</sup>*<sup>α</sup>* <sup>&</sup>gt; 0, for *<sup>α</sup>* <sup>∈</sup> (0, 1], the equation *<sup>K</sup>*1*<sup>t</sup>* <sup>2</sup> + *L*1*t* + *M*<sup>1</sup> has positive unique root such that
$$0 < t\_1 = \frac{-L\_1 + \sqrt{\Delta}}{2K\_1} < 1\_{\nu}$$
Therefore, for *d* ∗ <sup>1</sup> = √ *t*1, it follows that |*AB*| = |*C*| |*B*| − 4|*A*|
Moreover, |*AB*| ≤ |*C*| |*B*| − 4|*A*| , when *d*<sup>1</sup> ∈ (0, *d* ∗ 1 ], and |*AB*| ≥ |*C*| |*B*| − 4|*A*| , when *d*<sup>1</sup> ∈ [*d* ∗ 1 , 1).
.
Then for *d*<sup>1</sup> ∈ (0, *d* ∗ 1 ], we can write from (14) and Lemma 2, we obtain
$$\left|\gamma\_1\gamma\_3 - \gamma\_2^2\right| \le \frac{a^2 d\_1 (1 - d\_1^2)}{3} \left(-|A| + |B| + |\mathbb{C}|\right) = \Phi(d\_1).$$
where
$$
\Phi(d\_1) = \frac{a^2}{36} \left( -\left(4 - a^2\right) d\_1^4 + 3\left(1 + 2a\right) d\_1^2 \left(1 - d\_1^2\right) + 9\left(1 - d\_1^2\right) \right).
$$
Since
$$\Phi'(d\_1) = \frac{-\alpha^2 d\_1}{9} \left[ (7 + 6\alpha - \alpha^2) d\_1^2 + 3(1 - \alpha) \right] < 0,$$
for *d*<sup>1</sup> ∈ [0, *d* ∗ 1 ], Φ is a decreasing function on [0, *d* ∗ 1 ]. This implies that
$$\left|\gamma\_1\gamma\_3 - \gamma\_2^2\right| \le \Phi(0) = \frac{\alpha^2}{4}$$
.
3 (e) Next consider the case *d*<sup>1</sup> ∈ [*d* ∗ 1 , 1]. Using the last case of Lemma 2,
$$\left|\gamma\_1\gamma\_3 - \gamma\_2^2\right| \le \frac{\alpha^2 d\_1 (1 - d\_1^2)}{3} \left( \left( |A| + |\mathbb{C}| \right) \sqrt{1 - \frac{B^2}{4AC}} \right) = \Psi(d\_1)$$
where
$$\Psi(d\_1) = \frac{\alpha^2}{18} [9 + (1 - \alpha^2)d\_1^4 - 6d\_1^2] \sqrt{\frac{(1 - \alpha^2)d\_1^2 + 3}{(4 - \alpha^2)(d\_1^2 + 3)}}.$$
To find the maximum of the function Ψ(*d*1) on the interval *d*<sup>1</sup> ∈ [*d* ∗ 1 , 1], let us investigate the derivative of Ψ(*d*1):
$$\begin{aligned} \Psi'(d\_1) &= \frac{-d\_1^2 a^2}{18(4-a^2)(d\_1^2+3)^2} \sqrt{\frac{(4-a^2)(d\_1^2+3)}{(1-a^2)d\_1^2+3}} \\ &\times \left[ 4(3-(1-a^2)d\_1^2)(d\_1^2+3)((1-a^2)d\_1^2+3) + 3a^2(9+(1-a^2)d\_1^4-6d\_1^2) \right] < 0, \end{aligned}$$
since
$$4(3 - (1 - \alpha^2)d\_1^2 \ge 8 + 4\alpha^2 > 0$$
and
$$9 + (1 - \alpha^2)d\_1^4 - 6d\_1^2 \ge 9 - d\_1^2 \left(6 - (1 - \alpha^2)d\_1^2\right) = 3 + (1 - \alpha^2)d\_1^2 > 0$$
for *α* ∈ (0, 1] and *d*<sup>1</sup> ∈ [*d* ∗ 1 , 1]. Thus Ψ is a decreasing function on [*d* ∗ 1 , 1]. Furthermore, Φ(*d* ∗ 1 ) = Ψ(*d* ∗ 1 ). This implies that
$$\left|\gamma\_1\gamma\_3 - \gamma\_2^2\right| \le \Psi(d\_1) \le \Psi(d\_1^\*) = \Phi(d\_1^\*) \le \Phi(0) = \frac{a^2}{4}.$$
Summarizing parts from Cases 1–3, it follows the desired inequality.
In order to show that the inequality is sharp, let us set *c*<sup>1</sup> = 0 and *d*<sup>2</sup> = 1 into (8). Then, we obtain *c*<sup>2</sup> = 2 and *c*<sup>3</sup> = 0. Hence by (12) we have *a*<sup>2</sup> = *a*<sup>4</sup> = 0 and *a*<sup>3</sup> = *α*. This shows that equality is attained for the function given in (10).
This completes the proof of the theorem.
For *α* = 1 we obtain the bounds for the class S ∗ of starlike functions given in [25].
**Corollary 1.** *Let f*(*z*) ∈ S<sup>∗</sup> *. Then*
$$\left|\gamma\_1\gamma\_3 - \gamma\_2^2\right| \le \frac{1}{4}.$$
*The inequality is sharp.*
**3. Second Hankel Determinant of Logarithmic Coefficients for the Class** K*c*(*α*)
**Theorem 2.** *Let α* ∈ (0, 1]*. If f* ∈ K*c*(*α*)*, then*
$$\left|\gamma\_1\gamma\_3 - \gamma\_2^2\right| \le \begin{cases} \frac{a^2}{36}, & 0 < a \le \frac{1}{3} \\\\ \frac{a^2(13a^2 + 18a + 17)}{144(a^2 + 6a + 4)}, & \frac{1}{3} < a \le 1. \end{cases} \tag{15}$$
*The inequalities in (15) are sharp.*
**Proof.** Let *α* ∈ (0, 1] and *f* ∈ K*c*(*α*) be of the form (1). Then, by (3), we have
$$1 + \frac{z f''(z)}{f'(z)} = \left( p(z) \right)^{\alpha}, \qquad z \in \mathbb{U}, \tag{16}$$
for some function *p* ∈ P of the form (4). As in the proof of Theorem 1, we may assume that *c*<sup>1</sup> ∈ [0, 2] (i.e., in view of (8) that *d*<sup>1</sup> ∈ [0, 1]). Equating the coefficients, we obtain
$$\begin{aligned} a\_2 &= \frac{\alpha}{2} c\_1\\ a\_3 &= \frac{\alpha}{6} \left( c\_2 - \frac{1 - 3\alpha}{2} c\_1^2 \right) \\ a\_4 &= \frac{\alpha}{144} \left( (17\alpha^2 - 15\alpha + 4)c\_1^3 + 6(5\alpha - 2)c\_1 c\_2 + 12c\_3 \right) .\end{aligned} \tag{17}$$
Hence, by using (6)–(8) we obtain
$$\begin{split} \gamma\_1 \gamma\_3 - \gamma\_2^2 &= \frac{1}{4} \Big( a\_2 a\_4 - a\_3^2 + \frac{1}{12} a\_2^4 \Big) \\ &= \frac{a^2}{2304} \Big[ \left( a^2 - 6a + 4 \right) c\_1^4 + 4(3a - 2) c\_1^2 c\_2 + 24 c\_1 c\_3 - 16 c\_2^2 \Big] \\ &= \frac{a^2}{144} \Big[ (2 + a^2) d\_1^4 + 6a (1 - d\_1^2) d\_1^2 d\_2 - (1 - d\_1^2) \left[ 6d\_1^2 + 4(1 - d\_1^2) \right] d\_2^2 \\ &\quad + 6(1 - d\_1^2)(1 - |d\_2|^2) d\_1 d\_3 \Big]. \end{split} \tag{18}$$
Now, we may have the following cases on *d*1: **Case 1.** Suppose that *d*<sup>1</sup> = 1. Then, by (18) we obtain
$$\left| \gamma\_1 \gamma\_3 - \gamma\_2^2 \right| = \frac{\alpha^2}{144} (2 + \alpha^2)^2$$
**Case 2.** Suppose that *d*<sup>1</sup> = 0. Then, by (18) we obtain
$$\left|\gamma\_1\gamma\_3 - \gamma\_2^2\right| = \frac{\alpha^2}{36} |d\_2|^2 \le \frac{\alpha^2}{36}.$$
**Case 3.** Suppose that *d*<sup>1</sup> ∈ (0, 1). By the fact that |*d*3| ≤ 1, applying the triangle inequality to (18) we can write
$$\begin{aligned} \left| \gamma\_1 \gamma\_3 - \gamma\_2^2 \right| &= \left| \frac{a^2}{144} \left[ (2 + a^2) d\_1^4 + 6a(1 - d\_1^2) d\_1^2 d\_2 \right. \\ &\left. - (1 - d\_1^2) \left[ 6d\_1^2 + 4(1 - d\_1^2) \right] d\_2^2 + 6(1 - d\_1^2)(1 - |d\_2|^2) d\_1 d\_3 \right] \right| \\ &\leq \frac{a^2 d\_1 (1 - d\_1^2)}{24} \left[ \left| \frac{(2 + a^2)}{6(1 - d\_1^2)} d\_1^3 + 4a d\_1 d\_2 - \frac{4 + 2d\_1^2}{6d\_1} d\_2^2 \right| + 1 - |d\_2|^2 \right] \\ &= \frac{a^2 d\_1 (1 - d\_1^2)}{24} \left[ \left| A + B d\_2 + \mathbb{C} d\_2^2 \right| + 1 - |d\_2|^2 \right] \end{aligned} \tag{19}$$
where
$$A = \frac{2 + \alpha^2}{6(1 - d\_1^2)} d\_1^3 \qquad B = \alpha d\_1 \qquad \mathcal{C} = -\frac{2 + d\_1^2}{3d\_1}.$$
Since *AC* < 0, we apply Lemma 2 only for the case II. We consider the following sub-cases. 3 (a) Note that
$$\begin{aligned} -4AC\left(\frac{1}{C^2} - 1\right) - B^2 &= \frac{-d\_1^2 \left[d\_1^2 (7a^2 - 4) + 26a^2 + 16\right]}{9(d\_1^2 + 2)} \\ &= \frac{-d\_1^2 \left[a^2 (7d\_1^2 + 26) + 4(4 - d\_1^2)\right]}{9(d\_1^2 + 2)} \le 0. \end{aligned}$$
for *d*<sup>1</sup> ∈ (0, 1) and *α* ∈ (0, 1]. On the other hand, we have
$$|B| - 2(1 - |C|) = \frac{d\_1^2(3\alpha + 2) - 6d\_1 + 4}{3d\_1}.$$
Since <sup>∆</sup> <sup>=</sup> <sup>4</sup>(<sup>1</sup> <sup>−</sup> <sup>12</sup>*α*) <sup>≤</sup> 0 for <sup>1</sup> <sup>12</sup> ≤ *α* < 1, we have
$$d\_1^2(3\alpha + 2) - 6d\_1 + 4 \ge 0.$$
Further, since <sup>∆</sup> <sup>=</sup> <sup>4</sup>(<sup>1</sup> <sup>−</sup> <sup>12</sup>*α*) <sup>&</sup>gt; 0 for 0 <sup>&</sup>lt; *<sup>α</sup>* <sup>&</sup>lt; <sup>1</sup> <sup>12</sup> , the equation
$$d\_1^2(3\alpha + 2) - 6d\_1 + 4 = 0$$
has the roots
$$s\_{1,2} = \frac{3 \pm \sqrt{1 - 12\alpha}}{3\alpha + 2}$$
which are greater than 1. So
$$d\_1^2(3\alpha + 2) - 6d\_1 + 4 > 0$$
for *d*<sup>1</sup> ∈ (0, 1) and *α* ∈ (0, 1].
Consequently |*B*| < 2(1 − |*C*|) does not hold for *d*<sup>1</sup> ∈ (0, 1) and *α* ∈ (0, 1] . 3 (b) Since
$$4(1+|\mathbb{C}|)^2 = \frac{4(d\_1^2 + 3d\_1 + 2)^2}{9d\_1^2} > 0$$
and
$$-4AC\left(\frac{1}{C^2} - 1\right) = -\frac{2d\_1^2(4 - d\_1^2)(\alpha^2 + 2)}{9(d\_1^2 + 2)} < 0\_\lambda$$
we see that the inequality
$$
\alpha^2 d\_1^2 < \min \left\{ 4(1+|\mathcal{C}|)^2 \; , \; -4A\mathcal{C} \left( \frac{1}{\mathcal{C}^2} - 1 \right) \right\},
$$
is false for *d*<sup>1</sup> ∈ (0, 1). 3 (c) We can write
$$|\mathbb{C}| \left( |B| + 4|A| \right) - |AB| = \frac{1}{18(1 - d\_1^2)} (\mathbb{K}\_2 d\_1^4 + \mathbb{L}\_2 d\_1^2 + M\_2)^2$$
where
$$\begin{aligned} K\_2 &= -3\alpha^3 + 4\alpha^2 - 12\alpha + 8, \\ L\_2 &= 8\alpha^2 - 6\alpha + 16 \\ M\_2 &= 12\alpha. \end{aligned}$$
It is easy to see that *L*<sup>2</sup> > 0 and *M*<sup>2</sup> > 0, for *α* ∈ (0, 1]. There are two cases according to the sign of *K*2:
(i) If *K*<sup>2</sup> ≥ 0, then we have
$$|\mathbb{C}| \left( |B| + 4|A| \right) - |AB| = \frac{1}{18(1 - d\_1^2)} (K\_2 d\_1^4 + L\_2 d\_1^2 + M\_2) > 0.5$$
(ii) If *K*<sup>2</sup> < 0, then using the fact that *α* ∈ (0, 1] and *d*<sup>1</sup> ∈ (0, 1), we can write
$$\begin{aligned} |\mathbb{C}| \left( |\mathcal{B}| + 4|A| \right) - |AB| &= \frac{1}{18(1 - d\_1^2)} \left( K\_2 d\_1^4 + L\_2 d\_1^2 + M\_2 \right) \\ &> \frac{1}{18(1 - d\_1^2)} \left( K\_2 + L\_2 d\_1^2 + M\_2 \right) \\ &= \frac{1}{18(1 - d\_1^2)} \left( L\_2 d\_1^2 - 3a^3 + 4a^2 + 8 \right) \\ &\ge \frac{1}{18(1 - d\_1^2)} \left( L\_2 d\_1^2 + 5 + 4a^2 \right) > 0. \end{aligned}$$
Therefore, the inequality |*C*| |*B*| + 4|*A*| ≤ |*AB*| does not hold for *α* ∈ (0, 1] and *d*<sup>1</sup> ∈ (0, 1). 3 (d) We can write
$$\begin{aligned} \left( |AB| - |\mathbb{C}| \right) \left( |B| - 4|A| \right) &= \frac{\alpha (a^2 + 2)}{6(1 - d\_1^2)} d\_1^4 - \frac{d\_1^2 + 2}{3d\_1} \left( a d\_1 - 4 \frac{a^2 + 2}{6(1 - d\_1^2)} d\_1^3 \right) \\ &= \frac{1}{18(1 - t)} \left( \mathcal{K}\_3 t^2 + L\_3 t + M\_3 \right) \end{aligned}$$
where *t* = *d* 2 <sup>1</sup> ∈ (0, 1) and
$$\begin{aligned} K\_3 &= 3a^3 + 4a^2 + 12a + 8, \\ L\_3 &= 8a^2 + 6a + 16 \\ M\_3 &= -12a. \end{aligned}$$
It is easy to see that *K*<sup>3</sup> > 0, *L*<sup>3</sup> > 0 and *M*<sup>3</sup> < 0, for *α* ∈ (0, 1].
For the equation *K*3*t* <sup>2</sup> + *L*3*t* + *M*<sup>3</sup> = 0, we have ∆ > 0. Since *<sup>M</sup>*<sup>3</sup> *K*3 < 0 and *K*<sup>3</sup> + *L*<sup>3</sup> + *M*<sup>3</sup> > 0, for *α* ∈ (0, 1], the equation *K*3*t* <sup>2</sup> + *L*3*t* + *M*<sup>3</sup> = 0 has a unique positive root *t*<sup>1</sup> < 1. Thus, the inequality |*AB*| − |*C*| |*B*| − 4|*A*| ≤ 0 holds for (0, *d* ∗∗ 1 ], where *d* ∗∗ <sup>1</sup> = √ *t*1. So we can write from (19) and Lemma 2,
$$\begin{aligned} \left| \gamma\_1 \gamma\_3 - \gamma\_2^2 \right| &\leq \frac{\alpha^2 d\_1 (1 - d\_1^2)}{24} \left( -|A| + |B| + |C| \right) \\ &= \frac{\alpha^2}{144} \Phi\_1(d\_1) \end{aligned}$$
where
$$\Phi\_1(d\_1) = \left(Dd\_1^4 + Ed\_1^2 + 4\right).$$
and
$$\begin{aligned} D &= -(\alpha^2 + 6\alpha + 4) \\ E &= 6\alpha - 2. \end{aligned}$$
If Φ0 1 (*d*1) = 2*d*<sup>1</sup> 2*Dd*<sup>2</sup> <sup>1</sup> + *E* = 0, then *d* 2 <sup>1</sup> <sup>=</sup> <sup>−</sup> *<sup>E</sup>* 2*D* . So if *<sup>E</sup>* <sup>=</sup> <sup>6</sup>*<sup>α</sup>* <sup>−</sup> <sup>2</sup> <sup>&</sup>gt; 0, i.e., <sup>1</sup> <sup>3</sup> < *α* ≤ 1, then we have a critical point:
$$
\xi = \sqrt{-\frac{E}{2D}} = \sqrt{\frac{3\mathfrak{a} - 1}{\mathfrak{a}^2 + 6\mathfrak{a} + 4}}.\tag{20}
$$
Since
$$\begin{aligned} K\_3 \xi^4 + L\_3 \xi^2 + M\_3 &= K\_3 \left( \frac{3\alpha - 1}{a^2 + 6\alpha + 4} \right)^2 + L\_3 \left( \frac{3\alpha - 1}{a^2 + 6\alpha + 4} \right) + M\_3 \\ &= \frac{39a^5 + 28a^4 - 243a^3 - 296a^2 - 156a - 56}{(a^2 + 6\alpha + 4)^2} \\ &\le \frac{-243a^3 - 296a^2 - 89a - 56}{(a^2 + 6\alpha + 4)^2} \\ &< 0, \end{aligned}$$
we have 0 < *ξ* < *d* ∗∗ 1 ; therefore, we obtain
$$\begin{aligned} \left| \gamma\_1 \gamma\_3 - \gamma\_2^2 \right| &\leq \frac{\alpha^2}{144} \Phi\_1(\xi) \\ &= \frac{\alpha^2 (13\alpha^2 + 18\alpha + 17)}{144(\alpha^2 + 6\alpha + 4)} \end{aligned}$$
for <sup>1</sup> <sup>3</sup> < *α* ≤ 1.
Furthermore, if <sup>0</sup> <sup>&</sup>lt; *<sup>α</sup>* <sup>≤</sup> <sup>1</sup> 3 , then the function Φ1(*d*1) is decreasing on (0, *d* ∗∗ 1 ]. Thus we have
$$\begin{aligned} \left| \gamma\_1 \gamma\_3 - \gamma\_2^2 \right| &\leq \frac{\alpha^2}{144} \Phi\_1(d\_1) \\ &\leq \frac{\alpha^2}{36} .\end{aligned}$$
3 (e) Next consider the case *d*<sup>1</sup> ∈ [*d* ∗∗ 1 , 1]. Using the last case of the Lemma 2,
$$\left|\gamma\_1\gamma\_3 - \gamma\_2^2\right| \le \frac{\alpha^2 d\_1 (1 - d\_1^2)}{24} \left( (|A| + |\mathbb{C}|) \sqrt{1 - \frac{B^2}{4AC}} \right)$$
$$= \frac{\alpha^2}{144} \Psi\_1(d\_1)$$
where
$$\Psi\_1(d\_1) = (\alpha^2 d\_1^4 - 2d\_1^2 + 4)\sqrt{1 + \frac{9\alpha^2(1 - d\_1^2)}{2(\alpha^2 + 2)(d\_1^2 + 2)}}.$$
To find the maximum of the function Ψ1(*d*1) on the interval *d*<sup>1</sup> ∈ [*d* ∗∗ 1 , 1], let us investigate the derivative of Ψ1(*d*1):
$$\begin{aligned} \Psi\_1'(d\_1) &= \frac{-d\_1}{(a^2+2)(d\_1^2+2)^2} \sqrt{\frac{(a^2+2)(d\_1^2+2)}{(4-7a^2)d\_1^2+13a^2+8}} \times \\ &\left\{ 4(d\_1^2+2)\left(1-a^2d\_1^2\right) \left[ \left(4-7a^2\right)d\_1^2+13a^2+8 \right] + \left(a^2d\_1^4-2d\_1^2+4\right)27a^2 \right\}. \end{aligned}$$
Since for *d*<sup>1</sup> ∈ [*d* ∗∗ 1 , 1]
$$\left(4 - 7\alpha^2\right)d\_1^2 + 13\alpha^2 + 8 = \alpha^2(13 - 7d\_1^2) + 4(d\_1^2 + 2) > 0$$
and
$$\left(a^2d\_1^4 - 2d\_1^2 + 4\right) = 4 - d\_1^2(2 - a^2d\_1^2) \ge 4 - (2 - a^2d\_1^2) = 2 + a^2d\_1^2 > 0,$$
for *α* ∈ (0, 1] and *d*<sup>1</sup> ∈ [*d* ∗∗ 1 , 1]. Thus Ψ1(*d*1) is a decreasing function on the interval [*d* ∗∗ 1 , 1]. This implies that
$$\left|\gamma\_1\gamma\_3 - \gamma\_2^2\right| \le \frac{\alpha^2}{144} \Psi\_1(d\_1) \le \frac{\alpha^2}{144} \Psi\_1(d\_1^{\*\*}) = \frac{\alpha^2}{144} \Phi\_1(d\_1^{\*\*}).$$
Summarizing parts from Cases 1–3, it follows the desired inequalities.
To show the sharpness for the case 0 <sup>&</sup>lt; *<sup>α</sup>* <sup>≤</sup> <sup>1</sup> 3 , consider the function
$$p\_1(z) = \frac{1 - z^2}{1 + z^2}, \qquad (z \in \mathbb{U}).$$
It is obvious that the function *p*<sup>1</sup> is in P with *c*<sup>1</sup> = *c*<sup>3</sup> = 0 and *c*<sup>2</sup> = −2. The corresponding function *f*<sup>1</sup> can be obtained from (16). Hence, by (17) we have *a*<sup>2</sup> = *a*<sup>4</sup> = 0 and *<sup>a</sup>*<sup>3</sup> <sup>=</sup> <sup>−</sup>*<sup>α</sup>* 3 . From (18) we obtain
$$
\left| \gamma\_1 \gamma\_3 - \gamma\_2^2 \right| = \frac{\alpha^2}{36} \lambda
$$
for 0 <sup>&</sup>lt; *<sup>α</sup>* <sup>≤</sup> <sup>1</sup> 3
.
For the case <sup>1</sup> <sup>3</sup> < *α* ≤ 1, consider the function
$$p\_2(z) = \frac{1 - z^2}{1 - 2\mathfrak{F}z + z^2}, \qquad (z \in \mathbb{U})^2$$
where *ξ* is given in (20). From Lemma 1, it is obvious that the function *p*<sup>2</sup> is in P. The corresponding function *f*<sup>2</sup> can be obtained from (16), having the following coefficients:
$$\begin{aligned} a\_2 &= \alpha \xi\_{\prime} \\ a\_3 &= \frac{1}{3} a \left( (1 + 3\alpha) \xi^2 - 1 \right) , \\ a\_4 &= \frac{1}{18} \alpha \xi \left( (17\alpha^2 + 15\alpha + 4) \xi^2 - 15\alpha - 3 \right) .\end{aligned}$$
Hence from (18) we obtain
$$\left|\gamma\_1\gamma\_3 - \gamma\_2^2\right| = \frac{\alpha^2(13\alpha^2 + 18\alpha + 17)}{144(\alpha^2 + 6\alpha + 4)}.$$
This completes the proof.
For *α* = 1 we obtain the bounds for the class K of convex functions given in [25].
**Corollary 2.** *Let f*(*z*) ∈ K*. Then*
$$\left|\gamma\_1\gamma\_3 - \gamma\_2^2\right| \le \frac{1}{33}.$$
*The inequality is sharp.*
#### **4. Discussion**
In this work, we have obtained the sharp bounds for the second Hankel determinant of logarithmic coefficients of strongly starlike functions and strongly convex functions. Because of the importance of the logarithmic coefficients of univalent functions, our results provide a basis for research on the Hankel determinant of the logarithmic coefficients of the class of strongly starlike and strongly convex functions and other classes associated with these classes. Furthermore, our results could also inspire further studies taking other subclasses of S into consideration and/or obtaining the bounds for higher-order Hankel determinants.
**Author Contributions:** Conceptualization, S.S.E., B.¸S., B.Ç. and M.A.; methodology, S.S.E., B.¸S., B.Ç. and M.A.; writing—original draft preparation, S.S.E., B.¸S., B.Ç. and M.A.; investigation, S.S.E., B.¸S., B.Ç. and M.A. All authors have read and agreed to the published version of the manuscript.
**Funding:** This research received no external funding.
**Data Availability Statement:** Not applicable.
**Conflicts of Interest:** The authors declare no conflict of interest.
#### **References**
### *Article* **New Results about Radius of Convexity and Uniform Convexity of Bessel Functions**
**Lumini¸ta-Ioana Cotîrl˘a 1,\* , Pál Aurel Kupán <sup>2</sup> and Róbert Szász <sup>2</sup>**
**Abstract:** We determine in this paper new results about the radius of uniform convexity of two kinds of normalization of the Bessel function *J<sup>ν</sup>* in the case *ν* ∈ (−2, −1), and provide an alternative proof regarding the radius of convexity of order alpha. We then compare results regarding the convexity and uniform convexity of the considered functions and determine interesting connections between them.
**Keywords:** Bessel function; convex function; uniformly convex functions; radius of convexity
**MSC:** 33C10
#### **1. Introduction**
Let *U*(*r*) = {*z* ∈ C : |*z*| < *r*} be the disk, centered at zero, of radius *r*, where *r* > 0. We denote by *U*(*r*) = *U*(0,*r*).
We say that a function *f* of the form
$$f(z) = z + a\_2 z^2 + \dots \tag{1}$$
is convex on *U*(*r*) if and only if *f*(*U*(*r*)) is a convex domain in the set C and the function *f* is univalent.
We know that the function *f* is convex on *U*(*r*) if and only if
$$\operatorname{Re}\left(1+\frac{zf''(z)}{f'(z)}\right)>0,\ z\in\mathcal{U}(r).$$
We say that *f* is a convex function of order *α* on *U*(*r*) if
$$\operatorname{Re}\left(1+\frac{zf''(z)}{f'(z)}\right)>\alpha,\quad z\in\mathcal{U}(r).$$
The radius of convexity of order *α* for *f* is defined by the equality
$$r\_f^\varepsilon(\mathfrak{a}) = \sup \left\{ r \in (0, \infty) : \text{Re}\left( 1 + \frac{zf''(z)}{f'(z)} \right) > \mathfrak{a}, \ z \in \mathcal{U}(r) \right\}. \tag{2}$$
We say that *f* is uniformly convex in the disk *U*(*r*) if the function *f* has the form in (1), it is a convex function, and it has the property that the arc *f*(*γ*) is convex for every circular arc *γ* contained in the disk *U*(*r*) with center *ζ*, also in *U*(*r*). The function *f* is uniformly convex in the disk *U*(*r*) if and only if
$$\left| \operatorname{Re} \left( 1 + \frac{z f'''(z)}{f'(z)} \right) > \left| \frac{z f'''(z)}{f'(z)} \right|, \ z \in \mathcal{U}(r). \right|$$
**Citation:** Cotîrl˘a, L.-I.; Kupán, P.A.; Szász, R. New Results about Radius of Convexity and Uniform Convexity of Bessel Functions. *Axioms* **2022**, *11*, 380. https://doi.org/10.3390/ axioms11080380
Academic Editor: Georgia Irina Oros
Received: 6 July 2022 Accepted: 30 July 2022 Published: 31 July 2022
**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.
**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).
We know that the radius of uniform convexity is defined by
$$r\_f^{\mu c}(a) = \sup \left\{ r \in (0, \infty) : \operatorname{Re} \left( 1 + \frac{z f''(z)}{f'(z)} \right) > \left| \frac{z f''(z)}{f'(z)} \right|, \ z \in \mathcal{U}(r) \right\}.\tag{3}$$
The Bessel function of the first kind is defined by
$$J\_{\nu}(z) = \sum\_{n=0}^{\infty} \frac{(-1)^n}{n! \Gamma(n+\nu+1)} \left(z/2\right)^{2n+\nu}.$$
Consider the following normalized forms:
$$g\_{\nu}(z) = 2^{\nu} \Gamma(1+\nu) z^{1-\nu} l\_{\nu}(z) = z - \frac{1}{4(\nu+1)} z^3 + \dots,\tag{4}$$
and
$$h\_{\nu}(z) = 2^{\nu} \Gamma(1+\nu) z^{1-\nu/2} f\_{\nu}(z^{\frac{1}{2}}) = z - \frac{1}{4(\nu+1)} z^2 + \dots,\tag{5}$$
where *ν* is a real number and −2 < *ν* < −1, and *g<sup>ν</sup>* and *h<sup>ν</sup>* are entire functions.
This article can be considered a continuation of previous papers [1,2] which dealt with geometric properties of Bessel functions.
For more details about the geometric properties of Bessel functions, interested readers are referred to the following papers: [1,3–13].
The aim of this work is to determine the radius of convexity of order *α*, *r c f* (*α*) for *f* = *g<sup>ν</sup>* and *f* = *h<sup>ν</sup>* and the radius of uniform convexity *r uc f* (*α*) for the case *ν* ∈ (−2, −1) and to derive an interesting connection between the convexity and uniform convexity.
In the next section, we provide several results which are necessary later in this work.
#### **2. Preliminaries**
**Lemma 1** ([14], p. 483, Hurwitz)**.** *If ν* ∈ (−2, −1), *then Jν*(*z*) *has exactly two purely imaginary conjugate complex zeros, and all the other zeros are real.*
The zeros *z* −*ν Jν*(*z*) are taken to be ±*jν*,*n*, where *n* ∈ N<sup>∗</sup> = {1, 2, 3, . . .}. We may suppose, without restricting the generality, that *jν*,1 = *ia*, *a* > 0, and 0 < *a* < *jν*,2 < *jν*,3 < · · · < *jν*,*<sup>n</sup>* < · · · .
**Lemma 2** ([14], p. 502)**.** *The following equality holds*
$$\sum\_{n=1}^{\infty} \frac{1}{\hat{f}\_{\nu,n}^2} = \frac{1}{4(\nu+1)}.\tag{6}$$
**Lemma 3** ([8])**.** *In the notations of Lemma 2, we have*
$$\frac{z g\_{\nu}'(z)}{g\_{\nu}(z)} = 1 - 2 \sum\_{n=1}^{\infty} \frac{z^2}{f\_{\nu,n}^2 - z^2} \tag{7}$$
*and*
$$\frac{zh\_{\nu}'(z)}{h\_{\nu}(z)} = 1 - \sum\_{n=1}^{\infty} \frac{z}{f\_{\nu,n}^2 - z}. \tag{8}$$
*The series are uniformly convergent on every compact subset of* C \ {±*jν*,*<sup>n</sup>* : *n* ∈ N∗}.
**Lemma 4** ([9])**.** *If v* ∈ C*, δ* ∈ R*, and δ* > *ρ* ≥ |*v*|*, then*
$$\left|\frac{v}{\delta - v}\right| \le \frac{\rho}{\delta - \rho} \text{ and } \left|\frac{v}{(\delta - v)^2}\right| \le \frac{\rho}{(\delta - \rho)^2}.$$
**Proof.** The following implications hold
$$|\delta - v| \ge \delta - \rho \Rightarrow \frac{1}{|\delta - v|} \le \frac{1}{\delta - \rho} \Rightarrow \left| \frac{1}{(\delta - v)^2} \right| \le \frac{1}{(\delta - \rho)^2}.$$
If the last two inequalities are multiplied by the inequality |*v*| ≤ *ρ*, we obtain the desired results.
**Lemma 5.** *If v* ∈ C*, δ*, *γ* ∈ R, *γ* ≥ *δ* > *ρ* ≥ |*v*|*, then*
$$\left|\frac{v^2}{(\delta+v)(\gamma-v)}\right| \le \frac{\rho^2}{(\delta-\rho)(\gamma+\rho)}.\tag{9}$$
**Proof.** We can prove the second inequality of the following equivalence:
$$\left|\frac{1}{(\delta+\upsilon)(\gamma-\upsilon)}\right| \le \frac{1}{(\delta-\rho)(\gamma+\rho)} \Leftrightarrow (\delta-\rho)(\gamma+\rho) \le \left|(\delta+\upsilon)(\gamma-\upsilon)\right|\_{\prime} \tag{10}$$
where *γ* ≥ *δ* > *ρ* ≥ |*v*|.
We prove the inequality (10) in two steps.
Let *v* = *x* + *iy*; then, it is obvious that
$$| (\delta + \upsilon)(\gamma - \upsilon) | = \sqrt{[(\delta + \mathbf{x})^2 + y^2][(+y^2 + \gamma - \mathbf{x})^2]} \ge | (\gamma - \mathbf{x})(\delta + \mathbf{x}) |,\tag{11}$$
where *γ* ≥ *δ* > *ρ* ≥ p *x* <sup>2</sup> + *y* 2 .
On the other hand, a simple calculation results in
$$(\delta + \mathbf{x})(\gamma - \mathbf{x}) \ge (\delta - \rho)(\gamma + \rho), \ge \in [-\rho, \rho]. \tag{12}$$
It is easily seen that (11) and (12) imply the second inequality of (10). Finally, multiplying the inequality *ρ* <sup>2</sup> ≥ |*v*<sup>|</sup> <sup>2</sup> by the first inequality of (10), we obtain (9) and the proof is complete.
**Lemma 6.** *If v* ∈ C*, δ*, *γ* ∈ R*, and γ* ≥ *δ* > *ρ* ≥ |*v*|*, then*
$$\left|\frac{2v^2[2\gamma\delta + (\gamma - \delta)v]}{(\gamma - v)^2(\delta + v)^2}\right| \le \frac{2r^2[2\gamma\delta - (\gamma - \delta)\rho]}{(\gamma + \rho)^2(\delta - \rho)^2}.\tag{13}$$
**Proof.** The inequality obviously holds provided that *γ* = *δ* (see (10)), thus, we have to prove it in the case that *γ* > *δ*.
We can then prove the following inequality:
$$\left|\frac{2\gamma\delta + (\gamma - \delta)v}{(\delta + v)(\gamma - v)}\right| \le \frac{2\gamma\delta - (\gamma - \delta)\rho}{(\delta - \rho)(\gamma + \rho)}, \quad \gamma \ge \delta > \rho \ge |v|. \tag{14}$$
We define *z* = *x* + *iy* and define the mapping
$$\phi : [ -\rho, \rho ] \to \mathbb{R}, \ \phi(y) = \frac{(\omega + \mathfrak{x})^2 + y^2}{[(\delta + \mathfrak{x})^2 + y^2][(\gamma - \mathfrak{x})^2 + y^2]}, \ \omega = \frac{2\gamma\delta}{\gamma - \delta}.$$
Then, we have
$$\phi'(y) = 2y \frac{[(\delta + \mathbf{x})^2 + y^2][(\gamma - \mathbf{x})^2 + y^2] - [(\delta + \mathbf{x})^2 + (\gamma - \mathbf{x})^2 + 2y^2][(\omega + \mathbf{x})^2 + y^2]}{[(\delta + \mathbf{x})^2 + y^2]^2[(\gamma - \mathbf{x})^2 + y^2]^2}.$$
As *φ* 0 (*y*) < 0, *y* ∈ (0, *ρ*) and *φ* 0 (*y*) > 0, *y* ∈ (−*ρ*, 0), it follows that
$$\phi(y) \le \phi(0) = \frac{(\omega + \mathfrak{x})^2}{[(\delta + \mathfrak{x})^2][(\gamma - \mathfrak{x})^2]}, \ y \in [-\rho, \rho]. \tag{15}$$
We can determine the maximum of the function
$$\varphi: [ -\rho, \rho ] \to \mathbb{R}, \ \rho(\mathfrak{x}) = \frac{\omega + \mathfrak{x}}{(\delta + \mathfrak{x})(\gamma - \mathfrak{x})}.$$
We have
$$\phi'(x) = \frac{x^2 + 2\omega x - \gamma \delta}{(\delta + x)^2 (\gamma - x)^2}$$
The derivative *ϕ* 0 (*x*) = 0 has one positive root, *x*<sup>1</sup> = p *<sup>ω</sup>*<sup>2</sup> + *γδ* − *<sup>ω</sup>*, and one negative root, *x*<sup>2</sup> = − p *<sup>ω</sup>*<sup>2</sup> + *γδ* − *<sup>ω</sup>*. As *<sup>x</sup>*<sup>2</sup> < −*<sup>r</sup>* and *<sup>x</sup>*<sup>1</sup> ∈ (−*ρ*, *<sup>ρ</sup>*), it follows the inequality
$$\frac{\omega + \mathbf{x}}{(\delta + \mathbf{x})(\gamma - \mathbf{x})} = \varrho(\mathbf{x}) \le \max\{\varrho(-\rho), \varrho(\rho)\} = \varrho(-\rho) = \frac{\omega - \rho}{(\delta - \rho)(\gamma + \rho)}\tag{16}$$
for every *x* ∈ [−*ρ*, *ρ*]. From (15) and (16), we have (14). Finally, multiplying the inequalities (14), |*v* 2 | ≤ *ρ* <sup>2</sup> and the first inequality of (10), we infer (13).
**Lemma 7.** *If the functions g<sup>ν</sup> and h<sup>ν</sup> are defined by (4) and (5), respectively, then*
$$\frac{z g\_{\nu}^{\prime\prime}(z)}{g\_{\nu}^{\prime}(z)} = z \frac{z \mathbf{J}\_{\nu+2}(z) - \mathbf{3} \mathbf{J}\_{\nu+1}(z)}{\mathbf{J}\_{\nu}(z) - z \mathbf{J}\_{\nu+1}(z)}. \tag{17}$$
$$\frac{zh\_{\nu}^{\prime\prime}(z)}{h\_{\nu}^{\prime}(z)} = \frac{zf\_{\nu+2}(z^{\frac{1}{2}}) - 4z^{\frac{1}{2}}f\_{\nu+1}(z^{\frac{1}{2}})}{4f\_{\nu}(z^{\frac{1}{2}}) - 2z^{\frac{1}{2}}f\_{\nu+1}(z^{\frac{1}{2}})}.\tag{18}$$
.
**Proof.** We differentiate the equality (4), and at the second time we differentiate it logarithmically. After multiplying by *z*, we obtain the following equality:
$$\frac{z g\_{\nu}^{\prime\prime}(z)}{g\_{\nu}^{\prime}(z)} = \frac{z^2 f\_{\nu}^{\prime\prime}(z) + 2z(1-\nu)f\_{\nu}^{\prime}(z) + \nu(\nu-1)f\_{\nu}(z)}{z f\_{\nu}^{\prime}(z) + (1-\nu)f\_{\nu}(z)}.$$
The function *J<sup>ν</sup>* is a solution of the Bessel differential equation; thus, we can replace the function *z* 2 *J* 00 *<sup>ν</sup>* using the equality *z* 2 *J* 00 *ν* (*z*) = (*ν* <sup>2</sup> <sup>−</sup> *<sup>z</sup>* 2 )*Jν*(*z*) − *zJ*<sup>0</sup> *ν* (*z*), and it follows that
$$\frac{z g\_{\nu}^{\prime\prime}(z)}{g\_{\nu}^{\prime}(z)} = \frac{z(1 - 2\nu) f\_{\nu}^{\prime}(z) + (2\nu^2 - \nu - z^2) f\_{\nu}(z)}{z f\_{\nu}^{\prime}(z) + (1 - \nu) f\_{\nu}(z)}$$
In the second step, we use the following well-known equality: *zJ*0 *ν* (*z*) = *νJν*(*z*) − *zJν*+1(*z*), and infer
$$\frac{z\mathbf{g}\_{\nu}^{\prime\prime}(z)}{\mathbf{g}\_{\nu}^{\prime}(z)} = \frac{z(2\nu - 1)J\_{\nu+1}(z) - z^2 J\_{\nu}(z)}{J\_{\nu}(z) - zJ\_{\nu+1}(z)}.$$
Finally, we replace *zJν*(*z*) in the numerator by *zJν*(*z*) = 2(*ν* + 1)*Jν*+1(*z*) − *zJν*+2(*z*), and obtain (17).
We differentiate equality (5) twice, similarly to the case of the function *gν*, and obtain
$$\frac{zh\_{\nu}^{\prime\prime}(z)}{h\_{\nu}^{\prime}(z)} = \frac{\nu(\nu-2)f\_{\nu}(z^{\frac{1}{2}}) + (3-2\nu)z^{\frac{1}{2}}f\_{\nu}^{\prime}(z^{\frac{1}{2}}) + zf\_{\nu}^{\prime\prime}(z^{\frac{1}{2}})}{2(2-\nu)f\_{\nu}(z^{\frac{1}{2}}) + 2z^{\frac{1}{2}}f\_{\nu}^{\prime}(z^{\frac{1}{2}})}.$$
We use the equality *zJ*00 *ν* (*z* 1 <sup>2</sup> ) = (*ν* <sup>2</sup> <sup>−</sup> *<sup>z</sup>*)*Jν*(*<sup>z</sup>* 1 <sup>2</sup> ) − *z* 1 2 *J* 0 *ν* (*z* 1 <sup>2</sup> ), and obtain
$$\frac{zh\_{\nu}^{\prime\prime}(z)}{h\_{\nu}^{\prime}(z)} = \frac{(2\nu^2 - 2\nu - z)f\_{\nu}(z^{\frac{1}{2}}) + (2 - 2\nu)z^{\frac{1}{2}}f\_{\nu}^{\prime}(z^{\frac{1}{2}})}{2(2 - \nu)f\_{\nu}(z^{\frac{1}{2}}) + 2z^{\frac{1}{2}}f\_{\nu}^{\prime}(z^{\frac{1}{2}})}.$$
Now, using the equality *z* 1 2 *J* 0 *ν* (*z* 1 <sup>2</sup> ) = *νJν*(*z* 1 <sup>2</sup> ) − *z* 1 <sup>2</sup> *Jν*+1(*z* 1 <sup>2</sup> ), we infer
$$\frac{zh\_{\nu}^{\prime\prime}(z)}{h\_{\nu}^{\prime}(z)} = \frac{(2\nu - 2)z^{\frac{1}{2}}J\_{\nu+1}(z^{\frac{1}{2}}) - zJ\_{\nu}(z^{\frac{1}{2}})}{4J\_{\nu}(z^{\frac{1}{2}}) - 2z^{\frac{1}{2}}J\_{\nu+1}(z^{\frac{1}{2}})} \; .$$
and combining this with the equality *z* 1 <sup>2</sup> *Jν*(*z* 1 <sup>2</sup> ) = 2(*ν* + 1)*Jν*+1(*z* 1 <sup>2</sup> ) − *z* 1 <sup>2</sup> *Jν*+2(*z* 1 <sup>2</sup> ), (18) follows.
#### **3. Main Results**
**Theorem 1.** *If α* ∈ [0, 1) *and ν* ∈ (−2, −1)*, then the radius of convexity of order α for the mapping g<sup>ν</sup> is r<sup>c</sup> ν* (*α*) = *r*1, *where r*<sup>1</sup> *is the unique root of the equation*
$$1 + r \frac{I\_{\nu+2}(r) + 3I\_{\nu+1}(r)}{I\_{\nu+1}(r) + rI\_{\nu}(r)} = \alpha \tag{19}$$
*in the interval* (0, *a*).
**Proof.** According to the proof of Theorem 1 [2], the equalities
$$\frac{zg\_{\nu}'(z)}{g\_{\nu}(z)} = 1 - 2\sum\_{n=1}^{\infty} \frac{z^2}{j\_{\nu,n}^2 - z^2}, \\ \sum\_{n=1}^{\infty} \frac{1}{j\_{\nu,n}^2} = \frac{1}{4(\nu+1)}.$$
imply
$$\frac{z g\_{\nu}'(z)}{g\_{\nu}(z)} = 1 - \frac{a^2}{2(1+\nu)} \frac{z^2}{a^2 + z^2} - 2 \sum\_{n=2}^{\infty} \frac{a^2 + f\_{\nu,n}^2}{f\_{\nu,n}^2} \frac{z^4}{(a^2 + z^2)(f\_{\nu,n}^2 - z^2)}.$$
The logarithmic differentation of this equality leads to
$$\begin{split} 1 + \frac{z g\_{\nu}^{\prime\prime}(z)}{g\_{\nu}^{\prime}(z)} &= \\ 1 - \frac{a^2}{2(1+\nu)} \frac{z^2}{a^2 + z^2} - 2 \sum\_{n=2}^{\infty} \frac{a^2 + j\_{\nu,n}^2}{j\_{\nu,n}^2} \frac{z^4}{(a^2 + z^2)(j\_{\nu,n}^2 - z^2)} - \\ \frac{\frac{a^2}{1+\nu} \frac{a^2 z^2}{(a^2 + z^2)^2} + 2 \sum\_{n=2}^{\infty} \frac{a^2 + j\_{\nu,n}^2}{j\_{\nu,n}^2} \frac{2z^4 [2a^2 j\_{\nu,n}^2 + z^2 (j\_{\nu,n}^2 - a^2)]}{(a^2 + z^2)^2 (j\_{\nu,n}^2 - z^2)^2}}{1 - \frac{a^2}{2(1+\nu)} \frac{z^2}{a^2 + z^2} - 2 \sum\_{n=2}^{\infty} \frac{a^2 + j\_{\nu,n}^2}{j\_{\nu,n}^2} \frac{z^4}{(a^2 + z^2)(j\_{\nu,n}^2 - z^2)}}. \end{split} \tag{20}$$
It is proven in Theorem 1 [2] that the radius of starlikeness, *r* ∗ *gν* , for the function *g<sup>ν</sup>* is the smallest root of the equation
$$\begin{aligned} 1 + \frac{a^2}{2(1+\nu)} \frac{r^2}{a^2 - r^2} \\ -2 \sum\_{n=2}^\infty \frac{a^2 + j\_{\nu,n}^2}{j\_{\nu,n}^2} \frac{r^4}{(a^2 - r^2)(j\_{\nu,n}^2 + r^2)} = ir \frac{g\_\nu'(ir)}{g\_\nu(ir)} = 0, \end{aligned}$$
in the interval (0, *a*). Thus, we have
∗
).
$$0 < r\_{\mathcal{g}\_{\mathcal{V}}}^{\*} < a < j\_{\nu,2} < j\_{\nu,3} < \dots < j\_{\nu,\mathcal{U}} < \dots < \dots \tag{21}$$
Taking into account that *ν* + 1 < 0, the equality (20) implies the following inequality:
$$\begin{split} \text{Re}\left(1+\frac{z\mathbf{g}\_{\nu}^{\prime\prime}(z)}{\mathbf{g}\_{\nu}^{\prime}(z)}\right) &\geq \\ 1+\left.\frac{a^{2}}{2(1+\nu)}\right|\frac{z^{2}}{a^{2}+z^{2}}\left|-2\sum\_{n=2}^{\infty}\frac{a^{2}+j\_{\nu,n}^{2}}{j\_{\nu,n}^{2}}\right|\frac{z^{4}}{(a^{2}+z^{2})(j\_{\nu,n}^{2}-z^{2})}\right|-\text{(22)} \\ \leq \frac{-\frac{a^{2}}{1+\nu}\left|\frac{a^{2}z^{2}}{(a^{2}+z^{2})^{2}}\right|+2\sum\_{n=2}^{\infty}\frac{a^{2}+j\_{\nu,n}^{2}}{j\_{\nu,n}^{2}}\left|\frac{2z^{4}[2a^{2}j\_{\nu,n}^{2}+z^{2}(j\_{\nu,n}^{2}-a^{2})]}{(a^{2}+z^{2})(j\_{\nu,n}^{2}-z^{2})^{2}}\right|}{1+\frac{a^{2}}{2(1+\nu)}\left|\frac{z^{2}}{a^{2}+z^{2}}\right|-2\sum\_{n=2}^{\infty}\frac{a^{2}+j\_{\nu,n}^{2}}{j\_{\nu,n}^{2}}\left|\frac{z^{4}}{(a^{2}+z^{2})(j\_{\nu,n}^{2}-z^{2})}\right|} \end{split} \tag{22}$$
for every *z* ∈ *U*(*r*
*ν* Using *δ* = *a* 2 , *ρ* = *r* <sup>2</sup> and *v* = *z* 2 in Lemma 4, we obtain
$$\frac{a^2}{2(1+\nu)} \left| \frac{z^2}{a^2 + z^2} \right| \ge \frac{a^2}{2(1+\nu)} \frac{r^2}{a^2 - r^2} \text{ and} \frac{a^2}{2(1+\nu)} \left| \frac{z^2}{(a^2 + z^2)^2} \right| \ge \tag{23}$$
$$\frac{a^2}{(a^2 - r^2)^2} \frac{r^2}{2(1+\nu)}.$$
In a similar manner, Lemma 5 and Lemma 6 imply that
$$\begin{array}{c|c} \left| \frac{z^4}{(a^2+z^2)(j\_{\nu,n}^2-z^2)} \right| \le \frac{r^4}{(a^2-r^2)(j\_{\nu,n}^2+r^2)}\\ \left| \frac{2z^4[2a^2j\_{\nu,n}^2+z^2(j\_{\nu,n}^2-a^2)]}{(a^2+z^2)^2(j\_{\nu,n}^2-z^2)^2} \right| \le \frac{2r^4[2a^2j\_{\nu,n}^2-r^2(j\_{\nu,n}^2-a^2)]}{(a^2-r^2)^2(j\_{\nu,n}^2+r^2)^2}. \end{array} \tag{24}$$
Now, inequalities (22)–(24) imply the following inequality:
$$\begin{split} \text{Re}\left(1+\frac{zg\_{\nu}^{\mu}(z)}{g\_{\nu}^{\nu}(z)}\right) &\geq \\ 1+\frac{a^{2}}{2(1+\nu)}\frac{r^{2}}{a^{2}-r^{2}}-2\sum\_{n=2}^{\infty}\frac{a^{2}+j\_{\nu,n}^{2}}{j\_{\nu,n}^{2}}\frac{r^{4}}{(a^{2}-r^{2})(j\_{\nu,n}^{2}+r^{2})}-\\ -\frac{a^{2}}{1+\nu}\frac{a^{2}r^{2}}{(a^{2}-r^{2})^{2}}+2\sum\_{n=2}^{\infty}\frac{a^{2}+j\_{\nu,n}^{2}}{j\_{\nu,n}^{2}}\frac{2r^{4}[2r^{2}\_{\mu,n}-r^{2}(j\_{\nu,n}^{2}-a^{2})]}{(a^{2}-r^{2})^{2}(j\_{\nu,n}^{2}+r^{2})^{2}} &= 1+\frac{irg\_{\nu}^{\mu\prime}(ir)}{g\_{\nu}^{\prime}(ir)}=\Phi(r),\end{split} \tag{25}$$
provided that *a* > *r* ∗ *<sup>g</sup><sup>ν</sup>* > |*z*|, where *r* ∗ *gν* verifies the inequalities (21).
The following equalities hold: <sup>Φ</sup>(0) = 1 and lim*r*%*<sup>r</sup>* ∗ *gν* Φ(*r*) = −∞. Consequently, equation 1 + *irg*00 *ν* (*ir*) *g* 0 *<sup>ν</sup>*(*ir*) = *α* has a real root in the interval (0,*r* ∗ *gν* ). The smallest positive real root of the equation 1 + *irg*00 *ν* (*ir*) *g* 0 *<sup>ν</sup>*(*ir*) = *α* is denoted by *r c gν* (*α*), and this root is the radius of convexity of order *α* of the function *gν*. The first equality of Lemma 7 and the equality *Jν*(*iz*) = *i ν Iν*(*z*) imply that the equation 1 + *irg*00 *ν* (*ir*) *g* 0 *<sup>ν</sup>*(*ir*) = *α* is equivalent to (19).
We determine the radius of uniform convexity of the mapping *g<sup>ν</sup>* in the next theorem.
**Theorem 2.** *If ν* ∈ (−2, −1), *then the radius of uniform convexity for the mapping g<sup>ν</sup> is r* ∗ *ν* (*α*) = *r*2, *where r*<sup>2</sup> *is the smallest positive root of the equation*
$$\frac{1}{2} + r \frac{I\_{\nu+2}(r) + 3I\_{\nu+1}(r)}{I\_{\nu+1}(r) + rI\_{\nu}(r)} = 0 \tag{26}$$
*in the interval* (0,*r* ∗ *ν* ).
**Proof.** Equality (20) implies the following inequality:
$$\begin{array}{c|c} \left| \frac{zg\_{\nu}^{\mu}(z)}{g\_{\nu}^{\mu}(z)} \right| \leq \\ -\frac{a^{2}}{2(1+\nu)} \left| \frac{z^{2}}{a^{2}+z^{2}} \right| + 2\sum\_{n=2}^{\infty} \frac{a^{2}+j\_{\nu,n}^{2}}{j\_{\nu,n}^{2}} \left| \frac{z^{4}}{(a^{2}+z^{2})(j\_{\nu,n}^{2}-z^{2})} \right| + \\ -\frac{a^{2}}{1+\nu} \left| \frac{a^{2}z^{2}}{(a^{2}+z^{2})^{2}} \right| + 2\sum\_{n=2}^{\infty} \frac{a^{2}+j\_{\nu,n}^{2}}{j\_{\nu,n}^{2}} \left| \frac{2z^{4}[2a^{2}j\_{\nu,n}^{2}+z^{2}(j\_{\nu,n}^{2}-a^{2})]}{(a^{2}+z^{2})^{2}(j\_{\nu,n}^{2}-z^{2})^{2}} \right| \\ \phantom{\left| \frac{a^{2}}{1+\frac{a^{2}}{2(1+\nu)}} \right|} \frac{z^{2}}{a^{2}+z^{2}} \left| -2\sum\_{n=2}^{\infty} \frac{a^{2}+j\_{\nu,n}^{2}}{j\_{\nu,n}^{2}} \right| \frac{z^{4}}{(a^{2}+z^{2})(j\_{\nu,n}^{2}-z^{2})} \end{array} . \tag{27}$$
We can again use inequalities (22) and (23), and in combination with (27), we have
$$\begin{split} \left| \frac{zg\_{\nu}^{\prime\prime}(z)}{g\_{\nu}^{\prime}(z)} \right| &\leq -\frac{a^{2}}{2(1+\nu)} \frac{r^{2}}{a^{2}-r^{2}} + 2\sum\_{n=2}^{\infty} \frac{a^{2}+j\_{\nu,n}^{2}}{j\_{\nu,n}^{2}} \frac{r^{4}}{(a^{2}-r^{2})(j\_{\nu,n}^{2}+r^{2})} + 1 \\ &\frac{-\frac{a^{2}}{1+\nu}\frac{a^{2}r^{2}}{(a^{2}-r^{2})^{2}} + 2\sum\_{n=2}^{\infty} \frac{a^{2}+j\_{\nu,n}^{2}}{j\_{\nu,n}^{2}} \frac{2r^{4}[2a^{2}j\_{\nu,n}^{2}-r^{2}(j\_{\nu,n}^{2}-a^{2})]}{(a^{2}-r^{2})^{2}(j\_{\nu,n}^{2}+r^{2})^{2}} = -\frac{i\mathbf{r}g\_{\nu}^{\prime\prime}(ir)}{g\_{\nu}^{\prime\prime}(ir)}. \end{split}$$
Inequalities (25) and (27) imply
$$\left| \mathrm{Re} \left( 1 + \frac{z g\_{\nu}^{\prime\prime}(z)}{g\_{\nu}^{\prime}(z)} \right) - \left| \frac{z g\_{\nu}^{\prime\prime}(z)}{g\_{\nu}^{\prime}(z)} \right| \geq 1 + 2 \frac{i r g\_{\nu}^{\prime\prime}(ir)}{g\_{\nu}^{\prime}(ir)}, z \in \mathsf{U}(r\_{\nu}^{\*}). \tag{28}$$
The smallest positive root of the equation 1 + 2 *irg*00 *ν* (*ir*) *g* 0 *<sup>ν</sup>*(*ir*) = 0 in the interval (0,*r* ∗ *ν* ) is denoted by *r uc ν* . According to (28), the value *r uc ν* is the biggest with the property that
$$\left| \mathrm{Re} \left( 1 + \frac{z \mathcal{g}\_{\nu}^{\prime\prime}(z)}{\mathcal{g}\_{\nu}^{\prime}(z)} \right) - \left| \frac{z \mathcal{g}\_{\nu}^{\prime\prime}(z)}{\mathcal{g}\_{\nu}^{\prime}(z)} \right| > 0, \; z \in \mathcal{U}(r\_{\nu}^{\mu c}). \right.$$
Lemma 7 and the equality *Jν*(*iz*) = *i ν Iν*(*z*) imply that the equation 1 + 2 *irg*00 *ν* (*ir*) *g* 0 *<sup>ν</sup>*(*ir*) = 0 is equivalent to (26), completing the proof.
Theorems 1 and 2 imply the following result.
**Corollary 1.** *The mapping g<sup>ν</sup> is uniformly convex in the disk U*(*r*) *if and only if it is convex of order* <sup>1</sup> 2 .
**Theorem 3.** *If α* ∈ [0, 1) *and ν* ∈ (−2, −1), *then the radius of convexity of order α for the mapping h<sup>ν</sup> is r<sup>c</sup> hν* (*α*) = *r*3, *where r*<sup>3</sup> *is the smallest real root of the equation*
$$1 + \frac{rI\_{\nu+2}(r^{\frac{1}{2}}) + 4r^{\frac{1}{2}}I\_{\nu+1}(r^{\frac{1}{2}})}{4I\_{\nu}(r^{\frac{1}{2}}) + 2r^{\frac{1}{2}}I\_{\nu+1}(r^{\frac{1}{2}})} = a \tag{29}$$
*in the interval* (0,*r* ∗ *hν* ). **Proof.** According to the proof of Theorem 2 [2], the equalities
$$\frac{zh\_{\nu}'(z)}{h\_{\nu}(z)} = 1 - \sum\_{n=1}^{\infty} \frac{z}{j\_{\nu,n}^2 - z}, \quad \sum\_{n=1}^{\infty} \frac{1}{j\_{\nu,n}^2} = \frac{1}{4(\nu+1)}$$
imply
$$\frac{zh\_{\nu}'(z)}{h\_{\nu}(z)} = 1 - \frac{a^2}{4(\nu+1)} \cdot \frac{z}{a^2+z} - \sum\_{n=2}^{\infty} \frac{a^2 + f\_{\nu,n}^2}{f\_{\nu,n}^2} \cdot \frac{z^2}{(a^2+z)\left(f\_{\nu,n}^2 - z\right)}.$$
where *z* ∈ *U*(0,*r*).
The logarithmic differentiation of the equality leads to
$$1 + \frac{zh\_{\nu}^{\prime\prime}(z)}{h\_{\nu}^{\prime}(z)} = 1 - \frac{a^2}{4(\nu+1)} \cdot \frac{z}{a^2 + z} - \sum\_{n=2}^{\infty} \frac{a^2 + j\_{\nu,n}^2}{j\_{\nu,n}^2} \cdot \frac{z^2}{(a^2 + z)\left(j\_{\nu,n}^2 - z\right)} -$$
$$-\frac{\frac{a^2}{4(1+\nu)} \cdot \frac{a^2 z}{(a^2 + z)^2} + \sum\_{n=2}^{\infty} \frac{a^2 + j\_{\nu,n}^2}{j\_{\nu,n}^2} \cdot \frac{z^2 \left[2a^2 j\_{\nu,n}^2 + z \left(j\_{\nu,n}^2 - a^2\right)\right]}{\left(j\_{\nu,n}^2 - z\right)^2 \left(a^2 + z\right)^2}}{1 - \frac{a^2}{4a^2 + z} \cdot \frac{z}{(\nu+1)} - \sum\_{n=2}^{\infty} \frac{a^2 + j\_{\nu,n}^2}{j\_{\nu,n}^2} \cdot \frac{z^2}{\left(j\_{\nu,n}^2 - z\right) \left(a^2 + z\right)}}. \tag{30}$$
It is proven in [2] that the radius of starlikeness, *r* ∗ *hν* , for function *h<sup>ν</sup>* is the smallest root of the equation
$$\frac{-rh\_\nu'(-r)}{h\_\nu(-r)} = 0, \ r \in \left(0, a^2\right), \ z \in \mathcal{U}(0, r).$$
However,
$$\frac{-rh\_{\nu}^{\prime}(-r)}{h\_{\nu}(-r)} = 1 + \frac{a^2}{4(\nu+1)} \cdot \frac{r}{a^2 - r}$$
$$-\sum\_{n=2}^{\infty} \frac{a^2 + j\_{\nu,n}^2}{j\_{\nu,n}^2} \cdot \frac{r^2}{(a^2 - r)\left(j\_{\nu,n}^2 + r\right)} = 0, \ r \in \left(0, a^2\right).$$
Taking into the account that *ν* + 1 < 0, we obtain from relation (30)
$$\begin{split} \mathrm{Re}\left(1+\frac{zh\_{\nu}^{\mu\prime}(z)}{h\_{\nu}^{\prime}(z)}\right) &\geq \frac{a^2}{4(\nu+1)} \cdot \left|\frac{z}{a^2+z}\right| - \sum\_{n=2}^{\infty} \frac{a^2+j\_{\nu,n}^2}{j\_{\nu,n}^2} \cdot \left|\frac{z^2}{(a^2+z)\left(j\_{\nu,n}^2-z\right)}\right| - \\ &\quad -\frac{\frac{-a^2}{4(\nu+1)} \cdot \left|\frac{a^2z}{(a^2+z)^2}\right| + \sum\_{n=2}^{\infty} \frac{a^2+j\_{\nu,n}^2}{j\_{\nu,n}^2} \cdot \left|\frac{z^2\left|2a^2j\_{\nu,n}^2+z\left(j\_{\nu,n}^2-a^2\right)\right|}{(a^2+z)^2\left(j\_{\nu,n}^2-z\right)^2}\right|}{1+\frac{a^2}{4(\nu+1)} \cdot \left|\frac{z}{a^2+z}\right| - \sum\_{n} \frac{a^2+j\_{\nu,n}^2}{j\_{\nu,n}^2} \cdot \left|\frac{z^2}{(a^2+z)\left(j\_{\nu,n}^2-z\right)}\right|}\end{split} \tag{31}$$
and *z* ∈ *U*(0,*r*),*r* ∈ 0,*r* ∗ *hν* . We obtain from Lemmas 4 and 5 the following inequality:
$$\mathrm{Re}\left(1+\frac{zh\_{\nu}^{\prime\prime}(z)}{h\_{\nu}^{\prime}(z)}\right) \ge 1 + \frac{a^2}{4(\nu+1)} \cdot \frac{r}{a^2-r} - \sum\_{n=2}^{\infty} \frac{a^2+j\_{\nu,n}^2}{j\_{\nu,n}^2} \cdot \frac{r^2}{(a^2-r)(a^2+r)} - \cdots$$
$$-\frac{-\frac{a^2}{4(\nu+1)} \cdot \frac{a^2r}{(a^2-r)^2} + \sum\_{n=2}^{\infty} \frac{a^2+j\_{\nu,n}^2}{j\_{\nu,n}^2} \cdot \frac{r^2[2a^2j\_{\nu,n}^2 - r(j\_{\nu,n}^2 - a^2)]}{(a^2-r)^2(j\_{\nu,n}^2 + r)^2}}{1 + \frac{a^2}{4(\nu+1)} \cdot \frac{r}{a^2-r} - \sum\_{n=2}^{\infty} \frac{a^2+j\_{\nu,n}^2}{j\_{\nu,n}^2} \cdot \frac{r^2}{(a^2-r)(\frac{j\_{\nu,n}^2}{j\_{\nu,n}^2 - r})}} = \tag{32}$$
$$= 1 - \frac{r h\_{\nu}^{\prime\prime}(-r)}{h\_{\nu}^{\prime}(-r)} = \psi(r), \ a > r\_{h\_{\nu}}^{\ast}>|z|\_{\prime}$$
similarly to the proof of Theorem 1. The mapping
$$\psi : \left( 0, r\_{h\_{\mathcal{V}}}^{\*} \right) \to \mathbb{R}, \; \psi(r) = 1 + \frac{-rh\_{\mathcal{V}}^{\prime\prime}(-r)}{h\_{\mathcal{V}}^{\prime}(-r)} \lambda$$
is strictly decreasing, and *a* > *r* ∗ *hν* > |*z*|.
We then have lim *r*%*r* ∗ *ψ*(*r*) = −∞, *ψ*(0) = 1, and the equation
$$1 + \frac{-r h\_{\nu}^{\prime\prime}(-r)}{h\_{\nu}^{\prime}(-r)} = \alpha$$
has at least one real root in the interval 0,*r* ∗ *hν* .
*hν*
The smallest positive real root of the equation 1 − *rh*00 *ν* (−*r*) *h* 0 *<sup>ν</sup>*(−*r*) = *α* is denoted by *r c hν* (*α*), and this root is the radius of convexity of order *α* of the function *hν*. The second equality of Lemma 7 and the equality *Jν*(*iz*) = *i ν Iν*(*z*) imply that the equation 1 − *rh*00 *ν* (−*r*) *h* 0 *<sup>ν</sup>*(−*r*) = *α* is equivalent to (29).
**Theorem 4.** *If α* ∈ [0, 1) *and ν* ∈ (−2, −1), *then the radius of uniform convexity of h<sup>ν</sup> is r* ∗ *hν* (*α*) = *r*4, *where r*<sup>4</sup> *is the smallest positive root of the equation*
$$\frac{rI\_{\nu+2}(r^{\frac{1}{2}}) + 4r^{\frac{1}{2}}I\_{\nu+1}(r^{\frac{1}{2}})}{4I\_{\nu}(r^{\frac{1}{2}}) + 2r^{\frac{1}{2}}I\_{\nu+1}(r^{\frac{1}{2}})} = \frac{1}{2} \tag{33}$$
*in the interval* (0,*r* ∗ *hν* ).
>
**Proof.** Equality (30) implies the following inequality:
$$\begin{split} \left| \frac{zh\_{\nu}^{\mu\prime}(z)}{h\_{\nu}^{\prime}(z)} \right| &\leq -\frac{a^{2}}{4(\nu+1)} \cdot \left| \frac{z}{a^{2}+z} \right| + \sum\_{n=2}^{\infty} \frac{a^{2} + j\_{\nu,n}^{2}}{j\_{\nu,n}^{2}} \cdot \left| \frac{z^{2}}{(a^{2}+z)\left(j\_{\nu,n}^{2}-z\right)} \right| + \\ &+ \frac{\frac{-a^{2}}{4(\nu+1)} \cdot \left| \frac{a^{2}z}{(a^{2}+z)^{2}} \right| + \sum\_{n=2}^{\infty} \frac{a^{2} + j\_{\nu,n}^{2}}{j\_{\nu,n}^{2}} \cdot \left| \frac{z^{2}\left[2a^{2}j\_{\nu,n}^{2}+z\left(j\_{\nu,n}^{2}-a^{2}\right)\right]}{(a^{2}+z)^{2}\left(j\_{\nu,n}^{2}-z\right)^{2}} \right|} \right|. \end{split} \tag{34}$$
We obtain the following from the relation (31), Lemma 4, and the relation (34):
$$\left|\frac{zh\_{\nu}^{\prime\prime}(z)}{h\_{\nu}^{\prime}(z)}\right| \leq -\frac{a^2}{4(\nu+1)} \cdot \frac{r}{a^2 - r} + \sum\_{n=2}^{\infty} \frac{a^2 + \hat{f}\_{\nu,n}^2}{\hat{f}\_{\nu,n}^2} \cdot \frac{r^2}{(a^2 - r)\left(\hat{f}\_{\nu,n}^2 + r\right)} +$$
$$+\frac{\frac{a^2}{4(\nu+1)} \cdot \frac{a^2r}{(a^2 - r)^2} + \sum\_{n=2}^{\infty} \frac{a^2 + \hat{f}\_{\nu,n}^2}{\hat{f}\_{\nu,n}^2} \cdot \frac{r^2 \left[2a^2\hat{f}\_{\nu,n}^2 - r\left(\hat{f}\_{\nu,n}^2 - a^2\right)\right]}{\left(a^2 - r\right)^2 \left(\hat{f}\_{\nu,n}^2 + r\right)^2}}{1 - \frac{a^2}{4(a^2 - r)} \cdot \frac{r}{\nu+1} - \sum\_{n=2}^{\infty} \frac{a^2 + \hat{f}\_{\nu,n}^2}{\hat{f}\_{\nu,n}^2} \cdot \frac{r^2}{(a^2 - r)\left(\hat{f}\_{\nu,n}^2 + r\right)}} = \frac{r h\_{\nu}^{\prime\prime}(-r)}{h\_{\nu}^{\prime}(-r)}, |z| \leq r < a^2.$$
Inequalities (32) and (34) imply
$$\left| \operatorname{Re} \left( 1 + \frac{z h\_{\nu}^{\prime\prime}(z)}{h\_{\nu}^{\prime}(z)} \right) - \left| \frac{z h\_{\nu}^{\prime\prime}(z)}{h\_{\nu}^{\prime}(z)} \right| \geq 1 - \frac{2 r h\_{\nu}^{\prime\prime}(-r)}{h\_{\nu}^{\prime}(-r)}, \; z \in \mathcal{U}(r\_{h\_{\nu}}^{\*}). \tag{35}$$
The smallest positive root of the equation 1 − 2*rh*00 *ν* (−*r*) *h* 0 *<sup>ν</sup>*(−*r*) = 0 in the interval (0,*r* ∗ *hν* ) is denoted by *r uc hν* .
According to (35), the value *r uc hν* is the biggest with the property that
$$\left| \mathrm{Re} \left( 1 + \frac{z h\_{\nu}^{\prime\prime}(z)}{h\_{\nu}^{\prime}(z)} \right) - \left| \frac{z h\_{\nu}^{\prime\prime}(z)}{h\_{\nu}^{\prime}(z)} \right| > 0, \; z \in \mathcal{U}(r\_{h\_{\nu}}^{uc}).$$
The equation 1 − 2*rh*00 *ν* (−*r*) *h* 0 *<sup>ν</sup>*(−*r*) = 0 is equivalent to (33), completing the proof. Lemma 7 and the equality *Jν*(*iz*) = *i ν Iν*(*z*) imply that the equation 1 − 2*rh*00 *ν* (−*r*) *h* 0 *<sup>ν</sup>*(−*r*) = 0 is equivalent to (33).
From Theorems 3 and 4, we obtain the following corollary.
**Corollary 2.** *The function h<sup>ν</sup> is uniformly convex in the disk U*(*r*) *if and only if it is convex of order* <sup>1</sup> 2 .
**Author Contributions:** Conceptualization, L.-I.C., R.S. and P.A.K.; methodology, R.S.; software, L.- I.C., R.S. and P.A.K.; validation, R.S.; formal analysis, L.-I.C., R.S. and P.A.K.; investigation, L.-I.C., R.S. and P.A.K.; resources, L.-I.C., R.S. and P.A.K.; data curation, L.-I.C., R.S. and P.A.K.; writing—original draft preparation, L.-I.C., R.S. and P.A.K.; writing—review and editing, L.-I.C., R.S. and P.A.K.; visualization, L.-I.C., R.S. and P.A.K.; supervision, R.S.; project administration, L.-I.C., R.S. and P.A.K.; funding acquisition, L.-I.C., R.S. and P.A.K. All authors have read and agreed to the published version of the manuscript.
**Funding:** This research received no external funding.
**Institutional Review Board Statement:** Not applicable.
**Informed Consent Statement:** Not applicable.
**Data Availability Statement:** Not applicable.
**Acknowledgments:** The authors would like to express their sincere thanks to the referees for their valuable suggestions.
**Conflicts of Interest:** The authors declare no conflict of interest.
#### **References**
### *Article* **Cauchy Integral and Boundary Value for Vector-Valued Tempered Distributions**
**Richard D. Carmichael**
Department of Mathematics, Wake Forest University, Winston-Salem, NC 27109, USA; [email protected]
**Abstract:** Using the historically general growth condition on scalar-valued analytic functions, which have tempered distributions as boundary values, we show that vector-valued analytic functions in tubes *T <sup>C</sup>* = R*<sup>n</sup>* + *iC* obtain vector-valued tempered distributions as boundary values. In a certain vector-valued case, we study the structure of this boundary value, which is shown to be the Fourier transform of the distributional derivative of a vector-valued continuous function of polynomial growth. A set of vector-valued functions used to show the structure of the boundary value is shown to have a one–one and onto relationship with a set of vector-valued distributions, which generalize the Schwartz space D<sup>0</sup> *L* <sup>2</sup> (R*<sup>n</sup>* ); the tempered distribution Fourier transform defines the relationship between these two sets. By combining the previously stated results, we obtain a Cauchy integral representation of the vector-valued analytic functions in terms of the boundary value.
**Keywords:** analytic functions; vector-valued tempered distributions; boundary value; Cauchy integral
**MSC:** 32A26; 32A40; 46F12; 46F20
#### **1. Introduction**
Tillmann [1] introduced the analysis of analytic functions, which obtain tempered distributional boundary values in S 0 (R*<sup>n</sup>* ). In [1], Tillmann worked with scalar-valued analytic functions in tubes *T <sup>C</sup><sup>µ</sup>* <sup>=</sup> <sup>R</sup>*<sup>n</sup>* <sup>+</sup> *iCµ*, where the *<sup>C</sup><sup>µ</sup>* <sup>=</sup> {*<sup>y</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* : (−1) *<sup>µ</sup>jy<sup>j</sup>* > 0, *j* = 1, . . . , *n*} with *µ* = (*µ*1, *µ*2, . . . , *µn*) being any of the 2 *<sup>n</sup>* n-tuples, whose components are either 0 or 1 and characterize the growth conditions on the analytic functions, which obtain the S 0 (R*<sup>n</sup>* ) boundary values. This analysis by Tillmann was motivated by the work by K*o*¨the in [2,3].
Using a more restrictive growth on the analytic functions, we showed in [4] that vectorvalued analytic functions in tubes *T <sup>C</sup>* = R*<sup>n</sup>* + *iC*, where *C* is an open convex cone, having this more restrictive growth obtain vector-valued tempered distributions in S 0 (R*<sup>n</sup>* , X ), with X being a specified topological vector space. In this paper, our first objective is to generalize this result of [4] to the general growth form of Tillmann for the vector-valued analytic functions. We obtain this boundary value generalization in Section 4 of this paper.
Moreover, in Section 4, we study the structure of this boundary value in S 0 (R*<sup>n</sup>* , X ). To do this, we first restrict the topological vector space X by imposing certain conditions on it to ensure that the boundary value is the Fourier transform of a distributional derivative of a continuous vector-valued function **g**, which has polynomial growth in the norm of the space X . By further restricting X to be a Hilbert space, we show that function **g** is in a set that has a one–one and onto relationship with a set of vector-valued distributions, which generalize the D<sup>0</sup> *L* 2 (R*<sup>n</sup>* ) distributions of Schwartz. The relationship between these two sets is obtained using the tempered distribution Fourier transform; the proof of this relationship is proved in Section 3 of this paper. Using the relationships of these noted two sets, we are able to obtain an additional structure of the tempered distribution boundary value of the analytic functions in Section 4.
A few papers have been written concerning the construction of a Cauchy integral for tempered distributions. All of these papers concern scalar-valued analytic functions
**Citation:** Carmichael, R.D. Cauchy Integral and Boundary Value for Vector-Valued Tempered Distributions. *Axioms* **2022**, *11*, 392. https://doi.org/10.3390/ axioms11080392
Academic Editor: Georgia Irina Oros
Received: 12 July 2022 Accepted: 7 August 2022 Published: 10 August 2022
**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.
**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).
and scalar-valued tempered distributions. The first paper known to this author was by J. Sebastião e Silva [5] (Section 5) and concerned scalar-valued analytic functions and tempered distributions in one dimension. An associated analysis by Sebastião e Silva is contained in [6]. Carmichael [7] defined a Cauchy integral for tempered distributions in the C*n* setting corresponding to analytic functions in each of the 2 *<sup>n</sup>* quadrant tubes *T <sup>C</sup><sup>µ</sup>* <sup>⊂</sup> <sup>C</sup>*<sup>n</sup>* and showed that the analytic functions with growth, such as that of Tillmann in (C − R) *n* could be recovered as the defined Cauchy integral of the tempered distribution boundary value; the results of [7] can be extended to the vector-valued analytic functions in *T <sup>C</sup><sup>µ</sup>* and the tempered distribution setting considered in this paper by the same techniques as those of [7]. The Cauchy integrals introduced by Sebastião e Silva in [5] and by Carmichael in [7] are in fact equivalence classes of analytic functions defined by an integral involving the Cauchy kernel.
Vladimirov [8–10] defined a Cauchy integral for tempered distributions associated with analytic functions in general tubes *T <sup>C</sup>* <sup>=</sup> <sup>R</sup>*<sup>n</sup>* <sup>+</sup> *iC* <sup>⊂</sup> <sup>C</sup>*<sup>n</sup>* corresponding to open convex cones *C* with the functions satisfying a growth condition similar to that of Tillmann. Vladimirov has shown that the analytic functions that he has considered can be recovered by a Cauchy integral involving the tempered distribution boundary values of the analytic functions. An associated analysis by Vladimirov is contained in [11,12]. The works mentioned in this paragraph and the previous paragraph all concern scalar-valued analytic functions and scalar-valued tempered distributions.
In Section 5 of this paper, we build on our analysis of Sections 3 and 4 to obtain a Cauchy integral representation of the vector-valued analytic functions, which are shown to have tempered vector-valued distributions as the boundary values in Section 4. The proof of our result here and the form of the Cauchy integral representation are substantially different from any of the previous results concerning Cauchy integral representation of the analytic functions having tempered distribution boundary values.
#### **2. Definitions and Notation**
Throughout, X will denote a topological vector space with the stated appropriate properties corresponding to the results that we wish to prove. For X being a normed space, we denote the norm by N . Θ will denote the zero element of X ; and if X is a Hilbert space, we denote the space by H. For integration of the vector-valued functions and vector-valued analytic functions, we refer to Dunford and Schwartz [13]. For foundational information concerning vector-valued distributions, we refer to Schwartz [14,15].
The n-dimensional notation to be used in this paper will be the same as in [16,17]. Note 0 = (0, 0, . . . , 0) is the origin in R*<sup>n</sup>* . The information concerning cones *<sup>C</sup>* <sup>⊂</sup> <sup>R</sup>*<sup>n</sup>* needed is explicitly stated in [16] (Section 2) and [17] (Chapter 1). We do not repeat the definitions and notations concerning cones as stated in [16] (Section 2), and we ask the reader to refer to this reference.
The *L p* (R*<sup>n</sup>* , X ) functions, 1 ≤ *p* ≤ ∞, with values in a Banach space X and their norm |**h**|*<sup>p</sup>* [13] (p. 119) are noted in [13] (Chapter III). The Fourier transform on *L* 1 (R*<sup>n</sup>* ) or *L* 1 (R*<sup>n</sup>* , X ) is given in [17] (p. 3). All Fourier (inverse Fourier) transforms on scalar or vectorvalued functions will be denoted *<sup>φ</sup>*ˆ(*x*) = <sup>F</sup>[*φ*(*t*); *<sup>x</sup>*] (<sup>F</sup> <sup>−</sup><sup>1</sup> [*φ*(*t*); *x*]). As stated in [18,19], the Plancherel theory is not true for vector-valued functions, except when X = H, a Hilbert space. The Plancherel theory is complete in the *L* 2 (R*<sup>n</sup>* , H) setting in that the inverse Fourier transform is the inverse mapping of the Fourier transform with <sup>F</sup> <sup>−</sup>1<sup>F</sup> <sup>=</sup> *<sup>I</sup>* <sup>=</sup> F F <sup>−</sup><sup>1</sup> with *<sup>I</sup>* being the identity mapping.
We denote <sup>S</sup>(R*<sup>n</sup>* ) as the tempered functions with associated distributions being S 0 (R*<sup>n</sup>* ) or associated vector-valued distributions being S 0 (R*<sup>n</sup>* , X ). The Fourier (inverse Fourier) transform on S 0 (R*<sup>n</sup>* ) and S 0 (R*<sup>n</sup>* , X ) is the usual definition and is given in [14] (p. 73).
#### **3. Fourier and Inverse Fourier Transform on a Function Subset of** S 0 (R*<sup>n</sup>* **,** H)
Let X be a Banach space. We defined the space S 0 *p* (R*<sup>n</sup>* , X ), 1 ≤ *p* < ∞, in [16]. We repeat the definition here because of the importance of these functions for our results in this paper.
**Definition 1.** *For a Banach space* X , S 0 *p* (R*<sup>n</sup>* , X ), 1 ≤ *p* < ∞, *is the set of all measurable functions <sup>g</sup>*(*t*), *<sup>t</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup> , with values in* X *such that there exists a real number m* ≥ 0 *for which* (1 + |*t*| *p* ) <sup>−</sup>*mg*(*t*) <sup>∈</sup> *<sup>L</sup> p* (R*<sup>n</sup>* , X )*.*
Note that *m* can be taken as a nonnegative integer in Definition 1. As noted in [16], S 0 *p* (R*<sup>n</sup>* , X ) ⊂ S<sup>0</sup> (R*<sup>n</sup>* , X ), 1 ≤ *p* < ∞. The spaces S 0 *p* (R*<sup>n</sup>* , X ) will be important in this paper. Throughout this paper, the differential operator *D<sup>t</sup>* , *<sup>t</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* will take the form
$$D\_{t} = \frac{-1}{2\pi i} \left( \frac{\partial}{\partial t\_{1}}, \frac{\partial}{\partial t\_{2}}, \dots, \frac{\partial}{\partial t\_{n}} \right) \dots$$
Thus, for *α* being any n-tuple of nonnegative integers,
$$D\_t^\alpha = \begin{pmatrix} \frac{-1}{2\pi i} \end{pmatrix}^{|\alpha|} \begin{pmatrix} \frac{\partial^{\alpha\_1}}{\partial t\_1^{\alpha\_1}}, \frac{\partial^{\alpha\_2}}{\partial t\_2^{\alpha\_2}}, \dots, \frac{\partial^{\alpha\_{\text{II}}}}{\partial t\_n^{\text{III}}} \end{pmatrix}.$$
The goal of this section is to show a one–one and onto relationship between the set of functions S 0 2 (R*<sup>n</sup>* , H) and another subset of S 0 (R*<sup>n</sup>* , H), where H is a Hilbert space. This relationship is obtained by both the Fourier and inverse Fourier transforms in S 0 (R*<sup>n</sup>* , H). We define the space that has this stated relationship to S 0 2 (R*<sup>n</sup>* , H), as follows.
**Definition 2.** *Let m be any nonnegative integer. The set of Hilbert space* H*-valued generalized functions in* S 0 (R*<sup>n</sup>* , H) *of the form*
$$V\_t = \sum\_{|\alpha| \le m} D\_t^{\alpha} \mathbf{g}\_{\alpha}(t)$$
*where g<sup>α</sup>* ∈ *L* 2 (R*<sup>n</sup>* , <sup>H</sup>), <sup>|</sup>*α*| ≤ *m, will be denoted as L*2(R*<sup>n</sup>* , H)*.*
We emphasize that *L*2(R*<sup>n</sup>* , H) ⊂ S<sup>0</sup> (R*<sup>n</sup>* , <sup>H</sup>). When <sup>H</sup> <sup>=</sup> <sup>C</sup><sup>1</sup> , note that *L*2(R*<sup>n</sup>* , C<sup>1</sup> ) = D0 *L* 2 (R*<sup>n</sup>* ), the Schwartz space of distributions contained in S 0 (R*<sup>n</sup>* ) of the form of finite sums of distributional derivatives of *L* 2 (R*<sup>n</sup>* ) functions. For *φ* ∈ D*<sup>L</sup>* <sup>2</sup> (R*<sup>n</sup>* ), the Schwartz space that is the set of test functions for D<sup>0</sup> *L* 2 (R*<sup>n</sup>* ), the application h*V*, *φ*i, *V* ∈ D<sup>0</sup> *L* 2 (R*<sup>n</sup>* ), yields a complex number. In exactly the same way, for *<sup>V</sup>* <sup>∈</sup> *<sup>L</sup>*2(R*<sup>n</sup>* , H) and *φ* ∈ D*<sup>L</sup>* <sup>2</sup> (R*<sup>n</sup>* ), the application h*V*, *φ*i yields an element of H; and the algebraic and differentiation calculations on the form <sup>h</sup>*V*, *<sup>φ</sup>*<sup>i</sup> hold for *<sup>V</sup>* <sup>∈</sup> *<sup>L</sup>*2(R*<sup>n</sup>* , H), as usual, just as these calculations hold on the form h*V*, *φ*i for *V* ∈ S<sup>0</sup> (R*<sup>n</sup>* , <sup>H</sup>) and *<sup>φ</sup>* ∈ S(R*<sup>n</sup>* ). This is an important note in relation to our construction of the Cauchy integral (later in this paper).
We now obtain the relationship between S 0 2 (R*<sup>n</sup>* , <sup>H</sup>) and *<sup>L</sup>*2(R*<sup>n</sup>* , H) for any Hilbert space H.
**Lemma 1.** *The* S 0 (R*<sup>n</sup>* , H) *Fourier transform maps* S 0 2 (R*<sup>n</sup>* , <sup>H</sup>) *one-one and onto <sup>L</sup>*2(R*<sup>n</sup>* , H)*. The* S 0 (R*<sup>n</sup>* , <sup>H</sup>) *inverse Fourier transform maps L*2(R*<sup>n</sup>* , H) *one-one and onto* S 0 2 (R*<sup>n</sup>* , H)*.*
**Proof.** Let the function **g** ∈ S<sup>0</sup> 2 (R*<sup>n</sup>* , H). From Definition 1, there is a real number *m* ≥ 0 for which (1 + |*t*| 2 ) <sup>−</sup>*m***g**(*t*) <sup>∈</sup> *<sup>L</sup>* 2 (R*<sup>n</sup>* , H), and *m* can be taken as a nonnegative integer. Since **g** ∈ S<sup>0</sup> 2 (R*<sup>n</sup>* , H) ⊂ S<sup>0</sup> (R*<sup>n</sup>* , H), the Fourier transform of **g** in S 0 (R*<sup>n</sup>* , H) is an element of S 0 (R*<sup>n</sup>* , <sup>H</sup>); we put *<sup>V</sup><sup>x</sup>* <sup>=</sup> <sup>F</sup>[**g**]*x*. Let *<sup>φ</sup>* ∈ S(R*<sup>n</sup>* ), and let ∆ denote the Laplace operator in the variable *<sup>x</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* . Using integration by parts, we have
$$\begin{split} \langle V\_{\mathbf{x}'}\boldsymbol{\phi}(\mathbf{x})\rangle &= \langle \mathbf{g}(t), \mathcal{F}[\boldsymbol{\phi}(\mathbf{x}); t] \rangle \\ &= \langle \frac{\mathbf{g}(t)}{(1+|t|^2)^m}, \int\_{\mathbb{R}^n} \boldsymbol{\phi}(\mathbf{x})(1+|t|^2)^m e^{2\pi i \langle \mathbf{x}, t\rangle} d\mathbf{x} \rangle \\ &= \langle \frac{\mathbf{g}(t)}{(1+|t|^2)^m}, \mathcal{F}[(1-(4\pi^2)^{-1}\Delta)^m \boldsymbol{\phi}(\mathbf{x}); t] \rangle \\ &= \langle \mathcal{F}[\frac{\mathbf{g}(t)}{(1+|t|^2)^m}; \mathbf{x}], (1-(4\pi^2)^{-1}\Delta)^m \boldsymbol{\phi}(\mathbf{x}) \rangle. \end{split} \tag{1}$$
Since (1 + |*t*| 2 ) <sup>−</sup>*m***g**(*t*) <sup>∈</sup> *<sup>L</sup>* 2 (R*<sup>n</sup>* , H), then **h**(*x*) = F[(1 + |*t*| 2 ) <sup>−</sup>*m***g**(*t*); *<sup>x</sup>*] <sup>∈</sup> *<sup>L</sup>* 2 (R*<sup>n</sup>* , H). From (1), we have
$$
\langle V\_{\mathbf{x}} \varphi(\mathbf{x}) \rangle = \langle (1 - (4\pi^2)^{-1} \Delta)^m \mathbf{h}(\mathbf{x}), \phi(\mathbf{x}) \rangle,
$$
and *V<sup>x</sup>* = F[**g**]*<sup>x</sup>* = (1 − (4*π* 2 ) <sup>−</sup>1∆) *<sup>m</sup>***h**(*x*) <sup>∈</sup> *<sup>L</sup>*2(R*<sup>n</sup>* , H). Thus, the S 0 (R*<sup>n</sup>* , H) Fourier transform maps S 0 2 (R*<sup>n</sup>* , <sup>H</sup>) to *<sup>L</sup>*2(R*<sup>n</sup>* , H).
We now desire to prove that any element of *L*2(R*<sup>n</sup>* , H) is the S 0 (R*<sup>n</sup>* , H) Fourier transform of an element in S 0 2 (R*<sup>n</sup>* , <sup>H</sup>). Let *<sup>V</sup>* <sup>∈</sup> *<sup>L</sup>*2(R*<sup>n</sup>* , <sup>H</sup>) and *<sup>φ</sup>* ∈ S(R*<sup>n</sup>* ). By Definition 2, there is a nonnegative integer *m*, such that
$$V\_t = \sum\_{|\alpha| \le m} D\_t^{\alpha} \mathbf{g}\_{\alpha}(t)$$
with **g***<sup>α</sup>* (*t*) ∈ *L* 2 (R*<sup>n</sup>* , <sup>H</sup>), <sup>|</sup>*α*| ≤ *<sup>m</sup>*. Since *<sup>L</sup>*2(R*<sup>n</sup>* , H) ⊂ S<sup>0</sup> (R*<sup>n</sup>* , <sup>H</sup>), <sup>F</sup> <sup>−</sup><sup>1</sup> [*V*]*<sup>x</sup>* exists in S 0 (R*<sup>n</sup>* , H), and we have for the nonnegative integer *m*
hF <sup>−</sup><sup>1</sup> [*V*]*x*, *<sup>φ</sup>*(*x*)i = ∑ |*α*|≤*m* h*D α <sup>t</sup>* **g***<sup>α</sup>* (*t*), F −1 [*φ*(*x*); *t*]i = ∑ |*α*|≤*m* (−1) |*α*| h**g***<sup>α</sup>* (*t*), *D α t* Z R*n φ*(*x*)*e* −2*πi*h*x*,*t*i *dx*i = ∑ |*α*|≤*m* (−1) |*α*| h**g***<sup>α</sup>* (*t*),(−1/2*πi*) |*α*| Z R*n φ*(*x*)(−2*πi*) |*α*| *x α e* −2*πi*h*x*,*t*i *dx*i = ∑ |*α*|≤*m* h(−1) <sup>|</sup>*α*|**g***<sup>α</sup>* (*t*), Z R*n x <sup>α</sup>φ*(*x*)*e* −2*πi*h*x*,*t*i *dx*i = ∑ |*α*|≤*m* h(−1) <sup>|</sup>*α*|**g***<sup>α</sup>* (*t*), F −1 [*x <sup>α</sup>φ*(*x*); *<sup>t</sup>*]<sup>i</sup> = ∑ |*α*|≤*m* hF <sup>−</sup><sup>1</sup> [(−1) <sup>|</sup>*α*|**g***<sup>α</sup>* (*t*); *x*], *x <sup>α</sup>φ*(*x*)i.
For each *<sup>α</sup>*, <sup>|</sup>*α*| ≤ *<sup>m</sup>*, put **<sup>h</sup>***α*(*x*) = <sup>F</sup> <sup>−</sup><sup>1</sup> [(−1) <sup>|</sup>*α*|**g***<sup>α</sup>* (*t*); *x*]. We have **h***α*(*x*) ∈ *L* 2 (R*<sup>n</sup>* , H), |*α*| ≤ *m*, since each **g***<sup>α</sup>* (*t*) ∈ *L* 2 (R*<sup>n</sup>* , H); moreover, <sup>∑</sup>|*α*|≤*<sup>m</sup>* **<sup>h</sup>***α*(*x*) ∈ *<sup>L</sup>* 2 (R*<sup>n</sup>* , H). Thus, we have
$$
\langle \mathcal{F}^{-1}[V]\_{x'} \phi(x) \rangle = \sum\_{|a| \le m} \langle \mathbf{h}\_a(\mathbf{x}), \mathbf{x}^a \phi(\mathbf{x}) \rangle
$$
$$
= \langle \sum\_{|a| \le m} \mathbf{x}^a \mathbf{h}\_a(\mathbf{x}), \phi(\mathbf{x}) \rangle,
\tag{2}
$$
and <sup>F</sup> <sup>−</sup><sup>1</sup> [*V*]*<sup>x</sup>* = <sup>∑</sup>|*α*|≤*<sup>m</sup> <sup>x</sup> <sup>α</sup>***h***α*(*x*) in <sup>S</sup> 0 (R*<sup>n</sup>* , H). For the *L* 2 (R*<sup>n</sup>* , H) norm | · |<sup>2</sup> and the order *m* of the summation defining *V*, we consider
$$|(1+|\boldsymbol{x}|^{2})^{-m-2}\sum\_{|\boldsymbol{a}|\leq m}\boldsymbol{x}^{\boldsymbol{a}}\mathbf{h}\_{\boldsymbol{a}}(\boldsymbol{x})|\_{2}.\tag{3}$$
For |*α*| ≤ *m*, note that |*x α* | ≤ |*x*| <sup>|</sup>*α*<sup>|</sup> <sup>≤</sup> (<sup>1</sup> <sup>+</sup> <sup>|</sup>*x*|) <sup>|</sup>*α*<sup>|</sup> <sup>≤</sup> (<sup>1</sup> <sup>+</sup> <sup>|</sup>*x*|) *<sup>m</sup>*. Since (<sup>1</sup> <sup>+</sup> <sup>|</sup>*x*|) *<sup>m</sup>* <sup>≤</sup> <sup>2</sup> *<sup>m</sup>* if |*x*| ≤ 1 and (1 + |*x*|) *<sup>m</sup>* <sup>≤</sup> (<sup>1</sup> <sup>+</sup> <sup>|</sup>*x*<sup>|</sup> 2 ) *<sup>m</sup>* if <sup>|</sup>*x*| ≥ 1, then
$$\begin{aligned} &|\mathfrak{x}^{\mathfrak{a}}(1+|\mathfrak{x}|^{2})^{-m-2}| \leq (1+|\mathfrak{x}|)^{m}(1+|\mathfrak{x}|^{2})^{-m-2} \\ &\leq \max\_{\mathfrak{x}\in\mathbb{R}^{n}}\{2^{m},(1+|\mathfrak{x}|^{2})^{m}\}(1+|\mathfrak{x}|^{2})^{-m-2} \\ &\leq \max\{2^{m},1\}=2^{m} \end{aligned}$$
for |*α*| ≤ *m* since *m* ≥ 0 is a nonnegative integer. Thus, for the *L* 2 (R*<sup>n</sup>* , H) norm in (3), we have
$$\left| (1+|\boldsymbol{x}|^{2})^{-m-2} \sum\_{|\boldsymbol{a}| \leq m} \boldsymbol{\pi}^{\boldsymbol{a}} \mathbf{h}\_{\boldsymbol{a}}(\boldsymbol{\pi}) \right|\_{2} \leq 2^{m} \left| \sum\_{|\boldsymbol{a}| \leq m} \mathbf{h}\_{\boldsymbol{a}}(\boldsymbol{\pi}) \right|\_{2} < \infty \tag{4}$$
since <sup>∑</sup>|*α*|≤*<sup>m</sup>* **<sup>h</sup>***α*(*x*) ∈ *<sup>L</sup>* 2 (R*<sup>n</sup>* , <sup>H</sup>). Recalling (2), we have by (4) that <sup>F</sup> <sup>−</sup><sup>1</sup> [*V*]*<sup>x</sup>* = <sup>∑</sup>|*α*|≤*<sup>m</sup> <sup>x</sup> <sup>α</sup>***h***α*(*x*) <sup>∈</sup> S 0 2 (R*<sup>n</sup>* , <sup>H</sup>) for any *<sup>V</sup>* <sup>∈</sup> *<sup>L</sup>*2(R*<sup>n</sup>* , H); and *<sup>V</sup><sup>t</sup>* = F[∑|*α*|≤*<sup>m</sup> <sup>x</sup> <sup>α</sup>***h***α*(*x*)]*<sup>t</sup>* in S 0 (R*<sup>n</sup>* , H). Thus, the S 0 (R*<sup>n</sup>* , H) Fourier transform maps S 0 2 (R*<sup>n</sup>* , <sup>H</sup>) onto *<sup>L</sup>*2(R*<sup>n</sup>* , H); the fact that this mapping is one–one follows directly from the fact that the Fourier transform is a one–one mapping on S 0 (R*<sup>n</sup>* , H). The same statements and proofs as in this proof of Lemma 1 for the Fourier transform hold in exactly the same way for the inverse Fourier transform on S 0 (R*<sup>n</sup>* , H); and we have that the S 0 (R*<sup>n</sup>* , <sup>H</sup>) inverse Fourier transform maps *<sup>L</sup>*2(R*<sup>n</sup>* , H) one–one and onto S 0 2 (R*<sup>n</sup>* , H). The proof of Lemma 1 is complete.
Let *C* be a regular cone in R*<sup>n</sup>* ; that is, *C* is an open convex cone in R*<sup>n</sup>* , which does not contain any entire straight line. *C* <sup>∗</sup> <sup>=</sup> {*<sup>t</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* : h*t*, *y*i ≥ 0 for all *y* ∈ *C*} is the dual cone of *C*. We consider now the Cauchy kernel
$$K(z - t) = \int\_{\mathbb{C}^\*} e^{2\pi i \langle z - t, \mu \rangle} du, \; z \in T^{\mathbb{C}} = \mathbb{R}^n + i\mathbb{C}, \; t \in \mathbb{R}^n.$$
The ultradistributional test function spaces D(∗, *L p* ) ⊂ D*<sup>L</sup> <sup>p</sup>* (R*<sup>n</sup>* ), 1 < *p* ≤ ∞, where ∗ is Beurling (*Mp*) or Roumieu {*Mp*}, defined in [17] (Section 2.3, p. 21). For *C* being a regular cone, we proved in [17] (Section 4.1, Theorem 4.1.1) that *K*(*z* − ·) ∈ D(∗, *L p* ) ⊂ D*<sup>L</sup> <sup>p</sup>* (R*<sup>n</sup>* ) for *z* ∈ *T <sup>C</sup>*, 1 <sup>&</sup>lt; *<sup>p</sup>* <sup>≤</sup> <sup>∞</sup> , under specified conditions on the sequence *<sup>M</sup><sup>p</sup>* of positive numbers, which we assume here. (See [17] (pp. 13–14, Theorem 4.1.1) for assumptions on the sequence *Mp*.) The Schwartz space D<sup>0</sup> *L* 2 (R*<sup>n</sup>* ) consists of finite sums of distributional derivatives of *L* 2 (R*<sup>n</sup>* ) functions; thus, the space *L*2(R*<sup>n</sup>* , H|) is the extension of D<sup>0</sup> *L* 2 (R*<sup>n</sup>* ) to vector-valued distributions with values in H. Thus, for *p* = 2, we emphasize that the form h*Vt* , *K*(*z* − *t*)i, *z* ∈ *T <sup>C</sup>*, is well defined for *<sup>V</sup>* <sup>∈</sup> *<sup>L</sup>*2(R*<sup>n</sup>* , H), and yields an element of H; the algebraic and differentiation calculations on the form <sup>h</sup>*V*, *<sup>φ</sup>*<sup>i</sup> hold for *<sup>V</sup>* <sup>∈</sup> *<sup>L</sup>*2(R*<sup>n</sup>* , H) and *φ* ∈ D*<sup>L</sup>* <sup>2</sup> (R*<sup>n</sup>* ), as usual, just as these calculations hold for the form h*V*, *φ*i for *V* ∈ S<sup>0</sup> (R*<sup>n</sup>* , H) and *<sup>φ</sup>* ∈ S(R*<sup>n</sup>* ). We use this information in Section 5 of this paper.
#### **4. Boundary Values in** S 0 (R*<sup>n</sup>* **,** X )
Let *C* be an open convex cone in R*<sup>n</sup>* . In [4] (Theorem 8), we proved that an analytic function **f**(*z*), *z* ∈ *T <sup>C</sup>*, with values in a specified topological vector space <sup>X</sup> and satisfying a certain norm growth obtained a vector-value-tempered distributional boundary value, as *y* → 0, *y* ∈ *C* <sup>0</sup> ⊂⊂ *C*, for any compact subcone *C* 0 of *C*. The norm growth used in [4] (Theorem 8) was not as general as the growth of Tillmann [1] in which the original tempered distributional boundary value results in the scalar-valued case were obtained. In this section, we extend the result [4] (Theorem 8) by assuming a norm growth on the analytic function equivalent to that of Tillmann [1]; our result here also contains new information concerning the boundary value. As a corollary of our result, we obtain a precise representation of the boundary value when the conditions on the topological vector space X are restricted.
Following Vladimirov [11] (p. 230), we shall use the term "spectral function" but will extend the definition of this term to the vector-valued case. For an analytic function **f**(*z*), *z* ∈ *T <sup>C</sup>* <sup>=</sup> <sup>R</sup>*<sup>n</sup>* <sup>+</sup> *iC* <sup>⊂</sup> <sup>C</sup>*<sup>n</sup>* , with values in a topological vector space X , the spectral function of **f**(*z*) is that vector-valued distribution *V* ∈ D<sup>0</sup> (R*<sup>n</sup>* , X ), such that *e* <sup>−</sup>2*π*h*y*,*t*i*V<sup>t</sup>* <sup>∈</sup> S 0 (R*<sup>n</sup>* , X ), *y* ∈ *C*; and **f**(*x* + *iy*) = F[*e* <sup>−</sup>2*π*h*y*,*t*i*V<sup>t</sup>* ]*<sup>x</sup>* in S 0 (R*<sup>n</sup>* , X ) for *z* = *x* + *iy* ∈ *T C*.
We begin by assuming that the topological vector space X is locally convex, separable, and quasi-complete where quasi-complete is in the sense of Schwartz [15] (p. 198). We further assume that X is a normed space with norm N . These stated assumptions on X were the assumptions under which we obtained [4] (Theorem 8) and are the assumptions on the topological vector space X under which we obtain Theorem 1 below.
Throughout the paper, by *y* → 0, *y* ∈ *C*, we mean that *y* → 0, *y* ∈ *C* <sup>0</sup> ⊂⊂ *C* for every compact subcone *C* <sup>0</sup> ⊂⊂ *C*.
The following theorem generalizes and extends [4] (Theorem 8) for X , satisfying the properties noted above.
**Theorem 1.** *Let C be an open convex cone. Let <sup>f</sup>*(*z*) *be analytic in T<sup>C</sup> and have values in* <sup>X</sup> *. Let*
$$\mathcal{N}(f(\mathbf{x} + i\mathbf{y})) \le M(1 + |z|)^q |y|^{-r}, \; z = \mathbf{x} + i\mathbf{y} \in T^{\mathbb{C}}.\tag{5}$$
*where M* > 0 *is a real constant, q is a nonnegative integer, r* > 1 *is an integer, and M*, *q, and r are independent of z* = *x* + *iy* ∈ *T <sup>C</sup>. There exists an element U* ∈ S<sup>0</sup> (R*<sup>n</sup>* , X )*, such that*
$$\lim\_{y \to \overline{0}, y \in \mathbb{C}} f(x + iy) = \mathcal{U} \tag{6}$$
*in the weak and strong topologies of* S 0 (R*<sup>n</sup>* , X )*. Further, U* = F[*V*] *with V* ∈ S<sup>0</sup> (R*<sup>n</sup>* , X ) *being the spectral function of f*(*z*), *z* ∈ *T <sup>C</sup>, , such that supp*(*V*) <sup>⊆</sup> *<sup>C</sup>* ∗ *.*
**Proof.** We apply the proofs of [4] (Theorems 3 and 8). Note that in the second sentence of the proof of [4] (Theorem 8) that the value of *η* ≥ 1 is arbitrary but fixed; in the present proof, we simply take *η* = 1, where it is appropriate to use *η* = 1. Let *λ* > 0; put *<sup>ρ</sup>* <sup>=</sup> *<sup>σ</sup>* <sup>+</sup> *<sup>i</sup>λ*, *<sup>σ</sup>* <sup>∈</sup> <sup>R</sup><sup>1</sup> ; and define **f** 0 (*ρ*; *x*, *y*) = **f**(*x* + *ρy*), *y* ∈ *pr*(*C*), where *pr*(*C*) denotes the projection of *C*, which is the intersection of *C* with the unit sphere in R*<sup>n</sup>* . (Thus, |*y*| = 1 if *y* ∈ *pr*(*C*).) **f** 0 (*ρ*; *x*, *y*) is an analytic function of *ρ* in the half plane *λ* = Im(*ρ*) > 0 and has values in X . We have **f** 0 (*ρ*; *x*, *y*) = **f**(*x* + *ρy*) = **f**((*x* + *σy*) + *iλy*), *λ* > 0, for *z* = *x* + *iy* with *y* ∈ *pr*(*C*); and note that *λy* ∈ *C* for *λ* > 0 and *y* ∈ *pr*(*C*). Now for *y* = Im(*z*) ∈ *pr*(*C*) and 0 < *λ* ≤ *η* = 1 we have
$$\begin{split} &\mathcal{N}(\mathbf{f}(\boldsymbol{\varrho};\boldsymbol{x},\boldsymbol{y})) \leq &\mathcal{M}(1+|(\boldsymbol{x}+\sigma\boldsymbol{y})+i\lambda\boldsymbol{y}|)^{q}|\lambda\boldsymbol{y}|^{-r} \\ &=\mathcal{M}(1+(\lambda^{2}+|\boldsymbol{x}+\sigma\boldsymbol{y}|^{2})^{1/2})^{q}\lambda^{-r} \\ &\leq &\mathcal{M}(1+(1+(|\boldsymbol{x}|+|\sigma|)^{2})^{1/2})^{q}\lambda^{-r} \\ &\leq &\mathcal{M}(1+((1+|\boldsymbol{x}|+|\sigma|)^{2})^{1/2})^{q}\lambda^{-r} \\ &=\mathcal{M}(2+|\boldsymbol{x}|+|\sigma|)^{q}\lambda^{-r} \end{split} \tag{7}$$
which is of the form, with norm N replacing the absolute value, of [4] (15), which is used in exactly the same way in the proof of [4] (Theorem 8) as in the proof of [4] (Theorem 3). Thus, for *y* = Im(*z*) ∈ *pr*(*C*) and 0 < *λ* ≤ *η* = 1 the bound on N (**f** 0 (*ρ*; *x*, *y*)) is in the proper form to proceed with the proof of this present Theorem 1 exactly as in the form of the proofs of [4] (Theorems 3 and 8). We obtain the structured function of the form Λ(−*r*−1) **f** 0 (*ρ*; *x*, *y*), *y* ∈ *pr*(*C*), which satisfies the growth (similar to [4] (37))
$$\mathcal{N}(\Lambda^{(-r-1)}\mathbf{f}'(\rho; \mathbf{x}, y)) \le M^{(r+1)}(2 + |\mathbf{x}| + |\sigma|)^q (2 + |\sigma|)^{r+1}$$
for <sup>0</sup> <sup>&</sup>lt; *<sup>λ</sup>* <sup>≤</sup> *<sup>η</sup>* <sup>=</sup> <sup>1</sup> where *<sup>M</sup>*(*r*+1) is a positive constant, and obtains the representation (similar to [4] (38))
$$\mathbf{f}(\mathbf{x} + \rho y) = \mathbf{f}'(\rho; \mathbf{x}, y) = \frac{\partial^{r+1}(\Lambda^{(-r-1)}\mathbf{f}'(\rho; \mathbf{x}, y))}{\partial \sigma^{r+1}}, \; \sigma = \text{Re}(\rho).$$
Now, we proceed in our proof of Theorem 1 in exactly the same way as in [4] (Theorem 8) (p. 328) to obtain the desired boundedness properties leading to the existence of an element *V* ∈ D<sup>0</sup> (R*<sup>n</sup>* , X ), such that *e* <sup>−</sup>2*π*h*y*,*t*i*V<sup>t</sup>* ∈ S<sup>0</sup> (R*<sup>n</sup>* , X ), *y* ∈ *C*, and **f**(*z*) = F[*e* <sup>−</sup>2*π*h*y*,*t*i*V<sup>t</sup>* ]*x*, *z* = *x* + *iy* ∈ *T <sup>C</sup>* , in <sup>S</sup> 0 (R*<sup>n</sup>* , X ) from the results of Schwartz [14] (Prop. 22, p. 76). (These results of Schwartz [14] (Prop. 22, p. 76) were obtained in their original scalar-valued case in [20]; the related results were then obtained by Lions [21]). Thus, *V* ∈ D<sup>0</sup> (R*<sup>n</sup>* , X ) is the spectral function of **f**(*z*), *z* ∈ *T <sup>C</sup>* . The remainder of the proof of [4] (Theorem 8, pp. 329–330) and the succeeding discussion after the conclusion of the proof of [4] (Theorem 8) can be applied to the present proof of Theorem 1 in the same way to yield that, in fact, *V* ∈ S<sup>0</sup> (R*<sup>n</sup>* , X ) and that
$$\lim\_{y \to \overline{0}, y \in \mathbb{C}} \mathbf{f}(\mathbf{x} + iy) = \lim\_{y \to \overline{0}, y \in \mathbb{C}} \mathcal{F}[\mathfrak{e}^{-2\pi \langle y, t \rangle} V\_t] = \mathcal{F}[V] = \mathcal{U} \tag{8}$$
in the weak topology of S 0 (R*<sup>n</sup>* , <sup>X</sup> ). However, <sup>S</sup>(R*<sup>n</sup>* ) is a Montel space; thus, the convergence in (8) is in the strong topology of S 0 (R*<sup>n</sup>* , X ) as well. We emphasize that *V* ∈ S 0 (R*<sup>n</sup>* , X ) and that *U* = F[*V*] ∈ S<sup>0</sup> (R*<sup>n</sup>* , X ) is the desired boundary value in (6) as obtained in (8).
We now prove that supp(*V*) ⊆ *C* ∗ . Let *<sup>t</sup><sup>o</sup>* <sup>∈</sup> *<sup>C</sup>*<sup>∗</sup> <sup>=</sup> <sup>R</sup>*<sup>n</sup>* \ *<sup>C</sup>* ∗ ; *C*<sup>∗</sup> is an open set in R*<sup>n</sup>* since *C* ∗ is a closed set. From the definition of *C* ∗ , for *t<sup>o</sup>* ∈ *C*∗, there is a point *y<sup>o</sup>* ∈ *C*, such that h*yo*, *to*i < 0. Using the fact that *C*<sup>∗</sup> is open and the continuity of h*t*, *yo*i at *t<sup>o</sup>* ∈ *C*<sup>∗</sup> as a function of *t*, there is a fixed *τ* > 0 and a fixed neighborhood *<sup>N</sup>*(*to*; *<sup>γ</sup>*) = {*<sup>t</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* : |*t* − *to*| < *γ*, *γ* > 0} ⊂ *C*∗, such that h*t*, *yo*i < −*τ* < 0 for all *<sup>t</sup>* <sup>∈</sup> *<sup>N</sup>*(*to*; *<sup>γ</sup>*). Let *<sup>φ</sup>* ∈ D(R*<sup>n</sup>* ), such that supp(*φ*) ⊂ *N*(*to*, *γ*). Recall that *V* ∈ S<sup>0</sup> (R*<sup>n</sup>* , X ), such that *e* <sup>−</sup>2*π*h*y*,*t*i*V<sup>t</sup>* ∈ S<sup>0</sup> (R*<sup>n</sup>* , X ), *y* ∈ *C*, and **f**(*x* + *iy*) = F[*e* <sup>−</sup>2*π*h*y*,*t*i*V<sup>t</sup>* ]*x*, *z* = *x* + *iy* ∈ *T C*, in S 0 (R*<sup>n</sup>* , X ). Thus *e* <sup>−</sup>2*π*h*y*,*t*i*V<sup>t</sup>* <sup>=</sup> <sup>F</sup> <sup>−</sup><sup>1</sup> [**f**(*x* + *iy*)]*<sup>t</sup>* , *x* + *iy* ∈ *T <sup>C</sup>*, in <sup>S</sup> 0 (R*<sup>n</sup>* , X ); or *V<sup>t</sup>* = *e* <sup>2</sup>*π*h*y*,*t*i<sup>F</sup> <sup>−</sup><sup>1</sup> [**(***x* + *iy*)]*<sup>t</sup>* , *x* + *iy* ∈ *T <sup>C</sup>*, in <sup>S</sup> 0 (R*<sup>n</sup>* , X ). Let *y* = *βyo*, *y<sup>o</sup>* ∈ *C*, *β* > 0, now. We have *y* = *βy<sup>o</sup>* ∈ *C* and
$$\begin{split} \langle V, \Phi \rangle &= \langle e^{2\pi \langle \beta y\_o, t \rangle} \mathcal{F}^{-1} [\mathbf{f}(x + i \beta y\_o)\_{t \prime} \phi(t)] \\ &= \langle \mathcal{F}^{-1} [\mathbf{f}(x + i \beta y\_o)\_{t \prime} e^{2\pi \langle \beta y\_o, t \rangle} \phi(t)] \\ &= \langle \mathbf{f}(x + i \beta y\_o)\_{t \prime} \mathcal{F}^{-1} [e^{2\pi \langle \beta y\_o, t \rangle} \phi(t); x] \rangle \\ &= \int\_{\mathbb{R}^n} \mathbf{f}(x + i \beta y\_o) \int\_{supp(\phi)} e^{2\pi \langle \beta y\_o, t \rangle} \phi(t) e^{-2\pi i \langle x, t \rangle} dt d\mathbf{x} \end{split} \tag{9}$$
for the function *<sup>φ</sup>* ∈ D(R*<sup>n</sup>* ) chosen above. Using integration by parts and letting ∆ denote the Laplacian in the *<sup>t</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* variable, we have for any positive integer *<sup>m</sup>*
$$\mathcal{N}\left(\int\_{\mathbb{R}^n} \mathbf{f}(\mathbf{x} + i\beta y\_o) \int\_{\operatorname{supp}(\phi)} e^{2\pi \langle \beta y\_o, t \rangle} \phi(t) e^{-2\pi i \langle \mathbf{x}, t \rangle} dt d\mathbf{x}\right) \tag{10}$$
$$\begin{split} \mathcal{N} &= \mathcal{N}\left(\int\_{\mathbb{R}^n} \frac{\mathbf{f}(\mathbf{x} + i\beta y\_o)}{(1 + |\mathbf{x}|^2)^m} \int\_{\operatorname{supp}(\phi)} e^{2\pi \langle \beta y\_o, t \rangle} \phi(t) (1 + |\mathbf{x}|^2)^m e^{-2\pi i \langle \mathbf{x}, t \rangle} dt d\mathbf{x}\right) \\ &= \mathcal{N}\left(\int\_{\mathbb{R}^n} \frac{\mathbf{f}(\mathbf{x} + i\beta y\_o)}{(1 + |\mathbf{x}|^2)^m} \int\_{\operatorname{supp}(\phi)} (1 - \frac{\Delta}{4\pi^2})^m (e^{2\pi \langle \beta y\_o, t \rangle} \phi(t)) e^{-2\pi i \langle \mathbf{x}, t \rangle} dt d\mathbf{x}\right). \end{split}$$
(For the present, the positive integer *m* is arbitrary; later, we explicitly choose *m* to obtain the desired convergence of all integrals through Equation (15) below). For the interior integral over supp(*φ*) in (10), we note that by applying (1 − (∆/4*π* 2 ))*<sup>m</sup>* to the product *e* <sup>2</sup>*π*h*βyo*,*t*i*φ*(*t*) and then bounding the terms in the resulting sum, including the terms involving 2*π* or it powers, we obtain a finite sum of terms involving powers of *β*(*yo*)*<sup>j</sup>* , *j* = 1, . . . , *n*, multiplied by *e* 2*π*h*βyo*,*t*i , where (*yo*)*<sup>j</sup>* is the *j*th component of *yo*, *j* = 1, . . . , *n*, and multiplied by bounds on *φ*(*t*) or one of its partial derivatives with *e* 2*π*h*βyo*,*t*i in each term of the sum. Of course, the boundedness of *φ*(*t*) and any of its partial derivatives are valid because of the compact support of *φ*(*t*). Moreover, note that |*β*(*yo*)*<sup>j</sup>* | ≤ *β*|*yo*|, *j* = 1, . . . , *n*. Thus, since the interior
integral in (10) is over supp(*φ*) ⊂ *N*(*to*; *γ*), we obtain the following bound on this interior integral:
$$\left| \int\_{\operatorname{supp}(\phi)} (1 - \frac{\Delta}{4\pi^2})^m (e^{2\pi \langle \beta y\_o, t \rangle} \phi(t)) e^{-2\pi i \langle \mathbf{x}, t \rangle} dt \right| $$
$$\leq \int\_{\operatorname{supp}(\phi)} |(1 - \frac{\Delta}{4\pi^2})^m (e^{2\pi \langle \beta y\_o, t \rangle} \phi(t))| dt \tag{11}$$
$$\leq T\_{\operatorname{supp}(\phi)} (1 + \beta |y\_o|)^{4(m+1)} \sup\_{t \in \operatorname{supp}(\phi)} e^{2\pi \langle \beta y\_o, t \rangle}$$
where *Tsupp*(*φ*) is a positive constant depending only on supp(*φ*). Using (11) in (10), we have
$$\mathcal{N}\left(\int\_{\mathbb{R}^n} \frac{\mathbf{f}(\mathbf{x} + i\beta y\_o)}{(1 + |\mathbf{x}|^2)^m} \int\_{\operatorname{supp}(\phi)} (1 - \frac{\Delta}{4\pi^2})^m (e^{2\pi \langle \beta y\_o, t\rangle} \phi(t)) e^{-2\pi i \langle \mathbf{x}, t\rangle} dt d\mathbf{x}\right)$$
$$\leq T\_{\operatorname{supp}(\phi)} (1 + \beta |y\_o|)^{4(m+1)} \sup\_{t \in \operatorname{supp}(\phi)} e^{2\pi \langle \beta y\_o, t\rangle} \int\_{\mathbb{R}^n} \frac{N(\mathbf{f}(\mathbf{x} + i\beta y\_o))}{(1 + |\mathbf{x}|^2)^m} d\mathbf{x}\tag{12}$$
where *y<sup>o</sup>* ∈ *C*, *β* > 0 is arbitrary, and supp(*φ*) ⊂ *N*(*to*; *γ*) ⊂ *C*∗, *t<sup>o</sup>* ∈ *C*∗, *γ* > 0 and fixed. As noted before, since h*yo*, *to*i < 0 and *C*<sup>∗</sup> is open, by the continuity of h*t*, *yo*i at *t<sup>o</sup>* ∈ *C*<sup>∗</sup> as a function of *<sup>t</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* , the fixed *τ* > 0 is chosen and the fixed *N*(*to*; *γ*) ⊂ *C*<sup>∗</sup> is chosen, such that h*t*, *yo*i < −*τ* < 0 for all *t* ∈ *N*(*to*; *γ*) ⊂ *C*∗. Since supp(*φ*) ⊂ *N*(*to*; *γ*), we have
$$\sup\_{t \in \sup p(\phi)} e^{2\pi \langle \beta y\_\circ, t \rangle} \le e^{-2\pi \tau \beta} \rho$$
which yields from (12)
$$\mathcal{N}\left(\int\_{\mathbb{R}^n} \frac{\mathbf{f}(\mathbf{x} + i\beta y\_o)}{(1 + |\mathbf{x}|^2)^m} \int\_{\operatorname{supp}(\phi)} (1 - \frac{\Delta}{4\pi^2})^m (e^{2\pi \langle \beta y\_o, t \rangle} \phi(t)) e^{-2\pi i \langle \mathbf{x}, t \rangle} dt d\mathbf{x}\right)$$
$$\leq T\_{\operatorname{supp}(\phi)} e^{-2\pi \tau \beta} (1 + \beta |y\_o|)^{4(m+1)} \int\_{\mathbb{R}^n} \frac{\mathcal{N}(\mathbf{f}(\mathbf{x} + i\beta y\_o))}{(1 + |\mathbf{x}|^2)^m} d\mathbf{x}\tag{13}$$
where *y<sup>o</sup>* ∈ *C*, *τ* > 0, and *γ* > 0 are fixed and are independent of the arbitrary *β* > 0. We now bound the integral on the right of the inequality in (13) using the assumed growth (5) on **f**(*z*), *z* ∈ *T <sup>C</sup>*; (13) holds for all *β* > 0. To obtain the supp(*V*) containment result, we are going to let *β* → ∞ in (13); thus, we may assume that *β* > 1 in the remainder of this proof. By simple calculations and for *β* > 1 , we have
$$|1 + |\mathbf{x} + i\beta y\_o| = \beta(\frac{1}{\beta} + ((\frac{|\mathbf{x}|}{\beta})^2 + |y\_o|^2)^{1/2}) \le \beta(1 + (|\mathbf{x}|^2 + |y\_o|^2)^{1/2}),$$
and
$$(1 + |\mathbf{x} + i\beta y\_o|)^q \le \beta^q (1 + (|\mathbf{x}|^2 + |y\_o|^2)^{1/2})^q \le \beta^q (1 + |y\_o| + |\mathbf{x}|)^q$$
Hence, from (5),
$$\mathcal{N}(\mathbf{f}(\mathbf{x} + i\beta y\_o)) \le M\beta^q (1 + |y\_o| + |\mathbf{x}|)^q |\beta y\_o|^{-r}$$
and
$$\int\_{\mathbb{R}^n} \frac{\mathcal{N}(\mathbf{f}(\mathbf{x} + i\beta y\_o))}{(1 + |\mathbf{x}|^2)^m} d\mathbf{x} \le M \beta^{q-r} |y\_o|^{-r} \int\_{\mathbb{R}^n} \frac{(1 + |y\_o| + |\mathbf{x}|)^q}{(1 + |\mathbf{x}|^2)^m} d\mathbf{x}.\tag{14}$$
.
Combining (10), (12), (13), and (14) yields
$$\mathcal{N}\left(\int\_{\mathbb{R}^n} \mathbf{f}(\mathbf{x} + i\beta y\_o) \int\_{\operatorname{supp}(\phi)} e^{2\pi \langle \beta y\_o, t \rangle} \phi(t) e^{-2\pi i \langle \mathbf{x}, t \rangle} dt d\mathbf{x} \right) \tag{15}$$
$$\leq M T\_{\operatorname{supp}(\phi)} (1 + \beta |y\_o|)^{4(m+1)} \beta^{q-r} |y\_o|^{-r} e^{-2\pi \tau \beta} \int\_{\mathbb{R}^n} \frac{(1 + |y\_o| + |\mathbf{x}|)^q}{(1 + |\mathbf{x}|^2)^m} d\mathbf{x}.$$
The positive integer *m* in (15) was introduced in (10), and at that point in the proof, *m* was arbitrary. We now choose *m*, such that *m* > 2(*q* + *n* + 1). With this choice of *m*, the integral in (15) converges where *y<sup>o</sup>* ∈ *C* is a fixed point in *C*; further, with this choice of *m*, all calculations from (10) leading to (15) are valid and the integrals converge. Because of the exponential term *e* <sup>−</sup>2*πτβ*, where *τ* > 0 is fixed and now *β* > 1 is arbitrary, the right side of (15) has limit <sup>0</sup> as *<sup>β</sup>* <sup>→</sup> <sup>∞</sup>. Thus, from (9) <sup>h</sup>*V*, *<sup>φ</sup>*<sup>i</sup> <sup>=</sup> <sup>Θ</sup> for *<sup>φ</sup>* ∈ D(R*<sup>n</sup>* ), such that supp(*φ*) ⊂ *N*(*to*, *γ*) ⊂ *C*<sup>∗</sup> for *t<sup>o</sup>* being an arbitrary but fixed point in the open set *<sup>C</sup>*<sup>∗</sup> <sup>=</sup> <sup>R</sup>*<sup>n</sup>* \ *<sup>C</sup>* ∗ . That is, for each fixed point, *<sup>t</sup><sup>o</sup>* <sup>∈</sup> *<sup>C</sup>*<sup>∗</sup> <sup>=</sup> <sup>R</sup>*<sup>n</sup>* \ *<sup>C</sup>* ∗ , with *C*<sup>∗</sup> being an open set, there is a neighborhood *<sup>N</sup>*(*to*; *<sup>γ</sup>*) <sup>⊂</sup> *<sup>C</sup>*<sup>∗</sup> of *<sup>t</sup>o*, such that for all *<sup>φ</sup>* ∈ D(R*<sup>n</sup>* ) with supp(*φ*) ⊂ *N*(*to*; *γ*), we have h*V*, *φ*i = Θ. Thus, *V* vanishes on a neighborhood of each point of *<sup>C</sup>*∗; this proves that *<sup>V</sup>* vanishes on the open set *<sup>C</sup>*<sup>∗</sup> <sup>=</sup> <sup>R</sup>*<sup>n</sup>* \ *<sup>C</sup>* ∗ . Thus, supp(*V*) ⊆ *C* ∗ , which is a closed set in R*<sup>n</sup>* . The proof of Theorem 1 is complete.
Yoshinaga [22] (Proposition 3) provides a representation of the tempered vectorvalued distributions in the case of the topological vector space X being a complete space of type (DF). Yoshinaga's result is as follows for X , being a complete space of type (DF): *V* ∈ S<sup>0</sup> (R*<sup>n</sup>* , <sup>X</sup> ), if and only if there exists a continuous function **<sup>g</sup>** on <sup>R</sup>*<sup>n</sup>* with values in <sup>X</sup> , an integer *<sup>k</sup>* <sup>≥</sup> 0, and a n-tuple *<sup>α</sup>* of nonnegative integers, such that *<sup>V</sup>* <sup>=</sup> *<sup>D</sup>α***<sup>g</sup>** and {**g**(*t*)/(1 + |*t*| 2 ) *kn*; *<sup>t</sup>* <sup>∈</sup> <sup>R</sup>*n*} is a bounded subset of <sup>X</sup> . (In fact, in Yoshinaga's symbolism, *α* = (*k*, *k*, ..., *k*).)
The functions S 0 2 (R*<sup>n</sup>* , X ) of Definition 1 are an integral part of the following corollary to Theorem 1; recall that these functions are defined by the necessity for X being a Banach space. We know that a Banach space satisfies all of the conditions on X stated prior to Theorem 1 and also is a complete norm space of type (DF); since a Hilbert space is a Banach space, a Hilbert space also satisfies all of these stated conditions on X . Thus, the abovestated result of Yoshinaga and Theorem 1 of this paper both hold for X being a Banach or Hilbert space.
We obtain a corollary of Theorem 1 now in which more precise information is obtained concerning the spectral function *V* and the boundary value *U* of Theorem 1.
**Corollary 1.** *Let C be an open convex cone and* X *be a Banach space. Let f*(*z*) *be analytic in T <sup>C</sup>* <sup>=</sup> <sup>R</sup>*<sup>n</sup>* <sup>+</sup> *iC, have values in* <sup>X</sup> *, and satisfy* (5)*. There is a continuous function <sup>g</sup>* ∈ S<sup>0</sup> 2 (R*<sup>n</sup>* , X ) *with supp*(*g*) ⊆ *C* ∗ *a*.*e*. *and an n-tuple α of nonnegative integers, such that the spectral function V* ∈ S<sup>0</sup> (R*<sup>n</sup>* , <sup>X</sup> ) *of Theorem 1 has the form <sup>V</sup><sup>t</sup>* <sup>=</sup> *<sup>D</sup><sup>α</sup> t g*(*t*)*, and there is U* = F[*V*] ∈ S<sup>0</sup> (R*<sup>n</sup>* , X ) *such that*
$$\lim\_{y \to \overline{0}, y \in \mathbb{C}} f(x + iy) = \mathcal{U}$$
*in the weak and strong topologies of* S 0 (R*<sup>n</sup>* , X )*. Further, for* X = H *being a Hilbert space, we have* <sup>F</sup>[*g*] <sup>∈</sup> *<sup>L</sup>*2(R*<sup>n</sup>* , H)*; and the boundary value U* ∈ S<sup>0</sup> (R*<sup>n</sup>* , H) *has the form*
$$\mathcal{U}\_{\mathbf{x}} = \mathbf{x}^{a} \mathcal{F}[\mathbf{g}]\_{\mathbf{x}} = \mathbf{x}^{a} (1 - \frac{\Delta}{4\pi^{2}})^{m} h(\mathbf{x}) \tag{16}$$
*in* S 0 (R*<sup>n</sup>* , H) *where h* ∈ *L* 2 (R*<sup>n</sup>* , X ), *α is an n-tuple of nonnegative integers, and m* ≥ 0 *is a real number that can be taken to be a nonnegative integer.*
**Proof.** We apply the results of Theorem 1 and consider the spectral function *V* ∈ S<sup>0</sup> (R*<sup>n</sup>* , X ) obtained in Theorem 1 where X is a Banach space in this corollary. As per the result of Yoshinaga [22] (Proposition 3) stated above, there is a continuous function **g** on R*<sup>n</sup>*
with values in X , an n-tuple *α* of nonnegative integers, and an integer *k* ≥ 0, such that *V<sup>t</sup>* = *D<sup>α</sup> t* **g**(*t*) and { **g**(*t*) (1+|*t*| 2) *kn* ; *<sup>t</sup>* <sup>∈</sup> <sup>R</sup>*n*} is a bounded subset of <sup>X</sup> . (In Yoshinaga's symbolism, *α* is the n-tuple with all components being *k*.) Thus, there is a real constant *R* > 0, such that
$$\mathcal{N}\left(\frac{\mathbf{g}(t)}{(1+|t|^2)^{kn}}\right) = \frac{\mathcal{N}(\mathbf{g}(t))}{(1+|t|^2)^{kn}} \le \mathbb{R}, \ t \in \mathbb{R}^n.$$
For the integer *k* ≥ 0, we have
$$\begin{aligned} &\int\_{\mathbb{R}^n} (\mathcal{N}\left(\frac{\mathbf{g}(t)}{(1+|t|^2)^{(k+2)n}}\right))^2 dt \\ &= \int\_{\mathbb{R}^n} \left(\frac{1}{(1+|t|^2)^{2n}}\right)^2 (\mathcal{N}\left(\frac{\mathbf{g}(t)}{(1+|t|^2)^{kn}}\right))^2 dt \\ &\le R^2 \int\_{\mathbb{R}^n} \frac{1}{(1+|t|^2)^{4n}} dt < \infty \end{aligned}$$
which proves that **g** ∈ S<sup>0</sup> 2 (R*<sup>n</sup>* , X ). Further, supp(**g**) ⊆ *C* <sup>∗</sup> a.e. since supp(*V*) ⊆ *C* ∗ . From Theorem 1, the boundary value *U* ∈ S<sup>0</sup> (R*<sup>n</sup>* , X ) in (6) is *U* = F[*V*], the Fourier transform of the spectral function *V* ∈ S<sup>0</sup> (R*<sup>n</sup>* , X ) in S 0 (R*<sup>n</sup>* , X ). Moreover, from Theorem 1, the boundary value *U* is obtained in both the weak and strong topologies of S 0 (R*<sup>n</sup>* , X ).
Now, let X = H, a Hilbert space, in this Corollary 1. Since **g** ∈ S<sup>0</sup> 2 (R*<sup>n</sup>* , H), then <sup>F</sup>[**g**] <sup>∈</sup> *<sup>L</sup>*2(R*<sup>n</sup>* , H) in S 0 (R*<sup>n</sup>* , H) by Lemma 1. We know from the above that the boundary value *U* ∈ S<sup>0</sup> (R*<sup>n</sup>* , H) is *U* = F[*V*], and *V* ∈ S<sup>0</sup> (R*<sup>n</sup>* , <sup>H</sup>) has the form *<sup>V</sup><sup>t</sup>* <sup>=</sup> *<sup>D</sup><sup>α</sup> t* **g**(*t*) in S 0 (R*<sup>n</sup>* , <sup>H</sup>). Let *<sup>φ</sup>* ∈ S(R*<sup>n</sup>* ). We have
$$
\begin{split}
\langle\mathcal{U},\boldsymbol{\Phi}\rangle &= \langle\mathcal{F}[\boldsymbol{V}],\boldsymbol{\Phi}\rangle = \langle\boldsymbol{V},\boldsymbol{\hat{\Phi}}\rangle = \langle\boldsymbol{D}^{a}\_{t}\mathbf{g}(t),\boldsymbol{\hat{\Phi}}(t)\rangle \\ &= (-1)^{|\boldsymbol{\alpha}|} \int\_{\mathbb{R}^{\boldsymbol{u}}} \mathbf{g}(t) D^{a}\_{t} \int\_{\mathbb{R}^{\boldsymbol{u}}} \boldsymbol{\phi}(\boldsymbol{x}) e^{2\pi i \langle \boldsymbol{x},t\rangle} d\boldsymbol{x} dt \\ &= (-1)^{|\boldsymbol{\alpha}|} \int\_{\mathbb{R}^{\boldsymbol{u}}} \mathbf{g}(t) \int\_{\mathbb{R}^{\boldsymbol{u}}} \boldsymbol{\phi}(\boldsymbol{x}) (-1/2\pi i)^{|\boldsymbol{\alpha}|} (2\pi i)^{|\boldsymbol{\alpha}|} \boldsymbol{x}^{\boldsymbol{\alpha}} e^{2\pi i \langle \boldsymbol{x},t\rangle} d\boldsymbol{x} dt \\ &= \langle\boldsymbol{\Phi}(t),\mathcal{F}[\boldsymbol{x}^{\boldsymbol{a}}\boldsymbol{\phi}(\boldsymbol{x});t]\rangle = \langle\mathcal{F}[\boldsymbol{g}]\_{\boldsymbol{x}},\boldsymbol{x}^{\boldsymbol{a}}\boldsymbol{\phi}(\boldsymbol{x})\rangle = \langle\boldsymbol{x}^{\boldsymbol{a}}\mathcal{F}[\boldsymbol{g}]\_{\boldsymbol{x}},\boldsymbol{\phi}(\boldsymbol{x})\rangle.
\end{split}
$$
Thus, *U<sup>x</sup>* = *x <sup>α</sup>*F[**g**]*<sup>x</sup>* in <sup>S</sup> 0 (R*<sup>n</sup>* , H) with **g** ∈ S<sup>0</sup> 2 (R*<sup>n</sup>* , H). Since **g** ∈ S<sup>0</sup> 2 (R*<sup>n</sup>* , H), by definition there is a real number *m* ≥ 0, such that **g**(*t*)/(1 + |*t*| 2 ) *<sup>m</sup>* <sup>∈</sup> *<sup>L</sup>* 2 (R*<sup>n</sup>* , H), and *m* can be taken to be a nonnegative integer. We have—by the proof of Lemma 1—that F[**g**]*<sup>x</sup>* = (1 − (4*π* 2 ) <sup>−</sup>1∆) *<sup>m</sup>***h**(*x*) <sup>∈</sup> *<sup>L</sup>*2(R*<sup>n</sup>* , H) in S 0 (R*<sup>n</sup>* , H), where **h** ∈ *L* 2 (R*<sup>n</sup>* , H) and ∆ is the Laplace operator in the *<sup>x</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* variable. Combining equalities, we have
$$\mathcal{U}\_{\mathbf{x}} = \mathfrak{x}^{\mathfrak{a}} \mathcal{F}[\mathbf{g}]\_{\mathfrak{x}} = \mathfrak{x}^{\mathfrak{a}} (1 - \frac{\Delta}{4\pi^2})^m \mathbf{h}(\mathfrak{x})^\mathfrak{x}$$
in S 0 (R*<sup>n</sup>* , H) with **h** ∈ *L* 2 (R*<sup>n</sup>* , H), which is (16). The proof is complete.
#### **5. Cauchy Integral**
A Cauchy integral of tempered distributions S 0 (R*<sup>n</sup>* ) has been defined in one and many dimensions. Of course, the main problem in making such a definition is that the Cauchy kernel is not a tempered function in <sup>S</sup>(R*<sup>n</sup>* ); an arbitrary element of S 0 (R*<sup>n</sup>* ) applied to the Cauchy kernel is not well defined.
Let *C* be a regular cone in R*<sup>n</sup>* ; that is, *C* is an open convex cone that does not contain an entirely straight line. With *C* ∗ being the dual cone of *C*, the Cauchy kernel function is
$$K(z - t) = \int\_{\mathbb{C}^\*} e^{2\pi i \langle z - t, \mu \rangle} d\mu, \; z \in T^{\mathbb{C}}, \; t \in \mathbb{R}^n.$$
as defined in Section 3. For the tube *T <sup>C</sup>* being the upper or lower half-planes in C<sup>1</sup> or the tube defined by one of the 2 *<sup>n</sup>* quadrant cones *<sup>C</sup><sup>µ</sup>* <sup>=</sup> {*<sup>y</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* : (−1) *<sup>µ</sup>jy<sup>j</sup>* <sup>&</sup>gt; 0, *<sup>j</sup>* <sup>=</sup> 1, . . . , *<sup>n</sup>*} where *µ* is any of the 2 *<sup>n</sup>* n-tuples whose components are either 0 or 1, the Cauchy kernel takes the usual form. In order to generate an element of <sup>S</sup>(R*<sup>n</sup>* ) from the Cauchy kernel in the half plane setting in C<sup>1</sup> and the tube defined by a quadrant cone, one divides the Cauchy kernel by a certain specifically chosen polynomial.
Sebastião e Silva [5] introduced a Cauchy integral for tempered distributions in the half-plane setting. Carmichael [7] defined a Cauchy integral for tempered distributions in the C*<sup>n</sup>* setting corresponding to analytic functions in the quadrant cone setting *T Cµ* in C*<sup>n</sup>* and showed that the analytic functions in (C − R) *n* , which have distributional boundary values in S 0 (R*<sup>n</sup>* ), can be recovered as the Cauchy integral of the boundary value; the results of [7] can be extended to the vector-valued tempered distributions considered in this paper by the same techniques as those in [7]. The Cauchy integrals introduced by both Sebastião e Silva and Carmichael are in fact equivalence classes of analytic functions defined by an integral involving the Cauchy kernel. Vladimirov [8–10] has defined a Cauchy integral for tempered distributions associated with analytic functions in general tubes *T <sup>C</sup>* <sup>=</sup> <sup>R</sup>*<sup>n</sup>* <sup>+</sup> *iC* <sup>⊂</sup> <sup>C</sup>*<sup>n</sup>* corresponding to regular cones *C* similar to the analytic functions we considered in this paper. Vladimirov showed that the analytic functions that he considered can be recovered by a Cauchy integral involving the tempered distributional boundary values of the analytic functions. The papers mentioned in this paragraph all concern scalar-valued analytic functions and distributions.
In this section, we build on our analyses of Sections 3 and 4 to obtain a Cauchy integral representation of the vector-valued analytic functions, which we considered in Theorem 1 and in Corollary 1. The proof of our results here—and the forms of our results—are different from any of the previous results concerning the Cauchy integral of the tempered distribution representation of the analytic functions. By our technique here, we do not need to divide the Cauchy kernel or the boundary value in (16) by a specified form of the polynomial and do not need to apply other special features of proof previously used by the authors in order to obtain that our Cauchy integral is well defined and that the analytic function considered is represented by a Cauchy integral involving the boundary value.
The Cauchy integral representation of the analytic functions that we considered in this paper follows. Note that cone *C* in the following result is assumed to be a regular cone. In Theorem 1 and Corollary 1, we assumed that cone *C* was an open convex cone. However, an open convex cone could contain an entirely straight line; in this case, the dual cone has measure 0 and *K*(*z* − *t*) = 0, *z* ∈ *T <sup>C</sup>*, *<sup>t</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* . To avoid this triviality, we assume that cone *C* in the following Cauchy integral representation is a regular cone.
**Theorem 2.** *Let C be a regular cone in* R*<sup>n</sup> and* H *be a Hilbert space. Let f*(*z*) *be analytic in T <sup>C</sup>* <sup>=</sup> <sup>R</sup>*<sup>n</sup>* <sup>+</sup> *iC, have values in* <sup>H</sup>*, and satisfy* (5)*. There is a continuous function <sup>g</sup>* ∈ S<sup>0</sup> 2 (R*<sup>n</sup>* , H) *with supp*(*g*) ⊆ *C* ∗ *a.e. and an n-tuple α of nonnegative integers, such that*
$$f(z) = z^{\mathfrak{a}} \langle \mathcal{F}[\mathfrak{g}]\_{\nu} \mathcal{K}(z - \nu) \rangle, \; z \in T^{\mathbb{C}}, \tag{17}$$
*in* S 0 (R*<sup>n</sup>* , H)*. Further,*
$$
\langle \mathcal{F}[\mathbf{g}]\_{\nu} \, \mathsf{K}(z - \nu) \rangle = \Theta \, \mathsf{i} \, z \in T^{-\mathsf{C}} \, \mathsf{A} \tag{18}
$$
*in* S 0 (R*<sup>n</sup>* , H)*.*
**Proof.** From Theorem 1, there is an element *V* ∈ S<sup>0</sup> (R*<sup>n</sup>* , H), the spectral function of **f**(*z*), *z* ∈ *T <sup>C</sup>*, such that *e* <sup>−</sup>2*π*h*y*,*t*i*V<sup>t</sup>* ∈ S<sup>0</sup> (R*<sup>n</sup>* , H), *y* ∈ *C*; supp(*V*) ⊆ *C* ∗ ; and **f**(*z*) = F[*e* <sup>−</sup>2*π*h*y*,*t*i*V<sup>t</sup>* ]*x*, *y* ∈ *C*, in S 0 (R*<sup>n</sup>* , H). Further, by Corollary 1, there is a continuous function **g** ∈ S<sup>0</sup> 2 (R*<sup>n</sup>* , H) with supp(**g**) ⊆ *C* ∗ a.e. and an n-tuple *α* of nonnegative integers, such that *V<sup>t</sup>* = *D<sup>α</sup> t* **<sup>g</sup>**(*t*), *<sup>t</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* . Now, let *<sup>φ</sup>* ∈ S(R*<sup>n</sup>* ) and *z* = *x* + *iy* ∈ *T <sup>C</sup>*. Recall that we have defined the differential operator *<sup>D</sup>* to be *<sup>D</sup><sup>t</sup>* = (−1/2*πi*)( *<sup>∂</sup> ∂t*<sup>1</sup> , . . . , *∂ ∂tn* ). We have
h**f**(*x* + *iy*), *φ*(*x*)i = hF[*e* <sup>−</sup>2*π*h*y*,*t*i*V<sup>t</sup>* ]*x*, *φ*(*x*)i = h*e* <sup>−</sup>2*π*h*y*,*t*i*V<sup>t</sup>* , *<sup>φ</sup>*ˆ(*t*)<sup>i</sup> <sup>=</sup> <sup>h</sup>*V<sup>t</sup>* , Z R *φ*(*x*)*e* 2*πi*h*z*,*t*i *dx*i = h*D α <sup>t</sup>* **g**(*t*), Z R*n φ*(*x*)*e* 2*πi*h*z*,*t*i *dx*i (19) = (−1) |*α*| Z *C*∗ **g**(*t*) Z R*n φ*(*x*)(−1/2*πi*) |*α*| (2*πi*) |*α*| *z α e* 2*πi*h*z*,*t*i *dxdt* = Z *C*∗ **g**(*t*) Z R*n φ*(*x*)*z α e* −2*π*h*y*,*t*i *e* 2*πi*h*x*,*t*i *dxdt* = Z *C*∗ *e* <sup>−</sup>2*π*h*y*,*t*i**g**(*t*)F[*<sup>z</sup> <sup>α</sup>φ*(*x*); *t*]*dt* = h*z <sup>α</sup>*F[*IC*<sup>∗</sup> (*t*)*<sup>e</sup>* <sup>−</sup>2*π*h*y*,*t*i**g**(*t*)]*x*, *<sup>φ</sup>*(*x*)<sup>i</sup>
where *IC*<sup>∗</sup> (*t*) is the characteristic function of *C* ∗ . We have proven in [17] (Lemma 4.2.1, p. 62) that *IC*<sup>∗</sup> (*t*)*e* <sup>−</sup>2*π*h*y*,*t*<sup>i</sup> <sup>∈</sup> *<sup>L</sup> p* , *y* ∈ *C*, for all *p*, 1 ≤ *p* ≤ ∞. Since **g** ∈ S<sup>0</sup> 2 (R*<sup>n</sup>* , H), then <sup>F</sup>[**g**]*<sup>x</sup>* <sup>∈</sup> *<sup>L</sup>*2(R*<sup>n</sup>* , H) in S 0 (R*<sup>n</sup>* , H) by Lemma 1. Recall also from Section 3 that the Cauchy kernel *K*(*z* − ·) ∈ D(∗, *L p* ) ⊂ D*<sup>L</sup> <sup>p</sup>* (R*<sup>n</sup>* ), 1 < *p* ≤ ∞, for *z* ∈ *T <sup>C</sup>* with *C* being a regular cone and that an element of *L*2(R*<sup>n</sup>* , H) applied to *K*(*z* − ·), *z* ∈ *T <sup>C</sup>*, is a well-defined function of *z* ∈ *T <sup>C</sup>*. Continuing (19) and using convolution, we now have
$$\begin{split} \langle \mathbf{f}(\mathbf{x} + i\mathbf{y}), \boldsymbol{\phi}(\mathbf{x}) \rangle &= \langle z^{\mathbf{a}}(\mathcal{F}[\mathbf{g}] \ast \mathcal{F}[I\_{\mathbb{C}^\*}(t)e^{-2\pi \langle \mathbf{y}, t \rangle}])\_{\mathbf{x} \prime} \boldsymbol{\phi}(\mathbf{x}) \rangle \\ &= \langle z^{\mathbf{a}}(\mathcal{F}[\mathbf{g}]\_{\mathbb{V}^\prime} \mathcal{F}[I\_{\mathbb{C}^\*}(t)e^{-2\pi \langle \mathbf{y}, t \rangle}]\_{\langle \mathbf{x} - \mathbf{v} \rangle} \boldsymbol{\phi}(\mathbf{x}) \rangle \\ &= \langle z^{\mathbf{a}}(\mathcal{F}[\mathbf{g}]\_{\mathbb{V}^\prime} \int\_{\mathbb{C}^\*} e^{2\pi i \langle z - \mathbf{v}, t \rangle} dt \rangle, \boldsymbol{\phi}(\mathbf{x}) \rangle \\ &= \langle z^{\mathbf{a}}(\mathcal{F}[\mathbf{g}]\_{\mathbb{V}^\prime} \mathcal{K}(z - \mathbf{v})), \boldsymbol{\phi}(\mathbf{x}) \rangle \end{split} \tag{20}$$
where *IC*<sup>∗</sup> (*t*) is the characteristic function of *C* ∗ . Since **g** ∈ S<sup>0</sup> 2 (R*<sup>n</sup>* , H), then F[**g**] ∈ *L*2(R*<sup>n</sup>* , H) by Lemma 1; and as previously noted, F[**g**] applied to the Cauchy kernel is a well-defined function of *z* ∈ *T <sup>C</sup>* and is an analytic function of *<sup>z</sup>* <sup>∈</sup> *<sup>T</sup> <sup>C</sup>* with values in <sup>H</sup>. Thus, from (20) we have obtained
$$\mathbf{f}(z) = z^{\kappa} \langle \mathcal{F}[\mathbf{g}]\_{\nu} \, K(z - \nu) \rangle\_{\nu} \, z \in T^{\mathbb{C}}\_{\nu}$$
in S 0 (R*<sup>n</sup>* , H), and (17) is obtained.
To prove (18), first note that for a regular cone, *C*, −*C* is also a regular cone; and (−*C*) ∗ = −*C* ∗ . Thus, for *z* ∈ *T* <sup>−</sup>*<sup>C</sup>* and *<sup>φ</sup>* ∈ S(R*<sup>n</sup>* ),
$$\begin{split} \langle \langle \mathcal{F}[\mathbf{g}]\_{\nu}, K(z-\nu) \rangle, \boldsymbol{\phi}(\mathbf{x}) \rangle &= \langle \langle \mathcal{F}[\mathbf{g}]\_{\nu}, \int\_{-\mathbb{C}^{\*}} e^{-2\pi \langle \boldsymbol{y}, t \rangle} e^{2\pi i \langle \mathbf{x} - \boldsymbol{y}, t \rangle} d\boldsymbol{t} \rangle, \boldsymbol{\phi}(\mathbf{x}) \rangle \\ &= \langle \langle \mathcal{F}[\mathbf{g}]\_{\nu}, \mathcal{F}[I\_{-\mathbb{C}^{\*}}(t) e^{-2\pi \langle \boldsymbol{y}, t \rangle}]\_{(\mathbf{x}-\boldsymbol{y})}, \boldsymbol{\phi}(\mathbf{x}) \rangle \\ &= \langle \langle (\mathcal{F}[\mathbf{g}] \* \mathcal{F}[I\_{-\mathbb{C}^{\*}}(t) e^{-2\pi \langle \boldsymbol{y}, t \rangle}])\_{\mathbf{x}} \rangle, \boldsymbol{\phi}(\mathbf{x}) \rangle \\ &= \langle \mathcal{F}[I\_{-\mathbb{C}^{\*}}(t) e^{-2\pi \langle \boldsymbol{y}, t \rangle} \mathbf{g}(t)]\_{\mathbf{x}}, \boldsymbol{\phi}(\mathbf{x}) \rangle. \end{split} \tag{21}$$
Now *I*−*C*<sup>∗</sup> (*t*) = 0 if *t* ∈ − / *C* <sup>∗</sup> and, hence, if *t* ∈ *C* ∗ . This fact coupled with the fact that supp(**g**) ⊆ *C* <sup>∗</sup> a.e. yields *I*−*C*<sup>∗</sup> (*t*)*e* <sup>−</sup>2*π*h*y*,*t*i**g**(*t*) = <sup>Θ</sup> a.e. for *<sup>t</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* and *<sup>y</sup>* ∈ −*C*. Hence F[*I*−*C*<sup>∗</sup> (*t*)*e* <sup>−</sup>2*π*h*y*,*t*i**g**(*t*)]*<sup>x</sup>* <sup>=</sup> <sup>Θ</sup>, *<sup>x</sup>* <sup>∈</sup> <sup>R</sup>*<sup>n</sup>* , *y* ∈ −*C*, in (21). Thus, from (21), we have hF[**g**]*ν*, *K*(*z* − *ν*)i = Θ, *z* ∈ *T* <sup>−</sup>*C*, in <sup>S</sup> 0 (R*<sup>n</sup>* , H); and (18) is obtained.
#### **6. Conclusions**
Tillmann [1] obtained the original analysis concerning the scalar-valued tempered distributions S 0 (R*<sup>n</sup>* ) as boundary values of analytic functions. We proved a boundary value result concerning vector-valued tempered distributions S 0 (R*<sup>n</sup>* , X ) as boundary values of
vector-valued analytic functions in [4] (Theorem 8) but used a norm growth condition on the analytic functions, which was a special case for the growth of Tillmann. We desired to obtain a result, such as [4] (Theorem 8), but under the general norm growth on the analytic function, which was equivalent to the growth of Tillmann. We achieved this first goal of this paper in Theorem 1 for vector-valued analytic functions **f**(*z*) on tubes *T <sup>C</sup>* = R*<sup>n</sup>* + *iC* with *C* being an open convex cone. The values of the analytic functions and the tempered distributions were in a very general type of topological vector space. We achieved additional information in Theorem 1 concerning the spectral function of the analytic function.
We asked if additional information concerning the spectral function and the boundary value could be obtained if the topological vector space X was restricted somewhat. We obtained the desired information in Corollary 1 by restricting X to be a Banach space and then a Hilbert space; we showed the structure of the spectral function and the boundary value in these cases for X . Integral to this analysis was the Lemma 1 result, which proved the relation under the Fourier transform between two important subsets of S 0 (R*<sup>n</sup>* , H) for our results in Corollary 1. It is important to note that the reason to restrict to Hilbert space H (which we do in our results) is that the Plancherel theory for the Fourier transform of the functions holds if and only if the functions have value in the Hilbert space.
The second principal goal of this paper was to obtain a Cauchy integral representation of the analytic functions considered in Theorem 1 and Corollary 1. Sebastião e Silva, Carmichael, and Vladimirov have obtained and studied the Cauchy integral of tempered distributions S 0 (R*<sup>n</sup>* ) in the scalar-valued case and in one and several dimensions; see the papers of these authors in the references. Their analyses basically concerned dividing the Cauchy kernel or the boundary value by a suitable polynomial whose order was large enough to make the quotient when evaluated by the tempered distribution to be well defined, or used other special features of proof that we do not use here.
In Section 5 of this paper, we constructed our Cauchy integral used in the representation of the assumed analytic function in a different manner by using the general known structure of the spectral function and our proven structure of the tempered distributional boundary value in S 0 (R*<sup>n</sup>* , H) for H being a Hilbert space. The analytic function obtaining the boundary value in S 0 (R*<sup>n</sup>* , H) was shown to be equated to the product of a polynomial and the constructed Cauchy integral.
This paper concerns theoretical mathematics, yet the topics considered find applications in mathematical physics and in mathematics that are applied to physical problems. We survey historically some areas of application in the scalar-valued case. We recall the work of Streater and Wightman [23] in studying quantum field theory. In a field theory, the "vacuum expectation values" are tempered distributions, which are boundary values in the tempered distribution topology of analytic functions with the analytic functions being Fourier–Laplace transforms. In addition, a field theory can be recovered from its "vacuum expectation values"; see [23] (Chapter 3). A similar field theory analysis using boundary values of analytic functions is contained in the work by Simon [24]. We also reference Raina [25] concerning "form factor bounds" in particle physics in which tempered distributional boundary values, which are of a special form, imply that the analytic functions that obtain these boundary values are Hardy *H<sup>p</sup>* functions; this fact is then used in the analysis of the "form factor bounds". See also the associated papers listed in the references of [25].
As noted in Vladimirov [8], scalar-valued analytic functions of the type that we considered in this paper can arise in applying the Fourier–Laplace transform to convolution equations, which describe linear homogeneous processes with causality that find application in the quantum field theory, theory of electrical circuits, scattering of electromagnetic waves, and linear thermodynamic systems; refer to the list of references in [8]. We also note paper [26] by Vladimirov, concerning the linear conjugacy of scalar-valued analytic functions of several complex variables, which are again of the type that we considered in this paper with respect to growth. The linear conjugacy analysis involves scalar-valued tempered distributional boundary values of analytic functions represented as Fourier– Laplace integrals. Vladimirov [26] (p. 207) states that many problems arising in mathematical physics reduce to the problem of linear conjugacy involving tempered distributions; Vladimirov [26] provides examples of such problems.
The survey of applications above (concerning the type of analysis used in this paper) involve scalar-valued functions and distributions. Yet, a close consideration of the linear conjugacy problem of [26], together with the vector-valued analysis of this paper, leads one to believe that the linear conjugacy problem can be extended to the vector-valued case. Further, in an analysis of the stated applications above, one must sometimes obtain a distributional solution of a partial differential equation; such calculations can be extended to the vector-valued case. We suggest that the considerable related analyses to the results of this paper and the results of related references in this paper can be achieved in the vector-valued case and will work toward this end in the future.
**Funding:** This research received no external funding.
**Institutional Review Board Statement:** Not applicable.
**Informed Consent Statement:** Not applicable.
**Data Availability Statement:** Not applicable.
**Conflicts of Interest:** The author declares no conflict of interest.
#### **References**
## *Article* **On Special Fuzzy Differential Subordinations Obtained for Riemann–Liouville Fractional Integral of Ruscheweyh and S˘al˘agean Operators**
**Alina Alb Lupa¸s**
Department of Mathematics and Computer Science, University of Oradea, 1 Universitatii Street, 410087 Oradea, Romania; [email protected]
**Abstract:** New results concerning fuzzy differential subordination theory are obtained in this paper using the operator denoted by *D*−*<sup>λ</sup> <sup>z</sup> L n α* , previously introduced by applying the Riemann–Liouville fractional integral to the convex combination of well-known Ruscheweyh and S˘al˘agean differential operators. A new fuzzy subclass *DL*F *n* (*δ*, *α*, *λ*) is defined and studied involving the operator *D*−*<sup>λ</sup> <sup>z</sup> L n α* . Fuzzy differential subordinations are obtained considering functions from class *DL*F *n* (*δ*, *α*, *λ*) and the fuzzy best dominants are also given. Using particular functions interesting corollaries are obtained and an example shows how the obtained results can be applied.
**Keywords:** differential operator; fuzzy differential subordination; fuzzy best dominant; fractional integral
**MSC:** 30C45; 30A10; 33C05
#### **1. Introduction**
The concept of fuzzy set, introduced by Lotfi A. Zadeh in 1965 [1], has opened the way for a new theory called fuzzy set theory. It has developed intensely, nowadays having applications in many branches of science and technology.
The fuzzy set concept was applied for developing new directions of study in many mathematical theories. In geometric function theory, it was used for introducing the new concepts of fuzzy subordination [2] and fuzzy differential subordinations [3] as generalizations of the classical notion of differential subordination due to Miller and Mocanu [4,5]. The main aspects regarding the theory of differential subordination can be found in [6]. Steps in the evolution of the theory of fuzzy differential subordination can be followed in [7].
The general context of the study presented in this paper contains notions familiar to geometric function theory merged with fuzzy set theory. We first present the main classes of analytic functions involved and the definitions regarding fuzzy differential subordination theory.
*U* = {*z* ∈ C : |*z*| < 1} represents the unit disc of the complex plane and H(*U*) the space of holomorphic functions in *U*.
Consider A = { *f* ∈ H(*U*) : *f*(*z*) = *z* + *a*2*z* <sup>2</sup> <sup>+</sup> . . . , *<sup>z</sup>* <sup>∈</sup> *<sup>U</sup>*}, and <sup>H</sup>[*a*, *<sup>m</sup>*] = { *<sup>f</sup>* <sup>∈</sup> H(*U*) : *f*(*z*) = *a* + *amz <sup>m</sup>* + *am*+1*z <sup>m</sup>*+<sup>1</sup> <sup>+</sup> . . . , *<sup>z</sup>* <sup>∈</sup> *<sup>U</sup>*}, for *<sup>a</sup>* <sup>∈</sup> <sup>C</sup> and *<sup>m</sup>* <sup>∈</sup> <sup>N</sup>.
We remember the usual definitions needed for fuzzy differential subordination:
**Definition 1** ([8])**.** *A fuzzy subset of X is a pair* (*M*, *FA*)*, with M* = {*x* ∈ *X* : 0 < *FM*(*x*) ≤ 1} *the support of the fuzzy set and F<sup>M</sup>* : *X* → [0, 1] *the membership function of the fuzzy set. It is denoted M* = supp(*M*, *FM*)*.*
$$\textbf{Remark 1.}\text{ When }M\subset X\text{, we have }\mathsf{F}\_M(\mathsf{x})=\left\{\begin{array}{c}1,\mathsf{if}\ x\in M,\\0,\mathsf{if}\ x\notin M.\end{array}\right.$$
**Citation:** Alb Lupa¸s, A. On Special Fuzzy Differential Subordinations Obtained for Riemann–Liouville Fractional Integral of Ruscheweyh and S˘al˘agean Operators. *Axioms* **2022**, *11*, 428. https://doi.org/10.3390/ axioms11090428
Academic Editor: Radko Mesiar
Received: 18 July 2022 Accepted: 23 August 2022 Published: 25 August 2022
**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.
**Copyright:** © 2022 by the author. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).
*Evidently F*∅(*x*) = 0*, x* ∈ *X, and FX*(*x*) = 1*, x* ∈ *X.*
**Definition 2** ([2])**.** *Let D* ⊂ C *and let z*<sup>0</sup> ∈ *D be a fixed point. We take the functions f* , *g* ∈ H(*D*)*. The function f is said to be fuzzy subordinate to g and we write f* ≺<sup>F</sup> *g, if there exists a function F* : C → [0, 1] *such that f*(*z*0) = *g*(*z*0) *and Ff*(*D*) *f*(*z*) ≤ *Fg*(*D*) *g*(*z*)*, z* ∈ *D*.
**Remark 2.** *(1) If g is univalent, then f* ≺<sup>F</sup> *g if and only if f*(*z*0) = *g*(*z*0) *and f*(*D*) ⊂ *g*(*D*). |*z*| .
*(2) Such a function F* : C → [0, 1] *can be consider F*(*z*) = 1+|*z*| *, F*(*z*) = <sup>1</sup> 1+|*z*|
*(3) If D* = *U the conditions become f*(0) = *g*(0) *and f*(*U*) ⊂ *g*(*U*)*, which is equivalent to the classical definition of subordination.*
**Definition 3** ([3])**.** *Consider <sup>h</sup> an univalent function in <sup>U</sup> and <sup>ψ</sup>* : <sup>C</sup><sup>3</sup> <sup>×</sup> *<sup>U</sup>* <sup>→</sup> <sup>C</sup>*, such that h*(0) = *ψ*(*a*, 0; 0) = *a. When the fuzzy differential subordination*
$$F\_{\boldsymbol{\upPsi}(\mathbb{C}^{3}\times\mathcal{U})}\psi(p(\boldsymbol{z}),zp'(\boldsymbol{z}),\boldsymbol{z}^{2}p''(\boldsymbol{z});\boldsymbol{z}) \leq F\_{\boldsymbol{\upPsi}(\mathcal{U})}h(\boldsymbol{z}), \quad \boldsymbol{z}\in\mathcal{U}.\tag{1}$$
*is satisfied for an analytic function p in U, such that p*(0) = *a, then p is called a fuzzy solution of the fuzzy differential subordination. A fuzzy dominant of the fuzzy solutions of the fuzzy differential subordination is an univalent function q for which Fp*(*U*) *p*(*z*) ≤ *Fq*(*U*) *q*(*z*)*, z* ∈ *U, for all <sup>p</sup> satisfying (1). The fuzzy best dominant of (1) is a fuzzy dominant <sup>q</sup>*e*, such that <sup>F</sup>q*e(*U*) *q*˜(*z*) ≤ *Fq*(*U*) *q*(*z*)*, z* ∈ *U, for all fuzzy dominants q of (1).*
**Lemma 1** ([6])**.** *Consider <sup>h</sup>* ∈ A*. If Re zh*00(*z*) *h* <sup>0</sup>(*z*) + 1 <sup>&</sup>gt; <sup>−</sup><sup>1</sup> 2 *, <sup>z</sup>* <sup>∈</sup> *<sup>U</sup>*, *then* <sup>1</sup> *z* R *z* 0 *h*(*t*)*dt is a convex function, z* ∈ *U.*
**Lemma 2** ([9])**.** *Consider a convex function h with h*(0) = *a, and γ* ∈ C<sup>∗</sup> *such that Re γ* ≥ 0*. When <sup>p</sup>* ∈ H[*a*, *<sup>m</sup>*]*, <sup>ψ</sup>* : <sup>C</sup><sup>2</sup> <sup>×</sup> *<sup>U</sup>* <sup>→</sup> <sup>C</sup>*, <sup>ψ</sup>*(*p*(*z*), *zp*<sup>0</sup> (*z*); *z*) = <sup>1</sup> *γ zp*0 (*z*) + *p*(*z*) *is an analytic function in U and*
$$F\_{\Psi(\mathbb{C}^2 \times \mathcal{U})} \left( \frac{1}{\gamma} z p'(z) + p(z) \right) \le F\_{h(\mathcal{U})} h(z), \ z \in \mathcal{U}\_{\nu}$$
*then*
$$F\_{p(\mathcal{U})}p(z) \le F\_{\mathcal{G}(\mathcal{U})}g(z) \le F\_{h(\mathcal{U})}h(z), \quad z \in \mathcal{U}\_{\mathcal{A}}$$
*with the convex function g*(*z*) = *<sup>γ</sup> mz γ m* R *z* 0 *h*(*t*)*t γ <sup>m</sup>* −1 *dt*, *z* ∈ *U as the fuzzy best dominant.*
**Lemma 3** ([9])**.** *Consider a convex function g in U and define h*(*z*) = *mαzg*0 (*z*) + *g*(*z*), *z* ∈ *U*, *with m* ∈ N *and α* > 0*.*
*If p*(*z*) = *g*(0) + *pmz <sup>m</sup>* + *pm*+1*z <sup>m</sup>*+<sup>1</sup> <sup>+</sup> . . . , *<sup>z</sup>* <sup>∈</sup> *<sup>U</sup>*, *is a holomorphic function in U and*
$$F\_{p(\mathcal{U})}\left(p(z) + \alpha z p'(z)\right) \le F\_{h(\mathcal{U})}h(z), \quad z \in \mathcal{U}\_{\mathcal{H}}$$
*then we obtain the sharp result*
$$F\_{p(\mathcal{U})}p(z) \le F\_{\mathcal{g}(\mathcal{U})}\mathcal{g}(z), z \in \mathcal{U}.$$
The original results exposed in this paper are obtained using the well-known Ruscheweyh and S˘al˘agean differential operators combined with Riemann–Liouville fractional integral. The resulting operator was introduced in [10], where it was used for obtaining results involving classical differential subordination theory. The necessary definitions are reminded:
**Definition 4** (Ruscheweyh [11])**.** *The Ruscheweyh operator R<sup>n</sup> is introduced by R<sup>n</sup>* : A → A*,*
$$\begin{array}{rcl} \mathcal{R}^0 f(z) &=& f(z) \\ \mathcal{R}^1 f(z) &=& zf'(z) \\ &\cdots \\ (n+1)\mathcal{R}^{n+1} f(z) &=& n\mathcal{R}^n f(z) + z(\mathcal{R}^n f(z))' \end{array}$$
*for f* ∈ A, *n* ∈ N, *z* ∈ *U*.
**Remark 3.** *For a function f*(*z*) = *z* + ∑ ∞ *j*=2 *ajz <sup>j</sup>* ∈ A, *the Ruscheweyh operator can be written using the following form R n f*(*z*) = *z* + ∑ ∞ *j*=2 Γ(*n*+*j*) Γ(*n*+1)Γ(*j*) *ajz j , z* ∈ *U, where* Γ *denotes the gamma function.*
**Definition 5** (S˘al˘agean [12])**.** *The S˘al˘agean operator S<sup>n</sup> is introduced by S<sup>n</sup>* : A → A*,*
$$\begin{array}{rcl} S^0 f(z) &=& f(z), \\ S^1 f(z) &=& zf'(z), \\ & & \cdots \\ S^{n+1} f(z) &=& z(S^n f(z))', \end{array}$$
*for f* ∈ A*, n* ∈ N*, z* ∈ *U*.
**Remark 4.** *For a function f*(*z*) = *z* + ∑ ∞ *j*=2 *ajz <sup>j</sup>* ∈ A*, the S˘al˘agean operator can be written using the following form S<sup>n</sup> f*(*z*) = *z* + ∑ ∞ *j*=2 *j n ajz j , z* ∈ *U.*
**Definition 6** ([13])**.** *Define the linear operator L<sup>n</sup> α* : A → A, *given by*
$$L\_{\mathfrak{a}}^{\mathfrak{n}}f(z) = \mathfrak{a}\mathcal{S}^{\mathfrak{n}}f(z) + (1-\mathfrak{a})\mathcal{R}^{\mathfrak{n}}f(z), \ z \in \mathcal{U}\_{\mathfrak{n}}$$
*where α* ≥ 0*, n* ∈ N*.*
**Remark 5.** *For a function f*(*z*) = *z* + ∑ ∞ *j*=2 *ajz <sup>j</sup>* ∈ A, *the defined operator can be written using the following form L<sup>n</sup> α f*(*z*) = *z* + ∑ ∞ *j*=2 h *αj <sup>n</sup>* <sup>+</sup> (<sup>1</sup> <sup>−</sup> *<sup>α</sup>*) Γ(*n*+*j*) Γ(*n*+1)Γ(*j*) i *ajz j* , *z* ∈ *U*.
We also remind the definition of Riemann–Liouville fractional integral:
**Definition 7** ([14])**.** *The Riemann–Liouville fractional integral of order λ applied to an analytic function f is defined by*
$$D\_z^{-\lambda}f(z) = \frac{1}{\Gamma(\lambda)} \int\_0^z \frac{f(t)}{\left(z - t\right)^{1 - \lambda}} dt\_\lambda$$
*with λ* > 0.
In [10] we defined the Riemann–Liouville fractional integral applied to the operator *L n <sup>α</sup>* as follows:
**Definition 8** ([10])**.** *The Riemann–Liouville fractional integral applied to the differential operator L n α f is introduced by*
$$D\_z^{-\lambda}L\_a^n f(z) = \frac{1}{\Gamma(\lambda)} \int\_0^z \frac{L\_a^n f(t)}{(z-t)^{1-\lambda}} dt = $$
$$\frac{1}{\Gamma(\lambda)} \int\_0^z \frac{t}{(z-t)^{1-\lambda}} dt + \sum\_{j=2}^\infty \left( aj^n + (1-a)\frac{\Gamma(n+j)}{\Gamma(n+1)\Gamma(j)} \right) a\_j \int\_0^z \frac{t^j}{(z-t)^{1-\lambda}} dt,$$
*where α* ≥ 0*, λ* > 0 *and n* ∈ N*.*
**Remark 6.** *For a function f*(*z*) = *z* + ∑ ∞ *j*=2 *ajz <sup>j</sup>* ∈ A, *the Riemann–Liouville fractional integral of L<sup>n</sup> α f has the following form*
$$D\_z^{-\lambda}L\_a^{\mathfrak{n}}f(z) = \frac{1}{\Gamma(2+\lambda)}z^{1+\lambda} + \sum\_{j=2}^{\infty} \left[ \frac{aj^m\Gamma(j+1)}{\Gamma(j+\lambda+1)} + \frac{(1-a)j\Gamma(m+j)}{\Gamma(m+1)\Gamma(j+\lambda+1)} \right] a\_j z^{j+\lambda}/\lambda$$
*and D*−*<sup>λ</sup> <sup>z</sup> L n α f*(*z*) ∈ H[0, *λ* + 1].
The results exposed in this paper follow a line of research concerned with fuzzy differential subordinations which is popular nowadays, namely introducing new operators and using them for defining and studying new fuzzy classes of functions.
Fuzzy differential subordinations involving Ruscheweyh and S˘al˘agean differential operators were obtained in many studies, such as [15]. New operators introduced using fractional integral and applied in fuzzy differential subordination theory were studied in [16] where Riemann–Liouville fractional integral is applied for Gaussian hypergeometric function and in [17] where Riemann–Liouville fractional integral is combined with confluent hypergeometric function.
Motivated by the nice results obtained in fuzzy differential subordination theory using Ruscheweyh and S˘al˘agean differential operators and fractional integral applied to different known operators, the study presented in this paper uses the previously defined operator *D*−*<sup>λ</sup> <sup>z</sup> L n <sup>α</sup>* given in Definition 8 applied for obtaining new fuzzy differential subordinations. In the next section, a new fuzzy class will be defined and studied in order to obtain fuzzy differential subordinations inspired by recently published studies concerned with the same topic seen in [18–20].
The main results contained in Section 2 of the paper, begin with the definition of a new fuzzy class *DL*F *n* (*δ*, *α*, *λ*) for which the operator *D*−*<sup>λ</sup> <sup>z</sup> L n <sup>α</sup>* given in Definition 8 is used. The property of this class to be convex is proved and certain fuzzy differential subordinations involving functions from the class and the operator *D*−*<sup>λ</sup> <sup>z</sup> L n <sup>α</sup>* are obtained. The fuzzy best dominants are given for the considered fuzzy differential subordinations in theorems which generate interesting corollaries when specific functions with remarkable geometric properties are used as fuzzy best dominants. An example is also shown in order to prove the applicability of the new results.
#### **2. Main Results**
The usage of the operator *D*−*<sup>λ</sup> <sup>z</sup> L n α* seen in Definition 8 defines a new fuzzy subclass of analytic functions as follows:
**Definition 9.** *The class DL*F *n* (*δ*, *α*, *λ*) *is composed of all functions f* ∈ A *with the property*
$$F\_{\left(D\_z^{-\lambda}L\_a^nf\right)'(\mathcal{U})} ^{\prime} \left(D\_z^{-\lambda}L\_a^nf(z)\right)^{\prime} > \delta, \quad z \in \mathcal{U}\_{\prime}$$
*where n* ∈ N, *δ* ∈ [0, 1)*, α* ≥ 0*, λ* > 0*.*
We begin studying this subclass of functions:
**Theorem 1.** *DL*F *n* (*δ*, *α*, *λ*) *is a convex set.*
**Proof.** Taking the functions
$$f\_k(z) = z + \sum\_{j=2}^{\infty} a\_{jk} z^j, \quad k = 1, 2, \quad z \in \mathcal{U}\_2$$
belonging to the class *DL*F *n* (*δ*, *α*, *λ*), we have to prove that the function
$$h(z) = \gamma\_1 f\_1(z) + \gamma\_2 f\_2(z)$$
belongs to the class *DL*F *n* (*δ*, *α*, *λ*) with *γ*1, *γ*<sup>2</sup> ≥ 0, *γ*<sup>1</sup> + *γ*<sup>2</sup> = 1.
We have *h* 0 (*z*) = (*γ*<sup>1</sup> *f*<sup>1</sup> + *γ*<sup>2</sup> *f*2) 0 (*z*) = *γ*<sup>1</sup> *f* 0 1 (*z*) + *γ*<sup>2</sup> *f* 0 2 (*z*), *z* ∈ *U*, and *D*−*<sup>λ</sup> <sup>z</sup> L n <sup>α</sup>h*(*z*) 0 = *D*−*<sup>λ</sup> <sup>z</sup> L n α* (*γ*<sup>1</sup> *f*<sup>1</sup> + *γ*<sup>2</sup> *f*2)(*z*) 0 = *γ*<sup>1</sup> *D*−*<sup>λ</sup> <sup>z</sup> L n α f*1(*z*) 0 + *γ*<sup>2</sup> *D*−*<sup>λ</sup> <sup>z</sup> L n α f*2(*z*) 0 . and we can write *F* (*D* −*λ <sup>z</sup> L n <sup>α</sup> h*) 0 (*U*) *D*−*<sup>λ</sup> <sup>z</sup> L n <sup>α</sup>h*(*z*) 0 = *F* (*D* −*λ <sup>z</sup> L n <sup>α</sup>* (*γ*<sup>1</sup> *f*1+*γ*<sup>2</sup> *f*2)) 0 (*U*) *D*−*<sup>λ</sup> <sup>z</sup> L n α* (*γ*<sup>1</sup> *f*<sup>1</sup> + *γ*<sup>2</sup> *f*2)(*z*) 0 = *F* (*D* −*λ <sup>z</sup> L n <sup>α</sup>* (*γ*<sup>1</sup> *f*1+*γ*<sup>2</sup> *f*2)) 0 (*U*) *γ*1 *D*−*<sup>λ</sup> <sup>z</sup> L n α f*1(*z*) 0 + *γ*<sup>2</sup> *D*−*<sup>λ</sup> <sup>z</sup> L n α f*2(*z*) 0 = *F* (*<sup>γ</sup>*1*<sup>D</sup>* −*λ z L n α f*1) 0 (*U*) *γ*1(*D*−*<sup>λ</sup> <sup>z</sup> L n α f*1(*z*)) 0 +*F* (*<sup>γ</sup>*2*<sup>D</sup>* −*λ z L n α f*2) 0 (*U*) *γ*2(*D*−*<sup>λ</sup> <sup>z</sup> L n α f*2(*z*)) 0 <sup>2</sup> = *F* (*D* −*λ z L n α f*1) 0 (*U*) (*D*−*<sup>λ</sup> <sup>z</sup> L n α f*1(*z*)) 0 +*F* (*D* −*λ z L n α f*2) 0 (*U*) (*D*−*<sup>λ</sup> <sup>z</sup> L n α f*2(*z*)) 0 2 .
Having *f*1, *f*<sup>2</sup> ∈ *DL*<sup>F</sup> *n* (*δ*, *α*, *λ*) we get *δ* < *F* (*D* −*λ <sup>z</sup> L n <sup>α</sup> f*1) 0 (*U*) *D*−*<sup>λ</sup> <sup>z</sup> L n α f*1(*z*) 0 ≤ 1 and *δ* < *F* (*D* −*λ <sup>z</sup> L n <sup>α</sup> f*2) 0 (*U*) *D*−*<sup>λ</sup> <sup>z</sup> L n α f*2(*z*) 0 ≤ 1, *z* ∈ *U*.
In these conditions *δ* < *F* (*D* −*λ z L n α f*1) 0 (*U*) (*D*−*<sup>λ</sup> <sup>z</sup> L n α f*1(*z*)) 0 +*F* (*D* −*λ z L n α f*2) 0 (*U*) (*D*−*<sup>λ</sup> <sup>z</sup> L n α f*2(*z*)) 0 <sup>2</sup> ≤ 1 and we get *δ* < *F* (*D* −*λ <sup>z</sup> L n <sup>α</sup> h*) 0 (*U*) *D*−*<sup>λ</sup> <sup>z</sup> L n <sup>α</sup>h*(*z*) 0 ≤ 1, equivalently with *h* ∈ *DL*<sup>F</sup> *n* (*δ*, *α*, *λ*) and *DL*F *n* (*δ*, *α*, *λ*) is a convex set.
We give fuzzy differential subordinations obtained for the operator *D*−*<sup>λ</sup> <sup>z</sup> L n a* .
**Theorem 2.** *Considering a convex function g in U and defining h*(*z*) = *g*(*z*) + <sup>1</sup> *c*+2 *zg*0 (*z*), *with c* > 0, *z* ∈ *U*, *when f* ∈ *DL*<sup>F</sup> *n* (*δ*, *α*, *λ*) *and G*(*z*) = *<sup>c</sup>*+<sup>2</sup> *z c*+1 R *z* 0 *t c f*(*t*)*dt, z* ∈ *U*, *then*
$$\left(\mathrm{F}\_{\left(D\_z^{-\lambda}L\_u^{\mathrm{u}}f\right)'(\mathsf{U})}^{\prime}(\mathsf{U})\left(\mathrm{D}\_z^{-\lambda}L\_u^{\mathrm{u}}f(z)\right)' \leq \mathrm{F}\_{\mathbb{H}(\mathsf{U})}h(z), \quad z \in \mathsf{U},\tag{2}$$
*implies the sharp result*
$$\left(F\_{\left(D\_z^{-\lambda}L\_a^{\eta}G\right)'(\mathcal{U})}\right)'(\mathcal{U})\left(D\_z^{-\lambda}L\_a^{\eta}G(z)\right)' \le F\_{\mathcal{G}(\mathcal{U})}g(z), \ z \in \mathcal{U}.\,\,\,\xi$$
**Proof.** Differentiating relation
$$z^{\mathfrak{c}+1}G(z) = (\mathfrak{c}+\mathfrak{2})\int\_0^z t^{\mathfrak{c}}f(t)dt,$$
considering *z* as variable, we get (*c* + 1)*G*(*z*) + *zG*0 (*z*) = (*c* + 2)*f*(*z*) and
$$\left( (\mathfrak{c} + 1) D\_z^{-\lambda} L\_\mathfrak{a}^\mathfrak{n} G(z) + z \left( D\_z^{-\lambda} L\_\mathfrak{a}^\mathfrak{n} G(z) \right)' = (\mathfrak{c} + 2) D\_z^{-\lambda} L\_\mathfrak{a}^\mathfrak{n} f(z), \quad z \in \mathsf{U}\_\sigma$$
and differentiating it again with respect to *z*, we obtain
$$\left(D\_z^{-\lambda}L\_a^{\underline{n}}G(z)\right)' + \frac{1}{c+2}z\left(D\_z^{-\lambda}L\_a^{\underline{n}}G(z)\right)'' = \left(D\_z^{-\lambda}L\_a^{\underline{n}}f(z)\right)', \ z \in \mathcal{U}.\,\,\,\xi$$
and the inequality (2) representing the fuzzy differential subordination can be written
$$\left(F\_{D\_z^{-\lambda}L\_a^nG(\mathcal{U})}\left(\frac{1}{c+2}z\left(D\_z^{-\lambda}L\_a^nG(z)\right)'' + \left(D\_z^{-\lambda}L\_a^nG(z)\right)'\right) \le F\_{\mathcal{S}}(\mathcal{U})\left(\frac{1}{c+2}zg'(z) + g(z)\right).$$
Denoted
$$p(z) = \left(D\_z^{-\lambda} L\_a^n G(z)\right)', \ z \in \mathcal{U}\_\nu$$
where *p* ∈ H[1, *n*], we obtain
$$F\_{p(\mathcal{U})}\left(\frac{1}{\varepsilon+2}zp'(z)+p(z)\right) \le F\_{\mathcal{S}(\mathcal{U})}\left(\frac{1}{\varepsilon+2}zg'(z)+g(z)\right), \ z \in \mathcal{U}.$$
Applying Lemma 3, we get
$$F\_{\left(D\_z^{-\lambda}L\_{\mathfrak{a}}^{\mathfrak{n}}G\right)'(\mathcal{U})} ^{\prime} \left(D\_z^{-\lambda}L\_{\mathfrak{a}}^{\mathfrak{n}}G(z)\right)^{\prime} \leq F\_{\mathcal{G}(\mathcal{U})}g(z), \ z \in \mathcal{U}\_{\prime}$$
and *g* is the best dominant.
We give an inclusion result for the class *DL*F *n* (*δ*, *α*, *λ*):
**Theorem 3.** *Taking h*(*z*) = 1+(2*δ*−1)*z* 1+*z and G*(*z*) = *<sup>c</sup>*+<sup>2</sup> *z c*+1 R *z* 0 *t c f*(*t*)*dt, z* ∈ *U, with δ* ∈ [0, 1), *c* > 0*, n* ∈ N, *α* ≥ 0*, λ* > 0*, then*
$$\mathbb{G}\left[DL\_{\mathfrak{n}}^{\mathcal{F}}(\delta,\mathfrak{a},\lambda)\right] \subset DL\_{\mathfrak{n}}^{\mathcal{F}}(\delta^\*,\mathfrak{a},\lambda),\tag{3}$$
*where δ* <sup>∗</sup> = 2*δ* − 1 + 2(2 + *c*)(1 − *δ*) R 1 0 *t c*+1 *t*+1 *dt*.
**Proof.** Making the same steps such as in the proof of Theorem 2, taking account the hypothesis of Theorem 3 and that *h*(*z*) = 1+(2*δ*−1)*z* 1+*z* is a convex function, we obtain
$$F\_{p(\mathcal{U})}\left(\frac{1}{c+2}zp'(z)+p(z)\right) \le f\_{h(\mathcal{U})}h(z).$$
with *p*(*z*) = *D*−*<sup>λ</sup> <sup>z</sup> L n <sup>α</sup>G*(*z*) 0 , *z* ∈ *U*. Applying Lemma 2, we get
$$F\_{\left(D\_z^{-\lambda}L\_{\mathfrak{a}}^{\eta}G\right)'(\mathcal{U})} \left(D\_z^{-\lambda}L\_{\mathfrak{a}}^{\eta}G(z)\right)' \le F\_{\mathcal{G}(\mathcal{U})}\mathcal{g}(z) \le F\_{h(\mathcal{U})}h(z).$$
where
$$g(z) = \frac{2+c}{nz^{\frac{2+\zeta}{n}}} \int\_0^z t^{\frac{2+\zeta}{n}-1} \frac{1+(2\delta-1)t}{1+t} dt = (2\delta-1) + \frac{2(c+2)(1-\delta)}{nz^{\frac{\zeta+2}{n}}} \int\_0^z t^{\frac{2+\zeta}{n}-1} dt.$$
Since the function *g* is convex and *g*(*U*) is symmetric with respect to the real axis, we can write
$$F\_{D\_z^{-\lambda}L\_R^nG(\mathcal{U})}\left(D\_z^{-\lambda}L\_R^nG(z)\right)' \geq \min\_{|z|=1} F\_{\mathcal{S}}(\mathcal{U})\mathcal{g}(z) = F\_{\mathcal{S}}(\mathcal{U})\mathcal{g}(1)\tag{4}$$
$$\text{and } \delta^\* = g(1) = 2\delta - 1 + \frac{2(2+c)(1-\delta)}{n} \int\_0^1 \frac{t^{\frac{2+c}{n}-1}}{t+1}dt, \text{ that give the inclusion (3).} \quad \square$$
**Theorem 4.** *Taking a convex function g with the property g*(0) = 0, *define h*(*z*) = *g*(*z*) + *zg*0 (*z*), *z* ∈ *U*. *When f* ∈ A, *n* ∈ N, *α* ≥ 0, *λ* > 0*, and the fuzzy differential subordination holds*
$$\left(\mathrm{F}\_{\left(D\_{z}^{-\lambda}L\_{u}^{n}f\right)'(\mathsf{U})}^{\prime}\left(\mathrm{D}\_{z}^{-\lambda}L\_{u}^{n}f(z)\right)' \leq \mathrm{F}\_{h(\mathsf{U})}h(z), \quad z \in \mathsf{U},\tag{5}$$
*then we get the sharp result*
$$F\_{D\_z^{-\lambda}L\_a^n f(\mathcal{U})} \frac{D\_z^{-\lambda}L\_a^n f(z)}{z} \le F\_{\mathcal{S}(\mathcal{U})} \mathcal{g}(z), \quad z \in \mathcal{U}.$$
**Proof.** Considering *<sup>p</sup>*(*z*) = *<sup>D</sup>*−*<sup>λ</sup> <sup>z</sup> L n α f*(*z*) *<sup>z</sup>* ∈ H[0, *<sup>λ</sup>*], we can write *zp*(*z*) = *<sup>D</sup>*−*<sup>λ</sup> <sup>z</sup> L n α f*(*z*), *z* ∈ *U*, and differentiating it we get *zp*0 (*z*) + *p*(*z*) = *D*−*<sup>λ</sup> <sup>z</sup> L n α f*(*z*) 0 , *z* ∈ *U*.
The inequality (5) can be written as following
$$F\_{p(\mathcal{U})}\left(zp'(z) + p(z)\right) \le F\_{\mathbb{H}(\mathcal{U})}h(z) = F\_{\mathcal{g}(\mathcal{U})}\left(zg'(z) + \mathcal{g}(z)\right), \quad z \in \mathcal{U}\_{\mathcal{H}}$$
and applying Lemma 3, we get the sharp result
$$F\_{\left(D\_z^{-\lambda}L\_{\alpha}^nf\right)'(U)} ^{\prime} \frac{D\_z^{-\lambda}L\_{\alpha}^nf(z)}{z} \le F\_{\mathcal{S}(U)}g(z), \quad z \in \mathcal{U}.$$
**Example 1.** *Consider*
$$g(z) = \frac{-2z}{1+z}$$
*a convex function in U and we obtain that g*(0) = 0*, g*0 (*z*) = <sup>−</sup><sup>2</sup> (1+*z*) 2 *. Define*
$$h(z) = g(z) + zg'(z) = \frac{-2z}{1+z} - \frac{2z}{\left(1+z\right)^2} = \frac{-2z^2 - 4z}{\left(1+z\right)^2}.$$
*Take α* = 2*, n* = 1*, f*(*z*) = *z* + *z* 2 *, z* ∈ *U, and after a short computation we obtain*
$$L\_2^1 f(z) = z + 2z^2$$
*and*
$$D\_z^{-\lambda}L\_2^1f(z) = \frac{1}{\Gamma(\lambda)}\int\_0^z \frac{L\_2^1f(t)}{\left(z-t\right)^{1-\lambda}}dt = \frac{1}{\Gamma(\lambda)}\int\_0^z \frac{t+2t^2}{\left(z-t\right)^{1-\lambda}}dt$$
$$= \frac{1}{\Gamma(\lambda+2)}z^{1+\lambda} + \frac{4}{\Gamma(\lambda+3)}z^{2+\lambda}$$
*and differentiating it*
$$\left(D\_z^{-\lambda}L\_2^1f(z)\right)' = \frac{1}{\Gamma(\lambda+1)}z^{\lambda} + \frac{4}{\Gamma(\lambda+2)}z^{\lambda+1}.$$
*Applying Theorem 4 we get the following fuzzy differential subordination*
$$\frac{1}{\Gamma(1+\lambda)}z^{\lambda} + \frac{4}{\Gamma(2+\lambda)}z^{1+\lambda} \preccurlyeq \frac{-2z^2 - 4z}{\left(1+z\right)^2}, z \in \mathsf{U}\_{\lambda}$$
*induce the following fuzzy differential subordination*
$$\frac{1}{\Gamma(2+\lambda)}z^{\lambda} + \frac{4}{\Gamma(3+\lambda)}z^{1+\lambda} \prec\_{\mathcal{F}} \frac{-2z}{1+z}, \; z \in \mathcal{U}.$$
**Theorem 5.** *Taking a holomorphic function h, such that <sup>h</sup>*(0) = <sup>0</sup> *and* Re 1 + *zh*00(*z*) *h* 0(*z*) <sup>&</sup>gt; <sup>−</sup><sup>1</sup> 2 , *z* ∈ *U*, *when f* ∈ A, *n* ∈ N, *α* ≥ 0, *λ* > 0*, and the fuzzy differential subordination holds*
$$\left(F\_{\left(D\_z^{-\lambda}L\_u^n f\right)'(\mathcal{U})} \left(D\_z^{-\lambda}L\_u^n f(z)\right)' \le F\_{h(\mathcal{U})}h(z), \quad z \in \mathcal{U},\tag{6}$$
*then*
$$F\_{D\_z^{-\lambda}L\_a^n f(\mathcal{U})} \frac{D\_z^{-\lambda}L\_a^n f(z)}{z} \le F\_{q(\mathcal{U})}q(z), \quad z \in \mathcal{U}\_\lambda$$
*where the fuzzy best dominant q*(*z*) = <sup>1</sup> *z* R *z* 0 *h*(*t*)*dt is convex.* **Proof.** Considering Re 1 + *zh*00(*z*) *h* 0(*z*) <sup>&</sup>gt; <sup>−</sup><sup>1</sup> 2 , *z* ∈ *U*, and using Lemma 1, we deduce that *q*(*z*) = <sup>1</sup> *z* R *z* 0 *h*(*t*)*dt* is a convex function and it is a solution of the differential equation defining the fuzzy differential subordination (6) *zq*0 (*z*) + *q*(*z*) = *h*(*z*), therefore it is the fuzzy best dominant.
Differentiating *zp*(*z*) = *D*−*<sup>λ</sup> <sup>z</sup> L n α f*(*z*), we get *D*−*<sup>λ</sup> <sup>z</sup> L n α f*(*z*) 0 = *zp*0 (*z*) + *p*(*z*), *z* ∈ *U*, and (6) can be written
$$F\_{p(\mathcal{U})} \left( zp'(z) + p(z) \right) \le F\_{h(\mathcal{U})} h(z), \quad z \in \mathcal{U}.$$
Applying Lemma 3, we get
$$F\_{D\_z^{-\lambda}L\_a^nf(\mathcal{U})} \frac{D\_z^{-\lambda}L\_a^nf(z)}{z} \le F\_{q(\mathcal{U})}q(z), \ z \in \mathcal{U}.$$
**Corollary 1.** *Taking the convex function in U*, *h*(*z*) = <sup>1</sup>+(2*δ*−1)*<sup>z</sup>* 1+*z* , *with δ* ∈ [0, 1)*, when f* ∈ A *and the fuzzy differential subordination holds*
$$\left(\mathrm{F}\_{\left(D\_z^{-\lambda}L\_a^n f\right)'(\mathcal{U})}^{'}\left(\mathcal{U}\right)\left(\mathrm{D}\_z^{-\lambda}L\_a^n f(z)\right)' \leq \mathrm{F}\_{\hbar(\mathcal{U})}\hbar(z), z \in \mathcal{U},\tag{7}$$
*then*
$$F\_{D\_z^{-\lambda}L\_\alpha^n f(\mathcal{U})} \frac{D\_z^{-\lambda}L\_\alpha^n f(z)}{z} \le F\_{q(\mathcal{U})}q(z), z \in \mathcal{U}\_\alpha$$
*where the fuzzy best dominant q*(*z*) = 2*δ* − 1 + 2(1 − *δ*) ln(*z*+1) *z* , *z* ∈ *U*, *is convex.*
**Proof.** Taking *h*(*z*) = 1+(2*δ*−1)*z* 1+*z* , we obtain *h*(0) = 1, *h* 0 (*z*) = −2(1−*δ*) (1+*z*) <sup>2</sup> and *h* 00(*z*) = 4(1−*δ*) (1+*z*) 3 , therefore *Re zh*00(*z*) *h* <sup>0</sup>(*z*) + 1 <sup>=</sup> *Re* 1−*z* 1+*z* <sup>=</sup> *Re* 1−*ρ* cos *θ*−*iρ* sin *θ* 1+*ρ* cos *θ*+*iρ* sin *θ* = 1−*ρ* 2 1+2*ρ* cos *θ*+*ρ* <sup>2</sup> <sup>&</sup>gt; <sup>0</sup> <sup>&</sup>gt; <sup>−</sup><sup>1</sup> 2 . *n*
Following the same steps like in the proof of Theorem <sup>5</sup> with *<sup>p</sup>*(*z*) = *<sup>D</sup>*−*<sup>λ</sup> <sup>z</sup> L α f*(*z*) *z* , the fuzzy differential subordination (7) can be written
$$F\_{D\_z^{-\lambda}L\_{\mathfrak{a}}^{\mathfrak{n}}f(\mathsf{U})}\left(zp'(z)+p(z)\right)\leq F\_{\mathsf{h}(\mathsf{U})}\mathsf{h}(z),\quad z\in\mathsf{U}.$$
Applying Lemma 2 for *m* = 1 and *γ* = 1, we obtain
$$F\_{D\_z^{-\lambda}L\_\alpha^n f(\mathcal{U})} \frac{D\_z^{-\lambda}L\_\alpha^n f(z)}{z} \le F\_{q(\mathcal{U})}q(z),$$
where
$$q(z) = \frac{1}{z} \int\_0^z h(t)dt = \frac{1}{z} \int\_0^z \frac{1 + (2\delta - 1)t}{t + 1} dt = $$
$$2\delta - 1 + \frac{2(1 - \delta)}{z} \int\_0^z \frac{1}{t + 1} dt = 2\delta - 1 + 2(1 - \delta)\frac{\ln(z + 1)}{z}, \ z \in \mathcal{U}\_z$$
−2*z*
**Example 2.** *Consider*
*h*(*z*) = 1 + *z and we obtain that h*(0) = 0*, h*0 (*z*) = <sup>−</sup><sup>2</sup> (1+*z*) <sup>2</sup> *and h*00(*z*) = <sup>4</sup> (1+*z*) 3 *. Taking account that*
$$\operatorname{Re}\left(1+\frac{zh''(z)}{h'(z)}\right) = \operatorname{Re}\left(\frac{1-z}{1+z}\right) = \operatorname{Re}\left(\frac{1-\rho\cos\theta-i\rho\sin\theta}{1+\rho\cos\theta+i\rho\sin\theta}\right).$$
$$=\frac{1-\rho^2}{1+2\rho\cos\theta+\rho^2} > 0 > -\frac{1}{2}\rho$$
*h is a convex function in U.*
*Taking α* = 2*, n* = 1*, f*(*z*) = *z* + *z* 2 *, z* ∈ *U, as in Example 1, we have*
$$L\_2^1 f(z) = z + 2z^2$$
*and*
$$D\_z^{-\lambda}L\_2^1f(z) = \frac{1}{\Gamma(\lambda+2)}z^{1+\lambda} + \frac{4}{\Gamma(\lambda+3)}z^{2+\lambda}$$
*and differentiating it*
$$\left(D\_z^{-\lambda}L\_2^1f(z)\right)' = \frac{1}{\Gamma(\lambda+1)}z^{\lambda} + \frac{4}{\Gamma(\lambda+2)}z^{\lambda+1}.$$
*Additionally, we get*
$$q(z) = \frac{1}{z} \int\_0^z \frac{-2t}{1+t} dt = \frac{2\ln(1+z)}{z} - 2.$$
*Applying Theorem 5 we get the following fuzzy differential subordination*
$$\frac{1}{\Gamma(1+\lambda)}z^{\lambda} + \frac{4}{\Gamma(2+\lambda)}z^{1+\lambda} \preccurlyeq\_{\mathcal{F}} \frac{2z}{1+z'}, z \in \mathcal{U}\_{\lambda}$$
*induce the following fuzzy differential subordination*
$$\frac{1}{\Gamma(2+\lambda)}z^{1+\lambda} + \frac{4}{\Gamma(3+\lambda)}z^{2+\lambda} \preccurlyeq\_{\mathcal{F}} \frac{2\ln(1+z)}{z} - 2, \; z \in \mathcal{U}.$$
**Theorem 6.** *Taking a convex function g with the property g*(0) = 0 *and defining h*(*z*) = *zg*0 (*z*) + *g*(*z*)*, z* ∈ *U, when f* ∈ A, *n* ∈ N, *α* ≥ 0, *λ* > 0*, and the fuzzy differential subordination*
$$F\_{D\_z^{-\lambda}L\_a^n f(\mathsf{II})} \left( \frac{zD\_z^{-\lambda}L\_a^{n+1}f(z)}{D\_z^{-\lambda}L\_a^n f(z)} \right)' \le F\_{h(\mathsf{II})}h(z), \quad z \in \mathsf{II},\tag{8}$$
*holds, then we obtain the sharp result*
$$\mathbb{P}\_{D\_z^{-\lambda}L\_a^n f(\mathcal{U})} \frac{z D\_z^{-\lambda} L\_a^{n+1} f(z)}{D\_z^{-\lambda} L\_a^n f(z)} \le \mathbb{P}\_{\mathcal{S}(\mathcal{U})} g(z), \quad z \in \mathcal{U}.$$
**Proof.** Considering *<sup>p</sup>*(*z*) = *<sup>D</sup>*−*<sup>λ</sup> <sup>z</sup> L n*+1 *α f*(*z*) *D* −*λ <sup>z</sup> L n <sup>α</sup> f*(*z*) and differentiating it we get *zp*0 (*z*) + *p*(*z*) = *zLn*+<sup>1</sup> *α f*(*z*) *L n <sup>α</sup> f*(*z*) 0 . With this notation, inequality (8) can be written as
$$F\_{p(\mathcal{U})} \left( zp'(z) + p(z) \right) \le F\_{h(\mathcal{U})} h(z) = F\_{\mathcal{S}(\mathcal{U})} \left( zg'(z) + g(z) \right), \ z \in \mathcal{U}.$$
Applying Lemma 3, we get
$$F\_{D\_z^{-\lambda}L\_a^n f(\mathcal{U})} \frac{D\_z^{-\lambda}L\_a^{n+1}f(z)}{D\_z^{-\lambda}L\_a^n f(z)} \le F\_{\mathcal{S}(\mathcal{U})}g(z), \quad z \in \mathcal{U}.$$
**Example 3.** *Consider*
$$g(z) = \frac{-2z}{1+z}$$
*and*
$$h(z) = g(z) + zg'(z) = \frac{-2z^2 - 4z}{\left(1 + z\right)^2}$$
*as given in Example 1.*
*Taking α* = 2*, n* = 1*, f*(*z*) = *z* + *z* 2 *, z* ∈ *U, as in Example 1, we get*
$$L\_2^1 f(z) = z + 2z^2$$
*and*
$$L\_2^2 f(z) = z + 2z^2$$
*and applying Riemann–Liouville fractional integral of order λ we have*
$$D\_z^{-\lambda}L\_2^1f(z) = \frac{1}{\Gamma(\lambda+2)}z^{1+\lambda} + \frac{4}{\Gamma(\lambda+3)}z^{2+\lambda} = D\_z^{-\lambda}L\_2^2f(z).$$
*Applying Theorem 6 we get the following fuzzy differential subordination*
$$1 \preccurlyeq\_{\mathcal{F}} \frac{-2z^2 - 4z}{\left(1 + z\right)^2}, \; z \in \mathcal{U}\_{\star}$$
*induce the following fuzzy differential subordination*
$$z \preccurlyeq \frac{-2z}{1+z}, \; z \in \mathcal{U}.$$
**Theorem 7.** *Taking a convex function g with the property g*(0) = 0 *and defining h*(*z*) = *λzg*0 (*z*) + *g*(*z*), *z* ∈ *U*, *α* ≥ 0, *λ*, *δ* > 0*, when f* ∈ A *and the fuzzy differential subordination*
$$\left(F\_{\mathcal{D}\_z^{-\lambda}L\_\mathcal{H}^q(\mathcal{U})}\left(\left(\frac{\mathcal{D}\_z^{-\lambda}L\_\mathcal{u}^\eta f(z)}{z}\right)^{\delta-1}\left(D\_z^{-\lambda}L\_\mathcal{u}^\eta f(z)\right)'\right) \le F\_{\mathcal{H}(\mathcal{U})}h(z) \quad z \in \mathcal{U},\tag{9}$$
*holds, then we obtain the sharp result*
$$F\_{D\_z^{-\lambda}L\_a^n f(\mathcal{U})} \left( \frac{D\_z^{-\lambda}L\_a^n f(z)}{z} \right)^\delta \le F\_{\mathcal{S}(\mathcal{U})} g(z), \quad z \in \mathcal{U}.$$
**Proof.** Considering *<sup>p</sup>*(*z*) = *D*−*<sup>λ</sup> <sup>z</sup> L n α f*(*z*) *z δ* ∈ H[0, *λδ*], and differentiating it we obtain
$$z p'(z) = \delta \left(\frac{D\_z^{-\lambda} L\_a^n f(z)}{z}\right)^{\delta - 1} \left(D\_z^{-\lambda} L\_a^n f(z)\right)' - \delta \left(\frac{D\_z^{-\lambda} L\_a^n f(z)}{z}\right)^{\delta}.$$
$$= \delta \left(\frac{D\_z^{-\lambda} L\_a^n f(z)}{z}\right)^{\delta - 1} \left(D\_z^{-\lambda} L\_a^n f(z)\right)' - \delta p(z),$$
$$p(z) + p(z) = \left(\frac{D\_z^{-\lambda} L\_a^n f(z)}{z}\right)^{\delta - 1} \left(D\_z^{-\lambda} L\_a^n f(z)\right)', z \in \mathsf{U}.$$
*δ z* Inequality (9) can be written
and <sup>1</sup>
*zp*0
$$\mathbb{P}\_{p(\mathcal{U})}\left(\frac{1}{\delta}zp'(z)+p(z)\right) \le \mathbb{P}\_{h(\mathcal{U})}h(z) = \mathbb{P}\_{\mathcal{S}(\mathcal{U})}\{\lambda z \mathbf{g}'(z) + \mathbf{g}(z)\}, \ z \in \mathcal{U}.$$
*α*
Applying Lemma 3 for *α* = <sup>1</sup> *δ* and *m* = *λδ*, we get
$$F\_{D\_z^{-\lambda}L\_\alpha^n f(\mathcal{U})} \left( \frac{D\_z^{-\lambda}L\_\alpha^n f(z)}{z} \right)^\delta \le F\_{\mathcal{G}(\mathcal{U})} \mathcal{g}(z), \quad z \in \mathcal{U}.$$
**Example 4.** *Consider*
*and*
$$h(z) = g(z) + zg'(z) = \frac{-2z^2 - 4z}{(1+z)^2}$$
−2*z* 1 + *z*
*g*(*z*) =
*as given in Example 1.*
*Taking α* = 2*, n* = 1*, f*(*z*) = *z* + *z* 2 *, z* ∈ *U, as in Example 1, we obtain*
> *L* 1 2 *f*(*z*) = *z* + 2*z* 2
*and*
$$D\_z^{-\lambda}L\_2^1f(z) = \frac{1}{\Gamma(\lambda+2)}z^{1+\lambda} + \frac{4}{\Gamma(\lambda+3)}z^{2+\lambda}$$
*and differentiating it*
$$\left(D\_z^{-\lambda}L\_2^1f(z)\right)' = \frac{1}{\Gamma(\lambda+1)}z^{\lambda} + \frac{4}{\Gamma(\lambda+2)}z^{\lambda+1}.$$
*Applying Theorem 7 we get the following fuzzy differential subordination*
$$\left(\frac{1}{\Gamma(\lambda+2)}z^{\lambda}+\frac{4}{\Gamma(\lambda+3)}z^{1+\lambda}\right)^{\delta-1}\left(\frac{1}{\Gamma(1+\lambda)}z^{\lambda}+\frac{4}{\Gamma(2+\lambda)}z^{1+\lambda}\right)\prec\_{\mathcal{F}}\frac{-2z^2-4z}{(1+z)^2},\;z\in\mathsf{U},\;\lambda\geqslant 1$$
*induce the following fuzzy differential subordination*
$$\left(\frac{1}{\Gamma(\lambda+2)}z^{\lambda} + \frac{4}{\Gamma(\lambda+3)}z^{1+\lambda}\right)^{\delta} \prec\_{\mathcal{F}} \frac{-2z}{1+z}, \; z \in \mathcal{U}.$$
**Theorem 8.** *Considering a holomorphic function h, such that <sup>h</sup>*(0) = <sup>0</sup> *and* Re 1 + *zh*00(*z*) *h* 0(*z*) > −1 2 , *z* ∈ *U*, *when f* ∈ A, *α* ≥ 0, *λ*, *δ* > 0*, and the fuzzy differential subordination*
$$\left(\mathcal{F}\_{\mathcal{D}\_z^{-\lambda}L\_{\mathfrak{a}}^nf(\mathsf{U})}\left(\left(\frac{\mathcal{D}\_z^{-\lambda}L\_{\mathfrak{a}}^nf(z)}{z}\right)^{\delta-1}\left(\mathcal{D}\_z^{-\lambda}L\_{\mathfrak{a}}^nf(z)\right)'\right) \le \mathcal{F}\_{\mathcal{H}(\mathsf{U})}h(z), \quad z \in \mathsf{U},\tag{10}$$
*holds, then*
$$F\_{D\_z^{-\lambda}L\_a^n f(\mathcal{U})} \left(\frac{D\_z^{-\lambda}L\_a^n f(z)}{z}\right)^\delta \le F\_{q(\mathcal{U})}q(z), \quad z \in \mathcal{U}\_\nu$$
*where the fuzzy best dominant q*(*z*) = <sup>1</sup> *z* R *z* 0 *h*(*t*)*dt is convex.*
**Proof.** Considering *<sup>p</sup>*(*z*) = *D*−*<sup>λ</sup> <sup>z</sup> L n α f*(*z*) *z δ* ∈ H[0, *λδ*], after differentiating it and making an easy computation, we get
$$\frac{1}{\delta}zp'(z) + p(z) = \left(\frac{D\_z^{-\lambda}L\_a^n(z)}{z}\right)^{\delta-1} \left(D\_z^{-\lambda}L\_a^n f(z)\right)' , \ z \in \mathcal{U}\_{\lambda}$$
and inequality (10) can be written
$$\left(F\_{p(\mathcal{U})}\left(\frac{1}{\delta}zp'(z)+p(z)\right)\right)\le F\_{\hbar(\mathcal{U})}h(z),\ z\in\mathcal{U}.$$
Applying Lemma 2, we obtain
$$F\_{D\_z^{-\lambda}L\_a^n f(\mathcal{U})} \left( \frac{D\_z^{-\lambda}L\_a^n f(z)}{z} \right)^\delta \le F\_{q(\mathcal{U})} q(z), \ z \in \mathcal{U}.$$
Taking into account that Re 1 + *zh*00(*z*) *h* 0(*z*) <sup>&</sup>gt; <sup>−</sup><sup>1</sup> 2 , *z* ∈ *U*, applying Lemma 1 we obtain that *q*(*z*) = <sup>1</sup> *z* R *z* 0 *h*(*t*)*dt* is a convex function and it is a solution of the differential equation of the fuzzy differential subordination (10) *zq*0 (*z*) + *q*(*z*) = *h*(*z*), thus it is the fuzzy best dominant.
**Example 5.** *Considering*
$$h(z) = \frac{-2z}{1+z},$$
*as in Example 2, a convex function which satisfy conditions from Theorem 8, and taking α* = 2*, n* = 1*, f*(*z*) = *z* + *z* 2 *, z* ∈ *U, we obtain*
$$L\_2^1 f(z) = z + 2z^2$$
*and*
$$D\_z^{-\lambda}L\_2^1f(z) = \frac{1}{\Gamma(\lambda+2)}z^{1+\lambda} + \frac{4}{\Gamma(\lambda+3)}z^{2+\lambda}$$
*and differentiating it*
$$\left(D\_z^{-\lambda}L\_2^1f(z)\right)' = \frac{1}{\Gamma(\lambda+1)}z^{\lambda} + \frac{4}{\Gamma(\lambda+2)}z^{\lambda+1}.$$
*Additionally, we get*
$$q(z) = \frac{1}{z} \int\_0^z \frac{-2t}{1+t} dt = \frac{2\ln(1+z)}{z} - 2.$$
*Applying Theorem 8 we get the following fuzzy differential subordination*
$$\left(\frac{1}{\Gamma(\lambda+2)}z^{\lambda}+\frac{4}{\Gamma(\lambda+3)}z^{1+\lambda}\right)^{\delta-1}\left(\frac{1}{\Gamma(1+\lambda)}z^{\lambda}+\frac{4}{\Gamma(2+\lambda)}z^{1+\lambda}\right)\prec\_{\mathbb{F}}\frac{2z}{1+z},\ z\in\mathsf{U}\_{\lambda}$$
*induce the following fuzzy differential subordination*
$$\left(\frac{1}{\Gamma(\lambda+2)}z^{\lambda} + \frac{4}{\Gamma(\lambda+3)}z^{1+\lambda}\right)^{\delta} \precneq \frac{2\ln(1+z)}{z} - 2, \; z \in \mathcal{U}.$$
**Theorem 9.** *Considering a convex function g with the property g*(0) = <sup>1</sup> *λ*+1 *and defining h*(*z*) = *zg*0 (*z*) + *g*(*z*), *z* ∈ *U, λ* > 0*, α* ≥ 0*, n* ∈ N, *when f* ∈ A *and the fuzzy differential subordination*
$$\left(\mathcal{F}\_{D\_z^{-\lambda}L\_a^nf(\mathcal{U})}\left(1-\frac{D\_z^{-\lambda}L\_a^nf(z)\left(D\_z^{-\lambda}L\_a^nf(z)\right)'}{\left[\left(D\_z^{-\lambda}L\_a^nf(z)\right)'\right]^2}\right) \le \mathcal{F}\_{h(\mathcal{U})}h(z), \quad z \in \mathcal{U}\_{\mathcal{U}}$$
*holds, then we obtain the sharp result*
$$F\_{D\_z^{-\lambda}L\_a^nf(\mathcal{U})}\left(\frac{D\_z^{-\lambda}L\_a^nf(z)}{z\left(D\_z^{-\lambda}L\_a^nf(z)\right)'}\right) \le F\_{\mathcal{S}(\mathcal{U})}\mathcal{g}(z), \ z \in \mathcal{U}.\mathcal{Z}$$
**Proof.** Differentiating *<sup>p</sup>*(*z*) = *<sup>D</sup>*−*<sup>λ</sup> <sup>z</sup> L n α f*(*z*) *z*(*D* −*λ <sup>z</sup> L n <sup>α</sup> f*(*z*)) <sup>0</sup> we obtain *zp*<sup>0</sup> (*z*) + *p*(*z*) = 1− *D*−*<sup>λ</sup> <sup>z</sup> L n α f*(*z*)(*D*−*<sup>λ</sup> <sup>z</sup> L n α f*(*z*)) 00 h (*D* −*λ <sup>z</sup> L n <sup>α</sup> f*(*z*)) 0 i2 , *z* ∈ *U*.
Using this notation, the fuzzy differential subordination can be written
$$F\_{p(\mathcal{U})} \left( zp'(z) + p(z) \right) \le F\_{h(\mathcal{U})}h(z) = F\_{\mathcal{g}(\mathcal{U})} \left( zg'(z) + g(z) \right), \ z \in \mathcal{U},$$
and applying Lemma 3, we obtain the sharp result
$$\left(F\_{D\_z^{-\lambda}L\_n^nf(\mathcal{U})}\left(\frac{D\_z^{-\lambda}L\_n^nf(z)}{z\left(D\_z^{-\lambda}L\_n^nf(z)\right)'}\right) \le F\_{\mathcal{S}(\mathcal{U})}g(z), \quad z \in \mathcal{U}.\right)$$
#### **3. Conclusions**
Applying the theory of fuzzy differential subordination, we studied a subclass of analytic function *DL*F *n* (*δ*, *α*, *λ*) newly introduced regarding the operator *D*−*<sup>λ</sup> <sup>z</sup> L n α* . Several interesting properties are obtained for the defining subclass *DL*F *n* (*δ*, *α*, *λ*). New fuzzy differential subordinations are obtained for *D*−*<sup>λ</sup> <sup>z</sup> L n α* . To show how the results would be applied it is give an example. The operator *D*−*<sup>λ</sup> <sup>z</sup> L n α* introduced in Definition 8 and the subclass *DL*F *n* (*δ*, *α*, *λ*) introduced in Definition 9 can be objects in other future studies. Other subclasses of analytic functions can be introduced regarding this operator and some properties for these subclasses can be investigated regarding coefficient estimates, closure theorems, distortion theorems, neighborhoods, and the radii of starlikeness, convexity, or close-to-convexity.
The dual theory of fuzzy differential superordination introduced in [21] could be used for obtaining similar results involving the operator *D*−*<sup>λ</sup> <sup>z</sup> L n <sup>α</sup>* and the class *L* F *n* (*δ*, *α*, *λ*) which could be combined with the results presented here for sandwich-type theorems, as seen in [17].
**Funding:** This research received no external funding.
**Institutional Review Board Statement:** Not applicable.
**Informed Consent Statement:** Not applicable.
**Data Availability Statement:** Not applicable.
**Conflicts of Interest:** The author declares no conflict of interest.
#### **References**
### *Article* **Applications of Beta Negative Binomial Distribution and Laguerre Polynomials on Ozaki Bi-Close-to-Convex Functions**
**Isra Al-Shbeil <sup>1</sup> , Abbas Kareem Wanas <sup>2</sup> , Afis Saliu 3,4 and Adriana C˘ata¸s 5,\***
- Kanifing Serrekunda P.O. Box 3530, The Gambia
**Abstract:** In the present paper, due to beta negative binomial distribution series and Laguerre polynomials, we investigate a new family FΣ(*δ*, *η*, *λ*, *θ*; *h*) of normalized holomorphic and bi-univalent functions associated with Ozaki close-to-convex functions. We provide estimates on the initial Taylor–Maclaurin coefficients and discuss Fekete–Szeg˝o type inequality for functions in this family.
**Keywords:** bi-univalent function; Laguerre polynomial; coefficient bound; Fekete–Szeg˝o problem; beta negative binomial distribution; subordination
#### **1. Introduction**
*f*
Consider the set A of functions *f* which are holomorphic in the unit disk D = {|*z*| < 1} in the complex plane C, of the form:
$$f(z) = z + \sum\_{n=2}^{\infty} a\_n z^n, \quad z \in \mathbb{D}.\tag{1}$$
Let S be the subset of A which contains univalent functions in D having the form (1). As we can see in [1], due to the Koebe one-quarter theorem, every function *f* ∈ S has an inverse *f* −1 such that *f* −1 (*f*(*z*)) = *z*, (*z* ∈ D) and *f*(*f* −1 (*w*)) = *<sup>w</sup>*, (|*w*<sup>|</sup> <sup>&</sup>lt; *<sup>r</sup>*0(*f*),*r*0(*f*) <sup>≥</sup> <sup>1</sup> 4 ). With *f* on the form (1), we have
$$W^{-1}(w) = w - a\_2 w^2 + \left(2a\_2^2 - a\_3\right) w^3 - \left(5a\_2^3 - 5a\_2 a\_3 + a\_4\right) w^4 + \cdots, \quad |w| < r\_0(f). \tag{2}$$
We called a function *f* ∈ A as bi-univalent in D, if both *f* and *f* <sup>−</sup><sup>1</sup> are univalent in D. The set of bi-univalent functions in D is denoted by Σ.
In recent years, Srivastava et al. [2] reconsidered the study of holomorphic and biunivalent functions. In this sense, we pursued a kind of surveys represented by those of Ali et al. [3], Bulut et al. [4], Srivastava et al. [5] and others (see, for example, [6–18]).
The polynomial solution *φ*(*τ*) of the differential equation (see [19])
$$
\pi \phi'' + (1 + \gamma - \pi) \phi' + n\phi = 0,
$$
consists on the generalized Laguerre polynomial *L γ <sup>n</sup>* (*τ*), where *γ* > −1 and *n* is nonnegative integers.
We defined by
$$H\_{\gamma}(\tau, z) = \sum\_{n=0}^{\infty} L\_n^{\gamma}(\tau) z^n = \frac{e^{-\frac{\tau z}{1-z}}}{(1-z)^{\gamma+1}},\tag{3}$$
**Citation:** Al-Shbeil, I.; Wanas, A.K.; Saliu, A.; C˘ata¸s, A. Applications of Beta Negative Binomial Distribution and Laguerre Polynomials on Ozaki Bi-Close-to-Convex Functions. *Axioms* **2022**, *11*, 451. https:// doi.org/10.3390/axioms11090451
Academic Editors: Hans J. Haubold and Georgia Irina Oros
Received: 7 July 2022 Accepted: 29 August 2022 Published: 2 September 2022
**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.
**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).
the generating function of generalized Laguerre polynomial *L γ <sup>n</sup>* (*τ*), where *τ* ∈ R and *z* ∈ D. Similarly, the generalized Laguerre polynomials is given by the following recurrence relations:
$$L\_{n+1}^{\gamma}(\tau) = \frac{2n+1+\gamma-\tau}{n+1}L\_n^{\gamma}(\tau) - \frac{n+\gamma}{n+1}L\_{n-1}^{\gamma}(\tau) \quad (n \ge 1),$$
with the initial conditions
$$L\_0^\gamma(\tau) = 1, \quad L\_1^\gamma(\tau) = 1 + \gamma - \tau \quad \text{and} \quad L\_2^\gamma(\tau) = \frac{\tau^2}{2} - (\gamma + 2)\tau + \frac{(\gamma + 1)(\gamma + 2)}{2}. \tag{4}$$
Obviously, if *γ* = 0 the generalized Laguerre polynomial implies the simple Laguerre polynomial, i.e., *L* 0 *n* (*τ*) = *Ln*(*τ*).
Consider two functions *f* and *g* holomorphic in D. We say that the function *f* is subordinate to *g*, if there exists a function *w*, holomorphic in D with *w*(0) = 0, and |*w*(*z*)| < 1, (*z* ∈ D) such that *f*(*z*) = *g*(*w*(*z*)). We denote this relation by *f* ≺ *g* or *f*(*z*) ≺ *g*(*z*) (*z* ∈ D). In addition, if the function *g* is univalent in D, then we get the following equivalence (see [20]), *f*(*z*) ≺ *g*(*z*) ⇐⇒ *f*(0) = *g*(0) *and f*(D) ⊂ *g*(D).
From a theoretical standpoint, the Poisson, Pascal, logarithmic, binomial and Borel distributions have all been examined in some depth in geometric function theory (see for example [21–26]).
For a discrete random variable *x*, we say that it has a beta negative binomial distribution if it takes the values 0, 1, 2, 3, · · · with the probabilities
$$\frac{B(\eta+\theta,\lambda)}{B(\eta,\lambda)},\ \;\;\theta\frac{B(\eta+\theta,\lambda+1)}{B(\eta,\lambda)},\ \;\;\frac{1}{2}\theta(\theta+1)\frac{B(\eta+\theta,\lambda+2)}{B(\eta,\lambda)},\ \cdots,\ \;\;\;\theta$$
respectively, where *η*, *θ* and *λ* are the parameters.
$$\begin{split} \text{Prob}(\mathbf{x} = \boldsymbol{\tau}) &= \binom{\boldsymbol{\theta} + \boldsymbol{\tau} - 1}{\boldsymbol{\tau}} \frac{B(\boldsymbol{\eta} + \boldsymbol{\theta}, \boldsymbol{\lambda} + \boldsymbol{\tau})}{B(\boldsymbol{\eta}, \boldsymbol{\lambda})} \\ &= \frac{\Gamma(\boldsymbol{\theta} + \boldsymbol{\tau})}{\boldsymbol{\tau}!\Gamma(\boldsymbol{\theta})} \frac{\Gamma(\boldsymbol{\eta} + \boldsymbol{\theta})\Gamma(\boldsymbol{\lambda} + \boldsymbol{\tau})\Gamma(\boldsymbol{\eta} + \boldsymbol{\lambda})}{\Gamma(\boldsymbol{\eta} + \boldsymbol{\theta} + \boldsymbol{\lambda} + \boldsymbol{\tau})\Gamma(\boldsymbol{\eta})\Gamma(\boldsymbol{\lambda})} \\ &= \frac{(\boldsymbol{\eta})\_{\boldsymbol{\theta}}(\boldsymbol{\theta})\_{\boldsymbol{\tau}}(\boldsymbol{\lambda})\_{\boldsymbol{\tau}}}{(\boldsymbol{\eta} + \boldsymbol{\lambda})\_{\boldsymbol{\theta}}(\boldsymbol{\theta} + \boldsymbol{\eta} + \boldsymbol{\lambda})\_{\boldsymbol{\tau}}\boldsymbol{\tau}!} \end{split}$$
where (*α*)*<sup>n</sup>* is the Pochhammer symbol defined by
$$(\mathfrak{a})\_{\mathfrak{n}} = \frac{\Gamma(\mathfrak{a} + \mathfrak{n})}{\Gamma(\mathfrak{a})} = \begin{cases} 1 & (\mathfrak{n} = 0), \\ \mathfrak{a}(\mathfrak{a} + 1) \dots (\mathfrak{a} + \mathfrak{n} - 1) & (\mathfrak{n} \in \mathbb{N}). \end{cases}$$
Wanas and Al-Ziadi [27] developed the following power series whose coefficients are beta negative binomial distribution probabilities:
$$\mathfrak{X}\_{\eta,\lambda}^{\theta}(z) = z + \sum\_{n=2}^{\infty} \frac{(\eta)\_{\theta}(\theta)\_{n-1}(\lambda)\_{n-1}}{(\eta + \lambda)\_{\theta}(\theta + \eta + \lambda)\_{n-1}(n-1)!} z^n \qquad (z \in \mathbb{D}; \ \eta, \lambda, \theta > 0).$$
By the well-known ratio test, we deduce that the radius of convergence of the above power series is infinity.
We recall the linear operator B*<sup>θ</sup> η*,*λ* : A −→ A, as can be found in (see [27])
$$\mathfrak{B}\_{\eta,\lambda}^{\theta}f(z) = \mathfrak{X}\_{\eta,\lambda}^{\theta}(z) \* f(z) = z + \sum\_{n=2}^{\infty} \frac{(\eta)\_{\theta}(\theta)\_{n-1}(\lambda)\_{n-1}}{(\eta + \lambda)\_{\theta}(\theta + \eta + \lambda)\_{n-1}(n-1)!} a\_{\theta} z^n \qquad z \in \mathbb{D}\_{\prime}$$
where (∗) represents the Hadamard product (or convolution) of two series.
#### **2. Main Results**
We open the main section by introducing the family FΣ(*δ*, *η*, *λ*, *θ*; *h*) as follows:
**Definition 1.** *Suppose that* <sup>1</sup> <sup>2</sup> ≤ *δ* ≤ 1*, η*, *λ*, *θ* > 0 *and h is analytic in* D*, h*(0) = 1*. We say that the function f* ∈ Σ *is in the family* FΣ(*δ*, *η*, *λ*, *θ*; *h*) *if the following subordinations hold:*
$$\left(\frac{2\delta - 1}{2\delta + 1} + \frac{2}{2\delta + 1} \left(1 + \frac{z\left(\mathfrak{B}\_{\eta,\lambda}^{\theta}f(z)\right)^{\prime\prime}}{\left(\mathfrak{B}\_{\eta,\lambda}^{\theta}f(z)\right)^{\prime}}\right) \prec h(z)\right)$$
*and*
$$\left(\frac{2\delta - 1}{2\delta + 1} + \frac{2}{2\delta + 1} \left(1 + \frac{w\left(\mathfrak{B}\_{\eta,\lambda}^{\theta} f^{-1}(w)\right)^{\prime\prime}}{\left(\mathfrak{B}\_{\eta,\lambda}^{\theta} f^{-1}(w)\right)^{\prime}}\right) \prec h(w)\lambda$$
*where f* <sup>−</sup><sup>1</sup> *is given by* (2)*.*
For *δ* = <sup>1</sup> 2 in Definition 1, the family FΣ(*δ*, *η*, *λ*, *θ*; *h*) reduces to the family SΣ(*η*, *λ*, *θ*; *h*) of bi-starlike functions such that the following subordinations hold:
$$1 + \frac{z\left(\mathfrak{B}\_{\eta,\lambda}^{\theta}f(z)\right)''}{\left(\mathfrak{B}\_{\eta,\lambda}^{\theta}f(z)\right)'} \prec h(z)$$
and
$$1 + \frac{w\left(\mathfrak{B}^{\theta}\_{\eta,\lambda} f^{-1}(w)\right)^{\prime\prime}}{\left(\mathfrak{B}^{\theta}\_{\eta,\lambda} f^{-1}(w)\right)^{\prime}} \prec h(w).$$
**Theorem 1.** *Suppose that* <sup>1</sup> <sup>2</sup> ≤ *δ* ≤ 1 *and η*, *λ*, *θ* > 0*. If f* ∈ Σ *of the form* (1) *is in the family* FΣ(*δ*, *η*, *λ*, *θ*; *h*)*, with h*(*z*) = 1 + *e*1*z* + *e*2*z* <sup>2</sup> <sup>+</sup> · · · *, then*
$$|a\_2| \le \frac{(2\delta + 1)\Gamma(\eta + \theta + \lambda + 1)\Gamma(\eta)\Gamma(\lambda)|e\_1|}{4\theta\Gamma(\eta + \theta)\Gamma(\lambda + 1)\Gamma(\eta + \lambda)} = \frac{|e\_1|}{Y} \tag{5}$$
*and*
$$|a\_{3}| \leq \min\left\{ \max\left\{ \left| \frac{e\_{1}}{\Phi} \right|, \left| \frac{e\_{2}}{\Phi} + \frac{\Psi e\_{1}^{2}}{\Upsilon^{2}\Phi} \right| \right\}, \max\left\{ \left| \frac{e\_{1}}{\Phi} \right|, \left| \frac{e\_{2}}{\Phi} - \frac{(2\Phi - \Psi)e\_{1}^{2}}{\Upsilon^{2}\Phi} \right| \right\} \right\},\tag{6}$$
*where*
$$\begin{array}{lcl} \mathbf{Y} &=& \frac{4\theta \Gamma(\eta+\theta)\Gamma(\lambda+1)\Gamma(\eta+\lambda)}{(2\delta+1)\Gamma(\eta+\theta+\lambda+1)\Gamma(\eta)\Gamma(\lambda)}, \\\\ \cline{2-4} \Phi &=& \frac{6\theta(\theta+1)\Gamma(\eta+\theta)\Gamma(\lambda+2)\Gamma(\eta+\lambda)}{(2\delta+1)\Gamma(\eta+\theta+\lambda+2)\Gamma(\eta)\Gamma(\lambda)}, \\\\ \mathbf{Y} &=& \frac{8\theta^2 \Gamma^2(\eta+\theta)\Gamma^2(\lambda+1)\Gamma^2(\eta+\lambda)}{(2\delta+1)\Gamma^2(\eta+\theta+\lambda+1)\Gamma^2(\eta)\Gamma^2(\lambda)}. \end{array} \tag{7}$$
**Proof.** Assume that *f* ∈ FΣ(*δ*, *η*, *λ*, *θ*; *h*). Then, there exist two holomorphic functions *φ*, *ψ* : D −→ D given by
$$\phi(z) = r\_1 z + r\_2 z^2 + r\_3 z^3 + \cdots \quad (z \in \mathbb{D}) \tag{8}$$
and
$$\psi(w) = s\_1 w + s\_2 w^2 + s\_3 w^3 + \cdots \quad (w \in \mathbb{D}), \tag{9}$$
with *φ*(0) = *ψ*(0) = 0, |*φ*(*z*)| < 1, |*ψ*(*w*)| < 1, *z*, *w* ∈ D such that
$$1 + \frac{2}{2\delta + 1} \frac{z \left(\mathfrak{B}\_{\eta,\lambda}^{\theta} f(z)\right)'}{\left(\mathfrak{B}\_{\eta,\lambda}^{\theta} f(z)\right)'} = 1 + e\_1 \phi(z) + e\_2 \phi^2(z) + \cdots \tag{10}$$
and
$$1 + \frac{2}{2\delta + 1} \frac{w \left(\mathfrak{B}\_{\eta,\lambda}^{\theta} f^{-1}(w)\right)^{\prime\prime}}{\left(\mathfrak{B}\_{\eta,\lambda}^{\theta} f^{-1}(w)\right)^{\prime}} = 1 + c\_1 \psi(w) + c\_2 \psi^2(w) + \cdots \ . \tag{11}$$
Using (8)–(11), one obtains
$$1 + \frac{2}{2\delta + 1} \frac{z \left(\mathfrak{B}\_{\eta,\lambda}^{\theta} f(z)\right)'}{\left(\mathfrak{B}\_{\eta,\lambda}^{\theta} f(z)\right)'} = 1 + e\_1 r\_1 z + \left[e\_1 r\_2 + e\_2 r\_1^2\right] z^2 + \cdots \tag{12}$$
and
$$1 + \frac{2}{2\delta + 1} \frac{w \left(\mathfrak{B}\_{\eta,\lambda}^{\theta} f^{-1}(w)\right)'}{\left(\mathfrak{B}\_{\eta,\lambda}^{\theta} f^{-1}(w)\right)'} = 1 + e\_1 s\_1 w + \left[e\_1 s\_2 + e\_2 s\_1^2\right] w^2 + \dotsb \,\_\circ \tag{13}$$
Since |*φ*(*z*)| < 1 and |*ψ*(*w*)| < 1, *z*, *w* ∈ D, we deduce
*rj* ≤ 1 and *sj* ≤ 1 (*j* ∈ N). (14)
In view of (12) and (13), after simplifying, we obtain
$$\frac{4\theta\Gamma(\eta+\theta)\Gamma(\lambda+1)\Gamma(\eta+\lambda)}{(2\delta+1)\Gamma(\eta+\theta+\lambda+1)\Gamma(\eta)\Gamma(\lambda)}a\_2 = e\_1r\_{1\prime}\tag{15}$$
$$\frac{8\theta(\theta+1)\Gamma(\eta+\theta)\Gamma(\lambda+2)\Gamma(\eta+\lambda)}{(2\delta+1)\Gamma(\eta+\theta+\lambda+2)\Gamma(\eta)\Gamma(\lambda)}a\_3 - \frac{8\theta^2\Gamma^2(\eta+\theta)\Gamma^2(\lambda+1)\Gamma^2(\eta+\lambda)}{(2\delta+1)\Gamma^2(\eta+\theta+\lambda+1)\Gamma^2(\eta)\Gamma^2(\lambda)}a\_2^2 \quad \text{(16)}$$
$b = e\_1r\_2 + e\_2r\_{1\prime}^2$
$$-\frac{4\theta\Gamma(\eta+\theta)\Gamma(\lambda+1)\Gamma(\eta+\lambda)}{(2\delta+1)\Gamma(\eta+\theta+\lambda+1)\Gamma(\eta)\Gamma(\lambda)}a\_2 = e\_1s\_1\tag{17}$$
and
$$\frac{8\theta(\theta+1)\Gamma(\eta+\theta)\Gamma(\lambda+2)\Gamma(\eta+\lambda)}{(2\delta+1)\Gamma(\eta+\theta+\lambda+2)\Gamma(\eta)\Gamma(\lambda)}\left(2a\_2^2-a\_3\right)-\frac{8\theta^2\Gamma^2(\eta+\theta)\Gamma^2(\lambda+1)\Gamma^2(\eta+\lambda)}{(2\delta+1)\Gamma^2(\eta+\theta+\lambda+1)\Gamma^2(\eta)\Gamma^2(\lambda)}a\_2^2\tag{18}$$
$a = e\_1s\_2 + e\_2s\_1^2$ .
From (15) and (17), we derive inequality (5). Applying (7), then (15) and (16) become
$$\Psi \Upsilon a\_2 = e\_1 r\_1, \quad \Phi a\_3 - \Psi a\_2^2 = e\_1 r\_2 + e\_2 r\_1^2 \tag{19}$$
which yields
$$\frac{\Phi}{e\_1} a\_3 = r\_2 + \left(\frac{e\_2}{e\_1} + \frac{\Psi e\_1}{\Upsilon^2}\right) r\_{1'}^2 \tag{20}$$
and on using the known sharp result ([28], p. 10):
$$|r\_2 - \mu r\_1^2| \le \max\{1, |\mu|\}\tag{21}$$
for all *µ* ∈ C, we obtain
$$\left|\frac{\Phi}{e\_1}\right| |a\_3| \le \max\left\{ 1, \left|\frac{e\_2}{e\_1} + \frac{\Psi e\_1}{\Upsilon^2}\right| \right\}.\tag{22}$$
Similarly, (17) and (18) become
$$-\Upsilon a\_2 = e\_1 s\_1, \quad \Phi(2a\_2^2 - a\_3) - \Psi a\_2^2 = e\_1 s\_2 + e\_2 s\_1^2. \tag{23}$$
These equalities provide
$$-\frac{\Phi}{e\_1}a\_3 = s\_2 + \left(\frac{e\_2}{e\_1} - \frac{(2\Phi - \Psi)e\_1}{\Upsilon^2}\right)s\_1^2. \tag{24}$$
Applying (21), we deduce
$$\left|\frac{\Phi}{\varrho\_1}\right| |a\_3| \le \max\left\{ 1, \left|\frac{e\_2}{e\_1} - \frac{(2\Phi - \Psi)e\_1}{\Upsilon^2}\right| \right\}.\tag{25}$$
Inequality (6) follows from (22) and (25).
Furthermore, we use the generating function (3) of the generalized Laguerre polynomials *L γ <sup>n</sup>* (*τ*) as *h*(*z*). As a consequence, from (4), we obtain *e*<sup>1</sup> = 1 + *γ* − *τ* and *e*<sup>2</sup> = *<sup>τ</sup>* 2 <sup>2</sup> − (*γ* + 2)*τ* + (*γ*+1)(*γ*+2) 2 , and then, Theorem 1 is reduced to the following corollary.
**Corollary 1.** *If f* ∈ Σ *of the form* (1) *is in the class* FΣ(*δ*, *η*, *λ*, *θ*; *Hγ*(*τ*, *z*))*, then*
$$|a\_2| \le \frac{(2\delta + 1)\Gamma(\eta + \theta + \lambda + 1)\Gamma(\eta)\Gamma(\lambda)|1 + \gamma - \tau|}{4\theta\Gamma(\eta + \theta)\Gamma(\lambda + 1)\Gamma(\eta + \lambda)} = \frac{|1 + \gamma - \tau|}{Y}$$
*and*
$$\begin{split} |a\_{3}| &\leq \min\left\{ \max\left\{ \left| \frac{1+\gamma-\tau}{\Phi} \right|, \frac{\left| \frac{\tau^{2}}{2} - (\gamma+2)\tau + \frac{(\gamma+1)(\gamma+2)}{2}} + \frac{\Psi(1+\gamma-\tau)^{2}}{\Upsilon^{2}\Phi} \right| \right\} \right\} \\ &\quad \max\left\{ \left| \frac{1+\gamma-\tau}{\Phi} \right|, \left| \frac{\frac{\tau^{2}}{2} - (\gamma+2)\tau + \frac{(\gamma+1)(\gamma+2)}{2}} - \frac{(2\Phi-\Psi)(1+\gamma-\tau)^{2}}{\Upsilon^{2}\Phi} \right| \right\} \end{split}$$
*for all δ*, *η*, *λ*, *θ such that* <sup>1</sup> <sup>2</sup> ≤ *δ* ≤ 1 *and η*, *λ*, *θ* > 0*, where* Υ, Φ, Ψ *are defined by* (7) *and Hγ*(*τ*, *z*) *is given by* (3)*.*
In the following theorem, we develop "the Fekete–Szeg˝o Problem" for the family FΣ(*δ*, *η*, *λ*, *θ*; *h*).
**Theorem 2.** *If f* ∈ Σ *of the form* (1) *is in the class* FΣ(*δ*, *η*, *λ*, *θ*; *h*)*, then*
*a*<sup>3</sup> − *ηa* 2 2 ≤ |*e*1| Φ min max 1, *e*2 *e*1 + (Ψ + *η*Φ)*e*<sup>1</sup> Υ2 , max 1, *e*2 *e*1 − (2Φ − Ψ − *η*Φ)*e*<sup>1</sup> Υ2 , (26)
*for all δ*, *η*, *λ*, *θ such that* <sup>1</sup> <sup>2</sup> ≤ *δ* ≤ 1 *and η*, *λ*, *θ* > 0*, where* Υ, Φ, Ψ *are defined by* (7)*.*
**Proof.** According to the notations from the proof of Theorem 1 and from (19) and (20), we obtain
$$
\mu a\_3 - \eta a\_2^2 = \frac{e\_1}{\Phi} \left( r\_2 + \left( \frac{e\_2}{e\_1} + \frac{(\Psi + \eta \Phi)e\_1}{\Upsilon^2} \right) r\_1^2 \right). \tag{27}
$$
Applying the well-known sharp result |*r*<sup>2</sup> − *µr* 2 1 | ≤ max{1, |*µ*|}, one obtains
$$|a\_3 - \eta a\_2^2| \le \frac{|e\_1|}{\Phi} \max\left\{ 1, \left| \frac{e\_2}{e\_1} + \frac{(\Psi + \eta \Phi)e\_1}{\Upsilon^2} \right| \right\}. \tag{28}$$
Similarly, from (23) and (24), we derive
$$a\_3 - \eta a\_2^2 = -\frac{e\_1}{\Phi} \left( s\_2 + \left( \frac{e\_2}{e\_1} - \frac{(2\Phi - \Psi - \eta \Phi)e\_1}{Y^2} \right) s\_1^2 \right) \tag{29}$$
and in view of |*s*<sup>2</sup> − *µs* 2 1 | ≤ max{1, |*µ*|}, we get
$$|a\_3 - \eta a\_2^2| \le \frac{|e\_1|}{\Phi} \max\left\{ 1, \left| \frac{e\_2}{e\_1} - \frac{(2\Phi - \Psi - \eta \Phi)e\_1}{Y^2} \right| \right\}. \tag{30}$$
Inequality (26) follows from (28) and (30).
**Corollary 2.** *If f* ∈ Σ *of the form* (1) *is in the class* FΣ(*δ*, *η*, *λ*, *θ*; *Hγ*(*τ*, *z*))*, then*
$$\begin{aligned} & \left| a\_3 - \eta a\_2^2 \right| \\ & \le \quad \frac{|1 + \gamma - \tau|}{\Phi} \min \left\{ \max \left\{ 1, \left| \frac{\frac{\tau^2}{2} - (\gamma + 2)\tau + \frac{(\gamma + 1)(\gamma + 2)}{2}}{1 + \gamma - \tau} + \frac{(\Psi + \eta \Phi)(1 + \gamma - \tau)}{\mathbf{Y}^2} \right| \right\}, \end{aligned} $$
$$ \max \left\{ 1, \left| \frac{\frac{\tau^2}{2} - (\gamma + 2)\tau + \frac{(\gamma + 1)(\gamma + 2)}{2}}{1 + \gamma - \tau} - \frac{(2\Phi - \Psi - \eta \Phi)(1 + \gamma - \tau)}{\mathbf{Y}^2} \right| \right\}, $$
*for all δ*, *η*, *λ*, *θ such that* <sup>1</sup> <sup>2</sup> ≤ *δ* ≤ 1 *and η*, *λ*, *θ* > 0*, where* Υ, Φ, Ψ *are given by* (7) *and Hγ*(*τ*, *z*) *is given by* (3)*.*
#### **3. Conclusions**
In the present survey, we considered a certain class of bi-univalent functions, denoted by FΣ(*δ*, *η*, *λ*, *θ*; *h*), representable in the form of a Hadamard product of two power series. The coefficients of the first one, developed by Wanas and Al-Ziadi in [27], are beta negative binomial distribution probabilities. Furthermore, the Fekete–Szeg˝o Problem was developed, by making use of the newly introduced family. Consequently, inequalities of Fekete–Szeg˝o type were obtained in the special case of generalized Laguerre polynomials.
**Author Contributions:** Conceptualization, I.A.-S., A.K.W. and A.C.; Formal analysis, I.A.-S., A.K.W., A.C. and A.S.; Investigation, I.A.-S., A.K.W., A.C. and A.S.; Methodology, I.A.-S., A.K.W. and A.C.; Validation, I.A.-S., A.K.W. and A.C.; Writing—original draft, I.A.-S., A.K.W.; Writing—review and editing, I.A.-S., A.K.W. and A.C. All authors have read and agreed to the published version of the manuscript.
**Funding:** This research received no external funding.
**Institutional Review Board Statement:** Not applicable.
**Informed Consent Statement:** Not applicable.
**Data Availability Statement:** Not applicable
**Conflicts of Interest:** The authors declare no conflict of interest.
#### **References**
### *Article* **Geometric Study of 2D-Wave Equations in View of K-Symbol Airy Functions**
**Samir B. Hadid 1,2,† and Rabha W. Ibrahim 3,\* ,†**
**Abstract:** The notion of *k*-symbol special functions has recently been introduced. This new concept offers many interesting geometric properties for these special functions including logarithmic convexity. The aim of the present paper is to exploit essentially two-dimensional wave propagation in the earth-ionosphere wave path using *k*-symbol Airy functions (KAFs) in the open unit disk. It is shown that the standard wave-mode working formula may be determined by orthogonality considerations without the use of intricate justifications of the complex plane. By taking into account the symmetry-convex depiction of the KAFs, the formula combination is derived.
**Keywords:** analytic function; inequalities; univalent function; open unit disk; symmetric differential operator; airy functions; normalization; complex wave equation; k-symbol calculus
**MSC:** 30C45; 30C15; 33C10
#### **1. Introduction**
When Diaz and Pariguan [1] were assessing Feynman integrals, they introduced and researched *k*-gamma functions. Because they provide a generic integral representation of the relevant functions, these integrals are fundamentally important in high-energy physics [2]. *K*-gamma functions have since been developed which have a variety of consequences for mathematics and applications. In light of significant applications in quantum chemistry, Karwowski and Witek [3] employed *k*-special functions for determining the solution of the complex Schrodinger equation for the harmonium and similar designs. In their collected papers, there is a great deal of attention to the theory of measurement and combination versions for the *k*-maximizing factorial numbers that are used as examples as well as to the combinatorics of the Pochhammer *k*-symbol.
*K*-gamma functions were employed for combination analysis by Lackner and Lackner [4] in light of significant applications in statistics. Applications of various *k*-gamma function types have eliminated the major concerns, and, as a result, multiple publications analyzing *k*-gamma functions have been made available. Fractional calculus plays a vital role in simulating real-world issues [5]. It is perhaps surprising that *k*-gamma functions and associated *k*-Pochhammer symbols are also used in the field of fractional calculus functions. Fractional kinetic equations, including *k*-Mittag–Leffler functions, have been solved by Agarwal et al. [6]. In [7], Set et al. employed the *k*-calculus equivalent of the Riemann– Liouville singular kernel. More in-depth discussion can be found in [8,9]. Review of the literature on *k*-gamma functions has led us to conclude that, on the one hand, *k*-gamma functions have stimulated the study of mathematical ideas using novel methods, and on
**Citation:** Hadid, S.B.; Ibrahim, R.W. Geometric Study of 2D-Wave Equations in View of K-Symbol Airy Functions. *Axioms* **2022**, *11*, 590. https://doi.org/10.3390/ axioms11110590
Academic Editor: Georgia Irina Oros
Received: 3 October 2022 Accepted: 17 October 2022 Published: 26 October 2022
**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.
**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).
the other hand, that the application of these functions in diverse situations is fundamental. The *k*-symbol calculus has recently been proposed as a tool for modifying, generalizing, and analyzing classes of analytic functions, such as differential, integral, and convolution operators in the open unit disk [10–13].
Airy functions (AFs), which are the solutions of ℵ <sup>00</sup>(*ξ*) − *ξ*ℵ(*ξ*) = 0, and Legendre functions, are frequently used in place of the propagating wave functions in the approximate solution due to their asymptotic expansions. In their investigation on the optics of a raindrop, Olivier and Soares provided a thorough justification for the Airy hypothesis [14]. The theory of electromagnetic diffraction, the propagation of radio waves, the propagation of light, and physical optics are all fields in which AFs play a vital role. Additionally, they are often employed in research, as described in [15]. Applications of AFs are discussed in relation to the two characteristics of symmetry and convexity. Studies using radiation exploit the symmetry characteristic (see [16–18]). The convexity feature is used in lens research (see [19–21]).
To solve a complex *k*-symbol wave equation on the open unit disk, we use the characteristics of *k*-symbol Airy functions. We first give the *k*-symbol Airy functions in the normalized form in order to describe how the solution of the wave equation behaves. Investigation of the geometric characteristics is made easier by this. We establish that the normalized formula has several interesting special functions. We then locate the symmetryconvex representation of the KAFs to investigate the propagation of two-dimensional waves in a complicated domain. To acquire the univalent solution, which is crucial for solving the complex wave equation, we seek to demonstrate a set of necessary conditions. It is demonstrated that the fundamental working formula for the wave theory may be derived from orthogonality considerations without the need for a thorough explanation in the complex plane. The formula is coupled with consideration of the symmetry-convex representation of the KAFs. The approach is presented in Section 2, the findings are detailed and discussed in Section 3, and conclusions are drawn in Section 4.
#### **2. Approaches**
Different ideas that are considered in the conclusion are covered below.
#### *2.1. Normalized Airy Function*
The Airy functions are formulated by the integral structure
$$\aleph(\xi) = \int\_{-\infty}^{+\infty} \exp(i[\xi t + t^3/3])dt$$
achieving the power series
$$\begin{split} \mathfrak{N}\_{1}(\xi) &= \left(\frac{1}{3^{2/3}\pi}\right) \sum\_{n=0}^{\infty} \left(\frac{3^{n/3}\Gamma(\frac{n+1}{3})\sin\left(\frac{2(n+1)\pi}{3}\right)}{\Gamma(n+1)}\right) \mathfrak{E}^{n} \\ &= \left(\frac{1}{3^{2/3}\pi}\right) \left(\Gamma(\frac{1}{3})\sin\left(\frac{2\pi}{3}\right)\right) + \left(\frac{1}{3^{2/3}\pi}\right) \left(3^{1/3}\Gamma(\frac{2}{3})\sin\left(\frac{4\pi}{3}\right)\right) \mathfrak{E}^{n} \\ &+ \left(\frac{1}{3^{2/3}\pi}\right) \sum\_{n=2}^{\infty} \left(\frac{3^{n/3}\Gamma(\frac{n+1}{3})\sin\left(\frac{2(n+1)\pi}{3}\right)}{\Gamma(n+1)}\right) \mathfrak{E}^{n} \\ &= \frac{1}{(3^{2/3}\Gamma(2/3))} - \frac{\mathfrak{E}}{(3^{1/3}\Gamma(1/3))} + \frac{\mathfrak{E}^{3}}{(6\times 2^{2/3}\Gamma(2/3))} - \frac{\mathfrak{E}^{4}}{(12(3^{1/3}\Gamma(1/3)))} + O(\mathfrak{E}^{5}) \end{split}$$
and
$$\begin{split} \aleph\_{2}(\xi) &= \left(\frac{1}{3^{1/6}\pi}\right) \sum\_{n=0}^{\infty} \left(\frac{3^{n/3}\Gamma\left(\frac{n+1}{3}\right) \left|\sin\left(\frac{2(n+1)\pi}{3}\right)\right|}{\Gamma(n+1)}\right) \xi^{n} \\ &= \left(\frac{1}{3^{1/6}\pi}\right) \left(\Gamma(\frac{1}{3}) \Big|\sin\left(\frac{2\pi}{3}\right)\Big|\, \Big| + \left(\frac{1}{3^{1/6}\pi}\right) \Big(3^{1/3}\Gamma(\frac{2}{3}) \Big|\sin\left(\frac{4\pi}{3}\right)\Big|\, \Big|\, \xi\right) \\ &+ \left(\frac{1}{3^{1/6}\pi}\right) \sum\_{n=2}^{\infty} \left(\frac{3^{n/3}\Gamma(\frac{n+1}{3}) \left|\sin\left(\frac{2(n+1)\pi}{3}\right)\right|}{\Gamma(n+1)}\right) \xi^{n} \\ &= \frac{1}{3^{1/6}\Gamma(2/3)} + \frac{3^{1/6}\xi}{\Gamma(1/3)} + \frac{\xi^{3}}{6\times 3^{1/6}\Gamma(2/3)} + \frac{\xi^{4}}{4\times 3^{5/6}\Gamma(1/3)} + O(\xi^{5}). \end{split}$$
By setting *g*(0) = 0 and *g* 0 (0) = 1, we aim to normalize Airy functions. We can examine the geometrical structure of these functions using this technique. The normalized power series are as follows:
$$\begin{aligned} \mathbb{Y}\_1(\boldsymbol{\xi}) &= \left( \frac{\aleph\_1(\boldsymbol{\xi}) - \left( \frac{1}{(3^{2/3}\Gamma(2/3))} \right)}{\left( - \frac{1}{(3^{1/3}\Gamma(1/3))} \right)} \right) \\ &= \boldsymbol{\xi} - \frac{\hat{\mathsf{S}}^3 \Gamma(1/3)}{(6(3^{1/3}\Gamma(2/3)))} + \dots \\ &:= \boldsymbol{\xi} + \sum\_{n=2}^{\infty} y\_n \hat{\mathsf{s}}^n \end{aligned}$$
where
$$y\_n := \left(\frac{3^{(n-1)/3} \Gamma(\frac{n+1}{3}) \sin\left(\frac{2(n+1)\pi}{3}\right)}{\Gamma(\frac{2}{3}) \sin\left(\frac{4\pi}{3}\right) \Gamma(n+1)}\right)$$
$$= -\frac{2 \times 3^{n-3/2} \sin(2/3\pi(n+1)) \Gamma((n+1)/3)}{(\Gamma(2/3)\Gamma(n+1))};$$
and
$$\begin{split} \mathbb{V}\_{2}(\xi) &= \left( \frac{\aleph\_{2}(\xi) - \left( \frac{1}{3^{1/6}\pi} \right) \left( \Gamma\left(\frac{1}{3} \right) \left| \sin\left(\frac{2\pi}{3} \right) \right| \right)}{\left( \frac{1}{3^{1/6}\pi} \right) \left( 3^{1/3} \Gamma\left(\frac{2}{3} \right) \left| \sin\left(\frac{4\pi}{3} \right) \right| \right)} \right) \\ &= \widetilde{\xi} + \sum\_{n=2}^{\infty} \left( \frac{3^{(n-1)/3} \Gamma\left(\frac{n+1}{3} \right) \left| \sin\left(\frac{2(n+1)\pi}{3} \right) \right|}{\Gamma\left(\frac{2}{3} \right) \left| \sin\left(\frac{4\pi}{3} \right) \right| \Gamma(n+1)} \right) \widetilde{\xi}^{n} \\ &= \widetilde{\xi} + \frac{\widetilde{\xi}^{3} \Gamma(1/3)}{\left( 6 \times 3^{1/3} \Gamma(2/3) \right)} + \dots \\ &= \widetilde{\xi} + \sum\_{n=2}^{\infty} |y\_{n}| \widetilde{\xi}^{n}. \end{split}$$
#### *2.2. K-Symbol Calculus*
The *k*-symbol gamma function Γ*<sup>k</sup>* , often known as the motivate gamma function, is formulated as follows [1]:
$$\Gamma\_k(\mathfrak{f}) = \lim\_{n \to \infty} \frac{n! k^n (nk)^{\frac{\mathfrak{f}}{k} - 1}}{(\mathfrak{f})\_{n,k}},$$
where
$$(\mathfrak{f})\_{n,k} := \mathfrak{f}(\mathfrak{f}+k)(\mathfrak{f}+2k)\dots(\mathfrak{f}+(n-1)k).$$
and
$$(\mathfrak{f})\_{n,k} = \frac{\Gamma\_k(\mathfrak{f} + nk)}{\Gamma\_k(\mathfrak{f})}.$$
Based on the definition of Γ*<sup>k</sup>* ,, we present the normalized *k*-symbole functions as follows:
$$\begin{aligned} [\mathbb{Y}\_1]\_k(\xi) &= \xi - \frac{\xi^3 \Gamma\_k(1/3)}{(6(3^{1/3} \Gamma\_k(2/3)))} + \dots \\ &:= \xi + \sum\_{n=2}^{\infty} [y\_n]\_k \xi^n \end{aligned}$$
where
$$[y\_n]\_k := -\frac{2 \times 3^{n-3/2} \sin(2/3\pi(n+1))\Gamma\_k((n+1)/3))}{(\Gamma\_k(2/3)\Gamma\_k(n+1))}\mu$$
and
$$\begin{aligned} [\mathbb{Y}\_2]\_k(\xi) &= \xi + \frac{\xi^3 \Gamma\_k(1/3)}{(\mathfrak{G} \times \mathfrak{Z}^{1/3} \Gamma\_k(2/3))} + \dots \\ &= \xi + \sum\_{n=2}^{\infty} |[y\_n]\_k| \xi^n. \end{aligned}$$
The following outcomes demonstrate some characteristics of the *k*−symbol Airy functions (see Figure 1).
**Proposition 1.** *The following outcomes are accurate for k-special functions* •
$$\begin{split} \mathbb{E}[\mathbb{Y}\_{1}]\_{k}(\xi) &= \frac{\mathsf{G}\_{\mathsf{k}}(4/3)\mathsf{3}^{1/3}/\mathsf{G}\_{\mathsf{k}}(1/3)}{\mathsf{G}\_{\mathsf{k}}(5/3)\mathsf{3}^{2/3}/\mathsf{G}\_{\mathsf{k}}(2/3)} \\ &- \frac{\mathsf{G}\_{\mathsf{k}}(4/3)\mathsf{3}^{1/3}/\mathsf{G}\_{\mathsf{k}}(1/3)}{\mathsf{3}} \left( [I\_{-1/3}]\_{\mathsf{k}}(\frac{2\xi^{3/2}}{3})(\xi^{3/2})^{1/3} - \frac{\xi[I\_{1/3}]\_{\mathsf{k}}((2\xi^{3/2})/3)}{(\xi^{3/2})^{1/3}} \right) \end{split}$$
*where G<sup>k</sup> is the k-Barnes function satisfying G<sup>k</sup>* (*n*) = (Γ*<sup>k</sup>* (*n*))*n*−<sup>1</sup> *κ*(*n*) *(κ is the κ function) and* [*In*]*<sup>k</sup>* (*ξ*) *is the k-modified Bessel function.*
$$\text{The first-order coupling between the two-dimensional } \mathcal{N} \text{-matrices is the only possible } \mathcal{N} \text{-matrices with } \mathcal{N} = \{0, 1, 2, \dots, N\} \text{ and } \mathcal{N} = \{0, 1, 2, \dots, N\}.$$
•
$$\begin{split} [\mathbb{Y}\_{1}]\_{k}(\xi) &= -\Big( 1/3 [I\_{-1/3}]\_{k} (\mathbf{2}/\mathfrak{z}(-\xi)^{3/2}) ((-\xi)^{3/2})^{1/3} \Big) (\mathbb{G}\_{k}(4/3)\mathbf{3}^{1/3}/\mathbb{G}\_{k}(1/3) \\ &- \frac{\mathsf{G}\_{k}(4/3)\mathbf{3}^{1/3}/\mathsf{G}\_{k}(1/3)}{\mathsf{G}\_{k}(5/3)\mathbf{3}^{2/3}/\mathsf{G}\_{k}(2/3)} - \frac{\xi [I\_{1/3}]\_{k} (\mathbf{2}/\mathfrak{z}(-\xi)^{3/2}) (\mathsf{G}\_{k}(4/3)\mathbf{3}^{1/3}) / \mathsf{G}\_{k}(1/3)}{\mathfrak{z}((-\xi)^{3/2})^{1/3}} \Big). \end{split}$$
*where* [*Jn*]*<sup>k</sup>* (*ξ*) *indicates the k-Bessel function.* •
$$\mathbb{E}[\mathbb{Y}\_2]\_k(\xi) = \frac{\frac{\mathfrak{F}\_0[\mathbb{F}\_1]\_k(:4/3;\mathfrak{F}^3/\mathfrak{H})3^{1/6}}{\Gamma\_k(1/3)} - \frac{1}{\Gamma\_k(2/3)3^{1/6}} + \frac{[\_0\Gamma\_1]\_k(:2/3;\mathfrak{F}^3/\mathfrak{H})}{\Gamma\_k(2/3)3^{1/6}}}{3^{1/6}/\Gamma\_k(1/3)}$$
*where* [ <sup>0</sup>*F*1]*<sup>k</sup> represents the k-hypergeometric function.*
**Figure 1.** The graph of the normalized Airy functions Y<sup>1</sup> , Y2, respectively.
#### *2.3. K-Airy Differential Operator*
Using the normalized *k*-Airy functions, we then define the symmetric-convex differential operator. For an analytic function normalized in the open unit disk Λ := {*ξ* ∈ C : |*ξ*| < 1}, we have the following structure:
$$v(\mathfrak{f}) = \mathfrak{f} + \sum\_{n=2}^{\infty} a\_n \mathfrak{f}^n \mathfrak{f}$$
The following power series is produced using the convoluted operator (∗) and the normalized Airy function [Y1]*<sup>k</sup>* (*ξ*)
$$(\upsilon \* [\mathbb{Y}\_1)]\_k(\mathfrak{f}) = ([\mathbb{Y}\_1]\_k \* \upsilon)(\mathfrak{f}) = \mathfrak{f} + \sum\_{n=2}^{\infty} a\_n [y\_n]\_k \mathfrak{f}^n, \quad \mathfrak{f} \in \Lambda.$$
By considering the above convoluted product, we define the following normalized *k*-Airy symmetric-convex differential operator (KASCO):
$$\begin{split} [\Omega\_{\beta}]\_{k}(\xi) &= (1-\beta)\xi(\upsilon\*[\mathbb{Y}\_{1}]\_{k})'(\xi) - \beta\xi(\upsilon\*[\mathbb{Y}\_{1}]\_{k})'(-\xi) \\ &= (1-\beta)\left(\xi+\sum\_{n=2}^{\infty}na\_{n}[y\_{n}]\_{\xi}\xi^{n}\right) - \beta\left(-\xi+\sum\_{n=2}^{\infty}na\_{n}[y\_{n}]\_{k}(-1)^{n}\xi^{n}\right) \\ &= \tilde{\xi} + \sum\_{n=2}^{\infty}na\_{n}[y\_{n}]\_{k}[(1-\beta)+\beta(-1)^{n+1}]\xi^{n} \\ &:= \tilde{\xi} + \sum\_{n=2}^{\infty}na\_{n}[y\_{n}]\_{k}\omega\_{n}(\beta)\xi^{n} \quad \tilde{\xi} \in \Lambda\_{\prime} \end{split}$$
where
$$
\omega\_n(\beta) := [(1 - \beta) + \beta(-1)^{n+1}].
$$
The *m*-dimensional KASCO is illustrated as follows:
$$\begin{split} [\Omega\_{\beta}]^{2}\_{k}(\boldsymbol{\xi}) &= [\Omega\_{\beta}]\_{k}([\Omega\_{\beta}]\_{k})(\boldsymbol{\xi}) \\ &= (1-\beta)\xi([\Omega\_{\beta}]\_{k})^{\prime}(\boldsymbol{\xi}) - \beta\xi([\Omega\_{\beta}]\_{k})^{\prime}(-\boldsymbol{\xi}) \\ &= (1-\beta)\left(\boldsymbol{\xi} + \sum\_{n=2}^{\infty}n^{2}a\_{n}[y\_{n}]\_{k}\boldsymbol{\alpha}\_{n}(\boldsymbol{\beta})\boldsymbol{\xi}^{n}\right) - \beta\left(-\boldsymbol{\xi} + \sum\_{n=2}^{\infty}n^{2}a\_{n}[y\_{n}]\_{k}\boldsymbol{\alpha}\_{n}(\boldsymbol{\beta})(-1)^{n}\boldsymbol{\xi}^{n}\right) \\ &= \boldsymbol{\xi} + \sum\_{n=2}^{\infty}na\_{n}[y\_{n}]\_{k}\boldsymbol{\alpha}\_{n}(\boldsymbol{\beta})[(1-\boldsymbol{\beta}) + \beta(-1)^{n+1}]\boldsymbol{\xi}^{n} \\ &= \boldsymbol{\xi} + \sum\_{n=2}^{\infty}n^{2}a\_{n}[y\_{n}]\_{k}\boldsymbol{\alpha}\_{n}^{2}(\boldsymbol{\beta})\boldsymbol{\xi}^{n} \quad \boldsymbol{\xi} \in \Lambda. \end{split}$$
Generally, the *m*-formula is given by (see Figure 2)
$$[\Omega\_{\beta}]\_{k}^{m}(\xi) = \xi + \sum\_{n=2}^{\infty} n^{m} a\_{n} [y\_{n}]\_{k} \varpi\_{n}^{m}(\beta) \xi^{n} \quad \xi \in \Lambda. \tag{1}$$
Note that, under the consideration data *k* = 1, *β* = 0 and [*yn*]*<sup>k</sup>* ≈ 1,, this implies the Salagean differential operator [22].
**Figure 2.** The graph of KASCO, when *m* = *k* = 1, *β* = 1/2, 1/4, 3/4 accordingly.
*2.4. Univalent Solution of the k-Wave Equation*
In an effort to develop the wave equation, we suggest utilizing the parametric Koebe function. The Koebe function is an extreme function that belongs to the family of convex univalent functions. The Koebe function *σ*(*ξ*) = *ξ*/(1 − *ξ*) <sup>2</sup> maps the unit disk conformally onto the complex plane C with a slit along the disk |*ξ*| < 1/4. We utilize the rotate Koebe function of the structure
$$\sigma\_t(\xi) = \frac{\xi}{(1 - e^{it}\xi)^2} = \xi + \sum\_{n=2}^{\infty} n e^{i(n-1)t} \xi^n, \quad \xi \in \Lambda.$$
The operator Ω*<sup>k</sup> <sup>α</sup>* acts on *σ*(*ξ*), producing the following expansion
$$[\Omega\_{\beta}]\_{k}^{m}(\mathfrak{f};t) = \mathfrak{f} + \sum\_{n=2}^{\infty} n^{1+m} e^{i(n-1)t} [y\_n]\_k \varpi\_n^m(\beta) \mathfrak{f}^n \quad \mathfrak{f} \in \Lambda. \tag{2}$$
Using the operator (2), we proceed to formulate the complex wave equation. The complex wave equation is considered in the formula
$$
\left(\frac{\partial^2}{\partial t^2} + \varepsilon^2 \frac{\partial^2}{\partial \xi^2}\right) [\Omega\_{\beta}]\_k^m (\xi; t) = \Sigma(\xi),
\tag{3}
$$
where [Ω*β*] *m k* (*ξ*; *t*) indicates the *m*-iterative wave amplitude in Λ with the convex parameter *β* ∈ [0, 1] and Σ is known as the non-linear functional of the wave under consideration owing Σ(0) = 0 and Σ 0 (0) = 1 (normalized function in Λ). A unique instance is examined in [23], when Σ(*ξ*) = 0 and [Ω*β*] *m k* (*ξ*; *t*) = [Ω*β*] *<sup>m</sup>*(*ξ*; *t*).
We provide a univalent outcome to the wave equation. The univalent result is significant in wave equations (see [24–27]). The wave peaks necessarily travel faster than the troughs and ultimately reach these levels since the solutions to the wave equations are known to be erroneous for infinite layers as they are not univalent functions. The primary requirement to achieve an analytic univalent solution fulfilling the inequality is covered in the next section <([Ω*β*] *m k* (*ξ*; *t*) 0 ) > 0 where 0 = *d*/*dξ*). Alternatively, the answer is a complex domain Λ with a limited rotation function. In this instance, the gradients continue to increase, but eventually these effects start to occur, slowing this expansion. The precise behavior of the solution in Λ, which cannot be predicted from the wave equation, depends on the form of the dissipation components that are taken into account.
#### **3. Results and Discussion**
This section describes our findings for the univalent solution of Equation (3) for various hypotheses concerning Σ.
**Proposition 2.** *Consider Equation* (3)*. If the operator* [Ω*β*] *m k* (*ξ*; *t*) *fulfils the symmetrical inequality*
$$\Re \left( \frac{\mathfrak{z}[\Omega\_{\beta}]\_{k}^{m}(\mathfrak{z};t)'}{[\Omega\_{\beta}]\_{k}^{m}(\mathfrak{z};t) - [\Omega\_{\beta}]\_{k}^{m}(-\mathfrak{z};t)} \right) > 0 \tag{4}$$
*then* [Ω*β*] *m k* (*ξ*; *t*) *is a univalent outcome for Equation* (3)*.*
**Proof.** The normalization formula of [Ω*β*] *m k* (*ξ*; *t*) yields [[Ω*β*] *m k* (0; *t*) = 0 and [Ω*β*] *m k* (0; *t*) 0 = 1. Replacing −*ξ* by *ξ* in the inequality (4), we get
$$\Re \left( \frac{\mathfrak{F}[\Omega\_{\beta}]\_{k}^{m}(-\mathfrak{f};t)'}{[\Omega\_{\beta}]\_{k}^{m}(\mathfrak{f};t) - [\Omega\_{\beta}]\_{k}^{m}(-\mathfrak{f};t)} \right) > 0. \tag{5}$$
Combining inequalities (4) and (5), we receive
$$\left\|\Re\left(\frac{\mathfrak{F}\left([\Omega\_{\beta}]\_{k}^{m}(-\mathfrak{F};t)'-[\Omega\_{\beta}]\_{k}^{m}(-\mathfrak{F};t)'\right)}{[\Omega\_{\beta}]\_{k}^{m}(\mathfrak{F};t)-[\Omega\_{\beta}]\_{k}^{m}(-\mathfrak{F};t)}\right)\right\| > 0. \tag{6}$$
This shows that [Ω*β*] *m k* (*ξ*; *t*) − [Ω*β*] *m k* (−*ξ*; *t*) is univalent in Λ. In view of the Kaplan Theorem of uni-valency [28], we obtain [Ω*β*] *m k* (*ξ*; *t*) is a univalent outcome of Equation (3).
Different conditions for [Ω*β*] *m k* (*ξ*; *t*) to be univalently solvable are shown in the following outcomes.
**Proposition 3.** *For Equation* (3)*, assume that the operator* [Ω*β*] *m k* (*ξ*; *t*) *violates the relation*
$$\Re\left( [\Omega\_{\beta}]\_{k}^{m} (\mathfrak{f};t)^{\prime} + \lambda(\mathfrak{f}) [\Omega\_{\beta}]\_{k}^{m} (\mathfrak{f};t)^{\prime\prime} \right) > 0 \tag{7}$$
*where λ*(*ξ*) *is an analytic function in* Λ *with a non-negative real part. Then* [Ω*β*] *m k* (*ξ*; *t*) *is a univalent outcome for Equation* (3)*.*
**Proof.** Assume that (7) is a true inequality. Formulate an admissible function <sup>∆</sup> : <sup>C</sup><sup>2</sup> <sup>→</sup> <sup>C</sup>, as follows:
$$
\Delta(\rho,\mathfrak{g}) = \rho(\mathfrak{f}) + \lambda(\mathfrak{f})\mathfrak{g}(\mathfrak{f}).
$$
In view of the assumption (7), and by letting
$$\rho(\mathfrak{f}) := [\Omega\_{\beta}]\_{k}^{m}(\mathfrak{f};t)', \quad \mathfrak{g}(\mathfrak{f}) := \mathfrak{f}[\Omega\_{\beta}]\_{k}^{m}(\mathfrak{f};t)'', \mathfrak{f}$$
we have that
$$\mathfrak{R}\left(\Delta\left([\Omega\_{\beta}]\_{k}^{m}(\mathfrak{f};t)',\mathfrak{f}[\Omega\_{\beta}]\_{k}^{m}(\mathfrak{f};t)''\right)\right) > 0.$$
According to Theorem 5 of [29], we conclude that
$$\mathfrak{R}\left([\Omega\_{\beta}]\_{k}^{m}(\mathfrak{J};t)'\right) > 0,$$
which leads to [Ω*β*] *m k* (*ξ*; *t*) is a univalent solution of Equation (3).
Extra conditions on [Ω*β*] *m k* (*ξ*; *t*) to be univalent. The following outcome is a relation between [Ω*β*] *m k* (*ξ*; *t*) and Σ(*ξ*) in Equation (3).
**Proposition 4.** *Assume that Equation* (3)*, where* Σ(*ξ*) *is a bounded function in* Λ*, with*
$$\inf \left( \frac{\Sigma(\mathfrak{f}\_1) - \Sigma(\mathfrak{f}\_2)}{\mathfrak{f}\_1 - \mathfrak{f}\_2} \right) > 0, \quad \mathfrak{f}\_1, \mathfrak{f}\_2 \in \Lambda.$$
*If*
$$\left| \frac{\mathfrak{f}}{[\Omega\_{\beta}]\_{k}^{m}(\mathfrak{f};t)} - \frac{\mathfrak{f}}{\Sigma(\mathfrak{f})} \right| \leq \frac{2 \inf \left( \frac{\Sigma(\mathfrak{f}\_{1}) - \Sigma(\mathfrak{f}\_{2})}{\mathfrak{f}\_{1} - \mathfrak{f}\_{2}} \right)}{[\sup\_{\mathfrak{f} \in \Lambda} (\Sigma(\mathfrak{f}))]^{2}};$$
*which leads to* [Ω*β*] *m k* (*ξ*; *t*) *is a univalent solution for Equation* (3)*.*
**Proof.** Let [Ω*β*] *m k* (*ξ*; *t*) = *ξ* + ∑ ∞ *<sup>n</sup>*=<sup>2</sup> *ϑnζ <sup>n</sup>* and Σ(*ξ*) = *ξ* + ∑ ∞ *<sup>n</sup>*=<sup>2</sup> *ϕnξ n* . Formulate the function *F* : Λ → Λ, as follows: .
$$F(\mathfrak{f}) = [\frac{\mathfrak{f}}{[\Omega\_{\beta}]\_{k}^{m}(\mathfrak{f};t)} - \frac{\mathfrak{f}}{\Sigma(\mathfrak{f})}]''$$
Clearly, *F*(*ξ*) is analytic in Λ. Integrating both sides, we get
$$[\frac{\mathfrak{J}}{[\Omega\_{\beta}]\_{k}^{m}(\mathfrak{J};t)} - \frac{\mathfrak{J}}{\Sigma(\mathfrak{J})}]' = \varrho\_{2} - \mathfrak{\theta}\_{2} + \int\_{0}^{\mathfrak{J}} F(\tau)d\tau.$$
Consequently, we have
$$\mathbb{E}\left[\frac{\mathfrak{F}}{[\Omega\_{\beta}]\_{k}^{m}(\mathfrak{F};t)} - \frac{\mathfrak{F}}{\Sigma(\mathfrak{F})}\right] = (\varphi\_{2} - \vartheta\_{2})\mathfrak{F} + \int\_{0}^{\mathfrak{F}} ds \int\_{0}^{s} F(\tau)d\tau.\ \mathcal{J}$$
Therefore, a calculation gives that
$$[\Omega\_{\beta}]\_{k}^{m}(\boldsymbol{\xi};t) = \frac{\boldsymbol{\Sigma}(\boldsymbol{\xi})}{1 + (\varphi\_{2} - \vartheta\_{2})\boldsymbol{\Sigma}(\boldsymbol{\xi}) + \boldsymbol{\Sigma}(\boldsymbol{\xi})(f(\boldsymbol{\xi})/\boldsymbol{\xi})} \boldsymbol{\xi}$$
where
$$f(\xi) = \int\_0^{\xi} ds \int\_0^s F(\tau)d\tau.$$
A calculation yields that
$$\left(\frac{f(\xi)}{\widetilde{\xi}}\right)' = \frac{1}{\widetilde{\xi}^2} \int\_0^\zeta t f''(t) dt = \frac{1}{\widetilde{\xi}^2} \int\_0^\zeta \tau F(\tau) d\tau.$$
By virtue of the assumption, we have
$$\begin{split} \left| \frac{f(\tilde{\xi}\_{2})}{\tilde{\xi}\_{2}} - \frac{f(\tilde{\xi}\_{1})}{\tilde{\xi}\_{1}} \right| &= \left| \int\_{\tilde{\xi}\_{1}}^{\tilde{\xi}\_{2}} \left( \frac{f(\tilde{\xi})}{\tilde{\xi}} \right)' d\tilde{\xi} \right| \\ &\leq \left( \frac{2 \inf \left( \frac{\Sigma(\tilde{\xi}\_{1}) - \Sigma(\tilde{\xi}\_{2})}{\tilde{\xi}\_{1} - \tilde{\xi}\_{2}} \right)}{[\sup\_{\tilde{\xi} \in \Lambda} (\Sigma(\tilde{\xi}))]^{2}} \right) \left( \frac{|\tilde{\xi}\_{2} - \tilde{\xi}\_{1}|}{2} \right) . \end{split}$$
where *ξ*<sup>1</sup> 6= *ξ*2. The next step is to prove that [Ω*β*] *m k* (*ξ*1; *t*) 6= [Ω*β*] *m k* (*ξ*2; *t*) or
$$\left| \left[ \Omega\_{\beta} \right]\_{k}^{m} (\mathfrak{f}\_{1}; t) - [\Omega\_{\beta}]\_{k}^{m} (\mathfrak{f}\_{2}; t) \right| > 0, \quad \mathfrak{f}\_{1} \neq \mathfrak{f}\_{2}.$$
[Ω*β*] *m k* (*ξ*1; *t*) − [Ω*β*] *m k* (*ξ*2; *t*) = Σ(*ξ*1) − Σ(*ξ*2) + Σ(*ξ*2)Σ(*ξ*1) *f*(*ξ*2) *ξ*2 − *f*(*ξ*1) *ξ*1 1 + (*ϕ*<sup>2</sup> − *ϑ*2)Σ(*ξ*1) + Σ(*ξ*1) *f*(*ξ*1) *ξ*1 1 + (*ϕ*<sup>2</sup> − *ϑ*2)Σ(*ξ*2) + Σ(*ξ*2) *f*(*ξ*2) *ξ*2 > <sup>|</sup>Σ(*ξ*1) <sup>−</sup> <sup>Σ</sup>(*ξ*2)| − inf Σ(*ξ*1) − Σ(*ξ*2) *ξ*<sup>1</sup> − *ξ*<sup>2</sup> (*ξ*<sup>2</sup> − *ξ*1) 1 + (*ϕ*<sup>2</sup> − *ϑ*2)Σ(*ξ*1) + Σ(*ξ*1) *f*(*ξ*1) *ξ*1 1 + (*ϕ*<sup>2</sup> − *ϑ*2)Σ(*ξ*2) + Σ(*ξ*2) *f*(*ξ*2) *ξ*2 ≥ 0.
Consequently, we obtain that [Ω*β*] *m k* (*ξ*; *t*) is a univalent solution of Equation (3) in Λ.
Some unique examples of Proposition 4 are as follows:
**Corollary 1.** *If*
$$\left| \left( \frac{\xi}{[\Omega\_{\beta}]\_{k}^{m}(\xi;t)} \right)'' \right| \le 2\epsilon$$
*then* [Ω*β*] *m k* (*ξ*; *t*) *is a univalent solution.*
**Proof.** By putting Σ(*ξ*) = *ξ* in Proposition 4, we have the result. Note that this result is sharp when
$$[\Omega\_{\beta}]\_{k}^{m}(\mathfrak{f};t) = \frac{\mathfrak{f}}{(1+\mathfrak{f})^{2+\ell}}\,'$$
where
$$\left| \left( \frac{\mathfrak{z}}{[\Omega\_{\beta}]\_{k}^{m}(\mathfrak{z};t)} \right)'' \right| = (2+\ell)(1+\ell)(1+\xi)^{\ell}, \quad \ell > 0.$$
By Corollary 1, we have
**Corollary 2.** *If*
*where*
$$\begin{aligned} [\Omega\_{\beta}]\_k^{\mathfrak{M}}(\check{\xi};t) &= \frac{5}{1 + \sum\_{n=1}^{\infty} b\_n \check{\xi}} \\\\ \sum\_{n=1}^{\infty} n(n-1)|b\_n| &\le 2, \end{aligned}$$
*n* ,
(*ξ*; *<sup>t</sup>*) = *<sup>ξ</sup>*
*m*
*n*=2
*then* [Ω*β*] *m k* (*ξ*; *t*) *is a univalent solution.*
The concluding remarks are presented below.
#### **Remark 1.**
#### **4. Conclusions**
A symmetric-convex differential formula of normalized Airy functions in the open unit disk was developed. This equation was taken into account as a differential operator working on a class of normalized analytic functions. The proposed operator (KASCO) was shown to be a solution to a wave equation in the following phase of this inquiry. We provided the necessary requirements for KASCO to be a univalent solution because we sought to analyze the geometric shape of the solution (symmetry and convexity). Based on the theory of the wave equation of a complex variable, the univalent solution is a particularly delicate property. Based on the theory of geometric functions, this characteristic leads to several geometric presentations for the solution.
**Author Contributions:** Conceptualization, R.W.I. and S.B.H.; methodology, R.W.I.; software, S.B.H.; validation, R.W.I. and S.B.H.; formal analysis, R.W.I.; investigation, S.B.H.; writing—original draft preparation, R.W.I. and S.B.H.; funding acquisition, S.B.H. All authors have read and agreed to the published version of the manuscript.
**Funding:** This research was funded by Ajman University Fund: 2022-IRG-HBS-8.
**Institutional Review Board Statement:** Not applicable.
**Informed Consent Statement:** Not applicable.
**Data Availability Statement:** Not applicable.
**Conflicts of Interest:** The authors declare no conflict of interest.
#### **References**
### *Article* **Certain New Class of Analytic Functions Defined by Using a Fractional Derivative and Mittag-Leffler Functions**
**Mohammad Faisal Khan <sup>1</sup> , Shahid Khan 2,\* , Saqib Hussain <sup>3</sup> , Maslina Darus <sup>4</sup> and Khaled Matarneh <sup>5</sup>**
**Abstract:** Fractional calculus has a number of applications in the field of science, specially in mathematics. In this paper, we discuss some applications of fractional differential operators in the field of geometric function theory. Here, we combine the fractional differential operator and the Mittag-Leffler functions to formulate and arrange a new operator of fractional calculus. We define a new class of normalized analytic functions by means of a newly defined fractional operator and discuss some of its interesting geometric properties in open unit disk.
**Keywords:** analytic functions; fractional derivative operator; Mittag-Leffler function; convolution
**MSC:** 05A30; 30C45; 11B65; 47B38
#### **1. Introduction and Definitions**
Let A denote the class of functions *η* of the form
$$\eta(z) = z + \sum\_{n=2}^{\infty} a\_n z^n \tag{1}$$
which are analytic in the open unit disk
*U* = {*z* ∈ C : |*z*| < 1}
and satisfy the normalization condition
$$
\eta(0) = \eta'(0) - 1 = 0.
$$
Furthermore, we denote by S the subclass of A consisting of functions of the form (1), which are also univalent in *U*.
For two functions *η*, *y* ∈ A, we say that *η* subordinated to *y*, written as
$$\eta(z) \prec y(z)\_{\prime}$$
or equivalently
$$
\eta(z) = y(k(z))\_
\prime
$$
where, *k*(*z*) is the Schwarz function in *U* along with the condition, (see [1])
$$k(0) = 0 \text{ and } |k(z)| < 1.$$
**Citation:** Khan, M.F.; Khan, S.; Hussain, S.; Darus, M.; Matarneh, K. Certain New Class of Analytic Functions Defined by Using a Fractional Derivative and Mittag-Leffler Functions. *Axioms* **2022**, *11*, 655. https://doi.org/ 10.3390/axioms11110655
Academic Editor: Georgia Irina Oros
Received: 13 October 2022 Accepted: 14 November 2022 Published: 18 November 2022
**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.
**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).
If *y* is univalent in *U*, then
$$
\eta(z) \prec y(z) \Longleftrightarrow \eta(0) = y(0) \text{ and } \eta(\mathcal{U}) \subset y(\mathcal{U}).
$$
The majorization of two analytic function (*η y*) if and only if
$$\eta(z) = k(z)y(z), \ z \in \mathcal{U},$$
and also the coefficient inequality is satisfied
$$|a\_n| \le |b\_n|.$$
There exists a wide formation between the subordination and majorization [2] in *U* for established different classes including the the class of starlike functions (S ∗ ):
$$\operatorname{Re}\left(\frac{z\eta'(z)}{\eta(z)}\right) > 0, \ z \in \mathcal{U}$$
and convex functions (C):
$$1 + \operatorname{Re} \left( \frac{z \eta''(z)}{\eta'(z)} \right) > 0, \ z \in \mathcal{U}.$$
Related to classes S <sup>∗</sup> and C, we define the class P of analytic functions *m* ∈ P, which are normalized by
$$m(z) = 1 + \sum\_{n=1}^{\infty} c\_n z^n \rho$$
such that
$$\operatorname{Rem}(z) > 0 \text{ in } \mathcal{U} \text{ and } \operatorname{m}(0) = 1.$$
The convolution (∗) of *η* and *y*, defined by
$$(\eta \* y)(z) = \sum\_{n=0}^{\infty} a\_n b\_n z^n.$$
where,
$$y(z) = \sum\_{n=0}^{\infty} b\_n z^n, \qquad (z \in \mathcal{U}).$$
Srivastava et al. [3] geometrically explored the class of complex fractional operators (differential and integral) and Ibrahim [4] provided the generality for a class of analytic functions into two-dimensional fractional parameters in *U*. Number of authors used these operators to illustrate various subclasses of analytic functions, fractional analytic functions and differential equations of complex variable [5–7].
**Definition 1.** *Pochhammer symbol* (*α*)*<sup>n</sup> can be defined as:*
$$(\alpha)\_n = \mathfrak{a}(\mathfrak{a}+1)\dots(\mathfrak{a}+n-1) \text{ if } n \neq 0$$
*and*
$$(\mathfrak{a})\_n = 1 \text{ if } n = 0.$$
**Definition 2.** *The* (*α*)*<sup>n</sup> can be expressed in terms of the Gamma function as:*
$$(\mathfrak{a})\_{\mathfrak{n}} = \frac{\Gamma(\mathfrak{a} + \mathfrak{n})}{\Gamma(\mathfrak{a})}, \ (\mathfrak{n} \in \mathbb{N}).$$
In [8], Mittag-Leffler introduced Mittag-Leffler functions H*α*(*z*) as:
$$\mathcal{H}\_{\alpha}(z) = \sum\_{n=0}^{\infty} \frac{1}{\Gamma(\alpha n + 1)} z^n, \ (\alpha \in \mathbb{C}, \operatorname{Re}(\alpha)) > 0,$$
and its generalization H*α*,*β*(*z*) introduced by Wiman [9] as:
$$\mathcal{H}\_{\mathfrak{a},\mathfrak{\beta}}(z) = \sum\_{n=0}^{\infty} \frac{1}{\Gamma(\mathfrak{a}n + \beta)} z^n, \text{ (a, } \beta \in \mathbb{C}, \text{Re}(\mathfrak{a}), \text{Re}(\beta)) > 0. \tag{2}$$
Now we define the normalization of Mittag-Leffler function M*α*,*β*(*z*) as follows:
$$\begin{array}{rcl}\mathcal{M}\_{\mathfrak{a},\mathfrak{\beta}}(z)&=&z\Gamma(\mathfrak{\beta})\mathcal{H}\_{\mathfrak{a},\mathfrak{\beta}}(z)\\\mathcal{M}\_{\mathfrak{a},\mathfrak{\beta}}(z)&=&z+\sum\_{n=2}^{\infty}\frac{\Gamma(\mathfrak{\beta})}{\Gamma(\mathfrak{a}(n-1)+\mathfrak{\beta})}z^{n},\end{array} \tag{3}$$
where, *z* ∈ *U*, Re*α* > 0, *β* ∈ C\{0, −1, −2, . . . }).
A function *f* ∈ A is called bounded turning if it satisfies the condition
$$\operatorname{Re}\left(\eta^{'}(z)\right) > 0.$$
For <sup>0</sup> <sup>≤</sup> *<sup>ν</sup>* <sup>&</sup>lt; 1, let *<sup>B</sup>*(*v*) denote the class of functions *<sup>η</sup>* of the form (1), so that Re *η* 0 > *v* in *U*. The functions in *B*(*v*) are called functions of bounded turning (c.f. [1], Vol. II). Nashiro–Warschowski Theorem (see, e.g., [1], Vol. I) stated that the functions in *B*(*v*) are univalent and also close-to-convex in *U*. Now recall the definition of class R of bounded turning functions and can be defined as:
$$\mathcal{R} = \left\{ \eta \in \mathcal{A} : \eta'(z) \prec \frac{1+z}{1-z'}, \ z \in \mathcal{U} \right\}.$$
In [3], Srivastava and Owa gave definitions for fractional derivative operator and fractional integral operator in the complex *z*-plane C as follows:
The fractional integral of order *δ* is defined for a function *η*(*z*), by
$$I\_z^\delta \eta(z) \equiv I\_z^{-\delta} \eta(z) = \frac{1}{\Gamma(\delta)} \int\_0^z (z - t)^{\delta - 1} \eta(t) d(t), \ (\delta > 0).$$
The fractional derivative operator *D<sup>z</sup>* of order *δ* is defined by
$$\begin{aligned} D\_z^\delta \eta(z) &= \quad D\_z I\_z^{1-\delta} \eta(z) \\ &= \quad \frac{1}{\Gamma(1-\delta)} D\_z \int\_0^z \frac{\eta(t)}{\left(z-t\right)^\delta} d(t), \ (0 \le \delta < 1). \end{aligned}$$
where, the function *η*(*z*) is analytic in the simply-connected region of the complex *z*-plane C containing the origin, and the multiplicity of (*z* − *t*) −*δ* is removed by requiring log(*z* − *t*) to be real when (*z* − *t*) > 0.
Let *δ* > 0 and *m* be the smallest integer, and the extended fractional derivative of *η*(*z*) of order *δ* is defined as:
$$D\_z^\delta \eta(z) = D\_z^m I\_z^{m-\delta} \eta(z),\ 0 \le \delta,\ n > -1,\tag{4}$$
provided that it exists. We find from (4) that is
*D δ z z <sup>n</sup>* = Γ(*n* + 1) Γ(*n* + 1 − *δ*) *z n*−*δ* , (0 ≤ *δ* < 1, *n* > −1)
and
$$I\_z^\delta z^n = \frac{\Gamma(n+1)}{\Gamma(n+1-\delta)} z^{n+\delta}, \ (0 < \delta, \ n > -1).$$
Owa and Srivastava [10], defined the differential integral operator Ω*<sup>δ</sup> z* : A → A in the term of series:
$$\begin{aligned} \Omega\_z^\delta \eta(z) &= \frac{\Gamma(2-\delta)}{\Gamma(2)} z^\delta D\_z^\delta \eta(z) \\ &= \quad z + \sum\_{n=2}^\infty \frac{\Gamma(2-\delta)\Gamma(n+1)}{\Gamma(2)\Gamma(n+1-\delta)} a\_n z^n \end{aligned} \tag{5}$$
where,
$$(\delta < 2 \text{, and } z \in \mathcal{U}).$$
Here, *D<sup>δ</sup> <sup>z</sup>η*(*z*) represents the fractional integral of *η*(*z*) of order *δ* when −∞ < *δ* < 0 and a fractional derivative of *η*(*z*) of order *δ* when 0 ≤ *δ* < 2.
Now, by using the definition of convolution on (3) and (5), we define fractional differential integral operator D *δ*,*α*,*β z* : A → A, associated with normalized Mittag-Leffler function M*α*,*β*(*z*) as follows:
$$\mathfrak{D}\_z^{\delta,a,\beta}\eta(z) = z + \sum\_{n=2}^{\infty} \left( \frac{\Gamma(2-\delta)\Gamma(n+1)}{\Gamma(2)\Gamma(n+1-\delta)} \right) \left( \frac{\Gamma(\beta)}{\Gamma(a(n-1)+\beta)} \right) a\_n z^n \Big|\_{\beta}$$
where,
$$(\delta < 2, \text{ Re}\omega > 0, \,\beta \in \mathbb{C} \backslash \{0, -1, -2, \dots\}), \, z \in \mathcal{U}.$$
It is noted that
$$
\mathfrak{D}\_z^{0,0,1}\eta(z) = \eta(z).
$$
Again, by using fractional differential integral operator D *δ*,*α*,*β z* , we also define a linear multiplier fractional differential integral operator *<sup>α</sup> β* ∆ *δ*,*m λ* as follows:
$$
\Delta\_{\beta}^{\alpha} \Delta\_{\lambda}^{\delta, m} \eta(z) = \_{\beta}^{\alpha} \Delta\_{\lambda}^{\delta, 1} \left( \_{\beta}^{\alpha} \Delta\_{\lambda}^{\delta, m - 1} \eta(z) \right),
\tag{6}
$$
.
where,
$$\_\beta^\alpha \Delta\_\lambda^{\delta,0} \eta(z) = \eta(z)\_\lambda$$
and
$$
\pi^{\alpha}\_{\beta} \Delta^{\delta, 1}\_{\lambda} \eta(z) = (1 - \lambda) \mathfrak{D}^{\delta, \alpha, \beta}\_z \eta(z) + \lambda z \mathcal{D} \left( \mathfrak{D}^{\delta, \alpha, \beta}\_z \eta(z) \right),
$$
It is seen from *η*(*z*) given by (1) and from (6), we have
$${}^{\alpha}\_{\beta} \Delta\_{\lambda}^{\delta, m} \eta(z) = z + \sum\_{n=2}^{\infty} A(\lambda, \delta, \alpha, \beta, m, n) a\_n z^n,\tag{7}$$
where,
$$A(\lambda, \delta, a, \beta, m, n) = \left[ \left( \frac{\Gamma(2-\delta)\Gamma(n+1)}{\Gamma(2)\Gamma(n+1-\delta)} \right) \left( \frac{\Gamma(\beta)}{\Gamma(a(n-1)+\beta)} \right) (1-\lambda+n\lambda) \right]^m$$
and
$$(\delta < 2, \ m \in \mathbb{N}, \ \lambda \ge 0, \ \text{Re}\alpha > 0, \ \beta \in \mathbb{C} \\
\langle \{0, -1, -2, \dots\} \rangle, \ z \in \mathcal{U}.$$
**Remark 1.** *When, δ* = 0*, α* = 0, *and β* = 1*, in (7) then it is reduced to the operator given by Al-Oboudi [11].*
**Remark 2.** *For, δ* = 0, *λ* = 1, *α* = 0, *and β* = 1 *in (7) then it is reduced to the operator given by Salagean [12].*
**Definition 3.** *A function <sup>η</sup>* ∈ A, *is in the class <sup>β</sup> α*S ∗*δ*,*m λ* (*σ*) *if and only if*
$$\mathcal{A}\_{\alpha}^{\beta} \mathcal{S}\_{\lambda}^{\*\delta, m}(\sigma) = \left\{ \eta \in \mathcal{A} : \frac{z \left( \, ^{\alpha}\_{\beta} \Delta\_{\lambda}^{\delta, m} \eta(z) \right)' }{\, ^{\alpha}\_{\beta} \Delta\_{\lambda}^{\delta, m} \eta(z)} \prec \sigma(z), \; \sigma(0) = 1 \right\}.$$
**Definition 4.** *A function <sup>η</sup>* ∈ A, *is in the class <sup>α</sup> β J δ*,*m λ* (*L*, *M*, *b*) *if and only if*
$$\iota\_{\beta}^{a}I\_{\lambda}^{\delta,m}(L,M,b) = \left\{ \eta \in \mathcal{A} : 1 + \frac{1}{b} \left( \frac{2\left(\frac{a}{\beta}\Delta\_{\lambda}^{\delta,m}\eta(z)\right)}{\frac{a}{\beta}\Delta\_{\lambda}^{\delta,m}\eta(z) - \frac{a}{\beta}\Delta\_{\lambda}^{\delta,m}\eta(-z)} \right) \prec \frac{1+Lz}{1+Mz} \right\}.$$
The following lemmas will be use to prove our main results.
**Lemma 1** ([13])**.** *For \$* ∈ C *and a positive integer n*, *the class of analytic functions is given by*
$$\mathcal{H}(\eta, n) = \left\{ \eta : \eta(z) = \varrho + \varrho\_n z^n + \varrho\_{n+1} z^{n+1} + \dots \right\}.$$
*(i) Let l* ∈ R. *Then*
$$\operatorname{Re}\left(\eta(z) + lz\eta'(z)\right) > 0 \longrightarrow \operatorname{Re}(\eta(z)) > 0.$$
*Moreover, l* > 0 *and η* ∈ H(1, *n*), *then there is constant δ* > 0 *and k* > 0, *such that*
$$k = k(l, \delta, n)$$
*and*
$$
\eta(z) + lz\eta'(z) \prec \left(\frac{1+z}{1-z}\right)^k \longrightarrow \eta(z) \prec \left(\frac{1+z}{1-z}\right)^\delta.
$$
*(ii) For η* ∈ H(1, *n*), *and for fixed real number l* > 0 *and let c* ∈ [0, 1)*, so that*
$$\operatorname{Re}\left(\eta^2(z) + 2\eta(z)(zD\eta(z))\right) > c \longrightarrow \operatorname{Re}(\eta(z)) > l.\cdot$$
*(iii) Let η* ∈ H(*η*, *n*), *with Re*(*η*) > 0, *then*
$$\operatorname{Re}\left(\eta(z) + z\eta'(z) + z^2\eta''(z)\right) > 0,$$
*or for* N : *U* → R, *such that*
$$\operatorname{Re}\left(\eta(z) + \left(\frac{z\eta'(z)}{\eta(z)}\right)\mathbb{N}(z)\right) > 0.$$
*Then*
$$\operatorname{Re}(\eta(z)) > 0.$$
#### **2. Main Results**
To make use of Lemma 1, first of all, we illustrate differential integral operator *α β* ∆ *δ*,*m λ η*(*z*) is also bounded turning function.
**Theorem 1.** *Let η* ∈ *A, and*
$$(i)\ \;\_{\beta}^{\kappa} \Delta\_{\lambda}^{\delta, m} \eta(z) \text{ is of bounded turning function.}$$
$$(ii)\ \;\_{\beta}^{\kappa} \Delta\_{\lambda}^{\delta, m} \eta(z) \Big|\_{}^{} \prec \left(\frac{1+z}{1-z}\right)^{k}, \; k > 0, \; z \in \mathcal{U}.$$
$$(iii)\ \;Re\left(\left(\frac{\kappa}{\beta} \Delta\_{\lambda}^{\delta, m} \eta(z)\right)' \left(\frac{\frac{\kappa}{\beta} \Delta\_{\lambda}^{\delta, m} \eta(z)}{z}\right)\right) > \frac{c}{2}, \; c \in [0, 1).$$
$$(iv)\ \;Re\left(z \left(\frac{\kappa}{\beta} \Delta\_{\lambda}^{\delta, m} \eta(z)\right)' - \left(\frac{\kappa}{\beta} \Delta\_{\lambda}^{\delta, m} \eta(z)\right)' + 2\left(\frac{\frac{\kappa}{\beta} \Delta\_{\lambda}^{\delta, m} \eta(z)}{z}\right)\right) > 0.$$
$$(v)\ \;Re\left(\left(z \left(\frac{\kappa}{\beta} \Delta\_{\lambda}^{\delta, m} \eta(z)\right)' \Big|\_{}^{} \wedge\_{\alpha}^{\kappa} \Delta\_{\lambda}^{\delta, m} \eta(z)\right) + 2\left(\frac{\kappa}{\beta} \Delta\_{\lambda}^{\delta, m} \eta(z)/z\right)\right) > 1.$$
*Then*
$$\left(\frac{\,^{\alpha}\_{\beta}\Delta^{\delta,m}\_{\lambda}\eta(z)}{z}\right)\in\mathcal{P}(\lambda),\text{ for some }\lambda\in[0,1).$$
**Proof.** Define a function *m*(*z*) as follows:
$$m(z) = \frac{\,\_{\beta}^{\alpha} \Delta\_{\lambda}^{\delta, m} \eta(z)}{z}, \quad z \in \mathsf{U}. \tag{8}$$
Then computation implies that
$$zm^\prime(z) + m(z) = \left(^\alpha\_\beta \Delta^{\delta, m}\_\lambda \eta(z)\right)'.$$
From the first inequality (i), we have *<sup>α</sup> β* ∆ *δ*,*m λ η*(*z*) is bounding turning function, and this give us
$$\operatorname{Re}\left(zm'(z) + m(z)\right) > 0.$$
Thus, Lemma 1, part (i) implies that
$$\text{Re}(m(z)) > 0.$$
Hence (i) is proved. Accordingly, part (ii) is confirmed. By the virtue of Lemma 1 and part (i), let *l* > 0, such that *k* = *k*(*l*) and
$$\frac{\,\_{\beta}^{\alpha} \Delta\_{\lambda}^{\delta, m} \eta(z)}{z} \prec \left(\frac{1+z}{1-z}\right)^{l}.$$
This indicates that
$$\operatorname{Re}\left(\frac{\,^\alpha \Delta^{\delta,m}\_{\lambda}\eta(z)}{z}\right) > \lambda, \quad \lambda \in [0,1).$$
Suppose that
$$\begin{split} &\operatorname{Re}\left(m^{2}(z) + 2m(z)zm'(z)\right) \\ &= \quad \text{2Re}\left(\frac{\frac{\kappa}{\beta}\Delta\_{\lambda}^{\delta,m}\eta(z)}{z}\left(\left(\frac{\kappa}{\beta}\Delta\_{\lambda}^{\delta,m}\eta(z)\right)' - \frac{\frac{\kappa}{\beta}\Delta\_{\lambda}^{\delta,m}\eta(z)}{2z}\right)\right) > c, \; c \in [0,1). \end{split} \tag{9}$$
From the Lemma 1 and part (ii), there exists a fixed real number *l* > 0 and satisfying the condition
$$\operatorname{Re}(m(z)) > l$$
and
$$
\mathfrak{m}(z) = \frac{\, ^\alpha \Delta\_\lambda^{\delta, m} \eta(z)}{z} \in \mathcal{P}(\lambda).
$$
It follows from (9) that
$$\operatorname{Re}\left(\left(\prescript{\alpha}{\beta}\Delta\_{\lambda}^{\delta,m}\eta(z)\right)\right)' > 0.$$
Taking the derivative (8), we then obtain
$$\begin{aligned} &\operatorname{Re}\left(m(z) + zm'(z) + z^2m''(z)\right) \\ &= \quad \operatorname{Re}\left(z\left(\,^{\alpha}\_{\beta}\Delta\_{\lambda}^{\delta,m}\eta(z)\right)'' - \left(\,^{\alpha}\_{\beta}\Delta\_{\lambda}^{\delta,m}\eta(z)\right)' + 2\left(\frac{\,^{\alpha}\_{\beta}\Delta\_{\lambda}^{\delta,m}\eta(z)}{z}\right)\right) > 0. \end{aligned}$$
Hence, Lemma 1 (ii) implies that
$$\operatorname{Re}\left(\frac{\,^\alpha \Delta\_\lambda^{\delta,m} \eta(z)}{z}\right) > 0.$$
The logarithmic differentiation of (8) yields
$$\begin{aligned} &\operatorname{Re}\left(m(z) + \frac{zm'(z)}{m(z)} + z^2 m''(z)\right) \\ &= \quad \operatorname{Re}\left(\frac{z\left(\frac{\alpha}{\beta}\Delta\_{\lambda}^{\delta,m}\eta(z)\right)'}{\frac{\alpha}{\beta}\Delta\_{q,\lambda}^{\delta,m}\eta(z)} + 2\left(\frac{\frac{\alpha}{\beta}\Delta\_{\lambda}^{\delta,m}\eta(z)}{z}\right) - 1\right) > 0. \end{aligned}$$
Hence, Lemma 1 (iii) implies, where *N*(*z*) = 1,
$$\operatorname{Re}\left(\frac{\,^{\alpha}\_{\beta}\Delta^{\delta,m}\_{\lambda}\eta(z)}{z}\right) > 0.$$
Now we find the upper bounds of the operator *<sup>α</sup> β* ∆ *δ*,*m λ η*(*z*) by using the exponential integral in *U*, which provided *η* ∈ *β α*S ∗*δ*,*m λ* (*σ*) .
**Theorem 2.** *Let η* ∈ *β α*S ∗*δ*,*m λ* (*σ*) , *where σ*(*z*) *is convex in U. Then,*
$$\delta^a\_\beta \Delta^{\delta, m}\_\lambda \eta(z) \prec z \exp \int\_0^z \frac{\sigma(\phi(w)) - 1}{w} dw \,\tag{10}$$
*where, φ*(*z*) *is analytic in U having condition*
$$
\phi(0) = 0 \text{ and } |\phi(z)| < 1.
$$
*Furthermore, for* |*z*| = *ξ*, *we have*
$$\mathbb{E}\exp\int\_0^1 \frac{\sigma(\phi(-\tilde{\xi})) - 1}{w} d\tilde{\xi} \le \left| \frac{\,^\alpha\_\beta \Delta^{\delta, m}\_\lambda \eta(z)}{z} \right| \le \exp\int\_0^1 \frac{\sigma(\phi(\tilde{\xi})) - 1}{w} d\tilde{\xi}.$$
**Proof.** By the hypothesis we received the following conclusion:
$$\begin{array}{rcl} \frac{z\left(\prescript{\alpha}{\beta}\Delta\_{\lambda}^{\delta,m}\eta(z)\right)'}{\prescript{\alpha}{\beta}\Delta\_{\lambda}^{\delta,m}\eta(z)} & \prec & \sigma(z) \\\\ \frac{z\left(\prescript{\alpha}{\beta}\Delta\_{\lambda}^{\delta,m}\eta(z)\right)'}{\prescript{\alpha}{\beta}\Delta\_{\lambda}^{\delta,m}\eta(z)} & = & \sigma(\phi(z)), \; z \in \mathsf{U}\_{\lambda} \end{array}$$
and
$$\left(\frac{^\alpha\!\!\!\Delta^\delta\_\lambda}{^\beta\!\!\!\Delta^\delta\_\lambda}\eta(z)\right)' - \frac{1}{z} = \frac{\sigma(\phi(z)) - 1}{z}.\tag{11}$$
Consequently, integrating (11), we obtain
$$\log\left(\frac{\frac{a}{\beta}\Delta\_{\lambda}^{\delta,m}\eta(z)}{z}\right) = \int\_0^z \frac{\sigma(\phi(w) - 1)}{w} dw.\tag{12}$$
By the definition of subordination we attain
$$\iota\_{\beta}^{\alpha} \Delta\_{\lambda}^{\delta, m} \eta(z) \prec z \exp \int\_{0}^{z} \frac{\sigma(\Psi(w) - 1)}{w} dw.$$
Hence (10) is proved.
Note that the function *σ*(*z*) convex and symmetric with respect to real axis. That is
$$
\sigma(-\zeta|z|) \le \text{Re}\{\sigma(\Psi(\xi z))\} \le \sigma(\xi|z|) \quad (0 < \xi < 1, \ z \in \mathcal{U}),
$$
then we have the inequalities
$$
\sigma(-\mathfrak{F}) \le \sigma(-\mathfrak{F}|z|), \sigma(\mathfrak{F}|z|) \le \sigma(\mathfrak{F}).
$$
Consequently, we obtain
$$\int\_0^1 \frac{\sigma(\Psi(-\tilde{\xi}|z|)) - 1}{\tilde{\xi}} d\tilde{\xi} \le \text{Re} \int\_0^1 \frac{\sigma(\Psi(\tilde{\xi})) - 1}{\tilde{\xi}} d\tilde{\xi} \le \int\_0^1 \frac{\sigma(\Psi(\tilde{\xi}|z|)) - 1}{\tilde{\xi}} d\tilde{\xi}.$$
In the sight of Equation (12), we obtain
$$\left| \int\_0^1 \frac{\sigma(\Psi(-\mathfrak{f}|z|)) - 1}{\mathfrak{f}} d\mathfrak{f} \right| \le \log \left| \frac{\,\_0^\alpha \Delta\_\lambda^{\delta, m} \eta(z)}{z} \right| \le \int\_0^1 \frac{\sigma(\Psi(\mathfrak{f}|z|)) - 1}{\mathfrak{f}} d\mathfrak{f}.$$
which implies that
$$\left| \exp \int\_0^1 \frac{\sigma(\Psi(-\mathfrak{f}|z|)) - 1}{\mathfrak{f}} d\mathfrak{f} \le \left| \frac{\mathfrak{a}\_\lambda \Delta\_\lambda^{\delta, m} \eta(z)}{z} \right| \le \exp \int\_0^1 \frac{\sigma(\Psi(\mathfrak{f}|z|)) - 1}{\mathfrak{f}} d\mathfrak{f} .\right|$$
Hence, we have
$$\left| \exp \int\_0^1 \frac{\sigma(\Psi(-\tilde{\xi})) - 1}{\tilde{\xi}} d\tilde{\xi} \le \left| \frac{\,^a\_\beta \Delta^{\delta, m}\_\lambda \eta(z)}{z} \right| \le \exp \int\_0^1 \frac{\sigma(\Psi(\tilde{\xi})) - 1}{\tilde{\xi}} d\tilde{\xi} \right| $$
Now we investigate the sufficient condition of *η* to be in the class *<sup>α</sup> β* S ∗,*δ*,*m λ* (*σ*), where *σ* is convex univalent satisfying *σ*(0) = 1.
**Theorem 3.** *If η* ∈ *A, satisfies the inequality*
$$\left(\frac{z\left(\mathop{\boldsymbol{\alpha}}\nolimits^{\delta,m}\_{\boldsymbol{\lambda}}\eta(z)\right)'}{\mathop{\boldsymbol{z}}^{\boldsymbol{\alpha}}\_{\boldsymbol{\beta}}\nolimits^{\delta,m}\eta(z)}\left(\mathbbm{2}+\frac{z\left(\mathop{\boldsymbol{\alpha}}\nolimits^{\delta,m}\_{\boldsymbol{\lambda}}\eta(z)\right)'}{\left(\mathop{\boldsymbol{\alpha}}\nolimits^{\delta,m}\_{\boldsymbol{\lambda}}\eta(z)\right)'}\right)-\left(\frac{z\left(\mathop{\boldsymbol{\alpha}}\nolimits^{\delta,m}\_{\boldsymbol{\lambda}}\eta(z)\right)'}{\mathop{\boldsymbol{\alpha}}^{\boldsymbol{\alpha}}\_{\boldsymbol{\beta}}\nolimits^{\delta,m}\eta(z)}\right)' \prec \sigma(z)\prime$$
*then, η* ∈ *α β* S ∗,*δ*,*m λ* (*σ*).
**Proof.** Let
$$m(z) = \frac{z \left( {}^{\alpha}\_{\beta} \Delta\_{\lambda}^{\delta, m} \eta(z) \right)'}{{}^{\alpha}\_{\beta} \Delta\_{\lambda}^{\delta, m} \eta(z)}$$
and *m*(*z*) = 1 in the inequality
$$
\mathfrak{m}(z) + \mathfrak{m}(z) \left( z \mathfrak{m}'(z) \right) \prec \sigma(z),
$$
then, we obtain
$$\begin{aligned} &m(z) + m(z) \left(zm'(z)\right) \\ &= \frac{z \left(\frac{a}{\beta} \Delta^{\delta, m}\_{\lambda} \eta(z)\right)'}{\left(\frac{a}{\beta} \Delta^{\delta, m}\_{\lambda} \eta(z)\right)'} \times \left(2 + \frac{z \left(\frac{a}{\beta} \Delta^{\delta, m}\_{\lambda} \eta(z)\right)'}{\left(\frac{a}{\beta} \Delta^{\delta, m}\_{\lambda} \eta(z)\right)'} - \left(\frac{z \left(\frac{a}{\beta} \Delta^{\delta, m}\_{\lambda} \eta(z)\right)'}{\frac{a}{\beta} \Delta^{\delta, m}\_{\lambda} \eta(z)}\right)\right) \prec \sigma(z). \end{aligned}$$
This implies that
$$m(z) = \frac{z\left(^{\kappa}\_{\beta}\Delta\_{\lambda}^{\delta,m}\eta(z)\right)'}{^{\kappa}\_{\beta}\Delta\_{\lambda}^{\delta,m}\eta(z)} \prec \sigma(z)\lambda$$
that is
$$
\eta \in \left( {}^{\alpha}\_{\beta} \mathcal{S}\_{\lambda}^{\*,\delta,m}(\sigma) \right).
$$
**Corollary 1.** *Let the assumption of Theorem 3. Then,*
$$\frac{z\left(\mathop{\boldsymbol{\alpha}}\_{\boldsymbol{\beta}}\Delta\_{\lambda}^{\delta,m}\eta(z)\right)'}{\mathop{\boldsymbol{\alpha}}\_{\boldsymbol{\beta}}\Delta\_{\lambda}^{\delta,m}\eta(z)} \times \left(1 + \frac{z\left(\mathop{\boldsymbol{\alpha}}\_{\boldsymbol{\beta}}\Delta\_{\lambda}^{\delta,m}\eta(z)\right)''}{\left(\mathop{\boldsymbol{\alpha}}\_{\boldsymbol{\beta}}\Delta\_{\lambda}^{\delta,m}\eta(z)\right)'}\right) - \left(\frac{z\left(\mathop{\boldsymbol{\alpha}}\_{\boldsymbol{\beta}}\Delta\_{\lambda}^{\delta,m}\eta(z)\right)'}{\mathop{\boldsymbol{\alpha}}\_{\boldsymbol{\beta}}\Delta\_{\lambda}^{\delta,m}\eta(z)}\right)' \ll \boldsymbol{\sigma}'(z).$$
**Proof.** Let
$$m(z) = \frac{z \left( {}^{\alpha}\_{\beta} \Delta\_{\lambda}^{\delta, m} \eta(z) \right)^{\prime}}{{}^{\alpha}\_{\beta} \Delta\_{\lambda}^{\delta, m} \eta(z)}$$
.
In the view of Theorem 3, we have
$$\frac{z\left(^{\alpha}\_{\beta}\Delta^{\delta,m}\_{\lambda}\eta(z)\right)'}{^{\alpha}\_{\beta}\Delta^{\delta,m}\_{\lambda}\eta(z)} \prec \sigma(z)\_{\prime}$$
where, *σ* ∈ C. Then, by [2] (Theorem 3), we obtain
$$\boldsymbol{m}'(z) \ll \boldsymbol{\sigma}'(z)$$
for some *z* ∈ *U*, where
$$m'(z) = \frac{z\left(\mathop{\boldsymbol{\alpha}}\_{\beta}^{\boldsymbol{\alpha}}\Delta\_{\lambda}^{\delta,m}\boldsymbol{\eta}(z)\right)'}{\mathop{\boldsymbol{\alpha}}\_{\beta}^{\boldsymbol{\alpha}}\Delta\_{\lambda}^{\delta,m}\boldsymbol{\eta}(z)}\left(1 + \frac{z\left(\mathop{\boldsymbol{\alpha}}\_{\beta}^{\boldsymbol{\alpha}}\Delta\_{\lambda}^{\delta,m}\boldsymbol{\eta}(z)\right)''}{\left(\mathop{\boldsymbol{\alpha}}\_{\beta}^{\boldsymbol{\alpha}}\Delta\_{\lambda}^{\delta,m}\boldsymbol{\eta}(z)\right)'}\right) - \left(\frac{z\left(\mathop{\boldsymbol{\alpha}}\_{\beta}^{\boldsymbol{\alpha}}\Delta\_{\lambda}^{\delta,m}\boldsymbol{\eta}(z)\right)'}{\mathop{\boldsymbol{\alpha}}\_{\beta}^{\boldsymbol{\alpha}}\Delta\_{\lambda}^{\delta,m}\boldsymbol{\eta}(z)}\right).$$
It is well known that the function *σ*(*z*) = *e θz* , 1 <sup>&</sup>lt; <sup>|</sup>*θ*<sup>|</sup> <sup>≤</sup> *<sup>π</sup>* 2 is not convex in *U*, where the domain *σ*(*U*) is lima-bean (see [13], p. 123). Now, we can find the same result of Theorem 3 as follows:
**Theorem 4.** *If η* ∈ *A, it satisfies the inequality*
$$1 + \frac{z \left( {}^{\alpha}\_{\beta} \Delta\_{\lambda}^{\delta, m} \eta(z) \right)^{\prime}}{\left( {}^{\alpha}\_{\beta} \Delta\_{\lambda}^{\delta, m} \eta(z) \right)^{\prime}} \prec e^{\theta z}.$$
*Then,*
$$
\eta \in \left( {}^{\alpha}\_{\beta} \mathcal{S}\_{\lambda}^{\*,\delta,m}(e^{\theta z}) \right).
$$
**Proof.** Let
$$m(z) = \frac{z \left( {}^{\alpha}\_{\beta} \Delta\_{\lambda}^{\delta, m} \eta(z) \right)^{\prime}}{{}^{\alpha}\_{\beta} \Delta\_{\lambda}^{\delta, m} \eta(z)}$$
.
After some simple computation implies that
$$\begin{split} &m(z) + \frac{zm'(z)}{m(z)}\\ &= \quad \left(\frac{z\left(\frac{a}{\beta}\Delta^{\delta,m}\_{\lambda}\eta(z)\right)'}{\frac{a}{\beta}\Delta^{\delta,m}\_{\lambda}\eta(z)}\right)' + \frac{\left(\frac{z\left(\frac{a}{\beta}\Delta^{\delta,m}\_{\lambda}\eta(z)\right)'}{\frac{a}{\beta}\Delta^{\delta,m}\_{\lambda}\eta(z)}\right)' \left(1 + \frac{z\left(\frac{a}{\beta}\Delta^{\delta,m}\_{\lambda}\eta(z)\right)''}{\left(\frac{a}{\beta}\Delta^{\delta,m}\_{\lambda}\eta(z)\right)'}\right)}{\frac{z\left(\frac{a}{\beta}\Delta^{\delta,m}\_{\lambda}\eta(z)\right)'}{\frac{a}{\beta}\Delta^{\delta,m}\_{\lambda}\eta(z)}}\right) \\ &= \quad \left(1 + \frac{z\left(\frac{a}{\beta}\Delta^{\delta,m}\_{\lambda}\eta(z)\right)''}{\left(\frac{a}{\beta}\Delta^{\delta,m}\_{\lambda}\eta(z)\right)'}\right) \prec e^{\theta z}. \end{split}$$
This implies that (see [13], p. 123)
$$m(z) = \frac{z\left(\,^{\alpha}\_{\beta}\Delta^{\delta,m}\_{\lambda}\eta(z)\right)'}{\,^{\alpha}\_{\beta}\Delta^{\delta,m}\_{\lambda}\eta(z)} \prec e^{\theta z}$$
that is
$$
\eta \in \left( {}^{\alpha}\_{\beta} \mathcal{S}\_{\lambda}^{\*,\delta,m}(e^{\theta z}) \right).
$$
**Theorem 5.** *If η* ∈ *α β J δ*,*m λ* (*L*, *M*, *b*) , *then*
$$\mathcal{M}(z) = \frac{1}{2} [\eta(z) - \eta(-z)]$$
*satisfies*
$$1 + \frac{1}{b} \left( \frac{z \left( \, \_\beta \Delta\_{\lambda}^{\delta, m} \mathcal{B}(z) \right)'}{\left( \, \_\beta \Delta\_{\lambda}^{\delta, m} \mathcal{B}(z) \right)} \right) \prec \frac{1 + Lz}{1 + Mz'},$$
$$\text{Re}\left( \frac{z \mathcal{B}'(z)}{\mathcal{B}(z)} \right) \ge \frac{1 - \vartheta^2}{1 + \vartheta^2}, \quad |z| = \vartheta < 1.$$
**Proof.** Let *η* ∈ *α β J δ*,*m λ* (*L*, *M*, *b*) , then there occurs a function *J*(*z*) such that
$$\begin{array}{rcl} b(f(z)-1) &=& \frac{2z\left(\frac{\alpha}{\beta}\Delta\_{\lambda}^{\delta,m}\eta(z)\right)'}{\frac{\alpha}{\beta}\Delta\_{\lambda}^{\delta,m}\eta(z)-\frac{\alpha}{\beta}\Delta\_{\lambda}^{\delta,m}\eta(-z)},\\ b(f(-z)-1) &=& \frac{2z\left(\frac{\alpha}{\beta}\Delta\_{\lambda}^{\delta,m}\eta(-z)\right)'}{\frac{\alpha}{\beta}\Delta\_{q,\lambda}^{\delta,m}\eta(-z)-\frac{\alpha}{\beta}\Delta\_{q,\lambda}^{\delta,m}\eta(z)}.\end{array}$$
This confirm that
$$1 + \frac{1}{b} \left( \frac{z \left( \, \_\beta \Delta \_\lambda ^{\delta, m} \mathcal{G} (z) \right)'}{\left( \_\beta \Delta \_\lambda ^{\delta, m} \mathcal{G} (z) \right)} - 1 \right) = \frac{J(z) + J(-z)}{2}.$$
However, *J* satisfies
$$J(z) \prec \frac{1+Lz}{1+Mz}.$$
which is univalent, then we get
$$1 + \frac{1}{b} \left( \frac{z \left( {}^{\alpha}\_{\beta} \Delta\_{\lambda}^{\delta, m} \mathcal{G}(z) \right)'}{{}^{\alpha}\_{\beta} \Delta\_{\lambda}^{\delta, m} \mathcal{G}(z)} - 1 \right) \prec \frac{1 + Lz}{1 + Mz}.$$
Additionally, G(*z*) is starlike in *z*, and which implies that
$$h(z) = \frac{z\mathcal{G}(z)}{\mathcal{G}(z)} \prec \frac{1 - z^2}{1 + z^2}.$$
Hence, their exist a Schwarz function *w*(*z*), such that |*w*(*z*)| ≤ |*z*| < 1, *k*(0) = 0, we get
$$h(z) \prec \frac{1 - w(z)^2}{1 + w(z)^2}.$$
which leads to
$$w(\zeta)^2 = \frac{1 - h(\zeta)}{1 + h(\zeta)}, \ \zeta \in z, \ |\zeta| = r < 1.$$
A simple calculation yields
$$\left|\frac{1 - h(\zeta)}{1 + h(\zeta)}\right| = \left|w(\zeta)\right|^2 \le \left|\zeta\right|^2.$$
Therefore, we get the following inequalities:
$$\left| h(\zeta) - \frac{1 + |\zeta|^4}{1 - |\zeta|^4} \right|^2 \le \frac{4|\zeta|^4}{\left( 1 - |\zeta|^4 \right)^2}.$$
$$\left| h(\zeta) - \frac{1 + |\zeta|^4}{1 - |\zeta|^4} \right| \le \frac{2|\zeta|^2}{\left( 1 - |\zeta|^4 \right)^4}.$$
Thus, we have
$$\operatorname{Re}\left(\frac{z\mathcal{G}'(z)}{\mathcal{G}(z)}\right) \ge \frac{1-\vartheta^2}{1+\vartheta^2}, \quad |\mathcal{G}| = \vartheta < 1.$$
This completes the proof of Theorem 5.
**Example 1.** *Let*
$$\begin{array}{rcl} \frac{z\eta'(z)}{\eta(z)} & = & \frac{z\left(^{\alpha}\_{\beta}\Delta^{\delta,m}\_{\lambda}\eta(z)\right)'}{^{\alpha}\_{\beta}\Delta^{\delta,m}\_{\lambda}\eta(z)},\\ {^{\alpha}\_{\beta}\Delta^{\delta,m}\_{\lambda}\eta(z)} & = & \frac{z}{\left(1-z\right)^{2}}, \; \eta \in \mathcal{A}. \end{array}$$
*Then the solution of <sup>z</sup><sup>η</sup>* 0 (*z*) *<sup>η</sup>*(*z*) <sup>=</sup> <sup>1</sup>+*<sup>z</sup>* 1−*z is formulated as follows:*
$$
\pi^{\alpha}\_{\beta} \Delta^{\delta, m}\_{\lambda} \eta(z) = \frac{z}{\left(1 - z\right)^2}, \quad \eta \in \mathcal{A}.
$$
*Moreover, the solution of the equation*
$$
\eta(z) + \frac{z\eta'(z)}{\eta(z)} = \frac{1+z}{1-z}
$$
*is approximated to*
$$
\eta(z) = \frac{z}{1 - z}.
$$
#### **3. Conclusions**
Many researchers have discussed some applications of fractional differential operator in different areas of mathematics. In this paper, we combined fractional differential operator and the Mittag-Leffler functions and formulated a new operator of fractional calculus for a class of normalized functions in the open unit disk. We considered this operator on the two classes of analytic functions and investigated some of its applications in the field of geometric function theory. The suggested operator can be utilized to define some more classes of analytic functions or to generalize other types of differential operators.
**Author Contributions:** Conceptualization, S.K. and M.D.; data curation, K.M.; formal analysis, S.K. and K.M.; investigation, M.F.K.; methodology, M.F.K. and M.D.; supervision, S.H.; visualization, S.H.; writing—original draft, S.H.; writing—review and editing, M.D. All authors have read and agreed to the published version of the manuscript.
**Funding:** This study supported by UKM, under grant number: FRGS/1/2019/STG06/UKM/01/.
**Data Availability Statement:** No data were used to support this study.
**Conflicts of Interest:** The authors declare no conflict of interest.
#### **References**
MDPI St. Alban-Anlage 66 4052 Basel Switzerland Tel. +41 61 683 77 34 Fax +41 61 302 89 18 www.mdpi.com
*Axioms* Editorial Office E-mail: [email protected] www.mdpi.com/journal/axioms
MDPI St. Alban-Anlage 66 4052 Basel Switzerland
Tel: +41 61 683 77 34
www.mdpi.com ISBN 978-3-0365-6344-2 | doab | 2025-04-07T03:56:58.527298 | 2-2-2023 17:39 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "00356cdb-56f1-45a1-b97a-312d2e8aacab",
"url": "https://mdpi.com/books/pdfview/book/6636",
"author": "",
"title": "New Developments in Geometric Function Theory",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036563459",
"section_idx": 0
} |
0039f385-c9b0-4fa4-b06c-9a5b9a2397e4.0 | # **Materials and Processes for Photocatalytic and (Photo)Electrocatalytic Removal of Bio-Refractory Pollutants and Emerging Contaminants from Waters**
Edited by Annalisa Vacca Printed Edition of the Special Issue Published in *Catalysts*
www.mdpi.com/journal/catalysts
**Materials and Processes for Photocatalytic and (Photo)Electrocatalytic Removal of Bio-Refractory Pollutants and Emerging Contaminants from Waters**
| doab | 2025-04-07T03:56:58.565883 | 6-5-2022 13:34 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "0039f385-c9b0-4fa4-b06c-9a5b9a2397e4",
"url": "https://mdpi.com/books/pdfview/book/5221",
"author": "",
"title": "Materials and Processes for Photocatalytic and (Photo)Electrocatalytic Removal of Bio-Refractory Pollutants and Emerging Contaminants from Waters",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036535593",
"section_idx": 0
} |
0039f385-c9b0-4fa4-b06c-9a5b9a2397e4.1 | **Materials and Processes for Photocatalytic and (Photo)Electrocatalytic Removal of Bio-Refractory Pollutants and Emerging Contaminants from Waters**
Editor
**Annalisa Vacca**
MDPI • Basel • Beijing • Wuhan • Barcelona • Belgrade • Manchester • Tokyo • Cluj • Tianjin
*Editor* Annalisa Vacca Universita di Cagliari ` Italy
*Editorial Office* MDPI St. Alban-Anlage 66 4052 Basel, Switzerland
This is a reprint of articles from the Special Issue published online in the open access journal *Catalysts* (ISSN 2073-4344) (available at: https://www.mdpi.com/journal/catalysts/special issues/ Electrocatalytic (Photo)Electrocatalytic Removal Pollutants).
For citation purposes, cite each article independently as indicated on the article page online and as indicated below:
LastName, A.A.; LastName, B.B.; LastName, C.C. Article Title. *Journal Name* **Year**, *Volume Number*, Page Range.
**ISBN 978-3-0365-3559-3 (Hbk) ISBN 978-3-0365-3560-9 (PDF)**
© 2022 by the authors. Articles in this book are Open Access and distributed under the Creative Commons Attribution (CC BY) license, which allows users to download, copy and build upon published articles, as long as the author and publisher are properly credited, which ensures maximum dissemination and a wider impact of our publications.
The book as a whole is distributed by MDPI under the terms and conditions of the Creative Commons license CC BY-NC-ND.
| doab | 2025-04-07T03:56:58.566008 | 6-5-2022 13:34 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "0039f385-c9b0-4fa4-b06c-9a5b9a2397e4",
"url": "https://mdpi.com/books/pdfview/book/5221",
"author": "",
"title": "Materials and Processes for Photocatalytic and (Photo)Electrocatalytic Removal of Bio-Refractory Pollutants and Emerging Contaminants from Waters",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036535593",
"section_idx": 1
} |
0039f385-c9b0-4fa4-b06c-9a5b9a2397e4.2 | **Contents**
## **About the Editor**
**Annalisa Vacca**, a graduate in Chemical Engineering and PhD in Industrial Engineering, was granted the position of full professor in the field of "Fundamentals of chemical technology" in 2022. She carries out didactic and scientific research in the Department of Mechanical, Chemical and Materials Engineering at the University of Cagliari. Her research activities are focused on the field of electrochemical engineering applied to the study of processes for environmental remediation and energy conversion. In particular, her studies cover key aspects such as the catalytic activity of electrode materials and the identification of reaction mechanisms, as well as practical aspects such as the design and characterization of electrochemical reactors.
### *Editorial* **Materials and Processes for Photocatalytic and (Photo)Electrocatalytic Removal of Bio-Refractory Pollutants and Emerging Contaminants from Waters**
**Annalisa Vacca**
Dipartimento di Ingegneria Meccanica, Chimica e dei Materiali, Università di Cagliari, Piazza D'armi, 09123 Cagliari, Italy; [email protected]
This volume is focused on materials and processes for the electro- and photoelectrochemical removal of biorefractory pollutants and emerging contaminants from waters to show the importance of electrochemistry and photoelectrochemistry in offering solutions to current environmental problems. In addition, we highlight their interdisciplinarity and emphasize the fundamental and applied aspects of these methods.
The research for innovative methods for removing pollutants from water has grown along with the detection of new contaminants in water bodies, the so-called emerging pollutants (EP), that can affect both flora and fauna and human health [1]: they include products used daily in households, industry, pharmaceuticals and personal care products, gasoline additives, plasticizers and microplastics [2]. Two main issues of EP are their dynamic character, which is also connected to the improvement of detection techniques, and the difficulty of removal by conventional wastewater treatment technologies. Moreover, emerging pollutants constitute a threat—even at a trace level—because their real impact on human health is unknown.
Although there are no discharge limits for most EP up to now, the European Commission has drawn up and implemented a watch list containing several chemical contaminants that must be monitored with the aim to generate high-quality data on their concentrations in the aquatic environment and to support the risk assessments that underpin the identification of priority substances [3].
During recent years, electro- and photoelectrochemical processes have demonstrated their capacity to efficiently oxidize many of these compounds. Starting from the early 1980s, research on the electrochemical methods for treated wastewater has grown significantly, and thousands of papers now appear in the literature. Although several tests demonstrate the effectiveness of pollutant removal from synthetic and real matrices, this technology is still far from full-scale applications. Its TRL (technology readiness level) is between 4 (technology validated in the lab) and 5 (technology validated in a relevant environment) [4].
More recently, photoelectrochemical processes in which electrochemical and photochemical processes are combined has attracted increasing interest, thanks to the synergy of the two processes: the application of a bias potential improves the photochemical process and the electrochemical process is more efficient since the photo-potential generated on the semiconductor allows for the depolarizing of the cell. This is why, in the last two decades, the number of articles on photochemical wastewater treatment has quickly increased, and the publication of these articles in specific journals indicates that the technology is moving from the fundamentals to real applications [5]. Nevertheless, the TRL of the photoelectrochemical treatment of wastewater is still at the lab scale, and much more efforts are required to push this technology toward applications in the field.
Thus, this special issue contributes to this context, addressing the synthesis, characterization, and application of new materials, as well as the study of catalytic processes and reaction kinetics.
**Citation:** Vacca, A. Materials and Processes for Photocatalytic and (Photo)Electrocatalytic Removal of Bio-Refractory Pollutants and Emerging Contaminants from Waters. *Catalysts* **2021**, *11*, 666. https:// doi.org/10.3390/catal11060666
Received: 28 April 2021 Accepted: 21 May 2021 Published: 24 May 2021
**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.
**Copyright:** © 2021 by the author. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).
I thank all of the authors for their valuable contribution to this Special Issue and the editorial team at *Catalyst* for their kindness and constant support.
**Funding:** This research received no external funding.
**Conflicts of Interest:** The authors declare no conflict of interest.
#### **References**
### *Article* **Enhanced Photocatalytic Activity of Au**/**TiO2 Nanoparticles against Ciprofloxacin**
**Pedro Martins 1,2,\*, Sandro Kappert 3, Hoai Nga Le 3,4, Victor Sebastian 5,6, Klaus Kühn 3, Madalena Alves 1, Luciana Pereira 1, Gianaurelio Cuniberti 3,7,8, Manuel Melle-Franco <sup>9</sup> and Senentxu Lanceros-Méndez 1,10,11,\***
Received: 14 January 2020; Accepted: 11 February 2020; Published: 15 February 2020
**Abstract:** In the last decades, photocatalysis has arisen as a solution to degrade emerging pollutants such as antibiotics. However, the reduced photoactivation of TiO2 under visible radiation constitutes a major drawback because 95% of sunlight radiation is not being used in this process. Thus, it is critical to modify TiO2 nanoparticles to improve the ability to absorb visible radiation from sunlight. This work reports on the synthesis of TiO2 nanoparticles decorated with gold (Au) nanoparticles by deposition-precipitation method for enhanced photocatalytic activity. The produced nanocomposites absorb 40% to 55% more radiation in the visible range than pristine TiO2, the best results being obtained for the synthesis performed at 25 ◦C and with Au loading of 0.05 to 0.1 wt. %. Experimental tests yielded a higher photocatalytic degradation of 91% and 49% of ciprofloxacin (5 mg/L) under UV and visible radiation, correspondingly. Computational modeling supports the experimental results, showing the ability of Au to bind TiO2 anatase surfaces, the relevant role of Au transferring electrons, and the high affinity of ciprofloxacin to both Au and TiO2 surfaces. Hence, the present work represents a reliable approach to produce efficient photocatalytic materials and an overall contribution in the development of high-performance Au/TiO2 photocatalytic nanostructures through the optimization of the synthesis parameters, photocatalytic conditions, and computational modeling.
**Keywords:** Au-TiO2; antibiotics; emergent contaminants; nanocatalyst; photocatalysis; GFN-xTB
#### **1. Introduction**
The resilience of specific emerging pollutants such as pharmaceuticals to the traditional wastewater treatments makes them spread in variable concentrations in surface and groundwater [1]. Dissemination of antibiotics in nature is one of the most significant environmental concerns as they affect biological metabolism and induce the presence of bacterial resistance among drinking water sources [2]. Photocatalysis has received considerable attention from the scientific community as a possible solution to degrade these compounds [3,4].
Typically, the photocatalytic process takes place when a catalyst is UV irradiated and electron-hole pairs are created that will react with H2O, OH−, and O2 to generate oxidizing species such as the hydroxyl radical (OH•), superoxide radical anions (O2•−), and hydrogen peroxide (H2O2). These species will initiate a series of reactions that will degrade pollutants into harmless compounds (e.g., CO2 and H2O).
Photocatalysis presents several advantages when compared with other methods, such as the low cost, and the eco-friendly and straightforward processing conditions [5,6]. Many photocatalysts have been reported in the last decades [7,8]. Among them, titanium dioxide (TiO2) is the most studied and applied in photocatalysis, mainly because of its remarkable optical and oxidizing properties, superhydrophilicity, chemical stability, and durability [9,10]. Despite the compelling advantages of TiO2, there are also some drawbacks. One of the main hurdles is the low spectral activation of TiO2, caused by its wide bandgap (3.0–3.2 eV) excitation that only occurs under radiation in the UV or near the UV region (410–387 nm) [11].
For this reason, solar radiation cannot be efficiently used because only less than 5% of this radiation corresponds to UV [3]. Additionally, the process becomes less cost effective as the UV lamps are required to provide the radiation. Another limitation is the electron-hole pair recombination that decreases the photocatalytic efficiency [12,13].
The research developed in the last decades has been mainly devoted to surpassing those limitations by producing new and more efficient photocatalytic materials. Strategies for metallic and nonmetallic doping, co-doping [14,15], dye sensitization, semiconductor combination, co-catalyst loading, and nanocomposite materials [16,17] have been used and tested. These approaches allow us to reduce the electron-hole recombination rate and enhance the absorption of visible radiation of TiO2 by introducing intermediate energy levels inside the bandgap [18]. In this scope, several works have reported the functionalization of TiO2 nanoparticles surfaces with metals such as Au [19], Cu [20], Co [21], and Ag [22]. When irradiated, noble metals nanoparticles at the TiO2 surface can receive electrons and prevent the recombination of the photo-generated electron-hole pairs [23,24].
Metals such as Au and Ag can increase visible light absorption due to the surface plasmon resonance effect [25,26]. Gold (Au) nanoparticles have attracted considerable attention, mainly because they possess exceptional stability, nontoxicity, and biocompatibility [3]. Their properties are highly dependent on the size and shape of the nanoparticles, allowing a broad range of applications [27,28]. For instance, the literature shows that gold nanoparticles in the range of 5 to 10 nm present an enhanced catalytic activity [29,30]. In this sense, some works focused on the photocatalytic activity of Au/TiO2 nanocomposite have been published, including interesting review articles [3,29,31].
Different physical-chemical techniques have been exploited to produce Au/TiO2 nanocomposites with enhanced catalytic properties. For instance, chemical vapor deposition [32], sol-gel [33], spray pyrolysis [34], electrophoretic approach [35], deposition-precipitation (DP) [36], deposition-precipitation using urea [37], impregnation [38], hybridization [39], and surface functionalization [40], among others [41,42]. However, many of these techniques are time-consuming, and few of them have focused on the optimization of the nanocomposite and the computational modeling of its nanostructure. Thus, this work focused on the optimization of a DP, converting the Au/TiO2 nanocomposite production into a cost-effective and straightforward technique, with enhanced photocatalytic activity, under UV and visible radiation. The method optimization aims for cost reduction, using the lowest Au loading that endows visible spectra photocatalytic activity to the nanocomposite. The computational studies
provide further information about the electronic mechanism behind the enhanced photocatalytic activity of the Au/TiO2 nanocomposite, as well as the interaction with the target compound.
The target compound is the fluoroquinolone ciprofloxacin (CIP) (chemical formula in Supplementary Material, Figure S1), belonging to a class of synthetic broad-spectrum antibiotics [43], which is mostly used in medicine (e.g., tuberculosis, pneumonia, or digestive disorders). It is also one of the most prescribed fluoroquinolones in the world and studies has shown its presence in potable water and wastewater, as well as in sewage sludge at variable concentrations from milligrams to nanograms per liter [2,44].
In this work, photocatalytic efficiency during the degradation of CIP under UV and visible illumination was assessed. To the best of our knowledge, this is the first work that combines an optimization process of Au/TiO2 nanocomposite with photocatalytic experiments for CIP degradation and computational modeling that addresses the interaction between Au and TiO2 nanoparticles, as well as the interaction of CIP with the produced nanocomposites.
#### **2. Results and Discussion**
#### *2.1. Nanocomposite Characterization*
The Au/TiO2 nanocomposites were produced by nanoprecipitation method, and the temperature (25, 60, and 80 ◦C) and the Au loading (ranging from 0.025 to 0.5 wt. %) were changed to understand how these parameters affect the morphology of the nanocomposites and relate it to the photocatalytic efficiency. In this sense, scanning transmission electron microscopy-high-angle annular dark-field imaging (STEM-HAADF) analysis was performed, and the micrographs of the different nanocomposites are displayed in Figure 1.
**Figure 1.** Scanning transmission electron microscopy-high-angle annular dark-field imaging micrographs of Au/TiO2 nanocomposites synthesized with different Au loadings at 60 ◦C (**a**–**c**), and Au/TiO2 nanocomposites obtained at different temperatures with an Au loading of 0.05 wt. % (**d**–**f**).
The Au loading study (Figure 1a–c) was assessed producing different nanocomposites using the same experimental conditions (temperature = 60 ◦C) and changing the loading of gold exclusively, from 0.025 to 0.5 wt. %. The STEM-HAADF micrographs show that for the sample with 0.025 wt. % of Au (Figure 1a), the presence of Au nanoparticles over the surface of the TiO2 nanoparticles was almost inexistent (Figure 1a). With the increase of Au loading to 0.05 wt. % (Figure 1b), it was possible to observe a homogeneous distribution of predominantly small Au nanoparticles (bright contrast
nanoparticles below 5 nm in diameter) over the TiO2 nanoparticles. Similar results were obtained for 0.1 wt. % (data not shown). For the concentrations of 0.25 and 0.5 wt. % (Figure 1a–c), agglomerates of Au over theTiO2 nanoparticles (brightest areas of the micrograph) were identified as well as large Au nanoparticles. Analogously, the effect of temperature on the synthesis product was also performed maintaining all the synthesis parameters (Au loading = 0.05 wt. % yielded a homogeneous distribution and size of Au nanoparticles) and changing the temperature of the different samples. STEM-HAADF images (Figure 1d–f) indicate that although the used Au loading was the same in the three temperatures tested when the nanocomposite was synthesized at 80 ◦C, larger Au nanoparticles appeared more frequently on the nanocomposite (Figure 1f). Conversely, at lower temperatures (25 and 60 ◦C), the Au nanoparticles size was smaller (Figure 1d,e).
The study of the effect of Au loading and temperature in the nanocomposites morphology indicates that the samples produced at 60 ◦C and with an Au loading of 0.05 wt. % possessed the more homogeneous distribution and size of Au nanoparticles. In this way, a more detailed STEM-HAADF analysis (Figure 2) was performed on this sample. Figure 2a,b reveal a homogeneous dispersion of Au nanoparticles (white arrows) over TiO2 nanoparticles' surface. The representation of the sphere-like shape of Au nanoparticles in Figure 2c, where an high-resolution scanning transmission electron microscopy – high-angle annular dark field shows that single-crystal nanoparticles with high crystallinity were produced by the proposed method. Size distribution, ranging from 1 to 7 nm, and the average size of 3.2 ± 1.13 nm (Figure 2d), were quantified using Image J software applied to 400 nanoparticles. The size distribution of Au nanoparticles for synthesis at 25 ◦C and 80 ◦C is provided in Supplementary Material (Figure S2). All the images show Au nanoparticles with similar sizes, which is in good agreement with the size distribution histogram that presents a sharp size distribution.
**Figure 2.** STEM-HAADF micrographs of Au/TiO2 nanocomposites (produced at 60 ◦C and Au loading of 0.05 wt. %) at different scales (**a**) 50 and (**b**) 200 nm; detail of Au nanoparticle over TiO2 nanoparticles' surface and single Au nanoparticle amplification (inset) (**c**); size distribution of 400 Au nanoparticles with the respective average size (**d**).
The STEM-HAADF- energy-dispersive X-ray spectroscopy (EDX) measurements allowed us to identify the elements present in the Au/TiO2 sample in two different points, 1 and 2 (signaled in Figure 3a). STEM-HAADF-EDX spectra in Figure 3b in point 1 indicate the presence of Au and Cu (copper), which can be respectively addressed to Au nanoparticles and copper grid. In point 2, the signatures of Ti (titanium) and O (oxygen) were identified, corresponding to TiO2 nanoparticles. Thus, EDX measurements confirmed the presence of all the elements of the Au/TiO2 nanocomposite.
**Figure 3.** The STEM-HAADF- energy-dispersive X-ray spectroscopy (EDX) image of Au/TiO2 nanocomposites with the identification of the measured points: Au (1) and TiO2 (2) (**a**), EDX spectra with elemental identification (Au, Ti, O, and C) for points 1 and 2 (**b**). The Au/TiO2 nanocomposite synthesized at 60 ◦C and with an Au loading of 0.05 wt. % was used.
X-ray diffraction was performed to assess the crystal structure of the pure TiO2 nanoparticles and Au/TiO2 nanocomposite, Figure 4a. Both samples show the typical reflexes from anatase (25.3◦, 37.8◦, and 48.0◦) and rutile (27.49◦). There was no significant difference between the intensities or positions of the reflexes from both samples. Moreover, no reflexes of Au were detected, which can be explained by the low amount of Au present in the nanocomposite (below detection limit). Figure 4b shows the study of hydrodynamic size for TiO2 and Au/TiO2 nanocomposites obtained by dynamic light scattering (DLS). The results indicated nanoparticles diameters of 1023 nm and 342 nm, for the pristine TiO2 and the Au/TiO2 nanocomposites, respectively. The results suggest that the presence of Au nanoparticles over TiO2 nanoparticles surface may prevent the formation of nanoparticles' aggregates. On the other hand, the size distribution was broader for the nanocomposites regarding the pristine TiO2. Previous work equally showed that the presence of erbium (Er) on TiO2 nanoparticles contributed to reducing the hydrodynamic size when compared with bare TiO2 [15].
The zeta potential was studied at different pH values (3, 5, 7, 9, and 11) for TiO2 and Au/TiO2 samples and the results are displayed in Figure 4c. The pristine and the Au/TiO2 presented very similar profiles, with higher zeta potential values ≈ |20| mV for pH below 3 and 9. These data were in good agreement with the literature [45], with positive zeta potential values for acidic conditions and negative values for basic pH. The more significant difference between the two samples occurred at pH = 7, with the nanocomposite presenting higher zeta potential values than the pure TiO2. Higher zeta potential values mean that nanoparticles possess higher periphery surface charge, which promotes nanoparticles' repulsions, avoiding aggregates' formation and enhanced stability [46]. In this context, and relating it with DLS-obtained results, the smaller hydrodynamic size was probably obtained for the Au/TiO2 because repulsions endowed by Au on TiO2 nanoparticles surface prevented the formation of the aggregates.
**Figure 4.** X-ray diffraction reflexes of pristine TiO2 and Au/TiO2 nanocomposite and identification of the representative peaks for anatase (A) and rutile (R) phases (**a**); dynamic light scattering, intensity size distribution of the pristine TiO2 and the Au/TiO2 nanocomposite and respective Z-average hydrodynamic size (**b**); zeta potential measurements, performed at different pHs (3, 5, 7, 9, and 11) for pristine TiO2 nanoparticles and Au/TiO2 nanocomposite (**c**); UV–vis reflectance spectra of pristine TiO2 and Au/TiO2 and (inset) the estimation of the bandgap for both samples at (F(R))1/2 = 0 (**d**). The Au/TiO2 nanocomposite synthesized at 60 ◦C and with an Au loading of 0.05 wt. % was used.
To understand the differences in the photocatalytic performance of TiO2 and Au/TiO2 nanocomposite, the optical properties of these materials were studied by UV-visible diffuse reflectance spectra (DRS), depicted in Figure 4d. In the visible range (400–800 nm), the pure TiO2 nanoparticles reflect the radiation almost entirely (≈ 95%). However, the nanocomposite displays reflectance below 64% for the same range. Additionally, a minimum reflectance (≈ 44%) was obtained at 545 nm, indicating a maximum of absorbance band that can be associated with the surface plasmon of Au nanoparticles, typically in the wavelength range between 520 and 560 nm [47,48]. These results show that the nanocomposite presented a broad absorbance spectrum when compared to the pristine TiO2 nanoparticles, which is also consistent with the purple/pink color exhibited by the produced nanocomposite. In the ultraviolet range (200–400 nm), both samples showed similar behavior.
From DRS spectra it was possible to estimate the band gap, shown in the inset graph of Figure 4d, for pure TiO2 and Au/TiO2 nanocomposite was converting the reflectance to Kubelka–Munk units through Equation (1) and Equation (2). The obtained values show that the nanocomposites possessed a lower bandgap (2.84 eV) than the pristine TiO2 nanoparticles (2.96 eV). The decrease of the bandgap in Au/TiO2 was related to the shift absorption to longer wavelengths. Similar results have been reported in the literature [49,50].
#### *2.2. Nanocomposites' Optimization and Photocatalytic Experiments*
The photocatalytic activity of all the produced Au/TiO2 nanocomposites was assessed by monitoring the degradation of CIP under artificial UV and visible irradiations. Process conditions were varied depending on the studying purposes.
#### Nanocomposite Optimization
As gold is a noble metal, cost-effectiveness should be considered, and the amount of gold used in the nanocomposite is one of the most paramount parameters. In this study, Au loading was varied by using different concentrations of the gold precursor. The tested Au loadings were 0.025, 0.05, 0.1, 0.25, and 0.5 wt. %. These nanocomposites were employed for the photocatalytic degradation of CIP under both UV and simulated visible radiation.
Figure 5a shows the data of photocatalytic experiments under UV light. Accordingly, all produced samples and the pristine TiO2 used as a control showed photocatalytic activity, proven by the decrease of CIP concentration along with the irradiation time. As confirmed by the diffuse reflectance spectroscopy (Section 2.1), the bandgap of the nanocomposites was 2.84 eV, corresponding to the wavelength of 437 nm. Here, the used UV lamp had the mode wavelength of 365 nm, which was shorter than the bandgap. It means that the photon energy was adequate to excite the photocatalytic materials, and photocatalytic reaction occurred in all experiments. Pristine TiO2 was compared with the synthesized photocatalysts. After 30 min, 77% of CIP was degraded in the presence of pristine TiO2, whereas higher degradation of 80–90% was achieved in the same time of irradiation, using the synthesized photocatalysts, 0.05 wt. %. This efficiency can be assigned to the presence of gold particles on the surface of the photocatalysts, confirmed by the TEM and EDX characterization (Section 2.1). The further quantitative inspection was obtained using the Langmuir–Hinshelwood kinetics (Equation (3)), and data are shown in Table 1. The apparent reaction rate constant k of the experiment with the bare TiO2 was found to be 0.047 min−1, while the decoration with gold particles improved the photocatalytic activity by 2–3 times. As predicted, in the presence of gold, the excited electrons may be conducted to the gold particles, and the electron-hole recombination may be reduced, which prolongs the lifetime of generated holes [51,52].
Consequently, the photocatalytic activity of the composites increased. Additionally, the increase of the used chloroauric acid concentration might induce a more significant number of gold particles distributed on the TiO2 surface. In other words, the number of electron absorption centers was increased, which explains the increase of k from 0.078 to 0.131 min−<sup>1</sup> when increasing the Au loading from 0.025 to 0.5 wt. %. However, the further increase in the Au loading caused a decrease in k. These results can be addressed to the loss of photocatalytic active sites on the surface of TiO2 nanoparticles. Based on the TEM images shown in Figure 1, when the Au loading was very high both the amount and the size of Au nanoparticles over the surface of TiO2 nanoparticles were larger, which contributed to a reduction of the adsorption and probably to mitigate the radiation absorbance by the catalytic nanoparticles. Together, these limitations contributed to reducing the photocatalytic efficiency of the nanocomposite towards the samples with lower amounts of Au and demonstrated the relevance of optimizing the Au loading.
**Figure 5.** Photocatalytic degradation of ciprofloxacin (5 mg/L) with bare TiO2 and Au/TiO2 nanocomposite with different Au concentrations under 30 and 180 minutes of UV (**a**) and visible (**b**) radiation. The degradation with bare TiO2 and Au/TiO2 nanocomposites synthesized at different temperatures and Au loading of 0.05 wt. % under 30 and 180 minutes of UV (**c**) and simulated visible light radiation (**d**), respectively.
**Table 1.** Apparent reaction rates (k) for photocatalytic degradation of ciprofloxacin (CIP) (5 mg/L) with bare TiO2 and Au/TiO2 nanocomposite with different Au loadings, over 30 and 180 minutes of UV and simulated visible radiation, respectively.
The photocatalytic assays performed under visible illumination are shown in Figure 6b. Regarding these assays, it is essential first to mention the controls (Supplementary Material, Figures S3 and S4), which have shown that the CIP solution was stable under simulated visible radiation, demonstrating its photostability. Moreover, another control was performed by adding the Au/TiO2 nanocomposites to CIP solution in the dark for 180 minutes. In this case, approximately 11% of CIP was removed from the solution by adsorption to the Au/TiO2 nanocomposites.
**Figure 6.** Degradation efficiency (%) (**a**) and ln (C/C0) vs. time (**b**) for different initial ciprofloxacin concentrations (5, 10, and 25 mg/L), using Au/TiO2 nanocomposites produced at 60 ◦C and with an Au loading of 0.05 wt. %, under 3 hours of simulated visible radiation. Photocatalytic degradation of ciprofloxacin (5 mg/L) in 45 mL of aqueous solution with different Au/TiO2 concentrations (0.1, 0.3, 1.0, and 1.3 g/L). The Au/TiO2 nanocomposite synthesized at 60 ◦C and with an Au loading of 0.05 wt. % was used. The tests were performed over 30 minutes under UV irradiation (**c**).
With the information from controls, it is possible to understand the photocatalytic efficiency of the tested materials better. Similarly, to the UV light experiments, the degradation rates of all produced nanocomposites were faster than that with the bare TiO2. TiO2 could remove ≈ 33% of CIP after 180 min of simulated visible irradiation. This CIP removal may be assigned to adsorption, confirmed by controls performed in the dark (as above mentioned). Additionally, the sun simulator device had a small percentage (≈ 3%) of UV radiation (to mimic sunlight radiation). This radiation can induce a low photocatalytic activity on bare TiO2, which, together with the adsorption of CIP, is responsible for its removal from the solution.
More importantly, the decoration of gold particles on the TiO2 surface resulted in the faster degradation rate of CIP under visible radiation. The bandgap of the composites was lowered, from 2.96 eV to 2.84 eV (Section 2.1). Similar results were obtained for methylene blue degradation using Au/TiO2 nanoparticles. The authors obtained higher degradation efficiencies and ability to use visible radiation [37]. Thus, the materials could absorb the longer wavelength in the visible range (up to 437 nm). The reaction rates' constant increased from 0.073 h<sup>−</sup>1, without Au, to 0.195−0.224 h<sup>−</sup>1, with different Au loadings (Table 1).
The obtained results, for UV and visible radiation, confirmed that the photocatalytic efficiency of the TiO2 nanoparticles was enhanced with the Au loading, until a specific plateau. When the Au loading was higher than 0.1 and 0.05 wt. %, respectively, for UV and visible radiation, the gold nanoparticles can block the surface-active sites of TiO2 nanoparticles [53,54]. Furthermore, an excessive amount of Au nanoparticles can play as recombination centers for photo-induced electrons and holes. Both situations can contribute to a significant reduction of pollutant adsorption and, consequently, the photocatalytic efficiency [55]. The remaining assays of this study will be performed with an Au loading of 0.05 wt. %.
Another critical parameter that is worth to stress and study is temperature, which can affect the surface charge phenomenon and the dispersity of the TiO2 particles in the solution during the synthesis. It can also influence the nucleation and the gold particles' crystal growth on the TiO2 nanoparticle surface. In this study, the synthesis was operated at 25, 60, and 80 ◦C, and the photocatalytic degradation of CIP, with the nanocomposites produced at different temperatures, was performed under UV and visible radiation (Figure 5c,d and Table 2). Regardless of the synthesis temperature, the photocatalytic activity of the nanocomposites (Au loading = 0.05 wt. %) was equal or higher than that of the bare TiO2. Here, the synthesis at the room and medium temperatures (25 and 60 ◦C) yielded the more efficient photocatalytic materials, for UV and visible radiation, towards higher temperature synthesis (80 ◦C).
**Table 2.** Apparent reaction rates (k) for photocatalytic degradation of CIP (5 mg/L) with bare TiO2 and Au/TiO2 nanocomposite synthesized at different temperatures, over 30 and 180 minutes of UV and simulated visible radiation, respectively. The Au loading of 0.05 wt. % was used for the tested materials.
For both types of radiation, the sample obtained at 60 ◦C presented higher degradation efficiencies (Table 2), 91% and 49% of CIP degradation under UV (30 min) and visible radiation (180 min), respectively. On the other hand, the synthesis performed at 80 ◦C revealed lower degradations rates of 80% and 40% for UV and visible radiation, respectively. Other works have reported that higher temperatures accelerate the reduction process and yield broader Au nanoparticles size distributions [56].
In this context, when the synthesis occurred at 80 ◦C, the size of Au nanoparticles produced was larger than the sizes obtained with 25 ◦C and 60 ◦C (in good agreement with STEM-HAADF micrographs, Figure 1). Similarly to what happened with the Au loading, when the amount of Au on the surface of TiO2 was too high, the active sites were blocked and the pollutant adsorption can be limited. Compared with bare TiO2, these results corresponded to a degradation efficiency increase of approximately 13% and 145% for UV and visible radiation, respectively.
Both under UV and visible radiation, another control was performed (Figure 5c,d) by testing single Au nanoparticles at the very same amount of Au (corresponding to 0.05 wt. % obtained at 60 ◦C) and TiO2 nanoparticles on CIP degradation. The results confirmed that the photocatalytic efficiency obtained by the nanocomposites should be assigned to the interface between Au and the TiO2 surface.
#### *2.3. Photocatalytic Degradation*
The rate of photocatalytic degradation depends on the availability of the catalyst surface for the photo-generation of electron-hole pairs that produce hydroxyl radicals. Thus, in these experiments, the amount of catalyst was kept constant, and the number of hydroxyl radicals generated remained the same, while CIP concentration increased. The influence of CIP initial concentration of 5, 10, and 25 mg/L was studied under visible irradiation. It was observed that the CIP concentration impacted by the degradation rate and efficiency (Figure 6). With the lowest CIP concentration, 40% of CIP degradation was obtained after 30 min. With the increase of concentrations by 2 and 5 times, the efficiencies achieved were 22% and 8%, respectively. In these tests, while using the photocatalyst concentration of 0.3 g/L, the adsorption of the CIP on the Au/TiO2 nanoparticles surface might be halted due to surface saturation. Additionally, the presence of organic compounds such as CIP can generate an increased number of intermediates and products, which will compete with CIP for adsorption on the photocatalyst surface [57]. This competition caused a lower reaction rate for high CIP concentration. The following assays, focused on the photocatalytic activity of the produced nanocomposites, were performed using the lowest CIP concentration, 5 mg/L.
In short, the ratio between hydroxyl radical/CIP molecules decreased with higher concentrations, causing lower photocatalytic activity. Moreover, higher CIP concentrations may also reduce radiation harvesting by TiO2 nanoparticles surface, which will also contribute to decreasing the number of hydroxyl radicals formed. Figure 6b displays the plot of ln (C/C0) vs. time at different initial CIP concentrations. Linear plots were observed, and the R<sup>2</sup> values were higher than 0.9, confirming that the photocatalytic degradation of CIP obeyed pseudo first-order kinetics.
The optimal photocatalyst concentration was assessed through degradation of CIP with the different amounts of Au/TiO2 nanocomposites, from 5 to 60 mg, which corresponded to a photocatalyst concentration of 0.1 and 1.3 g/L, respectively. Experimental results are shown in Figure 6c.
In general, with the photocatalytic concentrations of 0.1–1.0 g/L, ≈ 90% of the CIP in solution was degraded, while with the higher concentration of 1.3 g/L, only 50% CIP was degraded after exposure to the same UV irradiation time. At the lowest photocatalytic concentrations of 0.1–0.3 g/L, the photocatalytic degradation increased significantly with the amount of nanocomposite. Indeed, the increased amount of photocatalyst resulted in higher surface coverage, owing to the highest number of active sites [58]. However, when increasing the concentration to 1.0 g/L, the degradation rate remained unchanged. With the highest concentration of Au/TiO2, 1.3 g/L, the degradation was slowest likely because of excessive turbidity that induced light extinguishment after penetrating a short distance from the illuminated surface [59]. Most of the light might be extinguished after penetrating a short distance from the illuminated surface of the suspension. Thus, photocatalytic particles in an inner region could not be activated. The result agrees with other reports [58,60]. For this reason, the concentration of 0.3 g/L (15 mg in 45 mL), which yielded the highest photocatalytic efficiency, was chosen for the following photocatalytic activity assays.
The reproducibility of the nanocomposites efficiency was also tested using three independent syntheses, performed under the same conditions. The produced samples were then used in the photocatalytic degradation of CIP under visible radiation (results in the Supplementary Material, Figure S5). The apparent reaction rate constant of the three experiments fluctuated around the value of 0.219 <sup>±</sup> 0.022 min−1. The standard deviation of 10% proved the reproducibility of the method to enhance the photocatalytic activity of pure TiO2 nanoparticles and endow them with visible light activity. It is also important to clarify that the Au concentrations used in this study were nominal, as no inductively coupled plasma atomic emission spectroscopy (ICP-AES), or similar characterization, was performed. However, given the considerable reproducibility of the method, the putative loss of gold would by similar for all the Au concentrations tested, making them comparable.
#### **3. Computational Modeling: Gold on Titanium Dioxide and Charge Transfer**
A computational study was performed to rationalize the effect of gold nanoparticles on TiO2. The GFN-xTB (Geometry, Frequency, Noncovalent, eXtended Tight-Binding) was used. GFN-xTB is a new semiempirical method developed by Grimme et al. [61] that allows computing efficiently systems with thousands of atoms. The GFN-xTB software used (version 5.4.6) did not allow us to compute systems with periodic boundary conditions, so a finite system composed of a gold nanoparticle adsorbed on a larger TiO2 nanoparticle was used. We chose a cuboctahedral (TiO2)97 anatase nanoparticle, which was found to produce bulk-like electronic properties [62] and had two large, equal, flat surfaces of ~ 1.2 <sup>×</sup> 1.5 nm<sup>2</sup> on which a cuboctahedral (Au55) <sup>−</sup>3, 10 nm diameter, gold nanoparticle was adsorbed [63]. Several adsorption modes were possible, yet an exhaustive search of these was beyond the scope of this study. We positioned the gold nanoparticle on four, arbitrary, different orientations so that, in all cases, one of the faces of (Au55) <sup>−</sup><sup>3</sup> was parallel to the anatase flat surface and minimized. In this minimization, the TiO2 coordinates were held fixed while the Au first neighbors' distances were constrained harmonically to an equilibrium value to force the gold nanoparticle to keep its initial shape while retaining some flexibility. Two different pH conditions were considered: (1) A neutral/basic pH represented by the bare (TiO2)97 anatase nanoparticles (i.e., without protonation), and (2) an acidic pH, where the eight under-coordinated oxygen atoms in the anatase nanoparticle were protonated, (TiO2)97H8 <sup>+</sup>8, which should represent better the experimental conditions (Figure 7).
**Figure 7.** The (Au55(TiO2)97H8) <sup>+</sup><sup>5</sup> with standard atom coloring (left) with color-rendered atomic charges (right), so that red represents negatively charged atoms, blue positively charged atoms, and white neutral atoms.
To study the charge transfer between (Au55) <sup>−</sup><sup>3</sup> and (TiO2)97 and ((TiO2)97H8) <sup>+</sup>8, we computed atomic charges specifically suited for condensed phases [64]. The analysis of the charges in (Au55(TiO2)97) <sup>−</sup><sup>3</sup> and (Au55(TiO2)97H8) <sup>+</sup><sup>5</sup> (Figure 7) show that the Au nanoparticle transferred electrons to the anatase, namely 4.7 electrons for neutral TiO2 and 4.3 electrons for the protonated TiO2. Most of this charge was transferred directly to Ti atoms, namely 3 and 3.3 electrons for the neutral and protonated case, respectively. So, the Au acted as an electron donor populating the Ti(d) states that were responsible for the photocatalytic activity of the material. Consequently, Au may increase the catalytic activity of the composite material through interfacial electron transfer. Interestingly, the anatase surface also polarized the gold nanoparticle so that all atoms in direct contact with TiO2 were more oxidized, i.e., had larger positive charges, see Figures 7 and 8.
This effect was observed in all cases, for neutral and acidic pH and when the harmonic constraint on Au atoms was released, which indicates that the observed charge transfer is a fundamental process of the Au-TiO2 interface. This corroborated the experimental controls shown in Figure 5c,d, showing that separation of Au and TiO2 nanoparticles yielded lower efficiencies. Moreover, this finding also fit previous DFT (density functional theory) calculations which found that an Au nanorod on a rutile TiO2 (110) surface might act as an activator for molecular oxygen through charge transfer to nearby Ti+<sup>4</sup> atoms [65]. Mechanistically, the presence of Au activated superficial Ti+<sup>4</sup> atoms nearby for catalysis via direct charge transfer, which rationalized our experimental observation that lower Au loading and small gold nanoparticles had larger catalytic activity than larger loadings and nanoparticles on TiO2.
**Figure 8.** The (Au55(TiO2)97H8) <sup>+</sup><sup>5</sup> with only the titanium subnetwork and gold atoms with their corresponding point charges.
#### *Adsorption of Ciprofloxacin on Au55(TiO2)97H8* +*58*
We also explored the different energetics of a CIP molecule interacting with (Au55(TiO2)97H8) +58 on four different adsorption sites (Figure 9). All the adsorption modes were found to be binding, i.e., exothermic, with energies ranging between 0.7 and 2.4 eV. The strongest binding was observed for the structure with a large contact between the oxygen atom of the carboxylic group and the gold nanoparticle, 2.4 eV. This binding was similar in energy to the adsorption on the anatase clean surface (2.1 eV), which indicates that both processes might be competitive and that CIP molecules might also adsorb near the gold/anatase interface where the Ti atoms were activated through electron transfer.
**Figure 9.** Four possible adsorption geometries of CIP on (Au55(TiO2)97H8) <sup>+</sup><sup>58</sup> with corresponding binding energies. The positive binding energies indicate that the process is exothermic.
#### **4. Materials and Methods**
#### *4.1. Materials*
P25 TiO2 nanoparticles were kindly provided by Evonik (Essen, Alemanha). Gold(III) chloride trihydrate, 99.9% CAS: 16961-25-4 (liquid solution) was purchased from Sigma-Aldrich (St. Louis, Missouri, EUA). Sodium hydroxide (NaOH) was obtained from VWR (Radnor, Pensilvânia, EUA) Millipore Milli-Q-system ultra-pure (UP) water was used in all the experiments.
#### *4.2. Nanocomposite Synthesis*
The Au/TiO2 nanocomposites were synthesized, as illustrated in Figure 10, dispersing 200 mg of TiO2-P25 nanoparticles in 40 mL of ultra-pure (UP) water in a sonication bath for 30 min. Afterwards, this solution was placed under agitation in a water bath at different temperatures (25, 60, and 80 ◦C), using a thermostat to precisely control and stabilize the temperature, avoiding thermal gradients. When the dispersion solution reached the desired temperature, different volumes from the chloroauric solution (10 μL of Gold(III) chloride trihydrate in 100 mL of UP water) were added to achieve the Au loadings of 0.025, 0.05. 0.1, 0.25, and 0.5 wt. %. The solution was then stirred for 10 minutes to achieve a homogeneous distribution of gold precursor solution. Later, several volumes of a 0.1 M sodium hydroxide solution (NaOH) were added dropwise and mixed for 10 minutes to obtain a pH = 9. The solution was then centrifuged at 23,000 rpm, the supernatant discarded, and the nanocomposite pellet redispersed in UP water with the ultrasonication for 1 minute, and this washing procedure was repeated one more time. The last step was to dry the nanocomposite at 80 ◦C in an oven overnight and grind it with a pestle and mortar to obtain a fine powder.
**Figure 10.** Schematic representation of the main steps to synthesize Au/TiO2 nanocomposites trough nanoprecipitation.
#### *4.3. Characterization*
The morphology of the nanocomposites was assessed by transmission electron microscopy (TEM), a Tecnai T20 from FEI (Hillsboro, Oregon, EUA). For the analysis, the nanocomposite samples were sonicated for 5 minutes to achieve good dispersion and afterwards a drop of the suspension was placed on a copper grid and dried at room temperature for the analysis. Particle size histograms were obtained after measuring at least 200 nanoparticles using Image J 1.50i software. Aberration-corrected scanning transmission electron microscopy (Cs-corrected STEM) images were acquired using a high-angle annular dark field detector in an FEI XFEG TITAN (Hillsboro, Oregon, EUA) electron microscope operated at 300 kV equipped with a Spherical Aberration Corrector for Transmission Electron Microscopes (CETCOR) Cs-probe corrector from CEOS Company (Heidelberg, Germany), allowing the formation of an electron probe of 0.08 nm. Elemental analysis was carried out with an EDX (energy-dispersive X-ray spectroscopy) detector, which allows performing EDX experiments in the scanning mode.
The crystallographic phases of the pure TiO2 and the Au/TiO2 nanocomposite were evaluated by X-ray diffraction using a D8 Discover diffractometer with incident Cu Kα (40 kV and 30 mA), from Bruker (Billerica, Massachusetts, EUA).
The average hydrodynamic diameter was assessed by dynamic light scattering (DLS) in a Zetasizer NANO ZS-ZEN3600, Malvern (Malvern Instruments Limited, United Kingdom), equipped with a He–Ne laser (wavelength 633 nm) and backscatter detection (173◦). The samples were dispersed (0.1 mg/L) in ultrasonication bath at 22 ◦C for 30 minutes to avoid aggregates, and each sample was measured 10 times. The zeta (ζ) potential was measured in the same device, and TiO2 nanoparticles were equally suspended in ultra-pure water and solutions at different pHs (2, 4, 7, 9, and 12) were prepared with HCl (1M) and NaOH (1M) solutions. The results were obtained using the Smoluchowski theory approximation, and each sample was measured 10 times at 22 ◦C. The manufacturer software (Zetasizer 7.12) was used to assess particles diameter (intensity distribution), the polydispersity index (PDI), and z-potential values.
The optical properties of the pristine TiO2 and the Au/TiO2 nanocomposite were assessed by UV–vis reflectance, using a Shimadzu UV-2501-PC (Kyoto, Japan) equipped with an integrating sphere. The spectra were acquired in reflectance, and the bandgap was estimated via the Kubelka–Munk Equation (1) [52] and the Tauc plot represented by Equation (2).
$$F(\mathbb{R}) = (1 - R\_{\infty})^2 / (2R\_{\infty}) \tag{1}$$
where *R*<sup>∞</sup> (*R*Sample/*R*BaSO4) corresponds to the reflectance of the sample and *F*(*R*) is the absorbance.
$$\left[F(R)h\nu\right]^{1/n}\text{ versus }h\nu\tag{2}$$
where *h* is the Planck constant (6.626 <sup>×</sup> 10−<sup>19</sup> J), *n* is the frequency, and *n* is the sample transition (indirect transition, *n* = 2) [66].
#### *4.4. Photocatalytic Degradation*
The photocatalytic activity of all the produced samples and pristine TiO2 was assessed by performing (CIP) degradation tests, under artificial ultraviolet (UV) or visible illumination. First, a solution of 5 mg/L of CIP was prepared. The CIP solution was adjusted to pH = 3, to ensure the solubility, by using 0.1 mL hydrochloric acid (HCl) 1 M.
Before the degradation assays (UV or visible radiation), the Au/TiO2 or P25 nanoparticles were stirred in the dark for 30 min to achieve an adsorption-desorption equilibrium.
The UV degradation of CIP was performed in a chamber with six Philips 8 W mercury fluorescent lamps with the mode wavelength of 365 nm. The suspensions of photocatalysts and CIP were kept stirred in a container under the illumination from the top. The distance between the beaker and the lamp was 13.5 cm, and the intensity coming to the system was 15−17 W/m2. The samples were irradiated for 30 min.
The visible light tests were performed in a visible chamber fabricated by Ingenieurbüro Mencke & Tegtmeyer GmbH©, Hameln, Alemanha. According to the manufacturer, the visible light spectrum was equivalent to that of the natural solar light. The light source had an intensity of 98 W/m2. The visible light test was performed similarly to the UV test. Here, the container was placed at 21 cm from the light source, and the samples were irradiated continuously for 180 minutes.
The first photocatalytic activity tests were performed to determine the optimal ratio of CIP/catalyst. For this purpose, 5, 15, 45, and 60 mg of Au/TiO2 nanocomposite were dispersed in a borosilicate beaker of 80 mL with 45 mL of CIP solution (5 mg/L). The effect of Au loading on the photocatalytic efficiency was also assessed, under UV and simulated visible radiation. The impact of the synthesis temperature (25, 60, and 80 ◦C) on CIP photocatalytic degradation was equally evaluated using both types of illumination. The photocatalytic reproducibility tests were performed using nanocomposites produced in different batches but under the same synthesis conditions.
The bare TiO2 nanoparticles were used as controls in all the assays. Additionally, to prove the relevance of the Au/TiO2 nanocomposites' structure and interface, the photocatalytic degradation of CIP was assessed using the same amounts of Au and TiO2 nanoparticles, not as a nanocomposite, but separately added to the solution.
The photocatalytic efficiencies were tested by degrading CIP in aqueous solution under UV and visible radiation and monitoring the maximum absorption peak (277 nm) using a Shimadzu UV-2501PC UV/Vis spectrophotometer. The degradation fit the Langmuir-Hinshelwood model, expressed by Equation (3):
$$\text{C}/\text{C}\_{0} = \exp^{-kt} \tag{3}$$
where *C*<sup>0</sup> and *C* represent the concentration of the pollutant at time 0 min and at time *t*, respectively, and *k* is the first-order rate constant of the reaction.
#### **5. Conclusions**
An Au/TiO2 nanocomposite was produced, characterized, and applied in the photocatalytic degradation of ciprofloxacin (CIP). The characterization results changing the synthesis conditions (temperature and Au loading) indicated that the synthesis performed at 60 ◦C with the Au loading of 0.05 wt. % yielded the most homogeneous distribution of Au nanoparticles (≈3 nm) over TiO2 nanoparticles surface, after TEM inspection. Additionally, these samples absorbed more radiation in the visible range (≈66% at 545 nm) and presented a lower bandgap (2.84 eV vs. 2.96 eV from bare TiO2). The photocatalytic results confirmed that all the manufactured nanocomposites possessed higher photocatalytic efficiency in the UV and simulated visible radiation towards the pristine TiO2. It was also possible to understand the impact of the synthesis parameters envisaging the optimal photocatalytic efficiency conditions. In this way, with the Au/TiO2 nanocomposite, it was possible to enhance the photocatalytic degradation efficiency in 13% and 145% under UV and simulated visible light radiation, respectively. The gold nanoislands played a paramount role transferring electrons from Au to the anatase from TiO2 nanoparticles. Additionally, Au endowed the nanocomposite with the ability to absorb the visible radiation.
Computational modeling supported the experimental data, showing the ability of Au to bind TiO2 anatase surfaces as well as the relevant role of Au transferring electrons. The fundamental importance of the interface between TiO2 and Au nanoparticles regarding the enhanced photocatalytic activity was also rationalized. Moreover, models indicated a high affinity of CIP to both Au and TiO2 surfaces, which favors the adsorption process and consequently may also be cause for enhanced photocatalytic efficiency in the presence of Au nanoparticles.
According to the results obtained through systematic experimental data and modeling results, the simple method herein presented constitutes a reliable approach to produce efficient photocatalytic materials.
**Supplementary Materials:** The following are available online at http://www.mdpi.com/2073-4344/10/2/234/s1: Figure S1: Size distribution of Au nanoparticles for synthesization at 25 and 80 ◦C; Figure S2: Photostability of CIP solution under UV; Figure S3: Photostability of CIP solution under visible radiation; Figure S4: Synthesis reproducibility on CIP degradation.
**Author Contributions:** Conceptualization, P.M. and S.L.-M.; data curation, P.M., S.K., and H.N.L.; formal analysis, H.N.L., L.P., M.M.-F., and S.L.-M.; investigation, P.M., S.K., H.N.L., and M.M.-F.; methodology, P.M., S.K., and V.S.; project administration, S.L.-M.; resources, V.S., M.A., G.C., and S.L.-M.; software, M.M.-F.; supervision, K.K., M.A., G.C., and S.L.-M.; validation, V.S.; visualization, P.M. and M.M.-F.; writing—original draft, P.M. and H.N.L.; writing—review & editing, P.M., M.A., and S.L.-M. All authors have read and agreed to the published version of the manuscript.
**Funding:** The authors acknowledge funding from the Basque Government Industry Department under the ELKARTEK Program and the Spanish Ministry of Economy and Competitiveness (MINECO) through the project MAT2016-76039-C4-3-R (AEI/FEDER, UE) (including the FEDER financial support). This work was also supported by the Graduate Academy of the Technische Universität Dresden. Centro de Investigacion Biomédica en Red – Bioengenharía, Biomateriales e Nanomedicina (CIBER-BBN) is an initiative funded by the 6th National R&D&i Plan 2008–2011, Iniciativa Ingenio 2010, Consolider Program, and CIBER Actions and financed by the Instituto de Salud Carlos III (Spain) with assistance from the European Regional Development Fund. S. Kappert and H.N. Le acknowledge fruitful discussions with Nadia Licciardello.
**Acknowledgments:** This work was supported by the Portuguese Foundation for Science and Technology (FCT) in the framework of the strategic projects UID/FIS/04650/2013 by Fundo Europeu de Desenvolvimento Regional (FEDER) funds through the COMPETE 2020—Programa Operacional Competitividade e Internacionalização (POCI) with the reference project POCI-01-0145-FEDER-006941, project PTDC/CTM-ENE/5387/2014, as well as UID/BIO/04469 unit through COMPETE 2020 (POCI-01-0145-FEDER-006684) and BioTecNorte operation (NORTE-01-0145-FEDER-000004) funded by the European Regional Development Fund under the scope of Norte2020—Programa Operacional Regional do Norte. P.M. Martins thanks the FCT for the grant SFRH/BD/98616/2013 and Luciana Pereira for the grant SFRH/BPD/110235/2015. M. Melle-Franco would like to acknowledge support from Centro de Investigação em Materiais Cerâmicos e Compósitos (CICECO)—Aveiro Institute of Materials, POCI-01-0145-FEDER007679 (UID/CTM/50011/2013) and the FCT (IF/00894/2015).
**Conflicts of Interest:** The authors declare no conflict of interest.
#### **References**
© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
### *Article* **TiO2 and Active Coated Glass Photodegradation of Ibuprofen**
**Samer Khalaf 1,2,\*, Jawad H. Shoqeir 1, Filomena Lelario 2, Sabino A. Bufo 2, Rafik Karaman 2,3 and Laura Scrano <sup>4</sup>**
Received: 29 January 2020; Accepted: 25 February 2020; Published: 18 May 2020
**Abstract:** Commercial non-steroidal anti-inflammatory drugs (NSAIDs) are considered as toxic to the environment since they induce side effects when consumed by humans or aquatic life. Ibuprofen is a member of the NSAID family and is widely used as an anti-inflammatory and painkiller agent. Photolysis is a potentially important method of degradation for several emerging contaminants, and individual compounds can undergo photolysis to various degrees, depending on their chemical structure. The efficiency oftitanium dioxide (TiO2) and photocatalysis was investigated for the removal of ibuprofen from the aquatic environment, and the performance of these different processes was evaluated. In heterogeneous photocatalysis, two experiments were carried out using TiO2 as (i) dispersed powder, and (ii) TiO2 immobilized on the active surface of commercial coated glass. The kinetics of each photoreaction was determined, and the identification of the photoproducts was carried out by liquid chromatography coupled with Fourier-transform ion cyclotron resonance mass spectrometry (LC-FTICR MS). The overall results suggest that the TiO2 active thin layer immobilized on the glass substrate can avoid recovery problems related to the use of TiO2 powder in heterogeneous photocatalysis and may be a promising tool toward protecting the environment from emerging contaminants such as ibuprofen and its derivatives.
**Keywords:** ibuprofen; advanced oxidation process; TiO2; photocatalysis; active glass
#### **1. Introduction**
Emerging contaminants resulting from the presence and circulation of pharmaceuticals (PhCs) were the focus of many environmental chemists over the last few decades. In the aquatic environment, PhCs are introduced anthropogenically through pharmaceutical or conventional plants [1]. PhCsare found in tiny concentrations in surface waters, indicating insufficient treatment of such entities during the standard sewage treatment processes (STPs). The occurrence of these toxic drugsin wastewater effluent, along with their metabolites which may be much more harmful than their parent compounds [2,3], has the potential to be a great health problem since they are endocrine-disrupting agents, thus posing a significant barrier to the use of water recycling [4].
Ibuprofen (IBP), (*RS*)-2-(4-(2-methylpropyl) phenyl) propanoic acid, shown in Figure 1, is a non-steroidal anti-inflammatory drug(NSAID) belonging to the class of propanoic acid derivatives used as pain relief for several inflammation conditions, including rheumatoid arthritis, asan analgesic for pain relief in general, and as an antipyretic to help in fever conditions [5]. IBP enters the aquatic environment through effluents exiting secondary wastewater treatment plants, which are inefficient in removing a variety of small organic molecules, particularly pharmaceuticals [6]. Reported studies showed that the concentrations of IBP found in rivers and other environmental waters range between 10 ng·L−<sup>1</sup> and 169 <sup>μ</sup>g·L−<sup>1</sup> [7].
**Figure 1.** Chemical structure and ultraviolet (UV) absorbance of ibuprofen (IBP).
The fact that the current conventional wastewater treatment technologies such as those based on biological, thermal, and physical treatment processes are not efficient in removing or degrading small-molecular-weight pharmaceuticals with low biodegradability and high chemical stability such as ibuprofen [8] encouraged us to devote considerable effort toward developing a novel purification method that can efficiently remove this recalcitrant organic contaminant from the water environment.
Recently, we found that the integration of separation technologies, consisting of sequential elements of ultra-filtration (UF), activated carbon filtration (AC), and reverse osmosis (RO), as well as adsorption technology based on a surface modified clay minerals, was efficient in removing IBP and other pharmaceuticals to a safe level [9–13]. Nevertheless, the operating principles of these tools are only based on phase-transfer technologies, whereby the contaminant is retained on the filter or adsorbent without being degraded or destroyed to non-toxic compounds. Furthermore, some of the technologies used, such as UF and RO, are too expensive to be adopted in most real environmental situations. For the abovementioned reasons, we successfully attempted to find a good alternative method (degradation via photocatalysis) to these technologies for removing such pollutants from the aquatic environment.
The growing awareness of the risk arising from the occurrence of toxic organic contaminants in the aquatic environment promoted the development of technologies, such as photodegradation, and other advanced oxidation processes (AOPs), for efficient destruction of organic toxic compounds that exist in water and wastewater, including PhCs [14–16].
AOPs, i.e., processes based on highly reactive species such as hydroxyl radical (•OH), can oxidize and mineralize practically every organic entity [3], yielding CO2 and inorganic ions, thus resulting in total destruction of the target pollutant [3,8,16]. Advanced oxidation processes (AOPs) involve several homogeneous and heterogeneous processes such as photolysis, photocatalysis, ozonation, electrochemical oxidation, photo-Fenton, wet air oxidation, and sonolysis [8,15,16]. The most popular and effective type of AOP employed in water and wastewater treatment is heterogeneous photocatalysis with semiconductors [17–19].
Heterogeneous photocatalysis is a process via which a photoreaction is accelerated by the presence of a catalyst (usually semiconductor). In order for this to occur, the dispersed solid particles of the semiconductor in the treated solutionshould absorb significant portions of the UV light, and, under radiation, they may be photo-excited and produce oxidizing agents from water and oxygen [19]. Generally, TiO2 is considered the most efficient semiconductor to be employed in photocatalysis because of several factors including its low cost, low toxicity, chemical stability, large band gap, and high photosensitivity [8,16–19].
Under ultraviolet irradiation, TiO2 as a semiconductor causes the jump of an electron (e−) from the valence band (VB) to the conduction band (CB), resulting in the formation of a positive hole (h+) at the site of the electron. In the presence of aqueous suspended TiO2, the hole and electron can produce radicals of hydroxyl and superoxide that are very potent in the oxidation of many kinds of organic entities found in water sources, thereby leading to total degradation of these toxic organic agents. [20,21]. Equations (1)–(3) depict the formation reactions of the superoxide and hydroxyl radicals upon catalysis with TiO2
$$\rm{TiO}\_2 + \rm{hv} \rightarrow \rm{e}\_{\rm{CB-}} + \rm{h}\_{\rm{VB+}}.\tag{1}$$
$$\rm H\_2O + h\_{VB+} \rightarrow OH^\bullet + H\_{aq+} \tag{2}$$
$$\rm O\_{2(ads)} + e\_{CB\cdot} \rightarrow O\_2{\bullet}^\bullet \text{-(ads)} + H\_2O + OH^\bullet. \tag{3}$$
In some processes, a complete degradation of organic pollutants requires the presence of a radiation source, oxidizing agent, and a semiconductor. Oxidation and reduction processes are promoted by photo-generated charge carriers resulting from the excitation of TiO2 via photons with higher energy. Currently, this kind of photocatalysis is utilized to purify water [8,16–19].
Although photocatalytic degradation using suspensions of TiO2 particles was extensively employed to catalyze different contaminants, such as drugs, and although it achieved good results in recent years, this technique fails to be widely used because of the high cost and difficulty in isolating the semiconductor from the mixture after degradation [8,16–19]. To find a way around the need for catalyst recovery via filtration, a different approach, consisting of catalyst immobilization on a stationary support, should be assessed. For fulfilling this aim, TiO2 immobilized on different materials instead of the traditional powder was advocated and tested for obtaining a promising clean treatment method with a low cost [22–24].
In this study, the efficiency of two different systems, direct photolysis and heterogeneous photocatalysis (TiO2 powder and TiO2 immobilized on active glass), was investigated using simulated solar irradiation for the removal of IBP and its major photoproducts from the aqueous phase. The comparative performance of the adopted process was analyzed under the same experimental conditions, and the kinetics for each photodegradation reaction was evaluated. Moreover, major photoproducts were detected and identified via liquid chromatography coupled with Fourier-transform ion cyclotron resonance mass spectrometry (LC-FTICR MS).
It should be emphasized that the study on the photodegradation of IBP using an immobilized TiO2 system can be regarded as representative of a process for the degradation of a variety of pollutants which impose a risk to the environment.
#### **2. Results and Discussion**
#### *2.1. Characterization of Active TiO2-Coated Glass*
Figure 2A,B depict the SEM image of the active glass and the TiO2 coating comb geometry on the glass surface, respectively. The TiO2 film thickness was 397.2 nm, as shown in Figure 2A.
**Figure 2.** (**A**) SEM image of the blue glass cross-section illustrating the position of the TiO2 layer immobilized on the glass surface; (**B**) SEM image of the blue glasssurface illustrating the fine-tooth comb nature of the TiO2 coating.
Table 1 lists the elements in the glass surface and core class. As shown in the table, TiO2is present only in the glass surface along with other elements, whereas it is absent in the core glass.
**Table 1.** Scanning electron microscopy EDX analysis of the glass surface coated with TiO2 as compared to the glass core composition.
Sheel et al. (1998) reported the presence of cobalt oxide in a concentration of less than 75 μg/g [24], whereas we did not succeed in detecting any amount of cobalt oxide using our technique. It is believed that the blue color of the glass is due to the presence of cobalt oxide. For this reason, we report here the TiO2-coated active glass using the abbreviation "blue glass".
The glass coating is based on photocatalytic anatase TiO2, which is the most effective known photocatalyst [22,23]. For the preparation of the glass coating, TiO2 nanocrystallinefilm is deposed onto float glass using an atmospheric pressure chemical vapor deposition technique (APCVD) as described in Reference [24]. Anatase is considered a more efficient photocatalyst than rutile because of its slower rate of recombination [23,24]. In the coated surface of the active glass, the catalyzing efficiency is improved by the presence of Fe3O4, which favors the formation of multiple band gaps, enlarging the wavelength range that can be absorbed by the glass surface [22,23].
#### *2.2. Photodegradation Experiments*
#### 2.2.1. Preliminary Study
Prior to photodegradation of IBP, which involves photolysis and photocatalysis experiments, thermal (45 ◦C) or hydrolytic reactions in pure IBP solution and the adsorption of the pharmaceutical on the catalyst surface in TiO2 powder suspension were assessed. No significant loss of IBP occurred in dark conditions due to thermal reactions or hydrolysis. The adsorption equilibrium on TiO2 powder was reached within 60 min, and a slight decrease (not more than 4.3%) of free IBP concentration in the 0.2–0.3 g·L−<sup>1</sup> TiO2 suspensions was achieved. Adsorption increased to 6.9% and more (7.8%) when the amount of TiO2 powder was augmented to 0.4–0.5 g·L−1. The mentioned values are the average of three replicates.
#### 2.2.2. Photolysis Experiment
The uranyl oxalate method [25,26] was used to assess the light emission effectiveness of the irradiation system prior to the experimental work. The disappearance of oxalate was 7.2 <sup>×</sup> <sup>10</sup>−<sup>4</sup> mol·s–1.
Despite the high efficiency of the irradiation system (irradiance to exposed surface of the reactor, 500 W/m2), IBP concentration during this experiment was decreased only by 8.4% after more than 20 h. This result indicates that IBP is stable during photolysis because it has great chemical stability and a reduced molar adsorption coefficient above 280 nm [19] (Figure 1).
#### 2.2.3. Photocatalysis Experiments
TiO2 commercial powder and TiO2-coated active glass were employed separately. The degradation efficiency via the two different methods was compared.
#### Photocatalytic Degradation Using TiO2Powder
The concentration of IBP during the photocatalysis reaction was monitored using High Performance Liquid Chromatography-Ultra Violet HPLC-UV. The standard solution used showed a peak of IBP at 4.1 min retention time. After 5 min of sample irradiation, a marked decrease in the IBP concentration was observed (21.8% of the initial concentration) along with the appearance of a new photoproduct (Figure 3A,B).
After 30 min, a new peak due to the formation of another derivative was observed, and an approximately 60% reduction in IBP initial concentration was obtained (Figure 3C). Complete disappearance of IBP was achieved after 270 min (Figure 3D), while complete depletion including derivatives was observed after approximately 23hours (Figure 3E).
The combined results demonstrate that IBP was completely degraded by photocatalytic oxidation using TiO2 powder under simulated solar irradiation, and the efficiency of its removal was more than 87% within 80 min. The degradation of IBP occurred as a result of a photo-irradiation of the semiconductor, causing an electron transfer to the conduction band which subsequently formed a hole in the valence band, which led to photo-induced charge separation on the semiconductor surface and an exchange of electrons on the water semiconductor interface. This led to the formation of •O2 − via interactions of adsorbed oxygen molecules with the photo-generated conduction band electrons, whereas the •OH generated from the oxidation of the adsorbed water or hydroxyl anions by the valence band hole oxidized the adsorbed IBP molecules [17,19].
**Figure 3.** HPLC-UV separation of photodegraded solution using TiO2 powder, IBP (1), and IBP photoproducts (2 and 3) (**A**) at time zero (initial standard solution), (**B**) after 5 min, (**C**) after 30 min, (**D**) after 270 min with complete disappearance of IBP, and (**E**) after 23 h with complete disappearance of IBP photoproducts.
Photocatalytic Degradation Using TiO2-Coated Active Glass
During the first 2 h, a slight decrease in IBP concentration (2.7%) was achieved, while the appearance of a unique photoproduct was accomplished after 3 h (Figure 4A,B). Following a further 9hof irradiation, the concentration of IBP was decreased to 50%, and, after 24 h, IBP and its photoproduct disappeared (Figure 4C).
**Figure 4.** HPLC-UV separation of photo-degraded solution using TiO2-coated blue glass, IBP (1), and IBP photoproducts (2) (**A**) after 180 min, (**B**) after 10 h, and (**C**) after 24 h with complete disappearance of IBP and IBP photoproducts.
#### *2.3. Kinetics Studies*
#### 2.3.1. Experimental Observations
No degradation of IBP was observed in the dark in all aqueous environments adopted for the experiments. Direct photolysis under simulated sunlight did not achieve the desired goal. Accordingly, we can conclude that the direct interaction of IBP with sunlight (both via thermal hydrolytic reactions and photolysis) cannot lead to IBP's quick degradation. However, in the presence of TiO2, complete removal of this NSAID was obtained although a xenon lamp with low UV energy was used for irradiation aiming at the simulation of sunlight effect.
Different amounts (0.1–0.5g·L−1) of TiO2 micro-particles were added to a solution of 25 mg·L−<sup>1</sup> IBP to determine the efficiency of the catalytic process. The half-life (experimentally observed) of the mother molecule was reduced upon increasing the concentration of the catalyst from 0.1 to 0.2 g·L<sup>−</sup>1, and it remained constant upon adding an amount of 0.3 g·L<sup>−</sup>1, while it increased when concentrations of 0.4 or 0.5 g·L−<sup>1</sup> were tested. The rationale behind such behavior is that the number of IBP molecules and photons absorbed on the TiO2 particles increased with a moderate increase in the catalyst loading; however, with further addition of the semiconductor (powder), the phenomenon of light scattering took place and the number of useful photons per mass unit of TiO2 was reduced. The disappearance of IBP at the highest concentrations of TiO2 powder was mostly due to its mere physical adsorption onto the surface of semiconductor particles.
Figure 5A,B illustrates the depletion trend of IBP measured as Ct/C0 versus irradiation time (A) and evolution of photoproducts (B) using different photodegradation methods. In both photocatalysis
processes, IBP underwent complete disappearance via the formation of one or two intermediates that were subsequently removed within 24 h.
**Figure 5.** Evaluation of IBP degradation measured as Ct/C0 versus irradiation time (**A**); evolution of photoproducts using different photodegradation methods (**B**).
In the photocatalysis experiment with TiO2-coated active glass, the reaction was apparently slower than degradation obtained using TiO2 powder, but a satisfactory depletion of IBP and its derivatives was reached in approximately the same time.
#### 2.3.2. Kinetic Parameters
To find the kinetics model, kinetic parameters were calculated using integrated equations describing zero-, first-, and second-order (Langmuir-Hinshelwood) order equations [27]. According to Snedecor and Cochran (1989) [28], the least square method should be utilized to find the best fit.
Table 2 summarizes the kinetic parameters of IBP degradation under the photocatalysis experiment with TiO2 powder.
**Table 2.** Kinetic parameters of IBP degradation under photocatalysis experiments.
ΣLSq, sum of least squares = Σn (Cexp–Ccalc)2; Cexp, experimental values of concentrations; Ccalc, value of concentrations calculated from rate equations; n, number of experimental observations; k, kinetic constant; t1/2, half-life.
It must be taken into account that, owing to the dissimilar units associated with them, the values of kinetic constants calculated by equations describing reactions of different order cannot be compared. For this reason, it is useful to consider the values of half-life, which are always expressed in time units. Table 2 shows dissimilar values of the half-life when calculated using different equations applied to the same system. The least square method of estimation is a powerful method to assess the equation that can best fit the experimental data.
Apparently, the measured reaction rate of IBP under irradiation conditions using TiO2 powder as a catalyst was best fit by a Langmuir–Hinshelwood-type equation [29].
$$\mathbf{C}\mathbf{t} = \mathbf{C}\_0 \,\mathbf{t}\_{1/2}/(\mathbf{t} + \mathbf{t}\_{1/2}),\tag{4}$$
where C0 is the initial amount (mg) of IBP per liter of solution, Ct is the remaining concentration at time t, and t1/<sup>2</sup> is the half-life of the reactant.
Equation (4) shows the minimum value of the sum of least squares, based on the number of observations (ΣLSq)/n, and describes a second-order reaction governed by the kinetic law.
$$\mathbf{v} = -\mathbf{d}\mathbf{C}\_t \text{/dt} = \mathbf{k}\mathbf{C}\_t \text{\textdegree } \tag{5}$$
where ν is the reaction rate, and k is the rate (or kinetic) constant [27,29], which in our case can be calculated as
$$\mathbf{k} = 1/(\mathbb{C}\_0 \mathbf{t}\_{1/2}).\tag{6}$$
Equation (5) represents a double dependence of the reaction velocity on the concentration.
The rationale behind such a finding may be due to the total reaction rate during the photocatalytic process being affected by two sorption states, both depending on the dissolved concentration of the pharmaceutical. The amount of reactant disappearing at each time t is affected by its free concentration in the powder suspension, as well as by the amount adsorbed on the catalyst particles, which depends on the remaining free concentration of IBP.
The half-life value for a second-order reaction, calculated by means of the linearized form of Equation (4) (Table 2), was just 11.8 min, while, after 80 min, 87% of IBP was converted.
The second-order kinetics shown in Figure 6A was confirmed by the linear behavior of (C0/Ct) as a function of irradiation time (Figure 6B).
**Figure 6.** (**A**) Photodegradation of ibuprofen catalyzed by TiO2 powder; Ct calc, values calculated using Equation (4); Ct exp, experimental values; error bars represent the standard deviations of three replicate experiments. (**B**) Trend of second-order linearized equation used for the calculation of kinetic parameters reported in Table 2.
From the results, it can be remarked that the initial degradation rate was high; however, it decreased rapidly as the reaction proceeded. The degradation was fast during the first 20 min, and then it gradually decreased; this trend is typical of second-order reactions. Several observations can be related to such a behavior: (1) the high concentration of IBP at the beginning of the reaction facilitates the useful attack by the hydroxyl radicals, resulting in high degradation rate; however, when IBP concentration gradually decreases, the degradation rate subsequently decelerates due to the dilution effect that reduces the possibility of useful collisions with the hydroxyl radicals; (2) the competitive reactions of the hydroxyl radicals with IBP degradation products that are produced during the reaction; (3) the recombination reactions of radical–radical.
The photodegradation reaction of IBP catalyzed by TiO2 immobilized on active glass surface achieved 85% compound disappearance after 24 h of simulated solar light irradiation. By attempting to fit the concentration values vs. time using various-order integrated kinetic equations, it was found that the data best fit the first-order kinetic equation
$$\mathbf{C}\_{\mathrm{I}} = \mathbf{C}\_{\mathrm{O}} \stackrel{-\mathrm{k}t}{\mathrm{e}}.\tag{7}$$
In Figure 7A, the best fit of experimental data calculated using Equation (7) is represented clearly as confirmed by the high value of the determination coefficient (*R*2) obtained for the linearized form of Equation (7) (Table 2, Figure 7B).
**Figure 7.** (**A**) Photodegradation of ibuprofen catalyzed by TiO2-coated blue glass; Ctcalc, values calculated using Equation (6); Ct exp, experimental values; error bars represent the standard deviations of three replicate experiments. (**B**) Trend of first-order linearized equation used for the calculation of kinetic parameters reported in Table 2.
In this case, the half-life can be calculated as
$$\mathbf{t}\_{1/2} = \text{L.r2/k.}\tag{8}$$
The value obtained was 575 min (Table 2), which is far from the half-life resulting from the experiment with TiO2 powder; nevertheless, it is satisfactory if we consider the high stability shown by IBP molecules not only in the darkness (for testing thermal and hydrolysis degradation), but also under light irradiation (photolysis degradation). Moreover, from Equation (8), it is possible to notice that, unlike the case of the second-order reaction, the half-life for the first-order reaction does not depend on the initial concentration of the reactant. This means that the reaction catalyzed by TiO2 immobilized on the active glass surface does not suffer from the same limitations encountered in the case of the second-order kinetics shown by IBP under the photoreaction catalyzed by TiO2 powder dispersion (light scattering, radical–radical recombination reactions, and dilution effect). Furthermore, it should be emphasized that (i) the number and persistence of derivatives was reduced in the case of coated active glass, and (ii) the time needed for an efficient degradation of the mother drug and its derivatives was approximately the same.
We have to mention that the degradation of IBP is also influenced by the pH value of the medium [30]. The production of hydroxyl radicals is generally increased in an alkaline medium, since high concentrations of OH− result in the formation of hydroxyl radicals, which are produced from the reaction of OH<sup>−</sup> with TiO2 on its surface's holes [31]. The pH value can also affect the charge on the catalyst particles; consequently, the electrostatic interactions between the charged surface of TiO2 and the pollutant molecules can be largely influenced, thus leading to a change in the adsorption level of these molecules on the catalyst surface and interfacial electron transfer [32].The most dominant factor affecting the adsorption of pollutant on catalyst surface is the catalyst zero-point charge (zpc), which is defined as the pH at which the surface of the catalyst has neutral charge [33].
For TiO2 P25 Evonik-Degussa, the zero-point charge value is 6.9. Therefore, the surface of TiO2 is positively charged in acidic media and negatively charged in basic media [33]. Accordingly, the effect of pH depends mainly on the type of the pollutant and the zero = point charge of the semiconductor. As IBP is weakly acidic in nature, it is expected to be negatively charged at pH higher than 3 [30], while the TiO2surface is positively charged at pH less than 6.9 [34]. Therefore, at pH = 4.5 where the photocatalytic experiment took place, the adsorption of IBP and, consequently, its photocatalytic oxidation were favored [30].
#### *2.4. Identification of Intermediate Photoproducts*
For the identification of IBP degradation byproducts, samples were collected at various time intervals and analyzed and identified by LC-FTICR MS system in the *m*/*z* range of 50–1000 in negative ionization mode. The results indicate the formation of two major photoproducts (Table 3). In addition, they reveal that the hydroxyl radicals attacked both the propionic acid and isobutyl substituent in IBP, resulting in the formation of two products, 2-[4-(1-hydroxyisobutyl) phenyl] propionic acid (2) and 4-(1-hydroxy isobutyl) acetophenone (3). Figure 8 depicts the proposed reaction pathways. The peak that appears at a nominal *m*/*z* value of 221, showing the formation of a mono-hydroxylated product of IBP, corresponds to 1-hydroxy IBP (2). Furthermore, product 2 was converted into another derivative with a nominal *m*/*z* value of 191, which corresponds to 4-(1-hydroxy isobutyl) acetophenone (3).
**Table 3.** Identification of ibuprofen and its photoproducts during photocatalytic degradation as deprotonated molecules, [M–H]−, by high-resolution LC-ESI-FTICR MS.
<sup>a</sup> Number used to identify each compound in the chromatograms of Figures 4 and 5. <sup>b</sup> Chromatographic retention time of compounds eluted under the experimental conditions described in Section 3. <sup>c</sup> Molecular formula of deprotonated compound. <sup>d</sup> Accurate *<sup>m</sup>*/*<sup>z</sup>* value of deprotonated molecules. <sup>e</sup> Mass error in parts per million <sup>=</sup> <sup>10</sup><sup>6</sup> × (accurate mass − exact mass)/exact mass.
**Figure 8.** By-products generated by TiO2 photocatalytic processes identified by LC-FTICR MS system in negative ion mode and proposed photodegradation pathway.
It is worth noting that, in the photodegradation method using TiO2 immobilized on the active glass surface, only one by-product, compound 2, was detected.
#### **3. Materials and Methods**
#### *3.1. Chemicals and Analytical Methods*
Ibuprofen (MW, 206.3 g·mol<sup>−</sup>1; pKa, 5.0) pure standard (purity, 99%) was purchased from Sigma Aldrich (Munich, Germany); acetonitrile, formic acid, and water for analysis were HPLC grade and purchased from Sigma Aldrich; TiO2 Degussa P-25 was a kind compliment from Evonik Industries (Steinheim, Germany); TiO2-coated active glass (Figure 9) was obtained as a gift from Pilkington (UK) (Sheel et al. 1998). PTFE (polytetrafluoroethylene) filters, 0.2 μm pore size, filter-Ø: 25mm, were purchased from Macherey-Nagel GmbH & Co. KG (Duren, Germany). Daily fresh working solutions were prepared using ultra-pure water from a bi-distilled purification system.
**Figure 9.** TiO2-coated blue glass.
To avoid microbial contamination, all glass apparatus was heat-sterilized by autoclaving for 60 min at 121 ◦C before use. Aseptic handling materials and laboratory facilities were used throughout the study to maintain sterility.
IBP concentrations were monitored using high-performance liquid chromatography (HPLC) (1200 series, Agilent Technologies, Santa Clara, USA) equipped with an Eclipse XDB-C18 (3 μm particle size, 4.6 × 150 mm) column (Phenomenex, Torrance, USA) using a diode array detector (DAD) at a wavelength of 230 nm. The mobile phase consisted of 40% of 1% formic acid solution/60% acetonitrile. The flow rate was 1.0 mL·min−1. Several aqueous solutions (from 0.5 to 25.0 mg·L−1) of IBP were filtered, and 20 μL of the filtrate was injected and analyzed. Peak areas vs. concentration of IBP were plotted, and the calibration curve was obtained with a determination coefficient (*R*2) of 0.9986. The limit of detection (LOD) of IBP for this method (using DAD) was 0.2 mg·L<sup>−</sup>1, and the limit of quantitation (LOQ) was 0.6 mg·L<sup>−</sup>1. The identification of IBP photoproducts was performed using the LC-FTICR MS system (Thermo Fisher Scientific, Bremen, Germany), in the same separation conditions. Negative ion ESI-MS mode was used for the detection of the compounds of interest. Full-scan experiments were performed in the ICR trapping cell in the range *m*/*z* 50–1000. Mass-to-charge ratio signals (*m*/*z*) were acquired as profile data at a resolution of 100,000 full width at half maximum (FWHM) at *m*/*z* 400. The limit of detection for mass spectrometric method was a few pmols.
The photodegradation experiments were performed using a solar simulator device Heraeus Sun-test CPS+ (Atlas, Chicago, USA), equipped with a 1500-W xenon arc lamp protected with a quartz filter (total passing wavelength: 280 nm < λ < 800 nm). The irradiation chamber was maintained at 20 ◦C by circulating water from a thermostatic bath and through a conditioned airflow.
#### *3.2. Characterization of the TiO2-Coated Active Glass*
The active glass was obtained via a coating process using a nanocrystalline film of TiO2 on a 4-mm-thin glass sheet. Some cross-sections obtained from theTiO2-coated active glass were analyzed. The scanning electron analysis for the TiO2-coated active glass was accomplished using a scanning electron microscope (SEM) of LEO model EVO50XVP, Carl Zeiss AG-EVO® 50 Series (Germany). The thin sections were grafted with 30-nm-thick carbon films. Semi-quantitative analyses of the
elemental composition of the different layers were obtained using a Ge ED Oxford-Link detector equipped with a super atmosphere thin window. Operating conditions of the SEM were as follows: 15 kV accelerating potential, 500 pA probe current, and about 10 mm working distance (WD). Thin sections of glass were prepared by the Department of Health and Environmental Science, Bari University. Samples were embedded in resin epoxy plugs and then polished.
#### *3.3. Photolysis Experiment*
Aqueous IBP solution of initial concentration 25 mg·L−<sup>1</sup> was prepared by dissolving a determined quantity of standard IBP in ultrapure water. The measured pH of the solution was 4.5. The photolysis treatment was carried out in a glass Pyrex®batch reactor closed at the top with a quartz cover. IBP solution (250 mL) was placed into the reactor; then, thereactor was placed into the irradiation oven inside the solar simulator, which reproduced the spectral distribution of natural solar irradiation. The IBP solution was continuously remixed during the experiment by magnetic stirring, and samples were taken (1 mL for each sample) at determined intervals and then analyzed using the HPLC system according to the analysis method in Section 3.1. Three experiments of direct photolysis were performed in triplicate.
#### *3.4. Photocatalysis Experiment with TiO2Powder*
A solution of ibuprofen was prepared as described in Section 3.3, but 50 mg of TiO2 powder (0.2 g·L−1) was added as the optimized amount in the reactor vessel. The aqueous suspension was mixed continuously in the dark for 2 h to ensure that the adsorption equilibrium of IBP on the catalyst surface was reached; then, the reactor was transferred into the solar simulator and exposed to solar irradiation, and samples were taken (1 mL for each sample) at determined intervals, then filtered and analyzed by HPLC. Three experiments were performed in triplicate.
#### *3.5. Photocatalysis Experiment with TiO2Immobilized on Active Glass*
Seven active glass sheets were placed vertically to cover the full surface of the inner wall of the reactor; then, the solution of IBP, prepared as in the previous two experiments, was added, and the reactor was transferred into the solar simulator and exposed to xenon lamp irradiation with continuous mixing; samples were taken (1 mL for each sample) at determined intervals, and then filtered and analyzed. Three experiments were performed in triplicate.
#### **4. Conclusions**
IBP is very stable under direct photolysis conditions due to its high chemical stability and low molar adsorption coefficient in the range of wavelengths provided by solar irradiation. On the other hand, our experiments showed that effective destruction of IBP and its photoproducts is possible by photocatalysis in the presence of TiO2 powder suspension or using TiO2immobilizedon the surface of active glass. Two intermediate photo-products were detected and identified by LC-FTICR MS. As there is substantial equivalence in the long-term efficacy of photocatalys is in the presence of TiO2 both as powder suspension and as glass coating, the use of active glass instead of TiO2 suspension could be a promising technique for the removal of pharmaceutical residues such as IBP and its photoproducts from aquatic environments not requiring the recovery of the catalyst after photodegradation. To increase the effectiveness of the technique described herein, a modification of TiO2immobilization on the glass surface using supports with a more complex geometry is essential.
**Author Contributions:** S.K. This work is a part of his PhD thesis (Chapter 4). Therefore all experiments and results and manuscript writing were referring to his own efforts; J.H.S. Supervision and reviewing results, reviewing the manuscript English writing; F.L. LCMS analysis; L.S. SEM analysis; S.A.B. Supervision and reviewing results, reviewing the manuscript English writing; R.K. Reviewing the manuscript English writing. All authors have read and agreed to the published version of the manuscript.
**Funding:** This work was supported by the European Union in the framework of the Project "Diffusion of nanotechnology-based devices for water treatment and recycling; NANOWAT" (ENPI CBC MED I-B/2.1/049, Grant No. 7/1997).
**Acknowledgments:** Many thanks to Jawad H. Shoqueir, the head of Soil and hydrology Lab at Al-Quds University, for his support to partially cover the publication fee from his own budget. Results reported in this article were partially presented by Samer Khalaf at the Second International Conference on Recycle and Reuse, 4–6 June 2014, Istanbul, Turkey and published in the book of abstracts.
**Conflicts of Interest:** The authors declare no conflict of interest.
#### **References**
© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
### *Article* **Photocatalytic Degradation of Chlorpyrifos with Mn-WO3**/**SnS2 Heterostructure**
#### **Charlie M. Kgoetlana, Soraya P. Malinga and Langelihle N. Dlamini \***
Department of Chemical Sciences, University of Johannesburg, Doornfontein Campus, P.O. Box 17011, Doornfontein, Johannesburg 2028, South Africa; [email protected] (C.M.K.); [email protected] (S.P.M.)
**\*** Correspondence: [email protected]; Tel.: +27-011-559-6945
Received: 21 May 2020; Accepted: 4 June 2020; Published: 21 June 2020
**Abstract:** Tungsten trioxide (WO3) is a photocatalyst that has gained interest amongst researchers because of its non-toxicity, narrow band gap and superior charge transport. Due to its fast charge recombination, modification is vital to counteract this limitation. In this paper, we report on the fabrication of Mn-doped WO3/SnS2 nanoparticles, which were synthesised with the aim of minimising the recombination rates of the photogenerated species. The nanomaterials were characterised using spectroscopic techniques (UV-Vis-diffuse reflectance spectroscopy (DRS), Raman, XRD, photoluminescence (PL) and electrochemical impedance spectroscopy (EIS)) together with microscopic techniques (FESEM-EDS and high resolution transmission electron microscopy selected area electron diffraction (HRTEM-SAED)) to confirm the successful formation of Mn-WO3/SnS2 nanoparticles. The Mn-doped WO3/SnS2 composite was a mixture of monoclinic and hexagonal phases, confirmed by XRD and Raman analysis. The Mn-WO3/SnS2 heterojunction showed enhanced optical properties compared to those of the un-doped WO3/SnS2 nanoparticles, which confirms the successful charge separation. The Brunauer–Emmett–Teller (BET) analysis indicated that the nanoparticles were mesoporous as they exhibited a Type IV isotherm. These nanomaterials appeared as a mixture of rectangular rods and sheet-like shapes with an increased surface area (77.14 m2/g) and pore volume (0.0641 cm3/g). The electrochemical measurements indicated a high current density (0.030 mA/cm2) and low charge transfer resistance (157.16 Ω) of the Mn-WO3/SnS2 heterojunction, which infers a high charge separation, also complemented by photoluminescence with low emission peak intensity. The Mott–Schottky (M-S) plot indicated a positive slope characteristic of an *n*–*n* heterojunction semiconductor, indicating that electrons are the major charge carriers. Thus, the efficiency of Mn-WO3/SnS2 heterojunction photocatalyst was monitored for the degradation of chlorpyrifos. The effects of pH (3–9), catalyst loading (0.1–2 g) and initial chlorpyrifos concentration (100 ppb–20 ppm) were studied. It was observed that the degradation was purely due to photocatalysis, as no loss of chlorpyrifos was observed within 30 min in the dark. Chlorpyrifos removal using Mn-WO3/SnS2 was performed at the optimum conditions of pH = 7, catalyst loading = 1 g and chlorpyrifos concentration = 1000 ppb in 90 min. The complete degradation of chlorpyrifos and its major degradation by-product 3,5,6-trichloropyridin-2-ol (TCP) was achieved. Kinetic studies deduced a second order reaction at 209 <sup>×</sup> <sup>10</sup>−<sup>3</sup> <sup>M</sup><sup>−</sup>1s−1.
**Keywords:** heterojunction; charge separation; photocatalysis; chlorpyrifos
#### **1. Introduction**
The fabrication and modification of photocatalysts has sparked interest amongst researchers due to their wide applications. Photocatalysts are used in applications ranging from water splitting, the degradation of pollutants in water, gas sensing and optoelectronic devices [1]. These can be *n*-type (electrons are the major charge carriers) or *p*-type (holes are the major charge carriers) photocatalysts [2]. The most widely studied photocatalysts are TiO2, WO3 (*n*-type) and ZnO, CdS (*p*-type) [3,4]. The photocatalytic efficiency of these materials is limited to a certain extent, primarily due to two major limitations. Firstly, they are prone to fast electron–hole recombination, which reduces the photocatalytic reactivity of the semiconductor. Secondly, they have wide band gaps that absorb only in the ultraviolet (UV) region, which accounts for 4% of the solar spectrum [5].
Modifications of photocatalysts to suit specific applications have been proposed. These include the use of metal dopants to form Schottky barriers and fusion with other semiconductor photocatalysts, resulting in heterojunctions [2]. The metal dopants that have been employed include magnesium (Mg), manganese (Mn), copper (Cu) and yttrium (Y) [6–9]. The metal doping of photocatalysts results in shifting the absorption band edge of the material to absorb the readily available visible region of the solar spectrum. They also separate photogenerated charges by forming electron traps, although a high concentration of the metal dopant may result in the creation of recombination centres, which leads to an increased recombination rate.
Heterojunctions that have shown enhanced optical and photocatalytic properties include BiVO4/WO3, CdS/ZnO and TiO2/SnO2 [10–13]. The formation of Type II heterojunctions using two different photocatalysts is sufficient to reduce the recombination rate of photogenerated charges. This occurs by the accumulation of photoexcited electrons in the conduction band (CB) of one semiconductor while photogenerated holes accumulate in the valence band (VB) of another semiconductor in the heterojunction system, which effectively leads to charge separation. Therefore, photo-oxidation and photo-reduction occur in different semiconductor surfaces of the heterojunction system due to the different migration points of the charges [14].
Tungsten trioxide (WO3) is a visible light photocatalyst with a band gap energy of 2.5–2.8 eV [15]. It is classified as an *n*-type semiconductor, wherein electrons are the major carriers. Due to its narrow band gap, WO3 absorbs light radiation in the visible range, and it has been used in a wide range of applications such as fuel production and combating water pollution [16,17]. This semiconductor exists in different polymorphs, which include monoclinic, triclinic, tetragonal and orthorhombic. The monoclinic phase of WO3 is the most stable and most photocatalytic compared to all the other phases.
Like most photocatalysts, WO3 suffers from limitations such as high electron–hole charge recombination. To overcome the intrinsic limitation of pristine WO3, different methods have been used, including metal doping and loading another semiconductor photocatalyst to form a heterojunction [10,18,19].
We, however, report the synthesis and characterisation of a material that fuses metal doping and a heterojunction (Mn-WO3/SnS2) that exhibits improved optical properties and minimises electron–hole recombination compared to current photocatalysts. Owing to the charge mobility in SnS2 to efficiently facilitate electron transfer to the WO3 CB, there results a high number of electrons for the oxidation reaction and separated holes that accumulate on the VB of SnS2. The Mn2<sup>+</sup> ions also separate charges by trapping electrons in the WO3, thereby increasing their lifetime, and act as reaction centres. This photocatalyst can be applied in water remediation, energy production and sensing. Thus, this study assessed the photo-efficiency of the heterostructure in the photodegradation of chlorpyrifos, an organophosphate pesticide.
Organophosphate pesticides have been used extensively in South Africa and the world at large due to their ability to combat a vast spectrum of pests [20,21]. Chlorpyrifos (O,O-diethyl O-[3,5,6, -trichloro-2-pyridyl] phosphorothionate) (CPF) is an organophosphorus pesticide extensively used in agricultural and domestic applications [22]. Chlorpyrifos agricultural application occurs throughout the year for a variety of fruits and vegetables [23]. It, however, does not readily dissolve in water, yet adsorbs strongly to soil particles.
Chlorpyrifos is an enzyme acetylcholinesterase inhibitor and persistent pesticide pollutant. It is a class II (moderately hazardous pesticide) pollutant, with a half-life of 60 days [24]. The pesticide is toxic to humans and other animals when ingested or inhaled; this is attributed to its lipophilic nature. It causes delayed peripheral neuropathy in humans and badly affect neuro-development in children at high doses [25,26].
Due to the numerous human and environmental effects caused by chlorpyrifos, different ways to remove this pesticide from the environment have been studied. These include advanced oxidation processes and biological treatment (with fungal and bacterial strains).
Bacterial strains have displayed high chlorpyrifos removal from water of up to 98% [27]. Though it is efficient, the method is strenuous, as bacteria require controlled specific conditions such as pH and temperature and a host for optimal function. On the other hand, Ismail et al. [24] in 2013 reported the use of advanced oxidation processes (AOPs) that yield 100% removal of chlorpyrifos by using 60Co γ-rays of 30–575 Gy [24].
However, γ-rays are harmful to human health; therefore, this led to the implementation of a better and safer method requiring the use of a photocatalyst to degrade chlorpyrifos under light irradiation. To date, zinc and titanium oxides have been used to degrade chlorpyrifos. The results were satisfactory, with up to 95% chlorpyrifos removal for TiO2 and 85% for ZnO under UV light [28]. However, the photocatalysts suffer from charge recombination and the use of UV light is not viable due to the insufficient amount of UV available (4%). Therefore, visible light-absorbing photocatalysts were discovered such as BiVO2, SnS2 and WO3.
To the best of our knowledge, no work has been reported to date on the fabrication of a metal-doped heterojunction (Mn-WO3/SnS2) photocatalyst.
#### **2. Results and Discussion**
#### *2.1. X-ray Di*ff*raction and Raman Analyses*
The phase and crystallographic properties of the nanomaterials were elucidated using XRD and Raman spectroscopy. Figure 1a shows the XRD pattern of WO3, which confirms the monoclinic nature of the WO3 (*m*-WO3). The *m*-WO3 was indexed and matched to the miller indices (002), (020), (200), (120), (112), (022), (202), (122), (222), (004), (040), (400), (042) and (420) (JCPDS Card No. 00-043-1035). Doping the *m*-WO3 with Mn2<sup>+</sup> did not distort the phase of the WO3, which implies that it had been intrinsically inserted into the WO3 crystal lattice as depicted in Figure 1b. The XRD pattern of the WO3/SnS2 heterojunction showed the presence of both monoclinic (*m*-WO3) and hexagonal (*h*-WO3) phases, and the hexagonal phase of SnS2 could be indexed (JCPDS Card No. 00-023-0677), as shown in Figure 1c. Again, the structural integrity of the manganese-doped WO3/SnS2 heterojunction (Figure 1d) was not distorted by the incorporation of Mn2<sup>+</sup> in the system. The average crystallite sizes of the nanomaterials were determined using the Debye–Scherrer equation and are tabulated in Table S1, all with an average size of 40 nm, with SnS2 having a crystallite size of less than 20 nm.
The nature of the phases was further confirmed with Raman analysis. Figure 2a illustrates Raman bands at 717 and 818 cm−<sup>1</sup> and less intense bands at 212 and 313 cm−<sup>1</sup> corresponding to O-W-O stretching and bending in the molecule, respectively, which confirms a monoclinic WO3; this finding was also reported by Simelane et al. 2017 and Xie et al. 2012 [3,15]. As in XRD analyses, the doping of Mn2<sup>+</sup> had no effect on the phase of WO3, as depicted in Figure 2b. No secondary bands resulting from Mn-oxides were observed. The heterojunction (WO3/SnS2) was successfully formed and confirmed by the Raman band at 317 cm−<sup>1</sup> corresponding to the A1g mode of hexagonal phase SnS2 as observed by Ma et al. 2015 (Figure 2c) [29]. Figure 2d displays the Raman band of Mn-WO3/SnS2 with no distortion due to Mn2<sup>+</sup> and SnS2. Therefore, the Raman band in Figure 2e corresponds to the pristine hexagonal phase of SnS2 due to the A1g band at 317 cm<sup>−</sup>1.
**Figure 1.** XRD patterns of (**a**) WO3, (**b**) Mn-WO3, (**c**) WO3/SnS2, (**d**) Mn-WO3/SnS2 and (**e**) SnS2.
**Figure 2.** Raman spectra of (**a**) WO3, (**b**) Mn-WO3, (**c**) WO3/SnS2, (**d**) Mn-WO3/SnS2 and (**e**) SnS2.
#### *2.2. Morphological Studies*
The morphological studies were conducted using microscopic techniques such as FESEM and HRTEM. Figure 3a is the FESEM image of pristine *m*-WO3 with rectangular sheets, rods and cubes, and the composition was confirmed by EDX (inset). The shapes of the nanomaterials did not change upon the insertion of Mn2<sup>+</sup> or the formation of the heterojunction as illustrated in Figure 3b,c and Figures S2 and S3. The EDX spectra displayed the elemental composition of the respective heterojunctions (WO3/SnS2) and Mn-WO3/SnS2 (inset).
**Figure 3.** FESEM images (inset is the corresponding energy-dispersive X-ray (EDX) spectrum) of (**a**) pristine WO3, (**b**) WO3/SnS2 and (**c**) Mn-WO3/SnS2, and TEM images (inset is the corresponding selected area electron diffraction (SAED) image) of (**d**) WO3, (**e**) WO3/SnS2 and (**f**) Mn-WO3/SnS2.
The HRTEM image of *m*-WO3 also showed rectangular sheets and rods (Figure 3d) and was further elucidated using SAED (inset) obtained through a 1–10 zone axis. The spots were indexed to (002), (220) and (112) corresponding to monoclinic WO3 as confirmed by XRD analysis. The HRTEM images (Figure 3e,f) displayed rectangular rods and sheet-like shapes as observed in Figure 3d, which implies that no shape distortion had occurred through metal doping and the formation of the heterojunction. The SAED image (inset) displays spots and rings characteristic of monoclinic WO3 and the SnS2 hexagonal phase, respectively. Furthermore, the SAED image (inset) illustrates spot (202, 200) and ring (101, −103) indices corresponding to the WO3 monoclinic phase and SnS2 hexagonal phase, respectively, captured through a 0–10 zone axis using the CrysTBox software [30] (Figure 3f). All the SAED indices correspond to the reported XRD patterns, which further confirms the successful formation of our nanomaterial. In the Mn-WO3/SnS2, the estimated percentage of Mn was 2.5%, with 47.5% of WO3 and 50% of SnS2.
#### *2.3. Optical Properties*
Ultraviolet-visible spectroscopy in diffuse reflectance mode was used to determine the optical properties of the synthesised nanoparticles. All the synthesised nanomaterials showed a shift of absorption to be in the visible region, which is in abundance. Pristine *m*-WO3 displayed a band gap of 2.71 eV, with a corresponding absorption wavelength of 466 nm, as shown in Figure 4a and Figure S4, respectively. The value obtained agrees with the value reported by Simelane et al. in 2017 [15]. The insertion of Mn2<sup>+</sup> in the *m*-WO3 lattice introduced impurities and thus a shift in the Fermi level below the conduction band, promoting the red-shifting of WO3 on the absorption spectrum (Figure 4b);
this was observed by Harshulkhan et al. in 2017 using magnesium as a dopant [6]. The band gap of the WO3 was reduced upon the insertion of Mn and decreased further after the formation of a heterojunction with SnS2 (Figure 4a,b,d,e). The Mn-doped heterojunction had the lowest band gaps (2.08 eV and 2.34 eV) amongst the nanomaterials (the others were WO3, SnS2, Mn-WO3 and WO3/SnS2), which correspond to a high light absorption wavelength (red-shift) (Figure S4); this was due to visible light absorption enhancement by both the Mn ion and SnS2.
**Figure 4.** Tauc plots indicating the band gaps of (**a**) WO3, (**b**) Mn-WO3, (**c**) SnS2, (**d**) WO3/SnS2 and (**e**) Mn-WO3/SnS2.
The diagram in Figure 5 illustrates the change in the band edge potential of the synthesised semiconductor photocatalysts. The valence band edge potential (*E*VB) and the conduction band edge potential (*E*CB) were calculated using Equations (5) and (6). A slight decrease in both *E*CB and *E*VB was observed during the introduction of Mn and fusion with SnS2 (Mn-WO3/SnS2). The conduction band edge potential shifted to be more positive (by 0.2 eV), and the valence band edge potential moved to a less positive potential (by 0.2 eV); this was due to the insertion of an ion with a high ionic radius, which reduces the band gap by pulling the band edges closer, resulting in band edge shifts. The change in the position of the band edges enhances the absorption wavelength of the material. The heterojunction (Mn-WO3/SnS2) enhances charge separation by the movement of electrons from the SnS2 CB to the WO3 CB through the interface, thereby leaving holes in the VB of the SnS2. This effectively separates the electrons and holes as they accumulate in the CB of WO3 and the VB of SnS2, respectively.
**Figure 5.** Diagram of the band *gap*, *valence band* and *conduction* band *edge positions* vs. the NHE of the (**a**) WO3, (**b**) SnS2, (**c**) Mn-WO3, (**d**) WO3/SnS2 and (**e**) Mn-WO3/SnS2 photocatalysts.
#### *2.4. Electrochemical and Photoluminescence Measurements*
The electrochemical impedance spectroscopy (EIS) measurements were carried out to study the interfacial reactions occurring between the photoelectrode and the electrolyte. Figure 6a illustrates the EIS spectrum (Nyquist plot) with a suppressed semicircle with a large diameter. At low frequency, the current density is in phase with the potential deviation of the system, resulting in a straight line at an angle of 45◦ to the *X*-axis. The large diameter of the semicircle at high frequency corresponds to the high charge transfer impedance of WO3. This relates to the high charge recombination rate as observed in Figure 6a. The charge transfer impedance was reduced after WO3 was doped with the Mn2<sup>+</sup> ion (Figure 6b), due to the reduced charge recombination rate and increased charge mobility. This was due to the Mn2<sup>+</sup> ions acting as charge collection sites, thereby serving as an electrical conduction pathway, allowing ion/electron mobility on the electrode [7]. The small diameter of the semicircle of the Mn-WO3/SnS2 spectrum indicates decreased electrode–electrolyte charge-transfer resistance/impedance compared to that in the WO3/SnS2, Mn-WO3, SnS2 and WO3 in the 0.1 M Na2SO4 electrolyte. The sloping straight line in the low-frequency region corresponds to oxygen diffusion within the electrode (Figure 6a). The low charge transfer resistance of Mn-WO3/SnS2 arises from the enhanced charge carrier separation induced by the Mn2<sup>+</sup> ion dopant and SnS2 semiconductor heterojunction with WO3. The charge transfer resistance and recombination rate decreased for WO3, Mn-WO3, SnS2 and WO3/SnS2, with the lowest rate observed in the Mn-WO3/SnS2 (Figure 6a,b).
**Figure 6.** (**a**) Electrochemical impedance spectra (Nyquist plot), (**b**) photoluminescence spectra, (**c**) linear sweep voltammetry, and (**d**) Mott–Schottky plot of (a) WO3, (b) Mn-WO3, (c) SnS2, (d) WO3/SnS2 and (e) Mn-WO3/SnS2.
Photoluminescence (PL) measurements (Figure 6b) support the EIS findings that the WO3 has a strong PL intensity, which indicates high charge carrier recombination, and it was reduced by the introduction of Mn and fusion with SnS2. A decrease in PL intensity was observed in the Mn-WO3/SnS2, indicating low charge carrier recombination, which implied that it would be a good photocatalyst in photocatalytic applications. This is attributed to the longer charge carrier lifetimes and enhanced charge carrier mobility provided by the Sn-S bond, thereby minimising the electron–hole recombination.
Upon the introduction of Mn2<sup>+</sup> and SnS2, the photocurrent density of WO3 was observed to be improved by up to 0.030 mA/cm<sup>2</sup> for Mn-WO3/SnS2 NPs (Figure 6c). This implied that there was high electron flow between the photocatalyst and the electrolyte produced from the photocatalyst upon light irradiation. The Mn2<sup>+</sup> acts as an electron sink and reaction side, which in turn supplies electrons for interfacial reactions, and upon illumination, SnS2 helps in the production of electrons and their separation from holes, which increases the current density.
Mott–Schottky plots were used to study the interfacial capacitance of the nanomaterials. The positive slopes obtained from Figure 6d confirmed that the synthesised nanomaterials are all *n*-type semiconductors, which use electrons as major charge carriers. The positive slope for the heterojunction WO3/SnS2 and Mn-WO3/SnS2 NPs further inferred the formation of an *n–n* type heterojunction system. Upon the introduction of the Mn2<sup>+</sup> and formation of the heterojunction, there was no significant change in the slope of the curves.
The flat-band potential (*V*fb) was obtained by extrapolating a line on the slope of the graph to the *x*-intercept (1/C2=0). The flat-band potentials were found to be 0.214 V, 0.159 V, <sup>−</sup>0.209 V, <sup>−</sup>0.103 V and −0.039 V, corresponding to WO3, Mn-WO3, SnS2, WO3/SnS2 and Mn-WO3/SnS2, respectively. The flat-band potential in *n*-type semiconductors corresponds to the bottom of the conduction band of the semiconductor photocatalyst, which was observed to decrease upon doping and the formation of the heterojunction (Figure 6d). The obtained flat-band potential (*V*fb) values were found to correspond to the calculated conduction band edge potentials (*E*CB) from UV-Vis DRS.
The Randles equivalent circuit model was used to fit the obtained EIS data. The Randles equivalent circuit models (Figure 7a,b) corresponding to the graphs show that the impedance was a contribution of three forms of resistance, namely the solution resistance, the electrode resistance due to the film composition of the nanomaterials, and charge-transfer resistance occurring at the electrolyte–electrode interface.
**Figure 7.** Randles equivalent circuit models corresponding to (**a**) WO3 and Mn-WO3, and (**b**) SnS2, WO3/SnS2 and Mn-WO3/SnS2.
The Warburg impedance is due to solid-state ion diffusion during the electrochemical reaction. The Warburg element manifests itself in EIS spectra as a straight line with a slope of 45◦ in the low-frequency region. Different slopes of the straight-line part in the low-frequency region indicate that the electrodes have different Warburg impedances and solid-state ion diffusion behaviors.
The equivalent circuit model was obtained after fitting the data to the Randles model: R1 is the solution resistance, R2 is the thin layer resistance, R3 is the charge transfer resistance, W2 is the Warburg resistance and Q1, Q2 and Q3 are the constant phase elements. The Warburg impedance relates to solid-state ion diffusion during the electrochemical reaction in the solution. This favours photocatalytic activity by utilising the separated charges during the reaction and, consequently, reduces charge recombination. The slope of the Warburg transition line also indicates the reactivity of the nanoparticles. Furthermore, the charge transfer impedance was found to be 631.80, 498.50, 310.55, 173.65 and 157.16 Ω for WO3, Mn-WO3, SnS2, WO3/SnS2 and Mn-WO3/SnS2, respectively.
#### *2.5. BET Analysis*
The BET analysis revealed that the nitrogen adsorption isotherms obtained for the nanoparticles were Type IV isotherms, according to the IUPAC (International Union of Pure and Applied Chemistry) classification (indicated in Figure S6). A Type IV isotherm is typical of mesoporous materials (IUPAC definition: pore size 2–50 nm), suggesting that the nanomaterials consist of agglomerates. The Mn-WO3/SnS2 nanoparticles were found to have the highest BET surface area (77.14 m2/g) and pore volume, of 0.0641 cm3/g, compared to pristine materials (Table 1). This suggests that Mn-WO3/SnS2 would have improved photocatalytic activity due to the adsorption capacity provided by its large specific surface area during photocatalysis. The large pore volume would allow the efficient trapping of pollutants during adsorption for degradation to take place.
**Table 1.** The specific surface area and pore volume of NPs.
#### *2.6. Surface Charge of Nanoparticles*
The stability of the nanomaterials in suspensions was studied using the electrophoretic light scattering technique. The zeta potentials of the nanomaterials are illustrated in Figure 8. showing a steady but gradual change in zeta potential from positive to negative as the pH increased from 2 to 11 for all the photocatalysts (Figure 8).
**Figure 8.** Surface charge of the nanoparticles.
The point of zero charge (pzc) for pristine WO3 was observed at pHpzc 2.5, which corresponds to what is reported in the literature.
A slight shift of the pzc to higher pH was observed for Mn-WO3 (pHpzc = 3.2). The shift is due to the substitution of W6<sup>+</sup> by Mn2<sup>+</sup> metal ions, consequently changing the overall charge of the material. Therefore, species adsorbed onto the surface of the photocatalyst change the surface charge and shift the point of zero charge of the suspended nanoparticles.
The point of zero charge for pristine SnS2 was found to be at pH 5.5, as reported in literature. Furthermore, the heterojunction (WO3/SnS2) displayed a point of zero charge (2.7) at a lower pH than SnS2 but higher than WO3; this was attributed to synergistic effects from both counterparts (WO3 and SnS2) in the heterojunction.
Furthermore, introduction of Mn in the heterojunction (Mn-WO3/SnS2) shifted the point of zero charge to 2.1, much lower than for all the other photocatalysts.
#### *2.7. Degradation of Chlorpyrifos*
The photodegradation of chlorpyrifos using the synthesized nanoparticles is showed in Figure 9. The degradation profile for chlorpyrifos indicated an increase in removal by the nanoparticles from the WO3 to Mn-WO3/SnS2 photocatalysts. The Mn-WO3/SnS2 nanoparticles showed high removal of chlorpyrifos due to high charge separation and lower charge impedance. Therefore, Mn-WO3/SnS2 represented the best performing photocatalyst with up to 95.90% chlorpyrifos removal, calculated using Equation (10).
**Figure 9.** Degradation of chlorpyrifos (1000 ppb) using different photocatalysts at pH = 5 and 0.1 g of photocatalyst.
Figure 10 displays the percentage removal of chlorpyrifos in water within a period of 60 min. The removal efficiency for chlorpyrifos by using the nanoparticles resulted in 56.80%, 60.20%, 75.00%, 84.88% and 95.90% removal for WO3, Mn-WO3, SnS2, WO3/SnS2 and Mn-WO3/SnS2, respectively.
**Figure 10.** Percentage removal for chlorpyrifos (1000 ppb) using 0.1 g of (**A**) WO3, (**B**) Mn-WO3, (**C**) SnS2, (**D**) WO3/SnS2 and (**E**) Mn-WO3/SnS2.
The reaction kinetics correspond to the percentage chlorpyrifos removal. The rate constants (K) of the reactions using the respective photocatalysts are presented (Figure 11), which were 9.3 <sup>×</sup> <sup>10</sup>−<sup>3</sup> <sup>M</sup><sup>−</sup>1min−<sup>1</sup> and 209 <sup>×</sup> <sup>10</sup>−<sup>3</sup> <sup>M</sup><sup>−</sup>1min−<sup>1</sup> for WO3 and Mn-WO3/SnS2, respectively.
**Figure 11.** Rate constants of (**A**) WO3, (**B**) Mn-WO3, (**C**) WO3/SnS2, (**D**) Mn-WO3/SnS2 and (**E**) SnS2.
The photodegradation reaction was fitted to Equation (10) from which the rate constant k(M−1min−1) was calculated from the gradient of the plot of 1/[C] against time (t). The reaction kinetics leading to the determination of the rate constant followed a second order reaction pathway.
The rate constants were 9.3 <sup>×</sup> 10−<sup>3</sup> M−1min−1, 14.3 <sup>×</sup> 10−<sup>3</sup> M−1min−1, 25.0 <sup>×</sup> 10−<sup>3</sup> M−1min−1, 47.4 <sup>×</sup> <sup>10</sup>−<sup>3</sup> <sup>M</sup><sup>−</sup>1min−<sup>1</sup> and 209.5 <sup>×</sup> <sup>10</sup>−<sup>3</sup> <sup>M</sup><sup>−</sup>1min−1, corresponding to WO3, Mn-WO3, SnS2, WO3/SnS2 and Mn-WO3/SnS2, respectively (Figure 11).
The linear plot for the Mn-WO3/SnS2 nanoparticles kinetic studies is illustrated in Figure 12. The rate constant is 209.5 <sup>×</sup> <sup>10</sup>−<sup>3</sup> <sup>M</sup><sup>−</sup>1min−1, and the R2 is 0.9656.
**Figure 12.** Photodegradation kinetics for chlorpyrifos using Mn-WO3/SnS2.
#### *2.8. E*ff*ect of pH on the Photocatalytic Activity*
The surface charge of the nanoparticles in a suspension is influenced by the pH of the solution. The photodegradation of chlorpyrifos using Mn-WO3/SnS2 nanoparticles increased with an increase in the pH of the initial solution, as illustrated in Figure 13. The point of zero charge for Mn-WO3/SnS2 is at pHpzc = 2.13 and above that is increasingly negative, as displayed by Equations (1) and (2).
$$\rm{M-OH} + \rm{H}^+ = \rm{M-OH}\_2^+ \ (\rm{pH} < \rm{pzc}) \tag{1}$$
$$\text{M-OH} = \text{M-O}^{-} + \text{H}^{+} \text{ (pH} > \text{pzc)} \tag{2}$$
The increase in the removal was also caused by the increase in the level of deprotonation of the nanoparticles at high pH, which influences the negative charge on the surface of the photocatalyst, consequently leading to high chlorpyrifos adsorption. That was also favoured by the positive charge of chlorpyrifos in alkaline solutions from pH 5, as reported in literature. There is a transfer of holes from the inner part of the nanoparticles to the surface, whereby OH− ions scavenge photogenerated holes and therefore yield very oxidative species such as •OH radicals. The percentage removal of chlorpyrifos achieved in 60 min was 85.6%, 94.3%, 99.8% and 99.0% at pH 3, pH 5.8, pH 7 and pH 9, respectively (Figure 13). Therefore, pH 7 was the optimum pH for chlorpyrifos removal using Mn-WO3/SnS2 nanoparticles and was used in the next sections. Hou et al. [31] in 2018 also reported pH 7 for optimum chlorpyrifos removal [20].
The increase in the removal was due to the increased electrostatic attraction between the photocatalyst and the chlorpyrifos that occurs when the pH is increased [28]. This causes an easy surface attachment, which implies that holes can oxidize chlorpyrifos directly and creates hydroxyl and superoxide radicals for further oxidation. As pH increased, the surface charge of the nanoparticles
also became more negative, which caused increased electrostatic attraction between the nanoparticles and chlorpyrifos.
**Figure 13.** Degradation of 1000 ppb chlorpyrifos using 0.1 g of Mn-WO3/SnS2 at different pH values.
#### *2.9. E*ff*ect of Initial Concentration*
The effect of initial chlorpyrifos concentration on the photocatalytic removal was studied, and the results are shown in Figure 14. The highest removal of 99.99% was achieved at a 100 ppb chlorpyrifos concentration, followed by 99.95% at 1000 ppb, compared to 94.40%, 87.51% and 84.38% at 5 ppm, 10 ppm and 20 ppm, respectively. The concentration of 1000 ppb was chosen as the best, because it is the highest concentration for which a high percentage removal was achieved.
**Figure 14.** Effect of the initial concentration on the removal of chlorpyrifos (1000 ppb) at pH 7 using 0.1 g of Mn-WO3/SnS2.
The decrease in the removal of chlorpyrifos was alluded to the opacity caused by the high chlorpyrifos concentration, which prevented the photocatalyst from utilising the irradiated light to produce reactive species for degradation. Again, the high concentration scatters the light, thereby inducing screening effects [28].
#### *2.10. E*ff*ect of Initial Photocatalyst Loading*
The initial photocatalyst loading's effect on the photoactivity was studied, and the results are presented in Figure 15. The photoactivity of Mn-WO3/SnS2 increased when 0.5 g of photocatalyst was used, then further increased when the photocatalyst loading was 1 g. The increase is a result of increased reactive surfaces, which further increase the rate and amount of chlorpyrifos removal [28].
**Figure 15.** Effect of the initial photocatalyst loading on the photodegradation of chlorpyrifos.
A high concentration of nanoparticles results in agglomeration, which further causes light scattering and screening effects, which reduce the specific activity of the photocatalyst. This further causes opacity, which prevents the further illumination of the photocatalyst.
Therefore, a decrease in chlorpyrifos removal was observed when 2 g of Mn-WO3/SnS2 was used, reaching up to 85%. This is compared to 0.1 g, 0.5 g and 1 g removing up to 99.95%, 99.98% and 99.99%, respectively. Thus, 1 g was the best performing, as it reached 98% removal within 30 min of reaction time.
#### *2.11. Mechanistic Pathway*
The mechanistic and proposed degradation pathway was evaluated, and the results are shown in Figure 16. The products obtained were 3,5,6-trichloropyridin-2-ol (TCP) and O,O-dihydrogen phosphorothioite. Only the O,O-dihydrogen phosphorothioite compound and no other by-product was observed, which implies that there was a complete degradation of chlorpyrifos and TCP in the synthetic water (Figure S7).
**Figure 16.** Proposed degradation pathway for chlorpyrifos.
#### **3. Materials and Methods**
#### *3.1. Materials*
Tungsten boride (WB) (≥97.0%), Tin(IV) chloride pentahydrate (SnCl4.5H2O, (98%)), manganese(II) chloride tetrahydrate (MnCl2•4H2O, (≥98%)), nitric acid (≥65%, Puriss), poly (vinylidene fluoride) (PVDF), *N*-methyl-2-pyrrolidone (NMP), silver paste, sodium sulphate (Na2SO4), silver/silver chloride (Ag/AgCl) electrode, sodium sulphide (Na2S), chlorpyrifos (PESTNATAL, 99.9%), methanol and formic acid were all supplied by Sigma-Aldrich (Pty) Ltd., Johannesburg, South Africa. The chemicals were used as received.
#### Synthesis of Nanomaterial
Pristine WO3 NPs were synthesised following the method developed by Xie et al. (2012) with slight modification [3]. Tungsten boride (4.12 mmol) was dissolved in 1 M HNO3 (56.0 mL) under constant stirring and then transferred into a 100.0 mL Teflon-lined stainless steel autoclave. The autoclave was then sealed and placed in an oven at 190 ◦C for 12 h with a heating rate of 16 ◦C per hour; thereafter, a yellow solution was obtained. The yellow solution was further centrifuged and washed with deionised water and dried at 100 ◦C for 12 h in an oven, resulting in a yellow solid product of WO3. The same procedure was followed to obtain Mn-WO3 via the one-pot synthesis of MnCl2•4H2O (10.00 mmol) with WB (4.12 mmol) in 56.0 mL of HNO3.
Pristine SnS2 NPs were synthesised by dissolving 2.73 mmol of SnCl4.5H2O in 40.0 mL of deionised water under continuous stirring at 60 ◦C for 10 min. Thereafter, Na2S (2.73 mmol) was added to the solution, and the mixture was then stirred for 10 min. The final mixture was then transferred into a 100.0 mL Teflon-lined stainless steel autoclave and heated at 180 ◦C for 12 h with a heating rate of 15 ◦C/h. The resultant solution was then centrifuged and washed with deionised water and dried at 100 ◦C for 12 h to obtain SnS2 nanoparticles.
The heterojunction WO3/SnS2 was synthesised stepwise using a hydrothermal method. The first step was adapted from the method for synthesising pristine WO3 and followed by the synthesis of SnS2 NPs on the surface of the dispersed WO3 NPs in 40.0 mL of deionised water. The SnS2 synthesis was adopted from the synthesis of pristine SnS2 NPs to obtain the heterojunction (WO3/SnS2).
Furthermore, the synthesis of the Mn-doped heterojunction WO3/SnS2 was carried out using hydrothermal treatment in a multistep method. Firstly, MnCl2•4H2O (10.0 mmol) and tungsten boride (4.12 mmol) were dissolved in 1 M HNO3 (56.0 mL) under constant stirring, and thereafter, the same procedure as in the synthesis of WO3 was followed to obtain a yellow solid product of Mn-WO3. Furthermore, Mn-WO3 (1.74 mmol) was dispersed in 40.0 mL of deionised water under continuous stirring and heating at 60 ◦C using a magnetic stirring hotplate. This was followed by the synthesis of SnS2 NPs on the surface of Mn-WO3 NPs adopted from the synthesis of pristine SnS2 nanomaterials to form Mn-doped WO3/SnS2 heterojunction composite nanoparticles.
#### *3.2. Characterization Techniques*
The synthesised nanoparticle phases were characterized using X-ray powder diffraction (XRD) (*PANalytical X'Pert* Pro-MPD Powder Diffractometer, Almelo, Netherlands) with CuKα radiation (0.1540 nm) and a monochromator beam in a 2θ scan range from 20◦–80◦. The instrument power settings used were 40 kV and 40 mA with a step size of 2θ (0.0080) and a scan step time of 87.63 s. The average crystallite size was calculated using the Debye–Scherrer equation, Equation (3):
$$L = \frac{K\lambda}{\beta \cos \theta} \tag{3}$$
where β is the full width at half maximum, λ is the X-ray wavelength (0.1541 nm) for CuKα, *K* = 0.89, and θ is the diffraction angle.
Raman spectroscopy (RamanMicro™ 200 PerkinElmer Inc., Waltham, MA, USA) with a single monochromator, a holographic notch filter and a cooled TCD, was used to detect and characterise the polymorphic forms of the NPs. The Raman spectra of the NPs were measured in a back-scattering geometry using an Ar-ion laser line (514.5 nm). Dark-field imaging was used with a power output of below 0.5 mW and an exposure time of 4.0 s. The morphological properties of the NPs were examined using high resolution transmission electron microscopy (HRTEM) (JOEL-TEM 2010) at an acceleration voltage of 200 kV. The ethanol-dispersed nanoparticles were deposited on a carbon-coated copper grid. Furthermore, selected area electron diffraction (SAED) images of the nanoparticles were captured and indexed using the CrysTBox software [29]. A field emission scanning electron microscope (FESEM) (TESCAN Vega TC instrument with VEGA 3 TESCAN software; TESCAN, Brno, Czech Republic) coupled with energy-dispersive X-ray (EDX) operated at 5.0 kV under a nitrogen gas atmosphere was used to further study the morphology and the elemental composition of the NPs. The optical properties were investigated using a UV-Vis spectrophotometer (Shimadzu UV-2450, Shimadzu Corporation, Kyoto, Japan) using diffuse reflectance spectroscopy (DRS) and BaSO4 as the reference material. The band gap (*Eg*) of the nanomaterials and a graph of (α*h*ν) against photon energy (*h*ν) was extrapolated following Equation (4):
$$
\alpha \text{lrb} = A(\text{lrb} - E\_{\%})^{\text{n/2}} \tag{4}
$$
where α is the absorption coefficient, *h*ν is the energy of the incident photon, *A* is a constant, and *Eg* is the band gap energy.
The value of *n* depends on the semiconductor transition type, which is a direct transition when *n* equals 0.5 and an indirect transition when *n* equals 2. The valence band edge potential (*EVB*) and the conduction band edge potential (*ECB*) were calculated using Equations (5) and (6):
$$E\_{CB} = \chi - E^{\text{e}} - 0.5E\_{\%} \tag{5}$$
$$E\_{VB} = E\_{CB} + E\_{\mathcal{X}} \tag{6}$$
where *ECB* and *EVB* are the conduction and valance band edge potentials, respectively; χ is the electronegativity of the semiconductor (the geometric mean of the electronegativities of all the constituent atoms); *E*<sup>e</sup> is the energy of free electrons on the hydrogen scale (4.5 eV); and *Eg* is the band gap energy of the semiconductor.
The photoluminescence spectra of the nanomaterials were obtained using a PerkinElmer fluorescence spectrometer (Model LS 45, PerkinElmer Inc., Waltham, MA, USA). A 300 W xenon lamp was used as a light source. The spectra were obtained at an excitation wavelength of 319 nm. The excitation and emission wavelengths were set at 319 nm and 605 nm, respectively. Specific surface area and pore volumes were determined using the Brunauer–Emmett–Teller (BET) method. Nitrogen was used as the adsorbate, and the nitrogen adsorption isotherms of the samples were obtained at 77K using a Micromeritics ASAP 2020 adsorption analyser (Micromeritics Instrument Corporation, Norcross, Georgia, USA). The samples were degassed before the analysis at 100 ◦C for 10 h. The pore volume was calculated from the amount of nitrogen adsorbed at the relative pressure (*P*/*P*o) of 0.980.
#### *3.3. Electrochemical Measurements*
The electrochemical measurements were conducted using a potentiostat (Gamry Interface 1000 potentiostat, Gamry Instruments, Philadelphia, PA, USA) in a standard three-electrode system employing Ag/AgCl (3.0 M KCl) as the reference electrode and Pt wire as the counter electrode. The working electrodes were the prepared nanomaterial mixed with polyvinylidene fluoride (PVDF) as a binder in a 10:1 ratio respectively, dispersed in 1 mL of N-methylpyridinium (NMP) solution and ultrasonicated for 30 min to obtain a homogeneous mixture. The obtained homogeneous mixture was drop casted onto the fluorine-doped titanium oxide (FTO-glass) substrate forming a thin film. The prepared electrodes were heated at 80 ◦C for 12 h in air. A copper wire was thereafter attached using a silver paste for charge transfer to the potentiostat from the paste and dried in air for 24 h.
The prepared electrodes were then applied in a three-electrode system for electrochemical impedance spectroscopy (EIS) at a frequency range of 10 kHz to 0.1 Hz at an AC voltage of 10 mV rms and DC voltage of 0.45 V vs. Ag/AgCl. The current density of the working electrode was determined by running a linear sweep voltammetry scan at a scan rate of 50 mV/s. The flat-band potential (*V*fb) values of the nanomaterials were obtained from Mott–Schottky plots (Equation (7)) at a frequency of 1000 Hz under the applied voltages of −2 to 2 V and a step voltage of 0.1 V.
$$\frac{1}{C^2} = \frac{2}{\left(\varepsilon \varepsilon\_o A^2 \varepsilon N\_D\right)} \left[ V - V\_{fb} - \left(\frac{k\_b T}{\varepsilon}\right) \right] \tag{7}$$
where *C* is the interfacial capacitance, *A* is the surface area of the electrode, *N*<sup>D</sup> is the donor density, *V* is the applied potential, and *Vfb* represents the flat-band potential. The temperature with dielectric constant and permittivity of free space are represented as *T*, ε and ε0, respectively. The charge of the electron (*e*) is 1.602 <sup>×</sup> 10−<sup>19</sup> C, and the Boltzmann constant (*k*B) is 8.617 <sup>×</sup> 10−<sup>5</sup> eV·K−1. All the electrochemical measurements were conducted in 0.1 M sodium sulphate (Na2SO4) solution as the electrolyte, and the values for the electrode potentials were recorded with reference to Ag/AgCl. A 300 W xenon lamp was used as the light source.
#### *3.4. Surface Charge*
Surface charge measurements were obtained using electrophoretic light scattering (ELS) with a Zetasizer NanoZS (Malvern) instrument. Zeta potential measurements were obtained using electrophoretic light scattering (ELS) to understand the surface charge of the nanomaterials as a function of the pH of the solution. The nanomaterials were suspended at 30 mg/L in deionized (DI) water. The pH of the suspensions was adjusted to a pH range of 2–10 using 1M NaOH and 1M HCl.
#### *3.5. Degradation of Chlorpyrifos*
#### 3.5.1. Chlorpyrifos Standard Preparations
A stock solution of Chlorpyrifos (0.01 g) was prepared in 1 L of deionized water, followed by a serial dilution to make 75, 50, 25, 12.5, 6.25, and 3.125 ppb solutions. The prepared solutions were thereafter transferred into 2 mL LC-MS vials, and 1 mL of deionized water was added. The working solution was maintained at pH 5.
#### 3.5.2. Photocatalytic Degradation of Chlorpyrifos
The photocatalytic activity of the nanomaterials was tested through the photodegradation of chlorpyrifos in synthetic water samples under visible light irradiation (Photoreactor, Lelesil Innovative Systems). The volume of the working solution was kept at 500 mL of chlorpyrifos solution. Initially, the concentration of the chlorpyrifos solution was 1 ppm and a photocatalyst loading of 0.1 g was used at pH 5. The photodegradation reaction occurred under continued magnetic stirring for 90 min under regulated temperatures of 20–25 ◦C, subjected to a cooling jacket using ice cubes.
The photocatalyst suspension containing chlorpyrifos was kept in the dark for 30 min before irradiation to allow equilibration. The samples were collected from the batch reaction before and after irradiation at set time intervals (10 min, 10 mL aliquots), filtered through a 0.45 μm PTFE membrane filter and transferred into a 2 mL LC-MS sample vial for analysis.
Furthermore, the optimization of reaction conditions such as the pH, initial chlorpyrifos concentration and initial photocatalyst loading were carried out. Therefore, the pH of the chlorpyrifos solution was adjusted to 3, 5, 7 and 9; the initial concentration of the pesticide (chlorpyrifos), to 100 ppb, 1 ppm, 5 ppm, 10 ppm and 20 ppm; and the photocatalyst loading, to 0.1, 0.5, 1 and 2 g.
#### *3.6. LC-MS Measurement*
Samples were analyzed using a triple quad UHPLC-MS/MS 8030 (Shimadzu Corporation) to monitor the removal of chlorpyrifos. The LC-MS/MS was fitted with a Nexera UHPLC upgrade with the capability to obtain 500 multiple reaction monitoring readings per second. The oven was equipped with a RaptorTM ARC-18 column (Restek Corporation) with a 2.7 μm pore diameter and length of 100 mm × 2.1 mm, maintained at 40 ◦C. The mobile phase consisted of 0.1% formic acid in water/methanol (9:1%, v/v) at a flow rate of 0.200 mL/min with a 10 μL injection volume. The ion source was electrospray ionisation (ESI) and was operated in positive mode. Meanwhile, LC-MS/MS data for the degradation intermediates were obtained after the full scan mode was run for 12 min at flow rate of 0.3 mL/min.
The percentage removal of chlorpyrifos from the synthetic water samples was calculated using Equation (8) below:
$$\% \text{ chlorpyryfos removal} = \left(1 + \frac{\text{C}}{\text{C}\_0}\right) \times 100 \tag{8}$$
where *C*<sup>0</sup> is the initial concentration and *C* is the final concentration of chlorpyrifos. The degradation products were determined by analysing the samples for a period of 60 min. The degradation pathway was then deduced from the mass/ion ratio obtained from the MS spectrum. The reaction kinetics of chlorpyrifos degradation were studied, the results were fitted to a second order model fitted, and a plot based on the calculated (1/[*C*]) versus reaction time was obtained following Equation (7).
$$\frac{1}{[\mathbb{C}]\_t} = kt + \frac{1}{[\mathbb{C}]\_0} \tag{9}$$
where *k* is the rate constant, *t* is time taken for the reaction, [*C*]t is the concentration of chlorpyrifos when time is equal to t, and [*C*]0 is the initial concentration of chlorpyrifos. The reaction rate is thus given by Equation (10):
$$rate = k[\mathbb{C}]^2\tag{10}$$
#### **4. Conclusions**
The Mn-doped WO3/SnS2 photocatalyst was successfully synthesized, resulting in a highly crystalline structure. Rectangular rods and sheet-like shapes were observed in the composite, confirming that no shape distortion had occurred in the heterojunction photocatalyst. The composite comprises both hexagonal and monoclinic phases that correspond to SnS2 and WO3, respectively, as confirmed by XRD patterns and Raman spectra. As shown in the UV-Vis spectra of the composite, a shift in the *band edge* (*absorption band edge*) from the UV to the visible region (red shift) was observed in the Mn-doped WO3/SnS2 photocatalyst relative to that for the pristine photocatalysts. The surface area of the WO3 was improved by more than 10 times by intrinsic doping with the Mn2<sup>+</sup> ion and the formation of the heterojunction with SnS2 to form the Mn-doped WO3/SnS2 photocatalyst. The Mn-doped composite was fully characterised using microscopic and spectroscopic techniques, which confirmed the synthesised composite to be Mn-WO3/SnS2. The Mn-doped WO3/SnS2 showed good electrochemical performance, ascribed to its high current density and lower interfacial charge transfer resistance, observed using electrochemical measurements (EIS), which correspond to high charge separation and a low photogenerated charge carrier recombination rate, observed using photoluminescence (PL) measurements. Chlorpyrifos has been applied extensively in agriculture, both in South Africa and other parts of the world, to fight against pests, therefore finding its way into water systems. Chlorpyrifos removal from synthetic water was investigated using Mn-WO3/SnS2 nanoparticles. The removal was due to the enhanced charge separation, high charge transfers and high electrostatic attraction between the nanoparticles and chlorpyrifos.
After the optimization of the reaction conditions, the chlorpyrifos removal achieved was 99.99% at pH 7 with 1 g of Mn-WO3/SnS2 and a 1000 ppb concentration.
The degradation pathway was also investigated, for which 3,5,6-trichloropyridin-2-ol and O,O-dihydrogen phosphorothioite were observed. Furthermore, after 60 min of the reaction, only O,O-dihydrogen phosphorothioite was detected. This implies that both chlorpyrifos and TCP were completely degraded. The results suggest that our material, Mn-WO3/SnS2, can completely degrade chlorpyrifos and its major degradation product.
**Supplementary Materials:** The following are available online at http://www.mdpi.com/2073-4344/10/6/699/s1, Table S1: Average crystallite sizes of nanomaterials; Figure S2: (a) FESEM image of pristine Mn-WO3, (b) TEM image of Mn-WO3, (c–e) elemental mapping, and (f) EDX spectrum of Mn-WO3 nanoparticles; Figure S3: (a) FESEM image of pristine SnS2, (b) TEM image of SnS2 (inset is the corresponding SAED image), (c,d) elemental mapping, and (e) EDX spectrum of SnS2 nanoparticles; Figure S4: Absorption spectra of WO3, SnS2, Mn-WO3, WO3/SnS2, and Mn-WO3/SnS2; Figure S5: EIS spectra showing the fitted spectra when obtaining the Randles circuit for (a) WO3 and Mn-WO3, and (b) SnS2, WO3/SnS2 and Mn-WO3/SnS2; Figure S6: (a–e) N2 adsorption-desorption isotherm of (a) WO3, (b) Mn-WO3, (c) Mn-WO3/SnS2, (d) WO3/SnS2, and (e) SnS2 (insets are pore volume graphs); Figure S7: Calibration curve of chlorpyrifos standards from 3.125 to 75 ppb; Figure S8: Mass spectra showing m/z ratios from 0 to 60 min; Figure S9: Fitted second order reaction kinetics graphs of the nanoparticles.
**Author Contributions:** Conceptualization, L.N.D., methodology, C.M.K.; validation, L.N.D., S.P.M. and C.M.K.; formal analysis, C.M.K.; investigation, C.M.K.; resources, L.N.D.; data curation, C.M.K.; writing—original draft preparation, C.M.K.; writing—review and editing, L.N.D., S.P.M.; supervision, L.N.D., S.P.M.; project administration, L.N.D.; funding acquisition, L.N.D. All authors have read and agree to the published version of the manuscript.
**Funding:** This research was funded by THUTHUKA NATIONAL RESEARCH FOUNDATION, grant number 15060-9119-027" and "The APC was funded by UNIVERSITY OF JOHANNESBURG-Accelerated Academic Mentorship Programme".
**Acknowledgments:** The authors would like to extend their gratitude to the University of Johannesburg, Faculty of Science, National Research Foundation (NRF) (TTK 15060-9119-027), TESP Eskom and the Centre for Nanomaterials Science Research, University of Johannesburg.
**Conflicts of Interest:** The authors declare no conflict of interest.
#### **References**
© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
*Article*
| doab | 2025-04-07T03:56:58.566142 | 6-5-2022 13:34 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "0039f385-c9b0-4fa4-b06c-9a5b9a2397e4",
"url": "https://mdpi.com/books/pdfview/book/5221",
"author": "",
"title": "Materials and Processes for Photocatalytic and (Photo)Electrocatalytic Removal of Bio-Refractory Pollutants and Emerging Contaminants from Waters",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036535593",
"section_idx": 2
} |
0039f385-c9b0-4fa4-b06c-9a5b9a2397e4.3 | **Fast Microwave Synthesis of Gold-Doped TiO2 Assisted by Modified Cyclodextrins for Photocatalytic Degradation of Dye and Hydrogen Production**
**Cécile Machut 1,\*, Nicolas Kania 1, Bastien Léger 1, Frédéric Wyrwalski 1, Sébastien Noël 1, Ahmed Addad 2, Eric Monflier <sup>1</sup> and Anne Ponchel <sup>1</sup>**
Received: 29 June 2020; Accepted: 16 July 2020; Published: 18 July 2020
**Abstract:** A convenient and fast microwave synthesis of gold-doped titanium dioxide materials was developed with the aid of commercially available and common cyclodextrin derivatives, acting both as reducing and stabilizing agents. Anatase titanium oxide was synthesized from titanium chloride by microwave heating without calcination. Then, the resulting titanium oxide was decorated by gold nanoparticles thanks to a microwave-assisted reduction of HAuCl4 by cyclodextrin in alkaline conditions. The materials were fully characterized by UV-Vis spectroscopy, X-Ray Diffraction (XRD), Transmission Electron Microscopy (TEM), and N2 adsorption-desorption measurements, while the metal content was determined by Inductively Coupled Plasma Optical Emission Spectroscopy (ICP-OES). The efficiency of the TiO2@Au materials was evaluated with respect to two different photocatalytic reactions, such as dye degradation and hydrogen evolution from water.
**Keywords:** photocatalysis; photodegradation; nanoparticles; gold; TiO2; cyclodextrins
#### **1. Introduction**
During the past decades, photocatalysis received extensive research interest for both limiting toxic wastes and developing clean and renewable sources of energy. Indeed, the association of a semiconductor with the sunlight in order to remove pollutants [1,2] or to produce hydrogen fuel by water splitting [3] could provide a sustainable solution to the crucial problems of environmental pollution and energy shortages [4].
Among a large number of photocatalysts, TiO2 has been extensively investigated due to its good properties such as low cost, non-toxicity, and good stability [5,6]. Anatase phase is particularly recognized for its high photocatalytic efficiency. However, its large band gap (3.2 eV) combined with a high recombination rate of the photogenerated electron/hole pairs (e−/h+) reduce the photon-to-charge carriers conversion efficiency, but also limit the use in photochemical applications under visible or solar light.
In order to overcome these drawbacks and improve the photocatalytic performance of semiconductors, one of the promising strategies consists in introducing noble metals at the surface of TiO2, such as gold nanoparticles [7]. Indeed, the combination with gold nanoparticles aims at inhibiting the electron-hole pair recombination by trapping electrons and facilitating the transfer of holes on the TiO2 surface [8]. Gold nanoparticles (Au NPs) are also known to enhance the activity
of TiO2 under visible-light irradiation due to the localized surface plasmon resonance of Au NPs in the visible light spectrum [9]. However, it is well accepted that the photocatalytic activity of such TiO2@Au composites can strongly depend on the particle size of Au NPs [10] and optimal synthetic conditions must be found, especially to prevent the aggregation of gold nanoparticles [11,12].
Over the last decade, cyclodextrins and derivatives have received great interest in the field of synthesis and stabilization of metallic nanoparticles in aqueous medium [13]. These macrocyclic oligosaccharides, which are well-known to form inclusion complexes with numerous guest molecules via supramolecular interactions [14], can also be used as capping agents to stabilize zerovalent metal nanoparticles, such as Au NPs. Owing to the numerous hydroxyl groups attached to the CD rims, they can also act as efficient reducing agents for the synthesis of Au NPs [15–19]. However, to the best of our knowledge, the use of cyclodextrins to prepare TiO2@Au composites through simple methods of synthesis has been scarcely investigated. Most synthetic routes involve the use of chemically modified cyclodextrins bearing thiol pendant groups as metal binding sites. Their preparations require multistep and complex synthetic procedure as well as the use of time-consuming purification methods. For instance, Zhu et al. developed a method to synthesize TiO2 decorated by the assembly of per-6-thio-β-cyclodextrin and gold nanoparticles. The resulting composite showed very good efficiency for the degradation of methyl orange (MO) under UV light [20]. More recently, TiO2 nanosheets consisting of the combination of Au nanoparticles and mono-6-thio-β-cyclodextrin were prepared for the electrochemical detection of trace of methyl parathion pesticide [21].
Recently, we have reported a sol-gel method using cyclodextrins as both structure-directing agents and metal-complexing agents to self-assemble titania and gold colloids in composite materials with controlled porosity and uniform metal dispersion [22]. Among the various cyclodextrins examined, the TiO2@Au material prepared using the commonly used randomly methylated β-CD (RAME-β-CD) have shown, after calcination, the best catalytic performance for the photodegradation of organic pollutants in water under visible light, due to a good compromise between its textural properties, crystallinity, and Au particle size. However, the preparation of such plasmonic photocatalysts involved a multistep process that occurred over several days (including acid hydrolysis, peptization, maturation, drying, and finally calcination at a high temperature of 500 ◦C to form Au NPs).
In recent years, microwave (MW) irradiation techniques have received considerable attention in the field of nanomaterial synthesis by inducing or enhancing chemical reactions [23–25]. The use of microwave heating may offer several advantages over conventional heating, such as shorter reaction times, higher heating rates as well as higher uniformities of the products. In the literature, a few articles were already devoted to microwave-assisted synthesis of gold nanoparticles protected by cyclodextrin derivatives [16,26,27]. As a matter of fact, Aswathy et al. synthesized β-cyclodextrin capped Au NPs with a mean diameter of 20 nm within a few minutes [16]. More recently, Stiufiuc et al. used native cyclodextrins as reducing and capping agents during the microwave reduction of the gold precursor and obtained stable monodispersed gold nanospheres covered with either α-, β- or γ-CD [26]. However, to the best of our knowledge, the stabilization and anchorage of Au NPs on titania support thanks to CDs under microwave irradiation have never been explored.
In this context, we reported hereby a novel method for elaborating TiO2@Au materials from a two-step microwave-assisted synthetic route without the need for high temperature calcination. Herein, the TiO2 support is synthesized using a microwave-method by hydrolysis of titanium tetrachloride while the cyclodextrins are employed afterwards to produce size-controlled gold metallic nanoparticles anchored on the support, once again under microwave irradiation. We have focused our efforts on randomly methylated-β-CD (RAME-β-CD) and 2-hydroxypropyl-β-CD (HP-β-CD), which are both highly water-soluble and readily available commercially at relatively low cost. The impact of the nature of the carbohydrate precursor is investigated and discussed on the basis of different physicochemical characterizations, including X-ray diffraction (XRD), N2 adsorption-desorption analysis, transmission electron microscopy (TEM), thermogravimetry analyses (TGA), and diffuse reflectance UV-Vis spectroscopy (DRUV-Vis). Finally, the efficiency of these photocatalysts is examined
with respect to two photocatalytic reactions carried out under near-UV-light irradiation (λ > 365 nm), i.e., the oxidative photodegradation of methyl orange and the hydrogen evolution reaction (HER).
#### **2. Results and Discussion**
As described in the Experimental Section, gold-doped TiO2 materials have been synthesized at 150 ◦C with a fast microwave heating using cyclodextrins as reducing agent of the metal precursor and stabilizer of Au NPs. The synthetic procedure is schematically depicted in the Figure 1.
**Figure 1.** Schematic illustration of the two-step microwave (MW) procedure used for the TiO2@Au materials synthesis.
Note that a bare TiO2 control was also prepared in the same conditions as those described for the first step. These conditions were selected based on preliminary experiments, by varying the duration and power of the microwave irradiation, in order to optimize the crystallinity of our titania support. Indeed, the crystallinity is known to be a key factor in the photoactivity of TiO2 particles. The XRD patterns of titania materials prepared from different heating programs (10, 30, and 45 min) and powers (320 and 600 W) are given for comparison in the Figure 2. With increasing the duration of heating at 320 W, we observe that the intensity of the XRD lines progressively increases and narrows, suggesting a growth in crystallite size. The planes (101), (004), (200), (105), and (211) associated to 2θ = 25.3◦, 37.7◦, 48◦, and 55.2–55.9◦ respectively correspond to the anatase phase (Ti-A, JCPDS 21-1272). No XRD signals related to the presence of other crystalline phases such as rutile and brookite are detected. However, the most interesting effects are produced with the power of 600 W, which offers a very good compromise between crystallinity state and rate of anatase formation since this crystalline phase was obtained after only 10 min, this duration being considerably shorter than that applied for conventional sol-gel synthesis [28]. In line with this first optimization, the heating power of microwave irradiation was set to maximum (600 W) for all the further investigations, with a duration of temperature rise of 2 min from room temperature to 150 ◦C (isothermal step-time of 10 min).
The impact of the addition of gold by microwave-assisted reduction of the TiO2 support was further investigated using mixtures of HAuCl4 and modified cyclodextrins in alkaline conditions (see Figure 1, second step). We decided to use the randomly methylated β-cyclodextrin (RAME-β-CD) and the hydroxypropylated β-cyclodextrin (HP-β-CD) to stabilize Au NPs. Indeed, we particularly focused on these two CDs because of their high solubility in water and their beneficial effect on previously described gold-doped TiO2 [22]. RAME-β-CD and HP-β-CD have also the advantages to
offer a number of available hydroxyl groups (8.4 per RAME-β-CD and 21 per HP-β-CD), which are known to play an important role in the reduction processes of metal cations [16,29].
The XRD patterns of these microwave-prepared titania@Au materials are reported in the Figure 3. For comparison, the XRD pattern of a control gold-doped TiO2 prepared in ethanol (selected as model reducing agent), but without cyclodextrin, was also included (TiO2@Au).
It can be noticed, that in addition to the reflections of anatase (Ti-A, JCPDS 21-1272), TiO2@Au, TiO2@Au-RB, and TiO2@Au-HP present broad and low intense peaks at 2θ = 38.2◦, 44.2◦, 64.3◦, and 78.1◦, which could be respectively indexed to the (111), (200), (220), and (311) planes of gold with face-centered cubic crystalline structure phase (JCPDS 04-0784). The Au crystallite sizes could have been estimated from the line broadening of the (200) diffraction peak at 2θ = 44.2◦ by the Debye-Sherrer equation. Interestingly, for the control-doped TiO2 material prepared using ethanol, the size of gold crystallites is ca. 15 nm, while it significantly decreases to ca. 8–10 nm for the materials prepared with HP-β-CD and RAME-β-CD.
The textural characteristics of the titania-based materials were then evaluated by N2 adsorption-desorption analysis. All the samples exhibit type IV adsorption isotherms with distinct hysteresis loops appearing at P/P◦ ≈ 0.5–0.8, thus supporting the mesoporous character of the samples with a monomodal pore size of 4 nm (Figure S1 ESI and Table 1). The specific surface areas of the bare TiO2 (TiO2-control) is close to those prepared by gold-doped TiO2 (240–260 m2.g−1). In the same way, the pore volumes and pore size values are substantially the same whether there is gold or not. The detailed surface properties and the gold loading determined by Inductively Coupled Plasma Optical Emission Spectroscopy (ICP-OES) measurements are summarized in the Table 1. ICP-OES analysis was used to quantitatively determine the gold content in our composites. Interestingly, the gold loading (≈2 wt %) corresponds to a gold incorporation efficiency around 80–90% of the initial amount of metal used during the synthesis.
**Figure 2.** X-Ray Diffraction (XRD) patterns of titania-based materials prepared by microwave heating with different programs of heating (**a**) 320 W 3 min ramp then 10 min at 150 ◦C, (**b**) 320 W 3 min ramp then 30 min at 150 ◦C, (**c**) 320 W 3 min ramp then 45 min at 150 ◦C, and (**d**) 600 W 2 min ramp then 10 min at 150 ◦C.
**Figure 3.** XRD patterns of titania materials prepared by microwave heating: (**a**) bare TiO2, (**b**) TiO2@Au prepared with ethanol, (**c**) TiO2@Au-RB, and (**d**) TiO2@Au-HP.
**Table 1.** Surface properties and gold loading of the different TiO2 materials.
<sup>a</sup> Specific surface area determined by the BET (Brunauer, Emmett et Teller) method in the relative pressure range of 0.1−0.25. <sup>b</sup> Pore volume computed by BJH. <sup>c</sup> Pore size determined by BJH. <sup>d</sup> Gold loading determined by ICP-OES analysis.
The morphology and structure of the TiO2@Au materials were further characterized by TEM analyses (Figure 4). Whatever the materials, the presence of Au NPs deposited onto the surface of TiO2 is observed. When the synthesis is performed under cyclodextrin-free conditions, with ethanol as reducing agent, TEM images (Figure 4a,b) show the presence of gold nanoparticles with a mean diameter of 13.5 nm but with a relatively broad size distribution ranging from 5 to 30 nm and a standard deviation of 5.3 nm (see histogram in Figure 4c). Note that larger gold nanoparticles with diameter ranging from 44 to 78 nm can be also observed (See Figure S2 in ESI).
Although a modest decrease in the mean particle size is noticed when modified β-cyclodextrins (12.5 nm for HP-β-CD and 12.9 nm for RAME-β-CD) are introduced during the microwave-assisted synthesis, it can be seen that, for these two catalysts, gold nanoparticles are more uniformly dispersed over the TiO2 support. Narrower size distributions with standard deviations as low as 2.5–2.8 nm (see histograms in Figure 4f,j) can be clearly found, evidencing the stabilization of small and well-dispersed spherical Au NPs, as can be seen at high magnification (See Figure S3 in ESI). It provides an intimate contact between Au NPs and the TiO2 mesoporous support. Conversely to what was observed with the ethanol procedure, no aggregation or formation of larger particles were observed over TiO2@Au-HP and TiO2@Au-RB.
**Figure 4.** Transmission electron microscopy (TEM) images at magnification of ×25,000 (Scale bar = 100 nm) and ×62,000 (Scale bar = 50 nm) and size distribution of (**a**–**c**) TiO2@Au, (**d**–**f**) TiO2@Au-RB, (**h**–**j**) TiO2@Au-HP.
As previously observed by several teams, cyclodextrins can stabilize metallic nanoparticles in aqueous solution [13]. Because of different types of interactions between the metal and the CDs (hydrophobic-hydrophobic interactions [30], non-covalent interactions between metal ions and hydroxyl groups of the CD [15]) the aggregation of gold nanoparticles can be avoided and it will result in a smaller particle size. As already observed with native CDs, RAME-β-CD and HP-β-CD are able to reduce Au3<sup>+</sup> thanks to their hydroxyl groups and then interact with the gold nanoparticles in order to prevent their agglomeration [31]. To the best of our knowledge, it is the first time that modified cyclodextrins are employed as both reducing agent of gold precursor and also stabilizing agent of gold nanoparticles.
Our materials were then characterized by UV-visible diffuse reflectance spectroscopy experiment. Figure 5 shows UV–Visible absorbance spectra and Tauc plots of TiO2-control, TiO2@Au, TiO2@Au-RB and TiO2@Au-HP materials. All the titania samples exhibit a broad absorption band around 330 nm corresponding to the charge transfer from O 2p valence band to Ti 3d conduction band [32]. Thus, the large band gap energy (Eg) of 3.20 eV estimated for the unmodified TiO2 is in agreement with typical values reported in the literature for anatase structures. However, it is worth noting that, for the gold-doped TiO2 samples prepared from cyclodextrins, a slight red-shift of the absorption edge of the TiO2 semiconductor toward higher wavelengths was observed compared to pure TiO2. The following sequence can be established in terms of Eg: TiO2@Au-RB (2.70 eV) < TiO2@Au-HP (2.95 eV) < TiO2@Au (3.20 eV) = TiO2 (3.20 eV).
As previously reported, the electrons can be transferred from the excited TiO2 to the metallic nanoparticles and the electron accumulation increases the Fermi level of the nanoparticle to more negative potentials. Therefore, the involved edge energy in the electron transfer from TiO2 to the metallic nanoparticles is lower than bare TiO2 [33]. The lowest band gap values for the TiO2@Au-RB and the TiO2@Au-HP materials suggest that the contact between the two inorganic phases (gold and TiO2) is enhanced when cyclodextrin is used during the Au NPs synthesis and this result is in good agreement with the TEM observations. However, the smallest value was found for the TiO2@Au-RB so that we can suppose that the use of the RAME-β-CD promotes the most intimate contact between the semiconductor and the metal. Further, another band is revealed at approximately 550 nm, confirming the presence of gold particles embedded in the TiO2 matrix [34]. When neither cyclodextrin nor ethanol is added to the gold salt in the second step, no reduction of Au3<sup>+</sup> was noticed, the resulting powder remained white and its UV-Vis spectra was similar to that obtained for the bare TiO2 (see Figure S4 in ESI).
**Figure 5.** Diffuse Reflectance UV-Vis (DRUV-Vis) spectra of titania-based materials prepared by microwave heating: (**a**) bare TiO2, (**b**) TiO2@Au, (**c**) TiO2@Au-HP, (**c**) and (**d**) TiO2@Au-RB. In the inset, Tauc plots for the determination of the band gap values Tauc (indirect bad gap energy).
UV-Vis experiment and TEM images proved that modified CDs can act as both reducing agent of the metal precursor and capping agent of well-dispersed homogeneously dispersed Au NPs even in the presence of titanium dioxide. But to further characterize our materials and specially to know if cyclodextrins still remained in the TiO2@Au-RB and the TiO2@Au-HP samples, thermogravimetric analyses (TGA) were performed. The thermal profiles of TiO2@Au, TiO2@Au-RB, and TiO2@Au-HP are shown in Figure 6.
The thermal patterns of the bare TiO2 and the TiO2@Au exhibit a one-step decomposition process with a weight loss in the 50–400 ◦C temperature range corresponding to the desorption of physically adsorbed water. The total weight loss for these samples are estimated to be 6.0 and 6.7 wt.%, respectively. The thermal profile of TiO2@Au-RB exhibits a two-step decomposition process with a total weight loss of ca. 10.4 % at 1000 ◦C. The first weight loss (≈4%) in the 50–250 ◦C temperature range corresponds to the removal of physically adsorbed water, whereas the second weight loss (≈6%) in the 250–450 ◦C temperature range with a major weight loss at ca. 380 ◦C attributed to the thermal decomposition of the modified β-CD (Figure S5 in ESI). A similar profile was obtained with the TiO2@Au-HP (Figure 6c) since this sample exhibited also a two-step decomposition process attributed to the removal of physically adsorbed water (≈6%) and to the thermal decomposition of residual HP-β-CD or its residues (≈11%) (see Figure S5 in ESI for the thermal profile of HP-β-CD alone). These thermal analyses proved that a small amount of saccharidic compounds (≈6 wt.% for the TiO2@Au-RB and 11 wt.% for the TiO2@Au-HP) remains adsorbed on our composite materials prepared with modified CDs even after the washing cycles. This result could be explained by the ability of CD derivatives to interact both with the gold nanoparticles and with the titania support. As previously described, we can suppose that after the
microwave reduction, cyclodextrin derivatives could be linked to the gold nanoparticles through weak interactions and covered the outer surface of the Au NPs [16,26]. In addition, cyclodextrins are known to be able to interact with the titanium dioxide through hydrogen bounds [35,36]. In fact, the hydroxyl groups located at the exterior of the torus favored the interactions of the cyclic oligosaccharides with the surface OH groups of titania. This latter hypothesis could also explain why the amount of organic compounds is higher in the TiO2@Au-HP than in the TiO2@Au-RB composite: the quantity of saccharidic compounds adsorbed on titania increases with the number of hydroxyl groups of the CD [37]. Because of a higher number of hydroxyl groups (21 vs. 8.4), the HP-β-CD is more adsorbed on the titania support than the RAME-β-CD. This hypothesis could also explain the larger Eg observed for the TiO2@Au-HP compared to the TiO2@Au-RB composite: the residual organic compounds may reduce the contact between the Au NPs and the titanium dioxide [38].
**Figure 6.** Thermogravimetric profiles for (**a**) bare TiO2, (**b**) TiO2@Au, (**c**) TiO2@Au-RB, and (**d**) TiO2@Au-HP.
According to the textural and structural studies, our titania-based materials exhibited interesting characteristics for photocatalytic applications. Indeed, the catalytic efficiency is known to be linked to two major physical properties: crystallinity and surface area of the photocatalysts [39].
With this microwave synthesis, only the anatase crystalline phase was obtained at low temperature (150 ◦C) without any additional calcination (or another thermal treatment) and this phase is known for its good activity in photocatalysis. On the other hand, good textural properties in terms of specific surface area, pore size, and pore volume could facilitate adsorption and diffusion of the target molecules onto the surface of the catalyst [40].
To confirm these hypotheses, the photocatalytic performances of the microwave gold-doped TiO2 materials have been investigated through two different experiments. The redox properties of these materials have been firstly evaluated in the photodegradation of methyl orange (MO) in water. Briefly, an aqueous solution of MO (50 ppm) in the presence of the semi-conductors was irradiated at 365 nm and the concentration of the residual dye was regularly quantified by HPLC measurements. Prior to the photocatalytic study, the photostability of the organic dye was checked in a preliminary test without photocatalyst (Figure S6), and it was found that the concentration of MO remained unchanged during
the 1h test period. The performances of TiO2@Au-RB and TiO2@Au-HP are reported in the Figure 7. For comparison, TiO2@Au prepared with ethanol was also tested (Figure 7a).
**Figure 7.** Photocatalytic performances of the gold-doped titania materials prepared by microwave heating for the degradation of methyl orange in near UV (λ = 365 nm): (**a**) Conversion of methyl orange after 60 min of irradiation (**b**) Evolution of the methyl orange concentration during one hour of irradiation for TiO2@Au-RB (yellow) and TiO2@Au-HP (green).
After 60 min under near UV irradiation, the dye was hardly degraded in the presence of the TiO2@Au prepared without CD by microwave heating and this result is similar to thus obtained with bare TiO2 (Figure S6). In contrast, after one hour of irradiation, the MO concentration was close to zero for the tests realized with the gold-doped TiO2 prepared with modified cyclodextrins (TiO2@Au-RB and TiO2@Au-HP). The addition of modified CD during the synthesis of Au NPs in the presence of TiO2 improved drastically the performances of the photocatalyst and this result is probably linked to the good dispersion of nanosized gold nanoparticles obtained from CDs over the support. In fact, small and well-dispersed metal islands deposited on the TiO2 core are known to provide a favorable geometry for facilitating the interfacial charge transfer under UV irradiation [41]: the electrons of the titanium oxide are excited from the valence band to the conduction band and then migrate to Au clusters, which prevent the direct recombination of electrons and holes. For the TiO2@Au-RB and TiO2@Au-HP samples, we can suppose that the small and spherical Au NPs observed on the surface of the semi-conductor by TEM experiments act as electron sink to favor the oxidation and the reduction reactions. Conversely, large particles of metal are often harmful to the photocatalytic activity so that the TiO2@Au prepared with ethanol as reducing agent was less efficient in our conditions [42]. Logically, large nanoparticles mobilize more gold atoms than small ones. With an equal metal loading, materials doped with large Au NPs offer fewer electronic reservoirs than those with small particles.
Additionally, the Figure 7b showed that the decrease of the MO amount was significantly faster in the presence of the microwave-assisted gold-doped TiO2 prepared with RAME-β-CD compared to that prepared with HP-β-CD (Figure 7b). The lowest efficiency of the TiO2@Au-HP compared to the TiO2@Au-RB might be correlated to the highest band gap (as evidenced by DRUV-Vis experiment) and also to the amount of CDs residues in the final material (as evidenced by thermogravimetric analysis). In fact, we can suppose that the residual organic compounds decrease the contact between the semi-conductor and the gold and so reduce the electron transfer. Furthermore, the CDs residues could maybe mask some of the active sites of the semi-conductor or reduce the potential adsorption of the MO [43].
The recyclability and reuse of the most efficient photocatalyst (TiO2@Au-RB) was also evaluated in the degradation of the MO. From Figure S7, it can be seen that the photocatalytic activity is stable during at least 3 runs. This study clearly showed the robustness of the catalyst and the strong embedment of the Au NPs onto the TiO2 support.
Finally, we studied the behavior of the gold-doped TiO2 in the production of hydrogen by photoreduction of water. Aqueous suspensions of the TiO2@Au, TiO2@Au-HP, and TiO2@Au-RB were
irradiated at 365 nm in the presence of ethanol as the sacrificial agent. The result of the amount of hydrogen produced by photoreduction of water is reported in the Figure 8a.
When the TiO2@Au-HP and the TiO2@Au-RB were irradiated in water, hydrogen was quickly detected and the amount of H2 was quantified as about 160 and 300 μmol.h−1.g−<sup>1</sup> of catalyst, respectively. Compared to other TiO2 catalysts in the literature [2,44], these amounts of produced hydrogen are promising since the power of our lamp is very low in comparison to Xe lamp usually used in such photocatalytic experiments. Moreover, the yield of hydrogen produced with our gold-doped catalyst was very high in comparison with that obtained with commercial anatase TiO2 (<2 μmol.h<sup>−</sup>1.g−1) or with TiO2@Au prepared with ethanol as reducing agent in the same conditions (about 3 μmol.h<sup>−</sup>1.g−1). As observed with the first photocatalytic test, the TiO2@Au-RB was also more efficient than theTiO2@Au-HP to produce hydrogen from water, probably due to the same reasons discussed above (i.e., twice as many organic compounds on the surface of the photocatalyst for the TiO2@Au-HP than for the TiO2@Au-RB). Finally, the amount of hydrogen produced is reproducible after several cycles of illumination (see for example Figure 8b with TiO2@Au-RB) and stable during more than 10 h (ESI, Figure S8). This catalytic result proved that the introduction of small and uniform gold nanoparticles thanks to CDs reduction leads to a real boost of the photocatalytic performances of titanium dioxide even under UV irradiation and clearly confirmed the need of intimate contact with TiO2 and Au to enhance the electron transfer between them.
**Figure 8.** (**a**) Amount of hydrogen produced by photoreduction of water in the presence of gold-doped TiO2 prepared by microwave heating process (100 mg of photocatalyst, 80 mL water, 20 mL ethanol, λ = 365 nm) (**b**) Evolution of the hydrogen production by photoreduction of water in the presence of TiO2@Au-RB during 3 cycles of illumination.
#### **3. Materials and Methods**
#### *3.1. Chemicals*
Randomly methylated β-cyclodextrin (RAME-β-CD) with an average degree of substitution of 1.8 methyl groups per glucopyranose unit (MW 1310 g.mol−1) was a gift from Wacker Chemie GmbH (Lyon, France). Hydroxypropyl-β-cyclodextrin (HP-β-CD) with an average substitution of 0.6 CH2CH(OH)-CH3 groups per glucopyranose unit (MW 1380 g.mol<sup>−</sup>1) was purchased from Roquette (Lestrem, France). Ethanol, methyl orange (MO) and TiCl4 were purchased from Sigma-Aldrich (Quentin-Fallavier, France) while HAuCl4 (49 wt.%) was provided by Strem Chemicals (Bischheim, France). All these reagents were used without purification.
#### *3.2. Preparation of the Au*/*TiO2 Materials with Cyclodextrins*
In a typical preparation, TiO2 was prepared from TiCl4 by microwave heating (CEM Mars instrument, Power 600 W) inspired by a method previously described by Wang et al. [45]. TiCl4 (0.9 mL, 8.21 mmol) was quickly added to ethanol (25 mL) and stirred at room temperature during 10 min. Then, the yellow solution was introduced in a Teflon microwave reactor equipped with temperature and pressure probes and heated to 150 ◦C during 10 min. The white suspension was centrifuged at 3000 rpm during 5 min. The supernatant was evacuated and the resulting white powder of TiO2 was added to 20 mL of an aqueous solution of HAuCl4 (3.73 <sup>×</sup> 10−<sup>5</sup> mmol) and cyclodextrin (4.04 <sup>×</sup> <sup>10</sup>−<sup>4</sup> mmol). Note that this CD/Au molar ratio of about 10 has been chosen to promote the synthesis of spherical gold nanoparticles, in line with a previous work reported in the literature [46]. NaOH (0.5 M) was slowly added to the solid suspension in order to adjust the pH value at about 9. Then the mixture was transferred in a Teflon microwave reactor and was finally heated under microwave irradiation with the same program used to prepare TiO2 from TiCl4 (600 W, 150 ◦C, 10 min). To promote the synthesis of gold nanoparticles. At the end of the heating microwave program, the suspension was centrifuged and the purple powder was thoroughly washed with water before overnight drying at 100 ◦C. The gold-doped TiO2 materials synthesized by microwave heating from RAME-β-CD and HP-β-CD were named as TiO2@Au-RB and TiO2@Au-HP, respectively. Additionally, note that a control gold-doped TiO2 (denoted as TiO2@Au) was also prepared in a very similar manner as the above described procedure, by substituting cyclodextrin for ethanol during the reduction process. The syntheses and characterizations have been reproduced several times.
#### *3.3. Characterization Methods*
#### 3.3.1. Powder X-ray Diffraction
Powder X-ray diffraction data were collected on a Siemens D5000 X-ray diffractometer (Bruker, Palaiseau, France) in a Bragg-Brentano configuration with a Cu Kα radiation source. Scans were run over the angular domains 10◦ < 2θ < 80◦ with a step size of 0.02◦ and a counting time of 2 s/step. Crystalline phases were identified by comparing the experimental diffraction patterns to Joint Committee on Powder Diffraction Standards (JCPDS) files for anatase. The treatment of the diffractograms was performed using the FullProf software [47] and its graphical interface WinPlotr [48]. The average crystallite size D was calculated from the Scherrer formula, D = Kλ/(β cos θ), where K is the shape factor (a value of 0.9 was used in this study, considering that the particles are spherical), λ is the X-ray radiation wavelength (1.54056 Å for Cu K), β is the full width at half-maximum (fwhm), and θ is the Bragg angle.
#### 3.3.2. Nitrogen Adsorption-Desorption Isotherms
Nitrogen adsorption-desorption isotherms were collected at −196 ◦C using an adsorption analyzer Micromeritics Tristar 3020 (Merignac, France). Prior to analysis, 200–400 mg samples were outgassed at 100 ◦C overnight to remove the species adsorbed on the surface. From N2 sorption isotherms, specific surface areas were calculated by the BET method while pore size distributions were determined using the BJH model assuming a cylindrical pore structure. The relative errors were estimated to be the following: SBET, 5%; pore volume (pv) (BJH), 5%; pore size (ps) (BJH), 20%.
#### 3.3.3. Diffuse Reflectance UV-Visible
Diffuse reflectance UV-visible spectra were collected using a Shimadzu UV-Vis NIR spectrometer (Marne-la-Vallée, France). BaSO4 was used as the reference. Tauc plot analysis was performed for the calculation of the band gap energy (Eg). In fact, the Eg can be estimated by plotting (F(R) hν) <sup>n</sup> vs. hν and extrapolated from linear part of the curve to the hν x-axis intercept. To determine values of these forbidden energies, the absorption data were fitted to the Tauc relation for indirect band-gap transitions (*n* = <sup>1</sup> <sup>2</sup> ) [49].
#### 3.3.4. Thermogravimetric Analysis (TGA) Coupled with Differential Scanning Calorimetry (DSC)
Thermogravimetric Analysis (TGA) coupled with Differential Scanning Calorimetry (DSC) analyses were performed using a Mettler Toledo TGA/DSC3+ STARe system unit (Viroflay, France). The samples were placed in aluminum oxide crucibles of 70 μL and heated from 40 to 1000 ◦C at 10 ◦C.min−<sup>1</sup> under a 50 mL.min−<sup>1</sup> air flow.
#### 3.3.5. ICP Optical Emission Spectrometry
ICP optical emission spectrometry was performed on an iCAP 7000 Thermo Scientific spectrometer (Les Ulis, France). For the quantification of gold loading, 10 mg of the Au/TiO2 materials were introduced in 20 mL of aqua regia and then heated to 130 ◦C during one hour. Then the remaining TiO2 was removed using a 0.2 μm pore filter. The resulting solution is finally diluted with pure water up to a final volume of 100 mL. The amount of gold incorporated in the material was determined using an external calibration with a gold ICP standard solution.
#### 3.3.6. Transmission Electron Microscopy (TEM)
Transmission Electron Microscopy (TEM) bright field observations were performed on a Tecnai G2 microscope (FEI, Hillsboro, Oregon, USA) operating at an accelerating voltage of 200 kV. The Au/TiO2 powder was deposited directly on a carbon coated copper grid. Metal particle size distributions have been determined from the measurement of about 200 Au NPs. The nanoparticles were found in arbitrarily chosen area of the images using the program ImageJ software.
#### *3.4. Photocatalytic Experiments*
#### 3.4.1. Photodegradation of Methyl Orange
The photocatalytic efficiency of the titania-based materials was first evaluated in the photodegradation of methyl orange (MO) carried out using quartz reactors of 5 mL. In a typical experiment, 10 mg of photocatalyst was added to 4 mL of a solution of methyl orange (50 ppm). After 30 min in the dark, UV irradiation was performed using a led UV light lamp (Opsytec λ = 365 nm, beam size = 0.785 cm2, power of 0.2 W.cm−2). Aliquots were centrifuged at regular intervals and the MO concentration in the supernatant was determined by high-performance chromatography (HPLC, PerkinElmer, Villebon-sur-Yvette, France) analyses using a PerkinElmer Pecosphere C18 (83 mm length × 4.6 mm diameter) column. An aqueous mixture of acetonitrile (20% (*v*/*v*)) was used as the mobile phase at a flow rate of 1 mL.min−1. Aliquots of 50 μL of the sample was injected and analyzed using a photodiode array detector. The MO conversion given in percentage refers to the difference in the MO concentration before irradiation (C0) and after 1 h of irradiation (C) divided by the MO concentration before irradiation (i.e., 100 × (C0 − C)/C0).
#### 3.4.2. Production of Hydrogen by Photoreduction of Water
Photocatalytic measurements for H2 generation were carried out in a cylindrical pyrex reactor equipped with a quartz window by irradiating the titania-based materials in a 20 vol% ethanol-water solution (ethanol was used as hole-scavenger). As light source, we used the same LED UV light as that employed for the photocatalytic degradation of MO experiments described above in Section 3.4.1. The reactor operated at room temperature and atmospheric pressure and was kept under stirring at a constant speed of 1250 rpm. In a typical experiment, 100 mg of photocatalyst was added to a 100 mL of ethanol-water solution in the reactor. The catalytic solid suspension was then flushed with argon gas (420 mL.h<sup>−</sup>1) for 60 min prior to photocatalysis. The amount of H2 produced was measured on-line using a micro gas chromatograph (Micro-GC Agilent 490, Les Ulis, France) equipped with a thermal conductivity detector and two separating columns (Microsieve 10 m (5 Å) and 8 m-Paraplot U) operating with backflush injection (Ar as carrier gas).
#### **4. Conclusions**
In this work, an easy and fast preparation of Au loaded TiO2 without calcination step is described. The addition of common modified cyclodextrin (methylated or hydroxypropylated) during the
microwave reduction of a gold precursor in the presence of TiO2 led to an efficient photocatalyst both for pollutant photodegradation and photoreduction of water under near UV irradiation. The saccharidic macrocycle was responsible for a good stabilization of gold nanoparticles in aqueous solution so that these latter could not aggregate during the microwave synthesis and were deposited uniformly on the TiO2 surface. Because of its lowest number of hydroxyl groups, the RAME-β-CD seems to be less adsorbed onto the surface of the final composite after the gold reduction and represents the most promising photocatalyst. It could be now interesting to study the photocatalytic performances of our materials under solar simulated lamp. However, this new and fast synthetic approach offers promising perspectives for photocatalytic depollution process and green energy production.
**Supplementary Materials:** The following are available online at http://www.mdpi.com/2073-4344/10/7/801/s1, Figure S1. N2 adsorption desorption isotherms of TiO2-control (a) gold decorated titania materials prepared without CD (TiO2@Au) (b) gold decorated titania materials prepared with HP-β-CD (TiO2@Au-HP) (c) gold decorated titania materials prepared with RAME-β-CD (TiO2@Au-RB) (d), Figure S2. TEM images of TiO2@Au catalyst at magnification of ×62,000, Figure S3. TEM images of (a) TiO2@Au-RBand (b) TiO2@Au-HB at magnification of ×490,000, Figure S4. UV-vis spectra of titania materials prepared by a two-step microwave heating procedure with HAuCl4 in a second step but without CD and without ethanol, Figure S5. TGA profiles for the RAME-β-CD and the HP-β-CD, Figure S6. Evolution of methyl orange concentration under irradiation (λ = 365 nm) as a function of time in the absence (open circle) or presence of the bare TiO2 prepared by microwave process (filled circle). Reaction conditions: TiO2, m = 10 mg; methyl orange solution, V = 4 mL (50 ppm) Figure S7. Performance of TiO2@Au-RB in three consecutive tests with reuse of the catalyst. Reaction conditions: 4 mL of a solution of methylorange (50 ppm), 10 mg of TiO2@Au-RB (λ = 365 nm, t = 10 min), Figure S8. Production of hydrogen by photoreduction of water (80 mL) in the presence of TiO2@Au-RB (100 mg) and ethanol (20 mL) as sacrificial agent (λ = 365 nm).
**Author Contributions:** Synthesis, catalytic tests, ICP, and UV experiments, C.M.; N2 adsorption-desorption measurements and IR spectroscopy, N.K.; TEM analysis, B.L. and A.A.; XRD, A.P. and F.W.; catalysis, S.N.; supervision and reviewing results, reviewing the manuscript, English writing, C.M., E.M., and A.P. All authors have read and agreed to the published version of the manuscript.
**Funding:** This research received no external funding.
**Acknowledgments:** The TEM facility in Lille (France) is supported by the Conseil Régional du Nord Pas de Calais and the European Regional Development Fund (ERF). Chevreul Institute (FR 2638), Ministère de l'Enseignement Supérieur, de la Recherche et de l'Innovation, Région Hauts-de-France and FEDER are acknowledged for supporting and funding partially this work. The authors are grateful to the University of Artois for supporting this research through the Quality Research Bonus (Micro-GC Agilent 490 in 2018).
**Conflicts of Interest:** The authors declare no conflict of interest.
#### **References**
© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
### *Article* **Effect of Potential and Chlorides on Photoelectrochemical Removal of Diethyl Phthalate from Water**
**Laura Mais, Simonetta Palmas, Michele Mascia and Annalisa Vacca \***
Dipartimento di Ingegneria Meccanica, Chimica e dei Materiali, Università degli Studi di Cagliari, Via Marengo 2, 09123 Cagliari, Italy; [email protected] (L.M.); [email protected] (S.P.); [email protected] (M.M.)
**\*** Correspondence: [email protected]
**Abstract:** Removal of persistent pollutants from water by photoelectrocatalysis has emerged as a promising powerful process. Applied potential plays a key role in the photocatalytic activity of the semi-conductor as well as the possible presence of chloride ions in the solution. This work aims to investigate these effects on the photoelectrocatalytic oxidation of diethyl phthalate (DEP) by using TiO2 nanotubular anodes under solar light irradiation. PEC tests were performed at constant potentials under different concentration of NaCl. The process is able to remove DEP following a pseudo-first order kinetics: values of kapp of 1.25 <sup>×</sup> <sup>10</sup>−<sup>3</sup> min−<sup>1</sup> and 1.56 <sup>×</sup> <sup>10</sup>−<sup>4</sup> min−<sup>1</sup> have been obtained at applied potentials of 1.8 and 0.2 V, respectively. Results showed that, depending on the applied potential, the presence of chloride ions in the solution affects the degradation rate resulting in a negative effect: the presence of 500 mM of Cl− reduces the value of kapp by 50 and 80% at 0.2 and 1.8 V respectively.
**Keywords:** diethyl phthalate; photoelectrochemical degradation; persistent organic pollutants; chloride ions; TiO2 nanotubes
#### **1. Introduction**
The application of photoelectrochemical process for polluted waters and wastewaters has been gaining more and more attention thanks to the possibility to obtain electrical energy from renewable energy sources, rather than from fossil fuels [1]. The technique exploits the synergy between photochemistry and electrochemistry: from one side, the photochemical process increases its efficiency as the bias potential lowers recombination of the photogenerated charges, from the other side the photo-potential generated on the semiconductor depolarizes the cell improving the yield of the electrochemical process [2].
Considering the application to real matrices, the effect of the composition of the water to be treated plays a crucial role, with particular regard to the presence of chlorides, which are ubiquitous ions in water and wastewater. Several studies on the photochemical process using TiO2 highlighted a negative effect of the presence of chloride: the inhibiting effect has been ascribed both to the competitive adsorption between the pollutant molecules and Cl− towards the surface-active sites of TiO2, or to the scavenging function of chloride ions towards holes and hydroxyl radicals [3,4]. Piscopo et al. [5] showed different effects on the degradation rate of two pollutants depending on the chloride concentration, the nature of the organics and the pH: in the case of poorly adsorbed molecules, if the pH favored the adsorption of Cl−, even low concentration of chloride strongly affected the degradation.
Several papers evidenced the key role of pH in the photocatalytic degradation using TiO2: point of zero charge (pHpzc) plays a crucial role in determining the surface charge of photocatalyst and, in turn, its interaction with charged molecules or ions. When the pH is higher than the pHpzc, the polarity of TiO2 surface is negative and the electrostatic repulsion toward anionic compounds dominates [6–8]. Moreover, since hydroxyl radicals can be formed by the reaction between hydroxide ions and positive holes, the hydroxyl radicals are
**Citation:** Mais, L.; Palmas, S.; Mascia, M.; Vacca, A. Effect of Potential and Chlorides on Photoelectrochemical Removal of Diethyl Phthalate from Water. *Catalysts* **2021**, *11*, 882. https://doi.org/10.3390/catal11080882
Academic Editor: Bruno Fabre
Received: 17 June 2021 Accepted: 19 July 2021 Published: 22 July 2021
**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.
**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).
considered as the predominant species at neutral or high pH, while at low pH the holes are considered the major oxidizing species [9]. Regarding the scavenging effect, chloride can react with HO• radicals and holes, allowing the formation of less reactive chloride radical (Cl•) and dichloride radicals (Cl2•−) [10–12]: the oxidized chloride may also recombine with photogenerated electrons quenching the photogenerated charge carriers [13].
Different considerations may be made when photoelectrocatalysis is considered: in this case, heterogenous photocatalysis can be improved by the application of a bias potential to obtain a more effective separation of photogenerated charges, thereby increasing the lifetime of electron–hole pairs. In the photoelectrocatalytic process, the increases of the applied potential can accelerate the photogenerated electrons toward the external circuit, generating the bending of the conduction and valence bands, with the consequent formation of a space charge layer. Thus, the recombination of the e−/h<sup>+</sup> pairs may be decreased or totally prevented, improving the photocatalytic performance [14,15]. Moreover, increase in the potential can empty the defects where the photogenerated charges are trapped, enhancing the photoactivity [16].
The presence of chloride in a photoelectrochemical process exerts a different effect with respect to the photochemical one: in fact, unlike the inhibitory effects found in photocatalysis, in photoelectrochemical removal of pollutants, enhancing effect in the degradation process has been often highlighted. Zanoni et al. [17] reported the highest discoloration rate and TOC removal for solution containing Remazol Brilliant Orange 3R at pH 6.0 in presence of 0.5 M of NaCl applying +1.0 V (SCE) to the TiO2 photoanode. Also, in the case of other dyes or organics, the presence of Cl− has been found beneficial to accelerate the degradation rate [18,19]. The improvement in the degradation has been explained by the synergistic action of the strong oxidizing species HO•, chlorine-based radicals Cl• and Cl2•−, and active chlorine species like HClO and Cl2 that can give a bulk contribution [20,21]. Moreover, at the anode the adsorption of negative charged ions, such as chloride, can be enhanced both by the polarization and the promotion of reactions that can generate local acidic pH variation near the anodic surface.
In this framework, our work is devoted to study the photoelectrochemical degradation of a persistent organic pollutant at two levels of applied potentials and in the presence of different concentrations of chloride under simulated solar light conditions, using TiO2 nanotubular electrodes. The pollutant selected for the study is the diethylphthalate (DEP). Phthalate esters (PAEs) are a group of widely used plasticizers that can lead to endocrine system disorders, affecting reproductive function, and inducing some tumors [22–24]. Due to their wide utilization and the difficulty to completely remove them with conventional treatment processes, PAEs are ubiquitous persistent organic pollutants in the environment, being the short chain phthalate as DEP, the most detected in surface marine waters, freshwaters, and sediments [25–27]. To the best of our knowledge, only few papers reported on the photoelectrochemical degradation of the diethyl phthalate [28,29]. Moreover, the influence of the presence of chloride during their treatment and the effect of the applied potential are not yet presented by the literature.
#### **2. Materials and Methods**
#### *2.1. Preparation of TiO2 Nanotubes*
TiO2 nanotube electrode (TiO2-NT) used for the photoelectrochemical degradation of DEP was prepared by electrochemical anodization as reported in our previous work [30]. Briefly, Ti foils (0.25 mm thickness, 99.7% metal basis, Aldrich, St. Louis, MO, USA) were cut in circular disks of 5 cm diameter. After ultrasonic treatments in acetone, isopropanol and methanol (10 min each), Ti was rinsed with deionized water, and dried with a nitrogen stream. The anodization was performed in a two-electrode cylindrical cell made by Teflon (inner dimension: diameter = 4.4 cm and height = 5 cm). The working electrode was located at the bottom of the cell where the electrical contact was an aluminum disc. The exposed geometrical area of the Ti electrodes was 15 cm2. A platinum titanium grid placed in front of the anode at 1 cm distance constituted the counter electrode.
The anodization was performed in (10%) deionized water/(90%) glycerol solution with 0.14 M of NH4F at room temperature. A potential ramp was imposed from open circuit voltage (OCV) to 20 V with a scan rate of 100 mVs−1; then the applied potential was maintained at this fixed value for 4 h. TiO2-NT was annealed in air atmosphere at 400 ◦C for 1 h to transform the amorphous structure into crystalline one. The phase transformation depends on both the structure morphology and annealing temperature: it has been shown that the anatase-to-rutile transformation starts near 430 ◦C for the 500 nm long nanotubes [31], while the same transformation has been reported to occur at 550 ◦C for nanotubes up to 200 nm [32]. In our case, after 1 h at 400 ◦C, a unique anatase phase was present [33]. The morphological characterization of TiO2-NT was presented in [30]: the average diameter of tubes ranged between 40–50 nm, while the tube length of around 700 nm was measured.
#### *2.2. Photoelectrochemical Tests*
Photoelectrochemical tests were performed in a three-electrode beaker cell using TiO2 nanotubes as photoanode, a platinized titanium grid as cathode, and a saturated calomel electrode (SCE) as reference. The cell was filled with 100 mL of solution and connected with a potentiostat-galvanostat (Metrhom Autolab 302N, Metrohm, Herisau, Switzerland) controlled by Nova software. The photoanode was irradiated by UV-vis light using a 300 W xenon lamp equipped with air mass (AM) 0 and 1.5 D filters to simulate the solar irradiation.
Photocurrent measurements were carried out by linear sweep voltammetric (LSV) runs, starting from the OCV to 2.5 V at a scan rate of 10 mVs−1, with hand-chopped light. The photocurrent-time measurements were recorded applying a constant potential in the dark for 10 min; afterward, the electrode was exposed to light for 200 s, followed by dark condition.
Photoelectrochemical oxidation of diethyl phthalate was performed under potentiostatic conditions at 0.2 and 1.8 V vs. SCE. The initial concentration of the organic compound was 40 mg dm−<sup>3</sup> and 0.1 M NaClO4 was used as supporting electrolyte. Moreover, different amount of NaCl (1, 100, 500 mM) were added to the solution, to investigate on the effect of chloride concentration during the photoelectrochemical oxidation of DEP. The pH of the solution was neutral. During degradation experiments, samples of electrolyte were withdrawn for qualitative and quantitative analyses of the model organic compound.
#### *2.3. Analytical Methods*
Analyses of the model organic compound were carried out by HPLC (Waters), equipped with a column Varian C18 and a dual band UV detector set to 283 and 229 nm. The mobile phase was Acetonitrile and aqueous solution 0.1% H3PO4 = 40:60 with a flow rate of 1 mL min<sup>−</sup>1.
The oxidant concentration, expressed as μM of active chlorine, was measured using the N,N-diethyl-p-phenylenediamine (DPD) colorimetric method. DPD oxidizes to form a red-violet product, the concentration of which is determined measuring the absorbance at 515 nm.
The trend of mineralization was monitored by measuring the total organic carbon (TOC) by a Shimatzu TOC 500L instrument.
For each sample a repeatability within ±5% has been evaluated.
#### **3. Results and Discussion**
Figure 1 shows the trend of polarization curve performed at the TiO2-NT electrode during LSV in aqueous solution of DEP under irradiation and in the dark.
**Figure 1.** LSV of TiO2-NT performed at 10 mV s−<sup>1</sup> of scan rate, under dark and irradiation condition. Blue symbols indicate the potentials selected for the degradation runs.
A typical trend is observed, with an onset potential of −0.25 V, followed by an ohmic behavior of the system, in which the positive influence of the potential is strictly connected to the increase in the space charge depletion region of the semiconductor; in the central range of potential (0.7–1.8 V) the saturation of the current is reached, in which increase of the potential is no more effective in terms of a corresponding increasing of the current. In the final range, at potentials higher than the value of band gap of the semiconductor, the barrier breakdown effect could be responsible for the sharp rising in the photocurrent along with the dark current contribution [30].
The degradation tests have been performed selecting two applied potentials: the first one in the ohmic region and the second one in the saturation region. The two blue diamonds in Figure 1 indicate the values of potential selected.
Figure 2a shows the trend with time of the DEP concentration, normalized with respect to the initial concentration, during electrolysis at the two different potentials. For comparison, the trend with time of the DEP concentration at the open circuit potential in the dark was also reported in the same figure: no significant adsorption of DEP on the electrode surface was detected that can be explained considering the neutral pH of the solution, the iso-electrical point of TiO2 located around pH = 6, and the non-ionic nature of the molecule of DEP. When the runs were performed in potentiostatic conditions and under illumination, the concentration of DEP decreased, being the highest reaction rate achieved at 1.8 V.
**Figure 2.** (**a**) Trends with time of the concentration of DEP, normalized to the initial concentration C0, during runs performed with solutions containing 40 mg dm−<sup>3</sup> DEP in 0.1 M NaClO4 as supporting electrolyte at different applied potentials. (**b**) Fraction of reactant removed as a function of the specific charge supplied during the related runs.
However, since the mean current intensity measured during the potentiostatic runs was 0.1 mA at 0.2 V and 1.2 mA at 1.8 V, it could be useful to compare the trend of fraction of the removed reactant as a function of the specific supplied charge (Figure 2b): in this case, the highest yield of the removal process is measured at the lowest potential, indicating that most of the charge passed at 1.8 V has been used for the side reaction of water oxidation.
An analogous behavior was observed in our previous work, where the photo-electrocatalytic degradation of 2,4-dichlorophenoxyacetic acid was investigated: higher efficiency and slower kinetics of degradation were detected in the ohmic region of the polarization curve with respect to those in the saturation region [30].
Degradation curves of DEP at various chloride concentration at the two applied potentials are shown in Figure 3a,b as semilogarithmic plots. A linear trend of ln(C/C0) vs. time is observed under all the experimental conditions, indicating that a pseudo-first order kinetics could be used to interpret the data, as follows:
$$\text{dC/dt} = -\mathbf{k\_{app}} \,\mathrm{C} \tag{1}$$
**Figure 3.** Trends with time of lnC/C<sup>0</sup> during photoelectrochemical degradations using solutions containing 40 mg dm−<sup>3</sup> DEP, 0.1 M NaClO4, and different chloride concentration. (**a**) Applied potential: 0.2 V; (**b**) applied potential: 1.8 V.
The values of the apparent kinetic constant kapp, evaluated from the slope of each straight line at the relevant operative conditions are reported in Figure 4, as a function of the chloride concentration. As already observed in absence of chloride, the fastest kinetics of the reactant removal are obtained at 1.8 V for each level of chloride concentration. Moreover, at 0.2 V, the increase of chloride concentration scarcely affects the reaction rate, except for 500 mM of Cl−, which halves the kapp. At 1.8 V, the effect of chloride is more evident: at 1 mM of Cl− the kapp is reduced by 40% while at 500 mM of Cl− by 80%, in respect to the kapp evaluated without chloride.
Figure 5 shows the trend of the ratio between kapp evaluated at 1.8 V and that at 0.2 V measured at different chloride concentrations. In absence of chloride, an increment of one order of magnitude is obtained, while in presence of the highest concentration of chloride kapp increases of two-fold when the potential values change from 0.2 to 1.8 V. This behavior indicates that the higher the potential, the higher is the negative effect of the concentration of chloride.
**Figure 4.** Pseudo-first order kinetic constants of the reactant removal process performed in solutions of 40 mg dm−<sup>3</sup> of DEP, 0.1 M NaClO4, and different chloride concentration. (**a**) Applied potential: 0.2 V; (**b**) applied potential: 1.8 V.
**Figure 5.** Ratio between the apparent kinetic constant evaluated at 1.8 and 0.2 V for different chloride concentrations.
The inhibiting effect observed in presence of Cl− agrees with observations reported for photocatalytic processes at TiO2-based materials. Several mechanisms have been proposed to explain the inhibiting effect on the photocatalytic degradation [13]:
Moreover, due to the complexity of the processes, a combination of mechanisms is often claimed to explain the inhibiting effect [5,13,39–41].
In the case of a photo-electrochemical process, also the effect of the applied potential should be considered, as well as the pH modification due to the side reactions that occur to a greater or lesser extent depending on the applied potential.
In order to verify the effect of the concentration of chloride and the applied potential on the behavior of the semiconductor, photocurrent transients have been recorded applying different potential and varying the chloride concentration during chopped light chronoamperometries.
Figure 6 shows the results obtained without chloride. For TiO2 nanotubes, the thickness of the wall can be determinant for the extension of the space charge depletion layer; this in turn, can be relevant for the recombination phenomena, which are strictly connected to the applied potential. As can be seen, at the lowest potential, a typical spike of the anodic current is observed, followed by an exponential decrease of the photocurrent with time until a stationary value is reached. The positive spike is no more visible at the highest potential. According to the literature [42], the positive current transient when the light is turned on represents the accumulation of holes at the electrode/electrolyte interface without injection to the electrolyte. Since any fast faradic reaction is occurring, the charge recombination is responsible for the subsequent decrease of the measured current.
**Figure 6.** Potentiostatic tests performed with solution containing 0.1 M NaClO4 at 0.2 (**pink**) and 1.8 V (**black**).
At low potentials, when we operate in the ohmic region of the polarization curve, where the charge depletion layer thickness is not fully developed inside the nanotubes wall, the photogenerated holes may rapidly recombine in the regions of the material that do not experience beneficial space charge effects, i.e., that are non-depleted of the majority carriers (electrons). When the experiment is performed at the highest potential (in the saturation region of the polarization curve) the depletion layer extends in the whole wall of nanotubes and the recombination is suppressed.
Photocurrent transients in presence of chloride are reported in Figure 7.
**Figure 7.** Potentiostatic tests performed with solution containing 0.1 M NaClO4 and different concentration of chloride ions at 0.2 and 1.8 V.
At low applied potential, the rate of the photocurrent decreasing (i.e., the rate of the charge recombination process) is scarcely influenced by the presence of chloride, when they are present at low concentration levels: overlapped curves are obtained related to the runs performed at 0, 1, and 100 mM of Cl− ions. Only at 500 mM of Cl−, slower decay of the photocurrent can be observed, indicating an inhibition of the recombination processes. Moreover, when the light was turned off, negative current transients were observed for high chloride concentration. Negative spikes were often detected during photocurrent transient of semiconductors and can be related to slower electron/hole pairs recombination due to the presence of holes trapped in the surface [43].
These transients in photocurrent can be explained as follows: at lower chloride concentration, the charge recombination prevails since chloride is poorly adsorbed onto the semiconductor electrode, so it is not able to react with the photogenerated holes faster than the electrons. However, at the highest concentration of chloride, it is likely that the adsorption effect would predominate, so that chloride can act as hole scavenger, according to the following adsorption phenomena:
$$\rm TiO\_2\rm \rm \rm h^+ + Cl^- = TiO\_2\rm \rm Cl\_{ads} \tag{2}$$
This process promotes the separation of electron-hole pair limiting the charge recombination as suggested by other authors [20,44,45].
At the highest potential, the recombination is suppressed, and positive transient and negative spikes disappear also in presence of high chloride ions. Moreover, very small increment in the steady state photocurrent was observed, increasing the concentration of chloride. So, at 0.2 V, the highest variation in the value of kapp obtained at 500 mM of chloride, can be connected to the blocking effect of adsorbed Cl− and the competitive adsorption, with respect to water molecules, which reduces the formation of HO• radicals.
Similar considerations should be done also to explain the result at 1.8 V, but, as we noticed, the inhibiting effect at this potential is evident also at low concentration of chloride. This can be explained by considering two aspects connected to the applied potential: the electrode works in a region of potential where the oxygen evolution reaction occurs to a large extent, so that a local acidic pH near the surface can generate a positive charge (pH < isoelectric point). Moreover, the application of high anodic potentials can generate a build-up of a positive surface charge. In this condition, the competitive adsorption or
blocking of active surface sites by chloride anions will be favored due to the electrostatic attraction of Cl−, also at low concentration of chloride.
The adsorbed chloride can react to form chlorine by the following reaction [17,44]:
$$\text{TiO}\_2\text{-Cl}\_{\text{ads}} + \text{Cl}^- \rightarrow \text{Cl}\_2 + \text{TiO}\_2 + \text{e}^- \tag{3}$$
Dissolved chlorine reacts with water to give hypochlorous acid and hypochlorite ions (Equations (3) and (4)), being the distribution of the three forms of active chlorine dependent on pH:
$$\text{Cl}\_2 + \text{H}\_2\text{O} \rightarrow \text{Cl}^- + \text{HClO} + \text{H}^+ \tag{4}$$
$$\text{HClO} \leftrightarrow = \text{H}^+ + \text{ClO}^-\tag{5}$$
Chlorine-based oxidants (active chlorine) have been detected during the photoelectrochemical degradation of DEP in different operating conditions.
At 0.2 V after 130 C dm−<sup>3</sup> of supplied charge, 2.0 and 4.2 μM of active chlorine concentrations were detected at 100 and 500 mM of Cl−, respectively. These small amounts agree with the poor adsorption of chloride at this value of applied potential. At 1.8 V, higher concentration of active chlorine was detected. As an example, the trend with time of the concentration of active chlorine obtained during DEP degradation in presence of 100 mM of Cl− is reported in Figure 8. The higher amount of active chlorine confirms a better reactivity of chloride with the positively charged surface of TiO2 at 1.8 V.
**Figure 8.** Trends with time of the concentration of active chlorine produced during a degradation run at 1.8 V with solution containing 40 mg dm−<sup>3</sup> of DEP, 0.1 M NaClO4, and 100 mM of Cl−.
The formation of chlorine-based oxidants during photoelectrochemical treatment of water containing chloride has been studied by several authors: some of them found that the presence of Cl− suppressed the degradation rate of organic pollutants, while others found opposed result [44].
The positive effect was generally observed when the active chlorine was able to give a bulk contribution to the reaction, i.e., in the cases where the organic pollutants can be oxidized also by active chlorine. For example, during photo-electrochemical discoloration of solutions containing Methylene Blue, low pH, and high concentration of Cl− were beneficial [18]. Also, Zanoni et al. [17] found the highest TOC removal for solution containing Remazol Brilliant Orange 3R, working at pH 6.0, 1.0 M NaCl, when the photoelectrode was biased at +1 V (versus SCE).
In our case, the formation of active chlorine seems not sufficient to contribute to the overall reaction rate at such an extent to make up for the negative effect.
Some specific tests were performed to evaluate the effectiveness of the photo-electrogenerated active chlorine on the DEP degradation. To this aim, during photoelectrocatalytic degradation runs, the light was turned off and the application of bias potential was stopped. In
this condition, in the solution, 25 μM of active chlorine accumulated, and residual 26 mg dm−<sup>3</sup> of DEP were present: the solution was monitored by following the concentration of the residual DEP with time. Negligible variation in the concentration of DEP was found after two hours indicating that the HO• radicals may be considered as the main factor responsible for the degradation, while active chlorine seems to give a not significant contribution to the overall oxidation rate. Similar behavior was found during the electrochemical degradation of the dimethyl phthalate ester on a fluoride-doped Ti/β-PbO2 anode: the lower removal of the pollutant in the presence of chloride ions was explained considering the lower reactivity of dimethyl phthalate with chlorine radical species in respect to hydroxyl radicals. Also, the active chlorine can react with HO• radicals thus reducing their availability for organic oxidation [46].
The low reactivity of active chlorine towards DEP obtained in our experimental conditions may indicate that the formation of harmful chlorinated intermediates is unlikely, even if the possible reaction of DEP intermediates with active chlorine during the runs cannot be excluded. Table 1 reports the ratio (*ϕ*) between the removal percentages of TOC and DEP evaluated at the end of each run, which indicates the level of total mineralization as defined by the following equation [47]:
$$\varphi = \frac{\% \text{[TOC]}\_{\text{removal}}}{\% \text{[}DEP\text{]}\_{\text{removal}}} \tag{6}$$
**Table 1.** TOC removal and ϕ evaluated at the end of each run.
At 1.8 V, a higher degree of mineralization was evaluated at the end of the runs, in which *ϕ* approached the unity. However, at 0.2 V, the high values of *ϕ* also indicate that the possible intermediates are almost completely removed.
#### **4. Conclusions**
In this work, the photoelectrochemical degradation of diethyl phthalate has been studied at two levels of applied potentials and in the presence of different concentrations of chloride under simulated solar light conditions, using TiO2 nanotubular electrodes. The process is able to remove DEP following a pseudo-first order kinetics: values of kapp of 1.25 × <sup>10</sup>−<sup>3</sup> min−<sup>1</sup> and 1.56 × <sup>10</sup>−<sup>4</sup> min−<sup>1</sup> were obtained at applied potentials of 1.8 and 0.2 V, respectively. Higher current efficiency and slower kinetics of degradation were detected in the ohmic region of the polarization curve at 0.2 V. The presence of chloride ions in the solution affects the degradation rate to different extents depending of the applied potential: the higher the potential, the higher the negative effect of the increase of chloride concentration. The presence of 500 mM of Cl− halves the kapp at 0.2 V, while at 1.8 V its value decreases to 1.56 × <sup>10</sup>−<sup>4</sup> min<sup>−</sup>1. This behavior can be connected to the blocking effect of adsorbed Cl− and the competitive adsorption, with respect to water molecules, which reduces the formation of HO• radicals: at 0.2 V, the adsorption of chloride predominates only at the highest concentration of chloride, while at 1.8 V, the positive surface charge due to the applied potential and the possible acidification of the anodic layer allow the adsorption also at low chloride concentrations.
**Author Contributions:** Conceptualization, A.V. and S.P.; methodology, M.M.; validation, L.M., A.V. and S.P.; formal analysis, L.M.; investigation, L.M.; writing—original draft preparation, A.V. and S.P.; writing—review and editing, S.P., L.M. and M.M. All authors have read and agreed to the published version of the manuscript.
**Funding:** This paper is part of the research project funded by P.O.R. SARDEGNA F.S.E. 2014–2020— Axis III Education and Training, Thematic Goal 10, Specific goal 10.5, Action partnership agreement 10.5.12—"Call for funding of research projects—Year 2017".
**Data Availability Statement:** Data is contained within the article.
**Conflicts of Interest:** The authors declare no conflict of interest.
#### **References**
### *Article* **Analysis of Photocatalytic Degradation of Phenol with Exfoliated Graphitic Carbon Nitride and Light-Emitting Diodes Using Response Surface Methodology**
**Adeem Ghaffar Rana 1,2 and Mirjana Minceva 1,\***
**Abstract:** Response surface methodology (RSM) involving a Box–Benkhen design (BBD) was employed to analyze the photocatalytic degradation of phenol using exfoliated graphitic carbon nitride (g-C3N4) and light-emitting diodes (wavelength = 430 nm). The interaction between three parameters, namely, catalyst concentration (0.25–0.75 g/L), pollutant concentration (20–100 ppm), and pH of the solution (3–10), was examined and modeled. An empirical regression quadratic model was developed to relate the phenol degradation efficiency with these three parameters. Analysis of variance (ANOVA) was then applied to examine the significance of the model; this showed that the model is significant with an insignificant lack of fit and an R<sup>2</sup> of 0.96. The statistical analysis demonstrated that, in the studied range, phenol concentration considerably affected phenol degradation. The RSM model shows a significant correlation between predicted and experimental values of photocatalytic degradation of phenol. The model's accuracy was tested for 50 ppm of phenol under optimal conditions involving a catalyst concentration of 0.4 g/L catalysts and a solution pH of 6.5. The model predicted a degradation efficiency of 88.62%, whereas the experimentally achieved efficiency was 83.75%.
**Keywords:** g-C3N4; photocatalysis; response surface methodology; wastewater treatment; phenol
#### **1. Introduction**
For all living beings, water is considered to be the most important resource. Easy access to clean water is one of the biggest challenges for mankind. In the last few decades, advancements in science, technology, and industrialization have led to considerable benefits to mankind but at the cost of a more polluted environment, particularly water [1]. There are multiple categories of pollutants in water, such as heavy metals, dyes, pesticides, pharmaceuticals, and other organic pollutants. Amongst organic pollutants, phenolic compounds, with ~3 million tons of global production, are an emerging contaminant detected in water [1–4].
Phenols or phenolics are essential because of their wide range of applications in the processing and manufacturing industry. However, the ecosystem's contamination by phenolics is concerning because of the adverse implications on human health such as their endocrine-disrupting abilities and carcinogenic behavior [1,5,6]. Moreover, these chemicals cause environmental issues such as water hardness, pH change, and a decrease in dissolved oxygen level. Furthermore, the Environmental Protection Agency (EPA) and the European Union (EU) have included a few phenols in their priority pollutants list. It is necessary to make this polluted water containing phenols and other pollutants suitable for human use and aquatic life using certain techniques to minimize the usage of these chemicals [5].
The removal of phenolic compounds from wastewater has attracted considerable attention from researchers [5]. Many biological, chemical, and physical techniques such
**Citation:** Rana, A.G.; Minceva, M. Analysis of Photocatalytic Degradation of Phenol with Exfoliated Graphitic Carbon Nitride and Light-Emitting Diodes Using Response Surface Methodology. *Catalysts* **2021**, *11*, 898. https:// doi.org/10.3390/catal11080898
Academic Editor: Annalisa Vacca
Received: 2 July 2021 Accepted: 23 July 2021 Published: 25 July 2021
**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.
**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).
as membrane filtration, coagulation–flocculation, adsorption [7,8], ion exchange, bacterial and fungal biosorption [9], aerobic and anaerobic processes [10] are used for phenol removal. In these processes, there are many constraints such as high cost, and low efficiency; furthermore, these methods do not completely remove phenol from wastewater [11,12]. Moreover, using these techniques, phenol is transferred from wastewater to a solid phase that requires treatment for safe disposal, which leads to additional cost for the whole process. Thus, it is necessary to develop an alternative effective and cost-efficient method for phenol removal from wastewater.
Advanced oxidative processes (AOP) are successful for achieving the complete removal of pollutants [13]. The degradation process using AOP can be performed in several ways, such as using only oxidizing agents, light irradiance in addition with oxidizing agents, and photocatalysis [14]. For all these processes, the degradation process is conducted using OH− radicals that are generated during the oxidation reaction. Among these processes, photocatalysis has attracted considerable interest because it can harvest solar light with the help of semiconductor materials (catalysts). The catalysts can help solve environmental issues related to water contaminations; these semiconductor materials are nontoxic and efficient. Note that different semiconductor materials such as ZnO [15], TiO2 [16], SiO2, Al2O3 [8], and g-C3N4 [17,18], are used for environmental applications in photocatalysis; these have considerable advantages because of the large surface areas, adsorption capacities, and better absorption of light. Among these materials, g-C3N4 offers improved visible light absorption [17,19–21].
g-C3N4, a polymeric semiconductor, composed of C, N, and H, has gained considerable interest from researchers for novel generation of photocatalysts because of its widespread catalytic uses in oxidation and reduction processes, such as pollutant degradation, water splitting, and CO2 reduction. These materials have been extensively used for environmental remediation because they are easy to synthesize, metal-free, inexpensive, and easily available [22–24]. Furthermore, g-C3N4 possesses higher thermal and chemical stability because of π-conjugated frameworks connecting the 2D layered structure of tri-striazine building blocks. g-C3N4 can be activated by visible light of 420–460 nm because of its low bandgap energy (2.7 eV) [25,26]. There are, however, certain challenges associated with the application of g-C3N4 in phenol removal such as low surface area, fast recombination rate, and low conductivity, thus resulting in lower efficiency. To overcome these limitations, multiple strategies have been used to improve the surface electronic structures and activity of the bulk g-C3N4 in visible light. To improve the activity of pristine g-C3N4, strategies such as metal and non-metal doping, exfoliation, hard and soft templating, and metal oxide heterojunctions have been used [27–31].
Factors affecting the removal efficiency can be tuned by the morphology and/or chemistry of the catalyst and by optimizing the operating parameters. Multiple operating parameters play an important role in the photocatalytic degradation process, thus making their optimization important for achieving good photocatalytic degradation of the target pollutant. Response surface methodology (RSM) is one of the most commonly applied optimization techniques; it is a powerful optimization tool for an experimental design that efficiently helps in systemic analysis [5,11,14]. RSM uses mathematics and statistics to analyze the relative significance of influencing factors on the response of the studied system. RSM is suitable for predicting the effect of individual experimental operating parameters, in addition to locating interactions between parameters and their impact on a response variable. RSM uses a systematic technique to simultaneously vary all parameters and evaluate the influence of these parameters on photocatalytic degradation [32,33]. The greatest advantage of RSM lies in the systematic approach for the experimental design, which mostly requires fewer experiments, thus reducing the time required and thereby being more economical. For designing these experiments, a central composite design (CCD) [3] and Box–Benkhen design (BBD) [11,12] are most commonly used. For the same number of parameters, BBD requires fewer experiments than CCD [3]; therefore, in this study, BBD is selected as a preferred design approach.
92
The objective of this study was to analyze the photocatalytic degradation of phenol with metal-free g-C3N4 and visible LED light and to model the process using RSM. In this study, the operating parameters considered were catalyst concentration, phenol concentration, and pH of the solution. BBD was used for the experimental design and RSM was applied to determine the mathematical relationship between operating parameters and phenol degradation. Finally, the correlation determined by RSM was experimentally validated.
#### **2. Materials and Methods**
#### *2.1. Chemicals and Materials*
Melamine (C3H6N6, 99%) was purchased from Alfa Aesar. Phenol (C6H5OH, 99%) was purchased from Merck. Acetonitrile (C2H3N, 99.99%) and ultra-pure water for highperformance liquid chromatography (HPLC) were purchased from Sigma Aldrich. NaOH and HCl were purchased from VWR chemicals. All chemicals used were of analytical grade and used as-received without any further purification.
#### *2.2. Photocatalyst Synthesis*
Photocatalyst was prepared as per the procedure used in our previous study [18]; the synthesis process is briefly reported here. Melamine was placed in a muffle furnace (Carbolite Gero, GPC 1200, Derbyshire, UK) in a closed crucible to prepare bulk g-C3N4 using thermal decomposition. The synthesis process comprised two steps: A heating ramp rate of 2 ◦C min−<sup>1</sup> was programmed up to 450 ◦C; this temperature was maintained for 2 h. Then, the temperature was increased to 550 ◦C using a heating ramp rate of 2 ◦C min−<sup>1</sup> and then maintained for 4 h. The material synthesized was crushed in mortar after cooling, then rinsed with ultrapure water, and dried overnight at 80 ◦C. The exfoliation process was conducted in an open crucible at 500 ◦C for 2 h at a heating ramp rate of 2 ◦C min−<sup>1</sup> in a muffle furnace.
#### *2.3. Characterization of the Photocatalyst*
Fourier transform infrared (FTIR) measurements (4000–400 cm−1) were performed on a Spectrum Two FT-IR Spectrometer (PerkinElmer, Switzerland) with a universal ATR (UATR Two) cell equipped with a ZnSe single crystal. The acquisition performed using 60 scans and the resolution was set to 4 cm−1. Zetasizer Nano ZEN5600 (Malvern, UK) was used to measure the zeta potential of the synthesized material. SU8030 (Hitachi, Japan) SEM-type microscope operated at an acceleration voltage of 10 kV and a probe current of 15 pA was used to examine the morphology of the material with scanning electron microscopy (SEM).
#### *2.4. RSM with Box–Behnken Experimental Design*
The influence of three independent operating parameters, i.e., catalyst concentration (A), phenol initial concentration (B), and pH of the solution (C), was considered in RSM. The remaining reaction conditions, namely, the airflow rate (50 mL/min) and reaction time (3 h), was kept constant in the experiment based on previous study [18]. The degradation efficiency of phenol (Equation (1)) was set as a response variable. Note that a previous study [18] was conducted to obtain the upper and lower limits of the parameters. Table 1 shows the ranges and levels of independent parameters A, B, and C. BBD was used to examine the combined effect of these three variables. Section 3.3 lists the set of experiments in table; it includes a replication of experiments at the central point. Regression analysis was the performed using OriginPro 2021 9.8.0.200 (OriginLab Corporation, Northampton, MA, USA) software. The suggested model's data were analyzed for significance and suitability using analysis for variance (ANOVA).
**Table 1.** Independent parameters and their ranges and levels.
#### *2.5. Photocatalytic Experiments*
Figure 1 shows the photocatalytic experiments that were conducted in a jacketed glass reactor (working volume 225 mL) (Peschl Ultraviolet GmbH, Mainz, Germany) with a safety cabinet. The reactor was irradiated from inside using a custom-made LED immersion lamp; the LED has maximum emission at 430 nm. Glass reactor was then sonicated with a reaction mixture for uniform dispersion, followed by stirring with continuous airflow to maintain adsorption–desorption equilibrium for 30 min. Subsequently, lights were turned on, which is considered as zero time (to). Nine to ten samples (1 mL) were periodically collected from the reaction mixture. After centrifugation and filtration, the samples were analyzed using HPLC. For acidic and basic reaction conditions, the pH of the mixture was adjusted using 0.1 M HCl and NaOH. The phenol degradation efficiency was determined using the following Equation:
$$\text{Depradiation efficiency } (\%) = \frac{\text{C}\_o - \text{C}}{\text{C}\_o} \times 100 \tag{1}$$
where *Co* is the initial phenol concentration and *C* is the residual phenol concentration in the solution at an irradiation time t.
**Figure 1.** Photocatalytic reactor setup.
The reduction of the reaction mixture volume due to the sampling was less than 5% at the end of the experiments and was therefore not considered in the calculation of the phenol degradation efficiency.
#### *2.6. Analytical Techniques*
A prominence HPLC system from Shimadzu (Kyoto, Japan) was used for analyzing the samples obtained from the reactor. The system is equipped with a binary pump (Model LC-20AB), an autosampler (Model SIL-20A), a degasser (Model DGU-20A3,) and a diodearray detector (Model SPD-M20A). Phenomenex (C18, 150 × 4.6 mm, 3 μm) column was used with a fixed flow rate of 0.8 mL/min, with the mobile phase gradient of water (A) and acetonitrile (B): starts with 15% B, followed by 60% B in 7 min and back to 15% B in 8 min;
injection of 5 μL; UV light of 254 nm. Phenol was analyzed at a maximum absorption wavelength (λmax) of 270 nm.
#### **3. Results and Discussion**
#### *3.1. Photocatalyst Characterization*
The metal-free g-C3N4 used in this study was synthesized and characterized in our previous study [18] using transmission electron microscopy (TEM), Brunauer–Emmett–Teller isotherms (BET), X-ray diffraction (XRD), X-ray photoelectron spectroscopy (XPS), photoluminescence (PL), and UV-Vis spectroscopy. In this study, scanning electron microscopy (SEM), Fourier transform infrared spectroscopy (FTIR), and zeta potential analyses were performed. Table 2 lists the physical properties of metal-free g-C3N4 before and after its exfoliation.
**Table 2.** Summary of characterization results [18].
The exfoliated material has a significantly higher surface area than the bulk material, while the average pore size of both materials is almost the same (Table 2 and Figure S1). Using XRD, the material shows two characteristic peaks of g-C3N4 (Figure S4) [34,35]. The strong and weak peaks of N1s and C1s observed in XPS confirm the chemical state of g-C3N4 (Figure S3) [17,36–41]. Table 2 lists the maximum absorption wavelength and bandgap of the material, which are presented in Figure S2 [42,43].
In Figure 2, the selected SEM images of bulk and exfoliated g-C3N4 are presented. The thermal exfoliation transformed the stacked and aggregated structure of bulk g-C3N4 in a porous nanosheet structure. The reduction in layer thickness (Figure 2b) leads to an increase in the specific surface area of g-C3N4 [17,44–46].
**Figure 2.** SEM images of the bulk (**a**) and exfoliated (**b**) g-C3N4.
Figure 3 shows the catalysts' FTIR spectra. A broad peak is observed between 3200 and 3000 cm−1, which can be attributed to the stretching vibrations of N–H bonds from
residual amino groups and adsorbed H2O. The sharp peak that appears at 806 cm−<sup>1</sup> can be attributed to the breathing mode of triazine units [47,48], whereas the strong bands between 1636 and 1242 cm−<sup>1</sup> belong to the C=N and C–N bonds of heterocyclic rings. Because the spectra of both materials show the same absorption bands, the chemical structure remained unaltered after treatment.
**Figure 3.** Fourier transform infrared spectra of bulk and exfoliated g-C3N4.
Figure 4 shows the effect of pH on the zeta potential of the exfoliated g-C3N4. The catalyst surface is positively charged at acidic pH (3) and negatively charged at natural (6) and basic pH (10).
**Figure 4.** Zeta potential at different pH of the synthesized exfoliated g-C3N4. Reproduced with permission from [18].
The optical properties (PL/UV-Vis) and surface area (BET) of the material have changed with exfoliation; however, the chemical state (XPS), phase (XRD), and the chemical structure (FTIR) remained the same after exfoliation.
#### *3.2. Photodegradation Studies*
The photodegradation efficiency of exfoliated g-C3N4 photocatalyst was evaluated under visible light irradiation using 430 nm wavelength LEDs. The influence of individual operation parameters, catalyst concentration, phenol concentration, and pH of the solution, in their preselected ranges (Table 1), was examined. For all experiments, an adsorption time of 30 min was used before the light irradiation was started. Moreover, the photolysis experiment was performed to verify the removal of phenol in the absence of the catalyst. Phenol removal with adsorption in the dark and photolysis is insignificant compared to the removal of phenol obtained in the presence of light (Figure 5a). Figure 5a shows the effect of g-C3N4 photocatalyst concentration in the range of 0.1–0.75 g/L on phenol degradation, which increased with the increase in catalyst concentration up to 0.75 g/L because of an increased number of active sites available for the reaction to occur. However, there is no significant increase at >0.5 g/L because an additional increase of the catalyst concentration might cause light scattering and hindrance in light absorption. The effect of phenol concentration on the performance of the catalyst on phenol degradation was examined for three concentrations between 20 and 100 ppm and is shown in Figure 5b. The phenol degradation efficiency decreased as the concentration increased because of the higher number of molecules for adsorption on the available active sites, which hinders the absorption of light. Figure 5c shows the effect of different pH on phenol degradation. Increasing the pH decreases the degradation efficiency of exfoliated g-C3N4. Note that acidic pH is most favorable for phenol degradation because as per the zeta potential (Figure 3) and the surface charge of the catalyst is positive at an acidic pH, which helps attract OH– ions produced in the solution due to dissociation of H2O2 to the surface and improves the degradation efficiency.
**Figure 5.** Phenol degradation at preselected (**a**) catalyst concentration (at 20 ppm and natural pH) (**b**) pollutant concentration (at 0.5 g/L and natural pH), and (**c**) pH of the solution (at 0.5 g/L and 20 ppm); airflow = 50 mL/min. Reproduced with permission from [18].
#### *3.3. Response Surface Methodology*
#### 3.3.1. Model Equation
To analyze the combined effect of three variables: catalyst concentration (A), phenol concentration (B), and pH of the solution (C) on the degradation efficiency of phenol (Equation (1)), a three-variable BBD was used in the experimental design for RSM. Table 3 lists the set of performed experiments and the obtained phenol degradation (in 3 h and under an airflow of 50 mL/min).
**Table 3.** Box–Behnken design with experimental and predicted phenol degradation efficiency values with Equation (2).
Experimental data were fitted with four different models: two-factor interaction (2FI), linear, quadratic, and cubic model to obtain regression equations. Three different tests, namely, the sequential model sum of squares, lack of fit, and model summary statistics, were conducted to determine the adequacy of various models; the results are presented in Table 4. The response surface model is then used to select the best model based on the following criterion: the highest-order polynomial with additional significant terms and the model is not aliased (Table 4). The cubic model has the highest polynomial model because there are no sufficient unique design points to independently estimate all terms for that model. The aliased model results in unstable and inaccurate coefficients and graphs. Thus, the aliased model cannot be selected [49,50]. The criteria used in the lack of fit test is the non-significant lack of fit (*p*-value > 0.05) based on which a quadratic model is selected. Moreover, multiple summary statistics are calculated to compare models or to confirm the adequacy of the model. These statistics include adjusted R2, predicted R2, and prediction error sum of squares (PRESS). A good model will have a largely predicted r2, and a low PRESS. According to the aforementioned criteria, adjusted R2 (0.967) and predicted R<sup>2</sup> (0.805) are in reasonable agreement with each other and have a low PRESS. Thus, the quadratic model is finally selected to build the response surface.
**Table 4.** Adequacy of the models tested.
Based on regression coefficients from Table 5, the following empirical second-order polynomial equation was obtained:
#### Degradation Efficiency (%)
<sup>=</sup> 85.72 <sup>+</sup> 6.36 A <sup>−</sup> 24.86 B <sup>−</sup> 15.22 C <sup>+</sup> 3.71 AB <sup>−</sup> 2.83 AC <sup>−</sup> 2.38 BC <sup>−</sup> 5.05 A2 <sup>−</sup> 5.22 B2 <sup>−</sup>16.07 C2 (2)
> where, A, B, and C are the catalyst concentration, phenol concentration, and pH of the solution, respectively.
**Table 5.** Coefficients of the second-order polynomial (quadratic) equation.
The influence of model terms on the degradation of phenol as per p-values (Table 5) is in the following order B < C < C2 <A<B<sup>2</sup> < A<sup>2</sup> < AB < AC < BC. The mixed interaction terms AB, AC, and BC are not significant because their *p*–value is > 0.05 and may be removed from Equation (2).
An ANOVA of the second-order polynomial (Equation (2)) for phenol degradation was conducted; the results are shown in Table 6. In statistics, the significance of the model can be confirmed by a large F-value (53.31) and a small *p*-value (<0.0001). Furthermore, the significance of the model can be confirmed by the lack of fit test. In this study, the lack of fit is not significant because its *p*-value is >0.05. The accuracy of the model is confirmed by the low coefficient of variation (CV) value of 5.79%. The results showed that the signal-to-noise ratio of 24.89 is adequate.
**Table 6.** Analysis of variance ANOVA of the second-order polynomial (Equation (2)).
Furthermore, the coefficient of determination R2 confirmed the fit of the model. For the used model, the value of the predicted R<sup>2</sup> = 0.810 (Table 6) is in agreement with adjusted R<sup>2</sup> = 0.967, which indicates that the obtained model is significant.
Equation (2) provides a suitable relationship (R<sup>2</sup> = 0.810) between the response (degradation efficiency) and the parameters, which can be seen in Figure 6. In this figure, the experimental values of phenol degradation are plotted against the predicted values obtained from the RSM model; these values of the percentage phenol degradation fit well.
**Figure 6.** The experimental phenol degradation efficiency (%) plotted against the predicted values from the RSM model.
3.3.2. Interaction Effects of Independent Operating Parameters
Three dimensional (3D) response surface and contour plots were generated using the regression model (Equation (2)) to visualize the influence of the independent operating parameters on phenol degradation; they are presented in Figures 7–9. In surface and contour plots, one parameter is maintained constant at its zero levels, whereas the other two are varied in the studied range reported in Table 1.
**Figure 7.** Effect of catalyst concentration and pH on the degradation of phenol: pollutant concentration was kept constant at 60 ppm.
**Figure 8.** Effect of pollutant concentration and pH on the degradation of phenol: catalyst concentration was kept constant at 0.5 g/L.
**Figure 9.** Effect of catalyst concentration and pollutant concentration on the degradation of phenol: pH was kept constant at 6.5.
Figure 7 shows the influence of pH and catalyst concentration on the degradation efficiency of phenol at a constant phenol concentration of 60 ppm. The contour lines show a decrease in the degradation efficiency with an increase in pH; there is no considerable increase in efficiency, even at higher catalyst concentrations. However, an increase in degradation efficiency with a decrease in pH is observed. These results demonstrate that pH has a significant effect on phenol degradation and a low pH favors the degradation process. This phenomenon is linked with the zeta potential of the catalyst surface [18]. There is a positive charge at the surface of the catalyst at an acidic pH (Figure 2), which attracts the OH− ions produced in the solution due to dissociation of H2O2 and significantly increases the degradation process. However, at a basic pH, the surface charge is negative and there could be electrostatic repulsion that reduces the efficiency of the degradation process.
Figure 8 shows the influence of pH and pollutant concentration on phenol degradation at a constant catalyst concentration of 0.5 g/L. For selecting the catalyst concentration, the effect of initial pollutant concentration is important. The contour lines demonstrate that simultaneously increasing both parameters (pH and phenol concentration) considerably decreases the degradation efficiency of phenol (33%), which is 62% at a low pH. As shown in Figure 5b, at low pH and low pollutant concentration, 100% degradation is achieved in a considered reaction time of 3 h. An increase in degradation efficiency from high to low pH can then be associated with catalyst surface charge. However, a decrease in efficiency at low pH from low to high phenol concentration is attributed to the increased number of pollutant molecules compared with the available active sites.
Figure 9 shows the effect of catalyst concentration and pollutant concentration at a constant pH of 6.5. The contour lines demonstrate that both parameters independently affect the degradation efficiency. By increasing the catalyst concentration at a lower pollutant concentration, phenol degradation increases; however, at a higher pollutant concentration, the degradation efficiency decreases. This can be attributed to the availability of active sites on the catalyst surface for OH− radicals, as well as phenol molecules. The electron–hole pair generated from the catalyst surface improves the degradation rate.
#### 3.3.3. Experimental Validation of RSM Model
To demonstrate the applicability of the model, a hypothetical case study for water with a phenol concentration of 50 ppm was considered. The model equation was used to identify the optimum catalyst concentration and pH, leading to maximal phenol degradation in 3 h under an airflow rate of 50 mL/min. According to the model prediction, maximal phenol degradation of 88.62% is achievable using 0.4 g/L of catalyst concentration and operating at a pH of 6.5. To examine the accuracy of the model prediction, an experiment was conducted under these conditions. The experimentally obtained phenol degradation was 83.75%, which is less than a 5% deviation from the predicted value. Thus, the optimum operating point obtained by RSM was successfully confirmed; this suggests that RSM can be a useful tool for optimizing photocatalytic processes. Similarly, the model developed can be used for minimizing the catalyst amount or for maximizing the degradation efficiency of phenols for any set of parameters in range.
#### **4. Conclusions**
Metal-free g-C3N4 was used for the photocatalytic degradation of phenol from an aqueous solution. The morphology of the catalyst was confirmed by SEM, and the surface charge was confirmed using zeta potential. Based on zeta potential, the catalyst surface was confirmed to have a positive surface charge under acidic conditions and a negative surface charge under basic conditions; therefore, acidic pH favors the degradation process. A RSM based on the BBD was used to analyze the degradation efficiency of phenol. The influence of experimental parameters, namely, catalyst concentration, pollutant concentration, and pH of the solution, and their interaction at a different level was examined for phenol degradation. An empirical regression quadratic model was developed for the response variable. Analysis of variance (ANOVA) demonstrated that the model is significant with an insignificant lack of fit and a high coefficient of determination (R2) of 0.96, which can be helpful to navigate the design space. Furthermore, an optimized degradation efficiency of 83.75% was achieved for phenol concentration of 50 ppm, catalyst concentration of 0.4 g/L, and a solution pH of 6.5 pH (in 3 h and under an airflow of 50 mL/min). Thus, the results suggest that the RSM can be used for the optimization of parameters for maximizing the photocatalytic degradation of phenol using g-C3N4 and LEDs.
**Supplementary Materials:** The following are available online at https://www.mdpi.com/article/10 .3390/catal11080898/s1, Figure S1 N2 adsorption-desorption isotherms of bulk and exfoliated g-C3N4. The inset shows the corresponding BJH pore size distribution curves of the sample, Figure S2 (**a**) UV-Vis absorption spectra and (**b**) PL spectra of bulk and exfoliated g-C3N4; insets of (**a**) showing the Tauc plots, Figure S3 XPS spectra of bulk and exfoliated g-C3N4 C1s, N1s, Figure S4 X-ray diffraction patterns of bulk and exfoliated g-C3N4.
**Author Contributions:** Conceptualization, A.G.R.; Formal analysis, A.G.R.; Investigation, A.G.R.; Methodology, A.G.R.; Resources, M.M.; Supervision, M.M.; Writing—original draft, A.G.R.; Writing review and editing, M.M. All authors have read and agreed to the published version of the manuscript.
**Funding:** This research received no external funding.
**Acknowledgments:** A.G.R. acknowledges the financial support from the Higher Education Commission, Pakistan, and Deutscher Akademischer Austauschdienst (DAAD), Germany.
**Conflicts of Interest:** The authors declare no conflict of interest.
#### **References**
| doab | 2025-04-07T03:56:58.578118 | 6-5-2022 13:34 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "0039f385-c9b0-4fa4-b06c-9a5b9a2397e4",
"url": "https://mdpi.com/books/pdfview/book/5221",
"author": "",
"title": "Materials and Processes for Photocatalytic and (Photo)Electrocatalytic Removal of Bio-Refractory Pollutants and Emerging Contaminants from Waters",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036535593",
"section_idx": 3
} |
0039f385-c9b0-4fa4-b06c-9a5b9a2397e4.4 | **On the Role of the Cathode for the Electro-Oxidation of Perfluorooctanoic Acid**
**Alicia L. Garcia-Costa 1,2,\*, Andre Savall 2, Juan A. Zazo 1, Jose A. Casas <sup>1</sup> and Karine Groenen Serrano 2,\***
**\*** Correspondence: [email protected] (A.L.G.C.); [email protected] (K.G.S.)
Received: 22 July 2020; Accepted: 6 August 2020; Published: 8 August 2020
**Abstract:** Perfluorooctanoic acid (PFOA), C7F15COOH, has been widely employed over the past fifty years, causing an environmental problem because of its dispersion and low biodegradability. Furthermore, the high stability of this molecule, conferred by the high strength of the C-F bond makes it very difficult to remove. In this work, electrochemical techniques are applied for PFOA degradation in order to study the influence of the cathode on defluorination. For this purpose, boron-doped diamond (BDD), Pt, Zr, and stainless steel have been tested as cathodes working with BDD anode at low electrolyte concentration (3.5 mM) to degrade PFOA at 100 mg/L. Among these cathodic materials, Pt improves the defluorination reaction. The electro-degradation of a PFOA molecule starts by a direct exchange of one electron at the anode and then follows a complex mechanism involving reaction with hydroxyl radicals and adsorbed hydrogen on the cathode. It is assumed that Pt acts as an electrocatalyst, enhancing PFOA defluorination by the reduction reaction of perfluorinated carbonyl intermediates on the cathode. The defluorinated intermediates are then more easily oxidized by HO• radicals. Hence, high mineralization (xTOC: 76.1%) and defluorination degrees (xF −: 58.6%) were reached with Pt working at current density *j* = 7.9 mA/cm2. This BDD-Pt system reaches a higher efficiency in terms of defluorination for a given electrical charge than previous works reported in literature. Influence of the electrolyte composition and initial pH are also explored.
**Keywords:** perfluorooctanoic acid; emerging contaminant; defluorination; platinum; electro-oxidation
#### **1. Introduction**
Perfluoroalkyl substances (PFAS), such as perfluorooctanoic acid (PFOA, C7F15COOH) are widely used in the chemical industry because of their amphiphilicity, stability, and surfactant property. They are employed in the synthesis of fluoropolymers and fluoroelastomers, as surfactants in fire-fighting foams, and in textile and paper industries to produce water and oil repellent surfaces [1]. Nevertheless, despite their practical interest, these substances present a high toxicity due to their potential bioaccumulation, and common occurrence in water resources. PFOA has been recognized as an emerging environmental pollutant and has been included in the European Candidate List of Substances of Very High Concern ("SVHC") [2]. Hence, the current challenge is to develop highly efficient and cost-effective processes for the elimination of perfluoroalkyl substances at source.
The main issue in PFOA degradation is to break the C-F bond, one of the strongest bonds known (≈460 kJ/mol) [3]. This confers a high stability and resistance to PFAS which cannot be degraded by direct hydrolysis, photolysis, or through conventional biological treatments [4]. As a result, PFAS have been detected in natural water streams [5], sediments [6], and even in tap and bottled water in
concentrations up to 640 ng/L [7]. So far, adsorption onto carbonaceous materials [8], alumina [9], or other sorbents [10] have been successfully applied for PFAS removal. Nonetheless, this technology implies the transfer of the pollutant to another phase, the sorbent, which becomes a new residue after use. To overcome this drawback, advanced oxidation processes (AOP) are being explored for PFAS removal. AOP are based on the use of strong oxidizing radicals to degrade, most commonly, organic pollutants in aqueous phase [11]. The most extended AOP are those based on the use of hydroxyl radicals (HO•) to attack organic pollutants by hydrogen abstraction [12]. Consequently, the substitution of all organic hydrogen for fluorine in PFOA makes these compounds inert to this kind of AOP. The non-reactivity of PFOA to HO• attack has been confirmed by various studies [13–15]. As a matter of fact, Maruthamuthu et al. have shown that the reactivity of hydroxyl radicals on acetate decreases considerably with increasing halogen substitution [13]. Using the Fenton process, known to generate hydroxyl radicals by the action of Fe (II) on hydrogen peroxide, no degradation was observed when the Fenton reagent (0.2 mM, Fe2<sup>+</sup>: H2O2, molar ratio = 1:1) was mixed with PFOA (0.02 mM) at room temperature [14]. Similar results were obtained by Santos et al. with only 10% PFOA removal and without any C-F bond cleavage [16].
More recently, photocatalytic treatments have been applied for PFOA degradation. This technology achieved high PFOA removal (xPFOA > 90%) when using modified TiO2 photocatalysts such as Cu-TiO2 [17], Pb-TiO2 [18], and rGO-TiO2 [19].
Besides PFOA removal, defluorination (xF2212−) is a very important parameter to evaluate the process efficiency. xF <sup>−</sup> defined as the ratio of the fluoride concentration (CF <sup>−</sup>, measured) released by PFOA degradation with respect to the initial content of fluoride in the initial amount of PFOA molecule (CF,PFOA 0) is expressed in percentage as shown in Equation (1).
$$\chi\_{\text{F}^{-}} = \frac{\mathbf{C}\_{\text{F}^{-}, \text{measured}}}{\mathbf{C}\_{\text{F}, \text{ PFOA}\_{\text{A}}}} \cdot 100 \tag{1}$$
In photochemical oxidation of PFOA, defluorination is usually low (xF − < 25%), with the average xF <sup>−</sup>/xPFOA ratio around 0.26 [20].
Another technique for PFAS remediation is electrochemical degradation. PFOA electrooxidation has been successfully carried out in different systems using boron-doped diamond (BDD) as anode (Table 1). Under the studied conditions, PFOA removal ranged from 60% to 100%. It should be noted that defluorination values were very different, suggesting that either the operating conditions (electrolyte, pH, etc.,) or the cathode reduction reactions may play a key role in the PFOA degradation mechanism. The cleavage of the C-F bonds to form F− ions is interesting because F− ions readily combine with Ca2<sup>+</sup> to form environmentally harmless CaF2, as reported by Hori et al. [3].
xF <sup>−</sup> and xPFOA ratios obtained by electrochemical treatment (up to 80–85%, Table 1) are higher than those reported in photo-oxidation (<25%). Nonetheless, all previous electrooxidation studies were conducted employing a high supporting electrolyte concentration, which makes difficult to dispose the treated wastewater after reaction. Therefore, this work aims to gain knowledge on the role of the cathode as electrocatalyst in PFOA electrooxidation working at low electrolyte concentration (3.5 mM). For this purpose, BDD was chosen as the anode and BDD, Pt, Zr, and stainless steel were tested as cathodes in the degradation of 100 mg/L PFOA.
**Table 1.** Perfluorooctanoic acid (PFOA) electrooxidation with boron-doped diamond (BDD) anode.
#### **2. Results and Discussion**
In order to test the influence of the cathode material on the degradation process, a BDD anode was successively coupled with cathodes made of BDD, Pt, Zr, and stainless steel. Results of electrolysis runs conducted at 7.9 mA/cm2, at 25 ◦C, for the treatment of 100 mg/L PFOA solutions (namely, 0.242 mol/m3), are presented in Figure 1a. For each couple of electrodes, the curves show that PFOA concentration followed, from CPFOA,0 = 100 mg/L to CPFOA,t ≈ 25 mg/L, a similar decrease, characteristic of a pseudo-first order kinetics. For experiments presented in Figure 1a the applied current density was higher than the limiting current density. Considering a pure mass transport controlled reaction for the first exchange of charge between a molecule of PFOA and the anode surface, the limiting current density calculated using the equation established from the Nernst diffusion model: *j*lim = *n*·*F*·*km*·CPFOA,0 equals to 0.63 A/m<sup>2</sup> for *n* = 1, *F* = 96,485 C/mol, *km* = 2.7·10−<sup>5</sup> m/s (determined experimentally using the ferri/ferro system, as described elsewhere [26]) for a flow rate of 0.360 m3/h [27], CPFOA,0 = 0.242 mol/m3. This value of the limiting current density is more than 100 times lower than that applied during electrolysis (*j* = 79 A/m2). Under these conditions, the decay of concentration from CPFOA,0 to CPFOA,t depends upon the mass transfer coefficient *km*, the surface area (A) of the electrode, and the volume (V) of electrolyte [28], as follows:
$$\mathbf{C}\_{\text{FFOA},t} = \mathbf{C}\_{\text{FFOA},0} \mathbf{e}^{\left(\frac{-t}{\pi}\right)} \tag{2}$$
where the constant of time, τ, is defined by: τ = V/(*km* A). For its calculation we considered the following values V <sup>=</sup> <sup>10</sup>−<sup>3</sup> <sup>m</sup>3, A <sup>=</sup> <sup>63</sup>·10−<sup>4</sup> <sup>m</sup>2, and *km* <sup>=</sup> 2.7·10−<sup>5</sup> <sup>m</sup>/s, obtaining a time constant (τ) equal to 5800 s. According to Equation (2), the PFOA theoretical concentration at t = 2 h is around 29 mg/L, which is in agreement with the experimental results (≈25 mg/L), as shown in Figure 1a.
**Figure 1.** Influence of the cathode material in (**a**) PFOA removal (symbols: experimental data, lines: kinetic fitting), (**b**) TOC depletion, (**c**) released fluorine (symbols: experimental data, lines: kinetic fitting). Operating conditions: [PFOA]0: 100 mg/L, j: 7.9 mA/cm2, electrolyte: 3.5 mM Na2SO4, T: 25 ◦C, pH0: 4.
The excess of charge on the first oxidation step of PFOA is used for other electron transfers and by the action of HO• radicals during the degradation of the numerous intermediates. PFOA removal after 6 h was 100%, 98.1%, 97.9%, and 97.6% for Pt, steel, Zr, and BDD, respectively. Values of the rate constants and regression coefficients are collected in Table 2. It should be noted that the Pt cathode has the best results with a PFOA degradation rate 39% faster than the other tested materials, which exhibit a similar behavior between them. This enhancement is also reflected in the mineralization degree (Figure 1b), with a 76.1% Total Organic Carbon (TOC) removal with the BDD-Pt system.
**Table 2.** PFOA degradation and fluoride release kinetics.
As previously explained, one of the main challenges in PFOA oxidation is the effective breakdown of the C-F bond. PFOA defluorination was followed along the reaction by means of ionic chromatography, as depicted in Figure 1c. The trend for defluorination was Pt > BDD > Zr > steel. In this case, the cathode also played an important role, reaching a 58.6% in the case of Pt, against 42–49% for BDD, Zr, and steel. Moreover, fluoride release follows a first order, as reflected in Figure 1 and Table 2, where the function of Pt as electrocatalyst is confirmed.
Pt is a common catalyst in hydrodehalogenation reaction of organic molecules, because of its capacity to adsorb hydrogen, providing a catalytic site were the dehalogenation takes place [29]. H2 generation by water electrolysis on the cathode's surface may be responsible for PFOA hydrodefluorination, following the reaction mechanism shown in Figure 2.
**Figure 2.** PFOA electrooxidation mechanism.
Because PFOA is inert to hydroxyl radicals, its degradation is initiated on the anode by a direct electron transfer reaction to form a perfluoro radical C7F15COO• (Equation (3)). This radical loses its carboxylic group (Equation (4)) and reacts with HO• leading to the generation of C7F13−CF2OH (Equation (5)), as previously described by Zhang et al. [30].
$$\text{BDD} + \text{C\gamma} \text{F}\_{15} \text{COO}^{-} \rightarrow \text{BDD} + \text{C\gamma} \text{F}\_{15} \text{COO}^{\bullet} + \text{e}^{-} \tag{3}$$
$$\rm C\_7F\_{15}COO^\bullet \rightarrow \rm C\_7F\_{15}\uparrow + CO\_2 \tag{4}$$
$$\rm C\_7F\_{15}\rm \rm ^\bullet + HO^\bullet \to \rm C\_7F\_{15}OH \tag{5}$$
This alcohol then reacts according to three pathways, (for clarity reasons, only the first one (i) is illustrated in Figure 2):
(i) With adsorbed hydrogen generated by water electro-reduction at the cathode, releasing 2 F− (Equation (6)).
$$\rm C\_6F\_{13}CF\_2OH + 4H\_{ads} \rightarrow \rm C\_6F\_{13}CH\_2OH + 2HF \tag{6}$$
As the first carbon in the alkyl chain is now defluorinated, HO• can attack it once again leading to the formation of C6F13COOH. This mechanism is similar to that presented for PFOA photocatalytic degradation byWang et al. [20] and theoretic quantum calculations and experimental data collected by Trojanowicz et al. [31]. Hence, this step depends strongly on the cathode material.
(ii) With hydroxyl radicals leading to the formation of COF2, as related by Niu et al. [31] and Zhang et al. [28], following Equations (7)–(9):
$$\rm C\_7F\_{15}OH + HO^\bullet \rightarrow C\_7F\_{15}O^\bullet + H\_2O \tag{7}$$
$$\text{C}\_7\text{F}\_{15}\text{O}^\bullet \rightarrow \text{C}\_6\text{F}\_{13}\text{"}+\text{COF}\_2\tag{8}$$
$$\text{COF}\_2 + \text{H}\_2\text{O} \rightarrow \text{CO}\_2 + 2\text{HF} \tag{9}$$
According to George et al. [30] hydrolysis of carbonyl fluoride COF2 in the aqueous phase is extremely fast since its half-life is 0.7 s at T = 273 K.
(iii) Giving the perfluorocarbonyl fluoride (Equation (10)) for which hydrolysis leads the formation of perfluorocarboxylic acid and HF (Equation (11)) [30,32].
$$\rm C\_7F\_{15}OH \rightarrow \rm C\_6F\_{13}COF + HF \tag{10}$$
$$\rm C\_6F\_{13}COF + H\_2O \rightarrow C\_6F\_{13}COO^- + HF + H^+ \tag{11}$$
Considering this complex reaction mechanism, it should be noted that TOC decay was faster within the first hour of reaction, then it slowed down (Figure 1b). This is related to the generation of short-chain fluorinated acids (decarboxylation step), which are less active to electro-oxidation processes. In fact, pH value in the Pt system decreased from 4 to 3.2 in 120 min, maintaining this pH until the end of the reaction, which evidences the generation of these acidic species.
Data displayed in Figure 1 allow to determine the fluoride concentrations produced with respect to the degraded carbon in the form of CO2 (F−/CO2) or with respect to the PFOA eliminated over time (F−/PFOA). Figure 3 shows that the PFOA defluorination leads to the formation of 1 to 1.5 fluorine ions per removed atom of carbon in the first three hours of electrolysis. According to the proposed mechanism, this value close to the one at the beginning of the electrolysis is related to decarboxylation which leads to the formation of Rf-COF. The kinetics of this step are probably faster than that of the defluorination stages according to Equations (6)–(9). Besides, part of the process can be attributable to the electrocatalytic hydrogenation of the perfluorocarbonyl fluoride C6F13COF that forms simultaneously the hydrofluoric acid and the 1,1-dihydroperfluoroalkyl alcohol C6F13CH2OH. This alcohol is stable but easily oxidizable on the BDD anode [33] (cf. Figure 2). This process is slowed down by the diffusion of the species to the cathode. Not all molecules undergo the loss of two fluoride atoms, which would explain the value of 1.5 instead of the usual ratio 1.9 present in the initial PFOA molecule.
In addition, Figure 3 shows the variation of the ratio between the concentration of fluoride ions released and the concentration of the removed PFOA. This ratio varies from 7.7 to 9 for 360 min of electrolysis. These values highlight the high, yet incomplete PFOA defluorination. Finally, the ratio between the carbon loss (in the form of CO2) and the removed PFOA (CO2/PFOA) is in the order of 6–7, slightly less than 8, i.e., the theoretical value for C7F15COOH, confirming the formation of reaction intermediates. This ratio decreases during electrolysis: the degradation being faster at the beginning of the reaction, until t:150 min.
At this time more than 85% of PFOA has been eliminated. PFOA depletion slows down both the defluorination and carbon skeleton breaking. In addition, shorter molecular chains could display slower kinetics.
**Figure 3.** Variation of fluorine ions (full symbols) and carbon removal (empty symbols) during electrolysis with respect to carbon removal and degraded PFOA. Operating conditions: [PFOA]0: 100 mg/L, j: 7.9 mA/cm2, electrolyte: 3.5 mM Na2SO4, T: 25 ◦C, pH0: 4.
Figure 4 shows the ratio of fluorine and carbon atoms contained in the chemical intermediates. The molar concentration of F and C atoms contained in the intermediates are defined, respectively, as follows:
$$\mathcal{L}\_{\text{F,intermediate}} = 15 \cdot (\mathcal{C}\_{\text{FPCA},0} - \mathcal{C}\_{\text{FPCA},t}) - \mathcal{C}\_{\text{F}^{-}t} \tag{12}$$
$$\mathbf{C\_{C,inermodies}} = \mathbf{TOC\_t} - \mathbf{TOC\_{FPOA,t}} \tag{13}$$
where CPFOA,0 and CPFOA,t refer to the molar concentration of PFOA at initial time and at time t, respectively; CF−,t is the molar concentration of fluorine ions at t; TOCt and TOCPFOA,t are the total carbon molar concentration and the carbon molar concentration in the PFOA, respectively.
From Figure 1, after 360 min of electrolysis, the defluorination rate is 59% whereas more than 98% of PFOA and 76% of TOC have been eliminated. Figure 4 highlights that in this moment, the intermediates still contain 24% of carbon and 41% of fluorine. Xiao et al. reached a 90% defluorination and mineralization working at high temperature (T: 80–120 ◦C), meaning they managed to degrade the short-chain acids [25]. This is in agreement with the results for degradation of phenol in heterogeneous Fenton at high temperature, where maleic, malonic, oxalic, and formic acids can be completely degraded [11], in contrast with room temperature processes [34]. Aiming to verify Xiao et al.'s results, an electrooxidation run at 80 ◦C was performed using Pt cathode. After 30 min reaction there was an overvoltage on the cell due to the damage on the cathode, probably because of the HF attack (Figure S1 of the Supplementary Material). Hence, high temperature electrooxidation could not be performed in our system and further runs were conducted at 25 ◦C.
**Figure 4.** Ratio of fluorine and carbon atoms in the chemical intermediates. Operating conditions: [PFOA]0: 100 mg/L, j: 7.9 mA/cm2, electrolyte: 3.5 mM Na2SO4, T: 25 ◦C, pH0: 4.
After selecting Pt as the best cathode, within the tested materials, different salts were used as electrolyte, NaClO4, KNO3, Na2SO4, Na2S2O8, at 3.5 mM. Results for these experiments can be found in Figure 5. As previously reported by Schaefer et al. [21], the influence of the electrolyte type on PFOA degradation is very low. Still, significant differences were found for TOC removal, where the removal efficiency followed this trend Na2SO4 (76.1%) > Na2S2O8 (72.6%) > KNO3 (70.5%) > NaClO4 (67.1%). Sulfate achieved both a slightly higher mineralization degree and defluorination. This can be explained by the fact that sulfate anions behave as an active electrolyte via the electrochemical generation of the strong oxidizing sulfate radicals (SO4 •−) on a BDD anode [35,36]. Indeed, the oxidation of water at the anode greatly decreases locally the pH at the surface leading to the formation of HSO4 <sup>−</sup> from SO4 2−. Then HSO4 − reacts with HO• radicals to form sulfate radicals [33,37].
$$\mathrm{HSO}\_{4}^{-}+\mathrm{HO}^{\bullet}\rightarrow\mathrm{SO}\_{4}^{\bullet-}+\mathrm{H}\_{2}\mathrm{O}\quad\mathrm{k}=6.9\cdot10^{5}\mathrm{M}^{-1}\mathrm{s}^{-1}\tag{14}$$
SO4 •− radical participates in electron transfer reactions and promotes the decarboxylation of carboxylic acids, contrary to HO• which rather acts in hydrogen abstraction or addition [38]. In addition, sulfate radicals are more stable than hydroxyl radicals (their half-life is 30–40 μs and 10−<sup>3</sup> μs, respectively).
Considering PFOA degradation with sulfate radicals, the literature review by Yang et al. highlights that the decomposition and defluorination efficiencies increase with a decrease in PFOA chain-length [39]. Besides the major role of hydroxyl radicals on PFOA oxidation, the presence of sulfate radicals helps to improve the degradation of the generated intermediates. Qian et al. [40] estimated the constant rate of PFOA degradation with sulfate radicals at 2.59·10<sup>5</sup> <sup>M</sup><sup>−</sup>1s−1. This is consistent with the higher TOC removal observed in our experiments in presence of sulfate. Furthermore, sulfate is a more environmentally friendly electrolyte, in comparison to perchlorate and nitrate, which can be considered pollutants by themselves. Thus, the rest of experiments were carried out using Na2SO4 3.5 mM.
Influence of initial pH (pH0) on PFOA degradation was also evaluated working at the natural pH of PFOA solution (pH: 4) and at pH values of 7 and 9. Results for these experiments are shown in Figure 6. As it may be seen in Figure 6d, reaction media is quickly acidified. This is related to both the generation of short chain acids and the reaction between sulfate radicals and water to produce hydroxyl radicals, which also generates protons, as depicted in Equation (15). PFOA decay (Figure 6a) was similar for all the runs. However, pH0 had a great influence on the initial rate for TOC abatement, related to the higher oxidation potential of sulfate radicals in alkaline media [41]. Despite achieving a higher mineralization degree at pH0: 9, the highest defluorination was reached when starting in acidic media.
$$\text{SO}\_4^{\bullet-} + \text{H}\_2\text{O} \rightarrow \text{H}^+ + \text{HO}^\bullet + \text{SO}\_4^{2-} \tag{15}$$
**Figure 5.** Influence of the electrolyte in PFOA (**a**) degradation, (**b**) mineralization, and (**c**) defluorination in electrooxidation using BDD/Pt electrodes. Operating conditions: [PFOA]0: 100 mg/L, j: 7.9 mA/cm2, electrolyte: 3.5 mM, T: 25 ◦C, pH0: 4.
So far, the role of the cathode as electrocatalyst in the degradation and defluorination of PFOA has been proved. Also, the influence of several operating conditions has been tested, demonstrating an overall great decontamination working at low electrolyte concentration at mild temperature. Nonetheless, in order to compare the obtained results with those reported in literature, we have compared the defluorination degree against the energetic requirements, measured as the applied charge, as shown in Figure 7. As may be seen, both Shaefer et al. [21] and Urtiaga et al. [22] boosted the defluorination degree when increasing the applied charge. However, the results presented in this work using Pt cathode at 7.9 mA/cm2, 3.5 mM Na2SO4 at pH0:4 and T:25 ◦C are the most competitive in terms of PFOA defluorination against electric charge. In this sense, cathode selection becomes a key point for both increasing the activity and reducing the energy requirements in PFOA electrooxidation.
**Figure 6.** Influence of the pH0 in PFOA (**a**) degradation, (**b**) mineralization, and (**c**) defluorination and (**d**) pH evolution in electrooxidation using BDD/Pt electrodes. Operating conditions: [PFOA]0: 100 mg/L, j: 7.9 mA/cm2, Na2SO4: 3.5 mM, T: 25 ◦C.
**Figure 7.** Process comparison in terms of defluorination against electric charge for PFOA electrooxidation with BDD anodes. Cathodes: Schaefer et al.—W [21], Urtiaga et al.—W [22], Zhuo et al.—BDD [23], Ochiai et al.—Pt [24], this work—Pt.
#### **3. Materials and Methods**
#### *3.1. Reactants*
Perfluoroctanoic acid (95 wt.%), Na2SO4, KNO3, NaClO4, Na2S2O8, acetonitrile (ACN), H2SO4, and NaOH were supplied by Sigma-Aldrich (Darmstadt, Germany). All reagents are of analytical grade and they were used as received without further purification. Working standard solutions of PFOA and fluoride (NaF from Sigma-Aldrich, Darmstadt, Germany) were prepared for calibration.
#### *3.2. Experimental Set-Up*
The electrochemical oxidation system consists in a 1-L thermoregulated glass reservoir connected to the cell through a centrifugal pump. PFOA solution was recycled in the system at 360 L/h flow rate and the temperature was set at 25 ± 1 ◦C. The electrochemical cell is a one-compartment flow filter-press reactor which was operated under galvanostatic conditions using an ELCAL 924 power supply (Italy). Electrodes present a 63 cm<sup>2</sup> active surface and the gap between them was set at 10 mm. A detailed scheme of the experimental set-up can be found elsewhere [42]. All experiences were performed with a BDD anode from Adamant Technologies (La Chaux-de-Fonds, Switzerland), which was elaborated by chemical vapor deposition on a conductive substrate of Si. BDD (Adamant Technologies, La Chaux-de-Fonds, Switzerland), Zirconium, Stainless steel, and Pt (5 μm) on titanium substrate (provided by MAGNETO special anodes B.V., Schiedam, Netherlands) were employed as cathodes. Before each electrolysis, the working electrodes were anodically pretreated (40 mA/cm2 for 30 min in 0.1 M H2SO4) to clean their surfaces of any possible adsorbed impurities. Then, the system was rinsed by ultrapure water.
In a typical reaction 1 L PFOA solution (100 mg/L) with 3.5 mM Na2SO4 as electrolyte at the natural pH of the solution (pH:4) was loaded to the reservoir, preheated to 25 ◦C and recycled through the system. Once the selected temperature was reached, the power supply was turned on and current intensity was set at 0.5 A, representing this as the reaction starting time. Samples were taken at regular intervals in the tank. The global volume of samples was less than 10% of the total volume. All runs were performed by triplicate with a deviation lower than 5% in all cases.
#### *3.3. Analytical Methods*
Samples were periodically withdrawn from the reactors, filtered through 0.2 μm nylon syringe plug-in filters and immediately analyzed, without any further manipulation. PFOA concentration was measured by high performance liquid chromatography connected with an ultraviolet-visible spectrometry detector (HPLC-UV Agilent 1200 Series HPLC, Santa Clara, USA). An ion-exclusion column (ZORBAX Eclipse Plus C18, 100 mm, 1.8 μm, Agilent, USA) was used as the stationary phase. As mobile phase mixture of ACN/4 mM H2SO4 aqueous solution with a ratio: 3/2 was employed and the column temperature was set to 50 ◦C. A 60% CAN—40% mixture was employed at 0.5 mL/min. The detection UV wavelength was set to 206 nm. Total organic carbon was quantified using a TOC analyzer (Shimadzu TOC-VSCH, Kyoto, Japan). Fluoride was analyzed in an ion chromatograph with chemical suppression (Metrohm 790 IC, Herisau, Switzerland) using a conductivity detector. A Metrosep A supp 5–250 column (25 cm long, 4 mm diameter, Herisau, Switzerland) was used as the stationary phase and 0.7 mL/min of a 3.2 mM/1 mM aqueous solution of Na2CO3 and NaHCO3, respectively, as the mobile phase.
#### **4. Conclusions**
PFOA electro-degradation follows a complex mechanism which involves both oxidation reactions on the anode surface and reduction reactions, responsible for the molecule's defluorination, which take place over the cathode. Electrocatalytic hydrogenation of the unsaturated acyl fluoride RfCOF can be a route for the degradation process. Atomic hydrogen produced in situ at the catalyst surface can form simultaneously the alcohol RFCH2OH and hydrofluoric acid.
In this work, different cathodes have been used, finding that its selection plays a key role in PFOA degradation. In this sense, Pt acts as an electrocatalyst because of its higher capacity to produce in situ atomic hydrogen, which seems efficient in hydrodefluorination. It has been also demonstrated that working at low electrolyte concentration (3.5 mM Na2SO4), complete PFOA removal can be reached with up to 76.1% TOC abatement and 58.6% defluorination working at the natural pH of the solution (pH0: 4). The kind of electrolyte employed did not have a significant impact on the overall reaction. Still, slightly better results were achieved using sulfate because of the generation of sulfate radicals. Regarding the influence of the starting pH, higher TOC removal was obtained working at pH0: 9, while at higher pH values PFOA mineralization was hindered. When comparing the results obtained in this work with those reported in literature, it must be remarked that the employed BDD-Pt system allows a higher defluorination degree with a lower energy consumption. In view to render the process economically viable to treat dilute solutions, further experiments are planned to combine the electrochemical process with a preconcentration step (such as filtration or adsorption).
**Supplementary Materials:** The following are available online at http://www.mdpi.com/2073-4344/10/8/902/s1, Figure S1. Damaged Pt cathode after high temperature PFOA electrooxidation (T: 80 ◦C).
**Author Contributions:** A.L.G.C.: conceptualization, investigation, data curation, methodology, writing: original draft. J.A.Z.: supervision, writing: review. A.S.: supervision, data curation, validation, writing: review. K.G.S.: formal analysis, supervision, validation, writing: review. J.A.C.: formal analysis, funding acquisition, supervision, validation, writing: review. All authors have read and agreed to the published version of the manuscript.
**Funding:** This research was funded by the Spanish Ministerio de Ciencia, Innovación y Universidades through project CTM2016-76454-R and by Comunidad de Madrid by P2018/EMT-4341 REMTAVARES-CM.
**Acknowledgments:** Authors thank the funding received from Ministerio de Ciencia, Innovación y Universidades through research project CTM2016-76454-R and Comunidad de Madrid for P2018/EMT-4341 REMTAVARES-CM. Alicia L. Garcia-Costa would like to thank Campus France for the mobility grant under the Make Our Planet Great Again (MOPGA) program and the Spanish Ministerio de Ciencia, Innovación y Universidades for mobility grant EST2019-013106-I. She would also like to thank both the Spanish Ministerio de Economía y Competitividad and the European Social Fund for the PhD grant BES-2014-067598.
**Conflicts of Interest:** The authors declare no conflict of interest.
#### **References**
© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
### *Article* **Simulated Ageing of Crude Oil and Advanced Oxidation Processes for Water Remediation since Crude Oil Pollution**
**Filomena Lelario 1, Giuliana Bianco 1, Sabino Aurelio Bufo 1,2,\* and Laura Scrano <sup>3</sup>**
**Abstract:** Crude oil can undergo biotic and abiotic transformation processes in the environment. This article deals with the fate of an Italian crude oil under simulated solar irradiation to understand (i) the modification induced on its composition by artificial ageing and (ii) the transformations arising from different advanced oxidation processes (AOPs) applied as oil-polluted water remediation methods. The AOPs adopted were photocatalysis, sonolysis and, simultaneously, photocatalysis and sonolysis (sonophotocatalysis). Crude oil and its water-soluble fractions underwent analysis using GC-MS, liquid-state 1H-NMR, Fourier transform ion cyclotron resonance mass spectrometry (FT-ICR-MS), and fluorescence. The crude oil after light irradiation showed (i) significant modifications induced by the artificial ageing on its composition and (ii) the formation of potentially toxic substances. The treatment produced oil oxidation with a particular effect of double bonds oxygenation. Non-polar compounds present in the water-soluble oil fraction showed a strong presence of branched alkanes and a good amount of linear and aromatic alkanes. All remediation methods utilised generated an increase of C5 class and a decrease of C6-C9 types of compounds. The analysis of polar molecules elucidated that oxygenated compounds underwent a slight reduction after photocatalysis and a sharp decline after sonophotocatalytic degradation. Significant modifications did not occur by sonolysis.
**Keywords:** crude oil; photocatalysis; sonolysis; sonophotocatalysis; FT-ICR/MS; Kendrick plot; van Krevelen diagram; water; pollution; remediation
#### **1. Introduction**
The composition of petroleum crude oil varies widely depending on the source and processing. Oil is a complex organic mixture counting for a high number of chemically distinct components, including unsaturated and saturated hydrocarbons, hetero-atoms (such as N, S, and O) and a minor percentage of metals predominantly vanadium, nickel, iron, and copper. Many oil constituents can be carcinogens, neurotoxins, respiratory irritants, hepatotoxins, nephrotoxins, and mutagens. Their toxic effects can be acute and chronic, causing many direct symptoms and major long-term injuries, including reproductive problems and cancer [1].
The hydrocarbon fraction can be as high as 90% by weight in light oils, compared to about 70% in heavy crude oil. A majority of the heteroatomic free constituents are side-byside paraffinic chains, naphthalene rings, and aromatic rings. Heteroatomic compounds constitute a relatively small portion of crude oils, less than 15%. However, they have significant implications since their presence, composition, and solubility, which depend on the origin of the crude oil, can cause either positive or negative effects in the transformation processes and are of environmental concern [2,3].
A significant consideration of the several processes affecting the crude oil spilt into the environment is needed to clarify the effects of increasingly widespread harmful events and
**Citation:** Lelario, F.; Bianco, G.; Bufo, S.A.; Scrano, L. Simulated Ageing of Crude Oil and Advanced Oxidation Processes for Water Remediation since Crude Oil Pollution. *Catalysts* **2021**, *11*, 954. https://doi.org/ 10.3390/catal11080954
Academic Editor: Fernando J. Beltrán Novillo
Received: 13 July 2021 Accepted: 4 August 2021 Published: 10 August 2021
**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.
**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).
predict the future fate of the oil. For this reason, the awareness of such phenomena will prove to be a valuable resource in the effort to develop innovative remediation technologies. The ecological impact of oil contamination in different environmental sections (marine, terrestrial and atmospheric) is a source of severe concern. Extraction techniques, transportation and refinery treatments of crude oil can originate pollution phenomena due to the dispersion of these compounds everywhere. These problems have attracted significant attention to understanding the fate of oil in the environment and the natural mechanisms of oil degradation and transformation to suggest a method to reduce the damages caused by original and derivative products [4–6].
As all xenobiotic substances, crude oil undergoes biotic (biotransformation by aquatic organisms such as algae, bacteria) and abiotic (hydrolysis, oxidation, photodegradation) processes, giving rise to many derivatives. In the same way as their parent molecules, these transformation products can lead to the contamination of terrestrial and aquatic environments due to oil deposition on soil and into the surface- and ground-water. Nevertheless, they can be more persistent and toxic than the parent compounds [1–3].
Extensive literature is already available on the microbiological degradation of crude oil, which received considerable attention from researchers. For example, since 1975, the biodegradability of crude oils has been studied and found to be highly dependent on their composition and incubation temperature [5]. Researchers also examined the ability of microorganisms to degrade a high number of hydrocarbons of a different structure in petroleum [6]. Furthermore, many authors have elucidated that the lighter fractions can undergo degradation more rapidly than the heavier ones, e.g., n-alkanes degraded more quickly than branched alkanes, and aromatics with two to three rings readily biodegraded through several pathways [4–8].
Photochemical processes are also essential contributors to pollutants' degradation and the removal of exogenous substances from the environment [9,10], especially in tropical and sub-tropical climates. In those areas, solar irradiation intensity is high, and the lack of nutrients hinders biological processes. Moreover, photochemical reactions are the primary cause of the compositional change of crude oil spilt in a marine environment [11–13]. Photolysis plays an essential role in the mousse formation that begins a few moments after an oil spill [12]. Due to sunlight, the interfacial tension of a crude oil film rapidly decreases, and chocolate mousse starts to form, which leads to the stabilisation of the waterin-oil emulsions [13,14]. The formation of emulsions seems to depend on the amount of asphaltene present in the oil film, and researchers reported that this amount increases upon irradiation [13]. Moreover, an increase in emulsion viscosity occurs due to the structural organisation of the asphaltenes [14].
The oxidised products resulting from the photochemical transformation significantly affect the viscosity, mousse formation, and weathered petroleum's physical properties. Moreover, photo-oxidation can lead to the destruction of existing toxic components, the generation of new toxic constituents and the formation of water-soluble products [10–14].
Since crude oil settles on the surface of water and soil, it undergoes solar irradiation. Solar degradation is a natural way for petroleum decontamination, also suggesting that techniques based on light irradiation could be helpful to the petroleum degradation processes. Light irradiation-based technologies have been improved using catalysts, the most effective and cheapest water purification tool being titanium dioxide (TiO2) [15,16]. Researchers have exploited combinations of different advanced oxidation processes (AOPs) for environmental detoxification in the last years, especially for wastewater treatment. The so-called sonophotocatalysis (SPC), the simultaneous use of ultrasound (US) and photocatalysis (PC) by semiconductors to degrade organic pollutants in water (e.g., the effluent of dye works) has been investigated, but combined AOPs methods were not applied to oil-polluted water remediation to our knowledge [17–22].
Among the analytical techniques available for structurally determining crude oil components or metabolites, gas chromatography combined with mass spectrometry (GC-MS) has been the best choice so far and most widely used [23,24]. The fractionation of crude
oil and subsequent GC-MS analysis has characterised nearly 300 components comprising aliphatic, aromatic, and biomarker compounds [25–28]. However, most crude oil fractions remain unidentified since many components cannot be resolved and appear as "hump" or "unresolved complex mixture (UCM)" in GC chromatograms [29,30].
Compositions of the saturated hydrocarbons have been better characterised by twodimensional gas chromatography coupled to mass spectrometry [29] and liquid chromatographymass spectrometry [31]. However, polar species appear poorly resolved due to their compositional complexity far exceeding the peak capacity of typical analytical techniques. High mass resolving power is necessary for the resolution of many compounds present in crude oil.
The development of Fourier transform ion cyclotron resonance mass spectrometry (FT-ICR MS) had provided the needed ultra-high resolving power (m/Δm**50%** > 100,000, in which Δm**50%** is peak width at half peak height), and the use of electrospray ionization (ESI) mass spectrometry had made possible to detect most polar species. Thus, the coupling of these two techniques, ESI and FT-ICR mass spectrometry, produces a powerful analytical tool for analysing these polar species without a preliminary chromatographic separation [32,33].
This work investigates the modifications that the artificial ageing induced on the composition of the polar fraction of an Italian crude oil (Basilicata region—Southern Italy, Val D'Agri countryside) under solar irradiation. Moreover, it explores the possibility of oil-polluted water remediation using AOPs, such as photocatalysis (UV + TiO2), sonolysis (US, ultrasound irradiation) and the simultaneous use of photocatalysis and sonolysis, i.e., sonophotocatalysis (UV + TiO2 + US).
#### **2. Results and Discussion**
The crude oil sample, collected from the Oil Centre sited in Val D'Agri (Basilicata), underwent simulated solar treatment. Information on the composition of the oil watersoluble fraction obtained through GC-MS, liquid state 1H NMR and FT-ICR-MS was the basis for this investigation. Liquid state 1H NMR spectroscopy accomplished helpful information on the oil composition. This technique recognised amounts of 59%, 19%, and 20% of total hydrocarbons as linear, cyclic (or branched), and aromatic compounds. NMR spectroscopy cannot discriminate branched from cyclic alkanes because both compounds have the same intramolecular environment. Figure S1 compares data obtained by NMR and GC-MS.
The use of real standards introduced by an electrospray ion source allowed the calibration of mass spectra. Recalibration was necessary for the identified homologous series in each sample [34]. A troubling complication in structural studies of crude oil has been its enormous complexity on a molecular scale. The ultrahigh-resolution of FT-ICR spectra can be highly complex: these spectra typically comprise many peaks at each "Nominal mass" and thousands of peaks in a whole spectrum. Each peak could represent a chemically diverse compound. This complexity poses an investigative challenge to the study of spectra for structural interpretation.
The univocal assignment of elementary composition, merely based on the high resolution and accuracy of the instrument, is not possible for all mass values. For values higher than 400–500 Da, it is necessary to validate the result differently. The Kendrick plot (Kendrick mass defect vs Kendrick Nominal mass or KMD vs KNM) offers an outstanding vehicle to visualise and categorise all of the peaks in a mass spectrum. Kendrick mass defect (KMD) breakdown has been effectively applied to ultra-high resolution mass spectra, consenting to categorise peaks into complex spectra based on their homologous similarities across a selected type of masses [35]. Bi-dimensional plots can discern compounds differing by masses associated with a structural unit (e.g., CH2, COOH, CH2O, etc.). In this drawing, the signals of structurally related moieties all lie on horizontal or diagonal straight lines. Such a method permits the extraction of peaks that are homogeneously associated. The method can effectively recognise groups of associated compounds in FT-ICR-MS of
petroleum samples [36]. The compounds of the same homologous series (having a different number of groups CH2) will fall in a single horizontal line of the diagram (KNM), with peaks separated from 14 Da and no difference of KMD.
Similarly, the signals relating to compounds of the same class but of different types will occupy points on a vertical line of the diagram, separated by a difference of 0.013 in the Kendrick mass defect. The conversion of mass spectra from the IUPAC mass scale (based on the 12C atomic mass as exactly 12 Da) to the Kendrick mass scale is the first step. The Kendrick mass scale poses CH2 = 14.0000 Da rather than 14.01565 Da. The Kendrick mass comes from the IUPAC mass, as shown in Equation (1) [35,36]:
$$\text{Kendrick mass} = \text{IIUPAC mass} \times \text{(14.00000/14.01565)} \tag{1}$$
Members of a homologous series (specifically, compounds that comprehend the same heteroatom and number of rings plus double bonds, but a different number of CH2 groups) have the same KMDs. They are thus quickly organised and selected from a list of all detected ion masses, as shown in Equation (2):
$$\text{KMD} = \text{KNM} - \text{KEM} \tag{2}$$
where KEM is the Kendrick exact mass.
By rounding the Kendrick mass up to the nearest whole number, the nominal Kendrick mass conveniently arises. Next, homologous series are parted based on even and odd Kendrick Nominal mass and KMD, as described elsewhere [36,37]. Finally, the Kendrick masses are sorted based on Kendrick mass defect and nominal-Z value and exported into an Excel spreadsheet in the second step. Then, a molecular formula calculator programme, limited to molecular formulas consisting of up to 12C 0–80 and 16O 0–10, assigns elemental compositions. Since members of a homologous series diverge only by integer multiples of CH2, the assignment of a single unit of such a series typically suffices to identify all higher-mass members of that series [36].
We also used the van Krevelen diagram for examining ultra-high resolution mass spectra. This kind of layout is used broadly in the geochemistry literature to study the evolution of coals or oil samples [38–40]. The molar hydrogen-to-carbon ratios (H/C) constitute the ordinate, and the molar oxygen-to-carbon ratio (O/C), the abscissa. As a result, each class of compounds plots in a specific location on the diagram. Researchers well recognised that they can identify the type of compounds from the position of their representative points in the van Krevelen plot [41–43].
In general, the chemical formula CcH2c(Z)NnOoSs can identify the crude oil composition. That is because the hydrogen deficiency index <Z> of the molecule is the same for all members of a homologous "type" series (i.e., the fixed number of rings plus double bonds). Every two-units decrease in <Z> value represents the addition of one ring or a double bond. Therefore, number-average molecular weight, *Mn*, and weight-average molecular weight *Mw* have a synthetic definition as:
$$\mathbf{M}\mathbf{n} = \Sigma \mathbf{i}\mathbf{M}\mathbf{i}/\Sigma\mathbf{N}\mathbf{i} \tag{3}$$
and
$$\mathbf{M}w = \Sigma \mathbf{N} \mathbf{i} \mathbf{M} \mathbf{i}^2 / \Sigma \mathbf{N} \mathbf{i} \mathbf{M} \mathbf{i} \tag{4}$$
where Ni is the relative abundance of ions of mass Mi [34].
The <Z> number plays an essential role for the general molecular formula CcH2c(Z)X of the corresponding neutral species, in which X denotes the constituent heteroatom (Nn, Oo, and Ss).
#### *2.1. Ageing Study of Crude Oil by FT-ICR-MS*
A solar simulator (Suntest®), equipped with a xenon lamp as the light source used for the ageing treatment, provided information about the crude oil's photochemical behaviour.
GC-MS spectra showed that the fraction present in the highest percentage shifted from the C8–C11 fractions to the C13 (Figure S2) in the irradiated sample. We observe an increased amount of the C13–C23 and a decreased amount of the C7–C12 fractions. In the natural (not irradiated) oil, the C8–C11 fractions represented 54.8% of all the compounds detected. Figure S2B depicts the distribution of the compounds as a function of their chemical type. The GC-MS analysis of the mixture deriving from solar simulator irradiation showed an increase in the relative amounts of both linear alkanes and aromatic compounds. At the same time, we observed a sharp decrease in the relative amounts of branched chains. After irradiation, we did not find cyclic alkanes and alkenes.
After the irradiation, the compositional analysis of the linear alkanes highlighted several changes (Figure S2C) compared to the not irradiated sample. Undecane was the hydrocarbon found in the highest percentage in the crude oil, while pentadecane was in the irradiated oil. A decrease in the C7–C12 and an increase in C13–C25 fractions is evident. All our analytical determinations agree with reducing the numberof branched alkanes in crude oil after irradiation. Figure S2D illustrates the modifications of the composition of this fraction. After the irradiation, branched alkanes underwent a sharp reduction and only C8, C9, C11, C12, and C13 fractions were present. Cyclic alkanes were not present after the solar simulator experiment (Figure S2E).
The percentage area of the aromatic compounds did not vary with solar irradiation (Figure S2B). However, a sharp decrease in benzene-like structures and an increase in naphthalenic ones have been observed (Figure S2F).
Figure 1a,b show the FT ICR-MS spectra of untreated and treated crude oil, respectively. The spectra show the distribution of multiple ions with a single charge comprised between m/z 150 and m/z 1400. Figure 1(c1–c3) show the scale-expanded segment of the mass spectrum in Figure 1a, revealing an average period of nominal 14 mass units. The signal intensities increased after light irradiation. The shift of maximum apex was not negligible in the treated crude oil sample, which was also more viscous. Thanks to the high accuracy of mass and the excellent resolution power of FT-ICR-MS, it was possible to carry out the non-ambiguous determination of the elementary composition of multiple isobaric picks.
The chemical formula CcH2c(Z)X generally expresses the composition of a hydrocarbon molecule; where, <c> is the number of carbon atoms, <Z> is the hydrogen deficiency (a measure of aromatic character), and X represents the constituent heteroatom (N, S, O) in the molecule. The heteroatom of interest is oxygen in this study. For simplification, Kendrick and van Krevelen diagrams of natural and irradiated crude oil shown in the figures report only the O3 class, which contains the most numerous groups of detected ions. Table 1 illustrates an example of homologous series extracted by the mass spectrum of the untreated sample, with a degree of unsaturation Z = −20 and class of oxygen O3, containing the most numerous groups of detected ions.
The compounds of the same homologous series, having a different number of CH2 groups, fall in a single horizontal line of the Kendrick plot with peaks separated from 14 Da and no difference in the Kendrick mass defect (Table 1, Figure 2). The compounds of the same class but different typology settle down on a vertical line of the diagram separated from a difference of 0.0134 in the value of KMD. Figure 2 compares Kendrick plots for positive-ion ESI FT-ICR mass spectra of natural (-) and irradiated (**X**) crude oil samples. Due to the high number of signals, the figure reports only the O3 class, containing the most numerous group of detected ions. Kendrick plot of crude oil sample for O3 class shows many compounds with a high degree of unsaturation (high value of KMD). In the low values of KMD, the highest percentage of compounds has a small alkylation series (limited number of -CH2- moieties).
**Figure 1.** (**a**) FT ICR-MS spectrum of the untreated oil sample; (**b**) FT ICR-MS spectrum of the same sample irradiated by xenon-lamp. The spectrum shows the distribution of multiple ions with a single charge comprised between m/z 150 and 1400; **c1**, **c2**, **c3** insets = mass scale-expanded segments of the full range crude oil mass spectrum in Figure 1a, revealing periodicities of 14.016 Da from compound series differing in the number of CH2 groups and 2.016 Da from compound series differing in the number of rings plus double bonds.
**Figure 2.** Kendrick mass plot of the O3 species found in natural (-) and irradiated oil (**X**). This plot illustrates the increase in the number of rings plus double bonds as the KMD increases (*y*-axis) and the alkylation series along the *x*-axis.
**Table 1.** Homologous series of O3 class with <Z> = −20.
This plot can visually sort up to thousands of compounds horizontally according to the number of CH2 groups and vertically according to class (heteroatom composition) and type (rings plus double bonds). Since these two classes have the same number of oxygen atoms, they have identical O/C ratios but distinguish themselves by different H/C ratios.
The attained results elucidate the transformation of oil components following irradiation. After irradiation with the xenon lamp (Suntest®), a slight shift of the peak to the higher masses appears in the recorded mass spectra, according to Griffiths et al. findings [31]. Therefore, it seems that a phenomenon of molecular polymerisation prevails on the destruction of the tri-, tetra- and penta-aromatic groups. Furthermore, since the increase in unsaturation correlates with the higher toxicity [44], our results could indicate higher toxicity for the oil after irradiation.
The plot of Figure 2 highlights the increase in the number of double bonds' rings as the KMD increases (*y*-axis) and the alkylation series along the *x*-axis. The solar irradiation causes a diminution of rings or double bonds (picks rarefaction in samples irradiated), a consequent Kendrick Nominal mass raising of 2 Da, and the Kendrick mass defect diminution. The irradiated crude oil sample shows an expansion of alkylation in compounds with a high degree of unsaturation and a reduced unsaturation number for molecules with a low alkylation degree.
Figure 3 shows the van Krevelen plot for the class of O3 compounds found in the natural crude oil. The compounds in homologous series, corresponding to varying degrees of alkylation, appear along lines that intersect the value of 2 on the H/C axis. Similarly, a vertical line connects homologous series differing in degree of unsaturation. In agreement with the results in the Kendrick plot, most compounds have a low number of oxygen atoms and a high degree of unsaturation.
**Figure 3.** van Krevelen plot of the O3 species found in natural (-) and irradiated oil (**X**). The compounds in homologous series, corresponding to varying degrees of alkylation, appear along lines that intersect the value of 2 on the H/C axis. Similarly, a vertical line connects homologous series differing in degree of unsaturation.
As the H/C ratio increases, the number of rings plus double bonds decreases. Thus, a slight shift to a lower H/C ratio (i.e., a higher number of rings plus double bonds) occurred. Figure 3 shows a minor shift of the data to the right due to increased oxidation and slight dehydration (the picks shift to the lower left) of hydrocarbons. Kendrick mass defect analysis has dramatically facilitated the interpretation of mass spectra, but it is still challenging to derive details for molecules that contribute to complex ultrahigh-resolution mass spectra.
Figure 4 shows the distribution of compounds associated with their number of oxygen atoms in natural and irradiated samples. In both samples, the number of total oxygenated compounds increases. The augmentation of oxygenated compounds should mainly refer to the O3 and O4 types present in the investigated model. The irradiation of crude oil in the solar simulator produces oil oxidation with a particular effect of double bonds oxygenation.
**Figure 4.** FT-ICR compositional analysis of natural and irradiated crude oil samples as a function of the number of oxygen atoms.
Figure 5 shows oxygen class Z-distributions for natural and irradiated samples, confirming a diminution of hydrogen deficiency index (~15–30% less) and augmentation of oxygen number after the light irradiation. Therefore, the light irradiation induces a manifest photo-oxidation of the crude oil composition. These results highlight toxicity as most of the new oxidised compounds are water-soluble, available in higher concentrations to the living organisms and probably more reactive and biologically active than their parent compounds [43,44].
**Figure 5.** Oxygen class <Z>-distributions for natural and irradiated samples.
#### *2.2. Remediation of Oil-Polluted Water*
Since crude oil lies over the surface of water and soil, it suffers solar irradiation. Solar degradation is one of the natural ways for petroleum decontamination, and, as a consequence, techniques based on light irradiation could be advantageous in the petroleum degradation processes. Enhanced light irradiation-based technologies are available, adopting different approaches for the scope [10–12,45].
The accidental dispersion of crude oil in water bodies forms a characteristic thin layer of not water-miscible compounds and a deeper layer of solubilised substances, which cannot easily separate from the aqueous solvent. In this direction, our approach was to prepare a water/oil suspension and investigate the efficiency of different cleaning methods. The water-soluble fraction of crude oil was undergone degradation by photocatalysis, sonolysis, and sonophotocatalysis, i.e., the simultaneous use of UV, titanium dioxide, and ultrasound emitter (UV + TiO2 + US). GC-MS, liquid-state NMR, fluorescence, and high-resolution mass spectrometry (FT-ICR) analyses elucidated the chemical nature of water-soluble organic compounds after degradation processes and liquid-liquid extractions (LLEs). The results obtained in this study are concisely readable in Table 2.
**Table 2.** Synthetic results obtained from the different photodegradation processes of crude oil and oil water-soluble fraction (WSF) under investigation.
2.2.1. Photocatalytic Degradation
In the photocatalytic process, the water/oil suspension was treated for 1 h with UV irradiation in the presence of titanium dioxide. GC-MS analysis of WSF (Figure S3) evidenced increased C5 compounds from 67% in not-treated WSF to 89% in the irradiated sample. Moreover, the amount of C6, C7, C8, and C9 compounds decreased. The analysis of chemical classes occurring in the irradiated WSF showed increased branched and cyclic
alkanes, from 50% to 65% (branched) and from 4% to 7% (cyclic), respectively. On the other hand, the number of linear alkanes underwent a slight decrease (from 22% to 14%), and the aromatic compounds had a sharp decline (from 23% to 13%).
1H-NMR spectra (Figures S4 and S5) confirmed a slight increase of linear and cyclic alkanes, and a sharp decrease in aromatics, as evidenced in the chromatographic analysis.
FT-ICR MS analysis showed a minor decrease in the total number of oxygenated compounds. The O1 and O2 classes prevailed over the other types (Figure 6).
**Figure 6.** FT-ICR MS analysis of natural and 1-h photocatalysed WSF of crude oil as a function of the number of oxygen atoms.
Comparison of Kendrick plots constructed for the untreated sample (Figure 7a) and the photodegraded model (Figure 7b) shows an increase in the number of compounds with low molecular weight and low degree of unsaturation. The formation of several homologous series, with KDM values of 0.124, 0.137, 0.150, and so on, is underlined in the Kendrick diagram plotted for the treated sample. The unsaturation degree of these homologous series falls in the range Z = −16 to Z = −20.
**Figure 7.** Kendrick mass plots of the O1-O10 species found in the untreated crude oil WSF (**a**) and the 1-h photocatalysed sample (**b**).
The van Krevelen diagram (Figure 8) shows an increase in O1 class, reduced O/C ratio, and decreased unsaturated compounds in the treated sample compared to the untreated one. After photocatalysis, the number of compounds with a low number of oxygen atoms increased. As shown in Figure 6, the O1 and O2 classes prevailed over the other types, resulting in a decrease in the O/C ratio.
**Figure 8.** The van Krevelen plots of the O1–O10 species found in the untreated crude oil WSF (**a**) and the 1-h photocatalysed sample (**b**).
From fluorescence spectra (Figure S6), it was possible to argue aromatic compounds' decrease after 1-h photocatalytic treatment. Essentially, the absolute intensity of the peak at 347 nm decreases from 92.07 mAU for the natural sample to 49.50 mAU for the treated sample, with a reduction of 46%.
#### 2.2.2. Ultrasonic Irradiation
In the sonolytic process, the water/oil suspension received 1-h ultrasound irradiation. GC-MS analysis (Figure S7) indicated that C5 compounds increased from 67% to 91% at the end of sonolysis, evidencing a behaviour analogous to photocatalysis. Furthermore, the C6, C7 and C8 compounds decreased, similarly to the photocatalytic process; otherwise, the C9 class disappeared. The analysis of functional groups evidenced that branched alkanes increased from 50% to 54% in the sonolysed WSF (from 50% to 65% in photocatalysis). Cyclic alkanes underwent a minor increase from 4% to 5% (like photocatalysis), but aromatic compounds slightly increased from 23% to 24% (decreased dramatically to 13% in photocatalysis). The number of linear alkanes decreased from 22% to 17% (22% to 14% in photocatalysis).
Liquid-state 1H-NMR spectra (Figures S8 and S9) evidenced a relatively equal amount of the three classes of compounds in the not-treated and sonolysed samples. In conclusion, no significant differences emerged in the composition of WSF before and after the processes of sonication and photocatalysis, with a unique exception for aromatic compounds, as mentioned above in the case of photocatalysis.
From FTICR MS analysis, the total number of oxygenated compounds registered a low increase in the sonicated sample. O1, O2 and O7 classes increased, but the other oxygenated types decreased (Figure 9).
**Figure 9.** FT-ICR analysis of natural and 1-h sonicated WSF of crude oil samples asa function of the number of oxygen atoms.
Analysis of the Kendrick plot (Figure 10) after the degradation treatment shows a sharp increase in the number of compounds with low molecular weight. Seventy percent of compounds stay in the range m/z 159–597, and many homologous series with a high degree of unsaturation are visible.
**Figure 10.** Kendrick mass plot of the O1–O10 species found in the not-treated WSF of crude oil (**a**) and after 1-h of US treatment (**b**).
The van Krevelen diagram (Figure 11) substantiates any differences between the natural and treated WSF samples.
The fluorescence study (Figure S10) confirms that the decrease of aromatic compounds is not so evident with US treatment. After 1-h of the sonolytic process, the absolute intensity of the maximum peak displays an insignificant drop from 99.96 mAU for the natural sample to 93.37 mAU for the treated one.
**Figure 11.** The van Krevelen plots of the O1-O10 species found in the not-treated WSF of crude oil (**a**) and after 1-h US treatment (**b**).
2.2.3. Sonophotocatalytic Degradation
The contemporary use of UV irradiation, titanium dioxide and ultrasound irradiation to treat the oil aqueous suspension shows results mainly similar to those obtained with sonolysis or photocatalysis.
GC-MS analysis (Figure S11) demonstrated that after 1-h of treatment, C5 compounds increased from 67% in not-treated WSF to 91% in the treated sample, while C6, C7 and C8 compounds decreased; C9 compounds were not detected (like the simple US). The analysis of functional groups in the sonophotocatalytic degradation evidenced an increase from 50% to 64% of branched alkanes (similar to photocatalysis) and from 4% to 9% of cyclic alkanes (higher than the other technologies). The number of linear alkanes underwent a slight decrease (from 22% to 19%, similar to the other technologies). In comparison, aromatic compounds showed the sharpest decline (from 23% to 7%), also proved by integrating NMR spectra (Figures S12 and S13). In the natural WSF, the aromatics alkanes occupied 19% of the whole spectral area, whilst in the treated sample, this amount decreases up to 3.3%. On the other hand, the amount of linear and cyclic alkanes increases by about 7–8%.
Figure 12 shows the trend of oxygenated compounds after sonophotocatalytic treatment. In this case, the total number of oxygenated compounds decreased from 1203 (not treated WSF) to 993 (treated WSF).
**Figure 12.** FT-ICR MS analysis of natural and 1-h sonophotocatalysed WSF sample of crude oil as a function of the number of oxygen atoms.
Comparison of Kendrick plots (Figure 13) obtained for the not-treated and treated samples showed an increased number of compounds with low molecular weight, especially in the range m/z 169–369, and compounds with a low unsaturation degree.
**Figure 13.** Kendrick mass plots of the O1-O10 species found in the not-treated WSF of crude oil (**a**) and after 1-h of sonophotocatalytic treatment (**b**).
The van Krevelen diagram (Figure 14) let us see an intensification of signals relative to oxygenated compounds with an O/C ratio in the range 0.10–0.25.
**Figure 14.** The van Krevelen plots of the O1-O10 species found in the not-treated WSF of crude oil (**a**) and after 1-h of sonophotocatalytic treatment (**b**).
Fluorescence spectra (Figure S14) substantiated the decreasing of aromatic compounds after 1-h sonophotocatalytic treatment. The absolute intensity of the peak at 347 nm decreased from 89.92 mAU for the natural sample to 47.95 mAU for the treated sample, reducing by 48%.
#### **3. Materials and Methods**
#### *3.1. Crude Oil and Chemicals*
The director of the Eni-Cova Oil Plant in Val d'Agri (Basilicata Region, Southern Italy) kindly provided the oil sample taken from the first step of oil purification after extraction, including dehydration and degasification. Table 3 accounts for the elemental composition reported in the label accompanying the sample delivered for this research.
**Table 3.** Elemental composition (%) of the oil sample taken from the first step of oil purification after extraction, including dehydration and degasification a.
<sup>a</sup> Metals (Ni and V) < 1000 ppm. <sup>b</sup> Obtained as the complement to 100.
All chemicals used were of analytical grade. TiO2 Degussa P-25, obtained as a gift from Evonik (Hanau, Germany), was the catalyst adopted. Table 4 reports a summary scheme of the investigation executed.
**Table 4.** Experiments performed and analytical methods used in this study.
#### *3.2. Photodegradation Apparatus*
The Suntest CPS+ (Heraeus Industrietechnik GmbH, Hanau, Germany), equipped with a xenon lamp of 1.1 kW, protected employing a quartz plate (total passing wavelength: 300 nm < λ < 800 nm), was the solar simulator adopted for photochemical reactions. The temperature of the irradiation chamber was 25 ◦C, maintained through both a thermostatic bath and a conditioned airflow. During the experiments, the crude oil samples were kept up in the horizontal position, creating a homogeneous film of 0.5 cm thickness.
#### *3.3. Photodegradation Process and Sample Preparation for ESI FT-ICR MS*
The protocol used for oil ageing experiments was: (i) the irradiation of the natural crude oil (10 mL) for a week in the borosilicate planar reactor; (ii) crude oil samples preparation by dissolving ~30 mg of material in 30 mL of toluene; (iii) withdrawal of 1 mL solution and its dilution with 0.5 mL methanol; addition of either 10 μL acetic acid (for positive ion ESI) or 10 μL ammonium hydroxide (for negative ion ESI) to facilitate protonation or deprotonation in the electrospray ionisation process, respectively.
#### *3.4. Ultrasonic Irradiation of WSF Samples*
A crude oil/water suspension was arranged in a borosilicate decanter (5 L) equipped with a Teflon tap at the bottom. The decanter was filled with 3.5 L of ultrapure water; the crude oil was added at the ratio of 1/20 (oil/water), and the suspension was magnetically stirred and then kept in the dark for 30 days at constant temperature (25 ◦C) to reach equilibrium and the separation of oil phase on the surface of the aqueous phase. Aqueous samples were drawn off through the Teflon tap without disturbing the oil/water separation surface. The collected aqueous sample (500 mL) underwent cotton filtration, to avoid the formation of an emulsion in the solution. The ultrasonic degradation tool was the immersible ultrasonic emitter Sinaptec Nexus P198-R (Sinaptec, Lezennes, France), an ultrasonic module furnished with a titanium sonotrode (S23-10-1/2, Sinaptec), an electrical signal of frequency close to 20 kHz, and a voltage of about 1 kV. In this configuration, the electric power provided by the generator (Nexus P198-R, same manufacturer) is adjustable between 7 W and 100 W, as indicated on a digital display panel. However, this electrical measurement does not determine with high precision the acoustic power dissipated in the liquid. The experimental temperature was fixed to 25 ◦C.
#### *3.5. Photocatalytic and Sonophotocatalytic Degradation of WSF Samples*
The photocatalytic method to degrade the water-soluble fraction of crude oil utilises a 125 W high-pressure mercury lamp (Philips-HPK, Philips, Turnhout, Belgium) that provides its maximum energy at 365 nm, with a range of emission from 195 to 580 nm. The catalyst was TiO2 (80% anatase–20% rutile). The simultaneous use of the mercury lamp, titanium dioxide and the ultrasound emitter (UV + TiO2 + US) permitted the sonophotocatalytic degradation. The experimental temperature was 25 ◦C for photocatalysis and sonophotocatalysis trials.
#### *3.6. Liquid–Liquid Extraction (LLE)*
The experimental protocol was (i) to collect samples of the oil WSF after 15, 30, 45, and 60 min of treatment; (ii) to extract in triplicate 30 mL of each sample in a separatory funnel (50 mL) with 3 mL dichloromethane for GC-MS analysis and (iii) another 30 mL with the same solvent for 1H-NMR spectroscopy; (iv) to perform fluorescence analysis using 5 mL of the aqueous solution without extraction.
The internal standard used for assessing the reproducibility of WSF extraction was 1 mL of 1,3-dibromopropane (26.7 mg L−1) added to the volume of dichloromethane needed for each liquid–liquid extraction. In addition, it was necessary to add 1.0 mL of 1-bromododecane (29.0 mg L<sup>−</sup>1) to evaluate the GC-MS analysis reproducibility at the end of each extraction. Thus, the injection volume was 1 μL extract for each GC-MS run.
#### *3.7. Analysis of Fluorescence*
A research-grade spectrofluorometer FP-6500 Jasco (Jasco Corporation, Cremella, Italy) was available for fluorescence analysis. This analysis was necessary to appreciate the aromatic compounds remaining in the water-soluble fraction of crude oil. The spectrofluorometer FP-6500 Jasco, adopting as emitting source a DC-powered 150 W xenon lamp (in a sealed housing), employs a photometric rationing system, which utilises a second photomultiplier tube to monitor and compensate for any variations in the intensity of the xenon source, thus ensuring maximum analytical stability. Furthermore, a concave holographic grating monochromator with optimised blaze angles provides maximum sensitivity over the entire wavelength range; (220–750 nm; 1 nm resolution).
The fluorescence optical path adopted was 1 cm in quartz cells (volume ca 5 cm3) at 237 and 320 nm excitation and 347 and 360 nm emission.
#### *3.8. 1H-NMR Analysis*
A Varian Oxford AS400 spectrometer (Palo Alto, CA, USA), operating at 400 MHz, was enough for the 1H-NMR spectra recording. The set temperature for the used 5 mm non-gradient broadband inverse (BBI) probe was 25 ◦C.
All the 1H-NMR spectra have tetramethylsilane (TMS) as reference under the acquisition parameters shown in Table 5.
In degradation experiments, liquid–liquid extraction with chloroform permitted to isolate the water-soluble fraction of crude oil. After the complete evaporation of the solvent in a rotary evaporator, the addition of 500 μL deuterated chloroform (CDCl3) permitted to recuperate the residual organic mixture.
**Table 5.** 1H-NMR acquisition parameters.
*3.9. Mass Spectrometry of Polar Components*
The instrument available to determine polar components was the micro ESI/FT-ICR/MS 7 T Thermo Electron (Waltham, MA, USA).
The method used for the routine analyses permitted a mass accuracy better than 2 ppm by external calibration, using the mixture of caffeine, MRFA, and Ultramark. The technique separated more than 6000 ion signals belonging to chemically different elemental compositions with a 200,000 resolving power (m/Δm**50%** at *m*/*z* 400) in positive electrospray mode. The robustness of this equipment, combined with unprecedented ease of use, ultra-high mass accuracy, high sensitivity, and excellent resolving power, make it an ideal instrument for analysis. The infusion of the samples at a flow rate of 5 μL/min permitted the best result in terms of spectrum resolution. ESI conditions were: needle voltage, +4.5 kV; heated capillary current 4 A; tube lens voltage 135.12; temperature 300 ◦C; N2 speed 2.33 u.a.; aux gas flow rate 0.73; scansions per second 1000.
#### **4. Conclusions**
In this study, we tried to characterise the ageing process of crude oil simulating solar irradiation on a thin layer of an oil sample. As a result, FT-ICR MS evidenced an augmentation of compounds with low molecular weight and a slight increase of the number of oxygen atoms in the oxygenated species, as depicted in Kendrick and van Krevelen diagrams. Furthermore, the simulated ageing produced the oxidation of crude oil with a particular effect on double bonds' oxygenation, as confirmed by the disappearance of alkenes in gas chromatographic analysis. The observed results seem to be recognisable because the energy irradiated with the xenon lamp could be enough for catalysing the reaction of olefins with the atmospheric oxygen following a bridge mechanism.
We experimented with different solutions for the cleaning treatment of oil-polluted water (photocatalysis, sonolysis, and sonophotocatalysis). GC-MS analyses of the watersoluble fraction of crude oil for both natural and treated samples discovered that only a few compounds are detectable in the aqueous solution, principally C5-organic chains (~50%). Low amounts of C6, C7, C8 and C9 chains were also present. Both GC-MS and liquid state 1H-NMR signals showed that the branched alkanes were the principal chemical class in the soluble fraction of oil, followed by a small amount of linear and aromatic alkanes.
With all the degradation methods utilised, an increase of the C5-class and a decrease of C6–C9 types of compounds was evident. Furthermore, the FT-ICR comparative analyses of oxygenated species elucidated that the total number of O-compounds in the treated WSF samples is different for all of the degradation methods experimented. The number of the oxygenated compounds slightly decreased with photocatalysis compared to the non-treated sample. An opposite trend appeared with the sonolysis treatment. The sonophotocatalytic method showed a sharp reduction in the number of oxygenated compounds, probably due to the volatilisation of small molecules formed during the oxidation process. It is conceivable that ultrasound can promote this volatilisation. The degradation of the watersoluble fraction of crude oil performed with photocatalysis and sonophotocatalysis led to an apparent decrease of aromatic compounds of 46% and 48%, respectively, for the two techniques, as also confirmed by the fluorescence analysis. With the use of sonolysis, there was no effect on the number of aromatic compounds. Nevertheless, all the degradation methods applied were capable of increasing the number of cyclic alkanes. Therefore, we could speculate that ultrasound in sonophotocatalytic technology can affect the rate of the
photocatalytic degradation of the organic pollutants due to a synergistic effect typically observed with an increase of the degradation process efficiency.
In conclusion, our results confirm the photo-oxidation effect caused by light irradiation either on crude oil (simulated ageing) or on the soluble oil fraction. Naturally, the behaviour of each oil type could be different, and then it is not possible to generalise our findings to all cases of oil spilling and environmental remediation. Therefore, it is necessary to check case by case before reaching specific solutions for more efficient remediation processes to avoid making the situation worse.
**Supplementary Materials:** The following are available online at https://www.mdpi.com/article/10 .3390/catal11080954/s1 Figure S1: Percentage of compounds in crude oil as recognized by GC-MS (blue column) and 1H-NMR (red column). Figure S2: GC-MS compositional analysis of crude oil (blue column) and solar simulator irradiated crude oil (red column), as a function of the number of carbon atoms (A); composition of crude oil as a function of the type of compounds, LH: linear aliphatic hydrocarbons; BH: branched aliphatic hydrocarbons; CH: cyclic aliphatic hydrocarbons; AH: aromatic hydrocarbons; AL: alkenes (B); composition of the linear aliphatic hydrocarbons as a function of the number of carbon atoms (C); composition of the branched aliphatic hydrocarbons fraction as a function of the number of carbon atoms (D); composition of the cyclic hydrocarbons as a function of the number of carbon atoms (E); composition of the aromatic hydrocarbons fraction as a function of the number of carbon atoms (F). Figure S3: GC-MS compositional analysis of WSF crude oil before (red column) and after (blue column) photocatalysis: distribution of hydrocarbons as a function of the number of carbon atoms (A) and distribution of the compounds as a function of chemical species (B). Figure S4: 1H-NMR spectra of WSF crude oil before (A) and after (B) photocatalysis. Figure S5: 1H-NMR compositional analysis of WSF crude oil before (red column) and after (blue column) photocatalysis: distribution of the compounds as a function of chemical species. Figure S6: Fluorescence spectra of WSF crude oil before (blu line) and after (red line) photocatalysis. Figure S7: GC-MS compositional analysis of WSF crude oil before (red column) and after (blue column) sonolysis: distribution of hydrocarbons as a function of the number of carbon atoms (A) and distribution of the compounds as a function of chemical species (B). Figure S8: 1H-NMR spectra of WSF crude oil before (A) and after (B) sonolysis. Figure S9: 1H-NMR compositional analysis of WSF crude oil before (red column) and after (blue column) sonolysis: distribution of the compounds as a function of chemical species. Figure S10: Fluorescence spectra of WSF crude oil before (blu line) and after (red line) sonolysis. Figure S11: GC-MS compositional analysis of WSF crude oil before (red column) and after (blue column) sonophotocatalysis: distribution of hydrocarbons as a function of the number of carbon atoms (A) and distribution of the compounds as a function of chemical species (B). Figure S12: 1H-NMR spectra of WSF crude oil before (A) and after (B) sonophotocatalysis. Figure S13: 1H-NMR compositional analysis of WSF crude oil before (red column) and after (blue column) sonophotocatalysis: distribution of the compounds as a function of chemical species. Figure S14: Fluorescence spectra of WSF crude oil before (blu line) and after (red line) sonophotocatalysis.
**Author Contributions:** Conceptualisation, F.L., S.A.B. and L.S.; Data curation, F.L., G.B. and S.A.B.; Formal analysis, L.S.; Investigation, L.S.; Methodology, G.B.; Resources, S.A.B.; Writing—original draft, F.L. and L.S.; Writing—review and editing, S.A.B. All authors have read and agreed to the published version of the manuscript.
**Funding:** This research received no external funding.
**Data Availability Statement:** Data supporting results reported in this paper can be found at the Department of Sciences, University of Basilicata, Via dell'Ateneo Lucano 10, 85100, Potenza, Italy. Due to the multitude of files produced for this investigation, a link will be provided under specific requests.
**Acknowledgments:** We acknowledge the Director of ENI-COVA Oil Plant in Val d'Agri (Basilicata Region, Southern Italy), who kindly provided the oil sample. Special thanks go to Mauro Tummolo, and to Jean-Marc Chovelon, University Claud Bernard Lyon 1, for the sonolysis and sonophotocatalysis data.
**Conflicts of Interest:** The authors declare no conflict of interest.
#### **References**
### *Article* **Visible Light Responsive Strontium Carbonate Catalyst Derived from Solvothermal Synthesis**
#### **Pornnaphat Wichannananon 1,2, Thawanrat Kobkeatthawin <sup>2</sup> and Siwaporn Meejoo Smith 2,\***
Received: 24 July 2020; Accepted: 11 September 2020; Published: 17 September 2020
**Abstract:** A single crystalline phase of strontium carbonate (SrCO3) was successfully obtained from solvothermal treatments of hydrated strontium hydroxide in ethanol (EtOH) at 100 ◦C for 2 h, using specific Sr:EtOH mole ratios of 1:18 or 1:23. Other solvothermal treatment times (0.5, 1.0 and 3 h), temperatures (80 and 150 ◦C) and different Sr:EtOH mole ratios (1:13 and 1:27) led to formation of mixed phases of Sr-containing products, SrCO3 and Sr(OH)2 xH2O. The obtained products (denoted as 1:18 SrCO3 and 1:23 SrCO3), containing a single phase of SrCO3, were further characterized in comparison with commercial SrCO3, and each SrCO3 material was employed as a photocatalyst for the degradation of methylene blue (MB) in water under visible light irradiation. Only the 1:23 SrCO3 sample is visible light responsive (Eg = 2.62 eV), possibly due to the presence of ethanol in the structure, as detected by thermogravimetric analysis. On the other hand, the band gap of 1:18 SrCO3 and commercial SrCO3 are 4.63 and 3.25 eV, respectively, and both samples are UV responsive. The highest decolourisation efficiency of MB solutions was achieved using the 1:23 SrCO3 catalyst, likely due to its narrow bandgap. The variation in colour removal results in the dark and under visible light irradiation, with radical scavenging tests, suggests that the high decolourisation efficiency was mainly due to a generated hydroxyl-radical-related reaction pathway. Possible degradation products from MB oxidation under visible light illumination in the presence of SrCO3 are aromatic sulfonic acids, dimethylamine and phenol, as implied by MS direct injection measurements. Key findings from this work could give more insight into alternative synthesis routes to tailor the bandgap of SrCO3 materials and possible further development of cocatalysts and composites for environmental applications.
**Keywords:** strontium carbonate (SrCO3); solvothermal method; photocatalysis; visible light
#### **1. Introduction**
Textile industries employ over 10,000 dyes and pigments in the manufacturing of cotton, leather, clothes, wool, silk and nylon products [1–3]. An estimated 700,000 tons or more of synthetic dyes are thought to be annually discharged into the environment [4], causing serious water pollution as many of these dyes are toxic, highly water soluble and highly stable against degradation by sunlight or increased temperature [5]. Therefore, effective treatments of dye-contaminated water have continuingly received great attention by academic and industrial sectors. Various wastewater treatment methods have been applied to remove toxic dyes from wastewater, such as coagulation–flocculation, adsorption, membrane separation, biodegradation and oxidation processes [6]. Among these methods, photocatalytic oxidation processes have been proven to be simple and effective at organic dye decomposition, forming relatively low toxic by-products with potential mineralization to generate CO2 and H2O [7–9]. In this process, under light irradiation a semiconducting catalyst absorbs photon energy promoting electron transfer from the valence band (VB) to the conduction band (CB), resulting in electron-hole pair generation. The generated holes (h+) further react with water molecules while the electrons (CB) react with oxygen, resulting in formation of active hydroxyl (•OH) and superoxide (•O2 <sup>−</sup>) radicals, respectively. The •OH radicals subsequently attack organic pollutants in water leading to oxidative degradation of pollutants.
Wide bandgap TiO2 [10] and ZnO [11] semiconducting materials have proven to be efficient catalysts for the photo-oxidation of organic pollutants in water. However, these require high-energy ultraviolet irradiation, which requires special and costly safety protocols to be in place for the use of these materials in wastewater treatment. An attractive alternative is to use harmless visible light sources in the photoreactor, employing a visible light responsive photocatalyst for pollutant degradation. Such visible light responsive photocatalysts need to promote the photo-oxidative degradation of pollutants using sunlight (7% UV and 44% visible light emission, and other low-energy radiations [12]) to ensure wastewater treatment is a sustainable process. Strontium carbonate (SrCO3) is a common starting material for the manufacture of colourants in fireworks, glass cathode-ray tubes and computer monitors [13,14]. While commercially available SrCO3 material is commonly derived from celestine (SrSO4) mineral via calcination followed by Na2CO3 treatment (the black ash method) [15], synthetic SrCO3 can be obtained using calcination and wet chemical methods under ambient [16,17] or high-pressure [18] atmospheres. Table 1 summarizes the key features (synthesis conditions, characteristics and the bandgap energy) of synthetic SrCO3 in the literature. Methylene blue, a cationic organic dye and a common colouring agent used in cotton, wood, silk [19] cosmetics, and textile [20] dying is a frequently utilized representative dye pollutant mimicking those present in industrial effluents. Song and coworkers reported effective methylene blue (MB) degradation under visible light irradiation (λ > 400 nm) after 3 h treatment with SrCO3 obtained from the calcination of synthetic Sr(OH)2 [21], while Molduvan and coworkers reported the removal of MB from aqueous solutions using a commercial natural activated plant-based carbon [22]. Other works have utilized SrCO3 as a cocatalyst incorporated in photocatalyst composites, e.g., Ag2CO3/SrCO3 [23], TiO2/SrCO3 [24] and SrTiO3/SrCO3 [25], to expand the photoresponsive range of the material and to improve its catalytic activity and reaction selectivity.
**Table 1.** Synthesis, key characteristics and bandgap energy of synthetic SrCO3.
This work investigated the effects of precursor concentrations (Sr:ethanol mole ratios), solvothermal temperatures and treatment times on the properties of SrCO3 materials and their photocatalytic degradation of MB in water under visible light irradiation, as a function of pH and temperature. Kinetic and mechanistic studies of the MB degradation process were carried out through reaction rate determination and identification of the end-products. The photocatalytic performance of synthesized SrCO3 was compared with that of commercially available material, in order to derive insights into the relationships between properties and catalytic activity.
#### **2. Results and Discussions**
#### *2.1. E*ff*ects of Synthesis Conditions*
Solvothermal treatments of strontium nitrate in ethanol (EtOH) were carried out at various temperatures (80, 100 and 150 ◦C), treatment times (0.5, 1, 2 and 3 h) and Sr:EtOH mole ratios (1:13, 1:18, 1:23 and 1:27). From powder X-ray diffraction (PXRD) results in Figure 1, a single phase of SrCO3 was obtained from two conditions: 2 h solvothermal treatment at 100 ◦C using a Sr:EtOH mole ratio of 1:18 or 1:23. These samples are denoted as 1:18 SrCO3 and 1:23 SrCO3 in further discussions. Notably, mixed phases of SrCO3 and hydrated strontium hydroxides (Sr(OH)2·xH2O, where x is the number of molar coefficient of water in strontium hydroxide solid) were obtained from all other synthesis conditions (results shown in Supplementary Materials: Figures S1 and S2). Typical diffraction peaks correspond well with (110), (111), (021), (002), (012), (130), (220), (221), (132) and (113) orthorhombic SrCO3 lattice planes [34,36], whereas other diffraction peaks match with those of previously reported Sr(OH)2·H2O [37] and Sr(OH)2·8H2O phases [38]. The formation of Sr(OH)2 xH2O is possibly due to adsorbed alcohol, promoting the addition of OH functional groups on the solid surface [39], upon solvothermal crystallization of Sr-containing products.
**Figure 1.** PXRD patterns of Sr-containing samples derived from 2 h solvothermal treatments of hydrated strontium hydroxide in ethanol at various Sr:EtOH mole ratios (1:13, 1:18, 1:23 or 1:27) at 100 ◦C.
FTIR spectra of the prepared Sr-containing samples are shown in Figure 2. The absorption bands located within 1700–400 cm−<sup>1</sup> regions were attributed to the vibrations in CO3 <sup>2</sup><sup>−</sup> groups. The strong broad absorption at 1470 cm−<sup>1</sup> was considered to be due to an asymmetric stretching vibration, and the sharp absorption bands at 800 cm−<sup>1</sup> and 705 cm−<sup>1</sup> can be specified to the bending out-of-plane vibration and in-plane vibration, respectively. The weak peak at 1770 cm−<sup>1</sup> indicated a combination of vibration modes of the CO3 <sup>2</sup><sup>−</sup> groups and Sr2<sup>+</sup>. The sharp peak at 3500 cm−<sup>1</sup> was assigned to the stretching mode of –OH- in Sr(OH)2, and the broad absorption peak around 2800 cm−<sup>1</sup> was assigned to the stretching mode of H2O in Sr(OH)2·H2O and Sr(OH)2·8H2O. These results are consistent with the commercial SrCO3 and the 1:18 SrCO3 and 1:23 SrCO3 samples being of similar chemical composition.
**Figure 2.** FTIR spectra of Sr-containing samples derived from 2 h solvothermal treatment of hydrated strontium hydroxide in ethanol at various Sr:EtOH mole ratios (1:13, 1:18, 1:23 or 1:27) at 100 ◦C.
SEM images of the obtained SrCO3 materials (derived from Sr:EtOH mole ratios of 1:18 or 1:23) are compared with those of commercial SrCO3 in Figure 3. Whisker-like SrCO3 and spherical particles were obtained under these respective synthesis conditions. Figure 3c highlights the relatively large rod-like particles of commercial SrCO3. Variation in particle sizes was observed in solvothermally obtained SrCO3, with particle sizes being smaller for the 1:18 SrCO3 samples. Notably, commercial SrCO3 contains much larger particles than those of the synthesized material. From literature [26,27], SrCO3 production plants utilize two common methods, the black ash method and the soda method, in conversion of celestine ore (SrSO4) to SrCO3 (Table 1). The black ash method involves high-temperature calcination of the ore to obtain SrS, with crystalline SrCO3 solid being formed after dissolving the SrS in aqueous Na2CO3, followed by precipitation. The soda method produces SrCO3 through the two-step decomposition reaction between celestine and aqueous Na2CO3, to obtain precipitated SrCO3. From this information, as the formation of commercial SrCO3 does not require high temperatures (>150 ◦C) for solvent evaporation and precipitation of SrCO3, the larger grain size of the commercial SrCO3 sample is probably due to the fast solvent evaporation during the precipitation processes.
**Figure 3.** SEM images of SrCO3 derived from 2 h solvothermal treatment at 100 ◦C, using Sr:EtOH mole ratios of (**a**) 1:18 and (**b**) 1:23 compared with (**c**) commercial SrCO3.
Thermogravimetric analysis (TGA) plots (Figure 4) suggest thermal stability of all SrCO3 samples up to 600 ◦C. Slight weight loss (<1%) was likely due to moisture or solvent residue [40]. The 1:23 SrCO3 sample gives a relatively high weight loss of 0.21%, which corresponds to the removal of surface adsorbed moisture and ethanol (weight loss upon heating up to 400 ◦C) and the loss of ethanol from the SrCO3 lattice at ca. 450 ◦C. Decomposition of SrCO3 takes place at temperatures above 800 ◦C as a result of conversion to SrO.
**Figure 4.** Thermogravimetric analysis (TGA) plots of SrCO3 samples prepared using Sr:EtOH mole ratios of 1:18 and 1:23, compared with that of commercial SrCO3.
Based on the PXRD and TGA results, chemical transformation of hydrated strontium hydroxide in the presence of ethanol under solvothermal treatments leads to the formation of SrCO3 and ethanol incorporated SrCO3 materials, as proposed by the reactions below. In general, CO2 in air can react with strontium hydroxide to form SrCO3, which precipitates after the sonication step and solvothermal treatments. Ethoxide could be formed under basic conditions, resulting in an CH3CH2O···Sr2+···OCH2CH3 intermediate, which is subsequently transformed to ethanol incorporated in SrCO3. Note that the amount of ethanol incorporated within the SrCO3 is sufficiently low, such that a single phase of SrCO3 was observed in PXRD pattern of the 1:23 SrCO3 sample.
$$\text{Sr(OH)}\_{2} + \text{CO}\_{2} \rightarrow \text{SrCO}\_{3} + \text{H}\_{2}\text{O}$$
$$\text{Sr(OH)}\_{2} + \text{CO}\_{2} \rightleftharpoons \text{Sr}^{2+} + \text{HCO}\_{3}^{-} + \text{OH}^{-}$$
$$\mathrm{HCO\_3^- + OH^- \to CO\_3^{2-} + H\_2O}$$
$$\mathrm{CH\_3CH\_2OH + OH^- \rightleftharpoons CH\_3CH\_2O^- + H\_2O}$$
$$\mathrm{Sr(OH)\_2 + H\_2O \rightleftharpoons Sr^{2+} (aq) + OH^- (aq)}$$
$$\mathrm{Sr^{2+} + 2CH\_3CH\_2O^- \to Sr^{2+} \cdots 2OCH\_2CH\_3}$$
$$\mathrm{Sr^{2+} \cdot \mathrm{OCH\_2CH\_3} + HCO\_3^- \to SrCO\_3 \cdots HOCH\_2CH\_3}$$
#### *2.2. Optical Properties*
UV–VIS diffuse reflectance spectra of the 1:18 SrCO3, 1:23 SrCO3 and commercial SrCO3 in Figure 5a showed that the characteristic absorption edge of the 1:23 SrCO3 sample is located in the visible light region (473 nm), whereas the spectral response of other SrCO3 samples was observed in the UV region, with absorption band edges of 268 and 381 nm for the 1:18 SrCO3 sample and commercial SrCO3, respectively. The band gap energy values suggested by Kubelka–Munk plots (Figure 5b) are 4.63 and 3.25 eV for 1:18 SrCO3 and commercial SrCO3, respectively. By contrast, the bandgap energy of the 1:23 SrCO3 sample is 2.62 eV, and its visible response is possibly due to the presence of incorporated ethanol in the solid sample, as suggested by TGA results.
**Figure 5.** (**a**) UV-visible diffuse reflectance spectra and (**b**) Kubelka–Munk plots of the SrCO3 synthesized at Sr:EtOH mole ratios of 1:18 and 1:23, compared with those of commercial SrCO3.
#### *2.3. Decolourisation of Methylene Blue (MB)*
Figure 6a illustrates the colour removal efficiencies of 10 ppm MB aqueous solutions in the dark and under visible light irradiation after 1 h treatment with SrCO3. Similar colour removal efficiencies from treatment of MB(aq) with 1:18 SrCO3 in the dark and under light illumination suggested major adsorption processes occurred due to the wide bandgap of the 1:18 SrCO3 sample. On the other hand, the visible responsive 1:23 SrCO3 and commercial SrCO3 gave higher colour removal efficiencies under irradiation conditions than those from dark experiments, implying both adsorption and photodegradation of MB are of importance. Therefore, from these catalyst screening tests, the colour removal efficiencies of aqueous MB solutions strongly depend on the bandgap energy of SrCO3 materials and that the 1:23 SrCO3 is the most active catalyst. Figure 6b demonstrates that only low colour removal efficiencies occur due to adsorption (in the dark) and photolysis (irradiation and no SrCO3). Treatments of dye solutions with 1:23 SrCO3 is much less effective (low colour removal efficiency) under dark conditions in comparison to decolourisation under visible light irradiation. These results suggest that the main process of MB colour removal is caused by photocatalytic treatment by using the SrCO3 photocatalyst rather than adsorption.
**Figure 6.** (**a**) Colour removal efficiencies of 10 ppm methylene blue (MB) aqueous solution in the dark, and under visible light irradiation in the presence of SrCO3. (**b**) Absorption spectra of MB in the dark or under visible light irradiation by SrCO3 (1:23 and 1:18). All decolourisation experiments were performed at 30 ◦C using SrCO3 with catalyst loadings of 4.0 g·L−<sup>1</sup> with 1 h treatment.
The percentage of MB colour removal after treatment with SrCO3 photocatalyst (sample 1:23) is shown in Figure 7a. When a suspension of SrCO3 in 10 ppm fresh MB solution was kept in the dark for 3 h, the concentration of dye slightly decreased, while the colour of the dye solution remained unchanged. It was observed that the absorption capacity of MB on the SrCO3 surface is negligible because the specific area of the prepared SrCO3 photocatalyst is low (9.23 m2·g−1). Upon visible irradiation, the prepared SrCO3 gave a high percentage of MB colour removal (>99% after 3 h visible irradiation).
**Figure 7.** (**a**) Colour removal efficiencies of 10 ppm MB aqueous solution (pH 5.5) as a function of time in the dark or under visible light irradiation; adsorption of MB into SrCO3 (loading 4 g·L<sup>−</sup>1) in the dark, MB photolysis and photocatalysis of MB under visible light illumination catalyzed by SrCO3 (loading 4 g·L<sup>−</sup>1). (**b**) Effects of a scavenger (tert-BuOH) on the colour removal efficiency of 10 ppm MB after 3 h treatment with the 1:23 SrCO3 (4 g·L<sup>−</sup>1).
In order to prove that hydroxyl radicals (•OH) are the active species in the photocatalytic degradation process, experiments were conducted in the presence of a radical scavenging reagent. One such reagent, tert-butyl alcohol (tert-BuOH), if present, should significantly inhibit the oxidation of MB [41]. The result in Figure 7b indicates that after treatment for 3 h, adding tert-BuOH resulted in poor colour removal efficiencies (6.90%), whereas in the absence of the reagent very high colour removal efficiencies (>99%) were achieved. The formation of a product arising from the reaction
between tert-BuOH and •OH as ascribed through a radical pathway [41] thus resulted in the poor activity, confirming that hydroxyl radicals are the important active species assisting MB degradation.
The effect of pH on the MB decolourisation under visible light irradiation was examined over a range of pH 3–9. The colour removal efficiency reached 73% after 1 h treatment at pH 3, while lower colour removal efficiencies were obtained at pH 5.5 (51%), pH 7 (42%) and pH 9 (29%) over the same time period, as shown in Figure 8a. In addition, the natural logarithm of the MB concentrations was plotted as a function of irradiation time, affording a linear relationship, as presented in Figure 8b. Using the first-order model, the highest rate constant of MB colour removal was obtained at pH 3, with the degradation being slowest at pH 9. The decreasing rate constants of MB decolourisation with increasing pH may be the result of the presence of carbonate (CO3 <sup>2</sup>−) and hydroxide (OH−) ions, which are radical scavengers [42,43]. At pH 5.5–10, the low colour removal efficiencies may be due to the following reactions.
$$\mathrm{CO\_3^{2-}} + \bullet\mathrm{OH} \rightarrow \mathrm{CO\_3^{\bullet-}} + \mathrm{OH^-}$$
$$\cdot \text{OH}^- + \bullet \text{OH} \rightarrow \text{H}\_2\text{O} + \text{O}^-$$
**Figure 8.** (**a**) Colour removal efficiencies of 10 ppm MB aqueous solution with time, using SrCO3 as photocatalyst. (**b**) Kinetics of MB decolourisation catalyzed by SrCO3 as a function of pH. All decolourisation experiments were performed using SrCO3 with catalyst loading of 4.0 g·L−<sup>1</sup> at 30 ◦C from pH 3–9.
The effect of temperature on the degradation of MB as a function of time is discussed in Figure 9. From Figure 9a, it can be observed that higher temperatures result in higher MB colour removal efficiencies. Under visible light irradiation, the MB colour removal efficiency reached 100% after 1 h treatment at 70 ◦C. In all cases MB concentrations decrease with irradiation time. The linear plots between the natural logarithm of the MB concentration versus irradiation time are shown in Figure 9b, which indicate that the decolourisation process follows first-order kinetics. The rate constants of MB decolourisation increased with temperature, indicating that MB removal by 1:23 SrCO3 is overall endothermic. The 1:23 SrCO3 sample is rather stable during the photocatalytic MB degradation reaction, as only negligible concentrations of Sr (<10 ppm) were detected in the treated MB solution.
**Figure 9.** (**a**) Colour removal efficiencies of 10 ppm MB aqueous solution (pH 5.5) over time using SrCO3 as photocatalyst. (**b**) Kinetics of MB decolourisation catalyzed by SrCO3 as a function of temperature. All decolourisation experiments were performed using 1:23 SrCO3 with catalyst loading of 4.0 g·L−<sup>1</sup> at temperatures 20–70 ◦C.
#### *2.4. Degradation Products*
Figure 10 highlights mass spectra generated from the MB degradation products with the mass-to-charge ratios (m/z) of 77, 122, 234, 284 and 303, reported with the possible fragmented ions shown accordingly.
**Figure 10.** Mass spectra of intermediates from the MB degradation after treatment for (**a**) 10 min, (**b**) 25 min and (**c**) 60 min. All experiments were performed by suspending 1:23 SrCO3 in 10 ppm MB (4 g·L−<sup>1</sup> of MB solution) followed by visible light illumination.
The proposed reaction pathway of MB photooxidation over SrCO3 photocatalyst is outlined in Figure 11. The detected degradation products, as identified from fragments based on m/z ratio, are illustrated in blue, while undetectable but expected intermediates [44,45] are presented in black. These results are in general agreement with previous works that report the generated intermediates during the MB photodegradation process [44,45].
**Figure 11.** Proposed photocatalytic degradation pathway of MB. Detected degradation products are illustrated in blue, while expected but undetectable [44] species are presented in black.
#### **3. Materials and Methods**
#### *3.1. Chemicals*
All reagents were used without further purification. Chemicals of HPLC grade were acetic acid (C2H3O2, Merck, Darmstadt, Germany) and acetonitrile (C2H2N, J.T. Baker, CA, USA). Chemicals of AR grade were ethanol (C2H5OH, Merck, Germany), potassium bromide (KBr, Merck, Germany), tert-butanol (C4H10O, Merck, Germany), ammonium acetate (C2H2ONH4, Rankem, Gurugram, India), sodium hydroxide (NaOH, Rankem, India), methylene blue (C16H18N3Cl, Fluka, Saint Louis, MO, USA), strontium carbonate (SrCO3, Fluka, USA), concentrated hydrochloric acid (HCl, Lab Scan, Ireland), mercury(II) sulphate (HgSO4, QRëc, Newzaland), nitric acid (HNO3, Mallinckrodt Chemicals, Phillipsburg, NJ, USA), potassium dichromate (K2Cr2O7, Unilab, Mandaluyong, Philippines), potassium hydrogen phthalate (C8H5KO4, Univar, Redmond to Downers Grove, IL, USA), silver sulphate (AgSO4, Carlo Erba, Barcelona, Spain), strontium hydroxide octahydrate (Sr(OH)2·8H2O, Sigma Aldrich, Saint Louis, MO, USA) and concentrated sulfuric acid (H2SO4, Lab supplies, Spain).
#### *3.2. Synthesis of Strontium Carbonate (SrCO3)*
Strontium carbonate (SrCO3) was synthesized by a solvothermal method modified from the procedure of Zhang et al. [34]. A suspension of 20 g Sr(OH)2·8H2O in ethanol (100 mL) was sonicated in an ultrasonic bath for 20 min, followed by solvothermal treatment in an autoclave at 80, 100, 120 or 150 ◦C for 2 h. The reaction mixtures were left at room temperature to cool down to room temperature. Then, the precipitates were washed with deionized water to remove Sr(OH)2·xH2O, dried and kept in a dry condition at room temperature. After obtaining the optimal treatment temperature, the reaction time was investigated through the above procedure by fixing the treatment temperature at 100 ◦C and varying reaction time between 0.5, 1, 2 or 3 h. The strontium-based samples were prepared by varying the Sr(OH)2·8H2O: ethanol mole ratio as either 1:13, 1:18, 1:23 or 1:27, and then the above procedures were followed using a treatment temperature of 100 ◦C for 2 h.
#### *3.3. Materials Characterisation*
The crystallinity and the phase structure of the samples were investigated using X-ray diffractometry (PXRD, Bruker AXS model D8 advance). The measurements were examined with CuKα radiation between 2<sup>θ</sup> values of 10–80 degrees, at a scan rate of 0.075 degree·min−<sup>1</sup> using accelerating voltage and currents of 40 kV and 40 mA, respectively. Chemical composition and bonding information were probed using Fourier transform infrared spectrophotometry (FT-IR, Elmer model lamda 800). Diffusion reflectance spectra were measured on a UV–VIS spectrophotometer (Agilent Cary 5000) using a scanning rate of 200–1100 nm. Sample morphologies were investigated using scanning electron microscopy (SEM). The thermal decomposition of SrCO3 was monitored using a thermogravimetric analyzer (TGA, TA instruments SDT 2960 Simultaneous DSC-TGA).
#### *3.4. Catalyst Performance Examinations*
SrCO3 samples were dispersed in 10 mL of 10 ppm MB aqueous solution in order to observe the change in colour under dark and visible light irradiation conditions. Before illumination, the suspensions were stirred in the dark for 5 min. Then, suspensions were irradiated using an LED (16 × 12 V EnduraLED 10 W MR16 dimmable 4000 K with λ > 400 nm) [46]. The colour removal efficiency of MB was monitored as a function of degradation time by measuring the absorbance of the dye solution after treatment. In order to terminate the reaction, the photocatalyst was filtered off using a syringe filter (0.45 μm). The absorbance of the dye was then measured, and the concentration of remaining MB was quantified using the absorbance at maximum wavelength (around 664.5 nm) using the Beer Lambert law.
The colour removal efficiency of MB was calculated via Equation (1):
$$\text{Color removal efficiency} = \left(\frac{\text{C}\_0 - \text{C}\_l}{\text{C}\_0}\right) \times 10\tag{1}$$
where *C*<sup>0</sup> is the concentration of fresh MB solution, and *Ct* is the concentration of dye residue after treatment at *t* minutes.
Leaching of strontium ions may be a major cause of photocatalyst deactivation. Therefore, the amount of strontium ions in the filtered MB solution was quantified by flame atomic absorption spectrometry (FAAS, Perkin Elmer, Waltham, MA, USA).
A mass spectrometer (micro TOF MS, Bruker, Billerica, MA, USA) equipped with electrospray ionization (ESI) source was employed to detect MB degradation products. For this, direct injection of the treated MB solution (with 1:23 SrCO3) under visible light irradiation was carried out, with fragments examined over the range m/z 50–700.
#### **4. Conclusions**
In this work, a solvothermal method without any calcination step was employed to prepare a single crystalline phase of strontium carbonate (SrCO3). Ethanol incorporated SrCO3, a visible light responsive SrCO3 material having a bandgap energy of 2.62 eV, was obtained from the solvothermal treatment of hydrated strontium hydroxide in ethanol at Sr:EtOH of 1:23. Nevertheless, the synthesis conditions strongly influence the bandgap energy of SrCO3, as UV responsive SrCO3 material can also be obtained by varying the precursor concentration. The narrow bandgap SrCO3 material can be utilized as a photocatalyst for decolourisation of methylene blue in water under visible light irradiation. Effective decolourisation of 10 ppm methylene blue aqueous solutions was achieved with >99% colour removal efficiencies after 3 h treatment, under visible light irradiation over the 1:23 photocatalyst, using a catalyst loading of 4 g·L−1. The decolourisation is mainly due to photocatalytic processes. The rate constant values showed a direct correlation with temperature, but decolourisation was most rapid at low pH. In addition to the conventional uses of SrCO3 in pyrotechnics and frit manufacturing, synthesized SrCO3 materials have their place as semiconductors and cocatalysts employed in energy
and environmental applications. The key findings of this work highlight that incorporated ethanol in the SrCO3 structure results in a narrowing of the energy bandgap in SrCO3, with the material being a visible light responsive semiconductor and active photocatalyst in dye degradation. Results from this work may suggest alternative synthesis routes to obtain visible responsive SrCO3 materials, for further development of new composites and cocatalysts in broader applications.
**Supplementary Materials:** The following are available online at http://www.mdpi.com/2073-4344/10/9/1069/s1. Figure S1: PXRD patterns of Sr-containing samples derived from solvothermal treatments of hydrated strontium hydroxide in ethanol (a) at various solvothermal temperatures, 2 h, Sr:EtOH mole ratios of 1:23 and (b) at various solvothermal treatment times, 100 ◦C, Sr:EtOH mole ratios of 1:23; Figure S2: FTIR spectra of Sr-containing samples derived from solvothermal treatment of hydrated strontium hydroxide in ethanol (a) at various solvothermal temperatures, 2 h, Sr:EtOH mole ratios of 1:23 and (b) at various solvothermal treatment times, 100 ◦C, Sr:EtOH mole ratios of 1:23.
**Author Contributions:** Formal acquisition, investigation and writing—original draft, P.W. writing—review, editing, T.K.; funding acquisition, writing—review, editing and supervision, S.M.S. All authors have read and agreed to the published version of the manuscript.
**Funding:** M.Sc. Student scholarship (for P.W.) was provided by the Center of Excellence for Innovation in Chemistry (PERCH-CIC). This work was partially supported by the Thailand Research Fund (Grant No. RSA5980027 and IRN62W0005) for T.K. and S.M.S., the National Research Council of Thailand for P.W, and by the CIF, Faculty of Science, Mahidol University.
**Acknowledgments:** The authors are grateful for partial financial support from the Thailand Research Fund (Grant No. RSA5980027 and IRN62W0005), the National Research Council of Thailand, and the CIF, Faculty of Science, Mahidol University. PP is thankful for an M.Sc. student scholarship from the Center of Excellence for Innovation in Chemistry (PERCH-CIC).
**Conflicts of Interest:** The authors declare no conflict of interest.
#### **References**
© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
MDPI St. Alban-Anlage 66 4052 Basel Switzerland Tel. +41 61 683 77 34 Fax +41 61 302 89 18 www.mdpi.com
*Catalysts* Editorial Office E-mail: [email protected] www.mdpi.com/journal/catalysts
MDPI St. Alban-Anlage 66 4052 Basel Switzerland
Tel: +41 61 683 77 34 Fax: +41 61 302 89 18
www.mdpi.com | doab | 2025-04-07T03:56:58.585447 | 6-5-2022 13:34 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "0039f385-c9b0-4fa4-b06c-9a5b9a2397e4",
"url": "https://mdpi.com/books/pdfview/book/5221",
"author": "",
"title": "Materials and Processes for Photocatalytic and (Photo)Electrocatalytic Removal of Bio-Refractory Pollutants and Emerging Contaminants from Waters",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783036535593",
"section_idx": 4
} |
003cf560-3166-422f-9a1c-98b1e971efa6.0 | **Dietary Polyphenols and Human Health**
• Anna Tresserra-Rimbau
| doab | 2025-04-07T03:56:58.598976 | 1-5-2021 17:49 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "003cf560-3166-422f-9a1c-98b1e971efa6",
"url": "https://mdpi.com/books/pdfview/book/3241",
"author": "",
"title": "Dietary Polyphenols and Human Health",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783039437931",
"section_idx": 0
} |
003cf560-3166-422f-9a1c-98b1e971efa6.1 | **Dietary Polyphenols and Human Health**
Printed Edition of the Special Issue Published in *Nutrients* Anna Tresserra-Rimbau Edited by
| doab | 2025-04-07T03:56:58.599062 | 1-5-2021 17:49 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "003cf560-3166-422f-9a1c-98b1e971efa6",
"url": "https://mdpi.com/books/pdfview/book/3241",
"author": "",
"title": "Dietary Polyphenols and Human Health",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783039437931",
"section_idx": 1
} |
003cf560-3166-422f-9a1c-98b1e971efa6.3 | **Dietary Polyphenols and Human Health**
Editor
**Anna Tresserra-Rimbau**
MDPI • Basel • Beijing • Wuhan • Barcelona • Belgrade • Manchester • Tokyo • Cluj • Tianjin
*Editor* Anna Tresserra-Rimbau University of Barcelona Barcelona
*Editorial Office* MDPI St. Alban-Anlage 66 4052 Basel, Switzerland
This is a reprint of articles from the Special Issue published online in the open access journal *Nutrients* (ISSN 2072-6643) (available at: https://www.mdpi.com/journal/nutrients/special issues/Dietary Polyphenols and Human Health).
For citation purposes, cite each article independently as indicated on the article page online and as indicated below:
LastName, A.A.; LastName, B.B.; LastName, C.C. Article Title. *Journal Name* **Year**, *Volume Number*, Page Range.
**ISBN 978-3-03943-793-1 (Hbk) ISBN 978-3-03943-794-8 (PDF)**
Cover image courtesy of Oriol Pages. `
c 2020 by the authors. Articles in this book are Open Access and distributed under the Creative Commons Attribution (CC BY) license, which allows users to download, copy and build upon published articles, as long as the author and publisher are properly credited, which ensures maximum dissemination and a wider impact of our publications.
The book as a whole is distributed by MDPI under the terms and conditions of the Creative Commons license CC BY-NC-ND.
| doab | 2025-04-07T03:56:58.599100 | 1-5-2021 17:49 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "003cf560-3166-422f-9a1c-98b1e971efa6",
"url": "https://mdpi.com/books/pdfview/book/3241",
"author": "",
"title": "Dietary Polyphenols and Human Health",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783039437931",
"section_idx": 3
} |
003cf560-3166-422f-9a1c-98b1e971efa6.5 | **About the Editor**
**Anna Tresserra-Rimbau** is Serra-Hunter tenure teacher in the Department of Nutrition, Food Science and Gastronomy of the Faculty of Pharmacy and Food Science of the University of Barcelona. She is a member of the INSA-UB (Institute for Research on Nutrition and Food Safety) and the CIBEROBN consortium (Spanish Biomedical Research Centre in Physiopathology of Obesity and Nutrition). Her main research topic is the influence of polyphenols and polyphenol-rich foods on chronic diseases. She has participated in many different national and international projects, mainly human intervention studies such as the PREDIMED, the PREDIMEDplus, and the SI! Program.
### *Editorial* **Dietary Polyphenols and Human Health**
### **Anna Tresserra-Rimbau 1,2**
Received: 18 September 2020; Accepted: 21 September 2020; Published: 22 September 2020
Plant-based foods are the main source of phytochemicals, including polyphenols, a large family of compounds with highly diverse chemical structures. The impact of polyphenols, ranging from simple gallic acid to the most complex proanthocyanidins, on different biological processes has been irrefutably demonstrated by numerous studies [1].
Multiple approaches, each with their strengths and weaknesses, have been used to investigate the effects of polyphenols, all making an important and complementary contribution to the field. In vitro and in vivo experimental models play a vital role in the elucidation of the mechanisms of action underlying the health benefits observed in human trials. However, their results cannot always be easily extrapolated to human beings, partly because of considerable interindividual variability and other external factors. For instance, potential effect-modulating variables, such as sex, age, smoking habits, body mass index, and hormone levels, need to be identified, as does the influence of other foods, nutrients and even culinary techniques [1,2]. Additionally, we should not forget the importance of gut microbiota and genetic polymorphisms, which lead to varied circulating metabolites with different biological activities and health impacts [3].
A more recent approach is the use of omics, an integration of disciplines such as metabolomics, genomics, epigenomics, and foodomics based on cutting-edge experimental techniques, including mass spectrometry. The comprehensive ultra-large data sets they generate allow the scientific community to answer new and complex questions [4].
Daily dietary intake of polyphenols is thought to be approximately 1 g, although this estimate is based on subjective food frequency questionnaires, in which participants tend to overestimate the consumption of healthier items. Moreover, despite the availability of useful and comprehensive databases on polyphenol content in food, the concentrations depend on a wide range of factors, including plant variety, ripeness, environmental conditions, cropping systems, cooking, and storage, all of which add to the complexity of calculating intake [5].
In this Special Issue on "Dietary Polyphenols and Human Health", a series of 10 papers are presented, including three literature reviews [6–8] and seven original research papers [9–15]. The described research contributes to filling some of the gaps in our knowledge about the beneficial effects of dietary polyphenols on chronic health conditions, notably cardiovascular disease, type 2 diabetes, neurological impairment, and also certain risk factors.
In their review, Sandoval et al. describe the molecular mechanisms and signaling pathways involved in the metabolic impact of each group of flavonoids on obesity and related disorders, focusing on the liver, white and brown adipose tissue and central nervous system [6]. Márquez-Campos et al. have collected and summarized the available literature on the antidiabetic effects of both parent flavan-3-ol compounds and their microbial metabolites. The role of microbiota is especially relevant, as flavan-3-ols are poorly absorbed and their metabolization and absorption largely depend on the activity of colonic bacteria [7]. In the third review, Domínguez-López et al. explore the effects of phytoestrogens on human hormone-dependent outcomes throughout the human lifespan, divided into stages of pregnancy, childhood, adulthood, and the pre- and post-menopause [8].
Individual phytoestrogens are also the subject of a cross-sectional study by Sun et al., who are interested specifically in their impact on sleep quality. The association between urinary phytoestrogens (enterolactone, enterodiol, daidzein, O-desmethylangolensin, equol, and genistein) and sleeping disorders and sleep duration was examined in adults from the National Health and Nutrition Examination Survey 2005–2010. Discrepant results were found, depending on the metabolites and the race and sex of the participants, revealing the need for further studies with prospective cohorts and clinical trials [9].
Two of the other papers report clinical trials on the effect of polyphenols on the brain. In a study on psychological well-being (the PPhIT study), Kontogianni et al. concluded that participants with a high polyphenol intake had fewer depressive symptoms and better general mental and physical health compared to those on a low-phenolic diet [10]. The crossover study on mood and cognitive function performed by Wightman et al., where healthy participants received a single dose of a polyphenol-rich extract obtained from mango leaves (<60% mangiferin), revealed no significant results for mood, but cognitive function was enhanced [11].
Taking on the challenge of assessing polyphenol intake, Martini et al. used food frequency questionnaires to compare the nutrients afforded by two different dietary patterns (polyphenol rich and control) in older participants of the MaPLE study. Their ultimate goal is to develop dietary guidelines to increase the intake of these bioactive compounds [12]. Castro-Barquero et al. also used food frequency questionnaires to make a detailed estimation of the polyphenol intake in high cardiovascular risk participants of the PREDIMEDplus study. Monitoring metabolic syndrome symptoms, they found that some phenolic groups were inversely associated with better values of blood pressure, fasting plasma glucose, HDL cholesterol, and triglycerides [13]. Interestingly, both MaPLE and the PREDIMEDplus studies gave similar values for polyphenol intake.
The final two publications shed light on the mechanism of action of polyphenols. Saji et al. explore how a rice bran phenolic extract could target metabolic pathways associated with Type 2 diabetes mellitus, concluding that it modulated the expression of genes involved in β-cell dysfunction and insulin secretion through different mechanisms [14]. Focusing on the pathogenesis of cardiovascular diseases, Nignpense et al. performed a clinical trial with healthy volunteers to evaluate the effect of ingesting a sorghum extract. Although oxidative stress-related endothelial dysfunction and platelet aggregation were not reduced, a beneficial impact on platelet activation and platelet microparticle release was observed [15].
The growth of publications on bioactive compounds in the last years reflects the considerable interest of the scientific community in the field, but a great deal of research still needs to be done. A better understanding of the health benefits of polyphenols and their mechanisms of action will lead to improved (and perhaps individualized) nutritional recommendations aimed at enhancing human health.
**Funding:** This research received no external funding.
**Acknowledgments:** A.T.-R. thanks all the authors for their contributions to this Special Issue, all the reviewers for evaluating the submitted articles, and the editorial staff of the journal *Nutrients*, especially C-W, for always being so kind and helpful.
**Conflicts of Interest:** The author declares no conflict of interest.
### **References**
© 2020 by the author. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
*Article*
| doab | 2025-04-07T03:56:58.599199 | 1-5-2021 17:49 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "003cf560-3166-422f-9a1c-98b1e971efa6",
"url": "https://mdpi.com/books/pdfview/book/3241",
"author": "",
"title": "Dietary Polyphenols and Human Health",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783039437931",
"section_idx": 5
} |
003cf560-3166-422f-9a1c-98b1e971efa6.6 | **Estimated Intakes of Nutrients and Polyphenols in Participants Completing the MaPLE Randomised Controlled Trial and Its Relevance for the Future Development of Dietary Guidelines for the Older Subjects**
**Daniela Martini 1,**†**, Stefano Bernardi 1,**†**, Cristian Del Bo' 1, Nicole Hidalgo Liberona 2,3, Raul Zamora-Ros 2,4, Massimiliano Tucci 1, Antonio Cherubini 5, Marisa Porrini 1, Giorgio Gargari 1, Raúl González-Domínguez 2,3, Gregorio Peron 2,3, Benjamin Kirkup 6, Paul A. Kroon 6, Cristina Andres-Lacueva 2,3, Simone Guglielmetti <sup>1</sup> and Patrizia Riso 1,\***
Received: 28 July 2020; Accepted: 13 August 2020; Published: 15 August 2020
**Abstract:** The evaluation of food intake in older subjects is crucial in order to be able to verify adherence to nutritional recommendations. In this context, estimation of the intake of specific dietary bioactives, such as polyphenols, although particularly challenging, is necessary to plan possible intervention strategies to increase their intake. The aims of the present study were to: (i) evaluate the nutritional composition of dietary menus provided in a residential care setting; (ii) estimate the actual intake of nutrients and polyphenols in a group of older subjects participating in the MaPLE study; and (iii) investigate the impact of an eight-week polyphenol-rich dietary pattern, compared to an eight-week control diet, on overall nutrient and polyphenol intake in older participants. The menus served to the participants provided ~770 mg per day of total polyphenols on average with small variations between seasons. The analysis of real consumption, measured using weighed food diaries, demonstrated a lower nutrient (~20%) and polyphenol intake (~15%) compared to that provided by the menus. The feasibility of dietary patterns that enable an increase in polyphenol intake with putative health benefits for age-related conditions is discussed, with a perspective to developing dietary guidelines for this target population.
**Keywords:** nursing home; residential care; aging; menu; flavonoids; phenolic acids
### **1. Introduction**
It is well recognized that nutrition plays an important role in health status, with increasing evidence of associations between intake of specific dietary components and risk of many non-communicable diseases (NCDs), such as cardiovascular diseases (CVDs), type 2 diabetes, and some types of cancer. For instance, the Global Burden of Diseases has recently indicated that high intake of sodium, low intake of whole grains, and low intake of fruits are the leading dietary risk factors for deaths and disability-adjusted life-years (DALYs) worldwide [1]. These findings have been widely used to prepare national and international dietary guidelines aimed both at recommending the adequate intake of energy and nutrients for different targets of population and possibly at reducing the risk for the most common NCDs [2].
The ageing process affects the nutrient needs of older subjects, whose requirements for some nutrients may be reduced or increased with respect to younger adults. In this life-stage, a variety of factors such as sensory losses, chewing and swallowing problems, and medications may compromise dietary intake and lead to nutritional deficiencies and malnutrition, which has been contributing to the progression of many diseases and common syndromes in older people [3].
For this reason, specific recommendations have been proposed to meet the nutritional requirements of this target group; for instance, energy, protein and fibre intake should be individually adjusted by considering their nutritional status and physical condition and accounting for the presence of specific disease [4]. In addition to macronutrients, micronutrients also play a fundamental role in promoting health and preventing NCDs and their deficiencies are often common in aged people for a number of reasons including reduced food intake or lack of a varied diet, but they are also associated with the vicious cycle promoted by diseases and pharmacological treatments.
It is noteworthy that these factors may also affect the intake, absorption and/or metabolism of bioactive compounds such as polyphenols. In this regard, data on polyphenol intake in different older target groups are not univocal, possibly due to differences in geographical area considered, and in the individual characteristics in terms of health/disease status, and living conditions, as previously evidenced [5]. The interest in the assessment of polyphenol intake and the study of their potential impact on older subjects has been growing by considering several findings suggesting the protective role they can play against age-related diseases and in the promotion of healthy aging [6]. Regarding the changes on polyphenol intake with age, conflicting results have been reported so far, with some studies showing an increased intake [7,8] while others reporting no differences depending on age [9,10].
For the above-mentioned reasons, the nutritional assessment of older people represents a critical issue, which may be particularly true for those living in residential care settings where the prevalence of malnutrition has been reported to be extremely variable, ranging from 1.5 to 66.5% [11]. This represents a current clinical and public health concern at both the individual and population level [12,13]. Several methods have been developed for the assessment of energy and nutrient intake, including food-frequency questionnaires, food diaries and 24-h dietary recalls, all having pros and cons to be considered when choosing the best method to use in each specific context [14]. The estimation of micronutrients and bioactives like polyphenols is particularly challenging, mainly due to methodological issues, including the tool and the database used for the evaluation, as well as the type of polyphenol under consideration (e.g., total polyphenols versus single classes and subclasses of polyphenols) [5]. Being able to make accurate estimates of actual polyphenol intake is a fundamental requirement of developing a better understand of the role of these compounds and their relationship with health or disease conditions. In addition, this information is crucial to define potential polyphenol exploitation for the development of dietary strategies to prevent against age-associated diseases.
Based on these premises, the aim of this research was to evaluate the nutritional composition of nursing home dietary menus and to estimate the actual intake of nutrients and polyphenols in a group of older subjects living in a residential care setting. The assessments were performed as part of the MaPLE (Microbiome mAnipulation through Polyphenols for managing Leakiness in the Elderly) project, funded within the European Joint Programming Initiative "A Healthy Diet for a Healthy Life" (JPI HDHL), with the aim to investigate benefits of a polyphenol-enriched diet on intestinal permeability in older subjects. An increased gut permeability, often associated with dysbiosis and inflammation, could play a role in the development of some age-related conditions. In this regard, it has been suggested that the intake of polyphenols may represent a promising strategy to improve intestinal permeability (IP) as demonstrated mainly in experimental studies suggesting the involvement of these bioactives in both direct and indirect modulatory mechanisms [15]. In this context, a more accurate estimation of the intake of polyphenols in a vulnerable target such as older subjects, in terms of amount, sources and distribution across the day and even in different seasons, can be of relevance. This could enable a better understanding of their potential benefits and the development of specific recommendations based on findings from dietary intervention studies.
### **2. Materials and Methods**
### *2.1. Study Design and Population*
The study design of theMaPLE randomized controlled trial (RCT) has been previously reported [16]. Briefly, the central hypothesis that this study sought to address was that a polyphenol-enriched dietary pattern would reduce IP and systemic inflammation and cause beneficial changes in various biomarkers of cardiometabolic health, and that this would be associated with changes in the gut microbiota in these older subjects. To this aim, volunteers were randomized to consume a polyphenol-rich diet (PR-diet) or a control diet (C-diet) for 8 weeks following a cross-over design separated by an 8-week wash-out. The development of the PR-diet and C-diet has been reported previously [16]. During the intervention, subjects were given three small portions of polyphenol-rich foods daily as substitutes for foods with lower polyphenol contents that were part of the C-diet (developed by analyzing the regular menus provided to the study participants and specifically assessing the nutrient and polyphenol composition). The characteristics and polyphenol content of the servings provided in the PR-diet for each product are reported in Table 1. The amount of polyphenols provided was more than double that deriving from the replaced products. Data shown include total polyphenol content (i.e., TPC) quantified by analysing products through the Folin–Ciocalteau method [17] and estimates of total polyphenols (i.e., TP). The estimation of TP was calculated as the sum of flavonoids, phenolic acids, lignans, stilbenes and other polyphenol classes expressed in mg (aglycone/100 g). The estimations were performed using an in-house ad hoc database of food composition on polyphenols, compiled from the USDA (fdc.nal.usda.gov/) for databases (for flavonoids, isoflavones and proanthocyanidins) and the Phenol-Explorer (PE; www.phenol-explorer.eu) database (for phenolic compounds lignans, stilbenes and other minor polyphenol classes) through a computer application developed that uses the relational database system. This methodology has been used and previously described [18–21]. Polyphenols were expressed as mg of aglycones per 100 g.
For the intervention study, all the participants were recruited from residents at Civitas Vitae, a large residential care setting (OIC Foundation including both nursing homes and independent residencies for older subjects, Padua, Italy) according to specific inclusion and exclusion criteria. Among inclusion criteria, subjects had to be aged 60 years and to have increased intestinal permeability evaluated by serum zonulin level as previously reported [16].
All the participants recruited into the study were self-sufficient and were in good cognitive health. The Ethics Committee of the Università degli Studi di Milano approved the study protocol (15/02/2016; ref.: 6/16/CE\_15.02.16\_Verbale\_All-7). All the participants were provided with detailed information about their involvement in the study and gave their informed consent before beginning the intervention. The trial was registered in the ISRCTN Registry on 28 April 2017; ISRCTN10214981.
**Table 1.** Polyphenol content and composition of each serving of MaPLE products included in the dietary intervention, expressed as mg per serving.
§ Frozen whole blueberry product was thawed and prepared before consumption; <sup>و</sup> Blueberry purée was a ready-to-eat product; ◦ Cocoa powder was dissolved in hot milk or water; \* Green tea was prepared by solubilization of 200 mg of green tea extract in 200 mL of hot water; <sup>+</sup> Renetta apple purée was prepared in controlled conditions and stored at −18 ◦C until consumption. TPC, total polyphenol content by Folin–Ciocalteau assay; TP, total polyphenols determined by USDA and Phenol Explorer databases.
### *2.2. Nutritional and Polyphenol Composition of the Menus*
To estimate the energy and nutrient composition of the planned meals regularly provided, the weekly menus during different seasons (summer, mid-season and winter) were evaluated (i.e., covering the whole intervention study). To this aim, Metadieta ® software (Me.te.da srl, S. Benedetto del Tronto, Italy) was used to include all the recipes and to estimate the nutritional composition of the different menus.
In addition, the TPC content of the menus was estimated by PE databases with the addition of our own data (characterized products in Table 1 used for the intervention) and other literature sources for those ingredients that were not available in those databases [22–24]. TP was instead estimated through the PE/USDA database, as also described in Section 2.1.
### *2.3. Evaluation of Actual Energy, Nutrient and Polyphenol Intake*
During both intervention periods, weighed food records (WFR) were used to estimate food, energy, nutrient and polyphenol intake as reported in Section 2.2. In particular, up to six detailed daily diaries (recording the amount of foods provided and the amount actually consumed by weighing the leftovers) were analysed for each subject during the two intervention periods. In addition, one diary was filled in by participants at baseline and scheduled the day of blood drawings and sampling according to what was previously reported [16].
### *2.4. Statistical Analysis*
Statistical analysis was conducted using the Statistical Package for Social Sciences software (IBM SPSS Statistics, Version 26.0, IBM corp., Chicago, IL, USA) and R statistical software (version 3.6.). One-way ANOVA was applied to analyse differences between the winter, mid-season and summer menus provided during the intervention in terms of nutrients and polyphenol composition. The nonparametric Wilcoxon–Mann–Whitney test with Benjamini–Hochberg correction pairing the data when possible was performed to ascertain differences at baseline between men and women in terms of actual intake and to verify the impact of treatment (PR vs. C-diet) and gender (men vs. women) on both nutrient and polyphenol intake in participants. The level of significance was set at *p* ≤ 0.05. All results were expressed as mean ± standard deviation (SD).
### **3. Results**
Fifty-one older subjects (22 men; 29 women; age ≥ 60 y) successfully completed the entire study, and the data reported here are for those 51 participants. Dropouts were not due to side effects of the dietary intervention itself.
### *3.1. Nutritional Composition of Menus*
The nutritional composition of the nursing home menus provided during the intervention is reported in Table 2. The average estimated daily energy content of the summer menu was 140 kcal higher than for the winter menu. No differences were detected for the nutrients among seasonal menus, both when expressed as net quantity or as percentage of energy provided.
**Table 2.** Mean energy and nutrient composition of the nursing home menus across three seasons and overall mean composition.
Data represent the daily amounts with the units given in parentheses and are shown as mean ± standard deviation. Data have been calculated through the Metadieta ® software. Data with different letters in the same row are significantly different (*p* < 0.05). CHO, carbohydrates; SFA, saturated fatty acids; MUFA, monounsaturated fatty acids; PUFA, polyunsaturated fatty acids; ω-3, omega-3 fatty acids; ω-6, omega-6 fatty acids.
Regarding the polyphenol composition of the menu, as shown in Figure 1, no significant differences were observed among the different seasonal menus, which had an estimated mean TPC of about 770 mg/day.
**Figure 1.** Box plot (panel **A**) showing polyphenol content in the seasonal menus, estimated through PE/USDA databases and other published data (TP in light blue) and by Folin–Ciocalteau data as reported in the PE database and other sources (TPC in red); percentage distribution of polyphenol classes (panel **B**) in the seasonal menus. Dots represent mild outliers that are more extreme than Q1 − 1.5 \* IQR or Q3 + 1.5 \* IQR but are not extreme data (where Q1=quartile 1; Q3=quartile 3; IQR=interquartile range).
### *3.2. Actual Energy, Nutrient and Polyphenol Intake at Baseline and during the Intervention*
The actual energy, nutrient and polyphenol intake estimated at baseline for women, men and the whole group of participants is shown in Table 3. Overall, energy intakes, and accordingly nutrient intakes, were lower than calculated for the estimates based on the foods consumed from the menus, in keeping with the fact that not all the food was consumed for any particular meal. There were no significant differences between women and men for any of the dietary variables that were assessed at baseline. This was also confirmed by analysing the data obtained during the intervention study (Supplementary Materials Figure S1), except for simple carbohydrates in women and for total lipids and PUFA in men when comparing intake measured during the PR-diet and the C-diet (*p* < 0.05). Finally, differences were observed in ω-6 fatty acids, iron and calcium intake following the PR-diet in both women and men.
All data are presented as mean ± standard deviation (SD); Data with *p* < 0.05 are significantly different. CHO, carbohydrates; SFA, saturated fatty acids; MUFA, monounsaturated fatty acids; PUFA, polyunsaturated fatty acids; ω-3, omega-3 fatty acids; ω-6, omega-6 fatty acids. † Comparison between women and men using Wilcoxon–Mann–Whitney test.
Regarding polyphenols, flavonoids and phenolic acids were the most consumed classes and were comparable between women and men.
### *3.3. Polyphenol Intake at Baseline and during Intervention*
Figure 2 shows the polyphenol intake at baseline and in the two intervention periods. At baseline, the intake of TPC was 663.4 ± 147.5 mg/d and comparable between women (669.2 ± 160.1 mg/d) and
men (655.2 ± 130.8 mg/d). The consumption of PR-products significantly (*p* < 0.0001) increased the intake of TPC by about 600 mg/d compared to the C-diet and was comparable in both men and women.
**Figure 2.** Polyphenol intake at baseline and in polyphenol (PR)-diet and control (C)-diet in the whole group of subjects and stratified by gender. The intake has been estimated using the PE and USDA databases and other published data (TP in light blue) and by Folin–Ciocalteau data as reported in the PE database and other sources (TPC in red). Dots represent mild outliers that are more extreme than Q1 − 1.5 \* IQR or Q3 + 1.5 \* IQR but are not extreme data. Asterisks are extreme data that are more extreme than Q1 − 3 \* IQR or Q3 + 3 \* IQR (where Q1=quartile 1; Q3=quartile 3; IQR=interquartile range).
Table 4 shows the contribution of different polyphenol classes to the total polyphenol intake during the PR and C-diet. Flavonoids were the main subclass increased in the PR-rich diet and accounted for 74.6%, followed by phenolic acids (23.3%), while lignans and other polyphenols accounted for the remainder. A treatment effect (*p* < 0.0001) for total flavonoids and phenolic acids was observed (Table 4), while a gender effect was observed for stilbenes showing a higher intake in men compared to women (*p* = 0.033).
**Table 4.** Intake of total polyphenols and classes (according to PE/USDA databases) during the PR-diet and the C-diet.
All data are expressed as mean ± standard deviation (SD); PR, polyphenol-rich diet; C, control diet. † Comparison between PR-diet vs. C-diet in women. ¥ Comparison between PR-diet vs. C-diet in men. # Comparison between women and men in PR-diet and C-diet. § Comparison between PR-diet vs. C-diet in all subjects. Comparisons have been performed using the Wilcoxon–Mann–Whitney test.
Considering the total polyphenol (TP) contribution from the different meals, in the PR-diet, ~50% of polyphenol intake derives from snacks and the remaining ~50% from breakfast, lunch and dinner (Figure 3). In particular, there is a significant contribution to mid-morning and afternoon snacks from the intake of PR-products. Conversely, during the C-diet, only ~15% of the total polyphenols consumed were derived from snacks.
**Figure 3.** Total phenolic contribution from different meals during the polyphenol (PR)-diet and the control (C)-diet expressed as percentage (Panel **A**), or expressed as amount in mg during the PR-diet (in red) and the C-diet (in blue) as estimated through PE/USDA databases and other published data (Panel **B**). Dots represent mild outliers that are more extreme than Q1 − 1.5 \* IQR or Q3 + 1.5 \* IQR but are not extreme data. Asterisks are extreme data that are more extreme than Q1 − 3 \* IQR or Q3 + 3 \* IQR (where Q1=quartile 1; Q3=quartile 3; IQR=interquartile range).
Overall, through the analysis of the menu items provided to the volunteers and recorded in the WFRs during the two intervention periods and by considering the frequencies of consumption of the single ingredients, we estimated the main polyphenol sources contributing to the different meal times. During the PR-diet, the main foods providing polyphenols at breakfast were fruit and fruit-derived products (e.g., orange, grape, orange juice, apricot jam, etc.), followed by barley coffee and minor contributions from coffee and tea. Polyphenol-rich products on the PR-rich diet were occasionally consumed at breakfast, where green tea, pomegranate juice, chocolate callets and blood orange juice were the most commonly consumed. For lunch and dinner, the main sources during the PR-diet were vegetables (e.g., chard, asparagus, broccoli, carrots), extra virgin olive oil, legumes and spices. A few participants occasionally consumed white wine in small portions (usually 1 glass), which also made a contribution to the polyphenol intake. PR-rich products were mainly consumed as mid-morning and mid-afternoon snacks, as reported in Figure 3. During the C-diet, we found similar foods providing polyphenols at breakfast, lunch and dinner compared to the PR-diet, except for the introduced PR-rich products. Major differences between the two treatments were largely due to the snack foods because only fruits and fruit-based products (i.e., juices), cakes (including sometime chocolate-based cakes) or yogurt were available during the C-diet, whereas a more extensive range of PR-foods were available as snacks on the PR-rich diet.
### **4. Discussion**
The evaluation of the adequacy of diets in older subjects is of utmost importance not only to identify possible deviations from desirable nutritional targets but also to contribute to the development of new recommendations that address gaps in the current guidance. In this context, the MaPLE project has given us the unique opportunity to assess dietary intake in a well-controlled setting where it is also possible to analyse the daily menus provided to the residents, while considering all the recipes and ingredients used for the preparation of the meals. At the same time, long-term residences often have facilities enabling the measurement of food intake (e.g., by collecting multiple weighed food records) and this represents the best procedure to estimate actual consumption. Menu planning in residential care involves modifications of recipes during the year to take account of seasonal changes in ingredient availability and this may partially affect not only nutritional characteristics in terms of macro- and micro- nutrients but also food sources of bioactive compounds with potential impact on host metabolism and other functions.
In the present study, the evaluation of three different menus showed that overall they were comparable in terms of nutritional composition, and also that they were in line with the dietary recommendations for older subjects in Italy (i.e., Italian Reference Intake) [2], with some dissimilarities that are worth highlighting. In regards to total energy, menus provided suitable amounts for the target population, at least in consideration of the main Italian guidelines developed for dietary management in residential care [25]. Some studies carried out in nursing homes showed lower energy provided by menus [26,27], while others reported data higher or similar to our observation [28–30]. The distribution in macronutrients was consistent with the recommendations: carbohydrates accounted for ~47% of total energy intake on average (reference intake range: 45–60% energy (E)), although we found there was a higher intake of simple carbohydrate in comparison with the recommendations (20% E vs. < 15% E) due to the wide use of fruit juices and hot beverages with added sugars as has been commonly reported in this target population. Protein intake derived mainly from animal sources (about two-thirds) and was higher in comparison with the suggested dietary target (1.1 g/kg/day), while total lipid intake was within the reference intake range (20–35% E). Specifically, SFAs were in accordance with the national/international recommendation (<10% E), while total PUFAs were slightly lower than 5% E due to the low intake of ω-6 in favour of higher MUFAs, as can often be found in the Mediterranean areas. The amount of fibre provided by the menus was slightly lower than the suggested dietary target of 25 g per day defined by Italian and international guidelines [2,31]. Regarding micronutrients, iron contribution was adequate while, as also reported in the literature, calcium content in the three menus was lower than the population reference intake (PRI, 1200 mg for both women and men ≥ 60 years) [2]. However, it is worth noting that these data included only calcium derived from recipes and did not consider contributions from other sources such as water and
supplements. Vitamin B1, B6 and B12 provided by menus were higher than reference values, while folates were slightly lower than the established population reference intake of 400 μg per day. With regard to antioxidants, vitamins E and C were both adequate, in particular vitamin C largely exceeded the PRI levels (i.e., 85 mg and 105 mg per day for women and men respectively). Overall, the results on the nutritional composition of the menus suggest that, although they are generally developed following specific guidelines, it is still possible to improve the content of critical nutrients such as fibre, specific micronutrients and bioactives, above all in institutionalised subjects as also reported in the literature [29,30].
Notably, actual food intake in older subjects can be significantly lower with respect to that provided by the menus. For these reasons, we also estimated the actual food consumption through the analysis of detailed and repeated weighed food records. Measured energy and nutrient intake were indeed lower than that provided through the menus (by about 20%), with no differences between women and men. In this regard, it is underlined that the subjects enrolled in the present study generally had a good nutritional status, evidenced also by their anthropometric characteristics (BMI <sup>=</sup> 26.8 <sup>±</sup> 5.5 kg/m2).
The energy intakes we have reported here (mean approximately 1580 kcal) were slightly lower than those found in the InCHIANTI study, performed on about 1200 free-living older subjects (>65 years) in Tuscany, in which mean energy intakes ranged from 1764 to 2260 kcal/d and from 1521 to 1793 kcal/d in men and women, respectively [32]. However, despite the higher energy intake, in the InCHIANTI study, a large group of subjects reported inadequate intakes of protein, calcium and other nutrients, which have been independently associated with frailty [33]. In our assessments, the lower food intake was associated with reduced protein intake (about 0.9 g/kg day on average), increasing the rate of inadequate intake above all in male subjects (about 22% with intake ≤ 0.71 g/kg per day and only 18% with intake ≥ 1.1 g/kg per day as defined by the suggested dietary target). The consumption of simple carbohydrates in older subjects was confirmed to be higher than the suggested values, while the fat intake appeared to be within the suggested intake range, although the amount of ω-6 fatty acids remained lower than recommended values, as did the intake of calcium, vitamins B1, B6 and folates. These results confirmed previous observations of a potential risk of long-term inadequate intake of nutrients that are fundamental for maintenance of functional and metabolic integrity in older subjects, and that these inadequate intakes are likely due to the actual food intake being significantly less than the amount of food provided to the care home residents in each meal (i.e., incomplete meal consumption is likely a major cause). Moreover, there is not only a problem related to overall food intake but also to specific classes of products that appear to be consumed in lower amounts with respect to others, for example justifying a low intake of fibre that has been found for most, if not all, the subjects under study. This is an underestimated consideration that should be a target for future multidisciplinary research that is able to finally implement guidelines for the achievement of nutritional targets through traditional or possibly alternative strategies.
A major focus in this study was polyphenols because these compounds have the potential to provide further specific benefits to the target population under study. It has been reported that there is a large variation in the polyphenol content of foods available in different periods of the year [34–36], and for this reason we specifically analysed recipes and ingredients used to develop seasonal menus and the results obtained showed a relatively comparable amount of these bioactive compounds (about 770 mg per day on average as TPC) among the different seasons. We could not find other data on the impact of seasonality on polyphenol content of dietary plans provided in long-term residences for older people, while more literature is available in free-living older subjects. In this regard, in the Blue Mountains Eye Study, a longitudinal study performed in Australia [35], the authors found that season did not affect the overall total flavonoid intake in a group of adult and older subjects; however, it was relatively higher in spring and lower in autumn in line with our results. Conversely, Tatsumi et al. [37] showed that total antioxidant intake in a Japanese population (39–77 years) was highest in winter and lowest in summer. The authors attributed this difference to the participants' selection of food (in particular fruits and vegetables) but also beverages across seasons.
In our study, the assessment of actual food consumption at baseline indicated a mean TPC intake of ~660 mg/d (i.e., evaluated by Folin–Ciocalteau through the PE database and specific literature), about 15% less than the amount estimated in the menus served to the study participants. Although a thorough comparison with other published data must be done cautiously because of the differences in the populations under study and the methods and databases used for estimating the intakes of total polyphenols and polyphenol classes, the overall actual intake estimated in the present study seems to be comparable with mean intake observed in the InChianti study [20], but lower with respect to others previously reported.
In fact, assessments in older subjects estimated polyphenol intakes from 333 mg/day up to 1492 mg/day, as reported previously [5]. For example, in the PREDIMED study evaluating a big cohort of Spanish older subjects aged 55–80 years, a mean polyphenol intake of 820 ± 323 mg/day expressed as glycosides was estimated through the PE database, by analysis of food consumption data obtained from FFQs [38]. With regard to the contribution of the classes, total flavonoid intake is generally the larger part of the intake, while data available in some studies suggest that up to 30–40% of the total polyphenol intake can be represented by phenolic acids [5]. Results from the EPIC cohort showed that older subjects tended to have increased intake of flavonoids, stilbenes, lignans, and other polyphenols with respect to younger individuals, while no differences were found for total polyphenol intake [7], and similar findings were reported by Karam and colleagues [8], also showing an impact of gender. In our study in a controlled setting, the data confirmed that the flavonoid subclass was the greatest contributor to total polyphenol intake followed by phenolic acids, while no differences were detected between men and women. Some studies have suggested a higher total and subclass polyphenol intake in females compared to males [8,10], above all when standardized by energy intake, and this may also be the reason for the lack of differences in our study. In addition, it is relevant that the overall lower availability of food alternatives for selection in controlled, with respect to a free-living condition, may have affected eating behaviour, increasing the comparability of the dietary intake.
With regard to polyphenol food sources, tea and coffee have been underlined as the main polyphenol contributors in northern European older subjects, while red wine, extra virgin olive oil and fruit are the main sources in Southern Europe [7,39]. In our evaluation, fruit and fruit juices, vegetable and extra virgin olive oil represent the main food categories providing polyphenols. In addition, we could not demonstrate a different selection of polyphenol sources depending on gender, despite some studies having reported a higher contribution from fruit and vegetables in females compared to males [8,34]. It is noteworthy that in the nursing home, the intake of coffee and wine was strongly limited, if not denied, to limit risks associated with caffeine and alcohol consumption and this may represent an important behavioural difference with respect to what may be observed in free-living older subjects.
The evaluation of habitual polyphenol intake in the older target group was a fundamental step in the process of developing a reliable and evidence based polyphenol rich dietary pattern to use for the intervention trial. In particular, the aim was to approximately double the habitual polyphenol intake of the nursing home residents when on the PR-rich diet in order to reach amounts in the highest quantile of intake identified in previous observational studies, where older subjects were included or specifically considered [7,21,40].
Indeed, the main objective of the MaPLE study was to investigate whether the increased intake of polyphenols might cause a reduction in intestinal permeability (IP) and inflammation associated with an improved intestinal microbial ecosystem, also affecting metabolic and functional activities in the older subjects [16]. In particular, the intervention was developed by replacing three portions per day of low polyphenol foods/beverages with specific products rich in polyphenols. The selection of the products was performed by considering different aspects: (i) the total amount of polyphenols provided, (ii) the contribution of the different polyphenol classes, (iii) the adequate portion of food able to provide a reliable high dose of polyphenols, and (iv) the possible food preparation in order to ensure polyphenol bioavailability. Additionally, foods selection was carried out by considering the characteristics of the target group and their specific needs in terms of acceptability and suitability in the context of residential care settings. Through the administration of the selected foods, we provided mainly flavonoids (approximately four times higher compared to the amount introduced through the C-diet) and phenolic acids. These bioactives have been suggested as potential modulators of critical factors and specific targets regulating IP, including the impact on microbiota composition and activities [15,41]. Overall, our results demonstrate that it is possible to obtain a significant increase in polyphenol intake in older subjects, through the use of small amounts of well-accepted polyphenol-rich food products. Moreover, it has been demonstrated that the intake is well tolerated and without undesirable effects. Participants appreciated the products and were interested in continuing with the dietary protocol after the end of the trial, suggesting that older people can change their diet if it does not dramatically modify their eating habits.
An interesting observation highlighted was that older subjects preferred the consumption of PR-products during the intervention as mid-morning and afternoon snacks. In fact, the protocol adopted did not fix the timing for the PR-food intake, but the products should have been consumed within the day according to preferences and/or habits. For this reason, our results give an important contribution to the development of dietary guidelines for this target population. At the same time, the analysis of the pattern of consumption of polyphenol-rich foods may also contribute to a better understanding of chronobiological aspects related to the effect of bioactive compounds. In this regard, it has been suggested that the inclusion of polyphenols within the meals may have an impact on related metabolic responses, e.g., through reduction of glucose and lipid levels, inflammation, oxidative stress, and blood pressure, associated with food intake [42–44]. Consuming most of the polyphenols outside of the main meals could also affect their bioavailability for direct absorption and their use as substrates for microbial transformation.
This work has several strengths mainly related to the well-controlled setting of the intervention, enabling both the evaluation of the nutrient and bioactive content of the menus and the actual intake during the whole intervention, ensuring high adherence to dietary instructions. Conversely, possible study limitations include the small sample size and the partial generalizability to free-living community dwelling older subjects. Finally, the limited food choices available in the main standard menus provided could have reduced the possibility of showing gender differences.
### **5. Conclusions and Perspectives**
In conclusion, the assessments performed within the MaPLE project have further underlined the need for a careful revision of dietary menus addressed for older subjects not only to optimize the intake of essential nutrients, but also of bioactive compounds, such as polyphenols, in order to lower the risk of chronic diseases and improve specific metabolic and functional activities during aging. In this context, we have shown that there is a possibility to develop feasible and reliable polyphenol-rich dietary patterns that can be appreciated and consumed by the older population with excellent compliance, while assuring a significant increase in the intake of these bioactive compounds. Moreover, the products and preparations included in the dietary menu have been easily managed in the residential care setting and this is a practical aspect of relevance for the success of new recommendations.
Further studies are needed to: (i) improve tools available to better estimate polyphenol intake and enable comparison of different data in the literature, as previously reported [5]; and (ii) improve dietary recommendations by defining the amount of polyphenol needed in order to obtain, if confirmed, the postulated health benefits in the older subjects. This is not an easy task and imply a strong research effort that needs to consider the potential impact of these results for the development of evidence-based dietary guidelines for the management of age-related conditions.
**Supplementary Materials:** The following are available online at http://www.mdpi.com/2072-6643/12/8/2458/s1, Figure S1: comparison of percentage energy and nutrient intake during 8-week polyphenol-rich diet (PR-diet) and control diet (C-diet) in older women and men.
**Author Contributions:** P.R. and S.G. designed the trial and in collaboration with A.C., C.A.-L. and P.A.K. optimised the study protocol including the development of the polyphenol rich dietary pattern. D.M., S.B., C.D.B. and P.R. drafted the manuscript. S.B., M.T., supervised by M.P. and P.R., evaluated the nutritional composition of dietary plans and estimated the actual nutrient intake. N.H.L., R.Z.-R. and C.A.-L. estimated the actual dietary polyphenol intake in collaboration with R.G.-D. and G.P., D.M., C.D.B., B.K. and G.G. performed the statistical analysis. All authors have read and agreed to the published version of the manuscript.
**Funding:** This research was undertaken as part of the MAPLE project (Gut and blood microbiomics for studying the effect of a polyphenol-rich dietary pattern on intestinal permeability in the elderly), which was funded through the European Joint Programming Initiative "A Healthy Diet for a Healthy Life" (JPI-HDHL-http://www. healthydietforhealthylife.eu/) with national funding support provided by Mipaaf (Italy; D.M. 8245/7303/2016), MINECO (Spain, PCIN-2015-238) and the BBSRC (UK, BB/N023951/1). In addition, C.A.-L. thanks 2017SGR1546 from AGAUR, CIBERFES (co-funded by the FEDER Program from EU) and ICREA Academia award 2018.
**Acknowledgments:** C.D.B. is grateful for support granted by "Piano di sostegno alla ricerca- Linea 2, azione A-grant number "PSR2019-CDELB". P.R. and C.D.B. acknowledge the European Cooperation for Science and Technology (COST Action) CA16112 "NutRedOx: Personalized Nutrition in Aging Society: Redox Control of Major Age-related Diseases". P.R. thanks also Coordinated Research Center (CRC) "Innovation for Well-Being and Environment (I-WE). R.Z.-R. was supported by the "Miguel Servet" program (CP15/00100) from the Institute of Health Carlos III (Co-funded by the European Social Fund (ESF)-ESF investing in your future). PAK and BK are grateful for additional support from the BBSRC (UK) through an Institute Strategic Programme Grant ('Food Innovation and Health'; Grant No. BB/R012512/1) and its constituent project BBS/E/F/000PR10346 (Theme 3, Digestion and Fermentation in the Lower GI Tract) to Quadram Institute Bioscience. We are grateful to the valuable contribution and dedication of our older volunteers and to the nursing and medical staff working at OIC Foundation (Padua, Italy). We would like to specifically acknowledge Alberto Fantuzzo, Chiara Cavazzini, Lorella Pinton, Paolo Bergantin, Rosanna Ceccato, Pamela Soranzo, and Silvana Giraldini and Guido Masnata. We are also grateful to all the physicians (Michela Rigon, Lorena D'Aloise, Antonio Merlo, Elisabetta Bernardinello, Nadia Malacarne, Silvana Bortoli, Fabiola Talato, Agostino Corsini, Maria Licursi, Nicoletta Marcon, Angela Sansone), the nurses and other personnel at the residential care who were essential to complete the study successfully. Giulia Minto and Nicola Fassetta are acknowledged for their help during the evaluation of food intake. Finally, the authors are grateful to Barry Callebaut, Indena, Melinda, Oranfrizer, Roberts Berrie, Zuegg for their kind contribution of products used for the intervention.
**Conflicts of Interest:** The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.
### **References**
© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
| doab | 2025-04-07T03:56:58.599774 | 1-5-2021 17:49 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "003cf560-3166-422f-9a1c-98b1e971efa6",
"url": "https://mdpi.com/books/pdfview/book/3241",
"author": "",
"title": "Dietary Polyphenols and Human Health",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783039437931",
"section_idx": 6
} |
003cf560-3166-422f-9a1c-98b1e971efa6.7 | *Article* **A High Polyphenol Diet Improves Psychological Well-Being: The Polyphenol Intervention Trial (PPhIT)**
**Meropi D. Kontogianni 1, Aswathy Vijayakumar 2, Ciara Rooney 2, Rebecca L. Noad 2,3, Katherine M. Appleton 4, Danielle McCarthy 5, Michael Donnelly 2, Ian S. Young 2,5, Michelle C. McKinley 2,5, Pascal P. McKeown 2,3 and Jayne V. Woodside 2,5,\***
Received: 29 May 2020; Accepted: 12 August 2020; Published: 14 August 2020
**Abstract:** Mental ill health is currently one of the leading causes of disease burden worldwide. A growing body of data has emerged supporting the role of diet, especially polyphenols, which have anxiolytic and antidepressant-like properties. The aim of the present study was to assess the effect of a high polyphenol diet (HPD) compared to a low polyphenol diet (LPD) on aspects of psychological well-being in the Polyphenol Intervention Trial (PPhIT). Ninety-nine mildly hypertensive participants aged 40–65 years were enrolled in a four-week LPD washout period and then randomised to either an LPD or an HPD for eight weeks. Both at baseline and the end of intervention, participants' lifestyle and psychological well-being were assessed. The participants in the HPD group reported a decrease in depressive symptoms, as assessed by the Beck Depression Inventory-II, and an improvement in physical component and mental health component scores as assessed with 36-Item Short Form Survey. No differences in anxiety, stress, self-esteem or body image perception were observed. In summary, the study findings suggest that the adoption of a polyphenol-rich diet could potentially lead to beneficial effects including a reduction in depressive symptoms and improvements in general mental health status and physical health in hypertensive participants.
**Keywords:** polyphenols; fruits; berries; vegetables; dark chocolate; psychological well-being; depression; physical health; mental health
### **1. Introduction**
Mental ill health, manifesting itself in a wide range of conditions such as depression, anxiety and stress [1], represents one of the leading causes of burden of disease worldwide, also substantially increasing the risk of cardiovascular disease, diabetes and cancer [2–4] and adversely affecting quality of life (QoL), relationships and the ability to work [5]. Northern Ireland has the highest prevalence of mental illness within the UK, and psychiatric morbidity is 25% higher than in the UK [6].
Thus, research is required in order to establish inexpensive and effective techniques to reduce the incidence of mental health problems and to improve the psychological well-being of the population. Alongside genetic and biological factors, researchers have increasingly begun to examine the role of lifestyle factors, including dietary intake, in the promotion of psychological well-being and the
prevention of mental illness [7,8]. Studies that have explored potential associations between nutrient intake (namely carbohydrates, B vitamins and antioxidants such as vitamins C, E and polyphenols) or foods rich in these nutrients (e.g., fruits, vegetables, legumes, coffee, chocolate) and psychological well-being have produced conflicting results [9–12].
Polyphenols, in particular, have gained increasing attention from health researchers in recent years due to their biological properties, as well as their abundance within the human diet [13]. A growing number of epidemiological studies support a role for polyphenols in the prevention of chronic non-communicable diseases such as cardiovascular disease (CVD) [14], cancer [15] and neurodegenerative diseases [14,16]. Furthermore, animal studies have demonstrated the ability of polyphenols to improve cognitive performance and memory [17,18] and, more recently, these results have been replicated in human studies [19,20]. Regarding mental health, a growing body of data from animal and human studies has emerged supporting the role of a variety of dietary polyphenols in affecting behaviour and mood through anxiolytic and antidepressant-like properties, mediated through multiple molecular and cellular pathways [21]. Moreover, given that recent studies have demonstrated the pathophysiological role of oxidative stress and inflammation in the onset and progression of depression, polyphenols have been examined both in vitro and in vivo as a potential antidepressant treatment, although randomised controlled trials are still scarce in the field [22,23]. The richest sources of polyphenols in the human diet include fruits (e.g., berries, grapes, apples and plums), vegetables (e.g., cabbage, eggplant, onions, peppers), plant-derived beverages including tea, coffee, red wine and fruit juices (e.g., apple juice), seeds, nuts and chocolate (particularly dark chocolate) [24,25]. In terms of a food-based approach, several of the above-mentioned foods have been studied both in observational and intervention studies for potential effects on outcomes related to mental well-being, mood, psychological distress and life satisfaction [26], although, potentially due in part to the great variation in study design, results are not consistent. Studying diet on a dietary pattern level will be beneficial in allowing potential complicated or cumulative intercorrelations, interactions and synergies to be revealed, given that different polyphenols may have different effects on outcomes of mental health [27–29].
The aim of the present study was to assess the effect of a polyphenol-rich dietary pattern (comprising fruits, including berries; vegetables and dark chocolate) in comparison to a control diet (low fruits and vegetables, <2 portions/day, and no dark chocolate) on aspects of psychological well-being and mental health status including mood, QoL, body image perception and self-esteem as secondary outcomes measured within the Polyphenol Intervention Trial (PPhIT) [30].
### **2. Materials and Methods**
### *2.1. Setting and Study Population*
PPhIT was a randomised, controlled, parallel-group, single-blinded dietary intervention trial, primarily designed to test whether increasing overall polyphenol dietary intake would affect microvascular function and a range of other markers of CVD risk, such as systolic blood pressure and lipid profile, in patients with hypertension. All participants underwent a full assessment at baseline (week 0) (described below); then, they entered a washout period, during which they consumed a low polyphenol diet, and afterwards were randomised to either a low polyphenol diet (LPD) or a high polyphenol diet (HPD) group for 8 weeks (Figure 1). A full assessment was repeated for all the participants at the end of the 8-week intervention (week 12), while at the end of the washout period (week 4), participants also underwent a dietary intake assessment, anthropometric measurements and blood and urine sample collection.
**Figure 1.** CONSORT diagram summarizing flow of participants through the study.
Participants aged 40–65 years, with documented grade I (140–159/90–99 mmHg) or grade II (160–179/100–109 mmHg) hypertension, were eligible. Participants with diabetes mellitus, acute coronary syndrome or transient ischaemic attack within three months, pregnancy or lactation, fasting triglyceride concentration>4 mmol/L, alcohol consumption (>28 units/week for men and>21 units/week for women), oral anticoagulant therapy or antioxidant supplements, dietary restrictions that would limit ability to comply with the study diets, body mass index >35 kg/m<sup>2</sup> or with an impalpable brachial artery were excluded from the study. Recruitment for PPhIT began in February 2011 and was completed by January 2013. All participants were informed about the aims and procedures of the study and gave their written consent. The study had ethical approval from the Office of Research Ethics Committee Northern Ireland (ref 10/NIR03/39) and was registered at ClinicalTrials.gov (ref NCT01319786). Details of the primary aim of the study, population, design, recruitment procedures and main findings have been published elsewhere [30]. Below, we provide some additional details on selected aspects of the evaluation that pertain to the analyses reported in this manuscript.
### *2.2. Dietary Intervention*
The intervention commenced with a four-week "washout period" for all participants, during which they were asked to consume two portions or less of fruits and vegetables (F&V) per day and to exclude berries and dark chocolate (LPD). At the end of this period, subjects were randomised to either continue with the above LPD for a further 8-week "intervention period" or to consume an HPD of six portions of F&V (including one portion of berries per day) and 50 g of dark chocolate per day (Figure 1). A portion of fruit and vegetables was quantitatively defined using household measures as outlined by UK guidelines (https://www.nhs.uk/live-well/eat-well/5-a-day-portion-sizes/), i.e., 1 apple, 1 orange, half a grapefruit or one glass (150 mL) fruit juice, 3 tablespoons of vegetables [31]. All participants in the HPD had a self-selected weekly delivery of F&V and dark chocolate (Lindt® 70% cocoa) free of charge to their homes from a local supermarket and were provided with written material regarding
F&V portion sizes, recipes and sample diet plans. In addition, each participant, regardless of dietary allocation, was also contacted by telephone at weekly intervals to provide support and encouragement and to discuss potential barriers encountered in relation to achieving the dietary goals.
Dietary intake and compliance with the intervention were assessed through 4-day food diaries completed on four occasions: on the four days leading up to the week 0 visit (baseline measurement), on the four days leading up to the week 4 visit (washout period measurement), at week 8 (intervention measurement) and on the four days leading up to the final week 12 visit (a second intervention period measurement). Circulating blood and urine levels of a panel of nutritional biomarkers, with detailed methodology given below, were also used to assess compliance. Self-reported F&V, berries and dark chocolate consumed per day (as recorded in the food diaries) were extracted and entered into a Microsoft Excel spreadsheet. The spreadsheet contained pre-determined formulae which transformed the actual amounts of F&V and berries consumed into "portions" according to the "5-a-day" message.
### *2.3. Other Lifestyle Parameters*
A "lifestyle and medical" questionnaire was used at week 0 to record participant demographic, lifestyle and medical information. The questionnaire had 16 items in total and assessed several aspects including vitamin and mineral supplement usage, smoking and alcohol habits, history of education, current occupational status, current medication, history of steroid use and, for females, use of hormone replacement therapy and details of menstrual cycle. Information regarding changes in medication use, smoking and alcohol patterns, as well as infections/illnesses were also recorded throughout the study.
Participants' physical activity levels were recorded at weeks 0, 4 and 12 to ensure that habitual activity levels were not altered for the duration of the study. The Recent Physical Activity Questionnaire (RPAQ), designed by the Medical Research Council (MRC Epidemiology Unit, Cambridge, UK), was used to measure physical activity. The questionnaire assesses physical activity within the preceding four weeks based on three primary areas: activity at home, activity at work (including travel to and from work) and recreational activities. The RPAQ has been shown to be a valid instrument for calculating total energy expenditure, physical activity energy expenditure and physical activity in healthy adults [32]. In terms of analysis, physical activity as recorded in the questionnaire was converted to total metabolic equivalent of task (MET) hours per day of sedentary, light, moderate and vigorous activity.
### *2.4. Anthropometric, Clinical and Biochemical Assessments*
Participants attended the Royal Victoria Hospital (Belfast, Northern Ireland, UK) for assessments on three occasions throughout the study: baseline (week 0), washout period (week 4) and intervention (week 12). Body weight of participants was measured to the nearest 100 g and height to the nearest 0.5 cm. Body mass index (BMI) was calculated as weight (kg) divided by height squared (m2). Waist circumference (WC) and hip circumference (HC) were tape-measured to the nearest 0.1 cm. Blood pressure (BP) was measured using an Omron M5-1 automatic BP monitor (Omron Healthcare, Hoofddorp, The Netherlands). Three consecutive readings were recorded, and a mean BP was calculated from the 2nd and 3rd readings. To measure the primary endpoint of PPhIT (microvascular function), venous occlusion plethysmography was conducted on participants by determining forearm blood flow during incremental intra-arterial infusions of acetylcholine and sodium nitroprusside, as previously described [30]. Blood samples were also collected. Fasting serum lipid profiles (total cholesterol, high density lipoprotein (HDL) and triglycerides) were assessed using standard enzymatic colorimetric assays on an automated Cobas® 8000 Modular system biochemical analyser (Roche Diagnostics Ltd., West Sussex, UK). Low-density lipoprotein (LDL) cholesterol was calculated using a standard Friedewald formula [33]. Blood and urine markers of micronutrient status were assessed at weeks 0 and 12 to objectively measure compliance to the intervention diet. Plasma vitamin C was measured on a FLUOstar Optima plate reader (BMG Labtech, Ortenberg, Germany), adapted from the method of Vuilleumier and Keck [34]. Serum concentrations of six carotenoids
(α-carotene, β-carotene, β-cryptoxanthin, lutein, lycopene and zeaxanthin) were measured by reverse phase high performance liquid chromatography (HPLC), as described by Craft [35]. Urine collected from the volunteers between evening meal and midnight the evening before each study visit was analysed, including an enzymatic hydrolysis step, to quantify total epicatechin content, using an Agilent Technologies 1100 series HPLC (Agilent Technologies, Stockport, UK) directly linked to a Waters Micromass Quattro Ultima Platinum API triple quadrupole mass spectrometer (Waters, Dublin, Ireland).
### *2.5. Psychological Well-Being, Self-Esteem and Body Image*
Aspects of psychological well-being and mental health status were evaluated through several scales and questionnaires that were completed at weeks 0 and 12. The decision to use these questionnaires only twice was made for three reasons: (i) the study wished to investigate the effect of the intervention diet in comparison to normal psychological well-being, rather than psychological well-being under the controlled conditions of the washout period; (ii) to reduce participants' burden at week 4 visits, which were already long (2.5 h) in duration due to vascular function and dietary assessments; (iii) distributing the surveys at three time points may have been disadvantageous in terms of allowing participants to become familiar with their format, which may have influenced responses. All questionnaires are commonly used for assessing various aspects of mental health and psychological well-being in the general population.
The Positive and Negative Affect Schedule (PANAS) was used for evaluating subjective mood. The questionnaire measures two distinctive dimensions: positive affect (PA) and negative affect (NA) [36]. PA is associated with pleasurable engagement with the environment, including feelings of enthusiasm and alertness as well as feeling active. NA refers to unpleasurable engagement with the environment, comprising feelings of anger, contempt, disgust, guilt, fear and nervousness. Whilst related, PA and NA represent two distinct and independent dimensions of mood. Participants were asked to respond to 10 items representing PA and 10 items representing NA on a five-point scale. Higher scores represent higher positive and negative affect, respectively. This was the only questionnaire assessing psychological well-being that was also completed at week 4, in order to monitor psychological well-being at the end of the washout period.
Depressive symptomology was assessed with the Beck Depression Inventory-II (BDI-II), a 21-item, self-report questionnaire developed by Beck and colleagues [37]. Each item on the BDI-II has four statements which relate to the severity of a particular depressive symptom, and respondents are asked to choose the one statement which best describes how they have been feeling in the preceding two weeks. Higher scores indicate higher levels of depression (scores 0–13 = minimal; 14–19 = mild; 20–28 = moderate; 29–63 = severe). The shorter version (21 items) of the Depression Anxiety Stress Scale (DASS-21) was also completed to measure depression, anxiety and stress [38]. The DASS-21 questionnaire was introduced nine months into the recruitment of the participants. DASS-21 has seven items per subscale and asks participants to rate the extent to which they experienced each emotional state the preceding week using a four-point Likert scale (0 = Did not apply to me at all, 3 = Applied to me very much or most of the time). Higher scores are indicative of higher levels of depression, anxiety and stress.
The Rosenberg Self-Esteem Scale was used as a global measure of self-esteem [39,40]. The questionnaire consists of a ten-item Likert scale, completed using a four-point scale from strongly agree to strongly disagree. Scores can range from 0 to 30, with higher scores indicating higher self-esteem. Finally, body image satisfaction was assessed through the Multidimensional Body Self-Relations Questionnaire—Appearance Scales (MSRQ-AS), a 34-item validated measure of body image perception for use in general populations (www.body-images.com) [41]. This version contains five subscales: appearance evaluation (satisfaction with ones looks), appearance orientation (levels of investment in one's appearance), overweight preoccupation (weight anxiety, vigilance, dieting etc), self-classified weight (how one perceives and labels one's weight) and body area satisfaction (satisfaction with areas of body). The questionnaire contains a series of statements and asks participants to indicate
the extent to which each statement applies to them personally, with higher scores generally indicating greater body image satisfaction.
### *2.6. Mental and Physical Health*
Mental and physical health were assessed with the RAND Medical Outcomes Study 36-item Short Form Health Survey (SF-36) [42,43]. A total of 36 questions are included in the RAND SF-36 survey and eight key areas are explored in the SF-36 including physical functioning, role limitations due to physical problems, pain, general health, energy/fatigue, social functioning, role limitations due to emotional problems and emotional well-being. The raw data were recoded using the RAND SF-36 scoring instructions available online. Additionally, the eight areas were combined to obtain the scoring for physical and mental health components. As the eight different components consist of different numbers of questions, the normal scores were transformed to T-scores, as described by Hays et al. 1993 [44] and Hays et al. 1995 [45]. Physical functioning, role limitations due to physical problems, pain and general health were combined to obtain the physical health component and role limitations due to emotional problems, energy/fatigue, emotional well-being and social functioning were combined to obtain the mental health component.
### *2.7. Statistical Analyses*
The sample size calculation was based on the PPhIT primary outcome, namely microvascular function. According to this, detection of a 33% difference between groups in microvascular function, measured by forearm blood flow responses to an endothelium-dependent vasodilator, with 90% power, using a 2-tailed test at the 5% significance level, would require 50 participants per group. The current analysis reports secondary outcomes, for which power calculations were not performed.
Results are expressed as mean ± standard deviation for normally distributed continuous variables and as medians and interquartile ranges for continuous skewed variables. Categorical variables are presented as absolute (*n*) and relative frequencies (%). The normality of variables was checked through the Shapiro–Wilk test and graphically through histograms. Concentration measures of micronutrients were logarithmically transformed and were summarised as geometric median and interquartile range. The principal analysis for each outcome variable was a between-group comparison of change using independent sample t tests or Mann–Whitney U test for continuous parametric and non-parametric variables, respectively, and chi-squared test for categorical variables. Within-group comparisons were performed using paired sample t tests or Wilcoxon signed-rank test for parametric and non-parametric continuous variables, respectively. Statistical significance was set at *p* ≤ 0.05. All statistical methods were conducted using PASW Statistics 18 for Windows (SPSS Inc., Chicago, IL, USA).
### **3. Results**
### *3.1. General Results*
Ninety-nine participants completed the PPhIT study, including 53 (53.5%) males. Participants had a mean age of 54.9 ± 6.9 years, with ages ranging from 40 to 65 years. The majority (52%) of the sample were obese (BMI <sup>≥</sup> 30 kg/m2). In total, 12.1% were current smokers, and 43.4% stated that they had smoked in the past. Baseline characteristics according to dietary group (LPD versus HPD) are shown in Table 1. Overall, the groups were similar upon entering the study, with no statistically significant differences in anthropometric, lifestyle and basic clinical characteristics.
**Table 1.** Baseline (week 0) participant characteristics according to the Polyphenol Intervention Trial (PPhIT) study group allocation.
HDL—high-density lipoprotein; LDL—low-density lipoprotein. Continuous variables are summarised as mean ± SD or medians and interquartile ranges. Categorical variables are summarised as *n* (%). \* Between-group comparisons were made using independent sample t-tests (*p* < 0.05) or Mann–Whitney U test for continuous variables and chi-squared tests (*p* < 0.05) for categorical variables.
During the washout period, no changes were recorded in participants' physical activity habits, weight status, smoking habits, medication use or clinical condition compared to baseline in both HPD and LPD groups (data not shown). Additionally, mood evaluation according to PANAS questionnaire did not record any change between baseline and end of washout period (both *p* > 0.05) (data not shown). F&V intake per day declined significantly during washout, from 2.67 portions at week 0 to 1.38 portions at week 4 within the overall sample (*p* < 0.001), and significant reductions in blood levels of vitamin C (*p* < 0.001) and β-cryptoxanthin (*p* = 0.05), but not in any of the other carotenoids measured, were also recorded (data not shown).
Dietary intake of food groups and micronutrients, as well as weight status and physical activity levels both at baseline and at the end of the intervention period, are presented in Table 2, per intervention group. At baseline, there was no significant difference in intake of F&V, berries and dark chocolate and concentration of micronutrients between the LPD and HPD group. By the end of the intervention, there was a significant increase in intake of F&V, berries and dark chocolate in the HPD group, and the differences in change in intake between the two groups were statistically significant. Furthermore, there was a significant increase in the concentration of biomarkers, plasma vitamin C, serum lutein, β-cryptoxanthin, α-carotene and lycopene and urinary epicatechin over the course of the intervention in the HPD group, and the differences in the change in the concentration between the LPD and HPD group were statistically significant. These results indicate good compliance with the intervention diet, with significant between-group differences in change in all biomarkers measured except β-carotene. No differences were recorded in change in physical activity and weight status between the two intervention groups during the intervention.
independent
geometric median change (IQR).
sample t-tests and
Mann–Whitney
U test (*<sup>p</sup>* < 0.05); 6 all variables are
logarithmically
transformed
and summarised
as geometric medians (IQ range) and change as
### *3.2. Changes in Aspects of Psychological Well-Being*
Changes in measures of psychological well-being between baseline and intervention are illustrated in Table 3. There were no significant differences in scores of BDI-II, DASS-21 or PANAS between the LPD and HPD groups at baseline. There was a significant between-group difference (*p* = 0.01) in change in depressive symptoms as assessed with BDI-II, but no other significant effects were found between groups with regards to depression, anxiety or stress measured using the DASS-21 or positive and negative affect measured with PANAS. Regarding within-group changes, a borderline significant (*p* = 0.05) result was detected for a reduction in stress measured by DASS-21 within the HPD group, as well as an improvement in subjective mood (positive affect) (*p* = 0.03) measured by PANAS.
### *3.3. Changes in Self-Esteem and Body Image Perception*
There were no significant differences in self-esteem or body image perception scores between the LPD and HPD groups at baseline. As shown in Table 3, there were also no significant differences between the HPD and LPD in self-esteem or body image perception scores at the end of the intervention.
### *3.4. Changes in Health-Related Quality of Life*
There were no significant differences between groups at baseline with regards to health-related quality of life measured using the SF-36. There were statistically significant between-group differences in change in different component scores (general health (*p* = 0.03) and energy/fatigue (*p* = 0.02)) and the overall summary scores for the physical health component (*p* = 0.04) and mental health component (*p* = 0.01), with more positive changes demonstrated in the HPD group. In the HPD group, there were also within-group improvements in role limitations due to physical health (*p* = 0.04), general health (*p* = 0.00), energy/fatigue (*p* = < 0.001), emotional well-being (*p* = < 0.001) and social functioning (*p* = 0.02)
Data are presented as mean ± SD or medians and interquartile ranges (IQR). There were no significant between-group differences in baseline values; median change was calculated as week 12- week 0 and is presented as median change (IQR); 3 within-group comparisons were performed using paired sample t test or Wilcoxon signed-rank test (*<sup>p</sup>* < 0.05); 4 between-group comparisons were made using Mann–Whitney U test for continuous variables (*<sup>p</sup>* < 0.05); 5 *n* = 57 (LPD (*<sup>n</sup>* = 27), HPD (*<sup>n</sup>* = 30)); \* BDI-II; Beck Depression Inventory Second Edition. Scores 0–13 = minimal depression, 14–19 = mild depression, 20–28 = moderate depression, 29–63 = severe depression; \*\* DASS-21; Depression Anxiety and Stress Scale 21 items. Depression score 0–9 = normal, Anxiety score 0–7 = normal, Stress score 0–14 = normal; \*\*\* PANAS; Positive and Negative Affect Scale. Higher score indicates higher positive and negative affect; † Rosenberg Self-Esteem Score; scores range from 0 to 30. Higher scores are indicative of higher self-esteem; †† MBSRQ-AS; Multi-Dimensional Body-Self Relations Questionnaire—Appearance Scales. Higher scores indicative of higher body image satisfaction; ††† Mental and physical health assessed using the RAND 36-Item Short Form Survey (SF-36). Physical functioning scores: "low" = limited a lot in performing all physical activities including bathing or dressing, "high" = performs all types of physical activities including the most vigorous without limitations due to health; Role limitations due to physical problems: "low" = problems with work or other daily activities as a result of physical health, "high" =no problems with work or other daily activities as result of physical health, past 4 weeks; Pain: "high" = very severe and extremely limiting pain, "low" = no pain or limitations due to pain, past 4 weeks; General health perceptions: "high" = believes personal health is poor and likely to get worse, "low" = believes personal health is excellent. Physical health component =sum of physical functioning, role limitations—physical health, pain and general health. Role limitations due to emotional problems: "high" = problems with work or other daily activities as a result of emotional problems, "low" = no problems with work or other daily activities as result of emotional problems, past 4 weeks; Energy/fatigue: "low" = feels tired and worn out all of the time, "low" = feels full of pep and energy all of the time, past 4 weeks; Emotional well-being: "high" = feelings of nervousness and depression all of the time, "low" = feels peaceful, happy and calm all of the time, past 4 weeks; Social functioning: "low" = extreme and frequent interference with normal social activities due to physical and emotional problems, "high" = performs normal social activities without interference due to physical or emotional problems, past 4 weeks. Mental health component = sum of role limitations—emotional health, energy/fatigue, emotional well-being and social functioning.
### **4. Discussion**
Given the high prevalence of mental health problems and the potential effect of dietary patterns on their onset and/or treatment, the aim of the present study was to assess the effect of a polyphenol-rich dietary pattern (comprising fruits, including berries; vegetables and dark chocolate) on aspects of psychological well-being or mental health status, including mood, self-esteem and body image perception, as secondary outcomes of the PPhIT study. Despite some heterogeneity, the study findings suggest that the adoption of such a polyphenol-rich diet could potentially lead to beneficial effects on certain outcomes including depressed mood and physical and mental health in hypertensive participants.
There was a significant difference in change in depressive symptoms assessed with BDI-II between the HPD group and the LPD group, indicating a positive effect of the HPD, which is in agreement with a number of other observational studies focusing on the same outcome measure and polyphenol-rich foods. In the HPD group, a 66.6% reduction in BDI-II score was observed after the intervention. Button et al. 2015, using data from three randomised controlled trials (RCT) with a sample of *n* = 1039, identified that a 17.5% reduction in score was necessary to observe minimally clinically important differences [46]. Oliveria et al. (2019) found a negative association between depressive symptoms measured by BDI and high intake of polyphenol food items [47]. In the Finnish general population (*n* = 2011), daily intake of tea was associated with reduced risk of depressive symptoms defined by BDI scores [48]. Similarly, in the Mediterranean healthy eating, lifestyle and aging (MEAL) study (*n* = 1572), the dietary intake of phenolic acid, flavanones and anthocyanin were negatively associated with depressive symptoms measured using the Center for Epidemiologic Studies Depression Scale (CES-D-10) [49]. The positive effects observed in the present study may be attributable to other nutrients found in F&V, berries and dark chocolate which may work independently or synergistically to influence health outcomes. Brody (2002) found that vitamin C intake over a 14 day period led to a moderate reduction in depressive symptoms amongst 42 healthy young adults [50]. In our study, there was a significant difference in plasma vitamin C status between the LPD and HPD group. Similarly, there were significant increases in serum carotenoids, lutein, zeaxanthin, β-cryptoxanthin and urinary epicatechin within the HPD group, and some studies have suggested a link between these nutrients and improvements in psychological well-being including depression [51]. The antidepressant effects of polyphenols may be associated with both their antioxidant and anti-inflammatory properties, whereby there is a reduction in free radicals and cytokine dysregulation [12]. Lua and Wong (2012) found that the consumption of 50 g dark chocolate (70% cocoa) for three days was associated with significant improvement in depressed mood [52].
The primary outcome of the PPhIT study was to identify whether high consumption of F&V, berries and dark chocolate could improve microvascular function in hypertensive subjects [30]. High intake of polyphenol, specifically including F&V, berries and dark chocolate in diet, resulted in significant improvements in endothelium-dependent (acetylcholine) vasodilator [30]. Depression is often observed among individuals with vascular diseases such as hypertension, peripheral vascular disease and coronary artery disease, known as "vascular depression hypothesis" [53]. Studies have reported morphological changes in vascular structure and altered expression of endothelial cell molecules such as nitric oxide in patients with depression [53]. In the current study, the improvements in endothelium-dependent vasodilatation might have also resulted in improvements in depressed mood.
In light of the findings from the BDI-II, it is interesting that no notable effects of the polyphenol-rich diet were observed on depressed mood measured using the DASS-21 questionnaire in this study. The DASS-21 questionnaire was introduced as an amendment to PPhIT, given concerns that BDI-II is used to screen for depression in normal populations or to assess severity of depression in clinical populations, and therefore it was thought possible by the research team that the tool may not have been sensitive enough to pick up changes due to diet. Page et al. (2007) showed that DASS-21 has good psychometric properties and is moderately sensitive to changes that result from the treatment [54]. However, this resulted in a considerably smaller sample size for the analysis of the
DASS-21 questionnaire (*n* = 57) compared to BDI-II, which may have had implications in terms of the associated power available to detect differences between the two diet groups. The DASS-21 also showed no statistically significant differences in change between groups with regards to stress or anxiety. Furthermore, for both measures, scores on all scales at the start of the study are low, and negative affect scores for the PANAS are also low. These low scores are unsurprising in a volunteer sample for a study intended to improve health but may also have limited our chances of finding effects. Further study in groups with higher levels of poor psychological health, e.g., those with diagnoses of depression or anxiety, may be of value.
In the present study, significant improvements in quality of life between the HPD group and the LPD group measured using the SF-36 health survey questionnaire were found. There were statistically significant improvements in both physical and mental health components in the HPD group when compared with the LPD group. Data showing the effect of dietary interventions and especially of polyphenols/antioxidants on quality of life parameters are sparse and mainly limited to patients with chronic diseases such as multiple sclerosis, chronic fatigue syndrome and depression. Steptoe et al. (2004) found that a higher intake of fruits and vegetables through behavioural and nutrition education counselling was positively associated with physical health status but not mental health status measured using SF-36 among adults in a low-income neighbourhood [55]. A sub-study of the DASH trial also found that adhering to a fruit and vegetable-rich diet was associated with improved perception of quality of life [56]. It is important to acknowledge that while the self-reported improvements in physical and mental health scores observed within the current sample may be attributed to the foods consumed, they may also be wholly or partly influenced by taking part in the intervention and increased positivity that may come from making positive dietary changes. As pointed out in the study by Plaisted et al. (1999), improvements in QoL might be attributable to participants' awareness that they are consuming a healthy diet, which could have contributed to improved self-ratings of general health and mental health component [56]. In addition, given that depressive symptoms were improved in the HPD group, the improvements in mental health component may simply mirror these findings.
It is important to consider the results of this study in light of a number of methodological limitations. Firstly, as the primary purpose of PPhIT was to test the effect of a polyphenol-rich dietary pattern on microvascular and platelet function, the outcomes discussed here are secondary endpoints. Hence, as mentioned previously, it is possible that the study may not have been adequately powered to detect differences between the dietary groups, which may account for some of the null findings demonstrated. Secondly, the study sample comprised mildly hypertensive participants, which limits the generalisability of these results to the wider population. Furthermore, it is possible that selection bias exists within the current sample, given that the volunteers for this study were on the whole well-educated, and, as is the case with most clinical trials, are likely to have been more motivated with regard to improving their health than the general population. The participants in the HPD group were provided with the key intervention foods on a weekly basis, whereas the LPD group received no food provision as their diet was to remain unchanged. This may have increased the likely compliance of the HPD group with the intervention. Another limitation of this study was the use of self-report measures to measure psychological outcomes. Self-report measures can be disadvantageous in that they can be affected by forms of bias, including response, recall and social desirability bias, which can lead to inaccurate responses and conclusions [57]. However, given the subjective nature of psychological well-being, self-reporting is the most suitable method of obtaining information on individuals' personal experiences and emotions. The current study employed validated and previously used measures to collect information on individuals' personal experiences and emotions [58,59]. Additionally, it must be noted that the questionnaires described in this study were distributed at week 0 (baseline) and week 12 (intervention). It is possible that the washout period (week 0 to week 4) could have potentially affected people's psychological well-being and thus it may have been useful to additionally measure the endpoints at week 4. However, the decision to distribute the questionnaires at week 0 and week 12 was made for three main reasons: (i) to reduce participant burden at week 4 visits, which were already long
(2.5 h) in duration due to vascular function assessment and the dissemination of dietary advice; (ii) the study wished to investigate the effect of the intervention diet in comparison to normal psychological status, rather than psychological states under controlled conditions, which would have limited the applicability of the results; (iii) distributing the outcome measures at three time points may have been disadvantageous in terms of allowing participants to become familiar with their format, which may have induced response bias. Another limitation common to most studies analysing self-reported questionnaire data is the number of variables assessed, which may have increased the chance of type one errors (identification of the false positive) associated with hypothesis testing.
In contrast, one of the most obvious strengths of this study is its RCT study design. However, as the randomisation according to the groups only occurred at week 4, the presentation of week 0 data based on the allocated groups is rather artificial, and this must be considered a limitation. As further strengths, the study implemented a variety of techniques to encourage and monitor compliance. As a result of such efforts, participants were demonstrated to have good compliance with both diets, which was assessed both subjectively and objectively. Furthermore, the study had good retention of participants, with a less than 5% (*n* = 5) drop out level, all of which were due to reasons unrelated to the study.
### **5. Conclusions**
In conclusion, the results from the present RCT trial showed heterogeneous findings regarding the effect of a polyphenol-rich dietary pattern on aspects of psychological well-being, with positive effects demonstrated on depressive symptoms and both the physical and mental health status components of the SF-36 quality of life measure. Further studies with psychological well-being impacts as primary endpoints, with appropriate study design and sample sizes, are needed in order to confirm the benefits of a polyphenol-rich dietary pattern on these outcomes.
**Author Contributions:** Conceptualization, C.R., R.L.N., K.M.A., I.S.Y., M.C.M., P.P.M. and J.V.W.; formal analysis, M.D.K., A.V., C.R., D.M. and J.V.W.; investigation, C.R. and R.L.N.; writing—original draft, M.D.K., A.V., C.R. and J.V.W.; writing—review and editing, M.D.K., A.V., C.R., R.L.N., K.M.A., D.M., M.D., I.S.Y., M.C.M., P.P.M. and J.V.W. All authors have read and agreed to the published version of the manuscript.
**Funding:** This study was funded by a Northern Ireland Health and Social Care Research and Development doctoral fellowship award (ref: EAT/4195/09) and Northern Ireland Chest Heart and Stroke scientific research grant (ref: 2010\_17).
**Acknowledgments:** We acknowledge Margaret Cupples and the Northern Ireland Clinical Research Network (Primary Care) for their assistance with participant recruitment, Lesley Hamill for conducting the epicatechin analysis and Sarah Gilchrist for performing vitamin C and carotenoid analysis. We also thank all of the participants in the study for their time, interest, cooperation and contribution to the research.
**Conflicts of Interest:** The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, and in the decision to publish the results.
### **References**
© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
*Article*
| doab | 2025-04-07T03:56:58.603074 | 1-5-2021 17:49 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "003cf560-3166-422f-9a1c-98b1e971efa6",
"url": "https://mdpi.com/books/pdfview/book/3241",
"author": "",
"title": "Dietary Polyphenols and Human Health",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783039437931",
"section_idx": 7
} |
003cf560-3166-422f-9a1c-98b1e971efa6.8 | **Acute E**ff**ects of a Polyphenol-Rich Leaf Extract of** *Mangifera indica* **L. (Zynamite) on Cognitive Function in Healthy Adults: A Double-Blind, Placebo-Controlled Crossover Study**
### **Emma L. Wightman 1,2, Philippa A. Jackson 2, Joanne Forster 2, Julie Khan 2, Julia C. Wiebe 3, Nigel Gericke 3,4 and David O. Kennedy 2,\***
Received: 17 June 2020; Accepted: 21 July 2020; Published: 23 July 2020
**Abstract:** Extracts made from the leaves of the mango food plant (*Mangifera indica* L., *Anacardiaceae*) have a long history of medicinal usage, most likely due to particularly high levels of the polyphenol mangiferin. In rodent models, oral mangiferin protects cognitive function and brain tissue from a number of challenges and modulates cerebro-electrical activity. Recent evidence has confirmed the latter effect in healthy humans following a mangiferin-rich mango leaf extract using quantitative electroencephalography (EEG). The current study therefore investigated the effects of a single dose of mango leaf extract, standardised to contain >60% mangiferin (Zynamite®), on cognitive function and mood. This study adopted a double-blind, placebo-controlled cross-over design in which 70 healthy young adults (18 to 45 years) received 300 mg mango leaf extract and a matched placebo, on separate occasions, separated by at least 7 days. On each occasion, cognitive/mood assessments were undertaken pre-dose and at 30 min, 3 h and 5 h post-dose using the Computerised Mental Performance Assessment System (COMPASS) assessment battery and the Profile of Mood States (POMS). The results showed that a single dose of 300 mg mango leaf extract significantly improved performance accuracy across the tasks in the battery, with domain-specific effects seen in terms of enhanced performance on an 'Accuracy of Attention' factor and an 'Episodic Memory' factor. Performance was also improved across all three tasks (Rapid Visual Information Processing, Serial 3s and Serial 7s subtraction tasks) that make up the Cognitive Demand Battery sub-section of the assessment. All of these cognitive benefits were seen across the post-dose assessments (30 min, 3 h, 5 h). There were no interpretable treatment related effects on mood. These results provide the first demonstration of cognition enhancement following consumption of mango leaf extract and add to previous research showing that polyphenols and polyphenol rich extracts can improve brain function.
**Keywords:** cognition; attention; memory; brain; polyphenols; mangiferin; mango leaf extract
### **1. Introduction**
The roots, leaves, fruit and bark of the food plant *Mangifera indica* (mango) have a long history of therapeutic use within traditional medicinal systems for a wide range of conditions. For example, extracts, teas and infusions made from mango leaves have been used for the treatment of diabetes, malaria, diseases of the digestive system, lungs, and kidneys, and as a topical treatment for wounds and burns [1]. The bioactivity of mango leaf extracts may be due to particularly high levels [2] of xanthones. This group of polyphenols are found in a restricted group of plant species [3], including members of the *Hypericum* genus that provide us with a number of medicinal herbal extracts [4], but they are rarely consumed in the diet, with only a few exceptions other than mango itself (e.g., [5]). The predominant member of this structural group in mango leaf is mangiferin, a xanthone glucoside that has been shown to have potential anti-inflammatory, antioxidant, immunomodulatory, neuroprotective, antiproliferative, antidiabetic, DNA protective, and hypoglycaemic properties [6–10].
Whilst structurally distinct from the flavonoids and other polyphenols that are ubiquitous in plant derived foods, mangiferin [8,11–14] likely owes its beneficial bioactivity to some similar mechanisms of action as found in the wider polyphenol group class [15], including interactions with, and modulation of, diverse components of a wide range of mammalian cellular signal transduction pathways. These pathways, in turn, control gene transcription and a plethora of cellular responses, including cell proliferation, apoptosis, and the synthesis of growth factors, and vasodilatory and inflammatory molecules. In the central nervous system, specific additional interactions attributed to polyphenols include direct neurotransmitter and neurotrophin receptor and signalling pathway interactions, and increased synthesis of neurotrophins and vasodilatory molecules, which, in turn, foster angiogenesis/neurogenesis [15–20]. These mechanisms potentially underlie the observation of consistent beneficial cardiovascular effects from meta-analyses of multiple intervention studies [21–23], and demonstrations of improved cognitive function [24–28], following diverse polyphenols.
In line with these mechanistic cellular effects, rodent studies have demonstrated that a single administration of mangiferin can improve memory in uncompromised rats [29] and that either single doses or extended supplementation with mangiferin can attenuate the memory deficits or depressive/anxiety behaviours associated with a range of brain insults and challenges. This includes the cholinergic antagonist scopolamine [30], sleep deprivation [31], the injection of lipopolysaccharides [32] and aluminium chloride-induced neurotoxicity in mice [9]. Consistent ex vivo evidence focussing on the hippocampus also shows that mangiferin can protect rodent neuronal tissue from the increase in inflammatory cytokines [9,30–32] and the decrease in neurotrophins such as brain-derived neurotrophic factor (BDNF) [9,31], associated with multifarious brain insults. Similarly, mangiferin has been shown to protect the rodent brain from lead-induced structural damage and decrease oxidative stress via interactions within the Nrf2 signalling pathways in rats [10].
A number of recent studies have assessed the potential efficacy of a mango leaf extract standardized to a minimum of 60% mangiferin (Zynamite®). In terms of physical performance, several of these studies have assessed the ergogenic effects in humans of both acute [33–36] and longer-term supplementation [34] with this mango leaf extract combined with the polyphenols luteolin or quercetin. This research has demonstrated an improved performance during high intensity exercise [33–35], increased brain oxygenation [33,34], maximal aerobic capacity [33], increased muscle oxygen extraction [34,35] and the attenuation of muscle damage and improvements in the time course of decreased muscle performance [37].
With regard to brain function, in rats, oral administration of mango leaf extract attenuated electroencephalography (EEG) power measured via implanted electrodes (frontal cortex, hippocampus, striatum, reticular formation) across the spectra and brain regions under investigation, with the most striking findings in the alpha and beta wavebands. These effects were synergistically increased by the co-administration of caffeine. A concomitant ex vivo study also demonstrated that 7 days supplementation with the mango leaf extract lead to increased hippocampal pyramidal cell excitability [38]. In a subsequent multi-disciplinary series of studies [39], both the ex vivo hippocampal excitability and the attenuation of EEG spectral power across brain regions in rats were confirmed both for mango leaf extract and mangiferin, confirming this polyphenol as the likely active component of the extract. In two subsequent pilot studies (also reported in [39]), both involving 16 healthy young
humans, quantitative EEG was employed at rest and during cognitive task performance 90- and 60-min post-dose respectively. In the first study, in comparison to control, mango leaf extract resulted in modest reductions in 'eyes open' power in delta and theta power, and a more pronounced increase in power during cognitive task performance, with significant increases in all wavebands across scalp electrodes interrogating the association cortex. These results were supported by more modest EEG changes in the second study, but no evidence of a synergistic relationship with caffeine. Cognitive task performance and mood were not significantly modulated by mango leaf extract.
The extant literature demonstrating functional benefits following polyphenol consumption, and the previous rodent and pilot human studies assessing the effects of mangiferin and mango leaf extract described above, suggest that a mango leaf extract with high levels of the polyphenol mangiferin may exert beneficial effects on human brain function, including the enhancement of cognitive function. The current exploratory, double-blind, placebo-controlled, balanced crossover study therefore assessed the effects of a single dose of mango leaf extract (Zynamite®) on cognitive function and psychological state 30 min, 3 h and 5 h post-dose in a large sample of healthy adults.
### **2. Methods**
### *2.1. Design*
This study adopted a randomised, double-blind, placebo-controlled, balanced crossover design, in which the acute effects of a single dose of 300 mg mango leaf extract and placebo were assessed on cognitive function and psychological state/mood at 30 min, 3 h and 5 h post-dose. All study procedures were reviewed and approved by Northumbria University's Department of Psychology Ethics Committee (Ref: 17741) and were conducted according to the principles of the Declaration of Helsinki. The trial was pre-registered at ClinicalTrials.gov (NCT04299217).
### *2.2. Participants*
The required sample size for the study (N = 72) was calculated (GPower 3.0) on the basis of delivering adequate power (0.8) to detect a small effect size (f = 0.1). The power to detect the anticipated medium effect size (f = 0.25) exceeded 0.95.
A total of 75 participants were randomised. Three participants subsequently withdrew from the study after completing one testing visit. Two participants were removed from the dataset during blind data review due to a persistent inability to achieve performance criteria across tasks.
The final per-protocol analysis sample therefore comprised 70 participants (F 37/M 33; mean age 26.9 years, range 18–45 years; 5 vegetarians and 1 vegan). All participants self-reported that they were healthy and free from any relevant medical condition or disease, including psychiatric and neurodevelopmental disorders; that they were not taking any prescription or illicit drugs, food supplements or nicotine containing products; that they were not pregnant, lactating or seeking to become pregnant. Participants were also excluded if they consumed >500 mg caffeine per day (>6 × 150 mL cups of filter coffee), had high blood pressure (>systolic 159 mm Hg or diastolic 99 mm Hg) or had a body mass index outside of the range 18.5–35 kg/m2. Participant dispositions are shown in Figure 1.
The final number of participants' data points (excluding missing data and data points removed during blind data review) included in the analysis of data from each individual outcome are shown in the relevant tables.
**Figure 1.** Participant disposition.
### *2.3. Treatments*
Zynamite® mango leaf extract is comprised of components within the following ranges: mangiferin—60–65%; homomangiferin—3–5%; isomangiferin—up to 1%; leaf polysaccharides—6–20%; hydrolysable and non-hydrolysable tannins—up to 1%; fibre, minerals, moisture—6 to 15%. Details of the manufacturing process are provided elsewhere [39].
Participants were randomly allocated to receive 300 mg mango leaf extract or placebo (maltodextrin) in methylcellulose capsules of identical appearance, during each of their two assessment days. Testing days were separated by a minimum of 7 days to ensure washout. The order in which participants received the two interventions was counterbalanced across the group via random allocation to a counterbalancing schedule. Individual treatments were delivered to the trial facility in individual sealed plastic envelopes, labelled with the participants' randomisation numbers and visit (1 or 2) according to the computer-generated double-blind randomisation schedule.
There were no significant adverse events that could be linked to administration of the treatments and no significant difference in the incidence of minor adverse events (e.g., mild headache) between the placebo and mango leaf extract treatments.
### *2.4. Psychological Measures*
### 2.4.1. Cognitive Tasks
All of computerised cognitive/mood assessments were identical, and were carried out via laptop computers and response boxes using the Computerised Mental Performance Assessment System (COMPASS, Northumbria University, Newcastle upon Tyne, UK). This software platform incorporates the presentation of classic and custom computerised cognitive tasks, with fully randomised parallel versions of each task delivered at each assessment for each individual. A similar selection of tasks has previously been shown to be sensitive to diverse nutritional interventions [40–43]. Within the 60-min assessment the participants also completed a 30-min component known as the Cognitive Demand Battery (CDB), which comprises the prolonged repetition of a series of demanding tasks that assess working memory, executive function and attention. The objective of this battery is to assess the impact of treatment on speed/accuracy and mental fatigue during continuous performance of cognitively demanding tasks. The CDB has also been shown to be sensitive to modulation by a wide range of nutritional interventions [43–46]. The individual tasks making up the cognitive assessment (including
the CDB) are shown in Figure 2 and described in more detail in the online Supplementary Materials (Section I).
Figure 2 also shows the contribution of individual tasks to the principal performance measures, which were derived by averaging the data (either msec for speed, or % correct/maximum score for accuracy) from individual tasks into the following global performance outcomes: 'Speed of Performance' and 'Accuracy of Performance'; and the following cognitive domain factor scores 'Speed of Attention', 'Accuracy of Attention', 'Speed of Memory', 'Working Memory', and 'Episodic Memory'. The derivation of the global scores and cognitive factors are described in more detail in the online Supplementary Materials (Section II). These global measures and cognitive domain factors have been shown to be sensitive to nutritional manipulations previously [40–42].
**Figure 2.** Cognitive assessments. The running order of tasks and their contribution to the cognitive factors (to the right) and global performance measures (to the left) derived from the overall battery. The same assessment was completed at the pre-treatment baseline and at 30 min, 3 h and 5 h post-dose on each assessment day. The selection of tasks took a total of 60 min to complete, with the Cognitive Demand Battery comprising 30 min of this. The individual tasks are described in more detail in the supplementary online materials (Section I). Rapid Visual Information Processing task (RVIP). Visual analogue scale (VAS).
### 2.4.2. Mood and Psychological State
Before each cognitive assessment, participants completed the Profile of Mood States (POMS-2) Adult Short Form [47]. As part of the COMPASS battery, and before the cognitive tasks, participants completed the Visual Analogue Mood Scales (VAMS), a set of 18 visual analogue scales anchored by pairs of antonymic mood/state adjectives (e.g., Alert–Inattentive; Lethargic–Energetic). Participants rated where they would position themselves between the adjectives anchoring each line according to how they felt at that moment. The individual item scores were combined to give an average (% along the line) score on three factors that had previously been derived by factor analysis:
'Alertness', 'Tranquillity' and 'Stress'. After the cognitive tasks participants also completed a further four stress visual analogue scales (S-VAS) that required them to rate their current psychological state between 'not at all' and 'extremely' with regard to their levels of stress, anxiety, calmness and relaxation. These were combined into two scores 'stress/anxiety' and 'calm/relaxed' with a higher score (average % along the line) representing more of the descriptor.
### *2.5. Procedure*
Participants were required to attend the Brain, Performance and Nutrition Research Centre (Northumbria University) for three visits. The first visit comprised a screening and training session where, once written informed consent had been obtained, participants were screened according to the inclusion/exclusion criteria. Eligible participants then provided lifestyle and demographic data and their height, weight, waist to hip ratio and blood pressure were measured. They completed a short training session in which they practiced the cognitive tasks. Practice took the form of three repetitions of shortened versions of the COMPASS cognitive tasks, followed by the completion of the full-length, 60-min battery twice. During and at the end of the practice session, participants' performance was checked against standard minimum performance criteria and additional guidance was provided as necessary. At the end of this visit, participants were briefed as to what to expect on testing visits and were provided with pre-testing instructions.
Within four weeks of the screening visit, participants returned to the laboratory for their first testing visit at an agreed time in the morning that remained consistent across all testing visits. A maximum of 5 participants were tested on any day, and all participants were visually isolated in individual testing booths. Participants arrived at the laboratory having refrained from alcohol for 24 h, caffeine overnight and having consumed a simple breakfast of cereal and/or toast at home no later than an hour before arrival. Once participants arrived at the lab, they were not permitted to eat any food (aside from food items provided by the study staff) or drink (except for water) or chew gum. Continued compliance with the inclusion/exclusion criteria was assessed. This was followed by completion of the POMS and a 60-min computerised cognitive and mood assessment (COMPASS—including the 30-min Cognitive Demand Battery (CDB), Visual analogue mood scales (VAMS) and stress visual analogue scales (S-VAS)—see Figure 2.). Cognitive tasks were completed with the participants visually isolated from each other. After the first cognitive/mood assessment, participants consumed their treatment for the day and completed cognitive/mood assessments, identical to the above, commencing at 30 min, 3 h and 5 h post-dose. An additional, brief, 5-min assessment investigating the participants' response to a laboratory stressor, plus pre/post-dose blood sampling for half of the participants (for quantification of neurotrophins and catecholamines), took place after the pre-treatment and 30-min post-dose cognitive/mood assessments (For methodology see [48]), the results of this theoretically distinct investigation are to be reported elsewhere). All participants were scheduled to return to the laboratory 7 days later, with a maximum allowable leeway of an additional 7 days should exceptional circumstances arise in the meantime. This second testing day was identical to the previous day, with the exception that participants consumed a different treatment on each of the two days. The timelines of the testing day are presented in Figure 3.
**Figure 3.** The timelines of the testing day for individual participants, showing the core cognitive assessment schedule. Profile of Mood States (POMS), 5 min Observed Multi-Tasking Stressor (\*OMS) (methodology and results to be reported elsewhere).
Participants were provided with a standardised lunch (comprising a cheese sandwich on white bread, crisps and a custard pot) between the 180 and 300 min post-dose assessments and were given the option of a snack (hot decaffeinated tea or coffee and digestive biscuits) after completion of the stressor following the 30-min post-dose assessment. No alternative drinks, snacks or lunches were permitted.
### *2.6. Analysis*
The study statistical analysis plan was formulated before the completion of data collection. Given the exploratory nature of the study, and the lack of any relevant human data, a small sub-set of primary outcomes was not pre-defined. Given the study intervention and objectives, a per protocol analysis was deemed the most appropriate.
All outcomes were analysed using SPSS (version 24.0, IBM corp., Armonk, NY, USA). During blind data review a number of participants' individual task datasets were removed due to technical or performance issues (for details of the issues and number of datasets involved see supplementary online materials). Prior to the primary analysis of the effects of treatment, pre-dose baseline differences between treatment were investigated by one-way (treatment group [placebo v mango leaf extract]) paired t tests, or in the case of the Cognitive Demand Battery (CDB) two-way (treatment x repetition) ANOVA. There were no significant differences between treatment groups at baseline.
For all cognitive and mood measures, the primary analysis of post-dose data was by Linear Mixed Models (LMM) using the MIXED procedure in SPSS (version 22.0, IBM corp.) with pre-dose baseline data for each outcome included as a covariate. For all LMM analyses, the 'compound symmetry' covariance structure provided the best fit, with the exception of 'mental fatigue' from the CDB for which an autoregressive covariance structure (AR1) was more appropriate.
For the cognitive outcomes derived from the COMPASS battery and the mood outcomes, terms were fitted for treatment (placebo/mango leaf extract) and assessment (30 min, 3 h, 5 h) and their interaction. For the CDB measures an additional 'repetition' term was added along with the appropriate interactions. Given that the treatment orders were balanced across the sample, or exactly or nearly balanced with regard to the participants contributing to each outcome (and given that treatment carry-over effects were highly unlikely), treatment order was not included as a factor in the analysis.
In order to establish the time course of any effects, pre-defined planned comparisons were conducted between treatments at each assessment time point (30 min, 3 h, 5 h) with a Bonferroni adjustment for the number of comparisons undertaken per outcome (i.e., 3). Only those planned comparisons conducted on data from outcomes that evinced a significant treatment related main or interaction effect are reported below.
### **3. Results**
### *3.1. Cognitive Task Global and Factor Outcomes*
The global outcomes and cognitive factors derived from the COMPASS battery showed that mango leaf extract resulted in significantly improved accuracy of performance across tasks and throughout the testing day (i.e., at 30 min, 3 h and 5 h post-dose). See Figure 4 below. There was a main effect of treatment on the global Accuracy of Performance measure (representing data from the eleven tasks that return % accuracy/maximum score data) (F (1, 335) = 22.8, *p* < 0.001). Reference to the planned comparisons at each assessment showed that this effect was evident throughout the post-dose testing period (30 min *p* = 0.03, 3 h *p* = 0.02, 5 h *p* = 0.009). There were also significant main effects in terms of improved accuracy following mango leaf extract on the Accuracy of Attention factor (F (1, 315) = 16.697, *p* < 0.001) and the Episodic Memory factor (F (1, 345) = 6.94, *p* = 0.009). With regard to the time course of these effects, whilst the Bonferroni adjusted comparisons of Episodic Memory scores did not reach significance during the individual assessments, Accuracy of Attention was improved at both the 3h(*p* = 0.048) and 5 h (*p* = 0.01) post-dose assessments. Data (plus F score and *p*) for the cognitive
outcomes derived from the COMPASS battery are presented in the online Supplementary Materials (Table S1.).
**Global Performance Measures**
**Figure 4.** The effects of mango leaf extract on the global outcome measures and factor scores derived from the Computerised Mental Performance Assessment System (COMPASS) cognitive tasks. Left-hand panels show the main effect of treatment averaged across assessments; middle panels show the pre-dose baseline scores; right-hand panels show time course data from each post-dose assessment for those measures that saw significant effects on the planned comparisons (Bonferroni). The global Accuracy of Performance measure represents averaged data from the eleven tasks from the battery that return % accuracy/maximum score data: Accuracy of Attention represents averaged % accuracy data from the five attention tasks; and Episodic Memory represents averaged % accuracy/recall across the four long-term memory tasks. \*, *p* < 0.05; \*\*, *p* < 0.01, \*\*\*, *p* < 0.001 versus placebo. Number of participants contributing to the measure: Accuracy of Performance, *n* = 68, Episodic Memory, *n* = 70, Accuracy of Attention, *n* = 64.
### *3.2. Cognitive Demand Battery (CDB)*
In keeping with the improved accuracy seen across the COMPASS task factors, performance in all three CDB tasks was improved across the testing day following mango leaf extract. See Figure 5. The Rapid Visual Information Processing task (RVIP) was improved across assessments in terms of
% of targets accurately detected (F (1, 1071) = 23.186, *p* < 0.001) with planned comparisons showing that these effects were apparent at the 30 min (*p* = 0.047) and 5 h (*p* = 0.001) assessments, with a trend towards the same effect at 3 h post-dose (*p* = 0.059). Performance was also improved on both the Serial 3s task (F (1, 1156) = 10.9, *p* < 0.001) and Serial 7s task (F (1, 1156) = 9.642, *p* = 0.002) in terms of number of correct subtractions across the testing day. Comparisons at each assessment showed that while the differences between groups did not reach significance during any individual assessment for the Serial 7s task, Serial 3s performance was enhanced at the 3 h assessment (*p* = 0.014), with a trend towards the same at 30 min post-dose (*p* = 0.088). There was no effect on ratings of mental fatigue during completion of the battery. Data (plus F score and *p*) for the CDB outcomes are presented in the online Supplementary Materials (Table S2.).
**Figure 5.** The effects of mango leaf extract on the Cognitive Demand Battery outcomes. Each task was repeated three times per assessment (total Cognitive Demand Battery (CDB) completion time, 30 min per assessment). Left-hand panels show the main effect of treatment averaged across assessments/repetitions; middle panels show the pre-dose baseline scores averaged across the three repetitions; right-hand panels show time course data from each post-dose assessment (averaged across the three repetitions per assessment) for those measures that saw significant effects on the planned comparisons (Bonferroni) of mango leaf extract versus placebo. t, *p* < 0.1; \*, *p* < 0.05; \*\*\*; *p* < 0.001 in comparison to placebo. Number of participants contributing to the measure: RVIP, *n* = 64, Serial 3s/7s, *n* = 69.
### *3.3. Mood and Psychological State*
There were no effects of treatment on any mood parameter (VAMS, S-VAS, POMS), with the exception of reduced calm/relaxed ratings on the S-VAS following mango leaf extract across testing assessments (F (1, 345) = 5.44, *p* = 0.02). See Figure 6. There were no significant differences on the comparisons made at each assessment for this outcome. Data from the POMS, VAMS and S-VAS data are presented in the online Supplementary Materials (Table S3.).
**Figure 6.** The effects of mango leaf extract on the calm/relaxed stress visual analogue scales (S-VAS) measure. There were no significant differences on the planned comparisons of data from each assessment. \*, *p* < 0.05 in comparison to placebo.
### **4. Discussion**
In the current study a single dose of mango leaf extract (Zynamite®) lead to significant, broad improvements in performance across a battery of cognitive tasks throughout the 6 h following consumption. There were no interpretable benefits found for any measure of mood/psychological state.
Cognitive improvements were seen on the global Accuracy of Performance measure, which comprised averaged % accuracy or % maximum score data from 11 computerised tasks. It was also seen more specifically in the cognitive sub-factors 'Accuracy of Attention', representing the overall % accuracy whilst performing the five attention tasks (excludes simple reaction time) within the battery and 'Episodic Memory', which represents the % recall or accuracy of the four long-term memory tasks. Performance benefits were also seen across all three of the tasks that make up the 30-min Cognitive Demand Battery, with improved RVIP accuracy and increased numbers of correct subtractions generated by participants on both the Serial 3s and Serial 7s tasks. These cognitive effects, taken as a whole, were evident as main effects across the post-dose testing day, which comprised 60 min assessments starting at 30 min, 3 h and 5 h post-dose, without any clear pattern of augmentation or attenuation over time. There were no benefits seen in terms of increased speed of task performance on the timed tasks, or indeed on the mood and psychological state measures.
Clearly, one question raised by these results is whether the effects seen here represent a truly global improvement in accuracy across cognitive domains, or whether they simply reflect the consequences of improved attention. Certainly, attention and episodic memory are inter-related, with enhanced attention leading to improved encoding and retrieval of information. It has been suggested that episodic memory processes are themselves, to an extent, 'acts of attention' [49]. As the attention and episodic memory tasks comprised the majority of the tasks that contributed to the global accuracy measure, it is possible that the improvements to the latter are simply a reflection of broad improvements to attention. However, the improvements in Serial 3s and Serial 7s subtraction task performance would be more difficult to accommodate solely within an attention framework. Whilst both subtraction tasks have attentional components, they draw more heavily on both working memory and executive function, particularly the more difficult Serial 7s, which requires greater executive resources in order to carry out the more complex manipulation of numbers [24]. Enhanced performance on these tasks, alongside improved accuracy across the tasks, therefore, seems to confirm that the benefits of mango leaf extract were seen broadly across cognitive domains.
The results also suggest that the modulation of cerebro-electrical activity (measured using EEG) seen in healthy adults following a single dose of Zynamite mango leaf extract [39] is most likely indicative of a benefit to brain function. The cognitive benefits seen here are broadly in line with previous demonstrations of improved cognitive function following both acute [24–26] and chronic administration [27,28] of polyphenol rich extracts. Several polyphenol studies also employed the Cognitive Demand Battery used here (but at a single post-dose time point), with demonstrations of improved performance across all three tasks following cocoa-flavanols [24], improved Serial 3s performance following fruit flavanols [50], but no benefits following resveratrol [51]. Of note, the global performance measures derived from the cognitive tasks utilised here have proved sensitive to the acute and chronic administration of a Nepalese pepper extract [42] and acute administration of a green oat extract to middle-aged adults [41]. However, both of these interventions contain other potentially bioactive phytochemicals alongside polyphenols, and in both cases global speed of performance was enhanced, rather than the improved global accuracy seen in the current study.
Previous research has demonstrated similarities in EEG cerebro-electrical response following both mango leaf extract and caffeine in rodents [38], but somewhat different responses to these two individual treatments in humans [39]. The cognitive effects of caffeine comprise modest but consistent improvements that are restricted to the performance of tasks measuring attention, with no reliable effect on other cognitive domains including long-term (episodic) memory [52–55]. Similarly, the duration of the effects seen following mango leaf extract do not follow the time course of caffeine's effects, which would become apparent by 30 min post-dose and would be expected to attenuate by 6 h post-dose. It is therefore notable that the pattern of cognitive benefits seen in the present study following the mango leaf extract are broader and longer lasting than those that would be expected after caffeine.
In terms of mechanism of action, a recent study investigating receptor binding and brain relevant enzyme inhibition found that mangiferin only significantly inhibited catechol-O-methyl transferase (COMT), the enzyme responsible for the degradation of catecholamine neurotransmitters [39]. Several other polyphenols that also feature a catechol moiety, including flavanols and oleacein, have also been shown to inhibit COMT [56,57]. COMT's catabolic pathway is most prevalent in brain tissue with low concentrations of catecholamine reuptake transporters, and therefore COMT inhibition predominantly affects dopaminergic function in the prefrontal cortex and hippocampus [58], potentially leading to improved working memory, selective attention, and executive function [59]. Clearly, the benefits seen in the current study correspond with these cognitive domains. However, whilst there is some evidence that COMT inhibitors may modulate these aspects of cognitive function, the overall pattern is for their effects to be bidirectionally moderated by COMT genotype (val158met polymorphism) [59–61]. COMT inhibition per se is therefore unlikely to be the primary mechanism underpinning the straightforward cognitive benefits seen here across a sample of mixed COMT genotypes.
Other potential 'direct' brain-relevant mechanisms of action previously established for mangiferin include acetylcholinesterase (AChE) inhibition [30,62] or other potential cholinergic mechanisms of action [63]. Increased acetylcholine activity would be expected to have a beneficial, inter-related effect on both focussed attention and memory consolidation/retrieval [64] and, therefore, could encompass many of the effects seen in the current study. However, it is equally likely that the effects seen here may be related to 'indirect' interactions within mammalian cellular signal transduction pathways, a property that mangiferin shares with other polyphenols [8,11–14]. These interactions potentially drive downstream modulation of neuroinflammation, neurotransmission, neurotrophin receptor and signalling pathway interactions, and increased synthesis of neurotrophins and vasodilatory molecules, leading to increased angiogenesis/neurogenesis and local cerebral blood flow [15–19]. These indirect cellular interactions may underlie the consistent demonstrations in humans of increased cerebral
blood-flow [51,65–69] and peripherally measured brain-derived neurotrophic factor [26] seen following diverse polyphenols. Again, potentially diffuse beneficial effects within the brain could be conceived as potentially leading to broad benefits to cognitive function across domains, as seen here.
Clearly, a strength of the current study is that it represents the first concerted investigation of the effects of mangiferin, or indeed any xanthone glycoside, on human cognitive function. Conversely, this was, by its nature, an exploratory study, and the absence of pre-defined primary outcomes, due to a lack of previous data to guide their formulation, could be considered a limitation. Certainly the absence of primary endpoints allows a greater freedom for the interpretation of the results than will be enjoyed in future research, and it is hoped that the results of the current study will be useful in terms of directing the research questions and outcomes addressed by more studies involving this compound.
It should also be acknowledged that the results herein relate to a molecule, or group of molecules (xanthones) that are unlikely to be encountered in meaningful quantities in the typical diet, and therefore the results can only realistically be extrapolated to supplementation with mangiferin-rich extracts. Whilst the results tell us little about the benefits of polyphenols consumed as part of the everyday diet it might be noteworthy that the dose of 300 mg employed here contained an amount of polyphenols that is achievable through the consumption of polyphenol rich foods.
In conclusion, a single dose of mango leaf extract (Zynamite®) with high levels of the polyphenol mangiferin, lead to broad improvements in cognitive function that were seen across assessments spanning from 30 min to 6 h post-dose. These benefits were seen most strikingly in terms of participants' improved attention and long-term memory task performance and in their extended performance of cognitively demanding tasks, including those requiring executive function resources.
**Supplementary Materials:** The following are available online at http://www.mdpi.com/2072-6643/12/8/2194/s1. Section I: Individual COMPASS and CDB cognitive task descriptions. Section II: Derivation of the global outcome measures and cognitive domain factors, including notes on lost data. Section III: Table S1. (data from global measures and cognitive factors), Table S2. (data from Cognitive Demand Battery), Table S3. (data from the mood measures).
**Author Contributions:** D.O.K., E.L.W., P.A.J., J.C.W., N.G. formulated the research question and overall methodology. All of the authors (D.O.K., E.L.W., P.A.J., J.F., J.K., J.C.W., N.G.) were actively involved in the practical planning of the research described herein. J.F and J.K. supervised the collection of the data. D.O.K. analysed the data and E.L.W. and D.O.K. compiled the first draft of the paper. All authors contributed to and reviewed the subsequent drafts and final publication. All authors have read and agreed to the published version of the manuscript.
**Funding:** The study was sponsored by Nektium Pharma.
**Acknowledgments:** The following people were involved in the day to day running of the study and/or data collection: Amy Ferguson, Jennifer Webster, Fiona Dodd, Michael Patan, Ellen Smith, Rian Elcoate, Jessica Greener, Lucy Keeler, Charlotte Kenney, Faye Williams, Evan Davies, Leah Smith, Veronika Rysinova.
**Conflicts of Interest:** Nektium Pharma produced the Mangifera indica extract and sponsored the study. J.C.W. is employed by Nektium Pharma, and N.G. works as a consultant for Nektium Pharma. However, neither J.C.W., N.G. nor any other representative of Nektium Pharma had any role in the running of the study, or the analysis, or the interpretation of the data. None of the other authors have a conflict of interest.
### **References**
© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
| doab | 2025-04-07T03:56:58.605736 | 1-5-2021 17:49 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "003cf560-3166-422f-9a1c-98b1e971efa6",
"url": "https://mdpi.com/books/pdfview/book/3241",
"author": "",
"title": "Dietary Polyphenols and Human Health",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783039437931",
"section_idx": 8
} |
003cf560-3166-422f-9a1c-98b1e971efa6.9 | *Article* **Associations of Urinary Phytoestrogen Concentrations with Sleep Disorders and Sleep Duration among Adults**
### **Jing Sun 1, Hong Jiang 2,\*, Weijing Wang 1, Xue Dong <sup>1</sup> and Dongfeng Zhang <sup>1</sup>**
Received: 29 June 2020; Accepted: 13 July 2020; Published: 16 July 2020
**Abstract:** Current evidence on the relationship of phytoestrogens with sleep is limited and contradictory. In particular, studies on individual phytoestrogens and sleep have not been reported. Thus, this study aimed to appraise the associations of individual phytoestrogens with sleep disorders and sleep duration. This cross-sectional study comprising 4830 adults utilized data from the National Health and Nutrition Examination Survey 2005–2010. Phytoestrogens were tested in urine specimens. Sleep disorders and sleep duration were based on a self-reported doctor's diagnosis and usual sleep duration. The main analyses utilized logistic and multinomial logistic regression models and a restricted cubic spline. In the fully adjusted model, compared with tertile 1 (lowest), the odds ratios (95% confidence intervals (CIs)) of sleep disorders for the highest tertile of urinary concentrations of enterolactone, enterodiol, and O-desmethylangolensin were 0.64 (0.41–1.00), 1.54 (1.07–2.21), and 1.89 (1.26–2.85), respectively. Linear inverse, approximatively linear positive, and inverted L-shaped concentration–response relationships were found between enterolactone, enterodiol, and O-desmethylangolensin and sleep disorders, respectively. Compared with normal sleep (7–8 h/night), the relative risk ratio (RRR) (95% CI) of very short sleep for enterolactone was 0.56 (0.36–0.86), and the RRR (95% CI) of long sleep risk for genistein was 0.62 (0.39–0.99). Furthermore, negative associations of genistein with sleep disorders and enterolactone with long sleep risk, as well as positive associations of enterodiol with both long and very short sleep, were observed in the stratified analysis by age or gender. Finally, a notable finding was that urinary O-desmethylangolensin concentration was positively related to sleep disorders in both females aged 40–59 years and non-Hispanic Whites but inversely associated with sleep disorders in both females aged 60 years or over and other Hispanics. Our findings suggested that enterolactone and genistein might be beneficial for preventing sleep disorders or non-normal sleep duration among adults, and enterodiol might be adverse toward this goal. However, the association of O-desmethylangolensin with sleep disorders might be discrepant in different races and females of different ages.
**Keywords:** sleep disorders; sleep duration; urinary phytoestrogens; concentration–response; NHANES
### **1. Introduction**
Sleep disorders, classified into insomnia, central disorders of hypersomnolence, sleep-related breathing disorders, parasomnias, sleep-related movement disorders, circadian rhythm sleep–wake disorders, and other sleep disorders [1], are common conditions and can seriously harm human health and quality of life [2]. Studies have found that poor sleep (short or long sleep duration, or other sleep problems) was associated with obesity, cardiovascular disease, diabetes, hypertension, cancer, and higher mortality [3–8]. Exploration of modifiable factors for reducing the risk of sleep disorders is urgently required. Estrogen has been a focus of attention due to its influences on the central nervous system that is involved in regulating sleep [9]. Randomized trials have revealed that estrogen therapy could reduce sleep disturbances and improve sleep quality [10–12]. Despite the significant beneficial effect on sleep, the use of estrogen therapy is still very cautious in virtue of its potentially serious health risks [13–17]. Therefore, as the naturally occurring mimetic agents of estrogen, phytoestrogens have aroused great interest in their possibility of being estrogen substitutes.
Phytoestrogens are a group of plant-derived bioactive compounds that have estrogenic and anti-estrogenic effects due to their structures resembling estradiol [18,19] and they are also considered to be endocrine disruptors [20]. The two principal groups of phytoestrogens comprise lignans and isoflavones. Lignans mainly originate from oilseeds, dried seaweeds, and whole-grain cereals, and are metabolized into enterolactone and enterodiol by bacteria in the colon [21,22]. Isoflavones, which are primarily derived from soya beans, soy products, and legumes, consist of genistin and daidzin, which can be hydrolyzed into genistein and daidzein, respectively, where daidzein is further metabolized into O-desmethylangolensin (O-DMA) or equol by gut microbiota [22]. Different individuals have different capabilities to produce phytoestrogens via microbial synthesis due to the complex interaction of the colonic environment with internal and external factors [23,24]; for example, only approximately 30–50% of individuals can produce equol via gut bacterial metabolism [25]. Additionally, these metabolites can also be directly obtained from some animal products, such as dairy [26,27].
Thus far, several trial studies have explored the relationship between total isoflavone supplementation and sleep in climacteric women or androgen-deprived prostate cancer patients, with mixed findings, including improvement of sleep problems [28–31], no significant association [32–34], and even insomnia aggravation [35]. Only two observational studies have evaluated the association of total isoflavone consumption with sleep status among general adults and results showed that total isoflavone consumption was positively related to sleep quality and optimal sleep duration in Japanese adults and inversely associated with long sleep duration in Chinese adults [36,37]. However, no epidemiological study to date has appraised the relationship between lignans and sleep. Only several animal experiments have found the sedative and hypnotic effects of the lignan component [38,39]. Moreover, individual phytoestrogens have unequal biological activities and estrogen receptor (ER) affinities [40,41]. Studies have reported divergent associations between individual phytoestrogens and disease [42–46]. Thus, individual phytoestrogens may also have diverse impacts on sleep, and investigating the relationships with sleep among individual phytoestrogens may be more meaningful. Meanwhile, investigating phytoestrogens via dietary evaluation made it difficult to include all food origins and did not take into account the metabolic transformation of intestinal flora, causing inexact individual exposure; therefore, it is necessary to assess phytoestrogens based on biomarkers to reflect the true exposure. Additionally, the concentration–response relationships of phytoestrogens with sleep were also unknown. Therefore, the present study aimed to: first, appraise the associations between the urinary concentrations of individual phytoestrogens and sleep disorder risk among U.S. adults by utilizing data from the National Health and Nutrition Examination Survey (NHANES) 2005–2010; second, to explore the concentration–response relationships of them; third, to explore the gender, age, and race differences in the associations; and finally, to evaluate the associations of individual phytoestrogens with sleep duration.
### **2. Materials and Methods**
### *2.1. Study Population*
The NHANES is a cross-sectional, complex, multistage, and stratified probability sampling design representing the non-institutionalized U.S. civilian population, which aims to investigate the health status and nutrition condition of Americans [47,48]. The NHANES collects data via examinations implemented in the mobile examination center (MEC) and via household interviews. All of the participants provided informed consent and the protocol of investigation was authorized by the Research Ethics Review Board of the National Center for Health Statistics.
This study chose three-cycle data (NHANES 2005–2006, 2007–2008, and 2009–2010) to create the current sample because the data of sleep and urinary phytoestrogens were measured simultaneously only in the three cycles. A total of 31,034 individuals were enrolled in the NHANES 2005–2010, where the number of participants aged 18 years and over was 18,318. A subsample of approximately one-third of all NHANES participants aged six years or over was chosen to measure urinary phytoestrogens, leaving 5496 participants. Among them, 666 individuals were further ruled out, including participants with missing sleep data (*n* = 24), lactating or pregnant females (*n* = 199), females with both ovaries removed (*n* = 269), and individuals using sedative-hypnotic drugs (*n* = 174). Ultimately, 4830 participants (age ≥ 18 years) with phytoestrogen data were analyzed in the current study (Figure 1).
**Figure 1.** Flowchart of the screening process of eligible participants from the National Health and Nutrition Examination Survey 2005–2010.
### *2.2. Urinary Phytoestrogens Measurement*
Spot urine specimens were collected in the MEC and urinary concentrations of individual phytoestrogens (enterolactone, enterodiol, daidzein, O-DMA, equol, and genistein) were tested in the one-third subsample of all NHANES participants aged six years or over by utilizing high-performance liquid chromatography–atmospheric pressure photoionization–tandem mass spectrometry during the NHANES 2005–2010. More details are available in the laboratory procedure manual [49]. Studies have revealed that phytoestrogen concentrations in spot urine are reliable biomarkers for phytoestrogen intake [23,50,51].
To correct for urine dilution, phytoestrogen concentrations were creatinine-standardized and expressed as μg/g creatinine [52]. Urine creatinine was measured by utilizing the Beckman CX3 during 2005–2006 but the Roche ModP was used since 2007; therefore, we adjusted creatinine 2005–2006 for the comparability of creatinine between 2005–2006 and 2007–2010 using the equations recommended by the NHANES [53].
### *2.3. Sleep Disorders and Sleep Duration Assessments*
The sleep disorder investigations were administered by utilizing a computer-assisted personal interviewing system by trained interviewers in the home. Participants were classified into the sleep disorder groups based on a self-reported doctor diagnosis, and this classification method was used by prior published studies [54,55]. Sleep duration was categorized as 7–8 h/night (normal sleep), <5 h/night (very short sleep), 5–6 h/night (short sleep), and ≥9 h/night (long sleep) based on the self-reported usual sleep duration at night [54,56].
### *2.4. Covariates*
The covariates gender, age, marital status, race, occupation, family income, educational level, body mass index (BMI), smoking status, alcohol consumption, use of female hormones, physical activity, caffeine intake, C-reactive protein, hypertension, depressive symptoms, and diabetes were chosen based on previous literature to control for potential confounding effects [12,36,57]. The details of the classification and criteria for covariates are displayed in Table S1.
### *2.5. Statistical Analysis*
To explicate the complexity of the sampling design and create the estimated values of national representativeness, a primary sampling unit, strata information, and specific sampling weights for the one-third subsample were utilized in the present analyses. According to the analytical guideline [58], the new six-year weights were generated by dividing the two-year environmental weights by three and were applied in this study due to the combination of three two-year NHANES cycles.
The distribution types of continuous variables were identified using Kolmogorov–Smirnov normality tests. Numbers (percentages) and medians (interquartile ranges) were used to describe the qualitative data and non-normal quantitative data, respectively. Chi-square tests and Mann–Whitney *U* tests were performed to compare percentages for qualitative data and averages for non-normal quantitative data between the non-sleep disorders group and the sleep disorders group, respectively. Urinary concentrations of individual phytoestrogens were segmented into tertiles based on their distributions in the present population, with tertile 1 (lowest) being the referent. First, logistic regression analyses were conducted to appraise the relationships between urinary phytoestrogens and sleep disorders, along with calculating the odds ratios (ORs) and 95% confidence intervals (CIs). Six phytoestrogens were entered into model 1 simultaneously for controlling for potentially confounding effects. Model 2 additionally adjusted for gender and age, and model 3 was further adjusted for marital status, race, occupation, family income, educational level, BMI, smoking status, alcohol consumption, use of female hormones, physical activity, caffeine intake, C-reactive protein, hypertension, depressive symptoms, and diabetes. The concentration–response relationships of urinary phytoestrogens with sleep disorders were evaluated using restricted cubic spline functions of three knots (the 25th, 50th, and 75th percentiles of the exposure distributions) in model 3. Second, multinomial logistic regressions were utilized to evaluate the associations between urinary phytoestrogens and sleep duration in model 3, with normal sleep (7–8 h/night) being the referent. Third, considering
the significant differences in sleep between different age groups and genders [59,60], we conducted stratified analyses by age (18–39, 40–59, and ≥60 years) and gender, as well as by age group separately for females and males, respectively. Finally, given that the intestinal metabolism of phytoestrogens may differ depending on the race [61], a stratified analysis by race (Mexican American, non-Hispanic Whites, non-Hispanic Blacks, and other Hispanics) was also performed. Statistical analyses were performed utilizing Stata 15.0 (Stata Corporation, College Station, TX, USA), and *p* < 0.05 (two-sided) suggested statistical significance.
### **3. Results**
The characteristics of the participants in this study across sleep disorders are displayed in Table 1. In total, 4830 eligible individuals were analyzed, and the prevalence of sleep disorders was 6.25%. Except for family income and alcohol consumption, other characteristics were significantly different between the sleep disorders group and the non-sleep-disorders group. Individuals with the following characteristics were more likely to experience sleep disorders: older, male, non-Hispanic White, married or living with a partner, smoker, obese, hypertension, depressive symptoms, diabetes, higher education level, less physical activity, higher C-reactive protein concentration, higher caffeine intake, without work, and women using female hormones. Except for family income, alcohol consumption, and smoking status, other characteristics were significantly different between the male sleep disorders group and the male non-sleep-disorders group. However, there were significant differences between the female sleep disorders group and the female non-sleep-disorders group only in terms of age, BMI, use of female hormones, hypertension, depressive symptoms, and diabetes.
Table 2 presents the weighted ORs with 95% CIs for sleep disorders according to the tertiles of urinary phytoestrogen concentrations. In model 1, the urinary concentrations of enterolactone and genistein were inversely associated with sleep disorders, while the urinary O-DMA concentration was positively related to sleep disorders. After an additional adjustment for gender and age in model 2, the results were concordant with model 1. In the fully adjusted model (model 3), the negative association between the genistein concentration and sleep disorders was no longer significant, and the enterolactone concentration was still inversely associated with sleep disorders, while the urinary concentrations of enterodiol and O-DMA were positively related to sleep disorders. Compared with tertile 1 (lowest), the fully adjusted ORs (95% CIs) for sleep disorders for the highest tertile of urinary concentrations of enterolactone, enterodiol, and O-DMA were 0.64 (0.41–1.00), 1.54 (1.07–2.21), and 1.89 (1.26–2.85), respectively.
The concentration–response relationships of the urinary concentrations of enterolactone, enterodiol, and O-DMA with sleep disorders are depicted in Figures 2–4, respectively. The urinary enterolactone concentration was linearly negatively associated with sleep disorders (*p*-nonlinearity = 0.849). The association began to show statistical significance when the enterolactone concentration reached around 904 μg/g creatinine (OR: 0.66, 95% CI: 0.43–0.99) (Figure 2). However, an approximately linear positive relationship was observed between the urinary enterodiol concentration and sleep disorders (*p*-nonlinearity = 0.274), and when the enterodiol concentration reached around 86 μg/g creatinine (OR: 1.68, 95% CI: 1.00–2.83), the relationship began to present statistical significance (Figure 3). There was a nonlinear positive (inverted L-shaped) association between urinary O-DMA concentration and sleep disorders (*p*-nonlinearity = 0.033). The OR of sleep disorders increased with increasing urinary O-DMA concentrations, and it arrived at a plateau when the O-DMA concentration was above approximately 13 μg/g creatinine (OR: 1.90, 95% CI: 1.09–3.31) (Figure 4).
**Table 2.** Weighted odds ratios (95% confidence intervals) for sleep disorders across tertiles of urinary phytoestrogens concentrations (National Health and Nutrition Examination Survey 2005–2010).
<sup>a</sup> Six phytoestrogens (tertiles) were entered into model 1 simultaneously. <sup>b</sup> Model 2 additionally adjusted for age and gender. <sup>c</sup> Model 3 further adjusted for race, education, marital status, occupation, family income, body mass index, physical activity, alcohol use, smoking status, depressive symptoms, diabetes, hypertension, use of female hormones, C-reactive protein (mg/dL), and caffeine intake (mg/day). \* *p* < 0.05, \*\* *p* < 0.01.
**Figure 2.** Concentration–response relationship of the urinary enterolactone concentration with sleep disorders. The solid line represents the estimated odds ratios (ORs) and the dashed lines represent their 95% confidence intervals. The relationship was adjusted for gender, age, marital status, race, occupation, family income, educational level, body mass index, smoking status, alcohol consumption, use of female hormones, physical activity, caffeine intake, C-reactive protein, hypertension, depressive symptoms, diabetes, and the other five phytoestrogens (tertiles).
**Figure 3.** Concentration–response relationship of the urinary enterodiol concentration with sleep disorders. The solid line represents the estimated odds ratios (ORs) and the dashed lines represent their 95% confidence intervals. The relationship was adjusted for gender, age, marital status, race, occupation, family income, educational level, body mass index, smoking status, alcohol consumption, use of female hormones, physical activity, caffeine intake, C-reactive protein, hypertension, depressive symptoms, diabetes, and the other five phytoestrogens (tertiles).
**Figure 4.** Concentration–response relationship of the urinary O-desmethylangolensin concentration with sleep disorders. The solid line represents the estimated odds ratios (ORs) and the dashed lines represent their 95% confidence intervals. The relationship was adjusted for gender, age, marital status, race, occupation, family income, educational level, body mass index, smoking status, alcohol consumption, use of female hormones, physical activity, caffeine intake, C-reactive protein, hypertension, depressive symptoms, diabetes, and the other five phytoestrogens (tertiles).
The associations of urinary phytoestrogen concentrations with sleep disorders stratified by age and gender are shown in Tables 3 and 4, respectively. In the fully adjusted model, the urinary enterolactone concentration was still inversely associated with sleep disorders in females (OR: 0.45, 95% CI: 0.21–0.94). Meanwhile, the urinary enterodiol concentration was still positively related to sleep disorders among middle-aged adults (40–59 years) (OR: 2.56, 95% CI: 1.12–5.84) and females (OR: 3.79, 95% CI: 1.83–7.87), and the O-DMA concentration was still positively associated with sleep disorders in young adults (18–39 years) (OR: 4.16, 95% CI: 1.58–10.97). Additionally, there was a negative association between the genistein concentration and sleep disorders in middle-aged adults (40–59 years) (OR: 0.34, 95% CI: 0.13–0.88). There were no significant associations between the urinary phytoestrogen concentrations and sleep disorders in males and older adults (≥60 years). To further describe the gender difference in the associations of the urinary concentrations of enterolactone and enterodiol with sleep disorders, the concentration–response relationships of them for males and females are presented in Figure 5a,b and Figure 6a,b, respectively. The urinary enterolactone concentration was linearly negatively associated with sleep disorders in females (*p*-nonlinearity = 0.283) (Figure 5a) and the enterodiol concentration was positively related to sleep disorders for females in a nonlinear manner (*p*-nonlinearity < 0.001) (Figure 6a), whereas the associations were not significant in males (Figures 5b and 6b).
**Table 3.** Weighted odds ratios (95% confidence intervals) for sleep disorders across tertiles of urinary phytoestrogens concentrations stratified by age (National Health and Nutrition Examination Survey 2005–2010).
**Table 3.** *Cont.*
<sup>a</sup> Six phytoestrogens (tertiles) were entered into model 1 simultaneously. <sup>b</sup> Model 2 additionally adjusted for gender. <sup>c</sup> Model 3 further adjusted for race, education, marital status, occupation, family income, body mass index, physical activity, alcohol use, smoking status, depressive symptoms, diabetes, hypertension, use of female hormones, C-reactive protein (mg/dL), and caffeine intake (mg/day). \* *p* < 0.05, \*\* *p* < 0.01.
**Table 4.** Weighted odds ratios (95% confidence intervals) for sleep disorders across tertiles of urinary phytoestrogen concentrations stratified by gender (National Health and Nutrition Examination Survey 2005–2010).
<sup>a</sup> Six phytoestrogens (tertiles) were entered into model 1 simultaneously. <sup>b</sup> Model 2 additionally adjusted for age. <sup>c</sup> Model 3 further adjusted for race, education, marital status, occupation, family income, body mass index, physical activity, alcohol use, smoking status, depressive symptoms, diabetes, hypertension, use of female hormones (only in females), C-reactive protein (mg/dL), and caffeine intake (mg/day). \* *p* < 0.05, \*\* *p* < 0.01, \*\*\* *p* < 0.001.
**Figure 5.** Concentration–response relationship of the urinary enterolactone concentration with sleep disorders for females (**a**) and males (**b**). The solid line represents the estimated odds ratios (ORs) and the dashed lines represent their 95% confidence intervals. The relationship adjusted for age, marital status, race, occupation, family income, educational level, body mass index, smoking status, alcohol consumption, use of female hormones (only in females), physical activity, caffeine intake, C-reactive protein, hypertension, depressive symptoms, diabetes, and the other five phytoestrogens (tertiles).
**Figure 6.** Concentration–response relationship of the urinary enterodiol concentration with sleep disorders for females (**a**) and males (**b**). The solid line represents the estimated odds ratios (ORs) and the dashed lines represent their 95% confidence intervals. The relationship adjusted for age, marital status, race, occupation, family income, educational level, body mass index, smoking status, alcohol consumption, use of female hormones (only in females), physical activity, caffeine intake, C-reactive protein, hypertension, depressive symptoms, diabetes, and the other five phytoestrogens (tertiles).
Furthermore, the associations between the urinary phytoestrogen concentrations and sleep disorders stratified by age group separately for males and females are shown in Table 5. The urinary O-DMA concentration was positively associated with sleep disorders in males aged 18–39 years (OR: 6.57, 95% CI: 2.06–20.99) and females aged 40–59 years (OR: 15.14, 95% CI: 2.99–76.65), whereas the O-DMA concentration was inversely related to sleep disorders in females aged 60 years or over (OR: 0.25, 95% CI: 0.10–0.58). The urinary enterodiol concentration was positively related to sleep disorders in females aged 18–39 (OR: 6.52, 95% CI: 1.69–25.23) and 40–59 years (OR: 13.66, 95% CI: 2.06–90.70). Furthermore, the urinary equol concentration was inversely associated with sleep disorders in females aged 18–39 years (OR: 0.08, 95% CI: 0.01–0.84), and daidzein concentration was positively related to sleep disorders in females aged 60 years or over (OR: 10.67, 95% CI: 1.88–60.44).
**Table 5.** Weighted odds ratios (95% confidence intervals) for sleep disorders across tertiles of urinary phytoestrogen concentrations stratified by age for males and females (National Health and Nutrition Examination Survey 2005–2010).
**Table 5.** *Cont.*
<sup>a</sup> Adjusted for education, marital status, occupation, family income, body mass index, physical activity, alcohol use, smoking status, depressive symptoms, diabetes, hypertension, use of female hormones (only in females), C-reactive protein (mg/dL), caffeine intake (mg/day), and the other five phytoestrogens (tertiles). \* *p* < 0.05, \*\* *p* < 0.01.
The associations between the urinary phytoestrogen concentrations and sleep disorders stratified by race are displayed in Table 6. An interesting finding was that the urinary O-DMA concentration was positively related to sleep disorders in non-Hispanic Whites (OR: 2.16, 95% CI: 1.31–3.55) but inversely associated with sleep disorders in other Hispanics (OR: 0.13, 95% CI: 0.02–0.86). Furthermore, the urinary enterolactone concentration was negatively related to sleep disorders (OR: 0.56, 95% CI: 0.33–0.95), while the enterodiol concentration was positively associated with sleep disorders (OR: 1.89, 95% CI: 1.13–3.13) in non-Hispanic Whites and the equol concentration was positively related to sleep disorders in other Hispanics (OR: 3.22, 95% CI: 1.11–9.30).
**Table 6.** Weighted odds ratios (95% confidence intervals) for sleep disorders across tertiles of urinary phytoestrogen concentrations stratified by race (National Health and Nutrition Examination Survey 2005–2010).
**Table 6.** *Cont.*
**Table 6.** *Cont.*
<sup>a</sup> Adjusted for age, gender, education, marital status, occupation, family income, body mass index, physical activity, alcohol use, smoking status, depressive symptoms, diabetes, hypertension, use of female hormones, C-reactive protein (mg/dL), caffeine intake (mg/day), and the other five phytoestrogens (tertiles). \* *p* < 0.05, \*\* *p* < 0.01.
Table 7 presents the weighted relative risk ratios (RRRs) with 95% CIs of sleep duration according to tertiles of the urinary phytoestrogen concentrations. In the fully adjusted model, compared with normal sleep (7–8 h/night), the urinary enterolactone concentration was negatively associated with very short sleep (<5 h/night) (RRR: 0.56, 95% CI: 0.36–0.86), and the genistein concentration was inversely related to a long sleep risk (≥9 h/night) (RRR: 0.62, 95% CI: 0.39–0.99).
The associations of urinary phytoestrogen concentrations with sleep duration stratified by age and gender are shown in Tables 8 and 9, respectively. In the fully adjusted model, compared with normal sleep, the urinary enterolactone concentration was still negatively associated with very short sleep among young adults (18–39 years) (RRR: 0.24, 95% CI: 0.09–0.61) and a long sleep risk among older adults (≥60 years) (RRR: 0.49, 95% CI: 0.29–0.84). The urinary genistein concentration was still inversely associated with a long sleep risk in middle-aged adults (40–59 years) (RRR: 0.10, 95% CI: 0.02–0.56) and males (RRR: 0.51, 95% CI: 0.30–0.87). The urinary enterodiol concentration was positively related to very short sleep among young adults (18–39 years) (RRR: 2.73, 95% CI: 1.20–6.21) and a long sleep risk in females (RRR: 2.14, 95% CI: 1.11–4.13).
**Table 7.** Weighted relative risk ratios (95% confidence intervals) for sleep duration (reference, 7–8 h/night) across tertiles of the urinary phytoestrogen concentrations (National Health and Nutrition Examination Survey 2005–2010).
<sup>a</sup> Adjusted for age, gender, race, education, marital status, occupation, family income, body mass index, physical activity, alcohol use, smoking status, depressive symptoms, diabetes, hypertension, use of female hormones, C-reactive protein (mg/dL), caffeine intake (mg/day), and the other five phytoestrogens (tertiles). \* *p* < 0.05, \*\* *p* < 0.01.
**Table 8.** Weighted relative risk ratios (95% confidence intervals) for sleep duration (reference, 7–8 h/night) across tertiles of the urinary phytoestrogen concentrations stratified by age (National Health and Nutrition Examination Survey 2005–2010).
**Table 8.** *Cont.*
**Table 8.** *Cont.*
<sup>a</sup> Adjusted for gender, race, education, marital status, occupation, family income, body mass index, physical activity, alcohol use, smoking status, depressive symptoms, diabetes, hypertension, use of female hormones, C-reactive protein (mg/dL), caffeine intake (mg/day), and the other five phytoestrogens (tertiles). \* *p* < 0.05, \*\* *p* < 0.01, \*\*\* *p* < 0.001.
**Table 9.** Weighted relative risk ratios (95% confidence intervals) for sleep duration (reference, 7–8 h/night) across tertiles of the urinary phytoestrogen concentrations stratified by gender (National Health and Nutrition Examination Survey 2005–2010).
**Table 9.** *Cont.*
<sup>a</sup> Adjusted for age, race, education, marital status, occupation, family income, body mass index, physical activity, alcohol use, smoking status, depressive symptoms, diabetes, hypertension, use of female hormones (only in females), C-reactive protein (mg/dL), caffeine intake (mg/day), and the other five phytoestrogens (tertiles). \* *p* < 0.05.
### **4. Discussion**
To our knowledge, the current study was the first evaluation of the associations between the urinary concentrations of individual phytoestrogens (enterolactone, enterodiol, daidzein, O-DMA, equol, and genistein) and sleep disorders and sleep duration among general American adults. This study was based on data from the NHANES 2005–2010 and found that the urinary enterolactone concentration was linearly inversely associated with the risk of sleep disorders, while the enterodiol concentration was positively related to sleep disorders in an approximately linear manner and the O-DMA concentration was positively associated with sleep disorders in a nonlinear (inverted L-shaped) manner. The urinary enterolactone concentration was negatively associated with very short sleep, and the genistein concentration was inversely related to a long sleep risk. Furthermore, a negative association of the urinary genistein concentration with sleep disorders was observed in middle-aged adults (40–59 years). The urinary enterolactone concentration was inversely associated with a long sleep risk among older adults (≥60 years), while the enterodiol concentration was positively related to a long sleep risk in females and very short sleep in young adults (18–39 years). Finally, an interesting finding was that the urinary O-DMA concentration was positively related to sleep disorders in both females aged 40–59 years and non-Hispanic Whites, but was inversely associated with sleep disorders in both females aged 60 years or over and other Hispanics.
There has been no epidemiological study that has appraised the relationship of total lignans or individual lignans (enterolactone or enterodiol) with sleep disorders to date. However, animal experiments have indicated that a lignan component could increase sleep duration and decrease sleep latency via modulating the γ-aminobutyric acid (GABA)-ergic system [38,39], which supports our findings that enterolactone was negatively associated with sleep disorders and very short sleep, and enterodiol was positively related to a long sleep risk in females. We also found that enterolactone was inversely associated with a long sleep risk among older adults (≥60 years), while enterodiol was positively associated with sleep disorders in the whole population and very short sleep in young adults (18–39 years). This seems to suggest that enterolactone might be beneficial for preventing both long and very short sleep, whereas enterodiol might be adverse toward this goal. However, further comparisons were difficult due to limited prior studies.
Some studies have evaluated the association between total isoflavone consumption and sleep but the results were contradictory. A study with a cross-sectional design performed in Japanese adults found that the intake of isoflavone was positively related to sleep quality and optimal sleep duration [36]. Another longitudinal study of Chinese adults reported that isoflavone intake was inversely related to falling asleep in the daytime in females and long sleep duration in both genders [37], which was similar to our result regarding genistein and a long sleep risk. Several trial studies also showed that isoflavone supplementation alleviated insomnia or sleep disorders among climacteric women [28–31]. However, other trial studies of isoflavone supplementation in climacteric women or androgen-deprived males found no significant improvement in insomnia or sleep quality [32–34]. Meanwhile, another randomized, placebo-controlled, double-blinded trial over six months performed on climacteric women indicated that insomnia was more frequent in the isoflavone supplementation group [35], and the longitudinal study reported that soy milk (one of the main food sources for isoflavone intake in the study) was positively related to falling asleep in the daytime among males [37]. Similarly, we found a positive association between O-DMA and sleep disorders.
Thus far, little is known about individual isoflavones (daidzein, O-DMA, equol, and genistein) and sleep. The current study found discrepant associations of them with sleep disorders, where O-DMA was positively associated with sleep disorders, and genistein was negatively associated with sleep disorders in middle-aged adults (40–59 years). Likewise, there were also differential relationships between them and sleep duration, where only genistein was significantly inversely related to a long sleep risk, and no significant associations were found between daidzein, O-DMA, and equol and sleep duration in this study. Variable abilities in different individuals regarding the metabolic transformation of isoflavones [23], diverse biologic activities, and ER affinities of these metabolites (O-DMA and equol), as well as other individual isoflavones (daidzein and genistein) [40,41], and discrepant associations between them and sleep disorders or sleep duration may partially lead to the contradictory findings of prior studies focusing only on total isoflavones. Furthermore, potential gender, age, and race differences regarding these associations may also partially contribute to the prior contradictory findings.
The underlying mechanisms of the associations of phytoestrogens with sleep are unclear, where the following are several possible explanations. Estrogen can influence the synthesis and transport of serotonin [62] that is involved in the regulation of wakefulness and sleep [63], and studies have revealed that estrogen therapy alleviates sleep disturbances and improves sleep quality [10,11]. Therefore, phytoestrogens may affect the sleep–wake cycle through their estrogenic or anti-estrogenic effects [19], which may also explain the discrepant relationships of individual phytoestrogens with sleep disorders or sleep duration. A notable finding in our study was that the urinary O-DMA concentration was positively related to sleep disorders in females aged 40–59 years but was inversely associated with sleep disorders in females aged 60 years or over, which may be partly attributed to the anti-estrogenic and estrogenic effects of phytoestrogens, depending on the level of endogenous estrogen [64]. We also found that the urinary O-DMA concentration was positively related to sleep disorders in non-Hispanic Whites but was inversely associated with sleep disorders in other Hispanics, which may be partially due to the race difference in the compositions of gut microbiota and daidzein-metabolizing phenotypes [61,65]. Equol has estrogenic activities [66], which may partly contribute to the inverse association of equol with sleep disorders in females aged 18–39 years, while the mechanisms for the positive association between equol and sleep disorders in other Hispanics need to be investigated. Furthermore, the inter-individual variation of phytoestrogen metabolism by gut microbiota [67] may also contribute to the complexity of
these results. The above-mentioned mechanisms may partly explain our findings, and further studies on the related mechanisms remain necessary.
This study has several advantages. First, the phytoestrogen assessments were based on urinary biomarkers, which reflected all food origins of phytoestrogens and took into account the metabolic transformation of intestinal flora, representing true and biologically effective exposures. Second, the relationships with sleep disorders or sleep duration were evaluated for individual phytoestrogens (enterolactone, enterodiol, daidzein, O-DMA, equol, and genistein), which made up for the fact that previous epidemiological studies did not investigate the relationship of lignans with sleep and focused only on total isoflavones while ignoring the potential differences of individual isoflavones. Third, our findings were based on data from the NHANES, which was carefully designed, of high quality, and nationally representative. Fourth, concentration–response relationships between individual phytoestrogens and sleep disorders were appraised in this study. Fifth, we further explored gender, age, and race differences in the associations of individual phytoestrogens with sleep disorders. Additionally, we adjusted for a wide range of confounders to control for potential confounding bias.
However, several potential limitations should also be considered. First of all, the cross-sectional design precludes the possibility of causal inference. Second, although phytoestrogen concentrations in spot urine are reliable biomarkers for phytoestrogen intake [23,50,51], it might be difficult to accurately reflect long-term intake information because spot urine was collected only once. Nonetheless, studies have demonstrated that phytoestrogen concentrations in spot urine are relatively stable and significantly related to dietary phytoestrogen intake over the long term [68,69]. Third, the sleep disorders assessment was via a self-reported doctor diagnosis, which might involve recall bias, and self-reported usual sleep duration might be not objective enough; however, objective sleep measurement, such as polysomnography, may be difficult to implement in large-scale surveys. Fourth, although three-cycle data with national representativeness were combined and the sample was relatively large, the case group might still be not enough due to the lower prevalence, which might lead to bias. Furthermore, although we adjusted multiple covariates, measurement errors and other factors affecting sleep quality might influence the present findings. However, the present results were approximate in the three models, suggesting that the results might be robust and exposure factors might be independently associated with sleep. Finally, we could not further explore the relationships between phytoestrogens and specific types of sleep disorders in virtue of the limited sleep disorders data.
### **5. Conclusions**
The current findings suggested that the urinary enterolactone concentration was linearly inversely associated with the risk of sleep disorders among American adults, whereas the enterodiol and O-DMA concentrations were positively related to sleep disorders in approximately linear and inverted L-shaped manners, respectively. Meanwhile, the urinary enterolactone concentration was negatively associated with very short sleep and the genistein concentration was inversely related to a long sleep risk. Furthermore, a negative association of urinary genistein concentration with sleep disorders was observed in middle-aged adults. The urinary enterolactone concentration was inversely associated with a long sleep risk among older adults, while the enterodiol concentration was positively related to a long sleep risk in females and very short sleep in young adults. Finally, a notable finding was that the association of the urinary O-DMA concentration with sleep disorders was different between females aged 40–59 years and females aged 60 years or over, as well as between non-Hispanic Whites and other Hispanics. It may be meaningful and vital to choose more specific individual phytoestrogens, not choosing total lignans or isoflavones for preventing or improving sleep problems, and to match target groups given the potentially differential associations of individual phytoestrogens with sleep, as well as the gender, age, and race differences in the associations. Prospective cohort or trial studies evaluating the relationships of individual phytoestrogens with sleep are warranted to confirm the current findings.
**Supplementary Materials:** The following are available online at http://www.mdpi.com/2072-6643/12/7/2103/s1, Table S1. The classifications of the covariates.
**Author Contributions:** Conceptualization, J.S. and D.Z.; methodology, J.S., W.W., and D.Z.; data curation, J.S., X.D., and H.J.; writing—original draft preparation, J.S.; writing—review and editing, H.J. and D.Z. All authors have read and agreed to the published version of the manuscript.
**Funding:** This research received no external funding.
**Acknowledgments:** This study used data from the National Health and Nutrition Examination Survey (NHANES). The authors would like to thank all participants and contributors to the NHANES. This work was supported by the Taishan Scholars Construction Project.
**Conflicts of Interest:** The authors declare no conflict of interest.
### **References**
© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
| doab | 2025-04-07T03:56:58.608551 | 1-5-2021 17:49 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "003cf560-3166-422f-9a1c-98b1e971efa6",
"url": "https://mdpi.com/books/pdfview/book/3241",
"author": "",
"title": "Dietary Polyphenols and Human Health",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783039437931",
"section_idx": 9
} |
003cf560-3166-422f-9a1c-98b1e971efa6.10 | *Article* **Rice Bran Phenolic Extracts Modulate Insulin Secretion and Gene Expression Associated with** β**-Cell Function**
### **Nancy Saji 1,2, Nidhish Francis 1,3, Lachlan J. Schwarz 1,4, Christopher L. Blanchard 1,2 and Abishek B. Santhakumar 1,2,\***
Received: 10 May 2020; Accepted: 20 June 2020; Published: 24 June 2020
**Abstract:** Oxidative stress is known to modulate insulin secretion and initiate gene alterations resulting in impairment of β-cell function and type 2 diabetes mellitus (T2DM). Rice bran (RB) phenolic extracts contain bioactive properties that may target metabolic pathways associated with the pathogenesis of T2DM. This study aimed to examine the effect of stabilized RB phenolic extracts on the expression of genes associated with β-cell function such as glucose transporter 2 (*Glut2*), pancreatic and duodenal homeobox 1 (*Pdx1*), sirtuin 1 (*Sirt1*), mitochondrial transcription factor A (*Tfam*), and insulin 1 (*Ins1*) in addition to evaluating its impact on glucose-stimulated insulin secretion. It was observed that treatment with different concentrations of RB phenolic extracts (25-250 μg/mL) significantly increased the expression of *Glut2*, *Pdx1*, *Sirt1, Tfam,* and *Ins1* genes and glucose-stimulated insulin secretion under both normal and high glucose conditions. RB phenolic extracts favourably modulated the expression of genes involved in β-cell dysfunction and insulin secretion via several mechanisms such as synergistic action of polyphenols targeting signalling molecules, decreasing free radical damage by its antioxidant activity, and stimulation of effectors or survival factors of insulin secretion.
**Keywords:** rice bran; phenolic extracts; β-cell function; gene expression; insulin secretion
### **1. Introduction**
Glucose homeostasis is regulated by a sequence of events within the pancreatic β-cells, which result in the secretion of insulin [1]. Typically, in the postprandial state, increased levels of glucose in plasma can initiate pancreatic β-cells to secrete insulin, consequently suppressing hepatic glucose output and increasing peripheral tissue glucose uptake [2]. However, impairment of glucose-stimulated insulin secretion as a result of oxidative stress and inflammation can result in β-cell dysfunction and insulin resistance, subsequently leading to the pathogenesis of type 2 diabetes mellitus (T2DM) [3]. There are several essential genes involved in insulin secretion pathways that are specifically expressed in pancreatic β-cells. They are known to be involved in the processes leading to insulin release from the initial glucose entry into the β-cells followed by mitochondrial adenosine triphosphate (ATP) generation and potassium and calcium membrane depolarization leading to exocytosis events [4]. They include, glucose transporter 2 (*Glut2*) [5], pancreatic and duodenal homeobox 1 (*Pdx1*) [6], sirtuin 1 (*Sirt1*) [7], mitochondrial transcription factor A (*Tfam*) [8], and insulin 1 (*Ins1*) [9].
The *Glut2* gene is located in the pancreatic plasma membrane and functions as a glucose transporter as part of the glucose-sensing mechanism for the stimulation of insulin secretion [2]. The *Pdx1* gene plays an important role in mitochondrial embryonic development and β-cell differentiation and is known to regulate the expression of a variety of different pancreatic endocrine genes, including *Glut2* [10]. The *Sirt1* gene is known to serve as a key energy redox sensor involved in generating ATP that helps promote glucose-stimulated insulin secretion in pancreatic β-cells and potentially contribute to β-cell adaptation in response to insulin resistance [1]. In the liver, skeletal muscles, and white adipose tissues, the *Sirt1* gene has key functions that include regulation of glucose production, improvement in insulin sensitivity via fatty acid oxidation, and control of the production of adipokines [7]. The *Tfam* gene plays an essential role in the maintenance of mitochondrial DNA (mtDNA) and replication [11]. Altered mitochondrial function is known to result in a defective oxidative metabolism, which seems to be involved in visceral fat gain and the development of insulin resistance [12]. Moreover, the *Tfam* gene is also involved in insulin exocytosis events by maintaining appropriate ADP/ATP ratio [4]. The *Ins1* gene and its transcription factors are regulated by the circulating levels of glucose [9]. It encodes the production of insulin that plays a vital role in the regulation of carbohydrate and lipid metabolism [13]. Therefore, a disruption in the function of these genes (*Glut2*, *Pdx1*, *Sirt1*, *Tfam,* and *Ins1*) in the pancreatic β-cells is known to impair insulin secretion and result in the development of T2DM.
Recent studies have shown the potential of plant-derived phenolic compounds in ameliorating β-cell dysfunction via their antioxidant and free radical scavenging properties [14–16]. Exposure of polyphenols to β-cells has also been responsible for the modulation of several signalling proteins, including transcription factors, protein kinase, and ion channels [17].
Rice bran (RB), a by-product of the rice milling process, is usually discarded or used as animal feed [18]. However, the bran layer is composed of several bioactive phytochemicals, including polyphenols and phenolic acids [19]. Although RB phenolic extracts are believed to target metabolic pathways associated with T2DM, the mechanisms behind its effect on gene expression under normal and diabetic conditions have not been investigated. This study aimed to determine the effect of RB phenolic extracts on the expression of genes (*Glut2*, *Pdx1*, *Sirt1*, *Tfam,* and *Ins1*) associated with insulin secretion pathways and on glucose-stimulated insulin secretion under normal and high glucose conditions.
### **2. Materials and Methods**
### *2.1. Chemicals and Reagents*
All chemicals and reagents used in this study were purchased from Promega Corporation (Madison, WI, USA), Bio-Rad (Hercules, CA, USA), or Sigma-Aldrich (St Louis, MO, USA).
### *2.2. Rice Bran Phenolic Extract Preparation*
Commercially stabilized RB (drum-dried), from an Australian grown Reiziq rice variety, was obtained from SunRice Australia, courtesy of their milling plant in Leeton, NSW, Australia and subsequently stored at 4 ◦C until further analysis. Phenolic compounds were extracted from stabilized RB using an acetone/water/acetic acid (70:29.5:0.5, v/v) mixture, the characterization of which has been described elsewhere [18]. It is known to contain several bioactive compounds including ferulic acid, p-coumaric acid, caffeic acid, vanillic acid, syringic acid, sinapic acid, feruloyl glycoside, shikimic acid, ethyl vanillate, tricin, and their isomers. The extract was reconstituted in 50% dimethyl sulfoxide (DMSO) and stored at −20 ◦C before starting cell culture studies.
### *2.3. Cell Culture Conditions*
INS-1E cells were maintained in RPMI 1640 media containing 11.1 mM glucose and supplemented with 2 mM L-Glutamine, 1 mM sodium pyruvate, 10 mM HEPES, 0.05 mM β-mercaptoethanol, 10% Fetal Bovine Serum, and 1% 10,000 U/mL Penicillin–10 mg/mL Streptomycin from Sigma-Aldrich (St Louis, MO, USA) at 37 ◦C in 5% CO2 and used before reaching passage 45.
### *2.4. Cytotoxicity Assay*
The cytotoxicity of RB phenolic extracts was examined using a resazurin red cytotoxicity assay wherein INS-1E cells were seeded into 96-well plates at a density of 50,000 cells per well and incubated for 24 h in the RPMI 1640 complete media. The cell count for experimental seeding was achieved with a Muse® Cell Analyzer from Luminex Corporation (Austin, TX, USA). INS-1E cells were then treated with 200 μL of freshly prepared RB phenolic extracts at various concentrations (25, 50, 100, 250, 500, 750, and 1000 μg/mL) for 6 h. Hydrogen peroxide (5 mM) was used as a positive control and 0.25% DMSO served as a negative control. Subsequently, all the treatment wells were emptied before adding 200 μL of resazurin red solution (14 mg/L) to each well and incubated for an additional 4 h at 37 ◦C in 5% CO2. The absorbance was measured on a microplate reader (FLUOstar Omega microplate reader, BMG Labtech, Offenburg, Germany) at 570 and 600 nm against a resazurin red blank. The percentage of cell viability was calculated as described by Saji, Francis [20]. Each treatment was measured in octuplicate.
### *2.5. Expression of Genes Associated with* β*-Cell Function*
### 2.5.1. Experimental Design
Two experimental conditions simultaneously tested were normal glucose treatment (11.1 mM) to represent a normal β-cell function and high glucose treatment (25 mM) to represent β-cell dysfunction under glucotoxic stress [4,15]. INS-1E cells were seeded at a density of 500,000 cells per well into 6-well plates and incubated for 24 h. To induce glucotoxicity, INS-1E cells were further subjected to 48 h incubation in RPMI 1640 complete media containing 25 mM glucose. Cells under both normal and high glucose conditions were treated for 6 h with RB phenolic extracts (25, 50, 100, and 250 μg/mL) and 0.125% DMSO served as the negative control. Each treatment was measured in quintuplicate.
### 2.5.2. Gene Expression Analysis
Total ribonucleic acid (RNA) extraction was conducted using the SV Total RNA Isolation System according to the manufacturer's instructions (Promega, Madison, WI, USA). RNA quality was determined using a NanoDrop™ 2000 c Spectrophotometer from Thermo Fisher Scientific (Waltham, MA, USA). Then, cDNA synthesis was conducted using a GoScript™ Reverse Transcriptase, according to the manufacturer's instructions (Promega, Madison, WI, USA).
Primers used for quantitative real-time polymerase chain reaction (qPCR) examinations are listed in Table 1. All of the qPCR primers were adapted from [21], designed using Primer3 software, and synthesized by Sigma-Aldrich (St Louis, MO, USA). The amplification efficiency was determined to be between 90–110% for all the primers before starting qPCR.
Gene expression was conducted in the CFX96 Touch™ Real-Time PCR Detection System (Bio-Rad) using SsoAdvanced™ Universal SYBR® Green Supermix (Bio-Rad) detection according to the manufacturer's instructions. The cycling conditions comprised 95 ◦C for 3 min, 95 ◦C for 10 s, and 60 ◦C for 30 s repeated for 39 cycles. The melt curve was generated at 65 ◦C for 5 s and 95 ◦C for 50 s. The endpoint or cycle threshold (Ct) values were obtained for all genes tested. The mean normalized expression of genes was determined using the Q-gene software application, as described by Muller, Janovjak [22]. *TfII*β served as the reference gene.
**Table 1.** The nucleotide sequences of the PCR primers used to assay gene expression by qPCR.
### *2.6. Glucose-Stimulated Insulin Secretion*
The preparation of supernatant for the glucose-stimulated insulin secretion assay was adapted from a previous study conducted by Bhattacharya, Oksbjerg [15] with slight modifications. Briefly, INS-1E cells were seeded into a 24-well plate at a density of 1 <sup>×</sup> 10<sup>5</sup> cells/well and incubated for 24 h or until 70–80% confluency was reached. The cells were treated with DMSO control and RB extracts at different concentrations (25–250 μg/mL) and incubated for 6 h. Cells were then starved with a Krebs-Ringer bicarbonate buffer (125 mM NaCl, 5.9 mM KCl, 1.28 mM CaCl2, 1.2 mM MgCl2, 25 mM HEPES, and 0.1% BSA at pH 7.4) containing 5 mM glucose for 1 h. Glucose-stimulated insulin secretion was then induced by treating cells with a Krebs-Ringer bicarbonate buffer containing either 11.1 mM or 25 mM glucose for 1 h. The supernatant containing secreted insulin was collected and stored at −20 ◦C until further analysis. Insulin secretion was measured using a Rat *Ins1*/Insulin ELISA Kit purchased from Sigma-Aldrich (St Louis, MO, USA) according to the manufacturer's instructions. Each treatment was measured in sextuplicate.
### *2.7. Statistical Analysis*
Statistical analysis was performed by one-way analysis of variance (ANOVA), followed by post-hoc Tukey's multiple comparisons test using GraphPad Prism 7 software (GraphPad Software Inc, San Diego, CA, USA) at a level of *p* < 0.05. The results are reported as mean ± standard deviation (SD).
### **3. Results**
### *3.1. Cytotoxicity of RB Phenolic Extracts on INS-1E Cells*
The cell viability of INS-1E cells 6 h post-exposure (Figure 1) to various concentrations of RB phenolic extracts did not display any cytotoxic effect on the INS-1E cells at the lower concentrations tested (25–250 μg/mL). However, higher concentrations (500–1000 μg/mL) displayed a reduction in cell viability. Optimal, non-toxic concentrations of RB extract were determined to be between 25–250 μg/mL under both normal and high glucose conditions.
**Figure 1.** INS-1E cell viability 6 h post-exposure to various concentrations of RB phenolic extracts. (**a**) Normal glucose conditions, (**b**) High glucose conditions (*n* = 8). Data are presented as Mean ± SD. Dimethyl sulfoxide, DMSO; Hydrogen Peroxide, H2O2; Insulin-secreting rat insulinoma cell, INS-1E; Rice bran, RB.
### *3.2. E*ff*ect of RB Phenolic Extracts on Expression of Genes Associated with* β*-Cell Function*
### 3.2.1. Expression of the *Glut2* Gene
Under normal glucose conditions, a significant increase (*p* < 0.01) in the expression of the *Glut2* gene was observed after treatment with 50 and 100 μg/mL of RB phenolic extracts when compared to that of the control. Under high glucose conditions, a significant increase (*p* < 0.0001) in the expression of the *Glut2* gene was also observed after treatment with 25–250 μg/mL of RB phenolic extracts when compared to that of the control (Figure 2).
**Figure 2.** Effect of RB phenolic extracts on the expression of the *Glut2* gene under normal and high glucose conditions in INS-1E cells. (**a**) Normal glucose conditions, (**b**) High glucose conditions (*n* = 5). The level of significance is indicated by the asterisks, whereby \* *p* < 0.05, \*\* *p* < 0.01. Data are presented as Mean ± SD. Dimethyl sulfoxide, DMSO; Insulin-secreting rat insulinoma cell, INS-1E; Glucose transporter 2, *Glut2*; Rice bran, RB.
### 3.2.2. Expression of the *Pdx1* Gene
A significant increase (*p* < 0.05) in the expression of the *Pdx1* gene was observed after treatment with 50 and 100 μg/mL of RB phenolic extracts under normal glucose conditions when compared to that of the control. A significant increase (*p* < 0.001) in the expression of the *Pdx1* gene was also observed after treatment with 25–250 μg/mL of RB phenolic extracts under high glucose conditions when compared to that of the control (Figure 3).
**Figure 3.** Effect of RB phenolic extracts on the expression of the *Pdx1* gene under normal and high glucose conditions in INS-1E cells. (**a**) Normal glucose conditions, (**b**) High glucose conditions (*n* = 5). The level of significance is indicated by the asterisks, whereby \* *p* < 0.05, \*\* *p* < 0.01, \*\*\* *p* < 0.001. Data are presented as Mean ± SD. Dimethyl sulfoxide, DMSO; Insulin promoter factor 1, *Pdx1*; Insulin-secreting rat insulinoma cell, INS-1E; Rice bran, RB.
### 3.2.3. Expression of the *Sirt1* Gene
There was no significant increase in the expression of the *Sirt1* gene observed under normal glucose conditions when compared to that of the control. However, under high glucose conditions, a significant increase (*p* < 0.0001) in the expression of the *Sirt1* gene was observed after treatment with 25–250 μg/mL of RB phenolic extracts when compared to that of the control (Figure 4).
**Figure 4.** Effect of RB phenolic extracts on the expression of the *Sirt1* gene under normal and high glucose conditions in INS-1E cells. (**a**) Normal glucose conditions, (**b**) High glucose conditions (*n* = 5). The level of significance is indicated by the asterisks, whereby \* *p* < 0.05, \*\* *p* < 0.01, \*\*\* *p* < 0.001, \*\*\*\* *p* < 0.0001. Data are presented as Mean ± SD. Dimethyl sulfoxide, DMSO; Insulin-secreting rat insulinoma cell, INS-1E; Rice bran, RB; Sirtuin 1, *Sirt1*.
### 3.2.4. Expression of the *Tfam* Gene
Under normal glucose conditions, a significant increase (*p* < 0.001) in the expression of the *Tfam* gene was observed after treatment with 25–250 μg/mL of RB phenolic extracts when compared to that of the control. However, under high glucose conditions, a significant increase (*p* < 0.05) in the expression of the *Tfam* gene was only observed after treatment with 25 μg/mL of RB phenolic extracts when compared to that of the control (Figure 5).
**Figure 5.** Effect of RB phenolic extracts on the expression of the *Tfam* gene under normal and high glucose conditions in INS-1E cells. (**a**) Normal glucose conditions, (**b**) High glucose conditions (*n* = 5). The level of significance is indicated by the asterisks, whereby \* *p* < 0.05, \*\* *p* < 0.01, \*\*\* *p* < 0.001. Data are presented as Mean ± SD. Dimethyl sulfoxide, DMSO; Insulin-secreting rat insulinoma cell, INS-1E; Mitochondrial transcription factor A, *Tfam*; Rice bran, RB.
### 3.2.5. Expression of the *Ins1* Gene
RB extract did not alter the expression of the *Ins1* gene under normal glucose treatment. However, under high glucose conditions, a significant increase in the expression of the *Ins1* gene was observed after treatment with 50 μg/mL (*p* < 0.01) and 100 μg/mL (*p* < 0.001) of RB phenolic extracts when compared to that of the control (Figure 6).
**Figure 6.** Effect of RB phenolic extracts on the expression of the *Ins1* gene under normal and high glucose conditions in INS-1E cells. (**a**) Normal glucose conditions, (**b**) High glucose conditions *(n* = 5). The level of significance is indicated by the asterisks, whereby \*\* *p* < 0.01, \*\*\* *p* < 0.001. Data are presented as Mean ± SD. Dimethyl sulfoxide, DMSO; Insulin 1, *Ins1*; Insulin-secreting rat insulinoma cell, INS-1E; Rice bran, RB.
### *3.3. Glucose-Stimulated Insulin Secretion*
Glucose-stimulated insulin secretion was observed to significantly increase after treatment with 25 (*p* < 0.0001), 50 (*p* < 0.0001), 100 (*p* < 0.0001), and 250 (*p* < 0.05) μg/mL of RB phenolic extracts under normal glucose conditions when compared to that of the control. Similarly, under high glucose conditions, a significant increase in glucose-stimulated insulin secretion was also observed after treatment with 25 (*p* < 0.0001), 50 (*p* < 0.05), and 100 (*p* < 0.05) μg/mL of RB phenolic extracts when compared to that of the control (Figure 7).
**Figure 7.** Effect of RB phenolic extracts on glucose-stimulated insulin secretion under normal and high glucose conditions in INS-1E cells. (**a**) Normal glucose conditions, (**b**) High glucose conditions (*n* = 6). The level of significance is indicated by the asterisks, whereby \* *p* < 0.05, \*\*\*\* *p* < 0.0001. Data are presented as Mean ± SD. Dimethyl sulfoxide, DMSO; Insulin-secreting rat insulinoma cell, INS-1E; Rice bran, RB.
### **4. Discussion**
Prolonged exposure of pancreatic β-cells to a high glucose environment is known to result in oxidative stress, consequently leading to the downregulation of pancreatic genes, in turn causing impaired β-cell function and insulin secretion [16]. Plant-derived phenolic compounds via their antioxidant, free radical scavenging and metal chelating properties have been observed to target metabolic pathways associated with the pathogenesis of T2DM [14]. The present study demonstrated that RB phenolic extracts effectively alter β-cell function in insulin-secreting cells by modulating the expression of genes and insulin secretion. It was observed that RB phenolic extracts upregulated the expression of key genes associated with β-cell function, including *Glut2*, *Pdx1*, *Sirt1*, *Tfam,* and *Ins1* both under normal and high glucose-induced stress conditions (Figures 2–6).
The *Glut2* gene primarily acts as a glucose transporter and the decreased expression of the *Glut2* gene is directly proportional to the loss of glucose-stimulated insulin secretion [5]. In this study, a significant increase in the expression of the *Glut2* gene was observed under normal conditions compared to that in high glucose conditions. This may have been caused by the increase in glucotoxic stress created by the high glucose environment, resulting in a reduced ability to maintain normal functioning as a glucose transporter. Nevertheless, a significant up-regulation of the *Glut2* gene was observed under both conditions compared to those of the respective controls after treatment with varying concentrations of RB extract (Figure 2). Similarly, studies in which phenolic compounds derived from *M. pumilum var. alata* extracts and purified phenolic compounds such as resveratrol were tested improved β-cell function, and insulin signalling was observed as a result of increased expression of the *Glut2* gene in the pancreas [21,23]. This is most likely due to the polyphenols targeting the exchange of calcium ions resulting in the exocytosis of insulin-containing granules, thereby favourably modulating β-cell function [5,24].
*Pdx1* gene expression is essential for the homeostatic regulation of the glucose-sensing system in β-cells [6]. It is also essential for survival and differentiation ofβ-cells as it primarily acts by upregulating the transcription of several β-cell-specific genes, including the *Ins* and *Glut2* genes [25]. Results obtained from this study show that under both normal and high glucose conditions, a significant upregulation of the *Pdx1* gene was evident after treatment with RB phenolic extracts (Figure 3). Upregulation of the *Pdx1* gene has been observed elsewhere, in which administration of *Teucrium polium* extract, known to contain phenolic compounds with strong antioxidant and anti-inflammatory effects, was found to reverse the symptoms of streptozotocin-induced diabetes in rats [26]. Another study, wherein the effect of gallic acid against glucolipotoxicity and insulin secretion was examined, showed that pre-treatment with different concentrations of gallic acid was found to increase insulin secretion and resulted in the upregulation of the *Pdx1* gene in RINm5F β-cells [27]. Reduction in insulin secretion has been attributed to the c-Jun N-terminal kinase (JNK) pathway activation under oxidative stress conditions. JNK activation as a result of oxidative stress results in forkhead box protein O1 (FOXO1) phosphorylation, and the nuclear localization of the FOXO1protein leads to a reduction in the expression of the *Pdx1* gene [28]. As an adequate expression of the pancreatic *Pdx1* gene is essential to maintain the proper function of insulin-producing β-cells, inhibition of the JNK pathway is crucial. As phenolic compounds are recognized to modulate the regulation of the JNK pathway [26], it is likely that the observed upregulation of the *Pdx1* gene by RB-derived phenolic extracts resulted from an inhibition of the JNK pathway.
The *Sirt1* gene is known to be a major contributor to the metabolic regulation of a cell via lipid metabolism and insulin secretion [7]. In the current study, under high glucose conditions, a significant increase in the expression of the *Sirt1* gene was observed after treatment with RB phenolic extracts (Figure 4). Sun, Zhang [29] demonstrated that resveratrol improved insulin sensitivity by repressing the protein tyrosine phosphatase (PTP) constitute and PTP1B transcription at the chromatin level (on the *Sirt1* gene) under normal versus insulin-resistant conditions. Hence, it is believed that upregulation of the *Sirt1* gene as a result of treatment with RB phenolics can potentially target PTP1B ranscription consequently improving insulin sensitivity.
Any disruption to the *Tfam* gene in the pancreatic β-cell is known to result in impaired insulin secretion, reduced β-cell mass, and, consequently, the development of T2DM [8]. The current study shows a significant increase in the expression of the *Tfam* gene under normal and high glucose conditions post-treatment with RB phenolic extracts (Figure 5). In an in vivo study where rats were gavaged with pterostilbene, *Tfam* gene expression was significantly increased in addition to improvements to glycaemic control and insulin resistance [30]. Furthermore, the treatment of INS-1E cells with resveratrol also displayed marked potentiation of glucose-stimulated insulin secretion as a result of the up-regulation of *Tfam* [21]. From the above studies, it is believed that RB phenolics have the potential to enhance the efficiency of mitochondrial function via interaction with transcription factors such as *Tfam*.
Appropriate regulation of the *Ins1* gene is essential for central insulin signalling as it is an anorectic gene that encodes for the production of the insulin hormone that plays a vital role in the regulation of carbohydrate and lipid metabolism [31]. Chronic exposure to high glucose conditions can reduce the expression of the *Ins1* gene in β-cells and is often accompanied by the decreased binding activity of the β-cell-specific transcription factor, *Pdx1* [32]. In the current study, although there was no significant increase in *Ins1* gene expression after RB extract treatment under normal glucose conditions, the expression of the *Ins1* gene was significantly upregulated under high glucose conditions (Figure 6). Similarly, an in vivo study by the author of [33] also demonstrated blueberry-leaf extract rich in chlorogenic acid and flavonol glycosides attenuates glucose homeostasis and improves pancreatic β-cell function by increasing the expression of several genes including *Ins1.* Polyphenols present in common spices, such as cinnamon, cloves, turmeric, and bay leaves, due to their doubly-linked procyanidin type-A polymers, have also shown an insulin-like activity in vitro [34]. The mechanism of cinnamon's insulin-like activity may be in part due to increases in the amounts of insulin receptor β and *Glut4* expression [34]. Some of the polyphenols present in cinnamon include caffeic, ferulic, *p*-coumaric, protocatechuic, and vanillic acids [35], a similar phenolic profile observed in the RB samples used in this study [18]. Therefore, it is likely that the effects observed in this study may be due to the insulin-like activity displayed by the polyphenols present in RB individually or via synergistic bioactivity.
Hormones such as insulin and amylin are co-secreted by β-cells in a fixed molecular ratio that provides circulating energy in the form of glucose and stored energy in the form of visceral adipose tissue [36]. However, conditions such as obesity, T2DM, and pancreatic cancer result in an increase in the amount of amylin relative to the insulin, which can disturb the regulation of energy homeostasis [36]. It was observed that under normal and high glucose-induced conditions, RB phenolic extracts significantly increased glucose-stimulated insulin secretion (Figure 7). Bhattacharya, Oksbjerg [15] also observed a similar trend where caffeic acid, naringenin, and quercetin significantly increased glucose-stimulated insulin secretion under hyperglycaemic and glucotoxic conditions in INS-1E cells. Similarly, several other phenolic compounds such as ferulic acid [37] and *p*-coumaric acid [38] have also been shown to increase insulin secretion both in vitro and in vivo, respectively. In this study, it was observed that RB phenolic compounds increase the expression of both the *Ins1* gene and the secretion of insulin in INS-1E cells under high glucose conditions. Since the *Ins1* gene is known to encode for the production of insulin hormone, this may indicate that there may be a correlation between insulin secretion and the expression of the *Ins1* gene.
Furthermore, it was observed that lower doses of the RB extract used in this study favourably modulated β-cell function associated gene expression and insulin secretion when compared to the higher doses in vitro. This phenomenon may be explained through the effect of hormesis, a biphasic dose-response to an environmental agent, wherein glucose-stimulated insulin secretion was observed to have a stimulatory or beneficial effect at low doses and an inhibitory or toxic effect at high doses of RB extract [39]. Dietary polyphenols are known to have strong cytoprotective effects, however, the hormetic role of dietary antioxidants in free radical-related diseases have demonstrated that under
uncontrolled nutritional supplementation, gene induction effects and the interaction with detoxification responses can result in a negative response by generating more reactive and harmful intermediates [40].
As a result of hindrance by cereal matrices, most of the bound phenolic compounds present in cereal grains are usually not readily accessible by digestive enzymes, leading to low bioavailability [41]. Studies have demonstrated that this could be improved by increasing their accessibility through suitable processing techniques, for example, thermal treatments [18,41]. The RB sample examined in this study was previously studied with respect to several thermal treatments. Of the treatments studied, drum drying resulted in the optimal antioxidant activity and was therefore selected for the current investigation [18]. The drum-dried RB samples resulted in a total free phenolic content of 362.17 ± 34.16 gallic acid equivalent (GAE)/100 g of RB with antioxidant activity of 975.33 <sup>±</sup> 20.24 Fe2+/100 g of RB and a total bound phenolic content of 160.65 ± 5.52 GAE/100 g of RB with antioxidant activity of 551.91 <sup>±</sup> 8.82 Fe2+/100 g of RB. This was much higher compared to that of a non-treated sample that had a total free phenolic content of 238.26 ± 30.34 GAE/100 g of RB with antioxidant activity of 621.76 <sup>±</sup> 26.76 Fe2+/100 g of RB and a total bound phenolic content of 222.94 <sup>±</sup> 3.74 GAE/100 g of RB with antioxidant activity of 712.37 <sup>±</sup> 14.57 Fe<sup>2</sup>+/100 g of RB [18].
### **5. Conclusions**
This study has demonstrated that RB phenolic compounds, under both normal and glucotoxic conditions, significantly increase the expression of genes associated with β-cell function, in addition to increasing glucose-stimulated insulin secretion. RB phenolic compounds could play an important role in modulating the expression of genes involved in β-cell dysfunction and insulin secretion via several mechanisms, including (1) Synergistic action of polyphenols and phenolic acids by targeting signalling molecules, including transcription factors, consequently modulating mitochondrial potential; (2) Reducing free radical damage related to β-cell dysfunction via their antioxidant activity; and (3) Stimulation of effectors or survival factors of insulin secretion. RB phenolic extracts present as a promising preventive/therapeutic target in the treatment of glucotoxicity induced β-cell dysfunction. More in vivo studies are warranted to confirm the bioactivity of RB phenolic compounds.
**Author Contributions:** N.S. performed the experiments outlined in this study and drafted the manuscript. N.F., L.J.S., C.L.B., and A.B.S. were involved in the experimental study design, preparation, and review of this manuscript. All authors have read and agreed to the published version of the manuscript.
**Funding:** This study was funded by the Australian Research Council Industrial Transformation Training Centre for Functional Grains (Project ID 100737) and from AgriFutures, Australia (PRJ-011503). We would like to acknowledge the Faculty of Science, Charles Sturt University, for providing funding towards the publication cost of this article.
**Acknowledgments:** The authors acknowledge SunRice, Australia, for providing the rice bran samples used in this study. We would like to acknowledge Kathryn Aston-Mourey, Head of Islet Biology Laboratory at Deakin University, Australia, for kindly donating the INS-1E cells.
**Conflicts of Interest:** The authors declare no conflicts of interest.
### **Abbreviations**
c-Jun N-terminal kinase (JNK), Dimethyl sulfoxide (DMSO), Forkhead box protein O1 (FOXO1), Gallic acid equivalent (GAE), Glucose transporter 2 (*Glut2*), Insulin 1 (*Ins1*), Insulin-secreting rat insulinoma cell (INS-1E), Mitochondrial DNA (mtDNA), Mitochondrial transcription factor A (*Tfam*), One-way analysis of variance (ANOVA), Pancreatic and Duodenal Homeobox 1 (*Pdx1*), Protein tyrosine phosphatase (PTP), Quantitative real-time polymerase chain reaction (qPCR), Ribonucleic acid (RNA), Rice bran (RB), Sirtuin 1 (*Sirt1*), Standard deviation (SD), Type 2 diabetes mellitus (T2DM).
### **References**
© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
| doab | 2025-04-07T03:56:58.610740 | 1-5-2021 17:49 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "003cf560-3166-422f-9a1c-98b1e971efa6",
"url": "https://mdpi.com/books/pdfview/book/3241",
"author": "",
"title": "Dietary Polyphenols and Human Health",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783039437931",
"section_idx": 10
} |
003cf560-3166-422f-9a1c-98b1e971efa6.11 | *Article* **Black Sorghum Phenolic Extract Modulates Platelet Activation and Platelet Microparticle Release**
### **Borkwei Ed Nignpense 1, Kenneth A Chinkwo 1,2, Christopher L Blanchard 1,2 and Abishek B Santhakumar 1,2,\***
Received: 20 May 2020; Accepted: 10 June 2020; Published: 12 June 2020
**Abstract:** Platelet hyper-activation and platelet microparticles (PMPs) play a key role in the pathogenesis of cardiovascular diseases. Dietary polyphenols are believed to mimic antiplatelet agents by blunting platelet activation receptors via its antioxidant phenomenon. However, there is limited information on the anti-platelet activity of grain-derived polyphenols. The aim of the study is to evaluate the effects of sorghum extract (Shawaya short black 1 variety), an extract previously characterised for its high antioxidant activity and reduction of oxidative stress-related endothelial dysfunction, on platelet aggregation, platelet activation and PMP release. Whole blood samples collected from 18 healthy volunteers were treated with varying non-cytotoxic concentrations of polyphenol-rich black sorghum extract (BSE). Platelet aggregation study utilised 5 μg/mL collagen to target the GPVI pathway of thrombus formation whereas adenine phosphate (ADP) was used to stimulate the P2Y1/P2Y12 pathway of platelet activation assessed by flow cytometry. Procaspase-activating compound 1 (PAC-1) and P-selectin/CD62P were used to evaluate platelet activation- related conformational changes and degranulation respectively. PMPs were isolated from unstimulated platelets and quantified by size distribution and binding to CD42b. BSE treatment significantly reduced both collagen-induced platelet aggregation and circulatory PMP release at 40 μg/mL (*p* < 0.001) when compared to control. However, there was no significant impact of BSE on ADP-induced activation-dependent conformational change and degranulation of platelets. Results of this study suggest that phenolic rich BSE may confer cardio-protection by modulating specific signalling pathways involved in platelet activation and PMP release.
**Keywords:** black sorghum; polyphenols; platelets; platelet microparticles; atherosclerosis
### **1. Introduction**
According to a World Health Organisation report, cardiovascular diseases accounted for an estimated 31% of deaths globally with majority being a result of stroke or heart attack [1,2]. In clinical settings, treatment involves blunting the activity of platelets using antiplatelet drugs. These drugs interfere with the thrombotic pathophysiology—wherein a rupture of an atherosclerotic plaque triggers platelet hyper-activation resulting in unwanted clot formation and occlusion of the blood vessel. Macrovesicles referred to as platelet microparticles (PMPs) are released following platelet activation and can contribute to the thrombotic situation [3,4].
The several signalling pathways involved in platelet activation and thrombus formation include receptor-agonist pathways such as P2Y1/P2Y12-ADP, GPVI-collagen, PAR1-thrombin and the COX-1-thromboxane [5]. An agonist such as collagen when exposed by atherosclerotic plaque may activate nearby platelets by binding to their GPVI receptor resulting in complex intracellular signalling
that produce a conformational change (indicated by GPIIb/IIIa receptor expression), degranulation (indicated by P-selectin secretion) and subsequent formation of platelet aggregates [5]. In addition, PMP released upon activation possess adhesive and pro-coagulant platelet-derived receptors that further enhance thrombus formation, thereby acting as biomarkers of platelet activation [3]. The common antiplatelet agents, clopidogrel and aspirin, used in clinical treatments inhibit platelet activation and its circulating biomarkers by selectively targeting P2Y1/P2Y12-ADP and COX-1-thromboxane respectively [6]. Unfortunately, because of the unresponsiveness and side effects associated with administration there have been considerable research in dietary bioactive agents known as polyphenols [7].
One such example of a polyphenol-rich functional food is sorghum whole grain. Although mainly used as animal feed, studies have demonstrated that it possesses anti-inflammatory, anti-cancer and antioxidant properties which add value to it as a food for human consumption [8]. Sorghum of different types exist that are classified based on the pigmentation of the pericarp and vary in their phenolic content [9]. The polyphenols found in sorghum that contribute to its bioactivity include flavonoids, hydroxybenzoic acids and hydroxycinnamic acid [8]. Furthermore Francis et al. [10] recently demonstrated that black sorghum rich in catechins and their derivatives may confer cardioprotective properties. The treatment of human umbilical vein cells with flavonoid-rich extract was found to prevent oxidative stress-related endothelial dysfunction through the modulation of gene expression.
Furthermore, these cardio-protective benefits of polyphenols apply in the context of platelet function. Several studies have demonstrated that polyphenols may inhibit platelet activation, adhesion, degranulation and aggregation by targeting specific thrombogenic pathways for example P2Y1/P2Y12-ADP, GPVI-collagen, PAR1-thrombin and the COX-1-thromboxane. As reviewed by Ed Nignpense et al. [11] many of the studies that investigate the polyphenol impact on platelet function and PMP generation utilise aggregometry and flow cytometry. However there is limited research on sorghum-derived polyphenols in modulating biomarkers of platelet activation. This study aims to investigate the impact of black sorghum derived polyphenol extracts on collagen-induced platelet aggregation, ADP-induced platelet activation and PMP generation.
### **2. Materials and Methods**
### *2.1. Research Ethics*
The study protocol was approved by the Charles Sturt University Human Research Ethics Committee (HR17012) and the Institutional Biosafety Committee (19HB02). The study was performed in compliance with relevant laws and institutional guidelines.
### *2.2. Volunteer Recruitment*
Eighteen healthy volunteers between 18–65 years of age (9 males and 9 females) were recruited from Charles Sturt University and the local community. Informed consent was obtained from all participants prior to commencement of the study. The criteria for recruitment involved normal health status with no history of conditions such as cardiovascular, metabolic, liver or lung disease. Other parameters that could affect the integrity of the analysis such as alcohol consumption, smoking, pregnancy, allergies or venepuncture difficulty were considered during the recruitment process. A health screening questionnaire was used to assess the already mentioned parameters. A dietary questionnaire (adapted from WINTEC and NZ academy of sport) was used to assess the usual dietary intake of volunteers and to avoid recruitment of participants on a high antioxidant diet. The cut-off figure for each type of food listed in the questionnaire was based on nutrient reference ranges for Australia and New Zealand—recommended daily nutrient intake values.
### *2.3. Blood Collection and Processing*
After fasting for at least 8 h, whole blood was collected from each participant by a trained phlebotomist into a tri-potassium ethylene diamine tetra-acetic acid (EDTA-1.8 mg/mL concentration) anticoagulant tube (Vacuette Greiner Bio-one, Interpath Services, Heidelberg West, VIC, Australia) and a tri-sodium citrate (28.12 g/L concentration) anticoagulant tube (Becton, Dickson and Company, North Ryde, NSW, Australia). A 20-mL syringe (Becton, Dickson and Company, North Ryde, NSW, Australia) and 21-gauge 1.5-inch needle (Terumo Medical Corporation, Macquarie Park, Australia) were used to draw blood from the median cubital vein. The purpose of choosing a larger gauge was to avoid the activation of platelets while drawing or dispensing blood. Utmost care was taken to ensure samples were not obtained through a traumatic collection and that none contained obvious clots. In addition, care was taken to ensure minimal specimen handling and agitation in order to prevent artefactual platelet activation. The first 2 mL of blood was discarded before drawing into the tri-sodium citrate tubes in order to avoid the risk of collecting platelets activated by venepuncture. Tri-sodium tubes were used for aggregometry and flow cytometry assays whereas the EDTA was used to perform full blood examinations.
### *2.4. Full Blood Examination*
Using an Abbott CELL-DYN Emerald 22 Haematology Analyser (Abbott Diagnostics, Illinois, USA), a full blood examination (FBE) was performed on all samples. The FBE results of volunteers indicated that the blood cell parameters were within normal reference ranges as determined by the Royal College of pathologists of Australia. Individuals with cell counts, especially platelet counts, outside of the reference range were excluded from the study. Quality control validation and maintenance were all performed according to the Abbott CELL-DYN Emerald 22 Haematology Analyser manual.
### *2.5. Extraction of Black Sorghum Polyphenols*
Sorghum (*Sorghum bicolor*) samples of six different pericarp varieties were obtained from glasshouse trials conducted by Curtin University, Perth, Western Australia. Six pigmented varieties of sorghum were cultivated under the same conditions, grown in a glasshouse equipped with Lumisol Clear AF cover (200 μm thick, at a transparency of ca UV-A 94%, UV-B 84% and photosynthetic active radiation (PAR, 400-700 nm) 93% [8]. Extraction and analysis of phenolic composition and antioxidant activity were performed previously using methods described by Rao et al. [8]. Among the different sorghum varieties, the black pericarp variety (Shawaya short black 1) was selected for this study because of its relatively high antioxidant activity when analysed with ferric reducing antioxidant power (FRAP; 20.19 ± 2.69 mg/g TE) and 2,2-dipheny-1-picrylhydrazyl (DPPH; 18.04 ± 3.53 mg/g TE) antioxidant assays (Supplementary Figure S1, Tables S1 and S2). The highest level of polyphenols found in the BSE included catechin derivatives, catechins and pentahydroxyflavanone-(3->4)-catechin-7-O- glucoside (Supplementary Table S2). Stock concentrations of BSE (20 mg/mL in 50% DMSO) were diluted in phosphate buffered saline (PBS) to achieve desired concentrations (5 μg/mL, 20 μg/mL and 40 μg/mL) in whole blood. Desired concentrations were selected based on viability studies done by Francis et al. [10].
### *2.6. Whole Blood Platelet Aggregometry*
Platelets in whole blood were stimulated for aggregation using 5 μg/mL collagen exogenous platelet agonists (DSKH Australia Pty. Ltd., Hallam, VIC, Australia) to investigate the effect of BSE treatment on the platelet aggregation. Five hundred microliters of citrated whole blood were added to 100 μL of 0.1% DMSO control (Sigma-Aldrich, Castle Hill, NSW, Australia) or BSE stock concentrations (5 μg/mL, 20 μg/mL and 40 μg/mL) and mixed with 400 μL of saline. The sample was then incubated at 37 ◦C for 20 min. Using a Chrono-log model 700 lumi-aggregometer (DKSH Australia Pty. Ltd., Hallam, VIC, Australia) the samples were analysed by means of electrical impedance (ohms) to determine the amount of platelet aggregation occurring in the sample over a 6-min time period (Supplementary Figure S2).
### *2.7. Flow Cytometry*
### 2.7.1. Standardisation
Flow-check fluorospheres were run as quality control for optimal laser alignments. Antibody capture beads (Anti-Mouse Ig, K CompBeads, BD Biosciences, North Ryde, NSW, Australia) were used for single colour compensation controls in order to achieve optimal compensation. Megamix beads (0.1 μm, 0.3 μm, 0.5 μm, 1 μm) from Biocytex, Marseille, France were used as per manufacturer's instructions to set up an appropriate gating to detecting microparticles. They were run before each PMP analysis.
### 2.7.2. Measurement of Platelet Activation-Dependent Conformational Change and Degranulation
The effects of BSE on ADP-induced platelet activation were analysed and interpreted using a Gallios flow cytometer (Beckman Coulter, Inc., Lane Cove NSW, Australia). The protocols were adopted from the method described by Santhakumar et al. [12] with some modifications. Platelet activation and thrombogenic indicators were assessed via activation-dependent platelet monoclonal antibodies (mAbs) purchased from Becton, Dickinson and Company, North Ryde, NSW, Australia. Procaspase activating compound-1 (PAC-1)-fluorescein isothiocyanate-fluorescein isothiocyanate was used to detect platelet activation-related conformational change and P-selectin/CD62P-allophycocyanin highlighted activation dependent degranulation. CD42b-phycoerythrin identified the GPIb-IX-V receptor, a common receptor found on the surface of all platelets, activated and resting included. A decreased expression of mAb exhibits alleviation of thrombogenesis. Within 5 min of collection tri-sodium citrated whole blood was used for assay preparation to avoid artefactual activation of platelets. A volume of 40 μL of blood was incubated with DMSO control or the various BSE concentration for 20 min at 37 ◦C in the dark. A 10-μL mixture of all three monoclonal antibodies (3.33 μL each of PAC-1, CD62P and CD42b) was added to blood samples and incubated for 20 min at room temperature in the dark. To induce platelet activation, 10 μM ADP (Helena laboratories Pty. Ltd., Mt Waverly, VIC, Australia) was added, and samples were incubated for a further 15 min in the dark at room temperature, after which erythrocytes were lysed (575 μL of 10 % lysing solution). Samples were thoroughly vortexed to ensure homogeneity and incubated in the dark at room temperature for a further 15 min and then analysed. In all, 10,000 platelet events were acquired, gated based on light scatter and CD42b mAb expression and activated platelets were articulated as mean fluorescence intensity (MFI) (Supplementary Figure S3).
### 2.7.3. Measurement of Circulatory PMPs
Using the microparticle gating established with Megamix beads, PMPs were identified and quantified by size distribution and binding to CD42b (Supplementary Figure S4). The protocol for circulatory PMP analysis was adapted from Lu et al. [13]. A volume of 1 mL whole blood was added to micro-centrifuge tubes in the presence of PGE1 (120 nmol/L; Sigma-Aldrich, Castle Hill, NSW, Australia). PGE1 was added to prevent artefactual activation during centrifugation. The blood samples were incubated with the respective 100 μL BSE concentrations and DMSO control for 20 min at 37 ◦C in the dark. Each sample was then centrifuged for 15 min at a 1000 rpm and the resultant platelet rich plasma (PRP) was discarded. The remaining blood was spun a further 15 min at 3000 rpm. The supernatant rich in PMPs (40 μL) was collected into flow tubes and incubated with 4 μL of CD42b and 6 μL of stain buffer (Becton, Dickson and Company, North Ryde, NSW, Australia) in dark room for 15 min. Four percent formaldehyde was used to fix any activation of platelets left in the supernatant. After a 10-min incubation period the samples were run for PMP analysis on the flow cytometer.
### *2.8. Statistical Analysis*
A two-way ANOVA following Tukey's post comparison test was performed using GraphPad Prism version 8.0 for Windows (GraphPad Software, La Jolla, California, USA). A minimum sample size of 14 participants in total was required for 80% power to detect a 5% variation in the laboratory parameters measured where a 3–5% standard deviation exists in the population, assuming an alpha error of 0.05. All the data were expressed as mean ± standard deviation (SD). Differences between the groups were significant when *p* < 0.05. Any significant statistical interactions were included in the analysis where applicable.
### **3. Results**
The baseline parameters including full blood counts for all 18 participants were within normal reference ranges set by the Royal College of Pathologists of Australasia (Table 1) [14].
**Table 1.** Baseline full blood counts of participants.
Values are represented as mean ± Standard deviation (SD). RBC, red blood cell, PCV, packed cell volume, MCV, mean cell volume, MCH, mean cell haemoglobin, MCHC, mean cell haemoglobin concentration, RDW, red cell distribution width, MPV, mean platelet volume.
### *3.1. E*ff*ect of BSE on Whole Blood Platelet Aggregation and Platelet Activation*
BSE at 40 μg/mL concentration significantly reduced platelet aggregation stimulated by collagen by 19 % (*p* = 0.0004) (Figure 1). BSE at lower concentrations did not exhibit any significant reduction in aggregation.
It was observed that whole blood treatment with the varying concentrations of BSE did not significantly affect ADP-induced platelet conformational change and degranulation indicated by PAC-1 and P-selectin expression respectively (Supplementary Figures S5 and S6).
### *3.2. E*ff*ect of BSE on Circulatory PMPs*
BSE at a concentration of 40 μg/mL significantly reduced the amount of circulatory PMPs in whole blood by 47% (*p* = 0.0008). Lower concentrations of BSE did not exhibit any significant reduction to the amount circulatory PMPs (Figure 2).
**Figure 1.** The effect of varying concentrations of BSE on collagen-induced aggregation. BSE at 40 μg/mL significantly reduced platelet aggregation (5.3 ± 1.3; *p* value = 0.0004). BSE at 5 μg/mL and 20 μg/mL did not reduce platelet aggregation when compared to control (*p* value > 0.1). N = 18 and data is represented in aggregation (Ohms) versus BSE concentrations. \*\*\* signifies statistical significance *p* < 0.001 compared to control. Error bars expressed as mean ± SD.
**Figure 2.** The effect of varying concentrations of BSE on circulatory PMP production in vitro. BSE at 40 μg/mL significantly reduced the amount of circulatory PMPs (<24190 ± 4935, *p* = 0.0008). BSE at 5 μg/mL and 20 μg/mL did not reduce platelet aggregation when compared to control (*p* value > 0.1). N = 14 and data are represented in number of PMP events versus BSE concentrations. \*\*\* signifies statistical significance *p* < 0.001 compared to control. Error bars expressed as mean ± SD.
### **4. Discussion**
There is growing interest in understanding the therapeutic benefits of functional foods. Sorghum for example is one of the functional foods that is showing promise in this area. With sorghum-derived polyphenols already having demonstrated anti-inflammatory, anti-cancer and antioxidant properties, the current study aimed to evaluate the effects of polyphenol-rich BSE on platelet function in terms of aggregation, conformational change, degranulation and circulatory PMP production [8–10]. It was observed that BSE significantly inhibited collagen-induced platelet aggregation and decreased the release of circulatory PMPs but did not have a significant effect on ADP-induced platelet conformational change or degranulation. Although these results do not reflect a typical dose-dependent inhibition, they suggest a potential role of BSE polyphenols at optimum concentrations to interfere with pathways in the GPVI-collagen signalling and the release of circulatory PMPs but little or no effect on P2Y1/ P2Y12-ADP pathway.
To the best of our knowledge only a few studies have investigated the antiplatelet effects of sorghum extracts. Li, Yu and Fan et al. [15] extracted alditols and monosaccharides from sorghum vinegar to evaluate their anti-aggregation activity using the turbidimetric method. Results from their study indicated a significant dose-dependent inhibition of aggregation via multiple agonists, arachidonic acid, collagen, ADP and thrombin. Furthermore, a different study by Fan et al. [16] reported in vitro inhibition of ADP- and thrombin- induced rabbit platelet aggregation by methanolic extracts of aged sorghum vinegar with the half maximal inhibitory concentrations (IC50) of 1.7 ± 0.3 and 8.9 ± 1.9 mg/mL respectively. When rats were orally administered the extracts (>100 mg/kg), both collagen- and epinephrine-induced pulmonary thrombosis were inhibited. In comparison with the present study it is to be noted that these studies employed platelet-rich plasma rather than whole blood hence not accounting for the possible involvement of other blood cells and extracellular mediators involved in thrombus formation. In addition, sorghum vinegar extracts were used at higher concentrations; milligrams compared to micrograms used in this study. This raises the question of bioavailability and the importance of employing physiological concentrations of extracts.
Although the BSE concentration of 40 μg/mL at which antiplatelet effect were observed is relatively lesser in concentration than used in the other studies, the most bioactive compounds with respect to antioxidant activity were catechins and other flavonoids which are usually considered to have low bioavailability [8]. It has been suggested that the total plasma polyphenol concentration rarely exceeds 1 μM and that their antiplatelet effects are only found at high non-physiological concentrations (greater than 50 μM) [17,18]. However, it is likely that these plasma concentrations are underestimations due to the ability of polyphenols to bind to the surface of red blood cells and thereby exert their bioactivity [19]. Furthermore polyphenols (structurally related to catechin) and their metabolites have been shown to inhibit platelet function in vitro [20]. This highlights the possibility of sorghum catechins and their metabolites having antiplatelet effects in vivo despite bioavailability concern. Interestingly, an in vivo human dietary intervention trial compared consumption of red and white wholegrain sorghum-based pasta to a control pasta in order to investigate its acute effect on the total phenol content and antioxidant activity in the plasma of healthy subjects [21]. Results showed that when compared to the control pasta, the red sorghum pasta showed significantly increased net plasma phenolic content and antioxidant activity post consumption (from 216.90 ± 2.62 at baseline to 269.40 ± 2.33 at 2 h; *p* < 0.001), thus demonstrating a plausible correlation between antioxidant activity and sorghum polyphenol consumption—which in turn may contribute to antiplatelet effects.
The antiplatelet effects was observed with BSE-included inhibition of collagen-induced aggregation and circulatory PMP production but no effect on ADP-induced platelet activation. The absence of antiplatelet effects on the P2Y1/ P2Y12-ADP activation pathway suggests that BSE polyphenols are not mimicking the action of drugs such as clopidogrel that antagonise P2Y12 receptor activation [22]. However, the inhibition of collagen-induced aggregation suggest that BSE polyphenols interfere with GPVI-collagen signalling pathways by either blunting the GPVI receptor directly or by other mechanisms [5]. Previous studies have demonstrated that flavonoids, specifically quercetin and catechin, can act synergistically to inhibit collagen-induced aggregation by blunting the associated burst of H2O2 and subsequent PLC activation [23,24]. Thus, a possible mechanism of inhibition BSE flavonoids may be a synergistic antagonism of the positive feedback activation of intracellular signals triggered by H2O2. Moreover, it has been shown that the phosphorylation cascade initiated by collagen can be inhibited by flavonoids [25,26]. Flavones, especially apigenin and luteolin, by virtue of a double bond in the C2-C3 and the keto group in C4 can also inhibit collagen-induced activation by antagonizing the TxA2 receptor activation which is also involved in the positive feedback loop [27]. Besides inhibition of the GPVI-collagen signalling, BSE polyphenols showed inhibition of the circulatory PMP production.
To the best of our knowledge, this is the first study investigating the effect of sorghum-derived polyphenols on PMP production. In contrast to this study, other PMP studies have employed the use of Annexin V as well as the platelet specific antibody CD42b, to identify pro-coagulant PMPs by their phosphatidylserine expression and to limit background noise [28]. However, because of the heterogeneity of PMPs, not all PMPs express phosphatidylserine [29]. Moreover, the measurement of CD42b-positive PMPs alone is significant as its increase has been associated with an increased risk of coronary heart disease [30]. From the current study, the significant inhibition of CD42b-positive circulatory PMPs observed in vitro may be attributed to the antioxidant properties of BSE polyphenols. It is believed that the inhibition of PMP generation may be the result of neutralisation of H2O2, scavenging of other free radicals or interaction with intracellular signalling leading to PMP release.
The juxtaposition of both the present study and that of an earlier study by Francis et al. [10] highlights the multifaceted role of BSE polyphenols in cardio-protection. The group investigated the effects of BSE polyphenols on the expression of antioxidant- and inflammatory-linked genes involved in endothelial dysfunction under oxidative stress. Results indicated that BSE polyphenols alleviate oxidative stress–induced damage to endothelial cells. Since vascular dysfunction is a precursor to cardiovascular diseases, the current study builds upon earlier findings by exhibiting the antiplatelet effects of BSE. In the context of endothelial dysfunction, platelet activation and circulatory PMPs play central roles in the pathogenesis of atherothrombosis. The disruption of the plaque exposes collagen that binds to the GPVI receptor resulting in platelet activation and subsequent thrombus formation [31]. Circulatory PMPs may contribute to thrombosis via GPIb-IX-V receptor binding and have a pro-inflammatory effect to promote the development of the plaque [32]. Therefore, by reducing collagen-induced platelet aggregation and circulatory PMP generation, BSE polyphenols may be displaying the potential to augment thrombosis.
### **5. Conclusion and Future Considerations**
In summary, the present study contributes to the growing body of literature on bioactivity of sorghum polyphenols and highlights possible mechanisms of antiplatelet action that may result in cardiovascular health benefits. Because of the ability to reduce collagen-induced platelet activation and circulatory PMP generation, BSE polyphenols demonstrate the potential to interfere with pathological processes involved in vascular disorders and thrombotic complications. However, a larger panel of agonist for the flow cytometry and aggregometry studies will aid to further elucidate antiplatelet mechanisms. Because of the bioavailability concerns, well-controlled dietary intervention trials using larger sample sizes to evaluate the antiplatelet effects of sorghum consumption in healthy and pro-thrombotic populations are warranted to justify our findings. Because of the varied phenolic profiles of the different sorghum varieties, further research comparing the antiplatelet therapeutic potential of different grains is also warranted. Furthermore, this study attest to the measurement of circulatory PMPs as a biomarker of platelet activation to assess the bioactivity of functional foods.
**Supplementary Materials:** The following are available online at http://www.mdpi.com/2072-6643/12/6/1760/s1, Figure S1: Characterisation of phenolic compounds in BSE. Ultra-high-performance liquid chromatography (UHPLC) was employed to quantify the different phenolic compounds identified by the peak (on top). An online 2,2 -azino-bis(3-ethylbenzothiazoline-6-sulfonic acid) (ABTS) was coupled with UHPLC to quantify the relative antioxidant activity (peaks below) of each compound identified, Figure S2: A report derived from the collagen induced platelet aggregation study using the Chrono-log model 700 lumi-aggregometer (DKSH Australia Pty. Ltd, Hallam, VIC, Australia). The blue tracing represents the control (whole blood with no BSE) and the black tracing represents the whole blood pre-treated with 5 μg/mL BSE. The addition of BSE reduced the maximum platelet aggregation expressed in Ohms from 15 ohms to 11 ohms, Figure S3: A report from the ADP-induced platelet activation analysis using Kaluza Flow Cytometry Software (Beckman Coulter, Brea, CA, USA). Results indicate the gating of whole platelet population (CD42b positive events) and the proportion of activated platelets indicated by PAC-1 and P-selectin expression, Figure S4: A report from the PMP analysis using Kaluza Flow Cytometry Software (Beckman Coulter, Brea, CA, USA). Microparticle gating was established using Megamix beads of standard sizes. PMPs were distinguished from other microparticles by size (0.5 μm - 0.9 μm) and expression of CD42b. The number of CD42b positive events in the microparticle gate was used to quantify the PMPs, Figure S5: The effect of varying concentrations of BSE on PAC-1 expression. BSE did not significantly reduce ADP-induced platelet conformational change detected by PAC-1 expression (*p* values > 0.1 compared to control) N=14 and data is represented in mean fluorescence intensity (MFI) versus BSE concentrations. Error bars expressed as mean ± SD, Figure S6: The effect of varying concentrations of BSE on P-selectin expression. BSE did not significantly reduce ADP-induced platelet degranulation detected by P-selectin expression (*p* values > 0.1 compared to control ) N=14
and data is represented in mean fluorescence intensity (MFI) versus BSE concentrations Error bars expressed as mean ± SD, Table S1: Phenolic composition and antioxidant activity of sorghum varieties on as is basis, Table S2: List of top ten phenolic compounds identified in the black sorghum phenolic rich extracts by Q-TOF LC/MS and quantified using UHPLC-Online ABTS system (Adapted from Rao et al., 2018).
**Author Contributions:** B.E.N. conducted the experiments outlined in this study and drafted the manuscript. K.A.C., C.L.B. and A.B.S. were involved in the designing and critical reviewing of the manuscript. All authors have read and agreed to the published version of the manuscript.
**Funding:** This study was funded by the Australian Research Council Industrial Transformations Training Centre for Functional Grains (Project ID 100737).
**Acknowledgments:** The authors would like to acknowledge Graham Centre for Agricultural Innovation for providing funding towards open access publication of this article. Borkwei Ed Nignpense is a recipient of a PhD Scholarship from the Australian Government Research Training Program and also a recipient of a top-up scholarship by the Graham Centre.
**Conflicts of Interest:** The authors declare no conflict of interest.
### **Abbreviations**
### **References**
© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
*Article*
| doab | 2025-04-07T03:56:58.612576 | 1-5-2021 17:49 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "003cf560-3166-422f-9a1c-98b1e971efa6",
"url": "https://mdpi.com/books/pdfview/book/3241",
"author": "",
"title": "Dietary Polyphenols and Human Health",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783039437931",
"section_idx": 11
} |
003cf560-3166-422f-9a1c-98b1e971efa6.12 | **Dietary Polyphenol Intake is Associated with HDL-Cholesterol and A Better Profile of Other Components of the Metabolic Syndrome: A PREDIMED-Plus Sub-Study**
**Sara Castro-Barquero 1,2,**†**, Anna Tresserra-Rimbau 2,3,4,5,**†**, Facundo Vitelli-Storelli 6, Mónica Doménech 1,2, Jordi Salas-Salvadó 2,3,4,5, Vicente Martín-Sánchez 6,7, María Rubín-García 6, Pilar Buil-Cosiales 2,8,9, Dolores Corella 2,10, Montserrat Fitó 2,11, Dora Romaguera 2,12, Jesús Vioque 7,13, Ángel María Alonso-Gómez 2,14, Julia Wärnberg 2,15, José Alfredo Martínez 2,16,17, Luís Serra-Majem 2,18, Francisco José Tinahones 2,19, José Lapetra 2,20, Xavier Pintó 2,21, Josep Antonio Tur 2,12,22, Antonio Garcia-Rios 23, Laura García-Molina 7,24, Miguel Delgado-Rodriguez 13,25, Pilar Matía-Martín 26, Lidia Daimiel 17, Josep Vidal 27,28, Clotilde Vázquez 2,29, Montserrat Cofán 2,30, Andrea Romanos-Nanclares 8, Nerea Becerra-Tomas 2,3,4,5, Rocio Barragan 2,10, Olga Castañer 2,11, Jadwiga Konieczna 2,12, Sandra González-Palacios 7,13, Carolina Sorto-Sánchez 2,14, Jessica Pérez-López 2,15, María Angeles Zulet 2,16,17, Inmaculada Bautista-Castaño 2,18, Rosa Casas 1,2, Ana María Gómez-Perez 2,19, José Manuel Santos-Lozano 2,20, María Ángeles Rodríguez-Sanchez 21, Alicia Julibert 2,12,22, Nerea Martín-Calvo 2,8, Pablo Hernández-Alonso 2,3,4,5,31, José V Sorlí 2,10, Albert Sanllorente 2,11, Aina María Galmés-Panadés 2,12, Eugenio Cases-Pérez 32, Leire Goicolea-Güemez 2,14, Miguel Ruiz-Canela 2,8, Nancy Babio 2,3,4,5, Álvaro Hernáez 1,2, Rosa María Lamuela-Raventós 2,33 and Ramon Estruch 1,2,34,\***
Received: 7 February 2020; Accepted: 29 February 2020; Published: 4 March 2020
**Abstract:** Dietary polyphenol intake is associated with improvement of metabolic disturbances. The aims of the present study are to describe dietary polyphenol intake in a population with metabolic syndrome (MetS) and to examine the association between polyphenol intake and the components
of MetS. This cross-sectional analysis involved 6633 men and women included in the PREDIMED (PREvención con DIeta MEDiterranea-Plus) study. The polyphenol content of foods was estimated from the Phenol-Explorer 3.6 database. The mean of total polyphenol intake was 846 ± 318 mg/day. Except for stilbenes, women had higher polyphenol intake than men. Total polyphenol intake was higher in older participants (>70 years of age) compared to their younger counterparts. Participants with body mass index (BMI) >35 kg/m<sup>2</sup> reported lower total polyphenol, flavonoid, and stilbene intake than those with lower BMI. Total polyphenol intake was not associated with a better profile concerning MetS components, except for high-density lipoprotein cholesterol (HDL-c), although stilbenes, lignans, and other polyphenols showed an inverse association with blood pressure, fasting plasma glucose, and triglycerides. A direct association with HDL-c was found for all subclasses except lignans and phenolic acids. To conclude, in participants with MetS, higher intake of several polyphenol subclasses was associated with a better profile of MetS components, especially HDL-c.
**Keywords:** polyphenols; metabolic syndrome; Mediterranean diet; glignans; stilbenes; HDL-cholesterol
### **1. Introduction**
Polyphenols are plant-derived molecules characterized by the presence of one or more aromatic rings and attached hydroxyl groups [1]. They are classified into five subclasses according to their chemical structure, including flavonoids and nonflavonoids subclasses defined as phenolic acids, stilbenes, lignans, and other polyphenols. These bioactive compounds are responsible for some health and sensory properties of foods, such as bitterness, astringency, and antioxidant capacity. The intake of phenolic compounds and their food sources is highly variable and depends on dietary patterns, sex, socioeconomic factors, and the native foods of each region [2]. The Mediterranean diet (MedDiet) is characterized by a high intake of phenolic compounds because MedDiet interventions promote the intake of phenolic rich and plant-based products, such as legumes, vegetables, fruits, nuts and wholegrain cereals, and promote the use of extra virgin olive oil as the main source of fat. It has been suggested that phenolic compounds are partly responsible for the beneficial effects attributed to the MedDiet [3].
The metabolic syndrome (MetS) is defined as a cluster of metabolic disturbances, which include impaired glucose metabolism, elevated blood pressure, and low level of HDL-c, dyslipidemia, and abdominal obesity [4]. Sedentary lifestyle, smoking, and unbalanced diets are well-known modifiable risk factors for MetS, and lifestyle interventions in those areas, especially dietary interventions based on the MedDiet [3–6], might improve this condition. Considering the chronic low-grade inflammation and oxidative stress observed in MetS, polyphenols are good candidates to improve the condition because of their antioxidant and anti-inflammatory properties [7]. Moreover, several epidemiological studies have observed a negative association between polyphenol intake and MetS rates [8]. Regarding MetS components, an adequate intake of phenolic compounds has been shown to improve lipid profile and insulin resistance, and decrease blood pressure levels and body weight [8,9].
Despite the fact that phenol-rich dietary patterns are effective in improving some MetS components, there is no single phenolic compound or extract able to improve all the components of MetS [10]. Nevertheless, given the complexity of MetS and the heterogeneity of polyphenols, more large randomized trials with MetS patients are needed to evaluate the effect of polyphenol intake in reducing MetS complications, and whether intake of the different polyphenol subclasses could be associated with improvements in MetS components, because each subtype has different absorption and metabolism [11].
Therefore, the aims of our study were firstly to describe polyphenol intake in 6633 participants with MetS from the PREvención con DIeta MEDiterranea-Plus (PREDIMED-Plus) trial and to identify the main food sources of polyphenols in those participants, and secondly to examine whether higher intakes of some polyphenol sub-classes are associated with MetS components in this population.
### **2. Materials and Methods**
### *2.1. Design of the Study*
A cross-sectional analysis of the baseline data of participants included in the PREvención con DIeta MEDiterranea-Plus (PREDIMED-Plus) study was performed. The profile of the cohort, recruiting methods, and data collection processes have been described elsewhere [12] and on the website http://predimedplus.com. The study protocol was approved by the 23 recruiting centers Institutional Review Boards and registered in 2014 at the International Standard Randomized Controlled Trial Number registry (http://www.isrctn.com/ISRCTN89898870). All participants provided written informed consent before joining the study.
### *2.2. Participants*
A total of 6874 subjects were recruited and randomized in the 23 recruiting centers between September 2013 and December 2016. Primary care medical doctors from primary care centers of the National Health System assessed potential participants for eligibility. Eligible participants were men (aged 55–75 years) and women (aged 60–75 years) with overweight or obesity (body mass index [BMI] <sup>≥</sup>27 and <40 kg/m2) and at least three components of MetS according to the comprehensive definition of the International Diabetes Federation; National Heart, Lung, and Blood Institute; and American Heart Association (2009) [4]. Exclusion criteria were documented history of cardiovascular diseases (CVD), having a long-term illness, drug or alcohol use disorder, a BMI of 40 or higher, a history of allergy or intolerance to extra virgin olive oil or nuts, malignant cancer, inability to follow the recommended diet or physical activity program, history of surgical procedures for weight loss, and obesity of known endocrine disease (except for treated hypothyroidism). Of the total sample of 6874 randomized participants, 241 participants were excluded from the current analysis (Figure 1): 53 without food-frequency questionnaire (FFQ) data at baseline, and 188 participants who reported energy intake values outside the predefined limits (<3347 kJ [800 kcal]/day or >17,573 kJ [4000 kcal]/day for men; <2510 kJ [500 kcal]/day or >14,644 kJ [3500 kcal]/day for women) [13].
**Figure 1.** Flowchart of the participants.
### *2.3. Estimation of Dietary Polyphenol Intake*
The total dietary polyphenol intake and polyphenol subclasses were obtained at baseline by the 143-item FFQs used in the PREDIMED-Plus study. As described elsewhere [14], dietary polyphenol intake was estimated following these steps: (1) All foods from the FFQ with no polyphenol content, or only traces, were excluded; (2) recipes were calculated according to their ingredients and portions using traditional MedDiet recipes; (3) when an item from the FFQ included several foods (e.g., oranges and tangerines), the proportion of intake was calculated according to data available in the national survey; (4) no retention or yield factors were used to correct weight changes during cooking because this was already taken into account in the FFQ; (5) the polyphenol content in 100 g of each food item was obtained from the Phenol-Explorer database (version 3.6) [15]; (6) finally, the individual polyphenol intake from each food was calculated by multiplying the content of each polyphenol by the daily consumption of each food. Total polyphenol intake was calculated as the sum of all individual polyphenol intakes from the food sources reported in the FFQ.
The data used to calculate polyphenol intake was obtained by chromatography of all the phenolic compounds, except proanthocyanidins, the content of which was obtained by normal-phase high-performance liquid chromatography. In the case of lignans and phenolic acids in certain foods (i.e., swiss chard, chickpeas, plums, and strawberry jam), data corresponding to chromatography after hydrolysis was also collected, since these treatments are needed to release phenolic compounds that could otherwise not be analyzed. Total and polyphenol subclass intakes were adjusted for energy intake (kcal/day) using the residual method [13].
### *2.4. Measurements and Outcome Assessment*
Data on age, sex, educational levels, anthropometric measurements, dietary habits and lifestyle were collected at baseline. Anthropometric measurements were measured according to the PREDIMED-Plus protocol. Weight was recorded with participants in light clothing without shoes or accessories using a high-quality calibrated scale. Height was measured with a wall-mounted stadiometer. Waist circumference was measured midway between the lowest rib and the iliac crest. The BMI was calculated as weight (kg) divided by the square of height (m2).
Physical activity and sedentary behaviors were evaluated using the validated Regicor Short Physical Activity Questionnaire [16] and the validated Spanish version of the Nurses' Health Study questionnaire [17], respectively.
Information related to sociodemographic and lifestyle habits, individual and family medical history, smoking status, medical conditions, and medication use was evaluated using self-reported questionnaires. Sociodemographic and lifestyle variables were categorized as follows: age (three categories: <65, 65–70, or >70 years), educational level (three categories: primary, secondary, or high school), physical activity level (three categories: low, moderate, or high), BMI (three categories: 27.0–29.9, 30.0–34.9, or <sup>≥</sup>35 kg/m2), and smoking status (three categories: never, former, or current smoker).
Blood samples were collected after overnight fasting. Biochemical analyses were performed to determine plasma glucose (mg/dL), glycated hemoglobin (%), HDL-c (mg/dL), and triglyceride (mg/dL) levels using standard laboratory enzymatic methods. Low-density lipoprotein cholesterol (LDL-c; mg/dL) was calculated using the Friedewald formula whenever triglyceride levels were less than 300 mg/dL. Blood pressure measurements were obtained after the participant had rested for five minutes. Each measurement was obtained with a validated semiautomatic oscillometer (Omron HEM-705CP), ensuring the use of the proper cuff size for each participant.
### *2.5. Statistical Analysis*
Descriptive statistics were used to define the baseline characteristics of the participants. The database used was the PREDIMED-Plus baseline database generated in September 2018. Continuous variables are expressed as mean ± SD. Categorical variables are expressed as number (*n*) and percentage (%). Comparisons among quartiles of dietary polyphenol intake used the Pearson chi square test (χ2) for categorical variables or one-way ANOVA for continuous variables. The associations between dietary polyphenol intake and MetS components were analyzed by linear regression models to determine differences between quartiles of polyphenol subclass intake. The results of the regression models are expressed as unstandardized β-coefficients. For regression models, polyphenol and polyphenol subclasses are expressed as quartiles of energy-adjusted dietary intake. We used robust variance estimators to account for intra-cluster correlation in all linear models, considering members of the same household as a cluster. All regression models were adjusted for potential confounders. Model 1 was adjusted for sex, age, recruiting center, and members of the same household. Model 2 was additionally adjusted for physical activity level, BMI (except for waist circumference criteria), smoking status, and educational level. We additionally adjusted for anti-diabetic treatment when assessing glycemia and antihypertensive treatments when assessing blood pressure. Lastly, model 3 was additionally adjusted for total energy intake (continuous, kcal/day), saturated fatty acids (g/day), and distilled drinks alcohol intake (g/day). In model 3, the analysis of glycemia was additionally adjusted for dietary simple sugar intake (g/day), whereas the analysis of systolic and diastolic blood pressures was also adjusted for dietary sodium intake (mg/day). The normality of the continuous outcomes and standardized residuals was assessed with the Shapiro–Wilk test. Values are shown as 95% confidence interval (CI) and significance for all statistical tests was based on bilateral contrast set at *p* < 0.05. The P value for linear trends was computed by fitting a continuous variable that assigned the median value for each quartile in regression models. The descriptive analyses shown in Tables 1–3 were performed using SPSS software version 22.0 (Chicago, IL, USA) and the regression analysis was performed using Stata software version 16 (StataCorp LP, College Station, TX, USA).
1
Cardiovascular
expressed as mean (<sup>±</sup> SD). Categorical variables are expressed as number (*n*) and percentage (%). Comparisons
test for categorical variables or one-way ANOVA for continuous variables. For glycated hemoglobine
was computed by fitting a continuous variable that assigned the median value for each quartile in regression models.
diseases (CVD), body mass index (BMI), high-density
lipoprotein-cholesterol
(HDL-c) and nonsteroidal
among quartiles of dietary polyphenol intake with Pearson's chi square
parameter, 9% of participants had no values available. The P value for linear trend
anti-inflammatory
drugs (NSAIDs). Continue variables are
### **3. Results**
The present study was conducted on 6633 participants from the PREDIMED-Plus study. The mean age was 65.0 <sup>±</sup> 4.9 years, and mean BMI was 32.5 <sup>±</sup> 3.44 kg/m2. Table <sup>1</sup> shows the main characteristics of the participants according to quartiles of dietary total polyphenol intake. We observed that participants included in the highest quartile of polyphenol intake (>1019.3 mg/day) were mainly men and former smokers with a higher educational level (all three *p* < 0.001). We observed an inverse trend in the relationship between polyphenol intake and BMI (*p* = 0.02), whereas this trend was direct for waist circumference (*p* = 0.01) and body weight (*p* < 0.001). Moreover, fewer participants with insulin and nonsteroidal anti-inflammatory drug treatment were observed in the highest quartile of polyphenol intake (both *p* = 0.01).
Total polyphenol intake was 846 ± 318 mg/day, of which 58.0% were flavonoids (491 ± 253 mg/day), 33.1% phenolic acids (280 ± 131 mg/day), and the rest other polyphenols, stilbenes, and lignans (70.8 ± 41.5, 2.13 ± 3.92, and 1.53 ± 0.56 mg/day, respectively). The mean of the total polyphenol aglycone intake was 620.9 ± 273.5 mg/day. Table 2 shows the contribution (%) of each polyphenol subclass and polyphenol aglycones. The highest contributor to total polyphenol intake was hydroxycinnamic acids (30.9%). Regarding flavonoids, flavanols were the main contributors (24.1% from proanthocyanidins, 3.32% catechins, and 0.08% of theaflavins), followed by flavanones (9.83%), flavones (8.65%), flavonols (6.40%), and anthocyanins (5.14%). Additionally, tyrosols represented 6.19% of the total polyphenol intake, being the most abundant polyphenol classified within the group of other polyphenols.
The main food sources for each polyphenol subclass are also shown in Table 2. In the case of flavonoids, the most important contributors to the intake of proanthocyanidins were fruits and chocolate and its derivatives. Fruits (mainly oranges and orange juice) were the greatest contributors of flavanones, while vegetables (mainly onion, spinach, and lettuce) were the greatest contributors of flavones. Red wine, olives, tea, and wholegrain cereals were also important contributors to the remaining subclasses. Coffee was the most significant contributor of phenolic acids, especially of hydroxycinnamic acids, followed by olives and red wine. Stilbenes were mainly provided by red wine (91.9%). Lignans were widely distributed among foods, with extra virgin olive oil, fruits, and vegetables the main contributors. The main contributors of other polyphenols were olives, olive oil, cereals, coffee, and alcoholic beverages (mainly beer and red wine).
Table 3 shows the energy-adjusted intake of total polyphenols and the main subclasses by sex, age, BMI, level of physical activity, educational level, and smoking status. Total polyphenol intake was significantly higher in women due to their high intake of flavonoids (*p* < 0.001), whereas men consumed more phenolic acids (*p* = 0.003), stilbenes, and other polyphenols. The intake of total polyphenols, flavonoids, and lignans increased with age (*p* = 0.002, *p* < 0.001, and *p* = 0.006, respectively). Interestingly, participants with the highest BMI (>35 kg/m2) showed the lowest total polyphenol (*p* = 0.042), flavonoid (*p* = 0.004), and stilbene intake (*p* < 0.001), whereas phenolic acid intake was significantly higher in this group (*p* = 0.006). The level of physical activity was directly associated with total polyphenol intake (*p* < 0.001) and with all polyphenol classes except for phenolic acids (*p* < 0.001 in all cases except *p* = 0.03 for other polyphenols). Participants with a higher educational level (high school) showed higher total polyphenol, flavonoid, and stilbene intake (*p* < 0.001 in all cases). Current smokers reported a significantly higher intake of coffee than non-smokers (*p* < 0.001) and, consequently, showed a significantly higher intake of phenolic acids (*p* < 0.001). Otherwise, the smokers group showed significantly lower intake of flavonoids and lignans than their counterparts (*p* < 0.001, both).
The associations between dietary polyphenol intake and MetS components after full adjustment are shown in Figure 2. High flavonoid and low phenolic acid intake were associated with lower waist circumference (*p* = 0.02 and *p* < 0.001, respectively). The highest intake of other polyphenols was significantly and inversely associated with systolic (*p* = 0.001) and diastolic blood pressure levels (*p* = 0.002). An inverse association was found between fasting plasma glucose levels and lignans (*p* = 0.04). Positive associations were found between HDL-c levels and all polyphenol classes except for phenolic acid and lignan intake. Lastly, triglyceride concentration was inversely associated with lignans and stilbenes (*p* = 0.006 and *p* = 0.004, respectively). Changes in the linear regression models after adjustment are shown in the Supplementary Table (Supplementary Table S1).
**Figure 2.** Energy-adjusted subclasses of dietary polyphenol intake by metabolic syndrome components (standardized β-coefficients [95% Confidence Intervals]).
### **4. Discussion**
In this cross-sectional study of the PREDIMED-Plus study, we showed that high intake of some polyphenol subclasses was inversely associated with levels of the MetS components. These associations were especially observed for the subclasses whose contribution to total polyphenol intake was lower, such as other polyphenols, lignans, and stilbenes. Previous epidemiological studies have investigated the association between dietary polyphenol intake and MetS components in healthy populations or those at high risk of CVD, but to our knowledge there are no previous studies on these associations in subjects previously diagnosed with MetS.
In our study, the polyphenol intake was 846 ± 318 mg/day, and the intake was highest for flavonoids (58% of total), followed by phenolic acids (33.1%), similar to results of other Spanish cohorts [14,18]. By contrast, the total polyphenol intake was considerably lower than the intake observed in Mediterranean countries of the EPIC Study (1011 mg/day) [19], the SU.VI.MAX cohort study (1193 mg/day) [20], and the data from other studies conducted in non-Mediterranean countries, such as the UK National Diet and Nutrition Survey Rolling Programme for participants of similar age (1053 mg/day) [21]. The main noteworthy difference between our results and those of other countries was the relevant contribution of seeds, olives and olive oil, and red wine [14,20], while coffee, tea, and cocoa products are the foods most commonly observed in non-Mediterranean countries [22–24].
In addition to the differences observed according to geographical location and dietary habits, sociodemographic and lifestyle habits significantly influence the quantity and profile of intake of polyphenol subclasses. The intake of total polyphenols, particularly flavonoids and lignans, increased with age compared to younger participants (<65 years), although Grosso et al. reported the opposite observation [23]. In addition, BMI was inversely associated with total polyphenol intake, mainly with lower flavonoid and stilbene intake. This finding was also reported in the TOSCA.IT and EPIC studies [19,25].
The intake of polyphenol subclasses has been reported to have an impact on MetS components [26,27]. Even though flavonoids were the principal contributors of total polyphenol intake in our study, no associations were found with any of the MetS components, except for an inverse association with waist circumference. Similar findings were described in the HELENA study [28], where flavonoid intake was associated with lower BMI. Research on the mechanisms of action involved in the anti-obesogenic properties of flavonoids suggests that the improvements in glucose homeostasis are promoted by reducing insulin resistance and decreasing oxidative stress levels [29]. Phenolic acid intake was associated with higher fasting plasma glucose levels and waist circumference. These results are opposite from those observed in the HAPIEE cohort study, which described the beneficial effects of phenolic acid on the overall risk of developing MetS and lowering blood pressure [30]. Nevertheless, it must be taken into account that the dietary intake of phenolic acids and total polyphenol in the mentioned study doubled the amount estimated in our results, probably because of the higher intake of tea and its contribution to phenolic acid intake compared to our study population [23]. In Mediterranean countries, dietary intake of stilbenes is relatively high compared to other countries [19], with red wine being their main source (>90%). In this setting, higher stilbene intake was associated with higher HDL-c levels, but since HDL-c is the best-established cardiovascular protective factor by alcohol consumption, we cannot exclude that the alcohol content of red wine may interfere with this result [31]. In the PREDIMED study, the intake of red wine was associated with improvements in four out of five MetS criteria (i.e., elevated abdominal obesity, low HDL-c levels, high blood pressure, and high fasting plasma glucose levels) [32]. Other studies also found an inverse association between abdominal adiposity and stilbene intake, BMI, and waist circumference [30,33]. As a mechanistic pathway for stilbenes, resveratrol has shown potential anti-obesogenic effects decreasing adipocyte proliferation while activating lipolysis and β-oxidation [34]. However, the association with lower body weight and waist circumference observed in the present study and the promising effects against obesity associated with polyphenol intake observed in other studies were not clinically relevant [35]. Our results showed an inverse association between fasting glucose and lignans, and
an increase in HDL-c levels and lower levels of systolic and diastolic blood pressure measurements for other polyphenols. The same finding was described in a Brazilian cohort for hypertension and other polyphenols [36]. In contrast with our results, flavonoids, mainly anthocyanins, showed greater antihypertensive effects in another study [37]. Finally, the association between lignan intake and fasting glucose levels was not demonstrated to be linked with the diagnosis of type 2 diabetes (T2D) in the EPIC study [38], but this inverse association aligned with the results observed in the PREDIMED cohort and PREDIMED-Plus study [39,40]. The potential mechanism of action underlying this association might be explained by the improvements observed in gut microbiota. This assumption was also observed in a study of US women [41], showing an inverse association between levels of gut microbiota metabolites from dietary lignan intake and T2D incidence.
Interestingly, in our study we found an association between intake of all polyphenol subclasses except phenolic acids and lignans and higher HDL-c levels. These associations were also found with total polyphenol intake in the TOSCA.IT study in T2D subjects [25] and in a similar cohort of participants at high cardiovascular risk [42]. We also observed that triglyceride levels were inversely associated with stilbene and lignan intake. Despite the fact that the antioxidant properties of polyphenols for the prevention of LDL-c oxidation are well described, the effects of dietary polyphenols on the reduction of total cholesterol levels or triglycerides are controverted [43].
The major strengths of the present study are its large sample size, the multicenter design, and the use of the Phenol-Explorer as the most comprehensive food composition database on dietary polyphenols [15]. In prior studies, the FFQ was validated to evaluate total polyphenol intake in both clinical and cross-sectional studies [44]. Our study has also some limitations. First, it used a cross-sectional design which does not allow attributing conclusions to plausible causes. In order to establish causality, a randomized controlled trial based on the intake of different polyphenol subclasses should be performed. Second, potential residual confounding and the lack of generalizability of the results to other populations than middle-aged to elderly people with higher BMI and MetS are limitations. Third, the use of the FFQ may have led to a misclassification of the exposure due to self-reported information of food intake and to the fact that some polyphenol-rich foods are grouped in the same item (e.g., spices). Nevertheless, the FFQ used has been validated in the adult Spanish population and showed good reproducibility and validity [45]. Fourth, other factors that affect food polyphenol content, such as bioavailability, variety, ripeness, culinary technique, storage, region, and environmental conditions, were not collected.
Even though recent research postulates that polyphenols are effective in improving MetS, no single phenolic compound or food has an impact on all the MetS components, suggesting that healthy and polyphenol-rich dietary patterns such as the MedDiet may be an adequate strategy for MetS management. This research might be useful for setting dietary and health counseling for MetS patients, especially those with low HDL-c levels. The use of a consensus methodology and polyphenol database might facilitate this in future studies. Future large-scale clinical trials are needed to clarify the underlying mechanisms of action and establish safe doses for the potential health effects described.
### **5. Conclusions**
This study provides detailed information about the relationship between polyphenol intake and the components of MetS in a population of overweight or obese adults. Higher intake of all the subclasses of polyphenols was associated with a better profile of the components of MetS, especially with HDL-c levels.
**Supplementary Materials:** The following are available online at http://www.mdpi.com/2072-6643/12/3/689/s1, Table S1: Energy-adjusted sub-classes of dietary polyphenol intake by metabolic syndrome criterias.
**Author Contributions:** Conceptualization, J.S.-S., D.C., M.F., D.R., J.V. (Jesús Vioque), J.W., J.A.M., L.S.-M., F.J.T., J.L., X.P., J.A.T., L.G.-M., M.D.-R., P.M.-M., L.D., J.V. (Josep Vidal), C.V., and R.E.; methodology, J.S-S, D.C., M.F., D.R., J.V. (Jesús Vioque), J.W., J.A.M., L.S.-M.,F.J.T., J.L., X.P., J.A.T., L.G.-M., M.D.-R., P.M.-M., L.D., J.V. (Josep Vidal), C.V., and R.E.; validation, R.E.; formal analysis, S.C.-B. and A.T.-R.; investigation, J.S.-S.; funding acquisition, J.S-S, D.C., M.F., D.R., J.V. (Jesús Vioque), J.W., J.A.M., L.S.-M., F.J.T., J.L., X.P., J.A.T., L.G.-M., M.D.-R., P.M.-M., L.D., J.V. (Josep Vidal), C.V., and R.E.; data curation, S.C.-B., A.T.-R. and F.V.-S.; writing—original draft preparation, S.C.-B. and A.T.-R.; writing—review and editing, F.V.-S., M.D., J.S.-S., V.M.-S., M.R.-G., P.B.-C., D.C., M.F., D.R., J.V. (Jesús Vioque), Á.M.A.-G., J.W., J.A.M., L.S.-M., F.J.T., J.L., X.P., J.A.T., A.G.-R., L.G.-M., M.D.-R., P.M.-M., L.D., J.V.(Josep Vidal), C.V., M.C., A.R.-N., N.B.-T., R.B., O.C., J.K., S.G.-P., C.S.-S., J.P.-L., M.A.Z., I.B.-C., R.C., A.M.G.-P., J.M.S.-L., M.Á.R.-S., A.J., N.M.-C., P.H.-A., J.V.S., A.S., A.M.G.-P., E.C.-P., L.G.-G., M.R.-C., N.B., Á.H., R.M.L.-R. and R.E.; visualization, S.C.-B. and R.E.; supervision, R.E.; project administration, J.S.-S. All authors have read and agreed to the published version of the manuscript.
**Funding:** The PREDIMED-Plus trial was supported by official Spanish institutions for funding scientific biomedical research, CIBER Fisiopatología de la Obesidad y Nutrición (CIBERobn) and Instituto de Salud Carlos III (ISCIII), through the Fondo de Investigación para la Salud (FIS), which is co-funded by the European Regional Development Fund (four coordinated FIS projects led by J.S.-S. and J.Vi., including the following projects: PI13/00673, PI13/00492, PI13/00272, PI13/01123, PI13/00462, PI13/00233, PI13/02184, PI13/00728, PI13/01090, PI13/01056, PI14/01722, PI14/00636, PI14/00618, PI14/00696, PI14/01206, PI14/01919, PI14/00853, PI14/01374, PI14/00972, PI14/00728, PI14/01471, PI16/00473, PI16/00662, PI16/01873, PI16/01094, PI16/00501, PI16/00533, PI16/00381, PI16/00366, PI16/01522, PI16/01120, PI17/00764, PI17/01183, PI17/00855, PI17/01347, PI17/00525, PI17/01827, PI17/00532, PI17/00215, PI17/01441, PI17/00508, PI17/01732, and PI17/00926), the Special Action Project entitled: Implementación y evaluación de una intervención intensiva sobre la actividad física Cohorte PREDIMED-Plus grant to J.S.-S., the Recercaixa grant to J.S.-S. (2013ACUP00194), a grant from the Fundació la Marató de TV3 (PI044003), grants from the Consejería de Salud de la Junta de Andalucía (PI0458/2013, PS0358/2016, and PI0137/2018),grants from the Generalitat Valenciana (PROMETEO/2017/017, APOSTD/2019/136), a SEMERGEN grant, a CICYT grant provided by the Ministerio de Ciencia, Innovación y Universidades (AGL2016-75329-R) and funds from the European Regional Development Fund (CB06/03). The Spanish Ministry of Science Innovation and Universities for the Formación de Profesorado Universitario (FPU17/00785) contract. Food companies Hojiblanca (Lucena, Spain) and Patrimonio Comunal Olivarero (Madrid, Spain) donated extra virgin olive oil, and the Almond Board of California (Modesto, CA), American Pistachio Growers (Fresno, CA), and Paramount Farms (Wonderful Company, LLC, Los Angeles, CA) donated nuts. This call is co-financed at 50% with charge to the Operational Program FSE 2014-2020 of the Balearic Islands.
**Acknowledgments:** We thank all the volunteers for their participation and medical professionals for their contribution to the PREDIMED-Plus trial. CIBEROBN, CIBERESP, and CIBERDEM are initiatives of the Instituto de Salud Carlos III (ISCIII), Madrid, Spain. A.T.R. and P.H.A. thanks the Ministry of Science Innovation and Universities for the Juan de la Cierva-formación contract. J.K. is grateful to the Fundación Instituto de Investigación Sanitaria Illes Balears (call financed by 2017annual plan of the sustainable tourism tax and at 50% with charge to the ESF Operational Program 2014–2020 of the Balearic Islands) for the postdoctoral contract for the 'FOLIUM' programme within the FUTURMed.
**Conflicts of Interest:** R.E. reported receiving grants from Instituto de Salud Carlos III and olive oil for the trial from Fundacion Patrimonio Comunal Olivarero during the conduct of the study and personal fees from Brewers of Europe, Fundación Cerveza y Salud, Interprofesional del Aceite de Oliva, Instituto Cervantes, Instituto Cervantes, Pernaud Richar, Fundación Dieta Mediterránea, Wine and Culinary International Forum; nonfinancial support from Sociedad Española de Nutrición and Fundación Bosch y Gimpera; and grants from Uriach Laboratories outside the submitted work.. R.M.L.-R. reports personal fees from Cerveceros de España, personal fees and other from Adventia, other from Ecoveritas, S.A., outside the submitted work. The rest of authors have no conflict of interest. None of the funding sources took part in the design, collection, analysis, or interpretation of the data or in the decision to submit the manuscript for publication.
### **References**
© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
| doab | 2025-04-07T03:56:58.614815 | 1-5-2021 17:49 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "003cf560-3166-422f-9a1c-98b1e971efa6",
"url": "https://mdpi.com/books/pdfview/book/3241",
"author": "",
"title": "Dietary Polyphenols and Human Health",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783039437931",
"section_idx": 12
} |
003cf560-3166-422f-9a1c-98b1e971efa6.13 | *Review* **E**ff**ects of Dietary Phytoestrogens on Hormones throughout a Human Lifespan: A Review**
### **Inés Domínguez-López 1, Maria Yago-Aragón 1, Albert Salas-Huetos 2, Anna Tresserra-Rimbau 1,3,4,5,\* and Sara Hurtado-Barroso 1,3**
Received: 7 July 2020; Accepted: 12 August 2020; Published: 15 August 2020
**Abstract:** Dietary phytoestrogens are bioactive compounds with estrogenic activity. With the growing popularity of plant-based diets, the intake of phytoestrogen-rich legumes (especially soy) and legume-derived foods has increased. Evidence from preclinical studies suggests these compounds may have an effect on hormones and health, although the results of human trials are unclear. The effects of dietary phytoestrogens depend on the exposure (phytoestrogen type, matrix, concentration, and bioavailability), ethnicity, hormone levels (related to age, sex, and physiological condition), and health status of the consumer. In this review, we have summarized the results of human studies on dietary phytoestrogens with the aim of assessing the possible hormone-dependent outcomes and health effects of their consumption throughout a lifespan, focusing on pregnancy, childhood, adulthood, and the premenopausal and postmenopausal stages. In pregnant women, an improvement of insulin metabolism has been reported in only one study. Sex hormone alterations have been found in the late stages of childhood, and goitrogenic effects in children with hypothyroidism. In premenopausal and postmenopausal women, the reported impacts on hormones are inconsistent, although beneficial goitrogenic effects and improved glycemic control and cardiovascular risk markers have been described in postmenopausal individuals. In adult men, different authors report goitrogenic effects and a reduction of insulin in non-alcoholic fatty liver patients. Further carefully designed studies are warranted to better elucidate the impact of phytoestrogen consumption on the endocrine system at different life stages.
**Keywords:** isoflavones; soy; dietary flavonoids; lignans; flaxseeds; endocrine; stages of life; estrogenic; polyphenols; health
### **1. Introduction**
Phytoestrogens are polyphenolic molecules with a structural similarity to endogenous human hormones, hence their estrogenic activity. The main dietary source of these plant secondary metabolites is legumes (particularly soy), and to a lesser extent fruits, vegetables, and cereals [1]. Figure 1 shows the most common phytoestrogens in diet. According to their origin, lignins are classified into plant lignans (e.g., pinoresinol, secoisolariciresinol, matairesinol, and sesamin) and enterolignans (e.g., enterodiol and enterolactone), which are metabolized from plant lignans by intestinal bacteria [1]. Although ingested in lower quantities than isoflavones and lignans, prenylflavonoids from beer and coumestans from soy are also regarded as polyphenols with estrogenic activity.
**Figure 1.** Classification and examples of the most common dietary phytoestrogens. Images are the chemical structures of genistein, coumestrol, and enterodiol.
The intake of phytoestrogens has increased due to the widespread use of soy products for human consumption and as cattle food [2]. In Europe, the lowest average intake of phytoestrogens occurs in Mediterranean countries, whereas consumption in Northern countries is 0.76 mg/day [3]. The highest soy-derived isoflavone intakes worldwide are still in China and Japan, where the population consumes an average of 15–50 mg per day, compared to only about 2 mg per day in Western countries [4,5]. The promising health effects of soy have driven some people in developed countries to consume it as an alternative to meat or dairy products.
Dietary phytoestrogens are digested in the small intestine, where they are poorly absorbed. Those that reach the liver are conjugated and circulate in the plasma until excretion in urine. Thos that are not absorbed are metabolized by the gut microbiota into lower weight compounds [1]. The diversity of food matrices (from pure compounds to complex foods) used in clinical studies could also lead to different results although interindividual variability seems more determinant. It has been demonstrated that phytoestrogen extraction from complex food matrices, such as those with high content of sugars and proteins, is more difficult in in vitro studies; however, no clear differences regarding food matrices were observed in humans [6]. Nevertheless, results using pure compounds must be extrapolated carefully because not only is the matrix different, but also the concentration, which is higher in pure extracts.
Results from human studies suggest that phytoestrogens may lower the risk of osteoporosis, some cardiometabolic diseases, cognitive dysfunction, breast and prostate cancer, and menopausal symptoms bymodulating the endocrine system (Figure 2). However, some authors describe phytoestrogens as endocrine disruptors and believe their beneficial effects have been overestimated [2,5,7]. This ambiguity could be partially due to the variability of published studies, as the beneficial or harmful effects of phytoestrogens depend on the exposure (type, amount consumed, and bioavailability), ethnicity, hormonal status (age and sex and physiological condition), and health status of the consumer [2,5,7].
**Figure 2.** Summary of potential health outcomes of phytoestrogens through the modulation of the endocrine system in (**a**) thyroids, (**b**) liver, (**c**) ovaries, (**d**) bones, (**e**) hypothalamic–pituitary–gonadal axis, (**f**) pancreas, (**g**) fat tissue, (**h**) prostate. FSH: follicle-stimulating hormone; GnRH: gonadotropinreleasing hormone; IGF-1: insulin growth factor 1; LH: luteinizing hormone; OC: osteocalcin; PE: phytoestrogens; PSA: prostate-specific antigen; SHBG: stimulating hormone-binding globulin; T3: triiodothyronine; T4: thyroxine; TPO: thyroperoxidase.
A plausible mechanism of action for phytoestrogens is estrogen receptor (ER) binding. The effects of isoflavones, which have a five-fold greater affinity for β-ER than α-ER [8], on the endocrine system may be through modulation of the hypothalamic-pituitary axis [9]. However, not all the biological effects of phytoestrogens involve estrogen receptors. They can also activate serotonergic and insulin-like growth factor (IGF) receptors 1, induce free radical binding and modify tyrosine kinases, cycle adenosine monophosphate (cAMP), phosphatidylinositol-3 kinase (PI3K)/Akt, mitogen-activated protein (MAP) kinases, transcription of nuclear factor-kappa β (NF-Kβ), as well as promote DNA methylation and affect histone and RNA expression. In addition, phytoestrogens can act as intracellular regulators of the cell cycle and apoptosis. Thus, due to their antioxidant, antiproliferative, antimutagenic, and antiangiogenic roles, phytoestrogens can improve health [10]. In addition, some authors observed that estrogen and androgen seem to be involved in breast and prostate cancer regulating proliferative and migratory signaling, such as Src/PI3K. Hormonal therapy response may vary depend on interactions between estrogen or androgen receptors and proteins, according to hormone levels [11–13].
This integrative review aims to synthesize the results obtained by human studies and assess the potential hormone-related health effects of dietary phytoestrogens throughout the human lifespan.
### **2. E**ff**ects of Phytoestrogen Intake on Sex Hormones**
The anti-estrogenic activity of phytoestrogens is due to their structural similarity with 17-βestradiol (E2), the main female sex hormone [5]. As well as interacting with ERs, phytoestrogens can affect the secretion of gonadotropin-releasing hormone (GnRH) [14]. Phytoestrogens could disrupt the endocrine system by interfering with the hypothalamic–pituitary–gonadal axis, which controls estrogen secretion. The hypothalamus releases GnRH and stimulates the pituitary to produce follicle-stimulating hormone (FSH) and luteinizing hormone (LH), gonadotropins that promote the secretion of estrogen, progesterone, and testosterone by the ovaries or testes. Low estrogen levels are a signal for the hypothalamus to release GnRH, whereas high levels provide a negative feedback [15]. Therefore, the presence of exogenous compounds structurally similar to E2 may interfere with this system.
Some studies have focused on how phytoestrogen affects urinary estrogen metabolites, some of which may be involved in the development of breast cancer. In particular, the ratio of 2-hydroxyestrone (2-OH-E1) to 16α-hydroxyestrone (16α-OH-E1) (2:16α-OH-E1) is considered a useful biomarker of estrogen-related cancer risk. A major 2:16α-OH-E1 ratio is related to lower risk of breast cancer. Previously, it was observed that a higher concentration of 16α-OH-E1 was associated with breast and endometrial cancer, while an increase of metabolite 2-OH-E1 seems to inhibit the carcinogenesis [16].
Phytoestrogens have also been reported to affect sex hormones through ER-independent mechanisms of action, such as by altering hormone-binding globulin (SHBG) levels. Circulating estrogens and androgens are mostly bound to albumin and SHBG, with only a small fraction remaining free. As estrogens and androgens are only biologically active in their free form, SHBG affects steroidal activity. In vitro studies have shown that isoflavonoids stimulate the synthesis of SHBG by liver cancer cells [17], but available data from human studies are inconclusive [18,19]. In addition, phytoestrogens inhibit aromatase and other enzymes involved in the synthesis of steroid hormones [20].
Preclinical studies have suggested that phytoestrogens influence sexual function and the incidence of cancer associated with the reproductive system such as ovarian and breast cancer [21], but the results of cross-sectional studies and clinical trials are conflicting [22,23]. In addition to factors such as dose, type, and bioavailability, the effects of phytoestrogens on sexual function could also depend on the life stage of the consumer, as explained below.
### *2.1. Pregnancy*
The results of a longitudinal study that measured E2, estriol (E3), testosterone, and isoflavones in urine and serum from 194 pregnant women weakly support the initial hypothesis that genistein and daidzein would reduce levels of E2 and testosterone at the 10th week of gestation. Additionally, sex hormones quantified in umbilical cord serum were not related to isoflavones (genistein, daidzein, and equol) measured at delivery [24].
### *2.2. Children*
Dietary phytoestrogens seem to be transferred from maternal blood to the fetus, but there is no evidence that they alter sex hormones in infants [25–27]. Although isoflavone bioavailability in this sensitive period may be higher than in adults [28], no estrogenic effects were observed in infants fed with a soy formula [29–31]. Nevertheless, a cross-sectional study carried out in children aged 3–6 years reported an increase of androgens in girls and a decrease of estrogens in boys consuming higher amounts of soy and isoflavones [25]. In a crossover trial conducted in girls aged 8–14 years, the consumption of a high-soy diet for 8 weeks significantly increased dehydroepiandrosterone (DHEA) concentrations but not other sexual hormones. Although the level of all sex hormone metabolites excreted was very low, positive correlations with the intervention were found, being higher for total androgens than for estrogens and pregnanediol [32].
### *2.3. Men*
A cross-sectional study in randomly selected Japanese men found a negative association between soy product consumption and E2 serum concentrations, but no link was observed with peripheral concentrations of androgen hormones [33]. In a randomized clinical study in Japanese healthy male volunteers consuming 60 mg per day of soy isoflavones, no changes in serum levels of E2 and total testosterone were observed compared to the baseline at the end of the 3-month intervention. However, serum levels of SHBG increased and free testosterone and dihydrotestosterone decreased [34].
There is weak evidence that phytoestrogens contribute to reducing the risk of prostate cancer (PCa). Several observational studies have found a negative association between the consumption of phytoestrogens (soy and its isoflavones) and the levels of prostate-specific antigen (PSA) in blood [35]. PSA, a protein produced by the prostate gland, is used as a marker to detect PCa, although its levels also increase with benign prostate hypertrophy. In a randomized controlled trial, a reduction in PSA levels was observed in men with PCa after consuming soy isoflavones for a mean of 23 days [36]. However, longer studies (minimum 3 weeks and maximum 12 months) did not find beneficial effects on PSA levels after a soy isoflavone intervention [37–40], nor were changes in PSA plasma levels observed in PCa patients that consumed rye bran bread for 3 weeks [41].
There is a relationship between sex hormones and the pathogenesis of PCa. High levels of androgens, which promote prostate cell growth, may contribute to the risk of PCa in some men. Some epidemiological studies suggest that low levels of testosterone are associated with a lower risk of PCa [42,43]. A meta-analysis of 32 studies published in 2010 by Hamilton-Reeves et al. reported that the consumption of isoflavones had no significant effect on circulating testosterone or free testosterone levels in men [44], in agreement with other clinical trials evaluating phytoestrogen intake in PCa patients [37,39–41]. Nor have effects on dihydrotestosterone been described [36,40]. However, in an open-labeled, non-randomized clinical trial of men with higher levels of PSA, free testosterone was depleted after 12 months of daily consumption of 141 mg of isoflavones in soy milk [38].
Recent studies have pointed to a protective role of estrogens in PCa development and progression, alone or in synergy with androgens [45]. Several studies have focused on the beneficial effect of soy isoflavones, specifically genistein and daidzein, as these components can act as weak estrogens. A 6-month randomized controlled study evaluating the effects of isoflavone on men at high risk of developing advanced PCa found an increase in concentrations of the estrogen hormones estrone (E1), E2, 2-hydroxi-estradiol (2-OH-E2), and 16α-OH-E1. An increase in the 2:16α-OH-E1 ratio was also reported, which is related to a reduced risk of estrogen-mediated cancer. No differences were observed for 2-methoxyestradiol (2-ME2), 1-methoxyestrone (2-ME1), E3 and 2-OH-E1 [46]. Conversely, Bylund A. et al. reported that levels of E2, FSH, and LH in PCa patients remained unaltered after a 3-week rye bran bread intervention [41].
### *2.4. Premenopausal Women*
In agreement with the potential anti-estrogenic effect of phytoestrogens, some authors have observed a significant decrease in estrogen levels after the consumption of soy products [18,22,47–49]. In a randomized controlled cross-over trial conducted in 12 healthy premenopausal women, those consuming a high-soy diet for three menstrual cycles had lower urinary concentrations of total estrogens (E1, E2, E3), and some metabolites compared to individuals on a low-soy diet [22]. Although significant correlations were obtained between serum levels of unconjugated estrogens, and urinary conjugated and unconjugated estrogen metabolites, the large intra-subject variability in urinary estrogen levels limits its use as a biomarker [50]. Similarly, a fall in the circulating levels of E2 after the consumption of a soy-rich diet has been reported [47–49]. In a cross-sectional study, Kapiszewska M. et al. found an association between low salivary E2 concentrations and the intake of black tea (only or plus green tea), catechins, theaflavins, and epigallocatechin gallate (EGCG), being more pronounced in premenopausal women living in urban areas than in those living in rural areas [51]. As well as a decline in estrogens, a significant decrease in progesterone levels after phytoestrogen consumption has been observed [47–49,52]. Conversely, other clinical trials and observational studies do not report any modifications of sex hormones attributed to the consumption of dietary soy-isoflavones [14,19,23,53–58].
It has been proposed that changes in estrogen levels induced by dietary phytoestrogens could depend on the individual capacity to produce equol [59,60], mainly the S-equol enantiomer, due to its high affinity for β-ER [61]. Accordingly, Duncan et al. reported that premenopausal equol excretors had a lower risk of breast cancer compared to non-excretors [60].
In summary, it is still uncertain if a phytoestrogen-rich diet triggers an imbalance of estrogen and progesterone concentrations. In randomized controlled crossover trials, no significant changes were observed in the progesterone/E2 ratio in women who consumed a soy diet for two menstrual cycles [56], whereas the ratio increased after the intake of 10 g/day of flaxseeds for three menstrual cycles [62].
The status of sex hormones can also be an indicator of breast cancer risk. Although measuring estrogen and progesterone concentrations in nipple aspirate fluid may be better than using serum samples to detect dietary-associated changes in the breast, the correlation between dietary phytoestrogen and estrogens was poor in both matrices [58]. On the other hand, the scant evidence for an increase in the 2:16α-OH-E1 ratio after the consumption of soy and flaxseed suggests such high-phytoestrogen foods may have a protective role against breast cancer [22,48,63]. In the same vein, Xu et al. observed a decrease in the ratio of genotoxic metabolites (16α-OH-E1, 4-hydroxyestradiol (4-OH-E2), and 4-hydroxyestrone (4-OE-E1)) and total estrogens [22]. However, no effects on biomarkers related to breast cancer risk have been reported in other studies [23,56,59].
Inconclusive effects of phytoestrogen supplementation in the form of soy protein powder (low and high doses) on the concentrations of FSH and LH have been observed. Both hormones decreased after low- but not high-isoflavone diets [18], as well as after the consumption of soy products [52,64]. However, other studies found no significant changes in FSH and LH concentration after phytoestrogen supplementation [14,23,48,49,53–55].
Most authors have not found any effects of a phytoestrogen-rich diet on circulating levels of androgens [23,53,56,57,64,65]. Two clinical trials did report a decrease of DHEA-sulfate concentrations in healthy premenopausal women after 1 to 3 months on a diet high in soy and soy products [18,47]. In contrast, in a randomized controlled clinical trial (RCT) conducted in healthy premenopausal women consuming 10 g/day of flaxseeds for three menstrual cycles, an increase in serum levels of testosterone in the mid-follicular phase was observed [62]. Overall, there is no solid evidence supporting the influence of phytoestrogens on SHBG [19,23,49,52,53,55–57,62,64–66], although a weak increase has been described [18,67,68].
A prolonged menstruation after regular intake of phytoestrogens has been reported [62,68], but most studies indicate no significant changes in menstrual cycle length or concentration of prolactin [18,23,53,56,57,62,65,66].
### *2.5. Postmenopausal Women*
Menopausal transition is caused by the depletion of ovarian follicles and their responsiveness to the pituitary gonadotropins FSH and LH. This results in low serum levels of the ovarian hormones estrogen and progesterone, and also an increase in FSH concentrations due to the disruption in the negative feedback regulating the hypothalamic–pituitary–gonadal axis [69]. These hormonal changes are responsible for several menopausal symptoms, such as vasomotor symptoms, hot flushes, and vaginal dryness, as well as long-term disorders like osteoporosis, cardiovascular diseases, and breast cancer.
In postmenopausal women there is little evidence supporting the hypothesis that phytoestrogens affect sex hormone levels. Numerous studies have reported that phytoestrogens—including isoflavones, flavonoids, and lignans—do not affect estrogen or progesterone concentrations in postmenopausal women [7,14,19,70–95]. However, other clinical studies did find that isoflavone administration produced significant changes in E2 [96,97] or progesterone concentrations [98]. Two other postmenopausal studies suggested that flaxseed lignans may reduce E2 and E1 sulfate [99] in healthy women, and E1 concentrations in obese and overweight women [100]. Tormala R. et al. (2008) also reported lower E1 concentrations after soy protein supplementation in tibolone-using postmenopausal women who were equol producers [101]. In one epidemiological study, the relationship between isoflavone intake and peripheral E2 concentrations was assessed in postmenopausal women, and urinary excretion of daidzein, genistein, and glycitein and serum levels of daidzein and glycitein were associated with lower plasma E2 levels. Interestingly, these associations were stronger in 18 postmenopausal women with the CC genotype for ESR1 *Pvull* polymorphism, suggesting that genes influence diet effects [102].
Phytoestrogen effects on urinary estrogen metabolites have been studied in order to assess their potential protective role against breast cancer. One of the most recent studies reports an increase in the 2:16α-OH-E1 ratio after red clover-derived isoflavone supplementation in postmenopausal osteopenic women [103]. These results are consistent with prior research that found an increase in the 2:16α-OH-E1 ratio after flaxseed supplementation [104,105]. Other studies, however, did not find any difference in the ratio after phytoestrogen supplementation in postmenopausal women [81,95], and one even reports a lower ratio [106].
Other hormones affected by the disruption of the hypothalamic–pituitary–gonadal axis are the gonadotropins FSH and LH, which according to different clinical trials are not affected by phytoestrogen supplementation [14,79,82,83,107–109]. Only Crisafulli A et al. found lower gonadotropin levels after 54 mg/day of genistein supplementation for 6 months in 60 postmenopausal women compared to the control group [110].
Whereas some clinical studies found that isoflavone consumption increased SHBG levels in postmenopausal women [18,71,96,110], others concluded the opposite. Thus, Wu A.H. et al. (2012) and Uesugi S et al. report lower concentrations of SHBG in healthy postmenopausal women after supplementation with EGCG or isoflavones [91,97], but most of the studies found no association between SHBG levels and phytoestrogen intake [19,78,79,81,83,85,86,88,95,99,100,102,108,109,111–113]. Lastly, results from epidemiological studies support the hypothesis that some phytoestrogens may have a positive influence on SHBG. Monroe K.R. et al. showed that plasma enterolactone levels were associated with higher concentrations of SHBG in postmenopausal Latina women [114]. Low et al. also reported higher concentrations of SHBG in postmenopausal women with higher urinary excretion of lignans, but no association was found with the excretion of other phytoestrogens, such as isoflavones, equol, or O-desmethylangolesin (O-DMA) [115].
To date, most of the few human studies evaluating phytoestrogen effects on androgens in postmenopausal women have found no associations between phytoestrogen intake and androgen peripheral concentrations [74,82,85,86,91,109,116]. Yet Basaria S. et al. reported a decrease in testosterone levels after 12 weeks of isoflavone supplementation, results that were supported by Kapoor R. et al. only in normal-weight postmenopausal women consuming pomegranate for 3 weeks [111,113]. Bioavailable testosterone remained unchanged in both trials. Furthermore, Wu W.H. et al. also reported lower levels of DHEA-sulphate, an androgen precursor, in postmenopausal women after a 5-week intervention with sesame lignans [81].
### **3. E**ff**ect of Phytoestrogen Intake on Thyroid Hormones**
It is not clearly established if phytoestrogen consumption alters the hypothalamic–pituitary– thyroid axis and triggers goitrogenic effects in humans [9,117]. A randomized, double-blind, and crossover trial carried out in 60 patients with subclinical hypothyroidism and an adequate iodine intake reported an advance to overt hypothyroidism in 10% of cases (6 females) after administering soy protein with isoflavones for 8 weeks [118]. Nevertheless, soy isoflavones appear not to affect euthyroid populations with an optimal iodine status [9,117]. Controversial results have been obtained by studies in healthy humans, who did not experience anti-thyroid or any other effects after consuming dietary phytoestrogens, particularly isoflavones. The evidence for the impact of phytoestrogens on thyroid function according to the life stage is provided below.
### *3.1. Pregnant Women*
A cross-sectional study found no association between soy consumption in early pregnancy and the development of thyroid dysfunction or autoimmunity in 505 women living in areas with an optimal intake of iodine [119]. However, further studies in countries with iodine deficiency are needed.
### *3.2. Children*
A cross-sectional study in children aged 8–15 years in an iodine-deficient region of the Czech Republic did not obtain conclusive results. Although levels of free thyroxine (free-T4) increased after a higher intake of soy, a positive correlation was observed between serum daidzein levels and thyroid stimulating hormone (TSH) [120]. In a retrospective study, children with congenital hypothyroidism fed with soy formula had a higher concentration of TSH compared to those fed with non-soy formula [121]. Nevertheless, a study in 12 hypercholesterolemic children consuming toffee candies containing isoflavone extract for 8 weeks did not find any effect on TSH, triiodothyronine (T3), or T4 levels [122].
### *3.3. Men*
Two clinical trials showed that products with a high phytoestrogen content had anti-thyroid effects in male populations [123,124]. Hampl R. et al. observed a significant increase of serum TSH in young males after supplementation with 2 g/kg body weight/day of boiled unprocessed natural soybeans for 7 days [123]. Similarly, significant changes in TSH (increasing) and free-T4 (decreasing) were found in men with type-2 diabetes mellitus and compensated hypogonadism after consumption of 15 g/day of soy protein isoflavones for 3 months [124].
### *3.4. Premenopausal Women*
No significant changes in thyroid function were found in healthy and obese premenopausal women consuming soy isoflavones for periods of 1 week to 6 months [53,123,125], but reduced free-T3 levels were found in healthy young females following a high-soy diet for three menstrual cycles [126].
### *3.5. Postmenopausal Women*
A randomized, double-blind and parallel trial conducted in 120 postmenopausal women reported a significant increase in TSH and a decrease of free-T4 after the intake of 66 mg/day of soy isoflavones for 6 months [124]. Jayagopal V. et al. obtained similar results in a cross-over design trial involving 32 postmenopausal women with type-2 diabetes mellitus consuming twice the amount for half the duration [109]. Another parallel RCT conducted by Mittal et al. in oophorectomized women showed a decrease in free T3 after a 12-week intervention with 75 mg/day isoflavone [127]. Other authors found no significant differences in the level of thyroid hormones, but other thyroid-related parameters such as thyroxine-binding globulin (TBG) and the T3:T4 ratio were altered, indicating possible goitrogenic activity derived from phytoestrogen consumption [18,108]. In addition, Sosvorová et al. observed the presence of mono-iodinated derivatives of daidzein and genistein in urine after daily consumption of isoflavonoids for 3 months, which could explain the entry of genistein and daidzein in human thyroid follicles and thyroperoxidase modification [108]. However, other RCT observed no significant changes in thyroid function after isoflavone intake [84,128–130]. Nor was an effect reported in longer studies (1 to 3 years of duration) after the administration of different doses of isoflavones (2–200 mg/day) [74,93,131–133], possibly because adaptation to long-term changes in dietary isoflavone intake triggers endocrine autoregulation [95].
### **4. E**ff**ect of Phytoestrogen Intake on Cardiometabolic Risk-Related Hormones**
Cardiometabolic diseases encompass a set of dysfunctions affecting the cardiovascular system. These are not limited to hard cardiovascular events such as coronary heart disease, myocardial infarction, and stroke, but also include cardiovascular risk factors, namely obesity, insulin resistance, endothelial dysfunction, atherosclerosis, lipid profile, or non-alcoholic fatty liver disease, among others [134].
Obesity is usually the first risk factor that triggers chronic low-grade inflammation, which plays a crucial role in systemic metabolic dysfunction. Adipose tissue and adipocytes are dysfunctional in obese individuals, causing the secretion of pro-inflammatory adipokines that contribute to chronic inflammation and subsequently to the progression of cardiometabolic disorders like insulin resistance [135–137].
Another condition that increases the odds of suffering cardiovascular diseases is type-2 diabetes mellitus, which involves alterations in intestinal sensitivity to insulin and glucagon-like peptide-1 (GLP-1), as does its previous state, insulin resistance. GLP-1 has a variety of metabolic effects, including the glucose-dependent stimulation of insulin secretion, and is also involved in cardiovascular health [138]. Insulin, an anabolic hormone secreted in the pancreas, also regulates carbohydrate metabolism, participates in the storage of free fatty acids in adipose tissue, and enhances protein synthesis, increasing amino acid uptake by tissues [139].
Some authors have studied whether phytoestrogens can decrease the levels of pro-inflammatory adipokines such as leptin and resistin or increase adiponectin, an anti-inflammatory hormone, as well as regulate the secretion of insulin, glucagon, and ghrelin. Ghrelin, a recently discovered hormone related to cardiovascular health, is involved in feeding behavior, energy homeostasis, and carbohydrate metabolism. It therefore participates in body weight maintenance, which is crucial for vascular health [140].
### *4.1. Pregnancy*
Only one study has evaluated the relationship between phytoestrogens and cardiovascular health during pregnancy. Shi et al. analyzed the association between urinary concentrations of isoflavonoids and cardiometabolic risk markers using data from 299 pregnant women from the NHANES cohort [141]. Those in the fourth quartile of isoflavones had lower levels of insulin and insulin resistance compared to women in the first quartile. None of the individual isoflavonoids (daidzein, equol, and O-desmethylangolensin) in urine were significantly associated with insulin levels.
### *4.2. Adults*
To date, two American cross-sectional studies have examined the relationship between phytoestrogens and cardiovascular-related hormones in healthy adults. In one, only lignan intake seemed to be associated with lower fasting insulin in men [142], whereas the other study reported no significant differences [143]. Ferguson et al. took a different approach, inducing transient endotoxemia in young and healthy volunteers and then analyzing the ability of dietary phytoestrogens to reverse the inflammatory response. They found a significant decrease of insulin sensitivity with the higher intake of isoflavones. The participants were asked to follow a healthy diet but did not receive counseling related to soy food intake. Moreover, similar trends were found in two independent cohorts (MECHE and NHANES) [144].
Six RCTs have evaluated the effect of phytoestrogens on cardiovascular health. The participants in four studies were at high cardiovascular risk [145–148], in one they were men with increased risk of colorectal cancer [149], and in the other they were healthy men [150]. Interventions included isoflavones [149,150], soy nuts [145], daidzein [146], genistein [148], and S-equol [147] administered for periods ranging from 4 weeks to 6 months. Three of the trials found no significant effects on insulin, leptin, or adiponectin [145–147]. However, Amanat et al. report that a daily dose of 250 mg of genistein administered to non-alcoholic fatty liver patients for 8 weeks reduced insulin levels [148]. Maskarinec et al. concluded that men who consumed soy early in life had higher levels of leptin, although no association was observed with soy intake during adulthood [150]. Finally, the results of the trial with subjects at high colorectal cancer risk suggested that isoflavones might reduce the insulin-growth factor but only in equol producers [147].
### *4.3. Postmenopausal Women*
Most studies on phytoestrogen intake and cardiometabolic hormones have evaluated insulin and insulin resistance (HOMA-IR) in postmenopausal women. The largest observational study included 301 women from The Netherlands and used food frequency questionnaires to assess dietary isoflavone and lignan intake. Individuals with a high lignan diet had lower blood pressure but no significant associations with insulin were observed [151]. In another cross-sectional study women in the highest quartile of lignan or enterolactone intake had better anthropometric profiles and insulin sensitivity [152].
Among published clinical trials, a research group from Italy has carried out several studies monitoring cardiovascular risk factors in women receiving 54 mg of genistein. After a 6-month intervention in 60 healthy women, a decrease in insulin and insulin resistance was observed [110]. Similar results were obtained after 12 and 24 months of intervention in a related study in 389 osteopenic postmenopausal women, who received the same dose of genistein plus calcium and vitamin D [153], the values remaining consistent after an extra year of follow-up in a sub-cohort [154]. Examining the role of metabolic status, Villa et al. divided the intervention group into normo- and hyperinsulinemic patients and found that genistein improved insulin sensitivity indexes only in in the latter [71]. Similarly, women with metabolic syndrome who consumed 54 mg of genistein had lower levels of fasting insulin and HOMA-IR, and higher levels of adiponectin than the placebo group [155,156]. More recently, a research group from Iran assessed the effectiveness of 108 mg of genistein on different metabolic factors in 54 women with type-2 diabetes mellitus in a 12-week intervention. As in the other studies, genistein reduced insulin sensitivity [157].
Other RCTs have used soy isoflavones instead of genistein, administering daily doses of 40–160 mg, far higher than the phytoestrogen intake reported in observational studies. For example, in one RCT the mean daily isoflavone intake in the highest tertile was 11.4 mg [151] as opposed to a total mean intake of 0.06 mg in an observational study [152]. Most of the trials with isoflavones have been performed in healthy postmenopausal women for durations ranging from 8 weeks to 24 months. Generally, the studied vascular-related hormones are leptin, adiponectin, and insulin, although in few cases, ghrelin and resistin have also been evaluated. Results from these trials are ambiguous. Whereas most report no significant changes in any of the aforementioned hormones [158–160], others have found beneficial effects on insulin markers in the treatment group compared to the control [118,161–163], or a significant increase in adiponectin peripheral levels [164]. Overall, we can conclude that phytoestrogen therapy did not change hormone levels in obese postmenopausal women [70,165], whereas among diabetic women in a randomized cross-over trial there was a significant decrease in insulin resistance in the soy consumers compared to the placebo group [109].
### **5. E**ff**ect of Phytoestrogen Intake on Hormones Related to Stress Response**
No significant changes in cortisol were observed in healthy women or in those at cardiometabolic risk after consuming soy isoflavones for 2–6 months [18,23,53,95,126]. However, differences between equol excretors and non-excretors have been described in premenopausal women, levels being lower in those who produce this metabolite [60].
### **6. E**ff**ect of Phytoestrogen Intake on Hormones Related to Bone Remodeling**
Estrogen plays a key role in bone metabolism, contributing to bone mass acquisition in puberty and helping to maintain normal bone density in adulthood [166]. Given that phytoestrogens are structurally similar to estrogens, they can bind to ERs in bone and exert estrogenic actions [167].
Most studies examining the impact of phytoestrogens on bone health measure osteocalcin (OC), a metabolic regulatory hormone secreted by osteoblasts, as it is a sensitive biomarker for bone formation [168]. Parathyroid hormone (PTH), secreted by the parathyroid glands, plays an important role in calcium and phosphate metabolism. As well as stimulating bone turnover, there is increasing evidence that PTH may also promote bone formation [169].
### *6.1. Children*
Early-life exposure to soy protein formula did not produce any change in OC and PTH in a clinical study of 48 children [29]. Even though the available data suggest that phytoestrogen intake does not affect bone-related hormones in early stages of life, more studies are needed to clarify this relationship.
### *6.2. Premenopausal Women*
The reported effects of dietary phytoestrogen on bone health in premenopausal women are inconsistent. Kwak H.S. et al. found an increase in serum OC after the administration of 120 mg/day of soy-isoflavones for three menstrual cycles. They also observed that high genistein-excretors in the soy group had higher concentrations of OC, suggesting that individual variation may affect the metabolism and functions of isoflavones [54]. However, previous studies report unaltered OC levels [170,171], indicating a need for more research on the phytoestrogen effects on bone metabolism in premenopausal women.
### *6.3. Postmenopausal Women*
After menopause, estrogen concentrations decrease dramatically, triggering a greater risk of osteoporosis [172]. Phytoestrogens might improve bone health due to their estrogenic effects, and it has been hypothesized that they could reduce the risk of osteoporosis. Chiechi L.M. et al. and Scheiber M.D. et al. report an increase in OC concentrations in postmenopausal women who consumed a soy-rich diet for 6 and 3 months, respectively. Although uncorroborated by the majority of studies, these results indicate a stimulation of osteoblast activity and suggest that soy may have beneficial effects on bone health [173,174]. It has been suggested that longer treatments may be necessary to produce any change in bone metabolism, but to date neither shorter nor longer studies have reported any alterations in OC related to phytoestrogen intake [74–77,96,97,107,170,175–179].
In contrast, beneficial effects on bone metabolism through mechanisms of action not involving OC have been described in healthy postmenopausal women [77,97,175–179]. Lambert M.N.T. et al. demonstrated that red clover-derived isoflavones combined with probiotics attenuated estrogendeficient bone mineral density loss and improved bone turnover even in postmenopausal women with osteopenia [177]. Moreover, a recent meta-analysis and systematic review of RCT with perimenopausal and postmenopausal women concluded that isoflavones can be effective in preserving bone mineral density and attenuating accelerated bone resorption [103]. A possible explanation for these contrasting results could be that estrogens are predominantly antiresorptive agents, so the beneficial effects of phytoestrogens may arise from decreased bone resorption by osteoclasts rather than increased bone formation by osteoblasts.
Lastly, administration of isoflavones or genistein alone for 1 to 24 months did not alter PTH in postmenopausal women [78,92,180–183]. Only a cross-sectional study carried out with Chinese women found that postmenopausal women with a high intake of isoflavone had lower serum PTH levels [184].
### **7. E**ff**ect of Phytoestrogen Intake on Insulin-Like Growth Factors**
Insulin growth factor 1 (IGF-1) is part of the growth hormone (GH)—IGF-1 axis and is mostly produced in the liver in response to GH stimulation. Among many other functions, IGF-1 binds its receptor on osteoblasts and enhances bone formation, so any changes in this hormone will have an impact on bone health [185]. Consequently, some studies have used IGF-1 and its binding proteins IGFBP-1 and IGFBP-3 as bone turnover biomarkers.
In addition, several epidemiological studies have shown that higher levels of IGF-1 are associated with an increased risk of different types of cancer. IGF-1 exerts its actions by binding to the IGF-1 receptor, which is expressed in most tissues of the body and stimulates cell proliferation (Cohen DH 2012). Apart from higher levels of IGF-1, several cancers also overexpress its receptor IGF-1R, which has a negative impact on their progression. IGF-2 also appears to be associated with gastrointestinal and gynecological tumors [186].
It has been hypothesized that phytoestrogens may interfere with the IGF system through their effects on steroid hormone physiology or by disrupting GH and IGF signaling [187]. However, the limited evidence in humans is inconclusive, as studies have found both positive and negative results.
### *7.1. Premenopausal Women*
To date, only two RCTs have evaluated phytoestrogen effects on IGF-1 and its binding proteins in premenopausal women. In the first study, groups of 14 women consumed soy protein isolates providing 8 mg (control), 65 mg (low dose), or 130 mg (high dose) of isoflavones daily for three menstrual cycles. The low dose significantly increased IGF-1 concentrations compared to the high dose only in the periovulatory phase of the menstrual cycle, although no value was significantly different compared to the control group. A similar result was obtained with IGFBP-3; its concentrations were increased by the low dose diet compared with the high dose in the early follicular phase, but they did not differ from those of the control group [170]. The other study assessed the effects of red clover-derived isoflavone supplementation on IGF-1, IGFBP-1, and IGFBP-3 and its role in breast cancer prevention. This one-month intervention resulted in a non-significant reduction in IGF-1, but this was likely due to differences in IGF-1 levels at baseline between the placebo and the control group. Interestingly, the IGF status was found to be influenced by the stage of the menstrual cycle [188].
Epidemiological data does not support a phytoestrogen effect on IGF levels either. A Japanese cross-sectional study reported that there was no correlation between soy products and isoflavone intake and serum IGF-1 and IGFBP-3 in 261 premenopausal women [189]. Another observational study in women living in Japan and Hawaii also failed to find an association between tofu intake and IGF-1, IGFBP-3, and IGF-1 molar ratio in premenopausal women [190].
### *7.2. Postmenopausal Women*
Most clinical studies on IGFs in postmenopausal women have failed to find a protective effect of phytoestrogens against osteoporosis, breast cancer, or colorectal cancer. One of the most recent found no impact on IGF-1 in women with osteopenia after a 24-month intervention [183], which is consistent with other studies reporting that isoflavone supplementation did not alter the IGF system [170,188,191]. Similar results have been obtained for lignan consumption. After administering flaxseed lignans for 3 months, Lucas E.A. et al. found that IGF-I and IGFBP-3 levels were unaltered [88]. An RCT in which 103 postmenopausal women consumed 400 or 800 mg of EGCG for 2 months found no significant changes in IGF-1 or IGFBP-3, although the latter tended to increase in both groups [91].
In contrast with these results, a study comparing the effects of soy protein and milk-based protein reported that both supplements increased IGF-1 levels. Further stratification showed that soy protein had a more pronounced effect on women who were not on hormone replacement therapy [192]. A cross-sectional study found an association between phytoestrogens and growth factors, specifically an inverse association between tofu intake and IGF-1 levels and the molar ratio in postmenopausal
women, whereas no changes were observed in those who had never used hormone replacement therapy [190]. However, a similar observational study performed in participants of the Singapore Chinese Health Study did not find any association between soy intake and the IGF-1, IGFBP-3, and IGF molar ratio [193].
### *7.3. Men*
It is well established that higher circulating IGF-1 levels are associated with an increased risk of PCa [194], and most studies assessing phytoestrogen effects on adult men have consequently focused on PCa patients. In one of two clinical trials with PCa patients, no changes in IGF-1 or IGFBP-3 were observed after a 3–6-month intervention consisting of 200 mg/day of soy isoflavones [37] and in the other Bylund A. et al. (2003) also found that IGF-1 levels remained unaltered after the administration of rye bran bread for 3 weeks [41]. Conversely, a cross-sectional study with 312 men reported a positive association between soy intake and the IGF-1, IGFBP-3, and IGF molar ratio [193].
### **8. Conclusions**
This review has summarized the results of studies on the effects of dietary phytoestrogens on endocrine regulation in humans. Although preclinical studies (in vitro and in animal models) show phytoestrogens to be potentially estrogenic compounds, triggering anti-estrogenic effects in the organism, the results of epidemiological studies are ambiguous.
The impact of phytoestrogens can vary according to the life stage (Figure 3). There is particular concern about how they may affect pregnant women, as this has been poorly studied. Soy isoflavones appear not to have any influence on sex and thyroid hormones, bone remodeling and IGF. However, a study focused on cardiometabolic risk reported a decrease in the level of insulin and insulin resistance in pregnant women consuming higher amounts of isoflavones. Although phytoestrogens transfer from maternal blood to the fetus, no effects have been observed in early life. Nor have endocrine changes been found in infants fed with soy formula, except in a retrospective study carried out in the first year of life of infants with congenital hypothyroidism, which reported an increase of TSH but no conclusive effects on thyroid function. Nevertheless, consumption of phytoestrogens in conditions of insufficient iodine and hypothyroidism may negatively affect thyroid function and favor endocrine imbalance, although such effects have not been observed in euthyroid individuals living in areas with enough supply of iodine. In later stages of childhood, an increase of androgens and decrease of estrogens associated with dietary phytoestrogens have been observed in girls and boys, respectively.
In adulthood, endocrine changes arising from phytoestrogen consumption are unclear, although goitrogenic activity has been observed in men. Effects on sex hormones and IGFs in men are ambiguous, as studies report contradictory results. PCa risk in patients with PCa was unaltered, whereas equol producers with colorectal cancer risk showed a decrease of IGF. Results regarding cardiometabolic risk-related hormones are inconclusive in healthy subjects. Although higher levels of leptin have been reported in early life, no association has been identified in adulthood. However, a reduction in insulin levels was found in patients with non-alcoholic fatty liver.
In premenopausal women, usually studied separately from postmenopausal women, uncertain results have been obtained regarding sex hormones, breast cancer protection, and bone remodeling. Nor has evidence been provided for phytoestrogens affecting IGF levels. Whereas no significant changes in thyroid function were observed, a decrease of free-T3 was found in healthy young females. Among stress response-related hormones, no significant changes in cortisol are described in healthy women or in those at cardiometabolic risk, but a lower production of cortisol is reported in equol-excretors. In postmenopause, the results reported for sex hormones are also ambiguous. However, possible goitrogenic activity derived from phytoestrogen consumption opens up a path for future research. Apart from that, an ameliorative effect has been observed in the cardiometabolic profile of hyperinsulinemic patients, individuals with metabolic syndrome and diabetes. Regarding bone remodeling, the effects of phytoestrogens on OC concentrations are unclear, and their beneficial
impact may arise instead from reducing bone resorption by osteoclasts. The results obtained for PHT and IGF are unconvincing, precluding the drawing of any conclusions.
In general, the available evidence for an association between dietary phytoestrogens and endocrine biomarkers is inconclusive. The disparity in results may be due to differences in the type and concentration of the compounds administered and the variety of matrices, which could influence phytoestrogen bioavailability and consequently the effect on hormonal function. Also, while most studies analyze circulating hormones, others report the urinary excretion of metabolites. There is a clear need for further carefully designed studies to elucidate the effects of phytoestrogen consumption on the endocrine system.
**Figure 3.** Summary of the effects of dietary phytoestrogens at different life stages. NAFLD: non-alcoholic fatty liver disease.
### **9. Future Directions**
Based on the available literature, we can conclude that intake of phytoestrogens does have some physiological effects in humans related to hormone regulation, but like hormones, the benefits depend on the stage of life. Some factors such as dose and type of compounds, as well as matrices englobing these phytoestroestrogens (food, capsule, etc.) affect their bioavailability and, therefore, the observed results. Most of the research is focused on postmenopausal women and only some have explored the effects during pregnancy and early stages of life. For instance, the effect of phytoestrogen intake on pubertal development has been poorly studied and could lead to interesting results. In order to do that, well-designed intervention trials are key to shed some light on this topic, especially regarding associations that are controversial.
**Author Contributions:** Conceptualization, A.T.-R. and S.H.-B.; Bibliography searching methodology, S.H.-B.; Writing—original draft preparation, I.D.-L., M.Y.-A., A.S.-H., A.T.-R., and S.H.-B.; Writing—review and editing, A.T.-R. and S.H.-B.; Supervision, A.T.-R. and S.H.-B. All authors have read and agreed to the published version of the manuscript.
**Funding:** This research received no external funding.
**Conflicts of Interest:** The authors declare no conflict of interest.
### **References**
© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
| doab | 2025-04-07T03:56:58.617044 | 1-5-2021 17:49 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "003cf560-3166-422f-9a1c-98b1e971efa6",
"url": "https://mdpi.com/books/pdfview/book/3241",
"author": "",
"title": "Dietary Polyphenols and Human Health",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783039437931",
"section_idx": 13
} |
003cf560-3166-422f-9a1c-98b1e971efa6.14 | *Review* **Metabolic Impact of Flavonoids Consumption in Obesity: From Central to Peripheral**
**Viviana Sandoval 1,**†**, Hèctor Sanz-Lamora 1,2,**†**, Giselle Arias 1, Pedro F. Marrero 1,3,4, Diego Haro 1,3,4,\* and Joana Relat 1,2,4,\***
Received: 21 July 2020; Accepted: 5 August 2020; Published: 10 August 2020
**Abstract:** The prevention and treatment of obesity is primary based on the follow-up of a healthy lifestyle, which includes a healthy diet with an important presence of bioactive compounds such as polyphenols. For many years, the health benefits of polyphenols have been attributed to their anti-oxidant capacity as free radical scavengers. More recently it has been described that polyphenols activate other cell-signaling pathways that are not related to ROS production but rather involved in metabolic regulation. In this review, we have summarized the current knowledge in this field by focusing on the metabolic effects of flavonoids. Flavonoids are widely distributed in the plant kingdom where they are used for growing and defensing. They are structurally characterized by two benzene rings and a heterocyclic pyrone ring and based on the oxidation and saturation status of the heterocyclic ring flavonoids are grouped in seven different subclasses. The present work is focused on describing the molecular mechanisms underlying the metabolic impact of flavonoids in obesity and obesity-related diseases. We described the effects of each group of flavonoids in liver, white and brown adipose tissue and central nervous system and the metabolic and signaling pathways involved on them.
**Keywords:** non-alcoholic fatty liver disease; obesity; flavonoids; lipid metabolism; metabolic regulation; adipose tissue; brain
### **1. Introduction**
Overnutrition and unhealthy diets together with physical inactivity cause an impairment in the metabolic homeostasis that lead to the development of pathologies such as obesity, type 2 diabetes, cardiovascular diseases (CVD) and more recently this kind of lifestyle has also been linked to neuroinflammation and neurodegenerative diseases [1–5].
The metabolic syndrome (MetS) is the medical term used to define the concomitance in an individual of some of the following alterations: hyperglycemia and/or insulin resistance, arterial hypertension, dyslipidemia and central or abdominal obesity [6]. It is currently one of the main public health problems worldwide and its incidence increases significantly each year, affecting almost 25% of the adult population today and has been directly associated to a greater risk of suffering from CVD or type 2 diabetes among others [3].
Obesity is one of the most important trigger for many of the other alterations include in the MetS. Obesity is essentially caused by an imbalance between energy intake and energy expenditure that initially causes an expansion of the white adipose tissue (WAT) to store the overfeed as triglycerides (TG). Some evidences indicate that at some point, WAT fails to adequately keep the surplus of nutrients and together with an insufficient differentiation of new adipocytes lead to an off-WAT accumulation of lipids in peripheral relevant organs. This ectopic accumulation of lipids causes lipotoxicity that may be, at least in part, responsible of the metabolic obesity-related metabolic dysfunctions [7]. It seems obvious that defects in WAT functionality together with peripheral lipotoxicity are the key points in the onset of metabolic syndrome (MetS) [8]. Looking for a way to restore lipid homeostasis and reduce lipotoxicity but also to diminish adipose tissue inflammation and macrophage infiltration many research groups are focused on identifying specific dietary patterns or foods capable to counteract these effects to finally revert obesity and its comorbidities.
Furthermore, it has been described that long-term hyperglycemia and diabetes complications induce impairments in the hippocampal synaptic plasticity as well as cognitive deficits [9] and increase the risk for Alzheimer disease [10,11] and depressive illness [12]. On the other side, diet-induced hypothalamic inflammation and mitochondrial dysfunction result in the onset and development of obesity and related metabolic diseases. It has been shown that, in rats, high fat diet (HFD) induces metabolic inflammation in the central nervous system (CNS), particularity in the hypothalamus [13].
The prevention of MetS and obesity is primary based on the follow-up of a healthy lifestyle, which includes, among other recommendations, a healthy diet. In this context, the Mediterranean Diet (DietMEd) has shown beneficial effects on the prevention and treatment of MetS and obesity by reducing chronic low-grade inflammation, improving endothelial function and reducing cardiovascular risk [14–16]. The study of Prevention with Mediterranean Diet (Predimed) has shown that high adherence to this nutritional profile is effective in the primary and secondary prevention of CVD, diabetes and obesity [17–24]. DietMed is characterized by a high consumption of foods rich in bioactive compounds such as polyphenols to whose have been attributed a large part of the health effects of this diet [18,23,25–28].
In this review, we have summarized the current knowledge on the metabolic effects of a specific group of polyphenols, the flavonoids, and the molecular mechanisms underlying these effects.
Concretely, the main goal of the present work is to describe the molecular mechanisms underlying the anti-obesity effects of flavonoids in three target organs/tissues: liver, adipose tissues (WAT and brown adipose tissue (BAT)) and central nervous system (CNS).
We choose a high variety of obesity models, sources and doses of flavonoids to identify the metabolic and signaling pathways involved in the effects of each subclass of flavonoids (anthocyanins, flavanols, flavanones, flavonols, isoflavones, flavones and chalcones) in these tissues/organs. Only studies in humans and experimental approaches whit animal models from the last years have been included, thus avoiding cell culture experimental approaches except when relevant.
### **2. Polyphenols and Metabolism**
Polyphenols are the most abundant phytochemicals in nature. They are widely distributed in fruits, vegetables, and highly present in foods like legumes, cocoa, some cereals as well as in some beverages, such as tea, coffee and wine [29]. Polyphenols are not essential nutrients for humans but research in nutrition, including epidemiological studies, randomized controlled trials, in vivo and in vitro assays with animal models and cell lines, has shown that long-term and acute intakes can have beneficial effects on weight management and chronic diseases such as CVD, obesity, type 2 diabetes, the onset and development of some cancers and cognitive function [13,30–37].
The effects of polyphenols are directly related to their bioavailability. It is assumed that just the 5%-10% of the total dietary polyphenol intake is absorbed directly through the stomach and/or small intestine, the rest reaches the colon where they are transformed by the microbiota [38–40]. After being absorbed, polyphenols undergo phase I and II metabolism (sulfation, glucuronidation, methylation, and glycine conjugation) in the liver [29]. Polyphenol metabolites derived from liver metabolism may interact, among others, with adipose tissue, pancreas, muscle, and liver, where they exert their bioactivity.
Polyphenols have been divided in two main families: flavonoids and non-flavonoids, that are subdivided into several subclasses. For many years, the health benefits of polyphenols have been attributed to their anti-oxidant capacity as free radical scavengers. More recently it has been described that polyphenols activate other cell-signaling pathways that are not related to ROS production but rather involved in metabolic regulation [23,41].
### *Flavonoids*
Flavonoids are widely distributed in the plant kingdom when are used for vegetables for growing and defensing. They are structurally characterized by two benzene rings and a heterocyclic pyrone ring and based on the oxidation and saturation status of the heterocyclic ring flavonoids are grouped in seven different subfamilies (Table 1).
Flavonoids are abundant in food and beverages highly consumed by human population including fruits, vegetables, tea, cocoa or wine [42] and in global are the bioactive compounds more largely associated with a reduced risk of all-cause mortality, type 2 diabetes [43–46], CVD [36,47], obesity and its comorbidities such as non-alcoholic fatty liver disease (NAFLD) [48–50] and more recently they have been described as potential therapeutic agents against cognitive pathologies such as Alzheimer's disease (AD) [42,51,52] or cerebrovascular alterations [47].
The molecular mechanisms underlying the beneficial effects of flavonoids have been widely studied and, in many cases, involved the activation of the AMP-activated protein kinase (AMPK). AMPK is a key enzyme for the control of lipid metabolism and adipogenesis. AMPK phosphorylation and activation promote catabolic processes such as FAO, glucose uptake, or glycolysis as well as inhibits anabolic pathways such as fatty acid synthesis or gluconeogenesis [53].
### **3. Anthocyanins**
Anthocyanins are natural pigments and are responsible for the red-blue color of several flowers, fruits (mainly berries and grapes), roots, seeds (beans) but also of some leaves and cereal grains where they are found in low concentrations. Cyanidin, delphinidin, malvidin and their derivates are the most commonly studied anthocyanins [29,42,54–56].
Anthocyanins have shown antioxidant and anti-inflammatory properties but also positive effects in obesity and its comorbidities [57–60]. Several studies have demonstrated that the intake of anthocyanins by itself or of anthocyanins-rich foods such as berries is able to prevent CVD [61], to reduce body fat accumulation, to improve glucose tolerance/insulin sensitivity, to diminish the levels of fasting glucose, to control body weight in humans and rodents [57,59,62–72] and to increase energy expenditure and fatty acid oxidation (FAO) in mice and humans [59,73–76]. Globally, anthocyanins and anthocyanins-rich foods are able to improve metabolic homeostasis. More recently, anthocyanins have also revealed promising effects on cognitive function [51,77–79].
Part of the anthocyanins metabolic effects occur by regulating adipogenesis, increasing FAO, lipolysis, thermogenesis and mitochondrial biogenesis, regulating satiety and reducing lipogenesis in different tissues and organs and enhancing energy expenditure and body weight progression [74–76,80–83] Dietary supplementation with anthocyanins improves the lipid profile by favorably controlling the circulating levels of TG, total cholesterol, LDL-cholesterol and HDL-cholesterol [84].
### *3.1. Anthocyanins Improve the Metabolic Hemostasis in Obesity: The Liver Response*
Non-alcoholic fatty liver disease (NAFLD) is characterized by an excessive accumulation of lipids in the livers. Its onset is closely related to obesity where an imbalance between fatty acids input and output causes initially a hepatic steatosis that can progress to NAFLD, non-alcoholic steatohepatitis (NASH), fibrosis, cirrhosis and in some cases hepatocarcinoma. Anthocyanins and anthocyanins-rich foods extracts or juices have demonstrated in several studies their ability to reduce the hepatic content of TG and lipids [85,86] and their capacity to modulate hepatic metabolism to protect against NAFLD [62,87–89]. Although in most of the published approaches performed with rodent models of obesity or NAFLD, anthocyanins or anthocyanin-rich fruits or extracts significatively reduced the hepatic lipid content and ameliorated the hepatic steatosis profile of these animals [88,90–92] some ineffective approaches have also been described [93–95].
The beneficial effects of anthocyanins in the liver have been linked to the activation of the AMPK, the upregulation of glycolytic and FAO genes and the downregulation of the gluconeogenic and lipogenic genes among others [70–72,96,97].
Mulberry anthocyanin extract administration to type 2 diabetic mice increased the activity of AMPK/peroxisome proliferator-activated receptor gamma coactivator 1 alfa (PGC1α)/p38 mitogen-activated protein kinase (MAPK) and reduced the activity of the acetyl-CoA carboxylase enzyme (ACC), a rate-limiting enzyme of fatty acid synthesis, and of the mammalian target of rapamycin (mTOR) that is involved in protein synthesis regulation and insulin signaling [96]. Similar effects were described in HFD-fed hamsters, where Mulberry water extracts exerted anti-obesity effects by inhibiting lipogenesis (downregulation of fatty acid synthase (FASN) and 3-hydroxy-3-methylglutaryl-coenzyme A (HMG-CoA) reductase) and upregulating PPARα and CPT1A [81]. On its side, honeyberry (*Lonicera caerulea*) extract (HBE) also decreased lipid accumulation in the liver of HFD-obese mice. HBE downregulated the hepatic expression of lipogenic genes such as *sterol regulatory element-binding protein-1 (Srebp-1c), CCAAT*/*enhancer-binding protein alpha (C*/*ebp*α*), Ppar*γ*,* and *Fasn* as well as upregulated the mRNA and protein levels of CPT1a and PPARα, thus enhancing FAO. As mulberry anthocyanin extract, HBE treatment also increased the phosphorylation of AMPK and ACC thus activating and inhibiting these enzymes respectively [98]. On the other hand, in NAFLD-induced rats, blackberry extracts improved insulin sensitivity and dyslipidemia, ameliorated triglyceride and lipid peroxide accumulation and suppressed the mRNA expression of genes involved in fatty-acid synthesis (*Fasn* and *Srebp-1c*) [88]. Finally, purple sweet potato reduced the protein levels of FASN and of the cluster of differentiation 36 (CD36), inactivated the C/EBPβ, restored AMPK activity and increased the protein levels of CPT1a in livers of HFD-fed mice, thus indicating decreased lipogenesis and fatty acid uptake and enhanced FAO [62].
Regarding glucose metabolism, protein-bound anthocyanin compounds of purple sweet potato ameliorate hyperglycemia in obese and diabetic mice by regulating hepatic glucose metabolism. Anthocyanin compounds of purple sweet potato induced the hepatic protein levels of p-AMPK, glucose transporter type 2 (GLUT2), insulin receptor α (IRα), glucokinase (GK)*,* as well as the expression of *phosphofructokinase (Pfk)* and *pyruvate kinase (Pk),* while gluconeogenic genes, *glucose-6-phosphatase (G6Pase)* and *phosphoenolpyruvate carboxykinase (Pepck)* were downregulated [99]. Further, Saskatoon berry normalized liver expression of *Gk* and *glycogen phosphorylase* and increased *G6Ppase*in diet-induced MetS rats, thus suggesting that Saskatoon berry regulated glycolysis, gluconeogenesis and glycogenesis to improve MetS [100].
Although most of the experimental approaches have been done using anthocyanins-rich extracts, pure compounds have been also analyzed. Cyanidin-3-glucoside (C3G) administration to C57BL/6J obese mice fed a HFD and db/db mice diminished the triglyceride hepatic content and steatosis [73,101], through the blockade of the c-Jun N-terminal kinase activation (JNK) and the promotion of the phosphorylation and nuclear exclusion of the transcription factor Forkhead box protein O1 (FoxO1) [101].
All these data confirm the impact of anthocyanins and even in a more significative way of the anthocyanin-rich foods on metabolism. These effects can be added to their anti-inflammatory, antiapoptotic, pro-autophagic and antioxidant properties in steatotic livers [59,62,102–104].
### *3.2. Anthocyanins in Adipose Tissue: The Activation of BAT and the Browning of WAT*
The impairment of adipose tissue function is strongly associated with the development of obesity and insulin resistance (IR). The activation of BAT and the browning in WAT are considered potential strategies to counteract the metabolic alterations linked to the obese phenotype. Both actions are mechanisms to increase the energy expenditure (EE) through the induction of lipolysis, FAO and thermogenesis and consequently efficient ways to reduce the ectopic lipid accumulation and the lipotoxicity [105–108].
Part of the beneficial effects of anthocyanins on diet-induced obesity are due to their impact on adipose depots. Anthocyanidins regulate lipolysis, FAO, lipogenesis and adipose tissue development [76,109–111]. They affected the adipokines secretion [112], modified the adipocytes-gene expression [33,113,114]. Moreover, anthocyanins are able to improve WAT functionality, to induce browning in WAT [33,57,82,115] or to increase the BAT mass or its activity [57,109,115], thus regulating energy expenditure [59,73]. Moreover, in WAT, anthocyanins ameliorate the obesity-associated inflammation [57,59,116].
In WAT, an anthocyanin-rich bilberry extract ameliorated hyperglycemia and insulin sensitivity through the activation of AMPK that resulted in an increase of the glucose transporter 4 (GLUT4) [72]. On its side, C3G-enriched *Aronia melanocarpa* extract reduced food intake and WAT weight in HFD-fed mice but also suppressed adipogenesis. These animals showed a downregulating in the expression levels of *C*/*ebp*α, *Srebp1c, Acc, ATP-citrate lyase, Pgc1*α*, Fasn,* and *adipocyte protein 2 (Ap2)* as well as in the circulating levels of leptin [111]. In the same way, in HFD-induced obese mice model, the dietary supplementation with maqui (*Aristotelia chilensis*) improved the body weight gain and glucose metabolism at least in part by modifying the expression of the *carbohydrate responsive element binding protein* β (*Chrebp*β), *the fibroblast growth factor 21* (*Fgf21*) and *adiponectin* as well as of the lipogenic and FAO genes [82]. Globally, the maqui supplementation induced the browning of the subcutaneous WAT (scWAT) [82].
The induction of browning is a common phenotype in obese rodent models treated with anthocyanins or anthocyanins-rich foods. The thermogenic and mitochondrial markers were also increased in the inguinal WAT (iWAT) of high fat-high fructose (HF/HFD)-fed mice treated with C3G, thus indicating the browning of this adipose tissue depot and suggesting an increased heat production and energy expenditure (EE) [117]. In db/db mice, C3G and vanillic acid exerted similar effects: increased EE, limited weight gain and upregulated expression of *Ucp1* and other thermogenic and mitochondrial markers, thus indicating the induction of brown-like adipocytes development in the scWAT [73] or iWAT [115]. Freeze dried raspberry decreased WAT hypertrophy induced by HFD and promoted the browning of WAT as it is showed by a higher expression of beige markers such as *Ucp1, PR-Domain zinc finger protein 16 (Prdm16), Cytochrome C, Cell death inducing DFFA like e*ff*ector A (Cidea), and Fatty acid elongase 3(Elovl3),* elevated levels of PGC-1α and Fibronectin type III domain-containing protein 5 (FNDC5)/irisin, and an activation of the AMPK/Sirtuin 1 (SIRT1) pathway [33]. AMPK and Sirt1 are important sensors of the energy status that together with PGC-1α regulate energy homeostasis and stimulate FNDC5/irisin expression, thus inducing beige adipogenesis [118]. The regulation of adipogenesis through the AMPK/SIRT1 pathway has also been described in HFD fed mice treated with maize extract rich in ferulic acid and anthocyanins [119].
In WAT, anthocyanins and anthocyanin-rich foods also improve the inflammatory profile. The administration of a black soybean testa extracts (BBT) to diet-induced obese mice decreased fat accumulation, and the expression of *Acc* and *C*/*ebp*α and increased the levels of lipolysis proteins such as lipoprotein lipase (LPL), hormone-sensitive lipase (HSL) in mesenteric fat but also showed anti-inflammatory effects [109]. Similar effects were observed in humans where the administration of BBT to overweight or obese individuals decreased the abdominal fat measured as waist and hip circumference and improved the lipid profile [110]. The anti-inflammatory effects have been also achieved with sweet cherry anthocyanins and blueberry (*Vaccinium ashei)* anthocyanins. These anthocyanins reduced the body weight gain, the size of adipocytes and the leptin secretion
in HFD-fed mice but also expression of *Il-6* and *Tnfa* genes, thus indicating an amelioration of the deleterious effects of a HFD [114,120].
Besides their effects on WAT, anthocyanins and anthocyanins-rich food also impact on BAT where they promote its activity. In high fructose/HFD-fed animals, besides inducing the browning of WAT, C3G attenuated the development of obesity by promoting the tremorgenic capacity of BAT. C3G upregulated the expression of thermogenic markers such as *Ucp1,* induced the mitochondrial biogenesis and function and finally increased the EE [117]. In db/db mice, C3G and vanillic improved cold tolerance and enhanced BAT activity and induced mitochondrial biogenesis. In BAT, anthocyanin and anthocyanin-rich foods upregulated the expression of thermogenic markers (*Ucp1, Prdm16, Cidea* ... *)*, lipid metabolism *(Cpt1a, Hsl, adipose triglyceride lipase (Atgl)),* mitochondrial markers *(mitochondrial transcription factor A (Tfam), Nuclear Respiratory Factor 1* and *2 (Nrf1* and *Nrf2)* ... *)* and transcriptional regulators or coactivators of these processes (*Ppar*α, *Pgc1*β, *Pgc1*α... *)* [73,115].
### *3.3. In the Central Nervous System (CNS) Anthocyanins Have Been Related to Neuroprotective E*ff*ects as Well as in Feeding Behavior*
The neuroprotective activity of anthocyanins has been widely evidenced in several epidemiological studies and their potential for the prevention of many neurodegenerative diseases such as Parkinson's disease (PD) and Alzheimer's disease (AD) has been suggested [77,78]. The neuroprotective effects of anthocyanins and C3G correlate with the regulation of molecules upstream of nitric oxide (NO) production, neuroinflammatory response and oxidative stress [79,121–123].
It has been demonstrated that C3G and malvidin 3-O-glucoside (M3G) inhibited the hyperphosphorylation of Tau protein in Alzheimer's disease [124] and berries supplementation have shown neurocognitive benefits in older adults at risk for dementia with mild cognitive impairment [125]. Recent studies highlighted an anti-depressive effect of a maqui-berry extract in a mouse model of a post-stroke depression. In this case the maqui effects were associated to its antioxidant capacity [126]. Otherwise, anthocyanins extracted from dried fruits of *Lycium ruthenicum Murr* have demonstrated a protective role in cerebral ischemia/reperfusion injury in rats [127] by inhibiting cell apoptosis and reducing edema and inflammation.
Besides their role in neuroprotection, anthocyanins modulate the feeding behavior. In rats, anthocyanins from black soybean increase the expression of the gamma-aminobutyric acid B1 receptor (GABAB1R) and decrease the expression of neuropeptide Y (NPY) in the hypothalamus, thus modulating the food intake behavior/body weight control. The upregulation of GABABR1 is followed by a decrease of the activated protein kinase A (PKA) and the phosphorylated cAMP-response element binding protein (CREB), both located downstream of GABAR1 [83]. In a similar way, the administration of an anthocyanin-rich black soybean testa (Glycine max (L.) Merr.) to diet-induced obese mice decreased food intake [109].
### **4. Flavanols**
Flavanols are present in cocoa, tea, red wine, beer and several fruits such as grapes, apricots, apples where they are responsible for their astringency [128]. Flavanols exist as monomers named catechins or as polymers named proanthocyanins. The monomeric forms include: catechin (−)-epicatechin (EC), (−)-epigallocatechin gallate (EGCG), (−)-epigallocatechin (EGC), and (−)-epicatechin gallate (ECG). The proanthocyanins, also known as tannins, are more complex structures (dimers, oligomers, and polymers of catechins) and can be transformed to anthocyanins [29]. Like other flavonoids, flavanols are absorbed between the small intestine and the colon depending on their physicochemical properties and structure [129].
Flavanols possess a health claim related to their role in maintaining the elasticity of blood vessels that was approved in 2014 by the European Food Safety Authority (EFSA) [130].
In humans and animal models, flavanols or flavanols-rich foods (mainly, cocoa or tea derivates) have demonstrated the ability to reduce body weight, decrease waist circumference and fat percentages, improve glucose metabolism in individuals with type 2 diabetes, obesity or MetS and increase energy expenditure [75,131–139]. One of the most described molecular mechanism underlying theses effects are the activation of the AMPK enzyme [140].
Due to the high amount of publications including flavanols and metabolism we just included a representative group of the most recently published and the ones that deepen more on the molecular mechanisms underlying the beneficial effects of flavanols.
### *4.1. Flavanols Improve Hepatic Steatosis and Glucose*/*Lipid Metabolism in Obesity Models*
In humans and several rodent models of obesity, flavanols have been able to improve blood lipid profile and protect liver from excessive fat deposition and hepatic steatosis [136,141–146]. These effects have been related mostly with an activation of the AMPK and the protein kinase B (PKB/Akt) pathways that finally lead to the suppression of lipogenesis by modulating the expression of *Srebp1c*, *cAMP-response element-binding protein regulated transcription coactivator 2 (Crtc2),* and *stearyl coenzyme A dehydrogenase-1 (Scdh1)* or the activity of ACC, the inhibition of gluconeogenesis by affecting the levels of *PepcK* and *G6pase* and the increment of FAO by increasing the *Cpt1a* levels. Moreover, flavanols are able to improve cholesterol homeostasis through the regulation of several enzymes from the cholesterol synthesis and bile acids metabolism apart from the modulation of the mRNA expression of apolipoprotein B100 and ATP-binding cassette transporter A1. Most of the approaches included have been done using tea extracts or cocoa flavanols but other extracts with a more diverse composition of flavonoids have been also described in this section [137,143,147–151].
Theabrownin from Pu-erh tea in combination with swinging improved serum lipid profile and prevent development of obesity and insulin resistance in rats fed a high-fat-sugar-salt diet and subjected to a 30-min daily swinging. A transcriptomic analysis in the liver indicated that theabrownin together with exercise activated circadian rhythm, PKA, AMPK, and insulin signaling pathway, increased the levels of cAMP and accelerated the consumption of sugar and fat [142]. Similar results were obtained with HFD-fed mice supplemented with Yunkang green tea and subjected to treadmill exercise. These animals showed a reduction in the body weight gain and liver weight, a lower level of blood glucose, serum total cholesterol (TC), TG, insulin and ALT and an improvement in the fatty liver and hepatic pro-inflammatory profile compared to HFD group. Supplemented and exercised-animals showed a downregulation of the lipid synthesis genes (*Srebp1c, Fasn, Acc),* and an improvement of the hepatic insulin signaling [143].
Furthermore, in obese Zucker rats fed with a HFD and treated with green tea polyphenols a significant reduction on fasting insulin, glucose and lipids and an improvement of the NAFLD were observed. Livers of treated rats had lower levels of alanine aminotransferase (ALT) and aspartate aminotransferase (AST), of inflammatory markers and of TG content and exhibited less lipid droplets. These improvements have been related to an activation of the AMPK pathway and the inhibition of the hepatic lipogenesis (higher levels of the inactive p-ACC and lower levels of SREBP1c) [152]. These effects on lipid metabolism were also observed after the administration of Benifuuki (a tea that contains methylated catechins such as epigallocatechin-3-O-(3-O-methyl) gallate (EGCG3 'Me) to high fat/high sucrose diet-fed mice. Benifuuki treatment lowered the levels of TG and NEFA in serum and liver and reduced the expression of hepatic lipogenic genes (*Srebp-1c, Acc1, Fasn* and *Stearoyl-CoA desaturase 1(Scd1*)) [153]. In parallel the use of *Euterpe oleracea* Mart.-derived polyphenols, known by the popular name of açai and rich in catechin and polymeric proanthocyanins, when administered to HFD-fed mice [154] or a pistachio-diet supplementation to diet-induce obese mice exhibited similar impact on lipid metabolism and gene expression modulation [150].
Finally, Oliogonol, a flavanol-rich lychee fruit extract, significantly reduced hepatic lipid content (less lipid droplets and ballooning by downregulating the *Ppar*γ and, *Srebp1c* mRNA levels [155] probably via the inhibition of the mTOR activity promoted by the activation of the AMPK enzyme [156]. Moreover, oligonol improved hepatic insulin sensitivity by reducing the phosphorylation of glycogen synthase kinase 3a (GSK3a) and the phosphatase and tension homologue (PTEN) in HFD-induced
obese mice [155] as well as inhibiting the mTOR/S6K cascade. The activation of the mTOR/S6K phosphorylates and desensitizes the insulin receptor substrate 1 (IRS1) [157]. In a similar way, GC-(4→8)-GCG, a proanthocyanidin dimer from *Camellia ptilophylla* improved hepatic steatosis and hyperlipidemia in HFD-induced obese mice [158].
Besides on hepatic lipogenesis, tea extracts also impact in FAO. The administration of tea water extracts from green tea, yellow tea, white tea, black tea, raw pu-erh tea and oolong tea decreased TG and total cholesterol levels in serum and liver as well as the hepatic lipid content. Supplemented animals displayed less lipid droplets, the activation of the AMPK and the upregulation of the *Cpt1a* together with the inhibition of the FASN enzyme. These treatments also reduced the inflammation profile linked to HFD [149]. Similar results were obtained with grape seed procyanidin B2 (GSPB2) and a polyphenol extract from *Solanum nigrum* that contains among other different catechins. In db/db mice, GSPB2 decreased body weight and improved the lipid profile in serum (TG, total cholesterol and free fatty acids (FFA)) but also reduced hepatic lipid droplets and TG accumulation. The proposed mechanism implied the AMPK activation, the ACC phosphorylation and *Cpt1a* overexpression, thus inhibiting FA synthesis and increasing FAO [159]. In a similar way, the *Solanum nigrum* polyphenol extract inhibited lipogenesis and enhanced FAO (upregulation of *Cpt1a* and P*par*α) through the AMPK cascade [151].
In different animal models of obesity and insulin resistance, EGCG has shown the capacity to improve glucose homeostasis, to inhibit gluconeogenesis, FA and cholesterol synthesis and to increase FAO [147,148]. In HFD and STZ-induced type 2 diabetes, EGCG downregulated *Pepck* and *G6Pase* and inhibited SREBP1c, FASN and ACC1. The mechanism underlying these effects is not yet well understood but it has been suggested that EGCG would activate the PXR/CAR-mediated phase II metabolism that through a direct or indirect mechanism would suppress gluconeogenesis and lipogenesis [147]. Moreover, in HFD Wistar rats, EGCG diminished the liver weight, the hepatic hyperlipidemia, animals showed less lipid droplets, reduced serum levels of ALT and AST, TG, total cholesterol and better profile of LDL/HDL but also an ameliorated oxidative stress. In this case, EGCG activated SIRT1, FoXO1 and regulate SREBP2 activity to suppress hepatic cholesterol synthesis. These data point out the downregulation of SREBP2 expression under the SIRT1/FOXO1 signaling pathway as a mechanism to reduce the cholesterol content [148]. Furthermore, EGCG also decreased bile acid reabsorption, which decreased the intestinal absorption of lipids [160]. In the same way, EC administered to a high-fat high cholesterol diet rats reduced serum levels of total cholesterol, LDL and TG while increased HDL [161]. Moreover, EC intake also reduced serum levels of ALT and AST enzymes, the lipid peroxidation and the pro-inflammatory cytokines levels, thus indicating an improvement in the liver functionality. The proposed mechanism of EC included the downregulation of the nuclear receptor liver-X-receptor (LXR), the FASN enzyme and the SIRT1 protein but also the blockade of the Insig-1-SREBP-SCAP pathway that drives the SREBP2 maturation [161].
### *4.2. Flavanols in Adipose Tissue: Less Adiposity and More Energy Expenditure: The Browning E*ff*ect*
In humans, some studies described the capacity of green tea to reduce body weight and abdominal fat accumulation [162,163], influence on the body fat mass index, waist circumference, total fat mass and energy expenditure through the induction of browning or BAT activity [164–166] but also to regulate ghrelin secretion and adiponectin levels, to control appetite and decrease nutrient absorption [135,167].
In rodents, the administration of grape seed-derived proanthocyanins to Wistar rats reduced the body weight by limiting food intake and activating EE in scWAT [168] and it has been widely described that in rodent models of obesity, flavanols are able to affect the lipid metabolism of WAT and BAT. Global effects of flavanols in adipose tissues lead to a decrease in adiposity, specially of the WAT depots and in adipocyte size by reducing adipogenesis, the release of adipokines such as leptin and resistin, the modulation of lipid metabolism and the induction of browning [153,155,158,169–174]. In BAT, flavanols caused the activation of thermogenesis and FAO [172–176].
As has been mentioned before, in WAT, flavanols modified lipid metabolism. EGCG reduced the expression of genes related with *de novo* lipogenesis (*Acc1, Fasn, Scd1, C*/*ebp*β*, Ppar*γ and *Srebp1c*), increased the expression of genes involved in lipolysis (*Hsl*) and lipid oxidization (*Ppar*α, *Acetyl-CoA oxidase (Acox)2,* and medium-chain acyl-CoA dehydrogenase (*Mcad)*) in epididymal (eWAT) and scWAT and highly upregulated the expression of delta-9 desaturase, the enzyme responsible to convert saturated fatty acids to monounsaturated [177]. The activation of the AMPK in HFD-EGGC-treated mice indicated that at least in part the changes in lipid metabolism observed were due to the AMPK phosphorylation [177]. In scWAT, although EGCG increased lipolysis (*Hsl)* and FAO (*Cpt1a)* [168,178], some lipogenic genes (*Acc1, Fasn, Scd1, Ppar*γ*,* and *Srebp1*) has been detected upregulated at the mRNA level but no at protein level [178]. These data suggested that EGCG might have different effects in scWAT and eWAT. Finally, pistachio-diet supplementation to diet-induce obese mice also ameliorated the HFD-induced expression of *Srebp1c*, *Ppar*γ, and *Fatp* [150].
Besides its effects in the liver, the GC-(4→8)-GCG inhibited the expansion of all WAT depots in HFD fed mice. Adipocytes from eWAT were smaller and some of the main adipocyte-associated transcription markers were downregulated (*Srebp1c, C*/*ebp*α *and Ppar*γ), thus indicating a better WAT functionality [158]. The GC-(4→8)-GCG-supplemented mice showed an upregulation of the adiponectin and a downregulation of the leptin mRNA levels as well as an improved inflammatory profile with less macrophage infiltration [158].
Regarding the browning effect of flavanols it has been published that EC increased mitochondrial biogenesis, fatty acid metabolism and upregulated the expression of BAT-specific markers (*Prdm16*, *Dio2*, *Ucp1* and U*cp*2) in WAT in a way that depends on phosphorylation and deacetylation cascades [170]. The authors demonstrated that EC supplementation upregulated the mitochondrial related proteins p-SIRT1, SIRT1, SIRT3, PGC1α, PPARγ, TFAM, NRF1, NRF2, complex II, IV and V and mitofilin [170]. In a similar way, a polyphenolic extract from green tea leaves (GTE) ameliorated the body weight gain caused by a HFD with no changes in calorie intake but reducing the adiposity and the adipocyte size in WAT and BAT. GTE supplementation induced BAT markers in scWAT (higher mRNA levels of *Pgc1*α*, Cbp*/*p300-interacting transactivator 1 (Cited1)* and *Prdm16* and of UCP1 protein) and reduced HFD-induced whitening in BAT (lower expression of adipogenic markers *C*/*ebp*α and *Ap2* and upregulation of *Pgc1*α and *vascular endothelial growth factor-A(165)* (*Vegfa165*)) [171]. These animals also showed an improvement in the inflammatory profile in scWAT and BAT. Finally, a Grape pomace extract (GPE) showed the capacity to induce browning (upregulation of *Pgc1*α*, Ppar*γ*, Prdm16* and *Ucp1*) in the eWAT of HFD-fed rats [179,180].
Besides tea extracts also cacao components are able to induce browning and BAT activation. Concretely, theobromine alleviated diet-induced obesity in mice by inducing a brown-like phenotype in the iWAT and activated lipolysis and thermogenesis in BAT. In HFD fed mice theobromine inhibited phosphodiesterase-4 (PDE4D) activity in adipose tissue, thus increasing β3-adrenergic receptor (AR) signaling pathway and EE [172]. The inhibition of PDE increases the cellular levels of cAMP levels thus activating the β-AR cascade and finally PKA and UCP1 activity [181].
The capacity of flavanols on activating BAT has been described even with a single dose of a flavanol mixture that included catechins and B type procyanidins or by administering individual components by itself [182]. In these animals, *Ucp1* mRNA expression in BAT and levels of catecholamines in plasma were significantly increased via SNS stimulation but with varying efficacy depending on the stereochemical structure of flavanols [182]. It should be noted that prolonged ingestion of a catechin-rich beverage increased the BAT density with a decrease in extramyocellular lipids in humans [183]. EGCG-supplemented diet-induced obese mice exhibited higher body temperature and more mitochondrial DNA (mtDNA) content in BAT together with an upregulation of the genes related to fatty acid metabolism, thermogenesis and mitochondrial biogenesis (*Ucp1, Ucp2, Prdm16, Cpt1*β*, Pgc-1*α*, Nrf1,* and *Tfam*) [184,185] and a downregulation of *Acc*. These effects have been related to an increased activity of the AMPK in BAT [184].
Thermogenesis can also be induced by a polyphenol-rich green tea extract (PGTE) through a mechanism that depends on adiponectin signaling. The treatment with this extract reversed part of the obesity phenotype in WT mice but no in adiponectin KO mice (AdipoKO). PGTE treatment increased EE, BAT thermogenesis, and promoted browning phenotype in the scWAT of WT mice but these effects were blunted in AdipoKO mice [176].
Some data regarding BAT activation by catechins in humans have also described. Different approaches have been done to demonstrate the effects of green tea extract and caffeine over thermogenesis and body weight [186,187]. Short- and long-term effects have been studied with different results and effectiveness but suggesting that catechins and caffeine may act synergistically to control body weight and induce thermogenesis [175,188]. It has been proposed that the thermogenic response to green tea extracts or its components would be mediated, in BAT, by the direct stimulation of the β-adrenergic receptor (β-AR) cascade through the inhibition of the enzyme catechol-O-methyl transferase (COMT), which degrades catecholamines. On its side, caffeine inhibited PDE, thus inducing a sustained activation of the PKA and its downstream cascade [175].
### *4.3. Flavanols Consumption Induces Energy Expenditure in Peripheral Organs through the Sympathetic Nervous System Activation*
Part of the anti-obesity effects of flavanols have been also related to their influence on sympathetic nervous system (SNS) activity. The SNS activation by green tea catechins (GTC) has been associated to their capacity to inhibit COMT. The inhibition of COMT leads to a prolonged activation of the sympathetically-response and of the β-adrenergic cascade that produces cAMP and the activation of the PKA. Caffeine, in turn, is able to inhibit the PDE activity which drives to a sustained activation of the PKA and its downstream response [175]. Then, both effects act synergistically to increase EE, lipolysis and FAO as has been described in the above sections. Some other mechanisms to describe the anti-obesity effects of flavanols include the modulation of food intake. It has been demonstrated that grape-seed proanthocyanins extract (GSPE) reduced food intake in rats fed a cafeteria diet. These animals showed an activation of the STAT3 protein which upregulated the *pro-opiomelanocortin (Pomc)* expression, thus improving the leptin resistance [189].
Moreover, GSPE supplementation reduced the neuroinflammation and increased the expression of SIRT1 [189]. Flavanols has been described as active molecules against diet-induced neuroinflammation. The induction of neuroinflammation and cognitive impairment in rats by feeding them with a high salt and cholesterol diet (HSCD) could be in part reversed by the treatment with different doses of an enriched-tannins fraction of the Indian fruit *Emblica o*ffi*cinalis.* Treatment with this tannin-enriched gooseberry reversed the HSCD-induced behavioral and memory disturbances, neuronal cell death and reduced the levels of cognitive impairment markers. [190]. In the same way, it has been published that, in mice, EGCG attenuated the neuronal damage and insulin resistance caused by a high fat/high fructose diet (HF/HFD). In this case, EGCG upregulated the IRS-1/AKT and the extracellular-signal-regulated kinase (ERK)/CREB/Brain-derived neurotrophic factor (BDNF) signaling pathways. In longer nutritional interventions with the HF/HFD, EGCG was capable to inhibit the MAPK and NF-κB pathways, as well as the expression of inflammatory mediators, such as TNF-α to reverse the neuroinflammation [191]. Similar results were obtained with EGCG-HFD dietary supplementation. The authors demonstrated that EGCG ameliorated the HFD-induced obesity in part by attenuating hypothalamic inflammation through the inhibition of NF-kB and Signal transducer and activator of transcription 3 (STAT3) phosphorylation, as well as the expression and release of inflammatory cytokines, such as TNF-a, IL-6, and IL-1b [185].
Finally, EGCG alleviated part of the cognitive deficits in a mixed model of familial Alzheimer's disease (AD) and type 2 diabetes mellitus (T2DM). The AD mice model APP/PS1 fed with a HFD showed an improvement in peripheral parameters such as insulin sensitivity but also in central memory deficits when treated with EGCG. Synaptic markers and CREB phosphorylation were increased because of an amelioration in the unfolded protein response (UPR) activity via a downregulation of the activation factor 4 (ATF4) levels. Moreover, EGCG decreased brain amyloid β (Aβ) production and plaque burden by increasing the levels of α-secretase (ADAM10) and reduced the neuroinflammation
in these animals [192]. Finally, green tea extracts can modulate the redox status of the CNS in obese and lean rats [193].
### **5. Flavanones**
Flavanones are a subfamily of flavonoids widely distributed in *citrus* fruits such as grape, tomatoes, and oranges and are the responsible of the bitter taste of their peel and of their juice. As other flavonoids, flavanones show strong health benefits due to its antioxidant activity but also exhibit antiviral, antimicrobial, antiatherogenic, anti-inflammatory antidiabetic and anti-obesity properties [45,48,75,194,195]. Flavanones are mainly found as aglycones or as glycosylated derivatives [196]. The most studied flavanones are hesperidin, naringenin but also eriodyctiol, isosakuranetin and taxifolin.
Hesperidin and its aglycone, hesperetin are found in citrus fruits, such as limes and lemons, tomatoes and cherries and have demonstrated antidiabetic, neuroprotective, antiallergic, anti-inflammatory anticarcinogenic besides their well-established antioxidant capacity [45,197] Naringenin and its aglycone naringin are found to be more abundant in citrus fruits such as grapefruit orange, lemon but also in tomatoes. Naringenin and derivates have been associated with beneficial effects in cardiovascular diseases, osteoporosis, cancer and have showed anti-inflammatory, antiatherogenic, lipid-lowering, neuroprotective, nephroprotective, hepatoprotective and antidiabetic properties [198,199].
### *5.1. Flavanones-Dietary Supplementation Ameliorates the NAFLD in Humans*
Frequently, liver diseases are initiated by oxidative stress, inflammation and lipid accumulation that lead to an excessive production of extracellular matrix followed by a progression to fibrosis, cirrhosis and hepatocellular carcinoma [200]. In the last years, several studies have demonstrated the capacity of different flavanones to ameliorate liver diseases.
To analyze the positives effects of flavanones in liver different approaches have been used. Some authors worked with hepatic chemical-induced damage being the most used the streptozotocin injection to mice or rats [199,201]. Other authors induced liver damage with diet [199] or worked with genetically obese models. Although flavanones demonstrated positive effects in the different approaches, in this review we focused on the experimental approaches where the liver disease has been induced by diet or where genetically obese-models has been used. Experiments with naringenin, hesperidin and eriodyctiol has been done to evaluate the impact of this flavanones' consumption in NAFLD or liver steatosis.
Naringenin has showed the capacity to restore the activities of liver hexokinase, PK, G6Pase and Fructose 1,6-bisphosphatase from rats fed a high fructose diet to levels similar to healthy non-diabetic animals [202]. In this animal model, naringenin also enhanced liver protein tyrosine kinase (PTK), while reduced protein tyrosine phosphatase (PTP) activity [202]. In addition, administration of naringenin to HF/HSD-fed rats increased the protein levels of PPARα, CPT1a and UCP2 [203]. In a similar way, naringenin increased FAO and the AMPK activity in HFD fed mice where ameliorated the metabolic alterations caused by diet [204]. Similar results were obtained in high-fat/high-cholesterol (HFHC) fed Ldlr -/- mice. In lean Ldlr -/- mice, naringenin induced weight loss and reduce calorie intake, enhanced EE and increased hepatic FAO by upregulating *Pgc1*α*, Cpt1a and Hsl*, thus indicating that naringenin is also effective in non-obese models [195]. In HFD fed Ldlr -/-, naringenin increased FAO and reduced lipogenesis. Hepatic *Srebp1c* and *Acox1* mRNA levels were downregulated, while *Fgf21, Pgc1*α*, and Cpt1a* were upregulated by naringenin [205]. Later on, it was published that naringenin prevented obesity, hepatic steatosis, and glucose intolerance in an FGF21-independent way [206]. More recently, it has been described that in obese-mice naringin decreased hepatic liver content (TG and total cholesterol) and activated the AMPK enzyme resulting in reduced expression and protein levels of liver SREBP1C, SREBP2, but increased LDLR. Moreover, these mice showed reduced plasma
levels of proprotein convertase subtilisin/kexin type 9 (PCSK9), leptin, insulin, and LDL-C compared to obese non-treated mice [207].
Besides naringenin, naringin and hesperidin effects in liver have also been evaluated. Hesperidin and naringin supplementation in *db*/*db* and *ob*/*ob* mice regulated hepatic gluconeogenesis and glycolysis, as well as lipid metabolism [208]. Hesperidin stimulated PPARγ, increased the hepatic GK activity and glycogen concentration and reduced the hepatic levels of *Glut2* as well as increased the expression of *Glut4* in WAT [46,208,209]. Moreover, hesperidin prevented hepatic steatosis in western diet-fed rats by preventing the upregulation of lipogenesis-related genes *Srebf1,* and *Scd1* caused by Western diet and the downregulation of *Ppar*α and *Cpt1a* expression and CPT1a protein levels [210]. Most of these effects were blunted when hesperidin is combined with capsaicin [210].
In diet-induced obese mice treated with neohesperidin the expression and secretion of FGF21 and the activity of the AMPK/SIRT1/PGC-1α axis were improved [211]. Treatment with neohesperidin improved the steatotic state (less and smaller lipid droplets), reversed the downregulation of hepatic *Ppar*α levels while increased the levels of the hepatic *Fgf21* expression and its plasma levels. Finally, neohesperidin treatment phosphorylated AMPK, resulting in a rise of the HFD-downregulated proteins SIRT1 and PGC1α [211]. On its side, eriodyctiol has also demonstrated effects on diet-induced obesity. Diet-induced obese mice supplemented with eriodyctiol showed a reduction of hepatic TG, fatty acids and the size and number of lipid droplets accompanied with an increased fecal excretion of cholesterol and fatty acids [212]. It is worth to mention that eriodyctiol decreased the enzymatic activity of malic enzyme (ME), FASN, phosphatide phosphohydrolase (PAP) and downregulated the expression of *Srepb1c, Acc and Fasn* [212]. These data indicate that eriodyctiol improved the hepatic steatosis caused by a HFD by decreasing hepatic lipogenesis and increasing the hepatic FAO. On the other hand, alpinetin, an O-methylated flavanone, improved HFD-induced NAFLD via ameliorating oxidative stress, inflammatory response and lipid metabolism. Alpinetin decreased *Scd1, Fasn, Srebp1c, Lxr*α*, Elovl2* and *Irs1* expressions, and increased PPARα levels [213].
In humans a randomized placebo-controlled, double-blind clinical trial with NAFLD patients shown the effect of hesperidin supplementation [214]. Patients who follow healthy lifestyle habits and supplemented their diet with hesperidin have a significant reduction of ALT, glutamyl-transferase, total cholesterol, hepatic steatosis, C reactive protein and TNFα, proving the scope of hesperidin [214]. One of the possible mechanisms underlying the effects of flavanones on metabolism goes through the FGF21 and AMPK/Sirt1/PGC1α signaling axis.
### *5.2. Flavanones Induce Browning in Adipose Tissue*
As other flavonoids, flavanones can also modulate lipid metabolism in adipose tissue as well as induce browning inWAT, and activate in BAT [166] as well as reduce the characteristic obese-macrophage infiltration in adipose tissue [215].
In HFD fed mice, hesperetin supplementation on its side showed metabolic health effects in adipose tissue, concretely is able to reduce mesenteric adipose weight and decrease leptin levels [216]. In this case, lipid metabolism was not changed nor in liver nor inWAT. On the other hand, a characteristic of obesity is the recruitment of immune cells by adipose tissue that leads to metabolic disorders such as insulin resistance. In a short-term HFD mice model, naringenin can suppress neutrophil and macrophage infiltration into adipose tissue [215]. Concretely it can inhibit the expression of several chemokines like MCP-1 and MCP-3 [217]. Eriodyctiol (ED) supplementation on its side lowered the adiposity in diet-induced obese mice by regulating gene expression. ED-supplemented mice showed reduced weight of all the WAT depots but also a downregulated expression of adipocyte genes involved in lipid uptake (*Cd36*, and *Lpl*) and lipogenesis (*Srebp1, Acc, and Scd1*), an upregulation of the *Ucp1*, with no changes in FAO genes such as *Adrb3, Cpt2, Pgc1*α*, Pgc1*β*,* and *Cox8b* genes [212].
Another beneficial effect of flavanones in adipose tissue is related to EE and thermogenesis. It has been demonstrated that in human white adipocytes and in scWAT a treatment with naringenin increased the expression of genes associated with thermogenesis and FAO, including *Atgl* and *Ucp1* as
well as *Pgc1*α and *Pgc1*β that can mediate the PPARδ-dependent transcriptional responses involved in mitochondrial biogenesis and uncoupling phenotype. Moreover, naringenin administration increased the expression of insulin sensitivity-related proteins such as *Glut4*, *adiponectin*, and *Chrebp* [218]. These data indicate that naringenin may promote the conversion of human WAT to a brown/beige adipose tissue. Similarly, in HFD-obese mouse model, the induction of brown-like adipocyte formation on WAT was described by supplementing the diet with a flavanones-rich extract from *Citrus reticulata* [219]. The main phytochemical components of a water extraction of *Citrus reticulata* in were synephrine, narirutin, hesperidin, nobiletin, and tangeretin. Among flavanones, citrus also contain synephrine that is an alkaloid which binds to β3AR in adipose tissue promoting lipolysis and thermogenesis [220]. Dietary supplementation with this citrus extract reduced body weight gain, epididymal fat weight, fasting blood glucose, serum levels of TG and total cholesterol, and lipid accumulation in liver and WAT as well as activated FAO and induced the browning phenotype [219]. These animals showed increased levels of *Ucp1* in the iWAT and an upregulation of *Prdm16*, *transmembrane protein 26 (Tmem26), cluster of di*ff*erentiation 137 (CD137),* and *Cidea* [219].
In the same way it has been published that hesperidin induced browning in retroperitoneal WAT (rWAT) but not in iWAT of Western diet-fed rats. Hesperidin decreased the size of adipocytes and induced the formation of multilocular and positive-UCP1 and CIDEA brown-like adipocytes. Besides the induction browning, hesperidin also enhanced the expression of *Ucp1* in BAT [221]. In contrast, it has been recently published a study where not hesperidin but its monoglycosyl has the capacity to induce brown-like adipocyte formation in HFD-fed mice [222]. In this case, α-monoglucosyl hesperidin increased EE and reduced body fat accumulation by stimulating the browning phenotype in the iWAT. iWAT adipocytes of supplemented mice exhibited a multilocular phenotype and were UCP1-positive cells. The iWAT of these animals also showed increased levels of COXIV. No effects were observed in BAT nor in other WAT depots [222].
In a human randomized double-blind placebo-controlled trial with moderate high BMI subjects, it's shown that glycosylated hesperidin decreased significantly abdominal and subcutaneous fat area when is supplemented with caffeine [223].
### *5.3. Flavanones Are Neuroprotective against Several CNS Injuries*
There is low information about the effects of flavanones on CNS to combat obesity. It has been demonstrated that quercetin, naringenin and berberine can modulate glucose homeostasis in the brain of STZ-induced diabetic rats through the regulation of glucose transporters and other key components of insulin signaling pathway [224].
Most of the studies that show the neuroprotective role of flavanones have been performed using animal with CNS-induced injuries. In a rat model of global cerebral ischemia reperfusion (I/R), pinocembrin (a honey flavanone) exerted antioxidant, anti-inflammatory and anti-apoptotic effects. [225] as well as inhibited autophagy on the hippocampus [226]. Moreover, naringenin and eriodyctiol exert effects in ischemic stroke, promoting cortical cell proliferation, inhibiting apoptosis and reducing oxidative stress in rodent models [227,228]. In a similar way, the induction of neurotoxicity by lipopolysaccharide (LPS) administration in mice can be ameliorated by the coadministration of hesperetin or naringenin that reduced the expression of inflammatory cytokines, attenuated the generation of reactive oxygen species/lipid peroxidation and enhanced the antioxidant capacity in CNS [229,230]. Furthermore, hesperetin enhanced synaptic integrity, cognition and memory processes by increasing the levels p-CREB, postsynaptic density protein-95 (PSD-95) and syntaxin proteins [229] and naringenin decreased the acetylcholinesterase (AChe) activity [230]. Other mental stresses such as social defeat stress, depression and autistic-like behaviors can also be counteract with flavanones in rodent models [231–233]. Hesperidin and naringenin have demonstrated positive effects by increasing the resilience through a reduction in the levels of interleukins and corticosterone thus suppressing the chronic inflammation caused by kynurenine pathway related to depression [234] and inhibiting the AChe activity, the oxidative stress as well as neuroinflammation [235].
### **6. Flavonols**
Flavonols are widely distributed in plants and are present as minor compound in many polyphenol-rich foods. Their synthesis is stimulated by light and they accumulate in the skin of fruits and vegetables being absent in the flesh. The main dietetic flavonols are quercetin, kaempferol, isorhamnetin, fisetin, and myricetin [48,236,237].
Quercetin is found in capers, lovage (*Levisticum o*ffi*cinale*) apples, seeds of tomatoes, berries, red onions, grapes, cherries, broccoli, pepper, coriander, citrus fruits, fennel, flowers, leaves pepper and teas (*Camellia sinensis*) and it is the skeleton of other flavonoids, such as hesperidin, naringenin, and rutin. Rutin, rutoside or sophorin are the glycosylated form of quercetin and can be extracted from buckwheat, oranges, grapes, lemons, limes, peaches, and berries [238]. Kaempferol is abundant in apples, grapes, onions, tomatoes, teas, potatoes, beans, broccoli, spinaches, and some edible berries. Isorhamnetin is commonly found in medicinal plants such as ginko (*Ginkgo biloba)*, sea-buckthorn *(Hippophae rhamnoides*) and *Oenanthe javanica.* Myricetin is found in teas, wines, berries, fruits and vegetables. Fisetin is abundant in apples, grapes, persimmon, cucumber, onions and strawberries. Finally, morin is present in *Prunus dulcis, Chlorophora tinctoria L*., and fruits such as guava and figs [45].
As other groups of flavonoids, flavonols have shown healthy effects. They exhibit anticarcinogenic, anti-inflammatory, and antioxidant activities but also anti-obesity and antidiabetic properties in animal models and in humans where flavonols consumption has been associated to a lower risk of type 2 diabetes [43,236–243]. Some flavonols inhibited carbohydrate absorption thus lowering postprandial blood glucose mainly through the inhibition of the α-glucosidase activity but also by inhibiting glucose transporters (GLUT2, SGLT1) or other enzymes such as maltase or saccharase [236]. Finally, a combination of quercetin and resveratrol have shown the capacity to reduce obesity in HFD-fed rats by modulating gut microbiota [244].
Due to the high number of publications and previous reviews [45,48,238], in the present work only the most recent data have been included.
### *6.1. Flavonols Exert Beneficial E*ff*ects on Lipid Steatosis by Regulating Lipid Metabolism, Inflammation and Oxidative Stress*
Quercetin enhanced hepatic insulin sensitivity and reduced liver fat content and ameliorated hepatic steatosis [245]. Quercetin diminished the mRNA and protein levels of CD36 and MSR1, upregulated the levels of LC3II and downregulated p62 and mTOR thus suggesting an autophagy lysosomal degradation as the potential hepatoprotective mechanism of quercetin [245]. From another point of view the effects and mechanisms of quercetin against NAFLD were analyzed through a metabolomic approach [246]. Treatment with quercetin decreased AST and ALT levels in serum and reduced lipid droplets and hepatocyte swelling in rats fed a high fat/high sucrose diet. A metabolomic analysis indicated that quercetin modified fatty acid- inflammation- and oxidative stress-related metabolites among others. In this case, the effects of quercetin were more evident in 30-day NAFLD induction than in 50 days, thus indicating that dietary quercetin may be beneficial in early stages of NAFLD development [246]. Besides the effects of quercetin alone there are several studies where quercetin is used in combination with other compounds. The beneficial effects of quercetin in NAFLD development increased synergistically when quercetin is administered within benifuuki, a tea that contains EGCG. Both compounds administered to rats fed high fat/high cholesterol diet were more effective to downregulate *Fasn* and *Scd1* showing higher effects on their lipid-lowering effects alone [247]. In a similar way, the combination of quercetin with resveratrol ameliorated fatty liver in rats by improving the antioxidant capacity of the liver [248]. Finally, a combination of borage seed oil (as a source of linoleic (18:2n-6; LA) and gamma-linolenic (18:3n-6; GLA) acids and quercetin improved liver steatosis in obese rats [249].
On its side, isoquercetin (IQ), a glucoside derivative of quercetin has demonstrated beneficial effects in NAFLD by improving hepatic lipid accumulation via an AMPK dependent way in HFD-induced NAFLD rats [250]. Concretely, IQ treatment enhanced the phosphorylation of AMPK and ACC and reversed the downregulation of liver kinase β1 (LKβ1) and Calcium/calmodulin-dependent protein kinase kinase-1 (CaMKK1) caused by HFD. The activation of AMPK modulated the expression of lipogenic and lipolytic genes, such as *Fasn, Srebp1c, Ppar*γ *and Cpt1a*. Moreover, IQ supplementation upregulated PPARα and downregulated nuclear factor-kB (NF-kB) protein levels [250].
As quercetin, kaempferol is also able to reduce lipid accumulation in liver of obese rodent models. In dyslipidemia-induced mice, kaempferol inhibited PKB (Akt) and SREBP-1 activities and blocked the Akt/mTOR pathway, thus inducing hepatic autophagy and decreasing hepatic lipid content [251]. Similarly, in ApoE deficient mice fed with a HFD, kaempferol attenuated metabolic syndrome via interacting with LXR receptors and inhibiting posttranslational activation of SREBP-1. Both effects contributed to the reduction of plasma and serum TG [252].
Other flavonols with positive effect in the liver are fisetin, dihydromyricetin or rutin. Obese rats fed with a high fat/high sucrose diet and supplemented with fisetin showed a decreased in body weight and hepatic lipid content as well as an improvement in the lipid profile (low levels of TG, total cholesterol, LDL) and liver functionality (reduced levels of ALT and AST). The hepatic nuclear receptor 4α (HNF4α) has been pointed out as the key factor in the hepatic effects of fisetin. Fisetin upregulated *Hnf4a* gene expression, increased nuclear lipin-1 levels. Moreover, fisetin promoted FAO, diminished FASN activity, enhanced hepatic antioxidant capacity and decreased the hepatic poly (ADP-ribose) polymerase 1 (PARP1) activity, a DNA repair enzyme, and thioredoxin-interacting protein (TXNIP) that is important for maintaining the redox status [253]. Through the regulation of SIRT3 signaling, dihydromyricetin have showed the ability to ameliorate NAFLD in HFD-fed mice. Dihydromyricetin increased *Sirt3* expression via activation of the AMPK/PGC1α/estrogen-related receptor α (ERRα) cascade thus improving mitochondrial capacity and restored redox homeostasis [254]. In a similar way, rutin lowered TG content and the abundance of lipid droplets in NAFLD-induced HFD fed mice. Rutin treatment restored the expression of *Ppar*α and *Cpt1a* and *Cpt2*, while downregulated *Srebp-1c*, *diglyceride acyltransferase 1 and 2 (Dgat-1 and 2* and *Acc*. These effects enhanced FAO and diminished lipid synthesis. In addition, rutin repressed the autophagy in the liver [255]. On its side, the rutin derivate, troxerutin (TRX), has also demonstrated effectiveness against metabolic disorders in a rat model of hereditary hypertriglyceridemia (HHTg) non-obese model of MetS [256]. The treatment with TRX lowered the levels of hepatic cholesterol and reduced the expression of cholesterol and lipid synthesis genes (*Hydroxymethylglutaryl-CoA reductase* (*Hmgcr), Srebp2* and *Scd1*) as well as decreased lipoperoxidation and increased the activity of antioxidant enzymes [256]. Moreover, these animals exhibited higher levels of adiponectin in serum [256].
Besides the effects of flavonols by itself, favonols-rich extracts have also been tested in fatty liver-associated diseases. A *Sicyos angulatus* extract that contains kaempferol as the main flavonol administered to a HFD-induced obese mice lowered plasma levels of ALT and AST and the hepatic lipid content. The *Sicyos angulatus* extract impacted on lipid metabolism by repressing the expression of genes related to fatty acid and TG synthesis (*Acc1, Fasn Scd1* and *Dgat*) and of the key transcription factors that regulate lipogenesis (*Srebp-1c* and *Ppar*γ) [257]. Another source of kaempferol, quercetin and derivates is Sanglan Tea (SLT), a Chinese medicine-based formulation consumed for the effective management of obesity-associated complications. It has been demonstrated that dietary SLT supplementation prevented body weight gain and fatty liver and ameliorated insulin resistance in HFD-induced obese mice. SLT improved the serum lipid profile (lower levels of TG, Total cholesterol and LDL) and reduced the ALT and AST circulating levels. The liver of these animals displayed less lipid droplets and a downregulation of the lipogenic genes *(Lxr*α*, Fasn, Acacb, Srebf-1,* and *Scd1)* and the adipogenesis-related genes (*Ppar*γ*, C*/*ebp*α and *Ap2*) that are induced under HFD [258].
In a similar way, the flower of *Prunus persica* commonly known as peach blossom has demonstrated that capacity to reduce body weight, abdominal fat mass, serum glucose, ALT, AST, and liver and spleen weights compared to a HFD fed mice. This flower is rich in flavonoids and phenolic phytochemicals with chlorogenic acid, kaempferol, quercetin and its derivatives as its major compounds. The supplementation with this flower suppressed hepatic expression of lipogenic genes (*Scd1, Scd2,*
*Fasn*) and increased the mRNA levels of FAO genes (*Cpt1a*), thus modifying he lipid metabolism in HFD-fed mice [259]. Furthermore, a mulberry leaf powder also showed effects on liver gene expression in a mice model of hepatic steatosis induced by a western diet. Liver weight, plasma TG and liver enzymes ALT and AST were reduced in treated-animals. A global hepatic gene expression analysis revealed that supplemented mice displayed a downregulation in inflammation-related genes and an upregulation in liver regeneration-related genes [260]. Finally, a 70% ethanol extract from leaves of *Moringa oleifera (MO)* that contains different flavonols and flavones such as quercetin and kaempferol and their derivates. reduced glucose and insulin but also the total cholesterol, TG and LDL serum and increased the HDL in high-fat diet obese rats as well as downregulated hepatic expression of *Fasn* and *Hmgcr* [261].
Through a network pharmacological approach Nie et al. [262] highlighted that Chaihu shugan powder (CSP) may exert its beneficial effects against NAFLD through the interaction of its main compounds with nuclear receptors. Through a molecular docking approach, they screened PPARγ, FXR, PPARα, RARα and PPARδ and quercetin, kaempferol, naringenin, isorhamnetin and nobiletin interactions. To confirm the results of docking, an in vivo approach was done using NAFLD-induced rats. The NAFLD-induced rats treated with CSP exhibited ameliorated effects in body weight, hepatic histopathology and serum and liver lipids. Moreover, the mRNA levels of *Ppar*γ*, FXR, Ppar*α and *Rar*α were modified suggesting nuclear receptors regulation as a potential molecular mechanism underlying the effects of CSP [262].
Adiponectin signaling and AMPK activation have been also pointed out as possible mechanisms underlying the effects of flavonols in the liver. An extract of black soybean leaves (EBL), which mainly contains quercetin glycosides and isorhamnetin glycosides was administered to HFD-fed mice. EBL supplementation reduced body weight, fasting glucose, TG, total cholesterol and non-esterified fatty acid levels as well as hepatic steatosis. EBL supplementation increased the levels of adiponectin and the expression of adiponectin-receptors in the liver (AdipoR1 and AdipoR2) thus restoring adiponectin signaling pathway [263]. Downstream of the adiponectin signaling there is the activation of AMPK and FAO, the suppression of fatty acid synthesis and the improvement of insulin signaling [264]. Moreover, the mRNA levels of *Pgc1, Ppar*α*, Ppar*δ*, Ppar*γ*, Acc, Fasn, Cpt1a, Glut2, FoxO1 and Irs1* were partially or totally normalized in HFD-EBL-supplemented animals [263].
Finally, it has been described that part of the mechanisms involving the hepatic beneficial effects of flavonols may be mediated by gut microbiota. An experimental approach of gut microbiota transplantation revealed a gut–liver axis where the *Akkermansia* genus have a key role on the quercetin protecting effects against obesity-associated NAFLD development. [247]. In a similar way, kaempferol blunted part of the effects of HFD in gut microbiota diversity. HFD fed mice displayed a reduced microbial diversity that it is mostly reversed by kaempferol [265]. Furthermore, IQ combined with inulin attenuated weight gain, improved glucose tolerance and insulin sensitivity and reduced lipid accumulation in the liver, adipocyte hypertrophy in WAT and diminished the circulating levels of leptin in HFD-fed mice probably through the modulation of gut microbiota [266].
### *6.2. Flavonols Impact on WAT Where They Modulate Lipid Metabolism and Induce Browning*
Several studies with animal models showed that flavonols can protect mice or rats from HFD obesity by reducing body weight gain and lipid accumulation in WAT via reducing inflammation, modifying lipid metabolism, increasing EE, inducing browning of WAT and activating BAT [174,242,267–269].
Quercetin and quercetin-rich red onion (ROE) ameliorated diet-induced WAT expansion and inflammation in HFD-fed mice [270]. Quercetin and ROE ameliorated adipocyte size and number compared to HFD fed mice in WAT depots and induced a multilocular phenotype typical of BAT [270]. Moreover, quercetin and ROE diminished the HFD-increased levels of leptin. Besides its impact on adipose tissue phenotype, quercetin and ROE supplementation also attenuated the inflammatory profile induced by HFD in WAT [270]. Similarly, a quercetin-rich supplement administered to diet-induced obese rats decreased body fat and adipocyte size of the perirenal WAT as well as increased adiponectin
circulating levels [271]. Quercetin-rich supplement attenuated the upregulation of genes related to lipid synthesis such as *Acc, Fasn*, *HMG-CoA reductase*, *Lpl, Ap2,* and *Fatty acid transporter protein 1 (Fatp1)* caused by HFD; and upregulated the HFD-downregulated genes such as *Atgl, Hsl, Ampk, Acox, Ppar*α*,* and *Cpt1a* [271]. In diet-induced obese mice quercetin administration decreased plasma TG levels without affecting food intake, body composition, or EE [272]. Quercetin enhanced the uptake of [3H]-oleate derived from labeled lipoprotein-like particles in the scWAT [272]. On the other side Perdicaro et al. demonstrated that quercetin attenuated adipose tissue hypertrophy, reduced the adipocyte size but activated the adipogenesis in HFD-fed rats. Quercetin supplemented rats showed increased levels of angiogenic (*Vascular endothelial growth factor 1* and *2 (Vegf1, Vegf2*) and adipogenic (*Pparg* and *C*/*ebpa*) markers but also mitigated inflammation, and reticulum stress [273].
Together with their capacity to modulate lipid metabolism, flavonols are also able to induce browning in WAT depots. Quercetin treatment increased the expression of *Ucp1, Pgc1*α and *Elovl3* in WAT [272,274]. In a similar way, the administration of onion peel extract (rich in quercetin) to HFD-fed mice upregulated markers of BAT (*Prdm16, Pgc1*α*, Ucp1, Fgf21, Cidea*) in perirenal and scWAT [275]. It has been described that the induction of browning was mediated at least in part through the activation of the AMPK and the SIRT1 or via sympathetic stimulation. The quercetin-supplemented HFD-fed mice displayed higher levels of plasma norepinephrine and of PKA protein levels in scWAT [274]. Besides the activation of PKA signaling, it has been described that quercetin also increased SIRT1 protein levels and pAMPK in visceral WAT [276]. Although most of the studies showed positive effects of quercetin, this flavonol did not induce significant effects on the adipose tissue weights of rats fed an obesogenic diet except when combined with resveratrol (RSV). The treatment with quercetin and RSV but not with just quercetin or RSV promoted multilocular UCP1-positive adipocytes that also displayed increased levels of browning markers (*Cidea, bone morphogenic protein 4 (Bmp4), Homeobox C9 (Hoxc9), Solute Carrier Family 27 Member 1 (Slc27a1), Tmem26* and *proton*/*amino acid symporter (Pat2))* and genes related to catabolic pathways (*Atgl* and *ATP synthase subunit delta (Atp5d))* in perirenal WAT. Regarding BAT, the supplementation with RSV and quercetin upregulated *Cidea* and *Ucp1* expression, thus indicating more thermogenic capacity in this tissue [277].
It is worth to mention that quercetin effectiveness is specie dependent. Studies in rats usually showed more effects than in mice whilst in humans the results are still unclear. In rodent models the levels of quercetin reached after its administration are higher than in humans [269]. Similar to quercetin, isoquercetin (IQ), a quercetin glycoside with greater bioavailability than quercetin, also exerts positive effects in WAT. In normal diet-fed mice IQ supplementation decreased WAT weight and increased pAMPK levels in WAT as well as in liver and muscle. Moreover, IQ reduced the expression of *Ppar*γ*, C*/*ebp*α*, C*/*ebp*β and *Srebp1* whilst increased the expression of *Ucp2, Pgc1*α*, Prdm16, Sirt1* and *Cpt1a* in WAT, suggesting less adipogenesis, enhanced FAO and browning [278].
On its side, rutin administration to db/db mice and diet-induced mice reduced body weight gain and improved adiposity (smaller lipid droplets) mainly by increasing EE [279]. These animals exhibited higher core temperature when submitted to a cold environment indicating enhanced BAT activity. Rutin-treated animals overexpressed BAT markers (*Ucp1, Cidea, Prdm16*), FAO-related genes (*Cpt1a, Mcad, Ppar*α *and Pgc1*α), mitochondrial biogenic transcription factors (*tfam, Nrf1, Nrf2)* and more copies of mitochondrial DNA in BAT [279]. Besides BAT, rutin also affected scWAT, where induces browning (upregulation of BAT-specific genes, including *Ucp1, Pgc1*α*, Pgc1*β*, Cpt1a, Ppar*α*, Tfam, Nrf1* and *Nrf2*...) [279]. The molecular mechanism underlying these effects may go through the Sirt1 activation. It has been demonstrated that rutin was able to directly bind to Sirt1 protein and activate the SIRT/PGC1α/NRF2/Tfam signaling pathway [279]. On the other hand, rutin combined with exercise (treadmill running) in diet-induced obese mice increased the mRNA levels of *adiponectin*, the protein levels of PPARγ, the binding immunoglobulin protein (BIP), and the phosphorylated form of c-Jun terminal quinase (JNK) and reduced disulfide-bond A oxidoreductase-like protein (DsbA-L). These profile indicated an improvement on the ER stress and on adipose tissue functionality [280].
When instead of flavonols, plant extracts were used similar effects were observed. A 70% ethanol extract of *Moringa oleifera (MO)* that mainly contains quercetin, kaempferol and their derivates induced the expression of *Glut4*, *adiponectin, omentin* and upregulated *Ppar*α and *melanocortin-4 receptor (MC4R)* on the WAT of diet-induced obese rats. [261]. *Cuscuta pedicellata* and some of its isolated compounds, including kaempferol, quercetin and some derivates were suggested to have an anti-obesity effect in HFD-fed rats. Supplemented animals showed a reduction in HOMA-IR and oxidative stress as well as exhibited an upregulation of *Ucp1* and*Cpt1a* expression in BAT [281]. Finally, through a high-throughput metabolomic approach it has been described that the consumption of a hawthorn ethanol extract that contains chlorogenic acid, hyperoside, isoquercetin, rutin, vitexin, quercetin, and apigenin affected several metabolic pathways including: fatty acid biosynthesis, galactose metabolism, biosynthesis of unsaturated fatty acids, arginine and proline metabolism, alanine, aspartate and glutamate metabolism, glycerolipid metabolism and steroid biosynthesis [282].
### *6.3. Flavonols: Neuroprotection in Neurodegenerative Diseases*
Flavonols have shown neuroprotective effects in neurodegenerative diseases. Quercetin, rutin and some other flavonols have exhibited positive effects against pathologies such as Alzheimer's Disease (AD), Parkinson's disease, Huntington's Disease, multiple sclerosis, brain ischemic injury, epilepsy neurotoxins but also for aging cognitive alterations [238,283–288]. Furthermore, flavonols have also demonstrated beneficial effects in the CNS alterations caused by HFD.
It is well-known that HFD induces oxidative stress in brain that may lead to neurodegenerative diseases. In HFD-fed mice, quercetin ameliorated the cognitive and memory impairment and enhanced the expression of *phosphatidylinositol-4,5-bisphosphate 3-kinase (PI3K), PKB*/*Akt, Creb,* and *brain-derived neurotrophic factor (Bdnf)* [289]. In a similar way, in HFD-fed mice, *Acer okamotoanum* and its main bioactive compound isoquercitin improved cognitive function by inhibiting the ROS production, the lipid peroxidation and nitric oxide formation, thus reducing oxidative stress [290]. Furthermore, it has been described that obesity induces hypothalamic inflammation and activates microglia. In diet-induced obese mice, quercetin supplementation reduced the levels of inflammatory cytokines and microglia activation markers in the hypothalamus [291]. Quercetin has also showed positive effects in streptozotocin (STZ)-induced AD rats where improved memory impairment and the anxiogenic-like behavior induced by STZ. In these rats, quercetin prevented the acetylcholinesterase (AChE) overactivity and the increased malondialdehyde levels caused by STZ [292]. Finally, quercetin showed capacity to modulate several kinases signaling cascades involved in synaptic plasticity such as the PI3K/Akt, protein kinase C (PKC) and mitogen-activated protein kinase (MAPK) [293].
### **7. Isoflavones**
Isoflavones, also known as phytoestrogens, are flavonoids with a limited distribution in plant kingdom. They are found in leguminous plants such as soybean, kudzu, red clover, fava beans, alfalfa, chickpeas or peanuts but also soy-based foods (tofu, soymilk, miso ... ) and some pants such the *Puerariae* genus [42,294]. Genistein and daidzein are the most representative dietary isoflavones.
Although there are several human clinical studies studying soy isoflavone consumption and diabetes the data obtained are not conclusive. Some evidence suggests that long-term intake of isoflavones may improve insulin resistance in type 2 diabetic patients and have anti-obesity effects [295–299]. In animal studies, isoflavones have showed antidiabetic and anti-obesity activities [45,236,297,300]. The beneficial effects of isoflavones include the improvement of insulin sensitivity, lipid profile and adiposity [45,49,301–303].
### *7.1. Isoflavones Reduced H Steatosis by Modulating Lipid Metabolism*
Like many of the other flavonoids, isoflavones also exert an hepatoprotective action [49]. A recent publication using data of the National Health and Nutrition Examination Survey from 1999 to 2010 in the USA describes an inverse correlation between urinary genistein levels and serum ALT levels in males but not in females [304]. On the other hand, in NAFLD-rodent models, genistein supplementation decreased fat accumulation, inflammation, hepatic steatosis and liver fibrosis in animal models and in humans [302]. These effects on hepatic steatosis have been described both in short- and long-term interventions [305].
One of the mechanisms proposed is the blockade of aldose reductase (AR)/polyol pathway. It has been described that some isoflavones are AR inhibitors. The inhibition of the AR/polyol pathway reduces fructose production and hepatic fat accumulation in high glucose diets as well as improved PPARα activity and enhanced FAO, thus attenuating liver steatosis in HFD-obese models [306]. Moreover, the blockade of AR/polyol pathway reduced the CYP2E1-mediated oxidative stress [306]. Other mechanism suggested for isoflavones is the downregulation of PPARγ and fat-specific protein 27 (FSP27) together with a reduction of fatty acid synthesis and increased lipolysis [307]. This mechanism was described in female rats fed with a 20% casein-diet and supplemented with soy isoflavones [307].
Effects via the activation of AMPK has been also described for genistein [308,309]. Hepatic activation of AMPK drives to an inhibition of cholesterol and fatty acid synthesis and an enhancement of FAO [310]. In high fat/high sucrose-fed rats, genistein improved lipid metabolism and ameliorated hepatic lipid accumulation. P-AMPK and p-ACC were increased while SREBP1 protein levels were decreased. Moreover, genistein downregulated the expression of *Fasn, glycerol-3-phosphate acyltransferase (Gpat)* as well as upregulated *Ppar*α*, Cpt1a* and *Acox* [309]. A similar effect on NAFLD has been described with Puerarin, a major bioactive isoflavone compound isolated from the roots of the *Pueraria lobata.* Puerarin attenuated NAFLD development in high fat/high sucrose-fed mice via the activation of the Poly(ADP-ribose) polymerase 1 (PARP-1)/PI3K/Akt signaling pathway and lately the improvement of the mitochondrial function [311]. In HFD-obese mice, puerarin reduced TG, total cholesterol and leptin serum levels as well as decreased the hepatic lipid content. Puerarin inactivated FASN and activated AMPK, CPT and HSL as well as increased the protein levels of PPARγ. These data indicated that puerarin regulated lipid metabolism by reducing lipid synthesis and enhancing lipid consumption [312].
Positive effects on NAFLD has been also observed by combining soluble soybean polysaccharides and genistein. This combination increased the bioavailability of genistein and administered to HFD-fed mice prevented weight gain, oxidative stress inflammation and dyslipidemia. These effects on lipid profile have been related to an activation of AMPK and PPARα/PPARγ pathways and changes in the mRNA levels of *Fasn, Acc, Srebp1c* and *adipose di*ff*erentiation-related protein (Adrp)* [313].
Besides genistein some of its derivatives are also active. Sophoricoside, a genistein derivate isolated from the *Sophora japonica* L, has been tested in high fructose-fed mice. Administration of sophoricoside diminished body and liver weight as well as reduced hepatic cholesterol and TG and serum levels of ALT, AST and LDL whilst increased the levels of circulating HDL. Moreover, the livers of treated-mice displayed a better inflammatory profile and an increased antioxidant capacity [314]. Calycosin, an o-methylated isoflavone showed positive effects against NAFLD-induced in HFD-fed mice. Calycosin improved insulin sensitivity, decreased the levels of ALT and AST and increased the levels of adiponectin. In the liver, calycosin blocked gluconeogenesis and lipogenesis by suppressing PEPCK G6Pase, SREBP1c and FASN, as well as induced the expression of *Gsk3*β*, Glut4,* increased the phosphorylation of Irs1 and Irs2 and activated farnesoid X receptor (FXR) [315].
Similar to isolated compounds, soy isoflavones (that includes genistein, daidzein and glycitein) or a soy protein preparation also reverted hepatic steatosis when administered to obese female Zucker or HFD-obese rats. Soy isoflavones reduced hepatic lipid accumulation, improved serum levels of ALT and downregulated *Srebp1c* and *Fasn* levels as well as increased the protein levels of PPARα indicating less lipogenesis and more FAO [316]. In a similar way, the intake of soy protein with isoflavones decreased the liver steatosis, reduced the levels of AST and ALT and increased the levels of leptin in female Zucker obese rats [305].
Apart from the effects of isoflavones on lipid metabolism they also exhibit anti-inflammatory properties. Genistein protected against NAFLD by targeting the arachidonic acid cascade that is responsible for the chronic inflammation [317]. Genistein supplementation to HFD-fed mice blocked the synthesis of ciclooxigeanse-1 activity and thromboxane A2 [317]. Other mechanism to explain the anti-inflammatory effect of genistein is the promotion of miR-451 [318]. In humans a randomized controlled trial described that genistein supplementation improved the inflammatory state in NAFLD patients [319].
### *7.2. Isoflavones Ameliorate the Weight Gain in Diet-Induced Obesity Models and Improve Lipid Metabolism in Adipose Tissue*
It has been widely described that isoflavones are able to control food satiety and appetite, to ameliorate the body weight gain and fat accumulation in rodent models of obesity, to modulate fatty acid metabolism and to induce browning and BAT activation which make its use in nutritional interventions as a promising approach for weight management therapies [269]. Isoflavones reach and affect adipose tissue as it was demonstrated through a whole-transcriptome microarray analysis of the perigonadal WAT from mice fed either control diet or a soybean extract diet containing a genistein/daidzein mix. This study described the impact of soy isoflavones on adipose tissue describing 437 downregulated genes and 546 upregulated [320].
In HFD-fed rats, soy isoflavones attenuated diet-induced obesity mainly by reducing the visceral WAT depot (lower hypertrophy and less lipid accumulation). Soy isoflavones supplementation downregulated fat synthesis (reduced SREBP1 protein levels) and upregulated lipolysis (increased ATGL protein levels) in visceral WAT via the activation of AMPK and the inhibition of SREBP1 [321]. In a similar way, 6,8-diprenylgenistein (DPG), a major isoflavone of *Cudrania tricuspidata* fruits decreased the body weight of HFD-induced obese mice at least in part by the suppression of de novo lipogenesis via the AMPK activation [322]. This isoflavone reduced the expression of lipogenic genes by regulating Pparγ and C/EBPα transcriptional activity as well as leptin and adiponectin levels. DPG also regulated ACC and HMGCR [322].
Isoflavones are also present in fermented soy products. The heathy properties of these products have been also evaluated. Fermented soybean meal (SBM) administered to HFD-fed rats showed positive effects on the obese profile of these animals. The body weight gain, as well as weights of abdominal and epididymal fat were reduced. Also, the lipid profile was improved. Supplemented rats exhibited lower levels of TG, total cholesterol and LDL and higher levels of HDL compared to HFD-non supplemented rats. Moreover, in WAT, there were a decrease on the hepatic lipogenesis (downregulation of *Fasn* and *Acc)* and an increase on lipolysis (upregulation of *Lpl*) [323].
Besides their effects on lipid metabolism, isoflavones also induce browning and BAT activation [166]. Genistein administration to HFD-fed mice reduced body weight gain and scWAT mass and induced the expression of *Ucp1* and *Cidea* in WAT, indicating a browning phenotype [324]. Genistein may induce the browning phenotype by a direct upregulation of *Ucp1* expression or through an indirect pathway that would imply irisin signaling. Irisin is a myokine that induces the expression of *Ucp1* and *Tmem26* in preadipocytes [325]. This indirect mechanism describes an induction of the PGC-1α/FNDC5 pathway in skeletal muscle that lead to an increase of irisin production and secretion [325].
Formononetin and puerarin also modulate adipogenesis and thermogenesis. Formononetin attenuated visceral fat accumulation and increased EE in HFD-fed mice [326,327]. In vitro, this isoflavone downregulated P*par*γ*, C*/*ebp*α and *Srebp1* probably via AMPK/β-catenin signal transduction pathway that drove its antiadipogenic effect [326]. Moreover, formononetin induced *Ucp1* expression in primary culture of mouse adipocytes [327]. In a similar way, *Puerariae lobata* root extracts (PLR) activated browning in iWAT and regulated BAT activity [328]. PLR treatment caused weight loss and improved glucose metabolism in diet-induced obese mice as well as increased EE. In BAT, PLR upregulated *Ucp1* expression (but no other thermogenic markers) and in iWAT
induced the expression of BAT markers (*Ucp1, Ppar*γ*1, Ppar*γ*2* and *Ppar*α*),* thus indicating a brown-like phenotype [328].
Several studies focused on describing the mechanisms underlying the isoflavones' effects have been performed in ovariectomized mice or rats. These models mimic menopausal stage in humans and are useful to analyze the potential role of isoflavones to counteract the increase of the adipose tissue that takes place during this period of life. In these rodent models, isoflavones exert positive effects on body weight gain and food intake as well as in fat pats enlargement [297]. In HFD-fed ovariectomized rats the administration of genistein decreased the body weight gain, improved insulin sensitivity and reduced plasma TG and cholesterol [329]. In liver, genistein blocked the lipogenic pathway by inhibiting p-ACC, SREBP-1, FASN and CD36 proteins. In retroperitoneal WAT, genistein diminished adiposity and adipocyte hypertrophy, inflammatory phenotype and induced browning. In iWAT, genistein-supplemented rats exhibited higher levels of UCP1, PRDM16, PGC-1α and CIDEA proteins and *Ppargc1a* and *Ucp-1* mRNAs [329]. Furthermore, isoflavones supplementation can modulate the metabolic effects of estradiol treatments in ovariectomized rats [330]. Finally, calycosin has demonstrated positive effects perivascular adipose tissue of obese mice. Through the adiponectin/AMPK/ endothelial nitric oxide synthase (eNOS) pathway, calycosin is able to restore at least in part the perivascular adipose tissue functionality [331].
### *7.3. Isoflavones Have Become Engaging Flavonoids in Neuronal Diseases due to Their Estrogenic-Like Structure and Its High Antioxidant Capacity*
Obesity is a risk factor for neurodegenerative diseases essentially because it causes the neuroinflammation and oxidative stress. Isoflavones can ameliorate part of these effects as well as affect food intake and feeding behavior.
It has been described that daidzein administered to HFD-fed rats reduced food intake and attenuated body weight gain as well as improved glucose tolerance, adiponectin and leptin levels and increased the 17b-estradiol. In rat hippocampus, daidzein enhanced cell proliferation and reduced apoptosis and gliosis, thus exerting a neuroprotective effect against the brain injuries caused by diet [332]. On the other side, doenjang, a Korean traditional fermented soybean pastry alleviated hippocampal neuronal loss and enhanced cell proliferation in HFD-fed mice as well as reduced oxidative stress markers (less oxidative metabolites and lower levels of oxidative stress- and neuroinflammation-related genes). Dietary doenjang reduced Aβ and tau phosphorylation [333]. Furthermore, genistein has shown the capacity to improve metabolism and induce browning via hypothalamus gene expression regulation. Through a transcriptome analysis it was identified that the hypothalamic expression of *urocortin 3 (Ucn3), decidual protein induced by progesterone (Depp*), and *stanniocalcin1 (Stc1)* correlated with the browning markers in WAT and with insulin sensitivity [324].
Regarding neurodegenerative diseases isoflavones have shown protective properties. An extract of soybean isoflavone reduced the elevated oxidative stress parameters and reversed the overproduction of Aβ in rats with colchicine-induced neuronal damage [334]. In the same way, daidzein alone or mixed with genistein and glycitin isoflavones could reverse the cognitive impairments produced by scopolamine injection by activating the cholinergic system and the BDNF/ERK/CREB signaling pathway in mice [335,336], thus reinforcing the idea that soy isoflavones may be a good candidate for the treatment of neurodegenerative diseases. Besides the BDNF/ERK/CREB signaling pathway, it has been postulated that the Nrf2 signaling pathway can also be underlying the neuroprotective effects of isoflavones [337].
### **8. Flavones**
Flavones is one of the largest groups of flavonoids with a high degree of chemical diversity. Some of the richest sources of flavones are parsley, celery, peppermint, and sage, which predominantly contain apigenin and luteolin as well as maize and citrus fruits. In general, flavones are found as glucosides in citrus fruits, vegetables, herbs and grains and although they represent a small fraction of
the total flavonoid intake, they have shown health effects and anti-obesity properties [338,339]. As it is going to described latter, most of the studies that investigate the beneficial effects of flavones use them as aglycone and a scarce number of approaches deepen on the effects of flavones when consumed within the whole food and a feasible doses or in combination with other bioactive compounds.
### *8.1. Flavones Improved Liver Steatosis and Hepatic Inflammation*
Flavones such as apigenin, luteolin, baicalin, vitexin, nobiletin among others prevented NAFLD and hepatic steatosis mainly by modulating lipid metabolism (increasing FAO and decreasing lipogenesis) and reducing oxidative stress and inflammation [340–345].
As many other flavonoids, some flavones also exert their hepatic effects by activating the AMPK enzyme. Vitexin, an apigenin flavone glucoside, for instance, when administered to HFD-fed mice reduced body and liver weight, triglyceride and cholesterol content in serum and liver and circulating levels of ALT and AST. Moreover, vitexin regulated lipid metabolism suppressing de novo lipogenesis by downregulating the expression of *Ppar*γ*, C*/*ebp*α*, Srebp1c, Fasn,* and *Acc* and enhancing FAO and lipolysis by increasing the expression of *Ppar*α*, Cpt1a* and *Atgl*) in an AMPK-dependent way that has been suggested may be activated by the binding of vitexin to the Leptin receptor [345].
In a similar way, luteolin, the principal yellow dye compound from *Reseda luteola*, or luteolin-enriched artichoke leaf extract alleviated hepatic alterations caused by a HFD by exerting anti-inflammatory activities and modulating lipid metabolism. Luteolin treatment of HFD-fed mice reduced hepatic lipotoxicity by improving the inflammatory profile, decreasing the extracellular matrix, enhancing the antioxidant capacity of the liver and increasing the FFA flux between liver and WAT [346]. A crosstalk between adipose tissue and liver has been suggested to explain the effects of luteolin on hepatic steatosis [347]. Moreover, luteolin and luteolin-enriched artichoke leaf extract administered to HFD-fed mice prevented hepatic steatosis (less and smaller lipid droplets, lower levels of C*idea*) and insulin resistance by suppressing lipogenesis and gluconeogenesis (suppression of PEPCK and G6Pase activities) and increasing FAO (more CPT1a activity and higher expression of *Ppar*α*, Pgc1*α *and Pgc1*β*)* [342]. The repression of hepatocyte nuclear factor 4a and of LXR/SREBP1c signaling pathway has been described as putative molecular mechanisms for luteolin improvement of liver steatosis and NAFLD [348,349].
Regarding the capacity of flavones to modulate FAO, it has been described through a quantitative proteomic study that baicalin may act as an allosteric activator of CPT1a enzyme thus increasing the FA entrance to the mitochondria to undergo the β-oxidation in the liver [343]. Moreover, baicalin attenuated liver alterations by regulating the AMPK/ACC pathway in diet-induced obese mice [350]. Finally, baicalin is also a potent anti-inflammatory and antioxidant compound in a way that as other flavones also implied the nuclear erythroid 2-related factor 2 (Nrf2) activity in a cholestatic mice model [351].
It has been described that some flavones exert their hepatoprotective effects via the activation of the Nrf2 transcription factor. Nrf2 is a positive regulator of the expression of genes involved in the protection against oxidative stress as well as a negative regulator of genes that promote hepatic steatosis [352,353]. In this context, apigenin and scutellarin exerted their hepatoprotective activity via the activation of Nrf2. Scutellarin is a natural compound of *Erigeron breviscapus*(vant.) that in a HFD-fed mice attenuated obesity. It repressed lipogenesis and promoted FAO and cholesterol output besides its anti-inflammatory activity [340]. Moreover it has been described that scutellarin increased mRNA and/or protein levels of PPARγ, PGC1α, Nrf2, haem oxygenase-1 (HO-1), glutathione S-transferase (GST), NAD(P)H quinone dehydrogenase 1 (NQO1) and PI3K and AKT, whilst reduced nuclear factor kappa B (NF-κB), Kelch-like ECH-associated protein 1 (Keap1) [354,355]. By contrast, apigenin administration to HFD-fed mice inhibited the expression of PPARγ target genes via the translocation to the nucleus and activation of the Nrf2 transcription factor that seems to block PPARγ activity. Apigenin treatment downregulated the expression of genes related to lipid droplet formation (*Cidea, Plin2, fat storage inducing transmembrane protein 1 and 2 and)* as well as genes involved in FA uptake
(*Fabp1* and *Lpl*), FAO (*Cpt1a, Pdk4, Acox1, Acaa2*) and lipogenesis *(Fasn, Scd11, Acaca)* [341]. On the other side, apigenin may act as a PPARγ modulator in a mouse model of obesity where it activated the p65/PPARγ complex translocation into the nucleus, thereby decreasing the NF-κB activation and favoring the M2 macrophage polarization [356] or blocking NLRP3 inflammasome assembly and the ROS production [357]. The capacity of flavones to modulate PPARγ activity and induce macrophage polarization to M2 phenotype has also been described for Chrysin in a HFD-fed mice model [358].
Finally, wogonin have shown beneficial effects on the liver steatosis development in a mice NAFLD model [359]. Concretely wogonin administration to HFD fed mice ameliorated the NAFLD progression via enhancing the PPARα/Adiponectin receptor R2 (AdipoR2) pathway. Wogonin induced the hepatic activity of PPARα and upregulated the levels of the AdipoR2. Moreover, wogonin also reduced the inflammatory profile and alleviated the hepatic oxidative stress [359].
Besides their effects alone, the combination of flavones with other bioactive compounds or polyphenols-rich extracts have also shown positive effects against hepatic steatosis [360].
### *8.2. Flavones Improved the Adipose Tissue Inflammation and Reduced the Macrophages Infiltration as Well as Enhanced the Thermogenic Capacity*
Although flavones have been widely studied for their antioxidant and anti-inflammatory properties [338] their capacity to impact on adipose tissue metabolism and functionality cannot be underestimated.
Besides its reduction of the inflammatory phenotype in adipose tissue, apigenin administration to diet-induced obese mice ameliorated the body weight increment, reduced the visceral adiposity by inhibiting the adipogenesis via a STAT3/CD36 signaling pathway [361], decreased leptin and increased adiponectin [362] and induced energy expenditure mainly by promoting lipolysis and FAO as well as browning of WAT [363]. In scWAT, apigenin-treated mice exhibited a downregulation of adipogenic genes (*Ppar*γ*, Lpl* and *aP2)* and of genes involved in lipogenesis (*Fasn* and *Scd1*) and a promotion of lipolysis by increasing the mRNA levels of *Atgl, Hsl, Forkhead box protein O1 (FoxO1) and Sirt1.* In BAT there is an increment of the p-AMPK and p-ACC levels, thus indicating that FAO is enhanced in this fat depot after apigenin administration. Finally, apigenin activated the thermogenesis in BAT (upregulation of *Ucp1* and *Pgc1*α) and induced the browning phenotype in scWAT (upregulation of *Ucp1*, *Pgc1*α*, Tmem26, Cited1*) [363]. Similar results were obtained with vitexin. Vitexin administration reduced the adipocyte size of HFD-fed mice and increased the p-AMPK levels in eWAT followed by a downregulation of C/EBPa and FASN protein levels [364].
In the case of nobiletin and luteolin, their administration to HFD-fed mice improved the fibrotic and inflammatory profile in adipose tissue and reduced the macrophage infiltration and polarization [344,346,365,366]; but in contrast with other flavones they increased the mRNA expression of FAO- (*Ppar*α, *Cox8b*, and *Cpt1a*) and lipogenic (*Ppar*γ, *Srebp1c, Fasn* and *Scd1)* -related genes simultaneously [342,344] as well as CPT1 and FASN activity [344] in WAT. The simultaneously activation of both metabolic pathways in adipose tissues has been demonstrated as a way to maintain thermogenesis in BAT [367,368] and as a marker of browning in WAT [82]. In the case of luteolin, its administration either in HFD-fed or low-fat-fed mice activated browning and thermogenesis in mice via the AMPK/PGC1α cascade. Under the AMPK/PGC1α signal, luteolin increased energy expenditure in HFD-fed mice and upregulated the mRNA levels of *Pgc1*α*, PPAR*α*, Cidea* and *Sirt1* in BAT as well as *Ucp1 Pgc1*α*, Tmem26, Cidea, PPAR*α*, Sirt1, Elovl3 and Cited1* in scWAT [369]. Moreover, the increased of PPARγ protein levels in WAT has been linked to an alleviation of the hepatic lipotoxicity in HFD-fed mice [347]. Similar effects were observed with baicalein that administered to HFD-fed mice decreased pP38MAPK, pERK and PPARγ levels and increased pAKT, PGC1α and UCP1 as well as the presence of GLUT4 in cell membranes of the eWAT. Globally, baicalein reversed the glucose intolerance and insulin resistance produced by HFD [370].
Besides the effects of each compound by itself some flavones-rich extracts or foods or combinations of different bioactive compounds have been evaluated regarding their potential therapeutic role against obesity and its metabolic and inflammatory features [371,372].
### *8.3. Flavones and Obesity in the CNS: No Clear Evidences*
There are few studies describing the potential role of flavones in obesity-related central alterations. Just luteolin has been demonstrated a protective effect against HFD-induced cognitive effects in obese mice. Luteolin administration alleviated neuroinflammation, oxidative stress and neuronal insulin resistance as well as improved the Morris water maze (MWM) and step-through task and increased the levels of BDNF [373]. Other effects of flavones described recently are anxiolytic-like activity [374], neuroprotection against gamma-radiation [375] treatment of glioblastoma [376], amelioration of the hypoxia-reoxygenation injury [377] or inhibition of the neuroinflammation caused by LPS [378].
### **9. Chalcones**
Chalcones is a group of polyphenolic compounds with a broad structural diversity. Chalcones are precursors of other flavonoids and responsible for the golden yellow pigments found in flowers, fruits, vegetables, spices, teas and different plant tissues. Although their metabolism in the gastrointestinal tract and their rate of absorption are not still completely known, chalcones have shown a wide variety of biological activities. Several studies have demonstrated that, either from natural sources or synthetic, chalcones can impact on glucose and lipid metabolism and their health benefits have been studied in relation to type 2 diabetes [379]. Chalcones have shown hypoglycemic capacity, the ability to modulate food intake and activate AMPK, as well as antioxidant, anti-inflammatory, anticancer, anti-obesity, hepatoprotective and neuroprotective properties [380–392] Although there are no many studies in humans the effects of chalcones in the obese phenotype in animal models are similar to the ones described for other flavonoids, thus suggesting a potential therapeutic role of these group of bioactive compounds.
### *9.1. The Hepatoprotective Role of Chalcones*
Chalcones have hepatoprotective properties in NAFLD, alcoholic fatty liver, drug- and toxicant-induced liver injury, and liver cancer [381]. It has been described that chalcones are able to inhibit the synthesis of triglycerides and the lipogenesis, to increase FAO, and to modulate adiponectin production and signaling.
Licochalcone F, a novel synthetic retrochalcone, has shown anti-inflammatory properties when administered to diet-induced obese mice. Licochalcone F inhibited TNFa-induced NF-kB activation and the mRNA expression of several pro-inflammatory markers. In the liver licochalcone F alleviated hepatic steatosis, by decreasing lipid droplets and glycogen deposition [380]. On its side, Licochalcone A, a chalcone isolated from *Glycyrrhiza uralensis*, administered to HFD-fed mice, reduced body weight, decreased serum triglycerides, LDL free fatty acids and fasting blood glucose, ameliorated hepatic steatosis, reduced lipid droplet accumulation [393]. In the liver, licochalcone A downregulated the protein levels of SREBP1c, PPARγ, and FASN as well as increased the phosphorylation of HSL, ATGL and ACC enzymes [393]. Moreover, licochalcone A increase the protein levels of CPT1A and stimulated SIRT1 and AMPK activity [393]. Taken together, licochalcone A ameliorated obesity and NAFLD in mice at least in part by reducing the fatty acid synthesis and increasing lipolysis and FAO via the activation of the SIRT1/AMPK pathway.
In a mouse model of HFD-induced obesity, *trans*-chalcone reduced the ALT levels and increased the HDL [394]. Similarly, in a mouse model of non-alcoholic steatohepatitis KK-Ay mice, xanthohumol, the chalcone from beer hops (*Humulus lupulus* L.), diminished hepatic inflammation and prevented from the expression of profibrogenic genes in the liver [395] as well as lowered hepatic fatty acid synthesis through the downregulation of *Srebp1c* expression and promoted FAO by upregulating the mRNA expression of *Ppar*α in KK-Ay mice [396]. Moreover, in HFD-fed mice, xanthohumol prevented
body weight gain; decreased glycemia, triglyceride and cholesterol, and improved insulin sensitivity. Xanthohumol activated the hepatic and skeletal muscle AMPK, downregulated the expression of *Srebp1c* and *Fasn* and inhibited the activity of ACC, thus reducing the lipogenic pathway [386,397].
According to these data, aspalathin a C-glucosyl dihydrochalcone present in rooibos tea from *Aspalathus linearis,* also activated AMPK and reduced the expression of hepatic enzymes and transcriptional regulators that are associated with either gluconeogenesis and/or lipogenesis (*Acc, Fasn, Scd)* in diabetic *ob*/*ob* mice [388,398]. Furthermore, Aspalathin-enriched green rooibos extract (GRE) improved hepatic insulin resistance via the regulation of the PI3K/AKT and AMPK Pathways [399]. In obese insulin resistant rats GRE upregulated the expression of *Glut2*, *insulin receptor (Insr), Irs1* and *Irs2*, as well as *Cpt1a* [399]. Finally, Isoliquiritigenin at a low dose ameliorated insulin resistance and NAFLD in diet-induced obese mice. Isoliquiritigenin administration to HFD-fed mice decreased body fat mass and plasma cholesterol as well as alleviated hepatic steatosis (smaller lipid droplets) with no changes in TG and FFA serum levels [400]. It has been described that isoliquiritigenin suppressed the expression of lipogenic genes (*Fasn* and *Scd1*) and increased FAO activity. Moreover, isoliquiritigenin improved the insulin signaling in the liver and muscle [400].
Besides chalcones, chalcones-enriched products like Safflower yellow or ashitaba have demonstrated hepatoprotective properties. In mice fed with HFD, Safflower yellow improved lipid profile and alleviated fatty liver in a mechanism that has been associated to a reduction of the biosynthesis of intracellular cholesterol. Safflower yellow significantly reduced the levels of total cholesterol, triglycerides, LDL-cholesterol and the LDL/HDL ratio [401]. On its side, ashitaba (*Angelica keiskei)* extract showed hepatoprotective activity in fructose-induced dyslipidemia due to increased expression of FAO genes in the liver. Treatment with this extract upregulated the expression of the *Acox1, Mcad, ATP-binding membrane cassette transporter A1 (ABCA1)* and *apolipoprotein A1 (Apo-A1)* [402]. In a similar way, this extract exerted hepatoprotective effects in HFD-fed mice. Ashitaba extract reduced plasma levels of cholesterol, glucose, and insulin, lowered triglyceride and cholesterol content in the liver, inhibited hepatic lipogenesis by downregulating *Srebp1* and *Fasn* and activated FAO by upregulating the expression of *Cpt1A* and *Ppar*α [403]. The proposed mechanism underlying this hepatic metabolic effects is an activation of the AMPK enzyme in the liver [403].
In some of the studies the hepatoprotective role of chalcones has been linked to the adiponectin production. Concretely, trans-chalcone administration to high cholesterol diet-induced liver fibrosis increased the serum levels of adiponectin and the hepatic antioxidant enzymes, thus alleviating liver damage [404]. Similarly, xanthohumol and ashitaba extract or licochalcone A also increased the adiponectin expression and secretion [393,403,405].
### *9.2. Chalcones in the Adipose Tissue, Upregulation of Adiponectin, Induction of Browning and Enhancement of Energy Expenditure*
As has been mentioned above, chalcones induce adiponectin expression and secretion but also improve adipocytes function and reduce fat depots. Different molecular mechanisms underlying these effects has been described.
The treatment of obese mice with licochalcone F to reduced adipocyte size and ameliorated macrophage infiltration in WAT depots as well as enhanced Akt signaling and reduced p38 MAPK pathway [380]. On its side, the administration of Licochalcone A, isoliquiritigenin or a *Glycyrrhiza uralensis* extract containing licochalcone A, isoliquiritigenin, and liquiritigenin to diet-induced obese mice reduced body weight gain and adipose tissues depots [393,400,406]. In this case, Licochalcone A and *Glycyrrhiza uralensis* extract induced the browning phenotype in the iWAT this fat depot [393,406] as it is demonstrated by the enhanced expression of brown fat markers such as *Ucp1, Prdm16* and *Pgc1*α [406]. By contrast, isoliquiritigenin elevated energy expenditure by increasing the expression of thermogenic genes (*Ucp1* and *Prdm16*) as well as *Sirt1* that is linked to mitochondrial biogenesis [407] in interscapular BAT [400].
Finally, butein, besides its anti-inflammatory activity via the p38 MAPK/Nrf2/HO-1 pathway that leads to a reduction of the adipocyte hypertrophy [408] is also capable to enhance energy expenditure and increase thermogenesis. Butein induced the browning phenotype in the iWAT (upregulation of *Ucp1, Prdm16, cytochrome C oxidase 8b,* and *Cidea)* and increased the UCP1 protein levels in BAT in HFD-fed mice as well as in lean mice. The proposed molecular mechanism underlying these effects is the induction of the PR domain containing 4 (Prdm4) and the activation of the PI3Kα/Akt1/PR domain containing 4 (Prdm4) axis [409,410]. The browning effect of butein was not observed in other mice models such as ThermoMouse strain nor in methionine- and choline-deficient diet-fed mice [411]. Butein actions have also been linked to its capacity to downregulate PPARγ expression [387,410].
Finally, chalcone-rich extracts such as Safflower yellow or Ashitaba extract have also demonstrated effects in adipose tissues. Concretely, in mice fed with HFD, Safflower yellow administration exerts anti-obesity and insulin-sensitizing effects by upregulating the expression of *Pgc1*α that may indicate a browning phenotype of the scWAT as well as activating the protein levels of AKT and GSK3β in visceral WAT [412]. On its side, Ashitaba extract suppressed the HF diet-induced body weight gain and fat deposition in WAT, increased the adiponectin level and the phosphorylation AMPK, inhibited lipogenesis by downregulating *Ppar*γ*, CCAAT*/*enhancer-binding protein* α *(C*/*ebp*α*)* and *Srebp1* [403].
### *9.3. Chalcones in CNS: A Potential Neuroprotective Role*
The antioxidant and anti-inflammatory properties of chalcones has been linked to some of their neuroprotective effects [382,383,389] but no studies with obesity-related neuronal damage has been found. Further studies are needed to identify the potential therapeutic role of chalcones on this obesity side effect.
### **10. Concluding Remarks**
Undoubtedly flavonoids are potential therapeutic agents against metabolic disorders such as obesity, type 2 diabetes or NAFLD. Their impact in CNS, liver, and adipose tissue has been extensively studied and the results let us to be optimistic. Several metabolic effects and signaling pathways have been described underlying the anti-obesity effects of flavonoids specially in liver, EAT and BAT but also in CNS. Globally theses effects go to control body weight, improve insulin sensitivity, reduce fat accumulation in adipose tissues as well in ectopic depots and to increase energy expenditure (Figure 1). Furthermore, the data presented in this review highlight that:
**Figure 1.** Summary of the metabolic and signaling pathways underlying the anti-obesity effects of flavonoids. Molecular mechanisms underlying the beneficial effects of flavonoids have been widely studied and, in many cases, involved the activation of the AMP-activated protein kinase (AMPK). AMPK is a key enzyme for the control of lipid metabolism and adipogenesis. AMPK phosphorylation and activation promote catabolic processes such as FAO, glucose uptake, or glycolysis as well as inhibits anabolic pathways such as fatty acid synthesis or gluconeogenesis.
Even so more research is needed to confirm their therapeutically functionality in humans, the doses and times needed for their effectiveness or the better combination of bioactive compounds. Nowadays is still difficult to answer some crucial questions such as what is the effective dose of polyphenols; and for how long do we need to intake them to get positive effects? It is obvious that differences among experimental diets to induce fatty liver, dosages of bioactive compounds as well as the presence of other food compounds or the use of isolated or extracted polyphenols could influence the outcomes obtained. Furthermore, the use of flavonoids as a preventive or for treatment also show different results. Usually, the doses used in published papers are much higher than the ones reached from fruits and vegetables consumed as a whole.
The Predimed study determined that Spanish adults should intake around 820 ± 323 mg of polyphones/day in a 2000 Kcal diet to get their beneficial effects [25,27] but probably these effects at this dose are closely related to the MedDiet lifestyle. It is evident that, as MedDiet, some other dietary patterns include high amounts of fruits, vegetables or polyphenols-rich beverages that make possible to reach the optimal doses of polyphenols and by extension of flavonoids. Then, the question is: Are the effects of polyphenols linked to the dietary pattern where they are included? Two recent systematic reviews analyzed if there are enough evidence to define a health promoting polyphenol-rich dietary pattern and concluded that the high variability in the experimental approaches and methods used to evaluate polyphenols intake and health outcomes make difficult to stablish specific polyphenol intake recommendations and to clarify whether total flavonoids or rather individual subclasses may exert beneficial effects [30,36].
Moreover, low is known about the effects of combining different bioactive compounds from different families. Are they going to have synergic, additive or antagonic effects? And not less important is the need to identify the role of the food matrix on polyphenols and flavonoid effects.
**3.**Metaboliceffectsandsignalingpathwaysunderlyingtheanti-obesityeffectsofflavonoidsintheadipose
The bioavailability of polyphenols is low and not just their basic chemical structures (aglycons) are key but also the attachment of additional groups. There are described around 8000 structures of polyphenols with different physiological impact and several chemical structures, but all of them with at least one a phenolic ring with one or more hydroxyl groups attached [38,413,414]. The polyphenols absorption in human body is dose- and type-dependent and their effects are related to their bioavailability and pharmacokinetics. They show a low absorption rate and limited stability during pass through the intestinal tract where microbiome may contribute to their absorption. Once absorbed, polyphenols enter portal circulation and are metabolized in the liver. This first pass metabolism modifies the polyphenol structure and in consequence its bioavailability and bioactivity [415,416]. Finally, the conjugate metabolites reach the bloodstream and the target tissues [415–418].
Several studies have demonstrated that the bioavailability and safety of polyphenols changed when they are included in a food matrix [419–421]. Although most of the assays has been done with in vitro models of digestion [422] it seems that the food matrices protect bioactive compounds from intestinal degradation [420,423]. Finally, also cooking processes would have an impact in the polyphenols content and bioavailability of some preparations [424–426]. On the other side, it has been described that bioactive compounds with antioxidant properties are safe and beneficial but that exogenous supplementation with isolated compounds can be toxic [427].
The role of intestinal digestion and microbiota impact on polyphenols' effects must be also considered. Besides their direct action in the liver, some flavonoids may exert their metabolic effects through the gut microbiota modulation. An experimental approach with rabbits described that procyanidin b2 may downregulated fatty acid synthesis genes and protected against obesity and NAFLD by increasing the ratio of *Bacteroidetes* and *Akkermansia* [159]. Similar results were obtained with green tea oolong tea and black tea water extracts that administered to HFD-fed mice improved the glucose tolerance and reduced the weight gained caused by the HFD. Moreover, these animals showed a better hepatic lipid profile and a reduced mass of the WAT. These effects were accompanied by a reduction in plasma LPS, thus indicating less production and a significant increase in the production of short-chain fatty acids (SCFAs). A metagenomic analysis indicated that the tea extracts changed the gut microbiota's composition [428]. In the same way also flavones 'effects on obesity has been linked to gut microbiota modifications [338]. Oral hydroxysafflor yellow A (HSYA) reversed the HFD-induced gut microbiota dysbiosis and reduced the obese phenotype [429].
**Author Contributions:** V.S., H.S.-L., G.A. and J.R. performed the literature research and wrote the first draft of the manuscript; P.F.M., D.H. and J.R. evaluated the information, reviewed and edited the manuscript to define its last version. All authors have read and agreed to the published version of the manuscript.
**Funding:** This study was supported by the Ministerio de Economía y Competitividad [grants AGL2017-82417-R to PFM and DH, by the Generalitat de Catalunya [grants 2017SGR683, VS was supported by Conicyt's fellowship from the Government of Chile. The APC was funded by the University of Barcelona.
**Acknowledgments:** We acknowledge Jacques Truffert for the images used in Table 1 and Ursula Martínez-Garza for the images used in Figure 1. Chemical structures of flavonoids' subclasses from Table 1 have been done with Chemdraw®.
**Conflicts of Interest:** The authors declare no conflict of interest.
### **References**
intake and incidence of cardiovascular events in the PREDIMED study. *Nutr. Metab. Cardiovasc. Dis.* **2014**, *24*, 639–647. [CrossRef] [PubMed]
effects of the flavonoids quercetin, hesperetin, epicatechin, apigenin and anthocyanins in high-fat-diet-fed mice. *Genes Nutr.* **2015**, *10*, 1–13. [CrossRef]
behavior in diabetic rats: Role of ectonucleotidases and acetylcholinesterase activities. *Biomed. Pharmacother.* **2016**, *84*, 559–568. [CrossRef] [PubMed]
© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
| doab | 2025-04-07T03:56:58.620465 | 1-5-2021 17:49 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "003cf560-3166-422f-9a1c-98b1e971efa6",
"url": "https://mdpi.com/books/pdfview/book/3241",
"author": "",
"title": "Dietary Polyphenols and Human Health",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783039437931",
"section_idx": 14
} |
003cf560-3166-422f-9a1c-98b1e971efa6.15 | *Review* **Antidiabetic E**ff**ects of Flavan-3-ols and Their Microbial Metabolites**
### **Estefanía Márquez Campos, Linda Jakobs and Marie-Christine Simon \***
Department of Nutrition and Food Sciences, Nutrition and Microbiota, University of Bonn, 53115 Bonn, Germany; [email protected] (E.M.C.); [email protected] (L.J.)
**\*** Correspondence: [email protected]; Tel.: +49-228-73-36-80
Received: 25 April 2020; Accepted: 26 May 2020; Published: 29 May 2020
**Abstract:** Diet is one of the pillars in the prevention and management of diabetes mellitus. Particularly, eating patterns characterized by a high consumption of foods such as fruits or vegetables and beverages such as coffee and tea could influence the development and progression of type 2 diabetes. Flavonoids, whose intake has been inversely associated with numerous negative health outcomes in the last few years, are a common constituent of these food items. Therefore, they could contribute to the observed positive effects of certain dietary habits in individuals with type 2 diabetes. Of all the different flavonoid subclasses, flavan-3-ols are consumed the most in the European region. However, a large proportion of the ingested flavan-3-ols is not absorbed. Therefore, the flavan-3-ols enter the large intestine where they become available to the colonic bacteria and are metabolized by the microbiota. For this reason, in addition to the parent compounds, the colonic metabolites of flavan-3-ols could take part in the prevention and management of diabetes. The aim of this review is to present the available literature on the effect of both the parent flavan-3-ol compounds found in different food sources as well as the specific microbial metabolites of diabetes in order to better understand their potential role in the prevention and treatment of the disease.
**Keywords:** polyphenol; diabetes; flavonoids; catechins
### **1. Introduction**
Diabetes can be classified into type 1 diabetes (T1D), type 2 diabetes (T2D), and gestational diabetes mellitus (GDM). Its prevalence has increased over the last decade, with 463 million people registered as suffering from it in 2019 (9.3% of the global population) [1]. In the case of T2D, whose prevalence constitutes around 90% of the total number of diabetes cases, its increase is directly related to ageing, increased urbanization, and obesogenic environments [1]. A rising prevalence of T1D has also been observed, but in this case the causes are not completely clear [2].
In general terms, glucose homeostasis involves glucose absorption in the intestine, glucose uptake and metabolism by organs and tissues, and glucose hepatic production [3]. In T2D, peripheral glucose uptake, mainly in muscle, is decreased. This, together with an increased endogenous glucose production, leads to a hyperglycemic status. Moreover, lipolysis is increased and the resulting free fatty acids (FFAs) and intermediary lipid metabolites all lead to a more pronounced glucose output, decreased glucose utilization, and impaired activity of beta cells. Pancreatic beta cells are stimulated to compensate the hyperglycemic state by secreting insulin, but this function deteriorates over time. Glucagon secretion by pancreatic alpha cells is, moreover, impaired. A deterioration in the incretin effect could be the cause of both the impaired insulin and glucagon secretion since there is an inadequate release of, or response to, the gastrointestinal incretin hormones post-prandially. Moreover, renal tubular glucose reabsorption is increased [3].
Due to the adverse effects that the most commonly used antidiabetic drugs can have [4], finding natural substances for preventing or treating T2D has become an attractive potential alternative. Flavan-3-ols, the most commonly ingested flavonoids [5], have been related to different health promoting outcomes such as the prevention of cardiovascular disease [6] and cancer [7]. Regarding their effects on T2D, epidemiological data show that some foods rich in flavan-3-ols, such as green tea, could lower the risk of the disease [8–10].
This review presents in vitro, in vivo, and clinical studies regarding the effects of flavan-3-ols on diabetes both in their original form and their microbial metabolites in order to better comprehend the underlying molecular mechanisms on diabetes prevention.
### **2. Search Criteria**
A literature search was performed in Medline via PubMed for in vitro, in vivo, and human intervention trials published between 2005 and 2019 investigating the protective role of flavan-3-ols and their colonic metabolites on diabetes. Search terms included flavan-3-ol, flavanol, catechin, epicatechin, epigallocatechin, gallocatechin, procyanidin, theaflavin, γ-valerolactone, valeric acid, 3,4-dihydroxyphenyl propionic acid, 3-hydroxyphenyl propionic acid, 3-hydroxyphenylacetic acid, 3,4-dihydoxyphenylacetic acid, homovanillic acid, protocatechuic acid, 3-hydroxybenzoic acid, green tea, grape seed extract, cacao, diabetes, glucose, insulin, insulin resistance, beta cell, pancreas, glucagon, incretin effect, and vasodilation. In vitro and in vivo studies included both diabetic models and non-diabetic models. Only human trials with a study population presenting an impaired glucose metabolism (type 1 or type 2 diabetes mellitus, gestational diabetes, or pre-diabetes) were considered. The focus was on studies that primarily investigated effects on glucose metabolism.
### **3. Flavan-3-ols: Intake and Metabolism**
Flavan-3-ols constitute a flavonoid subclass naturally present in food as monomers (catechin (C) and epicatechin (EC)), oligomers, polymers (proanthocyanidins), and other derived compounds (such as theaflavins and thearubigins) [11].
Monomeric forms of flavan-3-ols are commonly present in cocoa beans, nuts, and fruits such as berries, stone fruits, apples, and pears [12]. Cocoa, berries, and nuts are also rich in proanthocyanidins [12]. Green tea is rich in gallocatechins while fermented black and oolong teas are sources of theaflavins and thearubigins [13].
The mean flavan-3-ol intake seems to range between 77 mg/day and 182 mg/day depending on the region, representing a much higher intake than that of other polyphenols [5]. Although the intake of flavan-3-ols is the highest among other polyphenols, the amount as well as the subtype ingested differ among countries. For example, the UK was shown to be the country with the highest total flavan-3-ol consumption in Europe, which is probably due to the widespread and high consumption of tea [14]. Therefore, monomer (especially epigallocatechin-3-gallate (EGCG)) and theaflavin (TF) intake were the highest in the UK [5,14]. Nevertheless, proanthocyanidin intake was statistically higher in Mediterranean countries, with the main sources there being stone and pome fruits [5,14].
After ingestion, the monomeric forms of the flavan-3-ols are absorbed directly in the small intestine by passive diffusion before undergoing reactions lead by the phase II enzymes [11]. These enzymatic reactions, which first take place in the enterocyte and later in the liver, are performed by uridine-5'-diphosphate glucuronosyltransferases (UGT), catechol-O-methyltransferases (COMT), and sulfotransferases (SULT). The conjugated metabolites (glucuronides, O-methyl-esters, and sulphates, respectively) are then released [11]. The conjugated metabolites are water-soluble and can circulate through the human body via the systemic blood stream or be removed from the body in the urine and bile [11,15,16]. When the conjugated metabolites are eliminated via the bile, they can be recycled because they can be transported to the duodenum, where they will undergo enzymatic modifications and be reabsorbed [15].
The remaining unabsorbed ingested oligomeric and polymeric forms of flavan-3-ols, as well as a fraction of the structures already absorbed in the small intestine, go to the colon [11]. There, the microbiota can perform metabolic transformations of the flavan-3-ols aided by hydrolysis reactions
(O-deglycosylation and ester hydrolysis), cleavage (C-ring cleavage, delactonization, demethylation), and reductions (dehydroxylation and double bond reduction) [17,18]. Specific colonic metabolites for flavan-3-ols are γ-valerolactones, while further phenolic compounds are also common after the microbial catabolism of other flavonoids [11].
After absorption, flavan-3-ols' colonic metabolites go through phase II metabolism in the liver and their conjugated forms reach the organs and tissues, where they exert their potential positive effects [11]. Since the microbial metabolites could be the active substances with beneficial physiological effects in addition to their precursor compounds, flavan-3-ol-derived metabolites formed by the colonic microbiota have been given significant attention [11].
### **4. Antidiabetic E**ff**ects of Flavan-3-ols: In Vitro and In Vivo Studies**
Flavan-3-ols and their colonic metabolites can modulate the molecular mechanisms involved in the pathogenesis of diabetes, including the glucose absorption rate in the gut, glucose peripheral uptake, glucose secretion, the modulation of beta cell function, the modulation of insulin secretion, and the modulation of the incretin effect (Figure 1).
**Figure 1.** Potential molecular mechanisms underlying the antidiabetic properties of flavan-3-ols. ↑: increase; ↓: decrease; Akt: protein kinase B; AMPK: 5' adenosine monophosphate-activated protein kinase; G-6-Pase: glucose-6-phosphatase; GLUT4: glucose transporter type 4; GS: glycogen synthase; GSK3: glycogen synthase kinase 3; hIAPP: human islet amyloid polypeptide; IKK: IκB kinase; IR: insulin receptor; IRS-1: insulin receptor substrate 1; JNK: c-Jun N-terminal kinases; mRNA: messenger RNA; PEPCK: phosphoenolpyruvate carboxykinase; PI3K: phosphoinositide 3-kinase; PKC: protein kinase C; PPARγ: peroxisome proliferator-activated receptor-γ; PTP1B: protein-tyrosine phosphatase 1B.
### *4.1. Glucose Absorption in the Gut*
The first factor contributing to the postprandial glycemic level in the plasma is the absorption of glucose in the gastrointestinal tract. This process is regulated by key enzymes such as α-glucosidase, which releases glucose from complex carbohydrates. Inhibition of α-glucosidase activity by a green tea water extract, a green tea polyphenol mixture, and EGCG has been shown to be
stronger than by acarbose (half maximal inhibitory concentration (IC50) values were 4.421 ± 0.018, 10.019 ± 0.017, and 5.272 ± 0.009 μg/mL for flavan-3-ols, respectively, and 4822.783 ± 26.042 μg/mL for acarbose) [19] (Table 1). In addition, grape seed extract (GSE) (86% gallic acid equivalents) inhibited α-glucosidase activity (IC50 = 1.2 ± 0.2 μg/mL) more strongly than acarbose (IC50 = 91.0 ± 10.8 μg/mL), and of the individual catechin 3-gallates, EGCG was the one with the strongest inhibitory effect (IC50 = 0.3 ± 0.1 μg/mL) [20].
### **Table 1.** In vitro studies on antidiabetic e ffect of flavan-3-ols and their microbial metabolites 1.
Similarly, epicatechin-3-*O*-(3-*O*-methyl) gallate (ECG3"Me), epigallocatechin-3-*O*-(3-*O*-methyl) gallate (EGCG3"Me), EGCG, and epicatechin-3-*O*-gallate (ECG) inhibited α-glucosidase, and in this case EGCG3"Me had the strongest effect. Their IC50 values were 14.7, 8.1, 13.3, and 61.1 μM respectively [21]. C was also shown to inhibit α-glucosidase stronger than acarbose (IC50 = 87.55 μg/mL vs. 199.53 ± 1.12 μg/mL, respectively) [22].
Interestingly, isolated procyanidins B2, B5 (dimeric), and C1 (trimeric) also had stronger α-glucosidase inhibitory activities than acarbose (IC50 = 4.7 ± 0.2, 5.5 ± 0.1, and 3.8 ± 0.2 μg/mL, versus IC50 = 130.0 ± 20.0 μg/mL, respectively), suggesting that the inhibitory activity could be correlated to the molecular weight of the compound [23].
For α-amylase, another digestive enzyme responsible for starch hydrolysis, GSE (86% gallic acid equivalents) inhibited its activity (IC50 = 8.7 ± 0.8 μg/mL), with the same potency as acarbose (IC50 = 6.9 ± 0.8 μg/mL) [20]. However, α-amylase was not strongly inhibited by tea extracts and individual catechin 3-gallates [20].
These effects have also been observed in mice fed with proanthocyanidins with different degrees of polymerization [52] (Table 2). Mice fed with proanthocyanidins with a high degree of polymerization showed a stronger inhibition of α-amylase activity both in the small intestine and in the pancreas than those fed with a low degree of polymerization proanthocyanidins. The rates of inhibition compared to the control group were 41% in the small intestine and 45% in the pancreas for high degree of polymerization proanthocyanidins, and 21% and 26% for low degree of polymerization proanthocyanidins [52].
**Table 2.** In vivo studies on antidiabetic effects of flavan-3-ols and their microbial metabolites
2.
**Table 2.** *Cont.*
**Table 2.** *Cont.* density lipoprotein-cholesterol; MDA: malondialdehyde; NAFLD: non-alcoholic fatty liver disease; NADPH: nicotinamide adenine dinucleotide phosphate; NOD: non-obese diabetic;OGTT: oral glucose tolerance test; PI3K: phosphoinositide 3-kinase; PKC: protein kinase C; PTP1B: protein-tyrosine phosphatase 1B; SOD: superoxide dismutase; STZ: streptozotocin;TAC: total antioxidant capacity; T1D: type 1 diabetes; TC: total cholesterol; TG: triglycerides; QUICKI: quantitative insulin sensitivity check index; w: week. Flavan-3-ols and theirmicrobial metabolites: C: catechin; CG: catechin gallate; EC: epicatechin; ECG: epicatechin gallate; EGC: epigallocatechin; EGCG: epigallocatechin gallate; GC: gallocatechin; GCG:
gallocatechin
gallate; PA4-1:
EC-(4β–6)-EC-(4β–8)-EC-(4β–8)-EC;
VL: valerolactone.
### *4.2. Insulin Signaling Pathways and Glucose Peripheral Uptake*
Due to the polar nature of glucose, its transport into the cell requires the use of transporter proteins in the cell membrane. These glucose transporters have different tissue distributions and a specific affinity for carbohydrates [69]. The insulin-regulatable glucose transporter type 4 (GLUT4) is found in insulin-sensitive tissues: skeletal muscle, cardiomyocytes, and adipocytes. Under physiological conditions, the insulin-mediated translocation of intracellular GLUT4 from the cytoplasm to the plasma membrane results in the uptake of glucose. This process is influenced by phosphoinositide 3-kinase (PI3K), protein kinase B (PKB or Akt), and protein kinase C zeta type (PKCζ). In short, insulin binding to insulin receptor (IR) leads to the phosphorylation of the beta subunit which, at the same time, phosphorylates the insulin receptor substrate (IRS). Upon tyrosine phosphorylation, which could be inhibited by serine phosphorylation of insulin receptor substrate 1 (IRS-1), PI3K binds to IRS and activates the Akt/PKB and the PKCζ cascades. Activated Akt induces glycogen synthesis via the inhibition of glycogen synthase kinase (GSK-3). Eventually, the Rab GTPase-activating protein AS160 (Akt substrate of 160 kDa) is activated, leading to the translocation of GLUT4 to the plasma membrane and glucose uptake [70].
When the translocation of intracellular GLUT4 to the plasma membrane is impaired, insulin resistance (IRes) takes place. T2D develops when both IRes and defects in insulin secretion occur [3].
Since approximately 80% of insulin-stimulated glucose uptake in the postprandial state takes place in the skeletal muscle, this tissue plays a key role in maintaining glucose homeostasis; therefore, many studies have focused on the effect of flavan-3-ols on GLUT4 translocation in skeletal muscle.
In vitro, a cacao liquor procyanidin (CLP) extract (1–10 μg/mL), consisting of EC, C, and other procyanidins, dose-dependently enhanced glucose uptake and promoted GLUT4 translocation to the plasma membrane of L6 myotubes after 15 min of incubation [53].
A mixture of TF, theaflavin-3-gallate (TF-3-G), theaflavin-3 -gallate (TF-3 -G), and theaflavin-3,3 -digallate (TFDG) (2.5–10 μg/mL, 24 h treatment) improved IRes induced by palmitic acid in HepG2 cells, as measured by the increase in 2-(N-(7-nitrobenz-2-oxa-1,3-diazol-4-yl)amino)-2-deoxyglucose (2-NBDG) uptake using metformin as a positive control [24]. Total GLUT4 and protein levels of GLUT4 bound to the membrane were increased by theaflavins in a dose-dependent manner [24]. They reversed the reduction of the phosphorylation level of Akt induced by palmitic acid and led to an increased phosphorylation of IRS-1 (Ser307) in HepG2 cells [24].
Interestingly, Ojelabi et al. showed that EGCG and ECG dose-dependently inhibited sugar uptake by glucose transporter type 1 (GLUT1), which was measured using 3-*O*-methylglucose uptake. It was found that low concentrations of the flavan-3-ols activated sugar uptake, while higher concentrations inhibited sugar uptake and noncompetitively inhibited sugar exit [25].
Glucose uptake in induced insulin-resistant 3T3-L1 adipocytes significantly increased after incubation with EGCG at 5 μM. Moreover, EGCG dose-dependently reversed the dexamethasone (Dex) and tumor necrosis factor (TNFα)-induced increase of c-Jun N-terminal kinases (JNK) phosphorylation levels and promoted GLUT4 translocation (1 μM) [26].
In vivo studies showed similar results. KK-Ay mice, when supplemented with green tea catechins (98% pure) at a low as well as at high concentrations (150 mg/kg/day and 300 mg/kg/day), showed a reduced JNK phosphorylation in adipose tissues when compared to untreated animals and an increased GLUT4 content in the plasma membrane [26].
Yamashita et al. administered a CLP extract as a single dose (250 mg/kg) to mice at the Institute of Cancer Research (ICR). After carbohydrate ingestion, CLP suppressed the hyperglycemic response and improved GLUT4 translocation in skeletal muscle [53]. In fact, the GLUT4 translocation was approximately 3.9-fold higher in comparison with the control group, who were only administered water and no glucose [53]. These results were further confirmed by a consecutive administration of a CLP-supplemented (0.5%) diet to C57BL/6 mice for 7 days, which had the same effects on skeletal muscle GLUT4 after glucose load [53]. Similarly, procyanidins (both low and high degree of polymerization, 10 mg/kg) from a CLP extract prevented hyperglycemia through the promotion of GLUT4 translocation in the skeletal muscle of ICR mice [54]. This could be explained by the significantly increased phosphorylation of 5' adenosine monophosphate-activated protein kinase (AMPK), ß-subunit of IR (IRβ), IRS-1, and PI3K by procyanidins with both low and high degrees of polymerization [54].
After the oral administration of EC, procyanidin B2, procyanidin C1, EC-(4β–6)-EC-(4β–8)-EC-(4β–8)-EC (PA4-1), and cinnamtannin A2 (PA4–2) (10 μg/kg) to ICR mice, GLUT4 translocation in skeletal muscle significantly increased compared to in control mice [55]. Trimeric and tetrameric procyanidins significantly promoted phosphorylation of PI3K, and PA4-1 was able to significantly induce phosphorylation of Akt1 at both serine 473 and threonine 308 [55]. The latter compound was the only one able to significantly promote the phosphorylation of IRS-1 as well as increase the insulin plasma level. Similarly, all compounds significantly induced phosphorylation of AMPK [55].
In a model of T1D, streptozotocin (STZ)-induced rats were administered a green tea extract (GTE) for 12 days composed of the following catechins: C, EC, (−)-gallocatechin (GC), (−)-epigallocatechin (EGC), (−)-catechin gallate (CG), ECG, (−)-gallocatechin gallate (GCG), EGCG, and caffeine [56]. After an oral glucose tolerance test (OGTT), high blood glucose induced by STZ was significantly reduced with the GTE treatment when compared to the control group. When the possible mechanisms were investigated, the authors found that the GTE treatment increased the translocation of GLUT4 in the skeletal muscle to a normal level when compared to untreated rats. In contrast, the level of the IRß was not changed. These results imply that the green tea improved hyperglycemia in T1D rats without having an influence on insulin secretion from pancreatic beta cells, by promoting GLUT4 translocation in skeletal muscle. In addition to these findings, the degree of protein glycation induced by STZ measured by fructosamine and glycated hemoglobin (HbA1c) significantly decreased after the treatment with the GTE. This result suggests not only a protective role of green tea against the manifestation of diabetic complications but also an ability to improve those already presenting [56].
In a parallel experiment, an OGTT in KK-Ay mice was also performed, but in this case, mice were treated with GTE for 63 days (one group) or for 42 days directly after the appearance of hyperglycemia (another group). The authors found that the blood glucose after green tea intake was significantly lower when compared to the control and GLUT4 translocation in the skeletal muscle was significantly increased when compared to the control, but the level of IRß remained unaltered [56]. Another result from this experiment is the significant reduction of protein glycation and triacylglycerol by green tea [56].
In a study from Cremonini et al., EC supplementation (20 mg/kg) in high-fat-diet-induced obese and diabetic C57BL/6 mice improved insulin sensitivity and glucose homeostasis when compared to non-supplemented and control mice. The impairment of the insulin signaling cascade in the liver and the adipose tissue induced by the high-fat diet was prevented and the upregulation/activation of proteins which inhibit the insulin pathway (IκB kinase (IKK), protein kinase C (PKC), JNK, and protein-tyrosine phosphatase 1B (PTP1B)) was prevented [57].
Bettaieb et al. found that the supplementation of the diet of high-fructose-fed rats with EC (20 mg/kg) for 8 weeks mitigated the IRes induced by the high fructose concentrations, and it reversed both the impaired activation of the insulin signaling cascade (IR, IRS-1, Akt, and extracellular signal–regulated kinases 1/2 (ERK1/2)) as well as the upregulation of negative regulators (PKC, IKK, JNK, and PTP1B) in the liver and adipose tissue [58].
Glucose uptake has been shown to be promoted not only by the flavan-3-ols in their original form but also by some of their microbial metabolites. Specifically, 5-(3,5-dihydroxyphenyl)-γ-valerolactone promoted GLUT4 translocation in L6 skeletal muscle cells and soleus muscle by phosphorylation of the AMP-activated protein kinase (AMPK) signaling pathway both in vitro and in vivo at concentrations of 1–3 μM and 32 mg/kg, respectively [27]. At 32 mg/kg it caused suppression of hyperglycemia after an OGTT, while a higher dosage of 64 mg/kg only influenced AMPK phosphorylation [27].
Other microbial metabolites unspecific to flavan-3-ols have also been shown to modulate molecular mechanisms related to diabetes. Scazzocchio et al. investigated whether protocatechuic acid exerted an effect on glucose transport in adipocytes [28]. Incubation of the metabolite at 100 μM for 18 h with human and murine adipocytes treated with oxidized low density lipoprotein (oxLDL) significantly improved glucose uptake, GLUT4 translocation, and adiponectin secretion. These effects were observed after stimulation with insulin and also without it [28]. Glucose uptake was significantly and dose-dependently enhanced in non-oxLDL-treated human and murine adipocytes without the presence of insulin up to 40% and 60%, respectively [28]. These results indicate an insulin-like activity. A reversion of the oxLDL-induced diminishment of mRNA expression and activity of the peroxisome proliferator-activated receptor-γ (PPARγ) was also observed, and its inhibition impeded both the adiponectin and GLUT4 upregulation suggesting its implication in the insulin-like activity [28].
Both EC at 10 μM and 2,3-dihydroxybenzoic acid (2,3-DHB) at 20 μM increased IR and IRS-1 tyrosine phosphorylated and total protein levels in rat renal NRK-52E cells. In addition, phosphorylated levels of Akt and GSK-3 increased and those of glycogen synthase (GS) decreased [29]. Similarly, after treatment of renal tubular NRK-52E cells with high glucose levels and either EC at 5–20 μM or 3,4-dihydroxyphenylacetic-acid (3,4-DHPA) at 10–20 μM, the induced impairment of glucose uptake was restored. At 10 μM, EC and 3,4-DHPA increased tyrosine phosphorylated levels and total levels of IR, reversed the inhibition of the PI3K/Akt pathway involved in the insulin signaling cascade, and prevented the high-glucose-induced downregulation of AMPK phosphorylation [30].
### *4.3. Beta Cell Viability and Function*
In the situation of IRes, pancreatic beta cells try to maintain glucose levels by enhancing insulin production and increasing islet size and beta cell mass. However, an increased insulin response does not mean that beta cells are functioning normally. In fact, beta cells in this situation are kept under a high workload which, when maintained over time, results in functional exhaustion, dedifferentiation, and eventually beta cell death [71]. Apoptosis of beta cells is mainly induced by glucotoxicity, lipotoxicity, and deposits of islet amyloid polypeptide (IAPP) [72,73].
Glucose-stimulated insulin secretion (GSIS) in the beta cell line INS-1D after treatment with catechins was studied by Kaneko et al. [31]. Both EGCG at 10 μM as well as GCG at 30 μM significantly inhibited the GSIS. Furthermore, at 100 μM they almost eliminated GSIS. EC and C did not modify GSIS at concentrations up to 100 μM. At 10 μM, EGC nearly eliminated GSIS, while GC and ECG partially inhibited it. CG did not alter GSIS at concentrations up to 100 μM. Apart from this, EGCG, and not EC, inhibited the variation of intracellular Ca2<sup>+</sup> concentration. These results suggest that, at concentrations higher than physiological levels, some catechins have an inhibitory effect on GSIS, which is induced by the structure-dependent inhibition of voltage-dependent Ca2<sup>+</sup>-channels [31].
Supporting these results, a treatment with EC at a physiological dose of 0.3 μmol/L but not at 30 μmol/L improved GSIS of saturated fatty acid (SFA)-impaired INS-1 cells [32]. This was thought to be due to a modulation of the cell secretory capacity via the activation of the Ca2+/calmodulin-dependent protein kinase II (CaMKII) pathway and possibly through the GPR40 receptor [32].
In humans and animals, beta cell functionality can be measured by several methods. Some of the most commonly used methods include the homeostasis model assessment (HOMA), OGTT or intravenous glucose tolerance tests and the hyperglycemic clamp procedure [74]. The ability of flavan-3-ols to affect these has been as well assessed. In a study from Othman et al., treatment of diabetic rats with EGCG (2 mg/kg) every other day over one month significantly decreased the HOMA of insulin resistance (HOMA-IR) value and increased insulin levels when compared to untreated diabetic rats [59].
In a model where male Wistar rats were contrived to be obese through a cafeteria diet, a 21-day treatment with grape seed procyanidin extract (GSPE) at 25 mg/kg (defined composition) improved IRes measured by HOMA-IR [60]. The HOMA of beta cell function (HOMA-β) index also decreased. Insulin gene expression in the pancreas tended to decrease in treated rats, and a significant decrease in
the expression of carboxypeptidase E (Cpe) was also shown [60]. On the other hand, treatment with GSPE enhanced the increase in the Bcl-2-associated X protein (Bax) levels induced by the cafeteria diet, which suggests an increased apoptosis in the pancreas in contrast to results from other studies [60].
Gan et al. suggested that EGCG dose-dependently improved IRes in high-fat diet non-alcoholic fatty liver disease (NAFLD) mice by enhancing the insulin clearance of the hepatic insulin degrading enzyme (IDE) [61]. In this study, NAFLD mice were administered 10, 20, and 40 mg/kg EGCG intraperitoneally. Hyperglycemia, hyperinsulinemia, and IRes observed in mice fed a high-fat diet without EGCG were reversed by the polyphenol [61].
Insulin deficiency and IRes have been described in ß-thalassemia patients with iron overload, which is probably a secondary effect of a diminished pancreatic beta cell function. The incubation of iron-loaded rat insulinoma pancreatic β-cells with a GTE (2.29 μg EGCG equivalent) increased insulin secretion levels 2.5-fold and decreased cellular levels of iron and reactive oxygen species (ROS) [33].
The effect of flavan-3-ols directly on beta cell viability was also assessed. Cinnamtannin B1, procyanidin C1, and cinnamtannin D1 from cinnamon extracts were shown to dose-dependently protect INS-1 cells from palmitic acid and H2O2-induced reduction in terms of cell viability [34]. At 25 μmol/L, they enhanced insulin secretion in lipotoxic INS-1 cells [34]. However, the flavan-3-ols EC and procyanidin B2 had no significant effects [34].
In *db*/day*b* mice, treatment with EGCG (10 g/kg diet, 1% (*w*/*w*)) for 10 weeks improved glucose tolerance and additionally increased GSIS similarly to rosiglitazone, although no significant effect was found in IRes (HOMA-IR and quantitative insulin sensitivity check index (QUICKI)). This effect may be mediated by changes in pancreatic islets, since the number and size of pancreatic islets increased, together with a reduction of islet endoplasmic reticulum stress markers ex vivo [62].
The literature suggests that human islet amyloid polypeptide (hIAPP) fibril formation contributes to T2D by causing beta cell dysfunction and apoptosis. For this reason, the inhibition of the formation of toxic hIAPP oligomers and fibrils may be a good therapeutic strategy for the management of T2D. Some authors have therefore tried to elucidate the role of flavan-3-ols in the prevention of their formation. In hemizygous non-diabetic hIAPP transgenic mice treated with EGCG (0.4 mg/mL) for three weeks, EGCG reduced amyloid fiber intensity suggesting a beneficial effect on pancreatic amyloid fibrils in vivo [63]. However, there was no effect on diabetic hIAPP transgenic mice. This, therefore, suggests that EGCG would be effective as an early therapeutic method.
Mo et al. went further and examined the molecular process by which EGCG could inhibit hIAPP aggregation [35]. The authors found that in vitro EGCG could block the inter-peptide hydrophobic/aromatic interactions responsible for inter-peptide β-sheet formation and the intra-peptide interaction related to ß-hairpin formation. Thus, the three-stranded β-sheet structures were removed and loosely packed coil-rich conformations were formed. This EGCG-induced conformational shift of the hIAPP dimer was related to hydrophobic, aromatic stacking, cation-π, and H-bonding interactions [35].
Adding to these results, Meng et al. proved that EGCG inhibited in vitro amyloid formation by IAPP and disaggregated IAPP amyloid fibrils. At the same time, EGCG protected cultured rat INS-1 cells against IAPP-induced toxicity at 30 μM [36]. EGCG (2–32 μM) was also shown to inhibit the nucleation and fibrillation of hIAPP by forming hIAPP amorphous aggregates instead of ordered fibrils [37]. Moreover, a complex of Al(III)/EGCG was able to inhibit hIAPP fibrillation more effectively than the flavan-3-ol alone [37].
T-cell-mediated destruction of pancreatic beta cells leads to insulin deficiency in T1D. In addition, inflammation is known to play a role in the pathogenesis of T1D [3]. In this regard, EGCG prevented the onset of T1D in non-obese diabetic (NOD) mice when administered at 0.05% in drinking water (60–90 mg/kg body weight (b.w.), equivalent to 4.5–6.8 g/day by a 75 kg person) for 32 weeks [64]. Compared to control mice, plasma insulin levels were higher, HbA1c concentrations were lower, and circulating anti-inflammatory cytokine interleukin 10 (IL-10) levels were increased. However, no effect on pancreatic insulitis was observed. When human pancreatic islets were incubated with
inflammatory cytokines, addition of EGCG (1 and 10 μM) promoted islet viability [64]. Similarly, the administration of EC at 0.5% in drinking water (equivalent to an intake of 250 g dark chocolate containing 6% EC) for 32 weeks also delayed the development of T1D [65]. Importantly, pancreatic islet mass was preserved and the lymphatic infiltration into islets was lower meaning an improvement in the insulitis. Anti-inflammatory cytokine IL-10 levels increased [65]. HbA1c concentrations were, in this case, significantly lower and plasma insulin levels were significantly higher in mice treated with EC than in untreated mice [65].
The effect of low molecular weight phenolics produced after colonic metabolism of flavan-3-ols on beta cell functionality and viability has also been assessed. Fernández-Millán et al. found out a significant increase in GSIS in INS-1E pancreatic beta cells and isolated rat islets after treatment with 3,4-DHPA and 3-hydroxyphenyl propionic acid (3-HPP) at low concentrations (5 and 1 μM, respectively) [38]. Under oxidative stress induced by tert-butyl hydroperoxide (*t*-BOOH), both metabolites restored GSIS to control levels and significantly decreased cell death [38]. PKC and ERK could play a role in producing the observed effect, since their phosphorylation levels increased after treatment [38].
3,4-DHPA (250 μM) could also prevent the diminished insulin secretion induced by high cholesterol on Min6 pancreatic beta cells [39]. Moreover, it dose-dependently prevented cholesterol-induced cytotoxicity and apoptosis. Oxidative stress and mitochondrial dysfunction were also prevented [39].
5-Phenylvaleric acid, hippuric acid and homovanillic acid improved GSIS in beta cells more effectively than EC at concentrations up to 100 μM [40]. In addition to stimulating beta cell function, the microbial metabolites enhanced glucose utilization in skeletal muscle [40].
### *4.4. Endogenous Glucose Production*
The liver's inability to perceive insulin signals directly after glucose ingestion leads to the continuing production of glucose and, therefore, importantly contributes to a hyperglycemic status [3]. The maintained glucose output by the liver can be a consequence of two processes: gluconeogenesis and glycogenolysis [3]. However, the latter has a less important role in the increased glucose production of T2D patients [75]. The mechanisms responsible for the increase in hepatic gluconeogenesis include hyperglucagonemia, higher circulating levels of gluconeogenic precursors (lactate, alanine, and glycerol), elevated FFA oxidation, enhanced sensitivity to glucagon, and reduced sensitivity to insulin [3].
Increased activity of insulin-influenced phosphoenolpyruvate carboxykinase 1 (PCK1) and glucose-6-phosphatase (G-6-Pase) seems to contribute to the accelerated rate of hepatic glucose production [3]. In this sense, studies have shown how flavan-3-ols affect the expression of key regulators of the gluconeogenesis pathway.
Waltner-Law et al. studied the effects of green tea compounds on insulin signaling pathways, gene expression, and glucose production [41]. The authors found that EGCG had insulin-like activities in hepatoma cells. At 25 μM, EGCG reduced glucose production to basal levels in a similar way to insulin (10 nM) and these effects were already significant at lower concentrations (12.5 μM). When studying the impact of the flavan-3-ol on the expression of genes encoding gluconeogenic enzymes, EGCG reduced phosphoenolpyruvate carboxykinase (PEPCK) mRNA in a dose-dependent manner (12.5–100 μM) and both PEPCK mRNA and G-6-Pase in a phosphoinositide 3-kinase (PI3K)-dependent manner [41]. In addition, 50 μM EGCG could activate PI3K within 10 min, similar to insulin (10 nM), but the activation of other kinases such as PKB and p70s6k was much slower and not significant. The authors suggested that EGCG has a similar mechanism to insulin in reducing glucose production and expressing the PEPCK and G-6-Pase genes by modulation of the redox state of the cell [41].
Smaller amounts of EGCG (0.25–1 μM) suppressed gluconeogenesis in mouse cyclic adenosine monophosphate dexamethasone (cAMP-Dex)-stimulated hepatocytes and blocked the expression of the PEPCK and G-6-Pase genes [42]. However, no effect on the stimulation of tyrosine phosphorylation of IRS-1 or Akt, nor an influence of the PI3K inhibitor LY294002, was found suggesting an independent mechanism to the insulin signaling pathway [42]. The other known suppressor of hepatic gluconeogenesis, apart from the insulin signaling, is AMPK. In this case, EGCG increased the AMPK and acetyl-CoA carboxylase (ACC) phosphorylation in a time- and dose-dependent manner, and the suppression of AMPK resulted in the reversion of the effect of EGCG on the expression of the PEPCK and G-6-Pase genes in a calcium/calmodulin-dependent protein kinase kinase (CaMKK)- and ROS-dependent manner [42].
Yadollah et al. showed that 40 μM EGCG significantly reduced the expression of PEPCK and G-6-Pase in insulin-resistant HepG2 cells by 53% and 67%, respectively [43]. This effect was similar to that of 10 μM pioglitazone, which is a medication used to treat T2D. A combination of EGCG and pioglitazone induced a stronger reduction in the expression of PEPCK and G-6-Pase. The authors also proved that glucose production in HepG2 cells was significantly reduced by 50% by EGCG, by 55% by pioglitazone, and by 69% by a combination of both EGCG and pioglitazone [43].
Aside from the liver, the kidneys are also involved in glucose homeostasis and gluconeogenesis. EC (5–20 μM) and 2,3-DHB (20 μM) reduced cellular glucose uptake in rat renal NRK-52E cells similarly to the sodium-glucose cotransporter-2 (SGLT-2) antagonist phlorizin, leaving the expression of SGLT-2 and glucose transporter type 2 (GLUT2) unaltered [29]. A reduction in glucose production and PEPCK levels was also observed [29]. Moreover, the authors showed that Akt was involved in the modulation of both PEPCK levels and glucose production in NRK-52E cells [29].
Treatment of renal tubular NRK-52E cells with EC (10–20 μM) and 3,4-DHPA (10 μM) separately alleviated the alterations in glucose production and the upregulation of PEPCK induced by high glucose [30]. However, the protective effect disappeared when Akt and AMPK were inhibited. Therefore, both Akt and AMPK seem to be key molecules in the modulation of the glucose homeostasis and the preservation of renal tubular functionality [30].
### *4.5. Incretin E*ff*ect*
Incretin hormones include glucose-dependent insulinotropic polypeptide (GIP) and glucagon-like peptide-1 (GLP-1). They are gut peptides secreted after the intake of nutrients, such as glucose, and are responsible for the incretin effect, which is the increased stimulation of insulin secretion by oral glucose rather than by intravenous glucose infusion. This effect is impaired in patients with T2D due to the reduced insulinotropic effect of GIP and GLP-1 [76]. In addition to the insulinotropic activity, the incretin hormones work together to regulate glucagon secretion: GIP stimulates glucagon secretion while GLP-1 inhibits glucagon secretion by alpha cells. In diabetic patients, glucagon secretion is altered since it is not inhibited in hyperglycemic conditions [70].
Yamashita et al. studied if isolated dimeric, trimeric, and tetrameric procyanidins from cacao liquor administered as a single-dose (10 μg/kg) in mice could influence GLP-1 and insulin levels in plasma [66]. The tetrameric procyanidin cinnamtannin A2 was the only compound able to increase the plasma insulin level without a glucose load as well as significantly increase the GLP-1 secretion levels in plasma 60 min after oral administration [66]. In vitro experiments revealed an increased phosphorylation of proteins IRß and IRS-1 in the soleus muscle as a result of the action of insulin. Procyanidins (low and high degree of polymerization, 10 mg/kg) from a CLP-rich extract increased GLP-1 secretion with or without glucose load in mice [54].
González-Abuín et al. evaluated the modulation of the mechanisms that have an influence on GLP-1 secretion in STC-1 cells by GSPE [44]. The authors found out that 0.05 mg/L GSPE induced depolarization, while 50 mg/L induced hyperpolarization in enteroendocrine cells [44]. This high extract concentration suppressed GLP-1 secretion by around 40%. Under nutrient-stimulated conditions, 50 mg/L GSPE reduced the membrane depolarization induced by nutrients and reduced GLP-1 secretion by 20% in glucose- and proline-stimulated cells. These results indicate the importance of the GSPE concentration in depolarization and GLP-1 secretion by STC-1 cells, as well as the influence that nutrients have on GLP-1 secretion by enteroendocrine cells [44].
Glycogen synthesis is also one of the functions of incretin hormones. Its secretion rate in muscle is controlled by GS, which is also enhanced by insulin. Therefore, this stimulates a cascade of phosphorylation-dephosphorylation reactions [3]. Glycogen synthase phosphatase (PP1) is activated by the phosphorylation of serine phosphorylation site 1 in the regulatory subunit (G) of PP1 by insulin and this phosphorylation is catalyzed by insulin-stimulated protein kinase 1 (ISPK-1). Phosphorylation of site 2 by cAMP-dependent kinase (PKA) leads, on the contrary, to its inactivation [3].
Some authors have studied how flavan-3-ols influence glycogen synthesis. Kim et al. showed that green tea polyphenols consisting of 68% EGCG were able to enhance glycogen synthesis by up to a factor of 2 (10 μM) in high glucose treated HepG2 cells under 100 nM insulin stimulation [45]. The molecular mechanism can involve the regulation of enzymes such as glycogen synthase kinase 3-beta (GSK3ß) and GS since expression of phospho-GSK3β (Ser9) and phospho-GS (Ser461) were enhanced by EGCG [45].
### *4.6. Other Mechanisms*
The production of cellular oxidants may affect insulin sensitivity via the negative regulation of insulin signaling pathways (JNK, IKK), the promotion of sustained chronic inflammation, and oxidative stress. Flavan-3-ols are known to have antioxidative functions, and these could exert a protective effect against diabetes and its complications via controlling the oxidative stress. Cinnamtannin B1, procyanidin C1, and cinnamtannin D1 (12.5–50 μmol/L) from cinnamon extracts inhibited H2O2-induced ROS generation as well as increased cell viability of INS-1 cells [34]. Similarly, under *t*-BOOH-induced oxidative stress, the microbial metabolites 3,4-DHPA and 3-HPP (5 and 1 μM, respectively) significantly decreased rat pancreatic beta cell death and ROS and carbonyl group production [38]. While EC at a low dose of 0.3 μmol/L, but not at a higher dose of 30 μmol/L, improved GSIS of SFA-impaired INS-1 cells, only the highest dose of EC significantly reduced ROS after treatment with H2O2 and high glucose [32].
Bettaieb et al. found that the supplementation of high-fructose-fed rats with EC (20 mg/kg) for 8 weeks, mitigated the IRes induced by high fructose concentrations. EC supplementation (20 mg/kg) in high-fructose-fed rats showed an ability to inhibit the expression and activity of NADPH oxidase and the activation of redox-sensitive signals [58].
Treatment of STZ-induced diabetic rats with C (20 and 40 mg/kg/day) significantly decreased glucose levels, while superoxide dismutase (SOD), catalase (CAT), and glutathione S-transferase (GST) levels increased in a concentration-dependent manner, especially after treatment with 80 mg/kg/day [67].
Haidari et al. showed that a GTE given to STZ-induced diabetic rats at 200 mg/kg for 4 weeks, significantly decreased their serum glucose levels as well as the serum and hepatic malondialdehyde (MDA) concentration when compared to the diabetic control group. Total antioxidant capacity (TAC) was significantly increased after treatment [68].
Plasma glucose levels could also be controlled by the modulation of lipid digestion and the reduction of hyperlipidemia [77]. C treatment of STZ-induced diabetic rats dose-dependently decreased the serum levels of total cholesterol (TC), triglycerides, LDL, apoprotein B, and glucose levels, while it increased the serum levels of high density lipoprotein (HDL) and apoprotein A-I (20–80 mg/kg) [67]. EGCG dose-dependently reversed increased serum lipid levels including TC, TG, and LDL, and increased HDL in high-fat diet NAFLD mice compared with control mice [61]. EC (20 mg/kg body weight) prevented the high-fat-diet-induced increase in plasma TG and FFA in C57BL/6 mice [57]. Treatment of HepG2 cells with 100 nM insulin and 0.1–10 μM EGCG reduced lipogenesis to 65% compared to cells treated with insulin alone through increased expressions of phosphor-AMPKα and phosphor-ACC [45].
Inflammation contributes to impaired glucose management by adipocytes, hepatocytes, and muscle cells and interferes with insulin production and insulin signaling [78]. TNFα plays an important role in the activation of signaling cascades in adipocytes related to inflammation and IRes. In this context, EC (0.5–10 μM) has been shown to dose-dependently reduce TNFα-mediated JNK, ERK1/2, and *p*-38
phosphorylation, and nuclear AP-1-DNA binding in 3T3-L1 adipocytes [46]. It also inhibited the activation of the nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB) signaling cascade preventing the p65 nuclear transport and nuclear NF-κB-DNA binding. Moreover, EC reversed the TNFα-mediated downregulation of PPARγ expression and reduced nuclear DNA binding. The altered transcription of genes involved in inflammation and insulin signaling (monocyte chemoattractant protein 1 (MCP-1), IL-6, TNFα, resistin and protein-tyrosine phosphatase 1B) mediated by TNFα was inhibited by EC [46].
EC supplementation (20 mg/kg) in high-fructose-fed rats inhibited the expression of NF-κB regulated pro-inflammatory cytokines and chemokines [58].
The colonic metabolites of flavan-3-ols have also been shown to exert beneficial effects in diabetes other than those directly related to glycemia. One of them is the positive effect on vascular function, which is known to be directly linked to diabetes [79]. As reported in several studies, low molecular weight phenolics such as 2,3-dihydroxybenzoic acid (2,3-DHB), 3-HPP, and 3,4-DHPA could exert vasodilatory activities by stimulating NO production [47–49]. Apart from this, dihydroferulic acid, 3-hydroxyphenylacetic acid (3-HPA), 3,4-DHPA, and homovanillic acid could reduce the formation of advanced glycation end-products (AGEs) [50,51], which are thought to be linked to the development of diabetes and insulin resistance and to the occurrence of diabetic complications [80].
### **5. Antidiabetic E**ff**ects of Flavan-3-ols: Clinical Intervention Trials**
Some authors have studied the effect of the supplementation of pure flavan-3-ols on antidiabetic effects in order to exclude potential interactions with other compounds and with other flavonoids present in flavan-3-ol-rich food. Zhang et al. investigated the effects of a daily intake of EGCG (500 mg/day) in women with a diagnosed GDM at the beginning of the third quarter of pregnancy (29 weeks). HOMA-IR, HOMA-β fasting blood glucose and insulin levels decreased whereas the insulin sensitivity as measured by QUICKI increased due to the intervention. Furthermore, neonatal complications at birth, such as low birth weight or hypoglycemia, were significantly reduced in the intervention group [81] (Table 3).
**Table3.***Cont.*
QUICKI: quantitative insulin sensitivity check index; RCT: randomized controlled trial; RD: respiratory distress; SAE: small artery elasticity; SBP: systolic blood pressure; TAOS: total
antioxidant status; TC: total cholesterol; TG: triglycerides; week; y: year. Flavan-3-ols and their microbial metabolites:
TP: total phenols; T2D: type 2 diabetes; w: women; sig.: significant; WC: waist
EC: epicatechin; EGCG:
epigallocatechin
gallate.
circumference;
WHR: waist to hip ratio; wk:
Hsu et al. found no statistical differences in several parameters (fasting glucose, insulin, HOMA-IR, HbA1c, lipoproteins, hormones (leptin, ghrelin, adiponectin), blood pressure, anthropometrics) between a decaffeinated GTE-supplemented group (3 × 500 mg/day; 856 mg EGCG) of T2D obese patients and the placebo group. However, 16 weeks of treatment led to a significant reduction of HbA1c, HOMA-IR index and the insulin level from the baseline to the end of the treatment (within-group changes) [84].
A randomized controlled trial (RCT) assessed the effect of a daily consumption of 100 g of flavanol-rich dark chocolate (FRC; 1008 mg total phenols and 36.12 g C) for 15 days on IRes and showed a significant reduction of HOMA-IR, an enhancement of the insulin sensitivity and an increase in the beta cell activity in hypertensive individuals with impaired glucose tolerance [82]. Furthermore, the consumption of FRC decreased TC and LDL when compared to the baseline values and the control, but it did not affect HDL and TG. High-sensitivity C-reactive protein (hsCRP) did not change either [82]. Similarly, the daily consumption of 27 g FRC (850 mg flavan-3-ols and 90 mg EC) and 100 mg isoflavones for one year reduced HOMA-IR, LDL, and the TC:HDL ratio and increased QUICKI and the HDL:LDL ratio in postmenopausal women with T2D. These metabolic improvements resulted in a lower 10-year total coronary heart disease (CHD) risk compared to the control [83].
A clinical trial investigating the effects of the daily intake of green tea (340 mL) and GTE (582.8 mg catechins/day) for 12 weeks showed an increased insulin level and an increase in the adiponectin level (only within-group changes) in subjects with T2D. Furthermore, there was a reduction of FFA compared to the baseline and a decrease of the TC level when compared to the control. Fasting blood glucose and HbA1c remained unchanged [85].
The daily supplementation with a GTE (one packet/day; 544 mg polyphenols, 456 mg C) for two months improved the HbA1c value in individuals with glucose abnormalities when compared to the baseline. No other parameters of glucose metabolism (fasting blood glucose, insulin, HOMA-IR) or lipid metabolism (TC, LDL, HDL, TG) were affected by the intervention [86]. Furthermore, the daily intake of a GTE powder did not improve the hsCRP level [86]. In individuals with borderline T2D or T2D, GTE (544 mg polyphenols, 456 mg C) decreased the IRes, as measured by HOMA-IR, the fasting blood glucose, the insulin levels and the HbA1c when compared to the baseline. No significant differences between the intervention and the control were observed [87].
However, not all studies showed unambiguous protective effects against diabetes. A daily intake of 9 g green tea in 900 mL hot water for four weeks did not affect the IRes, the fasting blood glucose or the insulin concentration in subjects suffering from T2D. Furthermore, no beneficial effects on the lipid metabolism, hsCRP or IL-6 were shown [88]. Similarly, daily supplementation of GSE (2 × 300 mg/day) for four weeks did not improve the IRes [89]. Fasting blood glucose, insulin level, and HOMA-IR remained unchanged in individuals with T2D and a high cardiovascular risk. Moreover, the supplementation did not result in an improved lipoprotein status apart from a decrease in TC level. However, the regular intake of the GSE significantly decreased fructosamine concentration, decreased hsCRP, and increased the reduced glutathione (GSH) compared to the baseline value. Total antioxidant status (TAOS) and the concentration of oxidized glutathione (GSSG) remained unchanged [89].
A daily intake of 2.5 g cacao powder (ACTICOA TM; 207.5 mg flavanols) for 12 weeks did not enhance the glucose or lipid metabolism in T2D hypertensive patients [90]. The consumption of two cacao beverages per day (2 × 28 g cacao powder/day) containing 180 mg, 400 mg or 900 mg flavanols on five consecutive days did not affect fasting and postprandial glucose parameters in obese individuals who were at risk of IRes either. However, hsCRP, 8-isoprostane, and IL-6 decreased as the dose of flavanols increased. These effects were only significant when compared to the baseline values but not when compared to the control [91].
Acute cacao studies also showed no distinct improvement of postprandial glycemia and the insulin response in participants with T2D. The acute supplementation of a cacao beverage (960 mg polyphenols; 480 mg flavanols) with a high-fat fast-food-style breakfast (766 kcal, 50 g fat) elicited a higher insulin response and a decreased IRes, as measured by HOMA-IR. HDL increased while the concentration of TC, LDL, TG, and hsCRP remained unchanged [92]. No effects could be observed after an acute supplementation of 2.5 g cacao powder (ACTICOATM; 40.4 mg EC) with a diabetic-suitable breakfast in hypertensive, overweight or obese subjects with T2D [93].
### **6. Concluding Remarks**
Increasing evidence suggests that flavan-3-ols are responsible for the protective role of certain foods, such as green tea, against diabetes. Possible molecular mechanisms by which they could prevent or treat diabetes include the promotion of beta cell functionality and viability, the amelioration of glucose transport in muscle and adipose tissue by the promotion of the insulin signaling pathway, the enhancement of the incretin effect, and the decrease of endogenous glucose production.
Microbial metabolites of flavan-3-ols are suggested to be the actual active form by which these compounds exert their potential health benefits, such as the antidiabetic effect. However, the evidence regarding this is still scarce and only few studies assessed the effects of flavan-3-ol specific microbial metabolites.
One of the determinants of the possible antidiabetic effect of flavan-3-ols seems to be their concentration. In this regard, when interpreting the results of in vitro studies, it is necessary to consider not only the bioavailability of the compounds investigated but also their physiological plasma concentration after absorption. Some of the studies used higher concentrations than those found in human plasma after ingestion. Conjugated flavan-3-ols have been detected in plasma in low nanomolar ranges [94,95]. Their colonic metabolites phenyl-γ-valerolactones and phenylvaleric acids have been detected in plasma at concentrations under 1 μM [95–99]. The lower weight phenolics are usually found in plasma at concentrations lower than 0.5 μM [96–99], although phenylacetic acid, protocatechuic acid, and hippuric acid were also detected at concentrations ≈ 40 μmol/L [96–100].
Although many studies used physiological concentrations of the compound and microbial metabolites, some tested higher concentrations and suggested that, in some cases, supraphysiological ranges could induce an opposite response to that from lower ranges. A careful evaluation of the flavan-3-ol dose would, therefore, be needed when used as nutraceutical.
Under physiological concentrations, flavan-3-ols and their microbial metabolites exert different biological activities at varying concentrations. The concentration needed to exert a specific function by a particular compound might not be the same than the one needed to exert another biological activity. For example, 3,4-DHPA IC50 on vasodilation is lower than on AGEs formation [101,102]. In the prevention and management of diabetes, low concentrations of flavan-3-ols and their microbial metabolites could influence determined molecular mechanisms while higher concentrations could be needed to positively influence other mechanisms, as shown in many of the presented studies.
Although increasing evidence supports the stated mechanisms of action, the results of many studies are inconclusive, with some of them exhibiting contradictory outcomes or even negative effects. Possible reasons for the varying results could be not only the different compound concentrations used, but also the different methodologies used in each study. As for the in vitro and animal studies, a variety of models for diabetes was used. In the literature described in this review, the most frequent in vitro techniques used insulin-secreting cell line INS-1 and pancreatic beta cell lines, and they also included in vitro studies on glucose uptake mainly in skeletal muscle cells but also in 3T3-L1 adipocytes. In addition, assays on α-amylase inhibition and inhibition of α-glucosidase activity were performed. In vivo studies included spontaneous diabetic obese animal models, mice genetically predisposed to obesity and T2D, and others used chemicals to induce the disease, mainly streptozotocin. However, not all models were specific for diabetes mellitus. Therefore, although many of these studies showed positive effects on diabetic parameters, concluding that they are beneficial for the treatment of diabetes mellitus would not be appropriate. For the proper elucidation of the effect of a substance regarding diabetes management suitable models are required. For this reason, more in vitro and animal studies using the adequate models for the disease should be performed.
Moreover, it is worth mentioning that some of the used treatment samples often include other bioactive components. Therefore, the flavan-3-ol fraction does not always represent the exclusive component present in the sample, which must be taken into account when attributing positive effects to these compounds.
In the case of human clinical trials, the compliance of the patients to the treatment was not measured in all cases and the diet during the intervention was not always recorded. In addition, other confounders, such as body composition, were often insufficiently registered. Therefore, these factors could have influenced the studies' results. Moreover, not only the concentration of the flavan-3-ols might be a determinant of the possible antidiabetic effect, but also, the duration of the intervention could be a determinant. Further methodological weaknesses in some of the human trials presented are the absence of a wash-out phase, or not choosing the dietary restrictions of the control groups appropriately.
It is known that the individual gut microbiota composition has an influence on both the bioavailability and the metabolization of flavan-3-ols [3]. However, none of the included clinical trials investigated the bioavailability and the metabolization of the ingested flavan-3-ols in the study population. Therefore, no conclusion can be reached about whether there is a positive association between the blood concentration of the flavan-3-ols or their metabolites, the administered dose, and the putative effect. That is, it is not possible to know to what extent different metabolic effects are related to a different bioavailability and metabolization of the flavan-3-ols by individuals. In order to better understand the effects of flavan-3-ols and their metabolites on the prevention and management of diabetes, it is relevant to record not only metabolic parameters after treatment but also the pharmacokinetics of these substances.
For all these reasons, the use of homogeneous and more appropriate methods is essential for the clarification of flavan-3-ol's antidiabetic effect and mechanisms of action.
**Author Contributions:** E.M.C., L.J., and M.-C.S. wrote the paper; M.-C.S. conceived and supervised the project. All authors have read and agreed to the published version of the manuscript.
**Funding:** This work was funded by the Department of Nutrition and Food Sciences, Nutrition and Microbiota, University of Bonn and supported by the grant no 01EA1707 of the German Federal Ministry of Education and Research.
**Acknowledgments:** The authors would like to thank Katherine Macmillan for the professional English editing service.
**Conflicts of Interest:** The authors declare no conflicts of interest.
### **References**
© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
MDPI St. Alban-Anlage 66 4052 Basel Switzerland Tel. +41 61 683 77 34 Fax +41 61 302 89 18 www.mdpi.com
*Nutrients* Editorial Office E-mail: [email protected] www.mdpi.com/journal/nutrients
MDPI St. Alban-Anlage 66 4052 Basel Switzerland
Tel: +41 61 683 77 34 Fax: +41 61 302 89 18 | doab | 2025-04-07T03:56:58.628684 | 1-5-2021 17:49 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "003cf560-3166-422f-9a1c-98b1e971efa6",
"url": "https://mdpi.com/books/pdfview/book/3241",
"author": "",
"title": "Dietary Polyphenols and Human Health",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783039437931",
"section_idx": 15
} |
003cf94e-81c7-4214-b0fe-134420ebdba8.0 | *Edited by Elmer P. Dadios*
Fuzzy Logic is becoming an essential method of solving problems in all domains. It gives tremendous impact on the design of autonomous intelligent systems. The purpose of this book is to introduce Hybrid Algorithms, Techniques, and Implementations of Fuzzy Logic. The book consists of thirteen chapters highlighting models and principles of fuzzy logic and issues on its techniques and implementations. The intended readers of this book are engineers, researchers, and graduate students interested in fuzzy logic systems.
ISBN 978-953-51-0393-6
Fuzzy Logic - Algorithms, Techniques and Implementations
Photo by Nattakit / iStock
## Fuzzy Logic Algorithms, Techniques and Implementations
*Edited by Elmer P. Dadios*
**FUZZY LOGIC –**
Edited by **Elmer P. Dadios**
**ALGORITHMS, TECHNIQUES**
**AND IMPLEMENTATIONS**
| doab | 2025-04-07T03:56:58.645422 | 20-4-2021 17:29 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/3.0/",
"book_id": "003cf94e-81c7-4214-b0fe-134420ebdba8",
"url": "https://mts.intechopen.com/storage/books/2274/authors_book/authors_book.pdf",
"author": "",
"title": "Fuzzy Logic",
"publisher": "IntechOpen",
"isbn": "9789535103936",
"section_idx": 0
} |
003cf94e-81c7-4214-b0fe-134420ebdba8.1 | **FUZZY LOGIC – ALGORITHMS, TECHNIQUES AND IMPLEMENTATIONS**
Edited by **Elmer P. Dadios**
### **Fuzzy Logic - Algorithms, Techniques and Implementations**
http://dx.doi.org/10.5772/2663 Edited by Elmer P. Dadios
### **Contributors**
Kazuhisa Takemura, Toshihisa Sato, Motoyuki Akamatsu, Hashim Habiballa, Thoedtida Thipparat, Amin Parvizi, Agnes Achs, Volodymyr I Ponomaryov, Heydy Castillejos, Marcelo Moutinho, Arianit Maraj, Adrian Shehu, Adriana Florescu, Claudiu Oros, Anamaria Radoi, Jorge Ropero, Ariel Gomez, Alejandro Carrasco, Carlos Leon, Joaquin Luque, Arturo Tellez, Luis Villa, Heron Molina, Elsa Rubio, Ildar Batyrshin, Jorma Kalevi Mattila
### **© The Editor(s) and the Author(s) 2012**
The moral rights of the and the author(s) have been asserted.
All rights to the book as a whole are reserved by INTECH. The book as a whole (compilation) cannot be reproduced, distributed or used for commercial or non-commercial purposes without INTECH's written permission. Enquiries concerning the use of the book should be directed to INTECH rights and permissions department ([email protected]).
Violations are liable to prosecution under the governing Copyright Law.
Individual chapters of this publication are distributed under the terms of the Creative Commons Attribution 3.0 Unported License which permits commercial use, distribution and reproduction of the individual chapters, provided the original author(s) and source publication are appropriately acknowledged. If so indicated, certain images may not be included under the Creative Commons license. In such cases users will need to obtain permission from the license holder to reproduce the material. More details and guidelines concerning content reuse and adaptation can be foundat http://www.intechopen.com/copyright-policy.html.
### **Notice**
Statements and opinions expressed in the chapters are these of the individual contributors and not necessarily those of the editors or publisher. No responsibility is accepted for the accuracy of information contained in the published chapters. The publisher assumes no responsibility for any damage or injury to persons or property arising out of the use of any materials, instructions, methods or ideas contained in the book.
First published in Croatia, 2012 by INTECH d.o.o. eBook (PDF) Published by IN TECH d.o.o. Place and year of publication of eBook (PDF): Rijeka, 2019. IntechOpen is the global imprint of IN TECH d.o.o. Printed in Croatia
Legal deposit, Croatia: National and University Library in Zagreb
Additional hard and PDF copies can be obtained from [email protected]
Fuzzy Logic - Algorithms, Techniques and Implementations Edited by Elmer P. Dadios p. cm. ISBN 978-953-51-0393-6 eBook (PDF) ISBN 978-953-51-5686-4
| doab | 2025-04-07T03:56:58.645563 | 20-4-2021 17:29 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/3.0/",
"book_id": "003cf94e-81c7-4214-b0fe-134420ebdba8",
"url": "https://mts.intechopen.com/storage/books/2274/authors_book/authors_book.pdf",
"author": "",
"title": "Fuzzy Logic",
"publisher": "IntechOpen",
"isbn": "9789535103936",
"section_idx": 1
} |
003cf94e-81c7-4214-b0fe-134420ebdba8.2 | We are IntechOpen, the world's leading publisher of Open Access books Built by scientists, for scientists
4,000+
Open access books available
116,000+
International authors and editors
120M+
Downloads
Our authors are among the
Top 1%
most cited scientists
12.2%
Contributors from top 500 universities
Selection of our books indexed in the Book Citation Index in Web of Science™ Core Collection (BKCI)
## Interested in publishing with us? Contact [email protected]
Numbers displayed above are based on latest data collected. For more information visit www.intechopen.com
| doab | 2025-04-07T03:56:58.646053 | 20-4-2021 17:29 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/3.0/",
"book_id": "003cf94e-81c7-4214-b0fe-134420ebdba8",
"url": "https://mts.intechopen.com/storage/books/2274/authors_book/authors_book.pdf",
"author": "",
"title": "Fuzzy Logic",
"publisher": "IntechOpen",
"isbn": "9789535103936",
"section_idx": 2
} |
003cf94e-81c7-4214-b0fe-134420ebdba8.3 | **Meet the editor**
Prof Elmer P. Dadios finished his doctoral degree at Loughborough University, United Kingdom in 1996. He was a recipient of the Philippines Department of Science and Technology (DOST) 50 Men and Women of Science and Technology; DOST Scholar Achievers); The National Research Council of the Philippines Basic Research Achievement Award; The National Academy of Science
and Technology (NAST) Outstanding Scientific Paper Award; The De La Salle University Miguel Febres Cordero Research Award. Currently, Dr. Dadios is a University Fellow of the De La Salle University and holds the University's highest faculty rank of Full Professor 10. He is the president of the NEURONEMECH Inc.
He has been a consultant for Robotics and Automation in the Philippine government and private corporations. He is the founder of the IEEE Computational Intelligence Society - Philippines Chapter. He is the Founder and President of the Mechatronics and Robotics Society of the Philippines.
Contents
**Preface IX**
**Part 1 Hybrid Fuzzy Logic Algorithms 1**
**Fuzzy Set Model and Data Analysis 3**
Chapter 4 **Standard Fuzzy Sets and some Many-Valued Logics 75**
**System in Supply Chain Management Evaluation 115**
**Multivalued Knowledge-Base 25**
Chapter 3 **Resolution Principle and Fuzzy Logic 55**
Chapter 5 **Parametric Type-2 Fuzzy Logic Systems 97** Arturo Tellez, Heron Molina, Luis Villa,
Chapter 6 **Application of Adaptive Neuro Fuzzy Inference**
**Algorithms in Wavelet Domain 127** Heydy Castillejos and Volodymyr Ponomaryov
Chapter 8 **Fuzzy Logic Approach for QoS Routing Analysis 149**
**Part 2 Techniques and Implementation 147**
Adrian Shehu and Arianit Maraj
Elsa Rubio and Ildar Batyrshin
Chapter 1 **Ambiguity and Social Judgment:**
Kazuhisa Takemura
Chapter 2 **From Fuzzy Datalog to**
Agnes Achs
Hashim Habiballa
Jorma K. Mattila
Thoedtida Thipparat
Chapter 7 **Fuzzy Image Segmentation**
### Contents
### **Preface** XI
- **Part 2 Techniques and Implementation 147**
### Preface
*Algorithm* is used to define the notion of decidability. It is a set of rules that precisely defines a sequence of operations. This is essential for computers to process information. Computer programs contain algorithms that detail specific instructions a computer should perform to carry out a specified task. The traditional computer program performs specific instructions sequentially, and uses crisp values of information which do not support uncertainties. Thus, when a problem is getting harder and becoming more complex, alternative algorithms are required in order to obtain accurate solutions. To this date, the quest of discovering new algorithms is in a race. The fuzzy logic algorithm is one of very strong contender s in this race because fuzzy logic exhibits reasoning power similar to how humans reason out. Fuzzy logic is able to process incomplete data and provide approximate solutions to problems other methods find difficult to solve. Fuzzy logic was first proposed by Lotfi A. Zadeh of the University of California at Berkeley in 1965. This is based on the idea that humans do not think in terms of crisp numbers, but rather in terms of concepts. The degree of membership of an object in a concept may be partial, with an object being partially related with many concepts. By characterizing the idea of partial membership in concepts, fuzzy logic is better able to convert natural language control strategies in a form usable by machines.
This book presents Algorithms, Techniques, and Implementations of fuzzy logic. It is categorized into two sections, namely:
In section one, there are seven chapters that focus on hybrid fuzzy logic algorithms and methodology:
- Application of Adaptive Neuro Fuzzy Inference System in Supply Chain Management Evaluation
- Fuzzy Image Segmentation Algorithms in Wavelet Domain
In section two, there are seven chapters that focus on fuzzy logic modeling and implementations, particularly:
The contributions to this book clearly reveal the fuzzy logic models, techniques, and implementation which are very important for the development of new technologies. I hope the readers of this book will find it a unique and significant source of knowledge and reference for the years to come.
> **Elmer P. Dadios** University Fellow and Full Professor, Department of Manufacturing Engineering and Management, De La Salle University, Philippines
## **Part 1**
**Hybrid Fuzzy Logic Algorithms**
**1**
*Japan*
Kazuhisa Takemura *Waseda University,*
**Ambiguity and Social Judgment:**
**Fuzzy Set Model and Data Analysis**
Comparative judgment is essential in human social lives. Comparative judgment is a type of human judgment procedure, in which the evaluator is asked which alternative is preferred (e.g., "Do you prefer Brand A to Brand B?" or "How do you estimate the probability of choosing Brand A over Brand B when you compare the two brands? "). This type of judgment is distinguished from absolute judgment, in which the evaluator is asked to assess the attractiveness of an object (e.g., "How much do you like this brand on
The ambiguity of social judgment has been conceptualized by the fuzzy set theory. The fuzzy set theory provides a formal framework for the presentation of the ambiguity. Fuzzy sets were defined by Zadeh(1965) who also outlined how they could be used to characterize complex systems and decision processes ( Zadeh, 1973). Zadeh argues that the capacity of humans to manipulate fuzzy concepts should be viewed as a major asset, not a liability. The complexities in the real world often defy precise measurement and fuzzy logic defines concepts and its techniques provide a mathematical method able to deal with thought processes which are often too imprecise and ambiguous to deal with by classical
This chapter introduces a model of ambiguous comparative judgment (Takemura,2007) and provides a method of data analysis for the model, and then shows some examples of the data analysis of social judgments. Comparative judgments in social situations often involve ambiguity with regard to confidence, and people may be unable to make judgments without some confidence intervals. To measure the ambiguity (or vagueness) of human judgment, the fuzzy rating method has been proposed and developed (Hesketh, Pryor, Gleitzman, & Hesketh, 1988). In fuzzy rating, respondents select a representative rating point on a scale and indicate higher or lower rating points, depending on the relative ambiguity of their judgment. For example, fuzzy rating would be useful for perceived temperature, with the evaluator indicating a representative value and lower and upper values. This rating scale allows for asymmetries and overcomes the problem, identified by Smithson (1987), of researchers arbitrarily deciding the most representative value from a range of scores. By making certain simplifying assumptions (which is not uncommon in fuzzy set theory), the rating can be viewed as an L-R fuzzy number, thereby making the use of fuzzy set
**1. Introduction**
a scale of 0 to 100?").
mathematical techniques.
## **Ambiguity and Social Judgment: Fuzzy Set Model and Data Analysis**
Kazuhisa Takemura *Waseda University, Japan*
### **1. Introduction**
Comparative judgment is essential in human social lives. Comparative judgment is a type of human judgment procedure, in which the evaluator is asked which alternative is preferred (e.g., "Do you prefer Brand A to Brand B?" or "How do you estimate the probability of choosing Brand A over Brand B when you compare the two brands? "). This type of judgment is distinguished from absolute judgment, in which the evaluator is asked to assess the attractiveness of an object (e.g., "How much do you like this brand on a scale of 0 to 100?").
The ambiguity of social judgment has been conceptualized by the fuzzy set theory. The fuzzy set theory provides a formal framework for the presentation of the ambiguity. Fuzzy sets were defined by Zadeh(1965) who also outlined how they could be used to characterize complex systems and decision processes ( Zadeh, 1973). Zadeh argues that the capacity of humans to manipulate fuzzy concepts should be viewed as a major asset, not a liability. The complexities in the real world often defy precise measurement and fuzzy logic defines concepts and its techniques provide a mathematical method able to deal with thought processes which are often too imprecise and ambiguous to deal with by classical mathematical techniques.
This chapter introduces a model of ambiguous comparative judgment (Takemura,2007) and provides a method of data analysis for the model, and then shows some examples of the data analysis of social judgments. Comparative judgments in social situations often involve ambiguity with regard to confidence, and people may be unable to make judgments without some confidence intervals. To measure the ambiguity (or vagueness) of human judgment, the fuzzy rating method has been proposed and developed (Hesketh, Pryor, Gleitzman, & Hesketh, 1988). In fuzzy rating, respondents select a representative rating point on a scale and indicate higher or lower rating points, depending on the relative ambiguity of their judgment. For example, fuzzy rating would be useful for perceived temperature, with the evaluator indicating a representative value and lower and upper values. This rating scale allows for asymmetries and overcomes the problem, identified by Smithson (1987), of researchers arbitrarily deciding the most representative value from a range of scores. By making certain simplifying assumptions (which is not uncommon in fuzzy set theory), the rating can be viewed as an L-R fuzzy number, thereby making the use of fuzzy set
Ambiguity and Social Judgment: Fuzzy Set Model and Data Analysis 5
applied for the analysis of the former model, and a fuzzy logistic regression model
*X1 =* (*X11, X12,…,X1n*), *X2* = (*X21, X22,…,X2n*),…, *Xm =* (*Xm1, Xm2,…,Xmn*), where *Xij* (*i = 1.m; j = 1.,n*) is the value of alternative *Xi* on dimension *j.* Note that the components of *Xi* may be
The relational structure < X, > is a weak order if, and only if, for all *Xa, Xb, Xc*, the
However, the weak order relation is not always assumed in this paper. That is, transitivity
As a classical preference relation is a subset of X × X , is a classical set often viewed as a
Note that "iff" is short for "if and only if" and {0,1} is called the valuation set. If the valuation set is allowed to be the real interval [0,1], is called a fuzzy preference relation.
*µa*: X × X → [0,1].
Ambiguous preference relations are defined as a fuzzy set of *X* ×*X* × *S*, where *S* is a subset of one-dimensional real number space. *S* is interpreted as a domain of preference strength. *S*
:: X × X × *S* → [0,1].
1 iff X X 0 iff not(X X )
a b a b
β
is defined as:
.
*Xn* be a set of multidimensional alternatives with elements of the form
(Takemura, 2004) was proposed for the analysis of the latter model.
**2.2.1 Definition 1: Set of multidimensional alternatives**
ambiguous linguistic variables rather than crisp numbers.
Let be a binary relation on X, that is, is a subset of X × X*.*
1. Connectedness (Comparability): *Xa, Xb* or *Xb Xa,* 2. Transitivity: If *Xa Xb* and *Xb Xc*, then *Xa Xc.*
**2.2.3 Definition 3: Fuzzy preference relation**
characteristic function *c* from X × X to {0,1} such that:
That is, the membership function *µa* is defined as:
**2.2.4 Definition 4: Ambiguous preference relation**
or connectedness may be violated in the preference relations.
*c*(Xj Xk) =
may be bounded, for example, *S* = [0,1]. The membership function *µ*
*µ*β
**2.2.2 Definition 2: Classic preference relation**
**2.2 Assumptions of the model**
following two axioms are satisfied.
Let X *= X1*
× *X2* × *….* ×
theoretical operations possible (Hesketh et al., 1988; Takemura, 2000). Lastly, numerical illustrations of psychological experiments are provided to examine the ambiguous comparative judgment model (Takemura, 2007) using the proposed data analysis.
### **2. Model of ambiguous comparative judgment**
### **2.1 Overview of ambiguous comparative judgment and the judgment model**
Social psychological theory and research have demonstrated that comparative evaluation has a crucial role in the cognitive processes and structures that underlie people's judgments, decisions, and behaviors(e.g.,Mussweiler,2003). People comparison processes are almost ubiquitous in human social cognition. For example, people tend to compare their performance of others in situations that are ambiguous (Festinger,1954). It is also obvious that they are critical in forming personal evaluations, and making purchase decisions (Kühberger,,.Schulte-Mecklenbeck, & Ranyard, 2011; Takemura,2011).
The ambiguity or vagueness is inherent in people's comparative social judgment. Traditionally, psychological and philosophical theories implicitly had assumed the ambiguity of thought processes ( Smithson, 1987, 1989). For example, Wittgenstein (1953) pointed out that lay categories were better characterized by a " family resemblance" model which assumed vague boundaries of concepts rather than a classical set-theoretic model. Rosch (1975) and Rosch & Mervice(1975) also suggested vagueness of lay categories in her prototype model and reinterpret-ed the family resemblance model. Moreover, the social judgment theory (Sherif & Hovland,1961) and the information integration theory (Anderson,1988) for describing judgment and decision making assumed that people evaluate the objects using natural languages which were inherently ambiguous. However, psychological theories did not explicitly treat the ambiguity in social judgment with the exception of using random error of judgment.
Takemura (2007) proposed fuzzy set models that explain ambiguous comparative judgment in social situations. Because ambiguous comparative judgment may not always hold transitivity and comparability properties, the models assume parameters based on biased responses that may not hold transitivity and comparability properties. The models consist of two types of fuzzy set components for ambiguous comparative judgment. The first is a fuzzy theoretical extension of the additive difference model for preference, which is used to explain ambiguous preference strength and does not always assume judgment scale boundaries, such as a willing to pay (WTP) measure. The second type of model is a fuzzy logistic model of the additive difference preference, which is used to explain ambiguous preference in which preference strength is bounded, such as a probability measure (e.g., a certain interval within a bounded interval from 0 to 100%).
Because judgment of a bounded scale, such as a probability judgment, causes a methodological problem when fuzzy linear regression is used, a fuzzy logistic function to prevent this problem was proposed. In both models, multi-attribute weighting parameters and all attribute values are assumed to be asymmetric fuzzy L-R numbers. For each model, A method of parameter estimation using fuzzy regression analysis was proposed. That is, a fuzzy linear regression model using the least squares method (Takemura, 1999, 2005) was applied for the analysis of the former model, and a fuzzy logistic regression model (Takemura, 2004) was proposed for the analysis of the latter model.
### **2.2 Assumptions of the model**
4 Fuzzy Logic – Algorithms, Techniques and Implementations
theoretical operations possible (Hesketh et al., 1988; Takemura, 2000). Lastly, numerical illustrations of psychological experiments are provided to examine the ambiguous
Social psychological theory and research have demonstrated that comparative evaluation has a crucial role in the cognitive processes and structures that underlie people's judgments, decisions, and behaviors(e.g.,Mussweiler,2003). People comparison processes are almost ubiquitous in human social cognition. For example, people tend to compare their performance of others in situations that are ambiguous (Festinger,1954). It is also obvious that they are critical in forming personal evaluations, and making purchase decisions
The ambiguity or vagueness is inherent in people's comparative social judgment. Traditionally, psychological and philosophical theories implicitly had assumed the ambiguity of thought processes ( Smithson, 1987, 1989). For example, Wittgenstein (1953) pointed out that lay categories were better characterized by a " family resemblance" model which assumed vague boundaries of concepts rather than a classical set-theoretic model. Rosch (1975) and Rosch & Mervice(1975) also suggested vagueness of lay categories in her prototype model and reinterpret-ed the family resemblance model. Moreover, the social judgment theory (Sherif & Hovland,1961) and the information integration theory (Anderson,1988) for describing judgment and decision making assumed that people evaluate the objects using natural languages which were inherently ambiguous. However, psychological theories did not explicitly treat the ambiguity in social judgment with the
Takemura (2007) proposed fuzzy set models that explain ambiguous comparative judgment in social situations. Because ambiguous comparative judgment may not always hold transitivity and comparability properties, the models assume parameters based on biased responses that may not hold transitivity and comparability properties. The models consist of two types of fuzzy set components for ambiguous comparative judgment. The first is a fuzzy theoretical extension of the additive difference model for preference, which is used to explain ambiguous preference strength and does not always assume judgment scale boundaries, such as a willing to pay (WTP) measure. The second type of model is a fuzzy logistic model of the additive difference preference, which is used to explain ambiguous preference in which preference strength is bounded, such as a probability measure (e.g., a
Because judgment of a bounded scale, such as a probability judgment, causes a methodological problem when fuzzy linear regression is used, a fuzzy logistic function to prevent this problem was proposed. In both models, multi-attribute weighting parameters and all attribute values are assumed to be asymmetric fuzzy L-R numbers. For each model, A method of parameter estimation using fuzzy regression analysis was proposed. That is, a fuzzy linear regression model using the least squares method (Takemura, 1999, 2005) was
comparative judgment model (Takemura, 2007) using the proposed data analysis.
**2.1 Overview of ambiguous comparative judgment and the judgment model**
(Kühberger,,.Schulte-Mecklenbeck, & Ranyard, 2011; Takemura,2011).
**2. Model of ambiguous comparative judgment**
exception of using random error of judgment.
certain interval within a bounded interval from 0 to 100%).
### **2.2.1 Definition 1: Set of multidimensional alternatives**
Let X *= X1*× *X2* × *….* × *Xn* be a set of multidimensional alternatives with elements of the form *X1 =* (*X11, X12,…,X1n*), *X2* = (*X21, X22,…,X2n*),…, *Xm =* (*Xm1, Xm2,…,Xmn*), where *Xij* (*i = 1.m; j = 1.,n*) is the value of alternative *Xi* on dimension *j.* Note that the components of *Xi* may be ambiguous linguistic variables rather than crisp numbers.
### **2.2.2 Definition 2: Classic preference relation**
Let be a binary relation on X, that is, is a subset of X × X*.*
The relational structure < X, > is a weak order if, and only if, for all *Xa, Xb, Xc*, the following two axioms are satisfied.
However, the weak order relation is not always assumed in this paper. That is, transitivity or connectedness may be violated in the preference relations.
### **2.2.3 Definition 3: Fuzzy preference relation**
As a classical preference relation is a subset of X × X , is a classical set often viewed as a characteristic function *c* from X × X to {0,1} such that:
$$c(\mathsf{X}\_{\mathsf{j}} \succ \mathsf{X}\_{\mathsf{k}}) = \begin{cases} 1 & \text{iff} \\ 0 & \text{iff} \end{cases} \qquad \begin{array}{c} \mathsf{X}\_{\mathsf{a}} \succ \mathsf{X}\_{\mathsf{b}} \\ \text{not}(\mathsf{X}\_{\mathsf{a}} \succ \mathsf{X}\_{\mathsf{b}}) \cdot \mathsf{X}\_{\mathsf{c}} \end{array}$$
Note that "iff" is short for "if and only if" and {0,1} is called the valuation set. If the valuation set is allowed to be the real interval [0,1], is called a fuzzy preference relation. That is, the membership function *µa* is defined as:
$$
\mu\_a \colon \mathcal{X} \times \mathcal{X} \to [0, 1].
$$
### **2.2.4 Definition 4: Ambiguous preference relation**
Ambiguous preference relations are defined as a fuzzy set of *X* ×*X* × *S*, where *S* is a subset of one-dimensional real number space. *S* is interpreted as a domain of preference strength. *S* may be bounded, for example, *S* = [0,1]. The membership function *µ*βis defined as:
$$
\mu\_{\land} \dots \mathbb{X} \times \mathbb{X} \times \mathbb{S} \to [0, 1].
$$
Ambiguity and Social Judgment: Fuzzy Set Model and Data Analysis 7
*abn* ⊗ *(Xan*○-*-Xbn*
where *log*, ○÷ ,⊗ ,○+, and ○- are logarithmic, division, product , additive, and difference
The second model of the equation (2) is the model for [0,1]. However, the model could apply to not only the interval [0,1] but also any finite interval [a,b](a<b). Therefore, the model of
Non-comparability and intransitivity properties are explained if a threshold of comparative judgment is assumed, if intransitivity is indicated by the necessity measure of fuzzy comparative relation resulting from the existence of the threshold, and if a necessity
(1
difference model and the logistic regression model, respectively. Assuming the above relation of (3) or (4), it is clear that intransitivity and non-comparability hold in the
Traditional approaches to the measurement of social judgment have involved methods such as the semantic differential, the Likert scale, or the Thurstone scale. Although insights into the ambiguous nature of social judgment were identified early in the development of measurement of social judgment, the subsequent methods used failed to capture this ambiguity, no doubt because traditional mathematics was not well developed for dealing
In order to measure the vagueness of human judgment, the fuzzy rating method has recently been proposed and developed (Hesketh et al.,1988; Takemura,1996). In the fuzzy rating method, respondents select a representative rating point on a scale and indicate lower or upper rating points if they wish depending upon the relative vagueness of their judgment (see Figure 2). For example, the fuzzy rating method would be useful for measuring perceived temperature indicating the representative value and the lower or upper values. This rating scale allows for asymmetries, and overcomes the problem, identified by Smithson (1987), of researchers arbitrarily deciding most representative value from a range of scores. By making certain simplifying assumptions ( not uncommon within fuzzy set theory), the rating can be viewed as a L-R fuzzy number, hence making possible the use of
**3. Fuzzy data analysis for the ambiguous comparative judgment model**
*)* is a necessity measure, and *θ, and Pθ* are threshold parameters for the additive
*Xa Xb iff Nes ( v(Xa Xb)*
)(2)
>θ
○-*p(Xa Xb) )*
> *P*θ
*)* (3)
*)* (4)
*1nXb1)ype correction and drawing figures.mments and*
)=
A
operations based on the extension principle for the fuzzy set, respectively.
the equation (2)is considered to be a special case for the finite interval model.
measure for fuzzy relation does not always lead to comparability. That is,
*Xa Xb iff Nes( p(Xa Xb)* ○÷
l
or
where *Nes (*
・
comparative judgment.
**3.1 Fuzzy rating data and fuzzy set**
fuzzy set theoretic operations).
with vagueness of judgment (Hesketh et al.,1988).
*og ( p(Xa Xb)* ○÷
*Aab0*○+ *Aab1* ⊗
(1
*Xa1*○-*Xb1*
(
○-*p(Xj Xk)*
)○+…○+
**2.2.7 Explaining non-comparability and intransitivity**
Ambiguous preference relation is interpreted as a fuzzified version of a classical characteristic function *c*(*Xa Xb*).
Therefore, the ambiguous preference relation for *Xa Xb* is represented as the fuzzy set *v(Xa Xb)*. For simplicity, *v(Xa Xb)* will be assumed to be an asymmetrical L-R fuzzy number (see Figure 1).
Fig. 1. Example of Ambiguous Preference Relation
### **2.2.5 Additive difference model of ambiguous comparative judgement**
The ambiguous preference relation *v(Xa Xb)* for *Xa Xb* is represented as the following additive difference model using L-R fuzzy numbers:
*v(Xa Xb)*= *Aab0*○+ *Aab1* ⊗ (*Xa1*○-*Xb1*)○+…○+Aabn ⊗ (*Xan*○-*-Xbn*)(1)
where ⊗ , ○+, and ○-are the product, additive, and difference operation based on the extension principle for the fuzzy set, respectively.
The parameter *Ajk*0 involves a response bias owing to presentation order, context effects, and the scale parameter of the dependent variables. The parameter *Ajk*0 would be a fuzzy variable and larger than *Aab0* if *Xa* were more salient than *Xb*. This model can be reduced to the Fuzzy Utility Difference Model (Nakamura, 1992) if multi-attribute weighting parameters are assumed to be crisp numbers, and reduced to the Additive Difference Model (Tversky, 1969) if multi-attribute weighting parameters and the values of multi-attributes are assumed to be crisp numbers.
#### **2.2.6 Logistic model of ambiguous comparative judgement**
Let an ambiguous preference relation that is bounded (e.g., fuzzy probability in [0,1]) be *p*(*Xa Xb*) for *Xa Xb*. *p*(*Xa Xb*) and be represented as the following logistic model using L-R fuzzy numbers:
l*og ( p(Xa Xb)* ○÷ (1 ○-*p(Xj Xk)*)= *1nXb1)ype correction and drawing figures.mments and Aab0*○+ *Aab1* ⊗ (*Xa1*○-*Xb1*) ○+…○+A*abn* ⊗ *(Xan*○-*-Xbn*)(2)
where *log*, ○÷ ,⊗ ,○+, and ○- are logarithmic, division, product , additive, and difference operations based on the extension principle for the fuzzy set, respectively.
The second model of the equation (2) is the model for [0,1]. However, the model could apply to not only the interval [0,1] but also any finite interval [a,b](a<b). Therefore, the model of the equation (2)is considered to be a special case for the finite interval model.
### **2.2.7 Explaining non-comparability and intransitivity**
Non-comparability and intransitivity properties are explained if a threshold of comparative judgment is assumed, if intransitivity is indicated by the necessity measure of fuzzy comparative relation resulting from the existence of the threshold, and if a necessity measure for fuzzy relation does not always lead to comparability. That is,
$$\text{Xa} \succeq \text{Xb} \text{ } \text{iff} \text{ Nes} \left( v(\text{Xa} \succeq \text{Xb}) \ge \theta \right) \tag{3}$$
or
6 Fuzzy Logic – Algorithms, Techniques and Implementations
Ambiguous preference relation is interpreted as a fuzzified version of a classical
Therefore, the ambiguous preference relation for *Xa Xb* is represented as the fuzzy set *v(Xa Xb)*. For simplicity, *v(Xa Xb)* will be assumed to be an asymmetrical L-R fuzzy number
**Ambiguous preference relation**
∈*S*
**as fuzzy set :** *v(Xa Xb)*
characteristic function *c*(*Xa Xb*).
*μβ(Xa Xb)*
1
0
*Preference strength s*
**2.2.5 Additive difference model of ambiguous comparative judgement**
The ambiguous preference relation *v(Xa Xb)* for *Xa Xb* is represented as the following
*Xan*○-*-Xbn*
where ⊗ , ○+, and ○-are the product, additive, and difference operation based on the
The parameter *Ajk*0 involves a response bias owing to presentation order, context effects, and the scale parameter of the dependent variables. The parameter *Ajk*0 would be a fuzzy variable and larger than *Aab0* if *Xa* were more salient than *Xb*. This model can be reduced to the Fuzzy Utility Difference Model (Nakamura, 1992) if multi-attribute weighting parameters are assumed to be crisp numbers, and reduced to the Additive Difference Model (Tversky, 1969) if multi-attribute weighting parameters and the values of multi-attributes
Let an ambiguous preference relation that is bounded (e.g., fuzzy probability in [0,1]) be *p*(*Xa Xb*) for *Xa Xb*. *p*(*Xa Xb*) and be represented as the following logistic model using L-
)(1)
Fig. 1. Example of Ambiguous Preference Relation
additive difference model using L-R fuzzy numbers:
)○+…○+Aabn ⊗ (
extension principle for the fuzzy set, respectively.
**2.2.6 Logistic model of ambiguous comparative judgement**
(see Figure 1).
*v(Xa Xb)*
*Aab0*○+ *Aab1* ⊗
=
(*Xa1*○-*Xb1*
are assumed to be crisp numbers.
R fuzzy numbers:
$$X\_a \succ X\_b \text{ iff } \text{Yes}(\begin{array}{c} p(X\_a \succ X\_b) \oplus \end{array} \begin{pmatrix} \text{ } I \stackrel{\frown}{\ominus} p(X\_a \succ X\_b) \end{pmatrix} \succeq P\_\theta) \tag{4}$$
where *Nes (*・*)* is a necessity measure, and *θ, and Pθ* are threshold parameters for the additive difference model and the logistic regression model, respectively. Assuming the above relation of (3) or (4), it is clear that intransitivity and non-comparability hold in the comparative judgment.
### **3. Fuzzy data analysis for the ambiguous comparative judgment model**
### **3.1 Fuzzy rating data and fuzzy set**
Traditional approaches to the measurement of social judgment have involved methods such as the semantic differential, the Likert scale, or the Thurstone scale. Although insights into the ambiguous nature of social judgment were identified early in the development of measurement of social judgment, the subsequent methods used failed to capture this ambiguity, no doubt because traditional mathematics was not well developed for dealing with vagueness of judgment (Hesketh et al.,1988).
In order to measure the vagueness of human judgment, the fuzzy rating method has recently been proposed and developed (Hesketh et al.,1988; Takemura,1996). In the fuzzy rating method, respondents select a representative rating point on a scale and indicate lower or upper rating points if they wish depending upon the relative vagueness of their judgment (see Figure 2). For example, the fuzzy rating method would be useful for measuring perceived temperature indicating the representative value and the lower or upper values. This rating scale allows for asymmetries, and overcomes the problem, identified by Smithson (1987), of researchers arbitrarily deciding most representative value from a range of scores. By making certain simplifying assumptions ( not uncommon within fuzzy set theory), the rating can be viewed as a L-R fuzzy number, hence making possible the use of fuzzy set theoretic operations).
Ambiguity and Social Judgment: Fuzzy Set Model and Data Analysis 9
Firstly, the convexity of the fuzzy subset is defined as follows: A fuzzy subset A ⊆ R is
Aα= {x| μA(x) ≧ α}, α∈[0,1],
Secondly, the normality of the fuzzy subset is defined as follows: A fuzzy subset A ⊆ R is
One of the most well known fuzzy numbers is the L-R fuzzy number (Dubois & Prade,1980).
where L((x - m)/u) is a increasing monotonic function, R((x - m)/v) is a decreasing
An example of the fuzzy rating scale and of the representation of the rating data using L-R fuzzy number are shown in Figure 3. Note in Figure 3 that representations of variables are
0 100
*M ij x*
*R ij x*
*x*
*R ij x*
Fuzzy rating data
*Xij* :
L-R fuzzy number (Membership function)
abbreviated as follows: *xijL* for *xi j (0lL*, *xijR* for *xi j (0l<sup>R</sup>* , *xijM* for *xi j (1lL* = *xi j (1lR*.
*L ij x*
Fig. 3. Fuzzy Rating Data and Its Representation by L-R Fuzzy Numbers
*M ij x*
*L ij x*
convex if and only if every ordinary
The L-R fuzzy number is defined as follows:
μA(x) = L((x - m)/u), - ∞ < x < m,
= R((x - m)/v), m < x < ∞,
monotonic function, u>0, and v>0.
μ
0.1
normal if and only if ∀x ∈R, max μA(x) = 1.
<sup>x</sup>
∀x ∈R:
= 1, x=m,
subset is convex( That is, in the case of a closed interval of R).
Fig. 2. Example of Fuzzy Rating
A fuzzy set A is defined as follows. Let X denote a universal set, such as X={x1,x2,....,xn}. Then, the membership function μA⊆X by which a fuzzy set A is defined has the form
$$\mathsf{u} \mathsf{A} \,\,\, :\, \mathsf{X} \to \{0, 1\} \,\, \,\, \mathsf{A}$$
where [0,1] denotes the interval of real numbers from 0 to 1, inclusive.
The concept of a fuzzy set is the foundation for analysis where fuzziness exists (Zadeh, 1965). a fuzzy set may be expressed as:
$$\begin{aligned} \mathbf{A} &= \mathfrak{\mu} \mathbf{A}(\mathbf{x}\_{\mathrm{l}}) / \chi\_{\mathbf{1}} \oplus \mathfrak{\mu} \mathbf{A}(\mathbf{x}\_{\mathrm{2}}) / \chi\_{\mathbf{2}} \oplus \dots \quad \dots \quad \oplus \mathfrak{\mu} \mathbf{A}(\mathbf{x}\_{\mathrm{n}}) / \chi\_{\mathbf{n}} \\ &= \mathfrak{\underline{\mathfrak{\mu}}} \quad \mathfrak{\mu} \mathbf{A}(\mathbf{x}\_{\mathrm{i}}) / \chi\_{\mathbf{i}} \end{aligned}$$
where μA(xi) represents the "grade of membership" of Xi in A, or the degree to which Xi satisfies the properties of the set A. It should be noted that here the symbol '"+ " does not refer to the ordinary addition.
μA is called a membership function, or a possibility function. The Xi values are drawn from a global set of all possible values, X. Grade of membership take values between 0 and 1. The membership function has a value of 0 when the properties of the fuzzy set are not at all satisfied, and 1 when the properties of fuzzy set are completely satisfied.
Hesketh et al.(1988) pointed out that fuzzy rating data can be represented as fuzzy sets by making certain implifying assumptions, which are not uncommon within fuzzy set theory. According to Hesketh et al.(1988), those assumptions are:
Making those assumptions, fuzzy rating data in this study can be expressed as a fuzzy number which is a kind of fuzzy set. The concept of the fuzzy number can be defined from the concept of the fuzzy subset(Kaufman & Gupta,1985). The properties of fuzzy numbers are the convexity and the normality of a fuzzy subset.
Firstly, the convexity of the fuzzy subset is defined as follows: A fuzzy subset A ⊆ R is convex if and only if every ordinary
$$A\_{\mathfrak{a}} = \{ \mathbf{x} \mid \mathfrak{p}A \left( \mathbf{x} \right) \ge \mathfrak{a} \}, \mathfrak{a} \in [0, 1],$$
subset is convex( That is, in the case of a closed interval of R).
Secondly, the normality of the fuzzy subset is defined as follows: A fuzzy subset A ⊆ R is normal if and only if
∀x ∈R, max μA(x) = 1.
<sup>x</sup>
One of the most well known fuzzy numbers is the L-R fuzzy number (Dubois & Prade,1980).
The L-R fuzzy number is defined as follows:
∀x ∈R:
8 Fuzzy Logic – Algorithms, Techniques and Implementations
A fuzzy set A is defined as follows. Let X denote a universal set, such as X={x1,x2,....,xn}.
The concept of a fuzzy set is the foundation for analysis where fuzziness exists (Zadeh,
where μA(xi) represents the "grade of membership" of Xi in A, or the degree to which Xi satisfies the properties of the set A. It should be noted that here the symbol '"+ " does not
μA is called a membership function, or a possibility function. The Xi values are drawn from a global set of all possible values, X. Grade of membership take values between 0 and 1. The membership function has a value of 0 when the properties of the fuzzy set are not at all
Hesketh et al.(1988) pointed out that fuzzy rating data can be represented as fuzzy sets by making certain implifying assumptions, which are not uncommon within fuzzy set theory.
3. The fuzzy membership function takes its maximum value, one, at the point on the
4. The extent of the fuzzy support is represented by the horizontal lines to either side
5. The fuzzy membership function tapers uniformly from its value of one at the representative point to a value of zero beyond the fuzzy support or the left and right
Making those assumptions, fuzzy rating data in this study can be expressed as a fuzzy number which is a kind of fuzzy set. The concept of the fuzzy number can be defined from the concept of the fuzzy subset(Kaufman & Gupta,1985). The properties of fuzzy numbers
extensions. The membership value of the lower point and the upper point is 0.
Then, the membership function μA⊆X by which a fuzzy set A is defined has the form
where [0,1] denotes the interval of real numbers from 0 to 1, inclusive.
satisfied, and 1 when the properties of fuzzy set are completely satisfied.
1) Low ambiguity
2)High ambiguity
A = μA(x1)/x1 ⊕ μA(x2)/x2 ⊕ ... ⊕ μA(xn)/xn
According to Hesketh et al.(1988), those assumptions are:
2. The global set X is represented along the horizontal axis.
fuzzy support represented by the representative point.
1. The fuzzy set has a convex membership function.
are the convexity and the normality of a fuzzy subset.
Fig. 2. Example of Fuzzy Rating
1965). a fuzzy set may be expressed as:
μA :X→[0, 1],
= Σ μA(xi)/xi,
refer to the ordinary addition.
of evaluated point.
n
i=1
$$\begin{cases} \mu \text{A(x)} = \text{L}((\text{x} - \text{m})/\text{u}), \text{ } \cdot \text{ } \infty \text{ } \text{x} \le \text{m}, \\\\ \text{ } = \text{1}, \text{ } \text{x} \text{=} \text{m}, \\\\ \text{ } = \text{R}((\text{x} - \text{m})/\text{v}), \text{ } \text{m} \le \text{x} \le \text{w}, \end{cases}$$
where L((x - m)/u) is a increasing monotonic function, R((x - m)/v) is a decreasing monotonic function, u>0, and v>0.
An example of the fuzzy rating scale and of the representation of the rating data using L-R fuzzy number are shown in Figure 3. Note in Figure 3 that representations of variables are abbreviated as follows: *xijL* for *xi j (0lL*, *xijR* for *xi j (0l<sup>R</sup>* , *xijM* for *xi j (1lL* = *xi j (1lR*.
Fig. 3. Fuzzy Rating Data and Its Representation by L-R Fuzzy Numbers
Ambiguity and Social Judgment: Fuzzy Set Model and Data Analysis 11
= ( ) *R k x*0 α
fuzzy coefficient for the *j*-th attribute and the α-level set of fuzzy input data *Xjk*, () ()
j jk a x α α is defined in the same manner, respectively. ( )
assumed to be 1 (a crisp number) for the purpose of estimation for the fuzzy bias parameter
To define the dissimilarity between the predicted and observed values of the dependent
L 2 <sup>k</sup> z ) <sup>α</sup> +( ( ) R <sup>k</sup> y <sup>α</sup> - ( )
The definition in Equation (12) can be applied to interval data as well as to L-R fuzzy numbers. That is, Equation (12) represents the sum of squares for the distance between
To generalize, a dissimilarity indicator representing the square of the distance for L-R fuzzy
In the case of a triangular fuzzy number with *wj =* 1, the above equation is approximately
The proposed method is to estimate fuzzy coefficients using minimization of the sum of *Dk*
R 2 k 1 z ) +( ( ) R k 0 y - ( )
<sup>N</sup> <sup>2</sup> k k 1 Min D =
L 2 k j z ) <sup>α</sup> +( ( ) R <sup>k</sup> <sup>j</sup> y <sup>α</sup> - ( )
=1 (11)
L
<sup>k</sup> z ) <sup>α</sup> (12)
k j z ) <sup>α</sup> ) (13)
k 0 z ) (14)
*2*
0k x <sup>α</sup> and ( )
L R j jk a x α α ,
R 0k x <sup>α</sup> are
j jk a x α α is a product between the lower value of the α-level
R 2
R 2
R 2
(15)
j(h) <sup>1</sup> a 0, ≥ ∈j J (16)
j(h) j(h) <sup>2</sup> a 0, a 0, ≤ ≥∈j J (17)
j(h) <sup>3</sup> a 0, ≤ ∈j J (18)
j(h) j(h) −+ ≥ aa0 (19)
( ) *L k x*0 α
L <sup>k</sup> y <sup>α</sup> - ( )
L L
variable, the following indicator *Dk* ( ) <sup>α</sup> *<sup>2</sup>*was adopted:
n
j=0
L k 0 y - ( )
wj(( ( ) L <sup>k</sup> <sup>j</sup> y <sup>α</sup> - ( )
> L 2 k 0 z ) +( ( ) L k 1 y - ( )
Objective function:
Subject to: <sup>L</sup>
L R
R
L R
where α*j = jh/n, j = 0,...,n*, *h* is an equal interval, and *wj* is a weight for the *j*-th level.
*Dk* ( ) <sup>α</sup> *<sup>2</sup>*=( ( )
numbers can be written as follows:
Dk2=
Dk2 =( ( )
In the above Equation (9), () ()
R R
() () R L
interval data.
represented as:
respecting *k*. That is,
*A0.*
j jk a x α α , or () ()
### **3.2 Analysis of the additive difference type model**
The set of fuzzy input-output data for the *k*-th observation is defined as:
$$\left(\mathbf{Y}\_{\text{abk}}; \mathbf{X}\_{\text{a1k}'} \; \mathbf{X}\_{\text{a2k}'}..., \mathbf{X}\_{\text{ank}} \; \mathbf{X}\_{\text{b1k}'} \mathbf{X}\_{\text{b2k}'}..., \mathbf{X}\_{\text{bnk};}\right) \tag{5}$$
where *Yabk* indicates the *k*-th observation's ambiguous preference for the *a-*th alternative (a) over the *b-*th alternative (b), which represented by fuzzy L-R numbers, and *Xajk* and *Xbjk* are the *j*-th attribute values of the alternatives (a and b) for observation *k*.
Let *Xabjk* be *Xajk* - *Xbjk*, where - is a difference operator based on the fuzzy extension principle, and denote *Xk.* as the abbreviation of *Xabk* in the following section. Therefore, a set of fuzzy input-output data for the *i*-th observation is re-written as:
$$(\text{\textquotedblleft}\_{\text{k}}; \text{\textquotedblright}\_{\text{1k}}, \text{\textquotedblleft}\_{\text{2k}}, \dots, \text{\textquotedblleft}\_{\text{nk}}\text{\textquotedblright}\_{\text{\textquotedblleft}}), k = 1, 2, \dots, N\text{\textquotedblright}\tag{6}$$
where *Yk* is a fuzzy dependent variable, and *Xjk* is a fuzzy independent variable represented by L-R fuzzy numbers. For simplicity, assume that *Yk* and *Xjk* are positive for any membership value, α ∈ (0,1).
The fuzzy linear regression model (where both input and output data are fuzzy numbers) is represented as follows:
$$\overline{\mathbf{Y}}\_{\mathbf{k}} = \mathbf{A}\_0 \oplus \mathbf{A}\_1 \otimes \mathbf{X}\_{1\mathbf{k}} \oplus \dots \oplus \mathbf{A}\_n \otimes \mathbf{X}\_{n\mathbf{k}} \tag{7}$$
where is a fuzzy estimated variable, A*j*(*j = 1,…,n*) is a fuzzy regression parameter represented by an L-R fuzzy number, ⊗ is an additive operator, and ⊕ is the product operator based on the extension principle.
It should be noted that although the explicit form of the membership function of Yk cannot be directly obtained, the α-level set of Yk can be obtained from Nguyen's theorem (Nguyen, 1978).
Let ( ) L <sup>k</sup> z <sup>α</sup> be a lower value of Yk , and ( ) R <sup>k</sup> z <sup>α</sup> be an upper value of Yk .
Then,
$$\mathbf{Z}\_{\mathbf{k}} = \left[ \mathbf{z}\_{\mathbf{k}(a)}^{\mathrm{L}}, \mathbf{z}\_{\mathbf{k}(a)}^{\mathrm{R}} \right], \quad \alpha \in \{0, 1\} \tag{8}$$
Where
$$\mathbf{z}\_{\mathbf{k}(\alpha)}^{\mathcal{L}} = \sum\_{\mathbf{j}=0}^{n} \left[ \min \left( \mathbf{a}\_{\mathbf{j}(\alpha)}^{\mathcal{L}} \mathbf{x}\_{\mathbf{jk}(\alpha)}^{\mathcal{L}} \mathbf{a}\_{\mathbf{j}(\alpha)}^{\mathcal{L}} \mathbf{x}\_{\mathbf{jk}(\alpha)}^{\mathcal{R}} \right) \right] \tag{9}$$
$$\mathbf{z}\_{\mathbf{k}(\alpha)}^{\mathbb{R}} = \sum\_{\mathbf{j}=0}^{n} \left[ \max \left( \mathbf{a}\_{\mathbf{j}(\alpha)}^{\mathbb{R}} \mathbf{x}\_{\mathbf{jk}(\alpha)}^{\mathbb{L}} \mathbf{a}\_{\mathbf{j}(\alpha)}^{\mathbb{R}} \mathbf{x}\_{\mathbf{jk}(\alpha)}^{\mathbb{R}} \right) \right] \tag{10}$$
10 Fuzzy Logic – Algorithms, Techniques and Implementations
where *Yabk* indicates the *k*-th observation's ambiguous preference for the *a-*th alternative (a) over the *b-*th alternative (b), which represented by fuzzy L-R numbers, and *Xajk* and *Xbjk* are
Let *Xabjk* be *Xajk* - *Xbjk*, where - is a difference operator based on the fuzzy extension principle, and denote *Xk.* as the abbreviation of *Xabk* in the following section. Therefore, a set
where *Yk* is a fuzzy dependent variable, and *Xjk* is a fuzzy independent variable represented by L-R fuzzy numbers. For simplicity, assume that *Yk* and *Xjk* are positive for any
The fuzzy linear regression model (where both input and output data are fuzzy numbers) is
where is a fuzzy estimated variable, A*j*(*j = 1,…,n*) is a fuzzy regression parameter represented by an L-R fuzzy number, ⊗ is an additive operator, and ⊕ is the product
It should be noted that although the explicit form of the membership function of Yk cannot be directly obtained, the α-level set of Yk can be obtained from Nguyen's theorem (Nguyen,
> ( ) { ( ) () () () () } <sup>n</sup> <sup>L</sup> LL LR k j jk j jk
> ( ) { ( ) () () () () } <sup>n</sup> <sup>R</sup> RL RR k j jk j jk
z min a x ,a x α α <sup>α</sup> <sup>α</sup> <sup>α</sup>
z max a x ,a x <sup>α</sup> αααα
R
j 0
=
j 0
=
( ) Y ;X , X , ,X X ,X , ,X abk a1k a2k ank; b1k b2 k bnk; (5)
( ) Y ;X ,X , ,X k 1k 2k nk , *k=1,2,….,N* (6)
YAAX A X <sup>k</sup> = ⊕ ⊗ ⊕⊕ ⊗ 0 1 1k n nk (7)
<sup>k</sup> z <sup>α</sup> be an upper value of Yk .
() () ( ] L R Z z ,z , 0,1 <sup>k</sup> k k α α <sup>=</sup> α ∈ (8)
<sup>=</sup> (9)
<sup>=</sup> (10)
**3.2 Analysis of the additive difference type model**
membership value, α ∈ (0,1).
operator based on the extension principle.
<sup>k</sup> z <sup>α</sup> be a lower value of Yk , and ( )
represented as follows:
1978).
Then,
Where
Let ( ) L
The set of fuzzy input-output data for the *k*-th observation is defined as:
the *j*-th attribute values of the alternatives (a and b) for observation *k*.
of fuzzy input-output data for the *i*-th observation is re-written as:
$$\mathfrak{X}\_{0k(a)}^L = \mathfrak{X}\_{0k(a)}^\mathbb{R} = 1 \tag{11}$$
In the above Equation (9), () () L L j jk a x α α is a product between the lower value of the α-level fuzzy coefficient for the *j*-th attribute and the α-level set of fuzzy input data *Xjk*, () () L R j jk a x α α , () () R L j jk a x α α , or () () R R j jk a x α α is defined in the same manner, respectively. ( ) L 0k x <sup>α</sup> and ( ) R 0k x <sup>α</sup> are assumed to be 1 (a crisp number) for the purpose of estimation for the fuzzy bias parameter *A0.*
To define the dissimilarity between the predicted and observed values of the dependent variable, the following indicator *Dk* ( ) <sup>α</sup> *<sup>2</sup>*was adopted:
$$D\_{\mathbf{k}}\ (a) \ \mathbf{\color{red}{2}} = (\mathbf{y}\_{\mathbf{k}\langle a\rangle}^{\mathbf{L}} \cdot \mathbf{z}\_{\mathbf{k}\langle a\rangle}^{\mathbf{L}})^2 + (\mathbf{y}\_{\mathbf{k}\langle a\rangle}^{\mathbf{R}} \cdot \mathbf{z}\_{\mathbf{k}\langle a\rangle}^{\mathbf{R}})^2 \tag{12}$$
The definition in Equation (12) can be applied to interval data as well as to L-R fuzzy numbers. That is, Equation (12) represents the sum of squares for the distance between interval data.
To generalize, a dissimilarity indicator representing the square of the distance for L-R fuzzy numbers can be written as follows:
$$\mathbf{Dk}^2 = \sum\_{\mathbf{j}=0}^n \mathbf{w} \mathbf{j} ((\mathbf{y}\_{\mathbf{k}(\neq \mathbf{j})}^\mathcal{L} \mathbf{z}\_{\mathbf{k}(\neq \mathbf{j})}^\mathcal{L})^2 + (\mathbf{y}\_{\mathbf{k}(\neq \mathbf{j})}^\mathcal{R} \mathbf{z}\_{\mathbf{k}(\neq \mathbf{j})}^\mathcal{R})^2) \tag{13}$$
where α*j = jh/n, j = 0,...,n*, *h* is an equal interval, and *wj* is a weight for the *j*-th level.
In the case of a triangular fuzzy number with *wj =* 1, the above equation is approximately represented as:
$$\mathbf{Dk}^2 = (\mathbf{y}\_{\mathbf{k}(0)}^\mathcal{L} - \mathbf{z}\_{\mathbf{k}(0)}^\mathcal{L})^2 + (\mathbf{y}\_{\mathbf{k}(1)}^\mathcal{L} - \mathbf{z}\_{\mathbf{k}(1)}^\mathcal{R})^2 + (\mathbf{y}\_{\mathbf{k}(0)}^\mathcal{R} - \mathbf{z}\_{\mathbf{k}(0)}^\mathcal{R})^2 \tag{14}$$
The proposed method is to estimate fuzzy coefficients using minimization of the sum of *Dk 2* respecting *k*. That is,
$$\text{Objective function: Min } \sum\_{\mathbf{k}=1}^{N} \mathbf{D}\_{\mathbf{k}} \stackrel{\ast}{}^{\mathbf{2}} \tag{15}$$
$$\text{Subject to: } \operatorname{a}^{\mathcal{L}}\_{j(h)} \ge 0, \text{ j} \in \mathcal{J}\_1 \tag{16}$$
$$\mathbf{a}\_{\mathbf{j}(\mathbf{h})}^{\mathcal{L}} \le \mathbf{0}, \; \mathbf{a}\_{\mathbf{j}(\mathbf{h})}^{\mathcal{R}} \ge \mathbf{0}, \; \mathbf{j} \in \mathcal{J}\_2 \tag{17}$$
$$\mathbf{a}\_{j(h)}^{\mathbb{R}} \le \mathbf{0}, \ j \in \mathbb{J}\_3 \tag{18}$$
$$-\mathbf{a}\_{j(h)}^{\mathcal{L}} + \mathbf{a}\_{j(h)}^{\mathcal{R}} \ge 0 \tag{19}$$
Ambiguity and Social Judgment: Fuzzy Set Model and Data Analysis 13
highest value (such as 1). The present study develops the concept of logistic regression for the crisp numbers, and then proposes the fuzzy version of logistic regression analysis for
where *Pabk* indicates the *k*-th observation's ambiguous preference for the *a-*th alternative (a) over the *b-*th alternative (b), which is represented by fuzzy L-R numbers, and *Xajk* and *Xbjk*
Let *Xabjk* be *Xajk* ○- *Xbjk*, where ○- is a difference operator based on the fuzzy extension principle, and denote *Xk.* as the abbreviation of *Xabk* in the following section. Therefore, a set
where *Pk* is a fuzzy dependent variable, and *Xjk* is a fuzzy independent variable represented by L-R fuzzy numbers. For simplicity, I assume that *Pk* and *Xjk* are positive for any
The fuzzy logic regression model (where both input and output data are fuzzy numbers) is
*log*(*Pk* ○÷ (1 ○-*Pk*)) = ⊗ ⊕ ⊗ ⊕⊕ ⊗ AX AX A X 0 i0 1 i1 m im (24)
where *log*(*Pk*○÷ (1○-*Pk*)) is the estimated fuzzy log odds, ○÷ is the division operator, ○- is the difference operator, ⊗ is the product operator, and ⊕ is the additive operator based on the
It should be noted that although the explicit form of the membership function of *log*(*Pk*○÷ (1○-*Pk*)) cannot be directly obtained, the α -level set of *log*(*Pk*○÷ (1○-*Pk*)) can be
bound. Then, the α level set of the fuzzy dependent variable *Pk* can be represented as
[*log*(*Pk*○÷ (1○-*Pk*))]<sup>α</sup><sup>=</sup>
() () () () LL RR kk kk [min(log(P /(1 P )),log(P /(1 P ))) αα αα − −
() () () ()
LL RR max(lo kk kk g(P /(1 P )),log(P /(1 P )))] αα αα − − (25)
kP <sup>α</sup> be the lower bound of the dependent fuzzy variable and ( )
Therefore, the α level set of the left term in Equation (24) is as follows:
( ) P ;X ,X , ,X ;X ,X , ,X abk a1k a2k ank b1k b2k bnk (22)
( ) P ;X ,X , ,X k 1k 2k nk , k=1,2,….,N (23)
R
kP <sup>α</sup> be the upper
The set of fuzzy input-output data for the *k-*th observation is defined as:
are the *j*-th attribute values of the alternatives (a and b) for observation *k*.
of fuzzy input-output data for the *i*-th observation is re-written as:
fuzzy input and output data.
membership value, α ∈ (0,1).
() () <sup>α</sup> ( ] L R <sup>k</sup> k k P P ,P , 0,1 α α = α <sup>∈</sup> .
extension principle for the fuzzy set, respectively.
obtained using Nguyen's theorem (Nguyen, 1978).
represented as follows:
Let ( ) L
Where { } <sup>123</sup> j∈ =∪∪ 0,....,n J J J , 12 23 31 J J ,J J ,J J , ∩ = ϕ ∩ = ϕ ∩ = ϕ
$$\mathbf{z}\_{\mathbf{k}}^{\mathcal{L}}(a) = \sum\_{\mathbf{j}\_a \mathbf{l}\_1} \mathbf{a}\_{\mathbf{j}(a)}^{\mathcal{L}} \mathbf{x}\_{\mathbf{jk}(a)}^{\mathcal{L}} + \sum\_{\mathbf{j}\_a \mathbf{l}\_2 \mathbf{l}\_3} \mathbf{a}\_{\mathbf{j}(a)}^{\mathcal{L}} \mathbf{x}\_{\mathbf{jk}(a)}^{\mathcal{R}} \tag{20}$$
$$\mathbf{z}\_{\mathbf{k}}^{\mathbb{R}}(\boldsymbol{\alpha}) = \sum\_{\mathbf{j}\_{\mathbf{k}} \mathbf{J}\_{1} \mathbf{J}\_{12}} \mathbf{a}\_{\mathbf{j}(\boldsymbol{\alpha})}^{\mathbb{R}} \mathbf{x}\_{\mathbf{jk}(\boldsymbol{\alpha})}^{\mathbb{R}} + \sum\_{\mathbf{j}\_{\mathbf{k}} \mathbf{J}\_{3}} \mathbf{a}\_{\mathbf{j}(\boldsymbol{\alpha})}^{\mathbb{R}} \mathbf{x}\_{\mathbf{jk}(\boldsymbol{\alpha})}^{\mathbb{L}} \tag{21}$$
The estimated coefficients can be derived through quadratic programming. The proposed fuzzy least squares method is also shown in Figure 4.
Fig. 4. Fuzzy Least Squares Regressions Analysis for Fuzzy Input and Output Data
### **3.3 Analysis of the logistic type model**
Although the fuzzy linear regression analysis in the fuzzy additive difference model can give satisfactory results, these fuzzy regression analyses may fail to interpret psychological judgment data that have bounds on a psychological scale. For example, a perceived purchase probability has [0,1] interval and cannot be greater than 1 or less than 0. For such data, these fuzzy regression analyses may predict the values that are greater than 1 or less than 0. It may happen that the predicted values are greater than the highest bound or less than the lowest bound, and this causes a significant problem if the predicted values are used in a subsequent analysis. Therefore, the present study also attempted to solve this problem by setting predicted values to be greater than the lowest value (such as 0) or less than the
12 Fuzzy Logic – Algorithms, Techniques and Implementations
() () () () () 1 2 3 L LL LR k j jk j jk j J jJJ z ax ax ∈ ∈
( ) () () () () 1, 12 3 R R R R L k j jk j jk jJ J j J z ax ax ∈ ∈
The estimated coefficients can be derived through quadratic programming. The proposed
*z <sup>L</sup> z*
*j* +
*M y*
*M z*
*L y*
*n*
*j*
11
<sup>2</sup>
Fig. 4. Fuzzy Least Squares Regressions Analysis for Fuzzy Input and Output Data
Although the fuzzy linear regression analysis in the fuzzy additive difference model can give satisfactory results, these fuzzy regression analyses may fail to interpret psychological judgment data that have bounds on a psychological scale. For example, a perceived purchase probability has [0,1] interval and cannot be greater than 1 or less than 0. For such data, these fuzzy regression analyses may predict the values that are greater than 1 or less than 0. It may happen that the predicted values are greater than the highest bound or less than the lowest bound, and this causes a significant problem if the predicted values are used in a subsequent analysis. Therefore, the present study also attempted to solve this problem by setting predicted values to be greater than the lowest value (such as 0) or less than the
==
*N*
*k*
α αα αα = + (20)
α αα αα = + (21)
μ
*R*
*<sup>j</sup>* + <sup>2</sup>
*j*
*Y k*
*R y*
2
Where { } <sup>123</sup> j∈ =∪∪ 0,....,n J J J , 12 23 31 J J ,J J ,J J , ∩ = ϕ ∩ = ϕ ∩ = ϕ
fuzzy least squares method is also shown in Figure 4.
0.1
→
*Min*
**3.3 Analysis of the logistic type model**
highest value (such as 1). The present study develops the concept of logistic regression for the crisp numbers, and then proposes the fuzzy version of logistic regression analysis for fuzzy input and output data.
The set of fuzzy input-output data for the *k-*th observation is defined as:
$$\left(\mathbf{P}\_{\rm abk}; \mathbf{X}\_{\rm a1k}, \mathbf{X}\_{\rm a2k}, \dots; \mathbf{X}\_{\rm ank}; \mathbf{X}\_{\rm b1k}, \mathbf{X}\_{\rm b2k}, \dots; \mathbf{X}\_{\rm bnk}\right) \tag{22}$$
where *Pabk* indicates the *k*-th observation's ambiguous preference for the *a-*th alternative (a) over the *b-*th alternative (b), which is represented by fuzzy L-R numbers, and *Xajk* and *Xbjk* are the *j*-th attribute values of the alternatives (a and b) for observation *k*.
Let *Xabjk* be *Xajk* ○- *Xbjk*, where ○- is a difference operator based on the fuzzy extension principle, and denote *Xk.* as the abbreviation of *Xabk* in the following section. Therefore, a set of fuzzy input-output data for the *i*-th observation is re-written as:
$$\left(\mathbb{P}\_{\mathbf{k}}; \mathbb{X}\_{1\mathbf{k}}, \mathbb{X}\_{2\mathbf{k}}, \dots, \mathbb{X}\_{n\mathbf{k}}\right), \mathbf{k} = \mathbf{1}, \mathbf{2}, \dots, \mathbf{N} \tag{23}$$
where *Pk* is a fuzzy dependent variable, and *Xjk* is a fuzzy independent variable represented by L-R fuzzy numbers. For simplicity, I assume that *Pk* and *Xjk* are positive for any membership value, α ∈ (0,1).
The fuzzy logic regression model (where both input and output data are fuzzy numbers) is represented as follows:
$$\overline{\log(P\_k \oplus (1 \odot P\_k))} = \mathbf{A}\_0 \oplus \mathbf{X}\_{i0} \oplus \mathbf{A}\_1 \otimes \mathbf{X}\_{i1} \oplus \dots \oplus \mathbf{A}\_{\mathfrak{m}} \otimes \mathbf{X}\_{\text{im}} \tag{24}$$
where *log*(*Pk*○÷ (1○-*Pk*)) is the estimated fuzzy log odds, ○÷ is the division operator, ○- is the difference operator, ⊗ is the product operator, and ⊕ is the additive operator based on the extension principle for the fuzzy set, respectively.
It should be noted that although the explicit form of the membership function of *log*(*Pk*○÷ (1○-*Pk*)) cannot be directly obtained, the α -level set of *log*(*Pk*○÷ (1○-*Pk*)) can be obtained using Nguyen's theorem (Nguyen, 1978).
Let ( ) L kP <sup>α</sup> be the lower bound of the dependent fuzzy variable and ( ) R kP <sup>α</sup> be the upper bound. Then, the α level set of the fuzzy dependent variable *Pk* can be represented as () () <sup>α</sup> ( ] L R <sup>k</sup> k k P P ,P , 0,1 α α = α <sup>∈</sup> .
Therefore, the α level set of the left term in Equation (24) is as follows:
$$[\log(P\_k \ominus)(1 \ominus P\_k)]\_{\circ} -$$
$$\begin{aligned} \text{[min}(\overline{\log(\mathcal{P}\_{\mathbf{k}(a)}^{\mathcal{L}} / (1 - \mathcal{P}\_{\mathbf{k}(a)}^{\mathcal{L}}))}) & \overline{\log(\mathcal{P}\_{\mathbf{k}(a)}^{\mathcal{R}} / (1 - \mathcal{P}\_{\mathbf{k}(a)}^{\mathcal{R}}))}) \\\\ \max(\overline{\log(\mathcal{P}\_{\mathbf{k}(a)}^{\mathcal{L}} / (1 - \mathcal{P}\_{\mathbf{k}(a)}^{\mathcal{L}}))}, \overline{\log(\mathcal{P}\_{\mathbf{k}(a)}^{\mathcal{R}} / (1 - \mathcal{P}\_{\mathbf{k}(a)}^{\mathcal{R}}))})) \end{aligned} \tag{25}$$
Ambiguity and Social Judgment: Fuzzy Set Model and Data Analysis 15
The participant also rated the desirability of the attribute information for each computer using a fuzzy rating method. The fuzzy rating scale of desirability ranged from 0 point to 100 points. (see Figure 6). That is, the participant answered the lower value, the
0 100
100
The fuzzy coefficients were obtained by fuzzy linear regression analysis using the least squares under constraints, as shown in Tables 1 and 2. The dependent variable of Table 1 was the same as that in Table 2. However, the independent variables in Table 1 are objective values measured by crisp numbers, whereas in Table 2 the independent variables are fuzzy rating values measured by an L-R fuzzy number. The parameter of *Ajk*0 involves a response bias owing to presentation order, context effects, and the scale parameter of the dependent variables. The parameter *Ajk*0 would be a fuzzy variable and larger than *Aab0* if *Xa* were more salient than *Xb*. This model can be reduced to the Fuzzy Utility Difference Model (Nakamura, 1992) if multi-attribute weighting parameters are assumed to be crisp numbers, and reduced to the Additive Difference Model (Tversky, 1969) if multi-attribute weighting parameters and the values of multi-attributes are assumed to be crisp numbers as explained before. According to Tables 1 and 2, the preference strength concerning comparative judgment was influenced most by whether the target computer was new or used. The impact of the hard disks' attributes was smaller than that of the new-used dimension.
The participant was a 43-year-old adult. The participant rated the ambiguous probability of preferring a certain computer (DELL brand) out of seven different computers. Three types of attribute information (hard disk: 100 or 60 GB; memory: 2.80 or 2.40 GHz; new or used product) were manipulated in the same manner as in the previous judgment task.. That is, the participant answered the lower value, the representative value , and upper value for the probability that superior alternative is preferred to inferior alternative. The participant used the fuzzy rating method to provide representative, lower, and upper values of probabilities
representative value , and upper value for each attribute value.
Fig. 6. Example of a Fuzzy Desirability Rating.
1) Low ambiguity
2)High ambiguity 0
**4.1.1.2 Analysis and results**
**4.1.2 Example of the logistic model 4.1.2.1 Participant and procedure**
(see Figure 7 ).
Let ( ) L <sup>k</sup> z <sup>α</sup> be a lower value of [*log*(*Pk*○÷ (1○-*Pk*))]α, and ( ) R <sup>k</sup> z <sup>α</sup> be an upper value of [*log*(*Pk*○÷ (1○-*Pk*))]<sup>α</sup>
where
$$\mathbf{z}\_{\mathbf{k}(\alpha)}^{\mathcal{L}} = \sum\_{\mathbf{j}=0}^{n} \left[ \min \left( \mathbf{a}\_{\mathbf{j}(\alpha)}^{\mathcal{L}} \mathbf{x}\_{\mathbf{jk}(\alpha)}^{\mathcal{L}} \mathbf{a}\_{\mathbf{j}(\alpha)}^{\mathcal{L}} \mathbf{x}\_{\mathbf{jk}(\alpha)}^{\mathcal{R}} \right) \right] \tag{26}$$
$$\mathbf{z}\_{\mathbf{k}\left(\alpha\right)}^{\mathrm{R}} = \sum\_{\mathbf{j}=0}^{\mathrm{n}} \left[ \max \left( \mathbf{a}\_{\mathbf{j}\left(\alpha\right)}^{\mathrm{R}} \mathbf{x}\_{\mathbf{jk}\left(\alpha\right)}^{\mathrm{L}} \mathbf{a}\_{\mathbf{j}\left(\alpha\right)}^{\mathrm{R}} \mathbf{x}\_{\mathbf{jk}\left(\alpha\right)}^{\mathrm{R}} \right) \right] \tag{27}$$
$$\mathcal{X}\_{0k}^{L}(\alpha) = \mathcal{X}\_{0k}^{R}(\alpha) = \mathbf{1} \tag{28}$$
In the above Equation (26), is a product between the lower value of the �-level fuzzy coefficient for the *j*-th attribute and the α-level set of fuzzy input data *Xjk*, , or is defined in the same manner, respectively. and are assumed to be 1 (a crisp number) for the purpose of estimation for the fuzzy bias parameter *A0*. The parameter estimation method is basically the same as the fuzzy logistic regression method and a more concrete procedure is described in Takemura (2004).
### **4. Numerical example of the data analysis method**
To demonstrate the appropriateness of the proposed data analysis methods, the detail numerical examples are shown for the individual level analysis (Takemura,2007) and group level analysis (Takemura, Matsumoto, Matsuyama, & Kobayashi, 2011) of ambiguous comparative judgments.
### **4.1. Individual level analysis of ambiguous comparative model**
### **4.1.1 Example of additive difference model**
### **4.1.1.1 Participant and procedure**
The participant was a 43-year-old faculty member of Waseda University. The participant rated differences in WTP for two different computers (DELL brand) with three types of attribute information (hard disk: 100 or 60 GB; memory: 2.80 or 2.40 GHz; new or used product). The participant compared a certain alternative with seven different alternatives. The participant provided representative values and lower and upper WTP values using a fuzzy rating method. (see Figure 5)
The participant was asked the amount of money he would be willing to pay to upgrade the inferior from inferior alternative to superior alternative using fuzzy rating method. That is, the participant answered the lower value, the representative value, and upper value for the amount of money he would be willing to pay.
$$\begin{array}{cccc} \text{Lower Value} & \text{Representative Value} & \text{Upper Value} \\\\ \text{(} & ) \text{ Yen} & \text{(} & ) \text{ Yen} & \text{(} & ) \text{ Yen} \end{array}$$
Fig. 5. Example of a Fuzzy Rating in WTP Task.
The participant also rated the desirability of the attribute information for each computer using a fuzzy rating method. The fuzzy rating scale of desirability ranged from 0 point to 100 points. (see Figure 6). That is, the participant answered the lower value, the representative value , and upper value for each attribute value.
Fig. 6. Example of a Fuzzy Desirability Rating.
### **4.1.1.2 Analysis and results**
14 Fuzzy Logic – Algorithms, Techniques and Implementations
( ) { ( ) () () () () } <sup>n</sup> <sup>L</sup> LL LR k j jk j jk
( ) { ( ) () () () () } <sup>n</sup> <sup>R</sup> RL RR k j jk j jk
In the above Equation (26), is a product between the lower value of the �-level fuzzy coefficient for the *j*-th attribute and the α-level set of fuzzy input data *Xjk*, , or is defined in the same manner, respectively. and are assumed to be 1 (a crisp number) for the purpose of estimation for the fuzzy bias parameter *A0*. The parameter estimation method is basically the same as the fuzzy logistic regression method and a more concrete procedure is described in
To demonstrate the appropriateness of the proposed data analysis methods, the detail numerical examples are shown for the individual level analysis (Takemura,2007) and group level analysis (Takemura, Matsumoto, Matsuyama, & Kobayashi, 2011) of ambiguous
The participant was a 43-year-old faculty member of Waseda University. The participant rated differences in WTP for two different computers (DELL brand) with three types of attribute information (hard disk: 100 or 60 GB; memory: 2.80 or 2.40 GHz; new or used product). The participant compared a certain alternative with seven different alternatives. The participant provided representative values and lower and upper WTP values using a
The participant was asked the amount of money he would be willing to pay to upgrade the inferior from inferior alternative to superior alternative using fuzzy rating method. That is, the participant answered the lower value, the representative value, and upper value for the
> Lower Value Representative Value Upper Value ( ) Yen ( ) Yen ( ) Yen
z min a x ,a x α α <sup>α</sup> <sup>α</sup> <sup>α</sup>
z max a x ,a x <sup>α</sup> αααα
( ) *L k x*0 α = ( ) *R k x*0 α
R
<sup>=</sup> (26)
<sup>=</sup> (27)
<sup>k</sup> z <sup>α</sup> be an upper value of [*log*(*Pk*○÷
=1 (28)
<sup>k</sup> z <sup>α</sup> be a lower value of [*log*(*Pk*○÷ (1○-*Pk*))]α, and ( )
j 0
=
j 0
=
**4. Numerical example of the data analysis method**
**4.1. Individual level analysis of ambiguous comparative model**
**4.1.1 Example of additive difference model**
amount of money he would be willing to pay.
Fig. 5. Example of a Fuzzy Rating in WTP Task.
**4.1.1.1 Participant and procedure**
fuzzy rating method. (see Figure 5)
Let ( ) L
where
(1○-*Pk*))]<sup>α</sup>
Takemura (2004).
comparative judgments.
The fuzzy coefficients were obtained by fuzzy linear regression analysis using the least squares under constraints, as shown in Tables 1 and 2. The dependent variable of Table 1 was the same as that in Table 2. However, the independent variables in Table 1 are objective values measured by crisp numbers, whereas in Table 2 the independent variables are fuzzy rating values measured by an L-R fuzzy number. The parameter of *Ajk*0 involves a response bias owing to presentation order, context effects, and the scale parameter of the dependent variables. The parameter *Ajk*0 would be a fuzzy variable and larger than *Aab0* if *Xa* were more salient than *Xb*. This model can be reduced to the Fuzzy Utility Difference Model (Nakamura, 1992) if multi-attribute weighting parameters are assumed to be crisp numbers, and reduced to the Additive Difference Model (Tversky, 1969) if multi-attribute weighting parameters and the values of multi-attributes are assumed to be crisp numbers as explained before. According to Tables 1 and 2, the preference strength concerning comparative judgment was influenced most by whether the target computer was new or used. The impact of the hard disks' attributes was smaller than that of the new-used dimension.
### **4.1.2 Example of the logistic model**
### **4.1.2.1 Participant and procedure**
The participant was a 43-year-old adult. The participant rated the ambiguous probability of preferring a certain computer (DELL brand) out of seven different computers. Three types of attribute information (hard disk: 100 or 60 GB; memory: 2.80 or 2.40 GHz; new or used product) were manipulated in the same manner as in the previous judgment task.. That is, the participant answered the lower value, the representative value , and upper value for the probability that superior alternative is preferred to inferior alternative. The participant used the fuzzy rating method to provide representative, lower, and upper values of probabilities (see Figure 7 ).
Ambiguity and Social Judgment: Fuzzy Set Model and Data Analysis 17
0% 100%
100% 0%
The fuzzy coefficients were obtained by fuzzy linear regression analysis using least squares under constraints, as shown in Tables 3 and 4. However, in Table 3 the independent variables are objective values measured by crisp numbers, whereas in Table 4 the independent variables are fuzzy rating values measured by an L-R fuzzy number. The parameter *Ajk*0 involves a response bias owing to presentation order, context effects, and the scale parameter of the dependent variables. According to Tables 3 and 4, the bounded preference strength was influenced most by whether the target computer was new or used. Interestingly, the impact of the attribute for memory was slightly greater than was the case
Attribute Value
Hard Disk(M) Representative 0.000
Fuzzy Memory(L) Lower 1.781 Coefficient Memory(M) Representative 1.781 Memory(R) Upper 1.881
(R) Upper 1.443
Hard Disk (L) Lower 0.000
Hard Disk (R) Upper 0.009
New or Used(R) Lower 1.791 New or Used(M) Representative 2.097 New or Used(L) Upper 2.777
(L) Lower 0.847
(M) Representative 1.201
1) Low ambiguity
2)High ambiguity
Fig. 7. Example of Fuzzy Probability Rating.
Note: The independent variables are crisp numbers.
Table 3. Coefficients of Fuzzy Logistic Regression Analysis
*Ajk* <sup>0</sup>
*A*jk0
*A*jk0
**4.1.2.2 Analysis and results**
in Tables 1 and 2.
Note: The independent variables are crisp numbers.
Table 1. Coefficients of Fuzzy Regression Analysis
Note: The independent variables are fuzzy L-R numbers.
Table 2. Coefficients of Fuzzy Regression Analysis
Fig. 7. Example of Fuzzy Probability Rating.
### **4.1.2.2 Analysis and results**
16 Fuzzy Logic – Algorithms, Techniques and Implementations
Hard Disk(L) Lower 78.5
Fuzzy Memory(L) Lower 0.0 Coefficient Memory(M) Representative 0.0
(R) Upper 33 111.2
Hard Disk(L) Lower 33.9
Fuzzy Memory(L) Lower 0.0 Coefficient Memory(M) Representative 0.0
(R) Upper 48 004.0
Hard Disk (M) Representative 85.7 Hard Disk (R) Upper 986.8
Memory(R) Upper 0.0 New or Used(R) Lower 22 332.5 New or Used (M) Representative 22 332.5 New or Used(L) Upper 22 332.5
Hard Disk (M) Representative 33.9 Hard Disk (R) Upper 33.9
Memory(R) Upper 0.0 New or Used(R) Lower 446.1 New or Used(M) Representative 446.1 New or Used(L) Upper 446.1
(L) Lower 36 082.1
(M) Representative 36 082.1
(L) Lower 25 450.8
(M) Representative 29 420.1
Attribute Value
Note: The independent variables are crisp numbers. Table 1. Coefficients of Fuzzy Regression Analysis
*Ajk*<sup>0</sup>
*A*jk0
*A*jk0
Note: The independent variables are fuzzy L-R numbers. Table 2. Coefficients of Fuzzy Regression Analysis
*Ajk* <sup>0</sup>
*A*jk0
*Ajk* <sup>0</sup>
Attribute Value
The fuzzy coefficients were obtained by fuzzy linear regression analysis using least squares under constraints, as shown in Tables 3 and 4. However, in Table 3 the independent variables are objective values measured by crisp numbers, whereas in Table 4 the independent variables are fuzzy rating values measured by an L-R fuzzy number. The parameter *Ajk*0 involves a response bias owing to presentation order, context effects, and the scale parameter of the dependent variables. According to Tables 3 and 4, the bounded preference strength was influenced most by whether the target computer was new or used. Interestingly, the impact of the attribute for memory was slightly greater than was the case in Tables 1 and 2.
Note: The independent variables are crisp numbers.
Table 3. Coefficients of Fuzzy Logistic Regression Analysis
Ambiguity and Social Judgment: Fuzzy Set Model and Data Analysis 19
Then, please estimate the amount of money you would be willing to pay to upgrade the inferior alternative from inferior alternative to superior alternative using fuzzy rating method. That is, the participants answered the lower value, the representative value, and
The fuzzy coefficients were obtained by fuzzy linear regression analysis using the least squares under constraints, as shown in Tables 5 for the digital camera data and Table 6 for mobile phone data. The independent variables in Table5 and Table 6 are objective values measured by crisp numbers. The parameter of *Ajk*0 involves a response bias owing to presentation order, context effects, and the scale parameter of the dependent variables.According to Tables 5, the preference strength concerning comparative judgment was influenced most by whether the target digital camera was 2.5 or 5.0 inches. The impact of the memory's attribute was smaller than those of display size and weight dimensions. According to Tables 6, the preference strength concerning comparative judgment was influenced most by whether the target mobile phone was 2.8 or 3.0 inches. The impact of the pixel number's attribute was smaller than those of display size and weight dimensions. The participants also rated the desirability of the attribute information for each computer using a fuzzy rating method. The fuzzy rating scale of desirability ranged from 0 point to 100 points. (see Figure 6). That is, the participant answered the lower value, the representative value ,
Which alternative do you prefer ? Please circle the superior alternative.
upper value for the amount of money you would be willing to pay.
Minimum: 2,000 yen ----- Maximum: 10, 000 yen
**Brand A Brand B** Weight: 130g Weight: 160g Memory: 25MB Memory: 50MB Display: 50 inches Display: 25 inches
**Difference**
Fig. 8. Example of Fuzzy WTP Rating
and upper value for each attribute value.
**4.2.1.2 Analysis and results**
Representative Value: 5,000 yen
**Question:**
Note: The independent variables are fuzzy L-R numbers.
Table 4. Coefficients of Fuzzy Logistic Regression Analysis
### **4.2 Group level analysis of ambiguous comparative model**
### **4.2.1 Example of additive difference model**
### **4.2.1.1 Participants and procedure**
The participant s were 100 undergraduate university students (68 female and 32 male students) enrolled in an economic
psychology class at Waseda University. They were recruited for an experiment investigating "consumer preference ".
Their average age was 21.3 years old. The participants rated differences in WTP for two different digital cameras with three types of attribute information (weight: 130 gram or1 60 gram; memory: 25 or 50 MB; display size:2.5 or 5.0 inches). The participants compared a certain alternative with seven different alternatives. The participants also rated differences in WTP for two different mobile phones with three types of attribute information (weight: 123 gram or132 gram; pixel number:3,200,000 or 5,070,000 pixels; display size:2.8 or 3.0 inches). The participants compared a certain alternative with seven different mobile phones. The participant provided representative values and lower and upper WTP values using a fuzzy rating method. The participants were asked the amount of money he would be willing to pay to upgrade the inferior from inferior alternative to superior alternative using fuzzy rating method. That is, the participants answered the lower value, the representative value , and upper value for the amount of money he would be willing to pay. An example of fuzzy WTP rating is illustrated in the Figure 8.
### **Question:**
18 Fuzzy Logic – Algorithms, Techniques and Implementations
Hard Disk (M) Representative 0.000 Hard Disk (R) Upper 0.000
New or Used(R) Lower 0.043 New or Used(M) Representative 0.043 New or Used(L) Upper 0.043
(M) Representative 1.806
Hard Disk(L) Lower 0.000
Fuzzy Memory(L) Lower 0.008 Coefficient Memory(M) Representative 0.008 Memory(R) Upper 0.008
(L) Lower 1.806
(R) Upper 1.806
The participant s were 100 undergraduate university students (68 female and 32 male
psychology class at Waseda University. They were recruited for an experiment investigating
Their average age was 21.3 years old. The participants rated differences in WTP for two different digital cameras with three types of attribute information (weight: 130 gram or1 60 gram; memory: 25 or 50 MB; display size:2.5 or 5.0 inches). The participants compared a certain alternative with seven different alternatives. The participants also rated differences in WTP for two different mobile phones with three types of attribute information (weight: 123 gram or132 gram; pixel number:3,200,000 or 5,070,000 pixels; display size:2.8 or 3.0 inches). The participants compared a certain alternative with seven different mobile phones. The participant provided representative values and lower and upper WTP values using a fuzzy rating method. The participants were asked the amount of money he would be willing to pay to upgrade the inferior from inferior alternative to superior alternative using fuzzy rating method. That is, the participants answered the lower value, the representative value , and upper value for the amount of money he would be willing to pay. An example of fuzzy
Attribute Value
Note: The independent variables are fuzzy L-R numbers.
**4.2.1 Example of additive difference model**
**4.2.1.1 Participants and procedure**
students) enrolled in an economic
WTP rating is illustrated in the Figure 8.
"consumer preference ".
Table 4. Coefficients of Fuzzy Logistic Regression Analysis
*Ajk* <sup>0</sup>
Ajk0
*Ajk* <sup>0</sup>
**4.2 Group level analysis of ambiguous comparative model**
Which alternative do you prefer ? Please circle the superior alternative.
Then, please estimate the amount of money you would be willing to pay to upgrade the inferior alternative from inferior alternative to superior alternative using fuzzy rating method. That is, the participants answered the lower value, the representative value, and upper value for the amount of money you would be willing to pay.
Fig. 8. Example of Fuzzy WTP Rating
### **4.2.1.2 Analysis and results**
The fuzzy coefficients were obtained by fuzzy linear regression analysis using the least squares under constraints, as shown in Tables 5 for the digital camera data and Table 6 for mobile phone data. The independent variables in Table5 and Table 6 are objective values measured by crisp numbers. The parameter of *Ajk*0 involves a response bias owing to presentation order, context effects, and the scale parameter of the dependent variables.According to Tables 5, the preference strength concerning comparative judgment was influenced most by whether the target digital camera was 2.5 or 5.0 inches. The impact of the memory's attribute was smaller than those of display size and weight dimensions. According to Tables 6, the preference strength concerning comparative judgment was influenced most by whether the target mobile phone was 2.8 or 3.0 inches. The impact of the pixel number's attribute was smaller than those of display size and weight dimensions. The participants also rated the desirability of the attribute information for each computer using a fuzzy rating method. The fuzzy rating scale of desirability ranged from 0 point to 100 points. (see Figure 6). That is, the participant answered the lower value, the representative value , and upper value for each attribute value.
Ambiguity and Social Judgment: Fuzzy Set Model and Data Analysis 21
The participant s were 100 undergraduate university students (68 female and 32 male students). Their average age was 21.3 years old. The participants rated the ambiguous probability of preferring a certain digital camera out of seven different digital cameras. The three types of attribute information (weight: 130 gram or1 60 gram; memory: 25 or 50 MB; display size:2.5 or 5.0 inches) were manipulated in the same manner as in the previous individual judgment task. They also rated the ambiguous probability of preferring a certain mobile phone out of seven different mobile phones. The three types of attribute information (weight: 123 gram or132 gram; pixel number:3,200,000 or 5,070,000 pixels; display size:2.8 or 3.0 inches) were manipulated in the same manner in the previous judgment task. The participant provided representative values and lower and upper values of probabilities. That is, the participants answered the lower value, the representative value , and upper value for the probability that superior alternative is preferred to inferior alternative. The participants used the fuzzy rating method to provide representative, lower, and upper
The fuzzy coefficients were obtained by fuzzy logistic regression analysis using the least squares under constraints, as shown in Tables 7 for the digital camera data and Table 8 for mobile phone data. The independent variables in Table 7 and Table 8 are objective values measured by crisp numbers. The parameter of *Ajk*0 involves a response bias owing to presentation order, context effects, and the scale parameter of the dependent variables. According to Tables 7, the bounded preference strength was influenced most by whether the target digital camera was 2.5 or 5.0 inches. The impact of the memory's attribute was smaller than those of display size and weight dimensions. According to Tables 8, the bounded preference strength t was influenced most by whether the target mobile phone was 2.8 or 3.0 inches. The impact of the weight's attribute was smaller than those of display
Weight(L) Lower 0.035
Fuzzy Memory(L) Lower 0.003 Coefficient Memory(M) Representative 0.003 Memory(R) Upper 0.003
(R) Upper 1.072
Table 7. Coefficients of Fuzzy Logistic Regression Analysis for Digital Camera Data
Weight (M) Representative 0.038 Weight (R) Upper 0.054
Display Size(R) Lower 2.625 Display Size (M) Representative 2.625 Display Size(L) Upper 2.625
> (L) Lower -0.122 (M) Representative 0.459
**4.2.2 Example of the logistic model 4.2.2.1 Participants and procedure**
values of probabilities (see Figure 7 ).
size and pixel number dimensions.
Note: The independent variables are crisp numbers.
Attribute Value
*Ajk* <sup>0</sup>
Ajk0 Ajk0
**4.2.2.2 Analysis and results**
Note: The independent variables are crisp numbers.
Table 5. Coefficients of Fuzzy Regression Analysis for Digital Camera Data
Note: The independent variables are crisp numbers.
Table 6. Coefficients of Fuzzy Regression Analysis for Mobile Phone Data
### **4.2.2 Example of the logistic model**
### **4.2.2.1 Participants and procedure**
20 Fuzzy Logic – Algorithms, Techniques and Implementations
Weight (M) Representative 48.57 Weight (R) Upper 68.33
Memory(R) Upper 14.62 Display Size(R) Lower 223.10 Display Size (M) Representative 4791.98 Display Size(L) Upper 4791.98
(L) Lower 11361.25
(M) Representative 11361.25
Weight(L) Lower 48.57
Fuzzy Memory(L) Lower 8.29 Coefficient Memory(M) Representative 8.29
(R) Upper 15447.54
Weight(L) Lower 28.84
Fuzzy Pixel Number(L) Lower -12.12 Coefficient Pixel Number(M) Representative 28.55
(R) Upper 12569.35
Table 6. Coefficients of Fuzzy Regression Analysis for Mobile Phone Data
Weight (M) Representative 28.84 Weight (R) Upper 53.44
Pixel Number(R) Upper 28.55 Display Size(R) Lower -233.73 Display Size(M) Representative 190.29 Display Size(L) Upper 190.29
(L) Lower 7758.98
(M) Representative 8234.94
Table 5. Coefficients of Fuzzy Regression Analysis for Digital Camera Data
Attribute Value
*Ajk* <sup>0</sup>
*Ajk*<sup>0</sup>
*A*jk0
Attribute Value
Note: The independent variables are crisp numbers.
*Ajk* <sup>0</sup>
*A*jk0 *A*jk0
Note: The independent variables are crisp numbers.
The participant s were 100 undergraduate university students (68 female and 32 male students). Their average age was 21.3 years old. The participants rated the ambiguous probability of preferring a certain digital camera out of seven different digital cameras. The three types of attribute information (weight: 130 gram or1 60 gram; memory: 25 or 50 MB; display size:2.5 or 5.0 inches) were manipulated in the same manner as in the previous individual judgment task. They also rated the ambiguous probability of preferring a certain mobile phone out of seven different mobile phones. The three types of attribute information (weight: 123 gram or132 gram; pixel number:3,200,000 or 5,070,000 pixels; display size:2.8 or 3.0 inches) were manipulated in the same manner in the previous judgment task. The participant provided representative values and lower and upper values of probabilities. That is, the participants answered the lower value, the representative value , and upper value for the probability that superior alternative is preferred to inferior alternative. The participants used the fuzzy rating method to provide representative, lower, and upper values of probabilities (see Figure 7 ).
### **4.2.2.2 Analysis and results**
The fuzzy coefficients were obtained by fuzzy logistic regression analysis using the least squares under constraints, as shown in Tables 7 for the digital camera data and Table 8 for mobile phone data. The independent variables in Table 7 and Table 8 are objective values measured by crisp numbers. The parameter of *Ajk*0 involves a response bias owing to presentation order, context effects, and the scale parameter of the dependent variables. According to Tables 7, the bounded preference strength was influenced most by whether the target digital camera was 2.5 or 5.0 inches. The impact of the memory's attribute was smaller than those of display size and weight dimensions. According to Tables 8, the bounded preference strength t was influenced most by whether the target mobile phone was 2.8 or 3.0 inches. The impact of the weight's attribute was smaller than those of display size and pixel number dimensions.
Note: The independent variables are crisp numbers.
Table 7. Coefficients of Fuzzy Logistic Regression Analysis for Digital Camera Data
Ambiguity and Social Judgment: Fuzzy Set Model and Data Analysis 23
marketing research, risk perception research, and human judgment and decision-making research. Empirical research using possibilistic analysis and least squares analysis will be
Results of these applications to psychological study indicated that the parameter estimated in the proposed analysis was meaningful for social judgment study. This study has a methodological restriction on statistical inferences for fuzzy parameters. Therefore, we plan further work on the fuzzy theoretic analysis of social judgment directed toward the statistical study of fuzzy regression analysis and fuzzy logistic regression analysis such as
This work was supported in part by Grants in Aids for Grant-in-Aid for Scientific Research on Priority Area, The Ministry of Education, Culture, Sports, Science and Technology(MEXT). I thank Matsumoto,T., Matsuyama,S.,and Kobayashi,M.. for their
Anderson,N.H.(1988). A functional approach to person cognition. In T.K.Srull & R.S. Wyer
Dubois D. & Prade,H. (1980). Fuzzy sets and systems: Theory and applications, New York:
Festinger, L. (1954). A theory of social comparison processes. *Human Relations,* 7, 114–140. Hesketh, B., Pryor, R., Gleitzman, M., & Hesketh, T. (1988). Practical applications and
(Ed.), *Fuzzy sets in psychology* (pp. 425–454). New York: North Holland. Kühberger,A.,.Schulte-Mecklenbeck,M. & Ranyard,R. (2011). Introduction: Windows for
Mussweiler, T. (2003). Comparison processes in social judgment: Mechanisms and
Nakamura, K. (1992). On the nature of intransitivity in human referential judgments. In V.
Nguyen, H. T. (1978). A note on the extension principle for fuzzy sets. *Journal of Mathematical*
Rosch,E. (1975). Cognitive representation of semantic categories. Journal of Experimental
Rosch,E., & Mervis,C.B. (1975). Family resemblances: Studies in the internal structure of
Sakawa, M., & Yano, H. (1992). Multiobjective fuzzy linear regression analysis for fuzzy
(Eds.), *Advances in social cognition.* vol.1. Hiisdale, New Jersey: Lawrence Erlbaum
psychometric evaluation of a computerised fuzzy graphic rating scale. In T. Zetenyi
understanding the mind, In M.Schulte-Mecklenbeck, A.Kühberger, & R. Ranyard(Eds.), A handbook of process tracing methods for decision research,New
Novak (Ed.), *Fuzzy approach to reasoning and decision making, academia* (pp. 147–162).
statistical tests of parameters, outlier detection, and step-wise variable selection.
assistance, and the editor and the reviewers for their valuable comments.
needed to examine the validity of these models.
**6. Acknowledgment**
**7. References**
Associates, pp.37-51.
Yorrk: Psychologgy Press, pp.3-17.
Prague: Kluwer Academic Publishers.
*Analysis and Application,* 64, 369–380.
Psychology: General, 104, 192-233.
categories. *Cognitive Psychology,* 7, 573-603.
input-output data. *Fuzzy Sets and Systems,* 47, 173–181.
consequences*. Psychological Review,* 110, 472–489.
Academic Press.
Note: The independent variables are crisp numbers.
Table 8. Coefficients of Fuzzy Logistic Regression Analysis for Mobile Phone Data
### **5. Conclusion**
This chapter introduce fuzzy set models for ambiguous comparative judgments, which do not always hold transitivity and comparability properties. The first type of model was a fuzzy theoretical extension of the additive difference model for preference that is used to explain ambiguous preference strength. This model can be reduced to the Fuzzy Utility Difference Model (Nakamura, 1992) if multi-attribute weighting parameters are assumed to be crisp numbers, and can be reduced to the Additive Difference Model (Tversky, 1969) if multi-attribute weighting parameters and the values of multi-attributes are assumed to be crisp numbers. The second type of model was a fuzzy logistic model for explaining ambiguous preference in which preference strength is bounded, such as a probability measure.
In both models, multi-attribute weighting parameters and all attribute values were assumed to be asymmetric fuzzy L-R numbers. For each model, parameter estimation method using fuzzy regression analysis was introduced. Numerical examples for comparison were also demonstrated. As the objective of the numerical examples was to demonstrate that the proposed estimation might be viable, further empiric studies will be needed. Moreover, because the two models require different evaluation methods, comparisons of the psychological effects of the two methods must be studied further.
In this chapter, the least squares method was used for data analyses of the two models. However, the possibilistic linear regression analysis (Sakawa & Yano, 1992) and the possibilistic logistic regression analysis (Takemura, 2004) could also be used in the data analysis of the additive difference type model and the logistic type model, respectively. The proposed models and the analyses for ambiguous comparative judgments will be applied to marketing research, risk perception research, and human judgment and decision-making research. Empirical research using possibilistic analysis and least squares analysis will be needed to examine the validity of these models.
Results of these applications to psychological study indicated that the parameter estimated in the proposed analysis was meaningful for social judgment study. This study has a methodological restriction on statistical inferences for fuzzy parameters. Therefore, we plan further work on the fuzzy theoretic analysis of social judgment directed toward the statistical study of fuzzy regression analysis and fuzzy logistic regression analysis such as statistical tests of parameters, outlier detection, and step-wise variable selection.
### **6. Acknowledgment**
22 Fuzzy Logic – Algorithms, Techniques and Implementations
Weight (M) Representative 0.002 Weight (R) Upper 0.009
Pixel Number(R) Upper 0.024 Display Size(R) Lower 0.161 Display Size(M) Representative 0.165 Display Size(L) Upper 0.232
(L) Lower -0.871
(M) Representative 0.030
Weight(L) Lower 0.002
Fuzzy Pixel Number(L) Lower 0.012 Coefficient Pixel Number(M) Representative 0.017
(R) Upper 0.887
This chapter introduce fuzzy set models for ambiguous comparative judgments, which do not always hold transitivity and comparability properties. The first type of model was a fuzzy theoretical extension of the additive difference model for preference that is used to explain ambiguous preference strength. This model can be reduced to the Fuzzy Utility Difference Model (Nakamura, 1992) if multi-attribute weighting parameters are assumed to be crisp numbers, and can be reduced to the Additive Difference Model (Tversky, 1969) if multi-attribute weighting parameters and the values of multi-attributes are assumed to be crisp numbers. The second type of model was a fuzzy logistic model for explaining ambiguous preference in which preference strength is bounded, such as a probability
In both models, multi-attribute weighting parameters and all attribute values were assumed to be asymmetric fuzzy L-R numbers. For each model, parameter estimation method using fuzzy regression analysis was introduced. Numerical examples for comparison were also demonstrated. As the objective of the numerical examples was to demonstrate that the proposed estimation might be viable, further empiric studies will be needed. Moreover, because the two models require different evaluation methods, comparisons of the
In this chapter, the least squares method was used for data analyses of the two models. However, the possibilistic linear regression analysis (Sakawa & Yano, 1992) and the possibilistic logistic regression analysis (Takemura, 2004) could also be used in the data analysis of the additive difference type model and the logistic type model, respectively. The proposed models and the analyses for ambiguous comparative judgments will be applied to
psychological effects of the two methods must be studied further.
Table 8. Coefficients of Fuzzy Logistic Regression Analysis for Mobile Phone Data
Attribute Value
*Ajk* <sup>0</sup> *Ajk* <sup>0</sup>
*Ajk*<sup>0</sup>
Note: The independent variables are crisp numbers.
**5. Conclusion**
measure.
This work was supported in part by Grants in Aids for Grant-in-Aid for Scientific Research on Priority Area, The Ministry of Education, Culture, Sports, Science and Technology(MEXT). I thank Matsumoto,T., Matsuyama,S.,and Kobayashi,M.. for their assistance, and the editor and the reviewers for their valuable comments.
### **7. References**
**2**
Agnes Achs
*Hungary*
**From Fuzzy Datalog to**
*University of Pecs Faculty of Engineering,*
**Multivalued Knowledge-Base**
Despite the fact that people have very different and ambiguous concepts and knowledge, they are able to talk to one another. How does human mind work? How can people give answers to questions? Modelling human conversation and knowledge demands to deal with
Human knowledge consists of static and dynamic knowledge chunks. The static ones include the so called lexical knowledge or the ability to sense similarities between facts and between predicates. Through dynamic attainments one can make deductions or one can give answers to a question. There are several and very different approaches to make a model of human knowledge, but one of the most common and widespread fields of research is based
Fuzzy sets theory, proposed by Zadeh (1965), is a realistic and practical means to describe the world that we live in. The method has successfully been applied in various fields, among others in decision making, logic programming, and approximate reasoning. In the last decade, a number of papers have dealt with that subject, e.g. (Formato et al 2000, Sessa 2002, Medina et al 2004, Straccia et al 2009). They deal with different aspects of modelling and handling uncertainty. (Straccia 2008) gives a detailed overview of this topic with widespread references. Our investigations have begun independently of these works, and have run parallel to them. Of course there are some similar features, but our model differs from the
As a generalization of fuzzy sets, intuitionistic fuzzy sets were presented by Atanassov (Atanassov 1983), and have allowed people to deal with uncertainty and information in a much broader perspective. Another well-known generalization of an ordinary fuzzy set is the interval-valued fuzzy set, which was first introduced by Zadeh (Zadeh 1975). These generalizations make descriptions and models of the world more realistic, and practical.
In the beginning, our knowledge-base model was based on the concept of fuzzy logic, later on it was extended to intuitionistic and interval-valued logic. In this model, the static part is a background knowledge module, while the dynamic part consists of a Datalog based deduction mechanism. To develop this mechanism, it was necessary to generalize the Datalog language
and to extend it into fuzzy and intuitionistic direction. (Achs 1995, 2007, 2010).
**1. Introduction**
on fuzzy logic.
uncertainty and deductions.
others detailed in literature.
## **From Fuzzy Datalog to Multivalued Knowledge-Base**
Agnes Achs *University of Pecs Faculty of Engineering, Hungary*
### **1. Introduction**
24 Fuzzy Logic – Algorithms, Techniques and Implementations
Sherif,M.,& Hovland,C,I. (1961). *Social judgment: Assimilation and contrast effects in communication and attitude change.* New Haven: Yale University Press. Smithson, M. (1987). *Fuzzy set analysis for the behavioral and social sciences*. New York:
Takemura, K. (1999). A fuzzy linear regression analysis for fuzzy input-output data using
Takemura, K. (2000). Vagueness in human judgment and decision making. In Z. Q. Liu & S.
Takemura, K. (2004). Fuzzy logistic regression analysis for fuzzy input and output data.
Takemura, K. (2005). Fuzzy least squares regression analysis for social judgment study.
Takemura, K. (2007). Ambiguous comparative judgment: Fuzzy set model and data analysis.
Takemura, K. ,Matsumoto,T.,Matsuyama,S.,& Kobayashi,M., (2011). Analysis of consumer's
Takemura,K. (1996). *Psychology of decision making.* Tokyo:Fukumura Syuppan. (in Japanese). Takemura,K. (2011) Model of multi-attribute decision making and good decision.
Zadeh,A. (1973). Outline of a new approach to the analysis of complex systems and decision processes, *IEEE Transactions on Systems, Man and Cybernetics, SMC* 3(1), 28-44.
the least squares method under linear constraints and its application to fuzzy rating
Miyamoto (Eds), *Soft Computing for Human Centered Machines* (pp. 249–281). Tokyo:
Proceedings of the joint 2nd International Conference on Soft Computing and Intelligent Systems and the 5th International Symposium on Advanced Intelligent
ambiguous comparative judgment. *Discussion Paper, Department of Psychology,*
Smithson,M.(1989) Ignorance and uncertainty. New York: Springer-Verlag-
data. *Journal of Advanced Computational Intelligence,* 3, 36–40.
*Journal of Advanced Computational Intelligence,* 9, 461–466.
Operations Research,56(10),583-590 (In Japanese)
Zadeh,A. (1965). Fuzzy sets, *Information and Control*, 8, 338-353.
Tversky, A. (1969). Intransitivity of preferences. *Psychological Review,* 76, 31–48. Wittgenstein,L. (1953). *Philosophical investigations.* New York:MacMillan.
Systems 2004 (WE8-5), Yokohama, Japan.
*Japanese Psychology Research,* 49, 148–156.
Springer-Verlag.
Springer Verlag.
*Waseda University*.
Despite the fact that people have very different and ambiguous concepts and knowledge, they are able to talk to one another. How does human mind work? How can people give answers to questions? Modelling human conversation and knowledge demands to deal with uncertainty and deductions.
Human knowledge consists of static and dynamic knowledge chunks. The static ones include the so called lexical knowledge or the ability to sense similarities between facts and between predicates. Through dynamic attainments one can make deductions or one can give answers to a question. There are several and very different approaches to make a model of human knowledge, but one of the most common and widespread fields of research is based on fuzzy logic.
Fuzzy sets theory, proposed by Zadeh (1965), is a realistic and practical means to describe the world that we live in. The method has successfully been applied in various fields, among others in decision making, logic programming, and approximate reasoning. In the last decade, a number of papers have dealt with that subject, e.g. (Formato et al 2000, Sessa 2002, Medina et al 2004, Straccia et al 2009). They deal with different aspects of modelling and handling uncertainty. (Straccia 2008) gives a detailed overview of this topic with widespread references. Our investigations have begun independently of these works, and have run parallel to them. Of course there are some similar features, but our model differs from the others detailed in literature.
As a generalization of fuzzy sets, intuitionistic fuzzy sets were presented by Atanassov (Atanassov 1983), and have allowed people to deal with uncertainty and information in a much broader perspective. Another well-known generalization of an ordinary fuzzy set is the interval-valued fuzzy set, which was first introduced by Zadeh (Zadeh 1975). These generalizations make descriptions and models of the world more realistic, and practical.
In the beginning, our knowledge-base model was based on the concept of fuzzy logic, later on it was extended to intuitionistic and interval-valued logic. In this model, the static part is a background knowledge module, while the dynamic part consists of a Datalog based deduction mechanism. To develop this mechanism, it was necessary to generalize the Datalog language and to extend it into fuzzy and intuitionistic direction. (Achs 1995, 2007, 2010).
From Fuzzy Datalog to Multivalued Knowledge-Base 27
this model it is impossible to make any true fact false and still have a model consistent with
An interpretation assigns truth or falsehood to every possible instance of the program's predicates. An interpretation is a model, if it makes the rules true, no matter what assignment of values from the domain is made for the variables in each rule. Although there are infinite many implications, it is proved that it is enough to consider only the Herbrand
The Herbrand universe of a program *P* (denoted by *HP*) is the set of all possible ground terms constructed by using constants and function symbols occurring in *P*. The Herbrand base of *P* (*BP*) is the set of all possible ground atoms whose predicate symbols occur in *P* and
In general, a term is a variable, a constant or a complex term of the form *f(t1, …, tn)*, where *f* is a function symbol and *t1, …, tn* are terms. An atom is a formula of the form *p(t)*, where *p* is a predicate symbol of a finite arity (say *n*) and *t* is a sequence of terms of length *n* (arguments). A literal is either an atom (positive literal) or its negation (negative literal). A term, atom or literal is ground if it is free of variables. As in fuzzy extension, we did not deal with function symbols, so in our case the ground terms are the constants of the program.
In the case of Datalog programs there are several equivalent approaches to define the semantics of the program. In fuzzy extension we mainly rely on the fixed-point base aspect. The above concepts are detailed in classical works such as (Ceri et al 1990, Loyd 1990,
In fuzzy Datalog (fDATALOG) the facts can be completed with an uncertainty level, the rules with an uncertainty level and an implication operator. With the use of this operator and these levels deductions can be made. As in classical cases, logical correctness is extremely important as well, i.e., the consequence must be a model of the program. This means that for each rule of the program, the truth-value of the fuzzy implication following
β
≥ *0),*
*A1,…,An (n*
where *A* is an atom (the head of the rule), *A1,…,An* are literals (the body of the rule); *I* is an
For getting a finite result, all the rules in the program must be safe. An fDATALOG rule is safe if all variables occurring in the head also occur in the body, and all variables occurring in a negative literal also occur in a positive one. An fDATALOG program is a finite set of
∈ (0,1] (the level of the rule).
*; I*, where r is a formula of the form
the rule has to be at least as large as the given uncertainty level.
*A* ←
**2.1.1 Syntax and semantics of fuzzy datalog**
**Definition 1**. An fDATALOG rule is a triplet *r;*
β
More precisely, the notion of fuzzy rule is the following:
interpretation defined on the Herbrand universe and the Herbrand base.
the database.
Ullman 1988).
**2.1 Fuzzy Datalog**
implication operator and
safe fDATALOG rules.
whose arguments are elements of *HP*.
In many frameworks, in order to answer a query, we have to compute the whole intended model by a bottom-up fixed-point computation and then answer with the evaluation of the query in this model. This always requires computing a whole model, even if not all the facts and rules are required to determine answer. Therefore a possible top-down like evaluation algorithm has been developed for our model. This algorithm is not a pure top-down one but the combination of top down and bottom up evaluations. Our aim is to improve this algorithm and perhaps to develop a pure top down evaluation based on fuzzy or multivalued unification algorithm. There are fuzzy unification algorithms described for example in (Alsinet et al 1998, Formato et al 2000, Virtanen 1994), but they are inappropriate for evaluating our knowledge-base.
However, the concept of (Julian-Iranzo et al 2009, 2010) is similar but not identical with one of our former ideas about evaluating of special fuzzy Datalog programs (Achs 2006). Reading these papers has led to the assumption that this former idea may be the base of a top-down-like evaluation strategy in special multivalued cases as well. Based on this idea, a multivalued unification algorithm was developed and used for to determine the conclusion of a multivalued knowledge-base.
In this chapter this possible model for handling uncertain information will be provided. This model is based on the multivalued extensions of Datalog. Starting from fuzzy Datalog, the concept of intuitionistic Datalog and bipolar Datalog will be described. This will be the first pillar of the knowledge-base. The second one deals with the similarities of facts and concepts. These similarities are handled with proximity relations. The third component connects the first two with each other. In the final part of the paper, an evaluating algorithm is presented. It is discussed in general, but in special cases it is based on fuzzy, or multivalued unification, which is also mentioned.
### **2. Extensions of datalog**
When one builds a knowledge-base, it is very important to deal with a database management system. It is based on the relational data model developed by Codd in 1970. This model is a very useful one, but it can not handle every problem. For example, the standard query language for relational databases (SQL) is not Turing-complete, in particular it lacks recursion and therefore concepts like transitive closure of a relation can not be expressed in SQL. Along with other problems this is why different extensions of the relational data model or the development of other kinds of models are necessary. A more complete one is the world of deductive databases. A deductive database consists of facts and rules, and a query is answered by building chains of deductions. Therefore the term of deductive database highlights the ability to use a logic programming style for expressing deductions concerning the contents of a database. One of the best known deductive database query languages is Datalog.
As any deductive database, a Datalog program consists of facts and rules, which can be regarded as first order logic formulas. Using these rules, new facts can be inferred from the program's facts so that the consequence of a program will be logically correct. This means that evaluating the program, the result is a model of the formulas belonging to the rules. On the other hand, it is also important that this model will contain only those true facts, which are the consequences of the program; that is, the minimality of this model is expected, i.e. in 26 Fuzzy Logic – Algorithms, Techniques and Implementations
In many frameworks, in order to answer a query, we have to compute the whole intended model by a bottom-up fixed-point computation and then answer with the evaluation of the query in this model. This always requires computing a whole model, even if not all the facts and rules are required to determine answer. Therefore a possible top-down like evaluation algorithm has been developed for our model. This algorithm is not a pure top-down one but the combination of top down and bottom up evaluations. Our aim is to improve this algorithm and perhaps to develop a pure top down evaluation based on fuzzy or multivalued unification algorithm. There are fuzzy unification algorithms described for example in (Alsinet et al 1998, Formato et al 2000, Virtanen 1994), but they are inappropriate
However, the concept of (Julian-Iranzo et al 2009, 2010) is similar but not identical with one of our former ideas about evaluating of special fuzzy Datalog programs (Achs 2006). Reading these papers has led to the assumption that this former idea may be the base of a top-down-like evaluation strategy in special multivalued cases as well. Based on this idea, a multivalued unification algorithm was developed and used for to determine the conclusion
In this chapter this possible model for handling uncertain information will be provided. This model is based on the multivalued extensions of Datalog. Starting from fuzzy Datalog, the concept of intuitionistic Datalog and bipolar Datalog will be described. This will be the first pillar of the knowledge-base. The second one deals with the similarities of facts and concepts. These similarities are handled with proximity relations. The third component connects the first two with each other. In the final part of the paper, an evaluating algorithm is presented. It is discussed in general, but in special cases it is based on fuzzy, or
When one builds a knowledge-base, it is very important to deal with a database management system. It is based on the relational data model developed by Codd in 1970. This model is a very useful one, but it can not handle every problem. For example, the standard query language for relational databases (SQL) is not Turing-complete, in particular it lacks recursion and therefore concepts like transitive closure of a relation can not be expressed in SQL. Along with other problems this is why different extensions of the relational data model or the development of other kinds of models are necessary. A more complete one is the world of deductive databases. A deductive database consists of facts and rules, and a query is answered by building chains of deductions. Therefore the term of deductive database highlights the ability to use a logic programming style for expressing deductions concerning the contents of a database. One of the best known deductive
As any deductive database, a Datalog program consists of facts and rules, which can be regarded as first order logic formulas. Using these rules, new facts can be inferred from the program's facts so that the consequence of a program will be logically correct. This means that evaluating the program, the result is a model of the formulas belonging to the rules. On the other hand, it is also important that this model will contain only those true facts, which are the consequences of the program; that is, the minimality of this model is expected, i.e. in
for evaluating our knowledge-base.
of a multivalued knowledge-base.
**2. Extensions of datalog**
database query languages is Datalog.
multivalued unification, which is also mentioned.
this model it is impossible to make any true fact false and still have a model consistent with the database.
An interpretation assigns truth or falsehood to every possible instance of the program's predicates. An interpretation is a model, if it makes the rules true, no matter what assignment of values from the domain is made for the variables in each rule. Although there are infinite many implications, it is proved that it is enough to consider only the Herbrand interpretation defined on the Herbrand universe and the Herbrand base.
The Herbrand universe of a program *P* (denoted by *HP*) is the set of all possible ground terms constructed by using constants and function symbols occurring in *P*. The Herbrand base of *P* (*BP*) is the set of all possible ground atoms whose predicate symbols occur in *P* and whose arguments are elements of *HP*.
In general, a term is a variable, a constant or a complex term of the form *f(t1, …, tn)*, where *f* is a function symbol and *t1, …, tn* are terms. An atom is a formula of the form *p(t)*, where *p* is a predicate symbol of a finite arity (say *n*) and *t* is a sequence of terms of length *n* (arguments). A literal is either an atom (positive literal) or its negation (negative literal). A term, atom or literal is ground if it is free of variables. As in fuzzy extension, we did not deal with function symbols, so in our case the ground terms are the constants of the program.
In the case of Datalog programs there are several equivalent approaches to define the semantics of the program. In fuzzy extension we mainly rely on the fixed-point base aspect. The above concepts are detailed in classical works such as (Ceri et al 1990, Loyd 1990, Ullman 1988).
### **2.1 Fuzzy Datalog**
In fuzzy Datalog (fDATALOG) the facts can be completed with an uncertainty level, the rules with an uncertainty level and an implication operator. With the use of this operator and these levels deductions can be made. As in classical cases, logical correctness is extremely important as well, i.e., the consequence must be a model of the program. This means that for each rule of the program, the truth-value of the fuzzy implication following the rule has to be at least as large as the given uncertainty level.
### **2.1.1 Syntax and semantics of fuzzy datalog**
More precisely, the notion of fuzzy rule is the following:
**Definition 1**. An fDATALOG rule is a triplet *r;* β*; I*, where r is a formula of the form
$$A \gets A\_1, \ldots, A\_n \qquad \quad (n \ge 0),$$
where *A* is an atom (the head of the rule), *A1,…,An* are literals (the body of the rule); *I* is an implication operator and β∈ (0,1] (the level of the rule).
For getting a finite result, all the rules in the program must be safe. An fDATALOG rule is safe if all variables occurring in the head also occur in the body, and all variables occurring in a negative literal also occur in a positive one. An fDATALOG program is a finite set of safe fDATALOG rules.
From Fuzzy Datalog to Multivalued Knowledge-Base 29
*T1 = T(T0)*
*Tn = T(Tn-1)* ...
It is clear that *NTP* is inflationary transformation over *L*, and if *P* is negation-free, then *NTP* is monotone as well. (A transformation *T* is inflationary if *X* ≤ *T(X)* for every *X* ∈ *L* and it is
In (Ceri et al 1990) it is shown that an inflationary transformation over a complete lattice has a fixed point and if it is monotone then it has a least fixed point (Loyd 1990). Therefore *NTP* has a fixed point, i.e. there exists an *X* ∈ *F(BP)* for which *NTP(X)* = *X*. If *P* is positive, then *X*
The fixed point of the transformation will be denoted by *lfp(NTP)*. It can be shown (Achs
α*Ai*
β
*A* the condition *I(*
α
In the case of c because of the construction of *T0 Ai* is not negative, that is *Ai* is not among the
**Proof** In the case of a positive Datalog program, the least fixed point is the least model (Ceri et al 1990, Ullman 1988). In the case of fuzzy Datalog, according to the definition of the consequence transformation, the level of the rule's head is the least value satisfying the criterion of modelness. The application of the transformations may arise only one problem. A lower level would be ordered to the same rule's head, but according to the definitions we should accept the higher value. But such a case can arise only in the case of programs
According to the above statements, the meaning of the programs can be defined by this
**Definition 3.** *lfp(NTP)* is the nondeterministic semantics of fDATALOG *P* program.
*A)* ∈ *lfp(NTP)* and *(|Ai|,*
α*body,* α*<sup>A</sup> ) = 1* ≥
**Proposition 1**. For negation-free fDATALOG program *P lfp(NTP)* is the least model.
*)* ∉ *lfp(NTP).*
α*Ai*
*body = 0*, so *I(*
containing any negation. Therefore the proposition is true.
An ordering relation can be defined over *F(BP)*. For *G; H : BP* → [*0; 1*]; *G* ≤ *H* iff *(*
ω } if ω
is a limit ordinal.
*)* ∈ *lfp(NTP), 1*≤ *i* ≤ *n*.
α*body,* α*<sup>A</sup> )* ≥ β
, namely *lfp(NTP)* is a model.
∀*d* ∈ *BP)*
is realized.
*=* least upper bound of { *Tn | n* <
*G(d)* ≤ *H(d)*. It easily can be seen that *L = (F(BP),* ≤ *)* is a complete lattice.
is the least fixed point. (That is for any *Z=T(Z)* : *X* ≤ *Z*.)
**Proof** In *ground(P)* there are rules in the next forms:
*; I);* ∃*i : (|Ai|,*
α
In the case of a, b because of the construction of
Moreover, the next proposition is true as well.
1995) that this fixed point is a model of *P*.
**Theorem 1**. *lfp(NTP)* is a model of *P*.
β*; I); (A,* α
β
*= 0*, therefore
and let
a/ *(A* ←*;*
facts, so
fixed point:
β*; I).* b/ *(A* ← *A1, ..., An;*
c/ *(A* ← *A1, ..., An;*
α*Ai* *T*ω
monotone if *T(X)* ≤ *T(Y)* if *X* ≤ *Y*).
There is a special type of rule, called fact. A fact has the form *A* ←*;* β*; I*. From now on, the facts are referred as (*A,*β), because according to implication *I*, the level of *A* easily can be computed and in the case of the implication operators detailed in this chapter it is β.
For defining the meaning of a program, we need again the concepts of Herbrand universe and Herbrand base, but this time they are based on fuzzy logic. Now a ground instance of a rule *r;* β*; I* in *P* is a rule obtained from *r* by replacing every variable in *r* with a constant of *HP*. The set of all ground instances of *r;* β*; I* is denoted by *ground(r);* β*; I*. The ground instance of *P* is *ground(P) =* ∪ *(r; I;* β*)*∈*P (ground(r); I;* β*)*.
An interpretation of a program *P* is a fuzzy set of the program's Herbrand base, *BP*, i.e. it is: ∪ *<sup>A</sup>*∈*BP (A;* α*A).* An interpretation is a model of *P* if for each *(A* ← *A1,…,An;* β*; I)* ∈ *ground(P)*
$$I(\alpha\_{A1\_{A\gets\mathcal{A}}An}, \alpha\_{\mathcal{A}}) \ge \beta.$$
A model *M* is least if for any model *N*, *M* ≤ *N*. A model *M* is minimal if there is not any model *N*, where *N* ≤ *M*.
To be short α *A1*∧*...*∧ *An* will be denoted as αbody and αA as αhead.
In the extensions of Datalog several implication operators are used, but all cases are restricted to min-max conjunction and disjunction, and to the complement to 1 as negation. So: α*A*∧*B = min(*α*A,* α*B),* α*A*∨*B = max(*α*A,* α*B)* and α¬*A = 1* − α*A*.
The semantics of fDATALOG is defined as the fixed points of consequence transformations. Depending on evaluating sequences two semantics can be defined: a deterministic and a nondeterministic one. Further on only the nondeterministic semantics will be discussed, the deterministic one is detailed in (Achs 2010). It was proved that the two semantics are equivalent in the case of negation- and function-free fDatalog programs, but they differ if the program has any negation. In this case merely the nondeterministic semantics is applicable. The nondeterministic transformation is as follows:
**Definition 2.** Let *BP* be the Herbrand base of the program *P*, and let *F(BP)* denote the set of all fuzzy sets over *BP*. The consequence transformation *NTP* : *F(BP)* → *F(BP)* is defined as
$$NT\_F(\mathcal{X}) = \{ (\mathcal{A}, \mathcal{\alpha}\_{\mathcal{A}}) \} \cup \mathcal{X} \tag{1}$$
where
$$(\mathcal{A} \gets \mathcal{A}\_{\mathcal{l}}, \dots, \mathcal{A}\_{\mathcal{n}}; \mathcal{\mathcal{J}} \mathcal{I}) \in \operatorname{ground}(\mathcal{P}), \ (\lvert A\_{i} \rvert, \lVert \mathcal{A}\_{\mathcal{l}i} \rangle \in \mathcal{X}, \ (\mathsf{1} \leq \mathsf{i} \leq \mathsf{n}); \mathcal{I})$$
α*A = max(0, min{*γ *| I(*α*body,* γ*)* ≥ β*}).*
It can be proved that this transformation has a fixed point. To prove it, let us define the powers of a transformation:
For any *T : F(BP)* → *F(BP)* transformation let
$$T\_0 = \{ \cup \{ (\mathbf{A}, \alpha\_{\mathbf{A}}) \} \mid (\mathbf{A} \leftarrow \text{; I}; \mathfrak{f}) \in \text{ground}(\mathbf{P}), \alpha\_{\mathbf{A}} = \max(0, \min\{ \gamma \mid \mathbf{I}(1, \gamma) \ge \beta \}) \mid \cup \{ (\mathbf{A}, \bullet) \mid \exists \ (\mathbf{A}, \bullet) \in \text{ground}(\mathbf{P}) \}$$
and let
28 Fuzzy Logic – Algorithms, Techniques and Implementations
For defining the meaning of a program, we need again the concepts of Herbrand universe and Herbrand base, but this time they are based on fuzzy logic. Now a ground instance of a
An interpretation of a program *P* is a fuzzy set of the program's Herbrand base, *BP*, i.e. it is:
A model *M* is least if for any model *N*, *M* ≤ *N*. A model *M* is minimal if there is not any
In the extensions of Datalog several implication operators are used, but all cases are restricted to min-max conjunction and disjunction, and to the complement to 1 as negation.
The semantics of fDATALOG is defined as the fixed points of consequence transformations. Depending on evaluating sequences two semantics can be defined: a deterministic and a nondeterministic one. Further on only the nondeterministic semantics will be discussed, the deterministic one is detailed in (Achs 2010). It was proved that the two semantics are equivalent in the case of negation- and function-free fDatalog programs, but they differ if the program has any negation. In this case merely the nondeterministic semantics is
**Definition 2.** Let *BP* be the Herbrand base of the program *P*, and let *F(BP)* denote the set of all fuzzy sets over *BP*. The consequence transformation *NTP* : *F(BP)* → *F(BP)* is defined as
> α*A )}* ∪
*ground(P), (|Ai|,*
γ *| I(*α*body,* γ*)* ≥ β*}).*
It can be proved that this transformation has a fixed point. To prove it, let us define the
*ground(P),*
α*A1 ,…,* α*An).*
α
*body* = *min(*
α*Ai)* ∈ *X, (1* ≤ *i* ≤ *n);*
*A= max(0, min{*
*ground(P)}*
γ *| I(1,* γ*)* ≥ β*}) }* ∪
*NTP(X) = {(A,*
*A = max(0, min{*
α
*; I* in *P* is a rule obtained from *r* by replacing every variable in *r* with a constant of
*; I* is denoted by *ground(r);*
β
β
β
*X ,* (1)
), because according to implication *I*, the level of *A* easily can be
*; I*. From now on, the
β.
*; I*. The ground instance
*; I)* ∈ *ground(P)*
There is a special type of rule, called fact. A fact has the form *A* ←*;*
*P (ground(r); I;*
*I(*α *A1*∧*...*∧ *An ,*α*A)* ≥ β*.*
α*A,* α*B)* and α¬*A = 1* − α*A*.
applicable. The nondeterministic transformation is as follows:
*A1,…,An;*
*Ai*, if *Ai* is negative) and
For any *T : F(BP)* → *F(BP)* transformation let
β*; I )* ∈
α
← *; I;* β*)* ∈
> ∃ *(B* ← *...*¬ *A...; I;* β*)* ∈
*{(A, 0) |*
*An* will be denoted as αbody and αA as αhead.
computed and in the case of the implication operators detailed in this chapter it is
β
β*)*.
*A).* An interpretation is a model of *P* if for each *(A* ← *A1,…,An;*
β
β*)*∈
*HP*. The set of all ground instances of *r;*
facts are referred as (*A,*
of *P* is *ground(P) =* ∪ *(r; I;*
α
model *N*, where *N* ≤ *M*.
α *A1*∧*...*∧
> α*A,* α*B),* α*A*∨*B = max(*
> > *(A* ←
rule *r;* β
∪ *<sup>A</sup>*∈*BP (A;*
To be short
So: α*A*∧*B = min(*
where
literal, and
¬
*T0 = {* ∪*{(A,*α*A)} | (A*
powers of a transformation:
$$T\_1 = T(T\_0)$$
$$T\_n = T(T\_{n-1})$$
$$\dots$$
$$T\_{\alpha} \text{=least upper bound of } \{ T\_n \mid n \le \alpha \} \text{ if } \alpha \text{ is a limit ordinal.}$$
An ordering relation can be defined over *F(BP)*. For *G; H : BP* → [*0; 1*]; *G* ≤ *H* iff *(*∀*d* ∈ *BP) G(d)* ≤ *H(d)*. It easily can be seen that *L = (F(BP),* ≤ *)* is a complete lattice.
It is clear that *NTP* is inflationary transformation over *L*, and if *P* is negation-free, then *NTP* is monotone as well. (A transformation *T* is inflationary if *X* ≤ *T(X)* for every *X* ∈ *L* and it is monotone if *T(X)* ≤ *T(Y)* if *X* ≤ *Y*).
In (Ceri et al 1990) it is shown that an inflationary transformation over a complete lattice has a fixed point and if it is monotone then it has a least fixed point (Loyd 1990). Therefore *NTP* has a fixed point, i.e. there exists an *X* ∈ *F(BP)* for which *NTP(X)* = *X*. If *P* is positive, then *X* is the least fixed point. (That is for any *Z=T(Z)* : *X* ≤ *Z*.)
The fixed point of the transformation will be denoted by *lfp(NTP)*. It can be shown (Achs 1995) that this fixed point is a model of *P*.
**Theorem 1**. *lfp(NTP)* is a model of *P*.
**Proof** In *ground(P)* there are rules in the next forms:
a/ *(A* ←*;* β*; I).* b/ *(A* ← *A1, ..., An;* β*; I); (A,* α*A)* ∈ *lfp(NTP)* and *(|Ai|,* α*Ai )* ∈ *lfp(NTP), 1*≤ *i* ≤ *n*. c/ *(A* ← *A1, ..., An;* β*; I);* ∃*i : (|Ai|,* α*Ai )* ∉ *lfp(NTP).*
In the case of a, b because of the construction of α*A* the condition *I(*α*body,* α*<sup>A</sup> )* ≥ β is realized. In the case of c because of the construction of *T0 Ai* is not negative, that is *Ai* is not among the facts, so α*Ai = 0*, therefore α*body = 0*, so *I(*α*body,* α*<sup>A</sup> ) = 1* ≥ β, namely *lfp(NTP)* is a model.
Moreover, the next proposition is true as well.
**Proposition 1**. For negation-free fDATALOG program *P lfp(NTP)* is the least model.
**Proof** In the case of a positive Datalog program, the least fixed point is the least model (Ceri et al 1990, Ullman 1988). In the case of fuzzy Datalog, according to the definition of the consequence transformation, the level of the rule's head is the least value satisfying the criterion of modelness. The application of the transformations may arise only one problem. A lower level would be ordered to the same rule's head, but according to the definitions we should accept the higher value. But such a case can arise only in the case of programs containing any negation. Therefore the proposition is true.
According to the above statements, the meaning of the programs can be defined by this fixed point:
**Definition 3.** *lfp(NTP)* is the nondeterministic semantics of fDATALOG *P* program.
From Fuzzy Datalog to Multivalued Knowledge-Base 31
As the next examples show, there some problems would arise if the program had any
*M1 = {(p(a), 0.7) } and M2 = {(q(b), 1) }.*
*lfp(NTP) = {(r(a), 0.8), (p(a), 0.6), (q(a), 0.5)},*
*lfp(NTP) = {(r(a), 0.8), (p(a), 0.5), (q(a), 0.5)}.* According to the above examples, in the case of programs containing negation there are problems with the model's minimality. However, the nondeterministic semantics – *lfp(NTP)* – is minimal under certain conditions. These conditions are referred to as stratification. Stratification gives an evaluating sequence in which the literals are evaluated before
To stratify a program, it is necessary to define the concept of dependency graph. This is a directed graph, whose nodes are the predicates of *P*. There is an arc from predicate *p* to predicate *q* if there is a rule whose body contains *p* or ¬*p* and whose head predicate is *q*. A program is recursive, if its dependency graph has one or more cycles. A program is stratified if whenever there is a rule with head predicate *p* and a negated body literal ¬q,
The stratification of a program *P* is a partition of the predicate symbols of *P* into subsets
b/ if *p* ∈ *Pi* and *q* ∈ *Pj* and there is a rule with the head *p* whose body contains ¬*q*, then *i* > *j*. Stratification specifies an order of evaluation. The rules whose head-predicates are in P*1* are evaluated first, then those whose head-predicates are in P*2* and so on. The sets *P1,..., Pn* are
*q(b); IG; 0.7.*
So *lfp(NTP) = {(p(a), 0.8), (r(b), 0.6), (q(a, b), 0.6), (q(b, a), 0.5), (s(a), 0.7), (s(b), 0.7) }.*
*p(a)* ← ¬
**Example 3.** This example shows that there is a difference between the fixed points.
**Example 2.** Look at the next one-rule program:
This program has no least model, only two minimal ones:
The result depends on the evaluation order. If it is 1., 2., 3., 4., then
(The result of the above fixed point algorithm is *M1*.)
negation.
1. (r(a), 0.8).
while in the order 1., 3., 2., 4.
**2.1.2 Stratified fuzzy datalog**
there is no path in the dependency graph from *p* to *q*.
*P1,..., Pn* such that the following conditions are satisfied:
a/ if *p* ∈ *Pi* and *q* ∈ *Pj* and there is an edge from *q* to *p* then *i* ≥ *j*;
negating them.
2. p(x) ← r(x),¬ q(x); 0.6; IG. 3. q(x) ← r(x); 0.5; IG. 4. p(x) ← q(x); 0.8; IG.
To compute the level of rule-heads, we need the concept of uncertainty level function.
**Definition 4**. The uncertainty-level function is:
$$f(\mathsf{I}, \mathsf{\alpha}, \mathsf{\beta}) = \min \left( \| \, \mathsf{\gamma} \mid \mathsf{I} \, (\mathsf{\alpha}, \mathsf{\gamma}) \geq \mathsf{\beta} \right).$$
According to this function the level of a rule-head is: α*head = f(I,*α*body,* β*).*
It is an extremely important question whether the fixed-point algorithm terminates or not. It depends on the feature of uncertainty level function:
**Proposition 2**. If *f(I,* α*,* β*)* ≤ α for∀α∈ [0; 1] then the fixed point algorithm terminates.
**Proof** As *P* is finite, therefore in the fixed point there are only finite many ground atoms. The only problem may occur with the level of recursive predicates, but according to the above property of the uncertainty-level function, the level of the rule's head cannot be greater than any former one, so this algorithm must terminate.
In former papers (Achs 1995, Achs 2006) several implications were detailed (the operators are detailed in (Dubois et al, 1991)), for now three are chosen from these. The values of their uncertainty-level functions can be easily computed. They are the following:
$$\begin{array}{ll} \text{Gödel} & I\_{\mathbb{C}}(\alpha,\gamma) = \begin{cases} 1 & \alpha \le \gamma \\ \gamma & \text{otherwise} \end{cases} & f(\mathbb{I}\_{\mathbb{C}},\alpha,\beta) = \min(\alpha,\beta) \\\ \text{Lukasisiewicz} & I\_{\mathbb{L}}(\alpha,\gamma) = \begin{cases} 1 & \alpha \le \gamma \\ 1 - \alpha + \gamma & \text{otherwise} \end{cases} & f(\mathbb{I}\_{\mathbb{L}},\alpha,\beta) = \max(0, \alpha + \beta - 1) \\\ \text{Kleene-Dienes} & I\_{\mathbb{K}}(\alpha,\gamma) = \max(1 - \alpha, \gamma) & f(\mathbb{I}\_{\mathbb{K}},\alpha,\beta) = \begin{cases} 0 & \alpha + \beta \le 1 \\ \beta & \alpha + \beta > 1 \end{cases} & f(\mathbb{I}\_{\mathbb{C}},\alpha,\beta) = \begin{cases} 1 & \alpha = \beta \\ \beta & \alpha + \beta > 1 \end{cases} \end{array}$$
It is obvious that *IG* and *IL* satisfy the condition of Proposition 2, and it is easy to see that in the case of *IK* the fixed point algorithm terminates as well. (Among the operators of (Dubois et al, 1991) there is one for which the algorithm does not terminate and one for which the uncertainty-level function does not exists.)
**Example 1.** Let us consider the next program:
$$\begin{array}{l} \text{( $p(a)$ , $0.8$ ).}\\ \text{( $r(b)$ , $0.6$ ).}\\ q(\text{x}, y) \leftarrow p(\text{x}); \; r(y); \; 0.7; \; I\_{G.}\\ q(\text{x}, y) \leftarrow q(y, \text{x}); \; 0.9; \; I\_{L.}\\ s(\text{x}) \leftarrow q(\text{x}, y); \; 0.7; \; I\_{K.} \end{array}$$
Then *T0* = {*(p(a), 0.8), (r(b), 0.6)* } and the computed atoms are:
$$\begin{aligned} \text{( $q(a,b)$ , min(min(0.8, 0.6), 0.7) = 0.6 $);}\\ \text{($ q(b,a) $, max(0, 0.6 + 0.9 - 1) = 0.5$ );}\\ \text{( $s(a)$ )'} \begin{cases} 0 & 0.6 + 0.7 \le 1 \\ 0.7 & 0.6 + 0.7 > 1 \end{cases} = 0.7); \\ \text{( $s(b)$ )'} \begin{cases} 0 & 0.5 + 0.7 \le 1 \\ 0.7 & 0.5 + 0.7 > 1 \end{cases} = 0.7); \end{aligned}$$
So *lfp(NTP) = {(p(a), 0.8), (r(b), 0.6), (q(a, b), 0.6), (q(b, a), 0.5), (s(a), 0.7), (s(b), 0.7) }.*
As the next examples show, there some problems would arise if the program had any negation.
**Example 2.** Look at the next one-rule program:
$$p(a) \leftarrow \neg q(b); \text{ ló: } 0.7.$$
This program has no least model, only two minimal ones:
*M1 = {(p(a), 0.7) } and M2 = {(q(b), 1) }.*
(The result of the above fixed point algorithm is *M1*.)
**Example 3.** This example shows that there is a difference between the fixed points.
1. (r(a), 0.8).
30 Fuzzy Logic – Algorithms, Techniques and Implementations
γ *| I (*α*,* γ *)* ≥ β *}).*
It is an extremely important question whether the fixed-point algorithm terminates or not. It
**Proof** As *P* is finite, therefore in the fixed point there are only finite many ground atoms. The only problem may occur with the level of recursive predicates, but according to the above property of the uncertainty-level function, the level of the rule's head cannot be
In former papers (Achs 1995, Achs 2006) several implications were detailed (the operators are detailed in (Dubois et al, 1991)), for now three are chosen from these. The values of their
γ
*otherwise*
It is obvious that *IG* and *IL* satisfy the condition of Proposition 2, and it is easy to see that in the case of *IK* the fixed point algorithm terminates as well. (Among the operators of (Dubois et al, 1991) there is one for which the algorithm does not terminate and one for which the
> *q(x, y)* ← *p(x), r(y); 0.7; IG. q(x, y)* ← *q(y, x); 0.9; IL. s(x)* ← *q(x, y); 0.7; IK.*
*(q(a, b), min(min(0.8, 0.6), 0.7) = 0.6); (q(b, a), max(0, 0.6 + 0.9 - 1) = 0.5);* 0 0.6 +0.7 1 ( ( ), 0.7); 0.7 0.6 0.7 1
<sup>≤</sup> <sup>=</sup> + >
0 0.5+0.7 1 ( ( ), 0.7); 0.7 0.5 0.7 1 *s b* <sup>≤</sup> <sup>=</sup> + >
<sup>≤</sup> <sup>=</sup> − + *f(IL,*
*f(IG,*
*) K*
*otherwise*
α*,* γ
α α*head = f(I,*
α*body,* β*).*
> α*,* β*) = min(*
*) = max(0,*
( ,,) <sup>1</sup> α β
βα
*f I +*
α*,* β
α β α*,* β*)*
α *+* β *-1)*
*+*
<sup>≤</sup> <sup>=</sup> <sup>&</sup>gt;
0 1
β
∈ [0; 1] then the fixed point algorithm terminates.
To compute the level of rule-heads, we need the concept of uncertainty level function.
*) = min ({*
**Definition 4**. The uncertainty-level function is:
According to this function the level of a rule-head is:
depends on the feature of uncertainty level function:
α*,* β*)* ≤ α for∀α
Gödel *<sup>G</sup>*
Lukasiewicz *<sup>L</sup>*
Kleene-Dienes *IK(*
uncertainty-level function does not exists.) **Example 1.** Let us consider the next program:
**Proposition 2**. If *f(I,*
*f(I,* α*,* β
greater than any former one, so this algorithm must terminate.
*I*
1
(,)
α γ
*I*
α γ
(,) <sup>1</sup>
α*,*γ
Then *T0* = {*(p(a), 0.8), (r(b), 0.6)* } and the computed atoms are:
*s a*
uncertainty-level functions can be easily computed. They are the following:
1
<sup>≤</sup> <sup>=</sup>
γ
α γ
α γ
*) = max(1-*
*(p(a), 0.8). (r(b), 0.6).*
The result depends on the evaluation order. If it is 1., 2., 3., 4., then
$$\text{lfp}(\text{NT}\_P) = \{ (r(a), 0.8), (p(a), 0.6), (q(a), 0.5) \}.$$
while in the order 1., 3., 2., 4.
$$\text{lfp}(\text{NT}\_P) = \{(r(a), 0.8), (p(a), 0.5), (q(a), 0.5)\}.$$
According to the above examples, in the case of programs containing negation there are problems with the model's minimality. However, the nondeterministic semantics – *lfp(NTP)* – is minimal under certain conditions. These conditions are referred to as stratification. Stratification gives an evaluating sequence in which the literals are evaluated before negating them.
### **2.1.2 Stratified fuzzy datalog**
To stratify a program, it is necessary to define the concept of dependency graph. This is a directed graph, whose nodes are the predicates of *P*. There is an arc from predicate *p* to predicate *q* if there is a rule whose body contains *p* or ¬*p* and whose head predicate is *q*. A program is recursive, if its dependency graph has one or more cycles. A program is stratified if whenever there is a rule with head predicate *p* and a negated body literal ¬q, there is no path in the dependency graph from *p* to *q*.
The stratification of a program *P* is a partition of the predicate symbols of *P* into subsets *P1,..., Pn* such that the following conditions are satisfied:
a/ if *p* ∈ *Pi* and *q* ∈ *Pj* and there is an edge from *q* to *p* then *i* ≥ *j*;
b/ if *p* ∈ *Pi* and *q* ∈ *Pj* and there is a rule with the head *p* whose body contains ¬*q*, then *i* > *j*.
Stratification specifies an order of evaluation. The rules whose head-predicates are in P*1* are evaluated first, then those whose head-predicates are in P*2* and so on. The sets *P1,..., Pn* are
From Fuzzy Datalog to Multivalued Knowledge-Base 33
By induction it will be shown that *Ln* is a minimal model of *P*. For this purpose, we need the
**Definition 5**. An fDATALOG program *P* is semi-positive if its negated predicates are solely
**Proof.** A semi-positive program is almost the same as a positive one, because if *p* is a negated predicate of a rule-body, then it can be replaced by the fact q = ¬p. As p is a fact predicate, therefore the uncertainty level of q may be easily calculated. So the negation can be eliminated from the program and this program has a least fixed point which is the least
because according to the stratification each negative literal of the i-th strata belongs to a
This means that evaluating the rules in the order of stratification, the least fixed point of the program's nondeterministic transformation is the minimal model of the program as well. So: **Proposition 3.** For stratified fDATALOG program *P*, there is an evaluation sequence, in
As shown in Example 4, a program can have more then one stratification. Will the different stratifications yield the same semantics? Fortunately, the answer is yes. (Ceri et al 1990) declares, (Abiteboul et al 1995) proves the theorem, according to which for stratified Datalog programs the resulting minimal model is independent of the actual stratification. That is, two stratifications of a Datalog program yield the same semantics on all inputs. As the order of stratification depends only on the predicates of the program and it is not influenced by the uncertainty levels, therefore this theorem is true in the case of fDATALOG programs as
**Theorem 3.** Let *P* be a stratifiable fDATALOG program. The least fixed point according to
**Example 5.** In Example 3. the right stratified order is 1., 3., 2., 4.; so the least fixed point of
In fuzzy theory, uncertainty is measured by a single value between zero and one, and negation can be calculated as its complement to 1. However, human beings sometimes hesitate expressing these values, that is, there may be some hesitation degree. This illuminates a well-known psychological fact that linguistic negation does not always correspond to the logical one. Based on this observation, as a generalization of fuzzy sets, the concept of intuitionistic fuzzy sets was introduced and developed by Atanassov in 1983 and later (Atanassov 1983, 1999, Atanassov & Gargov 1989). In the next paragraphs some
an arbitrary order of stratification is a unique minimal model of the program.
the program is: *lfp(NTP)* = {*(r(a), 0.8), (p(a), 0.5), (q(a), 0.5)* }*.*
possible multivalued extensions of Datalog will be discussed.
**Theorem 2**. If *P* is a stratified fDATALOG program then *Ln* is a minimal model of *P*.
*\** is semi-positive,
\* , which is minimal model
**Lemma 1.** A semi-positive program *P* has a minimal model: *L = lfp(NTP).*
According to the lemma, *L1* is the least fixed point for *P1\**. Generally *Li-1* ∪ *Pi*
predicate of a lower level strata. So Li is the least fixed point for Pi
for the given stratification. Therefore the next theorem is true:
which *lfp(NTP)* is a minimal model of *P*.
next definition and lemma.
facts.
model.
well.
**2.2 Multivalued datalog**
called the strata of the stratification. A program *P* is called stratified if and only if it admits stratification. There is a very simple method for finding stratification for a stratified program in (Ceri et al 1990, Ullman 1988). Because this algorithm groups the predicates of the program, this is suitable for the fDATALOG programs as well.
The first of the following Datalog programs is not stratified, the other one has more distinct stratifications.
**Example 4**. Consider the one-rule program:
$$p(\mathbf{x}) \leftarrow \neg p(\mathbf{x}).$$
This is not stratified.
The next program has more stratification (Abiteboul et al 1995):
1. *s(x)* ← *r1(x),* ¬*r(x).* 2. *t(x)* ← *r2(x), r(x).* 3. *u(x)* ← *r3(x), t(x).* 4. *v(x)* ← *r4(x), s(x), u(x).*
The program has five distinct stratifications, namely:
```
{1.}, {2.}, {3.}, {4.}
{2.}, {1.}, {3.}, {4.}
{2.}, {3.}, {1.}, {4.}
{1., 2.}, {3.}, {4.}
{2.}, {1., 3.}, {4.}
```
These lead to five different ways of reading the program. As will be seen later, each of them yields the same semantics.
Let *P* be a stratified fDATALOG program with stratification *P1,..., Pn*. Let *Pi \** denote the set of all rules of *P* corresponding to stratum *Pi* , that is the set of all rules whose head-predicate is in *Pi*. Let
$$L\_1 = \text{lfp}(NT\_{P\_1}\text{})\_\prime$$
where the starting point of the computation is *T0* defined earlier.
$$L\_2 = lfp \text{ (NT}\_{P\_2} \text{:)}\_{\prime}$$
where the starting point of the computing is *L1*.
$$L\_n = l! p \text{ (NT } {}\_{P\_n}\text{)}\_{\prime}$$
where the starting point is *Ln-1*.
In other words: the least fixed point - *L1* - corresponding to the first stratum of *P* is computed at first. Once this fixed point has been computed, we can take a step to the next strata.
32 Fuzzy Logic – Algorithms, Techniques and Implementations
called the strata of the stratification. A program *P* is called stratified if and only if it admits stratification. There is a very simple method for finding stratification for a stratified program in (Ceri et al 1990, Ullman 1988). Because this algorithm groups the predicates of the
The first of the following Datalog programs is not stratified, the other one has more distinct
{1.}, {2.}, {3.}, {4.} {2.}, {1.}, {3.}, {4.} {2.}, {3.}, {1.}, {4.} {1., 2.}, {3.}, {4.} {2.}, {1., 3.}, {4.} These lead to five different ways of reading the program. As will be seen later, each of them
all rules of *P* corresponding to stratum *Pi* , that is the set of all rules whose head-predicate is
*L1 = lfp(NTP1*
*L2 = lfp (NTP2*
*Ln = lfp (NT Pn*
In other words: the least fixed point - *L1* - corresponding to the first stratum of *P* is computed at first. Once this fixed point has been computed, we can take a step to the next
*\*),*
*\*),*
*\*),*
*\** denote the set of
Let *P* be a stratified fDATALOG program with stratification *P1,..., Pn*. Let *Pi*
where the starting point of the computation is *T0* defined earlier.
where the starting point of the computing is *L1*.
where the starting point is *Ln-1*.
*p(x)* ← *p(x).*
program, this is suitable for the fDATALOG programs as well.
The next program has more stratification (Abiteboul et al 1995):
The program has five distinct stratifications, namely:
**Example 4**. Consider the one-rule program:
stratifications.
This is not stratified.
2. *t(x)*
3. *u(x)*
4. *v(x)*
1. *s(x)* ← *r1(x),* ¬*r(x).*
← *r2(x), r(x).*
← *r3(x), t(x).*
← *r4(x), s(x), u(x).*
yields the same semantics.
in *Pi*. Let
strata.
By induction it will be shown that *Ln* is a minimal model of *P*. For this purpose, we need the next definition and lemma.
**Definition 5**. An fDATALOG program *P* is semi-positive if its negated predicates are solely facts.
**Lemma 1.** A semi-positive program *P* has a minimal model: *L = lfp(NTP).*
**Proof.** A semi-positive program is almost the same as a positive one, because if *p* is a negated predicate of a rule-body, then it can be replaced by the fact q = ¬p. As p is a fact predicate, therefore the uncertainty level of q may be easily calculated. So the negation can be eliminated from the program and this program has a least fixed point which is the least model.
According to the lemma, *L1* is the least fixed point for *P1\**. Generally *Li-1* ∪ *Pi \** is semi-positive, because according to the stratification each negative literal of the i-th strata belongs to a predicate of a lower level strata. So Li is the least fixed point for Pi \* , which is minimal model for the given stratification. Therefore the next theorem is true:
**Theorem 2**. If *P* is a stratified fDATALOG program then *Ln* is a minimal model of *P*.
This means that evaluating the rules in the order of stratification, the least fixed point of the program's nondeterministic transformation is the minimal model of the program as well. So:
**Proposition 3.** For stratified fDATALOG program *P*, there is an evaluation sequence, in which *lfp(NTP)* is a minimal model of *P*.
As shown in Example 4, a program can have more then one stratification. Will the different stratifications yield the same semantics? Fortunately, the answer is yes. (Ceri et al 1990) declares, (Abiteboul et al 1995) proves the theorem, according to which for stratified Datalog programs the resulting minimal model is independent of the actual stratification. That is, two stratifications of a Datalog program yield the same semantics on all inputs. As the order of stratification depends only on the predicates of the program and it is not influenced by the uncertainty levels, therefore this theorem is true in the case of fDATALOG programs as well.
**Theorem 3.** Let *P* be a stratifiable fDATALOG program. The least fixed point according to an arbitrary order of stratification is a unique minimal model of the program.
**Example 5.** In Example 3. the right stratified order is 1., 3., 2., 4.; so the least fixed point of the program is: *lfp(NTP)* = {*(r(a), 0.8), (p(a), 0.5), (q(a), 0.5)* }*.*
### **2.2 Multivalued datalog**
In fuzzy theory, uncertainty is measured by a single value between zero and one, and negation can be calculated as its complement to 1. However, human beings sometimes hesitate expressing these values, that is, there may be some hesitation degree. This illuminates a well-known psychological fact that linguistic negation does not always correspond to the logical one. Based on this observation, as a generalization of fuzzy sets, the concept of intuitionistic fuzzy sets was introduced and developed by Atanassov in 1983 and later (Atanassov 1983, 1999, Atanassov & Gargov 1989). In the next paragraphs some possible multivalued extensions of Datalog will be discussed.
From Fuzzy Datalog to Multivalued Knowledge-Base 35
The next question is whether this fixed point is a model of *P*. The fixed point is an
Similarly to the proof of Theorem 1, it can easily be proved that this fixed point is a model of the program. For negation-free iDATALOG this is the least model of the program. (Achs
In fDATALOG a fact can be negated by completing its membership degree to 1. In iDATALOG the uncertainty level of a negated fact can be computed according to negators. A negator on *LF* or *LV* is a decreasing mapping ordering *0FV* and *1FV* together (Cornelis et al 2004). The applied negators are relevant for the computational meaning of a program, but they have no influence on the stratification. So for a stratified iDATALOG program *P* there is an evaluation sequence in which *lfp(iNTP)* is a unique minimal model of *P*. Therefore
After defining the syntax and semantics of extended fuzzy Datalog, it is necessary to examine the properties of possible implication operators and the extended uncertainty-level functions. A number of intuitionistic implications are discussed in (Cornelis et al 2004, Atanassov 2005, 2006) and other papers, four of them are the extensions of the above three fuzzy implication operators. Now these operators will be presented and completed by the suitable interval-value operators and the uncertainty-level functions. The computations will
The coordinates of intuitionistic and interval-valued implication operators can be determined by each other. The uncertainty-level functions can be computed according to the applied implication. The connection between *IF* and *IV* and the extended versions of
> α*,*γ
α*',*γ*');*
The studied operators and the related uncertainty-level functions are the following:
*) = (max(1-*
α*',*γ*');* γ*'=(*γ*1,1*γ*2).*
α*2,* γ*1), min(*α*1,* γ*2))*
α*2,* γ
*) = (IV1, IV2);*
α*'=(*α*1,1*α*2);*
*1 }), max ({*
*1 }), min ({*
*1), max(1-*
α*1,* γ*2))*
γ*2 | IF2 (*α*,* γ *)* ≤ β*2 }))*
γ*2 | IV2 (*α*,* γ *)* ≥ β*2 }))*
*ground(P), IFV(*
α*,* γ *)* ≥*FV* β*.*
interpretation of *P*, which is a model, if for each
*A* ←
2010).
*A1 ,…, An;*
*lfp(iNTP)* can be regarded as the semantics of iDATALOG.
uncertainty-level functions are given below: For
*) = (min ({*
*) = (min ({*
**2.2.1.1 Extension of Kleene-dienes implication**
*f (IF,* α*,* β
*f (IV,* α*,* β
not be shown here, only the starting points and results are presented.
*IV(*
*IV1=IF1(*
γ*1 | IF1 (*α*,* γ *)* ≥ β
γ*1 | IV1 (*α*,* γ *)* ≥ β
One possible extension of Kleene-Dienes implication for IFS is:
*IFK(*α*,* γ*) = (max(*
The appropriate computed elements are the following:
*IVK(*α*,* γ
*IV2=1-IF2(*
β*; IFV* ∈
### **2.2.1 Intuitionistic- and interval-valued extensions of datalog**
In intuitionistic fuzzy systems (IFS) and interval-valued systems (IVS) the uncertainty is represented by two values, μ = *(*μ*1,* μ*2)* instead of a single one. In the intuitionistic case the two elements must satisfy the condition μ*1+*μ*<sup>2</sup>* ≤ *1*, while in the interval-valued case the condition is μ*<sup>1</sup>* ≤ μ*2*. If μ = *(*μ*1,* μ*2)* belonging to a predicate *p* is an IFS level, then *p* is definitely true on level μ*1* and definitely false on level μ*<sup>2</sup>*, while in IVS the truth value is between μ*1* and μ*2*. It is obvious that the relation μ*'1* = μ*1*, μ*'2 = 1* − μ*<sup>2</sup>* creates a mutual connection between the two systems. (The equivalence of IVS and IFS was stated first in (Atanassov & Gargov 1989).)
The fixed point theory of programming is based on the theory of lattices. So does the theory of fuzzy Datalog as well, which is based on the lattice of fuzzy sets. The extension of the programs into an intuitionistic and interval-valued direction needs the extension of lattices as well.
**Definition 6.** *LF* and *LV* are lattices of IFS and IVS respectively, where
$$\mathbf{L}\_{F} = \{ (\mathbf{x}\_{1}, \mathbf{x}\_{2}) \in [0, 1]^{2} \mid \mathbf{x}\_{1} + \mathbf{x}\_{2} \le 1 \}, \ (\mathbf{x}\_{1}, \mathbf{x}\_{2}) \lessapprox (y\_{1}, y\_{2}) \Leftrightarrow \mathbf{x}\_{1} \le y\_{1} \text{ and } \mathbf{x}\_{2} \ge y\_{2}.$$
$$\mathbf{L}\_{V} = \{ (\mathbf{x}\_{1}, \mathbf{x}\_{2}) \in [0, 1]^{2} \mid \mathbf{x}\_{1} \le \mathbf{x}\_{2} \}, \ (\mathbf{x}\_{1}, \mathbf{x}\_{2}) \lessapprox\_{V} (y\_{1}, y\_{2}) \Leftrightarrow \mathbf{x}\_{1} \le y\_{1} \text{ and } \mathbf{x}\_{2} \le y\_{2}$$
It can be proved that both *LF* and *LV* are complete lattices (Cornelis et al 2004), so it can be the base of intuitionistic Datalog (ifDATALOG) and interval-valued Datalog (ivDATALOG) as well. (If the distinction is not important, both of them will be denoted by iDATALOG.)
The so called i-extended DATALOG is defined on these lattices, and the necessary concepts are generalizations of the ones presented in Definition 1 and Definition 2. Let us continue to denote by *BP* the Herbrand base of the program *P*, and let *FV(BP)* the set of all IFS or IVS sets over *BP*.
**Definition 7.** The i-extended Datalog program (iDATALOG) is a finite set of safe iDATALOG rules *r;* β*; IFV*;
> α*A = maxFV (0FV, minFV{*γ *| IFV(*α*body,* γ*)* ≥*FV* β *});*
$$f(\underline{\mathsf{I}}\_{FV}\underline{\mathsf{Q}},\underline{\mathsf{Q}}) = \min\_{\mathsf{V}} \iota\_{\mathsf{V}}(\{\underline{\mathsf{y}} \mid \mathsf{I}\_{FV}(\underline{\mathsf{Q}},\underline{\mathsf{y}}) \triangleq\_{\mathsf{V}} \underline{\mathsf{B}})),$$
where α, β, γ are elements of *LF*, *LV* respectively, *IFV = IF* or *IV* is an implication of *LF* or *LV*; *maxFV* = *maxF* or *maxV*; *minFV* = *minF* or *minV* are the max or min operator of *LF* or *LV*; *0FV* is *0F = (0,1)* or *0V = (0,0)* and ≥*FV* is ≥*F* or ≥*V*.
As *iNTP* is inflationary transformation over the complete lattices *LF* or *LV*, thus according to (Ceri et al 1990) it has an inflationary fixed point denoted by *lfp(iNTP)*. If *P* is positive (without negation), *iNTP* is a monotone transformation, so *lfp(iNTP)* is the least fixed point.
34 Fuzzy Logic – Algorithms, Techniques and Implementations
In intuitionistic fuzzy systems (IFS) and interval-valued systems (IVS) the uncertainty is
connection between the two systems. (The equivalence of IVS and IFS was stated first in
The fixed point theory of programming is based on the theory of lattices. So does the theory of fuzzy Datalog as well, which is based on the lattice of fuzzy sets. The extension of the programs into an intuitionistic and interval-valued direction needs the extension of lattices
*1}, (x1,x2)*
It can be proved that both *LF* and *LV* are complete lattices (Cornelis et al 2004), so it can be the base of intuitionistic Datalog (ifDATALOG) and interval-valued Datalog (ivDATALOG) as well. (If the distinction is not important, both of them will be denoted by iDATALOG.)
The so called i-extended DATALOG is defined on these lattices, and the necessary concepts are generalizations of the ones presented in Definition 1 and Definition 2. Let us continue to denote by *BP* the Herbrand base of the program *P*, and let *FV(BP)* the set of all IFS or IVS sets
**Definition 7.** The i-extended Datalog program (iDATALOG) is a finite set of safe
γ *| IFV(*α*body,* γ*)* ≥*FV* β *});*
γ *| IFV(*α*,* γ *)* ≥*FV* β *}),*
*maxFV* = *maxF* or *maxV*; *minFV* = *minF* or *minV* are the max or min operator of *LF* or *LV*; *0FV* is *0F*
As *iNTP* is inflationary transformation over the complete lattices *LF* or *LV*, thus according to (Ceri et al 1990) it has an inflationary fixed point denoted by *lfp(iNTP)*. If *P* is positive (without negation), *iNTP* is a monotone transformation, so *lfp(iNTP)* is the least fixed point.
are elements of *LF*, *LV* respectively, *IFV = IF* or *IV* is an implication of *LF* or *LV*;
*x2}, (x1,x2)*
≤*F (y1,y2)*
≤*V (y1,y2)* ⇔ *x1* ≤
⇔ *x1* ≤
μ*1+*μ
*1* and definitely false on level
≤
≤
*2)* instead of a single one. In the intuitionistic case the
*2)* belonging to a predicate *p* is an IFS level, then *p* is
μ
μ*'1* = μ*1*, μ*'2 = 1* − μ
*<sup>2</sup>* ≤ *1*, while in the interval-valued case the
*y1 and x2*
*y1 and x2*
≥ *y2,*
≤ *y2*
*<sup>2</sup>*, while in IVS the truth value is
*<sup>2</sup>* creates a mutual
**2.2.1 Intuitionistic- and interval-valued extensions of datalog**
*2*. It is obvious that the relation
**Definition 6.** *LF* and *LV* are lattices of IFS and IVS respectively, where
*[0,1]2 | x1+x2*
*[0,1]2 | x1*
μ = *(*μ*1,* μ
represented by two values,
μ*<sup>1</sup>* ≤ μ*2*. If μ = *(*μ*1,* μ
(Atanassov & Gargov 1989).)
*LF = {(x1, x2)*
*LV = {(x1, x2)*
β*; IFV*;
α
*= (0,1)* or *0V = (0,0)* and ≥*FV* is ≥*F* or ≥*V*.
*f(IFV,*α*,*β
*A = maxFV (0FV, minFV{*
*) = minFV({*
definitely true on level
μ*1* and μ
condition is
between
as well.
over *BP*.
where α, β, γ
iDATALOG rules *r;*
as *NTP* in (1) except:
two elements must satisfy the condition
μ
∈
∈
The next question is whether this fixed point is a model of *P*. The fixed point is an interpretation of *P*, which is a model, if for each
$$A \gets A\_1, \dots, A\_n; \underline{\mathcal{B}} \text{ } \underline{\mathcal{I}}\_{FV} \in \operatorname{ground}(\mathcal{P}), \operatorname{I}\_{\mathbb{P}V}(\underline{\mathcal{Q}}, \underline{\mathcal{Y}}) \not\simeq\_{\mathbb{V}V} \underline{\mathcal{B}}.$$
Similarly to the proof of Theorem 1, it can easily be proved that this fixed point is a model of the program. For negation-free iDATALOG this is the least model of the program. (Achs 2010).
In fDATALOG a fact can be negated by completing its membership degree to 1. In iDATALOG the uncertainty level of a negated fact can be computed according to negators. A negator on *LF* or *LV* is a decreasing mapping ordering *0FV* and *1FV* together (Cornelis et al 2004). The applied negators are relevant for the computational meaning of a program, but they have no influence on the stratification. So for a stratified iDATALOG program *P* there is an evaluation sequence in which *lfp(iNTP)* is a unique minimal model of *P*. Therefore *lfp(iNTP)* can be regarded as the semantics of iDATALOG.
After defining the syntax and semantics of extended fuzzy Datalog, it is necessary to examine the properties of possible implication operators and the extended uncertainty-level functions. A number of intuitionistic implications are discussed in (Cornelis et al 2004, Atanassov 2005, 2006) and other papers, four of them are the extensions of the above three fuzzy implication operators. Now these operators will be presented and completed by the suitable interval-value operators and the uncertainty-level functions. The computations will not be shown here, only the starting points and results are presented.
The coordinates of intuitionistic and interval-valued implication operators can be determined by each other. The uncertainty-level functions can be computed according to the applied implication. The connection between *IF* and *IV* and the extended versions of uncertainty-level functions are given below: For
*IV(*α*,*γ*) = (IV1, IV2); IV1=IF1(*α*',*γ*');* α*'=(*α*1,1*α*2); IV2=1-IF2(*α*',*γ*');* γ*'=(*γ*1,1*γ*2). f (IF,* α*,* β *) = (min ({* γ*1 | IF1 (*α*,* γ *)* ≥ β*1 }), max ({* γ*2 | IF2 (*α*,* γ *)* ≤ β*2 })) f (IV,* α*,* β *) = (min ({* γ*1 | IV1 (*α*,* γ *)* ≥ β*1 }), min ({* γ*2 | IV2 (*α*,* γ *)* ≥ β*2 }))*
The studied operators and the related uncertainty-level functions are the following:
### **2.2.1.1 Extension of Kleene-dienes implication**
One possible extension of Kleene-Dienes implication for IFS is:
$$\underline{\mathbf{J}}\_{\mathbb{F}\mathbb{K}}(\underline{\alpha}, \underline{\gamma}) = (\max(\alpha\_{\mathbb{L}}, \gamma\_{\mathbb{I}}), \min(\alpha\_{\mathbb{I}}, \gamma\_{\mathbb{Z}})) $$
The appropriate computed elements are the following:
*IVK(*α*,* γ*) = (max(1*α*2,* γ*1), max(1*α*1,* γ*2))*
From Fuzzy Datalog to Multivalued Knowledge-Base 37
or rather "not in all cases". For example, in the case of the Kleene-Dienes and the Lukasiewicz intuitionistic operators the levels of a rule-head satisfy the condition of intuitionism only if the sum of the levels of the rule-body is at least as large as the sum of the levels of the rule. That is, the solution is inside the scope of IFS, if the level of the rulebody is less "intuitionistic" than the level of the rule. In the case of the first Gödel operator, the solution is inside the scope of IFS only if the level of the rule-body is more certain than the level of the rule (Achs 2010). However, for the second Gödel operator the next
proposition can easily be proven:
α = *(*α*1,* α
*if* α*1+*α*2* ≤ *1, β1+β<sup>2</sup>*
α
**2.2.2 Bipolar extension of datalog**
*if* α*1*≤ α*2, β1*≤
*2), β = (β1, β2)*
≤
*1 then f1(IFG2,*
Similarly to Proposition 2, it can be seen that the fixed-point algorithm terminates if *f(IFV,*
The intuitive background of intuitionistic levels is some psychological perception. Experiments have shown that when making decisions people deal with positive and negative facts in different ways (Dubois et al 2000, 2005). Continuing this idea, it can be stated that there would be differences not only in the scaling of truth values, but in the way of concluding as well. This means that in a way similar to the facts, positive and negative inferences can be separated. The idea of bipolar Datalog is based on the above observation: two kinds of ordinary fuzzy implications are used for positive and negative deduction, namely, a pair of consequence transformations is defined instead of a single one. Since in the original transformations lower bounds are used with degrees of uncertainty, therefore starting from IFS or IVS facts, the resulting degrees will be lower bounds of membership and non-membership respectively, instead of the upper bound for non-membership. However, if each non-membership value μ is transformed into membership value μ' = 1 − μ, then both members of head-level can be deduced similarly. So the appropriate concepts are
**Definition 8.** The bipolar Datalog program (bDATALOG) defined on *LF* or *LV* is a finite set of
The elements of the bipolar nondeterministic consequence transformation *bNTP = (NTP1,*
γ*'2 | I2(*α*'body2,* γ*'2)* ≥ β*'2});*
γ*1 | I1(*α*1,* γ*1)* ≥ β*1});*
*β2 then f1(IVG2,*
∈ *LFV* (Achs 2010). G2 satisfies this condition, so:
**Proposition 5**. In the case of G2 operator the fixed-point algorithm terminates.
α
α*, β)* ≤ *f2(IVG2,* α*, β)*
*, β) + f2(IFG2,*
α*, β)* ≤ *1*
> α*,*
**Proposition 4**. For
for each
β*)* ≤*FV* α
as follows.
safe bDATALOG rules *r; (*
β*1,* β*2); (I1,I2).*
α
The uncertainty-level function is: *f = (f1, f2)* where
where α'body2=min FV2 (α'A12,…,α'An2).
*NTP2)* are similar to *NTP* in (1) except in *NTP2* the level of rule's head is:
*'2 = max FV2 (0, min FV2 {*
*f1 = min FV2 ({*
$$f\_1(\underline{I}\_{FK}, \underline{\underline{\alpha}}, \underline{\beta}) = \begin{cases} 0 & \alpha\_2 \ge \beta\_1 \\ \beta\_1 & \text{otherwise} \end{cases}$$
$$f\_1(\underline{I}\_{VV}, \underline{\underline{\alpha}}, \underline{\underline{\beta}}) = \begin{cases} 0 & 1 - \alpha\_2 \ge \beta\_1 \\ \beta\_1 & \text{otherwise} \end{cases}$$
$$f\_2(\underline{\mathcal{I}}\_{FK}, \underline{\mathcal{Q}}\_{\prime} \underline{\mathcal{B}}) = \begin{cases} 1 & \alpha\_1 \le \beta\_2 \\ \beta\_2 & \text{otherwise} \end{cases}$$
$$f\_2(\underline{\mathcal{I}}\_{VV}, \underline{\mathcal{Q}}\_{\prime} \underline{\mathcal{B}}) = \begin{cases} 0 & 1 \cdot \alpha\_1 \ge \beta\_2 \\ \beta\_2 & \text{otherwise} \end{cases}$$
### **2.2.1.2 Extension of Lukasiewicz implication**
One possible extension of Lukasiewicz implication for IFS is:
$$\underline{\mathbf{I}}\_{\rm FL}(\underline{\mathbf{a}}, \underline{\mathbf{y}}) = (\min(\alpha\_{\mathbf{\tilde{z}}}, \underline{\mathbf{y}}), \min(\alpha\_{\mathbf{l}}, \underline{\mathbf{y}})) $$
The appropriate computed elements are as follows:
$$\begin{array}{rcl} \underline{I}\_{\rm VL}(\underline{\alpha}, \underline{\beta}) &=& (\max(1, \alpha\_{\rm L} + \underline{\gamma}), \max(0, \alpha\_{\rm l} + \underline{\gamma}\_{\rm l} - 1)) \\\\ f\_{\rm l}(\underline{\underline{I}\_{\rm FL}}, \underline{\alpha}, \underline{\beta}) &=& \min(1 - \alpha\_{\rm L}, \max(0, \beta\_{\rm l} - \underline{\alpha})) \\\\ f\_{\rm l}(\underline{\underline{I}\_{\rm FL}}, \underline{\alpha}, \underline{\beta}) &=& \max(1 - \alpha\_{\rm l}, \min(1, \ 1 - \alpha\_{\rm l} + \beta\_{\rm l})) \\\\ f\_{\rm l}(\underline{\underline{I}\_{\rm VL}}, \alpha, \beta) &=& \max(0, \alpha\_{\rm l} + \beta\_{\rm l} - 1) \\\\ f\_{\rm l}(\underline{\underline{I}\_{\rm VL}}, \alpha, \beta) &=& \max(0, \alpha\_{\rm l} + \beta\_{\rm l} - 1) \end{array}$$
### **2.2.1.3 Extensions of Gödel implication**
There are several alternative extensions of Gödel implication, two of them are presented here:
$$\underline{I}\_{\mathbb{FG}1}(\underline{\alpha}, \underline{\gamma}) = \begin{cases} (1,0) & \alpha\_1 \le \gamma\_1 \\ (\gamma\_1, 0) & \alpha\_1 > \gamma\_1, \alpha\_2 \ge \gamma\_2 \\ (\gamma\_1, \gamma\_2)\alpha\_1 & \gamma\_1, \alpha\_2 < \gamma\_2 \end{cases} \qquad \underline{I}\_{\mathbb{FG}2}(\underline{\alpha}, \underline{\gamma}) \quad = \begin{cases} (1,0) & \alpha\_1 \le \gamma\_1, \alpha\_2 \ge \gamma\_2 \\ (\gamma\_1, \gamma\_2) & \text{otherwise} \end{cases}$$
The appropriate computed elements are:
$$\begin{aligned} \underline{I}\_{\text{VGL}}(\underline{\alpha},\underline{\gamma}) &= \begin{cases} (1,1) & \alpha\_{1} \le \gamma\_{1} \\ (\gamma\_{1},1) & \alpha\_{1} > \gamma\_{1}, \alpha\_{2} \ge \gamma\_{2} \\ (\gamma\_{1},\gamma\_{2})\alpha\_{1} & \gamma\_{1}, \alpha\_{2} < \gamma\_{2} \end{cases} & \underline{I}\_{\text{VGL}}(\underline{\alpha},\underline{\gamma}) &= \begin{cases} (1,1) & \alpha\_{1} \le \gamma\_{1}, \alpha\_{2} \le \gamma\_{2} \\ (\gamma\_{1},\gamma\_{2}) & \text{otherwise} \end{cases} \\\\ \underline{f}(\underline{I}\_{\text{FGL}},\underline{\alpha},\underline{\beta}) &= \min(\alpha,\beta\_{1}) & \underline{f}\_{2}\left(\underline{I}\_{\text{FGL}},\underline{\alpha},\underline{\beta}\right) &= \begin{cases} 1 & \alpha\_{1} \le \beta\_{2} \\ \max(\alpha\_{2},\beta\_{2}) & \text{otherwise} \end{cases} \\\\ \underline{f}(\underline{I}\_{\text{VGL}},\underline{\alpha},\underline{\beta}) &= \min(\alpha\_{\text{b}},\beta\_{1}) & \underline{f}\_{2}\left(\underline{I}\_{\text{VGL}},\underline{\alpha},\underline{\beta}\right) &= \begin{cases} 0 & \alpha\_{1} \le \beta\_{2} \\ \min(\alpha\_{2},\beta\_{2}) & \text{otherwise} \end{cases} \\\\ \underline{f}(\underline{I}\_{\text{VGL}},\underline{\alpha},\underline{\beta}) &= \min(\alpha,\beta\_{1}) & \underline{f}\_{2}(\underline{I}\_{\text{VGL}},\underline{\alpha},\underline{\beta}) &= \max(\alpha,\beta\_{2}) \end{cases} \end{aligned}$$
An important question is whether the resulting degrees satisfy the conditions referring to IFS and IVS respectively. Unfortunately, for implications other than G2, the answer is "no", 36 Fuzzy Logic – Algorithms, Techniques and Implementations
α*2,*γ*1), min(*α*1,*γ*2))*
> α*2 +* γ
α
α
α*2+ β1 -1)*
α*1+ β2 -1)*
There are several alternative extensions of Gödel implication, two of them are presented
*FG*
*VG*
*1, β1)* 1 2 2 1
*1, β1)* 1 2 2 1
An important question is whether the resulting degrees satisfy the conditions referring to IFS and IVS respectively. Unfortunately, for implications other than G2, the answer is "no",
2
α γ
> α β
α β
α
α
2
α γ
*) = (max(1,*
*, β) = min(1-*
*, β) = max(1-*
*, β) = max(0,*
*, β) = max(0,*
2
2
*1), max(0,*
α*2))*
α*1+ β2))*
*2, max(0, β1-*
*1, min(1, 1-*
α*1 +* γ*2 -1))*
*(,) I( )* 1 12 2
1 0 ,
≤ ≥
γ α
γ α γ
γ
α
1 1 ,
2 2
β
2 2
α β
α β
α*2, β2)*
α*2, β2)*
β
≤ ≤
1 2
*(,) I( )* 1 12 2
, = ( , ) otherwise α
1 2
<sup>≤</sup>
<sup>≤</sup>
γ γ
1 , , = max( , ) otherwise *FG f (I )* α
0 , , = min( , ) otherwise *VG f (I )* α
*, β) = max(*
*, β) = min(*
γ γ
, = ( , ) otherwise
*f I*
*- f I*
α β
α β 1 2
α β
1 2
β
2 1 ( , , )= otherwise *FK*
2
β
β
α
01 ( , , )= otherwise *VK*
<sup>≥</sup>
<sup>≤</sup>
2 1
α
One possible extension of Lukasiewicz implication for IFS is:
*IFL(*α*,*γ*) = (min(*
α*,* γ
α
α
α
α
γ
γ
γ
γ
*1, β1) f2(IFG2,*
*1, β1) f2(IVG2,*
β
2 1
β
1 0 ( , , )= otherwise *FK*
1
The appropriate computed elements are as follows:
( , , )= otherwise *VK*
**2.2.1.2 Extension of Lukasiewicz implication**
β
β
α
01
− ≥
<sup>≥</sup>
1
1
*f I*
*f I*
α β
*IVL(*
*f1(IFL,*
*f2(IFL,*
*f1(IVL,*
*f2(IVL,*
here:
*FG*
*VG*
*f1(IFG1,* α
*f1(IVG1,* α
*f1(IFG2,* α
*f1(IVG2,* α
*I( )*
α γ
*I( )*
α γ
**2.2.1.3 Extensions of Gödel implication**
1 1 1 12 2
The appropriate computed elements are:
1 1 1 12 2
<sup>≤</sup>
α
α
α
α
(,) ,
> ≥ > <
α
α
*(,)*
γ
γ γ α
*, β) = min(*
*, β) = min(*
*, β) = min(*
*, β) = min(*
11 , = ( ,1) ,
<sup>≤</sup>
α
> ≥ > <
α
10 , = ( ,0) , (,) ,
*(,)*
γ
γ γ α 1 1
γ
γ α
12 1 1 2 2
γ α
1 1
γ
γ α
γ α
12 1 1 2 2
α β or rather "not in all cases". For example, in the case of the Kleene-Dienes and the Lukasiewicz intuitionistic operators the levels of a rule-head satisfy the condition of intuitionism only if the sum of the levels of the rule-body is at least as large as the sum of the levels of the rule. That is, the solution is inside the scope of IFS, if the level of the rulebody is less "intuitionistic" than the level of the rule. In the case of the first Gödel operator, the solution is inside the scope of IFS only if the level of the rule-body is more certain than the level of the rule (Achs 2010). However, for the second Gödel operator the next proposition can easily be proven:
**Proposition 4**. For α = *(*α*1,* α*2), β = (β1, β2)*
$$\begin{array}{l} \text{if } \alpha\_1 + \alpha\_2 \le 1, \ f\_1 + \beta\_2 \le 1 \text{ then } f\_1(\underline{\mathcal{L}}\_{\text{FG}\_{\mathcal{D}}} \underline{\mathcal{Q}}, \underline{\mathcal{Q}}) + f\_2(\underline{\mathcal{L}}\_{\text{FG}\_{\mathcal{D}}} \underline{\mathcal{Q}}, \underline{\mathcal{Q}}) \le 1 \\\\ \text{if } \alpha\_1 \le \alpha\_2, \ f\_1 \le \beta\_2 \text{ then } f\_1(\underline{\mathcal{L}}\_{\text{VG}\mathcal{D}} \underline{\mathcal{Q}}, \underline{\mathcal{Q}}) \le f\_2(\underline{\mathcal{L}}\_{\text{VG}\mathcal{D}}, \underline{\mathcal{Q}}, \underline{\mathcal{Q}}) \end{array}$$
Similarly to Proposition 2, it can be seen that the fixed-point algorithm terminates if *f(IFV,* α*,* β*)* ≤*FV* α for each α∈ *LFV* (Achs 2010). G2 satisfies this condition, so:
**Proposition 5**. In the case of G2 operator the fixed-point algorithm terminates.
### **2.2.2 Bipolar extension of datalog**
The intuitive background of intuitionistic levels is some psychological perception. Experiments have shown that when making decisions people deal with positive and negative facts in different ways (Dubois et al 2000, 2005). Continuing this idea, it can be stated that there would be differences not only in the scaling of truth values, but in the way of concluding as well. This means that in a way similar to the facts, positive and negative inferences can be separated. The idea of bipolar Datalog is based on the above observation: two kinds of ordinary fuzzy implications are used for positive and negative deduction, namely, a pair of consequence transformations is defined instead of a single one. Since in the original transformations lower bounds are used with degrees of uncertainty, therefore starting from IFS or IVS facts, the resulting degrees will be lower bounds of membership and non-membership respectively, instead of the upper bound for non-membership. However, if each non-membership value μ is transformed into membership value μ' = 1 − μ, then both members of head-level can be deduced similarly. So the appropriate concepts are as follows.
**Definition 8.** The bipolar Datalog program (bDATALOG) defined on *LF* or *LV* is a finite set of safe bDATALOG rules *r; (*β*1,* β*2); (I1,I2).*
The elements of the bipolar nondeterministic consequence transformation *bNTP = (NTP1, NTP2)* are similar to *NTP* in (1) except in *NTP2* the level of rule's head is:
> α*'2 = max FV2 (0, min FV2 {*γ*'2 | I2(*α*'body2,* γ*'2)* ≥ β*'2});*
where α'body2=min FV2 (α'A12,…,α'An2).
The uncertainty-level function is: *f = (f1, f2)* where
*f1 = min FV2 ({*γ*1 | I1(*α*1,* γ*1)* ≥ β*1});*
From Fuzzy Datalog to Multivalued Knowledge-Base 39
*{(p(a), (0.7, 0.2)), (q(b), (0.65, 0.3)), (r(a,b), (0.7, 0.2)) }.*
*q(x, y)* ← *p(x, y); (0.85, 0.95); I1. q(x, y)* ← *p(x, z), q(z, y); (0.8, 0.9); I2.*
*1, β1) f2(IVG2,*
*2+ β1 -1) f2(IVL,*
α
X*0* = {*(p(a, b), (0.7, 0.8)), (p(a, c), (0.8, 0.9)), (p(b, d), (0.75, 0.8)), (p(d, e), (0.9, 0.95))*} X*1* =X*0* ∪ { *(q(a, b), (0.7, 0.8)), (q(a, c), (0.8, 0.9)), (q(b, d), (0.75, 0.8)), (q(d, e), (0.85, 0.95))*,
As fuzzy Datalog is a special kind of its multivalued extensions, so further on both fDATALOG and any of above extensions will be called multivalued Datalog (mDATALOG).
The facts of an mDATALOG program can be regarded as any kind of lexical knowledge including uncertainty as well, and from this knowledge other facts can be deduced according to the rules. Therefore a multivalued Datalog program is suitable to be the deduction mechanism of a knowledge base. Sometimes, however, it is not enough for getting answer to a question. For example, if there are rules describing the options of loving a good composer, and there is a fact declaring that Vivaldi is a good composer, what is the
Before showing the fixed point algorithm, two computations are set out. According to the
According to the second rule, the uncertainty of *q(a, d)* can be computed in this way:
α
*(q(a, d), (0.6, 0.6), (q(b, e), (0.6, 0.65))* }
α
α
α
*, β) = min(*
*, β) = max(0,*
*, β) = (min(*
*1+ β2 -1) = (max(0, 0.8 + 0.8-1), max(0, 0.7 + 0.9 – 1)) =*
α*2, β2)*
> α*1+ β2 -1)*
α
*body = min((0.7, 0.8),(0.75, 0.8)) =(0.7, 0.8);*
*1, β1), min(*
α*2, β2)) =*
Considering the other level of r*(a,b)*, its resulting level is *(max(0.4, 0.7), 1*
*(p(a, b), (0.7, 0.8)). (p(a, c), (0.8, 0.9)). (p(b, d), (0.75, 0.8)). (p(d, e), (0.9, 0.95)).*
Let *I1 = IVG2*, *I2 = IVL* , that is the appropriate uncertainty level functions are:
α
α
**Example 7.** Consider the next (recursive) IVS program:
α
α
*(min(0.7, 0.85), min(0.8, 0.95)) = (0.7, 0.8).*
The body of the rule is: *p(a, b), q(b, d), so*
α
X*2* =X*1* ∪ { *(q(a, e), (0.45, 0.5))* }
X*2* is fixed point, so it is the result of the program.
The steps of fixed point algorithm are:
**3. Multivalued knowledge-base**
*, β) = min(*
*, β) = max(0,*
first rule for *q(x, y)* the uncertainty level of *q(a, b)* is: *f(IVG2,*
*2+ β1 -1), max(0,*
*β2) = 1 – min(0.7, 0.8) = 0.3*. So the uncertainty level of rule's head is (0.4, 0.3).
−*max(1*−*0.3, 1*−*0.2)) =*
*min(*α*'2,1*−
*(0.7, 0.2)*, so the fixed point is:
*f1(IVG2,*
*f1(IVL,*
*, β) =( max(0,*
*f(IVL,* α
*(0.6, 0.6).*
$$f\_2 = 1 - \min\_{FV2} \left( \{ 1 - \underline{\gamma}\_2 \mid I\_2(1 - \alpha\_2, 1 - \underline{\gamma}\_2) \ge 1 - \underline{\beta}\_2 \} \right).$$
It is evident that applying the transformation μ'1 = μ1, μ'2 = 1 - μ2, for all IFS levels of the program, the above definition can be applied to IVS degrees as well. As a simple computation can show, contrary to the results of iDATALOG, the resulting degrees for most variants of bipolar Datalog satisfy the conditions referring to IFS:
**Proposition 5.** For α = (α1, α2), β = (β1, β2) and for implication-pairs I = (IG, IG); I = (IL, IL); I = (IL, IG); I = (IK, IK); I = (IL, IK);
$$\text{if } \alpha\_1 + \alpha\_2 \le 1, \beta\_1 + \beta\_2 \le 1 \text{ then } f\_1(\mathsf{l}\_1, \underline{\alpha}, \underline{\beta}) + f\_2(\mathsf{l}\_2, \underline{\alpha}, \underline{\beta}) \le 1.$$
From the construction of bipolar consequence transformations follows:
**Proposition 6.** The nondeterministic bipolar consequence transformation has a least fixed point, which is a model of program *P* in the following sense: for each *A*←*A1,...,An;* β*; I* ∈ *ground(P)*
$$I(\alpha\_{\text{body1}}, \alpha\_1) \ge \beta\_1 \colon I(\alpha'\_{\text{body2}}, \alpha'\_2) \ge \beta'\_2$$
As the termination of the consequence transformations based on these three implication operators was proven in the case of fDATALOG (Achs 2006) and since this property does not change in bipolar case, the bipolar consequence transformations terminate as well.
The bipolar extension of Datalog has no influence on the stratification, so the propositions detailed in the case of stratified fDATALOG programs are true in the case of bipolar fuzzy Datalog programs as well, that is, for a stratified bDATALOG program *P*, there is an evaluation sequence, in which *lfp(bNTP )* is a unique minimal model of *P*.
**Example 6.** Consider the next IFS program:
$$\begin{array}{l} (p(a), (0.7, 0.2)). \\ (q(b), (0.65, 0.3)). \\ (r(a, b), (0.7, 0.2)). \\ r(x, y) \leftarrow p(x), q(y); (0.75, 0.2); L. \end{array}$$
Let I = IFG2, then according to the rule r(a,b) is inferred and uncertainty can be computed as follows: αbody = minF((0.7, 0.2), (0.65, 0.3)) = (0.65, 0.3), f1(IFG2, α, β) = min(α1, β1) = min(0.65, 0.75) = 0.65, f2(IFG2, α, β) = max(α2, β2) = 0.3, that is, the level of the rule's head is (0.65, 0.3). There is a fact for r(a,b) as well, so the resulting level is the union of the level of the rule's head and the level of the fact: maxF ((0.65, 0.3), (0.7, 0.2)) = (0.7, 0.2). So the fixed point of the program is:
$$\{(p(a), (0.7, 0.2)), (q(b), (0.65, 0.3)), (r(a, b), (0.7, 0.2))\}.$$
Let us see the bipolar evaluation of the program. Let *I = (IL, IG)*. That is let the first element of the uncertainty level be computed according to the Lukasiewicz operator and the second one according to the Gödel operator. The Lukasiewicz operator defines the uncertainty level function *f (IL,* α*,* β*) = max(0,* α *+* β −*1)*. Then α*body1 = min(0.7, 0.65) = 0.65,* α*'body2 = min(1*−*0.3, 1*−*0.2) = 0.7*; *f1(IL,* α*1, β1) = max(0,* α*1+β1*−*1) = 0.65+0.75*−*1 = 0.4*, *f2(IG,* α*'2, β'2) = 1* −
38 Fuzzy Logic – Algorithms, Techniques and Implementations
It is evident that applying the transformation μ'1 = μ1, μ'2 = 1 - μ2, for all IFS levels of the program, the above definition can be applied to IVS degrees as well. As a simple computation can show, contrary to the results of iDATALOG, the resulting degrees for most
**Proposition 5.** For α = (α1, α2), β = (β1, β2) and for implication-pairs I = (IG, IG); I = (IL, IL);
*1 then f1(I1,*
**Proposition 6.** The nondeterministic bipolar consequence transformation has a least fixed point, which is a model of program *P* in the following sense: for each *A*←*A1,...,An;*
As the termination of the consequence transformations based on these three implication operators was proven in the case of fDATALOG (Achs 2006) and since this property does not change in bipolar case, the bipolar consequence transformations terminate as well.
The bipolar extension of Datalog has no influence on the stratification, so the propositions detailed in the case of stratified fDATALOG programs are true in the case of bipolar fuzzy Datalog programs as well, that is, for a stratified bDATALOG program *P*, there is an
*r(x, y)* ← *p(x), q(y); (0.75, 0.2); I.*
Let I = IFG2, then according to the rule r(a,b) is inferred and uncertainty can be computed as follows: αbody = minF((0.7, 0.2), (0.65, 0.3)) = (0.65, 0.3), f1(IFG2, α, β) = min(α1, β1) = min(0.65, 0.75) = 0.65, f2(IFG2, α, β) = max(α2, β2) = 0.3, that is, the level of the rule's head is (0.65, 0.3). There is a fact for r(a,b) as well, so the resulting level is the union of the level of the rule's head and the level of the fact: maxF ((0.65, 0.3), (0.7, 0.2)) = (0.7, 0.2). So the fixed point of the
*{(p(a), (0.7, 0.2)), (q(b), (0.65, 0.3)), (r(a, b), (0.7, 0.2)) }.* Let us see the bipolar evaluation of the program. Let *I = (IL, IG)*. That is let the first element of the uncertainty level be computed according to the Lukasiewicz operator and the second one according to the Gödel operator. The Lukasiewicz operator defines the uncertainty level
α
*1) = 0.65+0.75*
*body1 = min(0.7, 0.65) = 0.65,*
−
α
α
*1 = 0.4*, *f2(IG,*
*'body2 = min(1*
*'2, β'2) = 1*
−*0.3,*
−
*1)*. Then
α*1+β1*− −α*2, 1* −γ*2)* ≥ *1* −β*2}).*
α
*, β) + f2(I2,*
α*, β)* ≤ *1*
> β*; I* ∈
*f2 =1*−
I = (IL, IG); I = (IK, IK); I = (IL, IK);
*ground(P)*
program is:
function *f (IL,*
*0.2) = 0.7*; *f1(IL,*
*1*− α*,* β
*) = max(0,*
α
α *+* β −
*1, β1) = max(0,*
if α*1+*α*2* ≤ *1, β1+β<sup>2</sup>*
**Example 6.** Consider the next IFS program:
*min FV2 ({1*
variants of bipolar Datalog satisfy the conditions referring to IFS:
− γ*2 | I2(1*
≤
From the construction of bipolar consequence transformations follows:
evaluation sequence, in which *lfp(bNTP )* is a unique minimal model of *P*.
*(p(a), (0.7, 0.2)). (q(b), (0.65, 0.3)). (r(a, b), (0.7, 0.2)).*
*I(*α*body1,* α*1)* ≥ β*1; I(*α*'body2,* α*'2)* ≥ β*'2* *min(*α*'2,1*−*β2) = 1 – min(0.7, 0.8) = 0.3*. So the uncertainty level of rule's head is (0.4, 0.3). Considering the other level of r*(a,b)*, its resulting level is *(max(0.4, 0.7), 1*−*max(1*−*0.3, 1*−*0.2)) = (0.7, 0.2)*, so the fixed point is:
*{(p(a), (0.7, 0.2)), (q(b), (0.65, 0.3)), (r(a,b), (0.7, 0.2)) }.*
**Example 7.** Consider the next (recursive) IVS program:
*(p(a, b), (0.7, 0.8)). (p(a, c), (0.8, 0.9)). (p(b, d), (0.75, 0.8)). (p(d, e), (0.9, 0.95)). q(x, y)* ← *p(x, y); (0.85, 0.95); I1. q(x, y)* ← *p(x, z), q(z, y); (0.8, 0.9); I2.*
Let *I1 = IVG2*, *I2 = IVL* , that is the appropriate uncertainty level functions are:
*f1(IVG2,* α*, β) = min(*α*1, β1) f2(IVG2,* α*, β) = min(*α*2, β2) f1(IVL,* α*, β) = max(0,* α*2+ β1 -1) f2(IVL,* α*, β) = max(0,* α*1+ β2 -1)*
Before showing the fixed point algorithm, two computations are set out. According to the first rule for *q(x, y)* the uncertainty level of *q(a, b)* is: *f(IVG2,* α*, β) = (min(*α*1, β1), min(*α*2, β2)) = (min(0.7, 0.85), min(0.8, 0.95)) = (0.7, 0.8).*
According to the second rule, the uncertainty of *q(a, d)* can be computed in this way:
The body of the rule is: *p(a, b), q(b, d), so* α*body = min((0.7, 0.8),(0.75, 0.8)) =(0.7, 0.8); f(IVL,* α*, β) =( max(0,* α*2+ β1 -1), max(0,* α*1+ β2 -1) = (max(0, 0.8 + 0.8-1), max(0, 0.7 + 0.9 – 1)) = (0.6, 0.6).*
The steps of fixed point algorithm are:
X*0* = {*(p(a, b), (0.7, 0.8)), (p(a, c), (0.8, 0.9)), (p(b, d), (0.75, 0.8)), (p(d, e), (0.9, 0.95))*} X*1* =X*0* ∪ { *(q(a, b), (0.7, 0.8)), (q(a, c), (0.8, 0.9)), (q(b, d), (0.75, 0.8)), (q(d, e), (0.85, 0.95))*, *(q(a, d), (0.6, 0.6), (q(b, e), (0.6, 0.65))* } X*2* =X*1* ∪ { *(q(a, e), (0.45, 0.5))* }
X*2* is fixed point, so it is the result of the program.
As fuzzy Datalog is a special kind of its multivalued extensions, so further on both fDATALOG and any of above extensions will be called multivalued Datalog (mDATALOG).
### **3. Multivalued knowledge-base**
The facts of an mDATALOG program can be regarded as any kind of lexical knowledge including uncertainty as well, and from this knowledge other facts can be deduced according to the rules. Therefore a multivalued Datalog program is suitable to be the deduction mechanism of a knowledge base. Sometimes, however, it is not enough for getting answer to a question. For example, if there are rules describing the options of loving a good composer, and there is a fact declaring that Vivaldi is a good composer, what is the
From Fuzzy Datalog to Multivalued Knowledge-Base 41
Ann loves the music of Bach very much (*(love(Ann, Bach), (0.9, 0.95))*) and the concept of love is similar to the concept of like (*RSVS ("love", "like") = (0.8, 0.9)*) and the music of Bach is more or less similar to the music of Vivaldi (*RGVG (Bach, Vivaldi) = (0.7, 0.75)*) then how strongly can be stated that Ann likes Vivaldi, that is, what is the uncertainty of the predicate
To solve this problem, the concept of proximity-based uncertainty function will be introduced. According to this function, the uncertainty levels of "synonyms" can be computed from the levels of original fact and from the proximity values of actual predicates and its arguments. It is expectable that in the case of identity, the level must be unchanged, but in other cases it is should be equal or less than the original level or than the proximity values. Furthermore, this function is required to increase monotonically. This function will
Let *p* be a predicate symbol with *n* arguments, then *p/n* is called the functor of the atom,
*, 1FV, 1FV,…, 1FV) =*
Any triangular norm obeys the above constraints so they are appropriate proximity-based
**Example 8**. Let *(p(a), (0.7, 0.2))* be an *IFS* fact and *RSFS(p, q) = (0.8, 0.1), RGFG(a, b) = (0.7,*
**Example 9**. Let *(love(Ann, Bach), (0.9, 0.95)* be an *IVS fact* and *RSVS("love", "like") = (0.8, 0.9),*
⋅
⋅
⋅
*, λ, λ1, λ2) = minV(*
*0.7), min(0.95, 1, 1*
*1), min(0.95, 0.9, 1*
*0.7), min(0.95, 0.9, 1*
→
α
*[0FV, 1FV]*
*, λ, λ1,…, λn)*
*(0.7, 0.3))) = (min(0.7, 0.7), max(0.2, 0.3)) = (0.7, 0.3));*
*(1, 0))) = (min(0.7, 0.8), max(0.2, 0.1)) = (0.7, 0.2));*
*(0.7, 0.3))) = (min(0.7, 0.56), max(0.2, 0.37)) = (0.56, 0.37));*
α*, λ, λ1*⋅*λ2).* Then
⋅
⋅
⋅
*0.75) = (0.7, 0.75));*
*1) = (0.8, 0.9));*
*0.75) = (0.7, 0.75));*
α
*, λ, λ1,…, λn) : (0FV, 1FV]n+2*
*, λ, λ1,…, λn) ≤ min FV (*
*like(Ann, Vivaldi)*?
where
and ϕ*p(*α
*0.3)* and
uncertainty functions.
ϕ*p(*α
be ordered to each atom of a program.
characterized by this predicate symbol.
*, λ, λ1) =* minF*(*
*(p(b), (min((0.7, 0.2), ((1, 0)*
*(q(b), (min((0.7, 0.2), ((0.8, 0.1)*
*RGVG (Bach, Vivaldi) = (0.7, 0.75))* and
*(q(a), (min((0.7, 0.2), ((0.8, 0.1)*
The uncertainty levels of *p(b)*, *q(a)* and *q(b)* are:
(In IFS the product is defined as:
**Definition 12.** A proximity-based uncertainty function of *p/n* is:
ϕ*p(*α
μ ⋅ *λ* = *(*μ*1*⋅*λ1, 1-(1*μ*2)*⋅*(1- λ2))*.)
⋅
⋅
*(love(Ann, Vivaldi), (min(0.9, 1, 1*
*(like(Ann, Vivaldi), (min(0.9, 0.8, 1*
*(like(Ann, Bach), (min(0.9, 0.8, 1*
⋅
ϕ*love(*α
*, λ, λ1,…, λn)* is monotonically increasing in each argument.
ϕ*p(*α
> ϕ*p(*α
> > α*, λ*⋅*λ1).*
possible answer to the question inquiring about liking Bach? Getting an answer needs the use of synonyms and similarities. For handling this kind of information, our model includes a background knowledge module.
### **3.1 Background knowledge**
Some "synonyms" and "similarities" will be defined between the potential predicates and between the potential constants of a given problem, so it can be examined in a larger context. More precisely, proximity relations will be defined on the sets of the program's predicates and terms. These relations will serve as the basis for the background knowledge.
**Definition 9.** A multivalued proximity on a domain *D* is an IFS or IVS valued relation *RFVD* : *D* × *D* → [*0FV, 1FV*] which satisfies the following properties:
A proximity is similarity if it is transitive, that is
$$\underline{R}\_{FV\_D}(\mathbf{x}, z) \succeq\_{FV} \min\_{FV} \left( \underline{R}\_{FV\_D}(\mathbf{x}, y), \underline{R}\_{FV\_D}(y, \mathbf{x}) \right) \text{ \textquotedbl{}\forall x, y, z \in D.}$$
In the case of similarity, equivalence classifications can be defined over *D.* The effect of this classification is perhaps a simpler or more effective algorithm, but in many cases the requirement of similarity is a too strict constraint. Therefore this chapter deals only with the more general proximity.
Background knowledge consists of the "synonyms" of each terms and each predicates of the program. The "synonyms" of any element form the proximity set of the element, and all of the proximity sets compose the background knowledge. More precisely:
**Definition 10.** Let *d* ∈ *D* any element of domain *D*. The proximity set of *d* is an IFS or IVS subset over *D*:
$$\mathcal{R}\_{FV\_d} = \{ (d\_{1\prime} \,\underline{\lambda}\_{FV\_1}), \, (d\_{2\prime} \,\underline{\lambda}\_{FV\_2}), \dots, \, (d\_{n\prime} \,\underline{\lambda}\_{FV\_n}) \},$$
where *di* ∈ *D and λFVi = RFVD (d, di)* for *i = 1, … ,n.*
**Definition 11.** Let *G* be any set of ground terms and *S* any set of predicate symbols. Let *RGFVG* and *RSFVS* be any proximity over *G* and *S* respectively. The background knowledge is the set of proximity sets:
$$Bk = \{ \mathcal{R}\mathcal{G}\_{\mathbb{F}V\_{\mathcal{S}}} \mid \mathcal{g} \in \mathcal{G} \} \cup \{ \mathcal{R}\mathcal{S}\_{\mathbb{F}V\_{\mathcal{S}}} \mid s \in S \}$$
### **3.2 Computing uncertainties**
Up to now, the deduction mechanism and the background knowledge of a multivalued knowledge-base have been defined. Now, the question remains: how can the two parts be connected to each other? How can we find the "synonyms"? For example, if it is known that Ann loves the music of Bach very much (*(love(Ann, Bach), (0.9, 0.95))*) and the concept of love is similar to the concept of like (*RSVS ("love", "like") = (0.8, 0.9)*) and the music of Bach is more or less similar to the music of Vivaldi (*RGVG (Bach, Vivaldi) = (0.7, 0.75)*) then how strongly can be stated that Ann likes Vivaldi, that is, what is the uncertainty of the predicate *like(Ann, Vivaldi)*?
To solve this problem, the concept of proximity-based uncertainty function will be introduced. According to this function, the uncertainty levels of "synonyms" can be computed from the levels of original fact and from the proximity values of actual predicates and its arguments. It is expectable that in the case of identity, the level must be unchanged, but in other cases it is should be equal or less than the original level or than the proximity values. Furthermore, this function is required to increase monotonically. This function will be ordered to each atom of a program.
Let *p* be a predicate symbol with *n* arguments, then *p/n* is called the functor of the atom, characterized by this predicate symbol.
**Definition 12.** A proximity-based uncertainty function of *p/n* is:
$$(\underline{\mathfrak{g}}\_{\underline{\mathbb{P}}}(\underline{\alpha}\,\,\underline{\lambda}\,\,\underline{\lambda}\_{1}\,\,\underline{\lambda}\_{1},\ldots,\,\underline{\lambda}\_{n}) : (\underline{\Omega}\_{FV}\,\,\underline{\mathbf{1}}\_{FV}\,\|\_{\mathbb{P}^{\mathsf{T}}}\,\, \rightarrow \|\underline{\mathbf{Q}}\_{FV}\,\,\underline{\mathbf{1}}\_{FV}\|\_{\mathbb{P}})$$
where
40 Fuzzy Logic – Algorithms, Techniques and Implementations
possible answer to the question inquiring about liking Bach? Getting an answer needs the use of synonyms and similarities. For handling this kind of information, our model includes
Some "synonyms" and "similarities" will be defined between the potential predicates and between the potential constants of a given problem, so it can be examined in a larger context. More precisely, proximity relations will be defined on the sets of the program's predicates and terms. These relations will serve as the basis for the background knowledge. **Definition 9.** A multivalued proximity on a domain *D* is an IFS or IVS valued relation *RFVD* :
> *RFD (x, y) = λF(x, y) = (λ1, λ2), λ1 + λ<sup>2</sup> ≤ 1 RVD (x, y) = λV(x, y) =(λ1, λ2), 0 ≤ λ<sup>1</sup> ≤ λ<sup>2</sup> ≤ 1*
> > ∀*x* ∈
*FV min FV (RFVD (x, y), RFVD (y, x))*
In the case of similarity, equivalence classifications can be defined over *D.* The effect of this classification is perhaps a simpler or more effective algorithm, but in many cases the requirement of similarity is a too strict constraint. Therefore this chapter deals only with the
Background knowledge consists of the "synonyms" of each terms and each predicates of the program. The "synonyms" of any element form the proximity set of the element, and all of
**Definition 10.** Let *d* ∈ *D* any element of domain *D*. The proximity set of *d* is an IFS or IVS
*FVd = {(d1, λFV1 ), (d2, λFV2 ),… (dn, λFVn )},*
**Definition 11.** Let *G* be any set of ground terms and *S* any set of predicate symbols. Let *RGFVG* and *RSFVS* be any proximity over *G* and *S* respectively. The background knowledge is
Up to now, the deduction mechanism and the background knowledge of a multivalued knowledge-base have been defined. Now, the question remains: how can the two parts be connected to each other? How can we find the "synonyms"? For example, if it is known that
*D* (reflexivity)
*D* (symmetry).
∀*x, y, z* ∈ *D.*
∀*x* ∈
*D* → [*0FV, 1FV*] which satisfies the following properties:
*RFVD (x, y)= 1FV*
≥
R
*Bk = {*
A proximity is similarity if it is transitive, that is
*RFVD (x, z)*
more general proximity.
subset over *D*:
∈
the set of proximity sets:
*D and λFVi*
**3.2 Computing uncertainties**
where *di*
*RFVD (x, y)= RFVD (y, x)*
the proximity sets compose the background knowledge. More precisely:
*= RFVD (d, di)* for *i = 1, … ,n.*
RG
*FVg | g* ∈ *G}* ∪ *{*RS*FVs | s* ∈ *S}*
a background knowledge module.
**3.1 Background knowledge**
*D* ×
$$\mathfrak{gl}\_{\mathcal{V}}(\underline{\alpha}\,\,\underline{\lambda}\,\,\underline{\lambda}\_1,\dots,\,\lambda\_n) \le \min\,\,\_{FV}(\underline{\alpha}\,\,\underline{\lambda}\,\,\underline{\lambda}\_1\dots,\,\underline{\lambda}\_n)$$
ϕ*p(*α*, 1FV, 1FV,…, 1FV) =* α
and ϕ*p(*α*, λ, λ1,…, λn)* is monotonically increasing in each argument.
Any triangular norm obeys the above constraints so they are appropriate proximity-based uncertainty functions.
**Example 8**. Let *(p(a), (0.7, 0.2))* be an *IFS* fact and *RSFS(p, q) = (0.8, 0.1), RGFG(a, b) = (0.7, 0.3)* and ϕ*p(*α*, λ, λ1) =* minF*(*α*, λ*⋅*λ1).*
(In IFS the product is defined as: μ ⋅ *λ* = *(*μ*1*⋅*λ1, 1-(1*μ*2)*⋅*(1- λ2))*.)
The uncertainty levels of *p(b)*, *q(a)* and *q(b)* are:
*(p(b), (min((0.7, 0.2), ((1, 0)*⋅*(0.7, 0.3))) = (min(0.7, 0.7), max(0.2, 0.3)) = (0.7, 0.3));*
*(q(a), (min((0.7, 0.2), ((0.8, 0.1)*⋅*(1, 0))) = (min(0.7, 0.8), max(0.2, 0.1)) = (0.7, 0.2));*
*(q(b), (min((0.7, 0.2), ((0.8, 0.1)*⋅*(0.7, 0.3))) = (min(0.7, 0.56), max(0.2, 0.37)) = (0.56, 0.37));*
**Example 9**. Let *(love(Ann, Bach), (0.9, 0.95)* be an *IVS fact* and *RSVS("love", "like") = (0.8, 0.9), RGVG (Bach, Vivaldi) = (0.7, 0.75))* and ϕ*love(*α*, λ, λ1, λ2) = minV(*α*, λ, λ1*⋅*λ2).* Then
> *(love(Ann, Vivaldi), (min(0.9, 1, 1*⋅*0.7), min(0.95, 1, 1*⋅*0.75) = (0.7, 0.75));*
*(like(Ann, Bach), (min(0.9, 0.8, 1*⋅*1), min(0.95, 0.9, 1*⋅*1) = (0.8, 0.9));*
*(like(Ann, Vivaldi), (min(0.9, 0.8, 1*⋅*0.7), min(0.95, 0.9, 1*⋅*0.75) = (0.7, 0.75));*
From Fuzzy Datalog to Multivalued Knowledge-Base 43
The modifying algorithm is irrelevant to the evaluation sequence, so stratification can be applied with the same condition. That is, the modified consequence transformation has a least fixed point in the case of stratified programs as well. This transformation makes connections between an mDATALOG program, the background knowledge and the decoding-set of the program. So these four components can form a knowledge-base. However, there should be other transformations connecting the three other parts with each other, therefore the universal concept of a multivalued knowledge-base can be defined with
Φ
decoding-set of *P* and *dA* is any deduction algorithm connecting the three other part with each other. The least fixed point of the deduction algorithm is called the consequence of the
Because the actual deduction algorithm is the modified consequence transformation, now
Note: If it is important to underline that there are bound predicates in P, then *mKB* can be
Φ
**Example 10**. Let us suppose that an internet agent's job is to send a message to its clients if the cinema (*C*) presents a film, which its clients like. The agent knows that people generally go (*go*) to the cinema if they can pay (*cp*) for the ticket and are interested in (*in*) the film presented (*pr*) in the cinema. It also knows that people usually want to see (*ws*) a film if they like (*li*) its actor (*ac*). Moreover it knows that Paul (P) has enough money (*hm*) and he enjoys (*en*) Chaplin (*Ch*) very much. In the cinema, a film of Stan and Pan (*SP*) is presented. Should
This situation can be modelled, for example, by the following multivalued knowledge-base.
Let the IVS valued mDATALOG program and the background knowledge be as follows
*P, dA) = lfp(dA(P, Bk,*
*P, dA) = lfp(modNTP).*
where *P* is a multivalued Datalog program, *Bk* is a background knowledge,
*P, dA(P, Bk,*
Φ*P)),*
> Φ*P)).*
*P)),* where *Bp* is the set of bound predicates.
*pr(C, f), in(x, f), cp(x); (0.85, 0.95); IVL. (R1)*
*ac(f, y), li(x, y); (0.8, 0.85); IVG2. (R2)*
Φ*<sup>P</sup>* is a
**Definition 15.** A multivalued knowledge-base (*mKB*) is a quadruple
*C(Bk, P,*
*C(Bk, P,*
Φ
*go(x, C)*
*(hm(P), (0.75, 0.8)). (en(P, Ch), (0.9, 0.95)).*
*(pr(C, Film), (1,1)). (ac(Film, SP), (1,1)).*
*ws(f, x)*
*mKB = (P, Bk,*
Φ
Φ
the agent inform Paul about this film? How much will he want to go to the cinema?
←
←
According to their roll, *pr(C, Film)* and *ac(Film, SP)* have no alternatives.
*P, dA(P, Bp, Bk,*
an arbitrary deduction algorithm:
knowledge-base, denoted by
denoted by *mKB = (P, Bp, Bk,*
the consequence is
As the above examples show, the levels of "synonyms" can be computed according to proximity-based uncertainty functions. To determine all direct or indirect conclusions of the facts and rules of a program, a proximity based uncertainty function has to be ordered to each predicate of the program. The set of these functions will be the decoding-set of the program.
**Definition 13.** Let *P* be a multivalued Datalog program, and *FP* be the set of the program's functors. The decoding-set of *P* is: Φ*P =* {ϕ*p(*α*, λ, λ1,…, λn) |* ∀ *p/n* ∈ *FP* }*.*
### **3.3 Deduction with background knowledge**
The original deducing mechanism makes conclusions according to the rules of the program, but from now on the background knowledge must be considered as well. So the original mechanism has to be modified. This modified deduction consists of two alternating parts: starting from the facts, their "synonyms" are determined, then applying the suitable rules, other facts are derived, followed by their "synonyms" determined, and again the rules are applied, etc. To define it in a precise manner the concept of modified consecution transformation will be introduced.
The consequence transformation of a mDATALOG *P* program is defined over the set of all multivalued sets of *P*'s Herbrand base, that is, over *FV(BP)*. To define the modified transformation's domain, let us extend *P*'s Herbrand universe with all possible ground terms of the background knowledge, obtaining the so called modified Herbrand universe, *modHP*. The modified Herbrand base, *modBP* is the set of all ground atoms, whose predicate symbols occur in *P* ∪ *Bk* and whose arguments are elements of *modHP*.
However, it is possible, that there are some special predicates in *P*, which have no alternatives, even if their arguments have "synonyms". These predicates are named as bound predicates. For such predicates, the modified Herbrand base only includes atoms that are present in the original Herbrand base.
**Definition 14.** The modified consequence transformation *modNTP : FV(modBP)* → *FV(modBP )* is defined as
$$\begin{array}{c} \text{modNT}\_{\text{P}} \left( \mathbf{X} \right) = \left\{ \left( q(\mathbf{s}\_{1}, \dots, \mathbf{s}\_{n}) \right) \mid \underline{\mathfrak{Q}}\_{\text{p}} (\underline{\mathfrak{Q}}\_{\text{p}} \mid \underline{\mathfrak{A}}\_{\text{p}} \mid \underline{\mathfrak{A}}\_{\text{p}} \dots \mid \underline{\mathfrak{A}}\_{\text{p}}) \right) \\\\ \left( q, \underline{\lambda}\_{\mathfrak{Q}} \right) \in \underline{\operatorname{RS}}\_{\text{FV}\_{\text{P}'}} ; \left( \underline{\mathfrak{s}}\_{i} \underline{\lambda}\_{\mathfrak{s}\_{i}} \right) \in \underline{\operatorname{RG}}\_{\text{FV}\_{\text{V}\_{i}}} \cdot 1 \le i \le n \right\} \cup \mathbf{X}\_{\text{r}} \end{array}$$
where
$$(p(t\_1, \ldots, t\_n) \leftarrow A\_1, \ldots, A\_k; \underline{\mathcal{B}}\\_{\mathbb{L}}) \in \operatorname{ground}(\mathbb{P}),$$
$$(\lfloor A\_i \rfloor, \lfloor \underline{\mathcal{Q}}\_{A\_i} \rangle \in X, \; 1 \le i \le k) \qquad (\lfloor A\_i \rfloor \text{ is the kernel of } A\_i)$$
and α*<sup>p</sup>* is computed according to the actual extension of (1).
This transformation is inflationary over *FV(modBP)* and it is monotone if *P* is positive. So, according to (Ceri et al 1990) it has a least fixed point. If *P* is positive, this is the least fixed point. This fixed point is a model of *P*, but because *lfp(NTP)* ⊆ *lfp(mod NTP)*, it is not a minimal one (Achs 2010).
42 Fuzzy Logic – Algorithms, Techniques and Implementations
As the above examples show, the levels of "synonyms" can be computed according to proximity-based uncertainty functions. To determine all direct or indirect conclusions of the facts and rules of a program, a proximity based uncertainty function has to be ordered to each predicate of the program. The set of these functions will be the decoding-set of the
**Definition 13.** Let *P* be a multivalued Datalog program, and *FP* be the set of the program's
The original deducing mechanism makes conclusions according to the rules of the program, but from now on the background knowledge must be considered as well. So the original mechanism has to be modified. This modified deduction consists of two alternating parts: starting from the facts, their "synonyms" are determined, then applying the suitable rules, other facts are derived, followed by their "synonyms" determined, and again the rules are applied, etc. To define it in a precise manner the concept of modified consecution
The consequence transformation of a mDATALOG *P* program is defined over the set of all multivalued sets of *P*'s Herbrand base, that is, over *FV(BP)*. To define the modified transformation's domain, let us extend *P*'s Herbrand universe with all possible ground terms of the background knowledge, obtaining the so called modified Herbrand universe, *modHP*. The modified Herbrand base, *modBP* is the set of all ground atoms, whose predicate
However, it is possible, that there are some special predicates in *P*, which have no alternatives, even if their arguments have "synonyms". These predicates are named as bound predicates. For such predicates, the modified Herbrand base only includes atoms that
**Definition 14.** The modified consequence transformation *modNTP : FV(modBP)* → *FV(modBP )*
*)* ∈ *RGFVti*
*A1,…,Ak ;*
This transformation is inflationary over *FV(modBP)* and it is monotone if *P* is positive. So, according to (Ceri et al 1990) it has a least fixed point. If *P* is positive, this is the least fixed
ϕ*p(*α*p, λq, λs1*
β*; I )* ∈ *,…, λsn) ) |*
∪ *X,*
⊆
*lfp(mod NTP)*, it is not a
*, 1 ≤ i ≤ n}*
*k (|Ai| is the kernel of Ai)*
*ground(P),*
*, λ, λ1,…, λn) |*
∀ *p/n* ∈ *FP* }*.*
Φ*P =* {ϕ*p(*α
symbols occur in *P* ∪ *Bk* and whose arguments are elements of *modHP*.
*modNTP (X) = {(q(s1,…,sn),*
*(p(t1,…,tn)*
*<sup>p</sup>* is computed according to the actual extension of (1).
point. This fixed point is a model of *P*, but because *lfp(NTP)*
*RSFVp ; (si, λsi*
←
program.
functors. The decoding-set of *P* is:
transformation will be introduced.
are present in the original Herbrand base.
*(q, λq)* ∈
*(|Ai|,* α*Ai )* ∈ *X, 1* ≤ *i* ≤
is defined as
where
and α
minimal one (Achs 2010).
**3.3 Deduction with background knowledge**
The modifying algorithm is irrelevant to the evaluation sequence, so stratification can be applied with the same condition. That is, the modified consequence transformation has a least fixed point in the case of stratified programs as well. This transformation makes connections between an mDATALOG program, the background knowledge and the decoding-set of the program. So these four components can form a knowledge-base. However, there should be other transformations connecting the three other parts with each other, therefore the universal concept of a multivalued knowledge-base can be defined with an arbitrary deduction algorithm:
**Definition 15.** A multivalued knowledge-base (*mKB*) is a quadruple
$$
\dim \text{KB} = (P, \text{Bk}, \text{Qp}, \text{dA}(P, \text{Bk}, \text{Qp})),
$$
where *P* is a multivalued Datalog program, *Bk* is a background knowledge, Φ*<sup>P</sup>* is a decoding-set of *P* and *dA* is any deduction algorithm connecting the three other part with each other. The least fixed point of the deduction algorithm is called the consequence of the knowledge-base, denoted by
$$\mathbf{C}(\mathsf{Bk}, \mathsf{P}, \mathsf{Q}\_{\mathsf{P}}, \mathsf{d}\mathbf{A}) = \mathsf{l}\mathsf{f}\mathsf{p}(\mathsf{d}\mathbf{A}(\mathsf{P}, \mathsf{Bk}, \mathsf{Q}\_{\mathsf{P}})) .$$
Because the actual deduction algorithm is the modified consequence transformation, now the consequence is
$$\mathbf{C}(\mathbf{B}k, \mathrm{P}, \,\,\Phi\_{\mathrm{P}}, \,\mathrm{d}A) = \mathrm{l}\!\!\!p(mod \,\mathrm{NT}\_{\mathrm{P}})\,.$$
Note: If it is important to underline that there are bound predicates in P, then *mKB* can be denoted by *mKB = (P, Bp, Bk,*Φ*P, dA(P, Bp, Bk,*Φ*P)),* where *Bp* is the set of bound predicates.
**Example 10**. Let us suppose that an internet agent's job is to send a message to its clients if the cinema (*C*) presents a film, which its clients like. The agent knows that people generally go (*go*) to the cinema if they can pay (*cp*) for the ticket and are interested in (*in*) the film presented (*pr*) in the cinema. It also knows that people usually want to see (*ws*) a film if they like (*li*) its actor (*ac*). Moreover it knows that Paul (P) has enough money (*hm*) and he enjoys (*en*) Chaplin (*Ch*) very much. In the cinema, a film of Stan and Pan (*SP*) is presented. Should the agent inform Paul about this film? How much will he want to go to the cinema?
This situation can be modelled, for example, by the following multivalued knowledge-base.
Let the IVS valued mDATALOG program and the background knowledge be as follows
$$\text{go(x, C)} \leftarrow \text{pr(C, f)}, \text{in(x, f)}, \text{cp(x)}; \text{(0.85, 0.95)}; I\_{\text{VL}}. \tag{\mathbb{R}1}$$
$$ws(f, x) \leftarrow ac(f, y), \text{li(x, y): (0.8, 0.85); } I\_{VG\_2} \tag{R2}$$
*(hm(P), (0.75, 0.8)). (en(P, Ch), (0.9, 0.95)).*
*(pr(C, Film), (1,1)). (ac(Film, SP), (1,1)).*
According to their roll, *pr(C, Film)* and *ac(Film, SP)* have no alternatives.
From Fuzzy Datalog to Multivalued Knowledge-Base 45
atom. *q* may contain variables, and its levels may be known or unknown values. The
In standard Datalog, the most common approach in top-down direction is called query – sub-query framework. A goal, together with a program, determines a query. Literals in the body of any one of the rules defining the goal predicate are sub-goals of the given goal. Thus, a sub-goal together with the program yields a sub-query. In order to answer the query, each goal is expanded in a list of sub-goals, which are recursively expanded in their turn. That is, considering a goal, all rule-heads unifying with the goal are selected and the literals of the rule-body are the new sub-goals of given goal, which are evaluated one by one
The situation is the same with a multivalued knowledge-base as well, but in this case the algorithm is completed with the computation of the unification levels. However, it is possible that such rules do not exist, but perhaps they do exist for the synonyms. For example, the goal is to know if Ann likes Bach, but there are rules only for describing the options of loving somebody and there are facts only about Vivaldi. In such cases, the synonyms are used. Therefore, the algorithm has to consider the proximities and has to compute the uncertainty levels. It is a bidirectional evaluation: firstly, the uncertainty-free rules and the proximities are evaluated in a top-down manner, obtaining the required starting facts, then the computing of uncertainties is executed in the opposite direction, that
The uncertainty levels are not required in the top-down part of the evaluation, so, this part of the algorithm can be based on the concept of classical substitution and unification (Ceri et al 1990, Ullman 1988, etc.) However other kinds of substitutions may be necessary as well: to
From now on, for the sake of a simpler terminology, the terms "goal", "rule" and "fact" will refer to these concepts without uncertainty levels. An AND/OR tree arises during the evaluation, this is the searching tree. Its root is the goal; its leaves are either YES or NO. The parent nodes of YES are the facts, and uncertainty can be computed moving towards the direction of the root. This tree is built up by a periodic change of three kinds of steps: a
Proximity-based unification unifies the predicate symbols of sub-goals and the members of their proximity sets. Rule-based unification unifies the sub-goals with the head of suitable rules, and continues the evaluating by the bodies of these rules. The splitting step splits the rule-body into sub-goals if the body contains more literals or splits a literal of proximity sets
During the construction process, the edges are labelled by necessary information for computing the uncertainties. The searching graph according to its depth is build up in the
substitute some predicate *p* or term *t* with their proximity sets *RSFVp* and *RGFVt*
proximity-based unification, a rule-based unification and a splitting step.
in an arbitrary order. This procedure continues until the facts have been obtained.
is the fuzzy, the intuitionistic, the interval-valued or the bipolar level of the
α
*)*, where *q(t1, t2, …,*
, and to
enough to consider only a part of them. A goal is a pair *(q(t1, t2, …, tn);*
*tn)* is an atom,
α
evaluation algorithm gives answer to this query.
is, according to the fixed-point algorithm.
into literals of the suitable ground atoms.
following way.
**4.1 Evaluation of a general knowledge-base**
substitute some proximity sets with their members.
Let the proximities be:
$$
\underline{R}\_V \text{ (in, } wss) = (0.7, 0.8). \qquad \underline{R}\_V \text{ (Ch, SP)} = (0.8, 0.9).
$$
$$
\underline{R}\_V \text{ (li, } en) \quad = (0.8, 0.9).
$$
$$
\underline{R}\_V \text{ (cp, lom)} = (0.9, 0.95).
$$
According to the connecting algorithm, it is enough to consider only the proximity-based uncertainty functions of head-predicates. Let these functions be the minimum function:
$$\mathfrak{gl}\_{\mathbb{R}^3}(\underline{\mathbf{c}}\,\overline{\lambda}\,\,\underline{\lambda}\mathbb{T}\,,\,\underline{\lambda}\mathbb{T}) = \mathfrak{gl}\_{\mathbb{R}^3}(\underline{\mathbf{c}}\,\underline{\lambda}\,\,\underline{\lambda}\mathbb{T}\,,\,\underline{\lambda}\mathbb{T}) = \mathfrak{gl}\_{\mathbb{R}}(\underline{\mathbf{c}}\,\underline{\lambda}\,\,\underline{\lambda}\mathbb{T}\,,\,\underline{\lambda}\mathbb{T}) = \mathfrak{gl}\_{\mathbb{R}}(\underline{\mathbf{c}}\,\underline{\lambda}\,\,\underline{\lambda}\mathbb{T}\,\,\underline{\lambda}\mathbb{T}) = \mathfrak{gl}\_{\mathbb{R}}(\underline{\mathbf{c}}\,\underline{\lambda}\,\,\underline{\lambda}\mathbb{T}\,\,\underline{\lambda}\mathbb{T}) := \mathfrak{gl}\_{\mathbb{R}^3}(\underline{\mathbf{c}}\,\underline{\lambda}\,\,\underline{\lambda}\mathbb{T}\,\,\underline{\lambda}\mathbb{T}) := \mathfrak{gl}\_{\mathbb{R}^3}(\underline{\mathbf{c}}\,\underline{\lambda}\,\,\underline{\lambda}\mathbb{T}\,\,\underline{\lambda}\mathbb{T}) := \mathfrak{gl}\_{\mathbb{R}^3}(\underline{\mathbf{c}}\,\underline{\lambda}\,\,\underline{\lambda}\mathbb{T}\,\,\underline{\lambda}\mathbb{T}) := \mathfrak{gl}\_{\mathbb{R}^3}(\underline{\mathbf{c}}\,\underline{\lambda}\,\,\underline{\lambda}\mathbb{T}\,\,\underline{\lambda}\mathbb{T}) := \mathfrak{gl}\_{\mathbb{R}^3}(\underline{\mathbf{c}}\,\underline{\lambda}\,\,\underline{\lambda}\mathbb{T})$$
$$\underline{\mathfrak{g}}\_{\text{um}}(\underline{\mathfrak{g}}, \underline{\lambda}, \underline{\lambda}\_1) := \min\_{V} \left( \underline{\mathfrak{g}}, \underline{\lambda}, \underline{\lambda}\_1 \right),$$
The modified consequence transformation has the next steps:
*X0 =* {*(hm(P), (0.75, 0.8)), (en(P, Ch), (0.9, 0.95)), (pr(C, Film), (1,1)), (ac(Film, SP), (1,1))*} (according to the proximity) *X1 = modNTP(X0) = X0* ∪ {*(cp(P),* ϕ*hm((0.75, 0.8), (0.9, 0.95), (1,1)) = (min(0.75, 0.9, 1), min(0.8, 0.95, 1)) = (0.75, 0.8)), (en(P, SP),* ϕ*en((0.9, 0.95), (1,1), (1,1), (0.85, 0.9)) = (0.85, 0.9)), (li(P, Ch),* ϕ*en((0.9, 0.95), (0.8, 0.9), (1,1), (1,1)) = (0.8, 0.9)), (li(P, SP),* ϕ*en((0.9, 0.95), (0.8, 0.9), (1,1), (0.85, 0.9)) = (0.8, 0.9))*} (applying the rules – only (R2) can be applied) *X2 = modNTP (X1) = X1* ∪ {(*ws(Film, P), f(IVG2,* α*,* β *) = f(IVG2, minV((1,1), (0.8, 0.9)), (0.8, 0.85)) = minV((0.8, 0.9),(0.8, 0.85)) = (0.8, 0.85))* } (according to the proximity) *X3 = modNTP (X2) = X2* ∪ {*(in(Film, P),* ϕ*ws( (0.8, 0.85), (0.7, 0.8), (1,1), (1,1)) = (0.7, 0.8))*} (applying the rules – (R1) can be applied) *X4 = modNTP (X3) = X3* ∪ {*(go(P, C), f(IVL,* α*,* β *) = f(IVL, minV((1,1), (0.7, 0.8), (0.75, 0.8)), (0.85, 0.95)) = (max(0, 0.8 + 0.85 - 1), max(0, 0.7 + 0.95 - 1)) = (0.65, 0.65))*}
X*4* is a fixed point, so it is the consequence of the knowledgebase.
According to this result, the agent will know that the message can be sent, because Paul will probably enjoy Stan and Pan at a likelihood level of 85-90% (level (0.85, 0.9)), and there is a good chance (65%) that Paul will go to the cinema.
### **4. Evaluating algorithms**
The fixed point-query is a bottom-up evaluation algorithm, which may involve many superfluous calculations. However, very often, only a particular question is of interest and the answer to this question needs to be searched. If a goal (query) is specified together with the multivalued knowledge-base, it is not necessary to evaluate all facts and rules, and it is 44 Fuzzy Logic – Algorithms, Techniques and Implementations
According to the connecting algorithm, it is enough to consider only the proximity-based uncertainty functions of head-predicates. Let these functions be the minimum function:
*, λ, λ1, λ2)*
*X0 =* {*(hm(P), (0.75, 0.8)), (en(P, Ch), (0.9, 0.95)), (pr(C, Film), (1,1)), (ac(Film, SP), (1,1))*}
*en((0.9, 0.95), (1,1), (1,1), (0.85, 0.9)) = (0.85, 0.9)),*
*en((0.9, 0.95), (0.8, 0.9), (1,1), (0.85, 0.9)) = (0.8, 0.9))*} (applying the rules – only (R2) can be applied)
*ws( (0.8, 0.85), (0.7, 0.8), (1,1), (1,1)) = (0.7, 0.8))*} (applying the rules – (R1) can be applied)
According to this result, the agent will know that the message can be sent, because Paul will probably enjoy Stan and Pan at a likelihood level of 85-90% (level (0.85, 0.9)), and there is a
The fixed point-query is a bottom-up evaluation algorithm, which may involve many superfluous calculations. However, very often, only a particular question is of interest and the answer to this question needs to be searched. If a goal (query) is specified together with the multivalued knowledge-base, it is not necessary to evaluate all facts and rules, and it is
*en((0.9, 0.95), (0.8, 0.9), (1,1), (1,1)) = (0.8, 0.9)),*
*, λ, λ1, λ2),*
= ϕ*pr(*α
*hm((0.75, 0.8), (0.9, 0.95), (1,1)) = (min(0.75, 0.9, 1), min(0.8, 0.95, 1)) = (0.75, 0.8)),*
*) = f(IVG2, minV((1,1), (0.8, 0.9)), (0.8, 0.85)) =*
*) = f(IVL, minV((1,1), (0.7, 0.8), (0.75, 0.8)), (0.85, 0.95)) =*
*, λ, λ1, λ2)*
= ϕ*ac(*α
*, λ, λ1, λ2)*
:=
*RV (in, ws) = (0.7, 0.8). RV (Ch, SP) = (0.8, 0.9).*
= ϕ*en(*α
(according to the proximity)
(according to the proximity)
*(max(0, 0.8 + 0.85 - 1), max(0, 0.7 + 0.95 - 1)) = (0.65, 0.65))*} X*4* is a fixed point, so it is the consequence of the knowledgebase.
*minV (*α
*RV (li, en) = (0.8, 0.9).*
*RV (cp, hm) = (0.9, 0.95).*
*, λ, λ1, λ2)*
ϕ*hm(*α*, λ, λ1)* := *minV (*α*, λ, λ1),*
The modified consequence transformation has the next steps:
α*,* β
*minV((0.8, 0.9),(0.8, 0.85)) = (0.8, 0.85))* }
ϕ
α*,* β
good chance (65%) that Paul will go to the cinema.
Let the proximities be:
*, λ, λ1, λ2)*
*X1 = modNTP(X0) = X0* ∪
*X2 = modNTP (X1) = X1* ∪ {(*ws(Film, P), f(IVG2,*
*X3 = modNTP (X2) = X2* ∪
*X4 = modNTP (X3) = X3* ∪ {*(go(P, C), f(IVL,*
**4. Evaluating algorithms**
{*(in(Film, P),*
ϕ
ϕ
ϕ
ϕ
{*(cp(P),*
*(en(P, SP),*
*(li(P, Ch),*
*(li(P, SP),*
= ϕ*ws(*α
ϕ*go(*α enough to consider only a part of them. A goal is a pair *(q(t1, t2, …, tn);* α*)*, where *q(t1, t2, …, tn)* is an atom, α is the fuzzy, the intuitionistic, the interval-valued or the bipolar level of the atom. *q* may contain variables, and its levels may be known or unknown values. The evaluation algorithm gives answer to this query.
In standard Datalog, the most common approach in top-down direction is called query – sub-query framework. A goal, together with a program, determines a query. Literals in the body of any one of the rules defining the goal predicate are sub-goals of the given goal. Thus, a sub-goal together with the program yields a sub-query. In order to answer the query, each goal is expanded in a list of sub-goals, which are recursively expanded in their turn. That is, considering a goal, all rule-heads unifying with the goal are selected and the literals of the rule-body are the new sub-goals of given goal, which are evaluated one by one in an arbitrary order. This procedure continues until the facts have been obtained.
The situation is the same with a multivalued knowledge-base as well, but in this case the algorithm is completed with the computation of the unification levels. However, it is possible that such rules do not exist, but perhaps they do exist for the synonyms. For example, the goal is to know if Ann likes Bach, but there are rules only for describing the options of loving somebody and there are facts only about Vivaldi. In such cases, the synonyms are used. Therefore, the algorithm has to consider the proximities and has to compute the uncertainty levels. It is a bidirectional evaluation: firstly, the uncertainty-free rules and the proximities are evaluated in a top-down manner, obtaining the required starting facts, then the computing of uncertainties is executed in the opposite direction, that is, according to the fixed-point algorithm.
### **4.1 Evaluation of a general knowledge-base**
The uncertainty levels are not required in the top-down part of the evaluation, so, this part of the algorithm can be based on the concept of classical substitution and unification (Ceri et al 1990, Ullman 1988, etc.) However other kinds of substitutions may be necessary as well: to substitute some predicate *p* or term *t* with their proximity sets *RSFVp* and *RGFVt*, and to substitute some proximity sets with their members.
From now on, for the sake of a simpler terminology, the terms "goal", "rule" and "fact" will refer to these concepts without uncertainty levels. An AND/OR tree arises during the evaluation, this is the searching tree. Its root is the goal; its leaves are either YES or NO. The parent nodes of YES are the facts, and uncertainty can be computed moving towards the direction of the root. This tree is built up by a periodic change of three kinds of steps: a proximity-based unification, a rule-based unification and a splitting step.
Proximity-based unification unifies the predicate symbols of sub-goals and the members of their proximity sets. Rule-based unification unifies the sub-goals with the head of suitable rules, and continues the evaluating by the bodies of these rules. The splitting step splits the rule-body into sub-goals if the body contains more literals or splits a literal of proximity sets into literals of the suitable ground atoms.
During the construction process, the edges are labelled by necessary information for computing the uncertainties. The searching graph according to its depth is build up in the following way.
From Fuzzy Datalog to Multivalued Knowledge-Base 47
In this way each starting fact can be appointed. Then a solution can be determined by connecting the suitable unifications and computing in succession the uncertainties according to the labels of edges in the path from the symbol YES to the root of the graph. The union of
**Proposition 7**. For a given goal and in the case of finite evaluation graph, the top-down
**Example 11.** Consider the IFS program of Example 6, and let it be completed by proximities
*p(x), q(y); (0.75, 0.2); IFG2.*
The three facts-sets can easily be seen in the graph. From the first one, the uncertainty of
As *(r(a, b), (0.7, 0.2))* is a known member of the set, so knowing this uncertainty, the proximity-based function and the proximities of knowledge base, the uncertainties can be
*r (a1, b), (minF ((0.7, 0.2), (1, 0), (0.8, 0.1), (1,0))= (0.7, 0.2)).*
*r (a1, b1), (minF ((0.7, 0.2), (1, 0), (0.8, 0.1), (0.6,0.3))= (0.6, 0.3)).* Applying the next label of the path the uncertainty of *r1(a1, b)* and *r1(a1, b1)* can be found.
*(r1(a1, b), (0.7, 0.2))*
*(r1(a1, b1), (0.6, 0.3))*
ϕ*p(*α*, λ, λ1)*
ϕ*q(*α*, λ, λ1)*
ϕ*r(*α
*)*, where *x* is a variable. Then the evaluation graph is on the next
:= α⋅ *λ*⋅*λ1.*
:= *minF (*α*, λ*⋅*λ1).*
α
*, λ, λ1, λ2).*
*, λ, λ1, λ2) := minF (*
evaluation terminates and gives the same answer as the fixed point query.
←
these solutions is the answer to the given query. From the construction of searching graph follows:
and proximity-based uncertainty functions
*(p(a), (0.7, 0.2)).*
*(q(b), (0.65, 0.3)).*
*(r(a, b), (0.7, 0.2)).*
*RF (p, p1) = (0.7, 0.1).*
*RF (q, q1) = (0.8, 0.1).*
*RF (r, r1) = (0.75, 0.2).*
*RF (a, a1) = (0.8, 0.1).*
*RF (b, b1) = (0.6, 0.3).*
*r(a1, b)* and *r(a1, b1)* can be computed.
α
*r(x, y)*
Let the goal be: *(r1(a1,x);*
page.
determined:
They are as follows:
If the goal is on depth *0*, then every successor of any node on depth *3k+2 (k = 0, 1, …)* is in AND connection, and the others are in OR connection. The step after depth *3k (k = 0, 1, …)* is a proximity-based unification, after depth *3k+1 (k=0, 1, …)* is a rule-based unification and after depth *3k+2 (k=0,1,…)* is a splitting step. In detail:
If the atom *p(t1, t2, …, tn)* is in depth *3k (k = 0, 1, …)*, then the successor nodes let be all possible *p'(t1, t2, …, tn)*, where *p'* ∈ *RSFVp* . The edges starting from these nodes are labelled by the proximity-based uncertainty functions ϕ*p'*.
If the atom *L* is in depth *3k+1 (k=0, 1, …)*, then the successor nodes will be
That is, if the head of rule *M* ← *M1,…,Mn*, *(n>0)* is unifiable with *L*, then the successor of *L* will be *M1*θ*,…,Mn*θ, where θ is the most general unification of *L* and *M*. The edges starting from these nodes are labelled by the uncertainty functions belonging to the implication operator of the applied rules and by the uncertainty level of the rule.
If *n=0*, that is, in the program there is any fact with the predicate symbol of *L*, then the successors will be the unified facts. If *L = p(t1, t2, …, tn)* and in the program there is any fact with predicate symbol *p*, then let the successor nodes be all possible *p(t'1, t'2,…, t'n)*, where *t'i* = *ti* if *ti* is a ground term or *t'i* = *RGFVti*θ if *ti* is a variable and θ is a suitable unification. The edges starting from these nodes are not labelled.
According to the previous paragraph, there are three kinds of nodes in depth *3k+2 (k=0,1,…)*: a unified body of a rule; a unified fact the arguments of with are ordinary ground terms or proximity sets; or the symbol NO.
In the first case, the successors are the members of the body. They are in AND connection. The connected edges will not be labelled, but because of the AND connection, during the computation, the minimum value of the successor's levels will be regarded.
In the second case, the successors are the so called facts-sets. This means, that if the node is *p(t1, t2,…, tn)*, where *ti* is a ground term or a proximity set, then the facts-set is the set of all possible *p(t'1, t'2,…, t'n)*, where *t'i* ∈ *RGFVti* . The edges starting from these nodes are labelled by the proximity-based uncertainty functions ϕ*p*.
The facts-set has a further successor, the symbol YES.
The NO-node has no successor.
A solution can be achieved in the graph along the path ending in the symbol YES. According to the unification algorithm, one of the literals that are located at the parent node of YES can also be found among the original facts of the program. Knowing its uncertainty and using the proximity-based uncertainty function of the label leading to this facts-set, the uncertainty of other members of the facts-set can be computed. However, it is not necessary to compute all of them, only those ones, which are appropriate for the pattern of the literal being in the parent node of facts-set. This means, that if the literal is *p(t1, t2,…, tn)*, and *ti* is a ground term, then it is enough to consider only *p(t'1, t'2,…, ti ,…, t'n)* from the facts-set, but if *ti* is a proximity set, then it is necessary to deal with all *p(t'1, t'2,…, t'i ,…, t'n)*, where *t'i* ∈ *RGFVti* .
46 Fuzzy Logic – Algorithms, Techniques and Implementations
If the goal is on depth *0*, then every successor of any node on depth *3k+2 (k = 0, 1, …)* is in AND connection, and the others are in OR connection. The step after depth *3k (k = 0, 1, …)* is a proximity-based unification, after depth *3k+1 (k=0, 1, …)* is a rule-based unification and
If the atom *p(t1, t2, …, tn)* is in depth *3k (k = 0, 1, …)*, then the successor nodes let be all possible *p'(t1, t2, …, tn)*, where *p'* ∈ *RSFVp* . The edges starting from these nodes are labelled
> ϕ*p'*.
That is, if the head of rule *M* ← *M1,…,Mn*, *(n>0)* is unifiable with *L*, then the successor of *L*
from these nodes are labelled by the uncertainty functions belonging to the implication
If *n=0*, that is, in the program there is any fact with the predicate symbol of *L*, then the successors will be the unified facts. If *L = p(t1, t2, …, tn)* and in the program there is any fact with predicate symbol *p*, then let the successor nodes be all possible *p(t'1, t'2,…, t'n)*, where
According to the previous paragraph, there are three kinds of nodes in depth *3k+2 (k=0,1,…)*: a unified body of a rule; a unified fact the arguments of with are ordinary ground
In the first case, the successors are the members of the body. They are in AND connection. The connected edges will not be labelled, but because of the AND connection, during the
In the second case, the successors are the so called facts-sets. This means, that if the node is *p(t1, t2,…, tn)*, where *ti* is a ground term or a proximity set, then the facts-set is the set of all possible *p(t'1, t'2,…, t'n)*, where *t'i* ∈ *RGFVti* . The edges starting from these nodes are labelled
> ϕ*p*.
A solution can be achieved in the graph along the path ending in the symbol YES. According to the unification algorithm, one of the literals that are located at the parent node of YES can also be found among the original facts of the program. Knowing its uncertainty and using the proximity-based uncertainty function of the label leading to this facts-set, the uncertainty of other members of the facts-set can be computed. However, it is not necessary to compute all of them, only those ones, which are appropriate for the pattern of the literal being in the parent node of facts-set. This means, that if the literal is *p(t1, t2,…, tn)*, and *ti* is a ground term, then it is enough to consider only *p(t'1, t'2,…, ti ,…, t'n)* from the facts-set, but if *ti* is a proximity set, then
if *ti* is a variable and
θ
computation, the minimum value of the successor's levels will be regarded.
it is necessary to deal with all *p(t'1, t'2,…, t'i ,…, t'n)*, where *t'i* ∈ *RGFVti* .
is the most general unification of *L* and *M*. The edges starting
θ
is a suitable unification. The
If the atom *L* is in depth *3k+1 (k=0, 1, …)*, then the successor nodes will be
• the unified facts if L is unifiable with any fact of the program, or
operator of the applied rules and by the uncertainty level of the rule.
after depth *3k+2 (k=0,1,…)* is a splitting step. In detail:
by the proximity-based uncertainty functions
• NO, if there is not any unifiable rule or fact.
θ
• the bodies of suitable unified rules or
*t'i* = *ti* if *ti* is a ground term or *t'i* = *RGFVti*
terms or proximity sets; or the symbol NO.
by the proximity-based uncertainty functions
The NO-node has no successor.
The facts-set has a further successor, the symbol YES.
edges starting from these nodes are not labelled.
will be *M1*
θ*,…,Mn*θ, where In this way each starting fact can be appointed. Then a solution can be determined by connecting the suitable unifications and computing in succession the uncertainties according to the labels of edges in the path from the symbol YES to the root of the graph. The union of these solutions is the answer to the given query.
From the construction of searching graph follows:
**Proposition 7**. For a given goal and in the case of finite evaluation graph, the top-down evaluation terminates and gives the same answer as the fixed point query.
**Example 11.** Consider the IFS program of Example 6, and let it be completed by proximities and proximity-based uncertainty functions
```
(p(a), (0.7, 0.2)).
(q(b), (0.65, 0.3)).
(r(a, b), (0.7, 0.2)).
r(x, y) ← p(x), q(y); (0.75, 0.2); IFG2.
RF (p, p1) = (0.7, 0.1). ϕp(α, λ, λ1) := α⋅ λ⋅ λ1.
RF (q, q1) = (0.8, 0.1). ϕq(α, λ, λ1) := minF (α, λ⋅ λ1).
RF (r, r1) = (0.75, 0.2). ϕr(α, λ, λ1, λ2) := minF (α, λ, λ1, λ2).
RF (a, a1) = (0.8, 0.1).
RF (b, b1) = (0.6, 0.3).
```
Let the goal be: *(r1(a1,x);* α*)*, where *x* is a variable. Then the evaluation graph is on the next page.
The three facts-sets can easily be seen in the graph. From the first one, the uncertainty of *r(a1, b)* and *r(a1, b1)* can be computed.
As *(r(a, b), (0.7, 0.2))* is a known member of the set, so knowing this uncertainty, the proximity-based function and the proximities of knowledge base, the uncertainties can be determined:
*r (a1, b), (minF ((0.7, 0.2), (1, 0), (0.8, 0.1), (1,0))= (0.7, 0.2)).*
*r (a1, b1), (minF ((0.7, 0.2), (1, 0), (0.8, 0.1), (0.6,0.3))= (0.6, 0.3)).*
Applying the next label of the path the uncertainty of *r1(a1, b)* and *r1(a1, b1)* can be found. They are as follows:
```
(r1(a1, b), (0.7, 0.2))
(r1(a1, b1), (0.6, 0.3))
```
From Fuzzy Datalog to Multivalued Knowledge-Base 49
As mentioned earlier, in the case of standard Datalog, the heart of the evaluation algorithm is the unification process. Our special evaluation of multivalued knowledge is based on unification as well, but on multivalued unification. The multivalued unification consists of two parts, one is the alternation of a rule-based-unification and the other is a proximity-based one. Both of them are the extensions of the classical unification algorithm. Now, the splitting step is inside these unifications: evaluating a fact, the last proximity-based unification unifies the fact with its facts-set, and a rule-based unification splits these sets into their members.
This unification algorithm is similar to the classical one, that is, the goal can be unified with the body of any one of the rules defining the goal predicate – if the body is not empty. The level of the unification is the level of the rule defining the goal predicate. In that case, a variable can be substituted with other variable or with a constant; a constant can be substituted with itself only. The next sub-goal of the evaluation process will be the first member of the body. It is possible that during the evaluation a variable of a later member is substituted by a proximity set. In such a case, in the course of later evaluation, this
If the predicate symbol of the goal is the predicate symbol of a fact, its arguments can be
• The variables of the goal can be substituted with the proximity set of the constants being the corresponding arguments of the fact. (E.g.: if *q(a,b)* is a fact predicate and the
There is a special case of unification: the facts-set of a predicate is unified with its members
According to the previous unifications, between the literals of these facts-set there is one from the facts of the program. Knowing its uncertainty and the level of proximities, the uncertainty of other members can be computed. Then these members can be unified
The level of this unification is the level of the fitting member, and the former proximity set-
set of unification is {*p(a), p(b)*}, which is unified with empty clauses . The substitutions for *x* are *x|a* and *x|b*, and the levels are the computed levels of *p(a)* and *p(b)* respectively. )
• with the empty clause, if there is no other literal to evaluate in the parent-node;
.)
= {*a, b*}, then the facts-
goal is *q(x, y)*, then *x* can be substituted with *RFVa* and *y* with *RFVb*
• with their proximity set if the goal does not contain any variable. • The proximity set argument of a goal can be substituted with itself.
If there is no fact with the same predicate symbol, the unification process fails.
• with themselves, if the goal contains any variable or
• with the remaining part of the body to be evaluated.
substitution of a variable is replaced by the suitable member of this set.
(E.g. : if there is a former *x|RFVa* substitution for literal *p(x)*, and *RFVa*
**4.2.1 Rule-based unification**
substituted as follows:
proximity set will be substituted with itself.
• The constants of the goal can be substituted
The level of the unification is *1FV*.
in the following way:
respectively:
Fig. 1. The evaluation graph of Example 11.
In the second facts-set *(p(a), (0.7, 0.2))* is the known fact, from this *(p(a1), (0.56, 0.28))*. Similarly *(q(b), (0.65, 0.3)), (q(b1), (0.6, 0.3))*. Applying the min, fFG2 and ϕr functions:
```
(r1(a1, b), (0.56, 0.3))
(r1(a1, b1), (0.56, 0.3))
```
As the answer is the union of the different solutions, the final answer is:
```
(r1(a1, b), (0.7, 0.2))
(r1(a1, b1), (0.6, 0.3))
```
### **4.2 Special evaluation based on multivalued unification**
The necessity of bidirectional evaluation is derived from the generality of implications and proximity functions, because their values can be computed only from known arguments, namely in bottom-up manner. However, in special cases, computation can be realized parallel with the evaluation of rules and proximities, so the algorithm could be a more efficient pure top-down evaluation. This is the situation if all of the functions are the minimum function. That is, all implications that are used are the second Gödel implication (G2) and all proximity-based uncertainty functions are the minimum function.
As mentioned earlier, in the case of standard Datalog, the heart of the evaluation algorithm is the unification process. Our special evaluation of multivalued knowledge is based on unification as well, but on multivalued unification. The multivalued unification consists of two parts, one is the alternation of a rule-based-unification and the other is a proximity-based one. Both of them are the extensions of the classical unification algorithm. Now, the splitting step is inside these unifications: evaluating a fact, the last proximity-based unification unifies the fact with its facts-set, and a rule-based unification splits these sets into their members.
### **4.2.1 Rule-based unification**
48 Fuzzy Logic – Algorithms, Techniques and Implementations
In the second facts-set *(p(a), (0.7, 0.2))* is the known fact, from this *(p(a1), (0.56, 0.28))*.
Similarly *(q(b), (0.65, 0.3)), (q(b1), (0.6, 0.3))*. Applying the min, fFG2 and ϕr functions:
As the answer is the union of the different solutions, the final answer is:
(G2) and all proximity-based uncertainty functions are the minimum function.
**4.2 Special evaluation based on multivalued unification**
*(r1(a1, b), (0.56, 0.3))*
*(r1(a1, b1), (0.56, 0.3))*
*(r1(a1, b), (0.7, 0.2))*
*(r1(a1, b1), (0.6, 0.3))*
The necessity of bidirectional evaluation is derived from the generality of implications and proximity functions, because their values can be computed only from known arguments, namely in bottom-up manner. However, in special cases, computation can be realized parallel with the evaluation of rules and proximities, so the algorithm could be a more efficient pure top-down evaluation. This is the situation if all of the functions are the minimum function. That is, all implications that are used are the second Gödel implication
Fig. 1. The evaluation graph of Example 11.
This unification algorithm is similar to the classical one, that is, the goal can be unified with the body of any one of the rules defining the goal predicate – if the body is not empty. The level of the unification is the level of the rule defining the goal predicate. In that case, a variable can be substituted with other variable or with a constant; a constant can be substituted with itself only. The next sub-goal of the evaluation process will be the first member of the body. It is possible that during the evaluation a variable of a later member is substituted by a proximity set. In such a case, in the course of later evaluation, this proximity set will be substituted with itself.
If the predicate symbol of the goal is the predicate symbol of a fact, its arguments can be substituted as follows:
- with themselves, if the goal contains any variable or
- with their proximity set if the goal does not contain any variable.
The level of the unification is *1FV*.
If there is no fact with the same predicate symbol, the unification process fails.
There is a special case of unification: the facts-set of a predicate is unified with its members in the following way:
According to the previous unifications, between the literals of these facts-set there is one from the facts of the program. Knowing its uncertainty and the level of proximities, the uncertainty of other members can be computed. Then these members can be unified respectively:
The level of this unification is the level of the fitting member, and the former proximity setsubstitution of a variable is replaced by the suitable member of this set.
(E.g. : if there is a former *x|RFVa* substitution for literal *p(x)*, and *RFVa* = {*a, b*}, then the factsset of unification is {*p(a), p(b)*}, which is unified with empty clauses . The substitutions for *x* are *x|a* and *x|b*, and the levels are the computed levels of *p(a)* and *p(b)* respectively. )
From Fuzzy Datalog to Multivalued Knowledge-Base 51
*ac(f, y), li(x, y); (0.8, 0.85); IVG2. (R1)*
*su(f, y), li(x, y); (0.75, 0.85); IVG2. (R2)*
*ws(f, x)*
*ws(f, x)*
(en(P, Ch), (0.9, 0.95)). (en(P, H), (0.6, 0.7)).
(ac(F1, SP), (1, 1)). (to(F2, W), (0.9, 0.95)). (to(F3, W), (0.55, 0.6)).
RV (to, su ) = (0.9, 1).
Fig. 2. The evaluation graph of Example 12.
Let each proximity-based uncertainty function be the minimum function.
applied substitution is *x|F1* and the minimum of levels is *(0.7, 0.8)*.
According to the left path of the graph one can see, that *(in (P, F1), (0.7, 0.8))* because the
The other paths are only half-drawn, and they are continued in a partial graph, because this part of evaluation is similar in all cases. The only differences are in uncertainty levels.
Then the evaluation graph is Fig.2.
←
←
According to its roll, *ac(F1, SP)* has no alternatives. Let the other proximities be:
RV (in, ws) = (0.7, 0.8). RV (Ch, SP) = (0.8, 0.9). RV (li, en) = (0.8, 0.9). RV (H, W) = (0.6, 0.8).
### **4.2.2 Proximity-based unification**
This unification serves for handling proximities. When these steps are implemented, the following substitutions can be realized:
### **4.2.3 The unification algorithm**
The evaluation algorithm combines the two kinds of unification. It starts with the proximity based unification and after it is finished, they alternate. The query is successful if the unification algorithm ends with an empty clause or a failure. In the first case the variables get the values defined during the substitutions. If all uncertainties are regarded as a minimum value, the actual level of unification can be computed as the minimum of former levels and when the algorithm reaches the empty clause, its level will be the level of the goal as well.
If the unification algorithm ends with a failure, there is no answer on this path.
If more answers arise during the evaluation, their union will be the resolution of the query.
According to the construction of unifications, the following proposition is true.
**Proposition 8.** For a given goal, and in the case of a finite evaluation graph, the above topdown evaluation gives the same answer as the fixed point query.
Notes:
**Example 12.** Let us consider a part of Example 10, and let it be completed with a new rule and new facts. That is now the internet agent knows that people usually want to see (*ws*) a film if they like (*li*) its actor (*ac*), or they like more or less the subject (*su*) of the film. Moreover, it knows that Paul (P) enjoys (*en*) Chaplin (*Ch*) very much and mostly enjoys the historical films (*H*). In the cinema, a film (*F1*) of Stan and Pan (*SP*) is presented. There are two other films (*F2, F3*). Both films' topics (*to*) are the war (*W*), but in different manner. The first one's central message is the war, the second one play in wartime, but it is only a background. From the former example it is known, that the agent wants to know the interest of Paul. Therefore let our goal be *(in(P,x);* α*)*.
Let the IVS valued mDATALOG program and the background knowledge be as follows
50 Fuzzy Logic – Algorithms, Techniques and Implementations
This unification serves for handling proximities. When these steps are implemented, the
• A predicate symbol can be substituted with the elements of its proximity set. The level
• A proximity set can be substituted with itself except in the last step of the evaluation of a literal. In this case, that is if each argument of the literal is a proximity set, the literal
The evaluation algorithm combines the two kinds of unification. It starts with the proximity based unification and after it is finished, they alternate. The query is successful if the unification algorithm ends with an empty clause or a failure. In the first case the variables get the values defined during the substitutions. If all uncertainties are regarded as a minimum value, the actual level of unification can be computed as the minimum of former levels and when the algorithm reaches the empty clause, its level will be the level of the goal
If more answers arise during the evaluation, their union will be the resolution of the query.
**Proposition 8.** For a given goal, and in the case of a finite evaluation graph, the above top-
• Although this algorithm was described for a knowledge-base based on a negation free program, it is similar in the case of stratified programs; the only difference is the calculation of the uncertainty of the negated sub-goal, but the computing of minimum
• With a good depth limit this algorithm is suitable for evaluating recursive programs or
**Example 12.** Let us consider a part of Example 10, and let it be completed with a new rule and new facts. That is now the internet agent knows that people usually want to see (*ws*) a film if they like (*li*) its actor (*ac*), or they like more or less the subject (*su*) of the film. Moreover, it knows that Paul (P) enjoys (*en*) Chaplin (*Ch*) very much and mostly enjoys the historical films (*H*). In the cinema, a film (*F1*) of Stan and Pan (*SP*) is presented. There are two other films (*F2, F3*). Both films' topics (*to*) are the war (*W*), but in different manner. The first one's central message is the war, the second one play in wartime, but it is only a background. From the former example it is known, that the agent wants to know the interest
α*)*. Let the IVS valued mDATALOG program and the background knowledge be as follows
**4.2.2 Proximity-based unification**
**4.2.3 The unification algorithm**
as well.
Notes:
remains the same.
infinite graphs as well.
of Paul. Therefore let our goal be *(in(P,x);*
following substitutions can be realized:
• A constant or a variable can be substituted with itself only.
can be unified with its facts-set. The level of the unification is *1FV*.
If the unification algorithm ends with a failure, there is no answer on this path.
According to the construction of unifications, the following proposition is true.
down evaluation gives the same answer as the fixed point query.
of the unification is the current proximity value.
$$\begin{aligned} \text{ws}(\mathbf{f}, \mathbf{x}) &\leftarrow \text{ac}(\mathbf{f}, \mathbf{y}), \text{li}(\mathbf{x}, \mathbf{y}); (0.8, 0.85); \text{I}\_{\text{V}\mathbf{G}}. \end{aligned} \tag{R1}$$
$$\begin{aligned} \text{ws}(\mathbf{f}, \mathbf{x}) &\leftarrow \text{su}(\mathbf{f}, \mathbf{y}), \text{li}(\mathbf{x}, \mathbf{y}); (0.75, 0.85); \text{I}\_{\text{V}\mathbf{G}}. \end{aligned} \tag{R2}$$
$$\begin{aligned} \text{(en(P, Ch), (0.9, 0.95))}. \\ \text{(en(P, H), (0.6, 0.7))}. \\\\ \text{(ac(F1, SP), (1, 1))}. \\ \text{(to(F2, W), (0.9, 0.95))}. \\ \text{(to(F3, W), (0.55, 0.6))}. \end{aligned}$$
According to its roll, *ac(F1, SP)* has no alternatives. Let the other proximities be:
Then the evaluation graph is Fig.2.
Let each proximity-based uncertainty function be the minimum function.
According to the left path of the graph one can see, that *(in (P, F1), (0.7, 0.8))* because the applied substitution is *x|F1* and the minimum of levels is *(0.7, 0.8)*.
The other paths are only half-drawn, and they are continued in a partial graph, because this part of evaluation is similar in all cases. The only differences are in uncertainty levels.
From Fuzzy Datalog to Multivalued Knowledge-Base 53
The deduction mechanism can be any of the extensions of Datalog. These extensions are the fuzzy Datalog, based on fuzzy logic, the intuitionistic- or interval-value Datalog, based on the suitable logics and bipolar Datalog, which is some kind of coexistence of the former ones.
The semantics of Datalog is a fixed-point semantics, so the algorithm, which connects the two main pillars of the knowledge-base is the generalization of the consequence transformation determining this fixed-point. This transformation is defined on the extended Herbrand base of the knowledge-base, which is generated from the ground terms of
Applying this transformation it is necessary to compute the uncertainty levels of "synonyms". The proximity-based uncertainty functions can do it, giving uncertainty values from the levels of the original fact and from the proximity values. The set of this kind of
Two possible evaluation strategies were presented as well. One of them evaluates a general knowledgebase with arbitrary proximity-based uncertainty functions and arbitrary implication operators. The other one allows minimum functions only as proximity based uncertainty functions, and the special extension of Gödel operator, but in this case a multivalued unification algorithm can be determined. This strategy is based on the
The improvement of this strategy and/or the deduction algorithm and/or the structure of
A well structured multivalued knowledge-base and an efficient evaluating algorithm determining its consequence could be the basis of making decisions based on uncertain information, or it would be useful for handling argumentation or negotiation of agents. An
Abiteboul, S.; Hull, R. and Vianu, V. (1995). *Foundations of Databases*. Addison-Wesley
Achs, A. & Kiss, A. (1995). Fuzzy extension of datalog, *Acta Cybernetica Szeged*, Vol.12, pp.
Achs, A. (2006). Models for handling uncertainty, *PhD thesis*, University of Debrecen, 2006. Achs, A. (2007). From Fuzzy- to Bipolar- Datalog, In: *Proceedings of 5th EUSFLAT Conference*,
Achs, A. (2010). A multivalued knowledge-base model, *Acta Universitatis Sapientiae,*
Alsinet, T. & Godo, L. (1998). Fuzzy Unification Degree, In: *Proceedings 2nd Int Workshop on*
Atanassov, K. (1983). Intuitionistic fuzzy sets, *VII ITKR's Session, Sofia* (deposed in Central Science-Technical Library of Bulgarian Academy of Science, 1697/84). Atanassov, K. & Gargov, G. (1989). Interval-valued intuitionistic fuzzy sets, *Fuzzy Sets and*
Atanassov, K. (1994). Remark on intuitionistic fuzzy expert systems. *BUSEFAL*, Vol.59, pp.
*Logic Programming and Soft Computing '98, in conjunction with JICSLP'98,*
implementation of this model would be an interesting future development as well.
knowledge base and its background knowledge.
alternating rule-based- and proximity-based unification.
background knowledge is a subject of further investigations.
Publishing Company, Reading, Massachusetts.
Ostrava, Czech Republic, pp. 221-227.
*Informatica*, Vol.2, No.1, pp. 51-79.
*Systems*, Vol.31, No.3, pp. 343-349.
Manchester, UK, pp. 23-43.
functions is the decoding set.
**6. References**
153-166.
71-76.
So, according to these paths the other answers for the query are:
*(in (P, F2), (0.6, 0.7)), (in (P, F3), (0.55, 0.6))*
Fig. 3. The enlarged evaluation graph of Example 12.
### **5. Conclusion**
In this chapter, a possible multivalued knowledge-base was presented as a quadruple of background knowledge, a deduction mechanism, a decoding set and an algorithm connecting the background knowledge to the deduction mechanism.
The background knowledge is based on proximity relations between terms and between predicates and it serves as a mechanism handling "synonyms".
52 Fuzzy Logic – Algorithms, Techniques and Implementations
So, according to these paths the other answers for the query are:
*(in (P, F2), (0.6, 0.7)), (in (P, F3), (0.55, 0.6))*
Fig. 3. The enlarged evaluation graph of Example 12.
connecting the background knowledge to the deduction mechanism.
predicates and it serves as a mechanism handling "synonyms".
In this chapter, a possible multivalued knowledge-base was presented as a quadruple of background knowledge, a deduction mechanism, a decoding set and an algorithm
The background knowledge is based on proximity relations between terms and between
**5. Conclusion**
The deduction mechanism can be any of the extensions of Datalog. These extensions are the fuzzy Datalog, based on fuzzy logic, the intuitionistic- or interval-value Datalog, based on the suitable logics and bipolar Datalog, which is some kind of coexistence of the former ones.
The semantics of Datalog is a fixed-point semantics, so the algorithm, which connects the two main pillars of the knowledge-base is the generalization of the consequence transformation determining this fixed-point. This transformation is defined on the extended Herbrand base of the knowledge-base, which is generated from the ground terms of knowledge base and its background knowledge.
Applying this transformation it is necessary to compute the uncertainty levels of "synonyms". The proximity-based uncertainty functions can do it, giving uncertainty values from the levels of the original fact and from the proximity values. The set of this kind of functions is the decoding set.
Two possible evaluation strategies were presented as well. One of them evaluates a general knowledgebase with arbitrary proximity-based uncertainty functions and arbitrary implication operators. The other one allows minimum functions only as proximity based uncertainty functions, and the special extension of Gödel operator, but in this case a multivalued unification algorithm can be determined. This strategy is based on the alternating rule-based- and proximity-based unification.
The improvement of this strategy and/or the deduction algorithm and/or the structure of background knowledge is a subject of further investigations.
A well structured multivalued knowledge-base and an efficient evaluating algorithm determining its consequence could be the basis of making decisions based on uncertain information, or it would be useful for handling argumentation or negotiation of agents. An implementation of this model would be an interesting future development as well.
### **6. References**
**0**
**3**
Hashim Habiballa *University of Ostrava Czech Republic*
**Resolution Principle and Fuzzy Logic**
Fuzzy Predicate Logic with Evaluated Syntax (FPL) (Novák, V.) is a well-studied and wide-used logic capable of expressing vagueness. It has a lot of applications based on robust theoretical background. It also requires an efficient formal proof theory. However the most widely applied resolution principle (Duki´c, N.) brings syntactically several obstacles mainly arising from normal form transformation. FPL is associating with even harder problems when trying to use the resolution principle. Solutions to these obstacles based on the non-clausal
In this article it would be presented a natural integration of these two formal logical systems into fully functioning inference system with effective proof search strategies. It leads to the refutational resolution theorem prover for FPL (*RRTPFPL*). Another issue addressed in the paper concerns to the efficiency of presented inference strategies developed originally for the proving system. It is showed their perspectives in combination with standard proof-search strategies. The main problem for the fuzzy logic theorem proving lies in the large amount of possible proofs with different degrees and there is presented an algorithm (Detection of Consequent Formulas - DCF) solving this problem. The algorithm is based on detection of
The article presents the method which is the main point of the work on any automated prover. There is a lot of strategies which makes proofs more efficient when we use refutational proving. We consider well-known strategies - orderings, filtration strategy, set of support etc. One of the most effective strategies is the elimination of consequent formulas. It means the check if a resolvent is not a logical consequence of a formula in set of axioms or a previous resolvent. If such a condition holds it is reasonable to not include the resolvent into the set of resolvents, because if the refutation can be deduced from it, then so it can be deduced from
For the purposes of (*RRTPFPL*) it will be used generalized principle of resolution, which is defined in the research report (Bachmair, L.). There is a propositional form of the rule defined at first and further it is lifted into first-order logic. It is introduced the propositional form of
*F*[*G*] *F*�
[*G*]
*<sup>F</sup>*[*G*/⊥] <sup>∨</sup> *<sup>F</sup>*�[*G*/�] (1)
resolution (Bachmair, L.) were already proposed in (Habiballa, H.).
such redundant formulas (proofs) with different degrees.
the original resolvent, which it implies of.
**General resolution - propositional version**
**2. First-order logic**
the general resolution.
**1. Introduction**
### **Resolution Principle and Fuzzy Logic**
Hashim Habiballa *University of Ostrava Czech Republic*
### **1. Introduction**
54 Fuzzy Logic – Algorithms, Techniques and Implementations
Atanassov, K. (2005). Intuitionistic fuzzy implications and modus ponens. *Notes on*
Atanassov, K. (2006). On some intuitionistic fuzzy implications. *Comptes Rendus de*
Ceri, S.; Gottlob, G. & Tanca, L. (1990). *Logic Programming and Databases*, Springer-Verlag,
Cornelis, C.; Deschrijver, G. & Kerre, E.E. (2004). Implication in intuitionistic fuzzy and
Dubois, D. & Prade, H. (1991). Fuzzy sets in approximate reasoning, Part 1: Inference with possibility distributions, *Fuzzy Sets and Systems,* Vol.40, pp. 143-202. Dubois, D.; Hajek, P. & Prade, H. (2000). Knowledge-Driven versus Data-Driven Logics,
Dubois, D.; Gottwald, S.; Hajek, P.; Kacprzyk, J. & Prade, H. (2005). Terminological
Formato, F.; Gerla, G. & Sessa, M.I. (2000). Similarity-based Unification, *Fundamenta*
Julian-Iranzo, P. & Rubio-Manzano, C. (2009). A declarative semantics for Bousi~Prolog,
Julian-Iranzo, P. & Rubio-Manzano, C. (2010). An Efficient Fuzzy Unification Method and its
Medina, J.; Ojeda-Aciego, M. & Vojtas, P. (2004). Similarity-based unification: a multi-adjoint
Sessa, M. I. (2002). Approximate reasoning by similarity-based SLD resolution, *Theoretical*
Straccia, U. (2008). Managing Uncertainty and Vagueness in Description Logics, Logic
Straccia, U.; Ojeda-Aciego, M. & Damasio, C.V. (2009). On Fixed-points of Multi-valued
Ullman, J.D. (1988). *Principles of Database and Knowledge-base Systems*, Computer Science
Virtanen, H.E. (1994) Fuzzy Unification, *Proc. of IPMU'94, Paris (France)*, pp. 1147–1152. Zadeh, L. A. (1975) The concept of a linguistic variable and its application to approximate
*International Journal of Approximate Reasoning*, Vol.35, pp. 55-95.
*Journal of Logic, Language, and Information*, Vol.9, pp. 65-89.
Lloyd, J.W. (1990). *Foundations of Logic Programming*, Springer-Verlag, Berlin
approach, *Fuzzy Sets and Systems*, Vol.146, No.1, pp. 43-62.
Programs, *SIAM Journal on Computing*, Vol.8, pp. 1881-1911.
*Computer Science*, Vol.275, No.1-2, pp. 389-426.
al. (Eds), pp. 54-103, Springer-Verlag, Berlin
interval-valued fuzzy set theory: construction, classification, application,
difficulties in fuzzy set theory – The case of Intuitionistic Fuzzy Sets, *Fuzzy Sets and*
*PPDP '09: Proceedings of the 11th ACM SIGPLAN conference on Principles and practice*
Implementation into the BousiProlog System, *WCCI2010 IEEE World Congress On*
Programs and Description Logic Programs, In: Reasoning Web 2008, C. Baroglio et
Functions on Complete Lattices and their Application to Generalized Logic
reasoning (I–II-III), *Information Sciences*, Vol.8, pp. 199-249; 301-357; Vol.9 pp. 43-80.
*Intuitionistic Fuzzy Sets,* Vol. 11, pp. 1-5.
*Systems*, Vol.15, pp. 485-491.
*of declarative programming*.
Press, Rockville
*Informaticae*, Vol.41, pp. 393-414.
*Computational Intelligence, Barcelona*.
Berlin
*l'Academie Bulgare des Sciences, Tome*, Vol. 59, pp. 19-24. Atanassov, K. (1999). *Intuitionistic Fuzzy Sets*, Springer-Verlag, Heidelberg
> Fuzzy Predicate Logic with Evaluated Syntax (FPL) (Novák, V.) is a well-studied and wide-used logic capable of expressing vagueness. It has a lot of applications based on robust theoretical background. It also requires an efficient formal proof theory. However the most widely applied resolution principle (Duki´c, N.) brings syntactically several obstacles mainly arising from normal form transformation. FPL is associating with even harder problems when trying to use the resolution principle. Solutions to these obstacles based on the non-clausal resolution (Bachmair, L.) were already proposed in (Habiballa, H.).
> In this article it would be presented a natural integration of these two formal logical systems into fully functioning inference system with effective proof search strategies. It leads to the refutational resolution theorem prover for FPL (*RRTPFPL*). Another issue addressed in the paper concerns to the efficiency of presented inference strategies developed originally for the proving system. It is showed their perspectives in combination with standard proof-search strategies. The main problem for the fuzzy logic theorem proving lies in the large amount of possible proofs with different degrees and there is presented an algorithm (Detection of Consequent Formulas - DCF) solving this problem. The algorithm is based on detection of such redundant formulas (proofs) with different degrees.
> The article presents the method which is the main point of the work on any automated prover. There is a lot of strategies which makes proofs more efficient when we use refutational proving. We consider well-known strategies - orderings, filtration strategy, set of support etc. One of the most effective strategies is the elimination of consequent formulas. It means the check if a resolvent is not a logical consequence of a formula in set of axioms or a previous resolvent. If such a condition holds it is reasonable to not include the resolvent into the set of resolvents, because if the refutation can be deduced from it, then so it can be deduced from the original resolvent, which it implies of.
### **2. First-order logic**
For the purposes of (*RRTPFPL*) it will be used generalized principle of resolution, which is defined in the research report (Bachmair, L.). There is a propositional form of the rule defined at first and further it is lifted into first-order logic. It is introduced the propositional form of the general resolution.
### **General resolution - propositional version**
$$\frac{F[G] \quad F'[G]}{F[G/\perp] \lor F'[G/\top]} \tag{1}$$
*7. F is positive in a clause C if it is an element of C.*
**General ordered resolution with selection** *O*�
**Lemma 1.** *Lifting lemma*
*is an inference in O*�
*is an inference in O*�
*1.* ¬*a* ∨ ¬*b* ∨ *c (axiom), 2. a (axiom), 3. b (axiom)*
*premise respectively)* ⇒ *c*
*positive premise respectively)* ⇒ ¬*b* ∨ *c*
*such that*
of General resolution based on orderings applied to clauses.
*Let M be a set of clauses and K* = *G*(*M*) *(set of ground instances). If*
*SM* (*K*) *then there exist clauses C*�
*i*
level ordered) in formula 1 (it means 1. is a negative premise).
**Example 2.** *General resolution with equivalence*
*4.* [*a* ∧ ⊥] ∨ [*a* ∧ �] *(resolvent from (2), (2) on c)* ⇒ *a 5.* [*a* ∧ ⊥] ∨ [*a* ∧�↔ *b* ∧ *d*] *((2), (1) on c)* ⇒ *a* ↔ *b* ∧ *d*
*6.* ⊥ ∨ [� ↔ *b* ∧ *d*] *((4), (5) on a)* ⇒ *b* ∧ *d 7.* ⊥ ∧ *d* ∨�∧ *d ((6), (6) on b)* ⇒ *d 8. b* ∧⊥∨ *b* ∧ � *((6), (6) on d)* ⇒ *b*
Consider following table showing various cases of resolution on clauses.
*1. a* ∧ *c* ↔ *b* ∧ *d (axiom), 2. a* ∧ *c (axiom), 3.*¬[*b* ∧ *d*] *(axiom) - negated goal*
**Example 1.** *General resolution - polarity based selection*
*<sup>S</sup>* (*M*)*, Ci* = *C*�
Note that this proposition applies both to formulas and clauses and allows us to determine polarity of any subformula in a formula. It is safe to *select any sequence of negative atoms* in a general clause, since a negative atom cannot be false in an interpretation the clause is false. With the notion of the polarity as a selection function there is possible to state another notion
Resolution Principle and Fuzzy Logic 57
*S C*1(*A*1)...*Cn*(*An*) *D*(*A*1, ..., *An*)
where (i) either *A*1, ..., *An* is selected by *S* in *D*, or else *S*(*D*) is empty, *n* = 1, *A*<sup>1</sup> is maximal in
According to the (Bachmair, L.) an inference system based on this rule is refutationally complete. When trying to extend this into the first-order case we to use lifting lemma.
> *C*1...*Cn C*<sup>0</sup> *C*
> > *C*�
*σ.*
*4.* ⊥ ∨ ¬� ∨ ¬*b* ∨ *c (a is a negative atom in (1) - selected in (1) as negative premise, and (2) as*
*5.* ⊥ ∨ ¬� ∨ *c (b is a negative atom in (4) - selected in (4) as negative premise, and (3) as positive*
In the example we used the notion of polarity as a selection function. For example in the line 4 we select the atom a upon negative polarity (according the proposition criteria 1, 3 and 2 -
Further we can observe the behavior of the rule within the frame of clausal form resolution.
*<sup>i</sup> in M, a clause C*�
*, and a ground substitution σ*
*D*, (ii) each atom *Ai* is maximal in *Ci*, and (iii) no clause *Ci* contains a selected atom.
*C*� 1...*C*� *<sup>n</sup> C*� 0
*σ, and C* = *C*�
*<sup>C</sup>*1(⊥)...*Cn*(⊥) *<sup>D</sup>*(�, ..., �) (2)
where the propositional logic formulas *F* and *F*� are the premises of inference and *G* is an occurrence of a subformula of both *F* and *F*� . The expression *F*[*G*/⊥] ∨ *F*� [*G*/�] is the resolvent of the premises on *G*. Every occurrence of G is replaced by false in the first formula and by true in the second one. It is also called F the positive, F' the negative premise, and G the resolved subformula.
The proof of the soundness of the rule is similar to clausal resolution rule proof. Suppose the Interpretation I in which both premises are valid. In I, G is either true or false. If G (¬*G*) is true in I, so is *F*� [*G*/�] (*F*[*G*/⊥]).
Revised version of the paper which forms the core of the handbook (Bachmair, L.) is closely related with notion of selection functions and ordering constraints. By a selection functions it is meant a mapping *S* that assigns to each clause *C* a (possibly empty) multiset *S*(*C*) of negative literals in *C*. In other words, the function S selects (a possibly empty) negative subclause of *C*. We say that an atom *A*, or a literal ¬*A*, is selected by *S* if ¬*A* occurs in *S*(*C*). There are no selected atoms or literals if *S*(*C*) is empty. Lexicographic path ordering can be used as an usual ordering over a total precedence. But in this case the ordering is admissible if predicate symbols have higher precedence than logical symbols and the constants � and ⊥ are smaller than the other logical symbols. It means the ordering is following *A* �≡�⊃� ¬ � ∨ � ∧ � � � ⊥. The handbook also addresses another key issues for automated theorem proving - the efficiency of the proof search. This efficiency is closely related with the notion of *redundancy*.
If we want to generalize the notion of resolution and lift it into first-order case we have to define first the notion of selection function for general clauses. General clauses are multisets of arbitrary quantifier-free formulas, denoting the disjunction of their elements. Note that we can also work with a special case of such general clause with one element, which yields to a standard quantifier-free formula of first-order logic. A (general selection) function is a mapping *S* that assigns to each general clause *C* a (possibly empty) set *C* of non-empty sequences of (distinct) atoms in *C* such that either *S*(*C*) is empty or else, for all interpretations *I* in which *C* is false, there exists a sequence *A*1, ..., *Ak* in *S*(*C*), all atoms of which are true in *I*. A sequence *A*1, ..., *Ak* in *S*(*C*) is said to be *selected* (by *S*).
We have to define the notion of polarity for these reasons according to the handbook (Bachmair, L.). It is based on the following assumption that a subformula *F*� in *E*[*F*� ] is *positive* (resp. *negative*), if *E*[*F*� /�] (resp. *E*[*F*� /⊥]) is a tautology. Thus, if *F*� is *positive* (resp. *negative*) in *E*, *F*� (resp. ¬*F*� ) logically implies *E*. Even it should seem that determining of the polarity of any subformula is NP-complete (hard) problem, we can use syntactic criteria for this computation. In this case the complexity of the algorithm is linear (note that we base our theory on similar syntactic criteria below - structural notions definition).
### **Proposition 1.** *Polarity criteria*
2 Will-be-set-by-IN-TECH
where the propositional logic formulas *F* and *F*� are the premises of inference and *G* is an
resolvent of the premises on *G*. Every occurrence of G is replaced by false in the first formula and by true in the second one. It is also called F the positive, F' the negative premise, and G
The proof of the soundness of the rule is similar to clausal resolution rule proof. Suppose the Interpretation I in which both premises are valid. In I, G is either true or false. If G (¬*G*) is
Revised version of the paper which forms the core of the handbook (Bachmair, L.) is closely related with notion of selection functions and ordering constraints. By a selection functions it is meant a mapping *S* that assigns to each clause *C* a (possibly empty) multiset *S*(*C*) of negative literals in *C*. In other words, the function S selects (a possibly empty) negative subclause of *C*. We say that an atom *A*, or a literal ¬*A*, is selected by *S* if ¬*A* occurs in *S*(*C*). There are no selected atoms or literals if *S*(*C*) is empty. Lexicographic path ordering can be used as an usual ordering over a total precedence. But in this case the ordering is admissible if predicate symbols have higher precedence than logical symbols and the constants � and ⊥ are smaller than the other logical symbols. It means the ordering is following *A* �≡�⊃� ¬ � ∨ � ∧ � � � ⊥. The handbook also addresses another key issues for automated theorem proving - the efficiency of the proof search. This efficiency is
If we want to generalize the notion of resolution and lift it into first-order case we have to define first the notion of selection function for general clauses. General clauses are multisets of arbitrary quantifier-free formulas, denoting the disjunction of their elements. Note that we can also work with a special case of such general clause with one element, which yields to a standard quantifier-free formula of first-order logic. A (general selection) function is a mapping *S* that assigns to each general clause *C* a (possibly empty) set *C* of non-empty sequences of (distinct) atoms in *C* such that either *S*(*C*) is empty or else, for all interpretations *I* in which *C* is false, there exists a sequence *A*1, ..., *Ak* in *S*(*C*), all atoms of which are true in
We have to define the notion of polarity for these reasons according to the handbook (Bachmair, L.). It is based on the following assumption that a subformula *F*� in *E*[*F*�
polarity of any subformula is NP-complete (hard) problem, we can use syntactic criteria for this computation. In this case the complexity of the algorithm is linear (note that we base our
*2. If* ¬*G is a positive (resp. negative) subformula of F, then G is a negative (resp. positive) subformula*
*5. If G* → *H is a positive subformula of F, then G is a negative subformula and H is a positive*
*3. If G* ∨ *H is a positive subformula of F, then G and H are both positive subformulas of F. 4. If G* ∧ *H is a negative subformula of F, then G and H are both negative subformulas of F.*
/�] (resp. *E*[*F*�
theory on similar syntactic criteria below - structural notions definition).
*6. If G* → ⊥ *is a negative subformula of F, then G is a positive subformula of F.*
. The expression *F*[*G*/⊥] ∨ *F*�
[*G*/�] is the
] is
/⊥]) is a tautology. Thus, if *F*� is *positive* (resp.
) logically implies *E*. Even it should seem that determining of the
occurrence of a subformula of both *F* and *F*�
[*G*/�] (*F*[*G*/⊥]).
closely related with the notion of *redundancy*.
*positive* (resp. *negative*), if *E*[*F*�
**Proposition 1.** *Polarity criteria 1. F is a positive subformula of F.*
*of F.*
*subformula of F.*
*negative*) in *E*, *F*� (resp. ¬*F*�
*I*. A sequence *A*1, ..., *Ak* in *S*(*C*) is said to be *selected* (by *S*).
the resolved subformula.
true in I, so is *F*�
### *7. F is positive in a clause C if it is an element of C.*
Note that this proposition applies both to formulas and clauses and allows us to determine polarity of any subformula in a formula. It is safe to *select any sequence of negative atoms* in a general clause, since a negative atom cannot be false in an interpretation the clause is false. With the notion of the polarity as a selection function there is possible to state another notion of General resolution based on orderings applied to clauses.
#### **General ordered resolution with selection** *O*� *S*
$$\frac{\mathbb{C}\_1(A\_1)...\mathbb{C}\_n(A\_n) \quad D(A\_1,...,A\_n)}{\mathbb{C}\_1(\bot)...\mathbb{C}\_n(\bot) \quad D(\top,...,\top)}\tag{2}$$
where (i) either *A*1, ..., *An* is selected by *S* in *D*, or else *S*(*D*) is empty, *n* = 1, *A*<sup>1</sup> is maximal in *D*, (ii) each atom *Ai* is maximal in *Ci*, and (iii) no clause *Ci* contains a selected atom.
According to the (Bachmair, L.) an inference system based on this rule is refutationally complete. When trying to extend this into the first-order case we to use lifting lemma.
**Lemma 1.** *Lifting lemma*
*Let M be a set of clauses and K* = *G*(*M*) *(set of ground instances). If*
$$\frac{C\_1...C\_n \quad C\_0}{C}$$
*is an inference in O*� *SM* (*K*) *then there exist clauses C*� *<sup>i</sup> in M, a clause C*� *, and a ground substitution σ such that*
$$\frac{C\_1'...C\_n'\quad C\_0'}{C'}$$
*is an inference in O*� *<sup>S</sup>* (*M*)*, Ci* = *C*� *i σ, and C* = *C*� *σ.*
#### **Example 1.** *General resolution - polarity based selection*
*1.* ¬*a* ∨ ¬*b* ∨ *c (axiom),*
*2. a (axiom), 3. b (axiom)*
*4.* ⊥ ∨ ¬� ∨ ¬*b* ∨ *c (a is a negative atom in (1) - selected in (1) as negative premise, and (2) as positive premise respectively)* ⇒ ¬*b* ∨ *c*
*5.* ⊥ ∨ ¬� ∨ *c (b is a negative atom in (4) - selected in (4) as negative premise, and (3) as positive premise respectively)* ⇒ *c*
In the example we used the notion of polarity as a selection function. For example in the line 4 we select the atom a upon negative polarity (according the proposition criteria 1, 3 and 2 level ordered) in formula 1 (it means 1. is a negative premise).
Further we can observe the behavior of the rule within the frame of clausal form resolution. Consider following table showing various cases of resolution on clauses.
#### **Example 2.** *General resolution with equivalence*
*1. a* ∧ *c* ↔ *b* ∧ *d (axiom), 2. a* ∧ *c (axiom), 3.*¬[*b* ∧ *d*] *(axiom) - negated goal 4.* [*a* ∧ ⊥] ∨ [*a* ∧ �] *(resolvent from (2), (2) on c)* ⇒ *a 5.* [*a* ∧ ⊥] ∨ [*a* ∧�↔ *b* ∧ *d*] *((2), (1) on c)* ⇒ *a* ↔ *b* ∧ *d 6.* ⊥ ∨ [� ↔ *b* ∧ *d*] *((4), (5) on a)* ⇒ *b* ∧ *d 7.* ⊥ ∧ *d* ∨�∧ *d ((6), (6) on b)* ⇒ *d 8. b* ∧⊥∨ *b* ∧ � *((6), (6) on d)* ⇒ *b*
the ordering with respect to the scope of variables (which is also essential for skolemization simulation - unification is restricted for existential variables). Polarity enables to decide the global meaning of a variable (e.g. globally an existential variable is universal if its quantification subformula has negative polarity). Sound unification requires further definitions on variable quantification. We will introduce notions of the corresponding quantifier for a variable occurrence, substitution mapping and significance mapping (we have to distinguish between original variables occurring in special axioms and newly introduced
Resolution Principle and Fuzzy Logic 59
*Let F be a formula of FOL, G* = *p*(*t*1, ..., *tn*) ∈ *Sub*∗(*F*) *atom in F and α a variable occurring in ti. Variable mappings Qnt(quantifier assignment), Sbt (variable substitution) and Sig(significance) are*
*A variable α occurring in F* ∈ *LAx* ∪ *SAx is significant w.r.t. existential substitution, Sig*(*α*) = 1 *iff*
Example: ∀*x*(∀*xA*(*x*) → *B*(*x*)) - *Qnt*(*x*) = ∀*xA*(*x*), for *x* in *A*(*x*) and *Qnt*(*x*) = ∀*x*(∀*xA*(*x*) →
Note that with Qnt mapping (assignment of first name matching quantifier variable in a formula hierarchy from bottom) we are able to distinguish between variables of the same name and there is no need to rename any variable. Sbt mapping holds substituted terms in a quantifier and there is no need to rewrite all occurrences of a variable when working with this mapping within unification. It is also clear that if *Qnt*(*α*) = ∅ then *α* is a free variable. These variables could be simply avoided by introducing new universal quantifiers to F. Significance mapping is important for differentiating between original formula universal variables and newly introduced ones during proof search (an existential variable can't be bounded with it). Before we can introduce the standard unification algorithm, we should formulate the notion of global universal and global existential variable (it simulates conversion into prenex normal
Example: *F* = ∀*y*(∀*xA*(*x*) → *B*(*y*)) - *x* is a global existential variable, *y* is a global universal
It is clear w.r.t. skolemization technique that an existential variable can be substituted into an universal one only if all global universal variables over the scope of the existential one have been already substituted by a term. Skolem functors function in the same way. Now we can define the most general unification algorithm based on recursive conditions (extended
*Let F be a formula without free variables and α be a variable occurrence in a term of F.*
*1. <sup>α</sup> is a global universal variable* (*<sup>α</sup>* ∈ *Var*∀(*F*)) *iff* (*Qnt*(*α*) = ∀*α<sup>H</sup>* ∧*Pol*(*Qnt*(*α*)) = 1) *or* (*Qnt*(*α*) = ∃*αH* ∧ *Pol*(*Qnt*(*α*)) = −1) *2. <sup>α</sup> is a global existential variable* (*<sup>α</sup>* ∈ *Var*∃(*F*)) *iff* (*Qnt*(*α*) = ∃*α<sup>H</sup>* ∧*Pol*(*Qnt*(*α*)) = 1) *or* (*Qnt*(*α*) = ∀*αH* ∧ *Pol*(*Qnt*(*α*)) = −1) *Var*∀(*F*) *and Var*∃(*F*) *are sets of global universal and existential variables.*
� *.*
**Definition 2.** *Variable assignment, substitution and significance*
] *is a substitution of term t*� *into α in F* ⇒ *Sbt*(*α*) = *t*
*Qnt*(*α*) = *QαH*, *whereQ* = ∃ ∨ *Q* = ∀, *H*, *I* ∈ *Sub*∗(*F*), *QαH* ∈ *Sup*∗(*G*),
ones in the proof sequence).
∀*QαI* ∈ *Sup*∗(*G*) ⇒ *Lev*(*QαI*) < *Lev*(*QαH*)*.*
*variable is significant, Sig*(*α*) = 0 *otherwise.*
**Definition 3.** *Global quantification*
unification in contrast to standard MGU).
*defined as follows:*
*B*(*x*)), for *x* in *B*(*x*).
*F*[*α*/*t* �
form).
variable.
Table 1. Clausal resolution in the context of the non-clausal resolution
*9.* ⊥∨¬[� ∧ *d*] *((8), (3) on b)* ⇒ ¬*d 10.* ⊥ ∨ ¬� *((7), (9) on d)* ⇒ ⊥ *(refutation)*
When trying to refine the general resolution rule for fuzzy predicate logic, it is important to devise a sound and complete unification algorithm. Standard unification algorithms require variables to be treated only as universally quantified ones. We will present a more general unification algorithm, which can deal with existentially quantified variables without the need for those variables be eliminated by skolemization. It should be stated that the following unification process does not allow an occurrence of the equivalence connective. It is needed to remove equivalence by rewrite rule: *A* ↔ *B* ⇔ [*A* → *B*] ∧ [*B* → *A*].
We assume that the language and semantics of FOL is standard. We use terms - individuals (*a*, *b*, *c*, ...), functions (with n arguments) (*f* , *g*, *h*, ...), variables (*X*,*Y*, *Z*, ...), predicates(with n arguments) (*p*, *q*,*r*, ...), logical connectives (∧, ∨, →, ¬), quantifiers (∃, ∀) and logical constants (⊥, �). We also work with standard notions of logical and special axioms (sets LAx, SAx), logical consequence, consistency etc. as they are used in mathematical logic.
### **Definition 1.** *Structural notions of a FOL formula*
*Let F be a formula of FOL then the structural mappings Sub (subformula), Sup (superformula), Pol (polarity) and Lev (level) are defined as follows:*
*Sup*(*F*) = ∅ ⇒ *Lev*(*F*) = 0, *Pol*(*F*) = 1*,*
*Sup*(*F*) �= ∅ ⇒ *Lev*(*F*) = *Lev*(*Sup*(*F*)) + 1
*For mappings Sub and Sup reflexive and transitive closures Sub*∗ *and Sup*∗ *are defined recursively as follows:*
*1. Sub*∗(*F*) ⊇ {*F*}*, Sup*∗(*F*) ⊇ {*F*} *2. Sub*∗(*F*) ⊇ {*H*|*G* ∈ *Sub*∗(*F*) ∧ *H* ∈ *Sub*(*G*)}*, Sup*∗(*F*) ⊇ {*H*|*G* ∈ *Sup*∗(*F*) ∧ *H* ∈ *Sup*(*G*)}
Example: *A* → *B* - *Pol*(*A*) = −1, *Pol*(*B*) = 1, *Lev*(*A*) = 1
These structural mappings provide framework for assignment of quantifiers to variable occurrences. It is needed for the correct simulation of skolemization (the information about a variable quantification in the prenex form). Subformula and superformula mappings and its closures encapsulate essential hierarchical information of a formula structure. Level gives 4 Will-be-set-by-IN-TECH
When trying to refine the general resolution rule for fuzzy predicate logic, it is important to devise a sound and complete unification algorithm. Standard unification algorithms require variables to be treated only as universally quantified ones. We will present a more general unification algorithm, which can deal with existentially quantified variables without the need for those variables be eliminated by skolemization. It should be stated that the following unification process does not allow an occurrence of the equivalence connective. It is needed
We assume that the language and semantics of FOL is standard. We use terms - individuals (*a*, *b*, *c*, ...), functions (with n arguments) (*f* , *g*, *h*, ...), variables (*X*,*Y*, *Z*, ...), predicates(with n arguments) (*p*, *q*,*r*, ...), logical connectives (∧, ∨, →, ¬), quantifiers (∃, ∀) and logical constants (⊥, �). We also work with standard notions of logical and special axioms (sets LAx, SAx),
*Let F be a formula of FOL then the structural mappings Sub (subformula), Sup (superformula), Pol*
*F* = *G* ∧ *H or F* = *G* ∨ *H Sub*(*F*) = {*G*, *H*}*, Sup*(*G*) = *F, Sup*(*H*) = *F*
*F* = *G* → *H Sub*(*F*) = {*G*, *H*}*, Sup*(*G*) = *F, Sup*(*H*) = *F*
*Pol*(*G*) = −*Pol*(*F*)
*For mappings Sub and Sup reflexive and transitive closures Sub*∗ *and Sup*∗ *are defined recursively as*
*2. Sub*∗(*F*) ⊇ {*H*|*G* ∈ *Sub*∗(*F*) ∧ *H* ∈ *Sub*(*G*)}*, Sup*∗(*F*) ⊇ {*H*|*G* ∈ *Sup*∗(*F*) ∧ *H* ∈ *Sup*(*G*)}
These structural mappings provide framework for assignment of quantifiers to variable occurrences. It is needed for the correct simulation of skolemization (the information about a variable quantification in the prenex form). Subformula and superformula mappings and its closures encapsulate essential hierarchical information of a formula structure. Level gives
*Pol*(*G*) = *Pol*(*F*)*, Pol*(*H*) = *Pol*(*F*)
*Pol*(*G*) = −*Pol*(*F*)*, Pol*(*H*) = *Pol*(*F*)
Premise1 Premise2 Resolvent Simplified Comments *a* ∨ *b b* ∨ *c* (*a* ∨ ⊥) ∨ (� ∨ *c*) � no compl. pair *a* ∨ ¬*b b* ∨ *c* (*a* ∨ �) ∨ (� ∨ *c*) � redundant inference *a* ∨ *b* ¬*b* ∨ *c* (*a* ∨ ⊥) ∨ (⊥ ∨ *c*) *a* ∨ *c* clausal resolution *a* ∨ ¬*b* ¬*b* ∨ *c* (*a* ∨ �) ∨ (⊥ ∨ *c*) � no compl. pair
Table 1. Clausal resolution in the context of the non-clausal resolution
to remove equivalence by rewrite rule: *A* ↔ *B* ⇔ [*A* → *B*] ∧ [*B* → *A*].
**Definition 1.** *Structural notions of a FOL formula*
*(polarity) and Lev (level) are defined as follows:*
*Sup*(*F*) = ∅ ⇒ *Lev*(*F*) = 0, *Pol*(*F*) = 1*, Sup*(*F*) �= ∅ ⇒ *Lev*(*F*) = *Lev*(*Sup*(*F*)) + 1
*1. Sub*∗(*F*) ⊇ {*F*}*, Sup*∗(*F*) ⊇ {*F*}
*follows:*
logical consequence, consistency etc. as they are used in mathematical logic.
*F* = ¬*G Sub*(*F*) = {*G*}*, Sup*(*G*) = *F*
*F* = ∃*αG or F* = ∀*αG Sub*(*F*) = {*G*}*, Sup*(*G*) = *F*
*(α is a variable) Pol*(*G*) = *Pol*(*F*)
Example: *A* → *B* - *Pol*(*A*) = −1, *Pol*(*B*) = 1, *Lev*(*A*) = 1
*9.* ⊥∨¬[� ∧ *d*] *((8), (3) on b)* ⇒ ¬*d 10.* ⊥ ∨ ¬� *((7), (9) on d)* ⇒ ⊥ *(refutation)* the ordering with respect to the scope of variables (which is also essential for skolemization simulation - unification is restricted for existential variables). Polarity enables to decide the global meaning of a variable (e.g. globally an existential variable is universal if its quantification subformula has negative polarity). Sound unification requires further definitions on variable quantification. We will introduce notions of the corresponding quantifier for a variable occurrence, substitution mapping and significance mapping (we have to distinguish between original variables occurring in special axioms and newly introduced ones in the proof sequence).
### **Definition 2.** *Variable assignment, substitution and significance*
*Let F be a formula of FOL, G* = *p*(*t*1, ..., *tn*) ∈ *Sub*∗(*F*) *atom in F and α a variable occurring in ti. Variable mappings Qnt(quantifier assignment), Sbt (variable substitution) and Sig(significance) are defined as follows:*
*Qnt*(*α*) = *QαH*, *whereQ* = ∃ ∨ *Q* = ∀, *H*, *I* ∈ *Sub*∗(*F*), *QαH* ∈ *Sup*∗(*G*), ∀*QαI* ∈ *Sup*∗(*G*) ⇒ *Lev*(*QαI*) < *Lev*(*QαH*)*. F*[*α*/*t* � ] *is a substitution of term t*� *into α in F* ⇒ *Sbt*(*α*) = *t* � *. A variable α occurring in F* ∈ *LAx* ∪ *SAx is significant w.r.t. existential substitution, Sig*(*α*) = 1 *iff variable is significant, Sig*(*α*) = 0 *otherwise.*
Example: ∀*x*(∀*xA*(*x*) → *B*(*x*)) - *Qnt*(*x*) = ∀*xA*(*x*), for *x* in *A*(*x*) and *Qnt*(*x*) = ∀*x*(∀*xA*(*x*) → *B*(*x*)), for *x* in *B*(*x*).
Note that with Qnt mapping (assignment of first name matching quantifier variable in a formula hierarchy from bottom) we are able to distinguish between variables of the same name and there is no need to rename any variable. Sbt mapping holds substituted terms in a quantifier and there is no need to rewrite all occurrences of a variable when working with this mapping within unification. It is also clear that if *Qnt*(*α*) = ∅ then *α* is a free variable. These variables could be simply avoided by introducing new universal quantifiers to F. Significance mapping is important for differentiating between original formula universal variables and newly introduced ones during proof search (an existential variable can't be bounded with it).
Before we can introduce the standard unification algorithm, we should formulate the notion of global universal and global existential variable (it simulates conversion into prenex normal form).
#### **Definition 3.** *Global quantification*
*Let F be a formula without free variables and α be a variable occurrence in a term of F.*
*Var*∀(*F*) *and Var*∃(*F*) *are sets of global universal and existential variables.*
Example: *F* = ∀*y*(∀*xA*(*x*) → *B*(*y*)) - *x* is a global existential variable, *y* is a global universal variable.
It is clear w.r.t. skolemization technique that an existential variable can be substituted into an universal one only if all global universal variables over the scope of the existential one have been already substituted by a term. Skolem functors function in the same way. Now we can define the most general unification algorithm based on recursive conditions (extended unification in contrast to standard MGU).
**Definition 5.** *General resolution for first-order logic* (*GRFOL*)
<sup>1</sup>,..., *G*�
⇒ *Sig*(*α*) = 1 *in F or F*� *iff Sig*(*α*) = 1 *in Fσ*[*G*/⊥] ∨ *F*�
*A* = {*G*1,..., *Gk*, *G*�
*identical.*
*resolvent of the premises on G.*
refutational theorem prover for FOL.
**Example 3.** *Variable Unification Restriction*
*F*<sup>0</sup> : ∀*X*∃*Yp*(*X*,*Y*)*. F*1(¬*query*) : ¬∃*Y*∀*X p*(*X*,*Y*)*.*
*simply unifiable since the variables are the same. Non-trivial cases:*[*F*1&*F*0] *: no resolution is possible.*
*F0 :* ∃*Y*∀*X p*(*X*,*Y*)*. F1 (*¬*query) :* ¬∀*X*∃*Yp*(*X*,*Y*) *In this case we can simply derive a refutation:*
*R*[*F*1&*F*0] *:* ⊥ ∨ ¬�(*ref utation*)
[*F*0&*F*1] *: no resolution is possible (the same reason as above). No refutation could be derived from F*<sup>0</sup> *and F*<sup>1</sup> *due to VUR.*
*Further we would like to prove* ∃*Y*∀*X p*(*X*,*Y*) � ∀*X*∃*Yp*(*X*,*Y*)*.*
*variable over the scope of X in F*1*; Sbt*(*X*) = *X and Sbt*(*Y*) = *Y.*
of conjunction, disjunction etc. to be bound with Łukasiewicz operators.
**3. Fuzzy predicate logic and refutational proof**
*F*[*G*1, , ..., *Gk*] *F*�
*negative premise, G represents an occurrence of an atom. The expression Fσ*[*G*/⊥] ∨ *F*�
*where σ* = *MGU*(*A*) *is the most general unifier (MGU) of the set of the atoms*
[*G*� <sup>1</sup>, ..., *G*� *n*]
*<sup>n</sup>*} *, G* = *G*1*σ. For every variable α in F or F*�
Resolution Principle and Fuzzy Logic 61
Note that with Qnt mapping we are able to distinguish variables not only by its name (which may not be unique) but also with this mapping (it is unique). Sig property enables to separate variables, which were not originally in the scope of an existential variable. When utilizing the rule it should be set the Sig mapping for every variable in axioms and negated goal to one. We present a very simple example of existential variable unification before we introduce the
*We would try to prove if* ∀*X*∃*Yp*(*X*,*Y*) � ∃*Y*∀*X p*(*X*,*Y*)*? We will use refutational proving and therefore we will construct a special axiom from the first formula and negation of the second formula:*
*There are 2 trivial and 2 non-trivial combinations how to resolve F*<sup>0</sup> *and F*<sup>1</sup> *(combinations with the same formula as the positive and the negative premise could not lead to refutation since they are consistent): Trivial cases: R*[*F*1&*F*1] : ⊥∨� *and R*[*F*0&*F*0] : ⊥∨�*. Both of them lead to* � *and the atoms are*
*<sup>Y</sup>* ∈ *Var*∀(*F*1) *and Y* ∈ *Var*∃(*F*0) *can't unify since VUR for* (*Y*,*Y*) *does not hold - there is a variable <sup>X</sup>* ∈ *Sup*∗(*Qnt*(*Y*))*(over the scope), X* ∈ *Var*∀(*F*0), *Sbt*(*X*) = <sup>∅</sup>*); the case with variable X is*
*<sup>X</sup>* ∈ *Var*∀(*F*0) *and X* ∈ *Var*∃(*F*1) *can unify since VUR for* (*X*, *<sup>X</sup>*) *holds - there is no global universal*
The fuzzy predicate logic with evaluated syntax is a flexible and fully complete formalism, which will be used for the below presented extension (Novák, V.). In order to use an efficient form of the resolution principle we have to extend the standard notion of a proof (provability value and degree) with the notion of refutational proof (refutation degree). Propositonal version of the fuzzy resolution principle has been already presented in (Habiballa, H.). We suppose that set of truth values is Łukasiewicz algebra. Therefore we assume standard notions
*<sup>F</sup>σ*[*G*/⊥] <sup>∨</sup> *<sup>F</sup>*�*σ*[*G*/�] (3)
*,* (*Sbt*(*γ*) = *α*) ∩ *σ* = ∅
*σ*[*G*/�] *is the*
*σ*[*G*/�]*. F is called positive and F' is called*
### **Definition 4.** *Most general unifier algorithm*
*Let G* = *p*(*t*1, ..., *tn*) *and G*� = *r*(*u*1, ..., *un*) *be atoms. Most general unifier (substitution mapping) MGU(G, G') = σ is obtained by following atom and term unification steps or the algorithm returns fail-state for unification. For the purposes of the algorithm we define the Variable Unification Restriction (VUR).*
### *Variable Unification Restriction*
*Let F*<sup>1</sup> *be a formula and α be a variable occurring in F*1*, F*<sup>2</sup> *be a formula, t be a term occurring in F*<sup>2</sup> *and β be a variable occurring in F*2*. Variable Unification Restriction (VUR) for (α,t) holds if one of the conditions 1. and 2. holds:*
#### *Atom unification*
#### *Term unification* (*t* � , *u*� )
*MGU*(*A*) = *σ for a set of atoms A* = {*G*1,..., *Gk*} *is computed by the atom unification for* (*G*1, *Gi*), *σ<sup>i</sup>* = *MGU*(*G*1, *Gi*), ∀*i*, *σ*<sup>0</sup> = ∅*, where before every atom unification* (*G*1, *Gi*)*, σ is set to σi*−1*.*
With above defined notions it is simple to state the general resolution rule for FOL (without the equivalence connective). It conforms to the definition from (Bachmair, L.).
6 Will-be-set-by-IN-TECH
*Let G* = *p*(*t*1, ..., *tn*) *and G*� = *r*(*u*1, ..., *un*) *be atoms. Most general unifier (substitution mapping) MGU(G, G') = σ is obtained by following atom and term unification steps or the algorithm returns fail-state for unification. For the purposes of the algorithm we define the Variable Unification*
*Let F*<sup>1</sup> *be a formula and α be a variable occurring in F*1*, F*<sup>2</sup> *be a formula, t be a term occurring in F*<sup>2</sup> *and β be a variable occurring in F*2*. Variable Unification Restriction (VUR) for (α,t) holds if one of the*
*1. α is a global universal variable and t* �= *β, where β is a global existential variable and α not*
*2. α is a global universal variable and t* = *β, where β is a global existential variable and* ∀*F* ∈ *Sup*∗(*Qnt*(*β*))*, F* = *QγG, Q* ∈ {∀, ∃}*, γ is a global universal variable, Sig*(*γ*) = 1 ⇒
*2. if n* > 0 *and p* = *r then perform term unification for pairs* (*t*1, *u*1),...,(*tn*, *un*)*; If for every pair*
*4. if t*� = *a, u*� = *b are individual constants anda=b then for (t',u') unifier exists (success-state).*
) = *σ obtained during term unification (success state).*
) ∈ *σ then perform term unification for* (*v*�
) ∈ *σ then perform term unification for* (*t*
) *holds then unifier exists for* (*t*
) *holds then unifier exists for* (*t*
� <sup>1</sup>, *u*�
) *(success-state for an already substituted variable).*
) *(success-state for an already substituted variable).*
<sup>1</sup>)*, ...,* (*t* � *<sup>n</sup>*, *u*�
*<sup>n</sup>*) *are function symbols with arguments and f* = *g then*
� , *u*�
, *u*�
� , *v*�
*<sup>n</sup>*) *(success-state).*
� , *u*�
� , *u*� ) *(success-state)*
)*; The unifier*
)*; The unifier*
) *holds and σ* =
) *holds and σ* =
) ∈ *σ, r*� *is a term (existential substitution).*
*1. if u*� = *α, t*� = *β are variables and Qnt*(*α*) = *Qnt*(*β*) *then unifier exists for* (*t*
, *u*�
<sup>1</sup>, ..., *u*�
) *exists iff unifier exists for every pair* (*t*
� , *u*�
the equivalence connective). It conforms to the definition from (Bachmair, L.).
, *t* �
*MGU*(*A*) = *σ for a set of atoms A* = {*G*1,..., *Gk*} *is computed by the atom unification for* (*G*1, *Gi*), *σ<sup>i</sup>* = *MGU*(*G*1, *Gi*), ∀*i*, *σ*<sup>0</sup> = ∅*, where before every atom unification* (*G*1, *Gi*)*, σ is set*
With above defined notions it is simple to state the general resolution rule for FOL (without
� , *v*�
*1. if n* = 0 *and p* = *r then σ* = ∅ *and the unifier exists (success-state).*
**Definition 4.** *Most general unifier algorithm*
*occurring in t (non-existential substitution)*
*3. In any other case unifier does not exist (fail-state).*
) *exists iff it exists for* (*v*�
) *exists iff it exists for* (*t*
*<sup>m</sup>*)*, u*� = *g*(*u*�
) *(success-state).*
) *(success-state). 8. In any other case unifier does not exist (fail-state).*
*Restriction (VUR).*
*conditions 1. and 2. holds:*
(*Sbt*(*γ*) = *r*�
*Atom unification*
*Term unification* (*t*
*for* (*t* � , *u*�
*for* (*t* � , *u*�
*5. if t*� = *f*(*t*
*to σi*−1*.*
*unifier for* (*t*
*σ* ∪ (*Sbt*(*α*) = *u*�
*σ* ∪ (*Sbt*(*α*) = *t*
� <sup>1</sup>, ..., *t* �
� , *u*�
*Variable Unification Restriction*
*unifier exists then MGU*(*G*, *G*�
� , *u*� )
*(occurrence of the same variable). 2. if t*� = *α is a variable and* (*Sbt*(*α*) = *v*�
*3. if u*� = *α is a variable and* (*Sbt*(*α*) = *v*�
*6. if t*� = *α is a variable and VUR for* (*t*
*7. if u*� = *α is a variable and VUR for* (*u*�
�
**Definition 5.** *General resolution for first-order logic* (*GRFOL*)
$$\frac{F[\mathcal{G}\_{1\prime}, \dots, \mathcal{G}\_{k}]}{F\sigma[\mathcal{G}/\perp] \lor F'\sigma[\mathcal{G}/\top]}\tag{3}$$
*where σ* = *MGU*(*A*) *is the most general unifier (MGU) of the set of the atoms A* = {*G*1,..., *Gk*, *G*� <sup>1</sup>,..., *G*� *<sup>n</sup>*} *, G* = *G*1*σ. For every variable α in F or F*� *,* (*Sbt*(*γ*) = *α*) ∩ *σ* = ∅ ⇒ *Sig*(*α*) = 1 *in F or F*� *iff Sig*(*α*) = 1 *in Fσ*[*G*/⊥] ∨ *F*� *σ*[*G*/�]*. F is called positive and F' is called negative premise, G represents an occurrence of an atom. The expression Fσ*[*G*/⊥] ∨ *F*� *σ*[*G*/�] *is the resolvent of the premises on G.*
Note that with Qnt mapping we are able to distinguish variables not only by its name (which may not be unique) but also with this mapping (it is unique). Sig property enables to separate variables, which were not originally in the scope of an existential variable. When utilizing the rule it should be set the Sig mapping for every variable in axioms and negated goal to one. We present a very simple example of existential variable unification before we introduce the refutational theorem prover for FOL.
### **Example 3.** *Variable Unification Restriction*
*We would try to prove if* ∀*X*∃*Yp*(*X*,*Y*) � ∃*Y*∀*X p*(*X*,*Y*)*? We will use refutational proving and therefore we will construct a special axiom from the first formula and negation of the second formula: F*<sup>0</sup> : ∀*X*∃*Yp*(*X*,*Y*)*. F*1(¬*query*) : ¬∃*Y*∀*X p*(*X*,*Y*)*.*
*There are 2 trivial and 2 non-trivial combinations how to resolve F*<sup>0</sup> *and F*<sup>1</sup> *(combinations with the same formula as the positive and the negative premise could not lead to refutation since they are consistent): Trivial cases: R*[*F*1&*F*1] : ⊥∨� *and R*[*F*0&*F*0] : ⊥∨�*. Both of them lead to* � *and the atoms are simply unifiable since the variables are the same.*
*Non-trivial cases:*[*F*1&*F*0] *: no resolution is possible.*
*<sup>Y</sup>* ∈ *Var*∀(*F*1) *and Y* ∈ *Var*∃(*F*0) *can't unify since VUR for* (*Y*,*Y*) *does not hold - there is a variable <sup>X</sup>* ∈ *Sup*∗(*Qnt*(*Y*))*(over the scope), X* ∈ *Var*∀(*F*0), *Sbt*(*X*) = <sup>∅</sup>*); the case with variable X is identical.*
[*F*0&*F*1] *: no resolution is possible (the same reason as above). No refutation could be derived from F*<sup>0</sup> *and F*<sup>1</sup> *due to VUR.*
*Further we would like to prove* ∃*Y*∀*X p*(*X*,*Y*) � ∀*X*∃*Yp*(*X*,*Y*)*. F0 :* ∃*Y*∀*X p*(*X*,*Y*)*. F1 (*¬*query) :* ¬∀*X*∃*Yp*(*X*,*Y*) *In this case we can simply derive a refutation: R*[*F*1&*F*0] *:* ⊥ ∨ ¬�(*ref utation*)
*<sup>X</sup>* ∈ *Var*∀(*F*0) *and X* ∈ *Var*∃(*F*1) *can unify since VUR for* (*X*, *<sup>X</sup>*) *holds - there is no global universal variable over the scope of X in F*1*; Sbt*(*X*) = *X and Sbt*(*Y*) = *Y.*
### **3. Fuzzy predicate logic and refutational proof**
The fuzzy predicate logic with evaluated syntax is a flexible and fully complete formalism, which will be used for the below presented extension (Novák, V.). In order to use an efficient form of the resolution principle we have to extend the standard notion of a proof (provability value and degree) with the notion of refutational proof (refutation degree). Propositonal version of the fuzzy resolution principle has been already presented in (Habiballa, H.). We suppose that set of truth values is Łukasiewicz algebra. Therefore we assume standard notions of conjunction, disjunction etc. to be bound with Łukasiewicz operators.
**Definition 8.** *Evaluated proof, refutational proof and refutation degree*
*rsyn*(*Ai*<sup>1</sup> , ..., *Aim* ), *i*1, ..., *im* < *n or ai*
Resolution Principle and Fuzzy Logic 63
*Let T be a fuzzy theory and A* ∈ *FJ a formula. We write T* �*<sup>a</sup> A and say that the formula A is a*
*<sup>T</sup>* �*<sup>a</sup> <sup>A</sup>* iff*<sup>a</sup>* <sup>=</sup> {*Val*(*w*)<sup>|</sup> w is a proof of A from LAx <sup>∪</sup> SAx} (5)
D |= *T* if ∀*A* ∈ LAx : LAx(*A*) ≤ D(*A*), *A* ∈ SAx : SAx(*A*) ≤ D(*A*) (6)
*<sup>T</sup>* <sup>|</sup>=*<sup>a</sup> <sup>A</sup>* iff *<sup>a</sup>* <sup>=</sup> {D(*A*)|D|<sup>=</sup> *<sup>T</sup>*} (7)
*An evaluated refutational formal proof of a formula A from X is w, where additionally a*<sup>0</sup>
*We write T* |=*<sup>a</sup> A and say that the formula A is true in the degree a in the fuzzy theory T.*
*rMP* :
possible to introduce following notion of resolution w.r.t. the modus ponens.
*a* ⊗ *b*
*where σ* = *MGU*(*A*) *is the most general unifier (MGU) of the set of the atoms*
**Definition 11.** *General resolution for fuzzy predicate logic* (*GRFPL*)
*rGR* : *a*
*G represents an occurrence of an atom. The expression Fσ*[*G*/⊥] ∨ *F*�
<sup>1</sup>,..., *G*�
*F or F*� *iff Sig*(*α*) = 1 *in Fσ*[*G*/⊥] ∨ *F*�
*a A*, *b A*⇒*B*
> *a* ⊗ *b*
*where from premise A holding in the degree a and premise A*⇒*B holding in the degree b we infer B*
In classical logic *rMP* could be viewed as a special case of the resolution. The fuzzy resolution rule presented below is also able to simulate fuzzy *rMP*. From this fact the completeness of a system based on resolution can be deduced. It will only remain to prove the soundness. It is
*F*[*G*1, , ..., *Gk*], *b*
*F*� [*G*� <sup>1</sup>, ..., *G*� *n*]
*<sup>n</sup>*} *, G* = *G*1*σ. For every variable α in F or F*�
<sup>∼</sup> *FJ is a finite sequence of evaluated*
*Ai.*
> *A*<sup>0</sup> :=
*An such that An* := *A and for each i* ≤ *n, either there exists*
*Ai* := *X*(*Ai*)
*<sup>B</sup>* (8)
*<sup>F</sup>σ*[*G*/⊥]∇*F*�*σ*[*G*/�] (9)
*σ*[*G*/�]*. F is called positive and F' is called negative premise,*
*,* (*Sbt*(*γ*) = *α*) ∩ *σ* = ∅
*σ*[*G*/�] *is the resolvent of the*
*An evaluated formal proof of a formula A from the fuzzy set X* ⊂
*We will denote the value of the evaluated proof by Val*(*w*) = *an.*
¬*A and An* := ⊥*. Val*(*w*) = *an is called refutation degree of A.*
*theorem in the degree a, or provable in the degree a in the fuzzy theory T.*
The fuzzy modus ponens rule could be formulated:
*A*1, ..., *an*
*formulas w* := *a*<sup>0</sup>
*ai*
1 *A*0, *a*<sup>1</sup>
**Definition 9.** *Provability and truth*
**Definition 10.** *Fuzzy modus ponens*
*holding in the degree a* ⊗ *b.*
*A* = {*G*1,..., *Gk*, *G*�
⇒ *Sig*(*α*) = 1 *in*
*premises on G.*
*an m-ary inference rule r such that*
*Ai* :<sup>=</sup> *<sup>r</sup>evl*(*ai*<sup>1</sup> , ..., *aim* )
We will assume Łukasewicz algebra to be
$$\mathcal{L}\_{\mathbb{E}} = \langle [0,1], \wedge, \vee, \otimes, \to, 0, 1 \rangle$$
where [0, 1] is the interval of reals between 0 and 1, which are the smallest and greatest elements respectively. Basic and additional operations are defined as follows:
*a* ⊗ *b* = 0 ∨ (*a* + *b* − 1) *a* → *b* = 1 ∧ (1 − *a* + *b*) *a* ⊕ *b* = 1 ∧ (*a* + *b*) ¬*a* = 1 − *a*
The biresiduation operation ↔ could be defined *a* ↔ *b* =*d f* (*a* → *b*) ∧ (*b* → *a*), where ∧ is infimum operation. The following properties of L<sup>Ł</sup> will be used in the sequel: *a* ⊗ 1 = *a*, *a* ⊗ 0 = 0, *a* ⊕ 1 = 1, *a* ⊕ 0 = *a*, *a* → 1 = 1, *a* → 0 = ¬*a*, 1 → *a* = *a*, 0 → *a* = 1
The syntax and semantics of fuzzy predicate logic is following:
Graded fuzzy predicate calculus assigns grade to every axiom, in which the formula is valid. It will be written as *a A* where A is a formula and *a* is a syntactic evaluation. We use several standard notions defined in (Novák, V.) namely: inference rule, formal fuzzy theory with set of logical and special axioms, evaluated formal proof.
#### **Definition 6.** *Inference rule*
*An n-ary inference rule r in the graded logical system is a scheme*
$$r: \frac{a\_1 / A\_1, \dots, a\_{\text{fl}} / A\_{\text{fl}}}{r^{\text{evl}}(a\_1, \dots, a\_{\text{fl}}) / r^{\text{syn}}(A\_1, \dots, A\_{\text{fl}})} \tag{4}$$
*using which the evaluated formulas a*<sup>1</sup> *A*1, ..., *an An are assigned the evaluated formula revl*(*a*1, ..., *an*) *rsyn*(*A*1, ..., *An*)*. The syntactic operation rsyn is a partial n-ary operation on FJ and the evaluation operation revl is an n-ary lower semicontinous operation on L (i.e. it preserves arbitrary suprema in all variables).*
**Definition 7.** *Formal fuzzy theory A formal fuzzy theory T in the language J is a triple*
$$T = \langle \text{LAx}, \text{SAx}, R \rangle$$
*where* LAx ⊂ <sup>∼</sup> *FJ is a fuzzy set of logical axioms,* SAx <sup>⊂</sup> <sup>∼</sup> *FJ is a fuzzy set of special axioms, and R is a set of sound inference rules.*
8 Will-be-set-by-IN-TECH
LŁ = �[0, 1], ∧, ∨, ⊗, →, 0, 1� where [0, 1] is the interval of reals between 0 and 1, which are the smallest and greatest
*a* ⊗ *b* = 0 ∨ (*a* + *b* − 1) *a* → *b* = 1 ∧ (1 − *a* + *b*) *a* ⊕ *b* = 1 ∧ (*a* + *b*) ¬*a* = 1 − *a*
The biresiduation operation ↔ could be defined *a* ↔ *b* =*d f* (*a* → *b*) ∧ (*b* → *a*), where ∧ is
• predicates with *p*1, ..., *pm* are syntactically equivalent to FOL ones. Instead of 0 we write ⊥ and instead of 1 we write �, connectives - & (Łukasiewicz conjunction), ∇ (Łukasiewicz disjunction), ⇒ (implication), ¬ (negation), ∀*X* (universal quantifier),∃*X* (existential quantifier) and furthermore by *FJ* we denote set of all formulas of fuzzy logic
• FPL formulas have the following semantic interpretations (D is the universe): Interpretation of terms is equivalent to FOL, D(*pi*(*ti*<sup>1</sup> , ..., *tin* )) = *Pi*(D(*ti*<sup>1</sup> ), ..., D(*tin* )) where *Pi* is a fuzzy relation assigned to *pi*, D(a) = *a* for *a* ∈ [0, 1], D(*A* & *B*) = D(*A*) ⊗ D(*B*),
• for every subformula defined above *Sub*, *Sup*, *Pol*, *Lev*, *Qnt*, *Sbt*, *Sig* and other derived properties defined for classical logic hold (where the classical FOL connective is presented
Graded fuzzy predicate calculus assigns grade to every axiom, in which the formula is valid.
standard notions defined in (Novák, V.) namely: inference rule, formal fuzzy theory with set
*A*1, ..., *an*
*the evaluation operation revl is an n-ary lower semicontinous operation on L (i.e. it preserves arbitrary*
*T* = �LAx, SAx, *R*�
*An*
*rsyn*(*A*1, ..., *An*)*. The syntactic operation rsyn is a partial n-ary operation on FJ and*
*rsyn*(*A*1, ..., *An*)
*An are assigned the evaluated formula*
<sup>∼</sup> *FJ is a fuzzy set of special axioms, and R is a*
(4)
*revl*(*a*1, ..., *an*)
*A*1, ..., *an*
*A* where A is a formula and *a* is a syntactic evaluation. We use several
D(*A*∇*B*) = D(*A*) ⊕ D(*B*), D(*A*⇒*B*) = D(*A*) → D(*B*), D(¬*A*) = ¬D(*A*), D(∀*X*(*A*)) = D(*A*[*x*/*d*]|*<sup>d</sup>* ∈ *<sup>D</sup>*), D(∃*X*(*A*)) = D(*A*[*x*/*d*]|*<sup>d</sup>* ∈ *<sup>D</sup>*)
*a* ⊗ 1 = *a*, *a* ⊗ 0 = 0, *a* ⊕ 1 = 1, *a* ⊕ 0 = *a*, *a* → 1 = 1, *a* → 0 = ¬*a*, 1 → *a* = *a*, 0 → *a* = 1
elements respectively. Basic and additional operations are defined as follows:
infimum operation. The following properties of L<sup>Ł</sup> will be used in the sequel:
The syntax and semantics of fuzzy predicate logic is following:
the Łukasiewicz one has the same mapping value).
of logical and special axioms, evaluated formal proof.
*An n-ary inference rule r in the graded logical system is a scheme*
*<sup>r</sup>* : *<sup>a</sup>*<sup>1</sup>
We will assume Łukasewicz algebra to be
• terms *t*1, ..., *tn* are defined as in FOL
in language *J*
It will be written as *a*
*revl*(*a*1, ..., *an*)
*where* LAx ⊂
**Definition 6.** *Inference rule*
*suprema in all variables).*
*set of sound inference rules.*
*using which the evaluated formulas a*<sup>1</sup>
**Definition 7.** *Formal fuzzy theory*
*A formal fuzzy theory T in the language J is a triple*
<sup>∼</sup> *FJ is a fuzzy set of logical axioms,* SAx <sup>⊂</sup>
### **Definition 8.** *Evaluated proof, refutational proof and refutation degree*
*An evaluated formal proof of a formula A from the fuzzy set X* ⊂ <sup>∼</sup> *FJ is a finite sequence of evaluated formulas w* := *a*<sup>0</sup> *A*0, *a*<sup>1</sup> *A*1, ..., *an An such that An* := *A and for each i* ≤ *n, either there exists an m-ary inference rule r such that*
*ai Ai* :<sup>=</sup> *<sup>r</sup>evl*(*ai*<sup>1</sup> , ..., *aim* ) *rsyn*(*Ai*<sup>1</sup> , ..., *Aim* ), *i*1, ..., *im* < *n or ai Ai* := *X*(*Ai*) *Ai. We will denote the value of the evaluated proof by Val*(*w*) = *an. An evaluated refutational formal proof of a formula A from X is w, where additionally a*<sup>0</sup> *A*<sup>0</sup> := 1 ¬*A and An* := ⊥*. Val*(*w*) = *an is called refutation degree of A.*
### **Definition 9.** *Provability and truth*
*Let T be a fuzzy theory and A* ∈ *FJ a formula. We write T* �*<sup>a</sup> A and say that the formula A is a theorem in the degree a, or provable in the degree a in the fuzzy theory T.*
$$T \vdash\_a A \text{ if } \text{if } a = \bigvee \{Val(w) | \text{ w is a proof of A from LAx} \cup \text{SAx} \}\tag{5}$$
*We write T* |=*<sup>a</sup> A and say that the formula A is true in the degree a in the fuzzy theory T.*
$$\mathcal{D} \vdash T \text{ if } \forall A \in \text{LAx} : \text{LAx}(A) \leq \mathcal{D}(A), A \in \text{SAx} : \text{SAx}(A) \leq \mathcal{D}(A) \tag{6}$$
$$T \vdash\_a A \text{ iff } a = \bigwedge \{ \mathcal{D}(A) \mid \mathcal{D} \mid = T \} \tag{7}$$
The fuzzy modus ponens rule could be formulated:
**Definition 10.** *Fuzzy modus ponens*
$$r\_{MP}:\ \frac{a \, \slash\!\!\/ A \, \slash\!\/ A \Rightarrow B}{a \otimes b \, \slash\!\/ B} \tag{8}$$
*where from premise A holding in the degree a and premise A*⇒*B holding in the degree b we infer B holding in the degree a* ⊗ *b.*
In classical logic *rMP* could be viewed as a special case of the resolution. The fuzzy resolution rule presented below is also able to simulate fuzzy *rMP*. From this fact the completeness of a system based on resolution can be deduced. It will only remain to prove the soundness. It is possible to introduce following notion of resolution w.r.t. the modus ponens.
### **Definition 11.** *General resolution for fuzzy predicate logic* (*GRFPL*)
$$r\_{GR}:\frac{a\left/F\left[\mathbf{G}\_{1\prime\prime}\dots\mathbf{G}\_{k}\right]\,,\not{b}/F\left[\mathbf{G}\_{1\prime\prime}^{\prime}\dots\mathbf{G}\_{n}^{\prime}\right]}{a\odot\not{b}/F\sigma\left[\mathbf{G}/\perp\right]\nabla F^{\prime}\sigma\left[\mathbf{G}/\top\right]}\tag{9}$$
*where σ* = *MGU*(*A*) *is the most general unifier (MGU) of the set of the atoms A* = {*G*1,..., *Gk*, *G*� <sup>1</sup>,..., *G*� *<sup>n</sup>*} *, G* = *G*1*σ. For every variable α in F or F*� *,* (*Sbt*(*γ*) = *α*) ∩ *σ* = ∅ ⇒ *Sig*(*α*) = 1 *in*
*F or F*� *iff Sig*(*α*) = 1 *in Fσ*[*G*/⊥] ∨ *F*� *σ*[*G*/�]*. F is called positive and F' is called negative premise, G represents an occurrence of an atom. The expression Fσ*[*G*/⊥] ∨ *F*� *σ*[*G*/�] *is the resolvent of the premises on G.*
**Definition 12.** *Refutational resolution theorem prover for FPL*
*rs*<sup>∇</sup> :
**Lemma 3.** *Provability and refutation degree for GRFPL*
*a* ⊥∇*A a A*
proof of we can construct refutational proof as follows (*Val*(*w*) ≤ *a*):
<sup>¬</sup>*A*∇*A*, *ai*<sup>+</sup><sup>2</sup>
<sup>¬</sup>*A*, ..., *<sup>a</sup>*
can be replaced by *rGR*, *rs*∇, *rs*⇒. Indeed, let *<sup>w</sup>* be a proof:
*B*{*rs*⇒}
*<sup>T</sup>* �*<sup>a</sup> <sup>A</sup>* iff *<sup>a</sup>* = {*Val*(*w*)| w is a refutational proof of A from LAx ∪ SAx}
If *<sup>a</sup>* = {*Val*(*w*)| w is refutational proof of A from LAx ∪ SAx}(*Val*(*w*) ≤ *<sup>a</sup>*):
a since the formulas are either axioms or results of application of resolution.
**Theorem 1.** *Completeness for fuzzy logic with rGR, rs*∇*, rs*<sup>⇒</sup> *instead of rMP*
*<sup>A</sup>*⇒*<sup>B</sup>* {proof *wA*⇒*B*}, *<sup>a</sup>* <sup>⊗</sup> *<sup>b</sup>*
*<sup>A</sup>*⇒*B*{*proo f wA*⇒*B*}, *<sup>a</sup>* <sup>⊗</sup> *<sup>b</sup>*
*Ai*, *ai*<sup>+</sup><sup>2</sup>
**Definition 13.** *Simplification rules for* ∇, ⇒
*Ai*, 1
*Ai*, 1
and the formula ¬*A*∇*A* may be removed.
*A*0, ..., *ai*
*w* := *a*
*w* := *a*<sup>0</sup>
*w* := *a*
proof: *w* := *a*
*a* ⊗ *b* The proof *w*�� := *a*<sup>0</sup>
There is a proof: *w*� := *a*<sup>0</sup>
*A* {proof *A*}, 1
*A*0, ..., *ai*
*A*0, ..., *ai*
*set of formulas T* �*<sup>a</sup> A iff T* |=*<sup>a</sup> A.*
*A* {proof *wa*}, *b*
*<sup>A</sup>*{*proo f wa*}, *<sup>b</sup>*
**4. Implementation and efficiency**
�⇒*B*{*rs*∇, *<sup>a</sup>* <sup>⊗</sup> *<sup>b</sup>*
*Refutational non-clausal resolution theorem prover for FPL* (*RRTPFPL*) *is the inference system with the inference rule GRFPL and sound simplification rules for* ⊥*,* � *(standard equivalencies for logical constants). A refutational proof by definition 8 represents a proof of a formula G (goal) from the set of special axioms N. It is assumed that Sig*(*α*) = 1 *for* ∀*α in F* ∈ *N* ∪ ¬*G formula, every formula in a*
Resolution Principle and Fuzzy Logic 65
*and rs*<sup>⇒</sup> :
⊥, where *A*0, ..., *Ai* are axioms.
⊥∇*A*.
*Ai*<sup>+</sup>2∇*A*, ..., *<sup>a</sup>*
*Ai*<sup>+</sup>2∇*A*, ..., *<sup>a</sup>*
All the schemes of the type *Aj*∇*A* , *j* > *i* could be simplified by sound simplification rules
*Formal fuzzy theory, where rMP is replaced with rGR, rs*∇*, rs*⇒*, is complete i.e. for every A from the*
*Proof.* The left to right implication (soundness of such formal theory) could be easily done from the soundness of the resolution rule. Conversely it is sufficient to prove that the rule *rMP*
Using the last sequence we can easily make a proof with *rMP* also with the proposed *rR* and simplification rules. Since usual formal theory with *rMP* is complete as it is proved in (Novák, V.), every fuzzy formal theory with these rules is also complete. Note that the non-ground case (requiring unification) could be simulated in the same way like in the proof of soundness.
The author also currently implements the non-clausal theorem prover into fuzzy logic as an extension of previous prover for FOL (GEneralized Resolution Deductive System - GERDS) (Habiballa, H.). Experiments concerning prospective inference strategies can be performed with this extension. The prover called Fuzzy Predicate Logic GEneralized Resolution
⊥∇[�⇒*B*]{*rGR*},
*Proof.* If *<sup>T</sup>* �*<sup>a</sup> <sup>A</sup>* then *<sup>a</sup>* = {*Val*(*w*)| w is a proof of A from LAx ∪ SAx} and for every such a
<sup>¬</sup>*<sup>A</sup>* {member of refutational proof}, *<sup>a</sup>* <sup>⊗</sup> <sup>1</sup>
*a* �⇒*A a A*
> ⊥ {*rGR*}
*A* is a correct proof of A in the degree
*B* {*rMP*}. Then we can replace it by the
*proof has no free variable and has no quantifier for a variable not occurring in the formula.*
### **Lemma 2.** *Soundness of rGR*
*The inference rule rGR for FPL based on* L*Ł is sound i.e. for every truth valuation* D*,*
$$\mathcal{D}(r^{syn}(A\_1, \ldots, A\_{\text{fl}})) \ge r^{\varepsilon \text{vl}}(\mathcal{D}(A\_1), \ldots, \mathcal{D}(A\_{\text{fl}})) \tag{10}$$
*holds true.*
*Proof.* Before we solve the core of *GRFPL* we should prove that the unification algorithm preserves soundness. But it could be simply proved since in the classical FPL with the rule of Modus-Ponens (Novák, V.) from the axiom � (∀*x*)*A*⇒*A*[*x*/*t*] and � (∀*x*)*A* we can prove *A*[*x*/*t*]. For *rGR* we may rewrite the values of the left and right parts of equation (10):
$$\mathcal{D}(r^{\text{sym}}(A\_1, \dots, A\_n)) = \mathcal{D}[\mathcal{D}(F\_1[G/\perp])\nabla \mathcal{D}(F\_2[G/\top])]$$
$$r^{\text{col}}(\mathcal{D}(A\_1), \dots, \mathcal{D}(A\_n)) = \mathcal{D}(F\_1[G]) \otimes \mathcal{D}(F\_2[G])$$
It is sufficient to prove the equality for ⇒ since all other connectives could be defined by it. By induction on the complexity of formula |*A*|, defined as the number of occurrences of connectives, we can prove:
Let premises *F*<sup>1</sup> and *F*<sup>2</sup> be atomic formulas. Since they must contain the same subformula then *F*<sup>1</sup> = *F*<sup>2</sup> = *G* and it holds
$$\mathcal{D}[\mathcal{D}(F\_{\mathbb{T}}[G/\perp])\nabla \mathcal{D}(F\_{\mathbb{Z}}[G/\top])] = D(\perp \nabla \top) = 0 \oplus 1 = 1 \geq \mathcal{D}(F\_{\mathbb{T}}[G]) \otimes \mathcal{D}(F\_{\mathbb{Z}}[G])$$
Induction step: Let premises *F*<sup>1</sup> and *F*<sup>2</sup> be complex formulas and let *A* and *B* are subformulas of *F*1, *C* and *D* are subformulas of *F*<sup>2</sup> and *G* is an atom where generally *F*<sup>1</sup> = (*A*⇒*B*) and *F*<sup>2</sup> = (*C*⇒*D*). The complexity of |*F*1| = |*A*| + 1 or |*F*1| = |*B*| + 1 and |*F*2| = |*C*| + 1 or |*F*2| = |*D*| + 1. Since they must contain the same subformula and for *A*, *B*, *C*, *D* the induction presupposition hold it remain to analyze the following cases:
1. *F*<sup>1</sup> = *A*⇒*G F*<sup>2</sup> = *G*⇒*D* : D[D(*F*1[*G*/⊥])∇D(*F*2[*G*/�])] = D([*A*⇒⊥]∇[�⇒*D*]) = D(¬*A*∇*D*) = 1 ∧ (1 − *a* + *d*)
We have rewritten the expression into Łukasiewicz interpretation. Now we will try to rewrite the right side of the inequality, which has to be proven.
D(*F*1[*G*]) ⊗ D(*F*2[*G*]) = D(*A*⇒*G*) ⊗ D(*G*⇒*D*) = 0 ∨ ((1 ∧ (1 − *a* + *g*)) + (1 ∧ (1 − *g* + *d*)) − 1) = 1∧ (1 − *a* + *d*) The left and right side of the equation (10) are equal and therefore
$$\mathcal{D}[\mathcal{D}(F\_1[G/\perp])\nabla \mathcal{D}(F\_2[G/\top])] \ge \mathcal{D}(F\_1[G]) \otimes \mathcal{D}(F\_2[G])$$
for this case holds.
By induction we have proven that the inequality holds and the *rR* is sound. The induction of the case where only one of the premises has greater complexity is included in the above solved induction step.
10 Will-be-set-by-IN-TECH
*Proof.* Before we solve the core of *GRFPL* we should prove that the unification algorithm preserves soundness. But it could be simply proved since in the classical FPL with the rule of Modus-Ponens (Novák, V.) from the axiom � (∀*x*)*A*⇒*A*[*x*/*t*] and � (∀*x*)*A* we can prove *A*[*x*/*t*]. For *rGR* we may rewrite the values of the left and right parts of equation (10):
<sup>D</sup>(*rsyn*(*A*1, ..., *An*)) = <sup>D</sup>[D(*F*1[*G*/⊥])∇D(*F*2[*G*/�])]
*<sup>r</sup>evl*(D(*A*1), ..., <sup>D</sup>(*An*)) = <sup>D</sup>(*F*1[*G*]) ⊗ D(*F*2[*G*]) It is sufficient to prove the equality for ⇒ since all other connectives could be defined by it. By induction on the complexity of formula |*A*|, defined as the number of occurrences of
Let premises *F*<sup>1</sup> and *F*<sup>2</sup> be atomic formulas. Since they must contain the same subformula then
D[D(*F*1[*G*/⊥])∇D(*F*2[*G*/�])] = *D*(⊥∇�) = 0 ⊕ 1 = 1 ≥ D(*F*1[*G*]) ⊗ D(*F*2[*G*]) Induction step: Let premises *F*<sup>1</sup> and *F*<sup>2</sup> be complex formulas and let *A* and *B* are subformulas of *F*1, *C* and *D* are subformulas of *F*<sup>2</sup> and *G* is an atom where generally *F*<sup>1</sup> = (*A*⇒*B*) and *F*<sup>2</sup> = (*C*⇒*D*). The complexity of |*F*1| = |*A*| + 1 or |*F*1| = |*B*| + 1 and |*F*2| = |*C*| + 1 or |*F*2| = |*D*| + 1. Since they must contain the same subformula and for *A*, *B*, *C*, *D* the induction
1. *F*<sup>1</sup> = *A*⇒*G F*<sup>2</sup> = *G*⇒*D* : D[D(*F*1[*G*/⊥])∇D(*F*2[*G*/�])] = D([*A*⇒⊥]∇[�⇒*D*]) =
We have rewritten the expression into Łukasiewicz interpretation. Now we will try to
D(*F*1[*G*]) ⊗ D(*F*2[*G*]) = D(*A*⇒*G*) ⊗ D(*G*⇒*D*) = 0 ∨ ((1 ∧ (1 − *a* + *g*)) + (1 ∧ (1 − *g* + *d*)) − 1) = 1∧ (1 − *a* + *d*) The left and right side of the equation (10) are equal and therefore
D[D(*F*1[*G*/⊥])∇D(*F*2[*G*/�])] ≥ D(*F*1[*G*]) ⊗ D(*F*2[*G*])
2. *F*<sup>1</sup> = *A*⇒*G F*<sup>2</sup> = *C*⇒*G* : D[D(*F*1[*G*/⊥])∇D(*F*2[*G*/�])] = D([*A*⇒⊥]∇[*C*⇒�]) = 1 ≥
3. *F*<sup>1</sup> = *G*⇒*B F*<sup>2</sup> = *G*⇒*D* : D[D(*F*1[*G*/⊥])∇D(*F*2[*G*/�])] = D([⊥⇒*B*]∇[�⇒*D*]) = 1 ≥
4. *F*<sup>1</sup> = *G*⇒*B F*<sup>2</sup> = *C*⇒*G* : D[D(*F*1[*G*/⊥])∇D(*F*2[*G*/�])] = D([⊥⇒*B*]∇[*C*⇒�]) = 1 ≥
By induction we have proven that the inequality holds and the *rR* is sound. The induction of the case where only one of the premises has greater complexity is included in the above
presupposition hold it remain to analyze the following cases:
rewrite the right side of the inequality, which has to be proven.
<sup>D</sup>(*rsyn*(*A*1, ..., *An*)) <sup>≥</sup> *<sup>r</sup>evl*(D(*A*1), ..., <sup>D</sup>(*An*)) (10)
*The inference rule rGR for FPL based on* L*Ł is sound i.e. for every truth valuation* D*,*
**Lemma 2.** *Soundness of rGR*
connectives, we can prove:
*F*<sup>1</sup> = *F*<sup>2</sup> = *G* and it holds
D(¬*A*∇*D*) = 1 ∧ (1 − *a* + *d*)
for this case holds.
D(*F*1[*G*]) ⊗ D(*F*2[*G*])
D(*F*1[*G*]) ⊗ D(*F*2[*G*])
D(*F*1[*G*]) ⊗ D(*F*2[*G*])
solved induction step.
*holds true.*
### **Definition 12.** *Refutational resolution theorem prover for FPL*
*Refutational non-clausal resolution theorem prover for FPL* (*RRTPFPL*) *is the inference system with the inference rule GRFPL and sound simplification rules for* ⊥*,* � *(standard equivalencies for logical constants). A refutational proof by definition 8 represents a proof of a formula G (goal) from the set of special axioms N. It is assumed that Sig*(*α*) = 1 *for* ∀*α in F* ∈ *N* ∪ ¬*G formula, every formula in a proof has no free variable and has no quantifier for a variable not occurring in the formula.*
**Definition 13.** *Simplification rules for* ∇, ⇒
$$r\_{s\nabla} \colon \frac{a/\perp \nabla A}{a/A} \quad \text{and} \quad r\_{s\Longrightarrow} \colon \frac{a/\top \Rightarrow A}{a/A}$$
**Lemma 3.** *Provability and refutation degree for GRFPL <sup>T</sup>* �*<sup>a</sup> <sup>A</sup>* iff *<sup>a</sup>* = {*Val*(*w*)| w is a refutational proof of A from LAx ∪ SAx}
*Proof.* If *<sup>T</sup>* �*<sup>a</sup> <sup>A</sup>* then *<sup>a</sup>* = {*Val*(*w*)| w is a proof of A from LAx ∪ SAx} and for every such a proof of we can construct refutational proof as follows (*Val*(*w*) ≤ *a*): *w* := *a A* {proof *A*}, 1 <sup>¬</sup>*<sup>A</sup>* {member of refutational proof}, *<sup>a</sup>* <sup>⊗</sup> <sup>1</sup> ⊥ {*rGR*} If *<sup>a</sup>* = {*Val*(*w*)| w is refutational proof of A from LAx ∪ SAx}(*Val*(*w*) ≤ *<sup>a</sup>*): *w* := *a*<sup>0</sup> *A*0, ..., *ai Ai*, 1 <sup>¬</sup>*A*, ..., *<sup>a</sup>* ⊥, where *A*0, ..., *Ai* are axioms. There is a proof: *w*� := *a*<sup>0</sup> *A*0, ..., *ai Ai*, 1 <sup>¬</sup>*A*∇*A*, *ai*<sup>+</sup><sup>2</sup> *Ai*<sup>+</sup>2∇*A*, ..., *<sup>a</sup>* ⊥∇*A*. All the schemes of the type *Aj*∇*A* , *j* > *i* could be simplified by sound simplification rules and the formula ¬*A*∇*A* may be removed.
The proof *w*�� := *a*<sup>0</sup> *A*0, ..., *ai Ai*, *ai*<sup>+</sup><sup>2</sup> *Ai*<sup>+</sup>2∇*A*, ..., *<sup>a</sup> A* is a correct proof of A in the degree a since the formulas are either axioms or results of application of resolution.
### **Theorem 1.** *Completeness for fuzzy logic with rGR, rs*∇*, rs*<sup>⇒</sup> *instead of rMP*
*Formal fuzzy theory, where rMP is replaced with rGR, rs*∇*, rs*⇒*, is complete i.e. for every A from the set of formulas T* �*<sup>a</sup> A iff T* |=*<sup>a</sup> A.*
*Proof.* The left to right implication (soundness of such formal theory) could be easily done from the soundness of the resolution rule. Conversely it is sufficient to prove that the rule *rMP* can be replaced by *rGR*, *rs*∇, *rs*⇒. Indeed, let *<sup>w</sup>* be a proof:
*w* := *a A* {proof *wa*}, *b <sup>A</sup>*⇒*<sup>B</sup>* {proof *wA*⇒*B*}, *<sup>a</sup>* <sup>⊗</sup> *<sup>b</sup> B* {*rMP*}. Then we can replace it by the proof:
$$w := a / A \{
frac{
sigma\_a}{
rho} \}, b / A
Rightarrow
{a
odot
w\_{A
Rightarrow B}}, a
⊗ b /
\perp
∇ [
⌊
TR]
or
{a
odot b}, a
⊗ b /
\top
{a
odot b}, b
$$
Using the last sequence we can easily make a proof with *rMP* also with the proposed *rR* and simplification rules. Since usual formal theory with *rMP* is complete as it is proved in (Novák, V.), every fuzzy formal theory with these rules is also complete. Note that the non-ground case (requiring unification) could be simulated in the same way like in the proof of soundness.
#### **4. Implementation and efficiency**
The author also currently implements the non-clausal theorem prover into fuzzy logic as an extension of previous prover for FOL (GEneralized Resolution Deductive System - GERDS) (Habiballa, H.). Experiments concerning prospective inference strategies can be performed with this extension. The prover called Fuzzy Predicate Logic GEneralized Resolution
resolve all possible combinations of an atom. It uses the following scheme:
**Fuzzy predicate Logic redundancy-based inefficient knowledge bases**
Example: *Rnew* = *p*(*a*), *Rold* = ∀*x*(*p*(*x*)), ¬(∀*x*(*p*(*x*)) → *p*(*a*)) MGU: *Sbt*(*x*) = *a*, *Res* = ¬(⊥→⊥) ∨ ¬(�→�) ⇒ ⊥ We have proved that *Rnew* is a logical consequence of *Rold*.
*b* ≥ *a* then Kill *Rold* (resolvent is removed).
processor as described below.
evaluation degree.
*a* ∧ *b*1⇒*z,*
*a* ∧ *b*<sup>1</sup> ∧ *b*1⇒*z,*
*a* ∧ *b*<sup>1</sup> ∧ *b*<sup>1</sup> ∧ *b*1⇒*z,*
*a* ∧ *b*<sup>1</sup> ∧ *b*<sup>1</sup> ∧ *b*<sup>1</sup> ∧ *b*1⇒*z,*
*...,* 0.51
0.61
0.71
0.81
**Example 4.** *Redundant knowledge base Consider the following knowledge base (fragment):*
limitation of the algorithm and it will not affect the completeness of the *RRTPFOL*.
*Rold* � *Rnew* ⇔ ¬(*Rold* → *Rnew*) � ⊥ Even the usage of this teachnique is a semidecidable problem, we can use time or step
Resolution Principle and Fuzzy Logic 67
In FPL we have to enrich the DCF procedure by the limitation on the provability degree. if *U* �*<sup>a</sup> Rold* ∧ *U* �*<sup>b</sup> Rnew* ∧ *b* ≤ *a* then we can apply DCF. DCF Trivial check performs a symbolic comparison of *Rold* and *Rnew* we use the same provability degree condition. In other cases we have to add *Rnew* into the set of resolvents and we can apply DCF Kill procedure. DCF Kill searches for every *Rold* being a logical consequence of *Rnew* and if *U* �*<sup>a</sup> Rold* ∧ *U* �*<sup>b</sup> Rnew* ∧
We will now show some efficiency results concerning many-valued logic both for Fuzzy Predicate Logic. We have used the above mentioned application FPLGERDS and originally developed DCF strategy for FPL. It is clear that inference in *RRTPFPL* and *RRTPFDL* on general knowledge bases is a problem solved in exponential time. Nevertheless as we would like to demonstrate the need to search for every possible proof (in contrast to the two-valued logic) will not necessarily in particular cases lead to the inefficient theory. We have devised knowledge bases (KB) on the following typical problems related to the use of fuzzy logic. We have performed experimental measurements concerning efficiency of the presented non-clausal resolution principle and also DCF technique. These measurements were done using the FPLGERDS application (Habiballa, H.). Special testing knowledge bases were prepared and several types of inference were tested on a PC with standard Intel Pentium 4
As it was shown above in the theorem proving example the problem of proof search is quite different in FPL and FDL in comparison with the two-valued logic. We have to search for the best refutation degree using refutational theorem proving in order to make sensible conclusions from the inference process. It means we cannot accept the **first successful** proof, but we have to check **"all possible proofs"** or we have to be sure that every omitted proof is **worse** that some another one. The presented DCF and DCF Kill technique belong to the third sort of proof search strategies, i.e. they omit proofs that are really worse than some another (see the explication above). Proofs and formulas causing this could be called redundant proofs and redundant formulas. Fuzzy logic makes this redundancy dimensionally harder since we could produce not only equivalent formulas but also equivalent formulas of different
Deductive System (Fig. 1) - FPLGERDS provides standard interface for input (knowledge base and goals) and output (proof sequence and results of fuzzy inference, statistics).
Fig. 1. Fuzzy Predicate Logic GEneralized Resolution Deductive System
There are already several efficient strategies proposed by author (mainly Detection of Consequent Formulas (DCF) adopted for the usage also in FPL). With these strategies the proving engine can be implemented in real-life applications since the complexity of theorem proving in FPL is dimensionally harder than in FOL (the need to search for all possible proofs - we try to find the best refutation degree). The DCF idea is to forbid the addition of a resolvent which is a logical consequence of any previously added resolvent. For refutational theorem proving it is a sound and complete strategy and it is emiprically very effective. Completeness of such a strategy is also straight-forward in FOL:
$$(R\_{old} \vdash R\_{new}) \land (\mathsf{U}, R\_{new} \vdash \bot) \Rightarrow (\mathsf{U}, R\_{old} \vdash \bot)$$
Example: *Rnew* = *p*(*a*), *Rold* = ∀*x*(*p*(*x*)), *Rold* � *Rnew*.
DCF could be implemented by the same procedures like General Resolution (we may utilize self-resolution). Self-resolution has the same positive and negative premise and needs to 12 Will-be-set-by-IN-TECH
Deductive System (Fig. 1) - FPLGERDS provides standard interface for input (knowledge
base and goals) and output (proof sequence and results of fuzzy inference, statistics).
Fig. 1. Fuzzy Predicate Logic GEneralized Resolution Deductive System
Completeness of such a strategy is also straight-forward in FOL:
Example: *Rnew* = *p*(*a*), *Rold* = ∀*x*(*p*(*x*)), *Rold* � *Rnew*.
There are already several efficient strategies proposed by author (mainly Detection of Consequent Formulas (DCF) adopted for the usage also in FPL). With these strategies the proving engine can be implemented in real-life applications since the complexity of theorem proving in FPL is dimensionally harder than in FOL (the need to search for all possible proofs - we try to find the best refutation degree). The DCF idea is to forbid the addition of a resolvent which is a logical consequence of any previously added resolvent. For refutational theorem proving it is a sound and complete strategy and it is emiprically very effective.
(*Rold* � *Rnew*) ∧ (*U*, *Rnew* � ⊥) ⇒ (*U*, *Rold* � ⊥)
DCF could be implemented by the same procedures like General Resolution (we may utilize self-resolution). Self-resolution has the same positive and negative premise and needs to resolve all possible combinations of an atom. It uses the following scheme:
$$R\_{old} \vdash R\_{new} \Leftrightarrow \neg (R\_{old} \to R\_{new}) \vdash \bot$$
Even the usage of this teachnique is a semidecidable problem, we can use time or step limitation of the algorithm and it will not affect the completeness of the *RRTPFOL*. Example: *Rnew* = *p*(*a*), *Rold* = ∀*x*(*p*(*x*)), ¬(∀*x*(*p*(*x*)) → *p*(*a*)) MGU: *Sbt*(*x*) = *a*, *Res* = ¬(⊥→⊥) ∨ ¬(�→�) ⇒ ⊥ We have proved that *Rnew* is a logical consequence of *Rold*.
In FPL we have to enrich the DCF procedure by the limitation on the provability degree. if *U* �*<sup>a</sup> Rold* ∧ *U* �*<sup>b</sup> Rnew* ∧ *b* ≤ *a* then we can apply DCF. DCF Trivial check performs a symbolic comparison of *Rold* and *Rnew* we use the same provability degree condition. In other cases we have to add *Rnew* into the set of resolvents and we can apply DCF Kill procedure. DCF Kill searches for every *Rold* being a logical consequence of *Rnew* and if *U* �*<sup>a</sup> Rold* ∧ *U* �*<sup>b</sup> Rnew* ∧ *b* ≥ *a* then Kill *Rold* (resolvent is removed).
We will now show some efficiency results concerning many-valued logic both for Fuzzy Predicate Logic. We have used the above mentioned application FPLGERDS and originally developed DCF strategy for FPL. It is clear that inference in *RRTPFPL* and *RRTPFDL* on general knowledge bases is a problem solved in exponential time. Nevertheless as we would like to demonstrate the need to search for every possible proof (in contrast to the two-valued logic) will not necessarily in particular cases lead to the inefficient theory. We have devised knowledge bases (KB) on the following typical problems related to the use of fuzzy logic.
We have performed experimental measurements concerning efficiency of the presented non-clausal resolution principle and also DCF technique. These measurements were done using the FPLGERDS application (Habiballa, H.). Special testing knowledge bases were prepared and several types of inference were tested on a PC with standard Intel Pentium 4 processor as described below.
### **Fuzzy predicate Logic redundancy-based inefficient knowledge bases**
As it was shown above in the theorem proving example the problem of proof search is quite different in FPL and FDL in comparison with the two-valued logic. We have to search for the best refutation degree using refutational theorem proving in order to make sensible conclusions from the inference process. It means we cannot accept the **first successful** proof, but we have to check **"all possible proofs"** or we have to be sure that every omitted proof is **worse** that some another one. The presented DCF and DCF Kill technique belong to the third sort of proof search strategies, i.e. they omit proofs that are really worse than some another (see the explication above). Proofs and formulas causing this could be called redundant proofs and redundant formulas. Fuzzy logic makes this redundancy dimensionally harder since we could produce not only equivalent formulas but also equivalent formulas of different evaluation degree.
**Example 4.** *Redundant knowledge base Consider the following knowledge base (fragment):*
*...,* 0.51 *a* ∧ *b*1⇒*z,* 0.61 *a* ∧ *b*<sup>1</sup> ∧ *b*1⇒*z,* 0.71 *a* ∧ *b*<sup>1</sup> ∧ *b*<sup>1</sup> ∧ *b*1⇒*z,* 0.81 *a* ∧ *b*<sup>1</sup> ∧ *b*<sup>1</sup> ∧ *b*<sup>1</sup> ∧ *b*1⇒*z,*
**Search DCF** Code **Description** Breadth Trivial BT Complete Breadth DCF BDC Complete Breadth DCF Kill BDK Complete Mod. Linear Trivial MT Incomplete (+) Mod. Linear DCF MDC Incomplete (+) Mod. Linear DCF Kill MDK Incomplete (+) Linear Trivial LT Incomplete Linear DCF LDC Incomplete Linear DCF Kill LDK Incomplete
Resolution Principle and Fuzzy Logic 69
it is computationally very simple and forms necessary essential restriction for possibly infinite inference process. The next method of DCF technique enables do detect the equivalency of a formula (potential new resolvent) by the means described above. DCF Kill technique additionally tries to remove every redundant resolvent from the set of resolvents. The important aspect of the theorem DCF lies in its simple implementation into an automated theorem prover based on general resolution. The prover handles formulas in the form of syntactical tree. It is programmed a procedure performing general resolution with two formulas on an atom. This procedure is also used for the implementation of the theorem. A "virtual tree" is created from candidate and former resolvent (axiom) connected by negated implication. Then it remains to perform self-resolution on such formula until a logical value is obtained. Let us compare the efficiency of standard strategies and the above-defined one. We have built-up 9 combinations of inference strategies from the mentioned proof search and DCF heuristics. They have different computational strength, i.e. their completeness is different for various classes of formulas. Fully complete (as described above) for general formulas of FPL and FDL are only breadth-first search combinations. Linear search strategies are not complete even for two-valued logic and horn clauses. Modified linear search has generally bad completeness results when an infinite loop is present in proofs, but for guarded knowledge bases it can assure completeness preserving better space efficiency than breadth-first search. We tested presented inference strategies on sample knowledge bases with redundancy level 5 with 20, 40, 60, 80 and 100 groups of mutually redundant formulas (total number of formulas in knowledge base is 120, 240, 360, 480 and 600). At first we have tested their time efficiency for inference process. As it could be observed from figure 2, the best results have **LDK and LDC** strategies. For simple guarded knowledge bases (not leading to an infinite loop in proof search and where the goal itself assures the best refutation degree) these two methods are **very efficient**. DCF strategies significantly reduces the proof search even in comparison with LT strategy (standard), therefore the usage of any non-trivial DCF heuristics is significant. Next important result concludes from the comparison of BDK and MDK, MDC strategies. We can conclude that MDK and MDC strategies are relatively comparable to BDK and moreover BDK
Space complexity is even more significantly affected by the DCF heuristics. There is an interesting comparison of trivial and non-trivial DCF heuristics in figure 3. Even BDK strategy brings significant reduction of resolvents amount, while LDK, LDC, MDK, MDC strategies have minimal necessary amount of kept resolvents during inference process. The second examined redundancy level 10 shows also important comparison for increasing redundancy in knowledge bases. Tested knowledge bases contained 10, 20, 30, 40 and 50 groups of 10 equivalent formulas (the total number of formulas was 110, 220, 330, 440 and 550 formulas).
Table 4. Inference strategies
preserves completeness for general knowledge bases.
Table 2. Proof search algorithms
Table 3. DCF heuristics
0.91 *<sup>a</sup>* <sup>∧</sup> *<sup>b</sup>*<sup>1</sup> <sup>∧</sup> *<sup>b</sup>*<sup>1</sup> <sup>∧</sup> *<sup>b</sup>*<sup>1</sup> <sup>∧</sup> *<sup>b</sup>*<sup>1</sup> <sup>∧</sup> *<sup>b</sup>*1⇒*z,* <sup>1</sup> *b*1*,*
*...,* 0.52 *a* ∧ *b*2⇒*z,* 0.62 *a* ∧ *b*<sup>2</sup> ∧ *b*2⇒*z,* 0.72 *a* ∧ *b*<sup>2</sup> ∧ *b*<sup>2</sup> ∧ *b*2⇒*z,* 0.82 *a* ∧ *b*<sup>2</sup> ∧ *b*<sup>2</sup> ∧ *b*<sup>2</sup> ∧ *b*2⇒*z,* 0.92 *<sup>a</sup>* <sup>∧</sup> *<sup>b</sup>*<sup>2</sup> <sup>∧</sup> *<sup>b</sup>*<sup>2</sup> <sup>∧</sup> *<sup>b</sup>*<sup>2</sup> <sup>∧</sup> *<sup>b</sup>*<sup>2</sup> <sup>∧</sup> *<sup>b</sup>*2⇒*z,* <sup>1</sup> *b*2*, ...,*
```
Goal: ? − a⇒z
```
*Searching for the best proof of a goal will produce a lot of logically equivalent formulas with different degrees. These resolvents make the inference process inefficient and one of the essential demands to the presented refutational theorem prover is a reasonable inference strategy with acceptable time complexity.*
We have compared efficiency of the standard **breadth-first search**, **linear search** and **modified linear search** (starting from every formula in knowledge base) and also combinations with DCF and DCF-kill technique (Habiballa, H.). We have prepared knowledge bases of the size 120, 240, 360, 480 and 600 formulas. It has been compared the time and space efficiency on the criterion of 2 redundancy levels. This level represents the number of redundant formulas to which the formula is equivalent (including the original formula). For example the level 5 means the knowledge base contain 5 equivalent redundant formulas for every formula (including the formula itself). The basic possible state space search techniques and DCF heuristics and their combinations are presented in the following tables.
We use standard state space search algorithms in the FPLGERDS application - Breadth-first and Linear search. Breadth-first method searches for every possible resolvent from the formulas of the level 0 (goal and special axioms). These resolvents form formulas of the level 1 and we try to combine them with all formulas of the same and lower level and continue by the same procedure until no other non-redundant resolvent could be found. Linear search performs depth-first search procedure, where every produced resolvent is used as one of the premises in succeeding step of inference. The first produced resolvents arises from the goal formula. Modified linear search method posses the same procedure as linear one, but it starts from goal and also from all the special axioms.
DCF methods for reduction of resolvent space are basically three. The simplest is trivial DCF method, which detects redundant resolvent only by its exact symbolic comparison, i.e. formulas are equivalent only if the are syntactically the same. Even it is a very rough method, 14 Will-be-set-by-IN-TECH
Breadth B Level order generation, start - special axioms + goal
Modified-Linear M Resolvent ⇒ premise, start - goal + special axioms
DCF DC Potential resolvent is consequent (no addition) DCF Kill DK DCF + remove all consequent resolvents
*Searching for the best proof of a goal will produce a lot of logically equivalent formulas with different degrees. These resolvents make the inference process inefficient and one of the essential demands to the presented refutational theorem prover is a reasonable inference strategy with acceptable time complexity.*
We have compared efficiency of the standard **breadth-first search**, **linear search** and **modified linear search** (starting from every formula in knowledge base) and also combinations with DCF and DCF-kill technique (Habiballa, H.). We have prepared knowledge bases of the size 120, 240, 360, 480 and 600 formulas. It has been compared the time and space efficiency on the criterion of 2 redundancy levels. This level represents the number of redundant formulas to which the formula is equivalent (including the original formula). For example the level 5 means the knowledge base contain 5 equivalent redundant formulas for every formula (including the formula itself). The basic possible state space search techniques and DCF
We use standard state space search algorithms in the FPLGERDS application - Breadth-first and Linear search. Breadth-first method searches for every possible resolvent from the formulas of the level 0 (goal and special axioms). These resolvents form formulas of the level 1 and we try to combine them with all formulas of the same and lower level and continue by the same procedure until no other non-redundant resolvent could be found. Linear search performs depth-first search procedure, where every produced resolvent is used as one of the premises in succeeding step of inference. The first produced resolvents arises from the goal formula. Modified linear search method posses the same procedure as linear one, but it starts
DCF methods for reduction of resolvent space are basically three. The simplest is trivial DCF method, which detects redundant resolvent only by its exact symbolic comparison, i.e. formulas are equivalent only if the are syntactically the same. Even it is a very rough method,
Linear L Resolvent ⇒ premise, start - goal
Trivial T Exact symbolic comparison
*b*1*,*
*b*2*,*
heuristics and their combinations are presented in the following tables.
**Search method Description**
**DCF Method Description**
Table 2. Proof search algorithms
*<sup>a</sup>* <sup>∧</sup> *<sup>b</sup>*<sup>1</sup> <sup>∧</sup> *<sup>b</sup>*<sup>1</sup> <sup>∧</sup> *<sup>b</sup>*<sup>1</sup> <sup>∧</sup> *<sup>b</sup>*<sup>1</sup> <sup>∧</sup> *<sup>b</sup>*1⇒*z,* <sup>1</sup>
*<sup>a</sup>* <sup>∧</sup> *<sup>b</sup>*<sup>2</sup> <sup>∧</sup> *<sup>b</sup>*<sup>2</sup> <sup>∧</sup> *<sup>b</sup>*<sup>2</sup> <sup>∧</sup> *<sup>b</sup>*<sup>2</sup> <sup>∧</sup> *<sup>b</sup>*2⇒*z,* <sup>1</sup>
from goal and also from all the special axioms.
Table 3. DCF heuristics
*a* ∧ *b*2⇒*z,*
*a* ∧ *b*<sup>2</sup> ∧ *b*2⇒*z,*
*a* ∧ *b*<sup>2</sup> ∧ *b*<sup>2</sup> ∧ *b*2⇒*z,*
*a* ∧ *b*<sup>2</sup> ∧ *b*<sup>2</sup> ∧ *b*<sup>2</sup> ∧ *b*2⇒*z,*
0.91
0.62
0.72
0.82
0.92
*Goal:* ? − *a*⇒*z*
*...,*
*...,* 0.52
Table 4. Inference strategies
it is computationally very simple and forms necessary essential restriction for possibly infinite inference process. The next method of DCF technique enables do detect the equivalency of a formula (potential new resolvent) by the means described above. DCF Kill technique additionally tries to remove every redundant resolvent from the set of resolvents. The important aspect of the theorem DCF lies in its simple implementation into an automated theorem prover based on general resolution. The prover handles formulas in the form of syntactical tree. It is programmed a procedure performing general resolution with two formulas on an atom. This procedure is also used for the implementation of the theorem. A "virtual tree" is created from candidate and former resolvent (axiom) connected by negated implication. Then it remains to perform self-resolution on such formula until a logical value is obtained. Let us compare the efficiency of standard strategies and the above-defined one. We have built-up 9 combinations of inference strategies from the mentioned proof search and DCF heuristics. They have different computational strength, i.e. their completeness is different for various classes of formulas. Fully complete (as described above) for general formulas of FPL and FDL are only breadth-first search combinations. Linear search strategies are not complete even for two-valued logic and horn clauses. Modified linear search has generally bad completeness results when an infinite loop is present in proofs, but for guarded knowledge bases it can assure completeness preserving better space efficiency than breadth-first search. We tested presented inference strategies on sample knowledge bases with redundancy level 5 with 20, 40, 60, 80 and 100 groups of mutually redundant formulas (total number of formulas in knowledge base is 120, 240, 360, 480 and 600). At first we have tested their time efficiency for inference process. As it could be observed from figure 2, the best results have **LDK and LDC** strategies. For simple guarded knowledge bases (not leading to an infinite loop in proof search and where the goal itself assures the best refutation degree) these two methods are **very efficient**. DCF strategies significantly reduces the proof search even in comparison with LT strategy (standard), therefore the usage of any non-trivial DCF heuristics is significant. Next important result concludes from the comparison of BDK and MDK, MDC strategies. We can conclude that MDK and MDC strategies are relatively comparable to BDK and moreover BDK preserves completeness for general knowledge bases.
Space complexity is even more significantly affected by the DCF heuristics. There is an interesting comparison of trivial and non-trivial DCF heuristics in figure 3. Even BDK strategy brings significant reduction of resolvents amount, while LDK, LDC, MDK, MDC strategies have minimal necessary amount of kept resolvents during inference process. The second examined redundancy level 10 shows also important comparison for increasing redundancy in knowledge bases. Tested knowledge bases contained 10, 20, 30, 40 and 50 groups of 10 equivalent formulas (the total number of formulas was 110, 220, 330, 440 and 550 formulas).
Fig. 2. Time complexity for redundancy level 5 (seconds)
Time efficiency results shows that higher redundancy level causes expected increase in the necessary time for the best proof search (figure 4). The approximate increase is double, while the proportion shows good results for MDK, MDC and LDK, LDC (linear search based) strategies. This property also holds for space complexity as shown in figure 5. Performed experiments shows the significance of originally developed DCF strategies in combination with standard breadth-first search (important for general knowledge bases - **BDK**). We also outlined high efficiency for linear search based strategies (mainly **LDK**). Even this strategy is not fully complete and could be used only for guarded fragment of FDL, this problem is already known in classical (two-valued) logic programming and automated theorem proving. We also use these highly efficient linear search strategies, even they are not complete.
Fig. 3. Space complexity for redundancy level 5 (resolvents)
Resolution Principle and Fuzzy Logic 71
16 Will-be-set-by-IN-TECH
Fig. 2. Time complexity for redundancy level 5 (seconds)
Time efficiency results shows that higher redundancy level causes expected increase in the necessary time for the best proof search (figure 4). The approximate increase is double, while the proportion shows good results for MDK, MDC and LDK, LDC (linear search based) strategies. This property also holds for space complexity as shown in figure 5. Performed experiments shows the significance of originally developed DCF strategies in combination with standard breadth-first search (important for general knowledge bases - **BDK**). We also outlined high efficiency for linear search based strategies (mainly **LDK**). Even this strategy is not fully complete and could be used only for guarded fragment of FDL, this problem is already known in classical (two-valued) logic programming and automated theorem proving.
We also use these highly efficient linear search strategies, even they are not complete.
Fig. 3. Space complexity for redundancy level 5 (resolvents)
Fig. 5. Space complexity for redundancy level 10 (resolvents)
Resolution Principle and Fuzzy Logic 73
Fig. 4. Time complexity for redundancy level 10 (seconds)
18 Will-be-set-by-IN-TECH
Fig. 4. Time complexity for redundancy level 10 (seconds)
Fig. 5. Space complexity for redundancy level 10 (resolvents)
Jorma K. Mattila
*Finland*
**4**
*Lappeenranta University of Technology*
**Many-Valued Logics**
**Standard Fuzzy Sets and some**
The aim of this chapter is to consider the relationship between standard fuzzy set theory and some many-valued logics. Prof. Lotfi A. Zadeh introduced his theory of fuzzy sets in sixties, and his first paper that circulated widely around the world is "Fuzzy Sets" (Zadeh, 1965). In
After Zadeh has introduced his theory, many-valued logic began to have a new interest. Especially, Łukasiewicz logic was enclosed quite closely in fuzzy sets. There is a strong opinion that Łukasiewicz infinite-valued logic has the role as the logic of fuzzy sets, similarly as classical logic has the role as the logic of crisp sets. But actually, it seems that Kleene's 3-valued logic was the closest logic connecting to fuzzy sets, when Zadeh created his theory. We will discuss this thing later. In the books Rescher (Rescher, 1969) and Bergmann
In Section 2 we consider the main concepts of fuzzy set theory. We will not do it completely, because our purpose is not to present the whole theory of standard fuzzy sets. We restrict our consideration on those things we need when we are "building a bridge" between fuzzy sets
In Section 3 we consider De Morgan algebras in general in order to have a formal base to our consideration. There are many sources for this topic. One remarkable one is Rasiowa's book
In Section 4 we introduce an algebraic approach for standard fuzzy set theory by applying De Morgan algebras. We choose an algebra from the infinite large collection of De Morgan algebras that fits completely to standard fuzzy set theory. We call this De Morgan algebra by the name *Zadeh algebra*. The concept "Zadeh algebra" was introduced by the author in an international symposium "Fuzziness in Finland" in 2004. Also Prof. Zadeh attended this event. In the same year, a more comprehensive article about Zahed algebra (*cf.* (Mattila, 2004)) was published by the author. This algebra gives a tool for studying connections between standard fuzzy sets and certain many-valued logics. Two of these logics are Kleene's logic and Łukasiewicz logic. Some analysis about Łukasiewicz and Kleene's logic is given for example in Mattila (Mattila, 2009). Especially, connections to modal logic are considered in that paper.
the long run, this theory was began to call by the name *theory of standard fuzzy sets*.
and some closely related logics. The section is based on Zadeh (Zadeh, 1965).
(Bergmann, 2008) descriptions about Kleene's logic are given.
**1. Introduction**
(Rasiowa, 1974).
### **5. Conclusions and further research**
The *Non-clausal Refutational Resolution Theorem Prover* forms a powerful inference system for automated theorem proving in fuzzy predicate logic. The main advantage in contrast with other inference systems lies in the possibility to utilize various inference strategies for effective reasoning. Therefore it is essential for practically successful theorem proving.
The Detection of Consequent Formulas algorithms family brings significant improvements in time and space efficiency for the best proof search. It has been shown results indicating specific behavior of some combinations of the DCF and standard proof search (breadth-first and linear search). DCF strategies (BDC, BDK) have interesting results even for fully general fuzzy predicate logic with evaluated syntax, where the strategy makes the inference process practically manageable (in contrast to unrestricted blind proof-search). However it seems to be more promising for practical applications to utilize incomplete strategies with high time efficiency like LDK (even for large knowledge bases it has very short solving times). It conforms to another successful practical applications in two-valued logic like logic programming or deductive databases where there are also used efficient incomplete strategies for fragments of fully general logics.
It has been briefly presented some efficiency results for the presented automated theorem prover and inference strategies. They show the significant reduction of time and space complexity for the DCF technique. Experimental application FPLGERDS can be obtained from URL:// *http://www1.osu.cz/home/habibal/files/gerds.zip*. The package contains current version of the application, source codes, examples and documentation. This work was supported by project DAR (1M0572).
### **6. References**
## **Standard Fuzzy Sets and some Many-Valued Logics**
Jorma K. Mattila *Lappeenranta University of Technology Finland*
### **1. Introduction**
20 Will-be-set-by-IN-TECH
74 Fuzzy Logic – Algorithms, Techniques and Implementations
The *Non-clausal Refutational Resolution Theorem Prover* forms a powerful inference system for automated theorem proving in fuzzy predicate logic. The main advantage in contrast with other inference systems lies in the possibility to utilize various inference strategies for effective
The Detection of Consequent Formulas algorithms family brings significant improvements in time and space efficiency for the best proof search. It has been shown results indicating specific behavior of some combinations of the DCF and standard proof search (breadth-first and linear search). DCF strategies (BDC, BDK) have interesting results even for fully general fuzzy predicate logic with evaluated syntax, where the strategy makes the inference process practically manageable (in contrast to unrestricted blind proof-search). However it seems to be more promising for practical applications to utilize incomplete strategies with high time efficiency like LDK (even for large knowledge bases it has very short solving times). It conforms to another successful practical applications in two-valued logic like logic programming or deductive databases where there are also used efficient incomplete strategies
It has been briefly presented some efficiency results for the presented automated theorem prover and inference strategies. They show the significant reduction of time and space complexity for the DCF technique. Experimental application FPLGERDS can be obtained from URL:// *http://www1.osu.cz/home/habibal/files/gerds.zip*. The package contains current version of the application, source codes, examples and documentation. This work was supported by
Bachmair, L., Ganzinger, H. (1997). A theory of resolution. Technical report:
Bachmair, L., Ganzinger, H. (2001). Resolution theorem proving. In Handbook of Automated
Duki´c, N., Avdagi´c, Z. (2005). Fuzzy Functional Dependency and the Resolution Principle.
Habiballa, H. (2000). Non-clausal resolution - theory and practice. Research report: University
Habiballa, H., Novák, V. (2002). Fuzzy General Resolution. In Proc. of Intl. Conf. Aplimat 2002.
Habiballa, H. (2006). Resolution Based Reasoning in Description Logic. In Proc. of Intl. Conf.
Habiballa, H.(2006a). Fuzzy Predicate Logic Generalized Resolution Deductive System.
Hájek, P. (2000). Metamathematics of fuzzy logic. Kluwer Academic Publishers - Dordrecht,
Hájek, P. (2005). Making fuzzy description logic more general. Fuzzy Sets and Systems
Novák, V., Perfilieva, I., Moˇckoˇr, J. (1999). Mathematical principles of fuzzy logic. Kluwer, 1999.
of Ostrava, 2000, http://www.volny.cz/habiballa/files/gerds.pdf
rep. at http://ac030.osu.cz/irafm/ps/rep47.ps
http://ac030.osu.cz/irafm/ps/rep66.ps.gz.
In Informatica, Vilnius: Lith. Acad. Sci. (IOSPRESS), 2005, Vol.16, No. 1, pp. 45 - 60,
Bratislava, Slovak Technical University, 2002. pp. 199-206, also available as research
ZNALOSTI 2006, Univ. of Hradec Kralove, 2006, also available as research rep. at
Technical Report, Institute for Research and Application of Fuzzy Modeling,
reasoning. Therefore it is essential for practically successful theorem proving.
**5. Conclusions and further research**
for fragments of fully general logics.
Max-Planck-Institut, 1997.
Reasoning, MIT Press, 2001.
University of Ostrava, 2006.
154(2005),pp. 1-15.
project DAR (1M0572).
2005.
2000.
**6. References**
The aim of this chapter is to consider the relationship between standard fuzzy set theory and some many-valued logics. Prof. Lotfi A. Zadeh introduced his theory of fuzzy sets in sixties, and his first paper that circulated widely around the world is "Fuzzy Sets" (Zadeh, 1965). In the long run, this theory was began to call by the name *theory of standard fuzzy sets*.
After Zadeh has introduced his theory, many-valued logic began to have a new interest. Especially, Łukasiewicz logic was enclosed quite closely in fuzzy sets. There is a strong opinion that Łukasiewicz infinite-valued logic has the role as the logic of fuzzy sets, similarly as classical logic has the role as the logic of crisp sets. But actually, it seems that Kleene's 3-valued logic was the closest logic connecting to fuzzy sets, when Zadeh created his theory. We will discuss this thing later. In the books Rescher (Rescher, 1969) and Bergmann (Bergmann, 2008) descriptions about Kleene's logic are given.
In Section 2 we consider the main concepts of fuzzy set theory. We will not do it completely, because our purpose is not to present the whole theory of standard fuzzy sets. We restrict our consideration on those things we need when we are "building a bridge" between fuzzy sets and some closely related logics. The section is based on Zadeh (Zadeh, 1965).
In Section 3 we consider De Morgan algebras in general in order to have a formal base to our consideration. There are many sources for this topic. One remarkable one is Rasiowa's book (Rasiowa, 1974).
In Section 4 we introduce an algebraic approach for standard fuzzy set theory by applying De Morgan algebras. We choose an algebra from the infinite large collection of De Morgan algebras that fits completely to standard fuzzy set theory. We call this De Morgan algebra by the name *Zadeh algebra*. The concept "Zadeh algebra" was introduced by the author in an international symposium "Fuzziness in Finland" in 2004. Also Prof. Zadeh attended this event. In the same year, a more comprehensive article about Zahed algebra (*cf.* (Mattila, 2004)) was published by the author. This algebra gives a tool for studying connections between standard fuzzy sets and certain many-valued logics. Two of these logics are Kleene's logic and Łukasiewicz logic. Some analysis about Łukasiewicz and Kleene's logic is given for example in Mattila (Mattila, 2009). Especially, connections to modal logic are considered in that paper.
**2. Zadeh's theory of standard fuzzy sets**
The *power set* of all fuzzy subsets of the set *X* is
function
as a special case.
(Zadeh, 1965).
following operations:
1996)).
For considering the standard system of fuzzy sets, the range of fuzzy sets (i.e., that of membership functions) is the unit interval **I** = [0, 1]. We give the definition of the concept *fuzzy set* using Zadeh's original definition. However, some symbols have been changed. Usually, the symbol of a fuzzy set, in general, is denoted by *μ*. A membership function of
Standard Fuzzy Sets and some Many-Valued Logics 77
**Definition 2.1** (Standard fuzzy set)**.** A *fuzzy subset A* of a set *X* is characterized by a *membership function* A(*x*) which associates with each point *x* in *X* a real number in the interval [0, 1], with the value of A(*x*) at *x* representing the "grade of membership" of *x* in *A*. Thus, the nearer the
This definition means that a fuzzy subset *A* of a universe of discourse *X* is represented by a
A : *X* −→ **I**.
An important subset of the set of all membership functions (2.1) is the set of functions taking
**<sup>2</sup>***<sup>X</sup>* <sup>=</sup> { *<sup>f</sup>* <sup>|</sup> *<sup>f</sup>* : *<sup>X</sup>* −→ {0, 1}}
It is also a well-known fact that **I** and **I***<sup>X</sup>* are partially ordered sets. (Actually, **I** is a totally ordered set, but hence it is also prtially ordered.) In fact, they are also distributive complete lattices. Generally, some main properties of **I** can be embedded to **I***<sup>X</sup>* (*cf.* e.g. Lowen (Lowen,
We consider operations, properties, and some concepts involved in fuzzy sets given by Zadeh
**Definition 2.2** (Basic operations)**.** Let <sup>A</sup>, <sup>B</sup> <sup>∈</sup> **<sup>I</sup>***<sup>X</sup>* and *<sup>x</sup>* <sup>∈</sup> *<sup>X</sup>*. In **<sup>I</sup>***<sup>X</sup>* there are defined the
A(*x*) = 1 − A(*x*) *complementarity*
∀*x* ∈ *X*, A(*x*) = B(*x*).
A fuzzy set *A* is *contained* in a fuzzy set *B*, i.e., *A* is a *subset* of *B*, denoted by A ⊆ B, if their
∀ *x* ∈ *X*, A(*x*) ≤ B(*x*)
(A ∨ B)(*x*) = max{A(*x*), B(*x*)} *union* (A ∧ B)(*x*) = min{A(*x*), B(*x*)} *intersection*
Two fuzzy sets <sup>A</sup>, <sup>B</sup> <sup>∈</sup> **<sup>I</sup>***<sup>X</sup>* are *equal*, denoted by <sup>A</sup> <sup>=</sup> <sup>B</sup>, if
membership functions satisfy the condition
only values 1 or 0, i.e., the set of all characteristic functions of the crisp subsets of *X*
**<sup>I</sup>***<sup>X</sup>* <sup>=</sup> { <sup>A</sup> <sup>|</sup> <sup>A</sup> : *<sup>X</sup>* −→ **<sup>I</sup>** } (2.1)
a fuzzy set *A* in a reference set *X* can be written as *μA*(*x*) or A(*x*) where *x* ∈ *X*.
value of A(*x*) to unity, the higher the grade of membership of *x* in *A*.
In Section 5 we make some analysis about the essence of fuzziness from the formal point of view. We try to find the original point where fuzziness appears and how it "moves" from its hiding-place making some concepts fuzzy.
In Section 6 we give the definition of *propositional language* by introducing its alphabet and how the expressions, i.e., *wellformed formulas* (or *formulas*, for short) can be formed from the alphabet. This formal language can be used as classical propositional logic and as many-valued propositional logic, too. We do not consider any other logical properties here, because they are not necessary for our purpose. In addition to the formal language, only the concept *valuation* and *truth-function* are needed. About the truth value evaluation, we consider the common things for several logics. The counterparts are obtainable also from Zadeh algebra. We also construct a *propositional algebra* that appears to be a Zadeh algebra.
In Section 7 an important logic for fuzzy sets is Kleene's 3-valued logic, as we already noticed above. Hence, the consideration of this logic deserves its own section. We tell about Kleene's motivation for constructing his 3-valued logic and give the truth value evaluation rules for the basic connectives. These rules fit completely well to the fuzzy set operations Zadeh introduced. We also explain the connections between standard fuzzy sets and this logic from Zadeh's point of view. In the end of this section, we give a short description about *Kleene-Dienes many-valued logic* that is an extension of Kleene's 3-valued logic into infinite-valued logic.
In Section 8 we consider the main features of Łukasiewicz ifinite-valued logic. Our main problem is included in this section. Łukasiewicz chose the connectives negation and implication as primitive connectives and derived the connectives conjunction, disjunction, and equivalence from these primitives. This starting point does not fit together with the operations of Zadeh algebra. Only the counterpart of negation (the complementarity operation) is included in Zadeh algebra but implication does not appear in it. in Łukasiewicz logic the two other connectives, disjunction and conjunction, belongs to the derived connectives. But they have such a form that their truth value evaluation rules are exactly the same as the corresponding operations in Zadeh algebra. So, using the set negation, disjunction, and conjunction of Łukasiewicz logic's connectives, we have to derive the connective Łukasiewicz implication. Actually, for this task we need only negation and disjunction, as is seen in Proposition 8.2 and its proof. Our final result is presented in Proposition 8.3. Some considerations on this topic can be found in Mattila (Mattila, 2005).
In Section 9 we consider briefly MV-algebras and give some hints how the connection between standard fuzzy sets and Łukasiewicz logic can be found. MV-algebras and their applications to fuzzy set theory and soft computing are widely studied, and the study of this topic actually forms a mainstream in this research area. Three books are mentioned in References representing this topic, namely M. Bergmann (Bergmann, 2008), R. L. O. Cignoli et al. (Cignoli et al., 2000), and P. Hájek (Hájek, 1998). These books belongs to a quite central literature of the topic.
MV-algebras are more general than De Morgan algebras, but formally it can be proved that De Morgan algebras belong to MV-algebras as a special case. But according to our problem, the used ways to apply general MV-algebras seems to give a circuitous route rather than a straightforward bridge between standard fuzzy set theory and Łukasiewicz logic.
In Section 10 we point out the main results and other concluding remarks.
### **2. Zadeh's theory of standard fuzzy sets**
For considering the standard system of fuzzy sets, the range of fuzzy sets (i.e., that of membership functions) is the unit interval **I** = [0, 1]. We give the definition of the concept *fuzzy set* using Zadeh's original definition. However, some symbols have been changed. Usually, the symbol of a fuzzy set, in general, is denoted by *μ*. A membership function of a fuzzy set *A* in a reference set *X* can be written as *μA*(*x*) or A(*x*) where *x* ∈ *X*.
**Definition 2.1** (Standard fuzzy set)**.** A *fuzzy subset A* of a set *X* is characterized by a *membership function* A(*x*) which associates with each point *x* in *X* a real number in the interval [0, 1], with the value of A(*x*) at *x* representing the "grade of membership" of *x* in *A*. Thus, the nearer the value of A(*x*) to unity, the higher the grade of membership of *x* in *A*.
This definition means that a fuzzy subset *A* of a universe of discourse *X* is represented by a function
$$\mathcal{A}: \mathcal{X} \longrightarrow \mathbf{I}.$$
The *power set* of all fuzzy subsets of the set *X* is
$$\mathbf{I}^{X} = \{ \mathcal{A} \mid \mathcal{A} : X \longrightarrow \mathbf{I} \}\tag{2.1}$$
An important subset of the set of all membership functions (2.1) is the set of functions taking only values 1 or 0, i.e., the set of all characteristic functions of the crisp subsets of *X*
$$\mathfrak{2}^X = \{ f \mid f: X \longrightarrow \{ 0, 1 \} \}$$
as a special case.
2 Will-be-set-by-IN-TECH
In Section 5 we make some analysis about the essence of fuzziness from the formal point of view. We try to find the original point where fuzziness appears and how it "moves" from its
In Section 6 we give the definition of *propositional language* by introducing its alphabet and how the expressions, i.e., *wellformed formulas* (or *formulas*, for short) can be formed from the alphabet. This formal language can be used as classical propositional logic and as many-valued propositional logic, too. We do not consider any other logical properties here, because they are not necessary for our purpose. In addition to the formal language, only the concept *valuation* and *truth-function* are needed. About the truth value evaluation, we consider the common things for several logics. The counterparts are obtainable also from Zadeh algebra. We also construct a *propositional algebra* that appears to be a Zadeh algebra. In Section 7 an important logic for fuzzy sets is Kleene's 3-valued logic, as we already noticed above. Hence, the consideration of this logic deserves its own section. We tell about Kleene's motivation for constructing his 3-valued logic and give the truth value evaluation rules for the basic connectives. These rules fit completely well to the fuzzy set operations Zadeh introduced. We also explain the connections between standard fuzzy sets and this logic from Zadeh's point of view. In the end of this section, we give a short description about *Kleene-Dienes many-valued logic* that is an extension of Kleene's 3-valued logic into
In Section 8 we consider the main features of Łukasiewicz ifinite-valued logic. Our main problem is included in this section. Łukasiewicz chose the connectives negation and implication as primitive connectives and derived the connectives conjunction, disjunction, and equivalence from these primitives. This starting point does not fit together with the operations of Zadeh algebra. Only the counterpart of negation (the complementarity operation) is included in Zadeh algebra but implication does not appear in it. in Łukasiewicz logic the two other connectives, disjunction and conjunction, belongs to the derived connectives. But they have such a form that their truth value evaluation rules are exactly the same as the corresponding operations in Zadeh algebra. So, using the set negation, disjunction, and conjunction of Łukasiewicz logic's connectives, we have to derive the connective Łukasiewicz implication. Actually, for this task we need only negation and disjunction, as is seen in Proposition 8.2 and its proof. Our final result is presented in Proposition 8.3. Some considerations on this topic can be found in Mattila (Mattila, 2005).
In Section 9 we consider briefly MV-algebras and give some hints how the connection between standard fuzzy sets and Łukasiewicz logic can be found. MV-algebras and their applications to fuzzy set theory and soft computing are widely studied, and the study of this topic actually forms a mainstream in this research area. Three books are mentioned in References representing this topic, namely M. Bergmann (Bergmann, 2008), R. L. O. Cignoli et al. (Cignoli et al., 2000), and P. Hájek (Hájek, 1998). These books belongs to a quite central literature of the
MV-algebras are more general than De Morgan algebras, but formally it can be proved that De Morgan algebras belong to MV-algebras as a special case. But according to our problem, the used ways to apply general MV-algebras seems to give a circuitous route rather than a
straightforward bridge between standard fuzzy set theory and Łukasiewicz logic.
In Section 10 we point out the main results and other concluding remarks.
hiding-place making some concepts fuzzy.
infinite-valued logic.
topic.
It is also a well-known fact that **I** and **I***<sup>X</sup>* are partially ordered sets. (Actually, **I** is a totally ordered set, but hence it is also prtially ordered.) In fact, they are also distributive complete lattices. Generally, some main properties of **I** can be embedded to **I***<sup>X</sup>* (*cf.* e.g. Lowen (Lowen, 1996)).
We consider operations, properties, and some concepts involved in fuzzy sets given by Zadeh (Zadeh, 1965).
**Definition 2.2** (Basic operations)**.** Let <sup>A</sup>, <sup>B</sup> <sup>∈</sup> **<sup>I</sup>***<sup>X</sup>* and *<sup>x</sup>* <sup>∈</sup> *<sup>X</sup>*. In **<sup>I</sup>***<sup>X</sup>* there are defined the following operations:
Two fuzzy sets <sup>A</sup>, <sup>B</sup> <sup>∈</sup> **<sup>I</sup>***<sup>X</sup>* are *equal*, denoted by <sup>A</sup> <sup>=</sup> <sup>B</sup>, if
$$\forall \mathfrak{x} \in X, \quad \mathcal{A}(\mathfrak{x}) = \mathcal{B}(\mathfrak{x}).$$
A fuzzy set *A* is *contained* in a fuzzy set *B*, i.e., *A* is a *subset* of *B*, denoted by A ⊆ B, if their membership functions satisfy the condition
$$\forall \, \mathbf{x} \in \mathbf{X}, \quad \mathcal{A}(\mathbf{x}) \le \mathcal{B}(\mathbf{x})$$
Consider the unit interval lattice **I** = ([0, 1], ≤). Sometimes we write **I** = [0, 1], for short. As
Standard Fuzzy Sets and some Many-Valued Logics 79
Hence, we can write the lattice **I** into the form **I** = ([0, 1], ∨, ∧). We will prove it is a distributive lattice. We consider it in the proof of Theorem 3.1 when we prove that **I** forms a De Morgan algebra. Especially, the order relation ≤ is a total order on [0, 1] because it is an order and any two elements from the interval [0, 1] are comparable with each other under it,
The interval [0, 1] is a metric space with the natural metric *distance* between two points of [0, 1]
We will see that this equality measure can be used in Łukasiewicz infinite-valued logic as the
**Theorem 3.1.** *The system L***<sup>I</sup>** = �**I**, ∨, ∧, ¬, 0, 1� *is De Morgan algebra, where for all x* ∈ [0, 1]*,*
*Proof.* First, we show that **I** is a distributive lattice. It is clear that **I** is a lattice. For showing distributivity, we choose arbitrarily elements *a*, *b*, *c* ∈ [0, 1]. Without loss of generality, we can
(*<sup>a</sup>* <sup>∨</sup> *<sup>b</sup>*) <sup>∧</sup> (*<sup>a</sup>* <sup>∨</sup> *<sup>c</sup>*) = *<sup>b</sup>* <sup>∧</sup> *<sup>c</sup>* <sup>=</sup> *<sup>b</sup>* <sup>=</sup><sup>⇒</sup> *<sup>a</sup>* <sup>∨</sup> (*<sup>b</sup>* <sup>∧</sup> *<sup>c</sup>*)=(*<sup>a</sup>* <sup>∨</sup> *<sup>b</sup>*) <sup>∧</sup> (*<sup>a</sup>* <sup>∨</sup> *<sup>c</sup>*)
Similarly, we have *a* ∧ (*b* ∨ *c*)=(*a* ∧ *b*) ∨ (*a* ∧ *c*). Hence, **I** = ([0, 1], ∨, ∧) is a distributive
¬¬*a* = 1 − (1 − *a*) = *a*
<sup>¬</sup>*<sup>a</sup>* ∧ ¬*<sup>b</sup>* = (<sup>1</sup> <sup>−</sup> *<sup>a</sup>*) <sup>∧</sup> (<sup>1</sup> <sup>−</sup> *<sup>b</sup>*) = <sup>1</sup> <sup>−</sup> *<sup>b</sup>* if *<sup>a</sup>* <sup>≤</sup> *<sup>b</sup>* <sup>=</sup>⇒ ¬(*<sup>a</sup>* <sup>∨</sup> *<sup>b</sup>*) = <sup>¬</sup>*<sup>a</sup>* ∧ ¬*<sup>b</sup>*
*x* ∨ *y* = *y x* ∧ *y* = *x*
*d*(*x*, *y*) = |*x* − *y*| , *x*, *y* ∈ [0, 1] (3.2)
*x* ∨ *y* = max{*x*, *y*} (3.3) *x* ∧ *y* = min{*x*, *y*} (3.4)
*L***<sup>I</sup>** = �**I**, max, min, ¬, 0, 1� (3.5)
(3.1)
is well known, the order relation ≤ and the operations ∨ and ∧ have the connection
<sup>∀</sup>*x*, *<sup>y</sup>* <sup>∈</sup> *<sup>X</sup>*, *<sup>x</sup>* <sup>≤</sup> *<sup>y</sup>* ⇐⇒
i.e., for any *x*, *y* ∈ **I**, we can state whether the order *x* ≤ *y* either holds or not.
given by the condition
¬*x* = 1 − *x.*
lattice.
evaluation rule for the connective *equivalency*.
suppose that *a* ≤ *b* ≤ *c*. Then, by (3.1) we have
*a* ∨ (*b* ∧ *c*) = *a* ∨ *b* = *b*
(DM1) holds because for all *a* ∈ [0, 1],
(DM2) holds because for all *a*, *b* ∈ [0, 1],
Hence, by Def. 3.1, *L***<sup>I</sup>** is a De Morgan algebra.
¬(*a* ∨ *b*) = 1 − (*a* ∨ *b*) = 1 − *b* if *a* ≤ *b*
From the ordering property (3.1) it follows that for all *x*, *y* ∈ **I**
Hence, we can express the algebra of Theorem 3.1 in the form
Zadeh also shows that the operations max and min are associative, distributive to each other, and De Morgan's laws hold, and they have the form
$$1 - \min\{\mathcal{A}(\mathbf{x}), \mathcal{B}(\mathbf{x})\} = \max\{1 - \mathcal{A}(\mathbf{x}), 1 - \mathcal{B}(\mathbf{x})\}\tag{2.2}$$
$$1 - \max\{\mathcal{A}(\mathbf{x}), \mathcal{B}(\mathbf{x})\} = \min\{1 - \mathcal{A}(\mathbf{x}), 1 - \mathcal{B}(\mathbf{x})\}\tag{2.3}$$
Actually, Zadeh gives the building materials for an algebra in his paper (Zadeh, 1965). However, he did not think any algebras when he created his paper "Fuzzy Sets". He thought the problem from another point of view. We return to this matter in the end of Section 4.
Finally, we present the following theorem due to C. V. Negoit ˘a and D. A. Ralescu (Negoit ˘a & Ralescu, 1975).
**Theorem 2.1.** *The set* **I***<sup>X</sup> is a complete distributive lattice.*
*Proof.* The reference set *X* has the membership function
$$
\mu\_X(\mathfrak{x}) = 1, \quad \mathfrak{x} \in X
$$
and the empty set ∅ the membership function
$$
\mu\_{\mathcal{D}}(\mathfrak{x}) = 0, \quad \mathfrak{x} \in X.
$$
This corresponds to the fact that **<sup>1</sup>**, **<sup>0</sup>** <sup>∈</sup> **<sup>I</sup>***<sup>X</sup>* where **<sup>1</sup>**(*x*) = 1 and **<sup>0</sup>**(*x*) = 0 for any *<sup>x</sup>* <sup>∈</sup> *<sup>X</sup>*. Hence, the result follows by the definition of complete lattice and the order properties of the unit interval.
### **3. On De Morgan algebras**
To get an algebra of standard fuzzy sets we start by considering the concept of De Morgan algebras. The main source is Helena Rasiowa's book (Rasiowa, 1974).
**Definition 3.1** (De Morgan algebra)**.** An abstract algebra A = �*A*, ∨, ∧, ¬, **1**� is called *De Morgan algebra*, if (*A*, ∨, ∧) is a distributive lattice with unit element **1** (the neutral element of ∧ operation), and ¬ is a unary operation on *A* satisfying the following conditions:
(DM1) for all *a* ∈ *A*, ¬¬*a* = *a*,
$$\text{(DM2)}\qquad\text{for all }a,b\in A,\quad\neg(a\lor b)=\neg a\land\neg b.$$
It is easy to prove that in any De Morgan algebra �*A*, ∨, ∧, ¬, **1**� the following properties hold:
(DM3) there is a zero element **0** (the neutral element of ∨ operation),
(DM4) ¬**0** = **1** and ¬**1** = **0**,
(DM5) ¬(*a* ∧ *b*) = ¬*a* ∨ ¬*b*.
The unit element is the greatest element and the zero element the least element of *A*. By (DM3), we sometimes add the zero element of a De Morgan algebra into the component list of the entities belonging to the algebra: A = �*A*, ∨, ∧, ¬, **0**, **1**�.
4 Will-be-set-by-IN-TECH
Zadeh also shows that the operations max and min are associative, distributive to each other,
Actually, Zadeh gives the building materials for an algebra in his paper (Zadeh, 1965). However, he did not think any algebras when he created his paper "Fuzzy Sets". He thought the problem from another point of view. We return to this matter in the end of Section 4.
Finally, we present the following theorem due to C. V. Negoit ˘a and D. A. Ralescu (Negoit ˘a &
*μX*(*x*) = 1, *x* ∈ *X*
*μ*∅(*x*) = 0, *x* ∈ *X*.
This corresponds to the fact that **<sup>1</sup>**, **<sup>0</sup>** <sup>∈</sup> **<sup>I</sup>***<sup>X</sup>* where **<sup>1</sup>**(*x*) = 1 and **<sup>0</sup>**(*x*) = 0 for any *<sup>x</sup>* <sup>∈</sup> *<sup>X</sup>*. Hence, the result follows by the definition of complete lattice and the order properties of the
To get an algebra of standard fuzzy sets we start by considering the concept of De Morgan
**Definition 3.1** (De Morgan algebra)**.** An abstract algebra A = �*A*, ∨, ∧, ¬, **1**� is called *De Morgan algebra*, if (*A*, ∨, ∧) is a distributive lattice with unit element **1** (the neutral element
It is easy to prove that in any De Morgan algebra �*A*, ∨, ∧, ¬, **1**� the following properties hold:
The unit element is the greatest element and the zero element the least element of *A*. By (DM3), we sometimes add the zero element of a De Morgan algebra into the component list
of ∧ operation), and ¬ is a unary operation on *A* satisfying the following conditions:
algebras. The main source is Helena Rasiowa's book (Rasiowa, 1974).
(DM3) there is a zero element **0** (the neutral element of ∨ operation),
of the entities belonging to the algebra: A = �*A*, ∨, ∧, ¬, **0**, **1**�.
1 − min{A(*x*), B(*x*)} = max{1 − A(*x*), 1 − B(*x*)} (2.2)
1 − max{A(*x*), B(*x*)} = min{1 − A(*x*), 1 − B(*x*)} (2.3)
and De Morgan's laws hold, and they have the form
**Theorem 2.1.** *The set* **I***<sup>X</sup> is a complete distributive lattice.*
*Proof.* The reference set *X* has the membership function
and the empty set ∅ the membership function
Ralescu, 1975).
unit interval.
**3. On De Morgan algebras**
(DM1) for all *a* ∈ *A*, ¬¬*a* = *a*,
(DM4) ¬**0** = **1** and ¬**1** = **0**, (DM5) ¬(*a* ∧ *b*) = ¬*a* ∨ ¬*b*.
(DM2) for all *a*, *b* ∈ *A*, ¬(*a* ∨ *b*) = ¬*a* ∧ ¬*b*.
Consider the unit interval lattice **I** = ([0, 1], ≤). Sometimes we write **I** = [0, 1], for short. As is well known, the order relation ≤ and the operations ∨ and ∧ have the connection
$$\forall \mathbf{x}, y \in \mathbf{X}, \; \mathbf{x} \le y \iff \begin{cases} \mathbf{x} \lor y = y \\ \mathbf{x} \land y = \mathbf{x} \end{cases} \tag{3.1}$$
Hence, we can write the lattice **I** into the form **I** = ([0, 1], ∨, ∧). We will prove it is a distributive lattice. We consider it in the proof of Theorem 3.1 when we prove that **I** forms a De Morgan algebra. Especially, the order relation ≤ is a total order on [0, 1] because it is an order and any two elements from the interval [0, 1] are comparable with each other under it, i.e., for any *x*, *y* ∈ **I**, we can state whether the order *x* ≤ *y* either holds or not.
The interval [0, 1] is a metric space with the natural metric *distance* between two points of [0, 1] given by the condition
$$d(\mathbf{x}, y) = |\mathbf{x} - y|\,,\quad \mathbf{x}, y \in [0, 1] \tag{3.2}$$
We will see that this equality measure can be used in Łukasiewicz infinite-valued logic as the evaluation rule for the connective *equivalency*.
**Theorem 3.1.** *The system L***<sup>I</sup>** = �**I**, ∨, ∧, ¬, 0, 1� *is De Morgan algebra, where for all x* ∈ [0, 1]*,* ¬*x* = 1 − *x.*
*Proof.* First, we show that **I** is a distributive lattice. It is clear that **I** is a lattice. For showing distributivity, we choose arbitrarily elements *a*, *b*, *c* ∈ [0, 1]. Without loss of generality, we can suppose that *a* ≤ *b* ≤ *c*. Then, by (3.1) we have
$$\begin{cases} a \lor (b \land c) = a \lor b = b\\ (a \lor b) \land (a \lor c) = b \land c = b \end{cases} \implies a \lor (b \land c) = (a \lor b) \land (a \lor c)$$
Similarly, we have *a* ∧ (*b* ∨ *c*)=(*a* ∧ *b*) ∨ (*a* ∧ *c*). Hence, **I** = ([0, 1], ∨, ∧) is a distributive lattice.
(DM1) holds because for all *a* ∈ [0, 1],
$$\neg\neg a = 1 - (1 - a) = a$$
(DM2) holds because for all *a*, *b* ∈ [0, 1],
$$\begin{cases} \neg(a \lor b) = 1 - (a \lor b) = 1 - b & \text{if } a \le b \\\neg a \land \neg b = (1 - a) \land (1 - b) = 1 - b & \text{if } a \le b \end{cases} \implies \neg(a \lor b) = \neg a \land \neg b$$
Hence, by Def. 3.1, *L***<sup>I</sup>** is a De Morgan algebra.
From the ordering property (3.1) it follows that for all *x*, *y* ∈ **I**
$$\mathbf{x} \lor \mathbf{y} = \max\{\mathbf{x}, \mathbf{y}\} \tag{3.3}$$
$$x \wedge y = \min\{x, y\} \tag{3.4}$$
Hence, we can express the algebra of Theorem 3.1 in the form
$$L\_{\mathbf{I}} = \langle \mathbf{I}, \max, \min, \neg, 0, 1 \rangle \tag{3.5}$$
is given in the proof of Theorem 4.1.) This thing is analogous to the classical set complement expressed by subtraction a set *A* to be complemented from the universe of discourse *X*, i.e.,
Standard Fuzzy Sets and some Many-Valued Logics 81
The operations max and min are clearly commutative. Based on the fact that the algebra (3.7) is De Morgan algebra, the algebra (4.1) is De Morgan algebra, too. We call this algebra *Zadeh algebra* because it is an algebraic description of standard fuzzy set theory, similarly as in classical set theory, a certain Boolean algebra (set algebra or algebra of characteristic functions) is the algebraic description of the system of classical sets. Now, we have the following
**Theorem 4.1.** *Zadeh algebra* <sup>Z</sup> <sup>=</sup> �**I***X*, max, min, <sup>¬</sup>, **<sup>0</sup>**, **<sup>1</sup>**� *is an algebraic approach to standard fuzzy*
(i) The operations max and min are exactly the same as in Zadeh's theory by Def. 2.2.
max{*μ*, *ν*} = max{*ν*, *μ*} and max{*μ*, max{*ν*, *τ*}} = max{max{*μ*, *ν*}, *τ*}
for all *<sup>μ</sup>*, *<sup>ν</sup>*, *<sup>τ</sup>* <sup>∈</sup> **<sup>I</sup>***<sup>X</sup>* because these laws clearly hold for the elements of **<sup>I</sup>**, and these laws can be embedded to **<sup>I</sup>***<sup>X</sup>* by pointwice calculation of values of the functions *<sup>μ</sup>* <sup>∈</sup> **<sup>I</sup>***<sup>X</sup>* (*cf.*
(max{*μ*, **0**})(*x*) = max{*μ*(*x*), **0**(*x*)} = max{*μ*(*x*), 0} = *μ*(*x*)
(min{*μ*, **0**})(*x*) = *μ*(*x*)
(¬*μ*)(*x*)=(**1** − *μ*)(*x*) = **1**(*x*) − *μ*(*x*) = 1 − *μ*(*x*)
taking values from the unit interval [0, 1]. Hence, <sup>¬</sup>*<sup>μ</sup>* <sup>∈</sup> **<sup>I</sup>***X*, and <sup>¬</sup> is the complementarity
(vi) Clearly, Zadeh algebra <sup>Z</sup> satisfies the condition **<sup>0</sup>** �<sup>=</sup> **<sup>1</sup>**, by (iv). Hence, **<sup>2</sup>***<sup>X</sup>* <sup>⊂</sup> **<sup>I</sup>***X*. The constant functions **0** and **1** are the zero element and unit element of the algebra.
In classical set theory, an element either is or is not an element of a given set. In fuzzy set theory, we have three possibilities: a membership grade of an element in a given fuzzy set
For practical use, we may postulate Zadeh algebra by collecting the nevessary properties together. This means that we build Theor. 4.1 again using the main laws and properties
(v) For any membership function *<sup>μ</sup>* <sup>∈</sup> **<sup>I</sup>***X*, there exists <sup>¬</sup>*<sup>μ</sup>* <sup>∈</sup> **<sup>I</sup>***X*, such that for any *<sup>x</sup>* <sup>∈</sup> *<sup>X</sup>*,
(iii) From Theorem 2.1, distributive laws follows for max and min on **I***<sup>X</sup>* because (**I***X*, max, min) is a distributive lattice and Zaheh-algebra (4.1) is De Morgan algebra.
(ii) The operations max and min are commutative and associative on **I***X*, i.e.,
Lowen (Lowen, 1996)). The same properties hold for min, too.
(iv) For all *<sup>μ</sup>* <sup>∈</sup> **<sup>I</sup>***X*, max{*μ*, **<sup>0</sup>**} <sup>=</sup> *<sup>μ</sup>* and min{*μ*, **<sup>1</sup>**} <sup>=</sup> *<sup>μ</sup>*, because for any *<sup>x</sup>* <sup>∈</sup> *<sup>X</sup>*,
Zadeh (Zadeh, 1965) has also proved these laws.
Similarly, for any *x* ∈ *X*,
operation of Zadeh's theory.
equals either to zero or one, or is between them.
like postulates. The result is as follows.
This competes the proof.
*<sup>A</sup><sup>c</sup>* <sup>=</sup> *<sup>X</sup> <sup>A</sup>* where *<sup>A</sup>* <sup>⊂</sup> *<sup>X</sup>* and *<sup>A</sup><sup>c</sup>* is the complement of *<sup>A</sup>*.
*set theory.*
*Proof.*
Let *X* be a nonempty set. Consider a set of functions *μ* : *X* −→ **I**, i.e., the function set
$$\mathbb{I}^{X} = \{ \mu \mid \mu: X \longrightarrow \mathbb{I} \}\tag{3.6}$$
We extend the algebra of Theorem 3.1 into an *algebra of functions* (3.6)
$$L\_{\mathbf{I}^X} = \langle \mathbf{I}^X, \vee, \wedge, \neg, \mathbf{0}, \mathbf{1} \rangle \tag{3.7}$$
by pointwise calculation. Here **0** and **1** are constant functions, such that
$$\forall \mathbf{x} \in \mathbf{X}, \quad \mathbf{0}: \mathbf{x} \mapsto \mathbf{0}, \quad \mathbf{1}: \mathbf{x} \mapsto \mathbf{1} \tag{3.8}$$
The algebra (3.7) is a De Morgan algebra by its construction. This means that we calculate expressions *μ*(*x*) ∨ *ν*(*x*), *μ*(*x*) ∧ *ν*(*x*), ¬*μ*(*x*) etc. pointwise for any *x* ∈ *X*. Hence, the formulas (3.3) and (3.4) are applicable also in the function algebra (3.7).
As a special case, the algebra (3.7) has a subalgebra
$$L\_{\{0,1\}^X} = \langle \{0,1\}^X \rangle \text{max} \langle \text{min} \rangle \ \neg \text{,0} \ \mathbf{1} \rangle \tag{3.9}$$
being an algebra of characteristic functions of classical sets, *f* : *X* −→ {0, 1}. Sometimes we write **2** instead of {0, 1}, so, especially,
$$\mathbf{2}^{X} = \{ f\_{A} \mid f\_{A} : X \longrightarrow \{ 0, 1 \}, A \subset X \}$$
is the classical power set of a set *X* expressed by characteristic functions. The characteristic function of a given set *A* ⊂ *X*, *fA*, is the function
$$f\_A(\mathfrak{x}) = \begin{cases} 1 & \text{if} \quad \mathfrak{x} \in A\_{\sigma} \\ 0 & \text{if} \quad \mathfrak{x} \notin A\_{\sigma} \end{cases}$$
This function indicates by the value *fA*(*x*) = 1 that the element *x* ∈ *X* is an element of *A* and all the elements of *X* having the value *fA*(*x*) = 0 are elements of the complement of *A*. As a subalgebra of the algebra (3.7), the algebra (3.9) is a special De Morgan algebra, namely a *Boolean algebra*.
### **4. Algebra of standard fuzzy sets**
Consider the algebra (3.7). We may give a new label to it and use operation symbols max and min instead of ∨ and ∧, respectively, by the formulas (3.3) and (3.4). Hence, we have
$$\mathcal{Z}\_{\mathbb{N}\_1} = \langle \mathbb{I}^X, \mathbf{max}, \mathbf{min}, \neg, \mathbf{0}, \mathbf{1} \rangle \tag{4.1}$$
The subscript <sup>ℵ</sup><sup>1</sup> means the cardinality of continuum, so, **<sup>I</sup>***<sup>X</sup>* is a continuum because **<sup>I</sup>** is continuum, too. For short, we may refer to Zℵ<sup>1</sup> by Z, without the subscript, if there is no possibility for confusion. The complementarity operation ¬ is a mapping
$$\neg : \mathbb{I}^X \longrightarrow \mathbb{I}^X \text{, } \mu \mapsto \mathbf{1} - \mu \tag{4.2}$$
Hence, the complement function of a function *μ* is **1** − *μ*, such that for all *x* ∈ *X*, (**1** − *μ*)(*x*) = **1**(*x*) − *μ*(*x*) = 1− *μ*(*x*). (The proof, that ¬ defined in this way is really a membership function, is given in the proof of Theorem 4.1.) This thing is analogous to the classical set complement expressed by subtraction a set *A* to be complemented from the universe of discourse *X*, i.e., *<sup>A</sup><sup>c</sup>* <sup>=</sup> *<sup>X</sup> <sup>A</sup>* where *<sup>A</sup>* <sup>⊂</sup> *<sup>X</sup>* and *<sup>A</sup><sup>c</sup>* is the complement of *<sup>A</sup>*.
The operations max and min are clearly commutative. Based on the fact that the algebra (3.7) is De Morgan algebra, the algebra (4.1) is De Morgan algebra, too. We call this algebra *Zadeh algebra* because it is an algebraic description of standard fuzzy set theory, similarly as in classical set theory, a certain Boolean algebra (set algebra or algebra of characteristic functions) is the algebraic description of the system of classical sets. Now, we have the following
**Theorem 4.1.** *Zadeh algebra* <sup>Z</sup> <sup>=</sup> �**I***X*, max, min, <sup>¬</sup>, **<sup>0</sup>**, **<sup>1</sup>**� *is an algebraic approach to standard fuzzy set theory.*
*Proof.*
6 Will-be-set-by-IN-TECH
The algebra (3.7) is a De Morgan algebra by its construction. This means that we calculate expressions *μ*(*x*) ∨ *ν*(*x*), *μ*(*x*) ∧ *ν*(*x*), ¬*μ*(*x*) etc. pointwise for any *x* ∈ *X*. Hence, the formulas
being an algebra of characteristic functions of classical sets, *f* : *X* −→ {0, 1}. Sometimes we
**<sup>2</sup>***<sup>X</sup>* <sup>=</sup> { *fA* <sup>|</sup> *fA* : *<sup>X</sup>* −→ {0, 1}, *<sup>A</sup>* <sup>⊂</sup> *<sup>X</sup>*}
is the classical power set of a set *X* expressed by characteristic functions. The characteristic
This function indicates by the value *fA*(*x*) = 1 that the element *x* ∈ *X* is an element of *A* and all the elements of *X* having the value *fA*(*x*) = 0 are elements of the complement of *A*. As a subalgebra of the algebra (3.7), the algebra (3.9) is a special De Morgan algebra, namely a
Consider the algebra (3.7). We may give a new label to it and use operation symbols max and
The subscript <sup>ℵ</sup><sup>1</sup> means the cardinality of continuum, so, **<sup>I</sup>***<sup>X</sup>* is a continuum because **<sup>I</sup>** is continuum, too. For short, we may refer to Zℵ<sup>1</sup> by Z, without the subscript, if there is no
Hence, the complement function of a function *μ* is **1** − *μ*, such that for all *x* ∈ *X*, (**1** − *μ*)(*x*) = **1**(*x*) − *μ*(*x*) = 1− *μ*(*x*). (The proof, that ¬ defined in this way is really a membership function,
min instead of ∨ and ∧, respectively, by the formulas (3.3) and (3.4). Hence, we have
possibility for confusion. The complementarity operation ¬ is a mapping
1 if *x* ∈ *A*, 0 if *x* ∈/ *A*
*fA*(*x*) =
**<sup>I</sup>***<sup>X</sup>* <sup>=</sup> {*<sup>μ</sup>* <sup>|</sup> *<sup>μ</sup>* : *<sup>X</sup>* −→ **<sup>I</sup>**} (3.6)
*<sup>L</sup>***I***<sup>X</sup>* <sup>=</sup> �**I***X*, <sup>∨</sup>, <sup>∧</sup>, <sup>¬</sup>, **<sup>0</sup>**, **<sup>1</sup>**� (3.7)
∀*x* ∈ *X*, **0** : *x* �→ 0, **1** : *x* �→ 1 (3.8)
*<sup>L</sup>*{0,1}*<sup>X</sup>* <sup>=</sup> �{0, 1}*X*, max, min, <sup>¬</sup>, **<sup>0</sup>**, **<sup>1</sup>**� (3.9)
Zℵ<sup>1</sup> <sup>=</sup> �**I***X*, max, min, <sup>¬</sup>, **<sup>0</sup>**, **<sup>1</sup>**� (4.1)
<sup>¬</sup> : **<sup>I</sup>***<sup>X</sup>* −→ **<sup>I</sup>***X*, *<sup>μ</sup>* �→ **<sup>1</sup>** <sup>−</sup> *<sup>μ</sup>* (4.2)
Let *X* be a nonempty set. Consider a set of functions *μ* : *X* −→ **I**, i.e., the function set
We extend the algebra of Theorem 3.1 into an *algebra of functions* (3.6)
by pointwise calculation. Here **0** and **1** are constant functions, such that
(3.3) and (3.4) are applicable also in the function algebra (3.7).
As a special case, the algebra (3.7) has a subalgebra
function of a given set *A* ⊂ *X*, *fA*, is the function
write **2** instead of {0, 1}, so, especially,
**4. Algebra of standard fuzzy sets**
*Boolean algebra*.
max{*μ*, *ν*} = max{*ν*, *μ*} and max{*μ*, max{*ν*, *τ*}} = max{max{*μ*, *ν*}, *τ*}
for all *<sup>μ</sup>*, *<sup>ν</sup>*, *<sup>τ</sup>* <sup>∈</sup> **<sup>I</sup>***<sup>X</sup>* because these laws clearly hold for the elements of **<sup>I</sup>**, and these laws can be embedded to **<sup>I</sup>***<sup>X</sup>* by pointwice calculation of values of the functions *<sup>μ</sup>* <sup>∈</sup> **<sup>I</sup>***<sup>X</sup>* (*cf.* Lowen (Lowen, 1996)). The same properties hold for min, too.
$$(\max\{\mu, \mathbf{0}\})(\mathbf{x}) = \max\{\mu(\mathbf{x}), \mathbf{0}(\mathbf{x})\} = \max\{\mu(\mathbf{x}), \mathbf{0}\} = \mu(\mathbf{x})$$
Similarly, for any *x* ∈ *X*,
$$(\min\{\mu, \mathbf{0}\})(\mathbf{x}) = \mu(\mathbf{x})$$
(v) For any membership function *<sup>μ</sup>* <sup>∈</sup> **<sup>I</sup>***X*, there exists <sup>¬</sup>*<sup>μ</sup>* <sup>∈</sup> **<sup>I</sup>***X*, such that for any *<sup>x</sup>* <sup>∈</sup> *<sup>X</sup>*,
$$(\neg \mu)(\mathbf{x}) = (\mathbf{1} - \mu)(\mathbf{x}) = \mathbf{1}(\mathbf{x}) - \mu(\mathbf{x}) = 1 - \mu(\mathbf{x})$$
taking values from the unit interval [0, 1]. Hence, <sup>¬</sup>*<sup>μ</sup>* <sup>∈</sup> **<sup>I</sup>***X*, and <sup>¬</sup> is the complementarity operation of Zadeh's theory.
(vi) Clearly, Zadeh algebra <sup>Z</sup> satisfies the condition **<sup>0</sup>** �<sup>=</sup> **<sup>1</sup>**, by (iv). Hence, **<sup>2</sup>***<sup>X</sup>* <sup>⊂</sup> **<sup>I</sup>***X*. The constant functions **0** and **1** are the zero element and unit element of the algebra.
This competes the proof.
In classical set theory, an element either is or is not an element of a given set. In fuzzy set theory, we have three possibilities: a membership grade of an element in a given fuzzy set equals either to zero or one, or is between them.
For practical use, we may postulate Zadeh algebra by collecting the nevessary properties together. This means that we build Theor. 4.1 again using the main laws and properties like postulates. The result is as follows.
$$\mathbf{0}$$
Zadeh-algebra as a special case of De Morgan algebras give rise to closer analysis. Here we have done some part of it. The author thinks that Prof. Zadeh did not necessarily think about De Morgan algebras, when he created his crucial paper "Fuzzy Sets" (Zadeh, 1965). He thought the problem from another point of view, as can be seen in the construction of the paper. His leading idea was to model things in the eventful real world. In any way, it was a happy event that Prof. Zadeh's ideas met such a mathematical frame we have considered here. No others have been so successful to find such a *right interpretation* to some formal tools for modeling real world incidences. In the same time the *problem of interpretation* of many-valued logic got a solution. Many-valued logic began to give meaningful tools for analyzing and modeling things in real world. The role of many-valued logics were very nominal before Prof. Zadeh invented fuzzy set theory. After this, the study of many-valued logic met a new rise. Fuzzy set theory and fuzzy logic has helped the researchers to find new aspects from already
Standard Fuzzy Sets and some Many-Valued Logics 83
existing mathematical theories. This kind of work is now going on very strongly.
For example, imagine a set of beautiful women. Let us denote this set by *A*. There are women who do not belong to *A* with the highest grade 1. So, such a woman *does not* have some features which would make her beautiful. But she may have some of those features anyway. An intuitive hint about a possible answer to the question: "Where is the hiding-place of fuzziness?" can be found just on the second line above: "... *does not* ..." It seems that a partial
Let us compare Zadeh algebra with a general Boolean algebra with a supposition that the binary operations are associative because associativity holds in Zadeh algebra. The definition
**Definition 5.1.** Let ∧ (*meet*) and ∨ (*join*) be binary operations, and � (*complement*) a unary operation on a set *B*(�= ∅), and let **0** and **1** be the elements of *B*, such that the following
∀*x*, *y*, *z* ∈ *B*, *x* ∧ (*y* ∨ *z*)=(*x* ∧ *y*) ∨ (*x* ∧ *z*),
(BA4) ∀*x* ∈ *B*, *x* ∨ **0** = *x* and *x* ∧ **1** = *x*, i.e., **0** and **1** are the neutral elements (or identity
(BA5) For every element *x* ∈ *B* there exists an element *x*� ∈ *B*, such that *x* ∨ *x*� = **1** and
The only *structural difference* between these algebras is that between the axioms Z5 and BA 5. BA 5 is characteristic for complement operation, but Z5 does not satisfy the conditions of complement. So, fuzziness lies in the axiom Z5. The influence of this axiom is that also other
Hence, the set *B* together with these operations forms a *Boolean algebra* B = (*B*, ∨, ∧, �
*x* ∨ (*y* ∧ *z*)=(*x* ∨ *y*) ∧ (*x* ∨ *z*).
, **0**, **1**).
(BA1) ∧ and ∨ are commutative in *B*, i.e., ∀*x*, *y* ∈ *B*, *x* ∨ *y* = *y* ∨ *x* ja *x* ∧ *y* = *y* ∧ *x*;
**5. Where is the hiding-place of fuzziness?**
axioms hold:
*x* ∧ *x*� = **0**.
complementarity is somehow involved in this problem.
of this kind of Boolean algebra can be postulated as follows.
(BA2) The operations ∧ and ∨ are associative in *B*; (BA3) The operations ∧ and ∨ are distributive, i.e.,
elements) of the operations ∨ and ∧.
(BA6) For the elements **0** and **1** of *B* the condition **0** �= **1** holds.
**Proposition 4.1.** *Let* **<sup>I</sup>***<sup>X</sup>* <sup>=</sup> {*<sup>μ</sup>* <sup>|</sup> *<sup>μ</sup>* : *<sup>X</sup>* −→ **<sup>I</sup>**} *be the set of all functions from X to* **<sup>I</sup>***, where the operations* max *and* min *are pointwise defined between membership functions, and* ¬*μ* def = **1** − *μ. Then* <sup>Z</sup> <sup>=</sup> �**I***X*, max, min, <sup>¬</sup>, **<sup>0</sup>**, **<sup>1</sup>**� *is Zadeh algebra if it satisfies the conditions*
$$(\mathcal{Z}6) \quad \mathbf{0} \neq \mathbf{1}.$$
**Definition 4.1** (Kleene algebra)**.** De Morgan algebra is *Kleene algebra* if it satisfies the additional condition
(K) *x* ∧ ¬*x* ≤ *y* ∨ ¬*y*.
**Theorem 4.2.** *Zadeh algebra (4.1) is a Kleene algebra.*
*Proof.* Zadeh algebra is De Morgan algebra. The condition (K) in Zadeh algebra has the form
$$\min\{\mu, \neg \mu\} \le \max\{\nu, \neg \nu\}$$
for all *<sup>μ</sup>*, *<sup>ν</sup>* <sup>∈</sup> **<sup>I</sup>***X*.
To prove this, we can easily show that always min{*μ*, <sup>¬</sup>*μ*} ≤ <sup>1</sup> <sup>2</sup> and <sup>1</sup> <sup>2</sup> ≤ max{*ν*, ¬*ν*} for arbitrary *<sup>μ</sup>*, *<sup>ν</sup>* <sup>∈</sup> **<sup>I</sup>***X*, where the result follows immediately.
An alternative way is an easy task to check the four cases: (1◦) *<sup>μ</sup>* <sup>≤</sup> <sup>1</sup> <sup>2</sup> , *<sup>ν</sup>* <sup>≤</sup> <sup>1</sup> <sup>2</sup> , (2◦) *<sup>μ</sup>* <sup>≤</sup> <sup>1</sup> <sup>2</sup> , *ν* > 1 <sup>2</sup> , (3◦) *<sup>μ</sup>* <sup>&</sup>gt; <sup>1</sup> <sup>2</sup> , *<sup>ν</sup>* <sup>≤</sup> <sup>1</sup> <sup>2</sup> , and (4◦) *<sup>μ</sup>* <sup>&</sup>gt; <sup>1</sup> <sup>2</sup> , *<sup>ν</sup>* <sup>&</sup>gt; <sup>1</sup> <sup>2</sup> , and find out that each of these cases satisfies the condition (K).
Zadeh algebra Zℵ<sup>1</sup> has subalgebras which are Zadeh algebras, too. A range of membership functions can be a suitable subset of the unit interval [0, 1], such that the postulates of Prop. 4.1 are satisfied. Here the suitability means that the set is closed under the operations of the algebra.
**Example 4.1.** Consider a set *A* = {0, 1} which is a subset of [0, 1] consisting of the extreme cases of the unit interval. The algebra �*AX*, max, min, <sup>¬</sup>, **<sup>0</sup>**, **<sup>1</sup>**� satisfies the conditions of Zadeh algebra. This algebra is really an extreme case, because it is the Boolean algebra of characteristic functions of strict (i.e., usual) sets. It is a subalgebra of Zℵ<sup>1</sup> .
**Example 4.2.** Consider a set *<sup>A</sup>* <sup>=</sup> {0, <sup>1</sup> <sup>2</sup> , 1} being a subset of [0, 1]. The set *A* is the range of functions *<sup>μ</sup>* : *<sup>X</sup>* −→ *<sup>A</sup>* where *<sup>X</sup>* �<sup>=</sup> <sup>∅</sup> is a set. These functions belong to the set **<sup>I</sup>***X*, by means of which *A* is a subset of **I***X*. The conditions of Prop. 4.1 are clearly satisfied. Hence, <sup>Z</sup><sup>3</sup> <sup>=</sup> �*AX*, max, min, <sup>¬</sup>, **<sup>0</sup>**, **<sup>1</sup>**� is a 3-valued Zadeh algebra, and hence, a subalgebra of Zℵ<sup>1</sup> .
**Example 4.3.** Consider a set *A* consisting of all the rationals from the unit interval [0, 1]. The number of the elements of *A* is countable, but infinite. Hence, the cardinality of *A* is ℵ0. Making similar considerations as in the previous example, we verify that Zℵ<sup>0</sup> = �*AX*, max, min, <sup>¬</sup>, **<sup>0</sup>**, **<sup>1</sup>**� is a subalgebra of Zℵ<sup>1</sup> .
8 Will-be-set-by-IN-TECH
**Proposition 4.1.** *Let* **<sup>I</sup>***<sup>X</sup>* <sup>=</sup> {*<sup>μ</sup>* <sup>|</sup> *<sup>μ</sup>* : *<sup>X</sup>* −→ **<sup>I</sup>**} *be the set of all functions from X to* **<sup>I</sup>***, where the*
(Z4) *The neutral elements of the operations* max *and* min *are* **<sup>0</sup>** *and* **<sup>1</sup>***, respectively, i.e., for all <sup>μ</sup>* <sup>∈</sup> **<sup>I</sup>***X,*
**Definition 4.1** (Kleene algebra)**.** De Morgan algebra is *Kleene algebra* if it satisfies the additional
*Proof.* Zadeh algebra is De Morgan algebra. The condition (K) in Zadeh algebra has the form
min{*μ*, ¬*μ*} ≤ max{*ν*, ¬*ν*}
Zadeh algebra Zℵ<sup>1</sup> has subalgebras which are Zadeh algebras, too. A range of membership functions can be a suitable subset of the unit interval [0, 1], such that the postulates of Prop. 4.1 are satisfied. Here the suitability means that the set is closed under the operations of the
**Example 4.1.** Consider a set *A* = {0, 1} which is a subset of [0, 1] consisting of the extreme cases of the unit interval. The algebra �*AX*, max, min, <sup>¬</sup>, **<sup>0</sup>**, **<sup>1</sup>**� satisfies the conditions of Zadeh algebra. This algebra is really an extreme case, because it is the Boolean algebra of
of functions *<sup>μ</sup>* : *<sup>X</sup>* −→ *<sup>A</sup>* where *<sup>X</sup>* �<sup>=</sup> <sup>∅</sup> is a set. These functions belong to the set **<sup>I</sup>***X*, by means of which *A* is a subset of **I***X*. The conditions of Prop. 4.1 are clearly satisfied. Hence, <sup>Z</sup><sup>3</sup> <sup>=</sup> �*AX*, max, min, <sup>¬</sup>, **<sup>0</sup>**, **<sup>1</sup>**� is a 3-valued Zadeh algebra, and hence, a subalgebra of Zℵ<sup>1</sup> . **Example 4.3.** Consider a set *A* consisting of all the rationals from the unit interval [0, 1]. The number of the elements of *A* is countable, but infinite. Hence, the cardinality of *A* is ℵ0. Making similar considerations as in the previous example, we verify that Zℵ<sup>0</sup> =
<sup>2</sup> and <sup>1</sup>
<sup>2</sup> , and find out that each of these cases satisfies the
<sup>2</sup> , 1} being a subset of [0, 1]. The set *A* is the range
<sup>2</sup> , *<sup>ν</sup>* <sup>≤</sup> <sup>1</sup>
<sup>2</sup> ≤ max{*ν*, ¬*ν*} for
<sup>2</sup> , (2◦) *<sup>μ</sup>* <sup>≤</sup> <sup>1</sup>
<sup>2</sup> , *ν* >
(Z5) *For any membership function <sup>μ</sup>* <sup>∈</sup> **<sup>I</sup>***X, there exists* <sup>¬</sup>*<sup>μ</sup>* <sup>∈</sup> **<sup>I</sup>***X, such that* (¬*μ*) = <sup>1</sup> <sup>−</sup> *<sup>μ</sup>;*
def = **1** − *μ.*
*operations* max *and* min *are pointwise defined between membership functions, and* ¬*μ*
*Then* <sup>Z</sup> <sup>=</sup> �**I***X*, max, min, <sup>¬</sup>, **<sup>0</sup>**, **<sup>1</sup>**� *is Zadeh algebra if it satisfies the conditions*
(Z1) *The operations* max *and* min *are commutative on* **<sup>I</sup>***X;* (Z2) *The operations* max *and* min *are associative on* **<sup>I</sup>***X;*
max{*μ*, **0**} = *μ and* min{*μ*, **1**} = *μ;*
**Theorem 4.2.** *Zadeh algebra (4.1) is a Kleene algebra.*
To prove this, we can easily show that always min{*μ*, <sup>¬</sup>*μ*} ≤ <sup>1</sup>
An alternative way is an easy task to check the four cases: (1◦) *<sup>μ</sup>* <sup>≤</sup> <sup>1</sup>
characteristic functions of strict (i.e., usual) sets. It is a subalgebra of Zℵ<sup>1</sup> .
<sup>2</sup> , *<sup>ν</sup>* <sup>&</sup>gt; <sup>1</sup>
arbitrary *<sup>μ</sup>*, *<sup>ν</sup>* <sup>∈</sup> **<sup>I</sup>***X*, where the result follows immediately.
<sup>2</sup> , and (4◦) *<sup>μ</sup>* <sup>&</sup>gt; <sup>1</sup>
(Z6) **0** �= **1***.*
(K) *x* ∧ ¬*x* ≤ *y* ∨ ¬*y*.
for all *<sup>μ</sup>*, *<sup>ν</sup>* <sup>∈</sup> **<sup>I</sup>***X*.
<sup>2</sup> , (3◦) *<sup>μ</sup>* <sup>&</sup>gt; <sup>1</sup>
condition (K).
algebra.
<sup>2</sup> , *<sup>ν</sup>* <sup>≤</sup> <sup>1</sup>
**Example 4.2.** Consider a set *<sup>A</sup>* <sup>=</sup> {0, <sup>1</sup>
�*AX*, max, min, <sup>¬</sup>, **<sup>0</sup>**, **<sup>1</sup>**� is a subalgebra of Zℵ<sup>1</sup> .
1
condition
(Z3) *The operations* max *and* min *are distributive to each other;*
Zadeh-algebra as a special case of De Morgan algebras give rise to closer analysis. Here we have done some part of it. The author thinks that Prof. Zadeh did not necessarily think about De Morgan algebras, when he created his crucial paper "Fuzzy Sets" (Zadeh, 1965). He thought the problem from another point of view, as can be seen in the construction of the paper. His leading idea was to model things in the eventful real world. In any way, it was a happy event that Prof. Zadeh's ideas met such a mathematical frame we have considered here. No others have been so successful to find such a *right interpretation* to some formal tools for modeling real world incidences. In the same time the *problem of interpretation* of many-valued logic got a solution. Many-valued logic began to give meaningful tools for analyzing and modeling things in real world. The role of many-valued logics were very nominal before Prof. Zadeh invented fuzzy set theory. After this, the study of many-valued logic met a new rise. Fuzzy set theory and fuzzy logic has helped the researchers to find new aspects from already existing mathematical theories. This kind of work is now going on very strongly.
### **5. Where is the hiding-place of fuzziness?**
For example, imagine a set of beautiful women. Let us denote this set by *A*. There are women who do not belong to *A* with the highest grade 1. So, such a woman *does not* have some features which would make her beautiful. But she may have some of those features anyway. An intuitive hint about a possible answer to the question: "Where is the hiding-place of fuzziness?" can be found just on the second line above: "... *does not* ..." It seems that a partial complementarity is somehow involved in this problem.
Let us compare Zadeh algebra with a general Boolean algebra with a supposition that the binary operations are associative because associativity holds in Zadeh algebra. The definition of this kind of Boolean algebra can be postulated as follows.
**Definition 5.1.** Let ∧ (*meet*) and ∨ (*join*) be binary operations, and � (*complement*) a unary operation on a set *B*(�= ∅), and let **0** and **1** be the elements of *B*, such that the following axioms hold:
$$\begin{aligned} \forall \mathbf{x}, \boldsymbol{y}, \boldsymbol{z} \in B, \quad & \mathbf{x} \wedge (\boldsymbol{y} \vee \boldsymbol{z}) = (\mathbf{x} \wedge \boldsymbol{y}) \vee (\mathbf{x} \wedge \boldsymbol{z}),\\ \mathbf{x} \vee (\boldsymbol{y} \wedge \boldsymbol{z}) = (\mathbf{x} \vee \boldsymbol{y}) \wedge (\mathbf{x} \vee \boldsymbol{z}).\end{aligned}$$
Hence, the set *B* together with these operations forms a *Boolean algebra* B = (*B*, ∨, ∧, � , **0**, **1**).
The only *structural difference* between these algebras is that between the axioms Z5 and BA 5. BA 5 is characteristic for complement operation, but Z5 does not satisfy the conditions of complement. So, fuzziness lies in the axiom Z5. The influence of this axiom is that also other
propositional wff's. Starting with propositional letters, we can combine them by connectives in the way shown by the production system (6.1). And finally, we can combine any formulas
Standard Fuzzy Sets and some Many-Valued Logics 85
We can refer to the propositional letters also by lower case letters in general, and to combined formulas by lower case Greek letters or by usual capital letters. These letters belong to the metalanguage we use when we discuss and describe object language. Here we use English
The definitions 6.1 and 6.2 above defines the language of propositional logic. Second, we consider some central semantical concepts being necessary for our consideration. This means that we will not present the whole machinery of formal semantics of standard many-valued
We have two important functions, *valuation* and *truth-function* we need in our consideration.
that associates truth values to propositional letters. **Prop** is a set of propositional variables.
where *n* is the number of propositional variables *p*, *q*, . . . in the formula defining a truth
In general, truth-functions are functions of several variables defined on the *n*-tuple of the
subindices *n* = 1, 2, 3, . . . are usually dropped. Actually, a propositional formula *ϕ* itself is a truth-function. Suppose that a formula *ϕ* consists of the propositional letters *p*, *q*, and *r*. Then we may write *ϕ* = *V*(*p*, *q*,*r*). The equality sign is used only between truth-functions and
In connected formulas, valuations of propositional variables give the values for the variables of the corresponding truth-function presented by the connected formula. Hence, a "valuation" of a connected formula is the value of the corresponding truth-function. Hence, to evaluate a truth value of the whole connected formula corresponding a given valuation for propositional variables, we calculate the value of the truth-function where the given valuation *v* first determines the values of the arguments of the truth-function. We may denote the truth value of a connected formula *ϕ* by *V*(*ϕ*), being like a valuation depending on a given valuation *v*
**Example 6.1.** Evaluate the truth value of a formula *p* ∧ (¬*q* ∨ *r*) with regard to a given valuation *v* for *p*, *q*, and *r*. Actually the formula is a truth-function *f*(*p*, *q*,*r*) = *p* ∧ (¬*q* ∨ *r*) where *p*, *q*, and *r* obtain their values from [0, 1]. These values are *v*(*p*), *v*(*q*), and *v*(*r*). Now,
*V*(*p* ∧ (¬*q* ∨ *r*)) = *v*(*p*) ∧ (¬*v*(*q*) ∨ *v*(*r*)
the truth value of the formula,given by this valuation *v*, is
*v* : **Prop** −→ **I** (6.2)
*Vn* : **<sup>I</sup>***<sup>n</sup>* −→ **<sup>I</sup>**, *<sup>n</sup>* <sup>=</sup> 1, 2, 3, . . . , (6.3)
*<sup>n</sup>* where the independent variables are proposition letters. The
equipped by some formal symbols (so-called meta-symbols) as a metalanguage.
according to the production system.
They are defined as follows.
*Truth-function* is a function
set of truth values, [0, 1]
for propositional letters.
**Definition 6.3.** *Valuation v* is a function
logics.
function.
truth values.
values can be considered as membership degrees than only 0 and 1. In Boolean algebras with the universe of discourse {0, 1} the postulate BA5 do not cause conflicts, like the intermediate values may do if these values are added to the universe.
Because complement operation satisfies the conditions of strong negation, a Boolean algebra B = ({0, 1}, ∨, ∧, � , **0**, **1**) is a special case of Z, i.e., the classical case is included in Z. Trivially, {0, 1} ⊂ [0, 1]. This means that crisp sets are special cases of fuzzy sets, as they should be also according to Zadeh's own theory. See also the proof of Theorem 2.1.
We may conclude that formally the core hiding-place of fuzziness is the statement Z5 in Proposition 4.1. In a concept being fuzzy there is always something that *does not* hold, i.e., some missing particle the concept does not have. Hence, the complementarity is somehow involved in a fuzzy concept.
### **6. Common features of many-valued logics based on Zadeh algebra**
We consider here some preliminary things being common for several many-valued logics. The main purpose is to find a connection between Zaheh algebra and the structures of some many-valued logics. We consider only some propositional logics, because the main concepts we consider here are basic to higher order many-valued logics, too. We restrict our considerations only to structural properties.
First, we need a formal language for our considerations. This language is that of propositional logic.
**Definition 6.1.** A propositional lanuage L consists of
These symbols are the *aplhabets* of the propositional language.
Usually only the so-called *primitive connectives* belong to the alphabet, but it is possible to choose some other connectives to the alphabet, too. Hence, we could drop either conjunction or disjunction from the alphabet if we like. Primitive connectives are connectives from which we can derive the other connectives.
In a standard nonclassical propositional language, the meanings of the connectives '∧' (*conjunction*), '∨' (*disjunction*), and '¬' (*negation*) can be given as follows: negation ¬ is a stong negation (i.e., it is a negation with involution property ¬¬*p* ≡ *p*), conjunction ∧ is *glb* (greatest lower bound), and disjunction ∨ is *lub* (least upper bound). The symbol '≡' and ⇐⇒ are used as a *meta-symbols* of equivalency, i.e., this symbol does not belong to the alphabet of the *object language* which is the language of a formal logic under consideration.
**Definition 6.2.** Well-formed formulas of L are given as follows:
$$\mathfrak{a} \implies \mathcal{p}\_{\mathbb{k}} \mid \neg \mathcal{q} \mid \mathfrak{q} \wedge \psi \mid \mathfrak{q} \vee \psi \,. \tag{6.1}$$
In this recursive production system of well formed formulas (wff's, for short) the symbol *pk* represents any propositional letter and lower case Greek letters are labels of any atomic or connected wff's. Hence, *α* is a label for any wff, and similarly *ϕ* and *ψ* represent any 10 Will-be-set-by-IN-TECH
values can be considered as membership degrees than only 0 and 1. In Boolean algebras with the universe of discourse {0, 1} the postulate BA5 do not cause conflicts, like the intermediate
Because complement operation satisfies the conditions of strong negation, a Boolean algebra
{0, 1} ⊂ [0, 1]. This means that crisp sets are special cases of fuzzy sets, as they should be also
We may conclude that formally the core hiding-place of fuzziness is the statement Z5 in Proposition 4.1. In a concept being fuzzy there is always something that *does not* hold, i.e., some missing particle the concept does not have. Hence, the complementarity is somehow
We consider here some preliminary things being common for several many-valued logics. The main purpose is to find a connection between Zaheh algebra and the structures of some many-valued logics. We consider only some propositional logics, because the main concepts we consider here are basic to higher order many-valued logics, too. We restrict our
First, we need a formal language for our considerations. This language is that of propositional
Usually only the so-called *primitive connectives* belong to the alphabet, but it is possible to choose some other connectives to the alphabet, too. Hence, we could drop either conjunction or disjunction from the alphabet if we like. Primitive connectives are connectives from which
In a standard nonclassical propositional language, the meanings of the connectives '∧' (*conjunction*), '∨' (*disjunction*), and '¬' (*negation*) can be given as follows: negation ¬ is a stong negation (i.e., it is a negation with involution property ¬¬*p* ≡ *p*), conjunction ∧ is *glb* (greatest lower bound), and disjunction ∨ is *lub* (least upper bound). The symbol '≡' and ⇐⇒ are used as a *meta-symbols* of equivalency, i.e., this symbol does not belong to the alphabet of the *object*
In this recursive production system of well formed formulas (wff's, for short) the symbol *pk* represents any propositional letter and lower case Greek letters are labels of any atomic or connected wff's. Hence, *α* is a label for any wff, and similarly *ϕ* and *ψ* represent any
*α* ::= *pk* | ¬*ϕ* | *ϕ* ∧ *ψ* | *ϕ* ∨ *ψ*. (6.1)
, **0**, **1**) is a special case of Z, i.e., the classical case is included in Z. Trivially,
values may do if these values are added to the universe.
according to Zadeh's own theory. See also the proof of Theorem 2.1.
**6. Common features of many-valued logics based on Zadeh algebra**
B = ({0, 1}, ∨, ∧, �
logic.
involved in a fuzzy concept.
considerations only to structural properties.
we can derive the other connectives.
**Definition 6.1.** A propositional lanuage L consists of 1. a set of propositional letters *p*0, *p*1,..., *pk*, . . . and 2. the truth-functional connectives '∧', '∨', and '¬'.
These symbols are the *aplhabets* of the propositional language.
*language* which is the language of a formal logic under consideration.
**Definition 6.2.** Well-formed formulas of L are given as follows:
propositional wff's. Starting with propositional letters, we can combine them by connectives in the way shown by the production system (6.1). And finally, we can combine any formulas according to the production system.
We can refer to the propositional letters also by lower case letters in general, and to combined formulas by lower case Greek letters or by usual capital letters. These letters belong to the metalanguage we use when we discuss and describe object language. Here we use English equipped by some formal symbols (so-called meta-symbols) as a metalanguage.
The definitions 6.1 and 6.2 above defines the language of propositional logic. Second, we consider some central semantical concepts being necessary for our consideration. This means that we will not present the whole machinery of formal semantics of standard many-valued logics.
We have two important functions, *valuation* and *truth-function* we need in our consideration. They are defined as follows.
**Definition 6.3.** *Valuation v* is a function
$$v: \mathbf{Prop} \longrightarrow \mathbb{I} \tag{6.2}$$
that associates truth values to propositional letters. **Prop** is a set of propositional variables.
*Truth-function* is a function
$$V\_{\mathfrak{n}} : \mathbb{I}^{\mathfrak{n}} \longrightarrow \mathbb{I}, \quad \mathfrak{n} = 1, 2, 3, \dots, \tag{6.3}$$
where *n* is the number of propositional variables *p*, *q*, . . . in the formula defining a truth function.
In general, truth-functions are functions of several variables defined on the *n*-tuple of the set of truth values, [0, 1] *<sup>n</sup>* where the independent variables are proposition letters. The subindices *n* = 1, 2, 3, . . . are usually dropped. Actually, a propositional formula *ϕ* itself is a truth-function. Suppose that a formula *ϕ* consists of the propositional letters *p*, *q*, and *r*. Then we may write *ϕ* = *V*(*p*, *q*,*r*). The equality sign is used only between truth-functions and truth values.
In connected formulas, valuations of propositional variables give the values for the variables of the corresponding truth-function presented by the connected formula. Hence, a "valuation" of a connected formula is the value of the corresponding truth-function. Hence, to evaluate a truth value of the whole connected formula corresponding a given valuation for propositional variables, we calculate the value of the truth-function where the given valuation *v* first determines the values of the arguments of the truth-function. We may denote the truth value of a connected formula *ϕ* by *V*(*ϕ*), being like a valuation depending on a given valuation *v* for propositional letters.
**Example 6.1.** Evaluate the truth value of a formula *p* ∧ (¬*q* ∨ *r*) with regard to a given valuation *v* for *p*, *q*, and *r*. Actually the formula is a truth-function *f*(*p*, *q*,*r*) = *p* ∧ (¬*q* ∨ *r*) where *p*, *q*, and *r* obtain their values from [0, 1]. These values are *v*(*p*), *v*(*q*), and *v*(*r*). Now, the truth value of the formula,given by this valuation *v*, is
$$V(p \land (\neg q \lor r)) = v(p) \land (\neg v(q) \lor v(r))$$
the evaluation rules (6.4), (6.5), and (6.6), the operations max, min, and ¬ exist at least in the logics having these evaluation rules. Let us compare the power set of fuzzy sets of the set *X*, i.e., the set **<sup>I</sup>***X*, to the set of all valuations *<sup>v</sup>* : **Prop** −→ **<sup>I</sup>**. Hence, the set of all valuations is **IProp**. Especially, **1** and **0** are constant valuations, such that **1** gives the truth value 1 to every propositional letters, and similarly, **0** gives the truth value 0. Hence, we also have the neutral elements corresponding those in Zadeh algebra. It seems that if we replace the set **I***<sup>X</sup>* by **IProp**
Standard Fuzzy Sets and some Many-Valued Logics 87
The values of valuations are truth values and those of membership functions are membership grades. Can these two interpretations for the elements [0, 1] be considered to be anyhow similar? According to formal consideration, we say *yes*. The both values are obtainable from the same set, namely from the unit interval [0, 1], and the construction of the both algebras are exactly the same. On the other hand, membership grades are in principle subjective opinions about the membership of an element in a given set. About truth values, a *degree of truth* of a given propositional letter in a given situation depends on the state of affairs associated to this situation. But there is a valuation for every state of affairs in any situation representing a suitable degree of truth expressed by a number obtained from [0, 1]. Hence, these degrees of truth correspond to suitable membership grades even so that for any valuation there exists a membership function that is identical with the valuation. Hence, these two apparently different interpretations can be considered to be the same. This means that we can interpret the values of the functions of the algebra (6.7) as truth values, or more accurately, degrees of
For historical reasons, we consider a piece of Kleene's 3-valued logic. S. C. Kleene was Zadeh's logic teacher, and it is natural that Zadeh compared his concept of fuzzy set with Kleene's
"Note that the notion of belonging", which plays a fundamental role in the case of ordinary sets, does not have the same role in the case of fuzzy sets. Thus, it is not meaningful to speak of a point *x* "belonging" to a fuzzy set *A* except in the trivial sense of *fA*(*x*) being positive. Less trivially, one can introduce two levels *α* and *β* (0 < *α* < 1, 0 < *β* < 1, *α* > *β*) and agree to say that (1) "*x* belongs to *A*" if *fA*(*x*) ≥ *α*; (2) "*x* does not belong to *A*" if *fA*(*x*) ≤ *β*; and (3) "*x* has an intermediate status relative to *A*" if *β* < *fA*(*x*) < *α*. This leads to a three-valued logic (Kleene, 1952) with three truth
The symbols of the truth values of Kleene's 3-valued logic are *T* (*true*), *U* (*unknown*), and *F* (*false*). In the literature, there are also some alternative symbols for the intermediate truth
Kleene introduced his 3-vaued logic in 1938. We denote it by **K**3. In order to describe Kleene's
"In Kleene's system, a proposition is to bear the third truth-value *I* not for fact-related, ontological reasons but for knowledge-related, epistemological ones: it is not to be excluded that the proposition may *in fact* be true or false, but it is merely *unknown*
3-valued logic. Zadeh ((Zadeh, 1965) p. 341-342) gives the following comment:
values *T* (*fA*(*x*) ≥ *α*), *F* (*fA*(*x*) ≤ *β*), and *U* (*β* < *fA*(*x*) < *α*).
value. For example, Rescher (Rescher, 1969) uses the symbol *I*.
logic, we refer to Rescher (Rescher, 1969), p. 34 - 36. He writes:
or undeterminable what its specific truth status may be.
Lℵ<sup>1</sup> <sup>=</sup> �**IProp**, max, min, <sup>¬</sup>, **<sup>0</sup>**, **<sup>1</sup>**�. (6.7)
then we have a special Zadeh algebra, namely, say, *propositional algebra*
truth.
**7. Description of Kleene's logic**
Suppose that *v* is a valuation where *v*(*p*) = 0.5, *v*(*q*) = 0.3, and *v*(*r*) = 1. Hence,
$$V(p \land (\neg q \lor r)) = 0.5 \land ((1 - 0.3) \lor 1) = 0.5 \land (0.7 \lor 1) = 0.5 \land 1 = 0.5$$
If we have two wff's representing the same state of affairs we use meta-equivalence sign '≡' or '⇐⇒' between them because the formulas are equivalent to each other, not identical. For example, the wff *ϕ* is the formula *p* ∧ (¬*q* ∨ *r*). So, we can write *ϕ* ≡ *p* ∧ (¬*q* ∨ *r*), or *ϕ* ⇐⇒ *p* ∧ (¬*q* ∨ *r*) to denote that we use an abbreviation *ϕ* for the formula *p* ∧ (¬*q* ∨ *r*). Another case is that we have two formulas being equivalent to each other, for example, ¬*p* ∨ ¬*q* ≡ ¬(*p* ∧ *q*). This equivalency describes one of De Morgan's laws. However, the expression
$$\forall \mathbf{x}, y \in [0, 1], \quad \neg \mathbf{x} \lor \neg y = \neg (\mathbf{x} \land y)$$
emphasizes that two truth-functions ¬*x* ∨ ¬*y* and ¬(*x* ∧ *y*) are identical.
Instead of propositional letters, we prefer to use "usual" variable symbols as variables of a truth-function, because of possible confusions.
We are interested in the logics where the *evaluation rules* for these connectives are
$$V(p \lor q) = \max\{v(p), v(q)\}\tag{6.4}$$
$$V(p \wedge q) = \min\{v(p), v(q)\}\tag{6.5}$$
$$V(\neg p) = 1 - v(p) \tag{6.6}$$
where *v*(*p*), *v*(*q*) ∈ [0, 1], or *v*(*p*), *v*(*q*) ∈ *A* where *A* is a suitable subset of [0, 1].
Evaluation rules are rules for evaluating truth values to connected logical formulas in a given logic.
We must remember that all the logics are not truth-functional. For example, modal logics are non-truth-functional.
In practice, we need some other connectives, too. Two of them are the connectives *implication* and *equivalency*.
The way to choose implication separates the logics based on Zadeh algebra or any De Morgan algebra. Hence, implication must be presented by means of disjunction and negation, or by means of conjunction and negation. There are several ways to define different implications from other connectives, depending on the logic in question. We consider these things in the case of each logic to be considered below.
The formulas (6.4), (6.5), and (6.6) somehow emphasize the relationship to algebraic construction. We have two binary operations and one unary operation defined on a nonempty set just as in usual algebraic system. Additionally, the binary operations are combined together, for example, being distributive. The formulas (6.4), (6.5), and (6.6) are the same as Zadeh's operations defined on the set of fuzzy sets. The bridge between standard fuzzy sets and some many-valued logics seems to be obvious. Having got this kind of motivation, we continue our construction of the bridge between standard fuzzy sets and some many-valued logics.
Consider Zadeh algebra Zℵ<sup>1</sup> <sup>=</sup> �**I***X*, max, min, <sup>¬</sup>, **<sup>0</sup>**, **<sup>1</sup>**� (*cf.* formula (4.1)). Now the question is wether there is a counterpart to this algebra in the scope of many-valued logic. According to 12 Will-be-set-by-IN-TECH
*V*(*p* ∧ (¬*q* ∨ *r*)) = 0.5 ∧ ((1 − 0.3) ∨ 1) = 0.5 ∧ (0.7 ∨ 1) = 0.5 ∧ 1 = 0.5
If we have two wff's representing the same state of affairs we use meta-equivalence sign '≡' or '⇐⇒' between them because the formulas are equivalent to each other, not identical. For example, the wff *ϕ* is the formula *p* ∧ (¬*q* ∨ *r*). So, we can write *ϕ* ≡ *p* ∧ (¬*q* ∨ *r*), or *ϕ* ⇐⇒ *p* ∧ (¬*q* ∨ *r*) to denote that we use an abbreviation *ϕ* for the formula *p* ∧ (¬*q* ∨ *r*). Another case is that we have two formulas being equivalent to each other, for example, ¬*p* ∨ ¬*q* ≡ ¬(*p* ∧ *q*).
∀*x*, *y* ∈ [0, 1], ¬*x* ∨ ¬*y* = ¬(*x* ∧ *y*)
Instead of propositional letters, we prefer to use "usual" variable symbols as variables of a
Evaluation rules are rules for evaluating truth values to connected logical formulas in a given
We must remember that all the logics are not truth-functional. For example, modal logics are
In practice, we need some other connectives, too. Two of them are the connectives *implication*
The way to choose implication separates the logics based on Zadeh algebra or any De Morgan algebra. Hence, implication must be presented by means of disjunction and negation, or by means of conjunction and negation. There are several ways to define different implications from other connectives, depending on the logic in question. We consider these things in the
The formulas (6.4), (6.5), and (6.6) somehow emphasize the relationship to algebraic construction. We have two binary operations and one unary operation defined on a nonempty set just as in usual algebraic system. Additionally, the binary operations are combined together, for example, being distributive. The formulas (6.4), (6.5), and (6.6) are the same as Zadeh's operations defined on the set of fuzzy sets. The bridge between standard fuzzy sets and some many-valued logics seems to be obvious. Having got this kind of motivation, we continue our construction of the bridge between standard fuzzy sets and some many-valued
Consider Zadeh algebra Zℵ<sup>1</sup> <sup>=</sup> �**I***X*, max, min, <sup>¬</sup>, **<sup>0</sup>**, **<sup>1</sup>**� (*cf.* formula (4.1)). Now the question is wether there is a counterpart to this algebra in the scope of many-valued logic. According to
*V*(*p* ∨ *q*) = max{*v*(*p*), *v*(*q*)} (6.4) *V*(*p* ∧ *q*) = min{*v*(*p*), *v*(*q*)} (6.5) *V*(¬*p*) = 1 − *v*(*p*) (6.6)
Suppose that *v* is a valuation where *v*(*p*) = 0.5, *v*(*q*) = 0.3, and *v*(*r*) = 1. Hence,
This equivalency describes one of De Morgan's laws. However, the expression
We are interested in the logics where the *evaluation rules* for these connectives are
where *v*(*p*), *v*(*q*) ∈ [0, 1], or *v*(*p*), *v*(*q*) ∈ *A* where *A* is a suitable subset of [0, 1].
emphasizes that two truth-functions ¬*x* ∨ ¬*y* and ¬(*x* ∧ *y*) are identical.
truth-function, because of possible confusions.
case of each logic to be considered below.
logic.
logics.
non-truth-functional.
and *equivalency*.
the evaluation rules (6.4), (6.5), and (6.6), the operations max, min, and ¬ exist at least in the logics having these evaluation rules. Let us compare the power set of fuzzy sets of the set *X*, i.e., the set **<sup>I</sup>***X*, to the set of all valuations *<sup>v</sup>* : **Prop** −→ **<sup>I</sup>**. Hence, the set of all valuations is **IProp**. Especially, **1** and **0** are constant valuations, such that **1** gives the truth value 1 to every propositional letters, and similarly, **0** gives the truth value 0. Hence, we also have the neutral elements corresponding those in Zadeh algebra. It seems that if we replace the set **I***<sup>X</sup>* by **IProp** then we have a special Zadeh algebra, namely, say, *propositional algebra*
$$\mathcal{L}\_{\aleph\_1} = \langle \mathbf{I}^{\mathbf{Prop}}, \mathbf{max}, \min, \neg, \mathbf{0}, \mathbf{1} \rangle. \tag{6.7}$$
The values of valuations are truth values and those of membership functions are membership grades. Can these two interpretations for the elements [0, 1] be considered to be anyhow similar? According to formal consideration, we say *yes*. The both values are obtainable from the same set, namely from the unit interval [0, 1], and the construction of the both algebras are exactly the same. On the other hand, membership grades are in principle subjective opinions about the membership of an element in a given set. About truth values, a *degree of truth* of a given propositional letter in a given situation depends on the state of affairs associated to this situation. But there is a valuation for every state of affairs in any situation representing a suitable degree of truth expressed by a number obtained from [0, 1]. Hence, these degrees of truth correspond to suitable membership grades even so that for any valuation there exists a membership function that is identical with the valuation. Hence, these two apparently different interpretations can be considered to be the same. This means that we can interpret the values of the functions of the algebra (6.7) as truth values, or more accurately, degrees of truth.
### **7. Description of Kleene's logic**
For historical reasons, we consider a piece of Kleene's 3-valued logic. S. C. Kleene was Zadeh's logic teacher, and it is natural that Zadeh compared his concept of fuzzy set with Kleene's 3-valued logic. Zadeh ((Zadeh, 1965) p. 341-342) gives the following comment:
"Note that the notion of belonging", which plays a fundamental role in the case of ordinary sets, does not have the same role in the case of fuzzy sets. Thus, it is not meaningful to speak of a point *x* "belonging" to a fuzzy set *A* except in the trivial sense of *fA*(*x*) being positive. Less trivially, one can introduce two levels *α* and *β* (0 < *α* < 1, 0 < *β* < 1, *α* > *β*) and agree to say that (1) "*x* belongs to *A*" if *fA*(*x*) ≥ *α*; (2) "*x* does not belong to *A*" if *fA*(*x*) ≤ *β*; and (3) "*x* has an intermediate status relative to *A*" if *β* < *fA*(*x*) < *α*. This leads to a three-valued logic (Kleene, 1952) with three truth values *T* (*fA*(*x*) ≥ *α*), *F* (*fA*(*x*) ≤ *β*), and *U* (*β* < *fA*(*x*) < *α*).
The symbols of the truth values of Kleene's 3-valued logic are *T* (*true*), *U* (*unknown*), and *F* (*false*). In the literature, there are also some alternative symbols for the intermediate truth value. For example, Rescher (Rescher, 1969) uses the symbol *I*.
Kleene introduced his 3-vaued logic in 1938. We denote it by **K**3. In order to describe Kleene's logic, we refer to Rescher (Rescher, 1969), p. 34 - 36. He writes:
"In Kleene's system, a proposition is to bear the third truth-value *I* not for fact-related, ontological reasons but for knowledge-related, epistemological ones: it is not to be excluded that the proposition may *in fact* be true or false, but it is merely *unknown* or undeterminable what its specific truth status may be.
Clearly, the algebraic approach to **K**<sup>3</sup> is Kleene algebra, i.e., a 3-valued Zadeh algebra with the
Standard Fuzzy Sets and some Many-Valued Logics 89
*Kleene-Dienes many-valued logic* is an extension of **<sup>K</sup>**<sup>3</sup> into **<sup>K</sup>**ℵ<sup>1</sup> having the set of truth values [0, 1]. The evaluation rules for conjunction, disjunction, and negation are the same as in the
def
being in accordance with the implication of **K**3. This means that the evaluation rule for
The equations (6.4), (6.5), (6.6), (7.7), and (7.9) are the truth value evaluation rules for disjunction, conjunction, negation, implication, and equivalence, respectively, of
The implication operation of Kleene-Dienes many-valued logic is a typical example about a so-called *S-implication*. Another example is the implication operation of classical logic. The
We begin with Łukasiewicz' many-valued logic Łℵ<sup>1</sup> having the closed unit interval [0, 1] as the
As we know, Łukasiewicz chose the connectives of *negation* and *implication* as primitives. This is a remarkable difference, for example, between Kleene's logic and Łukasiewicz logic. Hence, the connection between standard fuzzy set theory and Łℵ<sup>1</sup> cannot be seen immediately.
Let *<sup>v</sup>* be any valuation of Łℵ, then the truth value evaluation rules for negation and implication
⇐⇒ ¬*p* ∨ *q* (7.6)
*x* → *y* = max {1 − *x*, *y*} (7.7)
⇐⇒ (*p* → *q*) ∧ (*q* → *p*) (7.8)
∀*x*, *y* ∈ [0, 1], *x* ↔ *y* = min{*x* → *y*, *y* → *x*} (7.9)
*v*(¬*p*) = 1 − *v*(*p*) (Neg.) *v*(*p* → *q*) = min{1, 1 − *v*(*p*) + *v*(*q*)} (Impl.)
*p* → *q*
def
property (K).
*Kleene-Dienes many-valued logic*:
formulas (6.4), (6.5), and (6.6) above.
implication is as follows. For any *x*, *y* ∈ [0, 1],
Hence, the evaluation rule for equivalence is
**8. On Łukasiewicz' many-valued logic**
<sup>2</sup> *Cf.* Rescher (Rescher, 1969), p.36, and 337.
set of truth values.2
are
Implication of Kleene-Dienes many-valued logic is defined by
Now, the connective *equivalence* is defined in the usual way:
*p* ↔ *q*
Kleene-Dienes many-valued logic with the set of truth values [0, 1].
general principle for S-implication is just the formula (7.6).
In **K**3, we have the following truth value evaluation rules for the connectives negation ¬, conjunction ∧, and disjunction ∨:
$$V(\neg p) = T - v(p)\_{\prime} \tag{7.1}$$
$$V(p \wedge q) = \min\{v(p), v(q)\},\tag{7.2}$$
$$V(p \lor q) = \max\{v(p), v(q)\}. \tag{7.3}$$
Two of these connectives can form a set of *primitive connectives*. One of the primitives must be negation. Hence, for example, the connectives negation and disjunction can be chosen as primitives, and all the other connectives can be defined by means of these primitive connectives. The alternative case for primitives are negation and conjunction. Hence, we can define the nonprimitive one by negation and the fact that disjunction and conjunction are dual (i.e., by using a suitable De Morgan's law). The implication defined by means of negation and disjunction is given by the formula (7.4).
We see immediately that there is a strong analogy between the basic operations of fuzzy sets (*cf.* Def. 2.2) and these three connectives of Kleene's 3-valued logic. In addition to this, an analogy can be found between Kleene's valuations and Zadeh's membership functions, too, although they have different ranges. Zadeh's comment above connects these two concepts "membership" (*μ* : *X* −→ **I**) and "valuation" (*v* : **Prop** −→ {*F*, *U*, *T*}) together. We can compare the set {*F*, *<sup>U</sup>*, *<sup>T</sup>*} with the set {0, <sup>1</sup> <sup>2</sup> , 1} where *<sup>F</sup>* <sup>=</sup> 0, *<sup>U</sup>* <sup>=</sup> <sup>1</sup> <sup>2</sup> , and *T* = 1. Hence, by analogy of the sets, we can understand some arithmetic operations in some evaluation rules.
Kleene defined the implication of his 3-valued logic, denoted by , analogously to material implication:
$$p \ni q \stackrel{\text{def}}{\iff} \neg p \lor q \tag{7.4}$$
hence, the evaluation rule of *p q* is
$$p \ni q = \max\{T - v(p), v(q)\}. \tag{7.5}$$
We may construct the truth tables according to the evaluation rules. Rescher tells that Kleene motivated the construction of his truth tables in terms of a mathematical application. He has in mind the case of a mathematical predicate *P* (i.e., a propositional function) of a variable *x* ranging over a domain *D* where "*P*(*x*)" is defined for only a part of this domain. For example, we might have the condition
$$P(\mathfrak{x}) \quad \text{iff} \quad 1 \le \frac{1}{\mathfrak{x}} \le 2.$$
Here *P*(*x*) will be:
Kleene presented his truth tables to formulate the rules of combination by logical connectives for such propositional functions. He writes:
"From this standpoint, the meaning of *Q* ∨ *R* is brought out clearly by the statement in words: *Q* ∨ *R* is true, if *Q* is true (here nothing is said about *R*) or if *R* is true (similarly); false, if *Q* and *R* are both false; defined only in these cases (and hence undefined, otherwise)."<sup>1</sup>
<sup>1</sup> Kleene, *Introduction to Metamathematics* (1952).
Clearly, the algebraic approach to **K**<sup>3</sup> is Kleene algebra, i.e., a 3-valued Zadeh algebra with the property (K).
*Kleene-Dienes many-valued logic*:
14 Will-be-set-by-IN-TECH
In **K**3, we have the following truth value evaluation rules for the connectives negation ¬,
Two of these connectives can form a set of *primitive connectives*. One of the primitives must be negation. Hence, for example, the connectives negation and disjunction can be chosen as primitives, and all the other connectives can be defined by means of these primitive connectives. The alternative case for primitives are negation and conjunction. Hence, we can define the nonprimitive one by negation and the fact that disjunction and conjunction are dual (i.e., by using a suitable De Morgan's law). The implication defined by means of negation
We see immediately that there is a strong analogy between the basic operations of fuzzy sets (*cf.* Def. 2.2) and these three connectives of Kleene's 3-valued logic. In addition to this, an analogy can be found between Kleene's valuations and Zadeh's membership functions, too, although they have different ranges. Zadeh's comment above connects these two concepts "membership" (*μ* : *X* −→ **I**) and "valuation" (*v* : **Prop** −→ {*F*, *U*, *T*}) together. We can
analogy of the sets, we can understand some arithmetic operations in some evaluation rules. Kleene defined the implication of his 3-valued logic, denoted by , analogously to material
def
We may construct the truth tables according to the evaluation rules. Rescher tells that Kleene motivated the construction of his truth tables in terms of a mathematical application. He has in mind the case of a mathematical predicate *P* (i.e., a propositional function) of a variable *x* ranging over a domain *D* where "*P*(*x*)" is defined for only a part of this domain. For example,
> 1 *x* ≤ 2.
*P*(*x*) iff 1 ≤
<sup>2</sup> to 1,
Kleene presented his truth tables to formulate the rules of combination by logical connectives
"From this standpoint, the meaning of *Q* ∨ *R* is brought out clearly by the statement in words: *Q* ∨ *R* is true, if *Q* is true (here nothing is said about *R*) or if *R* is true (similarly); false, if *Q* and *R* are both false; defined only in these cases (and hence undefined,
*p q*
<sup>2</sup> , 1} where *<sup>F</sup>* <sup>=</sup> 0, *<sup>U</sup>* <sup>=</sup> <sup>1</sup>
<sup>2</sup> , and *T* = 1. Hence, by
⇐⇒ ¬*p* ∨ *q*, (7.4)
*p q* = max{*T* − *v*(*p*), *v*(*q*)}. (7.5)
*V*(¬*p*) = *T* − *v*(*p*), (7.1) *V*(*p* ∧ *q*) = min{*v*(*p*), *v*(*q*)}, (7.2) *V*(*p* ∨ *q*) = max{*v*(*p*), *v*(*q*)}. (7.3)
conjunction ∧, and disjunction ∨:
and disjunction is given by the formula (7.4).
compare the set {*F*, *<sup>U</sup>*, *<sup>T</sup>*} with the set {0, <sup>1</sup>
hence, the evaluation rule of *p q* is
(1) *true* if *x* lies within the range from <sup>1</sup>
(2) *undefined* (or undetermined) if *x* = 0,
for such propositional functions. He writes:
<sup>1</sup> Kleene, *Introduction to Metamathematics* (1952).
we might have the condition
(3) *false* in all other cases.
Here *P*(*x*) will be:
otherwise)."<sup>1</sup>
implication:
*Kleene-Dienes many-valued logic* is an extension of **<sup>K</sup>**<sup>3</sup> into **<sup>K</sup>**ℵ<sup>1</sup> having the set of truth values [0, 1]. The evaluation rules for conjunction, disjunction, and negation are the same as in the formulas (6.4), (6.5), and (6.6) above.
Implication of Kleene-Dienes many-valued logic is defined by
$$p \to q \stackrel{\text{def}}{\iff} \neg p \lor q \tag{7.6}$$
being in accordance with the implication of **K**3. This means that the evaluation rule for implication is as follows. For any *x*, *y* ∈ [0, 1],
$$\mathbf{x} \to \mathbf{y} = \max\left\{1 - \mathbf{x}\_{\prime}\mathbf{y}\right\} \tag{7.7}$$
Now, the connective *equivalence* is defined in the usual way:
$$p \leftrightarrow q \stackrel{\text{def}}{\Longleftrightarrow} (p \to q) \land (q \to p) \tag{7.8}$$
Hence, the evaluation rule for equivalence is
$$\forall \mathbf{x}, y \in [0, 1], \quad \mathbf{x} \leftrightarrow y = \min \{ \mathbf{x} \to y, y \to \mathbf{x} \} \tag{7.9}$$
The equations (6.4), (6.5), (6.6), (7.7), and (7.9) are the truth value evaluation rules for disjunction, conjunction, negation, implication, and equivalence, respectively, of Kleene-Dienes many-valued logic with the set of truth values [0, 1].
The implication operation of Kleene-Dienes many-valued logic is a typical example about a so-called *S-implication*. Another example is the implication operation of classical logic. The general principle for S-implication is just the formula (7.6).
### **8. On Łukasiewicz' many-valued logic**
We begin with Łukasiewicz' many-valued logic Łℵ<sup>1</sup> having the closed unit interval [0, 1] as the set of truth values.2
As we know, Łukasiewicz chose the connectives of *negation* and *implication* as primitives. This is a remarkable difference, for example, between Kleene's logic and Łukasiewicz logic. Hence, the connection between standard fuzzy set theory and Łℵ<sup>1</sup> cannot be seen immediately.
Let *<sup>v</sup>* be any valuation of Łℵ, then the truth value evaluation rules for negation and implication are
$$v(\neg p) = 1 - v(p) \tag{\text{Neg.}}$$
$$v(p \to q) = \min\{1, 1 - v(p) + v(q)\}\tag{\text{Impl.}}$$
<sup>2</sup> *Cf.* Rescher (Rescher, 1969), p.36, and 337.
**Proposition 8.1.** *Suppose x*, *y* ∈ [0, 1]*. Then De Morgan's laws hold for* min{*x*, *y*} *and* max{*x*, *y*}*.*
Standard Fuzzy Sets and some Many-Valued Logics 91
*Proof.* Consider the operation min(*x*, *y*) and max(*x*, *y*), where *x* and *y* are variables taking their values from the interval [0, 1]. Using the arithmetical formula for min operation (i.e., the
> <sup>=</sup> <sup>2</sup> <sup>−</sup> <sup>1</sup> <sup>−</sup> <sup>1</sup> <sup>+</sup> *<sup>x</sup>* <sup>+</sup> *<sup>y</sup>*− | <sup>1</sup> <sup>−</sup> <sup>1</sup> <sup>+</sup> *<sup>x</sup>* <sup>−</sup> *<sup>y</sup>* <sup>|</sup> 2
> <sup>=</sup> <sup>1</sup> <sup>−</sup> <sup>1</sup> <sup>+</sup> <sup>1</sup> <sup>−</sup> *<sup>x</sup>* <sup>−</sup> *<sup>y</sup>*<sup>+</sup> <sup>|</sup> <sup>1</sup> <sup>−</sup> <sup>1</sup> <sup>+</sup> *<sup>x</sup>* <sup>−</sup> *<sup>y</sup>* <sup>|</sup> 2
From the formula (8.5), by replacing *x* by 1 − *x* and *y* by 1 − *y*, and then solving max(*x*, *y*), the
The formulas (8.5) and (8.6) show that DeMorgan laws hold for max and min, and they are
Łukasiewicz knew that the operations max and min are dual of each other. Actually, this property is easily found in the classical special case, i.e. using characteristic functions in presenting crisp sets. But the general proof for this is easily done by using the expressions (8.4) for max and min in such cases where a distance metric is defined in the universe of
Second, consider the connection between max and Łukasiewicz implication using Zadeh algebra (similar considerations are done in Mattila (Mattila, 2005), but the following
*Proof.* Consider disjunction operation *x* ∨ *y* = max(*x*, *y*). Because 0 ≤ *x*, *y* ≤ 1, using the
<sup>Ł</sup> *<sup>y</sup>*) <sup>→</sup>
<sup>Ł</sup> *<sup>y</sup>*, (8.7)
max(*x*, *y*)=(*x* →
<sup>=</sup> <sup>1</sup> <sup>−</sup> (<sup>1</sup> <sup>−</sup> *<sup>x</sup>*)+(<sup>1</sup> <sup>−</sup> *<sup>y</sup>*)+ <sup>|</sup> (<sup>1</sup> <sup>−</sup> *<sup>y</sup>*) <sup>−</sup> (<sup>1</sup> <sup>−</sup> *<sup>x</sup>*) <sup>|</sup> 2
= 1 − max{ (1 − *x*),(1 − *y*) }. (8.5)
max{*x*, *y*} = 1 − min{ 1 − *x*, 1 − *y*}. (8.6)
expression for min in (8.4)), we have
following formula follows:
dual of each other.
*where x* →
This completes the proof.
discourse. This always holds at least for real numbers.
proposition 8.2 is not completely proved).
<sup>Ł</sup> *y is Łukasiewicz implication.*
arithmetical formula (8.4) for max, we have
**Proposition 8.2.** *For all x*, *y* ∈ [0, 1]*,*
min{*x*, *<sup>y</sup>*} <sup>=</sup> *<sup>x</sup>* <sup>+</sup> *<sup>y</sup>*− | *<sup>x</sup>* <sup>−</sup> *<sup>y</sup>* <sup>|</sup>
2
By means of these connectives, Łukasiewicz defined the other connectives by the rules
$$p \lor q \stackrel{\text{def}}{\iff} (p \to q) \to q \tag{\text{Dist.}}$$
$$(p \land q \stackrel{\text{def}}{\iff} \neg(\neg p \lor \neg q) \tag{\text{Conj.}}$$
$$p \leftrightarrow q \stackrel{\text{def}}{\Longleftrightarrow} (p \to q) \land (q \to p) \tag{Eq.}$$
The truth value evaluation rules for these derived connectives are
$$\max\{v(p), v(q)\}\tag{8.1}$$
$$\min\{v(p), v(q)\} \tag{8.2}$$
1 − |*v*(*p*) − *v*(*q*)| for *p* ↔ *q* (8.3)
for any valuation *<sup>v</sup>* of Łℵ<sup>1</sup> .
In Zadeh algebra we have the operations representing disjunction, conjunction, and negation as given. Negation in Zadeh algebra has the same construction as that in Łukasiewicz' logic <sup>Ł</sup>ℵ<sup>1</sup> , so, we need not to do anything with it. Now our task is to derive algebraically the *implication* of Łℵ<sup>1</sup> by means of these three other connectives. For this we use the operations of Zadeh algebra. Actually, we need only complementarity and max operations in our solution. After succeeding to solve this problem we know that standard fuzzy sets and Łℵ<sup>1</sup> fits together completely, i.e., we can derive all the connectives of Łℵ<sup>1</sup> in terms of Zadeh algebra. The final result is given in Proposition 8.3. This is the main task in this section.
Consider again the special case of Zadeh algebra (6.7)
$$\mathcal{L}\_{\aleph\_1} = \langle \mathbf{I}^{\mathbf{Prop}}, \mathbf{max}, \min, \neg, \mathbf{0}, \mathbf{1} \rangle.$$
From the considertations above, we know that
We observed in Section 3 that [0, 1] is a metric space with the natural metric distance (3.2)
$$d(\mathfrak{x}, y) = |\mathfrak{x} - y|\, \, \, \, \mathfrak{x} \, y \in [0, 1].$$
This formula satisfies the general definition of the concept *metric*. We need it in the following consideration where we manipulate expressions involving maxima and minima.
In manipulating maxima and minima, the consideration can sometimes be done easier by using the following expressions for max and min operations:
$$\max\{\mathbf{x}, y\} = \frac{\mathbf{x} + \mathbf{y} + |\mathbf{x} - \mathbf{y}|}{2}, \qquad \min\{\mathbf{x}, y\} = \frac{\mathbf{x} + \mathbf{y} - |\mathbf{x} - \mathbf{y}|}{2} \tag{8.4}$$
These formulas hold on the set of real numbers **R**, and especially on the unit interval [0, 1].
First, consider the case where the operations min and max are used in the form of the formulas (8.4), and ¬ is defined in the usual way: ¬*x* = 1 − *x*.
16 Will-be-set-by-IN-TECH
In Zadeh algebra we have the operations representing disjunction, conjunction, and negation as given. Negation in Zadeh algebra has the same construction as that in Łukasiewicz' logic <sup>Ł</sup>ℵ<sup>1</sup> , so, we need not to do anything with it. Now our task is to derive algebraically the *implication* of Łℵ<sup>1</sup> by means of these three other connectives. For this we use the operations of Zadeh algebra. Actually, we need only complementarity and max operations in our solution. After succeeding to solve this problem we know that standard fuzzy sets and Łℵ<sup>1</sup> fits together completely, i.e., we can derive all the connectives of Łℵ<sup>1</sup> in terms of Zadeh algebra. The final
Lℵ<sup>1</sup> <sup>=</sup> �**IProp**, max, min, <sup>¬</sup>, **<sup>0</sup>**, **<sup>1</sup>**�.
• The unary operation ¬ is a complementarity operation with the property of involution. We observed in Section 3 that [0, 1] is a metric space with the natural metric distance (3.2)
*d*(*x*, *y*) = |*x* − *y*| , *x*, *y* ∈ [0, 1]. This formula satisfies the general definition of the concept *metric*. We need it in the following
In manipulating maxima and minima, the consideration can sometimes be done easier by
These formulas hold on the set of real numbers **R**, and especially on the unit interval [0, 1]. First, consider the case where the operations min and max are used in the form of the formulas
<sup>2</sup> , min{*x*, *<sup>y</sup>*} <sup>=</sup> *<sup>x</sup>* <sup>+</sup> *<sup>y</sup>* <sup>−</sup> <sup>|</sup>*<sup>x</sup>* <sup>−</sup> *<sup>y</sup>*<sup>|</sup>
<sup>2</sup> (8.4)
consideration where we manipulate expressions involving maxima and minima.
⇐⇒ (*p* → *q*) → *q* (Disj.)
⇐⇒ ¬(¬*p* ∨ ¬*q*) (Conj.)
⇐⇒ (*p* → *q*) ∧ (*q* → *p*) (Eq.)
max{*v*(*p*), *v*(*q*)} for *p* ∨ *q*, (8.1) min{*v*(*p*), *v*(*q*)} for *p* ∧ *q*, (8.2) 1 − |*v*(*p*) − *v*(*q*)| for *p* ↔ *q* (8.3)
By means of these connectives, Łukasiewicz defined the other connectives by the rules
def
def
def
*p* ∨ *q*
*p* ∧ *q*
*p* ↔ *q*
result is given in Proposition 8.3. This is the main task in this section.
• Lℵ<sup>1</sup> is a special Zadeh algebra, namely propositional algebra.
• The operations max and min are distributive to each others.
using the following expressions for max and min operations:
max{*x*, *<sup>y</sup>*} <sup>=</sup> *<sup>x</sup>* <sup>+</sup> *<sup>y</sup>* <sup>+</sup> <sup>|</sup>*<sup>x</sup>* <sup>−</sup> *<sup>y</sup>*<sup>|</sup>
(8.4), and ¬ is defined in the usual way: ¬*x* = 1 − *x*.
• The binary operations max and min are commutative and associative.
Consider again the special case of Zadeh algebra (6.7)
From the considertations above, we know that
The truth value evaluation rules for these derived connectives are
for any valuation *<sup>v</sup>* of Łℵ<sup>1</sup> .
**Proposition 8.1.** *Suppose x*, *y* ∈ [0, 1]*. Then De Morgan's laws hold for* min{*x*, *y*} *and* max{*x*, *y*}*.*
*Proof.* Consider the operation min(*x*, *y*) and max(*x*, *y*), where *x* and *y* are variables taking their values from the interval [0, 1]. Using the arithmetical formula for min operation (i.e., the expression for min in (8.4)), we have
$$\min\{x, y\} = \frac{x + y - \lfloor x - y \rfloor}{2}$$
$$= \frac{2 - 1 - 1 + x + y - \lfloor 1 - 1 + x - y \rfloor}{2}$$
$$= 1 - \frac{1 + 1 - x - y + \lfloor 1 - 1 + x - y \rfloor}{2}$$
$$= 1 - \frac{(1 - x) + (1 - y) + \lfloor (1 - y) - (1 - x) \rfloor}{2}$$
$$= 1 - \max\{ (1 - x), (1 - y) \}. \tag{8.5}$$
From the formula (8.5), by replacing *x* by 1 − *x* and *y* by 1 − *y*, and then solving max(*x*, *y*), the following formula follows:
$$\max\{\mathbf{x}, y\} = 1 - \min\{1 - \mathbf{x}, 1 - y\}. \tag{8.6}$$
The formulas (8.5) and (8.6) show that DeMorgan laws hold for max and min, and they are dual of each other.
This completes the proof.
Łukasiewicz knew that the operations max and min are dual of each other. Actually, this property is easily found in the classical special case, i.e. using characteristic functions in presenting crisp sets. But the general proof for this is easily done by using the expressions (8.4) for max and min in such cases where a distance metric is defined in the universe of discourse. This always holds at least for real numbers.
Second, consider the connection between max and Łukasiewicz implication using Zadeh algebra (similar considerations are done in Mattila (Mattila, 2005), but the following proposition 8.2 is not completely proved).
**Proposition 8.2.** *For all x*, *y* ∈ [0, 1]*,*
$$\max(\mathbf{x}, y) = (\mathbf{x} \underset{\mathbf{L}}{\rightarrow} y) \underset{\mathbf{L}}{\rightarrow} y,\tag{8.7}$$
*where x* → <sup>Ł</sup> *y is Łukasiewicz implication.*
*Proof.* Consider disjunction operation *x* ∨ *y* = max(*x*, *y*). Because 0 ≤ *x*, *y* ≤ 1, using the arithmetical formula (8.4) for max, we have
The case 4 follows from the case 2 by Prop. 8.2 as follows. When we consider the equation (8.10) in the proof of Prop. 8.2, we find two min-stuctures corresponding to the evaluation rule of implication, so that one of them is an inside part of the whole formula. If we denote the inner min-structure min(1, 1 − *x* + *y*) by *z* then the outer min-structure is min(1, 1 − *z* + *y*), i.e., the min-structures are formally the same. The implication operations in (8.9) are situated
Standard Fuzzy Sets and some Many-Valued Logics 93
<sup>=</sup> (*<sup>x</sup>* <sup>→</sup> *<sup>y</sup>*)+(*<sup>y</sup>* <sup>→</sup> *<sup>x</sup>*) <sup>−</sup> <sup>|</sup>(*<sup>x</sup>* <sup>→</sup> *<sup>y</sup>*) <sup>−</sup> (*<sup>y</sup>* <sup>→</sup> *<sup>x</sup>*)<sup>|</sup> 2
<sup>=</sup> min(1, 1 <sup>−</sup> *<sup>x</sup>* <sup>+</sup> *<sup>y</sup>*) + min(1, 1 <sup>−</sup> *<sup>y</sup>* <sup>+</sup> *<sup>x</sup>*) 2 <sup>−</sup> min(1, 1 <sup>−</sup> *<sup>x</sup>* <sup>+</sup> *<sup>y</sup>*) <sup>−</sup> min(1, 1 <sup>−</sup> *<sup>y</sup>* <sup>+</sup> *<sup>x</sup>*) 2 <sup>=</sup> <sup>4</sup> <sup>−</sup> <sup>2</sup> <sup>|</sup>*<sup>x</sup>* <sup>−</sup> *<sup>y</sup>*<sup>|</sup> <sup>−</sup> |−2*<sup>x</sup>* <sup>+</sup> <sup>2</sup>*<sup>y</sup>* <sup>−</sup> <sup>|</sup>*<sup>x</sup>* <sup>−</sup> *<sup>y</sup>*<sup>|</sup> <sup>+</sup> <sup>|</sup>*<sup>x</sup>* <sup>−</sup> *<sup>y</sup>*|| 4
<sup>4</sup> <sup>=</sup> <sup>1</sup> <sup>−</sup> <sup>|</sup>*<sup>x</sup>* <sup>−</sup> *<sup>y</sup>*<sup>|</sup> .
These cases are similar to the truth value evaluation rules for connected formulas in Łℵ<sup>1</sup> . Hence, if we want to use algebraic approach for Łℵ<sup>1</sup> we need not necessarily to follow the mainstream described in Section 9 using the operations of MV-algebras for studying the
However, in the next section, we give also a very brief description about the alternative approach starting from the definition of general MV-algebra. It is the mainstream in this
We open another way a little for creating an algebra for Łukasiewicz logic. We adopt the definition and some properties of MV-algebras from Cignoli et. al. (Cignoli et al., 2000). Other
**Definition 9.1.** An *MV-algebra* is an algebra �*A*, ⊕, ¬, 0� with a binary operation ⊕, a unary
<sup>Ł</sup> *<sup>y</sup>*, by (Impl.).
in the same way. Hence, min(1, 1 − *x* + *y*) must be the evaluation rule of *x* →
*x* ↔ *y* = min{(*x* → *y*),(*y* → *x*)}
<sup>=</sup> <sup>4</sup> <sup>−</sup> <sup>4</sup> <sup>|</sup>*<sup>x</sup>* <sup>−</sup> *<sup>y</sup>*<sup>|</sup>
research topic, but a circuitous route in the case of Łukasiewicz logic.
sources are Bergmann (Bergmann, 2008) and Hájek (Hájek, 1998).
operation ¬ and a constant 0 satisfying the following equations:
A non-empty set *A* is the universe of the MV-algebra �*A*, ⊕, ¬, 0�.
**9. The relationship between Łukasiewicz logic and MV-algebras**
Hence, the connectives of Łℵ<sup>1</sup> can be created by Zadeh algebra.
connections between the connectives in Łℵ<sup>1</sup> .
(MV1) *x* ⊕ (*y* ⊕ *z*)=(*x* ⊕ *y*) ⊕ *z*
(MV6) ¬(¬*x* ⊕ *y*) ⊕ *y* = ¬(¬*y* ⊕ *x*) ⊕ *x*
(MV2) *x* ⊕ *y* = *y* ⊕ *x* (MV3) *x* ⊕ 0 = *x* (MV4) ¬¬*x* = *x* (MV5) *x* ⊕ ¬0 = ¬0
The case 5 is deduced as follows:
$$\begin{split} \max(\mathbf{x}, y) &= \min\{1, \max(\mathbf{x}, y)\} = \min\{1, \frac{\mathbf{x} + y + \mid \mathbf{x} - y\mid}{2}\} \\ &= \min\left\{1, \frac{2 - 1 - 1 + \mathbf{x} + 2y - y + \mid 1 - 1 + \mathbf{x} - y\mid}{2}\right\} \\ &= \min\left\{1, 1 - \frac{1 + (1 - \mathbf{x} + y) - \mid 1 - (1 - \mathbf{x} + y)\mid}{2} + y\right\} \\ &= \min\left\{1, 1 - \min(1, 1 - \mathbf{x} + y) + y\right\} \end{split} \tag{8.8}$$
On the other hand, in Łℵ<sup>1</sup> disjunction is defined by(Disj.), i.e., by the formula
$$\mathbf{x} \lor \mathbf{y} = (\mathbf{x} \underset{\mathbf{L}}{\rightarrow} \mathbf{y}) \underset{\mathbf{L}}{\rightarrow} \mathbf{y} \tag{8.9}$$
When we apply the evaluation rule of implication (Impl.) to the right side of the equation (8.9) we get the equation
$$\mathbf{x} \lor y = (\mathbf{x} \to y) \underset{\mathbf{L}}{\to} y = \min\left\{1, 1 - \min(1, 1 - \mathbf{x} + y) + y\right\} = \max(\mathbf{x}, y) \tag{8.10}$$
by (8.8). Hence, the assertion (8.7) follows, and the proof is complete.
Of course, Łukasiewicz must have known the connection between maximum operation and his truth evaluation formula (Impl.) of the implication because without any knowledge about this, he would have not been sure that everything fits well together in his logic. But how he has inferred this is not known. Maybe, he has shown this in some special cases by truth tables with *n* truth values where *n* is finite.
The result of the proof of the formula (8.7) shows that from the join operation max of our algebra we deduce a formula that expresses the rule of Łukasiewicz' implication, and this formula is the truth value evaluation rule in Łℵ<sup>1</sup> . Hence, we have shown that from our algebra (6.7) it is possible to derive similar rules as the truth value evaluation rules in Łℵ<sup>1</sup> .
Hence, we may conclude our main result in a formal way:
**Proposition 8.3.** *If the cases*
*hold, then the other cases*
$$\begin{aligned} \text{3. } &\begin{aligned} \text{3. } &\begin{aligned} \text{x} \wedge y = \min(\text{x}, y); \\ \text{4. } &\begin{aligned} \text{x} \rightarrow y = \min(1, 1 - \text{x} + y); \\ \text{5. } &\begin{aligned} \text{x} \rightarrow y = (\text{x} \rightarrow y) \wedge (y \rightarrow \text{x}) = 1 - |\text{x} - y|. \end{aligned} \end{aligned} \end{aligned} \end{cases}$$
*can be derived based on Zadeh algebra (6.7).*
*Proof.* The case 3 follows from the case 2 by duality. (Actually, this operation already belongs to Zadeh algebra, and hence to Łℵ<sup>1</sup> .)
The case 4 follows from the case 2 by Prop. 8.2 as follows. When we consider the equation (8.10) in the proof of Prop. 8.2, we find two min-stuctures corresponding to the evaluation rule of implication, so that one of them is an inside part of the whole formula. If we denote the inner min-structure min(1, 1 − *x* + *y*) by *z* then the outer min-structure is min(1, 1 − *z* + *y*), i.e., the min-structures are formally the same. The implication operations in (8.9) are situated in the same way. Hence, min(1, 1 − *x* + *y*) must be the evaluation rule of *x* → <sup>Ł</sup> *<sup>y</sup>*, by (Impl.).
The case 5 is deduced as follows:
18 Will-be-set-by-IN-TECH
1, <sup>2</sup> <sup>−</sup> <sup>1</sup> <sup>−</sup> <sup>1</sup> <sup>+</sup> *<sup>x</sup>* <sup>+</sup> <sup>2</sup>*<sup>y</sup>* <sup>−</sup> *<sup>y</sup>*<sup>+</sup> <sup>|</sup> <sup>1</sup> <sup>−</sup> <sup>1</sup> <sup>+</sup> *<sup>x</sup>* <sup>−</sup> *<sup>y</sup>* <sup>|</sup> 2
1, 1 <sup>−</sup> <sup>1</sup> + (<sup>1</sup> <sup>−</sup> *<sup>x</sup>* <sup>+</sup> *<sup>y</sup>*)− | <sup>1</sup> <sup>−</sup> (<sup>1</sup> <sup>−</sup> *<sup>x</sup>* <sup>+</sup> *<sup>y</sup>*) <sup>|</sup>
<sup>Ł</sup> *<sup>y</sup>*) <sup>→</sup>
= min { 1, 1 − min(1, 1 − *x* + *y*) + *y* } (8.8)
<sup>Ł</sup> *<sup>y</sup>* <sup>=</sup> min { 1, 1 <sup>−</sup> min(1, 1 <sup>−</sup> *<sup>x</sup>* <sup>+</sup> *<sup>y</sup>*) + *<sup>y</sup>* } <sup>=</sup> max(*x*, *<sup>y</sup>*) (8.10)
2 }
<sup>2</sup> <sup>+</sup> *<sup>y</sup>*
<sup>Ł</sup> *<sup>y</sup>* (8.9)
max(*x*, *<sup>y</sup>*) = min{ 1, max(*x*, *<sup>y</sup>*) } <sup>=</sup> min{ 1, *<sup>x</sup>* <sup>+</sup> *<sup>y</sup>*<sup>+</sup> <sup>|</sup> *<sup>x</sup>* <sup>−</sup> *<sup>y</sup>* <sup>|</sup>
On the other hand, in Łℵ<sup>1</sup> disjunction is defined by(Disj.), i.e., by the formula
by (8.8). Hence, the assertion (8.7) follows, and the proof is complete.
Hence, we may conclude our main result in a formal way:
*x* ∨ *y* = (*x* →
When we apply the evaluation rule of implication (Impl.) to the right side of the equation (8.9)
Of course, Łukasiewicz must have known the connection between maximum operation and his truth evaluation formula (Impl.) of the implication because without any knowledge about this, he would have not been sure that everything fits well together in his logic. But how he has inferred this is not known. Maybe, he has shown this in some special cases by truth tables
The result of the proof of the formula (8.7) shows that from the join operation max of our algebra we deduce a formula that expresses the rule of Łukasiewicz' implication, and this formula is the truth value evaluation rule in Łℵ<sup>1</sup> . Hence, we have shown that from our algebra
*Proof.* The case 3 follows from the case 2 by duality. (Actually, this operation already belongs
(6.7) it is possible to derive similar rules as the truth value evaluation rules in Łℵ<sup>1</sup> .
= min
= min
we get the equation
*x* ∨ *y* = (*x* →
with *n* truth values where *n* is finite.
**Proposition 8.3.** *If the cases*
*4. x* → *y* = min(1, 1 − *x* + *y*)*;*
*5. x* ↔ *y* = (*x* → *y*) ∧ (*y* → *x*) = 1− | *x* − *y* |*,*
*can be derived based on Zadeh algebra (6.7).*
to Zadeh algebra, and hence to Łℵ<sup>1</sup> .)
*1.* ¬*x* = 1 − *x; 2. x* ∨ *y* = max(*x*, *y*)*; hold, then the other cases 3. x* ∧ *y* = min(*x*, *y*)*;*
<sup>Ł</sup> *<sup>y</sup>*) <sup>→</sup>
$$\begin{split} x \leftrightarrow y &= \min\{ (\mathbf{x} \to y), (y \to \mathbf{x}) \} \\ &= \frac{(\mathbf{x} \to y) + (y \to \mathbf{x}) - |(\mathbf{x} \to y) - (y \to \mathbf{x})|}{2} \\ &= \frac{\min(1, 1 - \mathbf{x} + y) + \min(1, 1 - y + \mathbf{x})}{2} \\ &- \frac{\min(1, 1 - \mathbf{x} + y) - \min(1, 1 - y + \mathbf{x})}{2} \\ &= \frac{4 - 2|\mathbf{x} - y| - |-2\mathbf{x} + 2y - |\mathbf{x} - y| + |\mathbf{x} - y||}{4} \\ &= \frac{4 - 4|\mathbf{x} - y|}{4} = 1 - |\mathbf{x} - y|. \end{split}$$
Hence, the connectives of Łℵ<sup>1</sup> can be created by Zadeh algebra.
These cases are similar to the truth value evaluation rules for connected formulas in Łℵ<sup>1</sup> . Hence, if we want to use algebraic approach for Łℵ<sup>1</sup> we need not necessarily to follow the mainstream described in Section 9 using the operations of MV-algebras for studying the connections between the connectives in Łℵ<sup>1</sup> .
However, in the next section, we give also a very brief description about the alternative approach starting from the definition of general MV-algebra. It is the mainstream in this research topic, but a circuitous route in the case of Łukasiewicz logic.
### **9. The relationship between Łukasiewicz logic and MV-algebras**
We open another way a little for creating an algebra for Łukasiewicz logic. We adopt the definition and some properties of MV-algebras from Cignoli et. al. (Cignoli et al., 2000). Other sources are Bergmann (Bergmann, 2008) and Hájek (Hájek, 1998).
**Definition 9.1.** An *MV-algebra* is an algebra �*A*, ⊕, ¬, 0� with a binary operation ⊕, a unary operation ¬ and a constant 0 satisfying the following equations:
$$(\mathbf{M}\mathbf{V}\mathbf{1})\qquad \mathfrak{x}\oplus(\mathfrak{y}\oplus\mathfrak{z})=(\mathfrak{x}\oplus\mathfrak{y})\oplus\mathfrak{z}$$
$$(\mathsf{MV2})\qquad \mathfrak{x} \oplus \mathfrak{y} = \mathfrak{y} \oplus \mathfrak{x}$$
A non-empty set *A* is the universe of the MV-algebra �*A*, ⊕, ¬, 0�.
can be expressed in MV-algebra in the form (*cf.* Cignoli et al. (Cignoli et al., 2000), p. 78)
Standard Fuzzy Sets and some Many-Valued Logics 95
The equation (9.6) shows that between Łukasiewicz implication ant the operation ⊕ there is a similar connection as in S-implications which are defined by means of disjunction. But ⊕ is not disjunction in Łukasiewicz logic. In many cases ⊕ is interpreted as disjunction, but defined on the unit interval it gives different values as Łukasiewicz disjunction operation max. However, it is possible to define the operations max and min by means of the operations of MV-algebra, but then the result usually is a logic with additional operations having no reasonable interpretations (*cf.* for example, the logic Fuzzy*<sup>L</sup>* in Bergmann's book (Bergmann,
Wajsberg created an algebra, called by *Wajsberg algebra* (*W-algebra*, for short) which is know to serve an algebraic approach to Łukasiewicz infinite-valued logic. The following lemma (*cf.* Cignoli et al. (Cignoli et al., 2000), p. 83) gives the connection between Wajsberg algebras and
The binary operation of W-algebra is implication operation. In this algebra a unary operation is ¬ because it is needed to create the unit element 1. The zero element 0 belongs to this algebra because it implies the unit element by means of negation. Hence, the algebra is in a suitable form according to Łukasiewicz logic. Now, we have counterparts of the primitive connectives of Łukasiewicz logic as the operations of W-algebra. The other connectives can be created in
One consequence from this consideration is that in MV-algebras the operations max and min can be created by the operations of W-algebra, i.e., by the primitive connectives of Łukasiewicz
The main problem we considered here is to find connections between standard fuzzy sets and Łukasiewicz logic Łℵ<sup>1</sup> and to find a suitable algebra for it, especially, because the primitive connectives are negation and implication. In De Morgan algebras the counterparts for the logical connectives disjunction, conjunction, and negation appear as the algebraic operations. It cannot immediately be seen how Łukasiewicz implication, that belongs to the primitive connectives, are derived from the disjunction (max) and negation (¬*x* = 1 − *x*, *x* ∈ **I**). We have done it here using a special De Morgan algebra, namely, Zadeh algebra. Hence, the connection between standard fuzzy sets and Łukasiewicz logic Łℵ<sup>1</sup> becomes clear. The key result, where Łukasiewicz implication is derived algebraically from disjunction and negation,
Kleene's logic is considered because of its close connection to standard fuzzy sets, already motivated by Zadeh. The sections 2, 3, 4, 6, and 8 gives the method we have used for creating
def
<sup>=</sup> <sup>¬</sup>*<sup>x</sup>* <sup>⊕</sup> *y and* <sup>1</sup> def
= ¬0*. Then* �*A*, →, 1� *is a*
2008). Some comments on Fuzzy*<sup>L</sup>* is given in (Mattila, 2010)).
the similar way in W-algebra as Łukasiewicz has introduced them.
**Lemma 9.1.** *Let A be an MV-algebra, and put x* → *y*
logic by means of the evaluation rule of (9.6).
whence,
MV-algebras.
**10. Conclusion**
is given in Proposition 8.3.
our results.
*W-algebra.*
*x* → *y* = ¬*x* ⊕ *y* (9.6)
*x* ⊕ *y* = ¬*x* → *y*. (9.7)
In particular, axioms (MV1) - (MV3) state that �*A*, ⊕, 0� is an *abelian monoid*.
Given an MV-algebra *<sup>A</sup>* and a set *<sup>X</sup>*, the set *<sup>A</sup><sup>X</sup>* of all functions *<sup>f</sup>* : *<sup>X</sup>* −→ *<sup>A</sup>* becomes an MV-algebra if the operations ⊕ and ¬ and the element 0 are defined pointwise. It is obvious that the unit interval [0, 1] is an MV-algebra. The continuous functions from [0, 1] into [0, 1] form a subalgebra of the MV-algebra [0, 1] [0,1] .
On each MV-algebra *A* we define the constant 1 and the operations � and � as follows:
1 def = ¬0 , (9.1)
$$\mathfrak{x} \ominus \mathfrak{y} \stackrel{\text{def}}{=} \neg(\neg \mathfrak{x} \oplus \neg \mathfrak{y}) \, , \tag{9.2}$$
$$
\mathfrak{x} \ominus \mathfrak{y} \stackrel{\text{def}}{=} \mathfrak{x} \ominus \neg \mathfrak{y}.\tag{9.3}
$$
An MV-algebra is nontrivial is and only if 0 �= 1. The following identities are immediate consequences of (MV4):
$$(\mathbf{M}\mathbf{V}\mathbf{\bar{\mathbf{\bar{\mathbf{\bar{\mathbf{\bar{\mathbf{\bar{\mathbf{\bar{\mathbf{\bar{\mathbf{\mathbf{\bar{\mathbf{\mathbf{\bar{\mathbf{\mathbf{\mathbf{\bar{\mathbf{\mathbf{\mathbf{\cdot}}}}}}}}}}}}}}}}}}}}}}}}}$$
}}
(MV8) *x* ⊕ *y* = ¬(¬*x* � ¬*y*).
Axioms (MV5) and (MV6) can now be written as:
(MV5� ) *x* ⊕ 1 = 1 , (MV6� ) (*x* � *y*) ⊕ *y* = (*y* � *x*) ⊕ *x* .
Setting *y* = ¬0 in (MV6) we obtain
$$(\mathsf{MV9})\qquad \mathfrak{x} \oplus \neg \mathfrak{x} = 1\_{\mathsf{M}}$$
In the MV-algebra �[0, 1], ⊕, ¬, 0� we have
$$\mathbf{x} \odot \mathbf{y} = \max(0, \mathbf{x} + \mathbf{y} - 1) \tag{9.4}$$
$$
\pi \ominus y = \max(0, \pi - y) \tag{9.5}
$$
*Notation*: Following common usage, we consider the ¬ operation more binding than any other operation, and the � operation more binding than ⊕ and �.
Consider the question about the connection between Łukasiewicz implication and operations in MV-algebra.
Given an MV-algebra �*A*, ⊕, ¬, 0� and a set *X*, the set
$$A^X = \{ f \mid f: X \longrightarrow A \}$$
becomes an MV-algebra if the operations ⊕ and ¬ and the element 0 are defined pointwice (Cignoli et al. (Cignoli et al., 2000), p. 8). To define 0 pointwice means here that the result is a constant function **0** : *x* �→ 0 for any *x* in the universe of that algebra.
Further, Łukasiewicz implication
$$x \to y \stackrel{\text{def}}{=} \min\{1, 1 - x + y\}$$
can be expressed in MV-algebra in the form (*cf.* Cignoli et al. (Cignoli et al., 2000), p. 78)
$$
\mathfrak{x} \to \mathfrak{y} = \neg \mathfrak{x} \oplus \mathfrak{y} \tag{9.6}
$$
whence,
20 Will-be-set-by-IN-TECH
Given an MV-algebra *<sup>A</sup>* and a set *<sup>X</sup>*, the set *<sup>A</sup><sup>X</sup>* of all functions *<sup>f</sup>* : *<sup>X</sup>* −→ *<sup>A</sup>* becomes an MV-algebra if the operations ⊕ and ¬ and the element 0 are defined pointwise. It is obvious that the unit interval [0, 1] is an MV-algebra. The continuous functions from [0, 1] into [0, 1]
An MV-algebra is nontrivial is and only if 0 �= 1. The following identities are immediate
*Notation*: Following common usage, we consider the ¬ operation more binding than any other
Consider the question about the connection between Łukasiewicz implication and operations
*<sup>A</sup><sup>X</sup>* <sup>=</sup> { *<sup>f</sup>* <sup>|</sup> *<sup>f</sup>* : *<sup>X</sup>* −→ *<sup>A</sup>*}
becomes an MV-algebra if the operations ⊕ and ¬ and the element 0 are defined pointwice (Cignoli et al. (Cignoli et al., 2000), p. 8). To define 0 pointwice means here that the result is a
= min{1, 1 − *x* + *y*}
= ¬0 , (9.1)
= ¬(¬*x* ⊕ ¬*y*), (9.2)
= *x* � ¬*y* . (9.3)
*x* � *y* = max(0, *x* + *y* − 1) (9.4) *x* � *y* = max(0, *x* − *y*) (9.5)
[0,1] .
On each MV-algebra *A* we define the constant 1 and the operations � and � as follows:
1 def
*x* � *y* def
*x* � *y* def
In particular, axioms (MV1) - (MV3) state that �*A*, ⊕, 0� is an *abelian monoid*.
form a subalgebra of the MV-algebra [0, 1]
consequences of (MV4):
(MV8) *x* ⊕ *y* = ¬(¬*x* � ¬*y*).
Setting *y* = ¬0 in (MV6) we obtain
Further, Łukasiewicz implication
) *x* ⊕ 1 = 1 ,
(MV9) *x* ⊕ ¬*x* = 1 ,
in MV-algebra.
Axioms (MV5) and (MV6) can now be written as:
) (*x* � *y*) ⊕ *y* = (*y* � *x*) ⊕ *x* .
In the MV-algebra �[0, 1], ⊕, ¬, 0� we have
operation, and the � operation more binding than ⊕ and �.
constant function **0** : *x* �→ 0 for any *x* in the universe of that algebra.
*x* → *y*
def
Given an MV-algebra �*A*, ⊕, ¬, 0� and a set *X*, the set
(MV7) ¬1 = 0 ,
(MV5�
(MV6�
$$
\mathfrak{x} \oplus \mathfrak{y} = \neg \mathfrak{x} \to \mathfrak{y}.\tag{9.7}
$$
The equation (9.6) shows that between Łukasiewicz implication ant the operation ⊕ there is a similar connection as in S-implications which are defined by means of disjunction. But ⊕ is not disjunction in Łukasiewicz logic. In many cases ⊕ is interpreted as disjunction, but defined on the unit interval it gives different values as Łukasiewicz disjunction operation max. However, it is possible to define the operations max and min by means of the operations of MV-algebra, but then the result usually is a logic with additional operations having no reasonable interpretations (*cf.* for example, the logic Fuzzy*<sup>L</sup>* in Bergmann's book (Bergmann, 2008). Some comments on Fuzzy*<sup>L</sup>* is given in (Mattila, 2010)).
Wajsberg created an algebra, called by *Wajsberg algebra* (*W-algebra*, for short) which is know to serve an algebraic approach to Łukasiewicz infinite-valued logic. The following lemma (*cf.* Cignoli et al. (Cignoli et al., 2000), p. 83) gives the connection between Wajsberg algebras and MV-algebras.
**Lemma 9.1.** *Let A be an MV-algebra, and put x* → *y* def <sup>=</sup> <sup>¬</sup>*<sup>x</sup>* <sup>⊕</sup> *y and* <sup>1</sup> def = ¬0*. Then* �*A*, →, 1� *is a W-algebra.*
The binary operation of W-algebra is implication operation. In this algebra a unary operation is ¬ because it is needed to create the unit element 1. The zero element 0 belongs to this algebra because it implies the unit element by means of negation. Hence, the algebra is in a suitable form according to Łukasiewicz logic. Now, we have counterparts of the primitive connectives of Łukasiewicz logic as the operations of W-algebra. The other connectives can be created in the similar way in W-algebra as Łukasiewicz has introduced them.
One consequence from this consideration is that in MV-algebras the operations max and min can be created by the operations of W-algebra, i.e., by the primitive connectives of Łukasiewicz logic by means of the evaluation rule of (9.6).
### **10. Conclusion**
The main problem we considered here is to find connections between standard fuzzy sets and Łukasiewicz logic Łℵ<sup>1</sup> and to find a suitable algebra for it, especially, because the primitive connectives are negation and implication. In De Morgan algebras the counterparts for the logical connectives disjunction, conjunction, and negation appear as the algebraic operations. It cannot immediately be seen how Łukasiewicz implication, that belongs to the primitive connectives, are derived from the disjunction (max) and negation (¬*x* = 1 − *x*, *x* ∈ **I**). We have done it here using a special De Morgan algebra, namely, Zadeh algebra. Hence, the connection between standard fuzzy sets and Łukasiewicz logic Łℵ<sup>1</sup> becomes clear. The key result, where Łukasiewicz implication is derived algebraically from disjunction and negation, is given in Proposition 8.3.
Kleene's logic is considered because of its close connection to standard fuzzy sets, already motivated by Zadeh. The sections 2, 3, 4, 6, and 8 gives the method we have used for creating our results.
**5**
*Mexico*
**Parametric Type-2 Fuzzy Logic Systems**
The use of Fuzzy Logic Systems (FLS) for control applications has increased since they became popular from 80's. After Mendel in 90's showed how uncertainty can be computed in order to achieve more robust systems, Type-2 Fuzzy Logic Systems (T2FLS) are in the
At same time, Batyrshin et al demonstrated that parametric conjunctions can be useful for tuning a FLS in order to achieve better performance beyond the set parameter tuning. In signal processing and system identification, this fact let the designer to add freedom degrees
This chapter presents the parametric T2FLS and shows that this new FLS is a very useful option for sharper approximations in control. In order to verify the advantages of the parametric T2FLS, it is used the Ball and Plate System as a testbench. This study case helps us to understand how a parametric conjunction affects the controller behavior in measures like response time or overshoot. Also, this application let us observe how the controller
A Parametric Type-2 Fuzzy Logic Systems (PT2FLS) is a general FLS which can be fully adjusted through a single or multiple parameters in order to achieve a benefit in its general performance. It means that a PT2FLS has several options to adjust set parameters (i.e. membership function parameters), rule parameters and output set parameters. Fig. 1 shows
In this figure the Defuzzification stage comprises the Output Processing Block and the Defuzzifier as Mendel stated in (Karnik, Mendel et al. 1999). For Interval Type-2 Fuzzy Logic Systems (IT2FLS) this block represents only the centroid calculation for example considering the WM Algorithm (Wu and Mendel 2002). As it can be seen, a dashed arrow
A general Fuzzy System is a function where all input variables are mapped to the output variables according to the knowledge base defined by rules. Rule Set represents the
crosses every stage; this means that every stage is tunable for optimization purposes.
focus of researchers and recently they became a new research topic.
the structure of a PT2FLS which it is almost equal to a general T2FLS.
**1. Introduction**
to adjust a general FLS.
works in noise presence.
**2. Parametric T2FLS**
configuration of the T2FLS.
Arturo Tellez, Heron Molina, Luis Villa,
Elsa Rubio and Ildar Batyrshin
*IPN, CIC, Mexico City,*
The section 9 tells very briefly how the others consider this topic. That way is different and alternative to ours. MV-algebra is quite general, and many algebras, like Boolean algebras and also De Morgan algebras belong to its scope. The reader may become familiar with this topic, for example, by reading the Bergmann's, Cignoli's et. al, and Hájek's books mentioned in References. A lot of other material is available, too.
Our alternative way we have considered the topic here, is not totally new, because these things are considered in (Mattila, 2004), (Mattila, 2005), and (Mattila, 2010), but our key result, Proposition 8.3 is. Proposition 8.2 is the core of this result. It makes the connection of max operation and Łukasiewicz inmplication clear by means of Zadeh algebra. Using the expressions (8.4) for the operations max and min is usually not used in general, but it makes the consideration easy. As we can see from the used references, De Morgan algebras have already been well known relatively long time having, in the long run, different alternative names, like "quasi Boolean algebras", and "soft algebras".
H. Rasiowa has considered *implicative algebras* and *implication algebras* in her book ((Rasiowa, 1974)). Hence, the future research policy may be based on these algebras. Also, the connections between implicative/implication algebras and De Morgan algebras or MV-algebras restricted to many-valued or modal logics are included to the future research.
### **11. References**
## **Parametric Type-2 Fuzzy Logic Systems**
Arturo Tellez, Heron Molina, Luis Villa, Elsa Rubio and Ildar Batyrshin *IPN, CIC, Mexico City, Mexico*
### **1. Introduction**
22 Will-be-set-by-IN-TECH
96 Fuzzy Logic – Algorithms, Techniques and Implementations
The section 9 tells very briefly how the others consider this topic. That way is different and alternative to ours. MV-algebra is quite general, and many algebras, like Boolean algebras and also De Morgan algebras belong to its scope. The reader may become familiar with this topic, for example, by reading the Bergmann's, Cignoli's et. al, and Hájek's books mentioned
Our alternative way we have considered the topic here, is not totally new, because these things are considered in (Mattila, 2004), (Mattila, 2005), and (Mattila, 2010), but our key result, Proposition 8.3 is. Proposition 8.2 is the core of this result. It makes the connection of max operation and Łukasiewicz inmplication clear by means of Zadeh algebra. Using the expressions (8.4) for the operations max and min is usually not used in general, but it makes the consideration easy. As we can see from the used references, De Morgan algebras have already been well known relatively long time having, in the long run, different alternative
H. Rasiowa has considered *implicative algebras* and *implication algebras* in her book ((Rasiowa, 1974)). Hence, the future research policy may be based on these algebras. Also, the connections between implicative/implication algebras and De Morgan algebras or MV-algebras restricted to many-valued or modal logics are included to the future research.
Bergmann, M. (2008). *An Introduction to Many-Valued and Fuzzy Logic*, Cambridge University Press, New York, Melbourne, Madrid,Cape Town, Singapore, São Paulo, Delhi. Cignoli, L. & D'Ottaviano, M. & Mundici, D. (2000). *Algebraic Foundations of Many-valued Reasoning*, Kluwer Academic Publishers, Dordrecht, Boston, London. Hájek, P. (1998). *Metamathematics of Fuzzy Logic*, Kluwer Academic Publishers, Dordrecht,
Lowen, R. (1996). *Fuzzy Set Theory. Basic Concepts, Techniques and Bibliography*, Kluwer
Mattila, J. K. (2004). Zadeh algebras as a syntactical approach to fuzzy sets, *Current Issues in Data and Knowledge Engineering*, Springer-Verlag, Warszawa, Poland, pp. 343–349. Mattila, J. K. (year 2005). On łukasiewicz modifier logic, *Journal of Advanced Computational*
Mattila, J. K. (2009). Many-valuation, modality, and fuzziness, *in* Seising R. (ed.), *Views on*
Mattila, J. K. (2010). On Łukasiewicz' infinite-Valued logic and Fuzzy*L*, *in*, *KES 2010*
Negoit ˘a, C. V. & Ralescu, D. A., (1975). *Applications of Fuzzy Sets to Systems Analysis*,
Rescher, N. (1969). *Many-Valued Logic*, McGraw-Hill, New York, St. Louis, San Francisco,
*Fuzzy Sets and Systems from Different Perspectives. Philosophy and Logic, Criticisms and*
*Knowledge-Based Intelligent Information and Engineering Systems*, Part IV, LNAI 6279,
Academic Publishers, Dordrecht, Boston, London.
Springer-Verlag, Berlin, Heidelberg, pp. 108–115.
London, Sydney, Toronto, Mexiko, Panama. Zadeh, L. A. (1965). Fuzzy Sets, *Information and Controll* 8.
*Intelligence and Intelligent Informatics* Vol. 9(No. 5): 506–510.
*Applications*, Springer-Verlag, Berlin, Heidelberg, pp. 271–300.
Rasiowa, H. (1974). *An Algebraic Approach to non-classical Logics*, North-Holland.
in References. A lot of other material is available, too.
names, like "quasi Boolean algebras", and "soft algebras".
**11. References**
Boston, London.
URL: *www.fujipress.jp*
Birkhäuser, Basel, Stuttgart.
The use of Fuzzy Logic Systems (FLS) for control applications has increased since they became popular from 80's. After Mendel in 90's showed how uncertainty can be computed in order to achieve more robust systems, Type-2 Fuzzy Logic Systems (T2FLS) are in the focus of researchers and recently they became a new research topic.
At same time, Batyrshin et al demonstrated that parametric conjunctions can be useful for tuning a FLS in order to achieve better performance beyond the set parameter tuning. In signal processing and system identification, this fact let the designer to add freedom degrees to adjust a general FLS.
This chapter presents the parametric T2FLS and shows that this new FLS is a very useful option for sharper approximations in control. In order to verify the advantages of the parametric T2FLS, it is used the Ball and Plate System as a testbench. This study case helps us to understand how a parametric conjunction affects the controller behavior in measures like response time or overshoot. Also, this application let us observe how the controller works in noise presence.
### **2. Parametric T2FLS**
A Parametric Type-2 Fuzzy Logic Systems (PT2FLS) is a general FLS which can be fully adjusted through a single or multiple parameters in order to achieve a benefit in its general performance. It means that a PT2FLS has several options to adjust set parameters (i.e. membership function parameters), rule parameters and output set parameters. Fig. 1 shows the structure of a PT2FLS which it is almost equal to a general T2FLS.
In this figure the Defuzzification stage comprises the Output Processing Block and the Defuzzifier as Mendel stated in (Karnik, Mendel et al. 1999). For Interval Type-2 Fuzzy Logic Systems (IT2FLS) this block represents only the centroid calculation for example considering the WM Algorithm (Wu and Mendel 2002). As it can be seen, a dashed arrow crosses every stage; this means that every stage is tunable for optimization purposes.
A general Fuzzy System is a function where all input variables are mapped to the output variables according to the knowledge base defined by rules. Rule Set represents the configuration of the T2FLS.
$$\begin{aligned} F^l &= \left[ \overline{F}^l; \underline{F}^l \right] = \left[ \overline{\mu}\_{\tilde{A}\_l^l} (\boldsymbol{\chi}\_l); \,\underline{\mu}\_{\tilde{A}\_l^l} (\boldsymbol{\chi}\_l) \right] \\\\ \overline{F}^l &= \overline{\mu}\_{\tilde{A}\_1^l} (\boldsymbol{\chi}\_1) \wedge \overline{\mu}\_{\tilde{A}\_2^l} (\boldsymbol{\chi}\_2) \wedge \dots \wedge \overline{\mu}\_{\tilde{A}\_m^l} (\boldsymbol{\chi}\_m) \\\\ \underline{F}^l &= \underline{\mu}\_{\tilde{A}\_1^l} (\boldsymbol{\chi}\_1) \wedge \underline{\mu}\_{\tilde{A}\_1^l} (\boldsymbol{\chi}\_2) \wedge \dots \wedge \underline{\mu}\_{\tilde{A}\_m^l} (\boldsymbol{\chi}\_m) \end{aligned}$$
$$\overline{F}^l = \mathbf{T}\left(\overline{\mu}\_{\vec{A}\_1^l}(\mathbf{x}\_1), \overline{\mu}\_{\vec{A}\_2^l}(\mathbf{x}\_2), \dots, \overline{\mu}\_{\vec{A}\_m^l}(\mathbf{x}\_m); \mathbb{p}\_r\right) = \mathbf{T}\left(\overline{\mu}\_{\vec{A}\_l^l}(\mathbf{x}\_l); \mathbb{p}\_r^l\right) \tag{1}$$
$$\underline{F}^l = \mathbf{T}\left(\underline{\mu}\_{\tilde{A}\_1^l}(\mathbf{x}\_1), \underline{\mu}\_{\tilde{A}\_2^l}(\mathbf{x}\_2), \dots, \underline{\mu}\_{\tilde{A}\_m^l}(\mathbf{x}\_m); \mathbf{p}\_r\right) = \mathbf{T}\left(\underline{\mu}\_{\tilde{A}\_l^l}(\mathbf{x}\_l); \mathbf{p}\_r^l\right) \tag{2}$$
$$\overline{\mu}\_{\mathcal{B}\_1^{\ell}}(\mathbf{y}) = \sqcup\_{\mathbf{x} \in \mathcal{X}} \left( \mathbf{T} \left( \overline{\mu}\_{\mathcal{A}\_l^{\ell}}(\mathbf{x}\_l); \mathbb{p}\_r \right) \right) \tag{3}$$
$$\underline{\mu}\_{\mathcal{B}\_1^{\ell}}(\mathcal{y}) = \sqcup\_{\mathbf{x} \in \mathcal{X}} \left( \mathbf{T} \left( \underline{\mu}\_{\mathcal{A}\_l^{\ell}}(\mathbf{x}\_l); \mathbb{p}\_r \right) \right) \tag{4}$$
$$T(a,b) = T\_{ll}(a,b) \text{ if } (a,b) \in D\_{lj}; t, j \in G \tag{5}$$
$$T\_{l,l} = T\_{l,l} \tag{6}$$
$$T(a,b,p) = \begin{cases} T\_{11}(a,b), & (a \le p) \land (b \le p) \\ T\_{21}(a,b), & (a > p) \land (b \le p) \\ T\_{12}(a,b), & (a \le p) \land (b > p) \\ T\_{22}(a,b), & (a > p) \land (b > p) \end{cases} \tag{7}$$
$$T(a,b,p) = \begin{cases} T\_{11}(a,b), & (a \le p) \land (b \le p) \\ T\_{21}(a,b), & (a > p) \land (a \le 1-p) \land (b \le p) \\ T\_{12}(a,b), & (a > 1-p) \land (b \le p) \\ T\_{12}(a,b), & (a \le p) \land (b > p) \land (b \le 1-p) \\ T\_{22}(a,b), & (a > p) \land (a \le 1-p) \land (b > p) \land (b \le 1-p) \\ T\_{32}(a,b), & (a > 1-p) \land (b > p) \land (b \le 1-p) \\ T\_{13}(a,b), & (a \le p) \land (b > p) \land (b \le 1-p) \\ T\_{23}(a,b), & (a > p) \land (a \le 1-p) \land (b > p) \\ T\_{33}(a,b), & (a > p) \land (b > p) \end{cases} \tag{8}$$
$$H\_{p-2} \xrightarrow{p} \begin{bmatrix} 1 & p & t-p \\ D\_{12} & D\_{22} & D\_{32} \\ D\_{13} & D\_{23} & D\_{33} \\ & H\_{12} & D\_{12} & D\_{23} \\ H\_{12} & H\_{21} & D\_{22} & D\_{33} \end{bmatrix} t-p$$
$$H\_{p-1} \xrightarrow[h \to H\_{p-1}]{H\_{p-1} \xrightarrow[h \to H\_{p-1}]{H\_{p-1} \xrightarrow[h \to H\_{p-1}]}} 1 \quad H\_{p-2} \xrightarrow[h \to H\_{p-2}]{H\_{p-1} \xrightarrow[h \to H\_{p-1}]} 1$$
Parametric Type-2 Fuzzy Logic Systems 105
ball, over the axis X. If it is assumed that the velocity and the acceleration of the ball are constant at shorter values of �, it is possible to estimate the next position of ball with
NA NM Z PM PA
NA NM Z PM PA PA NM NM NM PA PA PA Z NA NA Z PA PA PM NA NA NA PM PM PA NA NA NM Z PM
��(���) and finally the desired position �� and the estimated error �(���).
**Error**
Rubio-Espino et al. 2010) used for the PT2FLC purposes.
Fig. 7. IT2FLS Simulator for B&P System
negative" (NA). Rule set is described in Table 1.
corresponds to every axis.
**Tilt Change**
Table 1. Optimal Rule Set of B&P System with T1FLC described in (Moreno-Armendariz,
It is proved that B&P System is a decoupled system over its two axes (Moreno-Armendariz, Rubio-Espino et al. 2010). So, (10) are similar for the axis Y. Fig 5 shows that B&P system block has two inputs and two outputs for our control purposes; so, every in-out pair
T1FLC proposed by (Moreno-Armendariz, Rubio-Espino et al. 2010) has two inputs and one output. Every variable has 5 FS (Fig. 6) associated to linguistic variables "high positive" (PA), "medium positive" (PM), "zero or null" (Z), "medium negative" (NM), and "high
The T1FLC controls the tilt of plate using the information that the FPGA takes form the camera and calculates the current position using (10) under perfect environment conditions. But what happens when some external forces (e.g. weather) complicate the system stability? Some equivalent phenomena may be introduced to the plate. For example, the illuminating
in axis Y. Servomotors perform the adequate tilt over both axes. Every tilt value is calculated by its corresponding PT2FLC using the error position and the position change. Position change is the differential of the feedback of the plant, i.e. the current position.
It is noteworthy that PT2FLC hardware has not been implemented and tested for this application. Only simulations are performed in order to show all advantages of the use of PT2FLC in control applications. Mechanical model proposed in (Moreno-Armendariz, Rubio-Espino et al. 2010) has the characteristic of designing and testing new and improved controllers, which it is a suitable future work, because of the flexibility of FPGA.
$$A = \begin{bmatrix} 0 & 1 & 0 & 0 & 0 & 0 & 0 & 0 \\ 0 & 0 & -9.81 & 0 & 0 & 0 & 0 & 0 \\ 0 & 0 & 0 & 1 & 0 & 0 & 0 & 0 \\ -6.1313 \times 10^4 & 0 & 0 & 0 & 0 & 0 & 0 & 0 \\ 0 & 0 & 0 & 0 & 0 & 1 & 0 & 0 \\ 0 & 0 & 0 & 0 & 0 & 0 & -9.81 & 0 \\ 0 & 0 & 0 & 0 & 0 & 0 & 0 & 1 \\ 0 & 0 & 0 & 0 & -6.1313 \times 10^4 & 0 & 0 & 0 \end{bmatrix}$$
$$B = \begin{bmatrix} 0 & 0 & 0 \\ 0 & 0 & 0 \\ 5.6618 \times 10^4 & 0 \\ 0 & 0 & 0 \\ 0 & 0 & 0 \\ 0 & 0 & 0 \\ 0 & 5.6618 \times 10^4 \end{bmatrix} \tag{9}$$
$$C = \begin{bmatrix} 1 & 0 & 0 & 0 & 0 & 0 & 0 \\ 0 & 0 & 0 & 0 & 0 & 0 & 0 \\ 0 & 0 & 0 & 1 & 0 & 0 & 0 \end{bmatrix}$$
$$D = \begin{bmatrix} 1 & 0 & 0 & 0 & 0 & 0 & 0 \\ 0 & 0 & 0 & 0 & 1 & 0 & 0 \end{bmatrix}$$
$$x' = Ax + Bu$$
$$y = Cx$$
The characteristics of this B&P System (9) is a linearized state-space model, the same as described in (Moreno-Armendariz, Rubio-Espino et al. 2010). With (10), it can be calculated the current velocity, acceleration and position in axis x.
$$\begin{aligned} v(k) &= \frac{\{\mathbf{x}(k) - \mathbf{x}(k-1)\}}{T} \\ a(k) &= v(k) - v(k-1) \\ \mathbf{x}\_e(k+1) &= \mathbf{x}(k) + \frac{v(k)}{T} + \frac{a(k)T^2}{2} \\ e(k+1) &= \mathbf{x}\_d - \mathbf{x}\_e(k+1) \end{aligned} \tag{10}$$
The vision system described in (Moreno-Armendariz, Rubio-Espino et al. 2010) uses a sampling time � (50ms), which captures and processes a single image in that period. After the vision system process the image, FPGA calculates the current position of the ball in axis X, �(�), where � is the current sample. Once it is known the position, it is possible to find the current velocity component �(�) and the current acceleration component �(�) of the 104 Fuzzy Logic – Algorithms, Techniques and Implementations
in axis Y. Servomotors perform the adequate tilt over both axes. Every tilt value is calculated by its corresponding PT2FLC using the error position and the position change. Position
It is noteworthy that PT2FLC hardware has not been implemented and tested for this application. Only simulations are performed in order to show all advantages of the use of PT2FLC in control applications. Mechanical model proposed in (Moreno-Armendariz, Rubio-Espino et al. 2010) has the characteristic of designing and testing new and improved
change is the differential of the feedback of the plant, i.e. the current position.
controllers, which it is a suitable future work, because of the flexibility of FPGA.
� � � � � � �
the current velocity, acceleration and position in axis x.
� 0 100 0 000 0 0 −9.81 0 0 0 0 0 0 001 0 000 −6.1313 × 10� 000 0 000 0 000 0 100 0 0 0 0 0 0 −9.81 0 0 000 0 001 0 0 0 0 −6.1313 × 10� 000�
> � 0 0 0 0 0 0 5.6818 × 10� 0 0 0 0 0 0 0 0 5.6818 × 10��
�=�<sup>10000000</sup> 00001000�
> � = [0] �� = �� + �� � = ��
�(�) <sup>=</sup> ��(�) − �(�−1)� �
�(�+1) = �� − ��(�+1)
��(�+1) = �(�) +
The characteristics of this B&P System (9) is a linearized state-space model, the same as described in (Moreno-Armendariz, Rubio-Espino et al. 2010). With (10), it can be calculated
> �(�) � <sup>+</sup>
The vision system described in (Moreno-Armendariz, Rubio-Espino et al. 2010) uses a sampling time � (50ms), which captures and processes a single image in that period. After the vision system process the image, FPGA calculates the current position of the ball in axis X, �(�), where � is the current sample. Once it is known the position, it is possible to find the current velocity component �(�) and the current acceleration component �(�) of the
� � � � � � �
� = (9)
(10) �(�) = �(�) − �(�−1)
�(�)�� 2
� � � � � � �
� =
� � � � � � � ball, over the axis X. If it is assumed that the velocity and the acceleration of the ball are constant at shorter values of �, it is possible to estimate the next position of ball with ��(���) and finally the desired position �� and the estimated error �(���).
Table 1. Optimal Rule Set of B&P System with T1FLC described in (Moreno-Armendariz, Rubio-Espino et al. 2010) used for the PT2FLC purposes.
Fig. 7. IT2FLS Simulator for B&P System
It is proved that B&P System is a decoupled system over its two axes (Moreno-Armendariz, Rubio-Espino et al. 2010). So, (10) are similar for the axis Y. Fig 5 shows that B&P system block has two inputs and two outputs for our control purposes; so, every in-out pair corresponds to every axis.
T1FLC proposed by (Moreno-Armendariz, Rubio-Espino et al. 2010) has two inputs and one output. Every variable has 5 FS (Fig. 6) associated to linguistic variables "high positive" (PA), "medium positive" (PM), "zero or null" (Z), "medium negative" (NM), and "high negative" (NA). Rule set is described in Table 1.
The T1FLC controls the tilt of plate using the information that the FPGA takes form the camera and calculates the current position using (10) under perfect environment conditions. But what happens when some external forces (e.g. weather) complicate the system stability? Some equivalent phenomena may be introduced to the plate. For example, the illuminating
Parametric Type-2 Fuzzy Logic Systems 107
Fig. 8. Second approximation of PT2FLC modifying the FOU of sets
optimal set distribution.
Fig. 9. Parametric Fuzzy Conjunction using (p) −Monotone Sum with parameter p = 0.25
In first experiments, (Fig. 6) it was re-adjusted the FOU of every set, leaving the set distribution intact, so it was found that only for a very thin FOU in every input set it is gotten a good convergence without overshoot and other phenomena. But, what is the sense of having a very short FOU like T1FS if they will not capture the associated uncertainties of the system? So, there should be a way of tuning the PT2FLC without changing this initial
variation due to light incidence over the plate, an unbalanced motor tied to the plate, a low quality image sensor or some interference noise added to the processed image, may be introduced as external disturbances.
Table 2. Phenomena associated with the FOU of every set in system
In initial experiments, noise-free optimization is performed and similar results are achieved in order to compare it with T1FLC. For noise tests it is only considered an unbalanced motor tied to the plate that makes it tremble while a sine trajectory is performed, analyzing a single axis. This experiment helps us to verify the noise-proof ability of the T2FLC.
### **4. Experimental results**
FS distribution, i.e., FS shape parameters may arise several characteristic phenomena that expert must take into account when designing applied-to-control fuzzy systems, so-called Fuzzy Logic Controller (FLC). As described in (Moreno-Armendariz, Rubio-Espino et al. 2010), authors found an optimal FS distribution where FLC shows a great performance in 3.8 seconds. However, when it is used this same configuration some phenomena arises when it is introduced T2FS.
Starting from the initial optimal set distribution and without considering any possible noise influence, it was tested several configurations modifying every set FOU, starting from a T1FS (without FOU) and increasing it as much as possible; or starting from a very wide FOU and collapsing it until it becomes a T1FS. Some phenomena are related to them as described in Table 2, but in general, when it is introduced a T2FS a certain level of overshoot is found, no matter which variable was modified; so, if every variable has a T2FS, then the expert has to deal with the influence of nonlinear aggregation of overshoot, steady-state error or offset (SSE) and ripple, when tuning a PT2FLC, which might be a complicated task.
For every experiment it was used an implemented simulator for IT2FLS. With some instructions it can be constructed any parametric IT2FLS and expert may choose from set shape, several parametric conjunctions and defuzzification options (Fig. 7).
106 Fuzzy Logic – Algorithms, Techniques and Implementations
variation due to light incidence over the plate, an unbalanced motor tied to the plate, a low quality image sensor or some interference noise added to the processed image, may be
which set FOUs are increasing from zero. Yes No No
In initial experiments, noise-free optimization is performed and similar results are achieved in order to compare it with T1FLC. For noise tests it is only considered an unbalanced motor tied to the plate that makes it tremble while a sine trajectory is performed, analyzing a single
FS distribution, i.e., FS shape parameters may arise several characteristic phenomena that expert must take into account when designing applied-to-control fuzzy systems, so-called Fuzzy Logic Controller (FLC). As described in (Moreno-Armendariz, Rubio-Espino et al. 2010), authors found an optimal FS distribution where FLC shows a great performance in 3.8 seconds. However, when it is used this same configuration some phenomena arises when it
Starting from the initial optimal set distribution and without considering any possible noise influence, it was tested several configurations modifying every set FOU, starting from a T1FS (without FOU) and increasing it as much as possible; or starting from a very wide FOU and collapsing it until it becomes a T1FS. Some phenomena are related to them as described in Table 2, but in general, when it is introduced a T2FS a certain level of overshoot is found, no matter which variable was modified; so, if every variable has a T2FS, then the expert has to deal with the influence of nonlinear aggregation of overshoot, steady-state error or offset (SSE) and ripple, when tuning a PT2FLC, which might be a
For every experiment it was used an implemented simulator for IT2FLS. With some instructions it can be constructed any parametric IT2FLS and expert may choose from set
shape, several parametric conjunctions and defuzzification options (Fig. 7).
**Experiment Overshoot SSE Ripple**
No No Yes
Yes No No
No Yes No
introduced as external disturbances.
are wide as much as it can be possible.
are wide as much as it can be possible.
**4. Experimental results**
is introduced T2FS.
complicated task.
When all sets in every variable are T1, except the variable
When all sets in input variable error are T2 and their FOU are decreasing until they become T1. All other variable sets
When all sets in input variable change are T2 and their FOU are decreasing until they become T1. All other variable sets are wide as much as it can be possible.
When all sets in output variable tilt are T2 and their FOU are decreasing until they become T1. All other variable sets
Table 2. Phenomena associated with the FOU of every set in system
axis. This experiment helps us to verify the noise-proof ability of the T2FLC.
Fig. 8. Second approximation of PT2FLC modifying the FOU of sets
Fig. 9. Parametric Fuzzy Conjunction using (p) −Monotone Sum with parameter p = 0.25
In first experiments, (Fig. 6) it was re-adjusted the FOU of every set, leaving the set distribution intact, so it was found that only for a very thin FOU in every input set it is gotten a good convergence without overshoot and other phenomena. But, what is the sense of having a very short FOU like T1FS if they will not capture the associated uncertainties of the system? So, there should be a way of tuning the PT2FLC without changing this initial optimal set distribution.
Parametric Type-2 Fuzzy Logic Systems 109
**shoot SSE Ripple**
No No No
No No Yes
Yes No No
No Yes No
No No Yes
**Parameter Description Over**
non-parametric conjunctions.
parameter value of rule 18.
7b, 19c These rules increase or decrease the offset of
8b This rule help to stretch the ripple slightly but
Fig. 10. Rule 18 parameter distribution for 43 experiments
conjunctions also.
overshoot.
These rules have no influence with the final response, so their parameter values might be any. These rules may be quantified using
These rules have a very slight influence with the final response. Some of them reduce the ripple, but they are negligible. These rules may be quantified using non-parametric
These rules have a very positive influence with the final response, especially the
the final response, but could add some
also might be useful to reduce small ripple.
Table 3. Phenomena associated with the rule operator of every implication in inference
**Rule**
1, 2, 5, 6, 10, 16, 17, 20, 21, 22, 25
3, 4, 9, 11, 12,
13b, 14c, 18a, 23b, 24b
15
In second experiments, it was moved the FOU of every set in every variable and found a very close approximation of time response as described in Fig. 8. This configuration has wider FOU in every input and output variable as much as necessary (with uniform spread) for supporting variations in error until 0.0075 radians, in change until 0.01 radians per second and in tilt until 0.004 radians, all around the mean of every point of its corresponding set and variable.
As it can be seen, every set exhibits a wider FOU and its time response has increased over 5 seconds. Also, some overshoot and ripple are present, but reference is reached, so SSE is eliminated. This is the first best approximation using the same optimal distribution of sets, although it does not mean that there could not be any other set distribution for this application.
As it is sated in (Batyrshin, Rudas et al. 2009), a parametric operator may help to tune a T1FLC through the inference step, so every rule of the knowledge base related with the implication of the premises might be a parametric conjunction. In third experiments, it is used commutative (�) −monotone sum of conjunctions (11), where it is assigned to every section the following conjunctions: ��� = �� is the drastic intersection, ��� = ��� = �� is the product and ��� = �� is the minimum, using (7) as follows:
$$T(\mathbf{x}, \mathbf{y}, p) = \begin{cases} T\_d(\mathbf{x}, \mathbf{y}), \ (\mathbf{x} \le p) \land (\mathbf{y} \le p) \\ T\_p(\mathbf{x}, \mathbf{y}), \ [(\mathbf{x} > p) \land (\mathbf{y} \le p)] \lor [(\mathbf{x} \le p) \land (\mathbf{y} > p)] \\ T\_m(\mathbf{x}, \mathbf{y}), \ (\mathbf{x} > p) \land (\mathbf{y} > p) \end{cases} \tag{11}$$
In (10), it is possible to assure that when parameter �=0 then the conjunction in (11) will have a minimum t-norm behavior, but when parameter �=1, it will be a drastic product tnorm behavior as it can be seen in Fig. 9. If � has any other value between the interval (0,1), then it will have a drastic, product o minimum t-norm behavior depending on the membership values of operands. Resulting behavior of this monotone sum might help to diminish the fuzzy implication between two membership degrees of premises and therefore to reduce the resulting overshoot of system and then reach the reference faster. Now another task is to choose the values of every parameter of conjunctions.
Moreover the optimal FS distribution, it is used the same rule set of (Moreno-Armendariz, Rubio-Espino et al. 2010) as shown in Table 1 in order to show that any T1FLC can be extended to a PT2FLC. So, � = 25 rules define the T1FLC configuration, it means that there are 25 parametric conjunctions and therefore 25 parameters. When searching for an optimal value of every �, it is recommended to use an optimization algorithm in order to obtain optimal values and the resulting waste of time when calculating them manually.
According to (11), the initial values of ��, make the conjunctions to behave like min, i.e.
$$\mathbb{p}\_r = \{0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0\}$$
It is proposed some values when optimization was performed with heuristics to get optimal rule parameters, i.e.
$$\mathfrak{p}\_r = \{0, 0.25, 0.25, 0, 0, 0, 0.2, 0, 0, 1, 0, 1, 0, 0, 0.2, 0, 0, 0, 0, 0.25, 0.25, 0\} \tag{12}$$
108 Fuzzy Logic – Algorithms, Techniques and Implementations
In second experiments, it was moved the FOU of every set in every variable and found a very close approximation of time response as described in Fig. 8. This configuration has wider FOU in every input and output variable as much as necessary (with uniform spread) for supporting variations in error until 0.0075 radians, in change until 0.01 radians per second and in tilt until 0.004 radians, all around the mean of every point of its
As it can be seen, every set exhibits a wider FOU and its time response has increased over 5 seconds. Also, some overshoot and ripple are present, but reference is reached, so SSE is eliminated. This is the first best approximation using the same optimal distribution of sets, although it does not mean that there could not be any other set distribution for this
As it is sated in (Batyrshin, Rudas et al. 2009), a parametric operator may help to tune a T1FLC through the inference step, so every rule of the knowledge base related with the implication of the premises might be a parametric conjunction. In third experiments, it is used commutative (�) −monotone sum of conjunctions (11), where it is assigned to every section the following conjunctions: ��� = �� is the drastic intersection, ��� = ��� = �� is the
��(�, �), [(���) ∧ (���)] ∨ [(���) ∧ (���)]
(11)
��(�, �), (���) ∧ (���)
��(�, �), (���) ∧ (���)
In (10), it is possible to assure that when parameter �=0 then the conjunction in (11) will have a minimum t-norm behavior, but when parameter �=1, it will be a drastic product tnorm behavior as it can be seen in Fig. 9. If � has any other value between the interval (0,1), then it will have a drastic, product o minimum t-norm behavior depending on the membership values of operands. Resulting behavior of this monotone sum might help to diminish the fuzzy implication between two membership degrees of premises and therefore to reduce the resulting overshoot of system and then reach the reference faster. Now another
Moreover the optimal FS distribution, it is used the same rule set of (Moreno-Armendariz, Rubio-Espino et al. 2010) as shown in Table 1 in order to show that any T1FLC can be extended to a PT2FLC. So, � = 25 rules define the T1FLC configuration, it means that there are 25 parametric conjunctions and therefore 25 parameters. When searching for an optimal value of every �, it is recommended to use an optimization algorithm in order to obtain
optimal values and the resulting waste of time when calculating them manually.
According to (11), the initial values of ��, make the conjunctions to behave like min, i.e.
�� = [0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0] It is proposed some values when optimization was performed with heuristics to get optimal
�� = [0,0.25,0.25,0,0,0,0,0.2,0,0,1,0,1,0,1,0,0,0.2,0,0,0,0,0.25,0.25,0] (12)
product and ��� = �� is the minimum, using (7) as follows:
task is to choose the values of every parameter of conjunctions.
�(�, �, �) = �
corresponding set and variable.
application.
rule parameters, i.e.
Table 3. Phenomena associated with the rule operator of every implication in inference
Fig. 10. Rule 18 parameter distribution for 43 experiments
Parametric Type-2 Fuzzy Logic Systems 111
Suppose a PT2FLC where it is only modified the parameter value of rule 18 and a set of parameters that can be spread randomly around the mean of its value ��(18) = 0.7. For this experiment, it was performed 43 iterations in order to show how the variation of ��(18)
Table 4 shows some results about the transient when trying to reach a tilt = 0.125 rads. Other phenomena can be analyzed for all 43 iterations. Also, in Fig.12 it can be seen that overshoot is attenuated drastically when ��(18) → 1, if it is only modified this rule. Time response (rise time, peak time and settling time) is also compromised due to parametric conjunctions. It can be seen also that drastic attenuation of overshoot occurs for ��(18) ≲ 0.7. Greater values do not affect it meaningfully. As it can be seen in (12), rule parameter proposed as the optimal for rule 18 is near to 1, which might be different with other configurations. This is because of the influence of the rest of rule parameters. However, this optimal configuration does not compromise the response time but it does eliminate the
affects the overshoot attenuation and also other phenomena (Fig. 10-11).
Fig. 13. Final approximation of T2FLC modifying rule parameters
Fig. 14. Comparison of response between T1FLC and parametric T2FLC when reference is a
overshoot completely.
noisy sine signal
Fig. 11. Transient response for several values of parameter of rule 18
Table 4. Transient characteristics for parameter variation of rule 18
Fig. 12. Histograms for transient measures (overshoot, rise time, peak time and settling time) for rule parameter 18
Also, it was found that every rule parameter has a full, medium or null influence with final response. Table 3 shows the analysis made with every implication. For example, with rule 18 it can be diminished the overshoot when PT2FLC is just trying to control the system to reach a specific tilt of plate.
110 Fuzzy Logic – Algorithms, Techniques and Implementations
**Transient Values Min Max** � �� Overshoot (rads) 0.0038 0.0529 0.0086 6.5267e-05 Delay Time (s) 0.95 0.95 0.95 0 Rise Time (s) 1.05 1.2 1.1465 8.5992e-04 Peak Time (s) 2.05 3.95 3.2162 0.1249 Settling Time (s) 2.35 3.95 3.2465 0.0774
Fig. 12. Histograms for transient measures (overshoot, rise time, peak time and settling time)
Also, it was found that every rule parameter has a full, medium or null influence with final response. Table 3 shows the analysis made with every implication. For example, with rule 18 it can be diminished the overshoot when PT2FLC is just trying to control the system to
Fig. 11. Transient response for several values of parameter of rule 18
Table 4. Transient characteristics for parameter variation of rule 18
for rule parameter 18
reach a specific tilt of plate.
Suppose a PT2FLC where it is only modified the parameter value of rule 18 and a set of parameters that can be spread randomly around the mean of its value ��(18) = 0.7. For this experiment, it was performed 43 iterations in order to show how the variation of ��(18) affects the overshoot attenuation and also other phenomena (Fig. 10-11).
Table 4 shows some results about the transient when trying to reach a tilt = 0.125 rads. Other phenomena can be analyzed for all 43 iterations. Also, in Fig.12 it can be seen that overshoot is attenuated drastically when ��(18) → 1, if it is only modified this rule. Time response (rise time, peak time and settling time) is also compromised due to parametric conjunctions. It can be seen also that drastic attenuation of overshoot occurs for ��(18) ≲ 0.7. Greater values do not affect it meaningfully. As it can be seen in (12), rule parameter proposed as the optimal for rule 18 is near to 1, which might be different with other configurations. This is because of the influence of the rest of rule parameters. However, this optimal configuration does not compromise the response time but it does eliminate the overshoot completely.
Fig. 13. Final approximation of T2FLC modifying rule parameters
Fig. 14. Comparison of response between T1FLC and parametric T2FLC when reference is a noisy sine signal
Parametric Type-2 Fuzzy Logic Systems 113
It is introduced a PT2FLC suitable for control system implementation using a new set of
Some phenomena are present when trying to tune a fuzzy system. Original B&P T1FLC was tuned to obtain the best results as in (Moreno-Armendariz, Rubio-Espino et al. 2010). When it was implemented a B&P PT2FLC with same set distribution in input and output with same rule set, as its counterpart, it was found that some phenomenon appears again. Final system response is related with all their variables, like set distribution, FOU width or conjunction parameters and they all have an implicit phenomenon which might be controlled, depending on the characteristics of the plant and the proposed rule set for a
A parametric conjunction to perform the implication can be applied to any fuzzy system, no matter if it is type1 or type 2. The usage of parametric conjunctions in inference help to weight the influence of premises and therefore it can be forced to obtain a certain crisp value desired. Finally it was obtained an optimal result when trying to control the B&P system,
When the PT2FLC is subjected to external perturbations, i.e. an extra level of uncertainty is aggregated to the system; the PT2FLC exhibits a better response over its T1 counterpart. Therefore, uncertain variations in inputs of a general FLC require sets with an appropriated
Therefore, the usage of PT2FLS for control purposes gives additional options for improving control precision and the usage of Monotone Sum of Conjunctions gives an opportunity to
Future research needs to examine the use of other parametric classes of conjunctions using simple functions. Moreover, this work can be extended using optimization techniques for calculating both better rule parameter selection and other parameters like set distribution and rule set. A hardware implementation is convenient in order to validate its behavior in
This work was supported by the Instituto de Ciencia y Tecnologia del Distrito Federal (ICyTDF) under project number PICCT08-22. We also thank the support of the Secretaria de Investigacion y Posgrado of Instituto Politecnico Nacional (SIP-IPN) under project number SIP-20113813 and project number SIP-20113709, COFFA-IPN and PIFI-IPN. Any opinions, findings, conclusions or recommendations expressed in this publication are those of the
Batyrshin, I. Z. and O. Kaynak (1999). "Parametric Classes of Generalized Conjunction and
Disjunction Operations for Fuzzy Modeling." IEEE Transactions on Fuzzy Systems
authors and do not necessarily reflect the views of the sponsoring agency.
parametric conjunction called (p) −monotone sum of conjunctions.
reaching the reference without overshoot, SSE nor ripple in 2.65 seconds.
**6. Conclusion**
particular solution.
real time applications.
**8. References**
**7**(5): 586-596.
**7. Acknowledgements**
FOU that can capture and support them.
implement PT2FLC in hardware for real time applications.
Once it has been chosen the right parameter values of every rule it is possible to see that the influence of premises over a consequent may be regulated using a parametric conjunction. Then, overshoot and ripple have been completely removed and time response has been improved also as it can be seen in Fig. 13.
Finally Fig. 14 depicts this response of T1FLC and PT2FLC using the optimal set and rule parameters when reference cannot be determined in presence of noise. In this last experiment, signal to follow is a noisy sine signal with noise frequency equal to 500 Hz (applied to a single axis of plate). PT2FLC follows this shape very similar to T1FLC. It can be seen that PT2FLC filters all drastic changes of this noisy signal unlike T1FLC.
### **5. Discussion**
Some of encountered problems and solutions are listed below.
### **5.1 Overshoot**
The best results were obtained when it was reduced the FOU of every set, but reducing their FOU to zero converts the T2FLC into a T1FLC, so, this system could not deal with the uncertainties that could exist in feedback of control system (e.g. noise in sensor or noise due to illumination of room). The use of parametric conjunction operators instead the common tnorm operators, e.g. min, is the best solution to reduce the reminding overshoot after considering to modify the FOU of the sets. Due to overshoot is present when the ball is nearby the reference, inertia pulls the ball over the reference and no suitable control action could be applied. In order to smooth this action it is possible to decrease its effect diminishing the influence of premises using a parametric conjunction. A suitable value of parameter � of certain rule let drop that excessive control action, and therefore decrease the overshoot. Parameters of rules 8 and 18 have the major influence on overshoot.
### **5.2 Steady-State Error**
There is not a precise solution to decrease the SSE. But expert can play with FOU widths of variables. For example, reducing the SSE having a big FOU in sets of variable error and decreasing all FOUs of variable change is a good option to reduce all SSE. Also it is possible to reduce it modifying the centroids of output variable tilt. Unfortunately those actions could generate additional nonlinearities so an expert must evaluate this situation.
### **5.3 Ripple**
Ripple can be controlled considering the FOU width of the variable error. Having a big FOU in sets of variable change can help to reduce the ripple.
### **5.4 Response time**
A simpler approximation is possible considering the values of parameters of rules 8 and 18. If �� = 1 then all reminding ripple is cleared and if ��� = 1 then almost all overshoot is eliminated, but time response is increased. Hence, if the expert has not any timing constraints then the usage of those rule parameters might help to reduce the undesired phenomenon considering this compromise.
### **6. Conclusion**
112 Fuzzy Logic – Algorithms, Techniques and Implementations
Once it has been chosen the right parameter values of every rule it is possible to see that the influence of premises over a consequent may be regulated using a parametric conjunction. Then, overshoot and ripple have been completely removed and time response has been
Finally Fig. 14 depicts this response of T1FLC and PT2FLC using the optimal set and rule parameters when reference cannot be determined in presence of noise. In this last experiment, signal to follow is a noisy sine signal with noise frequency equal to 500 Hz (applied to a single axis of plate). PT2FLC follows this shape very similar to T1FLC. It can be
The best results were obtained when it was reduced the FOU of every set, but reducing their FOU to zero converts the T2FLC into a T1FLC, so, this system could not deal with the uncertainties that could exist in feedback of control system (e.g. noise in sensor or noise due to illumination of room). The use of parametric conjunction operators instead the common tnorm operators, e.g. min, is the best solution to reduce the reminding overshoot after considering to modify the FOU of the sets. Due to overshoot is present when the ball is nearby the reference, inertia pulls the ball over the reference and no suitable control action could be applied. In order to smooth this action it is possible to decrease its effect diminishing the influence of premises using a parametric conjunction. A suitable value of parameter � of certain rule let drop that excessive control action, and therefore decrease the
There is not a precise solution to decrease the SSE. But expert can play with FOU widths of variables. For example, reducing the SSE having a big FOU in sets of variable error and decreasing all FOUs of variable change is a good option to reduce all SSE. Also it is possible to reduce it modifying the centroids of output variable tilt. Unfortunately those actions
Ripple can be controlled considering the FOU width of the variable error. Having a big FOU
A simpler approximation is possible considering the values of parameters of rules 8 and 18. If �� = 1 then all reminding ripple is cleared and if ��� = 1 then almost all overshoot is eliminated, but time response is increased. Hence, if the expert has not any timing constraints then the usage of those rule parameters might help to reduce the undesired
seen that PT2FLC filters all drastic changes of this noisy signal unlike T1FLC.
overshoot. Parameters of rules 8 and 18 have the major influence on overshoot.
could generate additional nonlinearities so an expert must evaluate this situation.
in sets of variable change can help to reduce the ripple.
phenomenon considering this compromise.
Some of encountered problems and solutions are listed below.
improved also as it can be seen in Fig. 13.
**5. Discussion**
**5.1 Overshoot**
**5.2 Steady-State Error**
**5.3 Ripple**
**5.4 Response time**
It is introduced a PT2FLC suitable for control system implementation using a new set of parametric conjunction called (p) −monotone sum of conjunctions.
Some phenomena are present when trying to tune a fuzzy system. Original B&P T1FLC was tuned to obtain the best results as in (Moreno-Armendariz, Rubio-Espino et al. 2010). When it was implemented a B&P PT2FLC with same set distribution in input and output with same rule set, as its counterpart, it was found that some phenomenon appears again. Final system response is related with all their variables, like set distribution, FOU width or conjunction parameters and they all have an implicit phenomenon which might be controlled, depending on the characteristics of the plant and the proposed rule set for a particular solution.
A parametric conjunction to perform the implication can be applied to any fuzzy system, no matter if it is type1 or type 2. The usage of parametric conjunctions in inference help to weight the influence of premises and therefore it can be forced to obtain a certain crisp value desired. Finally it was obtained an optimal result when trying to control the B&P system, reaching the reference without overshoot, SSE nor ripple in 2.65 seconds.
When the PT2FLC is subjected to external perturbations, i.e. an extra level of uncertainty is aggregated to the system; the PT2FLC exhibits a better response over its T1 counterpart. Therefore, uncertain variations in inputs of a general FLC require sets with an appropriated FOU that can capture and support them.
Therefore, the usage of PT2FLS for control purposes gives additional options for improving control precision and the usage of Monotone Sum of Conjunctions gives an opportunity to implement PT2FLC in hardware for real time applications.
Future research needs to examine the use of other parametric classes of conjunctions using simple functions. Moreover, this work can be extended using optimization techniques for calculating both better rule parameter selection and other parameters like set distribution and rule set. A hardware implementation is convenient in order to validate its behavior in real time applications.
### **7. Acknowledgements**
This work was supported by the Instituto de Ciencia y Tecnologia del Distrito Federal (ICyTDF) under project number PICCT08-22. We also thank the support of the Secretaria de Investigacion y Posgrado of Instituto Politecnico Nacional (SIP-IPN) under project number SIP-20113813 and project number SIP-20113709, COFFA-IPN and PIFI-IPN. Any opinions, findings, conclusions or recommendations expressed in this publication are those of the authors and do not necessarily reflect the views of the sponsoring agency.
### **8. References**
Batyrshin, I. Z. and O. Kaynak (1999). "Parametric Classes of Generalized Conjunction and Disjunction Operations for Fuzzy Modeling." IEEE Transactions on Fuzzy Systems **7**(5): 586-596.
**6**
**Application of Adaptive Neuro Fuzzy**
Each construction project has unique features that differentiate it from even resembling projects. Construction techniques, design, contract types, liabilities, weather, soil conditions, politic-economic environment and many other aspects may be different for every new commitment. Uncertainty is a reality of construction business. Leung et al. (2007) developed a model to deal with uncertain demand by considering a multi-site production planning problem. The inventory control problem and quantify the value of advanced demand information were examined (Ozer and Wei, 2004). Mula et al. (2010) proposed mathematical programming models to address supply chain production and transport planning problems. A model for making multi-criteria decision was developed for both the manufacturers and the distributors Dong et al. (2005). A stochastic planning model was constructed for a twoechelon supply chain of a petroleum company Al-Othman et al. (2008). Weng and McClurg (2003) and Ray et al. (2005) focused supply uncertainty along with demand uncertainty in supply chains. Bollapragada et al. (2004) examined uncertain lead time for random demand
A number of methods were developed by researches to solve problems associated with uncertainties, including scenario programming (Wullink et al., 2004; Chang et al., 2007), stochastic programming (Popescu, 2007; Santoso et al., 2005), fuzzy approach (Petrovic et al., 1999; Schultmann et al., 2006; Liang, 2008), and computer simulation and intelligent algorithms (Kalyanmoy, 2001; Coello, 2005). However, each method is suitable for particular situations. The decision makers have to select the appropriate method for solving a problem. For a uncertain construction project, the fuzzy atmosphere has been represented with the terms 'uncertainty' or 'risk' by construction managers and researchers, and they tried to control this systematically through risk management and analysis methods since the early 1990s (Edwards L., 2004). Some researchers like Flanagan et al. Flanagan R, Norman G. (1993) and Pilcher R. (1985) put differentiation between these two terms. They have mentioned that uncertainty represents the situations in which there is no historical data; and risk, in contrast, can be used for situations where success or failure is determined in probabilistic quantities by benefiting from the previous data available. Since such a
**1. Introduction**
and supply capacity in assembly Systems.
**Inference System in Supply Chain**
*Prince of Songkla Universit, Kohong Hatyai, Songkhla,*
**Management Evaluation**
Thoedtida Thipparat
*Thailand*
*Faculty of Management Sciences,*
## **Application of Adaptive Neuro Fuzzy Inference System in Supply Chain Management Evaluation**
Thoedtida Thipparat *Faculty of Management Sciences, Prince of Songkla Universit, Kohong Hatyai, Songkhla, Thailand*
### **1. Introduction**
114 Fuzzy Logic – Algorithms, Techniques and Implementations
Batyrshin, I. Z., I. J. Rudas, et al. (2009). "On Generation of Digital Fuzzy Parametric
Karnik, N. N., J. M. Mendel, et al. (1999). "Type-2 Fuzzy Logic Systems." IEEE Transactions
Moreno-Armendariz, M. A., E. Rubio-Espino, et al. (2010). Design and Implementation of a
Rudas, I. J., I. Z. Batyrshin, et al. (2009). Digital Fuzzy Parametric Conjunctions for
Wu, H. and J. M. Mendel (2002). "Uncertainty Bounds and their Use in the Design of Interval Type-2 Fuzzy Logic Systems." IEEE Transactions on Fuzzy Systems **10**(5): 622-639.
Conference on Reconfigurable Computing and FPGAs, Cancun, Mexico. Prometeo Cortes, Ildar Z. Batyrshin, et al. (2010). FPGA Implementation of (p)-Monotone
Sum of Basic t-norms. International Conference on Fuzzy Systems.
Computational Cybernetics, Budapest, Hungary.
Visual Fuzzy Control in FPGA for the Ball and Plate System. IEEE International
Hardware Implementation of Fuzzy Systems. IEEE International Conference on
Conjunctions." Studies in Computational Intelligence **243**: 79-89. E. P. Klement, R. Mesiar, et al. (2000). Triangular norms. Dordrecht, Kluwer.
on Fuzzy Systems **7**(6): 643-658
Each construction project has unique features that differentiate it from even resembling projects. Construction techniques, design, contract types, liabilities, weather, soil conditions, politic-economic environment and many other aspects may be different for every new commitment. Uncertainty is a reality of construction business. Leung et al. (2007) developed a model to deal with uncertain demand by considering a multi-site production planning problem. The inventory control problem and quantify the value of advanced demand information were examined (Ozer and Wei, 2004). Mula et al. (2010) proposed mathematical programming models to address supply chain production and transport planning problems. A model for making multi-criteria decision was developed for both the manufacturers and the distributors Dong et al. (2005). A stochastic planning model was constructed for a twoechelon supply chain of a petroleum company Al-Othman et al. (2008). Weng and McClurg (2003) and Ray et al. (2005) focused supply uncertainty along with demand uncertainty in supply chains. Bollapragada et al. (2004) examined uncertain lead time for random demand and supply capacity in assembly Systems.
A number of methods were developed by researches to solve problems associated with uncertainties, including scenario programming (Wullink et al., 2004; Chang et al., 2007), stochastic programming (Popescu, 2007; Santoso et al., 2005), fuzzy approach (Petrovic et al., 1999; Schultmann et al., 2006; Liang, 2008), and computer simulation and intelligent algorithms (Kalyanmoy, 2001; Coello, 2005). However, each method is suitable for particular situations. The decision makers have to select the appropriate method for solving a problem.
For a uncertain construction project, the fuzzy atmosphere has been represented with the terms 'uncertainty' or 'risk' by construction managers and researchers, and they tried to control this systematically through risk management and analysis methods since the early 1990s (Edwards L., 2004). Some researchers like Flanagan et al. Flanagan R, Norman G. (1993) and Pilcher R. (1985) put differentiation between these two terms. They have mentioned that uncertainty represents the situations in which there is no historical data; and risk, in contrast, can be used for situations where success or failure is determined in probabilistic quantities by benefiting from the previous data available. Since such a
Application of Adaptive Neuro Fuzzy Inference System in Supply Chain Management Evaluation 117
Section 2 reviews the literature on construction supply chain, supply chain performance evaluation and Agile Supply Chain (ASC); Section 3 represents the conceptual model using the capabilities of construction supply chain such as reliability, flexibility, responsiveness, cost, and asset, Section four contains an adaptive neuro fuzzy inference system (ANFIS) model which is proposed to evaluate flexibility in construction supply chains and the applicability of the proposed model has been tested by using construction companies in
Considering the construction industry, the client represents a unique customer with unique requirements. Stakeholders in the supply chain will provide these requirements. They must have the required primary competencies to make possible the fulfilment of these
In reality, organisations within a supply network delivering an office development will differ from those required to deliver a residential project. It may be useful to consider the chain as a network of organisations or a network organisations operating within the same market or industry to satisfy a variety of clients. Stakeholders involved in the construction supply-chain were classified into five categories related to the construction stages (H. Ismail & Sharif., 2005). The contract is the predominant approach for managing the relationship between organisations that operate in a construction project to deliver the client's required project. Although contracts are a sufficient basis for the delivery of a completed project, they are not sufficient to deliver a construction efficiently, at minimum
The definition of flexibility is still fuzzy, mainly because it largely deals with things already being addressed by industry and which are covered by existing research projects and programs. Many researchers provide conceptual over views, different reference and mature models of flexibility. For instance, Siemieniuch and Sinclair (2000) presented that to become a truly agile supply chain key enablers are classified into four categories: Collaborative relationship as the supply chain strategy, Process integration as the foundation of supply chain, Information integration as the infrastructure of supply chain and Customer /marketing sensitivity as the mechanism of supply chain. The aggregation of current approaches can be criticized as they haven't considered the impact of enablers in assessing supply chain flexibility and also the scale used to aggregate the flexibility capabilities has
Several papers present application of theories of measurement systems for managing performance of supply chain. However, there is no measurement system for managing performance of the entire supply chain. The adoption of metrics that cross of the borders of organization considering dimensions of performance related to inter and intra organization
Thailand. Finally, in section 5 the main conclusion of this study is discussed.
**2. Construction supply chain**
**2.1 Construction supply chain**
cost, and right first time'.
the limitations.
**2.2 Flexibility supply chain**
requirements.
separation is regarded as meaningless in the construction literature, risk turns out to be the most consistent term to be used for construction projects because some probability values can be attached intuitively and judgmentally to even the most uncertain events (Flanagan R, Norman G., 1993). The uncertainty represented quantitatively at some level is not the uncertainty any more; rather it is the risk henceforth and needs to be managed.
Construction companies are trying to make their supply chain more effective, and more efficient. Supply chain management has the potential to make construction projects less fragmented, improve project quality, reduce project duration, and hence reduce total project cost, while creating more satisfied customers. Construction companies need to respond of uncertain environment by using the concept of flexibility. Construction companies have recognized that flexibility is crucial for their survival and competitiveness. Several definitions of flexibility have been proposed since the construct is still in its initial stage of application to organizational phenomenon. Flexibility is defined as "the agility of a supply chain to respond to market changes in demand in order to gain or maintain its competitive advantage" (Bolstorff, P., Rosenbaum, R., 2007). The combination of Supply Chain Management (SCM) and flexibility is a significant source of competitiveness which has come to be named Agile Supply Chain (ASC). This paper argues that it is important to establish the flexibility of the construction supply chain. After embracing ASC an important question must be asked: How construction companies can evaluate flexibility in supply chains? This evaluation is essential for construction managers as it assists in achieving flexibility effectively by performing gap analysis between existent flexibility level and the desired one and also provides more informative and reliable information for decision making. Therefore, this study attempts to answer this question with a particular focus on measuring flexibility.
An approach based on Adaptive Neuro Fuzzy Inference System (ANFIS) for measurement of agility in Supply Chain was developed (Seyedhoseini, S.M., et al., 2010). The researchers used ANFIS to deal with complexity and vagueness of agility in global markets. ANFIS was applied order to inject different and complicated agility capabilities (that is, flexibility, competency, cost, responsiveness and quickness) to the model in an ambiguous environment. In addition, this study developed different potential attributes of ANFIS. Membership functions for each agility capabilities were constructed. The collected data was trained by using the functions through an adaptive procedure, using fuzzy concepts in order to model objective attributes. The proposed approach was useful for surveying real life problems. The proposed procedure had efficiently been applied to a large scale automobile manufacturing company in Iran. Statistical analysis illustrated that there were no meaningful difference between experts' opinion and our proposed procedure for supply chain agility measurement.
A procedure with aforementioned functionality must be develop to cope with uncertain environment of construction projects and lack of efficient measuring tool for flexibility of supply chain system. This study is to apply fuzzy concepts and aggregate this powerful tool with Artificial Neural Network concepts in favor of gaining ANFIS to handle the imprecise nature of attributes for associated concepts of flexibility. ANFIS is considered as an efficient tool for development and surveying of the novel procedure. Due to our best knowledge this combination has never been reported in literature before. This paper is organized as follows. Section 2 reviews the literature on construction supply chain, supply chain performance evaluation and Agile Supply Chain (ASC); Section 3 represents the conceptual model using the capabilities of construction supply chain such as reliability, flexibility, responsiveness, cost, and asset, Section four contains an adaptive neuro fuzzy inference system (ANFIS) model which is proposed to evaluate flexibility in construction supply chains and the applicability of the proposed model has been tested by using construction companies in Thailand. Finally, in section 5 the main conclusion of this study is discussed.
### **2. Construction supply chain**
116 Fuzzy Logic – Algorithms, Techniques and Implementations
separation is regarded as meaningless in the construction literature, risk turns out to be the most consistent term to be used for construction projects because some probability values can be attached intuitively and judgmentally to even the most uncertain events (Flanagan R, Norman G., 1993). The uncertainty represented quantitatively at some level is not the
Construction companies are trying to make their supply chain more effective, and more efficient. Supply chain management has the potential to make construction projects less fragmented, improve project quality, reduce project duration, and hence reduce total project cost, while creating more satisfied customers. Construction companies need to respond of uncertain environment by using the concept of flexibility. Construction companies have recognized that flexibility is crucial for their survival and competitiveness. Several definitions of flexibility have been proposed since the construct is still in its initial stage of application to organizational phenomenon. Flexibility is defined as "the agility of a supply chain to respond to market changes in demand in order to gain or maintain its competitive advantage" (Bolstorff, P., Rosenbaum, R., 2007). The combination of Supply Chain Management (SCM) and flexibility is a significant source of competitiveness which has come to be named Agile Supply Chain (ASC). This paper argues that it is important to establish the flexibility of the construction supply chain. After embracing ASC an important question must be asked: How construction companies can evaluate flexibility in supply chains? This evaluation is essential for construction managers as it assists in achieving flexibility effectively by performing gap analysis between existent flexibility level and the desired one and also provides more informative and reliable information for decision making. Therefore, this study attempts to answer this question with a particular focus on measuring
An approach based on Adaptive Neuro Fuzzy Inference System (ANFIS) for measurement of agility in Supply Chain was developed (Seyedhoseini, S.M., et al., 2010). The researchers used ANFIS to deal with complexity and vagueness of agility in global markets. ANFIS was applied order to inject different and complicated agility capabilities (that is, flexibility, competency, cost, responsiveness and quickness) to the model in an ambiguous environment. In addition, this study developed different potential attributes of ANFIS. Membership functions for each agility capabilities were constructed. The collected data was trained by using the functions through an adaptive procedure, using fuzzy concepts in order to model objective attributes. The proposed approach was useful for surveying real life problems. The proposed procedure had efficiently been applied to a large scale automobile manufacturing company in Iran. Statistical analysis illustrated that there were no meaningful difference between experts' opinion and our proposed procedure for supply
A procedure with aforementioned functionality must be develop to cope with uncertain environment of construction projects and lack of efficient measuring tool for flexibility of supply chain system. This study is to apply fuzzy concepts and aggregate this powerful tool with Artificial Neural Network concepts in favor of gaining ANFIS to handle the imprecise nature of attributes for associated concepts of flexibility. ANFIS is considered as an efficient tool for development and surveying of the novel procedure. Due to our best knowledge this combination has never been reported in literature before. This paper is organized as follows.
uncertainty any more; rather it is the risk henceforth and needs to be managed.
flexibility.
chain agility measurement.
Considering the construction industry, the client represents a unique customer with unique requirements. Stakeholders in the supply chain will provide these requirements. They must have the required primary competencies to make possible the fulfilment of these requirements.
### **2.1 Construction supply chain**
In reality, organisations within a supply network delivering an office development will differ from those required to deliver a residential project. It may be useful to consider the chain as a network of organisations or a network organisations operating within the same market or industry to satisfy a variety of clients. Stakeholders involved in the construction supply-chain were classified into five categories related to the construction stages (H. Ismail & Sharif., 2005). The contract is the predominant approach for managing the relationship between organisations that operate in a construction project to deliver the client's required project. Although contracts are a sufficient basis for the delivery of a completed project, they are not sufficient to deliver a construction efficiently, at minimum cost, and right first time'.
### **2.2 Flexibility supply chain**
The definition of flexibility is still fuzzy, mainly because it largely deals with things already being addressed by industry and which are covered by existing research projects and programs. Many researchers provide conceptual over views, different reference and mature models of flexibility. For instance, Siemieniuch and Sinclair (2000) presented that to become a truly agile supply chain key enablers are classified into four categories: Collaborative relationship as the supply chain strategy, Process integration as the foundation of supply chain, Information integration as the infrastructure of supply chain and Customer /marketing sensitivity as the mechanism of supply chain. The aggregation of current approaches can be criticized as they haven't considered the impact of enablers in assessing supply chain flexibility and also the scale used to aggregate the flexibility capabilities has the limitations.
Several papers present application of theories of measurement systems for managing performance of supply chain. However, there is no measurement system for managing performance of the entire supply chain. The adoption of metrics that cross of the borders of organization considering dimensions of performance related to inter and intra organization
Application of Adaptive Neuro Fuzzy Inference System in Supply Chain Management Evaluation 119
Delivery cycle
The neuro-fuzzy system attempts to model the uncertainty in the factor assessments, accounting for their qualitative nature. A combination of classic stochastic simulations and fuzzy logic operations on the ANN inputs as a supplement to artificial neural network is employed. Artificial Neural Networks (ANN) has the capability of self-learning, while fuzzy logic inference system (FLIS) is capable of dealing with fuzzy language information and simulating judgment and decision making of the human brain. It is currently the research focus to combine ANN with FLIS to produce fuzzy network system. ANFIS is an example of such a readily available system, which uses ANN to accomplish fuzzification, fuzzy inference and defuzzification of a fuzzy system. ANFIS utilizes ANN's learning mechanisms to draw rules from input and output data pairs. The system possesses not only the function of adaptive learning but also the function of fuzzy information describing and processing, and judgment and decision making. ANFIS is different from ANN in that ANN uses the connection weights to describe a system while ANFIS uses fuzzy language rules from fuzzy
The ANFIS approach adopts Gaussian functions (or other membership functions) for fuzzy sets, linear functions for the rule outputs, and Sugeno's inference mechanism (R.E. Spekman, J.W. Kamau! Jr., N. Myhr., 1998). The parameters of the network are the mean and standard deviation of the membership functions (antecedent parameters) and the coefficients of the output linear functions as well (consequent parameters). The ANFIS learning algorithm is then used to obtain these parameters. This learning algorithm is a hybrid algorithm consisting of the gradient descent and the least-squares estimate. Using this hybrid algorithm, the rule parameters are recursively updated until an acceptable level of error is reached. Each iteration includes two passes, forward and backward. In the forward pass, the antecedent parameters are fixed and the consequent parameters are obtained using the linear least-squares estimation. In the backward pass, the consequent parameters are fixed and the error signals propagate backward as well as the antecedent
time
Order fulfillment Total cost supply chain
Source cycle time Finance and
Make cycle time Inventory
management
planning cost
carrying cost
chain
Order
IT cost for supply
acquisition cost
management cost
Cash to cash
Days sales outstanding
Days payable outstanding
Inventory days of supply
Return of asset
Asset turns Net profit
Reliability Flexibility Responsiveness Cost Asset
Upside flexibility supply chain
flexibility
Upside make flexibility
Upside delivery flexibility
Perfect condition Material
Perfect order fulfillment
day
day
Accurate documentation
Delivery to commit
Delivery to commit
Orders in full Upside source
Table 1. Input/Output indicators
inference to describe a system.
**4. Neurofuzzy model**
processes (Lapide, L., 2000). The metrics developed by the SCOR model (Supply-Chain Council (SCC), 2011) were proposed to analyze a supply chain form three perspectives: process, metrics and best practice. The connections between the inter-organizational processes in each company in a supply chain are created based on the SCOR framework. The common and standardized language among the company within a supply chain is developed in order to compare supply chain performance as a whole.
There are five performance attributes in top level SCOR metric, namely reliability, responsiveness, flexibility, cost and asset management efficiency (Bolstorff, P., Rosenbaum, R., 2007). Reliability is defined as the performance related to the delivery, i.e., whether the correct product (according to specifications) is delivered to the correct place, it the correct quantity, at the correct time, with the correct documentation and the right customer. The definition of responsiveness is the speed at which a supply chain provides the products to customers. Flexibility is the agility of a supply chain to respond to market changes in demand in order to gain or maintain its competitive advantage. All the costs related to the operation of supply chain are included in the cost attribute. The asset management efficiency is the efficiency of an organization in managing its resources to meet demand. The management of all the resources (i.e., fixed and working capital) is considered.
The first limitation of supply chain flexibility evaluation is that the techniques do not consider the ambiguity and multi possibility associated with mapping of individual judgment to a number. The second limitation is the subjective judgment, selection and preference of evaluators having a significant influence on these methods. Because of the fact that the qualitative and ambiguous attributes are linked to flexibility assessment, most measures are described subjectively using linguistic terms, and cannot be handled effectively using conventional assessment approaches. The fuzzy logic provides an effective means of handling problems involving imprecise and vague phenomena. Fuzzy concepts enable assessors to use linguistic terms to assess indicators in natural language expressions, and each linguistic term can be associated with a membership function. In addition, fuzzy logic has generally found significant applications in management decisions. This study applies a fuzzy inference system for mapping input space (tangible and intangible) to output space in order to assist construction companies in better achieving an flexibility supply chain. The proposed Fuzzy Inference System (FIS) has been based on the experiences of experts to evaluate flexibility of construction supply chains.
### **3. Methodology**
To evaluate flexibility of the construction supply chain two main steps are performed. At the first step, measurement criteria are identified. A conceptual model is developed based on literature review. Capabilities of supply chain are employed to define supply chain performance in three basic segments: sourcing, construction and delivery. In this study the conceptual model involves four attributes: reliability, flexibility, responsiveness, cost, and asset. Twenty seven sub-attributes are the basis of the conceptual model as shown in Table 1. At the Second step, the design of an ANFIS architecture is performed by constructing an input-output mapping based on both human knowledge in the form of fuzzy if-then rules with appropriate membership functions and stipulated input-output data based- for deriving performance in supply chains.
Table 1. Input/Output indicators
### **4. Neurofuzzy model**
118 Fuzzy Logic – Algorithms, Techniques and Implementations
processes (Lapide, L., 2000). The metrics developed by the SCOR model (Supply-Chain Council (SCC), 2011) were proposed to analyze a supply chain form three perspectives: process, metrics and best practice. The connections between the inter-organizational processes in each company in a supply chain are created based on the SCOR framework. The common and standardized language among the company within a supply chain is
There are five performance attributes in top level SCOR metric, namely reliability, responsiveness, flexibility, cost and asset management efficiency (Bolstorff, P., Rosenbaum, R., 2007). Reliability is defined as the performance related to the delivery, i.e., whether the correct product (according to specifications) is delivered to the correct place, it the correct quantity, at the correct time, with the correct documentation and the right customer. The definition of responsiveness is the speed at which a supply chain provides the products to customers. Flexibility is the agility of a supply chain to respond to market changes in demand in order to gain or maintain its competitive advantage. All the costs related to the operation of supply chain are included in the cost attribute. The asset management efficiency is the efficiency of an organization in managing its resources to meet demand. The
The first limitation of supply chain flexibility evaluation is that the techniques do not consider the ambiguity and multi possibility associated with mapping of individual judgment to a number. The second limitation is the subjective judgment, selection and preference of evaluators having a significant influence on these methods. Because of the fact that the qualitative and ambiguous attributes are linked to flexibility assessment, most measures are described subjectively using linguistic terms, and cannot be handled effectively using conventional assessment approaches. The fuzzy logic provides an effective means of handling problems involving imprecise and vague phenomena. Fuzzy concepts enable assessors to use linguistic terms to assess indicators in natural language expressions, and each linguistic term can be associated with a membership function. In addition, fuzzy logic has generally found significant applications in management decisions. This study applies a fuzzy inference system for mapping input space (tangible and intangible) to output space in order to assist construction companies in better achieving an flexibility supply chain. The proposed Fuzzy Inference System (FIS) has been based on the experiences
To evaluate flexibility of the construction supply chain two main steps are performed. At the first step, measurement criteria are identified. A conceptual model is developed based on literature review. Capabilities of supply chain are employed to define supply chain performance in three basic segments: sourcing, construction and delivery. In this study the conceptual model involves four attributes: reliability, flexibility, responsiveness, cost, and asset. Twenty seven sub-attributes are the basis of the conceptual model as shown in Table 1. At the Second step, the design of an ANFIS architecture is performed by constructing an input-output mapping based on both human knowledge in the form of fuzzy if-then rules with appropriate membership functions and stipulated input-output data based- for
developed in order to compare supply chain performance as a whole.
management of all the resources (i.e., fixed and working capital) is considered.
of experts to evaluate flexibility of construction supply chains.
**3. Methodology**
deriving performance in supply chains.
The neuro-fuzzy system attempts to model the uncertainty in the factor assessments, accounting for their qualitative nature. A combination of classic stochastic simulations and fuzzy logic operations on the ANN inputs as a supplement to artificial neural network is employed. Artificial Neural Networks (ANN) has the capability of self-learning, while fuzzy logic inference system (FLIS) is capable of dealing with fuzzy language information and simulating judgment and decision making of the human brain. It is currently the research focus to combine ANN with FLIS to produce fuzzy network system. ANFIS is an example of such a readily available system, which uses ANN to accomplish fuzzification, fuzzy inference and defuzzification of a fuzzy system. ANFIS utilizes ANN's learning mechanisms to draw rules from input and output data pairs. The system possesses not only the function of adaptive learning but also the function of fuzzy information describing and processing, and judgment and decision making. ANFIS is different from ANN in that ANN uses the connection weights to describe a system while ANFIS uses fuzzy language rules from fuzzy inference to describe a system.
The ANFIS approach adopts Gaussian functions (or other membership functions) for fuzzy sets, linear functions for the rule outputs, and Sugeno's inference mechanism (R.E. Spekman, J.W. Kamau! Jr., N. Myhr., 1998). The parameters of the network are the mean and standard deviation of the membership functions (antecedent parameters) and the coefficients of the output linear functions as well (consequent parameters). The ANFIS learning algorithm is then used to obtain these parameters. This learning algorithm is a hybrid algorithm consisting of the gradient descent and the least-squares estimate. Using this hybrid algorithm, the rule parameters are recursively updated until an acceptable level of error is reached. Each iteration includes two passes, forward and backward. In the forward pass, the antecedent parameters are fixed and the consequent parameters are obtained using the linear least-squares estimation. In the backward pass, the consequent parameters are fixed and the error signals propagate backward as well as the antecedent
Application of Adaptive Neuro Fuzzy Inference System in Supply Chain Management Evaluation 121
For proving the applicability of the model and illustration, the proposed model was applied in twenty-five of the construction companies in Thailand. The first step to apply the model was to construct the decision team. The stakeholders involved in the construction stage became the decision team including main contractor, domestic subcontractors, nominated subcontractors, project manager, material suppliers, plant/equipment suppliers, designers, financial institution, insurance agency, and regulatory bodies. For training the ANFIS, a questionnaire was designed including the identified criteria. The decision team was asked to give a score to them, based on their knowledge associated with the construction stage. A Matlab programme was generated and compiled. The pre-processed input/output matrix which contained all the necessary representative features, was used to train the fuzzy inference system. Fig 2 shows the structure of the ANFIS; a Sugeno fuzzy inference system was used in this investigation. Based on the collected data, 150 data sets were used to train the ANFIS and the rest (50) for checking and validation of the model. For rule generation, the subtractive clustering was employ where the range of influence, squash factor, acceptance ratio, and rejection ratio were set at 0.5, 1.25, 0.5 and 0.15, respectively during the process of subtractive clustering. The trained fuzzy inference system includes 20 rules (clusters) as present in Fig 3. Because by using subtractive clustering, input space was categorized into 20 clusters. Each input has 20 Gaussian curve built-in membership functions. During training in ANFIS, sets of processed data were used to conduct 260 cycles
By inserting ANFIS output to the system the flexibility level of the supply chain management can be derived. In addition, the trend of training error and checking error has been shown in Fig 4. The researcher continued the training process to 500 epochs because the trend of checking error started to increase afterward and over fitting occurred. The value of checking error by 500 epochs was 1.45 which is acceptable. Then the value of supply chain flexibility is derived by a trained ANFIS. The ANFIS output in Thai construction
of learning.
companies is calculated.
Fig. 2. Network of innovation performance by the ANFIS
parameters are updated by the gradient descent method. An ANFIS architecture is equivalent to a two-input first-order Sugeno fuzzy model with nine rules, where each input is assumed to have three associated membership functions (MFs) (Z.Zhang., D.Ding., L.Rao., and Z.Bi., 2006). Sub-attributes associated with reliability, flexibility, responsiveness, cost, and asset are used as input variables; simultaneously, construction supply chain performance is considered as output variables. These input variables were used in the measurement of the supply chain performance by (G.M.D. Ganga, L.C.R. Carpinetti., 2011). Fig 1 is an ANFIS architecture that is equivalent to a two-input first-order Sugeno fuzzy model with nine rules, where each input is assumed to have three associated membership functions (MFs) (J. Jassbi, S.M. Seyedhosseini, and N. Pilevari., 2010).
Fig. 1. The ANFIS architecture for two input variables
120 Fuzzy Logic – Algorithms, Techniques and Implementations
parameters are updated by the gradient descent method. An ANFIS architecture is equivalent to a two-input first-order Sugeno fuzzy model with nine rules, where each input is assumed to have three associated membership functions (MFs) (Z.Zhang., D.Ding., L.Rao., and Z.Bi., 2006). Sub-attributes associated with reliability, flexibility, responsiveness, cost, and asset are used as input variables; simultaneously, construction supply chain performance is considered as output variables. These input variables were used in the measurement of the supply chain performance by (G.M.D. Ganga, L.C.R. Carpinetti., 2011). Fig 1 is an ANFIS architecture that is equivalent to a two-input first-order Sugeno fuzzy model with nine rules, where each input is assumed to have three associated membership
functions (MFs) (J. Jassbi, S.M. Seyedhosseini, and N. Pilevari., 2010).
Fig. 1. The ANFIS architecture for two input variables
For proving the applicability of the model and illustration, the proposed model was applied in twenty-five of the construction companies in Thailand. The first step to apply the model was to construct the decision team. The stakeholders involved in the construction stage became the decision team including main contractor, domestic subcontractors, nominated subcontractors, project manager, material suppliers, plant/equipment suppliers, designers, financial institution, insurance agency, and regulatory bodies. For training the ANFIS, a questionnaire was designed including the identified criteria. The decision team was asked to give a score to them, based on their knowledge associated with the construction stage. A Matlab programme was generated and compiled. The pre-processed input/output matrix which contained all the necessary representative features, was used to train the fuzzy inference system. Fig 2 shows the structure of the ANFIS; a Sugeno fuzzy inference system was used in this investigation. Based on the collected data, 150 data sets were used to train the ANFIS and the rest (50) for checking and validation of the model. For rule generation, the subtractive clustering was employ where the range of influence, squash factor, acceptance ratio, and rejection ratio were set at 0.5, 1.25, 0.5 and 0.15, respectively during the process of subtractive clustering. The trained fuzzy inference system includes 20 rules (clusters) as present in Fig 3. Because by using subtractive clustering, input space was categorized into 20 clusters. Each input has 20 Gaussian curve built-in membership functions. During training in ANFIS, sets of processed data were used to conduct 260 cycles of learning.
By inserting ANFIS output to the system the flexibility level of the supply chain management can be derived. In addition, the trend of training error and checking error has been shown in Fig 4. The researcher continued the training process to 500 epochs because the trend of checking error started to increase afterward and over fitting occurred. The value of checking error by 500 epochs was 1.45 which is acceptable. Then the value of supply chain flexibility is derived by a trained ANFIS. The ANFIS output in Thai construction companies is calculated.
Fig. 2. Network of innovation performance by the ANFIS
Application of Adaptive Neuro Fuzzy Inference System in Supply Chain Management Evaluation 123
Fig. 5. Network of construction supply chain performance by the ANFIS
No. Criteria
Table 2. Input values for the trained ANFIS
The rate of sub-attributes associated with flexibility, responsiveness & quickness, competency and cost and the output of ANFIS have been shown in Table 2 and 3, respectively. The twenty-five scenarios were used to test the performance the proposed method. The results indicate that the output values obtained from ANFIS are closer to the values given by experts in most scenarios being tested. The average and standard deviation of the differences between the estimated and the output values obtained from expert produced by ANFIS are calculated to be 12.6% and 8.75% respectively. As far as ANFIS is concerned, its biggest advantage is that there is no need to know the concrete functional relationship between outputs and inputs. Any relationship, linear or nonlinear, can be learned and approximated by an ANFIS such as a five-layer with sufficient large number of neurons in the hidden layer. In the case that the functional relationship between outputs and inputs is not known or cannot be determined, ANFIS definitely outperforms regression, which requires the relationship between output and inputs be known or specified. Another remarkable advantage of ANFIS is its capability of modelling the data of multiple inputs and multiple outputs. ANFIS has no restriction on the number of output. The relationships can be learned simultaneously by an ANFIS with multiple inputs and multiple outputs.
Reliability Flexibility Responsiveness Cost Asset
1 23 54 31 31 20 2 20 65 65 28 26 3 19 75 75 30 23 : : : : : : 98 20 65 65 28 31 99 26 70 70 32 65 100 23 70 25 31 75
Fig. 3. Trained main ANFIS surface of supply chain performance
Fig. 4. Trend of errors of trained fuzzy system
Fig 5 depicts a three dimensional plot that represents the mapping from reliability (in1) and flexibility (in2) to supply chain performance (out1). As the reliability and flexibility increases, the predicted supply chain performance increases in a non-linear piecewise manner, this being largely due to non-linearity of the characteristic of the input vector matrix derived from the collected data. This assumes that the collected data are fully representative of the features of the data that the trained FIS is intended to model. However the data are inherently insufficient and training data cannot cover all the features of the data that should be presented to the trained model. The accuracy of the model, therefore, is affected under such circumstances.
122 Fuzzy Logic – Algorithms, Techniques and Implementations
Fig 5 depicts a three dimensional plot that represents the mapping from reliability (in1) and flexibility (in2) to supply chain performance (out1). As the reliability and flexibility increases, the predicted supply chain performance increases in a non-linear piecewise manner, this being largely due to non-linearity of the characteristic of the input vector matrix derived from the collected data. This assumes that the collected data are fully representative of the features of the data that the trained FIS is intended to model. However the data are inherently insufficient and training data cannot cover all the features of the data that should be presented to the trained model. The accuracy of the model, therefore, is
Fig. 3. Trained main ANFIS surface of supply chain performance
Fig. 4. Trend of errors of trained fuzzy system
affected under such circumstances.
Fig. 5. Network of construction supply chain performance by the ANFIS
The rate of sub-attributes associated with flexibility, responsiveness & quickness, competency and cost and the output of ANFIS have been shown in Table 2 and 3, respectively. The twenty-five scenarios were used to test the performance the proposed method. The results indicate that the output values obtained from ANFIS are closer to the values given by experts in most scenarios being tested. The average and standard deviation of the differences between the estimated and the output values obtained from expert produced by ANFIS are calculated to be 12.6% and 8.75% respectively. As far as ANFIS is concerned, its biggest advantage is that there is no need to know the concrete functional relationship between outputs and inputs. Any relationship, linear or nonlinear, can be learned and approximated by an ANFIS such as a five-layer with sufficient large number of neurons in the hidden layer. In the case that the functional relationship between outputs and inputs is not known or cannot be determined, ANFIS definitely outperforms regression, which requires the relationship between output and inputs be known or specified. Another remarkable advantage of ANFIS is its capability of modelling the data of multiple inputs and multiple outputs. ANFIS has no restriction on the number of output. The relationships can be learned simultaneously by an ANFIS with multiple inputs and multiple outputs.
Table 2. Input values for the trained ANFIS
Application of Adaptive Neuro Fuzzy Inference System in Supply Chain Management Evaluation 125
Bollapragada, R., Rao, U.S., Zhang, J. (2004) Managing inventory and supply performance in
Bolstorff, P., Rosenbaum, R. (2007) *Supply chain excelence*. A handbook for Dramatic
Coello, C.A.C. (2005) An introduction to evolutionary algorithms and their applications.
Dong, J., Zhang, D., Yan, H., Nagurney, A., (2005) Multitiered supply chain networks:
Edwards L. (2004). *Practical risk management in the construction industry,* London: Thomas
Flanagan R, Norman G.(1993) *Risk management and construction*, Cambridge: Backwell
Ganga, G.M.D., Carpinetti, L.C.R. (2011) A fuzzy logic approach to supply chain
Ismail, H. and Sharif. (2005) Supply chain design for supply chain: A balanced approach to building agile supply chain, *Proceeding of the International conference on Flexibility*. Jassbi, J., Seyedhosseini, S.M. and Pilevari, N. (2010) An adaptive neuro fuzzy inference
Kalyanmoy, D. (2001) Multi-objective Optimization Using Evolutionary Algorithms. John
Lapide, L. (2000) What about measuring supply chain performance? *Achieving Supply Chain*
Leung, S.C.H., Tsang, S.O.S., Ng, W.L., Wu, Y., (2007) A robust optimization model for
Liang, T.F. (2008) Integrating production-transportation planning decision with fuzzy
Mula, J., Peidro, D., D ด az-Madro๑ero, M., Vicens, E., (2010) Mathematical programming
O zer, O., Wei, W., (2004) Inventory Control with Limited Capacity and Advance Demand
Petrovic, D., Roy, R., Petrovic, R. (1999) Supply chain modeling using fuzzy sets. International *Journal of Production Economics*, Vol. 59, No.1–3, pp.443–453. Popescu, I. (2007) Robust Mean-Covariance Solutions for Stochastic Optimization. *Operations*
Ray, S., Li, S., Song, Y. ( 2005) Tailored supply chain decision making under price sensitive
stochastic demand and delivery uncertainty. *Management Science*, Vol. 51, No.12,
system for supply chain flexibility evaluation, *International Journal of Industrial*
multi-site production planning problem in an uncertain environment, *European*
multiple goals in supply chains. *International Journal of Production Research*, Vol. 46,
models for supply chain production and transport planning, *European Journal of*
Improvement Using the SCOR Model. AMACOM, NY.
performance management, *Int. J. Production Economics*.
*Engineering & Production Research*, Vol. 20, pp.187-196.
*Journal of Operational Research,* Vol.181, No.1, pp. 224–238.
Information, *Operations Research,* Vol. 52, No.6, pp.988–1000.
*Excellence through Technology*, Vol. 2, pp. 287–297. Pilcher R. (1985) *Project cost control in construction*, London: Collins.
*Operational Research,* Vol.204, No.3, pp. 377–390.
*Research*, Vol. 55, No.1, pp.98–112.
Advanced Distributed Systems. Springer, Berlin, Heidelberg.
Vol.50, No.12, pp.1729–1743.
Vol.135, No.1, pp. 155–178.
Wiley and Sons, New York.
No.6, pp.1477–1494.
pp.1873–1891.
Telford.
Scientific.
assembly systems with random supply capacity and demand. *Management Science*
Multicriteria mecision-making under uncertainty. *Annals of Operations Research*,
Table 3. Possible value obtained form ANFIS method
### **5. Conclusion**
This paper has discussed the need for flexibility assessment of the construction supply chain. The particular features of construction supply chains highlighted. The need for and potential benefits of, construction supply chain flexibility assessment were then examined and the conceptual model of a flexibility assessment model for the supply chain presented. Case studies of the use of the model in assessing the construction organizations were also presented. The following conclusions can be drawn from the work presented in this paper: The way to improve the construction supply chain delivers projects is necessary to achieve client satisfaction, efficiency, effectiveness and profitability. It is important to perform the flexibility assessment of the construction supply chain in order to ensure that maximum benefit can be obtained. Since agile supply chain is considered as a dominant competitive advantage in recent years, evaluating supply chain flexibility can be useful and applicable for managers to make more informative and reliable decisions in anticipated changes of construction markets. The development of an appropriate flexibility assessment tool or model for the construction supply chain is necessary, as existing models are not appropriate in their present form. The results reveal that the ANFIS model improves flexibility assessment by using fuzzy rules to generate the adaptive neuro-fuzzy network, as well as a rotation method of training and testing data selection which is designed to enhance the reliability of the sampling process before constructing the training and testing model. The ANFIS model can explain the training procedure of outcome and how to simulate the rules for prediction. It can provide more accuracy on prediction.
Further research is necessary to compare efficiency of different models for measuring flexibility in supply chain. Although this study has been performed in the construction companies, the proposed methodology is applicable to other companies, e.g. consulting companies. Enablers in flexibility evaluation should be determined and the impact of them on capabilities must be studied in further researches. In addition, the relations between enablers should be considered in order to design a dynamic system for the supply chain management evaluation.
### **6. References**
Al-Othman, W.B.E., Lababidi, H.M.S., Alatiqi, I.M., Al-Shayji, K., (2008) Supply chain optimization of petroleum organization under uncertainty in market demands and prices*. European Journal of Operational Research*, Vol. 189, No.3, pp. 822–840.
124 Fuzzy Logic – Algorithms, Techniques and Implementations
1 11 72 15.4 2 20 75 15.08 3 19 75 19.67
98 18 66 10.17 99 26 30 28.46 100 28 55 8.77
This paper has discussed the need for flexibility assessment of the construction supply chain. The particular features of construction supply chains highlighted. The need for and potential benefits of, construction supply chain flexibility assessment were then examined and the conceptual model of a flexibility assessment model for the supply chain presented. Case studies of the use of the model in assessing the construction organizations were also presented. The following conclusions can be drawn from the work presented in this paper: The way to improve the construction supply chain delivers projects is necessary to achieve client satisfaction, efficiency, effectiveness and profitability. It is important to perform the flexibility assessment of the construction supply chain in order to ensure that maximum benefit can be obtained. Since agile supply chain is considered as a dominant competitive advantage in recent years, evaluating supply chain flexibility can be useful and applicable for managers to make more informative and reliable decisions in anticipated changes of construction markets. The development of an appropriate flexibility assessment tool or model for the construction supply chain is necessary, as existing models are not appropriate in their present form. The results reveal that the ANFIS model improves flexibility assessment by using fuzzy rules to generate the adaptive neuro-fuzzy network, as well as a rotation method of training and testing data selection which is designed to enhance the reliability of the sampling process before constructing the training and testing model. The ANFIS model can explain the training procedure of outcome and how to simulate the rules
Further research is necessary to compare efficiency of different models for measuring flexibility in supply chain. Although this study has been performed in the construction companies, the proposed methodology is applicable to other companies, e.g. consulting companies. Enablers in flexibility evaluation should be determined and the impact of them on capabilities must be studied in further researches. In addition, the relations between enablers should be considered in order to design a dynamic system for the supply chain
Al-Othman, W.B.E., Lababidi, H.M.S., Alatiqi, I.M., Al-Shayji, K., (2008) Supply chain
prices*. European Journal of Operational Research*, Vol. 189, No.3, pp. 822–840.
optimization of petroleum organization under uncertainty in market demands and
Expert's ANFIS Percent Difference
No Value of output
Table 3. Possible value obtained form ANFIS method
for prediction. It can provide more accuracy on prediction.
:
**5. Conclusion**
management evaluation.
**6. References**
**0**
**7**
*Mexico*
**in Wavelet Domain**
*National Polytechnic Institute of Mexico*
**Fuzzy Image Segmentation Algorithms**
The images are considered one of the most important means of information transmission; therefore the image processing has become an important tool in a variety of fields such as video coding, computer vision and medical imaging. Within the image processing, there is the segmentation process that involves partitioning an image into a set of homogeneous and meaningful regions, such that the pixels in each partitioned region possess an identical set of properties or attributes (Gonzalez & Woods, 1992). The sets of properties of the image may include gray levels, contrast, spectral values, or texture properties, etc. The result of segmentation is a number of homogeneous regions, each having a unique label. Image segmentation is often considered to be the most important task in computer vision. However, the segmentation in images is a challenging task due to several reasons: irregular and dispersive lesion borders, low contrast, artifacts in the image and variety of colors within the interest region. Therefore, numerous methods have been developed for image segmentation within applications in the computer vision. Image segmentation can be classified into three categories: A) *Supervised*.- These methods require the interactivity in which the pixels belonging to the same intensity range pointed out manually and segmented. B) *Automatic*.- This is also known as unsupervised methods, where the algorithms need some priori information, so these methods are more complex, and C) *Semi-automatic*.- That is the combination of manual and automatic segmentation. Some of practical applications of image segmentation are: the medical imaging tasks that consist of location of tumors and other pathologies, recognition of the objects in images of remote sensing obtained via satellite or aerial platforms, automated-recognition systems to inspect the electronic assemblies, biometrics, automatic traffic controlling systems, machine vision, separating and tracking the regions appearing in consequent frames of an sequence, and finally, the real time mobile robot
A lot of methods have been developed in the image segmentation. Let present brief
**1. Introduction**
applications employing vision systems. 1.
description of the several promising frameworks.
**2. Related work**
<sup>1</sup> (Gonzalez & Woods, 1992)
Heydy Castillejos and Volodymyr Ponomaryov
### **Fuzzy Image Segmentation Algorithms in Wavelet Domain**
Heydy Castillejos and Volodymyr Ponomaryov *National Polytechnic Institute of Mexico Mexico*
### **1. Introduction**
126 Fuzzy Logic – Algorithms, Techniques and Implementations
Santoso, T., Ahmed, S., Goetschalckx, M., Shapiro, A. (2005) A stochastic programming
Schultmann, F., Frohling, M., Rentz, O. (2006) Fuzzy approach for production planning and
Seyedhosseini, S.M., Jassbi, J., and Pilevari, N.(2010) Application of adaptive neuro fuzzy
Siemieniuch, C.E., and Sinclair, M.A. (2000) Implications of the supply chain for role
Spekman, R.E., Kamau Jr. J.W., Myhr, N. (1998) An empirical investigation into supply chain
Wullink, G., Gademann, A.J.R.M., Hans, E.W., Van Harten, A. (2004) Scenario-based
Zhang, Z., Ding, D., Rao, L., and Bi, Z. (2006) An ANFIS based approach for predicting the
*Distribution and Logistics Management*, Vol. 28, No. 8, pp. 630-650. Supply-Chain Council (SCC). (3 June 2011). Availableat: /http://www.supply-chain.org Weng, Z.K., McClurg, T., (2003) Coordinated ordering decisions for short life cycle products
*Operational Research*. Vol.167 , No.1, pp. 96–115.
*Ergonomics in Manufacturing*, Vol. 10, No.3, pp. 251-272.
*Production Research*, Vol. 42, No.24, pp.5079–5098.
*Research*, Vol. 44, No.8, pp. 1589–1612.
*Research*, No. 151, No.1, pp.12–24.
*Methods (GSP 153) ASCE*.
083-096.
approach for supply chain network design under uncertainty. *European Journal of*
detailed scheduling in paints manufacturing. *International Journal of Production*
inference system in measurement of supply chain agility: Real case study of a manufacturing company, *African Journal of Business Management*, Vol.4, No. 1, pp.
definitions in concurrent engineering, *International Journal of Human Factors and*
management\*A perspective on partnerships, *International Journal of Physical*
with uncertainty in delivery time and demand. *European Journal of Operational*
approach for flexible resource loading under uncertainty*. International Journal of*
ultimate bearing capacity of single piles, *Foundation Analysis and Design : Innovative*
The images are considered one of the most important means of information transmission; therefore the image processing has become an important tool in a variety of fields such as video coding, computer vision and medical imaging. Within the image processing, there is the segmentation process that involves partitioning an image into a set of homogeneous and meaningful regions, such that the pixels in each partitioned region possess an identical set of properties or attributes (Gonzalez & Woods, 1992). The sets of properties of the image may include gray levels, contrast, spectral values, or texture properties, etc. The result of segmentation is a number of homogeneous regions, each having a unique label. Image segmentation is often considered to be the most important task in computer vision. However, the segmentation in images is a challenging task due to several reasons: irregular and dispersive lesion borders, low contrast, artifacts in the image and variety of colors within the interest region. Therefore, numerous methods have been developed for image segmentation within applications in the computer vision. Image segmentation can be classified into three categories: A) *Supervised*.- These methods require the interactivity in which the pixels belonging to the same intensity range pointed out manually and segmented. B) *Automatic*.- This is also known as unsupervised methods, where the algorithms need some priori information, so these methods are more complex, and C) *Semi-automatic*.- That is the combination of manual and automatic segmentation. Some of practical applications of image segmentation are: the medical imaging tasks that consist of location of tumors and other pathologies, recognition of the objects in images of remote sensing obtained via satellite or aerial platforms, automated-recognition systems to inspect the electronic assemblies, biometrics, automatic traffic controlling systems, machine vision, separating and tracking the regions appearing in consequent frames of an sequence, and finally, the real time mobile robot applications employing vision systems. 1.
### **2. Related work**
A lot of methods have been developed in the image segmentation. Let present brief description of the several promising frameworks.
<sup>1</sup> (Gonzalez & Woods, 1992)
Ideally, the order in testing the region merging is when any test between two true regions occurs, which means that all tests inside each of the two true regions have previously occurred.
Fuzzy Image Segmentation Algorithms in Wavelet Domain 129
The most promising in segmentation of the images in general is the approach based on clustering. Cluster oriented-segmentation uses the multidimensional data to partition of the image pixels into clusters. Such kind of technique may be more appropriate than histogram-oriented ones in segmenting images, where each pixel has several attributes and is represented by a vector. Cluster analysis has attracted much attention since the 1960's and has been applied in many fields such as OCR (*Optical Character Recognition*) system. Below, let present three most successful frameworks based on this technique that we apply
K-Means algorithm is an unsupervised clustering algorithm that classifies the input data point into multiple classes based on their inherent distance from each other. The algorithm assumes that the data features from a vector space and tries to find natural clustering in them (Hartigan
2. Calculate new cluster membership. A feature vector *xj* is assigned to the cluster *Ci* if and
*mi* <sup>=</sup> <sup>1</sup>
In Fig.3, the segmentation process using the K-Means algorithm is exposed:
<sup>|</sup>*Ci*<sup>|</sup> <sup>∑</sup> *xj*∈*Ci*
4. If none of the cluster centroids has been changed, finish the algorithm. Otherwise, go to
*<sup>i</sup>* <sup>=</sup> *argmink*<sup>=</sup>1,,*K*�*xj* <sup>−</sup> *mk*�2. (2)
*xj*, (3)
& Wong, 1979). It works an iterative manner according to the following steps:
1. Choose initial centroids *m*1,..., *mk* of the clusters *C*1,..., *Ck*.
3. Recalculate the centroids for the clusters according to
where *xj* belong to data set *X* = *x*1, , *xi*, *xN*.
Fig. 3. Block diagram for K-Means algorithm.
**Clustering based segmentation**
in segmentation applications.
only if:
step 2.
**2.3 K-Means clustering algorithm**
### **2.1** *Adaptive thresholding (AT)*
In (Argenziano & Soyer, 1996), the automatic adaptive thresholding (AT) performs the image segmentation comparing the color of each a pixel with a threshold. The pixel is classified as a lesion if it is darker than the threshold, finally, presenting the output as a binary image. Morphological post-processing is then applied to fill the holes and to select the largest connected component in the binary image. For color images, an automatic selection of the color component based on the entropy of the color component *i* is used:
$$S(i) = -\sum\_{k=0}^{L-1} h\_i(k) \log[h\_i(k)],\tag{1}$$
where *hi*(*k*) is the histogram of the color component *i*. It is assumed that the image *Ii*(*x*, *y*) varies in the range 0, . . . , 255 and the histogram is computed using bins of length *L* = 25. The block diagram in Fig.1 explains in detail the operation for AT method.
Fig. 1. Block diagram of Adaptive thresholding.
### **2.2** *Statistical region merging*
In (M. Celebi, 2008), the authors use a variant of region growing and merging technique, called as statistical region merging (SRM). The authors propose the following strategy:
The SRM framework uses the image generation homogeneity property and performs as follows in Fig. 2:
Fig. 2. Block diagram of Statistical region merging.
Ideally, the order in testing the region merging is when any test between two true regions occurs, which means that all tests inside each of the two true regions have previously occurred.
### **Clustering based segmentation**
2 Will-be-set-by-IN-TECH
In (Argenziano & Soyer, 1996), the automatic adaptive thresholding (AT) performs the image segmentation comparing the color of each a pixel with a threshold. The pixel is classified as a lesion if it is darker than the threshold, finally, presenting the output as a binary image. Morphological post-processing is then applied to fill the holes and to select the largest connected component in the binary image. For color images, an automatic selection of the
> *L*−1 ∑ *k*=0
where *hi*(*k*) is the histogram of the color component *i*. It is assumed that the image *Ii*(*x*, *y*) varies in the range 0, . . . , 255 and the histogram is computed using bins of length *L* = 25. The
In (M. Celebi, 2008), the authors use a variant of region growing and merging technique, called
• Regions are defied as the sets of pixels with homogeneous properties that then are
• Region growing/merging techniques is used employing a statistical test to form the
The SRM framework uses the image generation homogeneity property and performs as
--
-
as statistical region merging (SRM). The authors propose the following strategy:
*hi*(*k*)*log*[*hi*(*k*)], (1)
color component based on the entropy of the color component *i* is used:
*S*(*i*) = −
block diagram in Fig.1 explains in detail the operation for AT method.
Fig. 1. Block diagram of Adaptive thresholding.
iteratively growing by combining smaller regions.
! "
Fig. 2. Block diagram of Statistical region merging.
&#\$% &
**2.2** *Statistical region merging*
merging of regions.
follows in Fig. 2:
-
**2.1** *Adaptive thresholding (AT)*
The most promising in segmentation of the images in general is the approach based on clustering. Cluster oriented-segmentation uses the multidimensional data to partition of the image pixels into clusters. Such kind of technique may be more appropriate than histogram-oriented ones in segmenting images, where each pixel has several attributes and is represented by a vector. Cluster analysis has attracted much attention since the 1960's and has been applied in many fields such as OCR (*Optical Character Recognition*) system. Below, let present three most successful frameworks based on this technique that we apply in segmentation applications.
### **2.3 K-Means clustering algorithm**
K-Means algorithm is an unsupervised clustering algorithm that classifies the input data point into multiple classes based on their inherent distance from each other. The algorithm assumes that the data features from a vector space and tries to find natural clustering in them (Hartigan & Wong, 1979). It works an iterative manner according to the following steps:
$$\dot{\mathbf{x}} = \arg\min\_{k=1,\dots,K} ||\mathbf{x}\_j - m\_k||^2. \tag{2}$$
3. Recalculate the centroids for the clusters according to
$$m\_i = \frac{1}{|\mathbb{C}\_i|} \sum\_{\mathbf{x}\_j \in \mathbb{C}\_i} \mathbf{x}\_{j\prime} \tag{3}$$
where *xj* belong to data set *X* = *x*1, , *xi*, *xN*.
4. If none of the cluster centroids has been changed, finish the algorithm. Otherwise, go to step 2.
In Fig.3, the segmentation process using the K-Means algorithm is exposed:
Fig. 3. Block diagram for K-Means algorithm.
weight factor. However, in many applications *k* = 2 is a common choice. In case of crisp clustering, k may be chosen as 1. The membership value is proportional to the probability that a pixel belongs to some specific cluster where the probability is only dependent on the distance between the pixel and each independent cluster center. So, the criterion E has minimal value when for the pixels that are nearby the corresponding cluster center, higher membership values are assigned, while lower membership values are assigned to the pixels that are far from a center. This algorithm runs with the clusters' number and initial center positions that should be done at beginning, and then, the algorithm determines how many pixels belong to
Fuzzy Image Segmentation Algorithms in Wavelet Domain 131
�*xi* − *cj*�
<sup>2</sup> (*k*−1) , (6)
. (7)
�*xi* − *cm*�
*<sup>j</sup>*=<sup>1</sup> *<sup>u</sup><sup>k</sup> ijxj*
∑*<sup>N</sup> <sup>j</sup>*=<sup>1</sup> *<sup>u</sup><sup>k</sup> ij*
1. The center is initialized with the first value *'t'* of the data to be equal to zero, and this value
*t* = *t* + 1� is changed and novel centers are computed using (7).
Criterion E approaches to minimum value when its variations are decreased according to the restriction that a user should decide. The algorithm also can be interrupted if a user
The FCM algorithm, which is one of the most commonly used procedures, has the following drawback: the number of clusters should be pre-determined by a user before it starts to work. Therefore, sometimes the correct number of clusters in the concrete application may not be the same that the number being chosen by a user. Therefore, a method that should add a process based on fuzzy logic to find the number of clusters to be used. To realize this, we take into consideration the difference between the max (*Vmax*) and the min (*Vmin*) values of intensity in an image *D* = *Vmax* − *Vmin*, these proportions determine the number of clusters. Following, obtained data are applied in the determination of the centers, reducing the operational time of the FCM algorithm. This value is the first data of our fuzzy system called 'Distance', that has six fuzzy sets, 'minimum', 'shorter', 'short', 'regular', 'large' and 'maximum' (see Tab. 1). For value of data of our fuzzy system called 'Size', that has five fuzzy sets, 'Min', 'Small', 'Medium', 'Big' and 'Max' (see Tab. 2). Finally, For value of data of our fuzzy system called 'Cluster', that has five fuzzy sets, 'Very few', 'Few', 'Some', 'Many' and 'Too Many' (see Tab.
each cluster. The membership function and centers are determined as follows:
*<sup>μ</sup>ij* <sup>=</sup> <sup>1</sup> *C* ∑ *m*=1
*ci* <sup>=</sup> <sup>∑</sup>*<sup>N</sup>*
2. The fuzzy partition membership functions *μij* are initialized according to (6).
determines that only a certain number of iterations to be done Bezdek (1981).
The FCM algorithm runs four simple steps:
**2.5 Cluster pre-selection fuzzy C-Means**
3. The value �
3)
is used as a counter for number of iterations.
4. The steps 2 and 3 run until criterion E convergence.
### **Image segmentation using fuzzy methods**
### **Preliminaries and background**
The conventional set theory is based on a binary valued membership, which implies that a particular element either belongs to a particular set or it does not belong to it. A crisp set is defined as one whose elements fully belong to the set and they possess well-defined common attributes, which can be measured quantitatively. In a crisp set the common attributes are equally shared by all the elements of the set. On the other hand, in fuzzy sets, the degree of membership of an element to the set is indicated by a membership value, which signifies the extent to which the element belongs to the set. The membership value lies between 0 and 1, with membership "0" indicating no membership and "1" indicating full membership of the element to the set. In a crisp set, the membership values of its elements are either 0 or 1. The membership of an element *z* in a fuzzy set is obtained using a membership function *μ*(*x*) that maps every element belonging to the fuzzy set *XF* to the interval [0, 1]. Formally, this mapping can be written as:
$$
\mu(\mathfrak{x}): \mathbf{X}\_{\mathcal{F}} \to [0, 1] \tag{4}
$$
The membership assignment is primarily subjective in the sense that the users specify the membership values.
*Selection of the Membership Function* The assignment of the membership function may be performed by several ways.
### **2.4 Fuzzy C-Means algorithm**
Details of fuzzy approach to supervised pattern classification and clustering may be found in (Bezdek, 1981) In fuzzy clustering, a pattern is assigned with a degree of belongings to each cluster in a partition. Here, let present the most popular and efficient fuzzy clustering algorithm: *Fuzzy C-Means Algorithm*. The algorithm should find the center of '*n*' number of clusters iteratively adjusting their position via evaluation of an objective function. Additionally, it permits more flexibility by introducing the partial membership to the other clusters. The classical variant of this algorithm uses the following objective function:
$$E = \sum\_{j=1}^{C} \sum\_{i=1}^{N} \mu\_{ij}^{k} ||\mathbf{x}\_{i} - \mathbf{c}\_{j}||^{2} \tag{5}$$
where *μ<sup>k</sup> ij* is the fuzzy membership of the pixel *xi*; here, the cluster is identified by its center *cj*, and *k* ∈ [1, ∞] is an exponent weight factor. There is no fixed rule for choosing the exponent 4 Will-be-set-by-IN-TECH
The conventional set theory is based on a binary valued membership, which implies that a particular element either belongs to a particular set or it does not belong to it. A crisp set is defined as one whose elements fully belong to the set and they possess well-defined common attributes, which can be measured quantitatively. In a crisp set the common attributes are equally shared by all the elements of the set. On the other hand, in fuzzy sets, the degree of membership of an element to the set is indicated by a membership value, which signifies the extent to which the element belongs to the set. The membership value lies between 0 and 1, with membership "0" indicating no membership and "1" indicating full membership of the element to the set. In a crisp set, the membership values of its elements are either 0 or 1. The membership of an element *z* in a fuzzy set is obtained using a membership function *μ*(*x*) that maps every element belonging to the fuzzy set *XF* to the interval [0, 1]. Formally, this mapping
The membership assignment is primarily subjective in the sense that the users specify the
*Selection of the Membership Function* The assignment of the membership function may be
• *Membership based on visual model*: The membership function may be assigned in accordance with the human visual perceptual model. We may model the variation of the membership values of the pixels in a linear fashion as the pixel gray value changes from 0 to L - 1 (for
• Statistical Distribution: The membership values of the pixels may be assigned on the basis of image statistics as a whole or on the basis of local information at a pixel calculated from the surrounding pixels. The probability density function of the Gaussian or gamma distribution may be used for assignment of membership values (Chaira & Ray, 2003).
Details of fuzzy approach to supervised pattern classification and clustering may be found in (Bezdek, 1981) In fuzzy clustering, a pattern is assigned with a degree of belongings to each cluster in a partition. Here, let present the most popular and efficient fuzzy clustering algorithm: *Fuzzy C-Means Algorithm*. The algorithm should find the center of '*n*' number of clusters iteratively adjusting their position via evaluation of an objective function. Additionally, it permits more flexibility by introducing the partial membership to the other
clusters. The classical variant of this algorithm uses the following objective function:
*N* ∑ *i*=1 *μk*
*ij* is the fuzzy membership of the pixel *xi*; here, the cluster is identified by its center *cj*,
and *k* ∈ [1, ∞] is an exponent weight factor. There is no fixed rule for choosing the exponent
*E* = *C* ∑ *j*=1
*μ*(*x*) : *XF* → [0, 1] (4)
*ij*�*xi* <sup>−</sup> *cj*�2, (5)
**Image segmentation using fuzzy methods**
**Preliminaries and background**
can be written as:
membership values.
performed by several ways.
an L level image).
**2.4 Fuzzy C-Means algorithm**
where *μ<sup>k</sup>*
weight factor. However, in many applications *k* = 2 is a common choice. In case of crisp clustering, k may be chosen as 1. The membership value is proportional to the probability that a pixel belongs to some specific cluster where the probability is only dependent on the distance between the pixel and each independent cluster center. So, the criterion E has minimal value when for the pixels that are nearby the corresponding cluster center, higher membership values are assigned, while lower membership values are assigned to the pixels that are far from a center. This algorithm runs with the clusters' number and initial center positions that should be done at beginning, and then, the algorithm determines how many pixels belong to each cluster. The membership function and centers are determined as follows:
$$\mu\_{ij} = \frac{1}{\sum\_{m=1}^{C} \left(\frac{||\mathbf{x}\_i - \mathbf{c}\_j||}{||\mathbf{x}\_i - \mathbf{c}\_m|| \left(\frac{2}{(k-1)}\right)}\right)},\tag{6}$$
$$c\_i = \frac{\sum\_{j=1}^{N} \boldsymbol{u}\_{ij}^k \boldsymbol{x}\_j}{\sum\_{j=1}^{N} \boldsymbol{u}\_{ij}^k}.\tag{7}$$
The FCM algorithm runs four simple steps:
Criterion E approaches to minimum value when its variations are decreased according to the restriction that a user should decide. The algorithm also can be interrupted if a user determines that only a certain number of iterations to be done Bezdek (1981).
### **2.5 Cluster pre-selection fuzzy C-Means**
The FCM algorithm, which is one of the most commonly used procedures, has the following drawback: the number of clusters should be pre-determined by a user before it starts to work. Therefore, sometimes the correct number of clusters in the concrete application may not be the same that the number being chosen by a user. Therefore, a method that should add a process based on fuzzy logic to find the number of clusters to be used. To realize this, we take into consideration the difference between the max (*Vmax*) and the min (*Vmin*) values of intensity in an image *D* = *Vmax* − *Vmin*, these proportions determine the number of clusters. Following, obtained data are applied in the determination of the centers, reducing the operational time of the FCM algorithm. This value is the first data of our fuzzy system called 'Distance', that has six fuzzy sets, 'minimum', 'shorter', 'short', 'regular', 'large' and 'maximum' (see Tab. 1). For value of data of our fuzzy system called 'Size', that has five fuzzy sets, 'Min', 'Small', 'Medium', 'Big' and 'Max' (see Tab. 2). Finally, For value of data of our fuzzy system called 'Cluster', that has five fuzzy sets, 'Very few', 'Few', 'Some', 'Many' and 'Too Many' (see Tab. 3)
where '*N*' represents the number of clusters to be created and '*j*' is a counter to define all the centers. This looks like a hard type of algorithm, but the centers are still a bit far from the final ones, therefore, there are still a certain number of iterations that should be applied to find them, but the number of iterations is a lot less than for original system, permitting to reduce the computation time. RGB image is discomposed into its three-color channels, and the Euclidean distance is employed (L.A. & Zadeh, 1965) to determine, which one is the difference
Fuzzy Image Segmentation Algorithms in Wavelet Domain 133
*P* ∑ *k*=1 (*x<sup>k</sup> red* <sup>−</sup> *<sup>x</sup><sup>k</sup>*
*P* ∑ *k*=1 (*x<sup>k</sup> red* <sup>−</sup> *<sup>x</sup><sup>k</sup>*
*P* ∑ *k*=1 (*x<sup>k</sup>*
Two distances that are more alike should be combined into one gray scale image, and it is processed as a correct image, then the method proposed is used to determine the number of
1. Divide RGB image into three different images, use (9) to find two images that are more
2. Calculate the distance between intensity levels in the image *D*, and obtain the size of an
3. Feed with these data the fuzzy pre selective system and obtain the number of centers to be
4. Use (8) to obtain the approximate centers. The initial value 't' is equal to zero and it is used
The proposed frameworks in segmentation are based on wavelet analysis, so let present some brief introduction in this part. The continuous wavelet transform (CWT) (Grossman & Morlet,
> *<sup>x</sup>*(*t*) <sup>1</sup> |*a*<sup>|</sup> *ψ*∗ *t* − *b a*
where *b* acts to translate the function across *x*(*t*), and the variable *a* acts to vary the time scale of the probing function, *ψ*. If value *a* is greater than one, the wavelet function, *ψ* is stretched
*blue*)2, (9)
*dt*, (10)
*green*)2,
*blue*)2.
*green* <sup>−</sup> *<sup>x</sup><sup>k</sup>*
*d*1(*xred*, *xblue*) =
*d*2(*xred*, *xgreen*) =
*d*3(*xgreen*, *xblue*) =
similar each to other and use them to create a new gray scale image.
5. The fuzzy partition membership functions *μ<sup>i</sup> j* are initialized according to (6).
6. Let the value be 't=t+1' and compute the new centers using (7). 7. The steps 5 and 6 should be done until criterion E converges.
*<sup>W</sup>*(*a*, *<sup>b</sup>*) = <sup>+</sup><sup>∞</sup>
−∞
clusters to be created. The CPSFCM consists of the next steps:
as a counter for the number of the iterations.
between three distances.
image.
created.
**3. Wavelet texture analysis**
**3.1 Continuous wavelet transform**
1985) can be written as follows:
Table 1. Member functions of "Distance"
Table 2. Member functions of "Size"
Table 3. Member functions of "Clusters"
Fig. 4. Pre-selection of the Number of Clusters.
In the second phase, the number of clusters and its centers are already known, simply dividing the difference *D* into the 'N' clusters and determining its center.
$$c\_{j} = j \frac{D}{N'} \tag{8}$$
$$j = 1,2,3,N\_{\prime} \tag{8}$$
6 Will-be-set-by-IN-TECH
**Fuzzy set Function Center Variance** Minimum Gauss 15 16 Shorter Gauss 53 24 Short Gauss 105 30 Regular Gauss 150 30 Large Gauss 222 45 Maximum Gauss 255 15
**Fuzzy set Function Center Variance** Min Gauss 9000 1.789e+005 Small Gauss 3.015e+005 1.626e+005 Medium Gauss 6.53e+005 1.968e+005 Big Gauss 9.728e+005 2.236e+005 Max Gauss 1.44e+006 2.862e+005
**Fuzzy set Function Center Variance** Very few Gauss 2 3 Few Gauss 7 3 Some Gauss 16 5 Many Gauss 23 5 Too many Gauss 33 7
In the second phase, the number of clusters and its centers are already known, simply dividing
*<sup>N</sup>* , *<sup>j</sup>* <sup>=</sup> 1, 2, 3, , *<sup>N</sup>*, (8)
Table 1. Member functions of "Distance"
Table 2. Member functions of "Size"
Table 3. Member functions of "Clusters"
Fig. 4. Pre-selection of the Number of Clusters.
the difference *D* into the 'N' clusters and determining its center.
*cj* = *j D* where '*N*' represents the number of clusters to be created and '*j*' is a counter to define all the centers. This looks like a hard type of algorithm, but the centers are still a bit far from the final ones, therefore, there are still a certain number of iterations that should be applied to find them, but the number of iterations is a lot less than for original system, permitting to reduce the computation time. RGB image is discomposed into its three-color channels, and the Euclidean distance is employed (L.A. & Zadeh, 1965) to determine, which one is the difference between three distances.
$$d\_1(\mathbf{x}\_{red}, \mathbf{x}\_{blue}) = \sqrt{\sum\_{k=1}^{P} (\mathbf{x}\_{red}^k - \mathbf{x}\_{blue}^k)^2},\tag{9}$$
$$d\_2(\mathbf{x}\_{red}, \mathbf{x}\_{green}) = \sqrt{\sum\_{k=1}^{P} (\mathbf{x}\_{red}^k - \mathbf{x}\_{green}^k)^2},$$
$$d\_3(\mathbf{x}\_{green}, \mathbf{x}\_{blue}) = \sqrt{\sum\_{k=1}^{P} (\mathbf{x}\_{green}^k - \mathbf{x}\_{blue}^k)^2}.$$
Two distances that are more alike should be combined into one gray scale image, and it is processed as a correct image, then the method proposed is used to determine the number of clusters to be created. The CPSFCM consists of the next steps:
### **3. Wavelet texture analysis**
#### **3.1 Continuous wavelet transform**
The proposed frameworks in segmentation are based on wavelet analysis, so let present some brief introduction in this part. The continuous wavelet transform (CWT) (Grossman & Morlet, 1985) can be written as follows:
$$\mathcal{W}(a,b) = \int\_{-\infty}^{+\infty} \mathbf{x}(t) \frac{1}{\sqrt{|a|}} \psi^\* \left(\frac{t-b}{a}\right) dt. \tag{10}$$
where *b* acts to translate the function across *x*(*t*), and the variable *a* acts to vary the time scale of the probing function, *ψ*. If value *a* is greater than one, the wavelet function, *ψ* is stretched
itself can be defined from the scaling function (Rao & Bopardikar, 1998):
**L**
*x*
**H**
Fig. 5. Structure of the analysis filter bank for 2-D image.
**4. Wavelet based texture analysis**
*x*
∞ ∑ *n*=−∞
√
Fuzzy Image Segmentation Algorithms in Wavelet Domain 135
where *d*(*n*) are the series of scalars that are related to the waveform *x*(*t*) and that define the discrete wavelet in terms of the scaling function. While the DWT can be implemented using the above equations, it is usually implemented using filter bank techniques. The use of a group of filters to divide up a signal into various spectral components is termed sub-band coding. The most used implementation of the DWT for 2-D signal applies only two filters for
**2**
**N**
**L**
**H**
**M/2**
**M/2**
**2**
A recent overview of methods applied to segmentation of skin lesions in dermoscopic images (M. Celebi & Stoecker, 2009) results that clustering is the most popular segmentation technique, probably due to their robustness. In the image analysis, texture is an important characteristic, including natural scenes and medical images. It has been noticed that the wavelet transform (WT)provides an ideal representation for texture analysis presenting spatial-frequency properties via a pyramid of tree structures, which is similar to sub-band decomposition. The hierarchical decomposition allows analyzing the high frequencies in the image, which features are importantin the segmentation task. Several works beneficially use the image features within a WT domain during the segmentation process.In paper (Bello, 1994), the image data firstly are decomposed into channels for a selected set of resolution levels using wavelet packets transform, then the Markov random field (MRF) segmentation is applied to the sub-bands coefficients for each scale, starting with the coarsest level, and propagating the segmentation process from current level to segmentation at the next level. Strickland et al. (Strickland & Hahn, 2009) apply the image features extracted in the WT
**N**
2*d*(*n*)*φ*(2*t* − *n*), (15)
**L***<sup>y</sup>*
**2**
**LL**<sup>1</sup>
**LH**<sup>1</sup>
**HL**<sup>1</sup>
**HH**<sup>1</sup>
**2**
**2**
**2**
**H***<sup>y</sup>*
**L***<sup>y</sup>*
**H***<sup>y</sup>*
*ψ*(*t*) =
rows and columns, as in the filter bank, which is shown in 5.
**M**
**N**
along the time axis, and if it is less than one (but still positive) it contacts the function. Wavelets are functions generated from one single function (basis function) called the prototype or mother wavelet by dilations (scalings) and translations (shifts) in time (frequency) domain. If the mother wavelet is denoted by *ψ*(*t*) , the other wavelets *ψa*,*b*(*t*) can be represented as:
$$
\psi\_{a,b}(t) = \frac{1}{\sqrt{|a|}} \psi^\* \left( \frac{t-b}{a} \right). \tag{11}
$$
The variables *a* and *b* represent the parameters for *dilations* and *translations*, respectively in the time axis. If the wavelet function *ψ*(*t*) is appropriately chosen, then it is possible to reconstruct the original waveform from the wavelet coefficients just as in the Fourier transform. Since the CWT decomposes the waveform into coefficients of two variables, a and b, a double summation en discrete case (or integration in continuous case) is required to recover the original signal from the coefficients (Meyers, 1993):
$$\mathbf{x}(t) = \frac{1}{\mathbb{C}} \int\_{a-\infty}^{+\infty} \int\_{b=-\infty}^{+\infty} \mathcal{W}(a,b)\boldsymbol{\upvarphi}\_{a,b}(t) da db,\tag{12}$$
where*C* = <sup>+</sup><sup>∞</sup> −∞ |Ψ(*ω*)| 2 <sup>|</sup>*ω*<sup>|</sup> *<sup>d</sup><sup>ω</sup>* and 0 <sup>&</sup>lt; *<sup>C</sup>* <sup>&</sup>lt; <sup>−</sup><sup>∞</sup> (so called a*admissibility* condition). In fact, reconstruction of the original waveform is rarely performed using the CWT coefficients because of its redundancy.
### **3.2 Discrete wavelet transforms**
The CWT has one serious problem: it is highly redundant. The CWT provides an oversampling of the original waveform: many more coefficients are generated than are actually needed to uniquely specify the signal. The discrete wavelet transform (DWT) achieves this parsimony by restricting the variation in translation and scale, usually to powers of two that is the case of the dyadic wavelet transform. The basic analytical expressions for the DWT is usually implemented using filter banks (Mallat, 1989):
$$\mathbf{x}(t) = \sum\_{k=-\infty}^{\infty} \sum\_{l=-\infty}^{\infty} d(k, l) 2^{-k/2} \psi(2^{-k}t - l). \tag{13}$$
Here, *k* is related to *a* as: *a* = 2*k* ; *b* is related to *λ* as *b* = 2*k* ; and *d*(*k*, *λ*) is a sampling of *W*(*a*, *b*) at discrete points k and *λ*. In the DWT, it is introduced the scaling function, a function that facilitates computation of the DWT. To implement the DWT efficiently, the finest resolution is computed first. The computation then proceeds to coarser resolutions, but rather than start over on the original waveform, the computation uses a smoothed version of the fine resolution waveform. This smoothed version is obtained with the help of the scaling function. The definition of the scaling function uses a dilation or a two-scale difference equation:
$$\phi(t) = \sum\_{n=-\infty}^{\infty} \sqrt{2}c(n)\phi(2t - n). \tag{14}$$
Here, *c*(*n*) are the series of scalars that define the specific scaling function. This equation involves two time scales (*t* and 2*t*) and can be quite difficult to solve. In the DWT, the wavelet 8 Will-be-set-by-IN-TECH
along the time axis, and if it is less than one (but still positive) it contacts the function. Wavelets are functions generated from one single function (basis function) called the prototype or mother wavelet by dilations (scalings) and translations (shifts) in time (frequency) domain. If the mother wavelet is denoted by *ψ*(*t*) , the other wavelets *ψa*,*b*(*t*) can be represented as:
> <sup>|</sup>*a*<sup>|</sup> *ψ*∗ *t* − *b a*
The variables *a* and *b* represent the parameters for *dilations* and *translations*, respectively in the time axis. If the wavelet function *ψ*(*t*) is appropriately chosen, then it is possible to reconstruct the original waveform from the wavelet coefficients just as in the Fourier transform. Since the CWT decomposes the waveform into coefficients of two variables, a and b, a double summation en discrete case (or integration in continuous case) is required to recover the
. (11)
*W*(*a*, *b*)*ψa*,*b*(*t*)*dadb*, (12)
*<sup>d</sup>*(*k*, <sup>l</sup>)2−*k*/2*ψ*(2<sup>−</sup>*kt* <sup>−</sup> <sup>l</sup>). (13)
2*c*(*n*)*φ*(2*t* − *n*). (14)
*<sup>ψ</sup>a*,*b*(*t*) = <sup>1</sup>
+∞ *a*−∞ +∞ *b*=−∞
reconstruction of the original waveform is rarely performed using the CWT coefficients
The CWT has one serious problem: it is highly redundant. The CWT provides an oversampling of the original waveform: many more coefficients are generated than are actually needed to uniquely specify the signal. The discrete wavelet transform (DWT) achieves this parsimony by restricting the variation in translation and scale, usually to powers of two that is the case of the dyadic wavelet transform. The basic analytical expressions for
Here, *k* is related to *a* as: *a* = 2*k* ; *b* is related to *λ* as *b* = 2*k* ; and *d*(*k*, *λ*) is a sampling of *W*(*a*, *b*) at discrete points k and *λ*. In the DWT, it is introduced the scaling function, a function that facilitates computation of the DWT. To implement the DWT efficiently, the finest resolution is computed first. The computation then proceeds to coarser resolutions, but rather than start over on the original waveform, the computation uses a smoothed version of the fine resolution waveform. This smoothed version is obtained with the help of the scaling function. The definition of the scaling function uses a dilation or a two-scale difference equation:
<sup>|</sup>*ω*<sup>|</sup> *<sup>d</sup><sup>ω</sup>* and 0 <sup>&</sup>lt; *<sup>C</sup>* <sup>&</sup>lt; <sup>−</sup><sup>∞</sup> (so called a*admissibility* condition). In fact,
original signal from the coefficients (Meyers, 1993):
where*C* = <sup>+</sup><sup>∞</sup>
−∞
**3.2 Discrete wavelet transforms**
because of its redundancy.
the DWT is usually implemented using filter banks (Mallat, 1989):
∞ ∑ *k*=−∞
*φ*(*t*) =
∞ ∑ *l*=−∞
∞ ∑ *n*=−∞
√
Here, *c*(*n*) are the series of scalars that define the specific scaling function. This equation involves two time scales (*t* and 2*t*) and can be quite difficult to solve. In the DWT, the wavelet
*x*(*t*) =
itself can be defined from the scaling function (Rao & Bopardikar, 1998):
$$\psi(t) = \sum\_{n = -\infty}^{\infty} \sqrt{2}d(n)\phi(2t - n),\tag{15}$$
where *d*(*n*) are the series of scalars that are related to the waveform *x*(*t*) and that define the discrete wavelet in terms of the scaling function. While the DWT can be implemented using the above equations, it is usually implemented using filter bank techniques. The use of a group of filters to divide up a signal into various spectral components is termed sub-band coding. The most used implementation of the DWT for 2-D signal applies only two filters for rows and columns, as in the filter bank, which is shown in 5.
Fig. 5. Structure of the analysis filter bank for 2-D image.
### **4. Wavelet based texture analysis**
A recent overview of methods applied to segmentation of skin lesions in dermoscopic images (M. Celebi & Stoecker, 2009) results that clustering is the most popular segmentation technique, probably due to their robustness. In the image analysis, texture is an important characteristic, including natural scenes and medical images. It has been noticed that the wavelet transform (WT)provides an ideal representation for texture analysis presenting spatial-frequency properties via a pyramid of tree structures, which is similar to sub-band decomposition. The hierarchical decomposition allows analyzing the high frequencies in the image, which features are importantin the segmentation task. Several works beneficially use the image features within a WT domain during the segmentation process.In paper (Bello, 1994), the image data firstly are decomposed into channels for a selected set of resolution levels using wavelet packets transform, then the Markov random field (MRF) segmentation is applied to the sub-bands coefficients for each scale, starting with the coarsest level, and propagating the segmentation process from current level to segmentation at the next level. Strickland et al. (Strickland & Hahn, 2009) apply the image features extracted in the WT domain for detection of microcalcifications in mammograms using a matching process and some a priori knowledge on the target objects.
Zhang et al. (Zhang & Desai, 2001) employ a Bayes classifier on wavelet coefficients to determine an appropriate scale and threshold that can separate segmentation targets from other features.
### **5. Proposed framework**
The idea of our approach is consisted in employing the feature extraction in WT space before the segmentation process where the main difference with other algorithms presented in literature is in usage the information from three color channels in WT space gathering the color channels via a nearest neighbour interpolation (NNI). Developed approach uses the procedure that consists of the following stages: a digital color image I[n,m] is separated in R, G and Bchannels in the color space, where each a color channel is decomposed calculating their wavelets coefficients using Mallat's pyramid algorithm (Mallat, 1989). For chosen wavelet family is being used, the original image is decomposed into four sub-bands (Fig.5). These sub-bands labeled as LH, HL and HH represent the finest scale wavelet coefficient (detail images), while the sub-band LL corresponds to coarse level coefficients (approximation image), noted below as *<sup>D</sup>*(2*<sup>i</sup>* ) *<sup>h</sup>* ,*D*(2*<sup>i</sup>* ) *<sup>v</sup>* ,*D*(2*<sup>i</sup>* ) *<sup>d</sup>* and *<sup>A</sup>*(2*<sup>i</sup>* ), respectively at given scale 2*<sup>j</sup>* , for *j* = 1, 2, . . . , *J*, where J is the numbers of scales used in the DWT (Kravchenko, 2009). Finally, the DWT can be represented as follows:
$$\mathcal{W}\_{\mathbf{i}} = |\mathcal{W}\_{\mathbf{i}}| \exp(j\Theta\_{\mathbf{i}}) \,\prime \,\tag{16}$$
$$|\mathcal{W}\_{i}| = \left(\sqrt{|D\_{h,i}|^2 + |D\_{v,i}|^2 + |D\_{d,i}|^2}\right)^2,\tag{17}$$
$$\Theta\_{\dot{l}} = \begin{cases} \mathfrak{a}\_{\dot{l}} & \text{if } D\_{\mathfrak{h},\dot{l}} > 0 \\ \pi - \Theta\_{\dot{l}} \text{ if } D\_{\mathfrak{h},\dot{l}} < 0 \text{ /} \end{cases} \tag{18}$$
Fig. 6. Block diagram of proposed framework.
In this section, let present the evaluation criteria focusing them in segmentation process in dermoscopic image. The same measures can be used for segmentation in other applications. Different objective measures are used in literature for the purpose of evaluation of the segmentation performance in dermoscopic images. For objective measures, there is needed
Fuzzy Image Segmentation Algorithms in Wavelet Domain 137
**6. Evaluation criteria**
$$
\Theta\_i = \tan^{-1} \left( \frac{D\_{v,i}}{D\_{h,i}} \right).
$$
Therefore, *Wi* is considered as a new image for each color channel. The following process employed in the wavelet transform space is consisted of the stages: the classic segmentation method is applied to images; the image segmented corresponding to the red channel is interpolated with the image segmented corresponding to the green channel, the found image after applying *NNI* process is interpolated with the image segmented corresponding to the blue channel using *NNI* again, finally, this image is considerers the output of the segmentation procedure, Fig. 6 shows the block diagram of the above. The importance of considering the information of the three-color channels is an advantage in the segmentation process as it is judged to clusters formed in each of them.
The block diagram in Fig. 7 explains the operations for: a) image segmentation if K-Means algorithm is used where the WT is applied, named as WK-Means; b) image segmentation if FCM algorithm is used where the WT is applied, named as W-FCM; finally c) image segmentation if CPSFCM algorithm is used where the WT is applied, named as W-CPSFCM.
10 Will-be-set-by-IN-TECH
domain for detection of microcalcifications in mammograms using a matching process and
Zhang et al. (Zhang & Desai, 2001) employ a Bayes classifier on wavelet coefficients to determine an appropriate scale and threshold that can separate segmentation targets from
The idea of our approach is consisted in employing the feature extraction in WT space before the segmentation process where the main difference with other algorithms presented in literature is in usage the information from three color channels in WT space gathering the color channels via a nearest neighbour interpolation (NNI). Developed approach uses the procedure that consists of the following stages: a digital color image I[n,m] is separated in R, G and Bchannels in the color space, where each a color channel is decomposed calculating their wavelets coefficients using Mallat's pyramid algorithm (Mallat, 1989). For chosen wavelet family is being used, the original image is decomposed into four sub-bands (Fig.5). These sub-bands labeled as LH, HL and HH represent the finest scale wavelet coefficient (detail images), while the sub-band LL corresponds to coarse level coefficients (approximation
1, 2, . . . , *J*, where J is the numbers of scales used in the DWT (Kravchenko, 2009). Finally,
*Dv*,*<sup>i</sup> Dh*,*<sup>i</sup>* .
*α<sup>i</sup>* if *Dh*,*<sup>i</sup>* > 0
Therefore, *Wi* is considered as a new image for each color channel. The following process employed in the wavelet transform space is consisted of the stages: the classic segmentation method is applied to images; the image segmented corresponding to the red channel is interpolated with the image segmented corresponding to the green channel, the found image after applying *NNI* process is interpolated with the image segmented corresponding to the blue channel using *NNI* again, finally, this image is considerers the output of the segmentation procedure, Fig. 6 shows the block diagram of the above. The importance of considering the information of the three-color channels is an advantage in the segmentation process as it is
The block diagram in Fig. 7 explains the operations for: a) image segmentation if K-Means algorithm is used where the WT is applied, named as WK-Means; b) image segmentation if FCM algorithm is used where the WT is applied, named as W-FCM; finally c) image segmentation if CPSFCM algorithm is used where the WT is applied, named as W-CPSFCM.
Θ*<sup>i</sup>* = tan−<sup>1</sup>
), respectively at given scale 2*<sup>j</sup>*
*Wi* = |*Wi*|*exp*(*j*Θ*i*), (16)
*<sup>π</sup>* <sup>−</sup> <sup>Θ</sup>*<sup>i</sup>* if *Dh*,*<sup>i</sup>* <sup>&</sup>lt; 0, , (18)
2
, for *j* =
, (17)
some a priori knowledge on the target objects.
other features.
**5. Proposed framework**
image), noted below as *<sup>D</sup>*(2*<sup>i</sup>*
the DWT can be represented as follows:
judged to clusters formed in each of them.
) *<sup>h</sup>* ,*D*(2*<sup>i</sup>*
) *<sup>v</sup>* ,*D*(2*<sup>i</sup>* ) *<sup>d</sup>* and *<sup>A</sup>*(2*<sup>i</sup>*
Θ*<sup>i</sup>* =
Fig. 6. Block diagram of proposed framework.
### **6. Evaluation criteria**
In this section, let present the evaluation criteria focusing them in segmentation process in dermoscopic image. The same measures can be used for segmentation in other applications. Different objective measures are used in literature for the purpose of evaluation of the segmentation performance in dermoscopic images. For objective measures, there is needed
the ground truth (GT) image, which is determined by dermatologist manually drawing the
Fuzzy Image Segmentation Algorithms in Wavelet Domain 139
Employing GT image Hance *et al*. (Hance, 1996) calculated the operation exclusive disjunction (XOR) measure, other metrics used in segmentation performance are presented in(Garnavi, 2011): the *sensitivity* and *specificity, precision and recall, true positive rate, false positive rate, pixel misclassification probability*, the *weighted performance index*, among others. Below, let consider the *sensitivity* and *specificity* measure. Sensitivity and specificity are statistical measures of the performance of a binary classification test, commonly used in medical studies. In the context of segmentation of skin lesions, sensitivity measures the proportion of actual lesion pixels that are correctly identified as such. Specificity measures the proportion of background skin pixels
that are correctly identified. Given the following definitions:
**TP** true positive, object pixels that are correctly classified as interest object.
**FP** false positive, background pixels that are incorrectly identified as interest object. **TN** true negative, background pixels that are correctly identified as background. **FN** false negative, object pixels that are incorrectly identified as background. In each of the above categories, the sensitivity and specificity are given by:
*sensitivity* <sup>=</sup> *TP*
*speci ficity* <sup>=</sup> *TN*
We also apply the *Receiver Operating Characteristic* (ROC) analysis (Fig. 8) that permits to evaluate the image segmentation quality in terms of the ability of human observer or a computer algorithm using image data to classify patients as "*positive*" or "*negative*" with respect to any particular disease. This characteristic represents the second level of diagnostic efficacy in the hierarchical model described by Fryback and Thornbury (Fryback DG, 1991). Fig. 8 presents the points of the ROC curve that are obtained by sweeping the classification threshold from the most positive classification value to the most negative. These points are desirable to produce quantitative summary measure using the ROC curve, called as an area
In the processing area of biomedical image processing, we applied the developed and existed segmentation techniques to dermoscopic images. Let present some definitions of commonly used terms in this application area. The term "skin cancer" refers to three different conditions
The two most common forms of skin cancer are basal cell carcinoma and squamous cell carcinoma. Together, these two are also referred to as nonmelanoma skin cancer.
that are from the least to the most dangerous can be presented as follows:
• Squamous cell carcinoma (the first stage of which is called *actinic keratosis*)
• Basal cell carcinoma (or basal cell *carcinomaepithelioma*)
*TP* <sup>+</sup> *FN* (19)
*FP* <sup>+</sup> *TN* (20)
border around the lesion.
under the ROC curve (AUC).
**7. Dermoscopic images**
• *Melanoma*
Fig. 7. Block diagram of the proposed algorithms: a) segmentation with WK-MEANS; b) segmentation with W-FCM; c) segmentation with W-CPSFCM.
12 Will-be-set-by-IN-TECH
Fig. 7. Block diagram of the proposed algorithms: a) segmentation with WK-MEANS; b)
segmentation with W-FCM; c) segmentation with W-CPSFCM.
the ground truth (GT) image, which is determined by dermatologist manually drawing the border around the lesion.
Employing GT image Hance *et al*. (Hance, 1996) calculated the operation exclusive disjunction (XOR) measure, other metrics used in segmentation performance are presented in(Garnavi, 2011): the *sensitivity* and *specificity, precision and recall, true positive rate, false positive rate, pixel misclassification probability*, the *weighted performance index*, among others. Below, let consider the *sensitivity* and *specificity* measure. Sensitivity and specificity are statistical measures of the performance of a binary classification test, commonly used in medical studies. In the context of segmentation of skin lesions, sensitivity measures the proportion of actual lesion pixels that are correctly identified as such. Specificity measures the proportion of background skin pixels that are correctly identified. Given the following definitions:
**TP** true positive, object pixels that are correctly classified as interest object.
**FP** false positive, background pixels that are incorrectly identified as interest object.
**TN** true negative, background pixels that are correctly identified as background.
**FN** false negative, object pixels that are incorrectly identified as background.
In each of the above categories, the sensitivity and specificity are given by:
$$sensitivity = \frac{TP}{TP + FN} \tag{19}$$
$$Specificity = \frac{TN}{FP + TN} \tag{20}$$
We also apply the *Receiver Operating Characteristic* (ROC) analysis (Fig. 8) that permits to evaluate the image segmentation quality in terms of the ability of human observer or a computer algorithm using image data to classify patients as "*positive*" or "*negative*" with respect to any particular disease. This characteristic represents the second level of diagnostic efficacy in the hierarchical model described by Fryback and Thornbury (Fryback DG, 1991). Fig. 8 presents the points of the ROC curve that are obtained by sweeping the classification threshold from the most positive classification value to the most negative. These points are desirable to produce quantitative summary measure using the ROC curve, called as an area under the ROC curve (AUC).
### **7. Dermoscopic images**
In the processing area of biomedical image processing, we applied the developed and existed segmentation techniques to dermoscopic images. Let present some definitions of commonly used terms in this application area. The term "skin cancer" refers to three different conditions that are from the least to the most dangerous can be presented as follows:
The two most common forms of skin cancer are basal cell carcinoma and squamous cell carcinoma. Together, these two are also referred to as nonmelanoma skin cancer.
Fig. 9. Block diagram of CAD system.
different nature used in this study.
Melanoma (lesion2)
segmentation. The dataset presents 24-bits color images in JPEG format with 600 x 600 pixel size. Below, we only expose five different images with different texture characteristics where the sensitivity and specificity are used as the evaluation criteria for segmentation accuracy. We also plotted the ROC curves to examine the classifier performance. Additionally, the diagnostic performance was quantified by AUC measure. Fig. 10 shows the images of
Fuzzy Image Segmentation Algorithms in Wavelet Domain 141
(a) (b) (c)
(d) (e)
The simulation results in Table present the values of AUC for the proposed framework based on different wavelet families confirming their better performance in comparison with classical techniques. The maximum value of AUC is obtained when WF Daubechies 4 is used, followed by the WAF *π*6. According to (Fryback DG, 1991) AUC measure should have values greater than 0.8 to consider a good test, but our study is focused in the best approximation of
Based on the objective quantity metrics and subjective visual results presented in Fig.4, one can see that the W-FCM presents borders that characterize the lesion (green color), in Fig.11
Fig. 10. Images used in this study:a) *Flower* b) *sea shell* c) *Tree* d)*Melanoma* (lesion1) e))
segmented image to GT, this means to get the value of AUC approximated to one.
Fig. 8. ROC curve.
Melanoma is generally the most serious form of skin cancer because it tends to spread (metastasize) throughout the body quickly. For a diagnosis, doctors usually remove all or a part of the growth by performing a biopsy but is considered an invasive technique. Alternative, dermatoscopy reduces the need for a biopsy applying a dermatoscope device, which magnifies the sub surface structures with the use of oil and illumination, also called epiluminescence. Dermatoscopy is a particularly helpful standard method of diagnosing the malignancy of skin lesions (Argenziano & Soyer, 2001). A mayor advantage is the accuracy of dermatoscopy is increased to 20% in the case of sensitivity and up to 10% in the case of specificity, compared with naked-eye examination, permitting to reduce the frequency of unnecessary surgical excisions of benign lesions (Vestergaard, 2001). Several instruments designed for a computer aided diagnosis (CAD) (Fig. 9 of skin lesions have been proposed, which usually work in four steps: data acquisition of skin (dermoscopic images), segmentation, feature extraction and classification. The most relevant step is segmentation process because it provides fundamental information to the next stages. Image segmentation is the process of adequately grouping pixels into a few regions, which pixels share some similar characteristics. Automated analysis of edges, colors, and shape of the lesion relies upon an accurate segmentation and is an important first step in any CAD system but irregular shape, nonuniform color, and ambiguous structures make the problem difficult.
### **8. Simulation results**
This section presents numerous experimental results in segmentation obtained by developed and existed techniques. The segmentation algorithms were evaluated on a set of 50 images of dermoscopic images obtained from http://www.dermoscopyatlas.com and http://www.wisdom.weizmann.ac.il. The GT images were found via human based
Fig. 9. Block diagram of CAD system.
14 Will-be-set-by-IN-TECH
Melanoma is generally the most serious form of skin cancer because it tends to spread (metastasize) throughout the body quickly. For a diagnosis, doctors usually remove all or a part of the growth by performing a biopsy but is considered an invasive technique. Alternative, dermatoscopy reduces the need for a biopsy applying a dermatoscope device, which magnifies the sub surface structures with the use of oil and illumination, also called epiluminescence. Dermatoscopy is a particularly helpful standard method of diagnosing the malignancy of skin lesions (Argenziano & Soyer, 2001). A mayor advantage is the accuracy of dermatoscopy is increased to 20% in the case of sensitivity and up to 10% in the case of specificity, compared with naked-eye examination, permitting to reduce the frequency of unnecessary surgical excisions of benign lesions (Vestergaard, 2001). Several instruments designed for a computer aided diagnosis (CAD) (Fig. 9 of skin lesions have been proposed, which usually work in four steps: data acquisition of skin (dermoscopic images), segmentation, feature extraction and classification. The most relevant step is segmentation process because it provides fundamental information to the next stages. Image segmentation is the process of adequately grouping pixels into a few regions, which pixels share some similar characteristics. Automated analysis of edges, colors, and shape of the lesion relies upon an accurate segmentation and is an important first step in any CAD system but irregular
shape, nonuniform color, and ambiguous structures make the problem difficult.
dermoscopic images obtained from http://www.dermoscopyatlas.com and
This section presents numerous experimental results in segmentation obtained by developed and existed techniques. The segmentation algorithms were evaluated on a set of 50 images of
http://www.wisdom.weizmann.ac.il. The GT images were found via human based
Fig. 8. ROC curve.
**8. Simulation results**
segmentation. The dataset presents 24-bits color images in JPEG format with 600 x 600 pixel size. Below, we only expose five different images with different texture characteristics where the sensitivity and specificity are used as the evaluation criteria for segmentation accuracy. We also plotted the ROC curves to examine the classifier performance. Additionally, the diagnostic performance was quantified by AUC measure. Fig. 10 shows the images of different nature used in this study.
Fig. 10. Images used in this study:a) *Flower* b) *sea shell* c) *Tree* d)*Melanoma* (lesion1) e)) Melanoma (lesion2)
The simulation results in Table present the values of AUC for the proposed framework based on different wavelet families confirming their better performance in comparison with classical techniques. The maximum value of AUC is obtained when WF Daubechies 4 is used, followed by the WAF *π*6. According to (Fryback DG, 1991) AUC measure should have values greater than 0.8 to consider a good test, but our study is focused in the best approximation of segmented image to GT, this means to get the value of AUC approximated to one.
Based on the objective quantity metrics and subjective visual results presented in Fig.4, one can see that the W-FCM presents borders that characterize the lesion (green color), in Fig.11
(a) (b)
Fuzzy Image Segmentation Algorithms in Wavelet Domain 143
(c) (d)
(e) (f)
(g) (h)
(i) (j)
Fig. 11. Image segmentation results under different algorithms using: a) Melanoma, b) Ground Truth, c) FCM, d) W-FCM with WF Coiflets 3, e) W-FCM with Daubechies 4, f) W-FCM with WF biorthogonal 6.8, g) W-FCM with WAF *up*2, h) W-FCM with WAF *π*6" i)
W-FCM with WAF *fup*2, j) W-FCM with WAF *e*2.
Table 4. AUC simulation results using different segmentation algorithms
c-f, it is easy to note that the segmentation procedure has performed only around the lesion. On other hand, in Fig. 11 g-j, where WAF results are presented, one can see that together with segmentation of lesion boarder there are some areas into the lesion segmented.
Figure 12 presents ROC curves for lesion 1 comparing the classic and proposed algorithms. In particular, Fig.11c) exposes the ROC curves for WK-means and K-Means algorithms where one can see superiority of proposed WK-Means algorithm that uses WAFp6 (see ROC curve in light green color), Fig.12 d) presents ROC curves for W-FCM and FCM algorithms where it is easy to observe the better performance of WK-Means that employs the WF biorthogonal 6.8 (seeROC curve in red color), and finally, in Fig. 12 e), theROC curves for W-CPSFCM and CPSFCM algorithms have confirmed the better performance of the first one for WF biorthogonal 6.8usage (see ROC curve in red color).
16 Will-be-set-by-IN-TECH
Table 4. AUC simulation results using different segmentation algorithms
biorthogonal 6.8usage (see ROC curve in red color).
segmentation of lesion boarder there are some areas into the lesion segmented.
c-f, it is easy to note that the segmentation procedure has performed only around the lesion. On other hand, in Fig. 11 g-j, where WAF results are presented, one can see that together with
Figure 12 presents ROC curves for lesion 1 comparing the classic and proposed algorithms. In particular, Fig.11c) exposes the ROC curves for WK-means and K-Means algorithms where one can see superiority of proposed WK-Means algorithm that uses WAFp6 (see ROC curve in light green color), Fig.12 d) presents ROC curves for W-FCM and FCM algorithms where it is easy to observe the better performance of WK-Means that employs the WF biorthogonal 6.8 (seeROC curve in red color), and finally, in Fig. 12 e), theROC curves for W-CPSFCM and CPSFCM algorithms have confirmed the better performance of the first one for WF
Fig. 11. Image segmentation results under different algorithms using: a) Melanoma, b) Ground Truth, c) FCM, d) W-FCM with WF Coiflets 3, e) W-FCM with Daubechies 4, f) W-FCM with WF biorthogonal 6.8, g) W-FCM with WAF *up*2, h) W-FCM with WAF *π*6" i) W-FCM with WAF *fup*2, j) W-FCM with WAF *e*2.
**9. Conclusion**
in comparison with traditional existed techniques.
**10. Acknowledgement**
**11. References**
for their support to realize this work.
Technology 17(1): 33–44.
URL: *www.intechweb.org*
*Letters* 12(24): 1943–1950.
New York.
publication.
Medicine and Biology Magazine 15(1): 104–111.
11(1): 88–94.
The segmentation process involves the partition of an image into a set of homogeneous and meaningful regions allowing the detection of an object of interest in a specific task, and is an important stage in the different problems such as computer vision, remote sensing, medical images, etc. In this chapter, we present a review of existed promising methods of image segmentation; some of them are popular because they are used in various applications. Novel approach in segmentation exposed here has generated several frameworks that use traditional and fuzzy logic techniques (WK-Means, W-FCM, W-CPSFCM), all of them involve the wavelet transform space and approximation procedures for inter color channels processing, permitting better extraction of the image features. Numerous simulation results summarize the performance of all investigated algorithms for segmentation in images of different nature exposing quality in form of ROC curves (sensitivity-specificity parameters) and AUC values. It has been justified sufficiently better performance of the developed frameworks (WK-Means, W-FCM, and W-CPSFCM) that apply different classic wavelets families and WAF
Fuzzy Image Segmentation Algorithms in Wavelet Domain 145
The authors thank the National Polytechnic Institute of Mexico and CONACYT (grant 81599)
Garnavi,R. (Aldeen, M. (Celebi, M. E. (2011). Weighted performance index for objective
Fryback, D. G. (Thornbury, J.R. (1991). The efficacy of diagnostic imaging, Med Decis Making
Hance, G. A (Umbaugh, S.E. ((Moss, R. H. (Stoecker, W. V. 1991). Unsupercised Color
Argenziano, G. & Soyer, H. (1996). Adaptive thresholding of wavelet coefficients, Computational Statistics and amp. Data Analysis Vol.22(No.4): 351 – 361. Argenziano, G. & Soyer, H. (2001). Dermoscopy of pigmented skin lesions-a valuable tool for
Bello, M. (1994). A combined markov random field and wave-packet transform-based approach for image segmentation, IEEE Trans. Image Processing 3(6): 834–846.
Bezdek, J. (1981). *Pattern Recognition with Fuzzy Objective Function Algorithms, Plenum press*,
Chaira & Ray, A. K. (2003). Fuzzy approach to color region extraction, *Pattern Recognition*
Fryback DG, T. J. (1991). The efficacy of diagnostic imaging, Med Decis Making 11(1): 88U94. ˝ Gonzalez, R. C. & Woods, R. E. (1992). *Digital Image Processing*, Addison Wesley, Place of
early diagnosis of melanoma, The Lancet Oncology 2(7): 443U449. ˝
evaluation of border detection methods in dermoscopy images, Skin Research and
Image Segmentation with Application to Skin Tumor Borders, IEEE Engineering in
Fig. 12. a) Lesion 1 Melanoma b) Ground Truth image; ROC curves for c) WK-Means algorithm d) FCM algorithm e) W-CPSFCM: for WF Daubechies 4(dark blue), for WF biorthogonal 6.8 (red), for WF Coiflets 3 (purple), for WAF *up*<sup>2</sup> (dark green), for WAF fup2 (aqua), for WAF *π*<sup>6</sup> (light green); FCM (black).
### **9. Conclusion**
18 Will-be-set-by-IN-TECH
(a) (b)
(c) (d)
(e)
Fig. 12. a) Lesion 1 Melanoma b) Ground Truth image; ROC curves for c) WK-Means algorithm d) FCM algorithm e) W-CPSFCM: for WF Daubechies 4(dark blue), for WF biorthogonal 6.8 (red), for WF Coiflets 3 (purple), for WAF *up*<sup>2</sup> (dark green), for WAF fup2
(aqua), for WAF *π*<sup>6</sup> (light green); FCM (black).
The segmentation process involves the partition of an image into a set of homogeneous and meaningful regions allowing the detection of an object of interest in a specific task, and is an important stage in the different problems such as computer vision, remote sensing, medical images, etc. In this chapter, we present a review of existed promising methods of image segmentation; some of them are popular because they are used in various applications. Novel approach in segmentation exposed here has generated several frameworks that use traditional and fuzzy logic techniques (WK-Means, W-FCM, W-CPSFCM), all of them involve the wavelet transform space and approximation procedures for inter color channels processing, permitting better extraction of the image features. Numerous simulation results summarize the performance of all investigated algorithms for segmentation in images of different nature exposing quality in form of ROC curves (sensitivity-specificity parameters) and AUC values. It has been justified sufficiently better performance of the developed frameworks (WK-Means, W-FCM, and W-CPSFCM) that apply different classic wavelets families and WAF in comparison with traditional existed techniques.
### **10. Acknowledgement**
The authors thank the National Polytechnic Institute of Mexico and CONACYT (grant 81599) for their support to realize this work.
### **11. References**
**Part 2**
**Techniques and Implementation**
## **Part 2**
**Techniques and Implementation**
20 Will-be-set-by-IN-TECH
146 Fuzzy Logic – Algorithms, Techniques and Implementations
Grossman, A. & Morlet, J. . (1985). *Mathematics and Physics: Lectures on Recent Results*, L. Streit,
Hartigan, A. & Wong, M. A. (1979). A k-means clustering algorithm, Applied Statistics
Kravchenko, V., M. H. P. V. . (2009). *Adaptive digital processing of multidimensional signals with*
L.A. & Zadeh (1965). Fuzzy approach to color region extraction, *Information and Control*
M. Celebi, H. Iyatomi, G. S. & Stoecker, W. (2009). Lesion border detection in dermoscopy images, Computerized Medical Imaging and Graphics 33(2): 148U153.
M. Celebi, H. Kingravi, H. I. e. a. (2008). Border detection in dermoscopy images using statistical region merging, Skin Research and Technology Vol. 3(No. 14): 347–353. Mallat, S. (1989). A theory for multiresolution signal decomposition: The wavelet
Rao, R. M. & Bopardikar, A. S. (1998). *Wavelet Transforms: Introduction to Theory and*
Strickland, R. N. & Hahn, H. I. (2009). Wavelet transform matched filters for the
Zhang, X. & Desai, M. (2001). Segmentation of bright targets using wavelets and adaptive
Meyers, Y. (1993). *Wavelet: Algorithms and Applications*, SIAM, Place of publication.
thresholding, IEEE Trans. Image Processing 10(7): 1020–1030.
*Applications.*, Addison-Wesley, Place of publication.
representation, IEEE Trans. on Pattern Analysis and Machine Intelligence 11(7): 338
detection and classification of microcalcifications in mammography, Proceedings of the International Conference on Image Processing, Washington 1(2): 422–425. Vestergaard, ME; Macaskill, P. H. P. M. (2001). Dermoscopy compared with naked eye
examination for the diagnosis of primary melanoma: a metaanalysis of studies performed in a clinical setting, British Journal of Dermatology 159(3): 669U76.
˝
˝
Place of publication.
*applications.*, FizMatLit, Place of publication.
28(1): 100–108.
8(3): 338 – 353.
– 353.
**8**
*Albania*
**Fuzzy Logic Approach for**
One of the main challenges nowadays for managing IP networks is guaranteeing quality of service. One of the proposed solutions is traffic management with MPLS protocol. However, requirement characterization and the network state are very difficult tasks, taking into account that requirements for different services are random, where as a result the network condition varies dynamically and randomly. This is reason why researches have used fuzzy logic for solving a lot of problems that can occur in very dynamic networks. In this chapter we will analyze MPLS network routing metrics using fuzzy logic. We will pay attention the most appropriate defuzzification methods for finding the path that fulfills the QoS
One of the key issues in providing end-to-end quality of service (QoS) guarantees in today's networks is how to determine a feasible route that satisfies a set of constraints. In general, finding a path subject to multiple constraints is an NP-complete problem that cannot be exactly solved in polynomial time. Accordingly, several heuristics and approximation algorithms have been proposed for this problem. Many of these algorithms suffer from
Selecting feasible paths that satisfy various QoS requirements of applications in a network is known as QoS routing [1]. In general, two issues are related to QoS routing: state distribution and routing strategy. State distribution addresses the issue of exchanging the state information throughout the network. Routing strategy is used to find a feasible path that meets the QoS requirements. In this chapter we will present the fuzzy logic approach for QoS routing analysis in a network which is able to offer multimedia services, such is MPLS network [2] [3] [4] [5] [6]. Fuzzy sets offer powerful mathematical structure that has to do with non-preciosity and uncertainty of real word. Linguistic variables allow representation of numerical values with fuzzy sets. Knowing that networks nowadays are very dynamic, which means that networks have parameters that are affected from unexpected overloads, failures and other concerns, fuzzy logic offers promising approach for addressing different network problems [7] [8]. Applying of fuzzy logic in telecommunication networks is done lately and is proved to be a very economic and efficient method compared with other methods used in automatic control. Recent researches on application of fuzzy logic in telecommunication networks have to do with: packet queuing, buffer management, call acceptance, QoS routing, channel capacity
**1. Introduction**
requirements for multimedia services.
sharing, traffic management etc.
either excessive computational cost or low performance.
**QoS Routing Analysis**
Adrian Shehu and Arianit Maraj *Polytechnic University of Tirana,*
### **Fuzzy Logic Approach for QoS Routing Analysis**
Adrian Shehu and Arianit Maraj *Polytechnic University of Tirana, Albania*
### **1. Introduction**
One of the main challenges nowadays for managing IP networks is guaranteeing quality of service. One of the proposed solutions is traffic management with MPLS protocol. However, requirement characterization and the network state are very difficult tasks, taking into account that requirements for different services are random, where as a result the network condition varies dynamically and randomly. This is reason why researches have used fuzzy logic for solving a lot of problems that can occur in very dynamic networks. In this chapter we will analyze MPLS network routing metrics using fuzzy logic. We will pay attention the most appropriate defuzzification methods for finding the path that fulfills the QoS requirements for multimedia services.
One of the key issues in providing end-to-end quality of service (QoS) guarantees in today's networks is how to determine a feasible route that satisfies a set of constraints. In general, finding a path subject to multiple constraints is an NP-complete problem that cannot be exactly solved in polynomial time. Accordingly, several heuristics and approximation algorithms have been proposed for this problem. Many of these algorithms suffer from either excessive computational cost or low performance.
Selecting feasible paths that satisfy various QoS requirements of applications in a network is known as QoS routing [1]. In general, two issues are related to QoS routing: state distribution and routing strategy. State distribution addresses the issue of exchanging the state information throughout the network. Routing strategy is used to find a feasible path that meets the QoS requirements. In this chapter we will present the fuzzy logic approach for QoS routing analysis in a network which is able to offer multimedia services, such is MPLS network [2] [3] [4] [5] [6]. Fuzzy sets offer powerful mathematical structure that has to do with non-preciosity and uncertainty of real word. Linguistic variables allow representation of numerical values with fuzzy sets. Knowing that networks nowadays are very dynamic, which means that networks have parameters that are affected from unexpected overloads, failures and other concerns, fuzzy logic offers promising approach for addressing different network problems [7] [8]. Applying of fuzzy logic in telecommunication networks is done lately and is proved to be a very economic and efficient method compared with other methods used in automatic control. Recent researches on application of fuzzy logic in telecommunication networks have to do with: packet queuing, buffer management, call acceptance, QoS routing, channel capacity sharing, traffic management etc.
Fuzzy Logic Approach for QoS Routing Analysis 151
MPLS supports traffic engineering for QoS provision and traffic prioritization, for example: provision of wider bandwidth and lower delays for gold customers who are able to pay more for better quality services. Another example a lot of paths can be defined through edge points by ensuring lower levels of interferences and backup services in case of any network failure. This is like using routing metrics in IP network to enforce traffic flowing in one or another direction, but in this case MPLS is much more powerful. An important aspect in MPLS is the priority concept of LSP [13]. LSPs can be configured with higher or lower priority. LSPs with higher priority have advantages in finding new paths compared with
Figure 1 shows MPLS network and its corresponding elements. The core part represents the MPLS network. MPLS combines the advantages of packet forwarding which is based on layer 2 and routing properties of the 3'd layer. MPLS also offers traffic engineering (TE). TE is process of selecting suitable routes for data transmission on the network, that has to do with efficient use of network resources and improving network performance, thus increasing network revenue and QoS. One of the main goals of TE is efficient and reliable functionality of the network. Also, TE calculates the route from the source to the destination based on different metrics such as channel capacity (bandwidth), delays and other
Routing metrics have a significant role, not just in complexity of route calculation but also in QoS. The use of multiple metrics is able to model the network in a more precise way, but the problem for finding appropriate path can become very complex [9] [10]. In general, there are
Fig. 1. MPLS network
those of lower priority.
administrative requirements.
3 types of metrics:
Multiplicative and
Additive,
Concave.
**4. Routing metrics in MPLS network**
Some problems can occur during multimedia service transmission, therefore it is a good idea to design some control mechanisms for solving such problems. As a result of the complex nature of control mechanisms, more and more is being done in designing intelligent controlled techniques. One of the intelligent controlled techniques that will be part of this chapter is Fuzzy Logic Controller, a technique that is based on fuzzy logic. In this chapter we will use main metrics of MPLS network as input parameters of FLC, and we will try to choose the most appropriate defuzzification method for finding better crisp values in aspect of link utilization, in the output of Fuzzy Logic controller.
In this chapter we will firs explain shortly QoS routing principle, MPLS technology in aspect of QoS routing metrics. Also, here we will give the main attention to the fuzzy logic approach, especially FLC used for QoS routing analysis in MPLS network. In this aspect we will try to find the best defuzzification method for gaining better crisp values for link optimization in MPLS network.
### **2. QoS routing**
The main goal of QoS based routing is to select the most suitable path according to traffic requirements for multimedia applications. Selection of suitable transmission paths is done through routing mechanisms based on existing network resources and QoS requirements. Multimedia applications might suffer degradation in quality in traditional networks such as Internet [9]. This problem can be solved in networks that contain dynamic path creation features with bandwidth-guaranteed and constrained delays [10]. Real–time applications impose strict QoS requirements. These application requirements are expressed by parameters such as acceptable and end-to-end delays, necessary bandwidth and acceptable losses. For example, audio and video transmissions have strict requirements for delay and losses. Wide bandwidth must be guaranteed for high capacity transmission. Real time traffic, video in particular, quite often utilizes most important quantities of network resources**.** Efficient management of network resources will reduce network service cost and will allow more applications to be transmitted simultaneously. The task of finding suitable paths through networks is treated by routing protocols. Since common routing protocols are reaching their acceptable complexity limits, it is important that complexity proposed by QoS based routing [11] should not damage scalability of routing protocols. MPLS is a multiple solution for a lot of current problems faced by Internet [12]. By a wide support for QoS and traffic engineering, MPLS is establishing itself as a standard of the next generation's network.
### **3. MPLS network**
MPLS is a data transmission technology which includes some features of circuit switched networks through packet switched network. MPLS actually works at both Layer 2 and Layer 3 in OSI model and it is often referred to as a Layer 2.5 technology. It is designed to provide transport possibilities of data for all users. MPLS techniques can be used as a more efficient tool for traffic engineering than standard routing in IP networks. Also, MPLS can be used for path control of traffic flow, in order to utilize network resources in an optimal way. Network paths can be defined for sensitive traffic, high security traffic etc, guaranteeing different CoS (Class of Service) and QoS. Main MPLS feature is virtual circuit configuration through IP network. These virtual circuits are called LSP.
Fig. 1. MPLS network
150 Fuzzy Logic – Algorithms, Techniques and Implementations
Some problems can occur during multimedia service transmission, therefore it is a good idea to design some control mechanisms for solving such problems. As a result of the complex nature of control mechanisms, more and more is being done in designing intelligent controlled techniques. One of the intelligent controlled techniques that will be part of this chapter is Fuzzy Logic Controller, a technique that is based on fuzzy logic. In this chapter we will use main metrics of MPLS network as input parameters of FLC, and we will try to choose the most appropriate defuzzification method for finding better crisp
In this chapter we will firs explain shortly QoS routing principle, MPLS technology in aspect of QoS routing metrics. Also, here we will give the main attention to the fuzzy logic approach, especially FLC used for QoS routing analysis in MPLS network. In this aspect we will try to find the best defuzzification method for gaining better crisp values for link
The main goal of QoS based routing is to select the most suitable path according to traffic requirements for multimedia applications. Selection of suitable transmission paths is done through routing mechanisms based on existing network resources and QoS requirements. Multimedia applications might suffer degradation in quality in traditional networks such as Internet [9]. This problem can be solved in networks that contain dynamic path creation features with bandwidth-guaranteed and constrained delays [10]. Real–time applications impose strict QoS requirements. These application requirements are expressed by parameters such as acceptable and end-to-end delays, necessary bandwidth and acceptable losses. For example, audio and video transmissions have strict requirements for delay and losses. Wide bandwidth must be guaranteed for high capacity transmission. Real time traffic, video in particular, quite often utilizes most important quantities of network resources**.** Efficient management of network resources will reduce network service cost and will allow more applications to be transmitted simultaneously. The task of finding suitable paths through networks is treated by routing protocols. Since common routing protocols are reaching their acceptable complexity limits, it is important that complexity proposed by QoS based routing [11] should not damage scalability of routing protocols. MPLS is a multiple solution for a lot of current problems faced by Internet [12]. By a wide support for QoS and traffic engineering,
MPLS is a data transmission technology which includes some features of circuit switched networks through packet switched network. MPLS actually works at both Layer 2 and Layer 3 in OSI model and it is often referred to as a Layer 2.5 technology. It is designed to provide transport possibilities of data for all users. MPLS techniques can be used as a more efficient tool for traffic engineering than standard routing in IP networks. Also, MPLS can be used for path control of traffic flow, in order to utilize network resources in an optimal way. Network paths can be defined for sensitive traffic, high security traffic etc, guaranteeing different CoS (Class of Service) and QoS. Main MPLS feature is virtual circuit configuration
values in aspect of link utilization, in the output of Fuzzy Logic controller.
MPLS is establishing itself as a standard of the next generation's network.
through IP network. These virtual circuits are called LSP.
optimization in MPLS network.
**2. QoS routing**
**3. MPLS network**
MPLS supports traffic engineering for QoS provision and traffic prioritization, for example: provision of wider bandwidth and lower delays for gold customers who are able to pay more for better quality services. Another example a lot of paths can be defined through edge points by ensuring lower levels of interferences and backup services in case of any network failure. This is like using routing metrics in IP network to enforce traffic flowing in one or another direction, but in this case MPLS is much more powerful. An important aspect in MPLS is the priority concept of LSP [13]. LSPs can be configured with higher or lower priority. LSPs with higher priority have advantages in finding new paths compared with those of lower priority.
Figure 1 shows MPLS network and its corresponding elements. The core part represents the MPLS network. MPLS combines the advantages of packet forwarding which is based on layer 2 and routing properties of the 3'd layer. MPLS also offers traffic engineering (TE). TE is process of selecting suitable routes for data transmission on the network, that has to do with efficient use of network resources and improving network performance, thus increasing network revenue and QoS. One of the main goals of TE is efficient and reliable functionality of the network. Also, TE calculates the route from the source to the destination based on different metrics such as channel capacity (bandwidth), delays and other administrative requirements.
### **4. Routing metrics in MPLS network**
Routing metrics have a significant role, not just in complexity of route calculation but also in QoS. The use of multiple metrics is able to model the network in a more precise way, but the problem for finding appropriate path can become very complex [9] [10]. In general, there are 3 types of metrics:
Fuzzy Logic Approach for QoS Routing Analysis 153
When the control processes are too complex to analyze by conventional quantitative
When the available sources of information are interpreted qualitatively or uncertainly. Fuzzy logic controller consists of: fuzzifier, rule base, fuzzy inference and defuzzifier (see
**Fuzzifier**: A fuzzifier operator has the effect of transforming crisp value to fuzzy sets. Fuzzifier is presented with *x*=*fuzzifier*(*x*0), where *x*0 is input crisp value; *x* is a fuzzy set and
**Rule-Base (Linguistic Rules): C**ontains IF-THEN rules that are determined through fuzzy
*Example*: if *x* is *Ai* and *Y* is *Bi* the *Z* is *Ci,* Where *x* and *y* are inputs and *z* is controlled output;
**Fuzzy Inference:** Is a process of converting input values into output values using fuzzy logic. Converting is essential for decision making. Fuzzy Inference process includes:
**Defuzzifier:** can be expressed by: *y*ou=*defuzzifier*(*y*), where *y* identifies fuzzy controller action, *y*ou identifies crisp value of control action and defuzzifier presents defuzzifier operator. Converting process of fuzzy terms in crisp values is called defuzzification. There are some defuzzification methods: COG (Centre of Gravity), COGS (Centre of Gravity for Singletons), COA (Centre of Area), LM (Left Most Maximum) and RM (Right Most
For solving QoS routing problem, we will use fuzzy logic approach. Fuzzy logic is proved to be very effective in a lot of applications, such as intelligent control, decision making process etc. Fuzzy logic is based in a set of metrics which can be or not connected with each other. Calculating the best route cannot be done using complex mathematical solutions, but is
techniques
Fig. 2. Fuzzy Logic Controller
fuzzifier represents a fuzzification operator.
membership functions and logic operations
Ai, Bi and Ci are linguistic terms, like: low, medium, high etc.
**7. MPLS network metrics and membership functions**
Figure 2).
logic.
Maximum).
based in intuitive rules.
They are defined as below:
If *m* (*n*1, *n*2) are metrics for link (n1, n2). For one path *P* = (*n*1, *n*2, …, *n*i, *n*j), metric *m* is (*n*1, *n*2, …, *n*i, *n*j):
For any path *p* = (*i, j, k, …, l, m*), we say metric d is additive if:
$$d\mathbf{d}\text{ (}p) = d(\mathbf{i}, \mathbf{j}) + d\text{ (}\mathbf{j}, \mathbf{k}) + \dots + d(\mathbf{l}, \mathbf{m})\tag{1}$$
$$d \text{ (\$p\$)} \equiv d(\text{l, j}) \ge d \text{ (\$j, k)} \ge \dots \ge d(\text{l, m}) \tag{2}$$
$$d(p) = \min\{d(\mathbf{i}, \mathbf{j}), \ d(\mathbf{j}, \mathbf{k}), \ \dots \ d(\mathbf{l}, \mathbf{m})\}\tag{3}$$
In MPLS network there are a lot of metrics that we can take into consideration, but in this chapter, for sake of simplicity, we will consider three main metrics: delay, losses and bandwidth. Those metrics play a direct role in quality of service in MPLS network. In order to consider multiple metrics simultaneously, we will use fuzzy logic controller. FLC is intelligent technique that can manipulate with two or more input parameters simultaneously without any problem.
### **5. Soft computing**
Soft Computing is more tolerable in uncertainty and partial truth than Hard Computing. The model in which soft computing is based in human mind. The main components of soft computing are: Fuzzy Logic, Neural Networks, Probabilistic reasoning and Genetic algorithms. The most important component of soft computing is Fuzzy logic, which will be part of this chapter. Fuzzy logic will be used for a lot of applications. Applications of fuzzy logic in telecommunications networks are recent. Fuzzy Logic is organized into three main efforts: modeling and control, management and forecasting, and performance estimation.
### **5.1 Fuzzy logic**
Idea for fuzzy logic has born in 1965. Lotfi Zadeh has published one seminar for fuzzy which was the beginning for fuzzy logic [14]. Fuzzy logic is tolerant in imprecise data, nonlinear functions and can be mixed with other techniques for different problems solving. The main principle of fuzzy logic is using fuzzy groups which are without crisp boundaries.
### **6. QoS routing analysis using FLC – Fuzzy Logic Controller**
As we have mentioned above, for QoS routing analysis we will use FLC as intelligent controlling technique. A Fuzzy Logic Controller [15] is a rule based system in which fuzzy rule represents a control mechanism. In this case, a fuzzy controller uses fuzzy logic to simulate human thinking. In particular the FLC is useful in two special cases [15]:
152 Fuzzy Logic – Algorithms, Techniques and Implementations
If *m* (*n*1, *n*2) are metrics for link (n1, n2). For one path *P* = (*n*1, *n*2, …, *n*i, *n*j), metric *m* is (*n*1, *n*2,
In MPLS network there are a lot of metrics that we can take into consideration, but in this chapter, for sake of simplicity, we will consider three main metrics: delay, losses and bandwidth. Those metrics play a direct role in quality of service in MPLS network. In order to consider multiple metrics simultaneously, we will use fuzzy logic controller. FLC is intelligent technique that can manipulate with two or more input parameters
Soft Computing is more tolerable in uncertainty and partial truth than Hard Computing. The model in which soft computing is based in human mind. The main components of soft computing are: Fuzzy Logic, Neural Networks, Probabilistic reasoning and Genetic algorithms. The most important component of soft computing is Fuzzy logic, which will be part of this chapter. Fuzzy logic will be used for a lot of applications. Applications of fuzzy logic in telecommunications networks are recent. Fuzzy Logic is organized into three main efforts: modeling and control, management and forecasting, and performance
Idea for fuzzy logic has born in 1965. Lotfi Zadeh has published one seminar for fuzzy which was the beginning for fuzzy logic [14]. Fuzzy logic is tolerant in imprecise data, nonlinear functions and can be mixed with other techniques for different problems solving. The main principle of fuzzy logic is using fuzzy groups which are without crisp boundaries.
As we have mentioned above, for QoS routing analysis we will use FLC as intelligent controlling technique. A Fuzzy Logic Controller [15] is a rule based system in which fuzzy rule represents a control mechanism. In this case, a fuzzy controller uses fuzzy logic to
simulate human thinking. In particular the FLC is useful in two special cases [15]:
**6. QoS routing analysis using FLC – Fuzzy Logic Controller**
*d (p) = d(i, j) + d (j,k) + … + d(l, m)* (1)
*d (p) = d(i, j) x d (j,k) x … x d(l, m)* (2)
*d(p) = min[d(i, j), d (j,k), … d(l, m)]* (3)
They are defined as below:
simultaneously without any problem.
**5. Soft computing**
estimation.
**5.1 Fuzzy logic**
For any path *p* = (*i, j, k, …, l, m*), we say metric d is additive if:
…, *n*i, *n*j):
Fuzzy logic controller consists of: fuzzifier, rule base, fuzzy inference and defuzzifier (see Figure 2).
Fig. 2. Fuzzy Logic Controller
**Fuzzifier**: A fuzzifier operator has the effect of transforming crisp value to fuzzy sets. Fuzzifier is presented with *x*=*fuzzifier*(*x*0), where *x*0 is input crisp value; *x* is a fuzzy set and fuzzifier represents a fuzzification operator.
**Rule-Base (Linguistic Rules): C**ontains IF-THEN rules that are determined through fuzzy logic.
*Example*: if *x* is *Ai* and *Y* is *Bi* the *Z* is *Ci,* Where *x* and *y* are inputs and *z* is controlled output; Ai, Bi and Ci are linguistic terms, like: low, medium, high etc.
**Fuzzy Inference:** Is a process of converting input values into output values using fuzzy logic. Converting is essential for decision making. Fuzzy Inference process includes: membership functions and logic operations
**Defuzzifier:** can be expressed by: *y*ou=*defuzzifier*(*y*), where *y* identifies fuzzy controller action, *y*ou identifies crisp value of control action and defuzzifier presents defuzzifier operator. Converting process of fuzzy terms in crisp values is called defuzzification. There are some defuzzification methods: COG (Centre of Gravity), COGS (Centre of Gravity for Singletons), COA (Centre of Area), LM (Left Most Maximum) and RM (Right Most Maximum).
### **7. MPLS network metrics and membership functions**
For solving QoS routing problem, we will use fuzzy logic approach. Fuzzy logic is proved to be very effective in a lot of applications, such as intelligent control, decision making process etc. Fuzzy logic is based in a set of metrics which can be or not connected with each other. Calculating the best route cannot be done using complex mathematical solutions, but is based in intuitive rules.
Fuzzy Logic Approach for QoS Routing Analysis 155
For losses we will use 3 membership functions. For two membership functions we will use triangular form (ACCEPTABLE and TOLERABLE), while for one of membership functions
In table x are given details about metrics in the input of the fuzzy system and fuzzy sets
**Fuzzy sets**
Delays {ACCEPTABLE, TOLERABLE, INTOLERABLE}- *ms* Losses { ACCEPTABLE, TOLERABLE, INTOLERABLE }-%
Mathematical relations of 3 input parameters can be given by the below expression:
Where *p* is the calculated path, *B(p)* is channel capacity, *D(p)* is the delay among
Packet switching networks commonly are used for transmission of multimedia services. This trend continued in MPLS network also. Real time traffic is sensitive to delays, such as: voice, video etc, constitute an important part of real time traffic. Such traffic has more strict requirements for quality of service (QoS), especially in the aspect of delays between two end points and packet losses. The table below shows standard requirements for QoS for
**Maximum rate Average rate Probability of packet loss**
*B p f p <sup>D</sup> <sup>p</sup> xL <sup>p</sup>* (4)
Channel capacity (bandwidth) {LOW, MEDIUM, HIGH}-*Mbps*
**c. Losses (L)**
we will use rectangular form (INTOLERABLE).
Fig. 4. Fuzzy triangular number for delay-D
Table 1. Input parameters and fuzzy sets
transmission, and *L(p)* is the probability of packet loss.
**8. Limits of fuzzy sets for MPLS network parameters**
Voice 32 KBits/sec 11.2 KBits/sec 0.05 Voice 11.6 MBits/sec 3.85 MBits/sec 10\*\*(-5)
corresponding such inputs.
**Input parameters for MPLS**
multimedia services.
Table 2. Bit rate for voice transmission
**network**
Fuzzy logic applies to all those routes that are candidates for being chosen whereas the chosen path in this way is the path that has better quality. In this chapter we will use the Fuzzy logic controller for solving QoS routing problems and the routing algorithm refers to the fuzzy logic (fuzzy routing algorithm). This algorithm is able to choose the path with better transmission parameters. For solving such a problem using fuzzy logic, first we have to take into consideration some input parameters, acting at the entrance of FLC, which in our case must be the MPLS network metrics. These input variables can be crisp or fuzzy values. Whereas the main disadvantageous of MPLS network consist in: losses, delay and bandwidth, then these three metrics will be taken as network parameters. These metrics match with the main factors which affect for choosing the best route for transmission of multimedia services. Each network metric has the value from 0 to 1. These metrics are:
### **a. Channel capacity (B)**
Channel capacity is one of the main MPLS network parameters. In this chapter, channel capacity is combined with linguistic data for selecting the optimal route from the source to destination. In this particular case we took three membership functions that indicate the potential scale of the channel capacity: LOW, MEDIUM and HIGH. Channel capacity is presented by triangular membership functions. Figure x shows the membership function for the channel capacity. Triangular number *B* = (*b*1, *b*2, *b*3) in limited in his left hand with value b1 and in his right hand with the value *b*3. In this way, decision taker can calculate that channel capacity in a certain link cannot be smaller than *b*1 or greater than *b*2. Figure 3 shows the fuzzy set of linguistic data for channel capacity.
Fig. 3. Fuzzy Triangular number for channel capacity - B
It can be seen from the figure x that each value has the upper and lower limit.
### **b. Delays (D)**
For most applications, especially real time applications, the delay for transmission of information between two points is one of the most important parameters for meeting QoS requirements. For delay we set 3 membership functions in triangular form to show the potential scale of the delays: ACCEPTABLE, TOLERABLE and INTOLERABLE. Figure 4 represents the membership function *s s D* for the delay. From this function it is indicated that the greatest value of membership is (=1) for *D d s s* .
### **c. Losses (L)**
154 Fuzzy Logic – Algorithms, Techniques and Implementations
Fuzzy logic applies to all those routes that are candidates for being chosen whereas the chosen path in this way is the path that has better quality. In this chapter we will use the Fuzzy logic controller for solving QoS routing problems and the routing algorithm refers to the fuzzy logic (fuzzy routing algorithm). This algorithm is able to choose the path with better transmission parameters. For solving such a problem using fuzzy logic, first we have to take into consideration some input parameters, acting at the entrance of FLC, which in our case must be the MPLS network metrics. These input variables can be crisp or fuzzy values. Whereas the main disadvantageous of MPLS network consist in: losses, delay and bandwidth, then these three metrics will be taken as network parameters. These metrics match with the main factors which affect for choosing the best route for transmission of multimedia services. Each network metric has the value from 0 to 1. These metrics are:
Channel capacity is one of the main MPLS network parameters. In this chapter, channel capacity is combined with linguistic data for selecting the optimal route from the source to destination. In this particular case we took three membership functions that indicate the potential scale of the channel capacity: LOW, MEDIUM and HIGH. Channel capacity is presented by triangular membership functions. Figure x shows the membership function for the channel capacity. Triangular number *B* = (*b*1, *b*2, *b*3) in limited in his left hand with value b1 and in his right hand with the value *b*3. In this way, decision taker can calculate that channel capacity in a certain link cannot be smaller than *b*1 or greater than *b*2. Figure 3 shows
**a. Channel capacity (B)**
**b. Delays (D)**
represents the membership function
the fuzzy set of linguistic data for channel capacity.
Fig. 3. Fuzzy Triangular number for channel capacity - B
that the greatest value of membership is (=1) for *D d s s* .
It can be seen from the figure x that each value has the upper and lower limit.
For most applications, especially real time applications, the delay for transmission of information between two points is one of the most important parameters for meeting QoS requirements. For delay we set 3 membership functions in triangular form to show the potential scale of the delays: ACCEPTABLE, TOLERABLE and INTOLERABLE. Figure 4
*s s D* for the delay. From this function it is indicated
For losses we will use 3 membership functions. For two membership functions we will use triangular form (ACCEPTABLE and TOLERABLE), while for one of membership functions we will use rectangular form (INTOLERABLE).
Fig. 4. Fuzzy triangular number for delay-D
In table x are given details about metrics in the input of the fuzzy system and fuzzy sets corresponding such inputs.
Table 1. Input parameters and fuzzy sets
Mathematical relations of 3 input parameters can be given by the below expression:
$$f\left(p\right) = \frac{B\left(p\right)}{D\left(p\right)\ge L\left(p\right)}\tag{4}$$
Where *p* is the calculated path, *B(p)* is channel capacity, *D(p)* is the delay among transmission, and *L(p)* is the probability of packet loss.
### **8. Limits of fuzzy sets for MPLS network parameters**
Packet switching networks commonly are used for transmission of multimedia services. This trend continued in MPLS network also. Real time traffic is sensitive to delays, such as: voice, video etc, constitute an important part of real time traffic. Such traffic has more strict requirements for quality of service (QoS), especially in the aspect of delays between two end points and packet losses. The table below shows standard requirements for QoS for multimedia services.
Table 2. Bit rate for voice transmission
Fuzzy Logic Approach for QoS Routing Analysis 157
Membership function for channel capacity, delays, losses and the output of the fuzzy system will be seen in the following figure (where the values are taken for real tie applications). Based on the above limits for fuzzy sets, using Matlab software we can create the
membership functions main parameters of the MPLS network.
Fig. 5. Fuzzy system and its integral components in MATLAB software
[System]
[Input1]
Name='bandwidth' Range=[0 1000] NumMFs=3
Name='fuzzy\_link' Type='mamdani' Version=2.0 NumInputs=3 NumOutputs=1 NumRules=4 AndMethod='min' OrMethod='max' ImpMethod='min' AggMethod='max'
DefuzzMethod='Centre of Gravity'
According to ITU recommendations for delay, packet loss and channel capacity, we have defined the boundaries of fuzzy sets.
Table 3. ITU recommendation for delays
### **Delays**
Delays up to 150 ms are acceptable Delays between 150 and 400 ms are tolerable for special applications Delays higher than 400 ms are intolerable
### **Packet loss percentage**
Lower than 2% - acceptable From 2 – 6% - tolerable Higher than 6 % - intolerable
### **Channel capacity:**
Low: from 0 *Mbps* to 200 *Mbps.* Medium: from 180 *Mbps* to 500 *Mbps.* High: from 470 *Mbps* to 1000 *Mbps.*
### **9. Fuzzy logic toolbox in Matlab software**
Fuzzy logic tool in Matlab is used for solving different problems dealing with Fuzzy Logic. Fuzzy logic is a very valuable tool for planning because it makes a very good for problems that have high importance and require high precision – something that human beings have done long time ago. Fuzzy logic tool allows users to do important jobs, but the most important thing is to allow users to create fuzzy conclusions (fuzzy inference). It is also possible to use fuzzy logic tool through command line, but in general it is easier to build a system through the GUI. There are five primary GUI tools for building, editing and reviewing systems in fuzzy logic toolbox:
The interactions of these tools can be seen in the figure below (Figure 5).
Rule viewer and surface viewer are used for survey, compared with FIS editor, which is used for editing. So, they are read-only tools. These GUI dynamically are connected with each other and changes in FIS can be seen in other open GUIs.
156 Fuzzy Logic – Algorithms, Techniques and Implementations
According to ITU recommendations for delay, packet loss and channel capacity, we have
Fuzzy logic tool in Matlab is used for solving different problems dealing with Fuzzy Logic. Fuzzy logic is a very valuable tool for planning because it makes a very good for problems that have high importance and require high precision – something that human beings have done long time ago. Fuzzy logic tool allows users to do important jobs, but the most important thing is to allow users to create fuzzy conclusions (fuzzy inference). It is also possible to use fuzzy logic tool through command line, but in general it is easier to build a system through the GUI. There are five primary GUI tools for building, editing and
Rule viewer and surface viewer are used for survey, compared with FIS editor, which is used for editing. So, they are read-only tools. These GUI dynamically are connected with
The interactions of these tools can be seen in the figure below (Figure 5).
each other and changes in FIS can be seen in other open GUIs.
**Delays in one direction Characterization of quality**
Delays between 150 and 400 ms are tolerable for special applications
0 to 150 *ms* Acceptable for most of applications 150 to 400 *ms* May impact in some applications Above 400 *ms* Unacceptable for most of applications
defined the boundaries of fuzzy sets.
Table 3. ITU recommendation for delays
Delays higher than 400 ms are intolerable
**9. Fuzzy logic toolbox in Matlab software**
reviewing systems in fuzzy logic toolbox: 1. FIS (Fuzzy Inference System) editor 2. Membership function editor
Delays up to 150 ms are acceptable
**Packet loss percentage**
**Channel capacity:**
3. Rules editor 4. Rule Viewer 5. Surface viewer
Lower than 2% - acceptable From 2 – 6% - tolerable Higher than 6 % - intolerable
Low: from 0 *Mbps* to 200 *Mbps.* Medium: from 180 *Mbps* to 500 *Mbps.* High: from 470 *Mbps* to 1000 *Mbps.*
**Delays**
Membership function for channel capacity, delays, losses and the output of the fuzzy system will be seen in the following figure (where the values are taken for real tie applications). Based on the above limits for fuzzy sets, using Matlab software we can create the membership functions main parameters of the MPLS network.
Fig. 5. Fuzzy system and its integral components in MATLAB software
[System] Name='fuzzy\_link' Type='mamdani' Version=2.0 NumInputs=3 NumOutputs=1 NumRules=4 AndMethod='min' OrMethod='max' ImpMethod='min' AggMethod='max' DefuzzMethod='Centre of Gravity'
[Input1] Name='bandwidth' Range=[0 1000] NumMFs=3
Fuzzy Logic Approach for QoS Routing Analysis 159
b) Delays in **ms**
c) Packet loss percentage
d) Output of fuzzy system
Fig. 6**.** a) Membership function of channel capacity, b) delays, c) packet loss percentage and
d) output of fuzzy system
MF1='low':'trimf',[-400 0 400] MF2='medium':'trimf',[100 500 900] MF3='high':'trimf',[600 1000 1400] [Input2] Name='delays' Range=[0 600] NumMFs=3 MF1='acceptable':'trimf',[-240 0 240] MF2='tolerable':'trimf',[60 300 540] MF3='highintolerable':'trimf',[360 600 840] [Input3] Name='losses'
Range=[0 10] NumMFs=3 MF1='acceptable':'trimf',[-4 0 4] MF2='tolerable':'trimf',[1 5 9] MF3='intolerable':'trimf',[6 10 14] [Output1] Name='link optimization' Range=[0 100] NumMFs=3 MF1='low':'trimf',[-40 0 40] MF2='medium':'trimf',[10 50 90] MF3='high':'trimf',[60 100 140]
### [Rules]
1 1 1, 1 (1) : 1 2 1 1, 2 (1) : 1 3 1 1, 3 (1) : 1 3 2 3, 1 (1) : 1
158 Fuzzy Logic – Algorithms, Techniques and Implementations
a) Channel capacity (bandwidth) in **Mbps**
MF1='low':'trimf',[-400 0 400] MF2='medium':'trimf',[100 500 900] MF3='high':'trimf',[600 1000 1400]
MF1='acceptable':'trimf',[-240 0 240] MF2='tolerable':'trimf',[60 300 540]
MF1='acceptable':'trimf',[-4 0 4] MF2='tolerable':'trimf',[1 5 9] MF3='intolerable':'trimf',[6 10 14]
Name='link optimization'
MF1='low':'trimf',[-40 0 40] MF2='medium':'trimf',[10 50 90] MF3='high':'trimf',[60 100 140]
MF3='highintolerable':'trimf',[360 600 840]
[Rules] 1 1 1, 1 (1) : 1 2 1 1, 2 (1) : 1 3 1 1, 3 (1) : 1 3 2 3, 1 (1) : 1
[Input2] Name='delays' Range=[0 600] NumMFs=3
[Input3] Name='losses' Range=[0 10] NumMFs=3
[Output1]
Range=[0 100] NumMFs=3
Fig. 6**.** a) Membership function of channel capacity, b) delays, c) packet loss percentage and d) output of fuzzy system
Fuzzy Logic Approach for QoS Routing Analysis 161
*C L* (7)
*C L* are membership functions for channel capacity, delays
Fig. 7. The structure of the fuzzy controller system for MPLS network analysis
*<sup>O</sup> L* - Membership function for output of the fuzzy system.
operator, which graphically will look like in figure below.
gaining more accurate values in the output.
The role of Linguistic rules is to connect these input parameters with output of fuzzy system, which in our case is link optimization. The output comprises from three membership functions: LOW, MEDIUM and HIGH. Each rule determines one fuzzy relation. In our case, each rule represents the relation between 3 input parameters and
Fuzzy controller considered here is Mamdani type and consists of: Fuzzifier, fuzzy
It is well known that in some cases the output of fuzzy process needs to be a single scalar value. Defuzzification is the process of converting the fuzzy quantity to a precise value. The output of a fuzzy process can be the union of two or more fuzzy membership functions. To see this better, we will take into consideration one example. Let suppose a fuzzy output comprises from two parts: 1) Triangular membership shape and 2) Triangular membership shape. The union of these two membership functions means that we will use the max
There are a lot of methods that have been proposed recently as defuzzification methods. We will explain shortly each of these methods and we will analyze which one is the best for
inference, linguistic rules and defuzzifier. This fuzzy controller is shown in figure x.
Where *C Acce ptable Tolerable Intolerable* , ,
*B D* and
required output that is written as:
Where *O Low Medium Hi* , , *gh*
**11. Defuzzification process**
While,
and losses.
*A B* ,
With this simple program are created the membership functions for abovementioned parameters. These membership functions are seen in the figure 6 (a, b, c and d)
Once the variable and membership functions are assigned, fuzzy rules can be written for corresponding variables.
### **Some of the fuzzy rules derived from Rule editor (Matlab) are listed as below:**
*Rule 1: If (bandwidth is low) and (delay is acceptable) and (loss is acceptable) then (link optimization is Low)*
*Rule 2: If (bandwidth is Medium) and (delay is acceptable) and (loss is acceptable) then (link optimization is Medium)*
*Rule 3: If (bandwidth is High) and (delay is Acceptable) and (loss is Acceptable) then (link optimization is High)*
*Rule 4: If (bandwidth is Low) and (delay is Tolerable) and (loss is Acceptable) then (link optimization is Low)*
*Rule 5: If (bandwidth is Medium) and (delay is Tolerable) and (loss is Tolerable) then (link optimization is Medium)*
*Rule 6: If (bandwidth is High) and (delay is Tolerable) and (loss is Intolerable) then (link optimization is Medium)*
*Rule 7: If (bandwidth is Low) and (delay is Intolerable) and (loss is Acceptable) then (link optimization is Low)*
*Rule 8: If (bandwidth is Medium) and (delay is Intolerable) and (loss is Tolerable) then (link optimization is Low)*
*Rule 9: If (bandwidth is High) and (delay is Intolerable) and (loss is Intolerable) then (link optimization is Low)*
### **10. Fuzzy relations for QoS routing analysis**
Here we will illustrate the relation between input fuzzy value and required output. The figure below shows the structure of the proposed solution. So, in the figure is shown fuzzy controller comprising of: inputs (which react in the input), fuzzy rules and outputs. Parameters acting on the input of this controller are: channel capacity (bandwidth), delays and losses.
Three input parameters are noted as:
$$A\begin{pmatrix}B\\ \end{pmatrix} \tag{5}$$
Where *A Low Medium Hi* , , *gh*
$$
\mu \text{B} (D) \tag{6}
$$
Where *B Acce ptable Tolerable Intolerable* , ,
160 Fuzzy Logic – Algorithms, Techniques and Implementations
With this simple program are created the membership functions for abovementioned
Once the variable and membership functions are assigned, fuzzy rules can be written for
*Rule 1: If (bandwidth is low) and (delay is acceptable) and (loss is acceptable) then (link optimization*
*Rule 2: If (bandwidth is Medium) and (delay is acceptable) and (loss is acceptable) then (link*
*Rule 3: If (bandwidth is High) and (delay is Acceptable) and (loss is Acceptable) then (link*
*Rule 4: If (bandwidth is Low) and (delay is Tolerable) and (loss is Acceptable) then (link optimization*
*Rule 5: If (bandwidth is Medium) and (delay is Tolerable) and (loss is Tolerable) then (link*
*Rule 6: If (bandwidth is High) and (delay is Tolerable) and (loss is Intolerable) then (link*
*Rule 7: If (bandwidth is Low) and (delay is Intolerable) and (loss is Acceptable) then (link*
*Rule 8: If (bandwidth is Medium) and (delay is Intolerable) and (loss is Tolerable) then (link*
*Rule 9: If (bandwidth is High) and (delay is Intolerable) and (loss is Intolerable) then (link*
Here we will illustrate the relation between input fuzzy value and required output. The figure below shows the structure of the proposed solution. So, in the figure is shown fuzzy controller comprising of: inputs (which react in the input), fuzzy rules and outputs. Parameters acting on the input of this controller are: channel capacity (bandwidth), delays
*A B* (5)
*B D* (6)
parameters. These membership functions are seen in the figure 6 (a, b, c and d)
**Some of the fuzzy rules derived from Rule editor (Matlab) are listed as below:**
corresponding variables.
*optimization is Medium)*
*optimization is Medium)*
*optimization is Medium)*
*optimization is Low)*
*optimization is Low)*
*optimization is Low)*
and losses.
**10. Fuzzy relations for QoS routing analysis**
Three input parameters are noted as:
Where *B Acce ptable Tolerable Intolerable* , ,
Where *A Low Medium Hi* , , *gh*
*optimization is High)*
*is Low)*
*is Low)*
$$
\mu \mathbb{C} \begin{pmatrix} \mathsf{L} \end{pmatrix} \tag{7}
$$
Where *C Acce ptable Tolerable Intolerable* , ,
While, *A B* , *B D* and *C L* are membership functions for channel capacity, delays and losses.
Fig. 7. The structure of the fuzzy controller system for MPLS network analysis
The role of Linguistic rules is to connect these input parameters with output of fuzzy system, which in our case is link optimization. The output comprises from three membership functions: LOW, MEDIUM and HIGH. Each rule determines one fuzzy relation. In our case, each rule represents the relation between 3 input parameters and required output that is written as:
*<sup>O</sup> L* - Membership function for output of the fuzzy system.
Where *O Low Medium Hi* , , *gh*
Fuzzy controller considered here is Mamdani type and consists of: Fuzzifier, fuzzy inference, linguistic rules and defuzzifier. This fuzzy controller is shown in figure x.
### **11. Defuzzification process**
It is well known that in some cases the output of fuzzy process needs to be a single scalar value. Defuzzification is the process of converting the fuzzy quantity to a precise value. The output of a fuzzy process can be the union of two or more fuzzy membership functions. To see this better, we will take into consideration one example. Let suppose a fuzzy output comprises from two parts: 1) Triangular membership shape and 2) Triangular membership shape. The union of these two membership functions means that we will use the max operator, which graphically will look like in figure below.
There are a lot of methods that have been proposed recently as defuzzification methods. We will explain shortly each of these methods and we will analyze which one is the best for gaining more accurate values in the output.
Fuzzy Logic Approach for QoS Routing Analysis 163
*Max*
*Min Max*
*U*
Where *U* is defuzzification result, *u* = *output variable*,
determines the crisp value after defuzzification.
Fig. 9. COG method
**12.2 Bisectorial method**
in figure below (Figure 10).
**12.3 Middle, smallest and largest of maximum methods**
determines Middle of maximum value in that zone.
*limit for defuzzification*, *Max*=*maximum limit for defuzzification*
*Min*
*u u du*
With formula (1) we can calculate the surface of zone that is shown in figure below and also we can find one central point in this zone. Projecting this point in the abscissa axis
This method divides a certain zone into two equal regions by a vertical line. This can be seen
In some cases MOM and LOM methods are better than COG method, but in general, for the most of cases, no matter what zone we will have, COG method shows better results. In this chapter we will analyze which method is better with MATLAB software in 3 D. LOM method determines the largest of maximum value in the zone that is obtained from membership functions with AND and OR logic operators whereas MOM method
*u du*
(8)
*membership function* , *Min*=*minimum*
Fig. 8. a) Triangular membership shape, b) Triangular membership shape c) The union of these two membership (a and b)
### **12. Selection of defuzzification method for finding crisp value for link optimization**
For finding the appropriate path for transmission of multimedia services one important role plays selection of defuzzification method. Using fuzzy logic technique, the "not accurate" data are presented by linguistic values which depend on user preferences.
There are 5 defuzzification methods: Centre of Gravity (COG), bisectorial, LOM (largest of maximum), MOM (middle of maximum) and SOM (smallest of maximum). Three most important methods are: COG, MOM and LOM. It is important to find which method gives better results in aspect of link optimization in MPLS network. To see which method is most suitable for defuzzification, first we will explain shortly each abovementioned method .
### **12.1 Centre of gravity**
This method determines the centre of zone that is gained from membership functions with AND and OR logic operators. Formula with which we can calculate the defuzzified crisp output *U* is given:
162 Fuzzy Logic – Algorithms, Techniques and Implementations
Fig. 8. a) Triangular membership shape, b) Triangular membership shape c) The union of
For finding the appropriate path for transmission of multimedia services one important role plays selection of defuzzification method. Using fuzzy logic technique, the "not accurate"
There are 5 defuzzification methods: Centre of Gravity (COG), bisectorial, LOM (largest of maximum), MOM (middle of maximum) and SOM (smallest of maximum). Three most important methods are: COG, MOM and LOM. It is important to find which method gives better results in aspect of link optimization in MPLS network. To see which method is most suitable for defuzzification, first we will explain shortly each abovementioned method .
This method determines the centre of zone that is gained from membership functions with AND and OR logic operators. Formula with which we can calculate the defuzzified crisp
**12. Selection of defuzzification method for finding crisp value for link**
data are presented by linguistic values which depend on user preferences.
a) b)
c)
these two membership (a and b)
**optimization**
**12.1 Centre of gravity**
output *U* is given:
$$\begin{aligned} \int\_{M} u \, \mu\left(u\right) du\\ \mathcal{U} = \frac{\text{Min}}{\text{Max}}\\ \int\_{\text{Min}} \mu(u) du \end{aligned} \tag{8}$$
Where *U* is defuzzification result, *u* = *output variable*, *membership function* , *Min*=*minimum limit for defuzzification*, *Max*=*maximum limit for defuzzification*
With formula (1) we can calculate the surface of zone that is shown in figure below and also we can find one central point in this zone. Projecting this point in the abscissa axis determines the crisp value after defuzzification.
Fig. 9. COG method
### **12.2 Bisectorial method**
This method divides a certain zone into two equal regions by a vertical line. This can be seen in figure below (Figure 10).
### **12.3 Middle, smallest and largest of maximum methods**
In some cases MOM and LOM methods are better than COG method, but in general, for the most of cases, no matter what zone we will have, COG method shows better results. In this chapter we will analyze which method is better with MATLAB software in 3 D. LOM method determines the largest of maximum value in the zone that is obtained from membership functions with AND and OR logic operators whereas MOM method determines Middle of maximum value in that zone.
Fuzzy Logic Approach for QoS Routing Analysis 165
Fig. 11. LOM, MOM and SOM defuzzification methods
rules gives us better result in aspect of link utilization. If we use rule number 3 (derived from rule editor):
Fig. 12. Rule viewer when it is used COG method
*optimization is High)*
**13. Analysis and examples for different defuzzification methods**
Here we will use some examples taking different rules derived from rule editor of Matlba's toolbox. Also we will analyze which of the defuzzification method is better and which of the
*If (bandwidth is High) AND (delay is Acceptable) AND (loss is Acceptable) THEN (link*
Whereas, if we take the values for channel capacity (bandwidth) as "high" (946 Mbps), Delays 25.3 ms and losses 0.663 %, then link optimization for MPLS network will be 86.4 %. This can be shown graphically using rule viewer. So, from this figure, it is clearly seen that link optimization is high (according to determination of link optimization using
Fig. 10. Bisectorial Method
These three methods have to do with the maximum value of the sum of membership functions. In this example, since in this graph (figure x) is a flat curve at the maximum point, then these three methods have different values from each other. In case when we have a single maximum point, then three methods have the same value.
Finding defuzzification values using the above-mentioned methods can e done like below:
```
x3 = defuzz(x,mf1,'mom')
x4 = defuzz(x,mf1,'som')
x5 = defuzz(x,mf1,'lom')
set([h2 t2],'Color',gray)
h3 = line([x3 x3],[-0.7 1.2],'Color','k');
t3 = text(x3,-0.7,' MOM','FontWeight','bold');
h4 = line([x4 x4],[-0.8 1.2],'Color','k');
t4 = text(x4,-0.8,' SOM','FontWeight','bold');
h5 = line([x5 x5],[-0.6 1.2],'Color','k');
t5 = text(x5,-0.6,' LOM','FontWeight','bold');
x3 = -5
x4 = -2
x5 = -8
```
These values are represented graphically like in figure 11.
Methods that are used mostly for defuzzification are: COGM MOM and LOM. Although, in some cases MOM and LOM methods give very favorable values, COG method gives better results whatever the case that we are analyzing. The performance comparison offered by three methods can be better seen through examples by surface viewer in 3D.
164 Fuzzy Logic – Algorithms, Techniques and Implementations
These three methods have to do with the maximum value of the sum of membership functions. In this example, since in this graph (figure x) is a flat curve at the maximum point, then these three methods have different values from each other. In case when we have a
Finding defuzzification values using the above-mentioned methods can e done like below:
Methods that are used mostly for defuzzification are: COGM MOM and LOM. Although, in some cases MOM and LOM methods give very favorable values, COG method gives better results whatever the case that we are analyzing. The performance comparison offered by
three methods can be better seen through examples by surface viewer in 3D.
single maximum point, then three methods have the same value.
Fig. 10. Bisectorial Method
x3 = defuzz(x,mf1,'mom') x4 = defuzz(x,mf1,'som') x5 = defuzz(x,mf1,'lom') set([h2 t2],'Color',gray)
x3 = -5 x4 = -2 x5 = -8
h3 = line([x3 x3],[-0.7 1.2],'Color','k');
h4 = line([x4 x4],[-0.8 1.2],'Color','k'); t4 = text(x4,-0.8,' SOM','FontWeight','bold'); h5 = line([x5 x5],[-0.6 1.2],'Color','k'); t5 = text(x5,-0.6,' LOM','FontWeight','bold');
t3 = text(x3,-0.7,' MOM','FontWeight','bold');
These values are represented graphically like in figure 11.
Fig. 11. LOM, MOM and SOM defuzzification methods
### **13. Analysis and examples for different defuzzification methods**
Here we will use some examples taking different rules derived from rule editor of Matlba's toolbox. Also we will analyze which of the defuzzification method is better and which of the rules gives us better result in aspect of link utilization.
If we use rule number 3 (derived from rule editor):
*If (bandwidth is High) AND (delay is Acceptable) AND (loss is Acceptable) THEN (link optimization is High)*
Fig. 12. Rule viewer when it is used COG method
Whereas, if we take the values for channel capacity (bandwidth) as "high" (946 Mbps), Delays 25.3 ms and losses 0.663 %, then link optimization for MPLS network will be 86.4 %. This can be shown graphically using rule viewer. So, from this figure, it is clearly seen that link optimization is high (according to determination of link optimization using
Fuzzy Logic Approach for QoS Routing Analysis 167
*c) LOM Method*
While, 3 D surface viewer using three main defuzzification methods is the same and is depicted in the figure below. From this graph, it is clearly seen that we have high bandwidth, but delays and percentage of packet losses are also high, resulting thus in low
In this case we have a very low link usage, and this is the reason why surface viewer is
*If (bandwidth is Medium) and (delay is Tolerable) and (loss is Tolerable) then (link optimization is*
approximately same for three defuzzification methods (low link optimization).
Fig. 13. Surface viewer for 3 defuzzification methods: a) COG, b) MOM and c) LOM
Fig. 14. Rule viewer when it is used COG method for rule 9
If we use rule number 5 (derived from rule editor):
link optimization.
*Medium)*
membership functions). This value is obtained using COG method. Based on the results obtained here, we will conclude that this method is very effective.
The 86.4 % value represents the value after defuzzification. While, the surface viewer in 3D for three main defuzzification methods will look like in figure below.
If we use rule number 9 (derived from rule editor):
*If (bandwidth is High) AND (delay is Intolerable) AND (loss is Intolerable) THEN (link optimization is Low)*
Then, link optimization will be 20 % (see Figure 14). As we can see, in this case, link optimization is very low. So we can freely conclude that in a case when we have intolerable delays and intolerable losses, we will not have high QoS. The reason is because multimedia applications are very sensitive in delays and losses.
*a) COG Method*
*b) MOM Method*
166 Fuzzy Logic – Algorithms, Techniques and Implementations
membership functions). This value is obtained using COG method. Based on the results
The 86.4 % value represents the value after defuzzification. While, the surface viewer in 3D
*If (bandwidth is High) AND (delay is Intolerable) AND (loss is Intolerable) THEN (link*
Then, link optimization will be 20 % (see Figure 14). As we can see, in this case, link optimization is very low. So we can freely conclude that in a case when we have intolerable delays and intolerable losses, we will not have high QoS. The reason is because multimedia
*a) COG Method*
*b) MOM Method*
obtained here, we will conclude that this method is very effective.
If we use rule number 9 (derived from rule editor):
applications are very sensitive in delays and losses.
*optimization is Low)*
for three main defuzzification methods will look like in figure below.
*c) LOM Method*
Fig. 13. Surface viewer for 3 defuzzification methods: a) COG, b) MOM and c) LOM
Fig. 14. Rule viewer when it is used COG method for rule 9
While, 3 D surface viewer using three main defuzzification methods is the same and is depicted in the figure below. From this graph, it is clearly seen that we have high bandwidth, but delays and percentage of packet losses are also high, resulting thus in low link optimization.
In this case we have a very low link usage, and this is the reason why surface viewer is approximately same for three defuzzification methods (low link optimization).
If we use rule number 5 (derived from rule editor):
*If (bandwidth is Medium) and (delay is Tolerable) and (loss is Tolerable) then (link optimization is Medium)*
Fuzzy Logic Approach for QoS Routing Analysis 169
a) COG method
b) MOM method
While, 3 D surface viewer for 3 main defuzzification methods will look like below:
Fig. 15. Surface viewer for rule 9, for defuzzification methods: COG, MOM and LOM
Fig. 16. Rule viewer when we use COG method for rule 5
In this case, the value for bandwidth is 548 Mbps, delays = 437 ms (tolerable) and losses are tolerable (2.83%). From the analysis we have shown that the link optimization is medium. This means that using the parameters above, we can transmit almost every multimedia service.
168 Fuzzy Logic – Algorithms, Techniques and Implementations
Fig. 15. Surface viewer for rule 9, for defuzzification methods: COG, MOM and LOM
In this case, the value for bandwidth is 548 Mbps, delays = 437 ms (tolerable) and losses are tolerable (2.83%). From the analysis we have shown that the link optimization is medium. This means that using the parameters above, we can transmit almost every multimedia
Fig. 16. Rule viewer when we use COG method for rule 5
service.
While, 3 D surface viewer for 3 main defuzzification methods will look like below:
a) COG method
b) MOM method
Fuzzy Logic Approach for QoS Routing Analysis 171
methods that are used mostly for defuzzification, that are: COGM MOM and LOM. The performance comparison offered by three methods we have described through examples by surface viewer in 3D. Through these analyses using Matlab's toolbox we have shown that
[1] Mario Marchese, "*QoS over heterogeneous networks*", Copyright © 2007 John Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, England [2] Pragyansmita Paul and S V Raghavan–"Survey of QoS Routing"- Proceedings of the 15th international conference on Computer communication, p.50-75, August 12-14 [3] Santiago Alvarez, "*QoS for IP/MPLS Networks*", Cisco Press, Pub Date: June 02, 2006,
[5] Monique Morrow, Azhar Sayeed, "*MPLS and Next-Generation Networks: Foundations for*
[6] Arianit Maraj, B. Shatri, I. Limani, A. Abdullahu, S. Rugova *"*Analysis of QoS Routing in
[7] Runtong Zhang and Jian Ma - Fuzzy QoS Management in Diff-Serv Networks- Systems,
[8] A .Vasilakos, C .Ricudis, K. Anagnostakis, W .Pedrycz, A. Pitsillides-"Evolutionary-
[9] A .Vasilakos, C .Ricudis, K. Anagnostakis, W .Pedrycz, A. Pitsillides-"Evolutionary-
[10] Balandin, S. Heiner, A.P, SPF protocol and statistical tools for network simulations in
[11] By Eric Osborne, Ajay Simha, "*Traffic Engineering with MPLS*", Cisco Press, Pub Date:
[12] Baolin Sun, Layuan Li, Chao Gui - Fuzzy QoS Controllers Based Priority Scheduler for
[13] B. Shatri, A.Abdullahu, S. Rugova, Arianit Maraj, "VPN Creation in IP/MPLS Network
[14] K. H. Lee, "Firs Course on fuzzy theory and applications" [book], pages: 253-279, ISBN
*NGN and Enterprise Virtualization*", Cisco Press, Pub Date: November 06, 2006,
MPLS Network in Kosova using Fuzzy Logic"-Proceedings of the 7th WSEAS International Conference on Signal Processing, Robotics and Automation (ISPRA '08)
Man, and Cybernetics, 2000 IEEE International Conference on Volume 5, Issue ,
Fuzzy Prediction for Strategic QoS Routing in Broadband Networks", 0-7803-4863-
Fuzzy Prediction for Strategic QoS Routing in Broadband Networks", 0-7803-4863-
NS-2"- Information Technology Interfaces, 2002. ITI 2002. Proceedings of the 24th
Mobile Ad Hoc Networks- This paper appears in: Mobile Technology, Applications and Systems, 2005 2nd International Conference on Publication Date: 15-17 Nov.
in Kosova" icn, pp. 318-323, Seventh International Conference on Networking (icn
3-540-22988-4 Springer Berlin Heidelberg NewYork, Springer-Verlag Berlin
COG method gives better results in all analyzed cases.
[4] Shigang Chen and Klara Nahrstedt -"Distributed QoS routing"
International Conference on Publication Date: 2002
ISBN: 1-58705-233-4
ISBN: 1-58720-120-8
2000 Page(s): 3752 - 757 vol.5
W98 @10.0001998 IEEE
W98 @10.0001998 IEEE
2005
2008), 2008
Heidelberg 2005
July 17, 2002, ISBN: 1-58705-031-5
**15. References**
c) LOM method
Fig. 17. Surface viewer for rule 5 when it is used: a) COG, b) MOM and c) LOM
From the above examples can be clearly seen that COG defuzzification method gives better result in aspect of link utilization. This can be seen with surface viewer and rule viewer as presented in the above examples using different rules.
### **14. Conclusion**
In the past, routing problem in communication networks was relatively simple. The applications have used a modest percentage of bandwidth and no one of those applications had QoS requirements. However, the existing routing protocols should be improved or replaced with algorithms that meet different QoS requirements. Thus, it is necessary to present an architecture that supports new services in Internet and guarantees QoS for multimedia applications.
In this chapter we introduced the MPLS technology as a multiple solution for a lot of current problems faced by Internet today. By a wide support for QoS and traffic engineering capability, MPLS is establishing itself as a standard of the next generation's network. In MPLS network, some problems can occur during multimedia service transmission, therefore it is a good idea to design some control mechanisms for solving such problems. As a result of the complex nature of control mechanisms, in this chapter we used intelligent controlled techniques. One of the intelligent controlled techniques that we analyzed here is Fuzzy Logic Controller, a technique that is based on fuzzy logic. We have shown that fuzzy logic approach is suitable for QoS routing analysis in MPLS network. In this chapter we used main metrics of MPLS network acting in the input of FLC and we found the most appropriate defuzzification method for finding better crisp values in the aspect of link utilization in MPLS network. Also, we have shown that the most important part of FLC is defuzzifier, which converts fuzzy values into crisp values. In this chapter we explained the methods that are used mostly for defuzzification, that are: COGM MOM and LOM. The performance comparison offered by three methods we have described through examples by surface viewer in 3D. Through these analyses using Matlab's toolbox we have shown that COG method gives better results in all analyzed cases.
### **15. References**
170 Fuzzy Logic – Algorithms, Techniques and Implementations
c) LOM method
From the above examples can be clearly seen that COG defuzzification method gives better result in aspect of link utilization. This can be seen with surface viewer and rule viewer as
In the past, routing problem in communication networks was relatively simple. The applications have used a modest percentage of bandwidth and no one of those applications had QoS requirements. However, the existing routing protocols should be improved or replaced with algorithms that meet different QoS requirements. Thus, it is necessary to present an architecture that supports new services in Internet and guarantees QoS for
In this chapter we introduced the MPLS technology as a multiple solution for a lot of current problems faced by Internet today. By a wide support for QoS and traffic engineering capability, MPLS is establishing itself as a standard of the next generation's network. In MPLS network, some problems can occur during multimedia service transmission, therefore it is a good idea to design some control mechanisms for solving such problems. As a result of the complex nature of control mechanisms, in this chapter we used intelligent controlled techniques. One of the intelligent controlled techniques that we analyzed here is Fuzzy Logic Controller, a technique that is based on fuzzy logic. We have shown that fuzzy logic approach is suitable for QoS routing analysis in MPLS network. In this chapter we used main metrics of MPLS network acting in the input of FLC and we found the most appropriate defuzzification method for finding better crisp values in the aspect of link utilization in MPLS network. Also, we have shown that the most important part of FLC is defuzzifier, which converts fuzzy values into crisp values. In this chapter we explained the
Fig. 17. Surface viewer for rule 5 when it is used: a) COG, b) MOM and c) LOM
presented in the above examples using different rules.
**14. Conclusion**
multimedia applications.
**Term Weighting for Information**
Jorge Ropero, Ariel Gómez, Alejandro Carrasco,
*Department of Electronic Technology, University of Seville,*
The rising quantity of available information has constituted an enormous advance in our daily life. However, at the same time, some problems emerge as a result from the existing difficulty to distinguish the necessary information among the high quantity of unnecessary data. Information Retrieval has become a capital task for retrieving the useful information. Firstly, it was mainly used for document retrieval, but lately, its use has been generalized for the retrieval of any kind of information, such as the information contained in a database, a web page, or any set of accumulated knowledge. In particular, the so-called Vector Space Model is widely used. Vector Space Model is based on the use of index terms, which represent some pieces of knowledge or Objects. Index terms have associated weights, which
It is important that the assignment of weights to every index term - called Term Weighting - is automatic. The so-called TF-IDF method is mainly used for determining the weight of a term (Lee et al., 1997). Term Frequency (TF) is the frequency of occurrence of a term in a document; and Inverse Document Frequency (IDF) varies inversely with the number of documents to which the term is assigned (Salton, 1988). Although TF-IDF method for Term Weighting has worked reasonably well for Information Retrieval and has been a starting point for more recent algorithms, it was never taken into account that some other aspects of index terms may be important for determining term weights apart from TF and IDF: first of all, we should consider the degree of identification of an object if only the considered index term is used. This parameter has a strong influence on the final value of a term weight if the degree of identification is high. The more an index term identifies an object, the higher value for the corresponding term weight; secondly, we should also consider the existance of join terms.
These aspects are especially important when the information is abundant, imprecise, vague and heterogeneous. In this chapter, we define a new Term Weighting model based on Fuzzy Logic. This model tries to replace the traditional Term Weighting method, called TF-IDF. In order to show the efficiency of the new method, the Fuzzy Logic-based method has been tested on the website of the University of Seville. Web pages are usually a perfect example of heterogeneous and disordered information. We demonstrate the improvement introduced by the new method extracting the required information. Besides, it is also
possible to extract related information, which may be of interest to the users.
represent the importance of them in the considered set of knowledge.
**1. Introduction**
**Retrieval Using Fuzzy Logic**
Carlos León and Joaquín Luque
*Spain*
[15] D. Driankov, H. Hellenndoorn and M. Reinfrank "An Introduction to fuzzy Control", Springer – Verlang, Berlin, New York, 1993 **9**
## **Term Weighting for Information Retrieval Using Fuzzy Logic**
Jorge Ropero, Ariel Gómez, Alejandro Carrasco, Carlos León and Joaquín Luque *Department of Electronic Technology, University of Seville, Spain*
### **1. Introduction**
172 Fuzzy Logic – Algorithms, Techniques and Implementations
[15] D. Driankov, H. Hellenndoorn and M. Reinfrank "An Introduction to fuzzy Control",
The rising quantity of available information has constituted an enormous advance in our daily life. However, at the same time, some problems emerge as a result from the existing difficulty to distinguish the necessary information among the high quantity of unnecessary data. Information Retrieval has become a capital task for retrieving the useful information. Firstly, it was mainly used for document retrieval, but lately, its use has been generalized for the retrieval of any kind of information, such as the information contained in a database, a web page, or any set of accumulated knowledge. In particular, the so-called Vector Space Model is widely used. Vector Space Model is based on the use of index terms, which represent some pieces of knowledge or Objects. Index terms have associated weights, which represent the importance of them in the considered set of knowledge.
It is important that the assignment of weights to every index term - called Term Weighting - is automatic. The so-called TF-IDF method is mainly used for determining the weight of a term (Lee et al., 1997). Term Frequency (TF) is the frequency of occurrence of a term in a document; and Inverse Document Frequency (IDF) varies inversely with the number of documents to which the term is assigned (Salton, 1988). Although TF-IDF method for Term Weighting has worked reasonably well for Information Retrieval and has been a starting point for more recent algorithms, it was never taken into account that some other aspects of index terms may be important for determining term weights apart from TF and IDF: first of all, we should consider the degree of identification of an object if only the considered index term is used. This parameter has a strong influence on the final value of a term weight if the degree of identification is high. The more an index term identifies an object, the higher value for the corresponding term weight; secondly, we should also consider the existance of join terms.
These aspects are especially important when the information is abundant, imprecise, vague and heterogeneous. In this chapter, we define a new Term Weighting model based on Fuzzy Logic. This model tries to replace the traditional Term Weighting method, called TF-IDF. In order to show the efficiency of the new method, the Fuzzy Logic-based method has been tested on the website of the University of Seville. Web pages are usually a perfect example of heterogeneous and disordered information. We demonstrate the improvement introduced by the new method extracting the required information. Besides, it is also possible to extract related information, which may be of interest to the users.
Term Weighting for Information Retrieval Using Fuzzy Logic 175
So firstly, terms that are mentioned frequently in individual documents or extracts from a document, appear to be useful for improving recall. This suggests the use of a factor known as Term Frequency (TF) as part of a Term Weighting system, measuring the frequency of occurance of a term in a document. The TF factor has been used for Term Weighting for years in automatic indexing environments. Secondly, the TF factor solely does not ensure an acceptable retrieval. In particular, when the high frequency terms are not concentrated in specific documents, but instead are frequent in the entire set, all documents tend to be recovered, and this affects the precision factor. Thus, there is the need to introduce a new factor that favours the terms that are concentrated in only a few documents in the collection. The Inverse Document Frequency (IDF) is the factor that considers this aspect. The IDF factor is inversely proportional to the number of documents (n) to which a term is assigned in a set of documents N. A typical IDF factor is log (N / n) (Salton & Buckley, 1996). So the best index terms to identify the contents of a document are those able to distinguish certain individual documents from the rest of the set. This implies that the best terms should have high term frequencies, but low overall collection frequencies. A reasonable measure of the importance of a term can be obtained, therefore, by the product of term frequency and inverse document frequency (TF x IDF). It is usual to describe the weight of a term *i* in a
wij = tfij x idfj
This formula was originally designed for the retrieval and extraction of documents. Eventually, it has also been used for the retrieval of any object in any set of accumulated knowledge, and has been revised and improved by other authors in order to obtain better results in Information Retrieval (Lee et al., 1997), (Zhao & Karypis, 2002), (Lertnattee &
In short, term weights must be related somehow to the importance of an index term in the
corresponding set of knowledge. There are two options for defining these weights:
To calculate the weight of a term, the TF-IDF approach considers two factors:
*retrieved relevant objects* Precision
Equation 2. Definition of precision
document *j* as shown in Equation 3.
Theeramunkong, 2003), (Liu & Ke, 2007).
Information Retrieval.
Equation 3. Obtention of term weights; general formula
subjective and is not able of being automated.
occurrence of the term Tk in document i.
*total number of retrieved objects*
### **2. Vector Space Model and Term Weighting**
In the Vector Space Model, the contents of a document are represented by a multidimensional space vector. Later, the proper classes of the given vector are determined by comparing the distances between vectors. The procedure of the Vector Space Model can be divided into three stages, as seen in Figure 1 (Raghavan & Wong, 1986):
Fig. 1. Vector Space Model procedure
In this chapter, we are focusing in the second stage. It was in the late 50's when the idea of text retrieval came up - a concept that was later extended to general information retrieval -. Text retrieval was founded on an automatic search based on textual content through a series of identifiers. It was Gerard Salton who laid the foundations for linking these identifiers and the texts that they represent during the 70's and the 80's. Salton suggested that every document could be represented by a term vector in the way D = (ti, tj,…, tp), where every tk identifies a term assigned to a document D. A formal representation of the vector D leads us not to consider only the terms in the vector, but to add a set of weights representing the term weight, it is to say, its importance in the document.
A Term Weighting system should improve efficiency in two main factors, recall and precision. Recall takes into account the fact that the objects relevant to the user should be retrieved. Precision considers the fact that the objects that are not wanted by the user should be rejected. In principle, it is desirable to build a system that rewards both high recall, - retrieving all that is relevant - and high precision - discarding all unwanted objects (Ruiz & Srinisavan, 1998). Recall improves using high-frequency index terms, i.e. terms which occur in many documents of the collection. This way, it is expected to retrieve many documents including such terms, and thus, many of the relevant documents. The precision factor, however, improves when using more specific index terms that are capable of isolating the few relevant articles of the mass of irrelevant. In practice, compromises are utilized; using frequent enough terms to achieve a reasonable level of recall without causing a too low value of precision. The exact definitions of recall and precision are shown in Equations 1 and 2.
> *total number of relevant objects retrieved relevant objects* Recall
Equation 1. Definition of recall
Equation 2. Definition of precision
174 Fuzzy Logic – Algorithms, Techniques and Implementations
In the Vector Space Model, the contents of a document are represented by a multidimensional space vector. Later, the proper classes of the given vector are determined by comparing the distances between vectors. The procedure of the Vector Space Model can
In this chapter, we are focusing in the second stage. It was in the late 50's when the idea of text retrieval came up - a concept that was later extended to general information retrieval -. Text retrieval was founded on an automatic search based on textual content through a series of identifiers. It was Gerard Salton who laid the foundations for linking these identifiers and the texts that they represent during the 70's and the 80's. Salton suggested that every document could be represented by a term vector in the way D = (ti, tj,…, tp), where every tk identifies a term assigned to a document D. A formal representation of the vector D leads us not to consider only the terms in the vector, but to add a set of weights representing the term
A Term Weighting system should improve efficiency in two main factors, recall and precision. Recall takes into account the fact that the objects relevant to the user should be retrieved. Precision considers the fact that the objects that are not wanted by the user should be rejected. In principle, it is desirable to build a system that rewards both high recall, - retrieving all that is relevant - and high precision - discarding all unwanted objects (Ruiz & Srinisavan, 1998). Recall improves using high-frequency index terms, i.e. terms which occur in many documents of the collection. This way, it is expected to retrieve many documents including such terms, and thus, many of the relevant documents. The precision factor, however, improves when using more specific index terms that are capable of isolating the few relevant articles of the mass of irrelevant. In practice, compromises are utilized; using frequent enough terms to achieve a reasonable level of recall without causing a too low value of precision. The exact
*total number of relevant objects*
*retrieved relevant objects* Recall
be divided into three stages, as seen in Figure 1 (Raghavan & Wong, 1986):
**2. Vector Space Model and Term Weighting**
order to improve the retrieval relevant to the user.
Fig. 1. Vector Space Model procedure
weight, it is to say, its importance in the document.
definitions of recall and precision are shown in Equations 1 and 2.
Equation 1. Definition of recall
So firstly, terms that are mentioned frequently in individual documents or extracts from a document, appear to be useful for improving recall. This suggests the use of a factor known as Term Frequency (TF) as part of a Term Weighting system, measuring the frequency of occurance of a term in a document. The TF factor has been used for Term Weighting for years in automatic indexing environments. Secondly, the TF factor solely does not ensure an acceptable retrieval. In particular, when the high frequency terms are not concentrated in specific documents, but instead are frequent in the entire set, all documents tend to be recovered, and this affects the precision factor. Thus, there is the need to introduce a new factor that favours the terms that are concentrated in only a few documents in the collection. The Inverse Document Frequency (IDF) is the factor that considers this aspect. The IDF factor is inversely proportional to the number of documents (n) to which a term is assigned in a set of documents N. A typical IDF factor is log (N / n) (Salton & Buckley, 1996). So the best index terms to identify the contents of a document are those able to distinguish certain individual documents from the rest of the set. This implies that the best terms should have high term frequencies, but low overall collection frequencies. A reasonable measure of the importance of a term can be obtained, therefore, by the product of term frequency and inverse document frequency (TF x IDF). It is usual to describe the weight of a term *i* in a document *j* as shown in Equation 3.
$$\mathbf{w}\_{\overline{\imath}\overline{\jmath}} = \mathbf{tf}\_{\overline{\imath}\overline{\jmath}} \ge \mathbf{id} \mathbf{f}\_{\overline{\jmath}}$$
Equation 3. Obtention of term weights; general formula
This formula was originally designed for the retrieval and extraction of documents. Eventually, it has also been used for the retrieval of any object in any set of accumulated knowledge, and has been revised and improved by other authors in order to obtain better results in Information Retrieval (Lee et al., 1997), (Zhao & Karypis, 2002), (Lertnattee & Theeramunkong, 2003), (Liu & Ke, 2007).
In short, term weights must be related somehow to the importance of an index term in the corresponding set of knowledge. There are two options for defining these weights:
To calculate the weight of a term, the TF-IDF approach considers two factors:
Term Weighting for Information Retrieval Using Fuzzy Logic 177
The TF-IDF method works reasonably well, but has the disadvantage of not considering two
This chapter describes, firstly, the operation of TF-IDF method. Then, the new Term Weighting Fuzzy Logic-based method is introduced. Finally, both methods are implemented for the particular case of Information Retrieval for the University of Seville web portal, obtaining specific results of the operation of both of them. A web portal is a typical example of a disordered, vague and heterogenous set of knowledge. With this aim, an intelligent agent was designed to allow an efficient retrieval of the relevant information. This system should be valid for any set of knowledge. The system was designed to enable users to find possible answers to their queries in a set of knowledge of a great size. The whole set of knowledge was classified into different objects. These objects represent the possible answers to user queries and were organized into hierarchical groups (called Topic, Section and Object). One or more standard questions are assigned to every object and some
The last step is Term Weigthing; the assigned weight depends on the importance of an index term for the identification of the object. The way in which these weights are assigned is the
As an example of the classical TF-IDF Term Weighting method functioning, we are using
**3. Term Weighting method comparison**
**3.1 Term Weighting methods**
aspects that we believe key:
with different objects.
index terms are extracted from them.
other tfik and nk for the Topic-.
At Topic hierarchic level:
main issue of this chapter. All the process is shown in Figure 3.
the term 'library', used in the example shown in Table 1.
another object.
Introducing standardization to simplify the calculations, the formula finally obtained for the calculation of the weights is defined in Equation 4 (Liu et al., 2001)
$$W\_{ik} = \frac{y\_{ik}^r \times \log(N/n\_k + 0.01)}{\sqrt{\sum\_{k=1}^m y\_{ik}^r \times \log(N/n\_k + 0.01)}^{2^r}}$$
Equation 4. Obtention of term weights. Used formula.
A third factor that is commonly used is the document length normalization factor. Long documents usually have a much larger set of extracted terms than short documents. This fact makes it more likely that long documents are retrieved (Van Rijsbergen, 1979), (Salton & Buckley, 1996). The term weight obtained using a length normalization factor is given by Equation 5.
$$W\_{ik} = \frac{w\_{ik}}{\sqrt{\sum\_{i=1}^{m} (w\_i)^2}}$$
Equation 5. Obtention of term weights using a length normalization factor
In Equation 5, wi correspond to the weights of the other components of the vector.
All Term Weighting tasks are shown in Figure 2.
Fig. 2. Term Weighting tasks
### **3. Term Weighting method comparison**
### **3.1 Term Weighting methods**
176 Fuzzy Logic – Algorithms, Techniques and Implementations
Introducing standardization to simplify the calculations, the formula finally obtained for the
*ik k*
*nNtf*
A third factor that is commonly used is the document length normalization factor. Long documents usually have a much larger set of extracted terms than short documents. This fact makes it more likely that long documents are retrieved (Van Rijsbergen, 1979), (Salton & Buckley, 1996). The term weight obtained using a length normalization factor is given by
1
*i*
2 )(
*w*
*ik*
*w*
*i*
*m*
In Equation 5, wi correspond to the weights of the other components of the vector.
*ik*
*W*
Equation 5. Obtention of term weights using a length normalization factor
<sup>2</sup> ))01.0/log(
)01.0/log(
*ik k*
calculation of the weights is defined in Equation 4 (Liu et al., 2001)
*ik*
Equation 4. Obtention of term weights. Used formula.
All Term Weighting tasks are shown in Figure 2.
Fig. 2. Term Weighting tasks
1
*k*
*m*
*nNtf <sup>W</sup>*
expression log (N / nk + 0.01).
Equation 5.
The TF-IDF method works reasonably well, but has the disadvantage of not considering two aspects that we believe key:
This chapter describes, firstly, the operation of TF-IDF method. Then, the new Term Weighting Fuzzy Logic-based method is introduced. Finally, both methods are implemented for the particular case of Information Retrieval for the University of Seville web portal, obtaining specific results of the operation of both of them. A web portal is a typical example of a disordered, vague and heterogenous set of knowledge. With this aim, an intelligent agent was designed to allow an efficient retrieval of the relevant information. This system should be valid for any set of knowledge. The system was designed to enable users to find possible answers to their queries in a set of knowledge of a great size. The whole set of knowledge was classified into different objects. These objects represent the possible answers to user queries and were organized into hierarchical groups (called Topic, Section and Object). One or more standard questions are assigned to every object and some index terms are extracted from them.
The last step is Term Weigthing; the assigned weight depends on the importance of an index term for the identification of the object. The way in which these weights are assigned is the main issue of this chapter. All the process is shown in Figure 3.
As an example of the classical TF-IDF Term Weighting method functioning, we are using the term 'library', used in the example shown in Table 1.
At Topic hierarchic level:
Term Weighting for Information Retrieval Using Fuzzy Logic 179
Consequently, 'Library' is relevant to find out that the Object is in Topic 6, but not very relevant to find out the definite Object, which should be found according to other terms in a
As said above, TF-IDF has the disadvantage of not considering the degree of identification of the object if only the considered index term is used and the existance of join terms. The FLbased method provides a solution for these problems: the solution is to create a table of all the index terms and their corresponding weights for each object. This table will be created in the process of extracting the index words from the standard questions. Imprecision practically does not affect the method due to the fact that Term Weighting is based on fuzzy logic. This
For example, in the case of a website, the own web page developer may define standard questions. These questions are associated with the object - the web page -. He also should define the index for each object and answer the two questions proposed above. This greatly
Fuzzy Logic based Term Weighting method is defined below. Four questions must be
fact minimizes the effect of possible variations of the assigned weights.
How does an index term define an object by itself?
answered to determine the weight of an Index Term:
Are there any join terms tied to the considered index term?
Furthermore, the Fuzzy Logic-based method provides two important advantages:
simplifies the process and leaves the possibility of using collaborative intelligence.
is not very relevant to distinguish the desired Section inside the Topic.
At Section hierarchic level:
At Object hierarchic level:
once in an Object -.
IDF factor -.
TF factor -.
user consultation.
Fig. 3. Information Retrieval process.
Table 1. Example of the followed methodology.
At Section hierarchic level:
178 Fuzzy Logic – Algorithms, Techniques and Implementations
Fig. 3. Information Retrieval process.
Step 1: Web page identified by
standard/s question/s
Step 2: Locate standard/s question/s in the hierarchical
structure.
As well, an example of the followed methodology is shown in Table 1.
idweb.html
Seville?
Topic 6: Library
Object 1.
Step 3: Extract index terms Index terms: 'Library', 'services', 'online'
Section 3: Online services
**STEP EXAMPLE**
Step 4: Term weighting Explained below
Table 1. Example of the followed methodology.
At Object hierarchic level:
Consequently, 'Library' is relevant to find out that the Object is in Topic 6, but not very relevant to find out the definite Object, which should be found according to other terms in a user consultation.
As said above, TF-IDF has the disadvantage of not considering the degree of identification of the object if only the considered index term is used and the existance of join terms. The FLbased method provides a solution for these problems: the solution is to create a table of all the index terms and their corresponding weights for each object. This table will be created in the process of extracting the index words from the standard questions. Imprecision practically does not affect the method due to the fact that Term Weighting is based on fuzzy logic. This fact minimizes the effect of possible variations of the assigned weights.
Furthermore, the Fuzzy Logic-based method provides two important advantages:
For example, in the case of a website, the own web page developer may define standard questions. These questions are associated with the object - the web page -. He also should define the index for each object and answer the two questions proposed above. This greatly simplifies the process and leaves the possibility of using collaborative intelligence.
Fuzzy Logic based Term Weighting method is defined below. Four questions must be answered to determine the weight of an Index Term:
Term Weighting for Information Retrieval Using Fuzzy Logic 181
Provided that there are 1114 index terms defined in our case, we think that 1 % of these words must mark the border for the value 0 (11 words). Therefore, whenever an index term appears more than 12 times in other subsets, we will give it the value of 0. Associated values
Associated value 1 0.9 0.8 0.7 0.64 0.59 0.53
Associated value 0.47 0.41 0.36 0.3 0.2 0.1 0
Between 0 and 3 times appearing - approximately a third of the possible values - , we consider that an index term belongs to the so called HIGH set. Therefore, it is defined in its correspondant fuzzy set with uniformly distributed values between 0.7 and 1, as may be seen in Figure 5. Analogously, we distribute all values uniformly according to different fuzzy sets. Fuzzy sets are defined by linguistic variables LOW, MEDIUM and HIGH. Fuzzy sets are triangular, on one hand for simplicity and on the other hand because we tested other more complex types of sets (Gauss, Pi type, etc), but the results did not improve at all.
On the other hand, given that at each hierarchical level, a different term weight is defined, it is necessary to consider other scales to calculate the fuzzy system input values for the other hierarchical levels. As for the level of topic was considered the top level - the whole set of knowledge - , for the level of Section we consider the number occurrences of an index term on a given topic. Keeping in mind that all topics are considered, we take as reference the
Table 3. Input values associated to Q1 for topic hierarchic level.
0 1 2 3 4 5 6
7 8 9 10 11 12 ≥13
for every Topic are defined in Table 3.
Number of appearances
Number of appearances
Fig. 5. Input fuzzy sets.
With the answers to these questions, a set of values is obtained. These values are the inputs to a fuzzy logic system, a Term Weight Generator. The Fuzzy Logic system output sets the weight of an index term for each hierarchical level (Figure 4).
Fig. 4. Term Weighting using Fuzzy Logic.
Next it is described how to define the system input values associated with each of the four questions (Qi). Qi are the inputs to the Fuzzy Logic system
### *Question 1*
Term weight is partly associated to the question 'How often does an index term appear in other subsets?'. It is given by a value between 0 – if it appears many times – and 1 - if it does not appear in any other subset -. To define weights, we are considering the times that the most used terms in the whole set of knowledge appear. The list of the most used index terms is shown in Table 2.
Table 2. List of the most used words.
180 Fuzzy Logic – Algorithms, Techniques and Implementations
With the answers to these questions, a set of values is obtained. These values are the inputs to a fuzzy logic system, a Term Weight Generator. The Fuzzy Logic system output sets the
Next it is described how to define the system input values associated with each of the four
Term weight is partly associated to the question 'How often does an index term appear in other subsets?'. It is given by a value between 0 – if it appears many times – and 1 - if it does not appear in any other subset -. To define weights, we are considering the times that the most used terms in the whole set of knowledge appear. The list of the most used index
**Number of order Index term Number of appearances in the**
Resources 12
1 Service 31 2 Services 18 3 Library 16 4 Research 15 5 Address 14 Student 14 7 Mail 13 Access 13 9 Electronic 12 Computer 12
12 Center 10 Education 10 Registration 10 Program 10
**accumulated set of knowledge**
weight of an index term for each hierarchical level (Figure 4).
Fig. 4. Term Weighting using Fuzzy Logic.
*Question 1*
terms is shown in Table 2.
Table 2. List of the most used words.
questions (Qi). Qi are the inputs to the Fuzzy Logic system
Provided that there are 1114 index terms defined in our case, we think that 1 % of these words must mark the border for the value 0 (11 words). Therefore, whenever an index term appears more than 12 times in other subsets, we will give it the value of 0. Associated values for every Topic are defined in Table 3.
Table 3. Input values associated to Q1 for topic hierarchic level.
Between 0 and 3 times appearing - approximately a third of the possible values - , we consider that an index term belongs to the so called HIGH set. Therefore, it is defined in its correspondant fuzzy set with uniformly distributed values between 0.7 and 1, as may be seen in Figure 5. Analogously, we distribute all values uniformly according to different fuzzy sets. Fuzzy sets are defined by linguistic variables LOW, MEDIUM and HIGH. Fuzzy sets are triangular, on one hand for simplicity and on the other hand because we tested other more complex types of sets (Gauss, Pi type, etc), but the results did not improve at all.
Fig. 5. Input fuzzy sets.
On the other hand, given that at each hierarchical level, a different term weight is defined, it is necessary to consider other scales to calculate the fuzzy system input values for the other hierarchical levels. As for the level of topic was considered the top level - the whole set of knowledge - , for the level of Section we consider the number occurrences of an index term on a given topic. Keeping in mind that all topics are considered, we take as reference the
Term Weighting for Information Retrieval Using Fuzzy Logic 183
For example, the developer of a web page would only have to answer "Yes", "Rather" or
Finally, question 4 deals with the number of index terms joined to another one. If an index term is joined to another one, its weight is lower. This is due to the fact that the term must be a join term to refer to the object in question. We propose term weight values for this question in Table 9. Again, the values 0.7 and 0.3 are a consequence of considering the
After considering all these factors, fuzzy rules must be defined. In the case of Topic and Section hierarchical levels, we must consider the four input values that are associated with questions Q1, Q2, Q3 and Q4. Four output fuzzy sets have been also defined: HIGH, MEDIUM-HIGH, MEDIUM-LOW AND LOW. For the definition of the fuzzy rules for the
The combination of the four inputs and the three input fuzzy sets provides 81 possible
In the object level (the last hierarchic level), Question 2 is discarded. Therefore, there is a change in the rules, although the criteria for the definition of fuzzy rules are similar to the
An example of the followed methodology is shown below. A comparision with the classical TF-IDF is done, starting from the definition of an object in the database of the Web portal of
Associated value 1 0.5 0
"No" to Question 3, without complicated mathematical formulas to describe it.
**Joined terms to an index term** 0 1 2 ≥ 3 Associated value 1 0.7 0.3 0
Term Weighting system, we have used basically the following criteria:
little importance (low Q3) or it is joined to many terms (low Q4).
previous case. An input less reduces the number of rules to twenty seven.
particular object, it is much easier to find the object.
terms. This fact causes a lower output value.
combinations, which are summarized in Table 10.
**3.2 Example of the followed methodology**
Yes Rather No
**Answer (Does the term itself define the**
Table 8. Input values associated to Q3.
Table 9. Input values associated to Q3.
for extracting information.
border between fuzzy sets.
**Object?)**
*Question 4*
value of the topic in which the index term appears more often. The process is analogous to the above described, obtaining the values shown in Table 4.
Table 4. Input values associated to Q1 for section hierarchic level.
To find the term weight associated with the object level, the method is slightly different. It is also based on the definition of fuzzy sets, but we do not take into account the maximum number of words per secion, but the value associated to Q1 directly passes the border between fuzzy sets when the number of objects in which it appears increases in one unit, as seen in Table 5.
Table 5. Input values associated to Q1 for object hierarchic level.
### *Question 2*
To find the imput value to the FL system of FL with question 2, the reasoning is analogous to the one for Q1, Though, we only have to consider the frequency of occurance of an index term within a single subset of knowledge, and not the frequency of occurrence in other subsets. Logically, the more times a term appears in a subset, the greater the probability that the query is related to it. Question Q2 corresponds to the TF factor.
Looking again at the list of index terms used in a topic, we obtain the values shown in Tables 6 and 7. It has been taken into account that the more times an index term appears in a topic or section, the greater should be the input value. These tables correspond to the values for the hierarchical levels of Topic and Section, respectively.
Table 6. Input values associated to Q2 for topic hierarchic level.
Table 7. Input values associated to Q2 for section hierarchic level.
Q2 is meaningless to determine the input value for the last hierarchical level. At this level, an index term appears only once on every object.
### *Question 3*
For Question 3, the answer is completely subjective. In this chapter, we propose the values "Yes", "Rather" and "No". Table 8, shows the input values associated with Q3. This value is independent of hierarchical level.
Table 8. Input values associated to Q3.
For example, the developer of a web page would only have to answer "Yes", "Rather" or "No" to Question 3, without complicated mathematical formulas to describe it.
### *Question 4*
182 Fuzzy Logic – Algorithms, Techniques and Implementations
value of the topic in which the index term appears more often. The process is analogous to
To find the term weight associated with the object level, the method is slightly different. It is also based on the definition of fuzzy sets, but we do not take into account the maximum number of words per secion, but the value associated to Q1 directly passes the border between fuzzy sets when the number of objects in which it appears increases in one unit, as
To find the imput value to the FL system of FL with question 2, the reasoning is analogous to the one for Q1, Though, we only have to consider the frequency of occurance of an index term within a single subset of knowledge, and not the frequency of occurrence in other subsets. Logically, the more times a term appears in a subset, the greater the probability that
Looking again at the list of index terms used in a topic, we obtain the values shown in Tables 6 and 7. It has been taken into account that the more times an index term appears in a topic or section, the greater should be the input value. These tables correspond to the values
Number of appearances 1 2 3 4 5 ≥ 6 Associated value 0 0.3 0.45 0.6 0.7 1
Number of appearances 1 2 3 4 5 ≥ 6 Associated value 0 0.3 0.45 0.6 0.7 1
Q2 is meaningless to determine the input value for the last hierarchical level. At this level,
For Question 3, the answer is completely subjective. In this chapter, we propose the values "Yes", "Rather" and "No". Table 8, shows the input values associated with Q3. This value is
Associated value 1 0.7 0.6 0.5 0.4 0.3 0
Number of appearances 0 1 2 ≥ 3 Associated value 1 0.7 0.3 0
0 1 2 3 4 5 ≥ 6
the above described, obtaining the values shown in Table 4.
Table 4. Input values associated to Q1 for section hierarchic level.
Table 5. Input values associated to Q1 for object hierarchic level.
the query is related to it. Question Q2 corresponds to the TF factor.
for the hierarchical levels of Topic and Section, respectively.
Table 6. Input values associated to Q2 for topic hierarchic level.
Table 7. Input values associated to Q2 for section hierarchic level.
an index term appears only once on every object.
independent of hierarchical level.
Number of appearances
seen in Table 5.
*Question 2*
*Question 3*
Finally, question 4 deals with the number of index terms joined to another one. If an index term is joined to another one, its weight is lower. This is due to the fact that the term must be a join term to refer to the object in question. We propose term weight values for this question in Table 9. Again, the values 0.7 and 0.3 are a consequence of considering the border between fuzzy sets.
Table 9. Input values associated to Q3.
After considering all these factors, fuzzy rules must be defined. In the case of Topic and Section hierarchical levels, we must consider the four input values that are associated with questions Q1, Q2, Q3 and Q4. Four output fuzzy sets have been also defined: HIGH, MEDIUM-HIGH, MEDIUM-LOW AND LOW. For the definition of the fuzzy rules for the Term Weighting system, we have used basically the following criteria:
The combination of the four inputs and the three input fuzzy sets provides 81 possible combinations, which are summarized in Table 10.
In the object level (the last hierarchic level), Question 2 is discarded. Therefore, there is a change in the rules, although the criteria for the definition of fuzzy rules are similar to the previous case. An input less reduces the number of rules to twenty seven.
### **3.2 Example of the followed methodology**
An example of the followed methodology is shown below. A comparision with the classical TF-IDF is done, starting from the definition of an object in the database of the Web portal of
Term Weighting for Information Retrieval Using Fuzzy Logic 185
**Method** - - - - 1
**Method** - - - - 0.01
**Method** - - - - 0.01
We may see the difference with the corresponding weight for the TF-IDF method - a value Wik = 0.01 had been obtained), but this is just what we were looking for: not only the desired object is found, but also the ones that are more closely related to it. The word 'library' has a small weight for the TF-IDF method because it can not distinguish between the objects of Section 6.3. However, in this case all the objects will be retrieved, as they are interrelated. The weights of other terms determine the object which has a higher level of certainty.
Tests were held on the website of the University of Seville. 253 objects were defined, and grouped in a hierarchical structure, with 12 topics. Every topic has a variable number of sections and objects. From these 253 objects, 2107 standard questions were extracted. More
0.53 1 0.35 0.66 0.56
0 0.6 0.375 0.63 0.13
0 - 0.5 0.57 0.33
**value**
**Term Weight**
A summary of the values for the index term 'library' is shown in Table 11.
**Hierarchic levels Q1 value Q2 value Q3 value Q4**
At Object hierarchich level:
associated to Q1 is 0.
associated to Q4 is 0.57.
**TF-IDF**
**Fuzzy Logicbased method**
**TF-IDF**
**Fuzzy Logicbased method**
**TF-IDF**
**Fuzzy Logicbased method**
Table 11. Comparison of Term Weight values.
**Topic level (Topic 6)**
**Section level (Section 3)**
**Object level (Object 1)**
**4. Tests and results**
**4.1 General tests**
the University of Seville. The following example shows the difference between applying the TF-IDF method and applying the Fuzzy Logic-based one.
Table 10. Fuzzy rules.
In the Web portal database, Object 6.3.1 (http://bib.us.es/index-ides-idweb.html) is defined by the following standard question:
*What online services are offered by the Library of the University of Seville?*
If we consider the term 'library':
At Topic hierarchic level:
At Section hierarchic level:
At Object hierarchich level:
184 Fuzzy Logic – Algorithms, Techniques and Implementations
the University of Seville. The following example shows the difference between applying the
R1 IF Q1 = HIGH and Q2 ≠ LOW At least MEDIUM-
R2 IF Q1 = MEDIUM and Q2 = HIGH At least MEDIUM-
R3 IF Q1 = HIGH and Q2 = LOW Depends on other
R4 IF Q1 = HIGH and Q2 = LOW Depends on other
R5 IF Q3 = HIGH At least MEDIUM-
R6 IF Q4 = LOW Descends a level
R8 IF (R1 and R2) or (R1 and R5) or (R2 and R5) HIGH R9 In any other case MEDIUM-LOW
In the Web portal database, Object 6.3.1 (http://bib.us.es/index-ides-idweb.html) is defined
value associated to Q3 is a weighted average: (7\*0.5 + 3\*0)/10 = 0.35.
HIGH
HIGH
Questions
Questions
HIGH
If the Output is MEDIUM-LOW, it descends to LOW
**Rule number Rule definition Output**
TF-IDF method and applying the Fuzzy Logic-based one.
R7 IF Q4 = MEDIUM
*What online services are offered by the Library of the University of Seville?*
Table 10. Fuzzy rules.
by the following standard question:
If we consider the term 'library':
value associated to Q1 is 0.53.
At Topic hierarchic level:
At Section hierarchic level:
associated to Q1 is 0.
Q3 is (3\*0.5 + 1\*0)/4 = 0.375.
value associated to Q4 is 0.63.
A summary of the values for the index term 'library' is shown in Table 11.
Table 11. Comparison of Term Weight values.
We may see the difference with the corresponding weight for the TF-IDF method - a value Wik = 0.01 had been obtained), but this is just what we were looking for: not only the desired object is found, but also the ones that are more closely related to it. The word 'library' has a small weight for the TF-IDF method because it can not distinguish between the objects of Section 6.3. However, in this case all the objects will be retrieved, as they are interrelated. The weights of other terms determine the object which has a higher level of certainty.
### **4. Tests and results**
### **4.1 General tests**
Tests were held on the website of the University of Seville. 253 objects were defined, and grouped in a hierarchical structure, with 12 topics. Every topic has a variable number of sections and objects. From these 253 objects, 2107 standard questions were extracted. More
Term Weighting for Information Retrieval Using Fuzzy Logic 187
**Method Cat1 Cat2 Cat3 Cat4 Cat5 Total**
The results obtained with the TF-IDF method are quite reasonable. 81.18% of the objects are retrieved among the top 5 choices and more than half of the objects are retrieved in the first place, Fuzzy Logic-based method is clearly better. 92.45% of the objects are retrieved and
In order to refine the conclusions about both Term Weighting methods, it is important to make a more thorough analysis of the results. We submitted to both Term Weighting methods to a comprehensive analysis according to the type of standard question. Results are
According to the results, the TF-IDF method works relatively well considering the number of objects retrieved. Though, the Fuzzy Logic-based method is more precise, retrieving 91.67% of the objects in the first place. On the other hand, good results for this type of questions are logical, since questions correspond to supposedly well-made user queries.
For synonymous standard questions, the conclusions are similar: the results obtained using the Fuzzy Logic-based method are better than those achieved with TF-IDF method, especially in regard to precision. Though, the TF-IDF method also ensures good results. However, queries are not precise, so the performance is worse for the TF-IDF method than it is for the Fuzzy Logic-based method. This fact gives an idea of fuzzy logic as an ideal tool for adding more flexibility to the system. Anyway, the results are quite similar to those obtained for the main standard questions. They are only slightly worse, since synonim
The difference is even more noticeable in regard to imprecise standard questions and specific standard questions. Imprecise standard questions are detected nearly as well as the main standard questions in the case of Fuzzy Logic-based method. This is another reason to confirm the appropriateness of using Fuzzy Logic. As for the specific standard questions, we
Table 13. Information Retrieval results of using both Term Weighting methods.
466 (50.98%) 223 (24.40%) 53 (5.80%) 79 (8.64%) 93 (10.18%) 914
710 (77.68%) 108 (11.82%) 27 (2.95%) 28 (3.06%) 41 (4.49%) 914
degree of certainty -excluding the previous case -.
degree of certainty - excluding the previous cases -.
more than three-quarters are retrieved in the first place.
**4.2 Tests according to the type of standard questions**
standard questions are similar to the main standard questions.
higher degree of certainty.
Results are shown in Table 13
**TF-IDF method**
**FL-based method**
shown in the Table 14.
than half of them were not used for these tests, as they were similar to others and did not contribute much to the results. Finally, the number of standard questions used for the tests was 914. Also, several types of standard questions were defined.
Depending on the nature of the considered object, we defined different types of standard questions, such as:
For our tests, we considered the types of standard questions shown in Table 12.
Table 12. Types of standard questions.
The standard questions were used as inputs in a Fuzzy Logic-based system. The outputs of the system are the objects with a degree a certainty greater than a certain threshold. To compare results, we considered the position in which the correct answer appears among the total number of answers identified as probable.
First of all, we shall define the thresholds to overcome in the Fuzzy Logic system. Thus, topics and sections that are not related to the object to be identified are removed. This is one of the advantages of using a hierarchical structure. Processing time is better as many subsets of knowledge are discarded. Anyway, it is desirable not to discard too many objects, in order to also obtain the related ones. The ideal is to retrieve between one and five answers for the user. The results of the consultation were sorted in 5 categories:
Results are shown in Table 13
186 Fuzzy Logic – Algorithms, Techniques and Implementations
than half of them were not used for these tests, as they were similar to others and did not contribute much to the results. Finally, the number of standard questions used for the tests
Depending on the nature of the considered object, we defined different types of standard
The standard questions were used as inputs in a Fuzzy Logic-based system. The outputs of the system are the objects with a degree a certainty greater than a certain threshold. To compare results, we considered the position in which the correct answer appears among the
First of all, we shall define the thresholds to overcome in the Fuzzy Logic system. Thus, topics and sections that are not related to the object to be identified are removed. This is one of the advantages of using a hierarchical structure. Processing time is better as many subsets of knowledge are discarded. Anyway, it is desirable not to discard too many objects, in order to also obtain the related ones. The ideal is to retrieve between one and five answers
for the user. The results of the consultation were sorted in 5 categories:
For our tests, we considered the types of standard questions shown in Table 12.
**Type of standard question Number of questions**
*Main standard questions* 252 *Synonim standard questions* 308 *Imprecise standard questions* 125 *Specific standard questions* 229 *Feedback standard questions* 0 **Total standard questions 914**
Table 12. Types of standard questions.
total number of answers identified as probable.
was 914. Also, several types of standard questions were defined.
questions, such as:
questions.
used.
questions are optional.
them synonim standard questions.
Table 13. Information Retrieval results of using both Term Weighting methods.
The results obtained with the TF-IDF method are quite reasonable. 81.18% of the objects are retrieved among the top 5 choices and more than half of the objects are retrieved in the first place, Fuzzy Logic-based method is clearly better. 92.45% of the objects are retrieved and more than three-quarters are retrieved in the first place.
### **4.2 Tests according to the type of standard questions**
In order to refine the conclusions about both Term Weighting methods, it is important to make a more thorough analysis of the results. We submitted to both Term Weighting methods to a comprehensive analysis according to the type of standard question. Results are shown in the Table 14.
According to the results, the TF-IDF method works relatively well considering the number of objects retrieved. Though, the Fuzzy Logic-based method is more precise, retrieving 91.67% of the objects in the first place. On the other hand, good results for this type of questions are logical, since questions correspond to supposedly well-made user queries.
For synonymous standard questions, the conclusions are similar: the results obtained using the Fuzzy Logic-based method are better than those achieved with TF-IDF method, especially in regard to precision. Though, the TF-IDF method also ensures good results. However, queries are not precise, so the performance is worse for the TF-IDF method than it is for the Fuzzy Logic-based method. This fact gives an idea of fuzzy logic as an ideal tool for adding more flexibility to the system. Anyway, the results are quite similar to those obtained for the main standard questions. They are only slightly worse, since synonim standard questions are similar to the main standard questions.
The difference is even more noticeable in regard to imprecise standard questions and specific standard questions. Imprecise standard questions are detected nearly as well as the main standard questions in the case of Fuzzy Logic-based method. This is another reason to confirm the appropriateness of using Fuzzy Logic. As for the specific standard questions, we
Term Weighting for Information Retrieval Using Fuzzy Logic 189
Obviously, groups 1 and 2 are more numerous, since it is less common that many questions have the same response. However, the objects from the groups 3 and 4 correspond to a wide range of standard questions, so they are equally important. In Table 15 the number of
Group 1 1 95 Group 2 2-5 108 Group 3 6-10 22 Group 4 > 10 28
To analyze the results, the position in which the required object is retrieved must be considered. We consider the retrieval of most of the standard questions that define that object. For example, if an object is defined by 15 standard questions and, for 10 of them, the object is retrieved in second place, it is considered that the object has actually been retrieved
In short, this study does not focus on the answers to standard questions, but on the correctly retrieved objects. This provides a new element for the system analysis. Results are shown in
For group 1, the results are almost perfect for the Fuzzy Logic-based method, as nearly all the objects are retrieved in the first place (about 94%). However, the TF-IDF method, though not as accurate, resists the comparison. This behaviour is repeated in group 2. The objects are often retrieved by both methods among the top three items. Though, the Fuzzy Logicbased method is better for its accuracy, retrieving over 92% of the objects in the first place. In view of the tests, we conclude that the results are very good for both methods when up to five standard questions are defined. Although the results are better for the novel Fuzzy Logic-based Term Weighting method, they are also quite reasonable for the classical TF-IDF
However, the largest advantage of using Fuzzy Logic for Term Weighting occurs when many standard questions per object are defined, i.e. when the information is confusing, disordered or imprecise. For the case of group 3, where objects are defined by among six and ten standard questions per object type, we observe that there is a significant difference between the TF-IDF classical method and the proposed Fuzzy Logic-based method. Although both methods retrieve all the objects, there is a big difference in the way they are retrieved, especially on the accuracy of the information extraction. 86% of the objects are retrieved in first place using the Fuzzy Logic-based method, while only 45% using the TF-
**object Number of objects**
**Group number Number of standard questions per**
Table 15. Groups according to the number of standard questions per object.
objects for each of these groups is defined.
in second place.
Term Weighting method.
IDF classical method.
Table 16.
Table 14. Information Retrieval results of using both Term Weighting methods, according to the type of standard question.
get the worst result by far among all classes of standard questions. This is a logical fact, considering that these questions are associated with the main standard question, but it is more concrete. In fact, it is usual for such specific questions to belong to a list within a whole. This way, there may be objects that are more related to the query than the requiered object itself. This is hardly a drawback, since both objects are retrieved to the user - the more specific one and the more general one -. The own user must choose which one is the most accurate. This case shows more clearly that the fact of using Fuzzy Logic allows the user to extract a larger number of objects.
### **4.3 Tests according to the number of standard questions**
Another aspect to consider in the analysis of the results is the number of standard questions assigned to every object. Obviously, an object that is well defined by a single standard question is very specific. Thus, it is easy to extract the object from the complete set of knowledge. However, there are objects that contain very vague or imprecise information, making it necessary to define several standard questions for every object. For this study, the objects are grouped into the following:
188 Fuzzy Logic – Algorithms, Techniques and Implementations
**question Cat1 Cat2 Cat3 Cat4 Cat5 Total**
(23.02%) 6 (2.38%) 6 (2.38%) 11 (4.37%) 252
(27.92%) 13 (4.22%) 15 (4.87%) 17 (5.52%) 308
(13.31%) 3 (0.97%) 5 (1.62%) 47 (2.27%) 308
(25.60%) 6 (4.80%) 1 (0.80%) 12 (9.60%) 125
(23.14%) 24 (10.48%) 23(10.04%) 22 (9.61%) 229
(22.71%) <sup>229</sup>
(91.67%) 13 (5.16%) 2 (0.79%) 0 (0.00 %) 6 (2.38%) 252
(88.80%) 5 (4.00%) 0 (0.00 %) 0 (0.00 %) 9 (7.20%) 125
(21.40%) 26(11.35%) 55(24.01%) <sup>52</sup>
58
86
41
32
49
53
Table 14. Information Retrieval results of using both Term Weighting methods, according to
get the worst result by far among all classes of standard questions. This is a logical fact, considering that these questions are associated with the main standard question, but it is more concrete. In fact, it is usual for such specific questions to belong to a list within a whole. This way, there may be objects that are more related to the query than the requiered object itself. This is hardly a drawback, since both objects are retrieved to the user - the more specific one and the more general one -. The own user must choose which one is the most accurate. This case shows more clearly that the fact of using Fuzzy Logic allows the user to
Another aspect to consider in the analysis of the results is the number of standard questions assigned to every object. Obviously, an object that is well defined by a single standard question is very specific. Thus, it is easy to extract the object from the complete set of knowledge. However, there are objects that contain very vague or imprecise information, making it necessary to define several standard questions for every object. For this study, the
**Type of standard**
TF-IDF Method
Fuzzy Logicbased method
TF-IDF Method
Fuzzy Logicbased method
TF-IDF Method
Fuzzy Logicbased method
TF-IDF Method
Fuzzy Logicbased method
the type of standard question.
extract a larger number of objects.
objects are grouped into the following:
171 (67.86%)
231
177 (57.46%)
252 (81.82%)
74 (59.20%)
111
46 (20.08%)
107 (46.72%)
**4.3 Tests according to the number of standard questions**
Main standard questions
Synonim standard questions
Imprecise standard questions
Specific standard questions
Obviously, groups 1 and 2 are more numerous, since it is less common that many questions have the same response. However, the objects from the groups 3 and 4 correspond to a wide range of standard questions, so they are equally important. In Table 15 the number of objects for each of these groups is defined.
Table 15. Groups according to the number of standard questions per object.
To analyze the results, the position in which the required object is retrieved must be considered. We consider the retrieval of most of the standard questions that define that object. For example, if an object is defined by 15 standard questions and, for 10 of them, the object is retrieved in second place, it is considered that the object has actually been retrieved in second place.
In short, this study does not focus on the answers to standard questions, but on the correctly retrieved objects. This provides a new element for the system analysis. Results are shown in Table 16.
For group 1, the results are almost perfect for the Fuzzy Logic-based method, as nearly all the objects are retrieved in the first place (about 94%). However, the TF-IDF method, though not as accurate, resists the comparison. This behaviour is repeated in group 2. The objects are often retrieved by both methods among the top three items. Though, the Fuzzy Logicbased method is better for its accuracy, retrieving over 92% of the objects in the first place. In view of the tests, we conclude that the results are very good for both methods when up to five standard questions are defined. Although the results are better for the novel Fuzzy Logic-based Term Weighting method, they are also quite reasonable for the classical TF-IDF Term Weighting method.
However, the largest advantage of using Fuzzy Logic for Term Weighting occurs when many standard questions per object are defined, i.e. when the information is confusing, disordered or imprecise. For the case of group 3, where objects are defined by among six and ten standard questions per object type, we observe that there is a significant difference between the TF-IDF classical method and the proposed Fuzzy Logic-based method. Although both methods retrieve all the objects, there is a big difference in the way they are retrieved, especially on the accuracy of the information extraction. 86% of the objects are retrieved in first place using the Fuzzy Logic-based method, while only 45% using the TF-IDF classical method.
Term Weighting for Information Retrieval Using Fuzzy Logic 191
neuro-fuzzy techniques represent a very interesting field, as they combine human reasoning provided by Fuzzy Logic and the connection-based structure of Artificial Neural Networks, taking advantage of both techniques. One possible application is the creation of fuzzy rules
Another possible future direction is to check the validity of this method in other
The difficulty to distinguish the necessary information from the huge quantity of unnecessary data has enhanced the use of Information Retrieval recently. Especially, the socalled Vector Space Model is much extended. Vector Space Model is based on the use of index terms. These index terms are associated with certain weights, which represent the importance of these terms in the considered set of knowledge. In this chapter, we propose the development of a novel automatic Fuzzy Logic-based Term Weighting method for Vector Space Model. This method improves the TF-IDF Term Weighting classic method for its flexibility. The use of Fuzzy Logic is very appropiate in heterogeneous, vague, imprecise,
Fuzzy Logic-based method is similar to TF-IDF, but also considers two aspects that the TF-IDF does not: the degree of identification of the object if a determined index term is solely used in a query; and the existance of join index terms. Term Weighting is automatic. The level of expertise required is low, so there is no need for an operator of any kind of knowledge about Fuzzy Logic. Therefore, an operator only has to know how many times an
Although the results obtained with the TF-IDF method are quite reasonable, Fuzzy Logicbased method is clearly superior. Especially when user queries are not equal to the standard query or they are imprecise, we observe that the performance declines more for the TF-IDF method than for the Fuzzy Logic-based method. This fact gives us an idea of how suitable is
Lertnattee, V. & Theeramunkong, T. (2003). Combining homogenous classifiers for centroid-
Lee, D.L., Chuang, H., Seamons, K., 1997. *Document ranking and the vector-space model*. IEEE
Liu, S., Dong, M., Zhang, H., Li, R. & Shi, Z. (2001). An approach of multi-hierarchy text
Raghavan, V.V. & Wong, S.K. (1986). A critical analysis of vector space model for
Ruiz, M. & Srinivasan, P. (1998). Automatic Text Categorization Using Neural Networks*.*
based text classification. *Proceedings of the 7th International Symposium on Computers*
classification. *Proceedings of the International Conferences on Info-tech and Info-net*.
information retrieval. *Journal of the American Society for Information Science*, Vol.37
*Advances in Classification Research vol. 8: Proceedings of the 8th ASIS SIG/CR*
index term appears in a certain subset and the answer to two simple questions.
the use of Fuzzy Logic to add more flexibility to an Information Retrieval system.
by means of an Artificial Neural Network system.
or not in order information environments.
*and Communications*, pp. 1034-1039.
Beijing. Vol 3, pp. 95 – 100.
(5), p. 279-87.
Software, Vol. 14, Issue 2, pp. 67 – 75.
**6. Conclusion**
**7. References**
environments containing inaccurate, vague and heterogeneous data.
Table 16. Information Retrieval results of using both Term Weighting methods, according to the number of standard questions per object.
The difference is even more marked when more than ten standard questions per object are defined. In this case, it is obvious that none of the questions clearly define the object, so that information is clearly vague. While using the Fuzzy Logic-based method, more than 96% of the objects are retrieved - with 75% of them in the first place -, with the TF-IDF method correctly, only 82% of the objects are retrieved. Furthermore, only 35.7% of these objects are extracted in the first place.
In view of the table, we observe that the more standard questions per object, the better the results of the Fuzzy Logic-based method, compared with those obtained with the classical TF-IDF method. Therefore, the obvious conclusion is that the more convoluted, messy and confusing is the information, the better the Fuzzy Logic-based Term Weighting method is compared to the classical one. This makes Fuzzy Logic-based Term Weighting an ideal tool for the case of information extraction in a web portal.
### **5. Future research directions**
We suggest the application of other Computational Intelligence techniques apart from Fuzzy Logic for Term Weighting. Among these techniques, we believe that the so-called neuro-fuzzy techniques represent a very interesting field, as they combine human reasoning provided by Fuzzy Logic and the connection-based structure of Artificial Neural Networks, taking advantage of both techniques. One possible application is the creation of fuzzy rules by means of an Artificial Neural Network system.
Another possible future direction is to check the validity of this method in other environments containing inaccurate, vague and heterogeneous data.
### **6. Conclusion**
190 Fuzzy Logic – Algorithms, Techniques and Implementations
**Cat1 Cat2 Cat3 Cat4 Cat5 Total**
74 (77.89%) 16 (16.84%) 1 (1.05%) 1 (1.05%) 3 (3.16%) 95
89 (93.68%) 3 (3.16%) 2 (2.10%) 0 (0.00 %) 1 (1.05%) 95
86 (79.63%) 21 (19.44%) 1 (0.93%) 0 (0.00 %) 0 (0.00 %) 108
100 (92.59%) 7 (6.48%) 0 (0.00 %) 0 (0.00 %) 1 (0.93%) 108
10 (45.45%) 9 (40.91%) 3 (13.63%) 0 (0.00 %) 0 (0.00 %) 22
19 (86.36%) 3 (13.63%) 0 (0.00 %) 0 (0.00 %) 0 (0.00 %) 22
10 (35.71%) 10 (35.71%) 3 (10.71%) 2 (7.14%) 3 (10.71%) 28
21 (75.00%) 4 (14.29%) 1 (3.57%) 1 (3.57%) 1 (3.57%) 28
Table 16. Information Retrieval results of using both Term Weighting methods, according to
The difference is even more marked when more than ten standard questions per object are defined. In this case, it is obvious that none of the questions clearly define the object, so that information is clearly vague. While using the Fuzzy Logic-based method, more than 96% of the objects are retrieved - with 75% of them in the first place -, with the TF-IDF method correctly, only 82% of the objects are retrieved. Furthermore, only 35.7% of these objects are
In view of the table, we observe that the more standard questions per object, the better the results of the Fuzzy Logic-based method, compared with those obtained with the classical TF-IDF method. Therefore, the obvious conclusion is that the more convoluted, messy and confusing is the information, the better the Fuzzy Logic-based Term Weighting method is compared to the classical one. This makes Fuzzy Logic-based Term Weighting an ideal tool
We suggest the application of other Computational Intelligence techniques apart from Fuzzy Logic for Term Weighting. Among these techniques, we believe that the so-called
**Type of standard**
TF-IDF Method
Fuzzy Logicbased method
TF-IDF Method
Fuzzy Logicbased method
TF-IDF Method
Fuzzy Logicbased method
TF-IDF Method
Fuzzy Logicbased method
extracted in the first place.
**5. Future research directions**
the number of standard questions per object.
for the case of information extraction in a web portal.
**question**
**Group 1**
**Group 2**
**Group 3**
**Group 4**
The difficulty to distinguish the necessary information from the huge quantity of unnecessary data has enhanced the use of Information Retrieval recently. Especially, the socalled Vector Space Model is much extended. Vector Space Model is based on the use of index terms. These index terms are associated with certain weights, which represent the importance of these terms in the considered set of knowledge. In this chapter, we propose the development of a novel automatic Fuzzy Logic-based Term Weighting method for Vector Space Model. This method improves the TF-IDF Term Weighting classic method for its flexibility. The use of Fuzzy Logic is very appropiate in heterogeneous, vague, imprecise, or not in order information environments.
Fuzzy Logic-based method is similar to TF-IDF, but also considers two aspects that the TF-IDF does not: the degree of identification of the object if a determined index term is solely used in a query; and the existance of join index terms. Term Weighting is automatic. The level of expertise required is low, so there is no need for an operator of any kind of knowledge about Fuzzy Logic. Therefore, an operator only has to know how many times an index term appears in a certain subset and the answer to two simple questions.
Although the results obtained with the TF-IDF method are quite reasonable, Fuzzy Logicbased method is clearly superior. Especially when user queries are not equal to the standard query or they are imprecise, we observe that the performance declines more for the TF-IDF method than for the Fuzzy Logic-based method. This fact gives us an idea of how suitable is the use of Fuzzy Logic to add more flexibility to an Information Retrieval system.
### **7. References**
**10**
Amin Parvizi *University of Malaya,*
*Malaysia*
**Artificial Intelligence Techniques of Estimating**
Switched reluctance motor (SRM) is one of the best candidates for industrial and household applications. Owing to its superior abilities such as high torque to inertia ratio, easy cooling, high speed capability and ease of repair, SRM has been taken into consideration by researchers. one of the major difficulties is the nonlinear relation between current, rotor position and flux linkage. Due to the mentioned nonlinearity, it is essential to have an accurate model to deal with nonlinear characteristics of SRM. The essence of this research work is to develop the SRM model based on artificial techniques (AI) such as fuzzy logic, adaptive neuro-fuzzy. In the papers (Chancharoensook& Rahman,2002;Geldhof&Van den Bossche& Vyncke&Melkebeek,2008; Mirzaeian-Dehkordi& Moallem, 2006; Gobbi, Ramar;2008; Rajapakse& Gole& Muthumuni& Wilson& Perregaux,2004; Wai-Chuen Gan& Cheung& Li Qiu,2008) SRM models
In this short communication, rule based system are considered in order to find a model to deal with nonlinear characteristics of SRM. We call it rule based due to have fixed data point. Fuzzy logic and adaptive neuro-fuzzy are employed to develop a comprehensive model for nonlinear characteristics of 8:6 SRM. Torque profile is simulated based on fuzzy logic, adaptive neuro-fuzzy techniques via MATLAB software. In the line above, error analysis is conducted for those models. Data is tabled and compared with the published data. The result of error analysis reflects the precision of the method and the capability of
Switched reluctance motor (SRM) is a type of synchronous machine. Figure 1 shows the classification of the SRM. This initial classification is made by considering the method of
Stator and rotor are two basic parts of SRM. One of the most important features of the SRM comes back to its simple structure. This type of electrical machine has no winding or magnet in rotor part. Both of stator and rotor have salient poles. Thus, it is named double salient
**1. Introduction**
presented based on the look-up tables.
the approach for the further simulation.
machine. Figure 2 shows the typical structure of SRM.
**2. Background theory**
movement.
**of Torque for 8:6 Switched Reluctance Motor**
*Classification Research Workshop*. Ed. Efthimis Efthimiadis. Information Today, Medford:New Jersey, pp 59-72.
Salton, G. (1988). *Automatic Text Processing*. Addison-Wesley Publishing Company.
Salton, G. & Buckley, C. (1996). Term Weighting Approaches in Automatic Text Retrieval*. Technical Report TR87-881, Department of Computer Science, Cornell University, 1987. Information Processing and Management* Vol.32 (4), pp. 431-443.
Van Rijsbergen, C.J. (1979). *Information retrieval.* Butterworths.
Zhao, Y. & Karypis, G. (2002). Improving precategorized collection retrieval by using supervised term weighting schemes. *Proceedings of the International Conference on Information Technology: Coding and Computing*, pp 16 – 21.
## **Artificial Intelligence Techniques of Estimating of Torque for 8:6 Switched Reluctance Motor**
Amin Parvizi *University of Malaya, Malaysia*
### **1. Introduction**
192 Fuzzy Logic – Algorithms, Techniques and Implementations
Salton, G. & Buckley, C. (1996). Term Weighting Approaches in Automatic Text Retrieval*.*
Zhao, Y. & Karypis, G. (2002). Improving precategorized collection retrieval by using
Salton, G. (1988). *Automatic Text Processing*. Addison-Wesley Publishing Company.
*Information Processing and Management* Vol.32 (4), pp. 431-443.
*Information Technology: Coding and Computing*, pp 16 – 21.
Medford:New Jersey, pp 59-72.
Van Rijsbergen, C.J. (1979). *Information retrieval.* Butterworths.
*Classification Research Workshop*. Ed. Efthimis Efthimiadis. Information Today,
*Technical Report TR87-881, Department of Computer Science, Cornell University, 1987.*
supervised term weighting schemes. *Proceedings of the International Conference on*
Switched reluctance motor (SRM) is one of the best candidates for industrial and household applications. Owing to its superior abilities such as high torque to inertia ratio, easy cooling, high speed capability and ease of repair, SRM has been taken into consideration by researchers. one of the major difficulties is the nonlinear relation between current, rotor position and flux linkage. Due to the mentioned nonlinearity, it is essential to have an accurate model to deal with nonlinear characteristics of SRM. The essence of this research work is to develop the SRM model based on artificial techniques (AI) such as fuzzy logic, adaptive neuro-fuzzy. In the papers (Chancharoensook& Rahman,2002;Geldhof&Van den Bossche& Vyncke&Melkebeek,2008; Mirzaeian-Dehkordi& Moallem, 2006; Gobbi, Ramar;2008; Rajapakse& Gole& Muthumuni& Wilson& Perregaux,2004; Wai-Chuen Gan& Cheung& Li Qiu,2008) SRM models presented based on the look-up tables.
In this short communication, rule based system are considered in order to find a model to deal with nonlinear characteristics of SRM. We call it rule based due to have fixed data point. Fuzzy logic and adaptive neuro-fuzzy are employed to develop a comprehensive model for nonlinear characteristics of 8:6 SRM. Torque profile is simulated based on fuzzy logic, adaptive neuro-fuzzy techniques via MATLAB software. In the line above, error analysis is conducted for those models. Data is tabled and compared with the published data. The result of error analysis reflects the precision of the method and the capability of the approach for the further simulation.
### **2. Background theory**
Switched reluctance motor (SRM) is a type of synchronous machine. Figure 1 shows the classification of the SRM. This initial classification is made by considering the method of movement.
Stator and rotor are two basic parts of SRM. One of the most important features of the SRM comes back to its simple structure. This type of electrical machine has no winding or magnet in rotor part. Both of stator and rotor have salient poles. Thus, it is named double salient machine. Figure 2 shows the typical structure of SRM.
Artificial Intelligence Techniques of Estimating of Torque for 8:6 Switched Reluctance Motor 195
Fig. 3. Operation of switched reluctance motor a)Phase c aligned b) Phase a aligned
potential contour.
**4. Single – Phase SRMs**
in unaligned position.
toward itself. This process subsequently will be continued.
Following figure shows the lamination profile of 8:6 SRM in align position with magnetic
Fig. 4. Lamination profile of 8:6 SRM (Parvizi, Hassani, Mehbodnia, Makhilef, & Tamjis, 2009)
During the past years, single-phase SRMs have attracted much attention due to resemblance to universal and single phase induction machines and also, single-phase SRMs are low-cost manufacture as well as induction and universal machines. Specific applications of singlephase SRMs come up in where high-speed motors are needed. When the stator and rotor poles are in front of each other which means the align position, the current that relevant to stator phase is turned off and the rotor keeps moving toward the adjacent stator pole due to kinetic energy which is stored. Adjacent stator phase is energized to attract the rotor pole
The major problem of single-phase SRMs operation come up when the rotor and stator are in align position at the instant of starting or the rotor at a position where the load torque at the starting is greater than the produced load. Permanent magnet has been used as a solution. It pulls the rotor away from the stator or at the right position in which motor can produce a torque greater than the load torque. As the figure 5 shows the rotor and stator are
Fig. 1. Classification of the SRM
The number under the configuration (6/4 or 8/6) means SRM with 6 or 8 poles on stator and 4 or 6 poles on rotor.
### **3. Operation of the SRMs**
The key of understanding rotor movement is rising from the tendency of rotor to place in minimum reluctance position at the instance of excitation. While two rotor poles are in front of two stator poles, called align position. In align position; another set of rotor pole is out of alignment position there for another set of stator pole will be excited to move the rotor poles until the time to reach minimum reluctance. Figure 3 shows a 6:4 SRM. In the figure, at the first situation, suppose that �� and �� are two poles of rotor and in align position with *́ c* and ć which are the stator poles. When *a* is excited in the direction that is shown, stator poles tends to pull the rotor poles toward itself. Therefore, ��and ��́ are in front of the a and á, respectively. After they are aligned, the stator current is turned off and the corresponding situation is shown in Figure 3(b). Now, b is excited and pulls the ��and �� toward b ́ ́ and b, respectively. Hence, the rotor is rotating in a clockwise direction.
Fig. 3. Operation of switched reluctance motor a)Phase c aligned b) Phase a aligned
Following figure shows the lamination profile of 8:6 SRM in align position with magnetic potential contour.
Fig. 4. Lamination profile of 8:6 SRM (Parvizi, Hassani, Mehbodnia, Makhilef, & Tamjis, 2009)
### **4. Single – Phase SRMs**
194 Fuzzy Logic – Algorithms, Techniques and Implementations
The number under the configuration (6/4 or 8/6) means SRM with 6 or 8 poles on stator and
The key of understanding rotor movement is rising from the tendency of rotor to place in minimum reluctance position at the instance of excitation. While two rotor poles are in front of two stator poles, called align position. In align position; another set of rotor pole is out of alignment position there for another set of stator pole will be excited to move the rotor poles until the time to reach minimum reluctance. Figure 3 shows a 6:4 SRM. In the figure, at the first situation, suppose that �� and �� are two poles of rotor and in align position with *́ c* and ć which are the stator poles. When *a* is excited in the direction that is shown, stator poles tends to pull the rotor poles toward itself. Therefore, ��and ��́ are in front of the a and á, respectively. After they are aligned, the stator current is turned off and the corresponding situation is shown in Figure 3(b). Now, b is excited and pulls the ��and �� toward b ́ ́ and b,
respectively. Hence, the rotor is rotating in a clockwise direction.
Fig. 1. Classification of the SRM
4 or 6 poles on rotor.
Fig. 2. SRM configuration
**3. Operation of the SRMs**
During the past years, single-phase SRMs have attracted much attention due to resemblance to universal and single phase induction machines and also, single-phase SRMs are low-cost manufacture as well as induction and universal machines. Specific applications of singlephase SRMs come up in where high-speed motors are needed. When the stator and rotor poles are in front of each other which means the align position, the current that relevant to stator phase is turned off and the rotor keeps moving toward the adjacent stator pole due to kinetic energy which is stored. Adjacent stator phase is energized to attract the rotor pole toward itself. This process subsequently will be continued.
The major problem of single-phase SRMs operation come up when the rotor and stator are in align position at the instant of starting or the rotor at a position where the load torque at the starting is greater than the produced load. Permanent magnet has been used as a solution. It pulls the rotor away from the stator or at the right position in which motor can produce a torque greater than the load torque. As the figure 5 shows the rotor and stator are in unaligned position.
Artificial Intelligence Techniques of Estimating of Torque for 8:6 Switched Reluctance Motor 197
Figure 7 shows a magnetization curve under specific condition. Rotor angle is locked in somewhere between aligned and unaligned position. Energy and co-energy are defined in
�� = � ���(�� �) ��
�� <sup>=</sup> � �(�� �)�� ��
�(�� �) represents the flux linkage as a nonlinear function of rotor position and current. As figure 7 shows, the area behind the magnetization curve until ��called stored field energy (��) that this energy is stored in the iron core (rotor and stator) and in the air gap. The area
In the next step, suppose that rotor is released. In this situation, rotor moves toward the adjacent stator pole until place in align position. For an infinitesimal movement� , suppose that �� is considered constant, thus the flux linkage changes from point A to point B as shown in figure 8. By considering the conservation of energy, the mechanical work �� which has been done by rotor during the � movement is equal to the change in the stored
The area �� equals to the change in co-energy because of the � movement. Thus, the
under the magnetization curve until �� called co-energy (�� ) .
Fig. 7. Concept of stored field energy and co-energy
field energy (��).
Fig. 8. Mechanical work area
mechanical energy can be stated as following:
� (1)
� (2)
any point of that respectively by:
Fig. 5. Single-phase SRM with permanent magnet
Maximum duty cycle of single-phase SRM is 0.5, thus, noise and high ripple torque are deduced from a torque discontinuity which arises from duty cycle. Applications, in which torque ripple and noise are not important, are good for this machine such as home appliances and hand tool.
### **5. Magnetization characteristic of SRM**
Due to saturation and varying reluctance with rotor position, there is no simple analytical solution to express the field which is produced by phase winding. Energy conversion approach that is presented in is used to analyze energy conversion.
Figure 6 shows a typical magnetization curve. Flux linkage is a function of both rotor position and excitation current and also, it is nonlinear function. One of the most important parameter which affects on flux linkage is air gap. As it can be seen clearly, in unaligned position, flux linkage is a linear function due to big air gap. In other words, the gap between stator pole and rotor pole is big. In contrast, in aligned position, due to small air gap, the magnetization curve is heavily saturated.
Fig. 6. Magnetization curve for SRM
Figure 7 shows a magnetization curve under specific condition. Rotor angle is locked in somewhere between aligned and unaligned position. Energy and co-energy are defined in any point of that respectively by:
$$\mathcal{W}\_f = \int\_0^{\lambda\_a} i d\lambda (i, \theta) \tag{1}$$
$$
\dot{W} = \int\_0^{l\_a} \lambda(\theta, l) dl \tag{2}
$$
�(�� �) represents the flux linkage as a nonlinear function of rotor position and current. As figure 7 shows, the area behind the magnetization curve until ��called stored field energy (��) that this energy is stored in the iron core (rotor and stator) and in the air gap. The area under the magnetization curve until �� called co-energy (�� ) .
Fig. 7. Concept of stored field energy and co-energy
In the next step, suppose that rotor is released. In this situation, rotor moves toward the adjacent stator pole until place in align position. For an infinitesimal movement� , suppose that �� is considered constant, thus the flux linkage changes from point A to point B as shown in figure 8. By considering the conservation of energy, the mechanical work �� which has been done by rotor during the � movement is equal to the change in the stored field energy (��).
Fig. 8. Mechanical work area
196 Fuzzy Logic – Algorithms, Techniques and Implementations
Maximum duty cycle of single-phase SRM is 0.5, thus, noise and high ripple torque are deduced from a torque discontinuity which arises from duty cycle. Applications, in which torque ripple and noise are not important, are good for this machine such as home
Due to saturation and varying reluctance with rotor position, there is no simple analytical solution to express the field which is produced by phase winding. Energy conversion
Figure 6 shows a typical magnetization curve. Flux linkage is a function of both rotor position and excitation current and also, it is nonlinear function. One of the most important parameter which affects on flux linkage is air gap. As it can be seen clearly, in unaligned position, flux linkage is a linear function due to big air gap. In other words, the gap between stator pole and rotor pole is big. In contrast, in aligned position, due to small air gap, the
Fig. 5. Single-phase SRM with permanent magnet
**5. Magnetization characteristic of SRM**
magnetization curve is heavily saturated.
Fig. 6. Magnetization curve for SRM
approach that is presented in is used to analyze energy conversion.
appliances and hand tool.
The area �� equals to the change in co-energy because of the � movement. Thus, the mechanical energy can be stated as following:
$$
\Delta W\_m = \Delta \dot{W} = \int\_0^{l\_a} \lambda(\theta\_B, l) di - \int\_0^{l\_a} \lambda(\theta\_A, l) di \tag{3}
$$
$$
\Delta W\_m = T \Delta \theta \tag{4}
$$
$$T = \frac{\Delta W\_{\rm m}}{\Delta \theta} = \frac{\int\_0^{l\_0} \lambda(\theta\_B, l) dl - \int\_0^{l\_0} \lambda(\theta\_A, l) dl}{\Delta \theta} \tag{5}$$
$$T = \frac{\partial}{\partial \theta} \int\_0^l \lambda(\theta, \mathbf{f}) d\mathbf{f} \tag{6}$$
$$\lambda(\theta, t) = L(\theta)t \tag{7}$$
$$T = \int\_0^l \frac{\partial \lambda(\theta, \mathbf{i})}{\partial \theta} d\mathbf{i} = \int\_0^l \frac{dL}{d\theta} \mathbf{i} d\mathbf{i} = \frac{dL}{d\theta} \int\_0^l \mathbf{i} \, d\mathbf{i} = \frac{1}{2} l^2 \frac{dL}{d\theta} \tag{8}$$
Artificial Intelligence Techniques of Estimating of Torque for 8:6 Switched Reluctance Motor 201
The most important part of the modeling is the constructing FIS rules because of the outcome of this part will define output fuzzy set. In other words, torque as the output is a fuzzy set that basically are formed by the results of the constructing FIS rules. Table 2 is used to constructing FIS rules. In Mamdani's type a set of if-then called rules. Thus, the
Degree of membership function is a value between 0 and 1 which is the output of the membership function. Now, degree of a rule can be defined the multiple of the degree of the
FIS editor is used to produce the rules which are shown in figure 15. Also, figure 15 shows a
inputs and output. For example, degree of the mentioned rule can be as following:
conditional statements are formulated by if-then form. For instance,
Rule 1: If current is s6 and rotor angle is s4 then torque is s10
part of rules for 8:6 SRM and the number of total rules are 80.
Fig. 12. Membership functions for current
Fig. 13. Membership functions for rotor angle
Fig. 14. Membership functions for torque
Degree (Rule 1) = µ (s6).µ (s4).µ (s10)
**6.4 Constructing FIS rule**
Table 2. Fuzzy rule base table for 8:6 SRM
It should be noted that wrong interpretation of each of rules will influence on overall output and as a result, wrong model will come up. Therefore, this part of work should be done carefully and without any wrong rule.
### **6.2 Formation of Fuzzy Inference System (FIS)**
For the formation of the fuzzy inference system, fuzzy logic toolbox of MATLAB is used. Figure 11 shows the FIS structure that current and rotor angle are the inputs and torque is the output.
Fig. 11. Fuzzy SRM FIS structure
### **6.3 Assigning the FIS membership functions**
Once the FIS structure is completed, membership functions for each of the inputs and the output will be formed. Toolbox of MATLAB has 11 built-in membership functions (MFs) which some of those are trimf, gbellmf, gaussmf, gauss2mf, sigmf, psigmf. One of these MFs that are formed by straight lines is called triangular MFs. These MFs are used here in account of simple structure and well suited for the modeling.
Current and rotor angle as the inputs have 8 MFs and 13 MFs respectively for itself and torque as the output has 21 MFs for 8:6 SRM. Figure 12 , figure 13 and figure 14 show the MF for current, rotor angle and torque, respectively.
Fig. 12. Membership functions for current
200 Fuzzy Logic – Algorithms, Techniques and Implementations
s4 s3 s2 s1 M b1 b2 b3 s6 s10 s10 s10 s10 s10 s10 s10 s10 s5 s10 s10 s10 s9 s9 s8 s8 s8 s4 s10 s9 s9 s8 s8 s7 s7 s6 s3 s9 s8 s8 s7 s6 s5 s3 s2 s2 s8 s7 s5 s2 b1 b3 b6 b9 s1 s8 s7 s4 s1 b2 b5 b7 b10 M s8 s7 s4 s1 b2 b5 b7 b10 b1 s8 s7 s4 s1 b1 b4 b7 b10 b2 s8 s7 s5 s2 b1 b4 b6 b8 b3 s8 s7 s5 s2 M b2 b4 b6 b4 s8 s7 s5 s4 s2 M b1 b2 b5 s9 s8 s6 s5 s5 s4 s3 s3 b6 s10 s10 s10 s10 s10 s10 s10 s10
It should be noted that wrong interpretation of each of rules will influence on overall output and as a result, wrong model will come up. Therefore, this part of work should be done
For the formation of the fuzzy inference system, fuzzy logic toolbox of MATLAB is used. Figure 11 shows the FIS structure that current and rotor angle are the inputs and torque is
Once the FIS structure is completed, membership functions for each of the inputs and the output will be formed. Toolbox of MATLAB has 11 built-in membership functions (MFs) which some of those are trimf, gbellmf, gaussmf, gauss2mf, sigmf, psigmf. One of these MFs that are formed by straight lines is called triangular MFs. These MFs are used here in
Current and rotor angle as the inputs have 8 MFs and 13 MFs respectively for itself and torque as the output has 21 MFs for 8:6 SRM. Figure 12 , figure 13 and figure 14 show the
Table 2. Fuzzy rule base table for 8:6 SRM
carefully and without any wrong rule.
Fig. 11. Fuzzy SRM FIS structure
**6.3 Assigning the FIS membership functions**
account of simple structure and well suited for the modeling.
MF for current, rotor angle and torque, respectively.
the output.
**6.2 Formation of Fuzzy Inference System (FIS)**
Fig. 13. Membership functions for rotor angle
Fig. 14. Membership functions for torque
### **6.4 Constructing FIS rule**
The most important part of the modeling is the constructing FIS rules because of the outcome of this part will define output fuzzy set. In other words, torque as the output is a fuzzy set that basically are formed by the results of the constructing FIS rules. Table 2 is used to constructing FIS rules. In Mamdani's type a set of if-then called rules. Thus, the conditional statements are formulated by if-then form. For instance,
Rule 1: If current is s6 and rotor angle is s4 then torque is s10
Degree of membership function is a value between 0 and 1 which is the output of the membership function. Now, degree of a rule can be defined the multiple of the degree of the inputs and output. For example, degree of the mentioned rule can be as following:
Degree (Rule 1) = µ (s6).µ (s4).µ (s10)
FIS editor is used to produce the rules which are shown in figure 15. Also, figure 15 shows a part of rules for 8:6 SRM and the number of total rules are 80.
Artificial Intelligence Techniques of Estimating of Torque for 8:6 Switched Reluctance Motor 203
By considering the graphical representative of torque (figure 9), rotor angle and current are defined as inputs and torque as output. Figure 17 shows the FIS editor for ANFIS that two
Once the data set are obtained from the torque curve, loading data starts. The loaded data set should be in three columns matrix format. First and second belong to the inputs and the
The training data appears in the plot in center of the ANFIS editor as a set of circle as shown
Once data set is loaded, next step is initializing the MFs. There are two partitioning method
**7.1 Forming ANFIS**
inputs and one output has been shown clearly.
Fig. 17. Neuro-Fuzzy SRM FIS structure
**7.2 Training scheme of FIS**
in figure 18 for 8:6 SRM.
third one present the torque data.
Fig. 18. ANFIS editor with training data loaded for 8:6 SRM
**7.3 Initializing and generating FIS**
for initializing the MFs:
Fig. 15. Constructing rules using rule editor
The surface viewer is used to show the dependency of the output to both of inputs. There for, it generates torque surface map. Figure 16 shows the torque for 8:6.
Fig. 16. Surface viewer of the FIS for 8:6 SRM
### **7. Torque estimation model using Adaptive Neuro-Fuzzy Inference System (ANFIS)**
ANFIS is an acronym of Adaptive Neuro-Fuzzy Inference System. Adaptive Neuro-Fuzzy is a technique which provides a learning method from the desired input and output to adjust the MFs parameters. During this process, back propagation and hybrid are two algorithms that are used so that the best parameters for the MFs will be achieved.
### **7.1 Forming ANFIS**
202 Fuzzy Logic – Algorithms, Techniques and Implementations
The surface viewer is used to show the dependency of the output to both of inputs. There
**7. Torque estimation model using Adaptive Neuro-Fuzzy Inference System**
that are used so that the best parameters for the MFs will be achieved.
ANFIS is an acronym of Adaptive Neuro-Fuzzy Inference System. Adaptive Neuro-Fuzzy is a technique which provides a learning method from the desired input and output to adjust the MFs parameters. During this process, back propagation and hybrid are two algorithms
for, it generates torque surface map. Figure 16 shows the torque for 8:6.
Fig. 15. Constructing rules using rule editor
Fig. 16. Surface viewer of the FIS for 8:6 SRM
**(ANFIS)**
By considering the graphical representative of torque (figure 9), rotor angle and current are defined as inputs and torque as output. Figure 17 shows the FIS editor for ANFIS that two inputs and one output has been shown clearly.
Fig. 17. Neuro-Fuzzy SRM FIS structure
### **7.2 Training scheme of FIS**
Once the data set are obtained from the torque curve, loading data starts. The loaded data set should be in three columns matrix format. First and second belong to the inputs and the third one present the torque data.
The training data appears in the plot in center of the ANFIS editor as a set of circle as shown in figure 18 for 8:6 SRM.
Fig. 18. ANFIS editor with training data loaded for 8:6 SRM
### **7.3 Initializing and generating FIS**
Once data set is loaded, next step is initializing the MFs. There are two partitioning method for initializing the MFs:
Artificial Intelligence Techniques of Estimating of Torque for 8:6 Switched Reluctance Motor 205
Figure 21 shows the ANFIS model structure for 8:6 SRM. There are two inputs (rotor angle
and current) and one output (torque). There are total 104 MFs for each of inputs.
The summarized modeling description is shown in table 3 for 8:6 SRM.
Modeling Description Setting Number of Inputs 2 Number of Output 1
Number of MFs 104 Optimized Method Hybrid Epochs 150
The mapping surface of 8:6 SRM using neuro-fuzzy technique is shown in figure 19.
Method Subtractive Clustering
Fig. 20. ANFIS training with hybrid method
Fig. 21. ANFIS model structure for 8:6 SRM
Table 3. Fuzzy rule base table for 8:6 SRM
**7.5 Viewing ANFIS structure**
Second method is employed in account of having one-pass algorithm which estimates the number of clusters.
Figure 19 shows the cluster parameters which are:
For varying both of inputs "range of influence" is set to 0.15. Other parameters remain in their previous value because those values are acceptable for training scheme. Once the parameters are set, the outcome FIS generates 104 numbers of MFs for both of the inputs, and output.
Fig. 19. Parameters set for subtractive clustering
### **7.4 ANFIS training**
In order to optimize the obtained parameters, two methods are available:
The first method is considered for data training. Error tolerance is established to create halt criterion. The error training will stop after certain epoch which is set. The number of epochs for both of 8:6 is 150.
The final error training is 3.014e -7 which is shown in figure 18 after 150 epochs.
Fig. 20. ANFIS training with hybrid method
### **7.5 Viewing ANFIS structure**
204 Fuzzy Logic – Algorithms, Techniques and Implementations
Second method is employed in account of having one-pass algorithm which estimates the
For varying both of inputs "range of influence" is set to 0.15. Other parameters remain in their previous value because those values are acceptable for training scheme. Once the parameters are set, the outcome FIS generates 104 numbers of MFs for both of the inputs,
1. Grid partition
number of clusters.
and output.
1. Range of influence 2. Squash factor 3. Accept ratio 4. Reject ratio
2. Subtractive clustering
Figure 19 shows the cluster parameters which are:
Fig. 19. Parameters set for subtractive clustering
In order to optimize the obtained parameters, two methods are available:
2. Back propagation: this method consists of steepest descend method for MFs.
The final error training is 3.014e -7 which is shown in figure 18 after 150 epochs.
1. Hybrid method: this method is a combination of least squares and back propagation
The first method is considered for data training. Error tolerance is established to create halt criterion. The error training will stop after certain epoch which is set. The number of epochs
**7.4 ANFIS training**
method.
for both of 8:6 is 150.
Figure 21 shows the ANFIS model structure for 8:6 SRM. There are two inputs (rotor angle and current) and one output (torque). There are total 104 MFs for each of inputs.
Fig. 21. ANFIS model structure for 8:6 SRM
The summarized modeling description is shown in table 3 for 8:6 SRM.
Table 3. Fuzzy rule base table for 8:6 SRM
The mapping surface of 8:6 SRM using neuro-fuzzy technique is shown in figure 19.
Artificial Intelligence Techniques of Estimating of Torque for 8:6 Switched Reluctance Motor 207
ANFIS is one the best approaches due to the capability of learning without dependency to human knowledge. In other worlds, in fuzzy logic approach, membership functions have been formed by the human knowledge but ANFIS because of having training algorithm and independency to human knowledge is more capable to produce accurate data. In this
section, error analysis shows the preciseness of the mode:
∑� ��) = Sum of the computed torque =2.71E+02 ∑� ��) = Sum of the measured torque =270.95 ∑� |�|) = Calculated total absolute error= 4.61 e-011
N = Number of data points= 104
� = Rotor angle in mechanical degree �� = Measured torque in Newton-meter �� = Computed torque in Newton-meter
���� ��∗�� ∗ 100%
<sup>=</sup> 4.61� � 011
FIS ANFIS
2.6058 ∗ 104 ∗ 100 = 1.7011� � 011
Error analysis is conducted for the two approaches. Table 4 reflects the average percentage
As it can been seen clearly, table above shows the ANFIS model is the best among those. ANFIS technique is used in order to develop predictive model for obtaining precision outcome. This approach can be used for any nonlinear function with arbitrary accuracy.
Torque profile of switched reluctance motor is a nonlinear function and the inherent nonlinear characteristics lead us toward artificial intelligence approaches. Due to the mentioned nonlinearity a predictive model is needed. ANFIS model owing to its abilities to predict is opted. The reason being is due to the ANFIS modeling approach possessing
5.7431% 1.701� � 011%
**8.2 Error analysis for torque estimation model using adaptive neuro-fuzzy inference**
**system technique**
Thus,
Mean ��= �.������
**9. Conclusion**
error of each models.
Error
Table 4. Error analysis result
Average% error= � ∑|�|
**From the results in appendix B:**
I = phase current(A)
��� <sup>=</sup>2.6058
Average Percentage
Fig. 22. Surface view of 8:6 SRM
### **8. Result and discussion**
### **8.1 Error analysis for torque estimation model using fuzzy logic technique**
Torque estimation based on fuzzy logic technique has been presented. Thus, 8 and 13 membership functions were formed for the inputs and 21 for the torque as the output for 8:6 RM. Error analysis is conducted to obtain the accuracy of the model. Appendix D shows the computed torque values in term of comparison with the desired measured values.
### **From the results in appendix A:**
### Thus,
Mean $T\_f = \frac{262.4614}{104} = 2.716$
\_Average\% error = $\left[\frac{\Sigma|\varepsilon|}{Mean \, T\_f \, \ast N}\right] \ast 100\%
$$= \frac{16.2222}{2.716 \ast 104} \ast 100\% = 5.7431$$
$
### **8.2 Error analysis for torque estimation model using adaptive neuro-fuzzy inference system technique**
ANFIS is one the best approaches due to the capability of learning without dependency to human knowledge. In other worlds, in fuzzy logic approach, membership functions have been formed by the human knowledge but ANFIS because of having training algorithm and independency to human knowledge is more capable to produce accurate data. In this section, error analysis shows the preciseness of the mode:
### **From the results in appendix B:**
Thus,
206 Fuzzy Logic – Algorithms, Techniques and Implementations
**8.1 Error analysis for torque estimation model using fuzzy logic technique**
computed torque values in term of comparison with the desired measured values.
Torque estimation based on fuzzy logic technique has been presented. Thus, 8 and 13 membership functions were formed for the inputs and 21 for the torque as the output for 8:6 RM. Error analysis is conducted to obtain the accuracy of the model. Appendix D shows the
Fig. 22. Surface view of 8:6 SRM
**8. Result and discussion**
**From the results in appendix A:**
I = phase current(A)
Thus,
Mean �� <sup>=</sup>���.����
���
Average% error = � ∑|�|
N = Number of data points= 104
= 2.716
���� ��∗�� ∗ 100%
<sup>=</sup> 16.2222
2.716 ∗ 104 ∗ 100% = 5.7431
� = Rotor angle in mechanical degree �� = Measured torque in Newton-meter �� = Computed torque in Newton-meter
∑� ��) = Sum of the measured torque ��=270.95 ∑� ��) = Sum of the computed torque ��=282.4614 ∑� |�|) = Calculated total absolute error=16.2222
Mean $T\_f = \frac{2.71 \text{E} + 02}{104} = 2.6058$
Average\% error=
$\left[\frac{\Sigma \|\epsilon\|}{Mean \ T\_f \ast N}\right] \ast 100\%
$$= \frac{4.61e - 011}{2.6058 \ast 104} \ast 100 = 1.7011e - 011$$
$
### **9. Conclusion**
Error analysis is conducted for the two approaches. Table 4 reflects the average percentage error of each models.
Table 4. Error analysis result
As it can been seen clearly, table above shows the ANFIS model is the best among those. ANFIS technique is used in order to develop predictive model for obtaining precision outcome. This approach can be used for any nonlinear function with arbitrary accuracy.
Torque profile of switched reluctance motor is a nonlinear function and the inherent nonlinear characteristics lead us toward artificial intelligence approaches. Due to the mentioned nonlinearity a predictive model is needed. ANFIS model owing to its abilities to predict is opted. The reason being is due to the ANFIS modeling approach possessing
Artificial Intelligence Techniques of Estimating of Torque for 8:6 Switched Reluctance Motor 209
2 0 0 7.99e-014 7.99E-14 4 0 0 6.49e-014 6.49E-14 6 0 0 6.61e-014 6.61E-14 8 0 0 3.72e-14 3.72E-14 10 0 0 2.14e-13 2.14E-13 12 0 0 2.36e-13 2.36E-13 14 0 0 3.11e-13 3.11E-13 16 0 0 3.08e-014 3.08E-14 . . . . . . . . . . 2 30 0 -1.18e-013 1.18E-13 4 30 0 -4.62e-012 4.62E-12 6 30 0 -6.92e-012 6.92E-12 8 30 0 3.25e-012 3.25E-12 10 30 0 1.15e-012 1.15E-12 12 30 0 3.34e-012 3.34E-12 14 30 0 1.36e-012 1.36E-12 16 30 0 2.43e-011 2.43E-11 936 1560 270.95 2.71E+02 4.61E-11
Chancharoensook, P.& Rahman, M.F. , "Dynamic modeling of a four-phase 8/6 switched
Geldhof, K.R. & Van den Bossche, A. & Vyncke, T.J. & Melkebeek, J.A.A. , "Influence of flux
Mirzaeian-Dehkordi, B. & Moallem, P. , "Genetic Algorithm Based Optimal Design of
Gobbi, R. & Ramar, K. , "Practical current control techniques for torque ripple minimization
Rajapakse, A.D. & Gole, A.M.; Muthumuni, D. & Wilson, P.L.; Perregaux, A. , "Simulation of
*Conference of IEEE* , vol., no., pp.1246-1251, 10-13 Nov. 2008
reluctance motor using current and torque look-up tables," *IECON 02 [Industrial Electronics Society, IEEE 2002 28th Annual Conference of the]* , vol.1, no., pp. 491- 496
penetration on inductance and rotor position estimation accuracy of switched reluctance machines," *Industrial Electronics, 2008. IECON 2008. 34th Annual*
Switching Circuit Parameters for a Switched Reluctance Motor Drive," *Power Electronics, Drives and Energy Systems, 2006. PEDES '06. International Conference on* ,
in SR motors," *Power and Energy Conference, 2008. PECon 2008. IEEE 2nd*
switched reluctance motors embedded in large networks," *Power System Technology,*
Torque |ε<sup>|</sup>
Current Rotor Angle Measured Torque Computed
**12. Appendix B**
**13. References**
vol.1, 5-8 Nov. 2002
vol., no., pp.1-6, 12-15 Dec. 2006
*International*, vol., no., pp.743-748, 1-3 Dec. 2008
*Error analysis for the ANFIS model of 8:6 SRM*
learning characteristic capability that allows it to learn from the data values through the training scheme, thus avoids on the dependency of human knowledge with regard to the systems(Parvizi.A&Hassani&Mehbodnia&Makhilef&Tamjis,2009) . Besides, ANFIS method dose not have the complexity of FIS method which makes it much easier to understand and utilize. Average percentage error shows that the outcome is in good agreement with the published data. Torque profile is simulated and results reveals that ANFIS modeling method is a trustable model for further research. In addition, this approch can be used in order to control the turn-off angle of the SRM which leades to a SRM with low torque ripples.
### **10. Acknowledgment**
This work is dedicated to my parents, Mohammad and Fatemeh for their kindness and support.The author like to thank Dr.Aris Ramlan, Mr.Peter Nicoll and Dr.M. Beikzadeh for reviewing and his right-on-target comments.
### **11. Appendix A**
Error analysis for torque Using Fuzzy logic Technique
### **12. Appendix B**
208 Fuzzy Logic – Algorithms, Techniques and Implementations
learning characteristic capability that allows it to learn from the data values through the training scheme, thus avoids on the dependency of human knowledge with regard to the systems(Parvizi.A&Hassani&Mehbodnia&Makhilef&Tamjis,2009) . Besides, ANFIS method dose not have the complexity of FIS method which makes it much easier to understand and utilize. Average percentage error shows that the outcome is in good agreement with the published data. Torque profile is simulated and results reveals that ANFIS modeling method is a trustable model for further research. In addition, this approch can be used in order to control the turn-off angle of the SRM which leades to a SRM with low torque
This work is dedicated to my parents, Mohammad and Fatemeh for their kindness and support.The author like to thank Dr.Aris Ramlan, Mr.Peter Nicoll and Dr.M. Beikzadeh for
Current Rotor Angle Measured torque Computed torque |ε| 2 0 0 0.1723 0.1723 4 0 0 0.1723 0.1723 6 0 0 0.1723 0.1723 8 0 0 0.1723 0.1723 10 0 0 0.1723 0.1723 12 0 0 0.1723 0.1723 14 0 0 0.1723 0.1723 16 0 0 0.1723 0.1723 . . . . . . . . . . 2 30 0 0.1723 0.1723 4 30 0 0.1723 0.1723 6 30 0 0.1723 0.1723 8 30 0 0.1723 0.1723 10 30 0 0.1723 0.1723 12 30 0 0.1723 0.1723 14 30 0 0.1723 0.1723 16 30 0 0.1723 0.1723 936 1560 270.95 282.4614 16.2222
ripples.
**10. Acknowledgment**
**11. Appendix A**
reviewing and his right-on-target comments.
Error analysis for torque Using Fuzzy logic Technique
*Error analysis for the ANFIS model of 8:6 SRM*
### **13. References**
**11**
*1Romania 2Switzerland*
**Engine Knock Detection Based on**
**Computational Intelligence Methods**
Adriana Florescu1, Claudiu Oros1 and Anamaria Radoi2
*Artificial intelligence* emerged from human thinking that has both logical and intuitive or subjective sides. The logical side has been developed and utilized, resulting advanced von Neumann type computers and expert systems, both constituting the *hard computing* domain. However, it is found that hard computing can't give the solution of very complicated problems by itself. In order to cope with this difficulty, the intuitive and subjective thinking of human mind was explored, resulting the *soft computing* domain (also called *computational intelligence*). It includes *neural networks*, *fuzzy logic* and *probabilistic reasoning*, the last gathering *evolutionary computation* (including *genetic algorithms* with related efforts in *genetic programming* and *classifier systems*, *evolution strategies* and *evolutionary programming*), *immune networks*, *chaos computing* and parts of *learning theory*. In different kind of applications, all pure artificial intelligence methods mentioned above proved to be rather complementary than competitive, so that combined methods appeared in order to gather the advantages and to cope with the disadvantages of each pure method. The scope of this chapter is to study and finaly compare some representative classes of pure and combined computational intelligence methods applied
The internal-combustion engine is one of the most used vehicle power generators in the world today. When looking at the characteristics of a vehicle - and therefore the ones of the engine that drives it - , some of the most important are the emissions, fuel economy and efficiency. All three of these variables are affected by a phenomenon that occurs in the engine called knock**.** *Engine knock* (also known as *knocking*, *self-combustion*, *detonation*, *spark knock* or *pinging*) in spark-ignition internal combustion engines occurs when combustion of the mixture of fuel and air in the cylinder starts off correctly because of the ignition by the spark plug, but one or more pockets of the mixture explode outside the normal combustion front. The importance of knock detection comes from the effects it generates; these can range from increased fuel consumption and pollution, the decrease of engine power and up to partial or complete destruction of the cylinders, pistons, rods, bearings and many other
**1. Introduction**
in engine knock detection.
damages around the engine bay.
*1University Politehnica of Bucharest, 2Ecole Politechnique Federale de Laussane,*
*2004. PowerCon 2004. 2004 International Conference on* , vol.1, no., pp. 695- 700 Vol.1, 21-24 Nov. 2004
## **Engine Knock Detection Based on Computational Intelligence Methods**
Adriana Florescu1, Claudiu Oros1 and Anamaria Radoi2
*1University Politehnica of Bucharest, 2Ecole Politechnique Federale de Laussane, 1Romania 2Switzerland*
### **1. Introduction**
210 Fuzzy Logic – Algorithms, Techniques and Implementations
Wai-Chuen Gan& Cheung, N.C. & Li Qiu , "Short distance position control for linear
Bhiwapurkar, N.; Jain, A.K.; Mohan, N.; , "Study of new stator pole geometry for
Parvizi.A&Hassani.M&Mehbodnia.A&Makhilef.S&Tamjis.M.R *"Adaptive Neuro-Fuzzy*
*the 2001 IEEE* , vol.4, no., pp.2329-2336 vol.4, 30 Sep-4 Oct. 2001
*International Conference on* , vol., no., pp.516-520, 15-15 May 2005
conference proceeding , page(s): 1-4,2009
21-24 Nov. 2004
*2004. PowerCon 2004. 2004 International Conference on* , vol.1, no., pp. 695- 700 Vol.1,
switched reluctance motors: a plug-in robust compensator approach," *Industry Applications Conference, 2001. Thirty-Sixth IAS Annual Meeting. Conference Record of*
improvement of SRM torque profile," *Electric Machines and Drives, 2005 IEEE*
*Approach of Estimating of torque for 8:6 Switched Reluctance Motor"* International Conference for Technical Postgraduates TECHPOS conference, Malaysia. IEEE
> *Artificial intelligence* emerged from human thinking that has both logical and intuitive or subjective sides. The logical side has been developed and utilized, resulting advanced von Neumann type computers and expert systems, both constituting the *hard computing* domain. However, it is found that hard computing can't give the solution of very complicated problems by itself. In order to cope with this difficulty, the intuitive and subjective thinking of human mind was explored, resulting the *soft computing* domain (also called *computational intelligence*). It includes *neural networks*, *fuzzy logic* and *probabilistic reasoning*, the last gathering *evolutionary computation* (including *genetic algorithms* with related efforts in *genetic programming* and *classifier systems*, *evolution strategies* and *evolutionary programming*), *immune networks*, *chaos computing* and parts of *learning theory*. In different kind of applications, all pure artificial intelligence methods mentioned above proved to be rather complementary than competitive, so that combined methods appeared in order to gather the advantages and to cope with the disadvantages of each pure method. The scope of this chapter is to study and finaly compare some representative classes of pure and combined computational intelligence methods applied in engine knock detection.
> The internal-combustion engine is one of the most used vehicle power generators in the world today. When looking at the characteristics of a vehicle - and therefore the ones of the engine that drives it - , some of the most important are the emissions, fuel economy and efficiency. All three of these variables are affected by a phenomenon that occurs in the engine called knock**.** *Engine knock* (also known as *knocking*, *self-combustion*, *detonation*, *spark knock* or *pinging*) in spark-ignition internal combustion engines occurs when combustion of the mixture of fuel and air in the cylinder starts off correctly because of the ignition by the spark plug, but one or more pockets of the mixture explode outside the normal combustion front. The importance of knock detection comes from the effects it generates; these can range from increased fuel consumption and pollution, the decrease of engine power and up to partial or complete destruction of the cylinders, pistons, rods, bearings and many other damages around the engine bay.
Engine Knock Detection Based on Computational Intelligence Methods 213
The first layer represents the input and is built with fuzzy input neurons, each one selecting a characteristic of the original sample vector. In the case of a two dimensional sample containing N1xN2 vector elements we will have a first layer that has N1xN2 neurons. For
[1] [1]
[1] [1]
all the input elements. The notation will be kept for neurons belonging to all the following
The second layer is built of N1xN2 neurons and its purpose is to perform the fuzzification of the input patterns by means of the weight function *wmn* ( ,) - also called the fuzzification
> <sup>222</sup> ( ) ( ,) *m n wmn e*
where parameters m=-(N1-1), …, +(N1-1), n=-(N2-1), …, +(N2-1) and β determines how much of the sample vector each fuzzy neuron sees. Each neuron from the second layer has M outputs, one for each neuron in the third layer. The output for the second layer neuron on
*ij ij ij szx* , (1)
*ij ij <sup>v</sup>*max *y sP* , (2)
*ij z* is it's input value, *ij x* is the value of the element (i, j) in the
*ij y* is its output value and *Pv*max is the maximum value of
*ij s* represents the state of the neuron on the (i, j)
, (3)
Fig. 1. The Fuzzy Kwan-Cai Neural Network structure
the neuron on the (i, j) position the equations are:
for i=1, 2, …, N1; j=1, 2, …, N2, where [1]
position for the first layer, [1]
function - , defined as:
position (p, q) is:
layers.
input sample pattern, ( 0 *ij x* ), [1]
Internal combustion engines present an optimum working cycle that is right on the edge of self-combustion or knock. If engine knock occurs and is detected in a cycle then the ignition timing (spark angle) needs to be modified so that the next cycle does not suffer from the same phenomenon. This is why the detection needs to be done in under a cycle (Bourbai, 2000; Li&Karim, 2004; Hamilton&Cowart, 2008; Erjavec, 2009).
Engine knock can be detected using a series of devices placed in and around the engine bay like: pressure sensors mounted inside each cylinder, devices that measure the ionization current in the spark plug or accelerometers mounted on the engine to measure vibrations etc. The best and most accurate information on knock is given by the *pressure sensors* but the easiest and less expensive way to detect it is by using *vibration sensors* mounted on the engine (Erjavec, 2009; Gupta, 2006; Bosch, 2004; Thomas et al., 1997; Ettefag, 2008, Fleming, 2001). The knock detection methods used so far for extracting information from the engine sensors include *time*, *frequency (spectrum)* or a diversity *time-frequency analysis (Wavelet)* based solutions (Adeli&Karim, 2005; Park&Jang, 2004; Radoi et al., 2009; Midori et al., 1999; Lazarescu et al., 2004; Jonathan et al., 2006). The restriction of average detection rates and the complexity of information needed for the Wavelet analysis support further developments and hybridization with mixed techniques that proved useful in other fields of application than the one explored in this chapter: *wavelet-fuzzy* (Borg et al., 2005), *waveletneural* (Zhang&Benveniste, 1992; Billings&Wei, 2005; Wu&Liu, 2009; Banakar&Azeem, 2008) and *wavelet-neuro-fuzzy* (Ylmaz&Oysal, 2010).
Among the pure computational intelligence methods described in (Wang&Liu, 2006; Prokhorov, 2008; Mitchell, 2010;Wehenkel, 1997), different types of neural network applications have been employed with better detection rates than the previous non-neural methods but no clear comparative analysis results have been presented so far for engine knock detection. The methods taken into account and finally compared in this chapter start with the *Fuzzy Kwan-Cai Neural Network* (Kwan&Cai, 1994) - for the application of which other neuro-fuzzy or fuzzy logic models were studied (Zhang&Liu, 2006; Ibrahim,2004; Liu&Li, 2004; Hui, 2011; Chen, 2005) -, expand to the *Kohonen Self-Organizing Map (SOM)* (Kohonen, 2000, 2002; Hsu, 2006; Lopez-Rubio, 2010) and end with *Bayes Classifier* (Larose, 2006) to which results of this chapter conforming with other work (Auld et al., 2007) published so far have proved needing hybridization.
Work started using two sizes of training and testing sample groups, both belonging to the Bosch Group database in order to see how data size can affect the results. The applications were built to handle both pressure and vibration samples in order to see which of them can supply the most valuable information. In addition, due to the lack of chapters available on this subject, through the analysis of the results, we can get a better impression of the nature of these types of signals, the coherence of samples and evolution of detection rates with every new sample added. Also, to complete the analysis, a comparison of the responses from pressure and vibration families of samples is made for the three methods.
### **2. Mathematical background of used computational intelligence methods**
### **2.1 Fuzzy Kwan-Cai neural network**
The Fuzzy Kwan-Cai neural network shown in Fig.1 has four layers, each of them being a fuzzy block represented by a different type of fuzzy neurons with their own specific purpose and functions (Kwan&Cai, 1994).
212 Fuzzy Logic – Algorithms, Techniques and Implementations
Internal combustion engines present an optimum working cycle that is right on the edge of self-combustion or knock. If engine knock occurs and is detected in a cycle then the ignition timing (spark angle) needs to be modified so that the next cycle does not suffer from the same phenomenon. This is why the detection needs to be done in under a cycle (Bourbai,
Engine knock can be detected using a series of devices placed in and around the engine bay like: pressure sensors mounted inside each cylinder, devices that measure the ionization current in the spark plug or accelerometers mounted on the engine to measure vibrations etc. The best and most accurate information on knock is given by the *pressure sensors* but the easiest and less expensive way to detect it is by using *vibration sensors* mounted on the engine (Erjavec, 2009; Gupta, 2006; Bosch, 2004; Thomas et al., 1997; Ettefag, 2008, Fleming, 2001). The knock detection methods used so far for extracting information from the engine sensors include *time*, *frequency (spectrum)* or a diversity *time-frequency analysis (Wavelet)* based solutions (Adeli&Karim, 2005; Park&Jang, 2004; Radoi et al., 2009; Midori et al., 1999; Lazarescu et al., 2004; Jonathan et al., 2006). The restriction of average detection rates and the complexity of information needed for the Wavelet analysis support further developments and hybridization with mixed techniques that proved useful in other fields of application than the one explored in this chapter: *wavelet-fuzzy* (Borg et al., 2005), *waveletneural* (Zhang&Benveniste, 1992; Billings&Wei, 2005; Wu&Liu, 2009; Banakar&Azeem, 2008)
Among the pure computational intelligence methods described in (Wang&Liu, 2006; Prokhorov, 2008; Mitchell, 2010;Wehenkel, 1997), different types of neural network applications have been employed with better detection rates than the previous non-neural methods but no clear comparative analysis results have been presented so far for engine knock detection. The methods taken into account and finally compared in this chapter start with the *Fuzzy Kwan-Cai Neural Network* (Kwan&Cai, 1994) - for the application of which other neuro-fuzzy or fuzzy logic models were studied (Zhang&Liu, 2006; Ibrahim,2004; Liu&Li, 2004; Hui, 2011; Chen, 2005) -, expand to the *Kohonen Self-Organizing Map (SOM)* (Kohonen, 2000, 2002; Hsu, 2006; Lopez-Rubio, 2010) and end with *Bayes Classifier* (Larose, 2006) to which results of this chapter conforming with other work (Auld et al., 2007)
Work started using two sizes of training and testing sample groups, both belonging to the Bosch Group database in order to see how data size can affect the results. The applications were built to handle both pressure and vibration samples in order to see which of them can supply the most valuable information. In addition, due to the lack of chapters available on this subject, through the analysis of the results, we can get a better impression of the nature of these types of signals, the coherence of samples and evolution of detection rates with every new sample added. Also, to complete the analysis, a comparison of the responses
from pressure and vibration families of samples is made for the three methods.
**2. Mathematical background of used computational intelligence methods**
The Fuzzy Kwan-Cai neural network shown in Fig.1 has four layers, each of them being a fuzzy block represented by a different type of fuzzy neurons with their own specific
2000; Li&Karim, 2004; Hamilton&Cowart, 2008; Erjavec, 2009).
and *wavelet-neuro-fuzzy* (Ylmaz&Oysal, 2010).
published so far have proved needing hybridization.
**2.1 Fuzzy Kwan-Cai neural network**
purpose and functions (Kwan&Cai, 1994).
Fig. 1. The Fuzzy Kwan-Cai Neural Network structure
The first layer represents the input and is built with fuzzy input neurons, each one selecting a characteristic of the original sample vector. In the case of a two dimensional sample containing N1xN2 vector elements we will have a first layer that has N1xN2 neurons. For the neuron on the (i, j) position the equations are:
$$\mathbf{x}\_{i\uparrow}^{\{1\}} = \mathbf{z}\_{i\downarrow}^{\{1\}} = \mathbf{x}\_{i\downarrow} \,\, \, \, \, \tag{1}$$
$$y\_{ij}^{[1]} = \mathbf{s}\_{ij}^{[1]} \Big/ P\_{v \cdot \max \prime} \,\tag{2}$$
for i=1, 2, …, N1; j=1, 2, …, N2, where [1] *ij s* represents the state of the neuron on the (i, j) position for the first layer, [1] *ij z* is it's input value, *ij x* is the value of the element (i, j) in the input sample pattern, ( 0 *ij x* ), [1] *ij y* is its output value and *Pv*max is the maximum value of all the input elements. The notation will be kept for neurons belonging to all the following layers.
The second layer is built of N1xN2 neurons and its purpose is to perform the fuzzification of the input patterns by means of the weight function *wmn* ( ,) - also called the fuzzification function - , defined as:
$$w(m,n) = e^{-\beta^2 \left(m^2 + n^2\right)},\tag{3}$$
where parameters m=-(N1-1), …, +(N1-1), n=-(N2-1), …, +(N2-1) and β determines how much of the sample vector each fuzzy neuron sees. Each neuron from the second layer has M outputs, one for each neuron in the third layer. The output for the second layer neuron on position (p, q) is:
Engine Knock Detection Based on Computational Intelligence Methods 215
Fig. 2. Flowchart for implemented Kwan-Cai algorithm
The Kohonen Self-Organizing Map (SOM) with the structure presented in Fig.3 is a neural network characterized by the fact that neighboring neurons (cells) communicate among themselves by mutual-lateral interactions transforming into detectors of specific classes when given input patterns. The learning can be unsupervised or supervised (Kohonen, 2000, 2002; Hsu, 2006; Lopez-Rubio, 2010) In this chapter the supervised learning algorithm was
**2.2 Kohonen Self-Organizing Map (SOM)**
used.
$$y\_{pqm}^{[2]} = q\_{pqm}^{[2]} \, , \tag{4}$$
for p=1, …, N1; q=1, …, N2; m=1,…,M, where [2] *ypqm* is the *th m* output of the second layer neuron on position (p,q) to the *th m* third level neuron. The output function *qpqm* is determined within the training algorithm. For a more simplified approach, we can choose isosceles triangles with the base α and the height 1, mathematically defined as:
$$\left| y\_{pqm}^{\left[2\right]} = q\_{pqm} \left( s\_{pq}^{\left[2\right]} \right) = \left| 1 - \frac{2\left| s\_{pq}^{\left[2\right]} - \theta\_{pqm} \right| \le \frac{\alpha}{2}}{\alpha}, \text{ for } \left| s\_{pq}^{\left[2\right]} - \theta\_{pqm} \right| \le \frac{\alpha}{2} \right. \tag{5}$$
where 0 , p=1, …, N1; q=1, …, N2; m=1, …, M. Parameter θpqm is the center of the isosceles triangle base. By means of the training algorithm, p, q and m values corresponding to α and θpqm are determined.
The third layer is made-up of M neurons each of them representing a learned pattern and so the value for M can only be determined at the end of the learning process. It can be seen as a fuzzy deduction (inference) layer. The output for the third layer neuron is:
$$y\_m^{[3]} = s\_m^{[3]} = \min\_{p=1\ldots N1} (\min\_{q=1\ldots N2} (y\_{p\eta m}^{[2]}) \, . \tag{6}$$
for m=1,…, M.
The fourth and final layer is the network's output layer and is made up of competitive neurons one for each pattern that is learned; it is the defuzzification layer. If an input pattern is more similar to the mth pattern that was learned, then the output of the mth comparative neuron will be attributed value 1 and the others value 0:
$$\mathbf{y}\_{m}^{\{4\}} = \mathbf{s}\_{m}^{\{4\}} = \mathbf{z}\_{m}^{\{4\}} \,\tag{7}$$
$$y\_{pqm}^{[4]} = q[s\_m^{[4]} - T] = \begin{cases} 0, \text{ if } s\_m^{[4]} < T \\ \mathbf{1}, \text{ if } s\_m^{[4]} = T \end{cases} \tag{8}$$
$$T = \max\_{m=1\ldots M} \text{(max}\_{j=1\ldots N/2} \{ y\_m^{[\} \} )\tag{9}$$
for m=1,…, M, where T is defined as the activation threshold for all the neurons in the forth layer.
The flowchart in Fig.2 summarizes the procedure of adapting and implementing the Fuzzy Kwan-Cai algorithm to the application proposed in the chapter. The differences from the standard theoretical algorithm are that the sample databases are first imported and validated for integrity and then separated into pressure and vibration, respectively training and testing classes. The standard classification steps follow and the algorithm ends with the calculation of the detection rate.
214 Fuzzy Logic – Algorithms, Techniques and Implementations
for p=1, …, N1; q=1, …, N2; m=1,…,M, where [2] *ypqm* is the *th m* output of the second layer neuron on position (p,q) to the *th m* third level neuron. The output function *qpqm* is determined within the training algorithm. For a more simplified approach, we can choose
<sup>2</sup> <sup>1</sup> , 2
[3] [3] [2] *y s m m* min (min ( ) *<sup>p</sup>*1... 1 1... 2 *<sup>N</sup> <sup>q</sup> <sup>N</sup> ypqm* , (6)
[4]
*if T*
[4] m
[3] *T y* max (max ( ) *mM jNm* 1... 1... 2 , (9)
[4] [4] [4] *ymmm s z* , (7)
, (8)
isosceles triangles with the base α and the height 1, mathematically defined as:
*y qs for s*
*other*
0,
fuzzy deduction (inference) layer. The output for the third layer neuron is:
[2] [2] [2] [2] 2
*pq pqm pqm pqm pq pq pqm*
*<sup>q</sup> <sup>s</sup>*
The third layer is made-up of M neurons each of them representing a learned pattern and so the value for M can only be determined at the end of the learning process. It can be seen as a
The fourth and final layer is the network's output layer and is made up of competitive neurons one for each pattern that is learned; it is the defuzzification layer. If an input pattern is more similar to the mth pattern that was learned, then the output of the mth comparative
[4] [4] m
0, s [ ] 1, s *pqm m*
*y qs T if <sup>T</sup>*
for m=1,…, M, where T is defined as the activation threshold for all the neurons in the forth
The flowchart in Fig.2 summarizes the procedure of adapting and implementing the Fuzzy Kwan-Cai algorithm to the application proposed in the chapter. The differences from the standard theoretical algorithm are that the sample databases are first imported and validated for integrity and then separated into pressure and vibration, respectively training and testing classes. The standard classification steps follow and the algorithm ends with the
, p=1, …, N1; q=1, …, N2; m=1, …, M. Parameter θpqm is the center of the isosceles triangle base. By means of the training algorithm, p, q and m values corresponding
neuron will be attributed value 1 and the others value 0:
where 0
for m=1,…, M.
layer.
calculation of the detection rate.
to α and θpqm are determined.
[2] [2] *y q pqm pqm* , (4)
, (5)
Fig. 2. Flowchart for implemented Kwan-Cai algorithm
### **2.2 Kohonen Self-Organizing Map (SOM)**
The Kohonen Self-Organizing Map (SOM) with the structure presented in Fig.3 is a neural network characterized by the fact that neighboring neurons (cells) communicate among themselves by mutual-lateral interactions transforming into detectors of specific classes when given input patterns. The learning can be unsupervised or supervised (Kohonen, 2000, 2002; Hsu, 2006; Lopez-Rubio, 2010) In this chapter the supervised learning algorithm was used.
Engine Knock Detection Based on Computational Intelligence Methods 217
After the winner determination process has finished the weights refining one is started and this must not have an effect on all the neurons but only in a certain vicinity \* *Vj* around the winner j\*. Outside this perimeter the influence of this process is considered null. The radius of this vicinity starts out big and keeps on getting smaller and smaller with the refining
The learning rate can have many expressions. In this application, the chosen expression was:
<sup>0</sup> ( ) exp[ / ] *k j*
where rj\* and rk are position vectors in the network representing the characteristic of the neural center of the vicinity and the neuron with the index k for which the refining process is taking place. Function η0=η0(t) decrease in time, representing the value of the learning rate
0() / *<sup>p</sup>*
The parameter σ controls the speed of decreasing the learning rate, depending on the radius
After the refining process for the current input vector is finished the next one is selected and so on until all the input vectors are used and the stop training condition is inspected. A useful stopping condition is the moment when the weights of the network cease being
( 1) ( ) , *wt wt ij ij*
The flowchart in Fig.4 summarizes the procedure of adapting and implementing the Kohonen Self-Organizing Map algorithm to the application proposed in the chapter. The differences from the standard theoretical algorithm are the same as those described for the
For the Bayes Classifier working with Gaussian classes (Larose, 2006) considering first the case of two (R=2) 1-dimensional classes (n=1), the density of probability being of Gaussian
<sup>1</sup> <sup>2</sup> () (| ) ( ) ( ), <sup>2</sup>
*r*
*r rr r*
*g x px P e P*
2 2 ( )
>
*r r x m*
(16)
*t rr*
\* 2
, (13)
, (15)
*t at* , (14)
process.
in the center of the vicinity:
refined (are no longer being modified):
where i=0, 1, …, n-1 and j=0, 1, …, M-1.
Fuzzy Kwan-Cai algorithm in Fig. 2.
**2.3 Bayes classifier**
nature can be defined:
where parameter *r* 1; 2 .
of the vicinity.
The network transforms similarities among vectors into neural vicinities (the similar input patterns will be found as neighbors).
Fig. 3. The SOM neural network
From a structural point of view, the Kohonen neural network is composed of two layers out of which the first one is an input layer made of transparent neurons with no processing functions. Its purpose is to receive the input pattern and send it to the second layer. This first layer has the same size as the input pattern.
The second layer contains M output neurons, a number equal or higher than the number of classes desired in order to classify the entry patterns. They can be arranged planar, linear, circular, as a torus or sphere, the training and performances being dependent on the network shape. The planar network can also be rectangular or hexagonal depending on the placement of neurons.
An input vector *<sup>n</sup> X R <sup>p</sup>* is applied in parallel to all the neurons of the network, each of them being characterized by a weight vector:
$$\mathcal{W}\_{j} \{ w\_{0j'} w\_{1j'} \dots w\_{n-11j} \}^T \in \mathbb{R}^n \, \, \, \, \tag{10}$$
for j=0, 1, …, M-1.
In order to choose the winning neuron j\* with its associated weight vector Wj\* for an input pattern we must calculate the Gaussian distance dj between that pattern and each of the neuron's weight vectors. The winner will be chosen by the lowest distance \* *<sup>j</sup> d* of all:
$$\|d\_j = \left\| X\_p - \mathcal{W}\_j \right\|\_{\nu} \tag{11}$$
$$d\_j^\* = \min\{d\_j\} \tag{12}$$
for j=0, 1, …, M-1.
216 Fuzzy Logic – Algorithms, Techniques and Implementations
The network transforms similarities among vectors into neural vicinities (the similar input
From a structural point of view, the Kohonen neural network is composed of two layers out of which the first one is an input layer made of transparent neurons with no processing functions. Its purpose is to receive the input pattern and send it to the second layer. This
The second layer contains M output neurons, a number equal or higher than the number of classes desired in order to classify the entry patterns. They can be arranged planar, linear, circular, as a torus or sphere, the training and performances being dependent on the network shape. The planar network can also be rectangular or hexagonal depending on the
An input vector *<sup>n</sup> X R <sup>p</sup>* is applied in parallel to all the neurons of the network, each of
In order to choose the winning neuron j\* with its associated weight vector Wj\* for an input pattern we must calculate the Gaussian distance dj between that pattern and each of the
neuron's weight vectors. The winner will be chosen by the lowest distance \*
0 1 11 ( , ,..., )*T n Www w R j j j nj* , (10)
*j pj d XW* , (11)
\* min{ } *j j d d* , (12)
*<sup>j</sup> d* of all:
patterns will be found as neighbors).
Fig. 3. The SOM neural network
placement of neurons.
for j=0, 1, …, M-1.
for j=0, 1, …, M-1.
first layer has the same size as the input pattern.
them being characterized by a weight vector:
After the winner determination process has finished the weights refining one is started and this must not have an effect on all the neurons but only in a certain vicinity \* *Vj* around the winner j\*. Outside this perimeter the influence of this process is considered null. The radius of this vicinity starts out big and keeps on getting smaller and smaller with the refining process.
The learning rate can have many expressions. In this application, the chosen expression was:
$$\eta(t) = \eta\_0 \exp\left[-\left\|r\_k - r\_j^\*\right\|/\sigma^2\right],\tag{13}$$
where rj\* and rk are position vectors in the network representing the characteristic of the neural center of the vicinity and the neuron with the index k for which the refining process is taking place. Function η0=η0(t) decrease in time, representing the value of the learning rate in the center of the vicinity:
$$
\eta\_0(t) = a \nmid t^p \,, \tag{14}
$$
The parameter σ controls the speed of decreasing the learning rate, depending on the radius of the vicinity.
After the refining process for the current input vector is finished the next one is selected and so on until all the input vectors are used and the stop training condition is inspected. A useful stopping condition is the moment when the weights of the network cease being refined (are no longer being modified):
$$\left\| w\_{ij}(t+1) - w\_{ij}(t) \right\| < \varepsilon\_{\prime \prime} \tag{15}$$
where i=0, 1, …, n-1 and j=0, 1, …, M-1.
The flowchart in Fig.4 summarizes the procedure of adapting and implementing the Kohonen Self-Organizing Map algorithm to the application proposed in the chapter. The differences from the standard theoretical algorithm are the same as those described for the Fuzzy Kwan-Cai algorithm in Fig. 2.
### **2.3 Bayes classifier**
For the Bayes Classifier working with Gaussian classes (Larose, 2006) considering first the case of two (R=2) 1-dimensional classes (n=1), the density of probability being of Gaussian nature can be defined:
$$\log\_r(\mathbf{x}) = p(\mathbf{x} \mid o\_r) \cdot P(o\_r) = \frac{1}{\sqrt{2\pi} \cdot \sigma\_r} \cdot e^{-\frac{\left(\mathbf{x} - \mathbf{w}\_r\right)^2}{2\sigma\_r^2}} \cdot P(o\_r), \tag{16}$$
where parameter *r* 1; 2 .
Engine Knock Detection Based on Computational Intelligence Methods 219
/2 1/2
*C*
*r*
where { } *m Ex r r* represents the means of vectors in class r, {( )( ) } *<sup>T</sup> C Exm xm rr r r* is the matrix of covariance for the vectors in class r and { } *Er* is an operator that determines the mean value and that is used to make estimations concerning mr and Cr based on a finite
1
*r*
*r r r r x C xx m m N*
Cr being a positive semi-defined symmetrical matrix. The discriminant function based on
1 1 <sup>1</sup> ( ) ln ( ) ln [( ) ( )] 2 2
The flowchart in Fig.5 summarizes the procedure of adapting and implementing the Bayes Classifier algorithm to the application proposed in the chapter. The differences from the standard theoretical algorithm are the same as those described for the Fuzzy Kwan-Cai
The algorithms treated in this chapter were tested on a Bosch Group database using two sizes of vector sample groups: one of 100 vectors and one of 1000, both of them containing pressure and vibration samples. In each case two thirds of the group was used for training
The vectors that make up the database represent samples taken from petrol engines, some corresponding to knock situations and some not. The signals related to these samples were taken from pressure and vibration sensors mounted in and around the engine bay. The change in pressure caused by knock activity is seen as an immediate rise in pressure due to causes outside the normal engine piston cycle. On the other hand, the vibration sensors will detect vibrations – knocking noises – representing abnormal combustion fields being
The applications have to declare knock or no knock for every sample vector received and, after testing the database, reach a verdict on the error of the process or in this case the identification rate. When knock is encountered actions can be taken to return the engine to a
*r r <sup>j</sup> rr r gx P C x m C x m*
*r x m x N*
*r*
1
algorithm in Fig.2 and the Kohonen Self-Organizing Map in Fig.4.
*r*
*T T*
*T*
, (20)
<sup>1</sup> (| ) (2 )
number Nr of patterns from ωr. Their formulae are:
the Gaussian density of probability will be:
**3. Experimental results for each method**
and one third for testing.
generated inside the pistons.
non-knock state.
**3.1 Methodology description, results and analysis**
*r n*
*p x e*
2
<sup>1</sup> <sup>1</sup> ( )( )
, (17)
, (18)
, (19)
*T <sup>r</sup> <sup>j</sup> <sup>r</sup> xm C xm*
Fig. 4. Flowchart for implemented Kohonen Self- Organizing Map algorithm
Making an expansion to the n-dimensional case, the formula (16) for the Gaussian dispersion becomes:
218 Fuzzy Logic – Algorithms, Techniques and Implementations
Fig. 4. Flowchart for implemented Kohonen Self- Organizing Map algorithm
dispersion becomes:
Making an expansion to the n-dimensional case, the formula (16) for the Gaussian
$$p(\mathbf{x} \mid \boldsymbol{\alpha}\_r) = \frac{1}{(2\pi)^{n/2} \left| \mathbb{C}\_r \right|^{1/2}} e^{-\frac{1}{2} (\mathbf{x} - \boldsymbol{m}\_r)^T \mathbb{C}\_j^{-1} (\mathbf{x} - \boldsymbol{m}\_r)}\tag{17}$$
where { } *m Ex r r* represents the means of vectors in class r, {( )( ) } *<sup>T</sup> C Exm xm rr r r* is the matrix of covariance for the vectors in class r and { } *Er* is an operator that determines the mean value and that is used to make estimations concerning mr and Cr based on a finite number Nr of patterns from ωr. Their formulae are:
$$\text{cm}\_r = \frac{1}{N\_r} \sum\_{\text{x} \neq \text{co}\_r} \text{x} \tag{18}$$
$$\mathbf{C}\_r = \frac{1}{N\_r} \sum\_{\mathbf{x} \in ao\_r} \mathbf{x} \mathbf{x}^T - m\_r m\_r^T \tag{19}$$
Cr being a positive semi-defined symmetrical matrix. The discriminant function based on the Gaussian density of probability will be:
$$\log\_r(\mathbf{x}) = \ln P(o\_r) - \frac{1}{2} \ln \left| \mathbf{C}\_j \right| - \frac{1}{2} [\left(\mathbf{x} - m\_r\right)^T \mathbf{C}\_r^{-1} (\mathbf{x} - m\_r)],\tag{20}$$
The flowchart in Fig.5 summarizes the procedure of adapting and implementing the Bayes Classifier algorithm to the application proposed in the chapter. The differences from the standard theoretical algorithm are the same as those described for the Fuzzy Kwan-Cai algorithm in Fig.2 and the Kohonen Self-Organizing Map in Fig.4.
### **3. Experimental results for each method**
### **3.1 Methodology description, results and analysis**
The algorithms treated in this chapter were tested on a Bosch Group database using two sizes of vector sample groups: one of 100 vectors and one of 1000, both of them containing pressure and vibration samples. In each case two thirds of the group was used for training and one third for testing.
The vectors that make up the database represent samples taken from petrol engines, some corresponding to knock situations and some not. The signals related to these samples were taken from pressure and vibration sensors mounted in and around the engine bay. The change in pressure caused by knock activity is seen as an immediate rise in pressure due to causes outside the normal engine piston cycle. On the other hand, the vibration sensors will detect vibrations – knocking noises – representing abnormal combustion fields being generated inside the pistons.
The applications have to declare knock or no knock for every sample vector received and, after testing the database, reach a verdict on the error of the process or in this case the identification rate. When knock is encountered actions can be taken to return the engine to a non-knock state.
Engine Knock Detection Based on Computational Intelligence Methods 221
The testing method for both algorithms (Fuzzy Kwan-Cai and Kohonen Self-Organizing Map) is the following: one parameter varies between its theoretical limits whereas the others remain constant. It is obvious that the difference between the bigger training group and the
The following tables contain only the significant part of the experimental results in order to
This type of neural network does not need training cycles because it learns as it studies the vectors it receives and builds its own classes in the testing process. In order not to get the wrong idea from the start we have to mention that the high number of classes observed in Table I and Table II for this neural network is due to the second nature of the application which acts like a "focusing lens", examining the internal structure of the two main classes. Therefore it must be stated that the number of classes we are interested in, for this experiment, is two. The significance and proper function limits of this application for parameters given in Table I and Table II are: α which is the base of isosceles triangles (α [1.5; 3.5]), β that determines how much of the sample vector each fuzzy neuron sees (β [0.1; 1.6]) and Tf that
The first vector generates a class of its own, the next ones either are found relatives of one of the vectors that have come before and therefore are put in the same class or start a new class. The maximum detection results in the tables mentioned above are outlined by being bolded.
Tables Ia and Ib present the pressure sample detection rate results for the Fuzzy Kwan-Cai neural network using the small sample database and the large sample database. According to Table Ia, the highest detection rate value of 68% was obtained for combination (3.5; 0.15; 1)
The same method has been used for Table Ib showing the combinations used by changing the parameter Tf while keeping constant the other two. Combinations are from (3.5; 0.35; 1) down to (3.5; 0.15; 1). A maximum correct detection rate of 93.40% was obtained for the (3.5;
The detection rate results in Tables Ia and Ib show that from this point of view the Fuzzy Kwan-Cai neural network is very stabile, small variations of its parameters not affecting the experimental outcome. It is clear from the results presented that an increase in the sample database leads to an increase in the detection rates, the network not being affected by
Tables IIa and IIb contain the vibration sample detection results. Table IIa represents the small sample database and Table IIb the large one. Table IIa uses the same method of parameter variation as Tables Ia and Ib but valid variations are not achieved because for a
The first part of Table IIb contains results obtained by using combinations in the same way as Tables Ia, Ib and IIa, the parameter that varies being Tf whereas the others are kept constant. Used combinations start at (3.5; 0.35; 1) and end at (3.5; 0.15; 1). In this first set the
smaller one should be the higher detection rate.
outline the highest detection rates obtained.
**3.2 Fuzzy Kwan-Cai neural network results**
represents the neural network's sensitivity to errors (Tf [0.1; 0.35]).
Unsatisfactory results with high detection rates are presented in italic.
sample vectors that are not cohesive in nature with the rest of their class.
result to be considered satisfactory it should at least be higher than 50%.
where parameters Tf and β are kept constant whereas α varies.
0.22; 1) group.
Fig. 5. Flowchart for implemented Bayes Classifier
The testing method for both algorithms (Fuzzy Kwan-Cai and Kohonen Self-Organizing Map) is the following: one parameter varies between its theoretical limits whereas the others remain constant. It is obvious that the difference between the bigger training group and the smaller one should be the higher detection rate.
The following tables contain only the significant part of the experimental results in order to outline the highest detection rates obtained.
### **3.2 Fuzzy Kwan-Cai neural network results**
220 Fuzzy Logic – Algorithms, Techniques and Implementations
Fig. 5. Flowchart for implemented Bayes Classifier
This type of neural network does not need training cycles because it learns as it studies the vectors it receives and builds its own classes in the testing process. In order not to get the wrong idea from the start we have to mention that the high number of classes observed in Table I and Table II for this neural network is due to the second nature of the application which acts like a "focusing lens", examining the internal structure of the two main classes. Therefore it must be stated that the number of classes we are interested in, for this experiment, is two. The significance and proper function limits of this application for parameters given in Table I and Table II are: α which is the base of isosceles triangles (α [1.5; 3.5]), β that determines how much of the sample vector each fuzzy neuron sees (β [0.1; 1.6]) and Tf that represents the neural network's sensitivity to errors (Tf [0.1; 0.35]).
The first vector generates a class of its own, the next ones either are found relatives of one of the vectors that have come before and therefore are put in the same class or start a new class. The maximum detection results in the tables mentioned above are outlined by being bolded. Unsatisfactory results with high detection rates are presented in italic.
Tables Ia and Ib present the pressure sample detection rate results for the Fuzzy Kwan-Cai neural network using the small sample database and the large sample database. According to Table Ia, the highest detection rate value of 68% was obtained for combination (3.5; 0.15; 1) where parameters Tf and β are kept constant whereas α varies.
The same method has been used for Table Ib showing the combinations used by changing the parameter Tf while keeping constant the other two. Combinations are from (3.5; 0.35; 1) down to (3.5; 0.15; 1). A maximum correct detection rate of 93.40% was obtained for the (3.5; 0.22; 1) group.
The detection rate results in Tables Ia and Ib show that from this point of view the Fuzzy Kwan-Cai neural network is very stabile, small variations of its parameters not affecting the experimental outcome. It is clear from the results presented that an increase in the sample database leads to an increase in the detection rates, the network not being affected by sample vectors that are not cohesive in nature with the rest of their class.
Tables IIa and IIb contain the vibration sample detection results. Table IIa represents the small sample database and Table IIb the large one. Table IIa uses the same method of parameter variation as Tables Ia and Ib but valid variations are not achieved because for a result to be considered satisfactory it should at least be higher than 50%.
The first part of Table IIb contains results obtained by using combinations in the same way as Tables Ia, Ib and IIa, the parameter that varies being Tf whereas the others are kept constant. Used combinations start at (3.5; 0.35; 1) and end at (3.5; 0.15; 1). In this first set the
Engine Knock Detection Based on Computational Intelligence Methods 223
α Tf β Rate [%] No.classes 3.5 0.35÷0.29 1 48% 2 3.5 0.28 1 48% 3 3.5 0.27 1 62% 6 3.5 0.26 1 62% 6 *3.5 0.25 1 68% 9 3.5 0.24 1 68% 9* 3.5 0.23 1 62% 9 3.3 0.3 1 48% 3 3.3 0.3 0.9 48% 3 3.3 0.3 0.8 48% 3 3.3 0.3 0.7÷0.3 48% 2 3.3 0.3 0.2 48% 1
Table IIa. Vibration detection rates- small database
Table IIb. Vibration detection rates- large database
Fig. 7. Plot of Tf (blue),Rate[%] (red) and No.classes (green) (Table IIa)
α Tf β Rate [%] No. classes 3.5 0.35÷0.24 1 93.40% 1 **3.5 0.24 1 93.40% 2** 3.5 0.23 1 93.40% 3 3.5 0.22 1 93.40% 5 3.5 0.21 1 93.40% 9 3.5 0.2 1 82.05% 10 3.5 0.19 1 93.13% 32 3.5 0.18 1 92.34% 57 3.5 0.17 1 90.23% 107 3.5 0.16 1 85.75% 143 3.5 0.15 1 85.75% 200 3.3 0.3 1 85.10% 3 3.3 0.3 0.9 85.10% 3 3.3 0.3 0.8 85.10% 3 **3.3 0.3 0.7÷0.3 85.10% 2** 3.3 0.3 0.2 85.10% 1
Table Ia. Pressure detection rates- small database
Fig. 6. Plot of α (blue) ,Rate[%](green) and No. classes (red) (Table Ia)
Table Ib. Pressure detection rates- large database
222 Fuzzy Logic – Algorithms, Techniques and Implementations
α Tf β Rate [%] No. classes 3.5 0.15 1 68% 2 3.4 0.15 1 64% 4 3.3 0.15 1 64% 4 3.2 0.15 1 64% 4 3.1 0.15 1 64% 4 3 0.15 1 48% 5 2.9 0.15 1 48% 5 2.8 0.15 1 48% 7 2.7 0.15 1 48% 7 2.6 0.15 1 48% 7 2.5 0.15 1 72% 9 2.4 0.15 1 60% 9 2.3 0.15 1 58% 11 2.2 0.15 1 58% 11 2.1 0.15 1 58% 11 2 0.15 1 68% 12
Table Ia. Pressure detection rates- small database
Table Ib. Pressure detection rates- large database
Fig. 6. Plot of α (blue) ,Rate[%](green) and No. classes (red) (Table Ia)
α Tf β Rate [%] No. classes 3.5 0.35÷0.23 1 93.40% 1 3.5 0.22 1 93.40% 2 3.5 0.21 1 93.40% 3 3.5 0.2 1 93.40% 3 3.5 0.19 1 93.40% 4 3.5 0.18 1 93.40% 5 3.5 0.17 1 93.40% 5 3.5 0.16 1 93.40% 8 3.5 0.15 1 93.40% 10 3.5 0.35÷0.23 1 93.40% 1
Table IIa. Vibration detection rates- small database
Fig. 7. Plot of Tf (blue),Rate[%] (red) and No.classes (green) (Table IIa)
Table IIb. Vibration detection rates- large database
Engine Knock Detection Based on Computational Intelligence Methods 225
What can be observed from the start is that the bigger sample group has almost equal detection times in both pressure and vibration cases to the smaller group, a significant increase being shown in the detection rates. The average detection times in Table III show that via optimization the network can be used in real–time knock applications with very
One can observe for the Fuzzy Kwan-Cai algorithm that different combinations of parameters can produce the same detection rates, so that a linear variation in any of the
The Kohonen–Self Organizing Map has a separate learning stage taking place before the detection process begins and being composed of epochs. After the learning stage has ended
For this neural network three sizes of neural maps were used – nine, one-hundred and fourhundred neurons –, as shown in Tables IV, V, VI. They were tested on both pressure and
Table IVa contains only the pressure sample detection rate results for the small vector database using the one hundred–neuron configuration. By keeping the number of epochs constant at 100 and the learning rate at 0.2 and by means of a variation of the neighborhood size from 90 down to 10, we obtained the following spike values: a detection rate of 80% marked bold-italic for the (100; 0.2; 83) group and the maximum value of the detection rate
Table IVb contains the pressure sample detection rates using the large database. From the start, using the nine-neuron map, an important fact appears: the nine-neuron map can not cope with the large database due to the small number of neurons that have to remember a large amount of samples, leading to confusion and very low detection rates. The variation methods are the same ones as in the complete version of Table IVa but, even by varying each of the parameters and keeping the other two constant, we can not obtain a spike value higher than 29.78% marked italic, value resulting from the combination (100; 0.4; 5). Performing the same variation techniques as in Table IVa, the maximum value for the detection rate in Table IVb results of 90.57% from the (400; 0.2; 400) and (500; 0.2; 400) combinations – both marked bold - , with lower but not less important spikes of 89.66% for
Table Va contains the vibration sample detection rates for the small database. The same variation methods as those in Tables IVa and IVb were used for the exact same values. The one-hundred-neuron network encounters its top value of 80% for the (100; 0.2; 95) combination and also a smaller spike of 74.28% for (100; 0.2; 60). The four-hundred-neuron network tops out at the 82.85% detection rate for the (300; 0.2; 400) combination of parameters. The same marking methods as in the previous tables were also used here and in the following ones.
The large database results for the vibration sample vectors are found in Table Vb. These values have come from the same methods of testing and values used in Tables IVa, IVb and Va. As in the case of the complete Table IVa (from which only the one-hundred neuron
good detection rates and with no prior in-factory learning processes.
**3.3 Kohonen Self–Organizing Map neural network results**
vibration samples.
parameters will not always lead to a linear variation in the detection rate.
it does not need to be repeated and the processing of the test batch begins.
for the small database 82.85% marked bold for the (100; 0.2; 82) combination.
(100; 0.2; 400) and (100; 0.3; 400) – marked bold-italic.
Fig. 8. Plot of Tf (blue) ,Rate[%](red) and No.classes (green) (Table IIb)
maximum correct detection rate of 93.40% is achieved for (3.5; 0.24; 1) – values bolded. Set two contains combinations from (3.3; 0.3; 1) down to (3.3; 0.3; 0.2), parameter β varying between 1 and 0.2 and the other two staying constant. A detection value not as high but equally as important as the maximum one obtained in the previous set is showed in combinations from (3.3; 0.3; 0.7) to (3.3; 0.3; 0.3). The value is 85.10% and presents interest because it is a much higher value than the ones constantly obtained and also represents a correct class detection of two classes.
The vibration situation presented in Tables IIa and IIb leads us to the same results revealed by Tables Ia and Ib, that an increase in the database size will lead to a substantial increase in the detection rate.
In the case of the large sample group shown in Tables Ia and Ib, respectively in Tables IIa and IIb, the neural network does not show any difference in maximum detection rates, differences being observed only for the small sample group. Both tables also present the same maximum detection rate, showing that the network can learn to identify both types of samples with the same accuracy.
Table III presents the time situation. It contains the average detection time situation for both pressure and vibration samples and also from a small and large database point of view. It is clear that the large database obtains better results with almost equally small detection times – 0.0022s for pressure and 0.0046s for vibration – and that pressure vectors have the tendency of being faster detected than vibration ones because the pressure group is more coherent and homogenous than vibration group.
Table III. Average detection times representing pressure and vibration for both small and large databases
224 Fuzzy Logic – Algorithms, Techniques and Implementations
maximum correct detection rate of 93.40% is achieved for (3.5; 0.24; 1) – values bolded. Set two contains combinations from (3.3; 0.3; 1) down to (3.3; 0.3; 0.2), parameter β varying between 1 and 0.2 and the other two staying constant. A detection value not as high but equally as important as the maximum one obtained in the previous set is showed in combinations from (3.3; 0.3; 0.7) to (3.3; 0.3; 0.3). The value is 85.10% and presents interest because it is a much higher value than the ones constantly obtained and also represents a
The vibration situation presented in Tables IIa and IIb leads us to the same results revealed by Tables Ia and Ib, that an increase in the database size will lead to a substantial increase in
In the case of the large sample group shown in Tables Ia and Ib, respectively in Tables IIa and IIb, the neural network does not show any difference in maximum detection rates, differences being observed only for the small sample group. Both tables also present the same maximum detection rate, showing that the network can learn to identify both types of
Table III presents the time situation. It contains the average detection time situation for both pressure and vibration samples and also from a small and large database point of view. It is clear that the large database obtains better results with almost equally small detection times – 0.0022s for pressure and 0.0046s for vibration – and that pressure vectors have the tendency of being faster detected than vibration ones because the pressure group is more
> Pressur e
Large database **0.0022 0.0046** Small database 0.0052 0.0056
Table III. Average detection times representing pressure and vibration for both small and
Vibratio n
Fig. 8. Plot of Tf (blue) ,Rate[%](red) and No.classes (green) (Table IIb)
correct class detection of two classes.
samples with the same accuracy.
coherent and homogenous than vibration group.
Average detection time [s]
the detection rate.
large databases
What can be observed from the start is that the bigger sample group has almost equal detection times in both pressure and vibration cases to the smaller group, a significant increase being shown in the detection rates. The average detection times in Table III show that via optimization the network can be used in real–time knock applications with very good detection rates and with no prior in-factory learning processes.
One can observe for the Fuzzy Kwan-Cai algorithm that different combinations of parameters can produce the same detection rates, so that a linear variation in any of the parameters will not always lead to a linear variation in the detection rate.
### **3.3 Kohonen Self–Organizing Map neural network results**
The Kohonen–Self Organizing Map has a separate learning stage taking place before the detection process begins and being composed of epochs. After the learning stage has ended it does not need to be repeated and the processing of the test batch begins.
For this neural network three sizes of neural maps were used – nine, one-hundred and fourhundred neurons –, as shown in Tables IV, V, VI. They were tested on both pressure and vibration samples.
Table IVa contains only the pressure sample detection rate results for the small vector database using the one hundred–neuron configuration. By keeping the number of epochs constant at 100 and the learning rate at 0.2 and by means of a variation of the neighborhood size from 90 down to 10, we obtained the following spike values: a detection rate of 80% marked bold-italic for the (100; 0.2; 83) group and the maximum value of the detection rate for the small database 82.85% marked bold for the (100; 0.2; 82) combination.
Table IVb contains the pressure sample detection rates using the large database. From the start, using the nine-neuron map, an important fact appears: the nine-neuron map can not cope with the large database due to the small number of neurons that have to remember a large amount of samples, leading to confusion and very low detection rates. The variation methods are the same ones as in the complete version of Table IVa but, even by varying each of the parameters and keeping the other two constant, we can not obtain a spike value higher than 29.78% marked italic, value resulting from the combination (100; 0.4; 5). Performing the same variation techniques as in Table IVa, the maximum value for the detection rate in Table IVb results of 90.57% from the (400; 0.2; 400) and (500; 0.2; 400) combinations – both marked bold - , with lower but not less important spikes of 89.66% for (100; 0.2; 400) and (100; 0.3; 400) – marked bold-italic.
Table Va contains the vibration sample detection rates for the small database. The same variation methods as those in Tables IVa and IVb were used for the exact same values. The one-hundred-neuron network encounters its top value of 80% for the (100; 0.2; 95) combination and also a smaller spike of 74.28% for (100; 0.2; 60). The four-hundred-neuron network tops out at the 82.85% detection rate for the (300; 0.2; 400) combination of parameters. The same marking methods as in the previous tables were also used here and in the following ones.
The large database results for the vibration sample vectors are found in Table Vb. These values have come from the same methods of testing and values used in Tables IVa, IVb and Va. As in the case of the complete Table IVa (from which only the one-hundred neuron
Engine Knock Detection Based on Computational Intelligence Methods 227
Fig. 10. Plot of No. Neurons (blue), Neighborhood (red) and Rate[%] (green)(Table IVb)
Fig. 11. Plot No. Neurons (blue) Neighborhood(red) and Rate[%] green)(Table Va)
Table Va. Vibration detection rates- small database
No. neurons Epochs Learning rate Neighborhood Rate [%] **100 100 0.2 95 80** 100 100 0.2 90 65.71 100 100 0.2 80 65.71 100 100 0.2 70 57.14 **100 100 0.2 60 74.28** 100 100 0.2 50 65.71 100 100 0.2 40 65.71 100 100 0.2 30 68.57 100 100 0.2 20 60 400 100 0.2 400 65.71 400 200 0.2 400 71.42 **400 300 0.2 400 82.85** 400 400 0.2 400 65.71 400 500 0.2 400 62.85 400 600 0.2 400 71.42
Table IVa. Pressure detection rates- small database
Fig. 9. Plot of No. Neurons (blue), Neighborhood (red) and Rate[%] (green)(Table IVa)
Table IVb. Pressure detection rates- large database
226 Fuzzy Logic – Algorithms, Techniques and Implementations
No. neurons Epochs Learning rate Neighborhood Rate [%] 100 100 0.2 90 68.57 *100 100 0.2 83 80* **100 100 0.2 82 82.85** 100 100 0.2 80 77.14 100 100 0.2 70 74.28 100 100 0.2 60 68.57 100 100 0.2 50 71.42 100 100 0.2 40 71.42 100 100 0.2 30 71.42 100 100 0.2 20 77.14 100 100 0.2 10 65.71
Fig. 9. Plot of No. Neurons (blue), Neighborhood (red) and Rate[%] (green)(Table IVa)
No. neurons Epochs Learning rate Neighborhood Rate [%] 9 100 0.2 5 23.03 9 100 0.3 5 27.35 *9 100 0.4 5 29.78 9 100 0.5 5 28.57* 9 100 0.6 5 19.75 9 100 0.7 5 20.06 *400 100 0.2 400 89.66 400 100 0.3 400 89.96* 400 100 0.4 400 87.84 400 100 0.5 400 88.75 400 100 0.6 400 89.96 400 100 0.7 400 88.75 *400 100 0.2 400 89.66* 400 200 0.2 400 89.36 400 300 0.2 400 88.75 **400 400 0.2 400 90.57 400 500 0.2 400 90.57**
Table IVa. Pressure detection rates- small database
Table IVb. Pressure detection rates- large database
Fig. 10. Plot of No. Neurons (blue), Neighborhood (red) and Rate[%] (green)(Table IVb)
Table Va. Vibration detection rates- small database
Fig. 11. Plot No. Neurons (blue) Neighborhood(red) and Rate[%] green)(Table Va)
Engine Knock Detection Based on Computational Intelligence Methods 229
Table VI represent the average detection times using both pressure and vibration vectors for both small and large databases. With values of 0.0023s (small database) and 0.0024s (large database) the pressure samples obtain smaller detection times than the vibration samples with 0.0027s (small database) and 0.0028s (large database). This situation is representative for the four-hundred-neuron network, this also being the slowest solution but with the highest detection rates. The nine-neuron network, even though it has the best detection times, can not be taken into account as a real application because it is not able to cope with large database. The one-hundred-neuron network is the best compromise between detection
As with the previous described algorithms, the SOM results shown in Tables IV and V that an increase in the sample group size (training set case) will lead to an increase in detection rates. In this case, the two separate groups are not separated by big detection rate gaps.
samples
SOM - 400 neurons 0.0023 0.0027 0.0024 0.0028 SOM - 100 neurons **0.000193 0.000478 0.000538 0.000498** SOM - 9 neurons 0.0000576 0.0000579 0.0000535 0.0000872 Table VI. Pressure and vibration average detection times for both small and large sample
As in theory, the experimental results in Tables IV, V and VI show that with the increase in neurons there is an increase in detection rates but a decrease in detection times because more neurons translate to more detail that can be remembered, so the distinction between knock and non-knock situations can be more precisely done - therefore a compromise must be made. Being interested not only in obtaining high detection rates but also detection times that would be coherent to the task at hand (samples must be processed in under an engine cycle so the modifications can be brought to the next one), the one-hundred-neuron map seems to be the best option from the three methods tested. The nine-neuron map, even if it produces very high detection times, has a very poor detection rate in both pressure and
The four-hundred-neuron map presented the highest detection rates for this neural network, values that are a little bit smaller than the Fuzzy Kwan-Cai but with detection times very similar to it, the only difference being that the SOM needs separate training. In this case, looking at the detection times in Table VI, the SOM does not seem to make any difference between pressure and vibration signals, the medium detection times showing very small variations. There is a small difference in detection rates between pressure and vibration
A very important factor in the good working of the Kohonen Self-Organizing Map is getting the number of epochs and the learning rate well calibrated. A greater than necessary number of epochs would lead to the situation where the network learns in the necessary time period but it is left with more epochs that are not used for learning. This situation, in combination with a high learning rate, would lead to the network learning everything very fast in the first epochs and then forgetting or distorting the knowledge in the following ones.
Pressure samples Vibration
vibration groups making it useless for any further applications.
samples; the SOM seems to handle both models very well.
Small database Large database
Pressure samples
Vibration samples
speed and detection rates as shown in this table.
Average detection time [s]
databases
Table Vb. Vibration detection rates- large database
Fig. 12. Plot of No. Neurons (blue), Neighborhood (red) and Rate[%] (green) (Table Vb)
section has been presented in this chapter), the nine-neuron network in the complete Table Vb is not suited for working with such a large database, the network becoming confused. This shows in constant results under 50% which can not be taken into account as valid experimental results. These values can only be used as examples of exceptional cases. The one-hundred-neuron network section presented in Table Vb obtains a maximum detection rate of 81.76% for combinations (100; 0.2; 50), another important value over 80% being of 81.15 % for (100; 0.2; 70) . The four-hundred-neuron network tops out at 89.66% for combinations (100; 0.2; 250) and present other important values of 89.36% for (100; 0.2, 325) and of 89,05% for (100; 0.2; 375).
228 Fuzzy Logic – Algorithms, Techniques and Implementations
Neighborho
od Rate [%]
rate
Fig. 12. Plot of No. Neurons (blue), Neighborhood (red) and Rate[%] (green) (Table Vb)
section has been presented in this chapter), the nine-neuron network in the complete Table Vb is not suited for working with such a large database, the network becoming confused. This shows in constant results under 50% which can not be taken into account as valid experimental results. These values can only be used as examples of exceptional cases. The one-hundred-neuron network section presented in Table Vb obtains a maximum detection rate of 81.76% for combinations (100; 0.2; 50), another important value over 80% being of 81.15 % for (100; 0.2; 70) . The four-hundred-neuron network tops out at 89.66% for combinations (100; 0.2; 250) and present other important values of 89.36% for (100; 0.2, 325)
100 100 0.2 95 79.63 100 100 0.2 90 79.93 100 100 0.2 80 79.02 *100 100 0.2 70 81.15* 100 100 0.2 60 78.11 **100 100 0.2 50 81.76** 100 100 0.2 40 75.98 100 100 0.2 30 79.93 100 100 0.2 20 76.59 400 100 0.2 400 87.53 *400 100 0.2 375 89.05* 400 100 0.2 350 88.75 *400 100 0.2 325 89.36* 400 100 0.2 300 86.83 400 100 0.2 275 86.62 **400 100 0.2 250 89.66** 400 100 0.2 225 88.75 400 100 0.2 200 88.44 400 100 0.2 175 88.44
No. neurons Epochs Learning
Table Vb. Vibration detection rates- large database
and of 89,05% for (100; 0.2; 375).
Table VI represent the average detection times using both pressure and vibration vectors for both small and large databases. With values of 0.0023s (small database) and 0.0024s (large database) the pressure samples obtain smaller detection times than the vibration samples with 0.0027s (small database) and 0.0028s (large database). This situation is representative for the four-hundred-neuron network, this also being the slowest solution but with the highest detection rates. The nine-neuron network, even though it has the best detection times, can not be taken into account as a real application because it is not able to cope with large database. The one-hundred-neuron network is the best compromise between detection speed and detection rates as shown in this table.
As with the previous described algorithms, the SOM results shown in Tables IV and V that an increase in the sample group size (training set case) will lead to an increase in detection rates. In this case, the two separate groups are not separated by big detection rate gaps.
Table VI. Pressure and vibration average detection times for both small and large sample databases
As in theory, the experimental results in Tables IV, V and VI show that with the increase in neurons there is an increase in detection rates but a decrease in detection times because more neurons translate to more detail that can be remembered, so the distinction between knock and non-knock situations can be more precisely done - therefore a compromise must be made. Being interested not only in obtaining high detection rates but also detection times that would be coherent to the task at hand (samples must be processed in under an engine cycle so the modifications can be brought to the next one), the one-hundred-neuron map seems to be the best option from the three methods tested. The nine-neuron map, even if it produces very high detection times, has a very poor detection rate in both pressure and vibration groups making it useless for any further applications.
The four-hundred-neuron map presented the highest detection rates for this neural network, values that are a little bit smaller than the Fuzzy Kwan-Cai but with detection times very similar to it, the only difference being that the SOM needs separate training. In this case, looking at the detection times in Table VI, the SOM does not seem to make any difference between pressure and vibration signals, the medium detection times showing very small variations. There is a small difference in detection rates between pressure and vibration samples; the SOM seems to handle both models very well.
A very important factor in the good working of the Kohonen Self-Organizing Map is getting the number of epochs and the learning rate well calibrated. A greater than necessary number of epochs would lead to the situation where the network learns in the necessary time period but it is left with more epochs that are not used for learning. This situation, in combination with a high learning rate, would lead to the network learning everything very fast in the first epochs and then forgetting or distorting the knowledge in the following ones.
Engine Knock Detection Based on Computational Intelligence Methods 231
Training vectors Test vectors Press. rate [%] Vib. rate [%] 11 90 65.50 55.55 12 89 56.17 58.42 **13 88 72.50 60.22** 15 86 51.16 53.48 21 80 63.75 66.25 28 73 54.79 64.38 35 66 59.09 60.60 41 60 68.33 70 47 54 59.25 68.51 55 46 56.52 73.91 61 40 67.50 70 65 36 52.77 77.77 67 34 68.57 74.28 75 26 50 76.92 80 21 42.85 76.19
Table VIIa. Pressure - vibration detection rates- small database
Table VIIb. Pressure - vibration detection rates- large database
Fig. 13. Test vectors (blue), pressure (red) and vibration rates(green) (Table VIIa)
Training vectors Test vectors Press. rate [%] Vib. rate [%] **371 629 93.64 90.30** 391 609 93.43 90.14 411 589 93.20 90.32 431 569 92.97 89.98 451 549 92.71 89.79 471 529 92.43 89.60 491 509 92.14 89.58 511 489 91.82 89.77 531 469 91.42 89.55 **551 449 91.09 89.08 571 429 90.67 89.04 591 409 91.44 89.48** 611 389 92.28 90.23 631 369 93.22 91.32 651 349 94.26 92.26
### **3.4 Bayes classifier results**
The Bayes Classifier, as described by its name, is not a neural network but has been included in this chapter as a basic reference point for the evaluation of the two neural networks. It uses a method of calculating the minimum distance from a sample to one of the knock or non-knock class centers - classes that are considered Gaussian by nature. That is why it presents the worst detection times, as shown in Table VIII.
Table VIIa represents the combined pressure and vibration detection rates status for the small database. The way the testing has been done for this algorithm is by progressively growing from a small comparison group (the batch of samples chosen to represent the known classes for testing) versus large test group situation, to a large comparison group versus small test group situation.
The process starts out with a balance of 11 training vectors and 90 testing ones, which leads to a detection rate starting from 65.50% for pressure and 55.55% for vibration and grows (for training vectors) versus shrinks (for testing vectors) in a progressive way to 85 training vectors and 16 testing vectors, leading to a detection rate ending at 43.75% for pressure and 81.25% for vibration. An interesting detail can be observed in this table: the pressure vectors seem to present a constant state even though more and more are added to the learning group every time the detection rates stay approximately between 50% and 72.50%, the last value being the highest pressure detection rate.
The change of state occurs at the end of the table where we can observe a decrease in the learning rate for the combinations of (80 training vectors; 21 testing vectors) with a detection rate of 42.85% and (85 training vectors; 16 testing vectors) with a detection rate of 43.75%.
This decrease is due to the inclusion in the learning group of vectors that are radically different from their stated class; therefore, the knock or non-knock distinction can not be made. In the case of the vibration sample vectors the progression is of almost uniform growth from 55.55% to 81.25%, the last being also the maximum detection rate for the small database experiment.
Table VIIb follows the same type of progression, only that the large database is used for both pressure and vibration samples. The progression goes from a combination of (371 training vectors; 629 testing vectors) with a detection rate of 93.64% for pressure and 90.30% for vibration samples to a combination of (671 training vectors; 329 testing vectors) with the maximum detection rate achieved in this table of 95.44% for pressure samples and 92.40% for vibration samples. Within this progression it can be seen more clearly that the pressure samples are very cohesive in nature and that, given enough samples, the algorithm goes past the problems it has with radically different sample vectors, maintaining a detection rate over 90% in every case.
Table VIII represents the average detection times for both the small and large databases using both pressure and vibration samples.
Being a simple comparative algorithm, we can see in Table VIII that an increase in the database size leads to a slowing down of the process because the comparison must be made with more vectors. In the case of the small database, pressure vectors are detected faster 230 Fuzzy Logic – Algorithms, Techniques and Implementations
The Bayes Classifier, as described by its name, is not a neural network but has been included in this chapter as a basic reference point for the evaluation of the two neural networks. It uses a method of calculating the minimum distance from a sample to one of the knock or non-knock class centers - classes that are considered Gaussian by nature. That is why it
Table VIIa represents the combined pressure and vibration detection rates status for the small database. The way the testing has been done for this algorithm is by progressively growing from a small comparison group (the batch of samples chosen to represent the known classes for testing) versus large test group situation, to a large comparison group
The process starts out with a balance of 11 training vectors and 90 testing ones, which leads to a detection rate starting from 65.50% for pressure and 55.55% for vibration and grows (for training vectors) versus shrinks (for testing vectors) in a progressive way to 85 training vectors and 16 testing vectors, leading to a detection rate ending at 43.75% for pressure and 81.25% for vibration. An interesting detail can be observed in this table: the pressure vectors seem to present a constant state even though more and more are added to the learning group every time the detection rates stay approximately between 50% and 72.50%, the last
The change of state occurs at the end of the table where we can observe a decrease in the learning rate for the combinations of (80 training vectors; 21 testing vectors) with a detection rate of 42.85% and (85 training vectors; 16 testing vectors) with a detection rate of 43.75%.
This decrease is due to the inclusion in the learning group of vectors that are radically different from their stated class; therefore, the knock or non-knock distinction can not be made. In the case of the vibration sample vectors the progression is of almost uniform growth from 55.55% to 81.25%, the last being also the maximum detection rate for the small
Table VIIb follows the same type of progression, only that the large database is used for both pressure and vibration samples. The progression goes from a combination of (371 training vectors; 629 testing vectors) with a detection rate of 93.64% for pressure and 90.30% for vibration samples to a combination of (671 training vectors; 329 testing vectors) with the maximum detection rate achieved in this table of 95.44% for pressure samples and 92.40% for vibration samples. Within this progression it can be seen more clearly that the pressure samples are very cohesive in nature and that, given enough samples, the algorithm goes past the problems it has with radically different sample vectors, maintaining a detection rate
Table VIII represents the average detection times for both the small and large databases
Being a simple comparative algorithm, we can see in Table VIII that an increase in the database size leads to a slowing down of the process because the comparison must be made with more vectors. In the case of the small database, pressure vectors are detected faster
**3.4 Bayes classifier results**
versus small test group situation.
database experiment.
over 90% in every case.
using both pressure and vibration samples.
value being the highest pressure detection rate.
presents the worst detection times, as shown in Table VIII.
Table VIIa. Pressure - vibration detection rates- small database
Fig. 13. Test vectors (blue), pressure (red) and vibration rates(green) (Table VIIa)
Table VIIb. Pressure - vibration detection rates- large database
Engine Knock Detection Based on Computational Intelligence Methods 233
<sup>68</sup> 82,85 68,57
Kwan- Cai SOM Bayes
Detection rates (Pressure samples) [%]
(a)
Kwan- Cai SOM Bayes
90,57
Detection rates (Pressure samples) [%]
(b) Fig. 15. Pressure sample detection rates using the small database (a) and the large database
An increase in the database size from one hundred to one thousand sample vectors will lead to a minimum increase of ten percent in the detection rates. For the small database, the Fuzzy Kwan-Cai neural network obtains maximum detection rates for the pressure samples at 68% that are higher than the ones for vibration samples at 48%, but after using the large data set the maximum pressure and vibration detection rates become equal at 93.40%. The difference in detection rates for the pressure and vibration samples using the small database shows that the pressure samples are more coherent and therefore easier to classify. The same evolution as shown by the Fuzzy Kwan- Cai is also true for the Kohonen Self-Organizing Map (SOM). Even more so, the increase in learning database size will lead to a
The second discussion will be based on the detection rate point of view. As shown in Fig.15 and Fig.16, the Bayes Classifier seems to show the best detection rates. Its fault is that it needs large amounts of comparison data in order to create classes that are comprehensive enough. Out of the three algorithms tested in this chapter, it is also the less stabile due to the fact that it calculates distances to the center of the comparison classes. If these classes are not well defined and separated, the detection rates fall dramatically. This can be seen in Table VIIb. The Fuzzy Kwan-Cai obtains the highest detection rates of all three algorithms - these being valid detection rates that are not influenced by the nature of learned vectors leading to the great stability of this method. The learning method used employs the automatic generation of learning classes as it goes through the sample set. The fuzzy logic creates a more organic representation of the knowledge classes than the boolean one. The Kohonen Self-Organizing Map (SOM) presents the second highest detection rates and a more controlled and stabile learning and training environment then the other two algorithms. Because the learning is done prior to the start of the testing process and in repetitive epochs, the neural network has the chance to go through the data set again and again until a
95,44
93,4
(b) for the Kwan- Cai, SOM neural networks and the Bayes Classifier
theoretical increase in the detection rate of the Bayes Classifier.
Fig. 14. Test vectors (red), training vectors (blue), pressure rates (green) and vibration rates (violet)
Table VIII. Pressure and vibration average detection times for both small and large sample databases
(0.0287s) than vibration samples (0.0297s). The large database experiments lead to almost equal average detection times between pressure (0.0948s) and vibration (0.094s) samples, with a tendency to better recognize vibration samples.
There is little relevance in the detection rates for the small sample group, even though a small variation between pressure and vibration can be seen. The increase in detection rates due to a bigger knowledge database can also be seen from Tables VIIa and VIIb.
The greatest importance of the Bayes Classifier in this chapter comes from its great sensitivity to change. When the knowledge group includes vectors that are incoherent with the others or that are more different, the detection rate goes down immediately. In this case, the algorithm can not classify properly because one or both classes contain vectors that are very far away from their centers and vectors from one class may get tangled up with the other one. By doing this the Bayes Classifier acts as a monitor for change in the constitution of the sample classes or a "magnifying glass" reacting to the internal composition of the data groups.
Given a big enough knowledge database that is also very coherent in the nature of its classes, the detection rates go up and can be comparable to the neural networks but at a great cost in speed.
### **4. Comparison among the three tested methods**
The first discussion will be based on the database size point of view. As we can see from Fig.15 and Fig.16 that summarize results in Tables I, II, IV, V and VII, the size of the learning, training or comparison database is very important in the good functioning of all three tested algorithms.
232 Fuzzy Logic – Algorithms, Techniques and Implementations
Fig. 14. Test vectors (red), training vectors (blue), pressure rates (green) and vibration rates
Small sample database **0.0287** 0.0297 Large sample database 0.0948 **0.094** Table VIII. Pressure and vibration average detection times for both small and large sample
(0.0287s) than vibration samples (0.0297s). The large database experiments lead to almost equal average detection times between pressure (0.0948s) and vibration (0.094s) samples,
There is little relevance in the detection rates for the small sample group, even though a small variation between pressure and vibration can be seen. The increase in detection rates
The greatest importance of the Bayes Classifier in this chapter comes from its great sensitivity to change. When the knowledge group includes vectors that are incoherent with the others or that are more different, the detection rate goes down immediately. In this case, the algorithm can not classify properly because one or both classes contain vectors that are very far away from their centers and vectors from one class may get tangled up with the other one. By doing this the Bayes Classifier acts as a monitor for change in the constitution of the sample classes or
Given a big enough knowledge database that is also very coherent in the nature of its classes, the detection rates go up and can be comparable to the neural networks but at a
The first discussion will be based on the database size point of view. As we can see from Fig.15 and Fig.16 that summarize results in Tables I, II, IV, V and VII, the size of the learning, training or comparison database is very important in the good functioning of all
due to a bigger knowledge database can also be seen from Tables VIIa and VIIb.
a "magnifying glass" reacting to the internal composition of the data groups.
[s] Pressure Vibration
Average detection time
with a tendency to better recognize vibration samples.
**4. Comparison among the three tested methods**
(violet)
databases
great cost in speed.
three tested algorithms.
Fig. 15. Pressure sample detection rates using the small database (a) and the large database (b) for the Kwan- Cai, SOM neural networks and the Bayes Classifier
An increase in the database size from one hundred to one thousand sample vectors will lead to a minimum increase of ten percent in the detection rates. For the small database, the Fuzzy Kwan-Cai neural network obtains maximum detection rates for the pressure samples at 68% that are higher than the ones for vibration samples at 48%, but after using the large data set the maximum pressure and vibration detection rates become equal at 93.40%. The difference in detection rates for the pressure and vibration samples using the small database shows that the pressure samples are more coherent and therefore easier to classify. The same evolution as shown by the Fuzzy Kwan- Cai is also true for the Kohonen Self-Organizing Map (SOM). Even more so, the increase in learning database size will lead to a theoretical increase in the detection rate of the Bayes Classifier.
The second discussion will be based on the detection rate point of view. As shown in Fig.15 and Fig.16, the Bayes Classifier seems to show the best detection rates. Its fault is that it needs large amounts of comparison data in order to create classes that are comprehensive enough. Out of the three algorithms tested in this chapter, it is also the less stabile due to the fact that it calculates distances to the center of the comparison classes. If these classes are not well defined and separated, the detection rates fall dramatically. This can be seen in Table VIIb. The Fuzzy Kwan-Cai obtains the highest detection rates of all three algorithms - these being valid detection rates that are not influenced by the nature of learned vectors leading to the great stability of this method. The learning method used employs the automatic generation of learning classes as it goes through the sample set. The fuzzy logic creates a more organic representation of the knowledge classes than the boolean one. The Kohonen Self-Organizing Map (SOM) presents the second highest detection rates and a more controlled and stabile learning and training environment then the other two algorithms. Because the learning is done prior to the start of the testing process and in repetitive epochs, the neural network has the chance to go through the data set again and again until a
Engine Knock Detection Based on Computational Intelligence Methods 235
It is clear from the information presented in this chapter that the best detection rates correlated to very good detection times belong to the Kohonen Self-Organizing Map with a
The SOM with a configuration of four-hundred-neurons obtains results almost similar to the Fuzzy Kwan-Cai. The difference between these two networks is that the SOM requires a separate training stage where the separated and well defined learning classes are given to it and the Fuzzy Kwan-Cai learns as it receives sample vectors and builds its own classes.
The Bayes Classifier is very useful for showing the nature of the knock and non-knock classes how well they are defined and separated due to its sensitivity to drastic variations in sample vectors. Its detection rate depends on the size of the knowledge database and its
From a real-world application point of view, in order to further maximize detection rates, it is clear that a parallel process composed of a pressure-vibration analysis and detection becomes necessary, based on the experimental results. Due to the developments in digital signal processing (DSP) technology, the parallel process would not lead to an increasing
In order to avoid overcrowding, this final chapter contains general concluding remarks due to the fact that details and accurate conclusions have already been widely presented in
Three methods of knock detection were studied and compared in this chapter. Testing was performed on a Bosch Group database. Two of the three algorithms used are of neural nature: Fuzzy Kwan-Cai neural network – presenting the unsupervised learning approach and fuzzy inference core - and Kohonen Self-Organizing Map (SOM) – with a separate
The three algorithms were either trained or had comparison classes and were tested on two different database sizes, one small of one hundred samples vectors and one large representing one thousand samples in order to show how the database size would affect the
Experiments were made on both pressure and vibration sample vectors in order to see which of these are more coherent in nature, leading to results that show an overall greater coherence with slightly more increased detection rates and how this coherence might affect the algorithms being tested. The experiments performed have led to results that prove the superiority of the neural methods in contrast to the normal classification – the situation being looked at from a rate-time point of view as seen in Fig.15, Fig.16, Fig.17, Fig.18.The difference between the neural and non neural methods is represented by an average scale factor of 0,001s in favour of the neural. This superiority should be seen also from a stability to errors point of view as seen in Table VIIb where a stray vector can distort the judgement
Comparisons were made between the algorithms leading to experimental results enabling us to draw conclusions on which methods are superior to others, in what way and also on
supervised learning stage - and the third is non-neural: Bayes Classifier.
of the non neural Bayes Classifier so that detection rates fall.
the properties and nature of the database used in the experiments.
configuration of one-hundred-neurons.
detection times.
**5. Concluding remarks**
chapters III and IV above.
detection outcome.
coherence making it useless in real-world applications.
complete image is formed. The two neural networks show no considerable preference between pressure and vibration samples and present high stability to drastic variations in training samples which in a non-neural method could cause a decrease in detection rates. The nature of these types of signals and their differences are outlined by the Bayes Classifiers sensitivity to unclear classes and the way in which the Fuzzy Kwan-Cai neural network works by showing the internal structure of the classes.
Fig. 16. Vibration sample detection rates using the small database (a) and the large database (b) for the Kwan-Cai, SOM neural networks and the Bayes Classifier
The third discussion will be based on the detection time point of view. As present in Fig.17 and Fig.18 that summarize results in Tables III, VI and VIII, it is clear at first glance that the neural networks are far superior to the normal non-neural classification algorithm. The Bayes Classifier obtains the longest detection times due to the process of comparing each new vector to the knowledge classes. The best, valid, detection times are shown by the Kohonen Self-Organizing Map with the one-hundred-neurons configuration. This configuration, given optimization of the code, can lead to detection times coherent to the engine combustion cycles in which the knock detection needs to take place. Any number of neurons under one hundred will make it hard for the network to give satisfactory detection rates even though the detection times will decrease dramatically. In this chapter we are interested in maximizing the balance between high detection rates and low detection times and not achieving the two extremes and having to compromise one outcome. The second best detection times that are also very close to one another belong to the Fuzzy Kwan-Cai and SOM with the configuration of four-hundred-neurons.
These two algorithms also show the highest detection rates from the methods tested in this chapter. In a real-time application there should not be any problem with the SOMs separate training stage because it would be performed only once inside the factory. The Fuzzy Kwan-Cai neural network presents a different advantage in that it can learn as it goes along, not needing a separate training stage and continuously receiving information and gaining knowledge.
It is clear from the information presented in this chapter that the best detection rates correlated to very good detection times belong to the Kohonen Self-Organizing Map with a configuration of one-hundred-neurons.
The SOM with a configuration of four-hundred-neurons obtains results almost similar to the Fuzzy Kwan-Cai. The difference between these two networks is that the SOM requires a separate training stage where the separated and well defined learning classes are given to it and the Fuzzy Kwan-Cai learns as it receives sample vectors and builds its own classes.
The Bayes Classifier is very useful for showing the nature of the knock and non-knock classes how well they are defined and separated due to its sensitivity to drastic variations in sample vectors. Its detection rate depends on the size of the knowledge database and its coherence making it useless in real-world applications.
From a real-world application point of view, in order to further maximize detection rates, it is clear that a parallel process composed of a pressure-vibration analysis and detection becomes necessary, based on the experimental results. Due to the developments in digital signal processing (DSP) technology, the parallel process would not lead to an increasing detection times.
### **5. Concluding remarks**
234 Fuzzy Logic – Algorithms, Techniques and Implementations
complete image is formed. The two neural networks show no considerable preference between pressure and vibration samples and present high stability to drastic variations in training samples which in a non-neural method could cause a decrease in detection rates. The nature of these types of signals and their differences are outlined by the Bayes Classifiers sensitivity to unclear classes and the way in which the Fuzzy Kwan-Cai neural
Fig. 16. Vibration sample detection rates using the small database (a) and the large database
The third discussion will be based on the detection time point of view. As present in Fig.17 and Fig.18 that summarize results in Tables III, VI and VIII, it is clear at first glance that the neural networks are far superior to the normal non-neural classification algorithm. The Bayes Classifier obtains the longest detection times due to the process of comparing each new vector to the knowledge classes. The best, valid, detection times are shown by the Kohonen Self-Organizing Map with the one-hundred-neurons configuration. This configuration, given optimization of the code, can lead to detection times coherent to the engine combustion cycles in which the knock detection needs to take place. Any number of neurons under one hundred will make it hard for the network to give satisfactory detection rates even though the detection times will decrease dramatically. In this chapter we are interested in maximizing the balance between high detection rates and low detection times and not achieving the two extremes and having to compromise one outcome. The second best detection times that are also very close to one another belong to the Fuzzy Kwan-Cai
These two algorithms also show the highest detection rates from the methods tested in this chapter. In a real-time application there should not be any problem with the SOMs separate training stage because it would be performed only once inside the factory. The Fuzzy Kwan-Cai neural network presents a different advantage in that it can learn as it goes along, not needing a
separate training stage and continuously receiving information and gaining knowledge.
(b) for the Kwan-Cai, SOM neural networks and the Bayes Classifier
and SOM with the configuration of four-hundred-neurons.
network works by showing the internal structure of the classes.
In order to avoid overcrowding, this final chapter contains general concluding remarks due to the fact that details and accurate conclusions have already been widely presented in chapters III and IV above.
Three methods of knock detection were studied and compared in this chapter. Testing was performed on a Bosch Group database. Two of the three algorithms used are of neural nature: Fuzzy Kwan-Cai neural network – presenting the unsupervised learning approach and fuzzy inference core - and Kohonen Self-Organizing Map (SOM) – with a separate supervised learning stage - and the third is non-neural: Bayes Classifier.
The three algorithms were either trained or had comparison classes and were tested on two different database sizes, one small of one hundred samples vectors and one large representing one thousand samples in order to show how the database size would affect the detection outcome.
Experiments were made on both pressure and vibration sample vectors in order to see which of these are more coherent in nature, leading to results that show an overall greater coherence with slightly more increased detection rates and how this coherence might affect the algorithms being tested. The experiments performed have led to results that prove the superiority of the neural methods in contrast to the normal classification – the situation being looked at from a rate-time point of view as seen in Fig.15, Fig.16, Fig.17, Fig.18.The difference between the neural and non neural methods is represented by an average scale factor of 0,001s in favour of the neural. This superiority should be seen also from a stability to errors point of view as seen in Table VIIb where a stray vector can distort the judgement of the non neural Bayes Classifier so that detection rates fall.
Comparisons were made between the algorithms leading to experimental results enabling us to draw conclusions on which methods are superior to others, in what way and also on the properties and nature of the database used in the experiments.
Engine Knock Detection Based on Computational Intelligence Methods 237
known value for each vector at a time and incremented into an error counter. The databases
This work was supported by CNCSIS – UEFISCSU, project number PNII – IDEI code
Adeli, H. & Karim , A. (2005). Wavelets in Intelligent Transportation Systems (1st edition) ,
Auld, T.; Moore, A.W. & Gull, S.F. (2007). Bayesian Neural Networks for Internet Traffic
Banakar A. & Azeem M. F. (2008). Articial wavelet neural network and its application in
Billings, S.A. & Wei H.L. (2005). A new class of wavelet networks for nonlinear system
Borg, J.M., Cheok K.C, Saikalis G. & Oho, S (2005). Wavelet-based knock detection with
Bosch, R. (2004). Bosch-Gasoline-Engine Management, Ed. Robert Bosch GmbH, ISBN-13:
Boubai, O. (2000). Knock detection in automobile engines, vol.3, issue 3, pp. 24-28, ISSN:
Chen, P.C. (2005). Neuro-fuzzy-based fault detection of the air flow sensor of an idling
Erjavec, J. (2009). Automotive Technology: A System Approach (5th edition), Ed. Delmar
Ettefagh, M M., Sadeghi, H., Pirouzpanah, V. H. & Arjmandi T. (2008). Knock detection in
Gupta, H.N. (2006). Fundamentals of Internal Combustion Engines, Ed. Prentice-Hall of
Hamilton, L J. & Cowart, J S. (2008). The first wide-open throttle engine cycle: transition into
Hsu, C.C. (2006). Generalizing self-organizing map for categorical data, vol. 17, issue.2, pp.
Hui, C.L. P. (2011). Artificial Neural Networks - Application, Publisher: InTech, ISBN 978-
Ibrahim, A. M. (2004). Fuzzy logic for embedded systems applications, Ed. Elsevier Science,
Jonathan, M.B., Saikalis, G., Oho, S.T. & Cheok, K.C. (2006). Knock Signal Analysis Using the Discrete Wavelet Transform, No. 2006-01-0226, DOI: 10.4271/2006-01-0226
spark ignition engines by vibration analysis of cylinder block: A parametric modeling approach, vol. 22, Issue 6, pp. 1495-1514, august 2008, ISSN: 0888-3270 Fleming, W.J. (2001). Overview of Automotive Sensors, vol.1, issue 4, pp.296-308, ISSN:
knock experiments with fast in-cylinder sampling, vol. 9, no. 2, pp. 97-109, ISSN
Cengage Learning, ISBN-13: 978-1428311497, Clifton Park NY USA
India Private Limited, ISBN-13: 978-8120328549, New Delhi India
neuro-fuzzy models, Appl. Soft Comput., vol. 8, no. 4, pp. 1463–1485, ISSN: 1568-4946
fuzzy logic, in IEEE International Conference on Computational Intelligence for Measurement Systems and Applications – CIMSA 2005, , pp.26-31, ISBN: 978-1-
Classification, vol. 18, issue. 1, pp. 223–239, ISSN: 1045-9227
identification, vol. 16, issue. 4, pp. 862 – 874, ISSN: 1045-9227
gasoline engine, vol.219, no. 4, pp.511-524, ISSN 0954-4070
were verified to be consistent of their description.
Ed. Wiley, ISBN-13: 978-0470867426, England
4244-2306-4, Sicily Italy 14-16 July 2005
978-0837611006
1094-6969
1530-437X
1468-0874
294 - 304, ISSN: 1045-9227
ISBN-13: 978-0750676052, MA USA
953-307-188-6, Croatia
**6. Acknowledgement**
1693/2008.
**7. References**
Fig. 17. Pressure sample detection times using the small database (a) and the large database (b) for the Kwan- Cai, SOM neural networks and the Bayes Classifier
Fig. 18. Vibration sample detection times using the small database (a) and the large database (b) for the Kwan- Cai, SOM neural networks and the Bayes Classifier
Suggestions for real-world applications were made in the prior chapter leading to further optimizations around the strengths and weaknesses of each algorithm.
The three algorithms and most of all the two neural networks have long been used for varied applications showing great robustness and stability. The versions of these applications used in this paper are presented and have been used and tested in their standard form as presented in (Kohonen, 2000, 2002) and (Kwan&Cai, 1994) using as method of verification direct comparison of the outcome of detection and the optimal known value for each vector at a time and incremented into an error counter. The databases were verified to be consistent of their description.
### **6. Acknowledgement**
This work was supported by CNCSIS – UEFISCSU, project number PNII – IDEI code 1693/2008.
### **7. References**
236 Fuzzy Logic – Algorithms, Techniques and Implementations
Kwan- Cai Small SOM 100 Small SOM 400 Small Bayes Small
0,0287
0,0948
0,0297
0,094
0,0052 0,000193 0,0023
0,0056 0,000538 0,0024
0,0022 0,000538 0,0027
0,0046 0,000498 0,0028
(b) for the Kwan- Cai, SOM neural networks and the Bayes Classifier
optimizations around the strengths and weaknesses of each algorithm.
Detection times (vibration samples) [s]
(a)
Kwan- Cai Large SOM 100 Large SOM 400 Large Bayes Large
Detection times (Vibration samples) [s]
(b) Fig. 18. Vibration sample detection times using the small database (a) and the large database
Suggestions for real-world applications were made in the prior chapter leading to further
The three algorithms and most of all the two neural networks have long been used for varied applications showing great robustness and stability. The versions of these applications used in this paper are presented and have been used and tested in their standard form as presented in (Kohonen, 2000, 2002) and (Kwan&Cai, 1994) using as method of verification direct comparison of the outcome of detection and the optimal
(b) for the Kwan- Cai, SOM neural networks and the Bayes Classifier
Detection times (Pressure samples) [s]
(a)
Kwan- Cai Small SOM 100 Small SOM 400 Small Bayes Small
Detection times (Pressure samples) [s]
(b) Fig. 17. Pressure sample detection times using the small database (a) and the large database
> Kwan- Cai Large SOM 100 Large SOM 400 Bayes Large
**12**
*Brasil*
**Fault Diagnostic of Rotating Machines Based**
**on Artificial Intelligence: Case Studies of**
*Centrais Elétricas do Norte do Brasil S/A – ELETROBRAS-ELETRONORTE,*
The efficiency of the maintenance techniques applied in energy generation power plants is improved when expert diagnosis systems are used to analysis information provided by the continuous monitoring systems used in these installations. There are a large number of equipments available in the power plants of the Centrais Elétricas do Norte do Brazil S/A - ELETROBRAS-ELETRONORTE (known as ELETRONORTE). These equipments operate continuously because are indispensable for the correct functioning of the generation and transmission systems of the company. Anomalies in the operation of these devices can be detected with the use of intelligent diagnosis tools which analysis the information of the continuous monitoring systems and, based in a set of qualitative rules, indicate the best
The best maintenance strategy used in each equipment operated by ELETRONORTE should consider factors as: equipments importance for the production process, acquisition cost and failure rate. To accomplish this task, one of the three maintenance techniques more used nowadays is chosen: corrective, preventive or predictive [1]. In the predictive maintenance, an operational report of the equipment's condition is emitted using the information collected by the continuous monitoring system. The formulation of such report is a task divided in the following stages: 1) Anomaly identification that can be occurring in the equipment; 2) Detection of the anomalous component; 3) Evaluation of the severity of the fault; and 4) Estimation of the remaining life time of the equipment. The predictive maintenance policies is an efficient practice to identify problems in hydrogenerators that will increase reliability, decrease maintenance costs, limit service failures and increase the
There is a vast literature on techniques for detection and identification of faults known to the FDI (Fault Detection and Isolation). A possible classification of these techniques that consider the aspects related to the type of information available about the process analysis defines three categories: methods based on quantitative models, methods based on qualitative models or semi-qualitative, and methods based on historical data [2]. The first two categories are
commonly named Model Based Fault Detection and Isolation (MBFDI) [3].
**1. Introduction**
life of the machines.
procedures to avoid the fail of the equipments.
**the Centrais Elétricas do Norte do**
**Brazil S/A – Eletrobras-Eletronorte**
Marcelo Nascimento Moutinho
## **Fault Diagnostic of Rotating Machines Based on Artificial Intelligence: Case Studies of the Centrais Elétricas do Norte do Brazil S/A – Eletrobras-Eletronorte**
Marcelo Nascimento Moutinho *Centrais Elétricas do Norte do Brasil S/A – ELETROBRAS-ELETRONORTE, Brasil*
### **1. Introduction**
238 Fuzzy Logic – Algorithms, Techniques and Implementations
Kohonen, T. (2000).Self-organizing Maps (3rd edition), Ed. Springer, ISBN-13: 978-
Kohonen, T. (2002). The self-organizing map, vol. 78, no. 9., pp. 1464-1480, ISSN: 0018-9219 Kwan, H.K. & Cai, T (1994). A fuzzy neural network and its applications pattern recognition, IEEE Transactions on Fuzzy Systems, vol.2, issue.3, pp. 185-193, ISSN: 1063-6706 Larose, D.T. (2006).Data Mining Methods and Models, Wiley-IEEE Press, ISBN-13: 978-
Lazarescu, D., Radoi, C. & Lazarescu, V. (2004). A Real-Time Knock Detection Algorithm
2004, Bucharest, pp. 65-68., ISBN: 0-7803-8533-0, Bucharest 20-24 June 2004 Li, H. & Karim , G. A. (July 2004). Knock in spark ignition hydrogen engines, vol. 29, issue
Liu, P. & Li, H.X. (2004). Fuzzy neural network theory and application, Publisher: World Scientific Publishing Company, ISBN-13: 978-9812387868, Singapore Lopez-Rubio E. (2010). Probabilistic Self-Organizing Maps for Continuous Data, vol.21,
Midori, Y., Nobuo, K. & Atsunori K. (1999). Engine Knock Detection Using Wavelet Transform, Dynamics & Design Conference, Issue B, , pp. 299-302, Tokio, 1999 Mitchell, T.M. (2010). Machine Learning (3rd edition), Ed. New York: McGraw Hill Higher
Park, S.T. & Jang J. (2004). Engine knock detection based on Wavelet transform, Proceeding
Radoi, A., Lazarescu V., & Florescu A. (2009). Wavelet Analysis To Detect The Knock On Internal Combustion Engines, tome 54, no.3, pp. 301-310, ISSN: 0035-4066 Thomas, J.H., Dubuisson, B. & M.A. Dillies-Peltier (1997). Engine Knock Detection from Vibration Signals Using Pattern Recognition, Mecanica, vol.32, no 5, pp. 431-439 Wang, F.Y.& Liu, D. (2006). Advances in Computational Intelligence: Theory And
Wehenkel, L.A. (1997) Automatic Learning Technique in Power Systems, Kluwer Academic
Wu J.D. & Liu, C.H. (2009). An expert system for fault diagnosis in internal combustion
Yilmaz, S. & Oysal, Y. (2010). Fuzzy Wavelet Neural Network Models for Prediction and
Zhang, Q. & Benveniste A. (1992). Wavelet networks, vol. 3, issue. 6, pp. 889–898, ISSN:
Zhang, H. & Liu, D. (2006). Fuzzy Modeling and Fuzzy Control (Control Engineering) (1st edition) , Ed. Birkhauser Boston, ISBN-13: 978-0817644918, MD USA
KORUS 2004, vol.3, pp. 80-83, ISBN: 0-7803-8383-4, 26 June-3 July 2004 Prokhorov, D.(2008). Computational Intelligence in Automotive Applications (1st Edition),
of the 8th Russian-Korean International Symposiom on Science and Technology –
Applications (1st edition), Ed. World Scientific Publishing Company,. ISBN-13: 978-
engines using wavelet packet transform and neural network, vol. 36, issue 3, pp.
Identification of Dynamical Systems, vol. 21 , issue 10, pp. 1599 – 1609, ISSN:
Based on Fast Wavelet Transform, in International Conference Communications
3540679219, Berlin
0471666561, USA
8, pp. 859-865, ISSN: 0360-3199
issue.10, pp. 1543 - 1554, ISSN: 1045-9227
Education, ISBN 0070428077, Oregon USA
Publishers, ISBN-13: 978-0792380689 , USA
Ed.Springer, ISBN 978-3-540-79256-7
9812567345, Singapore
4278-4286, ISSN: 0957-4174
1045-9227
1045-9227
The efficiency of the maintenance techniques applied in energy generation power plants is improved when expert diagnosis systems are used to analysis information provided by the continuous monitoring systems used in these installations. There are a large number of equipments available in the power plants of the Centrais Elétricas do Norte do Brazil S/A - ELETROBRAS-ELETRONORTE (known as ELETRONORTE). These equipments operate continuously because are indispensable for the correct functioning of the generation and transmission systems of the company. Anomalies in the operation of these devices can be detected with the use of intelligent diagnosis tools which analysis the information of the continuous monitoring systems and, based in a set of qualitative rules, indicate the best procedures to avoid the fail of the equipments.
The best maintenance strategy used in each equipment operated by ELETRONORTE should consider factors as: equipments importance for the production process, acquisition cost and failure rate. To accomplish this task, one of the three maintenance techniques more used nowadays is chosen: corrective, preventive or predictive [1]. In the predictive maintenance, an operational report of the equipment's condition is emitted using the information collected by the continuous monitoring system. The formulation of such report is a task divided in the following stages: 1) Anomaly identification that can be occurring in the equipment; 2) Detection of the anomalous component; 3) Evaluation of the severity of the fault; and 4) Estimation of the remaining life time of the equipment. The predictive maintenance policies is an efficient practice to identify problems in hydrogenerators that will increase reliability, decrease maintenance costs, limit service failures and increase the life of the machines.
There is a vast literature on techniques for detection and identification of faults known to the FDI (Fault Detection and Isolation). A possible classification of these techniques that consider the aspects related to the type of information available about the process analysis defines three categories: methods based on quantitative models, methods based on qualitative models or semi-qualitative, and methods based on historical data [2]. The first two categories are commonly named Model Based Fault Detection and Isolation (MBFDI) [3].
Fault Diagnostic of Rotating Machines Based on Artificial Intelligence:
where *A*(*q*-1) and *B*(*q*-1) are as follow:
where *e*(*k*) is a Gaussian white noise;
.., *N*, we obtain, in matrix notation:
where. ˆ () () () *<sup>T</sup>*
*k y k k*
1 Represented as follows <sup>1</sup> *q yk yk* ( ) ( 1)
model parameters. The vectors
ˆ
Case Studies of the Centrais Elétricas do Norte do Brazil S/A – Eletrobras-Eletronorte 241
where *y*(*k*) and *u*(*k*) are, respectively, the values of the output and input signals at the discrete time *k*, an integer multiple of the sampling interval *Ts*, *na* and *nb* are the number of regressors is the output and input signals, respectively, and 1 *d* is the output transport system delay as an integer multiple of the sampling interval. Using the discrete delay
1 1 ( )() ( )() *<sup>d</sup> Aq y t q B q u t* (2)
*a <sup>n</sup> <sup>A</sup> <sup>n</sup> <sup>q</sup> <sup>a</sup> <sup>q</sup> <sup>a</sup> <sup>q</sup> <sup>a</sup> <sup>q</sup>* (3)
*b <sup>n</sup> <sup>B</sup> <sup>n</sup> <sup>q</sup> b b <sup>q</sup> <sup>b</sup> <sup>q</sup> <sup>b</sup> <sup>q</sup>* (4)
(*k*) of Eq. (5). The objective of
(*k*) is the vector of regressors and
(*k*) are represented as follows:
*a b*
*T*
*t aa a bb b* (7)
**<sup>y</sup> <sup>Φ</sup>** (8)
(9)
. The following quadratic performance index must be minimized:
*k yk yk n uk d ut d n* (6)
*n n*
(5)
(*k*) is the vector of
operator, ( <sup>1</sup> *q* )1 , the following polynomial representation of Eq. (1) can be obtained:
1 12
1 12
affected by uncorrelated noise. Thus, the following representation can be obtained:
(*k*) and
*y y*
(*k*), which represents approximately the parameter vector
1 2 ( )1 ... *<sup>a</sup>*
01 2 ( ) ... *<sup>b</sup>*
It is interesting to add stochastic characteristics in the model representing as realistically as possible the nature of the process. This can be done considering that the output signal is
> () ()() () *<sup>T</sup> y k k k ek*
( ) ( 1) ... ( ) ( ) ... ( ) *<sup>T</sup>*
12 01 ( ) ... ... *a b*
The non-recursive least squares method [14] will be used in order to estimate the vector
the method is to minimize the sum of the squares of the prediction error between the estimated model output and the real output of the plant. Substituting in equation (5) *k* = 1,2,
> (1) (1) (1) (2) (2) (2) , ,
*T T*
( ) ( ) ( )
*y N N N*
ˆ **y Φ ε**
*T*
A MBFDI algorithm consists of two components: the residues generator and the process of decision making: the residues generator compares the current values of inputs, outputs or states of the process with the estimated model that describes the normal behavior; the process decision is the logic that converts the residue signal (quantitative knowledge) on a qualitative information (normal operating condition or abnormal). The bases of MBFDI algorithms are described in [3], [4] and [5]. The main difficulty in implementing a MBFDI algorithm lies in the fact that the fidelity of the model affects the sensitivity of the fault detection mechanism and the diagnosis precision. Many real systems are not susceptible to conventional modeling techniques due to: the lack of precise knowledge about the system, the strongly nonlinear behavior, the high degree of uncertainty, or the time-varying characteristics.
Recently, well successfully applications of predictive techniques have been reported. In [6 to 9] are presented intelligent systems for predictive maintenance addressed to the diagnosis in real-time of industrial processes. In [10] a fault detection and isolation scheme of sensor and actuator is presented. The project considers multivariate dynamic systems with uncertainties in the mathematical model of the process. Detailed studies on the robustness of anomalous systems of identification in presence of modeling errors is also reported in the survival paper [2].
Nowadays, the expert diagnosis technologies available in the market are in maturation process. The tools commercially available have restrictions in the information exchange with the company's legacy systems. The users normally can't change the software structure and don't know the conceptual data base model. Due to these limitations, the company who uses this kind of paradigm is in a difficult situation when software modifications, not considered in the initial project, are necessary to adjust it to a specific application.
In this chapter is described the procedures for designing and test MBFDI system. Two types of models will be used: autoregressive models and fuzzy models. The proposed system is evaluated experimentally using real monitoring data from a synchronous compensator and a synchronous generator. The synchronous compensator analyzed is in operation at Vila do Conde substation, located at Pará state, Brazil. The synchronous generator studied is in operation at Tucuruí Hydroelectric, located at Pará state too. Both equipments are operated by ELETRONORTE.
### **2. Fuzzy system and regression models for use in diagnosis systems**
To design the fault detection system proposed in this work mathematical models are used to describe the relationships between the variables monitored in the equipment analyzed. Two types of models will be used: autoregressive models and fuzzy models. The purpose of this section is describe the two structures used.
### **2.1 System identification with regression models**
The following structure, known in literature as Autoregressive model with exogenous inputs (ARX), will be used [14]:
$$\begin{aligned} y(t) + a\_1 y(k-1) + \dots + a\_{n\_a} y(k - n\_a) &= \\ b\_0 \mu(k - d) + b\_1 \mu(k - 1 - d) + \dots + b\_{n\_b} \mu(k - d - n\_b) \end{aligned} \tag{1}$$
where *y*(*k*) and *u*(*k*) are, respectively, the values of the output and input signals at the discrete time *k*, an integer multiple of the sampling interval *Ts*, *na* and *nb* are the number of regressors is the output and input signals, respectively, and 1 *d* is the output transport system delay as an integer multiple of the sampling interval. Using the discrete delay operator, ( <sup>1</sup> *q* )1 , the following polynomial representation of Eq. (1) can be obtained:
$$A(q^{-1})y(t) = q^{-d}B(q^{-1})u(t)\tag{2}$$
where *A*(*q*-1) and *B*(*q*-1) are as follow:
240 Fuzzy Logic – Algorithms, Techniques and Implementations
A MBFDI algorithm consists of two components: the residues generator and the process of decision making: the residues generator compares the current values of inputs, outputs or states of the process with the estimated model that describes the normal behavior; the process decision is the logic that converts the residue signal (quantitative knowledge) on a qualitative information (normal operating condition or abnormal). The bases of MBFDI algorithms are described in [3], [4] and [5]. The main difficulty in implementing a MBFDI algorithm lies in the fact that the fidelity of the model affects the sensitivity of the fault detection mechanism and the diagnosis precision. Many real systems are not susceptible to conventional modeling techniques due to: the lack of precise knowledge about the system, the strongly nonlinear
Recently, well successfully applications of predictive techniques have been reported. In [6 to 9] are presented intelligent systems for predictive maintenance addressed to the diagnosis in real-time of industrial processes. In [10] a fault detection and isolation scheme of sensor and actuator is presented. The project considers multivariate dynamic systems with uncertainties in the mathematical model of the process. Detailed studies on the robustness of anomalous systems of identification in presence of modeling errors is also reported in the
Nowadays, the expert diagnosis technologies available in the market are in maturation process. The tools commercially available have restrictions in the information exchange with the company's legacy systems. The users normally can't change the software structure and don't know the conceptual data base model. Due to these limitations, the company who uses this kind of paradigm is in a difficult situation when software modifications, not considered
In this chapter is described the procedures for designing and test MBFDI system. Two types of models will be used: autoregressive models and fuzzy models. The proposed system is evaluated experimentally using real monitoring data from a synchronous compensator and a synchronous generator. The synchronous compensator analyzed is in operation at Vila do Conde substation, located at Pará state, Brazil. The synchronous generator studied is in operation at Tucuruí Hydroelectric, located at Pará state too. Both equipments are operated
To design the fault detection system proposed in this work mathematical models are used to describe the relationships between the variables monitored in the equipment analyzed. Two types of models will be used: autoregressive models and fuzzy models. The purpose of this
The following structure, known in literature as Autoregressive model with exogenous
( ) ( 1 ) ... ( ) *a*
*b n a*
*n b*
(1)
( ) ( 1) ... ( )
*yt ayk a yk n buk d buk d b uk d n*
**2. Fuzzy system and regression models for use in diagnosis systems**
behavior, the high degree of uncertainty, or the time-varying characteristics.
in the initial project, are necessary to adjust it to a specific application.
survival paper [2].
by ELETRONORTE.
section is describe the two structures used.
inputs (ARX), will be used [14]:
**2.1 System identification with regression models**
1
0 1
$$A(q^{-1}) = 1 + a\_1 q^{-1} + a\_2 q^{-2} + \dots + a\_{n\_a} q^{-n\_a} \tag{3}$$
$$B(q^{-1}) = b\_0 + b\_1 q^{-1} + b\_2 q^{-2} + \dots + b\_{n\_b} q^{-n\_b} \tag{4}$$
It is interesting to add stochastic characteristics in the model representing as realistically as possible the nature of the process. This can be done considering that the output signal is affected by uncorrelated noise. Thus, the following representation can be obtained:
$$
\phi(k) = \phi^T(k)\theta(k) + c(k) \tag{5}
$$
where *e*(*k*) is a Gaussian white noise; (*k*) is the vector of regressors and (*k*) is the vector of model parameters. The vectors (*k*) and (*k*) are represented as follows:
$$\phi(k) = \begin{bmatrix} -y(k-1) \ \dots - y(k-n\_a) \ u(k-d) \ \dots \ u(t-d-n\_b) \ \end{bmatrix}^T \tag{6}$$
$$\boldsymbol{\Theta}(t) = \begin{bmatrix} a\_1 \ a\_2 \dots \ a\_{n\_a} b\_0 \ b\_1 \dots \ b\_{n\_b} \end{bmatrix}^T \tag{7}$$
The non-recursive least squares method [14] will be used in order to estimate the vector ˆ (*k*), which represents approximately the parameter vector (*k*) of Eq. (5). The objective of the method is to minimize the sum of the squares of the prediction error between the estimated model output and the real output of the plant. Substituting in equation (5) *k* = 1,2, .., *N*, we obtain, in matrix notation:
$$\mathbf{y} = \begin{bmatrix} y(1) \\ y(2) \\ \vdots \\ y(N) \end{bmatrix}, \mathbf{O} = \begin{bmatrix} \phi(1)^T \\ \phi(2)^T \\ \vdots \\ \phi(N)^T \end{bmatrix}, \mathbf{c} = \begin{bmatrix} \xi(1) \\ \xi(2) \\ \vdots \\ \xi(N) \end{bmatrix} \tag{8}$$
$$\mathbf{y} = \mathbf{O} \, \hat{\boldsymbol{\theta}} + \mathbf{c} \tag{9}$$
where. ˆ () () () *<sup>T</sup> k y k k* . The following quadratic performance index must be minimized:
<sup>1</sup> Represented as follows <sup>1</sup> *q yk yk* ( ) ( 1)
Fault Diagnostic of Rotating Machines Based on Artificial Intelligence:
*Vi*(*k*).
each of the *M* rules:
as follows:
**Systems (ANFIS)**
Case Studies of the Centrais Elétricas do Norte do Brazil S/A – Eletrobras-Eletronorte 243
Fig. 2. Membership functions of the *ni* fuzzy sets associated with the Linguistic variable
degree of precision, dynamic systems governed by nonlinear relationships.
*y k*
real data from a continuous monitoring system described in the next section.
**2.3.1 Synchronous compensator monitoring system – VIBROCOMP**
**2.3 Prediction techniques based on Adaptive-Network-based Fuzzy Inference**
This monitoring system was designed as predictive maintenance tool for Synchronous Compensators (SC). These equipments are large rotary machines of 150 MVAr where the constant evaluation of its physical parameters is critical. In the State of Pará, Eletronorte operates three SC that are part of its transmission system: two are installed in Vila do Conde substation, located in Para State, and one is installed in the Marabá substation. These three
The set of M rules defined by (14) describe a fuzzy system of Sugeno [16], a mathematical tool that can represent globally and approximately the system described in Eq (12). It is a universal nonlinear approximator, a mathematical function that represents with an arbitrary
According to the theory of fuzzy systems [16], the output signal *y*( ) *<sup>k</sup>* of the fuzzy system defined by the set of rules (14) is obtained by weighted average of the individual outputs of
1
( ) ˆ( ) *M l l l M l l*
The weights, ω*l*, are called Functions Validation. They are calculated in terms of the vector
According to Eq.(16), the value of the signal output of the model is a function of the ω*<sup>l</sup>* weights and functions *fl*(.). Therefore, for a given set of values of signal *y*(*k*) can be found an optimal setting of the parameters of fuzzy membership functions defined on each input and the parameters of the functions *fl*(.) that minimizes the difference *y*() () *k yk* for the entire set. The details of the procedure for identification of these parameters will be the subject matter of section IV which will be described a procedure for identifying models based on
1
1, 2 , , 1 2 *ll l* ( ( )) ( ( )) ( ( )) *<sup>i</sup> <sup>j</sup> k p l k VV V*
*Vt Vt Vt* (17)
*y k*
(16)
$$J(\hat{\theta}) = \frac{1}{2} \sum\_{k=1}^{N} \xi(t)^2 \tag{10}$$
The value of ˆthat minimizes the Eq. (10) is [15]:
$$\boldsymbol{\theta}\_{\boldsymbol{M}\boldsymbol{Q}} = (\boldsymbol{\Phi}^{\boldsymbol{T}}\boldsymbol{\Phi})^{-1}\boldsymbol{\Phi}^{\boldsymbol{T}}\boldsymbol{\mathbf{y}} \tag{11}$$
### **2.2 Identification of predictive models based on fuzzy logic**
In this subsection the structure of the fuzzy model used in this work will be described. The following discrete nonlinear system representation is used:
$$y(k) = f\left[\Psi(k-1)\right] \tag{12}$$
where *f*(.) is a nonlinear function of the *Information Vector* , defined as:
$$\Psi(k-1) = \left[ y(k-1)\dotsm y(k-n\_a)\dotsm u(k-d)\dots u(k-d-n\_b) \right]^\Gamma \tag{13}$$
where: *na* and *nb* represent the number of regressors of discrete output signals, *y*(*k*) and input *u*(*k*), respectively, *d* is the output transport delay as an integer multiple of the sampling interval *Ts*; *e*(k) is a random signal that supposedly corrupts the signals of the model is designed in a stochastic environment. This model is known as Non-linear Autoregressive with exogenous inputs (NARX).
We consider the existence of a measurable set of variables that characterize the operating conditions of the system (12) at every moment. Using these variables, you can define a set of rules that describe, approximately, the behavior of the function *y*(*k*):
$$\mathbf{R}^{0\text{-}} \mathbf{I} \mathbf{F} < V\_1 \text{ is } V\_{1,l}^l > \mathbf{A} \mathbf{N} \mathbf{D} < V\_2 \text{ is } V\_{2,j}^l > \mathbf{A} \mathbf{N} \mathbf{D} \dots \mathbf{A} \mathbf{N} \mathbf{D} \\ \le V\_k \text{ is } V\_{k,\eta}^l > \mathbf{T} \mathbf{H} \mathbf{E} \mathbf{N} \ y\_l(k) = f\_l(k) \qquad (14)$$
where: *l*=1,2, ..., *M*; *i* = 1,2, ..., *n*1; *j* = 1,2, ..., *n*2; and *p* = 1,2,..., *nk*. The terms, *V*1, *V*2, ..., *V*k are fuzzy linguistic variables that are part of the vector and were chosen to describe the system (12). The domain of these variables is uniformly partitioned into *ni* = *n*1, *n*2, ..., *nk* fuzzy sets (for example, the partitions of *Vi* are: *Vi*,1, *Vi*,2, ..., *Vi*,*ni*. In this work the function *fl*(.) is represented by the following linear combination:
$$f\_l(k) = c\_l^0 + c\_l^1 V\_1 + c\_l^2 V\_2 + \dots + c\_l^k V\_k \tag{15}$$
onde *<sup>i</sup> <sup>l</sup> c i*=1,2,...,*k* are coefficients to be estimated.
At a given instant of discrete time *k* each linguistic variable, *Vi*, will have a membership value , [ ( )] *V i i j V t* associated with the fuzzy set *j* (*j* = 1,2, ..., *ni*). For mathematical simplicity, the membership functions used to represent these sets are triangular and trapezoidal, with the trapezoidal used only in two extreme sets, as shown in Figure 2. It is easy to see that, for each fuzzy variable, at most two and at least one fuzzy set has a membership value different from zero and the sum of these values always is equal one.
242 Fuzzy Logic – Algorithms, Techniques and Implementations
1 <sup>1</sup> <sup>ˆ</sup> ( ) () <sup>2</sup> *N*
<sup>1</sup> ( ) *T T*
In this subsection the structure of the fuzzy model used in this work will be described. The
( 1) ( 1) ( ) ( ) ( ) *<sup>T</sup>*
where: *na* and *nb* represent the number of regressors of discrete output signals, *y*(*k*) and input *u*(*k*), respectively, *d* is the output transport delay as an integer multiple of the sampling interval *Ts*; *e*(k) is a random signal that supposedly corrupts the signals of the model is designed in a stochastic environment. This model is known as Non-linear Auto-
We consider the existence of a measurable set of variables that characterize the operating conditions of the system (12) at every moment. Using these variables, you can define a set of
where: *l*=1,2, ..., *M*; *i* = 1,2, ..., *n*1; *j* = 1,2, ..., *n*2; and *p* = 1,2,..., *nk*. The terms, *V*1, *V*2, ..., *V*k are fuzzy linguistic variables that are part of the vector and were chosen to describe the system (12). The domain of these variables is uniformly partitioned into *ni* = *n*1, *n*2, ..., *nk* fuzzy sets (for example, the partitions of *Vi* are: *Vi*,1, *Vi*,2, ..., *Vi*,*ni*. In this work the function *fl*(.)
> 01 2 1 2 ( ) *<sup>k</sup>*
At a given instant of discrete time *k* each linguistic variable, *Vi*, will have a membership
*V t* associated with the fuzzy set *j* (*j* = 1,2, ..., *ni*). For mathematical simplicity, the membership functions used to represent these sets are triangular and trapezoidal, with the trapezoidal used only in two extreme sets, as shown in Figure 2. It is easy to see that, for each fuzzy variable, at most two and at least one fuzzy set has a membership value different
*<sup>l</sup> V <sup>j</sup>* > **AND** ... **AND** < *Vk* is ,
*l ll l l k f k c cV cV cV* (15)
*a b k yk yk n uk d uk d n* (13)
*k J t*
that minimizes the Eq. (10) is [15]:
**2.2 Identification of predictive models based on fuzzy logic**
following discrete nonlinear system representation is used:
regressive with exogenous inputs (NARX).
R(l): **IF** < *V*1 is 1,
onde *<sup>i</sup>*
value , [ ( )] *V i i j*
*MQ*
where *f*(.) is a nonlinear function of the *Information Vector* , defined as:
rules that describe, approximately, the behavior of the function *y*(*k*):
*<sup>l</sup> V <sup>i</sup>* > **AND** < *V*2 is 2,
is represented by the following linear combination:
*<sup>l</sup> c i*=1,2,...,*k* are coefficients to be estimated.
from zero and the sum of these values always is equal one.
The value of ˆ
2
(10)
**ΦΦ Φ y** (11)
*yk f k* ( ) ( 1) (12)
*<sup>l</sup> Vk <sup>p</sup>* >**THEN** () () *l l y k fk* (14)
Fig. 2. Membership functions of the *ni* fuzzy sets associated with the Linguistic variable *Vi*(*k*).
The set of M rules defined by (14) describe a fuzzy system of Sugeno [16], a mathematical tool that can represent globally and approximately the system described in Eq (12). It is a universal nonlinear approximator, a mathematical function that represents with an arbitrary degree of precision, dynamic systems governed by nonlinear relationships.
According to the theory of fuzzy systems [16], the output signal *y*( ) *<sup>k</sup>* of the fuzzy system defined by the set of rules (14) is obtained by weighted average of the individual outputs of each of the *M* rules:
$$\hat{\boldsymbol{y}}(k) = \frac{\sum\_{l=1}^{M} \alpha\_l \boldsymbol{y}\_l(k)}{\sum\_{l=1}^{M} \alpha\_l} \tag{16}$$
The weights, ω*l*, are called Functions Validation. They are calculated in terms of the vector as follows:
$$
\mu\_l = \mu\_{V\_{1,i}^l}(V\_1(t)) \times \mu\_{V\_{2,j}^l}(V\_2(t)) \times \cdots \times \mu\_{V\_{k,p}^l}(V\_k(t)) \tag{17}
$$
According to Eq.(16), the value of the signal output of the model is a function of the ω*<sup>l</sup>* weights and functions *fl*(.). Therefore, for a given set of values of signal *y*(*k*) can be found an optimal setting of the parameters of fuzzy membership functions defined on each input and the parameters of the functions *fl*(.) that minimizes the difference *y*() () *k yk* for the entire set. The details of the procedure for identification of these parameters will be the subject matter of section IV which will be described a procedure for identifying models based on real data from a continuous monitoring system described in the next section.
### **2.3 Prediction techniques based on Adaptive-Network-based Fuzzy Inference Systems (ANFIS)**
### **2.3.1 Synchronous compensator monitoring system – VIBROCOMP**
This monitoring system was designed as predictive maintenance tool for Synchronous Compensators (SC). These equipments are large rotary machines of 150 MVAr where the constant evaluation of its physical parameters is critical. In the State of Pará, Eletronorte operates three SC that are part of its transmission system: two are installed in Vila do Conde substation, located in Para State, and one is installed in the Marabá substation. These three
Fault Diagnostic of Rotating Machines Based on Artificial Intelligence:
Table 2. Some signals monitored by VibroComp.
Fig. 4. Main Interface of Client Module of VibroComp
VibroComp are presented.
Case Studies of the Centrais Elétricas do Norte do Brazil S/A – Eletrobras-Eletronorte 245
The Data Acquisition Module is a client/server application that uses the TCP/IP protocol to send information to the Client Module and Database Module. The Client Module was developed in order to be the interface between the user and the Acquisition and Database modules. The client can get the waveforms of the measured signals from the acquisition module and also make the trend analysis and event analysis. The Expert Diagnosis System is used to analyze the information stored in the Database Module. The application runs on the client module and provides to the analyst the possibility of each fails of the equipment. To do this is used a Fuzzy Inference Engine. In Figures 4 and 5 some of the interfaces of
In the next section will present the procedure for the identification of predictive models
used in this work. The modeling techniques presented in section II will be used.
**Tag Description Unit Type** *Mlah* Vibr. Bearing Ring - Horizontal μm Vibration *Mlaa* Vibr. Bearing Ring - Axial μm Vibration *Mlav* Vibr. Bearing Ring - Vertical μm Vibration *Mlbh* Vibr. Pump Bearing - Horizontal μm Vibration *Mlba* Vibr. Pump Bearing - Axial μm Vibration *Mlbv* Vibr. Pump Bearing - Vertical. μm Vibration *Ldh1* Vibr. Left - Horizontal 1 μm Vibration *Leh2* Vibr. Left - Horizontal 2 μm Vibration *Ph2* Pressure of Cooling Hydrogen bar Pressure *Rot* Compensator Speed RPM Speed *P* Active Power MW Power *Q* Reactive Power MVAR Power *Tbea*<sup>87</sup> Temp. stator bars - slot 87 Cº Temperature *Tbea*<sup>96</sup> Temp. stator bars - slot 96 Cº Temperature *Tbea*<sup>105</sup> Temp. stator bars - slot 105 Cº Temperature *Taer* Temp. Cooling Water - Input Cº Temperature *Thsr* Temp. Cooling Hydrogen - Output Cº Temperature *Ther* Temp. Cooling Hydrogen - Input Cº Temperature
equipments are monitored by VibroComp. Figure 3 shows a photograph of CPAV-01, one of SC monitored in the substation of Vila do Conde. This equipment, a member of the National Interconnected System (SIN), is used for voltage regulation. The main features of CPAV-01 are presented in Table 1.
Fig. 3. Synchronous Compensator 01 of the substation Vila do Conde.
Table 1. Nominal Characteristics of CPAV-01.
The VibroComp system consists of the following parts:
- Sources, sensors, transmitters and signal conditioners;
- Aquisition Computers, Database Computers, data acquisition cards, serial cards, cables, etc..;
- Data Acquisition Module;
- Database Module;
- Expert Diagnosis System;
- Client Module
The signal conditioning hardware and monitoring software were developed at the Centro de Tecnologia da ELETRONORTE, known as the Laboratório Central (LACEN). Further details on the development of VibroComp can be obtained in [17].
To evaluate the operational condition of a SC, mechanical, electrical and thermal properties are monitored. Table 2 shows some of the signs monitored by VibroComp that are used in this work.
Table 2. Some signals monitored by VibroComp.
244 Fuzzy Logic – Algorithms, Techniques and Implementations
equipments are monitored by VibroComp. Figure 3 shows a photograph of CPAV-01, one of SC monitored in the substation of Vila do Conde. This equipment, a member of the National Interconnected System (SIN), is used for voltage regulation. The main features of CPAV-01
Fig. 3. Synchronous Compensator 01 of the substation Vila do Conde.
Table 1. Nominal Characteristics of CPAV-01.
1. *Ha*rdw*a*r*e:*
2. *S*of*t*w*are:*
this work.
cables, etc..;
Database Module; Expert Diagnosis System;
Client Module
Data Acquisition Module;
The VibroComp system consists of the following parts:
Sources, sensors, transmitters and signal conditioners;
details on the development of VibroComp can be obtained in [17].
Characteristics Value Power 150 MVAR Speed 900 RPM Voltage 13.8 KV Current 6.275 A Frequency 60Hz
Aquisition Computers, Database Computers, data acquisition cards, serial cards,
The signal conditioning hardware and monitoring software were developed at the Centro de Tecnologia da ELETRONORTE, known as the Laboratório Central (LACEN). Further
To evaluate the operational condition of a SC, mechanical, electrical and thermal properties are monitored. Table 2 shows some of the signs monitored by VibroComp that are used in
are presented in Table 1.
The Data Acquisition Module is a client/server application that uses the TCP/IP protocol to send information to the Client Module and Database Module. The Client Module was developed in order to be the interface between the user and the Acquisition and Database modules. The client can get the waveforms of the measured signals from the acquisition module and also make the trend analysis and event analysis. The Expert Diagnosis System is used to analyze the information stored in the Database Module. The application runs on the client module and provides to the analyst the possibility of each fails of the equipment. To do this is used a Fuzzy Inference Engine. In Figures 4 and 5 some of the interfaces of VibroComp are presented.
In the next section will present the procedure for the identification of predictive models used in this work. The modeling techniques presented in section II will be used.
Fig. 4. Main Interface of Client Module of VibroComp
Fault Diagnostic of Rotating Machines Based on Artificial Intelligence:
Fig. 6. Profile Auto-correlation function of the signal *Tbea87*(*k*).
*r*
where *u* is the average value of the signal *u*(k).
of the auto-correlation signal *Tbea87*(*k*).
following formulation was used:
the signs *Tbea87*(*k*), *Ph2*(*k*) and *Q*(*k*).
signal *Tbea87*(*k*).
Case Studies of the Centrais Elétricas do Norte do Brazil S/A – Eletrobras-Eletronorte 247
regressive Moving Average (ARMA) [18]. The profile of the ACF shows signs of a fixed pattern for the first time delays followed by a pattern composed of combinations of exponential and damped sinusoidal functions. In Figure 6, for example, is shown the profile
The CCF is used to assess correlations between two discrete signals *u*(*k*) and *y*(*k*). The
Delays
1
In the Figure 7 is presented the profile of the CCF between the signal *Taer*(*k*) and the signal *Tbea87*(*k*). The analysis of the CCF of the *Taer*(*k*) indicates that this signal is more correlated to
*k*
2
(19)
[ ( ) ][ ( ) ] [ () ]
*yt y*
*y k y uk u*
1
Fig. 7. Profile of the cross-correlation function between the signal *Taer*(*k*) and the
Delays
*N k yu N*
Fig. 5. Interface of the Expert Diagnosis System of VibroComp
### **3. Case studies 1: System modeling of a synchronous compensator**
### **3.1 Synchronous compensator predictive models**
In this section we present a case study where we identified the parameters of mathematical models that describe, approximately, the behavior of a SC operating in a normal condition.
The equipment CPAV-01, located in the Vila do Conde substation was examined. The models proposed in this work were estimated and validated with real data from the VibroComp monitoring system. The analyzed period was 03/01/2008 to 25/03/2008. The SC was under normal conditions without showing any anomaly.
The identification of mathematical models to describe the behavior of CPAV-01 in this period is a practical procedure that can be divided into the following steps:
The objective of the first step is to identify the correlations that exist in the monitored signals. In this work two mathematical functions are used: Autocorrelation Function (ACF) and Cross-Correlation Function (CCF):
The ACF was used to identify correlations in time of a discrete signal *y*(*k*). The formulation used is as follows:
$$r\_{\tau} = \frac{\sum\_{k=\tau+1}^{N} [y(k) - \overline{y}][y(k-\tau) - \overline{y}]}{\sum\_{k=1}^{N} [y(k) - \overline{y}]^2} \tag{18}$$
where: *y* is the average value of the signal *y*(*k*) and *k* is the discrete time, an integer multiple of the sampling interval, *Ts* .
The ACF analysis revealed that some of the monitored signals (*P*(*k*) and *Rot*(*k*)) behaves approximately like random and uncorrelated white noise. Other signs (*Taer*(*k*), *Ph2*(*k*), *Mlah*(*k*), *P*(*k*), *Tbea87*(*k*) and *Ldh1*(*k*)) are auto-correlated and can be characterized by models with Auto246 Fuzzy Logic – Algorithms, Techniques and Implementations
Fig. 5. Interface of the Expert Diagnosis System of VibroComp
SC was under normal conditions without showing any anomaly.
period is a practical procedure that can be divided into the following steps:
1. Statistic analysis of the monitored signals to identify dynamic relationships;
1
*N k*
*r*
**3.1 Synchronous compensator predictive models**
2. Choose the structure of the models; 3. Models Estimation and validation;
and Cross-Correlation Function (CCF):
multiple of the sampling interval, *Ts* .
used is as follows:
**3. Case studies 1: System modeling of a synchronous compensator**
In this section we present a case study where we identified the parameters of mathematical models that describe, approximately, the behavior of a SC operating in a normal condition. The equipment CPAV-01, located in the Vila do Conde substation was examined. The models proposed in this work were estimated and validated with real data from the VibroComp monitoring system. The analyzed period was 03/01/2008 to 25/03/2008. The
The identification of mathematical models to describe the behavior of CPAV-01 in this
The objective of the first step is to identify the correlations that exist in the monitored signals. In this work two mathematical functions are used: Autocorrelation Function (ACF)
The ACF was used to identify correlations in time of a discrete signal *y*(*k*). The formulation
1
where: *y* is the average value of the signal *y*(*k*) and *k* is the discrete time, an integer
The ACF analysis revealed that some of the monitored signals (*P*(*k*) and *Rot*(*k*)) behaves approximately like random and uncorrelated white noise. Other signs (*Taer*(*k*), *Ph2*(*k*), *Mlah*(*k*), *P*(*k*), *Tbea87*(*k*) and *Ldh1*(*k*)) are auto-correlated and can be characterized by models with Auto-
*N k*
2
(18)
[ ( ) ][ ( ) ] [() ]
*y k y y k y*
*yk y*
regressive Moving Average (ARMA) [18]. The profile of the ACF shows signs of a fixed pattern for the first time delays followed by a pattern composed of combinations of exponential and damped sinusoidal functions. In Figure 6, for example, is shown the profile of the auto-correlation signal *Tbea87*(*k*).
Fig. 6. Profile Auto-correlation function of the signal *Tbea87*(*k*).
The CCF is used to assess correlations between two discrete signals *u*(*k*) and *y*(*k*). The following formulation was used:
$$r\_{yu} = \frac{\sum\_{k=\tau+1}^{N} [\underline{\boldsymbol{y}}(k) - \overline{\boldsymbol{y}}] [\boldsymbol{u}(k-\tau) - \overline{\boldsymbol{u}}]}{\sum\_{k=1}^{N} [\underline{\boldsymbol{y}}(t) - \overline{\boldsymbol{y}}]^2} \tag{19}$$
where *u* is the average value of the signal *u*(k).
In the Figure 7 is presented the profile of the CCF between the signal *Taer*(*k*) and the signal *Tbea87*(*k*). The analysis of the CCF of the *Taer*(*k*) indicates that this signal is more correlated to the signs *Tbea87*(*k*), *Ph2*(*k*) and *Q*(*k*).
Fig. 7. Profile of the cross-correlation function between the signal *Taer*(*k*) and the signal *Tbea87*(*k*).
Fault Diagnostic of Rotating Machines Based on Artificial Intelligence:
( )1
rules.
( ) ( ) ( )
Fuzzy Model Topology 1: MFT1
Fuzzy Model Topology 2: MFT2
Fuzzy Model Topology 3: MFT3
Table 4. Structure of Fuzzy Models for Signal *Taer*(*k*)
Case Studies of the Centrais Elétricas do Norte do Brazil S/A – Eletrobras-Eletronorte 249
1 1
<sup>2</sup>
The third model proposed model uses a fuzzy inference system to represent the signal *Taer*(*k*). Table 4 presents details of the two topologies used. All models are Sugeno fuzzy systems with weighted average defuzzifier and number of outputs equal to the number of
Inputs Sets Function Parameters
*Taer*, *Tbea*<sup>87</sup> 2-3 *Bell* 50 ou 135
*Taer*, *Tbea*87, *Ph2* 2-3 *Gaussiana* 44 ou 96
The nomenclature used to identify the models is as follows: *MFT1* represents the Fuzzy Model Topology 1. The interpretation of other fields in the Table 4 is as follows: in each model are specified the inputs, the number of sets on each input and the type of membership function. The model *MFT1*, for example, uses two inputs with two or three Bell fuzzy sets in each input. The number of parameters in the model is 50 or 135, depending on
<sup>1</sup> (,,,)
1 *Bell <sup>b</sup> f xabc*
<sup>2</sup> (, ,)
The parameter estimation was performed in the MATLAB environment. To estimate the models of the linear equations (20) and (21) we used the System Identification Toolbox [19]. The estimation method used was the non-recursive least squares. The mass of data was divided into two parts: the first was used for the estimation of the model and the second part was used for validation. Figure 8 shows the time domain validation of the model of
*Gauss f xce*
2
2
(22)
(23)
*x c a*
*x c*
*Taer*, *Tbea*<sup>87</sup> 2 *Gaussian* 212
the chosen combination. The Bell and Gaussian function used are as follows.
*T*
(21)
2*q*
2 87 1 1 23 4 1 23 4
*h bea*
( )() ( ) () () () () () () () ()
*Aq yk Bq Uk k yk T k Uk P k T k Qk A q a q a q a q a q Bq B q B q B q Bq b bq bq Bq b bq bq Bq b bq b*
*aer*
1 111 012 1 12 0 00 01 02 1 12 1 10 11 12 1 1 2 20 21 2
() () () ()
A similar analysis realized for the signal *Taer*(*k*) was performed for all other signs in Table 2. The final result of the statistic analysis is presented in Table 3. The interpretation of this table is as follows: the signals in the left column are related to the central column signals at delays intervals specified in the right column. For example, the signal *Taer*(*k*), is selfcorrelated and is also related with the signs *Tbea87*(*k*), *Tbea87*(*k*-1), *Ph*2(*k*-1), *Ph*2(*k*-2), *Ph*2(*k*-3) and so on. The order of presentation of the signs in the center column is proportional to the intensity of the relationship with the signals in the left column.
Table 3. Correlations of Signals Monitored by VibroComp.
The choice of the model structure, the goal of the second step of the identification procedure was based on information in Table 3. The statistical characteristics of the signals indicate that Auto-regressive Moving Average with Exogenous Input (ARMAX) models are good alternatives to explain the dynamic relationships of the monitored signals. However, it is suspected that there are nonlinear relationships between the monitored signals. These relationships are better described by a universal nonlinear approximate operator. For comparison purposes in this paper will be use three types of mathematical models: Singleinput single-output (SISO) ARMAX, multi-inpult single-output (MISO) ARX and a MISO Sugeno fuzzy system. For exemplification purposes, details of the procedure for identification of the *Taer*(*k*) model will be presented. The other signs presented in Table 3. can be estimated by a similar procedure.
The first model analyzed for the sign *Taer*(*k*) is the SISO ARMAX with the following structure:
$$T\_{arr}(k) = \sum\_{i=1}^{4} a\_i T\_{arr}(k-i) + \sum\_{i=0}^{1} b\_i T\_{heat37}(k-i) + \varepsilon(k) + c\_1 \varepsilon(k-1) \tag{20}$$
where is an uncorrelated noise that supposedly corrupts the data, since the model is designed in a stochastic environment. For the sake of structural simplicity, only the signal *Tbea87*(*k*) was chosen as the input for this model. As shown in Table 3, this signal has higher values for the CCF with the *Taer*(*k*). Under an intuitive point of view, it is coherent to suppose that the temperature of cooling water is dependent on the temperature values of the stator bars of SC.
A second more complex MISO ARX model was proposed to explain the behavior of the signal *Taer*(*k*). In this case, the other relationships identified in Table 3 were used. The following structure was chosen:
248 Fuzzy Logic – Algorithms, Techniques and Implementations
A similar analysis realized for the signal *Taer*(*k*) was performed for all other signs in Table 2. The final result of the statistic analysis is presented in Table 3. The interpretation of this table is as follows: the signals in the left column are related to the central column signals at delays intervals specified in the right column. For example, the signal *Taer*(*k*), is selfcorrelated and is also related with the signs *Tbea87*(*k*), *Tbea87*(*k*-1), *Ph*2(*k*-1), *Ph*2(*k*-2), *Ph*2(*k*-3) and so on. The order of presentation of the signs in the center column is proportional to the
intensity of the relationship with the signals in the left column.
Table 3. Correlations of Signals Monitored by VibroComp.
be estimated by a similar procedure.
structure:
where
the stator bars of SC.
following structure was chosen:
Tag Correlations Delays
*Ldh1 Mlah*, *Tbea*87 e *Leh2* [1 7], [1 7] e [1 7] *Ph2 Q*, *Tbea*87 e *Taer* [1 3], [1 3] e [1 3] *Q Ph2*, *Tbea*87 e *Thsr* [1 3], [1 2] e [1 3] *Tbea*87 *Tbea105*, *Q*, e *Ther* [1 2], [1 2] e [1 2]
*Taer Taer*, *Tbea*87, *Ph2, Q* [1 4], [0 2], [0 2] e [0 2]
The choice of the model structure, the goal of the second step of the identification procedure was based on information in Table 3. The statistical characteristics of the signals indicate that Auto-regressive Moving Average with Exogenous Input (ARMAX) models are good alternatives to explain the dynamic relationships of the monitored signals. However, it is suspected that there are nonlinear relationships between the monitored signals. These relationships are better described by a universal nonlinear approximate operator. For comparison purposes in this paper will be use three types of mathematical models: Singleinput single-output (SISO) ARMAX, multi-inpult single-output (MISO) ARX and a MISO Sugeno fuzzy system. For exemplification purposes, details of the procedure for identification of the *Taer*(*k*) model will be presented. The other signs presented in Table 3. can
The first model analyzed for the sign *Taer*(*k*) is the SISO ARMAX with the following
<sup>87</sup> <sup>1</sup> 1 0 () ( ) ( ) ( ) ( 1) *aer i aer i bea i i T k aT k i bT k i k c k*
designed in a stochastic environment. For the sake of structural simplicity, only the signal *Tbea87*(*k*) was chosen as the input for this model. As shown in Table 3, this signal has higher values for the CCF with the *Taer*(*k*). Under an intuitive point of view, it is coherent to suppose that the temperature of cooling water is dependent on the temperature values of
A second more complex MISO ARX model was proposed to explain the behavior of the signal *Taer*(*k*). In this case, the other relationships identified in Table 3 were used. The
is an uncorrelated noise that supposedly corrupts the data, since the model is
(20)
4 1
$$\begin{aligned} A(q^{-1})y(k) &= B(q^{-1})LI(k) + \varepsilon(k) \\ y(k) &= T\_{\text{arc}}(k) \end{aligned}$$
$$\begin{aligned} \begin{aligned} \begin{aligned} \left[I(k) = \left[P\_{h2}(k)\int\_{\text{break}} T\_{\text{break}}(k)\right]Q(k)\right]^T\\ A(q^{-1}) &= 1 + a\_1q^{-1} + a\_2q^{-2} + a\_3q^{-3} + a\_4q^{-4} \end{aligned} \\ B(q^{-1}) &= \left[B\_0(q^{-1}) + B\_1(q^{-1}) + B\_2(q^{-1})\right] \end{aligned} \tag{21}$$
$$\begin{aligned} B\_0(q^{-1}) &= b\_{00} + b\_{01}q^{-1} + b\_{02}q^{-2} \\ B\_1(q^{-1}) &= b\_{10} + b\_{11}q^{-1} + b\_{12}q^{-2} \\ B\_2(q^{-1}) &= b\_{20} + b\_{21}q^{-1} + b\_{22}q^{-2} \end{aligned} \tag{32}$$
The third model proposed model uses a fuzzy inference system to represent the signal *Taer*(*k*). Table 4 presents details of the two topologies used. All models are Sugeno fuzzy systems with weighted average defuzzifier and number of outputs equal to the number of rules.
Table 4. Structure of Fuzzy Models for Signal *Taer*(*k*)
The nomenclature used to identify the models is as follows: *MFT1* represents the Fuzzy Model Topology 1. The interpretation of other fields in the Table 4 is as follows: in each model are specified the inputs, the number of sets on each input and the type of membership function. The model *MFT1*, for example, uses two inputs with two or three Bell fuzzy sets in each input. The number of parameters in the model is 50 or 135, depending on the chosen combination. The Bell and Gaussian function used are as follows.
$$f\_{Bell}(\mathbf{x}, a, b, c) = \frac{1}{\mathbf{1} + \left| \frac{\mathbf{x} - c}{a} \right|^{2b}} \tag{22}$$
$$f\_{\text{Gauss}}(\mathbf{x}, \sigma, \mathbf{c}) = e^{\left(\frac{-\mathbf{x} + \mathbf{c}}{\sqrt{2\sigma}}\right)^2} \tag{23}$$
The parameter estimation was performed in the MATLAB environment. To estimate the models of the linear equations (20) and (21) we used the System Identification Toolbox [19]. The estimation method used was the non-recursive least squares. The mass of data was divided into two parts: the first was used for the estimation of the model and the second part was used for validation. Figure 8 shows the time domain validation of the model of
Fault Diagnostic of Rotating Machines Based on Artificial Intelligence:
Table 6. Coefficients of the Linear MISO Model for Signal *Taer*(*k*).
be presented in the next section.
**3.2 Performance evaluation of predictive models**
estimated in Section IV-A. The criteria used are as follows:
Case Studies of the Centrais Elétricas do Norte do Brazil S/A – Eletrobras-Eletronorte 251
Fuzzy models presented in Table 6 were estimated with the algorithm ANFIS (Adaptive-Network-based Fuzzy Inference System) proposed by Jyh-Shing [20] and available on Fuzzy Systems Toolbox of MATLAB, MathWorks (2002). ANFIS is an algorithm for parameter adjustment of Sugeno fuzzy systems based on training data. In Figure 10 presents the results
of the comparison between the output of the model *MFT2* and the real signal *Taer*(*k*).
**Parameters Values Parameters Values** *a1* -0.546560 *b10* 0.624452 *a2* -0.194398 *b11* -0.531587 *a3* -0.031782 *b12* -0.296779 *a4* 0.035349 *b20* -0.024828 *b00* 0.829278 *b21* 0.002835 *b01* -0.450037 *b22* 0.013416 *b02* -0.145693 - -
Fig. 10. Comparison between the output signal of the *MFT2* model and the real signal *Taer*(*k*).
Time(hours)
A similar procedure to that described for the signal *Taer(k)* was performed for all other signals in Table 3. Annex A shows the identified models. The set of models obtained represents the normal behavior of CPAV-01. Comparing the behavior estimated by the standard model with the actual behavior of the equipment is possible to identify the occurrence of malfunctions. The performance of predictive models of the signal *Taer(k)* will
In this section we present the results of performance evaluation of predictive models
Equation (20). The sampling interval used in the model is *Ts* = 1 hour. The model can explain the dynamics of the signal in most of the time interval analyzed. The identified parameters are presented in the Table 5.
Fig. 8. Comparison between the output signal of the SISO Model and the real signal *Taer*(*k*).
Table 5. Coefficients of the Linear SISO Model for Signal *Taer*(*k*).
Figure 9 shows the time domain validation of the MISO model of equation (21). Table 5 shows the values of the estimated coefficients.
Fig. 9. Comparison between the output signal of the MISO Model and the real signal *Taer*(*k*).
250 Fuzzy Logic – Algorithms, Techniques and Implementations
Equation (20). The sampling interval used in the model is *Ts* = 1 hour. The model can explain the dynamics of the signal in most of the time interval analyzed. The identified parameters
Fig. 8. Comparison between the output signal of the SISO Model and the real signal *Taer*(*k*).
Time(hours)
Parameters Values Parameters Values *a1* -1.054051 *b0* 0.624452 *a2* 0.130458 *b1* -0.531587 *a3* 0.036976 *c1* -0.296779 *a4* 0.016368 - -
Figure 9 shows the time domain validation of the MISO model of equation (21).
Fig. 9. Comparison between the output signal of the MISO Model and the real signal *Taer*(*k*). Time(hours)
Table 5. Coefficients of the Linear SISO Model for Signal *Taer*(*k*).
Table 5 shows the values of the estimated coefficients.
are presented in the Table 5.
Fuzzy models presented in Table 6 were estimated with the algorithm ANFIS (Adaptive-Network-based Fuzzy Inference System) proposed by Jyh-Shing [20] and available on Fuzzy Systems Toolbox of MATLAB, MathWorks (2002). ANFIS is an algorithm for parameter adjustment of Sugeno fuzzy systems based on training data. In Figure 10 presents the results of the comparison between the output of the model *MFT2* and the real signal *Taer*(*k*).
Table 6. Coefficients of the Linear MISO Model for Signal *Taer*(*k*).
Fig. 10. Comparison between the output signal of the *MFT2* model and the real signal *Taer*(*k*).
A similar procedure to that described for the signal *Taer(k)* was performed for all other signals in Table 3. Annex A shows the identified models. The set of models obtained represents the normal behavior of CPAV-01. Comparing the behavior estimated by the standard model with the actual behavior of the equipment is possible to identify the occurrence of malfunctions. The performance of predictive models of the signal *Taer(k)* will be presented in the next section.
### **3.2 Performance evaluation of predictive models**
In this section we present the results of performance evaluation of predictive models estimated in Section IV-A. The criteria used are as follows:
Fault Diagnostic of Rotating Machines Based on Artificial Intelligence:
Table 7. Results of Fuzzy Models Training for the Signal *Taer*(*k*).
**compensator**
steps:
**4.1 Project of the fuzzy expert system**
simplifying the inference unit;
What are the faults to be detected?
must be specified for each variable;
**4. Case studies 2: Development of a Fuzzy expert system for a synchronous**
This section describes the project of a Fuzzy Expert System used to faults diagnosis of a SC based on a Mandan fuzzy system [16]. The design methodology is formed by the following
1. Selection of input variables - the choice depends on the quantity and quality of information provided by the monitoring system. The cause and effect relationships involved in the operation of the equipment helps in this selection. A detailed study of the correlation between variables can help eliminate redundancy of information
2. Selection of Output Variables - At this stage the following question must be answered:
3. Selection of Membership Functions – For each input and output, acceptable and not acceptable levels should be determined. In addition, the number of sets and the overlap
Case Studies of the Centrais Elétricas do Norte do Brazil S/A – Eletrobras-Eletronorte 253
**ID Training** *EMQT EMQV MFT1* 2 sets, 50 parameters, 8 rules 1 20 0.7071 1.029 2 150 0.6009 1.231 3 sets, 135 parameters, 27 rules 3 10 0.4936 3.4069 4 50 0.4762 6.1030 *MFT2* 2 sets, 44 parameters, 8 rules 5 10 0.6597 1.0083 6 250 0.6064 0.9085 3 sets, 96 parameters, 27 regras 7 10 0.4692 3.8152 8 20 0.4663 4.1179 *MFT3* 2 sets, 212 parameters, 32 rules 9 10 0.3434 4.4351 10 20 0.3405 4.3070 *ML1* 0 sets, 7 parameters, 0 rules 11 1 1.7053 4.4881 *ML2* 0 sets, 13 parameters, 0 rules 12 1 1.2130 3.9079
The Structural Complexity (SCO) can be evaluated by the total number of adjustable parameters. For the fuzzy models the number of rules and membership sets are also considered.
The Computational Effort for Estimation (CEE) can be measured by the number of training epochs until a good model is estimated. In this work the efficiency of the estimation method is not considered. Therefore, a simplifying assumption will be used to assume that the cost estimation is associated only to the number of training epochs until a certain level of accuracy of the model is achieved.
The quality of a model depends on the value of the Mean Square Error Training (*EMQT*) and the Mean Squared Validation (*EMQV*). In this work the following index will be used:
$$EMQ\_x = \frac{1}{N} \sum\_{k=0}^{N} \left[ \hat{T}\_{\text{aerx}}(k) - T\_{\text{aerx}}(k) \right]^2 \tag{24}$$
where <sup>ˆ</sup> ( ) *T k aerx* is the signal is estimated; ( ) *T k aerx* is the real measured signal, and *<sup>x</sup>* [*T V*] indicates the error is calculated with training or validation data.
In the Table 7 are presented the results of the training of the fuzzy models. In some situations the increase in the number of membership functions results in improved performance during the training but decreased performance in the validation. This observation can be proved for the model *MFT1* comparing lines 1 and 2 with lines 3 and 4 and for the model *MFT2* comparing lines 5 and 6 with rows 7 and 8.
The increase in the number of training epochs can also exert a deleterious effect on the *EMQV*. For the model *MFT1*, this effect is observed comparing lines 1 with 2 and 3 with 4. For the model *MFT2* this increase in *EMQV* is observed comparing the line 7 to line 8. The cause of this behavior is to decrease the generalize ability of the model during the training, phenomenon known as overfitting. The best performance in the training was obtained with the model *MFT3* on line 10 and the best performance in the validation phase was observed in line 6 with the model *MFT2*.
Comparing the model MFT3 with the models MFT1 and MFT2 it's observed that increasing the number of inputs improves performance in training data. However, this relationship was not observed when the validation data are analyzed.
The comparison between the models of Equations (20) and (21) shows that the MISO is beter. In this case the increase in the SCO resulted in better performance.
In all simulations the performance of the Fuzzy model was superior to linear models in the training data. However, when the validation data are considered this relationship is not always true. An example is the comparison between lines 12 and 4 where we observe an increase in the SCO and a degradation of performance in the validation data.
252 Fuzzy Logic – Algorithms, Techniques and Implementations
The Structural Complexity (SCO) can be evaluated by the total number of adjustable parameters. For the fuzzy models the number of rules and membership sets are also
The Computational Effort for Estimation (CEE) can be measured by the number of training epochs until a good model is estimated. In this work the efficiency of the estimation method is not considered. Therefore, a simplifying assumption will be used to assume that the cost estimation is associated only to the number of training epochs until a certain level of
The quality of a model depends on the value of the Mean Square Error Training (*EMQT*) and
<sup>1</sup> <sup>ˆ</sup> () () *<sup>N</sup>*
2
*<sup>N</sup>* (24)
the Mean Squared Validation (*EMQV*). In this work the following index will be used:
indicates the error is calculated with training or validation data.
and for the model *MFT2* comparing lines 5 and 6 with rows 7 and 8.
was not observed when the validation data are analyzed.
beter. In this case the increase in the SCO resulted in better performance.
increase in the SCO and a degradation of performance in the validation data.
0
*<sup>x</sup> aerx aerx <sup>k</sup> EMQ T k T k*
where <sup>ˆ</sup> ( ) *T k aerx* is the signal is estimated; ( ) *T k aerx* is the real measured signal, and *<sup>x</sup>* [*T V*]
In the Table 7 are presented the results of the training of the fuzzy models. In some situations the increase in the number of membership functions results in improved performance during the training but decreased performance in the validation. This observation can be proved for the model *MFT1* comparing lines 1 and 2 with lines 3 and 4
The increase in the number of training epochs can also exert a deleterious effect on the *EMQV*. For the model *MFT1*, this effect is observed comparing lines 1 with 2 and 3 with 4. For the model *MFT2* this increase in *EMQV* is observed comparing the line 7 to line 8. The cause of this behavior is to decrease the generalize ability of the model during the training, phenomenon known as overfitting. The best performance in the training was obtained with the model *MFT3* on line 10 and the best performance in the validation phase was observed
Comparing the model MFT3 with the models MFT1 and MFT2 it's observed that increasing the number of inputs improves performance in training data. However, this relationship
The comparison between the models of Equations (20) and (21) shows that the MISO is
In all simulations the performance of the Fuzzy model was superior to linear models in the training data. However, when the validation data are considered this relationship is not always true. An example is the comparison between lines 12 and 4 where we observe an
Structural Complexity (SCO);
accuracy of the model is achieved.
in line 6 with the model *MFT2*.
Mean Square Error (EMQ);
considered.
Computational Effort for Estimation (CEE);
Table 7. Results of Fuzzy Models Training for the Signal *Taer*(*k*).
### **4. Case studies 2: Development of a Fuzzy expert system for a synchronous compensator**
### **4.1 Project of the fuzzy expert system**
This section describes the project of a Fuzzy Expert System used to faults diagnosis of a SC based on a Mandan fuzzy system [16]. The design methodology is formed by the following steps:
Fault Diagnostic of Rotating Machines Based on Artificial Intelligence:
distribution has great influence on the behavior of the diagnostic system.
Fig. 11. Membership functions for variable F7, misalignment of bearings.
Universe [0
Below is one of the specified rules:
of input variables.
Case Studies of the Centrais Elétricas do Norte do Brazil S/A – Eletrobras-Eletronorte 255
around the values of precision defined by the expert engineers. Triangular functions with no more than 10% of base showed satisfactory results. Figure 11 is shown an example of distribution of membership functions of the output variable F7. It was observed that this
Fifteen rules were defined by the expert engineers, so that the diagnostic system can detect the faults described in Table 8. These rules use only vibration and temperature variables.
**Regra 1:** IF *Mlah* IS Alarme 1 AND *Mlbh* IS Alarme 1 THEN F1 IS 70% AND F2 IS 30% AND F3 IS 20% AND F4 IS 10% AND F5 IS 10% AND F6 IS 10% AND F7 IS 20%. In this and all other rules provided by the experts can be observed another unique feature: the antecedents are short combinations of the monitored signals and the consequents are long combinations of faults. Table 9 shows the characteristics of the membership functions
100] Function
Table 9. Membership Functions of Vibration Inputs: *Mlah*, *Mlav*, *Mlaa*, *Mlbh*, *Mlbv* and *Mlba*.
Changes made in the distribution of the membership functions in the output variables and the choice of the precision of the faults resulted in satisfactory performance. In the validation tests were observed differences of performance related to the defuzzifier used. This is a project choice and the most appropriate defuzzifier depends on the application.
The other operators of the inference unit have not great influence on the performance. The Fuzzy Expert System, in its current state of development, allows the use of the following methods: T-Norm, Min operator; Mandani implication and Maximum aggregation method.
Fuzzy Sets Type Interval Normal Trapezoidal [0 0 30 40] Alarm - 1 Triangular [40 45 50] Alarm - 2 Trapezoidal [50 60 100 100]
In the first stage two approaches have been proposed: the first strategy considers only the global values of the signals monitored by VibroComp as inputs and the second approach uses the spectral information of the vibration signals as inputs. In this paper, only the conventional approach will be used because the data base structure of the Expert System of VibroComp has not using spectrum information. The input signals to be used are: *Mlah*, *Mlav*, *Mlaa*, *Mlba*, *Mlbh* and *Mlbv*. A description of these abbreviations can be found in Table 2.
The output variables are the faults to be detected. For each fault the expert maintenance engineers of the company defined default probability values of the fault. Table 8 shows the outputs of the fuzzy expert system and the probabilities values defined.
Table 8. Outputs of Synchronous Compensator Fuzzy Expert System
The structure defined in the expert system outputs is so peculiar: for each fault are defined the expected possibilities. The table 8 was determined from the experience of the company's maintenance experts. The validation tests of the fuzzy expert system proposed show that this feature can be better used if each fault is described by a finite number of fuzzy sets equal to the number of possibilities provided by the experts. From a practical point of view, this project choice is based on the following argument: defining a finite number of fault possibilities ensures that the diagnostic system will present expected results.
This project choice, however, does not guarantee the accuracy of the diagnosis. The distribution of fuzzy membership sets in the output variables is a very important aspect of the fuzzy expert system. There are significant inconsistencies between the output values of the fuzzy expert system and the expected values when the membership functions uniformly distributed throughout the universe of discourse of the output variables. So uniformly in the distribution the membership functions, which is a common practice in most applications described in the literature [16], did not show satisfactory results for any kind of defuzzifier used. The solution was to specify non-overlapping fuzzy sets, located in a rather narrow 254 Fuzzy Logic – Algorithms, Techniques and Implementations
4. Formulation of Rules – Standard fuzzy IF-THEN rules that considers the normality
In the first stage two approaches have been proposed: the first strategy considers only the global values of the signals monitored by VibroComp as inputs and the second approach uses the spectral information of the vibration signals as inputs. In this paper, only the conventional approach will be used because the data base structure of the Expert System of VibroComp has not using spectrum information. The input signals to be used are: *Mlah*, *Mlav*,
The output variables are the faults to be detected. For each fault the expert maintenance engineers of the company defined default probability values of the fault. Table 8 shows the
> F2 Faulty bearing 10%, 20%, 70%, 90% F3 Rubbing Axis 20%, 30%, 50%, 70%
> F5 Oil Whirl 10%, 20%, 70%, 90% F6 A bent shaft 10%, 30%, 40%, 60%
The structure defined in the expert system outputs is so peculiar: for each fault are defined the expected possibilities. The table 8 was determined from the experience of the company's maintenance experts. The validation tests of the fuzzy expert system proposed show that this feature can be better used if each fault is described by a finite number of fuzzy sets equal to the number of possibilities provided by the experts. From a practical point of view, this project choice is based on the following argument: defining a finite number of fault
This project choice, however, does not guarantee the accuracy of the diagnosis. The distribution of fuzzy membership sets in the output variables is a very important aspect of the fuzzy expert system. There are significant inconsistencies between the output values of the fuzzy expert system and the expected values when the membership functions uniformly distributed throughout the universe of discourse of the output variables. So uniformly in the distribution the membership functions, which is a common practice in most applications described in the literature [16], did not show satisfactory results for any kind of defuzzifier used. The solution was to specify non-overlapping fuzzy sets, located in a rather narrow
10%, 20%, 70%, 90%
10%, 20%, 70%, 90%
10%, 20%, 30%, 70%
*Mlaa*, *Mlba*, *Mlbh* and *Mlbv*. A description of these abbreviations can be found in Table 2.
outputs of the fuzzy expert system and the probabilities values defined.
F1 Mechanical Unbalance
F4 Housing/Support Loose
F7 Misalignment of Bearings
Table 8. Outputs of Synchronous Compensator Fuzzy Expert System
possibilities ensures that the diagnostic system will present expected results.
**ID Fail Values**
5. Selection of Operators – Plausibility and continuity should be used for this selection; 6. Adjust of Rule Base – Simulation using trial and error procedure used to detect inconsistencies in the rule base. Mathematical models of the monitored system also can
conditions;
be used;
around the values of precision defined by the expert engineers. Triangular functions with no more than 10% of base showed satisfactory results. Figure 11 is shown an example of distribution of membership functions of the output variable F7. It was observed that this distribution has great influence on the behavior of the diagnostic system.
Fig. 11. Membership functions for variable F7, misalignment of bearings.
Fifteen rules were defined by the expert engineers, so that the diagnostic system can detect the faults described in Table 8. These rules use only vibration and temperature variables. Below is one of the specified rules:
### **Regra 1:** IF *Mlah* IS Alarme 1 AND *Mlbh* IS Alarme 1 THEN F1 IS 70% AND F2 IS 30% AND F3 IS 20% AND F4 IS 10% AND F5 IS 10% AND F6 IS 10% AND F7 IS 20%.
In this and all other rules provided by the experts can be observed another unique feature: the antecedents are short combinations of the monitored signals and the consequents are long combinations of faults. Table 9 shows the characteristics of the membership functions of input variables.
Table 9. Membership Functions of Vibration Inputs: *Mlah*, *Mlav*, *Mlaa*, *Mlbh*, *Mlbv* and *Mlba*.
Changes made in the distribution of the membership functions in the output variables and the choice of the precision of the faults resulted in satisfactory performance. In the validation tests were observed differences of performance related to the defuzzifier used. This is a project choice and the most appropriate defuzzifier depends on the application.
The other operators of the inference unit have not great influence on the performance. The Fuzzy Expert System, in its current state of development, allows the use of the following methods: T-Norm, Min operator; Mandani implication and Maximum aggregation method.
Fault Diagnostic of Rotating Machines Based on Artificial Intelligence:
MBFDI technique with Fuzzy *MFT2*.
technique with the SISO linear model.
with MISO Linear Model.
Case Studies of the Centrais Elétricas do Norte do Brazil S/A – Eletrobras-Eletronorte 257
Fig. 13. Increase in the stator bars temperatures - Results of the event analysis using the
Time(x1 hours)
Fig. 14. Increase in the stator bars temperatures - Results of the event analysis using MBFDI
Time(x1 hours)
Fig. 15. Increase in the stator bars temperatures - Results of the event analysis using MBFDI
Time(x1 hours)
deviation. At the beginning of the anomalous behavior, the residues signal increases in the three models, which allows a rapid and reliable detection of the failure. In the model the
### **4.2 Experimental evaluation of the fuzzy expert system**
To evaluate the fault detection methodology proposed in this work will be presented a case study where it was possible to detect an anomalous behavior based on the residual analysis of the reference models of the SC.
In the figure 12 is presented the analyzed event which was monitored by VibroComp on 11/03/2008 11:47:56 AM in CPAV-01 at *t* = 1560 hours of operation. In this situation was detected a considerable increase in the stator bars temperatures *Tbea87*(*k*), *Tbea96*(*k*) and *Tbea105*(*k*). We also observed an increase in the value of reactive power that reached the value of *Q*(*k*) = 148,5 MVAR, close to the nominal limit of apparent power of the equipment (150MVA). In Table 10 are presented the recorded values and the normal limits of each monitored signal. Is note the scope of this work explain the causes of the dynamic behavior observed in CPAV-01 based on laws of physics and mechanical models. To establish clearly these cause and effect relationships, mechanical engineering and dynamic vibration knowhow are required, The author of this article does not have this knowhow. The main objective of this work is not to explain. The intention is to describe the mechanical behavior of the studied system and classify this dynamic in patterns or signatures using mathematical models estimated and validated based on real monitoring data. Using these models a fault detection system is projected based on MBFDI techniques.
Fig. 12. Experimental evaluation of fuzzy expert system with case study in CPAV-01. Signals monitored during the event and normal limits.
Table 10. Values monitored by VibroComp during the stator temperature event.
Figures 13 to 15 are presented the results of the analysis of the event using the MBFDI technique presented in this work. *MFT2*, the SISO linear model of Equation (20) and the MIMO linear model of Equation (21) are used as reference models, respectively. In the condition of normality, the residual signal has a mean near zero and a low standard 256 Fuzzy Logic – Algorithms, Techniques and Implementations
To evaluate the fault detection methodology proposed in this work will be presented a case study where it was possible to detect an anomalous behavior based on the residual analysis
In the figure 12 is presented the analyzed event which was monitored by VibroComp on 11/03/2008 11:47:56 AM in CPAV-01 at *t* = 1560 hours of operation. In this situation was detected a considerable increase in the stator bars temperatures *Tbea87*(*k*), *Tbea96*(*k*) and *Tbea105*(*k*). We also observed an increase in the value of reactive power that reached the value of *Q*(*k*) = 148,5 MVAR, close to the nominal limit of apparent power of the equipment (150MVA). In Table 10 are presented the recorded values and the normal limits of each monitored signal. Is note the scope of this work explain the causes of the dynamic behavior observed in CPAV-01 based on laws of physics and mechanical models. To establish clearly these cause and effect relationships, mechanical engineering and dynamic vibration knowhow are required, The author of this article does not have this knowhow. The main objective of this work is not to explain. The intention is to describe the mechanical behavior of the studied system and classify this dynamic in patterns or signatures using mathematical models estimated and validated based on real monitoring data. Using these models a fault
Fig. 12. Experimental evaluation of fuzzy expert system with case study in CPAV-01. Signals
Time(x1 hours)
*Tbea87*(*k*) 82ºC 70º C *Tbea96*(*k*) 82,23ºC 70º C *Tbea105*(*k*) 82,53ºC 70º C *Q*(*k*) 148,5 MVAR 150 MVAR
Figures 13 to 15 are presented the results of the analysis of the event using the MBFDI technique presented in this work. *MFT2*, the SISO linear model of Equation (20) and the MIMO linear model of Equation (21) are used as reference models, respectively. In the condition of normality, the residual signal has a mean near zero and a low standard
Table 10. Values monitored by VibroComp during the stator temperature event.
Tag Value Limit of Normality
**4.2 Experimental evaluation of the fuzzy expert system**
detection system is projected based on MBFDI techniques.
monitored during the event and normal limits.
of the reference models of the SC.
Fig. 13. Increase in the stator bars temperatures - Results of the event analysis using the MBFDI technique with Fuzzy *MFT2*.
Fig. 14. Increase in the stator bars temperatures - Results of the event analysis using MBFDI technique with the SISO linear model.
Fig. 15. Increase in the stator bars temperatures - Results of the event analysis using MBFDI with MISO Linear Model.
deviation. At the beginning of the anomalous behavior, the residues signal increases in the three models, which allows a rapid and reliable detection of the failure. In the model the
Fault Diagnostic of Rotating Machines Based on Artificial Intelligence:
The structure of the MISO linear model for the signal *Tbea87* is as follows:
( )1
1 2 20 21
*Bq b bq*
In the figure 17 is presented the results of the estimation of the MISO model of Eq. (27).
Fig. 17. Comparison between the signal estimated by the linear MISO model and the real
Time(hours)
<sup>87</sup> <sup>1</sup> 1 0 () ( ) ( ) ( ) ( 1) *<sup>i</sup> i bea i i Q k aQ k i bT k i k c k*
In the figure 18 is presented the results of the estimation of the SISO model of Eq. (28).
(28)
The structure of the SISO linear model for the signal *Q*(*k*)is as follows:
7 1
signal *Tbea87*(*k*).
( ) ( ) ( )
1 1
105
Case Studies of the Centrais Elétricas do Norte do Brazil S/A – Eletrobras-Eletronorte 259
*T*
(27)
1 1 23 4 1 23 4
1 2
<sup>22</sup> *b q*
1 111 012 1 12 0 00 01 02 1 12 1 10 11 12
() () () ()
87
( )() ( ) () () () () () () () ()
*Aq yk Bq Uk k yk T k Uk T k T k Qk Aq aq aq aq aq Bq B q B q B q Bq b bq bq Bq b bq bq*
*bea her*
*bea*
increase is greater *MFT2* indicating that this model has a higher sensitivity for the detection of such failures. The two linear models have approximately the same level of residue during the event.
One of the rules used for the residues evaluation is shown:
$$\begin{aligned} \text{IF} \left( |r(k)| > r\_{\text{Tar}} \right) & \text{AND} \left( T\_{\text{bast}\mathcal{T}}(k) \text{Real} \right) \approx LT\\ \text{THEN} & \\ \text{FALLURE} = \text{Statement Temperature out of Range} \\ r(k) &= T\_{ar}(k) \text{Real} \quad \text{[} \quad T\_{ar}(k) \text{Estimated]} \end{aligned} \tag{25}$$
where: *r*(*k*) represents the residual of signal *Taer*(*k*); *rTaer* is the maximum allowed residue in the normal condition, and *LT* represents the thermal limit of the stator winding.
### **5. Conclusions**
In this paper we presented the design and experimental evaluation of a MBFDI system used for SC. The predictive models used in the proposed system were estimated based on real data obtained from the monitoring of the studied equipment during normal conditions. With these models is still possible to detect failures in a fledgling state from the comparison between model output and real signs monitored, as was presented in a case study where it was possible to detect the moment of occurrence of a failure.
### **6. Appendix A: Structure of the models of the signs** *Mlah***,** *Ldh1***,** *Ph2***,** *Q***,** *Tbea87*
The structure of the SISO linear model for the signal *Tbea87* is as follows:
$$T\_{\text{ben87}}(k) = \sum\_{i=1}^{4} a\_i T\_{\text{ben87}}(k-i) + \sum\_{i=0}^{3} b\_i T\_{\text{ben105}}(k-i) + \varepsilon(k) + c\_1 \varepsilon(k-1) \tag{26}$$
In the figure 16 is presented the results of the estimation of the SISO model of Eq. (26).
Fig. 16. Comparison between the signal estimated by the linear SISO model and the real signal *Tbea87*(*k*).
The structure of the MISO linear model for the signal *Tbea87* is as follows:
258 Fuzzy Logic – Algorithms, Techniques and Implementations
increase is greater *MFT2* indicating that this model has a higher sensitivity for the detection of such failures. The two linear models have approximately the same level of residue during
IF ( ( ) *Taer rk r* ) AND ( <sup>87</sup> ( ) *T k bea Real* > *LT*) THEN FAILURE = Stator Temperature out of Range (25) () () *aer rk T k* Real - ( ) *T k aer Estimated*
where: *r*(*k*) represents the residual of signal *Taer*(*k*); *rTaer* is the maximum allowed residue in
In this paper we presented the design and experimental evaluation of a MBFDI system used for SC. The predictive models used in the proposed system were estimated based on real data obtained from the monitoring of the studied equipment during normal conditions. With these models is still possible to detect failures in a fledgling state from the comparison between model output and real signs monitored, as was presented in a case study where it
**6. Appendix A: Structure of the models of the signs** *Mlah***,** *Ldh1***,** *Ph2***,** *Q***,** *Tbea87*
<sup>87</sup> <sup>87</sup> <sup>105</sup> <sup>1</sup> 1 0 ( ) ( ) ( ) ( ) ( 1) *bea i bea i bea i i T k aT k i bT k i k c k*
Fig. 16. Comparison between the signal estimated by the linear SISO model and the real
Time(hours)
In the figure 16 is presented the results of the estimation of the SISO model of Eq. (26).
(26)
the normal condition, and *LT* represents the thermal limit of the stator winding.
One of the rules used for the residues evaluation is shown:
was possible to detect the moment of occurrence of a failure.
The structure of the SISO linear model for the signal *Tbea87* is as follows:
4 3
the event.
**5. Conclusions**
signal *Tbea87*(*k*).
$$\begin{aligned} A(q^{-1})y(k) &= B(q^{-1})\mathcal{U}(k) + \varepsilon(k) \\ y(k) &= T\_{\text{bar}87}(k) \end{aligned}$$
$$\begin{aligned} \mathcal{U}(k) &= \left[T\_{\text{bar}105}(k)\,T\_{\text{bar}}(k)\,\,Q(k)\right]^T \\ A(q^{-1}) &= 1 + a\_1q^{-1} + a\_2q^{-2} + a\_3q^{-3} + a\_4q^{-4} \\ B(q^{-1}) &= \left[B\_0(q^{-1}) + B\_1(q^{-1}) + B\_2(q^{-1})\right] \\ B\_0(q^{-1}) &= b\_{00} + b\_{01}q^{-1} + b\_{02}q^{-2} \\ B\_1(q^{-1}) &= b\_{10} + b\_{11}q^{-1} + b\_{12}q^{-2} \\ B\_2(q^{-1}) &= b\_{20} + b\_{21}q^{-1} + b\_{22}q^{-2} \end{aligned} \tag{27}$$
In the figure 17 is presented the results of the estimation of the MISO model of Eq. (27).
Fig. 17. Comparison between the signal estimated by the linear MISO model and the real signal *Tbea87*(*k*).
The structure of the SISO linear model for the signal *Q*(*k*)is as follows:
$$Q(k) = \sum\_{i=1}^{7} a\_i Q(k-i) + \sum\_{i=0}^{1} b\_i T\_{\text{benS}7}(k-i) + \varepsilon(k) + c\_1 \varepsilon(k-1) \tag{28}$$
In the figure 18 is presented the results of the estimation of the SISO model of Eq. (28).
Fault Diagnostic of Rotating Machines Based on Artificial Intelligence:
signal *Q*(*k*).
signal *Ph2*(*k*).
Case Studies of the Centrais Elétricas do Norte do Brazil S/A – Eletrobras-Eletronorte 261
Fig. 19. Comparison between the signal estimated by the linear MISO model and the real
Time(hours)
2 2 <sup>1</sup> 1 0 ( ) ( ) ( ) ( ) ( 1) *<sup>h</sup> i h i aer i i P k aP k i bT k i k c k*
In the figure 20 is presented the results of the estimation of the SISO model of Eq. (30).
Fig. 20. Comparison between the signal estimated by the linear SISO model and the real
Time(hours)
The structure of the MISO linear model for the signal *Ph2*(*k*)is as follows:
30)
The structure of the SISO linear model for the signal *Ph2*(*k*)is as follows:
5 8
Fig. 18. Comparison between the signal estimated by the linear SISO model and the real signal *Q*(*k*).
The structure of the MISO linear model for the signal *Q*(*k*)is as follows:
$$\begin{aligned} A(q^{-1})y(k) &= B(q^{-1})LI(k) + C(q^{-1})\varepsilon(k) \\ y(k) &= Q(k) \end{aligned}$$
$$\begin{aligned} \mathcal{U}(k) &= \left[T\_{\text{new}\mathcal{G}}(k)\ T\_{\text{arr}}(k)\ \ T\_{\text{lsr}}(k)\right]^T \\ A(q^{-1}) &= 1 + a\_1 q^{-1} \\ B(q^{-1}) &= \left[B\_0(q^{-1}) + B\_1(q^{-1}) + B\_2(q^{-1})\right] \\ B\_0(q^{-1}) &= b\_{00} + b\_{01}q^{-1} + b\_{02}q^{-2} \\ B\_1(q^{-1}) &= b\_{10} + b\_{11}q^{-1} + b\_{12}q^{-2} \\ B\_2(q^{-1}) &= b\_{20} + b\_{21}q^{-1} + b\_{22}q^{-2} \\ \mathcal{C}(q^{-1}) &= 1 + c\_1q^{-1} + c\_2q^{-2} \end{aligned} \tag{29}$$
In the figure 19 is presented the results of the estimation of the MISO model of Eq. (29).
260 Fuzzy Logic – Algorithms, Techniques and Implementations
Fig. 18. Comparison between the signal estimated by the linear SISO model and the real
Time(hours)
87
( )1 () () () ()
1
*C q*
( ) ( ) ( ) ( )
*T*
(29)
1 1 1
1 2 1 2 <sup>1</sup> *cq cq*
*bea aer hsr*
1 111 012 1 12 0 00 01 02 1 12 1 10 11 12 1 12 2 20 21 22
*Bq b bq bq Bq b bq bq Bq b bq bq*
In the figure 19 is presented the results of the estimation of the MISO model of Eq. (29).
11 1
*A q y k B q Uk C q k yk Qk Uk T k T k T k Aq aq B Bq Bq Bq q*
( )() ( ) () ( )() () () () () () ()
The structure of the MISO linear model for the signal *Q*(*k*)is as follows:
signal *Q*(*k*).
Fig. 19. Comparison between the signal estimated by the linear MISO model and the real signal *Q*(*k*).
The structure of the SISO linear model for the signal *Ph2*(*k*)is as follows:
$$P\_{h2}(k) = \sum\_{i=1}^{5} a\_i P\_{h2}(k - i) + \sum\_{i=0}^{8} b\_i T\_{aer}(k - i) + \varepsilon(k) + c\_1 \varepsilon(k - 1) \tag{30}$$
In the figure 20 is presented the results of the estimation of the SISO model of Eq. (30).
Fig. 20. Comparison between the signal estimated by the linear SISO model and the real signal *Ph2*(*k*).
The structure of the MISO linear model for the signal *Ph2*(*k*)is as follows:
Fault Diagnostic of Rotating Machines Based on Artificial Intelligence:
SISO and MISO models Eq. (32) and Eq. (33).
signal *Ldh1*(*k*).
signal *Ldh1*(*k*).
Case Studies of the Centrais Elétricas do Norte do Brazil S/A – Eletrobras-Eletronorte 263
( )() ( ) () () () () () () () ()
*Aq yk Bq Uk k yk L k Uk T k M k L k Aq aq B Bq Bq Bq q*
() () () ()
*B q bq B q bq B q bq*
1 1
( )1
( ) ( ) ( )
*i i*
*i i*
*i i*
*bea lah eh*
In the figures 22 and 23 are presented the results of the time comparison of the estimated
Fig. 22. Comparison between the signal estimated by the linear SISO model and the real
Time(hours)
Fig. 23. Comparison between the signal estimated by the linear MISO model and the real
Time(hours)
*T*
(33)
*i*
*i*
*i*
1 87 2 1 1 1 1 111 012 1 7 0 0 0 1 7 1 1 0 1 7 2 2 0
*dh*
$$\begin{aligned} A(q^{-1})y(k) &= B(q^{-1})LI(k) + \varepsilon(k) \\ y(k) &= P\_{h2}(k) \end{aligned}$$
$$\begin{aligned} LI(k) &= \left[ Q(k) \ T\_{\text{beastS7}}(k) \ T\_{\text{arr}}(k) \right]^T \\ A(q^{-1}) &= 1 + a\_1 q^{-1} \end{aligned}$$
$$\begin{aligned} B(q^{-1}) &= \left[ B\_0(q^{-1}) + B\_1(q^{-1}) + B\_2(q^{-1}) \right] \\ B\_0(q^{-1}) &= b\_{00} + b\_{01}q^{-1} + b\_{02}q^{-2} + b\_{03}q^{-3} \\ B\_1(q^{-1}) &= b\_{10} + b\_{11}q^{-1} + b\_{12}q^{-2} + b\_{13}q^{-3} \\ B\_2(q^{-1}) &= b\_{20} + b\_{21}q^{-1} + b\_{22}q^{-2} + b\_{23}q^{-3} \end{aligned} \tag{31}$$
In the figure 21 is presented the results of the time comparison of the estimated MISO model of Eq. (31).
Fig. 21. Comparison between the signal estimated by the linear MISO model and the real signal *Ph2*(*k*).
The structure of the SISO linear model for the signal *Ldh1*(*k*) is as follows:
$$\begin{split} L\_{dl1}(k) &= \sum\_{i=1}^{2} a\_i L\_{dl1}(k-i) + \sum\_{i=0}^{2} b\_i T\_{bar87}(k-i) + \\ &+ \varepsilon(k) + \sum\_{i=1}^{2} c\_i \varepsilon(k-1) \end{split} \tag{32}$$
The structure of the MISO linear model for the signal *Ldh1*(*k*)is as follows:
262 Fuzzy Logic – Algorithms, Techniques and Implementations
1 1
( )1
*Bq b bq bq*
( ) ( ) ( )
of Eq. (31).
signal *Ph2*(*k*).
<sup>3</sup>
*bea aer*
*T*
(31)
<sup>23</sup> *<sup>b</sup> <sup>q</sup>*
2 87 1 1 1 1 111 012 1 123 0 00 01 02 03 1 123 1 10 11 12 13 1 12 2 20 21 22
*h*
() () () ()
*B q b b q b q b q B q b b q b q b q*
Fig. 21. Comparison between the signal estimated by the linear MISO model and the real
Time(hours)
2 2 1 1 87 1 0 2 1
*L k aL k i bT k i*
( ) ( 1) *dh i dh i bea i i*
*i i*
() ( ) ( )
(32)
*k ck*
The structure of the SISO linear model for the signal *Ldh1*(*k*) is as follows:
The structure of the MISO linear model for the signal *Ldh1*(*k*)is as follows:
In the figure 21 is presented the results of the time comparison of the estimated MISO model
( )() ( ) () () () () () () () ()
*Aq yk Bq Uk k yk P k Uk Qk T k T k Aq aq B Bq Bq Bq q*
$$\begin{aligned} A(q^{-1})y(k) &= B(q^{-1})L(k) + \varepsilon(k) \\ y(k) &= L\_{dh1}(k) \end{aligned}$$
$$\begin{aligned} L(k) &= \begin{bmatrix} T\_{ho87}(k) & M\_{lail}(k) \ L\_{ch2}(k) \end{bmatrix}^T \\ A(q^{-1}) &= 1 + a\_1 q^{-1} \\ B(q^{-1}) &= \begin{bmatrix} B\_0(q^{-1}) + B\_1(q^{-1}) + B\_2(q^{-1}) \end{bmatrix} \\ B\_0(q^{-1}) &= \sum\_{i=0}^7 b\_{0i} q^{-i} \\ B\_1(q^{-1}) &= \sum\_{i=0}^7 b\_{1i} q^{-i} \\ B\_2(q^{-1}) &= \sum\_{i=0}^7 b\_{2i} q^{-i} \end{aligned} \tag{33}$$
In the figures 22 and 23 are presented the results of the time comparison of the estimated SISO and MISO models Eq. (32) and Eq. (33).
Fig. 22. Comparison between the signal estimated by the linear SISO model and the real signal *Ldh1*(*k*).
Fig. 23. Comparison between the signal estimated by the linear MISO model and the real signal *Ldh1*(*k*).
**13**
*Japan*
**Understanding Driver Car-Following Behavior**
Recently, automatic systems that control driving speeds and headway distances while following a vehicle have been developed worldwide. Some products, such as adaptive cruise control systems, have already been installed in upper segments of passenger vehicles. Car following is an important operation in safe and comfortable driving on straight and/or curved roads. The number of traffic accidents involving rear-end collisions is the highest over the last decade in Japan (Iwashita et al., 2011). A rear-end collision occurs when the distance between two vehicles decreases due to deceleration of the lead vehicle and/or higher speed of the following vehicle. The automatic vehicle control system maintains a safe headway distance while following a vehicle and controls velocity according to the relative
If the system's automatic controls do not match the driver's manual controls, driver acceptance of the automatic vehicle control systems decreases, and the driver is not likely to use them. For example, when a lead vehicle speeds up and the inter-vehicle distance increases, one driver may accelerate strongly, whereas another driver may accelerate slightly; and other drivers may not accelerate. The system's automatic hard acceleration does not suit drivers whose acceleration is slight and those who do not accelerate, and they may regard such automatic systems as dangerous. Therefore, it is expected that drivers will accept longitudinal control systems that operate in a manner similar to their own usual carfollowing behavior. Drivers' car-following behavior must be investigated in a real roadtraffic environment to develop vehicle control and driver support systems that are
Car-following behavior consists of two aspects: how much distance drivers allow for a leading vehicle as an acceptable headway distance, and how they control acceleration according to the movements of the leading vehicle. Figure 1 presents an example of a typical following process. This car-following behavior data was recorded using an instrumented vehicle on a real motorway in Southampton (Sato et al., 2009a). The relative distance and speed were detected by microwave radar. The data length was 5min. Car-following behavior is a goal-seeking process depicted by several spirals as drivers attempt to maintain
the desired following headway behind a vehicle in a car-following situation.
speed of the leading vehicle, in order to avoid a rear-end collision.
compatible with drivers' typical car-following behavior.
**1. Introduction**
**Using a Fuzzy Logic Car-Following Model**
Toshihisa Sato and Motoyuki Akamatsu
*National Institute of Advanced Industrial Science*
*Human Technology Research Institute,*
*and Technology (AIST),*
### **7. References**
## **Understanding Driver Car-Following Behavior Using a Fuzzy Logic Car-Following Model**
Toshihisa Sato and Motoyuki Akamatsu
*Human Technology Research Institute, National Institute of Advanced Industrial Science and Technology (AIST), Japan*
### **1. Introduction**
264 Fuzzy Logic – Algorithms, Techniques and Implementations
[1] Nepomuceno, L. X.; *Técnicas de Manutenção Preditiva*; Editora Edgard Blücher Ltda;
[2] Venkat Venkatasubramanian, Raghunathan Rengaswamy, Kewen Yin, Surya N. Kavuri;
[3] Patton, R. J., Frank, P. M., and Clark, R. N.; *Fault Diagnosis in Dynamic Systems, Theory and Application*. Control Engineering Series. Prentice Hall, London 1989. [4] Chen, J. and Patton, R. J. *Robust Model Based Fault Diagnosis for Dynamic Systems*. Kluwer
[5] Basseville, M. and Nikiforov, I. V. *Detection of Abrupt Changes: Theory and Application*.
[6] Mari Cruz Garcia, Miguel A. Sanz Bobi and Javier del Pico *SIMAP - Intelligent System for*
[7] Moutinho, Marcelo N. *Sistema de Análise e Diagnóstico de Equipamentos Elétricos de Potência*
[8] Moutinho, Marcelo N. *Fuzzy Diagnostic Systems of Rotating Machineries, some*
[9] Moutinho, Marcelo N. *Classificação de Padrões Operacionais do Atuador Hidráulico do*
[11] Paul M. Frank, *Fault Diagnosis in Dynamic Systems Using Analytical and Knowledge-based*
[12] Alexandre Carlos Eduardo; *Diagnóstico de Defeitos em Sistemas Mecânicos Rotativos através*
[13] VibroSystM. *Zoom 5 Software Guia do Usuário*. P/N: 9476-26M1A-100. VibroSystM Inc.
[15] Karl J. Aström and Björn Wittenmark, Computer *Controlled Systems*, Prentice-Hall, 1984. [16] L. X. Wang, *A course in Fuzzy Systems and Control*, Prentice-Hall International, Inc.,1997, [17] Norberto Bramatti. *Desenvolvimento e Implantação de um Sistema de Monitoração on-line de*
[20] Jang, J.-S. R., *ANFIS: Adaptive-Network-based Fuzzy Inference Systems*, *IEEE Transactions on Systems, Man, and Cybernetics*, Vol. 23, No. 3, pp. 665-685, May 1993.
[14] Aguirre, L. A., *Introdução à Identificação de Sistemas*, 2ª edição, UFMG, 2004.
Computers and Chemical Engineering, 27, 2003, 293-311.
*gearbox*, Elsevier, Computers in Industry, vol 57, 2006, 552568;
System Application to Power Systems, Curitiba - Brazil, 2009.
*A review of process fault detection and diagnosis Part I: Quantitative model-based methods*.
*Predictive Maintenance Application to the health condition monitoring of a windturbine*
*- SADE*. II Semana Eletronorte do Conhecimento e Inovação (II SECI). 21 a 23 de
*ELETRONORTE's applications*. The 15th International Conference on Intelligent
*Distribuidor de um Hidrogerador Utilizando Técnicas de Estimação Paramétrica e Lógica Fuzzy - Resultados Experimentais*. XIX SNPTEE - Seminário Nacional de Produção e Transmissão de Energia Elétrica. Florianópolis, SC. 23 a 26 Outubro de 2011. [10] ]Zhengang Han, Weihua Li and Sirish L. Shah; *Fault detection and isolation in the presence of process uncertainties*; Elsevier, Control Engineering Practice 13, 2005, pag 587-599;
*Redundancy A Survey and Some New Results*, Automatica, Vol. 26, No. 3, pp. 459-474,
*da Análise de Correlações e Redes Neurais Artificiais*, Tese de doutorado apresentada à Comissão de Pós-Graduação da Faculdade de Engenharia Mecânica, como requisito para a obtenção do título de Doutor em Engenharia Mecânica. Campinas,
*Compensadores Síncronos*. Dissertação de Mestrado. Universidade Federal do Pará, Centro Tecnológico, Programa de Pós-graduação em Engenharia Elétrica. 2002. [18] Lennart Ljung, *System Identification - Theory for the User*, PTR Prentice Hall, Englewood
**7. References**
Volume 2; 1989
Academic, 1999.
1990;
2005.
2003, S.P. - Brasil.
Cliffs, New Jersey, 1987;
[19] Lennart Ljung; *System Identification Tolbox 7 User's Guide*.
Prentice Hall, 1993.
outubro de 2009, São Luís - MA.
Recently, automatic systems that control driving speeds and headway distances while following a vehicle have been developed worldwide. Some products, such as adaptive cruise control systems, have already been installed in upper segments of passenger vehicles. Car following is an important operation in safe and comfortable driving on straight and/or curved roads. The number of traffic accidents involving rear-end collisions is the highest over the last decade in Japan (Iwashita et al., 2011). A rear-end collision occurs when the distance between two vehicles decreases due to deceleration of the lead vehicle and/or higher speed of the following vehicle. The automatic vehicle control system maintains a safe headway distance while following a vehicle and controls velocity according to the relative speed of the leading vehicle, in order to avoid a rear-end collision.
If the system's automatic controls do not match the driver's manual controls, driver acceptance of the automatic vehicle control systems decreases, and the driver is not likely to use them. For example, when a lead vehicle speeds up and the inter-vehicle distance increases, one driver may accelerate strongly, whereas another driver may accelerate slightly; and other drivers may not accelerate. The system's automatic hard acceleration does not suit drivers whose acceleration is slight and those who do not accelerate, and they may regard such automatic systems as dangerous. Therefore, it is expected that drivers will accept longitudinal control systems that operate in a manner similar to their own usual carfollowing behavior. Drivers' car-following behavior must be investigated in a real roadtraffic environment to develop vehicle control and driver support systems that are compatible with drivers' typical car-following behavior.
Car-following behavior consists of two aspects: how much distance drivers allow for a leading vehicle as an acceptable headway distance, and how they control acceleration according to the movements of the leading vehicle. Figure 1 presents an example of a typical following process. This car-following behavior data was recorded using an instrumented vehicle on a real motorway in Southampton (Sato et al., 2009a). The relative distance and speed were detected by microwave radar. The data length was 5min. Car-following behavior is a goal-seeking process depicted by several spirals as drivers attempt to maintain the desired following headway behind a vehicle in a car-following situation.
Understanding Driver Car-Following Behavior Using a Fuzzy Logic Car-Following Model 267
Basically, the response is the acceleration (deceleration) rate of the following vehicle. This is a function of driver sensitivity and the stimulus. The stimulus is assumed to be the difference between the speed of the lead vehicle and that of the following vehicle. Driver sensitivity is a function of the spacing between the lead and following vehicles and the speed of the following vehicle. Several derived equations have been proposed in the last 20
However, one weakness of the General Motors Model is that the response of the following vehicle is determined by one stimulus, speed relative to the leading vehicle. When the relative speed between the two vehicles is zero, the acceleration or deceleration response is zero. This is not a realistic phenomenon, because a driver decelerates to increase intervehicle separation when the relative speed is zero but the spacing is too short. To overcome this problem, Helly developed a linear model that includes the additional stimulus term of
( ) [ ( ) ( )] {[ ( ) ( )] ( )}
where Dn(t+T) is the desired following distance at time t+T; and α, β, γ, C1, and C2 are
Another limitation is the assumption of symmetrical behavior under car-following conditions. For example, a lead vehicle has a positive relative speed with a certain magnitude, and another lead vehicle has a negative relative speed with the same magnitude. In these situations, the General Motors Model gives the same deceleration rate in the first case as the acceleration rate in second case. In a real road-traffic environment, deceleration
The Stopping-Distance Model assumes that a following vehicle always maintains a safe following distance in order to bring the vehicle to a safe stop if the leading vehicle suddenly stops. This model is based on a function of the speeds of the following and leading vehicles and the follower's reaction time. The original formulation (Kometani &
where Δx is the relative distance between the lead and following vehicles; v<sup>L</sup> is the speed of the lead vehicle; vF is the speed of the following vehicle; T is the driver's reaction time; and
The Stopping-Distance Model is widely used in microscopic traffic simulations (Gipps, 1981), because of its easy calibration based on a realistic driving behavior, requiring only the maximal deceleration of the following vehicle. However, the "safe headway" concept is not a totally valid starting point, and this assumption is not consistent with empirical
Δx(t-T) =αv2L(t-T)+ βlv2
( ) ( ) ( ) (2)
F(t)+ βvF(t)+b0 (3)
years (see Mehmood et al., 2001).
calibration constants.
Stopping-Distance Model:
α, β, βl, and b0 are calibration constants.
Sasaki, 1959) is:
observations.
Action-Point Model:
the desired headway distance (Helly, 1959):
in the second case is greater than acceleration to avoid risk.
In this chapter, the range of headway distances that drivers leave for leading vehicles is the "static" aspect of car-following behavior. A driver's acceleration controls based on the relationship between the driver's own vehicle and the leading vehicle is termed the "dynamic" aspect. Following distances, Time Headway (THW) (defined by the relative distance to a leading vehicle divided by the driving speed of driver's own vehicle), and Time to Collision (TTC) (defined by the relative distance to a leading vehicle divided by the relative speed between the leading and drivers' own vehicles) are indicators for evaluating the static aspect. A number of car-following models deal with the dynamic aspect.
Fig. 1. Example of car-following behavior data collected using an instrumented vehicle on an actual road. (For details of the data collection method, please see section 3.2.)
### **1.1 Brief review of car-following models**
Car-following models have been developed since the 1950s (e.g., Pipes, 1953). Many models describe the accelerative behavior of a driver as a function of inter-vehicle separation and relative speed. The following are representative car-following models (for details, please see Brackstone & McDonald, 1999).
General Motors Model:
The fundamental concept behind the General Motors Model is the stimulus-response theory (Chandler et al., 1958). Equation (1) presents a representative formulation.
$$\mathbf{a\_{F}(t+T)} = \alpha \left[ \frac{[\mathbf{v\_{F}(t)}]^{n}}{[\mathbf{x\_{f}(t) - \mathbf{x\_{F}(t)}]^{l}}} \right] [\mathbf{V\_{L}(t) - \mathbf{V\_{F}(t)}] \tag{1}$$
where aF(t+T) is the acceleration or deceleration rate of the following vehicle at time t+T; VL(t) is the speed of the lead vehicle at time t; VF(t) is the speed of the following vehicle at time t; XL(t) is the spacing of the lead vehicle at time t; XF(t) is the spacing of the following vehicle at time t; T is the perception-reaction time of the driver; and m, l, and α are constants to be determined.
Basically, the response is the acceleration (deceleration) rate of the following vehicle. This is a function of driver sensitivity and the stimulus. The stimulus is assumed to be the difference between the speed of the lead vehicle and that of the following vehicle. Driver sensitivity is a function of the spacing between the lead and following vehicles and the speed of the following vehicle. Several derived equations have been proposed in the last 20 years (see Mehmood et al., 2001).
However, one weakness of the General Motors Model is that the response of the following vehicle is determined by one stimulus, speed relative to the leading vehicle. When the relative speed between the two vehicles is zero, the acceleration or deceleration response is zero. This is not a realistic phenomenon, because a driver decelerates to increase intervehicle separation when the relative speed is zero but the spacing is too short. To overcome this problem, Helly developed a linear model that includes the additional stimulus term of the desired headway distance (Helly, 1959):
$$\mathbf{a}\_{\rm F}(\mathbf{t} + \mathbf{T}) = \mathbf{C}\_{\rm 1}[\mathbf{V\_{L}(t)} - \mathbf{V\_{F}(t)}] + \mathbf{C}\_{\rm 2}\{[\mathbf{X\_{L}(t)} - \mathbf{X\_{F}(t)}] - \text{Dn}(\mathbf{t} + \mathbf{T})\}$$
$$\mathbf{Dn}(\mathbf{t} + \mathbf{T}) = \mathbf{a} + \beta \mathbf{V\_{F}(t)} + \gamma \mathbf{a\_{F}(t)}\tag{2}$$
where Dn(t+T) is the desired following distance at time t+T; and α, β, γ, C1, and C2 are calibration constants.
Another limitation is the assumption of symmetrical behavior under car-following conditions. For example, a lead vehicle has a positive relative speed with a certain magnitude, and another lead vehicle has a negative relative speed with the same magnitude. In these situations, the General Motors Model gives the same deceleration rate in the first case as the acceleration rate in second case. In a real road-traffic environment, deceleration in the second case is greater than acceleration to avoid risk.
Stopping-Distance Model:
266 Fuzzy Logic – Algorithms, Techniques and Implementations
In this chapter, the range of headway distances that drivers leave for leading vehicles is the "static" aspect of car-following behavior. A driver's acceleration controls based on the relationship between the driver's own vehicle and the leading vehicle is termed the "dynamic" aspect. Following distances, Time Headway (THW) (defined by the relative distance to a leading vehicle divided by the driving speed of driver's own vehicle), and Time to Collision (TTC) (defined by the relative distance to a leading vehicle divided by the relative speed between the leading and drivers' own vehicles) are indicators for evaluating
the static aspect. A number of car-following models deal with the dynamic aspect.
aspect
Static
Fig. 1. Example of car-following behavior data collected using an instrumented vehicle on
Car-following models have been developed since the 1950s (e.g., Pipes, 1953). Many models describe the accelerative behavior of a driver as a function of inter-vehicle separation and relative speed. The following are representative car-following models (for details, please see
The fundamental concept behind the General Motors Model is the stimulus-response theory
[ ( ) ( )]
where aF(t+T) is the acceleration or deceleration rate of the following vehicle at time t+T; VL(t) is the speed of the lead vehicle at time t; VF(t) is the speed of the following vehicle at time t; XL(t) is the spacing of the lead vehicle at time t; XF(t) is the spacing of the following vehicle at time t; T is the perception-reaction time of the driver; and m, l, and α are constants
][ ( ) ( )] (1)
Dynamic
aspect
an actual road. (For details of the data collection method, please see section 3.2.)
0 10 20 30 40 50
Relative distance (m)
(Chandler et al., 1958). Equation (1) presents a representative formulation.
( ) [ [ ( )]
**1.1 Brief review of car-following models**
Brackstone & McDonald, 1999).
General Motors Model:
0
Relative speed (m/s)
2
4
to be determined.
The Stopping-Distance Model assumes that a following vehicle always maintains a safe following distance in order to bring the vehicle to a safe stop if the leading vehicle suddenly stops. This model is based on a function of the speeds of the following and leading vehicles and the follower's reaction time. The original formulation (Kometani & Sasaki, 1959) is:
$$
\Delta \mathbf{x}(\mathbf{t} \cdot \mathbf{T}) = \mathbf{c} \mathbf{v} \cdot \mathbf{\hat{L}}(\mathbf{t} \cdot \mathbf{T}) + \beta\_1 \mathbf{v}^2 \mathbf{\hat{F}}(\mathbf{t}) + \beta \mathbf{v}\_\mathbf{F}(\mathbf{t}) + \mathbf{b}\_0 \tag{3}
$$
where Δx is the relative distance between the lead and following vehicles; v<sup>L</sup> is the speed of the lead vehicle; vF is the speed of the following vehicle; T is the driver's reaction time; and α, β, βl, and b0 are calibration constants.
The Stopping-Distance Model is widely used in microscopic traffic simulations (Gipps, 1981), because of its easy calibration based on a realistic driving behavior, requiring only the maximal deceleration of the following vehicle. However, the "safe headway" concept is not a totally valid starting point, and this assumption is not consistent with empirical observations.
Action-Point Model:
Understanding Driver Car-Following Behavior Using a Fuzzy Logic Car-Following Model 269
Sugeno-type. The main difference between the Mamdani and Sugeno types is that the output membership functions are only linear or constant for Sugeno-type fuzzy inference. A
The fuzzy logic car-following model was developed by the Transportation Research Group (TRG) at the University of Southampton (Wu et al., 2000). McDonald et al. collected carfollowing behavior data on real roads and developed and validated the proposed fuzzy logic car-following model based on the real-world data (briefly mentioned in 2.2 and 2.3;
The fuzzy logic model uses relative velocity and distance divergence (DSSD) (the ratio of headway distance to a desired headway) as input variables. The output variable is the acceleration-deceleration rate. The DSSD is the average of the headway distance that is observed when the relative speeds between vehicles are close to zero. This model adopts fuzzy functions (fuzzy sets described by membership functions) as the formula for the input-output
**Fuzzy Inference System**
Rule 1 Rule 2 Rule 3 Rule 4 Rule 5 Rule 6
Output membership function
Rule 8 Rule 7
Rule 9 Rule 10 Rule 11 Rule 12 Rule 13 Rule 14 Rule 15 **Acceleration**
relationship. Figure 2 depicts the structure of the fuzzy logic car-following model.
Input membership function
Fig. 2. Structure of the fuzzy inference system in the fuzzy logic car-following model
Specifications of the fuzzy inference system are as follows.
Type of input membership function: Gaussian Type of output membership function: Constant
Type of inference system: Sugeno
**DSSD**
Headway distance the desired headway <sup>=</sup>
> **Relative Velocity**
Velocity of leading vehicle – velocity of driver's vehicle
=
The constant output membership function is obtained from a singleton spike (*p*=*q*=0).
typical rule in the Sugeno-type fuzzy inference (Sugeno, 1985) is:
If input *x* is *A* and input *y* is *B* then output *z* is *x*\**p*+*y*\**q*+*r;* where *A* and *B* are fuzzy sets and *p*, *q*, and *r* are constants.
please see Wu, 2003; Zheng, 2003 for further explanation).
**2.1 Overview**
The Action-Point Model is the first car-following model to incorporate human perception of motion. The model developed by Michaels suggests that the dominant perceptual factor is changes in the apparent size of the vehicle (i.e., the changing rate of visual angle) (Michaels, 1963):
$$\frac{d\theta}{dt} = \frac{4\ast\mathsf{W}\_{\mathrm{L}}}{4\ast\left[\mathsf{X}\_{\mathrm{L}}(\mathrm{t}) - \mathsf{X}\_{\mathrm{F}}(\mathrm{t})\right]^{2} + \mathsf{W}\_{\mathrm{L}}^{2}} \left[\mathsf{V}\_{\mathrm{L}}(\mathrm{t}) - \mathsf{V}\_{\mathrm{F}}(\mathrm{t})\right] \tag{4}$$
where WL is the width of the lead vehicle.
This model assumes that a driver appropriately accelerates or -decelerates if the angular velocity exceeds a certain threshold. Once the threshold is exceeded, the driver chooses to decelerate until he/she can no longer perceive any relative velocity. Thresholds include a spacing-based threshold that is particularly relevant in close headway situations, a relative speed threshold for the perception of closing, and thresholds for the perception of opening and closing for low relative speeds (a recent work suggests that the perception of opening and that of closing have different thresholds (Reiter, 1994)). Car-following conditions are further categorized into subgroups: free driving, overtaking, following, and emergency situation. A driver engages in different acceleration behaviors in different situations when the perceived physical perception exceeds the thresholds.
The Action-Point Model takes into account the human threshold of perception, establishing a realistic rationale. However, various efforts have focused on identifying threshold values during the calibration phase, while the adjustment of acceleration above the threshold has not been considered, and the acceleration rate is normally assumed to be a constant. Additionally, the model dynamic (switching between the subgroups) has not been investigated. Finally, the ability to perceive speed differences and estimate distances varies widely among drivers. Therefore, it is difficult to estimate and calibrate the individual thresholds associated with the Action-Point Model.
### **2. Fuzzy logic car-following model**
Drivers perform a car-following task with real-time information processing of several kinds of information sources. The car-following models discussed above have established a unique interpretation of drivers' car-following behaviors. A driver in a car-following situation is described as a stimuli-responder in the General Motors Model, a safe distancekeeper in the Stopping-Distance Model, and a state monitor who wants to keep perceptions below the threshold in the Action-Point Model. However, these models include non-realistic constraints to describe car-following behavior in real road-traffic environments: symmetry between acceleration and deceleration, the "safe headway" concept, and constant acceleration or deceleration above the threshold.
The fuzzy logic car-following model describes driving operations under car-following conditions using linguistic terms and associated rules, instead of deterministic mathematical functions. Car-following behavior can be described in a natural manner that reflects the imprecise and incomplete sensory data presented by human sensory modalities. The fuzzy logic car-following model treats a driver as a decision-maker who decides the controls based on sensory inputs using a fuzzy reasoning. There are two types of fuzzy inference system that uses fuzzy reasoning to map an input space to an output space, Mandani-type and Sugeno-type. The main difference between the Mamdani and Sugeno types is that the output membership functions are only linear or constant for Sugeno-type fuzzy inference. A typical rule in the Sugeno-type fuzzy inference (Sugeno, 1985) is:
If input *x* is *A* and input *y* is *B* then output *z* is *x*\**p*+*y*\**q*+*r;*
where *A* and *B* are fuzzy sets and *p*, *q*, and *r* are constants.
The constant output membership function is obtained from a singleton spike (*p*=*q*=0).
### **2.1 Overview**
268 Fuzzy Logic – Algorithms, Techniques and Implementations
The Action-Point Model is the first car-following model to incorporate human perception of motion. The model developed by Michaels suggests that the dominant perceptual factor is changes in the apparent size of the vehicle (i.e., the changing rate of visual angle) (Michaels,
>
This model assumes that a driver appropriately accelerates or -decelerates if the angular velocity exceeds a certain threshold. Once the threshold is exceeded, the driver chooses to decelerate until he/she can no longer perceive any relative velocity. Thresholds include a spacing-based threshold that is particularly relevant in close headway situations, a relative speed threshold for the perception of closing, and thresholds for the perception of opening and closing for low relative speeds (a recent work suggests that the perception of opening and that of closing have different thresholds (Reiter, 1994)). Car-following conditions are further categorized into subgroups: free driving, overtaking, following, and emergency situation. A driver engages in different acceleration behaviors in different situations when
The Action-Point Model takes into account the human threshold of perception, establishing a realistic rationale. However, various efforts have focused on identifying threshold values during the calibration phase, while the adjustment of acceleration above the threshold has not been considered, and the acceleration rate is normally assumed to be a constant. Additionally, the model dynamic (switching between the subgroups) has not been investigated. Finally, the ability to perceive speed differences and estimate distances varies widely among drivers. Therefore, it is difficult to estimate and calibrate the individual
Drivers perform a car-following task with real-time information processing of several kinds of information sources. The car-following models discussed above have established a unique interpretation of drivers' car-following behaviors. A driver in a car-following situation is described as a stimuli-responder in the General Motors Model, a safe distancekeeper in the Stopping-Distance Model, and a state monitor who wants to keep perceptions below the threshold in the Action-Point Model. However, these models include non-realistic constraints to describe car-following behavior in real road-traffic environments: symmetry between acceleration and deceleration, the "safe headway" concept, and constant
The fuzzy logic car-following model describes driving operations under car-following conditions using linguistic terms and associated rules, instead of deterministic mathematical functions. Car-following behavior can be described in a natural manner that reflects the imprecise and incomplete sensory data presented by human sensory modalities. The fuzzy logic car-following model treats a driver as a decision-maker who decides the controls based on sensory inputs using a fuzzy reasoning. There are two types of fuzzy inference system that uses fuzzy reasoning to map an input space to an output space, Mandani-type and
[ ( ) ( )] (4)
the perceived physical perception exceeds the thresholds.
thresholds associated with the Action-Point Model.
acceleration or deceleration above the threshold.
**2. Fuzzy logic car-following model**
where WL is the width of the lead vehicle.
[ ( ) ( )]
1963):
The fuzzy logic car-following model was developed by the Transportation Research Group (TRG) at the University of Southampton (Wu et al., 2000). McDonald et al. collected carfollowing behavior data on real roads and developed and validated the proposed fuzzy logic car-following model based on the real-world data (briefly mentioned in 2.2 and 2.3; please see Wu, 2003; Zheng, 2003 for further explanation).
The fuzzy logic model uses relative velocity and distance divergence (DSSD) (the ratio of headway distance to a desired headway) as input variables. The output variable is the acceleration-deceleration rate. The DSSD is the average of the headway distance that is observed when the relative speeds between vehicles are close to zero. This model adopts fuzzy functions (fuzzy sets described by membership functions) as the formula for the input-output relationship. Figure 2 depicts the structure of the fuzzy logic car-following model.
Specifications of the fuzzy inference system are as follows.
Understanding Driver Car-Following Behavior Using a Fuzzy Logic Car-Following Model 271
where Ŷ<sup>i</sup> is a predicted value using the fuzzy logic model at time increment i, Yi is raw data
All possible model formulations (a single variable, combination of two variables, and combination of three variables) were tested. The data were collected on real motorways using a TRG instrumented vehicle. Although a three-input model suggested better RMSE performance than a one-input model or a two-input model, the two-input model using relative speed and distance divergence was adopted because of the complexity of the model structure and its applicability to a wide range of car-following situations. For details of the
The developed fuzzy logic car-following model was validated in terms of reproducing a single vehicle's car-following behavior, as well as reproducing traffic flow under car-
The single vehicle's car-following behavior was evaluated from empirical data, and the average RMSE of acceleration was 0.20m/s2. The platoon behavior was evaluated using simulation. The response of a platoon of 20 vehicles to step changes of acceleration or deceleration of a lead vehicle was assessed in order to investigate the influence of the movement of the lead vehicle on a line of vehicles. The results validated that the fuzzy logic car-following model could reproduce both stable and unstable traffic behavior. For details of
**3. Case study 1: Car-following behavior comparison between the UK and**
This section introduces a case study focusing on a comparison of drivers' car-following behavior in the UK and in Japan (Sato et al., 2009b). The fuzzy logic car-following model was developed using naturalistic data collected in Southampton. We applied this model to behavioral data collected in Japan. One objective is to confirm whether Japanese carfollowing behavior can be described by the fuzzy logic model with the same structure as the UK model. Another objective is to investigate cross-cultural variations of the car-following
With increasing globalization of automotive markets, it is important to understand the differences between driving behavior in different countries. Car-following behavior may differ due to differences in nationality and the road traffic environments of different countries. The findings may contribute to designing human-centered automatic vehicle
An AIST instrumented vehicle and a TRG instrumented vehicle are used for behavioral data collection (Brackstone et al., 1999; Sato & Akamatsu, 2007). Both vehicles are equipped with
control systems based on international differences in driving behavior.
at time increment i, and N is the number of data.
input variable validation, refer to Zheng, 2003.
following conditions (a platoon of vehicles).
behaviors of drivers in the two countries.
the model validation, refer to Wu et al., 2003 and Zheng, 2003.
**2.3 Model validation**
**Japan**
**3.1 Motivation**
**3.2 Methods**
**3.2.1 Instrumented vehicles**
The parameter of the fuzzy inference system is estimated using the following combination of back-propagation and least square methods. The initial fuzzy inference system adopts the grid partition method in which the membership functions of each input are evenly assigned in the range of the training data. Next, the membership function parameters are adjusted using the hybrid learning algorithm. The parameters of output membership functions are updated in a forward pass using the least square method. The inputs are first propagated forward. The overall output is then a linear combination of the parameters of output membership functions. The parameters of input membership functions are estimated using back propagation in each iteration, where the differences between model output and training data are propagated backward and the parameters are updated by gradient descent. The parameter optimization routines are applied until a given number of iterations or an error reduction threshold is reached.
The input-output mapping specified by the fuzzy inference system has a three-dimensional structure. We focus on relative velocity-acceleration mapping in order to analyze the dynamic aspect of car-following behavior (i.e., drivers' acceleration controls based on the variation in relative speeds).
### **2.2 Input variable validation**
The following eight candidates were applied to the fuzzy inference system estimation in order to obtain satisfactory performance of the fuzzy logic model.
The performance of the fuzzy logic model was evaluated by the Root Mean Square Error (RMSE) of the model prediction:
$$\text{RMSE} = \sqrt{\frac{\sum\_{l=1}^{N} (\overline{Y\_l} - Y\_l)^2}{N}} \tag{5}$$
where Ŷ<sup>i</sup> is a predicted value using the fuzzy logic model at time increment i, Yi is raw data at time increment i, and N is the number of data.
All possible model formulations (a single variable, combination of two variables, and combination of three variables) were tested. The data were collected on real motorways using a TRG instrumented vehicle. Although a three-input model suggested better RMSE performance than a one-input model or a two-input model, the two-input model using relative speed and distance divergence was adopted because of the complexity of the model structure and its applicability to a wide range of car-following situations. For details of the input variable validation, refer to Zheng, 2003.
### **2.3 Model validation**
270 Fuzzy Logic – Algorithms, Techniques and Implementations
Number of partitions for input (Relative Velocity): 5 (closing+, closing, about zero,
The parameter of the fuzzy inference system is estimated using the following combination of back-propagation and least square methods. The initial fuzzy inference system adopts the grid partition method in which the membership functions of each input are evenly assigned in the range of the training data. Next, the membership function parameters are adjusted using the hybrid learning algorithm. The parameters of output membership functions are updated in a forward pass using the least square method. The inputs are first propagated forward. The overall output is then a linear combination of the parameters of output membership functions. The parameters of input membership functions are estimated using back propagation in each iteration, where the differences between model output and training data are propagated backward and the parameters are updated by gradient descent. The parameter optimization routines are applied until a given number of iterations
The input-output mapping specified by the fuzzy inference system has a three-dimensional structure. We focus on relative velocity-acceleration mapping in order to analyze the dynamic aspect of car-following behavior (i.e., drivers' acceleration controls based on the
The following eight candidates were applied to the fuzzy inference system estimation in
The performance of the fuzzy logic model was evaluated by the Root Mean Square Error
̂ )
(5)
√<sup>∑</sup> (
order to obtain satisfactory performance of the fuzzy logic model.
relative speeds between vehicles were close to zero.)
Learning algorithm: combination of back-propagation and least square methods
Number of partitions for input (DSSD): 3 (close, ok, and far) Initialization of fuzzy inference system: grid partition method
opening, and opening+)
Defuzzification method: weighted average
or an error reduction threshold is reached.
(RMSE) of the model prediction:
= 0.)
variation in relative speeds).
**2.2 Input variable validation**
The developed fuzzy logic car-following model was validated in terms of reproducing a single vehicle's car-following behavior, as well as reproducing traffic flow under carfollowing conditions (a platoon of vehicles).
The single vehicle's car-following behavior was evaluated from empirical data, and the average RMSE of acceleration was 0.20m/s2. The platoon behavior was evaluated using simulation. The response of a platoon of 20 vehicles to step changes of acceleration or deceleration of a lead vehicle was assessed in order to investigate the influence of the movement of the lead vehicle on a line of vehicles. The results validated that the fuzzy logic car-following model could reproduce both stable and unstable traffic behavior. For details of the model validation, refer to Wu et al., 2003 and Zheng, 2003.
### **3. Case study 1: Car-following behavior comparison between the UK and Japan**
### **3.1 Motivation**
This section introduces a case study focusing on a comparison of drivers' car-following behavior in the UK and in Japan (Sato et al., 2009b). The fuzzy logic car-following model was developed using naturalistic data collected in Southampton. We applied this model to behavioral data collected in Japan. One objective is to confirm whether Japanese carfollowing behavior can be described by the fuzzy logic model with the same structure as the UK model. Another objective is to investigate cross-cultural variations of the car-following behaviors of drivers in the two countries.
With increasing globalization of automotive markets, it is important to understand the differences between driving behavior in different countries. Car-following behavior may differ due to differences in nationality and the road traffic environments of different countries. The findings may contribute to designing human-centered automatic vehicle control systems based on international differences in driving behavior.
### **3.2 Methods**
### **3.2.1 Instrumented vehicles**
An AIST instrumented vehicle and a TRG instrumented vehicle are used for behavioral data collection (Brackstone et al., 1999; Sato & Akamatsu, 2007). Both vehicles are equipped with
Understanding Driver Car-Following Behavior Using a Fuzzy Logic Car-Following Model 273
roundabout junctions. The driving behavior data in Southampton was collected as part of an EC STARDUST project (Zheng et al., 2006). The field experiments at the two sites were
The passive mode was used for the data collected (Fig. 5), reflecting random drivers who followed the instrumented vehicle. The passive mode can collect and evaluate a large population of drivers, rather than just the participating driver in the instrumented vehicle, in a short period and at a lower level of detail (Brackstone et al., 2002). The measured data in the passive mode enable evaluation of car-following behavior trends in each country.
Urban road Bypass Trunk road Motorway
conducted during the morning from 9:00 to 10:45.
Fig. 4. Road environments used for car-following behavior analyses
Fig. 5. Active and passive modes in car-following conditions
Following vehicle
In the analysis, the car-following condition was defined as a situation in which a driver followed a leading vehicle with relative speeds between 15km/h and -15km/h. The relative distance to a following vehicle under car-following conditions was obtained from the
Random drivers who are not involved in experiments
"Passive mode" "Active mode"
Instrumented vehicle
Driver participating in experiments
Leading vehicle
**3.2.3 Variables**
Tsukuba (Japan)
Southampton (UK)
various sensors and driving recorder systems in order to detect the vehicle driving status and to measure the driver's operations. Velocity is measured using a speed pulse signal, and acceleration is detected by a G-sensor. The relative distance and relative speed to the leading and following vehicles are recorded with laser radar units (AIST instrumented vehicle) or microwave radar (TRG instrumented vehicle) that are fixed within the front and rear bumpers. Figure 3 presents an overview of the AIST instrumented vehicle. This vehicle collects the following data:
Fig. 3. AIST instrumented vehicle with sensors and a recorder system for detecting nearby vehicles
The velocity of the following vehicle was calculated based on the velocity of the instrumented vehicles and the relative speed. The visual image of the rear scenes was used for better understanding of the traffic conditions while driving and for clarifying uncertainties identified in the radars.
### **3.2.2 Road-traffic environment**
Figure 4 depicts the road environment in the Southampton (UK) and Tsukuba (Japan) routes. The driving route in Tsukuba was 15km long (travel time 30min). This route included urban roads with several left and right turns at intersections, with a traffic lane that was mostly one lane, and a bypass that had one and two traffic lanes. The driving route in Southampton included trunk roads and motorways with two and three lanes and roundabout junctions. The driving behavior data in Southampton was collected as part of an EC STARDUST project (Zheng et al., 2006). The field experiments at the two sites were conducted during the morning from 9:00 to 10:45.
### **3.2.3 Variables**
272 Fuzzy Logic – Algorithms, Techniques and Implementations
various sensors and driving recorder systems in order to detect the vehicle driving status and to measure the driver's operations. Velocity is measured using a speed pulse signal, and acceleration is detected by a G-sensor. The relative distance and relative speed to the leading and following vehicles are recorded with laser radar units (AIST instrumented vehicle) or microwave radar (TRG instrumented vehicle) that are fixed within the front and rear bumpers. Figure 3 presents an overview of the AIST instrumented vehicle. This vehicle
Relative distance and speed to the leading and following vehicles by laser radar units,
Visual images (forward and rear scenes, lane positions, and driver's face) by five CCD
Fig. 3. AIST instrumented vehicle with sensors and a recorder system for detecting nearby
Driving recorder system, sampling rate at 30Hz
Laser radar for relative distance and velocity to a leading vehicle
The velocity of the following vehicle was calculated based on the velocity of the instrumented vehicles and the relative speed. The visual image of the rear scenes was used for better understanding of the traffic conditions while driving and for clarifying
Figure 4 depicts the road environment in the Southampton (UK) and Tsukuba (Japan) routes. The driving route in Tsukuba was 15km long (travel time 30min). This route included urban roads with several left and right turns at intersections, with a traffic lane that was mostly one lane, and a bypass that had one and two traffic lanes. The driving route in Southampton included trunk roads and motorways with two and three lanes and
collects the following data:
cameras.
vehicles
Driving speed by speed pulse signal,
Geographical position by D-GPS sensor,
Position of driver's right foot by laser sensors,
Application of gas and brake pedals by potentiometers,
Vehicle acceleration by G-sensor, Angular velocity by gyro sensor,
Steering wheel angle by encoder, Turn signal activation by encoder, and
> Laser radar for relative distance and velocity to a following vehicle
uncertainties identified in the radars.
**3.2.2 Road-traffic environment**
The passive mode was used for the data collected (Fig. 5), reflecting random drivers who followed the instrumented vehicle. The passive mode can collect and evaluate a large population of drivers, rather than just the participating driver in the instrumented vehicle, in a short period and at a lower level of detail (Brackstone et al., 2002). The measured data in the passive mode enable evaluation of car-following behavior trends in each country.
Fig. 4. Road environments used for car-following behavior analyses
Random drivers who are not involved in experiments
Fig. 5. Active and passive modes in car-following conditions
In the analysis, the car-following condition was defined as a situation in which a driver followed a leading vehicle with relative speeds between 15km/h and -15km/h. The relative distance to a following vehicle under car-following conditions was obtained from the
Understanding Driver Car-Following Behavior Using a Fuzzy Logic Car-Following Model 275
Figure 7 presents the relative velocity–acceleration mapping obtained from the fuzzy inference specification in Tsukuba and Southampton. The two sites have similar traces (Southampton, 15; Tsukuba, 14) and data length (Southampton, 511.5sec; Tsukuba, 522.9sec). The RMSEs of the predicted acceleration and the measured data in the estimated fuzzy logic model were 0.15m/sec2 in Tsukuba and 0.17m/sec2 in Southampton. These findings indicate a satisfactory model-to-data fit compared to other published works (Wu et
The deceleration of Tsukuba drivers is greater than that of Southampton drivers when their vehicle approaches the leading vehicle. When the distance between vehicles is opening, the acceleration of Southampton drivers is greater than that of Tsukuba drivers. Thus, the acceleration-deceleration rate of Tsukuba drivers indicates a tendency opposite that of
Fig. 7. Results of fuzzy logic model specification: Relative velocity–acceleration mapping
*Strong acceleration*
*Deceleration*
Acceleration (m/sec2)
Southampton [UK]
*Closing* Relative Velocity (m/sec) *Opening*
The low RMSE of the Tsukuba acceleration rate suggests that the proposed fuzzy logic model is well-suited to Japanese car-following behavior. The findings imply that Japanese drivers use relative velocity and distance divergence for adjusting acceleration and
The THW of Tsukuba drivers was longer at slow velocity. When Tsukuba drivers approached a preceding vehicle in the same traffic lane, they decelerated more strongly. In addition, Tsukuba drivers accelerated less as the distance to the leading vehicle increased. Strong deceleration while moving toward the leading vehicle and weak acceleration when
Southampton drivers tended to adopt shorter THW when in car-following in the low driving speed range. The acceleration rate of Southampton drivers was higher than that of Tsukuba drivers when overtaking a vehicle. It is assumed that such strong acceleration
contributes to maintaining a short headway distances in car-following situations.
**3.3.2 Dynamic aspect**
Southampton drivers.
*Strong acceleration*
*Deceleration*
Acceleration (m/sec2)
between Tsukuba and Southampton
deceleration while following a vehicle.
following a preceding vehicle led to long headway distances.
*Closing* Relative Velocity (m/sec) *Opening* Tsukuba [Japan]
**3.4 Discussion**
al., 2003).
measured data. The rear distances collected were divided into two sets in terms of the associated driving speeds: 30 to 49km/h and 50 to 69km/h. The speed range of 30 to 49km/h corresponds to driving on an urban road (Tsukuba) and on a trunk road (Southampton), while the speed of 50 of 69km/h corresponds to driving on a bypass (Tsukuba) and on a motorway (Southampton).
The THW of the passive mode (defined by the relative distance between the following vehicle and the instrumented vehicle divided by the driving speed of the following vehicle) was calculated, and the distributions of the THW at each set were compared for analysis of the static aspect of car-following behavior.
In addition to the rear distances, the relative speeds and acceleration of the following vehicle were used for the fuzzy logic car-following model. Although this model can be used to describe individual drivers' acceleration-deceleration behavior, we applied the model to the passive mode data in order to compare general features of the dynamic aspect of carfollowing behavior between Tsukuba and Southampton. The continuous data for more than 20sec was input to the model specification within the measured car- following data.
### **3.3 Results**
### **3.3.1 Static aspect**
Figure 6 presents the distributions of the THW for each speed range and proportions of the time when drivers take the relevant THW to the total time while driving at the corresponding velocity.
In the lower speed range (30 to 49km/h), the proportion of Southampton drivers taking very short THW (0.5 to 1s) exceeds that of Tsukuba drivers. The proportion of Tsukuba drivers taking THW longer than 3s exceeds that of Southampton drivers. In the higher speed range (50 to 69km/h), no difference in THW between the two regions is observed. Both Tsukuba drivers and Southampton drivers spend more time with the short THW (0.5s to 1.5s). As mentioned in previous research (Brackstone et al., 2009), THW tends to decrease as velocity increases.
Fig. 6. Comparison of THW between two countries for each speed range
### **3.3.2 Dynamic aspect**
274 Fuzzy Logic – Algorithms, Techniques and Implementations
measured data. The rear distances collected were divided into two sets in terms of the associated driving speeds: 30 to 49km/h and 50 to 69km/h. The speed range of 30 to 49km/h corresponds to driving on an urban road (Tsukuba) and on a trunk road (Southampton), while the speed of 50 of 69km/h corresponds to driving on a bypass
The THW of the passive mode (defined by the relative distance between the following vehicle and the instrumented vehicle divided by the driving speed of the following vehicle) was calculated, and the distributions of the THW at each set were compared for analysis of
In addition to the rear distances, the relative speeds and acceleration of the following vehicle were used for the fuzzy logic car-following model. Although this model can be used to describe individual drivers' acceleration-deceleration behavior, we applied the model to the passive mode data in order to compare general features of the dynamic aspect of carfollowing behavior between Tsukuba and Southampton. The continuous data for more than
Figure 6 presents the distributions of the THW for each speed range and proportions of the time when drivers take the relevant THW to the total time while driving at the
In the lower speed range (30 to 49km/h), the proportion of Southampton drivers taking very short THW (0.5 to 1s) exceeds that of Tsukuba drivers. The proportion of Tsukuba drivers taking THW longer than 3s exceeds that of Southampton drivers. In the higher speed range (50 to 69km/h), no difference in THW between the two regions is observed. Both Tsukuba drivers and Southampton drivers spend more time with the short THW (0.5s to 1.5s). As mentioned in previous research (Brackstone et al., 2009), THW tends to decrease as
Tsukuba [Japan] Southampton [UK]
Fig. 6. Comparison of THW between two countries for each speed range
3.5
4~4.5
4.5
~5
5~
0% 5% 10% 15% 20% 25% 30% 35% 40%
THW (s) from 30km/h to 49km/h THW (s) from 50km/h to 69km/h
0
0~0.5
0.5
~1
1~1.5
1.5
2~2.5
2.5
~3
3~3.5
3.5
4~4.5
4.5
~5
5~
~4
~2
~4
Frequency rate Frequency rate
20sec was input to the model specification within the measured car- following data.
(Tsukuba) and on a motorway (Southampton).
the static aspect of car-following behavior.
**3.3 Results**
**3.3.1 Static aspect**
velocity increases.
0% 5% 10% 15% 20% 25% 30% 35% 40%
0
0~0.5
0.5
~1
1~1.5
1.5
2~2.5
2.5
~3
3~3.5
~2
corresponding velocity.
Figure 7 presents the relative velocity–acceleration mapping obtained from the fuzzy inference specification in Tsukuba and Southampton. The two sites have similar traces (Southampton, 15; Tsukuba, 14) and data length (Southampton, 511.5sec; Tsukuba, 522.9sec). The RMSEs of the predicted acceleration and the measured data in the estimated fuzzy logic model were 0.15m/sec2 in Tsukuba and 0.17m/sec2 in Southampton. These findings indicate a satisfactory model-to-data fit compared to other published works (Wu et al., 2003).
The deceleration of Tsukuba drivers is greater than that of Southampton drivers when their vehicle approaches the leading vehicle. When the distance between vehicles is opening, the acceleration of Southampton drivers is greater than that of Tsukuba drivers. Thus, the acceleration-deceleration rate of Tsukuba drivers indicates a tendency opposite that of Southampton drivers.
Fig. 7. Results of fuzzy logic model specification: Relative velocity–acceleration mapping between Tsukuba and Southampton
### **3.4 Discussion**
The low RMSE of the Tsukuba acceleration rate suggests that the proposed fuzzy logic model is well-suited to Japanese car-following behavior. The findings imply that Japanese drivers use relative velocity and distance divergence for adjusting acceleration and deceleration while following a vehicle.
The THW of Tsukuba drivers was longer at slow velocity. When Tsukuba drivers approached a preceding vehicle in the same traffic lane, they decelerated more strongly. In addition, Tsukuba drivers accelerated less as the distance to the leading vehicle increased. Strong deceleration while moving toward the leading vehicle and weak acceleration when following a preceding vehicle led to long headway distances.
Southampton drivers tended to adopt shorter THW when in car-following in the low driving speed range. The acceleration rate of Southampton drivers was higher than that of Tsukuba drivers when overtaking a vehicle. It is assumed that such strong acceleration contributes to maintaining a short headway distances in car-following situations.
Understanding Driver Car-Following Behavior Using a Fuzzy Logic Car-Following Model 277
driving styles within a few years. One aim of this study is to clarify how elderly drivers follow a lead vehicle, based on analysis of how car-following behavior changes with aging. We collected car-following behavior data of elderly drivers determined in one year and compared it with that determined five years later. The distributions of THW in the two field experiments were compared in order to investigate the static aspect of car-following behavior. For analysis of the dynamic aspect, the fuzzy logic car-following model was
Field experiments were conducted in 2003 (first experiment) and in 2008 (second experiment). The two experiments were conducted using the same instrumented vehicle, the same driving route, and the same participants. The AIST instrumented vehicle was used for the data collection (Fig. 3). Almost all the sensors and the driving recorder system were fixed inside the trunk, so that the participating drivers could not see them, in order to
The experiments were conducted on rural roads around Tsukuba. The route included several left and right turns, and the travel time was 25min (total distance 14km). The participant rode alone in the instrumented vehicle during the experiment trials. Before the recorded drives, the participants performed practice drives from the starting point to the
Four elderly drivers (three males and one female) with informed consent participated in the two experiments. Their ages ranged from 65 to 70 years (average 67.3 years) in the first experiment and from 70 to 74 years (average 72.0 years) in the second experiment. Their annual distance driven ranged from 5,000 to 8,000km in the first experiment and from 2,000
The participants were instructed to drive in their typical manner. In the first experiment, the recorded trip for each elderly participant was made once a day on weekdays, for a total of 10 trips. In the second experiment, the trial was conducted twice a day on weekdays, for a total of 30 trips. The participants took a break between the experiment trials in the second
Figure 8 depicts a target section for the analysis of elderly drivers' car-following behavior. We focused on a two-lane bypass, the same road environment as that in section 3.2.2. We included only drives with a leading vehicle, excluding drives without a leading vehicle on
The active mode (distance between the instrumented vehicle and the leading vehicle) was used in the analysis of the elderly drivers' car-following behavior. We also used the passive mode (distance between the instrumented vehicle and the following vehicle) to investigate traffic characteristics on the analyzed road. The latter is expected to indicate whether changes in elderly drivers' car-following behaviors are influenced by their functional
applied to compare elderly drivers' accelerative behavior while following a vehicle.
encourage natural driving behaviors during the experiment trials.
destination without using a map or an in-vehicle navigation system.
declines or by changes in traffic characteristics on the target road.
to 10,000km in the second experiment (average 6,500km in both experiments).
**4.2 Methods**
experiment.
**4.2.2 Data analysis**
the target section.
**4.2.1 Procedures**
Tsukuba car-following behavior data were collected on urban roads and a bypass. When driving on urban roads, a leading vehicle often has to decelerate suddenly due to other vehicles at crossroads, a change of traffic signals, and the emergence of pedestrians or bicycles. The leading vehicle might also slow down suddenly on the bypass because a merging car may cut in front of it. Drivers adopted longer headway distances and decelerated more strongly in closing inter-vehicle separations when driving on roads where they should pay more attention to the movements of the leading vehicle.
Southampton car-following behavior data were collected on major roads with two or three lanes. In the speed range of 30 to 69km/h, traffic was quite congested in the morning peak when the field experiments were conducted. The drivers kept short headway distances in order to avoid lane changes of vehicles in front of them, leading to strong acceleration with opening inter-vehicle distances.
The road traffic environment in which the behavior data are collected is an important factor in the differences between car-following behavior in Southampton and that in Tsukuba, indicating that the road-traffic environment influences car-following behavior, regardless of the country of data collection. These findings imply that a single operational algorithm would suffice even when using vehicle control and driver support systems in different counties, although different algorithms would be necessary for different road types (e.g., roads in a city and roads connecting cities).
### **4. Case study 2: Longitudinal study of elderly drivers' car-following behavior**
### **4.1 Motivation**
This section introduces another case study focusing on the assessment of elderly drivers' carfollowing behavior, using the proposed fuzzy logic car-following model. The number of elderly drivers who drive their own passenger vehicles in their daily lives has increased annually. Driving a vehicle expands everyday activities and enriches the quality of life for the elderly. However, cognitive and physical functional changes of elderly drivers may lead to their increased involvement in traffic accidents. Thus, it is important to develop advanced driver assistance and support systems that promote safe driving for elderly drivers. Automatic vehicle control systems are expected to enhance comfort as well as safety when elderly drivers follow a vehicle. Understanding elderly drivers' car-following behavior is essential for developing automatic control systems that adapt to their usual car-following behavior.
Various studies comparing physical and cognitive functions between young and elderly drivers have been conducted in order to investigate the influence of age-related functional decline on driving (e.g., Owsley, 2004). Driving behavior is influenced by several driver characteristics (e.g., driving skill and driving style); and individual drivers' characteristics differ, especially between young and elderly drivers. Thus, a comparison of the driving behaviors of young and elderly drivers includes the influence of drivers' characteristics as well as the impact of the age-related decline of cognitive functions.
We have been involved in a cohort study on the driving behaviors of elderly drivers on an actual road (Sato & Akamatsu, 2011). A cohort study conducted in real road-traffic environments is expected to focus on changes in elderly drivers' cognitive functions because their cognitive functional changes may be greater than changes in their driving skills or driving styles within a few years. One aim of this study is to clarify how elderly drivers follow a lead vehicle, based on analysis of how car-following behavior changes with aging. We collected car-following behavior data of elderly drivers determined in one year and compared it with that determined five years later. The distributions of THW in the two field experiments were compared in order to investigate the static aspect of car-following behavior. For analysis of the dynamic aspect, the fuzzy logic car-following model was applied to compare elderly drivers' accelerative behavior while following a vehicle.
### **4.2 Methods**
276 Fuzzy Logic – Algorithms, Techniques and Implementations
Tsukuba car-following behavior data were collected on urban roads and a bypass. When driving on urban roads, a leading vehicle often has to decelerate suddenly due to other vehicles at crossroads, a change of traffic signals, and the emergence of pedestrians or bicycles. The leading vehicle might also slow down suddenly on the bypass because a merging car may cut in front of it. Drivers adopted longer headway distances and decelerated more strongly in closing inter-vehicle separations when driving on roads where
Southampton car-following behavior data were collected on major roads with two or three lanes. In the speed range of 30 to 69km/h, traffic was quite congested in the morning peak when the field experiments were conducted. The drivers kept short headway distances in order to avoid lane changes of vehicles in front of them, leading to strong acceleration with
The road traffic environment in which the behavior data are collected is an important factor in the differences between car-following behavior in Southampton and that in Tsukuba, indicating that the road-traffic environment influences car-following behavior, regardless of the country of data collection. These findings imply that a single operational algorithm would suffice even when using vehicle control and driver support systems in different counties, although different algorithms would be necessary for different road types (e.g.,
**4. Case study 2: Longitudinal study of elderly drivers' car-following behavior**
This section introduces another case study focusing on the assessment of elderly drivers' carfollowing behavior, using the proposed fuzzy logic car-following model. The number of elderly drivers who drive their own passenger vehicles in their daily lives has increased annually. Driving a vehicle expands everyday activities and enriches the quality of life for the elderly. However, cognitive and physical functional changes of elderly drivers may lead to their increased involvement in traffic accidents. Thus, it is important to develop advanced driver assistance and support systems that promote safe driving for elderly drivers. Automatic vehicle control systems are expected to enhance comfort as well as safety when elderly drivers follow a vehicle. Understanding elderly drivers' car-following behavior is essential for
developing automatic control systems that adapt to their usual car-following behavior.
well as the impact of the age-related decline of cognitive functions.
Various studies comparing physical and cognitive functions between young and elderly drivers have been conducted in order to investigate the influence of age-related functional decline on driving (e.g., Owsley, 2004). Driving behavior is influenced by several driver characteristics (e.g., driving skill and driving style); and individual drivers' characteristics differ, especially between young and elderly drivers. Thus, a comparison of the driving behaviors of young and elderly drivers includes the influence of drivers' characteristics as
We have been involved in a cohort study on the driving behaviors of elderly drivers on an actual road (Sato & Akamatsu, 2011). A cohort study conducted in real road-traffic environments is expected to focus on changes in elderly drivers' cognitive functions because their cognitive functional changes may be greater than changes in their driving skills or
they should pay more attention to the movements of the leading vehicle.
opening inter-vehicle distances.
**4.1 Motivation**
roads in a city and roads connecting cities).
### **4.2.1 Procedures**
Field experiments were conducted in 2003 (first experiment) and in 2008 (second experiment). The two experiments were conducted using the same instrumented vehicle, the same driving route, and the same participants. The AIST instrumented vehicle was used for the data collection (Fig. 3). Almost all the sensors and the driving recorder system were fixed inside the trunk, so that the participating drivers could not see them, in order to encourage natural driving behaviors during the experiment trials.
The experiments were conducted on rural roads around Tsukuba. The route included several left and right turns, and the travel time was 25min (total distance 14km). The participant rode alone in the instrumented vehicle during the experiment trials. Before the recorded drives, the participants performed practice drives from the starting point to the destination without using a map or an in-vehicle navigation system.
Four elderly drivers (three males and one female) with informed consent participated in the two experiments. Their ages ranged from 65 to 70 years (average 67.3 years) in the first experiment and from 70 to 74 years (average 72.0 years) in the second experiment. Their annual distance driven ranged from 5,000 to 8,000km in the first experiment and from 2,000 to 10,000km in the second experiment (average 6,500km in both experiments).
The participants were instructed to drive in their typical manner. In the first experiment, the recorded trip for each elderly participant was made once a day on weekdays, for a total of 10 trips. In the second experiment, the trial was conducted twice a day on weekdays, for a total of 30 trips. The participants took a break between the experiment trials in the second experiment.
### **4.2.2 Data analysis**
Figure 8 depicts a target section for the analysis of elderly drivers' car-following behavior. We focused on a two-lane bypass, the same road environment as that in section 3.2.2. We included only drives with a leading vehicle, excluding drives without a leading vehicle on the target section.
The active mode (distance between the instrumented vehicle and the leading vehicle) was used in the analysis of the elderly drivers' car-following behavior. We also used the passive mode (distance between the instrumented vehicle and the following vehicle) to investigate traffic characteristics on the analyzed road. The latter is expected to indicate whether changes in elderly drivers' car-following behaviors are influenced by their functional declines or by changes in traffic characteristics on the target road.
Understanding Driver Car-Following Behavior Using a Fuzzy Logic Car-Following Model 279
Figure 10 compares the relative velocity–acceleration mapping of the first and second experiments. In the fuzzy logic model specification, there is a total of 27 traces (data length 770.5sec) in the first experiment and 29 traces (data length 1481.0sec) in the second
The RMSEs between the predicted and measured accelerations in the estimated fuzzy logic model were 0.25m/sec2 in the first experiment and 0.14m/sec2 in the second experiment, which are within adequate errors compared to those estimated based on other real-world
The deceleration when the elderly participants approach the lead vehicle was the same in the two experiments. However, the elderly drivers accelerate more strongly in the second experiment than in the first experiment, when the leading vehicle goes faster and the
Acceleration (m/sec2)
*Strong acceleration*
*Deceleration*
Fig. 10. Results of fuzzy logic model specification of elderly drivers: Relative velocity–
Comparison of THW to following vehicles between the first and second experiments indicates no change in traffic flow on the target section in five years. In contrast, THW to leading vehicles is longer in the second experiment than in the first experiment, suggesting that elderly drivers take longer THW and the static aspect of their car-following behaviors
Second experiment
*Closing* Relative Velocity (m/sec) *Opening*
The task-capability interface model (Fuller, 2005) helps clarify why elderly drivers' carfollowing behavior changes with aging. In this model, drivers adjust task difficulty while driving in order to avoid road accidents. Task difficulty can be described as an interaction between the driver's capability and task demands. When the driver's capability exceeds the task demands, the task is easy and the driver completes the task successfully. When the task demands exceed the driver's capability, the task is difficult and a collision or loss of control occurs because the driver fails to accomplish the task. Here, the driver's capability is determined by the individual's physical and cognitive characteristics (e.g., vision, reaction time, and information processing capacity), personality, competence, skill, and driving style.
acceleration mapping between the first and second experiments
*Closing* Relative Velocity (m/sec) *Opening* First experiment
**4.3.2 Dynamic aspect**
data (Wu et al., 2003).
**4.4 Discussion**
changes over five years.
headway distance is opening.
*Strong acceleration*
*Deceleration*
Acceleration (m/sec2)
experiment.
Fig. 8. Road section used for car-following behavior analysis
### **4.3 Results**
### **4.3.1 Static aspect**
Figure 9 presents the distributions of THW for leading and following vehicles. The THW distributions suggest the proportion of time experienced in each category to the total time of the car-following conditions. There were no differences in the distribution of THW to following vehicles between the first and second experiments.
The peak of the distribution of THW to leading vehicles is found in the category from 1 to 1.5s in the first experiment. However, the peak is found in the category from 1.5 to 2s in the second experiment, indicating that the THW in the second experiment exceeds that in the first experiment.
Fig. 9. Comparison of THW to leading and following vehicles between the first and second experiments
### **4.3.2 Dynamic aspect**
278 Fuzzy Logic – Algorithms, Techniques and Implementations
Figure 9 presents the distributions of THW for leading and following vehicles. The THW distributions suggest the proportion of time experienced in each category to the total time of the car-following conditions. There were no differences in the distribution of THW to
From origin
To destination
The peak of the distribution of THW to leading vehicles is found in the category from 1 to 1.5s in the first experiment. However, the peak is found in the category from 1.5 to 2s in the second experiment, indicating that the THW in the second experiment exceeds that in the
First experiment Second experiment
Fig. 9. Comparison of THW to leading and following vehicles between the first and second
THW (s) to following vehicles THW (s) to leading vehicles
0% 5% 10% 15% 20% 25% 30% 35% 40%
0
0~0.5
0.5
~1
1~1.5
1.5
2~2.5
2.5
~3
3~3.5
3.5
~4
4~4.5
4.5
~5
5~
~2
5~
Fig. 8. Road section used for car-following behavior analysis
Distance:1.8km (about 2-min drive )
Target section for car-following behaviour analyses
following vehicles between the first and second experiments.
**4.3 Results**
**4.3.1 Static aspect**
first experiment.
experiments
0% 5% 10% 15% 20% 25% 30% 35% 40%
0
0~0.5
0.5
~1
1~1.5
1.5
2~2.5
2.5
~3
3~3.5
3.5
~4
Frequency rate Frequency rate
4~4.5
4.5
~5
~2
Figure 10 compares the relative velocity–acceleration mapping of the first and second experiments. In the fuzzy logic model specification, there is a total of 27 traces (data length 770.5sec) in the first experiment and 29 traces (data length 1481.0sec) in the second experiment.
The RMSEs between the predicted and measured accelerations in the estimated fuzzy logic model were 0.25m/sec2 in the first experiment and 0.14m/sec2 in the second experiment, which are within adequate errors compared to those estimated based on other real-world data (Wu et al., 2003).
The deceleration when the elderly participants approach the lead vehicle was the same in the two experiments. However, the elderly drivers accelerate more strongly in the second experiment than in the first experiment, when the leading vehicle goes faster and the headway distance is opening.
Fig. 10. Results of fuzzy logic model specification of elderly drivers: Relative velocity– acceleration mapping between the first and second experiments
### **4.4 Discussion**
Comparison of THW to following vehicles between the first and second experiments indicates no change in traffic flow on the target section in five years. In contrast, THW to leading vehicles is longer in the second experiment than in the first experiment, suggesting that elderly drivers take longer THW and the static aspect of their car-following behaviors changes over five years.
The task-capability interface model (Fuller, 2005) helps clarify why elderly drivers' carfollowing behavior changes with aging. In this model, drivers adjust task difficulty while driving in order to avoid road accidents. Task difficulty can be described as an interaction between the driver's capability and task demands. When the driver's capability exceeds the task demands, the task is easy and the driver completes the task successfully. When the task demands exceed the driver's capability, the task is difficult and a collision or loss of control occurs because the driver fails to accomplish the task. Here, the driver's capability is determined by the individual's physical and cognitive characteristics (e.g., vision, reaction time, and information processing capacity), personality, competence, skill, and driving style.
Understanding Driver Car-Following Behavior Using a Fuzzy Logic Car-Following Model 281
In the cross-cultural study, we compared the car-following behavior gathered on roads where driving is on the left side of the road. Further research will be addressed to compare the car-following behavior between left-hand driving and right-hand driving (e.g., in the
In the longitudinal study, we investigated the car-following behavior of small samples. The next step is to collect and analyze more elderly driver car-following behaviors to validate the findings of this study. Additionally, further study should be conducted to examine individual differences in car-following behaviors to clarify which cognitive function influences changes in car-following behavior with aging. We will assess the relationship between car-following behavior on a real road and elderly drivers' cognitive functions (e.g., attention, working memory, and planning (Kitajima & Toyota, 2012)) measured in a laboratory experiment. Analysis of the relationship between driving behavior and a driver's cognitive functions will help determine how driver support systems may assist driving behavior and detect the driver's cognitive functions based on natural driving behavior.
The authors are grateful to Prof. M. McDonald of the University of Southampton and Prof. P. Zheng of Ningbo University for useful discussions on estimation methodologies and
Brackstone, M., McDonald, M., & Sultan, B.; (1999). Dynamic behavioural data collection
Brackstone, M. & McDonald, M.; (1999). Car-following: a historical review. *Transportation Research Part F,* Vol.2, No.4, (December 1999), pp. 181-196, ISSN 1369-8478 Brackstone, M., Sultan, B., & McDonald, M.; (2002). Motorway driver behaviour: studies on
Brackstone, M., Waterson, B, & McDonald, M.; (2009). Determinants of following headway
Chandler, R.E., Herman, R. & Montroll, E.W.; (1958). Traffic dynamics: Studies in car
Fuller, R.; (2005). Towards a general theory of driver behaviour. *Accident Analysis and*
Gipps, P.G.; (1981). A behavioural car following model for computer simulation. *Transportation Research Part B,* Vol.15, No.2, (April 1981), pp. 105-111, ISSN 0191-2615 Helly, W.; (1959). Simulation of Bottlenecks in Single Lane Traffic Flow. *Proceedings of the Symposium of Theory of Traffic Flow,* pp. 207-238, New York, USA, 1959 Iwashita, Y., Ishibashi, M., Miura, Y., & Yamamoto, M.; (2011). Changes of Driver Behavior
*Prevention,* Vol.37, No.3, (May 2005), pp. 461-472, ISSN 0001-4575
*accident,* Tokyo, Japan, September 5-9, 2011
using an instrumented vehicle. *Transportation Research Record,* No.1689, (1999), pp.
car following. *Transportation Research Part F,* Vol.5, No.1, (March 2002), pp. 31-46,
in congested traffic. *Transportation Research Part F,* Vol.12, No.2, (March 2009), pp.
following. *Operations Research,* Vol.6, No.2, (March 1958), pp. 165-184, ISSN 0030-364X
by Rear-end Collision Prevention Support System in Poor Visibility. *Proceedings of First International Symposium on Future Active Safety Technology toward zero-traffic-*
United States).
**7. Acknowledgments**
**8. References**
results of the fuzzy logic car-following model.
9-17, ISSN 0361-1981
ISSN 1369-8478
131-142, ISSN 1369-8478
Task demands are determined by the operational features of the vehicle (e.g., its control characteristics), environmental factors (e.g., road surface and curve radii), interactions with other road users (e.g., slowing down of a lead vehicle and crossing of pedestrians or bicycles), and human factors (e.g., choice of driving speeds, headway distances, and acceleration control). The longitudinal assessment in this study is conducted using the same participant, the same instrumented vehicle, and the same route. These experiment settings lead to no differences in driver personality affecting capability or in vehicle operational features and road traffic environments influencing task demands. The decline in physical and cognitive functions may lead to a decrease in the elderly driver's capability. Therefore, elderly drivers reduce task demands by adopting longer THW to a leading vehicle, and they seek to maintain capability higher than the reduced task demands.
The results of the fuzzy logic car-following model estimation suggest that the acceleration rate when the inter-vehicle distance is opening becomes higher after five years, although the deceleration rate while approaching the vehicle in front does not change. The stronger acceleration may be a compensating behavior for maintaining the driver's capability by increasing the task demand temporarily, because the driver's capability interacts with the task demands, and drivers can control the task demands by changing their driving behavior in order to improve their capability (e.g., increasing speed, to wake up when feeling sleepy while driving).
Our findings imply that when a leading vehicle drives faster and the headway distances are opening while driving on multi-traffic lanes or while approaching a merging point, information or warning about the movements of the surrounding vehicles is helpful to elderly drivers because they accelerate more strongly and the temporal task demand is higher in this situation.
### **5. Limitations**
The fuzzy logic car-following model deals mainly with two vehicles: a vehicle in front and the driver's own vehicle. When drivers approach an intersection with a traffic light under car-following conditions, they may pay more attention to the signal in front of the leading vehicle and manage their acceleration based on the traffic light. Drivers allocate their attention to the forward road structure instead of the leading vehicle when they approach a tight curve; thus, they may reduce their driving speed before entering the curve even if the headway distance is opening. The car-following behavior before intersections or tight curves can be influenced by environmental factors other than a lead vehicle. Further car-following models should be developed to reproduce the car-following behavior in these situations.
### **6. Conclusion**
This chapter describes the fuzzy logic car-following model, including a comparison with other car-following models. We introduce two case studies that investigate drivers' carfollowing behavior using the fuzzy logic car-following model. This model can determine the degree to which a driver controls longitudinal acceleration according to the relationship between the preceding vehicle and his/her vehicle. The fuzzy logic model evaluates the driver's acceleration and deceleration rates using a rule base in natural language. This model contributes to interpretation of the difference in headway distances between Tsukuba and Southampton and changes in elderly drivers' headway distances with aging.
In the cross-cultural study, we compared the car-following behavior gathered on roads where driving is on the left side of the road. Further research will be addressed to compare the car-following behavior between left-hand driving and right-hand driving (e.g., in the United States).
In the longitudinal study, we investigated the car-following behavior of small samples. The next step is to collect and analyze more elderly driver car-following behaviors to validate the findings of this study. Additionally, further study should be conducted to examine individual differences in car-following behaviors to clarify which cognitive function influences changes in car-following behavior with aging. We will assess the relationship between car-following behavior on a real road and elderly drivers' cognitive functions (e.g., attention, working memory, and planning (Kitajima & Toyota, 2012)) measured in a laboratory experiment. Analysis of the relationship between driving behavior and a driver's cognitive functions will help determine how driver support systems may assist driving behavior and detect the driver's cognitive functions based on natural driving behavior.
### **7. Acknowledgments**
The authors are grateful to Prof. M. McDonald of the University of Southampton and Prof. P. Zheng of Ningbo University for useful discussions on estimation methodologies and results of the fuzzy logic car-following model.
### **8. References**
280 Fuzzy Logic – Algorithms, Techniques and Implementations
Task demands are determined by the operational features of the vehicle (e.g., its control characteristics), environmental factors (e.g., road surface and curve radii), interactions with other road users (e.g., slowing down of a lead vehicle and crossing of pedestrians or bicycles), and human factors (e.g., choice of driving speeds, headway distances, and acceleration control). The longitudinal assessment in this study is conducted using the same participant, the same instrumented vehicle, and the same route. These experiment settings lead to no differences in driver personality affecting capability or in vehicle operational features and road traffic environments influencing task demands. The decline in physical and cognitive functions may lead to a decrease in the elderly driver's capability. Therefore, elderly drivers reduce task demands by adopting longer THW to a leading vehicle, and they
The results of the fuzzy logic car-following model estimation suggest that the acceleration rate when the inter-vehicle distance is opening becomes higher after five years, although the deceleration rate while approaching the vehicle in front does not change. The stronger acceleration may be a compensating behavior for maintaining the driver's capability by increasing the task demand temporarily, because the driver's capability interacts with the task demands, and drivers can control the task demands by changing their driving behavior in order to improve their capability (e.g., increasing speed, to wake up when feeling sleepy
Our findings imply that when a leading vehicle drives faster and the headway distances are opening while driving on multi-traffic lanes or while approaching a merging point, information or warning about the movements of the surrounding vehicles is helpful to elderly drivers because they accelerate more strongly and the temporal task demand is
The fuzzy logic car-following model deals mainly with two vehicles: a vehicle in front and the driver's own vehicle. When drivers approach an intersection with a traffic light under car-following conditions, they may pay more attention to the signal in front of the leading vehicle and manage their acceleration based on the traffic light. Drivers allocate their attention to the forward road structure instead of the leading vehicle when they approach a tight curve; thus, they may reduce their driving speed before entering the curve even if the headway distance is opening. The car-following behavior before intersections or tight curves can be influenced by environmental factors other than a lead vehicle. Further car-following models should be developed to reproduce the car-following behavior in these situations.
This chapter describes the fuzzy logic car-following model, including a comparison with other car-following models. We introduce two case studies that investigate drivers' carfollowing behavior using the fuzzy logic car-following model. This model can determine the degree to which a driver controls longitudinal acceleration according to the relationship between the preceding vehicle and his/her vehicle. The fuzzy logic model evaluates the driver's acceleration and deceleration rates using a rule base in natural language. This model contributes to interpretation of the difference in headway distances between Tsukuba
and Southampton and changes in elderly drivers' headway distances with aging.
seek to maintain capability higher than the reduced task demands.
while driving).
**5. Limitations**
**6. Conclusion**
higher in this situation.
282 Fuzzy Logic – Algorithms, Techniques and Implementations
Kitagima, M. & Toyota, M.; (2012). Simulating navigation behaviour based on the
Kometani, E. & Sasaki, T.; (1959). Dynamic behaviour of traffic with a nonlinear spacing-
Mehmood, A., Saccomanno, F., & Hellinga, B.; (2001). Evaluation of a car-following model
Michaels, R.M.; (1963). Perceptual factors in car following. *Proceedings of the Second International Symposium on the Theory of Road Traffic Flow,* pp. 44-59, Paris, France, 1963 Owsley, C.; (2004). Driver capabilities. *Transportation in an Aging Society A Decade of*
Pipes, L.A.; (1953). An operational analysis of traffic dynamics. *Journal of Applied Physics,*
Reiter, U.; (1994). Empirical studies as basis for traffic flow models. *Proceedings of the Second International Symposium on Highway Capacity,* pp. 493-502, Sydney, Australia, 1994 Sato, T., & Akamatsu, M.; (2007). Influence of traffic conditions on driver behavior before
Sato, T., Akamatsu, M., Zheng, P., & McDonald, M.; (2009a). Comparison of Car-Following
Sato, T., Akamatsu, M., Zheng, P., & McDonald, M.; (2009b). Comparison of Car Following
Sato, T, & Akamatsu, M.; (2011). Longitudinal Study of Elderly Driver's Car-Following
Sugeno, M.; (1985). *Industrial Applications of Fuzzy Control,* Elsevier Science Inc., ISBN
Wu. J., Brackstone, M., & McDonald, M.; (2000). Fuzzy sets and systems for a motorway
Wu, J., Brackstone, M., & McDonald, M.; (2003). The validation of a microscopic simulation
Zheng, P.; (2003). *A microscopic simulation model of merging operation at motorway on-ramps,*
Zheng, P., McDonald, M., & Wu, J.; (2006). Evaluation of collision warning-collision
*Conference 2009,* pp. 4155-4160, Fukuoka, Japan, August 18-21, 2009
*Dynamics Society,* Atlanta, Georgia, USA, July 23-27, 2001
Vol.24, No.3, (March 1953), pp. 274-281, ISSN 0021-8979
1027-1652, Washington, D.C., USA, 2004
(September 2003), pp. 397-413, ISSN 1369-8478
41-58, ISSN 1362-3001
New York, USA, 1959
August 9-14, 2009
September 5-9, 2011
0444878297, New Yorkm USA
2000), pp. 65-76, ISSN 0165-0114
(December 2003), pp. 463-479, ISSN 0968-090X
No.1944, (2006), pp. 1-7, ISSN 0361-1981
PhD Thesis, University of Southampton, Southampton, UK
architecture model Model Human Processor with Real-Time Constraints (MHP/RT). *Behaviour & Information Technology,* Vol.31, No.1, (November 2011), pp.
speed relationhip. *Proceedings of the Symposium of Theory of Traffic Flow,* pp. 105-119,
using systems dynamics. *Proceedings of the 19th International Conference of the System*
*Experience (Transportation Research Board Conference Proceedings 27),* pp. 44-55, ISSN
making a right turn at an intersection: analysis of driver behavior based on measured data on an actual road. *Transportation Research Part F,* Vol.10, No.5,
Behavior between Four Countries from the Viewpoint of Static and Dynamic Aspects. *Proceedings of 17th World Congress on Ergonomics (IEA 2009),* Beijing, China,
Behavior between UK and Japan. *Proceedings of ICROS-SICE International Joint*
Behavior in Actual Road Environments. *Proceedings of First International Symposium on Future Active Safety Technology toward zero-traffic-accident,* Tokyo, Japan,
microscopic simulation model. *Fuzzy Sets and Systems,* Vol.116, No.1, (November
model : a methodological case study. *Transportation Research Part C,* Vol.11, No.6,
avoidance systems using empirical driving data. *Transportation Research Record,*
### *Edited by Elmer P. Dadios*
Fuzzy Logic is becoming an essential method of solving problems in all domains. It gives tremendous impact on the design of autonomous intelligent systems. The purpose of this book is to introduce Hybrid Algorithms, Techniques, and Implementations of Fuzzy Logic. The book consists of thirteen chapters highlighting models and principles of fuzzy logic and issues on its techniques and implementations. The intended readers of this book are engineers, researchers, and graduate students interested in fuzzy logic systems.
Fuzzy Logic - Algorithms, Techniques and Implementations
Fuzzy Logic
Algorithms, Techniques and Implementations
*Edited by Elmer P. Dadios*
Photo by Nattakit / iStock | doab | 2025-04-07T03:56:58.646130 | 20-4-2021 17:29 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/3.0/",
"book_id": "003cf94e-81c7-4214-b0fe-134420ebdba8",
"url": "https://mts.intechopen.com/storage/books/2274/authors_book/authors_book.pdf",
"author": "",
"title": "Fuzzy Logic",
"publisher": "IntechOpen",
"isbn": "9789535103936",
"section_idx": 3
} |
003dfb30-3d3b-432a-9c4e-217966fc6cb3 | We are IntechOpen, the world's leading publisher of Open Access books Built by scientists, for scientists
Open access books available 5,300
130,000 155M
International authors and editors
Downloads
Our authors are among the
most cited scientists 154 TOP 1%
Selection of our books indexed in the Book Citation Index in Web of Science™ Core Collection (BKCI)
# Interested in publishing with us? Contact [email protected]
Numbers displayed above are based on latest data collected. For more information visit www.intechopen.com
# **Electromagnetic Sensing Techniques for Non-Destructive Diagnosis of Civil Engineering Structures**
Massimo Bavusi et al. \* *CNR-IMAA, Italy*
# **1. Introduction**
Health Assessment Methods (HAM) and Structural Health Monitoring (SHM) aim to improve the standard of knowledge regarding the safety and maintenance of structures and infrastructure acquiring information about geometrical, mechanical and dynamical characteristics of structures. In earthquake-prone areas, this activity has the double aim of assessing the buildings structural integrity and extracting information regarding their response during a seismic event in order to define appropriate activities for risk mitigation. A number of factors afflict buildings and infrastructure safety in seismic areas:
The seismic assessment of structures is performed in terms of the estimation of the earthquake intensity that would lead to a certain damage condition and/or collapse. The assessment of the seismic vulnerability of existing buildings is generally based on the knowledge of building characteristics and through a complex analysis of the possible collapse mechanisms in order to identify the most probable failure for the given structure (as example: Ansari, 2005; Douglas, 2007; Moustafa et al., 2010).
The methodological approach for the evaluation of a structure resistance is represented in Figure 1 where structural knowledge obtained through a series of test assessments is needed in order to define vulnerability and thus design suitable retrofit strategies.
*<sup>3</sup>Basilicata University/DiSGG, Italy*
<sup>\*</sup>Romeo Bernini 2 , Vincenzo Lapenna<sup>1</sup> , Antonio Loperte 1 , Francesco Soldovieri 2 , Felice Carlo Ponzo<sup>3</sup> , Antonio Di Cesare 3 and Rocco Ditommaso<sup>3</sup>
*<sup>1</sup>CNR-IMAA, Italy*
*<sup>2</sup>CNR-IREA ,Italy*
Since the level of reliability of the assessment method is related to the adequacy of the model and to the completeness of the information, all useful available data have to be collected in order to define the original structural characteristics such as: geometry of structural elements, characteristics and behaviour of the construction materials, presence of degradation, arrangement of longitudinal and transversal reinforcement.
The knowledge of an existing structure is never complete and the level and accuracy of construction details obviously corresponds proportionally with the available original design documentation, the time and funds available for in situ investigations and experimental tests on the structural elements.
Fig. 1. Methodological approach.
A reliable assessment of the vulnerability of buildings is also strictly connected to the evaluation of the mechanical characteristics of the constitutive materials. This can be particularly complex for concrete, due to the high variability of its resistance that depends on intrinsic factors such as the composition as well as the environmental and maturing conditions, and other factors attributable to the collection technique and reworking of the concrete sample and the test conditions in general (Barlet, 1994).
A number of tests and methods have been developed for evaluating the resistance of construction materials ranging from completely non-destructive tests (NDT), where there is no damage to the structural element, using methods where the concrete surface is only slightly damaged, to partially destructive or destructive tests (DT), where the structural element has to be repaired afterwards.
The classical NDT methods, generally used for Reinforced concrete (R/C) structures, are the surface hardness method coupled with the ultrasonic method. As these methods are influenced in different and/or opposing ways by some fundamental parameters, their combined use allows outputs with minimal dispersion to be obtained. It is generally not advisable to use a single non-destructive test to estimate the strength in situ of concrete.
The range of properties that can be assessed using the range of NDT methods is significant and includes fundamental parameters such as density, elastic modulus as well as surface hardness, reinforcement location and depth of cover concrete. In some cases it is also possible to check the quality of workmanship and structural integrity through the ability to detect voids, cracking and delamination.
Preliminary tests can be performed with a covermeter according to the procedures described in the British Standard 188:204. With this technique it is possible to determine the presence and size of reinforcing bars, laps, transverse steel and depth and position of reinforcement. The identification of the position of the reinforcement bars is also used as a preliminary to the other NDT (such as Ultrasonic Pulse Velocity, Schmidt Rebound Hammer, Pull-out UNI EN 12504-3, 2005) and also the DT (Core Extraction and Compression Test). In addition the partially destructive method of removing the cover concrete in some areas and measuring directly the diameter and type of the reinforcement can be performed.
The Schmidt rebound hammer test is principally a surface hardness tester and is carried out according to UNI EN 12504-2 (2001). The system works on the principle that the rebound of an elastic mass depends on the hardness of the surface against which the mass impinges. There is little apparent theoretical relationship between the strength of concrete and the rebound number of the hammer, however within limits, experimental correlations are established between strength properties and the rebound number. All of this cannot be generalized and should be calibrated for each type of existing concrete, for example using the results of compression tests. In the some cases the results of the hammer tests, taken as a rebound average (Ir) is used individually to assess the homogeneity of the concrete.
The Ultrasonic test is carried out in compliance with UNI EN 12504-4 (2005) and is aimed at determining the propagation velocity of a mechanical vibration pulse in concrete. By measuring the pulse crossing time and the distance between the two probes, the apparent propagation velocity can be calculated. This value can differ from the real value when the elastic waves undergo deviations from the path identified by the conjunction line between the two probes (RILEM 1972). The factors that affect the ultrasonic test the most are linked to the concrete composition, environmental and test conditions.
When interpreting the NDT results, special attention is needed regarding the presence of possible anomalies which can negatively influence the experimental assessment of the in situ concrete mechanical characteristics. Such anomalies are generally characterized through an evident correlation of the experimental datum with either a physical parameter of reference (usually compressive strength from DT) or with respect to the trend shown by the data acquired in the same context of structural homogeneity. These anomalies usually arise from improper execution of the test or from the fact that the test is carried out in non-ideal conditions.
Direct measure of the compressive strength of concrete in a structure is provided by the Concrete Core Extraction and Compression method (DT). The process of obtaining core specimens and interpreting the strength test results is often affected by various factors that influence either the in-place strength of the concrete or the measured strength of the test specimen (UNI EN 12504-1, 2002). In spite of such disturbance factors, values measured in this way are certainly the most reliable possible. Furthermore errors can be reduced using the A.C.I. 214.4R-03 guidelines which summarize current practices for obtaining cores and interpreting core compressive strength test results.
Immediately after extraction, the core concrete is tested for carbonation (also called depassivation). Carbonation penetrates below the exposed surface of concrete extremely slowly. The significance of carbonation is that the usual protection of the reinforcing steel generally present in the concrete due to the alkaline conditions caused by the hydrated cement paste is neutralized. Thus, if the entirety of the cover concrete is carbonated, corrosion of the steel will occur if moisture and oxygen can infiltrate the section.
The necessary destruction of the test object usually makes DT methods more expensive, and these testing methods can also be inappropriate in many circumstances. Therefore the use of NDT plays a crucial role in ensuring an economical operation. A general proportion of 1 core to 4 non-destructive investigations is recommended. Both the results from the DT and NDT are then combined in order to estimate the in situ concrete strength using the SonReb method. This method is the principal combination of Schmidt Rebound Hammer with Ultrasonic Pulse Velocity used for quality control and strength estimation of in situ concrete (Braga, 1992).
Another group of NDT methods are the Dynamic identification tests which can be used in order to assess fundamental dynamic properties (frequencies and/or modal shapes) of the structure and indirectly estimate the Young's modulus of the material (Ponzo et al., 2010). All dynamic characteristics can be estimated using two different approaches: classical methodologies based on Fourier analysis (Ditommaso et al., 2010a) or innovative methodologies based on time-frequency and interferometric analyses (Ditommaso et al., 2010b; Picozzi et al., 2011). These latter analyses are also useful to detect possible structural damage occurred after an earthquake (Ditommaso et al., 2011).
The results of the above testing methods (both NDT and DT) are used to calibrate numerical models. These models can then be compared to the likely seismic loading thus providing the overall vulnerability of the structure being considered (Ponzo et al. 2011).
Even if the methods described above can ensure (when correctly applied) a high level of structural knowledge a number of issues remain to be addressed. In order to do this innovative technologies and new methods must be developed: reduce uncertainty regarding core extraction points, improve the detection of deflection and deficiencies, detect water infiltration, improve reinforcement information, improve the depth under investigation, and reduce time and cost.
Due to rapid and flexible execution, high spatial resolution and deep investigation depth, electromagnetic sensing techniques are a group of NDT methods which can achieve these objectives. They can direct the use of classical NDT and DT methods and reduce uncertainties, coring number and the survey cost. Furthermore, their contribution to the structural knowledge allows the adoption of lower safety coefficients (through the increase in available information thus minimising spread) and thus higher calculation resistances. This in turn reduces the extent and cost of the actions required for the improvement or seismic retrofit of structures, if needed.
The building and infrastructure diagnostics can be take advantage from the use of new NDT techniques enabling larger investigation depths, spatial resolution, void and defect detection capacity, low cost and fastness. Electromagnetic sensing techniques can be an useful tool in order to achieve these objectives. They are based on injection of a form of electromagnetic energy (electrical current, radiowave, microwave, light, etc.) into the surveyed object and gathering of returned signal in order to measure electromagnetic properties (resistivity, electrical permittivity), reconstruct the inner structure, detect embedded defects.
In table 1 an overview of the advantages and disadvantages of several techniques (both classic and innovative) is presented, in terms of cost, speed of procedure, non intrusivity, accuracy of data obtained and degree of correlation with actual values.
Nevertheless, applying electromagnetic sensing techniques to man made structures, some adaptation needs to fit stringent requirements in terms of exploration depth, spatial resolution and signal/noise ratio. In fact, commonly used building materials pose challenging issues in terms of electrode impedance, coupling antennas, survey modalities, tomographic reconstruction, sensor size.
Electromagnetic sensing techniques suitable for civil infrastructures and building diagnostic such as Ground Penetrating Radar (GPR) and Electrical Resistivity Tomography (ERT) are presented in this chapter. Concerning the GPR we focus the attention on the possibility to improve the imaging at low and radio/microwave frequencies by using novel inversion approaches such as the Microwave Tomography (MT).
Then, a novel distributed fiber optic sensor technology able to monitor strain and temperature variations, is described. Finally, we discuss about the real contribution provided by electromagnetic sensing techniques in the building and infrastructure monitoring.
Table 1. Advantages and disadvantages of different health assessment techniques.
# **2. Electromagnetic sensing techniques**
Electromagnetic sensing techniques use is now rather diffuse in several earth science fields such as geology, hydrogeology, seismology, glaciology, stratigraphy of urban areas study, polluted areas study, landslides characterization, etc. Few years ago, non intrusiveness and quickness of these techniques suggested their use for investigating buildings and civil engineering structures.
The migration of these techniques towards the engineering can be identified with the Microgeophysics where specific issues are the sensor miniaturizing, signal/noise ratio improvement, exploitation of all available free surfaces for energizing and acquiring signals (Cosentino et al., 2011).
Electromagnetic sensing techniques are an useful tool for the diagnostics of civil infrastructures, such as transport ones, in the framework of their static and dynamic behavior before, during and after a crisis event such as an earthquake.
In fact, they ensure a fast and not-intrusive diagnosis useful in the pre-event stage since a precautionary diagnosis of strategic buildings and transport infrastructure can be a critical issue in the seismic risk prevention.
Moreover, they can represent, during the crisis, a valid tool for the rapid damage mapping of the civil buildings and infrastructures (bridges, roads, dams, assessment) in order to have preliminary estimations of those safe for rescue forces. Then, a rapid damage assessment for private buildings enable a correct estimate of the damaged houses and resources to be committed. Finally, in the post-event stage, restoration interventions can be driven by the electromagnetic sensing techniques in order to minimize the costs and maximize the results. The electromagnetic sensing techniques provides information about investigated materials in terms of amplitude and phase of the gathered signals, in turn function of the electromagnetic properties of the materials.
Amongst the requirements of the infrastructure diagnostics there is certainly the determination of the structural element thickness, rebar diameters, fractures and defects detection, water content or moisture (as indicators of chemical reactions occurrence), strain.
Therefore, the information obtained by the electromagnetic sensing techniques have to be converted in order to provide information directly usable by the engineers.
Not all electromagnetic sensing techniques are suitable in becoming NDT techniques. Sensing technique selection have to keep into account a certain degree of electromagnetic noise immunity, high spatial resolution and a suitable sensor size. Other aspects to be kept into account are:
Ground Penetrating Radar (GPR) and Electrical Resistivity Tomography (ERT) have these qualities since they are active techniques providing the control on the injected signal, an adjustable spatial resolution and an useful sensor size. Moreover a number of processing codes and inversion routines are available allowing well interpretable images, although personnel with certain degree of experience is required in the data processing and interpretation.
Another class of sensors is the distributed ones, ensuring the availability of measurements along the entire envelop of the sensor. Among these, Fiber Optic Distributed Sensors based on Brillouin scattering phenomenon is a promising experimental technique able to provide field of temperature and strain along the fiber which can be a standard low cost telecommunication fiber. Unlike other fiber optic sensors, this technique permits the remote and spatially continuous monitoring of the structure in terms of temperature and strain with the resolution of some tens of centimetres.
In the following paragraphs we describe those three techniques providing examples of application and highlighting their strengths and limitations.
### **2.1 Ground penetrating radar**
Ground Penetrating Radar is an electromagnetic sensing technique based on the same operating principles of classical radars (Daniels, 2004). In fact, it works by emitting an electromagnetic signal (generally modulated pulses or continuous harmonic waves) into the ground or another natural or manmade object; the electromagnetic wave propagates through the opaque medium and when it impinges on a non-homogeneity of the electromagnetic properties, in terms of dielectric permittivity and electrical conductivity, a backscattered electromagnetic field arises. Such a backscattered field is then collected by the receiving antenna located at the air/opaque medium interface and undergoes a subsequent processing and visualization, usually as a 2D image (Figure 2).
#### Electromagnetic Sensing Techniques for Non-Destructive Diagnosis of Civil Engineering Structures 461
Fig. 2. GPR survey design (upper panel) and radargram (lower panel).
Spatial resolution and investigation depth of GPR method are strictly dependent by the central frequency of the used antenna. In fact, antennas with low and mid frequency (40 MHz - 750 MHz) provided high investigation depth (10 m - 3 m) associated to a relative low spatial resolution (2 m - 10 cm). On the contrary, high central frequency antennas (900 MHz - 2.5 GHz) provide low investigation depths (1 m - 10 cm) and high spatial resolution (5 cm - 0.5 cm). Since the physical size of the antennas decreases as the frequency increases, the requirement of miniaturizing the sensors is naturally achieved for the GPR.
As a consequence, the GPR technique is useful for the study of bedrock stratigraphy and cavity detection (Lazzari et al. 2006), groundwater and pollution (Chianese et al. 2006), metal and plastic pipelines such as cables in urban areas, archaeological finds (Bavusi et al. 2009) when low central frequency antennas are used.
On the contrary, when high central frequency antennas are used, the GPR, more properly named in this case Surface Penetrating Radar (SPR), can be considered a NDT technique (McCann and Forde 2001) providing precious information about the presence of "embedded" objects such as, reinforced rebars (Shaw et al., 2005; Che et al., 2009), but also embedded "defects" such as voids and, by using special antennas (Huston et al, 2000; Forest and Utsi, 2004; Utsi et al., 2008), fractures. Moreover, the GPR technique can contribute to determine the concrete moisture content (Shaari et al., 2004; Hugenschmidt and Loser, 2008;
GPR survey design is a crucial issue since it determines not only the possibility to detect the target (rebar, defect, water infiltration, ecc.), but also the format output in terms or 2D (cross section, time-slice, depth slice) or 3D data volume, kind of processing and difficult of interpretation.
A proper GPR survey design have to keep into account:
When rebars are searched, the most used method of acquisition requires a regular orthogonal survey grid with a proper spacing (a few centimeters) in order to have a suitable spatial resolution. Figure 3 shows a the survey design performed in order to check the continuity of longitudinal and transversal rebars and check the degree of success of the concrete restoration intervention based on epoxy resin injection (Bavusi et al., 2010a).
Fig. 3. GPR survey design and results on a beam of a school of L'Aquila damaged by the Abruzzo earthquake of 6th April 2009. a) beam; b) detail of damaged area with gridding; c) regular 4 cm square grid drawn on the beam; d) longitudinal processed radargram n.3.
A radargram has been gathered along each longitudinal and transversal survey line by using a 1500 MHz antenna provided by survey wheel. This survey design allows to select proper cross-sections and built a data volume. In fact, transversal radargrams offer a view of longitudinal rebars, while longitudinal radargrams are useful for visualizing transversal rebars. Then, the interpolation of all radargrams allows to built a data volume and selects more significant time-slices or depth-slices in order to have a plan view of all rebars (Figure 4).
#### Electromagnetic Sensing Techniques for Non-Destructive Diagnosis of Civil Engineering Structures 463
Fig. 4. a) data volume built by interpolating all radargrams gathered along the survey lines of figure 3c. b) slices extracted at several depths.
Then, GPR method provides very impressive and effective images of the inner of a reinforced concrete structure. However, the main limitation is that deeper rebar layer is not well detected due to scattering phenomena and attenuation losses producing in turn a loss in spatial definition in depth. Moreover the upper layer of rebars produces a strong disturbance on the rest of radargram.
In order to overcome this drawback, several strategies can be applied:
First strategy can be effective, but increases the time consuming. The second one exploits the property of cross-polarized radargrams which are less sensitive to the rebars normal to the survey direction and more sensitive to the rebars parallel to the survey direction (Figure 5).
Fig. 5. a) Effect of the antenna polarization on the rebar reflection intensity. b) comparison between the normal-polarized and cross-polarized radargram gathered on the longitudinal survey lines n.9 of figure 2c.
Despite of the above mentioned advantages of GPR, one of the obstacles to its use regards the "low interpretability" of the radargram; therefore a an understandable "interpretation and visualization" of the investigated scene entails a high level operator's expertise and often a-priori information is required. This difficulty of the interpretation is further on affected in the case that no a priori information is available as, for example, it often happens in the case of historical heritage (Masini et al., 2010) where a lack of knowledge about the constructive modalities and materials of the structure can arise.
Therefore, a GPR data processing is often necessary to achieve more easily interpretable and reliable reconstructions of the scene, i.e. images that be easily understandable even by a not expert operator.
The usual radaristic approaches are based on migration procedures that essentially aim at reconstructing buried scattering objects from measurements collected above or just at the air/soil interface. These approaches were first based on graphical methods (Hagendoorn, 1954) based on high frequency assumption of the electromagnetic propagation and scattering; afterward this approach found a more consistent mathematical background based on the wave equation of the electromagnetic scattering (Stolt, 1978).
The absence of reflection in the concrete corresponding to the restored fracture indicates the success of the epoxy injection which filled all possible voids. Finally, a retrofit reinforcing intervention can be designed on the bases of the existent rebar arrangement.
Recently, new data processing based on the inverse scattering problem have been developed and implemented also in realistic situations for infrastructure monitoring (Catapano et al., 2006; Soldovieri and Orlando, 2009, Bavusi et al., 2011). In particular, the microwave tomography approaches have arose as the most suitable ones for the on field exploitation (Soldovieri and Solimene, 2010; Persico et al., 2005).
Such a class of approaches is based on the modeling of the electromagnetic scattering phenomena. According to this modelization, the imaging problem is cast as an inverse scattering problem where one attempts to infer the electromagnetic properties of the scattering object starting from the scattered field measured somewhere outside it.
The statement of the problem is then the following: given an incident field, Einc , which is the electromagnetic field existing in the whole space (the background medium) in absence of the scattering object and is generated by a transmitting antenna, by the interaction of the incident field with the embedded objects the scattered field E<sup>S</sup> arises; from the knowledge of the scattered field E<sup>S</sup> properties about the scattering targets, either geometrical and/or structural, have to be retrieved. The mathematical equations subtending the scattering phenomena to solve the above stated problem are in order.
To this end, we refer to a two-dimensional and scalar geometry. We consider a cylindrical dielectric object (i.e. invariant along the axis out-coming from the sheet) enclosed within the domain D illuminated by an incident field linearly polarized along the axis of invariance. The scattered field is observed over the domain (not necessarily rectilinear). Moreover, we denote by (r) e by <sup>b</sup> (r) the permittivity profile of the unknown object and of the background medium, respectively. In particular, the latter is not necessarily constant (i.e., a non-homogeneous background medium is allowed too) but has to be known. The magnetic permeability is assumed equal to that of the free space 0 everywhere. The geometry of the problem is detailed in Figure 6.
Fig. 6. Geometry of the subsurface prospecting problem
The problem, thus, amounts to retrieving the dielectric permittivity profile (r) of the unknown object(s) from the knowledge of the scattered field E<sup>S</sup> . The physical phenomenon is governed by the two equations (Chew, 1995)
$$\begin{aligned} \mathbf{E}(\underline{\mathbf{r}}, \underline{\mathbf{r}}\_{\mathcal{S}}; \mathbf{k}\_{\mathbf{b}}) &= \mathbf{E}\_{\mathrm{inc}}(\underline{\mathbf{r}}, \underline{\mathbf{r}}\_{\mathcal{S}}; \mathbf{k}\_{\mathbf{b}}) + \mathbf{k}\_{\mathbf{b}}^{2} \int\_{\mathcal{D}} \mathbf{G}(\underline{\mathbf{r}}, \underline{\mathbf{r}}; \mathbf{k}\_{\mathbf{b}}) \mathbf{E}(\underline{\mathbf{r}}, \underline{\mathbf{r}}\_{\mathcal{S}}; \mathbf{k}\_{\mathbf{b}}) \mathbf{z}(\underline{\mathbf{r}}) d\underline{\mathbf{r}} & \underline{\mathbf{r}} \in \mathbf{D} \\ \mathbf{E}\_{\mathrm{S}}(\underline{\mathbf{r}}\_{\mathcal{O}}, \underline{\mathbf{r}}\_{\mathcal{S}}; \mathbf{k}\_{\mathbf{b}}) &= \mathbf{k}\_{\mathbf{b}}^{2} \int\_{\mathcal{D}} \mathbf{G}(\underline{\mathbf{r}}\_{\mathcal{O}}, \underline{\mathbf{r}}; \mathbf{k}\_{\mathbf{b}}) \mathbf{E}(\underline{\mathbf{r}}, \underline{\mathbf{r}}\_{\mathcal{S}}; \mathbf{k}\_{\mathbf{b}}) \mathbf{z}(\underline{\mathbf{r}}) d\underline{\mathbf{r}} & \underline{\mathbf{r}}\_{\mathcal{O}} \in \Sigma \\ \end{aligned} \tag{1}$$
where EE E inc S is the total field, k<sup>b</sup> is the subsurface (background) wave-number and b (r) (r)/ 1 is the dimensionless contrast function. G( , ) is the pertinent Green's function (Leone and Soldovieri, 2003), <sup>O</sup> r is the observation point and S r is the position of the source.
In accordance to the volumetric equivalence theorem (Harrington, 1961), the above integral formulation permits to interpret the scattered field as being radiated by secondary sources (the "polarization currents") which are just located within the space occupied by the targets. The reconstruction problem thus consists of inverting the "system of equations (1)" versus the contrast function. However, since (from the first of the equations 1) the field inside the buried targets depends on the unknown contrast function, the relationship between the contrast function and the scattered field is nonlinear. However, the problem can be cast as a linear one if the first line equation is arrested at the first term of its Neumann expansion. After doing this EE inc is assumed within the targets and the so-called Born linear model is obtained (Chew, 1995). Accordingly, the scattering model becomes
$$\mathrm{E}\_{\mathrm{S}}(\underline{\mathbf{r}}\_{\mathrm{O}}, \underline{\mathbf{r}}\_{\mathrm{S}}; \mathrm{k}\_{\mathrm{b}}) = \mathrm{k}\_{\mathrm{b}}^{2} \int\_{\mathrm{D}} \mathrm{G}(\underline{\mathbf{r}}\_{\mathrm{O}}, \underline{\mathbf{r}}; \mathrm{k}\_{\mathrm{b}}) \mathrm{E}\_{\mathrm{inc}}(\underline{\mathbf{r}}, \underline{\mathbf{r}}\_{\mathrm{S}}; \mathrm{k}\_{\mathrm{b}}) \chi(\underline{\mathbf{r}}) d\underline{\mathbf{r}} \tag{2}$$
Let us just remark that, within the linear approximation, the internal field does not depend on the dielectric profile, which is the same as to say that mutual interactions between different parts of any object or between different objects are neglected. In other words, this means to consider each part of the object as an elementary scatterer that does not depend on the presence of the other scatterers.
Consequently, at this point the problem can be cast as the inversion of the linear integral equation (2) and the numerical implementation of the solution algorithm requires the discretization of eq. (2). This task is pursued by resorting to the method of moments (MoM) (Harrington, 1961).
One of the main feature of a GPR data is its ability to provide images of the inner structure of a building or infrastructure at all useful observation scales by exploiting antennas with several central frequencies. Concerning this, the bridge inspection, which is normally carried out by using classical DT and NDT methods, can derive benefit from the GPR technique (Scott et al., 2003; Hugenschmidt and Mastrangelo, 2006). Structural particulars of interest are inner rebars, tendons, boundary conditions, anchors, saddles and other internal elements. On the other hand, the observation can involve the entire deck of a bridge. Such observations became crucial when all original project documentations are partially or completely lost.
Figure 7 shows the survey design drown for the deck survey of the deck of Musmeci bridge in Potenza (Basilicata Region, Southern Italy) (Bavusi et al., 2011).
Electromagnetic Sensing Techniques for Non-Destructive Diagnosis of Civil Engineering Structures 467
Fig. 7. a) location of the Musmeci bridge in Potenza, Basilicata Region, Italy. b) location of the survey area respect to the bridge; d) surveyed lane; e) survey grid. From Bavusi M., Soldovieri F, Di Napoli R., Loperte A., Di Cesare A., Ponzo F.C. and Lapenna V (2010).Ground Penetrating Radar and Microwave Tomography 3D applications for the deck evaluation of the Musmeci bridge (Potenza, Italy). J. Geophys. Eng. 8 (2011) 1–14. Courtesy of IOP Publishing Ltd.
A such survey design is able to provide several depth-slices in order to observe the asphalt layer, the sects layer and the lower reinforced concrete plate of the deck. Figure 8 shows more significant depth-slices showing the deck structure at several depths.
Fig. 8. a-d) depth-slices built at 6 cm (a), 20 cm (b), 40 cm (c) and 50 cm (d) by interpolating all radargrams gathered on the survey lines of figure 5e; d-f) depth-slices built at 6 cm (d), 20 cm (e) and 40 cm (f), by interpolating the same radargrams inverted by means of the Microwave Tomography. ab: absorption zone, Gs: Gerber saddle, ps: pillar support, sw: stiffening wall, ls: longitudinal stiffening wall. From Bavusi M., Soldovieri F, Di Napoli R., Loperte A., Di Cesare A., Ponzo F.C. and Lapenna V (2010).Ground Penetrating Radar and Microwave Tomography 3D applications for the deck evaluation of the Musmeci bridge (Potenza, Italy). J. Geophys. Eng. 8 (2011) 1–14. Courtesy of IOP Publishing Ltd.
In particular the asphalt layer shows concentrated absorptive zones that can be related to water infiltration zones. In this case the precise positioning of traces is a critical issue since it can produce a staggering effect when rectilinear features are detected. The use of a survey wheel is mandatory such as a certain degree of care in the cart dragging in order to limit mispositioning errors. In this way a residual error can be subsequently reduced by using proper algorithms.
Among possible defects afflicting buildings and infrastructures, fractures are a very warring problem. In fact, fractures can be due to several causes: temperature (3 mm), dry up (0.4 mm), load (0.4 mm). They can involve loss of mechanical strength and represent a preferential way for the water infiltration which in turn can favour the developin g of chemical reactions (expanding salt crystallization, oxidation, carbonation, etc.).
Crack detection is an important issue in the field of non-destructive testing. Several techniques can be employed in order to check, localize and characterize fractures in manmade buildings: ultrasonic shear waves (De La Haza et al. 2008), elastic waves (Ohtsu et al. 2008), GPR (Utsi et al. 2008).
Due to their small size and variable orientation, fractures detection represents a very challenge for the GPR technique. In fact, the crack detection requires the exploitation of all spatial resolution available in the frequency range used. Moreover it requires fracture to be surveyed is filled by air, water or a material different from the host medium in order to produce a backscattered field (Grandjean and Gourry 1996). In addition, the geometry of the fracture with respect to the survey line plays a fundamental role (Tsoflias et al. 2004). For a vertically oriented fracture, a reflection hyperbola arises due to the bottom of the fracture and to each change in the direction of the fracture with respect to the vertical path (Forest and Utsi 2004). By exploiting this property, it is theoretically possible to detect a fracture by using a common GPR dipole antenna, even the use of specifically designed high vertical resolution antennas is very helpful (Forest and Utsi 2004; Utsi et al. 2008). Moreover, data processing plays a fundamental role to improve the 'imaging' and the focusing of the buried reflectors (Grandjean and Gourry 1996; Leucci et al. 2007).
Figure 9 shows a 1500 MHz GPR survey carried out on a fracture in the floor of the Prefecture of Chania (Crete Island, Greece) (Bavusi et al., 2010b).
Fracture zone, located at the middle point of the radargram, is detected by using a classical processing approach, but best performances in terms of spatial resolution can be obtained by using the Microwave Tomography.
## **2.2 Electrical resistivity tomography**
Electrical resistivity tomography (ERT) is an electromagnetic sensing technique used to obtain 2D and 3D images in terms of electrical resistivity of areas of complex geology (Griffiths and Baker 1993), landslides, watertable, basins, faults.
Technically, during an electrical resistivity measurement, the electric current is injected into the ground via two 30-40 cm × 1.5 cm steel electrodes and the resulting electrical voltage is measured between two other electrodes in line with current electrodes (Sharma, 1997). ERT can be carried out by using different electrode configurations such as dipole-dipole, Wenner, Schulumberger, pole-dipole, etc. (Figure 10).
At present, such configurations can be carried out by using multi-electrode systems enabling the automatic switch of all available electrodes previously fixed into the ground. The system manages the current injection and simultaneous potential measurements which can occur simultaneously at more potential electrodes in case of multichannel systems.
#### Electromagnetic Sensing Techniques for Non-Destructive Diagnosis of Civil Engineering Structures 469
Fig. 9. a) GPR survey design carried out at the Prefecture of Chania (Crete, Greece) on a fracture in the floor. Fracture zone is at the middle point of the radargrams. b) processed radargram; c) Microwave Tomography From Bavusi M., Soldovieri F., Piscitelli S., Loperte A., Vallianatos F. and Soupios P. (2010). Ground-penetrating radar and microwave tomography to evaluate the crack and joint geometry in historical buildings: some examples from Chania, Crete, Greece. Near Surface Geophysics, Vol.8, No. 5, pp. 377-387. Courtesy of EAGE Publications BV..
Fig. 10. a) Dipole-dipole array for the acquisition of a measure of electrical resistivity.
The result of an ERT survey is a distribution 2D or 3D of apparent resistivity where each data point is defined by two coordinates (x and z) depending on the position of the quadrupole (couple of current and potential dipoles) used and a value of apparent resistivity. Then, in order to reconstruct real resistivity distribution, an inversion routine is required. A number of algorithms are available in order to perform this reconstruction such as Res2DInv (Loke and Barker, 1996) for the automatic 2D inversion of apparent resistivity data was used. The inversion routine is based on the smoothness constrained least-squares inversion (Sasaki, 1992) implemented by a quasi-Newton optimization technique.
ERT surveys have been successfully applied in geology for stratigraphy and cavity detection (Lazzari et al., 2010), fault characterization (Caputo et al., 2007), landslide studies (Lapenna et al., 2005), in hydrogeology, in environmental problems for contaminant plume detection and waste dump characterization (Bavusi et al., 2006), for hydrogeology and coastal salt water intrusion detection (Satriani et al., 2011a), in agricultural for the root-zone characterization (Al Hagrey, 2007; Satriani et al, 2011b), in archaeology and cultural heritage studies (Bavusi et al., 2009). Figure 11 shows an example of ERT carried out on a piling in an area subjected to landslides. It is well visible the effect of the structure on the water distribution.
Fig. 11. Example of ERT carried out on a piling in an area subjected to landslides.
The ERT exhibits significant potentialities in terms of high resolution and flexibility of the investigation depth that can be varied in a simple way, by varying the electrode spacing. A large electrode spacing provides a high investigation dept and a low spatial resolution. On the contrary, a small electrode spacing allows to achieve a great spatial resolution but a low investigation depth.
This characteristic makes the ERT a candidate for the structure and infrastructure monitoring, even though some problems have to be enfaced. First, a structure or infrastructure survey requires an electrode spacing ranging between one centimeter and some decimeter, then sensors have to miniaturized in order to respect the assumption of geophysics stating transducers have to be punctual (i.e. small respect to dimension of the investigated volume) and reduce modeling errors (Cosentino et al. 2011; Athanasiou et al., 2007). Then, a low contact resistance have to be ensured in order to put an adequate current injection (Cosentino et al. 2011).
In order to meet these requirements, several devices can be used such as Cu flat-base electrodes with conductive gel (Athanasiou et al., 2007), Ag/AgCl medical electrodes and nails (Cosentino et al 2011), Cu-CuSO electrodes (Seppänen et al., 2009). Main limitation of these devices is the difficulty of put them on a vertical or steeply slope surface, even worse under a ceiling. Moreover, medical electrodes are not stable in time (Cosentino et al 2011), while flat-base electrodes are not suitable for the asphalt, where the only possibility to apply the ERT is making holes in order to put the electrode in the substratum.
In spite of these limitations, the ERT has been successfully applied on masonry, floors, artifacts in order to detect fractures, voids, previous restoration works, structural particulars, moisture. The application to reinforced concrete is possible in order to detect those targets and rebars, but experiments demonstrates target detection capacity in simple geometrical configurations (Seppänen et al., 2009; Karhunen et al., 2009).
In presence of complex rebar configurations such as reinforcement cages, the potential field produced by the injected current suffers a warping due to the circuit represented by the cage not easily modelable. A new class of inversion routines is then required in order to solve this problem. This technique appears still not adequate to be applied to structures and infrastructures, but technological development can provide technical solutions able to mitigate and overcame described limitations.
# **2.3 Distributed fiber optic sensors**
Typically standard NDT systems are based on the use of point sensors, however the Structural Health Monitoring (SHM) of large civil infrastructures like bridges, dams could require a large number of sensors. For these applications, there is an increasing interest towards the use of distributed optical fiber sensors. In these sensors is the optical fiber itself that acts as a sensor providing measurements all along the fiber. This approach permits to monitoring the whole structure by use of a single optical fiber avoiding the need of a huge amount of measurement points and lead to the comprehension of the real static behavior of the structure rather than a limited number of sensors. Furthermore, distributed sensors could play a fundamental role in civil engineering because no other tools allow the detection of local phenomena whose location is impossible to be predicted "a priori" like, for instance, for crack detection. Distributed fiber optic sensors are substantially different from other fiber optic sensors technologies being based on optical scattering mechanism (Raleigh, Raman, Brillouin) occurring during light propagation in common telecom optical fibres.
Spatial resolution is typically achieved by using the optical time domain reflectometry (OTDR) (Barnoski 1976), in which optical pulses are launched into an optical fiber and consequent variations in backscattering intensity is detected as a function of time. Alternative detection techniques such as frequency-domain approaches have been also demonstrated. Raleigh scattering based sensors were first developed, in order to locate fiber breaks or bad splices along a fiber link. However Rayleigh backscatter in standard fibers gives information only about optical attenuation, and it can not be related to other parameters such as temperature or strain. Distributed temperature sensing was first demonstrated by Hartog and Payne (1982), who used temperature-induced variation of the Rayleigh scattering coefficient along the length of liquid-core, but the low reliability of liquid-core fibers may restrict their use. A recent, very interesting approach makes use of the very high spatial resolution allowed by sweptwavelength interferometry, in order to correlate temperature and strain of the fiber with the spectrum of the Rayleigh backscatter spatial fluctuations (Measures 2001). This approach requires standard telecommunication fibers and very high spatial resolution (a few millimeters) has been demonstrated. On the other hand, the main disadvantages are the equipment cost (a tunable laser source is needed for the measurements), and the limited sensing length (70m). Dakin et al. (1985) demonstrated temperature profiles measurement using the variations in the Raman backscattering coefficients of anti-Stokes and Stokes light. The Raman approaches are very practical because conventional silica-based optical fibers can be used as the sensor. The anti-Stokes-Raman-backscattered light is about 30 dB weaker than the Rayleigh-backscattered light. However, its sensitivity to temperature is great. Therefore, systems based on Raman scattering have been commercialized by several manufacturers. Nevertheless, Raman scattering based systems do not allow performing deformation measurements. In 1989, it was reported that the frequency shift of the Stokes-Brillouinbackscattered light (the so-called Brillouin frequency shift) greatly varies with strain and temperature along the fiber (Horiguchi 1989, Culverhouse 1989). Since then, considerable attention has been paid to exploiting Brillouin scattering for distributed sensing. This is for the following reasons. First, strain is a very important parameter in the monitoring of the integrity of civil structures. Secondly, unlike the Raman technique, Brillouin frequency shift measurement does not require calibration of the optical-fiber loss. Furthermore, a very attractive feature of Brillouin-based sensors stems from the use of a standard telecommunications-grade optical fiber as the sensor head. The low-cost and low-loss nature of the sensor make possible to perform distributed measurements over distances of many kilometers. Finally, the tremendous developments in the optical telecommunications market have reduced considerably the cost and increased the performances of optical fibers and their associated optical components.
Distributed optical fiber sensors based on stimulated Brillouin scattering (SBS) rely on the interaction between two lightwaves and an acoustic wave in the optical fiber. The measurement principle is based on the characteristic that the Brillouin frequency of the optical fiber is shifted when strain as well as temperature changes occur. Spatial information along the length of the fiber can be obtained through Brillouin optical time domain analysis (BOTDA) by measuring propagation times for light pulses travelling in the fiber. This allows for continuous distributions of the parameter to be monitored.
The research in the beginning of distributed fiber optic strain sensing was mostly based on laboratory applications (Bernini 2005, Bernini 2008) only in last years in-field demonstration by a fully distributed sensor have been previously reported (Komatsu 2002).
About bridge structures recently the applications and the validation of distributed strain sensor during load test has been demonstrated (Matta 2008, Minardo 2011). Other examples are the installation of a distributed fiber optic strain sensing cable into the inspection gallery of a dam (Inaudi and Glisic 2005, Glisic and Inaudi, 2007) or the monitoring of extra-long tunnel, running over 150 km of seafloor geologic body with complicated topographic and geologic units (Shi, 2003).
However, the use of distributed fiber optic sensors for crack detection in concrete are rare. This is mainly due to the fact that the instrumentations available for in-field application have a limited spatial resolution (1m) (Deif 2010). In fact, distributed sensors measure the with average strain at each measurement point, where the strain is averaged over the length called spatial resolution. Today, new methods for distributed fiber optic strain sensing with sub-meter spatial resolution are being developed in order to increase the opportunities in NDT of civil structures especially for crack detection (Hotate 2002, Zou 2005, Bernini 2007).
As an example, we report the results obtained in a load test on a road-bridge (Minardo 2011). In particular, the tests were performed by an stimulate Brilouin Scattering portable sensor prototype with 3m-spatial resolution. The fiber employed for the measurements was a PVC-coated, single-mode, standard telecom optical fiber. The fiber was bonded along the lower flange of a 44-m-long, double-T steel beam, by use of a epoxy adhesive. Strain measurements were performed while loading the bridge with an increasing weight by use of gravel-loaded trucks (Figure 12a). During the loading test, data were also collected by other instruments for a cross-correlation. In particular, two vibrating wire (VW) strain gauges were previously spot-welded to the surface of the steel beam, so as to provide the strain at the quarter and the middle section of the loaded beam. Figure 10b depicts the results of the optical fiber measurements, for different load conditions. In particular, the solid lines refer to the bridge loaded by two, four, and five gravel trucks, respectively. Each truck had a weight of approximately 47 tons. The same figure also reports the results of a finite-element-method (FEM)-based numerical analysis (circles). Numerical data were obtained by modelling each gravel truck as three concentrated loads applied in correspondence of the three truck axles. A good agreement exists between the experimental and numerical data. The standard deviation was always less than 20 μ, corresponding to the nominal accuracy of the instrument. Moreover, the maximum strain provided by the optical fiber sensor ( 350 μ), is in good agreement with the value provided by the strain gauge placed at the middle beam section (330 μ). Another interesting feature is that the optical fiber sensor was able to reveal the right-shift of the center of gravity (CG), when loading the beam with five trucks. Actually, the fifth truck was not disposed symmetrically with respect to the middle of the bridge; rather it was closer to the right side. As a consequence, the section at which maximum strain occurs is shifted to the right in the final load test. Finally, in figure 12b are also reported, for comparison, the data obtained by the strain gauges (squares). As can be observed a good agreement between the two different observations is achieved.
Fig. 12. a) bridge used for the load test. b) Distributed strain measurement along the girder, for different load conditions (solid lines). FEM simulations (open circles). VW strain gauges (squares).
# **3. Conclusion**
Health monitoring of civil engineering structures is devoted to assess the structural integrity and dynamic behaviour during a seismic event in order to define appropriate activities for risk mitigation. This need is more stringent for aged infrastructures built following outdated codes of practice and where the chemical degradation of concrete and irons acted for more time.
The design of suitable retrofit strategies requires a series of test assessments in order to define the vulnerability. A number of destructive (DT) and non destructive tests (NTD) and methods has been developed and applied around the world in order to achieve all necessary information about construction material. An appropriate survey have to include a suitable proportion of DT and NDT methods in order to reduce possible damage to investigated structure, achieve a balanced combination of punctual and distributed information and reduce the global cost of the survey.
Then, the requirement of new non invasive technique is a stringent need that can be satisfied by the electromagnetic sensing techniques, this class of geophysical techniques which can be easily adapted to the specific requirements of civil infrastructures. This is the case of the Ground Penetrating Radar which at high frequencies provides the needed spatial resolution and a convenient small size of the antenna. Moreover, this technique can benefit on a new kind of processing based on the Microwave Tomography (MT) inversion. In this way the technique can focus small defects such as fractures and voids, detect rebars with great precision even if deep. The combination of GPR and MT will provide in future a new class of devices able to supply a not focused image ready for use by non expert users.
The Electrical Resistivity Tomography can be another suitable technique useful to depict embedded structural particulars and defects, but its systematic application requires to solve two problems. First is the design of appropriate non intrusive electrodes, stable in time and easy to install in any position. A solution can be provided by medical industries that have experience in the design of electrodes for human body applications. The latter is the lack of inversion routines able to model the effect of a rebar cage on the potential field and properly reconstruct the inner of a reinforced concrete structure. Anyway the ERT is successfully applied on floors, pavements, masonry and reinforced concrete structures having a simple inner arrangement. Moreover, the industry will provide in future new sensors, equipments and inversion routines able to mitigate or solve described problems.
Another class of sensors which will change the way of monitor a civil infrastructure is the distributed fiber optic sensor able to provide temperature and strain information along the fiber. This sensor is based on standard telecom fiber optic which is inexpensive and allows to design sensor sized on the infrastructure to be surveyed or on a particular. It allows to achieve in real time information by several points of a distributed civil infrastructure such as a railway, aqueduct, gas or oil pipeline without using transmission devices. In future optic fibers will be embedded in construction materials allowing a for life monitoring of an infrastructure. Moreover, a possible technological improvement of this technique can allow in future its application to the ambient vibration monitoring (Wenzel and Pichler, 2005).
# **4. Acknowledgment**
The research leading to these results concerning the Musmeci bridge has received funding from the European Community's Seventh Framework Programme (FP7/2007-2013) under Grant Agreement n° 225663 Joint Call FP7-ICT-SEC-2007-1.
Moreover, the authors would like to thank the "Soprintendenza per i Beni Architettonici e Paesaggistici della Basilicata" and the "Direzione Regionale per i beni Culturali e Paesaggistici della Basilicata" that have partially funded the activities. Finally, the authors would like to thank the owner of the Musmeci Bridge, ASI Consortium, for their interest to these research activities and the Municipality of Potenza for the granted authorization to access and work on the structure. Furthermore, the authors would like to thank EAGE Publications BV and IOP Publishing Ltd for copyright permissions granted.
Finally, the authors would like to thank TeRN Consortium for supporting this work.
# **5. References**
**Earthquake-Resistant Structures - Design, Assessment and Rehabilitation** Edited by Prof. Abbas Moustafa
ISBN 978-953-51-0123-9 Hard cover, 524 pages **Publisher** InTech **Published online** 29, February, 2012 **Published in print edition** February, 2012
This book deals with earthquake-resistant structures, such as, buildings, bridges and liquid storage tanks. It contains twenty chapters covering several interesting research topics written by researchers and experts in the field of earthquake engineering. The book covers seismic-resistance design of masonry and reinforced concrete structures to be constructed as well as safety assessment, strengthening and rehabilitation of existing structures against earthquake loads. It also includes three chapters on electromagnetic sensing techniques for health assessment of structures, post earthquake assessment of steel buildings in fire environment and response of underground pipes to blast loads. The book provides the state-of-the-art on recent progress in earthquake-resistant structures. It should be useful to graduate students, researchers and practicing structural engineers.
### **How to reference**
In order to correctly reference this scholarly work, feel free to copy and paste the following:
Massimo Bavusi, Romeo Bernini, Vincenzo Lapenna, Antonio Loperte, Francesco Soldovieri, Felice Carlo Ponzo, Antonio Di Cesare and Rocco Ditommaso (2012). Electromagnetic Sensing Techniques for Non-Destructive Diagnosis of Civil Engineering Structures, Earthquake-Resistant Structures - Design, Assessment and Rehabilitation, Prof. Abbas Moustafa (Ed.), ISBN: 978-953-51-0123-9, InTech, Available from: http://www.intechopen.com/books/earthquake-resistant-structures-design-assessment-andrehabilitation/electromagnetic-sensing-techniques-for-civil-engineering-structures-non-destructive-diagnostics
### **InTech Europe**
University Campus STeP Ri Slavka Krautzeka 83/A 51000 Rijeka, Croatia Phone: +385 (51) 770 447 Fax: +385 (51) 686 166 www.intechopen.com
#### **InTech China**
Unit 405, Office Block, Hotel Equatorial Shanghai No.65, Yan An Road (West), Shanghai, 200040, China Phone: +86-21-62489820 Fax: +86-21-62489821
© 2012 The Author(s). Licensee IntechOpen. This is an open access article distributed under the terms of the Creative Commons Attribution 3.0 License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. | doab | 2025-04-07T03:56:58.705621 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/3.0/",
"book_id": "003dfb30-3d3b-432a-9c4e-217966fc6cb3",
"url": "https://library.oapen.org/bitstream/20.500.12657/49020/1/30141.pdf||https://library.oapen.org/bitstream/20.500.12657/49020/1/30141.pdf",
"author": "Bernini, Romeo||Bavusi, M.||Lapenna, Vincenzo||Di Cesare, Antonio||Carlo Ponzo, Felice||Soldovieri, Francesco||Loperte, Antonio||Ditommaso, Rocco",
"title": "Chapter Electromagnetic Sensing Techniques for Non-Destructive Diagnosis of Civil Engineering Structures",
"publisher": "InTechOpen",
"isbn": "",
"section_idx": 0
} |
|
004623c8-cb6c-4a99-aee8-c83552616d6e | # **Ecological Impacts of Toxic Chemicals**
# **Editors**
# **Francisco Sánchez-Bayo**
*University of Technology Sydney, Australia*
# **Paul J. van den Brink**
*Alterra and Wageningen University, The Netherlands*
# **Reinier M. Mann**
*University of Technology Sydney, Australia*
© 2011 by the Editor / Authors. Chapters in this eBook are Open Access and distributed under the Creative Commons Attribution (CC BY 4.0) license, which allows users to download, copy and build upon published chapters, as long as the author and publisher are properly credited, which ensures maximum dissemination and a wider impact of our publications. The book taken as a whole is © 2011 Bentham Science Publishers under the terms and conditions of the Creative Commons license CC BY-NC-ND.
# **CONTENTS**
# **FOREWORD**
"Ecological Impacts of Toxic Chemicals" is a long-overdue, comprehensive coverage of chemical fate and effects in terrestrial and aquatic environments. The editors Sánchez-Bayo, van den Brink and Mann have brought together an excellent group of international experts to systematically cover this complex topic from the source of organic and metal compounds, to their fate and impacts on land and in our freshwater and marine ecosystems. The book is very readable, serving as an excellent introduction to the topic or as a useful supplement to courses and readings in the environmental sciences at any level. Indeed, it is appropriate for the general public, students, or scientists from outside the field of ecotoxicology.
The first two chapters, by Sánchez-Bayo (Chapter 1) and van de Meent, Hollander, Peijnenburg and Breure (Chapter 2) introduce the theme of the book, covering the sources and mode of action of environmental contaminants and the toxicity of various common pollutant categories: mining wastes, sewage, industrial and metropolitan discharges. The transport and fate of metal and organic pollutants in the environment is described from a modeler's perspective. The processes governing the movement of chemicals between air, land and water are described, along with biological transformations, including degradation and bioaccumulation. The understanding of the fate and ultimate exposure to biota is essential in ecotoxicology and risk assessment and management.
The following three chapters deal with terrestrial ecosystems. In Chapter 3, Mann, Vijver and Peijnenburg explain how naturally-occurring metals and metalloids can become contaminants when they bioaccumulate and result in sublethal to lethal effects on populations and food chains. They cover the key metals of toxicological concern which continue to be a problem world-wide: arsenic, cadmium, copper, lead, mercury, molybdenum, selenium and zinc. Agricultural pesticides have been widely used in developing and developed countries and because they are biocides, have resulted in a range of unintended adverse effects on non-target biota. Sánchez-Bayo discusses fungicides, insecticides and herbicides and how they have impacted virtually every level of the food chain, from the microbial level to birds and mammals. Thompson focuses on the forest industry's use of pesticides (herbicides and insecticides) and case examples of lab to field studies that have assessed the risk of these widely used compounds in pest management in the forest sector. These studies are then linked to the risk assessment and management process providing for a comprehensive perspective of multiple stakeholder concerns.
The final six chapters address the many issues of chemicals in marine and freshwater environments. Schäfer, van den Brink and Liess have an excellent review of pesticide impacts on freshwater ecosystems, from primary producers, up the food chain, to fish. They explain the many complex interactions that must be considered regarding pesticide mode-of-action, exposure (particularly consideration of peak concentrations), indirect effects, and the potential for recovery of populations and communities. They describe a range of useful techniques and approaches for assessing pesticide risk from the broad to local scales, and the need for incorporating ecological knowledge into the risk assessment process. A growing concern exists for the impacts of other, non-pesticide, organic chemicals in freshwater ecosystems which is dealt with in Chapter 7 by Sibley and Hanson. Persistent organic pollutants (POPs), such as polychlorinated biphenyls (PCBs), polychlorinated dioxins and furans, polycyclic aromatic hydrocarbons (PAHs), and emerging contaminants such as pharmaceuticals, polybrominated diphenyl ethers and perfluorinated surfactants are becoming common in freshwaters throughout the world. This is due to their resistance to degradation and ability to be transported between water, soil, air and biota. Their bioaccumulation through the food chain presents recognized risks, but these risks are difficult to ascertain from studies at the lower end of the food chain. Chemicals tend to accumulate in sediments, hence biota associated with sediments, the benthos, are particularly susceptible. Borja, Belzunce, Garmendia, Rodríguez, Solaun and Zorita describe this complex issue in Chapter 8, for coastal and marine benthic communities. Their coverage begins at the molecular effect level and progresses up the ladder of biological complexity to populations and communities, and the need for integrative assessments. They document how important it is to understand biological effects by looking at the different levels of biological organization. Dying coral reefs have been documented throughout the world. They are impacted by nutrients, metals, organic chemicals, climate change and ocean acidification. In Chapter 9, van Dam, Negri, Uthicke and Mueller explain the severity of this phenomenon and the tools available for evaluating adverse effects. Of critical importance is their coverage of how adverse effects and risk is tied to exposures, which vary from short-term, to pulse-like spills, to recurring incidents from effluent discharges to river flooding. These later, chronic and repetitive events are likely to decrease the resilience of reef organisms making them more susceptible to climate change and acidification. In contrast to the previous two chapters, Hylland and Vethaak in Chapter 10 focus on contaminant effects on water column organisms, often referred to as pelagic organisms which fuel the world's ocean ecosystems. The various ways of assessing pelagic effects are reviewed, along with the unique strengths and limitations in the context of making environmental management decisions. Better monitoring of the pelagic zone is critical for long term monitoring programs and effective ecosystem management. Finally, in Chapter 11, Moore, Kröger and Jackson inform the reader of how aquatic ecosystems are so efficient at transferring, transforming and sequestering pollutants, thus reducing their risk to organisms and ecosystems. They focus on the successful use of phytoremediation of organic and inorganic pollutants.
Together, these chapters provide a broad, timely and comprehensive review of the potential effects of chemical pollutants in terrestrial and aquatic ecosystems. Readers new to this field will not be disappointed and quickly made aware of the critical issues affecting our current and near-future world.
> *G. Allen Burton* University of Michigan, Ann Arbor
# **PREFACE**
Ecotoxicology is a multidisciplinary science that examines the effects of toxic chemicals on individual organisms, populations, communities and ecosystems. However, with a 40-year history, ecotoxicology is still in its infancy. Up until recently a lot of work has been done to describe the fate and effect of chemicals in the environment, but most of it has been performed in the laboratory, usually with a narrow suite of test organisms. However, over the last two decades more and more experiments and monitoring have been performed in man-made (so called microcosms and mesocosms) as well as natural aquatic and terrestrial ecosystems. Also the use of modelling has allowed us to predict the behaviour of chemicals and their consequent effects in the environment. Impacts of pollutants at an ecosystem level, however, are reported mostly in the specialized journal literature as scattered pieces of a larger puzzle. To date, no systematic work bringing all the information on this subject together is available, neither to researchers nor the general public. This book was conceived to fill this gap.
*Ecological Impacts of Toxic Chemicals* presents a comprehensive, yet readable account of the known disturbances caused by all kinds of toxic chemicals on both aquatic and terrestrial ecosystems. Topics cover the sources of toxicants, their fate and distribution through the planet, their impacts on specific ecosystems, and their remediation by natural systems. Each chapter is written by well-known specialists in those areas, for the general public, students, and even scientists from outside this field. The book intends to raise awareness of the dangers of chemical pollution in a world dominated by industry and globalization of resources. Because the problems are widespread and far reaching, it is hoped that confronting the facts may prompt better management practices at industrial, agricultural and all levels of management, from local to governmental, so as to reduce the negative impacts of chemical contaminants in our Earth.
The editors would like to thank Bentham Science Publishers for providing this opportunity to bring this science to the general public.
> **Francisco Sánchez-Bayo, Paul J. van den Brink, Reinier M. Mann,**
# **List of Contributors**
# **María Jesús Belzunce**
AZTI-Tecnalia, Marine Research Division, 20110 Pasaia, Spain
# **Ángel Borja**
AZTI-Tecnalia, Marine Research Division, 20110 Pasaia, Spain; Email: [email protected]
# **Ton Breure**
RIVM Laboratory for Ecological Risk Assessment, Bilthoven 3720 BA, The Netherlands
# **Paul J. van den Brink**
Alterra and Wageningen University, Wageningen University and Research Centre, P.O. Box 47, 6700 AA Wageningen, The Netherlands; Email: [email protected]
# **Joost W. van Dam**
Australian Institute of Marine Science, Townsville, Qld 4810, Australia; Email: [email protected]
# **Joxe Mikel Garmendia**
AZTI-Tecnalia, Marine Research Division, 20110 Pasaia, Spain
# **Mark L. Hanson**
Department of Environment and Geography, University of Manitoba, Canada R3T 2N2
# **Anne Hollander**
Radboud University Nijmegen, Nijmegen,The Netherlands
# **Ketil Hylland**
Department of Biology, University of Oslo, Blindern N-0316 Oslo, Norway; Email: [email protected]
# **Colin R. Jackson**
Department of Biology, University of Mississippi, Mississippi 38677, USA
# **Robert Kröger**
Department of Wildlife, Fisheries and Aquaculture, Mississippi State University, Mississippi 39762, USA
# **Matthias Liess**
Department System Ecotoxicology, UFZ – Helmholtz Centre for Environmental Research, Leipzig 04317, Germany
# **Reinier M. Mann**
Centre for Ecotoxicology, Department of Environmental Sciences, University of Technology Sydney, NSW 2007, Australia; Present Address: Hydrobiology, Brisbane, Australia; Email: [email protected]
# **Dik van de Meent**
RIVM Laboratory for Ecological Risk Assessment, Bilthoven 3720 BA, The Netherlands; Email: [email protected]
# **Matthew T. Moore**
USDA Agricultural Research Service, National Sedimentation Laboratory, Oxford, Mississippi 38655, USA; Email: [email protected]
# **Jochen F. Mueller**
The University of Queensland, National Research Centre for Environmental Toxicology, Coopers Plains, Qld 4108, Australia.
# **Andrew P. Negri**
Australian Institute of Marine Science, Townsville, Qld 4810, Australia
# **Willie J.G.M. Peijnenburg**
Laboratory for Ecological Risk Assessment, National Institute of Public Health and the Environment, 3720 BA Bilthoven, The Netherlands; Leiden University, Institute of Environmental Sciences, 2300 RA Leiden, The Netherlands
# **José Germán Rodríguez**
AZTI-Tecnalia, Marine Research Division, 20110 Pasaia, Spain
# **Francisco Sánchez-Bayo**
Centre for Ecotoxicology, University of Technology Sydney, NSW 2007, Australia; Department of Environment, Climate Change & Water NSW, 480 Weeroona Road, Lidcombe NSW 2141, Australia; Email: [email protected]
# **Ralf B. Schäfer**
RMIT University, Melbourne, Australia; Present address: Institute for Environmental Sciences, University Koblenz-Landau, Landau, Germany; Email: [email protected]
# **Paul K. Sibley**
School of Environmental Science, University of Guelph, Ontario, Canada N1G 2W1; Email: [email protected]
# **Oihana Solaun**
AZTI-Tecnalia, Marine Research Division, 20110 Pasaia, Spain
# **Dean G. Thompson**
Canadian Forest Service, Natural Resources Canada, Sault Ste. Marie, Ontario, Canada P6A 2E5; Email: [email protected]
# **Sven Uthicke**
Australian Institute of Marine Science, Townsville, Qld 4810, Australia
# **Martina G. Vijver**
Leiden University, Institute of Environmental Sciences, 2300 RA Leiden, The Netherlands
# **A. Dick Vethaak**
Deltares, Marine and Coastal Systems, 2600 MH Delft, The Netherlands, VU University Amsterdam, Institute for Environmental Studies, De Boelelaan 1105, 1081 HV Amsterdam, The Netherlands
# **Izaskun Zorita**
AZTI-Tecnalia, Marine Research Division, 20110 Pasaia, Spain
# **CHAPTER 1**
# **Sources and Toxicity of Pollutants**
# **Francisco Sánchez-Bayo\***
*Centre for Ecotoxicology, University of Technology Sydney, Australia*
**Abstract:** Modern living standards depend largely on the production and usage of thousands of chemicals, many of which are toxic and synthetically produced. These substances are discharged into the air, soil, water bodies and the sea through a variety of ways, becoming pollutants of our environment. The investigation of their fate and impacts they have on ecosystems is called ecotoxicology, a multidisciplinary science which intends to evaluate the nature of the discharge, the transformation and distribution of toxicants in the environment, exposure, lethality and sublethal effects on organisms, population responses, and changes in community structure and ecosystem function. The sources and mode of action of some of the most common groups of toxicants are described in this chapter, leaving their fate and effects in organisms and ecosystems for the subsequent chapters.
# **INTRODUCTION**
We are living in the Chemical Era. Indeed, the most distinctive characteristic of our modern society is the production and use of an enormous amount of chemical products. Currently, some 70,000 chemicals are utilised worldwide, while the rate of introduction of new substances can be estimated between 200 and 1000 each year [1]. Our civilization depends to a large extent on the search for new materials that are employed to develop technology, medicines, textiles and construction materials of all kinds. What would happen to us if such production were to stop suddenly?
Throughout history, civilizations have relied on the use of natural materials to manufacture tools, clothing and furnishings, while poisons and medicinal plants must have been known to the first humans. The development of agriculture during the Neolithic (11,000 BP) brought with it the manufacturing of textiles as well as dyes and paints made from minerals, plant and animal products, some of which are quite toxic. With the discovery of metals during the Bronze and Iron Ages (ca. 5000 and 3000 BP, respectively) came mining and consequently pollution by toxic metals. Alchemy started in Persia about 2500 BP, and since then each civilization in Asia looked to develop new substances, mainly for medicinal purposes, using the diverse array of natural products available to them. In Europe, alchemy laid the foundations of toxicology and modern chemistry in the 16th and 17th centuries respectively, which would result in the discovery and manufacturing of hundreds of entirely new substances. During the industrial revolution of the 19th century, mining and chemical companies were created specifically to exploit natural resources and create new products. The discovery of large deposits of crude oil in Baku (Azerbaijan) and North America in the 1850s [2], together with the realisation that petroleum could be used as fuel for combustion engines, boosted the mechanised and oil-dependent society we still live in. As a result of these activities, pollution on a large scale began at that time and still continues despite amelioration efforts by governments and industries in most developed countries.
The technological race that started in the 20th century, particularly since the end of World War II, included chemicals as an essential part of modern development. For instance, communications and transport have benefited enormously from the use of new metals and alloys to make transistors, batteries and more durable metallic products. Synthetic organic compounds, mostly derived from petroleum, underwent a revolution of their own: polychlorinated biphenyls (PCBs), used as insulating fluids in the electrical industry since their introduction in 1929; chlorofluorocarbons (CFCs) used in refrigeration and air conditioning systems; plastics to serve a wide range of uses, from building materials to household items and toys; pesticides to control insect and rodent pests, weeds and plant diseases; and the immense array of chemicals used to make paints, cleaning products, cosmetics and pharmaceuticals. Many of these new substances are toxic, have become environmental pollutants in air, water and soil, and created unforseen problems related to their waste and disposal.
Although the discovery of toxic substances dates from ancient times, their systematic study or toxicology began during the European Renaissance with Paracelsus (1493-1541), a medical doctor and alchemist who sought to understand the effects of toxicants and drugs used in medicine. However, it wasn't until the effects of new pollutants
**<sup>\*</sup>Address correspondence to Francisco Sánchez-Bayo:** Centre for Ecotoxicology, University of Technology Sydney, NSW 2007, Australia; Department of Environment, Climate Change & Water NSW, 480 Weeroona Road, Lidcombe NSW 2141, Australia; Email: [email protected]
### **4** *Ecological Impacts of Toxic Chemicals Francisco Sánchez-Bayo*
from the industrial revolution started to take a toll on ecosystems that people realised the dangers they posed to our environment and our own health. In Japan, a country which experienced the fastest transformation from a feudal to an industrial society, Tanaka Shozo (1841-1913) appealed to the Meiji Emperor in protest against fish kills due to careless discharges from the Ashio copper mine north of Tokyo [3]. It might have been the first time that a local politician tried to protect the environment and the lives of his community by demanding regulation of indiscriminate exploitation of resources. Japan would suffer dearly the consequences of such a rush for industrial development, with Minamata and '*itai-itai*' being added to the infamous list of modern diseases caused by pollutants [4]. These unintended problems prompted a rethink of treating the natural environment as a receptacle for untreated waste, and yet many other industrialised societies would have to endure a large human toll from *smog* before taking any action to regulate the burning of fossil fuels in their cities [5]. In this climate, the publication of *Silent Spring* in 1962 [6] brought to the attention of ordinary people in the street what scientists were still trying to comprehend: the negative effects that pesticides can pose to the environment. Such a book would mark the start of the environmental movement in America and the rest of the world.
The investigation of the ecological impacts that toxic pollutants have on ecosystems constitutes a new science, emerged in the 1970s, called ecotoxicology [7]. Its approach is multidisciplinary, combining the knowledge from chemistry, toxicology and ecology to reach an understanding of the complex interactions of toxicants in the environment. In a broader sense, ecotoxicology has the role of assessing, monitoring and predicting the fate of foreign substances in the environment [8], with the ultimate end of helping the regulatory authorities establish limits that protect human health and nature.
This first chapter aims at providing the reader with a glimpse of the different kinds of pollutants currently in existence, where they come from and how they exert their toxic effects on organisms. Subsequent chapters will examine the ways these chemicals move between air, soil, rivers and oceans, with special attention given to the overall impacts on specific communities and types of ecosystems.
# **TYPES AND SOURCES OF TOXICANTS**
Toxic chemicals that pollute the environment can be called ecotoxicants [1]. They can be natural or man-made substances, but a common characteristic to all of them is that they can exert a deleterious effect on living organisms at relatively small doses, measured in milligrams or micrograms per litre or per kilogram [9]. An important aspect to consider with ecotoxicants is whether they are available to organisms (see risk assessment below). Indeed, pollutants are discharged into the air and water or disposed of in or on the ground, where they may be absorbed by plants or taken up by animals, which may in turn be affected by their toxic activity. By contrast, the vast majority of naturally occurring toxicants (e.g. plant poisons) are stored in tissues that are only available to animals if eaten. In the case of crude oil, natural deposits are many metres underground and out of reach…except to humans! For this reason, biological toxins very rarely become pollutants – these originate mostly from human activities of our modern society (Table **1**). The following is a brief description of the most common types.
# **Toxins of Biological Origin**
Although our knowledge is still limited, the variety of plant and animal poisons is staggering [10], with many of them being utilised as medical drugs or in the production of pesticides – the toxins of the soil microbe *Bacillus thuringiensis*, for instance, are used for pest control, either directly or through transgenic plants [11]. Most natural toxins produced by organisms are used as defence tools in mechanisms that evolved over millions of years, but some animals produce toxic venoms to capture and kill their prey [12]. In any case, very few of these toxins are ecotoxicants: botulin produced by the soil bacterium *Clostridium botulinum*, mycotoxins produced by some species of fungi, cyanotoxins and microcystins produced by certain blue-green algae [13], and saxitoxins and brevetoxins produced by several species of dinoflagellates (e.g. *Alexandrium* sp.) are the most notorious [14], as they can cause fish deaths through algal-blooms and serious intoxication or health problems in humans.
# **Waste Products**
Natural ecosystems recycle the elements through a variety of pathways which end up in mineralization, thus ensuring that all organic wastes in soil, water and sediments are eliminated as soon as possible. Raw sewerage is processed by naturally occurring micro-organisms in waters provided its volume is within the capability of aquatic ecosystems, but
large cities discharge excessive volumes of refuse into rivers, lakes and coasts, which if insufficiently treated can lead to eutrophication of the waters and foster toxic algal blooms [15]. Moreover, stormwater runoff and waste discharges from cities often contain a variety of toxic chemicals, including metals, petroleum hydrocarbons, pharmaceuticals, pesticides [16,17], phenols, steroids and many others which have endocrine disrupting activity [18].
**Table 1:** Sources of toxic pollutants and their mode of action.
\* Some compounds occur naturally as well
# **Metals and other Elements**
All elements, whether they are toxic or not, occur naturally in soils, air, oceans and sediments, usually at very low, non-toxic levels. Some of them are called micronutrients, since they are essential for the synthesis of certain biomolecules: iron, cobalt, chromium, copper, iodine, manganese, selenium, zinc and molybdenum. However, human activities such as mining, manufacturing and transport have produced abnormally elevated levels of many toxic elements, thus causing detrimental effects in the environment [19]. For example, urban soils throughout the world contain high concentrations of lead due to the intense use of leaded-fuel in motor vehicles for many decades. Mine tailings contain very high concentrations of residual metals and constitute a high risk to the surrounding environments, with their accidental release causing enormous damage to aquatic ecosystems and associated fishing industries [20]. Smelters and factories that process large amounts of metals may also be sources of metal pollution through smoke stacks and discharges into waterbodies.
# **Synthetic and Natural Organic Toxicants**
Although the majority of organic toxicants that contaminate the environment are man-made, some can be produced by natural events. Bushfires, for instance, are major sources of air and water contamination by polycyclic aromatic hydrocarbons (PAHs), which are present in nature at very low concentrations – background or baseline levels [21]; however, their environmental levels have increased in recent years due to the large increase of fossil fuels usage worldwide. Crude oil, composed mainly of a cocktail of aliphatic hydrocarbons of biological origin, has been stored safely underground for millions of years [2]; the petroleum spillages that result from the accidental break up of oiltankers or oil-field shafts, even if they are not very toxic, may have other temporary impacts on the ecosystems exposed [22].
In addition to these natural ecotoxicants there is an immense range of synthetic compounds, man-made for a variety of purposes: pesticides and fertilizers used in agriculture, industrial chemicals such as PCBs and CFCs, chemical reagents, solvents, plasticizers, dyes, surfactants, detergents, pharmaceutical drugs and explosives used in warfare and industrial activity. All these chemicals can be released into the environment through manufacturing, usage, accidental spillage or inefficient disposal. For example, chlorodibenzodioxins (PCDD) and chlorodibenzofurans (PCDF) are by-products formed in the manufacturing of some chlorinated pesticides and also from incineration of chlorophenolic wastes [23].
# **Inorganic Toxicants**
Volcanic eruptions discharge enormous volumes of sulphur dioxide, carbon monoxide and other inorganic poisons into the air (Table **1**), where in addition to their inherent toxicity are subsequently transformed into acids by reacting either with atmospheric water vapour or in direct contact with water [24]. Even carbon dioxide can be toxic at concentrations higher than 2%. Once again, the levels of these toxicants in the environment have increased due to human activities such as fertiliser usage or burning of coal and petroleum fuels. On a different matter, they can affect the global weather patterns, with CO2 and NO2 contributing to the warming of the atmosphere while SO2 forms aerosols which have the opposite effect [25].
# **MODE OF ACTION OF TOXICANTS**
Toxicants are chemicals that have toxic effects on organisms. Such effects occur when the toxicant, after being taken up through the roots or leaves in plants, or through the skin, digestive or respiratory systems in animals, is transferred by the circulatory fluids to the site of action within the plant or animal body. Some of the original compound, or its biotransformed active products, may reach the site of action, while the remainder may be excreted and metabolised, usually to non-toxic products, or stored in lipid tissues [26].
For most chemicals, the site of toxic action is at the cellular level, and their activity translates into one or more physiological effects which are manifested as several toxicity symptoms in individual organisms. For example, mercurial compounds block the degradation pathway of catecholamines in neuronal cells, so the resulting excess of epinephrine (adrenaline) in the blood stream causes profuse sweating and hypersalivation, whereas excess of dopamine in the brain induces tachycardia and hypertension. But organisms are rarely affected in isolation; since toxic pollutants are spread over certain areas, they usually affect a number of organisms at the same time – effects at the population level can translate into reduction of numbers (mortality) or decreased reproduction success in certain species but not in others. In turn, these changes typically result in altered communities of animals and/or plants, which impact the ecosystem functionality (e.g. biomass productivity, nutrient and predator-prey dynamics) to a lesser or greater degree [27]. Effects at different levels of organization are not always observed: while all ecotoxicants have effects at the individual level, or at most at the population level in acute exposures, impacts on communities and ecosystems depend mainly on the persistence and/or bioaccumulation of the chemicals concerned. However, short pulses can also have a large, long-lasting impact when the recovery of the communities affected is slow or when the ecosystem has been pushed to an alternative stable state.
# **Inorganic and Elementary Compounds**
Inorganic toxicants usually disable the functionality of essential biomolecules; for example, CO binds to haemoglobin and prevents it from transporting oxygen in the blood. Even if some are essential micronutrients, toxic metals (As, Cd, Cr, Cu, Hg, Ni, Pb, Sb, Sn, Zn) tend to form covalent bonds with organic molecules and effectively disable their funcionality:
# **Organic Compounds**
Within the enormous variety of organic toxicants, the following are some of the most common mechanisms of toxicity:
# **BASIC TOXICOLOGICAL PRINCIPLES**
Paracelsus established that "All things are poison and nothing (is) without poison; only the dose permits something not to be poisonous" [37]. In toxicology, the degree of exposure or dose is as important as the nature of a chemical. Common salt (NaCl), for instance, can be toxic to most aquatic organisms at concentrations above 6%. For a typical toxicant, the response is represented by a sigmoid curve which indicates increasing toxic effects at increasing doses, usually on a logarithmic scale (Fig. **1**). For essential elements (*i.e.* copper), the relationship is parabolic: at lower than normal doses for growth and development, the organisms may die of nutrient deficiency; as the levels increase within a short range there are not harmful effects, whereas higher doses produce the normal toxic response. In aquatic environments the dose is replaced by the concentration of the toxicant in water, since the internal dose (which causes the effect) is a function of the uptake rate by the organism and the external concentration [38]. The toxicity of a compound is usually evaluated as the dose required to cause a 50 percent effect in an organism (EC50), which in the case of mortality is the median lethal dose or concentration: LD50 or LC50 respectively. The noobserved effect level (NOEL) and lowest-observed effect level (LOEL) are also used, but they are less reliable measures and sometimes difficult to obtain.
**Figure 1:** Dose-effect relationships.
In addition to dose, time is an important factor to be considered in toxicology. Indeed, the same effect can be produced by a large dose applied to an organism at once than by repeated small doses applied over a certain period of time (chronic exposure), as long as the toxicant is not degraded (Fig. **2**). The exposure time is in fact related to the actual internal dose, and both are linked by a linear relationship in the logarithmic scale [39]. Obviously this relationship is more important where a continuous input occurs or in the case of persistent ecotoxicants, which can linger in the environment for months or years, and can be accumulated in the tissues of the organisms affected. For example, many chlorinated organic toxicants (PCBs, PCDDs, DDT and other pesticides) are recalcitrant and nonpolar compounds with a tendency to be stored in fatty tissues because they are lipophilic; the remobilization of these toxins from the body fat during periods of starvation or exertion, such as long migration, can take them to the sites of action (e.g. neuronal system) and produce a toxic effect [40].
Finally, sublethal effects – those effects caused by doses smaller than the LD50 or acute NOEL – are commonly found after chronic exposure of organisms to low levels of ecotoxicants. These effects are usually unrelated to the specific mode of action of the chemicals, and therefore are unpredictable. For example, some organochlorine insecticides which are neurotoxic at relatively high doses (mg/kg body weight), can also produce endocrine disruption when present at very low doses (g/kg body weight) [41]. In other instances, such as DDT, the metabolite DDE causes the thinning of egg-shells and consequently can lower the hatching success of many bird species [42].
**Figure 2:** Time-to-effect relationships.
# **ECOLOGICAL RISK ASSESSMENT**
The hazardous substances that pollute our environment (Table **1**) can cause detrimental effects only when organisms are exposed to them at sufficient, toxic doses. Ecological risk assessments aim at determining the extent of harm caused to the environment by the release of toxic chemicals. Such assessments consider the exposure of organisms and inherent toxicity of the substances as the two main components, and usually indicate the impacts in terms of probabilities [43]. Most risk assessments refer to individual toxicants discharged over specific areas either in one event (e.g. accidental spills) or repeatedly (e.g. continuous industrial effluents, pulsed pesticide applications), and consequently tend to be site-specific.
The exposure of organisms to toxicants depends on many factors: distribution and concentration of the chemicals in water, air, soil and sediment; uptake through the roots and/or leaves in plants or through contact, ingestion and inhalation in animals; persistence of the residues in the environment and bioaccumulation in tissues; proximity to the source of emissions and probability of being affected. The toxicity component is based on experimental laboratory data, which derive the LD(C)50 or NOEL (acute or chronic) of the chemical for certain species of organisms such as algae, *Daphnia*, worms, insects, fish, birds and mammals. Typically, endpoints include mortality as well as sublethal effects. A chemical's hazard indicates the danger it may pose to organisms; risk implies the probability of being affected. The simplest way to assess the risk is by comparing the amounts of chemical present in water, air, soil and sediment with the toxic endpoints to each organism [44]:
$$HQ = \frac{\text{Predicted Enviornmental Concentration (PEC)}}{\text{Toxicity endpoint (LD(C)50 or NOEL)}}$$
If the hazard quotient (HQ) between the two numbers is greater than 1 the organisms would be at high risk, meaning that more than 50% of the individuals of a given species may die or experience sublethal effects. Obviously, the PEC can be replaced by actual measured concentrations. Based on years of experience and numerous field trials – mainly with insecticides – it is concluded that for an ecotoxicant to be considered 'safe' to organisms, the HQ should be smaller than 0.1 [45]. Since the toxicity endpoints are fixed for each chemical and species the main variable of the quotient is the environmental concentration – this forms the basis for establishing contaminant guidelines for air, soil/sediment and water quality, to ensure that levels of pollutants are below harmful thresholds. In any case, such quotients only indicate potential hazards to certain species, while additional information on the chemical's persistence, bioaccumulation factors and probability of actual exposure to a community of species, under normal and worst-case scenarios, are needed for a more comprehensive risk assessment to ecosystems [46].
# **REFERENCES**
### **12** *Ecological Impacts of Toxic Chemicals Francisco Sánchez-Bayo*
© 2011 The Author(s). Published by Bentham Science Publisher. This is an open access chapter published under CC BY 4.0 https://creativecommons.org/licenses/by/4.0/legalcode
# **Fate and Transport of Contaminants**
#### **Dik van de Meent1,2,\*, Anne Hollander2 , Willie Peijnenburg1,3 and Ton Breure1,2**
*1 National Institute for Public Health and the Environment (RIVM), Bilthoven, NL; <sup>2</sup> Radboud University Nijmegen, Nijmegen, The Netherlands and <sup>3</sup> Leiden University, Leiden, The Netherlands*
**Abstract:** Release of toxic chemicals into the environment cannot always be avoided completely. As a result organisms, man included, will be exposed to chemicals *via* the environment. Given the release of certain chemicals into the environment, their exposure concentrations in air, water and soil would depend on the rates at which they are removed from the environment. This chapter deals with the transport and transformation processes that affect concentrations in the environment, with emphasis on the modeller's perspective. Being interested primarily in the effects that processes have on concentrations of chemicals in environmental media, we focus on a quantitative description of the rates at which losses from the environment take place, and on how these rates differ for different chemicals. We systematically formulate process rate constants for each transport or transformation process. Eventually, the rate constants combine into a mass balance model which allows us to describe and predict how releases into the environment result in exposure concentrations of organisms.
# **PROCESSES AND MECHANISMS**
After entering the environment, chemicals are transported, distributed over the various environmental compartments and may be transformed into other chemicals. Transport can occur within a compartment, such as in air or in soil, or between compartments (e.g. between air and water, air and soil or water and soil). Transformation processes in the environment involve chemical degradation or biodegradation.
Process rates (*i.e.* the mass flows of substance that result from them) generally depend on two independent factors: (i) the concentration of the substance in the environmental medium (driving force) and (ii) the likelihood of occurrence of the process (rate constant). When process rates are directly proportional to concentrations, process kinetics are called first order (first power concentration). Non-linear relationships apply in cases of higher or lower order kinetics. In the case of first-order kinetics, the mass *M* of chemical in the environmental compartment of origin falls exponentially with time *t*:
$$\frac{dM}{dt} = -k\_{\rm loss} \cdot M \quad \text{or} \quad M = M\_0 e^{-k\_{\rm int} \cdot t} \tag{1}$$
The first-order loss process is characterized by one single parameter: the loss rate constant *kloss*. The loss mechanism causes the mass of chemical in the compartment to fall from its original value *M0* to half of that value in a constant time period: the half-life time *t1/2* of the chemical in the compartment. Mass keeps falling at the same half-life until the chemical has disappeared entirely. It can be deduced from Eq **1**, or seen from its graphical representation (Fig. **1**), that in case of first-order kinetics the half-life time *t1/2* has a constant value throughout the loss process:
$$t\_{1/2} = \frac{\ln 2}{k\_{\text{loss}}} \tag{2}$$
Throughout this chapter, we shall assume first-order kinetics to apply to all transport and transformation process. Although in reality this is certainly not always true, usually too little information is available to better describe the process rate-concentration dependence. Moreover, even in cases where higher order kinetics apply, assuming firstorder kinetics will not always lead to dramatically erroneous predictions, as often concentrations do vary only slightly, so that so-called pseudo-first-order kinetics apply.
**\*Address correspondence to Dik van de Meent:** RIVM Laboratory for Ecological Risk Assessment, Bilthoven 3720 BA, The Netherlands; Email: [email protected]
**Francisco Sánchez-Bayo, Paul J. van den Brink and Reinier M. Mann (Eds) © 2011 The Author(s). Published by Bentham Science Publishers**
This text is taken largely from chapters 3 and 4 of the textbook by Van Leeuwen and Vermeire [1], the authors of which are kindly acknowledged.
**Figure 1:** Loss of chemical from the environment through first-order kinetics results in constant half-life.
# **Equilibrium Partitioning Between Phases**
Obeying general laws of thermodynamics, chemicals tend to spontaneously migrate from one phase (environmental medium: air, water, soil) to another if the phases are not in equilibrium. Migration in multi-phase systems continues until equilibrium has been reached. In thermodynamics, equilibrium is characterized as the state in which the chemical potential, and the chemical's activity and fugacity have the same value in the different phases. This principle has successfully been applied [2] in mass balance models of environmental fate of chemicals, that have become known as "fugacity models" or "Mackay models". For most practical situations, the equilibrium condition can be expressed by stating that chemicals are driven towards equilibrium until the ratio of concentrations (*C*1 and *C*2) is equal to the intermedia equilibrium constant *K*, also known as partition coefficient:
$$K\_{12} = \frac{C\_1}{C\_2} \tag{3}$$
Departure from thermodynamic equilibrium forms the main driving force of intermedia transport of chemicals. This is why intermedia equilibrium constants play such an important role in quantitative mathematical descriptions of transport and fate of chemicals in the environment.
# *Solids-Water Equilibrium*
Equilibrium partitioning between water and solids is the result of adsorption of the chemical onto the surface of particles. For low concentrations of the chemical in water, the equilibrium ratio is usually a constant: *K*12 of Eq **3** is independent of the concentrations of the chemical. For higher concentrations, it is often observed experimentally that the equilibrium ratio does depend on the concentrations. In such cases, the equilibrium relationship between the concentrations is given by a non-linear sorption isotherm. The Freundlich-isotherm equation is often used (without making assumptions about the nature of the underlying mechanism) to fit experimentally observed non-linear sorption.
Commonly used estimation methods for solids-water partition coefficients *K*p are based on the assumption that there is a "hydrophobic sorption" mechanism. This mechanism is generally modelled based on the organic carbon content of the soil, sediment or suspended solids *f*oc and the octanol-water partition coefficient of the chemical *K*ow, using simple regression equations:
$$\log K\_p = \log(K\_{oc} \cdot f\_{oc}) = a \log K\_{OW} + b + \log f\_{oc} \tag{4}$$
where
*Koc* = organic carbon referenced solids-water partition coefficient (L/kg)
*a*, *b* = constants, specific for chemical classes
Given a value of *Kp*, the extent to which partitioning from water to solids occurs depends on the amount of solids present. At equilibrium, the mass fraction of chemical dissolved in water *dissolved* can be calculated as
$$\mathfrak{sp}\_{diosobed} = \frac{1}{1 + K\_p \cdot \text{TSS}} \tag{5}$$
where
*TSS* = mass concentration of suspended solids in water (= ~ 10-5 kg/L)
Normalization to the organic carbon content of particulate matter has become standard procedure in this field of research. This procedure is based on the experimental observation that the *Kp* of organic chemicals is often proportional to the organic matter content of the solid phase. The estimation method was derived originally for hydrophobic chemicals [3]. The method was extended to apply to various other classes of non-ionic organic chemicals [4] and, more recently, to ionizing organic acids and bases [5]. The method cannot be applied to metals and other inorganic ionizing substances.
Solids-water partition coefficients are commonly reported in units L/kg. The physical meaning of this dimension can be understood by reading it as "the volume of water (L) which contains the same amount of the chemical as one kilogram of solid material does". For many purposes, however, we are not just interested in the concentration ratio, but also in the mass distribution of the chemical over the phases. Obviously, this distribution depends on both the partition coefficient and the relative volumes of the phases. In surface water, the solids-water ratio is much smaller than in sediment and soil systems. As a result, the extent of partitioning of a certain chemical into the particle phase of sediment or soil is much greater than in surface water. Solids-water partition coefficients of chemicals *Kp* range from < 1 L/kg to > 10<sup>5</sup> L/kg, with resulting extents of partitioning into the solid phases ranging from negligible in surface water to near-complete in soil (Table **1**).
# *Air-Water Equilibrium*
Equilibrium between air and water is given by Henry's law, which states that in equilibrium, the partial pressure of a chemical in the gas phase is proportional to its concentration in water. The ratio of these, Henry's law constant *H*, can be obtained as the ratio of the saturated vapour pressure *P*<sup>s</sup> and water solubility *S* of the pure compound, provided that *P*<sup>s</sup> and *S* refer to the same physical state (liquid or solid) and to the same temperature *T*. The air-water concentration ratio *K*AW can be derived from Henry's law constant by reworking it into a "dimensionless" partition coefficient. Dimensionless air-soil concentration ratios can be obtained in the same way:
$$K\_{AW} = \frac{C\_{air}}{C\_{water}} = \frac{H}{RT} = \frac{P\_{L,S}^s}{S\_{L,S} \cdot RT} \tag{6}$$
where
*R* = gas constant (8.314 Pa.m3 /mol/K)
*T* = temperature at the air-water interface (K)
Air-water equilibrium constants of chemicals *KAW* range from < 10-10 to > 1 (Table **1**).
**Table 1:** Typical environmental values of intermedia partition parameters for selected chemicals.
1 "Dimensionless" ratios to be read as mmedium1<sup>3</sup> .mmedium2-3; for K-values, and as "molphase1/molphase2" for -values;
2 numbers in parentheses calculated according to Eq **8**; 3 >>: very large; <<: very small
# *Air-Aerosol Equilibrium*
The extent of association of chemicals with the aerosol phase of air is known to be inversely related to the chemical's vapour pressure. The fraction associated with the aerosol phase *aerosol* has successfully been described by Junge [6] with
$$\Phi\_{around} = \frac{c\Theta}{P\_L^t + c\Theta} \tag{7}$$
where
= aerosol surface area per volume unit of air (m2 /m3 )
*Ps <sup>L</sup>* = vapour pressure of the pure compound in the liquid state (Pa)
$$\mathcal{C} \quad \text{--- constant (Pa.m)}$$
The constant c depends on the heat of condensation and molecular weight for many organics, its value being approximately 0.17 Pa.m. The local pollution climate determines the aerosol surface density. A typical value for aerosol surface area under rural conditions is 3.5 x 10-4 m2 /m3 . For more polluted urban or industrialized areas is estimated to be 1.1 x 10-3 m2 /m3 . Substitution of these values in Eq **7** shows that gas-particle partitioning is important for organic compounds with a *Ps L* lower than approximately 10-3 Pa. Since *Ps <sup>L</sup>* is strongly temperature dependent, the fraction of a substance absorbed to particles will also be temperature dependent. For certain organics this may imply that in tropical regions the pollutant will be in the gas phase, whereas in arctic regions it will be in the particle phase. More recently, it has been shown that for many (hydrophobic) chemicals, the octanol-air partition coefficient *K*OA is a more accurate descriptor of aerosol-air partitioning [7].
$$\Phi\_{\text{gas}} = \frac{1}{1 + K\_{OA} \cdot \left(B \cdot T \text{SP}\right)} = \frac{1}{1 + \left(K\_{OW} \mid K\_{AW}\right) \cdot \left(B \cdot T \text{SP}\right)}\tag{8}$$
where
*B* = chemical- specific constant (= ~2x10-12 m3 g)
*TSP* = mass concentration of aerosol in air (= ~50 g/m3 )
Eq **8** predicts significant partitioning to the aerosol phase for chemicals with *KOA* greater than approximately 1010, which is the case for many polyaromatic cyclic hydrocarbons PAH (Table **1**).
# **TRANSPORT MECHANISMS**
Two kinds of transport mechanisms may be distinguished: (1) *intramedia transport*, which is transport away from a source in one environmental medium, and (2) *intermedia transport*, which is transport from one environmental medium to another. Intramedia transport is important in relation to the mobile environmental media: air, water and groundwater; intermedia exchange takes place between all media, but is most important for transport of chemicals to and from the stationary media: sediment and soil.
**Intramedia transport** takes place through the mechanisms of advection and dispersion. Advection causes a chemical to travel from one place to another as a result of the flow of the medium in which it occurs; locally emitted packages or "puffs" of a chemical are carried as far as the wind or water current can take it during the residence time in that medium. Dispersion mechanisms (molecular diffusion, eddy diffusion) make the chemical move down concentration gradients until the gradients have disappeared. The residence time of the chemical in the medium is an important factor since besides this other removal processes occur at the same time. If, for example, a chemical is emitted into air and its degradation in air is rapid, the effective residence time of the chemical in air is short. Consequently, there is little time for the advective and dispersive processes to take place. In one medium, advection and dispersion always operate together. If a chemical is emitted continuously into air or water, the combined operation of advection and dispersion results in the formation of a plume. At short distances from emission sources, concentrations are usually affected most by intramedia transport. Intramedia transport of chemicals is observed as dilution, which in many situations is the most important process affecting environmental concentrations of chemicals. In this chapter, we shall account for the effect of intramedia dilution on the concentrations of chemicals by lumping it into rate constants for advective and dispersive loss of chemicals due to transport. However, explanation of the complex aerodynamic and hydrodynamic processes of advective and dispersive spread of chemicals within air, surface water and groundwater will not be discussed here. Interested readers are referred to specialized text books on mathematical modelling of air, surface water and groundwater [8].
**Intermedia transport** (air-water, water-sediment, *etc.*) also takes place by advective and dispersive mechanisms. Advective intermedia transport takes place if a chemical is transported from one environmental compartment to another by a physical carrier. Examples are deposition of fog, raindrops and aerosol particles from air to water or soil, sedimentation and resuspension of particulate matter across the water-sediment interface, and percolation of water through soil. Advective transport is a one-way phenomenon: the chemical is carried by the medium in which it resides in the direction of the medium flows. Intermedia dispersion is also diffusive in nature and follows concentration gradients. Examples are volatilization and gas absorption (air-water and air-soil), the direction depending on the concentration difference between the media, and diffusive exchange of chemicals between sediment and water. The driving force of intermedia transport is the tendency of chemicals to seek equilibrium between different phases. Transport from one environmental medium to another is commonly described by taking the box/compartment modelling approach. Theoretical backgrounds and detailed quantitative descriptions of intermedia exchange processes can be found in other texts [2, 9-13]. The most important interfaces are described below.
# *Air-Water and Air-Soil Exchange*
Atmospheric deposition and volatilization processes transport chemicals between air and the earth's surface. It is customary to distinguish between wet (precipitation-mediated) deposition mechanisms and dry deposition mechanisms (Fig. **2**). Wet deposition is further split into rain-out (in-cloud processes) and wash-out (below-cloud processes). Dry deposition is the sum of aerosol deposition and gas absorption. In multimedia environmental
chemistry, the latter mechanism is usually treated as one part of a bi-directional exchange mechanism. Rain-out, wash-out and aerosol deposition are one-way advective transport processes: the chemical is carried from the atmosphere to water and soil. This is true even if the chemical has a greater fugacity in water or soil. Gas absorption is a diffusive mechanism. There is only net absorption of chemicals from the gas phase by water or soil if the fugacity in air is greater than the fugacity in water or soil. If the fugacity in water or soil is greater, the result will be the reverse: net volatilization. This will generally be the case if a chemical is emitted to water or soil, in which cases fugacities in these media will be highest of all. Deposition from air to water and soil occurs at all times, even when net volatilization occurs (see below). It should be noted that absorption and volatilization occur simultaneously, and it is the net difference that accounts for the effective intermedia transport.
**Figure 2:** Exchange mechanisms between atmosphere and the earth's surface. From Schwarzenbach [14], as referred to by Sijm *et al.* [15].
# *Deposition with Aerosol and Rain*
Chemicals adsorbed to aerosol particles are carried from the air compartment to the earth's surface by dry particle deposition. Aerosol particles can also be scavenged by rain drops as wet particle deposition. In addition, rain drops absorb chemicals from the gas phase and carry chemicals to the earth's surface by rain-out and wash-out.
Deposition rates depend on the physical parameters of the particle, of which the size is most important. Small particles tend to behave like gases; larger particles (> 2 µm) are efficiently removed from the atmosphere by deposition under the influence of gravity. Inertial impaction is important for particles with a diameter of between 0.1 and 10 µm. This effect greatly depends on the velocity of the air and the intensity of the turbulence, which varies with the properties of the landscape. Larger particles (> 10 µm) are deposited primarily by sedimentation and chemicals associated with them will, in general, be deposited close to the source. Typical aerosol deposition velocities range from 10-4 to 10-2 m/s. Some chemicals are associated predominantly with the larger, rapidly depositing particles, whereas other chemicals bind predominantly to the smaller particles and stay airborne for much longer times.
The efficiency of wet deposition varies greatly. It depends on meteorological factors such as the duration, intensity and type of precipitation (snow, rain, hail), as well as on the size and the number of droplets. Other specific parameters, like solubility in rain and snow, are important too. Wash-out is an efficient removal mechanism for chemical substances with low Henry's law constants, and for aerosols with a diameter greater than 1 µm. For less volatile chemicals (high Henry's law constants) the falling droplet will absorb only a very small amount of the compounds below the cloud. Wash-out plays an important role when concentrations below the cloud are much higher than the concentrations in the cloud, e.g., for plumes close to the source. In clouds the uptake of aerosols by cloud droplets is a very efficient process. For most purposes, it is sufficient to assume that the rain phase is in equilibrium with the gas phase. The extent of gas scavenging by falling rain drops can then be calculated from the air-water distribution ratio KAW and the rain intensity. As a practical approach to estimating the extent of aerosol
scavenging, Mackay [2] has suggested that during rainfall in the atmosphere, each drop sweeps through a volume of air about 200,000 times its own volume. First-order rate constants for removal of chemicals from air by atmospheric deposition are given in Table **2**.
**Table 2:** Influence of transport- and transformation processes on concentrations in AIR for selected chemicals, after a steady state has been reached.
# *Volatilization and Gas Absorption*
Transport of a chemical from water and soil into the gas phase of air and vice versa is commonly described by a two-resistance approach, as originally introduced 110 years ago by Whitman [16]. In this concept, the resistance to intermedia transfer is considered to be concentrated in two thin films on either side of the interface. Transport through this interfacial double layer has to take place by molecular diffusion and is, therefore, slow in comparison with transport to and from the interface. This concept was used by Liss and Slater [17] as a basis for modelling the transfer of gases across the air-sea interface. The direction of transport depends on the concentrations in air and water. In fugacity terminology: the net diffusion is from the compartment with the highest fugacity to the compartment with the lowest fugacity. The rate mass-transfer (depending on the direction referred to as either gas absorption or volatilization) is expressed by means of an "overall" mass-transfer coefficient *k OV*, in which transfer resistances on either side of the air-water interface are accounted for, can be expressed in terms of air- or water concentrations. The mass-transfer coefficient (m/s) can be looked upon as the velocity of a piston, pushing the chemical through the interface. The driving force can be positive or negative, leading to absorption or volatilization. The overall mass transfer coefficients *k OV* represent the resistance to mass transfer: the greater *k OV*, the smaller the resistance. The magnitude of *k OV* derives from the partial mass transfer coefficients of the stagnant films at either side of the air-water interface, through which the chemical must diffuse, and the intermedia equilibrium constant *KAW* can be estimated as:
$$\begin{aligned} k\_{\dot{a}\dot{r}}^{OV} &= \frac{\text{kaw}\_{\dot{a}\dot{r}} \cdot \text{kaw}\_{\text{water}}}{\text{kaw}\_{\dot{a}r} \cdot K\_{AW} + \text{kaw}\_{\text{water}}}\\ k\_{\text{water}}^{OV} &= \frac{\text{kaw}\_{\dot{a}r} \cdot \text{kaw}\_{\text{water}}}{\text{kaw}\_{\dot{a}r} + \text{kaw}\_{\text{water}} / K\_{AW}} \end{aligned} \tag{9}$$
where
*kawair* = partial mass-transfer coefficient air side of the air-water interface (m/s)
*kawwater* = partial mass-transfer coefficient water side of the air-water interface (m/s)
The partial mass transfer coefficients *kaw* represent the resistances of the stagnant air and water films. Thick films have greater resistances than thin films; substances with large diffusivities (small molecules) have smaller
resistances than substances with small diffusivities (big molecules). Since diffusivities of substances do not differ much, differences in *kaw* originate mainly from differences in thickness of the stagnant film through which the molecules must diffuse. Film thicknesses vary from water body to water body due to differences in turbulence. Typical values of *kaw* are 10-3 and 10-5 m/s for air and water films, respectively. As a result, differences in volatilization between substances arise from differences in the air-water equilibrium constant *KAW*. It can be seen from Eq **9** that *k OV* of water-loving chemicals (*KAW* << ~10-2) is proportional to *KAW*; resistance to volatilization for such chemicals originates entirely from slow diffusion through the air film. Air-loving chemicals (*KAW* >>10-2) volatilize independently of *KAW*; resistance to volatilization for such chemicals is limited only by slow diffusion through the water film. Advanced readers are referred to specialized textbooks on this subject [9, 10].
When the concentration in air is negligibly small, the net rate of volatilization depends only on the concentration in water, so that volatilization acts as a first-order removal process from water. Applying the mass balance concept of Eq **1**, it follows that
$$\begin{aligned} \text{VOL} & \approx A \cdot k\_{\text{water}}^{OV} \cdot C\_{\text{water}} = k\_{\text{vol}} \cdot V\_{\text{water}} \cdot C\_{\text{water}},\\ \text{with} \quad k\_{\text{vol}} &= \frac{A \cdot k\_{\text{water}}^{OV}}{V\_{\text{water}}} = \frac{1}{D\_{\text{water}}} \cdot \frac{\text{kaw}\_{\text{air}} \cdot \text{kaw}\_{\text{water}}}{\text{kaw}\_{\text{air}} + \text{kaw}\_{\text{water}} / K\_{AW}},\end{aligned} \tag{10}$$
where
*kvol* = first-order rate constant for removal from water by volatilization (1/s)
*Vwater* = volume of the water compartment (m3 )
*Dwater* = depth of the water compartment (m)
Note that, while the volatilization rate (mol/s) depends on the area *A* of the water compartment, the effect that this volatilization has on the concentration *Cwater* depends on the depth *Dwater* of the water compartment. Similarly, gas absorption with negligible concentration in water acts as a first-order removal process from air. The rate constant for removal from air by gas absorption is left to be worked out by the reader. Volatilization from and gas absorption to soil can be deduced along the same lines and is not treated here. First-order rate constants for exchange of some chemicals between air and the Earth's surface are given in Tables **3** and **4**.
**Table 3:** Influence of transport and transformation processes on concentrations in WATER for selected chemicals, after a steady state has been reached.
**Table 4:** Influence of transport and transformation processes on concentrations in SOIL for selected chemicals, after a steady state has been reached.
# *Soil Run-off*
Part of the rainwater that reaches the soil runs off to surface water, *i.e.* rivers, estuaries and coastal waters. In urban areas, where most of the surface is paved, nearly all the precipitation is collected in sewerage systems, from where it may either be redirected to a waste water treatment facility or discharged into surface water. In rural areas the rainwater runs off directly into the surface waters. With the run-off, soil particles are washed away (eroded). Chemicals dissolved in water or associated with the soil particles, are transported by these mechanisms from soil to water. Assuming the water which runs off from soil is in equilibrium with the soil, the mass flow of a chemical resulting from run-off can be quantified. However, for most chemicals it is often more practical and accurate to conduct field measurements on contaminated sites than applying models.
Rates of net precipitation and fractions of water that run off and infiltrate into the soil are often well known from meteorological monitoring. Rates of soil erosion are much harder to obtain. There is extensive literature on the dependence of soil erosion on rainfall and terrain conditions (e.g. slope), which is not treated here [18]. First-order rate constants for removal of some chemicals from soil by run-off to surface water are given in Table **4**.
# *Deposition and Resuspension of Sediment Particles*
The transport of chemicals across the sediment-water interface can be treated in the same manner as air-water and airsoil exchanges. In this case there is an advective transport component *i.e.*, sedimentation and resuspension, and a diffusive transport component, *i.e.*, direct adsorption onto and desorption from the sediment. To estimate the rate of advective transport from water to sediment by sedimentation of suspended particles, we need to know the concentration of the chemical on the particles and the rate at which they settle. For most purposes it is sufficient to assume equilibrium between the suspended particles and water phase. The concentration in the particles is then proportional to the concentration in water and the can be derived by means of Eq **5.** Settling rates of sediment particles can be obtained from field- or laboratory measurements, or can be estimated by theoretical means. Resuspension of freshly deposited sediment counteracts this removal from water. Resuspension rates are usually not known and must be derived as the difference between the sedimentation rate and the net sediment growth rate, which can be measured in the field, or deduced from mass balance calculations of incoming and outgoing sediment loads.
# *Exchange between Water and Sediment by Direct Adsorption and Desorption*
Diffusive transport between sediment and water, by direct adsorption and desorption across that interface, is analogous to diffusive transport across the air-water and air-soil interfaces and can be described with a two-film resistance mechanism. A value of ~3 x 10-6 m/s (0.01 m/h) may be taken for the mass-transfer coefficient on the waterside of the sediment-water interface *kwsdwater* [19]. The mass-transfer on the pore water side of the sedimentwater interface *kwsdsed* can be treated as molecular diffusion in the aqueous phase of a porous solid material, characterized by an effective diffusivity of 2 x 10-6 m2 /h and a diffusion path length of 2 cm. This gives *kwssed* a value of ~3 x 10-8 m/s (0.0001 m/h). It should be noted, however, that additional processes that are typically of a non-equilibrium nature, may greatly affect the net mass-transfer of all kinds of chemicals. For instance bioturbation and shipping can play a key role in the sediment side resistance, essentially eliminating it in some cases. As the extent of bioturbation is not governed by thermodynamic principles, and as, in general, very limited information is available on this and similar topics, it will not be extensively discussed here. Instead, readers are referred to the textbook by Thibodeaux [10]. First-order rate constants for exchange of some chemicals between water and sediment are given in Table **3**.
# *Removal by Transport*
Transport processes result in translocation of chemicals in the environment, but do not lead to their elimination. The advective exchange processes between environmental media as discussed above (e.g., sediment-water exchange by sedimentation and resuspension), are examples of such non-eliminating transport processes. Advective transport also occurs within environmental media (air, water). It is common to consider only a part of the environment, e.g., a world region or a layer of air, water or soil. In these open systems, transport does remove chemical from the environment. Transport across the system boundaries has the same effect on the concentration of chemical inside the system as real elimination (e.g., by chemical reaction).
Wind carries airborne chemicals out of the region considered to other parts of the world; water currents do the same function for waterborne chemical. First-order removal rate constants for the transport by advection processes that cause concentrations in air and water to decrease can be formulated by considering all incoming and outgoing air and water flows in relation to the volumes of their compartments.
Transport from the upper layer of the soil to the groundwater takes place through leaching with percolating water. If we choose to exclude groundwater from the system considered, soil leaching should be regarded as elimination from the system. Background information on transport in porous media can be found in Spitz and Moreno [20] and will not be considered in detail here. In many approximations, the process of soil leaching is simplified by assuming equilibrium between the solid phase and pore water phase at all times and in all places. It is clear that leaching is an important factor for chemicals with a small *K*p value. Analogous transport phenomena take place in sediment. Surface water may seep into the sediment, thereby carrying the chemical from the upper sediment layer down and vice versa.
An additional phenomenon occurs in areas where there is continuous sedimentation. In this situation sediment is continuously being buried under freshly deposited material. It is common practice (e.g., in water quality management) to consider the (mixed) top few centimetres of sediment only. Regarded this way, the concentration of chemical in this sediment top layer can best be understood by regarding sediment burial as a mechanism that removes chemical from the top layer, transporting it to the deeper sediment. First-order rate constants for removal of some chemicals from air, water, sediment and soil by advection, burial and leaching are given in Tables **2-4**.
# **Effect of Transport Processes on Concentrations of Chemicals in the Environment**
As explained earlier (Fig. (**1**), Eqs **1** and **2**), concentrations of chemicals in the environment are controlled entirely by the transport and transformation mass flows of chemical into and out of environmental compartments. As discussed above, transport rates vary greatly between chemicals in a way that can be understood and predicted from the differences in physical-chemical properties. In addition, transport rates vary with environmental conditions (geometry, temperature, wind, precipitation). The combined effect of all transport and transformation processes on concentrations of chemicals can be evaluated by comparing the rate constants for net transport from the compartment of interest. Tables **2-4** list the most important first-order rate constants for removal from (and addition to) air, water and soil, respectively, for a selection of different chemicals, together with the half-lives of change in concentration due to the process. For comparison, rate constants and half-lives for degradation (to be discussed in the following section) and rate constants and half-lives for all combined (net) losses are given. The process data in Tables **2-4** were derived from calculations with the multimedia fate model SimpleBox [12, 13], parameterized to reflect the regional spatial scale of EUSES.
It can be seen that
# **TRANSFORMATION PROCESSES**
Following its release into the environment, a chemical may undergo various biotic and abiotic processes which modify its chemical structure. Degradation or transformation of a compound refers to the disappearance of the parent compound from the environment by a change in its chemical structure. When this change is brought about by micro-organisms, the degradation process is called primary biodegradation or biotransformation. When chemicals are converted entirely to simple molecules and ions, such as carbon dioxide, methane, water and chloride, biodegradation is referred to as mineralization. Transformation of chemicals in the environment can also occur by abiotic processes. Four categories of abiotic transformation processes are distinguished:
Transformation and mineralization processes alter the physicochemical and toxicological properties and can reduce exposure concentrations of chemicals in the environment. The rate of degradation of a specific chemical depends on its intrinsic sensitivity to undergo chemical transformation (reactivity), the presence of reactants and the availability of the chemical to undergo reaction, *i.e.* the presence of the chemical in the gas phase of air or dissolved in water. Generally, the availability and reactivity of both the chemical and the reactant depend to a large extent on environmental conditions like pH, temperature, light intensity and redox conditions.
# **Hydrolysis**
In a typical hydrolysis reaction a hydroxyl group replaces another chemical group in a molecule. However, certain functional groups, including alkanes, alkenes, benzenes, biphenyls, (halogenated) polycyclic aromatics (e.g., PAHs and PCBs), alcohols, esters and ketones, are often inert to hydrolysis.
The importance of hydrolysis stems from the fact that the products formed are more polar and, consequently, more water soluble and less lipophilic than the parent compound. Hydrolysis reactions are commonly catalysed by hydrogen or hydroxide ions. Because the concentrations of hydrogen ion [H+ ] and hydroxide ion [OH- ] change with the pH of the water, the rate of hydrolysis directly depends on the pH. Hydrolysis rate constants *kh*, which generally obey pseudo first-order kinetics, are measured experimentally in laboratory tests, in which a known quantity of the compound is introduced into a solution of fixed pH and the disappearance of the compound is followed over time. As in Eq **1** and Fig. (**1**), the mass (and hence the concentration) of the chemical typically declines exponentially with increasing time. When plotted logarithmically, the loss rate constant is observed as the slope of the concentration-time plot
$$\ln\left(C\_t \mid C\_0\right) = -k\_h \cdot t$$
where
From the results of a series of such experiments at different pH levels, a pH rate profile can be constructed by plotting the logarithms of the observed rate constants as a function of the pH of the experimental solutions. Fig. (**3**) shows the pH rate profile of the hydrolytic transformation of phenyl acetate to yield acetic acid and phenol. Under acid conditions (pH < 3), specific acid catalysis is the predominant mechanism. In this pH region, the logarithm of *k*obs decreases by a unit slope -1 with increasing pH. At less acidic pH (pH > 4), the hydrogen ion concentration is so small that the specific acid catalysed hydrolytic reaction is too slow to be seen in the profile. Between pH 4 and 6, the neutral mechanism (independent of pH) predominates. Finally, at pH > 8, due to base catalysis, an increase of *kh* directly proportional with increasing OH concentration, becomes visible.
**Figure 3:** Hydrolysis pH rate profile of phenyl acetate. From Burns and Baughman [21] and Mabey and Mill [22], as referred to by Sijm *et al.* [15].
# **Oxidation**
Oxidation is the chemical process in which an electron-deficient particle (the oxidant) accepts electrons from the compound to be oxidized. Examples of oxidants that occur under environmental conditions in sufficiently high concentrations and also react rapidly with organic compounds are: alkoxy radicals (RO**.** ), peroxy radicals (RO2 **.** ), hydroxyl radicals (HO**.** ), singlet oxygen (<sup>1</sup> O2) and ozone (O3).
Most of these oxidants are directly or indirectly generated from chemicals that interact with solar radiation, forming an "excited state" of the molecule; oxidation with photochemically formed reactive oxidants is usually referred to as photo oxidation. Oxidations are the main transformation routes for most organic compounds in the troposphere and also transform various micro pollutants in surface waters [23]. Most radical oxidants exhibit similar chemistry for aliphatic and aromatic structures.
Although many different kinds of RO2 **.** or RO**.** radicals may be present in a natural system, the simplifying assumption can be made that the structure of R has little effect on its reactivity [24]. Rate constants for reactions of most radical oxidants are known for a large number of organic molecules. The concentrations of the major oxidants
in less heavily polluted aquatic and atmospheric systems are also known. By combining these data it can be derived that, in general, the hydroxyl radical is the only oxidant of importance in atmospheric systems. In aquatic systems the concentration of.OH is so low that its contribution is negligible compared with RO2 **.** or RO**.** . To illustrate the differences in reactivity of the hydroxyl radical to various organic chemicals, the half-lives for gas-phase oxidation of various classes of chemicals in the northern hemisphere are given in Table **5**.
**Table 5:** Half-lives (days) for tropospheric oxidation of various classes of organic compounds in the northern hemisphere.
From this table it is clear that chlorofluorohydrocarbons (CFCs or halomethanes), in particular, may remain in the troposphere for prolonged periods of time. This enables them to reach the stratosphere, where they pose a threat to the ozone layer.
# **Reduction**
Reduction is the chemical process by which electrons are transferred from an electron donor (reductant) to the compound to be reduced. The redox half-reactions leading to reduction of a 1,2-substituted alkane are shown as a diagram in Fig. (**4**). In this example, Fe2+ is used as the reductant. Following the transfer of 2 electrons from 2 molecules of Fe2+ to the halogenated compound, Fe3+, the free halide ion and the product of reduction (in this case ethene) are formed.
It has been shown that reductive reaction pathways can contribute significantly to the removal of several micro pollutants. Nitro aromatics, azo compounds, halogenated aliphatic and aromatic compounds (including PCBs and even dioxins) can be reduced under certain environmental conditions [25].
Reduction can take place in a variety of reducing (non-oxic) systems, including sewage sludge, anaerobic biological systems, saturated soil systems, anoxic sediments, reducing iron porphyrin systems, solutions of various chemical reagents, as well as in the gastronomic tract of invertebrate species. It has also been shown that the reduction rate of specific halogen compounds depends on environmental factors, such as the prevailing redox potential, temperature, pH and the physical and chemical properties of the micro pollutant to be reduced. As in hydrolytic transformation, usually more polar products are formed from the parent compound by reduction, which makes them more susceptible to further chemical attack and less likely to accumulate.
At present, insufficient information is available on the nature of the reductants responsible for the main reductive transformations in natural systems. Nevertheless, it has been shown in most studies that reductive transformations generally follow pseudo first-order reaction kinetics.
# **Photochemical Degradation**
Interaction with sunlight can initiate a wide variety of photolytic processes. The primary requirement for photochemical processes is the penetration of radiation (light, in particular UV light) in aqueous and atmospheric environments. Various categories of photochemical conversions can be distinguished:
Following absorption of a photon by a compound, the photon energy either needs to be transferred to the reactive site within the molecule or transferred to another molecule, which may subsequently undergo a photochemical transformation. Not every photon that is absorbed by a molecule induces a chemical reaction. The proportion of absorbed photons which causes reaction is the quantum yield, a number between 0 and 1. Quantum yields may vary largely, depending on the chemical structure of the molecule. In direct photoreactions, the reaction rate is proportional to the absorption of light at a specific wave length and the quantum yield. The rate of light absorption depends on the light intensity and the specific absorptivity (molar absorption coefficient) of the chemical. Both the molar absorption coefficient and the quantum yield are intrinsic properties of the chemical; light intensity is a property of the environment. Since the rates of all photochemical reactions are proportional to light intensity, it is evident that the significance of the photo-transformation of a certain chemical will change with time and place. In this process factors such as time of the day or year, location (climate) and weather (cloud cover) play a major role.
In the aquatic environment, an important fraction of sunlight is absorbed by dissolved and particulate matter. This clearly reduces the rates of direct photo-transformation, and changes the solar spectrum in deeper water layers. However, this dissolved and particulate matter is also capable of initiating indirect photo-conversions. Given the complexity of these indirect conversions, and the many variables that influence the rate of indirect photolysis, it has so far only been possible, to a limited extent, to derive general, mathematical equations for rate constants in natural water systems.
Given the various direct and indirect transformations that can take place due to interaction with solar radiation, a variety of primary and secondary photoproducts is often observed. Since penetration of light is usually only possible in oxic systems, most photo-products formed are in an oxidized state, compared with the parent compound.
# **Biodegradation**
For most xenobiotic organic chemicals, microbial degradation plays a key role in their removal from the environment. By contrast with non-biological elimination processes such as hydrolysis or photochemical degradation, biodegradation in the oxygen-containing biosphere is, generally, equivalent to conversion into inorganic end-products, such as carbon dioxide and water. This has been named ultimate biodegradation or mineralization and may be regarded as a true sink in aerobic compartments. In the anaerobic environment, microbial degradation processes are generally much slower and may not always result in complete mineralization. Transformation of the parent compound into another organic product (metabolite) is often referred to as primary degradation.
Heterotrophic micro-organisms are characterized by a high catabolic versatility. Mixed micro floras, rather than monocultures, are responsible for the elimination of substances from the biosphere, and because adaptation of the microbial ecosystem to a xenobiotic compound is so important, a more operational definition would be useful. Adaptation can be described as a change in the microbial community that increases the rate of biodegradation of a chemical as a result of prior exposure to that compound. This definition does not distinguish between mechanisms such as gene transfer or mutation, enzyme induction and population changes. The enzymatic machinery of microorganisms consists of constitutive enzymes, which are involved in fundamental metabolic cycles (e.g., hydrolysis), and adaptive or induced enzymes. These enzymes enable bacteria to utilize organic compounds which are not appropriate for immediate use.
Environmental factors affect the population distribution and biochemistry of bacteria. Sediment and soil are more or less aerobic unless the oxygen consumption by micro-organisms, due to an abundance of substrate, is higher than the oxygen supply by diffusion. Aerobic bacteria use oxygen both as a reactant for the oxidation of organic compounds, and as a terminal electron acceptor. The latter is necessary for the conversion of the organic compound into carbon dioxide. This reaction, also known as dissimilation, produces the energy required during the formation of biomass from the organic compound (assimilation). Facultative anaerobic bacteria use oxygen but have the capability to change to another electron acceptor if their environment turns anaerobic. Other electron acceptors are nitrate, utilized by denitrifying bacteria and sulphate, used by sulphate-reducing bacteria particularly in marine and wetland environments. Oxygen is very toxic to the obligate anaerobic bacteria, which can only use alternative electron acceptors. The methanogens (methane-producing bacteria) derive energy from the conversion of hydrogen and carbon dioxide into methane. The considerable decrease in energy supply by the different electron acceptors from oxygen to the organic compound itself explains why microbial processes are faster in the aerobic world.
Biodegradation of synthetic chemicals does not always result in bacterial growth. When exponential growth does not occur the degradation process is called co-metabolism, in which micro-organisms - while growing on another, widely available, substrate - also have the capacity to transform other compounds (xenobiotics) without deriving any benefit from that transformation [26].
# *Biodegradation Kinetics*
In first approximation, removal of chemicals from the environment by microbial degradation can be treated similar to other removal processes, *i.e.* by describing the process as following pseudo first-order kinetics and by formulating pseudo first-order reaction rate constants for it. In fact, the so-called second-order rate concept of microbial degradation in water was proposed as early as 1981 [27]. More recently, this almost forgotten concept has been reintroduced to quantitatively treat the subject of persistence of chemicals in the environment [28]. Following the assumption that the removal by microbial degradation obeys pseudo first-order kinetics, the pseudo first-order rate constant postulated to be proportional to the concentration of bacteria in the system:
$$k\_{bio} = k^{2nd} \cdot \lfloor BAC \rfloor \tag{12}$$
where
*kbio* = pseudo first-order rate constant for biodegradation (1/s) *k 2nd* = second-order rate constant for biodegradation (L/CFU/s)
[*Bact*] = number of colony-forming units of bacteria in water (CFU/L)
Although of attractive conceptual simplicity, this approach has limited predictive power. Quantitative bacterial counts can be made, but the degrading power of the microbial colonies is hard to asses or predict; the second-order rate constants obtained from field observations or laboratory experiments are not nearly as constant as required for extrapolating observed biodegradation rates to other environmental situations, let alone to other chemicals. Reality of biodegradation kinetics is complex.
# *Biodegradability and Biodegradation Rates*
Biodegradation rates are hard to predict. Despite major efforts, it has so far proved difficult to formulate generally applicable predictive theory, even for aerobic biodegradation in water and soil. At present, the re-interpretation of experimental studies is the only way to estimate rates of aerobic biodegradation. Most experimental data on microbial degradation originate from the standard tests of biodegradability of chemicals, as required by many regulatory agencies, e.g., the European Chemicals Agency ECHA. Biodegradability testing is commonly done according to standard methods published by the OECD [29, 30]. In the OECD hierarchy, three different levels of testing are distinguished as follows:
# *Recalcitrance*
So it seems that microbial communities in the natural environment are catabolically so versatile, that always one or more species capable of degrading any chemical is present in a specific environment. Why then do some man-made chemicals persist in the environment for such a long time? The rate and extent of biodegradation of a chemical depends on both its chemical structure and the prevailing environmental conditions. In general, the following properties or conditions have a significant influence on the biodegradation of synthetic chemicals:
Chemical structure. Type, number and position of substituents on aliphatic or aromatic structures may cause "violation of comparative biochemistry and enzyme specificity", as described by Alexander [31]. Effects of substitution of radicals have already been discussed in the three examples of major metabolic pathways for biochemical oxidation; aromatic rings, however, are hard to break and substitute. The influence of the molecular structure on its biodegradability in the aerobic environment is clear.
Environmental conditions. Temperature is an important factor and especially around and below 4°C, microbial processes become very slow. The optimum temperature for psychrophilic (cold-loving) bacteria is between 0 and 20°C and for mesophyllic (moderate temperature loving) bacteria it is between 20 and 40°C. In seawater 15°C is the borderline between different microbial ecosystems. The inorganic nutrient status of the surface water affects the biodegradation rate and in some coastal waters may even exceed the temperature effect. The presence of auxiliary organic nutrients may also play a role, and the occurrence of co-metabolism has already been mentioned. Failure of biodegradation may be due to the presence of other, more easily degradable compounds used in preference to the specific xenobiotic compound. This phenomenon is known as diauxism. Unlike seawater, which is a well-buffered system of pH 8, inland waters can vary up to 5 pH units in acidity, thereby determining the form in which some chemicals exist. The availability of some natural organic substrates may also facilitate co-metabolism of the pollutant. However, even if it were possible to find two aquatic ecosystems characterized by similar environmental parameters, the outcome of a biodegradability experiment might be quite different for the same chemical. The presence and influence of high population densities of "specialized" degraders is evident. Some aquatic ecosystems may have been previously exposed to a chemical or another pollutant which shares a common enzyme system of such a specific degrader. The presence and density of specific degraders is often highly decisive for biodegradation to occur within a limited period of time.
Bioavailability. If a chemical is trapped in micro sites, e.g., in inorganic material such as clay minerals or the organic matrix of sediment or soil, interaction with micro-organisms may be physically impossible, which impedes biodegradation.
# *Biodegradation in Sediment and Soil*
Biodegradation in sediment or soil is commonly reported to obey first-order kinetics. Experimentally observed halflives and first-order rate constants can be found in the literature for many chemicals in many sediment and soil systems. It is often claimed that biodegradation in sediment- and soil systems is described and explained best from the theory that degradation takes place entirely in the water phase; chemical bound to the solid phase is considered unavailable for attack by microbes and, therefore, non-reactive. Several studies have provided evidence that a chemical associated with sediment or soil particles is not available for biodegradation because micro-organisms only utilize dissolved chemicals [32]. Therefore, the overall rate of biodegradation in a solids-water system greatly depends on the extent of partitioning to the solid phase:
This again illustrates the difficulties associated with extrapolation of a laboratory-derived degradation rate to an environmental half-life of a laboratory-derived degradation rate. When assessing the environmental risk of a chemical, it is important to realize that even a relatively easily biodegradable chemical can become more or less persistent when it ends up in an environmental compartment where its bioavailability becomes limited.
# **MODELLING CONCENTRATIONS IN THE ENVIRONMENT**
Distribution of chemicals in the environment can be determined by measurement of concentrations in air, water, soil and sediments. Such measurements are expensive, since they usually include analyses of many chemicals at different places and usually over long periods of time. Modelling the fate of chemicals in the environment is a feasible alternative which renders similar results at much cheaper cost. Moreover, modelling is necessary when measurement is no option, e.g., when predictions are to be made of expected results of environmental management measures, or when making predictions for new chemicals that have not been released yet, so they can be used in regulatory risk assessment.
Many of the models used in risk assessment of toxic substances are compartment models, also referred to as box models or mass balance models. The environment is thought to be made up of homogeneous, well-mixed compartments. Compartments can represent segments of the environment, or even entire environmental media. Examples of the former are the spatially segmented air and water transport models and layered soil models. The latter is used in multimedia (air, water, soil, *etc.*) fate models and in physiology-based pharmaco-kinetic models (blood, tissue, *etc.*). Compartment models apply the principle of mass conservation: the mass of a substance in a compartment appears or disappears only as a result of mass flows of a substance into or out of the compartment. What compartment models have in common is that the mass balance equation is used as their basic instrument. Because mass balance modelling is used so widely in the environmental risk assessment of toxic substances, its principles will be explained here. We shall first derive a mass balance equation for one compartment, then a mass balance model for more compartments.
# **One Compartment**
If a substance is added to or taken from a compartment, the mass of that substance in the compartment changes. This change can be quantitatively expressed in a mass balance equation, in which all incoming and outgoing mass flows of the substance are accounted for
$$\frac{\Delta M}{\Delta t} \left(=\text{V}\frac{\Delta C}{\Delta t}\right) = \text{gains} - \text{losses} = \sum \text{mass flows} \tag{13}$$
where *ΔM* and *ΔC* and are changes in mass and concentration within a time interval Δt, respectively, and V is the (constant) volume of the compartment. Note that the change is in unit mass per unit time (e.g., kg/s): a sum of mass flows. If nothing is added or taken away, or if gains and losses match exactly, the mass of substance in the compartment does not change: a steady state. If *ΔM*, *ΔC* and *Δt* are infinitesimally small, Eq **13** becomes what is
mathematically known as a differential equation. Differential equations describe at what rate a variable (here: mass of a substance in a compartment) changes. If the mass at starting time (*t*=0) is known (the initial condition), a differential equation can be used to derive the mass at other times. The art of mass balance modelling is thus to properly quantify the mass flows of a substance going into and out of the compartments. For the purpose of mass balance modelling it is useful to distinguish between mass flows that take place independently of what happens in the compartment and mass flows that do depend on the conditions within the compartment. Emissions and imports are examples of the first category. The rate at which mass is brought into the compartment by these processes may be constant or time-dependent, and may relate to the mass of a substance outside the compartment, but bears no relationship to the mass of a substance within the compartment. These mass flows need to be specified to the model as so-called "forcings". If a constant emission of E (kg/s) is forced upon a compartment, which contains M0 kg of the substance at t = 0, and nothing else happens, the mass balance equation becomes:
$$\frac{dM}{dt}\left(=\mathbf{V}\frac{dC}{dt}\right) = \mathbf{E}\tag{14}$$
of which the integral form or solution is
$$M = \mathbf{M}\_0 + \mathbf{E} \cdot t \tag{15}$$
How this solution is obtained is not further explained here. Readers may want to refresh their knowledge of this mathematical calculation method by reviewing a standard text on differential calculus, e.g. Wikipedia [http://en.wikipedia.org/wiki/Differential\_equation]. The result of a constant inflow of a substance is that its mass in the compartment continuously increases. Note that this occurs at the constant rate of E kg/s (Fig. **5**).
Loss rates generally depend on the mass of a substance in the compartment (see Eq **2** and accompanying text). It should be noted that first-order reaction kinetics (see Eq **1**) are the exception, rather than the rule. Zero-order kinetics, in which the reaction is independent of *C* (formally proportional to *C*<sup>0</sup> ), second-order kinetics (reaction rate proportional to *C*<sup>2</sup> ) and broken order kinetics (proportional to *C*1.5) commonly occur. Second-order kinetics will generally apply when a substance reacts with a chemical agent: the reaction is first-order in relation to both the substance degraded and the reactant. It is only because the concentration of the reactant is often approximately constant that the reaction appears proportional only to *C*<sup>1</sup> . This is called pseudo first-order reaction kinetics. For instance, the loss due to reaction with chemical or microbial agents (degradation) is often characterized by pseudo first-order kinetics.
If degradation is the only process, the mass balance equation becomes:
$$
\left(\frac{dM}{dt}\right) = \mathbf{V}\frac{dC}{dt}\tag{16}
$$
the solution of which results in an exponential decrease of mass in the compartment (Fig. **5**):
$$M = \mathbf{M}\_0 \cdot e^{-\mathbf{k} \cdot t} \tag{17}$$
If both emission and degradation act on a compartment, the combined result will be:
$$\frac{dM}{dt}\left(=\mathbf{V}\frac{dC}{dt}\right) = \mathbf{E} - \mathbf{k} \cdot M \; ; \quad M = \mathbf{M}\_0 \text{ at } t = 0$$
the solution of which is:
$$\mathbf{M} = \mathbf{M}\_0 \cdot \mathbf{e}^{-\mathbf{k} \cdot \mathbf{t}} + \frac{\mathbf{E}}{\mathbf{k}} (\mathbf{l} - \mathbf{e}^{-\mathbf{k} \cdot \mathbf{t}}) \tag{19}$$
(see Fig. **5**). Eq **18** and **19** illustrate how the mathematical solution of the mass balance equation yields a mass-time profile of a substance in a compartment as a function of the initial conditions (here: mass M0 at *t* = 0), forcings (here: emission rate, E) and the parameters of the mass flow rate equations (here: the degradation rate constant, k). Note that eventually (at *t* = ∞), the mass of substance in the compartment will reach a level at which the loss by degradation, k.M (kg/s), exactly matches the constant emission, E (kg/s), so that the mass of substance in the compartment is maintained at the steady-state level of E/k (kg).
There are many other loss mechanisms that need to be accounted for in the mass balance equation, such as such as advective or diffusive outflow. Because losses due to all mechanisms i are proportional to *M*, and can each be represented by a first-order rate constant ki (1/s), the full mass balance equation keeps the same simple format of Eq **18**:
**Figure 5:** Elementary form of a one-compartment mass balance model, showing the differential mass balance equation and its solution for the cases of emission only (red), degradation only (blue) and both (green).
$$\frac{dM}{dt}\left(=\mathbf{V}\frac{dC}{dt}\right) = \text{gains}-\text{losses} = \mathbf{E} - \sum\_{i} \mathbf{k}\_{i} \cdot M \; ; \quad M = \mathbf{M}\_{0} \text{ at } t = 0 \tag{20}$$
and its solution takes the same format as Eq **19**:
$$M = \mathbf{M}\_0 \cdot \mathbf{e} \stackrel{-\sum \mathbf{k}\_{i'}}{\cdot} + \frac{\mathbf{E}}{\sum \mathbf{k}\_i} (\mathbf{l} - \mathbf{e} \stackrel{-\sum \mathbf{k}\_{i'}}{\cdot}) \tag{21}$$
# **More Compartments**
Models usually comprise many compartments and describe the transport of a substance in and between these compartments. Such multicompartment mass balance models contain one mass balance equation for each compartment in the model. As in the above situation for one compartment, losses are all assumed to obey first-order kinetics. Where more than one compartment is involved, losses may be due to degradation or export, but losses may also represent mass flows from one compartment to another.
For a set of n compartments, this leads to a set of n mass balance equations, all of which will have the same format as Eq **20**, with n unknown masses *Mi* and a suite of first-order rate constants which describes the losses from the compartments. An example for three compartments is shown in Fig. (**6**). Each of the compartments receives an emission – for the sake of simplicity, emissions will be assumed to be constant and imports considered to be included in the emission flows. The emission flows into the compartments i are denoted by Ei (kg/s). Degradation occurs in the three compartments – again, in the interests of readability, the degradation flows will be considered to include possible exports. The resulting mass flows from the compartments i, out of the system are characterized by pseudo first-order loss rate constants ki and denoted by ki.*Mi* (kg/s).
**Figure 6**: Diagram of a three-compartment mass balance model. Intercompartment mass-transfer represents a loss to the source compartment and a gain to the receiving compartment.
There are six intercompartment mass-transfer flows, each proportional to the mass in the source compartments denoted by ki,j.*Mi* (kg/s). On this basis, and assuming all initial masses to be zero, the three differential mass balance equations become:
$$\begin{aligned} \frac{dM\_1}{dt} &= \mathbf{E}\_1 - (\mathbf{k}\_1 + \mathbf{k}\_{1,2} + \mathbf{k}\_{1,3}) \cdot M\_1 + \mathbf{k}\_{2,1} \cdot M\_2 + \mathbf{k}\_{3,1} \cdot M\_3; \quad M\_1 = 0 \text{ at } t = 0 \\\frac{dM\_2}{dt} &= \mathbf{E}\_2 + \mathbf{k}\_{1,2} \cdot M\_1 - (\mathbf{k}\_2 + \mathbf{k}\_{2,1} + \mathbf{k}\_{2,3}) \cdot M\_2 + \mathbf{k}\_{3,2} \cdot M\_3; \quad M\_2 = 0 \text{ at } t = 0 \\\frac{dM\_3}{dt} &= \mathbf{E}\_3 + \mathbf{k}\_{1,3} \cdot M\_1 + \mathbf{k}\_{2,3} \cdot M\_2 - (\mathbf{k}\_3 + \mathbf{k}\_{3,1} + \mathbf{k}\_{3,2}) \cdot M\_3; \quad M\_3 = 0 \text{ at } t = 0 \end{aligned} \tag{22}$$
For this system of three compartments there is an equation equivalent to Eq **19**, *i.e.* the analytical solution of the one-compartment system, which expresses the mass of the substance at all times. It is not possible to formulate precisely how the three masses in the three compartments change with time. Solutions can be approximated quite well, however, with computer-based numerical techniques which will not be described here. As in the onecompartment system, the three-compartment system will eventually (at *t* = ∞) reach to a steady state in which emission is equally balanced by degradation (d*Mi* /d*t* = 0) and masses reach their constant steady state level, *M\* i*:
$$\text{balance}\_1 = \mathbf{E}\_1 - (\mathbf{k}\_1 + \mathbf{k}\_{1,2} + \mathbf{k}\_{1,3}) \cdot M\_1^\* + \mathbf{k}\_{2,1} \cdot M\_2^\* + \mathbf{k}\_{3,1} \cdot M\_3^\* = 0$$
$$\text{balance}\_2 = \mathbf{E}\_2 + \mathbf{k}\_{1,2} \cdot M\_1^\* - (\mathbf{k}\_2 + \mathbf{k}\_{2,1} + \mathbf{k}\_{2,3}) \cdot M\_2^\* + \mathbf{k}\_{3,2} \cdot M\_3^\* = 0$$
$$\text{balance}\_3 = \mathbf{E}\_3 + \mathbf{k}\_{1,3} \cdot M\_1 + \mathbf{k}\_{2,3} \cdot M\_2^\* - (\mathbf{k}\_3 + \mathbf{k}\_{3,1} + \mathbf{k}\_{3,2}) \cdot M\_3^\* = 0 \tag{23}$$
The set of steady-state masses for which the mass balance equations become zero can be derived directly from Eq **23** quite easily through simple algebraic manipulation. Solving sets of equations algebraically becomes increasingly tedious for larger sets, so linear algebra (matrix calculus) is used to obtain solutions to large sets of linear equations, as follows:
$$\mathbf{m} = \begin{bmatrix} M\_1 \\ M\_2 \\ M\_3 \end{bmatrix}, \quad \mathbf{e} = \begin{bmatrix} \mathbf{E}\_1 \\ \mathbf{E}\_2 \\ \mathbf{E}\_3 \end{bmatrix}, \quad \mathbf{A} = \begin{bmatrix} -(\mathbf{k}\_1 + \mathbf{k}\_{1,2} + \mathbf{k}\_{1,3}) & \mathbf{k}\_{2,1} & \mathbf{k}\_{3,1} \\ & \mathbf{k}\_{1,2} & -(\mathbf{k}\_2 + \mathbf{k}\_{2,1} + \mathbf{k}\_{2,3}) & \mathbf{k}\_{3,2} \\ & \mathbf{k}\_{1,3} & \mathbf{k}\_{2,3} & -(\mathbf{k}\_3 + \mathbf{k}\_{3,1} + \mathbf{k}\_{3,2}) \end{bmatrix} \tag{24}$$
Using this, the three mass balance equations of Eq **23** can be rewritten into a one-line linear-algebraic equation:
$$\mathbf{m} = -\mathbf{A}^{-1} \cdot \mathbf{e} \,. \tag{25}$$
Various standard software packages, such as Microsoft Excel, can be used to carry out matrix inversion.
# **Multimedia Modelling**
If a chemical is released into one medium and resides there until it is removed by degradation or advection, singlemedia models may be perfectly suitable for estimating the environmental concentration. If, however, a chemical is released into several compartments simultaneously, or after release into one compartment is transported to other compartments, it becomes necessary to account for the intermedia transport processes so that its ultimate fate in the overall environment can be assessed. Multimedia models are specifically designed to do this. This section on multimedia models starts with a short description of their features and the explicit and implicit assumptions usually made. The use of these models in exposure assessment is described together with their limitations. Subsequently, some information on data requirements and on the different models available is given, following which a number of sample calculations are presented to illustrate the use of these models.
**Figure 7**: Diagram of a multimedia mass balance model concept. 1 = Emission, 2 = Import, 3 = Export, 4 = Degradation, 5 = Leaching, 6 = Burial, 7 = Wet deposition, 8 = Dry aerosol deposition, 9 = Run-off, 10, 11 = Gas absorption and volatilization, 12 = Sedimentation and resuspension, 13 = Sorption and desorption.
Multimedia fate models are typical examples of compartment mass balance models. The total environment is represented as a set of spatially homogeneous (zero-dimensional) compartments; one compartment for each environmental medium in which the chemical is assumed to be evenly distributed (Fig. **7**). Typical compartments considered in models are: air, water, suspended solids, sediment, soil and aquatic biota. Multimedia mass balance modelling was initiated in the early 1980s by Mackay and co-workers [2, 33-36]. The example was soon followed by others [11, 37-39]. In Europe, the model SimpleBox used in The Netherlands was adopted as the basis for the risk assessment model EUSES [12, 13]. While the early models described a fixed, "unit world", which was meant to represent a global scale, later models have enabled users to customize the environment and define smaller and more open spatial scales. More recently, the use of spatially resolved multimedia fate models has become more common [40-50].
A typical regional multimedia model describes a region between 10<sup>4</sup> and 10<sup>5</sup> km2 . In this generic form, the models can account for emissions into one or more compartments, exchange by import and export with compartments "outside" the system (air and water), degradation in all compartments and intermedia transport by various mechanisms (Fig. **7**). Mass flow kinetics, formulated slightly differently in models by different authors, are usually defined as simply as possible: mass flows are either constant (emission, import) or controlled by (pseudo) first-order rate constants (degradation, intermedia transport), as in Eq **15**. In all the models, the user has to set parameter values for these mass flows to provide input for the model.
Using a number of criteria, such as equilibrium or non-equilibrium, steady-state or non-steady-state, and based on whether to take the degradation of the chemical into account in the calculation or not, Mackay and Paterson introduced a classification of multimedia models [34]. This classification begins with a Level I model which describes the equilibrium partitioning of a given amount of a chemical between the above media. The Level II model simulates a situation where a chemical is continuously discharged into a multimedia environment in which partitioning, advection and degradation take place. Transport between the media is assumed as infinitely rapid, so that thermodynamic equilibrium between the media is maintained. At Level III, realistic intermedia transport kinetics are assumed, so that media may not be in thermodynamic equilibrium. Level III models calculate steadystate concentrations in all compartments. Finally, Level IV models assume a non-steady-state and yield time-related chemical concentrations.
Level I calculation requires knowledge of intermedia partition coefficients (air-water, water-solids) only. Calculation at level II and above requires additional knowledge of degradation rate constants in air, water, sediment and soil. Unfortunately, measured partition coefficients and rate constants are not always available. In the absence of measured data, partition coefficients can be estimated from basic substance properties, using quantitative structureactivity relationships (Q)SAR. Easy to use software is available to support such estimates. The consequence of using estimated model input data is that the accuracy of the model output will also depend on the quality of the (Q)SAR methods that have been used. Very often biodegradation rate constants are extrapolated from standard degradation tests, or even estimated using (Q)SARs (e.g., BIOWIN). This may introduce another uncertainty into the outcome of the calculation, especially if precise data is not available for the degradation rate constants in compartments that serve as a "sink" for a specific chemical.
The principal utility of multimedia models, as a first step in exposure assessment, is to determine to what extent intermedia partitioning may occur. If it appears that no significant partitioning into secondary compartments is expected, further exposure assessments may focus on the primary compartment(s) only. As intermedia transfer is usually relatively slow, its effect on the fate of chemicals is significant only over longer periods of time, *i.e.* if the spatial scale is large or the chemical does not degrade rapidly. This brings us to one of the major applications of these models, which is the exposure assessment of chemicals on regional (usually 104 to 10<sup>5</sup> km2 ) and larger spatial scales. These models are particularly useful for calculating the predicted environmental concentration (PEC) especially of chemicals with a very diffuse release pattern. Results from Level III multimedia models are used in EU risk assessments for new and existing chemicals. In addition to calculating the regional concentration of a chemical, the results of Level III models can also be used as input for local models. When using such models, the actual concentration is greatly underestimated if the concentration of the chemical in air or water from "outside" is set to zero, especially in relation to high production volume chemicals with a widely distributed use pattern. Regional concentrations estimated from the release rates for a larger region fed into a regional multimedia model can then be used as boundary concentrations in local model calculations.
One of the key processes in multimedia models is the partitioning between aqueous and solid phases. Most models follow in the footsteps of the original Mackay models and estimate solids-water partitioning from the octanol-water partition coefficient *K*OW. This means that the models are particularly useful for organic chemicals whose *K*OW values can be accurately measured or estimated. Applying these models to ionisable compounds, surface-active chemicals, polymers, or inorganic compounds (including metals) should be done with great care. However, the models can be used for these chemicals, provided certain adaptations to specific physicochemical properties are made. Mackay and Diamond, for instance, used an "equivalent" based model to describe the fate of lead in the environment [50], while in the example calculation for cadmium parameters such as soil-water and sediment-water partition coefficients or the fraction of the chemical associated with aerosols, must be specifically entered by the user in order to overrule the standard estimation routines.
Naturally, representing the environment in the form of a unit world or unit region with homogeneous boxes is a major simplification of reality. However, this extreme degree of simplification in this model concept is both a weakness and a strength at the same time. By disregarding spatial variation, the modelling effort can focus on intermedia distribution and understanding the ultimate fate of a chemical. The concentrations calculated with multimedia models should therefore be interpreted as "spatially-weighted averages" of the concentrations that would be expected in real situations. However, the assumption of homogeneity brings with it a considerable risk that potentially more localized effects may be overlooked. The disadvantage of zero-dimensionality becomes evident with larger areas since, other than for air, it is difficult to identify any large-scale situations where the homogeneity of compartments would seem to be a realistic assumption. To overcome this problem the SimpleBox has introduced the concept of "nesting" [12]. In a nested model the input and output flows of a regional or smaller scale model are connected to a continental scale model which in turn, is connected to a global scale model. In this way, the specific environmental characteristics of the region can be taken into account when the overall fate of the chemical is assessed. While spatial scale nesting was originally introduced as a tool for assessing the overall persistence of a chemical in the environment, the concept soon found wider application in regional exposure assessment in EUSES [13].
Testing the validity of multimedia models is difficult and, until recently, had not been seriously addressed [51]. If a common evaluation environment with agreed fixed environmental characteristics is used, validation of the outcome becomes almost paradoxical since this generic environment does not actually exist in reality. However, the regional generic characteristics can be modified at a later stage and region-specific information on environmental parameters, as well as information on specific discharge rates can be introduced in order to "validate" a specific model setting [35, 52].
# **Multimedia Models in Use**
Multimedia fate models of the Mackay type have been produced by different authors, most of them for their own scientific use. Many of these have been documented and made available for end users, e.g., HAZCHEM [53], SimpleBox [12, 13] CemoS [38], CalTOX [39], ChemCAN [40], EQC [41], ChemRange [42], ELPOS [43], Globo-POP [44], CliMoChem [45], BETR North America [46], BETR World [47], IMPACT 2002 [48] and MSCE-POP [49]. The similarities between these models are more striking than the differences. When fed the same input, the models were shown to yield the same results [51]. The main differences lie in the number of compartments or subcompartments included and how they are handled in terms of computer calculation.
**Table 6:** Parameters used for steady-state calculations with SimpleBox
The mixing depth represents the thickness of the soil, water or sediment box. b
a
Residence time for air or water represents the time needed for air or water to flush through the air or water compartments, respectively.
# **ENVIRONMENTAL FATE OF CHEMICAL SUBSTANCES**
Examples of how to perform Level I, II and III calculations for a range of different chemicals have been presented by Mackay and others [2, 35, 36, 40, 41, 54]. To illustrate the utility of Level III and IV type multimedia modelling, let us consider the use of three chemicals, 1,1,1-trichloroethane, dieldrin and cadmium, in a system resembling The Netherlands, as simulated with SimpleBox [12, 13].
The system parameters are summarized in Table **6**. Let us assume that the background concentrations of these chemicals in air and water outside The Netherlands are equal to the quality standards or objectives set for environmental protection. After 10 years, with these background concentrations, domestic emissions of 1000 tonnes/year for each chemical start to occur: dieldrin to water, cadmium to air, and 1,1,1-trichloroethane to air, water and soil simultaneously (ratio 1:1:1). This situation continues for 40 years and then suddenly stops. What concentrations may be expected in the different environmental compartments, how are the chemicals distributed, and how long does it take to return to the original situation after the emissions stop? In order to evaluate the change in concentrations of the three chemicals in the different environmental compartments some chemical-specific information is needed. This is summarized in Table 7.
**Table 7:** Input parameters used in the multi-media model calculations for 1,1,1-trichloroethane, dieldrin and cadmium
Substitute for zero-value.
The Level III mode of the SimpleBox program is then used to generate the concentrations and intermedia distribution at steady-state. The concentrations in and distribution over the environmental compartments at steady-state are summarized in Table 8. The mass flows that support these steady-states are also shown in Fig. (**8**). The model calculation emphasizes the high volatility of 1,1,1-trichloroethane. Approximately all emissions to soil and water go to air through diffusive transport. Of the total mass in the system, however, a high percentage still resides in the soil.
**Table 8:** Steady-state distribution of 1,1,1-trichloroethane, dieldrin and cadmium in The Netherlands, calculated with SimpleBox [12, 13]. Numbers in parentheses represent a percentage of the total mass in the environment at steady-state.
**Figure 8:** Steady-state mass flows of trichloroethane (A), dieldrin (B) and cadmium (C), as a percentage of the total throughput of the system.
Remarkably, the relatively high volatility of dieldrin causes more than half of the total load of the water compartment to be transported to air, from where it is exported out of the system. The high hydrophobicity and low biodegradation rates of the chemical produce relatively high concentrations in sediment and soil. Cadmium does not degrade at all. When emissions go to air the most important fate process is advection out of the air compartment. However, due to atmospheric deposition, some 10% of the total load of the atmosphere is transported to soil and water. Atmospheric deposition to soil leads to a build-up of cadmium in the soil, from where it is eventually leached to the ultimate sink: the deeper groundwater. It should be borne in mind that this build-up may be slow. If, as in the case of cadmium in soil, all mass flows are small, it may take an extremely long time before the steady-state is achieved. This can be demonstrated with Level IV calculations using the SimpleBox model. Fig. (**9**) shows the change in concentrations in the different compartments according to the above emission scenario relative to the background concentrations which result when there are no domestic emissions. For cadmium, the compartments air, water and sediment are expected to respond relatively quickly, whereas a near linear increase in the concentration in soil is predicted over the 40-year exposure period.
After reducing the emissions, the soil concentration of cadmium shows little response (Fig. **8C**). For dieldrin exposure, for 40 years is almost long enough to reach a steady-state, even in the "slow" soil compartment; after reducing the emission to 10% of its original value, the concentrations decrease at the same rate (Fig. **8B**). For trichloroethane the situation is completely different. The steady-state situation is reached so quickly that plotting the concentrations against time on a 100-year scale would yield a block diagram. Therefore, the Level IV calculation was repeated over a time-scale of one year. The results as presented in Fig. (**8A**) show that concentrations in air, water and soil reach steady-state within one month. For sediment this takes a little longer, though probably not much longer than a year. These results demonstrate the usefulness of Level III and Level IV multimedia box model calculations. Where steady-state calculations can give information on the concentrations and distribution in the environment at a constant emission scenario, the results of a Level IV calculation elucidate the time scale in which this situation may be reached. In addition, changes in the emission scenario as a result of evolving risk reduction strategies can be evaluated in this way.
**Figure 9:** Change in concentrations of trichloroethane (A), dieldrin (B) and cadmium (C) after a change in emission rates. Note the shorter time scale in graph A.
# **Calculation of Overall Persistence in the Environment and Long-Range Transport Potential**
It is clear that the physical and chemical properties of substances greatly influence their concentrations and distributions in the environment. Not only does this have implications for the risks posed to humans and ecosystems, there are other ethical and scientific consequences to be considered [55]. Slow degradation and great mobility mean that substances disperse throughout the entire globe. This has been recognized internationally. Two international conventions: the UNEP Stockholm Convention [56] and the UN ECE POP protocol [57] now regulate substances on the basis of their persistence in the environment and their long-range transport potential. Both of these are indirect or "derived" substance properties.
Persistence reflects the resistance of a substance to degradation. This is indicated by the dynamic response to changes in emissions, as shown above. Alternatively, persistence can be quantified by the degradation half-life or reactive residence time during an emission episode [58, 59]. As degradation half-lives in air, water and soil differ greatly, it needs to be decided how to combine the different single-medium half-lives. Calculation of overall persistence in the environment Pov as the reciprocal of the overall degradation rate constant kov, or the mass-weighted average reactive residence time in the environmental media *M*i, has been proposed for this purpose [58-60]:
$$\mathbf{P}\_{\rm OV} = \frac{1}{\mathbf{k}\_{\rm OV}} = \frac{\sum\_{i} M\_{i}}{\sum\_{i} M\_{i} \cdot \mathbf{k}\_{i}} \tag{26}$$
In this derivation of POV, ki's are the first-order degradation rate constants in pure media and *Mi*'s are the masses in the media at steady-state. According to this derivation, substance properties other than degradation half-lives (partition coefficients and mass-transfer velocities) play a role in determining the "derived property" POV. Applied to the calculation results of the previous paragraph, this would yield POV values of 2.8 years, 20.8 years and ∞ for trichloroethane, dieldrin and cadmium, respectively.
The long-range transport potential (LRTP) reflects the tendency of a substance to be transported away from the location where it was emitted. There are different ways to capture this in a "derived property" [59, 60]. One is to take the fraction of the total emission exported out of an open regional environment, as shown in the previous paragraph:
$$\text{LRTP} = \frac{adv\_{air} + adv\_{water}}{\text{E}},\tag{27}$$
with *advair* and *advwater* denoting the advective mass flows by air and water, respectively and E the sum of emissions. The LRTP values (dimensionless) for trichloroethane, dieldrin and cadmium would be 0.99, 0.92 and 0.91, respectively, based on example model used. Another method is to use the Lagrangian characteristic travel distance. The distance travelled by a parcel in the period that the original mass is reduced exponentially to 37% (=1/e) of its original value is calculated as [59, 60]
$$\text{LRTP} = \frac{\text{u}}{\text{k}\_{\text{ov}}^{\*}},$$
in which u is the average velocity at which the parcel travels. Here, kOV\* considers non-reactive losses to ultimate sinks such as sediment burial, groundwater or deeper ocean layers as well as abiotic and biotic degradation processes.
What POV and LRTP have in common is that they cannot easily be determined by observation, but must be calculated from substance properties that can be measured (degradation rate constants, partition coefficients, masstransfer velocities), using a multimedia environmental fate model. This has raised the concern that the choice of model could play a role in the calculation result, which would be undesirable if POV and LRTP are to be used as a property of the substance in a regulatory context. This issue has been thoroughly studied by an international group of modelling experts for the OECD [60]. The experts concluded that indeed the absolute values of POV and LRTP obtained from different models differ greatly, as a result of different modelling objectives and model parameterization. However, the rankings of substances obtained appeared to be relatively insensitive to the model choice: models tend to put chemicals in roughly the same order of POV and LRTP. If properly processed, output of any well-designed multimedia model can be used to derive POV and LRTP [61, 62]. This was concluded from a comparison of the performance of existing models with respect to POV and LRTP calculation, which demonstrated that a simplified version of existing models could be constructed that differed as little from the existing models as the models differed among themselves. This consensus model is available from the OECD on their website [63].
# **REFERENCES**
© 2011 The Author(s). Published by Bentham Science Publisher. This is an open access chapter published under CC BY 4.0 https://creativecommons.org/licenses/by/4.0/legalcode
# **CHAPTER 3**
# **Metals and Metalloids in Terrestrial Systems: Bioaccumulation, Biomagnification and Subsequent Adverse Effects**
**Reinier M. Mann1,\*, Martina G. Vijver2 and Willie J.G.M. Peijnenburg2,3**
*1 Centre for Ecotoxicology, University of Technology Sydney, Australia; 2 Leiden University, Leiden, The Netherlands and <sup>3</sup> National Institute of Public Health and the Environment, Bilthoven, The Netherlands*
**Abstract:** Metals and metalloids are elemental substances that occur naturally in the Earth's crust, and are variously incorporated into biological systems as structural components or proteins. Imbalances in the environmental concentrations of several metals present a challenge to ecosystems because the species that form part of these ecosystems are often not equipped to regulate internal concentrations of these elements, or employ detoxification mechanisms that serve to biomagnify these elements in the food chain. This review examines the trophic movement of metals and metalloids within terrestrial ecosystems and the consequences of biomagnification and toxicity on populations. Several elemental contaminants are given special emphasis, including copper, zinc, arsenic, selenium, molybdenum, cadmium, mercury and lead. All these elements are of high historical importance and continue to be deposited within the biosphere.
# **INTRODUCTION**
Elemental chemicals have a tendency to stick to clayish, peaty or organic-rich materials, and contamination of terrestrial systems generally occurs because soils are capable of acting as a sink for metals and other elemental chemicals [1]. The elemental contaminants of most concern are predominantly metals such as cadmium, lead, mercury or copper, among others. However, a few, like selenium and arsenic are metalloid elements. For the sake of simplicity in this discussion we have used only the term 'metal' to describe both types, even though examples may include metalloid elements like selenium.
Metals in soils originate from two separate sources: from geogenic processes related to the occurrence of metalbearing geological formations, and from anthropogenic sources. Information on the sources, fate, transport and toxicity of metals and metalloids can be found in Chapter 1 and 2 of this book. Once bound within soils, metals are persistent, because elemental contaminants cannot degrade further (unlike complex organic pollutants), although they can undergo various reversible changes in speciation depending on the chemical environment [2].
Many trace elements are essential for life functions [3], and plants and animals possess various mechanisms for the accumulation of sufficient amounts of trace elements from their environment. These same mechanisms can also facilitate the uptake of non-essential metals [4, 5]. The uptake and retention of a metal (or any other chemical) by an organism is termed bioaccumulation. Bioaccumulation of essential as well as non-essential elements is dependent on both the chemical availability of the metals within the environment and the organism's capacity for uptake and subsequent excretion. A full overview on bioaccumulation is given in Hodson *et al.* [6], in which the different definitions of bioavailability currently in use are reviewed.
The severity of impact on ecosystems will reflect the concentrations of bioavailable metals in the soil. The concentrations of individual metals are dictated by the source of each contaminant. For example, mercury (Hg) contamination occurs predominantly as a consequence of atmospheric deposition, and is therefore rather diffuse. By contrast, a major source of cadmium (Cd) contamination has historically been through the application of rockphosphate fertilizers, thereby selectively elevating Cd concentrations in agricultural soils. Very high levels of soil contamination usually only occur in the proximity to metal smeltering activities, and an examination of studies conducted in these environments are instructive about the relative movements of metals within local ecosystems.
One such study was conducted by Hunter *et al.* [7-9] in the vicinity of a copper refinery within Merseyside in northwest England. Copper (Cu) and Cd content of the soils within a 1 km radius of the refinery typically exceeded 500
**Francisco Sánchez-Bayo, Paul J. van den Brink and Reinier M. Mann (Eds) © 2011 The Author(s). Published by Bentham Science Publishers**
**<sup>\*</sup>Address correspondence to Reinier M. Mann:** Centre for Ecotoxicology, Department of Environmental Sciences, University of Technology Sydney, NSW 2007, Australia; Present Address: Hydrobiology, Brisbane, Australia; Email: [email protected]
and 5 mg/kg, respectively [7]. In this highly impacted area, floral diversity was reduced to a few metal tolerant species compared to a reference site. However, invertebrate diversity, as represented through pitfall trapping, was not greatly affected, with the exception of a reduction in abundance of isopods (woodlice) and oligochaetes (earthworms) within a 1 km radius of the refinery. The site also supported small mammals; specifically field voles (*Microtus agrestis* L.), wood mice (*Apodemus sylvaticus* L.) and common shrews (*Sorex araneus* L.).
All organisms within this contaminated site accumulated metals to varying degrees, and following the ratio of Cu:Cd through the various trophic levels illustrates the variability in accumulation potential for different metals. The Cu:Cd ratio in the soil close to the refinery was 716:1. Vegetation at the refinery bioaccumulated both Cu and Cd, however the ratio was reduced to 37:1. The ratio of Cu:Cd in the herbivorous field voles was reduced further to 5:1. Among the various herbivorous, detritivorous and predatory invertebrate taxa the Cu:Cd ratio varied markedly, but within the diet of the carnivorous common shrew the ratio averaged 10:1 and was reduced further to 1:3 in the shrew itself [8, 9]. These changes in Cu:Cd ratios illustrate the element specific mobility of Cu and Cd within terrestrial food chains. In this example the change in ratio of Cu:Cd occurs because copper is an essential element that can be regulated by homeostatic mechanisms. In the study cited here [9], shrews within the vicinity of the refinery accumulated large body burdens of Cd, whereas their Cu burdens remained low. Unlike copper, cadmium is a nonessential metal and organisms have only limited capacity to eliminate it from their bodies and tend to pass it on to consumers/predators. It is notable that despite the large body burdens of Cd, shrews persisted in the contaminated environment, and this will be discussed further later in the chapter.
This chapter will examine the metal- and species-specific movements of metals in terrestrial ecosystems, and where examples exist, the consequences of bioaccumulation and biomagnification of metals for populations of terrestrial organisms. Illustrations of metal transfer through the aquatic food chain are included to provide a complete picture, thereby improving our ability to make general statements or to fill gaps of knowledge for terrestrial ecosystems.
# **What is Biomagnification?**
The process whereby pollutants are transferred from food to an organism resulting in higher concentrations compared with the source is called biomagnification. There are two main groups of substances that biomagnify:
These pollutants biomagnify along food chains because successive trophic levels consume relatively large quantities of biomass (food) to obtain the resources required for metabolic functioning. If that biomass is contaminated, the contaminant will be taken up in large quantities by the consumer. Lipophilic contaminants within consumed biomass are subsequently absorbed and stored in the bodies of the consumers rather than eliminated along with other waste products. If the consumer is eaten by another consumer organism, the fat tissue is digested and the contaminant is then stored in the tissues of the latter consumer. In this way, the contaminant builds up in the fatty tissues of the subsequent consumers and the concentration of the contaminant in their tissues becomes higher with each trophic level. Water-soluble pollutants usually do not biomagnify in this way because they would dissolve in the bodily fluids of the organism and be excreted. Thus the principle of biomagnification is based on the fact that the mass of the contaminant is largely conserved along the food chain, while the biomass decreases [10].
The extent to which biomagnification occurs between a consumer and its food/prey can be expressed as the biomagnification factor (BMF). The BMF can be used to predict ecological risk of chemicals [11, 12]. Determining the biomagnification for a food chain experimentally, by studying the transfer within a chain of prey and predators, is rather simple, although it may not be very practical as all possible chemical exposure routes to organisms (water, food and soil/sediment) must also be taken into account [13]. The total bioaccumulation of a metal in species of each trophic level within a specific food chain corresponds to the chemical concentration in the organism relative to the concentrations in its surrounding environment and in its diet, respectively. Hence, bioaccumulation (expressed as the bioaccumulation factor (BAF)) is the sum of two processes: bioconcentration, which is accumulation *via* the exposure medium (expressed as the bioconcentration factor (BCF)) and biomagnification which is uptake *via* food only (expressed as the biomagnification factor (BMF)). Although BAF is the sum of BCF and BMF, it should be noted that summing both factors in a numerical way requires much care because of the differences in units for BCF and BMF [14].
# **Are Metals Biomagnified?**
Hendriks and Heikens [15] modelled metal kinetics in a food chain by means of empirical regressions based on mean values. In this modelling, it was concluded that despite taxonomic variability, metal concentrations diminish with increasing trophic levels. Also for marine ecosystems, Gray [16] concluded that biomagnification of metals is not a universal rule. This is in agreement with earlier conclusions of Laskowski [17], who proposed that, based on the mean concentrations that accumulate in successive trophic levels, biomagnification of Cd and Cu does not lead to high concentrations in carnivorous predators. In contrast to this view, van Straalen and Ernst [18] suggested that trophic movement of metal cannot be examined by generalizing about the body burdens of different trophic levels (*i.e.* using the statistical means of body burdens), but must be examined by following the path of each metal through each species, because different species will have different capacities to accumulate metals. The difference between these two views is illustrated in Fig. (**1)**. Scheme A (Fig. **1A**) provides a simplistic hierarchical view of a trophic cascade in which all consumers can be allocated to discrete trophic levels. Alternatively, scheme B (Fig. **1B**) provides a model with more complicated interactions between different trophic levels, and when examined in this way, metal biomagnification may be manifest in some trophic pathways, but not in others.
**Figure 1:** Schematic diagram of two biomagnification models. **A** provides a simplistic hierarchical scheme which places all consumers in subsequent trophic levels. **B** provides a more realistic "food web" representation where each organism has an associated trophic index (TI). TI is calculated as TIi = 1+ Σ (TIj × DFi,j), where TIi is the Trophic Index of the species i, TIj is the Trophic Index of the species j and DFi,j is the fraction of the species j in the total diet of species i. The TI for primary producers is set at 0. Redrawn and modified from Alonso [21].
Croteau *et al.* [19] concluded that predictive relationships between metal concentrations in predators and prey was only possible if their prey can be identified, and if the concentrations of metals in their prey is known. This information can subsequently be used in a bioaccumulation model. In the field situation, stable isotope ratios of carbon and nitrogen can be used to ascribe trophic positions of species within a food web. These ratios provide insights into time-integrated energy flows and food web structures [20]. The isotope ratios of carbon can be used to identify food sources, whereas the nitrogen isotopes can be used to infer the trophic position of an organism. After defining these relationships, pollutant concentrations can be compared to trophic levels inferred with the isotopic techniques, and a better understanding of habitat-specific food webs can be gained.
There is ongoing discussion as to whether biomagnification of metals occurs within aquatic and terrestrial systems, and more importantly, whether it can occur to the point that there is a detrimental effect on middle or top-order predators. Food web complexity (*i.e.* as represented in Fig. **1B**) makes predicting the trophic movement of metals rather difficult. Up to now, most information about biomagnification in field situations, especially quantitative relationships, is limited to comparative assessments between food chains, because the complexity of the process and the variability of data increase when moving to food-chain assessments.
# **TROPHIC MOVEMENT OF METALS AND FACTORS INVOLVED**
Partitioning of metals is dictated by the environmental compartments present (especially solid and liquid phases) and the size of those compartments. Within the compartments, various environmental factors affect the bioavailability of metals for bioaccumulation [22-24] The environmental factors that influence the fate and partitioning of metals include pH, sorbing ligands in the exposure matrix, and the amounts of competing ions. Also, metal specificities affect bioaccumulation [25], as do species-specific characteristics such as the excretion capacity of organisms [10, 26] and the trophic level of the species.
# **Metal Specificity**
Some metals are known to biomagnify. The best understood example is that of mercury (Hg), which biomagnifies to a great extent, but only when present in the lipophilic organic form, methyl mercury (MeHg). Methyl mercury is formed under anoxic conditions through microbial methylation [27], is readily bioaccumulated by algal species [28] and fungi [29] and subsequently biomagnified through trophic transfer [28, 30]. High body burdens of Hg are usually found in toporder marine predators [31] like the toothed whales. Among terrestrial fauna, particularly in birds, high body burdens in Hg usually occur as a consequence of consumption of aquatic invertebrates [32-34] or fish [34]. Similarly, semiterrestrial mammalian predators in aquatic ecosystems (e.g. mink, otters, polar bears) also bioaccumulate Hg [35, 36], as do wholly terrestrial carnivores for which fish forms a large proportion of the diet [37].
The extent to which Hg biomagnification occurs in these food chains requires knowledge of the dietary preferences of the species involved. For example, polar bears are known to bioaccumulate Hg from their prey. However, the concentrations of Hg reported in polar bear liver, although high (between 1 and 200 g/g dry wt) is rarely higher than in their prey food (seals) [38]. The reason for the apparently low level of trophic transfer of Hg between seals and polar bears likely lies in the polar bear's dietary preference for the skin and fat which have relatively little of the bioavailable MeHg, and the low bioavailability of inorganic Hg in seal liver [39].
Biomagnification of other metals is more difficult to demonstrate, although there are theoretical reasons to suspect that Se and Cd could biomagnify under some circumstances [40]. Gray [16] reviewed 35 papers on the subject of biomagnification of metals in aquatic systems and judged that there was little evidence, with the exception of MeHg, for the biomagnification of metals, despite the fact that 28% of papers (2 of 7) that examined biomagnification of organotin, did demonstrate biomagnification. Recently Se biomagnification has been demonstrated in a short aquatic food chain in the field [41]. In terrestrial systems, Se has not been demonstrated to biomagnify [42] although elevated tissue concentrations of selenium have been found among small resident mammals and birds nesting in the vicinity of a Se contaminated site [43, 44].
In aquatic systems, and contrary to prevailing views, Croteau *et al.* [40] demonstrated that Cd is progressively enriched among trophic levels in discrete epiphyte-based food webs composed of macrophyte-dwelling invertebrates (the first link being epiphytic algae) and fishes (the first link being gobies) [40]. In the same food web, Cu was not similarly enriched. Biomagnification of Cd has also been demonstrated in terrestrial food chains (see below).
Nickel (Ni) and thallium (Tl) also have the potential for transference along aquatic food chains. Dumas and Hare [45] demonstrated that the majority of both metals (58 to 83%) was assimilated by predatory alderflies (*Sialis velata*) feeding on aquatic invertebrates that had previously accumulated Ni and Tl from contaminated sediment, and indicates that these metals are easily transferred along the aquatic food chain and that food is an important source for biomagnification of these elements. Thallium in particular is known to bioaccumulate in plants grown in contaminated soils [46, 47]. However, very little information is available about the biomagnification potential of Tl in higher trophic fauna, although high levels of Tl were reported in greater white-toothed shrew (*Crocidura russula*) 19 months after the collapse of a tailing dam at the Los Frailes Mine in Aznalcóllar that resulted in extensive contamination with various metals in Doñana National Park, Spain [48].
# **Dietary Exposure to Metals**
In aquatic systems, all organisms are subject to the diffusive mechanisms that allow metals to passively enter tissues. Bioconcentration occurs when organisms sequester those metals internally (*i.e.* when assimilation is higher than excretion), and thereby maintain an inward diffusion gradient. Bioconcentration is distinct from bioaccumulation. Bioaccumulation takes into account the internalization and retention of contaminants *via* all routes including ingestion and absorption across membranes such as gills, whereas bioconcentration is particular to aquatic organisms that accumulate contaminants across exposed membranes. All aquatic organisms, including fish, possess membranes that are exposed to the water column (e.g. gills) where diffusion of metals can occur. With this distinction in mind, Gray [16] observed that with passive uptake (e.g. fish) biomagnification does not occur as opposed to dietary uptake (e.g. birds) where biomagnification may be observed. In aquatic systems, up to the trophic level of fish, there is usually no need to assume that food is the major route for contaminant intake and therefore, that biomagnification is not so important. However, Gray [16] also observed that organisms that have aerial respiration (e.g. sea birds, reptiles and marine mammals) must take in contaminants *via* food rather than their body surface and are likely to show biomagnification. Therefore it can be concluded that aquatic systems react differently on the transfer of metals through the food chain than terrestrial food chains.
# *Transfer from Soil to Plant to Grazing Fauna*
Many plants are able to bioaccumulate metals and some plant species are even able to hyperaccumulate various metals at levels exceeding their concentrations in ores [49]. It has been suggested that metal accumulation by plants may be a defence strategy to discourage consumption by herbivores [50, 51], although it may be more accurate to say that avoidance of plants with high metal burdens establishes an evolutionary selection pressure for hyperaccumulation among plants [51]. Similarly, some species of herbivores have evolved to utilise metals bioaccumulated in the ingested plant biomass as a defence against subsequent predation [52]. The implication here is that some animal consumers/predators are able to detect and selectively ingest or avoid metals in prey/food items.
However, in the absence of metal avoidance behaviours, trophic biomagnification of metals might be expected. In a study examining a floodplain area in The Netherlands with elevated metal concentrations [53], the most dominant plant species was the stinging nettle *Urtica dioica*. The stinging nettle contained only very low metal concentrations, far below the maximum values found in plants from non-polluted sites. Nevertheless, the main herbivore feeding on these plants, the snail *Cepaea nemoralis*, did contain metal concentrations that were much higher than background values [53]. Cadmium in particular was accumulated to very high levels, with consequent negative effects on reproduction [53, 54]. Similarly, substantial Cd accumulation was also reported among snails (*Helix aspersa*) in mesocosm studies [55, 56]. However, these studies indicated that up to 40% of the accumulated Cd, and even higher proportions of accumulated lead (Pb) and zinc (Zn) are bioconcentrated directly from the soil [56]. Exclusively dietary accumulation of Cd has been demonstrated in aphids. In an examination of the trophic movement of Cd and Zn between wheat grown on Cd-contaminated soils, and aphids (*Rhopalosiphum padi* and *Sitobion avenae*), aphids were demonstrated to bioaccumulate both Cd and Zn up to ten times the concentrations in wheat [57, 58].
Bioaccumulation of metals through grazing is not only dependent on metal levels in the plants consumed, but effectively depends on a delicate interplay between internal processes regulating concentrations of both essential and non-essential elements below the concentrations at which toxicosis occurs. A well known example is a two-stage process leading to copper toxicity in sheep and cattle. Copper toxicity among domestic ruminants generally occurs as a consequence of consumption of Cu-contaminated water, vegetation contaminated with Cu-based insecticides or fungicides or pasture that has been top-dressed with Cu salts or swine and poultry manure [59]. Initially, in the first stage, there is the steady accumulation of copper in the liver over time. Under normal circumstances, Cu is absorbed from the diet and transported in the bloodstream to the liver for storage. Excess Cu from the diet is stored in the liver and is released into the blood as needed for regular body functions. The circulating Cu level tends to remain constant regardless of the amount of excess Cu accumulating in the liver. When dietary intake of Cu is high, Cu can build up in the liver over a matter of weeks, months, or more than a year depending on a variety of factors without any clinical signs. However, with increased accumulation above the detoxification capacity of the animal, there is a sudden release of copper from the liver into the bloodstream. This rush of Cu into the sheep's bloodstream causes causing massive hemolysis, renal and liver failure and ultimately death in two to five days [59]. Copper toxicity among ruminants is usually recognised following veterinary examination of domestic stock, whereas copper toxicity among wildlife is seldom documented.
Despite the capacity of most animals to obtain and regulate Cu within narrow limits, Cu deficiency has also been observed among wild ruminants and domestic stock as a consequence of the presence in the diet of elements like molybdenum (Mo) and sulphur (S). Copper deficiency occurs because of the formation of insoluble Cu-Mo-S complexes that are excreted by the grazers [60]. Molybdenosis, or molybdenum induced Cu deficiency is likely the cause of death and disease among wild moose (*Alces alces*) in Sweden [61]. In such cases, total Cu levels in blood plasma are an unreliable guide to Cu status. Similarly, Cu levels in plants or in soil in themselves are also not reflective of actual bioaccumulative levels of essential elements for grazers as levels of other elements affecting the effective Cu dose also need to be taken into account.
# *Transfer to Higher Predators*
Cadmium is the one metal that has been demonstrated to be biomagnified along terrestrial food chains. As indicated above, herbivorous invertebrates can biomagnify the Cd ingested from their food plants. High body burdens among herbivorous/detritivorous invertebrates occur because of high dietary-Cd assimilation efficiencies [up to 100%, 62, 63, 64] and low rates of elimination [e.g. 65]. With successive trophic levels in terrestrial food chains, the occurrence of biomagnification becomes less predictable. Some invertebrate predators have developed physiological mechanisms that allow them to avoid accumulating Cd from their prey. Using the example of aphids cited above, Merrington et al. [66] and Green et al. [67] demonstrated that two predators of aphid, lacewings (Mallada signata) and ladybird beetles (Coccinella septempunctata), did not biomagnify Cd contained in their aphid prey. In the case of the beetles, Cd was assimilated by the larvae, but was subsequently sequestered in pupal exuviae. Another example is the spider Dysdera crocata, which preys upon isopods. Isopods are known to accumulate large body burdens of metals, including Cd; however, when maintained exclusively on a diet of isopods with high body burdens of Cd, D. crocata did not assimilate Cd [68]. The absence of net assimilation in this spider occurs because of the breakdown of the digestive cells in the midgut diverticulae where metals temporarily accumulate, with subsequent release of metals into the lumen of the midgut prior to excretion [69]. In contrast, wolf spiders (Pirata piraticus), when provided with Cd contaminated fruit flies, assimilated nearly 70% of Cd from its prey without any elimination [70].
Among terrestrial vertebrates, Cd assimilation efficiencies are relatively low [<10%, 71, 72, 73], indicating that vertebrate digestive physiology presents an efficient barrier against Cd assimilation [74, 75]. However, overall bioaccumulation can still be expected to be high among some taxonomic groups, particularly homeothermic animals with high metabolic demands (and high food intakes) and long-lived animals. The species-specific differences in capacity to biomagnify Cd is best illustrated by an examination of the numerous studies that have reported accumulation of Cd among carnivorous shrews (*Sorex araneus*) and herbivorous voles (*Microtus agrestis* and *Myodes* [syn. *Clethrionomys*] *glareolus*). Shrews are particularly interesting because their high metabolic rates require them to consume >80% of their body weight each day. The field data presented in Fig. (**2)** comprise several studies and locations, mainly diffusively polluted floodplain soils and former mining areas. Studies were selected for inclusion if Cd concentrations in food items, *i.e.* earthworms and plants, were measured. As small mammals predominantly accumulate Cd *via* ingestion, absorption from water and inhalation could be disregarded as negligible. Binding of Cd to the metal-binding protein metallothionein, with subsequent storage in the liver and kidney is the
main detoxification mechanism of small mammals and results in very low elimination rates. Any decrease in the concentration of Cd in the body could be attributed to elimination *via* growth dilution only.
**Figure 2:** Cadmium accumulation in kidneys of carnivorous shrews (A) and herbivorous voles (B) (mg/kg dry weight) compared to total metal concentrations (mg/kg dry weight) in soils of various origins. **=** Biesbosch [76], = Rhine [77], = ADW [78, 79], = near closed smelter (Budel) and industrially polluted area (Arnhem) [80], = near Cd / Cu refinery, 1 km from refinery and reference location [9, 81], = near mine and reference location in UK [82, 83], lead mine (Frongoch) and reference site [84]. Full line represents empirical regression. Upper and lower dashed lines represent the 97.5th and 2.5th percentile of the field data, respectively. The error bars represent the 95% confidence intervals and were plotted where possible. Figure adapted from Veltman *et al.* [85].
Linear regression analysis was performed to relate Cd concentrations in kidney, liver and whole body of small mammals to total soil concentrations. Results show a significant relationship between total soil concentrations and Cd concentrations in kidney and liver of carnivorous shrew and herbivorous voles (Fig. (**2**), only kidney data shown). Cadmium concentrations in above ground parts of plants were generally lower than concentrations in earthworms when exposed to similar soil-Cd concentrations. This was in agreement with the observation that Cd concentrations in voles were generally lower than in the shrew. Additionally, a large variation in Cd levels in plants was observed when related to total soil concentrations. As a consequence, regressions of Cd accumulation in herbivorous voles have a lower explained variance compared to carnivorous shrews based on total soil concentrations.
# *Transfer from Soil to Higher Fauna*
Direct uptake of metals from soil by wildlife may be an important pathway for metal accumulation as well. Wildlife may ingest soil deliberately, or incidentally when they ingest soil-laden forage or organisms that contain soil in their intestine (e.g. earthworms) [e.g. 86]. Because the concentrations of metals in ingested soil may be higher compared to the metal concentrations in prey items, the soil can be an important pathway of exposure to predators as well.
Beyer *et al.* [87] provided an overview on the estimates of soil ingestion by wildlife based on experimental data, field captured animals and modelling. The authors found that between 3% and up to 30% of the ingested diet of wildlife consists of soils and/or sediments. Examples of carnivorous predators are raccoons (*Procyon lotor*) with 9% direct soil ingestion and red fox with 3%, whereas sandpipers consumed sediments at a rate of 7 to 30% of their diets. Highest rates among the herbivorous eaters studied were Canada geese (*Branta canadensis*) and the blacktailed prairie dog (*Cynomys ludovicianus*) (up to 8%).
# **Excretion of Metals by Predators**
The capacity for excretion differs between predatory species. A high excretion capacity implies that biomagnification is lower. Reinfelder *et al.* [88] suggested that trophic transfer potential could be described from biodynamic parameters - weight-specific ingestion rate, assimilation efficiency (AE), and a rate constant of loss. Hendriks and Heikens [15] based their studies on these parameters as well. Higher concentrations at higher trophic levels can be explained by the fact that elimination rates decrease with increased body size [12, 89, 90]. The main reason for an absence of biomagnification among aquatic trophic chains is the observation that most marine organisms can fairly easy eliminate metals [16].
# **Predator Specificity (Gender & Life-Span)**
Some authors indicate that females accumulate higher concentrations of metals than males of the same species. For example, some studies suggest that male and female moles (*Talpa europaea*) accumulate different heavy metal concentrations in their tissues. Kormanicki [91] observed that females have higher levels of Cd in the femur and stomach and higher levels of Zn in the gonads, spleen and skin. Pankakoski [92], in an examination of the same species in Finland indicated that Pb accumulation was also higher in females than males. The consequences of female biased accumulation for local population viability are not studied and remain unknown.
Gender variation of metal burdens was also observed by Deng *et al.* [93] in a study on metal levels in great tit (*Parus major*) and greenfinch (*Carduelis sinica*) in the western mountains of Beijing (China), but the differences did not follow a specific pattern. In general, trace metal concentrations in different body parts were similar between males and females in both species. However, in the liver and feathers, there were significant gender related differences, and the pattern was species specific. In the great tit, males possessed higher chromium (Cr) and Ni in the liver and Se in the feather than females, while females had higher levels of Cd in these body parts. In the greenfinch, males had higher concentrations of arsenic (As) and Zn than females in the feather, while females had higher concentrations of Zn and Se in the liver. Janssens *et al.* [94] found no general age- or gender-related differences in metal levels across a pollution gradient except for As and iron (Fe), where a significant interaction between site and gender was observed. Actually, the results of these authors suggest that feathers of great tits might be useful biomonitoring tools because they reflect the environmental contamination by heavy metals well.
Nam *et al.* [95] on the other hand observed no gender-related variation in two populations (Lake Biwa and Mie) of great cormorants (*Phalacrocorax carbo*) from Japan for most of the trace elements that they studied, except for higher hepatic strontium (Sr) concentrations in males from Lake Biwa. In both populations, some elements revealed tissue-specific accumulation. For example, most of the burden of Mo, silver (Ag) and Cd was in liver, Tl and Cd in kidney, Cu, rubidium (Rb) and caesium (Cs) in muscle, and vanadium (V), Sr and barium (Ba) in bone. Hepatic V, muscular Hg and Tl, and Cd in liver, kidney and muscle increased with growth. There was little gonad-specific accumulation of any metal. Thus this study countered the hypothesis that enhanced excretion of metal in the eggs laid by females reduces female metal burdens. Site specific variation in elemental concentrations in stomach contents also indicated that dietary sources tended to be the main factor in regional variations observed between the two colonies studied. Concentrations of toxic Hg and Cd in the liver of cormorants from the two colonies were lower than those from other areas, implying relatively low exposure to these metals. Concentrations of V, Co, Ag, Cd, Cs, Hg, Tl, Pb and Bi in liver remained more or less at the same level between 1993 and 2003, while hepatic Cr, Mn, Cu, Zn, Se, Rb, Sr and Ba showed apparent decrease.
The life span of organisms is also important when considering the bioaccumulation of metals because it affects the total period of time that an animal is exposed to contaminants. The older an organism is, the longer it has been exposed to any contaminant. Bioaccumulation will have a linear relationship with the exposure time of an organism, particularly when excretion is negligible. Overall, the life-span of an animal is likely to be a more important factor affecting metal burdens in specific predators than gender variation [96-98, but see 99].
# **Metal Sequestration in Prey**
The transfer of metal from prey to predator is dependent on two highly variable factors. First, the capacity of predators to minimize net assimilation of metals contained within their prey is variable. As indicated above, the digestive physiology of vertebrates and some invertebrate predators provides an effective barrier against metal assimilation and accumulation. Second, the bioavailability of metal sequestered in the prey is also variable. Some plants (e.g. members of the family Brassicaceae) and animals (e.g. crustacean, oligochaete species) are able to accumulate large tissue burdens of metal. They are able to avoid the damaging effects of reactive forms of essential and nonessential metals and to selectively control utilization of essential metals, by sequestering them in non-toxic forms. This is effectively a kind of elimination. In the case of plants, metals appear to be sequestered predominantly as granular deposits in cell vacuoles [100]. Similarly, soil invertebrates are able to store metals such as Cd in special hepatopancreatic cells as inert granules [101, 102].
The form in which metals are stored in prey species has implications for bioavailability of metals to higher trophic levels (predators) [103-105]. In an attempt to discriminate between the various forms of sequestered metal, Wallace *et al.* [106, 107] defined various sub-cellular partitions based on a centrifugal protocol that partitioned animal tissue into five separate fractions: metal rich granules (MRG), cellular debris, organelles, heat stable proteins (HSP) and heat denatured proteins (HDP). Using this protocol in an examination of the relationship between sub-cellular Cd distribution in an oligochaete and its trophic transfer to a predatory shrimp [108], the authors proposed that only metal present in the soluble fraction (*i.e.* organelles and protein fractions) of prey is available to the predator. Similar conclusions were drawn in a study using bivalves as prey; the authors again concluded that the metal partitioning to organelles, HDP and HSP, comprise a sub-cellular compartment that is "trophically available" to predators [106]. This model seems to have established a degree of acceptance and provides a pragmatic approach to resolving subcellular metal distribution, and has been shown to yield useful insights with a variety of species: e.g. bivalves, perch and earthworms. However, refinements to the protocol will be required before it can be used to predict bioavailability of prey-bound metal because subsequent studies have demonstrated that the soluble fractions were not always 100% bioavailable and that the insoluble fractions were not always 100% non-bioavailable [105, 109].
# **CONSEQUENCES OF BIOMAGNIFICATION FOR POPULATIONS DECLINES/SHIFTS OR ECOSYSTEM EFFECTS**
# **Direct Poisoning by Metals**
Occasionally, human activities can lead to deposition of metal in such high concentrations that biomagnification is not required before symptoms of toxicity can be observed in wildlife, and result in the direct poisoning of higher vertebrates with metals. The most obvious example of this is through the use of lead shot for hunting and lead fishing sinkers. A recent study by Mateo *et al.* [110] reported up to 148 lead shot pellets per square meter in wetlands in southern Spain. Poisoning occurs when animals ingest spent shot. Waterfowl can ingest lead shot directly while feeding, and galliform birds can ingest shot as grit for their gizzards. Also, predators and scavengers ingest lead shot or lead shot fragments that are embedded in carcasses or that persists in the gizzards of birds [111, 112].
The first studies of Pb poisoning in waterfowl were conducted in the USA in as early as 1959 [113]. Since then, various scientific studies have examined Pb pollution in wetlands and the possible consequences of direct or indirect Pb poisoning on water birds. In a laboratory study, a single lead shot experimentally imbedded into the crop of ringnecked ducks (*Aythya collaris*) caused mortality or severe symptoms of debilitating toxicosis among 87% of birds [114]. Lead is a non-specific poison affecting numerous body systems, and sublethal exposure has variously been reported to result in (i) weakness of contaminated birds making it vulnerable to predation; (ii) alteration of energy requirements that will consequently be a handicap for birds on migration; (iii) reductions in clutch size; and (iv) result in eggshell thinning and embryo malformations [111, 115]. Most of these toxicological studies were conducted on species of ducks and swans.
Lead poisoning among waterfowl has resulted in several population declines. Population declines of mute swans (*Cygnus olor*) in the UK between the 1960s late 1980s were associated with ingestion of lead sinkers. Their numbers
subsequently increased following the ban on sales of small lead sinkers in the late 1980s. Similarly, up to 50% of recorded mortalities among common loons (*Gavia immer*) in Canada were also associated with ingestion of lead fishing sinkers [116]. Additional investigations to see if Pb poisoning could have a similar effect on waders (common snipe *Gallinago gallinago* and jack snipe *Lymnocryptes minimus*) concluded that Pb poisoning also affects waders to a similar extent to that found in ducks, thus confirming previous studies on aquatic birds [117]. The use of lead sinkers for recreational fishing and lead shot for waterfowl hunting has been restricted in the USA and various other countries since the 1980s/1990s. However the incidence of Pb poisoning among raptors has not decreased in the USA and it seems likely that the continued use of lead shot for upland hunting has shifted the emphasis from water birds to predatory birds [116, 118].
Mercury poisoning among mammals is exemplified by Minamata disease. Minamata disease is manifested as severe neurological pathologies among victims exposed to Hg-contaminated food. In Minamata, Japan (the location from which the syndrome takes its name), poisoning resulted among human inhabitants following consumption of fish contaminated with biomagnified Hg (see below). However, similar neurological pathologies have been observed among people and birds following the use of organo-mercurial fungicides for the protection of grains and as a consequence of ingestion of contaminated seed [31].
Arsenic poisoning in animals and humans is caused by several different types of inorganic and organic arsenical compounds. Toxicity varies with factors such as oxidation state of the As, solubility, species of animal involved, and duration of exposure. Organic forms appear to have a lower toxicity to mammals than inorganic As. Research has shown that arsenites (trivalent forms) have a higher acute toxicity than arsenates (pentavalent forms) [119]. In mammals, As is known to promote cancer of the bladder, lung, and skin. Arsenic poisoning among the human population of Bangladesh occurs as a consequence of direct consumption of ground water, which has naturally high concentrations of As. However, because groundwater is also used for irrigation, contamination of rice (the staple crop in Bangladesh) is also a likely source of As [120]. Many plants are able to accumulate As, and some are able to hyperaccumulate this metalloid. For example, the Chinese brake fern (*Pteris vittata*) is able to hyperaccumulate As (more than 1000 mg As/kg of shoot dry weight) as As(V), reduce it to As(III), translocate it through the xylem with water and minerals as an As(III)-S compound, and then store it as As(III) in the fronds [121], thus making trivalent As available to herbivores. It is not known to what extent As poisoning occurs in animals that feed on plants that accumulate As, but because there is no threshold below which As intake is regarded as safe for humans, there is clearly potential for poisoning among other herbivorous species following the consumption of plants.
The occurrence of arsenic poisoning in Bangladesh is a similar scenario to that of *itai-itai* disease in post-WWII Japan. *Itai-itai* (meaning 'painful' in Japanese) disease was the name given to cases of cadmium poisoning among the human population living around the Jinzu River that had been contaminated with Cd originating from a zinc mine further upstream. *Itai-itai* disease is characterized by symptoms of osteoporosis and osteomalacia associated with renal dysfunction resulting from chronic Cd poisoning. Poisoning occurred because people either drank or cooked with the river water, or because they ate rice that had been irrigated with Jinzu River water [122]. Kobayashi *et al.* [122] performed multiple regression analyses, and found a strong correlation between consumption of Cdcontaminated rice and the occurrence of renal tubular dysfunction, indicating that Cd poisoning in the human population has occurred as a consequence of biomagnifications of Cd through rice.
# **Populations Under Threat from Biomagnification**
Bioaccumulation of metals as a consequence of biomagnification or bioconcentration is frequently observed among lower trophic levels; however, elevated tissue burdens of metals do not necessarily translate into shifts in community or population structure. Again the studies by Hunter *et al.* [7, 8] are instructive. As indicated earlier, the vegetation structure was only altered in the highly metal-contaminated zone around the copper refinery, which was dominated by a few metal tolerant plant species. Although de-vegetation and dominance of contaminated soils by metaltolerant plant species has been demonstrated in other sites with high levels of metal contamination [e.g. 123, 124, 125], moderate levels of metal contamination generally have little effect on plant communities.
Another notable difference between the highly contaminated zone near the refinery and less contaminated areas studied by Hunter *et al.* [8] was the lower abundances of isopods and oligochaetes in the most highly contaminated zone. Other studies have also shown earthworms to be sensitive to high levels of metal contamination in soils [126]. However, at sites where metal concentrations are moderate, soil community shifts are often subtle or masked by other environmental factors [127] and indicates that bioaccumulation of metals in lower trophic levels does not necessarily impact on invertebrate community structure [128].
However, impacts in vertebrate consumers/predators might be expected to be more pronounced because of the large volumes of prey taken. For example, godwits (*Limosa limosa*) are migratory waders that breed and feed at numerous contaminated sites in The Netherlands [129]. Although their residency at contaminated sites is temporary, godwits feeding on worms that were known to accumulate metals, in turn accumulated Pb, Hg and Cd from worms. *Limosa limosa* is listed in the so-called Dutch Red List (a list of endangered species that are protected by special legislation) and is therefore a conservation priority species. Numerous factors are implicated in the population declines of this species and it is not possible to isolate metal contamination as an important factor.
Arsenic poisoning has been implicated in declines of small mammals in alpine regions of the Snowy Mountains of Australia, although the causal links remain speculative. Bogong moths (*Agrotis infusa*) migrate annually from the agricultural plains to alpine regions to estivate and carry with them large body-burdens of As. Between the time when arsenic was first reported from two mountain estivation sites in the summer of 2000/2001 [130] and the following summer, the amount of arsenic found in moths increased by at least an order of magnitude [131]. Over this period, populations of the herbivorous broad-toothed rat (*Mastacomys fuscus*), the insectivorous dusky antechinus (*Antechinus swainsonii*) and the omnivorous mountain pygmy possum (*Burramys parvus*) declined, and As was detected in the faeces of *A. swainsonii* and another non-declining species, *Rattus fuscipes*. Because the declines in mammals were across several trophic groups (*i.e.* herbivores and insectivores), As poisoning is not likely to be the predominant causal factor, but may be a co-factor that needs to be considered in the declines of rare species like the mountain pygmy possum (*B. parvus*). The source of As in the moths was not known [130, 131], but it is presumed to come from the breeding grounds of the moths in the agricultural plains of southern Queensland and northern NSW [130]. As-based pesticides have been used in Australia since the early 1900s [132, 133] and continue to be used to a lesser extent in the form of the organoarsenic herbicides (e.g. monosodium methylarsonate; MSMA).
MSMA has also been applied widely in British Columbia, Canada to control outbreaks of Mountain Pine Beetle [134]. As a consequence, beetle larvae accumulated between 1.3 and 700 g As/g (dry weight). Subsequent feeding by insectivorous woodpeckers and other forest passerines breeding within 1 km of MSMA stands contained elevated blood concentrations of total arsenic (geometric mean = 0.18 g/g; range = 0.02 to 2.20 g As/g) [135]. This range of whole blood concentrations is similar to the range of concentrations found among zebra finches provided orally with monomethylarsonic acid (MMA(V)) at doses between 8 and 72 g/g/day over 14 days [136]. Oral doses of this magnitude were found to cause weight loss among adult finches after 14 days [136] and mortality among nestling zebra finches after 20 days [137], indicating that natural field populations of passerine birds may be affected in MSMA treated areas.
As indicated in the introduction, various small mammals are able to persist in highly contaminated sites. Of particular note is the persistence of common shrews (*Sorex araneus*) in sites with high levels of Cd contamination. Shrews are higher order predators in terrestrial environments, and because of their high metabolic demands, they consume large quantities of prey. In Cd-contaminated sites, shrews accumulate very large body burdens of Cd. In the study by Hunter *et al.* [81], shrews in the vicinity of a cadmium/copper refinery accumulated from their prey Cd residues in excess of 200 μg/g and 500 μg/g (dry weight) in the kidney and liver, respectively. Similar liver-Cd concentrations have also been reported by other authors [80, 138, 139]. It remains unclear if the health of shrews in these studies was compromised by the high body burdens of Cd. In an earlier report, Hunter *et al.* [140] described lesions in the kidney and liver of shrews from the same location and with similar Cd burdens. Later laboratory studies [73, 141] demonstrated reduced weights (but no mortality) among shrews exposed to Cd-contaminated diets for a period of 12 weeks. In those studies, test animals accumulated kidney and liver Cd burdens in excess of 1000 μg/g. It seems likely that shrews have evolved to tolerate high levels of metals [142].
Other species do not appear to be so tolerant. Damek-Poprawa and Sawicka-Kapusta [143] described histopathological changes in the liver and kidneys of herbivorous bank voles (*Clethrionomys glareolus*) with much lower tissue burdens of Cd in kidney (33 μg/g dry weight) and liver (16 μg/g dry weight). Cadmium is also known to compromise reproduction [144], and accumulation of heavy burdens of Cd may be limiting recoveries of European badgers (*Meles meles*). Van den Brink and Ma [145] found that the quantities of Cd and Zn in the kidneys of badgers in the different regions of The Netherlands were negatively correlated with the increase in the number of breeding dens (setts) in these regions and hence with the number of cubs born. The highest kidney Cd burdens (101 to 405 g/g dry weight) were found among adult female badgers living close to rivers where earthworms (a major prey species) accumulated large burdens of Cd.
Although studies like those conducted by Hunter *et al.* [9, 81] and Mertens *et al.* [138] provide evidence that small mammal communities can persist even in highly contaminated environments without apparent detriment, other studies show that community structure is likely to be altered in the face of severe metal contamination. The Tar Creek Superfund Site in Oklahoma, USA, has a long history of metal contamination as a consequence of mining activities, and elevated tissue concentrations of Cd have been documented in several non-mammalian inhabitants [146, 147]. Although tissue-metal concentrations were not reported, Phelps and McBee [148] reported reduced species diversity among small mammal assemblages at Tar Creek compared to reference locations. At reference sites, several small mammals were recorded; however, in the contaminated sites, white-footed mice (*Peromyscus leucopus*) predominated, and in greater numbers than found in reference sites. White-footed mice have previously been demonstrated to dominate small mammal assemblages in disturbed environments [149, 150].
The capacity for white-footed mice to dominate in disturbed environments is note-worthy. In the case of the study by Levengood and Heske [150], the authors cited very high soil concentrations for Cd, Hg and Se (as well as other metals) in a wetland site that had received sediments dredged from Lake DePue, Illinois, in 1982. The wetland is periodically inundated as a management strategy for the conservation of waterfowl. White-footed mice, because of their arboreal habit, were able to utilise this habitat and persist with tissue burdens of Se that have been shown to be toxic to rats in laboratory studies. The white-footed mouse is an omnivore, and likely consumes invertebrates as well as plant seeds. Tissue metal (Cd and Pb) concentrations in grasshoppers and crickets were high (2.5 to 4.8 μg/g), and it seems likely that mice were exposed to relatively high levels of metal in their diet. However, apart from Se, kidney- and liver-metal concentrations remained comparatively low, indicating that the white-footed mouse was not under threat as a consequence of biomagnification of toxic metals. The mechanism by which white-footed mice are able to avoid bioaccumulation of metals is unknown, although tolerance to high levels of metals and other contaminants is likely related to the efficiency by which this species can detoxify reactive oxygen species as well as highly efficient DNA repair mechanisms [151].
Some metals only enter the higher terrestrial food chain after commencing their trajectories through aquatic food chains. Selenium and Hg are two such metals. Selenium is bioaccumulated in aquatic habitats and organoselenium compounds can be bioconcentrated over 200,000 times by zooplankton when water concentrations are in the 0.5 to 0.8 μg Se/L range. Inorganic selenium bioaccumulates more readily in phytoplankton than in zooplankton. Phytoplankton can concentrate inorganic selenium by a factor of 3000. Further biomagnifications occurs along the food chain, as predators consume selenium rich prey. A water concentration of 2 μg Se/L is considered highly hazardous to sensitive fish and aquatic birds. Compounding this trophic biomagnification, selenium poisoning can also be passed from parents to offspring through the egg, and selenium poisoning may persist for many generations [152]. As indicated above, selenium migrates into terrestrial ecosystems when birds or other taxa feed on aquatic organisms. The best studied example comes from California in the USA.
In the 1980s Kesterson National Wildlife Refuge, CA was contaminated with Se as a consequence of subsurface irrigation drainage. Selenium and other trace elements were leached from agricultural soils in the San Joaquin Valley, and the excess water was transported to Kesterson as a wildlife management strategy. Elevated body burdens of selenium were found in all animal taxa examined, including mammals, birds and reptiles [152]. The concentrations of Se in livers taken from waterfowl typically ranged between 20 and 100 μg/g and were significantly higher than in similar birds sampled from a reference site [153]. Selenium, although an essential element required for normal development, is toxic even when exposure is only slightly higher than essential requirements. Among waterfowl in Kesterson, particularly among eared grebes (*Podiceps nigricollis*), very low hatching rates were ascribed to selenium-induced embryotoxicosis [154]. More recent studies in Kesterson have continued to report elevated level of Se in passerine species such as starlings (*Sturnus vulgaris*) and small mammals, but without associated effects on hatching success or population declines [43, 44].
As indicated above, mercury enters the terrestrial food chain when predators eat aquatic species that have bioaccumulated the metal. The best known example of Hg poisoning in a terrestrial mammal is that of Minamata disease among human inhabitants of Japan following consumption of contaminated fish downstream of a Hgcontaminated river. Minamata disease is characterized by severe neurological disorders, and similar neurological symptoms of poisoning among mammalian predators such as predatory birds and mink have also been reported [31, 39]. Among species of mink (*Mustela* sp.), a lowest observable adverse effects level (LOEL) for Hg of 5 g/g (wet weight) in the brain has been established in laboratory studies [36]. Brain Hg concentrations of this order of magnitude (0.5 to 5.0 g/g wet weight) are common among wild mink [36], and some authors have speculated that mercury poisoning may be responsible for declines in mink (*Mustela vison*) in Georgia, North Carolina, and South Carolina [155], although the co-occurrence of chlorinated organic compounds in mink tissues was also implicated.
Biomagnification of metals in higher predators, such as predatory birds, large carnivorous mammals and reptiles [156- 158], is more difficult to demonstrate. Predatory birds and large mammals are more sparsely distributed and far more mobile, having home ranges that go well beyond localised contamination. Contaminated prey, though likely forming part of their diets, will be diluted with prey from less contaminated sites. In the case of reptiles, a combination of low metabolic rate typical of poikilothermic animals and relatively low assimilation efficiencies [72, 159], is likely to reduce the risk of metal bioaccumulation. However, reptiles in general are under-represented in the ecotoxicology literature [160] and it would be premature to suggest that reptiles were not at risk from exposure to metals.
# **CONCLUSION**
Direct metal poisoning of higher vertebrates is found sporadically, and examples thereof are mostly found for Pb. Magnification in the food chain *via* prey having elevated body burdens of metals is shown more often. The level of biomagnification that can be expected within the field situation is difficult to predict, and depends on five factors, namely: (i) metal specificity; (ii) exposure route of the predator; (iii). excretion possibilities of organisms; (iv) predator specificity such as gender and life span, and (v) metal sequestration in the prey organism.
In general, carnivorous populations show higher biomagnification compared to herbivorous organisms. Aquatic food chains are less at risk than terrestrial food chain when it comes to biomagnification of metals and metalloids. To assess biomagnification effects at community level, unraveling ecosystem complexity is necessary before species most exposed and at risk can be identified. Therefore, for the purpose of setting environmental quality objectives, it is important to include biomagnification as well.
The final objective of studies that examine biomagnification and toxicity of metals is to provide sufficient protection to all biological organisation levels. Observations in the laboratory and field have demonstrated that secondary poisoning of, for example, worm-eating birds and mammals may be more critical than direct exposure of soil organisms. In such cases, quality criteria that are protective of lower trophic levels, may not provide protection for top predators.
# **FUTURE PROSPECTS**
Several metals are familiar as common contaminants in our environment by virtue of the fact that they have been used for many decades or centuries for industrial and agricultural purposes. Indeed, the various phases in human civilization are characterised by our use of specific metals. The Copper Age (c. 3200–2300 bc), the Bronze Age (2300–700 bc) and the Iron Age (700–1 bc), mark the discovery and adoption of these metals. The more recent industrial ages have seen the use of many other metals including those which now present a pollution hazard, such as Cd, Pb and Hg. Much of our knowledge about the environmental risks posed by metals comes from research on these metals. However, the industrial age is far from spent, and recent decades have seen increased reliance on several other metallic elements. In the immediate future the continued and increased reliance on coal combustion for electricity production will result in increased diffuse contamination of soils and oceans with various elements, including Hg, As, Cd, Cu, Pb, Se, Zn and various others because they are volatile species which are not retained by flue gas filtration systems [161]. Therefore, it can be expected that the environmental consequences of bioaccumulation and subsequent biomagnification of various metals, particularly Hg, Se and Cd are yet to be fully realised.
Other, less abundant metals are also increasing in our environment, due to use of a multitude of metals in innovative new applications like the broad area of nanotechnology and related fields (nanofood, nanocosmetics, nanopharmaceuticals, etc). Some of these new metal-based products will follow well understood metal uptake kinetics. However others may potentially find entry into biological systems *via* yet to be discovered modes of biouptake. Only time will tell if biomagnification and subsequent toxicity of these innovative combinations of metals and metalloids will occur.
# **ACKNOWLEDGMENT**
Martina Vijver was supported by a NWO VENI-grant, project number 863.08.023.
# **REFERENCES**
© 2011 The Author(s). Published by Bentham Science Publisher. This is an open access chapter published under CC BY 4.0 https://creativecommons.org/licenses/by/4.0/legalcode
# **CHAPTER 4**
# **Impacts of Agricultural Pesticides on Terrestrial Ecosystems**
# **Francisco Sánchez-Bayo\***
*Centre for Ecotoxicology, University of Technology Sydney, Australia*
**Abstract:** Pesticides are toxic chemicals used to control pests, weeds and pathogens. Three quarters of all pesticides are employed in agricultural production, particularly in developed countries, in an effort to mitigate crop damage endured by intensive agriculture. However, after more than 60 years of worldwide usage, their side-effects on terrestrial ecosystems – even when applied as recommended – are obvious. This chapter examines the ecological problems caused by specific chemicals/groups, so that this awareness may help improve agricultural practices through appropriate risk management. Fungicides alter the microbial-fungi communities responsible for the recycling of nutrients in the soil, and copper fungicides are toxic to earthworms and other animals. The routine application of herbicides has produced a net loss of plant biomass and biodiversity in many landscapes, which indirectly reduces the associated arthropod communities and leads to population declines in many species of birds, and possibly amphibians too, due to lack of food. Insecticides are very toxic to most invertebrates in the soil, birds and small mammals, causing significant reductions in their populations and disturbing the trophic structure of their communities. Persistent pesticides accumulate in soil and concentrate through the trophic chain, causing a plethora of sublethal effects which are negative for the survival of individuals as well as the viability of their populations; the long term effects of DDT and cyclodiene poisoning in birds is still an ecological issue despite more than 30 years of not being applied in most developed countries. While pesticides have increased our agricultural productivity and helped feed the current human population, the price of this productivity is being paid by the Earth's ecosystems at large.
# **INTRODUCTION**
Since Neolithic times, humanity has learnt to use agriculture to supply the food needed for its own sustenance. Agricultural practices first started with cereal crops in the Fertile Crescent about 11,000 years ago, and subsequently developed in other regions of the world, although a rather small suite of 35 domesticated plants and seven animals ended up established over the world because of their yield and nutrient characteristics [1]. For centuries, most of the staple plant foods have been cultivated as monocultures: unusual ecosystems in which no diversity of plants other than the crop is allowed to grow on the same land in order to maximize crop yields, and where all means possible are used to ensure this is the case; the unwanted, competing plants are called weeds. Because of this feature, monocultures are ideal targets for specialized consumer animals (usually insects, birds and rodents) that feed on them. Once such animals find a crop that suits them, they multiply explosively and become pests. With the exception, perhaps, of locust plagues all other agricultural pests are a product of monocultures, and from early times humanity has struggled to keep at bay the pest species that decimated our crops.
As with agriculture, the story of pesticides – the substances used to control and kill pests – started in the Middle East. The Persians found that the extract of certain chrysanthemum flowers (known as pyrethrum) was very effective in killing flies and other insects, so they used it to control agricultural pests [2]. Late in the 18th century, Erasmus Darwin found nicotine (the extract of Nicotiana tabacum) to be a powerful insecticide, and early in the 1900s arsenic salts were also used to control a wide range of pests, particularly in orchards. However, it wasn't until the 1940s that a revolution in pesticides took place, when the chemical industry started to massproduce synthetic toxic substances that were effective, not only in killing insects (insecticides) and other animal pests (rodenticides), but also weeds (herbicides) and fungal diseases (fungicides). The rapid development that ensued, especially in North America, Europe and Asia-Pacific, led to the establishment of a new kind of agriculture based on chemistry. The so-called Green Revolution involves the use of chemical pesticides and fertilizers together with increased irrigation and genetic improvement for agricultural production. Hailed as the saviour of human starvation, the Green Revolution practices were quickly adopted worldwide, particularly in densely populated countries of South East Asia such as Indonesia and the Philippines, where food shortages were soon replaced by bumper crop yields [3]. Indeed, the use of pesticides in agricultural production became so widespread that the term 'conventional agriculture' indicates a cropping system where the Green Revolution tools are applied routinely.
**Francisco Sánchez-Bayo, Paul J. van den Brink and Reinier M. Mann (Eds) © 2011 The Author(s). Published by Bentham Science Publishers**
**<sup>\*</sup>Address correspondence to Francisco Sánchez-Bayo:** Centre for Ecotoxicology, University of Technology Sydney, NSW 2007, Australia; Department of Environment, Climate Change & Water NSW, 480 Weeroona Road, Lidcombe NSW 2141, Australia; Email: [email protected]
### **64** *Ecological Impacts of Toxic Chemicals Francisco Sánchez-Bayo*
While the Green Revolution was producing 'miracles' everywhere, the newly developed pesticides applied to an increasing variety of crops started to have side effects in the surrounding natural ecosystems. Bioaccumulation of DDT and cyclodiene insecticides was first noticed in bird predators like the peregrine falcon (*Falco peregrinus*) despite the fact they had little relation to the sprayed crops [4]. Through a long and painstaking research that involved many experts in the areas of environmental chemistry, toxicology and agriculture [5], it was eventually revealed how these chemicals had secondary and indirect effects on non-target organisms, and their impacts on the structure and functionality of natural ecosystems rang the alarm in environmental circles. Even the direct effects of insecticides on arthropod communities, and the birdlife that depended on them, was brought into question by Rachel Carson as early as 1962. The birth of the environmental movement and ecotoxicology was thus linked from its very beginnings to the widespread use of synthetic pesticides in agriculture, forestry, and urban pest control. It was realised that all pesticides are toxic to a greater or lesser degree, so their release could not be without risks to some kind or other of organisms.
Pesticides are the only man-made contaminants released into the environment deliberately, for a purpose; whereas industrial chemicals, mining wastes, pharmaceutical residues and the large list of pollutants that humanity produces find their way into the air, rivers and oceans either unintentionally or because our technology is still unable to reduce their emissions, avoid accidents, and too inefficient to recycle the wastes.
# **PESTICIDES IN AGRICULTURE**
There are currently 835 chemical compounds used in all sorts of agricultural enterprises [6], comprising some 1300 registered products, of which 31% are herbicides, 21% insecticides, 17% fungicides, 9% acaricides and 2% rodenticides; the remaining 20% of products include a plethora of biocides for control of snails (molluscicides), algae (algicides) and nematodes (nematicides) as well as plant growth regulators (6%) and natural or artificial pheromones (5%). In addition, 610 products, including most of the infamous organochlorine (OC) insecticides, were used in the past but not nowadays – they were banned for safety and environmental reasons or because they were no longer efficient (due to resistance) and have been replaced by newer products. Despite using so many chemicals, world crop losses are estimated at 37% of agricultural productivity: 13% due to insects, 12% to weeds and 12% to diseases [7].
The toxicity and specificity of pesticides depends on the mode of action of the active ingredients (a.i.), while the effects on organisms depend on the dose they are exposed to (see Chapter 1). Thus, organochlorine, cholinesterase inhibitors (organophosphorus (OP) and carbamates), synthetic pyrethroid and neonicotinoid insecticides are neurotoxic substances that disrupt the nervous system of arthropods and other animals. Given the similarities in neuronal physiology among all kinds of animals, it is not surprising that insecticides are also toxic to aquatic and terrestrial arthropods and, to a lesser extent, vertebrates, whereas they are harmless to plants and the majority of microbial organisms. Other insecticides affect cellular or physiological mechanisms of animals (e.g. chlorfenapyr, arsenic salts). Herbicides are very toxic to plants and algae, as they target physiological pathways specific to plants such as the photosynthesis; however, herbicides can interfere with metabolic and reproductive processes in animals as well, often in ways that are unrelated to their specific mode of action in plants. Fungicides are considered in some countries to be medicine for the crops as they control fungal infections of the roots or other parts of the plant; many of them are antibiotics or metabolic inhibitors of certain fungi, while organomercurial compounds are neurotoxic and poisonous to many animals. Rodenticide poisons are usually anticoagulants, and consequently are very dangerous to humans and all vertebrates alike. Thus, the specificity of action of pesticides is not restricted to the target pest or weed species, but it is rather general, affecting large taxonomic groups often at the order or class level, even though within the same class of organisms some species are more susceptible than others due to differences in body size and/or physiological traits [8].
# **Pesticide Usage**
Global pesticide usage is estimated at 4 million tons per year [9], although its distribution throughout the world is very uneven [10], with Europe using one third and North America a quarter of the total market until recently (Table 1). Herbicides account for nearly half of the pesticides used in North America, insecticides 19%, fungicides 13%, with the remaining 22% including a variety of other products [11], whereas insecticides are prevalent in developing countries. Agricultural industries, i.e. crops and livestock, are the main users of pesticides in the USA and other countries (74% of
# *Impacts of Agricultural Pesticides on Terrestrial Ecosystems Ecological Impacts of Toxic Chemicals* **65**
the annual consumption), with gardening, golf courses, industry and urban uses making up 25% of the total amount whilst only 1% is being used in forestry [7], mainly in Canada and Scandinavian countries. DDT and lindane are still used in countries like India [12]; by necessity most of the DDT is to control mosquito-vectors of malaria and tse-tse fly in tropical countries and South Africa, where no other cost-effective chemicals are available. The distribution of pesticide types among crops differs widely: corn, soybean and cotton crops are the main users of herbicides in the USA (75%); orchards use mainly insecticides, while vineyards and vegetables use most of the fungicides [7].
**Table 1:** Annual pesticide usage in the world up to 1996. Source: [10]
\* USA and Canada only
Average pesticide application in developed countries is 4.4 kg/ha per year. Since almost one fifth of the Earth's land area is dedicated to agriculture (12% as cropland and 6-8% as pastureland [13]), the impact of agrochemicals on ecosystems is quite significant at a global scale. However, not all agricultural land is treated with pesticides: in the USA, for instance, some 38% of the acreage is not treated with chemicals [7].
# **Application of Pesticides**
Agricultural pesticides are typically applied directly on to the crop plants or fruit trees by spraying them in a liquid carrier (oil or water mixed with surfactants) that can be delivered by plane, helicopter, ground machinery or simply by hand-operated sprayer-guns. Some pesticides are applied as granules buried in the soil, or as seed-dressings to protect the growing seedlings.
The method of application greatly determines the exposure of non-target organisms to pesticides. For instance, 25- 50% of the pesticide sprayed from aircraft reaches the crop, or 65-90% if sprayed with ground machinery [14]; the remainder is scattered around the target crop/orchard, with the spray droplets reaching distances up to 1.5 km under established conditions for application, *i.e.* low flying path, wind speeds between 3 and 15 km/h and no air inversions. Further drift can occur whenever these requirements are not met, as often happens with inexperienced personnel especially in developing countries. Not surprisingly, wildlife populations are systematically being affected every year by direct exposure to insecticide sprays, specially birds that are present in agricultural areas at the time of insecticide spraying [15] and receive a high dose *via* droplets or concentrated toxic vapours [16]. Exposure of terrestrial animals to herbicide sprays is less hazardous because of their lower toxicity. However, aquatic ecosystems and susceptible crops in nearby land can be affected as well, so the adoption of buffer zones around the crops can substantially mitigate the drift onto surrounding areas. For example, unsprayed strips 3 m wide around agricultural fields in the Netherlands reduced drift onto irrigation ditches by 95% [17]. Under present management practices in that country using narrow unsprayed buffer zones and other measures, the impact of sprays on non-target insects are down to 41% for herbicides, 21% for insecticides and 14% for fungicides compared to impacts in the past [18].
Granular pesticides are designed to avoid the risks of spray drift to farmers/applicators and wildlife. Also, the granules release the active ingredient over time, thus increasing the efficacy of the product. Many water soluble herbicides and fungicides are applied as granules, as well as some OP (fensulfothion, terbufos, parathion, fonofos, disulfoton, phorate, diazinon) and systemic insecticides (aldicarb, bendiocarb, carbofuran, imidacloprid). Special machinery is used to bury the granules in the soil, but inevitably some granules remain exposed on the surface (from <1 to 50% depending on conditions), where birds and other animals may ingest them [19]. Birds are particularly fond of such granules, which they take as grit for their gizzards or simply mistake them as food, and consequently are more at risk from this formulation than small mammals [20]. In the case of insecticides, a single granule may contain a lethal dose (up to 20% a.i.), so the consequences are often dramatic: in North America, waterfowl were poisoned by eating fonofos granules they sifted from waterlogged fields six months after they were applied [21]. Seed-dressing was a common practice with OC insecticides such as aldrin, dieldrin and lindane, as well as organomercurial fungicides, and it poses similar risks as the granules, *i.e.* granivorous birds and mammals ingest the treated seeds often spilt around the edges of the crop and farm buildings. Poisoning incidents with seed dressings of cholinesterase inhibitors are still relatively frequent, especially in Europe [22]. The systemic insecticide imidacloprid is often used as a dressing for maize, sunflower and rape seeds; when the plants grow the insecticide is still present at concentrations ranging from 4.1 mg/kg in stems to 2.1 mg/kg in pollen, thus causing a great risk to honeybees [23]. Rodenticides are applied as baits spread around the farm buildings or near the crops where pest mice or voles congregate, posing a risk to other non-target vertebrates.
In irrigated crops, herbicides are often poured into the water channels either to allow an even distribution of the chemical throughout the irrigated field or simply to eliminate aquatic plants that may clog the channels and use up the water. Treated waters such as these invariably affect aquatic communities in agricultural landscapes (see Chapter 6), and are a constant source of contamination for many birds, frogs and mammals that bath in or drink from them.
Finally, some insecticides are used to control ectoparasites in domestic animals. In the 1950-60s it was common practice in many places to drench farm animals with solutions of DDT to combat cattle ticks. Today, the OC insecticides have been replaced by OPs (e.g. famphur), pyrethroids (e.g. cypermethrin), spinosad, cyromazine, avermectins and insect growth regulators (e.g. fluazuron) to control ticks, lice and blowfly maggots. Despite their lesser persistence and greater specificity, residues of the latter chemicals in dung from treated livestock affect dungbreeding insects and the degradation of faeces [24].
# **EXPOSURE OF ORGANISMS TO AGRICULTURAL PESTICIDES**
Animals and plants are exposed to all these toxicants in a variety of ways. It is important to realise that just as the target pests and weeds are killed by the pesticides, all other non-target organisms may suffer deleterious or deadly consequences when exposed to the same doses of those chemicals.
# **Animal Exposure**
The first route of pesticide exposure for most animals, vertebrates and invertebrates alike, is by direct deposition of the sprayed products on them, which is equivalent to a topical application on their skins/epidermis. Spray droplets are made of concentrated active ingredient in an oily or water-based carrier solution that sometimes contains an adjuvant. The tiny droplets (100-200 m in diameter [25]) deliver a concentrated dose of toxicant to the skin, hair and feathers of animals they fall upon. Thus, liphophilic insecticides are quickly absorbed through the skin, and the ensuing acute dermal toxicity is often enough to kill the animal. In fact, dermal deposition has been recognized as one of the most crucial routes of exposure in birds [26]. Animals that die as a consequence of direct pesticide spray deposition do so because they happen to be at the wrong place at the wrong time [15], but it is hard to imagine how this could be avoided since agricultural land and surrounding landscapes are the natural home to countless species of non-target organisms of all kinds. Inevitably, pesticides and fertilizers are applied during the crop growing season, which coincides with the breeding of insects, nesting of birds and breeding/metamorphosis of amphibians. Although application schedules are dictated by crop pest/weed infestation levels and other management practices, the timing of application can have different impacts. For instance, the OP insecticide dimethoate applied early (spring) to barley crops at maximum rates (0.4 kg/ha) was very harmful to seven non-target soil-dwelling breeding beetles, but the same rate has a reduced impact on populations of old beetles when sprayed in autumn [27].
Concomitant with the deposition of spray droplets, inhalation of the misty and vaporized pesticides brings the active ingredients directly into the lungs and bloodstream of terrestrial vertebrates, even if they were initially sheltered from the spray deposits. Volatilisation of lipophilic insecticides from soil and other surfaces is a source of constant air contamination in agricultural areas [15] even years after they were applied. For example, fluxes of DDE, toxaphene, dieldrin and trans-nonachlor from cotton soils in Alabama (USA) have been estimated between 325 and 7000 kg annually or 0.07-1.56 mg/kg per day for each of the respective chemicals [28]. Animals with a high rate of ventilation such as birds are at the highest risk. Nevertheless, it is difficult to separate the two kinds of exposure mentioned here – direct contact and inhalation – when an animal has been found paralysed or dead in the field. Most of the time it is the combination of several routes of exposure that accounts for the fatalities observed.
The third route of exposure is by direct consumption of contaminated plants, fruits, granules and coated seeds. This is known as primary poisoning to distinguish it from the secondary poisoning that occurs when a predator eats contaminated prey, insects or worms containing pesticide residues. Primary poisoning also occurs through drinking of contaminated waters from irrigation channels, drains, farm reservoirs, puddles, streams, rivers and lakes, which may contain high levels of pesticide residues, especially when they are in or near the agricultural fields that act as their source. A typical example is the case of DDT and cyclodiene insecticides used lavishly in the past; the persistence and lipophilic characteristics of these OCs resulted in their accumulation in granivorous birds and rodents that consumed seeds dressed with aldrin or dieldrin, in caterpillars that fed on leaves, and in worms of the treated soil – exposure through primary poisoning. In turn these animals were eaten by insectivorous birds and small predators, so the residues accumulated in their bodies as well. Larger predators such as falcons and eagles ate the contaminated prey and ended up with insecticide concentrations in their bodies which were several thousand times those found in the original seeds or treated plants – secondary poisoning. A parallel chain of contamination events occurred in the aquatic ecosystems where residues of these insecticides found their way through washoff from plants, runoff and drift [29]. Fortunately, most modern pesticides do not accumulate in organisms because they are either metabolized readily or eliminated in the urine and faeces. This does not mean they are all safe in regard to trophic contamination; *i.e.* woodlice consuming litter materials contaminated (0.1-500 g/g food) with parathion-ethyl and endosulfan-sulfate take up these insecticides and experience their toxic effects [30]. Nor does it mean that secondary poisoning is a phenomenon relegated to past use of OC insecticides; it still occurs wherever the land was treated with arsenates [31] and OCs, as well as in tropical regions where they are still in use. Evidence of regular wildlife contamination by ingestion has been demonstrated by analyzing the gut contents of passerine birds in Australia during the agricultural season; sublethal levels of OC insecticides were found in 41-63% of the birds sampled, the OP parathion-methyl in 22% and the herbicide diuron in 78% of the birds [32]. The distribution of residues among trophic levels suggests that insecticides were obtained through ingestion of food whereas the herbicide was acquired by drinking from polluted waters. The highest residues were DDT (35-1980 g/L) and its metabolite DDE (2-21 g/L) even if it had not been used in that country for 20 years. Although the bioavailability of such old residues in soil and sediment decreases considerably with time [33], the fact that many animals continue to show DDT/DDE in their body tissues decades after they were applied indicates that the movement of this insecticide through the food chain is still a current issue in ecotoxicology.
Organisms exposed through primary consumption of highly toxic insecticides, rodenticides and fungicides usually experience acute effects, which may result in death if sufficient amounts are ingested. There are numerous examples of this, including the squirrels, raccoons and white-tail deer that have died over the years in the state of New York as a consequence of ingesting anticoagulant rodenticide baits [34], or the geese poisoned by ingesting heptachlor and chlordane treated seeds, and the countless songbirds killed in similar circumstances [15]. However, for the majority of pesticide products in the market, chronic and sublethal effects are more common because of the low level of residues (see Chapter 1). Secondary poisoning typically leads to chronic toxicity and unforeseen side-effects, as in the case of eggshell thinning in birds of prey and fish-eating birds contaminated with OC insecticides [35]. Nonetheless, secondary poisoning can be lethal to the predator even at normal rates of application if the chemical is very toxic (e.g. OP and carbamates) [36], or when the contamination is severe due to misuse; for example, the inappropriate spraying of monocrotophos over alfalfa fields to control voles in Israel resulted in the killing of hundreds of kites, eagles, buzzards and owls in a few days because they fed on voles that had been affected by this OP insecticide [37].
# **Exposure of Plants to Pesticides**
Plants are affected by herbicides and fungicides only when these products are deposited directly onto them (contact) or are taken up through the roots. To avoid damaging the crop they intend to help, herbicides are usually applied prior to
### **68** *Ecological Impacts of Toxic Chemicals Francisco Sánchez-Bayo*
planting. Herbicide drift on to non-target areas may affect other crops and wild plants alike, and is a common cause of economic injury to neighbouring farmers, which can reach up to 10% yield losses in the case of canola [38]. For this reason, aerial sprays of 2,4-D on fields of cereal crops must be carefully planned to avoid drift onto nearby sensitive crops like cotton [10]. Granular formulations of herbicides are otherwise preferred. Irrigation waters containing residues of unwanted herbicides and other pesticides may also affect the performance of rotational crops grown on the same fields. However, water-borne residues of herbicides in runoff are more likely to affect aquatic plant communities growing along streams, rivers and marshes since their levels are at most sublethal to animals.
# **Effects from Exposure to Pesticides**
Toxicological effects depend on the doses exposed to, and such effects may occur at individual, population and community levels (see Chapter 1). The focus of this chapter is on the latter two effects, since they define the impacts on the ecosystem more clearly than any sublethal effect manifested on particular individuals. Besides, standard measurements of toxicity (e.g. LD50, EC10, NOEL) are determined with reference to populations. Community effects are typically described by the proportion of species eliminated or severely reduced in numbers within a collective group of species, but there is no standardized measurement to express this kind of impacts.
Dose is the amount taken up by the organism, which can be taken either all at once or through several episodic events. This distinction is important, particularly when dealing with pesticides, as most agrochemical products are recommended to be applied once or twice within the growing season of a crop; orchards usually require several applications. When a pesticide is applied only once, all non-target animals and plants that are directly exposed to it may experience short-term, acute toxic effects. In ecotoxicology this is called pulse exposure to distinguish it from constant exposure to pollutants in a given environment. After an initial shock, the affected organisms will be subject to decreasing exposure as the pesticide disappears progressively by natural decay, microbial degradation, and other dissipation routes (see Chapter 2). However, residues remaining in the plants, soil and water of the agricultural fields and surroundings can be taken up by animals moving into those areas any time after application. For non-persistent and biodegradable pesticides, those residual amounts are sufficiently low to ensure the LD50s for most species are not reached, although there is no guarantee they won't have any impact whatsoever – sublethal effects on some individuals may still take place.
In a different situation, when a pesticide persists in the environment for longer than one season (which occurs whenever half-lives are over 3 months) its residues are expected to build-up between consecutive annual applications. That is the case with most 'old' pesticides like OC insecticides and copper fungicides. In such circumstances, all organisms chronically exposed are at risk of accumulating the toxicant in their tissues, and with time the internal doses may be sufficient to cause either sublethal or lethal effects – the eggshell thinning due to DDE residues in birds is a classical example of this problem [5].
Mortality is the most obvious consequence of direct pesticide toxicity, reducing the populations of both target and non-target species affected. Such reduction in numbers is directly proportional to the toxic potency of the chemicals involved as measured by their LD50s. Since species live in communities rather than in isolation, the decrease in numbers of one species inevitably affects the other species with which it interacts. The resulting imbalance of populations is the most apparent direct effect of pesticides in biological communities. This usually takes place in the agricultural fields and small surrounding areas affected by drift and volatilization, whereas direct effects on aquatic ecosystems may take place beyond these boundaries since water-borne residues can be transported long distances (see Chapter 2). It is important to bear in mind that populations can recover once the toxicant levels drop or disappear, so the direct ecological disturbances caused by pesticides are temporary, not permanent.
It is also important to consider that organisms are not exposed to a single agricultural pesticide alone but rather to a suite of insecticides, herbicides and fungicides that are routinely applied to the crops, sometimes on the same day or even at the same time. Evidence that the combination of several toxic substances produces synergistic effects on the organisms exposed was first reported for mosquito larvae (Aeddes aegypti) and fruit flies (Drosophila melanogaster) exposed to the OP insecticide parathion and the herbicide atrazine [39]; the addition of the herbicide enhanced the lethal effect of parathion by a factor of 2 to 12 depending on the soil type used and other factors. The fungicide propiconazole enhances the activity of neonicotinoids [40], but the best known synergism is the enhancing effect of piperonyl butoxide on pyrethroids and cyano-substituted neonicotinoid insecticides, because the synergist inhibits the P450 enzymatic detoxification mechanism. Estrogenic effects of mixtures of OC insecticides which are innocuous individually, are also examples of synergism that may have profound environmental implications [41]. The synergistic interaction of atrazine has also been proven in combination with some OP insecticides applied to house flies (Musca domestica), and other interactions between different types of pesticides are well documented in aquatic ecosystems (see Chapter 6); however, most of the mixture effects of pesticides are additive rather than synergistic [42].
Finally, sublethal doses of pesticides may cause enough stress in the organisms exposed so as to trigger anomalous behaviour. Examples are the reduced predatory skills in frogs exposed to malathion [43], negligence of female starlings exposed to OP insecticides in looking after their nestlings [44], as well as depressed immunological responses that may result in higher than normal rates of parasitic infection [45].
# **Persistence of Residues and their Bioavailability**
Persistence indicates the ability of a toxicant to remain intact and active over long periods of time. The half-life is a useful measure of persistence: it is the time required for half of the chemical to disappear, usually by transformation into a non-active degraded product (metabolite). However, some metabolites can also be toxic (e.g. endosulfan sulphate, dieldrin, aldicarb sulfoxide and sulfone, heptachlor epoxide), in which case the total persistence of parent compound and metabolites should be considered in assessments of ecological impact.
Apart from a few exceptions, modern pesticides are not as persistent as those used in the past, and this together with specificity of action is a prominent feature of modern agrochemical products. Compared to the arsenates and OC insecticides of old, with half-lives in the environment of several years, most neurotoxic insecticides are easily degraded in the environment by chemical and biological processes. Modern herbicides and fungicides are also more degradable than their early products, even though these chemicals are generally more persistent than insecticides (Table **2**). Currently, over 50% of pesticides have half-lives in soil under a month, with only 14-20% having halflives over three months either in soil or water.
**Table 2:** Persistence of pesticides according to their average half-life in soil, and their proportion among the total number of registered products of the same type. Source [6]
Non-persistent = half-life under 30 days, equivalent to 1% or less residues remaining after half a year
Moderate = half-life between 1 and 3 months, equivalent to 1-5% residues after 1 year
Persistent = half-life over 3 months, equivalent to 5% or more residues after 1 year
Persistent pesticides are more efficacious due simply to their prolonged action over time. From an environmental point of view this is undesirable because the longer the residues stay in the environment, the more chances of being dispersed and the higher risk they pose to organisms as a result of their prolonged exposure and accumulation. Indeed, persistence of agrochemicals poses as much concern as their acute toxicity. A highly toxic and degradable substance may have short-term lethal effects, but it usually allows recovery of populations after its disappearance, whereas a persistent substance of low toxicity will undoubtedly accumulate in the environment and in non-target organisms, in which case sublethal and unknown side-effects are likely to appear in the future. Obviously, when a pesticide is both persistent and very toxic the consequences can be disastrous, as happens with the 'old' OCs, arsenic insecticides and copper fungicides.
Residue accumulation in tissues of both plant and animals occurs whenever the degradation rate of a chemical is lower than its rate of uptake. Since toxic effects are related to the doses exposed, the bioavailability of the pesticide
### **70** *Ecological Impacts of Toxic Chemicals Francisco Sánchez-Bayo*
residues is essential. For example, residues attached to soil particles may remain largely inaccessible to soil organisms, as if the residues were locked [33], and do not cause the effects one would expect. An extreme case is glyphosate: to be effective this herbicide must be absorbed by the plants, either by direct contact on the leaves or by uptake of the chemical in solution through the roots [46]. However, when glyphosate falls on bare ground it is immediately adsorbed onto the clay particles and humic substances in the soil, so it cannot be taken up by the plant roots – it remains effectively inactivated. In contrast, residues of most hydrophobic insecticides (e.g. pyrethroids, OCs and many OPs), systemic and soluble insecticides (e.g. imidacloprid; carbaryl) and herbicides (e.g. diuron) are adsorbed onto organic matter in the soil and remain available to earthworms and other soil microfauna even many years after being applied to the fields.
# **REVIEW OF PESTICIDE IMPACTS ON NON-TARGET COMMUNITIES**
# **Soil Communities**
The soil is a micro-ecosystem in its own right, and the organisms that make it or live in it play a crucial role in recycling nutrients, thus sustaining the soil fertility which allows ecosystem and agricultural productivity. Their diversity and heterogeneity are therefore necessary for long-term ecological resilience of the biosphere.
# *Micro-Organisms and Soil Metabolism*
Fungi, bacteria and protists metabolize decaying plant and animal matter and convert it to either organic waste products (e.g. CO2, methane and others) or minerals (e.g. nitrates, phosphates), which constitute the nutrients of plants. In addition, white-rot fungi have evolved to degrade lignin, an ability that enables them to degrade recalcitrant chlorinated pesticides such as toxaphene, lindane and pentachlorophenol [47].
Pesticides can affect these processes by altering the microbial composition of the soil. For example, applications of the systemic fungicide benomyl over many years reduced mycorrhizal root colonization by 80%, thereby indirectly reducing the abundance of fungal-feeding and predatory nematodes by 33% while increasing microbial substrateinduced respiration by 10% [48]. Generally, fungicides eliminate pathogenic root-rot or dumping-off fungi (e.g. *Pythium, Phythophthora, Rhizoctonia*), thus fostering the growth of competing bacteria while surviving and resistant strains of fungi become dominant (Fig. **1**). Among the latter are the actinomycetes *Aspergillus, Penicillium, Mucor, Pyrenochaeta* and *Trichoderma,* which are less susceptible [29]. Reduction of fungi affects negatively the decomposition of the surface litter by 25-36%, but increases the mineralization in the buried litter carried out mostly by bacteria [49]. This structural change is not always significant in single applications of chlorothalonil (15 g/kg soil) [50], and may be masked by quick recovery and other factors. Suppression of mycorrhizal symbiosis in crop plants treated with fungicides has been observed with normal rates of captan, carbofuran and mercury fungicides, resulting in stunted plant growth and yield reduction [51]. In contrast, typical application rates of some OP insecticides (trichlorfon, chlorpyrifos and quinalphos) may promote rhizosphere fungi temporarily until the suppressed bacterial populations recover in 45-60 days [52]. In soils contaminated with persistent arsenic and copper fungicides the regeneration of fungi is slow and takes many years [53], and this also reduces the ability of the indigenous soil microbial community to degrade DDT [54].
Soil basal respiration is generally reduced 30-50% after treatment with the fungicides benomyl and captan at field rates (51 and 125 mg/kg soil, respectively) [55], or under persistent residues (21-490 mg/kg) of copper fungicides [54], but carbendazim, even at dosages as high as 87.5 kg/ha, does not have significant impacts on soil nutrient cycling processes nor on soil microbial activity [56]. Suppressed basal respiration has also been observed after treatment with the herbicides 2,4-D, picloram and glyphosate, usually at concentrations higher than normally applied, whereas glyphosate applied at 2.2 mg/kg for several years in Brazil increased soil metabolism some 10–15% and fostered fungi while reducing bacterial counts [57]. Repeated application of the herbicides atrazine and metolachlor over 20 years altered the soil community structure in corn fields, in particular by reducing methanotrophic bacteria, but did not cause a decreased community function (methane oxidation) [29].
The mineralization of organic N to ammonium and then nitrate in soil, carried out by the nitrifying bacteria Nitrosomonas and Nitrobacter, is suppressed by the fungicide maneb and the herbicide picloram, but is unaffected by the continuous use of most pesticides either singly or in combination. However, the fungicides metalaxyl, mefenoxam, mancozeb and chlorothalonil, and the herbicide prosulfuron, increase ammonium and nitrate levels by indirectly fostering nitrifying and denitrifying bacteria which inhibit N2O and NO production [29]. Nitrogen fixation in rice paddies by Azospirillum bacteria can increase following application of recommended doses of carbofuran insecticide (2- 5 mg/L), but larger doses are inhibitory [58]. The herbicide glyphosate suppresses most soil bacteria, including nitrogenfixing Rhizobium, because it inhibits the biosynthesis of aromatic amino acids. Susceptibility of plants to pathogens is also increased by glyphosate treatment as biosynthesis of the proteins phytoalexin and glyceollin, which normally block infection, is also inhibited. However, significant impairment is only observed at high concentrations, since glyphosate itself is completely degraded to CO2 by other micro-organisms living in the same soil [59]. A general inhibitory effect of phosphatase (5-98%) in the presence of glyphosate has also been observed [60], whereas the herbicides oxyfluorfen and oxadiazon at 0.4 and 0.12 kg/ha, respectively, stimulate the population and activities of phosphate solubilizing microorganisms and also the availability of phosphorus in the rhizosphere [61].
Little is known about the impact of pesticides on soil protozoans, but it seems that they are just as sensitive as other soil micro-organisms, with insecticides being more toxic than herbicides. Soil protozoa can be critically disturbed as populations often do not fully recover within 60 days. Fungicides have rather varied effects: ciliates decrease slightly but testate amoeba species can be reduced by 50% in pesticide-treated agroecosystems, contrasting with the increased abundances and biomasses of soil protozoa found in ecofarming [62]. Transgenic Bt-crops, which produce the toxic Cry proteins from *Bacillus thuringiensis,* do not have much impact on microbial, nematodes and protozoan communities. Although some effects of Bt-plants on microbial soil communities have been reported, they were mostly the result of differences in geography, temperature, plant variety, and soil type and, in general, were transient and not related to the presence of the Bt-toxins [63].
# *Soil Mesofauna*
The contribution of mesofauna to the recycling of total carbon has been estimated in the range 0.4-11% for surface litter from non-tillage fields and 6-22% in buried litter from pesticide treated fields [49]. Typical applications on crops, particularly of insecticides, can decimate the minute animals that carry out this essential task and disrupt the complex structure of the soil, which they effectively form. However, no matter how drastic their impact may be, all these effects can be reversed once the toxic activity has disappeared because populations of these small organisms recover very quickly [64]. The following is a summary of direct impacts on the most important taxonomic groups of soil fauna as affected by normal application rates used in agriculture, unless specified otherwise.
**Figure 1:** Diagram showing the main impacts of pesticides on soil, plant and arthropod communities. Red arrows indicate decreases and blue arrows indicate increases; empty arrows indicate indirect effects.
# *Arthropods*
Since the early days of pesticide usage it was noted that OC insecticides had mixed effects on the animal communities of the soil [65]. On the one hand, aldrin, dieldrin, heptachlor, chlordane and DDT controlled well the insect pests of the crops, but on the other hand their residues in soil greatly reduced most species of springtails (Collembola), saprophagous mites, symphylids and pauropods (Myriapoda). DDT was less toxic than aldrin and dieldrin but killed higher percentages of predatory mites than other insecticides, the destruction of the latter resulting in indirect increases of their Collembola prey species. All OC insecticides had little or no effect on earthworms, enchytraeid worms and nematodes at low application rates (*i.e.* aldrin 2.5 kg/ha), whereas five times that dose, as applied for the control of *Phyllophaga* larvae, affected several earthworm species. Many of these early reports refer to field observations that are difficult to evaluate, but proper assessments carried out later confirmed those findings [27].
Of special significance are the impacts on populations of mites because these tiny organisms are the most numerous arthropods in soil; many of them are predators, others are saprophytic while some *Tetranychus* are crop pests. Among the 84 studies in a variety of crops reported by Edwards and Thompson [66], 56 showed a decrease in mite densities, nine reported increases and 19 did not show significant changes. Impacts occur across all ecological types of mites, with populations of predatory mites being negatively affected more frequently when treated with OC insecticides (e.g. DDT, endosulfan, aldrin, chordane and heptachlor), most OP insecticides and carbamate biocides (*i.e.* aldicarb, carbofuran) [67]. Although mites recover within six weeks or a few months, a single exposure to aldicarb (25 kg/ha) resulted in different successional outcomes over the subsequent four years because of the elimination of many Gamasina predacious mites, which are often the most susceptible [68]. Even natural extracts like neem (from *Azadirachta indica*) are more detrimental to oribatid mites than other mites and spiders [69]. Fumigants (gaseous pesticides) have devastating effects: the D-D mixture eliminates all mite populations, does not allow their recovery until two years later and eventually decreases the soil biodiversity [66]. Some mites are susceptible to the herbicides simazine, atrazine, monuron and DNOC, but most of the population changes observed in fields treated with herbicides appear to be from indirect effects on the flora [70]. Apart from mites, predatory arthropods of the soil include carabid and staphylinid beetles, earwigs, centipedes and spiders, all of which control many pests and are, therefore, beneficial species to agriculture. Centipede populations were reduced by DDT and aldrin in the past, as subsequently did most OP insecticides and carbamates, but reports on impacts of modern pesticides on this group of animals are very few [71].
Saprophytic arthropods such as springtails (Collembola), Pauropoda, most millipedes (Diplopoda), woodlice (Isopoda), certain mites, symphylids and Diptera lavae help desintegrate plant material that many soil microorganisms are unable to process directly [66]. Although the role of these soil organisms is not as important in agricultural fields as it is in forests and other ecosystems, the current agronomic trend of no-tillage draws its benefits in soil fertility mainly from the role of these animals. For instance, an 80% reduction of springtail numbers after applications of lindane (0.5 kg/ha) to corn crops in Africa resulted in reduced breakdown of organic matter by 45% [72]. Collembola species are not as susceptible to pesticides as mites are; in fact, their numbers usually increase when fields are treated with normal doses of insecticides, as these kill the predatory mites that prey on them [73], thus altering the dominance structure of the springtail community even if the species composition remains unchanged. Springtails are very susceptible to fumigants, carbamates and many OP insecticides [74]. The arsenic herbicides reduced springtail populations in barley by half [75], while DNOC, paraquat, dalapon-sodium and several triazines also reduce their populations when applied in large doses [76], but most herbicides affect springtail communities indirectly [70]. Only a few fungicides (e.g. benomyl) appear to impact negatively on populations of springtails and woodlice [77]. Among the tiny Myriapoda of the soil, the pauropods seem to be most susceptible to all kinds of insecticides, and some populations are completely eliminated by OP insecticides. Symphylids, by contrast, do not suffer drastic effects because they live buried in the deep soil layers where they feed on plant rootlets. Thus, non-leaching, hydrophobic insecticides (most OCs, pyrethroids and some OPs) hardly affect their populations, whilst systemic, hydrophylic insecticides and fumigants deeply penetrate the soil and cause serious population reductions in all taxa [66]. Millipedes are more tolerant, and even if their populations are reduced temporarily by OC and OP insecticides, the herbicide monuron, or the fungicide carbendazim, they recover within a few months [78]. However, persistent residues of DDT in soil of cabbage plots can progressively accumulate in millipedes and reduce their populations over the years [29].
Larvae of many Diptera species are agricultural pests, but the majority of them are not. In any case, they all play an important role in breaking down dead plant/animal matter, so the repeated application of insecticides and herbicides like simazine leads to a significant loss of Diptera larvae and a potential accumulation of dead organic material on the surface [66]. Larvae of dung beetles and flies in pastureland are also affected by residues of parasiticides found in the faeces of treated livestock. For example, emergence of the dung beetle *Liatongus minutus* and eight species of flies from cowpats in the first two weeks following ivermectin treatment at normal rates (0.5 mg/kg body weight) was significantly reduced, while Ceratopogonidae and Psychodidae species prospered [79]. These impacts occur while lethal levels of residues persist in the dung – usually 1-3 weeks for most pyrethroids and avermectins in cowpats [80] but shorter times in sheep dung [81]. By contrast, insect growth regulators like fluazuron and methoprene appear to have no such effects at normal rates of treatment [82, 83].
# *Other Invertebrates*
Parasitic nematodes are regularly controlled with fumigants, lindane, some OP and carbamate insecticides applied directly into the soil, but depending on the doses applied, populations of saprophytic and beneficial nematodes are also reduced [29, 66]. Most OC insecticides and fungicides do not affect nematode numbers. Among the latter chemicals, carbendazim increases omnivorous species and benomyl reduces them [66]. Under field conditions, the risk of indirect effects from fungicide application is usually much greater than that of direct effects. For example, by reducing total fungal biomass and activity, captan decreases the numbers of fungal-feeding nematodes [84]. Herbicides have mixed effects, and this is believed to result from the complex interplay of top-down and bottom-up forces in soil food webs. Another example: plant-root parasitic species increased in rice paddy plots treated with a mixture of thiobencarb and simetryne (2.8 and 0.6 kg/ha, respectively) while predaceous mononchids, which mostly live on the surface, were drastically decimated when chlormethoxyfen at 2.8 kg/ha [85] was added to that mixture.
More important, particularly in tropical agroecosystems, orchards and vegetable patches with litter, are the impacts on detritivorous earthworms, because they remove large amounts of leaves and stubble material, and in doing so increase soil fertility and lessen the ability of certain pathogens to overwinter in the fields [66]. Past applications of copper fungicides and arsenates have led to the formation of mats of undecayed organic matter on the surface of many orchards, because these highly toxic and persistent compounds decimate earthworms populations [86], increase their avoidance behaviour [53], and negatively affect their burrowing rate. The latter sublethal effects have also been observed with the insecticide imidacloprid at 0.5-1.0 mg/kg dry soil [87]. The majority of OC, OP and carbamate insecticides do not cause significant reduction of earthworm populations at normal application rates [66], but chlordane, heptachlor, phorate and carbofuran are extremely toxic to all worms and eliminate them completely [88]. Recovery times from carbofuran treatments can last 90-105 days, and that from the OC insecticide butachlor can be longer than a season. Phorate can also foster enchytraeid worms indirectly by eliminating their predators [89]. All fumigants are deadly to earthworms because they penetrate the deep layers of the soil [66]. Among fungicides, carbendazim at 1 kg/ha decreased the abundance of several *Lumbricus* species in terrestrial model ecosystem (TME) studies, as well as *Fridericia* enchytraeid worms and native earthworms in rubber plantations of the Amazonia [78]. Some herbicides (e.g. DNOC, chlorpropham, atrazine, simazine, monuron) reduce earthworm populations slightly, and paraquat appears to increase them [70], but most have no direct effect on them. In general, conventional agronomic practices in orchards seem to affect negatively detritivores such as earthworms and woodlice. However, some long-term studies have shown that insecticide-treated fields had no ecologically significant impacts in earthworm populations when compared to untreated fields, the differences being largely consistent with the expected effects of climate, soil types, crop types and cultivation practices [90].
# **Vegetation and its Arthropod Communities**
The soil is the substrate and nutrient source for the growth of plants, and the vegetation provides the basic structure on which most species of arthropods live. Both weeds and macro-invertebrates provide many valuable services to the agroecosystem – nitrification; soil aeration and water percolation; recycling of litter, dung and decay materials; pollination; and vectors of mycorrhizal spores, among others.
# *Impacts on Vegetation*
Weeds are the competitors of the crop for water and nutrients, and can reduce crop yields significantly. Broadspectrum herbicides are toxic to all kinds of plants alike, usually by inhibiting the photosynthesis (e.g. urea herbicides, triazines) or any other essential plant metabolic pathway (e.g. glyphosate), but others inhibit seedling development from the seed (e.g. trifluralin and pendimethalin). Selective herbicides are designed to inhibit metabolic processes common to either grasses (monocotyledons) or broad-leaf plants (dicotyledons). This feature allows them to be used on certain crops to control weeds of the opposite type; for example, 2,4-D is used in cereal crops because it only inhibits growth of broad-leaf plants. The effectiveness of herbicides in reducing plant biomass is often underestimated. They effectively exclude many annual plants from being established, and although vegetation communities may recover in the following season, the constant application of herbicides year after year leads to the depletion of soil seed banks. For example, it has been reported that after many years of intensive agricultural practices using a range of herbicides, the Hilly Country of Saxony has lost many landscapes and their associated flora diversity [91]. It appears that the time of their application in relation to plant seed production influences more the nature of vegetation changes than does the soil seed bank type. However, individual herbicides have minimal impacts; a review of the impacts of the broad-spectrum herbicide glyphosate on a variety of ecosystems found the shifts in species floral composition and structure of habitats were within the normal range of variation in natural ecosystems [92].
Indirect impacts of herbicides on soil fauna are often reported. Long-term studies carried out over several years in vegetable crops have revealed that the soil arthropod community structure is positively correlated with the weed community biomass, which varies with the use of specific herbicides and other management practices [93]. For example, the abundance and diversity of rove beetles (staphylinids) is dependent on weed community composition as well as ploughing, with the highest biodiversity being observed on fields with no-tillage and less pesticide use [94], whereas use of paraquat and trifluralin herbicides in tomato plots result in significant reductions in the density of ground beetles. The unintended consequences of such indirect impacts are illustrated by the reduction of weeds in orange groves in Spain: many years of herbicide applications have reduced the abundance and biodiversity of consumer ants to the point that fewer ant colonies made the soil progressively less porous and more compacted, thus enhancing rainfall erosion and slowly depleting the orchard's soil fertility [95].
Plant biodiversity is not considered to be important in crop monocultures, but it is relevant to the establishment of stable arthropod communities in or around the crop. These play an essential role in effective crop protection and also sustain populations of birds and other vertebrates. In many cases, the losses in yield caused by weed competition can be offset by the benefits that predatory arthropods bring to the crop. For example, cane and sugar yields averaged 19% higher in weedy sugarcane plantations than in the weed-free plantations in Lousiana, because broadleaf weeds enhanced the populations of beneficial carabids, ants and spiders that control the sugarcane borer (*Diatraea saccharalis*) [96]. Similarly, the combined use of Bt-cotton, lucerne strips and a nuclear polyhedrosis virus in Australian cotton farms reduced the use of synthetic pyrethroid insecticides by 50% without sacrificing yield and profitability [97]. Experience over the years in these and other crops have demonstrated the benefits of the appropriately named integrated pest management (IPM) strategies that promote the conservation of existing natural biological controls through major reductions in insecticide and herbicide use.
The introduction of recent transgenic herbicide-tolerant crops (TGHT) may encourage no-tillage practices which are beneficial for soil fertility, but there is concern that such crops may lead to a more intensive use of herbicides and the removal of many weeds that support populations of pollinators [98]. Pollination by bees is a very important ecological service provided to agriculture, as 25% of tropical crops and possibly up to 84% of temperate crops [99] depend on insect pollination. Thus, management and protection of pollinator populations and habitats of nectarproducing plants can be essential for some crops, and for plant biodiversity in the environment at large. However, there are no clear examples of low crop yields resulting from the effect of pesticides or transgenic Bt-plants on pollinators [98]. Although agricultural intensification and habitat loss are the most frequent cause of pollinator impoverishment (64% of cases), direct bee mortality by insecticides is evident and cannot be ignored either [100].
# *Arthropod Communities*
Insecticide sprays can wipe out 99% of the population of target pests as well as those of non-target species, just as chemotherapy kills both bad and good cells alike. Since the early years of the Green Revolution entomologists realized the limitations of this approach and looked for alternative methods of pest control. In nature, predatory arthropods keep the populations of phytophagous insects (most pests) in check: ladybird beetles, dragonflies, earwigs, some ants and crab spiders predate on eggs of pest species, while parasitic Hymenoptera play an essential part in controlling numerous pest larvae, so they are being used in biological pest control. A recent review of 39 ecosystems found that agrochemical pollutants negatively affect these parasitoids in 46% of cases [101], with persistent and systemic insecticides (e.g. cartap and imidacloprid) having the greatest impacts [102]. However, predatory arthropods are less susceptible than parasitoids and more variable in response to pesticides [103]. Although some predatory species are very tolerant to pesticides (e.g. the spider *Lycosa pseudoannulata*, the coccinellid *Cryptolaemus montrouzieri,* and the lacewing *Chrysopa carnea*), their initial elimination by insecticides and their slower recovery than that of the pest species they control often results in rebounds of pests (Fig. **1**) in the short and long term [104].
Early insecticide impacts in non-target arthropod communities were reported for orchards sprayed during three years with lead arsenate and nicotine. Ground-dwelling beetles, spiders and ants were reduced by 15%, and the proportion of eggs and larvae of the main apple pest – the coddling moth (*Laspeyresia pomonella*), which is parasitized by Hymenoptera species – decreased by 64-97%, allowing the moths to come back unopposed [29]. DDT sprays helped eliminate the coddling moth, but it created new pests among leaf-rollers, woolly aphids, red-spiders and *Tetranychus* mites that surged as a consequence of the lack of predators and the suppression of parasitism. Citrus orchards sprayed with DDT to control cottony-cushion scales and mealybug pests also eliminated the predatory ladybird beetles and parasites which control them – as a result, pest numbers not only did not decrease but rather surged exponentially [105]. Because of the persistence of DDT, restoration of a normal predator-prey relationship after cessation of sprays could take up to five years [106].
The annihilation of predatory and parasite arthropods in cotton, corn, rice and horticultural crops has created new community structures characterized by the absence of predator-prey relationships, one where pests species thrived for a while until the next insecticide spray decimates them, where resurgence became the norm and resistance to chemicals the final outcome [29, 107]. In America and Australia, early sprays of calcium arsenate to control the main cotton pest, the boll weevil (*Heliothis* spp.), boosted the populations of cotton aphids due to the elimination of predatory arthropods. Lindane was applied to control both the weevils and the aphids, but this resulted in outbreaks of *Tetranychus* mites, as more predators were also affected. To top it all, the application of OP and systemic carbamate insecticides to control leafworms (*Spodoptera* spp.) resulted in further outbreaks of boll weevils and mites due to a combination of two factors: total lack of predators and insecticide resistance developed within the pest species [29]. It is easy to understand that restoring these shattered communities usually takes a few years. Pest management plans in cotton agroecosystems continued to rely on the routinely, heavy use of pyrethroids, OPs, carbamates and new insecticides until the 1990s [108]. Recently, the introduction of transgenic Bt-cotton in some countries appears to have a positive effect on restoring the biodiversity of most predatory insects, spiders and birds in cotton fields, since insecticide applications are reduced 50% or more [109]. Similarly, the biodiversity of arthropods in Bt-corn crops is much higher than in fields treated with pyrethroids. Insecticide sprays on rice crops upset natural enemy control of pests such as plant hoppers (*Nilaparvata lugens*) and also create heavy selection pressure for strains of pests that can overcome previously resistant rice cultivars. Such circumstances create outbreaks of secondary pests and impair biological control of some key primary pests such as Pyralidae stem borers [104]. Typical applications of BHC and parathion significantly decreased densities of predatory dragonflies, spiders and parasitoids, thus increasing the herbivore:predator ratio among arthropods [110]. The insecticides imidacloprid and fipronil also change this ratio even if their main impact is on midge larvae (Chironomidae) [111]. In addition, herbicides applied to rice paddies foster the numbers of parasitic nematodes and alter the plankton communities [85]. Perhaps, the rich biodiversity of rice fields, with some 200 species of predatory arthropods, could be used in IPM programs to control the 55 species of pests found in this crop [112].
Ground-dwelling carabid beetles are essential in controlling many horticultural pests, and together with staphylinid beetles make up about 75% of the predaceous and/or parasitic insects on vegetable crops [113]. In the past, OC insecticides decimated their populations and allowed very slow recovery afterwards [29], whereas the OC endosulfan at 1 kg/ha appears not to cause major impacts on these arthropods [72]. The impact of cholinesterase inhibiting insecticides on carabid populations ranges widely among species [114, 115], but all allow their recovery within a few weeks [66], whereas pyrethroids and imidacloprid at recommended rates have minimal impacts in spite of their extreme toxicity to insects [72]. Most herbicides indirectly increase densities of carabids, ladybird beetles and linyphiid spiders [74], but 2,4-D and chlorpropham are toxic to carabids too [29]. TGHT sugar beet and Btcanola crops do not appear to have any significant effect on carabids, staphylinids nor spiders, but rather reduce the overall arthropod abundance through indirect effects on weed biomass [116].
Spiders and phytoseiid mites are important predators in all kinds of crops. Applications of OP, carbamate and pyrethroid insecticides in vineyards, orchards and other crops usually result in increases of pest *Tetranychus* mites because of reductions in the more susceptible phytoseiid predators [117]. In experimental plots, spiders were three times less abundant in apple orchards treated with insecticides than in untreated ones, and spiders and ants were reduced in numbers in 53% of the corn crops in Africa treated with lindane (0.5 ka/ha), an effect that lasted 2-3 weeks [72]. Lycosidae and linyphiid spider populations undergo a similar pattern – they are initially eliminated from cereal fields treated with OP insecticides, but their abundance may increase subsequently in response to rebound densities of unaffected prey like springtails [29]. Indirect effects of herbicide application on field margins often reduces the habitat for lycosid and linyphiid spiders, as border crop fields and hedges act as refuges for these and many other beneficial predatory invertebrates [118]. No-tillage practices and TGHT crops enhance spider populations through a more heterogeneous and diverse vegetation structure [119].
The direct impact of insecticides on honey bees (*Apis mellifera*) was recognized a problem since the calcium arsenate dust sprays killed entire hive colonies in the past [29]. They also affect the performance of the colonies, with impacts ranging from odour discrimination to the loss of foraging bees due to disruption of their homing behaviour [120]. Pyrethroid and OP insecticides such as triazophos and dimethoate continue to be very toxic and hazardous to bees [121]. Spray drift and volatilization are responsible for most of the incidents reported on hives [100], while impacts on wild bumblebees (*Bombus* spp.) are likely underestimated and non-reported. All bees are also affected by the poisoned nectar and pollen taken from plants treated with systemic insecticides such as carbamates and imidacloprid. Typical concentrations of imidacloprid of 6 mg/kg in male flowers (panicles) and 2 mg/kg in pollen from maize, sunflower and rape plants are sufficient to decimate honeybee colonies [23], especially when the pollen contains higher residues of other pesticides that could act simultaneously or synergistically. Besides mortality, imidacloprid appears to affect the brain (memory) and metabolism in bees, with the resulting impairment in the workers activity [122].
Crop diversification in conventional farming can help increase the biodiversity of arthropods while significantly reducing the densities of phytophagous pests by 60-70% [123]. In tropical rice crops particularly, which sustain a large biodiversity [112], pest management is best achieved using natural controls rarely supplemented by insecticides [104]. In ephemeral annual crops such as cereals, sugar cane, alfalfa or even cotton, leaving strips of grass and weeds on field margins, woody borders and other practices that attract and provide refuge to many arthropods can increase both biodiversity and abundance of natural predators [118]. It should be borne in mind, however, that any efficiency in controlling the pest populations through natural enemies depends very much on the identity of both predator and prey species, not on the diversity of predators *per se* [124].
# **Vertebrates**
Since invertebrates are small and not very mobile – except some insects –, pesticide impacts on their communities are restricted to the agricultural fields, orchards and the margins affected by spray drift. By contrast, vertebrates move around fields, nearby forests, wetlands, rivers, lakes and even far away places in the case of many bird species. Therefore, off-farm contamination is another source of exposure for vertebrates, even though it is much lower than on-farm exposure due to its lower residue levels [125]. For persistent chemicals, the possibility of bioaccumulation in the animal tissues introduces also a new and often unknown risk factor.
# *Direct Impacts*
The susceptibility of vertebrates to agricultural doses of pesticides is typically lower than that of invertebrates simply because of their size difference. Vertebrates are more tolerant to synthetic pyrethroids, neonicotinoids and OC insecticides, but very susceptible to cholinesterase inhibitors, whereas amphibians are generally very sensitive to pyrethroids and more tolerant of cholinesterase inhibitors than birds and mammals [126]. Reptiles appear to have either less or similar sensitivities to mammals in regard to neurotoxic compounds [127]. Mammals are more tolerant to certain pesticide groups than other vertebrates because they posses active detoxification mechanisms. However, small insectivorous mammals, such as shrews and moles, are very sensitive to neurotoxic anti-cholinesterase insecticides because of their high feeding and metabolic rates. Birds are more tolerant of pyrethroid and neonicotinoid insecticides, but are very susceptible to chlorfenapyr.
Killing of non-target organisms such as birds, lizards and small mammals is often observed at the time of insecticide applications [128], but most incidents are probably not reported [129]. Bird mortalities from direct exposure to insecticides can range from a few birds to several hundreds. Indeed, OC insecticides were blamed for many bird fatalities in the past, and cholinesterase inhibitor insecticides were responsible for 25-50% of bird mortality observed in farmland of the United Kingdom between 1975-1990s, of 17% of all birds poisoned in agricultural lands of the Netherlands, and 3-12% of all birds of prey found poisoned in the USA [15]. Levels of inhibition of brain acetyl-cholinesterase in birds below 20% are associated with sublethal effects and levels above 70% result in death [130], whereas in lizards the levels are typically below 40% and above 50% for the respective effects [131]. In contrast to OC insecticides, carbamate and OP insecticides do not accumulate in vertebrates as they can be readily metabolized, so their potent effects are usually short-lived. Even so, mortality of magpies by direct poisoning with famphur, applied to cattle as parasitic treatment, has been reported [132]. Lizards suffer similar effects as birds and mammals when exposed to the latter insecticides, but impacts on their populations and ecology are unknown [127]. Frogs, toads and tadpoles are common inhabitants of rice paddies, irrigation ditches and farm ponds as well as in surrounding wetlands and riverbanks, and so are exposed to direct pesticide applications on farm and drift sprays into their habitats. Although pesticide concentrations in agricultural waters are insufficient to cause frog mortality, the development of tadpoles is usually affected by low concentrations of many OPs in water (e.g. 4-8 mg/L fenitrothion) and herbicides like triclopyr (2.4-4.8 mg/L) [133].
Apart from mortality, sublethal effects on birds and small mammals exposed to these insecticides are more common, including reductions in food consumption and drinking activity that leads to noticeable weight losses [44], lack of aggressive behaviour, memory impairment that can compromise their survival ability, immobility on the ground which puts them at risk of predation [134], apathy in bird hatching, nest defence and care for the nestlings [44] and reduced fertility [135]. In amphibians, stress [43], suppression of immunity, and susceptibility to parasite infections [45] have been reported. Most of these effects are transient, but those affecting reproduction impact on the long-term viability of a species, even if there might not be apparent short-term population reductions. For example, direct exposure to OC insecticides reduced the breeding success of songbirds in apple orchards [136] and the recovery of vole populations in experimental plots. This is of concern because wildlife species rely on tight net reproductive rates to maintain their populations and cannot cope with such adverse effects. Thus, it has been suggested that reduced egg weight and hatchling success in caimans as a result of typical exposures to atrazine (15 g/egg) and endosulfan (0.15-1.5 mg/egg) may influence the populations of this species in the wild Amazon [137]. As the field assessment of such populations is difficult, models have been developed to predict the long-term effects caused by reproduction impairment. Balanced population densities are important in the case of rodents, where a delay in reproduction can give a competitive advantage to another species. In this regard, exposure to the OP azinphosmethyl applied on alfalfa at 0.9-3.6 kg/ha caused lower than normal pregnancy rates, or its delay, in both voles and mice [138], whereas similar rates on tall grasses did not have effect on populations of *Microtus canicaudus* voles [139]. Similarly, lack of aggressiveness after exposure to dimethoate (0.4-0.6 kg/ha) did not impede the populations of herbivorous prairie voles (*Microtus ochrogaster)* to increase five-fold because the survival of competitor, omnivorous deer mice (*Peromyscus maniculatus*) decreased significantly [140].
Endocrine disruption is another sublethal effect by which some pesticides and other contaminants may impair developmental growth and reproduction in vertebrates [141]. Altered thyroid hormone concentrations, which influence development and metamorphosis, have been observed in birds exposed to DDT, OP, carbamate and pyrethroid insecticides [142, 143], in goldfinches exposed to the herbicide linuron [144], in amphibians and fish exposed to endosulfan and other insecticides [145]. Abnormal sexual differentiation caused by herbicides like atrazine have been observed in frogs, although conflicting evidence also exists [146]. Confirmed cases of impaired reproduction refer to populations of bald eagles (*Haliaetus leucocephalus*) in the Great Lakes of North America [147] and alligators in Florida [148], both of them after many decades of exposure to DDE residues. In the second case, high residues of OC insecticides and other chemicals were found in alligator eggs from Lake Apopka, which was heavily contaminated by a spill of difocol and DDT in the nearby agricultural area, and though hatching success was lower than normal it appeared to be unrelated to the pesticide levels measured in eggs [149]. Subsequent studies found the levels of estrogen in female alligators from that lake were double than normal, while levels of testosterone in male alligators were three times lower than normal or similar to those found in females. In addition, males had poorly organized testes and abnormally small phalli and females exhibited abnormal ovarian morphology [148]. As a consequence, alligator populations in Lake Apopka are in decline.
# *Primary Poisoning*
More common among vertebrates is the exposure to pesticide residues through ingestion of contaminated food. Granivorous birds and rodents often ingest large quantities of seeds that often contain pesticide residues; grazing mammals may consume pasture contaminated with herbicides or insecticide spray drift; and birds of prey and scavengers often consume the guts of their prey and/or carcass, so the undigested granules of cholinesterase inhibitors and rodenticides found in the prey can result in fatalities among raptors [150]. The extent of this contamination can be assessed by the relative amount of residues found in animal tissues. Based on the residue levels of mirex across a large number of non-target animals [151], we know that insects accumulate more residues than other invertebrates, and among the vertebrates amphibians and reptiles had lower levels than birds and mammals, which possibly reflect their differences in feeding rate and metabolism. While most residues are metabolized and/or excreted by the animals, persistent and recalcitrant chemicals may accumulate in organs such as the liver and kidney, whereas lipophilic residues usually are stored in fatty tissues. Modern biomarker techniques make it feasible to investigate the poisoning level of live animals in a non-destructive way, *i.e.* using small samples of blood serum from reptiles, birds and mammals [152, 153].
# *Secondary Poisoning*
Insectivorous birds, frogs, lizards and mammals often consume insects contaminated with pesticides [32]. The ecological consequences of secondary poisoning differ markedly among vertebrate taxa and the role each species plays in the trophic structure of the ecosystem, and obviously depend on the chemical nature of the poison. Build-up of insecticide residues in primary consumers can make them more susceptible to predators and scavengers. Birds of prey feeding on these animals accumulate even higher residue levels and often die as a result [36]. Most of the fatalities in raptors due to secondary poisoning are associated with the illegal use of insecticides and rodenticides (e.g. to eliminate wild carnivores), but some result from the normal use of pesticides by farmers [154]. Indeed, secondary poisoning by non-persistent carbamate and OP insecticides has been attributed as the cause of mortality in barn owls (*Tyto alba*), American kestrels (*Falco sparverius*), red-tailed hawks (*Buteo jamaicensis*), great horned owls (*Bubo virginianus*) and bald eagles [36, 150], and it is probably more common than we think because most of the time the victims die without being noticed. The removal of vertebrate predators from an ecosystem leads to similar imbalances as described above for the insect communities in (Fig. 1), encouraging pest rodent species to multiply unrestrained.
Persistent OC residues bioaccumulate in the fatty tissues of all organisms, and are released slowly during periods of fasting or intense flying activity such as during migration [155]. As they are passed on from consumers to predators at the top of the trophic chain, the biomagnification factors can be staggering – up to 10,000 times or more [156]. Not surprisingly, consumption of invertebrates contaminated with OC insecticides causes the death of many insectivorous birds and bats [157], but the sublethal effects from this poisoning are more damaging in the long term. One of the first known impacts of OC insecticides was the reproduction impairment they caused in birds of prey and fish-eating birds, which was felt worldwide in less than two decades, and put some species on the brink of extinction [4]. The case is well documented for DDT, though cyclodiene insecticides like dieldrin produced similar effects [158]. Persistent residues in soil, plant forage, seeds, earthworms and other invertebrates accumulate up the trophic ladder because vertebrates consuming such contaminated foods cannot excrete them. Consequently, predators and scavenger birds concentrate large amounts of DDT in their bodies, where it is transformed into DDE, an equally recalcitrant compound which causes eggshell thinning by altering the calcium metabolism in birds [5]. This unforeseeable effect produced a high mortality of embryos and chicks in birds of prey such as the peregrine falcon (*Falco peregrinus*), sparrowhawk (*Accipiter nisus*), kestrels (*Falco* spp.), Spanish imperial eagles (*Aquila adalberti*) [159] and many fish-eating birds like herons, cormorants and pelicans [160]. Initially, the reduction in juveniles was compensated by higher reproduction rates because there was less competition for food, until the introduction of cyclodienes years later dealt a fatal blow and populations of raptors started to decline [161]. DDT and many other OC insecticides were banned in most countries during the 1970-80s, but their residues are still out there. Wildlife feeding in areas where DDT was applied for agricultural pest control continues to be affected by the persistent residues [32], which fortunately are now reduced to the point that raptor populations are no longer threatened with extinction and, on the contrary, are slowly recovering [162, 163].
Rodenticides are one of the most common causes of secondary poisoning in bird and mammal predators that feed on the target rodents. In particular the second generation of anticoagulant coumarin rodenticides are very persistent, and residues ingested with the carcasses of poisoned animals accumulate in the predators' bodies, causing internal or external bleeding and eventually death. Some 70% of the owls collected in Canada between 1988-2003 had residues of at least one rodenticide at levels up to 0.93 mg/kg (brodifacoum) or 1.01 mg/kg (bromadiolone) in their liver [164]. Birds of prey are being increasingly reported dead as a consequence of coumarin poisoning in America [34].
# *Indirect Effects*
Insecticides directly affect insectivorous vertebrates by reducing the insect prey base available to them, whereas herbicides indirectly affect their populations through a variety of pathways, including 1) the direct removal of the food base of granivorous species, 2) reduction in invertebrate abundance by removing the plants that invertebrates depend on for food or habitat, and 3) reduction in vegetative cover necessary for nesting/breeding and reproduction [165].
The best documented evidence of indirect pesticide effects on insects and bird populations is found in the United Kingdom, where declines of grey partridge (*Perdix perdix)* had been noticed by game hunters and ornithologists for some time – it was rightly attributed to the combined indirect effect of herbicides and insecticides that resulted in breeding failure as a consequence of chick starvation and low survival [90, 166]. Even if other contributing factors such as worm parasites have added to the partridge demise [167], the fact that pesticides are routinely sprayed on cereal and other crops everywhere has indirectly affected the populations of many other bird species as well, which are declining in European countries and North America [168]. Declining bird species (e.g. skylark, corn bunting, *etc.*) are not associated with particular foods, but with overall reductions in abundance and diversity of plants, seeds and insects [169, 170] resulting from intensive agriculture [171]. Granivorous species feed on cereal grain and seeds of many 'weeds' like knotgrasses (Polygonaceae), chickweeds (*Stellaria* spp.), goosefoots (*Chenopodium* spp.), and others, so their decline has been driven primarily by herbicide use and the switch from spring-sown to autumn-sown cereals [172], both of which have massively reduced the food supplies of these birds [173]. However, herbicides are not the only culprits, as other intensive management practices (including TGHT crops) also reduce farmland food and biodiversity. During the breeding season, grasshoppers, sawflies, spiders, leaf-beetles, weevils, butterflies/moths and their larvae, aphids, and crane-flies and their larvae are important foods for insectivorous and omnivorous birds; the first four taxa (which are sensitive to insecticides) are associated with the diet of most declining bird species [174]. Recovery of plant and insect densities can be achieved in a few years once the intensive management practices are abandoned [174], offering hope for the recovery of birds as well. Hedgerows with bushes and trees may also provide protection and nesting places for birds, but first the food supply needs to be restored to levels capable of sustaining their populations. Thus, bird densities and biodiversity can double in corn organic farms compared to conventional corn farms [175], despite some organic crops providing only slightly better food supplies.
It is reasonable to assume a similar fate in small insectivorous mammal and reptile populations, but at present evidence from field studies on these animal taxa is lacking. The fact that many amphibian population declines occur in intensive agricultural areas [176] has alerted some researchers. It appears that a combination of indirect effects from insecticides and herbicides, which introduce a cascade of events affecting negatively the feeding and growth of tadpoles, plus sublethal effects involving trematode infection [45] and other intensive farming practices, such as the use of fertilizers, may account for such declines [146]. However, pesticide-treated rice paddies continue to be a valuable haven for many species of frogs, since herons are not interested in preying in conventional fields because they have less foraging value than organic ones [111].
Apart from farmlands, indirect herbicide impacts are observed in wetlands that receive the outflows of agricultural waters, which often contain residual concentrations of atrazine, diuron and other persistent herbicides. For example, the constant use of herbicides for intensive rice production is thought to have contributed to the elimination of macrophyte vegetation in the lagoons of Ebro delta (Spain) during the 1980s, consequently reducing the populations of diving ducks and coot (*Fulica atra*) that depend on vegetative cover for nesting and feeding [177]. In a controlled experiment, density reductions of cattails (*Typha* spp.) after glyphosate sprays (5.8 L/ha) were well correlated with parallel reductions in the abundance of insectivorous and granivorous birds that depend on those plants for nesting [178]. Many wetland plants can take up and metabolize certain herbicide and insecticide residues found in waters (see Chapter 11), but they are still susceptible to the harmful effects of others. A recent study indicates that even if concentrations of individual herbicides may have a low risk to macrophytes, mixtures of bromacil, diuron, and norflurazon have a high risk [179]. At present, more field data are needed to assess the extent to which submerged and emergent (cattails, reeds, rushes and sedges) macrophytes in wetlands are exposed to harmful concentrations of herbicide from aerial spraying, drift from ground application, runoff or soil erosion.
# **CONCLUSION**
Although evidence indicates that 'conventional' chemically-based agriculture renders higher yields per area than 'organic' traditional practices, this has come at a price – high costs due to chemicals and fuel inputs to produce them [180], and multiple environmental impacts which in the long term can be detrimental [10]. Indeed the 'chemotherapy' applied to agriculture has had many side-effects and one wonders if it can go on forever without destroying the fabric of the biosphere. Here I have focused only on the problems, but an overall assessment must consider the benefits pesticides provide to humanity and the negative environmental consequences of not using them. The latter actions would reduce crop yields and lead to further deforestation in developing countries just to produce enough food to feed us all [181]. In this dilemma, the search for alternative agricultural practices that reduce the ecological risks of pesticides is an urgent necessity [182, 183]. The use of pheromone traps is, for example, a very effective alternative to control most insect pests, one that does not impact on non-target organisms and cannot induce resistance [184, 185].
This review has shown that impacts of pesticides on soil fertility are almost neutral, although the long-term crop sustainability is questionable [7]. Truly, fungicides protect the crops against certain pathogens but may destroy the beneficial mycorrhizal symbioses that increase nutrient uptake by the plants. Copper fungicides and certain insecticides are detrimental to earthworms and reduce the recycling capacity of the soil; in the end, soil fertility decreases and yields drop slightly.
Impacts on the prevalence of weeds and pests are mixed and negative in many cases. On the one hand, herbicides increase crop yields, but on the other hand they indirectly reduce the biodiversity and abundance of beneficial arthropods that carry out pollination and keep most pest species at bay. Insecticides are then applied to decimate the pests arising naturally under these circumstances, but eliminate the predators and parasitoids; this causes serious destabilizing effects on invertebrate communities which result in the rebound, promotion and increased resistance of all pests. After a few years, such futile efforts to contain the pest populations reach an unbearable cost, which could be avoided if integrated management practices that rely on natural means of weed and pest control were put in place [184, 186]. On the positive side, these effects are short-lived for the majority of the agrochemical products currently in use, so the ecosystem can recover within a year or two following cessation of pesticide application.
Finally, the impacts on terrestrial wildlife vertebrates are clearly negative – the death toll that certain insecticides have annually on non-target bird and small terrestrial vertebrates cannot be overlooked, even if such mortality may not reduce their populations in the long term due to compensatory effects [187]. More serious is the indirect impacts of routinely applied herbicides that cause declining population densities and biodiversity of birds and possibly amphibians. Equally, the secondary poisoning of consumer and predatory birds, reptiles and mammals by ingestion of pesticide-contaminated food is a real and present worry affecting individuals in various ways; unfortunately, long-term impacts on their populations usually take years to be noticed. Significant changes in current policies, institutions and practices are necessary to reconcile biodiversity conservation and food security [183]. The contribution of DDT and other persistent OC insecticides to the local extinction of birds of prey is undeniable, and also a reminder that persistent toxic chemicals should have no place in this world. Indeed, the contamination of the planet's ecosystems with these and other persistent pesticides is an ecological tragedy that will take many decades to be cleaned up.
# **REFERENCES**
### **82** *Ecological Impacts of Toxic Chemicals Francisco Sánchez-Bayo*
### **84** *Ecological Impacts of Toxic Chemicals Francisco Sánchez-Bayo*
© 2011 The Author(s). Published by Bentham Science Publisher. This is an open access chapter published under CC BY 4.0 https://creativecommons.org/licenses/by/4.0/legalcode
# **CHAPTER 5**
# **Ecological Impacts of Major Forest-Use Pesticides**
# **Dean G. Thompson\***
*Canadian Forest Service, Sault Ste Marie, Ontario, Canada*
**Abstract**: Assessing the potential for ecological impacts of pesticides requires a hierarchical approach with research ranging from simple laboratory to complex field experiments and operational monitoring. While all levels of study provide useful information, higher tier research has inherently greater environmental relevance and inference potential. In this chapter, selected higher tier studies relating to the use of herbicides glyphosate and triclopyr, as well as the insecticides *Bacillus thuringiensis* var. *kurstaki* (Btk) and diflubenzuron in the forest sector are reviewed. These case examples illustrate scenarios in which higher tier studies either negate or support the presumptions of risk derived from results of lower tier experiments. Specifically, assessment of the cases for glyphosate and Btk support their continued judicious use as environmentally acceptable components of integrated vegetation and insect pest management strategies. In contrast, higher level studies confirm risk postulates associated with typical forest-sector use patterns for triclopyr ester and diflubenzuron. Mitigation measures are required to ensure that use of these latter compounds do not pose undue risk to sensitive non-target organisms. In a broader context, the ecological implications of pesticide use in the forest sector must be considered in light of the fact that any management action, including the "no intervention" option, carries both economic and ecological risk. Strict adherence to the weight of scientific evidence principle, incorporation of knowledge gained from all levels of investigation, and a balanced assessment of relative risks of all potential options are considered primary requisites of comprehensive risk analysis and effective decision making.
# **INTRODUCTION**
Truhaut [1] described ecotoxicology as that branch of the discipline concerned with the toxic effects of natural or synthetic pollutants on the constituents of ecosystems. As noted by Butler [2], this concept carries the inherent requirement to consider how the toxicant is released, its potential transformation and its possible transport to other compartments, since these are the primary determinants of exposure and effect. Potential effects must be considered at multiple scales, including those of biological organization (organism, population, or community), space (local to landscape) and time (days to years). These concepts are particularly relevant to the assessment of ecological impacts of pesticides in the forest sector where they may be applied to assist in regeneration or protection of forest stands and where there is potential exposure of a diverse array of organisms within highly interconnected ecosystem compartments.
Ecotoxicological risks associated with modern forest-use pesticides are quite unlike those of historic compounds such as DDT. However, the potential for both direct and indirect effects exists, and such risks are often the dominant element of public concern and policies associated with this forest management practice, as well as with forest certification schemes. Forest pesticide use varies across the globe, principally in relation to the size and accessibility of the resource, primary crop species and the value of commodities derived there from. In some countries blessed with huge areas of natural forests (e.g. Canada, Russia, USA), pesticides are applied to only a very small proportion of the forest land base that is managed for commercial production of high value products such as sawn wood, panels or pulp and paper. In other countries (e.g. New Zealand, Australia, Finland, Sweden and south-eastern USA) relatively more intensive "plantation" management may be employed for the same general purpose and of course gradients of relative management intensity occur in most countries.
While the focus of this chapter is on ecotoxicological risks of forest-use pesticides, such risks must be considered within the broader context of assessing both risk and benefit of this or alternative forest management actions. Few, if any forest managers would choose to apply pesticides if there were not substantial benefits associated with such treatments. For example, herbicides are recognized as the most effective tool for controlling competing vegetation to favour partitioning of essential light, water, nutrients and growing space to the desired crop species rather than to weedy competitors [3]. Wagner *et al.* [4] recently reviewed results from 60 of the longest-term studies in Canada, the USA, South Africa, Brazil, New Zealand and Australia, documenting that the majority of studies show 30 to
**<sup>\*</sup>Address correspondence to Dean G. Thompson:** Canadian Forest Service, Natural Resources Canada, Sault Ste. Marie, Ontario, Canada P6A 2E5; Email: [email protected]
500% increases in wood volume as well as reduced rotation periods from effective vegetation control treatments. Positive outcomes are reflected in significantly enhanced regeneration success and overall sustainable management of forest resources. A diverse array of insect pest species are capable of causing significant economic or ecological damage in major plantations or natural forest stands [5, 6]. Both chemical and biological insecticides are applied to protect semi-mature or mature high value forest stands or to slow the spread of invasive species across the landscape and thus mitigate either economic or ecological losses. In cases where no effective chemical or biological controls are applied, devastating losses are typically the result. For example in Canada, no effective pesticides have been developed or applied to control the epidemic outbreak of mountain pine beetle in lodgepole pine stands. This single insect pest now affects a forest area in excess of 14.5 million ha in the province of British Columbia [7] an area essentially equivalent to that of England. As the beetle moves across the Rocky Mountain divide into Alberta it threatens stands of other pine species including jack pine that spans the boreal forest region across country with massive implications in terms of economic loss, carbon release to the environment and unknown ecological effects in a region not previously adapted to this pest species.
Pesticide risk should also be considered in relation to specific use patterns and proportional use. In comparison to agriculture, pesticide use in forestry involves substantially fewer active ingredients as well as dramatically lower use frequency and proportion of the total productive land area treated in any given year. For example, pesticide use in Canadian forestry accounts for only ~2% of total pest control products sold in that country. Only two active ingredients, the herbicide glyphosate and the microbial insecticide *Bacillus thuringiensis* var. *kurstaki* (Btk) have any significant degree of use, each comprising more than 90% of the total forest area treated with a herbicide or insecticide respectively, a determination based on 2007 statistics for pest control product use in that sector [8]. Similarly, pesticide use in plantation forestry in Australia accounts for only 0.7% of the total annual national expenditures on pesticides [9]. The latter report presents detailed analysis of pesticide expenditures in agricultural crops as compared to forestry. Results emphasize the dramatically higher use frequency and hence expenditures associated with pesticide use in agricultural crop production. To a large degree, this reflects the common practice of multiple pesticide applications on an annual basis to much of the agriculture land base. In contrast, individual forest stands rarely, if ever receive annual pesticide treatments and frequency of use is typically quite low. Even under intensive forest management regimes, the total number of pesticide applications during a rotation period is unlikely to exceed four; that is two herbicide treatments in the early regeneration phase and two insecticide treatments when trees are semi-mature to mature. However, rotation periods vary markedly with forest crop species ranging from as little as 8 to 10 years for example in short rotation eucalypt plantations of Australia, to 80 years or more for spruce stands in the boreal forests of Canada. The total proportion of the productive forest land base treated is also an important consideration in ecotoxicological risk assessments. Again, on a comparative basis, agricultural food crop production often involves essentially 100% of the land base receiving at least one pesticide treatment each year, whereas production of fibre typically involves pesticide application to only a very small proportion of the commercial forest land base annually. However, exceptional cases have been documented historically, including for example massive spruce budworm outbreak in New Brunswick, Canada where almost 4 million ha of forest land was treated with insecticides in one year [5]. While these statistics vary with jurisdiction, year and pesticide type, the point is well exemplified by herbicide use in Canadian forestry where <1% of the commercial forest land base is treated in any given year [10].
Considered in combination, and particularly in relation to agricultural pesticide use, the few active ingredients employed in forest management, their relatively low use frequency, the minor proportion of total forest land area treated and the resultant lower environmental loadings (*i.e.* mass of total pesticide applied per unit area), public concern over pesticide use in forestry seems disproportionately high. For example a poll of 2500 Canadians indicated 71% opposed the use of chemicals in the forest [11]. As noted by Guynn *et al.* [12], public perception of risks may contrast significantly with scientific conclusions based on the weight of scientific evidence from the cumulative primary literature. However, under current socio-political systems in most countries, public opinion carries significant influence over decision making and management policy, thus controlling the "social license to operate" on publicly owned lands. Current examples include restrictions or outright bans on chemical pesticide use in the forest-sector in certain political jurisdictions of Canada and the USA, despite registration and approval for these specific uses by federal regulatory agencies. Another example is the mandatory requirement to reduce or eliminate the use of chemical pesticides as a forest management option in some forest certification schemes [13], presumably reflecting the wishes of a more environmentally conscious and engaged consumer base.
# **Scope Statement**
Ecological risk estimation is generally considered as a tiered or hierarchical process which requires fundamental knowledge and data derived from scientific disciplines of environmental chemistry, biology, ecology and toxicology [14]. Production of these primary data is a legislative requirement of government regulatory bodies in many countries (e.g. the United States Environmental Protection Agency, the Canadian Pest Management Regulatory Agency and the Australian Pesticides and Veterinary Medicines Authority). Each of these regulatory agencies, as well as many other regional regulatory agencies, conduct independent reviews of the data prior to national registration and specific regional or sectoral use of pesticides. General discussion of the fundamental environmental fate and toxicology data requirements are discussed in chapters 1 and 2 of this text and will not be considered in detail here. Readers interested in more specific details on these fundamental toxicological data are directed to the Pesticide Information Profile briefs available on the EXTOXNET website [15], which provides convenient summaries for each pesticide. The United States Department of Agriculture – Forest Service documents also available *via* the internet [16] are another comprehensive source of data and information on how such data may be used directly in human health and environmental risk analysis.
Over and above these fundamental regulatory data requirements, numerous higher tier experiments and field investigations are conducted to inform the process. Among the various classes of pesticides that might be applied in forest management, herbicide and insecticide use predominates, with relatively minor amounts of fungicides being broadcast applied to plantations or natural forest stands [6]. As such, discussion in this chapter will be restricted to herbicidal and insecticidal compounds and based on four selected case examples (two for each pesticide class). Case examples were chosen as representative compounds most commonly used in the forest sector, or because they emphasize key ecotoxicological issues which are integral to the continuous debate over pesticide use both in the forest sector and more generally. The examples put forward in this chapter are intended to demonstrate the wealth of scientific information pertinent to possible ecological impacts of major forest-use pesticides and to emphasize the importance of higher tier manipulative field experiments and monitoring as critical components of the overall risk assessment process governing their regulation and use.
# **USE PATTERNS AND EXPOSURE ASSESSMENT FOR MAJOR FOREST-USE PESTICIDES**
The critical determinant of any toxicological effect is the dose; that is the level of the toxicant which occurs at the physiological site of activity within the organism. As such, toxicological effects are often directly proportional to environmental exposure concentrations with due consideration for modulating effects associated with the fundamental biology or behaviour of the receiving organism. For example, the feeding rate and preferences of different insects may influence exposure, while seasonal development of plant cuticles may act as a barrier to herbicide uptake in plants. In the case of forest-use pesticides, which are intentionally applied to known areas for very specific purposes, typical use patterns and application rates (Table **1**) are, in turn, the key determinants of potential environmental exposures. The actual application rate employed is selected by experienced forest managers based on the degree of infestation, susceptibility of the pest problem and cost considerations. Often the rates employed operationally may be less than the maximum allowed.
As competing vegetation and insect pest problems in the forest sector often occur at very large spatial scales and with substantial infestation intensities, broadcast techniques are often the only practically feasible method for applying the chemical to target sites. Forest-use herbicides, excepting soil active compounds, are applied with the specific intent of impinging the maximum possible mass of active ingredient on foliage of the competing vegetation canopy. Similarly, insecticides are typically applied such that they impinge predominantly within the crop tree canopy upon which many insect pests feed. Thus, non-target organisms residing or foraging in targeted plant canopies have the greatest likelihood of direct exposure [17, 18]. However, since not all of the depositing spray cloud is impinged within the target canopy, exposures of ground dwelling, soil or aquatic organisms may occur to some extent through either direct or indirect mechanisms (e.g. by rain-wash) and cannot be completely disregarded. Such exposures may be of particular importance in cases where highly sensitive or rare species are known to occur. The development and use of various new technologies including low drift nozzles, electronic guidance systems on spray aircraft and geographic information system mapping of spray blocks have greatly improved control and optimization of spray deposition. When used in conjunction with recently developed decision support systems such as SprayAdvisor, such advanced tools and techniques can substantially reduce the probability of depositing toxicologically significant levels of pesticide outside the targeted spray area[10].
**Table 1:** Comparative examples of maximal and typical use rates, as well as calculated and actual environmental concentrations observed in field research and monitoring studies for four major forest-use pesticides.
Superscripted numbers in brackets correlate directly to references from which data were obtained.
\* kg/ha unless otherwise noted
Accord®, Roundup Original®,Vision® and VisionMax® are registered products of the Monsanto Co., St. Louis, Missouri; Garlon 4®, and Release® are registered products of DowAgroSciences, Indianapolis IN; Foray 76®, Foray 48B® and Dipel® are registered products of Valent Biosciences, Toronto ON; Dimlin 4L® and Dimlin 25W® are registered products of Uniroyal Chemical Co., Bethany CT;
**Table 2:** Observed concentrations, primary mechanisms of degradation or dissipation and persistence estimates for major use pesticides in environmental compartments of various forest ecosystems.
Superscripted numbers in brackets correlate directly to references from which data were obtained.
Pesticides currently in widespread use in the forest sector may be generally characterized as non-persistent, susceptible to microbial degradation, photolysis, hydrolysis or other degradation mechanisms and nonbioaccumulatory. Extensive scientific knowledge on their fundamental physico-chemical properties and on their environmental fate under both laboratory and representative field scenarios exist. From field experiments under quasi-operational conditions, empirical estimates of initial concentrations observed in various environmental compartments as well as time to 50% dissipation (DT50 or half-life) estimates are also available as shown in Table **2**.
DT50 values provided in Table **2** indicate that residues of pesticides commonly used in the forest sector are relatively short-lived in all major environmental compartments. As such, exposure regimes are typically characterized by peak concentrations occurring shortly after application and with diminishing magnitude of exposure through time. The duration of exposures are often curtailed by the combined effect of environmental degradation and dissipation mechanisms which are active in these compartments. The resultant changes to chemical structure or bioavailability may significantly modulate exposure regimes and thus potential toxicological effect. Commonly, where wildlife exposures to pesticides occur, the exposure regime may be characterized as a pulse exposure of relatively short duration. In some cases, natural environmental exposure regimes differ markedly from those typically employed in standard tier 1 toxicity testing protocols in which test concentrations are artificially maintained at some constant high level. Considering all of the foregoing information, Tier 1 hazard quotient analyses, which are based on estimated exposure under the assumption of maximum labeled use rates and effect endpoints derived from atypical exposure regimes, should be considered as worst case risk estimates. Often the magnitude and duration of real world exposures, as well as toxicity observed in studies conducted *in situ*, are substantially lower than those predicted from simple hazard quotient analyses. Nonetheless, the majority of these types of risk analyses demonstrate that major forest-use pesticides do not pose substantial risks of direct toxicity to most wildlife species. Risks are generally greater in cases where the mechanism of activity is common to target and non-target organisms alike (e.g. acetylcholine esterase inhibition) and where both groups may be equivalently exposed (e.g. target and non-target insects in forest canopies) or where a particular group of organisms are uniquely sensitive to the pesticide or constituents of the pesticide formulation. While it is recognized that there are exceptions to most, if not all, generalities (some of which are described below), ecological impacts associated with the modern pesticides currently in widespread use for forest insect pest control and vegetation management are much more likely to occur through indirect mechanisms, such as changes in habitat or food availability, as opposed to direct acute toxicity.
# **FOREST-USE HERBICIDES**
Among countries leading international trade in forest-resource based products, only a handful of herbicidal active ingredients are registered and commonly used to control competing vegetation as a means of enhancing forest regeneration (Table **1**). Given that vegetative competition is most critical during the early establishment phase of forest regeneration [24], herbicide applications are typically made to prepare the site just prior to planting or in the very early stages (1-3 years) subsequent thereto. It is important to recognize that herbicide treatments therefore follow shortly after the major physical disturbances which result from harvesting and planting operations. In ecological terms, this is a transient stage in the cycle characterized by dynamic change and relatively rapid vegetative succession. Immediately following the physical disturbance of harvesting, sites typically become dominated by pioneer plant species well adapted to the high light intensities, disturbed soils and fluctuating temperatures which are often characteristic. As such, the potential changes in ecological structure and function that may be induced by herbicide treatments, must be considered in the context of the typical ecological dynamics of the sites to which they are applied and with due consideration to the dynamics in the broader forest landscape to which that specific site is connected [25-27].
The environmental fate and effects of herbicides used in forest vegetation management have been extensively investigated at experimental scales ranging from small laboratory studies to whole ecosystem manipulations. Several directly relevant reviews have been published previously [28-37, 52]. Independent regulatory reviews conducted in several countries (e.g. USA, Canada and Australia) with significant herbicide use in the forest sector consistently conclude that when applied in accordance with their specific product labels, such uses do not pose a substantive risk to wildlife or general environmental health. A special issue of the Wildlife Society Bulletin considers the transport and direct toxicity of many of many herbicides noted here [29]. A discussion of indirect influences of herbicide products used predominantly in the south-eastern USA on forest biodiversity [30] and wildlife habitat [12] is also included in the same publication. Collectively, the authors drew the following general conclusions from their review of the pertinent scientific literature:
These conclusions are drawn largely from studies conducted in the south-eastern USA. However, they are further supported by results derived from several higher-tier studies conducted in Canada and in other major forest regions and may thus be considered as generally applicable. Below, case study examples for two different herbicides (glyphosate and triclopyr) are presented to illustrate scenarios in which ecotoxicological field studies demonstrated substantially differing levels of risk and the value of conducting detailed studies under real-world conditions as a critical component of a hierarchical approach to ecotoxicological risk assessment.
# **Glyphosate**
As well as being the dominant herbicide in modern agriculture [31], glyphosate is also one of the most widely used herbicides in the forest sector around the globe including in Canada, the USA and Australia. For example, glyphosatebased products have accounted continuously for more than 93% of the total forest market in Canada for the 15 year period from 1992 through 2006 [8]. The knowledge base pertaining to the ecotoxicology of glyphosate is arguably the most extensive ever developed for a forest-use herbicide. The general environmental behaviour and toxicology of this herbicide has been the subject of several major independent reviews [32-35]. In addition, a seminal text presents much of the historical background and detailed information on all aspects of this unique compound in the early years post-discovery [36]. A search of several electronic databases provided several hundred records of primary scientific literature specific to the fate and effects of glyphosate in forest ecosystems. Many of these are field studies involving formulated end-use products applied at typical or maximal application rates and designed to examine the fate and effects under natural conditions typical of major forest uses. In addition, a number of other field studies are currently being conducted to address specific issues of scientific, public or operational forestry interests.
Since its discovery and introduction by Monsanto, numerous formulations of glyphosate have been registered and used in forest vegetation management globally. More recently, with the loss of patent control, multiple manufacturers are generating "generic" glyphosate products and more than 35 different formulations are used in the USA alone [37]. From both a use and ecotoxicological perspective, not all formulations are equivalent, largely owing to differences in the exact chemical composition of the products but also because of differences in application methods and rates as specified on the product labels. Several formulations contain different glyphosate salts or different surfactant blends which may significantly influence the uptake of the chemical in plants or potential ecotoxicological effects. However, in general, it is known that glyphosate is rapidly taken up by the plant following application of the formulated product and thereafter translocated to active growing tissues in both the aerial and root structures. As such, it is particularly effective for control of biennial or perennial species which self-propagate from basal sprouts, roots or rhizomes. Plants with this type of reproductive strategy are often the most problematic in forestry, particularly because they tend to be very poorly controlled by mechanical techniques. Often mechanical cutting actually stimulates more extensive growth, thereby exacerbating rather than alleviating competition with more desirable crop species. The mechanism of action for glyphosate involves blockage of a specific enzyme (5 enolpyruvyl-shikimate-3-phosphate synthetase or EPSPS) in the synthesis of aromatic amino acids. This biosynthetic pathway exists in both plants and microorganisms but not in higher animals [38, 39]. Owing to its highly plant-specific mode of action, direct effects of glyphosate on animals generally require much higher dose levels than would be typically encountered in natural environments, thus conferring a substantial level of safety for many wildlife species that may be potentially exposed. The environmental fate and persistence of glyphosate has been examined in vegetation, litter, soil, and water compartments of forest ecosystems ranging from the Pacific coastal forests in both the USA [40, 41] and in Canada [42, 43], to high latitude coastal and interior forest sites in Alaska [44], in southern and northern deciduous forests of the USA [45], in boreal forest sites of central Canada [46- 48] and in the Acadian forest region of eastern Canada [49, 50]. The results of these extensive field studies allow for broad inferences on the environmental fate of glyphosate in forest ecosystems. In general, it is known that glyphosate is effectively impinged within the target canopy, with relatively low residues in ground vegetation or in soils. In all compartments, glyphosate is susceptible to rapid microbial degradation and thus non-persistent. It binds strongly to essentially any organic substrate including organic matter and clay particles of sediments and soils, and thus shows essentially no tendency to leach or move laterally with surface runoff even though it has relatively high solubility in water. The time to 50% dissipation for glyphosate in these various environmental compartments is provided in Table **2**. The primary degradation product is aminomethylphosphonic acid (AMPA) and several studies indicate that AMPA is also non-persistent under typical forest environmental conditions. At least one assessment [51] has focused specifically on AMPA which suggests that it provides little risk to aquatic organisms.
The United States Department of Agriculture – Forest Service [52] provided the first comprehensive review on glyphosate fate and effects related to forest uses in 1984, with a subsequent workshop proceedings pertaining to uses in coastal forests of western Canada constituting a second review [53]. Both documents provide detailed estimates of environmental exposures following normal use and concluded that such levels would be expected to have neither acute or chronic toxic effects, nor reproductive effects in animals. Durkin [37] published a more recent review and risk assessment in 2003, pertaining to typical ground-based backpack spraying of glyphosate at rates of 2.24 kg a.i./ha. The risk assessment generally supported the conclusions reached by the U.S. EPA, indicating that based on the currently available data, effects on birds, mammals, fish and invertebrates are minimal. Sullivan and Sullivan [54] provided another review of more than 60 published studies on glyphosate in forestry, considering potential effects of this management practice as a disturbance agent in forest ecosystems and focusing on aspects relating to biodiversity. These authors concluded that species richness and diversity of vascular plants, songbirds and small mammals were either not affected or affected to only a minimal degree by glyphosate treatments. The degree of change observed in all cases was considered to be within natural fluctuations. For both avian and small mammal species, temporary declines did occur in some species, whereas in other species, abundance actually increased in treated sites. Such differential responses are largely attributable to the specific habitat preferences of the species in question. For those species whose preferred habitat is removed by the herbicide treatment the typical response is transient reduction in populations in these specific treated sites, followed by return when these habitat features become re-established on the site. The impact of glyphosate on large mammalian herbivores was measured by abundance of animals and food plants and by habitat use. Hares (*Lepus* spp.) and deer (*Odocoileus* spp. and *Capreolus capreolus*) were little affected, whereas reductions in plant biomass and related moose (*Alces alces*) forage and habitat use generally occurred for 1 to 5 years after treatment. Studies on terrestrial invertebrates covered a wide range of taxa with variable responses in abundance to glyphosate treatments. The authors noted that management for a mosaic of habitats, which provides a range of conditions for plant and animal species, are likely to ameliorate any short-term changes in species composition which might occur on specific sites treated with glyphosate to enhance regeneration success and plantation growth rates following forest harvesting.
Several major field studies, as well as a hierarchical suite of lab to field studies focused on the effects of glyphosate on amphibian species, have been completed. Results of these studies provide a substantial empirical basis which taken as a whole demonstrates very low potential for significant direct deleterious effects of formulated glyphosate products on non-target organisms in forest ecosystems. One of the earliest of these studies was a long-term investigation conducted in the Carnation Creek watershed of coastal British Columbia. This whole ecosystem experiment involved a fall aerial application of glyphosate (Roundup) in which the herbicide was applied at a rate of 2.0 kg a.i./ha to 41.7 ha of the watershed. General results were summarized by Reynolds and co-workers [53, 55] with more specific details provided in a series of published studies by several of the principal investigators involved. A key focus of the study was on comparative fate and effects of the herbicide in directly over-sprayed versus buffered stream channels. Feng *et al.* [43] documented maximum glyphosate residues of 162 μg/L in stream water, 6.8 μg/g dry mass in bottom sediments and <0.03 μg/L in suspended sediments of two intentionally over sprayed tributaries, dissipating to <1 μg/L within 96 h post application. Buffered streams were characterized by very low glyphosate residue levels <4 μg/L in stream water. Ratios of maximum stream water concentrations of glyphosate observed in buffered and over sprayed tributaries relative to literature toxicity values indicated a substantial margin of safety under either operational or worst case scenarios. Holtby and Baillie [56] examined the responses of coho salmon (*Oncorhynchus kisutch*) fingerlings and observed some stress and low mortality of 2.6% in caged fish located in the over-sprayed tributary. No similar stress or mortality were observed in other sites. Catch per unit effort in the over-sprayed tributary declined immediately after the application but recovered within 3 weeks. While this was taken to suggest that coho fingerlings had been stressed by some component of the herbicide spray, no treatment related changes in over-winter mortality, growth rates, probabilities of entering and leaving the tributary or timing of spring emigration were observed in the two years subsequent to herbicide treatment relative to results from 1 to 3 years of pre-spray monitoring. Kreutzweiser [57] also concluded that herbicide treatments did not unduly disturb stream invertebrates. While drift densities of most aquatic invertebrates did not increase in response to herbicide applications, the slight increase in drift response of *Gammarus* sp. and *Paraleptophlebia* sp. observed downstream of treated areas may have resulted from herbicide treatment. Feng *et al.* [42] also documented fate and persistence of glyphosate and its primary metabolite in terrestrial compartments. Residues in red alder and salmonberry foliage were 261.0 and 447.6 µg/g respectively and indicated good impingement on the target. Leaf litter residues, which averaged 12.5 µg/g for red alder and 19.2 µg/g for salmonberry initially, declined to less than 1 µg/g within 45 days post application (DT50< 14 days). In soils, glyphosate and AMPA residues were retained primarily in the upper organic layers of the profile, with >90% of total glyphosate residue in the 0 to 15cm layer. Distribution data for both glyphosate and AMPA suggested strong adsorption and a low propensity for leaching. Glyphosate soil residues dissipated with time resulting in estimated DT50 values ranging from 45 to 60 days. After 360 days, total soil residues of glyphosate were 6 to 18% of initial levels. Results of the Carnation Creek study were consistent with a similar study conducted by Newton *et al.* [41] in the Oregon Coast range. Additional work in the latter study examined the exposure of mammalian herbivores, carnivores, and omnivores. Results showed that retention of the herbicide varied with food preference; however, all species had visceral and body contents at or below observed levels in ground cover and litter, indicating that glyphosate did not accumulate appreciably in animal tissues.
The Fallingsnow ecosystem project conducted in the boreal forest of northwest Ontario is one of the few studies to comparatively examine the ecological consequences of herbicide treatments, including glyphosate, with other methods of vegetation management. In this experiment, treatments included aerial applications of triclopyr ester (Release) at 1.9 kg a.i./ha or glyphosate (Vision) at 1.5 kg a.i./ha with direct comparison to mechanical cutting using either brush saws or tractor-mounted cutting heads. Lautenschlager [58] concluded that herbicide treatments had relatively inconsequential effects on most ecological response parameters examined in this boreal forest site. As part of this multidisciplinary study, Simpson *et al.* [59] observed no substantial treatment-related differences in the movement of selected nutrients such as total organic N, NH4+, NO3-, K, Ca. Woodcock *et al.* [60] assessed the effects on songbird densities as determined by territory mapping, mist netting, and banding and observed 20 to 38 species breeding within various treatment blocks. First year post-treatment assessments revealed that mean densities of the 11 most common species increased by 0.35/ha on the control plots. In contrast, densities on treated plots decreased by 1.1/ha (brush saw), 1.6/ha (Silvana Selective), 0.14/ha (Release) and 0.72/ha (Vision). A point of emphasis here is that essentially any effective vegetation management technique will alter available habitat to some degree. In at least this one study, songbird densities were relatively less impacted by herbicide treatments as compared to mechanical treatments. Response to these habitat changes will vary with species, favouring certain species while resulting in out-migration of other species at least for some period of time. As a single species example, chestnut-sided warbler (*Dendroica pensylvanica*) had lower (*p* <0.05) mean densities on the brush sawtreated and Silvana Selective-treated plots than on the control plots and fewer (*p* <0.05) female birds were captured in the first post-treatment year.
A particular strength of the Fallingsnow ecosystem project was the detailed studies on plant communities where relative differences were tracked before and 1 to 5 years after treatments. Newmaster and Bell [61] showed that species richness and abundance of pteridophytes, bryophytes and lichens were reduced by all of the silvicultural treatments. Herbicide applications had the greatest initial effect on species richness, species abundance, and diversity indices. The authors noted that cryptogam diversity showed signs of recovery 5 years after treatment and that missed strips or untreated areas within a clearcut, provided a refuge for remnant communities and could play a key role in the rehabilitation of sites in terms of recovering the full suite of plant diversity. Bell and Newmaster [62] further reported that woody, herbaceous and graminaceous species showed transient declines in species richness, abundance or foliar cover, diversity indices, and rank abundance, as would be expected given the intent of the treatment. As a result, spruce trees proliferated in the regenerating plantations, but in no case did single layer monocultures occur. While herbicides had a relatively greater initial effect on plant community composition as compared to the two different cutting treatments, woody, herb, and grass layers showed substantial resilience to all treatments and recovered to pre-treatment levels within five years. Duchesne *et al.* [63] examined effects on total captures, species richness, diversity, and assemblages of adult carabids (Coleoptera: Carabidae) and found no effect on total capture rates but an increase in species richness and diversity in response to all treatments.
As noted by Guynn *et al.* [12] impacts of forest-use herbicides on amphibians is an area that has been historically understudied. In recognition of this general dearth of scientific knowledge and the potential for both aquatic and terrestrial life stages of amphibians to be directly exposed to formulated glyphosate products, Thompson and coworkers undertook a multi-tier, hierarchical project including both laboratory and field component studies [64]. Each tier of study provided unique and valuable data pertaining to overall risk assessment for amphibians. The authors also noted the need to consider potential multiple stress and multiple species interactions in ecotoxicological research. As the lead component study in this series, Edginton *et al.* [65] reported 96-h LC10 and LC50 estimates ranging from 0.85 to 3.5 mg a.i./L for early larval stages (Gosner 25) of *Rana clamitans* and *R. pipiens*. These endpoints remain among the lowest documented toxicity endpoints for amphibians exposed to formulated glyphosate products. The study confirmed that amphibian larvae were more sensitive than embryos and showed general equi-sensitivity among the four amphibian species tested. Results also demonstrated that larval amphibians are among the most sensitive of aquatic organisms when exposed to formulated products of glyphosate containing the POEA surfactant and thus the importance of testing end-use formulations. Surfactants are generally required to be used with glyphosate to allow effective uptake of this electronically charged molecule across plant cuticles. Inclusion of the surfactant also results in reduced losses from treated foliage *via* rain-wash [43, 66]. The inclusion of the POEA surfactant in many formulations is also very important from an ecotoxicological perspective. It is well recognized that POEA is the primary toxicant to aquatic species. POEA and other surfactants may affect membrane transport generally and often act as a general narcotic [32, 33]. As such POEA mediated toxicity is well established as a concern for aquatic organisms such as fish and amphibians for which transport of oxygen and other compounds across gill or skin membranes is a critical physiological function. Unfortunately, owing to the chemical complexity of the POEA surfactant and resultant difficulty in analysing for it in complex environmental matrices, the environmental behaviour of POEA in natural forest ecosystems has not been specifically studied. However, fate experiments conducted in the laboratory show that the surfactant is also readily degraded in soils with a half-life of less than 7 days, that desorption from soil surfaces is minimal, and that persistence in natural waters under laboratory conditions resulted in an estimated half-life of about 2 weeks. Results of these studies suggested that POEA would be lost from the water column following application by a combination of sorption to sediments and microbial metabolism [67]. The half-life of POEA in shallow waters (15 cm deep) in the presence of sediments has subsequently been reported as about 13 h [68], further supporting the concept that any potential direct effects of formulated products on organisms in natural waters are likely to occur very shortly post-treatment rather than as a result of chronic or delayed toxicity.
Tier II studies conducted by Chen *et al.* [69] confirmed the interaction of pH and Vision toxicity in *R. pipiens* larvae and showed parallel effects for zooplankton population response parameters, suggesting that the pH–Vision interaction is of general ecological significance. In addition, Tier II studies demonstrated that effects on zooplankton reproduction could also be exacerbated by food deprivation when presented as a concomitant stressor. *In situ* enclosure studies conducted by Wojtaszek *et al.* [70] in two different wetlands systems showed 96-h LC10 and LC50 values generally higher than those derived from laboratory studies. This result was attributed to reduced magnitude and duration of exposures resulting from natural degradation and dissipation mechanisms which are active in real-world systems. Results clearly demonstrated the importance of including *in situ* manipulative studies in ecotoxicological risk assessments. Contrary to the results of the lab-based studies, the *in situ* enclosure experiment lead to the conclusion that typical silvicultural applications of Vision would not be likely to generate significant direct mortality in native amphibian larvae. This conclusion was strongly supported by both chemical and biological monitoring studies as reported by Thompson *et al.* [71] as the fourth and final tier of the research program. Results from these Tier IV studies showed no statistically significant differences in mean mortality among larvae of two different amphibian species (*R. clamitans* and *R. pipiens*) differentially exposed in over-sprayed, adjacent, and buffered wetlands. Results of the operational monitoring study were consistent with concentration-response relations from both Tier I and III studies since 99% confidence limits for real-world exposure concentrations in all wetland cases were below both estimated LC50 and LC10 values. As a general conclusion, results of this tiered research program indicate that aerial applications of the herbicide Vision, as typically conducted for conifer release in forestry, do not pose a significant risk of acute effects to the most sensitive aquatic life stages of native amphibians in forest wetland environments. The conclusion was consistent with specific risk assessments for formulated glyphosate products in aquatic systems [33]. Results of ongoing field studies consistently support this conclusion, thus allowing researchers to refocus their attention on more subtle but equally important potential effects on amphibian populations associated with possible indirect or multiple stressor interactions [72, 73].
# **Triclopyr**
Triclopyr is the common name for ((3,5,6-trichloro-2-pyridinly)oxy)acetic acid, the active ingredient of formulated commercial products such as Garlon 3A and Garlon 4. These two products also represent two different chemical forms of triclopyr, that is the triethylamine salt and the butoxyethyl ester (BEE) respectively. Triclopyr mimics indole auxins as plant growth regulating hormones and causes plant mortality through induction of irregular cell growth, particularly in the stem tissues of vascular plants. Typical use rates for triclopyr are in the range of 4 kg a.i./ha, comparatively higher than those for glyphosate. Although triclopyr receives markedly less use in the international forest sector than glyphosate, it is a regionally important forestry herbicide in the southeastern USA and other areas where it is typically applied using ground-based techniques. The fate and effects of triclopyr in forest ecosystems have been previously reviewed [52]. In combination with data derived from several field studies conducted in a variety of forest ecosystems, it is well documented that triclopyr dissipates rapidly from foliage and soils. The primary degradation mechanism in soils is microbial and the principal metabolite is trichloropyridinol. Both laboratory and field study results suggest that triclopyr exhibits limited to moderate leaching or lateral mobility in soils [40, 74-76]. In aquatic compartments, BEE degrades *via* base-catalysed hydrolysis to yield triclopyr acid [77] which in turn may further degrade by either photolytic or biological means to yield the principal metabolite [52].
Wan [78], studied the comparative acute toxicity of Garlon 3A, Garlon 4, triclopyr, triclopyr ester, and their transformation products to juvenile Pacific salmonids, demonstrating that the ester was considerably more toxic than all other forms. The ester form of triclopyr is considered to be approximately 100 fold more toxic than the acid [79]. McCall [80], conducted simulations of the aquatic fate of triclopyr butoxyethyl ester emphasizing the importance of mechanisms converting the ester to less toxic forms as this is a critical determinant of potential toxic effects in fish such as coho salmon, as well as other aquatic organisms. Under low pH or cool temperature conditions, the transformation of ester to acid may be relatively slow and thus variations in these environmental parameters may strongly influence ecotoxicological outcomes. In this regard, toxicity of the ester form of triclopyr to fish, amphibians and aquatic invertebrates is the major concern in relation to potential ecological impacts and this aspect has received a substantial amount of scientific investigation.
Kreutzweiser and co-workers, conduced time-toxicity tests with rainbow trout (*Oncorhynchus mykiss*) under both laboratory and field studies. In flow-through toxicity tests [90] the effect of exposure time on the toxicity of triclopyr butoxyethyl ester (Garlon 4) to fish (rainbow trout, *Oncorhynchus mykiss*, and chinook salmon, *Oncorhynchus tshawytscha*) and stream insects (*Hydropsyche* sp. and *Isonychia* sp.). The toxicity of triclopyr ester to all species increased with increasing time of exposure to the ester. For example, median lethal concentrations for rainbow trout exposed for 1, 6, or 24 h were 22.5, 1.95, and 0.79 mg a.i./L of triclopyr ester. Results suggested that even under conditions where maximal predicted environmental concentrations (2.7 mg a.i./L) might occur, risk of acute toxicity would be very limited under typical exposure durations observed in flowing systems. In contrast, considerably higher risk of acute lethal effects could be predicted under conditions where the ester form might persist for more than 6 h, even when initial concentrations were as low as 0.7 mg a.i./L. The authors noted the aquatic organisms in lentic systems (such as wetlands, ponds and lakes) are likely to be most at risk. These relations were subsequently confirmed in various field studies.
A major multidisciplinary study focused on the ecotoxicology of triclopyr ester (Garlon 4) following aerial application at a rate of 3.84 kg a.i./ha that was conducted in a typical boreal forest watershed of northern Ontario, Canada. A particular focus of this study was on the fate and effects of the more toxic form of triclopyr (BEE) in the stream under a worst case scenario of direct overspray [81]. Results showed an average deposit at the stream surface of 3.67 kg a.i./ha with BEE residues in stream water exhibiting instantaneous maxima of <0.35 mg a.i./L. A series of diminishing pulses were observed resulting from direct inputs during overspray of the stream channel upstream. Average concentrations of the BEE in stream water ranged from 0.05 to 0.11 mg/L during the first 12 to 14 h monitoring period and were below limits of detection within 72 h. Both the average concentrations and exposure durations observed in this field study were substantially below levels generating acute lethal responses for various aquatic organisms in either lab or field studies [e.g. 83-87]. Initial whole body tissue residues in samples taken from fathead minnow cages *in situ* at the downstream location (43 mg a.i./kg) were similar to those predicted from simulation models [80]. No statistically significant mortality was observed in three species of aquatic organisms (yellow perch, caddisflies or fathead minnows) caged *in situ* either in treated or control areas. The authors concluded that natural dissipation mechanisms including photolysis, hydrolysis and microbial action limited exposures to sublethal levels and that based on this study, significant impacts to aquatic organisms would not be anticipated under operational conditions where such streams would be protected by buffer zones of 60 to 100 m. Similarly, in a field experiment in which trclopyr BEE (Garlon 4) was directly injected directly into a small headwater forest stream, intensive sampling [82] showed maximal aqueous concentrations of 0.848 and 0.949 mg a.i./L at the monitoring stations nearest two discrete injection points. Average BEE concentrations ranged from 0.32 mg a.i./L at stations nearest injection points to 0.02 mg a.i./L approximately 225 m downstream. Results demonstrated rapid conversion of the BEE to triclopyr acid in this system, as well as significant sorption of the chemical to natural allochthonous (deciduous leaf pack) materials. Resultant short-term, pulse-type exposures of BEE were observed with magnitude decreasing and duration slightly increasing with downstream distance. Resultant exposure regimes failed to induce any mortality of resident brook trout, nor were there significant effects on the growth of 1 or 2 year old brook trout.
In contrast to the results of lotic system experiments, substantial toxicity to a variety of aquatic organisms has been observed in lentic studies characterized by longer duration of exposure to the more toxic BEE form of triclopyr. Kreutzweiser *et al.* [83] conducted a dose-response study on fish caged within *in situ* enclosures in a northern Ontario lake. Results showed median dissipation times for aqueous residues ranging from 4 to 8 days. All caged rainbow trout exposed to initial concentrations greater than 0.69 mg a.i./L died within 3 days and 43% mortality was observed at 0.45 mg a.i./L whereas no mortality was observed at the 0.25 mg a.i./L level. Using similar *in situ* enclosures in two different wetland ecosystems, Wojtasek *et al.* [84] studied the effects of triclopyr BEE (Release) on mortality, avoidance response, and growth of larval amphibians (*Rana clamitans*, *Rana pipiens*). A range of treatment concentrations were applied to yield nominal concentrations ranging from 0.26 to 7.68 mg a.i./L. Concentrationdependent mortality and abnormal avoidance response were observed but there were no significant effects on growth. Toxicity for the two test species (*R. clamitans* and *R. pipiens*) were less than those observed in prior laboratory studies [85-87], probably due to the rapid dissipation of BEE which showed a DT50 of less than 1 day in both of these shallow wetlands. The authors noted that LC10 and EC10 endpoints approximated aqueous concentrations of 0.59 mg a.i./L that is within the range for expected environmental concentrations in small wetland amphibian breeding habitats under direct aerial overspray scenarios, thus presenting a potential risk of impacts for a small proportion of native amphibian larvae. This conclusion was consistent with results of laboratory microcosm studies in which Chen *et al.* [88] showed that triclopyr BEE (Release) at environmentally relevant test concentrations (0.25 and 0.50 mg a.i./L) resulted in significant decreases in survival of both larval life stages of *R. pipiens* and a common wetland zooplankton species *Simocephalus vetulus*. Moreover results indicated that effects on amphibians and zooplankton may be amplified by other concomitant stressors such as low food availability or low pH.
Overall, risk assessments for triclopyr BEE based on early tier experiments identified a substantial risk of acute toxicity to fish, amphibians, zooplankton and aquatic invertebrates, particularly in lentic systems where dissipation of the ester form is limited in some way. The presumption of risk was confirmed by subsequent field studies in scenarios where longer term exposure to the more toxic ester form occurred, but not in lotic scenarios where the duration of exposure to the ester was too short to attain toxic thresholds in aquatic organisms. Results emphasize the particular importance of understanding both the duration and magnitude of exposures that occur in real-world systems and the need for considering such natural exposure regimes when designing or interpreting research results and also when considering potential mitigative measures.
# **FOREST-USE INSECTICIDES**
As compared to herbicides, fewer insecticides find widespread use in the forest sector internationally. Among those most commonly in use are the biological control agent *Bacillus thuringiensis* var. *kurastaki* (Btk) and the unique chitin-formation inhibiting chemical pesticide diflubenzuron (Dimlin®) (Table **1**). The use pattern for these products is highly sporadic with amounts applied varying dramatically in relation to the extent and severity of major insect pest outbreaks. Unlike herbicides, applications of insecticides are typically made to protect semi-mature or mature high value timber stands. Defoliating insect pests of significance historically in North America include the gypsy moth, spruce budworm, western spruce budworm, blackheaded budworm, jack pine budworm and Douglas fir tussock moth. Data provided by the USDA-Forest Service [16] indicates that of the total area treated for gypsy moth in the northeast region, 77% received applications involving Btk, while approximately 22% of the area was treated with Dimilin. In Canada, Btk is by far the most commonly used product accounting for approximately 86% of forest insecticide use [8], with the remainder being primarily tebufenozide (MIMIC). In the UK only four active ingredients were registered in 2004 as chemical insecticides for use in forestry [89]. Selected case studies for Btk and diflubenzuron are presented below to illustrate specific ecotoxicological issues of interest associated with forest uses of these active ingredients.
Increasingly, invasive insect species such as the mountain pine beetle, emerald ash borer, Asian long-horned beetle and brown spruce longhorn beetle are posing new and significant ecological and economic risks to the forest sector in North America [90, 91]. Similar invasive insect pest problems threaten forests in other countries, and often these are occurring in urban forest environments presenting several unique issues. For example, broadcast insecticide applications, as typically used against the major defoliating insect species, may be ineffective or publicly unacceptable as controls for invasive wood boring species. These issues have prompted the development and use of novel systemic injection techniques, as well as natural product insecticides such as azadirachtin [91] as alternative control techniques within broader integrated pest management strategies. A recent review [92] documents the environmental fate and effects information associated with several of these compounds which are purported to represent "reduced risk".
# **Bacillus Thuringiensis Var. Kurstaki (Btk)**
Among several strains of *Bacillus thuringiensis* with notable insecticidal activity, the proteinaceous crystalline toxin of Btk is known to be highly specific to larval Lepidoptera [93, 94]. The mechanism of action of Btk in Lepidoptera is the result of toxin induced rupture of the midgut followed by spore germination and septicemia in the body cavity that eventually results in death [95]. Several different formulations of Btk (see some examples in Table **1**), are used extensively in the USA and Canada, as well as for the control of Lepidopteran insect pests worldwide. One example of the latter is the use of Btk in an attempt to eradicate the invasive painted apple moth in New Zealand for which a comprehensive impact assessment has been published [96]. In North America, for major pests such as gypsy moth, spruce budworm, jack pine budworm and hemlock looper, applications are typically made by aircraft. Unlike conventional pesticides, the potency of Btk formulations is determined based on standardized bioassay response and reported in terms of Billions of International Units (BIUs). Typical application rates for Btk range from approximately 60 to 90 BIU/ha.
Results of published risk assessments [18, 96] indicate that given their highly specific mode of action, Bt products are unlikely to pose a significant hazard to vertebrates, fish, birds or insects other than macrolepidopteran larvae. Bt occurs naturally in soils throughout the world. The vegetative form of Btk does not generally persist in soil; however, endospores can survive in most types of soils for extended periods with half-lives of spores usually in the range of 100 to 200 days [97]. As noted in the New Zealand environmental impact assessment document [96], estimates on persistence of Bt toxins vary widely and there is some evidence to suggest that binding of Bt toxins to humic acids, organic supplements or onto soil particles protects the toxins from microbial degradation, without eliminating their
insecticidal activity. Leaf litter and soil samples collected following aerial spray Foray 48B for control of whitespotted tussock moth in Auckland, showed significantly enhanced levels of Btk-like isolates up to two years postspray. Laboratory studies by Visser and other workers [98], had previously shown that formulated Btk products generally had no effect on functional parameters associated with soil microflora.
The New Zealand environmental impact assessment generally suggested that relative to all available options, Btk was likely to be the most acceptable approach for attempted eradication of painted apple moth in the urban area of Auckland, from a public, economic, efficacy and environmental perspective. The World Health Organization specifically concluded that "Bt products may be safely used for the control of insect pests of agricultural and horticultural crops as well as forests". While such comprehensive assessments typically support the use of Btk as environmentally acceptable, there are concerns associated with potential ecotoxicological impacts on non-target Lepidoptera and derivative indirect effects on insectivorous species, particularly birds, which may depend upon these organisms as a primary food source, as well as potential effects on non-target aquatic insects.
Several studies demonstrate that Btk causes immediate reductions in abundance and species richness of non-target larval Lepidoptera [99-102]. Butler and co-workers [103] conducted extensive studies on this aspect following applications of Btk for control of gypsy moth in oak forests of West Virginia, USA. During the treatment year, Btk produced significant decline of canopy-dwelling macrolepidopterous larvae. No differences in abundance of various caterpillar species were observed among treated and control plots during the weeks or months following treatment. Similarly, no difference between treated and control plots were observed in abundance of most species in 1992, the first post-treatment year. Non-lepidoptera species also appeared to be unaffected by the Btk treatment. The fact that abundance and richness of non-target lepidopteran larvae declined during each year of the three year study, even on non-treated plots, emphasizes the importance of using appropriate controls in field studies of this type. It also underscores the need to consider pesticide-induced perturbations in light of the natural variation in abundance that may occur due to both random and non-random factors which typically influence biological systems in natural environments.
Boulton *et al.* [99] assessed the impacts of Btk (50 BIU/ha as Foray 48B) on native, non-target Lepidoptera following treatments to 12,805 ha of Garry oak forests for control of gypsy moth in southeastern Vancouver Island, British Columbia, Canada. Significant variation in diversity among the Lepidoptera were not detected, but reduced richness and abundance on two different host plant species were observed. The authors noted potential concerns associated with such effects, particularly in highly fragmented forest stands such as those associated with urban or industrial areas. They also emphasized the importance of such effects on rare and endangered non-target lepidopteran species such as *Euchloe ausonides isulanus*, and *Euphydryas editha taylori* which are found only in oak meadows and rocky knolls. In a follow-up study, Boulton *et al.* [100] examined longer term recovery of non-target Lepidoptera noting that reductions were greatest one year post-treatment. Relative to the reference sites, each of 11 species that were initially reduced by the Btk applications showed an increase in the treatment sites within the next 3 years, by which time only four species remained significantly reduced in the treatment sites. The uncommon species were significantly reduced in the year of treatment but not one or three years post-treatment. Results of this study highlight the importance of long term monitoring following pesticide induced perturbations in relation to understanding the rate and process by which recovery in ecosystem structural or functional processes may occur. The study also emphasizes the important consideration of effects on rare and endangered species. In general, and somewhat surprisingly, this aspect has does not appear to have received sufficient scientific attention. In the case of Btk, such a concern has been raised for the Karner blue butterfly. Herms *et al.* [104] conducted a field survey and laboratory bioassay demonstrating significant dose-dependent mortality in response to Btk treatments and found that early and late instars were equally susceptible. The authors concluded that the Karner blue is both phenologically and physiologically susceptible to Btk as employed for gypsy moth suppression, even though the larval generation at risk and extent of phenological overlap may vary from year to year.
In cases where direct effects on species in one trophic level deleteriously affect those in other trophic levels, concerns over ecological implications are heightened. In the case of forest-use insecticides with highly specific modes of action, such as Btk, potential indirect effects of reduced prey or food availability on insectivorous predators are of particular interest. Two separate studies [105, 106] have examined such indirect effects on insectivorous birds and small mammals in Ontario, Canada. As would be anticipated, substantial reductions in Lepidoptera larvae were observed in response to treatment. Many adult male shrews apparently emigrated and were replaced by young males and females. Effects on Nashville warbler and hermit thrush, chicks of spruce grouse, and adult male masked shrew were all attributed to indirect effects associated with reduction in the primary insect food source. Holmes *et al.* [105] further examined the hypothesis that food reductions caused by forest spraying with Lepidoptera-specific insecticides would affect songbird behaviour and reproduction. The comparative study of Tennessee warbler nests and parental behaviour involved spray blocks treated with Btk, tebufenozide (MIMIC) or left as an untreated control area. Nestling survival and growth were unaffected by the insecticide treatments. Nests in the treated blocks had smaller clutches, smaller broods and lower hatch rates than nests in the control block, but these differences were not statistically significant. Nestling diets were similar in the MIMIC and control blocks. There were slight differences in the behaviour patterns of female Tennessee warblers in the MIMIC and control blocks, with those from MIMIC treated spending less time at the nest and more time foraging. The authors concluded that the indirect effects of forest spraying with Lepidoptera-specific insecticides pose little risk to forest songbirds. Differential results may reflect species-specific behaviours, food preferences and ability to prey switch or broaden foraging ranges. The equivocal nature of these field study results, suggests that strategic species-specific biomonitoring and population modeling in conjunction with operational spray programs may be warranted to provide more conclusive evidence with regard to possible ecological consequences at broader spatial scales. Other studies have also documented potential indirect effects of Btk spraying associated with reductions in natural food for breeding black-throated blue warblers [107] and an endangered species – the Virginia big-eared bat [108].
Kreutzweiser *et al.* [109, 110] conducted a series of studies to examine the effects of Btk, as two different aqueous formulations of Dipel, on several aquatic invertebrate species (various species of Ephemeroptera, Plecoptera or Trichoptera). Results showed no significant mortality, drift response or consumption of treated leaf disks at levels well above label rates or expected environmental concentrations using either flow-through laboratory experiments or outdoor stream channel experiments. Although trends of reduced decomposition activity in treated outdoor stream channels were observed, there were no significant differences in mass loss of leaf material between treated and control channels. These results from laboratory and controlled field experiments indicated that contamination of watercourses with Btk is unlikely to result in significant adverse effects on aquatic invertebrates or microbial community function in terms of detrital decomposition. In a confirmatory field study, Kreutzweiser *et al.* [111] treated a section of a natural forest stream with Btk at 10 times the expected environmental concentration (200 BIU/mL) to determine effects on the aquatic macroinvertebrate community. Invertebrate drift density increased slightly, but only during the 0.5-h application and only at the site 10 m below the application point. There were no significant changes in taxonomic richness of benthic invertebrates after the application, but there were short-term alterations in community structure at the treated site after the application, as measured by a dissimilarity index. In 11 of 12 benthic taxa for which there were sufficient data, changes in abundance after the application were not significant compared with changes in abundance at the reference site. The stonefly *Leuctra tenuis* (Pictet) was reduced by ~70% at the treated site 4 days after the application, and abundance of this stonefly remained considerably lower, but not significantly different, from the reference site for at least 18 days. A follow-up study demonstrated that under laboratory conditions, Btk on leaf material was not toxic to *L. tenuis.* The Btk application had no significant effect on the growth or survival of caged caddisfly larvae, *Pycnopsyche guttifer*, in the treated stream.
# **Diflubenzuron**
Diflubenzuron is a benzoyl-phenylurea insecticide that inhibits chitin deposition in arthropods. It is effective either as a stomach or contact insecticide. In forestry, diflubenzuron sees major use principally in the USA for suppression of gypsy moth. Two formulations (Dimilin 4L and Dimilin 25W) are registered in the USA and the active ingredient is also efficacious against tent caterpillar and several other forest insect species including pine false webworm [112] and eastern hemlock looper [113]. For suppression of gypsy moth, diflubenzuron may be applied *via* either ground or aerial methods at rates ranging from 9 to 75 g a.i./ha. Typical use scenarios under severe infestation conditions may involve multiple applications over several years.
In a synoptic review of potential environmental effects of diflubenzuron [114], adverse effects on crustacean growth, survival, reproduction, and behaviour have been observed at environmentally realistic levels ranging from 0.062 to 2 µg/L. Rebach and French [115] examined the effects of diflubenzuron on blue crabs and provide a review of potential effects in marine and estuarine environments. This review demonstrated substantial toxicity to these species. Surprisingly, there appear to have been no field studies investigating potential effects on freshwater crayfish which are likely to be directly exposed during forest use. Mayflies, chironomids, caddisflies, and midges also have known sensitivity to diflubenzuron at similar aqueous concentrations, showing low emergence and survival as typical impacts. Fischer and Hall [116] reviewed the environmental fate and effects data on diflubenzuron with particular emphasis on aquatic systems. Organic matter and aquatic macrophytes are major factors influencing the adsorption and degradation of the compound. Reardon [117] presented an overview of field experiments examining the potential impacts of diflubenzuron (Dimilin 4L) on selected non-target organisms in an experimental broadleaf forest in West Virginia. Five non-target groups were monitored, including: fungi, bacteria, and invertebrates in leaf litter and soil; aquatic macroinvertebrates; canopy arthropods; pollinating insects and aquatic and terrestrial salamanders. Initial concentrations of diflubenzuron and degradation of residues on tree surfaces, in leaf litter, in soil, and in water were also determined. Except for aquatic macroinvertebrates, canopy arthropods, and native pollinating insects, there were no detectable effects of the treatment on the non-target groups. Diflubenzuron treatments were shown to decrease the densities and survival of several species of mayflies, stoneflies and a cranefly and reduce richness and abundance of non-target terrestrial arthropods, primarily macrolepidopterans and yellow jackets. Durkin [118] characterized the scientific database supporting the risk assessment of diflubenzuron (Dimlin) in forestry as large and somewhat complex, but concluded that direct effects of diflubenzuron on mammals, birds, amphibians, fish, terrestrial and aquatic plants, microorganisms, and non-arthropod invertebrates were considered implausible, largely owing to the specific mode of action for this compound. Just as for Btk, the fact that diflubenzuron is an effective insecticide against Lepidoptera results in a substantial likelihood of effects on other non-target members of this group, as well as indirectly on insectivorous species such as birds, which may specifically rely on these insect populations as their primary food source. Potential effects on aquatic invertebrates were also considered possible depending upon site-specific conditions controlling deposition to surface waters and thus resultant exposure levels.
Numerous laboratory studies have demonstrated the sensitivity of aquatic invertebrate species to diflubenzuron. Hansen and Garton [119] showed that among complex stream faunal communities in the laboratory, mayflies and stoneflies were affected at 1.0 µg/L and that crustaceans were also particularly sensitive. These authors also noted that single species toxicity tests adequately predicted direct lethal effects, but not indirect effects resulting from altered interspecies interactions. Liber *et al.* [120] conducted an elegant field mesocosm experiment using a concentration-response design approach and derived field EC50 values for insect emergence inhibition ranging from 1.0 to 1.4 µg/L. Overall, they concluded that significant adverse effects on insect emergence could be expected at diflubenzuron concentrations of >1.0 µg/L with the time to recovery being concentration dependent. Boyle *et al.* [121] also employed outdoor mesocosms in a study exploiting the unique mode of action of diflubenzuron to examine the indirect responses following direct impacts at the primary consumer (*i.e.* invertebrate) trophic level. Direct reductions in invertebrate grazers caused indirect increases in algal biomass. Indirect effects including 50% reductions in biomass and in individual weight of juvenile bluegill occurred because of apparent decreases in invertebrate food resources. In contrast, no statistically significant impacts were observed on adult bluegill or largemouth bass for the duration of the experiment. Results indicated that diflubenzuron had both direct and indirect impacts on the experimental aquatic ecosystems under the conditions tested and although treatment regimes were not environmentally realistic in relation to forest use patterns, the study provides an excellent example of potential indirect secondary and tertiary effects in aquatic systems that would be difficult if not impossible to determine under laboratory conditions. In a semi-operational field study in mixed wood forests of central Ontario, Canada, Sundaram *et al.* [23] reported that the fate of diflubenzuron residues following aerial applications of Dimlin 25W at a rate of 0.07 kg a.i./ha. Results indicated that dissipation patterns differed among water, sediment and aquatic vegetation substrates, with reported DT50 values of <1.3 days in pond water and <14 days in all cases. Zooplankton and benthic invertebrate populations were monitored for up to 110-day post-spray in two over-sprayed ponds with comparison to control ponds. Significant mortality occurred in two groups of caged macroinvertebrates (amphipoda and immature corixidae) 1 to 6 days post-treatment. Three taxa of littoral insects (*Caenis*, *Celithemis* and *Coenagrion*) were also significantly reduced in abundance in the treated ponds 21 to 34 days post-treatment, but recovered to pre-treatment levels by the end of the season. Zooplankton (cladocerans and copepods) populations were reduced 3 days after treatment and remained suppressed for 2 to 3 months.
Harrahy *et al.* [122] noted that diflubenzuron may persist on hardwood leaves throughout the growing season up until the time of leaf fall. Non-target aquatic organisms that consume these fallen leaves may therefore be exposed to the pesticide for a significant period of time. Several field studies have further investigated the potential effects of diflubenzuron on sensitive non-target aquatic invertebrates and confirmed effects under environmentally realistic scenarios. Griffith *et al.* [123] used Malaise traps to monitor emergence and flight distances of adult Plecoptera and Trichoptera from headwater streams in two different catchments of an experimental forest in West Virginia, USA before and after application of diflubenzuron. Stonefly, *Peltoperla arcuata* emergence was reduced in the first 4 months after treatment, as compared with the untreated catchments, however no differences in emergence of other species were observed. In a follow-up study [124], the flight of the stonefly *Leuctra ferruginea* was reduced in the treatment watersheds compared with the reference watersheds during the year following abscission of the treated leaves. Adult flight of other species did not decrease in the treatment watersheds during 1993. These results suggest that among aquatic invertebrates, stoneflies may be particularly sensitive to the effects of diflubenzuron even under scenarios of a single application. The authors noted that multiple applications of diflubenzuron over several years, which often occurs during gypsy moth suppression programs, may present a significant risk to these aquatic species. Similarly, Hurd *et al.* [125] observed significant reductions in the abundance of several taxa in treatment as compared to control watersheds following aerial application of diflubenzuron (Dimlin 4L) at a rate of 70 g a.i./ha. Most affected taxa included the stoneflies, *Leuctra* sp. and *Isoperla* sp., the mayfly, *Paraleptophlebia* sp., and the cranefly, *Hexatoma* sp. In a functional context, shredders were the dominant group affected, with reduced mean densities in treatment watersheds whereas densities of species such as Oligochaeta and Turbellaria increased in streams in treated watersheds. The authors again emphasized that since most aquatic insects oviposit in the watershed from which they emerge, repeated applications of diflubenzuron could have longer-term localized effects on invertebrate fauna in treated streams. In contrast to these studies, Boscor and Moore [126] studied the impacts of Dimilin at 70 g a.i./ha to a one-half mile stretch of White Deer Creek in central Pennsylvania. No spray-induced, adverse effects were detected on the organisms sampled, principally Ephemeroptera, Chironomidae, Trichoptera, and Plecoptera, for a period of up to 28 days after treatment.
The potential effects of diflubenzuron on non-target terrestrial insects have also been extensively examined. Sample [127] reported that the operational application of Dimilin [70, 75 g a.i/ha] resulted in greatest impacts on Lepidoptera which displayed reduced abundance and species richness at treated sites. No effects were observed among Coleoptera, Diptera, or Hymenoptera. Butler *et al.* [128] summarized results of a 6-year study conducted to evaluate the impact of diflubenzuron on the diversity and abundance of arthropods in West Virginia. Based on foliar sampling, overall arthropod family diversity and abundance, numbers of macrolepidoptera and beetles were significantly reduced in treated watersheds. Total arthropod abundance and macrolepidoptera abundance remained at significantly lower levels up to 27 months post-treatment. As noted by Durkin [126] some secondary effects resulting from reduced Lepidoptera prey may include increased foraging range, relocation and lower body fat content among foraging birds species. For example, Whitmore *et al.* [129] showed significantly lower fat reserves in seven of nine tested bird species following Dimlin applications to forests in the USA. Possible causal factors were listed as reduction in food availability and decreased biomass ingestion, increased energetic expenditures required in obtaining scarce food and reduced food quality in treated as compared to control sites. The latter study is an example of an investigation on functional (community energetics) rather than structural effects of pesticide use in forest ecosystems, an area which is generally under-studied. Another example is the study by Paulus *et al.* [130], who compared three methods for assessing the impacts of forest-use insecticides diflubenzuron and Btk on biological activity of soil organisms. While results were dependent on the monitoring technique employed, overall findings demonstrated transient effects on biological activity of soil organisms exposed to diflubenzuron but not Btk.
# **CONCLUSION**
The cumulative wealth of scientific data available for modern forest-use herbicides and insecticides is extensive. Research spans multiple tiers of testing ranging from simple laboratory studies, through microcosm and *in situ* mesocosm studies and includes several comprehensive large scale field experiments. Higher tier field studies provide several unique benefits that are considered highly contributory to comprehensive ecotoxicological risk assessments. As many previous authors have suggested, it is impossible to replicate natural ecosystems, inclusive of all of their innate and interactive physical, chemical and biological components in the laboratory. In ecotoxicological risk estimation, direct use of data from any laboratory study carries the critical and highly questionable assumption of equivalence of the test system and the real world. As such, it is very prudent to continue the use of *in situ* mesocosm, manipulative field studies and long-term monitoring to confirm that extrapolative predictions based on early tier laboratory studies are in fact valid. In higher tier field studies, experimentation should be focused on typical operational as well as worst case maximal use rates with common end-use products such that results incorporate any potential effects associated with surfactants or other formulants contained therein. In terms of response variables, these should be focused on population or community level response and recovery time, as these are typically most relevant to regulatory and policy decision making and involve levels of biological organization and interaction mechanisms (e.g. predation, competition, commensalism) that cannot be effectively simulated in laboratory experiments. Examination of the case studies presented here highlight all of these unique benefits as well as the overriding value of large scale field experiments in terms of negating or confirming risk. While these benefits and values are particularly important in forestry scenarios owing to the typically large scale of operations, similar advantages apply to field experimentation in other sectors as well.
Given the relatively specific mode of action of many modern pesticides, their general high water solubility and facile environmental degradation and metabolism, environmental concerns associated with modern forest-use pesticides differ significantly from historic issues associated with mass mortality, long-term persistence and bioaccumulation. Potential ecotoxicological impacts associated with modern day synthetic, natural product and biological pest control agents are likely to be much more subtle and are commonly associated with indirect or secondary effects associated with habitat alteration, reduced food resources or multiple stress interactions. For glyphosate and Btk, respectively the dominant herbicide and insecticide used in the forest sector internationally, case study evaluations reviewed here support the conclusions of several more comprehensive risk assessments. Based on the weight of scientific evidence currently available, these data and risk assessments suggest that the judicious use of these products, in accordance with product labels, pose little risk to forest environments or non-target wildlife species. In contrast, higher tier field studies conducted with triclopyr ester and diflubenzuron, confirm specific risks under environmentally realistic or operational conditions imposing a requirement for mitigative actions sufficient to negate the risk. In such cases it is considered prudent to use adaptive management strategies, including operational chemical and biological monitoring of small scale operational test programs to ensure that mitigative actions do in fact protect sensitive values and general ecological integrity of receiving environments. A highly positive sidelight of detailed operational monitoring studies is the ability to generate both exposure and effects data critical to effective probabilistic risk analysis. Overall, case study analyses presented here support the continued judicious use of pesticides as part of sustainable forest management. In cases where risks are identified, appropriate mitigative measures may still allow them to be employed where no other effective options exist.
With emphasis that the scientific knowledge base associated with potential ecotoxicological effects of major forestuse pesticides is both extensive and detailed, there are, as always, some areas where further research would be considered particularly valuable. These focus areas include: (a) evaluation of potential interactive effects of tankmixed herbicides; (b) assessment of plausible multiple stressor interactions (e.g. chemical and concomitant drought stress); (c) investigations on impacts on key ecosystem functional processes; (d) development and application of cost-effective operational monitoring techniques applicable over broader spatial scales and longer time frames than typical empirical studies; (e) application of probabilistic analyses and; (f) risk assessments that include population modelling over larger spatial and temporal scales.
Relative to the available information base on forest-use pesticides, our scientific knowledge on potential impacts of alternative vegetation or insect pest control techniques is exceedingly weak. As noted by Scriber [131], all pest management programs carry some risk of negative environmental impacts; this includes the "do nothing" option. In general, it is inappropriate to assume that biological controls, natural pesticides or other non-chemical approaches pose no risks to ecological integrity of forest ecosystems. In fact there are several lines of evidence that clearly demonstrate this assumption to be invalid – see Thompson and Kreutzweiser [92] as but one example. It is imperative that all options with potential use in integrated pest management strategies be equally scrutinized against cost, effectiveness and environmental acceptability criteria. One, particularly valuable means of conducting such comparisons is through direct side-by-side multi-disciplinary field studies conducted at semi-operational or operational scales. Finally, from the ecological perspective alone, the potential deleterious effects associated with the "do nothing option" or with the use of ineffective options may in fact be greater than those associated with pesticide treatment or other pest control alternatives. As a generality there appears to be far too little scientific or policy attention paid to this aspect. Similarly, there may also be significant economic implications of weakly effective or non-intervention strategies. Multiple cases of exotic invasive plant or insect pests as currently extant in
### *Ecological Impacts of Major Forest-Use Pesticides Ecological Impacts of Toxic Chemicals* **105**
the North American forest sector and elsewhere around the globe are unfortunately providing demonstrable evidence supporting the point that ineffective or non-intervention options are often not acceptable in either ecological or economic terms. Within an integrated management strategy, those options which best meet the three fundamental criteria of efficacy, economics and environmental acceptability should be made available and used by resource managers in optimizing the twin goals of sustainable resource use and protection of ecological integrity. All organisms, including humans, as integral components of forests and other global ecosystems are ultimately dependent on the successful achievement of those goals.
# **ACKNOWLEDGEMENTS**
The author would like to acknowledge the financial contributions and support of the Canadian Forest Service, Department of Natural Resources Canada in making this chapter possible and to thank S. Holmes, D. Kreutzweiser, M. Coppens and two anonymous reviewers for their very helpful review and suggestions for improving the draft manuscript.
# **REFERENCES**
© 2011 The Author(s). Published by Bentham Science Publisher. This is an open access chapter published under CC BY 4.0 https://creativecommons.org/licenses/by/4.0/legalcode
# **CHAPTER 6**
# **Impacts of Pesticides on Freshwater Ecosystems**
**Ralf B. Schäfer1,\*, Paul J. van den Brink2,3 and Matthias Liess4**
*1 RMIT University, Melbourne, Australia; <sup>2</sup> Alterra - Wageningen University and Research Centre, Wageningen, The Netherlands; <sup>3</sup> Department of Aquatic Ecology and Water Quality Management, Wageningen University, Wageningen, The Netherlands and 4 UFZ – Helmholtz Centre for Environmental Research, Leipzig, Germany*
**Abstract:** Pesticides can enter surface waters via different routes, among which runoff driven by precipitation or irrigation is the most important in terms of peak concentrations. The exposure can cause direct effects on all levels of biological organisation, while the toxicant mode of action largely determines which group of organisms (primary producers, microorganisms, invertebrates or fish) is affected. Due to the interconnectedness of freshwater communities, direct effects can entail several indirect effects that are categorised and discussed. The duration of effects depends on the recovery potential of the affected organisms, which is determined by several key factors. Long-term effects of pesticides have been shown to occur in the field. However, the extent of the effects is currently uncertain, mainly because of a lack of large-scale data on pesticide peak concentrations. In the final section, we elucidate the different approaches to predict effects of pesticides on freshwater ecosystems. Various techniques and approaches from the individual level to the ecosystem level are available. When used complementary they allow for a relatively accurate prediction of effects on a broad scale, though the predictive strength is rather limited when it comes to the local scale. Further advances in the risk assessment of pesticides require the incorporation and extension of ecological knowledge.
# **INTRODUCTION**
Modern agricultural practices rely on the usage of synthetic pesticides (mainly herbicides, fungicides and insecticides) in order to prevent losses by pests [1]. The global pesticide production reached significant levels after the Second World War and rose sharply from approximately 500,000 t/a in the 1950s to over 3 million t/a at the beginning of the 21st century [2]. This trend will probably continue over the next decades because of a demand for higher food production as the human population increases, monocultural production for biofuels and potentially introduction of new pests in many areas associated with climate change [2, 3], though introduction of pest-resistant plants and an increase in organic farming and integrated pest management may counter this trend. Given the large amounts of pesticides applied globally and given the fact that they are designed to harm biota, there is a high potential for adverse environmental effects also on non-target communities [4]. When pesticides enter freshwater ecosystems, they do interact with the biotic and abiotic components of the ecosystem. Abiotic factors can lead to degradation (photo-decomposition by sunlight or hydrolysis by water) or adsorption of the compounds on sediment or organic matter. The interaction with the biotic parts comprises uptake, metabolisation and accumulation in organisms, which in turn may lead to adverse affects on the freshwater biotic community. These adverse effects are the topic of this chapter and will be delineated in depth after a brief overview of the entry routes of pesticides in freshwater ecosystems.
Once released into the environment, pesticides can be subject to airborne and waterborne entry in aquatic ecosystems. Airborne processes encompass wind drift during pesticide spraying (spray drift) and volatilisation after application with subsequent atmospheric transport that may lead to the deposition of compounds in remote ecosystems (thousands of kilometres) from their initial application. For example, organochlorine insecticides such as dichloro-diphenyltrichloroethane (DDT) and lindane (-HCH), which exhibited high usage patterns from the 1950s to the 1970s have become ubiquitous in the environment due to their high environmental persistence and potential for long-range atmospheric transport [5, 6]. Organochlorines are at present even detected in the polar regions, although they were never applied there [6]. However, for the majority of currently used pesticides atmospheric transport is confined to regional translocations within a 300 km radius as they are less persistent and have a lower potential for long-range transport [7, 8]. The waterborne translocation of compounds is driven by precipitation events or irrigation. Precipitation and irrigation can wash compounds from the field into adjacent surface waters via runoff or subsurface flow or into the groundwater.
**Francisco Sánchez-Bayo, Paul J. van den Brink and Reinier M. Mann (Eds) © 2011 The Author(s). Published by Bentham Science Publishers**
**<sup>\*</sup>Address correspondence to Ralf B. Schäfer:** RMIT University, Melbourne, Australia; Present address: Institute for Environmental Sciences, University Koblenz-Landau, Landau, Germany; Email: [email protected]
The quantitative relevance of the exposure route (airborne or waterborne) varies depending on physicochemical properties of the compound as well as the geographical, geological, hydrological and climatic conditions and crop type. For Germany, a modelling study on the exposure routes estimated the input of 65%, 10% and 25% of diffuse pesticide load by field runoff, flow in drainage channels and spray drift, respectively, though this study did not include atmospheric transport [9]. Spray drift is greatest when the spraying is conducted aerially using aircraft and for crops such as vineyards or orchards where the spraying occurs in a horizontal direction [10]. However, several studies emphasised the relevance of the waterborne exposure route concerning pesticide concentrations as reviewed in [11]. A study in North Germany showed that concentrations in a small headwater stream were elevated by several orders of magnitude during heavy rainfall events in the pesticide spraying period [12]. For cotton in Australia, the endosulfan concentrations in a creek and a river were approximately 10-fold higher during runoff events than as a consequence of spray drift [13]. Similar observations were made in the Lourens River in South Africa with runoffassociated pesticide concentrations at the first strong rain event after pesticide application being approximately 50 fold higher for two pesticides compared to the spray drift concentrations [14]. In irrigation farming of rice in Japan, short-term peak concentrations of pesticides in adjacent water bodies occurred in association with heavy rain or irrigation events [15, 16]. Overall, for water bodies in agricultural areas, intensive rainfall (> 10 mm per day) or irrigation with consequent runoff and subsurface flows after pesticide applications is recognised as the most important route of entry, both resulting in episodic short-term peak pesticide concentrations. In rivers that are fed by agricultural tributaries, the exposure may be more continuous due to dilution and overlapping input from different tributaries, but is still seasonal.
The pattern of episodic peak concentrations has to be considered in investigations on the effects of pesticides on aquatic biota. Although a large number of studies has been conducted on the effects of pesticides in surface waters, the majority failed to clearly link effects to exposure, partly because the study design did not include a sufficient quantification of pesticide peak concentrations [11]. Hence, an exposure monitoring method should be adopted that captures runoff events in the spraying period. Automatic continuous water sampling or automatic event-triggered water sampling can be employed, but these methods are cost- and labour-intensive [12, 17, 18]. Therefore, alternative methods have been suggested such as a less costly version of an event-triggered water sampler [19], a suspended matter sampler for particle-associated hydrophobic pesticides [20, 21] or passive sampling using adsorbent membranes if continuous background exposure is absent or negligible [22]. To sum up, the determination of pesticide peak concentrations in the water bodies is crucial in studies on the effects of pesticides on freshwater organisms.
# **DIRECT EFFECTS OF PESTICIDES**
The freshwater community consists of different groups of organisms such as fish, amphibians, invertebrates, plants or microorganisms. Pesticides can have direct and indirect effects on these organisms. Direct effects are caused by the physiological action of a pesticide within an organism. However, the biotic community is characterised by ecological interactions between species such as competition or predation and indirect effects refer to effects mediated via these interactions [23]. For example, mortality of water fleas as a direct result of exposure to a pesticide (direct effect) may lead to an increase of algae biomass due to a release from grazing pressure (indirect effect). We will firstly focus on direct effects as they are a prerequisite for understanding the indirect effects.
In general, direct effects of chemicals on an organism depend on the concentration; i.e. the dose determines the poison (Paracelsus). However, some further general factors influence the occurrence and magnitude of adverse effects:
These factors can, together with the biological and physicochemical characteristics of the exposed ecosystem, modify the potential direct effect of a pesticide entering a freshwater ecosystem. Therefore, the effects may differ between ecosystems with different modalities of these factors. For example, a certain concentration of a pesticide will most likely have stronger effects on a community of a forest stream that never received pesticide input in comparison to an agricultural stream that is subject to recurring pesticide exposure [42]. However, our knowledge about the comparative relevance and the differences in the modalities of these factors between natural ecosystems is very limited and does not allow for a generalisation of their influence on effects. Nevertheless, when studies are conducted within similar ecosystems (e.g. agricultural streams) or the factors vary only slightly between the sites, the pesticide concentration (or the derived toxicity) should be the most important predictor of the pesticide effect. Indeed, in field studies on the effects of pesticides on invertebrate communities in agricultural streams, most of the variability in community endpoints could be explained by pesticide concentration (or the derived toxicity) alone even across biogeographical regions [43-45]. The period and duration of exposure as well as the history of the communities were relatively similar in these studies. Nevertheless, the amplification of the effects of pesticides through these factors is one explanation why field effects have been reported at levels lower than expected based on laboratory or artificial stream experiments [42, 44]. Moreover, these factors are presumably most important when results from laboratory toxicity tests are extrapolated to the real world.
Pesticides can act on different endpoints that can be classified according to the level of organisation in biotic communities:
These levels are linked mechanistically in a hierarchical bottom-up order that can be described as a pyramid effect (Fig. **1**) [61, 62], and we outline this conceptual model as follows. A toxicant always acts on the physiological level of an organism first. Subsequently this may lead to individual-level effects such as delayed development or mortality. If the effect is strong and several individuals are impacted, the effect can propagate to the population level (e.g. the population growth rate may change). In cases of severe contamination, some or many populations or even whole groups of organisms (e.g. crustaceans) may go extinct and the niches may be occupied by other species. Thus, a change in community composition would be observed. For example, freshwater sediment microbial communities were exposed to a combination of fungicides, insecticides and herbicides, which lead to an alteration in bacterial community composition [59]. Finally, whether the specific functions of the affected species can be compensated for by other species (*i.e.* the degree of functional redundancy of the ecosystem [63]) ultimately determines the occurrence of an effect on an ecosystem process (e.g. breakdown of organic matter, primary production, nutrient recycling). In general, effects on all subjacent levels are a prerequisite for effects on higher levels of the pyramid (Fig. **1**). However, effects on ecosystem processes may occur without visible effects on the population or community level. For example, some studies in artificial ecosystems demonstrated that although the effects of herbicides on the rate of photosynthesis could be measured, no population effects could be detected among populations of primary producers [64, 65]. In most studies with artificial ecosystems, however, effects on ecosystem processes resulted from community and population level changes [66].
Conceptually, effects on lower levels of the pyramid (Fig. **1**) generally occur at lower concentrations than effects on higher levels since higher levels integrate several individuals or populations, of which some will be more tolerant. Thus, higher levels of biological organisation have a higher tolerance for pollution. Moreover, the effects on the different levels of organisation are not necessarily temporally synchronised as it may take from some hours to days for a suborganismic effect to affect individuals and from days to weeks for individual effects to affect populations, communities and ecosystems [61]. Studies on the effects of the pyrethroid fenvalerate demonstrated that population and community disturbance lagged behind and persisted while physiological or individual effects were not detectable anymore [36, 67].
Direct effects of pesticides have been reported on almost all biological endpoints for all groups of freshwater organisms. However, most studies were conducted in the laboratory whereas the focus of this book lies on effects in ecosystems [68]. Therefore, we mainly consider effects that have been observed under field or field-relevant conditions; *i.e.* either detected in field studies or in artificial ecosystems. In addition, we only include direct effects that were reported in the last two decades, so that the findings represent effects of currently used pesticides. Nevertheless, we briefly portray the effects on the aquatic wildlife that were reported since the 1950s and stimulated the writing of the historical book "Silent Spring" by Rachel Carson [69].
The significant agricultural use of organic pesticides, mainly organochlorine insecticides, started after the Second World War [70]. DDT, in particular, was used globally in large amounts for mosquito control and agricultural pest control (70-80% of total DDT used), reaching an annual use of approximately 400,000 t in the 1960s [71]. Reports of effects on wildlife and humans occurred soon after the widespread introduction of organochlorine insecticides [70, 72]. In the aquatic ecosystems, runoff of organochlorine insecticides following rain events in adjacent streams lead to severe fish kills and the eradication of the stream invertebrate fauna over stretches of several kilometres [70, 73]. Also aquatic and terrestrial birds in sprayed regions succumbed to lethal doses [72, 74]. More surprisingly, effects on the reproduction of several fish-eating birds were observed that comprised thinning of the egg shells resulting in the eggs being crushed during nesting, abandonment of nest and egg-eating by the parents [5]. The failure to reproduce affected bird colonies, resulting in population declines in species such as the brown pelican (*Pelicanus occidentalis*), great blue heron (*Ardea herodias*) and herring gull (*Larus argentatus*) [5]. Reproductive effects were even observed in regions with relatively low environmental exposure to DDT and occurred as a consequence of biomagnification of this highly persistent compound along the food chain, so that fish-eating birds at the top of the aquatic food web were finally exposed to biologically effective concentrations [5]. For example, after spraying of the Clear Lake in California for gnat-control, local populations of the western grebe (*Aechmophorus occidentalis*) declined, exhibiting fat concentrations of up to 100,000-fold the lake water concentrations, 400-fold the plankton concentrations and 200-fold the small fish concentrations in fat [75].
In the 1970s several persistent compounds (e.g. DDT, endrin, dieldrin) were banned in most countries due to their unacceptable effects on wildlife [76]. In agriculture, organochlorine insecticides were mainly substituted by the less persistent chemical families of pyrethroid and organophosphate insecticides. Although generally less widespread, effects on non-target organisms were frequently reported in the 1980s. Between 1977 and 1984 approximately 56% of 128 fish kills in the United States were attributed to pesticide pollution, primarily to the organochlorine endosulfan and the organophosphate malathion [77]. Historically a trend can be observed from compounds with a broad mode of action affecting many non-target species to compounds with a specific mode of action that are less toxic for the majority of non-target organisms. Currently used insecticides such as pyrethroids are characterised by a 1000-fold lower toxicity for mammals compared to organochlorine pesticides [78, 79]. For example, an accidental input of the pyrethroid cypermethrin lead to the complete eradication of invertebrates over a stretch of 3 km but no fish kill was observed [80]. Nevertheless, given that there are usually some organisms in the freshwater community that are physiologically related to terrestrial pest species (e.g. insects), present and future pesticide use is likely to continue posing a threat to aquatic ecosystems. The effects of pesticides used in the last two decades on the different groups of organisms are summarised in Table **1**. Effects were reported for most groups and endpoints, though there are some notable differences. Field studies that show effects on macrophytes, phytoplankton and benthic algae as well as other microorganisms are scarce and are almost entirely limited to artificial ecosystem experiments. By contrast, several field studies have demonstrated effects in freshwater ecosystems on macroinvertebrates and zooplankton [45, 68, 81-85], fish [86] and amphibians [52, 87-90]. Here, the frequency of reported effects on macroinvertebrate and zooplankton assemblages is much higher than for fish and amphibians, for which only a few field studies reported effects at the population or community level. In the case of amphibians this is not surprising as they mainly appear in lentic (standing) surface waters, which generally receive less pesticide input than lotic (running water) habitats [91].
**Table 1:** Effects of pesticides on the different groups of organisms under field or field-relevant conditions reported in the last two decades: frequency of reported effects, field relevance and examples.
a See first column for abbreviations: none: no studies; low: 1-5 studies; medium: 5 to 10 studies, high: > 10 studies
The dominance of reports on effects on macroinvertebrates and zooplankton followed by fish and amphibians compared to macrophytes, phytoplankton and benthic algae raises the question whether this is: a) an artifact of organism selection in biomonitoring; or b) observed impacts represent the real frequency of effects in these organism groups. There are several reasons that suggest an over-representation of effects on animals and specifically on macroinvertebrates and zooplankton:
We therefore conclude that the present picture concerning effects of pesticides on ecosystems is likely biased towards the fauna and especially towards invertebrates. In the next section we examine to what extent the biotic community is affected by different classes of pesticides.
# **Compound-Specific Effects on Different Groups of Organisms**
The early days of ecotoxicology were driven by the myth of a "most sensitive species" that could be used as a standard test organism to predict the impacts of toxicants in ecosystems; *i.e.* no effects should occur in the ecosystem as long as no direct effects occur in the most sensitive species [101]. This quest for the "most sensitive species" relied on several assumptions, one being that the most sensitive test species for a set of compounds would be most sensitive to other compounds as well. This assumption has received extensive criticism and we will give a brief overview of studies with pesticides that contradict this assumption. Concerning herbicides, a study of van den Brink *et al.* [99] showed that the sensitivity of test species varies with the chemical class of the compound and that there is no single most sensitive primary producer. For example, the common duckweed *Lemna minor* was more sensitive to the herbicides diquat and linuron but less sensitive to diuron and metamitron than the green algae *Selenastrum capricornutum* [99]. Similarly, crustaceans are among the most sensitive species to organochlorine and pyrethroid insecticides, whereas the recently introduced neonicotinoid insecticides exhibit orders of magnitude higher toxicity to insects than to crustaceans [102]. In addition, the sensitivity of a species to a toxicant depends on the life stage and a sensitivity ranking of several species may therefore in some cases vary with the selected life stage [24].
However, this does not mean that patterns of sensitivity are completely stochastic, a broad classification into sensitive and tolerant taxa is possible when compounds are grouped according to their mode of action [103, 104]. In the case of pesticides, some substances have specific sites of action; e.g. photosynthesis or ergosterol synthesis that are only present in certain groups of organisms [105] (see Chapter 1). For example, pesticides that target the hormonal reproductive system of insects are unlikely to affect aquatic primary producers, which have a completely different reproduction system [106]. Other pesticides have a wider activity and target processes that are present in all or many organisms (e.g. cytochrome oxidase [105]), which complicates the prediction of effects.
There are some general rules of thumb on the sensitivity of the groups of aquatic organisms to herbicides, insecticides and fungicides. Herbicides are mainly targeting organisms that perform photosynthesis. A meta-study of artificial ecosystem studies on herbicides showed that for this class of compounds, primary producers are more sensitive than aquatic animals [65]. Which group of primary producers (*i.e.* macrophytes, algae or microorganisms) is most sensitive depends on the mode of action of the herbicide [99].
A meta-study on insecticides showed that for this class of compounds, macroinvertebrates, zooplankton and fish are the most sensitive groups compared to other organisms [66]. Given that several currently used insecticides are designed to target invertebrates [1], macroinvertebrates and zooplankton are presumably more susceptible than fish. In fact, 16 currently used insecticides for which comparative toxicity data was available, exhibited highest toxicity to invertebrates and zooplankton [107]. In addition, several artificial ecosystem studies support this hypothesis since fish were less sensitive than invertebrates [108].
Fungicides are less extensively studied than herbicides and insecticides. Therefore sound generalisations about which groups are most sensitive to these compounds are problematic. Nevertheless, based on the mode of action, several fungicides should be most toxic to aquatic microorganisms, especially aquatic fungi [109]. An unpublished evaluation of experiments in artificial ecosystems did not indicate elevated toxicity by fungicides to the aquatic fauna and primary producers [109].
Recent studies have began to incorporate these differences in the modes of actions and proposed a new strategy for the search of sensitive species. A study of Wogram & Liess [103], extended and confirmed by von der Ohe & Liess [104], found that for macroinvertebrate species the variability of the sensitivity to organic chemicals is higher between taxonomic groups (primarily families and orders) than within groups and could be pooled in a relative toxicity ranking. For example, stoneflies were among the most sensitive taxa for organic chemicals, whereas gastropods are relatively tolerant to organic chemicals [104]. This study, however, did not distinguish toxicant mode of actions between organic chemicals and some studies demonstrated that such differences exist. Rubach *et al.* [110] showed that the relative sensitivity of macroinvertebrate taxa differed, albeit minor, between organophosphates, pyrethroids and carbamates. In addition, imidazole fungicides were more toxic to gastropods than many insect taxa [111, 112]. Furthermore, the differences in sensitivity ranking were even stronger for heavy metals and salinity [104, 113]. No relative sensitivity rankings have been established for other groups of organisms so far. Overall, we propose that with regard to toxicants with a similar mode of action a consistent, relative sensitivity hierarchy may be established at least for the different groups of organisms that can be used for risk assessment of pesticides, though no universal most sensitive species or group of species exists.
# **INDIRECT EFFECTS IN THE AQUATIC COMMUNITY**
In ecosystems, species interact with other species and their abiotic environment. Direct effects of pesticides on a species can alter these interactions and therefore have an indirect (also termed secondary) effect on species that are otherwise not directly affected. The following ecological relationships may lead to indirect effects via propagation of direct effects:
Indirect effects of pesticides have been reported frequently and have been summarised in different publications [65, 66, 114-116]. Here, we will give a brief overview on potential indirect effects of herbicides and insecticides. Fig. (**2)** sketches the direct and indirect effects of a herbicide and an insecticide in a freshwater ecosystem. As outlined in the last section, primary producers are generally at highest risk of being adversely impacted by herbicides. A reduction of the primary producers can lead to a decrease in the herbivore populations due to food limitation and/or habitat loss (Fig. **2**). For example, in a study on the effects of the herbicide atrazine on freshwater communities in artificial ponds, growth and reproduction of zooplankton (e.g. *Simocephalus serrulatus, Daphnia pulex*) decreased as a consequence of phytoplankton biomass reduction [117]. Similarly, amphibian tadpole biomass (e.g. *Rana catesbeiana*) decreased due to a reduction of food source (periphyton) and loss of macrophyte habitat (e.g. *Typha* *latifolia, Chara* sp.) [116]. The indirect effect on herbivores as a consequence of a reduction in food may be less pronounced in nutrient-rich ecosystems in the field. Where a reduction in herbivores occurs, effects may subsequently propagate to higher trophic levels; e.g. predators that prey on the herbivores (Fig. **2**). For example, herbicide-induced reductions in zooplankton and macroinvertebrates (e.g. Chironomidae spp.) due to loss of primary producers as food and habitat resulted in a decreased total biomass of bluegill sunfish (*Lepomis macrochirus*) [116]. This ecological effect chain represents bottom-up indirect effects because the lowest trophic level (primary producers) determines the effects on higher trophic levels. While these indirect effects are due to predatory ecological relationships, competitive relationships between primary producers promote an increase of tolerant primary producers when sensitive competitors are eliminated by a pesticide (Fig. **2**). For example, algal blooms of *Chlamydomonas* sp. were observed after linuron strongly reduced macrophyte populations of *Elodea nuttallii* [118].
**Figure 2:** Schematic representation of direct (solid line) and indirect (dashed and dotted lines) potential effects of pesticides in freshwater ecosystems. See text for further explanation.
Photosynthesis is an important ecosystem process that influences water quality, and the inhibition of photosynthesis by herbicides results in a lower concentration of dissolved oxygen (DO) and lower pH values during daytime (Fig. **2**). In an indoor mesocosm study, the highest linuron treatments of 50 and 150 µg/L reduced DO and pH by up to 40% and 25%, respectively [119]. If a herbicide causes acute mortality of macrophytes, decomposition by decomposers may further enhance the reduction in pH and DO concentration [65]. This deterioration of water quality can then have detrimental impacts on sensitive invertebrate species and this represents a case of indirect effects resulting from species-habitat relationships (Fig. **2**). For example, a strong reduction of cladoceran and copepodan populations was partially attributed to a reduction in DO to only 20% compared to controls, following a 10 mg/L contamination with hexazinone in lake enclosures [120].
The schematic effects of the insecticide in Fig. (**2)** illustrates a combination of top-down and bottom-up indirect effect [121]. We present a scenario, demonstrated by many studies [66], where an insecticide adversely affected invertebrates and zooplankton species. The bottom-up indirect effect is represented by the decrease of fish population density due to a reduction of invertebrate prey. For example, a significant reduction in macroinvertebrates such as ephemeropterans (mayflies) and dipterans as well as two zooplankton groups (Daphniidae and Cyclopidae) in outdoor ponds after treatment with methyl parathion led to decreased mean weights in rainbow trout (*Salmo gairdneri*) [122]. The topdown indirect effect commences with a release of primary producers from grazing pressure and may result in growth of their populations. In a mesocosm study on the effects of chlorpyrifos, the eradication (Insecta and Amphipoda) and reduction (Isopoda, Cladocera and Copepoda) of parts of the invertebrate community by the pesticide resulted in a twoto three-fold increase in periphyton chlorophyll-*a* accompanied by a bloom of *Oscillatoria* sp. [123]. Some tolerant invertebrate and zooplankton species may profit from the reduced competition with directly affected sensitive invertebrate species (Fig. **2**). In the aforementioned study on chlorpyrifos, Sphaeriidae molluscs and herbivorous rotifers (*Polyartha* sp.) increased as a consequence of reduced competition for food with more sensitive invertebrates [123]. Similar observations were made in a field study where, after pesticide exposure, sensitive species decreased and tolerant species increased [43]. The effects on ecosystem processes are ambiguous in this scenario. While the increase of primary producers increases the pH value and DO concentration, the decomposition of dead invertebrates/zooplankton by fungi or bacteria decreases these water quality parameters [66]. We assume that in larger freshwater systems and lotic systems the first mechanism would be more important. In the case of a strong reduction of the invertebrate fauna another important ecosystem process, leaf-litter decomposition can be inhibited [124]. A field study in 16 French streams demonstrated a three- to five-fold decrease in leaf-litter decomposition in streams with an insecticide-impaired invertebrate community [44]. Leaf-litter decomposition represents an important energy source in stream ecosystems and a reduction can even adversely impact river sections several kilometers downstream, since they rely on particulate organic matter input from upstream sections [125, 126]. Hence, indirect effects can occur a long way from the location of the direct effect.
Pesticides represent only one of the many disturbances (e.g. floods, droughts, land-use change, acidification, dredging *etc.*) that shape freshwater ecosystems [127] and other disturbances can also result in indirect effects [114]. The similarity of indirect effects of different disturbances depends on the disturbance type and selectivity of their effects on biota [128]. While some disturbances such as floods also occur in pulses, they presumably act less selectively on the trophic levels or groups of organisms in the biotic community; e.g. they are unlikely to only affect primary producers or invertebrates. By contrast, many of the currently used pesticides are relatively selective *i.e.* they act on a specific trophic level or group of organisms as outlined above. However, similar indirect effects have been reported for other contaminants such as pulses of heavy metals or accidental discharges of organic toxicants in freshwater ecosystems [114].
Indirect effects of pesticides have mainly been studied in artificial ecosystems because under field conditions a clear differentiation between direct and indirect effects is more difficult. In the field, the time and magnitude of pesticide input driven by precipitation events is unknown and the input usually comprises a mixture of pesticides that may directly affect several groups in the biotic community concurrently. Furthermore, the variation regarding environmental parameters, pesticide exposure and biotic community composition is usually rather high between sampling sites. Hence, the detection of indirect effects would require an extended Before-After-Control-Impact (BACI) sampling design [129] consisting of a spatially and temporally highly replicated monitoring of the biotic community in control and impacted sites before the pesticide input and immediately after the contamination to determine the direct effects and then a few times in weekly intervals to identify indirect effects. However, the efforts of such a study would be jeopardised by ignorance of the impact of the monitored runoff events. In the worst case scenario, any of the selected sites would be impacted (compare [44] where no impacts were detected in a field study in Finland). Even if direct and indirect effects could be detected using multivariate statistical techniques [130-133], causality could not be inferred and additional studies under standardised conditions would be needed [134]. The field studies conducted to date on the effects of pesticides did not aim to differentiate between direct and indirect effects. This may explain the lower effect thresholds that have been observed in field studies compared to artificial ecosystems [43, 44, 135], because the effects on invertebrates may not have resulted from direct toxicity of pesticides alone but as well may be a bottom-up indirect effect from depletion of primary producers or heterotrophic microorganisms. An alternative approach for assessing the direct and indirect effects in the field integrates ecological modelling [136]. The modelling is used to predict direct effects of an exposure event and the differences in the effects that are observed in the field are considered as indirect effects. However, currently used models do not allow for an integration of all the factors that can influence the strength of direct effects (see section "Direct effects of pesticides") and therefore gives rise to high uncertainty.
# **EFFECT, DURATION AND RECOVERY DYNAMICS**
Given the widespread application of pesticides, it is almost inevitable that some fraction enters freshwater ecosystems. Consequently, the authorisation for use of pesticides involves the passing of a value judgement on the question: "Which ecological effects are unacceptable?" [137]. For regulators in the European Union, long-term field effects on populations and communities are deemed unacceptable, though the operationalisation of "long-term" may vary on a case-by-case basis [138]. Similarly, the US EPA includes the assessment of recovery from pesticide stress in their risk assessment framework and regards potential irreversibility (*i.e.* permanent changes in the community structure or ecosystem processes) as an adverse effect [139]. From these perspectives follows that transient shortterm effects are considered acceptable. The underlying hypothesis is that a toxicant can have only transient effects on an ecosystem and that the ecosystem may subsequently recover to an initial or reference state; *i.e.* the community recovery principle [140]. This hypothesis has been subject to criticism. Landis *et al.* [39] argue that because of the dynamic nature of ecosystems, any pesticide-induced effects are irreversible, rejecting the concept of recovery. Even if recovery can be observed on one level of biological organisation, changes may persist on other levels; e.g. the gene pool can be impoverished [141]. For example, fish populations of the brown bullhead (*Ameiurus nebulosus*) in the Great Lakes have been observed to have different genetic structures in populations that have been exposed to a mixture of organic toxicants and metals [142]. The US EPA acknowledges this criticism by defining recovery as "the return of a population or community to some aspect(s) of its previous condition" [139]. We agree with the criticism of Landis *et al.* [39], but would rather integrate it in the evaluation of studies on recovery; e.g. by studying effects on suborganismal endpoints in affected populations. Hence, we advocate the use of studies on the recovery of an affected artificial ecosystem or field ecosystem as a useful tool to deliver information on the toxicity of a pesticide that can be used to evaluate the acceptability of effects.
So far, almost all studies on the recovery of an ecosystem from the effects of a pesticide were conducted in artificial ecosystems and, in general, community composition was selected as the endpoint. Complete recovery was assumed when significant differences between treated and non-treated communities were not detected anymore. While controlling the type I error rate (reject the null hypothesis that recovery occurred when it is true), this entails the risk of a type II error (fail to reject the null hypothesis that recovery occurred when it is false) that may be more interesting in studies on recovery. Unfortunately, most studies on artificial ecosystems lack an analysis of the probability of a type II error of the selected test, which may be relatively high given that sample sizes in these studies are usually low (< 5 replicates per treatment). Therefore, the time to complete recovery is probably underestimated.
A meta-analysis of artificial ecosystem studies with various insecticides highlighted that the initial acute toxic effect of the substance is a critical factor for the time to recovery [143]. Methodologically, the toxicity of different compounds can be compared using the Toxic Unit (TU) approach [144], in which the results of laboratory toxicity experiments for a specific test organism (usually LC50) are used as a benchmark to compare the toxicity of concentrations of different compounds. The TU is given by
$$\text{TU} = \frac{c\_{\text{i}}}{\text{LCS0}\_{\text{i,j}}}$$
where c is the concentration of compound i and j the benchmark organism. The benchmark organism should be selected according to its sensitivity for the study compounds. In the studies considered here, *Daphnia magna* or, in very few cases, a fish species (*Pimephales promelas, Oncorhynchus mykiss, Lepomis macrochirus*) was selected as standard test organism to compute the TU for insecticides. For herbicides, green algae (*Scenedesmus subspicatus, Selenastrum capricornutum, Chlorella vulgaris*) or macrophytes (*Lemna* spp.) were employed for TU calculation. For reasons of simplicity, throughout this chapter we use TU*Daphnia*, TU*fish* and TU*primprod* for the TUs based on *Daphnia magna*, fish and primary producers, respectively. Note that the TU approach assumes concentration addition *i.e.* the same concentration-response relationship for compounds, while this may differ in reality; e.g. one compound has no effects at a TU of 0.01 (1/100 of the LC50) while another compound may still have effects due to a flatter concentration-response curve [145].
In a meta-analysis of 26 artificial mesocosm studies with acetylcholinesterase-inhibiting insecticides and 18 studies with pyrethroid insecticides, no long-term community effects (> 8 weeks) were observed for a TU*Daphnia* < 1 for fish, microorganisms and primary producers [108]. Even with higher compound concentrations relating to a TU*Daphnia* between 1 and 100 only 4 of 15, 2 of 22 and 2 of 27 observations indicated long-term effects for fish, microorganisms and primary producers, respectively (observations with unknown recovery excluded). By contrast, freshwater insects, macrocrustaceans and microcrustaceans (Ostracoda, Cladocera and Copepoda) exhibited clear long-term effects above a TU*Daphnia* of 0.1 and even lower for pyrethroids, where concentrations between a TU*Daphnia* of 0.1 and 0.01 caused long-term effects in aquatic insects (1 of 10 observations) and macrocrustaceans (2 of 5 observations) [108].
A similar meta-analysis was performed for artificial ecosystem studies with photosynthesis-inhibiting herbicides, auxin-simulating herbicides and growth-inhibiting herbicides [65]. No long-term effects (> 8 weeks) were observed for molluscs over the whole range of tested concentrations (up to a TU*primprod* of 100). For zooplankton, long-term effects were reported in 3 of 16 cases with a TU*primprod* > 1. Long-term effects on fish and amphibians occurred in 4 of 13 cases at concentrations relating to a TU*primprod* > 0.1. For macrocrustaceans and insects no long-term effects were detected, except in 1 of 3 observations on auxin-inhibiting herbicides – but at a TU*primprod* of 0.01. All these effects were most likely indirect effects that resulted from the depletion of populations of primary producers or from the associated habitat degradation [65].
Phytoplankton and periphyton showed clear long-term effects for concentrations with a TU*primprod* > 1, and in 1 of 8 cases with a TU*primprod* between 0.1 and 1, long-term effects were reported for phytoplankton. Macrophytes were more sensitive, with clear long-term effects in several studies above a TU*primprod* of 0.1, with 2 of 5 observations on auxin simulators indicating long-term effects between a TU*primprod* of 0.001 and 0.1.
To sum up, based on artificial ecosystem studies, long-term effects may occur when the concentrations exceed concentrations relating to a TU*primprod* and TU*Daphnia* of 0.01 for insecticides and herbicides. This seems to be in general agreement with two field studies on 20 streams in North Germany and 29 streams in Spain where the macroinvertebrate community exhibited long-term alteration at a TU*Daphnia* of similar magnitude [43, 135].
Only a few studies have scrutinised the duration of effects that are classified as long-term and they only focused on effects on macroinvertebrates. Recently, an artificial stream ecosystem study with the neonicotinoid insecticide thiacloprid reported the persistence of adverse effects on sensitive macroinvertebrate species half a year after a pulse exposure with a TU*Daphnia* of 0.014 [146]. An artificial pond ecosystem study demonstrated that at concentrations associated with a TU*Daphnia* of 20, the invertebrate communities of control ponds and treated ponds were still significantly different after 2 years whereas there was recovery at lower concentrations [67]. Very high concentrations of the insecticide methoxychlor (TU*Daphnia* of 10,000), which may occur from direct spraying of water bodies (e.g. mosquito control), implicated a different community in the treated stream compared to a reference stream over 5 years in a field study [84]. In the before mentioned study on 20 streams in North Germany, no full recovery of the community was observed within one year for a TU*Daphnia* > 0.001 [43]. Relating these studies to generation times of invertebrates, which usually range from a few weeks to a year, illustrates that the recovery time is in the range of one to a few generations. Overall, these studies suggest that recovery in community endpoints can take over one year, and up to several years in cases of very high concentrations.
# **Factors Fostering Community Recovery Processes**
The duration of community recovery from pesticide stress depends to some extent on the magnitude of the effect which in turn is determined by the concentration, exposure duration and toxicity of the pesticide and its transformation products [143, 147]. However, some other factors also influence the time to recovery:
Ecological traits of species in the affected biotic community: In particular a short generation time, high reproduction rate, presence of resistant life stages and a high dispersal capacity of species augment recovery of populations [148]. For example, phytoplankton species generally recover faster from adverse effects of pesticides than macrophytes due to shorter generation times [65]. Similar observations were made for invertebrates with a short generation time [146, 149, 150] or high dispersal capacity [43, 151, 152]. Finally, microorganisms are presumably less vulnerable to pesticides due to short generation times and adaptability [41], though there may be exceptions; e.g. aquatic hyphomycetes [109].
To sum up, the recovery time of a freshwater community from pesticide stress is influenced by ecological, physicochemical, geographical and temporal factors.
# **How Frequent are Long-Term Effects Under Current Use Patterns?**
In the previous sections we outlined the concentration levels that may cause long-term effects in the field. This raises the question, how frequently these exposure concentrations occur in the real world. For insecticides, Schulz [11] reviewed the concentrations given in field studies since 1982 and reported the maximum and minimum concentrations detected for each compound in each study. We calculated the respective TU*Daphnia* for the observed maximum concentrations for this data using the LC50 for *Daphnia magna* as given in the Pesticide Manual [165] in order to allow for a comparison with the long-term effect thresholds derived above. In the 64 studies, 162 of the 194 compounds measured, comprising 39 different insecticides, had concentrations above the limit of quantification. For these 162 observations, the TU*Daphnia* associated with the maximum concentrations exceeded 0.01 in 94 cases. Hence, in 58% of the observations in the respective field studies, the substances exhibited maximum concentrations that may cause long-term effects. To put this into the right context, one has to consider that: 1) some compounds without detections may not have been reported, hence the number of observations is limited to positive detections; 2) the study regions were not randomly selected but presumably based on some prior knowledge on pesticide pollution; 3) each of the 162 observations amalgamated up to 29 sampling sites and several sampling episodes; and 4) some regions with insecticide detections were sampled repeatedly [11]. However, a recent study on 83 pesticides in 17 agricultural streams over 4 years in the United States also reported that between 10% to 25% of the samples exceeded a TU*Daphnia* of 0.01 [166]. In addition, approximately 50% of the concentrations in the US streams exceeded a TU*primprod* of 0.01. By contrast, the TU*fish* were rather low (most TU*fish* << 0.01). Since no event-driven water sampler was employed in this study, the peak concentrations were most likely underestimated. Thus, the reported TUs represent a conservative estimate of the real exposure. Overall, the results confirm our conclusion that invertebrates and primary producers are at highest risk of being affected by pesticides and suggest that long-term effects of pesticides on both groups are not isolated cases.
# **RISK ASSESSMENT AND PREDICTION OF EFFECTS OF PESTICIDES**
The beginning of widespread pesticide use in the middle of the 20th century was soon followed by reports of detrimental effects on ecosystems and human health [69]. Hence, today most countries require a pesticide risk
## *Ecological Impacts of Major Forest-Use Pesticides Ecological Impacts of Toxic Chemicals* **123**
assessment for ecosystems and human health before authorisation of a substance is granted. The risk assessment procedure comprises a fate and an effects assessment. The fate side uses models and experimental data to assess the exposure in the environment (see chapter 2 in this book). In this chapter, we focus on methods to assess the effects; *i.e.* we give an overview of the different methods used to predict effects on aquatic ecosystems and describe their advantages and disadvantages.
Earlier, we mentioned that every effect of a pesticide has a physiological basis. However, the current science of ecotoxicology is very distant from a "grand unifying theory" that would mechanistically integrate all levels of biological organisation (Fig. **1**) and allows for the prediction of effects on the top levels from the lower levels. This holds true especially for the linking of suborganismal effects to higher levels. Currently, a clear link between responses at the suborganismal level and the fitness of individuals is still missing but would be a prerequisite for a sound suborganismal endpoint to be considered in risk assessment [167, 168]. Hence, although appealing from a precautionary principle point of view, suborganismal effects are at present no valid basis to predict effects on populations, communities or ecosystems. Currently, the approaches for ecological risk assessment of pesticides rely predominantly (1) on the individual (single-species laboratory tests) and ecosystem level and (2) on experiments. In the following we will sketch these as well as some alternative approaches.
# **Methods Relying on Toxicity on the Individual Level**
The individual level has been the starting point of ecotoxicological research and still represents an important backbone supporting research on other levels of organisation. In fact, the vast majority of ecotoxicological data that has been produced to date (*i.e.* EC50 and LC50 data), originates from single-species toxicity tests. These tests allow for high replication and deliver relatively precise estimates of the toxicity endpoints (e.g. mortality or growth) under standardised conditions (temperature, water quality, age of test organisms *etc.*) [169]. For the first tier in pesticide risk assessment in a regulatory context, the acceptable concentration for a compound in the environment is derived by dividing the LC50 by a safety factor to account for uncertainties in the extrapolation from a single species in the laboratory to communities in the field. The uncertainties arise from abiotic and biotic factors that are not considered in single-species laboratory tests but may significantly modify the susceptibility of populations in the field such as ecological relationships within ecosystems [35, 170], recovery processes [36] and multiple stressors [32, 171]. In the European Union, a safety factor of 100 for acute toxicity tests for the invertebrate *Daphnia magna* and fish is applied to account for the above mentioned uncertainties [172]. The results from algal growth inhibition tests and chronic toxicity tests are divided by a safety factor of 10 to obtain the threshold concentration that should not be exceeded for the first tier in pesticide risk assessment. However, the use of this approach to predict effects in the field is relatively inefficient as it can be over- or underprotective [173, 174]. Underprotection of freshwater ecosystems can lead to losses of species and ecosystem services while overprotection may put unnecessary constraints on economic activities. Nevertheless, the single-species test is far less labour- and time-consuming than experiments on higher levels of biological organisation and are, therefore, an indispensable tool in the first tier risk assessment. Moreover, the results from these tests are critical for other research areas such as ecotoxicological modelling [175, 176], trait-based risk assessment [42, 94] or assessment of mixture toxicity [144, 177].
Alternative approaches to current single-species tests on the individual level can be categorised into: 1) those which use different methods to generate identical endpoints; and 2) those which generate different endpoints to assess the risk. Approaches of the first category include the use of different species or assays for testing and computational methods to predict toxicity data. Especially in the case of vertebrate testing, ethical concerns have promoted the development of alternatives such as the fish embryo test [178] or more recently the cell line test [179]. Another experimental development represents the rapid tolerance test that is a response to the scarcity of toxicity data for many compounds and species [113, 180]. Rapid testing involves simultaneous toxicity testing with field-collected taxa and sacrifices some precision in the determination of the toxicity endpoint in order to generate toxicity data on a wide array of species representative of natural communities [180].
A non-experimental method to obtain toxicity data is modelling. Quantitative structure activity relationship (QSAR) models represent a promising method to predict acute or chronic toxicity data [181, 182]. Here, the structure of compounds with known toxicity is used to predict the toxicity of unknown compounds. So far, QSARs have most successfully been applied to differentiate between narcotic and excess toxicity [183]. However, their value is currently tenuous for compounds where the activity relies on exotic or unknown functional chemical groups. Moreover, several ecological models examine the influence of different test conditions on the determination of acute or chronic toxicity to allow for the adjustment to the respective field conditions [184]. For example, Ashauer *et al.* successfully (77% to 96% of explained variance) incorporated fluctuating and episodic concentrations of a pesticide using a threshold damage model to predict effects of realistic exposure conditions on the invertebrate *Gammarus pulex* [185], whereas standard toxicity tests utilise a constant or pulsed exposure. Finally, individual-level models can be used to analyse results from acute toxicity tests and explore mechanisms. For example, the energy budget model describes characteristics of individuals such as growth, metabolism or reproduction in terms of energy budgets [186] and the most commonly used energy budget model in ecotoxicology is DEBtox (see: http://www.bio.vu.nl/thb/deb/deblab/debtox/). However, most ecological models require toxicodynamic and toxicokinetic data that is available for a few species only and are therefore not widely applicable.
Species-sensitivity distributions (SSDs) present an alternative approach of the second category (generation of different endpoints) that was introduced to generate more accurate environmental quality targets [187]. SSDs integrate the results of single-species tests for a respective compound (or a mixture of compounds) to establish a statistical distribution of the sensitivity. The distribution function links the fraction of potentially affected species to the concentration of a compound. This allows for the derivation of a threshold concentration that is assumed to protect a defined percentage of taxa in the community (usually 95%). SSDs have received attention by regulators and are now used for the setting of environmental threshold concentrations in several countries including the US [188], the Netherlands, Australia and New Zealand [189, 190]. For herbicides and insecticides, two studies compared the thresholds derived with SSDs to effect concentrations observed in artificial ecosystem studies [99, 107]. For herbicides, SSDs based on chronic no-effect concentrations (NOEC) delivered threshold concentrations that were protective for artificial ecosystems, except for one out of nine compounds [99]. Similarly, the threshold concentrations derived from SSDs for 16 insecticides were protective for artificial ecosystems, though not in cases with repeated insecticide exposure where a safety factor of at least 5 was suggested [107]. However, species in artificial ecosystems are often more tolerant than natural communities [146] and therefore it remains to be demonstrated that SSDs are also protective in the field. Furthermore, the accuracy of SSDs for the prediction of effects in natural ecosystems has been questioned because they typically rely on the results of a few test species that are often not representative of natural communities and like single-species tests do not incorporate ecological relationships, multiple stressors or recovery processes [191]. In addition, their application range is limited due the scarcity of available toxicity data for many compounds [180]. In fact, toxicity data are often restricted to a few test species [135], while a minimum of 15 to 55 taxa have to be included in SSDs to arrive at thresholds with acceptable confidence limits [192], though as few as six taxa can be sufficient to derive protective concentrations [99]. The aforementioned rapid tests have been advocated as an experimental solution for the lack of species data [180]. Similarly, the use of, 1) expert judgement regarding the sensitivity of higher taxonomic groups (e.g., orders) combined with Bayesian statistical methods [193], and 2) statistical techniques such as interspecies correlation models [194, 195], allow for the construction of SSDs despite sparse data. Overall, SSDs represent a powerful tool to extrapolate individual level toxicity date to the community level, but do not consider ecological relationships so that the accuracy of the prediction is contentious.
# **Methods Relying on Toxicity on the Population level**
Currently, impacts at the population level do not receive a lot attention in regulatory pesticide risk assessment [169]. However, since risk assessment is most interested in predicting effects on populations or communities, which are constituted of populations of many species, several authors have suggested that population level endpoints should inform risk assessment [196, 197]. Population level experiments can include ecological factors such as recovery, intraspecific competition and different life stages, and studies demonstrated that single species acute toxicity tests are a poor predictor for effects at this level [36, 198, 199]. The experiments can either be conducted for several populations in a community context such as artificial ecosystems that are discussed in the next section or for a single population. Yet, no consensus has been established on the experimental conditions in single population toxicity testing. More importantly, single population experiments are more labour and time-consuming and exhibit higher variability compared to single species tests [169], while not including ecological inter-species relationships and indirect effects. Therefore, advocates have focused on population level mathematical modelling for the prediction of effects [196]. The models can be broadly categorised into demographic models and individual-based models (for a more thorough treatment of ecotoxicological population models see [196, 200]). Demographic models globally assign parameters such as fecundity or growth to age classes or the whole population and derive population level endpoints, which is most commonly the population growth rate. For example, in a study of the effects of diquat bromide on the bluegill (*Lepomis macrochirus*), the different age classes were parameterised with survival probabilities and fecundities and the model was used to assess the effects on the population growth rate [201]. Moreover, demographic models can be coupled with energy-budget models such as DEBtox to translate effects from the individual level to the population level [202]. By contrast, individual based models describe each individual in a population independently and the effects on the population level emerge out of the interaction and responses of the individuals when exposed to pesticides. These models can deliver new insights into mechanisms of toxic effects on populations and also allow for an explicit incorporation of the spatial dimension. For example, Van den Brink *et al.* [156] used an individual-based model to predict the effects of an insecticide on populations of the isopod crustacean *Asellus aquaticus* under different spatial scenarios. The results highlighted the relevance of habitat connectivity and the dispersal abilities, such as drift, of the species for its subsequent recovery in the impacted water body [156].
The major problem for ecotoxicological modelling on the population level remains the paucity of ecological data for other than the few better studied species, which hampers the parameterisation of models. This limits their value for prediction of pesticide effects. Nevertheless, they may be used to investigate effect mechanisms of pesticides in populations (and higher levels) to extrapolate empirical results from the individual level to the more meaningful population level. In addition, by modelling taxa based on generalised ecological traits such as generation time or dispersal capacity, population models can be valuable to select sensitive species or groups of species for the inclusion in artificial ecosystem experiments or identify indicator taxa in biomonitoring [149, 203].
# **Methods Relying on Toxicity on the Community and Ecosystem Level**
The most frequently applied experimental method to predict effects of pesticides on the community level is ecotoxicological testing in replicated artificial ecosystems. Artificial ecosystem studies encompass different sizes and are accordingly differentiated into macrocosms, mesocoms and microcosms [204]. So far, most of the studies have been conducted in replicated mesocosms. Since mesocosms represent at least a part of an ecosystem, we do not draw a distinction here between the community level and the ecosystem level. In addition, mesocosm test systems possess all characteristics of real ecosystems such as ecological interactions, recovery processes, and depending on the experimental design, multiple stressors, though their configuration may be different.
**Figure 3:** Effect duration of a neonicotinoid insecticide in mesocosms for short-living and long-living taxa. Asterisks indicate significant (p <0.05, ANOVA, confirmed by both Games–Howell and Tamhane post-hoc tests) differences from the controls. Cdt = canonical coefficient for treatment d and week t. Modified and reprinted from [146] with permission from Elsevier.
Mesocosms can be constructed for both lentic (ponds) and lotic (stream) freshwater ecosystems. The construction of mesocosms starts with the physical containment structure, which is subsequently furnished with substrate and water. After allowing some time for stabilisation, sediments, plants and animals from natural ecosystems are introduced [204, 205]. The freshwater community should be established and replicate mesocosms should be similar in community composition before an experiment is run. Mesocoms can be constructed indoors or outdoors, with outdoor mesocosms being more realistic with exposure to field environmental conditions (rain, sun) but at the same time subject to higher variability due to seasonal variation and the risk of freezing during winter in temperate regions.
In general, experiments in mesocosm systems represent close-to-field conditions and include most biotic and abiotic factors that can influence the effects of a pesticide. Mesocosms have the advantage over field monitoring in that many factors can be controlled; the timing, duration and concentration(s) of exposure can be manipulated and the statistical power is higher since the abiotic factors between the replicated units are similar. Results from mesocosm experiments can be regarded as relatively accurate to predict effects in the field (effect thresholds) compared to other experimental methods and are therefore used as the highest tier in pesticide risk assessment. Nevertheless, there are some reasons why caution is warranted when predicting effects in the field from mesocosm experiments:
The derivation of general concentration-response relationships that could be used in a predictive manner from field studies in freshwater ecosystems is difficult. This is because most field studies do not cover an exposure gradient as they are limited to a few streams or ponds (sample size < 10) and/or were not designed to deliver a regression. In addition, the causality between pesticide exposure and effects is often not clear [11]. In fact, field investigations on the effects of pesticides face two severe problems that hamper the establishment of a confident concentrationresponse relationship [42]. Firstly, the natural variation between communities at field sites is high, whereas reference sites often differ from pesticide-disturbed sites by more factors than only pesticide exposure [43]. Secondly, pesticide input during runoff events is associated with co-occurring changes in other environmental variables such as an increase in current velocity resulting in hydrological stress or increased turbidity that may confound effects of pesticides. We will outline three approaches to tackle natural variation and/or confounding factors. The first approach is to experimentally test the influence of potential confounding factors. Liess and Schulz [210] constructed a bypass microcosm system connected to an agricultural stream and compared the effects of runoff with and without pesticide contamination on the dominant stream invertebrate populations. Only runoff events containing insecticides caused a significant decrease in the invertebrate populations [210]. In another study, the observed concentration range of environmental factors such as pesticide exposure and turbidity was investigated for effects on test species in laboratory experiments [211]. Since the test species showed only significant acute effects in response to pesticide contamination, the authors concluded that pesticides were the main cause of observed effects on invertebrate assemblages in the field [211, 212]. However, this approach is still confronted with natural variation when used to derive a concentration-response relationship.
A second approach to deal with natural variation and confounding factors represents the usage of ecological traits such as generation time or dispersal capacity to identify effects of pesticides [43]. The underlying ecological theory is that the sensitivity of taxa to a stressor and the occurrence of subsequent recovery patterns and indirect effects is dependent on their configuration of traits [161]. For example, species with a stream-lined body are more tolerant to hydrodynamic stress [213]. Hence, stressors can be regarded as a filter that selects taxa with a suitable trait configuration which results in an increase of these traits in the community [161]. Liess and von der Ohe [43] hypothesised that invertebrate taxa with a long generation time, low dispersal capacity, presence in water bodies during time of pesticide application and high physiological sensitivity would be most susceptible to pesticides and predicted a decrease of these "species at risk" (SPEAR) in the communities (see http://www.systemecology.eu/SPEAR/Start.html for online SPEAR calculator). In fact, they demonstrated a decrease of the fraction of sensitive taxa during the time of pesticide application in 20 streams in North Germany. A similar study was conducted in two regions of Finland and France and found a reduction in sensitive taxa with an increase in pesticide toxicity measured in terms of TU*Daphnia* [44]. Finally, an analysis of monitoring data comprising 28 tributaries in a Spanish river basin confirmed the relationship between pesticide input and decrease of sensitive taxa [135]. All studies established a significant relationship between pesticide toxicity and response of the sensitive taxa in the communities, or more technically speaking, between TU*Daphnia* and percentage of SPEAR species. Interestingly, the three concentration-response relationships were not significantly different and allowed for the derivation of an approximate effect threshold [42-44, 135]. According to these studies, slight effects occur already above a TU*Daphnia* of 0.001 and strong effects prevail above a TU*Daphnia* of 0.01 [44, 135]. This effect threshold lies approximately a factor of 10 below the effect threshold derived from mesocosm studies. The differences may be due to one of the following explanations:
Currently there is no consensus as to which explanation is most plausible. While some scientists argue that pesticide concentrations are usually underestimated in field monitoring [82] and consider the first explanation as the most likely, others emphasise the short-comings of mesocosm studies, which we have discussed earlier, and advocate explanation 2 or 3. However, if explanation 1 were true, this would mean that the different sampling methods (event-driven, grab sampling, passive sampling) employed in the three field studies (in some studies even in parallel) [22, 43, 44, 135] are subject to the same systematic error. More field studies would be needed to scrutinise this issue. Overall, the results from the abovementioned studies represent at least an accurate prediction of the order of magnitude at which effects may occur in the field. Nevertheless, the concentration-response relationship should be interpreted with care, when applying to, 1) areas with a different spectrum of applied pesticides, 2) lotic ecosystems, and 3) larger freshwater systems. In addition, the results can not be extrapolated to organisms other than invertebrates in the freshwater biotic community.
A third approach to tackle natural variation and confounding factors is the use of field-based sensitivity estimates. One method is similar to the SSD approach described before but uses sensitivity estimates from large field data sets to predict thresholds. For example, a Norwegian dataset with 4200 sampling sites was used to assess the individual sensitivity of the frequently occurring and abundant taxa to toxicants and subsequently construct a field species sensitivity distribution (f-SSD)[214]. As with SSDs, the f-SSDs can be used to predict thresholds that should protect a certain fraction of the taxa in the community [214]. Recently, a new method was proposed by Kefford *et al.* [215] using dissimilarity indices to assess changes in the species pool across a contamination gradient. The method was applied to a larger data set on invertebrate data for 360 streams in North Germany and found a significant change in species composition with increasing modelled pesticide exposure [215, 216]. These methods are presumably more accurate in the prediction of effect thresholds than conventional SSDs relying on single species toxicity data. However, to apply this approach, large field monitoring data sets with concurrent pesticide measurements and biomonitoring data are required, which are very rare and presumably only available for governmental monitoring programs [135]. To date, governmental pesticide monitoring programs relied mainly on point water samples and were not adapted to detect episodic events such as pesticide runoff [82]. Hence, the predictions derived from these methods are likely less accurate than those from field studies using event-driven water samplers (see first section of this chapter).
# **Ecosystem Modelling**
While many mechanistic models have been developed to predict the fate of contaminants in the environment, only a few mechanistic models target the effects of toxicants in freshwater ecosystems. These mechanistic effect models have only rarely been applied in ecotoxicology and are not thoroughly validated and compared to field data [217]. In one of the few published studies, Sourisseau *et al.* [217] calibrated and validated the ecosystem model Aquatox (http://www.epa.gov/waterscience/models/aquatox/) for control streams in a mesocosm experiment. The model was very sensitive to temperature parameters such as the optimal growth temperature for periphyton, filamentous algae and predatory invertebrates. Similarly, the optimal temperature for fish was a very sensitive parameter in another application of this model on the bioaccumulation of polychlorinated biphenyls [218]. Overall, these mechanistic models do not currently represent an alternative to other methods of prediction of effects and given the lack of ecological data for many species it is questionable if an adequate model can be used for the prediction of effects in the near future.
An alternative approach to mechanistic effect models is represented by statistical models that extrapolate observed concentration-response relationships from lower levels of biological organisation or from case studies to a larger scale. For example, Schriever *et al.* [219] combined a mechanistic fate model with a statistical model incorporating the concentration-response relationship observed in two central European regions to predict effects of pesticides on the European level [220]. However, due to data limitations and simplifications to allow for a wider prediction, such models do not provide accurate predictions of the effects, neither for specific pesticides nor on the small scale. Another example is the PERPEST model that predicts the magnitude and duration of effects of a certain concentration of a pesticide (and mixtures of pesticides) on various community endpoints simultaneously (e.g. community metabolism, phytoplankton and macro-invertebrates) [176, 221] and relies on a database containing results from freshwater mesocosm studies. A key advantage of PERPEST over single species/safety factor analyses is that it removes the need to extrapolate to the community level. However, the premise that the concentration-response relationships observed in a limited number of field case or mesocosm studies can be extrapolated to a wide range of freshwater ecosystems implicates uncertainties in the accuracy of the predictions. Nevertheless, due to the current limitations of mechanistic models and paucity of field studies, the statistical approach is certainly useful to identify potential hot spots of pesticide pollution and compounds of concern.
# **ACKNOWLEDGEMENTS**
The authors like to thank Ben Kefford, Peter von der Ohe and three anonymous reviewers for valuable comments that helped to improve the quality of the manuscript. Special thanks to Mikhail Beketov for providing Fig. (3). RBS received financial support (SCHA 1580/1-1) from the German Science Foundation (DFG).
# **REFERENCES**
© 2011 The Author(s). Published by Bentham Science Publisher. This is an open access chapter published under CC BY 4.0 https://creativecommons.org/licenses/by/4.0/legalcode
# **Ecological Impacts of Organic Chemicals on Freshwater Ecosystems**
# **Paul K. Sibley1,\* and Mark L. Hanson2**
*1 School of Environmental Sciences, University of Guelph, Canada and <sup>2</sup> Department of Environment and Geography, University of Manitoba, Canada*
**Abstract:** The ecological impacts of organic pollutants on freshwater ecosystems have attracted immense scientific, regulatory, and public attention over the past fifty years. In part, this reflects the significant role that freshwater ecosystems play as a repository for anthropogenic chemicals relative to other systems. Some of the most severe ecological impacts have been documented in freshwater ecosystems from persistent organic pollutants (POPs) such as polychlorinated biphenyls, polychlorinated dioxins and furans, and polycyclic aromatic hydrocarbons. Such chemicals can reside for long periods in freshwater sediments, which can then constitute a continual source to the environment even when direct inputs have ceased. Exposure of freshwater biota at lower trophic levels to persistent chemicals can result in transfer to, and ecological impacts at, higher trophic levels through bioaccumulation and biomagnification. In contrast to historically significant organic pollutants, the pervasive nature of new pollutant classes (e.g. pharmaceuticals, polybrominated diphenyl ethers, and perfluorinated surfactants) in global freshwater ecosystems is beginning to be recognized but the full spectrum of their ecological impacts is poorly understood. In this chapter we review documented and potential ecological impacts of organic chemicals in freshwater ecosystems. We focus predominantly on effects at the population, community, and ecosystem levels but, to the extent that our understanding of impacts at these higher levels is predominantly extrapolated from information derived at lower levels, we also include information at the organism and sub-organism level. In addressing each chemical class, impacts on microbial, plant, invertebrate, fish, and fish-eating bird populations are considered where data exists.
# **INTRODUCTION**
The conceptual basis of the ecosystem, *i.e.* a system that includes living organisms interacting with each other and the inorganic components, was defined by Tansley [1]. The hierarchical and interconnected flow of energy and cycling of materials within an ecosystem lead to the concept of trophic structure [2] and subsequent bioenergetic studies of ecosystem development, including relationships between biota (food webs), diversity, and nutrient cycles [3]. From this pioneering work emerged the concept of hierarchical levels of organization within ecosystems (Fig. **1**). This hierarchical framework may be instructive in understanding the ecological impacts of contaminants; that effects at a given level of biological organization can propagate upward, or cascade downward, to other levels [4]; and that effects at one level can be understood mechanistically from information derived at lower levels in the hierarchy and interpreted ecologically from information derived from higher levels [5]. However, the greater the distance between any two levels, the more difficult it is to establish cause-effect relationships and, coupled with the non-linearity of many trophic relationships, clear examples of effects propagating from sub-individual levels to higher levels of organization are rare.
Current regulatory structures for protecting aquatic ecosystems typically rely on extrapolation of data derived from lower levels of biological organization (e.g. whole-organism toxicity tests) as these are often the only data available for criteria-setting. This situation stands in stark contrast to the protection goals of regulatory authorities, and the fundamental premise of ecological risk assessment (ERA), which is the protection of populations and communities. To bridge the gap between regulatory practice and the protection goals of ERA, much effort has been expended on understanding how ecosystems respond to contaminants at different levels of biological organization. Depending on the intensity and duration of exposure to organic contaminants, ecological impacts in the field may include avoidance, extirpation/extinction, loss of diversity and function, and, under severe situations, ecosystem collapse. Such large-scale impacts were observed in Lake Erie in the 1950s and in Great Lakes lake trout and fish-eating bird populations in the 1960s and 1970s [6]. Of course, ecosystems can recover when ameliorative action is taken as was witnessed in Lake Erie in the 1960s when phosphate inputs were reduced. Although examples of population collapse, community disruption, and ecosystem impacts in the field exist, unequivocal cause-effect relationships
**<sup>\*</sup>Address correspondence to Paul K. Sibley:** School of Environmental Science, University of Guelph, Ontario, Canada N1G 2W1; Email: [email protected]
between individual contaminants and ecological effects at higher levels of biological organization are difficult to establish due to high biotic/abiotic complexity.
**Figure 1:** Schematic representation of the hierarchical organization of biological systems
Some understanding of the potential impacts of organic chemicals at higher levels of biological organization can be extrapolated from studies using non-chemical stressors to manipulate whole ecosystems [7, 8]. However, the practical, and arguably ethical, difficulty of manipulating whole ecosystems, communities or populations to understand the ecological impacts of contaminants limits public and scientific acceptance of such approaches. This may be partially overcome by conducting manipulative studies in model aquatic ecosystems (e.g. micro/mesocosms) and considerable knowledge about contaminant impacts at higher levels of biological organization have been derived through the use of such systems. Models can also help to understand potential ecological impacts of contaminants at higher levels of biological organization. Complex ecosystem simulation models have been developed (e.g. Comprehensive Aquatic Systems Model for understanding the impacts of pesticides) but have not been applied extensively because of the large number of explicit/implicit assumptions needed for parameterization, the large amount of data required about the fate and effects of the chemical(s) in an ecosystem, and difficulties related to model validation [4]. Better success has been met with population models and these are commonly applied in ERA [9, 10].
In this chapter, we review the potential ecological impacts of organic contaminants on freshwater ecosystems, focusing on microbes, plants, invertebrates and vertebrates. The chapter is organized by contaminant class, including those such as polychlorinated biphenyls, polychlorinated dioxins/furans, polycyclic aromatic hydrocarbons, and plasticizers (alkyphenol ethoxlylates, bisphenol A) with a long historical presence in the environment and those, such as polybrominated diphenyl ethers, fluorinated surfactants, and pharmaceuticals with a much shorter history. We exclude pesticides as these are covered in Chapter 6 of this book. The scope of this review is largely restricted to population, community, and ecosystem levels of biological organization, but we draw on information from lower levels of biological organization as needed.
# **HALOGENATED AROMATIC HYDROCARBONS**
Halogenated aromatic hydrocarbons (HAHs) are a diverse class of organic chemicals, within which occur some of the most ubiquitous and toxicologically significant chemicals in aquatic ecosystems including: polychlorinated biphenyls (PCBs), polychlorinated dioxins and furans (PCDDs/PCDFs), and polybrominated diphenyl ethers (PBDEs). The unique physicochemical properties of HAHs including hydrophobicity, low melting points, high octanol-water partition coefficients (Kow), and low volatility reflect the unique properties of halogens, which can comprise a significant percentage of the molecular weight of these compounds. Halogens have high electronegativity, a measure of how strongly atoms attract and hold electrons, and therefore the strength of covalent bonds. Fluorine, chlorine and bromine have electronegativity values of 4.0, 2.0, and 2.8, respectively, which are among the highest in the periodic table. The strong covalent bonds formed by halogens impart high molecular stability and hence a strong propensity to persist in the environment. The environmental persistence of these compounds increases the probability of exposure for environmental receptors while their hydrophobic nature can result in bioaccumulation/bioconcentration and subsequent biomagnification. For many HAHs, the combination of persistence and hydrophobicity has left an indelible imprint on many freshwater ecosystems and yielded a long history of scientific, regulatory, and public scrutiny.
# **Polychlorinated Biphenyls and Polychlorinated Dioxins/Furans**
# *Background and Chemistry*
Polychlorinated biphenyls were introduced in the 1920s as cooling and insulating fluids for industrial transformers and capacitors, fluorescent light ballasts, and as hydraulic fluids in the automotive and related industries [11]. The chemical properties of PCBs that made them ideal for these applications include low flammability and electrical conductance and high thermal and chemical stability. PCBs contain between 1 and 10 chlorine atoms attached in various configurations to biphenyl and were primarily marketed by the Monsanto Corporation between 1930 and 1977 under the trade name Aroclor. There are a total of 209 PCB congeners but only 100 to 150 occurred in formulations that were used and are now ubiquitously dispersed in the global environment [11]. PCBs were first reported in herring gulls and eagles in the mid-1960s [12], and have since been consistently identified in, among other matrices, human and animal adipose tissue, breast milk, and freshwater and marine sediments [13]. Evidence of chronic toxicity in humans and widespread effects in the environment led to implementation of the final PCB ban rule by the EPA in 1979, prohibiting the manufacture, processing, distribution and use of PCBs. PCBs have now been banned for 30 years, but it is estimated that approximately 70% of the PCBs manufactured remain in the environment [14]. In 2001 PCBs were listed as one of the "dirty dozen" POPs under the Stockholm Convention.
In contrast to PCBs, PCDDs/PCDFs have no commercial value and largely occur as historical by-products of the manufacture of organochlorine compounds (e.g. PCBs and pesticides such as 2,4,5-T and pentachlorophenol), the incineration of chlorine-containing substances such as polyvinyl chloride, and chlorine-based bleaching of wood pulp to make paper. Polychlorinated dioxins and furans are also created naturally as pyrolytic by-products of volcanic and forest fire activity. Polychlorinated dioxins and furans contain 1 to 8 chlorine atoms, yielding 75 and 135 possible congeners, respectively. Of these, the 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) congener is the most toxic and PCDD/PCDF congeners having chlorine atoms in the 2,3,7,8 positions appear to be most toxic and bioaccumulative. Once accumulated, clearance of the 2,3,7,8 congeners is extremely low and biomagnification occurs through food chains, at a rate of 3 to 10-fold for each trophic level [15].
# *Ecotoxicology*
The environmental toxicology of PCBs and PCDDs/PCDFs has been well documented [12-14,16-18]. Animal studies indicate that PCBs and PCDDs/PCDFs cause teratogenic, mutagenic, carcinogenic, immunotoxic, and hepatoxic effects and both are known to disrupt endocrine and growth factor systems, including effects on the developing immune, nervous, and reproductive systems [18]. Mechanistically, the toxicity of PCBs and PCDDs/PCDFs is mediated through the aryl hydrocarbon receptor (AhR) which is present in jawed fish, mammals, reptiles, and birds but not primordial fish and invertebrates [19]. For this reason, invertebrates are generally insensitive to PCBs and PCDDs/PCDFs [20] and effects are most commonly observed only in biota of greater evolutionary complexity, although the degree of sensitivity varies considerably among species. Although much effort has been expended to identify cause-effect relationships between specific PCB and PCDD/PCDF congeners and ecological impacts at higher levels of biological organization, this has proven difficult because individual congeners exhibit varying degrees of toxicity and they are taken up and metabolized at different rates as they are passed up food chains [21] leading to different contaminant profiles in aquatic biota over time.
PCBs and PCDDs/PCDFs are hydrophobic. In aquatic ecosystems they may be accumulated through bioconcentration but given their hydrophobic nature and predominant association with sediments and lipids, bioaccumulation through dietary sources is probably more significant. PCBs and PCDDs/PCDFs are generally metabolized and eliminated slowly from tissues so they not only accumulate but also increase in concentration as they are passed up food chains. Consequently, these chemicals are most commonly found in highest concentrations in predatory species at the top of food chains as documented in the Great Lakes and Arctic regions. In the Great Lakes, PCB concentrations have declined in water, sediments and biota by up to 95% from peak concentrations in the mid-1970s but they remain sufficiently high that they continue to be a major cause of fish consumption advisories [22].
PCBs and PCDDs/PCDFs appear to be relatively non-toxic to freshwater microbial communities at environmental concentrations. Salizzato *et al.* [23] found that the maximum concentration (0.90 g/g) of PCBs extracted from contaminated sediment, a value well above those typically detected, was below the limit of sensitivity of the Microtox assay. Numerous studies have demonstrated that PCB and PCDD/PCDF congeners can be degraded by microbial communities in contaminated freshwater sediments [24-27]. Aerobic degradation typically involves attack of the carbon ring and subsequent metabolism of the molecule [24] while anaerobic degradation occurs through reductive dechlorination [25]. In general, the rate of dechlorination depends on the relative number and position of chlorine atoms on the molecule and generally decreases with an increase in the number of chlorine substituents. Although rates of metabolism attained in laboratory studies are often high, in situ removal rates are generally exceptionally slow due to poor bioavailability and mass transfer [28]. Nonetheless, it is common to observe population-specific increases in abundances and shifts in microbial community structure in PCB or PCDD/PCDF-contaminated sediments [27].
There is limited information on bioaccumulation and toxicity of PCBs and PCDDs/PCDFs for freshwater plants and algae [29, 30] and evidence of population and community-level effects is rare. Patterson *et al.* [31] evaluated the historical response of diatoms and chlorophytes in sediment cores from a PCB-contaminated freshwater lake. During the period of maximum contamination (estimated peak bioavailable sediment concentrations were 0.5 g/L), minimal changes were observed in both diatom and chrysophyte assemblages. They hypothesized that the bioavailable fraction of PCBs in lake sediments was too low to cause detrimental effects in the limnetic phytoplankton communities. Kostel *et al.* [32] found that periphyton in a laboratory stream system accumulated up to one order of magnitude greater concentrations than sediment. Periphyton community structure shifted from a diverse diatom-based community to one co-dominated by fewer types of cyanobacteria. Yockim *et al.* [33] estimated bioconcentration factors up to 2083 for 2,3,7,8-TCDD for the freshwater alga *Oedegonium cardiacum* but did not indicate if effects on growth or population size occurred. Residues of up to 7000 ng/g in freshwater macrophytes (*Ceratophyllum* and *Elodea* spp.) were measured in a 30-day mesocosm study on 2,3,7,8-TCDD but no adverse effects were reported [34]. These limited data indicate that significant impacts of PCBs and PCDDs/PCDFs on plants and algae are unlikely at current environmental concentrations but they may serve as an important source of these compounds to higher trophic levels.
Information on the toxicity of PCBs and PCDDs/PCDFs in freshwater invertebrates is also relatively limited. Dillon and Burton [35] found that exposure to some PCB congeners killed 47 to 83% of freshwater fishes and invertebrates after 24 to 48 h at concentrations that were several orders of magnitude higher than those encountered under field conditions. However, most of the PCB congeners tested produced negligible mortality. In the freshwater cnidarian, Hydra aligactis, Adams and Haileselassie [36] estimated LC50s of 5 and 20 mg/L, and found bud regeneration was inhibited at 1 and 4 mg/L, for Aroclors 1016 and 1254, respectively. In these and other studies, effects occurred at concentrations much greater than those measured in freshwater sediments. West *et al.* [37] observed no toxicity in full life cycle exposures with the midge Chironomus tentans and the oligochaete Lumbriculus variegatus up to 9533 ng/g lipid of 2,3,7,8-TCDD. TCDD also had no effect on development and reproduction in the freshwater snail Physa sp. [38, 39], the water flea Daphnia magna [39, 40], the mosquito Aedes aegypti [38], and the aquatic oligocheate Paranais sp. [38]. In contrast, Ashley *et al.* [41] estimated LD50s between 0.03 and 1.5 ng/g body weight for 2,3,7,8-TCDD in a freshwater crayfish and toxicity was characterized by delayed mortality (15 to 40 days after treatment) and reduced activity.
Given the relative insensitivity of invertebrates to PCBs and PCDDs/PCDFs, evidence for population and communitylevel impacts is rare and often confounded by co-occurrence with other contaminants or non-contaminant factors. For example, De Lange *et al.* [42] found that sediment moderately contaminated by PCBs and PAHs affected the structure but not productivity of benthic macroinvertebrate communities, which they attributed to counteracting effects between contamination and an associated food surplus. Cooper *et al.* [43] compared two urbanized watersheds in Michigan USA, one whose sediments were heavily contaminated with PCBs, PAHs and metals compared to the other. They found fewer insect taxa, reduced invertebrate index of biotic integrity scores, and higher sediment toxicity in the more industrialized watershed. Although PCBs exceeded the probable effects level at one site, their contribution to the benthic community impacts appeared to be low relative to other contaminants.
Collectively, the weight-of-evidence indicates that significant population or community-level effects from PCBs and PCDDs/PCDFs in invertebrates are improbable at environmentally relevant concentrations. This conclusion is supported by the hazard assessment of Loonen *et al.* [44] who concluded that invertebrates experience reduced hazard relative to fish, fish-eating birds, and mammals. Mechanistically, this has been attributed to the absence of the Ah receptor in invertebrates. However, although effects of PCBs and PCDDs/PCDFs might not be predicted for most invertebrates based on receptormediated toxicity, one area that has not been evaluated extensively is potential multi-generational effects resulting from long-term, low-level exposures and this may warrant some consideration in future assessments of these compounds.
Because of their relative insensitivity, association with sediments and occurrence at the base of most food chains, algae, macrophytes and macroinvertebrates play a key role in the bioaccumulation and transfer of PCBs and PCDDs/PCDFs to higher trophic levels. As such, some have been proposed as reliable indicators of PCB contamination in freshwater systems [45] and numerous studies have investigated uptake, metabolism, and trophic transfer by invertebrates of PCBs [46-54] and PCDDs/PCDFs [55-57] in aquatic invertebrates. Zebra mussels (Dreissena polymorpha) accumulated PCB 77 from sediments, diet and water at a rate 10 times more efficient than Lampsilis siloquoidea, the mussel to which they are often attached [58]. Accordingly, high densities of zebra mussels likely influence PCB contaminant dynamics in Great Lakes ecosystems [59]. The freshwater crustacean Mysis relicta played a key role in PCB transfer from sediments into the Lake Champlain food web [60], and Sallenave *et al.* [61] showed that accumulation of PCB 153 in spiked plant material by downstream collectors was enhanced by the presence of both scrapers and shredders in stream mesocosms. Kidd *et al.* [62] estimated that PCBs were accumulated in lower trophic level organisms between 1000 and 100,000 times over surrounding water and sediment concentrations and Rasmussen *et al.* [63] showed that each trophic level contributed a 3.5-fold biomagnification factor for PCBs in Great Lakes lake trout. For TCDDs/TCDFs, Muir *et al.* [55] determined biota-sediment accumulation factors (BSAFs) of 24.6 and 18.6 for crayfish and mussels exposed to TCDF in an experimental lake mesocosm study and BSAFs ranging from 0.31 to 1.62 for uncaged aquatic insects exposed to pulp mill effluent. In laboratory exposures, Loonen *et al.* [56] determined BSAFs of 1.6 and 0.07 for TCDD and octachlorodibenzo-p-dioxin and Pickard and Clarke [57] determined BSAFs ranging from 0.04 to 2.42 for eleven TCDD/TCDF congeners in L. variegatus.
In the Great Lakes region, declines in populations of both fish and fish-eating birds have been causally linked to the presence of PCBs and PCDDs/PCDFs [64, 65] with corresponding though often poorly understood impacts at the community and ecosystem level [66]. Lake trout populations in the lower Great Lakes provide an excellent case study on the role that contaminants likely played in regulating population levels. Lake trout population declines began in the 1930s in response to increased fishing pressure, advancements in fishing technology (e.g. improved fishing line materials), increases in invasive sea lamprey populations, and changes in food web structure caused by invasive fish species [67]. By 1960, a virtual collapse of lake trout populations in the lower Great Lakes had occurred. Extensive restocking of fingerlings began in the 1950s and continues to this day but has met with poor success. Hypotheses to explain the slow recovery of Great Lakes lake trout include poor survival to spawning age after stocking and hence insufficient numbers to ensure successful annual recruitment, failure to locate natural spawning grounds due to loss of olfactory acuity in hatchery-reared fish exposed to inappropriate spawning substrates, changes in reproductive performance due to complex changes in prey fish population densities, and exposure to contaminants [67].
There is strong evidence supporting the role of contaminants, particularly PCBs and PCDDs/PCDFs, in lake trout population declines [68]. Initial evidence for the role of contaminants was provided by Mac *et al.* [69] who found that up to 97% of lake trout fry reared in hatcheries between 1978 and 1981 died when exposed to water from the upper Great Lakes. Further, a higher than expected frequency of blue sac disease, which can lead to the death of eggs, was observed in lake trout from Lake Ontario in the 1970s suggesting that maternal transfer of dioxin and dioxin-like compounds to the eggs may have been responsible for the effects. Numerous studies have since established causal relationships between early life stage mortality in lake trout and exposure to PCDDs/PCDFs/PCBs [70]. Although the precise cascade of events from initial exposure to early life stage mortality is not fully understood, as Ah receptor agonists, PCBs and PCDDs/PCDFs are known to affect reproduction and development via disruption of endocrine function [71]. Ankley and Giesy [65], using a weight-of-evidence approach, outlined a series of laboratory and field studies conducted throughout the 1990s on Lake Ontario lake trout which provide strong evidence to link PCDDs/PCDFs with lake trout population declines. These studies established that early life stage lake trout were exquisitely sensitive to PCDDs/PCDFs and other Ah receptor agonists, that these compounds were transferred maternally from adult fish to eggs, and that effects were consistent with observed pathologies, such as blue sac disease, in field-collected lake trout [70, 72-74]. Comparing the observed effects to residues of PCDDs/PCDFs and PCBs extracted from Lake Ontario sediment cores, Ankley *et al.* [48] concluded that the toxicity predictions are in excellent agreement with available historical data for lake trout population levels and suggest that evidence for recent improvement in natural reproduction is consistent with declining levels of persistent bioaccumulative chemicals in sediments and biota, a conclusion that is supported by Cook *et al.* [70].
With their position atop freshwater food chains, fish-eating birds often represent the most vulnerable group of organisms with respect to the effects of POPs. Population declines of fish-eating birds such as herring gulls, cormorants, and Caspian terns in the Great Lakes have been linked to contaminant-induced reproductive effects [75]. Keith [76] found that reproductive success in herring gulls in Lake Michigan was approximately one third of the normal rate and Gilbertson [77] found that reproductive success was about 10% of expected rates in nesting sites in Lake Ontario. Initially, reproductive failure in fish-eating bird populations was attributed to egg shell thinning resulting from exposure to the pesticide DDT but high rates (up to 30%) of embryo mortality among gull populations [66] and continued observation of developmental and reproductive abnormalities after the ban of DDT in the early 1970s indicated other chemicals were also contributing to population declines. Experiments in which eggs were transferred from clean to contaminated sites for herring gulls [78] and Foster's terns [79] found lower hatching success and behavioral anomalies in adults, thus supporting the contaminant-based theory of observed declines. Many studies documented numerous consistent symptoms including increased embryo and chick mortality, growth retardation, congenital deformities (*i.e.* cross-bill syndrome), feminization of embryos, and abnormal parenting behavior. These effects became known as the Great Lakes Embryo Mortality, Edema, and Deformities Syndrome (GLEMEDS; [80]), a syndrome previously documented in animal studies on known Ah receptor agonists.
Causal evidence that PCBs and PCDDs/PCDFs contributed to population declines in Great Lakes fish-eating birds was developed in the 1990s. Ludwig *et al.* [81] evaluated reproductive success in a population of Caspian terns in Saginaw Bay, Michigan following a one in a hundred-year flood in 1986 that released sediment-bound PCBs. PCBs accumulated rapidly in tern eggs, accounting for 98% of the toxic equivalents (TEQs), and concentrations in the eggs approached the lethal dose required to kill 95% of chicken embryos. The percent of chicks that hatched from first and second clutches the following year was 28% and 0%, respectively, compared to corresponding 5-year rates of 60% and 43% and the 3 year average hatchling deformity rate increased 163 times over the historic rate. Using a weight-of-evidence approach, Ludwig *et al.* [82] reviewed available data from studies on cormorant and Caspian tern populations around the Laurentian Great Lakes to test the hypothesis that deformities in embryos and chicks of these species were caused by contaminants measured as TEQs. Hatching deformities and abnormalities were comparable to those observed in chickens exposed to PCBs and dioxins and were correlated with concentrations of PCBs and TEQs, which were present at concentrations sufficient to cause the effects. Overall, they rejected the null hypothesis and concluded that there was a relationship between the incidence of deformities in both bird species and exposure to planar halogenated compounds measured as TEQs or total PCBs. Giesy *et al.* [83], also using a weight-of-evidence approach, concluded that lethality and deformities in embryos of colonial fish-eating Great Lakes birds were caused by multiple planar dioxin-like compounds which expressed their effects through a common mechanism of action.
With strong evidence for the role of contaminants in causing population declines in Great Lakes fish and fish-eating birds, effects at the community and ecosystem levels might be expected. For example, the primary route of exposure for affected bird populations is contaminated fish, so loss of predatory bird species might be expected to cause changes in fish, and possibly other organism, populations [80] *via* a cascade of trophic interactions. However, understanding trophodynamic changes and shifts in food web structure in the context of contaminants is difficult as non-contaminant drivers of change in ecosystem structure (e.g. introduction of exotic species) must also be considered. Some studies have attempted to address this complexity by examining expected shifts in contaminant profiles in food webs experimentally [49, 84] and *via* modeling (e.g. [51]) but empirical evidence of shifts in community structure and ecosystem function resulting specifically from POPs remains elusive [66].
Numerous studies have measured PCB and PCDD/PCDF residues in adult amphibian and reptile tissues but evidence for effects of PCBs and PCDDs/PCDFs at the population and community level is scant [85, 86]. A study at a PCB-contaminated site in Peducah, Kentucky found that tissue-borne PCBs were significantly higher in larvae than adults of various anuran species but found no evidence of adverse effects at the population level [87]. PCB tissue levels in frogs in five PCB-contaminated southwestern Michigan wetlands were lower than those in sediments and suggested that the apparent lack of effects on frog populations could be explained by limited contaminant accumulation [88]. Jung and Walker [89] estimated that embryos and tadpoles of green frogs (*Rana clamitans*), leopard frogs (*R. pipiens*), and American toads (*Bufo americanus*) are 100 to 1000-fold less sensitive to TCDDinduced lethality than most fish species. Reeder *et al.* [90] observed a significant shift in sex ratios favoring males, and increased prevalence of intersexuality, in field populations of cricket frogs (*Acris crepitans*) and concluded that the evidence suggested a strong association between population declines and TEQ body burdens. They suggested that amphibian populations could be affected at environmentally relevant concentrations; however, in most studies effects occur at concentrations significantly higher than those typically measured in water and sediments. Based on their work with *X. leavis*, Levine *et al.* [91], suggest that the apparent insensitivity of anurans to dioxins reflects low affinity binding by the Ah receptor. Bishop *et al.* [92] suggested that high mortality in snapping turtles (*Chelydra s. serpentina*) in 1984 and 1985 in Hamilton Harbor, Ontario was strongly correlated with tissue PCB concentrations. Bishop *et al.* [93] observed a significant increase in abnormal development with increasing HAH exposure in snapping turtles eggs at various sites in the lower Great Lakes; the strongest correlations were associated with PCDD/PCDF concentrations. Eisenreich *et al.* [94] found no evidence of immediate effects on embryonic development and hatching success of maternally-exposed snapping turtle eggs collected from the Hudson River, USA relative to those from a reference site; however, high mortality and lower growth rates, correlated with PCB concentrations in the eggs, were observed eight months after hatching.
# **Polybrominated Diphenyl Ethers**
# *Background and Chemistry*
Polybrominated diphenyl ethers (PBDEs) are a diverse group of chemicals that are structurally similar to PCBs, and like PCBs have 209 congeners. Produced as octa-, penta-, or deca-BDE formulations, PBDEs are used primarily as flame retardants in commercial products, including building materials, electronics, furnishings, motor vehicles, plastics, polyurethane foams, and textiles [95, 96]. Commercial PBDE products are typically composed of a complex mixture of congeners, although most mixtures are dominated by one or two specific congeners [97]. The three dominant forms used in industrial manufacturing (deca-, octa- and penta-BDEs) each have specific uses, *i.e.*, penta-BDE (polyurethane foams), octa-BDE (rigid plastics e.g. ABS) deca-BDE (textiles, resins, and rigid plastics). By 1990, worldwide production of PBDEs had surpassed peak production of PCBs; coupled with the fact that PBDEs are not chemically bonded to the materials with which they are associated, and are thus readily released during product use and disposal, they now occur widely in the global environment [98].
PBDEs have low water solubility (typically in the low g/L range) and log Kow values ranging from 5.9-6.2, 8.4-8.9, and 10 for the penta- octa- and deca-BDE, respectively [99]. The penta-, octa-, and deca-BDEs have thus been detected globally in a wide variety of matrices, both biotic and abiotic [97, 100]. In a recent study of marine water and sediments in Japan, concentrations in water, sediment, and fish and invertebrates were in the low pg/L range, ng/kg range, and ng/g range, respectively [101]. In sediments, deca-BDEs generally dominate the total PBDEs, while in biota tetra-BDE (BDE-47) tends to dominate, an observation consistent with other studies [100]. Not all PBDEs appear to biomagnify and those that do, do not appear to biomagnify to the same degree as PCBs [101, 102]. Further, metabolism, specifically debromination, occurs in a number of species, including fish and microbes [98, 103]. Regional comparisons of total and specific PBDE congener concentrations in the environment indicate that PBDE concentrations in Europe are approximately 10-fold below those in North America [100] and concentrations in arctic marine mammals are currently 10 to 100-fold lower than in temperate species [100, 104]. Currently, the penta-BDE mixture is banned in Europe, the main manufacturer of the penta- and octa-BDEs has begun a phase-out of these congeners, and the deca-BDE, despite industry arguments, is now being banned in jurisdictions across the globe [98]. A series of recent reviews have examined their history, chemistry and toxicology in marine and freshwater ecosystems [95-100, 104, 105].
# *Ecotoxicology*
Due to the chemical nature of PBDEs, and their similarities to PCBs, toxicological research has predominantly focused on freshwater and marine organisms. In mammals, PBDEs may act as hormone mimics affecting thyroid function, sexual development and behavior; be cytotoxic (increased apoptosis and necrosis); and may be linked to tumor formation and cancer [98]. The mode of action of PBDEs is uncertain; like PCBs and dioxins, PBDEs were originally thought to act *via* the Ah receptor but recent evidence suggests that this may not be the case as the biomarkers associated with AhR activation are not up-regulated following PBDE exposure [97, 99]. PDBEs have been implicated in disruption of thyroid hormone homeostasis in rodents [97, 99], though it is unclear if this mode is shared with other vertebrates and invertebrates, especially those with no analogous hormonal system. Recent work has found evidence for disruption of oxidative phosphorylation and inhibition of complex II of the mitochondrial electron transport chain in fish, especially for hydroxylated forms of PBDEs [106].
The limited toxicological research on PBDEs has generally focused on whole organism or molecular responses and studies investigating potential effects of PBDEs at the population/community level are rare. Indeed, the majority of work has focused on characterizing concentrations of PBDEs in fish, birds, marine mammals and invertebrates [102, 107-111]. When examining microbial populations for their ability to metabolize various PBDEs, a strain of bacteria (*i.e.*, *Burkholderia xenovorans* LB400) was shown to exhibit some toxicity when exposed to mono-BDE. The toxicity was attributed to a metabolite formed during the biotransformation process, and this strain also produced hydroxylated PBDEs which, based on work with vertebrates, are thought to be more toxic than the parent compounds [103, 106]. Several recent reviews [103, 106] reveal the paucity of acute toxicity information for aquatic organisms. However, the limited data that are available indicate that acute toxicity from PBDEs or their metabolites is unlikely. In embryonic zebra fish, the 72-h EC50 for developmental effects (e.g. developmental arrest, edema) was 14.5 g/L (25 nM) and the adult LC50 after 96 h was between 174 and 232 g/L (300 and 400 nM) for the hydroxylated BDE-47 [106]. Based on current environmental concentrations and appropriate safety factors, the authors concluded that concerns for wildlife for this PBDE metabolite are unwarranted. The 24-h LC50 to *D. magna* for PBDE congener 153 was >210 g/L with some chronic effects on reproduction at concentrations >12.5 g/L [112]. Due to their physicochemical properties (e.g. hydrophobicity), PBDEs are not ideal candidates for microcosm/mesocosm-based toxicity studies and because of their similarities to PCBs (chemical, toxicological, and in terms of co-occurrence) it would be difficult to attribute any observed effects at the population or ecosystem level in the field to PBDEs alone. This issue is highlighted by Jaspers *et al.* [110] who measured a variety of POPs (PCBs, PBDEs, HCHs and DDT) in predatory aquatic and terrestrial bird tissues. Of the seven species examined, six showed no relationship between tissue burden and condition index. The lone exception was the barn owl, which showed negative correlations with PCBs and DDT, which were 100 and 10-fold higher in concentration, respectively, than the PBDEs. The same relative difference between total PCB and PBDE burden in salmonid tissue concentrations have been reported (43,100 ng/g lipid and 2440 ng/g lipid, respectively) in Lake Michigan [102], meaning the attribution of ecological effects solely to PBDEs in the field is unlikely or not occurring at this time.
# **POLYCYCLIC AROMATIC HYDROCARBONS**
# **Background and Chemistry**
Polycyclic aromatic hydrocarbons (PAHs) are a class of POPs comprised of thousands of individual substances that contain two or more fused aromatic rings composed of carbon and hydrogen atoms. PAHs are formed through pyrogenic, petrogenic, diagenetic, and possibly biogenic sources [113]. Pyrogenic sources may be natural, such as forest fires and volcanoes, or industrial, such as the incomplete combustion of fossil fuels, fugitive losses during petroleum extraction, transport, and industrial emissions [114]. Petrogenic sources result from diagenetic processes – low temperature, high-pressure reactions of biogenic materials that occur over geological time scales that lead to the formation of petroleum and other fossil fuels. PAHs are generally hydrophobic and many interact strongly with sedimentary organic carbon [113] and bioaccumulate in aquatic biota, particularly those at lower trophic levels [115, 116]. As such, PAHs are commonly associated with sediments and particulate matter and ecotoxicological concerns have therefore focused on toxicity to aquatic benthic communities and impacts on associated food chains. Historically, only 16 PAHs have been prioritized as environmentally significant and thus received the focus of research; however, it is now recognized that aquatic communities may be exposed to, and potentially affected by, hundreds of PAHs [116] and information on their potential risks are poorly understood.
# **Ecotoxicology**
The toxicology of PAHs in aquatic environments has been well documented and numerous reviews/books are available in relation to bioavailability [113, 115], bioaccumulation [115, 117] and toxicity [114, 118, 119]. PAHs can adversely affect aquatic organisms physically (e.g. smothering, attenuation of light, habitat modification, and reduced food availability) and directly, *via* toxicity from parent or photosensitized PAHs [114]. The former is most commonly associated with accidental releases of petroleum while the latter results from exposure to PAHs associated with oil or derived from natural or industrial sources. The toxicological effects of PAHs are numerous (see Table **14**.1 in [114]). The primary mode of action of PAHs is narcosis [120]. However, some PAHs act as procarcinogens through metabolic formation of DNA adducts, a potentially critical initial step in carcinogenesis [121]. DNA adduct formation has been used as a biomarker of PAH exposure in aquatic organisms [119]. PAHs can also induce immunosuppression [122] as indicated by increased incidences of disease in Japanese medaka exposed to benzo[a]pyrene [123]. PAHs and their derivatives may also affect estrogenic activity. Rainbow trout hepatocytes exposed to anthracene exhibited anti-estrogenic activity, possibly mediated through binding to the Ah receptor [124]. Villeneuve *et al.* [125] found that several PAHs and hydroxylated or methylated PAH derivatives induced estrogenic responses in three separate cell lines.
A unique property of some PAHs is the ability to absorb energy from the ultraviolet spectrum of sunlight, resulting in excited state molecules that, through the subsequent loss of energy, can be several orders of magnitude more toxic than the parent molecules [126-128]. This phenomenon, referred to as phototoxicity, has been demonstrated in freshwater invertebrates [129-132], fish [133-134], and amphibians [135-138]. While most of these studies were conducted under laboratory conditions, PAH-UV interactions in the field have been observed [126, 138, 139] and it has been speculated that synergistic interactions between UV light and PAHs in aquatic habitats may be a contributing factor in amphibian population declines [140]. Others have argued that phototoxicity in the field is ecologically irrelevant because abiotic factors (e.g. dissolved organic carbon), physiological mechanisms (e.g. metabolism/excretion) and physical structures (e.g. integument, burrowing, larval cases) mitigate exposure to UV radiation [141].
Evidence for impacts of PAHs at higher levels of biological organization in the field is scant. Unlike many of the classic POPs, PAHs do not biomagnify [115]. Greatest PAH tissue residues appear to be associated with primary consumers and detritivores in sediments, and tissue concentrations generally decrease with increasing trophic level due to species-specific differences in toxicokinetics and increased biotransformation, especially in vertebrates [115, 142]. Thus, effects on populations and communities are more likely to result from direct exposure to PAH or indirect ecological effects than to food chain transfer and subsequent direct effects at higher trophic levels.
Freshwater microbial communities are both affected by and adaptable to PAHs. In a field-based microcosm study, Baker and Morita [143] found that glucose mineralization and phosphatase levels declined significantly but methane and CO2 production rates significantly increased in sediment bacterial communities after a 4-week exposure to crude oil designed to mimic a spill. Nitrogen fixation was not affected by 0.1% (v/v) oil, but was reduced after 8 weeks by 1.0% oil. In contrast, Nyman [144] found that exposure of wetland sediment microbial communities to two types of crude oil stimulated bacterial metabolic activity as indicated by measurements of redox potential and respiration. PAHs occur naturally, so it is not surprising that microbial communities have evolved the capacity to degrade them [145], and PAH-degrading capacity is much greater in contaminated soil where selection has favored bacteria capable of withstanding exposure [146]. However, in situations of heavy contamination (e.g. oil spill), ecosystem integrity and function may be affected due to lower microbial diversity as this reduction disrupts the tight coupling and interdependence among consortia, and between consortia and grazers. For example, Nyman [144] observed an increase in metabolic activity and oil degrading activity in their wetland sediment study, but this came at the expense of microbial diversity, with tolerant species becoming dominant as sensitive species declined in abundance.
The toxicity of PAHs to freshwater algae and macrophytes has been evaluated in laboratory and field studies. Bott and Rogenmuser [147] exposed algal communities to three oil extracts in stream microcosms for several weeks. No. 2 fuel oil extracts depressed algal biomass (measured as chlorophyll *a*), decreased diatom occurrence, and resulted in dominance by blue-green algae. Used crankcase oil extracts also depressed biomass, but Nigerian crude extracts did not, and both of these extracts had less effect on algal community composition than did the No. 2 extracts. Marwood *et al.* [148] observed effects of PAHs at environmentally relevant concentrations on photosynthesis in natural algal assemblages and attributed this to phototoxicity. Burk *et al.* [149] found that total plant cover, total and mean number of species, and Shannon diversity declined progressively for two years after an accidental oil spill in a marsh and eighteen species found before the spill were absent the following season. However, the vegetation of the marsh showed substantial recovery by the third and fourth years. McGlynn and Livingston [150] modeled adsorption/desorption and potential effects of sediment PAHs at low concentrations by rooted aquatic plants in field and laboratory experiments. The macrophytes' roots assimilated PAHs and the assimilation exhibited saturation. Growth of the macrophytes was inhibited by PAHs but at concentrations several orders of magnitude greater than threshold effects levels for aquatic animals.
Bestari *et al.* [151] and Sibley *et al.* [152, 153] exposed freshwater plankton communities to creosote in microcosms for 83 days at concentrations ranging from 0.06 to 109 mg/L. Creosote had no direct toxic effect on phytoplankton whose population densities and diversity in all treatments exceeded those in the controls and exhibited a parabolic relationship relative to both time and total PAH [152]. In contrast, zooplankton abundance and diversity was significantly reduced by creosote, with a 7-day community-level no-effect concentration of 5.6 g/L [153]. The zooplankton community was dominated by rotifers, which proliferated at the expense of more sensitive cladocerans and copepods. Recovery to pre-treatment abundance levels occurred in all concentrations by the end of the 83-day exposure. The growth of phytoplankton populations appeared to be stimulated by both indirect (lower grazing pressure from zooplankton) and direct (hormetic stimulation by PAHs) effects.
Several studies have examined the response of freshwater benthic macroinvertebrate communities to PAHs. Crunkilton *et al.* [154] monitored the response of benthic macro-invertebrates in a small Missouri, USA stream into which 1.5 million liters of domestic crude oil had been spilled. Sensitive members of the benthic community (aquatic insects, mussels, snails) declined to <0.1% of expected abundance 25 days after the spill and species diversity indices and the abundance of mayfly and stonefly genera were below water quality criteria for Missouri streams up to 11 months after the spill. The impacts were attributed to physical obstruction of both substrate and organisms, and PAH toxicity. West *et al.* [155] evaluated the effectiveness of a carbonaceous resin to reduce the bioavailability of PAHs in field-contaminated sediments as a basis for potential remediation using laboratory toxicity tests and field colonization studies. The resin significantly reduced pore water concentrations of eight measured PAHs in both laboratory and field sediments. In laboratory tests, bioaccumulation and phototoxicity in L. variegatus were significantly reduced; in the field-deployed sediments, the resin amendment also decreased pore water PAH concentrations but did not improve benthic invertebrate colonization. Den Besten *et al.* [115] investigated impacts of PAH-contaminated sediments on benthic macroinvertebrates in the Rhine-Muese Delta in The Netherlands. Highly contaminated sediments contained significantly fewer taxa, had lower species diversity compared to reference sites, and produced significant toxicity in sediment bioassays with the invertebrates C. riparius and D. magna. De Lange *et al.* [156] evaluated seasonal variation and bioavailability of PAH in contaminated floodplain lake sediments in relation to benthic invertebrate community structure. While sediment-associated PAH concentrations occurred at levels at which effects were predicted, biovailability was low and the PAHs were not associated with observed impacts on benthic community structure.
Cooper *et al.* [43] compared benthic community and fish population structure in two sub-watershed wetlands of a western Michigan lake, one of which is highly contaminated with PAHs and metals as a result of a long history of industrial activity. Significantly fewer insect taxa, reduced fish species richness and catch per unit effort, and lower invertebrate and fish index of biotic integrity scores were found in the industrialized watershed. Cormier *et al.* [157] used a formal strength-of-evidence methodology [158] to infer causes of impairments at two sites in the Little Scioto River, Ohio, USA, which is heavily contaminated by sediment PAH. At the upstream site, they concluded that impairment of the benthic community and fish populations was due to altered habitat substrate (predominance of fine-textured sediment) and low dissolved oxygen. At the downstream site, impacts included lower diversity and dominance by pollution-tolerant invertebrates and reduced fish growth, elevated PAH tissue concentrations and increased incidences of abnormalities in fish, all of which could be causatively explained by concentrations of sediment PAHs. Lesko *et al.* [159] assessed the effects of contaminated sediments on reproductive potential of female brown bullhead (*Ameiurus nebulosus*) collected from the Black and Cuyahoga Rivers, Ohio, both contaminated with metals, PAHs and PCBs. Females from the most contaminated (Cuyahoga) river had higher fecundity and the population size was larger compared to the reference river, which they attributed to an enhanced food supply due to reduced competition from predators. However, fish diversity in the Cuyohoga River was lower and incidences of tumors higher relative to the reference river [160]. Evidence that PAHs can act as endocrine disrupters has largely been developed for fish [71, 161, 162]. While studies to date have not linked endocrine effects directly to population or community-level effects, evidence that PAHs may impair reproduction in fish, either through altered sex steroid metabolism or biosynthesis (see [71] for examples), reduced growth or abnormalities in larval fish [163] suggest that endocrine-induced population effects are possible.
Studies investigating the effects of PAHs on amphibians and reptiles have largely focused on organism and physiological responses, either through direct exposure to PAHs [136, 164, 165] or synergistic exposure to UV light as described above. Few studies have examined the effects of PAHs at the population and higher levels of biological organization in amphibians. Physiologically, amphibian responses to PAHs are similar to other vertebrates [118]. Lefcort *et al.* [166] studied the effects of oil and silt on the growth and metamorphosis of larval mole salamanders, *Ambystoma opacum* and *A. tigrinum tigrinum* in oil-contaminated ponds and outdoor microcosms treated with used motor oil*.* In both test systems, both species had reduced size and weight compared to controls that was attributed to an indirect effect of reduced algal growth (salamander food) and direct toxic effects.
# **FLUORINATED SURFACTANTS**
# **Background and Chemistry**
Surfactants are surface-active materials that, at low concentrations, are capable of reducing the surface tension of a liquid *via* selective adsorption at the interface [167]. Surfactant molecules are amphiphilic, characterized by a hydrophilic (water-soluble) 'head' group attached to a hydrophobic (water-insoluble) 'tail' portion. In conventional, hydrocarbon-based surfactants, the hydrophobe is typically an oleophilic (lipid soluble) hydrocarbon. In perfluorinated surfactants (PFSs) fluorine atoms replace hydrogen atoms on the hydrophobe. The replacement of hydrogen with highly electronegative fluorine atoms on the hydrophobe renders PFSs both hydrophobic and oleophobic, capable of repelling both water and oils. Increasing the number of fluorine atoms in the hydrophobe increases chemical stability as bond strength generally increases with an increase in the number of fluorine constituents [168]. The exceptional persistence rendered by the high molecular stability has led to the detection of PFSs in a variety of biotic and abiotic matrices on a global scale [169-172].
The global pervasiveness of PFSs reflects both a long history of manufacture (since the 1950s) and widespread use as surface treatments for carpets, fabrics, and paper products to repel soil, oil, and water and applications such as fire fighting foams, adhesives, electronic insulators, cosmetics, cleaners, among others [167, 170]. Recently, concerns over the occurrence of perfluorooctane sulfonic acid (PFOS) in the environment, especially in sensitive Arctic regions, resulted in a cessation of production in 2000 by 3M Corporation and the recent inclusion of PFOS under Annex B of the Stockholm Convention on POPs, indicating that its use should be restricted.
# **Ecotoxicology**
Environmental concerns about PFSs have predominantly focused on two compounds: perfluorooctanoic acid (PFOA) and PFOS. Over the past decade, the toxicity of PFOS and PFOA to environmental receptors has been well studied as reviewed in [170, 171, 172]. However, with the exception of the microcosm studies described below, most of this work has been conducted at the organism level.
Several studies have assessed the toxicity of PFSs in freshwater macrophytes and algae. Boudreau *et al.* [174] and Boudreau [175] assessed the toxicity of PFOS and PFOA in the algae *Chlorella vulgaris* and *Pseudokirchneriella subcapitata*, and the aquatic plant *Lemna gibba*, under laboratory conditions at chain lengths of 4 to 7 carbons. In tests with PFOS, 96-h growth inhibition NOEC values were 5.3 and 8.2 mg/L for *P. subcapitata* and *C. vulgaris*, respectively, and 6.6 mg/L for *L. gibba* (wet weight). In tests with PFOA, laboratory EC10 values for growth ranged from 5.7 to 59.4 mg/L for *C. vulgaris* (96-h) and *L. gibba* (7 d), respectively [175]. Colombo *et al.* [176] calculated a NOEC value of 12.5 mg/L for growth inhibition in *P. subcapitata* exposed to the ammonium perfluorooctanoate. Liu *et al.* [177] assessed four perfluorocarboxylates and two sulfonates to the alga *Scenedesmus obliquus* and found that toxicity based on cell density ranged from none (PFOA) to 21.6 mg/L (perfluorotetradecanoic acid). Latal *et al.* [178] showed that perfluoro-hexanoic, -heptanoic, -octonoic, and -nonanoic acid were more toxic than PFOS and PFOA to three species of algae (LC50s range: 6.0-24.3 mg/L). Blue-green and diatom species were comparable in sensitivity but both were more sensitive than green algal species. In an outdoor microcosm study, Boudreau *et al.* [174] determined a 42-day NOEC (frond number) for PFOS of 0.2 mg/L for a population of *L. gibba*. Hanson *et al.* [179, 180] estimated NOEC values in excess of 0.3 mg/L and 23.9 mg/L for PFOS and PFOA, respectively, for two species of *Myriophyllum* in outdoor microcosm studies.
With one exception, freshwater invertebrates appear to be relatively insensitive to PFSs. Boudreau *et al.* [174] estimated NOEC values for immobility in 48-h exposures of 0.8 and 13.6 mg/L for *D. magna* and *D. pulicaria*. NOEC values in tests with PFOA for both *Daphnia* species indicated reduced toxicity relative to PFOS [175]. In both cases, daphnids were only sensitive to carbon chain lengths ≥ 8. Ji *et al.* [181] estimated LC50s of 17.95 mg/L for PFOS and 199.51 mg/L for PFOA for the daphnid *Moina macrocopa*, which is approximately twice the LC50 determined for *D. magna*. In a 7-day chronic test, *M. macrocopa* experienced significantly reduced reproduction at 0.31 mg/L for PFOS, which was approximately seven times lower than the effect concentrations observed over the 21-day exposure in *D. magna*. The greatest toxicity observed for PFOS in any aquatic species is the 20-day LC50 of 9.2 g/L reported for *C. tentans* [182]. In the same test, *C. tentans* did not respond to PFOA in 10-day exposures at concentrations up to 100 mg/L. In a series of indoor microcosm studies, PFOS caused a significant reduction in zooplankton abundance and altered community structure at concentrations ≥10 mg/L [183]. In a similar test with PFOA, Sanderson *et al.* [184] determined a lowest observed effect concentration (LOEC) of between 10 and 70 mg/L depending on taxonomic group. In these studies, zooplankton communities became dominated by rotifers with simultaneous declines in cladoceran and copepod species at the highest concentrations. In a 35-day outdoor microcosm study, Boudreau *et al.* [174] estimated a community-level NOEC of 3.0 mg/L for zooplankton, with significant declines in zooplankton abundance at 30 mg/L. Kannan *et al.* [185] estimated a bioconcentration factor of approximately 1000 for PFOS in Great Lakes benthic invertebrates and Higgins *et al.* [186] estimated lipidnormalized BSAF values of 33 and 42 for PFOA and PFOS indicating that both compounds may be accumulated from sediments and thus available for trophic transfer in freshwater food chains.
In fish, studies indicate that toxicity thresholds of PFSs are typically much higher than environmental concentrations. Du *et al.* [187] observed no mortality in zebra fish exposed to PFOS in a 70-day exposure but did observe significant declines in growth and various biochemical and genetic endpoints at concentrations as low as 50 g/L. Hagenaars *et al.* [188] also found no mortality in carp fry exposed to PFOS concentrations up to 1 mg/L but did observe significantly reduced condition factor and liver indices as low 0.1 mg/L. Colombo *et al.* [176] estimated a 96-h LC50 for the ammonium salt of PFOA of 400 mg/L. PFOS and PFOA are hepatotoxic, affecting hepatocyte membranes indicative of necrosis and interfere with fatty acid metabolism [189]. PFOS and PFOA exposures can also decrease circulating sex steroids in fish depending on species, age, and sex [189, 190]. Contrary to mammalian studies, PFOS and PFOA appear to be relatively weak peroxisome proliferators in fish [189]. In freshwater fish, PFOS and PFOA bind tightly to serum proteins [191] and bioaccumulate (from highest to lowest) in fish in the blood, kidney, liver, and gall bladder [192]. Kannan *et al.* [185] found that PFOS concentrations in Chinook salmon were 20 times greater than in their prey species and Furdui *et al.* [193] estimated log bioconcentration factors of 4.1 and 3.8 for PFOS and PFOA, respectively, in Great Lakes lake trout. These data provide evidence of food chain transfer of PFAs. Interestingly, Kannan *et al.* [195] found notable concentrations of PFOS in fish eggs suggesting oviparous transfer.
Numerous studies have measured residues of PFSs in aquatic fish-eating birds [169, 185] but few have assessed toxicity. Newsted *et al.* [194] determined a 5-day LD50 of 150 mg/kg body weight in young (2-day old) mallard ducks exposed to food-borne PFOS. The concentration of PFOS in mallard livers associated with mortality was at least 50-fold greater than the single maximum concentration that has been measured in livers of avian wildlife indicating low risk. Adult mallards fed PFOS up to 150 mg/kg feed showed no treatment-related effects [195]. Based on this work, they estimated an avian toxicity reference value of 0.021 mg/kg body weight per day. Kannan *et al.* [185] recorded the highest concentrations of PFOS from bald eagles in the Great Lakes and estimated a biomagnification factor of 10 to 20. Excretion of PFOS in bald eagles appears to be more rapid than classical POPs [185], but the potential for binding of PFOS with serum proteins and production of metabolites whose toxicity is poorly understood, warrants further investigation with respect to potential risks to fish-eating bird populations.
The occurrence and toxicity of PFSs in amphibians and reptiles is limited to only a few studies. PFS concentrations ranging from 137 to 250 ng/g (wet weight) have been measured in green frogs, yellow-blotched map turtles, and snapping turtles from the Great Lakes [185]. Ankley *et al.* [196] observed reduced growth and delayed metamorphosis, which can impact population stability, in northern leopard frogs at 3 mg/L PFOS and hypothesized that this may have been the result of impaired thyroid function.
The weight of evidence indicates that PFSs pose limited risks to freshwater organisms as toxicity thresholds are typically well above concentrations of PFAs measured in the field. Beach *et al.* [171] derived protective screeninglevel concentrations for PFOS of 2.3 mg/L for freshwater plants and algae and 1.2 g/L for aquatic invertebrates, the latter value reflecting the sensitivity of *C. tentans* [182]. A tissue-based threshold value of 87 mg/kg wet weight was determined to be protective of fish. Collectively, the evidence does not support a causal link between current PFS contamination and population or community-level impacts in aquatic systems.
# **PHARMACEUTICALS AND PERSONAL CARE PRODUCTS**
# **Background and Chemistry**
Pharmaceuticals and personal care products (PPCPs) encompass a wide variety of chemicals and applications, including the cure and prevention of disease in humans and livestock, diagnostic treatment (e.g. x-ray contrast media), growth promotion in livestock, and chemical additives (e.g. musk fragrances, antibacterial agents) in personal care products [197]. Thousands of PPCPs are produced and used daily around the world, in quantities that now approach those typical of agrochemicals [198]. Sources of PPCPs to freshwater environments include wastewater treatment effluents, run-off from agricultural fields amended with manure and sewage, direct addition *via* livestock excretion, and leaching from landfill sites.
In contrast to many of the legacy chemicals addressed in this chapter, knowledge of pharmaceuticals in the environment, and the attendant concerns for human and environmental health, emerged only in the late 1990s. While evidence of hormonally active pharmaceuticals date back to the 1960s [199], the pervasive nature of PPCPs has only recently been brought to light because of advancements in analytical technology that made it possible to detect PPCPs at the low concentrations at which they typically occur. In North America, the widespread occurrence of PPCPs in surface waters was documented by Kolpin *et al.* [200] who identified 82 compounds from surface waters of 139 streams with many compounds co-occurring. Many studies have since added to the list of PPCPs known to be present in the environment [201]. These studies show that PPCPs occur predominantly at sub-µg/L concentrations. However, although the majority of PPCPs occur at low concentrations and degrade rapidly under most environmental conditions, continual addition to the environment renders them effectively "pseudo-persistent" [202]. Moreover, pharmaceuticals are designed to elicit biological effects, which is the basis of their therapeutic activity [203]. The combination of pseudo-persistence and biological activity has lead to uncertainty about how PPCPs will behave toxicologically in the environment and legitimate questions about potential risks to environmental receptors.
# **Ecotoxicology**
The fate and effects of PPCPs in aquatic systems has been summarized in several reviews [203-209]. There is general consensus that acute exposure of aquatic organisms to PPCPs carries negligible risk [210, 211]. In a review of over 360 acute toxicity endpoints in freshwater organisms for 107 human PPCPs, Webb [212] found that <10% were toxic at concentrations <1 mg/L, a value approximately 3 to 4 orders of magnitude above concentrations typically measured in freshwater ecosystems. Thus, interest in PPCPs from an ecological impacts perspective is presently focused on potential impacts from chronic exposures. Here, ecological impacts, if any, will likely depend on the type of PPCP. For example, Fent *et al.* [210] found that chronic LOECs in laboratory test species are about two orders of magnitude greater than maximum concentrations in sewage treatment plant effluents. However, their assessment did not include antibiotics or hormones, both of which warrant additional detailed investigation regarding potential risks to aquatic biota [205, 206].
One of the best studied PPCPs is ethynylestradiol (EE2), a synthetic compound widely used in birth control formulations and commonly detected in sewage treatment plant effluents and biosolids. As a hormone mimic, EE2 is a potent endocrine disrupting agent in freshwater vertebrates and has been implicated in a number of cases of sexual disruptions reported in freshwater fish exposed to sewage treatment effluent [213-214]. EE2 induces synthesis of the egg yolk precursor vitellogenin in male and juvenile fish, can cause increased incidences of intersex (gonads possess features of both sexes), feminization of male fish and reduced fertilization success [215, 216]. However, few studies have linked these changes to actual changes at higher levels of biological organization. One exception is the study of Kidd *et al.* [217] who showed that chronic exposure to environmentally relevant concentrations of EE2 over several years led to the collapse of a population of fathead minnows in a whole lake exposure. In that study, Palace *et al.* [216] found evidence of intersex and inhibited development of testicular tissue in males of pearl dace (*Margariscus margarita*) and suggested a trend toward reduced population abundance and smaller young-of-the-year size classes in the EE2-treated lake. Interestingly, and reflective of the transient nature of PPCP contamination, fish populations were observed to recover after exposure was halted. Watts *et al.* [218] and Dussault *et al.* [219] found no evidence for effects of EE2 in *C. riparius* and *C. tentans*, respectively, in life cycle tests. However, Dussault *et al.* [220] showed that *C. tentans* and *H. azteca* accumulated EE2 (bioaccumulation factors of 31 and 142, respectively) and could therefore serve as a source of this compound to vertebrate receptors at higher trophic levels. Jensen *et al.* [221] estimated an EC50 of 0.011 µg/L for 17--trenbolone, an endocrine-active growth additive used in livestock, in fathead minnows; this concentration is comparable to those measured in beef cattle feedlot runoff [222].
Evidence for effects of non-endocrine PPCPs at higher levels of biological organization is rare but may occur in microbial populations exposed to antibiotics and veterinary medicines. In a recent review of the occurrence and toxicity of antibiotics in aquatic ecosystems, Kümmerer [207] suggests that bacteria and microalgae are 2 to 3 orders of magnitude more sensitive than organisms at higher trophic levels so the prospect for effects on microbial communities cannot be discounted. Indeed, Backhaus and Grimme [223], in a bioluminescence inhibition test with *Vibrio fischeri*, found toxic effect values (EC10) for two antibiotics in the range of concentrations expected in surface waters. Tetracycline was found to disrupt nitrification at concentrations found in some freshwater sediments [224]. Of particular concern with this class of pharmaceuticals is the potential for the development of resistance in freshwater bacteria and this has been demonstrated in natural bacterial populations in sediments associated with aquaculture [207, 225, 226].
Richards *et al.* [227] evaluated the effects of a PPCP mixture composed of ibuprofen, fluoxetine, and ciprofloxacin at individual concentrations of 10, 100, and 1000 µg/L on populations of macrophytes (*L. gibba* and *Myriophyllum sibiricum*), plankton, and bacterioplankton in a 35-day microcosm study. Significant decreases in growth of the plant species at intermediate and high concentrations and eventual loss of plant populations in the high treatment were observed. A significant increase in overall abundance and a significant decrease in diversity of phytoplankton occurred at the high concentration. The higher abundance reflected a large increase in one species that dominated the phytoplankton community; other phytoplankton species were unaffected or were eliminated, explaining the lower diversity. Similarly, zooplankton increased in abundance and had reduced diversity at the highest concentration. The mixture had no effect on bacterioplankton abundance. The authors concluded that the individual risks posed by these compounds in freshwater ecosystems were negligible.
In a 49-day microcosm study (34 days exposure and 14 days recovery), to a four-tetracycline mixture at individual concentrations of 10, 30, 100, and 300 µg/L, Wilson *et al.* [228] measured biomass production, community respiration, and primary productivity, as well as phytoplankton and zooplankton community responses. Phytoplankton abundance and community respiration decreased significantly at the two largest concentrations but primary productivity was unaffected. Community metabolism (ratio of productivity to respiration) decreased significantly at the two greatest concentrations due to significant increases in respiration. Zooplankton were not affected by the tetracycline mixture. The effects observed in this study are approximately 2 orders of magnitude greater than concentrations expected for tetracyclines in freshwater systems. Hillis *et al.* [229] evaluated the effect of monensin, an antibiotic commonly used in beef and poultry, on zooplankton communities in a 50-day microcosm study at concentrations ranging from 0.5 to 500 µg/L. The community-level NOEC (50 µg/L) was approximately 50 times greater than environmental concentrations. McGregor *et al.* [230] observed no impacts of monensin on macrophyte populations up to 100 µg/L in a microcosm study. Sanderson *et al.* [231] evaluated ivermectin, a commonly applied anti-helmenthic drug that has been shown to be highly toxic to aquatic invertebrates, in a longterm (250 day) microcosm study. They demonstrated that ivermectin could pose risks to aquatic organisms at or below the predicted environmental concentrations.
Overall, while there are some exceptions for classes of PPCPs such as estrogens and antibiotics, the weight of evidence indicates that the probability of acute or chronic effects at higher levels of biological organization in aquatic ecosystems is small. Indeed, based on a simple hazard approach comparing the ratio of the predicted effects concentration (PEC) and the predicted no effects concentration (PNEC), Tarazona *et al.* [232] suggest that the likelihood of observing ecosystem-level effects would be expected at ratios of approximately 10 or higher. Such high PEC/PNEC ratios for PPCPs are rarely observed in freshwater environments.
# **COMPOUNDS IN PLASTICS**
# **Background and Chemistry**
The production of plastic yields a variety of potential environmental contaminants. The two most common, which are addressed here, are bisphenol A (BPA, 4,4'-dihydroxy-2,2-diphenylpropane) and nonylphenol (NP), a member of the alkylphenol ethoxylates (APEs). BPA is a key building block used to produce polycarbonate plastics and epoxy resins [233]. Polycarbonates are incorporated into sheeting, glazing, bottles and storage containers and epoxy resins are used as protective coatings on buildings, boats and vehicles; collectively, this usage accounts for 95% of BPA in the plastics industry [234]. NP is the basis for non-ionic surfactants commonly used in the manufacture of industrial and domestic detergents, pesticide formulations, emulsifier and dispersing formulations, cosmetics, and paints.
BPA is not especially stable in the end product and has been observed to migrate into surrounding environmental matrices. Although BPA is relatively short-lived in the environment (the half-life in water ranges from several hours to a few days depending on initial conditions [235]), continuous inputs and the ubiquity of plastics in the environment, has resulted in routine detection of this compound [233, 236]. BPA is most commonly detected in waters downstream of wastewater treatment plants with concentrations in the effluent typically an order of magnitude greater than in receiving waters [235, 237]. In a review of BPA concentrations in rivers and groundwater, Sharma *et al.* [235] found BPA occurred predominantly at low ng/L and low g/L range, respectively. In a recent exposure analysis for North America and Europe, median BPA concentrations in freshwater systems were 0.081 and 0.01 g/L and 0.6 and 3.4 ng/g in sediments, respectively, [238]. Despite low persistence and generally reduced global presence relative to other organic contaminants, BPA has attracted attention from an ecotoxicological perspective because it can bioaccumulate and has been shown to act as an estrogen mimic. For example, BPA is structurally similar to the potent estrogen diethylstilbestrol and has been shown in the yeast estrogen assay to bind and activate the estrogen receptor in vertebrates [233, 239, 240].
NP enters the environment *via* industrial and commercial sources [241, 242]. Due to their occurrence in cleaning agents, high concentrations of NP ethoxylates enter wastewater treatment plants. Here, metabolic degradation leads to the production of NP, which is released into receiving waters [241]. Not surprisingly, NP has been widely detected in systems with wastewater inputs, with the resulting distribution between environmental compartments driven primarily by its physicochemical properties. Due to low water solubility and a Kow >4, fugacity modeling has shown that NP partitions preferentially into sediment, with concentrations downstream of inputs reaching the mg/Kg range compared to low g/L for water [241, 243]. NP undergoes significant degradation in the water column, with a half-life of a few days [244], but in sediments, half-lives >60 years have been reported [241]. The bioaccumulation potential of NP is generally low to moderate [242], although recent work with zebra fish estimated BCFs >1000 [243]. As an endocrine disruptor, NP can impair reproduction and sexual development as has been shown in fish [241, 245-247]. These estrogenic effects are more pronounced in NP relative to the parent ethoxylate forms and current environmental concentrations may result in population-level effects *via* effects on reproductive fitness [236, 246].
# **Ecotoxicology**
There is a reasonable body of literature examining higher-level effects of NP in aquatic ecosystems. The most extensive is a series of papers that describe the impacts of NP on sediment dwelling nematode, plankton and microbial communities in 230 L aquatic microcosms over an 8-week application phase and a 6-week dissipation phase. The nematode community initially had high abundance and low Shannon diversity, with dominance by one species, *Eumonhystera filiformis* [248]. At week 7, abundances declined and diversity increased but did not correspond to the NP concentrations. The maturity index was the only response that showed a treatment-related response; being significantly lower at the highest concentration (3.4 mg/kg sediment) relative to controls and other treatments. The relative insensitivity of nematodes was attributed to decreased bioavailability due to binding of the NP to the cationic groups in the sediment as a result of the pH of this particular test system. Changes in phytoplankton and periphyton species richness and diversity were not correlated with NP concentrations as the measured NP concentrations were approximately 10-fold lower than those known to cause direct toxicity [249]. However, changes in community composition of phytoplankton were noted, with Conjugatophyceae, which were dominant in all microcosms during the pre-treatment period, being dominant only in the controls and lowest NP concentration post-treatment. In contrast, Cyanophyceae came to dominate at intermediate and higher test concentrations. The authors interpreted this trend as evidence for differential grazing by zooplankton, specifically decreased grazing pressure at higher NP exposures due to direct toxicity on the zooplankton through estrogenic effects [250]. Abundances of copepod larvae were the most severely affected, with declines up to 95% at 200 mg/L NP, with no recovery observed during the 6-week post-treatment period in the three greatest concentrations. Cladocerans were less sensitive, recovering in all but the highest NP exposure, a response that may have been facilitated by the shift in phytoplankton to smaller species [250]. Time-dependent NOEC values for the community ranged from 19 to 44 mg/L. Interestingly, these NOEC values are lower than those of many single species laboratory invertebrate tests [242], though comparable to that estimated for *D. magna* (EC50 of 16.5 mg/L for population growth (*r*) in a 21-day test) [251]. In a subsequent study, using the same microcosms, Hense *et al.* [252] concluded that decreases in zooplankton reproduction, attributed to the endocrine effects of NP, resulted in a delayed shift in phytoplankton community structure and increases in rotifera, due to reduced grazing and competitive pressures. Microbial communities, specifically bacteria and microfungi in sediments, tended to increase in abundance with elevated NP concentrations in these test systems, with only a slight change in the overall microbial community [253]. This could be attributed to increased food resources due to zooplankton mortality at higher NP exposures, as has been observed elsewhere with mass zooplankton and benthos mortality [254]. In microbial microcosm test systems lacking sediment, zooplankton and benthos, water borne microbial communities showed no significant differences in diversity up to 5 mg/L NP [244].
In fish, NP has been shown to affect behavior and survival and, through estrogenic effects, reproduction [245]. Indeed, field populations of freshwater fish exposed to NP *via* wastewater effluents consistently show endocrine modulated effects such as vitellogenin expression, gonadal abnormalities, reductions in circulating testosterone and reproductive dysfunction, and reductions in the gonadosomatic index, all of which can result in population-level effects through impaired fecundity [241]. However, while some studies show a correlation between NP and these effects downstream of wastewater effluents [255], assigning direct causality to NP when many other contaminants that share a mode of action and input source co-occur is difficult. A modeling exercise using data from laboratory and field studies was conducted to examine the potential impacts on populations of brook trout (*Salvelinus fontinalis*) and fathead minnows exposed to NP for three years at 1 and 30 µg/L [245]. Depending on model parameterization, they predicted an increase in population size of 17% or a decline by 28% at 30 µg/L but no significant change at 1 µg/L NP. Fathead minnows showed a similar response, with population reductions up to 21% and a shortened spawning season at 30 µg/L, but full recovery was anticipated within two years after exposure.
There is a robust body of knowledge on the acute and chronic effects of BPA to a suite of aquatic organisms at the individual level under laboratory conditions (see review by Mihaich *et al.* [256]). In acute exposures, EC50 and LC50 values (24 to 96 h) are typically in the mg/L range for invertebrates (1.1 to 16 mg/L), while chronic testing for invertebrates found NOEC values ranging from 0.25 mg/L in the snail *Marisa cornuarietis* for female growth to >3 mg/L for *D. magna* reproduction [256]. Primary producers appear to be slightly less sensitive than invertebrates. For example, the EC50 for growth for the diatom *Skeletonema costatum* was 2.5 mg/L, while the EC50 for growth for *L. gibba* was 32 mg/L [256]. Based on currently measured environmental concentrations, BPA is unlikely to induce acute effects in these organisms. However, Oehlmann *et al.* [233] suggest that effects at higher levels of biological organization may occur through subtle impacts on reproduction and development in vertebrates and invertebrates. They summarized the toxicological literature for types of responses in organisms exposed to BPA and concluded that invertebrates were generally more sensitive than vertebrates such as fish. In snails significant increases in reproductive effects and super-feminization have been observed, which is consistent with the proposed mode of action of BPA as an estrogen mimic. For example, in a 180-day study with *M. cornuarietis*, the EC10 for egg production (increase) was 13.9 ng/L, which is within the range of some environmental concentrations. Other invertebrates appear to be less sensitive. A NOEC of 1 mg/L was determined for reproduction in *D. magna*, and conflicting results have been reported in marine copepods, with some showing inhibition of larval development and others showing accelerated growth, including increased egg production at 20 µg/L. In *C. riparius*, emergence of second-generation individuals was delayed at concentrations as low as 78 ng/L. In fish, the majority of papers report feminization effects *in vivo* and expression of the vitellogenin protein, but at concentrations in the high µg/L and well above what is typically observed in aquatic environments. However, some studies have reported changes in circulating concentrations of some hormones at more environmentally relevant concentrations. Oehlmann *et al.* [233] cite one study on brown trout that reports impacts on sperm quality, a delay in ovulation, and inhibition of ovulation in the low µg/L BPA range, with the interpretation that this could lead to delayed breeding in less favorable periods, with potential impacts at the population-level for these fish. Based on available data, Oehlmann *et al.* [233] felt that BPA could be contributing to adverse reproductive outcomes in populations of exposed fish, but to date, no studies have shown this causally in the field. Staples *et al.* [257] summarized the chronic laboratory data (growth, reproduction and mortality) for this compound, developing species sensitivity distributions and estimating chronic predicted no effects concentrations (PNEC or the 5th centile of the distributions), which are considered protective of populations, communities and ecosystems. They determined PNEC values of 11 to 71 µg/L BPA and concluded, based on current environmental concentrations, that higher-level effects are not anticipated.
# **CONCLUSION**
Global freshwater ecosystems have a long history of contamination from organic pollutants. While an enormous amount of research has been conducted to assess the impacts of organic contaminants on freshwater systems, much of this has been generated at lower levels of biological organization (organism and lower) and clear, cause-effect examples of contaminant-associated impacts at the population, community, and ecosystem level are generally rare (Table **1**).
**Table 1:** Relative state of current understanding, including cause-effect relationships, of the ecological impacts of the organic chemicals addressed in this review in relation to levels of biological organization. Relative rating based on evidence from laboratory, field, and cosm studies. XXX: clear evidence of impacts with some causal relationships established; XX: evidence of impacts but causal relationships not established or uncertain; X: possible evidence of impacts but causality not established; \_\_\_ no evidence of impacts.
1 There is some evidence that community-level effects could occur in microbial communities
Our present understanding of how freshwater ecosystems respond to contaminants is largely based on work with persistent, bioaccumulative legacy chemicals (e.g. PCBs, dioxins, and PAHs) and the information derived from this work has proved essential in developing protective regulatory criteria and implementing ecosystem-based management strategies to mitigate effects. However, there is much research that remains to be done. For example, future research should focus on the potential impacts of more recent chemical classes (e.g. PBDEs, PFSs, and pharmaceuticals), whose physicochemical properties, environmental behavior, and potential impacts on aquatic ecosystems, has not been fully elucidated. In addition, since contaminants rarely occur individually in the environment, there is a need to better evaluate the impacts of chemical mixtures in aquatic systems and potential risks that result from exposure to them. Like individual chemicals, the historical focus for chemical mixtures assessment has been at lower levels of biological organization and there is little information about their effects at the population or community level. The potential effects of mixtures should be considered in the context of cumulative impacts, with emphasis on interactions between both chemical and non-chemical (e.g. nutrients, sedimentation, *etc.*) stressors. In terms of population and community-levels assessments, one of the ideal tools to undertake such studies is model aquatic ecosystems such as microcosms or mesocosms as these facilitate evaluation under "close-to-field" conditions, including direct and indirect effects, both of which may be critical aspects to quantify the fate and effects of organic chemicals in freshwater systems. Finally, greater resources are needed for chemical and biological monitoring of aquatic systems for the purpose of assessing trends in exposure to, and impacts from, chemicals to provide a stronger foundation on which to support regulatory and research initiatives.
# **REFERENCES**
#### **160** *Ecological Impacts of Toxic Chemicals Sibley and Hanson*
### **162** *Ecological Impacts of Toxic Chemicals Sibley and Hanson*
© 2011 The Author(s). Published by Bentham Science Publisher. This is an open access chapter published under CC BY 4.0 https://creativecommons.org/licenses/by/4.0/legalcode
# **CHAPTER 8**
# **Impact of Pollutants on Coastal and Benthic Marine Communities**
### **Ángel Borja\* , María Jesús Belzunce, Joxe Mikel Garmendia, José Germán Rodríguez, Oihana Solaun and Izaskun Zorita**
# *AZTI-Tecnalia, Marine Research Division, Pasaia, Spain*
**Abstract:** In recent years, sources, types and levels of contaminants in the marine environment have increased as a consequence of anthropogenic activities worldwide. Chemical substances are usually present in the marine environment at different concentrations. They are accumulated in the tissues of marine organisms exerting damaging effects at different levels of organization, from organisms to communities and ecosystems. The understanding of their effects and distribution has increased substantially since the early reports on the biological effects of marine pollution and associated monitoring problems. This Chapter has been divided into five main sections: (i) molecular, cellular and tissue level biomarkers in assessing effects, reviewing the use of biomarkers in monitoring effects; (ii) biological effects at organism and population level; (iii) bioassay studies at organism level, focusing on ecotoxicology and discussing toxicity estimation and bioassay limitations; (iv) ecological effects at community level, including structural parameters, such as richness, diversity, *etc.*, but also the proportion of opportunistic and sensitive species, discussing the multiple pressure interactions; and (v) measuring pollutant effects in integrative assessments, both in evaluating risk and assessing the ecological status of the ecosystems. The main objective of the Chapter is to bring together the knowledge on the biological effects of pollution, developed in recent years, at different biological levels of organization.
# **INTRODUCTION**
In recent years, sources, types and levels of contaminants in the marine environment have increased as a consequence of anthropogenic activities worldwide [1, 2] (Table **1**). Chemical substances are present usually in the marine environment in different concentrations; they are not necessarily accumulated in the tissues of marine organisms, but they still can cause toxic effects (e.g. herbicides) even at very low concentrations, exerting damaging effects [3]. The understanding of their effects and distribution has increased substantially since the early reports on the biological effects of marine pollution and associated monitoring problems [4] (Table **1**). However, there is a need to develop methods for the identification, estimation, comparative assessment and management of the potential risks posed by chemical pollutants to aquatic living resources and marine ecosystems [5, 6]. Hence, the relationships between pollutants and direct and indirect effects on the marine environment, across different levels of organization, together with the operational framework for establishing causal-effect relationships are presented in Fig. (**1**).
Assessment of environmental pollution cannot be based solely upon chemical analyses, because this approach does not provide clear indication of the deleterious effects of contaminants [5, 7]. In addition, the increasing number and types of potential pollutants (*i.e.* polybrominated diphenyl ethers, endocrine disruptors, pharmaceuticals, *etc.*) entering the marine environment requires novel strategies for pollution assessment. Consequently, there is general agreement that the most appropriate approach for the assessment of environmental pollution is by integrating a suite of chemical and biological measurements [5, 7-9].
In this respect coastal organisms are of high importance in environmental toxicology as sentinel species, because they are can be used in the assessment of the effects of pollution through biological effect measurements [10]. Bivalve molluscs have been one of the most widely-used indicators to determine the existence and toxicity of chemical substances. Therefore, due to their sessile nature, wide geographical distribution and high bioaccumulation capacity, they have been considered as ideal for the detection of the biological effects of pollutants [11].
In order to analyse the extent of disturbances of a biological system and to quantify the state of health, the integration of several biological effects at different levels of biological organisation has been suggested by several authors [6, 12]. Biological effect measurements incorporate three approaches: (i) biomarker studies at molecular, cellular and tissue level; (ii) bioassay studies at whole organism level; and (iii) ecological surveys at community and population level [13].
**<sup>\*</sup>Address correspondence to Ángel Borja:** AZTI-Tecnalia, Marine Research Division, 20110 Pasaia, Spain; Email: [email protected]
**Table 1:** Effects on marine life, from different types and sources of pollution. Source: adapted from WorldWatch Institute (http://www.gdrc.org/oceans/marine-pollution.html).
This Chapter has been divided into five main sections: (i) molecular, cellular and tissue level biomarkers in assessing effects; (ii) biological effects at organism and population level; (iii) bioassay studies at organism level (as a part of the previous section, focusing on ecotoxicology); (iv) ecological effects at community level; and (v) measuring pollutant effects in integrative assessments. The main objective of the Chapter is to bring together knowledge on the biological effects of pollution developed recently at different biological levels of organization.
# **MOLECULAR, CELLULAR AND TISSUE LEVEL BIOMARKERS IN ASSESSING EFFECTS**
Recent studies have applied the "biomarker approach" to assess deleterious effects in biological systems [8, 16]. However, there is still some debate relating to their definition and, especially, their use in environmental risk assessment [17, 18]. According to McCarthy and Shugart [19], biomarkers are defined as measurements of body fluids, cells, or tissues, that indicate in biochemical or cellular terms, the presence of contaminants (exposure biomarkers) or the magnitude of the host response (effect biomarkers). Owing to the short time of response, biomarkers are used as early warning signals of biological effects caused by environmental pollutants in order to predict changes at higher levels of biological organisation; *i.e.* populations, communities or ecosystems [5]. In general, responses at lower biological organisation levels are more specific, sensitive, reproducible and easier to determine, but more difficult to relate with ecological changes. Conversely, responses at higher biological organisation levels are indicative directly of ecosystem health; hence, more relevant to environmental management. However, they are more difficult to determine, less specific and only manifest at a late stage, when environmental damages have already occurred [20, 21].
**Figure 1:** Relationships between pollutants and their direct and indirect effects on the marine environment along a gradient of ecological relevance and spatiotemporal scale, together with the operational framework for establishing causal-effect relationships using multiple lines of evidence (adapted and modified from Clements [14] and Adams [15]).
The biomarker approach should be applied as a 'multi-marker' approach, since the use of a single biomarker in environmental studies does not reflect the integrated response of the organism [5, 6]. Furthermore, for each biomarker, specificity, temporal relevance, inter- and intra- individual variability, baseline values and the existence of confounding biotic and abiotic factors should be defined before its application [8].
Despite the above mentioned shortcomings, biomarkers could be a very valuable tool to determine cause-effect relationships, in the assessment of environmental pollution in the operational and investigative monitorings defined in the European Water Framework Directive (WFD) [22]. Thus, against the background of the development of the European Marine Strategy and of relevant Directives, the biomarker approach has been recommended and introduced in monitoring programmes for the surveillance of the marine environment, as an approach that complements the use of diagnostic chemical analysis with the assessment of biological effects [7, 23].
Over the last 20 years, an increasing number of ecotoxicological papers or research dealing with pollutants have focused on the study of causality between pressure of xenobiotics and responses at the molecular, cellular and tissue level. The possible effects of various contaminants on the health status of several coastal organisms are reviewed herein. Metal toxicity occurs when the rate of metal uptake into the body exceeds the combined rate of excretion and detoxification of available metal [24]. The first measurable effect of metals is on the inducibility of metallothionein (MT) genes and MT protein synthesis [25, 26]. In mussels, two families of MT isoforms have been characterised recently [27]; it has been suggested that these isoforms may accomplish different functions in presence of different metals and stressors. Whereas the monomeric MT10 isoforms have been reported to be synthesised constitutively and involved in essential-metal regulation, particularly of Cu and Zn, the dimeric MT20 isoforms are inducible and are involved mainly in Cd and hydroxyl radical detoxification [26, 28].
Organic pollutants are taken up readily into the tissues of aquatic organisms; here, biotransformation via Phase I (functionalisation) and Phase II (conjugation) metabolism can in part, determine the fate and toxicity of the contaminants [29]. Of central importance to Phase I metabolism is the mixed function oxygenase system, whose terminal component cytochrome P450 (CYP) exists as a superfamily of proteins, capable of oxidising a wide variety of substrates in numerous aquatic species [30]. One specific form, CYP1A1, can be estimated in fish liver by the measurement of 7-ethoxyresorufin-O-deetilase (EROD) activity and has been used as a biomarker of exposure to organic aromatic xenobiotics [23, 31]. In contrast with fish, marine invertebrates, especially mussels, have a low ability to metabolise organic pollutants, due to the lack of efficient isoforms of P450, which are only found in terrestrial organisms. Thus, one of the most important sentinel species (mussels) does not possess this important biotransformation system, discouraging the utilisation of EROD induction in pollution biomonitoring programmes, using mussels as sentinels. However, recent studies have suggested that the assessment of peroxisome proliferation and multixenobiotic resistance (MXR) may be more appropriate, as biomarkers of exposure to organic xenobiotic compounds in molluscs [32]. In this respect, both laboratory and field studies have shown that organic xenobiotics (such as phthalate ester plasticizers, Polycyclic Aromatic Hydrocarbons (PAHs) and oil derivatives, Polychlorinated Biphenyls (PCBs), certain pesticides, bleached Kraft pulp and paper mill effluents, alkylphenols and estrogens) all provoke peroxisome proliferation in aquatic organisms [33, 34]. Accordingly, MXR-related gene expression has been detected, either in terms of induction or inhibition, in molluscs exposed to anthropogenic organic pollutants [35, 36].
It is important to note that biomarkers can be affected by mixtures of different chemicals present in the field, giving rise to additive, synergistic and/or antagonistic effects. Several studies have demonstrated decreases in the activities of several enzyme markers of organic compounds in molluscs from polluted sites, probably due to complex interactions occurring in the marine environment between mixtures of pollutants [16, 37]. Similarly, acetylcholinesterase activities, inhibited by some pesticides (in particular carbamates and organophosphorus compounds), can be inhibited also by metals [38].
At a cellular level, tolerance to metals is based upon sequestration by a range of cellular ligands, such as MTs, followed by compartmentalisation within lysosomes [25]. Thus, metals bound to metal-binding proteins may enter lysosomes, and follow the catabolic pathway as would any other cellular protein. However, excessive concentrations of metals can cause alterations of structure, permeability and integrity of the lysosomal membrane when storage capacity of the lysosomes is overloaded [39, 40]. Impairment of lysosomal functions and, thereby, of food assimilation, can result in severe alterations in the nutritional status of the cells and whole organisms, and could be indicative of disturbed health. Previous studies have reported lysosomal enlargement, increased intralysosomal accumulation of metals and enhanced production of lipofuscins in metal exposed mussels, as a detoxification mechanism to minimise toxic effects of excess metals [28, 41, 42]. Additionally, a decrease in lysosomal membrane stability has been documented upon exposure to a progressively higher Cu concentration in laboratory experiments [43, 44] and in a transplant study in a copper mine [45].
Nevertheless, lysosomal responses are not metal-specific and, thus, are considered as a biomarker of general stress [46]. For instance, destabilised lysosomal membranes have been found also in marine organisms exposed to organic compounds, or collected from sites with mixture of contaminants [16, 47, 48]. Likewise, enhanced lipofuscin deposition has been observed in the digestive cells of mussels exposed to PAHs under both laboratory and field conditions [47, 49]. Intralysosomal neutral lipid accumulation in aquatic organisms (*i.e. Mytilus edulis*) has been reported upon exposure to organic chemicals in laboratory experiments [50, 51] and in sites polluted with oil derivatives [41, 52].
There is evidence [48] that reactive oxygen species (ROS) are formed in the presence of a wide range of contaminants, such as metals (e.g. Cu, Fe), PCBs and some pesticides, provoking alterations in proteins, DNA and membrane structures/functions. ROS are detoxified by antioxidant enzymes and scavenger molecules. In marine mussels caged for 4 weeks, in an industrialised harbour of north-west Italy, a biphasic trend for single antioxidants (catalase, glutathione *S*-transferases, glutathione reductase, total glutathione) and the total oxyradical scavenging capacity was shown. There was no variation or increase during the first 2 weeks of exposure to the polluted site, followed by a progressive decrease up to a severe depletion of ROS in the final part of the experiment [48]. The decreased capacity to neutralise specific ROS has been shown to correlate with the occurrence of alterations at various sub-cellular targets, including lysosomal membranes and DNA [53, 54]. Oxidative DNA damage, revealed as high levels of the mutagen 8-oxo-7,8-dihydro-2'-deoxyguanosine and lipid peroxidation, measured in terms of high malondialdehyde levels, has been found in mussels from polluted areas [55, 56].
Genotoxic compounds, such as persistent organic pollutants (POPs), can alter the integrity of DNA structure, either directly or through their metabolites [57], causing mutagenesis [58]. Biomarkers of genotoxicity include DNA damage, which is based upon potentially pre-mutagenic lesions (such as DNA adducts, base modifications, DNA-DNA and DNA-proteins cross-linking and DNA strand breaks) and chromosomal damage [59]. The presence of micronuclei is an indicator of chromosome breakage or chromosome loss [60]; this technique has been used extensively in invertebrates [61, 62]. Caged mussels exposed to seawater polluted by aromatic hydrocarbons, displayed a continuous increase of micronuclei frequency in gill cells, reaching a plateau after a month of caging [63]. The incidence of micronuclei has also been linked to the induction of leukaemia cells in the clam *Mya arenaria*, suggesting that the micronucleus test is a very good indicator of the potentially life-threatening consequences of genotoxic exposure [64].
The effects of environmental pollution have been identified also at tissue level and, therefore, histopathological changes in target tissues have been proved to be sensitive markers of health status in aquatic organisms [65, 66]. In bivalve molluscs, highly significant correlations between tissue pathologies and contaminants (Pb, Hg and PCBs) have been observed in mussels from the east coast of the USA [67]. In the digestive gland of mussels exposure to Cu, Cd and the water-accommodated fraction of different oils induced alterations in cell type ratios of digestive tubules, with basophilic cells increasing in number, in relation to digestive cells [28, 68]. This morphological change of digestive tubules has been observed, accompanied by atrophy of the digestive epithelium in molluscs [45, 69], apparently involving augmented autophagic processes [70]. Following the *Prestige* oil spill [69] that occurred in November 2002 in the Atlantic, north-west coast of Spain, digestive gland atrophy was demonstrated in mussels, indicating disturbed health, due partly to disturbances in digestion and metabolism [71]. Another severe tissue alteration, the incidence of granulocytomas, has been observed in molluscs from highly polluted areas [72].
There is a growing concern that chemicals in the environment, either natural or synthetic, can interact with the endocrine system, causing reproductive disturbances in aquatic organisms that may affect recruitment and lead eventually to deleterious population effects [73, 74]. Endocrine–disrupting chemicals such as phytoestrogens, alkylphenols, synthetic estrogenic hormones and bisphenol A mimic estrogenic hormones and thus cause estrogenic or feminising effects, whereas other chemicals such as tri-butyltin (TBT), used in antifouling paints for ships, and synthetic androgenic hormones cause androgenic effects [73, 75]. In invertebrates, the endocrine regulation of reproduction and development is not as clear as in vertebrates [76]. Vitellogenin (VTG)-like proteins, yolk proteins produced only in sexually mature females, have been observed in male molluscs exposed to different xenoestrogens such as alkylphenols [34, 77, 78]. In contrast, reduced VTG-like protein levels have been described in female molluscs inhabiting PAH contaminated sites or, after exposure to North Sea oil, indicating a possible anti-estrogenic effect of PAHs [34, 79]. Moreover, exposure of females to certain metals, such as Cu, provoked an increase in VTGlike protein levels and accelerated spawning, maybe as a result of possible acute toxic effects, or as an effect on hormone regulation of gamete development [45, 80]. These controversial results in females highlight the need for more basic research to understand biomarker responses, before their implementation in monitoring studies. On the other hand, it should be noted that the development of more advanced techniques for biomarker quantification could also be helpful in the interpretation of the multi-marker approach.
Overall, the utility of stress indices, based upon molecular, cellular and tissue responses in sentinel species, could provide a comprehensive indication of the impact of chemical pollutants in coastal marine environments. However, the need to increase the knowledge on biomarker baseline values and confounding factors is also highlighted.
# **BIOLOGICAL EFFECTS AT ORGANISM AND POPULATION LEVEL**
The impacts of metallic and organic pollutants on marine biota cover a plethora of direct and indirect effects on organisms and populations. These effects have been reported in a wide variety of biota such as bacteria, fungi, plants and animals.
Among the inorganic metallic pollutants, Cd, Cu, Cr, Pb, Hg, Ni and Zn are some of the best studied in terms of speciation, toxicity, bioavailability or bioaccumulation in marine ecosystems. Other metals with toxic effects in marine communities are Al, Sb, As, Se or Ag. Unlike other contaminants, all of these metals occur naturally in the environment; some of them (e.g. Cu and Zn) have essential functions in several biota at low concentrations. The degree of increase in levels of these metals in the environment, with respect to the background levels, and the degree to which they have toxic effects on biota, depends upon a wide number of factors such as: (i) their geochemical behaviour; the physiology and condition of the target biota; (ii) chemical speciation and; (iii) the presence of other toxicants, or environmental conditions [81]. In order to assess the lethal responses of marine biota to the metallic pollution, several approaches have been carried out since the 20th Century. Laboratory bioassays have permitted comparisons of toxicity between different pollutants and among different species. Moreover, these bioassays have shown that the first developmental stages of several species of invertebrates are highly sensitive to several toxicants [82, 83]. Research has also been carried out in order to evaluate sublethal responses in marine biota to inorganic metals; this has included effects on growth, sexual maturity, diseases, luminescence, metabolism, *etc.*
The research undertaken on the biological effects of the organometallic pollutants has shown that some of these compounds are substantially more toxic than the inorganic metal to marine organisms. As an example, methylmercury is substantially more toxic to several biota than inorganic mercury, because it is more efficiently transported across the gut [84] and because of its ability to diffuse through lypophilic media and cross cell membranes. Another highly researched pollutant is TBT, which has toxic effects in a wide variety of biota, whereas inorganic tin is less toxic. TBT effects include lethal toxicity and effects on growth, reproduction, physiology, and behaviour [85]. Several of the negative effects are due to interferences with the endocrine function, and manifested as the imposex phenomenon. In dioecious gastropods, imposex presents as the superimposition of male organs in females [86]. In some species, such as the dogwhelk *Nucella lapillus,* this masculinisation has provoked the local extinction of populations in several polluted areas, during the 1970s and 1980s [87]. Because TBT has negative effects at a very low concentration (e.g. ~ 1 ng Sn/L [87]), it has been considered as one of the most toxic xenobiotics ever produced and introduced deliberately into the environment [88]. At those sites where measures were taken to reduce the input of TBT into marine ecosystems, the previously affected species have shown a recovery [89].
PAHs are relatively ubiquitous organic molecules, within marine ecosystems. There exists more than 100 different PAHs, with both natural (e.g., forest fires) and anthropogenic (e.g., burning processes or oil spills) inputs into the ecosystems. The fact that PAHs occur generally as complex mixtures in the environment makes the evaluation of toxicity more difficult [90, 91]. This difficulty is increased by the fact that some have shown higher toxicity as a consequence of photoactivation [92]. Although there is a need for research to clarify the biological effects of PAHs in marine ecosystems [90], toxic (carcinogenic) effects have been found in phytoplankton, zooplankton, invertebrates and vertebrates because they form DNA adducts. Most of these toxic effects are related to a reaction with macromolecules (like nucleic acids or proteins), or an interaction with lipids in cell membranes or other cellular constituents [93, 94]. These processes can cause diseases (including tumours) and negative effects on immunosystems, growth, reproduction and behaviour.
Chlorinated pollutants (*i.e.* PCBs, organochlorine pesticides, herbicides and fungicides) have effects on the physiology and behaviour of marine biota. Almost all of these compounds are not known to occur naturally in the environment and some are very persistent. Negative effects were found on growth, reproduction, luminescence, metabolism, *etc.* Some of these effects are due to endocrine disruption. Although some of these effects in marine organisms are relatively well known, others are still poorly understood [95].
Pollution effects on organisms can imply consequences at the population level. These can occur in different degrees, from changes in the population dynamics or genetic diversity, to the local extinction (as the above-mentioned case of some gastropod extinctions due to TBT pollution). The linkage between pollution and population can be more complex to assess than is the case for individual organisms. Nevertheless, the recent development of genetic techniques presents some assessments of this issue [96]. Populations might respond with increased genetic variation (e.g. resulting from new mutations), or decreased genetic variation (e.g. resulting from population bottlenecks) [97]. The loss of genetic diversity can imply a reduced adaptive potential of populations to changes in environmental characteristics or to the presence of new pollutants [96, 98, 99]. Because of the complexity of the processes involved, the loss of genetic diversity is not predictable based solely upon knowledge of the mechanism of toxicity of the chemical contaminants and the life cycle of the biota [97].
Several studies of the effects of contaminants in populations have been carried out with meiobenthic species. This approach is related to the fact that several meiobenthic species have short generation times; this permits a study for the duration of a full life cycle. Studies carried out near offshore platforms, combined with laboratory experiments, found genetic diversity in meiobenthic copepods (*Nitocra lacustris, Cletodes sp., Enhydrosoma pericoense, Normanella* sp.*, Robertsonia* sp.*, and Tachidiella* sp.) correlated inversely with the degree of sediment contaminants (hydrocarbons); however, other causes could be attributed [100, 101]. Laboratory assays have found that the exposure to polybrominated diphenyl ether (BDE-47) and copper can reduce genetic diversity and alter genotype composition without affecting population abundance of some meiobenthic species [102, 103].
Genetic studies have also been carried out on macrobenthic taxa. Research undertaken on mussel, barnacle, prawn and isopod species suggests that pollution may reduce genetic diversity, but other causes cannot be discarded. Moreover, long-term exposure to metal pollution does not necessarily result in decreased genetic diversity [104, 105].
The exposure of the populations to pollutants can select rapid genetic changes or small-scale evolutionary processes associated with a genetically-inherited increase in tolerance to the pollutants (a process called microevolution [106]). This microevolution can buffer, partially, the effects of pollution in populations. As an example, an estuarine oligochaete species (*Limnodrilus hoffmeisteri*) was found to be more resistant to cadmium and nickel pollution following two generations of selection in laboratory assays [107].
# **BIOASSAY STUDIES AT ORGANISM LEVEL**
An important component of pollutants arriving in the marine environment is retained in the bottom sediments, where they can reach concentrations several orders of magnitude higher than in the overlying waters [108]. Hence, in coastal systems, bottom sediments need to be characterized in environmental studies. Historically, such characterization has been limited to physicochemical analysis [109]. However, the chemical analyses by themselves do not provide evidence of biological effects on organisms; therefore, they do not assist in confirming the effect that they induce on ecosystems [110, 111]. Thus, toxicity tests on marine systems, among other biological methods of ecotoxicological evaluation, are necessary as a complement to physicochemical analyses to assess the potential effects of pollutants on organisms and biological communities [109].
Ecotoxicology, as the science that studies all of the adverse biochemically-mediated effects of all chemicals on all living organisms, including all their interactions within organisms and among species in the environment [112], is applied to the evaluation of the effects caused by pollution on the marine environment by means of bioassays.
# **Bioassays**
Bioassays are used to evaluate the environmental quality through the measures of toxicity in natural samples, and to predict the ecological risk of contamination. These tests show numerous advantages: (i) the test organisms only respond to the bioavailable fraction of a pollutant; (ii) also, as opposed to chemical analyses that detect only previously well-known compounds, bioassays can help identify new toxic elements whose noxious effects had not been described previously [113]. In addition, (iii) they offer quantitative information on sediment toxicity, which provides a basis for discriminating between impacted and unimpacted sites. The results from these tests are also relevant ecologically because they commonly use resident species. As such, the tests undertaken provide a way to compare the sensitivities of different organisms.
The toxicity tests under laboratory conditions are carried out with the purpose of establishing any relationship between exposure of pollutants and the effects caused on individual organisms. By means of these tests, dose-response relationships are established, determining the relationship between the concentration or dose of the toxin and the noxious effect on an organism. Following the proposal of the use of toxicity tests as an appropriate tool for the valuation of marine pollution [83], they have become a fundamental part of the evaluation of environmental risk; they provide a direct measure of toxic adverse effects, complementing the traditional physicochemical measures [114].
At present, the list of bioassays to assess toxic effects of exposure to marine bottom sediments, as a measure of risk to populations is very extensive [115-118], given the multiple combinations that exist between the elements and the conditions to be selected. These bioassays, carried out in highly defined, controlled and reproducible conditions, can be applied to total sediment, suspended sediment, elutriate, pore water and/or sediment extract. The response variables to be measured include long-term toxicity, acute toxicity, bioaccumulation, endocrine effects, effects on reproduction, carcinogenesis and mutagenesis. Different marine organisms belonging to different trophic levels (bacteria, algae, molluscs, echinoderms, annelids, fishes, *etc.*) and in different development phases can be used in bioassays. This approach offers a wide range of biological possibilities for investigation [119, 120].
The organisms used in these tests are selected on the basis of: (i) their sensitivity; (ii) their ecological, commercial or recreational relevance; (iii) their high availability and abundance; (iv) their ease of culture or maintenance in laboratory; (v) and the simplicity of the analysis of results [121-123]. Finally, the responses obtained in test species can be qualitative or quantitative, but should be unequivocal, easily observable, describable, measurable, biologically significant and reproducible [124].
In a typical bioassay, the response of an aquatic organism to a toxic substance is related to the toxic concentration in water/sediment, together with the time of exposure. A commonly used technique to measure the dose-response relationship requires exposure time to be held constant over a series of different concentrations in order to record the proportion of individuals that present a specific biological response; e.g. mortality-survival, fertilization-non fertilization, or mobility-non mobility. The final objective is to obtain a toxicity curve of a substance or compound for an organism that defines the relationship between concentration (dose) and response (see Chapter 1).
# **Toxicity Estimation**
To estimate the toxicity of a substance or sample, the EC50 is used extensively since it is a statistically-reliable measure. The EC50 is defined as the Effective Concentration that produces a specific effect on 50% of a population based upon experimental laboratory tests. It is used as a standard measure of toxicity and it permits a comparison of the toxicity of different compounds on an organism, or the toxicity of the same compound on different organisms. However, on the basis that the final objective of toxicological studies is to protect ecosystems, the EC50 alone is not sufficient. Therefore, it is necessary to obtain a second parameter that defines the toxicity threshold; *i.e.* the concentration above which they begin to show adverse effects. In this sense the NOEC (the highest experimental concentration in which the response does not present statistically-significant differences, with regard to the control) and LOEC (the lowest experimental concentration in which the observed response is significantly different from that of the control) are defined. These last reference parameters present a higher potential for utilisation from the point of view of their ecological application. Nevertheless, some debate exists presently on the suitability of NOEC and LOEC as estimates of the toxicity threshold because of their strong dependence on the experimental design, and an alternative frequently used is the concentration causing a lower level of effect, such as the EC10 [125, 126]. Moreover, the reason for not using LOEC and NOEC as endpoints in many instances (e.g. regulatory, water quality guidelines) is the statistical unreliability of the calculated values for such endpoints [125, 126].
# **Bioassays Limitations**
The laboratory bioassays have some limitations since they do not necessarily reproduce the range of potentially relevant environmental factors present in nature [127, 128]: (i) the species used are not necessarily part of the communities that inhabit the studied sediments and, as such, they may not be representative of the species found in the area of interest [129]; (ii) the laboratory conditions are controlled, whilst the factors (i.e. abiotic:. climate, temperature, hydrodynamics, quality of water and/or sediment; and biotic: development stage, reproductive state, health, presence of other individuals and/or species, etc.) are changing in the environment [125, 130]; (iii) biomagnification through trophic webs, effects of nutrients, habitat alteration, inter- and intraspecific predation or competition relationships are not taken into account [131]; and (iv) laboratory bioassays do not predict the indirect effects that often characterize the responses of the ecosystems to stress [132]. Therefore, although these bioassays have improved our understanding of the effects of pollutants, their results are difficult to extrapolate to the environment, because they lack ecological "realism" [133], since too many components exist in an ecosystem that make it impossible to predict accurately the effect that a toxic substance can exert. For example, the effect of a toxicant varies not only between species within an ecosystem, but also in the same species in different ecosystems. On the other hand, in natural environments, a substance may not produce adverse effects on a particular species, but does so in its predators or its food source, which influences finally the survival of the organism. Thus, in the absence of a thorough understanding of natural systems and their processes there is always a high degree of uncertainty [134]. Hence, it is necessary to obtain the highest amount of data on ecosystem dynamics, abiotic compartments and xenobiotics' toxic effect on the species. The analysis of these data permits the prediction, with a higher reliability, of the ecological risk associated with an episode of contamination (Environmental Risk Evaluation). Similarly, all this knowledge needs to be applied together for the restoration of ecological health in contaminated systems [135, 136]. This is the reason why ecotoxicological information is important when establishing environmental monitoring programmes [137].
Because of the necessity of obtaining more complete and useful information from an ecological perspective, testing on the basis of a single species has not been considered sufficient in recent years. When evaluating potentially contaminated samples from the environment, the use of a test battery, including species belonging to different habitats, development stages, multiple trophic and evolutionary levels, and (even) different times of exposure, is recommended as a more appropriate strategy [138, 139].
For the purpose of chemical registration, regulatory authorities require bioassays with individual chemicals.. These bioassays need to be conducted under standard conditions and there is no interest in reproducing the broad variability of natural ecosystems. On the other hand, bioassays with environmental samples are normally conducted in order to detect the presence of chemicals a priori unknown. In that case, the bioassay is a tool in the 'tool-box'. Once the hot-spots are identified, subsequent more expensive techniques may be applied to study the ecological effects.
# **EFFECTS ON COMMUNITIES AND ECOSYSTEMS**
There is general agreement amongst marine investigators that measuring a suite of indicators across levels of biological organization is often necessary to assess ecological integrity [14]. As the effects of pollutants on the marine environment may be identified at all levels of biological organization, these indicators should include biochemical, population, community, and ecosystem responses (Fig **1**). In previous sections, the most common approaches in assessing those effects at low organizational levels have been presented. In turn, warnings of pollutant effects at community and ecosystem level are scarcer. However, as ecological integrity requires the protection of a good structure and functional processes at those high levels [140, 141], demonstrating biochemical and physiological responses to pollutants may not be sufficient.
Following Clements [14], the key to predicting the effects of contaminants on communities and ecosystems, is to understand the underlying mechanisms. Thus, establishing a cause and effect relationship between stressors and responses at higher levels of organization is problematic. The structure and functioning of these systems may be altered for many reasons other than contaminant exposure. Hence, Clements [14] suggests that one of the major goals of ecotoxicology is to develop an improved mechanistic understanding of ecologically-significant responses, to contaminants.
A review of the effects of pollutants on aquatic ecosystems in different parts of the world can be seen in Islam and Tanaka [142]. These authors describe a decrease in species diversity, changes in community structure, degradation of habitats, decline in abundance and biomass, diminution in yield of marine resources, *etc.* Hence, Wolfe [143] and other authors (see also Fig. **1**) have systematized the bioindicators of pollution for marine monitoring programmes at community and ecosystem levels. These approaches include measurements of abundance, biomass, richness, dominance, similarity, ratio opportunistic:sensitive species, age-size spectra, trophic interactions, energy flow, productivity, and the loss of goods and services. The response to pollutants (metals, organic compounds), of some of these indicators, is examined below.
# **Richness, Diversity and Evenness**
In a recent review, Johnston and Roberts [144] make a meta-analysis of 216 papers in which the most frequently used measures of diversity and evenness were species richness (number of species per unit area), the Shannon– Wiener index and Pielou evenness (Margalef's richness and Simpson's diversity were used occasionally). The vast majority of the contributions concluded that there were significant negative effects of pollution upon species richness, with occasional increases in species richness and diversity associated with nutrient enrichment. Only 20% of the papers did not detect the effects of contamination upon diversity. When an effect was detected, its response ratio based upon species richness and Shannon–Wiener diversity tended to be greater than reductions in the Pielou evenness. Hence, a 30–50% reduction in species richness and diversity were identified in all habitats exposed to all contaminant types. In turn, Dauvin [145] does not consider species richness a good indicator of disturbance in estuaries, due to marked changes linked to salinity gradients.
There is also good evidence that offshore discharges of oil-based drilling fluids by the oil and gas industry have caused reductions in benthic species diversity or other changes in community structure at distances of <1–3 km from drilling locations [146, 147].
Although there is a large degree of variability when comparing laboratory toxicity values and benthic measures, Long *et al.* [148, 149] have demonstrated that (from almost 1500 samples) in 92% of the samples classified as toxic, at least one measure of benthic diversity or abundance was less than 50% of the average reference value. These findings have been used in the derivation of sediment quality guidelines (SQGs), as commented below.
Pollution of marine habitats has been associated with a reduction in biodiversity, either as a result of reduced species richness, increased dominance of tolerant species (i.e. decreased evenness), or a combination of both factors following the Pearson and Rosenberg paradigm [150]. However, the abovementioned meta-analysis [144] indicates a remarkable similarity in the response ratios across habitat and contaminant types. Hence, pollution was never associated with the complete exclusion of life from a particular location (commonly 50–70% of species were able to tolerate the contaminant load). In some cases, estuarine communities showed very high abundance and biomass values together with very high levels of contaminants [145, 151]. This paradox is explained by the delay between the period with the maximum runoff and the maximum contaminant input (at the end of autumn and during the winter) and the recruitment period for the dominant species (throughout the spring and summer), together with the absence of anoxic conditions.
# **Ratio Opportunistic/Sensitive Species**
From the previous Section, it is clear that the identity of pollution-tolerant and –intolerant species is of great interest. Pollution-tolerant and opportunistic species have long been recognized as potential bioindicators of impacted systems [130, 152, 153]. Macrobenthic communities respond to pollution by means of different adaptive strategies [130]: (i) r-selected species, with short life-span, fast growth, early sexual maturation and larvae throughout the year; (ii) k-selected species, with relatively long life, slow growth and high biomass; and (iii) T, stress-tolerant species, not affected by alterations.
These strategies have been used in developing different indices that can be used to assess the environmental quality status in estuarine and coastal systems (see reviews in [154, 155]). From the high number of indices based upon sensitive/opportunistic ratio of species, probably the most successful are AZTI Marine Biotic Index (AMBI) [156] and the Benthic Quality Index (BQI) [157], which are on the basis of many other methodologies, most of them used within the WFD [158].
From the extensive publications using these indices (and especially from that of AMBI), it can be observed that increases in metals [159-161], organic compounds and TBTs [146, 162, 163] produce a decrease in benthic community quality detected by this index. Hence, a primary mechanism driving these changes, as a result of exposure to contaminants, is the elimination of sensitive species and the subsequent monopolization of resources by tolerant and opportunistic species [144].
In some cases such as at Restronguet Creek in the Fal estuary system [147], copper and zinc pollution from mining is strongly suspected to have caused the exclusion or restriction of several species of bivalves, including *Cerastoderma edule, Macoma balthica, Mytilus edulis* and *Scrobicularia plana*, as well as the changes in nematode metal tolerance. However, some species (such as *Nereis diversicolor*) are able to adapt to pollution. Normally, in these extreme situations of metal pollution, infaunal communities are dominated typically by metal-tolerant opportunistic deposit-feeding polychaetes [144]. In fact, sediment metal chemistry and benthic infauna surveys undertaken over 33 years with sampling before, within and after tailings deposition from a metal (Pb, Zn) mine in Greenland [161] have shown dramatic changes of benthic fauna composition. Faunal recolonisation 15 years after closure of the mine was slow. Of the metals, Pb had the greatest impact, with deterioration of benthic communities above a threshold of 200 mg/kg, decreasing diversity and dominance of sensitive species, and increasing tolerant and opportunistic species; *i.e.* long-lasting effects on the biological system.
In turn, from the relatively few data on hard-bottom substrata communities, it has been suggested that macroalgal communities are relatively resilient to pollution [144]. However, some research has shown metal- and nutrientimpacted rocky shores to contain degraded communities of macroalgae. Opportunistic algal species with rapid growth rates, including *Ulva* and *Enteromorpha* dominated*,* replacing relatively diverse communities of large perennial algae and sessile filter feeders seen in more pristine areas [164-168].
# **Trophic and Other Interactions**
Pollution may affect diverse components of the ecosystem, through primary species structuring the community. Hence, Roberts *et al.* [169] review the ways in which the contamination of biogenic habitats may affect other compartments (*i.e.* epifauna), describing four pathways: (i) colonisation by mobile fauna; (ii) inhibition of larval settlement; (iii) feeding by herbivores and predators; and (iv) post-ingestive effects on fauna.
The effects of habitat-bound contaminants, on the abundance of epifauna, may be driven by the behavioural responses of dispersing organisms [169]. For instance, recruitment of epifauna is reduced to macroalgae experimentally-spiked with copper as a result of behavioural preferences for uncontaminated algal hosts. Thus, exposure to habitat-bound contaminants is likely to be spatially complex. Hence, small-scale variation in contaminant concentrations interacts with variation between organisms in their ability to disperse among alternate habitats.
In addition, the accumulation of metals by macroalgae and seagrasses represents a potentially important pathway of contaminant exposure to grazing organisms (herbivores and detritivores), which are responsible for much of the transference of metals to higher trophic levels [169]. In fact, the algae *Ulva lactuca* and *Enteromorpha intestinalis*, collected from contaminated sites and used to feed herbivorous gastropods, produced complete mortality of the latter organisms within 1–4 weeks of continuous dietary exposure [170]. Similar effects have been described at other trophic levels [169].
# **Interactions of Pressures within the Ecosystems**
The potential for interactions to occur between chemical contaminants and habitat factors (e.g., food and habitat availability) has been identified as being important for understanding the ecological effects of pollutants [171]. Thrush *et al.* [171] demonstrated a way of determining likely interactions and also that multiplicative effects, such as stressors, frequently interact across environmental gradients. This pattern suggests a strong role for regression-based analysis of field gradients in the determination of contaminant effects. This study highlights the potential variation in response to metal contaminants across ecological landscapes; it provides an insight into fitting ecotoxicological responses into ecosystems. The complex community effects, mediated by impacts on foundation or key species and ecosystem engineers, have been assessed by Thrush *et al.* [171], highlighting the need for improved integration of ecological patterns with contaminant-stress responses.
Moreover, Crain *et al.* [172] in an analysis of 171 studies that manipulated two or more stressors in marine and coastal systems, found that cumulative effects in individual studies were additive (26%), synergistic (36%), and antagonistic (38%), which is very close to what would be expected from a random distribution (33-33-33%). The overall interaction effect across all of the studies was synergistic, but interaction type varied in relation to response level (antagonistic for community, synergistic for population), trophic level (antagonistic for autotrophs, synergistic for heterotrophs), and specific stressor pair (seven pairs additive, three pairs each synergistic and antagonistic). Addition of a third stressor changed the interaction effects significantly in two-thirds of all of the cases; it doubled
the number of synergistic interactions. Hence, although pollutants can affect communities and ecosystems, their effects can be reinforced when other pressures or stressors (*i.e.* nutrient inputs, habitat loss, hypoxia, *etc.*) are present. Generally, organisms living under conditions close to their environmental tolerance limits appeared to be more vulnerable to additional chemical stress [173].
# **MEASURING POLLUTANT EFFECTS IN AN INTEGRATIVE ASSESSMENT**
Much discussion has taken place about the lack of a coherent terminology to differentiate the various assessment types and the diverse nature of aquatic environmental integrative tools and methods in assessing ecological integrity [141, 174]. Following Borja *et al.* [141], these approaches can be divided into two categories: (i) those evaluating risk and state of a particular system (*sensu* the Drivers-Pressures-State-Impacts-Response (DPSIR) approach); and (ii) those assessing the ecological integrity status of the whole ecosystem under an ecosystem-based approach.
# **Evaluating the Risk and State of a System**
The ecotoxicological effect measurements must be used, within the context of ecological risk assessment (ERA), as a tool to assess the likelihood of harm being caused to ecosystems, or their components through exposure to a specific concentration of a chemical.
Among the approaches used to overcome the limitations shown above, some authors propose the use of multispecies tests, or using different compartments of the system (*i.e.* chemical analysis, bioassays, impacts on benthic communities) in the assessment. This approach is developed within the context of an integrative assessment, considering several lines of evidence (LOE); *i.e.* sediment contamination, toxicity and benthic fauna.
Another relatively recent approach is the weight of evidence (WOE) approach, which is the result of combining different measures of environmental quality to establish an overall assessment of environmental health. The philosophy behind WOE is a preponderance/burden of evidence approach, where the conclusions drawn from individual components are considered not as a sum of these components, but relative to one another [175]. WOE determination incorporates judgements concerning the quality, extent, and congruence of the data contained in the different LOE. It includes also observational (e.g. ecology) and investigative or manipulative (e.g., toxicology used to determine cause-and-effect) components. Ideally, any WOE framework will be easily understandable by lay personnel or decision-makers; it will also appropriately differentiate between hazard (the possibility of impact) and risk (the probability of impact) [176].
One of the first sediment quality WOE frameworks was the sediment quality triad (SQT). The triad concept was conceived more than 20 years ago by Long and Chapman [177] to provide a sediment quality evaluation based upon three components: (i) chemistry, to determine chemical contamination; (ii) bioassays to evaluate toxicity; and (iii) benthic community structure to determine the status of resident fauna exposed to the sediment contaminants. These three original components provide the basis for the SQT, or contaminated sediment risk assessment [178]. However, the traditional SQT is based on correlation, not causation; it can provide definitive conclusions regarding the pollution status of contaminated sediments, but cannot provide definitive conclusions in all cases, and cannot derive causation without further studies. Hence, the traditional SQT can be considered as a screening-level ERA, with causation examined at a higher tier [179]. Hence, the SQT needs to include additional LOE to address all aspects of ERA.
Individual LOE involved in contaminated sediments evaluation should include [175]: (i) measures of sediment chemistry to determine the level and extent of pollution and modifying factors (e.g., grain size, total organic carbon) compared to SQGs, and to answer the question "are contaminants present at levels of concern?"; (ii) measures of resident benthic community structure to determine whether community structure has been altered, possibly due to pollution; and (iii) measures of toxicity to determine whether the contaminated sediments are affecting the biota.
Additional LOE can include: (i) measures of biomagnification, usually involving measurements of body burdens in sediment-dwelling invertebrates, and food chain modelling to answer the question "are any contaminant of concern capable of biomagnifying and likely to do so?"; (ii) measures of exposure such as biomarkers or body burdens (bioaccumulation) to determine which sediment contaminants, if any, are bioavailable and to try to determine causation; (iii) toxicity identification evaluations (TIE) to attempt to assign causation; and (iv) determinations of sediment stability to determine whether only surficial sediments should be evaluated or whether deeper sediments, which may be exposed during storm or other events, need to be evaluated by answering the question "is the sediment stable or is it prone to erosion resulting in exposure of deeper, more contaminated sediments and/or contamination down-current?"
In summary, a battery of different LOE selected for specific purposes should be developed, maximizing flexibility in the use of WOE within a wide variety of situations and locations which exist in the environment.
For sediment risk assessment, the recommended WOE is the tabular decision matrix (TDM); this is the most effective and logical basis for presenting WOE in a manner readily understandable. TDM was used under the SQT first proposed by Chapman [125] and improved by other authors [176, 180, 181]. Such a matrix must be based on a strong quantitative, statistical evaluation / summarisation prior to merging into more qualitative matrix tables. Each LOE is established on the basis of a graduation (a scoring system) to rate each measurement endpoint as indicative, moderate, or negligible/low ecological risk. These LOE are summarized in SQGs, toxicity test results and biotic indices. The classification of the toxicity tests to use in the ordinal ranking scheme is based upon comparison with sediment toxicity guidelines and/or standards established previously in national ring, or intercalibration tests using the same species of organisms [182]. The integration of data-reducing techniques is very useful to incorporate into a tabular matrix, as emphasised by Chapman [183]. Some steps in the SQT are assigned more weight than others based upon expert knowledge of the sediment assessment, system behaviour and factors interpretation computed from Best Professional Judgement. As mentioned previously, to develop a tabular matrix, a ranking scheme must be applied for categorisation. An example of this application is shown in Tables **2** and **3** [184].
**Table 2:** Ordinal ranking scheme applied for 'weight of evidence' categorisation. PIAE = Potential Impact for Adverse Effects.
1 For sediment pollution, two guideline values, obtained from previous studies undertaken in Spanish Ports [185], are used: Action Level 1 (AL1) and Action Level 2 (AL2).
2 Toxic thresholds for amphipods from DelValls *et al.* [182], whilst benthic reference values have been derived for Basque Country ecosystems based on Borja *et al.* [156].
In WOE assessments, there is no single correct way to relate sets of variables and other approaches are considered, such as multivariate analyses (e.g., Principal Component Analysis) which are used typically in WOE determinations [183, 186].
Some authors [183, 187] have also detailed a tiered scheme for WOE components. The tier testing approaches are recommended for regulatory purposes as it permits keeping risk assessment cost-effective and feasible. It allows an assessor to undertake only as much sampling and analysis as are needed to come to a reasonable decision. Moving through the tiers (steps), one moves from a broad to a more focused scope and from general benchmarks to more detailed, directed tests [188]. Although WOE assessments (such as the SQT) have improved since their initial development, future WOE applications should consider specific LOE in terms of risk assessment, ensuring that both exposure and effect assessment are addressed adequately, as are both causation and ecological relevance. As emphasised by Chapman *et al.* [180], WOE assessment (such as the SQT) should not be used to develop a single numerical index; any simplification of WOE should be site- or region-specific, not generic.
**Table 3:** Example of tabular matrix with the Sediment Quality Triad (Lines of Evidence, for management), using a port from the Basque Country (northern Spain).
Note: use of symbols provides for a convenient and rapid visual assessment of all endpoint results, as well as assessment of the concordance among endpoints, for a given site. Symbols indicate: +, contamination, effect or alteration are observed; +/-, moderate contamination, effect or alteration; -, no contamination, effect or alteration.
PAH = Polycyclic Aromatic Hydrocarbons; PCB = Polychlorinated Biphenyls
1 Benthic community *in-situ* alteration using the AZTI Marine Biotic Index (AMBI) [156].
2 For overall risk assessment (see Table **2**): two moderate (+/-) equal to one positive. However, the components of WOE assessments will change as new, more ecologically-relevant measurement endpoints are discovered and applied. In particular, it is expected that: (i) particular WOE "tools" will be refined and validated; (ii) interpretative guidelines will be more fully developed; and (iii) chronic toxicity tests and community responses will be further incorporated into WOE assessments.
In summary, environmental quality, assuming the persistence of a suitable habitat, can be determined only by the responses, or condition of multiple (never single) measures undertaken as part of integrative assessments.
The uncertainty and high variability, inherent in both ecosystems and methods of measurement, require a burden of evidence approach. The WOE approaches are, and will continue to be, most useful where they are flexible and responsive to study goals, ecological realities, and social concerns [175].
# **Assessing the Ecological Integrity of an Ecosystem**
One criticism of the sole use of diversity indices, as measures of ecological impacts is that such measures do not consider alterations to the structure of communities; therefore, they may mask more effects than they elucidate [144]. Following these authors, indices which consider taxonomic relatedness and multivariate analyses of community structure are more sensitive and powerful means of detecting ecological impacts than studies that consider diversity and species richness (or other indices) alone. These criticisms, together with the interactions and synergistic effects described above, have led to the use of multiple indices and integrative methods to assess ecological integrity in marine waters [189]. These methods take into account the concept of environmental or ecological status, which includes the structure, function and processes of marine ecosystems, bringing together natural physical, chemical, physiographic, geographic and climatic factors; subsequently, integrating these conditions within the anthropogenic impacts and human activities in the assessment.
Hence, the environmental status concept defines quality in an integrative way, using several biological parameters (*i.e.* macroalgae, macroinvertebrates, *etc.*) together with physico-chemical and pollution elements, including ecosystem attributes (such as food web dynamics, species diversity, and the distribution of life histories) that are not direct biological properties but functions of the entire ecosystem. They are important because they provide information about the functioning and status of the ecosystem; they have been perceived widely as additional and potentially useful indicators of environmental status [141]. This approach is intended to permit an assessment of the ecological status at the ecosystem level ('ecosystem-based approach' or 'holistic approach' methodologies), more effectively than can be carried out at a species or chemical level (*i.e.* quality objectives). However, there are few examples of pollutant impacts at the whole marine ecosystem level [142, 190], in an integrative way, because they are masked by other human pressures as commented upon above.
An overview of these kinds of integrative tools and methods in assessing ecological integrity in estuarine and coastal systems world-wide can be seen in Borja *et al.* [189]. Overall, the legislative measures world-wide tend to converge in defining environmental water quality in an integrative way. However, the degree of convergence is variable, based generally upon studies carried out in single systems, which do not permit generalisation.
In practical terms, managers and decision-makers need simple, but scientifically well-established methodologies capable of demonstrating to the general public the evolution of a zone (estuary, coastal area, *etc.*), taking into account pollution and other human pressures or recovery processes [155, 189]; likewise, capable of guiding the implementation of successful management. Within this context, there is a major scientific challenge to develop tools to define adequately the scale and present condition of marine ecosystems in terms of biological performance, as well as to monitor changes through time, and similarly, to identify and address, through management, the causes of observed impairments [189].
Some of these challenges have been addresses in assessing ecological status in large ecosystems in the USA and Europe [166, 191]. However, with the success of these tools come additional challenges. The proliferation of indices to assess the status adds an element of confusion back into what they had been intended to simplify [140]. Some of the confusion arises because of the different processes used for developing, calibrating and validating methods in different regions; this leads to inconsistencies in assessment across regions. Additional confusion results from indices developed for multiple types of biota, providing managers with multiple, often conflicting, answers for a single water body [140]. Whereas the last decade was characterized by an explosion of methods, the next decade should be one of consolidation and agreement [140].
# **REFERENCES**
<sup>© 2011</sup> The Author(s). Published by Bentham Science Publisher. This is an open access chapter published under CC BY 4.0 https://creativecommons.org/licenses/by/4.0/legalcode
# **CHAPTER 9**
# **Chemical Pollution on Coral Reefs: Exposure and Ecological Effects**
**Joost W. van Dam1,2,3,\*, Andrew P. Negri1 , Sven Uthicke1 and Jochen F. Mueller<sup>3</sup>**
*1 Australian Institute of Marine Science, Townsville, Australia; <sup>2</sup> The University of Queensland, Centre for Marine Studies, St. Lucia, Australia and <sup>3</sup> The University of Queensland, National Research Centre for Environmental Toxicology, Coopers Plains, Australia.*
**Abstract:** In this chapter we review the effects of anthropogenically derived chemical pollutants on tropical coral reef ecosystems. A wide range of compounds, including pesticides, trace metals and petroleum hydrocarbons enter reef systems through various pathways and affect different reef species and/or life history stages. Tools for evaluation of chemical stress on coral reefs consist of molecular, biochemical, physiological and ecological bioindicators, providing information at organismal or community levels. This chapter collates and assesses available information on different chemical stressors in the marine environment and the effects on reef-building corals. Ecological effects from chemical stressors are strongly dependent on exposure characteristics. Three probable pollution scenarios are discussed and their individual properties evaluated. Short-term, pulse-like pollution events including oil spills or antifoulant deposition through ship groundings often have a direct and severe impact upon multiple trophic levels of the system. However, these events are typically localised and possibly irrelevant on an ecosystem-wide scale. In contrast, recurring pollution events such as input from river floods or chronic pollution from land runoff (e.g. sewage treatment effluent or herbicides), may exert subtle effects on lower trophic levels of the system, affecting species fitness and driving adaptation. Effects from recurring or chronic pollution are more likely to combine and interact with other environmental factors, but remain poorly understood. Over time, chronic sub-lethal stress may decrease resilience of reef organisms to other forms of environmental stress like elevated sea surface temperatures and ocean acidification.
# **CORAL REEF ECOSYSTEMS AND SYMBIOTIC PRODUCTION**
Coral reefs are biogenic structures that often contribute significantly to the seaward section of tropical shorelines, buffering the coast from wave action and erosion [1]. Coral reefs are among the most biologically diverse and productive systems in the world and many coastal communities depend upon these economically and culturally important ecosystems as a source of income or resources. Coral reefs are crucial to tropical fisheries and tourism and provide many island populations with primary building materials. Reef-related tourism alone generates vast revenues in some parts of the world. For example, in the 1990s, an estimated \$140 billion was generated by Caribbean reefs annually [2]. The ecological value of coral reefs is also enormous. There are over 600 species of calcifying corals, and these contribute directly to the habitat of thousands of species of tropical fish, algae and invertebrates. The physical protection offered by coral reefs enables formation and persistence of associated ecosystems such as seagrass beds and mangrove forests, allowing for the existence of essential habitats, hatching grounds and fisheries [3]. These ecosystems are in turn crucial for the reefs, as they play an important part in the sustenance of the marine foodweb by providing detrital matter that helps maintain the lower trophic levels of the reef food web, by functioning as a sediment trap limiting particulate matter reaching reefs and by acting as nursery grounds for many juvenile fish that find their way out to the reef in adulthood [1].
Reef-like structures have existed on earth for over 500 million years, with modern reefs developing around 250 million years ago [4]. Most coral reefs are located in tropical oceans within 30° of the equator with water temperatures ranging between 18 and 30 °C (Fig. **1**). Coral reefs are primarily formed by calcification processes of scleractinian (hard) corals and coralline algae, providing a structural basis for reef-dwelling organisms. The extraordinary productivity of coral reefs that may seem remarkable within a marine environment harbouring low nutrient concentrations can be explained by the photosynthetic contribution of intracellular microalgae to host tissues [5]. Reef building corals all host endosymbiotic dinoflagellates (zooxanthellae) of the genus *Symbiodinium*. Coral hosts profit from this mutualistic relationship by obtaining high-energy photosynthetic products in the form of sugars, amino acids, carbohydrates and small peptides from the algae, while the symbiotic algae receive inorganic plant nutrients, refuge and protection within the polyp tissues [5, 6]. The symbiosis serves the main purpose of
**<sup>\*</sup>Address correspondence to Joost W. van Dam:** Australian Institute of Marine Science, Townsville, Qld 4810, Australia; Email: [email protected]
restricting nutrient outflow into the surrounding oligotrophic water column, and as a result the host is endowed with substantially more energy than would otherwise be available to heterotrophs, enabling corals to extract calcium carbonate from surrounding waters and secrete it as a skeleton [6, 7]. As coral symbiosis based upon algal primary production is the engine driving coral reef ecosystems, stressors that interfere with photosynthetic processes could undermine the basis of this biologically and economically important marine habitat with serious consequences [8].
**Figure 1:** Distribution of coral reefs around the world (source: http://www.nasa.gov/).
Most scleractinian corals reproduce by broadcast spawning, with many species releasing eggs and sperm into the water column simultaneously at annual spawning events [9]. Typically fertilisation of the highly buoyant eggs is external and mobile planula larvae develop over 48–72 h [10]. The swimming planulae are usually competent to undergo settlement on the substratum followed by metamorphosis into a juvenile coral polyp after 96 h. Larvae of many coral species are known to require a biochemical inducer from their preferred settlement substratum, crustose coralline algae, to trigger metamorphosis [11]. Juvenile coral polyps are generally less than 2 mm long and are susceptible to predation as well as smothering by sediments and other stressors [12]. It takes between five and ten years for many coral species to become reproductive and these early life histories, including fertilisation, larval development, larval metamorphosis and the juvenile stage, are all critical for the long term resilience and health of coral reefs and their vulnerability to anthropogenic stress needs to be considered [13, 14].
Over the last century and a half in which intensive agriculture, fishing practises and industries evolved, an estimated 30% of coral reefs worldwide have been severely depleted. Between 50 and 70% of coral reefs are thought to be under direct and immediate threat from climate change and human activities [15, 16]. Coral reef cover in the Caribbean has been reported to have declined over 80% in the last 30 years [17]. Climate related changes in ocean acidity and temperature, nutrients and chemical pollution are the main proposed reasons for coral reef declines and cause for concern [12, 16, 18-21]. In this chapter we aim to critically review and collate information about the possible ecological effects of chemical pollution on coral reefs. We assess potential pollutants, sources and the likely exposure of reef-building corals and their zooxanthellate symbionts. Furthermore, we evaluate what trophic level is most likely to be affected by various pollutants and in what particular way an organism will be impacted. Finally, we aim to assess evidence that links exposure to effect and identifies gaps in our knowledge concerning the effects of anthropogenic pollutants on coral reefs.
# **SOURCES OF MARINE POLLUTANTS**
Waterborne chemicals affecting tropical marine communities can have both point and non-point sources and may be transported to reefs from distant origins. Different classes of contaminants are often associated with a particular environmental compartment because of their physicochemical characteristics. Fig. (**2)** presents a conceptual model for potential contaminant routes reaching reef ecosystems. Transport, dispersion, and ultimately biological effects of pollutants in marine systems depend on the persistence of these chemicals under tropical conditions and their bioaccumulation and biodegradation rates. Typically pollutants with a higher solubility in surrounding waters will find their way further offshore. Association of pollutants with particulate matter may increase environmental persistence. Because of the rapid sorption of many contaminants to sediments, it is not surprising the largest reservoirs of chemical stressors will be found in estuaries, wetlands or nearby urban centres. Nevertheless, suspended sediments transported in monsoonal flood-plumes have the potential of contaminating sites further offshore. Additionally, volatilisation from surface waters, transport through the atmosphere and redeposition elsewhere can deliver residues far from their original sites of application [22]. Biota carrying accumulated loads of persistent chemicals in their tissues can also transport pollutants between ecosystems and far from their application or deposition sites.
Terrestrial runoff from rivers and streams contaminated by agricultural, industrial or urban activities is usually the most important route for chemicals to enter marine waters (Table **1**). The array of potential contaminants is very wide from organics such as pesticide residues, pharmaceuticals and hydrocarbons to industrial waste products, metals and organometallic compounds. In highly protected areas such as Australia's Great Barrier Reef (GBR), chemicals of concern are contemporary pesticides; most specifically herbicides originating from farming activities [23]. In more highly populated regions such as south-east Asia, a much wider range of urban and industrial contaminants also threaten coral reefs.
**Figure 2**: Conceptual model for pollutant pathways in marine systems.
Industrial activities including mining and smelting operations are sources of metals and dioxin-like compounds while shipping operations, refineries and oil extraction and explorative processes introduce hydrocarbons and trace metals to the marine environment [24, 25]. In addition, urban sources such as sewage outfalls, desalination plants and landfills can contribute considerably to contaminant loads [1]. The potential off-site movement of chemical residues depends on both physicochemical properties of compounds (e.g. partition properties) and environmentally specific factors such as hydrology, sediment composition, temperature and biological degradation parameters [26]. While the breakdown of organic compounds is likely to be more rapid in the tropics than in temperate regions, the warmer conditions may increase sorption of organics to particulates thereby increasing their persistence in some environmental compartments. Direct contamination of marine waters can result from inputs associated with boating (hydrocarbons and antifouling paint applications) or chemical-based fishing (e.g. cyanide, [27]). Oceanic waste disposal, ship groundings and incidental spills are additional routes of chemical pollution. Coral reefs are often in close proximity to shipping lanes where contaminated bilge water is disposed of or cargo is spilled. Dredging of channels can resuspend metals and other buried organic pollutants [28].
**Table 1:** Main contaminants, sources and concerns in regards to tropical coral reefs
# **STRESS EVALUATION**
Coral reef ecosystems are composed of a diverse community ranging from algae to mammals interacting in a variety of ways on multiple hierarchical levels. All reef organisms are likely to vary substantially in their sensitivity and response to individual pollutants or combinations thereof. Furthermore, susceptibility of species and possible stressor interactions may differ considerably depending on the life stage of exposed organisms. Even though reefbuilding corals are only one component of the ecosystem, they are considered fundamental for its existence and hence in this review we will concentrate on available knowledge and key questions related to potential effects of pollutants on reef-building corals and their endosymbionts.
The delicate interaction between coral host and *Symbiodinium* sp*.* is particularly vulnerable to water quality declines and stress from high sea surface temperatures and high light intensities [3, 12]. From a toxicological point of view, chemicals can affect the host animal, the algal symbionts, or both. While contact of host with contaminants is often direct, multiple membranes need to be crossed before a chemical can reach the intracellular algae. Studies in the past have suggested physiological differences between reef building organisms may determine their susceptibility to chemically induced stress. For example, branching coral species seem more vulnerable to some chemical contaminants than massive corals [29] and small-polyped species seem more susceptible to pollution stress than large-polyped corals [30]. Orientation, growth form, life stage, reproduction strategy, mucus production and lipid content are all factors that will determine how easily survival of a coral species is influenced by chemical stress [1]. A wide range of laboratory assays have been used to evaluate effects of chemical pollutants on corals over the past four decades. Table **2** provides an overview of assays used to evaluate stress in multiple life history stages of hard corals.
Prior to 2000 most studies focussed on the effects of metals, PAHs and older organochlorine pesticides on corals (for reviews see [1, 53]). Due to increased application of modern pesticides and altered pollution profiles, up to date information is now required on the distribution and impact of contemporary pollutants on coral reefs.
**Table 2:** Examples of bioassays and biomarkers used to assess the effects of contamination on scleractinian corals.
The purposes of toxicity testing are to: (I) determine cause and effect relationships; (II) determine thresholds for lethal and sub-lethal effects to enable the derivation of water quality guidelines; and (III) to use this information in combination with exposure concentrations to evaluate and compare hazards and ideally determine the magnitude of risk potential and mitigation options. While the majority of toxicological data for most chemical contaminants originates from temperate studies and species, in the past decade there has been an increase in the number of studies examining the effects of relevant contaminants on tropical marine species, including corals. A number of assays have been developed to evaluate sublethal effects of chemical pollution on scleractinian corals. These are molecular, cellular and physiological diagnostic indicators that provide a means of assessing both qualitative and quantitative responses to a variety of pressures, individually and collectively. Organisms respond to environmental changes by regulating metabolic pathways to prevent physiological damage. These expressions precede population-level changes and are useful indicators if linked to specific physiological or ecological events [54].
Often a combination of pressures can result in mortality or impaired biological function. Thus, the evaluation of environmental effects on biological systems associated with anthropogenic pressures such as pollutants must begin with understanding causal linkages between stressor and effect [45]. For adult corals, biomarkers of effect include molecular (e.g. gene regulation), biochemical (e.g. lipid content), physiological (e.g. reproduction, calcification) and ecological (e.g. bleaching due to loss of symbionts, distribution) endpoints. The effects of contaminants on zooxanthellate photosynthesis have been assessed using respirometry, 14C fixation and pulse amplitude modulated (PAM) fluorometry [55]. A technique is based on fluorescence quantum yield measurements as a function of photochemical efficiency of photosystem II (PSII) in phototrophic organisms, which can be applied *in situ* or in the laboratory (for a technical review, see [56]). The degree of quantum yield inhibition may be directly correlated to the stressor concentration the evaluated organism is exposed to [57].
Previous studies have demonstrated that different stressors may trigger similar responses through alternating pathways. For example, exposure to metals [32, 58], cyanide [27], herbicides [14], elevated temperature [59, 60] and ocean acidification [61] can all induce bleaching. It is also possible that a single stressor induces a variety of responses that can only be detected using an assortment of different biomarkers [62]. Molecular analysis of gene expression can evaluate the relative impact of stressors, by identifying specific responses to individual stressors [49, 54, 63]. The field of transcriptomics identifies transcription of certain genes at a given time and allows monitoring
of gene expression which in turn can identify gene function and mechanisms behind a particular biological response [47, 64], although establishment of a causal link between genomic and physiological responses has often proved difficult. In effect, while physiological or ecological endpoints indicate stress effects at organismal, population or even ecosystem levels, genetic tools may reveal subtle and specific effects at a genomic level, preceding physiological responses such as coral bleaching or impaired photosynthesis.
# **CHEMICAL STRESSORS ON CORAL REEFS**
In this section we separately investigate organic biocides and industrial pollutants, metals and oil hydrocarbons. We will discuss pathways of exposure and summarise known effects on scleractinian corals. Finally, we will describe ecological impacts associated with these contaminants and evaluate risk posed to coral reef communities.
# **Pesticides and Antifouling Agents**
The second half of the 20th century has seen a massive increase in population growth and a corresponding expansion of world food production. The acceleration of agricultural output can partly be ascribed to the introduction of organochlorine (OC) pesticides such as DDT, which eventually allowed for the shift from classical small-scale farming to industrialised agriculture. The use of pesticides in combination with fertilisers has been growing steadily and has become an essential element of modern agriculture, by preventing the spread of pests and exotic species and by enhancing growth potential, resulting in dramatic increases in crop yields. By the late 1960s environmental drawbacks arising from the use of DDT for agricultural purposes, or as a vector of disease control, became apparent and usage of selected persistent OC pesticides (e.g. DDT, HCB, dieldrin and chlordane) was slowly phased out in most developed countries. Other classes of pesticides such as organophosphates, carbamates, organotins and pyrethroids soon filled these roles. Today a wide variety of pesticides are used in specialised farming globally, with enormous annual application rates (see Chapter 4).
**Figure 3**: Sources of pesticides reaching coral reef sites.
Modern pesticides generally exhibit shorter half-lives in the environment when compared with their predecessors. Improved usage patterns and progressive application techniques, alongside highly specialised modes of action, make unintended side-effects to non-target species considerably less of a concern. However, care must be observed. The economies of countries in tropical areas are often based on agriculture, and depend on intensive use of pesticides to maintain and improve production [65]. As agricultural activities are primarily concentrated in river valleys and coastal plains, it is not surprising that agrochemical residues originating from terrestrial applications are ubiquitous in streams and rivers, eventually draining into estuaries and coastal seas [23, 66]. Multiple studies have shown that the main source of pesticides in nearshore coastal areas is agricultural application, transported via terrestrial runoff (especially during the monsoon season) [23, 67, 68]. The effects of pesticide residues are of great concern as many of these compounds and their breakdown products have been reported to influence both human and environmental health. It is therefore vital to characterise transport and fate of pesticides and their toxicity to non-target organisms to confidently assess risk associated with application, especially in tropical areas where pesticide usage patterns are generally much higher than in temperate zones [65]. Fig. (**3)** provides an overview of sources of biocides reaching reef sites.
In addition to terrestrial sources, biocides are intentionally released into the marine environment by incorporation into antifouling paint formulations where they function to prevent unwanted growth of a wide range of algae and invertebrates on boats or other marine structures. The organometallic biocide tributyltin (TBT) proved the most effective antifouling ingredient in the latter half of the 20th century and is a common contaminant in harbours globally. Although normally uncommon on coral reefs, TBT as well as metal biocide 'boosters' have been identified in extreme concentrations on reef rubble following ship groundings [69, 70]. The use of TBT has been regulated internationally since 1990 due to recognition of its severe impact on aquatic ecosystems. In 2003 the International Maritime Organization (IMO) and their Marine Environment Protection Committee (MEPC) banned application of TBT as an antifouling agent on ships completely. As of 2008, TBT-based antifouling paints must be removed or covered with a sealer-coat [71]. Since the phasing out of TBT in antifouling paints alternative coatings have been developed often using copper or zinc in combination with an organic booster biocide. These are typically herbicidal in nature, as primary colonisation of hull surfaces by microalgae allows for a convenient base for subsequent attachment and growth of seaweeds and invertebrates. This may result in elevated environmental concentrations in areas of high yachting activity, particularly in and around harbours [72]. Studies have shown that biocide dissipation from hulls of vessels is a function of leaching and degradation rates, water movement, sorption and hull treatment amongst other parameters [73]. The same studies have suggested two of the most popular booster biocides in use, diuron and Irgarol 1051, to persist in the water column while other booster biocides disappear more rapidly [73]. Although some data are available on biodynamics of antifouling agents in temperate regions, there is little comparable information on bioavailability in tropical waters.
Until 2000, most studies on pesticides in tropical marine ecosystems focussed on OC compounds, as these persistent chemicals have dominated the pesticide market for decades in tropical areas. Studies also focused on estuaries, as this is likely to be where most terrestrially derived sediments settle and therefore where most concentrated reservoirs of persistent chemicals will be detected. Furthermore, studies focused on commercial species such as fish that may affect human health. Apart from a few studies performed in the Florida Keys [74-76], Bermuda [34] and the Philippines [77], most available data on pesticide dynamics on coral reefs are associated with the Great Barrier Reef. The limited studies assessing banned persistent organochlorine substances on coral reefs found concentrations to be generally low [74-76, 78] or declining [53, 79]. While organochlorine residues (e.g. endosulfan) have been measured in recent times [78-80], it is the current generation of pesticides that are of greatest concern to inshore marine ecosystems. Organophosphorous (OP) insecticides such as chlorpyrifos, or herbicides such as glyphosate (Round-up®); triazine herbicides such as atrazine, simazine, ametryn and Irgarol 1051; and urea herbicides such as diuron and tebuthiuron are some of the key pesticides that may affect coral reef biota [23, 67, 81].
Recent studies have found contemporary herbicides to be ubiquitous in waters nearby coral reefs [23, 34, 68, 81, 82]. Systemic herbicides are of particular ecological concern to coral reef systems, as these compounds are developed for quick environmental uptake through the root system of plants and are therefore relatively water soluble. A chemical's solubility will determine whether it mainly exists in the water column or is associated with suspended particulate matter, and therefore more likely to sink to the bottom. In contrast to the latter fraction that precipitates nearshore and becomes incorporated in the sediment, chemicals dissolved in the water column can travel greater distances and exert adverse effects far from their application sites. Thus, apart from a greater potential to reach reef sites, herbicide pollution can have severe consequences for ecosystems dependent on primary production. Herbicides target a wide range of physiological processes; however the herbicides most commonly detected on coral reefs are the photosystem II (PSII) herbicides (Table **3**). This class of herbicide acts by inhibiting electron transport through the photosystem in chloroplasts by reversibly binding to a specific electron-acceptor protein (D1-enzyme in PSII). These herbicides outcompete the normal ligand for binding sites on this highly conserved protein vital for plant photosynthesis [83]. The D1-enzyme also forms part of the photosystem in symbiotic zooxanthellae of corals and is likewise affected by herbicide exposure. As far as the holobiont (host combined with symbiont) is concerned, direct effects of herbicide exposure are a decrease in algal photosynthetic efficiency, limiting energy flow from symbiont to host [84]. Secondary effects of restricted electron flow in PSII include a build-up in reactive oxygen leading to oxidative stress [85], a process intensified by high illumination [86]. Further effects include disruption of
membrane structure and chlorosis, as well as aberrations in energy dynamics due to the reduced availability of photosynthetic products, ATP, NADPH and ferredoxin [87]. As inhibition of photosynthesis can lead to decreased algal production and eventually expulsion of symbionts (bleaching), herbicide contamination may disturb the fragile keystone algal-coral relationship so important for all ecological processes on coral reefs.
# *Impact*
Table **3** provides an overview of studies on pesticide effects on hard corals. Little is known regarding the effects of OC pesticides on reef building corals. Suspected adverse effects caused by OCs range from carcinogenesis, interruption of neurological function, changes in cell metabolism and gene expression, to endocrine disruption and interference with reproduction [78].
In an early study on organochlorine pollution, McCloskey & Chesher [89] observed photosynthetic depression in a number of scleractinian corals after *in situ* exposure with DDT, dieldrin and a PCB with concentrations in the mg/L range. However, even these extremely high concentrations did not result in alterations of feeding behaviour, polyp expansion, sediment clearing or skeletal crystal formation. Olafson [90] explored bioaccumulative potential of OCs in reef biota and found DDT, chlordane and lindane able to accumulate in coral tissues. In a recent study assessing effects of short-term exposure (up to 96 h) to low insecticide concentrations on different life history stages of the branching coral *Acropora millepora*, endosulfan (OC), chlorpyrifos and profenofos (OPs) were found to affect photosynthetic performance and/or density of zooxanthellae within adult branches at relatively high concentrations [13]. In addition, profenofos-exposed branches expressed permanent tissue retraction. These neurotoxic insecticides did not inhibit fertilisation of gametes as may be expected from the absence of neurons in oocytes and sperm, yet larval metamorphosis was heavily impacted, with 50% effect concentrations (EC50) for inhibition of metamorphosis as low as 0.3-1.0 µg/L [13]. This study found the swimming behaviour of the larvae was not affected by the insecticides, an observation confirmed by Acevedo [91], who demonstrated that far greater concentrations of chlorpyrifos and carbaryl were required (mg/L) in order to cause mortality amongst larvae of the brooding coral *Pocillopora damicornis*. These findings suggest that the mode of impact of investigated insecticides in coral larvae involves specialised pathways instead of general neurotoxicity. In contrast, another study on adult colonies of *P. damicornis* demonstrated 50% mortality (LC50) to occur after 96-h exposure to 6 µg/L chlorpyrifos [92]. While most insecticides tested appear to negatively affect corals or their larvae, a recent study indicates that the "eco-friendly" larvicidal agent, *Bacillus thuringiensis* ssp *israelensis* (Bti), used extensively to control mosquitoes in the tropics, is harmless to coral larvae [93].
Herbicides readily penetrate coral tissues and rapidly (within minutes) reduce the photosynthetic efficiency of the endosymbiotic zooxanthellae. No apparent acute effects have been observed on host animals, fertilisation of gametes or metamorphosis of larvae after short-term exposures to diuron [14]. Nonetheless, bleaching of established recruits or adult coral branches is a common reaction to high concentrations or chronic exposures of PSII herbicides [14, 32, 94]. The dissociation of symbiosis is considered a sub-lethal stress response and a secondary effect most probably caused by oxidative stress in zooxanthellae as a result of chronic photoinhibition. Although the mechanism is still not entirely understood, the main hypothesis is that by expelling symbionts, host corals can reduce the number of damaged symbionts within their tissue, while at the same time limit their exposure to reactive singlet oxygen [86, 94].
In a study on the effects of the antifouling herbicide Irgarol 1051, Owen and co-workers [34] found isolated *in vitro* symbionts of *Madracis mirabilis* to cease incorporation of H14CO3 - after 4-8 h exposures to concentrations as low as 63 ng/L. The effects of PSII herbicides on coral symbionts has flow-on effects to the host, which suffers a proportional decrease in the energy translocated as sugars to the animal tissue [84] and this can lead in the long term to reduced reproductive output [31]. Jones [86] reviewed the toxicological effects of various PSII herbicides on a range of coral life history stages and isolated symbiotic algae and argued the most sensitive endpoint to be inhibition of photosynthesis in algal symbionts. Overall, in directly comparable exposure experiments using adult branches of multiple coral species, concentrations of tested herbicides reducing effective quantum photosynthetic yield (10-h EC50) of symbiotic algae *in hospite* ranged over three orders of magnitude (Fig. **4**), while significant reductions in photosynthetic yields were observed at concentrations as low as 50 ng/L (Irgarol 1051), 200 ng/L (diuron) or 300 ng/L (ametryn) [38, 39, 86]. These outcomes were supported by tests on isolated symbionts [88] and can be directly compared with toxicological results on other aquatic primary producers, including tropical estuarine microalgae [95], seagrass [96], crustose coralline algae [97], marine diatoms [98] and freshwater algae [99]. Conversely, recovery of normal photosynthetic rates occurred quickly after placement in clean exposure medium [38, 39]. This would suggest short-term exposure to PSII herbicides inflicts no permanent damage.
**Table 3:** Toxicological studies on the effects of selected pesticides on different life history stages of scleractinian corals
1. For methodology and exposure durations please refer to text or original publications
2. For quantitative toxicological data please refer to text or original publications
However, chronic exposure experiments of coral branches to low (1-10 µg/L) diuron concentrations showed, apart from decreased photosynthetic rates, severe visible bleaching, partial colony mortality, two to five fold reductions in tissue lipid content and significant reductions in fecundity [31]. In laboratory exposures, the widely used non-PSII herbicide 2,4-D (a growth inhibitor) failed to exert any effect on colour, tissue extension, photosynthesis and metabolism of adult corals except in extremely high concentrations (100 mg/L) [74, 77]. When exposed to a formula containing a 2,4-D amine salt including a 'wetting agent' (dispersant) though, severe toxic effects were observed at 100 µg/L [74].
**Figure 4**: EC50 for several herbicides impacting effective quantum PSII yield of zooxanthellae symbionts in tissues of different coral species. Lower values indicate greater toxicity with susceptibility ranging over three orders of magnitude. Data after [38, 39, 86].
The historic focus on persistent pesticides has found that the key concern is associated with accumulation in the food chain and potential effects in air-breathing top predators. A key finding in the last decade relates to PSII herbicides. There is increasing evidence of widespread relatively low level exposure of inshore reefs (e.g. on Australia's GBR) to these chemicals. Concentrations are typically below levels at which adverse effects may occur on photosynthesis (e.g. diuron <20 ng/L). However, during major flow events concentrations have been determined in flood plumes near inshore reefs that are sufficiently high for effects to be detectable (diuron 0.1-1 µg/L) [23]. Furthermore, as flood plumes often carry elevated concentrations of several pollutants simultaneously, it is likely that nearshore reef systems are exposed to combinations of chemical stressors. Herbicides are commonly detected in complex mixtures within fresh and seawater systems. As multiple factors potentially interfere with the same physiological mechanism (e.g. PSII electron flow), additive or even synergistic toxic effects may occur [100]. A recent study has confirmed that the phytotoxicity of PSII herbicides commonly detected in GBR waters towards benthic microalgae is additive [101]. Overall, the margin of safety between observed concentrations and measurable effects is relatively small and the potential risks from chronic exposure remain unclear. Herbicide-induced interference with primary producers may exert a bottom-up pressure on the system, potentially decreasing reef resilience to other environmental stressors as elevated temperatures and ocean acidification.
# **Industrial Organochlorines: PCBs, Dioxins and Furans**
Polychlorinated dibenzodioxins (PCDDs), dibenzofurans (PCDFs) and biphenyls (PCBs) are ubiquitous organic contaminants. This group of chemicals is extremely persistent, has a tendency to bioaccumulate in biotic tissues and includes some of the most toxicologically potent compounds known (see Chapter 7). Although a global treaty aimed at the reduction and possible elimination of PCBs and dioxin-like chemicals has been established in 2001, significant concentrations are still found in marine environments worldwide. Evidence exists that distribution of PCBs, PCDD/Fs into the aquatic environment is mainly due to atmospheric deposition of volatilised molecules [102]. Despite many studies in temperate and arctic regions and extensive reporting on the concentrations and some effects of PCBs and PCDD/Fs in tropical marine species including mammals [103-105], fish [106, 107] and invertebrates [79, 107], to our knowledge no studies have been performed assessing concentrations in tropical waters or on coral reefs and effects on reef-building corals remain speculative. Dioxin-like substances bind to the aryl hydrocarbon receptor in vertebrates and invertebrates, resulting in interference with a broad range of cellular processes. Concentrations of these compounds in water are presumably too low to cause direct effects on primary producers and environmental impact is mostly associated with bioaccumulation, affecting air-breathing organisms at the top of the food chain. Therefore, likely ecological impacts of dioxin-like pollutants on coral reefs will consist of top-down imbalance of the system.
# **Oil Hydrocarbons and PAHs**
Crude oil, refined petroleum and their combustion products all contain aromatic hydrocarbons, including some polycyclic aromatic hydrocarbons (PAHs). These compounds enter the marine environment through anthropogenic processes or have naturally occurring sources (Table **1**) [53]. A study by Capone and Bauer [108] suggested that in the late 1980s an estimated average of 6 million metric tons of petroleum products were released into our oceans annually. As most petroleum products are hydrophobic in nature, the majority of aromatic hydrocarbons introduced into the marine environment will associate themselves with particulate matter and be deposited in the sediment [108], where these compounds tend to persist. Benthic filter feeders or sessile organisms are at risk through direct contact or ingestion of oil compounds. Straughan [109] argued the biological consequences of oil spills should be determined by the nature and interaction of a multitude of factors, including type of oil, dosage, remedial action, prior exposure, presence of other stressors, differences between biota and many physical environmental, climatic and seasonal factors. Generally, a mixed product containing a broad spectrum of hydrocarbons is released to the marine environmental where it may affect a variety of biological processes [108].
**Table 4:** Studies on the effects of selected oil products on different life history stages of scleractinian corals.
1. For methodology and exposure durations please refer to text or original publications
2. For quantitative toxicological data please refer to text or original publications
3. PFW and drilling mud often contain toxic concentrations of both metals and hydrocarbons
The potential of oil contamination to coral reefs is high given their often intense commercial activity and proximity to shipping lanes [1]. Acute exposure of reef ecosystems to oil spills can occur through accidental discharge from ships or as a result of terrestrial runoff. Production platforms in the vicinity of coral reefs have the potential to contaminate surrounding waters through discharge of production formation water (PFW) [42], a complex mixture that may contain petroleum hydrocarbons, suspended solids, metals, naturally occurring radioactive materials, organic acids and inorganic ions amongst other substances [110]. Likewise, the drilling of oil-wells may introduce mud heavily contaminated with petroleum hydrocarbons and metals to the marine system, while refineries and pipelines are additional sources with potential for chronic pollution by oil products. Floating oil can be deposited on reef flats or to interfere with reproductive processes in buoyant gametes or larvae [111]. Large oil spills are often moderated by application of surface dispersants that dissolve slicks into smaller droplets. However, application of
dispersants is likely to increase hydrocarbon concentrations in the water column and thus increase exposure to benthic reef organisms [42, 112]. Several studies have demonstrated that a combination of dispersal and oil products expressed higher toxicity than either component separately [42, 113, 114], yet it has been suggested modern dispersants may be less toxic to marine biota [1]. Once in the water column petroleum hydrocarbons rapidly become associated with organic matter and suspended particles. Volatile components evaporate while non-volatile components are deposited into the sediment. This deposited fraction is unlikely to absorb, evaporate, dissolve or be biologically degraded [53, 108]. Oil products may also occur in globulised form dispersed through the water column and can settle onto the reef [115]. Aromatic hydrocarbons, either in the form of dispersed oil or as water soluble components, can be absorbed by coral tissues while oil globules can adhere to coral surfaces [116]. The high lipophilicity of aromatic hydrocarbons stimulates rapid passive uptake in the coral tissue, while detoxification can be slow [114, 117]. Residues of aromatic hydrocarbons have been reported to remain present in coral tissues months after exposure occurred [118], yet evidence suggests hydrocarbon deposits in sediments and coral tissues to be substantially reduced after two years at high-energy reef sites [76, 119].
# *Impact*
Adult coral colonies can be killed or injured by direct contact with oil or drilling mud [116, 124, 125]. Filter feeders and benthic organisms that cannot escape the oil or contaminated sediments will typically see bioaccumulation of toxic compounds, genetic mutations and metabolic disorder in their tissues. Corals exposed to hydrocarbons have been shown to exhibit loss of zooxanthellae (bleaching), impaired reproduction and tissue damage [116]. Field studies on mud discharges during oil well drilling found decreased coral growth rates in *Montastraea annularis* exposed to the fluids, while several years of exposure found 70 to 90% reduction in coral cover within one hundred meters from the drilling site [126]. These findings were supported by laboratory studies predicting decreased growth, metabolic aberrations and nutritional abnormalities [127, 128]. More recently, Raimondi and colleagues [124] observed tissue mortality in adult cup corals after exposure to drilling muds. Bak [129] observed decreased coral cover, diversity and local recruitment after chronic exposure to refinery petroleum on a Caribbean reef and argued chronic pollution to have more severe effects than single spills, yet comparable detrimental effects were observed after an acute major oil spill in Panama [130]. Guzmán and co-workers [130] also observed a strong correlation between injured corals and oil residues in sediment during five years of monitoring coral health after this spill. Aromatic hydrocarbons have been shown to decrease photosynthetic performance of dinoflagellate symbionts in some corals [113, 131]. Jones and Heyward [123] observed decreased photochemical efficiency and subsequent expulsion of zooxanthellae by host corals exposed to PFW as a result of photoinhibition in the symbionts. Freshly isolated symbionts were affected at slightly lower concentrations [123]. Tissue retraction as an environmental stress response has been observed after exposure to low concentrations of oil or dispersed oil, but normal tentacular expansion recovered within a week [114]. It has also be hypothesized that aromatic hydrocarbons may cause a reduction in coral tissue lipid contents, thereby limiting fat reserves necessary for increased mucus production or proliferation of mucus secretory cells [116, 118, 121].
Decreased reproductive success of both brooding and broadcasting corals after oil exposure has been observed in a number of studies. Two histological studies showed impaired gonadal development [121, 132], while in field studies infertility and decreased egg size as a result of increased injury occurrence were observed five years after exposure to a major oil spill [132]. Loya and Rinkevich [111] noted oil products could induce premature expulsion of larvae in a brooding coral. More recently, Negri and Heyward [42] suggested that early life stages may be more sensitive to pollution than established adult colonies. In annual mass spawning events, as experienced in the Indo-West Pacific or Australian waters, gametes of broadcast spawning coral species are released in the water during brief synchronised events. The highly buoyant eggs float to the surface where fertilisation occurs [133]. If the spawning event coincides with an oil spill, an entire year of reproductive effort is threatened while adverse environmental conditions endure with implications for larval settlement and parent colonies. A number of laboratory studies have reported a dose-dependent inhibition of fertilisation, metamorphosis and settlement of broadcast spawning scleractinian corals after exposures to oil and/or dispersant [42, 120, 134]. For example, crude oil inhibited larval metamorphosis at 82 µg/L and this was reduced to 33 µg/L in the presence of a non-ionic dispersant [42]. Despite these findings, effects of oil (whether or not dispersed) on gamete fertilization, embryogenesis, larval metamorphosis and settlement are unknown for the majority of both broadcast spawning and brooding species.
Marine pollution from PAHs and oil hydrocarbons is associated with localized events as runoff from urban centres, oil exploration and extraction activities and accidental spills. These chemicals are highly hydrophobic and therefore contamination typically remains relatively confined, although spillage from offshore drilling operations can travel for great distances. Likely effects on coral reefs will consist of overall disturbance of biological homeostasis by exerting effects over multiple trophic levels. The greatest impacts on coral reefs are likely to occur if hydrocarbons come into direct contact with coral spawn during mass reproduction or at low tide on shallow reefs. Interactions with other environmental stressors are improbable, as effects from spills and shipping incidents often cause severe mortality and will overwhelm subtle adverse effects from other factors.
# **Trace Metals and Metalloids**
Metals are a physical component of rocks and soils and enter the environment through natural weathering and erosion processes. Many metals are biologically essential, yet most have the potential to become toxic above certain threshold concentrations [135]. Industrial activities such as mining and smelting as well as agricultural applications (*i.e.* organometallic pesticides and fertilisers) and urban waste have substantially contributed to the release of elevated quantities of trace metals into the environment [53]. Even though recognition of toxic potential and legislation in the past have seen a great reduction in metal output, environmental contamination continues. Terrestrial runoff and sediment-bound transport through freshwater streams and rivers eventually delivers these contaminants to estuaries and inshore seas (Fig. **1**). Metals are strongly associated with particulate matter and therefore not usually directly available to aquatic biota. However, particulate metals in sediments can be solubilised by acidic juices in the gut of sediment-feeding organisms, and thus become available for accumulation in biotic matrices through passive uptake across permeable surfaces such as gills or the digestive tract [136, 137]. Biological availability and solubilisation rates of trace metals from particulate matter are dependent upon a variety of environmental variables, including sediment cation exchange capacity and organic content, dissolved oxygen concentrations, pH, salinity, temperature and redox potential amongst other factors [53, 137, 138]. Furthermore, remobilisation and resuspension of sediments may return metals to the water column [28]. High environmental metal concentrations are generally restricted to locations adjacent urban centres, industrialised areas or sites draining areas of intensive agriculture [139]. Trace metals and metalloids have a multitude of applications and sources, yet the most abundant metals entering the environment in elevated quantities as a result of agricultural activities are copper (Cu) and zinc (Zn), used as constituents of fertilisers or biocides; arsenic (As), cadmium (Cd) and mercury (Hg) as components of some fungicides. Lead (Pb), nickel (Ni), aluminium (Al), manganese (Mn) and iron (Fe) often enter marine waters as the results of mining activities (as do As and Hg), industrial or urban waste discharges and runoff. Tin (Sn) has generally been introduced into the environment as a biocide, principally as constituent of antifouling paint formulations, e.g. tributyltin (TBT) [139, 140].
Numerous studies exist on metal contamination and effect on corals [141-144]. In adult corals, metals might be absorbed and occur in various capacities. As early as 1971, Livingstone and Thompson [145] found trace metals to be incorporated into the aragonite (a carbonate mineral) of coral skeletons. Quantification of these built-in skeletal metals is currently used as a biomarker that reflects environmental conditions during the coral's lifetime [146-148]. Trace metals can also be found in skeletal cavities [149], integrated within the organic matrix of coral skeletons [150], or absorbed onto exposed surfaces of the skeleton [151]. Besides skeletal inclusion, several studies have demonstrated trace metals present in coral tissue [33, 152, 153]. Pathways through which corals absorb metals may vary. Brown and colleagues [151] found that corals retract their tissue in response to environmental stress, and thus may be more susceptible to direct uptake of metals by exposed skeletal spines. Another responsive action to physical or chemical stress is the excretion of high quantities of mucus with a high affinity to bind metals that may actively reduce metal uptake [127]. Additionally, it has been suggested corals are able to regulate internal metal concentrations through the physiology of their zooxanthellae endosymbionts. Studies with related symbiotic organisms like sea anemones indicated zooxanthellae to be responsible for the majority of metal uptake and accumulation [154]. In response to elevated metal concentrations, zooxanthellae can enhance calcification rates. Furthermore, zooxanthellae may be involved in the active uptake of trace metals, accumulating higher concentrations of metals than do host coral tissues [58, 143, 145, 155]. Subsequent stress-induced expulsion of symbionts by the coral host may act as a regulatory response mechanism in reaction to high metal concentrations [32].
# *Impact*
Bioavailability, physiological effects and fate of trace metals are highly dependent on the chemical form and oxidation state in which metals exist, as reflected by their toxicity [164]. Thus, it is of clear importance to distinguish between individual metal species present in a particular biological compartment [165]. Once introduced in a biotic matrix, trace metals have the potential to affect nutrient cycling, cell growth and regeneration, as well as reproductive cycles and photosynthetic potential [1, 53]. Table 5 summarises a range of effects on corals caused by exposures to metals and organometallic compounds. Elevated levels of copper, zinc and tin in the effluent of a tin smelter in Thailand caused reduced growth and calcification rates in branching corals [33]. In a study considering corals in a Hong Kong estuary exposed to elevated concentrations of metals, pesticides, nutrients, sewage effluents and suspended sediments over a prolonged period of time, it was argued metals were mainly responsible for declines in coral cover, diversity, abundance and growth rates [30]. Laboratory exposure of the massive coral Porites lutea to elevated iron concentrations resulted in bleaching. It was noted that corals that had been pre-exposed to an iron-enriched environment responded in a less drastic way, suggesting development of some form of iron tolerance [58]. Jones [32, 94] found elevated copper concentrations to induce rapid bleaching in the branching corals Acropora formosa and Seriatopora hystrix, while no inhibition of photosynthetic efficiency of zooxanthellate endosymbionts was observed. The author suggested copperinduced bleaching to occur without affecting the algal photosynthesis but may be related to effects on the host coral. However, in a longer term exposure experiment on Plesiastrea versipora, low concentrations of copper were observed to reduce symbiont response to a host signalling factor regulating photosynthesis, while stress responses as inhibition of photosynthetic efficiency or bleaching were not detected [166]. To further emphasise the toxic effects of copper, two separate studies showed how copper has a detrimental effect on the metabolism of both the branching coral Pocillopora damicornis and the massive coral Porites lutea [35, 167].
**Table 5:** Toxicological studies on the effects of selected trace metals and metalloids on different life history stages of scleractinian corals
1. For methodology and exposure durations please refer to text or original publications
2. For quantitative toxicological data please refer to text or original publications
In a recent study on the effects of an organometallic fungicide containing mercury on different life history stages in *Acropora millepora*, 2-methoxyethylmercuric chloride (MEMC) severely affected adult branches exposed to 10 µg/L MEMC. Branches bleached and some host tissue died at this concentration, while at a lower concentration (1 µg/L) polyps retracted and photosynthetic efficiency decreased [13]. Early life history stages proved very sensitive to MEMC exposure. Lowest observed effect concentrations (LOECs) inhibiting fertilisation of gametes and larval metamorphosis were established at 1 µg/L MEMC (EC50 values of 1.7 µg/L and 2.5 µg/L, respectively) [13]. In a series of experiments assessing effect of trace metals on fertilisation and settlement success of selected coral species, copper was found to be a highly effective inhibitor of fertilisation in all species tested, with fertilisation rates dropping proportionally with increasing copper exposure concentrations (EC50 = 15-40 µg/L Cu). Zinc, lead and cadmium were much less potent. However, high interspecific variety in sensitivity was observed [40, 41]. In contrast to the high sensitivity of hard-coral gametes to copper exposure, gametes of the soft coral *Lobophytum compactum* exhibited a surprising resistance to copper toxicity (EC50 = 261 µg/L) [168]. Settlement of *Acropora tenuis* larvae was significantly reduced at concentrations of 42 µg/L Cu (48-h EC50 = 35 µg/L) [44]. Negri and Heyward [160] found fertilisation of gametes and larval metamorphosis of *Acropora millepora* reduced when exposed to low concentrations of copper and TBT. Copper proved the most potent inhibitor of fertilisation in this study (4-h EC50 = 17.4 µg/L), while TBT proved more toxic towards larval metamorphosis (24-h EC50 = 2 µg/L). The same study also showed that surfaces coated with antifouling paints containing copper or TBT to inhibit both fertilisation and metamorphosis [160]. These findings were confirmed by Victor and Richmond [144] who exposed *Acropora* gametes in Guam to low copper concentrations and calculated 50% inhibition of gamete fertilisation after 12-h exposure to 11.4 µg/L Cu. TBT is known to inhibit protein synthesis [169] and some cnidarians such as the sea anemone *Aiptasia palliida* have been shown to exhibit reduced resistance to infection and decreased zooxanthellae densities after chronic exposure to very low concentrations (0.05 µg/L) of TBT [170]. In another study involving antifouling paint contamination, sediment polluted with a mixture of TBT, copper and zinc, all components of commercial antifouling formulations used until recently and acquired at a ship grounding site in Australia, demonstrated potential to interfere with normal larval behaviour [43]. Modern antifouling formulations that contain organometallic components used as replacements for TBT are often not exempt from environmental impact; e.g. the biocide zinc-pyrithione (Zpt) is detrimental to embryonic development of both sea urchins and mussels [171]; however, no toxicological studies of this compound have been performed on corals. In a review by Reichelt-Brushett and McOrist [143] on trace metal contamination within corals from around the world it was made evident that, although high interspecific, regional and temporal variance was observed, environmental concentrations of selected metals (especially copper) were often near or exceeding concentrations proven to exert detrimental ecological effects on scleractinian corals.
Trace metal pollution is often limited to areas adjacent to urban and industrial centres or near river deltas. As metals are relatively immobile in the marine environment, adverse effects are likely to be exerted on a localised scale. Copper and organometallic substances containing tin or mercury are significantly more potent than other trace metals and affect a broad range of variables in a variety of species. Consequences for the system will consist of both bottom-up and topdown effects. Chronic, low-level metal contamination may decrease resilience of marine organisms to other environmental stressors such as elevated temperatures, ocean acidification and other chemical pollutants.
# **Interactions of Multiple Stressors**
Nearshore marine pollution often occurs in combination with other natural and anthropogenic sources of stress for resident biota. A globally changing climate results in increasing sea surface temperatures and ocean acidification, arguably the most important factors when considering stress on coral reefs [3, 19, 35]. Moreover, in the tropics monsoonal rainfall delivers vast amounts of fresh water, nutrients and suspended sediments to estuaries and inshore reef systems [172]. At this time of the year seawater temperatures approach thermal tolerance limits for many coral species [6, 60]. Thus, during flooding events, inshore coral reefs can potentially face combinations of low salinity, high turbidity, nutrient and pesticide exposures during episodes of thermal stress. It is assumed that nearshore corals and other sensitive organisms may be at extreme risk to combinations of stressors but little research has been performed to test this hypothesis.
Of the studies performed on stressor interactions, most have dealt with temperature in combination with an additional stress factor. Temperature affects physicochemical properties of membranes, including permeability, fluidity and diffusion rates. This will have a likely effect on a chemical's toxicity that is dependent on target site delivery [38]. Simultaneously, temperature can affect toxin solubility, speciation and (bio)degradation rates but also an organism's sensitivity to a particular chemical [164]. The PSII herbicide diuron was observed to take longer at 20 °C than at 30 °C to reach a similar response in *Seriatopora hystrix*, while a decreasing sensitivity to the herbicide was observed at higher temperatures [38]. Another good illustrative example has recently been provided in a study where high dissolved inorganic nitrogen (DIN) concentrations were correlated with a decreased resilience of corals to high temperature [173]. Wooldridge and Done [174] subsequently proposed how combinations of high temperature and high DIN concentrations work synergistically and may be a causative mechanism for large-scale coral bleaching. However, the multiple sources of stress (often simultaneous events), coupled with the great complexity of marine ecosystems and their high variability obscures the establishment of simple causal relationships between stressors and observed effects, which greatly complicates assessment of tolerance, resilience and ecological implications of stress [175].
# **SYNTHESIS**
Tropical coral reefs are among the most biologically diverse and productive ecosystems in the world but their continued existence is threatened by a number of factors. Despite their extended geological subsistence, coral reefs appear very sensitive to changes in environmental conditions. Elevated ocean acidity and surface temperature, high turbidity and nutrient concentrations through terrestrial runoff and chemical pollution from various sources are the main pressures exerted on modern reef systems and among the proposed reasons for global coral reef declines [16, 19]. Community alterations or a shift from autotrophy to heterotrophy will eventually affect the entire reef community, and could possibly change the dominant ecological process from calcium carbonate deposition to erosion [176]. Pressures on coral reef ecosystems are likely to increase further as a result of expanding coastal agricultural practises and industrialisation, population growth and climate change. While limiting the effects of climate change is a global challenge, management approaches to minimise the effects of pollution pressures on nearshore coral reefs can contribute towards sustainable exploitation of our marine resources; curbed inflow of suspended sediments, nutrients and chemical stressors are potential means to protect our reef systems in a shifting environment [12, 177, 178]. We have shown substantial differences exist in sensitivity and response of scleractinian corals and their zooxanthellae symbionts to various types of chemical pollutants. Furthermore, susceptibility may differ considerably depending on its life history stage. Evaluation of environmental stress on coral reefs caused by chemical pollution is constrained by the limited number of assessment endpoints, implying further research is required on a broad range of species and life history stages.
Chemical pollutants can enter and affect a reef ecosystem in a number of ways. The type of exposure often determines the severity and scale of effects. Coral reef ecosystems may be chronically exposed to combinations of stressors at low concentrations, regular pulses of chemical stressors and/or be subject to acute exposures of specific chemicals at relatively high concentrations during an accidental pollution event (Table **6**). While accidental pollution events (e.g. oil spills) will most likely overshadow effects from other potential stressors and have profound effects at all trophic levels, they are often localised, not harming the overall structure and function of large ecosystems. In contrast, regularly recurring pollution events such as input from river floods or chronic pollution from land runoff (e.g. sewage treatment effluent or herbicides) is more likely to affect a larger area and exert subtle effects, driving the system towards genetic and ecological adaptability. In these scenarios, the combined effects of chemical mixtures may significantly increase total toxicity and associated risk.
Low-level chronic pollution and recurring pollution events may interfere with the resilience of lower trophic levels instead of directly impacting the ecosystem. Because adverse conditions persist for prolonged periods of time, fitness of species is affected in the long run and this may lead to an increased vulnerability to various other stressors, including those related to climate change.
Ecological risk assessment requires knowledge of the spatial and temporal distributions of key stressors in relation to the most vulnerable tropical species. Exposure is a function of the intensity (magnitude), timing, frequency and duration of adverse conditions. Thus, ecological exposure scenarios may become very complex, especially when multiple stressors are involved and exert pressures on interconnected biological compartments. With all relevant information taken into consideration, it is not likely that the chemicals discussed in this chapter will significantly impact reefs single-handedly as concentrations are generally low or localised. For example, a number of chemical pollutants have been identified at low concentrations on the GBR, including the herbicide diuron and the insecticide chlorpyrifos (both primarily used in coastal agriculture), TBT and copper (antifoulants) and PAHs from oil spills and boating/shipping activities. The toxic effect concentrations of these pollutants and their observed annual mean
(background) and peak event (e.g. oil spill, ship grounding, flood plume) concentrations can be combined as risk quotients (effect concentration divided by environmental concentration), providing semi-quantitative estimates of risk and safety margins for GBR species (Table **7**).
**Table 6:** Profiles and likely effect for three types of pollution scenarios on coral reefs.
LOEC=Lowest observed effect concentration.
1 Higher numbers indicate a greater safety margin.
2 Exposure scenario is indicative of which scenario is more likely for a given type of chemical to be relevant in terms of risk: (I) chronic exposure, (II) recurring event-style exposure, (III) random event-style exposure (Table **6**)
3 Concentrations of TBT are usually specified in µg/g sediment; to our knowledge no information is available for water concentrations found in Australia.
4 Concentrations found offshore in northern Australia[179]. Few chronic pollutants on the GBR approach concentrations that may cause harmful effects to corals as indicated by high risk quotients (>10 for background concentrations) for diuron, chlorpyrifos and copper. Only during short term events, such as in river plumes, does the concentration of diuron become great enough to affect corals (risk quotient = 0.3). However, risks posed by mixtures of pollutants may well exceed those presented by individual chemicals.
Despite the relatively low risk associated with chemical pollution on the GBR, evidence is emerging that pollution reduces the resilience of corals and other organisms to global climate change [174] and more research is required to document the increased sensitivity of corals to pollutants at elevated temperatures or under more acidic conditions. Even though the GBR is one of the most highly monitored regions in the tropics, risk quotients can only be calculated for a small number of pollutant types as indicated by data gaps in Table **7**. This general lack of data is more extreme in other tropical regions where other pollutant types may be more relevant and stronger links between regional water quality monitoring and relevant toxicological testing is required. In spite of our growing understanding of the effects of chemical stressors on reef species, there are still gaps in our knowledge which complicate assessment of ecological significance. To characterise a risk to an ecosystem, all relevant information concerning exposure and effect needs to be evaluated, including concentrations and distribution patterns of chemicals. Extremely limited information is available for concentrations of chemical pollutants on coral reefs worldwide. Obtaining these data is difficult as the majority of tropical coral reefs are situated in developing areas and most available data originates from more developed areas such as Florida or the GBR, where usage patterns of pesticides and discharge of industrial and urban waste is likely to differ from pollution patterns in South-east Asia, the Pacific islands or Africa.
Research in the last two decades has identified that coral reefs are threatened by a variety of stressors including elevated ocean temperatures, ocean acidification, overfishing, nutrient input and turbidity and that these pressures vary considerably between regions. While overfishing and nutrification are currently considered as the most detrimental local stressors to coral reefs, current knowledge on the distribution and effects of chemical pollutants is too limited to assess how these compare with other stressors. While laboratory studies have identified toxic thresholds of corals and other tropical organisms to a widening range of pollutants, further monitoring and research needs to be undertaken in several key areas:
# **ACKNOWLEDGEMENTS**
This work was supported by the Australian Government's Marine and Tropical Sciences Research Facility, implemented in North Queensland by the Reef and Rainforest Research Centre Ltd. JvD received financial support from the University of Queensland. Entox is a partnership between Queensland Health and The University of Queensland.
# **REFERENCES**
© 2011 The Author(s). Published by Bentham Science Publisher. This is an open access chapter published under CC BY 4.0 https://creativecommons.org/licenses/by/4.0/legalcode
# **CHAPTER 10**
# **Impact of Contaminants on Pelagic Ecosystems**
**Ketil Hylland1,**\* **and A. Dick Vethaak2,3**
*1 Department of Biology, University of Oslo, Blindern, Norway; <sup>2</sup> Deltares, Delft, The Netherlands and <sup>3</sup> VU University Amsterdam, Institute for Environmental Studies, Amsterdam, The Netherlands*
**Abstract:** Most of the primary production of the world's oceans takes place in the water column, thereby fuelling not only marine pelagic food-webs, but also most benthic communities. In addition, nearly all marine organisms depend on the pelagic zone for some part of their life-cycle. Although most contaminants have physico-chemical properties that cause them to associate with organic material particles and eventually be transported to sediments, direct contaminant inputs are predominantly to pelagic ecosystems. Taking both the ecological importance and the contaminant load into account, there is a surprising lack of scientific knowledge concerning the effects of contaminants in pelagic systems. The main reasons are presumably the difficulty in linking exposure with processes at a scale relevant for environmental management, and challenges involved in using pelagic fish and zooplankton species for experimental studies (excluding the 2-3 copepod species used for regulatory toxicity testing). Contaminants have been shown to affect primary producers as well as secondary producers-consumers, but there is very limited knowledge about ecological impacts. Top predators in marine ecosystems (piscivorous fish species, marine mammals, seabirds) will be particularly at risk from persistent organic contaminants since they will biomagnify. Although there is evidence of effects caused by such substances in the past, there is a need for continuous updates including "new" contaminants. Most relevant for lower trophic levels, micro- and mesocosm studies under controlled conditions are critical for increased understanding of processes and putative effects of contaminants in the pelagic zone. Some field-based strategies have been suggested and implemented to varying degrees for environmental management of contaminants in the water column, including riskbased modelling, bioassay-analyses of environmental samples or extracts (e.g., through the use of passive samplers), caging of organisms and, finally, collection and analyses of native organisms.
# **INTRODUCTION**
The pelagic zone of the oceans constitutes the single largest ecosystem of the world and contains the organisms that form the basis for most marine food chains and all fisheries resources. The characteristics of the marine pelagic ecosystem have been extensively reviewed [1]. Verity *et al.* [1] clearly indicate that the various forms of anthropogenic impacts on the seas, may result in, *i.e.* overexploitation, habitat changes, extinctions, increased disease, species replacements, and how an integrated understanding of resource availability and predation pressure is required for effective environmental management. As will become apparent later in this chapter, increased concentrations of contaminants may affect both bottom-up and top-down processes. Although causing less obvious effects than, for example, overfishing or habitat modification, contaminants are nevertheless important for our understanding and proper management of human interactions with marine pelagic ecosystems.
There are of course spatial and temporal variation of physical and chemical parameters in the pelagic zone, both vertically and horizontally, but it is comparatively stable compared to habitats in most terrestrial or freshwater ecosystems (e.g., Kaiser *et al.* [2]). However, in terms of productivity there are large differences between areas. Whereas coastal areas and shallow seas are among the most productive per area of any ecosystem on the planet, oceanic areas generally have low biomass and productivity [2]. Sunlight-driven primary production needs to take place in the upper reaches of the oceans, sometimes limited to the upper ten or twenty meters. The part of the pelagic zone with the highest primary production will in most cases also be the area that receives contaminant inputs and will have the highest concentrations of such substances. Although there is an extensive literature on oceanographic trace metals, including non-essential metals such as mercury, cadmium and lead, and their behaviour in relation to hydrographic processes and nutrients [3], there is limited data for organic contaminants (see [4]). Organic contaminants are generally thought to be associated with dissolved or particulate organic material, to some extent inorganic particles, and will thus be gradually removed from the water column through sedimentation. Contaminant exposure to pelagic organisms will therefore be from low concentrations in water, through ingestion of particles with
**<sup>\*</sup>Address correspondence to Ketil Hylland:** Department of Biology, University of Oslo, Blindern N-0316 Oslo, Norway; Email: [email protected]
somewhat higher concentrations, through uptake of organic material with associated contaminants or through trophic transfer (which would lead all the way from bacteria and protists to marine mammals, seabirds and humans). Although it should theoretically be simple to quantify the relative distribution and bioavailability of a given substance in pelagic waters by knowledge of its lipid-solubility (and hence affinity for organic material), complex biotic and abiotic processes results in concentrations of contaminants in water, particles or organisms that are difficult to predict (e.g., Ruus *et al.* [5], Vethaak *et al.* [6]). The available data support some general observations; for example, bioaccumulation and possible biomagnification of polycyclic aromatic hydrocarbons in invertebrate food chains, but not in vertebrates [7, 8, 9] (but see Berrojalbiz *et al.* [10]), trophic transfer of persistent organic contaminants [11, 12, 13, 14] and mercury [15, 16], and finally, more or less species- and exposure-dependent accumulation of other trace metals [17, 18].
Major sources of contaminant inputs to the pelagic zone are atmospheric deposition, riverine inputs, shipping activities, land run-off and point discharges. A small proportion of marine contaminants will be directly deposited on the seafloor through activities such as dredging or drilling operations. Sediment-associated contaminants may eventually be a source of input to the pelagic zone through diffusion, resuspension or trophic transfer, but there is limited knowledge about links between contaminants in benthic or demersal species and their predators in the pelagic zone. An emerging problem is the presence of plastic debris and associated contaminants. Contaminants can interact with both floating microplastics and plankton, and thus potentially enter food chains that may ultimately affect humans [19]. Preliminary data show that chemicals in plastic microparticles (<1 mm) are being taken up by marine organisms, including mussels [20]. There is as yet limited knowledge of any effects.
A distinction needs to be made between coastal and oceanic areas. Coastal areas are for natural reasons the waters of the world's oceans with the highest inputs and levels of contaminants, but at the same time areas with a high variability in environmental factors such as particle load, primary production, salinity and temperature. About 30% of oceanic primary production occurs in shelf and coastal environments, constituting less than 10% of the total area of the ocean [21]. The factors discussed above will affect the behaviour of contaminants and how they may impact marine ecosystems [22]. Oceanic areas are less variable than coastal areas and sources of contaminants are limited to atmospheric deposition, offshore oil and gas activities, shipping discharges and, to a lesser extent, the presence of plastic debris.
Over the last decade there has been an increasing number of studies reporting the concentrations of contaminants in surface and microlayer water [23, 24], associated with plastic resin pellets [25], passive samplers [26, 27], particulate material [24, 28] and caged or pelagic organisms [5, 24, 29]. As will be discussed in greater detail below, there are obvious problems in trying to assess the effective concentration of contaminants in water-masses, both due to the variable solubility, speciation, association with particles and bioavailability of contaminants and because water-masses move and mix.
There is even less data for contaminant-related effects in pelagic ecosystems. Nearly all marine model organisms for laboratory- or field-based studies on contaminant effects are benthic species, including blue mussel (Mytilus edulis; [30, 31]), dab (Limanda limanda; [32, 33]), eelpout (Zoarces viviparus; [34]) and flounder (Platichthys flesus; [35, 36], [37]). There are however, some studies that have targeted pelagic species or used caged species. The BECPELAG (Biological Effects of Contaminants in Marine Pelagic Ecosystems; [23]) workshop investigated effects and levels of contaminants in pelagic systems through field-collected organisms [38], caged organisms [39] and bioassays of water and passive sampler extracts [40]. Organisms studied ranged from invertebrates to fish. The results from the workshop clearly showed that levels and effects of contaminants in field-collected organisms were less clear than in organisms caged in the same area. Other studies have focused on species at the top of food chains such as swordfish, for which there are indications of relationships between contaminant levels and sublethal endocrine disrupting effects [41].
The aims of this chapter are to review the current understanding of how contaminants affect pelagic ecosystems, outline approaches and to suggest research directions.
# **CHALLENGES**
There are reasons why benthic organisms and systems have been preferred to pelagic systems in contaminant research. As hinted to above, ecological importance is certainly not the reason and many pelagic fish species are as economically important as benthic species. One reason for the preference of benthic species for research in general is accessibility – intertidal or shoreline species require less infrastructure for their collection and study than organisms in the water column. Secondly, benthic species are generally more amenable to being kept in the laboratory and there is hence much more general knowledge about their biology. Thirdly, and possibly most important, concentrations of contaminants are orders of magnitude higher in sediment than in the water column, at least in theory resulting in higher exposure levels for sediment-dwelling than for pelagic organisms. However, exposure levels in the two habitats will vary considerably for different groups of contaminants. Pelagic organisms will generally be exposed to higher levels of the more easily degradable substances than their benthic counterparts. Finally, there is a difference between benthic and pelagic organisms in our knowledge of their exposure history (or at least perceived knowledge). Whereas many benthic species, for example blue mussel, are sedentary and stationary, pelagic species move continuously. Although contaminants may enter marine ecosystems through pelagic waters, there is a feeling that it is easier to quantify exposure for benthic than for pelagic species. In enclosed water bodies such as fjords or estuaries this may be true, but in the open sea it is not obviously a clearer relationship between contaminants in abiotic matrices such as sediment and epibenthic organisms than between concentrations in water and pelagic organisms. Even for benthic organisms there are not obvious quantitative relationships between contaminants in sediment and the tissues of sediment-dwelling organisms [42, 43], and sediment-related factors such as black carbon strongly affects bioavailability even of organic contaminants [8, 44].
One major challenge for understanding contaminant exposure and effects for pelagic organisms concerns their presence in and exposure to different water masses. For planktonic organisms this is not necessarily the case as they will remain associated with a water mass for periods of time, but nekton such as fish will clearly be exposed to different levels of contaminants as they move through more or less contaminated water masses.
A relevant question here is how contaminant exposure in marine ecosystems can be most precisely estimated. For species with low metabolising capacity, accumulated concentrations of many organic contaminants and nonessential metals will be a reasonable estimate for long-term exposure. Other species, and particularly vertebrates, will to a larger extent regulate their intake and accumulation of non-essential metals and metabolise and excrete a variable fraction of absorbed organic contaminants. Although some organic contaminants have half-lives in the range of years in most organisms [45, 46], most are metabolised at least to some extent and some, such as alkylphenols and polycyclic aromatic hydrocarbons, to the extent that tissue residue analyses are less useful than analyses of metabolites in bile or other excretory fluids [47, 48, 49]. As mentioned above, there is a complex relationship between contaminant concentrations in abiotic matrices (sediment, water) and concentrations in tissues, particularly for mobile species. Contaminant exposure may therefore be most accurately determined from tissue concentrations for persistent substances and metabolite levels for others. There are however some other alternatives and we will focus particularly on the pelagic organisms here. There is limited knowledge about the ecotoxicology of this group of organisms, but zooplankton does not appear to metabolise organic substances efficiently [50] (but see Magnusson *et al.* [51]), and they accumulate a range of metals [52] as well organic contaminants [11, 53] and would therefore be a useful matrix by which to estimate exposure in any given water-mass. Using zooplankton for this purpose would however need to be part of a carefully designed experiment to ensure spatial representivity, and vertical migration patterns would need to be taken into account. In the photic zone phytoplankton could be used for the same purpose, although any vertical movement would have to be considered for the species used. A second alternative is to use passive samplers: a range of different materials have been used, including membranes with a lipid inside [54], silicone sheets [26, 55], various plastics [56], coated membranes [57] or polyurethane foam [58]. Common to most passive samplers as they have been deployed until now is the need for a mooring system. Passive samplers are generally deployed for a period of three to six weeks prior to extraction and chemical analyses.
# **PRIMARY PRODUCERS**
Phytoplankton forms the basis of marine food webs and embodies the carrying capacity of marine ecosystems. In the classical view, the main route for organic carbon was through zooplankton feeding on phytoplankton, but it is now well established that microzooplankton, bacteria and probably viruses play crucial roles in affecting the trophic dynamics and composition of plankton communities [59, 60]. Our knowledge of how and whether contaminants affect these organisms and interactions between them is limited.
### *Impact of Contaminants on Pelagic Ecosystems Ecological Impacts of Toxic Chemicals* **215**
The increase in primary production in coastal waters since the 1970s, at least to some extent due to increased nutrient inputs, has received much attention from the scientific community as well as from environmental managers. In many coastal systems, phytoplankton blooms are common events and a significant amount of this phytoplankton biomass will sediment through the water column, settle on the bottom and the nutrients be remineralised in surface sediments [61]. Increases in the occurrence of algal blooms have been linked to phenomena such as oxygen deficiency and mass kills of benthic fauna and fish as well as the formation of foam on beaches (produced by algae species such as Phaeocystis) and toxic shellfish.
To what extent will chemical stressors affect primary producers? Given the large amount of new, industrially produced substances, this is an important and relevant issue for the coming decades. Results from experimental studies indicate that certain chemicals may have a direct impact on plankton communities and food chains, and may thus potentially affect the carrying capacity of estuarine and coastal ecosystems. The most important compounds for causing toxic effects upon phytoplankton are pesticides and biocides, especially those with a herbicidal mode of action. The antifouling agent TBT has been shown to affect phytoplankton communities at concentrations that are present in coastal waters [62]. Effects include reductions in population development rate and shifts in species composition – *i.e.*, towards species that are more tolerant to TBT pollution. Worldwide measures to restrict TBT in antifouling paints (with a total ban by 2008) has lead to the development of alternative antifouling compounds such as zinc pyrithione (ZPT), copper pyrithione (CPT), Irgarol 1051 and diuron [63, 64, 65]. Residues of these novel antifouling agents are currently found worldwide, especially in estuarine and coastal waters near and in contaminated marinas. Irgarol 1051, like other triazine herbicides, is a strong inhibitor of photosystem II and reduces growth and productivity of sensitive phytoplankton species [66]. Some phytoplankton species appear to be more sensitive to Irgarol 1051 than others. For example, a 23-h exposure to Irgarol (112 ng/L) decreased the abundance of some eukaryotic species to less than half of the controls [67]. Zamora-Ley *et al.* [63] found in a marine harbour that Irgarol 1051 caused changes in several phytoplankton species with increasing herbicide concentrations.
Maraldo and Dahllöf [64] found that the acute toxicity of the antifouling agents ZPT and CPT among natural phytoplankton communities was similar to that of TBT [62], which in turn was higher than those reported for Zn and Cu alone [64]. The sensitivity towards ZPT and CPT was dependent on the phytoplankton community structure and the density of algae and suggested an enhanced effect of ZPT and CPT under phosphate-limiting conditions.
The effects of the herbicide atrazine on marine phytoplankton typical of the German Bight (North Sea) were demonstrated in mesocosm experiments [68]. The authors reported reduced photosynthesis accompanied by lower chlorophyll concentrations and reduced primary production. Other recent experimental work have demonstrated that the pharmaceutical clotrimazole can affect marine microalgal communities at picomolar concentrations, but the true potential for impact on marine primary producers has not been established [69].
The development of plankton communities in estuarine and coastal waters is governed by highly dynamic physical and chemical processes. This makes it hard to predict or establish the effect and ecological significance of chemical compounds on these communities. The potential impact of chemicals on phytoplankton and phytobenthos communities in coastal waters is known to depend on environmental factors such as salinity, temperature, nutrients, and exposure to UV-A and UV-B radiation and contaminants. Although contaminants may affect phytoplankton, any effects might be masked by other factors and interactions. To tackle this problem field studies complemented with mesocosm experiments should be conducted to improve control over factors and to improve the ecological relevance of the findings.
Another aspect of chemical stress on plankton and other organisms higher in the food chain are natural toxins produced by marine algae. As a consequence of changes in the coastal zone, the frequency and intensity of toxic algal blooms might increase, resulting in increased levels of natural toxins. The risk of toxic algal blooms can also increase as a result of unintended introductions of new invasive species, for example by ballast water releases. However, it remains difficult to quantify ecological impacts of such natural toxins because available toxicity data are limited. The relative contribution of anthropogenic chemical compounds and natural toxins on the total chemical pressure under field conditions is therefore unknown, and we lack insight into any interactions between these groups of chemicals.
# **SECONDARY PRODUCERS AND TERTIARY CONSUMERS**
Secondary production includes the consumption of primary producers and biomass generated by heterotrophs. Tertiary consumers include predatory fish and fish-eating mammals and birds. Long-term changes of offshore zooplankton appear to be mainly associated with climatic and hydrographic phenomena [70]. Any direct or indirect effects of contaminants on marine zooplankton are not well understood. Bioaccumulation of metals and organic contaminants in marine zooplankton including jellyfish has been reported, [71, 72]. An obvious challenge in this context is the identification and separation of different species in a sample. In a comprehensive study, Hoekstra and co-authors concluded that concentrations of organic contaminants in zooplankton predominantly reflected chemical partitioning and that there was limited biotransformation by the *Calanus* species investigated [71]. Although organochlorine contaminants do not appear to be metabolised extensively by zooplankton, there is some evidence that polycyclic aromatic hydrocarbons may be [10].
Toxicity information for zooplankton is limited, except for the few species used in toxicity testing (mainly *Acartia, Nitocra*, *Tisbe* and mysids, [73, 74, 75]), although there is some indication that, e.g. insecticides affect coastal zooplankton [76]. Toxic effects have been shown for TBT at concentrations present in coastal waters [77]. The observed effects included reduced population development rate and shifts in species composition.
A high potential for bioaccumulation of endocrine disrupting compounds (*i.e.*, organotins, flame retardants) and indications of endocrine disrupting effects have been demonstrated for the estuarine mysid *Neomysis integer* [72, 78]. This species plays a key role in the transfer of energy between phytoplankton and fish production in estuaries and along shallow coastal waters in northern Europe, and between benthic and pelagic food webs. Furthermore, some studies have investigated effects of contaminants on population-level effects in the ecologically very important copepod genus *Calanus* [79, 80]. A limited number of studies have evaluated the application of sublethal effect protocols and biomarkers, in phyto- and/or zooplankton species [78, 81]. However, there have been some recent studies using transcriptomic approaches for ecologically important *Calanus* species [82, 83, 84].
A number of studies indicate that eggs and larvae of pelagic and demersal fish that float in surface and subsurface layers may be particularly sensitive to diffuse contaminant exposure (including PAHs from oil pollution) and sublethal effects [85, 86, 87]. Unfortunately, the full impact of contaminants on critical life stages of fish and other nekton is still largely unknown.
Several studies have demonstrated effects of contaminants on sublethal responses in selected pelagic fish species. In studies with saithe (*Pollachius virens*) as part of the BECPELAG workshop, tissue-level effects were observed in fish collected close to a production platform in the North Sea [88]. A North Sea monitoring study using a predominantly demersal feeding species, haddock (*Melanogrammus aeglefinus*), reported a range of effects in this species linked to the presence of populations in or near areas with offshore activity [89]. There were substantially increased levels of DNA damage and changes in the lipid composition of membranes in haddock collected in areas with high offshore activity. The effects were corroborated by other biomarkers and showed a total picture of a population with increased DNA damage mainly due to PAH exposure (indicated through elevated PAH metabolite concentrations), but also increased oxidative stress resulting in changed lipid composition [89]). However, the ecological significance of the observed effects remains unresolved.
Fossi and co-workers [41] showed that large pelagic predators, bluefin tuna (*Thunnus thynnus*), swordfish (*Xiphias gladius*) and Mediterranean spearfish (*Tetrapturus belone*), contained increased levels of vitellogenin (VTG), a yolk precursor protein only expected to be present at appreciable quantities in female fish. Such levels are most likely caused by accumulation of endocrine-disrupting substances through their diet. Another study by De Metrio *et al.* [90] supported these findings and showed that close to a quarter of caught male Mediterranean swordfish (*Xiphias gladius*) displayed ovotestis (intersexuality), again possibly caused by endocrine-disrupting compounds (EDCs). Furthermore, elevated VTG levels were found in liver tissue. The causes of these phenomena are not yet known, but bioaccumulation of endocrinologically active substances is a possible explanation. The evidence of wide-spread EDC exposure in the marine environment is supported by studies of Scott and co-workers [91, 92], who observed offshore male cod (*Gadus morhua*) and male dab (*Limanda limanda*) with elevated levels of VTG.
Because of bioaccumulation and biomagnification processes in food webs, globally distributed persistent organic pollutants (POPs), including EDCs, may attain high concentrations, in pelagic top predators. Such substances may reach levels that result in effects on reproductive and/or immune systems. This has been well illustrated in field studies on Baltic grey and ringed seals, and semi-field studies with Wadden Sea harbour seals. Those studies have shown that reproduction and immune functions can be impaired in top predators following biomagnification of PCBs in the food chain (see review by Vos *et al.* [93]). Reproduction effects have resulted in population declines and may also have contributed to the mass mortalities observed in some European seal populations due to virus infections.
Numerous other cases refer to mass mortalities by infectious diseases, poor reproductive performance, immunosuppression, thyroid abnormalities and other non-reproductive disorders in marine mammals and fish-eating birds (for reviews, see Vos *et al.* [93] and Law *et al.* [94]). Such effects have to some extent been associated with the presence of POPs (e.g., organochlorine compounds, brominated flame retardants and metabolites) and other endocrine disrupting and/or immunotoxic compounds in the body fat [95]. Bennett *et al.* [96] found an association between chronic exposure to mercury and infectious disease in harbour porpoises. An increase in disease susceptibility in contaminant-exposed whale and dolphin populations has further fed speculation about a possible negative influence of contaminants on the immune system [97]. Accumulation of persistent and lipophilic contaminants, including polychlorinated dibenzo-p-dioxins (PCDDs), polychlorinated dibenzofurans (PCDFs), and coplanar polychlorinated biphenyls (coplanar PCBs), were found in several albatross species feeding in the open oceans, specially the North Pacific Ocean. Possible adverse effects of these compounds to these birds may be expected from toxic equivalent (TEQ) levels [98]. However, in most of these cases, it was not possible to confirm a cause-effect relationship between a specific chemical or group of chemicals and individual or population level effects. Studies over the last decade have shown high concentrations of a range of substances of concern in marine top predators, including TBT [99, 100], toxaphenes [101], polybrominated diphenyl ethers [101, 102, 103, 104], perfluorooctane sulfonates PFOS and perfluorooctanoic acid PFOA [105, 106]), nonyl- and octylphenol [107] and phthalate esters [108]. Single and combined impacts of food-chain accumulation of these contaminants and subsequent high concentrations in marine pelagic secondary producers and tertiary consumers has yet to be elucidated. In addition to the above, increasing levels of human pharmaceuticals, personal care products and aquaculture veterinary pharmaceuticals in coastal pelagic ecosystems is an area of concern with limited knowledge of any ecological impacts [109].
# **INTERACTIONS BETWEEN ENVIRONMENTAL FACTORS**
Three of the most obvious pressures from human activity in marine waters are eutrophication, oil and contaminant inputs. For eutrophication there is extensive data on nutrient and bloom dynamics in coastal areas [110]. There are large amounts of data on the environmental physiology of many algal species. There is also a substantial body of knowledge on how oil and offshore-related discharges affect marine ecosystems, not least from monitoring following accidental spills from, for example, Exxon Valdez [111] or Prestige [112]. Aspects of the consequences of offshore-related effluents were evaluated recently through the BECPELAG workshop [23]. Finally, there is a large literature on the presence and effects of contaminants in coastal ecosystems even at low exposure levels [113]. Although there is limited evidence of large-scale effects of contaminants in marine ecosystems, possibly with the exception of Puget Sound, USA [114] and the North Sea and Baltic in the 1970-80s [86, 115], there is reason to believe that chronic exposure to low levels of contaminants will affect pelagic organisms.
Eutrophication, oil and contaminant inputs are co-occurring features of most estuaries and harbours in industrialised countries. Organic enrichment, the presence of oil, contaminants and variable oxygen availability would be expected to interact in their effects on marine biota, but there are surprisingly few studies on whether and to what extent this is the case (but see Gunnarsson *et al.* [22] and Herman *et al.* [116]). Natural waters contain both dissolved (DOM) and particulate organic material (POM), both of which may act as "sponges" to mop up organic and many inorganic contaminants in the water column. Increased levels of organic material could therefore be expected to modulate effects of contaminants through decreased bioavailability in water or increased sedimentation and "co-precipitation" of contaminants. For filter-feeding organisms in the water column, association of contaminants with particles may actually increase exposure as both food and water will contain contaminants. For predators this process would decrease water-borne exposure, but increase exposure through the food chain. Water-soluble components of oil would behave as other contaminants in this context, whereas dispersed oil would be expected to behave like DOM. It is not clear how algal, bacterial and protist interactions may be affected, although specific effects from contaminants on any one group would be expected to affect energy and nutrient flows in the network. Association of contaminants with particles will generally decrease residence time in the water column and thus shift exposure from pelagic to sediment ecosystems.
Despite existing knowledge about eutrophication effects in pelagic systems, there is a need for further knowledge about how natural systems behave under conditions of varying nutrient or carbon availability and there is limited understanding about how oil or contaminants may interact in such systems. Small-sized organisms could be thought to be at greater risk since they would be expected to accumulate higher concentrations of contaminants, but organisms that accumulate non-limiting substrates may also have a high uptake [117]. The question remains whether organisms that accumulate high concentrations of contaminants are most sensitive to the effects of the contaminants. In addition to ecological consequences of modulating the systems themselves, changes in both small and medium scale pelagic processes could strongly affect fluxes and effects of contaminants in coastal ecosystems through affecting sedimentation and transfer to higher trophic levels.
Combined effects between UV radiation and contaminants on plankton community structure in coastal zones have been observed in several recent studies. Major coastal and marine contaminants that still often exceed environmental risk limits in estuarine and coastal waters, such as TBT, PAH, Irgarol or atrazine have phototoxic capacity and proven or suspected impact on planktonic species composition and communities. Microphytobenthos and phytoplankton might be especially sensitive to such phototoxic effects. What appeared to be a synergistic interaction between TBT exposure and UV-B radiation effects on a natural planktonic assemblage was found by Sargian [118] and Pelletier *et al.* [119] using a microcosm approach. Deleterious effects of TBT exposure were significantly more pronounced when cells were co-exposed to enhanced UVB levels. The same author also found a reduced bacterial production in the presence of TBT. Hjorth and co-workers [120] observed effects of the polycyclic aromatic hydrocarbon pyrene on a natural marine plankton community using a food-web approach in a mesocosm. Direct and indirect effects on the function and structure of bacteria, phytoplankton and to a lesser degree on zooplankton communities were found. The change in system function suggested that PAHs might be an important stress factor for pelagic systems, as a one-time exposure of a single compound changes the development of a pelagic community.
An important finding was recently reported by Echeveste *et al.* [121]. These authors performed *in situ* experiments on board of a research vessel in the NE Atlantic Ocean that determined the influence of complex mixtures of organic pollutants on oceanic phytoplankton populations. The results of these experiments suggest that current levels of POPs are only 20 times below the levels at which significant influence on ecosystem function (primary productivity) would be found.
**Table 1:** Alternative strategies for pelagic environmental assessment.
# **APPROACHES**
There are substantial logistical challenges involved in the study of how contaminants may affect pelagic systems or species. Micro- or mesocosm studies are required for detailed studies of specific effects or interactions between factors. For lower trophic levels, mesocosm studies are generally required to assume any kind of ecological relevance. In the field, four approaches have been used:
The five approaches all have weak and strong characteristics, outlined in Table **1**.
# **RESEARCH NEEDS**
As will be apparent from the above, there are large blank areas in our understanding of how and whether contaminants impact pelagic ecosystems. On the other hand, knowledge of the pelagic zone is clearly vital in the management of our oceans. In this context it is important not to view the pelagic zone in isolation, but remember that pelagic processes are important to both the surface layer and benthic ecosystems. Future research should be directed towards integrating and not dividing our understanding of different environmental compartments.
As for all other fields in ecotoxicology, we face a major challenge in developing methods to assess the effects of contaminant mixtures. For pelagic systems this may be particularly relevant since even the less persistent contaminants will be present in the water column near the source. In addition to contaminant mixtures, there is a scarcity of knowledge on how other factors modulate contaminant impacts or combination effects. Micro- and mesocosm model systems (see below) should be useful tools in this context.
It will be clear that there is a need for an improved understanding of how contaminants affect both primary producers and microbial loop components. Current knowledge is limited to effects on single algal species and there is virtually no knowledge of impacts in more complex systems that include bacteria and protists.
There is some understanding of how some contaminants affect a limited number of zooplankton species (e.g., calanoid copepods), but little is known about the wide range of mesozooplankton species, including metamorphosing stages and effects on their sensory systems [127].
It is inherently challenging to keep pelagic fish species and their early life stages for experimental studies due to the need for specialised sampling techniques and large volume aquarium systems. In contrast to primary producers and zooplankton, there is a substantial knowledge of general physiology and biochemistry that can be applied for fish, even though there may be species-dependent contaminant-associated effects. There are even larger obstacles involved in experimental studies of pelagic top predators.
In addition to experimental micro- or mesocosms, four approaches have been used for the assessment of contaminant effects in marine pelagic ecosystems: modelling, *in situ* extracts/passive samplers, caging and field collection. Both laboratory- and field-based methodologies are needed and they complement each other.
# **REFERENCES**
### **220** *Ecological Impacts of Toxic Chemicals Hylland and Vethaak*
### **222** *Ecological Impacts of Toxic Chemicals Hylland and Vethaak*
### **224** *Ecological Impacts of Toxic Chemicals Hylland and Vethaak*
© 2011 The Author(s). Published by Bentham Science Publisher. This is an open access chapter published under CC BY 4.0 https://creativecommons.org/licenses/by/4.0/legalcode
# **CHAPTER 11**
# **The Role of Aquatic Ecosystems in the Elimination of Pollutants**
**Matthew T. Moore1,\*, Robert Kröger2 , and Colin R. Jackson<sup>3</sup>**
*1 USDA Agricultural Research Service, Oxford, Mississippi, USA; <sup>2</sup> Mississippi State University, Mississippi, USA and <sup>3</sup> University of Mississippi, Mississippi, USA*
**Abstract:** Contamination of aquatic ecosystems is always of concern to environmental scientists; however, these systems also possess unique capabilities allowing them to eliminate or remediate certain levels of pollutants. Primarily through the presence of vegetation, aquatic ecosystems are known to be capable of removing or at least decreasing pollutant loads travelling through the aqueous phase. In addition to vegetation, soil/sediment and microbes play a significant role in transferring or transforming pollutants to acceptable levels in aquatic ecosystems. This chapter focuses on some of the primary literature describing phytoremediation of organic pollutants (e.g. hydrocarbons and pesticides) and inorganic pollutants (e.g. metals and nutrients). Research indicates the popularity and success of phytoremediation techniques used to clean up both organic and inorganic pollutants from the water column. While certain caution should always be exercised, phytoremediation continues to serve as a successful means of pollutant remediation in aquatic ecosystems.
# **INTRODUCTION**
Aquatic ecosystems are often receptacles of point- and non-point source pollutants from spills, sprays, or runoff events. While much emphasis is placed on aquatic ecosystem damage from pollutants, research has demonstrated these unique systems have resilience and assimilative capacity in the mitigation of such pollutants. This chapter will focus on aquatic ecosystem responses to metal, nutrient, and pesticide inputs, primarily discussing the concept of pollutant remediation *via* plants (phytoremediation) and microbes. Because various review articles have been published regarding specific phytoremediation techniques [1, 2], this chapter is not meant as an exhaustive literature review. Instead, it provides a broad understanding of some of the principle concepts involved in aquatic system remediation (through plants) of common pollutants.
# **PHYTOREMEDIATION**
Phytoremediation is generally defined as the use of plants and associated microbes to remove, contain or render harmless environmental pollutants [2, 3]. The nature of pollutants will affect their ability to successfully undergo phytoremediation. For example, while organic pollutants can be degraded, inorganic pollutants such as nutrients are unable to be degraded. Instead, through processes of phytoremediation, inorganic pollutants can be stabilized or sequestered. According to Susarla *et al.* [4], three general factors affect pollutant uptake and distribution within plants used in phytoremediation efforts: physicochemical properties of the pollutant (e.g. octanol water partition coefficient, vapor pressure, water solubility); environmental conditions (e.g. pH, temperature, soil moisture, organic matter); and plant characteristics (e.g. available enzymes and root systems). In addition to the factors affecting pollutant uptake, phytoremediation itself has five major mechanisms by which the process may operate [2, 4, 5, 6].
**Francisco Sánchez-Bayo, Paul J. van den Brink and Reinier M. Mann (Eds) © 2011 The Author(s). Published by Bentham Science Publishers**
**<sup>\*</sup>Address correspondence to Matthew T. Moore:** USDA Agricultural Research Service, National Sedimentation Laboratory, Oxford, Mississippi 38655, USA; Email: [email protected]
Not all mechanisms are equally effective for remediation of all pollutants. Phytoextraction, phytoaccumulation, and phytostabilization are efficient mechanisms for remediation of many metals, including cadmium, chromium, lead, nickel, and zinc. Mercury, selenium, and various chlorinated solvents are effectively remediated through phytovolatilization. Pollutants such as munitions, chlorinated solvents, and certain pesticides are best remediated through phytotransformation and phytodegradation. Rhizodegradation is an effective mechanism for remediation of radionuclides, certain organic chemicals, and metals.
# **AQUATIC SYSTEM REMEDIATION OF ORGANIC POLLUTANTS**
Studies of organic pollutant remediation in aquatic systems tend to focus on structures such as oxbow lakes, detention ponds, riparian buffer zones, vegetated drainage ditches and constructed wetlands. As with inorganic pollutants, remediation occurs not only in and around vegetation, but also within sediment and aqueous phases *via* chemical and microbial processes. Polarity and lipophilicity of pollutants give reliable indications on their ability to be remediated *via* vegetation. Limited plant-pollutant uptake will be achieved with chemicals which are extremely polar due to difficulty in crossing biomembranes [7]. On the other hand, extremely lipophilic pollutants quickly penetrate biomembranes, only to be sorbed to root material. It is the pollutants with intermediate lipophilicity which are best remediated by vegetation. These pollutants can be translocated to upper plant parts, rather than become concentrated in root material [7]. According to Chaudhry *et al.* [8], by reducing plant wax viscosity, uptake of nonpolar compounds will be enhanced. Additionally, factors which increase leaf cuticle hydration increase the permeability of hydrophilic compounds.
Once pesticides are absorbed in plant material, three main reactions are responsible for pollutant transformation [8]:
# **Specific Examples of Organic Pollutant Phytoremediation**
Euliss *et al.* [9] compared reduction of petroleum hydrocarbons found in sediments with sedge (*Carex stricta*), switchgrass (*Panicum virgatum*), and gamagrass (*Tripsacum dactyloides*) versus sediments under willow (*Salix exigua*), poplar (*Populus* spp.) or no vegetation. Significantly fewer residues of petroleum hydrocarbons (70%) were in sediments with sedge or grass; whereas only 20% fewer residual hydrocarbons were noted in the sediments containing trees or no vegetation. Two aquatic plants, *Juncus fontanesii* and *Lemna minor* have reportedly removed phenol concentrations ranging from 8 to 48 mg/L [10]. Polychlorinated biphenyls, another common organic pollutant, has been shown capable of being transferred from an aqueous spiked solution into plant material from the common reed (*Phragmites australis*) and rice (*Oryza sativa*) [11].
A great deal of phytoremediation literature addresses the ability to reduce pesticides. Many studies examine remediation capabilities within stream mesocosms, constructed wetlands, or vegetated drainage ditches. Several studies have examined the influence of vegetation on the reduction of pyrethroid insecticide concentrations in aqueous solution. In a mesocosm experiment, Moore *et al.* [12] reported *cis-*permethrin reduction ranging from 67 ± 6% in common cattails (*Typha latifolia*) to 71 ± 2% in cutgrass (*Leersia oryzoides*). Another study conducted by Moore *et al.* [13] examined permethrin mitigation in constructed ditches in Yolo County, California. The ditch distance needed to reduce permethrin concentrations to half of their original inflow concentration (D1/2) in non-vegetated ditches (50-55 m) was basically twice that of vegetated ditches (21-22 m). Bennett *et al.* [14] determined that in order for initial bifenthrin and lambdacyhalothrin aqueous inflow concentrations to be reduced to 0.1% of their initial value, a vegetated ditch 280 m would be necessary. Other vegetated ditch studies have reported 87% of the mean measured lambda-cyhalothrin was associated with plant material [15]. In a constructed wetland experiment, 49% of measured lambda-cyhalothrin was associated with vegetation, while 76% of cyfluthrin was found in vegetation [16].
Various studies have also examined the remediation of organophosphate insecticides and different herbicides with vegetation. After dosing a field-scale constructed wetland in the Mississippi Delta, USA, Moore *et al.* [17] reported 43% of the measured mass of the insecticide diazinon was associated with wetland plant material. In a California study, diazinon was amended into two constructed ditches, one vegetated and one non-vegetated. Ditch halfdistances (D1/2) were calculated (see previous paragraph for description) and results indicated a non-vegetated ditch would need three times the distance (158 m) of a vegetated ditch (55 m) to remediate the same diazinon concentration [13]. Comparing methyl parathion transport in vegetated versus non-vegetated constructed wetlands, Moore *et al.* [18] reported pesticide concentrations were detected in outflow samples of the non-vegetated wetland 30 min after initial dosing. During the same time sequence, methyl parathion concentrations in the vegetated wetland were only measured at 20 m (slightly less than half way through the system). Semi-permeable membrane devices deployed in both wetlands confirmed that, although methyl parathion concentrations reached the nonvegetated wetland outflow, no pesticide was detected in the vegetated wetland outflow [18]. Experimental constructed wetland mesocosms at the University of Mississippi Field Station were utilized for specific pesticide phytoremediation studies in the late 1990s. Results from those studies indicated that 25% and 10% of measured chlorpyrifos and metolachlor (herbicide), respectively, were associated with wetland plant material [19, 20]. A study examining atrazine mitigation in a vegetated drainage ditch populated with *Polygonum* spp., *Leersia oryzoides*, and *Sporobolus* spp. reported 61% of measured herbicide concentrations were associated with plant material [15]. Rice *et al.* [21] examined radiolabelled pesticide concentrations in aqueous solution in vegetated versus non-vegetated systems. In different systems vegetated with *Ceratophyllum demersum*, *Elodea canadensis,*and *Lemna minor*, 1%, 4%, and 23% of 14C-metolachlor, respectively, remained in aqueous solution, while 61% of the pesticide was present in non-vegetated system aqueous solutions. Likewise, 14C-atrazine was amended into identical systems. Percentage of pesticide remaining in aqueous solution was 41%, 63%, and 85% for *C. demersum*, *E. canadensis*, and *L. minor*, respectively. In non-vegetated systems, 85% of 14C-atrazine remained in aqueous solution [21]. Rose *et al.* [22] monitored reduction of the herbicide fluometuron in open and vegetated ponds for consecutive growing seasons. Significant differences (58% reduction in vegetated pond versus 41% reduction in open pond) was noted during the second incubation of the second season.
# **Microbial Remediation of Organic Pollutants**
Organic pollutants represent a vast range of chemicals with diverse properties and varying degrees of toxicity and recalcitrance to microbial remediation. Common pollutants include petroleum hydrocarbons, polycyclic aromatic hydrocarbons (PAHs), nitroaromatics, polychlorinated biphenyls, industrial solvents and various pesticides [23]. The biodegradation of these materials has received a substantial amount of interest over the last two decades and interactions between various organic pollutants and microorganisms have been examined from physiological, molecular, and even evolutionary perspectives [24, 25, 26, 27]. At a basic level organic materials can be separated into those that are biodegradable (*i.e.* are transformed by microorganisms into more innocuous products, ultimately into carbon dioxide and water), those that are persistent (materials which are not biodegradable in certain environments), and those that are recalcitrant (materials which are resistant to biodegradation in most situations). While individual microorganisms are capable of degrading simple organics, typically the complete biodegradation of organic pollutants may involve the metabolic activity of several microbial populations acting as a consortium [28].
As well as the fundamental capability of natural bacterial populations to degrade an organic pollutant, a number of other considerations are important and may impact the effectiveness of microbial remediation [28]. Organic pollutants typically occur as mixtures of different groups of chemicals and even within specific groups (e.g. PAHs) there is a great deal of variability on degradability [28, 29]. Different organic substrates (or their degradation byproducts) may interfere with the microbial pathways used to degrade other substrates, impairing effective remediation. As well as substrate variability, certain microorganisms might also require electron acceptors other than oxygen; for example, sulfate reducing bacteria may be particularly effective in the reductive dehalogenation of highly substituted materials such as organochlorine compounds [30]. Even if metabolically suitable microbial populations and electron acceptors are present, interactions between the microorganisms and pollutant may be limited. Organic compounds can become associated with polymers in the soil matrix limiting their accessibility to microbial populations [31], and the same could occur in aquatic sediments. At a more fundamental level, many organics are also insoluble in water, and this is likely to be a limiting factor in the degradation of materials such as polychlorinated biphenyls in aquatic ecosystems, although the microbial production of surfactants may overcome this to some extent [23]. Limited accessibility of microbial populations to organic pollutants also means that the concentration of pollutant that the microorganisms are exposed to may be substantially lower than the actual concentration in the system, and these lower concentrations may be below the threshold needed for induction of degradative enzyme systems even if the natural microbial populations contain them [32, 33]. The movement of microorganisms towards increasing concentrations of pollutants (*i.e.* positive chemotaxis) may be just as important as actual degradative ability in the microbial remediation of some organic contaminants [29].
Populations of the bacterium *Pseudomonas putida* can show both chemotaxis towards and the ability to degrade the two-ring PAH naphthalene if they possess the appropriate plasmid [34], and natural microbial populations are likely to show the same capabilities. Naphthalene is a common organic micropollutant in water and many bacteria capable of degrading naphthalene have been isolated [29]. Similarly a large number of bacteria appear to be capable of degrading three-rings PAHs such as phenanthrene, and as with naphthalene degraders, these bacteria represent a diverse range of bacterial taxa [35, 36, 37]. The capability of microorganisms to degrade higher molecular weight PAHs such as benzo[*a*]pyrene, a five-ring carcinogenic compound that is a commonly formed from combustion of organic material, is much more limited [37]. Those bacteria that can oxidize benzo[*a*]pyrene generally do so through cometabolism, requiring the presence of other organic substrates either for metabolism or to stimulate PAH degradation [38, 39]. Various fungi have been shown to be potential degraders of benzo[*a*]pyrene and other high molecular weight PAHs in terrestrial environments [37, 40], but the importance of fungi as degraders of PAHs in aquatic ecosystems is not known.
Nitroaromatic organic compounds released from incomplete fossil fuel combustion and as feedstock in the manufacture of materials such as pesticides are generally regarded as being fairly recalcitrant to bioremediation, especially through oxidative reactions [27, 41]. Few microorganisms are capable of using nitroaromatics substrates as their sole source of carbon and/or nitrogen, although many more appear to be capable of reducing nitroaromatics to corresponding aminoaromatics through the action of various nitroreductases [27]. This typically occurs under anaerobic conditions, and may be the major method by which poly-nitroaromatic compounds can be degraded [42]. Intermediate products, however, may be more toxic than the original pollutant, and effective mineralization in aquatic sediments is likely to require consortia of many different interacting microbial populations. Simpler monoand di-nitroaromatics are mineralized aerobically by some bacteria that potentially use them as a source of carbon, energy, and nitrogen [27]. Various actinomycetes and pseudomonads can hydroxylate the nitro groups in 2 nitrophenol and 4-nitrophenol, releasing nitrite and forming dihydroxybenzene which is subsequently mineralized [43, 44, 45]. These reactions are important in the microbial degradation of the pesticides parathion and methyl parathion which are first hydrolyzed to yield 4-nitrophenol, which is subsequently hydroxylated [46]. Monoxygenases and dioxygenases are involved in the hydroxylation of mono- and di-nitroaromatics, respectively, but other aerobic degradation mechanisms exist in some bacteria [27]. However, compared to many non-nitrogen containing organic pollutants, nitroaromatics are more resistant to microbial mineralization and the majority of studies have been at the bench- or laboratory-scale rather than in natural environments.
As with nitroaromatics, most organic molecules that contain substituted groups are more recalcitrant to microbial remediation than simpler hydrocarbons. This is especially true of halogenated organic pollutants, even those with relatively simple modifications of aromatic hydrocarbons such as chloro- and fluoro-benzene. However, while most bacteria in natural environments have no ability to degrade these compounds, continued exposure to simple halogenated aromatics encourages genetic exchange between bacterial populations and has been shown to result in the evolution of new degradative pathways [47, 48]. Organics with more extensive substitutions such as the pesticide 1,1,1-trichloro-2,2-bis(*p*-chlorophenyl)ethane (DDT) or polychlorinated biphenyls (PCBs) are more difficult to degrade. Microorganisms that degrade these materials do so through either an oxidative process, in which they use the pollutant as the main substrate for metabolism, or through reductive dehalogenation, in which the halogen groups are replaced with hydrogen [42]. Reductive dechlorination of PCBs appears to be common in contaminated aquatic sediments and typically involves populations of anaerobic microorganisms such as the sulfate reducing bacteria [30, 42]. However, while microbial consortia may process chlorinated organic compounds completely, most of the isolated bacteria that are capable of reductive dehalogenation do not do so completely so that end products of remediation may still contain chlorine groups. A notable exception is *Dehalococcoides ethenogenes* which can reductively dechlorinate the solvent tetrachloroethene to ethene, using perchloroethene as an electron acceptor for metabolism [49, 50]. *D. ethenogenes* is also interesting in that while most bacteria that have been studied from the perspective of pollutant remediation belong to well studied microbial groups such as the Proteobacteria or Firmicutes, *D. ethenogenes* is related to the Chloroflexi [51], a poorly studied group of unusual photosynthetic organisms. This illustrates the importance of considering the possible role of all microorganisms in natural ecosystems for bioremediation, not just those that have been previously studied. It suggests that other poorly studied groups of bacteria may have novel metabolic pathways for pollutant removal that are as yet undiscovered.
# **AQUATIC SYSTEM REMEDIATION OF INORGANIC POLLUTANTS: METALS**
Aquatic systems ranging from lotic (*i.e.*, rivers, streams) to lentic ecosystems (*i.e.*, wetlands, lakes, oxbows) have the ability to transform heavy metal pollutants from the water column through various ecological processes. Plants, sediments and microbes actively participate through biological and chemical processes to transform, remediate, and stabilize toxic metal pollutants. Marshes or constructed treatment wetlands are most often used for phytoremediation of metals [52]. However, the remediation capabilities of the system are not limited to plants. Sediments actively participate in forming metal complexes, reducing certain metal forms, and binding elements to particulate matter. Lentic conditions create ideal circumstances for decreases in soil redox, habitat for aquatic plants, and reducing toxic, soluble valence forms of metals to insoluble, reduced non-toxic forms.
Phytostabilization is the most common form of phytoremediation whereby assimilation and transformation of elements are restricted to the roots and there is no translocation of elements to the shoot. Often these plants are called root accumulators [53]. Plants with a higher concentration of element within the plant tissue than in the surrounding substrate (*i.e.*, water or sediment) are often considered hyperaccumulators and often exhibit luxury uptake. Luxury uptake is the ability to increase elemental concentrations within the plant tissue beyond the needs of the plant for normal metabolic functions [54]. The optimal plant for phytoextraction should not only be able to tolerate and accumulate high levels of heavy metals in its harvestable parts, but also have a rapid growth rate and the potential to produce a high biomass in the field. An ideal plant for rhizodegradation should have rapidly growing roots with the ability to remove toxic metals from solution over extended periods of time [5].
Studies in plant response to heavy metals have suggested that plants have evolved two different physiological mechanisms which enable them to tolerate metal toxicity: accumulators and excluders [55]. Accumulators concentrate sequestered metals in plant parts at low to high concentrations above background concentrations. Excluders have differential uptake and transport between root and shoot which result in constant low shoot/root levels over a wide range of external concentrations. In accumulators, root uptake and transport are more or less in balance, but metals can still accumulate in the roots. Excluders do not generally regulate metal uptake, with restriction of transport from root to shoot as the likely mechanism reducing metal toxicity. Studies have suggested that plants growing on metalliferous soils cannot prevent metal uptake, but can only restrict it and hence accumulate metals in root and shoot tissues at varying concentrations. Different plant species globally have shown considerable differences in their uptake ability for various metal species. Baker [55] highlighted 12 different wetland and upland species that had 18-fold variation for zinc, 240-fold difference for lead, and 273-fold for cadmium.
Phytoremediation of metals has several advantages:
# **Specific Examples of Metal Phytoremediation**
Maine *et al.* [56] identified two strategies of metal remediation depending on the plant species used. Submerged nonrooted *Eichornia crassipes* retained the majority (97%) of metals in macrophytic biomass, while a community cohabitated or completely dominated with *Typha domingensis* had the majority of metals associated with the sediments. This example illustrates the varying degrees of assimilatory capacity between aquatic plants. Water hyacinth (*E. crassipes*) has also been used to phytoremediate iron-rich wastewaters in constructed wetlands [57]. Iron removal by water hyacinth was largely due to the process of rhizofiltration and phytostabilization, since chemical precipitation of iron oxides was followed by flocculation and sedimentation. In this study, phytoremediation seemed to not be very substantial in iron accumulation in comparison to chemical precipitation. Rhizofiltration was the predominant mechanism of remediation of iron since a substantial portion was localized in the roots. Iron phytoextraction was possibly negligible due to the physiological barriers to iron transport to aerial tissues. Caution must be exercised with the use of *E. crassipes,* since it is considered a noxious weed in many countries.
Sharma and Gaur [58] examined the ability of *Lemna polyrhiza* to remediate zinc, lead, and nickel. It was noted that the plant had a rapid increase of metal assimilation within 12 hours, with subsequent assimilation reaching a plateau. It is hypothesized that within the initial 12 hours, rapid, passive uptake of metals occur, while thereafter the assimilation occurs at a slower rate due to metabolic control. A consequence of too great a concentration of heavy metals is the decline and inhibition of chlorophyll synthesis. Thus, most plants have an evolutionary and metabolic constraint to assimilation of certain elements.
Zazo *et al.* [59] examined two species, *Typha latifolia* and *Carex lurida* for their phytoremediation ability in reducing hexavalent chromium. Irrespective of the plant species, as there were no significant differences between species; hexavalent chromium removal was enhanced by plants, with a decrease in soil redox promoted by organic root exudates released by the plants. In low redox conditions, iron and sulfate reduction is increased. Additionally, concentrations of ferrous iron and sulfides increase in the sediment pore water which in turn reduces hexavalent chromium to CrIII. Soils high in organic humic substances will also possess the ability to transform and sequester toxic metals ions. Humic acids constitute a large organic carbon fraction and represent a significant electron donor reservoir for metal reduction and amelioration [59].
Often plants will significantly phytostablize contaminants whereby metals are reduced in and around the roots [60, 61]. The aquatic plant rhizosphere provides a particularly effective, locally oxidized/reduced environment for metal precipitation and adsorption outside the root. *Phragmites australis* roots have been shown to accumulate Fe, Cu, Zn, Pb and Cd, with little to no translocation of metals within the plant to rhizomes and shoots. Iron plaque formation of Fe-oxyhydroxides formed by oxygen evolution by the roots and microbial metabolism is believe to be a mechanism of avoiding toxicity of reduced forms of Fe and Mn to roots under flooded conditions. Vesk *et al.* [62] identified where various element species occurred within the roots of aquatic plants. Iron was often present at highest concentrations at the root surface and decreased within the cell, while trace metals (Cu, Zn, Pb) had highest concentrations occurring within the plant cell, and decreasing towards the root surface. Meyers *et al.* [63] examined the uptake and distribution of lead sequestered by hydroponically grown *Brassica juncea.* The study showed lead uptake was restricted to the root tissue suggested rhizofiltration, where the concentration of lead was always two to three orders of magnitude greater in roots than in shoots. Electron microscopy work revealed substantial and predominantly intracellular uptake at the root tip, while endocytosis of lead within the plasma membrane was not observed. Further experiments demonstrated uptake of lead increase as concentration of lead in solution increased.
In some instances an interaction occurs between metals. Studies have shown [64] that manganese absorption by plant tissue will be suppressed or depressed by high levels of iron precipitate or assimilation. For example *Juncus effusus* showed reduced concentrations of manganese in shoots as a result of high iron concentrations. Thus, phytostabilization of one element could result in deficiencies in other elements important to metabolic functions such as growth.
Plants can accelerate and promote bioremediation of metals and other contaminants by stimulating the growth and metabolism of microorganisms through the release of nutrients and oxygen*.* There is a significant amount of information concerning the influence of aquatic plants on metal fluxes at larger scales. There is also a substantial amount of information concerning small scale laboratory research addressing kinetics of metal uptake in aquatic plants. There is still a very relevant need for research understanding processes of metal accumulation and transformation in the field and how it affects larger scales.
# **Microbial Remediation of Metals and Metalloids**
The microbial remediation of metals differs from that of organic pollutants as metals are not degraded into what are ultimately innocuous products [65]. Rather, interactions between microorganisms and metals may change the redox state of the metal or alter its mobility in the environment. At a basic level, interactions between microorganisms and metal contaminants in aquatic ecosystems can be separated into four broad types: (1) microbial redox transformations that change the metals mobility; (2) volatilization or precipitation from the water column; (3) absorption of metals to microbial cells or cellular products (biosorption); and (4) microbial transformations of other chemicals that indirectly influence metal behavior [66]. Commonly, a number of these processes will be involved in the microbial remediation of metals and metalloids; for example, dissimilatory metal reduction as part of anaerobic respiration (a redox transformation) can result in a metals precipitation or biosorption. Aquatic ecosystems harbor appreciable numbers and diversity of bacteria that metabolize or are resistant to toxic metals and many of these organisms are capable of biotransforming elements into forms of different mobility and toxicity. Fig. (1) illustrates some of the microbial processes that can be involved in transforming metals in oxic and anoxic layers of aquatic environments.
Microbial interactions with arsenic are an example of naturally occurring metal-microbe processes that may have remediation potential. Studies suggest that arsenic resistant bacteria are a common component of both aquatic and terrestrial ecosystems, even those not suffering from arsenic pollution [67, 68]. Bacteria possess a number of genetic and physiological systems for dealing with arsenic toxicity including redox transformations and its incorporation into organic forms [69]. The metabolic process of arsenite (AsIII) oxidation converts arsenic to the less toxic arsenate (AsV) and has been shown to occur in a number of bacteria, either as a resistance mechanism or as a form of energy generating metabolism [70, 71, 72, 73]. The microbial oxidation of arsenite has been proposed as a bioremediation strategy for aquatic environments [70], as the resulting arsenate is much less soluble and can be more easily removed through steps such as alkaline precipitation with lime [74].
While inorganic arsenic becomes less mobile and toxic following oxidation, the opposite is true for other metals. The oxidized forms of chromium and the radioactive metals uranium and technetium are much more water soluble than their reduced forms [65, 75] so that, in contrast to arsenic, it is the microbial reduction of these metals which may be more beneficial to the remediation of aquatic environments. Differences in the mobility of various metals in different redox states also highlight a fundamental difference in remediation strategies for aquatic and terrestrial environments. In solid phase systems (soils) it is typically beneficial to increase metal mobility so they are removed from the system; however, in aquatic phases the opposite approach (to encourage microbial redox processes that decrease metals mobility and increase their precipitation or adsorption) is often more desirable [75].
The reduction of chromium from the toxic and highly soluble CrVI to relatively insoluble CrIII has been demonstrated in a wide range of microorganisms including bacteria, fungi, and algae [76, 77]. Chromate reducing enzyme systems vary greatly between bacteria and include both soluble enzymes, those associated with the cell membrane, and enzymes capable of reducing CrVI either aerobically or anaerobically [78, 79]. Such diversity suggests that bacterial chromate reductases may be useful in the bioremediation of chromium contaminated sites in a wide range of environments [80]. As is the case with arsenic, chromium resistant bacteria appear to be ubiquitous, having been recovered from both chromium impacted and non-polluted environments [81].
Toxic metals can also be removed from aquatic environments *via* their precipitation with the products of other microbial redox processes. Dissimilatory sulfate reducing bacteria, a common group of bacteria in aquatic sediments that utilize sulfate as the terminal electron acceptor in anaerobic respiration, produce hydrogen sulfide as a waste product. The sulfide produced can precipitate and immobilize arsenic [82] and abiotically reduce CrVI to CrIII [83]. Sulfate reducing bacteria have also been shown to reduce aqueous concentrations of cadmium, copper, iron, nickel and zinc through the formation of insoluble metal sulfides in model systems [84, 85] and similar processes are possible in natural environments. Dissimilatory iron-reducing bacteria convert Fe3+ to Fe2+ during their anaerobic respiration, and Fe2+ can also abiotically reduce and immobilize toxic metals such as chromium [83]. Some iron-reducing bacteria such as species of *Geobacter* and *Shewanella* are also capable of directly reducing other metals as part of their metabolism, and may be useful in the reductive remediation of technetium, cobalt, and uranium [86, 87, 88].
While sulfate and iron-reducing bacteria are typically found in anoxic aquatic sediments, they can also be important components of aquatic biofilms [89, 90]. Biofilms are naturally occurring communities of attached microorganisms found on any submerged surface, and they are characterized by a complex architectural structure often with both aerobic and anaerobic layers [91, 92]. They often support a diverse range of microbial populations in close proximity to each other, which can be important in bioremediation [93]. Furthermore, the cells within biofilms are enclosed within a matrix of extracellular polysaccharides or slime, which itself can remove metal contaminants from the surrounding water through the process of biosorption [94, 95, 96]. The combination of sulfate reduction and potential biosorption/precipitation of metal sulfides within the biofilm structure can be a particular effective method of remediation and has been shown to be effective for metals such as chromium, copper, and lead [97, 98, 99].
**Figure 1.** Microbial processes that can remediate metal pollution in aquatic ecosystems include: (a) the aerobic oxidation of metals such as arsenic which can reduce their mobility and toxicity; (b) the absorption of metals such as chromium and copper to biofilms associated with sediments and aquatic plants; (c) the microbial reduction of metals such as cobalt and uranium to less toxic forms by iron-reducing bacteria such as *Shewanella*; (d) indirect transformations resulting from microbial metabolism that result in metal precipitation, such as the reduction of chromium following the production of hydrogen sulfide by sulfate-reducing bacteria in anoxic sediments.
Sulfate-reducing bacteria in anoxic sediments and biofilms are important mediators of mercury methylation [100, 101, 102], which can have negative impacts on human activities because methylmercury is highly toxic and subject to biomagnification through aquatic food webs [103]. However, sulfate-bacteria may also play some role in the removal of aqueous mercury (Hg2+) *via* the production of hydrogen sulfide as a waste product which can react with Hg2+ to form the much less soluble mercuric sulfide [104]. Of more importance from a remediation aspect is the enzymatic reduction of Hg2+ to elemental mercury (Hg0 ) which is common and widespread throughout bacteria [105, 106]. Elemental mercury is insoluble and much less toxic than other forms. It is also volatile so that the microbial reduction of Hg2+ to Hg0 is a significant mechanism that can contribute to the removal of mercury from natural waters to the atmosphere [106]. Mercury contaminated environments select for microorganisms capable of carrying out this transformation [107], which is encoded for by a number of mercury resistance (*mer*) genes [105, 106].
Bacterial mercury resistance genes are usually located on plasmids and are often components of transposons [105, 106, 108, 109]. These mobile genetic elements can be passed between bacterial species *via* the process of horizontal gene transfer and the evolutionary history of *mer* genes suggest that this has been a relatively frequent occurrence in the past [110]. The same phenomenon has been shown for arsenic resistance genes, which are also often borne on plasmids [111].
Mobile resistance genes demonstrate the capability of natural microbial communities to respond and adapt to environmental pollution both from metals and organic pollutants [112]. From an applied perspective they present an excellent opportunity to incorporate biotechnology into bioremediation in that the genes can be transferred into specific bacterial species that may be suitable for a particular environment. Such an approach is likely to be particularly beneficial in environments where there are multiple contaminants, in that traits such as metal resistance may be passed onto to organic-degrading bacterial populations [23].
# **AQUATIC SYSTEM REMEDIATION OF INORGANIC POLLUTANTS: NUTRIENTS**
Unlike previously described organic pollutants and metals, nutrients are a vital component in aquatic systems. Productivity and trophic status of aquatic systems impart information on not only the stability, but also the relative ecological health of aquatic systems. Problems arise however, when nutrients in aquatic systems reach levels in excess of the natural system's capacity to utilize them. Nutrient concentrations must strike a fine balance between the needs of the aquatic system and excessive levels which will lead to ecological problems such as hypoxia or harmful algal blooms.
Several studies have examined abilities of aquatic plants to remediate excessive nutrient concentrations. Cronk and Fennessy [113] warn that nitrogen and phosphorus removal from water by vegetation is not the major pathway for nutrient remediation where concentrations are high. More success can be achieved with plants in a nutrient phytoremediation scenario when overall nitrogen and phosphorus concentrations have lower loads. Nutrient uptake by plants is also dependent on several factors including season, plant growth rate, plant biomass, and latitude [113]. In a two year study in northwest Mississippi, USA, Kröger *et al.* [114, 115] reported that vegetated drainage ditches reduced 53% and 43% of the dissolved inorganic nitrogen and maximum inorganic effluent phosphorus loads, respectively. Mesocosm scale studies reported 83 ± 3% and 40 ± 8% decrease in aqueous ammonia and nitrate concentrations, respectively, in systems vegetated with *Ludwigia peploides* [116]. Although it was least effective in decreasing ammonia and nitrate concentrations, the aquatic plant *Leersia oryzoides* was more effective than *L. peploides* at removing organophosphorus (29 ± 7%) [116]. Two separate studies examined the use of *Eichhornia crassipes* in remediating excessive nutrients from water. In a system with a 21 day hydraulic retention time, 100% removal of total nitrogen and phosphorus was achieved after nine weeks of treatment through *E. crassipes* [117]. Using a 31-day batch growth experiment with *E. crassipes*, reductions of total Kjeldahl nitrogen, ammonium, and total phosphorus were 92%, 99%, and 99%, respectively [118].
# **CONCLUSIONS**
Aquatic systems are resilient habitats which receive many point and non-point-source pollutants. Rather than focus on their contamination, this chapter was devoted to the abundance of literature demonstrating remediation capabilities – primarily phytoremediation – of these aquatic systems. Although the literature presented within this chapter is not an exhaustive review of all the research conducted in these specific areas, it provides a solid foundation for those interested in the ability of vegetation to clean waters receiving pollutants. Cautions exist, of course, when using phytoremediation for any pollutant. While benefits certainly exist, there are also drawbacks to using phytoremediation tools. For example, harvested biomass from metal phytoextraction may be a hazardous waste. Improper initial planning may lead to a potential food chain effect due to consumption of contaminated plants. Just as a mechanic cannot fix every problem with a wrench, phytoremediation should be considered as a valuable tool in practitioners' environmental toolbox. Remediation of aquatic systems is equivalent to an "ecological tagteam" of physical, chemical, and biological processes conducted in plants, sediment, and water.
# **REFERENCES**
© 2011 The Author(s). Published by Bentham Science Publisher. This is an open access chapter published under CC BY 4.0 https://creativecommons.org/licenses/by/4.0/legalcode
# **Concluding Remarks**
#### **Francisco Sánchez-Bayo**<sup>1</sup> **, Paul J. van den Brink**2,3 **and Reinier M. Mann**<sup>1</sup>
*1 Centre for Ecotoxicology, University of Technology Sydney, Australia; <sup>2</sup> Alterra - Wageningen University and Research Centre, Wageningen, The Netherlands and <sup>3</sup> Department of Aquatic Ecology and Water Quality Management, Wageningen University, Wageningen, The Netherlands*
# **THE EXTENT OF CONTAMINATION**
The new millennium started with a legacy of unprecedented contamination of the world ecosystems left in the wake of the various activities of humankind. Chemical pollutants have become so diverse (see Chapter 1) and widespread that there is hardly any region of the world that is not currently affected by their impacts. With the exception, perhaps, of the desert wilderness areas (for which information on pollution is still lacking), every other ecosystem on earth, from the polar regions to the tropics, whether on land or in the oceans, has been shown to contain residues or traces of organic and inorganic pollutants of anthropogenic origin.
Even if most of the contaminants originate in industrial, developed countries located mainly in temperate regions, it has become obvious that natural transport processes have carried many pollutants far from their sources to places like mountain tops, polar caps and remote islands in the oceans. Those transport routes have been thoroughly examined in Chapter 2, which shows the enormous progress achieved to date in this field of environmental research. As a consequence, current models on fate and transport of contaminants are becoming more accurate, and no doubt will minimise in future the expensive analytical monitoring that was once required to understand the movement and accumulation of chemical pollutants in ecosystems. However, monitoring will still remain necessary to determine whether the remedial actions taken as a result of risk assessments and regulation are effective. This is particularly true of recent contaminant classes (e.g., fluorinated surfactants, pharmaceuticals) as little is known about these in terms of potential ecosystem level effects (see Chapter 7). Therefore, the use and development of monitoring devices such as passive samplers will be crucial in future.
As our knowledge of pollution increases, so also do the efforts to eliminate the most toxic pollutants. Encouraging examples of bioremediation by aquatic plants and micro-organisms are already available for some pesticides and metallic pollutants, as explained in Chapter 11. The natural capacity of aquatic ecosystems to eliminate most kinds of chemical pollutants is already being fostered by natural selection, because of the increasing pressure that such ecosystems experience in our heavily contaminated world. Indeed, nature always moves towards a point of equilibrium, thereby mitigating the impacts caused by pollution. Phytoremediation systems harness the adaptivity of natural systems and are already being used to sequester metal contaminants in many places, and in future could become essential management tools to eliminate pesticide residues in agricultural regions once long-term maintenance issues are resolved adequately. However, before they become widely accepted we need first to investigate the microbe-plant interactions in order to enhance the capabilities of the system, and also to explore the possibility of phytoremediation of 'emerging' organic contaminants such as new pesticides and even persistent organic compounds like dioxins, perfluorinated compounds, *etc.* The recalcitrant nature of the latter chemicals is certainly a problem that may require the combined action of both natural bioremediating systems and artificial remediation initiatives (e.g. incineration at high temperatures).
# **Impacts on Terrestrial Ecosystems**
Throughout human history, metals and metalloids have accumulated within soils and surface waters of the earth, sometimes as a consequence of extraction from otherwise stable ore-bodies, redistribution of materials rich in metals (e.g., guano) or through combustion of fossil fuels. Our understanding of the ecological risks associated with increased environmental deposition of metals has improved as a consequence of the study of heavily polluted environments, where various elements such as Cd, Cu, Zn, Se, As and Hg have been deposited in high concentrations with severe and obvious detrimental effects on flora and fauna. However, we are only now beginning to appreciate the subtle and slow shifts in ecosystem function and dynamics that may occur through the movement of some elements through the food chains. Those flora and fauna that have been detrimentally affected through the bioaccumulation and biomagnification of elements such as Cd or MeHg, can only be detected by careful examination of their trophic movement through ecosystems, and by experimental manipulations. While the evidence to date has been summarised in Chapter 3, it is clear that more carefully planned studies in a wider selection of taxa (e.g., pelagic and coastal fish and invertebrates, insectivorous and predatory birds, and marine and terrestrial reptiles) are needed if we want to predict the overall consequences of increasing pollution by metallic and metalloid elements in ecosystems.
Among the most common organic pollutants are pesticides, which are routinely applied to the majority of the vast agricultural landscapes of the world as well as to many forested areas. Within the immense literature available on the subject, only effects at the community and ecosystem levels have been considered in this book. The known ecosystem impacts of pesticides have been studied mostly within the agricultural fields and pastureland with livestock, a summary of which is presented in Chapter 4. Our knowledge of pesticide impacts outside of those fields where pesticides are applied (the so-called off-crop areas) is scarce and in some cases anecdotal, which explains why the assessment of other terrestrial ecosystems surrounding the agricultural land, such as scrub or shrub landscapes and deserts, could not be considered in this book. This is precisely one of the areas where future studies on this topic should focus: the comparative assessment of impacts either of individual pesticides or their mixtures between the crop areas and the surrounding ecosystems (e.g., wetlands, scrubland, adjacent forests, etc). To a certain extent this has already been accomplished for bird communities in England and the USA, but much more work needs to be done in other countries (e.g., tropical ecosystems) to obtain a full picture of the effects at ecosystem level.
In parallel to scenarios in the agricultural field, Chapter 5 examines in detail four case studies of pesticides currently in use for forest management - two insecticides and two herbicides. Based on that evidence we are beginning to understand now the complex issues involved in these practices, but we are still unable to identify critical thresholds at which the system becomes unsustainable. Obviously, the damage that one single pest can cause in a monospecific boreal or temperate forest should not be measured only in economic terms, as it really affects entire animal and plant communities that depend on that forest. Here is where a balance must be struck between the need to control such pest, thus protecting those communities, and the ecological side-effects derived from the application of chemicals for controlling the pest. Thus, the need for identifying mitigative strategies under an adaptive management regime and their comparison with the 'do-nothing' approach. Since most of our knowledge on forest impacts comes from studies carried out in North America and Europe, it would be desirable to compare such impacts with those in other parts of the world, e.g., the more diverse forests of Japan, India or Australia.
# **Impacts on Aquatic Ecosystems**
Pesticides impacts on freshwater communities are quite well understood and have been studied more widely than in any other ecosystem (see Chapter 6). Under normal circumstances, most pesticide residues in soil move eventually into water bodies found in the landscape. However, studies on this topic are biased towards fish and macro-invertebrates, whereas field studies about the effects of pesticides on macrophytes and microorganisms are scarce. Moreover, fungicides have received very little study [1], even though they are routinely applied to agricultural systems [2]. Of course, the effects of mixtures and multiple stressors in aquatic systems are still an important topic of research which has not been developed sufficiently. One area that requires more attention in future investigations is the effect of pesticides on ecosystem function as well as the services provided by the aquatic systems, whose potential losses have not been quantified yet. In connection with this objective, studies on ecosystem-based traits have already been proposed [3]. Another area that deserves more attention is recovery from pesticide exposure. It is not known which processes weigh more in the recovery of communities in freshwaters (recolonisation, compensatory effects of nutrients, higher temperatures?), and how widespread is the occurrence of long-term effects in the field.
A common problem encountered when evaluating the impacts of toxic pollutants in ecological risk assessments, whether they are pesticides, pharmaceuticals or any other chemical, is that predictions are based on rough extrapolations from effects at low levels of organization (e.g., individual or population effects measured by EC50s or NOECs). At the same time, measurements such as NOEC are unreliable and should be replaced by more accurate alternatives such as the no-effect concentration (NEC) [4,5]. Direct testing of individual chemicals or mixtures of chemicals using micro- and mesocoms is one effective way to overcome our current deficient approaches, and thus reduce our dependence on extrapolating results from lower levels of biological organization for both risk assessment and regulation. This is particularly relevant in the case of the emergent pollutants of concern (see Chapter 7), for
## **240** *Ecological Impacts of Toxic Chemicals Sánchez-Bayo et al.*
which very few ecotoxicological data are available in the first place. It follows from this that we need to develop improved empirical and experimental methodologies for evaluating field-based observations (and mesocosms) of anthropogenic stressors at higher levels of biological organization in both terrestrial and aquatic systems. Only then we will be able to determine the real impacts at community and ecosystem levels without flawed extrapolations [6], and improve our risk assessment modelling [7].
If freshwater ecosystems are the recipient of the first load of residues emanating from industrial processes, urban wastes or agricultural practices, eventually all pollutants enter the oceans. Those which are persistent or bioaccumulate in organisms will almost inevitably end up at the top of the marine food chain, and therefore will impact our fisheries. From the start, coastal ecosystems receive the bulk of the pollution *via* the discharges of rivers into estuaries and adjacent coasts. Marine currents spread these contaminants not only between the pelagic and benthic communities but also up and down the coastline, reaching coral reefs and finally the open ocean, where pelagic ecosystems represent the ultimate frontier. The complexity of these interactions in regard to chemical pollution is well described in Chapter 8, where the effects of ecotoxicants are analysed from the cellular and tissue level to the community level in progressive steps, culminating in the integrative assessment of overall impacts using various methods. The lack of information on how individual responses are transferred to ecosystem level is the major hurdle in the latter assessments: although bioassays have improved our understanding of the effects of pollutants, it is difficult to extrapolate their results to the ecosystem, as they lack ecological 'realism' [8]. As indicated above, this calls for new approaches to evaluate the effects at the highest level of organization, as well as for agreed methods for integrative assessment.
A similar approach was followed in Chapter 9 to describe the effects of a wide range of pollutants in coral reefs, an ecosystem often described as the 'rainforest of the oceans' due to its high biodiversity. Here, more than in any other ecosystem, is where our ecotoxicological knowledge is still lacking considerably. Since reefs are often exposed to multiple stressors simultaneously and for prolonged periods there is a need for information on the effects of chronic exposure to pollutants and stressor interactions (e.g., climate change factors) on a broad range of species and life history stages. In particular, there is concern that prolonged exposure to herbicide mixtures may trigger a cascade of unknown effects in the delicate and complex food web associated with the corals [9]. At the same time, more effort needs to be directed towards characterisation of the reef system's potential to recover from and adapt to ever increasing anthropogenic contaminants in a globally changing environment.
Finally, because pelagic processes (see Chapter 10) are important to both the surface layer (plankton) and benthic communities, further research in the ecotoxicology of these ecosystems should be directed towards integrating, not dividing, our understanding of the effects of pollutants in the different environmental compartments. There is a need for more research on how contaminants affect both primary producers and microbial loop components. Current knowledge is limited to effects on single algal species, and there is virtually no knowledge of impacts in more complex systems that include bacteria and protists. Although there is a better understanding of how contaminants affect some zooplankton species (e.g., calanoid copepods), there is an urgent need for more research on mesozooplankton species, including metamorphosing stages of numerous invertebrate taxa.
The chapters in this book provide a comprehensive review of the known ecological impacts of a wide variety of pollutants in all major ecosystems of the planet at the present time. Ignorance about the world is indeed the greatest enemy of humankind, so by bringing these facts to the public we hope will encourage the new generations to take more care in managing our activities. Only by knowing where we stand can we start thinking about solutions to minimise the release of contaminants and device capable remediation systems that may reduce or eliminate completely the major chemical threats to the environment. It is from this perspective that this book was conceived in the first place.
# **ACKNOWLEDGEMENTS**
The editors would like to acknowledge the contribution of all authors of this book in preparing this concluding chapter.
# **REFERENCES**
[1] Maltby L, Brock TCM, Brink PJvd. Fungicide risk assessment for aquatic ecosystems: importance of interspecific variation, toxic mode of action and exposure regime. Environ Sci Technol 2009; 43:7556-7563.
© 2011 The Author(s). Published by Bentham Science Publisher. This is an open access chapter published under CC BY 4.0 https://creativecommons.org/licenses/by/4.0/legalcode
# **APPENDIX**
# **ADDITIONAL REFERENCES TO CHAPTER 4**
Included here are a number of references consulted to write this chapter, but that have not been cited in the text for reasons of insufficient space.
# **Introduction**
# **Pesticides in Agriculture**
# **Exposure of Organisms to Agricultural Pesticides**
**Francisco Sánchez-Bayo, Paul J. van den Brink and Reinier M. Mann (Eds) © 2011 The Author(s). Published by Bentham Science Publishers**
# **Review of Pesticide Impacts on Non-Target Communities**
# *Soil Communities*
# **244** *Ecological Impacts of Toxic Chemicals Francisco Sánchez-Bayo*
# *Vegetation and its Arthropod Communities*
### **246** *Ecological Impacts of Toxic Chemicals Francisco Sánchez-Bayo*
# **Vertebrates**
## **248** *Ecological Impacts of Toxic Chemicals Francisco Sánchez-Bayo*
© 2011 The Author(s). Published by Bentham Science Publisher. This is an open access chapter published under CC BY 4.0 https://creativecommons.org/licenses/by/4.0/legalcode
# **Index**
1,1,1-trichloroethane, 36, 37 17--trenbolone, 151 2,3,7,8-tetrachlorodibenzo-p-dioxin, 140 2,4,5-T, 140 2,4-D, 8, 68, 70, 74, 75, 195 2-methoxyethylmercuric chloride, 201 2-nitrophenol, 228 4-nitrophenol, 228 7-ethoxyresorufin-O-deetilase, 168
# **A**
absorption, 17, 19, 20, 21, 26, 33, 47, 48, 230, 231, 232 acaricides, 64 acids, 6, 8, 15 amino, 5, 8, 71, 94, 187 fatty, 8 humic, 99, 230 nucleic, 170 organic, 197 actinomycetes, 70, 228 active ingredient, 65, 66, 90, 97, 101 adaptation, 26, 203 adhesives, 148 adipose tissue, 140 adsorption, 14, 20, 21, 95, 102, 111, 147, 148, 230, 231 aerosols, 7, 18, 34 air conditioning, 3 albatross, 217 alchemy, 3 aldicarb, 65, 69, 72 aldrin, 66, 67, 71, 72 algae benthic, 115, 196 blue-green, 4, 5, 146 coralline, 187, 188, 194 epiphytic, 47 filamentous, 128 freshwater, 146, 149, 194 green, 116, 120 macroalgae, 175, 179 microalgae, 151, 187, 193, 194 symbiotic, 187, 194 algicides, 64 alkanes, 23 alkylphenol ethoxylates, 152 alligators, 77 alloys, 3 amelioration, 3, 230 aminomethylphosphonic acid, 94 ammonia, 178, 233
amoeba, 71 amphibians, 66, 76, 77, 78, 80, 96, 97, 98, 102, 112, 115, 116, 121, 146, 148, 149 analysis chemical, 167, 171, 176 meta-, 120, 174 molecular, 191 regression, 49, 175 risk, 90, 104 anemones, 199 anilines, 8 annelids. *See* worms anthracene, 8, 146 antibacterial agents. *See* biocides antibiotics, 64, 150, 151 anticoagulants, 64 antifouling paints, 169, 193, 201, 215 antimony, 7 antioxidants, 169 ants, 74, 75, 76 anuran species. *See* frogs aphids, 47, 48, 75, 79 apoptosis, 145 applications aerial, 95, 97, 102 herbicide, 74, 92, 95, 96 insecticide, 75, 77, 99 pesticide, 10, 68, 77, 89, 93, 112, 192 silvicultural, 97 aquaculture, 151, 166, 217 aquatic plants, 66, 102, 147, 226, 229, 230, 232, 233, 238 aragonite, 199 Aroclor, 140 arsenate calcium, 75, 76 lead, 75 arsenates, 52, 67, 69, 73, 231 arsenic, 7, 43, 50, 52, 53, 63, 64, 69, 70, 72, 199, 231, 232 arthropods canopy, 102 parasite, 75 predatory, 72, 74, 75 saprophytic, 72 terrestrial, 64, 72, 73, 102 assimilation, 27, 47, 48, 50, 51, 55, 147, 168, 229, 230 atmosphere, 7, 18, 19, 38, 189, 232 atrazine, 8, 68, 70, 72, 73, 77, 79, 117, 193, 215, 218, 227 atrophy, 169 azadirachtin, 99 azinphos-methyl, 77
# **B**
*Bacillus thuringiensis*, 4, 71, 89, 91, 99, 194 bacteria, 5, 8, 70, 119, 145, 151, 153, 170, 172, 214, 218, 219, 225, 228, 231, 240
aerobic, 27 anaerobic, 27 cyanobacteria, 141 denitrifying, 27, 71 iron-reducing, 231 methane-producing, 27 methanotrophic, 70 nitrifying, 70 psychrophilic, 28 resistant, 231 soil, 71 sulfate-reducing, 27, 227, 231, 232 badgers, 54 baits, 66, 67 barium, 50 barnacle, 171 bass, 102 bats, 78 batteries, 3, 5 Bayesian statistics, 124 bees, 74, 76 bumblebees, 76 honeybees, 66 beetles, 99 carabid, 75, 96 dung, 72 ladybird, 48, 74, 75 leaf-beetles, 79, 103 rove, 74 soil-dwelling, 66, 74, 75 staphylinid, 72, 75 weevils, 75, 79 bendiocarb, 65 benomyl, 8, 70, 72, 73 benzene, 8, 228 benzimidazoles, 6 benzo[a]pyrene, 146, 228 BHC, 75 bifenthrin, 226 bioaccumulation, 7, 10, 43, 44, 47, 48, 55, 76, 104, 140, 145, 147, 165, 170, 172, 196, 198, 213, 216 EDCs, 216 metals, 45, 46, 50, 53, 54, 216, 238 PCBs and PCDDs/PCDFs, 128, 141, 142 risk, 55 bioavailability, 29, 43, 46, 51, 67, 69, 92, 141, 145, 147, 152, 170, 193, 213, 214, 217 biocides, 8, 64, 72, 192, 193, 199, 215 biodiversity, 72, 74, 75, 76, 79, 80, 92, 94, 166, 174, 240 biofilms, 231, 232 bioindicators, 173, 174 biomagnification, 44, 46, 47, 48, 50, 51, 52, 54, 55, 56, 112, 114, 140, 172, 176, 213, 216, 232 metals, 45, 46, 47, 54, 55, 238 biomarkers, 145, 166, 167, 168, 176, 190, 191, 216 biomass, 7, 27, 74, 103, 151, 173, 174, 212, 215 algal, 102, 112, 146
fungal, 73 phytoplankton, 117, 215 plant, 47, 74, 75, 94, 229, 233 biomonitoring, 50, 101, 116, 125, 128, 168 bioremediation, 228, 229, 230, 231, 233, 238 biosorption, 231 biosphere, 26, 70, 80 biosynthesis, 5, 6, 8, 71, 147 bioturbation, 22 birds, 8, 10, 46, 47, 54, 55, 63, 65, 80, 99, 143, 145, 149, 239 aquatic, 52, 54, 114 fish-eating, 142, 143 galliform, 51 granivorous, 66, 67, 78, 79 insectivorous, 78, 100 passerine, 53 predatory, 52, 55, 64, 67 seabirds, 215, 217 songbirds, 103 waders, 52, 53 waterfowl, 51, 52, 54, 66 birds of prey. *See* raptors bisphenol A, 139, 151, 169 bladder, 52 blood, 7, 8, 29, 149 clotting, 8 residues in, 48, 53 samples, 78 stream, 7, 48, 66 bluegill, 102, 118, 125 body burdens, 44, 45, 46, 48, 49, 53, 54, 55, 144, 176 fat, 9, 103, 217 residues, 98 size, 50, 64 surface, 47 weight, 48 bone, 50 borer ash, 99 stem, 75 sugarcane, 74 botulin, 4 brain, 7, 8, 55, 76, 77 breast milk, 140 breeding, 66, 95, 101 delay in, 153 failure, 77, 79 grounds, 53, 54, 98 season, 79 brevetoxins, 4 brodifacoum, 79 bromacil, 79 bromadiolone, 79
bromine, 140 bromoxynil, 8 bryophytes, 96 budworm, 89, 91, 99 buffer zones, 65, 98, 226 bushfires, 5, 6, 140, 170 butachlor, 73 butane, 8 buzzards, 67
# **C**
cabbage, 72 cadmium, 7, 34, 36, 37, 38, 39, 43, 44, 52, 53, 171, 199, 201, 212, 226, 229, 231 caesium, 50 caimans, 77 calcium, 7, 75, 76, 78, 188, 202 cancer, 7, 52, 145 captan, 70, 73 carbaryl, 8, 70, 194, 195 carbendazim, 70, 72, 73 carbofuran, 65, 70, 71, 72, 73 carbon dioxide, 6, 23, 26, 27, 227 carbon monoxide, 6 carboxamides, 5 carcasses, 51, 79 carcinogens, 146 carnivores, 46, 78, 95 caterpillars, 67 cattails, 79, 226 cattle, 48, 77, 151, 166 cattle ticks, 66 cell division, 8 growth, 97, 200 membranes, 170, 231 metabolism, 194 vacuoles, 51 centipedes, 72 chemicals anthropogenic, 28 antifouling, 189, 215 elemental, 43 endocrine disrupting (EDCs), 8, 169, 216 hydrophilic, 20, 226 hydrophobic, 15, 16 industrial, 6, 64 organic, 15, 25, 26, 34, 117, 139, 154, 168, 226 persistent, 76, 80, 143, 189, 193 recalcitrant, 78 synthetic, 27, 28 toxic, 4, 5 volatile, 18, 20 chemotaxis, 228
chickweeds, 79 Chironomidae. *See* midges chlordane, 67, 71, 73, 192, 194 chlorfenapyr, 64, 76 chlorine, 140, 141, 228 chloroacetamides, 6, 8 chlorodibenzofurans (PCDF), 6 chlorofluorocarbons (CFCs), 3 chlorophyll *a*, 119, 146 chloroplasts, 8, 193 chlorosis, 194 chlorothalonil, 70, 71 chlorpropham, 73, 75 chlorpyrifos, 8, 70, 119, 193, 194, 195, 202, 203, 204, 227 chlorsulfuron, 8 chromium, 6, 7, 50, 226, 230, 231, 232 ciliates, 71 ciprofloxacin, 151 Cladocera. *See* waterfleas cladocerans, 102, 147 clams, 169 clay minerals, 28 particles, 70 climate change, 111, 188, 202, 204, 240 clotrimazole, 215 cnidarian, 141 coal, 5, 7, 55 coated seeds. *See* seed-dressings cobalt, 6, 231, 232 cod, 216 coefficient mass-transfer partition, 19, 21 molar absorption, 26 octanol-air partition, 16 octanol-water partition, 14, 34, 225 partition, 14, 15 solids-water partition, 15 Coleoptera. *See* beetles Collembola. *See* springtails commensalism, 104, 117 common salt, 9 communications, 3 communities algal, 146, 215 animal, 71, 102 aquatic, 66 arthropod, 64, 73, 74, 75 benthic, 145, 176, 240 bird, 239 coastal, 187 coral reef, 192 estuarine, 174 infaunal, 175
insect, 78 invertebrate, 80, 113, 121 lotic, 126 macrobenthic, 174 macroinvertebrate, 141, 147 mammal, 54 mesocosm, 126 microbial, 28, 113, 141, 146, 151, 152, 232 non-target, 111 phytobenthos, 215 phytoplankton, 141, 215 plankton, 147, 214, 215 plant, 52, 68, 74, 93, 96, 239 protozoan, 71 reptile, 93 soil, 70 zooplankton, 149, 151 competition, 74, 78, 92, 94, 104, 112, 117, 119, 124, 147, 173 compounds aliphatic, 25 aromatic, 25 arsenical, 52 dioxin-like, 142, 143, 189 genotoxic, 169 halogenated, 25, 140, 143 immunotoxic, 217 inorganic, 7, 34 ionisable, 34 mercurial, 7, 64 neurotoxic, 76 nitroaromatic, 228 non-polar, 9, 226 oil, 197 organic, 8, 16, 24, 168, 174, 189, 227, 238 organochlorine, 140, 193, 217, 227, 228 organometallic, 5, 8, 189, 200 organophosphorous, 168 organoselenium, 54 organotin, 7 perfluorinated, 5, 238 phenolic, 5 synthetic, 3, 6, 215 xenobiotic, 168 concentration effective, 151, 203, 213 environmental, 33, 101, 203 gradient, 17 lowest observed effect (LOEC), 149 median lethal (LC50), 9 no-effect (NEC), 151, 239 no-observed effect (NOEC), 124 peak, 92, 112 predicted environmental (PEC), 10, 34 threshold, 123, 124, 172
conjugation, 168, 226 conservation, 29, 53, 54, 74, 80 consortia, 146, 228 coolants, 5 coot, 79 Copepoda, 121 copepods, 102, 147, 153, 171, 219, 240 copper, 7, 43, 44, 48 coral bleaching, 191, 194, 195, 198, 200, 202 corals, 166, 240 branching, 200 calcifying, 187 cup, 198 hard, 190, 194 massive, 190 reefs, 188, 190, 196 scleractinian, 187, 188, 192, 194, 202 cormorants, 50, 78, 143 corn bunting, 79 corrosive, 7 cosmetics, 3, 148, 152 coumarin, 79 cowpats, 73 crayfish, 102, 141, 142 creosote, 147 crickets, 54 crop barley, 66 Bt-canola, 75 Bt-corn, 75 Bt-cotton, 74, 75 canola, 68 cereal, 63, 79 corn, 72, 76 cotton, 68 damage, 67, 239 horticultural, 75 losses, 64 pests, 72 production, 89 protection, 74 rice, 52, 71, 73, 75, 76, 77, 79, 226 soybean, 65 sustainability, 80 transgenic herbicide-tolerant (TGHT), 74 yields, 63, 73, 74, 80, 192 crude oil. *See* petroleum crustaceans, 102, 113, 116, 166 cutgrass, 226 cyanide, 189, 191 cyanotoxins, 4 cypermethrin, 8, 66, 115 cyromazine, 66 cytochrome C, 8
cytochrome P450, 7, 8, 68, 168 cytotoxic, 145
# **D**
dab, 213, 216 dalapon-sodium, 72 DDE, 9, 67, 68, 77, 78 DDT, 5, 8, 9, 64, 65, 66, 67, 70, 71, 72, 75, 77, 78, 80, 88, 111, 114, 143, 145, 166, 190, 192, 194, 228 decay, 68, 73 dechlorination, 141, 228 decomposers, 118 decomposition, 70, 101, 111, 115, 118, 119 deer, 67, 77, 94 deforestation, 80 deformities, 143 congenital, 143 embryo, 143 Great Lakes Embryo Mortality, Edema, and Deformities Syndrome (GLEMEDS), 143 degradation, 7, 92, 102, 104, 189 aerobic, 141, 228 anaerobic, 141 chemical, 13, 27, 34, 92 co-metabolism, 27, 28 metabolic, 152 microbial, 26, 27, 68, 92, 94, 99, 228 photo chemical, 23, 26, 111 primary, 26, 94, 97 dehalogenation, 227, 228 deposition, 17, 18, 19, 20, 21, 33, 38, 43, 51, 66, 90, 101, 102, 111, 168, 175, 189, 190, 196, 202, 203, 213, 238 desalination plants, 189 desorption, 20, 21, 29, 33, 96, 147 detergents, 6, 152 detoxification, 7, 8, 44, 48, 49, 69, 76, 168, 198 detritivores, 73, 146, 175 development, 3, 4, 8, 9, 54, 63, 73, 77, 90, 99, 104, 113, 123, 126, 138, 141, 142, 144, 145, 150, 151, 152, 153, 167, 169, 171, 172, 173, 178, 188, 198, 200, 201, 215, 216, 218, 238 diatoms, 141, 194 diazinon, 65, 227 dichlobenil, 8 dieldrin, 36, 37, 38, 39, 66, 67, 69, 71, 78, 114, 192, 194 diethylstilbestrol, 152 difenoconazole, 8 diffusion, 17, 19, 20, 22, 27, 47, 201, 213 diflubenzuron, 99, 101, 102, 103, 104 dihydroxybenzene, 228 dilution, 17, 49, 112 dimethoate, 66, 76, 77 dinitroanilines, 6, 8 dinoflagellates, 4, 5, 187 dioxygenases, 228 Diptera. *See* flies diquat, 8, 116, 125 discharges, 4, 5, 6, 119, 174, 198, 199, 213, 217, 240
diseases blue sac, 142, 143 fungal, 63 infectious, 166, 217 *itai-itai*, 4, 52 Minamata, 4 plant, 3, 64 disinfectants, 5 dispersants, 197 dissipation, 68, 91, 92, 94, 96, 98, 102, 152, 193 disulfoton, 65 diuron, 8, 67, 70, 79, 116, 193, 194, 195, 196, 202, 203, 204, 215 DNOC, 72, 73 dolphins, 217 dopamine, 7 dose, 9, 64, 90, 112 effective, 48 lethal, 66, 114, 143 median lethal (LD50), 9 response relationship, 98, 100, 172, 198 sublethal, 69 dredging, 119, 166, 213 drift invertebrate, 95, 101, 126 mitigation, 65 spray, 65, 67, 68, 76, 77, 78, 80, 90, 111, 112, 125 drugs, 3, 4, 6, 8 ducks, 51, 52, 79, 149 duckweed, 116 dyes, 3, 6 dysfunction, 52, 153
# **E**
eagles, 67, 77, 78, 140, 149 earwigs, 72 echinoderms, 172 ecosystem function, 143, 218, 238, 239 ecosystems agroecosystems, 71, 75 aquatic, 6, 46, 65, 67, 68, 69, 102, 111, 114, 115, 138, 139, 140, 151, 152, 173, 193, 227, 228, 230, 238 artificial, 113, 117, 119, 120, 124, 125 benthic, 217, 240 coastal, 166, 215, 217, 218 coral reef, 188, 190, 202 forest, 91, 93, 94, 96, 97, 103, 104 freshwater, 111, 119, 120, 126, 128, 139, 144, 150, 151, 154, 212, 240 lakes, 142 lentic, 98, 126, 229 lotic, 98, 115, 119, 126, 127 marine, 45, 165, 170, 179, 193, 202, 213, 214, 217 microbial, 28 natural, 4, 64, 103, 173, 229 pelagic, 212, 213, 219
stream, 119 terrestrial, 44, 54, 231, 239 tropical, 73, 239 wetland, 98 ecotoxicants, 4, 6, 7, 9, 240 ectoparasites, 66 eelpout, 213 effects acute, 67, 68, 97, 127, 153, 169 additive, 175, 218, 219 amplification of, 113, 122 androgenic, 169 carcinogenic, 140, 170, 172 cellular-level, 193, 194 chronic, 94 community, 68, 103, 115, 119, 120, 141, 142, 147, 151, 173, 175, 179, 218 compensatory, 80, 239 developmental, 145 direct, 64, 68, 94, 100, 102, 112, 114, 117, 119, 147, 148, 172, 193 duration, 121, 128 ecological, 118, 120, 139, 145, 175, 201 ecosystem, 113, 119, 143, 151, 174, 198, 199, 202, 217, 239 endocrine, 147, 153, 170, 172, 216, 217 environmental, 191 estrogenic, 69, 152, 153, 169 eutrophication, 218 evaluation of, 123, 165, 166, 172, 190, 219, 240 extrapolation of, 173 feminization, 153 genetic, 190 genomic, 192 growth, 148, 170 indirect, 64, 72, 73, 76, 79, 100, 101, 102, 104, 112, 117, 118, 119, 121, 124, 127, 146, 165, 170, 216, 218 individual-level, 113, 114 interactive, 104 lethal, 68, 69, 102, 170 long-term, 77, 120, 121, 122, 126, 142, 239 metal toxicity, 170, 199 mixtures, 154, 202, 239 organismal, 172 phototoxic, 218 phyisiological, 7 physiological, 114, 150, 170, 200 phytotoxic, 146, 151, 215 population, 113, 125, 146, 147, 152, 153, 169, 216, 217, 239 prediction of, 116, 119, 123, 124, 125, 126, 128, 173 reproductive, 47, 94, 114, 143, 145, 152, 153, 170, 216, 217 sedative, 8 side-, 64, 67, 69, 80, 192, 239 sublethal, 9, 67, 68, 73, 77, 78, 79, 170, 191, 202, 213, 216, 240 suborganismal, 123 synergistic, 68, 168, 179, 196, 218 teratogenic, 8, 140 threshold, 191
tissue-level, 169, 216 toxic, 4, 7, 67, 68, 69, 90, 140, 146, 165, 195, 200, 216 transient, 103, 120, 127, 194 effluents, 10, 150, 153, 168, 200, 217 egg shell thinning, 114, 143 electrical insulators, 6, 148 embryotoxicosis, 54 emissions domestic, 36, 38 fuel, 5 industrial, 145 emulsifiers, 152 endocrine disruption, 8, 9, 170, 194 endocytosis, 230 endosulfan, 67, 69, 72, 75, 77, 112, 115, 129, 193, 194 endpoints, 10, 92, 96, 98, 113, 114, 115, 120, 121, 123, 124, 127, 128, 149, 150, 172, 178, 179, 191, 192, 202 endrin, 114 enzymes, 7, 26, 169, 225, 231 Ephemeroptera. See mayflies epifauna, 175 epoxy resins, 152 equilibrium, 14, 15, 17, 238 air-aerosol, 16 air-water, 15, 20 intermedia, 19 solids-water, 14, 21, 22 eradication, 100, 114, 115, 119 ergosterol, 8, 116 essential elements, 9, 43, 48 estrogen, 77, 151 estuaries, 21, 174, 189, 192, 193, 199, 201, 214, 216, 217, 240 ethynylestradiol, 150 eucalypt plantations, 89 eutrophication, 5, 217, 218 excretion, 43, 44, 46, 47, 48, 50, 51, 55, 146, 150, 168, 199 experiments field, 90, 92, 98, 101, 102, 103, 104 *in-situ*, 218 laboratory, 27, 96, 101, 104, 120, 127, 147, 168, 171 mesocosm, 115, 117, 122, 123, 125, 126, 215 population-level, 124 stream, 98, 101, 113 explosives, 6 exposure acute, 92, 150, 171, 172, 191, 193, 194, 197, 200, 201, 215 aquatic, 112, 217 assessment, 34, 35 chronic, 9, 98, 150, 171, 195, 196, 198, 200, 201, 214, 217, 228, 240 dermal, 66 dietary, 47, 67, 143, 175, 217 direct, 55, 65, 77, 90, 112, 146, 148 frequency, 122 history, 214 inhalation, 67
monitoring, 112 pulse, 68, 72, 92, 121, 122, 124, 127, 218 repeated, 113 routes, 45, 55, 66, 112, 192 sublethal, 51 time, 9, 50, 52, 92, 97, 98, 99, 112, 113, 121, 122, 138, 172, 173 timing, 122 extinction, 78, 80, 138, 170 extrapolation, 29, 123, 138
# **F**
factors abiotic, 111, 126, 146, 167 bioaccumulation, 10, 142, 151 bioconcentration, 141, 149 biomagnification, 45, 78, 142, 149 biotic, 123 climatic, 179, 197, 240 confounding, 126, 127, 169, 178 ecological, 124 environmental, 25, 27, 46, 53, 127, 172, 213, 215 meteorological, 18 safety, 145 temporal, 122 falcons, 64, 78 famphur, 66, 77 farming, 76, 79, 111, 112, 166, 189, 192 fecundity, 125, 147, 153, 195 fenpropidin, 8 fensulfothion, 65 fenvalerate, 114 fertilisation, 188, 194, 198, 201, 203 fertilizers, 5, 6, 43, 63, 66, 79, 192, 199 filter feeders, 175, 197 fingerlings, 95, 142 fipronil, 75 fire fighting foams, 148 retardants, 6 fish kills, 4, 114, 115 fishing, 6, 51, 52, 142, 166, 188, 189 fjords, 214 flies, 63, 72, 73 alderflies, 47 blowfly, 66 crane-flies, 79, 103 fruit, 48, 68 house, 69 sawflies, 79 stonefly, 101, 103, 147 tse-tse, 65 flocculation, 229 flounder, 213
fluazuron, 66, 73 fluometuron, 227 fluorine, 140, 148 fluoxetine, 151 flusulfamide, 8 fonofos, 65 food web, 44, 45, 46, 47, 48, 54, 55, 112, 114, 140, 141, 142, 143, 145, 149, 166, 179, 187, 212, 213, 215, 238, 240 forests, 72, 76, 93, 102, 239 boreal, 89 coastal, 94 mangrove, 187 natural, 88 temperate, 239 formalin, 8 fossil fuels, 4, 6, 145, 238 fragrances, 150 Freundlich-isotherm, 14 frogs, 66, 69, 77, 78, 79, 144, 149 fumigants, 72, 73 fungi, 4, 5, 7, 8, 46, 64, 70, 102, 115, 117, 119, 170, 225, 228, 231 fungicides, 5, 64, 70, 71, 73, 117 application, 65 copper, 48, 68, 69, 70, 73, 80 imidazole, 117 organo-mercurial, 52, 66, 70 persistent, 69 strobilurin, 8 triazoles, 5 usage, 64
# **G**
gall bladder, 149 gamagrass, 226 gardening, 65 germination, 7, 8, 99 gills, 47, 166, 199 gizzard, 51, 66 glyceollin, 71 glyphosate, 70, 71, 73, 79, 89, 93, 94, 95, 96, 97, 104, 193 godwits, 53 goose, 50, 67 goosefoots, 79 grasshoppers, 54, 79 grazers, 48, 102, 146 great tit, 50 grebes, 114 Green Revolution, 63, 64, 74 greenfinch, 50 greenhouse gasses, 5 grouse, 101 guidelines, 10, 172, 174, 177, 179, 191
# **H**
haddock, 216 haemoglobin, 7 haemorrhaging, 8 half-life, 13, 14, 29, 39, 69, 92, 96, 152 harbours, 193, 217 hares, 94 hatcheries, 142 hatching, 9, 54, 77, 143, 144, 187 hawks, 78 hazard quotient, 10, 92 hemlock looper, 99, 101 hemolysis, 48 Henry's law, 15, 18 hepatocytes, 8, 146 heptachlor, 67, 69, 71, 72, 73 herbicides, 63, 74, 79, 88, 113, 116, 189, 191, 226 application, 65, 66, 67, 90 auxin-type, 121 broad-spectrum, 8, 73, 74 formulation, 68 growth inhibiting, 121 organoarsenic, 53, 72 persistent, 69, 79 phenoxy, 6 PSII-inhibiting, 8, 121, 193, 194, 196 pyridine, 8 residues, 68 selective, 73 silvicultural, 93 sulfonylurea, 8 systemic, 193 triazine, 72, 77, 193, 215 urea-derived, 73, 193 usage, 65, 74, 79, 80, 93 herbivores, 47, 52, 53, 118, 175 herons, 78, 79 hexazinone, 93, 118, 195 homeostasis, 145, 199 hormone mimic, 150 human health, 4, 90, 122, 123, 193 hunting, 51, 52 hydrocarbons aliphatic, 6, 8 aromatic, 8, 169, 197, 198, 227, 228 halogenated aromatic, 139 petroleum, 5, 192, 197, 198, 226, 227 polycyclic aromatic (PAH), 5, 6, 17, 139, 145, 197, 198, 213, 214, 216 hydrolysis, 23, 24, 26, 91, 92, 97, 98, 111, 226 Hymenoptera. *See* bees, ants, parasitoids hyperaccumulation, 47 hyphomycetes, 121 hypoxia, 176, 233
**I**
ibuprofen, 151 imazapyr, 93 imazethapyr, 8 imidacloprid, 8, 65, 70, 73, 75, 76 imidazoles, 8 immobility, 77, 148 impacts amphibians, 96 community, 4, 76 direct, 71, 76 ecological, 4, 88, 92 ecosystem, 7, 68 indirect, 74, 79, 80 long-term, 80 population, 72 temporary, 6, 100, 102, 103, 104, 116, 118, 119, 124, 138, 139, 140, 141, 142, 145, 146, 147, 150, 151, 152, 153, 154, 170, 175, 176, 179, 192, 197, 199, 212, 215, 217, 219, 232, 238, 239, 240 imposex, 7, 170 incineration, 6, 140, 238 indandiones, 6 index AZTI Marine Biotic Index (AMBI), 174 Benthic Quality Index (BQI), 174 biotic, 142, 147 dissimilarity, 101 gonadosomatic, 153 maturity, 152 Shannon-Wiener diversity, 146, 174 Trophic Index (TI), 45 induction, 8, 26, 97, 168, 169, 228 industry, 65 chemical, 6, 63, 144, 166 electrical, 3 fuels, 174 manufacturing, 166 pesticide, 5 pharmaceutical, 6 plastics, 152, 166 infection, 69, 71, 79, 201 ingestion, 10, 47, 48, 50, 51, 52, 67, 78, 80, 103, 197, 212 inhibition, 77, 92, 102, 115, 118, 123, 145, 148, 151, 153, 168, 175, 191, 194, 195, 197, 198, 200, 201, 203, 230 inhibitors acetyl-cholinesterase, 5, 64, 66, 76, 78 biosynthesis, 5 cell growth, 5 germination, 6, 8 metabolic, 64 nicotinic, 5 photosynthesis, 6, 8 respiration, 5, 8 insecticides, 5, 79, 80, 99, 113, 117, 195, 216 application, 72
avermectins, 66, 73 benzoylurea, 8 biological, 89 carbamate, 8, 72, 73, 75 cyclodiene, 64, 67, 78 hydrophobic, 72 lipophilic, 66 neonicotinoid, 64, 68, 76, 116 neurotoxic, 8, 69, 76, 194 organoarsenic, 69 organochlorine (OC), 5, 8, 9, 64, 66, 67, 68, 71, 72, 73, 75, 77, 78, 111, 114 organophosphorus (OP), 66, 69, 70, 72, 76, 77, 78, 114, 120, 193, 226 persistent, 69 pyrethroid, 8, 64, 69, 70, 72, 74, 75, 76, 77, 115, 120 selective, 101 systemic, 65, 70, 72, 75, 76 usage, 64 insectivores, 53 insects, 10, 63, 64, 66, 67, 76, 78, 79, 90, 99, 115, 116 aquatic, 97, 103, 121, 142, 147 dung-breeding, 66 littoral, 102 non-target, 65, 92, 100, 103 parasitic, 75 phytophagous, 74 pollinating, 102 predatory, 75 insulating fluids, 3, 140 integrated pest management (IPM), 74, 80, 104, 111 integrative assessments, 166, 179 intensive agriculture, 79, 188, 199 interactions, 215, 218 chemical, 154, 215 complex, 4, 45, 168, 227, 240 ecological, 125, 126 interspecies, 102, 112, 171, 214 microbial, 217, 230, 231, 238 multiple species, 96 multiple stressor, 97, 104, 190, 201, 240 synergistic, 69, 146, 176 trophic, 143, 173, 175 intersex. *See* imposex intestine, 49 intoxication, 4, 122 invertebrates, 140, 142 aquatic, 46, 47, 95, 97, 98, 101, 102, 103, 117, 141, 142, 144, 146, 148, 149, 151 benthic, 101, 147, 149, 176 detritivorous, 48 herbivorous, 48 macro-, 73, 117, 128, 239 marine, 7, 122, 145, 168, 187, 193, 239 pollution-tolerant, 147 predatory, 76, 128 terrestrial, 94
iodine, 6 Irgarol 1051, 193, 194, 195, 215 irrigation, 52, 54, 63, 65, 67, 77, 111, 112 isopods, 44, 48, 52 isotope ratios, 46 ivermectin, 73, 151
# **J**
jellyfish, 216
# **K**
kestrels, 78 kidney, 48, 49, 50, 53, 54, 78, 149 kinetics biodegradation, 27, 29 first-order, 13, 27, 28, 30, 31, 230 mass flow, 33 metal, 45, 56 pseudo first-order, 13, 23, 25, 27, 30 second-order, 30 transport, 34 kites, 67 knotgrasses, 79
# **L**
lacewings, 48 lagoons, 79 lakes, 5, 67, 76, 98, 226, 229 lambda-cyhalothrin, 226 lamprey, 142 landfills, 150, 166, 189 larvae, 174, 194, 197 amphibian, 96, 97, 98 copepod, 152 coral, 194, 198 Diptera, 72 fish, 216 insect, 48, 53, 72 midge, 75 mosquito, 68 planula, 188 leaching, 22, 72, 95, 97, 150, 193 lead, 7, 8, 34, 43, 47, 201, 212, 226, 229, 230, 232 in fuel, 6 shot, 51, 52, 213 sinkers, 51, 52 uptake, 230 leafworms, 75 Lepidoptera, 99, 100, 101, 102, 103 levels background, 170
baseline, 6 ecosystem, 173, 192, 239, 240 environmental, 6 exposure, 214, 217 infestation, 66 lowest-observed effect (LOEL), 9, 55 metal, 47, 49, 50, 170, 200, 229 non-toxic, 138 no-observed effect (NOEL), 9 of contaminants, 174, 176, 213, 214 of organization, 55, 104, 113, 114, 123, 128, 138, 139, 148, 165, 167, 173, 190, 239 pH, 24 protein, 169 residue, 76, 77, 78, 95, 143, 169 threshold, 147 trophic, 44, 45, 46, 48, 50, 51, 52, 53, 55, 67, 100, 118, 119, 126, 141, 142, 145, 146, 151, 172, 173, 175, 187, 199, 202, 218, 219 vitellogenin, 216 lichens, 96 ligands, 46, 168 lignin, 70 lindane, 8, 65, 66, 70, 72, 73, 76, 111, 194 lines of evidence (LOE), 104, 167, 176 lipid tissues. *See* adipose tissue lipofuscins, 168 litter, 67, 70, 71, 73, 94, 95, 100, 102, 119, 166 liver, 7, 46, 48, 49, 50, 53, 54, 78, 79, 149, 168, 216 livestock, 64, 66, 73, 150, 151, 166, 239 lizards, 77, 78 locusts, 63 loons, 52 lubricants, 6 lucerne, 74 lufenuron, 8 lungs, 66 lysosomes, 168
# **M**
macrophytes, 79, 102, 115, 116, 117, 118, 120, 121, 141, 142, 146, 148, 151, 239 magnesium, 7 magpies, 77 malaria, 65 malathion, 69, 115 malformations. *See* deformities malondialdehyde, 169 mammals, 52, 53, 76, 142, 144 carnivorous, 55, 80 grazing, 78 herbivorous, 94, 95 insectivorous, 76, 78 marine, 47, 144, 145, 190, 196, 213, 215, 217 ruminants, 48 small, 44, 46, 48, 49, 53, 54, 66, 77, 94, 100
management drift, 65 environmental, 29, 105, 154, 167, 179, 202, 212, 238 farm, 79 forest, 88, 89, 90, 104, 239 oceans, 219 pest, 66, 74, 75, 76, 99, 104 policies, 204 vegetation, 92, 93, 95 water quality, 22 weed, 74 wildlife, 54 mancozeb, 71 manganese, 6, 199, 230 manufacturing, 3, 6, 144 manure, 48, 150 marinas. *See* harbours marshes, 68 materials allochthonous, 98 biogenic, 145 building, 3, 144, 187 new, 3, 228 radioactive, 197 water-proof, 5 mealybugs, 75 mechanism detoxification, 44, 49, 69, 76, 168, 230 electron-transfer, 8, 196 exchange, 18 homeostatic, 44 of remediation, 225, 226, 230, 232 of toxicity, 7, 94, 99, 143, 171 sorption, 14 transport, 17 medaka, 146 medicines, 3, 151 mefenoxam, 70 memory impairment, 77 mercury, 7, 199, 201, 212, 213, 217, 226, 232 mesofauna, 71 metabolism Phase I, 168 Phase II, 168 metabolite, 9, 26, 67, 69, 95, 97, 145, 214, 216 metalaxyl, 70 metalloids, 43, 55, 56, 178, 199, 200, 230, 231, 238 metallothionein, 48, 168 metamitron, 116 metamorphosis, 66, 77, 148, 149, 188, 191, 194, 198, 201, 203 methane, 23, 27, 70, 146 methoprene, 73 methylation, 46, 232 methylmercury, 44, 170, 232
metolachlor, 8, 70, 227 metsulfuron methyl, 93 mice, 44, 54, 66, 77 microcystins, 4 micronutrients, 6, 7 micro-organisms, 4, 23, 26, 27, 28, 29, 71, 72, 238 microtubules, 6, 8 midges, 102 migration, 9, 51, 78, 95, 214 millipedes, 72 minerals, 3, 52, 70 mining, 3, 6, 48, 54, 64, 166, 174, 189, 190, 199 mink, 46, 55 minnows, 98 mirex, 78 mites, 72 oribatid, 72 phytoseiid, 76 predatory, 72 saprophagous, 72 *Tetranychus*, 75, 76 mitigation, 191, 225, 226, 227 mitochondria, 7, 8 models Aquatox, 128 bioaccumulation, 46 Comprehensive Aquatic Systems, 139 ecological, 124 energy budget, 124, 125 fate, 14, 29, 123, 128, 238 fugacity, 14, 34 individual-based, 124, 125 mass balance, 29, 33 multicompartment, 31 multimedia, 33, 34, 35 PERPEST, 128 pharmaco-kinetic, 29 population, 124, 125, 139 predictive, 77 Quantitative Structure-Activity Relationship (QSAR), 123 transport, 29 moles, 50, 76 molluscicides, 64 molluscs, 119, 121, 168, 169, 172 bivalve, 165, 169 gastropods, 7, 117, 170, 175 molybdenum, 6, 48 monensin, 151 monitoring, 4, 21, 90, 91, 95, 97, 98, 100, 103, 104, 112, 116, 119, 126, 127, 128, 154, 165, 167, 169, 173, 191, 198, 204, 216, 217, 238 monocultures, 26, 63, 74, 96 monosodium methylarsonate MSMA, 53 monoxygenases, 228 monuron, 72, 73
moose, 48, 94 morphine, 8 morpholines, 5, 8 moths bogong, 53 coddling, 75 gypsy, 99, 100, 101, 103 painted apple, 99, 100 tussock, 99, 100 multivariate analyses, 177, 179 statistical techniques, 119 muscle, 50 mussels, 142, 147, 168, 169, 171, 213, 214 mutations, 8, 171, 198 mutualism, 117 mycotoxins, 4 Myriapoda, 72, *See* millipedes mysids, 216
# **N**
nanotechnology, 55 naphthalene, 228 narcotics, 5 necrosis, 145, 149 neem. *See* azadirachtin nekton, 214, 216 nematicides, 64 nematodes, 64, 70, 71, 72, 73, 75, 152 nervous impulse, 8 nestlings, 69, 77 nettle, 47 niche, 122 nickel, 7, 171, 199, 226, 230, 231 nicotine, 8, 63, 75 nitrate, 27, 70, 233 non-target organisms, 64, 65, 66, 69, 77, 80, 90, 92, 94, 102, 115, 193 nonylphenol, 5, 151 norflurazon, 79 nozzles, 90 nuclear polyhedrosis virus, 74
# **O**
oak, 100 ocean acidification, 191, 196, 201, 204 offspring, 54 Oligochaeta. *See* worms oocytes, 194 orange groves, 74 orchards, 63, 65, 68, 73, 75, 76, 112 apple, 76, 77 citrus, 75
organelles, 51 organoselenium, 44, 54 organotin, 7, 44, 46 osteomalacia, 52 osteoporosis, 52 Ostracoda, 121 otters, 46 overspray, 98 owls, 67, 78, 79 oxadiazon, 71 oxidants, 5, 24 oxidation, 8, 23, 24, 25, 26, 27, 28, 52, 70, 199, 226, 231, 232 oxyfluorfen, 71 ozone, 5, 24, 25
# **P**
paper mill, 168 paralysis, 8 paraquat, 8, 72, 73, 74 parasiticides, 72 parasitoids, 74, 75, 80 parathion, 65, 67, 68, 75, 119, 227, 228 particles, 14, 16, 213, 217 aerosol, 17, 18 clay, 94 sediment, 21 size, 18 soil, 21, 29, 70, 99 suspended, 21, 198 partitioning, 14, 15, 16, 17, 29, 34, 46, 51, 88 carbon-water, 29 partridge, 79 passive samplers, 213, 214, 219, 238 pathogens, 71, 73, 80 Pauropoda, 72 pauropods, 72 pelicans, 78 pellets, 51, 213 pendimethalin, 73 pentachlorophenol, 70, 140 perch, 51, 98 percolation, 17, 73 perennial plants, 94 perfluorooctane sulfonic acid, 148 perfluorooctanoic acid, 148, 217 periphyton, 117, 119, 121, 128, 141, 152 permethrin, 195, 226 peroxisome, 149, 168 persistent organic pollutants (POPs), 44, 140, 143, 145, 146, 148, 149, 169, 216, 217, 218 personal care products (PPCPs), 150 pest control, 4, 64, 74, 78, 80, 89, 92, 104, 114 pest outbreaks, 99 pesticides
biodegradable, 68 chlorinated, 6, 70 granular, 65 modern, 67, 69, 72, 92, 104, 191, 192 natural, 104 persistent, 69, 80 synthetic, 64, 111 petroleum, 3, 5, 6, 7, 145, 146, 197, 198, 226, 227 pharmaceuticals, 3, 5, 139, 150, 151, 154, 165, 189, 217, 238, 239 phenanthrene, 228 pheromones, 64 phorate, 65, 73 phosphatase, 71, 146 phosphates, 70 phthalate ester, 168 phytoalexin, 71 phytoextraction, 225, 226 phytoremediation, 225, 226, 227, 229, 230, 233, 238 picloram, 70 piperonyl butoxide, 68 plankton phytoplankton, 54, 115, 116, 117, 121, 126, 128, 141, 147, 151, 152, 170, 214, 215, 216, 218 zooplankton, 54, 96, 98, 115, 116, 117, 118, 121, 126, 147, 149, 151, 152, 170, 214, 216, 218, 219, 240 plant hoppers, 75 plant tissues, 230 plasmids, 232 plasticizers, 6, 139, 168 plastics, 3, 144, 151, 152, 213, 214 polycarbonate, 152 rigid, 144 Plecoptera. *See* caddisflies plumes, 18, 189, 196, 204 poisoning, 78 by metalloids, 52, 53, 54 by metals, 51, 52, 55 by pesticides, 77, 78, 79, 80 primary, 67 secondary, 67, 78 polar bears, 46 pollination, 73, 74, 80 pollinators, 74 pollutants inorganic, 225, 226, 238 toxic, 4, 5, 7, 238, 239 polybrominated diphenyl ethers (PBDEs), 6, 139, 144, 165, 171, 217 polychlorinated biphenyls (PCBs), 3, 6, 139, 140, 196, 217, 227, 228 polychlorinated dioxins (PCDDs), 5, 139, 140 polyps, 188, 201 polyurethane, 144, 214 ponds, 98 poplar, 226 population decline, 51, 53, 54, 79, 114, 142, 143, 144, 146, 217 density, 113, 118
human, 52, 111 possums, 53 practices agricultural, 63, 73, 74, 76, 79, 80, 240 management, 65, 66, 74, 79, 80, 239 no-tillage, 74, 76 prairie dog, 50 prawns, 171 precautionary principle, 123 precipitation, 17, 18, 21, 22, 35, 111, 119, 217, 229, 230, 231, 232 predators, 51, 53, 55, 67, 75, 76, 78, 80, 173 carnivorous, 45, 50 insectivorous, 100 invertebrate, 48, 76 mammalian, 46, 55, 78 marine, 46, 175, 196, 213, 216, 217, 219 primary consumers, 78, 146 primary producers, 45, 113, 116, 117, 119, 120, 121, 122, 194, 196, 215, 219, 240 pristine areas, 175 processes abiotic, 23, 213 advection, 22 biochemical, 93 bioconcentration, 45 biological, 69, 197, 233 biosynthetic, 8 calcification, 187 cellular, 196 chemical, 215 degradation, 23, 26, 29, 39 diagenetic, 145 dispersive, 17 ecological, 194, 229 elimination, 26 erosion, 199 evolutionary, 171 geogenic, 43 hydrographic, 212 industrial, 240 metabolic, 8, 64, 73 microbial, 27, 28, 226, 231, 232 mineralization, 23 nutrient-cycling, 70 pelagic, 219, 240 phosphorylation, 7, 231 photolytic, 26 photosynthetic, 188 physiological, 8, 193 phytoremediation, 225 redox, 231 removal, 17, 27 reproductive, 197 transformation, 23 transport, 18, 22, 33, 238
production agricultural, 63, 192 biofuel, 111 chemical, 3, 34 electricity, 55 food, 111, 192 forests, 88 PBDEs, 144 PCBs, 144 pesticide, 4, 111 PFOS, 148 plastics, 151 primary, 113, 126, 188, 193, 212, 213, 215 secondary, 215 products agrochemical, 68, 69, 80 animal, 3 by-products, 6, 140 cleaning, 3 metallic, 3 natural, 3 pesticide, 64 waste, 44, 70, 189 profenofos, 194, 195 propellants, 5 prosulfuron, 71 Proteobacteria, 228 protists, 70, 71, 213, 217, 219, 240 pseudomonads, 228 pteridophytes, 96 Pyralidae. *See* stem borers pyrethrum, 8, 63 pyrithione copper, 215 zinc, 190, 201, 215
# **Q**
quinalphos, 70
# **R**
raccoons, 50, 67 rape seeds, 66 raptors, 52, 78 rates application, 70, 71, 90, 93, 99, 192 bioaccumulation, 189 biodegradation, 27, 37, 189 calcification, 191, 199, 200 degradation, 193, 201 elimination, 49, 50 growth, 94, 95, 144, 175, 198, 200 metabolic, 48, 76
second-order, 27 rats, 53 receptor aryl hydrocarbon (AhR), 8, 140, 142, 144, 145, 146, 196 cholinesterase, 8 estrogen, 152 GABA, 8 nicotinic, 8 recolonisation, 122, 126, 175, 239 recovery, 121, 123, 124, 125, 179 time, 104, 121, 122 red fox, 50 redundancy, 113, 126 reeds, 226 refinery, 43, 44, 49, 52, 53, 198 refrigeration, 3 refugia, 122, 126 regeneration, 8, 70, 88, 89, 92, 94, 141, 200 regulatory authorities, 4, 138, 173 rehabilitation, 96 remobilization, 9 reproduction impairment, 77, 78 residence time, 17, 39, 217 residues faeces, 53, 67, 73 feathers, 50, 66 resilience, 70, 93, 96, 188, 196, 201, 202, 204, 225 resistance, 19, 20, 21, 39, 64, 75, 80, 151, 168, 201, 231, 232 respiration, 113, 146, 151 aerobic, 47 anaerobic, 231 microbial, 70 soil basal, 70 resurgence, 75 resuspension, 17, 20, 21, 22, 33, 199, 213 retinol, 8 rhizodegradation, 229 rhizomes, 94, 230 rhizosphere, 70, 71, 225, 230 rodenticides, 8, 63, 64, 67, 78 rodents, 63, 67, 77, 78, 145 rotifers, 119, 147, 149 rubber, 5, 73 rubidium, 50 runoff, 67, 68, 80, 94, 111, 112, 114, 119, 126, 128, 131, 151, 174, 189, 190, 192, 197, 198, 199, 202, 203, 225 agricultural, 190 stormwater, 5 rushes, 80
# **S**
saithe, 216 salamanders, 102, 148 salinity, 117, 174, 199, 201, 213, 215
salmon, 95, 97, 149 salmonberry, 95 sandpipers, 50 scrapers, 142 sea urchins, 201 seagrass, 187, 194 seagulls, 140, 143 seals, 46, 216, 217 sedges, 80 Sediment Quality Triad (SQT), 176 sedimentation, 17, 18, 20, 21, 22, 154, 212, 217, 218, 229 seed-dressings, 65 selection pressure, 47, 75, 122 sequestration, 51, 55, 168 serum proteins, 149 sewerage, 4, 21 sheep, 48, 73 shellfish, 166, 215 shredders, 103, 142 shrews, 44, 48, 49, 53, 76, 101 silicone, 214 silver, 50 simazine, 72, 73, 193 skeleton, 188, 199 skin, 7, 46, 50, 52, 66, 96 skylark, 79 smelters, 6 smog, 4 snails, 47, 64, 147, 153 snipe, 52 sodium channels, 8 soil fertility, 70, 72, 73, 74, 80 soils agricultural, 43, 54 floodplain, 48 metalliferous, 229 urban, 6 solar radiation, 24, 26 solvents, 5, 6, 226, 227 sparrowhawk, 78 spawning, 142, 153, 169, 188, 198 spearfish, 216 species competitive, 122 endangered, 53, 100, 101 invasive, 89, 215 k-selected, 174 meiobenthic, 171 opportunistic, 174, 175 pelagic, 213, 214 r-selected, 174 sensitive, 116, 117, 119, 125, 146, 173, 174, 175 sentinel, 165, 168, 169 tolerant, 44, 119, 146, 174
species sensitivity distribution (SSD), 128 sperm, 153, 188, 194 Sphaeriidae. *See* molluscs spiders, 48, 72, 74, 75, 76, 79 spills, 6, 10, 189, 190, 198, 199, 217, 225 oil, 170, 197, 202 spinal cord, 8 spinosad, 66 spleen, 50 spores, 7, 73, 99 springtails, 72, 76 spruce, 89, 96, 99, 101 squirrels, 67 starlings, 54, 69 starvation, 9, 63, 79 steady-state, 31, 32, 34, 35, 36, 37, 38, 39 stomach, 50, 101 structure community, 53, 54, 70, 74, 101, 120, 141, 143, 147, 149, 153, 173, 174, 176, 179, 215, 218 ecosystem, 143 population, 52, 113, 147 trophic, 78, 138 vegetation, 52, 76 sub-cellular, 51, 169 substances hazardous, 10 man-made, 4 subsurface, 54, 111, 112, 216 sugarcane, 74 sulfides, 178, 230, 231, 232 sulfometuron, 93 sulfonamides, 5, 8 sulphur dioxide, 6 sunfish, 118 sunflower, 66, 76 surfactants, 6, 65, 104 non-ionic, 152 perfluorinated, 139, 148, 238 POEA, 96 susceptibility, 76, 77, 90, 113, 115, 122, 123, 190, 202, 217 suspended solids, 14, 15, 33, 35, 197 switchgrass, 226 swordfish, 213, 216 symbionts, 188, 190, 191, 194, 198, 199, 202 symbiosis coral, 187, 194 mycorrhizal, 70, 80 symphylids, 72 system atmospheric, 25 endocrine, 169 hormonal, 145 immunosystem, 170 nervous, 7, 64
photosystem II (PSII), 8, 191, 193, 215 respiratory, 7 root, 70, 73, 93, 193, 225, 226, 229, 230
# **T**
Tabular Decision Matrix (TDM), 177 tadpoles, 77, 79, 144 tebufenozide, 99, 101 technetium, 231 terbufos, 65 terns, 143 testosterone, 77, 153 tests acute toxicity, 123, 124 biodegradibility, 28 bioluminescence, 151 cell line, 123, 146 chronic toxicity, 123, 179 embryo, 123 flow-through toxicity, 97 laboratory toxicity, 113, 147 life cycle, 150 multispecies, 176 single species, 102, 123, 124 time-toxicity, 97 tetrachloroethene, 228 tetracycline, 151 textiles, 3, 5, 144 thermal stress, 201 thermodynamics, 14 thiacloprid, 121, 126 thiobencarb, 8, 73 thiocarbamates, 6 thrush, 101 thyroid, 77, 144, 149, 217 thyroxine, 8 tier studies, 93 tin inorganic, 170, 200, 201 organotins, 192, 216 toads, 77, 144 tolerance, 54, 113, 114, 123, 168, 171, 175, 176, 200, 201, 202 tomato, 74 tourism, 187 toxaphene, 67, 70 toxic equivalents (TEQs), 143, 217 toxicity dermal, 66 **mixture**, 123 toxicokinetics, 146 toxicosis, 48, 51 toxins algal, 4, 166, 215
biological, 4, 5 Bt-endotoxins, 71, 91, 99 microbial, 4 natural, 4, 215 trace elements, 43, 50, 54 trace metals, 189, 199, 200, 201, 212, 230 trait-based risk assessment, 123 translocation, 22, 91, 111, 229, 230 trans-nonachlor, 67 transport atmospheric, 111, 112 intermedia, 14, 17, 18, 33, 34 intramedia, 17 treated areas, 53, 95 triazophos, 76 tributyltin (TBT), 7, 193, 199 trichlorfon, 70 trichloropyridinol, 97 Trichoptera. *See* stoneflies triclopyr, 8, 77, 93, 95, 97, 98, 104 triclopyr ester, 97 trophic cascade, 45 trout, 97, 98, 119, 138, 142, 146, 149, 153 tuna, 216 Turbellaria, 103 turbidity, 126, 166, 201, 202, 204 turtles, 144, 149
# **U**
uptake dietary, 47 root, 229 uranium, 231, 232 UV radiation, 112, 146, 218
# **V**
vanadium, 50 vapour pressure, 15, 16 vertebrates, 8, 48, 51, 55, 64, 66, 74, 76, 77, 78, 79, 80, 99, 116, 126, 139, 145, 146, 148, 150, 152, 153, 169, 170, 196, 213, 214 vineyards, 65, 76, 112 vitamin A, 8 vitamin K, 8 vitellogenin, 150, 153, 216 volatilization, 17, 19, 20, 21, 33, 68, 76, 231 voles, 44, 48, 49, 53, 66, 67, 77
# **W**
warblers, 95, 101 wash-out, 17, 18 wastewater treatment, 21, 150, 152, 153, 166
water groundwater, 17, 22, 38, 39, 52, 111, 152 pore, 21, 22, 29, 147, 230 quality, 10, 22, 118, 119, 123, 147, 172, 179, 190, 191, 204 solubility, 15, 104, 144, 152, 225 surface, 15, 17, 21, 24, 28, 102, 111, 112, 115, 150, 151, 189, 238 vapour, 6 watershed, 95, 98, 103, 142, 147 water fleas, 112 water hyacinth, 229 webworm, 101 weeds, 3, 63, 64, 66, 73, 74, 76, 79, 80 wetlands, 51, 76, 77, 79, 96, 98, 144, 147, 189, 226, 227, 229, 239 whales, 217 willow, 226 woodlice, 44, 67, 72, 73 woodpeckers, 53 worms earthworms, 44, 48, 49, 51, 53, 54, 70, 72, 73, 78, 80 enchytraeid, 72, 73 oligochaetes, 44, 52 polychaetes, 175 worst-case scenario, 10
# **X**
xenobiotics, 27, 167, 168, 170, 173 xylem, 52
# **Y**
yolk proteins, 169
# **Z**
zinc, 6, 7, 47, 52, 174, 193, 199, 200, 201, 215, 226, 229, 230, 231 zooxanthellae, 187, 193, 194, 198, 199, 201, 202
© 2011 The Author(s). Published by Bentham Science Publisher. This is an open access chapter published under CC BY 4.0 https://creativecommons.org/licenses/by/4.0/legalcode | doab | 2025-04-07T03:56:58.715779 | 19-10-2023 18:40 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "004623c8-cb6c-4a99-aee8-c83552616d6e",
"url": "https://library.oapen.org/bitstream/20.500.12657/75863/1/9781608051212.pdf",
"author": "",
"title": "Ecological Impacts of Toxic Chemicals",
"publisher": "Bentham Science Publishers",
"isbn": "9781608051212",
"section_idx": 0
} |
004acd65-0f8a-4342-8761-8da04dfc73fe.0 | **Application of Liquid Chromatography in Food Analysis**
• Oscar Núñez and Paolo Lucci
| doab | 2025-04-07T03:56:58.778756 | 1-5-2021 17:44 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "004acd65-0f8a-4342-8761-8da04dfc73fe",
"url": "https://mdpi.com/books/pdfview/book/2994",
"author": "",
"title": "Application of Liquid Chromatography in Food Analysis",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783039433629",
"section_idx": 0
} |
004acd65-0f8a-4342-8761-8da04dfc73fe.1 | **Application of Liquid Chromatography in Food Analysis**
Printed Edition of the Special Issue Published in *Foods* Oscar Núñez and Paolo Lucci Edited by
www.mdpi.com/journal/foods
| doab | 2025-04-07T03:56:58.778838 | 1-5-2021 17:44 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "004acd65-0f8a-4342-8761-8da04dfc73fe",
"url": "https://mdpi.com/books/pdfview/book/2994",
"author": "",
"title": "Application of Liquid Chromatography in Food Analysis",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783039433629",
"section_idx": 1
} |
004acd65-0f8a-4342-8761-8da04dfc73fe.2 | **Application of Liquid Chromatography in Food Analysis**
| doab | 2025-04-07T03:56:58.778879 | 1-5-2021 17:44 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "004acd65-0f8a-4342-8761-8da04dfc73fe",
"url": "https://mdpi.com/books/pdfview/book/2994",
"author": "",
"title": "Application of Liquid Chromatography in Food Analysis",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783039433629",
"section_idx": 2
} |
004acd65-0f8a-4342-8761-8da04dfc73fe.3 | **Application of Liquid Chromatography in Food Analysis**
Editors
**Oscar N ´u ˜nez Paolo Lucci**
MDPI • Basel • Beijing • Wuhan • Barcelona • Belgrade • Manchester • Tokyo • Cluj • Tianjin
*Editors* Oscar Nu´ nez ˜ University of Barcelona (INSA-UB) Spain
Paolo Lucci University of Udine — UNIUD Italy
*Editorial Office* MDPI St. Alban-Anlage 66 4052 Basel, Switzerland
This is a reprint of articles from the Special Issue published online in the open access journal *Foods* (ISSN 2304-8158) (available at: https://www.mdpi.com/journal/foods/special issues/liquid chromatography).
For citation purposes, cite each article independently as indicated on the article page online and as indicated below:
LastName, A.A.; LastName, B.B.; LastName, C.C. Article Title. *Journal Name* **Year**, *Article Number*, Page Range.
**ISBN 978-3-03943-362-9 (Hbk) ISBN 978-3-03943-363-6 (PDF)**
c 2020 by the authors. Articles in this book are Open Access and distributed under the Creative Commons Attribution (CC BY) license, which allows users to download, copy and build upon published articles, as long as the author and publisher are properly credited, which ensures maximum dissemination and a wider impact of our publications.
The book as a whole is distributed by MDPI under the terms and conditions of the Creative Commons license CC BY-NC-ND.
| doab | 2025-04-07T03:56:58.778904 | 1-5-2021 17:44 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "004acd65-0f8a-4342-8761-8da04dfc73fe",
"url": "https://mdpi.com/books/pdfview/book/2994",
"author": "",
"title": "Application of Liquid Chromatography in Food Analysis",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783039433629",
"section_idx": 3
} |
004acd65-0f8a-4342-8761-8da04dfc73fe.4 | **Contents**
## **About the Editors**
**Oscar N ´u ˜nez** studied chemistry at the University of Barcelona where he also received his Ph.D. in 2004. He worked as visiting researcher for half a year in the development of on-line pre-concentration methods using micellar electrokinetic chromatography (MEKC) at the University of Hyogo (Japan) in collaboration with Professor Terabe (father of MEKC technique). Starting in October 2005, he joined the Kyoto Institute of Technology (Japan) as a two-year post-doc researcher working with Professor Tanaka developing monolithic silica capillary columns under a fellowship from the Japan Society for the Promotion of Science. In November 2007, he joined the University of Barcelona again, working in the Chromatography, Capillary Electrophoresis and Mass Spectrometry research group. Since December 2014, he has been a Serra Hunter Professor at the ´ Section of Analytical Chemistry (Department of Chemical Engineering and Analytical Chemistry, University of Barcelona). He has more than 130 scientific papers and book chapters to his name, and he is editor of three books on LC-MS/MS, sample preparation techniques in food analysis, and capillary electrophoresis. He has extensive experience in the development of liquid chromatography methods coupled to low- and high-resolution mass spectrometry, as well as sample treatment procedures, for environmental and food analysis. His major research areas nowadays involve the characterization, classification and authentication of food and natural products, as well as the prevention of food frauds.
**Paolo Lucci** attained his Ph.D. in 2008 at the SAIFET department of the Polytechnic University of Marche (Italy). In February 2009, he joined the NASCENT European project as an experienced researcher at POLYIntell SAS (France) and then in 2010 he spent one year as an experienced researcher at the Department of Analytical Chemistry of the University of Barcelona within the Carbosorb European Project. In April 2011, he joined the School of Sciences of the Pontificia Universidad Javeriana (Colombia) where he was named head of the research group of "Foods, Nutrition and Helath" in 2012, and then head of the Department of Nutrition and Biochemistry in 2014. Currently, he is a Senior Researcher at the Department of Agri-Food, Environmental and Animal Sciences at the University of Udine (Italy). He has published more than 70 scientific papers and book chapters, and he is editor of two books focused on LC-MS and sample preparation techniques in food analysis. He has extensive experience in liquid chromatography coupled with tandem mass spectrometry (LC-MS/MS), molecularly imprinted polymers (MIPs), as well as alternative treatment procedures for environmental and food sample analysis. Other research areas include the chemical characterization of natural extracts and the evaluation of their health-promoting effect through in vitro and in vivo studies.
### *Editorial* **Application of Liquid Chromatography in Food Analysis**
**Oscar Núñez 1,2,3,\* and Paolo Lucci 4,\***
Received: 9 September 2020; Accepted: 10 September 2020; Published: 11 September 2020
Food products are very complex mixtures consisting of naturally-occurring compounds and other substances, generally originating from technological processes, agrochemical treatments, or packaging materials. Several of these compounds (e.g., veterinary drugs, pesticides, mycotoxins, etc.) are of particular concern because, although they are generally present in very small amounts, they are nonetheless often dangerous to human health. On the other hand, food is no longer just a biological necessity for survival. Society, in general, demands healthy and safe food, but it is also increasingly interested in other quality attributes more related to the origin of the food, the agricultural production processes used, the presence or not of functional compounds, etc. In an increasingly populated world and with an increasingly demanding society regarding food quality, food production has become a completely global aspect on a global level. In addition, in this field, where the number of people involved in the food production process, from its origin to its consumption, is enormous, it is increasingly difficult to guarantee the integrity and, above all, the authenticity of foodstuffs. In consequence, improved methods for the determination of authenticity, standardization, and efficacy of nutritional properties in natural food products are required to guarantee their quality and for the growth and regulation of the market. Thus, food safety and food authentication are hot topics for both society and the food industry.
Nowadays, liquid chromatography with ultraviolet (LC-UV) detection, or coupled to mass spectrometry (LC-MS) and high-resolution mass spectrometry (LC-HRMS), are among the most powerful techniques to address food safety issues and to guarantee food authenticity in order to prevent fraud [1–8]. The aim of this Special Issue "Application of Liquid Chromatography in Food Analysis" was to gather review articles and original research papers focused on the development of analytical techniques based on liquid chromatography for the analysis of food. This Special Issue is comprised of six valuable scientific contributions, including five original research manuscripts and one review article, dealing with the employment of liquid chromatography techniques for the characterization and analysis of feed and food, including fruits, extra virgin olive oils, confectionery oils, sparkling wines and soybeans.
Cortés-Herrera et al. reviewed the potential of liquid chromatography for the analysis of common nutritional components in feed and food [9]. Food and feed share several similarities when facing the implementation of liquid-chromatographic analysis. Using the experience acquired over the years through the application chemistry in food and feed research, the authors selected and discussed analytes of relevance for both areas. This interesting review addresses the common obstacles and peculiarities that each analyte offers for the implementation of LC methods throughout the different steps of the method development (sample preparation, chromatographic separation and detection). The manuscript consists mainly of three sections: feed analysis (at the beginning of the food chain); food destined for human consumption determinations (the end of the food chain); and finally, assays shared by either matrices or laboratories. Polyphenols, capsaicinoids, theobromine and caffeine, cholesterol, mycotoxins, antibiotics, amino acids, triphenylmethane dyes, nitrates/nitrites, ethanol soluble carbohydrates/sugars, organic acids, carotenoids, and hydro and liposoluble vitamins are examined.
Several original research works reported the application of liquid chromatography-based analytical methodologies for the characterization of food products. Loizzo et al. characterized native Colombian fruits and their by-products by determining their phenolic profile, antioxidant activity and hypoglycaemic potential [10]. The use of ultra-high performance liquid chromatography-high resolution mass spectrometry (UHPLC-HRMS) with an Orbitrap mass analyzer revealed the presence of chlorogenic acid as dominant compound in Solanaceae samples. In addition, and based on the Relative Antioxidant Score (RACI) and Global Antioxidant Score (GAS) values, *Solanum quitoense* peel showed the highest antioxidant potential among Solanaceae samples, while *Passiflora tripartita* fruits exhibited the highest antioxidant effects among Passifloraceae samples. Considering that some of the most promising results were obtained by the processing waste portion, the authors highlighted that its use as functional ingredients should be considered for the development of nutraceutical products intended for patients with disturbance of glucose metabolism. Obyedul Kalam Azad et al. evaluated the effect of artificial LED light and far infrared (FIR) irradiation on phenolic compounds, isoflavones and the antioxidant capacity of soybean (*Glycine max* L.) sprouts [11]. Six isoflavones (daidzin, glycitin, genistin, daidzein, glycitein and genistein) were determined by LC. The authors applied artificial blue (470 nm) and green (530 nm) LED and florescent light (control) on soybean sprouts, from three to seven days after sowing in the growth chamber. Total phenolic content, antioxidant capacity and total isoflavones content were higher under blue LED compared to control. Thus, results suggested that blue LED was the most suitable light to steady accumulation of secondary metabolites in growing soybean sprouts. In another interesting research work, polyphenolic profiles obtained by high-performance liquid chromatography with ultraviolet/visible detection (HPLC-UV/Vis) and principal component analysis (PCA) were employed by Izquierdo-Llopart and Saurina for the characterization of sparkling wines [12]. Chromatographic profiles were recorded at 280, 310 and 370 nm to gain information on the composition of benzoic acids, hydroxycinnamic acids and flavonoids, respectively. The authors employed the obtained HPLC-UV/vis data, consisting of composition profiles of relevant analytes, to characterize cava wines produced from different base wine blends by using chemometrics. Other oenological variables, such as vintage, aging or malolactic fermentation, were fixed over all the samples to avoid their influence on the description. PCA and other statistic methods were able to extract the underlying information and provided an excellent discrimination of the analyzed samples according to their grape varieties and coupages. Finally, Santoro et al. performed the characterization and determination of interesterification markers (triacylglycerol regioisomers) in confectionery oils by liquid chromatography-high resolution mass spectrometry (LC-HRMS) [13]. In the confectionery industry, controlling the formation degree of positional isomers is important in order to obtain fats with the desired properties. The separation of triacylglycerol regioisomers is a challenge when the number of double bonds is the same and the only difference is in their position within the triglyceride molecule. The authors aimed to obtain a chromatographic resolution that might allow reliable qualitative and quantitative evaluation of triacylglycerol positional isomers within rapid retention times, and robustness in respect of repeatability and reproducibility by means of LC-HRMS using an LTQ-Orbitrap analyzer with atmospheric pressure chemical ionizacion (APCI). The time required for the global analysis was relatively short, the chromatographic resolution and efficiency were satisfactory and the mass detection allowed for identifying the isobaric components of each position isomer couple. In conclusion, the described method may well be considered a good diagnostic tool of interesterification consequences that are strictly connected to confectionery product quality.
HPLC-UV was also proposed by Carranco et al. for the authentication and quantitation of frauds in extra virgin olive oils (EVOOs) [14]. For that purpose, HPLC-UV chromatographic fingerprints recorded at 257, 280 and 316 nm were employed as sample chemical descriptors for the characterization and authentication of monovarietal EVOOs and other vegetable oils by chemometrics. PCA results showed a noticeable discrimination between olive oils and other vegetable oils using raw HPLC-UV fingerprints as data descriptors. However, the authors observed that selected HPLC-UV chromatographic time-window segments were able to improve the discrimination among the monovarietal EVOOs analyzed. In addition, partial least squares (PLS) regression was employed to tackle olive oil authentication of Arbequina EVOO adulterated with Picual EVOO, refined olive oil, and sunflower oil, achieving highly satisfactory results with overall errors in the quantitation of adulteration in the Arbequina EVOO (minimum 2.5% of adulterant) below 2.9%.
In summary, the Special Issue "Application of Liquid Chromatography in Food Analysis" demonstrated the great importance of liquid chromatography analytical methodologies to address the characterization and determination of targeted compounds in food, as well as to guarantee food integrity and authenticity. In addition, several authors have also demonstrated the requirement of advanced chemometric approaches as essential tools in combination with LC to achieve robust results, not only with the objective of characterizing food composition, but also to obtain satisfactory classification methods, and to prevent food fraud.
**Author Contributions:** Writing—original draft preparation, O.N.; writing—review and editing, O.N. and P.L.; All authors have read and agreed to the published version of the manuscript.
**Funding:** This research received no external funding.
**Conflicts of Interest:** The authors declare no conflict of interest.
#### **References**
© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
### *Review* **Liquid Chromatography Analysis of Common Nutritional Components, in Feed and Food**
#### **Carolina Cortés-Herrera 1, Graciela Artavia 1, Astrid Leiva <sup>2</sup> and Fabio Granados-Chinchilla 2,\***
Received: 14 September 2018; Accepted: 5 November 2018; Published: 20 December 2018
**Abstract:** Food and feed laboratories share several similarities when facing the implementation of liquid-chromatographic analysis. Using the experience acquired over the years, through application chemistry in food and feed research, selected analytes of relevance for both areas were discussed. This review focused on the common obstacles and peculiarities that each analyte offers (during the sample treatment or the chromatographic separation) throughout the implementation of said methods. A brief description of the techniques which we considered to be more pertinent, commonly used to assay such analytes is provided, including approaches using commonly available detectors (especially in starter labs) as well as mass detection. This manuscript consists of three sections: feed analysis (as the start of the food chain); food destined for human consumption determinations (the end of the food chain); and finally, assays shared by either matrices or laboratories. Analytes discussed consist of both those considered undesirable substances, contaminants, additives, and those related to nutritional quality. Our review is comprised of the examination of polyphenols, capsaicinoids, theobromine and caffeine, cholesterol, mycotoxins, antibiotics, amino acids, triphenylmethane dyes, nitrates/nitrites, ethanol soluble carbohydrates/sugars, organic acids, carotenoids, hydro and liposoluble vitamins. All analytes are currently assayed in our laboratories.
**Keywords:** food and feed analysis; liquid chromatography; challenges; nutritional analysis; additives; contaminants
#### **1. Introduction**
Food and feed analysis are paramount to assess both nutritional quality and safety of commodities. Interconnectivity of food sources [1,2] and new processing techniques [3] make for a more diverse and complex food supply. Legal thresholds have been stipulated that establish acceptable levels for individual chemical additives, residues, and contaminants in products [4,5]. Feed is a paramount target for analysis since it situates at the start of the food chain and poor feed quality can affect the yield on food-producing animals [6]. Understanding the complexities of food safety is the goal of approaches such as One Health [7], Farm-to-Fork [6], or MyToolbox [8]. Furthermore, feed contaminants carryover downstream can reach products such as meat, eggs, and milk (see for example the transference of aflatoxin M1 from aflatoxin B1-contaminated feed). Ingredients either destined for food or feed production (e.g., cereals) are among the fundamental constituents for several staple commodities. Other regulations require food and feed labeling to list ingredients relating to the nutritional content [9,10]. All stakeholders involved in the food and feed chain must be able to assess product quality and safety. Hence, it is imperative to rely on techniques that meet several
analytical performance parameters. More and more, food and feed analysis methods are based on LC (liquid-chromatography) [11,12], which has proven to be an optimal technology for screening, detection, and quantification of a vast variety of analytes (see Table 1). The reason behind this is related to the molecular affinity between the analyte and also: i. the mobile phase (which is usually a mixture of solvents) ii. stationary phase (modified silica and polymer scaffolds). Within the LC approach itself, several alternatives are available for a researcher to resolve a specific task at hand. Each analyte presents its own unique trials.
**Table 1.** Typical food and feed analytes assayed using HPLC (High-Performance Liquid-Chromatography).
**Table 1.** *Cont.*
To successfully analyze or isolate a compound, a researcher is faced with several questions: What is the problem to solve, the objective or purpose for the analysis? Is the required data qualitative or quantitative? Are there two or multiple compounds to be separated? What are the physicochemical characteristics of the target(s)? What matrix was the analyte recovered from and which interferences are expected? What is the amount of analyte expected to be recovered? What equipment is accessible in the laboratory?
Considering the above, a suitable column (Table 2) and detection system must be selected (Table 3). Sample preparation can aid to solve some of these issues, especially those regarding interferences and sensitivity but cannot solve issues with poor detector choice. For example, if sensitivity is a problem using the selected detection system on hand and no other system is available, the initial sample mass can be increased, or a concentration step (evaporation or solid phase extraction (SPE)) can be performed. Additionally, the sample injection volume can be expanded to improve sensitivity.
**Table 2.** General conditions required for each mode of chromatography.
Additionally, automation is relevant for conserving resources and reducing turnover times. An analyst can program an autosampler to increasingly adjust the volume of a standard with a fixed concentration. For example, to construct a calibration curve between 1000 and 62.5 μg L<sup>−</sup>1, one could use a 1000 μg L−<sup>1</sup> standard and instruct the sampler to take from the same vial 20 μL, 10 μL, 5 μL, 2.5 μL, 1.25 μL, consecutively. The sampler will construct a calibration curve without analyst intervention and this automation will reduce errors. Autosamplers are designed to inject small volumes without significant loss, with good precision, and adequate reproducibility. They can also inject variable amounts, dilute the sample prior to injection and perform precolumn derivatization [91]. If a sample is outside of calibration standard of higher concentration, an analyst can inject a different volume to ensure it will fit among the calibration curve range. However, injection volume has an impact on peak shape. The method must be validated to show this is a valid approach. (See for example, [92]). Reference for one example of the versatility of an LC system and capabilities for its automation. In this review, we intend to give the reader a thorough background on the common analyses performed, for quality assurance and safety, in food and feed laboratories. We will include the most recent and
relevant experience gathered for each test while pointing out the difficulties that each essay presents and the common ground shared by both types of laboratories.
**Table 3.** Characteristics of the most common detectors used in liquid chromatography.
#### **2. Measurements of Commonly Consumed Food Commodities**
#### *2.1. Polyphenols*
Polyphenols usually refers to several chemical compounds including flavanols (e.g., catechins and tannins from tea), flavanones (i.e., hesperidin from citrus fruits), flavonols (e.g., quercetin from tea, apples and onions), "chlorogenic acids" (including hydroxycinnamic acids caffeic, ferulic, *p*-coumarinic acids usually extracted from coffee), anthocyanins (which are partly responsible for imparting color to plant structures), and stilbenes (e.g., from berries, grape skins and peanuts) (Figure 1) [93].
These compounds, secondary metabolites from plants [94], have, among other functions, a protective capability within the vegetable tissue, structure, and support [94], and, even, pollinator attraction [95]. For example, chlorogenic acid (i.e., the esterification product between caffeic and quinic acid) is an intermediate in lignin biosynthesis [96]. Data suggests that long-term consumption of such compounds can have beneficial effects [94] as it can improve an organism's antioxidant capacity [93] which in turn relates, for example, to cognitive improvement [97] and reduction in adipogenesis and oxidative stress [98]. Fruits, especially berries, are [97–101] rich in these bioactive compounds, both extractable [102,103] and non-extractable [104]. From the technological standpoint, polyphenol safeguard is paramount to achieve functional foods [105] with added value
(e.g., beverages) and a bioactive capacity of compounds as close as those from the raw material. Several operation units have been applied to fruits to assess polyphenol retention after processing including nanofiltration [101], high hydrostatic pressure [106], and drying [107,108]. Method-wise, the solvent has a profound effect on the number and type of polyphenols extracted. Polyphenol analysis must first identify the type of matrix to be analyzed, the chemical nature of the polyphenols of interest, and different solvents and solvent systems should be examined. The most appropriate solvent for the case in hand (i.e., maximizing compound diversity and yield) should be the one selected [109]. For example, Flores and coworkers resuspended the methanolic extract in hexane, chloroform, ethyl acetate, and *n*-butanol and reanalyzed each fraction. Ethyl acetate fraction exhibited the best results [110]. Finally, though polyphenols are usually related to health applications [111,112], antinutritional effects should be considered [109]. Some examples of polyphenol analysis are included in Table 4.
**Figure 1.** Polyphenols structure and classification [97]. Highly functionalized structures account for the molecules radical scavenging, metal ion chelating, and enzyme inhibition. Hydrogen bonding can stabilize phenoxyl radicals.
Gordon and coworkers used accelerated solvent extraction (ASE) to characterize polyphenolic compounds in *Psidium guineense* Sw., *Syzygium cumini* (L.) Skeels, *Byrsomina crassifolia* (L.) Kunth, and *Pouteria macrophylla* (Lam.) Eyma. [113]. ASE techniques allow for multiple extractions simultaneously. Swifter assays are obtained which, in turn, expedite research results and minimize solvent waste [114] when compared to common extraction methods (e.g., Soxhlet, sonication). Anton and coworkers investigated the effect of ripening in tomato polyphenols content and antioxidant capability. A differential mass spectrometry approach allowed the authors to conclude that cultivar-dependent patterns are observed during ripening (e.g., maximum concentrations of polyphenols achieved half-ripe stage) [115]. Radovanovi´c and coworkers, associated polyphenols from berries to antibacterial activity [116]. Veljkovi´c and coworkers analyzed phenolic compounds in different types of tea. Nettle/pineapple, and bearberry/raspberry teas showed the lowest and highest phenolic contents, respectively [117]. Mileti´c assessed polyphenols in dried and candied fruit. In this particular case, acid hydrolysis was applied to the previously dispersed methanolic extracts to free matrix-bound polyphenols [118]. One g *tert*-butyl hydroquinone/100 mL was added during extraction as a radical sink to protect polyphenols. Kowalska and coworkers used preparative chromatography to remove non-phenolics [98].
Tentative screening for *Psidium friedrichsthalianum* (Berg) Niedenzu pulp showed 1,5-dimethyl citrate, 1-*trans*-cinnamoyl-β-D-glucopyranoside, sinapic aldehyde-4-*O*-β-D-glucopyranoside, 1,3-*O*-diferuloylglycerol, and 3,3 ,4-tri-*O*-methylellagic acid-4 -*O*-D-glucopyranoside [110]. Phenolic compounds from pink guava from Costa Rica have been recently reported, *n* = 60 phenolic compounds were characterized. The authors report for the first time in *P. guajava n* = 42 compounds in the fruit's peel and flesh, and *n* = 24 new compounds, e.g., phlorizin, nothofagin, astringin, chrysin-C-glucoside, valoneic acid bilactone, cinnamoyl-glucoside, and two dimethoxy cinnamoyl-hexosides [119]. During polyphenol analysis, HLB® SPE (Hydrophilic-Lipophilic,
Balance Solid Phase Extraction) cartridges are used routinely for clean-up. At least one research group has applied this approach to assay polyphenols and vitamin C in plant-derived materials [121]. Interestingly, when using the Folin–Ciocalteu spectrophotometric approach, ascorbate is considered interference and must be eliminated from the eluate (usually taking advantage of ascorbate thermolability) or else the measurements are overestimated. However, simultaneous retention of both analytes in the SPE cartridge can be exploited, if HPLC methods are used instead. We recommend that in countries in which fruits with high polyphenol content are readily available (and in considerable quantities), preparative separation of polyphenol fractions is a possibility for obtaining pure compounds (See for example, [122]). Finally, vanillic acid was reported in cocoa pod polyphenol-rich extracts. Interestingly, the application of 2000 mg L−<sup>1</sup> of this cocoa extract to a vegetable oil improved its oxidative stability and shelf-life [123].
#### Method Application Experience
In our laboratory, ultrasound-assisted extraction is preferred for reducing processing time and avoiding degradation of the compounds. Additionally, polyphenols are quite light sensitive, hence yellow lights are used during the extraction using acetone-water or methanol-water solutions. As the polyphenol family is extensive and chemically diverse, a surface response design is always recommended to assess the appropriateness of the solvent system (i.e., selecting a solvent that provides the highest yields). Samples with a high lipid content (i.e., > 5 g total fat/100 g) usually cause significant interferences and must be defatted previous to polyphenol extraction. It is usual to add additional antioxidants (e.g., ascorbic acid) to polyphenol extracts to protect them from oxidation. Finally, it is common to find natural existing polyphenols as adducts with protein or carbohydrate moieties. These adducts are usually formed by non-covalent interactions (e.g., salt bridges); therefore, by adjusting the extract ionic strength, one can remove these artifacts. Sugar adducts are considerably more difficult to analyze since only a few compounds are commercially available (e.g., cyanidin 3-*O*-glucoside chloride). Hydrolysis (mild acidic, basic or enzymatic) is the usual approach to circumvent the lack of these commercial standards. Availability of mass spectrometry or nuclear magnetic resonance (NMR) can help elucidate unknown compounds and adducts.
#### *2.2. Capsaicinoids*
Capsaicinoids are plant metabolites from the *Capsicum* genus which give pungency to chili peppers [124]. Scoville scale which measures the spiciness of the fruits (originally, tested by sensory assays) is reported in function of capsaicin concentration (i.e., mg capsaicin kg−<sup>1</sup> × 16 [125]). Today, the most reliable, rapid, and efficient method to identify and quantify capsaicinoids is HPLC. Measurement of this molecules is significant as a quality measure of chili pepper (22 domesticated varieties consumed regularly worldwide), a crop which is of significant cultural and global trade market value [126]. More than 20 different capsaicinoids have been described; the foremost capsaicinoids found in these plant structures include capsaicin and dihydrocapsaicin [127] (Figure 2).
**Figure 2.** Chemical structures for (**A**) capsaicin (8-methyl-*N*-vanillylamide) and (**B**) dihydrocapsaicin (8-methyl-*N*-vanillylnonamide), the aromatic vanillyl radical is shown in red.
2.2.1. Measurement of Capsaicin and Dehydrocapsaicin in Real Samples
Research reports have described capsaicinoid analysis; the most recent are summarized in Table 5. Garcés-Claver and coworkers determined capsaicin and dihydrocapsaicin in two different scenarios,
i.e., fruits grown in summer and then in spring [128]. The authors concluded that capsaicinoids varied largely among fruit families and that these families did not respond similarly to producing these capsaicinoids when their fruits were grown in the two seasons tested [128].
**Table 5.** Common chromatographic conditions used for capsaicinoid analysis.
Goll and coworkers optimized a cyclic solid support free liquid–liquid partition to separate a capsaicin and dehydrocapsaicin mixture into two sequentially collected product streams. This approach may serve as a base for compound purification before chemical characterization. With this optimization, the authors demonstrated theoretical and predictive tools are useful in preparative chemistry and process design [129].
The pretreatment of capsaicinoid determination (i.e., extraction steps) is usually straightforward, and the majority of methods are based on methanol-based extraction. However, Lu and coworkers reviewed several techniques that can be used to extract capsaicinoids successfully [136]. Ma and coworkers [131] used capsaicin and dihydrocapsaicin, and nonivamide [132] were selected as adulteration markers to authenticate vegetable oils. No capsaicinoid compounds were found in edible vegetable oils, thereby ruling out a possible adulteration source. The authors prepared immunosorbents by covalently coupling highly specific capsaicinoid polyclonal antibodies with CNBr-activated Sepharose 4B and packed into a polyethylene column [131]. This research is interesting, from the clean-up standpoint, since the authors adjusted the major parameters affecting the immunoaffinity column extraction efficiency (i.e., loading, washing, and eluting conditions) [131]. Schmidt and coworkers compared different chili peppers available in Austria and compared their contents of capsaicin and dihydrocapsaicin [133]. The authors used UPLC (Ultra-Performance Liquid-Chromatography) and hence obtained a reduced resolved chromatogram for both compounds of just 1.7 min. [133]. The authors also corroborated that the highest capsaicinoids content was in the fruits' placenta and the seeds. Similarly, Sganzerla and coworkers obtained a complete separation under 4 min [134]. The above examples correspond to high-throughput methods of analysis.
Finally, ingested capsaicinoids can persist in the bloodstream and can be determined in plasma using LC coupled with tandem mass spectrometry [137]. Intestinal absorption and metabolisms (via capsaicinoid glucuronides) have also been reported for a mammal [138]. At the same time,
SPE: Solid phase extraction. UPLC: Ultra-Performance Liquid-Chromatography.
dietary capsaicin has been linked to the browning of adipose tissue, which in turn, promotes energy expenditure [139].
#### 2.2.2. Method Application Experience
As shown, capsaicinoids can very well be measured by using a wavelength in the 200–400 nm UV range. However, fluorescence analysis can be performed (λex 280 nm λem 338 nm) improving sensitivity dramatically [134], an approach preferred by our laboratory for routine analysis. A short column with a smaller particle size seems to improve both resolution and sensitivity.
#### *2.3. Caffeine and Theobromine*
Caffeine and theobromine are naturally occurring methylxanthines with antioxidant potential [140] (Figure 3). There are some misconceptions regarding health effects caused by caffeine ingestion [140]. On the contrary, theobromine (and cocoa) consumption has demonstrated beneficial effects [141]. Coffee, cocoa, tea, and caffeine-containing beverages (e.g., soft and energy drinks) are widespread and relevant food commodities. For example, caffeine intake has been calculated at 25 and 50 mg per day for children and adolescents aged 2–11 and 12–17 years, respectively. The more relevant caffeine sources were soda and tea as well as flavored dairy (for children aged < 12 years) and coffee (for those aged 12 years and above). Similarly, caffeine consumption has been between 2.5–3 and 400 mg kg−<sup>1</sup> bw (body weight) day−<sup>1</sup> for children and adults, respectively [142,143]. The evidence is suggesting an alimentary impact as some nutrients are poorly absorbed when combined with alkaloids [140]. Caffeine analysis is common in the food industry (e.g., quality control in beverages) and research (e.g., alkaloid carrying plants); it has also been incorporated in academia and student curricula [144].
**Figure 3.** Chemical structures for (**A**) caffeine (1,3,7-trimethylxanthine), (**B**) theobromine (3,7-dimethylxanthine), (**C**) theophylline (1,3-dimethylxanthine), (**D**) paraxanthine (1,7-dimethylxanthine), and (**E**) antipyrine (2,3-Dimethyl-1-phenyl-3-pyrazoline-5-one or phenazone). (**F**) Caffeine biotransformation pathway is dependent on the CYP1A2 and CYP2A6 enzyme system. **1**. 1,3,7-trimethylxanthine **2**. 1,7-dimethylxanthine **3**. 7-methylxanthine **4**. 7-methyluric acid **5**. 1-mthyluric acid **6**. 5-acetylamino-6-formylamino-3-methyluracil **7**. 1,7-dimethyluric acid **8**. 5-acetylamino-6-amino-3-methyluracil [145].
#### 2.3.1. Alkaloid Analysis and Reported Application to Real Samples
Several methods have been developed for alkaloid analysis in food samples. Also, methods for studying the fate of these alkaloids have been documented (Table 6). For example, Gruji´c-Leti´c and coworkers, analyzed 12 commercial tea and coffee products, non-alcoholic energy drinks and foods (including mate, green tea, and black tea), 5 combined preparations of over the counter non-steroid anti-inflammatories and water samples collected from 7 representative locations of the Danube River [146]. This paper represents a clear example of method versatility, as a single analyte was recovered, from variable matrices, and assessed using a similar procedure. This analysis was not only used for characterization, but also demonstrated a potential for quality control in commercial products (e.g., compliance of the nutritional label) and water. In water samples, the highest caffeine concentration found was 306.120 ± 0.082 ng L−<sup>1</sup> during springtime. Gonçalves and coworkers recently demonstrated that caffeine might be a suitable chemical marker of domestic wastewater contamination in surface waters [147].
**Table 6.** Summary of conditions regarding alkaloid analysis.
Shrestha and coworkers developed a method for use as quality control. Concentrations of Nepalese tea and coffee ranged from 1.10 to 4.30 mg caffeine kg−<sup>1</sup> dry basis [156]. Fajara and Susanti also determined caffeine in coffee beverages; they found 109.7–147.7 mg caffeine kg−<sup>1</sup> per serving [157]. Gliszczy ´nska-Swigło and Rybicka used both a photodiode and fluorescence detector to monitor both ´ caffeine and water-soluble vitamins, simultaneously, in energy drinks [148]. A¸sçı and coworkers analyzed caffeine in soft drinks [158]. The authors used Behnken response surface design to optimize HPLC conditions. Optimized variables included pH, 6.0, flow rate, 1.0 mL min−<sup>1</sup> and a mobile phase ratio, 95% [158]. Similarly, preservatives sorbate and benzoate also can be determined with caffeine simultaneously in sports drinks [149]. Ortega and coworkers compared data from HPLC- and UPLC-MS/MS (MS/MS also known as tandem mass spectrometry). The authors analyzed procyanidin
oligomers (mono to nonamers) and catechin, epicatechin, caffeine, theobromine. The analysis was performed under 12.5 min [150]. Recently, Rodríguez-Carrasco and coworkers used to analyze polyphenols and alkaloids in cocoa-based products. Mainly, they compared three different coffee varieties including "Forastero", "Trinitario", and "Criollo". Mostly, theobromine was found in major quantities relative to caffeine except Criollo 70 and 75% where the theobromine/caffeine ratio is ca. 1:1. Of all samples examined, Criollo varieties showed the highest quantities of alkaloids. [151]. Interestingly, a positive association has been described between cacao polyphenol absorption and theobromine [159]. Other identifying markers, such as fatty acids, have also been reported as tools for discrimination among coffee varieties. The authors were able to discern *Coffea arabica* (Arabica) and *Coffea canephora* (Robusta) using ∑MUFA, 18:3n3, ∑MUFA/∑SFA [160].
#### 2.3.2. Alkaloid Bioavailability and Transference to Biological Samples
Caffeine is rapidly absorbed following oral consumption; maximum blood (plasma) levels are usually reached within 30 min [140]. Caffeine bioavailability studies have been performed in human plasma, for example, Alvi and Hummami monitored caffeine and antipyrine (Figure 3). Caffeine in human plasma was stable for at least 24 h at room temperature or 12 weeks at –20 ◦C [153]. Caffeine is a demonstrated therapeutic agent for apnea of prematurity. Hence, López-Sánchez developed a method to monitor caffeine in serum to demonstrate that the drug had achieved its therapeutic levels (i.e., 30 or 35 μg mL−1) [161]. Cleanup using SPE adapted in multiple well plates, as the one used in the former study, is an easy way to process several samples simultaneously, instead of the one-on-one cartridge approach. Only in 85% and 78% of the cases studied, maternal and newborn absorption of caffeine was demonstrated, respectively. Another research group investigated caffeine metabolism based on CYP1A2 enzyme activity. The presence and ratio of theophylline, paraxanthine, theobromine, and caffeine (Figure 3) was evaluated in human saliva [154]. The authors collected saliva of healthy subjects after consumption of a caffeinated beverage and obtained data of compared chromatographic profiles from the saliva of smoking (active xenobiotic hepatic metabolism) and non-smoking subjects [154]. Saliva, plasma, and urine already have been demonstrated valuable to intervention studies for cocoa [155,162]. Kobayashi used differential chromatogram analysis to narrow the signal width for caffeine, in urine samples, to improve separation demonstrating that peak enhancing posterior to injection is possible [163]. Finally, Ramdani and co-workers incorporated green and black tea powder into bovine diets demonstrating that alkaloids, catechins, and theaflavins diminished ammonia and methane productions without any detrimental effect on rumen functions *in vitro* [51].
Although theobromine is not a usual analyte for feed analysis, is noteworthy that the 2002/EC/32 regulation sets limits for the analyte at 300 mg kg−<sup>1</sup> for compound feed, except for adult cattle feed, where the threshold is laxer (i.e., 700 mg kg<sup>−</sup>1).
#### 2.3.3. Method Application Experience
Tea and coffee sample extracts are rich in tannins and other non-desired compounds that may generate matrix effects and reduce the shelf life of an analytical column. We have successfully used MgO to remove said interferences while increasing the extract pH. An alkaline medium ensures positively charged alkaloid molecules. Furthermore, defatting is vital for an adequate recovery when a lipid-rich sample is treated (e.g., cacao seeds, >30 g total fat/100 g), especially, if aqueous extracting is employed. We suggest the use of efficient organic solvents; *n*-hexane, petroleum benzine, for example, have been exploited. Minimal amounts possible should be used, as this otherwise generates waste. Chlorinated solvents and ethyl ether should be avoided, as alkaloids exhibit some degree of solubility in these solvents which, in turn, may affect recovery.
#### *2.4. Cholesterol*
Cholesterol ((3*S*,8*S*,9*S*,10*R*,13*R*,14*S*,17*R*)-10,13-dimethyl-17-[(2*R*)-6-methylheptan-2-yl]2,3,4,7,8, 9,11,12,14,15,16,17-dodecahydro-1*H*-cyclopenta[a]phenanthren-3-ol), is a waxy steroid metabolite found in the cell membranes and transported in the blood plasma of all animals [164]. This sterol plays a role in metabolic (e.g., precursor for bile acids and steroid hormones) and structural processes (e.g., regulates biological membrane fluidity) [165,166]. Cholesterol can be introduced to the metabolism through *de novo* synthesis or diet [162]. In plant structures, similar compounds are found such as phytosterols and stanols [167]. However, when analyzing cholesterol, one must consider that the amount of cholesterol made by many plants is not negligible [168]. Nutritional information regarding cholesterol content in food and intake through dietary sources is relevant, as overload can drastically increase plasma cholesterol levels and, hence, health risks. From a methodological standpoint, a considerable advantage in using the LC approach is that lipid oxidation is negligible, as measurements can be performed at relatively low temperatures. Herein are detailed some examples of cholesterol analysis in food samples (Table 7).
Albuquerque and coworkers compared both HPLC and UPLC for the analysis of eggs, egg yolks, sour cream, and chicken nuggets. The latter approach rendered a method with 8-fold less solvent waste and ca. 4-fold more sensitivity, with a decreased analysis time (i.e., 4 min) [166]. The initial sample mass used from the assay was optimized; 0.25 g and 1 g for samples with relative lower (e.g., sour cream) and higher (e.g., egg yolk) cholesterol contents. The authors also compared different cooking methods for the chicken nuggets (baked vs. deep frying). They found that cholesterol content was higher in the oven baked goods. This is a result of the processing as the meat loses water during baking. Meanwhile, water/oil exchange occurs during frying. Although several solvents were tested, the authors concluded that an acetonitrile/2-propanol solvent system was the most successful in eluting the cholesterol molecule [166]. Cholesterol analysis usually renders clean chromatograms since most interferences are eliminated by saponification. Saponification segregates the molecule of interest from the saponifiable lipid fraction (e.g., acylglycerols) and hydrolyzes cholesterol esters. This step has been considered critical for cholesterol analysis in food matrices [166]. Furthermore, Cruz and coworkers, quantitatively, compared several extraction methods on freeze dried and thawed seafood samples [169]. In this regard, the direct saponification and extraction considerably reduce solvent waste, while the Smedes method used non-chlorinated solvents (is a greener approach). Better recoveries for vitamin E are obtained when the analysis is performed before saponification step (e.g., modified Folch, Smedes). The authors were able to analyze α-tocopherol, cholesterol, and fatty acids all from the same extract and applied the optimized method to octopus, squid, mackerel, and sardine successfully. From the assayed samples, squid and sardine showed higher values of cholesterol and vitamin E, respectively. Interestingly, normal phase chromatography was used to assess vitamin E [169]. Saldanha and Bragagnolo also used normal phase chromatography. The authors used very mild conditions during saponification, which are paramount to avoid cholesterol oxidation. Also, they monitored cholesterol contents after heat treatment and demonstrated that it decreased significantly, with a simultaneous increase of the cholesterol oxides contents (i.e., 19-hydroxycholesterol, 24(*S*)-hydroxycholesterol, 22(*S*)-hydroxycholesterol, 25-hydroxycholesterol, 25(*R*)-hydroxycholesterol, and 7-ketocholesterol) [170]. Bauer and coworkers analyzed cholesterol and cholesterol oxides in milk samples using reversed-phase chromatography. [171]. The presence of cholesterol oxides can indicate the source and nature of the food, as well as the storage and processing conditions suffered by a commodity. The authors conclude that milk has physicochemical characteristics that make it more resistant to oxidation of cholesterol compared to other products of animal origin. In this regard, several sample preparation methods for cholesterol oxides have been detailed elsewhere [173]. Daneshfar and coworkers used dispersive liquid–liquid microextraction as an alternative to the extraction and clean-up steps in sample preparation [172]. In this case, ethanol was used as a disperser solvent and carbon tetrachloride as an extraction solvent [172]. This work is a fine example of parameter optimization during method validation; different dispersion (i.e., EtOH, acetone, and ACN)
and extraction (i.e., CS2, CH2Cl2, CHCl3, and CCl4) solvents were tested, as well as variables such as pH, volume and time. However, the authors fail to explain how they obtain total cholesterol from a complex matrix (for example, a method must be able to free cholesterol from its esterified form) when no hydrolysis is performed (i.e., ensuring not just the mere quantification of unbound/free cholesterol).
**Table 7.** Measurement techniques meant for cholesterol in food samples.
It should be pointed out that though the chlorinated solvents are used in very small quantities, they are still classified by the IARC (International Agency for Research on Cancer) as possible human carcinogens (group 2B). Finally, Robinet and coworkers used a cholesterol esterase in an unrelated matrix to avoid chemical saponification [174]. In this regard, cholesterol esterases (most active at pH 7.0, 37 ◦C, and in the presence of taurocholate) and lipases (most active at pH 7.7, and 37 ◦C [175]) are commercially available.
#### Method Application Experience
We suggest two major points: i. that it is recommendable to perform the saponification first and then the solvent-aided extraction ii. a response surface design may be useful to optimize the length of the saponification treatment.
#### **3. Determinations Designed for Feed and Feed Ingredients**
#### *3.1. Mycotoxins*
#### 3.1.1. Recent Approaches for the Determination of Mycoxotins in Feeds
Mycotoxins are secondary metabolites mainly by fungi *Aspergillus, Penicillium, Fusarium* and *Alternaria* species, in stress situations, which involve changes in temperature, moisture or pH in plants [58,176,177]. Currently there are more than 400 types of mycotoxins as ubiquitous contaminants in a wide variety of foods [178,179], such as, corn, cocoa, sorghum, wheat, oats, rye, cotton, peanuts, coffee, dairy products, eggs, among others [180]. Among the best known are ochratoxin (OTA), zearalenone (ZEA), trichothecenes, aflatoxin B1 (AFB1), fumonisin B1 (FB1) and their metabolites. The last two are listed as carcinogenic by the IARC [181]. Mycotoxins, in general, are teratogenic, mutagenic, carcinogenic, and can possess an immunosuppressive effect in both animals and humans [178,182], which can be aggravated by factors such as the animal species, the concentration of the toxin and synergism existing among them, in addition to the health and nutritional status of the animal [182,183]. Also, the direct effects on health, including decreased weight gain, feed conversion
inefficiency, reduced production, and a decrease of the food system profitability, the increase in feedstuff costs, medical treatments, and ineffectiveness when exploiting the genetic potential of animals [183].
At an organ level, in the liver, AFB1 can generate several metabolites, which include aflatoxin M1 (AFM1), which is transferred to milk, a complete food nutritionally, and which is vital in the development of the first years of life [184,185]. Also, the AFM1 is a compound declared as a carcinogen that is very resistant to pasteurization and freezing [180,183]. Therefore, being trawl compounds in the trophic chain, which involve the adverse effects on livestock production, with an obvious risk to the health of consumers, it stresses the need for laboratories to possess the ability to analyze a large number of analytes in a single sample. In this way, the amount of information can be increased, and a wider diagnosis can be made about the safety of the food and feed industry.
In this regard, Table 8 shows a summary of methods developed for the identification and quantification of mycotoxins, by different research groups, focused mainly on animal feed. For example, Njumbe Ediage and coworkers developed a technique capable of determining 25 mycotoxins in cassava meal, peanut cakes, cornmeal, and different sorghum varieties. The most exciting thing, in this case, is how the researchers solved the affinity fact of fumonisin and ochratoxin with the amino groups (due to the presence of carboxylic acid moiety, Figure 4) [177,186]. The researcher divided their extract into two portions, one to which formic acid and dichloromethane were added. After cleanup, the two independent shares were remixed evaporated at 40 ◦C, reconstituted with MeOH/H2O/CH3COOH, and 5 mmol L−<sup>1</sup> CH3COO<sup>−</sup> NH4 +. During MS-based mycotoxin separations, flows are usually kept low, so solvent nebulization and evaporation are performed swiftly. The mobile phase is generally accompanied by an acetic or formic acid buffer to improve ionization especially for those compounds without readily ionizable functional groups (e.g., aflatoxins). Also, the formate ion is added in both solvents as one solvent depletes during the gradient separation and the buffer must always be present in a similar proportion [177,186]. Dzuman and coworkers and Rasmussen and coworkers, used, as an extraction method, a modification of the QuEChERS method, (Quick, Easy, Cheap, Effective Rugged, and Safe usually used for pesticide analysis). Both research groups coincide that QuEChERS adaptations for mycotoxin analysis open the possibility toward the simultaneous assay of several and distinct groups of contaminants (e.g., pesticides and mycotoxins) [179,187].
**Table 8.** Measurement techniques meant for mycotoxins in feed samples.
**Figure 4.** Chemical structures for (**A**) ochratoxin A, (**B**) ochratoxin B, (**C**) ochratoxin C, blue colored circles represent changes in the structure between ochratoxins, loss of Cl and OH in ochratoxin B and C respectively render a more lipophilic molecule. Et = C2H5, and (**D**) are the general backbone of Fumonisins. FB1 = 721.83 g mol−<sup>1</sup> R1:HR2: OH R3: OH; FB2 = 705.84 g mol−<sup>1</sup> R1: OH R2:HR3: OH; FB3 = 705.84 g mol−<sup>1</sup> R1:HR2:HR3: OH; FB4 = 689.84 g mol−<sup>1</sup> R1:HR2:HR3: H. Functional groups colored in green and red represent a positively and negatively ionizable moiety, respectively.
#### 3.1.2. Agricultural by-Products as Feed Ingredients
Agricultural and food-industry residues are valuable to animal nutrition as they are rich in many bioactive and nutraceutical compounds, such as polyphenolics, carotenoids and dietary fiber among others [190]. Agro-byproducts, used in animal feed, originate from perishable crops and, as such, are susceptible to fungal infection [191]. Hence, mycotoxin surveillance of these materials contemplating the most common contaminants present in such matrices, but also considering emerging contaminants (e.g., beauvericin, enniantins, and fusaproliferin) [191,192] is paramount. The food industry generally includes practices that guarantee the safety of the product meant for human consumption. Residues destined for animal production may not be subject to the same scrutiny. For example, the wine industry with a production estimated at 27 million liters worldwide. Presence of OTA in wine has been widely investigated [193]. However, with the development of new methods, it has been possible to find up to 36 different mycotoxins. (See for example, [194]).
Countries where the production of wine is the predominant, compared to other types of industry, a considerable amount of waste must be repurposed. As such, this might be of use as a ruminant (such as cows and goats) feed ingredient, where the pulp, husks, and seeds, might offer to the animal diet: fiber, energy, fatty acids, and antioxidant compounds which improve ruminal health, echoing in the quality of meat and milk [195–197]. As yet another benefit from this waste processing, the use of grape seeds as mycotoxin adsorbents has been investigated both *in vitro* [198] and *in vivo* (e.g., pigs [199]).
#### *3.2. Antibiotics*
#### 3.2.1. Recent Multiresidue and Multi-Class Analysis of Antibiotics in Feeds
Antibiotics are bioactive substances used against bacteria as a therapeutic, metaphylaxis or prophylactic agent both in humans and animals [200–202]. In livestock, some antibiotics are included in animal diets as growth promoter (e.g., monensin, narasin, ractopamine), decrease feed conversion, improve feed efficiency, and overall cost-effectiveness of animal production systems [203,204]. Overuse of veterinary pharmaceuticals in livestock, aquaculture, and the feed industry is reflected in the incidence of residues found in animal-derived food products (e.g., meat, eggs, milk, and honey) [201,205–207]. Antibiotic biotransference through the food chain may contribute to allergic reactions, mutanogenic and cancerogenic effects, found in humans and animals; additional to the growing rates of antimicrobial resistance [208,209]. Considering these issues, organizations worldwide (e.g., European Commission, United States Food and Drug Administration, World Health Organization) have generated protocols that help control, regulate and surveil the use of antibiotics in food-producing
animals [208,210–212]. Hence, similar to mycotoxins, development of analytic methods that allow for identifying and quantitating a broad spectrum of compounds from a sample, directly contributes to surveillance programs for feedstuff manufacturing (raw materials or feed ingredients, compound feed, and premixes) and, similarly, those commodities derived from food-producing animals.
Table 9 shows a summary of the different characteristics of validated methods for the identification and quantification of veterinary antibiotics in different types of matrices. Molognoni and coworkers, optimized a method for the determination of spectinomycin, halquinol y zilpaterol in compound feed demonstrating once again the capabilities of mass spectrometry to assess two or more families of seemingly unrelated compounds. The authors tried both hydrophilic interaction and reverse-phase chromatography. Though HILIC (Hydrophilic Interaction Liquid Chromatography) offered good results, it requires a longer analysis time (i.e., up to 5 additional min), and is pH sensitive. Reverse-phase chromatography requires a relatively inexpensive column that is usually available in laboratories and which analytical instrumentation providers generally keep in stock. Additionally, a more effective separation was archived using heptafluorobutyric acid in the mobile phase [202].
**Table 9.** Measurement techniques meant for veterinary antibiotics in food and feed samples.
ACN: Acetonitrile.
#### 3.2.2. Multiresidue Analysis of Antibiotics in Foods
Barreto and coworkers developed a method to assay *n* = 14 different coccidiostats (i.e., lasalocid A, maduramicin, monensin, narasin, salinomycin, semduramicin, robenidine, diclazuril, toltrazuril, trimethoprim, chlopidol, amprolium, diaveridin y nicarbazin) in poultry muscle tissue and eggs; after testing several chromatographic columns, they selected the one that completed the separation under less time (i.e., 14 min). The authors used low temperature clean-up as an alternative to SPE, reducing costs, time and ion suppression. Internal standards where used to compensate intense matrix effects [207]. Regarding aquaculture, Kang and coworkers analyzed *n* = 41 antibiotics in fish muscle [205]. Similarly, Kumar Saxena and coworkers developed and validated *n* = 24 antibiotics (including quinolones, sulfonamides, and tetracyclines) in shrimp, and they preferred to use methanolic separation [206]. Finally, Shendy and coworkers identified *n* = 6 different classes of antibiotics in honey with a modified QuEChERs procedure simultaneously. Extraction was performed using ACN and MgSO4 and NaCl [214].
For both mycotoxins and antibiotics, a review was made of the wide variety of methods used in the food industry for the simultaneous, extraction of multiple analytes. For the identification and quantification of each chemical, a sensitive and selective tool is required. It is here that mass spectrometry has been useful, by reducing costs and response time. [185,202,209].
#### 3.2.3. Method Application Experience (Mycotoxins and Antibiotics)
A multitoxin (*n* = 26) analysis was applied to feedingstuffs using, as a reference, a method previously described by Wang and coworkers in cornmeal. ACN/CH3COOH/H2O (74:1:25) was used for extraction and cleanup we exploited the versatility of HLB cartridges (which allow the retention of a wide array of analytes with the least of interferences) [215]. When compared with immunoaffinity columns, this sorbent is less prone to fracturing and do not require low temperatures for storage. Later, the recovered extract was evaporated to dryness using vacuum at 60 ◦C and reconstituted with MeOH. The method relies on the 12.5-fold concentration of the original analyte to improve sensitivity. In the case of antibiotics (*n* = 23), we based our procedure on that described by Duelge and coworkers [216]. We extracted and eluted analytes using an ACN/MeOH solution. Again, we trusted the versatility of HLB SPE cartridges during cleanup. Both assays were single quadrupole equipped LC system using ESI<sup>+</sup> and relied on a reverse phase separation (Zorbax Eclipse Plus, <sup>100</sup> × 3 mm, 3.5 <sup>μ</sup>m). For mycotoxin separation, the mobile phase consisted of a gradient using acidified (0.1 mL/100 mL formic acid) ACN and H2O. For antibiotics, the gradient consisted of three different acidifed solvents ACN, H2O, and MeOH. In our experience, the first two-solvent gradient (starting with water) can separate most antibiotic families (β-lactams, tetracyclines, macrolides, streptogramins, lincosamides, aminoglycosides). Our gradient finishes with MeOH which is the only solvent capable of eluting coccidiostats (e.g., monensin and narasin). Efficient chromatographic separation was achieved under 35 min.
#### *3.3. Amino Acids*
Protein building blocks (i.e., amino acids), biologically, can be separated into two main groups. Exogenic/essential amino acids (i.e., Arg, Phe, His, Ile, Leu, Lys, Met, Thr, Trp, and Val), are not synthesized by the organism and must be provided in the diet to cover the requirement. The remaining amino acids are endogenic (i.e., Ala, Cys, Asp, Glu, Pro, Ser, Tyr, and Gly). Several of these amino acids (e.g., Lys, Met, Thr, and Trp) are prepared synthetically and are commercially available to use as feed additives. The purity of these additives must be routinely checked and adequately verified. Hence, methodological development is paramount for quality control for determination of amino acids in feed materials and feed mixtures. However, few reports have focused on feed. As a result; we intend to give an overview of the methods available in related matrixes.
#### 3.3.1. Fish Tissue
In a comprehensive research article, Mohanty and coworkers reported the complete amino acid profile (except tryptophan which was assessed spectrophotometrically and basic hydrolysis) for 27 different food fishes. [217]. Derivatization was performed using 6-aminoquinolyl-*N*-hydroxysucciminidyl carbamate (AQC), this specific reagent requires neutral pH to work. Adduct formation has the advantage of being stable and reacting with secondary amines. No variability among profiles was found in fishes of the same species from different locations. They also related the concentration of the amino acid found in the fish with the environment in which they live (e.g., marine and cold-water fishes showed relatively higher amounts of Met). At the same time, they recommend the consumption of certain fish species for several amino acids dietary deficiency in humans [217]. Example of methods suitable to analyze amino acids in diverse matrixes is shown in Table 10.
**Table 10.** Sample pretreatment, derivatization and measurement conditions for amino acids in feeds.
#### 3.3.2. Filamentous Cyanobacteria, *Spirulina* sp.
*Spirulina* sp. is a filamentous cyanobacterium that have been recognized for its nutritional value as a feed ingredient and supplement, and has been related to health benefits in humans [230]. Its nutrient profile has been reported previously, and it has even exhibited a higher amino acid value (except for Lys, Glu, Pro, His) when compared with that of soybean meal (a staple feed ingredient) [218]. Additionally, based on this profile, they calculated energy for a broiler diet. Nurcahya Dewi and coworkers applied different physical treatments (i.e., drying, sonication 30/60 min, reflux 60/90 ◦C, maceration in MeOH) to determine their effects on *Spirulina* sp. amino acid profile, which they concluded is rich in amino acids related to umami flavor (i.e., Asp and Glu). Drying and methanol maceration showed to be the treatment that delivered the highest (8.37 g/100 g) and lowest (2.34 g/100 g) contents of Glu, respectively [231].
Campanella and coworkers assayed total and free amino acids from *Spirulina* sp.; they found that freshwater *Spirulina* contained relatively high concentrations of non-essential amino acids. The authors indicate that the samples tested were lysine-rich and limited in sulfur-containing amino acids. Free amino acids constitute as high as 2% of the amino acid input. Method-wise, the authors used an oxidation-capable acid, this is chancy as it may contribute to analytes deterioration. Additionally, the mobile phase already included the derivatization agent [219].
Al-Dhabi and Arasu quantified polyunsaturated fatty acids, sugars, polyphenol and total and free amino acids in *Spirulina* sp. In contrast to the authors mentioned above, this research group used pre-column derivatization and a dedicated column for analysis. Total amino acids contents ranged from 11.49 to 56.14 mg/100 g; from which essential amino acids accounted for 17.00 to 39.18%. [220].
#### 3.3.3. Compound Feedstuff
For the specific case of feed, a time-reduced (i.e., complete separation in an eight-minute chromatographic run) analysis has been recently developed [225]. AOAC OMASM includes two different assays to determine amino acids based on LC; 992.12 design for pet foods using fluorescence and 999.13 include ninhydrin/Orto-phthalaldehyde (OPA) fluorescence or pulsed coulometric detection. Finally, a report made by Wang and coworkers described a successful simultaneous analysis of 20 amino acids without using derivatization using an evaporative light scattering detector [232]. More recently, underivatized amino acids have also been monitored using hydrophilic interaction liquid chromatography coupled with tandem mass spectrometry [233]. Herein, we included some examples of derivatization agents. However, we suggest the reader access a paper written by Masuda and Dohmae, which not only cites the four most commonly used reagents for amino acid derivatization, but also identifies their strengths and weaknesses [234].
#### 3.3.4. Bacterial Cell Walls, Peptidoglycan, and Food-Extracted Peptides
A less common application for LC, is to monitor the products from the hydrolysis of bacterial cell walls (using enzymatic physical, and chemical approaches) and posterior fragment analysis. Desmarais and coworkers design a method that included the digestion of Braun's lipoprotein. Muropeptide fragments (monomers-trimers), 3,3-diaminopimelic acid among others [226]. Kühner and coworkers developed a similar application; complete muropeptide hydrolysis was accomplished within 24 h. UPLC/MS was used to monitor fragments. After BH4 − reduction, both Gram-positive to Gram-negative bacteria can be evaluated after gradient modification [227]. In this regard, MSD (Mass Spectrometry Detectors) serve as a good reference for additional mass information, which will ease the peptidoglycan *in silico* reconstitution. This application has not found accommodation in food or feed, but it can correctly be adapted for bioreactions/fermentations or lactic bacteria.
Other applications for LC include, for example, the work by Marseglia and coworkers. They identified *n* = 44 different peptides from cocoa beans. The peptide fragmentation pattern in fermented cocoa samples was used to describe the geographic origin, different fermentation levels, and roasting. Vicilin, a storage protein, was identified in cocoa bean samples, information that can be useful to understand the biological activity of cocoa and to determine the aroma relevant peptides [228]. MS assisted analysis is advantageous as amino acids lack any distinctive chromophores and already have readily ionizable moieties. Prados and coworkers recently have described a method to isolate, characterize and identify peptides that can downregulate adipogenesis. The authors also used semipreparative fractionation to achieve the initial peptide separation [229].
#### 3.3.5. Method Application Experience
When facing fresh feed products (e.g., wet pet food, forages) additional operation units such as lyophilization is necessary before sample treatment (see, for example, [235]). To obtain individual amino acids, most applications require acid or alkaline hydrolysis. However, amino acids are extremely susceptible to oxygen during hydrolysis, to prevent quantitative losses, we recommend the sample hydrolysis steps suggested elsewhere for furosine [236]. Additionally, pyrogallol in 1 g/100 mL is also used as a radical receptor (i.e., a radical sink) to avoid amino acid degradation. Particularly, Trp, Thr, and Tyr are usually lost during acid hydrolysis, cysteine is oxidized to cysteic acid, and asparagine and glutamine (if present) will transform to their respective acids. Hydrolysis may be performed using a feed of known concentration in parallel as a reference [237].
From the sample preparation standpoint, we have applied a Supelco ENVI-Carb SPE cartridge for cleanup as hydrolysate retain undesired particulates. A translucent extract is obtained after SPE that will be suitable for both FLD (Fluorescence Detector) and UV-Vis (Ultraviolet-Visible) detection-based analysis. Also, cleaner chromatograms are obtained as interferences are significantly reduced. This cartridge will adsorb (including those responsible for coloration) a great range of molecules, while the (charged) amino acids will not be retained. Sodium azide is applied routinely for extract storage. However, best results are obtained when measurements are performed immediately after preparation steps.
We have used a method based on OPA pre-column derivatization adapted from an established method for biopharmaceuticals [238]. We also recommend the strict use of a C18 guard column to increase column lifespan. When applied to feed samples and feed ingredients, essential amino acids covered include Arg, His, Ile, Leu, Lys, Met, Phe Val, and non-essential Ala, Asp/Asn, Glx, Cys/CY2, Gly, and Ser for a total of 14 amino acids. OPA derivatization is only effective under alkaline conditions (usually performed using borate buffer pH 8–10). Therefore, the feed hydrolysate must be neutralized (to pH 7.0) before injection, as the buffer will not be able to compensate for the [H3O+] that results from the acid treatment. Furthermore, 9-fluorenylmethyl chloroformate (FMOC) must be included during derivatization (additional to OPA) to obtain proline and hydroxyproline amino acids (see, for example, [224]).
Method automatization (when an automatic sampler is available) can concede an advantage since the reaction occurs in situ within the needle. Automated precolumn derivatization is also useful for unstable adducts (e.g., OPA derivatives). A benefit of amino acid derivatization is that most adducts can be monitored using a UV/VWD (Ultraviolet/Variable Wavelength Detector) or DAD/PDA (Diode-Array Detection/Photodiode-Array Detection), so even if the fluorescence detector is not available, analysis can still be performed. Though, the fluorescent detector can filter interferences, begetting cleaner chromatographs. We have also used the same method to assess the purity of feed grade amino acids, and taurine. The technique can also be applied to energy drinks to evaluate taurine in as a very simple "dilute and shoot" method after sonication for sample degassing. Interestingly, ninhydrin and OPA can detect complementary analytes to methods based in ninhydrin (see, for example, [223]).
#### *3.4. Triphenylmethane Dyes*
Malachite green is a dye usually used in aquaculture as a fungicide and antiparasitic due to its low cost and effectiveness [239]. The widespread use of this substance is not without downsides, though, including residue accumulation in fish tissue and contamination of sediments and water bodies, which can affect non-target organisms downstream (see, for example, [240,241]).
Recent and improved methods have found acceptance to monitor these kinds of dyes in fish tissue. For example, Hidaya and coworkers already conducted a short review on techniques available for detection of malachite and leucomalachite green in the fish industry [242].
Within, this paper, several LC-based techniques are mentioned. Triphenylmethane dyes suffer from reversible redox reactions; each form can be oxidized or reduced to one another (see, for example, [243]; Figure 5).
**Figure 5.** Chemical structures for three triphenylmethane dyes which are sharing a common phenyl backbone sharing a methylidyne. Each molecule has extended π-delocalized electrons justifying their crystal coloration and visible light absorption (ca. 621 nm for malachite green).
Table 11 shows a summary of various methods for the extraction and identification of malachite green and its metabolites. Although it is a common contaminant in aquaculture production, and research focuses on fresh residues from aquaculture production animals (fish, shrimp, lobster, among others), the development of methods should also be extended to the analysis of feed [244] as fish and shrimp feed are made from marine by-products. Doses on fish or shrimp range from 0.05–0.2 mg L−<sup>1</sup> as an active ingredient have been used. Treatments for fish eggs include dosages of 5 mg L−<sup>1</sup> is usually suggested for fish tanks. Laboratories usually measure malachite green with equipment able to detect tissue residues below 2 μg kg−<sup>1</sup> (maximum permitted residue limit in fish tissue; [250]). A very interesting approach was made by Furusawa who developed a green chemistry method for malachite green and its metabolite [251].
**Table 11.** Measurement techniques meant for triphenylmethane dyes.
As previously mentioned, Wang and coworkers used solid-phase microextraction with the excellent result to assess malachite green, crystal violet and their respective metabolites using a monolithic fiber [245]. Bae Lee and coworkers homogenized fish tissue samples, and the extracted residues were partitioned into dichloromethane and an in situ oxidation with 2,3-dichloro-5,6-dicyano-1,4-benzoquinone. Afterward, cleaned-up was performed on neutral alumina and propyl sulfonic acid cation-exchange solid-phase extraction cartridges. Malachite green and crystal violet were determined at 618 and 588 nm using HPLC-Vis detector [246]. A common approach included analyzing dyes using traditional detectors and adding a step that included confirmation by MS. Chengyun and coworkers relied on Oasis® MCX (a strong cation exchange-based adsorbent) to perform clean-up. After a two-step, QuEChERS extraction, dispersive solid phase extraction coupled with, both, a reverse phase and strong anion exchange (as well as a mixed mode adsorbent) cleanup was tested. Residues of the dyes were evaluated in codfish [247]. However, we do not see how anion exchange favors dye-stationary phase interaction, since all parent compounds are positively charged. Noteworthy, usually reverse phase columns can resolve these types of dyes with ease, even if several analytes are to be evaluated simultaneously. Croatia and Iran are specific examples of countries which have stated have found residues of this dye in fish tissue [252,253]. Both cases demonstrate the need to assess these compounds in food items. However, both research groups used immunoassays to evaluate the contaminant. AOAC method OMASM 2012.25 is a reference based on LC-MS/MS to assess triphenylmethane dyes and their metabolites in aquaculture products.
Additionally, US FDA reference method is based on the isolation of malachite green using alumina/propyl sulfonic solid phase extraction cartridges previous to Non-Discharge Atmospheric Pressure Chemical Ionization and an LC-MSn; quantification was performed in salmon [248]. Finally, since fish and shrimp compound feed can also be based in aquaculture by-product meal, as a source of protein, contaminated tissue can reach the final product. Hence, the need for feed analysis is evident, as it shows, Abro and coworkers [249].
#### **4. The Common Ground among Measurements Performed in Food and Feed Laboratories**
#### *4.1. Nitrates and Nitrites*
Nitrates and nitrites are natural compounds that are part of the nitrogen cycle, but especially high dosages of these ions are registered because of anthropological activities [254,255]; they enter human diets by means of drinking water, leafy vegetables, and cured meats. Noteworthy, these ions have been authorized as additives in several countries including the European Community [256,257].
Though there is evidence that both ions have a relevant biological and physiological function, special attention has been paid to nitrates and nitrites and their metabolites such as *N*-nitrosamines and nitrous oxide as all these molecules may pose a health hazard [256,257]. For example, these compounds have been related to colorectal cancer [256–261]. Hence, risk management and assessment in food have been proved necessary [258]. Regarding the quantification of NO2 − and NO3 − using HPLC, there are two main approaches used i.e., ion exchange and reverse phase columns (Table 12).
#### 4.1.1. Ion Exchange Chromatography
When analyzing crops, one must consider that cultivar, and harvest date can affect the nitrate levels of selected vegetables. Hence, maximum levels have been set by European legislation accordingly [262]. For example, Brki´c and coworkers analyzed several leafy greens (*n* = 200) in two different seasons, in order to evaluate differences in ion content and encountered considerable differences among vegetable and sampling season [263]. Pardo-Marin and coworkers assessed vegetable-based baby foods, considering the levels found within these types of foods. They calculated ion ingestion between 13–18% of the acceptable daily intake for an infant. [264]. Quijano and coworkers assessed the nitrate content of vegetables (*n* = 533); they obtained values up to 3509 mg kg−<sup>1</sup> in chard samples. They calculated an intake of 490 mg kg−<sup>1</sup> bw day−<sup>1</sup> for a young population, values which tend to increase the risk of exceeding acceptable intake values [265]. The main advantages in using ion exchange columns are that the separation can be accomplished using aqueous buffers which are made up from relatively cheap salts, making the methods apt for green chemistry and avoid mobile phase drift [263].
#### 4.1.2. Ion Pairing and Reverse Phase Chromatography
Tetrabutylammonium salt has also been used as an ion-pairing agent coupled with reverse phase columns (Table 12). For example, Hsu and coworkers used a reverse phase approach to assess both ions in cured meats and vegetables. The authors found the highest values of NO3 − in spinach (4849.6 mg kg−1) and for NO2 <sup>−</sup> in hot dogs (78.6 mg kg−1) [266]. Nitrite tends to oxidate to NO3 −, the authors cite several factors affecting nitrate and nitrite recovery in foods (e.g., temperature, pH, metals). Usually, non-desired compounds found in greens differ from those found in meat products, for which proteins interfere significantly. Meat sample extracts will need pH adjustments and higher temperatures are needed to improve recovery. Some of these parameters must be monitored during analysis, especially when vegetables are subject of study [266]. Croituru used a similar approach to assess human, rabbit, rat urine as well as vegetables. However, they roduced adducts (an azo dye, HO3SC6H4–N=N–C10H6NH2) based on Greiss reaction (sulfanilic acid form a diazonium cation (HO3SC6H4–N≡N+) with NO2 − and then with 1-naphthylamine) for NO2 − that was measured at 520 nm [267,268]. Interestingly, the authors followed the reaction with mass spectrometry. We encourage the reader to pay special attention to this paper as highlights difficulties during method development. The author concluded that while useful, the use of Greiss reaction, spectrophotometrically, is unadvisable as several samples may exhibit additional confounding compounds that may behave similarly as the NO2 − ion adduct. However, is quite valuable as a derivatizing agent when coupled with HPLC; the method can work with samples of different origins without the need for further modifications [267]. Samples were decolorized with carbon and ZnSO4 was applied for protein precipitation to overcoming this matrix interference and enhance the sensitivity. Croituru and coworkers used a validated method to assess NO2 − and NO3 − in vegetables for self-consumption; toxicologically speaking, the NO2 − content found in the samples was deemed too low to represent a hazard [269].
Stationary phases containing only alkyl chains have been used, but it is also possible to find mixed stationary phases, for example, Abdulkair and coworkers assayed NO2 − and NO3 − using a stationary phase containing both alkyl groups and phenyl groups (Table 12) to separate both ions successfully after sonication [270].
Chou and coworkers assessed both ions in vegetables and observed a high concentration variability was observed which reflect differences in environmental conditions [271]. The authors also optimized critical chromatographic parameters such as pH, organic solvent fraction, and flow [271]. In this regard, the methanol fraction optimization was demonstrated to be paramount to improve octylammonium solubility and achieve an optimal resolution between both ions. In contrast, pH and flow variations tend to have an effect only on chromatographic run times and not so much in resolution.
**Table 12.** Common chromatographic approaches for the determination of nitrate and nitrite ion.
#### 4.1.3. Miscellaneous Methods for Nitrates and Nitrites
In contrast with ion pairing approaches, dos Santos and coworkers developed a method based on the reaction of the NO2 − with 2,3-diaminonaphthalene to yield a highly specific fluorescent 2,3-naphthotriazole adduct (λex 375 λem 415 nm), under acidic conditions, to assess the ions in beetroot [274]. Cassanova and coworkers have developed an application for HPLC derivatization based on VCl3, 4-nitroaniline, methanesulfonic acid, and *N*-(1-naphthyl)-ethylenediamine. Under these conditions, a post-column reduction of nitrate to nitrite can be accomplished [275].
#### 4.1.4. Method Application Experience
The preferred methodology used in our laboratories is based on the chromatographic determination of NO2 − and NO3 − anions simultaneously. Reverse phase (using a C18 column, i.e., Zorbax Eclipse 5.0 <sup>μ</sup>m, 4.6 mm × 150 mm, set at 30 ◦C and 0.6 mL min−1) HPLC-PDA or -VWD (213 nm as the absorption spectra maximum) is sufficient to perform the assay [266,271]. It is important to emphasize that for the detection and separation of inorganic anions, in this case NO3 − and NO2 −, the mobile phase must contain a complementary counter ion that interacts with it and with the bonded stationary phase of the column concurrently. In the absence of the counter ion, no interaction with the column is achieved and, as a result, no retention will be obtained at all, as the ions would come out in the void. In this scenario, a tetrabutylammonium salt (e.g., tetrabutylammonium hydrogen sulfate, TBAHS, 155837 Sigma-Aldrich) is a possibility (Figure 6B). In this case, the four alkyl chains from the reagent interact with the eighteen-carbon alkyl chains of the stationary phase and, at the same time, with the NO2 −/NO3 −. The elution order may be explained by considering a more delocalized negative charge (among three oxygen atoms) in NO3 − and the bent geometry of NO2 − due to the nitrogen atom-containing an electron lone pair. Interestingly, NO2 − is a larger anion (0.192 nm), when hydrated, than NO3 − (0.179 nm) [276]. Now, depending on the length of the column, the affinity of the this will not be sufficient to resolve peaks from the solvent front (specially the first peak; NO2 −), this issue is easily solved including acetonitrile in the mobile phase, using slower flows, a longer column or even an ion pair agent with longer alkyl chains (e.g., octylamine). The mobile phase used is 20% acetonitrile, 80% TBAHS 5 mmol L−1, at a 6.5 pH. Interestingly, when injecting a solution with both ions present and at the same concentration, the response (the signals obtained), is very similar in area/height and, as such, sensitivity is very close for both anions.
**Figure 6.** Schematic representation for the interaction of nitrite ion with (**A**) a cation exchange stationary phase or (**B**) interaction with TBAHS present in the mobile phase and stationary phase C18.
The same methodology has been used in feed to assay hay samples (Figure 7A,B) that were presumed as the source of intoxication in horses [277]. In this case, from ten samples assayed, three (average concentrations of 92.77 ± 60.88 mg kg−1) and six (average concentrations of 92.13 ± 47.55 mg kg<sup>−</sup>1) samples tested positive for NO2 − and NO3 −, respectively (unpublished data). Forage and swine compound feed samples (*n* = 10) have also been assayed with this method obtaining values from <5 to 23.69 and 2.30 to 4.96 and 925.15 to 1135.10 and 989.51 to 1479.71 mg kg−<sup>1</sup> for NO2 − and NO3 −, respectively on both accounts. In the case of feeds and fish meals, which suffer from severe matrix effects, SPE has been applied, with good results, as a cleanup and concentration step. Specifically, Oasis® MAX cartridges, conditioned with 2 mL methanol, and 4 mL water, load 1 mL sample, wash <sup>3</sup> × 1 mL water, elute with 2 mL 0.5 mol L−<sup>1</sup> NaCl solution. Chromatograms improve drastically when the elution from the cartridge is performed using the mobile phase.
**Figure 7.** Chromatograph of (**A**) an aqueous 10 mg L−<sup>1</sup> nitrite (4.95 min) and nitrate (6.26 min) standard (**B**) hay sample after extraction with hot water, SPE cleanup, and micropore filtration presence of nitrite (4.91 min) and nitrate (6.23 min) is evident.
#### 4.1.5. Legislation
Regulation 2002/EC/32 sets limits for NO2 <sup>−</sup> in fish meal (i.e., 60 mg NaNO2 kg<sup>−</sup>1) and complete feedingstuffs (i.e., 15 mg NaNO2 kg−1) excluding those intended for pets except birds and aquarium fish. We refer the reader to two thorough reviews that tackle regulatory as well as methodological topics [278,279].
#### *4.2. Carotenoids*
Chemically, carotenoids are conjugated hydrocarbons that may be further classified as carotenes (without any oxygen molecules) and xanthophylls (with one or more oxygen molecules). Carotenoids are widespread natural pigments, are recognizable from the bright colors (yellow, orange,
red, or purple) that they often confer on plant and animal organ. The molecules responsible for producing said coloration must be attained from dietary sources. For example, lutein and zeaxanthin are carotenoid pigments that impart yellow or orange color to various common foods such as cantaloupe, pasta, corn, carrots, orange/yellow peppers, fish, salmon and eggs, β-carotenoid and isomer are found in sweet potatoes, dark leafy greens, butternut squash, lettuce, red bell peppers, apricots, broccoli, and peas, and lycopene are in tomato. As molecules with a conjugated double bond system, carotenoids serve several physiological functions (e.g., antioxidants, immunostimulants, photoprotection, visual tuning, among others). This electron delocation causes them to be particularly unstable compounds, especially sensitive to light, heat, oxygen, and acids. Hence, several precautions have been taken while extracting carotenoids. For example, must be carried out in dim lighting; use rotary evaporation at low temperature and reduced pressure also it has to be carried out under a stream of nitrogen. Finally, samples should be stored in the dark, at about −20 ◦C [280,281].
Carotenoids are fat soluble but, because of the high moisture content of plant tissues, a preliminary extraction solvent miscible with water (e.g., methanol or ethanol) is generally necessary to allow for penetration of the extraction solvent.
Saponification is required to remove interference as neutral fats, chlorophylls, and chlorophyll derivatives. Usually, this procedure is carried out with potassium hydroxide in methanol. Then, it is necessary to perform liquid–liquid extraction using a water-immiscible solvent (e.g., ethyl acetate, ethyl ether, hexane) to obtain the unsaponifiable fraction, where carotenoids should be present. [280–298]. The identification and quantification require high-resolution techniques; the reversed-phase high-performance liquid chromatography has been used routinely to determine carotenoids because of its satisfactory separation efficiency. So, several factors must be evaluated to employ HPLC technique such as column type, mobile phase, chromatographic conditions. Several methods for carotenoid analysis are summarized in Table 13.
Regarding column type, the analysis can be performed using a C18 column. However, YMC C30 Carotenoid dedicated column provides excellent results, had better resolution than a C18 column for separation of carotenoids and their geometric isomers. The thirty-carbon alkyl chains interaction with the carotenoid lipophilic profile guarantee less peak distortion and better resolution [280,281]. Compounds such as α/β/γ/δ/ε-carotene, lutein, zeaxanthin, β-cryptoxanthin, dehydrolutein, anhydrolutein, astaxanthin, galloxanthin, α-doradexanthin, adonirubin, and canthaxanthin can all be separated using the aforementioned chromatographic column.
According to Huck and coworkers, the flow rate did not significantly influence the resolution, but it is essential to use an adequate flow to generate acceptable column back pressure. Also, they studied the effect of column temperature on the separation of lutein, zeaxanthin, β-cryptoxanthin, and β-carotene. The column temperature was varied between 21 and 80 ◦C; the best selectivity being achieved at 21 ◦C, at a temperature of 34 ◦C, zeaxanthin could not be easily separated from lutein. The authors concluded that maintaining a constant temperature during carotenoid analysis is critical as small changes in the ambient temperature can cause significant changes in the chromatographic selectivity of the carotenoids and at temperatures higher than 60 ◦C, the investigated carotenoids unstable.
In the case of the mobile phase, the same authors indicated that carotenoid selectivity was better using tetrahydrofuran, rather than ethyl acetate, and also better than MeOH and ACN. Carotenoids are sensitive to degradation on the stationary phase of the column by the presence of silanol groups.
**Table 13.** Common chromatographic approaches for the determination of carotenoids.
Solvent modifiers could be added to the mobile phase, for example triethylamine (TEA). Free silanol groups on the surface of silica deprotonate in the presence of the basic molecules, preventing the analyte from interacting with the medium. The TEA generates a positive impact on peak symmetry, reducing the peak tailing effect, reducing the retention time, and improving the recovery. The addition of triethylamine to the mobile phase can also have negative consequences, such as changes in the pH of the mobile phase; therefore, it is recommended that TEA be used in low concentration (less than 0.05 mL/100 mL) [299].
When using chlorinated solvents, the addition of ammonium acetate to the MeOH provides sufficient buffer capacity to prevent losses due to acid degradation of carotenoids. Some papers use MTBE as part of the mobile phase. The advantage in using this solvent, instead of chlorinated solvents, lies in the MTBE is less volatile (55.2 vs. 39.6 ◦C, respectively) and less toxigenic. Depending on the solvent system, a good compound separation may require a longer run time and poorer resolution compared with MeOH/ACN/H2O/CH2Cl2. Carotenoid content in tropical pigment-bearing fruits [281,295,300–302], and fish [302] have also been described.
#### Method Application Experience
The preferred methodology used in our laboratories is based on the work by Gayosso and coworkers with some modifications [282]. We use MTBE/MeOH as the mobile phase with a gradient system for 45 min with YMC C30 (150 × 3.0 mm, 3 <sup>μ</sup>m) at 0.6 mL min−<sup>1</sup> and 30 ◦C. These conditions were applied to identify and quantify carotenoids in food matrices such as palm oil, peach palm, sweet potatoes, papaya, and guava. We extracted the carotenoids from these matrices using a saponification procedure, followed by extraction with ethyl ether. This solvent evaporates at 40 ◦C and the residue is reconstituted in CHCl3. Undesired coextractants (e.g., waxes, sterol and tocophenol esters) are usually better solubilized with this solvent than MTBE saving from additional filtration steps and within-system precipitation. Optimization of injection volumes and initial composition of the mobile phase can somewhat mitigate the effect that injecting in a different solvent [303]. Analogous to polyphenols, carotenoid extraction methods must contemplate ester hydrolysis or other treatments to ensure the quantitation of overall amounts of carotenoids. For example, it is common to find carotenoid esters in food matrices, and these adducts present several intrinsic difficulties during carotenoid determination [295]. However, mass spectrometry-based LC is a powerful tool able to discriminate both parent compounds and their esters [295]. Recently, Wen and coworkers identified *n* = 69 carotenoids esters in *Physalis alkekengi* L. and *P. pubescens* L. fruits [297]. Additionally, BHA and BHT are common organic-solvent-soluble antioxidants to preserve carotenoid integrity [298]. Finally, our laboratory has also assessed carotenoid content in plasma from colored tropical frogs (*Agalychnis callidryas*).
#### *4.3. Carbohydrates and Sugars Soluble in Ethanol*
Animal feeds are, by definition, based on vegetable/plant sources that use carbohydrates as storage compounds, structure elements, and energy sources [10]. Then, carbohydrates form the most substantial portion of the organic matter in feeds; they can be divided into two main categories non-structural and structural carbohydrates. We encourage the reader to examine an excellent review of carbohydrate and organic acid in food commodities intended for human consumption by da Costa and Conte-Junior [304]. A great starting point for reviewing different approaches for carbohydrate analysis is the thesis written by de Goeij [305].
#### 4.3.1. Carbohydrate Measurement Using Amino-Based Columns
Xu and coworkers compared two methods for sample cleanup and extraction. A macroporous resin was compared to a solid phase sorbent based on alkyl chain. From the two approaches, SPE showed less analyte loss (11.32 vs. 0.69%). However, the discoloration ratio was similar for both methods. Sugar profile from molasses samples was obtained [305] after pigments, nitrogen compounds, and inorganic ions were removed. The analysis was performed using two NH2-based columns. Under the same conditions, it was concluded that the Zorbax Carbohydrate column showed better performance. Agius and coworkers recently developed a method to determine organic acid and sugars in tomato fruits [306]; the authors used ACN to improve peak shape. RID (Refractive Index Detector) is used for carbohydrate analysis since sugars do not have chromophores and alternative detectors (e.g., MS) are expensive. RID is the detector of choice in many labs for sugar profiling (Table 14) despite its relative lack of sensitivity. However, usual concentrations found in fruits counteract the issue.
4.3.2. Carbohydrate Measurement Using Amide-Based Columns
Koh and coworkers developed a method using an amide-based column, which is designed to retain polar molecules [308]. Contrary to their amino counterparts, these columns can retain analytes wide range of mobile phase pH. Thirteen sugars were separated including monosaccharides, disaccharides, sugar alcohols. This separation is impressive since it includes several molecules commonly used as sugar substitutes or replacement sweeteners. Organic amines within the mobile phase are used as stationary phase modifiers [308]. The authors recommended the use of a 150 mm column as the reduction of time of analysis using shorter lengths, compromise resolution. However, peaks obtained on longer columns are typically wider peaks resulting in lower sensitivity due to increased diffusion.
#### 4.3.3. Carbohydrate Measurement Using Ligand Exchange-Based Columns
Duarte-Delgado and coworkers assayed four different extraction methods for sucrose, glucose, and fructose, and demonstrated that a double aqueous MeOH extraction was the more efficient approach for the determination of these sugars [310]. The authors used SPE and guaranteed the removal less polar compounds and avoid possible co-elution with sugars during HPLC analysis. Extraction method seems to be more critical for mono than disaccharides, and starch gelification appears to be an interference when extraction is performed with hot water. Zielinski and coworkers a cation exchange gel in calcium form column to determine sucrose, D-glucose, fructose, and sorbitol in different ripe stages and during senescence of *Malus domestica* (Suckow) Borkh [309].
Senescent apple juice showed higher sugar concentration; a stage in which fruit is better suited for fermentation. Shindo and coworkers used recovered sugars from samples such as orange juice, yogurt, chewing gum, milk, and biscuits (this last matrix needed a triple extraction to obtain adequate recoveries). Additionally, the authors optimized column temperature and flow rate [311].
#### 4.3.4. Reverse Phase Columns and Sugar Derivatization Techniques
Several detection systems are used to detect carbohydrate after chromatographic separation, an approach commonly used is the pulsed electrochemical detection. A thorough review of this technique has been already written by Corradini and coworkers [321]. Evaporative light scattering detector has also been used to assay sugars. Dvoˇráckova and coworkers wrote a comprehensive ˇ review of this technique [322]. The most common detector for chromatographic analysis of sugars is refractive index. UV detection is usually inconvenient as the wavelength 210 nm (low range of the UV) has the disadvantage of exhibiting interferences. An easy way to circumvent this to derivatize using pyrazolones (e.g., 1-phenyl-3-methyl-5-pyrazolone) to form Schiff bases with reducing sugars and monitor using 248 nm. This approach only works for reducing sugars. Hence, sucrose will not be detectable. Additionally, a C18 column (usually readily available) can be used to separate the adducts. Canesin and coworkers analyzed sorbitol from lateral buds of fruit trees (e.g., black mulberry, peach, avocado, and pear) as a way to monitor primary photosynthesis products [312]. In this case, traditional detection systems are not useful as levels of sorbitol are in the μg per mg.
Hung and coworkers were able to add a fluorophore to aldol sugars assisting in their detection and mass fragmentation [313]. Naphthylimidazole fluorescent derivatives were obtained successfully for sugars (only for reducing aldoses) extracted from beverages such as fruit juice, yogurt, coffee drink, milk tea, and flavored milk. Additionally, oligosaccharides from a *Solanaceae* were identified using the approach above and NMR as an additional confirmatory tool. Recently, special attention has been drawn toward added sugars in food commodities; sterner regulations have been set in different countries due to population health concerns such as obesity, diabetes, and heart disease [323]. Hung and coworkers also used their approach to assess added sugar in the food items tested [313]. Carbohydrates analysis in food should contemplate, systematically, added sugars during chemical determinations [324].
#### 4.3.5. Aqueous Normal Phase Chromatography for Sugars
Interestingly, Valliyodan and coworkers used an aqueous normal phase approach based on a hydrophilic polymeric gel) to assess sugars from soybean. The addition of just 20–30 mL/100 mL of acetone to acetonitrile, in the mobile phase, permitted the successful separation of galactose from glucose [314].
#### 4.3.6. Complex Carbohydrates and Conjugates
Hydrolysis of complex (mainly structural) carbohydrates has been used previously to assess them [325,326] by indirect determination of their basic units and building blocks. Several approaches can be used to achieve this [325]. However, HPLC can be an attractive one since it provides high specificity and selectivity. As hydrolysis usually produces considerable concentrations of the monomer, usually sensitivity is not an issue. For example, we have used endo-1,4-β-mannanase (EC 3.2.1.78) to break down and indirectly determine mannan, monitoring mannose. Similarly, hydrolysis can be used to assess the quality of commercial mannanase. Mannanase is commonly used as a feed ingredient to improve nutrient absorption [10]. Here, an enzyme of known activity (a standard, see for example E-BMANN from Megazyme) is directly compared to the commercial one (the feed additive); a galactomannan polysaccharide (like guar gum) can be used as the substrate.
Weiß and Alt described an exhaustive method to assay sugars in plant materials and feeds. Separation of the following was achieved: inulin, verbascose, stachyose, raffinose, cellobiose, sucrose, isomaltose, maltose, lactose, glucose, xylose, galactose, rhamnose, arabinose, fructose, mannose, ribose, and mannitol [315]. Flow rate, temperature, mobile phase composition, and injection volume were optimized. From the series of columns tested, the Nucleosil® Sugar 682 Pb (Macherey-Nagel GmbH & Co. KG, Düren, Germany) was finally used at 85 ◦C, H2O at 0.4 mL min−1, and using 20 μL. Recent data show that inulin-rich diets can benefit gut microbiome, notwithstanding, routine inulin analysis in feeds is uncommon [327]. It was not until very recently that the minimal performance requirements were established for fructans analysis in feed, pet food, and their ingredients [328]. However, excess dietary fructans have demonstrated adverse health effects in equines [329]. AOAC® (Rockville, Maryland, USA) Official MethodSM 997.08 is available to assess fructans in food products using ion exchange chromatography with pulsed amperometric detection. The method is based on two-step hydrolysis using amyloglucosidase (to remove starch) and inulinase. Measurement of simple sugars in different food-derived extract fractions is performed using glucoheptose as an internal standard. The same principle has been used to assess fructose derived from fructans in pet food [330]. Verspreet and coworkers analyzed fructan from wheat grains after acid hydrolysis. Mild conditions used during hydrolysis avoid the release from other naturally occurring saccharides in wheat that would otherwise interfere during the fructan estimation (e.g., raffinose oligosaccharides) [316]. Correia and coworkers have developed a method to analyze fructooligosaccharides [317]. These sugars are dietary and are used as food ingredients (incorporated as dietary fibers in commodities). The authors monitored sucrose pathway fermentation products from *Aspergillus aculeatus* as a potential source of fructooligosaccharides; fructose, glucose, sucrose, 1-kestose, nystose, and 1F-Fructofuranosylnystose were monitored.
We have used enzymatic hydrolysis to obtain glucose from starch molecules present in feed and feed ingredients. Total and resistant starch was measured in several matrices including (e.g., silages) [318]. Bai and coworkers analyzed mono- and oligosaccharides from Hakka rice [319] as a measure of quality for sugars such as isomaltotriose, isomaltose, panose, maltose, and glucose. Finally, the determination of bacterial exopolysaccharides has also been reported [320,331].
#### 4.3.7. Method Application Experience
Amine-based columns (e.g., Zorbax® Carbohydrate (Agilent technologies, Santa Clara, USA), Ultisil® XB-NH2 (Welch Materials, Inc, Texas, USA)) are successful in separating mono and disaccharides in foods especially those containing lactose (such as dairy products). However, this type of stationary phase suffers easily from poisoning as amine functional groups form covalent bonding with several compounds (e.g., Schiff bases). As the amine functional group is sensitive to pH changes, extracts must be adjusted to avoid changes in the chemical form of the stationary phase functional
group as this may affect repeatability/reproducibility or even obliterate the column capacity for retention. Hence, the elimination of interferences is paramount. Additionally, when retention capacity is lost, it is possible to apply changes in the mobile phase composition and flow (e.g., to increase acetonitrile concentration and reduce flow).
A particular case is that of coffee samples. Amine-based columns especially suffer when analyzing coffee extracts as they contain phenolic acids (e.g., chlorogenic, syringic, ferulic, protocatechuic and hydroxybenzoic acid) and alkaloids (e.g., caffeine, caffeic acid, theophylline, trigonelline) [332]. Costa Rican regulations accept not more than 10 g/100 g sucrose in roasted coffee. Hence, monitoring sugar levels, as a quality standard, in these products is paramount. When routine quality control in coffee samples is necessary, we recommend to use stationary phases more resistant to pH changes (e.g., amide-based), include mobile phase modifiers (e.g., triethylamine), or intensive extract clean up.
In the case of animal compound feed, for example, suckling pigs feed usually contain lactose. Contrary to the amine-based column (Figure 8A,B and Figure 9A), ion exclusion (e.g., Agilent Hi-Plex Ca, Phenomenex® RezexTM RCU-USP Ca2+ (Torrance, California, USA)) is better equipped to deal with a larger range of samples and is less prone to deteriorate.
**Figure 8.** Chromatographs of (**A**) 2 g/100 mL standard mixture of four sugars including fructose (5.24 min), glucose (6.26 min), sucrose (9.12 min), and lactose (13.09 min) separated using amino column (Zorbax Carbohydrate, 0.7 mL min−1, 80 ACN: 20 H2O). (**B**) Sugar content of a molasses sample after hot water extraction, fructose (5.18 min) and glucose (6.31 min) signals are evident. (**C**) 1 g/100 mL standard solution for arabinose (3.89 min) (**D**) 1 g/100 mL standard solution for xylose (4.30 min) (**E**) 1 g/100 mL standard solution for ribose (4.76 min), and (**F**) 1 g/100 mL standard solution for mannose (5.42 min). Signal at ca. 1.80 min corresponds to the solvent front; constant in all injections.
**Figure 9.** Schematic representation of sugar interaction mechanism using (**A**) amine based (**B**) calcium ion-based ligand exchange column.
These types of columns are able to separate H+ (organic acids/monosaccharides), Na+, Ca2+ (sugars/alcohols), Ag+ and Pb2+ (oligosaccharides). Sugars and alcohols are separated using ligand exchange (Figure 9B) and organic acids by ion exchange. During ligand exchange, the more complex sugars elute first whereas simple ones, such as fructose, elute last, opposite to the elution order found in amine-based columns. Another advantage is that ion exchange columns need ultra-high purity water (type I) to segregate analytes while amino-based columns require acetonitrile in the mobile phase to perform. However, amino-based columns have the inherent advantage that complete chromatographic runs can be achieved under 12 min (while setting the column at 30 ◦C). Meanwhile, a good separation using exchange columns can be extended up to 30 min at 60 ◦C. As LC-MS usually use low flow rates to aid in solvent nebulization, it is harder to develop methods using this type of column due to their dimensions. When using ion exchange based-columns, EDTA can be used to sequestrate ions present on complex sample extracts reducing interferences. Peak tailing or fronting is also common when there is already wear of the chromatographic column. Finally, mathematic designs can be used to optimize method critical parameters and attributes, at least one research group has used this approach (i.e., Monte-Carlo simulation) to analyze sugars in herbs [333].
#### *4.4. Organic Acids*
Organic acids are common substances found naturally in several foods and result from fermentation processes [334]. As such, they are responsible for the particular flavor and aroma of commercially relevant commodities such as wine, vinegar, fermented meats, and yogurt, to name just a few. These substances have found widespread use in the food and feed industry as preservatives (increasing shelf-life, [335]) and antimicrobials (due to their bacteriostatic properties) (large-scale use of benzoic acid in beverages is a clear example, [335]). Organic acids can be determined by HPLC using ion exchange columns. Several anion exchange columns have already been mentioned in the previous section. Usually, though, measurement is commonly made using a UV detector at 210 nm (detection of absorption of carboxyl groups). Solvents and columns may vary slightly, but an isocratic method is sufficient to separate the analytes. The sample preparation for organic acid determination in beverages can be straightforward. In some cases, depending on the clarification, process suffered by the final product, it may consist of centrifugation and microfiltration. For an excellent primer for organic acids, we recommend the book edited by Vargas [336].
#### 4.4.1. Reverse Phase Chromatography Analysis in Foods
Neffe-Skoci ´nska and coworkers analyzed sugars, ethyl alcohol and organic acids in Kombucha tea beverages (a fermented brew) with an emphasis on glucuronic acid (which has been associated with health benefits) [337]. The fermentative profile was evaluated for 10 days at 3 different temperatures. Nour and coworkers use low temperature so they can separate 6 compounds (oxalic, tartaric, malic, lactic, citric, and ascorbic) in 13 min. The applied temperatures (i.e., 10 ◦C) in a 250 mm column using 0.7 mL min−<sup>1</sup> flow rates ensures optimal resolution of the compounds while obtaining adequate peak shapes within a reasonable time. A buffer adjusted at 2.8 pH guarantees that the compounds of interest are maintained during chromatography as protonated species [338].
Different citrus juices were tested finding concentrations of citric acid ranging from 7.39 × 104 to 6.89 × <sup>10</sup><sup>4</sup> mg L−1. Reverse phase separation of acids uses buffers (e.g., the H2PO4 −/HPO4 2− pair) or salts (e.g., Na2SO4) to accomplish separation. The advantage of this approach is that usually C8/C18 columns are readily available and are relatively inexpensive. The downside resides in that the use of this kind of mobile phases increase the possibility of crystal precipitation in the pump and capillaries. The buffer has to be prepared daily (to circumvent microbial growth), and pH values strictly supervised (to avoid retention time shifts). Lobo Roriz and coworkers determined organic acids in three different medicinal plants which are widely consumed as infusions. *Gomphrena globosa* L. showed the highest levels or organic acids (mainly malic and oxalic) [339].
*Pterospartum tridentatum* (L.) Willk. and *Cymbopogon citratus* (DC.) Stapf showed higher levels of citric and succinic acids, respectively. Acid content depends on inherent plant genetic characteristics and edaphoclimatic conditions. The authors also analyzed sugars (using HPLC-RID Eurospher 100–5 NH2 column and melezitose as internal standard) and, interestingly, tocopherols (α, γ, and δ-tocopherol, normal phase YMC Polyamide II column and fluorescence at λex and λem 290 nm and 330 nm). Scherer and coworkers used a reverse phase column to assess ascorbic acid stability in apple, orange and lemon juices. They also compared nutritional analysis reported within the food labels for ascorbic acid with that obtained experimentally [340].
#### 4.4.2. Ion Exchange Chromatography Analysis in Foods
Llano and coworkers analyzed sugars, acids, and furfural in pulp mill residue. The authors used a resin-based cross-linked gel column for low molecular-weight chain acids, alcohols, and furfurals. This method is particularly interesting since a comparison between 2 sets of columns for each application was tested (Table 15). The authors also included specific details for each column and optimize temperature, injection volume, and flow rate. Size exclusion also seem to have a role in sugar separation using ion exchange columns [341]. Though this application is not specifically for food, xylooligosaccharides (from revalorization alternatives for materials derived from the pulp mill enterprise) have found applications in the food industry and have been even linked to health benefits [342]. Saleh Zaky and coworkers reported a simultaneous analysis of chlorides, sugars, and acids [343]. However, the paper states that several inorganic ions are retained with the same strength within the column (all tested ions have different physicochemical properties; i.e., hydration spheres, charge among others). We have not been able to repeat this procedure. The authors did analyze sugars and acids (i.e., citric, lactic, acetic) and ethanol in a grand variety of food products including energy drinks, sodas, tomato juice and sauce, brine, milk, whey, cheese, and hummus. An interesting paper focused on the determination of organic acids from olive fruits. Different organic acid profiles were found for unique fruit varieties. They found oxalic, malic, succinic, and citric as main organic acids [344].
Mihaljevi´c and coworkers separated organic acids in wine. Organic acid profile (especially glucuronic and galacturonic acids levels) was able to distinguish among Traminer vs. Welsch produced Croatian wines. Mobile phase rate was reduced during chromatography when target acids were glucuronic, gluconic, galactaric, and galacturonic [346]. Diacids and citric acid considerably differ structurally (e.g., number of carbons). Meanwhile, the reduction of flow rate responds to the subtle differences among these intimately related structures, making them more difficult to resolve. Sánchez-Machado and coworkers preserved shrimp tissue through fermentation with lactic acid bacteria. A complete separation of lactic, citric and acetic acid was accomplished. Sonication time and initial sample mass were optimized during the assay [347]. Finally, though most of the tests regarding organic acids extraction-wise are straightforward, even in brightly colored samples (e.g., fruits [350]), still SPE cleanup has been applied, with adequate recoveries, to these extracts to remove interferences as anthocyanins and carbohydrates that may co-elute during acid analysis (especially relevant if a non-selective detector is used) [345].
#### 4.4.3. Ion Exclusion Chromatography Analysis in Foods
Fasciano and coworkers modified a reverse phase column, a C18 column was dynamically modified by running a solution of SDS through the column (Table 15). They separated organic acids after optimizing sulfuric acid concentration, flow rate, and pH; an example of how a reverse column can be made more versatile. A wide array of compounds in juices and sodas were analyzed using ion exclusion chromatography [349]. For a detailed description of ion exclusion chromatography, we encourage the reader to pay special attention to this paper introduction.
#### 4.4.4. Silages
The maturity of the crop governs silage quality at harvest. However, fermentation in the silo further influences the nutritive value of silage. Coblentz and Akins recently published a detailed discussion of silages [351]. Similarly, Khan and coworkers wrote a more specific review based on maize silages [352]. In both papers, references to silage quality based on organic acids are mentioned. Since silage is the result of this fermentation process, the organic acid analysis is used to monitor its quality. Concentrations of fermentation acids do not seem closely related to silage intake; however, they are decisive in the balance of volatile fatty acids produced in the rumen. In turn, affecting gluconeogenic metabolism and influencing milk and body composition in productive livestock. Several researchers have dedicated efforts to not only assess organic acid concentrations from silages but also have studied the effect that organic acid has on silage fermentation. For example, Ke and coworkers included malic or citric acid at concentrations of 0.1 to 0.5 g/100 g during alfalfa ensiling of alfalfa and concluded that these levels improved silage fermentation quality [348]. Additionally, both acids can be further used as feed additives that have proven to promote animal performance. Silva and coworkers determined the fermentation profile of alfalfa silages treated with microbial inoculants at different fermentation periods under tropical conditions [353]. From the strains tested *P. pentosaceus* showed the most efficiency suggesting its use as a silage inoculant. The sample pretreatment just consisted of extract acidification with metaphosphoric acid, gravity-aided filtration, and centrifugation.
#### 4.4.5. Method Application Experience
We have used ligand exchange-based analysis to routinely screen silage quality (Figure 10). Sample pretreatment consists of metaphosphoric acid extraction. We also have taken advantage of sample extraction for ammoniacal nitrogen (a modified version of method 941.04). In this type of columns, poly and diacids are eluted first. Monocarboxylic acids will elute later on during the chromatographic run in order of increasing alkyl chain length (i.e., formic, acetic, propionic, butyric). Finally, we have used liquid chromatography coupled with a variable wavelength detector set at 210 nm, a Hi-Plex H (300 × 7.7 mm and 8 μm particle size) column kept at 60 ◦C, and a 50 mmol L−<sup>1</sup> H2SO4 solution with a flow rate of 0.6 mL min−<sup>1</sup> to monitor ammonium propionate, added as a preservative, in dry dog foods (see for example, [354]). Average concentrations of (693.12 ± 75.63) mg kg−<sup>1</sup> have been obtained for local products. Sieved (at 0.5 mm particle size) dog food was treated with hot water to extract the propionate quantitatively. Both RID and UV detectors can be used for both organic acid and sugar (and alcohol) analysis. Using RID will enable the user to monitor all the compounds above simultaneously, but RID detectors suffer from low sensitivity when compared to others.
**Figure 10.** Chromatographs of (**A**) Mix of organic acid standards malic acid (9.24 min) methanoic acid (formic acid, 10.92 min), ethanoic acid (acetic acid, 11.65 min), propanoic acid (propionic acid, 12.62 min), lactic acid (14.92 min), 2-methylpropanoic acid (isobutyric acid, 17.22 min), butanoic acid (butyric acid, 18.52 min). (**B**) A silage sample after extraction with acid 0.01 mol L−<sup>1</sup> H2SO4. Fermentation products identified at 18.499 min, 14.903 min, 12.606 min. The signal at ca. 5.70 min corresponds to the solvent front.
#### *4.5. Vitamins*
Vitamins are essential micronutrients that humans and animals need for normal metabolism. The lack of these nutrients in dietary sources can cause serious disease; trace amounts of these
compounds are required for growth and reproduction. Based on their solubility, vitamins have been divided into two groups: those soluble in organic non-polar solvents and water-soluble vitamins. Thus, the vitamins from the B-complex and vitamin C are classified water soluble while the fat-soluble vitamins are isoprenoid compounds, namely vitamins A, D, E and K. This last group is found in small amounts on foodstuffs, associated with lipids; stored in the liver and fatty tissues, and are eliminated slower than water-soluble vitamins [45].
#### 4.5.1. Fat-Soluble Vitamins
Vitamin A is commonly expressed as retinol equivalents but can occur in different chemical forms, i.e., retinal, retinoic acid and retinyl esters. In foods, it is very common to find this vitamin as retinyl esters, more specifically, as acetate, propionate, or palmitate [355]. This vitamin is involved in immune function, vision, reproduction, and cellular communication [356]. Vitamin E is a term used to designate some related compounds as tocopherols and tocotrienols, it is found in fat products of vegetal origin, mainly oils. The most common tocopherols that can be found in food and feed are α/β/δ/γ-tocopherol in different proportion. Mixed tocopherols are considered the most effective lipid-soluble antioxidants [357]. Vitamin D is naturally present in very few foods like sea products, eggs, meat, and dairy products, the most commonly found members are vitamin D2 and D3, but it is also produced endogenously when ultraviolet rays from sunlight strike the skin and trigger vitamin D synthesis from 7-dehydrocholesterol [358]. It promotes the absorption of calcium, regulates bone growth and plays a role in immune function [359]. Lastly, vitamin K is an essential nutrient for animals and humans because it is required for functioning of the blood clotting cascade [360], just as vitamin D, vitamin K can be found in two forms i.e., phylloquinone (vitamin K1, found in green plant leaves e.g., spinach, collards, lettuce, and broccoli [361]) and menaquinone (vitamin K2, bacterium residing in the vertebrate intestine [362]).
Since these compounds are involved in metabolic pathways, and are paramount in health promotion in animals and humans, it is crucial to determine their content in food and feed to comply with daily requirements and quality control. That is why several studies have been conducted regarding the extraction and quantitative analysis of these vitamins, either individually or simultaneously [359,363–382].
#### Sample Preparation
Most of the analytical methods involve previous steps of sample preparation like saponification, solid-liquid or liquid–liquid extractions, followed by a concentration step before HPLC analysis (Table 16). The sample pre-treatment is critical for an accurate method. That is why there are many aspects that need to be controlled, Qian and Sheng have studied seven different variables to take into account for simultaneous analysis of vitamins in animal feed. These variables were related to the extraction procedure: (1) sample particle size, (2) solvent, (3) the ratio of sample to solvent, (4) extraction with and without N2 protection, (5) extraction time, (6) equipment and (7) the use of SPE for cleanup [367]. They evaluated how each of the variables affected both the coefficient of variation and the recovery of each of the vitamins in order to obtain extraction conditions that would allow them to satisfy each of the vitamins in a satisfying way.
#### **Table 16.** Determination of fat vitamins and in foods and feeds.
Regarding the first variable in meals and flours, subsample variability and homogeneity is closely linked to particle size. Qian and Sheng observed that large sample particle size causes an incomplete vitamin A extraction with high variability [367]. We have noted that for fresh products, with a total fat content greater than 10 g/100 g and high moisture content (i.e., greater than 85%) (e.g., avocado (*Persea americana* Mill.) and peach palm (*Bactris gasipaes* Kunth)), it is advisable to freeze dry the sample, before analysis, to promote homogenization and eliminate water that interferes with fat-soluble compound extraction.
Qian and Sheng developed the assay procedure without saponification, but nevertheless it is a widespread procedure used in the analysis of vitamins since it is an efficient way of removing interferences of lipid origin that can be found in the matrix [363,365,368,374,378,379,382]. Since it is based on an alkaline digestion (i.e., heated KOH or NaOH aqueous or alcoholic solutions) there is a disadvantage, the saponification could generate oxidation of the vitamins, which translate into a loss for vitamin degradation and low recovery percentage [366,368]. Some researchers have made use of the antioxidant ability of some compounds such as BHT, BHA, TBHQ, ascorbic acid or pyrogallol to reduce oxidation losses [363,365,367,368,371,374,379,380]. Nevertheless, saponification procedures take time, and the extractions procedures are not always straightforward, because emulsions are generated, as Lim and co-worker mentioned in their comparison of extraction methods for determining tocopherols in soybeans [369], these inconveniences introducing considerable variation, low recovery, and reproducibility, situations that we have also seen in the development of this type of methodologies.
Recently, alternatives to saponification process for the extraction of vitamins have been used, some papers use enzyme-catalyzed hydrolysis and alcoholysis of ester bonds in vitamin A and E esters to facilitate their determination in milk powder and infant formula. They assayed six lipase preparations and one esterase preparation using diisopropyl ether, hexanes/ethanol and supercritical CO2 containing ethanol. Three of the lipases' preparations from *Candida antarctica* (Novozyme 435), *Rhizomucor miehei* (Lipozyme IM) and *Pseudomonas cepacia,* showed considerably higher activity toward retinyl palmitate but there was no observed activity with α-tocopheryl acetate [372]. In feed, Xue and co-workers applied enzymolysis instead of saponification with a basic proteinase named Savinase in 30 min of incubation time at 40 ◦C getting good results in the determination of four fat-soluble vitamins (K3, A, D3, E) [366].
With respect to the solvent type and ratio solvent: sample, there is a wide variety of solvents available for the fat-soluble extraction, most of the methods use solvents such as hexane, heptane, chloroform, dichloromethane, ethyl acetate, tetrahydrofuran, ethyl ether, and the choice will depend on the type of matrix to work with (Table 16). For example, in the case of animal feed [367], a poor resolution was observed using hexane and chloroform, generating an overestimation of vitamin D, such mixture does not allow a good separation during centrifugation which produced a high %RSD.
If a mixture of acetone/CHCl3 (30:70) is used, the results in terms of variability and recovery of vitamins are outstanding, mainly for vitamin A. In low-fat matrices (less than 0.1 g/100 g, e.g., fruit juices), this solvent system has the disadvantage of generating emulsions and, hence, low recoveries. In the case of dairy and infant formulas where the presence of milk proteins is a hindrance, the extraction of the lipid part has been reported using saponification and extraction with hexane, leading to vitamin degradation in fat. It is also an extensive process [373,374]. For this reason, a group of researchers developed a fat extraction methodology using a mixture of CH2Cl2:EtOH 2:1 and separation at 4 ◦C with a centrifuge, giving satisfactory results for analysis of FAMES so it could be applied in the extraction of vitamins in these matrices [375].
Regarding extraction time and equipment, Qian and Sheng used vortex mixer for several minutes, rotatory mixer and supersonic mixer, these last two methods were not as effective for extraction of vitamin A and other vitamins due to low recoveries [367]. Hung used a rotatory mixer for extraction of vitamins D2 and D3 during one hour [376]. We have found, that for foodstuffs, the most efficient sample treatment is to rely on the combination of a vortex mixer for one minute, a rotatory mixer for 30 min or supersonic mixer for 15 min.
Effective extraction can be aided if the solvent contains a percentage of an appropriate antioxidant. N2 has been used in some protocols to the protection [377,380], of extracted vitamins from degradation because the solvent vapor that replaces air over the surface of extraction mixture has a protective antioxidant effect [367]. Qian and Sheng showed evidence that this protection did not influence in the mean values of vitamins A, D, and E and pro-vitamin D, but decreased the variation coefficient [367].
#### Chromatographic Analysis
The analytical method for the determination of vitamins in food and feed, is liquid chromatography (HPLC or UPLC) due to its, selectivity, short time of analysis, and high resolution. Methods based on chromatography can be easily automated and can determine several compounds at the same time.
Methods range from using normal phase chromatography with silica columns to reverse phase chromatography with C8, C18, and C30 columns (Table 16). Lee and coworkers studied three different columns to separate vitamin A and E: an NH2 column, C30, and C18. Concerning the resolution, they observed the β-tocopherol and γ-tocopherol peaks of vitamin E were not separated and appeared as a single overlapping peak when using a C18, but it could be separated using an NH2 column. Regarding detection and quantification limits the NH2 column presented values lower than C8 column but higher than C18.
The solvent systems to use as mobile phase vary depending on the selected approach, in the case of normal phase chromatography the solvents systems mostly used are 2-propanol/hexane in different proportion, but also can be use methanol/hexane/THF (97.25:2.5:0.25), or hexane/MTBE (96:4) [374]. In reverse phase the most common are MeOH-H2O, MeOH-ACN, both techniques can be used in gradient o isocratic mode.
As mentioned before, the liquid chromatography technique has a wide variety of monitoring techniques including PDA, FLD, ECD, ELSD or MSD. The most commonly used detector for vitamins is FLD, which is considerably more sensitive and selective than UV. Therefore, it is possible to carry out a simultaneous determination of vitamin A and E, for which a programming of the equipment is required so that at certain time intervals it uses the excitation wavelength (λex) and emission wavelength (λem) specifies for each vitamin, for vitamin E, λex = 285 and λem = 310 nm, for vitamin A the configuration at λex = 325 and λem = 470 nm, but no other vitamins could be detected such as K or D. Alternatively, PDA can work with multiple UV wavelengths and determine the four vitamins at the same time. Mass spectrometry coupled chromatography is usually the most versatile option. However, it requires that the laboratory has the resources for its acquisition. We have successfully applied mass spectrometry to assess tocopherols in feed supplements and animal biological samples (Figure 11A–F).
**Figure 11.** Single quadrupole LC/MS ESI<sup>+</sup> chromatographs of (**A**) Total ion chromatogram α-tocopherol (a 1 mg L−<sup>1</sup> solution in butanol) signal positively identified at 11.75 min (**B**) Mass spectra for α-tocopherol (a 1 mg L−<sup>1</sup> solution in butanol) using a cone energy of 120 V extracted from a signal with a retention time of 11.71 min (**C**) α-tocopherol (retention time 11.77 min) identified in a chicken plasma sample after extraction with chloroform and butanol (**D**) α-tocopherol in selected ion monitoring (SIM) mode using a cone energy of 120 V extracted from signal with a retention time of 11.82 min (**E**). α-tocopherol acetate in an injectable vitamin E solution for veterinary use using a "dilute and shoot" approach (16.32 min), and (**F**) α-tocopherol acetate in SIM mode using a cone energy of 60 V extracted from signal with a retention time of 16.34 min.
#### 4.5.2. Hydrosoluble Vitamins
One of the main issues that the hydrosoluble vitamins analysis exhibit is that each molecule is structurally different. Hence, to assess each vitamin, different conditions must be applied to the HPLC system to assess each vitamin. Kim published a paper in which ion pairing chromatography was used to monitor six different vitamins (nicotinic acid, nicotinamide, folic acid, and pyridoxine) in the feed [383]. Sodium hexanosulfate was used to with this approach; all six vitamins can be quantified using the same chromatographic run (using the same wavelength and column for all species) (Figure 12).
**Figure 12.** Hydrosoluble vitamin analysis based on ion pairing [383]. (**A**) Successful separation of 7 complex B vitamins including niacin (nicotinic acid, B3, 6.67 min), FMN (B2, 14.12 min), pyridoxal (B6, 17.007 min), pyridoxamine (B6, 18.607 min), pyridoxine (B6, 19.963 min), folic acid (B9, 20.630 min), and thiamine (B1, 25.074 min). (**B**) Analysis of a vitamin premix destined for feed formulation. Another advantage presented is that the separation can be performed using a reverse phase C18 column.
A usual problem that arises with the analysis of this compounds is the fact that pH changes affect their chemical behavior drastically both during extraction and chromatographic separation. The author circumvented this issue using a mobile phase and sample extraction solution spiked with 0.1 mL/100 mL acetic acid, maintaining all species protonated. Midttun and coworkers described an extensive two-phase analysis based on an LC-MS/MS (to determine fat soluble and water soluble). In the chloroform/isooctane phase all-*trans* retinol, 25-hydroxyvitamin D2, 25-hydroxyvitamin D3, α-tocopherol, γ-tocopherol, and phylloquinone were retained. The hydrophilic phase (in which water-soluble vitamins were found), was mixed with ethanol, water, pyridine, and methyl chloroformate as a derivatizing agent. In this assay there can be a third phase (i.e., the methyl chloroformate fraction) that it is reserved for gas chromatography analysis of amino acids [384]. As an excellent example of hydrosoluble vitamins analysis using LC-MS/MS in the food industry, is the determination of 15 compounds in beverages using a multi-mode column (SM-C18 column, 150 × 2.0 mm, 3 μm; Imtakt Co., Kyoto, Japan), which provided reverse-phase, anion- and cation-exchange capacities, and therefore improved the retention of highly polar analytes such as water-soluble vitamins. The use of this column removes the need for an ion pair reagent in the mobile phase [385]. Finally, we encourage the reader toward a recent and ample review regarding fat- and hydrosoluble vitamins, respectively [386,387].
#### 4.5.3. Method Application Experience
In our experience in the development of a methodology for the determination of fat-soluble vitamins in food matrices, the most challenging part has been the sample pretreatment. As mentioned above, several factors have to be considered. For the species retinyl acetate and palmitate, we chose to use a direct extraction to avoid decomposition by the saponification process. Extraction can be performed with hexane, ethyl acetate or chloroform; as the last solvent is far easier to eliminate, during concentration steps, is considered the most suitable option. Isopropanol is a useful solvent for reconstitution.
This method applies to dry matrices as flour, bakery products, freeze-dried pulps or fortified sugar. In the case of dairy products, we highly recommend the use of dichloromethane/ethanol. Separation is carried out using an HPLC-DAD set at 325 nm and a Zorbax Eclipse XDB-C8 (150 × 4.6 mm, 5 μm) column at 50 ◦C. Shifting the solvent system form a MeOH/H2O (90:10) to MeOH/2-propanol/ACN (95:1.5:3.5) saves up to 5 min of chromatographic run time and better peak shape, for the palmitate, is obtained (Figure 13A,B).
**Figure 13.** Chromatographs for vitamin A standards mixtures separated with C8 column at 325 nm and 50 ◦C, of (**A**) retinyl acetate (3.11 min) and retinyl palmitate (17.63 min) using MeOH/H2O (90:10) and (**B**) retinyl acetate (2.89 min) and retinyl palmitate (13.30 min) using MeOH/2-propanol/acetonitrile (95:1.5:3.5).
Monitoring the analytes at 295 nm, using MeOH/H2O (90:10) as solvents at a flow rate of 0.5 mL min−1, tocopherols can be successfully separated (i.e., retention times for δ/γ/α-tocopherol 5.35, 6.03 and 8.59 min, respectively; *Rs* 1.97). In samples with relatively high-fat content, saponification is necessary to eliminate interferences that usually share similar retention times that those for α-tocopherol. Food and feed samples can be fortified or can naturally contain both, vitamin D2 and D3. Therefore, for a method to be suitable, simultaneous detection of both analytes is a must. Under the conditions above, C8 stationary phases are incapable of resolving both species.
A C18 column heated at 30 ◦C and a MeOH/2-propanol/ACN (90:3:7) solvent system, with a flow set at 0.3 mL min<sup>−</sup>1, can achieve a resolution of 1.25 (Figure 14A,B). Though a mobile phase composed MeOH and H2O is highly desirable, a drawback using this solvent system in complex matrices is that α-tocopherol can be interference for the identification of vitamin D3 and vice versa (Figure 14C). A solvent gradient, a column with longer alkyl chains (e.g., C30) or the use an MS detector may be employed to solve this issue.
**Figure 14.** Separation for vitamin D2+D3 standards at 264 nm using (**A**)aC8 column and (**B**)aC18 column D2 (16.47 min) y D3 (17.24 min). Analysis performed at 30 ◦C using MeOH/2-propanol/ACN (90:3:7) (**C**) Superposed chromatograms for vitamin D2 + D3 (blue line) and δ/γ/α-tocopherol standards using a C18 column and MeOH/H2O (90:10), 30 ◦C.
#### **5. Conclusions**
LC is a powerful and versatile tool for food and feeds analysis, food and feed matrices are complex mixtures that on occasion present to the researcher difficulties as analytes of interest must be extracted and purified before injecting into the LC system. Advantages that chromatography provides when applied to food or feed analysis include sensitivity (determination of trace amounts, especially important in the case of contaminants, residues, controlled or undesired substances), automation and high throughput (reducing time and user dependence in laboratories with considerable workloads), simultaneous determination of multiple analytes. Food and feed chemists must make an effort to develop methods that provide a faster response and with few possible numbers of steps. Several current methods that are based on other step-full or less automated, or specific techniques can be reinvented, transformed and transplanted to LC analysis to improve sensitivity, specificity, and selectivity (For example, exchanging a spectrophotometric-based piperine analysis for a chromatography one). Within the myriad of alternatives, the LC approach delivers, each on its own seldom can solve a research problem, and each technique (i.e., each detector, each chromatographic column or sample treatment) has its shortcomings and limits as to which data is to provide. Usually, a multiphasic methodology is desirable to reach an appropriate conclusion. Hence, LC methods are more useful when are tailored to fit a purpose. Nowadays, mass spectrometry coupled liquid chromatography is an almost widespread technique that can provide molecule confirmation, ease trace analysis and allows the assay of multiple analytes simultaneously. Not only is a versatile tool for routine analysis but, research-wise, it provides more information about the target molecules and opens a valuable doorway toward a myriad of applications in food analysis including metabolomics, proteomics, and parvomics. Notwithstanding, as demonstrated above, traditional detectors are still the most commonly available in most laboratories. With a proper
sample pretreatment, some traditional-detector-based methods are, regarding analytical performance, comparable to those based on MS.
**Author Contributions:** Conceptualization: C.C.-H and F.G.-C.; Methodology: C.C.-H, G.A., F.G.-C., A.L.; Software: F.G.-C.; Investigation: C.C.-H, G.A., F.G.-C., A.L.; Resources: C.C.-H, G.A., F.G.-C., A.L.; Data Curation: F.G.-C.; Writing—original draft preparation: C.C.-H, G.A., F.G.-C., A.L., C.C.-H, G.A., F.G.-C., A.L.; Visualization: F.G.-C..; Supervision: F.G.-C, G.A; Funding Acquisition: C.C.-H.
**Funding:** This research was funded by the University of Costa Rica grants number A2502, B2062, B2066, B2659, B8042, B5084. B3097, ED-427 and ED-428 and the APC was funded by the Office of the Vice Provost for Research of the University of Costa Rica.
**Acknowledgments:** Marelyn Rojas Lezama is acknowledged for her help doing the experiments regarding nitrate and nitrite in the hay. Special thanks to Guy Lamoureux Lamontagne for his suggestions.
**Conflicts of Interest:** The authors declare no conflict of interest.
#### **References**
*Importance in Human Nutrition and Health*; Saura-Calixto, F., Pérez-Jiménez, J., Eds.; Royal Society of Chemistry: Cambridge, UK, 2018; ISBN 978-1-78801-106-8.
storage time on the stability on palm olein during thermoxidation. *J. Food Process. Preserv.* **2018**, *42*, e13592. [CrossRef]
© 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
| doab | 2025-04-07T03:56:58.779089 | 1-5-2021 17:44 | {
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"book_id": "004acd65-0f8a-4342-8761-8da04dfc73fe",
"url": "https://mdpi.com/books/pdfview/book/2994",
"author": "",
"title": "Application of Liquid Chromatography in Food Analysis",
"publisher": "MDPI - Multidisciplinary Digital Publishing Institute",
"isbn": "9783039433629",
"section_idx": 4
} |
Subsets and Splits
No saved queries yet
Save your SQL queries to embed, download, and access them later. Queries will appear here once saved.