The study was conducted on “Tombul” hazelnuts in 2018 from a single orchard, and nut samples were provided by Yavuz Gıda Sanayi ve Ticaret AŞ (Giresun, Turkey) (40°54´37.33˝N, 38°26´17.23˝E, and altitude 12 m). The average kernel moisture content was ∼27.5% at the time of harvest (August 4 - August 5, 2018). The clusters were spread on the grass ground and dehydrated for 3 days (August 9 to August 12, 2018) to allow moisture loss (∼21.52%). Then, the nuts were separated from their husks by hand. The samples were dried in the sun and the drying process continued until the moisture content was 6.45%.The samples (unshelled) were kept in a 2 kg vacuum polyethylene package (150 ± 8 µm thickness, 0.029 gm⁻²/day oxygen permeability, 5 gm⁻²/day water vapor permeability), and stored in a refrigerator (Bosch KDN53NW22N A, No-Frost, Germany) at 60-65% relative humidity and −5 °C temperature until oil extraction and further analysis.

Hazelnut oil was extracted using a Ceselsan cold press oil extraction system (AISI3004, Ceselsan, Giresun, Turkey) (compression force: 10,000 kgf, pressure: 34.7 MPa, temperature: −5 °C to +45 °C, and capacity: 250 g kernel) (Turan, 2018a). The extracted hazelnut oil was stored in the freezer at −18 °C until analysis (Bosch KDN53NW22N A, No-Frost, Germany).

The vitamin E (tocopherol) composition of the samples was determined using the standard method described in AOCS Ce 8-89 (AOCS, 1997). One gram of extracted hazelnut oil was diluted with 10 mL hexane and the resulting mixture was injected into the HPLC instrument using a 0.45 µm PTFE syringe filter. It was analyzed by Shimadzu-Prominence LC-20A under the following HPLC conditions: column: C8 (250×4 mm) 5 µm, flow rate: 1 mL min, mobile phase: Hexane:Isopropyl alcohol (99:1), wavelength: 295 nm, column temperature: 25 °C.

The total phenolic content was determined by modifying the Folin-Ciocalteu colorimetric method with a UV-visible spectrophotometer (Singleton et al., 1965). For analysis, 20 µL of sample extract was taken in a micro cuvette and 1.58 mL of purified water and 100 µL of Folin-Ciocalteu reagent were added. After 5 min, 300 µL of a saturated Na₂CO₃ solution were added. The solution was stored in a dark place for 2 h. After 2 h, the absorbance of the samples was determined at 760 nm simultaneously in triplicate. A calibration curve was created by using a set of solutions with 40, 80, 120, 160, 200, 250, and 300 mg gallic acid/L concentrations to calculate the results. The results were expressed as gallic acid equivalent (GAE).

The color ordinates of the hazelnut kernels were determined by Hunter Lab Color Flex Ez color instrument (HunterLab, USA) as L* (lightness), a* (redness), and b* (yellowness). The colors of the samples were read following calibration to X:79.05, Y:84.02 and Z:89.03 (Mexis and Kontominas, 2009). The browning index (BI) was measured based on CIE L*a*b* coordinates, using the following formula (Marzocchi et al., 2017):

To obtain fatty acid methyl esters (Turan, 2018a), 0.5 g oil was weighed in Erlenmeyer flasks, and 4 mL of iso-octane and 2 mL of methanolic KOH solution were added, followed by agitation for 30 s. The mixture was stored in a sealed container in a dark place for 6 min; 2 drops of 1% methyl orange indicator were added; and the mixture was then titrated with 1 M HCl solution until a pink color was developed. After it was kept for 15 min, the colorless layer formed on top of the mixture was put into glass vials and analyzed by GC. The composition of fatty acids was determined by gas chromatography with a flame ionization detector and TR-CN100 column (60 m × 0.25 mm I.D., 0.20 µm; Shimadzu GC-2010, Japan). Both injector temperature and detector temperature were set at 250 °C. The sample (1.0 µL) was injected and helium was used as the carrier gas at 200 kPa. The injection was made at a ratio of 1:100. The column temperature was held at 90 °C for 7 min and then increased to 240 °C at 5 °C/min. Finally, it was maintained at 240 °C for 15 min. Fatty acids were characterized by comparison of the FAME mixture consisting of 37 standard components (Supelco 37 Component FAME Mixture, Cat. No. 18919-1AMP, Bellefonte PA, USA) based on their elution times (Turan, 2018b). SFA (saturated fatty acid), UFA (unsaturated fatty acid), MUFA (monounsaturated fatty acid), and PUFA (polyunsaturated fatty acid) were calculated by the following equations:

Free fatty acidity (method Ca 5a-40) was determined by the AOCS Standard Method (AOCS, 2004), peroxide level (method Cd 8-53) by AOCS (AOCS, 2004) (Metrohm, Dosimat 799, Switzerland), and the Rancimat value by the Rancimat 743 device (Metrohm, Switzerland) (Velasco et al., 2004). The iodine value (IV) was calculated using the percentage of fatty acids (Belviso et al., 2017; Turan, 2019) (9).

Data from the fatty acid profile analyses was used to evaluate the nutritional composition of the lipid fraction. Three oil quality indices were used: Index of atherogenicity (IA), index of thrombogenicity (IT) and hypocholesterolemic/hypercholesterolenic fatty acid ratio (H/H). The IA (10) and IT (11) were calculated as described by Bezerra et al., (2017).

The H/H (12) index was determined by Fernandez et al. (2019).

The experiment was conducted with three replicates based on the randomized block design. Descriptive statistics were determined by SPSS v. 22.0 (Armonk, New York: IBM Corp.). Statistical tests were performed using SAS-JAMP v. 10.0 (SAS Institute Inc., Cary, North Carolina, USA). Statistical differences were determined using the t-test. The differences among the results were determined at the levels of p < 0.05, p ˂ 0.01, and p ˂ 0.001.

Tombul hazelnut takes the first place in terms of quality among 18 hazelnut cultivars in Turkey; whereas the other cultivars are regarded secondary with respect to quality (Alaşalvar et al., 2010). The kernel of the Tombul cultivar content was determined to comprise 15.01% protein, 59.83% fat, 22.77% carbohydrate, and 2.39% ash by Seyhan et al., (2007); 4.63% moisture, 64.60% fat, 17.5% protein, 383.60 mg/100 g vitamin E, and 726.5 mg/100 g total phenolic substance by Köksal et al., (2006); 15.35% protein, 61.21% fat, 17.30% carbohydrate, 3.90% moisture, 2.24% ash, and 631 kcal/100g energy by Alaşalvar et al., (2009); and 61% fat, 16% carbohydrate, 14.9% protein, and 5.3% moisture by Memoli et al., (2017). In our study, the good kernel (GK) content was determined to contain 4.70% moisture, 54.21% fat, 15.00% protein, 23.91 g/100 g carbohydrate, 2.39% ash, 64.30 mg/kg vitamin E, 196.68 mg GAE/100 g total phenols, and 613.27 kcal/100 g energy (Table 1). Differences between bug-damaged and good kernel values were found to be statistically significant except for humidity (p < 0.001; Table 1). While total fat, ash, vitamin E, and energy values were higher, protein, carbohydrate, and total phenolic contents were lower in GK than in GD samples. Memoli et al., (2017) reported that damage caused by bugs affects the nutrient content of hazelnuts and cimiciate causes significant damage to hazelnuts. This damage is predictably caused by secretion during feeding in the developmental stages of the kernel and delay in growth (Figure 1A). Oil oxidation begins (Figure 1B) in the deformed kernel and changes in nutrient content occur.

Color is known to be an important parameter in the evaluation of hazelnuts (Marzocchi et al., 2017; Deng et al., 2018). The differences among the color ordinates were found to be statistically significant (p < 0.001) except for the browning index (BI), which is presented in Table 1 in detail. It was found that L* and a* levels were higher in the good kernel, whereas the b* level was higher in the bug-damaged nuts. The reason is that oil oxidation begins and progresses at places where bug damage occurs. Therefore, an increase in yellowness (Figure 1B), an indicator of oxidation in nuts, was noted.

Alaşalvar et al., (2010) reported that there is a generally high level of MUFA (78.10-87.26%), moderate level of PUFA 83.92-13.86%), and low level of SFA (7.46-9.59%) in hazelnuts, which is consistent with the reports of other studies (Alaşalvar et al., 2006; Turan, 2018a; Turan, 2019). It is known that hazelnut oil is preferred over olive, corn, and sunflower oil as it contains a higher level of UFA (Köksal et al., 2006; Alaşalvar et al., 2010) based on the scientific evidence that these fatty acids have a protective effect on the cardiovascular system (Uribe et al., 2018).

In our study, a total of 13 fatty acids were determined in the “Tombul” hazelnut, although 8 fatty acids were under the limit of detection (< 0.001%; Table 2). Palmitic, stearic, oleic, and linoleic fatty acids formed the major group; whereas myristic, margaric, arachidic, behenic, palmitoleic, heptadecanoic, eicosenoic, nervonic, and linolenic fatty acids formed the minor group. The fatty acids in the major group formed approximately 99.39% of the total fatty acids; whereas those in the minor group formed approximately 0.6% (Table 2). The effect of bug damage on fatty acid composition was found to be statistically significant except for margaric, behenic, and nervonic fatty acids (p < 0.001), which is presented in Table 2.

The effect of bug damage on SFA was found to be significant (p < 0.001) at 7.27% in the good kernels and 7.71% in the damaged kernels. As expected, there was a difference in the levels of palmitic and stearic major fatty acids which form the SFA. The primary fatty acid level of MUFA was 81.19% in the good kernels and 80.32% in the bug-damaged kernels. It is known that the primary fatty acid in the PUFA is linoleic acid. The effect of bug damage on linoleic acid was found to be statistically significant (p < 0.001; Table 2), with 11.12% in the good kernels and 11.60% in bug-damaged kernels. Telahigue et al., (2019) indicated that pathogens caused a decrease in fatty acids and SFA ranged between 1.82 and 0.41%, MUFA between 0.70 and 0.11% and PUFA between 2.67 and 0.38%. It was reported that this difference in PUFA is caused by the oxidation of linoleic and linolenic fatty acids (Memoli et al., 2017). Therefore, it should not be consumed in specific diets (Mostafavi et al., 2019). The effect of bug damage on UFA levels was found to be significant (p < 0.001). It was higher in the good kernels than in the bug-damaged kernels (92.65 and 92.29%, respectively, Table 2). It was reported that diseases and pests have an effect on the UFA/SFA ratio in hazelnuts, with 14.82% in good kernels, 16.13% in cimiciate and 15.96% in mold-effected kernels (Memoli et al., 2017). In our study, it was also found that bug damage has an impact on the UFA/SFA ratio (p < 0.001). However, it was found to be higher in the good kernels (12.74-11.98%). This difference has probably resulted from the different nutrient contents and/or various types of damage. In addition, it was stated that these differences may have been caused by the interaction of several factors such as altitude, latitude, longitude, temperature, precipitation, cultural practices, harvest time, and drying method (Koyuncu et al., 1997; Amaral et al., 2006; Cristofori et al., 2008; Alaşalvar et al., 2010; Turan et al., 2018b; Mostafavi et al., 2019).

The oleic/linoleic acid ratio (O/L) is one of the essential characteristics used to evaluate the quality of hazelnut kernels, and linoleic acid is more sensitive to oxidation than oleic acid (Turan, 2019). Therefore, the high O/L ratio indicates resistance to oxidation (Belviso et al., 2017; Turan, 2018a). The effect of bug damage on the O/L ratio was found to be statistically significant (p < 0.001, Table 2), and O/L was higher in the good kernels (7.30%). Based on this, it can be concluded that bug damage (Figure 1A) causes oxidation in hazelnuts (Figure 1B). The iodine value (IV) is known as a measure of degree of unsaturation in fats and expressed as the amount of iodine absorbed (Belviso et al., 2017; Turan, 2018b). In addition, a high value of IV indicates that the content is unstable and more sensitive to oil oxidation. The effect of bug damage on IV was not found to be significant (p > 0.05). However, it was shown that good kernels had a lower level (93.59). In conclusion, it appears that good kernels have a longer shelf-life. Free fatty acids (FFA) are considered to be the first indicator of the lack of quality, and their level above FFA ≥1% indicates spoilage. In our study, there was a remarkable difference between FFA values in the good and bug-damaged kernels (oleic acid: 0.49, 2.62%, respectively, Table 2). Therefore, it can be concluded that bug damage results in oxidation in the nut, and these nuts cannot be purchased. The peroxide value (PV) is one of the crucial characteristics used to indicate the quality of products stored in the hazelnut industry (Turan, 2018a), and is also considered to be the most important indicator of PV oil oxidation. The effect of bug damage on PV was found to be statistically significant (p ˂ 0.001), which is presented in Table 2 in detail. PV levels were noted as 21.58 meqO₂·kg⁻¹ in the bug-damaged kernels and 15.18 meqO₂·kg⁻¹ in the good kernels. It was highlighted that bug damage increased oil oxidation in nuts (Figure 1B). The rancimat value (RV) is a characteristic used to determine the shelf-life of hazelnuts (Turan, 2019). Table 2 shows that the RV value is higher (8.21 h) in the good kernels. Thus, the rancimat value of the nuts is decreased due to bug damage and the shelf-life is expected to be shortened.

Generally, the effect of bug damage on the quality index was found to be significant, which is presented in Table 3 in detail. The effect of bug damage on the PUFA/MUFA ratio was not found to be significant (p > 0.05, Table 3). Consumption of low levels of SFA and high levels of PUFA/SFA is associated with a low risk of heart attack (Chan and Matanjun, 2017); therefore, this characteristic is used to determine the quality of the fat fraction in foods. The PUFA/SFA ratio is generally considered to indicate the quality of fats in a diet program (Telahigue et al., 2019), and values lower than 0.45 are not desirable owing to their ability to increase blood cholesterol. In our study, although the bug damage decreased the PUFA/SFA level, it was determined to be above the threshold (1.52-1.55) and higher in the good kernels (1.55, Table 3). This characteristic was reported to be 1.46 in fish (Telahigue et al., 2019); therefore, it can be confirmed that its amount is lower in fish than in hazelnuts. In fact, this aspect of the evaluation of hazelnuts suggested that they are a valuable nutrition source for humans. Atherogenicity (AI) and thrombogenicity index (TI) levels should approach zero (Bezerra et al. 2017) because this trend represents an increase in anti-atherogenic fatty acids, which has an effect on preventing heart disease. The AI (0.16) and TI (0.15) levels in the good kernels were determined to be lower than those in the bug-damaged kernels. It has been indicated that the H/H ratio is associated with the cholesterol mechanism (Fernandes et al., 2019) and a higher level of this ratio has a positive effect on human health. In our study, the H/H ratio (18.27) was found to be higher in the good kernels than that in the bug-damaged kernels. Based on the findings of this study, it is suggested that bug damage has a negative effect on the nut quality index values. Therefore, in case of bug-damaged hazelnuts being consumed, their expected effect on the cardiovascular system and cholesterol mechanism would not be observed.