Fatty acids composition and physical properties of stones and kernels from different peach cultivars as biomarker of origin and ripening time

The peach stones and kernels are easily available biowaste which could be useful for the extraction of nutritionally important compounds such as fatty acids. Except in industry, characterization of stones and kernels could be useful in pomology to describe different cultivars, and for selecting new parents in a breeding program. A total of 25 samples of stones and kernels from various peach cultivars that differed in origin and ripening time, but growing in the same climatic conditions, were characterized by fatty acids composition and physical properties. This work confirmed that unsaturated fatty acids (oleic and linoleic fatty acids) were the most represented in peach kernel oil and their content depended of peach genotype. Additionally, the fatty acids in combination with length, weight, and moisture of peach kernels could be used as a parameter of authenticity assessment. This research may contribute for the peach cultivar discrimination and recommendation of cultivars/genotypes with high kernel quality which could be used for the extractions of oil rich in unsaturated fatty acids and further use in food, pharmaceutical and cosmetic industry. Besides, selected cultivars could be used in breeding programs, for creating new genotypes for oil production.


Introduction
Peach (Prunus persicae L.) that is native to the region of Northwest China is one of the most important continental fruit species. Worldwide peach production is > 24 mil t, with China being a leading country (> 15 mil t, 62% of the world total production), and followed by Italy and Greece (~ 1 mil t each). In Serbia, peach production is organized on > 7000 ha, with up to 80,000 t [1]. Beside standard cultivars of peach and nectarine, a numerous genotypes of vineyard peaches are grown, especially in Serbia. Vineyard peach was primarily grown in old vineyards (after which it was named) and it was the only peach type in Serbia until 1938, but now, it is an indigenous population [2].
The production and consumption of fruit products are increasing and causes huge amount of by-products that create disposal and environmental problems. The peach kernels are by-products of industrial process which are not currently used in commercial purpose and are discarded as a waste [3]. In recent years, previous studies found that peach kernels are good source of phytochemicals such as essential oil, proteins, peptides, carbohydrates, phenolics, and antioxidant constituents. Since these phytochemicals exhibit human health benefits, peach kernels could be used for their extraction or to be food supplements [3][4][5]. The Prunus persica includes peach and nectarine cultivars, which kernels can supply significant amounts of oil with good therapeutic features [6,7]. Extracted peach kernel oils are used as biodiesel feedstock and in cosmetic industry as an ingredient in various bath products such as soaps, creams and shampoos. In addition, they have the potential to be used as a substitute for commonly used edible oils in food products, also for animal feeds, plant protection, fertilizers or even in the prevention and treatment of some diseases [8]. Since peach kernel has a similar fatty acid composition and oil content as almond, it is increasingly used for the production of persipan paste, a substitute for more expensive marzipan. Persipan is product of blanched and peeled apricot or peach kernels, has similar taste as marzipan which is made exclusively from almonds [9].
Several reports dealing with fatty acids composition of the peach oil showed that it mostly consists of unsaturated oleic (18:1, n-9c) and linoleic (18:2, n-6c) acids and their percentage in kernel oil depends on location of cultivation, climate factors, varieties, and fruit ripeness [10]. These unsaturated fatty acids are rare in vegetables oils and thanks to them peach kernel oil has unique characteristics [4,11,12]. Depending on the extracting agent of the peach kernels, oleic acid could be found in values between 21.90 and 75.0% and linoleic acid between 6.90 and 29.07% [4,12,13]. The peach kernel oil in lesser extent contained saturated fatty acids, mostly palmitic (16:0) and stearic acids (18:0) [6,7].
High content of unsaturated fatty acids and low amounts of saturated ones is probably the reason for such importance and nutritionally attraction of peach kernel oil, because some of human physiological and biological functions are regulated with unsaturated fatty acids. These metabolites are significant for formation of healthy cell membranes, maintenance the proper function of brain and nervous system, for regulating blood pressure, reducing glycemic index, total cholesterol and triglycerides [6,12]. Also, essential fatty acids are included in effective prevention of many diseases, such as inflammation, autoimmune disorders, some heart diseases (lowering the level of triacylglycerides and LDL cholesterol and increasing the HDL cholesterol), and even cancer risk by dramatically decreasing disease development [14,15]. Both linoleic and linolenic acids are not synthetized in the body, but they are needed for building and repairing cell walls, tissues in the nervous system and in the formation of prostaglandins [16]. Additionally, importance of unsaturated fatty acids reflected with increasing in stability of oil, because reduces oxidation reactions so products with peach kernel oil could be more stable than others [4].
Peach stone is a thick endocarp that promotes long dormancy of the kernels (seeds). Stone is organ that is often used to differentiate between various clones/genotypes and its size dependents on which parents were crossed. Stone size is likely a proxy for fruit size, while stone shape could indicate mesocarp thickness [17]. From plant's point of view, determination of the physical characteristics of peach stones and kernels are important due of their impact on germination and plant growth, but from another side, stone shape and size are useful for processing market for designing equipment for sorting, breaking, separation and transport [18,19]. The simple measurements of some peach fruit physical properties are already applied in industry but limited number of studies reported about the correlation of physical characteristics of stone and kernel and kernel's quality in different peach cultivars [19].
In addition, the consumers demand quality food with nice flavor, nutrient benefits and without unhealthy contaminants. The authentication of food products in the market is of great interest and analytical techniques for obtaining authenticity and traceability have been developing constantly. For the protection of the consumers and the classification of cultivars, determination of the germplasm variability of the stones and kernels could be important. Observation of chemical composition and oil quality of kernels could be useful to select cultivars with better attributes which are going to be used in fruit breeding programs and for the extraction of phytochemicals for cosmetic and pharmaceutics purpose [20,21].
In that sense, present study evaluated genotypic variation and effect of dissimilarity in origin and ripening time on the fatty acids composition and physical properties of stones and kernels from 25 cultivars/genotypes of peach (including standard cultivars, promising hybrids and vineyard peaches). For that purpose, traditional extraction with n-hexane and gas chromatography with flame ionization detector (GC-FID) analytical technique for detection and quantification were performed. Additionally, principal component analysis (PCA) and correlation of fatty acids composition with physical properties of stones and kernels were performed to determine which parameters could be responsible for differentiation of genotypes and possible establish valuable authenticity factors of origin and some biologic features.

Chemicals and materials
Hexane was supplied by Merck (KGaA, Darmstadt, Germany), potassium hydroxide, sodium chloride, sodium hydrogen sulfate monohydrate were purchased from Sigma-Aldrich (Steinheim, Germany). All chemicals were of analytical purity grade. Ultra-pure water (MicroPure water purification system, 0.055 μS/cm, TKA, Thermo Fisher Scientific, Niederelbert, Germany) and methanol (Fluka Chemie Ag, Buchs. Switzerland) were used to prepare reagent solutions. The standard for fatty acid methyl esters determination used for GC-FID analysis was purchased from Restek (37 components Food Industry FAME Mix, RESTEK, lot: #23889).

Plant material
This study included 25 peach cultivars/genotypes (Table 1) which were provided from the Experimental Station "Radmilovac" of the University of Belgrade-Faculty of Agriculture. Trees were trained as an open vase, under the non-irrigated standard cultivar practices with 4 m × 4 m planting distance. Harvesting was done when fruits started softening being suitable for the consumption. Each cultivar/genotype was represented with five trees. During full maturity, ten fruits were picked from each tree, from all around the canopy. A total of 50 peach fruits were harvested in the same year from each cultivar/genotype. The peach stones were separated from tissues, washed with tap water, air dried for 2 weeks, and stored in paper bags. Before grinding and oil extraction, the peach kernel samples were separated from stone shells and dried in the open air at room temperature until moisture content was below 10%. Immediately prior to analysis, collected kernels were ground in a coffee mill with the aim of providing better extraction of the analyzed components. The extractions were performed at least in duplicate per each cultivar.

Physical characterization of peach stone and kernel samples
For the purpose of physical characterization of peach stone and kernel samples, the average length of peach stones (ALS) and kernels (ALK), the average weight of the stone (AWS) and kernel (AWK), and water content (% Moisture) were determined. Measuring the length and calculating the mean value of three randomly selected stones/kernels was done using Nonius Vernier Caliper 0-150 mm. The average weight of the stone and kernel was obtained by measuring mass with Adventurer Ohaus electronic balance and calculating the mean value of three randomly selected stones or eight peach kernels. Determination of water content in peach kernels was performed according to the procedure for determining the total dry substance described in the Rulebook on sampling methods and performing chemical and physical analysis to control the quality of fruit and vegetable products (Official Gazette of SFRY, No. 29/83). For that purpose, samples were dried in an oven (Instrumentaria ST-05, Zagreb, Croatia) at 105 ºC to constant weight.

Oil extraction and fatty acid analysis
Crude oil was extracted from powder of peach kernels using ultrasonic maceration method with n-hexane. About 1 g of each sample was extracted using 30 mL of n-hexane in an ultrasonic bath at room temperature for 30 min. After extraction, the hexane layer was evaporated at 40 °C in a rotary evaporator under vacuum (IKA RV 05 Basic 1-B Distilling Rotary Evaporator). In addition, the residual traces of solvent were removed with an N 2 stream and the extracted oil was stored in the freezer prior to further analysis. Fatty acid methyl esters (FAME) were prepared using trans-methylation under alkaline conditions, following ISO 12966-2:2012. In a 10 mL screw-top test tube, approximately 0.1 g of the extracted oil was weighed and dissolved in 2 mL n-hexane. After the addition of 1 mL of 2 mol/L methanolic potassium hydroxide solution, the tube was vortexed for 2 min at room temperature, and centrifuged at 4000 rpm for 5 min. After 2 min, 2 mL of sodium chloride solution (40 g of sodium chloride in 100 mL of water) was added and the tube shaken briefly. The solution was neutralized by adding 1 g of sodium hydrogen sulfate monohydrate. After the salt had settled, 1 mL of the upper phase was transferred to a 2 mL vial for FAME analysis.
Fatty acid methyl esters were analyzed by gas chromatography, using GC-FID Agilent 7890B GC System with flame ionization detection (FID). A fused-silica capillary column type CP-Sil 88 for FAME 100 m × 0.25 mm df = 0.2 µm was used. The flow rate of nitrogen carrier gas was 1.0 mL/min. Injector and detector temperatures were 250 ℃ and 270 ℃, respectively. The oven temperature was programmed to start with a temperature of 80 ℃, then to rise to 220 ℃ at a rate of 4 ℃/min and to maintain that temperature for 5 min, then to rise to 240 ℃ at the rate of 4 ℃/min, and to maintain that temperature for more 10 min. The sample injection volume was 1 μL. Total run time for one cycle was 55 min. Fatty acid identifications were based on retention times by comparing with those of the standard FAME mixture. Quantification of individual fatty acids was based on the peak area obtained, without any corrections. Fatty acid analysis was performed in duplicate for single samples, and average values were reported. The average relative standard deviation (RSD) of repeatability for minor components (components present at less than 1%) was 5%, while the average RSD for the components present in percentages greater than 1% was 2-3%.

Statistical analysis
To verify the existence of statistically significant differences between samples, descriptive statistics, Kruskal-Wallis test and Mann-Whitney U test were applied using a demo version of NCSS statistical software (Hintze, 2001, Number Cruncher Statistical Systems, Kaysville, UT; www. ncss. com). Principal component analysis (PCA) were carried out by a PLS ToolBox, v.6.2.1, for MATLAB 7.12.0 (R2011a; MathWorks, Natick, MA, USA). All data were auto scaled prior to multivariate analysis. PCA was carried out as an exploratory data analysis using a singular value decomposition algorithm (SVD) and a 0.95 confidence level for Q and T2 Hotelling limits for outliers. The Pearson correlation was calculated for stone length and weight (ALS, AWS), kernel length and weight (ALK, AWK), kernel moisture (% Moisture), kernel oil content (Oil), palmitic acid (16:0), stearic acid (18:0), oleic acid (18:1 n-9c) and linoleic acid (18:0 n-6c).

Physical characterization of peach stone and kernel samples
The descriptive statistics for average length (ALS, ALK) and average weight (AWS, AWK) of peach stones and kernels, respectively, together with kernel water content (% Moisture) in peach groups that differ in origin (standard cultivars, promising hybrids and vineyard peach accessions) and two groups that differ in ripening time (early and late) are shown in Table 2. Based on physical characteristics of peach stones and kernels can be seen that promising hybrids have higher stone and kernel weight compared to vineyard peach genotypes and standard cultivars, which indicates an additive effect in the hybridization of peaches. Similar finding were reported by Zheng et al. [17] where cultivated peach stones differ from those of its known wild relatives in being significantly larger and less spherical. However, vineyard peaches have a higher weight fraction of kernel in the stone weight compared to other cultivars/hybrids (in standard cultivars kernel accounts for 3.8% of the stone weight; while in promising hybrids, it was 5.1% and in vineyard peach 9%). Kernel weight is an important trait since seedlings from heavy-weight kernels are more competitive, which is very important in fruit breeding [18].
Standard cultivars and promising hybrids had the same water content, while vineyard genotypes had a bit lower percentage of moisture. When comparing peach kernels between different ripening times (Table 2), early cultivars contained slightly more water than cultivars with late ripening time. That also means that late varieties have a higher percentage of dry matter which is essential for preserving the quality and prolonging the life of the seeds (kernels). This is expected since decrease of dry weight is a manifestation of physiological maturity of the seed (kernel), and if would not undergo this process its germination would be impossible [24]. Our results correspond to those obtained by Lipan et al. [25] who determined that almond kernels contained about 3.9% water, while the industry standards prescribe water content from 3 to 6% in raw almonds.
Statistical evaluation of differences among the peach cultivars/genotypes for each of the observed variables was checked by Kruskal-Wallis test taking the appropriate cultivar as a single factor (Table 3). Kruskal-Wallis multiplecomparison Z value test was performed when statistically significant difference between the medians were observed. Based on the Kruskal-Wallis test, ALS, ALK and AWS were variables that separate samples of promising hybrids from standard cultivars and vineyard genotypes, while AWK separates standard cultivars from other observed groups. Results of Mann-Whitney U test which assessed the similarities and differences between two groups of samples depending on ripening time showed that ALK and AWK lead to statistically significant differences between these two observed groups ( Table 3). The observations in this study agreed with the results of Femenia et al. [26] for sweet and bitter apricot kernels. Morphologically, tested peach cultivars  exhibited the main differences in the physical kernel properties, such as kernel weight, percentage of kernel in relation to the whole stone, and kernel length. On the other hand, the water content is not a good parameter for variety/genotype differentiation due to its high variability within the samples obtained from the same crops [26].

Oil and fatty acids determination in peach kernel samples
The total oil yields and fatty acids composition of the kernel's oil extracts obtained from 25 different peach cultivars/ genotypes divided in 3 groups based on its origin (standard cultivars, promising hybrids, and vineyard peaches) and ripening (early or late) are given in Table 4. According to the results from this study, a wide range of oil content can be observed for the investigated genetic pool. The lowest oil content was obtained in standard cultivars (mean value 24.42%), while the highest being found in vineyard peaches (mean 40.00%) which are also part of the late ripening cultivars. Contents of oil in standard cultivars were similar with those obtained by Wu et al. [4] who found 26% oil content in peach kernels extracted by hexane. On the other side, oil contents from kernels of promising hybrids and vineyard peaches were similar with 37.7% of kernel oil content reported by Ashraf et al. [27]. Observing the cultivars according to the ripening time (Table 4), it could be noticed that the cultivars with earlier ripening time had lower oil content (19.87%) compared to cultivars that mature later (38.71%). Extraction performed by hexane ultrasonic maceration method of peach kernels confirmed that all samples contained oleic acid, linoleic acid, palmitic acid, and stearic acid as the main fatty acids in the oil. The unsaturated fatty acids (UFA) of kernel oils were major contents, primarily oleic and linoleic acids, with relative concentration ranging from 89.50% (standard cultivars) to 91.58% (promising hybrids). Amounts of fatty acids in the peach kernel oils were in the order of MUFA > PUFA > SFA (Table 4) regardless of the origin and ripening time of peach. Thanks to the high content of monounsaturated fatty acids (MUFA) in oils, mainly oleic acid, peach kernel oils have unique characteristics, high stability and good health benefits [7,12]. According to this study, the peach kernel oil has a higher oleic acid content (mean value about 65% for all tested genotypes) compared to extra virgin olive oil (about 60%), peanut oil (about 48%) and ~ 33% in sunflower seed oil [25]. Also, the ratio of SFA and UFA was less than one in all samples, indicating on high nutrition level valuable for human health, and longer shelf life, because the rates of oxidation of fatty acids are approximately 1:10:100:200 for stearic, oleic, linoleic, and linolenic acid, respectively [28]. Obtained results agree with the previous studies which reported fatty acids composition in oils of apricot [29,30], almond [25,31] and peach [4,6,13]. The fatty acid composition was not genotype dependent, but the quantity of the most abundant fatty acids differ between the cultivars/genotypes (Table 4), so the oleic acid varied from 38.4 to 74.9%, linoleic acid from 15.7 to 44.6%, palmitic acid from 5.5 to 8.95% and stearic acid from 1.80% up to 6.40%. According to Wilson et al. [32], a desirably fatty acids composition of oils intended for frying is 7% saturates, 60% oleic acid, 31% linoleic acid and 2% linolenic acid, while desired fatty acid composition of oils intended for industrial use is 11% saturates, 12% oleic acid, 55% linoleic acid and 22% linolenic acid, which means that peach kernel oil can be intended for all kind of purposes.
Long time ago, Soler et al. [33] noticed remarkable changes regarding quantity and the content of fatty acids in dependence of the harvest stage in almond. According to the mentioned authors, the major part of oil is stored 2 months before harvest with high percentages of palmitic, linoleic, and linolenic acids in initial stages of ripening, followed by a diminution of these acids and a continuous increase of oleic acid in the final stage [33]. In peach, olive and walnut saturated fatty acids showed a decrease during fruit ripening while oleic and linoleic acids had an increase [34][35][36]. In pomegranate seeds, palmitic and linoleic acid levels increased significantly with the delay of harvest, while in avocado, oleic acid significantly increased with late harvest, while other fatty acids decreased [37,38].
For statistical evaluation of differences between the samples, the Kruskal-Wallis test was applied to three groups that differ in origin, and the Mann-Whitney U test was employed to two groups of peach genotypes that differ in the ripening time ( Table 5). Results of the Kruskal-Wallis test showed that standard cultivars differ from promising hybrids due to statistically significant differences in % oil, content of C18:0, SFA, UFA and S/U ratios. In addition, except to the content of saturated fatty acids C16:0, C20:0, C24:0 and unsaturated C16:1 acid, other tested variables related to oil and fatty acid content lead to the separation of standard peach cultivars from vineyard peach genotypes. The results of Kruskal-Wallis test indicated that standard cultivars differ from the promising hybrids and vineyard peach according to % oil in kernels and fatty acids content. According to the results of the Mann-Whitney U test (Table 5), statistically significant difference between the samples grouped by ripening time was confirmed by all variables related to the peach kernel oils and their fatty acids content, except for the amount of lignoceric acid, C24:0. The variables C17:0, C17:1 and C20:1 were not taken into account in statistical data processing due to the small and equal content in all samples. No matter that the previous knowledge showed that environmental conditions are playing very important role in kernel`s oil composition [39], the current study indicated the exceptional influence of ripening time on kernel fatty acid Table 4 Parameters of descriptive statistics of fatty acids content (%) in Standard cultivars, Promising hybrids, Vineyard peach accessions, peach cultivars with early and late ripening time  composition which goes in the same line with the finding of Summo et al. [40].

Principal component analysis
Among all classification methods, PCA is commonly used for exploratory data analysis, because it reduces the dimensionality and improves visualization of the data. The PCA was resulted in a four-component model which explains 82.16% of total variance. The first principal component, PC1, accounted for 51.60% of the overall data variance, the second one, PC2, for 15.80% and each next principal component explains less than 10% of total variance (PC3-8.75% and PC4-6.01%). Mutual projections of scores values and their loading vectors for the first two PCs are presented in Fig. 1a, 1b, respectively. Taking into account PC1 and PC2 score values (Fig. 1a), three distinctive groups of samples were obtained according to origin of the peach genotypes. The first principal component, PC1, explained separation of standard cultivars (SC) from perspective hybrids (PH) and vineyard peaches (VP). Majority of SC samples are widespread alongside the right side of PC1 axis, with the exceptions of samples 9 and 10. The separation of these two SC can be explained by their ripening (maturation) period where belong to cultivars that mature later (samples 9-25). On the other hand, PC2 explained separation two groups of local peach genotypes. On the positive side of PC2 axis (upper, left quadrant on Fig. 1a) group of PH samples was obtained, while on the negative side of PC2 axis (under, left quadrant on Fig. 1a) was located VP group of samples. From Fig. 1b, it can be seen that the separation of the SC samples is strongly influenced by saturated fatty acids (C16:0, C18:0, C24:0, and SFA), ratio S/U, linoleic acid (C18:2 n-6c) and moisture. In addition, the PH group of samples was separated based on physical characteristics of stone and kernel (ALK, ALS, AWS, AWK), while the VP samples was separated based on content of oleic acid C18:1 n-9c, thus and of MUFA and UFA. The score plot revealed that sample 4 was outlier, due the highest content of linoleic fatty acid (C18:2 n-6c).
Observing only results for fatty acids (PCA plots, Fig. 1), it can be concluded that local genotypes (PH and VP) were separated due to higher oil content in kernels, oleic acid, MUFA and UFA, while SC (except those which are part of

Correlation between variables
The correlation between parameters of appearance and parameters of nutritional value could be important and practical determinant of food quality [41]. The Pearson correlations were made to obtain whether there are some interdependencies between the stone length and weight (ALS, AWS), kernel length and weight (ALK, AWK), kernel moisture (% Moisture) as physical parameters and on the other side, kernel oil content (Oil) and four the most abundant fatty acids (palmitic-C16:0, stearic-C18:0, oleic-C18:1 n-9c and linoleic-C18:2 n-6c) as chemical composition parameters (Fig. 2). Observing results for all genotypes, SC samples, early and late ripening cultivars, it could be concluded that the analysis of correlation highlighted a positive interdependencies between all stone and kernel sizes, which was also determined by Kumar and Bhan and Milošević et al. [42,43] for wild genotypes and cultivated apricot cultivars, respectively. In the case of PH and VP, different relationships were determined.
For vineyard peach genotypes (Fig. 2) probably natural selection influenced the shape and weight of stones and kernels. In PH samples, planned hybridization, which was done to create new cultivars with higher fruit weight, did not changed kernel size according to the stone size, so just kernel weight correlated with stone length. This probably happened due to the additive effect during inheritance of some stone and kernel traits. All cultivars/genotypes studied and all separated groups (except for late peach cultivars, Fig. 2) showed positively correlations between oil content and physical traits of kernel. This direct proportion proposes that kernel enlargement would be associated with gaining total oil content. Results from this study coincided with data obtained by Deng et al. [44] for the bitter-and sweet-kernelled apricot cultivars who obtained positive correlation between kernel oil content and the physical parameters of the kernels. On the opposite, a significant and negative correlation was found between kernel weight and lipid content in almonds [40]. In current study, oil content showed positive correlation with the C18:1 n-9c, but negatively with C16:0, C18:0 and C18:2 n-6c. Our results partly correspond to those determined in Theobroma cacao by Mustiga et al. [45], and in soybean by Zhou et al. [46]. Such a contradiction between the studies could be due to the difference in material comprised in the mentioned experiments.
In our experiment, palmitic acid was positively correlated to stearic acid in all genotypes, SC samples and early ripening cultivars, but negatively correlated in PH samples and late ripening cultivars (Fig. 2). This means that in some peach kernel`s oil, the palmitic acid is not only converted into stearic acid and that stearic acid is not only dependent on palmitic acid, and vice versa [46]. Mustiga et al. [45] also found significant negative correlations between palmitic acid and stearic acid in cacao (Theobroma cacao L.) beans, while Islam et al. [47] found positive correlation between those two saturated fatty acids in rapeseed (Brassica napus) and mustard (B. juncea).
In all cultivars/genotypes and groups, palmitic acid was negatively correlated with oleic acid and positively correlated with linoleic acid (Fig. 2). Significant and positive correlation was observed between palmitic acid and oleic acid and between palmitic acid and linoleic acid in rapeseed and mustard [47]. Zhou et al. [46] determined negative correlation in C18:0/C18:1 n-9c and positive C18:0/C18:2 n-6c in soybean, while Mustiga et al. [45] both negative correlation for cacao beans. This means that synthesis and storage of fatty acids and its relationship in kernels/seeds is species dependent.
All separated groups of peaches (according both to the origin and ripening time, Fig. 2) oleic (C18:1 n-9c) and linoleic (C18:2 n-6c) fatty acids were strongly negatively correlated with each other. This negative correlation could be based on different pathways for biosynthesis of these two fatty acids [48]. On the other hand, this adverse relationship could be very interesting for the food industry and customers to produce oil with high oleic and low linoleic content [49]. According to Yu et al. [50], high-oleic-acid peanuts have an extended shelf time and have an increased health value for humans due higher oleic acid (which improves oxidative stability and increases oil functionality) and lower linoleic and lower palmitic acid compared with regular peanuts. Also, in all cultivars/genotypes, palmitic acid was in a strong positive correlation to linoleic acid, as it was previously determined in investigations done by Qin et al. [51] and Wu et al. [52]

Conclusion
In this study, both fatty acids and physical kernel and stone properties showed large divergence. No matter that high variability in total oil content among the studied cultivars and genotypes was observed, the peach kernel oil showed similar fatty acids composition regardless the origin and/or ripening time. The lowest oil content was obtained in standard cultivars, while the highest was found in vineyard peaches. The most abundant fatty acids were oleic acid (C18:1 n-9c) and linoleic acid (C18:2 n-6c) with less than 15% of total saturated fatty acids content (palmitic and stearic acids). The PCA revealed that standard cultivars were separated from promising hybrids and vineyard peach genotypes by different oil and fatty acids content, while promising hybrids were separated from vineyard peach genotypes based on physical stone and kernel properties. By observing a correlation, a positive interdependence between oil content, oleic acid with physical parameters of stone and kernel, on one side, and positive relationship between saturated fatty acids and linoleic acid, on other side, were determined. All this suggests that developing peach germplasm with multiple desired traits of vegetable oil is possible.
The findings of this study are clearly suggesting that there is a possibility of using fatty acids composition and physical properties of stones and kernels from different peach cultivars as a biomarker of origin and ripening time. Besides, knowledge of fatty acids composition in peach oil will contribute to its use in both human nutrition and nonfood applications.