Phenolic Composition and Antioxidant Capacity of Pistachio Seed Coats at Different Tree Ages Under Saline Irrigation Conditions

Abstract

Sustaining irrigated agriculture under drought conditions with alternative water sources such as saline groundwater requires understanding their effects on salt-tolerant crops like pistachio. During recent California droughts, pistachio trees planted in 2002, 2009, and 2011 were irrigated with high-saline water containing traces of boron (B) and selenium (Se). In 2018, irrigation was divided so that half of the trees received low-saline water, while the others continued under high-saline irrigation. Three years later, nuts were harvested to evaluate how irrigation quality affected seed coats, the main storage site of phenolic antioxidants. Sixty seed coat extracts from both irrigation treatments were analyzed for antioxidant capacity (ABTS, DPPH, FRAP and Folin–Ciocalteu assays). Nuts from the oldest trees (planted in 2002) had the highest antioxidant capacity. High-performance liquid chromatography (HPLC) identified gallic acid and nine flavonoids. Catechin, procyanidin B1, cyanidin-3-O-galactoside, and eriodictyol were most abundant in the oldest trees. Irrigation salinity significantly affected gallic acid, quercetin, and isoquercetin, with higher concentrations detected in seed coats from trees receiving continued high-saline irrigation. These compound-specific shifts, together with strong age-dependent patterns, provide insight into how long-term salinity exposure influences phenolic composition in pistachio seed coats.

1. Introduction

With climate change and recurring droughts, water becomes a limiting resource for irrigated agriculture [1]. Consequently, alternative water sources of varying quality must be considered to sustain irrigated agriculture. This strategy should include utilizing poor-quality waters such as saline groundwater or saline drainage water. Irrigation with saline waters, however, must be carefully managed to avoid over-salination of soil and downstream water sources [2]. In California’s western San Joaquin Valley (SJV), many of its agricultural soils were developed from marine and continental sediments, with fertile alluvial surface layers overlying low-permeability clay [3]. Irrigation in this region has mobilized naturally occurring salts, including selenium (Se) and boron (B), leading to their respective accumulation in shallow groundwater and drainage systems [4]. During the droughts in the last decade, water shortages forced growers to recycle poor-quality drainage water for irrigation to maintain irrigated agricultural production [5]. In this regard, the SJV is one of the most irrigated, productive agricultural regions in the world, but it is also one of the most vulnerable regions to drought, soil salinity, and water-quality problems. To sustain irrigated agriculture in this region of California, crop species that are tolerant to salt and B in soils and irrigation waters must be identified.

In this regard, pistachio (Pistacia vera L.) was proposed to be a nut-producing tree crop that can tolerate irrigation with salt and B-containing water [6]. Several studies have focused on the salt tolerance of pistachio seedlings and greenhouse-grown young trees [7,8,9,10,11]. However, less information is available about semi-mature and mature pistachio trees grown and irrigated with saline water under long-term field conditions [12,13,14]. Growing pistachio trees for nut production is a long-term investment, as they typically begin producing nuts after 7 to 10 years, with maximum yields around 15 years [15]. California produces the most significant share of global pistachios, with cultivated nut-producing acreage expanding annually up to 400,000 acres [16]. To support pistachio production in drought-prone California, it is vital to provide growers with relevant information about the effects of using irrigation water high in salt, B, and Se on pistachio trees at different ages of maturity.

Pistachio trees rely on several coordinated physiological mechanisms to maintain productivity under saline irrigation. Key among these are xylem ion exclusion and ion retrieval, which restrict the movement of sodium and chloride into photosynthetic tissues and favor their storage in roots and woody stems [6,12]. Both controlled rootstock studies and field observations show that pistachio can retain substantial amounts of Na⁺ and Cl⁻ in roots and basal stems while transporting comparatively smaller quantities to leaves, and that more tolerant rootstocks such as Pistacia atlantica maintain more favorable K⁺/Na⁺ and Ca²⁺/Na⁺ ratios in foliage under saline irrigation [5,6,12]. Pistachio also employs intracellular compartmentalization and osmotic adjustment, including increases in proline and soluble sugars under salt stress, which support water balance and limit oxidative injury [7,9]. These ion-management and metabolic responses, together with the buffering capacity of a large perennial woody biomass, help explain why pistachio generally tolerates higher salinity than many other nut crops, even though the exact efficiency of these mechanisms varies among rootstocks.

Micronutrients that commonly accompany salinity in drainage waters in the westside of California’s SJV, particularly B and Se, can further influence these responses. Moderate B levels support membrane stability, cell-wall function, and antioxidant activity [17], and B amendments have been shown to enhance glycine betaine accumulation, and reduce salt-induced leaf damage in pistachio seedlings [11]. Although Se has not been tested directly in pistachio, work in soybean and other crops shows that low concentrations of Se or Se + B can strengthen antioxidant defenses, improve leaf hydration, and reduce oxidative damage during salt exposure [18,19]. At non-toxic concentrations, both B and Se can promote the accumulation of endogenous protective metabolites, including proline, glycine betaine, and soluble sugars, and may therefore contribute positively to the overall resilience of pistachio orchards irrigated with moderately saline drainage waters.

Generally, the edible part of a pistachio nut consists of a kernel and a thin seed coat that surrounds the kernel like a skin. Pistachio nuts are regarded as a healthy snack due to their heart-healthy fatty acid profile and high content of dietary fiber, protein, minerals, and antioxidants [15,20]. Health benefits of pistachio consumption were proposed to include reduced risk of cardiovascular conditions, digestive disorders, and cancer [21,22,23,24]. The high antioxidant content of pistachios can decrease the impact of oxidative stress, as shown in human and animal studies [24,25,26]. A major contributor to the antioxidant capacity of pistachios are phenolic compounds, which are effective in scavenging reactive oxygen species (ROS) and other radicals [15,27]. Several studies demonstrated that the pistachio seed coat has a much higher antioxidant capacity and phenolic content than the kernel [28,29,30]. The seed coat protects the kernel from oxidative damage, microbial infections, and other environmental stress conditions [31,32]. It is still unknown to what extent the pistachio seed coat, as a major storage site for phenolic compounds, is affected by irrigation with high-saline waters.

For some plants, salinity stress was found to increase production of natural antioxidants, specifically phenolic compounds [33,34,35]. Abiotic stress, including salinity stress, results in the generation of ROS. Given the radical-scavenging properties of phenolic compounds [36], their enhanced biosynthesis can mitigate oxidative damage and improve plant stress tolerance [37]. Beyond modulating ROS concentration levels, phenolic compounds influence protein kinase pathways and the distribution of plant hormones such as abscisic acid and auxins [38,39]. Acting both as antioxidants and as signaling molecules, phenolic compounds are central to stress adaptation. Consequently, moderately salt-stressed plants may accumulate higher levels of health-promoting phenolics, rendering them more appealing for health-conscious consumers [35,40].

Phenolic compounds include flavonoids, phenolic acids, stilbenoids, and lignans [36]. Among these, flavonoids are the most abundant and diverse, with over 6000 known structures. They are further classified into subgroups such as flavones, flavonols, flavan-3-ols, flavanonols, flavanones, isoflavones, and anthocyanins. The structural features of flavonoids influence their functional properties. Among flavonoids, the dihydroxy flavonol quercetin exhibits the highest antioxidant capacity due to its multiple hydroxyl groups and conjugated electron systems that improve radical-scavenging efficiency [36]. The ortho-dihydroxy group in the flavonol quercetin also enhances its ability to bind to kinases and transport proteins, making it a potent modulator of plant signaling pathways [41,42]. Flavonols, such as quercetin and kaempferol, have gained the most attention as potential plant stress mediators [43]. Given the structural diversity and functional significance of phenolic compounds, a detailed compositional analysis—such as through high-performance liquid chromatography (HPLC)—is essential for understanding their roles in plant stress responses.

This study is part of a long-term investigation of saline irrigation management in mature pistachio orchards of different ages located in the Grassland Basin Water District near Firebaugh, CA, USA. The experiment offers a rare field-scale opportunity to evaluate pistachio responses after years of irrigation with agricultural drainage water followed by a gradual transition to improved water quality. The drainage water, typical of the westside of the SJV, contains naturally elevated sodium and chloride concentrations that are approximately tenfold higher than the good-quality canal water available for purchase when supplies permit, and reflects the saline conditions under which drought-susceptible orchards on the westside of the SJV were originally managed. Beginning in 2018, portions of each orchard block were switched to low-saline canal water as part of a gentle reclamation effort, while other sections continued to receive drainage water to maintain the long-term comparison.

Within this context, the present study focuses on the seed coat as the principal source of antioxidants contributing to pistachio nutritional quality. We evaluated the total antioxidant capacity of seed coat extracts and quantified specific phenolic compounds using HPLC to determine how metabolite levels varied with tree age under continued saline irrigation versus improved, low-saline conditions. Although several studies have explored pistachio salt tolerance physiologically, few have examined field-based, long-term transitions from high-saline to low-saline irrigation and their respective impact on phenolic metabolism and antioxidant capacity.

2. Materials and Methods

2.1. Field Design

Pistachio (Pistacia vera) trees (Kerman grafted to PG-1 rootstock) were planted in 2002, 2009, and 2011, respectively, on a large orchard (100–120 acres) located in the Grassland Basin Water District near Firebaugh, California (Figure 1). From their initial planting, all trees were irrigated with agricultural drainage water (3–7 dS/m) containing trace concentrations of selenium (Se) and boron (B). In 2018, a saline reclamation trial was initiated to evaluate gradual recovery under improved irrigation quality. For the research study, specific planting blocks were selected, and each planting block was divided into two sub-blocks, each comprising 105 trees: one continued receiving the same high-saline drainage water, while the other was switched to low-saline canal water (<1.0 dS m⁻¹). The treatments were maintained from 2018 to 2022.

The two natural water sources in the westside of California’s SJV differed markedly in ionic composition. The canal water contained 25–66 mg Cl L⁻¹, 31–72 mg Na L⁻¹, 11–41 mg Ca L⁻¹, 5–18 mg Mg L⁻¹, and 9–55 mg S L⁻¹ (pH 7.5–8.0), while the drainage water contained 575–950 mg Cl L⁻¹, 356–670 mg Na L⁻¹, 218–370 mg Ca L⁻¹, 56–112 mg Mg L⁻¹, and 270–545 mg S L⁻¹ (pH 7.8–8.4). The high-saline water also contained higher concentrations of B (4–9 mg B L⁻¹) and Se (75–150 µg Se L⁻¹), whereas the low-saline water contained 0.5–1.0 mg B L⁻¹ and 5–10 µg Se L⁻¹. These differences reflect the elevated ionic strength and salt load typical of reclaimed irrigation waters in the SJV.

Pistachio nuts were mass-harvested with commercial shakers. Three years after the initiation of the reclamation trial, seed coat extracts were prepared from nuts harvested in 2021 to assess biochemical responses to contrasting irrigation salinities.

2.2. Seed Coat Extract Preparation

Nut samples collected in September 2021 were separately packaged according to saline irrigation treatment and planting year, shipped on dry ice to California State University, East Bay (Hayward, CA, USA), and stored at −80 °C until analysis as described below.

Pistachio nuts (separated from drupes but still in their hard shell) were placed for five days in a drying oven set to 50 °C. A total of 60 samples were processed with 10–14 pistachio nuts per sample. After removing the hard shell with gloved hands, the nuts were placed in 50 mL Falcon tubes (Corning, Glendale, AZ, USA) and vortexed to separate the seed coat from the kernels. The additional seed coat was removed with gloved hands. The average weight of the seed coat was 0.7 ± 0.3 g per sample. The pistachio seed coat was crushed with a hand-held homogenizer that fitted into the 50 mL Falcon tubes. A volume of 20 mL HPLC-grade methanol was added portion-wise, while grinding the seed coat. Each ground sample was placed into a sonicator (Branson Ultrasonics Corporation, Brookfield, CT, USA) for 30 min. After letting the debris settle, the supernatant was further clarified by centrifugation at 15,000 rpm for 20 min. The supernatants were transferred into HPLC vials and 96-well sample blocks to conduct biochemical assays described below.

2.3. Antioxidant Capacity Assays

The procedures described by Thaipong and coworkers [44] were scaled down to a microplate format and slightly modified for our samples. The 2,2-azinobis (3-ethyl-benzothiazoline-6-sulfonic acid) (ABTS) reagent was prepared by dissolving 100 mg ABTS and 16 mg potassium persulfate in 50 mL dH₂O. After an overnight incubation in the dark at room temperature, the ABTS reagent was diluted 10-fold with deionized water. All seed coat samples were diluted 20-fold with methanol. A volume of 200 µL ABTS reagent and 10 µL 20-fold diluted sample or standard were mixed in a microplate and incubated for 60 min in the dark before recording the absorbance values at 734 nm with a Synergy H1 microplate reader from Biotek (Winooski, VT, USA). A ten-point standard curve ranging from 0 to 1 mM Trolox prepared in methanol was used for analysis.

The 2,2-diphenyl-1-picrylhydrazyl (DPPH) reagent was prepared by dissolving 5.0 mg DPPH in 100 mL ethanol. This reagent was stored at −20 °C. A volume of 200 µL DPPH reagent and 10 µL 20-fold diluted sample or Trolox standard (0–1 mM) were mixed in a microplate and incubated for 120 min in the dark before recording the absorbance values at 517 nm.

The reagent for the ferric ion reducing antioxidant power (FRAP) assay consisted of a 10:1:1 mixture of 300 mM acetate buffer, pH 3.6, 10 mM 2,4,6-tripyridyl-1,3,5-triazine (TPTZ) in 40 mM HCl, and 20 mM iron(III) chloride and was prepared fresh and pre-heated to 37 °C before use. A volume of 200 µL FRAP reagent and 10 µL 20-fold diluted sample or Trolox standard (0–1 mM) were mixed in a microplate and incubated for 30 min in the dark before recording the absorbance values at 593 nm.

2.4. Folin–Ciocalteu Assay

The Folin–Ciocalteu assay was performed according to the protocol described by Ainsworth and Gillespie [45]. All samples were diluted 20-fold with methanol. Gallic acid standards were prepared in methanol in a concentration range of 0 to 0.8 mM.

2.5. Aluminum(III) Chloride Complexation Assay

Two aluminum(III) chloride reagents were prepared with and without sodium acetate by mixing 1.2 mL 10% w/v aluminum(III) chloride solution with either 25.8 mL methanol or with a mixture of 8 mL saturated sodium acetate solution and 17.8 mL methanol. In 96-well microplates, 270 µL of these reagents were mixed with 30 µL 20-fold diluted seed coat extract or quercetin standard solutions prepared in a concentration range of 0–100 µg/mL. Absorption spectra from 350 nm to 700 nm were recorded with a Synergy H1 microplate reader from Biotek (Winooski, VT, USA).

2.6. HPLC

Phenolic standards, including gallic acid, catechin, leucocyanidin, quercetin, kaempferol, and rutin, were purchased from Fisher Scientific (Pittsburgh, PA, USA). Procyanidin B1, procyanidin B2, cyanidin-3-O-glucoside, cyanidin-3-O-galactoside, isoquercetin, myricetin, myricetin 3-galactoside, eriodictyol 7-glucoside, naringin (naringenin 7-neohesperidoside), taxifolin, and aromadendrin were obtained from Cayman Chemical Company (Ann Arbor, MI, USA). Cyanidin chloride, eriodictyol, naringenin, luteolin, and apigenin were purchased from Indofine Chemical Company (Hillsborough, NJ, USA). Phenolic standard stock solutions of 100 µg/mL were prepared in methanol or ethanol. Dilution series ranging from 0.2 to 20 µg/mL were used to generate calibration curves.

An InfinityLabPoroshell120 EC-C18 column (4.6 × 250 mm, particle size 4 μm) was attached to a 1260 Infinity system with an autosampler. The column and HPLC instrument were procured from Agilent (Santa Clara, CA, USA). A solvent gradient with 3% (v/v) acetic acid (solvent A) and methanol (solvent B) was applied: 0–5 min 100% solvent A, 5–65 min linear gradient 0–60% solvent B, 65–70 min 60% solvent B, 70–71 min linear gradient 60–0% solvent B, 71–80 min 100% solvent A. The flow rate was set to 1 mL/min. The temperature of the column compartment was maintained at 25 °C. The injection volume was 2 µL. The diode array detector recorded signals at 260, 280, 292, 320, 370, and 530 nm. The fluorescence detector was set to 276 nm for the excitation and 316 nm for the emission wavelength.

2.7. Data Analysis

Data analysis was performed with the program RStudio (Version 2025.09.2+418) [46]. We tested for normal distribution (Shapiro–Wilk test) and equal variances (Levene’s test). Because multiple variables deviated from normality and two showed unequal variances, we analyzed all responses using a two-way aligned rank transform ANOVA (ART ANOVA) [47]. Sidak-adjusted pairwise comparisons were used to test differences among factor levels [48]. Parametric two-way ANOVA yielded equivalent conclusions which is presented in the Supplementary Materials.

Pearson correlation coefficient calculations [49], a principal component analysis [50], and hierarchical cluster analysis with a heatmap [51] were performed to explore relationships among measured quantities.

3. Results

3.1. Antioxidant Capacity and FC Assays

Three different antioxidant capacity assays (ABTS, DPPH, and FRAP) and a Folin–Ciocalteu (FC) assay were performed with the sixty seed coat samples (Figure 2). According to the results of statistical tests from Table S1, no significant differences were observed for seed coat samples from trees with different saline irrigation treatments (all p > 0.025). Only the planting year (e.g., 2002, 2009, 2011) was a significant factor (p < 0.0001). The interaction term between treatment and planting year was not significant (p > 0.13). Pistachio seed coats from trees planted in 2002 had the highest antioxidant capacity and the highest phenolic content, irrespective of saline irrigation treatment. Seed coats from trees planted in 2009 and 2011 had virtually identical antioxidant capacity and phenolic contents.

Mean values across all samples were 149 ± 49 µmol Teq/g (ABTS), 134 ± 56 µmol Teq/g (DPPH), and 112 ± 32 µmol Teq/g (FRAP). Total phenolic content averaged 114 ± 44 µmol GAeq/g. All three antioxidant capacity assays (ABTS, DPPH, FRAP) and the FC assay showed very strong and highly significant pairwise correlations (p < 0.001). Pearson coefficients were 0.92 for DPPH and ABTS, 0.92 for DPPH and FC, 0.90 for DPPH and FRAP, 0.88 for ABTS and FRAP, 0.88 for ABTS and FC, and 0.78 for FRAP and FC. An analysis of the correlation coefficients separated by planting year can be found in Supplementary Figure S1.

We performed a quantification of total flavonoids using the aluminum(III) chloride complexation assay and observed considerable variability in color development across all samples, ranging from yellow to purple (see Supplementary Figure S2). Given its strong structural dependence, this assay was deemed unsuitable for quantifying total flavonoids in pistachio seed coats. This observation anticipated later HPLC results, which confirmed substantial differences in flavonoid composition across samples.

3.2. Effects of Planting Year and Saline Irrigation Treatment on Individual Phenolic Compounds

We quantified gallic acid and nine flavonoids in sixty seed-coat extracts. The HPLC parameters for the detected compounds are summarized in Supplementary Table S2. Mean concentrations (µg/g ± SD), listed in order of decreasing abundance, were: catechin (273 ± 101), gallic acid (239 ± 199), cyanidin-3-O-galactoside (92 ± 55), procyanidin B1 (45 ± 17), taxifolin (44 ± 32), epicatechin (32 ± 12), luteolin (28 ± 10), eriodictyol (27 ± 14), quercetin (14 ± 9), and isoquercetin (quercetin-3-O-glucoside; 7 ± 4). Cyanidin-3-O-glucoside was detected in 56 of 60 samples at 8.2 ± 4.1 µg/g and was excluded from further analyses due to incomplete detection. Other screened phenolics, including protocatechuic acid, epicatechin gallate, myricetin, myricetin-3-O-glucoside, rutin, kaempferol, eriodictyol-7-O-glucoside, naringenin, naringenin-7-O-neohesperidoside, and apigenin, were not detected in our samples.

To investigate whether planting year (2002, 2009, 2011) and/or saline irrigation treatment affected the content of the ten detected phenolic compounds, we applied an aligned rank transform (ART) ANOVA followed by Sidak-adjusted pairwise comparisons. The results are summarized in Figure 3 and in Supplementary Table S3.

The factor planting year had a statistically significant effect on five compounds: gallic acid, catechin, procyanidin B1, cyanidin-3-O-galactoside, and eriodictyol (p < 0.05). For all five compounds, seed coats from trees planted in 2002 contained significantly higher amounts than those from trees planted in 2009 and 2011. No significant differences were observed between the 2009 and 2011 planting years. In contrast, taxifolin, luteolin, and epicatechin did not vary significantly with planting year.

The factor irrigation treatment had a significant effect on three compounds: gallic acid, quercetin, and isoquercetin (p < 0.05). For all three compounds, the Sidak contrasts indicated higher concentrations in samples from trees that continued to receive high-saline irrigation, compared to those irrigated with low-saline water for the three-year reclamation period. The remaining seven compounds showed no significant treatment effects.

No significant interaction effects between planting year and irrigation treatment were detected for any of the ten quantified phenolic compounds.

3.3. Correlation and Cluster Analyses

Pearson correlation analysis (Figure 4A) revealed several strong and highly significant relationships among the ten detected phenolic compounds. The three most abundant flavonoids (catechin, procyanidin B1, and cyanidin-3-O-galactoside) formed the most prominent correlation group (highlighted in magenta in Figure 4), with pairwise correlations ranging from r = 0.71–0.87 (p < 0.001). Epicatechin showed moderate correlations with catechin (r = 0.50, p < 0.001) and procyanidin B1 (r = 0.47, p < 0.001). Since procyanidin B1 is a dimer containing epicatechin and catechin, some correlation among these three compounds is expected. In the cluster analysis, however, epicatechin grouped with luteolin (r = 0.54, p < 0.001) as indicated by the orange-colored box in Figure 4A. A second strong correlation pair was observed between eriodictyol and taxifolin (r = 0.71, p < 0.001), shown in teal in Figure 4A. The flavonols quercetin and isoquercetin exhibited one of the strongest correlation coefficients in the dataset (r = 0.80, p < 0.001); this close association is expected because isoquercetin is a glycosylated form of quercetin and typically follows similar abundance patterns. These two compounds appear in the blue cluster of Figure 4.

Across all pairwise comparisons, no strong negative correlations were detected among the ten phenolic compounds. The most negative values occurred between cyanidin-3-O-galactoside and quercetin (r = −0.10, p > 0.05) and between cyanidin-3-O-galactoside and gallic acid (r = −0.07, p > 0.05), but these weak and non-significant correlations indicate no meaningful inverse relationship. Thus, within the range of variations observed here, none of the phenolic compounds appeared to vary at the expense of another.

Correlations between phenolic compounds and antioxidant/FC assays showed a similar pattern. Catechin, procyanidin B1, and cyanidin-3-O-galactoside, which were the three most abundant flavonoids, were also most strongly associated with antioxidant capacity. Their correlations with ABTS and DPPH were r = 0.70–0.80 (p < 0.001), with FRAP were r = 0.58–0.77 (p < 0.001), and with FC were r = 0.68–0.76 (p < 0.001). Other compounds showed only moderate correlations with the antioxidant assays (r = 0.14–0.52, p < 0.05 to < 0.001). Although gallic acid was the second most abundant phenolic compound, it displayed comparatively weak correlations with ABTS, DPPH, FRAP, and FC (r = 0.23–0.29, p < 0.05), indicating a more limited connection to antioxidant variation than the major flavonoids.

Hierarchical clustering (Figure 4B) grouped the phenolic compounds into three main clusters based on similarities in their abundance patterns. The first group contained the closely related flavan-3-ols and the anthocyanidin glycoside (catechin, procyanidin B1, and cyanidin-3-O-galactoside) which also showed the strongest internal correlations. A second cluster consisted of the two flavanones, eriodictyol and taxifolin, which varied similarly across samples. The third cluster included the flavonols and gallic acid (quercetin, isoquercetin, gallic acid), which shared strong associations with one another. In addition to these three main groups, epicatechin and luteolin formed a smaller, separate cluster, indicating that their abundance patterns differed from both the flavan-3-ol cluster and the flavonol/flavanone clusters.

The agreement between Figure 4A (correlation matrix) and Figure 4B (dendrogram topology) highlights coordinated variation among compounds that are neighboring products within the flavonoid biosynthetic pathway. Principal component analysis (PCA) and heat map clustering (Supplementary Figures S3 and S4) did not reveal additional structure among the samples, but their compound-level clustering was consistent with Figure 4.

4. Discussion

4.1. Antioxidant Capacity Assays and Their Interpretation

Despite the large ionic contrast between drainage water and canal water, antioxidant capacity remained statistically unchanged across irrigation treatments. This indicates that the seed coat, already known to be a rich source of reducing agents and radical scavengers, maintains a robust and resilient antioxidant reservoir even after prolonged exposure to high salinity. All three antioxidant capacity assays (ABTS, DPPH, and FRAP) and the FC assay showed a strong effect for the factor planting year, with the highest values found in trees planted in 2002 and no difference for trees planted in 2009 and 2011. At the time of nut harvest and subsampling in 2021, trees planted in 2002, 2009, and 2011 were 19, 12, and 10 years old, respectively. The factor of tree age was also considered by Fabani and coworkers: slight age-dependent variations were detected for total flavonoid and anthocyanin content in pistachio nut extracts from Argentinian cultivars aged 5, 9, and 11 years [27].

The total phenolic content obtained in our study was 114 ± 44 µmol GAeq/g, corresponding to 19.4 ± 7.5 mg GAeq/g. Grace and coworkers obtained 29.12 ± 1.32 mg GAeq per g skin tissue for unsalted, roasted pistachios of the Kerman cultivar [29]. In contrast, Tomaino and coworkers reported higher values for their seed coat extracts from pistachios of the Bronte variety, with 116.32 ± 8.54 mg GAeq/g for the FC assay and 2.19 ± 0.14 mmoles Teq/g for the ABTS assay [28]. It is important to note that the FC method does not exclusively detect phenolic compounds, as other reducing substances can also contribute to color development [52,53]. Therefore, the FC values represent the overall reducing capacity of the extracts rather than phenolics alone. Moreover, different phenolic compounds do not respond equally in the FC assay; purified standards produce varying color intensities at identical molar concentrations [53,54].

The two assays based on quenching of an organic radical (ABTS and DPPH) yielded average values of 149 ± 49 and 134 ± 56 µmol Teq/g. The FRAP assay, which is based on the reduction of iron(III) to iron(II) and subsequent complexation of iron(II) by TPTZ, resulted in a lower average value of 112 ± 32 µmol Teq/g. In contrast to the single-electron transfer mechanism of FRAP, ABTS and DPPH operate through a mixed mode that encompasses both hydrogen atom transfer and single-electron transfer [55]. The outcome of each assay is further influenced by the reduction potential and steric constraints of the main assay reagent, as well as the polarity and pH of the reaction medium. Because samples contain complex mixtures of diverse antioxidants, antioxidant capacity is best assessed with multiple assays rather than a single method [55].

We were unable to use the aluminum(III) complexation assay to quantify flavonoids as we obtained a range of different colors for our samples. As demonstrated by Shraim and coworkers, who used quercetin, catechin, and rutin as model flavonoids, the ability to form colored complexes with aluminum(III) is highly dependent on flavonoid structure [56]. Given the structural specificity of this assay, the variation in color intensity and hue among our samples strongly suggests differences in their flavonoid composition. This was later corroborated via HPLC.

4.2. Phenolic Composition in the Context of Previous Studies

We based our identification of specific phenolic compounds on five published studies that employed HPLC analysis of pistachio seed coats [27,28,29,30,57], as well as on the KEGG metabolic map for flavonoid biosynthesis in Pistacia vera (https://www.kegg.jp/pathway/map00941, accessed on 6 October 2025). Figure 5 compares the average quantities of phenolic compounds quantified in our study alongside those reported in the five reference studies. Among the six studies (including ours), gallic acid and catechin were consistently detected at high concentrations (>75 µg/g) in five of them. A study on whole pistachio nuts also reported high levels of both compounds [58]. Cyanidins were only detected in glycosylated forms and at highly variable levels: cyanidin-3-O-galactoside ranged from 21 to over 5800 µg/g and cyanidin-3-O-glucoside from <1 to over 200 µg/g. We detected cyanidin 3-O-galactoside in all samples (92 ± 55 µg/g), while cyanidin-3-O-glucoside was only found in 56 of 60 samples (8.2 ± 4.1 µg/g). Phenolic compounds found in moderate concentrations (>10 µg/g) in at least three of the six studies, including ours, were epicatechin, eriodyctiol, luteolin, quercetin, isoquercetin, and a procyanidin dimer. Using procyanidin B1 and B2 standards, we identified the procyanidin dimer in our samples as procyanidin B1, which consists of epicatechin-(4β→8)-catechin, while procyanidin B2 (epicatechin-(4β→8)-epicatechin) was not present. Grace and coworkers also detected procyanidin B1 along with several other procyanidin conjugates and were able to quantify five different glycosylated quercetin derivatives in their samples [29].

Tomaino and coworkers reported high levels of naringin (118.82 ± 9.64 µg/g) and eriodictyol 7-glucoside (365.68 ± 13.56 µg/g) in seed coats from pistachio of the Bronte variety [28]. We screened for both compounds but did not detect them. Eriodictyol 7-glucoside was detected in moderate levels (59.0 ± 2.7 µg/g) by Grace and coworkers [29]. Without specifying the sugar entity, small levels (3.56 µg/g) of eriodictyol-7-hexoside were measured by Fabani and coworkers [27]. A novel finding of our study was the detection and quantification of taxifolin in pistachio seed coats. Compounds not detected in our analysis, but reported in at least one of the five comparison studies, included myricetin, kaempferol, rutin, naringenin, and apigenin. Additionally, several intermediates in the flavonoid biosynthesis pathway, such as leucocyanidin, un-glycosylated cyanidin, and aromadendrin (dihydrokaempferol), were included in our screening but not detected.

The phenolic profile of our samples most closely matched that of seed coats from unroasted Argentinean pistachios (Pistacia vera var. Kerman) analyzed by Fabani and coworkers [27]. As previously reported, phenolic composition in pistachios is strongly influenced by cultivar, geographic location, post-harvest processing, season, and other environmental factors [30,59]. Among all reported studies, the Sicilian Bronte pistachio seed coats stood out for their elevated levels of several phenolic compounds [28]. The remaining studies investigated Californian Kerman pistachios, but differing post-harvest treatments, extraction protocols, and analytical methods made detailed cross-study comparisons impractical.

4.3. Correlation and Cluster Analyses in Relation to Flavonoid Pathway Position

To interpret the correlation and clustering patterns in the context of biosynthetic relationships, it is useful to consider where each quantified metabolite sits within the flavonoid pathway. Figure 6 provides this pathway map along with the molecular structures. To highlight the correspondence between pathway position and statistical grouping, we applied the same color scheme used in Figure 4.

We quantified ten phenolic compounds, nine of which are flavonoids. Many of these flavonoids are directly connected within the metabolic network shown in Figure 6, and several share close biosynthetic relationships. For example, catechin, glycosylated cyanidin, procyanidin B1, and epicatechin were detected at medium to high concentrations, and all four share leucocyanidin as a common precursor. Although leucocyanidin itself was not detected, the accumulation of its downstream products indicates that substantial metabolic flux must have passed through this intermediate.

In most cases, metabolites placed near each other in the pathway (Figure 6) also clustered together statistically (Figure 4). A particularly strong example is the pair eriodictyol and taxifolin, which showed a significant positive correlation (r = 0.71, p < 0.001). Their biochemical connection is direct: flavanone 3-hydroxylase (F3H) converts eriodictyol to taxifolin in a single step. The tight co-clustering therefore reflects a true biosynthetic dependency rather than noise or co-occurrence by chance. One exception was epicatechin, which did not join the cluster containing catechin and procyanidin B1 despite sharing the same upstream precursor, leucocyanidin. This deviation suggests either differential downstream regulation or differences in turnover and glycosylation rates. Aside from this case, the clustering patterns (Figure 4) closely mirrored the pathway map (Figure 6), reinforcing the idea that the structure of the biosynthetic pathway influences the observed statistical patterns.

Several medium-level metabolites (luteolin, eriodictyol, taxifolin, and quercetin) are all generated by hydroxylation reactions mediated by either flavonoid 3′,5′-hydroxylase (F3′5′H) or flavonoid 3′-hydroxylase (F3′H). Notably, none of the known substrates of these enzymes (apigenin, naringenin, aromadendrin, or kaempferol) were detected in our samples. Because myricetin, a product specific to F3′5′H, was also absent, we speculate that F3′H activity dominates over F3′5′H in pistachio seed coats under our conditions.

A biosynthetic branching point of particular relevance to our dataset is the reaction catalyzed by flavonol synthase (FLS). FLS converts dihydroflavonols into flavonols such as quercetin and kaempferol. If F3′H activity is high, kaempferol (a 3′-hydroxylation product) would be rapidly converted to quercetin, preventing its accumulation. This is consistent with our data: quercetin and isoquercetin were the only two flavonoids that showed a subtle but statistically significant increase in trees irrigated with high-saline water compared to those under low-saline irrigation for three years. These increases may be linked to altered FLS flux or enhanced precursor availability under sustained salt stress. Other studies have shown that overexpression of FLS in plants such as alfalfa, cotton, and cress enhances flavonol accumulation and improves performance under saline or saline–alkali stress [60,61,62]. The parallels with our findings raise the possibility that FLS activity in pistachio may similarly contribute to salt-stress adaptation, either by modulating flavonol abundance or by influencing flux through key branches of the flavonoid pathway.

4.4. Selective Phenolic Adjustments Under Salinity: Roles of Quercetin, Isoquercetin, and Gallic Acid

Salinity influenced only a narrow subset of the quantified phenolic compounds, specifically quercetin, isoquercetin, and gallic acid. This pattern indicates that pistachio seed coats respond to prolonged saline irrigation not by expanding the total antioxidant pool, but by adjusting particular branches of the phenolic network. These selective shifts differ from the large and age-dependent variations observed for catechin, procyanidin B1, and cyanidin glycosides and are consistent with reports that chronic salinity or combined stressors often alter specific flavonol or phenolic acid branches rather than eliciting broad increases in all phenolics [33,63,64].

Quercetin and isoquercetin were the only flavonoids to show a consistent and statistically significant enrichment under high-saline irrigation. Although present at lower absolute concentrations than catechin or procyanidin B1, these flavonols have a chemical structure with two ortho hydroxy-groups, which enhances radical-scavenging activity [36] and enables interactions with proteins involved in stress-related signaling [41,42]. Similar compound-specific increases in quercetin-type flavonols have been reported under diverse oxidative or ionic stress conditions, including heat- and salt-challenged tomato [63], UV- and salinity-treated Ligustrum vulgare [64], and rice exposed to simultaneous heat and salt stress [65].

In pistachio, the seed coat acts as a protective barrier that limits oxidative and environmental damage to the kernel [31,32]. Within this barrier, flavonols such as quercetin and isoquercetin may help maintain local redox balance during prolonged salinity. Because total antioxidant capacity did not differ between irrigation treatments, the observed enrichment of these flavonols likely reflects a compositional shift toward phenolics that are effective in membrane stabilization, ROS scavenging, or modulation of stress-signaling pathways [39]. This interpretation aligns with studies showing that enhanced flux through the flavonol branch via increased FLS activity supports salinity or combined-stress tolerance in alfalfa, cotton, and cress [60,61,62]. Pistachio may rely on analogous localized adjustments to preserve seed-coat integrity under saline irrigation.

Gallic acid also increased modestly but significantly under high-saline irrigation, suggesting an additional adjustment outside the flavonol branch. As a hydroxybenzoic acid with high reducing capacity [36], gallic acid contributes to the antioxidant properties of the seed coat. Although its correlations with antioxidant assays were weaker than those of catechin or procyanidin B1, its salinity-linked increase supports the interpretation that pistachio seed coats undergo selective, rather than generalized, phenolic adjustments. In several nut and seed systems (e.g., cashew seed coats), free gallic acid can arise not only from de novo biosynthesis but also from hydrolysis or oxidative breakdown of galloylated flavan-3-ols [31]. Stress-related turnover of such compounds provides a plausible explanation for the observed increase in gallic acid even when total phenolic content remains unchanged.

Together, the increases in quercetin, isoquercetin, and gallic acid indicate that long-term saline irrigation leads to targeted changes within the seed-coat phenolic profile. Total antioxidant capacity remained unchanged, suggesting that pistachio relies on adjustments in specific compounds rather than expanding the overall antioxidant pool. These patterns align with pistachio’s established strategies for regulating Na⁺ and Cl⁻ through root-level exclusion, controlled xylem loading, and vacuolar sequestration, which together limit ion-induced oxidative pressure. When considered alongside these ion-homeostasis mechanisms, the compound-specific responses observed here offer insight into how pistachio maintains seed-coat integrity under saline irrigation. Quercetin-type flavonols, in particular, may serve as indicators of localized redox adjustments and salinity exposure in mature pistachio orchards.

4.5. Study Limitations and Perspectives

This study focused on a single harvest season and did not include direct measurements of ion accumulation or other physiological indicators of salt stress such as sodium, chloride, or oxidative stress markers. Because only the seed coat was analyzed, potential metabolic adjustments in other tissues—such as leaves, hulls, or roots—remain unresolved. The dataset therefore provides a biochemical snapshot of a long-term field system rather than a temporal profile of salinity adaptation. Field-based sampling is also influenced by environmental variability, including seasonal weather conditions, pest pressure, and soil heterogeneity, which may contribute to sample variability. Multi-year sampling that couples ion analysis with tissue-wide metabolite profiling will be needed to link phenolic composition more directly to physiological and agronomic outcomes.

Within these constraints, the present study shows that total antioxidant capacity in pistachio seed coats is largely maintained under prolonged saline irrigation, with selective adjustments confined to quercetin, isoquercetin, and gallic acid. These compound-specific patterns point to a salinity response that is compositional rather than global and highlight the value of examining individual phenolic components instead of relying solely on aggregate antioxidant metrics. Pairing metabolite profiling with transcript levels of key flavonoid biosynthetic genes could help identify the regulatory steps contributing to these shifts. After further validation of quercetin and isoquercetin as consistently enriched flavonols under salinity, these compounds may hold potential as indicators of localized redox adjustments in saline-stressed pistachios. Finally, relating seed-coat phenolic profiles to nut-quality traits and field performance will help clarify their relevance for irrigation management and for sustaining productivity in saline agroecosystems.

5. Conclusions

In our study with nuts harvested from pistachio trees planted in different years and grown in a real-world saline environment, we observed that planting year was the deciding factor that influenced the overall antioxidant capacity and concentrations of several major phenolic compounds, such as catechin and procyanidin B1, in the seed coats of pistachio nuts. The oldest trees planted in 2002, which were 19 years old at the time of nut harvest, yielded pistachio nuts with seed coats of the highest antioxidant consumer benefit. Continuous irrigation with high-saline water, compared to a three-year intervention or gentle reclamation by irrigation with low-saline water, did not influence overall antioxidant capacity, but instead caused a subtle shift in the phenolic composition of the seed coat samples. Gallic acid, quercetin, and isoquercetin were significantly higher in samples from trees maintained under high-saline irrigation, and these increases were most pronounced in the oldest trees. These patterns indicate that salinity affects specific branches of phenolic metabolism rather than inducing broad changes across all phenolic compounds.

Among the responsive compounds, quercetin derivatives show particular promise for tracking localized metabolic adjustments to salinity in mature pistachio orchards. Further work linking these compositional changes to physiological traits and field performance will help clarify their utility as biochemical indicators of salinity exposure and adaptation in long-lived perennial crops such as pistachio.