Effect of Fusarium proliferatum infection on physiological, phytochemical, and nutrient responses in garlic

Plant and fungal materials

Three garlic genotypes from different regions of Türkiye, where garlic has been grown for a long time, were studied. Plant samples were collected from garlic common storages, which were owned by garlic growers, in August–September 2022. Sampling method was done according to Mondani et al. (2021). Garlic genotypes were harvested at the end of July 2022, when plants reached their physiological maturity. After harvest, garlic were sundried on field ground for two weeks, then garlic bulb samples representing three different genotypes were collected from a total of 13 storage facilities across three provinces in Türkiye. Specifically, 64 bulbs of the Local-Konya genotype were obtained from five storage houses located in Konya, 46 bulbs of the Babaeski-Kırklareli genotype were collected from six storage houses in Kırklareli, and 49 bulbs of the Iranian-Balıkesir genotype were sourced from two storage houses in Balıkesir. Details about location, genotype characteristics and number of surveyed storages and samples were given in Table 1.

The disease ratio (%) was calculated by dividing the number of garlic bulbs identified as infected with F. proliferatum by the total number of bulbs examined for each genotype and then multiplying the result by 100. This value reflects the proportion of visibly or molecularly confirmed infected samples within each genotype group.

The basal part of garlic bulbs was used for the isolation of F. proliferatum. The disease exhibiting garlic cloves were first washed under tap water and the small explants containing symptomatic tissue were cut into 0.5–1 cm pieces with the help of sterile scalpel. Explants were soaked in 1.5% NaOCl solution (0.5 mL L⁻¹, Tween 20 added) for 5 min in a laminarflow cabinet in order to surface sterilization. For removing sterilant, they were rinsed with sterile distilled water 3 times. Then, they were put in sterile filter papers for 4–5 h to dry (Altınok & Can, 2010). Explants were cultured to Potato Dextrose Agar (PDA) and Fusarium Minimal Media (FMM) containing antibiotics (Stretomycin Sulphate, 100 µg mL L⁻¹) and incubated 24 ± 2 °C for 5–7 days. At the end of incubation, the fungal colonies developing around the plant tissues were purified and single spore isolation was done.

For the purpose of F. proliferatum characterization, single spored isolates were morphologically investigated according to the species identification key of Leslie & Summerell (2006). The characteristics of the colonies were analyzed after the isolates were sub-cultured onto Potato Sucrose Agar (PSA) to determine colony morphology and onto Synthetic Nutrient Agar (SNA) to investigate conidial structures according to Altınok & Can (2010).

In the molecular identification of the fungal isolates, translation elongation factor 1-α (Tef-1α), and RNA polymerase second large subunit (RPB2) regions were used (Dedecan et al., 2022).

After isolations and molecular diagnosis, garlic bulbs without any fungal development were considered as healthy and constituted the healthy garlic group in subsequent biochemical analyses, while the cloves from which F. proliferatum isolates were obtained formed the diseased group.

Total phenolic content extraction and measurement

The extraction of the phenolic compounds was done by homogenizing 10 g garlic cloves in 80 mL of 80% methanol solution using a high-speed homogenizator (IKA™, T25).The homogenate was stored at room temperature in the dark for 24 h and later centrifuged for 10 min at 4,000 g. The resulting supernatant was filtered to discard impurities, and total phenolic content was quantified using the Folin-Ciocalteu method. Absorbance of the resulting solution was measured at 765 nm. The total phenolic content was expressed as mg GAE g⁻¹ fresh weight, using a calibration curve prepared with standard gallic acid solutions (Erol, 2024a).

Extraction and 2,2-Diphenyl-1-picrylhydrazyl (DPPH) free radical scavenging method

For antioxidant activity, garlic cloves were extracted with 85% ethanol in an amount equal to two g. Homogenization was done for samples collected and kept away from light at room temperature for a period of 24 h. Then, centrifugation was done for 10 min with 5,000 g, after which solution was obtained for extraction. The 2,2-Diphenyl-1-picrylhydrazy (DPPH) radical scavenging activity was carried out using the procedure previously described with some modifications. To this end, 0.1 mM DPPH radical was reconstituted in methanol and 100 µL of the extract from each sample was added to three mL of the DPPH solution. The resulting mixture was thereafter kept in the dark for 30 min. The absorbance was consequently measured at an absorbance wavelength of 517 nm using a spectrophotometer while ascorbic acid was used as the positive control. The DPPH radical scavenging capacity was calculated using % inhibition formula according to Ozdenefe et al. (2024). (1)${\displaystyle \text{\%}\text{Inhibition}=\left(\text{Absorbance of control}-\text{Absorbance of sample}\right)/\left(\text{Absorbance of control}\right)\times 100.}$

Assays were done in triplicate, and data are expressed as mean ± standard deviation (SD). Data were analyzed by analysis of variance (ANOVA), and a probability of less than 0.05 was considered significant.

Protein analysis

The protein content in garlic cloves was determined by the Kjeldahl method. A 0.25 g sample was digested with sulfuric acid (H₂SO₄) and a catalyst, with combustion for 3.5 h. The digested solution was then distilled in an alkaline environment created by 40% sodium hydroxide, NaOH, and the ammonia collected with 4% boric acid. The distillate obtained was then titrated against 0.1 N hydrochloric acid (HCl) for protein content. The calculations were performed according to the guidelines laid out by Casado et al. (2004).

Mineral content analysis

The samples of garlic cloves were dried at 72 °C and ground into powder form for the analysis of their mineral content. A 0.25 g portion of each sample was then subjected to digestion using a mixture of nine mL nitric acid (HNO₃) and three mL hydrogen peroxide (H₂O₂) in a microwave digestion system operating with 200 W for 30 min. The digested samples were filtered and diluted to a final volume of 25 mL. The concentration of Ca, Mg, Fe, Mn, Zn, and Cu was done by atomic absorption spectroscopy, while for Na and K, flame photometry was done (Besirli et al., 2022). The level of N and P was determined in a UV spectrophotometer. All the measurements were carried out in three times for reliability as technical replicates (triplicates).

Extraction method and high-performance liquid chromatography (HPLC) analysis conditions of phenolic compounds

The phenolic compounds were extracted from garlic cloves using a method optimized for maximum recovery. About 10 g of each garlic sample was homogenized in 80 mL of 80% methanol and incubated at room temperature in the dark for 24 h. On incubation, the mixture was centrifuged at 4,000 g for 10 min to separate the supernatant. Filtration was subsequently carried out on the supernatant to remove impurities and improve the solubility of the phenolic compounds (Erol, 2024b).

The separation and quantification of phenolic compounds were performed on a reverse phase high-performance liquid chromatography (RP-HPLC) system fitted with a C18 column (4.6 × 250 mm, five µm particle size). The column temperature was kept at 30 °C, and a diode array detector (DAD) was set to monitor the phenolic peaks. A gradient elution system was performed when using two mobile phases: A, 0.1% phosphoric acid in water, and B, 100% acetonitrile. The volume of injection was 10 µL, and the flow rate was one mL/min. Identification and quantification of phenolic compounds were performed in accordance with comparison of retention times with those of external standards. Firstly, standard solutions with known concentrations of phenolic compounds were plotted in calibration curves. Results were given as milligrams of phenolic content per kilogram of garlic weight. All analyses were run in triplicate in order to assure precision and repeatability (Erol, 2024b).

Extraction method and HPLC analysis conditions of organic acids and sugars

Organic acids and sugars were extracted from garlic cloves. Approximately 2.5 g of finely crushed garlic samples were placed into 50 mL Falcon tubes, to which 25 mL of a deionized water/methanol mixture (7:3, v/v) was added. The mixture was thoroughly homogenized using a high-speed homogenizator (IKA™,T25). The homogenate was then incubated in a water bath at 40 °C for 30 min to aid extraction. The solution was then centrifuged for 10 min at a speed of 10,000 g at 4 °C and the supernatant filtered via a 0.45-micron syringe filter. The filtered extracts were kept under −20 °C until further analysis (Erol, 2024b).

Organic acids and sugars in the garlic extracts were determined by a high-performance liquid chromatography system. Separation of sugars was done on a Rezex RCM-Monosaccharide Ca²⁺ (8%) LC Column (300 × 7.8 mm). Column temperature was maintained at 80 °C, with analysis done in an isocratic mode by using ultrapure water as a mobile phase at a flow rate of 0.5 mL/min. The analysis was completed within 15 min, and the glucose, fructose, and sucrose concentrations were calculated from calibration curves prepared using standard sugar solutions. Results were presented as milligrams of sugar per gram of fresh weight, and three replicates of each sample were analyzed (Erol, 2024b).

Organically bound acid was determined by the same high-performance liquid chromatography (HPLC) system, only with a different column: Rezex ROA-Organic Acid H⁺ (8%) LC Column (300 × 7.8 mm). The column temperature was 50 °C, and the UV detector was set at 210 nm. A gradient elution system was used for separating a number of organic acids. The analytical results for organic acids were determined using calibration curves built from standard solutions of the target organic acids. Milligrams of organic acid were expressed per gram weight (Erol, 2024b).

Statistical analysis

All biochemical, elemental, and HPLC analyses were conducted using three biological replicates for each genotypes (for both healthy and diseased samples). Each replicate was prepared by pooling 20 cloves, ensuring representative sampling within the group. The variables that were considered in this correlation analysis including phenolic compounds, organic acids, sugars, antioxidant activity, elements, total phenolic content, and protein levels were calculated using a Pearson’s correlation coefficient. The data set on the analysis of elements was used to construct a correlation matrix by summarizing and visualizing it through a heatmap showcasing relationships as positive or negative. Then, principal component analysis (PCA) was applied to gauge the variation among phenolic compounds, reducing data dimensions into two main components. The distribution of genotypes and their effect regarding F. proliferatum infection on the selected phenolic compounds were shown in biplot chart. Correspondingly, changes in the organic acid concentration within different genotypes were compared by applying line charts for healthy and diseased genotypes. Mean comparisons were done by analysis of variance (ANOVA), followed by Tukey’s honestly significant difference (HSD) test to establish the significant difference among the samples. Any results with a p ≤ 0.05 were considered significant. Furthermore, PCA and other data visualization techniques were carried out using JMP 14 and OriginPro.

Isolation of Fusarium proliferatum from garlic bulbs

Upon examination of the garlic bulbs collected from storage houses, browning was observed at the basal parts of the cloves, along with rot progressing from the basal area towards the tip. A total of 159 garlic bulbs were studied and 2.52% of the garlic bulbs were classified as infected with F. proliferatum where three of the fungal isolates were deposited in the National Center for Biotechnology Information (NCBI) and accession numbers were assigned (Table 1). The morphological characteristics of Fusarium proliferatum isolates grown on PSA and SNA media were represented in Fig. 1. For the biochemical analysis, healthy and F. proliferatum infected garlic cloves were used.

Total phenolic contents

Total phenolic content was significantly higher in the infected than in the healthy plants (Fig. 2A, Table 2). Total phenolic content in the Local-Konya genotype of healthy plants was 5,400.96 mg kg⁻¹ and increased by 36.0% to 7,347.44 mg kg⁻¹ in infected plants (p ≤ 0.01). While the phenolic content was 5,264.72 mg kg⁻¹ in healthy plants for the Babaeski-Kırklareli genotype, the infection increased it by 66.9% and reached the value 8,790.07 mg kg⁻¹ (p ≤ 0.001). For the genotype Iranian-Balıkesir, the content of total phenolic was found to be 3,634.33 mg kg⁻¹ in healthy plants and increased by 110.9% to 7,664.61 mg kg⁻¹ in infected plants (p ≤ 0.001). According to the statistical analysis, it was found that the total phenolic compounds increased significantly in all the studied genotypes after F. proliferatum infection. It is obvious that the 110.9% increase in the Iranian-Balıkesir genotype might show a meaningful increase in phenolic compound production due to resistance against this disease. Overall, an increase in the total phenolic content was statistically significant (at least at p ≤ 0.05) in all studied genotypes.

Antioxidant activity

F. proliferatum infection enhanced the antioxidant activity of all genotypes (Fig. 2B, Table 2). While the DPPH radical inhibition was 39.07% in healthy plants of the Local-Konya genotype, it increased to 51.25% in the infected plants, reflecting a 31.1% increase (p ≤ 0.05). In the genotype Babaeski-Kırklareli, antioxidant activity increased from 33.63% in healthy plants to 41.57% after the infection, marking a 23.6% rise (p ≤ 0.05). In the Iranian-Balıkesir genotype, antioxidant activity increased from 29.94% in healthy plants to 46.57% in infected plants, which is a 55.5% increase (p ≤ 0.01). Statistical analysis carried out with respect to F. proliferatum infection showed that antioxidant activities increased significantly. It can be said that with the increase of 55.5% observed in the Iranian-Balıkesir genotype, this genotype developed a good antioxidant defense against oxidative stress caused by F. proliferatum infection. Increases in antioxidant activities in all genotypes were significant statistically at least at p ≤ 0.05.

Total protein content

F. proliferatum infection effects on the protein content of garlic plants showed lower protein contents of all genotypes in comparison to that of the healthy ones (Fig. 2C, Table 2). The protein content of healthy plants was 4.23% for the Local-Konya genotype and declined by 2.6% to 4.12% in the infected ones, significantly lower at p ≤ 0.05. In the healthy plants of the Babaeski-Kırklareli genotype, protein contents were determined as 4.93%, but it was observed to decrease by 7.5% to 4.56% due to F. proliferatum infection (p < 0.01). In the Iranian-Balıkesir genotype, it was 4.18% in the healthy plants, while it also decreased by 15.5% in the infected ones as in the first group to 3.53% (p < 0.001).

Whereas the reduction in protein content was observed for all genotypes, the Iranian-Balıkesir genotype showed a greater reduction. One-way ANOVA showed that these reductions were at a statistically significant level in all genotypes (p ≤ 0.05). Among the genotypes, the highest reduction was obtained in the Iranian-Balıkesir genotype. This result definitely shows that F. proliferatum infection may interfere with the synthesis and metabolism of proteins by the plants, hence causing such a sharp loss of the protein content.

Element analysis

F. proliferatum infection caused significant variations in both macro- and microelement levels among all the garlic genotypes tested (Table 3). The nitrogen (N) level was stable for both the healthy and infected Local-Konya plants at 0.81% (p > 0.05); regarding phosphorus, it increased by 43.6% to 430.62 mg/100 g (p ≤ 0.01). K decreased slightly by 1.4% for 5,607.23 mg kg⁻¹, whereas for Ca, there was a 13.3% decrease, which was 1,374.31 mg kg⁻¹ (p ≤ 0.05). In the Babaeski-Kırklareli genotype, N decreased from 0.96% to 0.91% (p ≤ 0.05), while P increased by 10.4% to 477.32 mg/100 g (p ≤ 0.01). K dropped by 7.5% to 6,472.62 mg kg⁻¹ (p ≤ 0.05), and Ca decreased by 16.4% to 1,712.5 mg kg⁻¹ (p ≤ 0.01). In the Iranian-Balıkesir genotype, N dropped to 0.67% (p ≤ 0.05), and K levels fell by 13.9% to 4,798.5 mg kg⁻¹ (p ≤ 0.01). P levels showed a slight increase to 498.35 mg/100 g (p > 0.05), while Ca levels decreased by 24.3% to 1,337.88 mg kg⁻¹ (p ≤ 0.01). These results highlight that F. proliferatum infection caused significant differences in macro element levels between genotypes, particularly in potassium and calcium levels, which showed notable reductions.

Regarding microelements, copper (Cu) levels varied significantly between genotypes (Table 3). In the Local-Konya genotype, Cu levels increased by 145.2% from 1.79 mg kg⁻¹ in healthy plants to 4.39 mg kg⁻¹ in infected plants (p ≤ 0.01). In the Babaeski-Kırklareli genotype, Cu levels showed a slight decrease to 2.28 mg kg⁻¹, which was not statistically significant (p > 0.05). In the Iranian-Balıkesir genotype, Cu levels dropped sharply from 2.24 mg kg⁻¹ to 0.20 mg kg⁻¹ (p ≤ 0.001). Iron (Fe) levels showed a slight increase to 16.7 mg kg⁻¹ in the Local-Konya genotype (p > 0.05), while they decreased by 52.3% to 13.41 mg kg⁻¹ in the Iranian-Balıkesir genotype (p ≤ 0.01). Magnesium (Mg) levels decreased across all genotypes, with the largest reduction observed in the Babaeski-Kırklareli genotype, where Mg levels fell by 34.2% to 279.71 mg kg⁻¹ (p ≤ 0.01). In the Iranian-Balıkesir genotype, Mg levels decreased by 25.6% to 204.97 mg kg⁻¹ (p ≤ 0.05). These findings indicate that F. proliferatum infection significantly affected microelement levels, particularly in copper and iron, with notable differences observed between genotypes.

HPLC analysis of phenolic compounds

The HPLC analysis of phenolic compounds revealed significant effects of F. proliferatum infection on phenolic content across the Local-Konya, Babaeski-Kırklareli, and the Iranian-Balıkesir genotypes (Table 4). In the Local-Konya genotype, chlorogenic acid levels in healthy plants were 210.99 mg kg⁻¹, but after F. proliferatum infection, they decreased by 29.0% to 149.78 mg kg⁻¹ (p ≤ 0.05). In the Babaeski-Kırklareli genotype, chlorogenic acid levels dropped by 65.7% from 220.46 mg kg⁻¹ in healthy plants to 75.48 mg kg⁻¹ in infected plants (p ≤ 0.01). Conversely, in the Iranian -Balıkesir genotype, chlorogenic acid increased by 55.7% from 88.35 mg kg⁻¹in healthy plants to 137.60 mg kg⁻¹ in infected plants (p ≤ 0.05). Catechin, which was undetectable in healthy plants, was found to be 276.43 mg kg⁻¹ in infected Babaeski-Kırklareli plants (p ≤ 0.001). Similarly, catechin levels rose to 252.55 mg kg⁻¹ in infected Iranian-Balıkesir plants, where it was also absent in healthy plants (p ≤ 0.001).

Caffeic acid was undetectable in the healthy Local-Konya plants but measured at 3.01 mg kg⁻¹ in infected plants (Table 4). In the Babaeski-Kırklareli genotype, caffeic acid levels increased by 47.3% from 0.91 mg kg⁻¹ in healthy plants to 1.34 mg kg⁻¹ in infected plants (p ≤ 0.05). In the Iran-Balıkesir genotype, caffeic acid rose slightly from 0.36 mg kg⁻¹ in healthy plants to 1.32 mg kg⁻¹ in infected plants (p ≤ 0.05). 4-hydroxybenzoic acid was only detected in infected Iranian-Balıkesir plants at 0.36 mg kg⁻¹, while it was absent in the other genotypes. Vanillin levels increased to 0.23 mg kg⁻¹ in the infected Local-Konya plants (p ≤ 0.05), with no significant difference observed in the other genotypes. Rutin levels remained low in the Iranian -Balıkesir genotype and showed no significant changes after F. proliferatum infection. T-ferulic acid, which was undetectable in the healthy Babaeski-Kırklareli plants, increased to 0.67 mg kg⁻¹ following infection (p ≤ 0.01), while remaining at low levels in the other genotypes. Naringin, found at 2.04 mg kg⁻¹ in the healthy Babaeski-Kırklareli plants, disappeared in infected plants, while only low levels were detected in the Iranian-Balıkesir genotype. O-coumaric acid levels in the Babaeski-Kırklareli genotype decreased by 23.1%, from 0.13 mg kg⁻¹ in healthy plants to 0.10 mg kg⁻¹ in infected plants (p ≤ 0.05), while in the Iranian -Balıkesir genotype, this compound showed a slight increase to 0.03 mg kg⁻¹ following infection (p ≤ 0.05).

Resveratrol levels increased dramatically in the Iranian-Balıkesir genotype, rising by 998.5% from 0.042 mg kg⁻¹ in healthy plants to 0.461 mg kg⁻¹ in infected plants (Table 4). A significant increase in resveratrol levels was also observed in the Local-Konya genotype, from 0.051 mg kg⁻¹in healthy plants to 0.144 mg kg⁻¹ in diseased plants, representing a 182.4% increase (p ≤ 0.01). Quercetin levels in the Babaeski-Kırklareli genotype decreased by 14.8%, from 1.89 mg kg⁻¹in healthy plants to 1.61 mg kg⁻¹in infected plants (p ≤ 0.05), with no significant changes in quercetin levels observed in the other genotypes. These results indicate differential responses in the biosynthesis of phenolic compounds by garlic plants according to their genotype after F. proliferatum infection. While in some phenolic compounds, such as resveratrol and catechin, significant enhancements were observed, in others, such as quercetin and O-coumaric acid, there was a reduction. Most of these changes were statistically significant (p ≤ 0.05) by statistical analysis. A strong defense response can be seen in compounds such as resveratrol against F. proliferatum infection.

HPLC analysis of organic acids and sugars

Organic acid profiles determined by HPLC analysis showed that F. proliferatum infection significantly affects garlic genotypes (Table 5, Fig. 3). The oxalic acid level increased from non-detected in healthy Local-Konya plants to 15.17 mg kg⁻¹ in infected plants. For the Iranian-Balıkesir genotype, too, the oxalic acid level increased from 7.01 mg kg⁻¹ in healthy plants to 10.56 mg kg⁻¹ in infected ones (p ≤ 0.05). Citric acid levels in the Babaeski-Kırklareli genotype increased from 4,175.95 mg kg⁻¹ in healthy plants to 6,291.02 mg kg⁻¹ following infection (p ≤ 0.01). In the Local-Konya genotype, citric acid levels rose from 4,322.72 mg kg⁻¹ in healthy plants to 4,781.87 mg kg⁻¹ in infected plants (p ≤ 0.05). Succinic acid showed significant increases in all genotypes, especially in the Local-Konya plants, where it rose from 9,281.79 mg kg⁻¹ in healthy plants to 11,579.81 mg kg⁻¹ in infected plants (p ≤ 0.01).

Among other organic acids, malic acid was not detected in healthy Babaeski-Kırklareli plants but measured at 305.87 mg kg⁻¹ in infected plants (p ≤ 0.05). Lactic acid levels also increased significantly; in the Babaeski-Kırklareli genotype, levels rose from 13,639.16 mg kg⁻¹ in healthy plants to 20,840.81 mg kg⁻¹ in infected plants (p ≤ 0.01), while in the Iranian-Balıkesir genotype, lactic acid increased to 2,031.90 mg kg⁻¹ in infected plants.

The sugar analysis revealed that F. proliferatum infection had significant effects on sucrose levels (Table 2, Fig. 2D). In the Local-Konya genotype, sucrose levels increased by 57.2% from 2.40 mg kg⁻¹ in healthy plants to 3.77 mg kg⁻¹ in infected plants (p ≤ 0.01). However, in the Babaeski-Kırklareli genotype, sucrose levels dropped by 59.5%, from 2.84 mg kg⁻¹ in healthy plants to 1.15 mg kg⁻¹ in infected plants (p ≤ 0.05). Similarly, in the Iranian-Balıkesir genotype, sucrose levels decreased by 38.5%, from 2.58 mg kg⁻¹ in healthy plants to 1.59 mg kg⁻¹ in infected plants (p ≤ 0.05). These results suggest that F. proliferatum infection influences the synthesis of organic acids and sugars differently depending on the genotype. Valuably enough, the increased levels detected at succinic acid, lactic acid, and sucrose upon infection might suggest providing a more direct indication of the metabolic response of the plant. Most of these changes have already been statistically proven (p ≤ 0.05) to be significant.

Multivariate analysis

Multivariate analyses were performed in order to obtain an overview of the interactions among the parameters studied, garlic genotypes and the pathogen. The heatmap very well demonstrated great changes between the healthy and diseased plants in a level of both macro- and microelements (Fig. 4), where red color indicates a higher and blue lower level, thus showing great differences among genotypes. In the case of the Local-Konya genotype, a significant decline in Ca and K levels can be observed between healthy and diseased plants. With respect to the Babaeski-Kırklareli genotype, high levels of elements in the healthy plants dropped to low values when under F. proliferatum infestation. In diseased plants, a serious decrease of copper, along with other microelements, was observed in the Iranian-Balıkesir genotype. The Ni levels were significantly reduced at p ≤ 0.001, hence showing a great deal of difference between healthy and diseased plants. Overall, this heatmap clearly presented the consequence of F. proliferatum infection on plant nutrient levels and the importance of such changes.

The result of PCA biplot analysis (Fig. 5) illustrates differentiation among genotypes because of F. proliferatum infection, as explained by two major principal components. Notably, in both healthy and diseased plants, the Local-Konya genotype clustered closely, indicating minimal changes in phenolic compound levels. On the other hand, the Babaeski-Kırklareli genotype showed clear differentiation between healthy and diseased plants. Among them, the variables chlorogenic acid, caffeic acid and resveratrol contributed a lot to genotype separation, whereas catechin and quercetin showed the highest variance especially under disease. The first principal component explained 58.9% of total variance, while the second principal component accounted for 24.5%, which indicates an increased production of these compounds upon F. proliferatum infection. Overall, this biplot effectively showed the variation in the effect of F. proliferatum infection on phenolic compounds regarding the variability among the genotypes.

Analysis provides some relationships of all the measured parameters, across three genotypes (Local-Konya, Babaeski-Kırklareli, and Iran-Balıkesir) (Fig. 6). In most cases, it was a positive correlation for the phenolic compounds and antioxidant activity. For instance, strong correlations were observed for compounds such as resveratrol and DPPH % inhibition. A weaker correlation was observed not only between the total phenolic content and protein level, but also within organic acids of the genotypes studied. In terms of elements, potassium (K) had a good positive relationship with some macro elements, especially in magnesium (Mg) and calcium (Ca).