Introduction

Grapes are rich in phenolic compounds, with their composition influenced by factors such as variety, environmental conditions, and ripeness (de-la-Rosa et al. 2019). Grape berries contain both primary and secondary metabolites (Bahar et al. 2018). Primary metabolites, essential for plant growth, include carbohydrates, proteins, lipids, minerals, vitamins, and amino acids (Crozier et al. 2008; Ramawat 2007). Secondary metabolites (SM), such as phenolic compounds, protect against abiotic and biotic stresses (Yadav et al. 2021). These compounds also benefit human health, with their biosynthesis triggered by stimuli such as injury (Billet et al. 2018). Key phenolic compounds include tannins, anthocyanins, organic acids, and flavonoids, which protect plants from diseases and pests (Shitan 2016). Ivanova et al. (2011) stated that anthocyanins, responsible for red grape color, accumulate from veraison, while proanthocyanidins accumulate before veraison. Kemp et al. (2009) highlighted the importance of tannins, which polymerize between veraison and harvest. Pagare et al. (2015) classified tannins as phenolic polymers with defensive properties. Plants produce diverse SM, including chemically heterogeneous phenolic compounds with phenol group, an aromatic ring with hydroxyl group.

SM support plant defense against abiotic and biotic stresses (Silva et al. 2023). Phenolic compounds are categorized into phenolic acids, stilbenes, flavonoids, tannins, and lignans (Tsimogiannis and Oreopoulou 2019). Mazid et al. (2011) grouped them as phenolics, terpenes, and nitrogen- and sulfur-containing compounds. Birt and Jeffery (2013) identified over 4000 flavonoids, which Vermerris and Nicholson (2007) noted as a significant portion of phenolics. Resveratrol, a phytoalexin known as 3,4′,5-trihydroxy-trans-stilbene, has two isomers, with the -trans form predominant in plants (Farhan and Rizvi 2023). Found in grape skins and wines (Rienth et al. 2021), resveratrol levels are affected by conditions such as water stress and temperature (Sun et al. 2023). Yaman et al. (2016) determined that resveratrol amounts are indirectly related to rainfall and directly influenced by sun exposure, and temperature.

Abiotic stress factors such as ultrasonication have been found to increase resveratrol levels in grape (Xiao et al. 2005). Ultrasonication for 5 min generates high resveratrol levels. Electrical stimulation via solar panels also increases total soluble solids (TSS), anthocyanin, and resveratrol contents in grapes (Mikami et al. 2017). Mechanical vibration causes weight loss and a decrease in TSS (Jung et al. 2018). Leaf removal before flowering increases TSS and anthocyanin (Poni et al. 2009), while removal 2 weeks before harvest slows sugar accumulation and improves quality (Alço 2019).

Ultraviolet C (UV-C) is shortwave radiation and a significant abiotic stressor (Liang et al. 2023). Applying 5 min of UV‑C to ‘Cabernet-Sauvignon’ shoots, starting 15 days after full bloom (DAF) and continuing every 10 days for 105 days, increased flavan-3-ols (Niu et al. 2021). UV‑C at 260–270 nm is ideal for maximum resveratrol biosynthesis (Gonzálvez et al. 2013). Post-harvest UV‑C treatment enhanced stilbene and resveratrol production in grapes (Maurer et al. 2017). UV‑C also reduced Botrytis cinerea infection by increasing resveratrol synthesis (Bavaresco et al. 1997) and significantly boosted resveratrol content, with reported increases of up to four-fold (Cantos et al. 2003; Gonzalez-Barrio et al. 2009). Water stress negatively affects resveratrol in grapes (Oraee and Tehranifar 2020). While irrigation decreases sugar accumulation and total anthocyanin content (Lopez et al. 2001), water stress increases total anthocyanin, polyphenol concentration, and TSS (Carbonneau and Bahar 2009). Berry composition varies among Vitis genotypes and is influenced by environment and viticultural practices.

The main objective of this study was to determine the effects of abiotic and biotic stress factors applied in the late-stage (5 days before harvest) on the primary (TSS, titratable acidity [TA], etc.) and secondary metabolites (anthocyanin, tannin, total phenolic index [TPI], total phenolic content [TPC], etc.) of grapes and the concentration of resveratrol. The most important reason for late-stage applications is to minimize the damage to trial vines caused by damaging procedures such as UV‑C, shock action, leaf injury, leaf removal, and vibration. Additionally, since all specified applications are repeated for 2 years, the health of the vines is prioritized.

Materials and Methods

The research took place in a Tekirdağ vineyard in Turkey situated 235 m above sea level, and grafted onto SO4 rootstock, with 13-year-old cv. ‘Cabernet Sauvignon’ and cv. ‘Merlot’ (Vitis vinifera L.) grapevines. The region has a Mediterranean climate with an average temperature of 14 °C and 582 mm of annual rainfall. Classified as Zone III by the Winkler Index, it accumulates 1893 degree-days and has a Branas Heliothermic Index long-term value of 6.24. The soil is clay loam with a slightly alkaline pH of 7.38. With a planting density of 427 vines · ha−1, the vines are trained to the bilateral cordon pruning system.

Stress Applications

The vines were treated twice daily, 5 days before harvest, at 08:00 h and 19:00 h, with harvest on 18.09.2016 and 27.09.2017. Applications included:

  • Control (A1 = C): No application.

  • Shock Action (A2): Thick branches and trunk were struck with a plastic hammer for 1 min.

  • UV‑C (A3): Vines were exposed to a 40-watt bulb emitting 254 nm light for 1 min in a covered cabin (Niu et al. 2021).

  • Vibration (A4): Mechanical vibration was applied with a drill for 1 min, using a rubber-isolated wedge against branches and trunk.

  • Leaf Injury (A5): Leaves were struck with a hard stick on both sides once.

  • Leaf Removal (A6): All leaves were removed once.

  • B. cinerea Pers ex. Fr. Inoculation (A7): Grapevine-derived Botrytis sp. isolate was sprayed, and clusters were isolated for 24 h (Leone and Heuvel 1987).

Grape Juice Analyses

TSS was determined as °Brix (20 °C) via a handheld refractometer (ATC-0-50, Istanbul, Turkey). TA in g L−1 (tartaric acid) was determined by titrimetric method as shown in Eq. 1:

$$\textit{Titration}\,\textit{acidity}(g/L)=(V)(f)(E)(1000)/M$$
(1)

pH values were determined potentiometrically via a digital pH meter (Hanna Instruments, HI-2210, Bedfordshire, England) (Cemeroğlu 2007). Sugar concentrations equivalent to the °Brix values of the samples were determined from the table (g L−1) (Blouin and Guimberteau 2000). The sugar content in the berries was determined via Eq. 2 (Carbonneau and Bahar 2009):

$$\textit{Sugar}\,\textit{amount}\,in\,\textit{berry}(mg/\textit{berry})=[1/1.3x\,\textit{Sugar}(g/L)]x[1/100x100\,\textit{berry}\,\textit{weight}(g)]$$
(2)

The milligrams of sugar per berry were determined via Eq. 3:

$$1g\,\textit{berry}\,\textit{sugar}\,\textit{content}(mg/g-\textit{berry})=\textit{Sugar}\,\textit{content}\,in\,\textit{berries}/\textit{berry}\,\textit{fresh}\,\textit{weight}$$
(3)

The maturity index were calculated in two ways. °Brix/TA (g L−1) according to Blouin and Guimberteau (2000), and the square of the pH measurements multiplied by the TSS were obtained; 260 pH2 × °Brix is ideal (Blouin and Guimberteau 2000).

Secondary Metabolites

All the samples were deseeded, crushed, and centrifuged. The total anthocyanin content (mg kg−1) was determined spectrophotometrically via Eq. 4 using the pH-differential method (Cemeroğlu 2007). Samples were weighed, placed into polypropylene tubes, and mixed with 0.1% HCl. After being left in the dark overnight, they were centrifuged at 4500 rpm for 10 min. The clear supernatant was stored at −18 °C until analysis. KCl (pH 1.0) and NaOAc (pH 4.5) buffer solutions were prepared, and two dilutions of the methanolic extract were made. Absorbances were measured at 520 nm and 700 nm after ~ 30 min of equilibration.

$$\textit{Total}\,\textit{anthocyanin}\,\textit{content}(mg/L)=(A)(MW)(Sf)(1000)/(\varepsilon)\lambda$$
(4)

A: Corrected absorbance difference, MW: molecular weight of the total anthocyanin to be used as a reference = Malvidin-3-glucoside MW = 493.5, Sf: dilution factor ε: molar absorptivity coefficient, ε = 28.000 for malvidin-3-glucoside Ɩ: cuvette path length = 1 cm.

Total tannin content (mg kg−1): About 40 µL of the diluted sample was placed in a spectrophotometer cuvette with 3.36 mL distilled water and 200 µL Folin–Denis reagent. After adding 400 µL saturated Na2CO3 solution, absorbance was measured at 760 nm after 30 min via spectrophotometer (UV-Mini 1240, Shimadzu, Kyoto, Japan) against distilled water blank.

In the vineyard, 200 berries were separated from each parcel and each was analysed individually. For TPI analysis, 1 mL of must sample was diluted with 50 mL of distilled water and then centrifuged at 8000 rpm for 10 min. The readings were taken at 280 nm via a spectrophotometer. The dilution factor was calculated by multiplying the absorbance value (INRA 2007).

For TPC determination (mg kg−1), 40 µL of extract was placed in a spectrophotometer cuvette with 3.16 mL distilled water and 200 µL Folin–Ciocalteu reagent (Merck, Germany), and left to stand for 1–2 min. Then, 600 µL of saturated NaCl solution (200 g L−1) was added, stirred, and left at room temperature for 2 h. Absorbance was measured at 765 nm using a spectrophotometer against a distilled water blank. TPC, expressed as gallic acid equivalent (GAE), was calculated using a gallic acid standard curve.

Two grape bunches from each vine were tested via the Shimadzu Prominence LC 20A HPLC system, with a fluorescence detector (RF-20A), used a gradient system with 300 nm excitation and 386 nm emission wavelengths. An Inertsil ODS‑3 guard column (5 μm, 10 × 4.0 mm) and an Inertsil ODS‑3 column (5 μm, 250 × 4.6 mm) were used. Mobile phases were CH3CN with 0.2% HCOOH (B) and ultrapure water with 0.2% HCOOH (A). Samples were filtered through a 0.45-μm membrane before high-performance liquid chromatography (HPLC) analysis. The flow rate was 1.5 mL min−1, the column oven was set to 30 °C, and the analysis time was 22 min. Samples were then filtered through a 0.45-μm PTFE syringe filter, placed in amber vials, and 5 μL was injected using an automatic injector. Resveratrol standard was from Sigma-Aldrich. Calibration was done with a resveratrol stock solution stored at −18 °C. A calibration curve (R2 = 0.99) was created, and sample concentrations were calculated using LC Solutions (Shimadzu, Japan) software.

Statistical Analysis

The findings obtained from abiotic and biotic stress applications were evaluated via variance analysis, for which the MSTAT‑C statistical package program was utilized. Differences between the results from 2 years were determined via the Least Significant Difference (LSD) test.

Results and Discussion

Primary Metabolites

The TSS values were not affected by the application, cultivar, or year. However, compared with the control, late-stage stress applications slightly increased the TSS (23.67°Brix). In both years, TSS values were found to be close to each other (‘Cabernet-Sauvignon’ 24.37°Brix and ‘Merlot’ 24.02°Brix). The TSS value, which is among the harvest criteria for wine grape varieties, is within the range of 20–25°Brix, similar to the research values (Blouin and Guimberteau 2000). The findings reported by Öner (2014), indicating TSS values ranging from 22.69–23.51°Brix in ‘Cabernet-Sauvignon’, are consistent with the results. In their analysis conducted 110 DAF in the ‘Cabernet-Sauvignon’, they measured the TSS value of the control as 20.64°Brix (Sun et al. 2023). This value was found to be considerably lower than the research results, thought to be due to different ecological conditions and soil structure.

Table 1 shows significant differences at the LSD0.1 level in 2‑year average TA values between varieties and at the LSD1 level among Cultivar × Application interactions. In the cv. ‘Cabernet-Sauvignon’, Öner (2014) reported that the TA values ranged from 5 to 5.09 g L−1, while Sun et al. (2023) indicated a TA value of 7.36 g L−1 for the control group in the analysis conducted 110 DAF. Conversely, Candar (2019) reported that the TA value in the cv. ‘Merlot’ ranged from 5.65 to 7.80 g L−1. These values are within a similar range as the research findings. Similarly, the TA value at maturity, which should be between 3.2 and 3.5 g L−1 in terms of TA, as emphasized by Blouin and Guimberteau (2000), is consistent with the research results.

Table 1 Effects of cultivar, application, year, and their interaction on titratable acidity (TA)
Full size table

The main effects and interactions of different stresses applied to ‘Cabernet-Sauvignon’ and ‘Merlot’ did not significantly affect the pH. The pH values of both cultivars were determined to be the same (3.26). Öner’s (2014) finding that pH in the ‘Cabernet-Sauvignon’ ranged from 3.30 to 3.33 aligns with the reported values.

The sugar concentration value in 2016 was 240.43 g L−1, while in 2017 it was 240.30 g L−1. There was no statistical difference between the Year Main Effect (YME), Application Main Effect (AME) and the Cultivar Main Effect (CME) (‘Cabernet-Sauvignon’ 242.36 g L−1 and ‘Merlot’ 238.37 g L−1). However, the A3 yielded a numerical value of 247.74 g L−1. The sugar concentration at maturity, ranging from 190–250 g L−1, as indicated by Blouin and Guimberteau (2000), is consistent with these findings.

The effect of applications on the sugar content in berries was found to be significant. The A3 (100.98 mg berry−1) reached the highest sugar content in berries, followed by the A2 (94.37 mg berry−1). The other applications were in the same significance group.

In 2016 and 2017, it was observed that the AME, CME, and YME, as well as their interactions, did not have a significant impact on the milligram sugar content per gram of berry. However, in the ‘Cabernet-Sauvignon’, this value was recorded as 80.83 mg 1 g berry−1, while in the ‘Merlot’, it was 79.87 mg 1 g berry−1. The applications also ranged from 83.55 mg 1 g berry−1 (A3) to 77.36 mg 1 g berry−1 (A1).

Maturity Index

The Year Main Effect was found to be significant for the ripening index represented by the TSS/TA ratio. Values of 3.56 g L−1 in 2016 and 3.24 g L−1 in 2017 were obtained. The findings were consistent with the report by Candar (2019) regarding the variation in TSS/TA values in ‘Merlot’ (ranging from 2.79 to 3.82 g L−1). Additionally, the Cultivar Main Effect was significant. As expected, there was a difference in the ripening index between the ‘Cabernet-Sauvignon’ and ‘Merlot’, which is related to their distinct characteristics. While the TSS/TA value for the ‘Merlot’ was 3.56 g L−1, it was 3.24 g L−1 for the ‘Cabernet-Sauvignon’. These results align with the ideal range of 3–4 g L−1 emphasized by Blouin and Guimberteau (2000).

Regarding pH2 × °Brix values, no significant relationship was found between applications, years, cultivars, and their interactions. However, pH2 × °Brix values were recorded as 259.12 g L−1 for ‘Cabernet-Sauvignon’ and 254.68 g L−1 for ‘Merlot’. Öner (2014) reported these values ranging from 237.77 to 272.10 g L−1 for ‘Cabernet-Sauvignon’, and similarly, Candar (2019) indicated that values for ‘Merlot’ ranged from 251.50 to 257.80 g L−1. These values are consistent with the research findings.

Secondary Metabolites

The effects of stress applications, years, and interactions were not significant, with differences observed only in anthocyanin values among cultivars. This aligns with previous studies, such as Maurer et al. (2017), which found no statistical differences in total anthocyanin content after post-harvest UV‑C application. Higher anthocyanin levels in ‘Cabernet-Sauvignon’ (110.77 mg kg−1) compared to ‘Merlot’ (93.67 mg kg−1) are consistent with findings by Shitan (2016) and Iacopini et al. (2008). The control’s total anthocyanin value of 0.77 mg g−1 in ‘Cabernet-Sauvignon’ 110 DAF also aligns with Sun et al. (2023).

Tannin content in 2016 and 2017 showed significant differences based on stress applications (Table 2). ‘Cabernet-Sauvignon’ (3.76 g kg−1) had higher tannin content than ‘Merlot’ (3.52 g kg−1), with 2017 showing higher tannin levels than 2016. The highest tannin values were recorded for A6 and A5, supporting findings by Kemp et al. (2009) and Wimalasiri et al. (2024) that vines with removed leaves produce higher tannin wines. The 2017 × A6 interaction had the highest tannin value, likely due to the factors mentioned above. Additionally, in 2016 and 2017, vines inoculated with B. cinerea synthesized 1.5 times more tannins compared to the control.

Table 2 Effects of cultivar, application, year, and their interaction on total tannin content
Full size table

The effects of late-stage stress applications on the TPI were found to be significant in terms of AME, CME, and Cultivar × Application interaction. Among the years, especially in 2016, the TPI value stood out (11.28). The highest TPI value among the applications was observed in the A7 (9.46), while the lowest value was detected in the control (Table 3). Candar (2019) reported that different shoot length applications in the cv. ‘Merlot’ resulted in TPI ranging from 8.57 to 13.70. Additionally, the finding by Sun et al. (2023) that the TPI value of control 110 DAF in ‘Cabernet-Sauvignon’ was 6.29 mg g−1 is consistent with the ranges observed in this study.

Table 3 Effects of cultivar, application, year, and their interaction on total phenolic index
Full size table

TPC varied with applications, with A5 (3312.08 mg kg−1) and A6 (3422.08 mg kg−1) showing the highest values (Table 4). Sunlight, as noted by Crippen and Morrison (1986), significantly impacts TPC levels. The increase in TPC with A5 and A6 applications is likely due to more sunlight exposure in ‘Cabernet-Sauvignon’ grapes. TPC was higher in 2017 compared to 2016, and ‘Cabernet-Sauvignon’ had a higher TPC than ‘Merlot.’ Gülcü (2016) observed that TPC can vary depending on the variety, even under the same vineyard conditions. Other studies, such as Šulc et al. (2005) and Ivanova et al. (2011), reported TPC levels in grape skins, with Küskü and Tahmaz (2023) finding TPC in ‘Cabernet-Sauvignon’ wine as 3451 mg L−1 and in ‘Merlot’ as 2874 mg L−1. UV‑C application increased TPC by about 20% (Maurer et al. 2017), consistent with this research. SM such as tannin, resveratrol, TPI, and total anthocyanin levels were lowest in the control application, similar to TPC.

Table 4 Effects of cultivar, application, year, and their interaction on total phenolic content
Full size table

Resveratrol (mg kg−1)

Late-stage stress applications significantly impacted resveratrol levels in terms of CME, AME, YME, Year × Application, and Year × Cultivar interactions (Table 5). Applications near harvest triggered the phytoalexin defense mechanism, boosting stilbene production and resveratrol accumulation. The highest resveratrol value was in the A7 application (6.73 mg kg−1), while the lowest was in the A1 (4.45 mg kg−1).

Table 5 Effects of cultivar, application, year, and their interaction on the resveratrol
Full size table

Hasan and Bae (2017) found that UV application increased resveratrol accumulation over 2000 times compared to the control, consistent with this study’s results (A1 = 4.45 mg kg−1 and A3 = 5.67 mg kg−1). Similar to Langcake and Pryce (1977) and Creasy and Coffee (1988), UV applications boosted phytoalexin synthesis in both years, with A3 grapes producing more resveratrol than the control. However, Yaman et al. (2016) found that ‘Merlot’ had slightly more resveratrol than ‘Cabernet-Sauvignon,’ although they were not significantly different, likely due to varying climate and practices. Additionally, consistent with Blanco-Ulate et al. (2015), who suggested B. cinerea promotes secondary metabolite accumulation, resveratrol increased with A7 in 2016 and with A5 and A4 in 2017. Billet et al. (2018) also noted that UV and biotic stress, particularly mechanical injury, raise resveratrol levels over time. Lamuela-Raventós et al. (2001) reported red grape varieties containing 0.69–14.47 mg L−1 resveratrol, matching this study’s findings. Çaylak et al. (2009) noted that the highest resveratrol levels in ‘Merlot’ occurred during a 2007 drought, aligning with the idea that ‘Merlot’ is frost-sensitive. Resveratrol values were higher in 2016 due to drought and moderate in 2017.

Conclusion

In ‘Cabernet Sauvignon’, the TSS, TA, sugar concentration, sugar content per berry, and sugar content per gram berry were found to be high. The pH level was highest in ‘Merlot’. In vines treated with UV‑C, TSS, sugar concentration, sugar content per berry, and sugar content per gram berry were detected at high levels. This suggests that stress applications created numerical differences in primary metabolites and maturity index, but these differences were not statistically significant.

When SM and the TPI were examined, ‘Cabernet Sauvignon’ presented relatively high levels of total tannins, total anthocyanins, TPC, and TPI, whereas ‘Merlot’ presented relatively high levels of resveratrol. In terms of abiotic and biotic stress applications, total tannin was most affected by leaf injury at 3.70 mg kg−1, total anthocyanins was most affected by UV‑C at 1194.74 mg kg−1, and TPC was highest in the leaf injury at 3405.4 mg kg−1. TPI was most concentrated in B. cinerea inoculation at 12.60. Resveratrol was highest in B. cinerea inoculation at 9.72 mg kg−1, followed by UV‑C at 6.96 mg kg−1. Although total tannin, TPC, and total anthocyanin were primarily influenced by abiotic stress, they were also affected by A7.

Statistical differences were observed in SM and TPI between abiotic and biotic stress applications. When considering variety and applications, ‘Merlot’, especially in terms of resveratrol, was found to be highly sensitive to the applied abiotic and biotic stimuli. The ‘Merlot’ variety was most affected by B. cinerea inoculation and, subsequently, by UV‑C. The most impactful were, in order, B. cinerea inoculation and leaf injury. In 2016, under the conditions of Tekirdağ, it was determined that abiotic and biotic stress applications had an effect on SM, particularly resveratrol, in ‘Cabernet Sauvignon’ and ‘Merlot’ vineyards.

In Tekirdağ conditions, it has been determined that B. cinerea inoculation is suitable for increasing resveratrol in cv. ‘Merlot’. However, leaf injury, UV‑C, and vibration applications, alone or in combination, are also expected to be effective. On the other hand, UV‑C application is recommended to increase the amount of resveratrol in cv. ‘Cabernet-Sauvignon’. Additionally, B. cinerea inoculation, leaf removal, and leaf injury applications were found to be effective.

As a result, the reason for abiotic and biotic stresses to be applied 5 days before harvest is to minimize the damage (UV-C) to the vine. In addition, the higher the secondary metabolites and resveratrol in a very short period of time, such as 5 days before harvest, the more functional the grape harvest will be. In conclusion, to increase resveratrol in grapes, it is beneficial to perform B. cinerea inoculation, leaf injury, UV‑C, and vibration applications 5 days before harvest.