Cold Hardiness Evaluation of 20 Commercial Table Grape (vitis Vinifera L.) Cultivars

ABSTRACT

The objectives of this study were to compare cold hardiness (CH) of 20 field-grown grapevine (Vitis vinifera L.) cultivars at four stages: Nov., Jan., Mar. and Apr., and to investigate the relationship between CH and soluble carbohydrates, proline, phenolic compounds and water content changes during acclimation and deacclimation stages. After exposure to various freeze test temperatures, cane LT50 values were estimated using electrolyte leakage and Tetrazolium stain test. CH of all cultivars studied increased with the overall trend of declining temperatures from Nov. through Jan., after which the canes began to deacclimate from Mar. and gradually lost CH in Apr.; overall LT50 means were -15.06, -23.73, -22.42 and -12.08, respectively. A relatively high variation of CH was shown in Jan., when ranged from -20.76 to -26.43ºC. In this sampling stage, cultivars were classified as hardy (CH = -24.5 to -26.5; ‘Moukhtchaloni’, ‘Volgo-Don’, ‘Druzhba’, ‘Khalili’, ‘Bidaneh Qermez’, ‘Bidaneh Sefid’ and ‘Monaqa’), moderately hardy (CH = -22.5 to -24.5; ‘Fakhri Zoudras’, ‘Asgari’, ‘Peykani’, ‘Siahe Qarabaq’, ‘Qareh Shahani’, ‘Galin Barmaqi’, ‘Mirzaei’, ‘Thompson Seedless’ and ‘Fiesta’) and least hardy (CH = -20.5 to -22.5; ‘Flame Seedless’, ‘Shirazi’, ‘Perlette’ and ‘Ruby Seedless’). The pattern of soluble carbohydrates, proline, and phenolic compounds changes were highly consistent with the LT50 profile and cold hardy cultivars accumulated higher osmoregulants compared to others. Our results are valuable to viticulturist to select suitable cultivars for sustainable production in cold zones and to the grape breeder to advance grape CH programs.

KEYWORDS:

• Acclimation
• electrolyte leakage
• freezing tolerance
• osmoregulants

Introduction

Cold stress usually affects fruit trees distribution and productivity. In cold climate, winter freezing damage and spring chilling injury are two main destructive environmental phenomena in vineyards affecting crop productivity, quality and even survival of vines (Karimi and Ershadi, 2015). In these climatic regions, winter minimal temperatures may plunge down −22ºC and exceptionally −30ºC for several hours, causing serious damages to vineyards. Under these harsh conditions, cold injury is expected in mid-winter, but it also often occurs in late fall or early spring. For example, when the temperature reached −21ºC for three consecutive days in 2007 in west of Iran, a severe winter frost led to a 20% loss in annual grape production, as compared to preceding years. For this reason, cold hardiness (CH) screening of grapevine cultivars would be useful to assess their potential for grapevine breeding programs.

The level of grapes CH varies according to the vines growth stage and vineyard temperature (Keller, 2010). Furthermore, plant genetics base and its interaction with environmental signals play an important role in vines survival and productivity in front to low temperatures condition (Fennell, 2004). The cold hardening and dehardening time and the level of seasonal variation in CH change with species, origin, and cultivar (Levitt, 1980). Plants experience some metabolic change during exposing to cold temperature, including rearrangement of cell membrane composition (Steponkus, 1984), increases in compatible solutes (Ghasemi et al., 2012; John et al., 2010; Karimi et al. 2017; Valle, 2002; Zhang et al., 2012), expression of cold-related genes and antifreeze proteins synthesis (Salzman et al., 1996; Thomashow, 1999), altered tissues water content (Salzman et al. 1996; Valle, 2002) and increasing in dormancy-related hormones concentration (Koussa et al., 1998). Compatible solutes such as soluble sugars and proline serve as cryoprotective compounds (Karimi 2017; Morin et al., 2007), which can induce osmotic adjustment, maintains turgor in dehydrating cells, and allows plants to tolerate dehydrative stresses (Ghasemi et al., 2012; Wang et al., 2003; Zhang et al., 2012). The amount of tissue water content is thought that mediate supercooling degree in woody plants (Salzman et al. 1996; Valle, 2002). Regarding the roles of these osmoregulants in CH levels determination of fruit trees during exposure to cold stress, it is possible to use changes in these compounds as physiological indicators for investigation of CH in different grapes cultivars.

Tissue injury assay after artificial subzero temperatures in the laboratory provides useful means for CH comparison in different grapes genotype and cultivars (Fennell, 2004; Jones et al., 1999). The accuracy of CH estimation can be improved by simultaneously using two or more viability tests and by combining the results of these tests (Sutinen et al., 1992). Among the methods of freezing injury measurement, electrolyte leakage (EL) and tetrazolium stain test (TST) are the most accurate ones (Ghasemi et al., 2012). After freezing temperature exposure, ionic leakage assay has been shown to be a useful indicator of differentiation between cold-damaged and -undamaged tissues and has been suggested for frost tolerance screening of woody plants germplasm (Ghasemi et al., 2012; Jones et al., 1999; Morin et al., 2007; Zhang et al., 2012).

Evaluation of redness degree of cold-experienced tissues after TST is a reliable method for quantifying the amount of damages in both field and laboratory tests of cold-tolerant assessment of grapevine and other woody plants (Ershadi et al., 2016; Ghasemi et al., 2012; Linden 2002). However, apart from visual field observations after winter frosts of survival, little data are available on seasonal changes in CH of different grapevine cultivars, especially with a different origin.

CH screening of grapes cultivars for identifying those with higher freezing tolerance is important for matching the cultivars appropriately with cold climate. Therefore, the objectives of this study were: (1) CH screening of 20 different grapevine cultivars using EL and TST measurement in four stages during acclimation and deacclimation periods. (2) to compare seasonal changes in soluble carbohydrates, proline, phenolic compound and water content of these cultivars during the hardening cycle. (3) to identify cold tolerance markers that could be implemented in grapevine breeding programs for marker-assisted selection.

Materials and Methods

Plant Materials

In this study, 20 different grapevine (Vitis vinifera L.) cultivars from 4-year-old own-rooted vines grown at research vineyard (#3) of Research Institute of Grape and Raisin (RIGR) of Malayer University, Iran (it is situated in the northern hemispheres) were investigated at four stages during autumn (10 November; Nov.), winter (12 January; Jan., 11 March; Mar.) and spring (10 April; Apr.) seasons of 2017–18. Grapevine cultivars included 12 Persian (Vitis vinifera L.; ‘Asgari’, ‘Bidaneh Qermez’, ‘Bidaneh Sefid’, ‘Fakhri Zoudras’, ‘Galin Barmaqi’, ‘Hoseini’, ‘Khalili’, ‘Monaqa’, ‘Peykani’, ‘Qareh Shahani’, ‘Siyah Qarabaq’, ‘Mirzaei’), five American (Vitis vinifera L.; ‘Fiesta’, ‘Flame Seedless’, ‘Perlette’, ‘Ruby Seedless’, ‘Thompson Seedless’) and three Russian (Vitis vinifera L.; Druzhba, ‘Moukhtchaloni’, ‘Volgo-Don’) that currently cultivated commercially in Iran.

Samples were collected from five randomly chosen vines for each cultivar, and at least five representative cuttings (20–30 cm in length from node positions 4–7 of mature canes) per each vine were collected at any sampling date. Samples were packed in polyethylene bags, and shipped on ice in Styrofoam boxes to the laboratory, and held in the laboratory at the same field air temperature (± 2°C) as occurred in the vineyard to prevent acclimation.

Controlled Freezing Tests

Cane segments were closed in polyethylene bags and were randomly assigned to predetermined test temperatures. Five replicates were assigned for each cultivar/test temperature. Freezing was accomplished in a programmable freezing chamber (Rad Electronic, Tehran, Iran) according to a stepwise lowering freezing program starting from the prevailing outdoor temperature. Cane segments were subjected to four different freeze test temperatures (−6, −10, −14 and −18ºC in Nov. and Apr., −16, −20, −24 and −28ºC in Jan. and Mar.). The control samples were kept in polyethylene bags at 4ºC. Freeze test temperatures were determined following preliminary tests on canes collected 2 days before each sampling date. The freezing rate was 2ºC h⁻¹, and samples were held at each test temperature for 75 min, before removing from the freezing chamber (Ershadi et al., 2016).

Electrolyte Leakage Assay

Canes were cut into 1-cm long segments and three pieces were placed into 70 mL plastic test tubes containing 40 ml of deionized water. The tubes were capped with aluminum foil and incubated at room temperature under constant shaking at 120 rpm. After 24 h of incubation, the electrical conductivity (EC₁) of the bathing solutions was measured using a Cond-720 conductimeter (WTW GmbH, Weilheim, Germany). Samples were subsequently autoclaved at 120ºC for 20 min, allowed to cool to room temperature, and electrical conductivity was measured again (EC₂). Electrical conductivity of bud samples remained at 4ºC for 24 h considered as control (EC₀). Relative electrolyte leakage (REL) was calculated using the formula: REL = (EC₁-EC₀)/(EC₂-EC₀) ×100.

Tetrazolium Stain Test

Freezing treatments were similar to those used in electrolyte leakage tests. After freezing treatments, cane samples were removed from freezing chamber and placed on ice for slow thawing, and subsequently, 5 mL from a solution of 1% TTC (2, 3, 5-Triphenyle Tetrazolium Chloride) was added to each tube. Samples were then kept for 24 h in the dark at room temperature and dead tissues monitored by stereomicroscope (Leica MS5; Wetzlar, Germany). Samples that did not turn into red color in phloem were considered as dead. The intensity of staining varied within the various cultivars and samples considered stained across a variety of red hues. Mortality percentage was calculated by dividing the number of dead stem samples by the total number of samples (Steponkus and Lanphear, 1967).

Soluble Carbohydrates

Soluble carbohydrates were determined based on the anthrone method (Yemm and Willis, 1954). Canes were oven-dried at 70°C for 2–3 days, ground in a coffee grinder to pass 40-mesh and stored in airtight containers at room temperature, in the dark, until analysis. Soluble carbohydrates were extracted three times from 1 g of ground tissue with 5 ml of 80% ethanol and centrifuged for 15 min at 3000 g_n. One ml of 0.2% anthrone reagent (2 g anthrone in 1 L of 72% sulfuric acid) was added to 100 µl of the ethanolic extract. The reaction mixture was heated in a boiling water bath for 10 min and then rapidly cooled on ice. Absorbance of the extract was read at 620 nm using a Cary WinUV 100 spectrophotometer (Varian, Australia). The concentration of soluble carbohydrates was finally calculated through a calibration curve and expressed as mg g⁻¹ dry weight (DW).

Proline

Proline concentration was determined as described by Bates et al. (1973). Samples of canes were ground in liquid nitrogen, and 0.5 g of ground tissue was homogenized in 10 ml of 3% (w/v) aqueous sulfosalicylic acid. Homogenate was then filtered through a Whatman No.1 filter paper. Two ml of filtered extract were taken for the analysis to which 2 ml ninhydrin and 2 ml glacial acetic acid were added. The reaction mixture was incubated in a boiling water bath for 1 h and the reaction was finished in an ice bath. Four ml of toluene was added to the mixture and the organic phase was extracted. Absorbance was spectrophotometrically measured at 520 nm, while toluene used as blank. Proline concentration was finally calculated through a calibration curve and expressed as µmol g⁻¹ fresh weight (FW).

Water Content

Canes were collected at four stages; from Nov. to Apr. Intermediate sections of each cane were excised and weighed before and after placing in an oven for 3 days at 70ºC. Water content was expressed as a percent of FW.

Total Phenolic Content

Total phenolics were determined colorimetrically using Folin-Ciocalteu reagent as described by Velioglu et al. (1998) with slight modifications. In brief, a volume of 0.3 mL from each diluted methanolic extract (10%) was mixed with 1 mL Folin-Ciocalteau reagent (10%) and vortexed. After 5 min, 1 mL of 7% sodium carbonate solution was added to the mixture. The final solution was shaken for 90 min at room temperature and then the absorbance was spectrophotometrically measured at 765 nm. The total phenolics were quantified by the calibration curve obtained from measuring the absorbance of a known concentration of gallic acid (GA) standard (20–150 mg/Lit). The concentrations were expressed in terms of gallic acid equivalent (mg g-1 of fresh weight). All samples were analyzed in triplicates.

Statistical Analyses

CH was expressed as LT₅₀ (lethal temperature at which 50% of the total ion leakage occurs; or, in the case of tetrazolium stain test, the lethal temperature at which 50% of the tissues are dead) by plotting conductivity or TST data against temperature, with a sigmoidal response curve and using the inflection point of the sigmoidal response curve to predict the lethal temperature (Fry et al., 1993). To evaluate differences among cultivars, results obtained using different methods of CH assessment (EL and TST assays) were subjected to analysis of variance using GLM procedure of SAS (SAS Institute, Cary, NC) and means separated by Duncan’s multiple range tests at p ≤ 0.05. Correlation analysis between soluble carbohydrates, proline, water content, and LT₅₀ values, estimated by EL and TST measurement, was performed using the CORR procedure of SAS.

Results

Cold Hardiness as Estimated by EL Measurement

Based on EL LT₅₀ values, significant differences in CH were observed among cultivars at each sampling date (Table 1). CH of all cultivars studied increased with the overall trend of declining temperatures from Nov. through Jan., after which the cultivars began to deacclimate from Mar. and gradually lost CH in Apr.; overall LT₅₀ means were −15.06, −23.73, −22.42 and −12.08, respectively (Table 1). Based on EL LT₅₀ values, a significant difference (p ≤ 0.001) was found in the CH of cultivars in Nov. (Table 1). In this sampling stage, the highest CH was related to ‘Moukhtchaloni’ (CH = −17.9ºC), which did not differ significantly with ‘Khalili’ cultivar (CH = −17.4ºC). The lowest CH was found to be related to ‘Flame Seedless’ (CH = −12.14ºC) which did not differ significantly with ‘Perlette’ cultivar (CH = −12.65ºC). The CH of other cultivars ranged from −13.48 to −16.80ºC (Table 1).

Table 1. LT₅₀ (lethal temperature at which 50% of the total ion leakage occurs) values, estimated by electrolyte leakage (EL) measurement in 20 Vitis vinifera cultivars in Nov., Jan., Mar. and Apr.

	Cold hardiness (EL-LT50)
	Autumn	Winter		Spring
Grapevine cultivars	10 Nov.	12 Jan.	11 Mar.	10 Apr.
Asgari	−15.61 cd	−23.89 c	−23.36 b	−13.63 c
Bidaneh Qermez	−16.52 bc	−25.13 b	−23.86 b	−14.23 b
Bidaneh Sefid	−16.15 c	−24.43 c	−23.83 b	−13.57 c
Druzhba	−16.52 bc	−25.43 b	−23.79 b	−13.82 c
Fakhri Zoudras	−15.12 d	−24.20 c	−22.70 bc	−13.57 c
Fiesta	−13.48 e	−22.51 e	−21.30 d	−9.91 de
Flame Seedless	−12.14 f	−21.42 fg	−20.10 e	−8.94 f
Galin Barmaqi	−13.95 e	−23.24 d	−21.42 d	−10.83 d
Khalili	−17.43 ab	−25.76 b	−25.36 a	−14.86 b
Mirzaei	−14.91 d	−23.12 d	−22.11 c	−10.73 d
Monaqa	−15.53 cd	−24.56 c	−24.23 a	−14.76 b
Moukhtchaloni	−17.91 a	−26.43 a	−26.12 a	−17.15 a
Perlette	−12.65 f	−21.76 f	−19.69 f	−9.40 ef
Peykani	−13.82 ef	−24.09 c	−23.06 b	−9.80 de
Qareh Shahani	−14.21 e	−23.50 d	−21.75 d	−10.43 d
Ruby Seedless	−14.70 d	−20.76 g	−18.91 g	−9.70 e
Shirazi	−14.10 e	−21.99 f	−20.46 a	−10.30 de
Siahe Qarabaq	−15.32 d	−23.82 c	−21.39 d	−11.83 d
Thompson Seedless	−14.64 de	−22.89 de	−20.03 e	−10.00 de
Volgo-Don	−16.80 b	−25.69 b	−24.93 a	−14.10 bc
Mean	−15.06	−23.73	−22.42	−12.08

Means followed by similar letters are not statistically different (P ≤ 0.05) as compared by Duncan’s

multiple range test.

Throughout the hardening cycle, from Nov. (early hardening) to Jan. (full hardening), CH of all cultivars considerably increased. However, cultivars did not show similar acclimation rates during this period. A relatively high variation of CH was shown in Jan., when ranged from −20.76 to −26.43ºC. In this sampling stage, cultivars were classified as hardy (CH = −24.5 to −26.5; ‘Moukhtchaloni’, ‘Volgo-Don’, ‘Druzhba’, ‘Khalili’, ‘Bidaneh Qermez’, ‘Bidaneh Sefid’ and ‘Monaqa’), moderately hardy (CH = −22.5 to −24.5; ‘Fakhri Zoudras’, ‘Asgari’, ‘Peykani’, ‘Siahe Qarabaq’, ‘Qareh Shahani’, ‘Galin Barmaqi’, ‘Mirzaei’, ‘Thompson Seedless’ and ‘Fiesta’) and least hardy (CH = −20.5 to −22.5; ‘Flame Seedless’, ‘Shirazi’, ‘Perlette’ and ‘Ruby Seedless’). CH of all cultivars decreased slightly (1.31ºC in average) in Mar. ‘Shirazi’, ‘Flame Seedless’, ‘Thompson Seedless’, ‘Perlette’, and ’Ruby Seedless’, less hardy cultivars in this period, reached EL LT₅₀ at ≥ −20.5ºC, while the remaining cultivars had EL LT₅₀ values lower than −21.3 ºC. By mid-Apr., all cultivars showed a high deacclimation rate, and their CH substantially decreased. Based on the EL LT₅₀ values, the highest CH in this stage was related to ‘Moukhtchaloni, while ‘Flame Seedless’ and ‘Perlette’ showed the least CH; other cultivars showed an intermediate CH. In general, a relatively high variation of CH was shown in Nov. and Apr., compared to the other sampling stages (Table 1).

Cold Hardiness as Estimated by TST Measurement

The CH of grapevine cultivars was also investigated with the TST assay in all sampling stages. Results of TST LT₅₀ are presented in Table 2. Significant difference (p ≤ 0.001) was found among studied cultivars. In Nov., TST LT₅₀ values were ≤ −15.8ºC in ‘Moukhtchaloni’ and ‘Druzhba’ cultivars. ‘Ruby Seedless’, ‘Flame Seedless’, ‘Thompson Seedless’ and ‘Perlette’ showed %50 TST at ≥ −12ºC, whereas the remaining cultivars showed an intermediate CH (Table 2). TST estimated cold hardiness of all cultivars increased considerably in Jan. According to the TST LT₅₀ estimations in this sampling stage, cultivars could be divided into three groups: (1) CH cultivars including ‘Volgo-Don, ‘Moukhtchaloni’ and ‘Khalili’ with LT₅₀ ≤ −23ºC, (2) cold susceptible cultivars consisting of ‘Flame Seedless, ‘Shirazi’, ‘Ruby Seedless’ and Perlette with LT₅₀ ≥ −20ºC and (3) other cultivars with an intermediate hardiness. Cold hardiness of all cultivars reduced considerably in Mar. and Apr. sampling stages and they showed an approximately similar ranking in CH, as seen in Jan.

Table 2. LT₅₀ (lethal temperature at which 50% of phloem tissues did not turn into red color or dead) values, estimated by tetrazolium stain test (TST) in 20 Vitis vinifera cultivars in Nov., Jan., Mar. and Apr.

	Cold hardiness (TST-LT50)
	Autumn	Winter		Spring
Grapevine cultivars	10 Nov.	12 Jan.	11 Mar.	10 Apr.
Asgari	−14.20 bc	−22.23 c	−19.05 e	−10.00 cd
Bidaneh Qermez	−14.60 b	−20.83 e	−22.25 b	−11.30 bc
Bidaneh Sefid	−14.00 bc	−21.83 d	−21.25 c	−11.70 bc
Druzhba	−15.80 a	−22.54 c	−20.25 d	−14.00 a
Fakhri Zoudras	−13.96 c	−22.03 cd	−22.05 b	−13.42 ab
Fiesta	−12.56 de	−20.59 e	−18.60 ef	−8.11de
Flame Seedless	−11.65 e	−19.85 f	−18.34 ef	−7.42 f
Galin Barmaqi	−12.24 e	−21.45 de	−19.34 de	−8.24 e
Khalili	−14.70 b	−23.03 b	−23.55 a	−13.70 ab
Mirzaei	−13.36 cd	−22.23 c	−22.45 b	−12.83 b
Monaqa	−14.00 bc	−21.53 de	−21.25 c	−10.65 c
Moukhtchaloni	−16.00 a	−23.03b	−22.25 b	−13.81 ab
Perlette	−10.32 f	−18.03 g	−17.95 f	−8.80 e
Peykani	−11.70 e	−21.83 d	−19.55 de	−12.22 b
Qareh Shahani	−12.85 d	−20.85 e	−19.83 d	−9.35 d
Ruby Seedless	−12.00 e	−18.51 g	−18.00 f	−9.50 d
Shirazi	−13.50 cd	−19.73 f	−19.25 e	−9.60 d
Siahe Qarabaq	−13.87 c	−21.52 de	−19.85 d	−9.24 d
Thompson Seedless	−11.80 e	−20.23 ef	−17.75 f	−10.00 cd
Volgo-Don	−15.20 b	−24.03 a	−23.75 a	−14.51 a
Mean	−13.42	−21.29	−20.33	−10.92

Means followed by similar letters are not statistically different (P ≤ 0.05) as compared by Duncan’s

multiple range test.

There was an overall agreement between results obtained from two methods (Table 7). TST LT₅₀ values were well correlated with EL LT₅₀ values in Nov. (r = 0.89, p ≤ 0.001, n = 60), Jan. (r = 0.86, p ≤ 0.001, n = 60) and Mar (r = 0.83, p ≤ 0.001, n = 86) and Apr. (r = 0.70, p ≤ 0.001, n = 60; Table 7); however, CH estimated by EL measurement was higher than those estimated through TST in both stages (Tables 1 and 2).

Table 3. Changes in soluble carbohydrates concentration in cane of 20 V. vinifera cultivars.

	Soluble carbohydrates (mg g^−1 DW)
	Autumn	Winter		Spring
Grapevine cultivars	10 Nov.	12 Jan.	11 Mar.	10 Apr.
Asgari	48.93 e	95.31 e	65.94 e	35.24 b
Bidaneh Qermez	48.97 e	110.5 b	75.67 b	30.24 c
Bidaneh Sefid	51.93 d	94.53 e	56.81 f	17.81 g
Druzhba	59.21 b	120.3 a	79.26 a	41.23 a
Fakhri Zoudras	56.44 c	107.7 bc	65.50 e	28.45 cd
Fiesta	43.65 fg	75.24 h	49.45 g	20.33 f
Flame Seedless	41.36 g	83.12 g	52.37 g	21.55 f
Galin Barmaqi	49.36 e	99.32 d	63.50 ef	25.90 e
Khalili	60.20 b	111.3 b	73.84 bc	36.42 b
Mirzaei	52.34 d	102.8 cd	58.31 f	21.35 f
Monaqa	47.36 e	105.4 c	75.11 b	27.56 d
Moukhtchaloni	64.75 a	123.5 a	81.44 a	38.91 a
Perlette	33.45 i	72.31 h	45.78 h	16.26 gh
Peykani	53.15 d	106.5 bc	71.82 c	28.77 cd
Qareh Shahani	51.42 d	102.9 cd	69.43 d	27.42 d
Ruby Seedless	35.41 hi	90.39 f	50.88 g	18.43 g
Shirazi	45.61 f	73.14 h	49.50 g	17.42 g
Siahe Qarabaq	55.23 c	112.6 b	77.85 ab	36.48 b
Thompson Seedless	37.40 h	106.2 bc	43.81 h	14.35 h
Volgo-Don	63.71 a	119.4 a	76.92 b	38.70 a
Mean	49.99	100.6	64.16	27.14

Means followed by similar letters are not statistically different (P ≤ 0.05) as compared by Duncan’s

multiple range test.

Table 4. Changes in proline concentration in cane of 20 V. vinifera cultivars during the hardening and dehardening period.

	Proline (µmol g^−1 FW)
	Autumn	Winter		Spring
Grapevine cultivars	10 Nov.	12 Jan.	11 Mar.	10 Apr.
Asgari	1.34 d	5.24 e	4.31 f	2.85 d
Bidaneh Qermez	1.76 bc	7.41 a	6.74 ab	3.69 b
Bidaneh Sefid	1.32 d	6.31 bc	5.81 c	3.39 bc
Druzhba	2.10 a	7.22 a	6.52 b	3.91ab
Fakhri Zoudras	1.81 b	6.96 b	5.75 c	3.64 b
Flame Seedless	1.25 d	5.36 e	4.47 ef	2.83 d
Fiesta	1.26 d	5.75 d	4.92 de	3.10 c
Galin Barmaqi	1.27 d	6.59 b	5.11 d	3.43 bc
Khalili	1.65 c	6.73 b	5.87 c	3.33 bc
Mirzaei	1.26 d	6.22 c	5.13 d	3.41 bc
Monaqa	1.63 c	6.19 c	5.49 cd	3.12 c
Moukhtchaloni	2.31 a	7.63 a	6.73 ab	3.87 ab
Perlette	1.28 d	5.33 e	4.95 de	2.12 e
Peykani	1.55 cd	5.97 d	5.21 cd	3.63 b
Qareh Shahani	1.90 ab	6.96 b	5.75 c	3.45 bc
Ruby Seedless	1.33 d	5.51 de	4.75 e	2.65 d
Shirazi	1.52 cd	5.78	4.63 e	3.19 c
Siahe Qarabaq	1.69 c	7.02 b	6.69 ab	4.11 a
Thompson Seedless	1.38 d	6.35 bc	5.27 cd	2.84 d
Volgo-Don	1.92 ab	7.73 a	7.01 a	4.21 a
Mean	1.58	6.41	5.56	3.34

Means followed by similar letters are not statistically different (P ≤ 0.05) as compared by Duncan’s

multiple range test.

Table 5. Changes in cane water content of 20 V. vinifera cultivars during the hardening and dehardening period.

	Water content (%)
	Autumn	Winter		Spring
Grapevine cultivars	10 Nov.	12 Jan.	11 Mar.	10 Apr.
Asgari	53.29 c	40.19 d	54.30 b	60.81 d
Bidaneh Qermez	49.15 d	33.25 fg	46.40 d	57.72 e
Bidaneh Sefid	53.92 c	36.42 f	53.14 b	64.88 bc
Druzhba	46.21 ef	31.35 g	42.49 f	58.42 e
Fakhri Zoudras	48.19 de	38.12 ef	50.18 c	60.49 d
Fiesta	46.82 ef	39.40 de	48.96 cd	57.60 e
Flame Seedless	59.45 a	48.12 ab	59.14 a	66.12 b
Galin Barmaqi	55.76 bc	44.32 c	55.30 b	63.23 c
Khalili	48.17 de	33.54 fg	46.27 d	60.12 d
Mirzaei	45.54 f	35.11 f	46.70 d	61.33 d
Monaqa	53.40 c	42.30 cd	49.32 c	61.39 d
Moukhtchaloni	41.35 g	28.45 h	44.12 e	55.12 f
Perlette	61.59 a	49.81 a	61.15 a	69.13 a
Peykani	57.30 b	47.42 b	54.60 b	63.45 c
Qareh Shahani	51.34 cd	43.52 c	50.07 c	58.92 e
Ruby Seedless	62.15 a	47.83 b	60.23 a	69.17 a
Shirazi	53.12 c	48.35 ab	50.75 c	63.01 c
Siahe Qarabaq	48.81 de	34.61 f	47.51 cd	58.13 e
Thompson Seedless	56.00 b	41.50 d	54.75 b	65.16 b
Volgo-Don	47.41 e	30.17 gh	44.81 e	55.47 f
Mean	51.95	39.69	51.01	61.48

Means followed by similar letters are not statistically different (P ≤ 0.05) as compared by Duncan’s

multiple range test.

Table 6. Changes in total phenol content of 20 V. vinifera cultivars during the hardening and dehardening period.

	Total phenol (mg g^−1 DW)
	Autumn	Winter		Spring
Grapevine cultivars	10 Nov.	12 Jan.	11 Mar.	10 Apr.
Asgari	10.88 ef	15.91 e	12.22 cd	9.41 b
Bidaneh Qermez	13.87 bc	17.30 c	12.42 c	10.29 a
Bidaneh Sefid	10.25 f	14.28 f	11.39 e	8.95 c
Druzhba	13.92 bc	17.43 c	13.12 b	10.25 a
Fakhri Zoudras	14.16 b	17.65 bc	12.10 d	9.68 ab
Fiesta	11.20 e	14.50 f	10.09 f	7.80 d
Flame Seedless	10.13 f	14.35 f	9.46 fg	6.29 e
Galin Barmaqi	11.25 e	15.92 e	13.32 b	9.24 b
Khalili	14.15 b	17.95 b	12.32 cd	10.14 a
Mirzaei	12.32 d	16.89 d	12.08 d	10.1 a
Monaqa	11.13 e	15.45 e	12.61 c	9.75 ab
Moukhtchaloni	14.35 b	18.95 a	14.21 a	10.24 a
Perlette	9.45 g	13.30 g	8.45 h	5.21 f
Peykani	11.36 e	15.84 e	12.50 c	8.94 c
Qareh Shahani	12.24 d	16.56 d	12.91 bc	10.50 a
Ruby Seedless	10.60 ef	14.21 f	9.40 fg	5.21 f
Shirazi	13.25 c	14.57 f	9.23 g	6.24 e
Siahe Qarabaq	13.50 c	17.48 c	12.48 c	10.21 a
Thompson Seedless	10.22 f	14.21 f	9.13 g	7.22 d
Volgo-Don	15.10 a	19.13 a	14.25 a	9.83 ab
Mean	12.17	16.09	11.68	8.78

Means followed by similar letters are not statistically different (P ≤ 0.05) as compared by Duncan’s

multiple range test.

Table 7. Pearson correlation coefficients among electrolyte leakage (EL) LT₅₀, tetrazolium stain test (TST) LT₅₀, and the concentrations of soluble carbohydrates, proline, water content and phenolic compound in canes of 20 Vitis vinifera cultivars.

	Autumn	Winter		Spring
variables	10th Nov.	12th Jan.	11th Mar.	10th Apr.
TST LT50
EL LT50	0.89**	0.86**	0.83**	0.70**
Soluble carbohydrates
EL LT50	-0.74**	-0.83**	-0.80**	-0.71**
TST LT50	-0.83**	-0.77**	-0.66**	-0.59**
Proline
EL LT50	-0.67**	-0.77**	-0.62**	-0.52**
TST LT50	-0.74**	-0.64**	-0.67**	-0.59**
Water content
EL LT50	0.67**	0.83**	0.73**	0.64**
TST LT50	0.81**	0.81*	0.74**	0.49**
Phenolic compounds
EL LT50	-0.69**	-0.81**	-0.80**	-0.68**
TST LT50	-0.79**	-0.79**	-0.71**	-0.58**

^{*, **} Significant at P ≤ 0.05 or 0.01, respectively.LT₅₀: lethal temperature at which 50% of the total ion leakage occurs; or, in the case of tetrazolium stain test, the lethal temperature at which 50% of the tissues are dead.

Soluble Carbohydrate

Soluble carbohydrates increased during cold acclimation from Nov. to Jan., then started to decrease in Mar. and were the lowest in Apr.; overall means were 46.47, 97.63, 59.95 and 23.33 mg g⁻¹ DW, respectively (Table 3). In Nov., significant differences (p ≤ 0.001) existed in the concentration of soluble carbohydrates among cultivars. Maximum amounts were related to ‘Moukhtchaloni’ and ‘Volgo-Don’, whereas, the least rate was observed in ‘Perlette’ and ‘Ruby Seedless’. From Nov. to Jan., the concentration of soluble carbohydrates increased considerably in all cultivars, this increase was concomitant with the LT₅₀ decrease. At this stage, maximum amounts of soluble carbohydrates (119–123 mg g⁻¹ DW) was detected in ‘Moukhtchaloni’, ‘Volgo-Don’, ‘Druzhba’, while the minimum amount (72–75 mg g⁻¹ DW) was found in ‘Shirazi’, ‘Perlette’ and ‘Fiesta’. A slight reduction was found in soluble carbohydrates in Mar., compared to Jan, which was consistent with the LT₅₀ profile in this sampling stage. Soluble carbohydrates were considerably dropped in Apr., and a significant difference was observed among cultivars at this stage. High correlations were found between EL LT₅₀ and TST LT₅₀, and soluble carbohydrate concentrations in all sampling stages (Table 7).

Proline

Proline concentrations had a narrow range of variation among cultivars in Nov., ranging from 1.25 to 2.10 µmol g⁻¹ FW (Table 4). A dramatic increase in proline was observed during the hardening period in all cultivars. The highest proline concentration in Jan. was found in ‘Moukhtchaloni’ (7.63 µmol g⁻¹ FW), followed by ‘Volgo-Don’, ‘Bidaneh Qermez’ and ‘Druzhba’. All cultivars showed a slight decrease in proline concentration in Mar., compared to Jan. (Table 4). In Apr., before bud break stage, amounts of proline decreased (3.34 µmol g⁻¹ FW in average) in all cultivars and those with lower LT₅₀ values had higher proline content. The highest amount of proline in Apr. was found in Russian cultivars, and the lowest in ‘Thomson Seedless’. There were high correlations between CH and proline concentrations of cultivars in all sampling stages (Table 7).

Water Content

There were significant differences (p ≤ 0.01) in water content among cultivars in all sampling stages. From Nov. through Jan., marked cane dehydration took place in all cultivars; while the water content of canes increased with the onset of warm temperatures in Mar., and reached the highest amounts in Apr. (Table 5). In Nov., the highest water content of canes was presented in ‘Ruby Seedless’, ‘Flame Seedless’ and ‘Perlette’ while the lowest amounts were observed in ‘Moukhtchaloni’. In Jan., the water content of canes decreased significantly in all cultivars (overall mean 39.69% FW); however, cultivars did not show similar rates of reduction. Water content showed a general increase throughout the deacclimation period, the greatest increment was seen in Apr. The lowest water content in Apr. was found in ‘Moukhtchaloni’ and ‘Volgo-Don’, while the highest amounts were observed in ‘Ruby Seedless’ and ‘Perlette’. Positive correlations were found between EL LT₅₀ and TST LT₅₀ values and water content of canes in all stages; the highest correlations were seen in Jan. (Table 6).

Total Phenolic Content

Total phenolic content increased during cold acclimation from Nov. to Jan., then started to decrease in Mar. and was the lowest in Apr.; overall means were 12.17, 16.09, 11.69 and 8.78 mg g⁻¹ DW, respectively (Table 6). In Nov., significant differences (p ≤ 0.001) existed in the concentration of total phenolic content among cultivars. Maximum amounts were related to ‘Volgo-Don’, whereas, the least rate was observed in ‘Perlette’. From Nov. to Jan., the concentration of total phenolic content increased considerably in all cultivars, this increase was concomitant with the LT₅₀ decrease. At this stage, maximum amounts of total phenolic content were detected in ‘Moukhtchaloni’, and ‘Volgo-Don’, while the minimum amount was found in ‘Perlette’. A slight reduction was found in total phenolic content in Mar., compared to Jan, which was consistent with the LT₅₀ profile in this sampling stage. Total phenolic content was considerably dropped in Apr., and a significant difference was observed among cultivars at this stage. High correlations were found between EL LT₅₀ and TST LT₅₀, and total phenolic content in all sampling stages (Table 7).

Discussions

Grapes are a worldwide economically important fruit crop, but display a great level of cold sensitivity and are frequently affected by winter damaging especially in cold climate. Frost injury is a serious problem which can damage vine canes and trunks, resulting in trunk cracking and crown gall infection, or even kill the whole vine (Fennell, 2004). Therefore, it is critical to understand CH of grapevines cultivars. Indeed, suitable cultivar selection is the first consideration to protect vines from winter injury (Fennell, 2004). Differences in the CH of cultivars may be small but will result in large differences in the field survival of these cultivars (Fennell, 2004; Wolf and Cook, 1992). In the present study, two laboratory methods of EL and TST were used for screening of CH of 20 grapevine cultivars. Significant variation in CH was observed among the 20 grapevine cultivars and EL and TST measurement allowed us to discriminate grapevine cultivars, CH determined by these methods correlated well (Table 6). The correlation analyses among the two methods used to estimate LT₅₀ showed that, in general, cultivars were ranked similarly across the methods. In Jan. sampling stage, ‘Moukhtchaloni’, ‘Volgo-Don’, ‘Druzhba’, ‘Khalili’, ‘Bidaneh Qermez’, ‘Bidaneh Sefid’ and ‘Monaqa’ were usually among the cultivars with the lowest LT₅₀ values, ‘Fakhri Zoudras’, ‘Asgari’, ‘Peykani’, ‘Siahe Qarabaq’, ‘Qareh Shahani’, ‘Galin Barmaqi’, ‘Mirzaei’, ‘Thompson Seedless’ and ‘Fiesta’ exhibited moderate LT₅₀ values, while ‘Flame Seedless’, ‘Shirazi’, ‘Perlette’ and ‘Ruby Seedless’ were usually among the cultivars with the highest LT₅₀ values.

EL LT₅₀ of all cultivars considerably decreased in Jan., and a relatively lower variation of CH was found among cultivars. There is evidence that cell wall properties change during cold acclimation, with an increase in lignification and suberization of cell walls during the fall (Griffith and Brown, 1982). The lower EL LT₅₀ of mid-winter samples of grapevine cultivars might be due to the changes in cell wall properties that provide resistance to the diffusion of electrolytes from the cells to the extracellular water. ‘Shirazi’ like with ‘Flame Seedless’, ‘Perlette’ and ‘Ruby Seedless ’was one of the least CH cultivars from Jan. through Apr. ‘Shirazi’ is one of the leading cultivars in southern regions of Iran (lat. 30–31° N) with relatively mild winters. The rest of the cultivars mentioned above are released in California which have moderately mild winter. Generally, species/accessions from the north are more cold-hardy than those from the south, with accessions from similar latitudes usually being a little different in this regard (Zhang et al., 2012).

Soluble carbohydrates increased during cold acclimation from Nov. to Jan., then started to decrease in Mar., and were the lowest in Apr. This result is consistent with previous studies on the seasonal changes of carbohydrates observed in other grape cultivars (Ershadi et al., 2016; Jones et al., 1999). Correlation coefficients between CH and soluble carbohydrates in all stages were highly significant (Table 6), and the strongest correlations were observed in Jan., the full acclimation stage. The Maximum amounts of soluble carbohydrates (119–123 mg g⁻¹ DW) was detected in ‘Moukhtchaloni’, ‘Volgo-Don’, ‘Druzhba’, as a cold hardy cultivars, which confirms the involvement of carbohydrates in freezing protection and dormancy status in grapevine (Hamman et al., 1996; Valle, 2002). Under cold stress, polysaccharides are hydrolyzed to soluble sugars, which increase the osmotic potential of the cytoplasm and lower the freezing temperature (Zhang et al., 2012). Soluble sugars can also function as protective substances and their concentration is positively correlated with cold hardiness (Jones et al., 1999; Zhang et al., 2012). Soluble carbohydrates decrease substantially in Apr. which may be due to its allocation to processes such as cell growth or conversion to starch during deacclimation stage (Morin et al., 2007).

There was a close relationship between proline content and CH of canes in Jan; however, this relationship was relatively weaker in other sampling stages. In higher plants, proline concentration is associated with serious abiotic stresses such as chilling stress (Taylor et al., 2004); it has also been shown that proline directly protects key cellular macromolecules, in particular, the lipid membranes and proteins such as enzymes (Verbruggen and Hermans, 2008; Zhang et al., 2012). Free proline can help maintain osmotic equilibrium between the symplast and apoplast and thus aid in resisting low-temperature damage by maintaining the functional integrity of the cellular membranes (Dionne et al., 2001). Under low-temperature stress, freezing-tolerant grape cultivars accumulate free proline, more than the less resistant ones (Dionne et al., 2001). Therefore, soluble carbohydrates and free proline contents may be considered as indicators for screening the CH of grape cultivars, especially in midwinter.

Phenolic compound concentration of all cultivars increased considerably during the acclimation phase and reached a peak in Jan., this increase was concomitant with the LT₅₀ decrease. At this stage, maximum amounts of total phenolic content were detected in ‘Moukhtchaloni’, and ‘Volgo-Don’, while the minimum amount was found in ‘Perlette’. The relationship between phenolic compounds and low-temperature tolerance was reported by Cansev et al. (2012) in olive and by Karimi et al. (2015) in grape. Biosynthesis of secondary metabolites such as phenolic compounds is partially affected by regulating the genetic expression and mainly by environmental factors such as low-temperature stress (Dixon and Paiva, 1995). Under low-temperature stress, phenylalanine amonialyse activity increases but the activity of phenolic compound oxidative enzymes such as peroxidase and polyphenol oxidase decreases (Pakkish et al., 2009; Rivero et al., 2001). This increases the accumulation of soluble phenol compounds and may act as an adaptive mechanism to overcome cold-induced oxidative stress in vines (Balasundram et al. 2007). In the present study, the maximum amount of phenolic compounds was observed in Jan. at the same time as the deep winter dormancy, which indicates the association of these compounds with cold tolerance and their protective role in dormant buds of the vine. During the dormant phase, a number of growth inhibitors including phenolic compounds (Caffeic acid, Naringin, Phloridzin, Quercitrin) accumulate in the buds of most trees and after getting enough cold in winter, the concentration of these compounds is minimized at bloom (Codignola et al., 1988). It seems that the precipitate of most phenolic compounds in the canes epidermal and buds scales in the cultivars prevents the development of ice crystals within the internal parts of the bud (Chalker-Scott, 1988).

The close inverse relationship was seen between CH and cane water content during all stages (Table 6), and reduced water content was associated with increased CH; confirms previous observations in grapevines (Ershadi et al., 2016; Salzman et al., 1996; Valle, 2002). The loss of water content in Jan. is very often explained in terms of cold adaptation via the concentration of the cell sap (Pogosyan et al., 1975). However, by spring the dehydrative resistance decreased concomitantly with an increase in water content which reduced CH especially in cultivars with lower chilling and heat requirements such as ‘Shirazi’ and ‘Ruby Seedless’.

In conclusion, the pattern of CH changes in different grapevine cultivars was consistent with canes tissue dehydration and accumulation of cryoprotectants, namely soluble carbohydrates, proline, and phenolic compounds during hardening and dehardening cycle. Thus, it is likely that the LT50 profile is a consequence of solute accumulation as well as cane dehydration. According to the two methods used to estimate FT, we can cluster the 20 cultivars into three groups from the CH (1) to the least CH (20). Cold hardy cultivars were (1) Moukhtchaloni, (2) Volgo-Don, (3) Druzhba, (4) Khalili, (5) Bidaneh Qermez, (6) Bidaneh Sefid and (7) Monaqa. Moderately cold hardy cultivars were (8) Fakhri Zoudras, (9) Asgari, (10) Peykani, (11) Siahe Qarabaq, (12) Qareh Shahani, (13) Galin Barmaqi, (14) Mirzaei, (15) Thompson Seedless and (16) Fiesta; Most cold-sensitive cultivars were (17) Flame Seedless, (18) Shirazi, (19) Perlette and (20) Ruby Seedless. The results of this study are valuable to the nursery industry and viticulturist to select suitable cultivars for sustainable production in cold climates and to the grape breeder to advance grape CH programs.

Acknowledgments

This work was supported by the Research Institute of Grape and Raisin (RIGR) of Malayer University Fund/Foundation. We acknowledge the technical assistance of vineyard and laboratory; Ahmad Asadi and Mehran Moghadampour from University of Malayer.

Additional information

Funding

This work was supported by the Malayer university [84.5-1-408].