Abstract
Cold injury is a major cause of economic loss in winegrapes (Vitis vinifera L.) grown at high latitudes. The objective of this study was to investigate the relationship between dormancy and cold hardiness transitions in two cultivars with differing freeze tolerance (Chardonnay and Cabernet Sauvignon). Cold hardiness was measured by differential thermal analysis, and a bud forcing bioassay was used to measure the stage and depth of dormancy. Canes were sampled from field-grown grapevines in Parma, ID at periodic intervals during two consecutive winters. Both cultivars transitioned into endodormancy each year in September when day length was ~12.5 hr. Cold acclimation in both cultivars occurred each year in October during endodormancy and steadily increased during ecodormancy to a max hardiness in December. Effective temperatures for release from endodormancy were lower for Chardonnay (−3°C) than for Cabernet Sauvignon (3°C), and, each year, Chardonnay transitioned to ecodormancy earlier than Cabernet Sauvignon. From October to December, Chardonnay buds were more cold hardy than Cabernet Sauvignon buds. The number of days to budbreak under forcing conditions increased steadily during endodormancy and decreased during ecodormancy. Resistance to deacclimation during ecodormancy was inversely related to the level of bud cold hardiness and the duration of time in ecodormancy, suggesting that the mechanisms that impart hardiness may interact with those involved in resumption of growth. Results from this study show that the influence of autumn weather events on dormancy and cold hardiness transitions can affect vulnerability to subsequent cold injury and have important implications under global climate change. Differences between cultivars in dormancy and cold hardiness transitions can be used to improve cultivar and site selection.
The European winegrape Vitis vinifera L. is a cold-sensitive woody perennial that utilizes dormancy, extracellular freezing, and supercooling as mechanisms to avoid injury when exposed to freezing temperatures (Keller 2015). Dormancy is a state of suspended growth and is categorized as para-, endo-, or ecodormancy according to the mechanism of growth restriction (Lang et al. 1987). Para- and ecodormant buds can resume growth under favorable environmental conditions. Paradormant buds transition into endodormancy in autumn in response to environmental cues, such as short photoperiod and/or cold nighttime temperatures (Fennell 2004). Endodormant buds require exposure to cold (i.e., chilling) before they can resume growth. The temperature and duration of cold exposure (accumulated chill-unit hours) required for transition from endo- to ecodormancy appear to vary according to grape cultivar, but overall, information in this area is lacking (Ferguson et al. 2014).
Cold hardiness—the ability of tissue to resist lethal intracellular ice nucleation during exposure to freezing temperature—is dynamic throughout the winter season (Ferguson et al. 2014), and bud and cane tissues acclimate and deacclimate in response to fluctuating winter temperatures (Hubackova 1996). The relationship between dormancy phases and bud cold hardiness has not been well characterized. Cold acclimation appears to depend on growth cessation; however, it remains unknown whether acquisition of cold hardiness is coincident with, contingent upon, or independent of, dormancy transitions.
Winter survival depends upon timely cold acclimation in autumn, resistance to deacclimation during transient winter warming events, and/or reacclimation when organs are exposed to subsequent cold temperatures (Kalberer et al. 2006). After ecodormant buds initiate budbreak, they lose the ability to reacclimate in response to cold temperatures (Ding and Gu 2001). The depth of dormancy and dormancy transitions may correspond to changes in bud cold hardiness and may influence the capacity of bud tissue to acclimate, deacclimate, and reacclimate (Arora et al. 2003, Arora and Rowland 2011). Wolf and Cook (1992) reported that max winter bud cold hardiness was associated with reduced resistance to deacclimation; however, much remains unknown about the relationship between de- and reacclimation processes during ecodormancy and resumption of growth.
In this study, we investigated the relationship between dormancy stage and seasonal changes in cold hardiness in the winegrape cultivars Cabernet Sauvignon and Chardonnay over two consecutive winters. We also compared the efficacy of different temperatures for fulfilling the chilling requirements of both cultivars for transition to ecodormancy. Finally, we investigated the relationship between dormancy transitions and the ability to resist deacclimation in response to winter warming events.
Materials and Methods
Field site
The grapevines used in this study were grown in an experimental vineyard at the University of Idaho Parma Research and Extension Center, Parma, ID (43°49′N; 116°56′W; elevation 760 m asl) in USDA plant hardiness zone 6a, in which the average annual min temperature was 23 to 21°C (1976 to 2005). Details of the research vineyard experimental design and edaphoclimatic conditions have been described by Shellie (2007). Chardonnay and Cabernet Sauvignon were planted on their own roots in 1998 as 8-vine panels in five replicate blocks. Rows were oriented north-to-south with vine spacing of 2.1 m within rows and 2.7 m between rows. The vines were cordon-trained and spur-pruned to ~32 buds/vine with vertical shoot positioning. Vines were drip-irrigated and managed according to local commercial practices as described by Shellie (2007) and Watson (1999).
Cold hardiness evaluation
Bud and cane cold hardiness were measured using differential thermal analysis (DTA) as described by Mills et al. (2006) and were evaluated using the temperature of the thermoelectric module (TEM) at the occurrence of a low-temperature exotherm (LTE). Bud cold hardiness was calculated as the LTE temperature at which 50% of sampled buds were killed (LT50) and as the average LTE. For a particular cultivar and sampling date, mean LTE values were first estimated for each block and used to conduct the statistical analysis, which computed the average LTE and standard error. The cold hardiness of phloem and xylem tissues was reported as the LTE temperature at which 10% of sampled canes displayed an LTE (LT10).
Field sampling of shoots or canes was performed monthly from July 2011 to Jan 2012 and biweekly from Jan to March 2012 and from Aug 2012 to March 2013. On each monthly sampling in 2011 to 2012 and bi-weekly sampling from Aug 2012 until Jan 2013, tissue was used to evaluate bud cold hardiness and stage of dormancy. Tissue from bi-weekly samplings from Jan until March 2012 and from Feb to March 2013 was only used for evaluation of bud cold hardiness. Shoots or canes were removed from random vines in each of the five replicate field blocks according to Mills et al. (2006). The canes sampled from each field replicate were randomly allocated for DTA or the bud forcing assay. For the canes used for DTA, 40 buds were excised from node positions three through eight according to Wolf and Poole (1987), and 15 internode segments were trimmed to 35 mm for measuring xylem and phloem hardiness. Three or four excised buds were placed into one of 13 TEM wells, and three internode segments were placed into one of five TEM wells. High-density foam was placed over the tissue in each TEM well prior to closure to hold tissue in place and ensure good tissue surface contact with the TEM. The trays were stacked inside a programmable freezer. The freezer temperature was held at 4°C for 1 hr and then decreased to −40°C at a rate of 4°C/hr, as described by Mills et al. (2006).
Dormancy evaluation
A bud forcing bioassay was used to categorize the stage and depth of dormancy. Endodormancy was defined as budbreak <50%, paradormancy (before endodormancy) or ecodormancy (after endodormancy) was defined as budbreak >50% after 60 days of exposure to bud forcing conditions (Dokoozlian et al. 1995, Dokoozlian 1999). Budbreak was defined as growth stage 4 (visible green leaf tips) of the modified E-L system (Coombe 1995). Depth of dormancy was calculated as the average number of days required for buds to break (ADBB), using the equation:
, where ni is the number of buds that reached stage 4 or beyond within consecutive days, Di is the number of days elapsed since the beginning of the test, and N is the total number of buds that reached stage 4 or beyond during the 60-day assay.
For the bud forcing assay, 20 single-node segments were sectioned at node positions three through six and placed into wet floral foam blocks inside stainless steel trays filled with deionized water. The trays were maintained under day/night conditions of 25/20°C (±2°C) with a 15-hr photoperiod in a greenhouse at Boise State University. The growth stage of the bud on each single-node cane segment was visually assessed every other day for up to 60 days, and the date of budbreak was recorded. At the end of the 60-day period, buds that had not reached growth stage 4 were dissected and categorized as either dead (brown) or alive (green). The sample size was adjusted for the number of dead buds to calculate percent budbreak.
Additional canes of each cultivar were sampled on 8 Oct 2012 and used to evaluate the effectiveness of different temperatures for transitioning to ecodormancy (chill-unit efficacy). A subset of canes was used immediately for determining percent budbreak using the 60-day bud forcing assay and served as the unchilled control. The remaining canes were grouped into bundles, moist towels were placed at either end, and the bundle was wrapped in polyvinyl chloride (PVC) film (All-Purpose Laboratory Wrap, Fisherbrand). The wrapped bundles were stored for 3 wks in temperature-controlled chambers at −8, −5, −3, 0, or 3°C (±0.5°C). After the 3-wk cold exposure, the canes from each chamber were sectioned into 60 single-node segments per cultivar and used to evaluate percent budbreak by the bud forcing assay described above. Cumulative percent budbreak during the 60-day bud forcing assay was used to measure chill-unit efficacy.
Deacclimation and reacclimation
Canes sampled biweekly from Jan to March in 2012 and 2013 were used for deacclimation and reacclimation studies. For each cultivar, five canes per field replicate were grouped into five bundles, each with one cane from each field replicate. One bundle was used for DTA on the day of field sampling. Canes in the remaining bundles were trimmed to include node positions three through eight; moist paper towels were placed at either end of each bundle, and the bundles were wrapped in PVC film. To induce deacclimation, the bundles were placed into a temperature-controlled chamber with a day/night temperature of 25/20°C (±0.5°C) and 15-hr photoperiod for two or four days. Reacclimation was induced in buds that had completed a four day deacclimation exposure by immediately transferring them to a temperature-controlled chamber at 0°C (± 0.5°C) in the dark for three or five days. For each cultivar, one cane bundle was removed after two and four days of deacclimation and after three and five days of reacclimation. Bud cold hardiness was measured using DTA as described previously (Mills et al. 2006).
Statistical analyses
The effects of cultivar, time of sampling, and the interaction between these factors on cold hardiness were evaluated using a generalized least squares (GLS) model in the nlme package in R (R version 3.3.0, http://www.r-project.org/). Similarly, the effects of cultivar, time of sampling, and the interaction between these factors on budbreak under forcing conditions were analyzed using a generalized linear model (GLM) with Poisson-distributed errors in R. Average days to budbreak under forcing conditions were analyzed using a GLS model. When appropriate, different variances were modeled using the GLS procedures to account for unequal variance among treatments. The Shapiro-Wilk test was used to ascertain the normality of residuals in the GLS model; goodness of fit of the GLM model was ascertained using a chi-square test. Post-hoc multiple comparisons were analyzed with Tukey’s significant difference test using the lsmeans package in R.
The Kaplan-Meier method was used to compare budbreak under different chilling temperatures (McNair et al. 2012). The Kaplan-Meier curve represents the probability that budbreak will not occur during exposure to chilling. Statistical differences among Kaplan-Meier curves were evaluated with the log-rank test and adjusted using the Holm-Sidak method for multiple comparisons in SigmaPlot 12 (Systat Software).
For a particular sampling date and cultivar, the rate of deacclimation was estimated from the slope of the linear regression between bud LT50 values and days of deacclimation (0, 2, and 4 days). Similarly, reacclimation rates were estimated from the slope of the linear regression between bud LT50 values and days of reacclimation (0, 3, and 5 days, where day 0 of reacclimation is day 4 of deacclimation). Statistical differences between cultivars or sampling dates were determined by ANCOVA in R. This approach involved using the rates of de- or reacclimation to test more than one hypothesis, which increased the chance of type I error. To reduce this problem, the significance of p values was evaluated after Bonferroni adjustment.
Results
Dormancy transitions and cold hardiness
The timing of harvest each year was based on fruit maturity (~23% soluble solids concentration and titratable acidity of ~6 g/L), and buds of both cultivars were endodormant when fruit were removed from the vine. In 2011, Cabernet Sauvignon was endodormant by 23 Aug and harvested on 19 Oct; and, in 2012, was endodormant by 10 Sept and harvested on 4 Oct. Chardonnay was endodormant by 20 Sept and harvested on 6 Oct in 2011; and, in 2012, was endodormant by 10 Sept and harvested on 15 Sept (Tables 1 and 2).
Percent budbreak and average days to budbreak (ADBB) for single-node cane segments repeatedly sampled from field-grown Cabernet Sauvignon (CS) and Chardonnay (CH) grapevines from July 2011 through Jan 2012. Percent budbreak was estimated after a 60-day exposure to day/night temperatures of 25/20°C with a 15-hr photoperiod. ADBB was calculated as:
, where ni is the number of buds that reached stage 4 or beyond within consecutive days, Di is the number of days elapsed since the beginning of the test, and N is total number of buds that reached stage 4 or beyond during the 60-days of the assay.
Percent budbreak and average days to budbreak (ADBB) for single-node cane segments repeatedly sampled from field-grown Cabernet Sauvignon (CS) and Chardonnay (CH) grapevines from Aug 2012 through Dec 2012. Percent budbreak was estimated after a 60-day exposure to day/night temperatures of 25/20°C with a 15-hr photoperiod. ADBB was calculated as: as:
, where ni is the number of buds that reached stage 4 or beyond within consecutive days, Di is the number of days elapsed since the beginning of the test, and N is total number of buds that reached stage 4 or beyond during the 60-days of the assay.
On some, but not all field-sampling dates, the cultivars differed (p < 0.0001) in percent budbreak and ADBB (Tables 1 and 2). Buds of both cultivars were paradormant on the first field-sampling date of each year; however, in 2011, the low percent budbreak in Cabernet Sauvignon suggested that it had begun to transition into endodormancy. On the second sampling date in 2011 (23 Aug), Cabernet Sauvignon was endodormant and Chardonnay was paradormant. Both cultivars were endodormant by the third sampling in 2011 (20 Sept) and by the second sampling in 2012 (10 Sept). Each year, Chardonnay began to transition into ecodormancy earlier than Cabernet Sauvignon, but both cultivars were ecodormant by 15 Dec in 2011 and 5 Nov in 2012. For both cultivars, the transition to ecodormancy occurred earlier in 2012 than in 2011.
In both years and for both cultivars, ADBB increased steadily during endodormancy and decreased during ecodormancy (Tables 1 and 2). ADBB increased during endodormancy in both cultivars each year by ~2-fold, and decreased during ecodormancy in both cultivars by ~50 and 20% in 2011 and 2012, respectively. Budbreak occurred earlier in Chardonnay than in Cabernet Sauvignon in four out of the seven sampling dates in 2011 and six out of the nine sampling dates in 2012.
In both years, bud cold hardiness was affected by sampling date, cultivar, and their interaction (p < 0.0001). Similar results were obtained with LT50 and average LTE values, and only the latter are presented. In both years and in both cultivars, cold hardiness of buds, xylem, and phloem tissue increased during October when vines were endodormant (Figures 1 and 2).
Field-grown Cabernet Sauvignon (CS) and Chardonnay (CH) grapevines repeatedly sampled from Sept 2011 until March 2012 in Parma, ID. (A) Seasonal daily temperature. (B) Seasonal average bud low temperature exotherm (LTE) values. Percent budbreak under forcing conditions illustrates temporal changes in dormancy relative to changes in cold hardiness. Stars above symbols indicate significant differences (p < 0.05) in LTE values between cultivars. (C) Lethal temperature for 10% (LT10) of sampled phloem and xylem tissue.
Field-grown Cabernet Sauvignon (CS) and Chardonnay (CH) grapevines repeatedly sampled from Aug 2012 until March 2013 in Parma, ID. (A) Seasonal daily temperature. (B) Seasonal average bud low temperature exotherm (LTE) values. Percent budbreak under forcing conditions illustrates temporal changes in dormancy relative to changes in cold hardiness. Stars above symbols indicate significant differences (p < 0.05) in LTE values between cultivars. (C) Lethal temperature for 10% (LT10) of sampled phloem and xylem tissue.
In September of both years, min daily temperatures remained above 0°C, and the average max and min temperatures were higher than in October. The number of days in September when min temperatures were between 0 to 7°C was greater in 2012 (15 days) than in 2011 (9 days). In October, weather events with daily temperatures below 0°C occurred earlier and more frequently in 2012 than in 2011. In 2011, the first weather event with min temperatures below 0°C began on 25 Oct and continued for five days. In 2012, the first of two weather events when min temperatures were below 0°C for six consecutive days occurred on 4 Oct and the second occurred on 21 Oct. Each year, the cold hardiness of both cultivars increased steadily in October, when vines were endodormant, and in November and December, when vines were ecodormant. Chardonnay buds were more cold hardy than Cabernet Sauvignon buds from Oct to Dec 2011 and from Sept to Nov 2012. Max seasonal bud cold hardiness in both cultivars occurred in December in both years, when buds were ecodormant. In January, min daily temperatures were lower in 2013 than in 2012, and Chardonnay buds were more cold hardy than Cabernet Sauvignon buds in 2013. In February of both years, there was little change in daily min and max temperatures or bud cold hardiness; however, phloem and xylem tissues of both cultivars began to lose cold hardiness, and the phloem and xylem tissues of Chardonnay were less cold hardy than Cabernet Sauvignon. In March of both years, daily min and max temperatures increased and bud cold hardiness decreased in both cultivars, and Chardonnay buds were less cold hardy than Cabernet Sauvignon buds in 2012. Budbreak under field conditions occurred earlier in Chardonnay (24 April 2012, 22 April 2013) than in Cabernet Sauvignon (28 April 2012, 1 May 2013). The level of cold hardiness in March relative to the preceding September was greater in buds and similar in phloem and xylem in both years, and phloem tissue was less cold hardy than xylem tissue from September until March in both cultivars.
Response to chilling temperatures
The temperature that best fulfilled chilling requirements differed for the two cultivars (Figure 3). For Cabernet Sauvignon, three weeks of exposure to a constant temperature of 3 or 0°C resulted in 80 and 60% budbreak, respectively, after ~45 days under bud forcing conditions. The probability of budbreak was similar after exposure to 3 or 0°C (p = 0.16), and was higher than in the unchilled control (p < 0.001) and all other tested chilling temperatures. In Chardonnay, three weeks of exposure to −3°C was associated with 100% budbreak after 21 days under bud forcing conditions. The log-rank test (p < 0.001) showed that the probability of budbreak at −3°C was higher than at any other exposure temperature and higher than the unchilled control that was subjected to the forcing bioassay on the day of field sampling. The median time to budbreak after exposure to 0, −3, or −5°C was less (p < 0.001) for Chardonnay than for Cabernet Sauvignon. Cumulative budbreak at 3°C was similar for both cultivars (p = 0.35); no budbreak was observed for either cultivar after three week exposure to −8°C, which resulted in >90% bud mortality. Bud mortality after three week exposure to −5°C was higher in Cabernet Sauvignon (40%) than in Chardonnay (5%).
Constant temperature exposure of canes with endodormant buds to 3, 0, −3, −5, or −8°C to evaluate fulfillment of chilling requirements for transition to ecodormancy. Canes were field-sampled on 8 Oct 2012 from Cabernet Sauvignon (A) and Chardonnay (B) grapevines grown in Parma, ID.
Deacclimation and reacclimation
The rate of deacclimation for Cabernet Sauvignon was significantly higher than zero on four out of five sampling dates in both years (Table 3). In Chardonnay, the rate of deacclimation was higher than zero on one out of five sampling dates (27 Feb) in 2012 and on every sampling date in 2013. For a particular sampling date, no differences in deacclimation rates were detected between cultivars, except on 28 Jan 2013, when the rate of deacclimation of Chardonnay was higher than that of Cabernet Sauvignon. Similarly, within a particular year and cultivar, there were no differences in deacclimation rates among sampling dates, except for Chardonnay in 2012, when the rate was higher on 27 Feb than on either sampling date in March. Clear differences in deacclimation rates occurred between the two years of the study: rates from January through March were higher in 2013 than in 2012. In particular, we compared within each cultivar deacclimation rates between buds that were collected nearly a year apart: 30 Jan 2012 versus 28 Jan 2013, 13 Feb 2012 versus 11 Feb 2012, 27 Feb 2012 versus 25 Feb 2013, and 12 March 2012 versus 11 March 2013. All of these comparisons revealed significant differences in deacclimation rates (p < 0.001) except for the January samplings of Cabernet Sauvignon.
Rates of deacclimation and reacclimation of ecodormant buds on canes repeatedly sampled from field-grown Cabernet Sauvignon (CS) and Chardonnay (CH) grapevines in Parma, ID from Jan through March in 2012 and 2013. Canes were deacclimated by exposure to a day/night temperature of 25/20°C for four days and immediately reacclimated by exposure to 0°C for five days. Rate of acclimation was determined from the slope of the linear relationships between acclimation time and bud LT50 values measured at the start, middle, and end of each acclimation exposure.
Reacclimation was evaluated immediately after deacclimation in both years; reacclimation rates did not differ from zero on any sampling date for either cultivar except for 30 Jan 2012, when the rate of reacclimation for Cabernet Sauvignon was less than zero. The prevalence of reacclimation values similar to zero indicates little to no reacclimation in either cultivar under the tested conditions.
Discussion
Dormancy transitions and seasonal changes in cold hardiness
Our results support the findings of others that showed a major role of photoperiod in the transition to endodormancy (Arora et al. 2003, Grant et al. 2013). Both cultivars were endodormant by 20 Sept 2011 and 10 Sept 2012 (Tables 1 and 2). Day length on these dates was 12.1 and 12.4 hr, respectively (www.solartopo.com/daylength.htm), similar to the day length Fennell and Hoover (1991) found effective for inducing endodormancy in V. labruscana under warm, controlled temperatures. The steady increase in ADBB in both cultivars from August to October in both years showed that the depth of endodormancy steadily increased through August to reach a max depth in late October (Tables 1 and 2). This increase in depth of endodormancy during August and September occurred when monthly chill-unit accumulation was negative due to a prevalence of exposure hours to temperatures ≥16°C (Richardson et al. 1974).
In each cultivar and each year of this study, cold hardiness increased during endodormancy after exposure to cold temperature, and the amount of increase corresponded to the amount of cold exposure (Figures 1 and 2). Average LTE values for each cultivar decreased each year during October. Chill-unit accumulation (Richardson et al. 1974) during October was less in 2011 (372) than in 2012 (637), and min daily temperatures below 0°C occurred later and less often in 2011 than in 2012. The average decrease in LTE values from September to October was less in 2011 (~2°C) than in 2012 (~4°C), and the average LTE value for each cultivar in October was ~4°C lower in 2012 than in 2011. These data support the notion that transition from para- to endodormancy is a prerequisite for subsequent cold acclimation, and that cold acclimation occurs mostly in response to decreasing autumn temperatures. Schnabel and Wample (1987) also showed that bud cold hardiness increased subsequent to the induction of endodormancy.
The transition to ecodormancy occurred approximately one month earlier in 2012 (Nov) than in 2011 (Dec) for both cultivars (Tables 1 and 2). Based on the model of Richardson et al. (1974), accumulated chill-unit hours during endodormancy in 2011 and 2012 were 179 and 244 for Cabernet Sauvignon and 588 and 34 for Chardonnay. Temperatures from 1.5 to 12.5°C are considered effective for chill-unit accumulation, and temperatures ≥16°C negate previously accumulated chill units (Richardson et al. 1974). The greater chill-unit accumulation in 2012 relative to 2011 for Cabernet Sauvignon during endodormancy was due to 53% fewer negative chill-unit hours (temperatures ≥16°C) and fewer incidences (35%) of positive chill units (temperatures equal to and between 1.5 and 12.5°C). For Chardonnay, the greater chill-unit accumulation during endodormancy in 2011 was associated with 58% more positive chill units and 28% less negative chill units. The most effective chill-unit temperature for Cabernet Sauvignon (3°C) was within the effective temperature range defined by Richardson et al. (1974); however, the temperature for Chardonnay (−3°C) was not, since temperatures lower than 1.5°C were considered ineffective. The higher number of calculated chill units for Chardonnay in 2011 was likely due to the lower incidence of days with temperatures lower than 1.5°C during the longer duration of endodormancy in 2011 relative to 2012. The earlier transition to ecodormancy observed in this study in both cultivars in 2012 corresponded with a higher incidence of days with temperatures lower than 1.5°C in September and October. The lack of correspondence between calculated chill units and transition to ecodormancy in Chardonnay further supports that temperatures lower than 0°C were effective for chill-unit accumulation.
There was no apparent influence of the timing of the transition into ecodormancy or the transition to ecodormancy on max midwinter cold hardiness in either cultivar. Buds continued to acclimate after the endo- to ecodormancy transition. Nevertheless, the steady decrease in ADBB of buds in December and January in both years for both cultivars (Tables 1 and 2) indicates that readiness to resume growth increased during ecodormancy. This finding supports that of Lavee and May (1997) who reported that the capacity to resume growth increased steadily during ecodormancy. Cold hardiness prediction models usually assume that ecodormancy primes buds for deacclimation (Ferguson et al. 2014); however, the coincident increase in cold hardiness and readiness to resume growth observed in this study during ecodormancy suggests that changes in hardiness may be independent of the ability to resume growth.
Rates of deacclimation tended to be lower in winter 2011 to 2012 than in winter 2012 to 2013 (Table 3), indicating higher resistance to deacclimation during the first year of the study. Two factors may have accounted for these differences in deacclimation rates: the LT50 values measured at the time of field sampling prior to experimental manipulation, and the duration of bud ecodormancy (Figure 4). Combining results from both cultivars and years, LT50 measured prior to experiment manipulation was negatively correlated with deacclimation rates (p = 0.018, r2 = 0.275). This correlation partially explains the difference in deacclimation rates between the years, because overall bud LT50 values were higher in the winter of 2011 to 2012 than in 2012 to 2013. Wolf and Cook (1992) reported that max winter bud cold hardiness was associated with reduced resistance to deacclimation. We did not observe clear differences in max cold hardiness between cultivars, but bud cold hardiness at the time of field sampling affected the resistance to deacclimation. Another difference between samples collected in the winter of 2011 to 2012 with those of 2012 to 2013 was the period that the vines had been ecodormant. This period was longer for buds sampled in winter 2012 to 2013 than in 2011 to 2012 due to the earlier transition into ecodormancy in autumn of 2012 to 2013. If the capacity to resume growth increases during ecodormancy, then midwinter samples collected in 2012 to 2013 should have greater capacity to deacclimate and resume growth than those collected the previous year. Results from a linear regression model that included LT50 values and duration of ecodormancy as predictors of deacclimation rates support this notion (Figure 4). Both predictors had a significant effect on deacclimation rates (p < 0.0001). The model that included LT50 values and duration of ecodormancy had an r2 value of 0.65, whereas prediction using only LT50 had an r2 value of 0.28. The effects of cold hardiness and dormancy status on resistance to deacclimation are poorly understood and likely to be affected by complex interactions (Arora et al. 2003, Wisniewski et al. 2014). Our results suggest that both the degree of cold hardiness and the timing of transition to ecodormancy influenced subsequent resistance to deacclimation.
Influence of duration of ecodormancy and bud cold hardiness on resistance to deacclimation. Symbols with drop-lines to the x–y plane indicate measured deacclimation rates, duration of ecodormancy, and cold hardiness of Cabernet Sauvignon and Chardonnay buds field-sampled biweekly from Jan through March in 2012 and 2013. Meshed area indicates predicted deacclimation rates (°C/day) from regression equation: −5.47 – 0.20 × LT50 + 0.019 × duration of ecodormancy. The p value for this model was 0.0001 with an r2 of 0.65.
The capacity to reacclimate is an important trait for tolerance of late-season cold events. Neither cultivar showed the ability to reacclimate under the controlled temperature conditions used in this study, and the explanation for this is not readily apparent. Daily DTA measurements have shown distinct cycles of acclimation in Merlot buds under field conditions (Keller 2015), and an exposure temperature of 0°C induced reacclimation in other species (Kalberer et al. 2007), but may be ineffective in grapevine. It is not likely that a lack of energy reserves inhibited reacclimation (Kalberer et al. 2007, Pagter et al. 2011), since there was no apparent decline in reacclimation ability over time (Table 3). Additional research using reacclimation temperature exposures below 0°C for longer periods under differing day/night photoperiods and temperatures is warranted.
Cultivar differences
The similar onset of endodormancy for both cultivars in this study, and the more rapid autumn cold acclimation of Chardonnay compared to Cabernet Sauvignon, suggests that day length response for induction of endodormancy was similar for the two cultivars, and that chill-unit accumulation and cold-hardiness transitions were more temperature-responsive in Chardonnay. Earlier and more rapid autumn cold acclimation in Chardonnay than in Cabernet Sauvignon was also observed under field conditions in eastern Washington (Ferguson et al. 2011, Wample et al. 2001), however, the relationship between cold hardiness and dormancy stage or depth was not evaluated. The results from this study suggest that autumn temperatures influence dormancy and cold hardiness transitions and may potentially reduce or increase vulnerability to cold injury. Chardonnay was more responsive than Cabernet Sauvignon to low autumn temperatures, and this led to more rapid chill-unit accumulation, earlier transition to ecodormancy, and greater cold hardiness. Resistance to deacclimation during ecodormancy was inversely related to the level of bud cold hardiness and the duration of ecodormancy. Chardonnay was more responsive to elevated temperatures and quicker to deacclimate than Cabernet Sauvignon.
The earlier transition to ecodormancy in Chardonnay compared to Cabernet Sauvignon under field conditions in each year of this study and the greater range of effective chill-unit temperatures observed under controlled temperature conditions, suggest that chill-unit requirements for transition to ecodormancy differ between these cultivars. The cumulative percent budbreak observed after exposure to 3°C was similar to that reported by Dokoozlian et al. (1995, 1999) for the table grape cultivar Perlette. We could not find other studies that evaluated chill-unit effectiveness at temperatures below 0°C, and such temperatures are not thought to contribute to chill-unit accumulation (Arora et al. 2003). Results from this study challenge that notion since −3°C contributed to chill-unit accumulation in both cultivars and −5°C contributed to chill-unit accumulation in Chardonnay.
Exposure to a constant temperature of −8°C for three weeks was lethal to buds of both cultivars. In January 2013, daily max temperatures were below −8°C, and min temperatures remained between −20 and −23°C for eight days (Figure 2). DTA-estimated cold hardiness was lower than these min temperatures, yet there was a high incidence of bud kill (data not shown). The number of dead buds subsequent to this prolonged cold event was greater in Cabernet Sauvignon than in Chardonnay. The discrepancy between DTA-estimated hardiness and actual bud survival might be explained by the fact that DTA measures acute cold stress due to intracellular freezing (i.e., instantly lethal temperatures), but does not account for the long-term stress of chronic exposure to cold and desiccation-associated injury due to extracellular freezing. No information is available about cultivar differences in susceptibility to injury during prolonged exposure to cold. At present, it is unclear whether the mechanisms that confer long-term cold tolerance are similar to the mechanisms involved in freeze avoidance during transient temperature fluctuations (Gusta and Wisniewski 2013, Wisniewski et al. 2014).
Conclusion
We monitored the progression of dormancy and cold hardiness under field conditions in Chardonnay and Cabernet Sauvignon, cultivars with different freeze tolerance. In both cultivars, cold acclimation was first apparent during endodormancy, and cold hardiness increased to a seasonal max during ecodormancy. This suggests that endodormancy may be a prerequisite for subsequent cold acclimation. The average number of days to budbreak under forcing conditions steadily increased during endodormancy and decreased during ecodormancy, suggesting that the capacity to resume growth increased during ecodormancy. Resistance to deacclimation during ecodormancy was inversely related to the level of bud cold hardiness and the duration of ecodormancy. This suggests an interaction between mechanisms that impart hardiness with those involved in resumption of growth. The influence of autumn temperatures on dormancy and cold hardiness transitions and resistance to deacclimation observed in this study has important implications for vulnerability to cold injury in changing global climates. Cold temperature events in autumn can influence subsequent susceptibility to injury during midwinter warming events by influencing the timing of transition to ecodormancy. Weather conditions during ecodormancy influence the level of cold hardiness attained. Cold temperatures that occur in both early autumn and midwinter increase the risk of injury after a midwinter warming event by decreasing the resistance to deacclimation. The differences in temperature responsiveness observed between cultivars show that there is an opportunity to mitigate potential cold injury through better matching of cultivar dormancy and hardiness transitions with the pattern of weather events at potential site locations.
Acknowledgments
Funding for this research was provided by the Idaho State Department of Agriculture Specialty Crop Block Grant Program and Agricultural Research Service project 5358-21000-034-00D entitled “Production Systems to Promote Yield and Quality of Grapes in the Pacific Northwest.” The authors thank John Ferguson, Alan Muir, and Monte Shields for their technical assistance. USDA is an equal opportunity provider and employer.
- Received August 2016.
- Revision received November 2016.
- Accepted December 2016.
Literature Cited
Vol 68 Issue 2
