Bud Hardiness and Deacclimation in Blueberry Cultivars With Varying Species Ancestry: Flowering Time May Not Be a Good Indicator of Deacclimation

Blueberry cultivars with varying percentages of species ancestry (V. corymbosum L., V. angustifolium Ait., V. ashei Reade, V. darrowi Camp) were assayed in mid-February to determine initial bud hardiness, and rates of deacclimation under constant temperature conditions. The LT50 (the temperature at which 50% lethality occurs) of detached shoots of field-grown plants of ‘Weymouth’, ‘Bluecrop’, ‘Legacy’, ‘Ozarkblue’, and ‘Tifblue’ were evaluated at Day 0 by controlled freezing in a glycol bath at temperatures from -1EC to -28EC, followed by visual evaluation after a 24h incubation at 23 EC. Similar shoots were deacclimated at a constant temperature of 20 EC and a new batch was evaluated daily for 6 days. Cultivars with any amount of southern germplasm (V. ashei or V. darrowi) were less hardy (LT50 = 20 to -21 EC) than northern highbush cultivars (LT50 = -24 EC) which are composed primarily of V. corymbosum with small percentages of V. angustifolium. Cultivars with greater amounts of southern germplasm (‘Legacy’, ‘Ozarkblue’, and ‘Tifblue’) started at less hardy levels, and deacclimated to a slightly less hardy level (LT50 = -12 to -14 EC) than did northern-adapted cultivars (‘Weymouth’ and ‘Bluecrop’) (LT50 = -15 EC). By Day 6, deacclimation appeared to plateau for all cultivars. Cultivar deacclimation was modeled using a log-linear regression model. Northern and southern cultivars differed in their regression parameters. ‘Ozarkblue’, an extremely late-flowering cultivar, would seem to be adaptable to northern climates, yet the data from this study suggest bud swell and flowering time may be poor measures of rates of deacclimation. Deacclimation under fluctuating field conditions is currently being evaluated. INTRODUCTION The development of more cold hardy and spring frost resistant cultivars is an important need to the blueberry industry in the United States (Moore, 1993). Over the past several years, the research group at the Beltsville Fruit Laboratory has focused on studying the genetics of mid-winter cold hardiness in the hope of using what is learned to aid in developing more cold hardy cultivars. In these studies, we have identified dehydrin genes and demonstrated that levels of these proteins are correlated with cold hardiness levels (Muthalif and Rowland, 1994) and have made progress toward mapping QTLs (quantitative trait loci) controlling mid-winter cold hardiness (Rowland et al., 1999) in blueberry. More recently, we have initiated a study to address the need to develop cultivars more resistant to spring frosts. For this, we have focused on deacclimation (the loss of cold hardiness with exposure to warm temperatures), as opposed to cold acclimation. For cultivars that exhibit early spring frost damage, several possibilities exist with respect to cause: 1) the cultivar may lack sufficient mid-winter cold hardiness, 2) the cultivar may be sufficiently cold hardy mid-winter but may deacclimate too quickly in the spring, or 3) a combination of both, the cultivar may lack sufficient cold hardiness and deacclimate quickly. In an effort to better understand deacclimation in blueberry, we have measured and developed models to describe deacclimation in several blueberry cultivars with different germplasm compositions and different phenological field responses. We then examined these models to determine their predictive value. Proc. XXVI IHC – Berry Crop Breeding Eds. P. Hicklenton and J. Maas Acta Hort. 626, ISHS 2003 Publication supported by Can. Int. Dev. Agency (CIDA) 40 MATERIALS AND METHODS Detached shoots of blueberry cultivars with varying percentages of species ancestry (V. corymbosum L., V. angustifolium Ait., V. ashei Reade, V. darrowi Camp) were assayed in mid-February in 2000 and 2001 to determine bud hardiness, and to evaluate rates of deacclimation under constant temperature conditions. The germplasm composition of the cultivars used, ‘Weymouth’, ‘Bluecrop’, ‘Legacy’, ‘Ozarkblue’, and ‘Tifblue’, are listed in Table 1. Germplasm composition was determined from previously published sources (Ehlenfeldt, 1994; Hancock and Siefker, 1982; Clark, et al., 1996). Shoots of ‘Weymouth’, ‘Bluecrop’, ‘Legacy’, and ‘Ozarkblue’ were collected from a commercial farm in Hammonton, New Jersey. Shoots of ‘Tifblue’ were collected from a field plot at the Henry A. Wallace Agricultural Research Station at Beltsville, Maryland. The LT50 values of shoots of were first measured at Day 0 by controlled freezing in a glycol bath (Forma Scientific, Marietta, Ohio) as previously described by Arora et al. (2000). For the freeze-thaw test, 3 shoots with at least 3 attached buds per treatment temperature (-1EC to -28EC, at 2 EC increments) were frozen, followed by visual evaluation after a 24h incubation at 23EC to determine percent damage. In 2000, similar sets of shoots were deacclimated at a constant temperature of 20EC for increasing 3-day intervals (i.e, one set deacclimated for 3 days, the next set deacclimated for 6 days, etc) up to 15 days, then evaluated similarly. In 2001, shoots were deacclimated for increasing 1-day intervals up to 6 days. Controls in both years consisted of similarly handled shoots with no exposure to glycol bath freezing regimes. LT50 values were estimated using Proc Probit in SAS version 8.2 by fitting percent damage versus temperature to a logistic (i.e. sigmoidal) regression model. The single LT50 estimates obtained at each day from these models were subsequently used to model the relationship between LT50 and days. For each year and genotype, followed by each genotype (combining years), a log-linear model was fit to the LT50 versus day data. Data on flowering times were collected from a cultivar plot located at the Philip E. Marucci Center for Blueberry and Cranberry Research and Extension at Rutgers University, Chatsworth, New Jersey from 1999 to 2002. The plot was surveyed once a week, usually on Tuesday, from the initiation of flowering of the earliest-blooming cultivar until the end of fruiting of the latest-ripening cultivar. The values represent a composite estimate across five representative plants of each cultivar. “Start of flowering” dates represent the week-long interval during which flowering initiated. The 50% bloom time was determined by interpolation from bracketing values. RESULTS AND DISCUSSION In 2000, it was observed that deacclimation occurred quickly at 20EC, and appeared to plateau for all cultivars by day 6. Therefore, in 2001 deacclimation assessments focused on this smaller, critical, 6-day period. During this time frame, ‘Legacy’, ‘Ozarkblue’, and ‘Tifblue’ started at a less hardy level (LT50 = -20 to -21 EC) and deacclimated to a slightly less hardy level (LT50 = -12 to-14EC) than did the cultivars, ‘Weymouth’ and ‘Bluecrop’. ‘Weymouth’ and ‘Bluecrop’ started at an LT50 of -24 to 25EC and deacclimated to an LT50 of -15 EC. (Figure 1). By the end of this period all cultivars had reached a comparable plateau level. The relationship between LT50 and days of deacclimation for all years and genotypes were initially modeled individually; however, after preliminary analysis, that showed no statistically significant differences between years, years were combined and reanalyzed over the critical interval from Day 0 to Day 6. Data from 0, 3, and 6 days from 2000 were combined with data from 0-6 days from 2001 and a single log-linear regression model developed for each genotype (Figure 2). ‘Legacy’ and ‘Ozarkblue’ could be described by identical models. ‘Bluecrop’ and ‘Weymouth’ models were similar to each other, but not identical, and the ‘Tifblue’ model differed from all of the others (Table 2). In a comparison of initial LT50 values (predicted from the models), significant differences were observed between the two groups of cultivars, one consisting of ‘Weymouth’ and ‘Bluecrop’, and the other consisting of ‘Ozarkblue’, ‘Legacy’, and