Shifts in Bud and Leaf Hardiness during Spring Growth and Development of the Cranberry Upright: Regrowth Potential as an Indicator of Hardiness

‘Stevens’ cranberry (Vaccinium macrocarpon Ait.) terminal bud freezing stress resistance was assessed by nonlinear regression utilizing relative scoring of the post-thaw bud growth and development based on defi ned bud stages 2 weeks following controlled freezing tests. Bud stages tested were chosen based on a phenology profi le from each sampling date throughout the spring season. Previous year (overwintering) leaf freezing stress resistance was evaluated after both 2 days (injury) and 2 weeks (survival). The Gompertz function with a bootstrapping method was used to estimate the tissues’ relative freezing stress resistance as the LT50. Bud injury levels (LT50) were expressed as the temperatures at which the mean potential regrowth capability was impaired by 50%, as compared with the unfrozen controls. In leaves, the LT50 is the temperature at which 50% injury (2-day evaluation) or survival (2-week evaluation) was modeled to occur. Dramatic changes in terminal bud relative freezing stress resistance occurred both within and between the tight and swollen bud stages. These results clearly show that seasonal changes in freezing stress resistance do not necessarily parallel changes in crop phenology and bud development. These results indicate that some physiological, biochemical, or fi ne anatomical changes may explain the seasonal loss in hardiness within a visual bud stage. Previous year leaves may possess the ability to recover from freeze-induced injury, as leaf survival was found to be the most reliable indicator of cranberry leaf hardiness. Major shifts in phenology and bud and leaf hardiness coincided with the rise of minimum canopy-level air temperatures to above freezing. The nonlinear regression technique utilized made it possible to estimate LT50 with data points comprising half or more of the sigmoidal dose response curve. Our study provides precise and quantitative estimates of the cold hardiness changes in cranberry terminal buds and leaves in spring. From precise estimates we were able to defi ne critical temperatures for the impairment of cranberry bud growth. This is the fi rst systematic study of cranberry terminal bud cold hardiness and spring bud development in relation to changes in the soil and air temperatures under natural conditions. Our study shows that regrowth assessment of the cranberry upright inherently describes the composite effects of freezing stress on plant health. Due to their low elevations, cranberry growing areas are often subject to frosts, events that are possible even in summer (Dana, 1990). Sprinkle irrigation is the main frost protection method used by Wisconsin growers. To protect the vines through the winter, cultivated cranberries are fl ooded to form a protective layer of ice around and above the vines. Bud hardiness is lost quickly in spring when development and growth resumes (Abdallah and Palta, 1989). Classifi cation schemes for stages of bud development and corresponding hardiness levels have been developed for apple (Malus xdomestica Borkh.), pear (Pyrus communis L.), sweet cherry (Prunus avium L.), peach (Prunus persica Batsch), prune (Prunus domestica L.), and apricot (Prunus armeniaca L.) (Proebsting and Mills, 1978), grape (Vitis vinifera L.) (Johnson and Howell, 1981a, 1981b), and rabbiteye blueberry (Vaccinium ashei Reade) (Spiers, 1978). Attempts to relate freezing stress resistance to spring developmental stages have been reported for cranberries grown in Massachusetts (DeMoranville, 1998; DeMoranville and DeMoranville, 1997; Roberts and DeMoranville, 1985) using potted sods collected from commercial beds. Stages of bud development identifi ed in these reports ranged Received for publication 30 June 2005. Accepted for publication 21 Feb. 2006. We thank Bjorn Karlsson for fi eld assistance and Peter Crump for writing the SAS programs. Teryl Roper provided the original drawing that was adapted for Fig. 1. This study was supported by a grant from Wisconsin Cranberry Board and by the College of Agriculture and Life Sciences, Univ. of Wisconsin–Madison. 1To whom reprint requests should be addressed. E-mail address: jppalta@wisc. edu from spring dormant buds to bloom, and shifts in hardiness were estimated to be from about –9 °C to about –1 °C, respectively. However, preliminary studies by Abdallah and Palta (1989) estimated much greater changes in freezing stress resistance using controlled freezing tests. Beside these general estimates, there is little published work on freezing stress resistance of the cranberry plant. Furthermore, there is little agreement on the hardiness of the cranberry plant, even in its dormant state (Doughty and Shawa, 1966; Eck, 1990; Roberts and DeMoranville, 1985). Due to the supercooling of various tissues, ice nucleation is necessary in freezing stress experiments to obtain precise estimates of freezing stress resistance; yet this has not been used in most cranberry studies (DeMoranville and DeMoranville, 1997; Eaton and Mahrt, 1977; Reader, 1979). The main objective of our study was to defi ne critical temperatures for the impairment of cranberry bud growth. A unique aspect of our study is the assessment of bud freezing stress resistance by nonlinear regression utilizing relative scoring of bud growth and development based on defi ned bud stages. Bud development and budbreak can be impaired by freezing temperatures (Spiers, 1978). Since multiple bud stages can be present in the fi eld at a given time, we assessed the effect of freezing temperatures for the leading bud stages at each test sample date. Plant tissue response to freezing stress is characterized by an “S-shape,” or sigmoidal, curve bounded by a lower and an upper asymptote (Zhu and Liu, 1987). In plant cold hardiness studies, the term LT50 has been used to denote the temperature at which 50% of the samples or plants were killed, 50% of the tissue was Book 1.indb 327 5/1/06 11:21:47 AM 328 J. AMER. SOC. HORT. SCI. 131(3):327–337. 2006. damaged, or 50% of the ions or other cell contents leaked from the cells. The LT50 injury level has been interpreted as the critical temperature for tissue recovery (Sakai and Larcher, 1987), although it has also been shown that 50% ion leakage does not necessarily relate to 50% cell death (Palta et al., 1977a, 1977b). Other arbitrary levels of injury, such as LT10, LT20, etc., have also been applied to freezing stress resistance data to indicate severity of damage. To be meaningful, these critical injury levels must be related to empirical knowledge regarding the impairment of cellular function or survivability of the tissue or plant (Luoranen et al., 2004; Stergios and Howell, 1973). Such methodology has not been employed for the study of cranberry plant freezing stress resistance. Empirical mathematical models have been used recently to characterize the sigmoidal plant response to environmental stress. Although straight line techniques have been used (Holt and Pellet, 1981), computer simulations using sigmoid functions have proven to be more applicable (Anisko and Lindstrom, 1995; Jánacek and Prásil, 1991; Lim et al., 1998; Luoranen et al., 2004; von Fircks and Verwijst, 1993; Zhu and Liu, 1987). In addition to the interpolation of various injury levels (e.g., LT10 or LT50), the infl ection point of the curve indicates the temperature at which the greatest rate of injury occurred (Tmax). Asymmetric functions can often better describe biological responses to stress than the logistic function since such data are often skewed in their distribution (von Fircks and Verwijst, 1993). The Richards function has been used to model the effect of plant freezing stress with both electrolyte leakage data (Anisko and Lindstrom, 1995; von Fircks and Verwijst, 1993) and shoot mortality data (von Fircks and Verwijst, 1993). Although the results of these studies produced seemingly meaningful results, use of the Richards function has been criticized on the grounds that it requires the estimation of an additional parameter and that it has high intrinsic nonlinearity, resulting in the parameter estimates to be severely biased (Ratkowsky, 1990). Lim et al. (1998) compared the suitability of the Richards and the Gompertz functions for use in evaluating freeze injury electrolyte leakage data. Although there was no signifi cant difference between the performances of the two functions, these authors chose the Gompertz function to estimate the LT50 and Tmax values because it had fewer parameters to estimate and that it fi t relatively better when compared with the general linear model. Our study is an attempt to evaluate the freezing stress resistance of the cranberry upright by quantifying the competence of terminal buds following a freeze-thaw stress. We used the Gompertz function with a bootstrapping method to evaluate the critical temperatures related to various levels of freezing stress damage. From precise measurements we attempted to determine systematic changes in cranberry bud hardiness in relation to bud development and changes in the soil and air temperatures. Changes in hardiness between and within bud stages were estimated. Materials and Methods PLANT MATERIAL. Samples of ‘Stevens’ uprights were collected from a bed in central Wisconsin throughout the spring seasons of 1997 and 1998. Preliminary fi eldwork was conducted in 1996. Each year the sampling period began when the ice cover was fully melted and beds could be accessed. This date was called the “ice-off” date. The sampling period continued from the ice-off date until the beginning of bloom. Samples were collected on 21 dates in 1997, and on 15 dates in 1998. On each sample date, cuttings were taken from nine random locations in the bed. For each cutting location, all of the uprights from an approximately 50 × 13-cm strip were cut (≈350 to 400 uprights). Samples were kept on ice for transport to Madison, Wis. Unless otherwise noted, hardiness values and patterns described are from the 1997 results, including 1998 data where relevant. Uprights were sorted according to previously defi ned terminal bud stages (Workmaster et al., 1997) (Fig. 1). The stage of “earlyhook” was added at the time of sorting. Controlled freezing tests were performed on uprights of the most advanced bud stages that each constituted about 10% or more of the bud stages present (Tables 1 and 2). Bud stage categories were tested from all 21 sampling dates in 1997, and from fi ve of the 15 sampling dates in 1998 (6 and 27 Apr.; 5, 7, and 14 May). B ud b re ak B ud e lo ng at io n