Assessment of Winter Injury in Grape Cultivars and Pruning Strategies Following a Freezing Stress Event

Bud injury assessment.

Extreme subfreezing temperatures occurred in January 2009 throughout the grapegrowing regions east of the Rocky Mountains, including Illinois, Indiana, Iowa, Minnesota, and Missouri. In Ohio, temperatures dropped below −22°C on 15–17 Jan and reached −31°C (Figure 1). These temperatures were the lowest recorded since 1994 (Bordelon et al. 1997). Temperatures recorded at the OSU weather stations ranged from −18 to −28°C in northeastern, −22 to −28°C in northwestern, −22 to −26°C in central, and −16 to −20°C in southern Ohio. At the two research vineyards, temperatures dipped to −26°C in Wooster and −24°C in Kingsville (Figure 1, Figure 2). In northeast Ohio, along the Lake Erie shores where the majority of commercial V. vinifera cultivars are grown, temperatures ranged from −23 to −29°C. These extreme low temperatures are considered critical for grapevine productivity and survival (Zabadal et al. 2007).

As expected and due to the colder temperature, the same cultivars in both vineyards sustained more severe bud injury in Wooster than in Kingsville (Figure 3). For example, Cabernet franc, Chardonnay, Pinot gris, and Traminette recorded 98, 93, 87, and 40% bud injury in Wooster and 54, 25, 9, and 11% bud injury in Kingsville, respectively (Figure 3). These differences in bud injury were attributed primarily to variation in minimum temperatures in both locations. Differences in primary bud injury associated with different temperatures in different locations were also reported in other regions (Bordelon et al. 1997, Wolfe 2001). A difference of two degrees Celsius between Wooster (colder site) and Kingsville resulted in a two- to ten-fold increase in bud injury for Cabernet franc and Pinot gris, respectively. In both sites, V. vinifera cultivars sustained the most injury followed by old and new hybrids; Vitis labrusca Concord sustained the least injury. In each site, cultivars and selections exhibited substantial variation in primary bud injury (Figure 3). In Wooster (−26°C), V. vinifera cultivars averaged 93% bud injury. Among the old hybrid cultivars, Chambourcin sustained the most injury (93%), which was similar to V. vinifera cultivars, followed by Vidal and Seyval. Newly developed hybrid cultivars and advanced selections from the grape breeding programs at Cornell University and University of Minnesota sustained bud injury of 12 to 35% and 8 to 14%, respectively. In Kingsville (−24°C), similar trends were observed, with V. vinifera cultivars being the most cold-sensitive followed by the new hybrid Traminette and the most cold-hardy V. labrusca Concord. We also observed a better separation of bud injury among V. vinifera cultivars. Another observation was the unexpected performance of V. vinifera with an average bud injury of only 42% after exposure to −24°C.

• Download figure
• Open in new tab

Figure 3

Primary bud injury of winegrape cultivars and selections grown at the OSU research vineyards in Wooster and Kingsville, OH, following freezing events in January 2009.

In addition, the survey data collected from commercial vineyards in the same region showed similar patterns of bud injury among species and cultivars (data not shown). Furthermore, cold hardiness of several genotypes based on observations from natural freezing events is in agreement with that from simulated freezing tests using thermal analysis (Wolf and Cook 1994). According to Zabadal et al. (2007), critical temperatures, measured as LT50, are usually −18 to −22°C for the tender cultivars (Chardonnay, Cabernet Sauvignon, Pinot gris, Pinot noir), −21 to −23°C for the moderately tender (Cabernet franc, Riesling), −23 to −26°C for the moderately hardy (Chardonel, Traminette), −26 to −29°C for the hardy V. labrusca (Concord), and −29 to −34°C for the very hardy new hybrids (Frontenac and LaCrescent). In contrast to trunk loss and vine dieback of most V. vinifera cultivars after exposure to −26°C in Washington (Wolfe 2001), minimal trunk damage or dieback was observed in the Wooster vineyard after exposure to the same −26°C. This better than expected freezing tolerance is attributed to the timing of the extreme minimum temperature which occurred at an acclimation stage when vines were at their maximum cold hardiness in mid-January (Figure 2). Furthermore, minimum temperatures were below freezing for many days (16 days below 0°C, 4 days below −5°C, and 11 days below −10°C in Wooster) prior to the extreme cold event. This has been shown to increase cold hardiness of vines when they were preconditioned with subfreezing temperature (Quamme 1986, Wolf and Pool 1987).

Finally, grapevines in both locations had ideal cold acclimation conditions during the previous fall of 2008 (Dami 2009). Pinot noir and Pinot gris performed the best among 10 V. vinifera cultivars grown in Kingsville, which may be attributed to the early to midseason ripening nature of these cultivars, whereby the postharvest leaves continued to produce photosynthates that may have contributed to an increase in the reserves needed for full cold acclimation prior to the killing frost. That may not be the case with late-ripening cultivars such as Cabernet franc, in which postharvest photosynthates make little contribution due to the short period between harvest and the killing frost. Similar observations were made in Michigan (Howell 2001).

Pruning adjustment.

In spring 2009, the no pruning treatment visually resulted in the earliest and most vigorous growth, whereas spur pruning had the latest and weakest, and hedge pruning treatments were intermediate (Dami and Ennahli 2010). In July and August 2009, we also observed the so-called midsummer vine collapse. Vine collapse occurred preveraison through postveraison and consisted of leaf wilting and shoot collapse in a portion of the cordon (partial collapse) or whole vine (total collapse) (Dami and Ennahli 2010). Such collapse is an indicator of vascular damage including phloem, xylem, and even cambium tissues (Goffinet 2001, Zabadal et al. 2007). Since precipitation was abundant (625 mm) during the growing season (growing degree days, base 10°C, = 1424) and adjacent and noninjured cultivars in the vineyard block did not collapse, it was concluded that shoot collapse was not a result of drought, but rather a freezing stress induced-injury in the vascular system of grapevines. The assessment of cordon, trunk, and whole vine injuries confirmed vascular damage (Table 1). Although nonpruned vines sustained the most injury, there were no statistical differences among treatments. This lack of statistical significance was attributed to the large variability of winter injury among treatment replications as a result of the sporadic nature of injury distribution and severity in the vineyard block. This random pattern of winter injury incidence was also observed in other regions in Ohio (Dami, author’s personal observations, 2009). Midsummer collapse due to vascular damage has been previously observed in other grape cultivars and regions (Zabadal et al. 2007). The loss of vascular function typically begins in the phloem tissues (outer to inner) and progresses to the xylem tissues (inner to outer). Phloem injury prevents carbohydrate movement while xylem injury leads to filling of vessels with gums, which block water and nutrient flow resulting in shoot wilting and collapse (Goffinet 2001, Zabadal et al. 2007). Our findings concur with those of Keller and Mills (2007), who concluded no effect of pruning on vine survival in cold-injured Merlot vines grown in Washington.

At harvest and as expected, nonpruned vines had the highest number of clusters and yield per vine, and spur-pruned and 2-node hedge pruned vines had the lowest; cluster number and yield in 5-node hedging were intermediate (Table 2). In other words, the higher the pruning severity, the lower the yield. These findings concur with previous reports from Washington where pruning severity after winter injury dictated yield (Keller and Mills 2007, Wolfe 2001). Since bud and vascular injuries were not different among treatments, it is suggested that the yield response is attributed to pruning severity rather than to cold injury.

As predicted, it took the least time to prune and remove wood from the trellis with the spur-pruning treatment and the most time with the nonpruned and 5-node hedged treatments (Table 2). Vine size, expressed as pruning weight of 1-year-old wood, was smallest in nonpruned vines and largest in the remaining treatments. Furthermore, most of the 2-year-old wood was removed from nonpruned vines and the weight decreased as pruning severity increased. However, the weight of total wood (1- and 2-year-old) removed during pruning was not different among all treatments. When expressed per linear cordon length, pruning weights of 1-year-old wood ranged from 0.22 to 0.49 kg/m. Smart and Robinson (1991) have suggested that a balanced vine has an optimal pruning weight of 0.3 to 0.6 kg/m of row length. Based on that premise, all pruned vines were within the optimal range except for nonpruned vines, which were below the optimal range and thus were considered out of balance.

The Ravaz index (RI) was also determined based on 1-year-old wood. Nonpruned vines had the highest RI and spur-pruned and 2-node hedged vines had the lowest (Table 2). It has been suggested that an RI value between 4 and 10 was ideal and indicated balanced vines, while a value outside that range was out of balance, with values >10 indicating overcropping and values <4 indicating undercropping or overvigor (Kliewer and Dokoozlian 2005, Smart and Robinson 1991). Based on those recommendations, it was concluded that nonpruned vines were overcropped, spur-pruned and 2-node hedged vines were undercropped, and 5-node hedged vines were balanced, as the RI was within the ideal range.

In year 2, the goal was to reestablish spurs in the vicinity of the cordons. Therefore, pruning involved removal of 1-year-old canes as well as misplaced 2-year-old wood. It was not always possible to have 2-node spurs spaced evenly on the cordons since some were dead. By focusing on bringing spurs as close to the cordon as possible, some treatments resulted in uneven numbers of spurs thus uneven numbers of buds per vine. As a result, count buds were the lowest in nonpruned vines and the highest in the 2-node hedged vines (Table 3). Total shoots per vine followed the same trend as count buds and were directly influenced by count shoots, as noncount shoot number was the same among all treatments. Total cluster count per vine was also influenced by the total shoots per vine and was lowest in nonpruned vines and highest in spur-pruned and 2-node hedged vines. However, bud fruitfulness in 2010 was not affected by pruning treatments applied in 2009. Therefore, it is suggested that there is no carryover effect of pruning type after winter injury that might negatively affect fruitfulness in the following season. Similar conclusions have been reported in Washington (Keller and Mills 2007).