Summary

The effect of simulated hail on Chardonnay grape production was evaluated from 1999 to 2002 at the Western Colorado Research Center near Grand Junction, CO. In the first year, the applicability of a hail device supplied by the National Crop Insurance Services to inflict "hail" damage was evaluated. Natural hail damage could be successfully replicated. Damage severity could be manipulated by varying the distance of the hail device from the treated vine. The closer the hail device, the greater the damage. Distances of 10 to 15 feet caused damages similar to that experienced with natural hail events. Damage due to early (pre-veraison) hail treatments did not differ significantly from those caused by late (postveraison) hail treatments in two out of three years. Differences in 2002 were mainly due to a lack of berry shriveling in the late hail treatment. Simulated hail damaged clusters only on the side of hail application. Hail did not penetrate through the canopy and did not damage clusters on the other side. Hail-affected bunches sustained close to 50 % berry damage. Theoretical yield reductions from hail are around 25 %. This is based on the assumption that clusters are evenly distributed on a vine and a 50 % berry damage on the hail-affected side. Actual yield reductions caused by hail were in the order of 6 to 30 %. Yield reductions were due to desiccation of damaged berries, as well as from berries, cluster parts and whole clusters being knocked off by hail. Damaged, desiccated berries accounted for 10-15 % of the total yield. As those shriveled, dried berries are of zero value from a winemaking perspective, juice yield from a load of hail-affected grapes at the press can be expected to be 10-15 % less than from a load of grapes without such damage. The yield reduction due to hail thus is greater than that predicted simply from harvest data. Under climatic conditions that favor bunch rots such as Botrytis cinerea, actual yield losses could be much higher than those found in the dry climate of Colorado.

Introduction

Hail storms are an annually recurring event in Western Colorado. These storms generally have a brief duration and are rather localized. Nevertheless, even short hail events can cause significant damage to many fruit crops due to the market demand for blemish-free fruit. For example, small skin ruptures and blemishes on fresh-market peaches or apples will reduce the market quality, and often the price received by the grower. The impact of hail on wine grapes is somewhat less than for fresh-market fruit such as peaches and apples because a) appearance is less important in a processed crop such as wine grapes; b) trellis systems with upward shoot positioning appear to provide more protection to the bunches from leaves above compared to the open tree canopies used for tree crops; and c) even when some berries on a bunch get damaged, the remaining undamaged berries retain full market value.

The severity of hail damage depends on many factors. First and foremost, damage depends on the intensity and duration of the hail event. Sustained events that produce only small hail pellets may be as damaging as short, intense events with big hailstones. Second, damage from similar hail events may differ due to the stage of crop development. For example, a hail event during tree or vine dormancy is likely to cause no or little damage while a similar event close to harvest may result in significant crop losses. And third, damage depends on the type of crop and its intended end use (e.g. fresh market or processed). Growers use different approaches to manage the risk of crop losses due to hail. Some orchardists in Western Colorado use overhead hail nets to protect their tree fruit. Those nets can also be used as shade nets to reduce sunburn of fruit. Other growers purchase hail insurance for parts or all of their crops, while some growers use neither approach. We started a series of experiments in 1999 to provide the National Crop Insurance Services with scientifically-based data on hail damage for wine grapes. The project objective was to quantify the effect of hail damage at different times of the season on wine grape production.

Material and Methods

Experimental setup

The experimental setup differed between years. From 1999 to 2001 we used own-rooted Chardonnay vines planted in 1991 at a spacing of 5' x 10'. Due to significant vine damage in the original block during the winter 2001/02, we used own-rooted Chardonnay vines planted in 1984 at a spacing of 8' x 12' in the final year. The row orientation was North-South, and all vines were trained to a bilateral cordon and spurpruned. From 1999-2001, we used 2-vine plots, while in 2002 treatments were applied to singlevine plots. There were five replications per treatment in 1999 and 2000, six in 2001, and seven in 2002.

Treatments

In 1999, hail was applied on 10 August (seven weeks prior to harvest) using a hail device supplied by National Crop Insurance Services. There were three hail treatments; hail was applied from a distance of 5', 10' or 15' to the vine row.

In 2000 and 2001, treatments were arranged in a split-plot design with time (Early or Late) as the main plot, and distance [Close (10') or Far (15')] as sub-plot. In 2002, hail was applied either early or late from a distance of 12' from the vine row. In all years, Early and Late treatments were applied approximately three weeks prior to or after veraison, respectively. In each year, damage caused by hail was compared to an untreated Control. A treatment summary is provided in Table 1.

Hail application

In 1999, the hail device was positioned on a pallet on the back of a tractor so that the device's delivery tube went through a 90º bend. The device was positioned at 5', 10' or 15' distances from the vines. Only vines on the outside row could be used, as the vine row spacing is 10' leaving insufficient room to operate the device inside the block. In the final three years, the hail device was positioned on a pallet at the front of a tractor located in the inter-row adjacent to the treatment rows. The machine was lifted to a height of about 6'6" so that the hail was delivered to the treatment vines at about 50-70º angles from above. As the delivery tube was flexible, the distance from the tube outlet to the treatment row could be altered.

From 1999 to 2001, 40 lb of solid tubed ice was applied to each 2-vine plot. In the 2002 season, 20 lb of solid tubed ice was applied to each single-vine plot. The operator applied the ice in approximately six sweeps of the vines taking about ten seconds per plot. In all years, hail was applied to the western side of the canopy, and plastic sheeting was used to protect adjacent vines both within the row and in adjacent rows.

Damage evaluation

In 1999 and 2000, damage caused by the hail application was evaluated at harvest only. Total cluster number and total cluster weight was determined separately for each plot. In 2001 and 2002, damage was evaluated two weeks after the early hail application and at harvest. For the early evaluation, six clusters per plot were harvested at random from the treated (west) side of the Early vines. Likewise, six clusters per plot were collected at random from the western side of three Control plots. Clusters were taken back to the laboratory to determine total cluster weight. Berries were then cut off and classified as either damaged or undamaged, and total number and total weight was recorded separately for each class.

At commercial harvest (24 September, 2001 and 9 September 2002), six clusters per plot were harvested at random from the Late treatment plots on the treated (west) side of the vines. Likewise, six clusters per plot were collected at random from the western side of three (2001) or four (2002) Control plots not used for the early evaluation (see above). These clusters were evaluated as described previously.

Statistical analysis

All data were analyzed by the general linear model procedure (SAS Institute, Cary, N.C.). Significance was determined at the p>0.05% level.

Results

1999 season

The severity of damage differed between the three hail treatments. Very severe damage was inflicted with a spacing of 5' from the row: berry skins were ruptured, leaves were severely shredded to the point of defoliation, and some damage to wood occurred. With 10' spacing, damage was severe with some berry skin rupture and shredding of leaves. At the furthest distance of 15', damage was classified as moderate with minimal rupture of berry skins and holes in the leaves. In all cases, damage only occurred to the side of the vine that was being treated. While the "hail" generated from the machine may have been larger than natural hail, the damage, especially with 10' and 15' distance, appeared similar to that experienced in nature.

As expected, the closer the distance the hail machine was positioned to the vines the greater the yield decrease (Table 2). However no statistical significance was found between any of the treatments.

2000 season

Two changes were made compared to the 1999 season. First, the damage observed with 5' spacing in 1999 was found to be too severe and this treatment was not repeated. Second, a platform was built to secure the hail device so that the device could be lifted above the vine rows using a tractor-mounted forklift. Thus, hail could be applied to the treatment vines from above at an angle of about 50-70º.

Although yields were reduced by hail applications before or after veraison compared to the Control, the differences were not significant (Table 3). Further, there was no significant effect of timing (Early or Late) or spacing (Close or Far).

2000 season

Two changes were made compared to the 1999 season. First, the damage observed with 5' spacing in 1999 was found to be too severe and this treatment was not repeated. Second, a platform was built to secure the hail device so that the device could be lifted above the vine rows using a tractor-mounted forklift. Thus, hail could be applied to the treatment vines from above at an angle of about 50-70º.

Although yields were reduced by hail applications before or after veraison compared to the Control, the differences were not significant (Table 3). Further, there was no significant effect of timing (Early or Late) or spacing (Close or Far).

The hail application significantly reduced the overall mean berry weight from 0.58 g for the control to 0.43 g and 0.48 g for EC and EF. However, the mean weight of undamaged berries was very similar between treatments (Fig. 2).

Late damage evaluation

Application of hail approximately four weeks after veraison resulted in a larger reduction of cluster weight than a similar application prior to veraison. Compared to the control, hail reduced the weight of clusters sampled at commercial harvest from the hailaffected side of the late-treatment vines by about
40 % and 30 % in the close and far treatment, respectively (Table 5; Fig. 3). Unlike the early evaluation, the differences in cluster weights were highly significant (P=0.0023).

About 36 % and 23 % of the berries were damaged in Late Close (LC) and Late Far (LF) clusters, respectively, and accounted for 15 % and 10 % of total cluster weight. It should be noted that we also found about 2 % damaged berries accounting for 1 % of cluster weight in control clusters. This damage is likely due to physical injury from wind or shoot rubbing, insect damage, or damage caused during canopy management or harvesting operations.

The mean berry weight of 0.99 g in the control treatment was significantly higher than the 0.80 g and 0.90 g for LC and LF. Again, the mean berry weight of undamaged berries did not differ between treatments, but damaged berries in the control were heavier than those of LC and LF (Table 5; Fig. 4).

Effect on yield

Yield reductions due to hail ranged from about 12% in the LF regime to 22% in LC (Table 6), but the differences were not significant. The number of clusters per vine was slightly less in the early treatments. The control had significantly more clusters with no damage than any of the hail treatments. There was no
difference in the number of clusters with or without damage between the hail treatments. Compared to the control, the mean cluster weight was reduced by about 10 % by the early and about 20 % by the late hail treatments.

2002 season

Application of hail approximately four weeks prior to veraison reduced the weight of clusters sampled two weeks later from the hailaffected side of the vine by about 53 % (Table 7; Fig. 4). Damaged berries accounted for 18% of total berry weight in the early treatment.

The hail application significantly reduced the overall mean berry weight from 0.46 g for the control to 0.24 g for the early treatment, representing a 49 % decrease. In contrast to previous years, the mean weight of undamaged berries was significantly lower in the early treatment than the control (Fig. 5). The mean weight of undamaged berries in the Early treatment was about 17 % lower than in Control.

Late damage evaluation

In contrast to the previous year, a hail application approximately four weeks after veraison caused less reduction of cluster weight than a similar application prior to veraison. Compared to the control, hail reduced the weight of clusters sampled at commercial harvest from the hail-affected side of the late treatment vines by about 25 %, but differences in cluster weight were not significant (Table 8; Fig. 6).

About 47 % of the berries were damaged in late treatment clusters, and accounted for 35 % of total cluster weight. It should be noted that we also found about 11 % damaged berries accounting for 12 % of cluster weight in control clusters. This damage is likely due to physical injury from wind or shoot rubbing, insect damage, or damage caused during canopy management or harvesting operations.