Abstract
Thirteen Chardonnay (Vitis vinifera L.) clones were evaluated for viticulture performance in a vineyard managed for sparkling wine production in the Los Carneros American Viticultural Area, Sonoma County, California. Two clones were of California origin (Foundation Plant Services [FPS] 4 and 72) and 11 were of French origin, reported to be from either Dijon (clones 75, 76, 78, 96, and 352) or Champagne (clones 118, 121, 124, 130, 131, and 132). Vine yield and yield components, fruit composition (soluble solids, pH, and titratable acidity), and vegetative growth were measured over four growing seasons, from 1998 through 2001. Clones were harvested on a soluble solids basis. Yield, berries per cluster, berry weight, cluster weight, and clusters per shoot differed significantly among clones. FPS 4 and French clones 96 and 352 had highest yields, and FPS 72 the lowest, with no great differences among other clones. Low yield of FPS 72 was due to very low berry weight and fewer berries per cluster. Yield components of other clones differed significantly over the four years of study. FPS 4 and FPS 72 retained the highest levels of titratable acidity at harvest. Vegetative growth was greatly influenced by both clone and year. FPS 72 had the highest pruning weights and clone 118 the lowest. Yield-to-pruning weight ratios were lowest for FPS 72 and highest for FPS 4.
Chardonnay (Vitis vinifera L.) is a highly regarded white wine cultivar, producing high-quality table wines, sparkling wines, and dessert-style wines. Chardonnay is well adapted to a range of acceptable environments and consequently is found throughout winegrowing regions in both the Old and New Worlds (Bernard 1986). It is widely planted in France, the United States, Italy, Australia, New Zealand, and Chile.
The earliest Chardonnay vineyards in California were planted in the late 19th century, but the cultivar did not receive much attention until the 1970s and 1980s when total Chardonnay plantings saw dramatic growth. Today, Chardonnay is the most widely planted winegrape in California, with 40,000 hectares under cultivation (CASS 2006). One reason for the increase in planted area has been the success of clonal selection in increasing yields while maintaining desired quality (Wolpert 1995).
As would be expected for such an important cultivar, Chardonnay clones have been studied internationally. Clonal evaluations of Chardonnay have been published in France (Bernard 1986), Italy (Battistutta et al. 1995), Australia (Cirami 1993), Canada (Reynolds et al. 2004), and South Africa (Visser 1995). Similar studies of Chardonnay clones have been conducted in the United States (Wolpert et al. 1994, Goldschmidt and Kenworthy 1995, Bettiga 2003), with varying degrees of heterogeneity found among clones. One study of Chardonnay clones in Australia found multiple differences among yield components of eight clones (Cirami 1993). However, yield did not differ among Chardonnay clones planted at three sites in Italy, although vegetative growth did (Battistutta et al. 1995).
Interpretation of clonal trials must consider whether grapes will be used for different wine styles and production goals. A clone that is ideal for production of highquality still wine may not be ideal for sparkling wine production. For this reason, in France, clonal selection of Chardonnay in Champagne has been conducted independently from clonal research on Chardonnay in Burgundy (Barillere et al. 1995). Sparkling wine producers typically desire clones with higher acidity and accept higher yield than their still wine counterparts (Bernard 1986). Additionally, still wine producers, in regions such as Mâcon, are tolerant of clones with Muscat-like flavor, a characteristic not desirable in sparkling wine (Bernard 1995). This trial is the first in the United States to evaluate the viticultural characteristics of Chardonnay clones grown for the production of sparkling wine.
Materials and Methods
Thirteen Chardonnay clones were evaluated for viticultural performance (Table 1⇓). Clones FPS 4 and 72 were widely planted in California. Clones 75, 76, 78, 96, commonly referred to as the “Dijon” clones, as well as clone 352 were largely planted for table wine production in France. Clones 118, 121, 124, 130, 131, and 132 were originally selected for sparkling wine production in Champagne, France, and were imported from Eperney by Gloria Ferrer Champagne Caves (Sonoma, CA). Authenticated French winegrape clones became available later to U.S. growers through ENTAV International, a program that distributes ENTAV-INRA selections. French clones used in this study arrived in California before the establishment of ENTAV and therefore carry the designation “reported to be” as a way to distinguish them from clones that arrived through the later trademarked program (Nelson-Kluk 2003). All clones tested in this experiment are currently available from Foundation Plant Services (FPS), UC Davis.
Dormant 5C rootstocks (Vitis berlandieri x V. riparia) were planted in spring 1993 at the Gloria Ferrer Champagne Caves within the Los Carneros American Viticultural Area (AVA). Rootstocks were field-budded in fall 1993. The soil was a Diablo clay loam (Miller 1972). Vines were planted to a 1.5 m x 3.0 m spacing (vine x row), equivalent to a vine density of 2,222 vines/ha. Row orientation was east-west. Vines were head-trained, canepruned, and trained to a vertical shoot-positioned trellis system. The trellis consisted of a fruiting wire at 90 cm and two pairs of movable shoot-positioning wires above the fruiting wire.
Mature vines were pruned to two 12-bud canes with four 2-bud renewal spurs, and vertically shoot-positioned. Shoot number varied from 24 to 26 shoots per vine and was controlled by spring shoot thinning. Applied water was delivered by drip irrigation. Cultural practices were typical for Chardonnay sparkling wine vineyards in the Los Carneros AVA.
Data collection began in 1998 when vines were six years old and continued through the 2001 growing season. Clones were harvested on a maturity basis with a target of 21 ± 0.5 Brix. Clusters were counted and crop weights measured individually for all data vines. Prior to harvest, a 100-berry sample was taken per treatment replicate. Samples were processed in the laboratory. After weighing, each sample was crushed and juice was filtered through cheesecloth. Soluble solids were measured as Brix with a hand-held, temperature-compensating refractometer. Juice pH was measured with an electronic pH meter, and 5.0-mL aliquots of juice were titrated with 0.1 N NaOH to a pH 8.2 end point to determine titratable acidity (TA).
During the dormant season, shoots on each vine were counted and dormant prunings weighed, from which average shoot weight was calculated. The Ravaz index (RI: yield:pruning weight ratio) was calculated (Ravaz 1911).
The experiment was designed as a randomized complete block, consisting of 10-vine clone plots randomized within each of five blocks. The trial was established as a separate vineyard block within the estate vineyard, and guard vines were used to separate the data vines from edge effects. For statistical purposes, the experiment was treated as a split-plot design, with clones as main plots and years as subplots. Data were analyzed by analysis of variance (ANOVA) with mean separation by Duncan’s multiple range test.
Results
Yield and yield components.
Over the four-year period, FPS 4 and clones 352 and 96 had the highest average yields, while FPS 72 yielded far less than all other clones (Table 2⇓). The 10 remaining clones showed a relatively narrow range in yield (5.6 to 6.2 kg/vine).
FPS 4 had the most berries per cluster, followed by clones 96 and 75 (Table 2⇑). FPS 72 averaged the fewest berries per cluster, while the remaining clones differed little, averaging between 86 and 91 berries.
Berry weight among 12 of 13 clones differed by 12% (1.37 to 1.54 g), the exception being FPS 72, which averaged less than 1 g per berry. The year-to-year difference in average berry size in the experiment, averaged over all clones, was large (0.26 g) especially between 1998 and the other three years. Cluster weight, as expected from berry number and weight, was highest for FPS 4 and lowest for FPS 72.
Clusters per shoot were greatest for clones 352 and 75 at 1.9, with seven clones being in a group at 1.8. FPS 4 and clone 96 had the fewest clusters per shoot with less than 1.7. Year-to-year variation was particularly high for this yield component ranging from 1.57 in 1999 to 1.95 in 2001. Differences in number of clusters per shoot (∼1.6 to 1.9) were a result of bud fruitfulness, not shoot number, because the latter was manipulated through dormant pruning and in-season shoot thinning.
Fruit composition.
Clones were harvested at similar soluble solids levels as determined by frequent sampling. With a target of 21 ± 0.5 Brix, clones were harvested on average over a four-day period, except for FPS 4 which was delayed on average by nine days (data not shown). Variation in pH was small, averaging between 3.11 and 3.18 (data not shown). There was greater variation in TA, with some clones maintaining greater acid levels than others, despite having higher soluble solids (Figure 1⇓). FPS 4 and 72 had significantly more acid than all other clones, while clones 118, 132, and 130 had the least acidity.
Vegetative growth.
FPS 72 had both greatest pruning weight and shoot weight while clone 118 had the lowest pruning weight and shoot weight (Table 3⇓). Vegetative growth decreased over the years that data were taken, from a high of 2.0 kg/vine in 1998 to 1.25 kg/vine in 2001, a 38% reduction, resulting in an increase in Ravaz index from <3 to ∼5.
Discussion
Significant differences were found in reproductive measures of Chardonnay clones for sparkling wine. Yield data (Table 2⇑) show that differences were greatest between the two American clones, high-yielding FPS 4 and lowyielding FPS 72. French clones as a group, whether from Burgundy or Champagne, were intermediate in yield and differed relatively little, ∼10% lowest to highest. Yield differences in Chardonnay clonal evaluations in Australia (Cirami 1993) and California (Wolpert et al. 1994, Bettiga 2003) were not seen in Italy (Battistutta et al. 1995); it is unknown whether the different conclusion of the latter evaluation was due to location of the experiment or to lack of genetic variability in the Italian clonal selections. However, in California, FPS 4 was found to be a relatively high-yielding clone (Bettiga 2003, Wolpert et al. 1994), a result seen again here (Table 2⇑).
In the case of FPS 72, low cluster weight, as explained by berries per cluster and berry weight, was responsible for low yields. In these three categories FPS 72 was significantly lower than all other clones evaluated. FPS 72, also known in California as the “Wente clone,” produces clusters comprised of normal-sized berries, as well as some very small berries, a characteristic known locally as “hens and chicks” or “pumpkins and peas.” This cluster morphology helps explain dramatic differences between FPS 72 and other clones. Other than FPS 4 and FPS 72, yields of other clones were closely grouped.
The number of clusters per shoot was not directly related to yield. High-yielding clone FPS 4 and French clone 96 had the lowest number of clusters per shoot, averaging 1.65 and 1.66, respectively. Conversely, clone 352 had the highest number of clusters per shoot (1.94), yet it was also among the highest-yielding clones. Although yield component differences among clones were small, they can be economically significant when multiplied over an entire vineyard.
All clones were harvested, on average, within four days of the first clone, except FPS 4, which took nine days to reach the desired level of soluble solids (data not shown). However, that is not considered critical. In the sparkling wine production areas in California’s North Coast, all clones typically reach desired maturity. Thus, the few days delay in harvesting FPS 4 is not a concern.
Sparkling winegrapes are typically harvested at lower sugar levels than still winegrapes, which was the case in this study where harvest occurred at 21 ± 0.5 Brix (Figure 1⇑). Although juice pH varied little, between 3.11 and 3.18 (data not shown), there were significant differences in TA, with FPS 4 and 72 having greater acidities than other clones. The Champagne clones from Eperney—118, 121, 124, 130, 131, and 132—had lower TA than other French and American clones (Figure 1⇑), in contrast to the observation that sparkling wine clones were chosen for higher acidities at a given soluble solids level (Bernard 1986). This discrepancy may reflect climatic differences between California and France or, alternatively, may reflect changing selection criteria in Champagne.
Vegetative growth also showed some distinct trends. The Dijon clones, with the exception of clone 75, had relatively high pruning weight and shoot weight. The newer Champagne clones produced relatively low pruning weights and shoot weights, similar to FPS 4 (Table 3⇑).
In Cabernet Sauvignon (Wolpert et al. 1995), Merlot (Benz et al. 2006), and Pinot noir for sparkling wine (Mercado-Martín et al. 2006), the clone with the lowest yield had the highest pruning weight, which might suggest an inverse relationship between yield and pruning weight. However, in this study, although FPS 72 had the highest pruning weight and the lowest yield, the remaining 12 clones showed no significant relationship between growth and yield (r2 = 0.26, data not shown).
FPS 72 produced an RI of 1.97 (Table 3⇑), indicating that FPS 72 was significantly undercropped, as RI values of less than 2 or 3 are widely regarded as too low (Smart and Robinson 1991). One way to ameliorate this lack of balance could be to increase the shoot number for FPS 72, which could potentially increase yield. Other clones had RI values that were considered low, primarily because of low plot-average RI values in 1998 and 1999 (Table 3⇑). Plot-average RI values returned to more acceptable levels (∼5) in 2000 and 2001.
Conclusions
This trial reports the first evaluation of Chardonnay clones for sparkling wine in the United States and the first viticultural performance data of a newly introduced group of clones imported from Champagne, France. Clones differed significantly in yield and growth components and maturity indices, with California clones being both the highest and lowest yielding clones. French clones were more uniform in their viticultural characteristics. Harvest date was not closely related to yield.
Footnotes
Acknowledgments: Research funding from the American Vineyard Foundation is gratefully acknowledged.
The authors thank Michael Crumly and Robert Iantosca of Gloria Ferrer Champagne Caves for superb cooperation, M. Jason Benz for data analyses, and G. Stanley Howell for critical comments on the manuscript.
- Received January 2007.
- Revision received April 2007.
- Revision received September 2007.
- Copyright © 2008 by the American Society for Enology and Viticulture