AVF-funded research —
Zinfandel Heritage Vineyard Update

By James A. Wolpert & Michael M. Anderson
Department of Viticulture & Enology
University of California, Davis
The assistance of UC Farm Advisors Rhonda Smith (Sonoma), Donna Hirschfelt (Amador and El Dorado), Ed Weber (Napa), Glenn McGourty (Mendocino), Paul Verdegaal (San Joaquin), Jack Foott (formerly, San Luis Obispo) is gratefully acknowledged. The cooperation of UC emeritus viticulturist Amand Kasimatis, Ridge Vineyards viticulturist David Gates, and the department’s Oakville Vineyard manager Jason Benz, is also greatly appreciated. The Zinfandel Heritage Vineyard is a collaboration between the Zinfandel Advocates and Producers and the American Vineyard Foundation. The authors are indebted to ZAP’s board of directors, and in particular, to its research committee, and its chair Joel Peterson, for their outstanding support. We thank the American Vineyard Foundation through which this research was funded.

Zinfandel is a grape variety of noble stature. In California, we are fortunate that it has no historical stylistic context elsewhere in the world that restricts our winemakers’ creativity.

Few New World wine regions have had the opportunity to create a great new wine. With other varieties, we have constantly compared our efforts to European standards. Zinfandel gives us an unparalleled opportunity to set our own standards for unique world-class wines.

However, some winemakers have commented — and even complained — for many years that the Zinfandel clones offered through the University of California’s Foundation Plant Service (FPS) were not high-quality selections. In short, the clones were disliked because they had large, tight clusters and large berries, leading to bunch rot at relatively low ripeness. The resultant wines were criticized for having low intensity of varietal character.

Before Zinfandel’s country of origin was discovered, it was not possible to improve the variety by importing better clones, as is the case for European varieties. However, the discovery that Croatia is the country of origin turned out to be of little help, because the variety is not widely planted there, so the amount of variability Croatian plant material can contribute would not be expected to be high.

Fortunately, however, many old Zinfandel vineyards in California survived Prohibition and often dated back to the late 1890s. The reputation of old Zinfandel vineyards had been highlighted by winemakers, especially since the middle and late 1980s.

Armed with these demonstrations of quality “in the bottle” throughout California, and with the regional knowledge of UC farm advisors and local vineyard managers, a concerted effort was made to improve the UC Davis Zinfandel selections in the early 1990s.

Collecting field selections
Zinfandel selections were collected from throughout California to increase the possibility of finding selections that distinguish themselves. Collections were made from Sonoma, Napa, Mendocino, Lake, Contra Costa, Alameda, Santa Clara, San Luis Obispo, San Joaquin, Amador, El Dorado, and Calaveras counties, and the Cucamonga region of southern California.

Selections were made from vineyards at least 60 years old and older, because vines of this age appeared free of virus (visual “red leaf”) symptoms. Attention was paid to finding vines with small berries and no disease symptoms. In addition to preserving these selections as a historical legacy, our goal was to choose from among them selections for distribution that will improve the quality of Zinfandel wines.

Planting the vineyard
The Zinfandel Heritage Vineyard, located in the Oakville Experimental Vineyard (Napa Valley) currently consists of 90 selections. Phase I, budded in 1995/96, consists of 63 selections and includes certified selections of Zinfandel (FPS 1A, 2 and 3) as well as three selections of Primitivo (FPS 3, 5, and 6).

In 1999, Phase II was added to the vineyard with an additional 27 selections. Dr. Carole Meredith and her associate Gerald Dangl confirmed through DNA analysis that all the selections in Phases I and II were indeed Zinfandel. The vineyard is planted at 9-ft x 8-ft spacing (row x vine) on a Gravelly Bale Loam. St. George was used as the rootstock, and the vines are head-trained and spur-pruned. Selections in the vineyard consist of seven-vine experimental units, without replication.

Planning for this vineyard was done with a strong appreciation that this was both a repository of plant material and a collection of historic material. Therefore, the vineyard was planted as traditionally as possible. The use of St. George rootstock, nearly square spacing and head-trained, spur-pruned vines supported only by split redwood stakes is a design used widely 100 years ago. One concession to modern viticulture was installation of a subsurface drip irrigation system.



Virus status
No evaluation of the Heritage selections can be done without knowing the status of each. Often people gather selections together and take data, not realizing that the selections differ in viral status, so it is impossible to know if the demonstrated differences are due to clone or virus. The goal here was to avoid that mistake.

Relying on visual inspections, every effort was made to select vines that were free of virus. However, tests would be needed to confirm the viral status of the selections. All selections were tested for grapevine fanleaf virus (GFLV) prior to budding.

Polymerase chain reaction (PCR) detection of virus in grapevines, completed in 1999, is now felt to be more reliable than traditional woody indexing. From this point forward, PCR will be used as the definitive virus test for this project. Of the viruses detected in plant material with use of PCR, grapevine leafroll (GLR) was by far the most common, with 46% of the selections infected with one or more of the GLR strains (data not shown).

Other detected viruses were generally found in combination with a GLR strain. Only one selection that was free of GLR was found to have one of the other viruses. The high number of selections testing positive for GLR was not anticipated, reminding us once again that the lack of red leaves in fall is far from assuring a negative virus status.

Viticultural data
In 1998, we began viticultural evaluations of the selections in Phase I. These measurements are taken at harvest and include Brix, pH, TA, berry weight, yield per vine, cluster weight, number of clusters, and pruning weight per vine. Collection of yield-per-vine and clusters-per-vine data began in 1999.

Table I reports mean data inclusive of all years. Yield, measured as the mean of three vines per selection, nearly doubled from a low of 3.5 kg to a high of 7.7 kg per vine. Mean yield was 5.0 kg per vine. Brix values ranged from a high of 26.3† to a low of 23.2† with the mean of 24.2†.

Average cluster weights varied almost two-fold, from 170 g to 336 g, a variation resulting from a combination of both the number of berries per cluster and berry weight. In order for a selection to be given advanced consideration, vines must perform uniquely and consistently over time relative to other selections.

The Primitivo selections have their origin in Italy and therefore may represent a line of Zinfandel different from those we collected in California. When compared to the Heritage Vineyard selections as a whole, several general observations can be made. Primitivo selections (Table I) all had values of yield, cluster weight, berry weight, and berries per cluster that were equal to or lower than the mean for all selections. Primitivo FPS 05 had both the lightest clusters and the smallest berries in the entire vineyard.

Additionally, the Primitivo selections had values for clusters per vine and soluble solids that were above the vineyard mean. At this point, the Primitivo selections, despite having more clusters per vine, seem to have lighter yields resulting from smaller berries and fewer berries per cluster that ripen earlier than most of the other selections in the Heritage Vineyard.

The FPS Zinfandel selections were singled out to investigate whether their reputation as large- berried, large-clustered, high-yielding selections was justified. The research showed that on no parameter did these selections set the high or low value for the vineyard. Nor did they fall outside the range set when looking at the mean ± one standard deviation. In general, we can say that up to this point, there is no data that distinguishes them from the Heritage selections.

Average yield data show a clustering between 4 and 6 kg per vine (Figure I). However, it is interesting to note that the data for some selections are quite variable, as seen by the length of the error bars, for example, on selections 25, 37, and 46. This means that the yield is not consistent from year to year. Virus status is identified with different symbols in the figure, and there is no correlation of virus status with either yield or variability of the data.



New replicated vineyard
As research progressed, we came to believe we would not learn all we wanted to about the Heritage selections without a fully replicated trial. The advantages of replication appeared to be two-fold. First, we would have statistically valid comparisons, which would increase our confidence when we eventually make recommendations. Second, we would have much more fruit for winemaking trials.

One problem in our planning for a replicated vineyard was that the large number of selections made a large vineyard impossible. So we proceeded on a modified scale.

In 2001, a new replicated vineyard was planted. Deciding what selections to include required constructing a logical scheme. Within the Heritage Vineyard, there are multiple selections made from the same vineyard, and some of the selections have been found to contain virus. Our scheme employed these facts. Using the criteria of virus-free and unique vineyards origin, the number of selections was reduced from 61 to 20. In the case of vineyards with multiple clean selections, one selection was arbitrarily chosen unless wine had been made from one of the choices previously.

The new vineyard, like the original Heritage Vineyard, is located in the Oakville Experimental Vineyard’s “Old Federal” vineyard. It consists of five replications of 18 vines occupying two acres. Once again, St. George rootstock was used. The vines are planted at a spacing of 6 x 8 and will be head-trained and spur-pruned.

It is anticipated that this vineyard will produce 450 kg (1,000 lbs) of fruit per selection, and that this will be sufficient to produce more than one barrel of wine per selection.

The rootstock was planted in 2001 and was budded in spring 2002. Drip irrigation was installed in 2001. This vineyard represents a commitment to continued research on Zinfandel. The project continues to expand the understanding of Zinfandel, and we are excited about what will be achieved.

The range shown in growth and yield parameters thus far fuels our hope that there is significant variability within the Zinfandel Heritage selections. Using the Heritage Vineyard as a base, we hope to identify Zinfandel selections that will achieve the status of clones and play important roles in the production of Zinfandel for years to come.