Crop estimation, an important practice for vineyardists and wineries, can be accurately done using a modified lag phase procedure developed for Washington State conditions, says winemaker and consultant Dr. Wade Wolfe.

Wolfe, of Thurston Wolfe Winery in Prosser, Washington, helped to adopt a modified lag phase procedure for crop estimates during his years at Hogue Cellars. He shared the formula during the annual meeting of the Washington Association of Wine Grape Growers which was recently held in Kennewick.

“Crop estimation is a critical cultural practice used to forecast grapes to sell or crush,” he said. Crop estimation helps growers adjust yield to achieve a desired crop level and helps wineries determine if tank capacity matches the coming harvest volumes.

Yield depends on the number of vines per acre, the number of clusters per vine, and the average cluster weight.


Techniques used to estimate the crop include eyeballing the vineyard, reviewing historical records of production, and measuring yield parameters (block density, percent of vine stand, average number of clusters per vine, and estimating cluster weight at harvest).

“Forecasting the harvest cluster weight is the most difficult parameter to come up with,” Wolfe said. He cited Washington State University research that shows that the number of berries per cluster is the biggest variable. Berry weight typically stays the same from year to year, but the number of berries within the cluster varies greatly.

Research conducted in Oregon found that berry weight at the end of lag phase, the third stage of development in which berry growth stops temporarily, is approximately one-half of its final weight. Wolfe said that in Oregon’s Willamette Valley the lag phase occurs about 50 to 55 days after bloom.

“By multiplying the berry weight at lag phase by two, Oregon’s wine industry can estimate a final berry weight at harvest,” he said.

“This is a useful tool for Pinot Noir in the Willamette Valley, but when I was at Hogue, we had 12 different varieties, 40-plus growers, and 200-some blocks of grapes. Because there is a short time period between lag phase and veraison, to measure all this at lag phase leaves a very short time for growers to adjust crop loads.”

In the late 1990s, after several years of variable yields due to drought and freezes, Wolfe and his associates at Hogue began to study the lag phase crop estimation to see if they could modify it to get earlier, more accurate information. They collected data for three years, from 1998 to 2000, on five varieties from two Yakima Valley vineyards that served as reference blocks. In terms of temperatures, the first year was representative of a hot year; 1999 was a cool year; and 2000 was average. Data were collected weekly from early July through harvest. Brix and heat units were also recorded.

“We wanted to know if lag phase is different for early and late varieties,” he said.

Cluster and berry growth

They found that lag phase occurs around the third week of July through the first week of August. There is some overlap between varieties, with Chardonnay being the earliest and Cabernet Sauvignon the latest.

“The lag phase onset dates don’t vary much for either the variety or year,” he said, adding that the average beginning date was July 20 to 24. “Lag phase starts about 45 to 50 days postbloom and occurs at about 50 percent of the season’s total heat units. Lag phase lasts for about seven days.”

In the three years of data, Wolfe also compared using Julian dates versus heat units to determine the onset of lag phase. He found that the Julian dates were better markers to identify the onset of lag phase than heat units. Generally, lag phase began around 200 to 210 Julian days.

Once lag phase was determined, they worked to see if they could assess berry weight one to three weeks before lag and still be accurate at harvest. This would give growers more time to adjust crop levels.

Cluster growth factor

Wolfe concluded that cluster growth factors or ratios outside lag phase are predictable and useful for extending crop estimation time. During 2000, the estimated Hogue crop numbers versus actual tonnage were within five percent.

By using the appropriate cluster growth factor, growers can predict their crop levels by multiplying the factor or ratio with the average cluster weight. The average number of clusters per vine and number of vines per acre must also be calculated when estimating the crop.