Off-the-shelf load cells used in Julie Tarara’s research automatically detect increases in the tension of a grape trellis wire every ten seconds, formulating an average every 15 minutes.

Off-the-shelf load cells used in Julie Tarara’s research automatically detect increases in the tension of a grape trellis wire every ten seconds, formulating an average every 15 minutes.

Dr. Julie Tarara is completing research that could revolutionize the way grape crops are managed by providing real-time information on the crop’s progress throughout the growing season. The technique could help growers to better manage the crop and be used to estimate crop tonnage.

The current method of estimating grape crops is time consuming, labor intensive, and not always accurate. Traditionally, the size of a grape crop is estimated by counting and weighing clusters and berries once or twice during the season—taking a snapshot of that moment in time—and comparing the data to past records.

Last year, California wine industry officials were way off in their crop projections and caught by surprise by the record crush of 4.3 million tons. The record crop represents a 20 percent increase from the previous year. Berries sized up larger than normal, with most of the extra tonnage developing at the end of the season.

Tarara, research horticulturist with the U.S. Department of Agriculture in Prosser, Washington, and colleagues John Ferguson and Paul Blom are employing new technology to measure the tension changes that occur in the cordon wires of vineyard trellis systems as the vine canopy and fruit grow. By continually collecting information from the trellis tension, growers and processors can track fruit development throughout the season, as well as other vineyard activities like irrigation and crop thinning.

Since 2001, Tarara has been working with commercially available electronic load cells to monitor changes in the grape crop. The load cells, which cost about $300 each, are spliced to the cordon wire and connected to a data logging device that measures the tension signal every ten seconds and stores the average value of the signals every 15 minutes, giving a total of 120 tension measurements every day. The measurements are downloaded to a computer twice a week for processing.


This season, the ARS research team is collecting the last of three years of field data from ten commercial juice grape vineyards and two research vineyards. They expanded the study in 2005 to include two research vineyards (a Concord juice and a Merlot wine grape block) at the Prosser station. The Merlot block allows the scientists to use the cell loads on a lower-yielding wine block with a different trellis configuration than the heavy-producing juice grapes.

The commercial blocks are the same ones used by Yakima Valley processors to annually forecast the juice grape crop.

To collect more sensitive measurements, Tarara installed steel braces in the vineyards as part of her study. The braces are placed next to two trellis posts in a central portion of a row to isolate a span of trellis. The brace allows the cordon wire between the steel end posts to be separate from the rest of the row’s cordon wire, with the remaining cordon wire channeled or bypassed above the brace and reconnected to the row’s wire. The brace was installed in all of the testing plots.

By placing the brace in the same location, they can replicate and compare what the processors are currently doing, she said, adding that the brace ensures that post movement isn’t skewing tension changes. Also, by decreasing the length of the cordon wire being measured, the changes in tension measurements are more accurate.

“The brace equipment was robust enough to withstand mechanical harvesters that were used to thin and harvest the crop last year,” Tarara said. Tests showed that they also withstood other normal over-the-row vineyard operations like spraying, mechanical pruning, and mechanical thinning.

At bud break, the tension is set at zero, she explained. By the time flowering takes place, the tension has increased by about one-fourth of the total amount gained in the season.

The temperature of the wire affects the tension measurement, because the wire stretches as it heats during the day and contracts when it cools at night.

The trellis wire temperature is collected at the research plots; data are later adjusted with an algorithm for the temperature effects.


While she believes that the tension measurements can be as accurate or more so than hand sampling—within 5 percent of the actual yield—the technology does not address the issue of where to sample or how many locations need sampling to provide numbers that are representative of the vineyard.

“This particular technology cannot overcome the sampling issue,” she said. “It can be used as an indicator. I can’t tell you how many rows you would need to monitor within a vineyard to give you a representative number. That depends on the variability of your vineyard.”

But variability is also a problem with the current hand sampling method. The crop projection is only as representative as the vines that it came from. Generally, on an industrywide basis, the hand counts tend to average out data from the low- and high-yielding vines.

“It’s a dilemma,” Tarara said in regard to representative sampling. “Some in the industry want a silver bullet. The technology is really a monitoring tool that gives growers a new way to look at growth every single day. That’s something that hand sampling can’t do.”

A grower could use only one load cell in the vineyard if the location was representative, she notes. But if the vineyard is highly variable, multiple locations will be needed.

Technology transfer

The last major hurdle facing the project is getting the load cell and data logger into a user-friendly commercial package for the grape industry.

Tarara believes that, once a package is fully developed, it could be adapted to vineyards around the world. The wire tension method could be applied to any grape vineyard that uses a trellis with at least one horizontal support wire. She envisions an agricultural service or consulting firm maintaining Internet postings of the season’s growth for wineries, processors, or growers.

It could even have application to the tree fruit industry in high-density, trellised orchards, depending on the wire or wires that bear most of the crop load. Up-to-date crop information would be valuable to tree fruit growers who depend on hand labor for harvesting.

But for now, as she moves into the final stage of the project, she hopes that a private company will show interest and transfer the technology to industry.

“We’re not a business enterprise,” she said, “but we welcome cooperative development agreements with private sector partners so that technologies developed within ARS may be transferred to the commercial realm.”

Everyone involved in the grape industry wants accurate crop information, from the grower who may be selling tonnage on the cash market, to the winemaker who must make sure enough tank space is available, to the juice or wine marketer who must manage inventory, she said.

Funding for the project has come from USDA, Washington State University’s Center for Precision Agriculture Systems, and most recently from a USDA Risk Management Agency grant awarded to the Washington Wine Industry Foundation to develop user-friendly information for growers and industry members.