Blueberry Forecasting

New Block Grant Could Help Predict Yields


The Florida blueberry industry is starting the first of a two-year yield study with the goal of giving growers and retailers better ways to predict how much product will be available throughout the season.

The Florida Blueberry Growers Association was awarded a $135,000 research grant in 2018 in order to develop a yield forecasting system for the Florida blueberry industry. Prior to this grant, there had been no efforts to forecast the state’s blueberry production with accuracy in advance of production. While some growers had sophisticated systems that could predict week-to-week harvest, that kind of predictability was not available to all growers, according to Brittany Lee, President of the Florida Blueberry Growers Association.

“Developing a forecasting model to predict the yield of the statewide Florida blueberry crop will enable Florida growers to realize and reclaim its place as the cornerstone of the U.S. blueberry industry as the first state with domestic blueberries,” says Lee, who authored the grant proposal. “Smooth relationships with retailers also required predictability of volume and timing of anticipated supply, and Florida’s short harvest season leaves little room for error in projected yields.”

While the Florida blueberry market has more than doubled from $37 million in sales in 2007 to over $82 million in 2016, it faces pressure from Georgia in the north and South America to the south. Add to those pressures the increased acreage devoted to the crop in Mexico, and the need for a reliable model becomes more apparent for the 1,000-plus blueberry growers in the state.

The first part of the study actually started this past season as the FBGA actively started to poll growers to try to compile information related to acreage by variety, age, and growing region. The commission also attempted to collect environmental data from each growing region including chill hours, heat units, rain, and humidity. For the environmental data, the study will use the Florida Automated Weather Network (FAWN) to obtain localized weather data to help develop a preliminary statistical forecasting model.

A total of six farms — two in each of Florida’s three blueberry-growing regions — were selected as participants in the study, according to Lee. These farms represent 15 to 20 percent of the Florida acreage and represent a broad survey of different varieties that will accurately capture the variation and production diversity of the state’s blueberry industry.

The Florida Blueberry Association contracted with Highland Precision Ag of Mulberry to conduct the survey. Chris Crockett, the director of research and grower services, will lead a team that is responsible for bud, bloom, and early berry surveys. The team is also the primary source of data analysis, model development, and industry-wide predictions from the collected data.

The collection phase of the study started in January. The researchers from Highland Precision Ag will visit each of the six participating farms during the 20-week window of blueberry season. Crockett will then report to an Advisory Committee of the Blueberry Association once a week. Once a month, according to Brittany Lee, the Association will send out a grower’s summary to all the farms in all of the regions, in order that the individual growers can compare the collected data with their own.

“We want to make sure all of our growers are seeing the same or similar numbers,” says Lee. “If not, those growers should contact us so that we can get someone to visit them and check on their crops in order to get as accurate data as possible.”

After the initial 20-week window ends in May 2019, Highlands Precision Ag compile all of the data, develop statistical models and extrapolate industry-wide numbers in order to compare the model with data received from growers. This comparison will then be fed into a statistical model that will be used to help predict the 2020 blueberry yield. By accumulating the data over a two-year period, the Association hopes to establish a predictive model that can adequately predict Florida blueberry yields in the future, according to the proposal. While the model will most likely variety and region-specific, says Lee, there will be a potential to a set of methods to apply to all current and emerging varieties of blueberries.

The hope is that the project will be self-sustaining after the two-year study has been completed. The Florida Blueberry Growers Association is exploring the possibility of forming a Commission, which will be charged with educating and informing growers of the reporting and modeling process in order to maintain the most-up-to-date and accurate data. The Commission would serve as a grower-funded source to continue the forecasting efforts started by the grant as well as a marketing tool that will promote the Florida blueberry crop to the global retailer and consumer. Research shows that consumers are more likely to choose blueberries from Florida when available in stores, according to the proposal.

For Lee and the rest of the Florida Blueberry Growers Association, the immediate needs are two-fold.

“What we need from growers is to first buy-in to the process,” says Lee, “and second to respond to the requests for acreage and varieties so that we can, as accurately as possible, predict the crop.

“We feel forecasting is in the best interest of the entire blueberry industry to help uphold our standing in the global marketplace.”

Share this post: