Kaylah Vasquez, Marc Fuchs, Monica Cooper, and Luca Brillante*
*Department of Viticulture and Enology, California State University Fresno, 2360 E Barstow Ave, Fresno, CA 93740 (lucabrillante@csufresno.edu)

The cost of virus infections per acre in vineyards with an average lifespan ranges from $10,000 to $16,000. With 880,000 acres, California has ~80% of the total vineyard acreage in the U.S. Early detection and eradication of infected plants are crucial to managing pathogen spread, as there is no cure for viruses in the field. Chemical control of insect vectors of viruses is of limited efficacy and often not an option in organic agriculture. Supported by preliminary data, our project used remote sensing coupled to machine learning to detect major grapevine diseases (red blotch and leafroll) more efficiently than traditional methods. In 2020, we monitored grapevine infections in two vineyards located in Madera and Rutherford, planted with Cabernet Sauvignon and Cabernet franc, respectively. Around 400 plants in each vineyard were imaged twice with a VIS-NIR hyperspectral, thermal, and high-resolution RGB camera mounted on a tripod, an ATV, or a UAV. Leaves were collected for further imaging in a dark room in the laboratory and then used for molecular analyses to assess virus presence. All plants in the vineyard blocks were also scouted to assess the presence of the infection and compared to high-resolution satellite imagery. The imagery was used to train machine-learning algorithms to predict grapevine infection, which show promising performances and will be further optimized in the second season of measurements.

Funding Support: CDFA Specialty Crop Block Grant Program