Project Report:
This project was successful in several regards. It provided support for Olena Sambucci as GSR, who ended up completing some of her dissertation on powdery mildew management in California grapes as a related by-product of her involvement in this project. Based on her contributions, we ended up generating a number of interesting insights and publishing an ARE Update article: “Behavioral Responses to Disease Forecasts: From Precision to Automation in Powdery Mildew Management” https://s.giannini.ucop.edu/uploads/giannini_public/57/a4/57a43d40-0972-40dc-8c8e-4fe5872498fd/v20n1_2.pdf. Unfortunately the real-time powdery mildew detection technology, which showed promise in the lab, delivered less successful results in practice. Based on this challenge, we pivoted in our work. For the reasons outlined in our ARE Update article, this would be an interesting topic to revive once the technology catches up to the evaluation.

The prevention of powdery mildew outbreaks is among grape growers' most important management practices. The explosive episodes of powdery mildew growth that are possible when optimal weather conditions prevail pose substantial production risks to growers. The Gubler-Thomas Powdery Mildew Index (PMI) models powdery mildew growth as a function of length of exposure to different temperature ranges in order to help growers to better anticipate outbreaks. This project aims to understand how growers' use of the PMI shapes their powdery mildew control strategies according to their operational environmental and how these potentially heterogeneous forecast responses determine the environmental impact of the PMI. We find that the environmental impact of the PMI may actually be negative because growers' adjust their powdery mildew management strategies along several dimensions in response to the information–including shifting to higher potency synthetic fungicides, increasing dosage rates, and using multiple products when the risk is high according to the PMI.