Modeling Uncertainty in Grapevine Powdery Mildew Epidemiology Using Fuzzy Logic
Powdery mildew is the most important disease of grapevines worldwide. Despite the potential for rapid spread by the causal pathogen, grape powdery mildew has been effectively managed using fungicide applications applied based on a calendar schedule or modeled disease risk index. Various epidemiological models for predicting disease development or risk have helped to improve disease management. The Gubler-Thomas (GT) risk index is a popular disease risk model used by many growers in the western U.S. We modified the GT risk index using fuzzy logic to address both biological and mechanical uncertainty in the pathosystem. The spraying schedule suggested by the fuzzy-modified GT risk index was tested in eight site-years. Overall, the fuzzy-modified risk index maintained comparable levels of disease control as both the original model and a calendar based treatment, and had significantly less disease than the untreated control. The fungicide use efficiency of the fuzzy-modified GT risk index suggests that the updated risk index was significantly more efficient with fungicide applications than both the calendar and original GT risk index.