Consecutive Heat Acclimation and Modeling Uncertainty for Grapevine Powdery Mildew (Erysiphe necator).

摘要

Grapevine powdery mildew, caused by the ascomycete Erysiphe necator, is a major threat to grapes worldwide. Despite its global impact on grape production,Erysiphe necator is sensitive to adverse environmental conditions, such as excess heat, free water and UV radiation. Using detached leaf co-culture assays, three-day-old single colonies of Erysiphe necator were exposed to one, two or three consecutive days of punctuated heat stress. While there was a consistent decrease in colony growth after a single heating event, there were little to no significant effects from subsequent heating events on colony growth. Similar effects were observed on the latent period, with a large initial effect from the first heat treatment and small marginal effects from subsequent heat treatments. Erysiphe necator colonies growing on live pot-grown plants were affected similarly by consecutive heat stress events. These data suggest that Erysiphe necator is more adaptable to environmental stress than previously thought.;Several epidemiological models have been developed to predict powdery mildew disease onset and disease intensity based on weather conditions, most notably the Gubler-Thomas (GT) model. The GT model uses hourly temperature data to predict the disease intensity based on optimum and lethal conditions for the pathogen, and results in an advisory forecast for disease management. However, scattered weather stations and microclimates lead to uncertainty in hourly weather data in specific locations. There is also uncertainty about how the pathogen behaves in the field in response to unfavorable conditions, although recent and current studies are helping to fill these knowledge gaps. Fuzzy logic is a multi-valued logic developed to help deal with uncertainty. In this study, we used fuzzy logic to modify the GT model to adapt it to uncertainty in weather data and pathogen biology. The fungicide spray programs suggested by the GT and the fuzzy GT models were tested at eight sites in California and Oregon. Based on regular disease severity measurements, we found that the fuzzy GT model performed as well as the original GT model, and could reduce the overall number of fungicide applications while maintaining comparable disease control. The future prospects for such models are discussed.