Abstract
Range shifts in plant disease distributions are sensitive to scaling processes; however, few crop case studies have included these considerations. High-quality wines are increasingly being produced in topographically heterogeneous river valleys, and disease models that capture steep relief gradients are especially relevant. Here, we show that nonlinear epidemiological models more accurately reflect the threat of an emerging grapevine pathogen in areas with significant spatial gradients. By comparing the results of simulations using climate data with different spatial resolutions, we identified an increased risk of Pierce's disease caused by the vector-borne bacterium Xylella fastidiosa in wine regions worldwide. Over 100,000 vineyards worldwide were analysed, with an increase from 21.8 to 41.2% of the area at risk in Europe, from 5.6 to 47.2% in South Africa, and to a lesser extent in other wine-growing regions. This general trend has been preceded by an accelerating rate of increase in risk within wine-growing areas. Our analysis demonstrates the importance of microclimatic conditions, highlighting previously unresolved risk zones in areas close to rivers and valleys and the insufficiency of lower-resolution datasets to capture complex climatic variations.