Multi-scale modeling of potato late blight epidemics

Keywords: Solanum, invasion, Gaussian plume model, functional connectivity, landscape design Proper landscape-scale deployment of disease resistant genotypes of agricultural crop species could make those crops less vulnerable to invasion by resistance breaking genotypes. Here we develop a multi-scale, spatio-temporal model of the potato late blight pathosystem (Phytophthora infestans - Solanum tuberosum) to investigate spatial strategies for the deployment of host resistance. The model comprises a landscape generator, a potato late blight model, and a suite of aerobiological models, including an atmospheric dispersion model. Spatial phenomena are solved using Fast Fourier transforms. Increasing the number of host genotypes caused the greatest reduction in epidemic extent, followed by reduction of the proportion of potato in the landscape, lowering the clustering of host fields, and reducing the size of host fields. Simulation results showed that spatial spread through short-distance “island hopping” is not a prerequisite for P. infestans invasions, and it appeared not possible to generate host free zones at the landscape level that were large enough to provide worthwhile levels of resilience against disease invasion from one host area to another. Deployment of host resistance in genotype mixtures had a large effect on disease invasion. A new functional connectivity parameter, characterizing the probability of successful infection following spore dispersal, proved to be useful in interpreting these results. Variation in simulation results revealed the importance of using an atmospheric dispersion model for dispersal, with large weather data sets, and many random landscape iterations. The specific coincidence in time and space between weather conditions and the geographic locations of source and target sites defined true landscape connectivity and determined model results regarding inoculum exchange between fields. Given the apparent capacity of P. infestans for long distance transport of viable inoculum, it can be concluded that spatial resistance deployment strategies that center on the creation of spatial barriers to disease at scales up to several kilometers may not be effective in mitigating invasions of virulent pathogen strains. Strategies that induce finer-grained spatial and genotypic heterogeneities in host populations are more limiting to epidemic spread. Genotype mixing was an effective option for generating agricultural landscapes that are comparatively resilient to disease invasion