Simulating crop-disease interactions in agricultural landscapes to analyse the effectiveness of host resistance in disease control: The case of potato late blight

Disease-resistant potato varieties can play a key role in sustainable control of potato late blight. However, when these varieties are more widely used, resistance breakdown can occur as a result of pathogen adaptation. Here we focussed on potato cultivation in the Netherlands, where new (single gene) resistant varieties have been introduced over the last ten years. This new generation of late blight resistant varieties has moderate yield levels and does not meet all market requirements. As a result, adoption rates for resistant varieties have been low so far. We developed a spatially explicit agent-based model to simulate potato production, disease spread and pathogen evolution at the landscape level. We analysed how late blight severity, resistance durability and potato yield are affected by the spatial deployment of a resistant variety, with a lower potential yield than susceptible varieties. The model was applied to an agricultural region in the Netherlands (596 km(2)) and was run for a period of 36 years using daily weather data as input for crop growth and disease dynamics. The short- and long-term effects of the deployment of a resistant variety were analysed with the model. With respect to short-term dynamics, years were analysed independently to study between year variation. The model demonstrated that in most years, susceptible fields without fungicide application suffered severe yield losses and resistant fields performed better despite their lower potential yield. Resistance breakdown was observed in a small fraction of fields with the resistant variety, but this did not affect mean potato yield or disease incidence in the short term since it occurred at the end of the growing season. Increasing the fraction of potato fields with the resistant variety strongly reduced late blight infection within a landscape. With respect to the long-term effects, the model showed the emergence and spread of a virulent strain over time. The virulent strain gradually took over the pathogen population, decreasing mean potato yields from fields with the resistant variety. This occurred in all landscape compositions where the resistant variety was deployed to different extents. It was found that low as well as high proportions of fields with the resistant variety could increase durability of resistance. With these findings, the model provided more insight into the opportunities and risks related to the use of plant resistance in disease control, an important and sustainable disease management strategy.