Interaction between pattern, process and scale in plant disease epidemics

Optimal arrangement of crops and crop varieties could make agricultural landscapes more resilient to plant pathogen invasions, but is there a 'correct' scale at which to introduce artificial heterogeneities in plant communities? For example, aggregation of host fields into clusters could serve to hamper disease spread by increasing the separation distances between host areas, but aggregation of clusters into super-clusters could lead to large crop losses if just one or two areas became infected. Here we present a spatiotemporal simulation framework to investigate the scale-dependence of pattern-process relationships in plant disease epidemics. Epidemics that result from transmission of both vectorborne and airborne infectious agents are simulated using models such as a standard SEIR model and Levy flights to describe vector movement, and a previously published model for potato late blight linked to an atmospheric dispersion model. We introduce a new class of neutral landscape model that facilitates the simultaneous study of host pattern-epidemic process relationships across a continuum of spatial scales, i.e., from plant-scale through to intercontinental-scale. In some model scenarios, landscape designs that reduced disease at one spatial scale led to increased disease at other spatial scales. These initial results indicate potential for the development of new scale-appropriate management strategies.