Manipulating host resistance structure reveals impact of pathogen dispersal and environmental heterogeneity on...

Understanding how variation in hosts, parasites, and the environment shapes patterns of disease is key to predicting ecological and evolutionary outcomes of epidemics. Yet in spatially structured populations, variation in host resistance may be spatially confounded with variation in parasite dispersal and environmental factors that affect disease processes. To tease apart these disease drivers, we paired surveys of natural epidemics with experiments manipulating spatial variation in host susceptibility to infection. We mapped epidemics of the wind-dispersed powdery mildew pathogen Podosphaera plantaginis in five populations of its plant host, Plantago lanceolata. At 15 replicate sites within each population, we deployed groups of healthy potted 'sentinel' plants from five allopatric host lines. By tracking which sentinels became infected in the field and measuring pathogen connectivity and microclimate at those sites, we could test how variation in these factors affected disease when spatial variation in host resistance and soil conditions was minimized. We found that the prevalence and severity of sentinel infection varied over small spatial scales in the field populations, largely due to heterogeneity in pathogen prevalence on wild plants and unmeasured environmental factors. Microclimate was critical for disease spread only at the onset of epidemics, where humidity increased infection risk. Sentinels were more likely to become infected than initially healthy wild plants at a given field site. However, in a follow-up laboratory inoculation study we detected no significant differences between wild and sentinel plant lines in their qualitative susceptibility to pathogen isolates from the field populations, suggesting that primarily non-genetic differences between sentinel and wild hosts drove their differential infection rates in the field. Our study leverages a multi-faceted experimental approach to disentangle important biotic and abiotic drivers of disease patterns within wild populations.