Biogeography, climate, and land use create a mosaic of parasite risk in native bumble bees

Host-parasite interactions are crucial to the regulation of host population growth, as they often impact both long-term population stability and ecological functioning. Animal hosts navigate a number of environmental conditions, including local climate, anthropogenic land use, and varying degrees of spatial isolation, all of which can mediate parasitism exposure. Despite this, we know little about the potential for these environmental conditions to impact pathogen prevalence at biogeographic scales, especially for key ecosystem service-providing animals. Bees are essential pollination providers that may be particularly sensitive to biogeography, climate, and land-use as these factors are known to limit bee dispersal and contribute to underlying population genetic variation, which may also impact host-parasite interactions. Importantly, many native bumble bee species have recently shown geographic range contractions, reduced genetic diversity, and increased parasitism rates, highlighting the potential importance of interacting and synergistic stressors. In this study, we incorporate spatially explicit environmental, biogeographic, and land-use data in combination with genetically derived host population data to conduct a large-scale epidemiological assessment of the drivers of pathogen prevalence across >1000 km for a keystone western US pollinator, the bumble bee Bombus vosnesenskii. We found high rates of infection from Crithidia bombi and C. expoekii, which show strong spatial autocorrelation and which were more prevalent in northern latitudes. We also show that land use barriers best explained differences in parasite prevalence and parasite community composition, while precipitation, elevation, and B. vosnesenskii nesting density were important drivers of parasite prevalence. Overall, our results demonstrate that human land use can impact critical host-parasite interactions for native bees at massive spatial scales. Further, our work indicates that disease-related survey and conservation measures should take into account the independent and interacting influences of climate, biogeography, land use, and local population dynamics.