Plant-pollinator interactions in Mediterranean semiarid ecosystems

Pollinators are fundamental for plant reproduction in natural and agricultural ecosystems. However, their populations are declining worldwide, threatening the functioning of the ecosystem service they provide. The factors driving this change are manifold, but land use changes and interspecific transmission of pathogens between managed and wild bees are prominent. In this context, most research efforts have focused on specific taxa and rarely at the community level, limiting our ability to fully understand the effects of global change on the functioning of plant-pollinator interactions in ecosystems.

Here, we investigate the impact of human activities (beekeeping and land use intensity) on the spread of an emergent pathogen (Vairimorpha ceranae) in Mediterranean wild bee communities inhabiting landscapes with varying levels of anthropogenic disturbance. Plant-pollinator interactions were sampled in nine one-hectare plots along a gradient of land use (urban structures, croplands and natural vegetation) and beekeeping intensity. We analised the impact of human disturbances on pollination networks and pathogen prevalence and applied a network approach to examine whether total effects of species in networks (i.e. direct plus indirect interactions) explain pathogen spread through bee communities.

We found that V. ceranae prevalence in honey bees is not a good predictor of the pathogen spread through bee communities. There seems to be a temporal mismatch between pathogen dynamics in managed and wild bees. Networks with more diversity of interactions and more plants showed less pathogen prevalence, but total effect analyses (i.e. combining direct and indirect interactions) failed to explain pathogen transmission across pollination networks. Croplands increased wild bee density, and interactions and species diversity in networks while shrublands had the opposite effect. Our results highlight the importance of studying pathogen dynamics at the community level and analysing species interaction patterns to improve our understanding of pathogen spread through communities.