Mixed infections reveal virulence differences between host-specific bee pathogens.

Dynamics of host-pathogen interactions are complex, often influencing the ecology, evolution and behavior of both the host and pathogen. In the natural world, infections with multiple pathogens are common, yet due to their complexity, interactions can be difficult to predict and study. Mathematical models help facilitate our understanding of these evolutionary processes, but empirical data are needed to test model assumptions and predictions. We used two common theoretical models regarding mixed infections (superinfection and co-infection) to determine which model assumptions best described a group of fungal pathogens closely associated with bees. We tested three fungal species, Ascosphaera apis, Ascosphaera aggregata and Ascosphaera larvis, in two bee hosts (Apis mellifera and Megachile rotundata). Bee survival was not significantly different in mixed infections vs. solo infections with the most virulent pathogen for either host, but fungal growth within the host was significantly altered by mixed infections. In the host A. mellifera, only the most virulent pathogen was present in the host post-infection (indicating superinfective properties). In M. rotundata, the most virulent pathogen co-existed with the lesser-virulent one (indicating co-infective properties). We demonstrated that the competitive outcomes of mixed infections were host-specific, indicating strong host specificity among these fungal bee pathogens.