The role of landscape factors in shaping bumble bee pathogen loads across regions of the eastern Nearctic

Pathogens and parasites are drivers of declines in bumble bees. Their levels can be influenced by numerous abiotic and biotic factors, thus managing disease in these bees requires understanding the relative impact of these factors on pathogen loads. We evaluated loads of black queen cell virus (BQCV) and deformed wing virus in bumble bees (Bombus impatiens) and honey bees (Apis mellifera), and loads of parasites Vairimorpha bombi and Crithidia bombi in B. impatiens, from sites varying in habitat type and quality across North Carolina, USA. Pathogen loads were assessed against metrics for land cover, floral quality, bee diversity and weather. Results were compared against similar data from Pennsylvania, USA, and other landscape studies on bumble bees in North America. Spatial variation in pathogen loads was lower in bumble bees in North Carolina than in Pennsylvania, which reduced the power to detect landscape effects. For example, Crithidia was fairly ubiquitous and Vairimorpha was not detected. Data from both states revealed that developed land and honey bees were most consistently positively correlated with viral loads, especially for BQCV, whereas forest and nesting habitat availability were often negatively correlated with loads. Multivariate models only supported a positive relationship between summer floral resources and BQCV loads in North Carolina bumble bees. Together with the broader literature, these data indicate that levels of developed land and honey bees are most associated with increased pathogen loads while floral availability and bee community composition show more context‐dependent effects. Bumble bees and honey bees across North Carolina, USA harbour more viruses in developed areas and areas with more summer forage, and less in forested areas. Cross‐state and literature surveys suggest development and honey bee levels most consistently increase bumble bee viruses, but floral resources and community have mixed effects. More ubiquitous abundance of Crithidia and lack of Vairimorpha in North Carolina compared with Pennsylvania, USA, suggest regional variation in pathogen incidence that impacts ability to infer landscape patterns.