Abstract
Bees (Apoidea: Anthophila) are recorded as common hosts of a diverse array of parasitic Diptera species, but information is widely scattered across the literature, limiting our knowledge of global bee-dipteran association patterns. Here, we retrieved published information on 453 bee-dipteran associations to test the role of life-history traits of hosts and parasites on such patterns. The 13 families and 37 genera of Diptera associated with bees are unevenly distributed across bee families within a specialized and modular network. The family Apidae was the most associated with parasitoids of adults, Colletidae were the most attacked by nest-entering larval parasitoids, Apidae and Megachilidae were the most attacked by non-nest-entering larval parasitoids, and cleptoparasites were more frequently observed attacking Andrenidae and Halictidae nests. The number of parasitic strategies, but not the richness of Diptera, was higher in fossorial species and both weakly depended on bee phylogeny. However, when considered singularly, the four parasitic strategies ((i) parasitoids of adults; (ii) nest-entering larval parasitoids; (iii) non-nest-entering larval parasitoids; and (iv) cleptoparasites) yielded moderate phylogenetic signals. Fossorial species were more associated with cleptoparasites, while the association rate of non-nest-entering larval parasitoids was higher in aerial-nesting bee species and in solitary bee species. In addition, fossorial bee species were found to be more specialised in their associate dipteran fauna. Despite the uneven distribution of studies across bee genera, our results suggest that these associations emerged through co-evolutionary processes, during which dipterans specialized and optimized their parasitic strategies based on the ecological traits of their host bees.