Why are there so many species of mining bees (Hymenoptera, Andrenidae)? The possible roles of phenology and Wolbachia incompatibility in maintaining species boundaries in the Andrena proxima-complex

Abstract

The factors leading to the extraordinary diversity of the bees of the genus Andrena Fabricius, the second most speciose genus among bees, remain largely unknown. To examine the pattern of diversification in this genus, we investigate species boundaries within the controversial Andrena proxima-complex using a dataset of ultra-conserved elements and various species delimitation analyses (admixture analyses, BPP, DAPC). Our results confirm the presence of three separate species in this species group (Andrena proxima (Kirby), A. ampla Warncke and A. alutacea Stöckhert) and suggest very low levels of interspecific gene flow. Andrena proxima and A. alutacea are regularly found in sympatry, suggesting an advanced stage of speciation. By contrast, A. ampla shows a parapatric distribution with both other species. Andrena alutacea harbours a unique strain of Wolbachia Hertig, absent in the other two species, which are infected by the two same strains of Wolbachia. In addition, the three species have distinct phenologies, with A. proxima flying earlier in the season and A. alutacea significantly later; the phenology of A. ampla is intermediate. Our results suggest that Wolbachia-induced incompatibilities and phenological differences maintain species boundaries in this group. The most advanced stage of speciation is observed in the two species showing putatively incompatible strains of Wolbachia and strongly distinct phenologies, A. alutacea and A. proxima. We propose that the smaller differences in phenological differentiation between A. ampla and both other species may explain the observed pattern of parapatric distribution. We discuss how these factors may underlie the high diversification rate observed in other groups of Andrena, a genus characterized by particularly variable and short flying periods.