The bee tree of life: a supermatrix approach to apoid phylogeny and biogeography

Background: Bees are the primary pollinators of angiosperms throughout the world. There are more than 16,000 described species, with broad variation in life history traits such as nesting habitat, diet, and social behavior. Despite their importance as pollinators, the evolution of bee biodiversity is understudied: relationships among the seven families of bees remain controversial, and no empirical global-level reconstruction of historical biogeography has been attempted. Morphological studies have generally suggested that the phylogeny of bees is rooted near the family Colletidae, whereas many molecular studies have suggested a root node near (or within) Melittidae. Previous molecular studies have focused on a relatively small sample of taxa (similar to 150 species) and genes (seven at most). Public databases contain an enormous amount of DNA sequence data that has not been comprehensively analysed in the context of bee evolution. Results: We downloaded, aligned, concatenated, and analysed all available protein-coding nuclear gene DNA sequence data in GenBank as of October, 2011. Our matrix consists of 20 genes, with over 17,000 aligned nucleotide sites, for over 1,300 bee and apoid wasp species, representing over two-thirds of bee genera. Whereas the matrix is large in terms of number of genes and taxa, there is a significant amount of missing data: only similar to 15% of the matrix is populated with data. The placement of the root as well as relationships between Andrenidae and other bee families remain ambiguous, as several alternative maximum-likelihood estimates fall within the statistically credible set. However, we recover strong bootstrap support for relationships among many families and for their monophyly. Ancestral geographic range reconstruction suggests a likely origin of bees in the southern hemisphere, with Melittidae ancestrally located within Africa, and Halictidae, Colletidae, and Apidae within the New World. Conclusions: Our study affirms the monophyly of each bee family, sister-taxa relationships between Apidae and Megachilidae (the 'long-tongued bees'), between Colletidae and Stenotritidae, and between Colletidae + Stenotritidae and Halictidae. Our analyses reject a Colletidae-basal hypothesis for family-level relationships and instead support Melittidae as sister to the remaining bees. Southern hemisphere vicariance likely played an important role in early diversification within many bee families.