SEARCH

SEARCH BY CITATION

Keywords:

  • Allodapini;
  • aridification;
  • Australia;
  • bees;
  • dispersal;
  • exoneurines;
  • Hill Gap;
  • historical biogeography

Abstract

Aim  Early diversification of allodapine bees occurred in Africa c. 50 Ma. They are most abundant in sub-Saharan Africa and Australia, and one of the oldest phylogenetic divergences in the tribe involves a split between an African + Malagasy clade and an Australian clade. The historical biogeographical scenario for this has been highly problematic, entailing an Eocene dispersal from Africa to Australia, followed by an unresolved, and apparently rapid, set of bifurcations leading to the Australian ‘exoneurine’ genera. Here we use an expanded taxon set of Australian species to explore the timing and historical biogeography of the exoneurine radiation.

Location  Australia, Africa, Madagascar.

Methods  One nuclear gene (F2 copy of elongation factor 1α) and two mitochondrial genes (cytochrome c oxidase subunit I and cytochrome b) were sequenced for 33 Australian exoneurine species from all five genera found on the continent, as well as for an additional 37 species from all non-parasitic genera in the remainder of the tribe. We used Bayesian inference analyses to study phylogenetic topology and penalized likelihood analyses to infer key dates of divergence within the tribe. We also used lineage-through-time (LTT) analyses and Bayesian analyses to explore the tempo of radiations and biogeographical history of the exoneurines.

Results  Results from the phylogenetic analyses were congruent with previous studies, indicating a single colonization event c. 34 Ma, too late for Gondwanan vicariance models, and too early for a Laurasian dispersal route. In contrast to earlier studies, we show that this colonization event did not result in an ancient rapid radiation. However, LTT patterns indicated a rapid radiation of the temperate-adapted genera Exoneura and Brevineura, but not of the xeric-adapted genus Exoneurella, from 10 to 6 Ma.

Main conclusions  Our results indicate a trans-oceanic dispersal event from Africa to Australia, most likely via Antarctica, with an accelerated diversification of temperate-adapted lineages during the major Late Miocene event referred to as the ‘Hill Gap’. This is the first study to link radiations in Australian bee faunal elements to changing climate, and differs from many other plant and insect phylogenetic studies by showing increased radiation of temperate clades, rather than xeric clades, with increasing aridification of Australia.