Search for: All records

ABSTRACT AimInsect brood parasites (i.e., cleptoparasites), like cuckoo bees, typically attack hosts within specific lineages, but seem to be less constrained by the biogeographical movements of their hosts compared to obligate parasites. Cuckoo bees depend on stable host populations, being particularly sensitive to environmental changes and thus valuable bioindicators of the bee community health. We here test the congruence between the biogeographical history of cuckoo oil bees and their oil bee hosts. LocationThe Americas. TaxonBees (Hymenoptera, Apidae). MethodsUsing phylogenomic and Sanger sequence data, we present new time‐calibrated phylogenies for cuckoo oil bees in the ericrocidine line and their oil bee hosts,CentrisandEpicharis.We estimate their ancestral ranges using six historical biogeographical models on a set of 100 trees, randomly sampled from the posterior distribution of phylogenies in each group, thus accounting for uncertainties in divergence time estimates and model selection. ResultsThe origin of the hosts stem in the Cretaceous precedes the origin of their cleptoparasite's stem in the Palaeocene. Cleptoparasite and host crown origins were synchronous in the Eocene, and both took place in tropical South America. While the pair Rhathymini‐Epicharisremained mostly associated within this region,Centrisand their cleptoparasites expanded their distribution to other parts of Neotropical and Nearctic regions in independent range expansions events. In all cases, host range shifts preceded the cleptoparasite shifts. Main ConclusionThe biogeographical history of cleptoparasitic oil bees and oil‐collecting hosts is generally congruent in time and space. Events of range expansion mainly occurred in the more species‐rich lineages of cleptoparasites. Range shifts in cleptoparasites followed the distribution of their hosts and coincided with the distribution of oil‐producing plants visited by the host bees. Our results broaden our understanding of the complex biogeography of interacting partners and on how changes in host distributions may impact cleptoparasitic bees. 

Abstract AimAn antitropical pattern is characterized by the occurrence of closely related taxa south and north of the tropics but absent or uncommonly represented closer to the equator, in contrast to most taxa, which tend to have their highest diversity in the tropical regions. We investigate the antitropical distribution of eucerine bees with the aim of contributing to the characterization and understanding of this pattern. LocationAll continents except Antarctica and Australia. TaxonEucerine bees (Hymenoptera: Apidae: Eucerinae). MethodsWe carried out phylogenomic dating under two different clock models and used multiple strategies to vary matrix composition, evaluating the overlapping of divergence times estimated across models using Bhattacharyya coefficients. Lastly, we reconstructed the biogeographic history of eucerine bees using a Bayesian implementation of the DEC model. ResultsEucerinae is estimated to have started diversifying during the Palaeocene, with all its tribes originating during the Palaeocene/Eocene transition in southern South America. At least two range expansions happened into North America before the full closure of the Isthmus of Panama. We show that divergence between closely related groups with disjunct distributions would have happened in periods when the climate favoured the expansion of open habitats and became isolated when the forests were re‐established. Main conclusionsWe describe the early diversification of eucerine bees, revealing an intimate association with southern South America. Events of range evolution of Eucerinae were likely affected by periods of global cooling and aridification, and palaeoclimatic and vegetational conditions probably have been more relevant to the formation of the antitropical distribution of Eucerinae than the consolidation of the Isthmus of Panama connecting the Americas. We also demonstrate that most uncertainty in divergence time estimation is not due to the amount of molecular data being used, but more likely other factors like fossil calibrations and violations of clock models.