Multilocus phylogeny and species delimitation within the genus Glauconycteris (Chiroptera, Vespertilionidae), with the description of a new bat species from the Tshopo Province of the Democratic Republic of the Congo

The genus Glauconycteris Dobson, 1875 currently contains 12 species of butterfly bats, all endemic to sub-Saharan Africa. Most species are rarely recorded, with half of the species known from less than six geographic localities. The taxonomic status of several species remains problematic. Here, we studied the systematics of butterfly bats using both morphological and molecular approaches. We examined 45 adult specimens for external anatomy and skull morphology, and investigated the phylogeny of Glauconycteris using DNA sequences from three mitochondrial genes and 116 individuals, which in addition to outgroup taxa, included nine of the twelve butterfly bat species currently recognized. Four additional nuclear genes were sequenced on a reduced sample of 69 individuals, covering the outgroup and Glauconycteris species. Our molecular results show that the genus Glauconycteris is monophyletic, and that it is the sister-group of the Asian genus Hesperoptenus. Molecular dating estimates based on either Cytb or RAG2 data sets suggest that the ancestor of Glauconycteris migrated into Africa from Asia during the Tortonian age of the Late Miocene (11.6-7.2Mya), while the basal diversification of the crown group occurred in Africa at around 6 +/- 2Mya. The species G.superba is found to be the sister-group of G.variegata, questioning its placement in the recently described genus Niumbaha. The small species living in tropical rainforests constitute a robust clade, which contains three divergent lineages: (i) the poensis group, which is composed of G.poensis, G.alboguttata, G.argentata, and G.egeria; (ii) the beatrix group, which contains G.beatrix and G.curryae; and (iii) the humeralis group, which includes G.humeralis and a new species described herein. In the poensis group, G.egeria is found to be monophyletic in the nuclear tree, but polyphyletic in the mitochondrial tree. The reasons for this mito-nuclear discordance are discussed.