On the origin and diversification history of the African genus Encephalartos.

• Our biogeographic analysis suggests that the genus Encephalartos may have diverged around 9 Ma from southern Africa and then colonized the region through dispersal. • The east-central-west African clade might have diverged from the southern African clade through vicariance. • The eastern rift system in eastern Africa and the West African Dahomey Gap may act as geographical barriers driving the vicariance observed. • Although there was rapid species accumulation 2.6 Ma, the shift in diversification may not be as dramatic as initially thought. Of the 10 known genera of cycads, the genus Encephalartos is one of the two endemic genera to Africa. Although a number of studies have explored the taxonomic relationships within the genus as well as its diversification history, we still have limited understanding of its historical biogeography. In the present study, using five gene markers, we first reconstructed a complete phylogeny of the genus and its historical biogeography based on S-DIVA model. We then fitted the COMET model to the phylogeny to test for significant shifts in evolutionary events that may have driven the current diversity in the genus. Overall, our phylogeny is well supported and reveals two major clades following species geographic origins, one southern African clade and one east-central-west African clade. Our biogeographic analysis suggests that the genus may have diverged around 9 Ma from southern Africa (100% probability) and then colonized the region through dispersal. Although the origin of the east-central-west African clade is uncertain (50% probability), the clade may have diverged from the southern African clade through vicariance. We suggest that the eastern rift system in eastern Africa and the west African Dahomey Gap – a dry corridor that breaks pan-African tropical rainforest into different blocks – are geographical barriers limiting dispersal into the rest of the continent whilst, at the same time, promoting the vicariance signature observed. Although most species accumulated in the last 2.6 My, there were no significant shifts in any of the evolutionary events, suggesting that an overall constant-rate diversification model is best suited for the genus and that the rapid species accumulation during the Pliocene–Pleistocene transition may not be as dramatic as initially thought. [ABSTRACT FROM AUTHOR]