Genetic variation of the most abundant forest‐dwelling rodents in Central Africa ( Praomys jacksoni complex): Evidence for Pleistocene refugia in both montane and lowland forests

International audience; AimWe investigate the Plio‐Pleistocene evolutionary history of one of the most abundant rodents in Afrotropical forests. Specifically, we ask how their diversification was influenced by climate change, topography and major rivers.LocationTropical Africa: Lower Guinea (including Cameroon volcanic line; CVL), Congolia, Albertine Rift (AR), Kenyan highlands (KH).TaxonMurine rodents of the Praomys jacksoni complex.MethodsWe used 849 genotyped individuals to describe the overall diversity and spatial genetic structure across a majority of their known distribution area. The combination of one mitochondrial and three nuclear markers was used to infer dated phylogenies using Bayesian and maximum likelihood approaches. Genetic structure was further assessed by multispecies coalescent species delimitation. Current and past distributions of particular taxa were predicted using environmental niche modelling.ResultsThe complex is composed of five major genetic clades (proposed species). Two of them are restricted to specific habitat types (either montane forests of AR or wetlands in lowland forests along the Congo River), three others have wide geographic distributions and lower levels of ecological specialization. The earliest divergence is dated to the Plio‐Pleistocene boundary and is in accordance with the separation of AR forests and Guineo‐Congolian forests. Further diversification of the complex is associated with Pleistocene climate changes. Relatively stable refugia of suitable climatic conditions were identified in lowland Congolia (for two species currently distributed only in lowland forests) as well as in montane forests of CVL, AR, KH (playing the role of reservoirs of diversity). Large rivers, especially the Congo River, are important barriers to gene flow for most taxa, but probably were not the primary cause of differentiation.Main conclusionsThe evolutionary history of the complex was primarily affected by Pleistocene climate changes and diversification in forest refugia. There is little support for ecological parapatric speciation or the riverine barrier hypothesis.