Recent divergence and lack of shared phylogeographic history characterize the diversification of neotropical savanna birds

Aim: Neotropical savanna birds occur north and south of, but mostly not in the Amazon Basin, except for a few isolated savanna patches. Here, we investigate the phylogeography of 23 taxa of Neotropical savanna birds co-distributed across multiple isolated savanna patches to assess to what extent these species have a shared history of spatial diversification. We explore the role of the forested Amazon Basin as a vicariant barrier separating northern and southern populations, particularly focusing on the role of the coastal savannas of Amapá as a potential corridor of gene flow between northern and southern populations. Location: Neotropical savannas. Taxon: Aves. Method: We employ 775 mtDNA samples of 24 co-distributed savanna bird taxa from all major savanna patches in South America to infer phylogeographic patterns. For this purpose, we use 24 genomic samples (UCEs) of a subset of 12 taxa in addition to the mtDNA samples to estimate timing of divergence across the Amazon Basin. We use phylogeographic concordance factors (PCF) to assess the level of phylogeographic congruence across co-distributed taxa. Finally, we assess to which level physical distance drives genetic structuring by estimating isolation-by-distance (IBD) effects. Results: We find that although the study taxa generally do not share similar diversification patterns geographically, many have at least two distinct genetic groups, one north and one south of the Amazon Basin, that have only recently diverged. The timing of divergence between both areas is generally centered in the late Pleistocene, but somewhat variable, indicating there is no single vicariant event responsible for driving diversification. Main conclusions: Variability in divergence times indicates that landscape processes have not led to shared phylogeographic responses, which indicates a relatively minor role for vicariance. Shallow divergences suggest that Neotropical grassland habitats may have recently been more connected or that gene flow has played an important role. We did not find evidence of a single dominant corridor of dispersal between savannas north and south of the forested Amazon Basin.