Open vegetation (campinas and campinaranas) associated with white sand patches occurs in the form of islands in a forested matrix throughout the Amazon basin. Bird species restricted to these habitats have patchy distributions, although connectivity may have been influenced by past glacial cycles as a result of the substitution of forest by savanna. Because these landscape changes are a matter of debate in the history of Amazonia, we studied the diversification of Xenopipo atronitens, a white sand specialist, aiming to infer the effects of past climate changes. The split of Xenopipo atronitens from its sister species, Xenopipo uniformis, may be related to Tepuis erosion and retreat of escarpments during the Miocene, or to a dispersal event. Compared with birds from terra firme forest, X. atronitens has low genetic structure. Low levels of unidirectional gene flow were found from the Guyana Shield to adjacent areas. Demographic expansion starting approximately 25 kyr BP was detected for some populations and is probably related to the Last Glacial Maximum and subsequent climate improvement. Landscape genetic analyses indicate that the forested (terra firme) matrix acts as a barrier for the dispersal of X. atronitens. The results of the present study indicate that glacial cycles have deeply influenced Amazonian biogeographical history, demonstrating a complex interaction between forest and nonforest habitats during the Pleistocene. © 2013 The Linnean Society of London, Biological Journal of the Linnean Society, 2013, 110, 60–76.

For two centuries evolutionary biologists have sought to explain elevated biodiversity in the Neotropics. Although different process are known to be important, it is still not uncommon for researchers to emphasize a single mechanism. Recently, arguments have highlighted the importance of dispersal shaping community structure and evolution across the region. We examine this hypothesis by visualizing spatial variation in community structure for the majority of South American passerines (Aves) across the northern half of South America. By sampling over a contiguous area we show how community structure varies widely across Amazonia and surrounding regions. Our results support a combination of processes including: the inability of species to disperse across geographic barriers, Andean uplift, and variation in habitat type. Although dispersal is a factor, our results emphasize a lack of dispersal, driven primarily by features of the landscape, coupled with historical changes in climate to be important drivers of Neotropical diversity.

No abstract