Molecular systematics, phylogeography, and population genetics of Xiphorhynchus (aves: Dendrocolaptidae) in the Amazon basin

Abstract

Among those few hypotheses amenable to falsification by phylogenetic methods concerning the diversification of the Amazonian biota, three can be singled out because of their verifiable predictions: the riverine barrier, the gradient, and the basal trichotomy hypotheses. I used phylogenetic and population genetics methods to reconstruct the diversification history of the genus Xiphorhynchus (aves: Dendrocolaptidae) in Amazonia. First, I estimated the phylogeny of the entire genus Xiphorhynchus to test a key prediction of the gradient hypothesis; secondly, I documented phylogeographies of a superspecies associated with upland forest (X. spixii / elegans) and two species linked to floodplain forest (X. kienerii and X. obsoletus) to evaluate predictions of the riverine barrier and basal trichotomy hypotheses. The phylogeny estimated for the genus Xiphorhynchus falsified an anticipated sister relationship between floodplain and upland forest species, as predicted by the gradient hypothesis. Phylogeographic and population genetics analyses of the upland forest superspecies (X. spixii / elegans), and floodplain forest species (X. kienerii and X. obsoletus) indicated that predictions of the riverine barrier hypothesis hold only for populations of the upland forest superspecies separated by rivers located on the Brazilian shield; in contrast, rivers located in western Amazonia did not represent areas of primary divergence for populations of X. spixii / elegans. As expected, populations of the floodplain forest species showed high levels of gene flow and no geographic structure throughout the entire Amazon basin, a pattern consistent with their expected capacity to overcome riverine barriers. In agreement with predictions of the basal trichotomy hypothesis, populations of the X. spixii / elegans superspecies found on the Brazilian shield were basal in the phylogeny, exhibiting xi some population genetics attributes typical of old populations having reached equilibrium. In contrast, populations found in western Amazonia were more recently derived and experienced a dramatic recent population expansion, probably colonizing the area from the geologically older Brazilian shield. The data presented herein supported important predictions of the basal trichotomy and riverine barrier hypotheses, indicating that they are not mutually exclusive, and may together account for the diversification of the genus Xiphorhynchus in Amazonia at different temporal and geographical scales.