Palm diversification in two geologically contrasting regions of western Amazonia

Aim: The subduction of the Nazca Plate and the eastward propagation of the Andean orogenic wedge in western Amazonia caused the formation of arches or ridges that have influenced the modern configuration of the upper Amazon drainage and the diversification of biota. We used a lineage of 15 palm species (Astrocaryum sect. Huicungo, Arecaceae) to test two biogeographical hypotheses for lowland plants: (1) that vicariance resulted from tectonically mediated geographical barriers (population contraction), and (2) that recurrent dispersal events (population expansion) produced geographical isolation and subsequent speciation. Location: Rain forests of South America. Methods: A total of 78 palm individuals were collected in the field, from which five chloroplast and two nuclear DNA fragments were sequenced. We reconstructed a Bayesian dated phylogeny and inferred the demographic history. We used a Bayesian phylogeographical spatial diffusion approach to propose a model of colonization. Results: We found a phylogeographical break at c. 5° S between two main clades with crown ages of c. 6.7 and 7.3 Ma located in the Fitzcarrald Arch (FA) and the subsiding northern Amazonian foreland basin (NAFB), respectively. These diversification times were close to the emergence of the FA in the late Miocene, and the coeval development of the transcontinental modern drainage and sedimentation plain of the NAFB. As expected for the recurrent-dispersal hypothesis, lineage delimitations were spatially inconsistent with the location of rivers or ridges, and we found some evidence of past ancestral population expansion supported particularly by the chloroplast sequences. Main conclusions: Our results support the biogeographical scenario whereby recurrent dispersal into western Amazonia produced spatial isolation of populations, followed by speciation within two areas of contrasting geological activity: tectonic uplift in the FA versus subsidence in the NAFB. We did not test and cannot rule out ecological speciation within western Amazonia or at finer spatial scales.