Leveraging evolutionary history to advance knowledge about the compositional patterns of Amazonian Flora

Abstract

The longitudinal tree species diversity gradient is a striking biodiversity pattern in Amazonia - western forests are more diverse than eastern forests. Western Amazonia was dominated by a mega-wetland ecosystem for at least 10 Ma during part of the Neogene. Since floods impose eco-physiological restrictions for trees, such wetland habitats can be considered non-ideal habitats for tree species to thrive. The rise of the Andean cordillera has gradually filled huge sedimentary basins in western Amazonia, creating elevational heterogeneity close to the Andean forelands, and shaping the climate of most of the South America continent. The geology of eastern Amazonia is much older, characterized by crystalline rock basements of the Brazilian and Guiana shields. Meanwhile, the climate of eastern Amazonia is driven by the seasonal arrival of winds bringing humidity from the Atlantic Ocean. Beyond the species richness gradient, the tree species composition of forests also changes in a longitudinal fashion, allowing the definition of bioregions with distinct tree species assemblages. Nevertheless, existing classifications of Amazonian vegetation are based mainly on physiographic - edaphic terrain attributes and structural - phenological attributes of canopy trees; only recently a floristic support for such classifications has been described. The upland non-flooded forests (terra-firme) are the floristic core of Amazonia, while the seasonal flooded forest types and the low stature forests growing on white-sand soils are marginal habitats. The interplay of four key processes influence the origin and maintenance of biological communities. Speciation adds new species into a continental-to-regional pool of species; dispersal allows species to spread among regions; selection modulates populations' ability to colonize and persist into different and changing habitats; and drift exerts a stochastic feature to speciation, dispersal, and species biotic interactions, via modulation of species population sizes, for example. To better understand the relative role of such processes on the assembly of biological patterns we must undertake cross-scale studies recognizing both the temporal and spatial scales involved. The aim of this project is to evaluate the tree species compositional pattern in Amazonia and the association of this pattern with the evolutionary history of species in local and regional communities, as well as the development of the Amazonian landscape. We will do this by using the most comprehensive database of tree species assemblages amassed for all of Amazonia coupled with a temporally calibrated and up-to-date phylogenetic hypothesis for the evolutionary history of Angiosperm tree species occurring in that region. Specifically we will analyze evolutionary dissimilarities and congruence among i) existing floristic compositional gradients, ii) the derived floristic bioregions, and iii) the four major forest habitats of Amazonia, as defined by their occupied area and richness. By doing this we expect to shed light on processes creating biodiversity patterns through the region and on the evolution of the most tree species rich biome on Earth. (AU)