A BIOGEOGRAPHIC HISTORY OF THE ROSACEAE (ANTHOPHYTA: EUDICOTYLEDONAE) USING FOSSIL-INFORMED PHYLOGEOGRAPHIC METHODS

Understanding the geographic origins of biodiversity is one of the goals of historical biogeography. Phylogenetic data and comparative phylogenetic methods are used to predict the location where clades diversify, but inferences made by these methods can be misled by paleoclimatic variation, extinction, and other factors that such methods cannot incorporate. Fossil assemblages can give us real paleodistributions, but are themselves snapshots of paleobiogeography, only allowing us to bracket where and when diversification happened.

The Rosaceae (Anthophyta: Eudicotyledonae) today have a worldwide distribution and are common in temperate and subpolar climates, and in tropical habitats at high altitude. Additionally, the Rosaceae constitute one of the more common families in the Cenozoic fossil record. We use the R package BioGeoBEARS (Matzke 2013) to compare the phylogeography of the living Rosaceae to the family’s fossil record compiled from the literature in order to understand the timing and geographic origins of the Rosaceae.

Although details of ancestral state reconstructions vary as to the paleodistributions of different genera, there is overwhelming agreement that fossil-informed reconstructions of the family point to the continent of origin being Eocene North America. While the plant fossil record is heavily biased in favor of North America and Europe, the fact that the Rosaceae have a largely temperate distribution can be interpreted to suggest that it is unlikely that they would have originated on a tropical continent in the Eocene. By contrast, application of the same methods of ancestral geographic state reconstruction without using fossil data provide an incoherent origin of the family. As such, fossil data continues to prove essential in reconstructing ancestral states, in this case of geographic distributions.