Reticulation, data combination, and inferring evolutionary history: an example from Danthonioideae (Poaceae).

We explore the potential impact of conflicting gene trees on inferences of evolutionary history above the species level. When conflict between gene trees is discovered, it is common practice either to analyze the data separately or to combine the data having excluded the conflicting taxa or data partitions for those taxa (which are then recoded as missing). We demonstrate an alternative approach, which involves duplicating conflicting taxa in the matrix, such that each duplicate is represented by one partition only. This allows the combination of all available data in standard phylogenetic analyses, despite reticulations. We show how interpretation of contradictory gene trees can lead to conflicting inferences of both morphological evolution and biogeographic history, using the example of the pampas grasses, Cortaderia. The characteristic morphological syndrome of Cortaderia can be inferred as having arisen multiple times (chloroplast DNA [cpDNA]) or just once (nuclear ribosomal DNA [nrDNA]). The distributions of species of Cortaderia and related genera in Australia/New Guinea, New Zealand, and South America can be explained by few (nrDNA) or several (cpDNA) dispersals between the southern continents. These contradictions can be explained by past hybridization events, which have linked gains of complex morphologies with unrelated chloroplast lineages and have erased evidence of dispersals from the nuclear genome. Given the discrepancies between inferences based on the gene trees individually, we urge the use of approaches such as ours that take multiple gene trees into account.