Estimating species trees: methods of phylogenetic analysis when there is incongruence across genes.

Discord among the gene trees of multilocus data has motivated the development of phylogenetic approaches that account for gene-tree heterogeneity in the estimation procedure. Rather than equating a gene tree with the phylogenetic history, the new approaches explicitly consider the relationships between gene trees and the underlying history of species divergence, providing direct estimates of species trees (Fig. 1). The inherent appeal of these approaches is 2-fold. Incorporating information contained in the distribution of gene trees not only extracts phylogenetic signal, but modeling the relationship between the gene trees embedded in a species tree also reveals the biological processes that have influenced the diversification history and shaped organismal genomes. In contrast, ignoring the variance in genealogical histories (e.g., concatenating loci into a single supermatrix) disregards an inescapable biological reality—gene trees differ for a variety of reasons (reviewed in Maddison 1997; Degnan and Rosenberg 2009). As such, when the natural variation in gene trees is not taken into account during phylogenetic estimation, the reliability of inferences from such approaches is drawn into question (Degnan and Rosenberg 2006; Kubatko and Degnan 2007; Huang and Knowles 2009), historical scenarios with recently diverged taxa that have not reached reciprocal monophyly has become intractable (Carstens and Knowles 2007), and interpretation of the support for bipartitions across taxa is problematic (Mossel and Vigoda 2005).