The status of platyrrhine phylogeny: A meta-analysis and quantitative appraisal of topological hypotheses

Abstract

Phylogenetic or species trees reflect the branching process of lineages and have direct and indirect interest for several branches of evolutionary anthropology. Estimating phylogenetic trees is a necessary first step toward understanding the factors responsible for the ecological and phenotypic diversification of a primate clade. The platyrrhines have become well known as a phylogenetic challenge. Since the 1990s, platyrrhine phylogenetic studies have increasingly analyzed DNA sequences, or other molecular datasets. Several researchers have claimed with confidence that platyrrhine phylogenetic history has been ‘resolved’ using these molecular data, but the concordance among these studies has never been quantified. Here, we perform a meta-analysis of published platyrrhine trees using topological information and multivariate methods. Specifically, we examine the claim that platyrrhine phylogeny has been determined and explore the relationships between phylogenies and dataset types used for phylogenetic inference (nuclear DNA, mtDNA, Alu sequences, morphology or mixed data). We compare topologies summarizing 31 major neontological studies of the platyrrhines produced since 1975. The analysis reveals that major disparities are rather common among the hypotheses regarding the higher-level relationships of platyrrhines. We also find that the global concordance that appears to emerge at the generic level is less impressive when one looks more finely at particular relationships. Moreover, correspondence among trees appears to be related to the ‘type’ of dataset analyzed, which suggests that the biological properties of distinct datasets have an inherent influence on the likelihood of producing similar reconstructions of phylogenetic relationships. This serves to remind us that the main questions surrounding the phylogeny reconstruction program begin with experimental design, for both molecular and morphological datasets. Thus, previous claims that platyrrhine genus-level topology have been ‘resolved’, or that calibrated molecular trees are sufficiently accurate representations of phylogenetic history that they overpower morphological interpretations of fossils, must be considered premature.