The 'species problem' and testing macroevolutionary hypotheses

Species lists change for a variety of reasons, including new information and preferences for different species concepts. Uncertainty over species numbers is potentially damaging to tests of proposed correlates of species richness, particularly if taxonomic changes are biased toward some clades over others. We investigate the effects of this error and bias by testing the same suite of macroevolutionary hypotheses in seven different arrangements of primate taxonomy. This is the first time that the effects of the ‘species problem’ have been systematically investigated in this way. Primates are an excellent model system for examining the effects of taxonomic uncertainty: species numbers have doubled in the past two decades, with the fastest growth in the Neotropics. We found that different variables were significantly associated with species richness in each taxonomic arrangement. However, there were no significant differences among taxonomies in the regression slopes for any predictor variable. We found no tendency for significant correlations to occur in taxonomies with more species, suggesting that the results cannot be explained by a lack of power in the smaller taxonomies. The findings are discussed with reference to the wider implications for testing macroevolutionary hypotheses.