Comment on Rieux and Balloux: calibration from tip-dating can compromise topological accuracy and evolutionary inference

Abstract

We contribute to the recent review of Rieux & Balloux, 2016, Mol. Ecol., 25, 1911 on inferences from tip-dated phylogenies by developing their discussion on the influence of population size (N_e ) under panmixia for the estimation of substitution rate (μ). We highlight how phylogenetic trees inferred with tip-dated sequences under large panmictic N_e tend to erroneously enforce an age-based coalescent pattern on the posterior distribution of trees, which in turn results in systematically inflated estimates of μ. We discuss the consequences of this and suggest how to accommodate the issue in the short term and long term.