Bayesian evolutionary rate estimate and dated trees
A root-to-tip regression analysis was undertaken using TempEst to establish the temporal signal of datasets II-VII. The GTR+G+I model was selected as the evolutionary model using JmodelTest, v. 2.1.7, while the mean evolutionary rate and the dated trees were made using a Bayesian Markov Chain Monte Carlo approach implemented in BEAST, v.1.10.4 (Suchard et al., 2018). Two coalescent priors (constant population size and exponential growth) and strict vs relaxed molecular clock models were tested by means of path sampling (PS) and stepping stone (SS) sampling. Convergence was assessed by estimating the effective sampling size (ESS) after a 10% burn-in, using Tracer software (http://tree.bio.ed.ac.uk/software/tracer/), and accepting ESS values of 200 or more. Uncertainty in the estimates was indicated by 95% highest posterior density (95% HPD) intervals. Posterior probability was used as statistical support for specific clades and clusters. The trees were summarized in Tree Annotator, and the tree with the maximum product of posterior probabilities (maximum clade credibility or MCC), after a 10% burn-in, chosen.