Bayesian relaxed molecular clock method was used. Ages in bold indicate the study in which the corresponding order was more extensively sampled. The result shows that increasing the sampling size pushes the ages towards older estimates. In this analysis rodents could not be taken into account because of sampling incongruences between the two studies.
a Paleontological time constraints used as calibrations.
b The results of MODELTEST were used to define the partitioning; the three studies are therefore directly comparable.