Abstract
The lemurs of Madagascar are a unique radiation of primates that show an extraordinary diversity of lifestyles, morphologies and behaviours. However, very little is known about the relative antiquity of lemuriform clades due to the lack of terrestrial fossils for the Tertiary of Madagascar. Here, we employ a Bayesian method to estimate divergence dates within the lemuriform radiation using several unlinked gene loci and multiple fossil calibrations outside the lemuriform clade. Two mitochondrial genes (cytochrome oxidase II and cytochrome b), two nuclear introns (transthyretin intron 1 and von Willebrand factor gene intron 11) and one nuclear exon (interphotoreceptor retinoid binding protein, exon 1) are used in separate and combined analyses. The genes differ in taxon sampling and evolutionary characteristics but produce congruent date estimates. Credibility intervals narrow considerably in combined analyses relative to separate analyses due to the increased amount of data. We also test the relative effects of multiple vs. single calibration points, finding that, when only single calibration points are employed, divergence dates are systematically underestimated. For the mitochondrial DNA data set, we investigate the effects of sampling density within the mouse lemur radiation (genus Microcebus). When only two representative species are included, estimated dates throughout the phylogeny are more recent than with the complete‐species sample, with basal nodes less affected than recent nodes. The difference appears to be due to the manner in which priors on node ages are constructed in the two analyses. In nearly all analyses, the age of the lemuriform clade is estimated to be approximately 62–65 Ma, with initial radiation of mouse lemurs and true lemurs (genus Eulemur) occurring approximately 8–12 Ma. The antiquity of the mouse lemur radiation is surprising given the near uniform morphology among species. Moreover, the observation that mouse lemurs and true lemurs are of similar ages suggests discrepancies in rates of morphological, behavioural and physiological evolution in the two clades, particularly with regard to characteristics of sexual signalling. These differences appear to correlate with the nocturnal vs. diurnal lifestyles, respectively, of these two primate groups.