Madagascar harbors four large adaptive radiations of endemic terrestrial mammals: lemurs, tenrecs, carnivorans, and rodents. These rank among the most spectacular examples of evolutionary diversification, but their monophyly and origins are debated. The lack of Tertiary fossils from Madagascar leaves molecular studies as most promising to solve these controversies. We provide a simultaneous reconstruction of phylogeny and age of the four radiations based on a 3.5-kb data set from three nuclear genes (ADRA2B, vWF, and AR). The analysis supports each as a monophyletic clade, sister to African taxa, and thereby identifies four events of colonization out of Africa. To infer the time windows for colonization, we take into account both the divergence from the closest non-insular sister group and the initial intra-insular radiation, which is a novel but conservative approach in studies of the colonization history of Madagascar. We estimate that lemurs colonized Madagascar between 60 million years ago (Mya) (split from lorises) and 50 Mya (lemur radiation) (70-41 Mya taking 95% credibility intervals into account), tenrecs between 42 and 25 Mya (50-20 Mya), carnivorans between 26 and 19 Mya (33-14 Mya), and rodents between 24 and 20 Mya (30-15 Mya). These datings suggest at least two asynchronous colonization events: by lemurs in the Late Cretaceous-Middle Eocene, and by carnivorans and rodents in the Early Oligocene-Early Miocene. The colonization by tenrecs may have taken place simultaneously with either of these two events, or in a third event in the Late Eocene-Oligocene. Colonization by at least lemurs, rodents, and carnivorans appears to have occurred by overseas rafting rather than via a land bridge hypothesized to have existed between 45 and 26 Mya, but the second scenario cannot be ruled out if credibility intervals are taken into account.

Platyrrhine primates and caviomorph rodents are clades of mammals that colonized South America during its period of isolation from the other continents, between 100 and 3 million years ago (Mya). Until now, no molecular study investigated the timing of the South American colonization by these two lineages with the same molecular data set. Using sequences from three nuclear genes (ADRA2B, vWF, and IRBP, both separate and combined) from 60 species, and eight fossil calibration constraints, we estimated the times of origin and diversification of platyrrhines and caviomorphs via a Bayesian relaxed molecular clock approach. To account for the possible effect of an accelerated rate of evolution of the IRBP gene along the branch leading to the anthropoids, we performed the datings with and without IRBP (3768 sites and 2469 sites, respectively). The time window for the colonization of South America by primates and by rodents is demarcated by the dates of origin (upper bound) and radiation (lower bound) of platyrrhines and caviomorphs. According to this approach, platyrrhine primates colonized South America between 37.0 +/- 3.0 Mya (or 38.9 +/- 4.0 Mya without IRBP) and 16.8 +/- 2.3 (or 20.1 +/- 3.3) Mya, and caviomorph rodents between 45.4 +/- 4.1 (or 43.7 +/- 4.8) Mya and 36.7 +/- 3.7 (or 35.8 +/- 4.3) Mya. Considering both the fossil record and these molecular datings, the favored scenarios are a trans-Atlantic migration of primates from Africa at the end of the Eocene or beginning of the Oligocene, and a colonization of South America by rodents during the Middle or Late Eocene. Based on our nuclear DNA data, we cannot rule out the possibility of a concomitant arrival of primates and rodents in South America. The caviomorphs radiated soon after their arrival, before the Oligocene glaciations, and these early caviomorph lineages persisted until the present. By contrast, few platyrrhine fossils are known in the Oligocene, and the present-day taxa are the result of a quite recent, Early Miocene diversification.