Molecular dates and the mammalian radiation.

In a perspectives article, Bromham et al.1 raise some important concerns about the use of both molecular and paleontological data in assessing the timing of diversification of extant mammalian orders. In Box 1, they describe well the need to differentiate crown (or more generally, nodebased) taxa from stem-based taxa. Unfortunately, their Fig. 2 is not clearly explained and thus confuses these kinds of taxa. As they note, the paleontologically based orders of extant placental mammals (except, perhaps, Insectivora) are recognized as appearing soon after the Cretaceous–Tertiary (K–T) boundary (thick lines in their Fig. 2). These dates are based on apomorphyor node-based intraordinal diversifications for the order in question. In contrast, the extensions into the Late Cretaceous of clades shown in their Fig. 2 (the thin lines), estimated by molecular data, are interordinal separations. Thus, in at least this comparison, the molecular data indicate nothing about ordinal origination and diversification, but rather argue only that stem-based clades extend into the Late Cretaceous. Furthermore, the possible Cretaceous record of primates they mention in their text is a single tooth originally assigned to the primatomorph Purgatorius, which was discovered at a site now regarded as Paleocene in age2. The authors’ biogeographical assessment of fossil taxa requires updating. The possible placental for the early Cretaceous of Australia is now regarded by most as symmetrodont3 or early therian. Thus, although marsupials are known for the early Eocene of Australia4, non-chiropteran placentals do not appear until the Pliocene5. In South America, all definite pre-Tertiary mammals are non-therians, with both marsupials and placentals appearing only after the K–T boundary6. As the authors note, the Late Cretaceous of North America and Asia have a good record of mammals, but except for, perhaps, Insectivora, no modern orders of placentals are known. In fact, the latest Cretaceous record7 is better known than the earliest Paleocene8. Europe is not well known but echoes what is known in Asia for placentals9. As the authors also note, Africa is a cipher. Unless, however, one wishes to make the unsubstantiated argument that all 18 orders of extant placentals arose in Africa, the claim is not valid that the biogeography of placentals is too poor to help in deciphering ordinal appearances.