Quantitative analysis of the timing of the origin and diversification of extant placental orders

Fossil evidence is consistent with origination and diversification of extant placental orders in the early Tertiary (Explosive Model), and with the possibility of some orders having stem taxa extending into the Cretaceous (Long Fuse Model). Fossil evidence that 15 of 18 extant placental orders appeared and began diversification in the first 16 m.y. of the Cenozoic is, however, at odds with molecular studies arguing some orders diversified up to 40 m.y. earlier in the Early Cretaceous (Short Fuse Model). The quality of the fossil record was assessed by tabulating localities of all mammals in the last 105 m.y. Global locality data (except Africa) for 105 m.y. of eutherian evolution indicate discernible biogeographic patterns by the last 15 m.y. of the Cretaceous. Eutherian genera increase from 11 in latest Cretaceous to 139 in earliest Tertiary, although both are represented by about 50 localities. Yet even in the Late Cretaceous of North America and Asia where eutherians are abundant, none of the 18 extant orders are definitely known. A series of Monte Carlo simulations test whether the rapid appearance of most mammalian orders is statistically significant, and if so, whether it is a radiation event or an artifact of a limited fossil record. Monte Carlo tests affirm that the clustering of appearances in the early Cenozoic is statistically significant. Quantitative analysis of the locality data suggests that the number of genera described is a function of the number of localities sampled. In contrast, the number of orders is not a simple function of localities and thus does not appear to be limited by localities. A second set of Monte Carlo simulations confirms that the increase in orders cannot be explained by the limited number of localities sampled. Even for best-fit simulations, the observed pattern of ordinal appearances is steeper than expected under a variety of null models. These quantitative analyses of the fossil record demonstrate that the rapid ordinal appearances cannot be ascribed to limited Late Cretaceous sample sizes; thus, early Tertiary ordinal diversification is real. Although the fossil record is incomplete, it appears adequate to reject the hypothesis that orders of placentals began to diversify before the K/T boundary. © 2001 Plenum Publishing Corporation.