Calibration and error in placental molecular clocks: a conservative approach using the cetartiodactyl fossil record.

The nature of the molecular and fossil record and their limitations must be ascertained in order to gain the most precise and accurate evolutionary timescale using genetic information. Yet the majority of such timescales are based on point estimates using fossils or the molecular clock. Here we document from the primary literature minimum and maximum fossil age estimates of the divergence of whales from artiodactyls, a commonly used anchor point for calibrating both mitogenomic and nucleogenomic placental timescales. We applied these reestimates to the most recently established placental timescale based on mitochondrial rRNA and several nuclear loci, and present an attempt to account for both genetic and fossil uncertainty. Our results indicate that disregard for fossil calibration error may inflate the power of the molecular clock when testing the time of ordinal diversification in context with the K-T boundary. However, the early history of placentals, including their superordinal diversification, remained in the Cretaceous despite a conservative approach. Our conclusions need corroboration across other frequently used fossil anchor points, but also with more genetic partitions on the linear relationship between molecular substitutions and geologic time. Copyright 2004 The American Genetic Association