NEW AND REFINED DATA ON ORDOVICIAN TO EARLY SILURIAN CRINOIDS YIELD A REVISED MACROEVOLUTIONARY HISTORY

Our knowledge of Phanerozoic biodiversity is largely derived from literature-based compilations of taxonomic first and last occurrences. There are many sources of both error and bias that can alter biological signals in global compilations of this sort. Nevertheless, these compilations are commonly interpreted as relatively undistorted records of biological phenomena. This is at least partly justified by assuming that errors are randomly distributed so as to introduce no systematic distortion of biological patterns. Unfortunately, very few studies have tested this assumption. Here we compare a new, comprehensive global database of Ordovician and Early Silurian crinoids to Sepkoski’s global genus compendium, which spans the putative end-Ordovician mass extinction. Approximately 44 % of the crinoid genera that were resolved to substage in Sepkoski’s compendium are inaccurate. Further, taxonomic and stratigraphic errors among Ordovician and Early Silurian crinoids in Sepkoski’s data are non-randomly distributed among substages, resulting in incorrect richness patterns and significantly overestimated rates of extinction among crinoids during the end-Ordovician. These results suggest that the role of extinction during the end-Ordovician has been exaggerated for crinoids and that taxonomic and stratigraphic range data in literature-based compilations need to be verified prior to being used to understand macroevolutionary changes.