PROXIES REVISITED: THE RELATIONSHIP BETWEEN GENERA AND SPECIES IN THE STUDY OF PHANEROZOIC DIVERSITY

Ever since Valentine’s seminal work of the 1960s, there has been considerable interest in relationships among Phanerozoic diversity patterns at different hierarchical levels, from phylum to species. Diversity trajectories at or above the order level differ systematically from those at the family level and below; the former stabilized early, reflecting the “front-loading” of the origination of major body plans, whereas the latter exhibited significant post-Paleozoic increases, related to the elaboration of many of these body plans. It has since been standard practice to use families and genera as proxies for species in large-scale studies of diversification and extinction. This rests on the assumption, implicit or explicit, that similar patterns would be observed at the species level if only adequate data were available.

But there continues to be some skepticism about the validity and utility of families and genera as proxies for species. We therefore think it worthwhile to reconsider this issue, in part to help clarify the actual intent of the use of higher taxa as proxies for species. The utility of genera as proxies, and their limitations in conveying meaningful information about their constituent species, can be shown with data from The Paleobiology Database. As demonstrated previously, we observe significantly higher mean longevities of stage-level genus cohorts originating in association with several mass extinctions, and many of these peaks are also seen at the species level. We see a similar pattern if we evaluate the maximum geographic ranges attained by genera and species in these cohorts.

Clearly, the genus-level pattern is a viable proxy for the species level. That said, when we study the characteristics of individual genera and their constituent species, we find, as have other researchers, that genus-level attributes do not carry over to their constituent species: long-lived or geographically wide-ranging genera do not simply consist of species that are long-lived or wide-ranging.

That one level is used as a proxy for another in studies of Phanerozoic diversity has never been meant to imply that characteristics of individual taxa match those of their daughter taxa, but that aggregate properties of sets of taxa at different levels agree and reflect common macroevolutionary and macroecological processes.