COMPARATIVE LONGEVITIES OF MARINE GENERA AND THEIR CONSTITUENT SPECIES AFTER MASS EXTINCTIONS: THE WHOLE DOES NOT NECESSARILY REFLECT ITS PARTS

Most investigations of Phanerozoic marine diversification and extinction have been conducted at the genus level or higher, and relationships of genus-level patterns to underlying, species-level dynamics have been open to speculation. We might expect a species-level diversity trajectory to mirror its genus-level counterpart, albeit with differences in volatility related to the hierarchical nature of classification. Further, if we find that a temporal trajectory is, indeed, similar at the species and genus levels, we might speculate that the macroevolutionary properties of the genera included in the analysis are, on average, shared by their constituent species.

Previously, we demonstrated that genus cohorts originating after several mass extinctions beginning with the end-Permian event were longer-lived, on average, than their counterparts originating at other times. Here, we analyze data extracted from the Paleobiology Database (PaleoDB; http://paleodb.org/), which includes species-level identifications, to determine whether this pattern is also exhibited by species and, if so, whether this is attributable directly to long-lived species belonging to genera that were also long-lived.

Relative to our previous analysis, conducted at sub-stage temporal resolution using Sepkoski’s global compendium, the present genus-level analysis, conducted at the stage level, is more nuanced, but nevertheless exhibits enhanced mean longevities associated with several mass extinctions. Inevitably, the species-level trajectory is more muted, but exhibits features observed at the genus-level, including mass-extinction-associated peaks, and parallel, longer-term trends. There is no indication, however, that the species contributing to longevity peaks belonged to the genera contributing to the genus-level peaks. More broadly, there is no tendency for long-lived genera throughout the study interval to be associated disproportionately with long-lived constituent species. This is not to say that observable genus-level properties, such as geographic occupancy, are not tied in meaningful ways to aggregate properties of their constituent species. But there is no evidence that genera are constituted disproportionately of species that share their relative temporal characteristics.