The only exception to Miller and Foote's basic finding was an interval of heightened extinction in the run-up to the big event at the end of the Permian, during which the extinction rate was higher in epicontinental seas. But this interval had already been fingered by previous researchers as one in which there was a big sea-level decline, so Miller and Foote's finding bolsters the view that a drop in sea level was uniquely important as a cause of extinction in that interval.
For their analysis, Miller and Foote looked at several hundred thousand occurrences of fossils throughout the world assembled from the Paleobiology Database and used paleogeographic maps on which they had delineated the boundaries of epicontinental seas to determine whether individual occurrences were from epicontinental seas or the open ocean. "We then determined from these occurrences whether a given genus had a statistically significant tendency to occur in one setting more often than in the other, and if so, it was classified as open-ocean 'loving' or epicontinental-sea 'loving,'" Miller explains. "Our assessments of extinction and origination rates for the two settings were based on these assignments."
Anticipating that biologists might ask if the results primarily reflected a property unique to bivalve mollusks, which comprise a plurality of the data, Miller and Foote ran their analysis again using everything but the bivalves and obtained the same pattern. Similarly, because epicontinental seas tend to be associated with tropical latitudes dominated by carbonate (biologically derived) sediments and open-ocean settings occur more frequently in nontropical latitudes dominated by terrigenous (land-derived) sediments, Miller and Foote extracted subsets of the data limited only to tropical carbonates and nontropical terrigenous sediments and then split them into epicontinental seas versus open oceans. In nearly all cases they got t
|Contact: Wendy Beckman|
University of Cincinnati