Mass extinctions are, by definition, dramatic events. But the one that punctuated the Permian Period some 250 million years ago was especially spectacular, extinguishing nearly 70 percent of all terrestrial species and 95 percent of marine-dwelling forms. Despite the magnitude of the die-off, it's a whodunit that has long vexed geologists. Numerous culprits have been proposed--from climate change to large-scale volcanism to an asteroid impact--but exactly how these hypotheses stack up against one another has been unclear. To that end, a new analysis of current models may prove insightful. A report detailing the findings appears in the current issue of the Proceedings of the National Academy of Sciences.

To evaluate the various theories, Robert Berner of Yale University focused on a certain geochemical signature associated with the extinction event--namely, a sharp drop in the carbon-13 isotope level relative to that of isotopically light carbon. He then assessed the predictions these hypotheses make concerning the global carbon cycle. Neither the release of toxic amounts of carbon dioxide from the ocean nor volcanic degassing could alone account for the observed carbon-13 drop, he found. But another suspect in the die-off--a massive release of methane from methane hydrates stored in sediments--did exactly that.

Berner notes, however, that although the methane scenario best explains the carbon-13 measurements, it cannot explain the higher levels of atmospheric carbon dioxide present at that time. Thus although methane may well have had a hand in the killing, it does not appear to have acted alone. In conclusion, Berner proposes that the extinction may have resulted from a combination of an asteroid impact, methane release and volcanic CO2 release.