Goner. This gorgonopsian carnivore disappeared as extinction accelerated in the late Permian, well before the main extinction event.

A Prolonged Demise

Although asteroid impacts have become popular scapegoats for mass extinctions, new fossil evidence suggests that our planet's greatest extinction may have resulted from a far more protracted crisis: thousands of years of volcanic eruptions.

Most scientists now agree that an asteroid or comet wiped out the dinosaurs 65 million years ago, paving the way for mammal evolution. So some have speculated that a similar impact 251 million years ago—at the boundary between the Permian and Triassic periods (P-T)--may have triggered the even bigger extinction that gave dinosaur ancestors their start.

But new fossil finds challenge this theory. In a paper published online this week by Science), paleontologist Peter Ward of the University of Washington, Seattle, and colleagues report on the skulls of 126 reptiles and mammal-like reptiles they collected during the past 7 years across the P-T boundary in the Karoo Basin of South Africa.

To determine the relative ages of the fossils, the team analyzed the rocks that held the fossils for changing carbon isotopic composition and records of Earth's flip-flopping magnetic field. The researchers found that, after 10 million years or more of relative stability, Permian creatures suffered more rapid extinction in the time during which the last 50 meters or so of Permian rock were deposited before Triassic rocks appear. As a result, Ward guesses that the extinction-driven decline of Permian taxa might have gone on for as long as 1 million years or as little as 10,000 years. Then a burst of extinctions occurred at the P-T boundary, lasting perhaps 10,000 years.

"We can definitely see it's different from the [dinosaur extinction]," says Ward. No decline in the diversity of dinosaurs preceded their demise or the extinction of other species; they all had a very bad day when the asteroid hit, and then they disappeared. The gradual decline before the boundary makes Ward "think there was no impact at all" at the P-T.

Paleontologist Desmond Maxwell of the University of the Pacific in Stockton, California, agrees that the gradual decline on land—which the new Karoo data strongly support—points to a noncatastrophic cause. Not that life would have been comfortable late in the Permian. In one scenario, eruption of the lavas of the great Siberian Traps at the time of the P-T boundary (Science, 21 November 2003, p. 1315) would have poisoned the air and water with acid and alternately chilled the world with a sun-screening haze and baked it with the greenhouse gas carbon dioxide. Hard times indeed.

Related sites
Peter Ward's homepage
More information on the P-T extinction