Recent analysis of South African rocks reveals that rivers suddenly became clogged with sediments 251 million years ago, indicating Earths worst mass extinction wiped out many trees and other plants that held soil in place.

Peter Ward, the studys first author, said he believes a huge comet or asteroid walloped Earth to cause the mass die-off at the end of the Permian Period and dawn of the Triassic although his study does not say so. Another scientist estimates the object was 9 to 12 miles (15 to 20 kilometers) wide.

Previous research showed the extinction wiped out nearly 90 percent of sea species and 70 percent of vertebrate animal species on land. That made it far worse than the Cretaceous-Tertiary extinction of the dinosaurs and many other creatures 65 million years ago an event often blamed on the impact of a perhaps 6-mile- (10-kilometer-) wide asteroid.

In the September 8 issue of the journal Science, Ward and colleagues concluded the Permian-Triassic catastrophe also stripped Earth of many rooted plants, triggering severe erosion.

As a result, sedimentary rocks from that time show that large meandering rivers throughout South Africas Karoo Basin took on a braided, multichannel appearance, resembling streams in areas devastated by Mount St. Helens big eruption or areas logged by clear-cutting.

Braided rivers are common in Alaska and mountainous areas where glaciers and streams erode rock quickly, filling streams with sediments.

The Permian-Triassic switch from meandering to braided streams once was thought due to mountain-building uplift and subsequent erosion. But mountain-building episodes in the Karoo Basin do not coincide with the Permian-Triassic extinction, Ward said.

"Lets pretend were looking for habitable planets from space," he said. "All you need to do is see a meandering river and you know you have higher plants, so go there."

The new study is the latest in a series showing the Permian-Triassic catastrophe was quick at least in geological terms and extremely nasty.

But it has not settled debate among those who advocate various theories of what caused the extinction: an object whacking Earth, floods of lava from the Siberian Traps, climate change, and/or deadly radiation from a nearby supernova or other cosmic explosion.

In July, Doug Erwin of the Smithsonian Institutions National Museum of Natural History published a study in Science in which marine rocks from China revealed the Permian-Triassic extinction happened in less than 160,000 years.

And in Julys issue of the journal Geology, a study of seafloor rocks now in the Austrian Alps concluded the extinction happened in less than 60,000 years and perhaps in less than 8,000 years, said the main author, planetary scientist Michael Rampino of New York University.

Because the rock layers do not permit finer dissection of time, the findings are consistent with the extinction being triggered by an impact that happened during "a single bad day," Erwin said.

"The mass extinction at the end of the Permian Period was catastrophic and sudden," ravaging sea and land life, Rampino said. "The only thing we know of that can cause an extinction like this is a large impact of an asteroid or comet. But we still havent found conclusive evidence that an impact occurred."

Rampino said the object "would have to be bigger one and a half to twice as big" as the 6-mile- (10-kilometer-) wide asteroid usually blamed for the dinosaurs extinction.

In 1997, University of Oregon paleontologist Gregory Retallack reported finding elevated iridium levels and shocked quartz crystals in Permian-Triassic rocks telltale signs of an impact.

"Unfortunately, we have not found a good candidate crater," he said.

Ward said a comet made of ice "would be almost invisible geologically."

Last April, Australian scientists said they found a 75-mile- (120-kilometer-) wide crater in western Australia that might be from an impact that caused the Permian-Triassic extinction or a later extinction at the end of the Triassic Period roughly 200 million years ago.

But Rampino said the age of that crater is so poorly known that "it is impossible to tie that impact to this [Permian-Triassic] extinction."

Nevertheless, Retallack favors impact as the cause, perhaps with the impact triggering an undersea release of methane that robbed the oceans of life-sustaining oxygen. He says the impact also may have triggered massive eruptions from volcanic vents named the Siberian Traps.

"There is no evidence of an impact" at Permian-Triassic time, Erwin said. "So while the data are consistent with an impact, there is nothing that tells us it was an impact," and massive volcanic eruptions may be a more likely cause.

Rampino said the Siberian volcanic eruptions lasted hundreds of thousands of years, so "if the extinctions were gradual over a half million years, we might suspect volcanism or changes in climate caused by volcanic eruptions. But the fact the extinctions were so sudden and catastrophic argues against a volcanic interpretation."