ALBUQUERQUE—After combing through layers of ancient lake sediments, paleoecologist Jacquelyn Gill of the University of Wisconsin–Madison says her team has found no evidence to support a controversial comet theory for an ice age extinction event.

"There's no physical trend to suggest that there was an impact event," Gill said Tuesday at the Ecological Society of America meeting held here this week. "If there was an impact event...it's not having the ecological effects [previously] suggested."

In 2007 Richard Firestone at the Lawrence Berkeley National Laboratory and colleagues published evidence suggesting that a comet exploding in the atmosphere 12,900 years ago near the Great Lakes set off massive fires in North America. These fires supposedly led to the rapid disappearance of the continent's Clovis culture as well as megafauna including mammoths, ground sloths and 33 other large mammal genera.

But skeptics point to the fact that no associated impact craters have been found, and evidence for continental forest fires and a rapid decline in human populations is sketchy. Besides, if such an episode had occurred, small mammals and birds somehow survived. (Another recent study has called into question the likelihood of comet impacts being responsible for more than one extinction event during Earth's history.)

Gill and her team decided to look for hints of the comet's impact not on land but in three lakes in Indiana and Ohio, where pollen and minerals have settled daily, creating an ecological record dating back millennia. She scoured core samples for evidence of ash, charcoal, magnetic grains, tiny silicate spheres, and elements such as titanium and chromium that could be associated with impact events. She did not look for rare earth minerals like iridium, which other researchers rely on as a signature of impacts.

The team failed to find a consistent signal that would indicate that a single catastrophe occurred around 12,900 years ago. At one lake, titanium actually decreased at the same time charcoal was increasing. "It's clearly not an impact event," Gill said.

Gill also questioned the view that animals died off right at the time of the proposed impact event.

Fungal spores called Sporormiella, associated with the dung of large mammals, actually begin to decline 14,600 years ago, soon after the end of the last ice age. The spores wink out of the record at one lake 13,600 years ago and only recover in the past few centuries with the rise in cattle grazing. At the same time, pollen from the plants that megafauna munched on—ash, ironwood and hop hornbeam—starts piling up, suggesting the plants' growth was no longer being kept in check by leaf-loving large mammals.

More recent megafauna fossils, dated between 13,600 and 12,900 years ago, were likely the last hangers-on as the Clovis people decimated species with their characteristic spears, Gill says.

But Firestone was not swayed by the new study, which he says "adds nothing to the argument." He says that he would not expect to find much evidence in lake sites because magnetic material would rust away, while the tiny spheres float and would not collect on the lake bottom. For him, the data still points toward an extraterrestrial impact. "As far as I'm concerned the debate is settled," he said.