• Dick writes about Earth and planetary science for Science magazine.

Mammoth-Killer Impact Rejected

30 August 2010 4:26 pm

Over the past 3 years, a loosely confederated group of researchers has argued that an asteroid or comet struck North America about 13,000 years ago, wiping out the woolly mammoth, the giant sloth, and other large animals. Experts say they have shot down most of the supposed evidence, but one finding remained: nano-scale diamond crystals that could have formed only under the extreme pressure of an impact. Now, a group of experts has dismissed this evidence as well, putting what many see as the final nail in the coffin of the mammoth-killer impact.

In an article published online 27 September 2007 in the Proceedings of the National Academy of Sciences (PNAS), a team headed by nuclear chemist Richard Firestone of Lawrence Berkeley National Laboratory in California claimed that a 4-kilometer-wide object struck an ice sheet that covered northern North America 12,900 years ago, at the onset of a cold snap known as the Younger Dryas. That would have set off continent-engulfing wildfires, killed off mammoths and other large animals, and disrupted the reigning Paleo-Indian culture. Such an event would have left its mark in all sorts of debris left over from that time, but analyses for geochemical impact markers—such as the exotic element iridium, which is delivered only by comets and asteroids—have come up empty-handed. "The geochemical story is finished, it's over," says impact geochemist Philippe Claeys of the Free University of Brussels.

That left a couple of other lines of evidence, including microscopic, chemically distinctive mineral grains and magnetic spherules in sediment. They are out there, says archaeologist Todd Surovell of the University of Wyoming in Laramie, but they are scattered through the geologic record. They are not concentrated in 12,900-year-old sediments, he says, as an impact would have left them.

Another problem: Experts have not found the charred, carbon-rich traces of impact-triggered wildfires across North America reported by proponents in the 2007 PNAS paper. In a paper published earlier this summer, a group of researchers said certain honeycombed spherules—touted as fire residue —were merely fungus and bug poop.

"None of these [impact proponents] has spent their lives looking at carbon material in sediments," says paleobotanist and fire scientist Andrew Scott of Royal Holloway, University of London, in the United Kingdom. "I've spent 35 years looking at it. I see no evidence of an exceptional fire event across North America."

In the eyes of impact specialists, that leaves nano-size bits of diamond called lonsdaleite, or hexagonal diamond, as a potential impact marker, something proponents claimed to have found in a PNAS paper published online 20 July 2009. Physicist Tyrone Daulton of Washington University in St. Louis and colleagues, who were not involved with proponents, searched sediment from the time of the supposed impact for hexagonal nanodiamond—which is formed in nature only in impacts—using transmission electron microscopy. They found none.

"I'm convinced there's no [hexagonal] diamond present," says Daulton. Instead, the group unearthed aggregates of sheetlike forms of carbon. "If you don't look too closely at it, you could convince yourself it is [hexagonal diamond]," says Daulton. "Theirs was a gross misidentification as lonsdaleite." The team reports its findings online today in PNAS.

Proponents of a Younger Dryas impact disagree. "The Daulton et al. claim that we have misidentified diamonds is false and misleading," says archeologist Douglas Kennett of the University of Oregon, Eugene. He argues that Daulton and others may not have sampled the correct layer of sediment and that sample processing may have destroyed key evidence—in this case, the nanodiamonds. "There's been a real problem of data quality," Kennett says.

Counter-criticism aside, outsiders are walking away from the mammoth-killer impact in increasing numbers. "I spent 16 months in the lab and found very little evidence to support their hypothesis," says Surovell. "I have other things to worry about."

Posted In: