The Day Everything Died

The big knockdown fight in science these days is a debate about a cataclysm that occurred 250 million years ago

By Karen Wright
Apr 28, 2005 5:00 AMNov 12, 2019 4:22 AM
Geologist Luann Becker argues that 250 million years ago an asteroid six miles wide crashed into the seafloor off the coast of Australia. She thinks the fallout from the impact provoked the greatest of the five known mass extinctions. | Illustration by Mondolithic Studio


Sign up for our email newsletter for the latest science news

One of the boldest assertions ever published in the scientific literature started with a single modest observation. In the late 1970s, geologist Walter Alvarez of the University of California at Berkeley and his father Luis, a Nobel Prize–winning physicist, found an unusual chemical signal in an ancient layer of Italian clay. The clay was enriched in iridium, a rare metal that comes mostly from meteorites, interplanetary dust, and other cosmic debris. The iridium spike appeared in sediments 65 million years old, at the so-called K-T boundary between the Cretaceous and the Tertiary periods. It coincided with the demise of the dinosaurs.

 Contamination from local sources or a glitch in the iridium-counting machine could have explained the finding. But the Alvarezes found an even bigger spike in another Cretaceous-Tertiary deposit in Denmark. Their interpretation, published in 1980, was heretical.

The clay at the K-T boundary was high in iridium, they said, because it was made of the ash and dust from a six-mile-wide asteroid that had crashed into Earth with the energy of

100 million megatons of TNT. The impact instantly killed every living thing within hundreds of miles. The animals that weren’t incinerated or gassed by fumes froze or starved to death soon after, when dust kicked up by the impact blotted out the sun for more than a year, killing plant life around the globe. Dinosaurs were only the most conspicuous casualty of an epic disaster that eradicated half of all the species on Earth.

“Their idea was met by instant ridicule and derision by most geologists and paleontologists,” recalls paleontologist Michael Benton of the University of Bristol in a recent book. It took another decade of evidence gathering, including the documentation of an impact crater off the Yucatán Peninsula, for the impact theory to win acceptance, he notes. Now “ ‘Extraterrestrial Cause for the Cretaceous-Tertiary Extinction’ is considered . . . one of the most influential publications in earth sciences in the twentieth century,” Benton writes in When Life Nearly Died: The Greatest Mass Extinction of All Time.

The book, however, is not about the Cretaceous-Tertiary impact. And the death of the dinosaurs was not the greatest mass extinction of all time. That superlative belongs to a more severe crisis at the P-T boundary, between the Permian and Triassic periods. Fossil records show that about 250 million years ago, 90 percent of the species on Earth were snuffed out in an abrupt event that spanned the globe. The extinction occurred a couple hundred million years before the dinosaurs died out, so its causes, like its sediments, are buried more deeply. No one has even come close to proving what happened.

But in the past five years, one scientist has dared to implicate a familiar culprit: an asteroid or comet comparable in size and speed to the K-T perpetrator. Geologist Luann Becker of the University of California at Santa Barbara has published a series of papers describing rocks from China, Japan, and Antarctica that have subtle and sometimes unorthodox signs of an impact, including extraterrestrial gases trapped in microscopic carbon cages and minerals deformed by shock waves. Last year, her research team delivered the coup de grâce: evidence of an impact crater off the northwest coast of Australia, hidden beneath two miles of sediment on the ocean floor.

Like the Alvarezes’ theory, Becker’s Permian extinction work has been greeted with hostility. It prompted a vitriolic exchange in the journal Science and a showdown at last December’s annual meeting of the American Geophysical Union. NASA has launched an investigation to explore Becker’s claims, and some of Becker’s peers are second-guessing any findings that fit her interpretation—even if the findings are their own. In January, for example, geologist Peter Ward of the University of Washington in Seattle revised his earlier thesis that the extinction had occurred suddenly, documenting new fossil successions that suggest a more prolonged die-off. Geologist Greg Retallack of the University of Oregon in Eugene is retracting evidence of impact-shocked minerals at the P-T boundary he reported in the late 1990s.

Compared with those researchers, Becker is young and relatively inexperienced, but she cannot be dismissed as a fringe figure. Her academic credentials are impeccable, and she publishes in the country’s most prestigious science journal with experts from top-flight universities as coauthors. Although highly qualified scientists often disagree, some insiders are baffled by the heat of this particular debate. Retallack, for one, still believes that an impact scenario is credible. “I don’t know why people are trashing Luann,” he says.

There could be two reasons: She’s wrong, or she’s right. If she’s wrong, say her detractors, her crusade is drawing focus away from investigators looking at other, more likely scenarios, such as the eruption of hundreds of volcanoes in prehistoric Siberia. “All this putative impact stuff is muddying the waters,” grumbles geophysicist Jay Melosh of the University of Arizona at Tucson.

If she’s right, then a newcomer who wound up studying the Permian extinction and its “putative impact” has bested paleontological stalwarts who have devoted decades to solving the puzzle of mass extinctions. “I’ve been all over the world looking for shocked minerals at the P-T boundary, and I haven’t found any at all,” says geologist Michael Rampino of New York University.

It’s the objection of a seasoned scientist, but it could just as easily be the complaint of a runner-up. Becker could be driving a discovery as profound as any in earth science, or she could be courting career-dashing disgrace. Depending on whom you ask, the cause of the greatest extinction of all time has been either finally identified or hopelessly obscured.

To understand the fuss over Becker’s claims, it helps to know that ideas about the Permian extinction have long been subject to scholarly caprice. Two centuries ago, the very concept of extinction itself was considered scandalous. The great thinkers of the early 1800s only grudgingly acknowledged that the fossils of mastodons, mammoths, and giant ground sloths unearthed in the previous century had no living counterpart remaining on Earth. Then they portrayed extinction as a gradual event. The preeminent British geologist Charles Lyell maintained that iguanodons, ichthyosaurs, and pterodactyls might stage a comeback if hospitable habitats and climates returned. Lyell also came out against any notion of sudden, indiscriminate cataclysm in the history of life. He branded such catastrophism muzzy-headed voodoo science.

The geologists who defined the fossil hallmarks of the Permian in the 1840s must have feared Lyell’s opprobrium, for they failed to mention the signs of mass extinction at the end of that period. It seems unlikely that they simply overlooked it. The Permian extinction obliterated ecosystems as complex as any on Earth today. On land, 10-foot-long saber-toothed reptiles succumbed, and grazing, root-grubbing, and insect-eating lizards vanished, along with the plants and bugs they ate. In the ocean, reefs teeming with life were reduced to bare skeletons. The Permian even finished off the lowly trilobite—perhaps the one celebrity species of the predinosaur era.

Even when geologists finally acknowledged these disappearances in the fossil record, they decided that, while thorough, the Permian die-off had been prolonged. The best estimates had it taking about 10 million years, which doesn’t seem terribly cataclysmic. A lot can slowly go wrong in 10 million years. The climate can grow too hot or too cold; sea levels can rise or fall; the amount of oxygen in the ocean or the atmosphere can change. Most plants and animals are exquisitely sensitive to such shifts, and many might not be able to adapt. But they would die out one by one over millennia, at such a stately pace that a hypothetical human would hardly notice.

Thus Lyell’s gradualism continued to prevail, and catastrophic change stayed taboo for most of the 20th century. That’s one reason the Alvarezes’ K-T impact theory seemed so radical, even in 1980. It invoked the sort of deus ex machina that Lyell had disparaged, and it conjured improbable images of instantaneous apocalypse.

But for once, an idea that plays well in the tabloids also turned out to be true. Emboldened by that example, geologists began to revisit other scenes of carnage in rock beds around the globe. In addition to the Cretaceous and Permian die-offs, they had identified three other episodes of mass extinction in the past 500 million years (a die-off is considered a mass extinction when 50 percent or more of all species are extirpated from the fossil record). Some specialists could not help but hope that a single uncomplicated cause might explain all five of Earth’s great extinctions. “A few years ago, we thought maybe they’re all impacts,” says Rampino. The pendulum Charles Lyell had pulled far to one side swung back just as far to the other. For a few years, it stayed there. Catastrophe became all the rage.

With the paradigm shifting, geologists admitted they could not prove that the Permian extinction had been gradual after all. Evidence to the contrary began to surface. In the early 1990s, geologists examined a rock section in China that bore the critical fossils of the P-T boundary interleaved with ashy volcanic layers suitable for isotopic dating. Called the Meishan section, this felicitous stratigraphy—along with advances in radiometric methods—allowed researchers to time the extinction better than ever before. In 1998 a Chinese and American group headed by geochronologist Sam Bowring of MIT nailed the date of the Permian extinction at 251 million years ago. A carbon signature in the Meishan section suggested the catastrophe had lasted at most 165,000 years. In other words, it had happened two orders of magnitude faster than the 10-million-year textbook estimate.

Armed with the new time line, fossil experts began weighing in. In a 2000 survey of 333 marine species in the Meishan section, paleobiologist Doug Erwin of the National Museum of Natural History in Washington showed that the extinction happened abruptly in the oceans. That same year Peter Ward documented a sudden die-off of vegetation on land in present-day South Africa. Lines of evidence were converging, and figures kept ratcheting downward. The duration of the Permian extinction went from hundreds of thousands of years to tens of thousands and, finally, to just thousands. Although they couldn’t resolve time in terms of days, weeks, or months, many experts came to believe that the whole doomed Permian assemblage—flora, fauna, and foraminifera—might have bought it overnight. The abrupt demise made an impact scenario seem even more plausible. Then Luann Becker came along.

In 1991 Becker was working toward her Ph.D. at the Scripps Institution of Oceanography in La Jolla, California, when her adviser, Jeffrey Bada, showed her an article about the discovery of a new form of carbon molecule called a fullerene. Fullerenes are hollow, closed lattices shaped like nanoscale soccer balls or geodesic domes (they’re also known as buckyballs, after Buckminster Fuller, the dome’s inventor). They had first been synthesized in the laboratory in 1985, but some scientists thought they might also be made in space, in the furnaces of stars.

If fullerenes are star dust, Bada reasoned, they could be among the cosmic debris that has fallen to Earth more or less constantly since the birth of the planet. The biggest payloads, of course, would arrive via meteorites. But would they survive an impact? Becker—who was planning to be an environmental geologist—got swept up in Bada’s enthusiasm. The two decided to search for fullerenes near known impact craters. They soon found them, in 1993, at an impact site in Canada nearly 2 billion years old. The molecular cages at the so-called Sudbury site might have been forged on Earth from the intense heat and pressure of the impact or in a common forest fire. Yet in their hollow centers the fullerenes held captive helium gas with an unearthly composition that was distinctive of some meteorites and interplanetary dust.

“We were absolutely taken aback,” Becker says. “What was in these little buckyballs was an extraterrestrial signature.”

Becker next succeeded in isolating fullerene molecules directly from meteorites. Encouraged that she had found a new way to trace impact events, she joined with geochemist Robert Poreda of the University of Rochester in New York, who had helped develop the technique to find trapped fullerene gases, to look for buckyballs at the sites of mass extinctions. First they found some at the K-T boundary. Then they found some at the P-T, in rocks from the Meishan section and at another site called Sasayama in Japan. In the first of several controversial papers, Becker and her colleagues reported that the P-T fullerenes contained trapped helium and argon gases with extraterrestrial compositions. The helium content of the Sasayama fullerenes, for example, is more than 50 times higher than background levels.

“Thus, it would appear that [extraterrestrial] fullerenes were delivered to Earth at the P-T [boundary], possibly related to a cometary or asteroidal impact event,” Becker and her colleagues concluded. “Our results are consistent with recent paleontological studies that now point to a very rapid extinction event.”

Becker’s fullerene report received guarded praise. True, the notion of alien gases trapped in microscopic carbon cages for millions and even billions of years strains credulity, especially when you imagine the force of the impact that supposedly delivered them. But by the time Becker’s work appeared, impact geologists were sorely in need of alternative tracers. Their two favorites from the K-T days—iridium spikes and shocked quartz—hadn’t turned up in any incriminating abundance in the rocks associated with other mass extinctions. So, fullerenes from outer space? Why not? “They looked like a possible winner in terms of a signature of an impact,” says Rampino, a coauthor of that first report.

Two years later, Becker and geochemist Asish Basu of the University of Rochester published another paper with still more unconventional evidence for a Permian impact. Becker’s group claimed to have found dozens of actual fragments of meteorites in rocks from the P-T boundary in Antarctica. That evidence is unconventional because meteoritic remains are so easily turned to dust. If they had somehow avoided being incinerated on entry or pulverized on impact, they would have disintegrated in a year of heavy rain—long before geologic processes could fold them into native rock. Less than half a dozen meteorite fragments have been found intact in rock layers the world over.

Becker’s fragments are intact and unweathered, although they are supposed to be a quarter of a billion years old. “The meteorite fragments . . . are so well preserved that their preservation must be due to rather unusual circumstances,” the authors themselves concede. But as far as they are concerned, “the two largest mass extinctions in Earth history at the K-T and P-T boundaries were both caused by catastrophic collisions with chondritic meteoroids.”

Seven months later, Science published Becker’s report of the proposed impact site. This time Becker, Basu, and four other coauthors described a submarine hump called Bedout High that is buried in ocean sediments 100 miles off the northwest coast of Australia. Geologist John Gorter of ENI Australia was surveying for offshore oil there when he spotted the plateau on a seismic profile of the seabed in the late 1990s. Becker hadn’t learned of Gorter’s find until 2002, but when she called him he said he could also get her rock cores drilled from the top and the flank of the structure’s uplifted bull’s-eye. “I got my rear end over there and started looking at those samples,” she says.

In those seafloor samples, Becker’s team reports finding shocked and melted minerals and glass that could be produced only by the intense heat and pressure of a bolide, or meteoric, crash. The researchers dated one of the mineral grains and got a familiar number: 250 million, give or take a few million. They say a gravity model of the site, a kind of topographical map of buried geologic structures, looks much like the gravity model of Chicxulub, the K-T impact structure. Becker and company say the signs of impact at Bedout are compelling enough to warrant further scrutiny. And they are getting it.

Maybe it’s because none of her coauthors are Nobel Prize winners. Maybe it’s because she and her colleagues are the only ones who know how to find a fullerene. Maybe it’s because evidence of an extinction four times older than the K-T is that much more difficult to find and interpret. For whatever reason, Becker’s latest paper—“as spectacular and annoying to some people as the Alvarez paper in the 1980s,” she says—has fared no better than that historical example. Except there’s no sign of eventual acceptance—even from a former coauthor and fan of the impact theory.

“The dates are not unequivocally 250 million years, the shocked minerals don’t look like shocked minerals, and the gravity anomaly doesn’t look like the gravity anomaly you’d get from an impact,” Rampino observes. “There’s no evidence of a crater, let alone a crater of that time period.”

Becker’s critics have aired their grievances in caustic missives to Science. A group led by British sedimentologist Paul Wignall of the University of Leeds writes about examining rocks cored from basin sediments 600 miles south of Bedout. “At no level in the core . . . is there evidence for a layer of impact ejecta or a tsunamite,” the authors contend. Becker’s team responds that Wignall’s core hasn’t even been proved to include material from the Permian-Triassic boundary. Another group headed by geochronologist Paul Renne of the Berkeley Geochronology Center in California notes that the gravity map of Bedout bears no resemblance whatsoever to another confirmed impact site called Vredefort. Becker admits that her gravity signature is a little irregular: In the map’s caption she calls it “significantly reduced and more subdued” than Chicxulub. But she says the reference to Vredefort—which is plainly visible on the surface of the South African desert—just shows how irrational her critics have become: “Comparing a crater that’s exposed at the surface to something that’s been buried under four kilometers of debris? Give me a break. I mean, hello!”

Renne and other investigators also charge that Becker’s supposedly shocked minerals don’t have the telltale patterns of impact-induced features: narrow, parallel bands crisscrossing at various angles, like a microscopic tartan weave. Rampino says Becker’s group offers nothing nearly as persuasive as the shocked minerals found in K-T boundaries across the globe in the years following the Alvarezes’ breakthrough paper. He still remembers the day in 1983 when he saw the first slides of K-T shocked quartz at a meeting. “I went down the hall to see it, and I came back convinced,” he recalls. “Had there been a picture like that in [Becker’s] paper, these questions wouldn’t have come up at all.”

Additional questions surround Becker’s impact glass, which Renne and others believe could be volcanic. In a recent online analysis, earth scientist Andrew Glikson of Australian National University asserts that Bedout is probably just a buried, burned-out volcano. The oil prospectors who originally collected the Bedout cores also assumed that the rocks were volcanic. But everyone thought Chicxulub was a volcanic crater, too, Becker says, until tests done in the late 1980s proved otherwise.

“It’s tectonically and geologically impossible for it to be [a volcanic crater],” she claims. “At the time this thing formed, it was in the middle of a basin that was nowhere near a subduction zone—it was nowhere near the kind of geologic activity that would cause a volcano to form.”

Even the fullerene tracers that were once warmly received have come under attack, because geochemist Ken Farley of Caltech in Pasadena found no helium in P-T rocks from Meishan when he tried to replicate Becker’s work.

“Nobody can reproduce her results,” says Melosh, who maintains that Becker’s means of isolating fullerenes could also be used to synthesize them. “Possibly she’s fooling herself because she’s making the fullerenes she’s detecting.”

To each of these charges, Becker has detailed and spirited retorts. She points out that Farley, for example, did not examine the same Meishan rock samples she did and that he looked for helium in bulk rather than isolating fullerenes first and then looking for gases trapped within them. “We’ve got everybody hounding us because it’s a spectacular claim,” says Becker. “They feel threatened. Why else would they make such absurd statements?”

At this point, it is reasonable to wonder how an extraterrestrial bolide the size of Mount Everest could plow into the planet without leaving an unambiguous trace. The answer is, in part, that 250 million years of heat and pressure can deal its own damage to rocks. The ocean floor, for example, recycles on a tectonic conveyor belt every 200 million years or so, erasing all signs of disturbance. Bedout sits just offshore on a continental shelf; otherwise its features—whether volcanic or extraterrestrial—would be history. Still, even that relatively stable continental crust erodes, subsides, uplifts, and deforms over the millennia, obscuring its original mien.

So scientists are left to reconstruct epochal events from nearly inscrutable remains. Retallack and others, for example, have found an iridium blip at the P-T boundary, but it’s one-tenth as large as the iridium spike reported by the Alvarezes and others at the K-T. That could imply a modest-size meteorite, not big enough to cause a worldwide extinction. But some meteorites contain very little iridium, and comets, which are mostly ice, don’t have any. If an impactor landed in the deep ocean, it wouldn’t create much shocked quartz either, because the ocean floor has less quartz in it than continental crust. If a king-size comet landed in the deep ocean, it would be like stabbing a man with an icicle: a murder with a weapon that vanishes.

There are other suspects. One is the Siberian Traps, a million-year-long volcanic eruption that flooded five time zones of Russia with basalt lava more than a mile deep. Over the past decade, ever more sophisticated dating of the ancient basalt has shown that the lava bed could be about the same age as the extinction, and recent studies have revealed that it covered twice as much area as previously supposed. “Knowing that this province was probably twice as big as we thought has a visceral effect,” says Renne. “We’re staring one of the significant coincidences [of the P-T boundary] right in the face.”

A million years of eruption might release massive clouds of sulfurous gases and carbon dioxide. “It probably wouldn’t have been a lot of fun to breathe,” says Renne. The oxidation of coal beds beneath the magma could release methane as well. The sulfur could produce torrents of acid rain, the carbon dioxide and methane could lead to rapid greenhouse warming, and life on Earth just might not have been worth living for a while.

The Siberian Traps hypothesis has been a favorite among Permian experts when times get tough with the impact theory. There has even been speculation that an impact caused the eruptions. But no one has described a convincing mechanism for an impact-induced eruption, especially not one lasting a million years. And some geologists now question whether the eruptions could have been disastrous enough to account for a global extinction. Basaltic eruptions are mild, like those in Hawaii, not spectacular like the pyrotechnic Mount Saint Helens.

“There are no big explosions,” says Melosh. “It’s very bad if you happen to be right under the lava.” As for exterminating life elsewhere on the planet, “I don’t think it’s enough.”

The one element of the Permian mystery that is certain is that something did indeed claim the lives of 9 out of every 10 species. The fossil signature of the Permian remains the only obvious signal of what went down 250 million years ago, but it, too, resists deciphering. Ward’s latest findings are a case in point: Though his 2000 report on South African plant fossils showed signs of an abrupt extermination at the P-T boundary, his new analysis of animal fossils suggests that a gradual extinction preceded that ultimate burst of fatalities.

“Different organisms have different reactions to different stresses, so if you knew the sequence of mortality you could get a handle on the sequence of events,” says Renne. Unfortunately, current dating methods aren’t precise enough to determine the exact order in which species disappear from the fossil record.

It does seem certain that the extinction was followed by an extraordinarily long recuperation, called a survival interval, of at least 4 million years. During that time, the fossil record shows that a handful of plants and creatures held on for dear life: humble things such as clams, “well adapted to living in lousy environments,” says Erwin. According to Benton, it would be fully 100 million years before the planet recovered the same level of biodiversity it had hosted before the end-Permian crises.

What then follows is oddly reminiscent of Lyell’s dubious notions on the resurrection of extinct forms. Paleontologists have documented a number of plants and animals that disappear at the end of the Permian, stay gone for millions of years, and re-emerge in the middle Triassic. They call these Lazarus species.

It’s difficult to build a convincing theory of mass extinction around such data. “I think there was definitely an impact,” says Retallack, “but I don’t think it caused the extinction necessarily.” Instead, Retallack imagines that the impact released methane stored in the seafloor when it struck. The methane essentially suffocated life. “The actual mode of death would’ve been coughing up a blood-specked frothy sputum,” says Retallack.

Rampino detects a note of desperation in such scenarios. “The search is widening from the standard cast of characters,” he says. “They’re pulling suspects off the street. It would be so much easier if we could just find a big crater in the ground, or a big, smoking volcano.”

But Renne, for one, would not mourn the loss of the impact theory or any other pat explanation for biocatastrophe. “Why should every major extinction have the same cause?” he says. “It would just be too tidy.”

There is nothing tidy about the p-t impact theory as it stands today. “Everybody is waiting for an ending to the story,” says Becker.

Well, not everybody. Some mass-extinction veterans, weary from 15 years of fruitless rock-wrangling at the Permian boundary, are throwing in the trowel.

“I really enjoyed the P-T field in the 1990s,” says Retallack. “Then it was fun. Now it’s gotten to be name-calling and acrimony. I don’t relish the kind of debates that will go on from here on out. I don’t intend to pursue further studies of impact tracers.” He says he would rather study Paleozoic paleosols—really, really old dirt.

Doug Erwin has a book in press that will bid his adieu to the Permian problem. When he wrote his first book on the extinction in 1993, few of his colleagues were interested. Then “everybody who was busy worrying about the K-T extinction got bored and decided to come down to the P-T,” he says. Now it’s just too crowded to be any fun. Rampino agrees: “You have to take a number to get to study any boundary. It’s like going to the grocery store.”

Those players still in the game can look forward to more fireworks later this year, when results from a NASA-funded effort to verify Becker’s findings are due to be released. Last fall, the NASA program sent Erwin, Becker, and Frank Kyte, a geologist at UCLA, to the Meishan section in China, where they could decide, on site and in person, which rocks to analyze and how to divvy them up. The rocks were distributed among several U.S. laboratories for independent testing.

“The expedition to do some definitive sampling is just what we need,” says Melosh.

And so the crisis in end-Permian science will continue a while longer, claiming careers and ravaging reputations. It remains to be seen whether Luann Becker and her putative impact will be survivors, casualties, or one of those mysteriously resurgent Lazarus species.

The top five extinctions

Ordovician 440 million years ago; eradicated small organisms that lived on the bottom of the ocean.

Devonian 365 million years ago; caused the loss of coral reefs and small marine life-forms.

Permian 250 million years ago; wiped out more than 90 percent of marine species and drastically affected lineages of four-footed animals on land.

Triassic 210 million years ago; killed off more than a fifth of animal lineages, both on land and in the ocean.

Cretaceous 65 million years ago; decimated nonavian dinosaurs as well as marine reptiles and numerous species of marine organisms.

1 free article left
Want More? Get unlimited access for as low as $1.99/month

Already a subscriber?

Register or Log In

1 free articleSubscribe
Discover Magazine Logo
Want more?

Keep reading for as low as $1.99!


Already a subscriber?

Register or Log In

More From Discover
Recommendations From Our Store
Shop Now
Stay Curious
Our List

Sign up for our weekly science updates.

To The Magazine

Save up to 40% off the cover price when you subscribe to Discover magazine.

Copyright © 2023 Kalmbach Media Co.