Ask someone how the dinosaurs died, and chances are they will tell you about an apocalyptically bad day 66 million years ago, when a huge asteroid slammed into Earth and triggered a nuclear winter. This event left behind distinct traces in the geologic record, as well as a vast crater off Mexico's Yucatán peninsula called Chicxulub. But since the 1980s, researchers have debated whether volcanoes in ancient India may have been an accessory to the mass extinction—or even its main driver.
Now, two independent research teams have created the best time lines of this volcanic activity ever assembled. Despite using different dating methods, their two studies—published on Thursday in the journal Science—agree in broad strokes on the timing of the ancient eruptions, helping clarify how they could have played a supporting role in wiping out non-avian dinosaurs. (Meet the dinosaurs that didn't die.)
“We agree more than we disagree, and that's a really important conclusion,” says Courtney Sprain, a postdoctoral researcher at the University of Liverpool who led one of the studies, working with a team from the University of California, Berkeley.
According to the two teams, the massive volcanoes, called the Deccan Traps, started erupting about 400,000 years before the Chicxulub impact and wrapped up about 600,000 years after the end of the Cretaceous period. At least half of the volcanoes' total erupted lava spilled out after the impact.
“That's a big improvement over 20 years ago, or even 15 years ago, where [the two teams' dating methods] couldn't agree to better than a few percent, which here would be millions of years,” says Princeton geochronologist Blair Schoene, who led the other study. “The level of agreement between the studies is pretty remarkable.”
However, the studies disagree on the eruptions' finer timing and tempo, which are crucial when assigning relative blame for the extinctions. One study claims that the Deccan Traps saw a major eruptive uptick in the hundred thousand years leading up to the impact, potentially stressing out some ecosystems before the asteroid's decisive blow. The other study, however, finds that most of volcanoes' lava erupted after the impact had occurred, implying a smaller role.
Wrinkles in time
The Deccan Traps were incomprehensibly huge by modern standards, expelling some 135,000 cubic miles of lava over a million-year period. That's enough lava to circle Earth with a rocky belt more than five miles wide and a mile tall. By contrast, the 2018 eruption of Hawaii's Kilauea volcano spewed about a fifth of a cubic mile of lava over a couple months.
These sorts of mega-volcanoes are suspects in other mass extinctions. Most infamously, huge volcanoes in what is now Siberia probably played a key role in Earth's deadliest mass extinction at the Permian-Triassic boundary about 252 million years ago. But how, exactly, the Deccan Traps contributed to the dinosaurs' demise depends on timing. (Here are some of the weirder theories once proposed for what killed the dinosaurs.)
That's why the Berkeley and Princeton teams independently traveled to the Western Ghats, the region in India where the Deccan Traps once roared. The region's dramatic, banded valleys are carved from the leftovers of these immense volcanoes. In some places, the rock layers formed from cooled lava are more than a mile thick.
“As someone who loves looking at rock, it's one of the few places you can go where everything you look at is the same thing and caused by the same event,” Sprain says. “It's just a really cool place to be.”
Sprain's team dated when minerals in the lava cooled below a few hundred degrees Fahrenheit, which told them directly when the lavas erupted and cooled. By contrast, Schoene's team dated crystals called zircons found in the sediments sandwiched between lava layers. These crystals formed in distant volcanoes and arrived in wind-borne ash. Based on the spread of zircon ages within a given sediment layer, Schoene's team could precisely estimate when the lava above it was laid down.
Despite agreeing on the big picture, the two time lines differ in the details. Schoene's team found signs of four distinct pulses of eruptions at the Deccan Traps, with the biggest coming about a hundred thousand years before the impact. That's suggestive of a larger role for the volcanoes in the extinction: The eruptions could have injected massive amounts of greenhouse gases and particles into the atmosphere, changing Earth's climate in ways that stressed out late Cretaceous life. Then, in a one-two punch, the impact's nuclear winter would have sharply cooled Earth and caused ecosystems to collapse.
However, Sprain's team doesn't see pulses, and they find that a full 75 percent of the Deccan lava erupted after the impact. This suggests that the Deccan Traps' lava couldn't have played a decisive role in the extinction event itself—and judging by life's rapid bounce back from the mass extinction, the eruptions didn't seriously hamper recovery, either. Instead, the smaller pre-impact eruptions may have been extremely gassy, causing the warming and cooling observed in the 300,000 years before the impact.
Sprain's data even suggest that the impact may have caused the Deccan Traps' larger, post-impact eruptions. Sprain and UC Berkeley geochronologist Paul Renne, one of her coauthors, have raised this idea in previous studies.
A unified theory
In large part, the two studies' differences stem from how hard it is to reconstruct Earth's past. Technology has improved to the point at which scientists can date the formation of certain rocks to within tens of thousands of years. In the long view, that's very precise: If Earth's 4.54-billion-year history were compressed onto a calendar year, we now can time geologic events down to a few minutes. And the teams that authored the new studies are considered the best at rock dating in the world.
“It's kind of a neat commentary on the progress of science,” says U.S. Geological Survey geochronologist Seth Burgess, who wrote about both studies in an accompanying article in Science. “With better techniques and more fieldwork, we're able to look much more closely of what happened when, and what was a more simple story has gotten more complex.”
The problem is that the Chicxulub impact was literally over in a day, so when trying to date the rocks that formed directly before or after, the sequence of events can remain stubbornly fuzzy. Case in point: the Poladpur Formation, a key layer of Deccan Traps lava that both teams examined. Sprain's data allow for the Poladpur to have formed alongside or after the impact. But Schoene's models suggest that the Poladpur formed in the hundred thousand years before the impact.
Whether the Poladpur lava flowed before or after the impact matters quite a bit. The formation consists of 35,000 cubic miles of lava, about a quarter of the total Deccan Traps volume. Did the Poladpur help prime Earth for death, or did it come after the fact?
“I find it difficult to know how we'll ever stop arguing about it,” says University of Leeds paleontologist Paul Wignall, who wasn't involved with either study. “As the timing has improved, in a way it's just almost worse, because it shows that the coincidence of the volcanism and the impact are very close—almost beyond resolution, really.”
Still, the researchers are already talking about how to put their competing ideas to the test. Schoene's analysis, for instance, predicts that the Deccan Traps went relatively quiet for up to 300,000 years, plenty of time for cooled lava on the surface to erode. If Schoene's team is right, then signs of this ancient weathering should be buried within the Deccan Traps' layers.
Along the way, both universities' teams remain committed to working together—a sign of unity after decades of disagreement. Berkeley was home to Luis and Walter Alvarez, the father-son team that famously advocated for the asteroid-impact model. Princeton, meanwhile, is home to Gerta Keller, a coauthor on Schoene's paper who has argued for decades that the Deccan Traps alone caused the dinosaurs' demise.
“We're getting closer to an answer we hope that all groups agree on,” Sprain says.