Archaeological studies have found that human colonization of North America by the so-called Clovis culture dates back more than 13,000 years ago, and recent archaeological evidence suggests that people could have been on the continent 14,700 years ago—and possibly even several millennia before that. The conventional thought has been that the first migrants who populated the North American continent arrived across an ancient land bridge from Asia once the enormous Cordilleran and Laurentide ice sheets receded to produce a passable corridor nearly 1,000 miles long that emerged east of the Rocky Mountains in present-day Canada.

Evolutionary geneticist Eske Willerslev, however, believed there was one aspect of the conventional theory that required further investigation. “What nobody has looked at is when the corridor became biologically viable,” says Willerslev, director of the Center for GeoGenetics at the University of Copenhagen. “When could they actually have survived the long and difficult journey through it?”

A pioneer in the study of ancient DNA who led the first successful sequencing of an ancient human genome, Willerslev specializes in extracting ancient plant and mammal DNA from sediments to reconstruct ancient history. According to a recent profile in the New York Times, “Willerslev and his colleagues have published a series of studies that have fundamentally changed how we think about human history,” and a new study published in the journal Nature co-authored by Willerslev may lead to a rethinking of how Ice-Age humans first arrived in North America.

The study’s international team of researchers travelled in the dead of winter to the Peace River basin in western Canada, a spot that based on geological evidence was among the last segments along the 1,000-mile corridor to become free of ice and passable. At this crucial chokepoint along the migration path the research team took nine sediment cores from the bottoms of British Columbia’s Charlie Lake and Alberta’s Spring Lake, remnants of a glacial lake that formed as the Laurentide Ice Sheet began to retreat between 15,000 and 13,500 years ago.

After examining radiocarbon dates, pollen, macro-fossils and DNA from the lake sediment cores, the researchers found that the corridor’s chokepoint was not “biologically viable” to have sustained humans on the arduous journey until 12,600 years ago—centuries after people were known to have been in North America. Willerslev’s team found that until that time the bottleneck area lacked the basic necessities for survival, such as wood for fuel and tools and game animals to be killed for sustenance by hunter-gatherers.

From the core samples, the researchers discovered that steppe vegetation first began to appear in the region 12,600 years ago followed quickly by the arrival of animals such as bison, wooly mammoths, jackrabbits and voles. Around 11,500 years ago there was a transition to a more densely populated landscape with trees, fish such as pike and perch and animals including moose and elk.

The research team used a technique called “shotgun sequencing” to test the samples. “Instead of looking for specific pieces of DNA from individual species, we basically sequenced everything in there, from bacteria to animals,” Willerslev says. “It’s amazing what you can get out of this. We found evidence of fish, eagles, mammals and plants. It shows how effective this approach can be to reconstruct past environments.

“The bottom line is that even though the physical corridor was open by 13,000 years ago, it was several hundred years before it was possible to use it,” Willerslev says. “That means that the first people entering what is now the U.S., Central and South America must have taken a different route. Whether you believe these people were Clovis, or someone else, they simply could not have come through the corridor, as long claimed.”

“There is compelling evidence that Clovis was preceded by an earlier and possibly separate population, but either way, the first people to reach the Americas in Ice-Age times would have found the corridor itself impassable,” adds study co-author David Meltzer, an archaeologist at Southern Methodist University.

While later groups may have used the passageway across the land bridge between Siberia and Alaska, the study’s authors say the first humans in North America likely migrated along the Pacific coast, although it is still not known exactly how.

“The route taken by first humans coming to America is still unknown, but much evidence points to the Pacific coast,” says study co-author Mikkel Winther Pedersen, a Ph.D. student at the Centre for GeoGenetics at the University of Copenhagen. “If this is the case, we could be looking at humans who adapted to survive by exploiting the marine resources, whether by boat or from sea-ice. They could have had a subsistence resembling what Inuit have had.”