Looking south through what was once the “ice-free corridor” in present-day Canada. A new study suggests that humans couldn’t have traversed through the corridor until about 12,600 years ago.

Looking south through what was once the “ice-free corridor” in present-day Canada. A new study suggests that humans couldn’t have traversed through the corridor until about 12,600 years ago.

Mikkel Winther Pedersen

Ancient DNA suggests the first Americans sidestepped the glaciers

Ask any schoolkid how the first people came to the Americas, and you might get some version of the following: They crossed a spit of land connecting Alaska and Siberia and made their way south between melting glaciers at the end of the last ice age. Until recently, science agreed. But mounting evidence has shown that the dry land exposed by the melted route—known as the ice-free corridor—may not have been passable until long after humans had already settled the Americas. So when did it become a viable route for people? Using ancient DNA, along with the remains of pollen, plants, and animals collected from lake sediments, a new study has an answer: about 12,600 years ago. This suggests that the earliest humans to make their homes in the New World, including people from the Clovis civilization, must have taken a very different route.

“This is a really neat and pioneering study,” says Stephen Jackson, a paleoecologist with the U.S. Geological Survey’s Southwest Climate Science Center in Tucson, Arizona, who was not involved in the work. Because this is the first study to take into account both so-called environmental DNA (eDNA) as well as more traditional types of data, he says, “we stand to learn a good deal more about how to interpret our records.”

Scientists have long thought that humans traveled across a region knowns as Beringia, a now-submerged area in the Bering Sea that was dry land during the lower sea levels of the last ice age.  What humans did next has always been a big question. Until recently, most researchers thought that the ice-free corridor (see map) was the most likely route south, once the glaciers began melting 14,000–15,000 years ago. For years, those dates fit the timing of the Clovis people, big game hunters thought to have inhabited the lower 48 about 13,000 years ago. But in the past decade, scientists have discovered even earlier settlements, revealing that humans made the journey to the Americas as early as about 15,000 years ago. The key question, says Eske Willerslev, a paleoecologist at the University of Copenhagen and a co-author of the study, is when the ice-free corridor really became viable for humans to cross it. At about 1500 kilometers long, large animal game would be essential to making the journey, he says.

Before 12,600 years ago, the researchers say, the more interior route into the Americas was blocked. Instead, early settlers may have taken a coastal route as early as 14,700 years ago.

Before 12,600 years ago, the researchers say, the more interior route into the Americas was blocked. Instead, early settlers may have taken a coastal route as early as 14,700 years ago.

Mikkel Winther Pedersen

To see into the corridor’s past, the new study looked at pollen, plant, and animal fossils from nine sediment cores taken from two lakes near what was thought to have been the narrowest bottleneck in the corridor—the last part to open. The modern lakes are the remnants of a giant prehistoric lake, known as glacial Lake Peace, which covered much of the area between the retreating glaciers. From the cores, the scientists also extracted eDNA—DNA lingering in the soil from, for example, plant leaves, rootlets, animal feces, urine, or even skin cells. Because DNA is electrically charged, it can bind to sediment particles, helping to preserve it from degradation over time.

Taken together, these data paint a detailed picture of the ecology of the corridor as it transitioned from icy wasteland to fertile forest. For about 700 years after the ice retreated, there was scant evidence of any life, Willerslev says. Then, about 12,600 years ago, steppe plants like aromatic sagebrush appeared, followed soon after by animals such as woolly mammoth, bison, and jackrabbits. By about 12,400 years ago, forests of Populus trees—such as aspen and poplar—began to dominate, after which elk and moose arrived. About 11,600 years ago, the region transformed again into a boreal forest of spruce and pine trees.

But this opening and flourishing occurred too late for the migrating humans who arrived in the Americas about 15,000 years ago, the authors report online today in Nature. It was even too late for the Clovis people, who arrived about 13,000 years ago, they say.

Instead, Willerslev says, both pre-Clovis and Clovis peoples may have taken the so-called coastal route, which would have taken them down along the western coast of North America, a much-discussed path with scant hard data to support it. He says he hopes to use eDNA to hunt for evidence of their passage. “The idea is that there was a land bridge a few thousand years earlier than the formation of the ice-free corridor,” he says. “That land is now covered by ocean, but there are some islands believed to be part of that route. It would be interesting to go and look for cores and try to do the same exercise there.”

Other recent evidence supports the idea that the earliest human migrants to the Americas—the pre-Clovis people—didn’t enter through the Canadian corridor. For example, a recent study of mitochondrial DNA in northern and southern populations of bison separated by the corridor suggest the passage opened up slightly earlier than the present study posits—13,000 years ago instead of 12,600 years ago. In terms of that timing, “it’s a pretty subtle difference,” says Duane Froese, a geoscientist from the University of Alberta, Edmonton, in Canada, and a co-author of that study.

But when it comes to estimating the opening of the ice-free corridor, the devil may be in those details—in part because the slightly earlier opening suggested by the bison mitochondrial DNA would have allowed the Clovis people to take it. Willerslev questions the reliability of using just one organism to record environmental history. But Froese points to several other fragments of data from the region that suggest the corridor was habitable earlier: a 13,700-year-old fragment of a poplar tree and a 13,100-year-old bison found near the bottleneck.

Jackson notes that the eDNA evidence is powerful and fills in many gaps left by conventional paleoecological data. For example, certain trees—Populus is a particularly well-known example—are poorly represented in the pollen record, even when they were certainly present. eDNA offers a detailed snapshot of local flora and fauna that passed through a particular location, from bison stepping into the lake to pikefish swimming in it. But, Jackson adds, eDNA paleoecological research is still new, and there remain lingering questions about how eDNA is represented in sediments and what exactly the data mean.

Froese is optimistic. “It’s great from a science point of view that we don’t completely agree,” he says of his team’s work and Willerslev’s. “I hope this will be good impetus for future research.”