A present day view south in the area where the retreating ice sheets created the ice free corridor more than 13,000 years ago.
(Mikkel Winther Pedersen)

At long last, the Ice Age is almost over. The vast glaciers that have formed an impenetrable barrier across the top of North America are finally beginning to recede. Between the two walls of ice, a corridor emerges, harsh and scrubby, but green enough to sustain life. In a chilly but habitable region of Alaska, the descendants of Asians who daringly crossed an ice bridge into the unknown get the opportunity they've been waiting for: Finally, a pathway south.

They step into the corridor. They slowly edge their way south. When they emerge on the other side, many decades and 1,000 miles later, they are in what is now Montana – the first human inhabitants of what is truly a new world.

Unless you closely follow the latest findings in American paleoarchaeology, this is probably the version of America's origin story you're accustomed to. But it's almost certainly wrong.

A growing body of evidence suggests that people had colonized the Americas several thousand years before the end of the last Ice Age — long before the corridor between the glaciers was even open.

"The school book story that most of us are used to doesn't seem to be supported," said Eske Willerslev, an evolutionary geneticist at the University of Copenhagen's Center for Geogenetics, and a co-author of a new study on the history of that legendary ice-free corridor. 

Writing in the journal Nature on Wednesday, Willerslev and his colleagues reconstruct what the corridor would have been like 13,000 years ago, when geologic records indicate it first opened up. In rock cores taken from the bottoms of former lakes, they find evidence of a rocky, barren place that no prehistoric human would have ventured into for several hundred years.

If and when people finally did migrate through the corridor, the scientists say, it seems likely that they traveled south to north as well north to south — adding to the evidence that humans were already in the lower 48 when the path through Canada emerged from the ice.

Twenty years ago, this finding would have been deemed outlandish. For most of the 20th century, it was archaeological orthodoxy that a society known as Clovis — for the characteristic fluted points made by human hands and discovered in Clovis, New Mexico — was the first to colonize the Americas 13,000 years ago. Evidence of their technological sophistication and dominant presence litters the prehistoric landscape: some 10,000 Clovis points have been discovered in locations as far flung as Florida and Washington state. Not coincidentally, those points' appearance in the archaeological record coincides perfectly with when geologists believed the ice sheets covering Canada first began to melt. For years, the "Clovis first" and "ice-free corridor" models for migration into the Americas went hand in hand — and straight into high school history textbooks.

But in the late 1990s, a group of prominent archaeologists examined a hunk of mastodon meat and the remains of huts uncovered at Monte Verde in southern Chile. They concluded the site had been established by people who predated Clovis by some 1,500 years. (Note that dating is a fraught subject in paleo-American studies; some have estimated Monte Verde is as old as 18,500 years.) Other discoveries of human activity — but so far, no actual human remains — at roughly a dozen sites scattered throughout two continents have bolstered support for a new model: the first Americans traveled along the Pacific coast by boat, bopping between intermittent ice-free beaches and living off the abundance of the sea until they finally reached the land beyond the glaciers.


Map outlining the opening of the human migration routes in North America. The yellow arrow indicates the coastal route, orange indicates the "ice free corridor." (Mikkel Winther Pedersen)

These days, it's rare find a scientist who promotes the "Clovis first" and "ice-free corridor" model. But the concept still looms large in American archaeology.

"It's really funny. After all these years a lot of people keep starting their papers with the 'Clovis first' model and then knocking it down," said Jennifer Raff, an anthropological geneticist at the University of Kansas who was not involved in the Nature study. "Sometimes it's like, 'You guys, we have to move on.'"

But moving on is hard to do — even when the incident in question happened more than a dozen millennia ago. If the very first Americans didn't come through the corridor, some wondered, perhaps the Clovis people took the route later.

No such luck, according to Willerslev.

"Even for Clovis people ... I find it very unlikely they could have come through the interior ice corridor," he said. "It simply wasn’t viable."

To reach this conclusion, Willerslev and his colleagues took core samples from nine former lake beds in British Columbia, where the vast Laurentide and Cordellian ice sheets parted. They used radiocarbon analysis to date each sediment layer, then tested for large fossils, pollen (evidence of trees) and DNA (evidence of, well, everything else).

Normally, it would be all but impossible to get coherent data from a random sediment sample — the strands of DNA are too jumbled and degraded to be isolated for sequencing. But the Nature team used a technique called "shotgun sequencing," which involves indiscriminately sequencing every bit of DNA in a clump of organic matter and matching the results to a database of known genomes to see signs of individual organisms emerge.

Laminated lake sediments, younger layers deposited on top of older layers, containing molecular and fossil evidence revealing the succession of plants and animals in the ice free corridor. (Mikkel Winther Pedersen) Laminated lake sediments, younger layers deposited on top of older layers, containing molecular and fossil evidence revealing the succession of plants and animals in the ice-free corridor.
(Mikkel Winther Pedersen)

"This allows us to analyze DNA from all types of organisms ... and start to characterize the environment through time," said Mikkel Pedersen, PhD student at the Center for Geogenetics and the first author on the paper.

In the deepest layers, from 13,000 years ago, Pedersen and his colleagues found almost no evidence of life.

"The land was completely naked and barren," he said.

About 400 years above that, in the 12,600 year old layer, they started seeing traces of bison, hare and sage brush — the things that would live in an emerging tundra steppe environment. At this point, the corridor was likely lush enough to support human travelers, he said.

By 10,000 years ago, the corridor would close again — not because of ice, but because of trees. Canada was blanketed in the dense boreal forest we know today.

This research doesn't put an end to the questions about human migration into the Americas — it just means scientists can focus on new ones. For example, how were the Clovis people related to ones who settled at early sites like Monte Verde. And if those first arrivals really did take a coastal route, where is the evidence?

"It's a complicated geological situation and a complicated archaeological situation," said David Meltzer, an author on the Nature paper and an archaeologist at Southern Methodist University. Sea levels have risen dramatically in the last 13,000 years — likely swamping any coastal settlements the first Americans might have established. 

Does that mean those settlements are now beyond our reach? He laughed. "Oh, one never bets against negative data in archaeology."

In other words, there's still hope.

Correction: An earlier version of this post misstated where Clovis is located. It is in New Mexico.

Read more:

Challenge to presidential candidates: Debate about science

Dear Science: Why do hangovers happen, and how can I make them go away?

Report: A scientific superstar at Smithsonian is fighting for his job

This 250,000-year-old tool was used to butcher a rhinoceros

Blame plate tectonics for the disappearance of this ancient sea