How Did People Reach the Americas?

From ancient DNA, Eske Willerslev teases out answers to prehistoric puzzles.

After years of spirited debate over how and when people first reached the Americas, scientists finally seem poised to reach agreement. The emerging consensus: In contrast to what was long held as conventional wisdom, it now seems likely that the first Americans did not wait for ice sheets covering Canada to melt some 13,000 years ago, which would have allowed them to traipse south over solid ground. Instead, early nomads might well have traveled by boat or at least along the coast from Siberia to North America, perhaps navigating arctic waters near today's Bering Strait. The telltale evidence: ancient DNA from those early people that's been coaxed, by powerful analytical technology, into revealing its secret.

Rewriting the prehistory of the Americas is perhaps the most remarkable discovery—but hardly the only one—so far achieved through the analysis of ancient DNA. Other new insights about the past are being drawn from the same emerging scientific discipline. In the past five years, the double helix has shed light, for example, on the vanished woolly mammoth, the flightless dodo, and even humanity's long-lost kin, the Neanderthals. Extracting and testing old DNA, once considered practically impossible because too little of the stuff survives the eons intact, are now at the cutting edge of archaeology, paleontology, and other fields, thanks to new techniques and more powerful technology.

"Archaeologists are used to stone tools and bones," says Ted Goebel of Texas A&M. "So for us to be presented with this kind of evidence is pretty intriguing." DNA, which contains the blueprints for organisms, degrades over time, breaking down into tiny pieces or disintegrating entirely. For years, the dearth of intact DNA in ancient samples—a chunk of mammoth bone, for instance, or a human hair—stymied researchers who were trying to analyze the material. But now, using a technique called polymerase chain reaction, or PCR, researchers can "unzip" minute fragments of surviving DNA and duplicate them millions of times over, until they have a sample large enough to test. Then, by comparing differences between the ancient material and modern samples of known provenance, they can analyze a long-extinct animal's genome.

The resulting data, in some cases, can resolve a long-standing scientific deadlock. For almost a century, most archaeologists believed that people arrived in the Americas between 13,000 and 13,500 years ago. The date was based on flint tools first found in Clovis, N.M., and later all over North America. From that evidence, archaeologists sketched out a scenario in which fur-clad "Clovis" hunters chased mammoths and other prey from Siberia to North America across a land bridge exposed by low sea levels. Then, the theory goes, they hunted along a path from Alaska down through Central America and all the way to Chile in just a few centuries.

Before Clovis. But the notion that Clovis came first has collapsed in the face of recent evidence, including DNA that pushes the arrival of the first humans in the Americas back at least 1,000 years, centuries before the ice that covered northern Canada at the time had melted enough to allow migration. Instead, some argue, the first Americans must have arrived by boat, skirting the coast from Siberia and sailing south along the American coast.

University of Oregon archaeologist Dennis Jenkins discovered the critical new evidence buried more than 4 feet below the floor of a dusty cave near Paisley, Ore. The "artifacts" were 14,300-year-old fossilized pieces of excrement, or coprolites. Jenkins, who has been digging in Oregon's high desert for decades, handed off bits of coprolite to geneticist Eske Willerslev of the University of Copenhagen. In Willerslev's laboratory, PCR pulled enough DNA from the ancient poop to prove it was human and even genetically link it to modern American Indians.

Announced in April in the journal Science, the find backs up evidence previously found at the other end of the Americas, at a site in Chile called Monte Verde. There, a full-fledged campsite was radiocarbon dated to 14,500 years ago, putting people in South America more than a millennium before those Clovis hunters supposedly crossed the Bering Strait. But carbon dating is inexact, and the Monte Verde find had not convinced some skeptics. With DNA analysis, says Jenkins, "we can directly date the item and verify it's human." The Oregon find has largely silenced the last few Clovis adherents. "It's pretty compelling stuff," says Goebel, a longtime Clovis supporter.