In contrast to ice-covered North America, parts of the land bridge were cold but not frozen, and dotted with brushy shrubs and even trees.
"It was an area where people could have had resources, lived and persisted through the last glacial maximum in Beringia," said O'Rourke in a statement. "That may have been critical for the people to subsist because they would have had wood for construction and for fires. Otherwise, they would have had to use bone, which is difficult to burn."
O'Rourke and his co-authors propose the ancient people may have lived in the shrubby lowland area from about 25,000 years ago to 15,000 years ago, when glacial ice sheets melted and they began making their way down to North America.
The theory could also help explain how American Indians evolved a genome separate from their Asian ancestors, he said.
"The only way to do that was for the population to be isolated," he said. "Most of us don't believe that isolation took place in Siberia because we don't see a place where a population could be sufficiently isolated. It would always have been in contact with other Asian groups on its periphery."
O'Rourke and his co-authors point to a study of mitochondrial DNA that found the unique American-Indian genome arose more than 25,000 years ago, but didn't spread through the Americas until closer to 15,000 years ago.
Written with archaeologist John Hoffecker, of the University of Colorado at Boulder, and paleoecologist Scott Elias, from the University of London, the Science Perspective column titled "Out of Beringia?" says that even though most of the archeological sites that could prove the idea are now underwater, a few may have survived and could be discovered with more research.