If you travel the meandering Sepik River of New Guinea, it quickly becomes apparent that from one bend to the next the people along the banks speak distinct languages. The island's remarkable linguistic diversity reflects real genetic differences, a research team reports this week in Science. More unexpected, the team concludes that this genetic variation dates back just 10,000 to 20,000 years, rather than to 50,000 years ago or so, when humans first arrived.
The island's independent invention of agriculture 10,000 years ago did not wipe out the genetic differences, as it did in Europe or parts of Asia. "With agriculture, you tend to get genetically homogenized societies," says team member Anders Bergström, a graduate student at the Wellcome Trust Sanger Institute in Hinxton, U.K. In Europe, farmers from Anatolia replaced local hunter-gatherers and erased much of their genetic contribution. That this did not happen on New Guinea "is a big surprise," says Sanger geneticist Chris Tyler-Smith, who led the team.
The researchers analyzed variation among 1.7 million DNA markers across the genomes of 381 Papua New Guinea (PNG) residents, and they also compared the complete genomes of another 39. They concluded that the people of New Guinea were isolated from Asians for most of prehistory, and that highland and lowland dwellers separated from each other 10,000 to 20,000 years ago. In the highlands, people split into three very distinct clusters of social groups within the past 10,000 years, soon after they began cultivating plants. In the lowlands, two main clusters arose in the north and south.
The best explanation for this pattern, says Bergström, is that once people began cultivating crops, they spread their genes across the island along with the technology. But soon afterward, their descendants apparently stopped mixing as much and evolved distinct local genotypes. Although researchers have long thought that the island's mountainous terrain kept highland groups isolated, this study finds that the clusters formed in both the highlands and lowlands—where the terrain is flatter. Cultural factors, such as warfare or within-group marriage, were more important than geographic barriers in preventing mixing, Bergström suggests.
But Polly Wiessner, an anthropologist at the University of Utah in Salt Lake City, finds it unlikely that farmers' genes overtook the island. She argues that trade networks, not just farmers, spread stone mortars and pestles to pound taro and other new crops between groups of hunter-gatherers. Their practice of marrying within groups amplified their genetic differences. "There is no evidence that I know of that agriculturalists replaced hunter-gatherers," she says.
Regardless of the cause, PNG's sharp genetic distinctions suggest that the spread of farmers may not be enough to homogenize DNA across large regions. Other, later waves of migration may have wiped out differences in Europe and Asia. "People in PNG did not experience the Bronze Age and Iron Age transformation," adds evolutionary geneticist Simon Easteal of the Australian National University in Canberra. The new study, he says, indicates that "there is a general erosion of local genomic diversity in human populations associated with technological change—a process that is continuing today."