The latest "made in China" item isn't a plastic widget or a pair of shoes. It's a dog. A new study suggests that wolves were first domesticated in Southeast Asia some 16,000 years ago. The work is the latest volley in a long-standing debate about just where canine companionship got its start.
Most researchers agree that dogs descended from wolves, but when and where has been hard to pin down because archaeologists have trouble telling wolf remains from dog remains. So in the past decade, geneticists have started to look at DNA for clues. In 2002, geneticist Peter Savolainen of the KTH-Royal Institute of Technology in Stockholm and colleagues analyzed mitochondrial DNA from 38 wolves and more than 500 dogs around the world. They found the most genetic diversity--a marker of a species' origin--in East Asia and concluded that dogs were domesticated there, and just once.
But last month, a study of African village dogs called that conclusion into question. Computational biologists Adam Boyko and Carlos Bustamante of Cornell University and their colleagues sampled more than 300 village dogs from Egypt, Namibia, and Uganda. The genetic diversity of the African village dogs was on par with that seen in East Asian dogs, they reported  online 3 August in the Proceedings of the National Academy of Sciences. Boyko and Bustamante don't think that dogs originated in Africa, however, because gray wolves, the dog's likely predecessor, are not found on that continent. But the work did seem to argue that Savolainen's genetic diversity data weren't strong enough to support his conclusions for an East Asian origin.
Now Savolainen, Ya-Ping Zhang of the Chinese Academy of Sciences' Kunming Institute of Zoology, and their colleagues have done an even more extensive survey of dog DNA. They looked at a small piece of mitochondrial DNA from more than 1500 dogs distributed across Europe, Africa, and Asia, with an emphasis on East Asia. Some were breeds with known geographic origins; others were working dogs in rural areas. The researchers also looked at 40 wolves. They then sequenced almost all of the mitochondrial genome from eight wolves and from 169 dogs representing the range of diversity identified in the initial 1500-plus animals.
The data reaffirm a single site for domestication and pinpoint the origin of the domesticated dog to a region south of the Yangtze River, where wolf taming was quite common, Savolainen's team reports  today in Molecular Biology and Evolution. That's where the largest number of similar groupings of DNA, called haplogroups, is found. As the researchers looked at dogs farther from this region, they saw fewer haplogroups; Europe had only four, for example. "The gene pool we are finding in Europe and Africa are a subset of the South Chinese gene pool," says Savolainen.
Savolainen's critics have noted that low diversity in European dogs might be the result of intense selective breeding, and thus it may not reflect the ancient diversity of dogs there. But the European dogs all lack the same six haplogroups, regardless of breed, indicating that the lower diversity preceded the development of purebred strains, Savolainen says.
The study "is the most compelling evidence for the origin of the dogs published to date," says Hannes Lohi, a veterinary molecular geneticist at the University of Helsinki. "The data here [are] very much similar to corresponding human data for the out of Africa evidence. It looks like dogs were 'created' in China."
Bustamante is impressed, albeit cautious. "It's a solid study and an important study," he says. "But the conclusion that there was one and only one center of domestication is premature." He and others point out that researchers need to see if the pattern of diversity holds up in nuclear genes, which might tell a different story. More wolves need to be analyzed as well to correlate their haplogroups with those seen in dogs.
Carles Vilà of the Biology Station of Doñana-CSIC in Seville, Spain, is even more cautious. He points out that other genetic studies suggest dogs date back at least 20,000 years and that archaeological remains of dogs in Europe are almost as old. He also worries about bias that might have arisen in Savolainen's study if the dogs were not sampled the same way in all locations. "I'm not convinced by the results," he says, "and I do not think this is the last that we will hear about the time and place of the domestication of dogs."