A Genetic Study Writes a New Origin Story for Dogs

But it probably won’t be the last word.

Erik de Castro / Reuters

“Not too many people have been bitten during the course of this project,” says Adam Boyko at Cornell University, “but it's not zero.”

Boyko's team, fronted by postdoc Laura Shannon, have spent the last seven years traveling the world and collecting blood samples from village dogs—free-ranging dogs that live near human settlements. Some are feral, some are strays, some hang around specific properties but aren't owned by a particular person. Whatever the case, it's these unleashed hounds, and not the familiar breeds at Crufts, that represent the full genetic diversity of our canine companions.

“There are millions of dogs in the world and the vast majority of them are not purebred,” says Boyko. “And very little is known about these free-ranging, village dogs.” His team rectified that by collecting blood samples from 549 village dogs in 38 countries across 6 continents. “I did my Ph.D. work with tropical butterflies; by comparison, working with dogs is fantastic,” he adds. “You don’t need to hunt them down with a net. You show up, you have food, there are dogs.”

After analyzing more than 185,000 genetic markers in these individuals—the largest ever survey of global canine diversity—the team concluded that dogs were first domesticated in Central Asia, somewhere near India or Nepal.

But this isn't a final answer. It's just the latest volley in a long-running debate about when, where, and why wild wolves first transformed into man's best friend. “It almost completes the set of possible areas where dogs may have been domesticated,” says Greger Larson from the University of Oxford in an email statement. “Previous genetic analyses have pointed a finger at, or at least gestured in the direction of, Europe, the Near East, Siberia, and China. Adding Central Asia now means that everyone with a favourite region can point to at least one paper that supports their suspicions.”

The oldest dog fossils are at least 15,000 years old and come from Western Europe and Siberia. But genetic studies have thrown forth hugely conflicting answers depending on whether they look at modern DNA from living dogs or ancient DNA from fossil ones, which parts of the world they collect samples from, and whether they look at full genomes, specific markers, just the Y chromosome, or just mitochondrial DNA (a secondary genome that sits outside the main one).

In 2013 alone, a whole-genome study of living wolves and dogs argued that domestication took place around 10,000 years ago during the Agricultural Revolution, when wolves that scavenged at humanity's scrapheap became more accustomed to life with us. Four months later, another whole-genome study argued that wolves were domesticated in East Asia around 32,000 years ago. Six months after that, yet another study—this one of mitochondrial DNA in both modern and fossil dogs—put the site of domestication in Europe and the time somewhere between 18,000 and 32,000 years ago. That's well before the Agricultural Revolution, and suggests that wolves may have accompanied European hunter-gatherers as either hunting partners or scavengers.

Boyko's study doesn't resolve this morass of conflicting results, but it does adds two valuable new things. First, the team focused on village dogs, which have been largely ignored till now. Second, they used a different method than previous genetic studies. They relied on a well-known phenomenon called linkage disequilibrium (LD), in which nearby genes, which are normally inherited as a single chunk, are more likely to break apart over time.

If you look at human populations, you see the least amount of LD in east Africa and increasing amounts in more distant parts of the globe—this reflects the African origin of our species, and our subsequent planetary expansion. Similarly, Boyko's team found the least amount of LD among village dogs in Central Asia, pinpointing this region as the cradle of domestication.

But Olaf Thalmann from the University of Oulu says that LD is a rather “crude estimator.” There are other explanations for the pattern that the team saw. For example, domestic dogs could have evolved in the Middle East or Europe and been carried to Central Asia by humans, before going extinct in their site of origin. Will this be the last word in dog domestication?” asks Thalmann. “I highly doubt it. I personally think that we reached a dead end with inferences from modern data.”

By which he means: We need to analyze more ancient DNA from dog and wolf fossils.

Larson agrees. “The origins of modern populations are extremely messy,” he says. “Thus, it is difficult, if not impossible, to infer the true deeper history of [dogs] by exclusively investigating extant populations, no matter how far flung or deeply sequenced they are.”

Just take our own species. No comparison of living people would have ever revealed that all non-Africans carry a small proportion of Neanderthal DNA, or that Melanesians carry DNA from another mysterious group of ancient humans called the Denisovans, or that a massive wave of European migrants headed into Africa around 3,000 years ago. It took ancient DNA to uncover those revelations about our history. The same applies to our dogs.

Larson is on the case. His team has spent years collecting samples from over 1,500 ancient dogs and wolves, and is now finally ready to analyze them all. “Once we have the ancient data, we can compare it against patterns of genetic variability in the modern populations to really get to grips with where and when dogs were domesticated and the dynamic and complicated pattern of dispersal and admixture that’s happened since,” he says.

Boyko is also excited about that project. “If both of our results lined up and pointed in the same place, then we could get a consensus in the field,” he says.

And if they don't?

“Then we need work out which of our assumptions is least well supported.”

Ed Yong is a former staff writer at The Atlantic. He won the Pulitzer Prize for Explanatory Reporting for his coverage of the COVID-19 pandemic.