Dogs were domesticated once from a lost population of wolves
Our history with dogs is complicated, according to a study of ancient dog DNA.

KIONA N. SMITH - 10/29/2020


Genomics researcher Anders Bergstrom and his colleagues recently sequenced the genomes of 27 dogs from archaeological sites scattered around Europe and Asia, ranging from 4,000 to 11,000 years old. Those genomes, along with those of modern dogs and wolves, show how dogs have moved around the world with people since their domestication.

All the dogs in the study descended from the same common ancestor, but that original dog population split into at least five branches as it expanded in different directions. As groups of people split apart, migrated, and met other groups, they brought their dogs along. Dog DNA suggests that their population history mirrors the story of human populations, for the most part.

“Understanding the history of dogs teaches us not just about their history, but also about our history,” said Bergstrom, of the Francis Crick Institute, in a statement.

We still don’t know who let the dogs out

We still don’t know exactly when or where dog domestication first happened; it already had a pretty complex history by 11,000 years ago. But it looks like it only happened once. The ancient genomes suggest that dogs all share a common ancestor, which they don’t share with modern wolves. According to Bergstrom and his colleagues, that probably means that dogs all descend from one group of wolves, and that group is now extinct.

Modern gray wolves don’t appear to be very closely related to any of the ancient or modern dogs in the study. That suggests that since domestication set them apart, wolves haven’t contributed much DNA to dog bloodlines.

The oldest dog in the study lived with Mesolithic hunter-gatherers about 10,900 years ago in what is now Sweden. Its DNA suggests that most of its ancestors were from an eastern branch of the dog family tree—the branch that gave rise to Siberian dogs, indigenous North American dogs, and even New Guinea singing dogs and Australian dingoes.

But some of the dog’s ancestry also came from the branch that had followed humans into the Levant and Southwest Asia. Those bits of DNA were probably picked up as souvenirs when the dog’s ancestors met dogs from another population. In other words, by 11,000 years ago, dogs had had time to become a species, divide into distinct populations as they moved apart, and then meet again and exchange DNA.

Have dogs, will travel

Bergstrom and his colleagues wanted to know how dogs’ population history lined up with that of humans. They compared their ancient dog data to what ancient human DNA tells us about how groups of people migrated and interacted during the last 12,000 years. Not too surprisingly, the timing of splits, mergers, and movements mostly matched up. That suggests that, as groups of people migrated, they brought their dogs with them, and the dogs got up to much the same things the humans got up to when they met new neighbors.

Ancient human DNA tells us that early farmers from what is now Turkey moved north and west into Europe around 8,000 years ago, and it took just a few centuries for them to completely replace the populations of hunter-gatherers who were already there.

“It’s not clear how these movements happened—whether by disease, or by violence, or by some kind of biased intermarriage process—but what the genetics shows unambiguously is that these changes did happen, and much more dramatically than any archaeologists expected,” said Reich back in 2018.

And DNA from ancient European dogs tells us that very similar things were happening between the Neolithic newcomers’ dogs and the ones (like the 11,000-year-old Swedish dog mentioned above) that were already there. In general, dogs found at archaeological sites in north and western Europe have more eastern ancestry, and less Levantine ancestry, than dogs found in the south and east—and vice versa.

Some dogs were on a very long leash

Dogs’ and humans’ stories match up, at least in the broad strokes. But Bergstrom and his colleagues found a few points where the story of dogs seemed “decoupled” from ours. Those differences are probably the result of disease, trading, preferences for particular dog types, or people moving to a new place without taking the dogs (which sounds awful, honestly). These “decoupled” population histories can tell us about how dogs fit into ancient human societies.

A few thousand years after the Neolithic takeover of Europe, another group of people swept westward from Central Asia. They probably brought along dogs like the 3,800-year-old animal recovered from an archaeological site on the Russian Steppes.

But while the pastoralists from the steppes added their DNA to the mix that makes up modern European populations, their dogs didn’t seem to mingle much with local dogs. Meanwhile, in China, the reverse happened. Steppe pastoralists expanded eastward, but modern people in east Asia don’t carry much of their DNA. Modern East-Asian dogs, however, get quite a bit of their ancestry from dogs like the 3,800-year-old Srubnaya dog.

“Perhaps there is sometimes also an element of chance in these processes, such that if we could replay the tape of human history many times, the outcome for dogs might not always be the same,” Bergstrom told Ars.

Old dogs and new genomes

Part of the reason the earliest years of dogs’ domestication are so fuzzy (not sorry) is that ancient dog DNA has been pretty scarce. Until the recent study, scientists had published just six prehistoric dog and wolf genomes. In case you’re keeping score, we had sequenced more Neanderthal genomes than prehistoric dog genomes—until now, that is.

“Ancient DNA is still a young field, and for most animals there have not yet been many studies of whole genomes,” Bergstrom told Ars. For him and his colleagues to add 27 ancient dog genomes to that list, it took an international effort by archaeologists and museum curators. The collaborators found ancient dog remains in museum and university collections and on lists of material excavated at archaeological sites.

According to Bergstrom, more ancient dog genomes, along with more archaeological evidence about how dogs fit into ancient cultures and economies, could help us understand the origin of dogs and the parts of our shared history that don’t seem to line up.

Perhaps someday we’ll even learn the answer to the most pressing question of all: “Who’s a good dog?”

Science, 2020 DOI: 10.1126/science.aba9572 (About DOIs).

KIONA N. SMITH  is a freelance science journalist at Ars Technica.