Researchers chasing the origin of modern dogs find that canines were domesticated once, between 20,000 and 40,000 years ago.

The results, published on 18 July in Nature Communications1, push back against a controversial 2016 study that suggested dogs were domesticated twice. The latest analysis also add weight to previous research that moves the timing of domestication back as far as 40,000 years ago.

Everyone has their own idea about where and when dogs originated, says Krishna Veeramah, a palaeogeneticist at Stony Brook University in New York and an author on the latest study. “Archaeologists suggest one and geneticists suggest another — people are always getting very different answers.”

Zeroing in on a precise time and location for dog domestication has long been challenging because of seemingly contradictory or incomplete evidence. A 14,700-year-old jawbone is the oldest undisputed fossil from a domesticated dog (Canis lupus familiaris), but dog-like remains date back as far as 35,000 years ago. Genetic data show that the ancestors of all modern dogs split into two populations: one that gave rise to East Asian breeds and another that would become modern European, South Asian, Central Asian and African dogs. Yet researchers still can’t pin down when this split occurred. And they can’t agree on whether dogs were domesticated once or twice.

Veeramah and his colleagues studied genomes from Neolithic dog fossils found in different parts of Germany — one from the start of the period, 7,000 years ago, and one from 4,700 years ago. They also looked at data from a 5,000-year-old dog specimen found in Ireland. The team then compared these ancient genome sequences with genetic data from 5,649 canids, including modern dogs and wolves.

The researchers estimate that dogs and wolves diverged genetically between 36,900 and 41,500 years ago, and that eastern and western dogs split 17,500–23,900 years ago. Because domestication had to have happened between those events, the team puts it somewhere from 20,000 to 40,000 years ago.

. . . .The authors of the latest work acknowledge that their study won’t settle the debate over when and where man’s best friend originated. “More ancient dog DNA from genomes will ultimately solve the problem,” Veeramah says.

“If we can add in other ancient samples from all around the world, it'll give us a more comprehensive picture of population history and likely dog origins,” says Adam Boyko, a geneticist at Cornell University in Ithaca, New York. But we need diverse samples not just geographically, but also through time, says Boyko, who is compiling a modern dog genetic database from canines all over the world.

. . . .“Dogs and humans have an important history together,” [Veeramah] says.

By around 2200–2000 BC, these horses had appeared outside the Western Eurasian steppes — first reaching Anatolia, the lower Danube, Bohemia and Central Asia, and then spreading across Eurasia, replacing all other local horse populations by about 1500 to 1000 BC. “We found that around 4,200 years ago, the horse reproductive pool dramatically expanded, indicating that this was when past breeders started to multiply such horses in large numbers to supply increasing demands for horse-based mobility,” says Orlando. Humans probably rode on horses’ backs before the invention of horse-drawn vehicles: the first spoke-wheeled chariots emerged around 2,000–1,800 BC.


The findings also challenge previously held ideas about the role of horses in some early human migrations. Analyses of ancient human genomes have revealed massive migrations from the Western Eurasian steppes into Europe during the third millennium BC, associated with a culture known as the Yamnaya. These people are thought to have helped to spread Indo-European languages into Europe, and have often been assumed to have ridden horses. “If these numerous people came with as many horses, then we should expect an equivalent shift in the horse ancestry profile,” says Orlando. But the analysis suggests that during this time, there were few domestic-horse ancestors outside the Western Eurasian steppes. This would rule out scenarios in which horses played a part in Yamnaya migration and in the initial spread of Indo-European languages.

“This radically changes our understanding of mass human movements from the steppe into western Europe in the Bronze Age,” says Outram. “It seems that those migrations were not, as had been commonly believed, facilitated by domestic horses.”

“Two variants of the GSDMC and ZFPM1 genes were selected early on during the domestication process, likely facilitating taming, increasing stress resilience and providing horses with a stronger back,” says Orlando. “These qualities may explain why the new horse type had such a global success.”