Horses may have been domesticated twice. Only one attempt stuck

Horse power may have revved up about four millennia ago.

Horses were domesticated at least twice, researchers report June 6 in Nature. Genetic data suggest Botai hunter-gatherers in Central Asia may have been the first to domesticate the animals for milk and meat around 5,000 years ago. That attempt didn’t stick. But other people living north of the Caucasian Mountains domesticated horses for transportation about 4,200 years ago, the researchers found.

Those latter horses took the equine world by storm. In just a few centuries, they replaced their wild cousins and became the modern domestic horse.

The findings call into question some long-held ideas about the when, why and who of horse domestication, says Ludovic Orlando, a molecular archaeologist and director of the Center for Anthropobiology and Genomics of Toulouse in France. For instance, ancient people from southwest Asia known as the Yamnaya have been credited with being the first horseback riders (SN: 3/3/23).

The Yamnaya were pioneers who hitched up cattle-drawn carts and left increasingly dry grasslands about 5,000 years ago to make new homes in Europe and Asia. Along the way, they helped build major Bronze Age cultures in Europe (SN: 11/15/17). They spread Indo-European languages and left a genetic legacy for modern people that includes increased risk of multiple sclerosis and Alzheimer’s disease (SN: 9/5/19; SN: 1/10/24).

But none of that happened on horseback, Orlando and colleagues argue. The timing just doesn’t work.

The researchers examined DNA from 475 ancient horses that lived as far back as 50,000 years ago and 77 modern horses. Combining that genetic analysis with carbon dating and archaeological data, the team established a timeline for horse domestication.

DNA data suggest horses were domesticated later than previously thought

Researchers already knew that domestic horses galloped off the steppes of what is now southwestern Russia and began spreading around Europe and Asia, replacing wild horses (SN: 10/20/21). The new genetic data show that happened about 4,200 years ago. Before then, “there are many bloodlines that you see around,” Orlando says. “But from 4,200 years ago, that bloodline that was north of the Caucasian range becomes global.” The speed of the spread suggests people domesticated horses with mobility in mind, he says.

If the Yamnaya people and the horses were migrating together, their genes would have spread at the same time, “because you would be literally on their backs,” Orlando says. But the horse genes didn’t start spreading until about 800 years after the Yamnaya migrated.

The idea that Yamnaya weren’t horseback riders and herders is “potentially a difficult pill to swallow for a lot in the science community,” says William Taylor, an archaeozoologist at the University of Colorado Boulder who was not involved in the study. The new research overturns the idea that Yamnaya were horse people. “The animals we know today as domestic horses did not have a presence in Yamnaya culture,” he says. “This is a hard reality that genetic evidence is able to provide.”

Some researchers say the finding ignores earlier evidence of horsemanship and attempts at domestication. And, says archaeologist Volker Heyd of the University of Helsinki, the Yamnaya would have needed horses in order to spread so quickly. “Our best scenario for the rapid and extensive Yamnaya expansions, covering 5,000 kilometers and more in 100 to 200 years, [is for it] to have been facilitated by wheel and wagon and on horseback.”

Not so, says archaeologist Ursula Brosseder of the Leibniz Zentrum für Archäologie in Mainz, Germany. “There is a general mistake in assuming that migration needs horses. [But] humans throughout history have done their migrations mostly not with horses but on foot.” Even walking, people can cover 1,000 kilometers in a month, she says.

Genetic evidence shows when people purposely started breeding horses

Using a new technique, Orlando’s team found that as the horses began to spread, their generation time fell from just over seven years to about four years. And other genetic evidence suggests that closely related horses bred. Neither of those things happen naturally, Orlando says. The evidence points to people controlling horse breeding to increase numbers and to select for certain traits, clear signs of domestication (SN: 7/6/17).

The researchers also found the shortened generation time in 5,000-year-old horse remains associated with the Botai culture of Central Asia. Previous research suggested that the Botai may have milked and bridled horses (SN: 3/5/09). The short generation time could be an indication that the Botai were domesticating prey horses to pump up their meat supply, Orlando says. If so, it could be the first attempt at horse domestication, although one that ultimately wasn’t successful. The only living relatives of Botai horses are wild Przewalski’s horses, six of which were included in the genetic analysis of modern animals (SN: 2/22/18).

Brosseder says that the shortened breeding time is “very convincing” evidence that the Botai were using horses for a specific purpose, which could be considered domestication.

But Taylor doesn’t buy that the Botai domesticated horses. He’s been “squinting at the available evidence and thinking about what the archaeology shows,” he says, and concludes that what was happening with Botai horses “was the last hurrah of a hunter-prey relationship with horses … that really didn’t have anything meaningful to do with domestication.”