This article was originally published by Knowable Magazine.
Updated at 11:21 a.m. ET on May 9, 2022.
They say dogs are man’s best friend, but horses could also claim that title.
Horses gave us a way to transport people and goods—literal horsepower. They changed warfare: drawing chariots, carrying cavalries. They’ve inspired artists from Stone Age cave painters to the makers of My Little Pony. Their role in industry may have waned in favor of machines, but they still maintain a place in sport, in leisure, and in our collective hearts. Horses have been intertwined with human culture since at least 2000 B.C. and were associated with certain human groups even earlier.
“Horses are the animal that has changed history,” says Ludovic Orlando, a molecular archaeologist at the University of Toulouse III-Paul Sabatier in France.
Today, horse breeds number in the hundreds, from the high-stepping Lipizzan horses of Austria to the Clydesdale draft horses of Budweiser commercials to the thoroughbreds of the Kentucky Derby. Despite their differences, these animals are all Equus caballus, joined in the modern equid family by donkeys, zebras, and the wild Przewalski’s horses (pronounced shuh-VAL-skees) of Central Asia (though some taxonomists prefer the name Equus ferus for wild horses, and classification of Przewalski’s horses can vary).
The evolutionary path leading up to Equus is a classic model of evolution—a thoroughly documented history that graces textbooks and museum exhibits. But until lately, the route to domestication by people has been a black box. The bones of E. caballus all look pretty much the same, whether wild or domestic, so they couldn’t answer a long-standing question: Where and when did humans first domesticate horses, linking the two species on a road that would lead to horse-drawn carriages, horse racing, and so much more?
Today, a revolution in the study of DNA, from both ancient and modern creatures, is providing answers. Applying the same approach used in a landmark 2010 study of Neanderthal DNA, scientists have learned much about the history of Equus caballus. They have tracked how ancient wild horses shared genes across the Bering Strait between Asia and North America, and revealed the surprising history of the Przewalski’s horse. And working with more modern samples, they have observed how recent management by people has undone much of the diversity in horse genomes, while adding a host of breed-specific features.
But there has never been quite enough ancient DNA to answer the question of domestication—until late 2021, when scientists reported their analysis of more than 250 ancient horse genomes.
“It’s great to have this big piece filled in, in the puzzle of where horses actually came from,” says Jessica Petersen, an animal geneticist at the University of Nebraska-Lincoln who wasn’t involved with that particular mystery. But, she adds, the domestication process was a complex series of events, and more intricate details will be difficult to uncover.
Sifting through fossil bones and teeth, paleontologists have traced the ancestry of horses back roughly 50 million years to a dog-sized, hoofed animal called Hyracotherium—a.k.a. eohippus, the “dawn horse.” The genus Equus, as we know it, probably emerged between 4 million and 4.5 million years ago in the continent that would become North America. (That’s well before the Homo lineage, which wouldn’t hit the scene for at least another million years.)
Fast-forward to the late Pleistocene, 11,700 to 129,000 years ago, and horses were trotting back and forth between Asia and the Americas on the Bering Land Bridge. The line leading to modern-day domestic horses and wild Przewalski’s horses split sometime in the middle of that epoch, from 35,000 to 50,000 years ago, Orlando says.
But about 11,000 years ago, around the time the Bering Land Bridge submerged for the last time, the North American horses went extinct, along with many other large species such as mammoths and giant beavers. While it’s hard to pinpoint a reason, a factor might have been climate, hunting, or a combination of the two, says Alisa Vershinina, a geneticist at LifeMine Therapeutics in Cambridge, Massachusetts, who investigated the Bering crossings while working as a researcher at UC Santa Cruz.
Early humans would have seen horses around, and they were clearly interested in the majestic animals: Horses are the top animal depicted in Stone Age, Western European cave art. But there’s a big difference between observing the animals for artistic inspiration and harnessing them for horsepower, transport, and sport. When, and where, did this relationship between human and beast undergo a dramatic change?
Horses were a late addition to the barnyard. Dogs were domesticated 15,000 years ago; sheep, pigs, and cattle, about 8,000 to 11,000 years ago. But clear evidence of horse domestication doesn’t appear in the archaeological record until about 5,500 years ago.
Horse remains from across Eurasia gave scientists several candidates for the first domestication event. For example: In 2018, scientists found a frozen, mummified horse in modern-day Siberia. It was dated to about 4,600 years ago.
Iberia, the peninsula containing modern-day Spain and Portugal, seemed promising because horses have continuously inhabited the region for the past 50,000 years and would have been available for potential domestication, Orlando says.
And in the part of Eastern Europe around the Caspian Sea, archaeologists noticed horse remains appearing alongside those of other domestic animals. Human burials about 6,000 years ago began to contain maces decorated with horse heads, perhaps indicating some change in human-horse relations. This area also got attention because of a long-term horse presence in the area.
But an archaeological site that captivated many horse-domestication researchers was the 3500 B.C. settlement at Botai, about 1,000 miles northeast of the Caspian, in modern-day Kazakhstan.
The diet of the people in Botai seems to have been “entirely focused on horses,” says Alan Outram, a zooarchaeologist at the University of Exeter in England. Aside from a few dog bones, those of horses make up the majority of nonhuman remains on the site. There’s evidence of fenced yards that might have held herds. Some skulls hint at slaughter by an ax-like tool, and some horse teeth exhibit “bit wear,” as if they’d been bridled. Pottery shards contain chemical traces of mare’s milk, which Outram says might have been consumed as butter, yogurt, or cheese.
Nonetheless, the site’s importance has been hotly debated. There’s no way to confirm that Botai inhabitants fully domesticated horses. Outram suspects that the Botai peoples treated the horses somewhat like how modern reindeer herders use their animals: They may have kept the horses near at hand for meat and milk, and maybe even rode a few of them to herd others. But they probably weren’t managing breeding or using the animals extensively as pack or transport animals.
And without enough ancient DNA, there was no way to be sure these were the horses that spread around the world as human-managed livestock.
Then Orlando, Outram, and colleagues analyzed a broad set of horse genomes, from as far back as about 42,800 years ago all the way up to 18 modern breeds, publishing the findings in the journal Science in 2018. The result: Today’s ponies, draft horses, and their ilk have little in common with the Botai horse bones. “They’re not the genetic origin for modern domestic horses,” says Outram.
The Botai lineage does live on, though. Unexpectedly, the researchers were able to draw a direct line between those 5,500-year-old bones and modern Przewalski’s horses. These stocky animals with short, bristly manes live on the steppes of Mongolia, where they’re called takhi, or “spirit,” and considered a national symbol.
In other words: Przewalski’s horses, once considered the remnants of an eternally wild population, may not be completely wild after all. Rather, they seem to be the feral descendants of horses that people at Botai might have managed, to some extent, but later lost control of. They’d have that in common with other feral populations such as the mustangs of the American West and Australian brumbies.
Przewalski’s horses aren’t much good for riding, notes Arne Ludwig, an evolutionary geneticist at the Leibniz Institute for Zoo and Wildlife Research in Germany. Perhaps, he speculates, that’s why they fell out of use.
Whatever happened after Botai, Przewalski’s horses had a rough go of it. They nearly went extinct, with the last wild animal disappearing in 1969. Today’s population, all descendants of a handful of animals that persisted in captivity, now number around 2,000 individuals in captivity or natural reserves. There are also a few modern domestic horses in their family tree.
Despite these advances, when Orlando documented genomic studies of horse domestication for the Annual Review of Genetics in 2020, he was forced to conclude that the “geographic origin of modern domestic horses is presently unknown.”
But the clues were building up. Scientists had already nixed the Iberian and Siberian candidates: When researchers looked at ancient DNA, they found that those horse populations had withered away, contributing little to the modern domestic lineage.
Getting to the true domestication site was a numbers game, says lead author Orlando: “We built the answer by narrowing down the evidence, little by little.” The more than 150 collaborating scientists, including Outram and Ludwig, kept adding more horse genomes, from across Eurasia and spanning about 50,000 to 200 B.C.
With 264 ancient horse genomes in hand, the answer was undeniable: The homeland of modern domestic horses was the part of Western Eurasia between and north of the Black and Caspian Seas, more specifically known as the lower Volga-Don region. The team reported their results in Nature in October 2021.
While the data point to a clear answer, there’s still plenty of room for interpretation and speculation. Pinpointing that spot near the Caspian doesn’t mean it was the only place—and time—that people bent horses to the bridle. The genomic and paleontological evidence from the other candidate regions suggests that horses may have been domesticated multiple times, at Botai and elsewhere, without leading to widespread horsemanship.
“It shows how important horses were to people, that so many groups of people independently domesticated them,” says Beth Shapiro, an evolutionary biologist at the University of California, Santa Cruz, and a co-author of the Nature study.
The 2021 analysis does suggest that the domestication in the Volga-Don was the only one that “took”—the only one that spread like horse-drawn wildfire. Why were horses one of the last animals to be domesticated, and why these horses in particular?
While it’s impossible to be certain, ancient genomes suggest tantalizing hypotheses. The lineage leading to modern domestic horses included a change near a gene called GSDMC. In people, alterations to this gene are linked to back problems. It’s possible that the domestic-horse version gave the animals stronger backs, suitable for long-distance riding.
The domestic horse line also includes a change near a gene called ZFPM1. This gene is important in mood regulation. Perhaps some domestic version of ZFPM1 made the animals in the region more docile, easier to tame. These changes could have been the key to long-term horse domestication—but that’s all speculation, says Shapiro.
As for who did the domesticating, it hasn’t been possible to narrow it down to one culture, says Orlando. Diverse people from the region may have started experimenting with these horses about 4,200 years ago. Domestic horses spread a bit from that point on, but things didn’t really take off until about 2000 B.C.
The scientists tentatively attribute the explosion in horse-based transport and technology to the warlike Sintashta culture, which inhabited the northern Eurasian steppes from 2100 to 1800 B.C. The Sintashta traveled back and forth between Asia and Europe, Outram says, and presumably picked up horses on one of these journeys. Later they got around by riding, or via the horse-drawn chariots found in their graves.
This was during the Bronze Age, and it’s thought that horses carried the people of these steppes far and wide, along with cultural accessories like advanced metalworking, lightweight spoked wheels, and Indo-European languages. Anthropologist David W. Anthony of Hartwick College in New York suggested in his 2007 book, The Horse, the Wheel, and Language, that steppe dwellers invented the spoked wheel that made their horses useful for carting cargo and chariot-based warfare. The prestige of the horses and metal goods, along with the chariots for raiding, would have helped these elements of steppe culture, and the Proto-Indo-European language, to spread.
The result was that in later centuries, horsepower and spoked wheels became commonplace, and languages as diverse as Punjabi, Polish, and Welsh can be traced back to the same root.
That domestication event was just the beginning of a relationship between people and horses—and between people and the horses’ DNA. Human management can do striking things to animal genomes over millennia.
For example, all the Y chromosomes of modern domestic horses—passed only through the male line—are nearly identical. To track how this happened, Ludwig, Orlando, and colleagues examined the Y chromosomes of 96 Eurasian stallions from the past 5,000 years. While Y chromosomes started out rather diverse, over time they became more similar, with big changes starting about 1,500 years ago. This corresponds to when certain bloodlines, such as Oriental horses, became popular for breeding, says Orlando.
But even that level of breeding is mild compared with what’s happened in the last 200 years. The diversity of the horse genome has since dropped further, even as specific breeds acquired genes that create their defining characteristics.
Petersen, of the University of Nebraska-Lincoln, has studied modern breeds to identify regions of the genome associated with color, speed, gait, and size. For example, changes to the gene for the muscle protein myostatin are known to occur in racing breeds such as thoroughbreds. Petersen also looked at the “gaited breeds” with unusual styles of locomotion—in her words, “horses that move funny”—which are often more comfortable to ride. These breeds typically show DNA changes in a particular spot, which acts as a sort of “master switch” for gait. That spot contains a gene called DMRT3; a shortened version of the protein it encodes has been linked to horse gait. And mice that lack this gene altogether have problems with how their spinal nerves function.
Genomes aside, there’s undoubtedly something special about the horse-human connection, says Orlando, who took up riding lessons in 2019.
“You have the feeling, when you ride, that this animal understands you and you understand this animal,” he says. “You also have this feeling of mastering this big animal—it makes you feel powerful.”
With that opportunity, how could our human ancestors possibly have resisted making horses their best friends, in peace and war, in work and leisure? It was a match for the ages.
* This article originally misidentified the direction of the Botai archaeological site.