In 1912, Norwegian explorer Roald Amundsen and his team fought more than exhaustion and cold on their famous quest to become the first humans to reach the South Pole. The rations they had packed for the arduous dog-sled journey across Antarctica proved insufficient, so Amundsen and his men decided to shoot and eat some of their dogs. The explorer later described the fare as delicious, adding that "it is anything but a real hardship to eat dog flesh."
Amundsen may have come up with the idea after hearing stories of aboriginal hunters in Greenland eating their sled dogs in winter. But just how long have humans regarded Fido as food? In a paper published online this month in the American Journal of Physical Anthropology, a team headed by geneticist Raul Tito at the University of Oklahoma in Norman reports finding a dog bone in 9260-year-old naturally preserved human feces. According to team member Samuel Belknap III, a graduate student at the University of Maine, Orono, it is the earliest incontrovertible evidence for domestic dogs in the New World. "And I feel fairly confident that it's the oldest direct evidence of human consumption of dog in the world," notes Belknap.
Belknap discovered the bone while identifying the contents of ancient human feces excavated from a lower layer at a rock shelter known as Hinds Cave in Texas. The ancient fecal material, or coprolite, was littered with seared prickly pear seeds—a food prepared and cooked by humans—and flecked with small bones from fish, birds, and rodents. Belknap was initially surprised to find the bone of a larger mammal. Further analysis suggested that it was part of a dog's skull—the occipital condyle, a knoblike structure on the back of the head, near the first vertebra—and studies by fellow University of Maine graduate student Robert Ingraham at the Harvard University Museum of Comparative Zoology showed that it closely matched that from a Native American dog collected in New Mexico. That dog weighed about 25 to 30 pounds in life and possessed a short nose. "Our dog probably represents the ancestor of those dogs," says Belknap.
The team obtained the date of 9260 years ago for the coprolite by radiocarbon dating one of the prickly pear seeds inside. But the early date, says Belknap, necessitated a more definitive identification of the bone. Research teams elsewhere had advanced claims for other dogs in this time range in the New World, but critics had frequently disputed the findings. The best evidence of an early dog in the New World came from a 9400-year-old skeleton from the Koster site in Illinois, but the dog was dated only in association with charcoal from a hearth, and the identification was based on just one line of evidence: skeletal characteristics.
To conclusively identify the Hinds Cave bone, Tito and his colleagues at the University of Oklahoma's Molecular Anthropology Ancient DNA Laboratory took two small samples from the bone for DNA testing. They sequenced a 368 base pair fragment from the mitochrondrial genome and then compared it to sequences from ancient dogs, as well as to sequences from modern dogs, wolves, and coyotes. The Hinds Cave bone fell into Clade 1, a group belonging solely to modern and ancient dogs. "This data predates any other genetic data on dogs in the New World by 7800 years," says team member Cecil Lewis Jr., a geneticist at the University of Oklahoma.
Jennifer Leonard, a geneticist at the University of Uppsala in Sweden who was not involved in the project, thinks the findings are solid. "I was impressed by the strength of their evidence that people were eating dogs at this time." But she is not terribly surprised that ancient humans saw man's best friend as more than a companion or hunting assistant. "Most Americans living in cities and suburbs don't see their dogs as food, but there are lots of historical records of Native Americans eating dogs," she says.