Attempt to Redirect Dog Evolution Falls Flat
Researchers use computer models to bolster their history of early dogs.
Posted Feb 10, 2015
A report in Discovery News and other media outlets that a trio of researchers had resurrected a nearly dead theory of dog domestication came across my computer screen on Thursday, along with several emails from readers questioning its validity. They had good reasons to do so.
The researchers’ analyses of three-dimensional computer reconstructions of the fossil skulls of two ancient canids contradicts the view that dog domestication occurred in the camps of hunters and foragers some 15,000 to 35,000 years ago, a period known in human prehistory as the Upper Paleolithic and in geological time as the Late Pleistocene. Instead, they claim in their paper, published online Thursday in the journal Scientific Reports, that dogs emerged relatively quickly in the Neolithic Age, meaning less than 10,000 years ago, when people settled into permanent settlements and took up agriculture. Wolves drawn to their garbage dumps became self-tamed dogs, they say, citing Raymond and Lorna Coppinger’s book Dogs  and the mid-20th century experiments of Russian geneticist Dimitry Belyaev in which Siberian farm foxes were bred to look and behave like obsequious dogs.
Without commenting on the validity of 3-D modeling in analyzing skulls and other skeletal structures, it is possible to say that the researchers appear profoundly to misunderstand and misrepresent the evidence for emergence of dogs in the Upper Paleolithic. That allows them to reach a conclusion unsupported by evidence.
Drake and her colleagues say that a canid skull from Goyet Cave in Belgium dating to around 36,000 years ago and another from Eliseevichi, Russia, dating to 13,500 years ago provide essential underpinning to the argument for the appearance of the dog in the Late Pleistocene. Remove them from consideration as early dogs and the argument collapses, they say.
But the Goyet dog, as it is known, and two other fossil skulls from the Belgian site have always been difficult to classify genetically and morphologically. They were not used to date the origins of the modern dog for a recent study by an international team of canine geneticists. The researchers sequenced mitochondrial DNA from ancient wolves and doglike canids, as well as modern dogs, and reported their findings in the November 15, 2013, issue of Science magazine [paywall].
“Given their mitochondrial distinctiveness, the Belgian canids, including the Goyet dog, may represent an aborted domestication episode or a phenotypically distinct, and not previously recognized population of gray wolf,” wrote Olaf Thalmann, a geneticist at the University of Turku, Finland, and first author on the team’s paper.
Nor does it matter, other than for purposes of identification, what kind of wolf or dog the Eliseevichi canid was. It was found in circumstances indicating that it had a relationship with human hunters. Indeed one of the difficulties in identifying early dogs is separating them physically from wolves because of frequent crossbreeding, especially in the early days of domestication. Indeed, Thalmann and his colleagues suggest that Eliseevichi canid might represent another failed attempt at domestication.
By their own admission, Drake and her colleagues did not make a three-dimensional representation of the 33,000-year-old skull from Razboinichya Cave in the Altai Mountains that some experts feel is an ancient dog. But because of its uncertain nature, it too is not factored into the search for the ancestors of dogs alive today. Rather, it is taken as strong evidence that attempts by humans and wolves to get together were probably fairly common, something I proposed in How the Dog Became the Dog.
Unlike what Drake and her colleagues say, the case for emergence of dogs in the Late Pleistocene/Upper Paleolithic does not rest on any of those disputed fossils or dates. Current genetic and archaeological evidence points to emergence of dogs 15,000 to 17,000 years ago in Northern Europe, as hunters, guards, food, sacrificial animals, haulers, and companions.
The Bonn-Oberkassel, Germany, archaeological site has produced remains of a dog buried with two people some 14,000 years ago—late in the Upper Paleolithic. For some years that has been taken as the earliest dog, but in a 2014 essay in the Journal of Archaeological Science, Darcy Morey, one of the top archaeologists specializing in early dogs, identifies a number of other dogs dating to 14,000 to 15,000 years ago from Germany, Switzerland, and Southwestern France [paywall]. He takes as transitional two dogs from Ukraine and Germany and concludes that 16,000 to 17,000 years ago represents a credible estimate for dog domestication.
Thalmann and his team had earlier said that the genetic and archaeological evidence favored a time frame older than 15,000 years ago, during the Last Glacial Maximum, in multiple places. The association of wolves and members of the Homo genus may go back further, especially if the transition to dogs took time.
In addition to misreading or ignoring the data on the emergence of dogs more than 15,000 years ago—and that date is not itself written in stone but will probably move back, as many dates of origin do—the nature of the dog progenitor remains an open question. Was it a now extinct gray wolf or a wolf who was ancestral to gray wolves and dogs?
Those questions seem of no interest to Drake and her collaborators; in fact, they seem even to doubt that wild wolves could have made common cause with humans for any length of time. Drake is quoted in a Discovery News story as saying: “Wolves are far too dangerous to have around without adequate means of controlling them.”
Raymond Coppinger tells Discovery News that wolves are hard for him to tame and thus they cannot be dogs’ forebears.
But there is a long history of humans and wolves working and living together—far longer than the history of dogs. In one recent example, Jim and Jamie Dutcher brilliantly recount in The Hidden Life of Wolves their six years living among wolves in Idaho’s Sawtooth Mountains. The photos in this blog are from the book, chosen to show that wolves do not kill or even harm everyone they meet. They also show that wolves choose whom they will meet and spend time with and when. They do not threaten or attack everyone they meet.
Drake et al.’s article and the press release about it—which serves as the basis for many of the "news" accounts—need much more data to justify their radical conclusions. As they now stand, they prove nothing except that even bad science can get published and, if outrageous enough in its claims, garner headlines.