When is a wolf not a wolf?
The skull of the ancient dog from Goyet Cave, Belgium
In today’s issue of Science
, one of the most prestigious journals, various scholars—some good friends of mine—write about the origin of dogs. Dogs are very good at interacting with humans, highly trainable, amusing, loyal, and loving. Sometimes they are very beautiful and they take great joy in simple things. Of course, dogs are notoriously our best friend—and wolves are not. Wolves are in fact such a feared and hated enemy that humans have tried repeatedly to wipe the species out. Wolves are clever, large, ferocious, efficient, wild and deadly. They are also beautiful, good parents
, amazing singers, and a symbol of wildness.
So how do you tell the difference?
We know by simple observation, genetic studies, and obvious resemblances that dogs are descended from wolves. They are a very special type of wolf that humans, over long periods of time, shaped to be companions, guardians, assistants, and something very much like family. Being a dog is a genetic issue and the great variety of dogs in the world show that humans have been able to select for many different traits – such fast running, ability to swim, an incredible sense of smell, size, color, and many other things. But all this took time and careful breeding, by humans that had no other example of what domestication could do. Dogs were the first—and some would argue, the best (though I would personally contend that cats and horses are quite as good)—domesticated animal ever.
While outstanding geneticists were trying to track down where wolves were domesticated into dogs, and estimate the time when dogs first arose, paleontologists who study the anatomy and morphology (shape) of fossils were asking a similar question but using different tools. Led by Belgian paleontologist Mietje Germonpre, a team of morphologists began working out statistical ways to measure shape differences in the skulls, and later in the jaws, of fossil canids to determine which features distinguished dogs from wolves. They were able to show that combining certain measurements of the skulls of canids neatly separated known modern dogs from known modern wolves and relatively recent fossil dogs from fossil wolves. And there, plunk in the space between prehistoric or fossil dogs and modern dogs was a group of fossil canids that made their own group, with a high degree of probability.
The biggest surprise came when those fossil canids, which looked for all the world like the earliest dogs, were dated and found to be 32,000 years old, not 15,000 years old which was everyone’s best guess about the timing of the origin of dogs. Additional work with fossil canids enlarged the sample of early dogs, creating a sort of chain of doggy specimens from 32,000 to 15,000 years ago. Of course, many specimens that were examined were judged to be wolves from this period too.
Enter the geneticists. Working with Robert K. Wayne, a renowned dog geneticist, Beth Shapiro, an expert in retrieving ancient DNA, and an international group, Otto Thalmann of the University of Turku in Finland began working on getting DNA out of the fossils. They already had a large number of DNA samples from modern dogs of a wide range of breeds and of modern and ancient wolves from many places. The fossils – as fossil often do—were difficult. Preservation was less than ideal. They abandoned the attempts to get nuclear DNA from those samples and turned to mitochondrial DNA (mtDNA).
The advantages of mtDNA are twofold. First, there is much more of it in each cell than there is nuclear DNA because there is only one copy of nuclear DNA while mtDNA is found in every mitochondrion, an energy-producing organ tht ias abundant in the cell. Second, mtDNA is transmitted down the female line, from mother to daughter to granddaughter and so on. So, although most mtDNA lineages go extinct in the course of time, there is no confusion caused by the recombination of mother’s DNA with father’s DNA. mtDNA gives you a solid and unbroken lineage of females back through the ages, if it survives at all.
Compiling their very large mtDNA dataset, the geneticists were able to produce a phylogenetic tree, a diagram of resemblances showing how the mtDNA evolved. Many fossils’ mtDNA grouped with different groups of modern dogs from Europe. This finding suggests strongly that the ancestors of modern dogs lay among European wolves.
However, the mtDNA of the three ancient Belgian canids was not matched by any other sample the team knew, out of hundreds. That mtDNA did not say “dog” and it didn’t say “wolf” either. It said something like “very early canid like no other we know.”
How do we decide if those animals were dogs or wolves? Genetically, they seem to be a sister group to (not an ancestor of) modern dogs and wolves. Given that mtDNA is passed through the female line only, the mtDNA evidence is entirely compatiable with a scenario in which the early domesticated male dogs interbred with female wolves, creating a new admixture that eventually led to dogs or eventually went extinct. Oddly enough, in one important sense what makes a dog a dog and not a wolf is its relationship to humans and it is hard to read behavior from fossils.
Dogs, as I said earlier, respond to people, watch them for cues to desired behaviors, work with them, follow their gestures, and so on. If these ancient canids were doglike in behavior, then it would swing the argument in favor of their being an early attempt at a domesticated dog. If they were wolflike in behavior, then they were functionally more like wolves.
What would we see in the archaeological record if these 32,000-year-old canids were dogs? Judging from studies of modern hunters, we’d see a dramatic increase in hunting success—more meat per hunt, more meat per hunter, less energy expenditure to take big animals. That is exactly what we do see.
In the Aurignacian, Gravettian, and later Epigravettian evidence left by early modern humans, we see bigger sites, more complex sites with different components (cooking areas, tool-making areas, areas for skinning and dismembering prey, and kill sites), and probably a longer occupation by more people. More startlingly, about the time that these unusual canids lived, we suddenly see a whole new type of site, one in which dozens of even hundreds of mammoths dominate the assemblage of prey species. The other highly unusual aspect of these remains is the presence of many Arctic foxes and hares—skinned for fur and meat—and many large canids. Probably most of the large canids are wolves, which is what the morphology of their skulls and jaws suggest, but some look like the Belgian canids: dogs or proto-dogs.
It is obvious why hunting with dogs or proto-dogs would be helpful in killing and retaining possession of the carcasses of enormous mammoths. Dogs find prey faster; they can hold a large animal in one spot by barking and charging while humans come up to complete the kill; they can stand guard over the campsite.
But why would living with early domesticated dogs result in more wolf kills? Because dogs, like wolves, are very protective of their “pack”—humans and all—and are ferociously aggressive with canids from other groups (strange wolves) that threaten their pack and its food supply. Yes, these wolves were skinned and probably eaten, too, in most cases. But more than that, they were threats to the wellbeing of the joint human-dog pack and had to be eliminated.
Whether or not these early and unusual group of canids were ancestral to all modern dogs, I argue that the evidence tells us they behaved like dogs, helping and protecting and living with early humans and enabling them to take far bigger prey than they ever could before.
You could say a dog is as a dog does.
When is a wolf not a wolf?