The Debate on Canine Domestication

Canine Origin Story
By Jane Brackman PhD, February 2017
Jawbone and teeth fragments held at the Oxford Museum of Natural History. DNA extracted from these and similar materials contributes to the effort to track the separation between wolf and dog.

Jawbone and teeth fragments held at the Oxford Museum of Natural History. DNA extracted from these and similar materials contributes to the effort to track the separation between wolf and dog.

Jennings Dog, 2nd century AD; Linnaeus’s Systema Naturae described a new classification system.

Jennings Dog, 2nd century AD; Linnaeus’s Systema Naturae described a new classification system.

Natufian burial of human with dog, Israel, 12.5 to 10.8 BCE; Illustration from Buffon’s Histoire Naturelle.

Natufian burial of human with dog, Israel, 12.5 to 10.8 BCE; Illustration from Buffon’s Histoire Naturelle.

Researchers have identified the origin of cattle, horses, pigs, sheep, camels, ducks, chickens, cats and goats. But the genesis of the domestic dog, our oldest companion and the most varied, numerous and widely distributed domestic animal on the globe? We’re still trying to figure out that one.

The study of patterns of diversity is called systematics, and it is a critical subdivision of evolutionary biology. Systematics researchers (earlier called naturalists and taxonomists) sort out species’ genealogical relationships and estimate the points at which populations diverged from one another. Traditionally, they relied on observations of differences in stable physical traits like teeth, skulls and sometimes fossils. More recently, genome-wide comparisons have been used to provide detailed information about species relationships, including the question of when and where wolves became dogs.

Canis lupus familiaris exhibits the most variability in shape, size, behavior and temperament of any mammal species living on earth. About one billion dogs, a population larger than any other domestic subspecies, roam the globe. Canine fossils, some dating to as long ago as 36,000 years, are found on every continent except Antarctica.

Add to that the unusual phenomenon that extreme variation can occur in as little as one generation—a sort of evolution at hyper speed—and we begin to understand why classifying domestic dogs has challenged many of the taxonomical systems that have been used to make sense of Canidae, a family that includes wolves, jackals, foxes and dogs.

Science Evolves

Historically, as far back as the fourth century BCE, theories of the descent of animals were the product of using philosophical approaches to relate organic life to the history of time. At first, fundamental ideas about species-change involved sorting out living beings by means of their common essential properties. Philosophers wanted to know how organic life forms were related, not where they came from.

The Greek philosopher Aristotle (384–322 BCE) endorsed the idea that natural beings were always here and always would be. He commented on the dog’s origin, not in respect to the animal’s continuous chronological past but rather, in terms of breed creation. In his view, the dog that nature created was bred to the fox to make small dogs and to bears to make big ones, perhaps making the point that breeds (although he was mistaken about cross-species hybridization) were created by humans. Still, in the Aristotelian view, dogs always existed.

As time went on, the earliest naturalists came to understand that species were related in more complicated ways, and began to devise orderly classification systems. The bigger picture of life, however, was explained within a theological context: a specific act of an omnipotent creator transformed all living things whole and complete. The revolutionary notion that every animal might not be a singular divine creation didn’t materialize until the late Middle Ages, a contradiction that had to be explored hypothetically to avoid conflict with religious doctrine.

In the late 18th century, France’s leading naturalist and the father of paleontology, Georges Cuvier (1769–1832), introduced a new way of looking at life and death. Although he was firmly in the camp of divine creationism, he theorized that animals eventually went extinct.

Earlier, 16th-century English cleric Edward Topsell (1572–1625), author of The History of Four-footed Beasts, whose worldview was defined by fire-and-brimstone religion, based his categories on morality. This was not as much of a stretch as it might seem from today’s vantage point; during Topsell’s time, people had real reason to fear wolves. For them, the predatory wolf and sagacious, noble dog provided excellent examples of two moral extremes.

Domesticated farm livestock had derived from prey species, and no other large predator had (or has) been domesticated. So it seemed illogical that the gentle, devoted dog could have evolved from the wolf. As one writer lamented, “How could such a noble animal as the dog be derived from the likes of the wolf? If evolution were true of dogs and wolves, wouldn’t every beast choose to live the noble life?” Indeed.

But as Darwin later observed, if organic beings didn’t possess an inherent tendency to vary, humans could do nothing. Unlike bears and lions, wolves, for reasons still scientifically unclear, possessed the variation necessary for the creation of the multiple hundreds of dog breeds recognized today.

The Shape of Things

Imagine how frustrating it must have been to try to make sense of how dogs were related and where they came from based on their appearance. Travel the world over and a cat will usually look like a cat, but dogs were a vexing contradiction.

The lack of understanding of the complexity of canine morphology made it difficult to unravel relationships among the ever-increasing numbers of dogs and dog-like animals being discovered on far-away, previously unexplored continents. In the Americas, many were likely Old World breeds introduced by European explorers, eventually returned to a feral state. Over time, they interbred with American Indian dogs, wolves and coyotes, defaulting to pariah-type dogs—a catchall term for semi-feral, free-ranging canines. But a misunderstanding of the distinct differences between wild, tame, domestic and feral dogs added to the confusion about how Canidae should be classified.

The father of modern taxonomy, Carl Linnaeus (1707–1778), assigned dogs both wild and domestic to groups based on their anatomy (muzzle, jaw, ear shape), tail carriage (dog tails curve when relaxed, wolf tails don’t), hair texture, limb length and behavior, criteria that are still used today.

Linnaeus’s contemporary, Georges- Louis Leclerc, Comte de Buffon (1707– 1788)—of whom Linnaeus sniffed, “Always eloquent, often incorrect”— suspected that changes in canine morphology were influenced by environmental pressures, such as climate. But, like his colleagues, Buffon did not consider change within an evolutionary context.

Dividing dogs into categories based on skull shape was Cuvier’s idea, and although his forward-thinking approach to paleontology and the history of organisms would seem to make him an advocate for evolution, he was not. Cuvier’s interest, after all, was in a species’ demise, not its origin. Nevertheless, his contributions greatly influenced Charles Darwin.

Darwin (1809–1882) believed that the dog had multiple origins: from wolves, jackals and at least one South American species. He supported the latter by referencing his observations of dogs in Patagonia who swam underwater and an unusual dog he had seen in Central America. He also advanced the idea of multi-regional domestication.

Darwin further imagined that these small populations of “inferior” native dogs were eventually supplanted by the incursion of more robust dogs introduced by Europeans, an analogy he used to demonstrate the idea of “survival of the fittest.”

Although the fundamental theory of origin is attributed to Darwin, other taxonomists previously proposed similar ideas and connections, including Jean-Baptiste Lamarck and Alfred Russel Wallace. Unlike Lamarck and Wallace, however, Darwin suggested that the evolutionary process occurred through natural selection.

Although Darwin used the breeding of dogs and other artificially selected animals as analogies to explain how natural selection worked, dogs continued to be an untidy group of animals —a puzzle that science has began to unlock by the use of genome-wide sequencing.

In the Genes

Once scientists discovered methods to explore origin at the molecular level, they began to test these historical theories. As early as the 1970s, research papers were published suggesting that dogs may have been derived from several different gray wolf populations, and that canine domestication may have happened much earlier than the fossil record’s 15,000 years ago. By the late 1990s, geneticists worldwide were working together to build a comprehensive map that would chart the evolutionary journey of domestic dogs.

The path was not smooth. Differences of opinion erupted and criticism of research methodologies undermined a delicately balanced collaboration process. Numerous studies argued for canine origin in places as diverse as East Asia, Mongolia, Siberia, Europe and Africa, with timing varying from somewhere between 15,000 and 135,000 years ago. Archeologists who’d studied ancient canine burials were relegated to the sidelines, their fossil records dismissed as “old school,” which created further dissention. Researchers struggled to find common ground, but without much success.

The debate ramped up in 2013, when UCLA evolutionary biologist Robert Wayne and his team published a comprehensive set of data suggesting that dogs evolved from a group of European wolves, now extinct, somewhere between 19,000 and 32,000 years ago.

Two years later, Peter Savolainen, a molecular biologist, and his colleagues at the Royal Institute of Technology in Stockholm published convincing results indicating that dogs originated in China, south of the Yangtze River. They estimated that this dog population split from wolves 33,000 years ago.

Both teams were sequencing DNA. Why were their findings literally all over the map?

Savolainen’s research team analyzed DNA samples from living global dog and wolf populations, then tracked DNA from least to most diverse, going back through time. The general rule is that the older a population of animals, the more diversity it has in its genome, which is a hallmark of ancient origin.

Whether these animals represented the first domesticated dogs or, rather, dogs who migrated to the region from elsewhere and split off from a more ancient dog population, is unresolved. Fossil remains of an ancestral and probably extinct population of wolves that would have been indigenous to the area would seal the deal, but researchers have yet to find them. As Savolainen notes, “We have access to some archaeological samples we are about to analyze. However, there has been quite little archaeological work, especially on animals, in the region.”

While Savolainen and his colleagues worked backward in time, Wayne’s group worked forward, tracking ancient DNA collected from prehistoric bones of wolves and wolf-like dogs, then measuring decreasing genetic diversity. As DNA becomes less diverse, it points to animals transitioning from wolves to dogs. A dead end indicates that a lineage became extinct in that particular region.

Wayne’s team sequenced ancient DNA on canid skulls and bone fragments discovered in present-day Siberia and the Czech Republic dating to between 27,000 and 33,000 years ago. The physical characteristics of the skulls—wider muzzles and foreshortened jaws—suggest that these were ancient proto-dogs, not wolves. The canids may have looked similar to today’s Arctic breeds (for example, the Siberian Husky and the Greenland Dog), but were probably much larger. Although their findings were met with skepticism, the team said their data showed that domestic dogs originated from different wolf populations at different times in different places, in a series of starts and stops. And, they added, living dogs are more closely related to ancient extinct wolves than they are to modern wolves.

In an interesting twist, Wayne’s findings reignited the theory of parallel and multi-regional proto-domestication, an idea that Darwin introduced in the 19th century and one that’s gone in and out of favor since.

Both studies have detractors. Some claim that diversity in Savolainen’s ancient dog population is a result of admixture with European dogs as people traversed the Silk Road. Those who criticize Wayne’s study maintain that he has no solid proof that the ancient bones he’s studying are definitively wolf or dog. Additionally, critics say, his study is geographically biased because he excluded samples from dogs in China based on his position that there are no ancient dogs there.

Although the two studies point in very different directions, Savolainen and Wayne may both be right. It’s possible that dogs were domesticated multiple times in different regions, and that most lineages died out when humans were faced with overwhelming challenges, like climate change. Their findings aren’t mutually exclusive.

Crunching the (Very Big) Numbers

One reason for the disparities, according to Oxford’s visionary evolutionary biologist Greger Larson, who was a part of a team that successfully mapped the origin of the pig, is that scientists studying the dog are not including enough ancient DNA in their studies.

Larson and colleague Keith Dobney, an archaeologist at the University of Aberdeen, had the idea to bring together all the evidence collected to date, find ancient canid specimens from museums, apply state-of-the-art technology and create a database bigger than anything produced before. All they had to do was convince scientists to agree to work together.

Fortunately for canine genetics, Larson was able to sell the idea that more cooperation and collaboration improves the outcome. As chief mediator and conciliator, and supported by substantial funding, he has persuaded more than 50 influential canine evolutionary scientists to join the project. Team members include archaeologists, paleobiologists, anthropologists, zooarchaeologists, paleogeologists and others.

The purpose of the study, which began in 2013 and is slated to wrap up this year, is to combine ancient DNA analysis and geometric morphometric techniques and apply them to archaeological canid remains. This, he suggests, will directly address where, when and how many times dogs were domesticated.

Geometric morphometrics, the study of form in two or three dimensions, is a powerful new way to visually quantify evolutionary relationships. It does this by correlating thousands of geometric points that identify exact places on bones—specifically, points of evolutionary significance that differ between very closely related animals such as the wolf and the dog.

Using a special camera, researchers take hundreds of 360-degree photographs. Software then transfers the pictures to a three-dimensional computerized image that emphasizes a set of tightly defined, very specific points on each bone. The process results in holographic- like images that show domestication in progress through space and time, much like a movie.

Additionally, scientists are isolating and examining ancient DNA collected from museum specimens, looking for changes in the degree of genetic diversity over long periods. This will provide a comprehensive overview of the wolf-to-dog transition from the beginning to the present.

No individual genetic fragment of DNA says This is a wolf or This is a dog. Rather, scientists tease the two apart by looking at strands of DNA and identifying and measuring similarities and differences. As differences become more extreme, the separation between wolf, proto-dog and, finally, dog is suggested.

The team hopes to isolate genetic fragments that can be linked to minor changes in the geometric morphometricimaged samples. Combining the two techniques will tell a deeper, more layered and detailed story about canine domestication.

Larson expects to analyze up to 7,000 specimens representing wolves, incipient canids and domestic dogs. “We’re taking samples from all over the world, sources in not only museums but from private collections, too. Curators are very agreeable when we ask for permission, and they’re usually very happy to have us take photos and DNA samples. They help us, and in turn, we provide more information for their collection.”

While Larson is enthusiastically optimistic about the outcome of this unprecedented project, some scientists not affiliated with the study think the findings will only add to the existing mishmash of conflicting hypotheses. But that’s how science works: come up with an answer and you invariably end up with a lot more questions.

The ongoing search to understand where, when and how many times dogs were domesticated continues to be a topic of active scholarly exploration. Besides the millions of dog lovers who are curious about the roots of our affectionate and unusual cross-species relationship, substantial scientific issues are at stake, issues that may profoundly alter the future of evolutionary theory.

Read about new developments.

Jane Brackman, PhD, is an authority on the cultural history of canine domestication and the author of two books on pets in 19th-century America. See her new pup, Barkley, and watch him grow on her blog.