Some We Love, Some We Hate, Some We Eat

by Hal Herzog

  The tools of the archaeologist’s trade are spades, dental picks, and dusty hiking boots, but molecular biologists trying to ferret out the mysteries of canine evolution wear lab coats and spend their days listening to the hum of DNA sequencers. The raw materials for their research are not fossilized bits of jaw and teeth, but fragile strands of genetic material, mitochondrial DNA that has been passed from mother to child over thousands of generations. Mitochondrial DNA is different from the nuclear DNA that makes up the genetic instructions you bestow on your kids when sperm meets egg. Mitochondrial DNA floats in a cell’s cytoplasmic goo, converting organic material into energy. Most important for evolutionary biologists, all your mitochondrial DNA comes from your mother. Unlike regular DNA, which is randomly shuffled during sex, ancestral lines of mitochondrial DNA remain unchanged from generation to generation, except for the occasional mutation.

  Geneticists use the rate of mutational change in mitochondrial DNA as a kind of molecular clock that gives an indication of when, where, and from what a species has emerged. In 1997, an article appeared in the journal Science that shook up the world of people who care about things such as when dogs made the great leap from wolves. A group of researchers led by Robert Wayne of UCLA analyzed the mitochondrial DNA of sixty-seven dog breeds as well as the DNA of wolves, coyotes, and jackals. Based on their interpretation of the molecular clock, they concluded that dogs emerged from gray wolves between 100,000 and 135,000 years ago, ten times earlier than the fossil chronology of the dog-wolf split. That is a huge discrepancy. What’s going on? The accuracy of a molecular clock, just like an alarm clock, depends on when you set it, and the UCLA researchers may have been overly generous when calibrating theirs. Most biologists now believe that dogs began to diverge from wolves between 15,000 to 40,000 years ago, a date that jibes better with the story the bones tell.

  WERE THE FIRST DOGS PETS OR DUMPSTER DIVERS?

  Anthrozoologists disagree about how humans and dogs came to throw in their lots together. One popular theory goes like this: Fred Flintstone, Stone Age hunter, stumbled upon a wolf den one afternoon while foraging for dinner and brought a baby wolf back to the cave to cook for the evening meal. But just as Fred was about to toss little Lobo on the campfire, his wife, Wilma, looked deeply into those adorable puppy eyes and her maternal instincts kicked in. She snatched the puppy from Fred and doted on it as if it were her child, maybe even suckling the pup. The wolf’s fierce nature was soon tamed, in Bill Monroe’s words, “by a good woman’s love.” Voilà, Lobo became the first domesticated companion animal, and soon produced puppies that, over time, came to look and act like Rin Tin Tin.

  Ray Coppinger, a biologist and dog-sled racer who teaches at Hampshire College, is skeptical of this idea, which he dismisses as the “Pinocchio Hypothesis.” Coppinger thinks that converting full-blown wolves into dogs is about as likely as the fairy godmother in Pinocchio waving her magic wand over the wooden puppet and turning it into a flesh-and-blood boy. The trouble with the wolf-into-puppy-dog theory, says Coppinger, is that pure-blooded wolves are essentially untamable. That’s why you don’t see them in circus animal acts. Coppinger admits that they can learn a few tricks and that some become habituated to walking on a leash, but he feels this is a far cry from being truly tame. Other researchers have found that even hand-reared wolves do not become attached to their caretakers the way dogs do. They are more likely to growl and literally bite the hand that feeds them.

  But if humans did not tame wolves, how did they become our pals? Coppinger says they tamed themselves. He traces the emergence of dogs to the time people began to give up the nomadic life and settle into permanent villages. Settlements generate piles of refuse, potential gold mines for opportunistic scavengers. Wolves who were less nervous around humans would be more efficient garbage-pickers, just like the dogs-gone-wild you can see today hanging around garbage dumps on the outskirts of Lagos or Mexico City or Istanbul. These animals would have been better nourished than their warier competitors, hence bearing more offspring—puppies that would share their parents’ genes for being less fearful of humans. Over generations, these tamer animals would have access to more food and eventually adapt to living around people. This self-domestication process would set the stage for unconscious selective breeding for functional traits like barking at strangers and chasing game.

  HUNTING LIZARDS WITH DOGS

  Coppinger may be right. I once got a glimpse of how humans may have used the first dogs to hunt small game. My graduate school friend Bev Dugan spent several years in the jungles of Panama figuring out the sexual strategies of green iguanas. She had a problem: All iguanas look more or less like reptilian versions of Keith Richards, and they spend most of their days in treetops. If her project was going to succeed, she needed to capture and mark the animals in order to tell them apart. But how do you catch a creature that lives in the canopy of a tropical forest?

  That’s where dogs come in.

  Another researcher introduced Bev to Pifo and Cesar, two local men who had perfected the art of hunting big lizards. When I visited her research site, they invited me to tag along on one of their forays. Pifo’s dog was a brownish mongrel, medium-sized, with short hair, a curly tail, and perky ears—the classic profile of dogs that go feral and are allowed to breed freely. Ray Coppinger calls these “natural breeds.” They usually resemble Australian dingos, and you find them running wild from Tanzania and Israel to the coast of South Carolina. Pifo’s dog was a companion as well as a hunting partner. He talked to his dog, petted him, and fed him table scraps.

  An iguana hunt begins when you spot a lizard on a limb basking in the sun. The acrobatic Cesar shimmies up the tree trunk and gingerly makes his way out on the branch. The dog, tense, stares up at the lizard. Cesar violently shakes the branch until the iguana bails and plunges straight down sixty feet into the underbrush. In a flash, the dog is after it. There is no way a human could chase the lizard down, but thanks to Pifo’s dog, no problema.

  Cesar and Bev take off after the dog. Clumsily, I do my best to keep up. The dog quickly corners the lizard. The trick for Bev is to subdue the iguana before the dog does any damage to it or it does any damage to her. Male iguanas can be six feet long. They will whip you with their tails and bloody your arms with their claws. You do not want to get bitten by one. Bev usually wins these woman-versus-lizard wrestling matches. Once she has the iguana under control, she weighs it, checks its reproductive status, marks it, and lets it go.

  It is an efficient way to hunt. With the aid of the dog, they can capture ten or twenty lizards a day. When Pifo and Cesar were not catching iguanas for Bev’s research project, they used their dog to hunt lizards for meat. Once they catch an iguana, they immobilize it by yanking a tendon from the iguana’s forefeet and use it to tie its legs together behind its back. Later, they will kill the lizard by inserting a long spiky thorn straight through the animal’s brain and into the spinal cord.

  Coppinger argues that pure-blooded wolves are too wild for this sort of cooperation between man and beast. After spending time with Nancy Brown’s wolf-dogs, I am inclined to agree. But the most convincing support for Coppinger’s theory of dog domestication comes from an extraordinary genetic experiment with foxes.

  FROM WILD TO TAME: HOW TO MAKE FOXES FRIENDLY

  In the mid 1950s, a geneticist named Dmitri Belyaev found himself on the losing end of a scientific dispute and was fired from his position at Moscow’s Central Research Laboratory of Fur Breeding. Belyaev wound up as head of an animal research facility in Siberia where, in 1959, he began a remarkable experiment on silver foxes that went on for forty years and involved over 45,000 animals. It changed the way scientists think about domestication.

  By selectively breeding foxes for tameness around humans, Belyaev sought to create a strain of captive foxes that would be easier to work with. At the fur farm, he tested fox kits for their reactions to people and then cross-bred the tamest males and females. Initially, only a few foxes me
t his strict friendliness-to-humans criteria. By the tenth generation, 20% of foxes were tame, and after forty generations, 80% of fox kits were tame. The selectively bred foxes would lick your face; the unselected line of animals would rip it off.

  So foxes selected for tameness became nice. Big deal. Here is the important part: An unintended by-product of selection for tameness was that the foxes began to look and act like dogs. Over the generations, their ears became floppy and their tails curly. Their coats began to show traces of brown (the probable color of early dogs), and some of them developed the white patches you sometimes see on the faces of dogs. The foxes’ faces became shorter, wider, cuter. The animals showed decidedly un-foxlike behaviors. They wagged their tails when people greeted them. The foxes’ physiology changed, too. Compared to normal foxes, their levels of the stress hormone cortisol were lower and their neurons produced more of the brain’s natural antidepressants. No wonder they were mellow.

  Why should traits as different as coat color, the curl of the tail, and the shape of the head be dragged along with having a nice personality? We don’t know, but behavior and color are genetically linked in species ranging from garter snakes to rats. Whether the changes that occurred in Belyaev’s tame foxes were produced by a small number of linked genes, or even a single gene, is unclear.

  CAN DOGS READ OUR MINDS?

  When I hear Doc Watson, the great flat picker from Deep Gap, North Carolina, lay out just the right notes in a fiddle tune guitar break, a tingle runs from my left shoulder and up the side of my neck and I get goose bumps. The Hawaiians call this “chicken-skin music.” As a scientist, I read a lot of articles in academic journals. Most of them put me to sleep, but every now and then, one gives me the chicken skin. It happened in 2002 when one of my colleagues handed me an unpublished manuscript that his old college roommate, Mike Tomasello, a developmental psychologist, sent him. The paper, which soon appeared in the journal Science, compared the ability of wolves, chimps, and dogs to understand human gestures. The lead researcher was a twenty-six-year-old graduate student named Brian Hare.

  To understand the importance of the study, you have to know something about chimpanzees. They are our closest, smartest, and most socially savvy relative. Like other nonhuman primates, however, chimps are lousy when it comes to following human signals to find food. Even chimpanzees that are hand-raised by people are not able to use gazes and finger points to locate food. Given that big-brained chimps are poor at this task, you would predict that wolves, which have much smaller brains than apes, would also fail the point-gaze test. You would be right. You would also, of course, think that domestic dogs would be dumber than wolves as their brains are roughly 25% smaller. Wrong.

  Hare, who now heads the Canine Cognition Center at Duke University, used a simple choice experiment to show that dogs are naturals at understanding human signals. He hid food under one of two bowls. A research assistant would tap, point, and gaze at the bowl with the hidden food. As expected, the chimps and wolves were hopeless. They randomly guessed the left or right bowl and were right about 50% of the time. The dogs did much better. When given all three cues, the dogs made the right choice 85% of the time.

  Is the superior performance of dogs the result of nature or nurture? He believes that evolution has endowed dogs with a special ability to read human signals. Recently, however, the origins of the differences between dogs and wolves have become a matter of hot debate. Clive Wynne, a dog researcher at the University of Florida, thinks Hare is wrong. He attributes the differences to socialization. The truth is probably somewhere in between. A research group in Hungary headed by Adam Miklósi has found that you can train wolf pups to attend to human signals, but it takes the wolves ten times longer to achieve a similar level of success and requires much more socialization.

  An increasing number of scientists are focusing their research on the canine mind, and dog behavior research centers are popping up everywhere. The remarkable ability of pet dogs to understand human intentions and their eagerness to participate in experiments makes them excellent subjects for studies of the evolution and development of communication and social behavior. As Yale University’s Paul Bloom says, in cognitive ethology, dogs have become the new chimpanzee.

  Unexpected examples of canine brain power appear every month in journals like Science, Animal Behavior, The Journal of Comparative Psychology, and Animal Cognition. A border collie named Rico tested by researchers in Germany knew the meaning of over 300 words and could learn the name of a new object in a single trial. Scientists in Brazil taught a dog to use a keyboard so she could indicate whether she wanted to play or eat, be petted, or go for a walk. Researchers at the University of Vienna found that dogs can tell when their owners are watching them or watching television—and they disobey more when their owners are glued to the TV set. Further, dogs are copycats. For instance, when their owners yawn, they tend to yawn. And when they try to solve a problem, dogs imitate the behavior of humans whose actions succeed rather than that of those whose actions fail.

  WHAT WOULD LASSIE DO?

  Studies like these show that dogs have some amazing abilities. But do we sometimes expect too much from an animal whose brain weighs a couple of ounces? For thousands of years, humans have used the intelligence of dogs to good effect—watch a border collie bring in a flock of sheep or a gun dog on point. In addition to hunting and herding, we now count on dogs to sniff out bombs, dope, and bladder cancer. They locate lost kids and rotting cadavers, warn deaf owners when the smoke alarm goes off, and lead the blind through city streets, even “intelligently disobeying” commands that would jeopardize their owners. A lot of people go further and believe that their dogs possess quasi-mystical abilities. Half of dog owners believe that their pets have a form of ESP that allows them to know in advance when they are about to return home after being away.

  Humans love the idea that a good dog knows when you are in a jam and will help you get out of it. When I was a kid, Lassie was my idea of what a pet should be. Bosco, our beagle, was fun to play with, but Lassie could save you from a cougar, fetch your dad if you fell down a well, and bark when the barn caught fire. (Many people still feel this way. Just Google “dog saves owner.”)

  Bill Roberts, a psychologist at the University of Western Ontario, was impressed by news accounts of dogs who rescue people. Bill knew that social psychologists investigate the circumstances in which human bystanders come to the aid of strangers by staging fake accidents. He thought that similar experiments might shed light on a dog’s ability to intentionally seek help when their owners are in dire circumstances. Fortuitously, one of his students, Krista Macpherson, was a dog breeder and trainer. She had access to lots of dogs and owners, and she and Bill came up with a way to test what I call the Lassie Get Help hypothesis.

  They asked dog owners to walk their pets through a field in which a bystander sat quietly in a chair reading a book. (The bystander was a part of the research team.) When the dog-owner pairs were thirty feet from the bystander, the owner would clutch his chest, collapse, and lie motionless on the ground. When faced with this situation, Lassie, of course, would sense that her owner was in dire straits, bark once or twice, and run over to the bystander and tug at his sleeve. In this situation, did the real dogs—dogs like my beagle Bosco—step up to the plate?

  No. Not one of the animals in the experiment made the slightest effort to solicit help for their injured owner. Nor did they try to get help when Bill repeated the experiment in a laboratory situation in which a fake bookshelf was rigged to fall and pin their owners to the floor. Even the collies in Bill’s study failed the Lassie Get Help test.

  Some breeds, however, are more adept than others at understanding humans and responding to human commands. For example, Adam Miklósi’s group in Budapest has found that breeds that have been selectively bred to work with humans, such as sheepdogs and retrievers, are better at using human hand points to locate food than breeds like hounds that were bred to work independently
. Not too surprising. What is surprising is that breeds with short snouts and wide faces like bulldogs, boxers, and pugs understand human signals better than long-nosed breeds like Dobermans, dachshunds, and greyhounds. The researchers suspect that this odd finding boils down to a difference in the placement of the eyes and the shape of the head. The eyes of the narrow-faced breeds are located more on the sides of their heads. While this trait gives them better peripheral vision, it also makes them more distractible.

  Researchers at the Center for the Interaction of Animals and Society at the University of Pennsylvania have examined breed differences in dog behaviors in a different way. They developed a Web-based owner survey to examine traits like the willingness to obey commands. Their Canine Behavioral Assessment and Research Questionnaire (C-BARQ for short) has been taken by thousands of dog owners and has provided a goldmine of data on differences in the behaviors of dog breeds. (If you have a dog and would like to participate in C-BARQ research, the survey Web site is w3.vet.upenn.edu/cbarq.)

  The researchers were particularly interested in a dog’s willingness to do what its owner wanted it to do—come when called, fetch, learn tricks, and pay attention. The researchers analyzed C-BARQ trainability scores for 1,500 dogs representing eleven breeds. They found that some types of dogs are easy to train while others are not politically smart. Labrador retrievers were the most trainable breed and basset hounds the least. Seventy percent of Labs were highly trainable, compared to only about 5% of basset hounds. And while none of the Labs were in the dense categories, half the basset hounds were.