If a giraffe is an animal designed by committee, a bulldog must have been put together by a vet facing large house payments. Newborn bulldogs typically have heads so out of proportion to a mother's birth canal that they have to be delivered by cesarean section. By the time they become adults, their skulls are as big around as the dogs' height at the shoulder. Bulldogs also frequently have bad hearts. "I have a friend who's been trying to breed bulldogs for years," says Janice Koler-Matznick, founder of the Primitive and Aboriginal Dog Society. "She'll get a litter of three puppies, all of which die, or they'll all have bad hearts, and she'll have to put them down. It's just heartbreaking."
Koler-Matznick isn't too keen on German shepherds, either, at least not the ones that win at dog shows. ("They have this wonderfully long rear end, which is also crippled," she says.) Or bloodhounds. ("They've been selected to be a giant nose on legs, but unfortunately those legs don't work too well.") Koler-Matznick isn't naturally cranky, and she doesn't hate purebreds. She has spent decades as a dog trainer, a professional handler at dog shows, and a dog breeder, among other things. But she likes her dogs healthy and hardy and has come to believe that much of dog breeding is bad for the species, engendering not just design flaws but also a terrible propensity toward genetic disorders. She's not alone. If you think of creating purebreds as a kind of genetic engineering—employing intensive artificial selection to increase the frequency of desired genes—then Koler-Matznick is part of a reverse engineering movement. Along with a few disillusioned breeders, veterinary researchers and even geneticists are looking for ways to undo the problems humans have created.
Many of these problems are entirely predictable, says conservation biologist Dan Wharton. As director of the Central Park Zoo in New York City and manager of the American Zoo and Aquarium Association's breeding programs for gorillas and snow leopards in North America, Wharton is professionally committed to the science of genetics. But he also has two Pembroke Welsh corgis at home, and a few years ago it struck him that his two interests were running head-on into each other. "Some of the concepts of dog breeding are left over from our understanding of genetics a hundred years ago," he says. "What breeders are doing and what they expect doesn't necessarily match up."
Take linebreeding, which involves mating closely related dogs with a desired trait—perfectly droopy ears, for example. Linebreeding works. It will, as breeders put it, "set" a trait, ensuring that at least some pups in the litter will carry droopy-ear genes, and only droopy-ear genes, to pass on to future generations. "But from a biological point of view, linebreeding is also what we call inbreeding," Wharton says. "And inbreeding has a downside."
After all, you shouldn't marry your sister, and not just so you'll be interested when she complains about her mother. Humans and dogs alike have two copies of each chromosome, giving them two full sets of genes. If Mom gives her kid a cracked version of some gene, the copy from Dad can usually do the job. But if Mom and Dad are brother and sister, they may well be carrying the same malfunctioning gene, and some of their children are likely to get a double dose. Call it linebreeding or call it inbreeding, it's a great way to increase the chance of genetic disease. And so basenjis are prone to a potentially deadly type of anemia, English mastiffs to blindness, and Bedlington terriers to a metabolic problem that can cause them to hoard enough copper from food to fatally damage their livers.
Admittedly, some dogs swim in smaller gene pools than others. Every topknotted, squash-faced shih tzu can be traced to a group of just 14 dogs brought out of China before the Cultural Revolution. But the gene pools of nearly all show breeds are closed: The bylaws of the American Kennel Club prevent breeders from registering a dog unless its parents were registered, which means no new genes can be introduced to freshen things up. Add to that the effects of the "popular sire" syndrome, when breeders use a showstopper of a dog to produce way more than his fair share of puppies, and not even the supernumerous breeds like Labrador and golden retrievers are immune to the dangers of genetic bottlenecking.
One remedy, Wharton says, would be to convince dog fanciers to change the way they think about breed purity—to focus their concern on specific genes rather than on the dog. After all, the vast majority of genes in both a Chihuahua and a Great Dane code for basic dog; the difference comes down to slight variations in a few genes that code for each breed's signature traits. "One can inbreed those genetic traits without inbreeding the whole dog," Wharton says.
Say a genetic disease was becoming too common in Australian shepherds. A breeder might mate one of his prize sires to a Border collie, which is also a herding dog and vaguely similar in appearance. When the pups got old enough, they could be backcrossed to the target breed—that is, mated to an Australian shepherd. The resulting litters will be three-quarters Aussie, genetically speaking, but in appearance they will be a mix-and-match assortment. Most of the pups will look like classic hybrids, but some will have the appearance and behavior of a purebred Australian shepherd, albeit with more genetic diversity and "hybrid vigor." Those are the ones a breeder can use to refresh the Aussie gene pool.
One of gene therapy's biggest successes has been in dogs with congenital blindness. In 2001 researchers inserted a working version of the flawed gene into a harmless virus, then injected the virus behind the retina of a blind puppy named Lancelot. Three months later, Lancelot could see.
"Breed integrity is not being challenged when you do genetic management," Wharton says. "It's just a matter of understanding that very few genes make up the integrity of your breed. It's not like pouring two liquids into a bottle, where they blend and you can't separate them. The genes are specific entities. They sort themselves at random, but you are in a position to pick and choose." Such a system would have costs. A breeder might have to "select back" a few generations before getting a dog that makes the grade. For a careful hobbyist who doesn't want to push his dogs to breed too quickly, that could mean years without a gold medal at dog shows.
Another strategy would scientifically reproduce the happy accident familiar to all breeders: the litter of champions born of so-so parents. Breeders would start with two lineages, each of which carries a few of the breed's prized traits. Crossing the lines could produce a hardy, genetically diverse dog of ideal size, proportions, coloring, and behavior. Unfortunately, that dog can't be used to produce the next generation or the two lines will quickly merge.
Chicken farmers use a similar method because they're picky about things like meaty breasts and good egg production but can't afford the health problems that come with inbreeding. Birds in the male line might be selected for their size and fast growth while the female line is selected for fertility and egg-laying capacity. A breeder can't simultaneously select for rapid growth and good egg production, as it happens, because fast-growing chickens tend to be poor layers. By crossing the lines, though, a chicken farmer can have both.
The drive for profit has turned chicken farming into a science. Dog breeding remains mostly a passion, which leaves Koler-Matznick pessimistic about reform. Wharton is more sanguine. "So far, I haven't heard of breeders being interested in reinventing the paradigm," he says. "But I think the time will come."
The Web site of the American Kennel Club has a listing of the 150 AKC-recognized breeds, along with photos: www.akc.org.