The dog days of August: How humans have shaped the canine sense of smell

One of the most unshakeable bits of conventional wisdom about smell is that the nose of the dog is vastly superior to our own. In my book, I took a few potshots at this shibboleth; I pointed out that (a) molecule for molecule, the sensitivity of the human nose is competitive with that of the dog, and (b) plenty of human nasal talents sound amazing if given equal hoopla. (Amazing human ability: people can identify the geographical origin of the wood in a popsicle stick by the flavor it leaves in the ice cream!)If we are on a par with dogs when it comes to odor detection under lab conditions, why don’t we outperform them on a daily basis? What dogs have going for them is that they devote more of their brain to analyzing and interpreting scent signals than we do. To a far greater degree, their life revolves around odors.

In What the Nose Knows, I proposed that human odorant receptors evolved to track our biological traits of cooking and seasoning food. I cited evidence that “in the last 5,000 to 10,000 years, genes for smell receptors, along with genes related to diet and metabolism, have been evolving faster than those in any other physiological system.” I speculated that this may have had consequences for our long friendship with dogs.

I also suspect that dogs are part of the whole story. Dogs were first domesticated by man somewhere in Siberia about 15,000 years ago, just as humans populations were shifting from a hunter-gatherer existence to sedentary village life. Increasingly preoccupied with the complex man-made aromas of the cooking pot, our ancestors began to rely on hunting dogs to locate the telltale scent of game. Having co-opted the canine nose, our own scent-tracking ability began to fade. Dogs became, in effect, our long-distance noses, while we specialized in the close-in smelling of food in the mouth.

Two recent studies on canine evolution got me thinking again about human-canine co-evolution. The first was a comparative analysis of cranial and brain anatomy of thirteen different dog breeds. It was conducted by a trio of Australian scientists led by Taryn Roberts at the University of Sydney. Her team notes that the grey wolf, the immediate ancestor of the domestic dog, is dolichocephalic; that is to say it is has a long, relatively narrow skull. Some breeds of domestic dog, such as greyhounds and Russian wolfhounds, still display this trait. Many other breeds are bradycephalic (short-skulled); they retain the flat-faced, snub-nosed look of puppies. (Think Maltese or Staffordshire bull terrier.) As the authors say, “canine bradycephaly is purely a human invention,” the result of selective breeding over the past few thousand years.

Using MRI images of skull and brain, the Roberts team calculated a Cephalic Index (skull width / skull length x 100) for each breed. They also charted the anatomy of the olfactory bulbs—the first processing point in the brain for odor information from the nose. As in humans, the dog’s olfactory bulbs are located just below the frontal lobes of the brain. It turns out that as one moves from long-skulled to short-skulled breeds, the brain tilts downward and the olfactory bulbs change their relative position—they move further below and to the rear of the frontal lobes. This remarkable rotation depends entirely on the Cephalic Index; a breed’s overall body size or weight have nothing to do with it.

How does this human-induced change in olfactory neuroanatomy affect each breed’s sense of smell? That’s the big question, one the authors say is “of intense interest for future research.” I hope we’ll be hearing more soon.

The other dog paper that caught my eye is called “Tracking footprints of artificial selection in the dog genome.” It’s a highly mathematical analysis of a large-scale DNA dataset: more than 21,000 gene loci taken from 275 dogs from 10 breeds. The researchers—from four different institutions across the United States—combed through this enormous pile of data using statistical filters that picked up genetic variation consistent with artificial selection, i.e., evidence of selective breeding. They found 155 genetic regions that met the criteria, including all five genes linked to breed-specific traits in previous, smaller scale studies. (One gene, for example, is associated with skin-wrinkling in the Shar-Pei.)

Although they don’t specifically mention olfactory gene regions that might vary between breeds, the authors raise another, rather provocative, possibility. They note that for many gene locations associated with positive selection in the dog, there are analogous locations in humans.

Although this result should be interpreted with caution, as the specific targets of selection are generally not known with certainty in either dogs or humans, it does raise the intriguing possibility that recent selection has influenced common loci in both the human and dog lineages.

Translation: to some extent dogs and humans have coevolved at the genetic level.

Would it be too big a stretch to think this might have to do with our shared food? Every cooked and spiced scrap tossed to the dogs might have helped shape their olfactory ability. And every gazelle felled by an arrow after being scented by hound would just keep the feedback loop going.