Compare the petite Chihuahua with the daunting Great Dane, or the lithe greyhound to the poofy Pomeranian. Many scientists agree that the domestic dog displays more morphological variation than any other known species, thanks to selective breeding by humans. But dog breeds differ in more than their outward appearances. A new study suggests that human preferences have dramatically altered the structure and position of the brain in certain dog breeds, potentially modifying their sense of smell and behavior as well.

In a study published online July 26 in PLoS ONE, neuroscientist Michael Valenzuela of the University of New South Wales in Australia investigated an aspect of canine anatomy that has not received much attention from earlier research: the position of the brain within the skull. All dogs, no matter what the breed, belong to the same subspecies (Canis lupus familiaris) of the gray wolf (Canis lupus) from which they were domesticated. Gray wolves have relatively long skulls and one gray wolf skull is more or less the same as another (age and gender differences notwithstanding). In contrast, the skulls of domestic dogs range from one extreme to another: from the long-snouted German shepherd and Russian wolfhound to the stubby-snouted Boston terrier and flat-faced pug, whose wrinkled muzzle resembles a collapsed accordion. Valenzuela and his colleagues wanted to determine whether the artificially selected differences in skull length between dog breeds also reorganized the canine brain.

The researchers used magnetic resonance imaging (MRI) to scan the brains of 11 recently euthanized dogs from a local Australian pound, along with two living English springer spaniels. The euthanized dogs made up a diverse group, including an Akita cross, a mastiff cross, a Maltese, a Staffordshire bull terrier, a shih tzu cross, a greyhound, an Australian cattle dog cross, a Jack Russell terrier and a pit bull cross. Once they acquired the brain images, the researchers analyzed the overall position of the brain in the skull and estimated the relative volumes of the olfactory lobe—a bulb of neural tissue responsible for processing scents that is approximately 40 times larger in dogs than in humans, relative to total brain size. The experimenters also calculated a cephalic index (CI) by dividing skull width by skull length and multiplying by 100. The higher the CI, the shorter the skull length and vice versa.

The dogs with the shortest skulls—such as the pit bull, Akita and shih tzu cross—demonstrated significant brain reorganization. In short-snouted dogs, not only had the brain’s cerebral hemispheres rotated forward by as much as 15 degrees, the olfactory lobes had shifted position from the front of the brain toward the base of the skull. The brains of puglike dogs did not sit inside the skull in the same way as brains of long-snouted dogs, whose skulls and brains more closely match those of the domestic dog’s wolf ancestor. In other words, the findings imply that when selective breeding by humans squashed the snouts of certain dog breeds, it also morphed their brains.

“The whole brain had rotated in more puglike dogs,” Valenzuela says, “and the olfactory lobe had shifted to the bottom of the skull. There was quite a high correlation, a linear relationship, between short-skulledness and forward rotation of brain. And although there were no differences between breeds in the volume of the olfactory lobe, the position of the olfactory lobe changed in short-skulled dogs.”

Because these effects were consistent across a wide range of skull shapes and were independent of body weight and brain size, the researchers determined that brain organization is most strongly linked to skull length as opposed to other anatomical variables. The researchers hypothesized that the olfactory lobe may have scooted into a new location to allow for proper development of the frontal cortex, but they did not test for this directly.

The research team also speculated as to the implications of morphed brains for smell—regulated by the olfactory lobe—and behavior, which is largely orchestrated by the frontal cortex. Selective breeding may have robbed short-snouted dogs of their keen sense of smell when it relocated their olfactory lobes. The researchers note that puglike dogs are rarely if ever recruited to track down scents. Instead, olfactory orienteering commonly falls to the long-snouted hounds.

“This is the next step of our work,” Valenzuela says. “At the moment we don’t have any direct correlations or linkages, but we have circumstantial evidence that puglike dogs don’t seem to have very good olfactory abilities—they’re never used in scent work. Breeding may have a consequence for their olfactory ability and general behavior, but we need to do more research.”