How a Snake Uses the Sense of Smell

Say the words “animal self-recognition,” and many scientists will think of chimpanzees, crows and elephants.

For the first time, researchers — employing an innovative twist on the mirror test — have found evidence that garter snakes can distinguish themselves from others, using not sight but scent.

“Reptiles are massively understudied,” said Noam Miller, a comparative psychologist at Wilfrid Laurier University in Ontario, Canada, and an author of the paper, published on Wednesday in the journal Proceedings of the Royal Society B. “There’s a bias out there that they’re these boring, not very cognitive animals, and that’s completely wrong. That’s one of the reasons why we got very interested in studying them and showing the complex cognitive things they can do.”

One traditional sign of animal cognition has generally been the mirror test, Dr. Miller said, or whether an animal can learn to recognize itself in a reflective surface, a trait thought to be a proxy for more sophisticated intelligence. Pioneered by primate researchers in the 1970s, the test typically involves marking an animal with paint somewhere that is visible only in the mirror and waiting to see if it investigates the change.

Similar tests have since been done with a range of species: elephants (passed), pandas (failed), roosters (passed) and even fish like the cleaner wrasse (passed).

But the mirror test is geared toward animals that are primarily visual. Many species — such as snakes — rely primarily on their sense of smell, Dr. Miller said. In 2017, researchers devised an olfactory version of the test for dogs. (They passed.)

Two different species of snakes were tested in the new study. In one corner: North American eastern garter snakes, predators of insects and fish with a surprisingly complex social life. In the other, African ball pythons, a largely solitary, sedentary snake that ambushes rodents.

Snakes, like humans, have oils in their skin that leave a scent trail. The team rubbed makeup removal pads along the undersides of both snakes to collect scent samples, some of which they doctored with olive oil. They placed the pads at either ends of long, narrow boxes and offered the snakes several choices: between their own odor and straight olive oil; their own odor modified with olive oil; and the modified or unmodified odors of other snakes of the same species.

The team measured the snakes’ interest by gauging how long they flicked their tongues to taste the air — longer indicated sustained interest, he said. The ball pythons showed no apparent distinction. But the garter snakes zeroed in on their own tampered smell and ignored variations of the other snakes’ smells.

“Essentially, it seems like if others smell weird, they don’t care,” Dr. Miller said. “If they smell weird, that’s something they need to investigate.”

Recent research has found that eastern garter snakes are remarkably social, gathering in large groups to hibernate in the winter and forming networks — complete with “friends” — during their active season.

As a more gregarious species, they may be more attuned toward a need to distinguishing themselves from others. One possible explanation of how self-recognition works is the ability to recognize the difference between self and “not-self,” Dr. Miller said. “That then ties it to social behaviors.”

It’s hard to say, however, whether ball pythons’ failure to pass the test is down to a lack of ability or a lack of interest, he added. Continuing research in his lab suggests that ball pythons, while more solitary, are socially complex.

But with over 5,000 species of living snakes inhabiting a range of different environments, he said, the family as a whole offers a wide array of opportunities to figure out what ecologies and behaviors might drive animals to actively distinguish themselves. Future tests might focus on tree-dwelling species, or on vipers like rattlesnakes, which recent research suggested prefer to den with kin and get less stressed around other snakes. Granted, the rattlesnake is also “harder to work with in a lab full of undergrads,” Dr. Miller said.

“In a lot of ways, I think their experimental paradigm is more powerful than the mirror tests,” said Rulon Clark, a biologist at San Diego State University who has researched snake social behavior and was not involved in the study. “A highly reflective mirrored surface doesn’t have a lot of ecological analogues. But encountering and understanding the importance of chemical cues left by yourself and your conspecifics is probably a deeply important aspect of the natural history of these animals.”

“Our research links how snakes experience themselves with how they experience the world around them,” said Morgan Skinner, a biologist at Wilfrid Laurier University and an author of the study. “It also demonstrates that when you can do this effectively in an experiment, you can find cognitive capabilities that some might find surprising.”

Little is known about the social structures of snakes and other reptiles, Dr. Miller said. “And if we want to understand the fundamental building blocks of social structure we need to study a wider range of species rather than just rats and pigeons all the time.”