Snakes Show Surprising Capacity for Learning


February 5, 1999

New findings suggest that when it comes to learning and cognition, the humble snake may be quite a bit more like humans than anyone had imagined. David Holtzman, a neuroscientist at the University of Rochester, has found that snakes have a much greater capacity for learning than earlier studies had indicated. His research also indicates that, like humans, many snakes rely on sight to get around, and that older and younger snakes differ in how they gather and decipher information about the world around them.

The findings appear in the January issue of Animal Behaviour.

Holtzman's study challenged 24 captive-bred corn snakes (species Elaphe guttata guttata) to escape from a black plastic tub the size of a child's wading pool. Cards mounted on the arena's walls and tape on its floor provided the snakes with visual and tactile cues to find their goal: holes in the tub's bottom that offer a dark, cozy spot to hide.

"These snakes appear to have a very strong aversion to the bright lights and open spaces found in the arena. When a snake is first placed in the arena, it tends to circle around the edge, looking for a way out," says Holtzman, an assistant professor of brain and cognitive science. His team found that given a nudge in the right direction, snakes are readily taught to find the exits -- and then recall how to use cues to find them in successive trials.

Simply stumbling into a hole isn't the only proof that the snakes are learning something, though. "Speed to find that goal is one of the measures which shows they're learning," Holtzman says. "On average, they take over 700 seconds to find the correct hole on the first day of training, and then go down to about 400 seconds by the fourth day of training. Some are actually very fast and find it in less than 30 seconds."

Studies dating back to the 1950s interpreted snakes' clumsiness with mazes as a poor reflection on their intelligence. Holtzman's peers regard his work as groundbreaking because unlike a maze, his arena confronts snakes with a situation that they're likely to encounter in the natural world.

"Early attempts to study snake navigation were awry because the studies used mazes as testing arenas -- as though snakes might be expected to run through mazes in the same way rats run through mazes," Peter Kareiva, a professor of zoology at the University of Washington, wrote last summer in Integrative Biology, of which he is editor-in-chief. "Of course, snakes do not encounter anything resembling mazes in nature, and they do not learn how to run mazes in laboratory conditions.

"The bottom line is that when tested in a biologically meaningful way, snakes exhibit spatial learning that rivals the learning abilities of birds and rodents," he concluded, "but the cues used by snakes [need] to match their ecology."

Holtzman found a few age-based differences in the cues snakes use to extricate themselves from the arena. Young snakes - - those up to three years old -- appear to be more adaptable and resourceful, using a variety of clues to find their way to the exit. But their elders seem to rely much more heavily on visual cues, becoming a bit befuddled if the brightly colored card marking the exit hole is tampered with.

"Actually, one of the interesting findings from our studies is that snakes use vision at all in locating places," says Holtzman. "They don't just rely on the chemical cues picked up by flicking their tongues out, as many snake biologists assume."

The experiments within the arena were surveyed by video cameras that can detect tiny foil hats fitted to the bright orange and red snakes, which can grow to lengths of four feet. The snakes can't be observed directly during experimentation, because the presence of a person might provide them a cue, disrupting the experiment. Researchers lurk just out of sight behind black curtains that wall off the arena, watching the snakes on closed-circuit television and using a computer to analyze and catalog their movements.

Holtzman hopes his work may someday have major implications for people, in the form of therapies to grow new neurons to compensate for brain damage.

"One of the most interesting discoveries in neuroscience and cognitive science is that new nerve cells, neurons, can be formed in some brain regions in adult higher-order vertebrates, including primates," says Timothy Nyberg, a Rochester undergraduate who joined Holtzman in the research. Neuroscientists know, for instance, that adult humans can produce limited numbers of new neurons related to the sense of taste and smell, as well as in the hippocampus, a brain region involved in memory and spatial learning. New neurons grow in other animals as well: Learning to store food engenders brain growth in birds, even doubling the number of hippocampal cells, and in lizards, if the hippocampus is removed, it can grow back -- and skills lost return as if by magic.

It's Holtzman's theory that what holds for snakes, lizards, and birds may also hold for their evolutionary descendants -- humans. If he and his colleagues come to understand how to control the mechanisms that govern neurogenesis in other animals, it could offer new therapies for the treatment of people afflicted by brain damage -- whether from accidents, strokes, or diseases like Alzheimer's or Parkinson's. The work could also help to better pin down the as-yet fuzzy notion that babies who grow up in more stimulating environments develop more robust brains.

Holtzman and Nyberg were joined in the research by Anita Stone, a former research assistant; Terrence Harris, now a physician at Washington University; Guillermo Aranguren, now a graduate student at the University of Texas at Tyler; and Elizabeth Bostock, a physician at the University of Rochester. The research was sponsored by the National Institutes of Health's National Center for Research Resources.

CONTACT: Tom Rickey, (716) 275-7954.