Caterpillar call-of-the-wild aims at ants.

Caterpillar call-of-the-wild aims at ants

Butterfly caterpillars don't have it easy. Some become victims of parasitic wasps and flies, which inject their eggs into the juicy larvae. These caterpillars suffer slow deaths as wasp hatchlings consume them from the inside out. Others get swept from their leafy perches by low-flying adult wasps in search of fast food.

But many caterpillars in the families Riodinidae and Lycaenidae -- which metamorphose into the beautiful, widely distributed butterflies commonly called "metalmarks" and "blues" -- don't have these problems. They line in relative comfort by surrounding themselves with bodyguard ants. They appease the ants by secreting sugary fluids; the ants, in turn, keep predators at bay.

But how do they attract the ants in the first place? It took a dedicated entomologist with a tiny microphone to find out.

Philip J. DeVries of the University of Texas in Austin used to a custom-designed recording system to tape the sounds made by riodinid caterpillars of the species Thisbe irenea. When he held the microphone against the suface upon which a caterpillar was walking, he detected a simple, repeated vibratory pulse. These were not footsteps, however. The sounds were made by the rapid tapping of tiny body structures, called vibratory papillae, against the surface.

Until now, the purpose of vibratory papillae has remained a mystery, but DeVries finds that the rhythmic drumming of these tiny appendages attracts ants. Indeed, the roughly 16-pulse-per-second, 896-hertz signal closely resembles some vibratory messages that other entomologists have identified as a component of ant communication. DeVries' study points to the intriguing possibility that some insects have evolved calls designed to summon unrelated species into symbiotic relationships.

In laboratory experiments and field studies in Panama, DeVries showed that caterpillars whose papillae he had surgically removed failed to produce calls and were tended by significantly fewer antts. Listening to 13 other riodinid species bearing vibratory papillae, he recorded calls similar to those of T. irenea. "Prior to this report, sound production was unknown from butterfly caterpillars, nor was it known from any symbiotic association with ants," DeVries writes in the June 1 SCIENCE.

His recordings of 44 other neotropical caterpillars, none of which associate with ants, revealed no vibratory signals. Interestingly, however, the entomologist did detect vibratory signals from 23 other ant-associated caterpillar species bearing no vibratory papillae. How these species produce calls remains unknown, he says.

"It's a pretty neat paper," says Robert Robbins, an entomologist specializing in caterpillar-ant relations at the National Museum of Natural History in Washington, D.C. He says entomologists have suspected that vibratory papillae may have a signaling role, and that caterpillars actively recruit protective coteries of ants rather than simply waiting for them to arrive. "This pits a lot of random observations into perspective," Robbins says. "And it may open up all kinds of research into kinds of mutualism based on sound communication."