Graduate student Eatai Roth pokes around the fish tanks in Ames Hall, trying to coax a knifefish out of its tube-even resting, they hang out in PVC pipes-and explains the naming scheme for the latest experimental batch: Hope, Maverick, Joe, and Rogue, for Obama, McCain, Biden, and Palin, respectively. A guy who likes robots, Roth is excited about a discovery made while studying mechanics in the fish: They show behavior suggestive of complex learning, something expected only in higher-order vertebrates.
Since knifefish share a lot of the same neural infrastructure with other species, their motion can offer clues as to how different complex animals control movement-humans included. “If I presented this at a conference for human behavior,” Roth says of the discovery, “it would get polite nods. But the really impressive thing is that we found it in fish!”
In examining the mechanics of the knifefish’s motion, Roth collaborated with his advisor, Noah Cowan, associate professor in the Department of Mechanical Engineering, and neural ethologist/ biologist Eric Fortune, associate professor in the Department of Psychological and Brain Sciences at the School of Arts and Sciences.
Their experiment took place in a tank with a high-speed camera looking up from beneath. In the water, the scientists moved a small PVC tube longitudinally over the body of a knifefish. (The pipe was cut off along the bottom and at both ends.) The fish swam forward and backward to stay hidden beneath the tube. Knifefish can swim both ways equally well because they move by sending undulations up and down a long ribbon fin on their underside, making the experiment and the math simple-no need to create counter-currents or account for motion in more than one linear dimension.
Knifefish, which live in the murky rivers of the Amazon, probably hide in reeds, rocks, and changing currents in order to avoid prey, which is why they innately stay inside the tube during the experiment, Roth explains. They are also electro-sensory, able to “see” their surroundings by generating an electric field and registering conductivity changes around themselves in electricity-sensing pores all over their skin. This sense, along with normal vision, helps control their locomotion.
“The funny thing about this study was that it was intended to be The Most Boring Study,” says Fortune, “and it turned out not to be.” Rather than just showing how the fish used their sensory systems to position themselves, the way they moved suggested that something, well, fishier was going on. So good at staying in the tubes were the fish that they must have been able to predict where they would be less than a second in advance.
Rather than simple behavior, the fish were “doing something clever and complicated,” says Roth-presumably complex learning. The knifefish appear not only to learn about their surroundings by sensing where the tube is and moving beneath it but to adapt to the tube’s trajectory as well, moving in response to the object’s way of moving. A correlation in people would be something akin to visually tracking the motion of different birds; after a while, a human will innately realize that hawks have very different trajectories from doves, and to keep following them, eye motion will auto-adapt to their flight paths.
Says Cowan of the experiment, “It makes it a great system for examining the how, as opposed to the why.” Even after observing the fish in the wild, Fortune can only speculate about their motivations for hiding in PVC pipes-a problem for behavioral biologists to take on later, perhaps.
The essence of a reverse-engineering study, the “how” will be key for future engineers trying to build environmentally adaptive robots or for neurologists interested in relating the brain to the mechanics of motion.