To study turkey instincts, consider robot turkeys

How does a bird or any other animal recognize another member of its species? In many creatures, this is done by imprinting on a parent. But a young Australian brush turkey can’t do that because by the time it hatches, its parents are long gone. The mama birds lay nests underground in mounds of decaying material and then take off. The baby birds, kept warm by the surrounding decomposition, hatch one by one, dig themselves out and start looking for food. Within a few days, young chicks begin to aggregate. But how do they know that these other chicks are like themselves? About a decade ago, scientists built robot turkeys to figure that out.

For all that brush turkeys look similar to our American turkeys, they’re not closely related. And behaviorally they’re very different. Australian brush turkeys are megapodes, a family of birds found in Australasia that are distinct for their mound-building behavior. When these birds hatch, they know more about how to be a bird than any other type of avian. At an age when robins, penguins and wild turkeys (of the American kind) are dependent on mom and/or dad to provide everything from food to shelter to protection from predators, young brush turkeys are figuring out how to do this all on their own, often through trial and error. They’d have to, since there’s no one around to teach them how to be a bird.

But you can’t ask a turkey how it learned to be a turkey. You’d have to be cleverer than that. So Macquarie University animal behavior scientists Ann Göth and Christopher Evans, who died in 2011, decided to build remote-controlled brush turkey chicks out of the skins of chicks that had died of natural causes and tiny motors taken from toy cars. Their study was published in 2004 in the Journal of Experimental Biology.

The researchers collected eggs from the wild then raised them to hatching. When the chicks were two days old, and before they could see another member of their species, they were placed at the end of T-shaped arena. In one short spur at the top was a robot chick performing a pecking motion, bobbing its head up and down. In the other spur was either a static chick or one performing a scanning motion, moving its head horizontally from side to side. Once released, the researchers watched where each chick went and what it did.

In both setups, the live chicks spent most of their time with the pecking robot. “Responding to pecking movement has a likely functional benefit, as this behaviour can indicate a potential food source,” Göth and Evans wrote. “Brush-turkey chicks are omnivorous. They hatch with a general tendency to respond to some common features of food objects … and while trial and error is important … a pecking conspecific indicating food might speed up the transition from trial and error searching to more selective pecking.”

Australian brush turkeys, in other words, are born attracted to that pecking motion. That’s actually not unique; young domestic chickens will also respond to a pecking motion, though one coming from a model of a hen. But brush turkeys are special in that it’s that pecking motion, plus the distinctive coloring of its compatriots, that indicate to a young bird that it has found a friend.

Robots can be a useful way of learning more about animal behavior. After all, it’s the rare animal that acts on cue. But with robots scientists can reproduce behaviors in exactly the same way over and over, repeat experiments, eliminate confounding factors and tease out how animals respond in specific situations. “Robots have increased our understanding of self-organization and the evolution of cooperative behaviour and communication,” scientists noted last year in Biological Reviews.

However, the use of robots has yet to take over in animal behavior studies. The field is still in its infancy in many ways, with results published in computer science and artificial life journals biologists are unlikely to read. And many studies have been designed not with hypotheses to test but simply with ideas to explore; many scientists may not have realized the robots’ potential value, the researchers contend. But with robots becoming ever cheaper to produce, that may change, allowing biologists to figure out even more about how and why animals do the things they do.