Closely observing how male Greater Sage-Grouse court females is difficult—unless you have a lady on wheels. Photo: Noppadol Paothong

Tech

Is That a Robo-Bird or Real Lady? For Male Sage-Grouse, Either Will Do

Gail Patricelli is one of a few researchers using robotic birds to study avian courtship. Her current work focuses on the Greater Sage-Grouse.

Showing off in a crowd to attract a mate, otherwise known as lekking, is hard work for male Greater Sage-Grouse, even under the best circumstances. They have to show up early, strut around a clearing in the sagebrush, pump and pop the booming air sacs in their chests, and fight off rivals—all in hopes that a female will deign to take interest. It’s enough to really wear a guy out. And as if that wasn’t bad enough, on a handful of leks in Wyoming, every once in awhile the female a lucky male is energetically courting might suddenly topple over, revealing a set of rubber wheels.

Behind the controls of that odd female sage-grouse is Gail Patricelli, an ecologist at the University of California, Davis. Patricelli's main focus of study is understanding how birds communicate, especially during courtship rituals. As she discovered early in her career, though, studying these behaviors in the wild in any sort of controlled way is hard. So two decades ago Patricelli turned to remote-controlled robots wrapped in taxidermied skins to impersonate female birds. For the past decade, her current research has focused on the Greater Sage-Grouse, and she calls the latest versions of her robotic birds “fembots.”

Patricelli compares her experiments to playback studies, which she calls “the oldest tool in the playbook for behavioralists.” In a standard playback study, a researcher records a sound made by an animal, then plays that sound back to another individual while observing how it responds. Researchers have added a visual component before—filming an animal or using a puppet, for example—but those techniques don’t necessarily offer the same experimental control.

A leader in her field, Patricelli is one of only a handful of researchers who have managed to successfully design robotic bird imposters, and she says she’s been surprised by how difficult interaction studies have been. Lizards and frogs are much simpler to fake with rubber molds, but for animals like birds, she says, “it’s really hard to build something that they’re going to fall for and that’s going to elicit natural behaviors.”

Patricelli first experimented with building robotic birds for her PhD research on Australian Satin Bowerbirds, which are known for the elaborate nests, called bowers, the males build and use as the base for their equally elaborate courtship displays. Previous scientists had noticed that although females are more likely to mate with males who have more intense displays, males don’t always give a display their all.

After a decade of using her 'fembots' in the field, Patricelli knows this much: Males don't discriminate. Photo: Noppadol Paothong

It turns out there's a good reason for that; in Satin Bowerbirds, a courtship display looks surprisingly like a threat, with the male hopping around, flinging his wings out dramatically, and making his feathers buzz loudly. Sometimes, instead of impressing an eligible bachelorette, an enthusiastic male startles her—and then has to start all over again with the next female to stop by. Patricelli thought males were reading behavioral cues given by the females to gauge how intensively they could display, and she decided to use robots to confirm her suspicions.

At the time, in the mid 1990s, there was only one example of the sort of experiment she wanted to design. A team of Danish and German researchers had created a robot that could imitate the vibrations and scents of honeybees’ trademark waggle dance. They succeeded in using the robotic bee to communicate with a hive, and the results of the experiment cemented the importance of the waggle dance in bee communication.

Patricelli had spent two summers before graduate school coordinating an annual robotics conference, where she became comfortable with the technology and made connections she could later call on to help make her experimental robotic birds a reality. “I knew other people had tried it in the past and just weren’t able to get it to work,” she says. “I had had that experience that made it seem entirely doable.” For bowerbirds, unlike bees, looks are important, so Patricelli knew she wanted to cloak the robots inside taxidermied female skins. Thanks to a background in art, she was confident she could combine all the various parts necessary to make a realistic bird. “I came into it with a love of building things,” she says.

With innards built by a robotics engineer at her university, Patricelli constructed three female Satin Bowerbirds, which are about the size of an American Crow. The imposters could turn their heads, raise themselves upright to mimic a startled female, and crouch down to signal interest in the male’s display. For each trial, Patricelli snuck the female onto the male’s carefully constructed lovenest and then waited for him to return and begin courting. Sometimes the faux female would be startled, sometimes she would stay engaged the whole time, and sometimes she would remain completely aloof. Not only was the experiment a success, but Patricelli’s results suggested that she had been right: Male Satin Bowerbirds can indeed tailor their courtship based on female responses. 

Whether robot interactions are a helpful experimental technique depends mostly on the animal’s own characteristics—mainly how large they are and how thoughtfully they interact with each other. Because of this, Patricelli says she has only tried the technique to fool male birds of lekking species. “That’s about as easy as you can get in terms of the choosiness of your target animal,” she says of the males pining for any attention they can get. 

David White, a psychologist at Wilfrid Laurier University in Canada, knows how hard robot work can be when the species are smaller and the mating ritual is more nuanced. He wanted to study how female Brown-headed Cowbirds influence males during courtship with small wing movements. First, he had to fit all the robot parts into a female cowbird’s body, which is only about seven inches tall. But even once he had a robotic cowbird up and running, he could tell something wasn’t quite right. The males either completely ignored her or tried to mate with her right away, rather than following the drawn-out courtship that lets birds gradually get to know each other. “There are so many social cues that these guys pay attention to that we could just never get to the point of simulating all of them,” White says.

Beneath the taxidermied hide of each remotely controlled bird is a sophisticated robot and four wheels. Photo: Gail L. Patricelli

He's moved onto other techniques, but he says the experiment still helped his research. Because he had to mimic the female birds, he also studied them more closely than ever before. “I learned things about their behavior that I’d never really paid any attention to,” he says—details like the fact that female birds move their heads almost constantly. 

Patricelli has avoided White’s challenges, since Greater Sage-Grouse males don’t really seem to have any dealbreakers. “Anything brown and round is fair game,” says Patricelli, who says she’s even seem them try to mate with dried cow poop when nothing more promising was around. “The bar is pretty low for us in trying to fool the males.”

Jackie Augustine, a behavioral ecologist at Ohio State University, Lima, also benefits from these lower standards. The newest researcher on the birdbot block, Augustine wanted to understand how Greater and Lesser Prairie-Chickens, which are similar to Greater Sage-Grouse in size and mating rituals, interact in places where the two species share leks. So she’s made two robotic versions of each species’ female. For Augustine, being able to run experiments without flushing birds unnecessarily and having realistic models to fool males are the key perks of working with robots.

Nevertheless, learning how to best use the robotic birds has taken a few field seasons. This spring, she finally got data she’s satisfied with, and one more year should provide her enough to start analyzing. Much of the learning curve until now has been about figuring out how to drive the critter like a Prairie-Chicken instead of like a robot. Augustine’s first instinct was to just drive the hen out there, but the males would fly away, “because that’s not what real females do,” she says. So instead, she’s learned to drive a few feet, then stop and look around for a bit before moving on, “kind of mimicking what real females actually do when they visit the lek.”

Through the years, Patricelli also had to develop a signature avian driving style. She has retired three older robots from fieldwork, including two that ran on set train tracks rather than wheels. She’s pretty satisfied with the two robots of the current model, except for their occasional propensity to topple over. (The bottom of each robot is circled by lead weights to counteract the top-heavy machinery.) Before even running an experiment, Patricelli has to scout out the lek—deciding on a path for the fembot to follow, trimming some unmanageable clumps of grass, flagging obstacles to avoid, and the like. Actually driving the robot takes two people, with Patricelli tucked in a blind and a PhD student, Ryane Logsdon, perched on a hill with binoculars and a radio to help navigate the lek.

Patricelli and Logsdon aren’t just contending with difficult terrain—they’re also trying to keep the fembot a safe distance away from males. Ideally, the robot is about 30 feet away from males at all times. That’s close enough for the male to respond to the female, but far enough away that the pair can evacuate the robot if needed. That’s the one downside of males with low standards: If they get close enough, they’ll likely try to actually mate with the fembot, which can cause wear and tear on the robot’s taxidermy disguise.

It's one thing to create a bird that can fool males, but driving them around the lek is a challenge that requires two people. Photo: Gail L. Patricelli

There's also lots of waiting around involved. To avoid interfering with any courtship that could be successful, the team only sends the robots out when there aren’t any female grouse on the lek. “Most fieldwork is a lot of trial and error, and that’s definitely true with the robots,” Logsdon says. “It’s waiting for the stars to align perfectly.”

If she can find the resources, Patricelli wants to build more of the fembots to better explore group dynamics on the lek. “What we would love to do is come up with a gaggle of robots,” she says. Often multiple females are on the lek evaluating males at the same time, and choices made by older females can influence which males younger females are interested in. With enough fembots, Patricelli could mimic those dynamics, artificially altering a male’s status and seeing how he responds.

Even with just the pair of robots Patricelli and Logsdon have now, there are plenty of questions left to tackle about sage-grouse behavior, including how information spreads on the lek and how a male decides between a female who isn’t responding and a new arrival. “The robots just give us an incredibly amazing opportunity to actually control half of the conversation,” Logsdon says. “That just opens up so many opportunities for us and makes it such a cool system.”

Although robots have proved helpful for Patricelli’s research, she says there will always be limits on how they can be used. Some species and some members of species will always be too difficult to fool. Even the female sage-grouse she mimics will never find themselves facing off a robot male. Reproducing all the male physical traits that females scrutinize so closely would be “an impossible task,” Patricelli says. “I think even if I had a Hollywood budget I couldn’t pull that off.”

***

Take Action! You can help protect the Greater Sage-Grouse. Tell Secretary Zinke you support sage-grouse conservation plans that work.    

“The views expressed in user comments do not reflect the views of Audubon. Audubon does not participate in political campaigns, nor do we support or oppose candidates.”