If wings came before flight, what were they for?

Flight may be one of evolution’s most iconic innovations, but zoologist Piotr Jablonski is convinced that early wings were first meant to be seen, not to fly.

The idea came to Jablonski after studying bird behavior in the American West. He noticed some birds would fling out their wings or fan out their tail feathers to lure insects into the open. Then the birds would catch and eat the bugs. If early winged dinosaurs were the ancestors of birds, maybe the dinos did the same, he thought.

For years, Jablonski and his colleagues have been putting that idea to the test, showing robotic and computer-simulated dinosaurs to real insects and recording their brain activity. The approach may seem unconventional, but the team is part of a growing group of scientists who want to experimentally reconstruct what remains elusive: how long-extinct animals behaved.

In the experiments, Jablonski, now at the Museum and Institute of Zoology of the Polish Academy of Sciences, Warsaw, and his colleagues focused on small, feathered dinosaurs called pennaraptorans. The animals’ “protowings” were unlikely to have supported flight, says Minyoung Son, a vertebrate paleontologist at the University of Minnesota in Minneapolis.

The surface area of the wings, he says, would have been too small to create the aerodynamic force needed to lift pennaraptorans off the ground, and the ranges of their wing joints would also have limited their movements. Plus, Son says, feathers need to have an asymmetrical shape to be aerodynamic, and based on the fossil record, these dinosaurs “don’t have the aerodynamic feathers yet.”

To test prehistoric hunting tactics, Jablonski, Son and their colleagues first built a robot dinosaur. They modeled it on the turkey-sized Caudipteryx, which is among the best-preserved “winged” pennaraptorans. The team added detachable wings to the robot’s arms to test insect responses to the arms alone or the arms with wings.

When the robot they named Robopteryx was complete, Jinseok Park took it to a paved path cutting through a natural area in Seoul, South Korea, where it could encounter wild grasshoppers (Oedaleus infernalis). The robot would confront the grasshoppers with what scientists call “flush displays,” opening its wings out to its sides or tipping toward the insect and catapulting its tail forward. Over hundreds of displays from two summers, Park, an ornithologist now at Max Planck Institute for Biological Intelligence in Planegg, Germany, compared how often grasshoppers would flee Robopteryx, both with and without its protowings. The flush displays were more effective with the protowings, according to the results of the study published in Scientific Reports in 2024.

But the researchers were still curious about different styles of flush displays with different kinds of protowing movements. Instead of building another robot, the researchers opted to animate a Caudipteryx and show it to insects on a computer screen. This time, the team focused on domesticated locusts and their neural responses to movement that drives escape.

The procedure was invasive. Park hooked one electrode to each locust’s nerve cord while pinning another to its abdomen. The electrodes were then connected to a specialized instrument that detected and recorded neural activity. The visual displays were more effective at eliciting a neural response from the insects when the animated Caudipteryx had protowings, instead of just bare forelimbs, the team reports in a paper posted April 7 on bioRxiv.

While there still is not enough information about pennaraptorans to conclude that their protowings were in fact used for flush displays, “what this shows, rather elegantly and persuasively, is that it’s possible,” says paleontologist Corwin Sullivan of the University of Alberta in Edmonton, Canada.

And that’s not all. “Let’s say that the early pennaraptoran feathers were being used for these flush displays,” Sullivan says. “That doesn’t mean they couldn’t have been used for other things.” Jablonski agrees. Both he and Sullivan suggest the dinos could have also waved their wings at potential mates.