Why Is This Ostrich Wearing an Extra Set of Wings?

A young ostrich sprints down a path at a Chinese farm. His neck bobs, his legs pump — and the artificial wings attached to his back flap up and down.

Alas, this ostrich will not fly. He’s filling in for another earthbound creature, a dinosaur called Caudipteryx. Some 125 million years ago, this theropod walked on two legs and bore a pair of feathery “proto-wings,” similar to the fake ones worn by the ostrich. Recently, researchers set out to study the long-extinct creature as they took on one of the greatest controversies in paleontology: how avian flight first evolved.

In a study published Thursday in PLOS Computational Biology, a team including mechanical engineer Jing-Shan Zhao and several paleontologists used the outfitted ostrich — along with mathematical and robot models — to argue that when Caudipteryx ran, its mini-wings flapped involuntarily. Eventually, the researchers proposed, the dinosaur’s descendants would have harnessed this trait and taken off from the ground for the first time.

Other experts are less than convinced, arguing that the study’s analogues don’t do justice to the complexity of the animal it purports to study. But the disagreement itself highlights the unsettled nature of debates over the evolution of flight among the world’s feathered inhabitants.

“You’d be hard-pressed to find a paleontologist who doesn’t have an opinion,” said Armita Manafzadeh, a doctoral student at Brown University who was not involved in the study.

Because modern birds evolved from a particular lineage of dinosaurs, the theropods, researchers look to them to figure out how flight began. Some promote a “top-down” theory, contending that winged dinosaurs learned to fly by climbing trees and gliding from the branches, not unlike flying squirrels. Others are on the “ground-up” side, positing that dinosaurs went airborne by running and flapping. Still others question the dichotomy itself, arguing that the two theories are not mutually exclusive.

For their entrance into the fray, Zhao’s group performed three experiments. All focused on Caudipteryx, a useful organism for studying “the evolution from the earliest dinosaur to the modern bird,” said Zhao, who teaches at Tsinghua University in Beijing. Caudipteryx was discovered in China in the late 1990s, and, in renderings, often resembles an angry, oversized chicken.

For the first experiment, the group used fossil analyses to develop a simplified mathematical model of the dinosaur. They then simulated a running motion in the model’s legs, and calculated how other parts of the body responded. They found that at a fast enough clip — between 5.5 and 12.9 mph, a range within Caudipteryx’s estimated capability — the model’s wings flapped.

For the second test, they built a robo-Caudipteryx and ran it on a treadmill. Here, too, they observed consistent oscillation, even as they increased and decreased wing length.

Then it was time for a live test. Zhao and his colleagues built a harness with force sensors and a set of wings. They augmentedmany birds, including ducks and geese, before settling on a 6-month-old ostrich.

“Its size and weight looked like the real size and weight of Caudipteryx,” Zhao said. (Because living dinosaurs aren’t available for paleontology studies, their descendants, from baby partridges to chickens with prosthetic tails, often serve as proxies.)

As the ostrich ran, flapping was evident, and the sensors measured a small amount of lift: The motion was counteracting some of the bird’s body weight. Zhao suggests that Caudipteryx might have noticed that the weird thing its wings were doing was actually making it easier to run, and leaned into it. In this way, the passive jostling provided “basic training for later flapping flight,” he said.

Is it conclusive proof that flapping came first? Other researchers, though impressed by the study’s three-pronged approach, expressed doubts.

Dennis Voeten, who studies flying dinosaurs at the European Synchrotron Radiation Facility, in France, and Sweden’s Uppsala University, thinks the group’s conclusions went too far. Observing oscillation “does not prove that this dynamic ‘trained’ early feathered dinosaurs in executing a birdlike flapping motion,” he said. Manafzadeh added that simplifying dinosaur anatomy has its pitfalls, as does using passive, plastic wings to stand in for active limbs.

Zhao responded that the benefits of flapping should have been enough to nudge evolution in that direction, and that the group’s experiments left room for various anatomical considerations.

“You can change the mass, or the stiffness of the muscles, anything,” he said. “The flapping motion will not change.”

This article originally appeared in The New York Times.