Dinosaurs Evolved Flight at Least Three Times

A new study finds that many feathered dinosaurs were more aerodynamic than previously thought

Microraptor Eating Fish
A Microraptor, a small four-winged dinosaur that could fly, eats a fish. Emily Willoughby / Stocktrek Images via Getty Images

Flight is a relatively rare ability. Many animals crawl, slither, burrow, walk and swim, but comparatively few have the ability to take to the air. There’s something about evolving the ability to fly that’s more difficult than other ways of getting around. Yet, despite these challenges, dinosaurs didn’t just evolve the ability to fly once, but several times.

The previously-unappreciated aerodynamic capabilities of many feathered dinosaurs allowed more of the “terrible lizards” to fly than previously thought. That’s the conclusion of a new study conducted by University of Hong Kong paleontologist Michael Pittman and colleagues, published earlier this year in Current Biology. Instead of flight evolving as a single process of greater aerodynamic ability in one lineage alone, the process was something that might be called experimental with many different feathered dinosaurs moving, flapping, fluttering and flying in different ways. “The current paradigm shift involves recognition that flight was arising independently from different, closely-related groups at nearly the same time,” Pittman says. “This moves away from the traditional idea that flight is a rare gem.”

Much of what we know about dinosaur flight comes from birds. That’s because all birds are living dinosaurs, the last remaining members of the family. The raptor-like ancestors of birds split off from their closest dinosaur relatives during the Jurassic, about 150 million years ago, and were just another part of the Age of Dinosaurs. When an asteroid strike sparked a mass extinction 66 million years ago, beaked birds were the only dinosaurs to survive the catastrophe and carry on the legacy of the terrible lizards to the present day.

But this picture is still relatively new. For decades, in books and museum displays, paleontologists differentiated dinosaurs from other ancient reptiles by the fact that dinosaurs didn’t fly or swim. “Flight is not something dinosaurs were traditionally expected to do,” says Pittman. The change came not only from new discoveries, including finds of feathered dinosaurs, but from new ways of analyzing and thinking about fossils. Beyond the gross anatomy of the fossils, paleontologists use an evolutionary classification called cladistics focused on which traits are shared between animals—a technique that allows a clearer picture of how each dinosaur species is related to others. Being able to discern who is most closely related to whom—such as which non-bird dinosaurs were most closely-related to the first birds—is an important part of reconstructing how feathery dinosaurs evolved the ability to fly. Paleontologists have also been able to borrow from engineering techniques to study the aerodynamic capabilities of feathered dinosaurs, allowing experts to better test which species could flap through the air and which were permanently grounded.

In the new Current Biology study, the evolutionary tree of dinosaurs related to birds lined up with what paleontologists have reported. The closest relatives to early birds, the study found, were deinonychosaurs—the family of raptor-like, feathered dinosaurs that contains the likes of Velociraptor and Troodon. But then the researchers went a step further. By looking at whether the dinosaurs could overcome some of the mechanical constraints needed to make flapping motions required for flying, the paleontologists found that the potential for deinonychosaurs to fly evolved at least three times.

Given that all living vertebrates capable of powered flight leap into the air—whether bats or birds—Pittman and colleagues hypothesize that dinosaurs did the same. Even though paleontologists previously debated whether dinosaurs evolved flight from the “ground up” by running and jumping, or from the “trees down” by gliding, the fact that living animals take off by leaping indicates that deinonychosaurs did, too, regardless of what substrate they pushed off from. “This isn’t exclusive to take off from the ground or from height,” Pittman notes, “so birds in a tree also leap to take off.”

Naturally, birds and their closest relatives—such as the small, magpie-colored deinonychosaur Anchiornis—had the anatomical hallmarks of powered flight. These dinosaurs were small, had lightweight bones, possessed long feathers along their arms and had strong legs that allowed the dinosaurs to leap after prey—and sometimes into the air. The researchers also looked at wing loading, or the size of each deinonychosaur’s wing relative to their body size. By comparing the wing loading estimates to figures derived from animals known to fly today, the researchers were able to narrow down which deinonychosaurs could likely fly and which could not.

In addition to the deinonychosaurs most closely-related to birds, the paleontologists found that two other deinonychosaur lineages had wings capable of powered flight. Within a group of Southern Hemisphere raptors called unenlagines, a small, bird-like dinosaur called Rahonavis would have been able to fly. On a different branch, the four-winged, raven-shaded dinosaur Microraptor shared similar abilities. More than that, the researchers found a few other species on varied parts of the deinonychosaur family tree—such as Bambiraptor and Buitreraptor—that were getting anatomically close to fulfilling the requirements for flight. Flight wasn’t just for the birds, in other words. Many non-bird dinosaurs were evolving aerodynamic abilities, but only a few were able to actually flap their wings and fly.

“The new paper is really exciting and opens new views on bird origins and the early evolution of flight,” says Bernardino Rivadavia Natural Sciences Argentine Museum paleontologist Federico Agnolin. So far, other studies haven’t found the same pattern of dinosaurs evolving flight more than once. Given that dinosaur family trees are bound to change with the discovery of new fossils, Agnolin adds, this might mean that the big picture of how many times flight evolved might change. All the same, he adds, “I think that the new study is really stimulating.”

The major question facing paleontologists is why so many feathered dinosaurs evolved the ability to fly, or got close to it. Flight has particular physical requirements—such as wings capable of generating enough lift to get the animal’s weight off the ground—and paleontologists have long proposed that what dinosaurs were doing on the ground might have had a role to play in opening up the possibility to flight. “Repeated evolution of powered flight is almost certainly related to feathery deinonychosaurs doing things that opened up the possibility of flight,” Pittman says. Feathers were important to display, insulation, fluttering to pin down prey, flapping to create more grip while running up inclines and other activities. Becoming more maneuverable on the ground, in other words, may have helped dinosaurs repeatedly stumble upon the ability to fly.

Getting a clearer picture of when and how flight evolved among dinosaurs surely rests on finding more fossils. Each one adds another paleontological puzzle piece in the effort to understand when and how dinosaurs evolved the ability to fly. Now flight seems to have evolved more than once, experts may very well find new dinosaurs that were not ancestors of birds but still took to the skies all the same. As paleontologists continue to scour rocky outcrops and museum collections for new clues, a new understanding of flight in the Age of Dinosaurs seems cleared for take off.

Get the latest Science stories in your inbox.