Aerodynamic feathers have turned up where paleontologists least expected them: on the hind limbs of small predatory dinosaurs. The birdlike plumes, discovered in half a dozen fossils from China, surprised paleontologists who believe that flight evolved in running dinosaurs that flapped their arms. Instead, the scientists who studied the finds say, the new specimens suggest that avian flight originated in small dinosaurs that glided from trees.
The six diminutive dinosaurs, known as microraptors, belong to the dromeosaurs, the group thought to be most closely related to birds. Their feathers are mostly arranged like those of modern birds. The body is covered with downy feathers, the tail has a tuft of longer feathers, and the hands have flightlike "primary" feathers. Along the front limb are shorter, "secondary" feathers. The hind limbs sport the same pattern of plumage, confirming a hint published last year when another team found a dromeosaur with a single hind-limb feather.
These rear feathers would make “a perfect airfoil” and probably allowed Microraptor to glide, argue Xing Xu and Zhonghe Zhou of the Institute of Vertebrate Paleontology and Paleoanthropology in Beijing. Although bird fossils predate these 125-million-year-old specimens, similar gliding dinosaurs could have been an evolutionary step from flightless dinosaurs to airborne birds, which would have eventually lost the rear wings, Zhou and Xu speculate in the 23 January issue of Nature. (Other recent evidence, however, favors such a ground-up scenario; see Science, 17 January, p. 329.)
Zhou's argument makes sense to Rick Prum, an ornithologist at the University of Kansas, Lawrence. “I don't think there's any way to interpret it other than as stunning new support for an arboreal stage in the evolution of flight,” he says. “At the moment of the actual taking to the air, dinosaurs are likely to have been four-winged.”
Others aren't so sure. Kevin Padian of the University of California, Berkeley, points out that known dinosaur hips didn't allow the legs to extend out laterally from the body, which would be necessary to spread the rear wings. Larry Witmer of Ohio University in Athens notes that even if Microraptor was a glider, it may have been a failed experiment among one branch of dromeosaurs; powered flight could have evolved independently elsewhere.