Even experts often resort to the word “bizarre” when describing pterosaurs, the winged dragons that ruled the skies for more than 160 million years. This is especially true of the group of short-tailed pterosaurs called anurognathids, which used to dart and bob through Mesozoic era forests like bats, hawking for insects.

Now it appears anurognathids and other pterosaurs may also have worn a weirdly varied coat of feather- and fur-like structures, according to a new study published Monday in Nature Ecology & Evolution. A team led by paleontologist Zixiao Yang from Nanjing University in China reached that conclusion based on two near-complete, pigeon-size anurognathid pterosaur specimens found in northern China.

The idea that pterosaurs (which lived from around 228 million years ago to the Cretaceous extinction 66 million years ago) may have had some kind of furlike coat is not by itself new. Researchers have proposed as much since the discovery of the first known pterosaurs in the 19th century. But the exact character of this covering has been difficult to determine from the short, filamentlike structures—called pycnofibers—preserved in pterosaur fossils. The new study set out to fill in that gap with the help of a battery of advanced technological tools. As a result, the authors characterize what they say are four different types of pycnofibers, distributed around the animal’s body in ways that suggest different types of pycnofiber performed different functions: thermal insulation on the neck and head, for example, or reducing drag on the wings. One type of pycnofiber is a simple, hairlike monofilament. But three others appear to be branched in a way the authors describe as “remarkably similar” to bird feathers. The similarities go beyond shape, or morphology, they say, to resemblance at chemical and cellular levels.

Based on this finding, the study also argues “featherlike branching integumentary structures” may have evolved first not in dinosaurs, as generally thought, but in some primordial archosaur—a common ancestor of both pterosaurs and dinosaurs, including modern birds. This would mean the ancestor even of decidedly nonavian dinosaurs like Stegosaurus might have been covered in quills, rather than scales. It would also push the origin of feathers out of the Jurassic period and back 60 million or 70 million years to the dawn of the Triassic period.

That early date for the appearance of feathers would fit, says Michael Benton, senior author of the new study and a paleontologist at the University of Bristol in England, with the transition from a sprawling to an upright stance and to warm-bloodedness in many animal groups—along with other evidence indicating “the pace of life sped up” as Triassic species struggled to recover from the Permian–Triassic mass extinction, in which 70 percent of terrestrial vertebrates had vanished some 252 million years ago. It would also fit with evidence most of the genes controlling feather production were present in vertebrates before the origin of dinosaurs.

The counterargument, Benton says, is that big dinosaurs like Stegosaurus or Brontosaurus lacked feathers. But that is no more strange, he says, than elephants or whales having little or no hair—even though both evolved long after the evolution of hair in mammals.

In a commentary published in the same issue, behavioral ecologist Liliana D’Alba of Ghent University in Belgium, who was not involved in the new study, remains skeptical. The study demonstrates that the chemical composition of the pycnofibers is similar to that of feathers, she wrote, and both scanning electron microscopy and energy dispersive x-ray spectroscopy show the fibers contain melanosomes—the same pigment packets that impart color to feathers and to mammal hairs. But the assertion some pycnofibers are branched like feathers is based, she says, on subjective interpretation of “gross filament morphology,” or shape. She notes that a previous attempt by other researchers to characterize pycnofibers as featherlike failed to persuade most paleontologists. It may require developments in other advanced technologies, she suggests, to resolve the question.

“Does this work show that archosaur skin was more complex than we knew? Yes,” says Yale ornithologist Richard Prum, whose extensive knowledge of feathers was the basis for his prize-winning 2017 book The Evolution of Beauty. (Prum also was not involved in the new study.) “Does it show that archosaurs grew all sorts of interesting stuff from their skin? You bet. All you have to do is look at a turkey beard to see that genuinely new stuff can evolve on the skin of an archosaur.” But Prum says the authors’ big conclusion is “flawed” because they overlook this evolutionary knack for novelty; just because pterosaurs produced some weird featherlike structures does not automatically imply feathers must have emerged in some common ancestor of pterosaurs and dinosaurs. “These pterosaur skin appendages are cool,” Prum says, “but their branched structure is not homologous with that of feathers”—that is, they do not have a shared evolutionary origin. “And they are probably not homologous with feathers at all,” he says. “In short, they ain't feathers.”

Most of them are not even pycnofibers, says pterosaur specialist David Unwin at the University of Leicester in England, who was not involved in the study. “These are fantastic specimens, and they did a brilliant job of imaging them,” he adds. But he contends the researchers are mistaken when they use keratin content to identify certain structures as external pycnofibers. Those structures, he says, are almost certainly pterosaur wing tissues called actinofibers, which may also contain keratin. None of the new study’s nine authors has experience with soft-tissue preservation in pterosaurs; perhaps as a result, Unwin says, they fail to reference other relevant studies—for instance, of pterosaur melanosomes. Their interpretation of the evidence, he adds, “is problematic, to say the least.”

Benton challenges critics to show “the structures in pterosaurs are morphologically or chemically different from feathers.” Meanwhile, he says, “we are taking a parsimonious view” that they are in fact feathers. He likens the new study to putting up a kite: “We are just setting up a hypothesis that can be tested.”

Attempts to bring down that kite—by rhetorical shotgun fire—have now commenced.