The one thing every pterosaur reconstruction needs is the one thing no pterosaur fossil preserves. The wing membrane, that thin sheet of skin and muscle stretched along an absurdly long fourth finger, almost never survives in a form you could lay flat and measure. It rots, it folds, it gets crushed by the weight of sediment until what remains is a smear rather than a shape. So the great wings you see in museums and textbooks, the ones carrying Quetzalcoatlus across a Cretaceous sky, are educated guesses dressed up as fact.
And according to a new study from the University of Bristol, those guesses are too similar to one another. Far too similar, in fact, for a group of animals this strange.
Pterosaurs were the first vertebrates to evolve powered flight, roughly 100 million years before birds got the hang of it, and they kept at it for the better part of 150 million years before vanishing alongside the dinosaurs. Over that vast stretch they sprawled into forms that beggar belief: anurognathids barely the size of your palm, flitting after insects at dusk, and azhdarchids like Quetzalcoatlus with wingspans north of 10 metres, comparable to a small aircraft. You would expect that kind of range to leave a mark on wing shape. A sparrow and an albatross do not fly the same way, and they do not wear the same wings.
Yet when Bristol palaeontologist Benton Walters and his colleagues went looking for that diversity in the scientific record, it simply wasn’t there.
Mapping every wing that could exist
The team gathered 79 published reconstructions, covering eight commonly drawn genera plus a couple of broader families, and ran them through a technique called theoretical morphospace. The idea is a bit like building a map of every wing that could possibly exist, optimal or hopeless, plausible or impossible, and then dropping the real reconstructions onto it to see where they land. Do the shapes spread out across the map the way a diverse group should? Or do they huddle together in one corner?
They huddled. “In living flying animals, such as birds and bats, different lifestyles are associated with distinct wing shapes and flight abilities,” says Walters. “The lack of comparable diversity in pterosaur reconstructions suggests that the reconstructions are missing important variation.”
The researchers set five tests of whether reconstructions behaved like real wings: whether drawings of the same animal cluster together, whether an illustrator’s personal style skews the result, whether reconstructions changed after the field reached a rough consensus in 2011, whether the shapes span as much ground as bone-based studies predict, and whether different proposed lifestyles produce different flight performance. Reconstructions passed exactly one of the five. Style, it turned out, barely mattered; with one notable exception (the controversial reconstructions of David Peters, which cluster anomalously tightly with their oddly narrow wings), individual illustrators scattered all over the place. What didn’t scatter was the animals. Drawings of the colossal Quetzalcoatlus and the tiny Anurognathus ended up sitting almost on top of each other, near the middle of the map, sharing space they had no business sharing.
The trouble runs deeper than artistic habit. When the team projected the pterosaur wings onto a space built from modern birds, nearly all of them piled into the quadrant occupied by albatrosses and other ocean-going soarers. Even Dimorphodon, an early, heavily built pterosaur that was probably a fairly poor flier happiest on the ground, came out looking like a seabird optimised for gliding over open water. That, frankly, is a bit daft.
An argument hiding in a single joint
So where does the sameness come from, if not from the artists? Mostly, it seems, from a quarrel about one attachment point. The rearmost edge of the main wing membrane, the brachiopatagium, has to anchor somewhere on the hindlimb, and researchers have never fully agreed where: the ankle, the knee, somewhere higher still. That single unresolved question drives much of the variation the analysis picks up, because an attachment higher on the leg pinches the wing thin, making a stout flier look like a slender oceanic one. “Reconstructions of pterosaur wings are commonly made using measurements of the bones which support the wing, and information about the soft tissues gleaned from a handful of exceptional fossils, but there is still a lot that cannot be definitively stated from these alone,” says Walters.
There is a tidy irony here. A 2011 paper proposed a consensus wing, attaching to the ankle, and yet reconstructions before and after it look much the same; the field agreed on paper and carried on drawing whatever it pleased.
None of this means the reconstructions are worthless, and the authors are careful not to overclaim. Theoretical morphospace can flag a wing that performs poorly, but it cannot prove a real pterosaur ever flew with an optimal one, and assuming evolution always lands on the best possible shape is a trap the team are keen to sidestep.
What better light might reveal
What the study really delivers is a measuring stick. For over a century the “pterodactyl” has been one of the most recognisable prehistoric animals going, plastered across films and lunchboxes and Jurassic Park sequels, and it remains, on the evidence, something we have not actually pinned down. “For a group of animals that existed for over 100-million years and includes both palm-sized and plane-sized animals, you would expect diversity in shape,” Walters notes. “But wing shape was similar regardless of the pterosaur they depicted.” The gap between what we draw and what once flew is wider than the confident silhouettes let on.
Walters reckons the picture will sharpen as newer methods spread, among them imaging fossils under wavelengths of light the human eye can’t see, which can coax faint traces of soft tissue out of the rock. “This research provides a helpful guide to show where the scientific understanding of pterosaur wings is lacking and will be used as a benchmark to test new and improved reconstructions of pterosaurs as our understanding of these amazing creatures improves,” he says.
Source: “Exploring the limits of wing design in pterosaurs” by B. Walters, E. J. Rayfield and P. C. J. Donoghue, Paleobiology (2026). DOI: 10.1017/pab.2026.10103.
Frequently Asked Questions
If we have pterosaur fossils, why can’t we just look at the wings?
Because the part that mattered most for flight, the membrane, was soft tissue, and soft tissue almost never survives the fossilisation process intact. The handful of specimens that do preserve membrane are folded, torn or flattened, so nobody has ever measured a fully spread pterosaur wing from a fossil. Everything else is reconstructed from the bones and a lot of careful inference, which is exactly where the uncertainty creeps in.
Why does it matter if all the reconstructions look a bit alike?
Wing shape is one of the few windows we have onto how an extinct animal actually flew, so if every reconstruction collapses toward the same shape, we lose the ability to tell a tiny insect-hunter apart from a giant ocean-crosser. The Bristol analysis found exactly that kind of flattening, which means our sense of how varied pterosaur flight really was is probably underselling the group. Pinning down the real range could rewrite how we picture some of the most famous animals in prehistory.
What’s actually causing the reconstructions to come out so similar?
Most of it traces back to a long-running disagreement over where the back of the wing membrane attached to the leg, whether at the ankle, the knee or higher up. Move that one attachment point and you change the whole shape of the wing, thinning out a broad flier until it resembles a slender seabird. Resolving that single question would do more for pterosaur reconstruction than any change in drawing style.
Could new technology fix the problem?
Possibly. Techniques like imaging fossils under wavelengths of light beyond what the human eye can see are starting to reveal faint soft-tissue traces that ordinary photography misses. As those methods spread, the study’s authors hope their map of wing shapes can serve as a benchmark to test whether the next generation of reconstructions captures the diversity the current ones miss.