Around 124 million years ago, in what's now China, a turkey-sized, feathered dinosaur called Caudipteryx zoomed across the ground through an early Cretaceous forest.

In many ways, it looked a lot like the early birds that were learning to navigate the skies of the world at the time.

It had wings and a tail featuring the same kind of feathers that those birds were using for flight ones that are long, broad, and vaned.

But while Caudipteryx had flight-type feathers, it couldn’t actually fly its wings were simply too small for that.

So the question of what non-flying dinosaurs like Caudipteryx were doing with their feathered wings and why they originally evolved in the first place has been a mystery since they were discovered.

But thanks to what might just be one of the strangest experiments in the recent history of paleontology involving a makeshift robot-dino a new hypothesis has emerged Early feathered proto-wings might have actually evolved to be scary.

This may have been a case of fright before flight.

Caudipteryx is a member of a group of dinosaurs called the Pennaraptorans, which includes modern birds as well as their close extinct relatives.

And the group gets its name for the unique kind of feathers they have: pennaceous feathers, often found on their forelimbs and tails.

Pennaceous feathers are rigid and flat, and they’re structured around a central shaft known as a rachis.

If you imagine a feather, you’re almost certainly picturing a pennaceous one.

But while today they seem like the classic form of feathers, and are thought of primarily as tools for flight, feathers didn’t evolve this structure until relatively late in their history.

And they weren’t originally used for flight at all.

The earliest feathers were simple and hairlike, and evolved at least as far back as the early Triassic Period, 250 million years ago, before the earliest true dinosaurs themselves even emerged.

It's thought that they formed a layer of fuzz primarily for insulation, and possibly for display, too a lot like the fur of mammals.

And when these simpler feathers transformed into pennaceous feathers in early pennaraptorans around 100 million years later, they first appeared in species that, like Caudipteryx, were clearly flightless.

Which means that, even though wings with pennaceous feathers today are about as synonymous with flight as legs are for walking, that clearly can’t have been their original purpose.

So why did such strange and complex structures evolve to begin with?

This mystery started in the 1990s when a trove of feathered non-avian dinosaur fossils started turning up in China and elsewhere, including Caudipteryx.

And for decades now, scientists have been arguing over the story of feathers, and why they still underwent so much development in dinosaurs that, in some cases, weren’t using them for flight.

Could it have been just a refinement of feathers' original role of insulation?

Well, probably not.

Pennaceous feathers aren’t much better at this than their earlier, fuzzier precursors, so that can’t explain the transition.

A more widely accepted idea is communication and display.

See, having sheets of broader, flatter pennaceous feathers might have provided a better canvas,’ allowing for more elaborate coloration and ornamentation.

But there’s still the question of whether communication and display alone could have exerted enough evolutionary pressure to drive such a radical jump in feather complexity.

Perhaps these were things that pennaceous feathers were only adapted for later.

Another idea is that, even if the pennaceous-feathered protowings of dinosaurs like Caudipteryx were too small for powered flight, they might’ve still provided some useful aerodynamic advantages.

For example, in 2003, a researcher proposed the Wing-Assisted Inclined Running hypothesis.

This essentially argued that by flapping their feathered protowings, those dinosaurs could have generated enough downforce to run up steep or even vertical obstacles, to chase prey or escape predators, for example.

And this kind of dinosaur parkour might’ve been a sort of stepping stone on the road to flight.

Since then, wing-assisted inclined running has been one of the leading hypotheses.

But what if these early birds and their ancient relatives were actually using these feathers in ways that are much harder to imagine?

More recently, in 2024, a team of researchers in South Korea proposed a new hypothesis.

One that combined elements of both the display hypothesis and the aerodynamics idea, but put them into a completely different, and much stranger, ecological context What if feathered protowings and tails evolved to scare and chase down small, hidden prey?

This is known as the flush-pursue hypothesis.

And, as the name suggests, it argues that wings and tails with pennaceous feathers would have enabled dramatic flourishes with contrasting patterns that could have startled prey like insects.

By triggering them to flee from their hiding places at the sight of these displays, they would have ended up in the open where they were more vulnerable.

Plus, if feathered wings really did help these dinosaurs maneuver, like the wing-assisted incline running hypothesis argues, this would have also come in useful as they chased down the prey that they’d flushed out of hiding.

It’s a hunting strategy that some modern birds still do today, and it seems to work pretty well for them.

And it’s based on a concept in ecology known as the rare enemy effect’.

See, most predators are not flush-pursuers.

So most prey, like insects, have evolved an escape response of fleeing when they’re startled by sudden movements a response that’s hardwired into their neural pathways.

Flush-pursuers exploit this anti-predator response, purposefully triggering it with dramatic displays to trick their prey into moving from their hiding spots out into the open.

Could this explain why so many non-flying dinosaurs like Caudipteryx had protowings with elaborate pennaceous feathers?

And could it potentially even be a reason that feathered wings evolved in pennaraptorans in the first place?

Well, while it’s plausible on paper, it's hard to actually prove.

These are animals that lived more than a hundred million years ago, and complex behaviors like this usually don’t leave any kind of tangible evidence behind in the fossil record.

So to actually test the flush-pursue hypothesis, the researchers had to, in a sense, resurrect Caudipteryx but in robot form.

They built the robot which they named Robopteryx based on the size, shape, and movement range of Caudipteryx, complete with folding wings and paper feathers.

They then unleashed Robopteryx on some unfortunate grasshoppers a kind of insect old enough to potentially have been preyed on by the real Caudipteryx back in the Cretaceous Period.

They performed a series of experiments on the grasshoppers, with Robopteryx approaching them using different variations of flushing behaviors Including with and without the use of its protowings, with and without highly-contrasting patterns on the wings, and with and without tail feathers.

Sure enough, their experiments showed that the use of wings, patterns, and tail feathers all significantly increased Robopteryx’s success rate of flushing out the grasshoppers the insects clearly found them more startling.

Plus, the researchers also showed the grasshoppers computer animations of a Caudipteryx-like dinosaur performing flushing displays, and measuring the activity of certain neurons in their brains involved in triggering the insects’ escape-jump reactions.

Here too, they found that the activity of these neurons was higher when the grasshoppers were exposed to animations involving proto-wing displays.

These results are all consistent with the idea that feathered wings and tails would have provided a major advantage to non-flying dinosaurs that hunted small prey using a flush-pursuit strategy.

Potentially, this might have been one of the earliest actual functions of pennaceous feathers, as well as the protowings that supported them.

So frightening bugs with dramatic displays could have laid the foundation for the later evolution of larger feathered wings that enabled powered flight to eventually emerge.

Now, as cool and scary Robopteryx is I mean just look at it the flush pursue hypothesis is still far from a sure thing.

Using modern-day recreations is often our best and only option to test deep time hypotheses, but it still can’t prove that’s the way things happened in the past.

And even if it is correct, it’s likely only part of the story of the evolution of feathered wings.

Complex traits rarely have single drivers behind them, and the early evolution of feathered wings was probably driven by a cocktail of multiple pressures and factors that contributed in different ways, at different times, for different reasons.

But this hypothesis at least expands the way that we think about their evolution Not just in terms of what effect they had on the dinosaurs themselves like insulation, locomotion, and display but also the effect that these structures had on the minds and behaviors of the species around them.

And while the scariest dinosaurs are often thought of as the tyrannosaurs, spinosaurs, and other ferocious giants, for the insects of the Cretaceous, the real dinosaur nightmare-fuel might just have been a group of small, flappy, feathered weirdos.