Three hundred million years ago, Earth was home to a strange creature the size of a small dog called Orobates pabsti. Most people won’t know its name, but scientists consider the animal to be a close cousin to the last common ancestor of reptiles, dinosaurs, birds, and even mammals.

Now, scientists not only know that O. pabsti walked the Earth, they also know how it would have walked.

Related: How leaf-blowers are helping us study natural selection Scientists collected small-bodied anole lizards and set up a hurricane simulation to study how natural disasters affect animal evolution.

You see, O. pabsti is special—it’s the oldest known critter for which we have found both well-preserved fossil skeletons and fossilized footprints. Neither clue is enough on its own, but together they provide a framework that has allowed scientists to re-create the ancient animal’s gait, first in a computer simulation and then using a real-life robot.

The work is helping biologists better understand when and why locomotion evolved in the ways it did, and it may be an early step in unraveling the ways our ancestors made the move from sea to land.

For instance, the scientists think O. pabsti would have looked a lot like a caiman as it scurried around what is now Germany. This is surprising, says John Nyakatura, an evolutionary biologist at Humboldt University in Berlin and lead author of the study, published today in Nature. Previously, scientists thought that the animals of this time would have walked more like salamanders, with their legs sprawled out and bodies close to the ground.

Caiman, on the other hand, hoist their torsos into the air when they move, an advanced form of locomotion that many scientists believed did not evolve until millions of years later, when egg-laying creatures called amniotes came on the scene.

“If the last common ancestor of the group that Orobates belongs to and the amniotes belong to may have already evolved this advanced locomotion, then it does not seem to be linked to the evolution of the egg,” Nyakatura says.

Following in Orobates’ footsteps

To better understand O. pabsti’s swagger, the scientists first created a three-dimensional digital model of the animal’s skeleton. They then ran computer simulations of what it would have looked like as O. pabsti walked in those fossilized footprints while using different gaits.

Many kinds of animals could walk in those footprints, but they each would do so using slightly different forms of locomotion, says study coauthor Kamilo Melo, a bio-roboticist at Ecole Polytechnique Fédérale de Lausanne in Switzerland. For instance, skinks keep all four limbs low to the ground, while iguanas elevate the front half of their bodies when they walk.

After 512 simulations, it was clear that some gaits would have been anatomically impossible, Nyakatura says, because they would have forced O. pabsti’s bones to clack against each other or caused its wrist joints to break.

Walk like a robot

However, certain factors, such as gravity, friction, and balance, can’t be accurately represented in a computer simulation, Melo says. So the team wanted to try them out in the real world. To do that, the scientists turned to OroBOT, a biomimetic robot.

Bringing a fossil to life: Reverse engineering locomotion

How an animal moves can tell scientists a lot about how it lives, and that is how researchers hope to glean more information about an ancient crocodile-like creature called Orobates pabsti.


OroBOT was able to rule out still more potential strides by tweaking certain parameters, such as how much the ancient animal’s spine would have bent as it moved, how wide its legs sprawled, and the range of motion found in its shoulder joints. Some gaits caused OroBOT to stumble and flip onto its back, Melo says. And that’s not even the worst of it.

“We broke one of the ribs, actually,” he says.

Eventually, the team settled on a caiman-like gait as the most probable match. But just to make sure their methodology was sound, they went back and did the same experiment for living species of caiman and salamander, to see if they could predict how those animals would move using only footprints and digital skeleton models. When these tests produced gaits like those seen in nature, they knew they were on to something.

“Throwing down the gauntlet”

Nyakatura emphasizes that all of these techniques have been tried before, but until now, no one had brought them all to bear in one exhaustive study. “I think we can be quite proud about our idea,” he says.

Julia Molnar, a comparative anatomist at the New York Institute of Technology, seems to agree with that assessment, saying that the paper sets a new standard for the field.

“They’re really throwing down the gauntlet in terms of investigating locomotion in extinct animals,” Molnar says. As for the bigger picture, Molnar says that lesser-known creatures like O. pabsti are really interesting because they can help us better understand the story of how our distant ancestors came to live on land.

“I think this is a very early chapter in a really cool story,” Molnar says.

Editor's Note: This story has been updated to correct Kamilo Melo's affiliation.