“Inelegant” is a generous way to describe how Polypterus moves on land.

A small brown fish with a wide flat head and strong pectoral fins, Polypterus wriggles powerfully when Valentina Di Santo, a fish physiologist from Stockholm University, places it on a mesh mat on the floor of Rowe Laboratory at the Marine Biological Laboratory (MBL).

“This one’s a jumper,” says Di Santo, a visiting scientist in MBL’s Whitman Center, which hosts researchers from around the world. She watches as the Polypterus writhes, periodically launching itself into the air in twists and spins. But then the fish calms down and begins to do something very un-fishlike. Contorting from side to side, Polypterus swings one pectoral fin forward, then the other, slowly dragging its body across the ground like a clumsy salamander. In a rudimentary way, it’s walking.

Walking is among the most important moments in vertebrate evolution. The first fish that crawled onto land diversified quickly to exploit an abundance of resources, giving rise to everything from amphibians and reptiles to birds and mammals.

This oft-repeated story, however, obscures the real moment of innovation. “[Walking] really happened long, long before the first fish ever thought about going onto land,” says Di Santo. Before fish walked on land, they were able to walk underwater.

The first to climb onto shore needed strong fins to support its weight. It’s probable that before land walking, fish would have adapted to “proto-walking” along the seafloor. In all likelihood, this underwater walking behavior is the origin of the transition to land.

Understanding how walking began is key to understanding our ancient past. Along with UChicago paleontologist Neil Shubin, Di Santo is working to answer a simple yet vexing question: Why would something that could swim ever choose to walk?

Di Santo and Shubin believe the answer is efficiency. Fish are inherently unstable underwater, especially at low speeds when they need to move their fins to control posture—and swimming slowly requires a lot of energy. Like riding a bike, swimming is much less stable at slow speeds. Being able to stroll along the seafloor might be like getting off the bike and walking.

At the front of the lab is a transparent three-foot-long tank shaped like a racetrack; inside is a treadmill with a high-speed camera underneath. This hybrid flow tank/treadmill allows Di Santo and Shubin to watch exactly what happens when a fish starts walking.

When the experiment begins—with the tank full of water and a fish at the starting line—the treadmill and flowing water start moving at the same speed. “We expect that fish may walk rather than swim at the lowest speeds,” Di Santo says. As the water flow velocity increases, though, there will be a point when the fish may choose to lift up in the water column and swim instead. All the while, the high-speed camera films how the fish moves its fins, and an oxygen meter measures how much energy the fish expends as it transitions from walking to swimming.

“We’re trying to understand the rules of walking,” Di Santo says—which fins the fish uses, how it moves its body, when it chooses to swim and to walk, whether there is a transitional period of hybrid “skipping” locomotion. The scientists will test at least 11 different species, including sharks, rays, gobies, and lungfish, trying to understand what defines their walking style. Polypterus will be tested, too, but only after it undergoes special training.

This training happens next to the flow tank, in what Di Santo calls a “Polypterus gym.” Some of the fish here live in tanks with no water, just mist coming down from pipes. Polypterus do just fine. They have lungs, so they can breathe air as long as misters keep them moist. Without water to swim in, though, they have to walk.

Di Santo and Shubin hope to test how these fish change their walking style—and the efficiency of it—as they get more familiar with life on land. In some tanks, they have even installed pebbly hills for the fish to climb to see if uneven terrain might drive changes in their walking style. The fish will remain in this enclosure for three months to a year. “There’s a chance that after so long out of water,” Di Santo says, “these fish may start to move more efficiently.”

This audacious idea—training Polypterus to behave like a salamander—is not the final step of the project. While Di Santo runs the flow tank experiments, Shubin will do anatomical studies to figure out the exact anatomy of these fishes’ fins, both normal Polypterus and those in the terrestrial tanks. Then, armed with Di Santo and Shubin’s data, roboticist Fumiya Iida of the University of Cambridge will try to construct a robotic Polypterus.

If Iida can make a robot that walks and swims like the real thing, he can try creating robot versions of other walking fish and compare them to their real-life counterparts.

They could even bring the fossil fish Tiktaalik to life.

Tiktaalik was a fish that lived 375 million years ago. It was long and crocodile-shaped, spending most of its life in water but occasionally walking on land. When Shubin codiscovered the fossil Tiktaalik in the Canadian Arctic in 2004, it was hailed as the “missing link” between sea and land vertebrates—our ancestor.

Unfortunately, fossils can only tell us so much. But if Iida could make a robot Tiktaalik, scientists like Di Santo and Shubin could examine how it might have moved. A realistic robotic Tiktaalik could reveal a lot about why fish decided to take to the land.

“Fishes are so much like us,” Di Santo says. Both fish and people try to save energy when they can, she says. This quest to save energy is so powerful that it might have laid the foundation for the evolutionary success of land-dwelling vertebrates: one day a fish grew tired of swimming and decided to take a walk along the seafloor, kicking off one of the most important events in the history of our planet. Thanks to Di Santo and Shubin’s research, we may one day better understand why that change came to be.