How squirrels learn to jump – Wissenschaft.de

When jumping from branch to branch, squirrels find trade-offs between jumping distance and branch flexibility. A new study shows how young acrobats make their decisions and make up for mistakes to prevent fatal falls. Both biomechanical adaptations to habitat aid the treetops and their ability to learn. The findings could also contribute to the development of robotics for challenging application areas.

Squirrels spend most of their lives high in trees – a complex environment in which they have to skillfully move through a maze of branches to find forage and nest and avoid predators. This requires a highly developed set of biomechanical skills and learned behaviours. But how exactly do squirrels determine which leap to risk and which one could lead to a potentially fatal downfall?

Croissants jumping tests

That’s what a team led by Nathaniel Hunt of the University of Nebraska looked at. “As a model organism for understanding the biological limits of balance and movement, I think squirrels are second to none,” Hunt says. “If we understand how squirrels move in treetops, perhaps we can discover general principles for navigating complex terrain that could also apply to the movements of other animals and robots.”

In order to observe the performance of the jump in the most natural environment possible, the researchers lure some fox squirrels (Sciurus niger) to the campus of the University of California, Berkeley. They used the nuts to motivate the tiny rodents to make bold jumps in a course designed by the researchers. They documented the movements of the test animals on video tape and then assessed their jumping position, flight path and landing maneuver. In their experiments, Hunt and colleagues used springboards of different lengths and flexibility. From here the croissants can reach a narrow branch into which the walnut is lured into a small bowl. The more they ventured onto the flexible jumping platform, the less they had to jump. Instead, the base of the take-off became more and more vibrating with each step, which made it difficult to take off.

acrobatic maneuvers

Observations showed: “When squirrels jump over a gap, they choose their jump point depending on the flexibility of the branch and the size of the gap,” Hunt says. They got too far on stable diving boards, while on wobbly boards they’d rather jump a little longer than venture too far forward. The flexibility of the jumping rule played a six times greater role in making the decision than jumping distance. “When they hit a branch with new mechanical properties, within a few jumps they learned to fine-tune the starting mechanics,” Hunt explains. “This flexibility of behavior, which adapts to the mechanics and geometry of jumping and landing structures, is important for jumping over a gap with precision and landing on a small target.”

“During the entire experiment, no squirrel fell while trying to jump,” the researchers said. They made mistakes and jumped either too far or too short to land optimally on their narrow target. But each time they managed to cling to the branch with their claws and pull them with acrobatic maneuvers. “If you jump through the air too quickly or too quickly, you can use a variety of landing maneuvers to compensate for that,” Hunt explains. “If they jump too far, they roll forward around the branch. If they jump too short, they land on their front legs and swing under the branch before pulling themselves onto the perch. This combination of adaptive planning behavior, learning control, and reactive stabilization maneuvers helps you move quickly across the branches without falling.”

In addition, the researchers noticed an unexpected tactic by people tested on animals: during difficult jumps, they often pushed themselves away from the wall to which the branches were attached, and in this way they adjusted their speed and direction – similar to human runners. Squirrels have adapted flexibly to the new possibilities of their environment and quickly learned to use them to make jumps more efficient and safe. The study shows that there is likely to be a synergy between biomechanical properties and trial-and-error learning that enables animals to move quickly and skillfully in a canopy. The findings could also aid the development of resilient robots that can safely traverse rough terrain.

Quelle: Nathaniel Hunt (University of Nebraska) et al., Science, doi: 10.1126/science. abe5753

© wissenschaft.de – Elena Bernard