Something in the way it moves

Some amphibious animals are able to move seamlessly from walking to swimming, and now scientists think they’ve found out how they do it.

Researchers from Japan, Switzerland and Canada say they have decoded the flexible motor control mechanism underlying how salamanders and centipedes, among others, coordinate their bodies and appendages during this “adaptive locomotion”.

They did it by watching the Chinese red-headed centipede (Scolopendra subspinipes mutilans), which is easy to watch because its homogeneous and segmented body structure puts all the changes on show.

This centipede walks on land by coordinating its many legs, but on entering water folds its legs and swims by bending its body trunk much as an eel does.

The research team, led by Akio Ishiguro from Japan’s Tohoku University, observed intact and nerve transected during this transition and suspected that interactions between the central nervous system, the peripheral nervous system, the body, and the environment could explain it.

They hypothesised that walking or swimming signals generated in the brain are sent posteriorly via distributed neural networks belonging to the central nervous system and located along the body, and that these signals can be overridden by sensory signals felt by the peripheral nervous system of the legs when they touch the ground during walking. 

They then described this multiple-signal mechanism mathematically and reproduced the behaviour of centipedes in different situations through computer simulations. 

Writing in the journal Scientific Reports, they report two main findings. The first was that walking can be initiated by mechano-sensory inputs alone, whereas swimming needs a descending control from the brain for initiation. 

The second was that sensory feedback from the legs can override a swimming pattern directed from the brain and elicit a walking pattern.

“This suggests two distinctly different roles of sensory feedback for swimming compared with walking,” says co-author Emily Standen, from Canada’s University of Ottawa. 

“This adds information about how animal nervous systems can integrate and use sensory feedback to display functional locomotion.”

The authors acknowledge some of the limitations in a study that built a highly abstract model based on behavioural experiments and tested it with simulations.

They suggest, however, that the simplicity of the essential control mechanism described in the model “can be a basis for discussing common principles of locomotion control among various animal species whose detailed structures of neural and mechanical systems are different but may operate on similar mechano-sensory feedback systems”.

They also see the potential to help develop robots that can move on various environments by flexibly changing body coordination patterns.

“In the field of amphibious robots, a centipede-like robot based on our model has the following two advantages,” they write. 

“First, it can utilise two different locomotion to achieve effective propulsion depending on the environment; legged locomotion on land and undulatory locomotion in water. Second, the robot can be fault-tolerant to physical damages owing to its redundant multi-segmented body and its decentralized control. 

“Therefore, robotic implementation of our centipede model will pave the way to realise a highly adaptive and resilient amphibious robot.”

CREDIT: Ishiguro-Kano Lab