Mechanosensory representation of wing deformations
Efficient representation of structural deformations is crucial for monitoring the instantaneous state of biological structures. Insects' ability to encode wing deformations during flight demonstrates a general morphological computing principle applicable across sensory systems in nature as well as engineered systems. To characterize how relevant features are encoded, we measured and modelled displacement and strain across dragonfly wing surfaces in tethered and free flight. Functional interpretations were supported by neuroanatomical maps, and ablation and perturbation experiments. We find that signal redundancy is reduced by non-random sensor distributions and that morphology limits the stimulus space such that sensory systems can monitor natural states with few sensors. Deviations from the natural states are detected by a flexible population of additional sensors with many distinguishable activation patterns.