TY - JOUR T1 - Complete neuroanatomy and sensor maps of Odonata wings for fly-by-feel flight control JF - bioRxiv DO - 10.1101/2021.04.11.439336 SP - 2021.04.11.439336 AU - Joseph Fabian AU - Igor Siwanowicz AU - Myriam Uhrhan AU - Masateru Maeda AU - Richard J Bomphrey AU - Huai-Ti Lin Y1 - 2021/01/01 UR - http://biorxiv.org/content/early/2021/04/11/2021.04.11.439336.abstract N2 - “Fly-by-feel” describes how flying animals capture aerodynamic information via their wings’ sensory system to implement or enhance flight control. Traditional studies on animal flight emphasized controlling body stability via visual or inertial sensory inputs. In line with this, it has been demonstrated that wing sensory systems can provide inertial state estimation for the body. What about the state estimation of the wings themselves? Little is known about how flying animals utilize their wing sensory systems to monitor the dynamic state of their highly deformable wings. This study is a step toward a comprehensive investigation of how a flying animal senses aerodynamic and aeroelastic features of the wings relevant to flight control.Odonates: dragonflies and damselflies, are a great model for this because they have excellent flight performance and their wing structure has been extensively studied. Here, we developed a strategy to map the entire sensory system of Odonata wings via confocal microscopy. The result is the first complete map of a flying animal’s wing sensory system, including both the external sensor morphologies and internal neuroanatomy. This complete search revealed over 750 sensors on each wing for one of the smallest dragonfly species and roughly half for a comparable size damselfly. We found over eight morphological classes of sensors, most of which resembled mechanosensors. Most sensors were innervated by a single neuron with an innervation pattern consistent with minimising wiring length. We further mapped the major veins of 13 Odonate species across 10 families and identified consistent sensor distribution patterns, with sensor count scaling with wing length. To explain the strain sensor density distribution along the major veins, we constructed a high-fidelity finite element model of a dragonfly wing based on micro-CT data. This flapping wing model revealed dynamic strain fields and suggested how increasing sensor count could allow encoding of different wing states. Taken together, the Odonate wing sensory system is well-equipped to implement sophisticated fly-by-feel flight control.Competing Interest StatementThe authors have declared no competing interest. ER -