Meeting Abstract

38.2  Monday, Jan. 5  Decoding Cockroach Antennal Tactile Navigation Using Naturalistic and White Noise Stimuli in a Control Theoretic Framework SPONBERG, S.*; MONGEAU, J.M.; MILLER, J.P.; FULL, R.J.; Univ. of CA, Berkeley; Univ. of CA, Berkeley; Montana St. Univ.; Univ. of CA, Berkeley

Control theoretic models for stabilization and navigation behaviors provide a framework for generating hypotheses of sensory encoding. Cockroaches demonstrate remarkable tactile tracking abilities using antennae to preview surfaces during high-speed wall following in low-light environments. A simple control model suggests a control strategy relying on proportional and derivative information of wall distance. Bulk recordings of all mechanoreceptive units in the flagellum of the cockroachs antenna are consistent with this model showing phasic and tonic components in the envelope neural response. However, it remains unclear if explicit velocity- and position-dependent signals from antennal deflection are encoded in the individual mechanoreceptors and if they are appropriately tuned to match stable wall-following. Here we record small numbers of individual units via (en passant ), suction electrode recordings of the antennal nerve within the head of the cockroach (Periplaneta americana ). The antenna flagellum is driven with a speaker to evoke naturalistic and band-limited white noise responses in antennal position. Recordings from both stimuli reveal derivative and position dependent firing. Activity across the population of mechanoreceptors shows a temporally filtered response that matches the time course of response kinematics, but is composed of differential responses in component mechanoreceptors. We integrate these results with a more anchored version of the earlier control theoretic model, incorporating within-stride lateral plane dynamics. This approach reveals sufficient encoding for stable wall-following and suggests that significant processing occurs at the primary mechanoreceptive afferents.