Meeting Abstract

3.4  Monday, Jan. 4  Motion feature detection in a biological gyroscope FOX, J.L*; DANIEL, T.L.; University of Washington; University of Washington jessfox@uw.edu

Sensory systems acquire information from the environment, filter it, and transmit it to the nervous system in the form of action potentials. For mechanoreceptors, the relevant information can take the form of numerous kinds of forces occurring in multiple directions. While it is known that halteres of dipteran insects experience large inertial forces during flight and mediate behavioral responses to perturbations in the flight path, less is known about the neural encoding of such forces. Here, we use band-limited white noise mechanical stimulation combined with single-cell recording to determine the features of the motion stimulus that elicit responses (neural feature selectivity). We recorded >1000 neural spikes from each of 36 cells in 7 animals. We find that two principal motion features, the second of which is the derivative of the first, can accurately predict the spiking activity of haltere neurons. Moreover, these features are similar between cells and differ only in their phases. We posit that a population of haltere neurons with similar feature selectivity and distinct phase responses could encode a variety of inertial forces, including Coriolis forces. During a body rotation, the phase of the wingstroke at which peak strain occurs will shift depending on the magnitude of the body rotation. Thus, the identity of the first neurons activated will also vary with the magnitude of body rotation. The central nervous system could therefore discriminate different force magnitudes by detecting the order in which the primary afferents fire. Our data suggest that this type of place code, supported by a population of fast and highly precise neurons with broad frequency sensitivity and sharp feature detection, may be a viable strategy for encoding the complex inertial forces associated with high-speed flight maneuvers.