Many sensory systems have some kind of representation in the brain that seems to form a kind of spatial map. One well-studied case of this is a map of directional responses from the tiny wind-sensitive hairs on organs called cerci at the back end of a cricket, which has been used as a relatively simple model system. The nerve fibers carrying information from these cerci come in to a central cluster called a ganglion that is part of the animal's central nervous system. Crickets, like many other invertebrates, often have single nerve cells that are characteristically identifiable from animal to animal, so it is possible to construct an atlas of such identifiable neurons, and relate their activity to the behavior of the animal, such as turning toward or away from a stimulus. This project builds on an existing large database of identified neurons in this system, to add dynamic properties about their physiological responses to the known spatial properties of their connectivity from the cercal inputs, and their outputs to other parts of the ganglion and the brain. Data on response properties including threshold sensitivity, frequency tuning, phase response, and rates of adaptation are incorporated and used to derive a set of mathematical functions. These functions, representing the response dynamics of the cells in the system, are projected on to the three-dimensional matrix of the spatial map. Advanced computational visualization algorithms will be implemented to represent the spatial patterns of activity within the map. The impact of this work will be not only on mechanosensory integration, but on general issues of information processing in nervous systems, combining the constraints of anatomical structure and the insights about computational algorithms used in neural maps, which allow the organism to respond with appropriate spatial behavior to environmental signals.
