Electric field maps and boundary element simulations of electrolocation in weakly electric fish

Weakly electric fish use electroreception - the generation and detection of electric currents - to explore the world around them. Neurophysiological studies of these fish have greatly increased our understanding of central electrosensory processing, and have significant implications for sensory processing in the cerebellum and cerebellar-like neural structures. This thesis addresses a particular deficiency in our understanding of electrosensory systems: the input pattern of currents stimulating the fish's electroreceptors has not yet been well defined. My goals were to quantitatively reconstruct the electric organ discharge (EOD) and electrosensory images detected by weakly electric fish. These were accomplished by mapping the EODs of six gymnotiform species, simulating the EODs in two (Eigenmannia and Apteronotus), and predicting the electrosensory input during exploratory behaviors. The EOD maps display a wide diversity of species-specific patterns, implying significant differences in field generation, sensory input patterns, possible behavioral strategies, and processing algorithms. Each fish must interpret electrosensory images which are highly dependent upon its own particular EOD pattern. To study electrolocation noninvasively during natural behaviors, I developed a 3-d electric fish simulator based on the boundary element method. The simulator solves Poisson's equation for the electric potential in and around the fish, modeled as an electrostatic boundary value problem. Models of two species were built and optimized to match the measured maps. By varying only a few parameters, I explored how the electric organ structure and activation generate a particular EOD pattern: Eigenmannia has a synchronous electric organ and dipolar EOD, while Apteronotus is better described as a propagating multipole. The simulator was used to reconstruct the EOD during a previously published tailprobing behavior of Eigenmannia, and from my own videotapes of Apteronotus exploring objects under infrared light. Simulations of selected exploratory behaviors revealed the EOD fields, modulations from body orientation and objects, and the resulting electrosensory patterns. The results (1) imply the fish control their body positions to regulate particular features of the electrosensory image, (2) predict features of the electrosensory input reaching the brain, and (3) suggest algorithms needed to extract useful signals from the electrosensory stream.