Upshall, Mary2025-05-082025-05-082025-05-08http://hdl.handle.net/10393/50423https://doi.org/10.20381/ruor-31087Weakly electric fish have an electric organ which generates an electric field surrounding their body. Electroreceptors embedded in the skin of these fish encode the field perturbations produced by environmental objects forming the basis for electric sensing. Interestingly, the spatial properties of the electric field change significantly with body pose, but how this affects sensing is not clear. Using novel computational and statistical techniques, this thesis addresses the impact of body-pose on electric sensing in the wave-type electric fish, Apteronotus albifrons. We developed an unsupervised behavioural identification pipeline to classify the movements which make-up the behavioural repertoire in freely-swimming fish. We then investigate the impact of different environments on the underlying statistics of the behavioural repertoire. We found that certain movements were context specific: scanning (quick back-and-forth swimming) is more prevalent in static environments, while tail-bending is more frequent in temporally varying environments. Finally, we simulate the fish’s electric field under the same conditions to estimate the available electrosensory information. These results suggest that body-pose may improve information acquisition in some contexts. Overall, this thesis combines behavioural and computational approaches to suggest that self-motion improves electrosensory acquisition in wave-type weakly electric fish.enAttribution 4.0 Internationalhttp://creativecommons.org/licenses/by/4.0/Weakly electric fishApteronotus albifronsBehavioural classificationUnsupervised machine learningElectrosensationActive sensingScanningAnimal behaviourThesis