Weakly electric fish use self-generated motion to discriminate object shape

Abstract

Body movements are known to play an active role in sensing. However, it is not fully understood what information is provided by these movements. The Peters elephantnose fish, Gnathonemus petersii sense their environment through active electrolocation during which they use epidermal electroreceptors to perceive object-induced distortions of a self-produced electric field. The analysis of electric images projected on their skin enables them to discriminate between three-dimensional objects. While we know the electric image parameters used to encode numerous object properties, we dont understand how these images encode object shape. We hypothesise that movement-induced modulations (MIMs) evoked by body movements might be involved in shape discrimination during active electrolocation. To test this, we trained fish to complete a shape discrimination task in a two-alternative forced-choice setup, and then manipulated the space available to individuals for scanning movements to see if this led to a change in their discrimination performance. We found that if enough space was available, fish were very good at discriminating objects of different shapes. However, performance decreased strongly when the space was reduced so that scanning movements were impaired. Our study demonstrates the importance of body movements for gaining complex environmental information such as object shape through active electrolocation. Movement can enhance perception by allowing the extraction of certain kinds of information. Similar observations have been made in other animals using different senses, suggesting that the core principles of sensory-motor integration might be valid for various sensory modalities.