Abstract

Fisheries independent research strives for new technology that can help remotely and unobtrusively quantify fish biomass. Some large fish species, such as groupers vocalize during reproductive behaviors. Fish sounds provide an innovative approach to assess fish presence and numbers during reproductive events. However, large datasets make the detection process by a human ear very tedious and lengthy. We have developed an algorithm based on machine learning and voice recognition methods to identify and classify fish sounds. This algorithm currently operates on a SV3 Liquid Robotics wave glider, which has been fitted to accommodate a passive listening device. Fish sounds detection and classification results, and location along with environmental data are transmitted in real-time enabling verification of the detections with divers or other in-situ methods. Results from deployments in the US Virgin Islands and Puerto-Rico confirmed the location of known aggregations and their specific species and revealed the presence of potential new ones.

PDF Preview