UVic Biologists Use AI to Decode Distinctive Fish Sounds

Researchers at the University of Victoria have made significant strides in understanding aquatic communication by using artificial intelligence to identify and differentiate the unique sounds made by various fish species. Their findings reveal that closely related species produce distinct sounds, which can be classified with remarkable accuracy.

The study focused on eight fish species native to Vancouver Island. By employing passive acoustic monitoring, the team collected underwater audio and video recordings. They developed a machine learning model capable of predicting which sounds belong to which species with an impressive 88 percent accuracy.

Darienne Lancaster, a PhD student and the lead researcher, explained the implications of this research. “We knew previously that many fish were making sounds in the wild, but we didn’t know which sounds belonged to which species, or if it was possible to tell these sounds apart,” she stated in a news release. “Now, just as we use bird song to identify specific bird species in the wild, we can also listen to fish sounds to identify specific fish species.”

The research highlighted notable examples of fish sounds. For instance, the black rockfish produces a long, growling noise akin to a frog’s croak, while the quillback rockfish emits a series of short knocks and grunts. Lancaster noted the excitement of discovering various fish species that communicate through sound, emphasizing that some species use specific calls in response to threats. “Some fish, like the quillback rockfish, make rapid grunting sounds when they’re being chased by other fish, so it’s likely a defensive mechanism,” she added.

Innovative Techniques for Acoustic Monitoring

To carry out this research, Lancaster used a sound localization array designed by Xavier Mouy, a former PhD student and collaborator. The process involved analyzing 47 different sound features, including duration and frequency, to detect subtle differences in species calls. These differences were crucial for grouping the sounds accurately.

The methodologies developed by Lancaster not only contribute to the understanding of fish communication but can also be adapted by scientists globally to decipher other fish calls. This research underscores the importance of sound in the underwater ecosystem and the potential for further studies in marine biology.

The project received funding from the Natural Sciences and Engineering Research Council of Canada and Fisheries and Oceans Canada, highlighting its significance in advancing scientific knowledge about aquatic life.

In summary, the University of Victoria’s innovative approach to studying fish sounds through AI represents a promising direction for future research, offering new insights into the complex lives of marine species.