Fish have long been known to grunt, knock and growl beneath the surface, but until now, much of this underwater soundscape has been something of a mystery. New research from the University of Victoria (UVic), however, has found that even closely related fish species produce distinctive sounds that can be reliably told apart – opening up new possibilities for identifying fish by sound alone.

Using passive acoustic monitoring in coastal waters around Vancouver Island, UVic researchers recorded and analysed the natural calls of eight common fish species. By combining these recordings with machine learning, the team was able to correctly identify which species produced which sound with up to 88% accuracy.

The findings point to a powerful, non-invasive tool for monitoring marine life – one that could transform how scientists study and protect fish populations.

“We’ve known for some time that many fish are vocal in the wild,” said Darienne Lancaster, a PhD candidate in biology at UVic and lead author of the study. “What we didn’t know was which sounds belonged to which species, or whether those sounds were distinct enough to tell apart. Now, just as bird song helps us identify birds on land, we can listen to fish sounds to identify fish underwater.”

The research, published in the Journal of Fish Biology, forms part of the broader Fish Sounds project led by the Juanes Lab at UVic.

Although fish bioacoustics has been studied for decades, most previous work relied on laboratory recordings. Whether fish produced unique, identifiable sounds in their natural environments had never been rigorously tested.

Lancaster’s fieldwork changed that. She identified species-specific sounds for eight fish commonly found along the coast of British Columbia: black rockfish, quillback rockfish, copper rockfish, lingcod, canary rockfish, vermilion rockfish, kelp greenling and pile perch. For canary and vermilion rockfish, these were the first documented sounds ever recorded – either in the lab or in the wild.

The recordings reveal a surprisingly rich acoustic world. Quillback rockfish emit rapid grunts when chased, likely as a defensive response, while copper rockfish produce repeated knocking sounds as they pursue prey along the seabed. Black rockfish, meanwhile, generate long, low growls reminiscent of a frog’s croak.

“It’s been fascinating to see how many species are vocal, and how closely these sounds are tied to behaviour,” Lancaster said.

To capture these sounds, Lancaster used passive acoustic monitoring – an approach that involves listening rather than emitting sound. Underwater audio and video were collected using a sound localisation array designed by former UVic PhD student and collaborator Xavier Mouy, allowing researchers to link specific sounds to individual fish.

Lancaster then trained a machine learning model using 47 different acoustic features, including sound duration and frequency. These subtle differences allowed the model to group calls by species, revealing consistent acoustic signatures beneath the waves.

According to Francis Juanes, professor of biology at UVic and principal investigator on the project, the implications are far-reaching.

“Passive acoustics offers a new way to estimate population sizes, monitor activity and assess the health of fish populations without disturbing them,” he said. “For conservationists and fisheries managers, that’s a significant step forward—especially for vulnerable or hard-to-study species.”

The techniques developed through this research can be applied globally, helping scientists decode the sounds of other fish species and ecosystems. The work feeds into the Juanes Lab’s wider Fish Sounds project, which includes FishSounds.net, an international database of peer-reviewed fish sound research; The Codcast, a podcast exploring the discovery of the Arctic cod’s grunt; a custom-built sound localisation array; and Fish Sounds Educate, a free programme designed to boost ocean literacy and inspire future conservation leaders across Canada.

Funded by the Natural Sciences and Engineering Research Council of Canada and Fisheries and Oceans Canada, the research also supports the United Nations Sustainable Development Goal 14: Life Below Water.

As scientists continue to tune in to the ocean’s hidden soundscape, studies like this suggest that listening carefully may be one of the most powerful ways to understand – and protect – life beneath the surface.

Click here for more from the Oceanographic Newsroom.