Biologists from the University of Victoria discovered that even closely related fish species make unique and distinctive sounds, and they determined that it’s possible to differentiate between the sounds of different species.
Using passive acoustics, researchers identified unique sounds for eight Vancouver Island fish species. They then developed a machine learning model that can predict which sounds belong to which species with 88 per cent accuracy.
“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. 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,” said Darienne Lancaster, UVic PhD student and lead researcher, in a news release.
For example, the black rockfish makes a long, growling sound similar to a frog croak, and the quillback rockfish makes a series of short knocks and grunts.
“It has been exciting to see how many different species of fish make sounds and the behaviours that go along with these calls,” noted Lancaster.
“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. Other times, fish, like copper rockfish, will repeatedly make knocking sounds as they chase prey along the ocean floor.”
Lancaster used a technique called passive acoustic monitoring to identify the fish sounds, where she collects underwater audio and video using a sound localization array designed by former UVic PhD student and project collaborator Xavier Mouy, and then used sound characteristics to identify differences in species calls.
Her AI machine learning model used a set of 47 different sound features, such as duration and frequency, to detect small differences in each species’ sounds that can be used to tell them apart. The model used these small differences in sound features to group species calls together.
The techniques that Lancaster developed can be adapted by scientists all over the world to decipher other fish calls.
The research was funded by the Natural Sciences and Engineering Research Council of Canada and Fisheries and Oceans Canada.