A recent study published in PLOS ONE examines the soundscape of a reef undergoing restoration at Cousin Island Special Reserve in Seychelles. Led by Emily Croasdale, in collaboration with Nature Seychelles, the research compared three reef sites: one being restored, one healthy, and one degraded. Recordings were captured in November 2023. To analyse the recordings, the team used two different methods: they manually listened to and counted the sounds made by fish and also harnessed artificial intelligence and machine learning to identify broader patterns within the soundscape.
Manual analysis indicated that fish sounds at the restored site were similar to those of the healthy reef, suggesting a positive response to restoration. However, machine learning placed the restored reef closer to the degraded one, highlighting that recovery rates of different reef components vary. These findings emphasize the importance of using multiple methods to assess restoration success.
The open access paper is available here: https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0347307
