Fin spotting: efficacy of manual and video-based visual assessments of reef fish swimming behaviour

• Both manual & video-based observations detect key trends in fish swimming behaviour
• Underwater video was most effective at recording subtle and/or rapid fin use
• Manual observations were best at detecting rare activities (e.g., chasing)
• Video requires greater sampling effort than manual census for the same precision

Here we report on the efficacy of manual and video-based underwater visual census (UVC) for recording the swimming behaviour of fishes within complex coral reef habitats. Focusing upon four common fish species from the Great Barrier Reef (Acanthurus nigrofuscus, Chaetodon lunulatus, Siganus doliatus and Sufflamen chrysopterus), we compared manual recordings by divers against post-field analysis of underwater video of the same species in the ability to detect the different fin types used by fishes while they foraged over the reef. Both video and manual UVC methods detected similar trends in the number of fins and the proportional use of each fin type by reef fishes engaged in different activities (e.g., travelling, searching, feeding). While estimates of fin use from the two methods were largely within 15% of each other, video UVC tended to record higher proportions of fin use within each species due to a better detection of rare and/or rapid fin movements. Such differences were most extreme for the balistiform-swimming S. chrysopterus, where video detected a significantly higher proportion of fins used by this species. Notably, the sampling effort required to improve precision and gain sufficient independent replicates was much higher for video-based versus manual UVC. Our findings reflect the relative strengths of each UVC method for studies of fish behavioural ecology in the wild, with the various benefits of each method likely to align with different research questions and fish species of interest, within the logistical constraints of funding and field safety.