Background/Aims: Accurate monitoring and research data are essential to understanding natural and human-driven changes in fish assemblages and to evaluate management strategies for mitigating anthropogenic pressures. Underwater visual census, along transects, is one of the most common survey methods used to collect fish data, however, video-based techniques such as diver and remotely operated video are becoming popular alternatives. We assess how commonly used indicators of fish condition collected by different transect-based methods compare and explore how to make these metrics more compatible for synthesis of temporal and spatial assessments of fish assemblages.
Methods: We examined 10 published studies that compared fish assemblages using underwater visual census (UVC), diver operated stereo-video (stereo-DOVs) or stereo remotely operated vehicle (stereo-ROV), to synthesise knowledge on common fish metrics collected using these methods. Combined with unpublished data, we assessed comparability across these metrics using effect sizes, and developed corrections to improve comparability across methods.
Results: We demonstrate that by applying relatively simple corrections, data collected by stereo-DOVs and UVC can be made compatible for assemblage composition, community temperature index, the abundance and biomass of targeted species, total biomass and the large reef fish index (B20). More comparisons of UVC and ROV are needed to enable a comprehensive assessment of their comparability, which could be improved with the application of consistent operating procedures. Total abundance and species richness were not comparable across methods.
Conclusion: We provide the first synthesis of method comparisons for transect-based sampling of fishes and corrections that enable UVC, DOVs and ROV data to be combined across a range of fish metrics. Our findings will enable more extensive spatial and temporal assessments of fishes, whilst also increasing the confidence of research and monitoring programs to adopt new technologies.