Cost-efficient sampling of fish assemblages: comparison of baited video stations and diver video transects

Baited remote underwater stereo-video (stereo-BRUV) stations and diver operated stereo-video (stereo-DOV) transects are increasingly used to sample both tropical and temperate fish assemblages. Compared to in situ visual census methods, the use of stereo-video reduces interobserver variability, improves definition of the sample unit area, increases accuracy of fish length estimates and provides a permanent record of the assemblage that can be validated where required or independently reanalysed. Previous studies have suggested that stereo-BRUV collects representative data on both carnivorous and herbivorous species and can be more cost-efficient than diverbased survey methods. This study compares estimates of common fish assemblage metrics obtained with stereo-BRUV stations and stereo-DOV transects across 3 biogeographic regions, and uses a costoptimization procedure to compare the efficiency of these 2 methods. Stereo-BRUV stations were found to sample greater species richness and obtain greater estimates of relative biomass of generalist carnivores, but no differences occurred in the biomass of herbivores sampled by the 2 techniques. Stereo-BRUV stations generally obtained estimates of assemblage metrics with less variance, resulting in greater power to detect spatial and temporal changes in the fish assemblage metrics. Cost -benefit analyses found that stereo-BRUV was generally more time efficient than stereo-DOV transects in terms of smaller standard error around the mean of the various metrics considered. However, across the 3 biogeographic regions sampled there was considerable variation in the magnitude of these differences. Results suggest that stereo-BRUV stations are, in general, a more cost-effective method for monitoring fish assemblages than stereo-DOV transects.