Evaluating the effect of soak time on bottomfish abundance and length data from stereo-video surveys

Abstract

Baited stereo-camera surveys of fish assemblages provide conservative estimates of abundance and length-frequency distributions. While underwater camera systems have numerous advantages over traditional fishing and diver surveys, limitations in sampling capacity, data processing time, and resultant data still exist. Previous studies have shown that shorter camera soak times can increase sampling efficiency and reduce per-sample data processing time without affecting overall data quality. Using data from stereo-video surveys of bottomfish in the main Hawaiian Islands, this study evaluates the effect of camera soak time on relative abundance metrics, fish length data, sampling efficiency, and power to detect differences in relative abundance and fish lengths. A soak time of 15min was found to be the shortest duration able to capture bottomfish abundance and length metrics while 30min generated data that did not significantly differ from the standard 40-min soak time. These shorter soak times allow for better survey efficiency and improved cost–benefit through increased levels of field sampling and reductions in video-processing time, while maintaining the power to detect differences in bottomfish relative abundance and lengths. The main drawback to shortening soak time was the concurrent reduction in the number of length measurements collected per species. An increased sample yield can alleviate this effect but only for bottomfish with a higher frequency of occurrence. Species-specific patterns in abundance were apparent in this study suggesting a strong influence of fish behavior on stereo-video abundance metrics. While a soak time of 15 to 30min was found to be sufficient for effectively sampling bottomfish, the cost–benefit of employing a given soak time in future stereo-video surveys should be assessed based on the target species and survey goals.