Stationary underwater cameras assess more efficiently clear‐water mangrove fish assemblages: A comparison of non‐extractive techniques

Abstract

AbstractResearch in aquatic ecology requires adequate sampling designs if the underlying processes that govern population and community dynamics are to be understood. Adequate sampling design is particularly important for organisms inhabiting environments such as mangroves. Thus, collecting quantitative ecological data remains a major challenge. Here, we analysed the structural and temporal dynamics of a tidal cycle and mangrove fish assemblages at 15 sites along the Brazilian central coast. Specifically, we used two non‐extractive sampling techniques, which included stationary underwater visual census (s‐UVC) and stationary underwater wide‐camera census (s‐UWC). Our results showed that by reducing potential observer effects, the s‐UWC protocol maximized fish counts, yielding more accurate and precise estimates. In contrast, although the s‐UVC technique was a useful approach, this technique did not yield precise density and richness estimates. Furthermore, the negative effect associated with the presence of an observer (i.e., diver) was a significant constraint to the observer‐based technique. Furthermore, the methods were more similar for the high tide samples and more divergent for the flood tide and ebb tide samples. These pronounced divergences among tidal periods allowed the application of a temporal mixed‐sampling protocol (i.e., unbalanced temporal samples) based on s‐UWC without the loss of statistical power. Methodological gains also were felt with a reduction in the costs associated with the presence of a trained observer in water and overcoming of constraints positively related to depth limitations. Overall, the s‐UWC protocol adequately captures species using different stages of the tide to enter the mangroves.