Quantitative Evaluation of the Efficiency of Visual Fish Surveys As a Function of Near-Seafloor Water Clarity and Ambient Light Levels

We used a stereo-video lander to study how varying conditions of seafloor water clarity and ambient light altered the efficiency of visual surveys.  We deployed the stereo lander into fish schools at different depths and calculated mean range of detection using the stereo calibration parameters.  Ambient light was measured using a Wildlife Computers TDRMK9 and water clarity was measured using a WetLabs EcoBB scattering meter.  At the deep-shelf reef, light was consistently very low and water clarity consistently high resulting in a mean range of fish detection of about 4.1 m.  At the mid-shelf reef, light varied with depth, but water clarity was consistently high, resulting in a mean detection range from 3.3-5.6 m. At the nearshore reef, light was consistently high, but water clarity was lower and highly variable, reducing the mean detection  range from a high of 3.4 m down to just 1.6 m under turbid conditions. Across sites, mean range of detection was negatively correlated with the backscattering index (P=0.0003), but uncorrelated with variation in light (P>0.05). The combination of stereo-video and in situ measurement of backscattering has the potential to improve quantification of visual survey data.