MEPS prepress abstract  -  DOI: https://doi.org/10.3354/meps12395

Do traditional Baited Remote Underwater Video Surveys (BRUVs) undercount sharks at high densities? A comparison using conventional and newly developed full-spherical camera technologies

James P. Kilfoil*, Aaron J. Wirsing, Matthew D. Campbell, Jeremy J. Kiszka, Kirk R. Gastrich, Michael R. Heithaus, Yuying Zhang, Mark E. Bond

*Email: jkilf001@fiu.edu

ABSTRACT: Video surveys are an essential tool for monitoring marine communities. Their use to study elasmobranch populations has dramatically increased over the last decade. However, the restricted field-of-view of traditional cameras in these surveys may bias abundance estimates in a number of ways, including saturation at high densities, and low detection probability for rare or cryptic species. This study investigated these potential biases using newly developed full-spherical (FS) camera technology. A comparison of 35 Baited Remote Underwater Video Surveys (BRUVs), using both FS and traditional cameras, was conducted from July-August 2016 in shallow waters (0.4-8.5m) of Tetiaroa, French Polynesia. Both blacktip reef (Carcharhinus melanopterus) and sicklefin lemon sharks (Negaprion acutidens) were quantified from traditional cameras using MaxN and MeanCount methods. These estimates were then regressed against FS cameras counts, which were assumed to more accurately represent site abundance, to test for gear saturation. Detection probabilities of the traditional and FS cameras were assessed using a Bayesian binomial model, with uninformed-uniform priors. Results indicated a significant effect of gear saturation for standard BRUVs as counts on FS cameras increased, regardless of the metric used. Furthermore, traditional cameras had a significantly lower detection probability (69.88±0.008%mean ±2SD) than FS cameras (81.20±0.007%). Our findings show that traditional cameras are unlikely to adequately discriminate differences in shark relative abundance at high densities. Therefore, standard BRUV techniques that use restricted FOV cameras are likely limited in their ability to provide accurate information to managers once populations have reached particular thresholds of abundance.