Using unbaited remote video cameras to assess the effect of marine reserves on fish populations

Abstract

Marine reserves are being increasingly used as conservation and fisheries tools, and thus need to be monitored to determine whether their objectives are met. Underwater video is a popular method for monitoring fish as it avoids many of the shortcomings associated with underwater visual censuses (UVC), the biggest one being that fish may either be repelled by, or attracted to, the diver doing the counts. Baited underwater video (BUV) is commonly used to monitor marine reserves and “control” sites because it is thought to avoid the influence of the diver on fish detectability, and provides statistically powerful results. However, there is growing concern that BUV, like UVC, is prone to variability in fish behaviour towards the presence of a foreign object in the water. The present study compared relative density of snapper Pagrus auratus measured using UVC, BUV, and an unbaited inconspicuous underwater video system (UUV) to test whether UUV was potentially better than the other two methods. The UUV required that fish lengths be estimated from a single camera image, to avoid the expense and complication of stereo video and other sizing methods. I found strong relationships between relative eye size and body length for snapper and five other common rocky reef fish species in northeastern New Zealand. The comparison between methods entailed sampling areas inside and outside two marine reserves in northeastern New Zealand, during each of two seasons. Relative densities (expressed as reserve:non-­‐reserve ratios) were highest for UVC, intermediate for BUV, and lowest for UUV, i.e., the ratios declined as the intrusiveness of the survey method decreased. Differences in reserve:non-­‐reserve (R:NR) ratios among methods were greatest for legal-­‐sized (≥270mm fork length) snapper, leading to the conclusion that UVC and BUV were underestimating abundances of larger snapper outside the reserve. This would mean that estimates of R:NR ratios made by BUV have been exaggerated in the past. The method of size estimation developed in this study in conjunction with the UUV system could provide a more simple and cost-­‐effective method of monitoring reef fish species than current methods.