Does breathing apparatus affect fish counts and observations? A comparison at three New Zealand fished and protected areas

Across three areas, open-circuit scuba (OC) and rebreather (RB) surveys produced similar results for the density and size distribution of fish species inside and outside marine reserves. At Tonga Island, more Notolabrus celidotus were counted with OC than with RB, independently of reserve status [log-scale response ratio of OC/RB (RR) = 0.7]. At Long Island, differences in abundance of Parapercis colias between sampling methods were small at reserve sites (RR = −0.1), but more were counted with scuba than with RBs at fished sites (RR = 0.5). RRs for Pagrus auratus were −1.0 in fished areas and 0.3 in the reserve at Leigh. We also sampled each site using a baited video system (BUV) to establish whether diver-transects sampled the full size range of target species. Most fish in BUV views were Parapercis colias at Long Island (97%), and Pagrus auratus at Leigh (77%). Size structures of Parapercis colias were similar among all three sampling methods within reserve and fished areas at Long Island (max. chi-squared distance = 0.11). BUV samples for Pagrus auratus at Leigh did not detect a prominent juvenile size class observed by divers, but size-frequency distributions of OC, RB, and BUV corresponded at sizes beyond 15 cm TL (max. chi-squared distance = 0.08). To investigate the effects of diver sound on fish behaviour at Long Island, we also compared fish activity when divers with RBs or scuba were present, when the sound of each breathing apparatus was replayed underwater, when no divers were present and no sound was replayed, and when bait was provided, within the reserve only. The lowest number of fish visits to the focal area (mean of 3.0 per 10 min) for Parapercis colias occurred with RB divers present. Maximum abundances of Parapercis colias in all speaker treatments averaged 4.1 per 10 min, whereas with scuba divers present maximum abundances were 5.7, and with baits the average was 38.0 per 10 min.