Monitoring of blue cod (Parapercis colias) at kapiti marine reserve, New Zealand; a comparison of survey methodologies

Abstract

Temporal and spatial patterns of relative abundance and mean total length (TL) of blue cod (Parapercis colias) were examined on the western side of kapiti Marine Reserve (KMR). Sampling was undertaken on four occasions over a nine-month period between April and December 2003. Four study sites were examined, two within and two outside KMR, and areas within each site were randomly sampled using three different survey methodologies. This study compares surveys of P. colias conducted using three methods (Underwater Visual Census (UVC), Experimental Angling (EA), and Baited Underwater Video (BUV)), as it is generally assumed that methodologies will give relatively accurate estimates, yet little comparative work has been undertaken to test this assumption to date. The ability to detect fish, standard deviation of estimates, statistical power and logistical and cost considerations were all examined. This study highlighted differences among survey methodologies, with the baited methodologies (BUV and EA) providing similar and highly significant differences among sites. Both baited methodologies had high ability to detect fish, high statistical power and small standard errors, despite a relatively low sampling effort. The BUV and the EA methodologies detected significant differences between reserve and non-reserve sites in both relative abundances and TL, indicating that the marine reserve sites were home to significantly higher relative abundances and significantly larger sized individuals than non-reserve sites. The UVC methodology differed in its estimates of P. colias relative abundance and TL than the BUV and EA methodologies. The UVC was unable to detect significant differences between reserve and non-reserve sites because power to detect fish was low, resulting in high standard errors and low statistical power. Results indicate that the traditionally used methodology in marine ecology, UVC, has limited effectiveness for sampling of P. colias at KMR. Mark and recapture techniques were also undertaken during the study to assess patterns of P. colias movement among sties at Kapiti Island (inside and outside KMR). Total distance travelled, distance vs. fish length and distance vs. time at liberty were all examined. This study revealed that fish within the reserve travelled significantly greater distances than non-reserve fish. However, the majority of recaptures/re-sighting (80% EA and 58% BUV) were made within 100 m of the tagging location. Very limited amounts of movement were examined across the reserve boundary, but 'spillover' may be occurring, even if at low frequencies, to fished areas because four individuals, originally tagged inside the reserve, were recaptured outside the reserve. No specific trends in movement were observed with regard to fish size. However, the EA highlighted the existence of a negative relationship between distance travelled and time at liberty. Recommendations and considerations for future monitoring of P. colias at KMR are discussed in light of these observed differences among methodologies. It is concluded that the most effective monitoring strategy would be a continued UVC programme used in conjunction with a BUV monitoring programme. The combined use of the two methodologies would allow a detailed archived investigation (BUV) into P. colias relative abundance and TL as well as an investigation (UVC) of non-carnivorous species not included in a baited survey such as BUV.