Detection of protection benefits for predatory fishes depends on census methodology

Abstract

Marine protected areas (MPAs) are used as fisheries management and conservation tools. Well-enforced no-take zones allow the rebuilding of natural populations of exploited species; however, there is still controversy on the role of buffer zones. The effectiveness of MPAs could be underestimated, as fish population assessments depend largely on traditional methodologies that have difficulties in detecting predatory fish because of their low abundances, their patchy distribution, and their reaction to the presence of divers. The performance of different census methods was compared in assessing the protection benefits for large predatory fishes under different protection levels (i.e. no-take and buffer zones) in five Mediterranean MPAs. Specifically, conventional strip transects (CSTs, 50 × 5 m2) and tracked roaming transects combined with distance sampling (TRT + DS, variable lengths) were compared, including a series of TRT-derived estimators with variable transect lengths and fixed widths of 20, 10, and 6 m (TRT20, TRT10, and TRT6, respectively). Additionally, the effectiveness of the MPAs studied and protection levels for conserving large predatory species was evaluated. Transects covering larger areas (i.e. TRT + DS and TRT20) allowed the detection of a greater number of species and yielded more accurate estimates of density and biomass than transects of narrower fixed widths, particularly the CSTs, which were associated with the lowest richness detection capability, accuracy, and precision. On average, both no-take zones and buffer zones appeared effective for the conservation of predatory fishes, indicating that multiple protection areas were ecologically effective. Differences between MPAs were also observed, however, probably arising from both local environmental and management factors. We suggest the implementation of methodologies with larger transects for the study of large predatory fish, combined with CSTs for the rest of the fish community, in order to avoid biases in predatory population assessments, which are key indicators of MPA effectiveness.