Search

This first survey of fish in the Betty's Bay Marine Protected Area (MPA), on the south coast of South Africa, was conducted using baited remote underwater video systems (BRUVs). A total of 58 deployments recorded 42 species in 20 km2, including reef, kelp and sand habitats in protected and exploited zones, at between 5 and 40 m depth. Chondrichthyans accounted for 28% of diversity.Teleost diversity was dominated by Sparidae, Cheilodactylidae, Sciaenidae and Ariidae. Diversity (H′) was highest in kelp and lowest over sand. Species composition differed among habitat and depths, but protection had no effect. Among four commercial species, onlyPachymetopon blochiiresponded positively to protection. The apparent failure of protection may attest to poor compliance, but an investigation into fish size might show an effect. Many species were detected at the western extreme of their range. Diversity in Betty's Bay was predictably lower than in the more eastward Stilbaai MPA, but also lower than in the westward Table Mountain National Park MPA. Fish diversity did not follow a linear increase eastwards from Cape Point. Betty's Bay includes the most easterly protected kelp forests and contains seven species not recorded in the other two areas, and is therefore an important element in the MPA network. [ABSTRACT FROM AUTHOR]

Global declines in chondrichthyan populations necessitate evidence‑based planning for their conservation. False Bay, in south‑west South Africa, has a high diversity of marine biota, with Atlantic and Indian Ocean influences, but its chondrichthyan fauna is poorly described ecologically. False Bay has a long fishing history and supports various harvesting methods, including shark‑directed fishing. This is the first baited remote underwater video system (BRUVs) survey of the relative abundance, diversity and seasonal distribution of chondrichthyans in False Bay. Nineteen species from 11 families were recorded across 185 sites at between 4 and 49 m depth. Diversity was greatest in summer, on reefs and in shallow water. Endemic scyliorhinids, especially Haploblepharus edwardsii, dominated everywhere, in summer and winter. Habitat and season were key predictors of species composition, and a significant interaction between depth and season indicated migration patterns for Mustelus mustelus. By avoiding the selectivity associated with catch records, BRUVs data are superior to fishery‑dependent catch‑per‑unit‑effort data to monitor chondrichthyan diversity and abundance. Baited remote underwater video systems access greater depths and more habitats than previously surveyed in False Bay and avoid taxonomic lumping typical of catch records in the region, making them more efficient than other methods of surveying chondrichthyan diversity. [ABSTRACT FROM AUTHOR]

Marine protected areas (MPAs) play an important role in coastal conservation, but there is presently no uniformly applied methodology for monitoring the efficacy of coastal fish protection. Whereas underwater visual census and controlled angling surveys have been used, their skilled-labour requirements and environmental impact have prevented their nation-wide application in South Africa. We used a baited remote underwater video system (BRUVs) to provide the first survey of reef fishes in the Stilbaai MPA (SMPA), and assessed the deployment time and sample size requirements of BRUVs for reef fish monitoring. Thirty-eight species, including 13 chondrichthyans, were recorded in one-hour deployments across 29 sites in the 11.3 km2no-take zone of the SMPA, in depths ranging from 5 to 41 m. Bait was limited to sardineSardinops sagaxhomogenate, but the species recorded by BRUVs comprised several feeding guilds, including herbivores. Optimal deployment time was 49 min, but 60 min is recommended as a conservative and practical standard. Over a five-year period, annual BRUVs sample sizes required to detect increases in abundance of 0.1 y−1with 80% power would be 28 deployments for abundant species and 129 for rare species. Detection of decreases of equivalent magnitude will require more samples. [ABSTRACT FROM PUBLISHER]