Towards standardised reef fish monitoring: an assessment of stereo-video techniques to sample shallow and deep reef fish assemblages

• Main Author: Halse, Sarah Jane
• Format: Master Thesis
• Language: English
• Published: Rhodes University 2017
• Subjects: Reef fishes > South Africa > Agulhas; Cape (Cape); Reef fishes > Ecology > South Africa > Agulhas; Marine biodiversity > South Africa > Agulhas; Biotic communities > South Africa > Agulhas; Sparidae > South Africa > Agulhas; Reef fishes > Size > South Africa > Agulhas; Underwater videography in wildlife monitoring > South Africa > Tsitsikamma; Baited remote underwater stereo-video systems (stereo-BRUVs); Crassostrea gigas
• Online Access: http://hdl.handle.net/10962/4635; http://vital.seals.ac.za:8080/vital/access/manager/Repository/vital:20702

Baited remote underwater stereo-video systems (stereo-BRUVs) were developed to determine fish abundance and size structure in a more unbiased, and relatively non-invasive manner across a broader range of depths and habitats than conventional sampling methods achieve. These characteristics make stereo-BRUVs particularly useful for research inside marine protected areas (MPAs) where researchers are required to minimize disturbance to the protected environment. Although stereo-BRUVs have been widely used around the world, they are novel to South Africa and standardised protocols for their application are required. As such, this study aimed to address i) the effect of different bait types, namely pilchard (Sardinops sagax), squid (Loligo reynaudi) and bivalves (Crassostrea gigas and Perna perna) on the observed reef fish assemblage structure and ii) the effect of artificial illumination (lighting) on the observed reef fish assemblage structure sampled with stereo-BRUVs. One key component of the stereo-BRUVs methodology is the application of bait to attract fish into the field of view. Globally, pilchard is the most often used bait type for stereo- BRUVs, however, its suitability has never been tested comparatively against other bait types for South African conditions. Significant differences in fish abundance and assemblage structure were recorded between stereo-BRUVs deployments when applying the different bait types. The bivalve baits, oyster and mussel, were typically poor when sampling the reef fish assemblages of the region. Pilchard and squid were sampled in similar communities and were able to attract the majority of the reef fish from the region, suggesting they are suitable bait types for stereo-BRUVs in the Agulhas ecoregion. However, a power analysis indicated that roughly twice the amount of samples are required to effectively sample the same abundances of Chrysoblephus laticeps (roman) and Cheilodactylidae (fingerfins) species when sampling with squid as opposed to pilchard. Both bait types can be locally sourced, however, squid is considerably more expensive than pilchard. Considering the difference in cost, together with the fact that twice as much squid bait is required, squid is not a cost-effective option relative to pilchard. As such this study recommends that pilchard is the most appropriate bait for the Agulhas ecoregion of South Africa. Due to low light levels, artificial lighting is required when conducting remote video research in the sub-photic zone. Blue light has a shorter wavelength than other colour lights and attenuates slower through water. In addition, blue lighting has no recorded effect on fish behaviour and has proven to be a successful light colour for underwater sampling. The use of lighting is unavoidable for sub-photic stereo-BRUVs sampling and is recommended for all stereo-BRUVs sampling for comparability of data from environments characterised by different light levels. A study was conducted to test the necessity of artificial lighting when sampling from environments characterised by ambient light levels adequate for stereo- BRUVs sampling, a comparative to measure the effect of light on the fish assemblage structure. Of the more commonly detected fish species (those observed in >50% of the samples), most were seen at similar abundances in samples using lighting and in those without lighting. No significant differences were observed in fish length data between samples collected with and samples collected without lighting. This suggests that standardised stereo-BRUVs sampling across photic zones on the continental shelf of South Africa can be collected without blue lighting when ambient light levels are sufficient to see the survey area.