A large scale BACI experiment to test the effects of prawn trawling on seabed biota in a closed area of the Great Barrier Reef Marine Park, Australia

A multiple Before-After-Control-Impact experiment was conducted to estimate the large scale effects of a single coverage by prawn trawlers on an offshore inter-reef area in the northern Great Barrier Reef that was closed to fishing. Prospective power and cost-benefit analyses facilitated the design and optimization of the experiment, the effect size of which was guided by prior publications at the time that had indicated impacts of 10–100×. The optimal design, given the number of factors to be tested and the constraint of available ship time (100 days) and resources needed to process samples, was for 12 control and 12 impact plots each measuring ∼2.8 km × 1.2 km. The final design specification was capable of detecting an average −80% effect size (5×). The experiment was divided over two seasons, wet and dry. During the experimental trawl impact phase, a total of 32 t of benthic biota was removed from six shallow plots (15–25 m depth) and 6 t from six deep plots (30–50 m depth). Sampling before and 6 months after impact was conducted by an epibenthic dredge, a prawn trawl and a fish trawl. Analysis of means before and after impact, relative to controls, showed very few significant differences. This indicated that the impact of a single prawn trawl was less than the design specification. The lower than expected impact, compared to previous studies, may have been partly because this study included all benthic groups, not just the most sensitive as earlier studies had focussed on. Comparison of catch rates from the prawn trawl and the dredge indicated the overall impact on total biomass was around −3% but ranged from close to 0% for some species to around −20% for sensitive sessile species. A review of effect sizes in 30 other recent trawl experiments indicated that this result was not exceptional; i.e. the reported effects of single trawls generally were not large, and prawn trawls appeared to have smaller effect sizes than fish trawls, beam trawls and scallop dredges—also, several studies reported recovery within ca. 6 months. Analysis of fishery effort data indicated that this result was appropriate for the majority of trawl fishing grounds, where effort is sparse and infrequent. Many published experiments had confounded designs and most had not used a priori power analyses or pre-specified effect sizes. Nevertheless, recurrent trawling can be expected to have cumulative impacts on benthos, as has been demonstrated by repeat-trawl depletion experiments. On the other hand, the spatial extent of fishing grounds trawled as intensively as these depletion experiments is quite limited.