Reef fish assemblages: a re-evaluation using enclosed rotenone stations

Abstract

The reef fish assemblage at Orpheus Island, Great Barrier Reef, was examined using visual censuses and the ichthyocide rotenone. Small 3.5 m2 quantitative rotenone samples, using a fine-mesh net to enclose the site, were compared with visual-point censuses (prior to the placement of the net), random-point censuses and strip censuses. Furthermore, the fishes collected inside and outside the net were examined to determine the relative efficiency of enclosed versus open rotenone samples. Rotenone samples comprised 128 species in 28 families. Of these, only 49% of the species overlapped between the enclosed (inside net) and open (outside net) samples. Only 17.7% of the species in the enclosed rotenone sample were seen prior to collection. Rotenone samples reveal that visual censuses underestimated the abundance of small taxa. Overall, rotenone samples increased the estimated number of species by 40.4% and the total fish abundance by 50.1% (75.3% for fishes <50 mm). Biomass increased by less than 1.0%. However, through relatively high energetic demands, high turnover rates and their role as potential prey, small fish taxa may play a greater role in reef processes than previously assumed. The limits of visual censusing techniques are highlighted, emphasising the value of small, enclosed, intensive rotenone samples for providing reliable quantitative samples of small taxa.