Direct versus indirect methods of quantifying herbivore grazing impact on a coral reef

Two methods were used to assess the grazing impact of roving herbivorous fishes across a coral reef depth gradient within Pioneer Bay, Orpheus Island, Great Barrier Reef. The first technique employed was a method traditionally used to quantify herbivory on coral reefs via the (indirect) inference of herbivore impact from biomass estimates and reported feeding rates. The second method (one of a range of direct approaches) used remote underwater video cameras to film the daily feeding activity of roving herbivores in the absence of divers. Both techniques recorded similar patterns and relative levels of herbivore biomass across five reef zones at the study site. Indirect estimates of the grazing impact across the reef depth gradient of the three predominant species of herbivore broadly coincided with levels quantified directly by remote underwater video, indicating that, to a large extent, presence does correspond to function. However, the video data suggested that, for individual species in particular reef zones, the absolute level of impact may be less than that inferred from presence. In the case of the parrotfish Scarus rivulatus, the video recordings suggested that, at the reef crest, an average of 52% (±18 SE) of each m² area of reef would be grazed each month, compared with an area of 109% (±41 SE) suggested by inferring grazing activity from presence alone. Potential biases associated with remote video recorders may explain some of the discrepancy between values. Overall, the results suggest that, for some fish groups, the indirect method of inferring function from presence can provide a good indication of relative levels of herbivore impact across a coral reef. Electronic supplementary material The online version of this article (doi:) contains supplementary material, which is available to authorized users.