Exclusion of predatory fish on a coral reef: The anticipation, pre-emption and evaluation of some caging artefacts

The use of cages as a technique to understand the dynamics of reef fish has often been treated with suspicion. Cages may cause artefacts, which cage controls may not always detect, possibly confounding the result (exclusion of predators) and secondary outcomes associated with the cage (such as altered algal growth). This study was concerned with the minimisation and evaluation of artefacts associated with cages used to exclude large predatory fish on a coral reef. The manipulations were specifically designed to avoid major artefacts that have previously confounded the effects of excluding fish from hard substrata: (i) abnormally high algal growth due to exclusion of large herbivorous fishes; (ii) abnormally high sedimentation due to changes in hydrodynamics; (iii) confusing the experimental effects with local scale patchiness of predators. For cases in which major artefacts could not be minimised or eliminated from the experimental design, controls were augmented with additional data to determine the importance of these artefacts. Special attention was given to assessing the importance of the differential attraction of the treatments to fish. Observational data indicated that the abundance of large predatory fish was unaffected by partial cages (control) and hence predation pressure between open plots and partial cages were commensurable. Observational data also indicated that the encounter rates of prey with predators that could pass through the mesh was even among all treatments. I concluded that caging artefacts were minimal and that experimental effects of excluding large predatory fish could be interpreted realistically. These results demonstrate that caging is not an inherently flawed technique, but its use requires careful planning and evaluation of artefacts that could confound interpretation of the treatments. (C) 1997 Elsevier Science B.V.