Visual coding for action violates fundamental psychophysical principles

(Plectropomus spp.), are specialized predators of fishes. The evidence that any exploited fishes are significant predators of A. planci is circumstantial or anecdotal. Fishing commonly removes large predators and has led to trophic cascades involving multiple trophic levels in other marine systems (for example [9]). A study comparing near-shore fish communities in no-take and in fished areas on the GBR [10] found that numbers of coral trout were higher in no-take areas, while a majority of likely prey species were less abundant, including the common benthic-feeding wrasse, Thalassoma lunare. A plausible positive link between commercially exploited fishes and predation on A. planci could involve higher numbers of large piscivores in no-take areas reducing densities of benthic carnivorous fishes such as wrasses, so causing ecological release of invertebrates that prey on very small A. planci. A. planci juveniles live hidden in rubble for 16–19 months after settlement [1] and have very high disappearance rates that are not due to emigration [11]. This implies that the invertebrate faunas in the rubble habitat of juvenile A. planci should also differ predictably between no-take and fished reefs. The GBRMP was re-zoned in mid-2004, increasing the no-take zones from 4.5% to 33% of the area of the park [7]. Whatever the underlying mechanism, this study suggests that this increase should reduce the overall impact of future waves of A. planci outbreaks. That effect may be amplified if fewer reefs with starfish outbreaks mean less effective propagation of outbreaks from reef to reef through the central GBR. More generally, the geographic range of A. planci includes the most biodiverse [12] as well as some of the most threatened reefs [13] on earth; this study provides an additional argument for establishment of effective MPAs across the range [14], as refuges from exploitation and other threats and as sources for recolonisation of damaged reefs to increase ecological resilience. Acknowledgments This project was funded by the Australian Institute of Marine Science and the CRC Reef Research Centre. I thank GBRMPA library staff for their help, L. Castell and A. Dolman and two referees for helpful comments on the manuscript.