No-take reserves protect coral reefs from predatory starfish

The crown-of-thorns starfish, Acanthaster planci, is a predator of corals that is a major management issue on coral reefs [1]. It occurs throughout the Indo–Pacific and shows boom–bust population dynamics with low background densities and intermittent outbreaks. Three waves of population outbreaks have affected Australia’s Great Barrier Reef (GBR) since the 1960s. The waves of outbreaks appear to start ~15 ̊S [2] and progress southward through the central GBR (Figure 1A), causing major losses of living coral on many reefs across a large area and dwarfing losses from other disturbances such as storms or coral bleaching over the same period [3]. Humans can potentially influence starfish population dynamics by exploiting predators, though evidence to date is circumstantial. Extensive surveys in the GBR Marine Park (GBRMP) show that protection from fishing affects the frequency of outbreaks: the relative frequency of outbreaks on reefs that were open to fishing was 3.75 times higher than that on no-take reefs in the midshelf region of the GBR, where most outbreaks occur, and seven times greater on open reefs if all reefs were included. Although exploited fishes are unlikely to prey on starfish directly, trophic cascades could favour invertebrates that prey on juvenile starfish. New starfish infestations arise through larval transport by the prevailing southward currents and outbreak populations die out after some years from starvation and disease [1]. Suggested (nonexclusive) causes of outbreaks [1] include greater survival of starfish larvae caused by phytoplankton blooms from nutrients in terrestrial runoff, and anthropogenic reduction of predator populations causing higher survival of juvenile and adult starfish. Two studies [4,5] have found negative relationships between outbreaks and the abundances of possible fish predators of starfish, leading to the suggestion that marine protected areas (MPAs) might reduce outbreak occurrence [6]. To address the question of whether MPAs provide protection from outbreaks of A. planci more directly, I compared the frequency of starfish outbreaks on no-take reefs and on reefs that were open to fishing on the GBR, based on results of an extensive monitoring program. The initial zoning plan for the GBRMP was fully implemented by 1989, with no-take zones covering 4.5% of the region [7]. Zoning largely followed existing uses. Where possible, significant areas for activities that did not remove natural resources were zoned ‘no-take’ and conflicting uses on individual reefs were resolved by split zoning [8]. The zoning of individual reefs was not affected by their history of starfish outbreaks (see Supplemental data available on-line with this issue). Because starfish outbreaks occur in waves, not all the reefs that were surveyed for A. planci in any year were equally likely to have outbreaks. For this reason, only reefs within the regions where outbreaks were present in each year were included in the analysis (see Supplemental data). There were fewer A. planci outbreaks in no-take zones. The majority of outbreaks occur on reefs in the mid-section of the continental shelf (Figure 1B); after allowing at least five years for zoning to take effect, surveys between mid-1994 and mid-2004 showed that proportionately fewer mid-shelf no-take reefs were affected by outbreaks of A. planci (20%), compared with mid-shelf reefs that were open to fishing (75%, Figure 1C). When all reefs were considered, the corresponding values were 8% and 57% (Figure 1D). The difference in frequency of outbreaks between no-take reefs and fished reefs is clear, but the ecological link between exploited fishes and A. planci remains uncertain. On the GBR, most outbreaks occur on mid-shelf and offshore reefs that are not accessible to most amateur fishers, while the primary target species of commercial fishers, coral trout