Table S1: Nearest neighbour analysis (index R)-variance to mean (VMR) approximations and corresponding nearest neighbour distance for 5, 15, 25, and 40 A. planci ha-1. Entries are given by R (VMR ±SEM) where means are computed over 5000 simulations and approximate an analytically computed R value (by Equation 1) with a VMR over a region of 4 ha in 10 m by 10 m quadrats. Only R>0 (0-1: random to highly aggregated) is considered as R<0 (regular spatial distributions) would cause the aggregation model to simulate individuals outside the model region.

No abstract

Management of coral predators, corallivores, is recommended to improve coral cover on tropical coral reefs under projected increasing levels of accumulated thermal stress, but whether corallivore management can improve coral cover, which is necessary for large-scale operationalisation, remains equivocal. Here, using a multispecies ecosystem model, we investigate intensive management of an invertebrate corallivore, the Crown-of-Thorns Starfish (Acanthaster cf. solaris), and show that culling could improve coral cover at sub-reef spatial scales, but efficacy varied substantially within and among reefs. Simulated thermal stress events attenuated management-derived coral cover improvements and was dependent on the level of accumulated thermal stress, the thermal sensitivity of coral communities and the rate of corallivore recruitment at fine spatial scales. Corallivore management was most effective when accumulated thermal stress was low, coral communities were less sensitive to heat stress and in areas of high corallivore recruitment success. Our analysis informs how to manage a pest species to promote coral cover under future thermal stress events.

No abstract