MEPS 176:1-10 (1999)  -  doi:10.3354/meps176001

ABSTRACT: Repeated outbreaks of crown-of-thorns starfish Acanthaster planci have been observed on the Great Barrier Reef (GBR) since the mid 1960s. Several authors have conjectured that long-term degradation of reef community structure is a possible consequence of such repeated outbreaks. In this paper we use the data from the Australian Institute of Marine Science's (AIMS) annual synoptic surveys (1985 to 1996) of the whole GBR as a statistical database with which to investigate this question. We use a simple mathematical model of transition between 3 reefs states--AO (active outbreak in progress), RE (recovering from recent outbreak), and NO (no evidence of a recent outbreak)--which we show is adequate to track the large-scale data over time. A more refined analysis, using Bayesian statistical methods, is then employed to generate a (posterior) probability distribution for a key model parameter, β, which represents the potential for non-stationary, temporal variation in the rate of recovery of reefs from outbreaks. If β = 0, the rate of recovery is constant, if β < 0, the rate of recovery increases with time, and if β > 0, the rate of recovery decreases with time. We conclude from this distribution that there is a clear signal showing that the average reef recovery time is lengthening over the period for which data is available, i.e. β > 0. We interpret this signal as evidence that it is harder for reefs to recover from outbreaks in later years than in earlier years, other things being equal, indicating that key features of reef community structure have been damaged over time.

KEY WORDS: Crown-of-thorns · Active outbreaks · Reef recovery · Community structure · Reef degradation · Mathematical model · State transition probabilities · Bayesian statistics