Abstract
One of the aims of environmental management is to achieve good ecosystem state. Assessing the state needs to be informed by thresholds above which state is defined as "good" for both the quality that defines good state, and the extent of the habitat that needs to be in such a quality. Operationalizing such thresholds has been carried out using a wide variety of approaches, with, often, haphazard and subjective outcomes. Here, we review approaches for setting "good-state" thresholds and evaluate their strengths and weaknesses for application to marine seabed ecosystems. Only two approaches defined a current ecologically meaningful good state and estimated thresholds quantitatively from data, while two other approaches ("avoid collapse" and "allow recovery") would result in a state that could recover to good in the future. Other methods were subjective in the choice of threshold or based on statistically detectable thresholds rather than thresholds between good and not good or degraded state. We argue that the most objective method for setting a good-state threshold is based on maintaining the state within the range of natural variation in undisturbed systems. Preliminary time-series analyses of marine seabed community biomass suggest this threshold is located between 54 and 79% of the undisturbed state.