Multiple attribute Bayesian analysis of adaptive ecosystem management

Adaptive ecosystem management is based on the premise that human understanding of ecosystems is incomplete and biophysical responses to management actions are highly uncertain. Uncertainty is reduced and learning occurs by conducting experiments that provide information about how ecosystems are likely to respond to alternative management actions. Implementation of adaptive ecosystem management is hampered by the lack of an implementation framework. This paper proposes a two-stage hierarchical framework for identifying effective management actions. The first stage uses Bayes rule to identify management actions that have the greatest likelihood of achieving a sustainable ecosystem state. Sustainability is measured by ecological productive capacity. The second stage evaluates the ecologically sustainable ecosystem states identified in the first stage based on their non-ecological attributes. The optimal management action is the one that has the greatest likelihood of achieving an ecologically sustainable state and provides the most preferred combination of non-ecological attributes. The two-stage framework requires a resource manager to identify and measure ecological and non-ecological attributes of ecosystems, specify prior probabilities for ecosystem states, estimate likelihood functions for ecosystem states and weights for non-ecological attributes, and select a procedure for ranking sustainable management actions based on their multiple non-ecological attributes.