Using Bayesian belief networks in adaptive management

Abstract

Bayesian belief and decision networks are modelling techniques that are well suited to adaptive-management applications, but they appear not to have been widely used in adaptive management to date. Bayesian belief networks (BBNs) can serve many purposes, from illustrating a conceptual understanding of system relations to calculating joint probabilities for decision options and predicting outcomes of management policies. We describe the nature and capabilities of BBNs, discuss their applications to the adaptive-management process, and present a case example of adaptive management of forests and terrestrial lichens in north-central British Columbia. We recommend that those unfamiliar with BBNs should begin by first developing influence diagrams with relatively simple structures that represent the system under management. Such basic models can then be elaborated to include more variables, the mathematical relations among them, and features that allow assessment of the utility of alternative management actions or strategies. Users of BBNs should be aware of several important limitations, including problems in representing feedback and time–dynamic functions. Nevertheless, when properly used, Bayesian networks can benefit most adaptive-management teams by promoting a shared understanding of the system being managed and encouraging the rigorous examination of alternative management policies.