STATE AND TRANSITION MODELS FOR RANGELANDS .7. BUILDING A STATE AND TRANSITION MODEL FOR MANAGEMENT AND RESEARCH ON RANGELANDS

State and transition models have recently emerged as flexible conceptu al frameworks for abstracting information concerning vegetation change in rangelands. They characterise the rangeland through both a partiti oning of the system in terms of multiple steady states, and the identi fication of transitions between those states. This paper examines the steps and some of the issues involved in building such models for reso urce management and research purposes. The number and nature of states defined depends on: the specific objective of management; the way in which the rangeland functions; and the state of existing knowledge abo ut its dynamics and structure. Each transition may be defined in terms of a suite of causes which explain the mechanisms of vegetation chang e involved. Each cause can then be defined in terms of a probability o f occurrence, a time-frame for the cause to be maintained to effect a transition, and a confidence rating in the prediction. State and trans ition models should be applied at spatial and temporal scales that are relevant to the scale of disturbance as well as to the context within which management decisions are made. Identifying the appropriate vari ables and processes that reflect these scales is critical. Developing these models should also be an iterative process involving progressive refinements to ensure validity, accuracy and usefulness.