Bayesian networks and adaptive management of wildlife habitat.

摘要: Adaptive management is an iterative process of gathering new knowledge regarding a system's behavior and monitoring the ecological consequences actions to improve decisions. Although concept originated in 1970s, it rarely actively incorporated into restoration. Bayesian networks (BNs) are emerging as efficient decision-support tools well suited adaptive management, but examples their application this capacity few. We developed BN within adaptive-management framework that focuses on managing effects feral grazing prescribed burning regimes avian diversity woodlands subtropical eastern Australia. constructed with baseline data predict bird abundance function habitat structure, pressure, burning. Results sensitivity analyses suggested pressure increased aggressive honeyeaters, which turn had strong negative effect small passerines. Management interventions reduce were then conducted, after we collected second set field test response passerines these measures. used data, changes may have resulted from interventions, validate update BN. The network predictions passerine under conditions very accurate. updated concluded first iteration will be planning next round interventions. unique belief-updating feature BNs provides land managers flexibility outcomes evaluate effectiveness