The extensive loss and degradation of North America’s grasslands presents a formidable challenge to resource managers. Much of the native mixed-grass and tallgrass prairies managed by the U.S. Fish and Wildlife Service (Service) in the northern Great Plains are extensively invaded by introduced cool-season grasses, principally smooth brome (Bromus inermis) and Kentucky bluegrass (Poa pratensis). Management to suppress these invasive plants has had poor to inconsistent success, mainly for lack of understanding of prairie restoration ecology and absence of systematic evaluation of management effects. The central challenge to managers is selecting appropriate management actions in the face of biological and environmental uncertainties.
In partnership with the Service, the USGS is developing an adaptive decision support system to assist managers in selecting management actions under uncertainty and maximizing learning from management outcomes. The framework is built around the practical constraints of refuge managers and includes identification of the management objective and strategies, analysis of uncertainty and construction of competing decision models, monitoring, and mechanisms for model feedback and decision selection. We describe the technical components of this approach, how the components integrate and inform each other, and how data feedback from individual cooperators serves to reduce uncertainty across the whole region.
Results/Conclusions
Twenty refuges, spanning five states of the northern Great Plains, are participating in the project. They share the same management objective (increase native prairie composition), available management strategies (fire, grazing, haying), and uncertainties (system response to management). Our models describe vegetation dynamics as a function of plant parameters that respond to management actions and uncontrolled effects. Given a set of vegetation input states and a management decision, the models predict future system states. Managers make decisions and monitor annually. At each decision cycle, we use adaptive stochastic dynamic programming to identify the optimal management decision that best pursues the objective, given the present vegetative state and our relative beliefs in the competing models. Monitoring provides the data needed to assess the predictive performance of our models and update their relative influence on the selection of the next management decision. Decision making under this framework is adaptive, as monitoring feedback increases understanding of the system and in turn determines the path of future decision making. While the scope is broad, the project interfaces with individual land managers who provide refuge-specific information and receive updated decision guidance that incorporates understanding gained from the collective experience of all cooperators.
