Dynamic programming for resource management

Resource management strategies, including management plans for endangered or invasive species, habitat enhancement and wildlife harvesting, involve balancing current costs and benefits with potential future costs and benefits that are affected by actions taken now. The Markov Decision Process (MDP) approach provides a rigorous and transparent approach for developing such strategies.The goal of this project is to make the process of specifying and solving MDP models easier and to develop tools that will help analysts work with decision makers and stakeholders. Specifically, software will be developed to provide a graphical framwork for specifying relevant variables and their interactions. This approach breaks model building efforts down into a set of small, clearly defined steps. Applications of the approach will include the development of strategies to manage recreational/breeding conflicts at golden eagle nesting sites and habitat management for endangered species including Florida scrub jays. In addition, the project will address difficulties that arise for management when the structure of the modeled system is not well understood or when key variables cannot be accurately observed. In particular, strategies will be examined that address if and when experimentation and/or monitoring programs will improve management goals.