January 11 2012

4:00 pm 154 BSRB

Louis Gross
NIMBioS, & Dept. of Ecology & Evolutionary Biology & Mathematics, University of Tennessee, Knoxville

Ecology and Big Projects: Lessons from Everglades Restoration Planning


As human-dominated systems spread across our planet, we need to evaluate how human modifications to natural systems impact the services provided by ecosystems. While this is an inherently global issue, much of our planning associated with environmental systems occurs at a regional spatial extent. Mathematical and associated computational models play a key role in regional planning since it is impractical to carry out
controlled experiments at these scales. Models can assess alternative scenarios, incorporate the trade-offs important to stakeholders with differing criteria for the objectives of management, and consider the impacts of uncertainty of future conditions. One of the largest attempts to date to evaluate alternative management plans at regional extent has been the massive project associated with the restoration of the Everglades of South Florida. I'll describe the process by which a set of models has been used to address biological and management questions for various ecosystem components, and give some lessons learned from this long-term project. This will be placed in a context of using models to to develop hypothetical "best" methods to manage a system, and connected to applications of optimal control methods in areas such as wildlife management and disease ecology that incorporate spatial variation.















































































































































































































































































































































































































































































































this is idtest: