143rd Annual Meeting of the American Fisheries Society

September 08 - 14, 2013

Ecological modeling techniques that link biological response with habitat conditions, whether mechanistic (e.g., bioenergetics) or statistical (e.g., regression), whether simple (e.g., correlation) or complex (e.g., individual-based), represent a toolbox that we can use to address research and fisheries problems.  Selecting a technique out of the toolbox should ideally be guided by the research or management question at hand; however, as certain techniques gain visibility and traction in the literature, or are touted as the ‘best available approach’, they are sometimes used without consideration of application potential or the tradeoffs between realism, precision, and generality.  We are organizing this symposium to generate discussion on modeling techniques, inviting presentations that pose one or more of the following questions: 1) What are the tradeoffs between model precision, realism and accuracy?, 2) Are mechanistic models better than statistical models for ecological forecasting?, 3) What can we learn from simple vs complex models?, 4) How useful are heuristic models?, and 5) How important is model validation and how do we use models that cannot be validated?  Following presentations, we will convene a panel that balances modelers and model users to address the overarching question – why would you use a certain kind of model for a certain research question or a management problem.