Coping with uncertainty in modeling environmental resources: practices, pitfalls & recommendations from modelers, restoration biologists & managers

Title: The Role of Uncertainty in Limiting Demographic Model Application in Oyster Management in Cheapeake Bay
Abstract: An oyster demographic model was developed for the purpose of comparing potential oyster population outcomes from several Chesapeake Bay oyster management alternatives evaluated in a Programmatice Environmental Impact Statement (PEIS) issued in June 2009. Model parameters for the native Eastern oyster were derived from management data and research findings collected over many decades. The intent was to use findings from a series of more current research studies of the Asian oyster to revise model parameters to be representative of that species and run the model for non-native species alternatives. A peer review panel critiqued the modeling effort throughout its development. An Oyster Advisory Panel (OAP) conducted a subsequent peer review of the use of model outcomes in the draft PEIS. The OAP concluded that the level of uncertainty associated with model projections was so great that it would be inadvisable to rely on the outcomes in making management decisions, even for native oyster alternatives. In particular, levels of uncertainty about benefits and risks associated with the Asian oyster were found to be unacceptable. Sources of uncertainty regarding native oyster outcomes included basic data inputs that had to be derived from disparate data sources and numerous assumptions required to construct the basic life history architecture of the model. The primary source of uncertainty regarding the Asian oyster was the absence of reliable life history data that would characterize a reproducing population of the species within the Chesapeake Bay environment. While the model has proven to be a useful scientific tool for investigating potential relationships among various life history and environmental factors influencing oyster populations in the Bay, unresolvable uncertainties precluded its use in making management decisions, particularly when being applied to highly controversial management alternatives. As a result, only very limited applications of demographic model outcomes were included the anlyses included in the Final PEIS.
Authors: Richkus, , , ,
Presenter: William Richkus - Versar, Inc.