Multi-model inference in support of marine ecosystem-based fisheries management

As marine ecosystem-based management (EBM) moves towards operational status, it is clear that models will play a key role in assessing the implications of alternative management actions.  Here we describe a suite of such models applied to elements of the Northeast U.S. Continental Shelf Large Marine Ecosystem. These models range in complexity from simple aggregate production models of community dynamics to detailed simulation models such as Atlantis, capable of incorporating physical, ecological and social dimensions.  These models can serve as operating models in management strategy evaluations (MSEs) embedded within integrated ecosystem assessments for the region.  MSEs model the entire process from system dynamics through the assessment process itself, incorporating the testing of management scenarios and accounting for risk and uncertainty. The use of multiple operating models permits comparison of management inferences from each to assess whether the overall conclusions are robust.  We provide examples of the use of these models for simulated scenarios of ocean use, largely focusing on fishery and protected resource species issues.  Our intended focus with these operating models is on strategic rather than tactical management advice.  This approach allows us to focus on tradeoffs which emerge in the evaluation of potentially competing management objectives.