When Tradeoffs Are Not What They Appear and Robustness May Not Exist: The Fisheries Challenge

Managing ecosystems poses a significant decision-analysis challenge, as these systems are often governed by highly nonlinear dynamics and threshold responses. Socio-ecological systems are typically governed by multiple stakeholders who either do not know or cannot converge on probability density functions for the parameters describing the system or even the system description (model) itself. This predicament has been referred to in the literature as "deep uncertainty" and it adds formidable complexity to the decision-making process, especially in cases with potentially irreversible tipping points. An example of such a socio-ecological system can be found in fisheries, the management of which must balance economic and environmental objectives. Management strategies need to be established to maximize fish harvest while avoiding population collapse. How such complex systems are modeled, in terms of which system components and interactions are included or excluded in the analysis, can yield severe and surprising outcomes. In this study we explore the effects of different model descriptions and management strategies on the acquired ecological-economic tradeoffs under different model parameterizations to aid in the discovery of robust management policies. Our results show that classical assumptions for the highly nonlinear fisheries game, with respect to the stability of attained tradeoffs and their ability to inform stakeholder preferences is questionable. Small changes in deeply uncertain factors can fundamentally reshape system dynamics, risks, and the validity of the candidate management strategies. Insights from this study highlight the importance of ensuring models capture deep uncertainties in fundamental system dynamics as well as a breadth of economic and ecological criteria when seeking to discover if robust management options even exist.