Development and application of ecosystem models to support fishery sustainability: A case study for the Gulf of Alaska

In this dissertation, I develop quantitative ecosystem models to be used in fisheries management applications. The current goal of fishery management is to sustain fisheries for individual exploited species. In the first chapter, I develop a concept of "ecosystem sustainability" which extends this goal to sustain relationships between species (including humans), and which views ecosystems as complex adaptive systems which are simultaneously robust to random disturbance and yet prone to rapid, irreversible state changes under certain conditions. There I review the history of commercial exploitation in the Gulf of Alaska from 1740 through the present day within this ecosystem sustainability context. In the following chapters I develop two types of models. In Chapter 2 I employ a "standard" modeling approach addressing static and dynamic system properties that has been applied in many other fished marine ecosystems. Using this standard approach, I identify and classify relationships between fisheries and fished species in the Gulf of Alaska food web, and dynamic relationships between predator-prey interactions, fishing, and climate change over the history of commercial exploitation. In Chapter 3 I use novel "complex systems" modeling approaches which have been applied in many biological, social, and physical systems, but have not yet been applied in fished marine ecosystems. I use the complex systems approaches to identify both static and dynamic sources of robustness in the Gulf of Alaska ecosystem, and also potential ecosystem thresholds where the state of the ecosystem is prone to rapid change. Some potential thresholds relate to food web structure, while other thresholds relate to fishing intensity in the ecosystem. In the final chapter, I synthesize the standard and novel modeling approaches in management scenarios developed specifically for the Gulf of Alaska, and argue that combining multiple modeling approaches with a realistic accounting of uncertainty can support fishery sustainability in the ecosystem context.