Behavioral games and emergent population dynamics

Large-scale models of ecosystems typically do not include explicit habitat choice. This is in spite of adaptive habitat choice being known to have a powerful influence on ecosystems, with indirect effects often being stronger than direct effects. The importance of behavior is particularly pronounced in the aquatic setting, where population dynamics are determined by the diel vertical migration. There is no general toolbox which a working ecologist can apply to add behavior to an ecosystem model. The goal of this thesis is to develop game-theoretic methods to address this shortcoming, and apply the tools to model aquatic ecosystems. This thesis only considers unstructured populations, ignoring the import of ontogeny.

Paper A develops a general method for implementing optimal habitat choice in ecosystems of Lotka-Volterra type, both with continuous and discrete habitats. We apply the method to a predator-prey system, modeling copepods and forage fish in the water column. Paper B focuses on the ecosystem impact of optimal behavior on a tri-trophic ecosystem with a refuge, and how behavior changes the impact of bottom-up and top-down forcing. Finally, the paper investigates the relationship between the Type II and Type III functional responses. Paper C develops a general method to study optimal habitat choice in population games. Paper D couples stochastic mean-field games to predator-prey population dynamics, revealing the emergence of diel migration patterns as a result of ecosystem productivity. Finally paper E is included to show the potential of the methods, but is still a draft. Paper E studies the impact of optimal behavior on a shelf ecosystem, which support the majority of the worlds fisheries.