Exploring anthropogenic activities and management decisions using a novel environmental agent based model

Lake whitefish (Coregonus clupeaformis) are an ecologically, economically, and culturally important species to the native and non-native fishers of Lake Huron, Canada. Studying the effects of anthropogenic activity on lake whitefish is of utmost importance to ensure this species remains viable in its environment for sustainable harvest. One analysis tool that is frequently used for ecological population risk assessments are agent-based models (ABMs), in which the population is represented as a network of heterogeneous individual agents that interact with one another and their environment. However, in an ABM that incorporates a high level of biological detail to model a large population moving within a spatial environment over time, significant computation is required to manage, manipulate, and store the relevant data for each agent over successive time iterations. We introduce a new approach to ABMs known as environmental ABMs (enviro-ABMs) to reduce this computational expense and simulation runtime. Specifically, we divide the environment into a collection of spatially indexed cells and treat each of these as a single agent, allowing fish to move from one contiguous cell to another. This reduces the computational requirements to a limited number of active cells. In addition to more predictable computational requirements, this method keeps all fish sorted by age and location for efficient mortality, spawning, and harvest operations, and reduces the amount of computational overhead needed. Applying the enviro-ABM to our case study in Lake Huron, we demonstrate how it can be used to model anthropogenic activities and stressors that may affect lake whitefish, and how the model can be used to facilitate fisheries management decision making. While the model is applied specifically to the case of whitefish in Lake Huron, it can be generalized to conduct risk assessment for other species in a variety of habitats.