PS 32-193 - Development and validation of a spatially-explicit agent-based model to simulate habitat use of juvenile Chinook salmon (Oncorhynchus tshawytscha) in the Sacramento River

Abstract: The juvenile life stage of Chinook salmon (Oncorhynchus tshawytscha) is characterized by freshwater residence prior to ocean migration, during which juveniles encounter diverse habitats increasingly being transformed by anthropogenic changes to river ecosystems. Realistically representing the habitat selection behavior and movement of juvenile salmonids in population models is therefore a critical step in assessing how salmonid populations will respond to ecosystem change and developing population management tools. The agent-based modeling (ABM) framework has been increasingly used to simulate realistic animal movement behavior by allowing habitat selection patterns to emerge from interactions between agents and between agents and their environment.

We developed a spatially explicit agent-based model to simulate habitat selection and movement of juvenile Chinook salmon rearing in and outmigrating from the Sacramento River. We simulated habitat selection on a daily scale and utilized rasters to represent the spatially heterogeneous external drivers of juvenile movement, including river velocity, depth, ground and canopy cover, and predator presence. These external drivers of movement were linked with mechanistic models of drift feeding, metabolism, and growth. Habitat use represented a response to changes in their physiological state, in which juveniles either balanced the energy availability and predation risks of their surrounding environment when selecting habitat or made a trade-off between predation risk and energy availability when selecting habitat.

Preliminary analyses indicate that some model outputs were consistent with observations from juvenile populations in the Sacramento River. For example, the outmigration timing of juvenile agents coincided with increases in river discharge, and the main mortality causes of agents were poor condition and predation. To assess the validity of the model more rigorously, future work will consist of using a pattern-oriented modelling approach to test the model’s ability to reproduce patterns found in real juvenile salmonid populations. Patterns of interest include responses of juveniles to fluctuating flows, responses to intraspecific competition, and long-term survival of rearing juveniles and migrants. Eventually, the model will be used to forecast how juvenile salmon might respond to habitat alteration changes.