Development and Evaluation of a Chinook Salmon Smolt Swimming Behavior Model

Hydrologic currents and swimming behavior influence routing and survival of emigrating Chinook salmon in branched migratory corridors. Behavioral particle-tracking models (PTM) of Chinook salmon can estimate migration paths of salmon using the combination of hydrodynamic velocity and swimming behavior. To test our hypotheses of the importance of management, models can simulate historical conditions and alternative management scenarios such as flow manipulation and modification of channel geometry. Swimming behaviors in these models are often specified to match aggregated observed properties such as transit time estimated from acoustic telemetry data. In our study, we estimate swimming behaviors at 5 s intervals directly from acoustic telemetry data and concurrent high-resolution three-dimensional hydrodynamic model results at the junction of the San Joaquin River and Old River in the Sacramento-San Joaquin Delta, California. We use the swimming speed dataset to specify a stochastic swimming behavior consistent with observations of instantaneous swimming. We then evaluate the effect of individual components of the swimming formulation on predicted route selection and the consistency with observed route selection. The PTM predicted route selection fractions are similar among passive and active swimming behaviors for most tags, but the observed route selection for some tags would be unlikely under passive behavior leading to the conclusion that active swimming behavior influenced the route selection of several tagged smolts.