Predict salmonid migration pathways from flow simulations in rivers

This project proposes a model that can be used to predict the movements of migratory salmonids (salmon and trout) in regulated rivers by analyzing flow dynamics alone. Most migratory fish rely on flow-related signals to navigate through rivers. A deeper understanding of how these signals guide, attract, or deter fish is crucial for improving the effectiveness of fish passage solutions. Until recently, it has been difficult to study the relationship between fish behavior and flow dynamics due to the lack of suitable technology for collecting reliable data.

Fish passages have long been used to provide migratory fish with alternative routes past barriers such as hydropower plants and dams. However, uncertainty remains regarding how these passages should be placed and designed to ensure that as many fish as possible can locate and use them. Many fish passages have low efficiency because fish struggle to find the entrance and instead swim into competing flow paths. These competing flows are often much stronger than the flows through the passages, making it challenging to create attractive guidance flows toward the desired migration routes. Conducting experimental field studies in rivers to evaluate hydraulic conditions and map fish movement patterns in relation to passage structures can also be difficult and costly. Hydraulic modeling can serve as a complementary tool to better understand fish movement by linking flow dynamics with observed migration behavior in wild fish.

By using newly developed high-resolution acoustic telemetry to track fish, along with detailed mapping of river flow dynamics, it is now possible to combine these data to develop models that help predict which migration routes fish are most likely to use. This is a valuable tool in planning new fish passages and optimizing the performance of existing ones under varying flow conditions. Ensuring cost-effective implementation of passage solutions for migratory fish in the coming decades is essential to meet the modern environmental requirements now placed on Swedish hydropower.

The development of the model in this project depends on linking positional data from tagged fish to the flow dynamics at the same location as the observed fish tracks. In the next step, environmental variables and behavioral rules of increasing complexity are incorporated to identify which factors best reproduce the observed movement patterns of tagged fish.