Experimental study of fish response to turbulent flow fields generated by in-stream structures

The influence of hydrodynamics on habitat choice and migration path of fish is of great interest for biologists and engineers, especially when considering the impact of in-stream structures like dams or fish passages during stream restoration projects. Research is needed in both laboratory and field environments to better understand how fish respond to different hydrodynamic conditions resulting from various aquatic environments, to better predict fish behavior in altered streams. While work has been done to observe fish response, including locomotion and tail beat behavior under increasingly complex flow conditions, models often rely on bulk or simplified flow parameters to correlate with fish behavior. The present study, uses a novel system to obtain high spatial and temporal resolution data that allows for correlation of mean and turbulent flow statistics with swimming behavior. Mean velocities, turbulent intensity, Reynolds stresses, and vorticity are investigated to explore both intensity and orientation of turbulent flow structures and their impact on swimming speed, acceleration, as well as tail beat frequency and amplitude in response to modified flows. A volumetric three-component velocimetry (V3V) system was used to investigate the 3D velocity field generated by various obstructions: horizontal cylinder, vertical cylinder, gravel bed, and splitter plate, in addition to a flat bed case for reference as control group. Videos of fish swimming behavior (Micropterus salmoides) for each scenario were analyzed to determine fish position, swimming speed, and fish tail beat behavior. Multiple comparisons were made between hydraulic parameters and fish swimming response, as well as among different scenarios to reveal the relationship between intensity and orientation of turbulent flow structures and fish swimming preferences. Based on this project, it is found that bulk mean velocity didn't show any consistent correlation with fish swimming preferences. However, our investigated species showed a clear preference for regions with low turbulent intensity, turbulent kinetic energy and vorticity, even with high levels of Reynolds stress. Studies on both vertical and horizontal obstructions also highlighted the relevance of coherent flow structure orientation, with horizontal eddies, generated by the vertical obstruction, being more easily utilized than those in a vertical plane. While this study provides valuable information on flow-fish interactions, further studies are required with a broader variable space and larger-scale facilities to find general correlations under more realistic scenarios