| Abstract: | To properly manage streams for fish habitat it is necessary to move beyond standard evaluations and the acceptance of relationships between variables that are derived either from rivers of much greater scale or those which are based on time-averaged conditions. Clearly, the relationship between moving fluid and channel boundary is the most important in smaller streams that are used as habitat by fish, as variations in speed patterns causing erosion or sedimentation of the bed will determine the success of spawning activity. To this end, an evaluation of fluid speed profiles is presented using field data collected from spawning streams in Ontario, Canada. The data have been subjected to an interpolation algorithm, minimum curvature, to deliver a more accurate view of fluid speed patterns. The results show that fluid speed patterns vary completely within the same cross-section under different flow regimes, even when the mean depth and mean speed of the flow are virtually unchanged, and that mean speed depiction of flow is inaccurate when considering particle transport/deposition. A comparison of data collection practices is presented which shows that flow measurement from the common 60% of the depth at mid-channel produces large errors in speed prediction on moving towards the bed and surface of the channel. Errors in the prediction of near-bed fluid speeds can result in the entrainment of particles within the preferred spawning range of salmonids, which was not predicted through the use of mean fluid speed for the cross-section or through the less intense data collection technique. The minimum curvature technique of data interpolation is shown to be an accurate method for visualizing fluid speed patterns. |
