A Combined Hydraulic-Habitat Model For Assessing Restoration of Fish Passage At A Low Head Dam

This project developed a combined hydraulic-habitat model for assessing fish passage options for the Long Tom River located in the Willamette Valley of western Oregon. This work was undertaken as part of a habitat restoration feasibility study in collaboration with the City of Monroe and the Confederated Tribes of Siletz Indians. In the 1940’s, the USACE’s Long Tom River channel rectification and improvement project included channel dredging, straightening, channel embankments, and construction of three drop structures to improve channel capacity and limit erosion for receiving outflows from Fern Ridge Dam (river mile 24). The Monroe Drop structure (river mile 7) is the first impediment to upstream fish passage by several salmonid species and Pacific lamprey. A city-owned park with wetland areas connected to the main channel via a network of culverts is adjacent to the structure. Restoring fish passage was the primary goal, and we also sought options for improving the connectivity between the channel and wetland areas in the city park area. Alternatives considered included dam removal, several bypass-channel options, rock ramp variations, as well as combinations of these features. A two-dimensional hydraulic model was developed for a 6-mile river reach covering 2.5 square miles that consisted of approximately 48,000 grid cells in the model. Cell sizes ranged from 15 feet in the river channel up to 100 feet in floodplain regions. Upstream fish passage potential was greatest during the winter months. Model flows included 72, 800, and 4360 cubic feet per second, the 95%, 50%, and 5% daily flow duration values, respectively, based on nearby gage data over an 80-year period. Habitat suitability index (HSI) curves were obtained from literature that focused on adult cutthroat and juvenile Chinook preferences with respect to water depth and velocity. At each model grid cell, the hydraulic model’s output of velocity and depth were used to calculate habitat value based on the HSI curves (four in total). Habitat scores were normalized by the area of the individual model cells and then summed over the total model domain for each HSI curve. Fish passage efficiency of each alternative, represented by a factor ranging from 0.5 for no action to 0.95 for alternatives that included full removal of the drop structure, was qualitatively assessed and applied to the aggregate habitat scores. Non-adjusted habitat scores typically scaled with the length of bypass channel added under each alternative for the median and high winter flows. Rock ramp features scored better under low flows where bypass channels no longer offered adequate flow-through capacity. Application of fish passage scaling factors favored alternatives that included removal of the drop structure, with greater overall scores for those that also increased access to side channel habitat. Results of our combined hydraulic-habitat model allowed us to take a tiered approach in evaluating habitat restoration alternatives and final habitat values typically increasing two-fold over the no action alternative for the top five ranking alternatives.