White River Bioenergetics

Science Center Objects

The White River Basin is located in western Washington and drains an area of about 500 square miles. Rivers in the White River Basin are fed by melt water from glaciers on Mt. Rainier, runoff from snowmelt and rain, and groundwater discharge. Beginning in the early to mid-twentieth century, the White River from river mile (RM) 9 to its confluence with the Puyallup River was extensively channelized with levees to prevent meandering and minimize flooding. These system changes affected fish habitat and growth potential. In recent years, setback levees have been built both to increase the carrying capacity of the rivers and help reduce flooding, and to potentially improve fish habitat and fish growth potential. Additional setback levees are being considered, however, prior to building any additional setbacks, management agencies are interested in getting estimates of potential changes in available fish habitat. Ideally, levee setbacks would be designed to reduce flooding and provide improved habitat for native fish like the federally listed Chinook salmon (Oncorhynchus tshawytscha). 


White River Salmon Habitat and Energetics Model

Problem - High sediment input to the White River System from Mt. Rainier has resulted in high rates of aggradation of selected reaches of the Lower White River System. This process reduces the carrying-capacity of rivers and recently resulted in unexpected flooding beyond the 100-year flood boundaries in the City of Pacific after water was released from storage from behind Mud Mountain Dam. To minimize future flooding, one possible river-system management option being considered is building setback levees. It is not known how effective the setback levees would be for controlling flooding nor what their impact would be on fish habitat and fish growth potential. Ideally, levee setbacks would be designed to reduce flooding and provide improved habitat for native fish like the federally listed Chinook salmon (Oncorhynchus tshawytscha). Currently, insufficient information is available to quantify the impact of flood-mitigation strategies, like setback levees, on fish habitat and growth potential.

Objectives - The overall objective of the study is to evaluate how current and potential future hydraulic conditions between the A Street Bridge and the 8th Street Bridge on the Lower White River affect the potential for fish habitat and growth.

Specific objectives of the study are to:

  1. Estimate the abundance and distribution of potential juvenile spring Chinook habitat in the study reach for current and simulated post-setback levee conditions under low- and high-flow conditions; and
  2. Estimate the abundance and distribution of energetically favorable locations under current and future conditions for low- and high-flow conditions using a nested 2-dimensional hydrologic, invertebrate-drift, and bioenergetics model.


Relevance and Benefits - Natural resource agencies throughout the Pacific Northwest are constantly weighing the pros and cons of various management actions on human well being and natural resources. Management actions like the construction of flood control levees have generally had a very positive effect on reducing the potential harmful property loss, but have often been associated with negative effects on natural physical, chemical and biological processes and resources. The potential of constructing levee setback appears to address both the need to maintain flood protection, but also the need to re-establish the linkage between rivers and their floodplains. The proposed White River levee setback theoretically could substantially increase spring Chinook habitat. However, the biological benefits of such a levee setback are unknown and limit the ability of resource managers to assess the potential benefits of such a project. The proposed project will develop an integrated modeling approach to assess the biological value of the proposed project for spring Chinook salmon. The approach that will be developed will be transferable to other stream and river systems and will benefit other resource managers.

Approach - Task 1. Temperature data and invertebrate drift (potential fish food) will be characterized for the study reach. Some information will be derived from the literature, but those data will be supplemented with new data. In summer 2010, self-contained temperature recorders will be installed in the study reach. These recorders will be located in such a way as to characterize the range of temperatures found in different geomorphic settings throughout the study reach. The goal of this effort will not be to fully characterize temperature in the study area, but to produce information appropriate for inclusion in the foraging and energetics model discussed below. In addition, drift nets will be installed from dusk to dawn to provide general insight into the composition and size structure of drifting invertebrates entering the study reach. Drift nets will be installed to capture representative samples from the full water column in an effort to characterize both aquatic and terrestrial invertebrates. As with the temperature sampling, the drift sampling will not be an exhaustive sampling of invertebrate drift entering the study reach but a generalization of the taxonomic composition and size structure of invertebrates entering the study reach suitable for inclusion in the drift and foraging models discussed below. Invertebrates in the drift samples will be identified to the order or family level and measured for length. The energy density for each invertebrate prey type will be based on literature values (e.g., Cummins and Wuycheck, 1971; McCarthy et al. 2009). All invertebrate information used in the modeling efforts discussed below will be consistent across all management- and flow-scenarios examined. These scenarios will include current conditions under high- and low-flows and a hypothetical setback-levee scenario under high- and low-flows.

Task 2. USGS will generate 3-dimensional flow and depth coverages for the study reach under different management and flow scenarios. The coverages will build upon results from a spatially-referenced, 2-dimensional velocity and depth model for the study reach for current conditions and for a hypothetical setback-levee scenario for different flow scenarios (i.e., high- and low-flow) that King County will be responsible for generating and providing to USGS. For the 3-dimensional analysis, a series of transects will be established in the reach (fig. 3); the actual number and locations of the transects will be based on the resolution of the 2-dimensional model and geomorphic- and fisheries-related information. After establishing the transects, a series of flowlines will be established for the study reach (fig. 4). The first flowline will connect the highest 2-dimensionally modeled flows at each transect and will generally follows the channel thalweg. Additional flowlines will connect successively lower flows. The total number of flowlines is based on geomorphic and fisheries related information and may be modified during the modeling. The stream transects and flowlines create a series of polygons with associated depth and velocity information based on the 2-dimensional model. Each polygon will then subdivided to create a 3-dimensional coverage of the study reach (fig. 5). The number of vertical subdivisions will again be based on geomorphic and fisheries related information. Given that the velocity information for each of the 2-dimensional polygons will be depth-averaged, a logarithmic velocity/depth profile approach will be used to generate velocities for each of the 3-dimensional cells (Gordon et al., 1992). The results from Task 2 will be include a 3-dimensional coverage of depth and velocities for each management scenario, for current and levee-setback conditions, and for high-flow and low-flow conditions.

Task 3. The a) abundance of preferred juvenile Chinook habitat, and b) the available juvenile Chinook "food" throughout a study reach will be determined for each management and flow scenario.


Task 3a. Depth and velocity values for each cell in the 3-dimensional coverages generated by Task 2 will be compared to published values of preferred depths and velocities for juvenile Chinook salmon. The published preferred depth and velocity values for juvenile Chinook will undoubtedly have a range of values. We will use this range of information to generate maps and tables of low, mean, and high abundances of preferred depth and velocity habitat grid cells. This information will be used to empirically compare the abundance of preferred juvenile Chinook habitat under each management and flow scenario.



Task 3b. The primary food sources for juvenile Chinook are aquatic and terrestrial invertebrates drifting downstream. We will quantitatively determine the density of specific drifting invertebrate taxa within each of the 3-dimensional cells described in Task 2 to assess the available juvenile Chinook "food" throughout the study reach under each of the management and flow scenarios. Based on existing literature and limited drift sampling, we will establish an expected invertebrate density, size composition, and taxa composition for each grid cell at the upstream boundary of the study reach. This information will represent the potential Chinook food entering the study reach. We will use the depth and velocity information generated in Task 1 and drift-dispersion algorithms of Rutherford (1994) to model the drift dispersion of the invertebrates from cell to cell within the study reach. Traditionally, invertebrate drift models have often assumed uniform drift densities which are known to be unrealistic (Stark et al., 2002). For this study, the approach described above will predict how variations in depth, flow, invertebrate abundance and behavior interact to affect drift density in a spatially explicit manner. For all management and flow scenarios, a table of taxa-specific drift densities will be produced for each cell.


Task 4. The depth and flow information from Task 2 and the species-specific drift-density information from Task 3b will be used to model the net energy intake of drift-feeding juvenile Chinook for each of the 3-dimensional cells for the study reach for each management and flow scenario. We will use the foraging algorithms developed by Hughes et al. (2003), wherein the net rate of energy intake equals the gross rate of energy intake minus the energy cost of steady swimming, and feed-and-hold positions. To implement these algorithms, a number of input variables are needed. Data for some of these input variables are available in the literature, while others will require some assumptions. The following data will be needed for this analysis:

  1. Fish length and weight for juvenile Chinook salmon (literature data);
  2. Minimum and maximum invertebrate prey size, which will be estimated as 1.15 percent and 45.2 percent of fish length (Wankowski, 1979; Hayes et al., 2000; Hughes et al., 2003); and
  3. Probability of prey detection and number of invertebrate prey "interrogations" by the Chinook. Values for these variables are available in the literature, but are based on fish located in clear-water streams. Given the high level of turbidity in the White River, both of these values will need to be reduced. While identifying appropriate values for both of these variables in a glacially-fed river are a unique challenge, values that will be used will be consistent across management and flow scenarios to facilitate appropriate empirical comparisons.

The results from Task 4 will be a table of net rate of energy intake for juvenile Chinook salmon for each cell of the 3-dimensional coverage for each of the management and flow scenarios. Empirical comparisons of the number of cells with positive net rates of energy intake will be made to evaluate the two management scenarios under high- and low-flow conditions. In addition, 2-dimensional graphical representations of the location and average value of the energy intake will also be developed.


Task 5. The final task will be to summarize the results in a technical report/journal article and provide all relevant data and model output to King County.