Scientific Investigations Report 2006–5230
U.S. GEOLOGICAL SURVEY
Scientific Investigations Report 2006–5230
Hydraulic and habitat simulation models contained in PHABSIM (Waddle, 2001) were used to characterize instream physical attributes (depth, velocity, and substrate) during expected summer (July through September) streamflow. Estimates of fish cover also were recorded, but were not used in the model. To estimate fish habitat available over a range of discharges, hydrologic and habitat data were collected at a few targeted discharges representing the range of discharges for the period of interest at each study site. These data were used to calibrate a hydraulic model, which then was used to predict the stream hydraulic attributes (depth and velocity) over the range of discharges of interest. The biological importance of the stream hydraulic attributes then was assessed with the suitability criteria for each species and life stage to produce a relation between habitat availability and discharge. The final output was expressed as WUA for a representative stream segment. To facilitate interpretation, the WUA results were normalized to a percentage of maximum for the range of discharges simulated.
The hydraulics portion of the PHABSIM model includes the water-surface elevations and velocity distributions. Data required in this part of the model and collected in the field are: channel geometry, Manning’s roughness (n) values, water‑surface elevations, water velocities, and stream discharges. Water-surface elevations can be calculated using one or any combination of the following methods: (1) stage-discharge relation or rating curve (STGQ), (2) Manning’s equation (MANSQ), or (3) step-backwater water-surface profile (WSP) (Waddle, 2001). In most cases, the stage-discharge relation method is used only when three or more discharges and corresponding water-surface elevations are available. In both the stage-discharge relation and Manning’s equation methods, the individual transects are independent of each other. In the WSP method, the individual transects are hydraulically connected.
The hydraulic portion of the PHABSIM model is calibrated in two steps. First, attempts are made to match simulated water-surface elevations with measured elevations for the calibration discharges. Calibration is done by adjusting the n values or related roughness variables within a realistic range as observed in the field until simulated water-surface elevations match or nearly match measured elevations. A difference of 0.02 ft or less between the simulated and measured values typically is desirable (Waddle, 2001). Second, attempts are made to match simulated velocities at each transect with measured velocities for the calibration discharges. This calibration is done by adjusting local n values in specific cells until simulated velocities match or nearly match measured velocities. It may be unrealistic to exactly simulate a measured velocity distribution. However, in relatively smooth, uniform channels, it may be possible to closely simulate a measured velocity. Velocity distributions for fairly rough, nonuniform channels are more difficult to simulate, and the final calibration values are based on the user’s selection of the simulation that best represents the measured values (J. Henriksen, U.S. Geological Survey, oral commun., 2004). Velocity adjustment factors that generally increase with increasing discharge (Waddle, 2001) also were used to evaluate model performance. Calibration data sets for each study site are available in IFG4 data files at http://id.water.usgs.gov/projects/salmon_streamflow.
PHABSIM use requires target species selection, life stages present during the period of stream use (periodicity), and habitat suitability criteria (HSC). This information was derived from previous SRA studies by the BIA and USFS in the Salmon River Basin (EA Engineering, Science, and Technology, Inc., 1991a, 1991b; R2 Resource Consultants, 2004; Rubin and others, 1991). Upon review of this information, the Interagency Technical Workgroup (ITWG) (see “Acknowledgments” for list of members) directed the USGS to target the ESA-listed species bull trout, Chinook salmon, and steelhead trout for juvenile, adult, and spawning life stages (J. Spinazola, Bureau of Reclamation, written commun., 2005). The endangered sockeye salmon was not selected as a target species because its habitat in the upper Salmon River Basin generally is not directly affected by diversions. The ITWG also directed the USGS to not include the fry life stage (<50 mm, or about 2 in.) because of the inability to accurately measure microhabitat parameters at a meaningful scale.
Species-specific HSC that accurately reflect habitat requirements during the life stage of interest are essential to developing meaningful and defensible instream flow recommendations. Suitability criteria quantify the relative importance of depth, velocity, and channel index (substrate) for specific life stages of each species. HSC are interpreted using a suitability index (SI) on a scale of 0 to 1, with zero being unsuitable and one being most used or preferred. The best approach is to develop site-specific HSCs for each species and life stage of interest. Alternatively, HSCs can be developed from existing literature.
Because time and budget constraints precluded developing stream-specific HSC, ITWG also directed the USGS to use existing HSC developed SRA processes. The fish species HSCs selected for this study were developed in the Pacific Northwest and Idaho. In addition, invertebrate HSCs published by Gore and others (2001) were used to simulate streamflow and habitat suitability for riffle dwelling Ephemeroptera, Plecoptera, and Trichoptera (EPT) taxa. These three orders of insects make up the predominant invertebrate fauna of riffle habitat and provide a major food source for stream dwelling trout and salmon (Gore and others 2001; Maret and others, 2001). Evaluating desirable streamflow needs for riffle habitat and associated invertebrates is particularly important because these are the first habitats to become dry when streamflows are reduced below baseflow. According to Jowett (2003) instream flow requirements for fish should also consider habitat for benthic invertebrates, particularly where food availability may limit fish numbers and (or) growth. The HSC and periodicity (period of stream use) for the various fish species and life stages targeted in this study can be accessed at http://id.water.usgs.gov/projects/salmon_streamflow/habitat_curves.
Maximum juvenile WUAs and median (Q.50) summer (July and August) streamflow data collected by Maret and others (2004) from Fourth of July, Pole, Elk, and Valley Creeks revealed that maximum preferred juvenile salmonid habitat predicted by the model often was less than summer median streamflow. For example, a summer streamflow comparison from streams in the upper Salmon Basin established on average that maximum WUA for juvenile Chinook salmon were only 33 and 63 percent of the July and August Q.50, respectively. Similar relations between streamflow and maximum WUA also were determined for juvenile steelhead and bull trout. Reasons for this likely result from HSCs that were developed during drought conditions (Rubin and others, 1991) and the potential inability to measure accurately microhabitat parameters at a scale that would be meaningful using PHABSIM. Therefore, modeling results for the juvenile life stage are not presented.
ITWG recommended a July through September study period because water is diverted for irrigation mostly during summer. High streamflows for channel maintenance generally have not been a problem in the upper Salmon River Basin (Bohn and King, 2000; M. Moulton, U.S. Forest Service, oral commun., 2003).
The habitat program HABTAE in PHABSIM was used to estimate WUA for the simulated discharges of interest. HABTAE uses the SI values derived from each cell in a transect for depth, velocity, and substrate. The geometric mean calculation was used to derive the composite index (CI) score for each cell at a transect. The CI was calculated as the geometric mean of the input variable:
CI = (SIdepth × SIvelocity × SIsubstrate ×...SIn)l/n,
where
| SIn | is the suitability index value for variable n, and |
| n | is the number of input variables (Waddle, 2001). |
Calculating the CI based on the geometric mean allows for more compensatory relations among variables than an arithmetic mean (J. Henriksen, U.S. Geological Survey, oral commun., 2003). For example, if two of three individual composite suitabilities are high (close to 1.0) and the third is low, the third individual composite suitability has a reduced effect on CI computation. The resulting CI value, combined with the surface area measured for various discharge scenarios, represents the weighted suitability, where a value of 1.0 indicates maximum habitat for the target species and life stage. The WUA is the sum of the products of CI values and surface area for all transect cells representing the study area.
Mean column velocities (0.6 ft of the depth) and default settings were used to compute SI scores for all species and life stages, except bull trout. Nose velocity settings were used for adult bull trout as recommended by EA Engineering, Science, and Technology, Inc. (1991b). Specific settings for nose velocity consisted of estimates of Manning’s n, which ranged from 0.04 to 0.06 for the study sites, 0.2-ft depth from the stream bottom, and use of a power law to calculate nose velocity from mean column velocity (Waddle, 2001).
For adult passage, the minimum depth criterion must be present greater than 25 percent of the total stream width and contiguous greater than at least 10 percent of the stream width at a representative transect (Thompson, 1972). This criterion represents a minimum depth over relatively short stream distances, generally less than 20 ft (Arthaud and others, 2001). The minimum depth criterion recommended by Thompson (1972) is 0.8 ft for Chinook salmon. According to SNRA biologists, this criterion is too high for marginally acceptable anadromous adult fish passage in the upper Salmon River Basin (Scott and others, 1981). Therefore, a 0.6-ft depth criterion (Scott and others, 1981) was used in this study to assess anadromous fish passage. Shallower water depths can allow passage. On August 15, 2002, adult Chinook salmon were observed moving through a shallow riffle that was 0.2-ft deep on Valley Creek. Depths that would provide marginal adult Chinook passage also would meet the passage requirements for other adult and juvenile fish.
A hydraulic parameter option in PHABSIM called AVDEPTH/AVPERM was used to characterize the hydraulic properties of each passage transect (Waddle, 2001). Stream depth criteria between 0.4 and 0.8 ft were used to evaluate the stream width available for passage at the simulated discharges for each transect. Simulated discharge results graphically display the relation between discharge and the specified depth criteria over stream width.
Stream temperature data were inspected for obvious errors such as data logger malfunction and exposure to air temperatures. Data collected prior to deployment and after retrieval were removed from the data set. Time-series plots and other graphical displays were used to inspect the data and to compare data sets. Temperature metrics, which characterize the thermal regime of stream temperatures, were calculated for all data sets and consisted of MDAT, MDMT, MWMT, and maximum weekly-average (7-day) temperature (MWAT). Maximum 7-day metrics were derived from the 7-day moving average of daily (maximum or average) temperatures.
To ensure that stream temperatures stay within the optimal range, State and Federal regulatory agencies have established stream temperature standards. IDEQ is tasked with establishing and enforcing water-quality standards, which include stream temperature criteria. In the early 1990s, the IDEQ established stream temperature criteria of 22.0ºC MDMT and 19.0ºC MDAT for the protection of coldwater biota, and 13.0ºC MDMT and 9.0ºC MDAT for the protection of salmonid spawning (Grafe and others, 2002). In addition to the Idaho water-quality standard stream temperature criteria, the USEPA imposed a site-specific rule on water bodies where they considered bull trout likely are present (40 CFR 131.E.1.i.d, 1997). This rule set a criterion of 10.0ºC MWMT during June through September for protection of bull trout spawning and juvenile rearing in natal streams.
Although these stream temperature criteria have been established, a single stream temperature criterion for all streams may not accommodate the natural temperature variation in and among streams or the existence of naturally warm water. Consequently, temperatures in Idaho streams commonly exceed the criteria (Essig, 1998; Maret and others, 2001; Donato, 2002; Ott and Maret, 2003; Maret and others, 2004, 2005).
The study results presented in this report summarize the hydrology, habitat, and temperature characteristics of each stream in the study area. PHABSIM, the primary analysis tool used, provides WUA output in relation to discharge for target species and life stages. WUA is thought to be proportional to habitat availability (Bovee and others, 1998). This output can be illustrated with a series of graphs showing curves for each life stage for the fish species of interest. The highest point on each curve represents the discharge at which WUA is maximized for adult or spawning life stages. These maximum values rarely coincide among life stages for any one species or for several species. Furthermore, the habitat/discharge relation does not address water availability. Even natural unregulated flow may not provide the discharge approaching the maximum WUA or water depth sufficient for adult passage. Also, WUA-discharge curves can be used to estimate how much habitat is gained or lost with incremental streamflow changes. In some cases, small streamflow changes can result in major habitat changes. WUA is an instantaneous representation of how much water it takes to create a certain amount of habitat. Seasonal, monthly, or daily streamflow regimes have to be applied to the instantaneous WUA curves to estimate how much habitat is actually present. The amount of WUA lost or gained can be determined by comparison with a reference, or unregulated, streamflow condition. Maximum, percentiles, or inflections typically are selected from these curves at the protection level desired or at points above which greater flow amounts provide only minor gains in usable habitat. In streams with more than one species of interest, study results should be reviewed to ensure that recommended flows are beneficial to all species and harmful to none.
Discharge/depth relations for adult fish passage were evaluated at each study site at selected transects across wide, shallow areas. These areas were identified during the stream mesohabitat typing phase and represent potential passage barriers or “bottlenecks.” If available, results from multiple passage transects can be averaged to represent overall passage conditions and streamflow needs for a particular stream segment. Relative percentage of mesohabitat types representing selected passage transects can be used to approximate the amount of potential passage habitat in various stream segments. This information may help identify those streams that have a relatively large amount of wide, shallow habitat that may restrict adult fish passage. Passage transects not representative of mesohabitats and (or) not perpendicular to the streamflow were not included in PHABSIM habitat modeling.
The mechanisms by which the various components are integrated and the relative importance they are assigned in the water-management decision process is a matter of professional judgment and beyond the scope of this study. Failure to provide adult fish passages connecting to the Salmon River would preclude success of improved conditions for spawning; therefore ensuring enough water for adult fish passage would be foremost in management priorities. Water depth for adult passage is an additional consideration for the adult life stage. If possible, target flows should not reduce the water depth less than that required for adult fish passage. In addition, providing streamflow for optimum protection of riffle habitat will ensure healthy invertebrate communities, which are a major food source for fish.
Discharge estimates providing maximum WUA for juvenile salmonid life stages are usually less than summer base flows, indicating a disconnect between the PHABSIM model simulation results and actual juvenile salmonid needs (Maret and others, 2004). PHABSIM studies on streams in Washington demonstrated that streamflows estimated to produce maximum WUA for juvenile Coho salmon (Oncorhynchus kisutch) were less than streamflows determined to actually increase juvenile recruitment (H. Beecher, Washington Department of Fish and Wildlife, oral commun., 2004). When estimated flow for maximum juvenile WUA is less than estimated unimpaired summer base flow, the unimpaired summer base flow should be considered optimum until stream‑reach‑specific fish population and streamflow relations can be obtained (J. Morrow, National Oceanic and Atmospheric Administration, written commun., 2004).
Reasons for the apparent disparity between juvenile WUA curves and actual fish population and flow relations may include: inability to accurately measure and (or) quantify habitat parameters such as velocity, cover (including escape cover), and substrate at a scale that is meaningful for small fish; inability to accurately quantify side channels, bank indentations, riparian wetlands, or other lateral habitat essential for rearing juvenile salmonids; inability to adequately incorporate temperature or other water-quality parameters into the model; and use of habitat suitability criteria that do not consider importance of high-velocity water in adjacent cells. Hampton (1988) determined that water velocity is the critical hydraulic parameter that determines microhabitat selection for juvenile Chinook salmon and steelhead trout. For example, juvenile Chinook salmon are strongly associated with pool habitat with little or no velocities (Hillman and others, 1987; Roper and others, 1994). However, stream salmonids have been observed to reside in, and forage from, shielded microhabitat locations, but adjacent to high-velocity water (Everest and Chapman, 1972). Likewise, foraging models that address improved foraging conditions associated with high-velocity flow near cover are correlated with growth and survival of juvenile Atlantic salmon (Salmo salar) (Nislow and others, 2004). Accurately modeling WUA for juvenile stream salmonids may require using habitat suitability criteria developed from foraging models (Baker and Coon, 1997) and (or) more comprehensive habitat parameter modeling.
To focus integration of the various modeling results and relevant species and life stages, a priority species and life stage ranking approach should be developed for each stream and period of concern. For example, the USFS prioritized ESA-listed anadromous species with the highest ranking, followed by Species of Special Concern, in their adjudication of water right claims for selected streams in central Idaho (Hardy, 1997). Prioritizing life stages present for the month or period of concern would benefit the target flow selection using the assumption that the priority life stage would require higher streamflows than other life stages. This priority ranking generally would be (from high to low) for small tributary streams of the upper Salmon River Basin: passage > spawning > adult > juvenile. The ranking approach should involve discussions among resource-management agency representatives familiar with the streams of interest (J. Spinazola, Bureau of Reclamation, written commun., 2005). Once the priority species and life stage are ranked, each study site should be examined to determine streamflow and passage conditions for the period of interest. Results from PHABSIM provide a science-based linkage between biology and river hydraulics; however, no one single answer can be determined from this approach.
Habitat results are presented for each target species and life stage over an incremental range of discharges, allowing flexibility in interpretation. Because the streams studied are relatively small tributaries (basin size <80 mi2 ) to the Salmon River, a greater discharge proportion is required to provide suitable water depths for fish habitat and connectivity for passage than larger streams (Hatfield and Bruce, 2000). Once an adequate number of sites have been characterized using PHABSIM, it may be feasible to develop habitat/discharge relations for streams with similar basin characteristics in specific geographic locations. This could provide a regional planning tool that could eliminate intensive, site-specific studies.
The natural hydrograph also needs to be considered when developing flow targets. In drought years, summer flows that provide maximum possible habitat may not be attainable because of the hydrologic limits on the stream. In addition, PHABSIM does not estimate flow or downstream migrants’ habitat needs or spring runoff conditions necessary for channel morphology maintenance or riparian zone functions. Arthaud and others (2001) have shown that downstream migrant survival can increase significantly with discharge; therefore, high spring flows that mimic the natural hydrograph should be considered in managing streamflows outside PHABSIM analysis.
U.S.
Department of the Interior | U.S. Geological
Survey
Persistent URL: https://pubs.water.usgs.gov/sir20065230
Page Contact Information: Publications Team
Page Last Modified: Thursday, 01-Dec-2016 19:18:32 EST
