5
Instream Flow Study

Results from the Natural Flow Study (NFS) reviewed in Chapter 4 were provided for the Klamath River basin Instream Flow Study (IFS) conducted by Hardy et al. (2006a). Hardy et al. subsequently produced recommendations for instream flows at the U.S. Geological Survey (USGS) stream gauge below Iron Gate Dam by conducting an elaborate series of investigations and model-building efforts, which are reviewed in this chapter. The following pages address the general technical elements of an IFS, and describe the procedures followed by Hardy et al. and the committee’s evaluation of those procedures. The chapter continues with an examination of the implications for implementing the instream flow recommendations, followed by brief comments regarding the larger context of those recommendations. The chapter closes with conclusions and recommendations.

TECHNICAL ELEMENTS OF AN INSTREAM FLOW STUDY

Instream flow is simply the water flowing in a stream channel (Annear et al. 2002). This simple concept belies the difficulty in determining what is the most appropriate instream flow regime when considering competing uses for water, such as irrigation, public supply, recreation, hydropower, and aquatic habitat; and the consequences of flow-level changes across seasons and years. Since natural-resource managers must make defensible decisions that balance competing demands for water, it is critical that appropriate and well-documented methods are used to quantify instream flow needs (Annear et al. 2002).

Instream flow programs involve technical and nontechnical compo-



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 135
5 Instream Flow Study Results from the Natural Flow Study (NFS) reviewed in Chapter 4 were provided for the Klamath River basin Instream Flow Study (IFS) conducted by Hardy et al. (2006a). Hardy et al. subsequently produced recommenda- tions for instream flows at the U.S. Geological Survey (USGS) stream gauge below Iron Gate Dam by conducting an elaborate series of investigations and model-building efforts, which are reviewed in this chapter. The fol- lowing pages address the general technical elements of an IFS, and describe the procedures followed by Hardy et al. and the committee’s evaluation of those procedures. The chapter continues with an examination of the im- plications for implementing the instream flow recommendations, followed by brief comments regarding the larger context of those recommendations. The chapter closes with conclusions and recommendations. TECHNICAL ELEMENTS OF AN INSTREAM FLOW STuDY Instream flow is simply the water flowing in a stream channel (Annear et al. 2002). This simple concept belies the difficulty in determining what is the most appropriate instream flow regime when considering competing uses for water, such as irrigation, public supply, recreation, hydropower, and aquatic habitat; and the consequences of flow-level changes across seasons and years. Since natural-resource managers must make defensible decisions that balance competing demands for water, it is critical that ap- propriate and well-documented methods are used to quantify instream flow needs (Annear et al. 2002). Instream flow programs involve technical and nontechnical compo- 13

OCR for page 135
136 HYDROLOGY, ECOLOGY, AND FISHES OF THE KLAMATH RIVER BASIN nents for developing and negotiating acceptable flows. Within a particular IFS, major technical elements include hydrology and hydraulics, geomor- phology and physical processes, aquatic resource biology, and water quality (Annear et al. 2004). Each of these technical elements may involve an inde- pendent study, though integration of these elements must occur to address connectivity, scaling, integration, quality assurance and quality control, and model testing. Nontechnical components of an instream flow program include legal, regulatory, and public-participation issues that are unique to a particular study. The following subsections provide overviews of the major elements of an instream flow program (NRC 2005a). Hydrology and Hydraulics The “natural flow regime” of a river is the characteristic pattern of flow quantity, timing, rate of change of hydrologic conditions, and vari- ability across time scales of hours, days, seasons, years, and multiple years, all without the influence of human activities (Poff et al. 1997). The natural flow regime in general has four components (not all of these necessarily occur in every river or even in every reach of a river) (NRC 2005a): subsis- tence flows, base flows, high-flow pulses, and overbank flows (Figure 5-1). 5-1.eps FIGURE 5-1 Illustration depicting flow regime components (subsistence flow, base fixed image—rules < 0.5 pt flow, overbank, high-flow pulse). SOURCE: Hardy et al. 2006a. Reprinted with permission; copyright 2006, Utah State University.

OCR for page 135
13 INSTREAM FLOW STUDY Subsistence flow is the minimum stream flow needed during critical drought periods to maintain tolerable water-quality conditions and to provide mini- mal aquatic habitat space for the survival of aquatic organisms. Base flow is the “normal” flow condition between storms. Base flow sustains habitat that supports diverse, native aquatic communities. Base flow also maintains the groundwater level that supports riparian vegetation. High-flow pulses are short-duration flows confined to the stream channel and occur during or immediately after storms. High-flow pulses flush fine-sediment deposits and waste products from the system, restore normal water quality follow- ing prolonged low flows, and provide longitudinal connectivity for species movement. Overbank flow is an infrequent, high-flow event that breaches riverbanks. Overbank flows may restructure the channel and floodplain, recharge groundwater tables, deliver nutrients to riparian vegetation, and connect the channel with floodplain habitats that provide additional food and space for aquatic organisms. In contrast to the once popular convention of developing a single, minimum flow or “flat-line” flow, current instream flow science advocates including all flow components in an instream flow recommendation. The many flow components, including maxima, minima, duration, frequency, and timing, are important to managers because various species of interest use habitats that are defined by these flow characteristics, and the best way to gain understanding of desirable attributes of flows is to understand “natural” flows that once supported useful, multi-species habitat. Table 5-1 demonstrates how the four components of a flow regime affect the major technical elements of an instream flow program (geomor- phology and physical processes, biological processes, and water quality). Stream flow has been described as the “master variable” (Poff et al. 1997), because the major technical elements depend on it and because it indirectly controls aspects of the physical, chemical, and biological environment, such as water temperature, hydraulic conditions, habitat, nutrient concentra- tions, aquatic vegetation, and connectivity. The goal of a hydrologic evaluation in an IFS is to understand and quantify those processes that affect stream flow quantity. This evaluation includes quantifying the magnitude, frequency, timing, and duration of the four flow-regime components; descriptive aspects of the hydrologic system, such as the location of springs, tributaries, and dams; and the impacts of land and water use on the flow regime (NRC 2004a). The purpose of hydraulic modeling is to define stream flow characteristics (for example, depths and velocities) as a function of discharge and channel and flood- plain geometry (NRC 2005a). Results from the hydrologic and hydraulic modeling, along with water-quality and fish-population modeling results (discussed later), facilitate assessing biological, water-quality, and physical processes for managing instream flows.

OCR for page 135
138 HYDROLOGY, ECOLOGY, AND FISHES OF THE KLAMATH RIVER BASIN TABLE 5-1 Flow Regime Components and Their Effect on Physical Processes, Biological Processes, and Water Quality Flow Regime Geomorphology and Component Physical Processes Biological Processes Water Quality Subsistence Increase in deposition Aquatic habitat is Dissolved oxygen flow of fine particulate restricted, vegetation decreases, temperature materials encroachment increases, establishing tolerance limits Base flow Base hydraulic Near optimal conditions for aquatic temperatures for habitat physiological processes High-flow Flushing of fine Recruitment events Increasing levels of pulse sediment, connection for water-propagating bacteria, total suspended to low-level off-channel species solids water bodies, channel maintenance, scouring pools, and uprooting vegetation Overbank Floodplain construction High connectivity Increases in total Flow and maintenance; between aquatic and suspended solids and connection to off- floodplain systems, sediment loads channel water yielding biotic bodies; bar building, exchanges between channel migration, channel and floodplain and alterations; and refuge from high large woody debris in-channel velocities recruitment and transport SOURCE: Adapted from NRC 2005a. One principle of an effective instream flow prescription is to mimic, to the extent possible, the processes characteristic of the natural flow regime (Annear et al. 2002). Meeting this objective requires data for evaluating historical and post-development stream flows, water budgets, and the like. Results from the data evaluation should serve to identify those aspects of the present conditions that must be preserved or enhanced. If there are no suitable stream flow data or if a sufficient period of record is not available, synthetic data generation may be required (Bovee et al. 1998, Wurbs and Sisson 1999). Hydrologic simulation models (for example, HEC-HMS, WMS) use information on watershed characteristics, precipitation, and run- off patterns to synthesize or extend stream flow records or create synthetic ones. If stream flow data are available from gauges in the region, runoff patterns for the watershed of interest can be synthesized by establishing statistical relationships with similar watersheds.

OCR for page 135
13 INSTREAM FLOW STUDY Poff et al. (1997) pointed out that the natural flow regime of virtually all rivers is inherently variable and this variability is critical to ecosystem functioning and native biodiversity. Year-to-year variation in flow regimes drives important physical and biological processes that periodically reset geomorphic conditions; temperature patterns across seasons; and important biological processes, such as fish-egg maturation, incubation, and growth rates characteristic of good and poor year classes (Trush et al. 2000). Therefore, to ensure sustained biological diversity and a functioning dy- namic ecosystem, both inter- and intra-annual flow regimes that attain the critical threshold levels necessary to drive important ecological processes must be maintained or provided through managed flow releases (Annear et al. 2004). Some aquatic species thrive during high-flow water years, while other species do well during years of drought. Generalist species flourish under wide-ranging flow conditions. High flows route coarse sediments, build bars, erode banks, flush fine sediments, scour vegetation, undercut and topple large woody riparian vegetation, all of which are necessary aspects of dynamic rivers that characterize the coastal salmon-rearing streams of the western United States. Typically, anadromous salmonids have successful year-classes during normal to below-normal water years when flow condi- tions are relatively steady during the spawning, incubation, and fry-rearing seasons. The most favorable habitat conditions usually develop the year after wet years, which scour pools, recruit large woody debris, flush fine sediments, and build bars. Geomorphology and Physical Processes Physical processes form and maintain the shape of the stream channel and floodplain (the strip of land that sometimes borders a stream chan- nel and that is normally inundated during seasonal floods, Bridge 2003). The form of a river channel results from interactions among discharge, sediment supply, sediment size, channel width, depth, velocity, slope, and roughness of channel materials (Leopold et al. 1964, Knighton 1998). Sedi- ment transport and deposition also shape the floodplain and riparian zone. Stream channels react to changes in sediment dynamics and either degrade or aggrade along the longitudinal gradient in response to sediment load. Channel form provides the physical structure for habitat for aquatic organ- isms. Human modifications, such as channelization, bank fortification, and reduction of coarse-sediment load due to dam installation influence the channel form and resulting habitat. Instream flow technical evaluations of physical processes are useful in documenting changes in channel structure, aquatic habitat composition, riparian vegetation, and other effects of the flow dynamics in river systems.

OCR for page 135
140 HYDROLOGY, ECOLOGY, AND FISHES OF THE KLAMATH RIVER BASIN Biology Instream flow studies historically were single-species oriented, and focused on periodicity of instream life history, physical-habitat suitability, water-quality tolerances, and temperature effects on reproduction, growth, and physiology. This approach led to flow prescriptions that ignored the needs of other fishes and other organisms such as benthic macroinverte- brates, aquatic macrophytes, and riparian species that are dependent on riverine processes (Annear et al. 2004). Given the connectivity among the elements of riverine systems (Vannote et al. 1980), a restricted species- habitat approach could have harmful effects within the very systems the instream flow advocate is trying to protect. Later advances in addressing fish-community needs led to the development of a guild approach to de- termine the physical-habitat needs of groups of fishes or invertebrates that use similar habitat types such as slow-moving pools or high-gradient riffles (Leonard and Orth 1988, Lobb and Orth 1991). Recently, instream flow prescriptions have addressed more- comprehensive aspects of riverine system ecology by attempting to meet the flow requirements of the entire aquatic community and the associated terrestrial (that is, riparian) community (Moyle et al. 1998; Annear et al. 2004; NRC 2005a). These holistic efforts are often hampered by limited data on habitat requirements, biology, and life history of aquatic organisms including federally listed endangered species and nongame fishes (Myrick and Cech 2000; Moyle 2002), and by the limited, but growing understand- ing of the intricate connections within the biological community of a river ecosystem. Hydraulic habitat (that is, flow depth, velocity, substrate, and instream cover components of a stream) is a key component of any instream flow prescription, but providing hydraulic habitat alone will not guarantee any particular state of the aquatic ecosystem (Annear et al. 2004, NRC 2005a). Appropriate physical habitat is necessary, but not sufficient on its own. The dynamic effects of varying levels of hydraulic habitat on biological processes, including competition, bioenergetics, predation, disease, and the recruitment of juveniles into the population, must be considered (Bartholow et al. 1993, Annear et al. 2004, NRC 2005a). Ecological and biological processes occur over variable scales of time and space, so an instream flow prescription should provide an appropriate level of spatial and temporal variability, to preserve the complexity of these processes. Water Quality Historically, the water-quality component of instream flow prescrip- tions was based on modeling efforts to ensure that water-quality standards

OCR for page 135
141 INSTREAM FLOW STUDY (for example, dissolved oxygen, temperature, and nutrient and contaminant concentrations) were not violated. This focus on water-quality standards alone, including the use of 7Q10 to establish minimum flows, often resulted in minimum flow prescriptions that ignored the physiological needs of fishes and other organisms, such as benthic macroinvertebrates, aquatic macrophytes, and riparian species, which also are dependent on riverine processes (Annear et al. 2004). Comprehensive methods address seasonal requirements for successful spawning, incubation and growth of important species of fish, or sometimes a group of fishes such as salmonids (Bovee et al. 1998, NRC 2005a). Issues of spatial and temporal scale and inter- and intra-annual vari- ability also are relevant to the water-quality component of an instream flow prescription. The influence of water quality on ecological and biological processes occurs at various scales, and responds differently to the flow regimes of wet and dry years. For example, degree-day accumulation deter- mines the timing and location of spawning, egg maturation, and the dura- tion of incubation, fry emergence, and growth for most riverine fishes. The flows specified by an instream flow prescription exert a significant influence on the temperature regime and water quality within the system, influencing directly or indirectly the biological, physical, and chemical characteristics (Annear et al. 2004, NRC 2005a). Ignoring any of these effects is fraught with risk, because the combined effects of the various components of water quality will influence the presence, abundance, and distribution of biota. The primary water-quality parameters considered by instream flow studies are sediment, total dissolved solids, dissolved oxygen, water temperature, pH, contaminants, and nutrients, but analyses focus on water temperature, contaminants, and sediment (Annear et al. 2004, NRC 2005a). Brett (1964) described water temperature as the “ecological master variable” because it affects all aspects of the biology of aquatic organisms including reproduction (Stonecypher et al. 1994), growth (Jobling 1997), susceptibility to disease (Antonio and Hedrick 1995), and migratory ability (Lee et al. 2003). Clearly, it is an important driver of riverine ecosystems. Any alteration of the flow regime also alters the temperature regime. As with the hydraulic component, the temperature is a necessary but not suffi- cient component of the habitat for successful instream flow prescriptions. A broad range of pollutants are present in riverine systems; their effects range from sublethal changes in physiological performance (Beaumont et al. 1995), to disruption of the endocrine system with subsequent changes in re- productive success (Jobling et al. 1998), to changes in community composi- tion (Hickey and Golding 2002), to acute toxic effects (Hamilton and Buhl 1990). The effects of suspended solids and sediments on stream-dwelling organisms are well documented (see Waters [1995] for a comprehensive review) and include loss of spawning habitat (Burns 1970), increased physi-

OCR for page 135
142 HYDROLOGY, ECOLOGY, AND FISHES OF THE KLAMATH RIVER BASIN ological stress (Servizi and Martens 1992), and direct mortality of fish (Servizi and Martens 1991) and other aquatic organisms. Sediments are a major pollutant in U.S. waters (EPA 1990) and should be an important con- sideration in instream flow studies. Predicted changes in the nutrient status of streams should also be considered in instream flow prescriptions. Streams of particular concern include those that are effluent-dominated or that flow through nutrient-rich landscapes. Nutrient enrichment (or impoverishment) can change the flow of energy in a river ecosystem (sometimes with positive effects; see Bilby et al. 1998), and can alter conditions sufficiently to shift the balance among competing species. Nutrient enrichment may also lead to changes in other water-quality parameters, including dissolved oxygen levels that will have additional negative impacts on river ecosystems. The development of an instream flow prescription will benefit from prior knowledge of water-quality conditions, but this information is not always available. Monitoring water quality in even river system of mod- est size is challenging because of the number of physical and chemical parameters that should be tracked, and the heterogeneous nature of those parameters, particularly when spatial and temporal variability, and the heterogeneous legal and jurisdictional framework are included (Woodling 1994). Nevertheless, because of the significant effects flows can have on water quality, instream flow prescriptions should attempt to address the key aspects of changes in water temperature, dissolved oxygen, nutrient concentrations, sediment loads, and contaminants. Connectivity Historically, low-flow fish passage for migratory species (for example, anadromous salmon and trout) was frequently the only consideration of biological connectivity when developing instream flow prescriptions. The temptation is strong to simplify riverine ecosystems as unidirectional, two- or three-dimensional systems. For some exploratory modeling exercises, this degree of simplification may be appropriate. However, such simplifica- tions do not incorporate connectivity in lateral, longitudinal, vertical, and temporal dimensions in a river system (Ward 1989). Connectivity is “the flow, exchange, and pathways that move organisms, energy, and matter through these [river] systems” (Annear et al. 2004). The presence of physi- cal, chemical, or biological barriers degrades the connectivity and function- ing of rivers. Common connections include nutrient and energy flow from headwaters to downstream (Vannote et al. 1980) and connections between surface flow and groundwater. Connections are severed by dams and cul- verts (Helfrich et al. 1999, Schlosser and Angermeier 1995), and by changes in flow regimes or water quality (Cherry et al. 1978). These connections should be considered in an instream flow evaluation to ensure that con-

OCR for page 135
143 INSTREAM FLOW STUDY nectivity is improved or not further degraded (for example, by prescribed releases of poor-quality water that effectively impedes upstream movements of fish and invertebrates). Spatial and temporal scale and intra- and inter-annual variability need to be considered in an assessment of connectivity (Kondolf et al. 2006). Biological connectivity, such as unimpeded upstream passage and access to the floodplain for fishes, does not have to be permanent, but needs to exist during critical phases of fish life history, such as spawning migrations or juvenile rearing (for example, Maslin et al. 1997). Furthermore, the natural flow regime may have included periods of intermittent flow that provided or created off-channel habitat for locally adapted species (Labbe and Fausch 2000); these periods of intermittent flow (and the challenging physical conditions that result) may also afford native species an advantage over invasive nonnative fishes. Scaling Physical, chemical, and biological processes affect stream ecosystems at different spatial and temporal scales (a more general discussion of scaling is in Chapter 3). The importance of spatial and temporal scales and the study of flow, temperature, and habitat time series during instream flow studies has been recognized for more than three decades, but seldom has been adequately incorporated into instream flow prescriptions (Bartholow and Waddle 1986, Milhous et al. 1990). Spatial scaling issues, such as specifying what length of a river must be studied, how study reaches are selected, and how data from study areas are extrapolated to unstudied areas, remain a major research focus for instream flow science, and effective methods for reconciling different scales are not well documented. Spatial Scaling Habitat-suitability modeling integrates results from hydraulic simu- lations and fish-distribution modeling. A basic assumption in habitat- suitability modeling is that the hydraulic variables (flow depth and velocity) and structural elements (bathymetry, substrate size, and cover distribution) are uniform within each of the simulated small areas within the gridded design for sampling stream reach. These simulated small grids are often referred to a cells or meshes by the modelers. Consequently, the spatial resolution is a critical element and must be compatible between the two modeling efforts. Often, the cell size determined from field measurements and used for fish-habitat simulations is selected to facilitate hydraulic mod- eling and calibration. This is especially true for one-dimensional hydraulic modeling when some analysts survey a single transect over individual

OCR for page 135
144 HYDROLOGY, ECOLOGY, AND FISHES OF THE KLAMATH RIVER BASIN mesohabitat types (for example, riffle, pool, and run). Failure to consider the spatial scale of fish use and the field-collection techniques used in devel- oping habitat-suitability criteria can result in widely spaced transects (used with one-dimensional hydraulic modeling) yielding rectangular cells many meters in length. Such excessively large cells violate the assumption of cell homogeneity. For fish, an appropriate cell size often is approximately a few square meters (depending the size of the fish and their degree of territorial- ity and movement), which requires that transects be placed close together (Bovee 1982, Bovee et al. 1998). Two-dimensional hydraulic modeling can more easily overcome this limitation when detailed bathymetry of the stream channel is known. The only assurance that model cell size is com- patible is by comparing habitat-model output (habitat suitability by cell) with independent field observations of fish distribution. Statistical tests for acceptance or rejection of model output should be made before using these types of models as input to habitat time series and fish-population model- ing. In addition, as described in Chapter 3, the scale for habitat-suitability criteria should be the same as that for hydraulic modeling. Temporal Scaling Time-series modeling of habitat and water quality facilitate the evalu- ation of the suitability of environmental conditions for supporting the completion of fish life stages and yield input to fish-population modeling. These models require weekly time steps at a minimum and ideally would use daily time steps. When rapid flow fluctuations may occur, as under hydropower peaking operations, hourly time steps are necessary. Integration Integration is the process of combining different technical studies into flow recommendations (NRC 2005a) at specified points within the river system. Integration requires that evaluations of hydrologic, hydraulic, geomorphologic, biological, and physical processes and water quality be compatible and at commensurate or complementary spatial and temporal scales. Integration of study results and model linkages is an important, complicated step, and although integration methods are being generated empirically, they are not well documented. There are few widely recog- nized general procedures for integration, and thus methods must be evalu- ated without the benefit of direct, observable, or historical evidence. The Instream Flow Incremental Methodology (IFIM) (Stalnaker et al. 1995) promotes integration by providing guidance for combining microhabitat (hydraulic habitat) with macrohabitat conditions (water quality) over lon- ger stream lengths through a river system (Bartholow and Waddle 1986,

OCR for page 135
14 INSTREAM FLOW STUDY Bartholow 1989, Bovee et al. 1998). The USGS SALMOD model integrates hydrologic processes, hydraulic-habitat carrying capacity, and degree-days as the principal drivers of a salmon population-dynamics model (Bartholow et al. 1993). The integration process is not achieved by simply prescribing flows that mimic the shape of the natural flow regime. Appropriate integra- tion methods link instream flow recommendations to desirable outcomes using scientifically defensible methods. An example of such integration and linkage can be found in the Trinity River Flow Evaluation report (US- FWS/HVT 1999), considered a founding document for the Trinity River Restoration Program. Quality Assurance and Quality Control As described in Chapter 2, quality assurance and quality control (QA/ QC) for ecosystem modeling requires a modeling plan, adherence to proce- dures, documentation, and model testing. Given the importance of evaluat- ing alternative management schemes, rigorous QA/QC protocols, including peer review, should be established, documented, and enforced. Also, if legal challenges are likely, consideration should be given to establishing and documenting model and data versions, how data were processed and modified when necessary, and the custody of information during the model- ing process. Model Testing Model testing refers to procedures used to evaluate the performance of predictive models (Chapter 3). Instream flow studies inevitably utilize a multitude of models including hydraulic (for predicting flow depth and velocity), water quality, water temperature, habitat suitability (for various species and life stages), water routing, and reservoir operation models. Model testing requires that the model produce testable predictions and that independent data are available for comparing model output to obser- vations. Model test conditions should span the entire range of conditions for which the model will be applied. In the case of hydraulic models, for example, comparison between observed and predicted velocities is often tested for flow within the channel banks, and comparisons are seldom made for overbank flows. Because a comprehensive IFS must examine overbank conditions, hydraulic models should be tested under those conditions. Habitat models rarely are tested against observed fish distributions in a modeled stream reach. Thus, the use of hydraulic-habitat-suitability models has become quite controversial and has been criticized. Testing of model output is essential to establishing that a model is reliable. Box 5-1 provides the committee’s recommendations for improving.

OCR for page 135
5-12.eps fixed image set broadside because rules reduce to < 5-pt at portrait view FIGURE 5-12 Comparison of modified Table 9 flows (NMFS 2002, USGS 2005) with Hardy et al. (2006a) instream flow recom- mendations. Bars represent modified Table 9 flows, as specified for each water-year type. Lines represent recommended instream flows corresponding to annual exceedance values from 5% to 95%. 18 SOURCE: Hardy et al. 2006a. Reprinted with permission; copyright 2006, Utah State University.

OCR for page 135
186 HYDROLOGY, ECOLOGY, AND FISHES OF THE KLAMATH RIVER BASIN FIGURE 5-13 Relationship between5-13.eps October-March precipitation and naturalized flows at Iron Gate Dam during April-September period.type fixed image—bitmapped Vertical lines correspond to annual exceedance levels for precipitation derived from the equivalent natural- could be improved ized flow values. COMpREHENSIVE ANALYSES AND INTEGRATION Hardy et al. (2006a) provide several important initial steps (including novel and valuable data) toward a comprehensive Klamath River basin management program. The temperature and hydraulic modeling compo- nents have been accomplished to the degree necessary for the main-stem Klamath River. Similar efforts are ongoing for the Trinity River. However, tributaries to the Klamath and Trinity rivers, which are important to the life history of coho salmon, have not been quantitatively addressed. As is true of all instream flow studies, a comprehensive hydrologic description of the river system (including tributaries) is fundamental to system under- standing and management. The existing hydrologic data are not sufficient for the level of analysis that is necessary. An understanding of the relations

OCR for page 135
18 INSTREAM FLOW STUDY among temporal and spatial variations in hydrologic conditions and the resulting habitat, temperature, and fish-population dynamics is essential for developing management prescriptions leading to sustained recovery of the anadromous species. Adequate management of Klamath River salmonids requires a net- work-level hydrologic model capable of generating daily time series of stream discharge at numerous points throughout the river system (includ- ing tributaries). This level of modeling is essential for using the spatial and temporal dynamics of the temperature regime and habitat modeling to elucidate possible “habitat bottlenecks” and as input to appropriate fish population-dynamics models (species-specific SALMOD). Comparing “synthesized natural and existing” hydrologic conditions and subsequent development of instream flow prescriptions for recovery of the salmonid species is only appropriate if the various models use the same time step. Such information is critical for recommending instream flows for a Klam- ath River basin water-management program. Unfortunately, there is little or no synergy between the two reports the committee was charged to re- view. The NFS presents monthly data below Link River Dam, whereas the temperature model used by Hardy et al. (2006a) provides daily data down the main-stem Klamath River, and the habitat and fish-population analyses used a monthly time step and SIAM model data. A second point of concern is the apparent lack of emphasis on the sedi- ment dynamics of the system. Changes in management of a river system require study of the sediment input (or lack thereof) and sediment transport and discharge relations. Any river system with dams and altered discharges will have changes in the sediment balance and perhaps alterations in the channel form and sediment distribution. Valuable data from such studies could address the present state of the river channel (aggrading, degrading, or in some stage of dynamic equilibrium) and better inform management decisions on maintaining the existing channel, if in a state of dynamic equi- librium, or driving the system toward a new equilibrium state. Hardy et al. (2006a) do not discuss this issue, although reference to the Ayers report (Ayers Associates 1999) implies an assumption that the present channel is not actively aggrading or degrading. The roles of overbank flows (channel forming, recruitment of nutrients, and fish refuge) and sediment-flushing flows are important components of a comprehensive IFS and should re- sult from sediment-transport and channel-dynamics studies. Hardy et al. (2006a) assumed that these flows would be present, but the necessary flows for maintaining channel dynamics were not quantified. Without such quan- tification, there may be no protection from future flow depletions resulting from diversion and storage of peak flows. Another apparent impediment to the integration of science and man- agement strategies is the jurisdictional separation of the Klamath River

OCR for page 135
188 HYDROLOGY, ECOLOGY, AND FISHES OF THE KLAMATH RIVER BASIN and Trinity River main stems into two separate USBR Area Management Offices. A thorough understanding of salmon stocks native to the whole Klamath River system and of how management of the water supply may help to achieve their recovery will require system-wide analyses and joint management. Following the mass mortality of adult salmon in the lower Klamath River in September 2002, there has been increased pressure to integrate efforts of the Trinity River Restoration and Klamath River Task Force. For example, summer releases from the Trinity River could pro- vide much-needed cooler water in the lower Klamath River, but if done routinely each year, summer releases might conflict with Trinity River adaptive-management objectives for attaining appropriate variability within inter-annual flow regimes. Integration of these two efforts is essential. The Trinity River Restoration Program is in the process of collecting habitat, temperature, and fish-population data similar to those presented by Hardy et al. (2006a), and a data-management system is being developed for use by both the Klamath River and Trinity River programs (R. Wittler, USBR, personal communication 2006). To fully integrate these two programs, the same models (such as flow, temperature, habitat, and fish-population dy- namics) with the same level of detail, time step, and linkage among models should be used on both sub-systems. With similar integrated modeling and data management, alternative management scenarios (involving reservoir releases, sediment augmentation, and so forth), using an assumed water supply, climate conditions, and salmon stock returns, could be quickly evaluated. Such linkage of models could greatly facilitate communication among various stakeholders and managers and lead to better adaptive-man- agement approaches. Integration of these efforts with ongoing activities in the Shasta and Salmon rivers also would be beneficial. CONCLuSIONS AND RECOMMENDATIONS Having reviewed the IFS (Hardy et al. 2006a) the committee finds that it enhances understanding of the Klamath River basin ecosystem and the flows required to sustain it. However, the flow recommendations were based on manipulations of the PHABSIM modeling and supplemented by existing flow records in ways that did not clearly derive from any theoreti- cal considerations. The steps used to derive the final flow recommendations were not pre-specified or thoroughly tested. In its present form, to the degree they are adopted, the recommended flows resulting from the study should be adopted on an interim basis pend- ing the model improvements outlined below and a more integrated assess- ment of the scientific needs of the basin as a whole. The recommended flow regimes offer improvements over existing monthly flows in that they include intra- and inter-annual variations and appear likely to enhance Chinook

OCR for page 135
18 INSTREAM FLOW STUDY salmon growth and young-of-the-year production. More detailed discus- sion of the study’s implications for the basin’s fishes is in the final section of this chapter. In this report, the committee is critical of many aspects of the IFS, but the committee also found substantial strengths in the study. The following paragraphs outline the substantial strengths and contributions of the IFS that managers and investigators may find useful for decision making. The IFS represented a state-of-the-art process for the modeling of temperature and bioenergetics for riverine fish species. Temperature mod- eling is especially important for fishes in the lower Klamath River because temperature may be a limiting factor in late summer for suitable habitat. Understanding how temperature varies, particularly in response to flows that likely result from dam releases, may provide support for more effective operating rules during low-flow periods. High temperatures affect the sur- vival of fishes directly if they are high enough to kill the fish; they also can affect them indirectly by affecting bioenergetic processes. A prominent fea- ture of the study is the Chinook salmon fry and juvenile growth modeling using the Wisconsin Bioenergetics Model (Stewart and Ibarra 1991, Hanson et al. 1997). Bioenergetic analysis of fish growth often is not included in instream flow assessments. The IFS is broadly consistent with the guidelines of Instream Incremen- tal Flow Methodology (IFIM). While the committee has reservations about the use of some models (including PHABSIM) that support the connection between flow and habitats, IFIM is widely used as a general approach to connecting research associated with flows to management decisions. Gen- erally speaking, the IFIM process includes legal and institutional analysis, strategy design, technical scoping, habitat modeling, definition of alter- natives, feasibility studies, and negotiated resolutions with stakeholders (USGS 2007b). The IFS for the Klamath River fits well into this approach in the habitat modeling portion of the overall process. Incorporation of salmon fry behavior and distance to escape cover is an exceptional advancement offered by the IFS. Salmon fry are a critical part of the salmon life cycle in the river environment, and their inclusion in the model provides increased confidence in the model predictions. By examin- ing the micro-geography of the stream channel, the IFS provides the ability to examine how the distribution of fishes interacts with the distribution of potential protective cover for them, a metric that is substantially variable over short lengths of river channel. Many instream flow studies do not as- sess this distance to cover because of the stringent data requirements, but the Klamath study successfully addressed this concern. The Klamath instream flow model makes a substantial contribution to decision making with its comparison of consequences of implemented flows with existing flows for smolt production, a critical step in the overall life

OCR for page 135
10 HYDROLOGY, ECOLOGY, AND FISHES OF THE KLAMATH RIVER BASIN cycle of the anadromous salmonid fishes. Most instream flow models create a set of predictions and stop at that point, but the Klamath study goes one important increment further by comparing the existing flow regime, which might be suspected of inhibiting smolt production, and drawing quantita- tive comparisons with a proposed flow regime that might improve smolt production. The result provides managers with useful insights into factors that are important for decision making and provides scientific insights to the general salmon survival process. In another comparison between predictions and observations, the IFS assesses predicted and observed fish locations in representative reaches of the river. Generally speaking, there was considerable agreement among predictions of fish distributions through habitats in the reaches that were examined. This agreement lends some credibility to the model, and further comparisons can show under what river conditions the model is strongest and under what conditions it is weakest. The committee finds that many similar studies fail to test their output, as was done in this case. Finally, an important strength of the IFS is its state-of-the-art applica- tion of flow models in simulation of habitat suitability. Flow models have gradually evolved over the past several decades from one-dimensional rep- resentations with assessments of flow variation only in the downstream direction. While useful, such assessments did not reflect the enormous complexity of river channels, particularly relatively large channels, such as the Klamath River. Recently, hydraulic researchers and engineers have made increasing use of two-dimensional flow models that can trace variation in the cross-channel direction in addition to the downstream direction. Few instream flow studies have yet to take full advantage of the newer two- dimensional models, partly because of their demand for large amounts of data describing the topography of the bed and banks of the channel as well as nearby riparian areas and flood plains. Two-dimensional models also require extensive computing capabilities. The IFS has made good use of new techniques and extended computing capability to improve the understand- ing of the fluvial complexities of salmon habitat in the river. Specific conclusions and recommendations follow. Conclusion 5-1. The goal of the IFS is to recommend “flow regimes that will provide for the long-term protection, enhancement, and recovery of the aquatic resources within the main stem Klamath River.” This study was limited because achieving that goal requires daily flow data and the USBR NFS provided monthly flow data. Recommendation 5-1. The IFS should be updated using daily flow data. The monthly data could be disaggregated into daily data before hydrologic time series are used in habitat modeling, or a daily watershed model that

OCR for page 135
11 INSTREAM FLOW STUDY represents each of the major sub-basins of the Klamath River basin under natural conditions could be developed. Such forecasts could be used to update the annual exceedance types through the summer and fall months. Future investigations could compare the unregulated flow regime of the river with present conditions and seek to restore as much of the functional river-based ecosystem as possible through managed flows benefiting a full complement of species. Conclusion 5-2. Although the methods used for habitat assessment in the IFS extend the standard practice for assessing habitat-flow relations by in- corporating escape cover and state-of-the-art hydraulic modeling, there are limitations in assessing primarily the hydraulic aspects of habitat, indepen- dent of other river-ecosystem attributes (for example, physical processes, temperature, water quality, bioenergetics, and fish production). Therefore, more integration of the analytical models is needed. Better-integrated mod- els also would facilitate adaptive management of the river system. Recommendation 5-2. Habitat modeling should integrate hydraulic anal- yses with geomorphic processes of sediment transport (for example, sediment-flushing and channel-forming flows), water quality (for example, temperature, dissolved oxygen, and contaminants), and fish-population dynamics. In addition, an adaptive management strategy should provide a clear context of management decisions to which habitat modeling results would be relevant. Conclusion 5-3. In addition to the integration recommended above, suc- cessful maintenance of aquatic resources in the Klamath River depends on several aspects of water quality in addition to temperature conditions, including dissolved-oxygen levels, nutrient concentrations, sediment loads, and contaminants. The IFS does not present analyses of these water-quality attributes, nor does it justify excluding them from analyses. Failure to ana- lyze the thermal conditions for the recommended flows during the critical autumn and spring migrations is another shortcoming of the study. Recommendation 5-3. An addendum to Hardy et al. (2006a) should be prepared that includes analyses of several aspects of water quality, such as dissolved-oxygen levels, nutrient concentrations, sediment loads, and contaminants, or the addendum should justify excluding these important considerations. Conclusion 5-4. Tributaries of the Klamath River are important for main- taining anadromous fish populations because they provide essential habitat and are sources of water and sediment that maintain main-stem habitats.

OCR for page 135
12 HYDROLOGY, ECOLOGY, AND FISHES OF THE KLAMATH RIVER BASIN Furthermore, although coho salmon and steelhead are found in the main stem, tributaries contain the most important habitat for producing juve- niles of these species. Since technical assessments conducted as part of the IFS were confined to the main-stem Klamath, the usefulness of the study for evaluating coho salmon and steelhead management options is severely limited. Recommendation 5-4. Future studies should include explicit analyses of the habitat, water, and sediment contributions of tributary streams in the context of the fish life histories (particularly coho and steelhead) and movements throughout the entire Klamath River basin. They should assess the ability of tributaries to facilitate juvenile fish production and thermal refugia and provide estimates of tributary habitat, their areal extent, and the extent of overcrowding of fish in them. Conclusion 5-5. The IFS lacks adequate statistical testing of model pre- dictions. Specifically needed are statistical analyses of the comparison of observed fish distributions with predicted distributions of usable habitat as defined by simulated hydraulic and temperature conditions, confidence lim- its on the fish-growth predictions, and statistical testing of hydraulic-model velocity predictions for the entire range of flow conditions in the channel and overbank. These analyses would increase confidence in the validity of modeling results and conclusions. Recommendation 5-5. An addendum to Hardy et al. (2006a) should be prepared that contains results from rigorous statistical testing of model outcomes, including a comparison of observed fish distributions, predicted temperatures, and confidence limits on growth predictions. Just as habitat simulations were made for the entire range of flow conditions, statistical analyses should be presented to support velocity predictions in the channel and overbank. Conclusion 5-6. The approach used in the IFS apparently assumes that physical habitat is an important limiting factor to recovery of the salmonid fishes. However, the study does not demonstrate when (or if) habitat may be limiting to the fish species and the identification of potential life-stage “bottlenecks” when comparing existing and naturalized flow time-series simulations. Habitat-duration curves alone are not sufficient to illustrate which life stages may be most vulnerable. SALMOD modeling would be more appropriately used for analyses of potential habitat limitations im- posed by hydraulic and temperature conditions, sensitivity analyses, and comparison of salmon-smolt growth and production between naturalized and existing flow regimes.

OCR for page 135
13 INSTREAM FLOW STUDY Recommendation 5-6. Investigators should provide a thorough analysis of the life stages of salmonid species, allowing for comparisons of seasonal differences in usable habitats between naturalized and existing flow simu- lations and for identifying possible habitat limitations imposed by present conditions and potential improvements provided through recommended flow regimes. Full integration of hydrology, habitat, temperature, and Chi- nook salmon life history with SALMOD is needed. SALMOD modeling should be used for identifying habitat limitations and developing alternative instream flow recommendations, not simply for post hoc testing. Testing of salmon movement, growth, and production through ongoing monitoring efforts should accompany these modeling efforts, as is being done on the Trinity River. Such integrated and tested modeling capabilities could prove useful for future adaptive-management efforts. Conclusion 5-7. Given the overall objectives of the IFS, a reasonable pro- cess and rationale were used to stratify the main stem Klamath River into five “homogeneous” study reaches and to identify seven representative study sites. However, the representativeness of the study sites was deter- mined by inter-agency group agreement and was not statistically assessed. If the mesohabitat distribution within the study sites is not shown to be representative of the larger study reaches, there will remain significant uncertainty about the efficacy of the node-weighting approach used to transform (that is, upscale) site-specific habitat-suitability criteria for the study reaches. Recommendation 5-7. The IFS should include an analysis that demon- strates that the study sites are representative of the respective study reaches by providing field measurements of basic channel properties at intervening locations among the study sites. In addition, analyses are needed to dem- onstrate the efficacy of upscaling site-level habitat modeling results to the study reaches. Conclusion 5-8. Hardy et al. (2006a) used a PARMA model to generate a set of naturalized flows for the Klamath River. There are three potential limitations in their approach. First, the PARMA model may not permit an accurate assessment of uncertainties in the naturalized flows simulated using the water-budget approach. Second, the model may not adequately capture serial (intra-annual or inter-annual) autocorrelation and spatial cor- relation in the data. Finally, investigators may have fitted the model without properly accounting for skewness in monthly flow distributions. Recommendation 5-8. The analyses required to address limitations to the PARMA model developed by Hardy et al. (2006a) should be conducted, and a formal assessment of model inaccuracies should be conducted.

OCR for page 135
14 HYDROLOGY, ECOLOGY, AND FISHES OF THE KLAMATH RIVER BASIN Conclusion 5-9. The IFS identifies five necessary instream flow compo- nents: overbank flows, high-flow pulses, base flows, subsistence flows, and ecological base flows. Among these, the discussion of overbank flows states that operation of the existing Klamath system provides large discharges during wet periods that meet or exceed the channel-maintenance require- ments, but there is no discussion of the importance of overbank flows, particularly their frequency and duration. In addition, the study defers discussion of high-flow pulses (citing ongoing testing) and assumes that the maintenance of intra- and inter-annual habitat values at or near “natural flow” accounts for base, subsistence, and ecological base flows. Recommendation 5-9. Specific high-flow recommendations for maintain- ing channel integrity are needed. Additional physical-process studies are also needed to address sediment transport and establish threshold levels of discharge that maintain channel dynamics. Results of these studies need to be incorporated into variable instream flow recommendations. MANAGEMENT IMpLICATIONS OF THE INSTREAM FLOW STuDY The basic conclusions of the IFS are recommended flows expressed as monthly target values for discharges below Iron Gate Dam on the Klamath River. The study adopted the Natural Flow Paradigm, and its primary in- put was the natural flows defined by the NFS. The IFS integrated the NFS flows and the historical flow records with water-temperature simulations to accommodate existing understanding that flow volume and water tem- perature are two primary controls on fish growth and survival. The most important outcome of the IFS was that it indicated that increases in existing flows downstream from Iron Gate Dam probably would benefit fish popula- tions through improved physical habitat associated with more water and through reduced water temperatures. The increased flows would reduce slack-water areas that are disease-prone areas for salmon. Tests conducted in the study led the authors to conclude that if the prescribed flows had defined the river’s hydrology instead of the actual regulated flows (mostly less than the prescribed flows) during the period 1949 to 2000, salmon production in the lower river would have been higher than it was under the prescribed flows. If these conclusions were borne out by studies incorporating experimen- tal flows and monitored responses, managers would be able to have greater confidence that decisions to increase flows would be likely to have a benefi- cial effect on anadromous fishes in the lower river. The authors of the IFS mention two caveats, and the committee agrees with them. First, the flow recommendations apply to the needs of the anadromous fishes in the lower

OCR for page 135
1 INSTREAM FLOW STUDY Klamath River, and they do not account for competing water demands for other purposes, such as agricultural needs or the needs of federally listed fishes in the upper basin. Second, the flow recommendations address the needs of all the anadromous species in the lower Klamath River. They are not targeted for any individual species (listed or otherwise), and it is not possible to evaluate the conclusions separately for individual species. The committee has additional caveats about the study’s results. The flow recommendations are limited to monthly values because of the nature of the input from the NFS. This characteristic implies for users that the recommendations are general in nature and that they are useful for general planning. Because there are no daily flow recommendations, and because of various limitations on the calculations outlined in Chapter 5, the recom- mended instream flows do not provide guidance for daily discharge rules for system operations. The recommended instream flows are applicable to management of all anadromous fishes in the lower Klamath River, and there is no specification by species (Hardy et al. 2006a, p. 215). Of at least equal importance, the study, like the NFS, does not address the Klamath River’s tributaries. Those tributaries are of great importance to the hydrologic re- gime in the Klamath River, especially below the confluence with the largest tributary, the Trinity River, but also above that point. To the degree that the anadromous fishes use the tributaries for spawning and rearing habitat, they are important to the productivity of those species. Finally, as the com- mittee has detailed in Chapter 5, there are sufficient uncertainties and flaws associated with the study to show that it cannot be used as a specific guide to specific flows with much confidence. Despite all the foregoing, it is extremely unlikely, in the committee’s judgment, that following the prescribed flows of the IFS Phase II would have adverse effects on any of the anadromous fish species. Based on gen- eral principles and the information developed in that study, following its prescribed flows probably would have some beneficial effects on the suite of anadromous fishes in the Klamath River considered as a whole, although not necessarily for every species. The conclusions and recommendations appearing in this and the pre- ceding chapter lead the committee to consider the NFS and the IFS of the Klamath River in a more general context. The river, its waters, species, and habitats along with the myriad of human-induced controls operate in a complex ecosystem defined geographically by the watershed. The next chapter considers these larger-scale issues as a way to better understand the validity of the science and engineering approaches to the NFS and IFS and provides a framework for integrating resource management, science, and stakeholder concerns.