Click for next page ( 86


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 85
Environmental Issues The American River basin possesses significant environmental values. The upper American River, known for its steep gradient, free-flowing white water, and relatively natural plant communities, was listed under the state's Wild and Scenic River classification system in 1978. In 1981, a 23-mile stretch of the lower American River was designated under the National Wild and Scenic River System as a sport fishery and a critical recreational resource. Sacramento owes its nickname, "River City," its identity, its sense of history, and its regional character to the American River, which flows through the city's center, and to the Sacramento River, which joins near the city's historic district. The 1991 American River Watershed Investigation (ARWI) raised a number of contentious environmental issues. A key concern was whether the description of the environmental impacts of the various proposed project alternatives was adequate. This chapter reviews how the 1991 ARWI and the associated environ- mental impact statement/environmental impact report (EIS/EIR', referred to here as the 1991 ARWI report (USAGE, Sacramento District, 1991), considered envi- ronmental impacts. It pays particular attention to the assessment of environmen- tal impacts associated with the proposed Auburn dry dam alternative and associ- ated inundation impacts in the American River canyon, but also addresses a range of other environmental considerations. In addition, the chapter discusses whether the planning process is adequate to produce plans that consider a full array of feasible alternatives that integrate social, environmental, and flood risk reduction factors. (Chapter 6 addresses planning issues in more detail.) 85

OCR for page 85
86 FLOOD RISK MANAGEMENT AND THE AMERICAN RIVER BASIN TREATMENT OF ENVIRONMENTAL ISSUES IN THE ARWI REPORT Overview of NEPA and CEQA The 1991 ARWI was prepared to meet the requirements of both the National Environmental Policy Act (NEPA) of 1969 and the California Environmental Quality Act (CEQA). NEPA Section 102 requires an environmental impact statement for "major federal actions significantly affecting the quality of the human environment." The Council on Environmental Quality (CEQ) oversees the implementation of the NEPA act. Guidelines prepared by CEQ and federal agencies and a large number of legal cases have established that the major legal issues associated with the EIS process are determining when one must be done (i.e., determining what is a major federal action) and determining the adequacy of a prepared statement with regard to the accuracy of description, the identification and quantification of probable environmental impacts, and the exploration of reasonable alternatives. CEQA passed a year later and, although based on NEPA, had some different provisions. It is implemented through a state agency, the Office of Planning and Research, and has its own guidelines for preparation of environmental impact reports. Both CEQA and NEPA guidelines urge the issu- ance of joint reports that satisfy both state and federal law (Heyman, 1974; Remy et al., 1994~. Like the federal act, CEQA was conceived primarily as a means to document and consider the environmental implications of actions. Unlike NEPA, CEQA is not merely a "procedural" statute but contains substantive provisions that agen- cies are to comply with. Under NEPA the federal government is required only to give "appropriate consideration" to environmental values and presumably can take actions causing environmental damage even if feasible and effective mitiga- tion measures could easily be implemented. In contrast, CEQA requires agencies to implement feasible mitigation measures and alternatives identified in the envi- ronmental impact reports. In addition to this important distinction, CEQA places an emphasis on any growth-inducing impacts associated with proposed actions and the potential of population growth to stress existing community service fa- cilities (Secretary for Resources, California, 1973; Remy, 1994~. Both NEPA and CEQA provide guidance on the development of alternatives in the EIS and EIR. CEQ guidelines state (Council on Environmental Quality, 1973~: A rigorous exploration and objective evaluation of the environmental impacts of all reasonable alternative actions, particularly those that might avoid some or all of the adverse environmental effects, is essential. Sufficient analysis of such alternatives and their environmental costs and impact on the environment should accompany the proposed action through the agency review process in order not to foreclose prematurely options which might have less detrimental effects.

OCR for page 85
ENVIRONMENTAL ISSUES 87 The American River is a significant recreational resource. The lower river is an important sport fishery, while the upper river is a steep, free-flowing river valued by boaters. (Ruth- erford H. Platt, University of Massachusetts, Amherst.) Examples of such alternatives include the alternative of taking no action, or of postponing action pending further study; alternatives requiring actions of a sig- nificantly different nature which would provide similar benefits with different environmental impacts (e.g., nonstructural alternatives to flood control pro- grams, or mass transit alternatives to highway construction). CEQA's guidance states, "Attention should be paid to alternatives capable of substantially reducing or eliminating any environmentally adverse impacts, even if these alternatives substantially impede the attainment of the project objectives, and are more costly" (Secretary for Resources, California, 19731. Environmental Information Deficiencies in the 1991 ARWI There were several areas where the lack of scientifically based descriptions of environmental impacts prevented the 1991 ARWI from serving as an adequate planning document to assess the impacts of proposed projects. The most signifi- cant information deficiencies were in the assessment of (1) potential impacts of periodic inundations from a dry dam on the plant communities located in the upper American River canyon, (2) potential impacts of inundation on canyon soils and geologic stability, and (3) potential impacts on ecosystems and regions. Information gathered since the issuance of the 1991 ARWI is still incomplete, so it is not possible to determine or quantify impacts from inundation by a dry dam in the American River canyon. An adequate EIS/EIR would have described the situation in these terms and made recommendations for a research plan.

OCR for page 85
88 FLOOD RISK MANAGEMENT AND THE AMERICAN RIVER BASIN NEPA and CEQA guidelines call for impact assessments to consider re- gional and ecological contexts, but readers of the 1991 ARWI report cannot discern the regional significance of the fisheries, plant communities, wildlife, or the landscape of the upper American River canyon. It is also unclear whether potential ecosystem shifts might result from the cumulative effects of inundation, unstable geology, and plant regeneration problems. The regional significance of such an ecosystem shift was not explored. For example, if the wild trout fishery were affected in the Middle Fork, how would this affect the status of other fisheries in the region or state? Environmental impact assessments must not only identify the probable effects of an action, but also estimate the magnitude and evaluate the importance of these effects. Numerous reports from the early 1970s observed that the assessment process suffered from lack of a systematic means of reporting impact significance and provided procedures to remedy the problem (Leopold et al., 1971; Warner et al., 1974; Dickert and Domeny, 1974~. Because of the scale of the controversy associated with building a dam on a scenic river, uncertainties about the significance of impacts will only hinder decisionmaking and efforts to gain public consensus. The measures that have been taken to improve the 1991 ARWI should provide significant benefit. The 1994 Alterna- tives Report (USAGE, Sacramento District, 1994a) contains a more thorough consideration of options to reduce flood damage, but this preliminary document does not provide analysis of the environmental impacts or gains associated with the different alternatives. A supplemental EIS/EIR was not available in time for this committee's review, so it cannot comment on the final planning document. Project Alternatives Assessment in the 1991 ARWI To meet the requirements of NEPA, and CEQA in particular, the 1991 ARWI reports should have provided substantial environmental analysis of the different alternatives that could be used to increase conveyance in the lower American River. Environmental restoration project components and geomorphological con- siderations involving sediment transport and deposition in the lower American and Sacramento rivers, weirs, and bypasses should have been integrated into project alternative scenarios. If geomorphological factors were judged to have no influence on managing or increasing conveyance capabilities in the lower river systems or bypass, then these conclusions should have been substantiated. Following the release of the 1991 ARWI, a number of significant changes occurred that enabled the 1994 Alternatives Report to present a fuller array of project alternatives. The 1991 ARWI focused strongly on reservoir related op- tions, while the 1994 report was oriented more toward river flood conveyance. Coordination with upstream hydroelectric reservoirs has been arranged to im- prove Folsom flood operations. The local project sponsor has hired consultants in geomorphology to integrate the rebuilding of lower American River levees with riparian habitat restoration. Before and during the preparation of the 1994

OCR for page 85
ENVIRONMENTAL ISSUES 89 Alternatives Report, the Lower American River Task Force, assembled by the local sponsor, met with the goal of actively soliciting the input of any public and agency stakeholders in the lower American River on environmental and public safety needs (SAFCA, 1994b). A more extensive risk and uncertainty analysis was conducted, which should help Sacramento officials better select risk reduc- tion alternatives that balance public safety, financial, and environmental costs. During the period between issuance of the two reports, local initiatives were begun in cooperation with the Sacramento District to restore wetlands in the Yolo Bypass. (The first Yolo Bypass wetland restoration projects date back to 1990.) The interagency Yolo Basin Working Group and the District began an assess- ment of how to integrate both flood protection and environmental restoration objectives in the Yolo Bypass. By 1994, interagency agreements had been ap- proved for the multi-objective management of the Bypass for endangered species protection, wetland and wildlife habitat restoration, and flood control. In 1994 the California Reclamation Board passed Resolution 94-3, requesting that USACE initiate a new reconnaissance study of the Sacramento River Flood Control Project for the purpose of cooperating with other federal and state agencies and public interests leading to a "comprehensive multi-objective river corridor management plan." The local sponsor has realized that the public interest in the environmental values of the upper American River canyons must be an integral component in any flood damage reduction plan and toward this end has reordered its project . . . priorities. The USACE goal is to design projects and conduct analyses of these designs. Land use planning traditionally has been a local responsibility and this division between the two levels of government often results in the omission of consider- ation of how the two elements project design and land use are interrelated. USACE environmental assessments do not generally deal with the secondary impacts of flood damage reduction alternatives, even though such impacts are possible and indeed probable. For example, in the American River case, if an Auburn dam were built as multi-purpose facility, it might open up significant suburban development in Sierra foothill counties such as Placer and E1 Dorado Counties. On the other hand, if a single-purpose flood control dam were built at the Auburn site, it might facilitate development in the Natomas Basin. The state of California's CEQA guidelines do require explicit attention to the relationships between land use and facilities development. A truly comprehensive EIR/EIS, might explore how development options could increase or reduce the need for an Auburn flood control facility, levee improvements, or other measures. It would identify the land use development options near Sacramento with the highest to lowest flood risks. In other words, the description of flood risk reduction alternatives would include land use alter- natives and how they can be used to lessen the risk exposure of population growth centers. A report with an adequate representation of alternatives would also describe

OCR for page 85
9o FLOOD RISK MANAGEMENT AND THE AMERICAN RIVER BASIN how nonstructural measures could be integrated with other measures to form flood risk reduction alternatives. Nonstructural alternatives alone will not pro- vide adequate flood risk reduction for most areas in the American River flood- plain, but they can provide important supplements to levee or dam construction scenarios. Easily placed and removed temporary dams for doorways and win- dows for central Sacramento structures could have been used during flood events such as 1986, when great uncertainties about levee stability and channel capacity on the American River threatened the area. New construction could require elevation of structures both in older developed areas and newly developing areas. A combined flood warning and flood proofing system could be used regionwide. Current efforts under way by the city of Sacramento to produce a compre- hensive floodplain management plan present an opportunity to integrate federal flood control facilities planning along with the National Flood Insurance Pro- gram and local plans. The city plan places new emphasis on floodplain zoning, land use planning, floodproofing, flood warning, and evacuation plans. Presum- ably, the supplemental EIS/EIR preparation is well timed with these local efforts for producing integrated plans. Chapter 5 addresses these floodplain manage- ment issues in greater detail. Limitations of the Environmental Impact Assessment Approach to Project Planning Good project planning should be more than just an exercise in disclosure of the potential consequences of a project. The EIS/EIR process can serve as a dynamic planning tool to facilitate development of project alternatives and com- munity consensus. Public and private stakeholders should be involved as co- participants in the identification of reasonable alternatives, rather than merely reviewing draft reports. The Sacramento Area Flood Control Agency's Lower American River Task Force, organized subsequent to release of the 1991 ARWI to plan levee improvement projects, seems to offer a successful model for im- proved project planning, although this effort is still in progress and thus it is premature to comment on its ultimate usefulness. The 1991 ARWI identified a preferred alternative plan and described its impacts. It then identified the manner and extent of mitigation that might take place. The concept of environmental mitigation provides an unsatisfactory frame- work for this water planning process. Environmental mitigation assumes that environmental factors are considered separately, after hydraulic, hydrologic, and engineering factors. One of the conceptual weaknesses of this planning frame- work is that the objectives of civil works projects and the objectives of restoration projects are viewed as mutually exclusive, competitive objectives or as trade- offs, rather than as mutually supportive objectives. This lack of integration contributed to the reason the 1991 ARWI report did not adequately address ways for managers concerned with reducing flood damage to find incremental gains in

OCR for page 85
ENVIRONMENTAL ISSUES 91 reducing flood risk using different management schemes for levees, weirs, and bypasses as well as reservoir reoperations. The pressures placed on the Fish and Wildlife Service (FWS) to provide quantification for plant mortality for a dry Auburn dam, when FWS staff felt quantification based on existing information was not scientifically defensible, is symptomatic of the mitigation-based empha- sis. The quantification was forced too early in the planning process for FWS to be able to develop a mitigation plan to make specific projects possible. In situations where trade-offs between hydraulic or other engineering con- cerns and natural resources need to occur, mitigation is an appropriate remedy. Mitigation is only as good as its implementation, however, which creates signifi- cant uncertainties. In comments on the 1991 ARWI, the public expressed disap- pointment in past mitigation performances (e.g., of USACE on its Warm Springs Dam and New Melones DamJ. In response to this sentiment, project mitigation strategies that address public distrust should be developed. If environmental features are not determined to be required for mitigation, they are treated as add-one, or "enhancements," to a project. The current practice of USACE is not to engage in enhancement projects and to relegate these as "separable" projects to the responsibility of local sponsors (Kiesck, 1994~. While federal participation in environmental restoration and enhancement projects is fully authorized, planning practice in 1991 and even in 1995 has not yet reflected these legislative policies. ASSESSING THE IMPACTS OF A DRY DAM The probable environmental impacts of an Auburn dry dam to the American River canyon are of central importance in determining the desirability of flood risk reduction alternatives. Two critical areas in which information has been deficient are the impacts of periodic inundation on canyon soils and geologic stability and impacts on plant mortality. The significance of these two issues is that the degree to which canyon hillslope failures or vegetation mortality do or do not damage scenic or ecological values has an important bearing on the public perception of the desirability of a dry dam. Landsliding can degrade scenic val- ues, alter the physical and ecological base for plant and wildlife communities, and degrade fisheries through sedimentation of stream channels. A combination of geological instability and inundation stress on plant communities could alter canyon scenery and ecology. Many Sacramento area residents assign high value to the aesthetic, environ- mental, and recreational attributes of the scenic canyons of the North and Middle Forks of the American River. To accommodate this public sensitivity, the Sacra- mento District and local flood control planners, with public input, proposed a "dry dam" an innovation at this scale. The dry dam at Auburn was conceived as an environmentally sensitive alternative to a previously planned multipurpose dam. As described in the 1991 ARWI, the dam would have impounded peak

OCR for page 85
92 FLOOD RISK MANAGEMENT AND THE AMERICAN RIVER BASIN flows during periods when upstream runoff exceeded the dam outlet capacity. The ungated outlet would have been sized to allow unrestricted passage of nor- mal streamflow volumes, and flow volumes exceeding outlet design would have been detained only for the length of time required to drain through the outlet. This alternative would presumably decrease impacts to vegetation, wildlife, aes- thetics and recreation that normally accompany permanent pool multipurpose dams (USAGE, Sacramento District, 1991~. The range of technical issues raised concerning the potential dry dam im- pacts includes how to predict impacts from unstable geology and soils in inun- dated zones; how to predict impacts from inundation of nonriparian chaparral, oak woodland, and digger pine communities, which are not covered by inunda- tion literature; how to determine the impacts of indirect, longer-term influences stemming from inundation; how to determine the overall ecosystem response to inundation; and how to collect valid data from existing sites with some similarity in plant communities that have undergone inundation events. Canyon Slope Stability There is a legitimate concern over hillslope stability in the upper American River canyon, given the frequency of hillslope failures in the region, the condi- tion of the slopes, and the importance of saturation to slope failures. The extent of impacts cannot be quantified because of a lack of empirical data on depths of anticipated drawdowns and long-term effects of vegetation changes on slope stability. This section evaluates what is known about potential effects on slope stability of deep inundation and rapid drawdowns as might occur behind a dry Auburn dam. Further study on the stability of slopes under current conditions and over extended time scales will be necessary if the dry dam option is pursued. Physical Mechanisms of Slope Failure Slope failures are common in the Sierra Nevada and west central California, where single storm events have generated numerous failures (Campbell, 1975; Ellen and Fleming, 1987; DeGraff, 1994~. Many conditions commonly associ- ated with landslide susceptibility (Cooke and Doornkamp, 1990) pertain to the Middle Fork canyon. Most landslides in California occur during the winter and spring rainy season, which is also the Deriod likely to coincide with deen inunda tion. ~-r Inundation and rapid dewatering affect slope stability. Rising-stage failures are not a dominant concern in cohesive materials because high pore pressures are offset by hydrostatic forces of the submerging water (Taylor, 1937; Chandler, 1986~. They can, however, be a concern in granular, noncohesive materials (Chandler, 1986), such as steep riverwash materials or unconsolidated roadbeds. Rapid dewatering, on the other hand, may pose a substantial problem in the

OCR for page 85
ENVIRONMENTAL ISSUES 93 canyon. Rapid drawdowns can generate slope failures due to increased effective shear stresses while pore pressures remain high (Chandler, 1986), especially in fine-grained materials (Brunsden, 19791. Excess pore pressure is not supported by grains and can be resisted only by soil cohesion (Taylor, 1937~. Experimental results indicate that grain-contact stresses may fall to zero, causing pore-water pressures to locally support the entire stress field (Iverson and LaHusen, 19891. Some failures occur progressively through the cumulative deterioration of friction elements in the matrix until resisting strength is seriously compromised (Brunsden, 1979; Chowdhury, 1992~. For example, clay mineral grains may become locally aligned along a failure plane, which facilitates subsequent fail- ures (Chandler, 1986~. Laboratory results indicate that pore-water fluctuations can propagate outward from existing shear zones, leading to expansion of shear zones (Iverson and LaHusen, 1989~. These factors imply that deep inundation and rapid drawdowns could decrease shear strengths and increase mass wasting hazards long after the inundation period. Past Hillslope Failures Landslides generated by rapid drawdown are common. For example, draw- downs during the failure of the Teton Dam in Idaho in 1976 resulted in the failure of about 3.6 million cubic yards of material from the canyon walls (Schuster and Embree, 1980; Cedergren, 1989~. In addition, major landslides can cause dam- ming of valleys (Evans, 1986) and dam failures downstream. The Vaiont slide in Italy, which was triggered in part by buoyant forces due to elevated ground water levels in response to reservoir filling, was catastrophic both in volumetric propor- tions and loss of life, because it led to failure of the dam below (Cedergren, 1989; James and Kiersch, 1991~. There is little empirical information on drawdowns approaching the rates or magnitudes that might occur behind an ungated Auburn dry dam (USAGE, Sac- ramento District, 1991), although rates could be controlled and damage lessened if controllable gates were employed. Drawdowns in the canyon following the cofferdam breach in 1986 generated numerous landslides and provide the most direct indicator of landslide potential at this location, although no slopes were monitored. The 1986 conditions are regarded as a worst-case scenario for a single, isolated geomorphic event, as drawdown was extremely rapid and was completed within a few hours. An inventory and analysis of landslides in the canyon identified and mapped at least 35 small new slides caused by the 1986 inundation and about 5 slides interpreted as reactivated older slides (USAGE, Sacramento District, 1991, Appendix M). In addition, two large ancient failures were recognized as having potential to fail, and it was recommended they be monitored. These slides are the River Mile 22.4 Slide, which had part of its toe removed by the 1986 flood, and the Cherokee Flat Slide. The Sacramento Dis- trict concluded that there have been several episodes of prehistoric slides and that

OCR for page 85
94 FLOOD RISK MANAGEMENT AND THE AMERICAN RIVER BASIN they make up only a small percentage of the total canyon area, but that it is impossible to determine the magnitude and frequency of sliding in the canyon. The nature of expected canyon inundation was provided explicitly in the form of depth-duration-frequency curves for the original dry dam (USAGE, Sac- ramento District, 1991), although frequencies should have been increased for events with recurrence intervals of less than 10 years due to derivation from an annual maximum rather than a partial duration flood series (Stedinger et al., 19931. A similar set of curves should be generated for any new planned structure. In the initial dry dam plan, the Sacramento District mapped inundation areas using the depth-frequency-duration curves and concluded that inundation im- pacts would be insignificant, with an estimated 1,927 acres of vegetation lost to combined inundation and mass wasting (USAGE, Sacramento District, 1991~. It concluded that following a period of inundation-induced slides, stability would soon be reached: "Most likely, each episode of filling and emptying should cause fewer failures as the unstable portions of the slopes are gradually removed and eventually the canyon walls should stabilize" (USAGE, Sacramento District, 1991, Appendix M). An alternative model, however, could be that failures will propagate upslope beyond the inundation upper limit (WRC-Environmental and Swanson, 1992) and that slope stability would not be reestablished as long as there is a substantial amount of colluvium at high gradients. FWS was critical of the proposed dry dam alternative. On the basis of analysis of aerial photographs in the lower canyon, it concluded that slope fail- ures would be substantial and would have adverse effects on vegetation and habitats in the canyon (FWS, 1991~. The California Department of Water Re- sources (CDWR) expanded on the FWS study by analyzing soil maps and aerial photographs further up the canyon, where soils are coarser "rained and more permeable (CDWR, 1991~. It found that permeable soils tended to remain stable through the 1986 inundation, while prominent scarps developed in impermeable soils. The study concluded that 35 percent of the inundation area is coarse riverwash not susceptible to failure. Of the remaining area, 50 percent (all permeable soils) would be stable under drawdown rates for the 400-year flood control dams but possibly unstable under rates for the 200-year dam (which were proposed at nearly twice those for the 400-year dam), and 15 percent (all imper- meable soils) could be unstable under either drawdown rate. These values have been criticized for underestimating potential failures because (1) the soil perme- ability model did not account for throughflow or rainfall excess contributions from upslope unsaturated zones, (2) the threshold head differential was overesti- mated, and (3) pool drawdown rates were underestimated (WRC-Environmental and Swanson, 1992~. The CDWR study modeled slope stability by comparing soil drainage rates to drawdown rates. The soil-water approximations, based on a ground water model, appear valid, but the criterion for interpreting those values has been criticized (WRC-Environmental and Swanson, 19921. A critical head differential

OCR for page 85
ENVIRONMENTAL ISSUES 95 was estimated between reservoir stage and soil piezometric head above which slope failures were presumed to occur. This threshold was set at 35 feet based on debris-flow scar lengths (including toeslope sediment accumulation areas' mea- sured from the landslide maps. This 35-foot critical head differential has been criticized as too large for several reasons (WRC-Environmental and Swanson, 1992): Failures often begin on upper slopes and propagate down across lower slopes, so slide lengths exceed initial failure planes (WRC-Environmental and Swanson, 1992; DeGraff, 1994~. Failure plane lengths do not necessarily equal head differences at the onset of sliding because failure planes can extend both above the saturated zone and below the reservoir water surface. Subsequent analysis located numerous small slides not recognized by the CDWR study and concluded most slides were less than 30 feet in length (WRC- Environmental and Swanson, 19929. Critical head differentials derived from the difference between reservoir stage and soil piezometric head were compared by CDWR (1991) to much slower drawdown rates than those anticipated (WRC-Environmental and Swanson, 19921. Fugro-McClelland and Leiser (1991) suggested that the CDWR estimates of areas susceptible to mass wasting were excessive because they neglected the stabilizing effects of vegetation roots. These authors subtracted the entire area of fine-grained soils (400 acres) on the premise that those slopes would ultimately fail whether inundated or not, and lowered the remaining area of susceptible coarse-grained soils by 50 percent, from 2,200 to 1,100 acres, on the premise of root stabilization. This assessment was arbitrary and inappropriate because the model did not analyze slope shear strengths, it evaluated slope stability using an empirical relation between head differentials and landslide scar lengths. There is no physical basis for decreasing the CDWR (1991) estimated areas of slope stability to account for vegetation roots. Landslide scar length distributions, a slope-elevation-frequency graph, draw- down analyses, and a slope failure frequency by elevation curve were presented by WRC-Environmental and Swanson (1992~. They concluded that stresses on slopes will be much greater than CDWR (1991) estimates, that drawdown rates greater than 3 ft/hr (3 times the CDWR threshold) will occur on more than 50 percent of the slopes, and that about 80 percent of the inundated slopes (2,300 acres) are "extremely likely to fail" (WRC-Environmental and Swanson, 19921. They called for analyses to establish critical drawdown rates, stable slope angles for this rate, and mapping of stable and rock rubble areas based on inundation depths, topography, and soils. The WRC-Environmental and Swanson (1992) report also estimated slope failure cumulative frequencies, but the frequency

OCR for page 85
ENVIRONMENTAL ISSUES 103 but nonetheless are important to highlight. They include impacts on riparian vegetation in the lower American River, the value of a geomorphic perspective, recreational conflicts, impacts on fish and wildlife resources, and the need for an ecosystem approach to environmental assessment. Impacts on the Lower American River Plant Community In addition to the environmental considerations assessed in the upper Ameri- can River, the flood control planners also had to take a cautious approach to flood damage reduction alternatives that could affect the popular and heavily used American River Parkway on the lower American River, which flows through urban Sacramento. The alternative involving the increase in releases from a reoperated Folsom Reservoir through the American River Parkway levees was considered a contentious environmental issue because of projected impacts on parkway vegetation. Levee reconstruction and/or clearing for channel capacity or levee safety could have impacts on the riparian resources and quality of the river environment. The historically concerned and well-organized constituency associated with the parkway put the planners in the position of balancing public opinion of potential impacts to the upper American River against public opinion of potential impacts to the lower American River a seemingly intractable posi- tion. The Detailed Report on Fish and Wildlife Resources prepared for the 1991 ARWI by the FWS shows a net loss of 679 acres of riparian forest, marsh, and shrub vegetation along the lower American River for the without-dam 150-year protection alternative. This alternative would have changed the operations of Folsom Reservoir to release flows up to 180,000 cfs through the American River Parkway and levee system located in Sacramento. These estimates were arrived at assuming the need to rebuild American River levees, remove vegetation, riprap them, and remove vegetation to increase channel capacities in the floodplains. These impact estimates also assumed short- to long-term loss of riparian vegeta- tion due to removal of the vegetation from the construction necessary to rebuild portions of the Sacramento Weir and Bypass and Yolo Bypass (Monty Knudsen, personal communication, FWS, August 1994~. New studies and institutional developments since the release of the 1991 ARWI reports change the impact assessments of this alternative on riparian and wetland species. The new institutional developments have served to relieve what was initially perceived as an untenable deadlock between upper American River public concerns and lower American River public concerns. As a result of innovative efforts by SAFCA to integrate the concerned public into its planning and design teams and of its openness to environmentally positive levee enhance- ment projects, the levee improvement projects can now be reclassified as envi- ronmentally beneficial projects. Levee improvement plans call for the integra- tion of native riparian plantings into the projects, thereby providing net benefits

OCR for page 85
104 FLOOD RISK MANAGEMENT AND THE AMERICAN RIVER BASIN for riparian plant communities compared to the existing situation (SAFCA, 1994b). Other institutional changes include the previously mentioned program initi- ated by the California Resources Agency to view the Yolo Bypass as an opportu nity for restoring and increasing riparian environments. In 1994 the agency announced new cooperative agreements to begin these restoration projects. The local and state agencies therefore are now viewing flood control improvement projects as opportunities to improve environmental values. Geomorphological Influences on Flood Control Water resources planning traditionally uses hydrologic data and hydraulic models as the focus of engineering studies. The realization that geomorphologi- cal influences, including stream dynamics, need to be routinely integrated into project designs and models has been advocated by fluvial geomorphologists for some time (Leopold, 1974), but only recently have hydraulic engineers and geo- morphologists made progress in integrating consideration of natural river dynamics into project design and hydraulic models (Shields, 1982; Cook and Doornkamp, 1990; Neill et al., 1990; USACE, 1992~. The 1991 ARWI, for the most part, considered geomorphology in terms of its potential influence on upper American River environmental impacts associ- ated with periodic flooding from a dry dam. But geomorphological issues should have received more consideration in the 1991 ARWI in analyzing the levee management options on the lower American River. Channel adjustment, which has the potential to increase the conveyance of floodwaters in the lower Ameri- can, and the sediment transport and deposition in weirs and bypasses, are both important considerations not examined in the 1991 ARWI. If geomorphological factors are not considered, options for increasing flood conveyance while limiting environmental impacts or increasing opportunities for environmental restoration can be overlooked. An increase in conveyance of floodwaters due to increased channel capacities (a result of channel degradation) may make it easier, for example, to allow more riparian restoration on the levees along the lower American River. An increase or decrease in storage capacities of bypasses also has implications for restoration opportunities. There may be systemwide benefits from reduction in flood damage if routine sediment removal at the Fremont and Sacramento weirs can improve the hydraulics of the Yolo Bypass and can lower water surface elevations upstream in the lower American. Sediment removal may represent a cost-effective and environmentally sensitive method of increasing lower river channel capacities. An understanding of stream dynamics is critical to the design of levee im- provements, particularly if there is a commitment to design the improvements with soil-bioengineering revegetation systems instead of traditional riprap. The anticipation of future channel adjustments becomes an integral part of multiple

OCR for page 85
ENVIRONMENTAL ISSUES 105 objective levee improvement programs that must balance conveyance capacities and structural reliability of levees with riparian restoration opportunities. Fi- nally, a better awareness of geomorphological processes provides a potential for incremental gains or losses of conveyance or storage capacities in the whole flood system. Given the recent emphasis on considering more management options for the lower American, the local sponsor has commissioned consultants to evaluate the geomorphology in the lower American River as it relates to bank and levee stability. Given the geomorphological processes acting on the lower American River, the possibility of future channel degradation in the lower river deserves more consideration. The lower river may not have yet attained an equilibrium state from past historical influences. These two factors deserve attention in the formulation of current and ongoing alternatives. Recreation Conflicts Although the 1991 ARWI included nominal consideration and analysis of recreation resources and interests, it became clear in subsequent complaints that planners had not sufficiently involved this segment of the population in real negotiation during the formulation and evaluation of alternatives. Opposition to the dry dam proposal was to a significant extent organized by these interests. The SAFCA Lower American River Task Force is a step in the direction of resolving the impasse, but only a partial solution. Recreational issues will continue to be unresolved without substantial effort by all parties. There are numerous recreational areas in the American River watershed that, by western standards, support heavy public use. For example, the American River is the most popular of all the white water rafting rivers in California (CSLC, 1994~. The development and heavy use of these areas are due largely to the proximity of the basin to the dense population centers of Sacramento and the San Francisco Bay area. Recreation areas are located throughout the basin, from Discovery Park in Sacramento, up the lower American River along the American River Parkway, to the Folsom Recreation Area, and beyond to the Auburn State Recreation Area above the proposed Auburn dam site. In the upper American River basin, the rivers act as natural corridors through the mountains and attracted human travel and activity long before contact with western civilization. The area is characterized by deep canyons with steep walls covered by chaparral, and narrow rugged valley bottoms and occasional rapids. White water rafting is a popular use of all three forks of the American River. The North and Middle Forks are particularly challenging, with many Class IV and V rapids, resulting in white water boating activities of state and national signifi- cance. The major rapids on the North and Middle Forks provide unique scenic features with minimal human intrusion. A dry dam in the upper canyon would significantly disrupt these activities and affect scenic and natural values.

OCR for page 85
106 FLOOD RISK MANAGEMENT AND THE AMERICAN RIVER BASIN The North Fork above the project area, from the Colfax Iowa-Hill bridge upstream to near Heath Springs, was designated a National Wild and Scenic River in 1978. In January 1993, the Bureau of Reclamation determined that the Middle Fork and the North Fork within the project area are eligible for Wild and Scenic designation, and a suitability study is under way. The Auburn State Recreation Area lies mainly within the projected inunda- tion zone of the originally proposed Bureau of Reclamation Auburn dam. The area is less than an hour from Sacramento and is visited by about half a million people each year. Because of its location and the diversity of opportunity, recre- ational use of this area will undoubtedly grow rapidly in the future. In 1972 the lower American River was included in the State Wild and Scenic River System. In 1981 the exceptional anadromous salmonid fishery and other important recreational values of this reach of the river led to its designation as a unit of the National Wild and Scenic River System. The recreational units of the lower basin are linked together by an award-winning trail system. The Jedediah Smith Trail includes bicycle, pedestrian, and equestrian trails from Discovery Park to Folsom Reservoir. Recreational facilities along the American River begin in Sacramento at Discovery Park, at the confluence of the American and Sacramento rivers. Above Discovery Park, the American River Parkway extends 23 miles upstream to the Folsom State Recreation Area at Nimbus Dam. The parkway is largely on the floodplain bordered by high levees that isolate it from the surrounding urban development. The parkway functions not only as a recreational area, but increas- ingly as an urban transportation artery for pedestrians and bicycles. The parkway was used by an estimated 5.5 million people in 1988, and annual use is expected to grow to 7.5 million by year 2000 and 9.6 million by 2020. A 1983 survey found that more than half of these visits were associated with water-enhanced activities such as jogging, nature study, hiking, and picnicking, and that about a third of the visits were associated with water-related activities such as swimming, boating, and fishing. About 12 percent of the recreational use on the lower American River is by boating primarily rafting, canoeing, and kayaking. These activities are highly seasonal in nature, with about 90 percent occurring between Memorial and Labor Days (USAGE, Sacramento District, l991J. Fishing continues to be the biggest recreational use of California rivers, and angling use of the lower American River is particularly important. About 55 percent of the total catch of chinook salmon in the freshwater of the entire Sacramento River basin for the year 1991 came from the American River. Catches of steelhead and American shad from the American River in the same year were also comparatively large, making up 48 and 44 percent, respectively, of the total Sacramento River basin harvest (CSLC, 19941. Because of the size of Folsom Reservoir and its proximity to the Sacramento metropolitan area, Folsom State Recreation Area is one of the most heavily used areas in the state park system. The recreation area begins at the upper end of the

OCR for page 85
ENVIRONMENTAL ISSUES 107 parkway at Lake Natomas, an afterbay formed by Nimbus Dam. Recreational activities include fishing, power boating, sail boating, and windsurfing; there are conflicts between power boating and windsurfing. About 2.1 million people visit Folsom Reservoir each year, mostly Central Valley residents during the summer (USAGE, Sacramento District, 1991J. Reoperation of Folsom Reservoir will result in a lower pool during part of the year, adversely affecting recreational opportunities there. Mitigation is included in the reoperation plan. Prior to urbanization and development, there was little public recreation development in Natomas, but bird hunting and watching on privately owned farmlands were common and continue today through the lease of hunting rights to hunting clubs. Much of the land along the Sacramento River in the Natomas area is privately owned, but the river channel is heavily utilized for recreational fishing and water sports including power boating, jet skiing, and kayaking. De- velopment in the Natomas Basin would limit these recreational opportunities. Impacts on Fish and Wildlife Resources Given the importance placed on fish and wildlife by Sacramento area resi- dents and visitors, potential impacts to these resources warrant careful review. Fish habitat in the project area of the North Fork has been degraded by a number of past actions, extending as far back as placer mining in the mid- 1 800s. Years of habitat degradation have combined with high summer water temperatures to limit the value of North Fork as a fishery resource in the reach that would be affected by a dam of any kind. Though the Middle Fork has also experienced some habitat degradation, the cool water outflow from Oxbow Dam supports a substantial population of large wild trout, both brown and rainbow. This population quali- fied as an "outstandingly remarkable" resource during the Wild and Scenic River Eligibility Assessment conducted by the Bureau of Reclamation in 1992. The fish resources of Folsom Reservoir consist of both warmwater and Goldwater species. The warmwater species, primarily bass, catfish, and sunfish, are adversely affected by fluctuations in surface elevation during the spawning season. These fluctuations, along with low nutrient levels in the reservoir, result in relatively low annual production for the warmwater fishes. The Goldwater species, trout and salmon, are maintained by stocking, though limited natural reproduction occurs in tributary streams. The once abundant chinook salmon resource of the Sacramento River basin has been reduced to a fraction of its original importance. Of the four distinct seasonal runs of this species, only the fall run now occurs in any numbers, and the winter-run fish is classified as endangered under the Federal Endangered Species Act (Fisher, 1994~. Historic runs of salmon in the American River were esti- mated above 130,000 and included both spring- and fall-run fish. Both races of chinook were nearly decimated by hydraulic mining and dam construction in the late 1800s and early 1900s (Gerstung, 1971~. The principal anadromous fish still

OCR for page 85
108 FLOOD RISK MANAGEMENT AND THE AMERICAN RIVER BASIN surviving in the American River is the fall-run chinook salmon, now limited to the reach below Nimbus Dam. This population supports the extensive sport fishery mentioned above and also a significant sport and commercial harvest in the ocean. Over the period 1967 to 1991 (the baseline for the Central Valley Project Improvement Act), the river supported an average run of 32,000 naturally spawning fall chinook adults, about 22 percent of the total Sacramento River run of 143,000. Returns to Nimbus Hatchery below Nimbus Dam for the same period averaged 7,300 fish, 35 percent of the average Sacramento River total of 21,000 hatchery returns. The "naturally spawning" portion of the chinook run is actually heavily influenced by hatchery fish. Some of the fish spawning in the river are progeny of hatchery parents that fail to return to the hatchery, and some of the naturally produced fish interbreed with hatchery stock. The natural run in the American River has declined in recent years (the average run in the past 5 years was about 50 percent of the 25-year average). Steelhead in the American River are substantially less abundant and nearly entirely supported by hatchery produc- tion (1967 to 1991 average returns to Nimbus Hatchery were about 1,700 fish). Fishery values in the Natomas Basin are much lower than those upriver. However, the Natomas area is highly significant for its wildlife values. Thou- sands of migratory waterfowl use the basin for feeding and resting. The Natomas basin reach of the Sacramento River supports one of the highest concentrations in California of nesting territories for the Swainson's hawk, a state-listed threatened species. And the southern portion of the American River basin in Sacramento and Sutter counties, including the Natomas basin, provides one of the most im- portant habitats remaining in California for the threatened giant garter snake (EIP Associates, 19921. Significant fish species in the zone of influence of the lower American River (principally in the San Francisco Bay and Delta, affected by flow releases from Folsom Reservoir' include the striped bass, which provides one of the most important sport fisheries of the state, and the endangered winter-run chinook salmon (reclassified from threatened status in January 1994), which passes the mouth of the American River on its way to spawning grounds in the upper Sacramento River basin. Other sensitive species affected by American River flows include the federally listed as threatened delta smelt, primarily resident in the bay and delta, and the Sacramento splittail (proposed for threatened status by the FWS in January 1994), which occurs both in the delta and in the lower reaches of the American River. Two other federally listed species occur in the project area, the bald eagle (federally listed as threatened) and the valley elderberry longhorn beetle (feder- ally listed as threatened). The eagle occurs in significant numbers only on Folsom Reservoir. The beetle occurs in association with elderberry shrubs primarily in riparian areas of the upper canyon and the lower American River. Owing to the extent of historical habitat degradation in the upper basin, impacts of a detention dam in the canyon are much more likely to be significant

OCR for page 85
ENVIRONMENTAL ISSUES 109 to recreational and rafting interests than to fisheries or fish habitat, particularly in the North Fork. However, the important wild trout fishery in the Middle Fork could be substantially affected if canyon wall sloughing following inundation is extensive. Under current operations the major limitation to success of the fishery re- source of the lower American River is the flow and temperature regime below Nimbus Dam. The period of major concern is during the spawning migration of the fall-run chinook salmon. Owing to low flows during that time and to inad- equate control of temperature of the releases from Folsom Dam, temperatures in the river often exceed those suitable for survival of incubating salmon embryos. Also of concern are temperatures and flows for rearing of juvenile salmon and steelhead during spring or summer. The salmon are less at risk because the juveniles leave the river by the early summer of their first year, before tempera- tures reach maximum levels. Steelhead, however, must rear one or two full years in the river before moving to the ocean. As a consequence of high summer temperatures and limited flows, natural rearing of steelhead has been virtually eliminated; more than 95 percent of returning fish are the result of hatchery rearing (Snider and Gerstung, 1986J. A number of State Water Resources Control Board rulings regulate releases from Folsom Reservoir but they are inadequate to protect fish habitat. High temperatures and substantial and rapid fluctuation in flows are a major limitation to significant natural production of salmonids in the lower river (Snider and Gerstung, 1986; Williams, 1995~. Some relief is potentially available as a result of a recent court decision, resolution of which is still evolving. The case involved the Environmental De- fense Fund et al. v. East Bay Municipal Utilities District et al. At issue was where EB MUD would be allowed to divert an annual 150,000 acre feet, for which it had contracted with the Bureau of Reclamation. The utility district wished to divert the water through the Folsom-South Canal, above Nimbus Dam. Environmental groups and others held that the diversion should occur lower down in the river system to protect the public trust resources of the river. In a decision handed down in January 1990, Judge Hodge of the Alameda County Superior Court allowed diversion through the Folsom-South Canal, provided that sufficient flow was available in the lower American River to support the anadro- mous fishery and other trust resources. The judge approved minimum flows for each season and mandated an ongoing research program. He also appointed a Special Master to oversee the research, which was to be directed toward reducing the overwhelming uncertainty that surfaced throughout the trial and also toward more accurately defining the required minimum flows (Williams, 19951. The decision was based in part on the Public Trust Doctrine (Sax, 1993) and has the potential to influence water management in the state for some time. Owing to provisions in the state constitution, members of the public in California have a special right to use navigable waters for all purposes. The Public Trust Doctrine gives the state particular responsibilities for protecting all beneficial uses of such

OCR for page 85
0 FLOOD RISK MANAGEMENT AND THE AMERICAN RIVER BASIN waters (CSLC, 1994), and it was this authority, in part, invoked in the Hodge . . c .eclslon. Given that reservoir operation already has a substantial detrimental impact on the fish populations, there seem to be no significant additional impacts on fisheries of the reservoir or the lower river from any of the alternatives in the original 1991 ARWI proposal. It also appears that no major additional effects would be associated with interim reoperation of Folsom Reservoir. In fact, the reoperation EIR/EA (SAFCA, 1994a) made several significant concessions to the anadromous fishery and to protection of endangered species. It ensures that if reoperation would require flow levels lower than the "Hodge flows," then Hodge flows would be met, provided that water were available. This obligation would be met by converting, to the extent possible, all potential environmental impacts to reductions in CVP water delivery. In a contract with the Bureau of Reclama- tion, signed in March 1995, SAFCA agreed to compensate the federal govern- ment for this water debt by acquiring sufficient water or water rights from other sources. SAFCA also agreed to finance modifications to the temperature control louvers in Folsom Dam to ameliorate high temperatures in the river, and to fund an evaluation of the impacts of reservoir reoperation on habitat of the Sacramento splittail in the lower river (SAFCA, 1994a). Direct impacts of flood reduction measures in the Natomas Basin appear to be insignificant, but the indirect impacts of the additional development that would be allowed by flood protection could be important to habitat of the threatened giant garter snake and Swainson's hawk. The giant garter snake, listed as threat- ened by the state and federally listed as threatened in October 1993, has recently been given a high profile by the National Biological Service. Development of a giant garter snake Habitat Conservation Plan has been named one of 12 new national priority ecosystem initiatives of the agency. Though contractors for SAFCA had already produced a draft Habitat Conservation Plan for both the Swainson's hawk and the giant garter snake (EIP Associates, 1992), the plan to be developed under this newer proposal will be critical to land development plans in the Natomas Basin. Another major source of uncertainty in the realm of aquatic resource issues is the question of how the additional water required to reduce salinity and improve habitat for endangered fish species in the bay and delta will be allocated. The complicated and interwoven set of circumstances surrounding this issue is dis- cussed in Chapter 6. Projecting Ecosystem Responses in Impact Assessments Since the preparation of the 1991 ARWI reports, federal resource manage- ment agencies have adopted new strategies to consider and evaluate potential impacts within the framework of whole ecological systems. This shift is an effort to correct for past practices. Too often in the past, the focus of environmental

OCR for page 85
ENVIRONMENTAL ISSUES 111 assessment has been on dominant plant or animal species at the expense of understanding the important role that interactions among species and their envi- ronment may have on the species and community survival and the role of corri- dors and linkages of natural environments. The 1991 ARWI report was, for the most part, no exception to this narrow focus, although the FWS did raise the possibility of significant shifts in ecological systems due to disturbances that a dry dam could trigger. It noted the absence of information on impacts to the plant communities (emphasis added). Ecosystem responses to a dry dam could include wildlife community shifts associated with plant community shifts because of habitat changes. Positive ecosystem changes could occur to the Lower American Parkway, in which levee rebuilding and associated revegetation projects could help reintroduce greater riparian species diversity. Because of the significance of potential impacts of inundation on the plant communities in the American River canyon, an ecosystem framework for de- scription of these probable impacts is particularly important. An adequate envi- ronmental assessment should attempt to provide descriptions on how ecosystem dynamics, function, and structure could react to changes made to the system. In the 1991 ARWI report the focus on potential inundation tolerances of individual species loses sight of this critical larger picture. The potential for ecosystem shifts in the American River canyon could be related to the direct impacts of periodic inundation on plants or to the indirect impacts previously discussed, such as changes to the composition of soils, soil microbiota, or community tree or shrub densities. The significance of such potential ecosystem losses and shifts needs to be discussed in a regional context. An example of a regional ecosystem approach to characterizing environmental impacts would be a discussion of the regional or statewide value of riparian and oak woodlands. The 1991 ARWI noted that of the state's original riparian habitat, less than 5 percent remains today. Moreover, less than 2 to 3 percent of the woody riparian habitat remains along the Sacramento River. What is the value of the riparian environment in the upper American River in this context? What is the value of the upland woodlands that could be lost through a combination of inundation and hillslope failure? An effort to frame impacts in a regional context could, for example, recog- nize that oak woodlands are an ecosystem of increasing concern to plant commu- nity ecologists. Studies of foothill oak populations indicate that they do not have the age distribution of healthy vigorous populations. Communities of valley, blue, and Engelmann oaks show a narrow cluster of middle-aged trees, with few young or old ones. The fear is that middle-aged oaks could reach the natural limit of their life span and disappear, to be replaced by other less productive and less diverse communities. Lack of reproductive success is attributed to a combination of factors including deforestation, overgrazing by gophers, deer, and cattle, intro- duction of exotic grasses, and alteration of fire cycles. It is estimated that the

OCR for page 85
2 FLOOD RISK MANAGEMENT AND THE AMERICAN RIVER BASIN state has lost over a million acres of oak woodland since the 1940s (Barbour et al., 1993~. The potential area of impact in the upper canyon support two of the state's three "oaks of special concern," the valley oak, Quercus lobata, and the blue oak, Quercus douglasii. Although sensitive species of oaks could be of concern, inventories of the Cosumnes River watershed suggest that rare landscape forms could be of even greater regional significance in the upper American River watershed. Inventories of rare landforms and vegetative patterns indicate that montane oak woodland and shrub communities are more widespread, for example, than the remnants of Central Valley riparian systems and associated oak woodlands. Nonetheless, a central issue to explore is, how many miles of canyons with free-flowing rivers, scour zones, steep canyon walls, and a diversity of ecosystems do we have (Hart, personal communication, 19951? Future impact evaluations should draw on the methods for inventorying and evaluating landscape uniqueness (Leopold, 1969; Riley, 1974~. Scenarios should be developed for the potential ecosystem shifts of each vegetative community that could be affected. Information so far suggests the loss of old and young individuals and shifts to deciduous species. Chaparral commu- nities disturbed by periodic inundation or landslides could shift to grasslands or even "communities" of invasive exotics. Inundation would likely result in the replacement of any native perennial grasses, ranked very rare by the state, by more weedy, nonnative annual grasses. Disturbances in lower canyon elevations could increase the extent of riparian zones, which typically occupy disturbed environments (Knudsen, 1991; Keeley, 1992; Hart et al., 1994; Meredith et al., 19941. While the forecasting of ecosystem shifts still remains in the realm of speculation, the potential for both positive and negative ecological and aesthetic changes should become part of the evaluation of impacts in future assessments. CONCLUSION The 1991 ARWI raised a number of contentious environmental issues, in- cluding debate over whether the descriptions of the environmental impacts of the various proposed alternatives was adequate. Based on its review, the committee determined that there were several areas of the 1991 ARWI where the lack of scientifically-based descriptions of environmental impacts prevented the report from serving as an adequate planning document. The most significant deficien- cies were in the assessment of impacts that might be caused by periodic inunda- tions from a dry dam on the plant communities in the upper American River canyon, the impacts of inundation on canyon soils and geologic stability, and an ecosystem and regionally-based assessment of impacts. Whether subsequent activities are filling some or perhaps all of these gaps is unclear at this time, but these questions should be resolved with the expected publication of the Sacra

OCR for page 85
ENVIRONMENTAL ISSUES 113 mento District's Draft Supplemental Information Report, expected in the summer of 1995. Public officials ultimately face a difficult decision: given the significant flood hazard to Sacramento, landslide hazards in the American River canyon may be deemed a necessary cost of flood protection. If this option is pursued, great sensitivity to environmental values should be incorporated. If a dry dam contin- ues to be included among the alternatives under consideration, the committee suggests that the following research needs and issues be given consideration: The Sacramento District should form a team of experts in plant physiol- ogy, plant ecology, and geomorphology to design a research plan that combines field experiments, observations, vegetation mapping, landscape uniqueness data, and modeling of landslide risks to develop a canyon inundation impact assess- ment that can secure acceptance and credibility from the scientific community. . The two large old slides should be thoroughly mapped, analyzed, and monitored to assess the potential hazards of catastrophic failure. . If dry dam outlets and storage are to be used they should be designed to accommodate high sediment loads in anticipation of a worst-case scenario of numerous hillslope failures. . Rates of drawdown should be minimized but should be sufficient to pre- vent substantial vegetation mortality until more is known about long-term veg- etation responses to inundation and slope responses to subtle vegetation changes. Gate design and operating policies should consider the depth and fre- quency of inundation while keeping drawdown rates low. These conflicting constraints prevent a dry dam from being used as a first line of defense and instead restrict it to use as a last resort to contain peak discharges from extreme rare events. This philosophy could be made explicit in the Folsom Dam operat- ing policy.