A. ANALYSIS AND DISCUSSION A literature review was performed on the subject of Baseflow Augmentation by Streambank Storage. Part A contains an analysis and discussion arising from the literature review. Part B contains annotated abstracts for ready reference. Part C lists all references reviewed in the course of this study. BASEFLOW AUGMENTATION BY STREAMBANK STORAGE General Aspects Although the subject of baseflow augmentation by streambank storage is not new (Alexander von Humboldt wrote about it in 1819), interest in it is relatively recent. This literature review was able to identify only a few references which dealt directly with the subject (see Part C: Bibliography by Subject: Baseflow Augmentation). In this report, Baseflow Augmentation by Streambank Storage is used to refer to the temporary storage of subsurface water in floodplains, streamsides, streambanks and/or streambottom during the wet season, either by natural or artificial means, for later release during the dry season to increase the magnitude and permanence of low flows. This type of storage can also be used to effect a change in the hydrologic character of a stream, from one that flows intermittently (i.e., seasonally) to one that flows perennially (year-round). In an effort to avoid repetition, and unless specifically stated otherwise, in this report the term "streambank storage" will be used to refer to floodplain, streamside, streambank, and/or streambottom (or streambed) storage. The importance of streambank storage and its effect on stream hydrology, ecology, and geomorphology is now becoming increasingly apparent to a broad spectrum of scientists and professionals, including biologists, ecologists, hydrologists, hydraulic and environmental engineers, and natural resource managers. The temporary storage of precipitation in subsurface soil strata adjacent to streams, for later release during the dry summer months, can directly benefit many stream uses and users. Among the perceived benefits of baseflow augmentation by streambank storage are:
Thus, the benefits of baseflow augmentation by streambank storage are many and varied, cutting across several knowledge areas. The following is a comprehensive list of disciplines impacted by baseflow augmentation by streambank storage:
Hydrologic Aspects The flow of water and moisture under the land surface occurs in two distinct forms or phases: (1) unsaturated, and (2) saturated. Unsaturated flow occurs beneath the land surface and above the groundwater table. The groundwater table forms the boundary between unsaturated subsurface flow (above it) and saturated subsurface flow (below it). In the unsaturated zone the preferred path of movement of moisture is vertical, by percolation, toward the saturated zone. In the saturated zone the preferred path of movement of moisture is horizontal, toward aquifer discharge areas [Mull, 1986]. Sustainable low flows in streams are largely due to aquifer discharge as baseflow. Therefore, the subject of baseflow augmentation can be readily redefined as that of the conversion of ephemeral and intermittent streams into perennial streams. Ephemeral streams are those that flow only in response to direct runoff, during and immediately following a major storm. Intermittent streams are those that flow during the wet season and dry up during the dry season (summer in the U.S. southwest). Perennial streams are those that flow year-round. The behavior of ephemeral, intermittent, and perennial streams can be explained in terms of the relative contributions of direct and indirect runoff. Direct runoff is that which flows on the land surface, is characterized by relatively short response times, and can lead to high peak flows. Indirect runoff is that which flows below the land surface, features longer response times than those of direct runoff, and correspondingly lower peak flows. Indirect runoff consists of two components: (1) interflow, and (2) groundwater flow. Interflow occurs in the soil layers immediately below the land surface, either as unsaturated flow or as isolated pockets of saturated flow moving in a predominantly lateral direction. Groundwater flow occurs below the watertable, driven by potential gradients which tend to follow the natural topography in a subdued way. Ephemeral streams are influent; that is, they serve as aquifer recharge areas. Conversely, perennial streams are largely effluent, serving as aquifer discharge areas. Intermittent streams are those that can change from effluent to influent, depending on the season. During the wet season, an intermittent stream is effluent, discharging subsurface water into the stream. As subsurface water and moisture are depleted, the stream gradually changes its character, from effluent to influent, losing water to the subsurface and eventually drying up. Given the proper topography and lithology, it may be possible for unsaturated flow to contribute to streamflow. In general, however, sustainable amounts of low flow appear to be possible only in streams that can remain effluent throughout the dry season. To assure that the stream remains effluent, the following conditions are necessary:
The first two conditions are related. Generally speaking, adequate aquifer replenishment leads to shallow groundwater tables. In turn, shallow watertables lead to effluent, i.e., perennial, streams. Therefore, adequate aquifer replenishment should cause streams to flow year-round. Aquifer replenishment is a very broad subject, encompassing several disciplines, including groundwater hydrology and hydrogeology, forest and range hydrology, and watershed management, to name a few. Furthermore, the spatial and temporal diversity of aquifer properties, the distributed nature of subsurface water use, and various other institutional and legal constraints contribute to increase the complexity of the subject. Notwithstanding this complexity, it has been widely recognized for some time that aquifer replenishment is directly related to the conservation of precipitation [Horton, 1937; U.S. Dept. of Agriculture, 1940). For a given climate, the larger the fraction of precipitation that is allowed to infiltrate into the ground, the more likely it is that the infiltrated water will eventually go on to replenish the local groundwater reservoirs (Stephens and Knowlton, 1986). Conversely, if most of the precipitation is kept from infiltrating into the ground, aquifer replenishment may be slowed down. In extreme cases, a lack of adequate seasonal aquifer replenishment can cause a lowering of the watertable and the associated depletion of groundwater resources. Water that does not infiltrate into the soil not only does not replenish groundwater, but also becomes available for surface runoff. In surface soils of low permeability, whether natural or human-induced, increased quantities of surface runoff invariably lead to floods and flood damages to individuals and property. Furthermore, increased amounts of surface runoff substantially enhance the flow's competence to entrain and transport sediment, resulting in negative impacts to water quantity (by reservoir sediment deposition) and water quality (nonpoint-source pollution). While aquifer replenishment is subject to management, the hydraulic properties of aquifers are largely determined by nature, with little or no human intervention. Therefore, it is possible to accomplish baseflow augmentation with a management strategy focused on effective and adequate seasonal aquifer replenishment. Moreover, the aquifer's size and hydraulic properties can be used to identify those which can be readily managed for baseflow augmentation. In general, large and relatively slow-draining aquifers are good candidates for baseflow augmentation. On the other hand, small and relatively fast-draining aquifers are not very promising candidates for baseflow augmentation.
Hydraulic Aspects Given adequate aquifer replenishment, baseflow augmentation hinges upon the characteristics of the aquifer, including its geometric and hydraulic properties. The aquifer's geometric features help establish its type, size, and boundaries, including the presence of constraining aquicludes. The hydraulic properties help establish the rate of drainage, which in turn determines whether or not the aquifer can continue to drain throughout the summer. A literature review on the subject of surface-subsurface flow analysis led to the references listed in part C under Bibliography by Subject. The interaction between surface and subsurface flow near streambanks is characterized by the rate of baseflow recession. In finite-width aquifers drained by intersecting streams, the recession of baseflow can be shown to follow an exponential decay curve [Cooper and Rorabaugh]: Q = Qo e-at . . . . . . . . . . . . . . . . . . . . . (1) in which Q is the baseflow at time t after Qo, and a is a recession constant equal to: a= (π2T) / (4SL2) . . . . . . . . . . . . . . . . . . . . . (2)
in which T = transmissivity, S = coefficient of storage, and L = aquifer width. In groundwater hydraulics, the ratio of transmissivity T to coefficient of storage S is referred to as aquifer diffusivity. Therefore, the rate of aquifer drainage increases with aquifer diffusivity (T/S) and decreases with the square of aquifer width (L). In theory, the smaller the aquifer diffusivity and the larger the aquifer width (measured in a direction perpendicular to the stream alignment), the smaller the rate of aquifer drainage and the greater the likelihood that the stream will remain effluent throughout the year. For practical applications, the recession constant a can be obtained from Eq. 1 using baseflow recession data. Moreover, given an estimate of aquifer width L, coefficient of storage S, and transmissivity T, Eq. 2 can be used to solve for the recession constant a. With the aquifer parameters either known, estimated, or obtained by calibration using streamflow data, Eq. 1 can be used in a predictive mode to calculate baseflow recession curves. In another study, Rorabaugh (1963) has shown that the aquifer response to an instantaneous water table increase (i.e., an aquifer recharge) can be related to the aquifer properties (see Part B). These analyses have shown that it is possible to predict baseflow response on the basis of aquifer hydraulic characteristics, which can be either estimated or determined from field tests. Notwithstanding current knowledge, the literature is inconclusive as to the behavior of complex stream-aquifer systems which can be effluent at one time and influent at another. While the governing physical processes appear to be the same, the initial and/or boundary conditions are likely to be different. Additional research is needed in this area of groundwater hydrology and hydraulics.
Vegetative Aspects The vegetative aspects of baseflow augmentation are now beginning to receive wide attention. This has closely followed a renewed public interest in riparian areas, their hydrology, ecology, and management. The literature of the last decade contains many studies reporting on all aspects of riparian area management. See Part C: Bibliography by Subject: Riparian Area Management, for a list of key references identified in this study. Riparian zones or areas are the often narrow strips of land that border creeks, rivers, and other bodies of water [Elmore and Beschta, 1987]. Riparian areas usually have varying amounts and diversity of riparian vegetation. In arid and semiarid regions, the latter are typically phreatophytes, or well plants, i.e., plants that are able to survive the dry summer months by drawing moisture from the subsurface and groundwater [Meinzer, 1927]. Up to the early 1970s, riparian vegetation was largely regarded as a nuisance, consuming large amounts of valuable water, particularly in the arid and semiarid regions of the western United States. The last decade, however, has seen a gradual change in the public's perception of the role of riparian areas [U.S. General Accounting Office, 1988]. Riparian areas are now broadly perceived as beneficial, positively impacting a wide range of stream functions, including water and sediment control, channel and streambank stability, fish and wildlife habitat, stream temperature, water quality, and stream aesthetics. Riparian vegetation serves as the catalyst for the storage of large amounts of water in streambanks and streambottoms, generally storing more water than it consumes. The amount of water consumed by riparian vegetation is seen as a small price to pay for the multiple benefits that can accrue from healthy riparian areas. The relationship between baseflow augmentation and riparian vegetation is unclear at the time of this writing, despite the many efforts to document the link between them [Heede, 1977; Stabler, 1985; Elmore and Beschta, 1987]. A plausible scenario supported by field observation appears to be the following: Increased amounts of subsurface moisture in streambanks, resulting from natural and/or artificial aquifer replenishment, encourage the establishment and growth of riparian vegetation and assure its survival from year to year. In turn, once established, the riparian vegetation acts to encourage sediment deposition, increase soil infiltrability and soil-moisture retention capacity, and reduce stream velocity, thereby further increasing the rate of subsurface moisture replenishment during high flows [Horton, 1937; U.S. Department of Agriculture, 1940]. Effective subsurface moisture replenishment then leads to saturated groundwater flow and to groundwater accretion and raises the watertable near the streambank. With an aquifer of the proper geometric and hydraulic properties, the rise of the watertable near the streambanks can change the character of the adjoining stream from intermittent to perennial. Moreover, the magnitude and duration of summer streamflows is a function of the aquifer properties and of the effectiveness and amount of aquifer replenishment. Most experts agree that sound riparian area management is the key to restoring degraded streams to their original (or pre-impact) conditions [Elmore and Beschta, 1987; U.S. General Accounting Office, 1988; DeBano and Schmidt, in press]. In the field of riparian area management, baseflow augmentation is perceived as a predictable byproduct, to be counted as an additional benefit of the treatment. Experience has shown time and again that degraded streams can lose their perennial character and become intermittent, while sound riparian restoration practices can help degraded streams regain their perennial character in time [Heede, 1977; Elmore and Beschta, 1987].
Irrigation Return Flow and Artificial Recharge Irrigation return flow is that fraction of the flow diverted from a stream or river to irrigate neighboring agricultural lands, which is in excess of that consumed by the crops and which is eventually returned to the nearby stream or river, either through surface or subsurface flow. Artificial recharge refers to the management of surface water with the aim of converting increased amounts of it to subsurface and groundwater, thereby replenishing local aquifers. Both irrigation return flow and artificial recharge can eventually lead to baseflow increases. For a given site, irrigation return flow amounts vary with the cropping patterns, mode of irrigation, and crop water application techniques [Brosz, 1986]. These amounts tend to fluctuate widely in a random manner and are, therefore, not readily subject to management. For this reason, irrigation return flow is not generally perceived to be a viable strategy for baseflow augmentation. Artificial recharge encompasses the methods and practices whose objective is to increase soil infiltrability, ponding time, and/or total infiltration. Methods of artificial recharge are varied, ranging from mechanical to structural practices [Helweg and Smith, 1978; Motts and O'Brien, 1981]. The literature on artificial recharge focuses on mechanical methods or other means of replenishing groundwater reservoirs, primarily to increase the yield of neighboring wells. Thus, the subjects of artificial recharge and baseflow augmentation are intrinsically related. Recharge methods are discussed in more detail in Sections 3.4 and 3.5 of this report. CASE STUDIES OF BASEFLOW AUGMENTATION The Camp Creek Experience The experience of Camp Creek is widely regarded as a classic example of successful baseflow augmentation following sound riparian management [Elmore and Beschta, 1987; Barber, 1988]. Camp Creek, a tributary of the Crooked River, is located about 43 km east of Prineville reservoir in Crook County, in semiarid central Oregon. It is about 64 km in length, including its south, middle, and west forks [Winegar, 1977]. Before extensive settlement in the mid-nineteenth century, the area drained by Camp Creek was a grassy wetland, featuring predominantly grassy vegetation and a shallow watertable maintained within root reach throughout the year. The grassy vegetation encouraged soil infiltrability and made possible an effective aquifer replenishment from year to year. Around the turn of the century, overgrazing of the meadow (loss of grassy vegetation coupled with reduced surface soil infiltrability caused by livestock trampling) led to increased quantities of surface runoff and incipient gully development. Under these conditions, severe floods triggered accelerated gully erosion and led to the development of the deeply incised channel with almost vertical banks which now cuts through the valley. Winegar (1977) has documented gully depths ranging from 4.5 to 7.5 m and widths from 7.5 to 30 m. In time, the lowering of the base level coupled with inadequate aquifer replenishment led to the draining of the meadow and the lowering of the watertable to the point where it remained permanently out of reach of most herbaceous vegetative species. Over a period of several decades, this unfortunate sequence of events led to the transformation of the Camp Creek meadow from a grassy wetland into an area that could support only upland species, with sagebrush taking over as the predominant mode of vegetative biota. The Camp Creek meadow and its environs, which originally provided a livelihood for approximately 75 people, is now only able to support only one family [Wayne Elmore, personal communication]. About 20 years ago, the Oregon Department of Fish and Wildlife and the USDI Bureau of Land Management took bold steps to reverse the desertification trend of the Camp Creek meadow, brought upon by decades of land abuse and mismanagement. Beginning about 1968, increasing lengths of the creek (up to approximately 6 km) were fenced to exclude livestock, with fencing continuing to date. Over the years since the original fencing, the exclusion of livestock has allowed the establishment of a healthy riparian area. This has led to a buildup of the streambed and streambanks, a deeper and more stable low flow channel, and a reduction in the amount of sediment being transported by the stream. Net aggradation of the streambed is continuing to date. A byproduct of the Camp Creek fencing decision has been the substantial increase in summer flows within the exclosure, even during drought years. Despite recent field research to document the groundwater system at Camp Creek, the mechanisms responsible for this phenomenon have not been clearly delineated. The presence of the exclosure, resulting in the establishment of riparian vegetation, net bank building, channel aggradation, and subsequent changes in stream gradient above and within the exclosure, may be responsible for raising the watertable and decreasing the exit gradient to the stream, thereby slowing the rate of aquifer drainage into the creek. Although the regional watertable has been shown to be an effluent system feeding the stream during low flow periods, the presence of an aquiclude underlying permeable strata may be partly responsible for the flow regimes documented in Camp Creek [Barber, 1988]. While the literature is inconclusive [see for instance, Oregon State University Water Resources Research Institute, 1986], it appears that the conversion of Camp Creek from an intermittent stream to a perennial stream within the exclosure is largely due to the rise in the watertable adjacent to the streambanks. An increase in base level within the exclosure, amounting to a maximum of 1.8 m since the start of fencing [Wayne Elmore, personal communication], may have reduced the exit gradient with which the regional watertable drains into the creek. Given the extent of the regional aquifer and the alluvium deposits prevailing at Camp Creek, it appears that the increase in base level may have contributed to an increase in aquifer storage, resulting in the maintenance of effluent stream-aquifer conditions throughout the year. The experience of Camp Creek is judged by many experts to be an excellent example of successful riparian restoration. The conversion of Camp Creek from an intermittent stream into a perennial stream is one of the many positive impacts brought about by the exercise of sound riparian management. Although the Camp Creek experience has some site-specific features, its value as an example of successful riparian restoration and stream improvement appears beyond doubt.
Sheep Creek Barrier Dam The Sheep Creek Resource Conservation Area Project was implemented from 1957 to 1966 to stabilize and rehabilitate the upper watershed of Sheep Creek, a tributary of the Paris River, in southern Utah. At the time, the USDI Bureau of Reclamation was considering construction of a dam on the Colorado River in Marble Canyon near Lee's Ferry, Arizona. Bureau officials were concerned that the small reservoir created by the dam would rapidly fill with sediment. The Paris River drainage produced the highest sediment concentrations of any subwatershed within the Colorado River system, and would be a major source of sediment for the Marble Canyon project. The Sheep Creek project was a cooperative effort involving six federal agencies, the Utah Division of Wildlife Resources, and private landowners. Rehabilitation and stabilization measures included construction of detention dams, dike water-spreader systems, gully plugs and check dams. In addition, numerous seeding and brush control projects were undertaken, and livestock grazing was more intensively managed [Hooper et al, 1987]. As part of the project, the Bureau of Reclamation constructed the Sheep Creek Barrier Dam, a large detention dam on the main stem of Sheep Creek at the lower end of the project area. This dam has been very successful in trapping sediment eroded from gullies and other headwater sources. In 1961, only one year after dam completion, the reservoir was completely filled with sediment to spillway crest elevation. Over the ensuing years, sediment has continued to accumulate behind the dam, creating a sediment wedge which in 1984 extended 1,424 m upstream to an elevation of 18 m above spillway crest. More than 75 percent of the total sediment deposit behind Sheep Creek Barrier dam is now above spillway crest [Van Haveren et al, 1987]. Healthy stands of riparian vegetation have become established on approximately 6.6 ha of surface area above the sediment deposit. A small channel, with an average bankfull capacity of 2.6 m3/s, is currently incised in the sediment wedge that has formed between the spillway and the upstream natural channel [Hooper et al, 1987]. A perennial flow at the dam has resulted from water draining slowly from the accumulated sediment deposit. Flooding occurs on the sediment wedge for all significant runoff events. Flooding serves to spread the flows and reduce runoff peaks, as compared to a gullied condition in which large flows are entirely confined within the banks. In addition, flooding reduces stream energy, encourages sediment deposition, and recharges the aquifer. The size and recharge characteristics of the sediment wedge at Sheep Creek Barrier Dam are apparently adequate to maintain perennial flow at the dam site.
Alkali Creek Watershed Rehabilitation Project Alkali Creek is located in the White River National Forest, 32 km south of Silt, Colorado. In 1958, the Alkali Creek Soil and Water Rehabilitation Project was initiated by the USDA Forest Service in response to increased land use pressure, documented as accelerated gully development throughout the first half of the twentieth century. The reason for the gully growth is attributed to the combined effect of drought and overgrazing, as well as the overuse of the agricultural lands below the watershed. Overuse of the lands led to greater incision of the channels, which resulted in the lowering of the base level for Alkali Creek [Heede, 1977]. The headwaters of Alkali Creek, about 2.5 square kilometers in area, were fenced from 1958 to 1966 to exclude cattle grazing. In 1963, the Forest Service constructed 133 check dams in about half of the gullies located within the project area. Four gullies with a total length of 580 m were converted to vegetation-lined waterways. Seven years later, the previously ephemeral flow became perennial, although there was no noticeable change in average annual precipitation during the period [Heede, 19771. The change from ephemeral to perennial flow at Alkali Creek is attributed to the establishment of vegetation in the gullies above the check dams. The vegetated channels and check dams were able to moderate flow velocities, arrest gully erosion, and encourage sediment deposition. In time, the sediment deposits behind the check dams acted as artificial aquifers, retaining flood waters and creating a sufficiently moist environment to encourage the establishment and growth of riparian vegetation. Once established, the riparian vegetation aided aquifer replenishment by reducing flow velocities, trapping sediment, and increasing soil infiltrability. The amounts of moisture stored in streambanks and streambottoms at and near the check dams are apparently sufficient to maintain perennial flow at the Alkali Creek watershed outlet.
Trout Creek Watershed Rehabilitation Project The Trout Creek watershed is located within the San Isabel National Forest, east of Buena Vista, Colorado. Following settlement in the mid-nineteenth century, the Trout Creek watershed was subjected to intensive exploitation for timber, ranching, and farming. Timber clearing, overgrazing, and numerous forest fires stripped the watershed of its natural vegetation, encouraged accelerated erosion, and led to the development of an extensive gully network. In 1933, the Trout Creek Civilian Conservation Corps was established to carry out watershed rehabilitation works in Trout Creek. The activities of the Trout Creek C.C.C. between 1933 and 1938 led to the construction of nine concrete dams, 53,996 temporary dams (rock, log, brush-wire), 405 acres of gully seeding and sodding, three acres of streambank protection, 2.5 miles of terracing, 29 acres of tree planting in gullies, and 5,825 acres of tree planting in upland areas, 38 percent of which were considered successful. Rehabilitation work continued until 1953, with grass seedings and shrub plantings [Jauch, 1957]. The many thousands of rock-log-brush type gully plugs built by the Civilian Conservation Corps produced a marked slowdown of runoff that contributed to gully stabilization. The varied soil and water conservation strategies tried at Trout Creek, including gully plugs, water-spreader dikes, contour trenches, seeding, etc., aided in the restoration of the vegetative cover, stabilization of soil and water, and in raising the watertable to restore meadows to their original (or pre-impact) condition. Moreover, the sediment deposits accumulated at the gully-control structures have acted as artificial aquifers, storing subsurface moisture for eventual release as baseflow [Van Haveren, 1986]. In the past decade, beaver have moved into the Trout Creek watershed, which now features many large beaver dams and ponds. These dams contribute to the replenishment of subsurface moisture and encourage the establishment of perennial flow in many of the Trout Creek drainages. Many of these drainages are now either completely healed or well along in the process of healing, pointing to the unqualified success of the combined structural and vegetative treatments implemented in the Trout Creek watershed. MANAGEMENT STRATEGIES FOR BASEFLOW AUGMENTATION Several strategies for baseflow augmentation have been proposed in the literature. Invariably, they focus on the replenishment of subsurface water and groundwater during the wet season, for later use during the dry season. The following strategies are judged to be of sufficient importance to warrant detailed discussion in this report:
Range Management Rangelands are lands on which the native vegetation is predominantly grasses, grass-like plants, forbs, or shrubs. Rangelands include natural grasslands, savannas, shrublands, most deserts, tundra, alpine communities, coastal marshes and wet meadows. Grazing is the consumption of standing forage by livestock or wildlife. Browsing is the consumption of leaf and twig growth of plants by livestock and wildlife. Grazing management is the manipulation of grazing and browsing animals to accomplish a desired result. Range management refers to the use of rangelands and range resources for a variety of purposes, including livestock grazing, soil and water conservation, wildlife habitat, recreation, and aesthetics. The hydrologic effects of livestock grazing have been recognized for several decades [U.S. Department of Agriculture, 1940]. Comprehensive reviews of the subject are given by Blackburn (1984), Skovlin (1984), and more recently, by Platts (1989). Blackburn (1984) focuses on the impacts of specialized grazing systems on watershed characteristics and response. Skovlin (1984) discusses the impacts of grazing on wetlands and riparian habitat. Platts (1989) evaluates grazing management strategies with respect to their compatibility with the requirements of the stream-riparian zone and a productive fisheries. Generally, livestock grazing affects watershed response by the removal of protective plant cover and by trampling. Removal of vegetation has the following effects: 1.It increases the impact energy of raindrops, encouraging splash erosion and dislodgement of soil particles from the surface. 2.It decreases soil organic matter, decreasing soil infiltrability. 3.It increases surface runoff, encouraging the entrainment and transport of fine suspended sediments and the eventual development of relatively impermeable surface crusts, further abetting surface runoff and sheet erosion. Livestock trampling has the following effects: 1.It destroys the protective cover of plant litter, decreasing infiltration and surface detention and increasing surface runoff. 2.It increases the bulk density of the soil beneath the surface, decreasing soil infiltration and increasing surface runoff. The net effect of grazing is increased surface runoff and erosion and a decreased rate of subsurface moisture replenishment. The question of 'to graze or not to graze' has been promoted in the literature as the key to effective soil and water conservation practices, in the interest of sound range management (Elmore and Beschta, 1987]. Throughout the years, numerous studies have shown that livestock grazing, if not managed properly, has a pronounced effect on soil and water conservation, particularly in the dry arid and semiarid regions of the U.S. southwest. However, prohibiting or substantially limiting livestock grazing in public lands does not appear to be a feasible solution at this time. Barring grazing exclusion, range management is currently seen as a viable alternative. Range management aims to control the intensity, duration, and season of use of vegetation by livestock. A range management plan prescribes rotational grazing strategies that control the distribution of livestock to meet management objectives. If given the option, livestock will tend to concentrate in riparian areas. Therefore, these areas often become overutilized, while forage on adjacent uplands remains underutilized. Range management systems are designed to more evenly distributed grazing pressure, to improve rangelands, and to ensure sustainability of renewable resources. Blackburn (1984) has examined the status of range management and its effect on western range hydrology. A review of the literature on range management led to the conclusion that hydrologic impacts of livestock grazing are a function of interactions between climate, vegetation, soil, and intensity and duration of livestock use. Livestock grazing should be managed to minimize its adverse hydrologic impacts. The exclusion of livestock from some of the nation's arid and steep rangelands is a desirable and viable option. However, extreme solutions such as heavy continuous grazing or complete grazing exclusion do not appear to be realistic management objectives. Programs of field research and demonstration are needed to establish the hydrologic impacts of specialized grazing systems, including:
Sound range management is a viable management strategy for baseflow augmentation. Experts are in general agreement that poorly managed rangelands invariably lead to watershed degradation, excessive surface runoff, floods, accelerated erosion, depletion of subsurface moisture and groundwater, and loss of baseflow. Conversely, numerous examples show the multiple benefits of sound range management, including effective soil and water conservation, reduced incidence of floods, enhanced erosion control, and adequate subsurface moisture replenishment, the latter often translating into net gains in baseflow [see, for instance, Copeland, 1960; Heede, 1977; Blackburn, 1984; and Elmore and Beschta, 1987]. Upland Vegetation Management The effects of upland vegetation management on baseflow augmentation are not very well defined to date. Past emphasis has been largely on the effects of vegetation clearing (forest and shrub) on water yield, rather than on baseflow augmentation. Numerous studies have attempted to relate vegetation clearing to water yield increases, with mixed results. For instance, Hibbert (1971) showed that under favorable conditions, water yield can be increased by conversion of chaparral to grass in Arizona. The amount of increase depends on the soil depth and moisture storage capacity, depth and distribution of annual rainfall, and types of vegetation present on the site before and after conversion. However, other studies have failed to establish a direct relationship between the clearing of certain types of forest vegetation (i.e., pinyon and juniper) and overall increases in water yield [Collings and Myrick]. Baker (1984) studied an herbicide-treated pinyon-juniper watershed in Arizona and concluded that the increase in water yield brought about by the treatment was not significant. Bosch and Hewlett (1982) reviewed a total of 94 catchment experiments designed to relate vegetation clearing to water yield and concluded that, while detailed quantitative relationships were elusive, it is possible to develop reasonable qualitative estimates for planning purposes. In general, vegetation clearing solely for the purpose of increasing water yield appears to be misdirected, since it is very likely to lead to increased erosion potential and to negatively impact water quality. There are, however, exceptions, particularly when the emphasis is shifted from water yield to baseflow augmentation. For instance, in central Oregon, juniper stands in their natural state are known to consume large amounts of water and to inhibit the growth of grassy vegetation, encouraging surface runoff and erosion. Partial clearing of these forests to allow the reestablishment of grassy vegetation leads to increased infiltration, reduced surface runoff and erosion, adequate soil moisture replenishment, and baseflow augmentation in nearby creeks [Wayne Elmore, personal communication]. Stabler (1985) has provided additional details on the central Oregon experience in baseflow augmentation. Streamflow increases were documented in five small Oregon watersheds that had undergone treatment by juniper cuttings. Livestock were excluded from the treated watersheds. In time, these watersheds underwent a change from a dry, erosive, juniper-dominated valley to a moist, stable, riparian environment in which the remaining juniper trees coexisted with newly established grasses. Increased release of groundwater to augment baseflow was attributed to the removal of juniper. At each site, extension of flows in both time and space occurred, and healthy stands of riparian vegetation became established. Along one of these previously intermittent drainages (Skull Hollow, Crooked River National Grassland), perennial flow was present throughout a mile-long study area and for 0.3 miles downstream four years after treatment [Stabler, 1985]. The somewhat limited experience with baseflow augmentation by upland vegetation management precludes broad generalizations. However, the central Oregon experience appears to hold some promise. More research is needed to clearly delineate the potential for baseflow augmentation as a result of upland vegetation management. The relationship between climate, topography, vegetation type, soil, surface and subsurface water amounts has a bearing on the potential for baseflow augmentation in a given upland site. Timber management is an important component of the management of upland vegetation. Timber harvesting usually requires the development of extensive road networks to transport logs and equipment. Timber management focuses on the proper location of roads and skid trails to minimize the adverse effects of timber harvesting on soil and water conservation. Lack of adequate timber management usually leads to accelerated erosion, increased flood hazard, loss of baseflow, and watershed degradation. Riparian Vegetation Management The literature provides ample evidence in support of the statement that baseflow augmentation in small upland streams is directly related to the enhancement of riparian vegetation. A comprehensive review of the subject has been given by DeBano and Schmidt (in press). Field observations indicate that riparian vegetation becomes established in streambanks where there is an ample supply of subsurface moisture year-round. Healthy stands of riparian vegetation act to reduce stream velocities, to encourage overbank flooding and sediment deposition, and to increase bank stability. In addition, they play a key role in moderating streamflows, serving as recharge areas during high flows and as discharge areas during low flows [Van Haveren and Jackson, 1986]. Thus, riparian vegetation acts as an effective agent for aquifer recharge, enabling the storage of significant amounts of water in streamsides, streambanks and streambottom, for gradual release as baseflow during the dry season. Interest in the multiple benefits of healthy riparian areas is relatively recent. Traditionally, riparian area management has focused on the clearing of phreatophytes for the purpose of conserving water resources, particularly in the semiarid Southwest U.S. There was a broad perception among researchers and managers alike that vast quantities of groundwater were being consumed by phreatophytes, a concept seemingly substantiated by early research showing that Saltcedar (Tamarix chinensis), for example, transpired from 4 to 10 acre-feet of water per year. However, more recent work has shown that phreatophytes transpire much less water than previously thought, and water savings due to phreatophyte clearing now appear to be minimal. The cost/benefit ratio for phreatophyte clearing operations is, therefore, unsatisfactory, and such efforts have been largely abandoned [Graf, 1980]. Dense growth of riparian vegetation has run counter to flood control objectives in natural channels draining major basins. These channels have tended to overflow their banks during floods due to the presence of riparian vegetation, which reduces stream conveyance, increasing the magnitude and frequency of floods. Where such floods have occurred, state and federal flood control officials have been requested to clear the flood channels of the riparian vegetation, to allow for the passage of flood waters. However, plans for clearing riparian vegetation to satisfy flood control objectives have been opposed by efforts to preserve the riparian habitat [Graf, 1980]. The multiple benefits of riparian vegetation to the hydrology and ecology of upland streams appear beyond doubt. However, for lowland streams, the issue of preservation of riparian vegetation vs flood control objectives is very complex, with hydraulic, biological, ecological, and legal implications. A stable, vegetated lowland stream system is probably able to absorb most of the impact of a flood, since infiltration rates and soil-water holding capacity are greater than in a comparable degraded, nonvegetated stream. However, vegetation stands decrease stream velocities, increasing flood stages and flood hazard. For lowland streams and rivers, the benefits of riparian vegetation must be reconciled with the loss of channel conveyance and increased flood hazard. Additional research is needed to formulate management strategies that are duly responsive to both stream functions. In the end, it is likely that some measure of compromise can be reached. Upland Runoff Detention and Retention The positive effects of upland runoff detention and retention on the conservation of soil and water, the replenishment of subsurface and ground water, and the stabilization of streamflows (i.e., reduction of peak flows and increase in baseflows) have been recognized for some time [see, for instance, Horton, 1937; and U.S. Department of Agriculture, 1940; Van Haveren, 1986]. In general, replenishment of subsurface water and groundwater is a necessary condition for baseflow augmentation. Therefore, upland runoff detention and retention is a viable management strategy to increase the magnitude and permanence of low flows. As early as 1937, Horton identified the following strategies for streamflow stabilization (Horton, 1937]:
In agricultural lands, substantial increases in infiltration capacity can be attained by keeping the soil in good tilth and cultivating promptly when the soil surface has become rain-packed or sun-baked. Wile recently plowed soil has an infiltration capacity equal to or greater than the highest rainfall intensities, it is normally not feasible to keep agricultural lands in top infiltration condition throughout the year. Therefore, for agricultural lands, streamflow stabilization by increases in infiltration capacity does not appear to be a viable management strategy. Increases in depression storage serve to moderate runoff, to lengthen the ponding time of surface water, to increase total infiltration, and to replenish subsurface moisture. Depression storage can be effectively increased by the construction of water-spreader dikes, which detain runoff in slopes and meadows, encouraging its infiltration into the soil [Hooper et al,1987]. Such dikes have been used for centuries by agricultural societies throughout the world for the purposes of conserving valuable water, particularly in arid and semiarid regions. Oosterbann (1982), for instance, describes the Kushkaba and Sailaba systems used in Baluchistan, Pakistan. The Kushkaba consists of a series of small earth embankments constructed across upland slopes for water harvesting purposes. The Sailaba system is similar in principle to the Kushkaba, but it is built on the flood plain, to retain overbank flood waters and encourage infiltration. In addition to dike structures, it is also possible to increase depression storage by agricultural practices such as strip cropping and contour furrowing. Decreases in the rate of overland flow can be accomplished by increasing surface roughness, decreasing surface slope, and/or increasing the length of overland flow [Horton, 1937]. Decreases in overland flow serve to slow down the rate of water exit from soil surfaces, increasing time of ponding and total infiltration. In natural vegetated slopes, increases in surface roughness are invariably related to the establishment and growth of grassy vegetation, which acts to impede overland flow, reducing its velocity. Decreases in surface slope can be accomplished through terracing. The use of grasses serves to reduce surface runoff as compared with areas under cultivation. Grasses provide a higher and better-sustained soil infiltration capacity. This is due to root, earthworm, and insect borings, and to the prevention of soil inwashing by the grass mat. Another, and probably more important reason for the observed low surface runoff from grass-covered areas is the greatly increased resistance to overland flow. The additional friction due to subdivision of the flow by plant stems and leaves lengthens overland flow detention time and increases total infiltration. Agricultural practices and other mechanical measures have long been identified as viable management strategies to conserve soil and water. Certain cropping practices and mechanical measures effectively reduce soil and water loss from croplands and other areas lacking adequate protective cover. The U.S. Department of Agriculture has identified the following practices for the proper management of soil and water conservation [U.S. Department of Agriculture, 1940]: (1) crop rotations, (2) strip cropping, (3) terracing, (4) contour furrowing, and (5) the use of structures. The effectiveness of crop rotation in reducing soil and water losses from agricultural areas has been thoroughly documented in the literature (see, for instance, U.S. Department of Agriculture, 1940]. This reduction is due to the cover provided during certain years of the rotation period and to the additional organic matter added to the soil by the rotated crop. Strip cropping is an effective measure to reduce soil and water losses from agricultural lands. In strip cropping, strips of close-growing crops are placed on the contour and alternated with strips of clean-tilled crops. The width of the strips is a function of the steepness and length of the slope, soil characteristics, and type of crop. On steeper slopes, the strips are narrower, and the proportion of sod-forming strips is greater. Terracing is very effective in conserving soil and water, particularly in arid and semiarid regions. In these regions, terracing can be laid out exactly on the contour, and all water is held in the field. Terracing increases average soil moisture, improves crop yields, and perhaps more importantly, is very effective in controlling soil erosion. Contour furrowing pasturelands are widely used mechanical practices for soil and water conservation. Contour furrowing of pastures consists of plowing furrows on the contour, usually from 2.5 to 4.5 m apart. Contour-furrowed pastures hold more moisture than comparable unfurrowed pastures, contributing to water flow retardation. The use of structures is an effective management tool for controlling runoff and erosion from upland areas. Check dams, gabion dams, earth embankments, water-spreader dikes, and other devices constructed of permanent or temporary materials can serve specific objectives of soil and water conservation in upland areas [Jauch, 1957; Heede, 1977; Hooper et al, 1987]. In summary, upland runoff detention and retention can serve the dual purpose of conserving soil and water, while ensuring adequate subsurface water replenishment to make possible baseflow augmentation. The literature offers ample evidence that adequate upland runoff management by retention and detention eventually leads to baseflow augmentation [see, for instance, Kennon, 1966; Heede, 1977; Stabler, 1985; and Van Haveren, 1986).
Use of Instream Structures Baseflow augmentation can also be accomplished by the use of instream structures. An instream structure is broadly defined as any natural or manmade structure, either temporary or permanent, which has the effect of detaining and retaining streamflows. By retaining flows, instream structures encourage local infiltration to streambed and streambanks, contributing to subsurface and ground water replenishment. Therefore, instream structures can be effective means of augmenting baseflow [Heede, 1977; Van Haveren, 1986; DeBano and Heede, 1987]. Natural instream structures are represented by log steps and beaver dams. In small mountain streams bordered by forests, trees and logs falling across the channel create log steps, which act as check dams and influence the hydraulic geometry. Sediment accumulates upstream of the log steps, producing gentler channel gradients; in time, a waterfall develops over the log step. During high flows, the log steps are submerged, contributing to channel roughness and decreasing flow velocities. Although log steps are temporary in nature, since they rot with time, streamside forests are generally able to provide sufficient material for log step replacement. The breaching of log steps during flood events can cause severe channel degradation and needs to be evaluated carefully. Beaver dams can have some benefits for baseflow augmentation, but they are not without pitfalls. These dams change the hydraulic regime of a stream, decreasing flow velocities and encouraging the growth of riparian vegetation, which can aid in the replenishment of subsurface and ground water. They also raise water levels, increasing infiltration and moisture storage in streambed and streambanks, for eventual release as baseflow. Beaver colonies generally require an adequate renewable source of large organic material (such as willows) to build dams and dens, and to eat. On a stable, well vegetated site beaver can be a significant asset, but on a degraded site, they can eat themselves out of house and home. Unkept dams may be washed out during floods, causing major blowouts and headcutting. Manmade instream structures are either large dams, check dams, or trap dams. Large dams substantially impact all major stream functions and uses, including stream hydrology, sedimentology, ecology, biology, fisheries biology, and aquatic biota. In addition, large dams serve to moderate streamflows, decreasing the magnitude and frequency of floods and increasing the permanence of low flows [DeBano and Schmidt, in press]. In time, this moderation of streamflows can have a significant impact on the morphology of downstream reaches. Among the long-term impacts of a large dam are: (1) the establishment and growth of riparian vegetation [DeBano and Heede, 1987; Turner and Karpiscak, 1980], (2) increased bank stability, and (3) a deeper and narrower downstream channel [Smith, 1984; Lewis, 19841. Check dams are small structures built primarily for the purpose of adjusting the base level of badly degraded streams. These dams can be either temporary (log, brush, and brush-wire dams), more-or-less permanent (rock dams and gabions), or permanent concrete structures [Heede, 1965 and 19661. Large numbers of check dams have been installed successfully in many western rangelands to detain and retain runoff, control accelerated gully erosion, revegetate creeks, and restore degraded watersheds to their original (pre-impact) condition [Jauch, 1957; Heede, 1977]. Baseflow augmentation has often been an unexpected byproduct of successful watershed restoration with check dams and other water detention and retention practices. In perhaps the better known of these experiences, Heede (1977) reported on the establishment of perennial flow in a previously ephemeral creek, seven years after the construction of 133 check dams within a 2.5 square kilometer upland watershed in western Colorado. Other similar experiences with baseflow augmentation following check dam and/or small dam construction have been reported in the literature [see, for instance, Stabler, 1985; Van Haveren, 1986; and Hooper, 1987]. Trap dams are low instream structures designed to fill with coarse sediment over a period of several years following construction. The sediment deposited behind the dam serves as an artificial aquifer for the storage of flood waters and their eventual release as baseflow. Therefore, trap dams serve to moderate streamflow, augmenting baseflow for downstream uses [Baurne, 1984]. The difference between check dams and trap dams is largely one of function. The primary function of check dams is to control base levels, and therefore, arrest channel downcutting and gully erosion. Baseflow augmentation is typically a byproduct of watershed treatment with check dams, occurring several years after construction, to allow for the saturation of streambanks and streambed and of the sediment deposited behind the check dam. On the other hand, a trap dam is built primarily for the purpose of augmenting water supplies during the low flow season. The dam is carefully engineered and built to trap only coarse (sand and gravel size) sediment, enabling the formation of a saturated artificial aquifer and its functioning as early as the beginning of the second season after start of construction. While trap dams are built to satisfy local water supplies, check dams are built for control of gully erosion. In summary, the use of instream structures is a viable management tool for baseflow augmentation. However, structures are known to accelerate streamflow impacts, both positive and negative. Structural solutions can often accelerate recovery and lead to relatively quick results, but are usually more costly than comparable nonstructural management alternatives. Cost, timing of results, and the possible short- and long-term impacts need to be evaluated carefully when considering instream structures as part of a strategy for baseflow augmentation [Elmore and Beschta, 1988]. SUMMARY AND OUTLOOK Summary A literature review was performed on the subject Baseflow Augmentation by Streambank Storage. Baseflow augmentation by streambank storage refers to the temporary storage of subsurface water in floodplains, streamsides, streambanks and/or streambottom during the wet season, either by natural or artificial means, for later release during the dry season to increase the magnitude and permanence of low flows. The subject of baseflow augmentation is intrinsically related to the type of streamflow regime, whether ephemeral, intermittent, or perennial, and to the characteristics of the stream-aquifer system, whether effluent (i.e., from aquifer to stream), or influent (from stream to aquifer). Sustainable amounts of low flow appear to be possible only in streams that can remain effluent throughout the dry season. To assure that the stream remains effluent, the aquifer feeding the stream should be:
Effective aquifer replenishment leads to shallow groundwater tables, which, in aquifers of sufficient size and suitable drainage characteristics, can cause perennial flow. While aquifer replenishment is subject to management, the hydraulic properties of aquifers are largely determined by nature. Therefore, it is possible to accomplish baseflow augmentation with a management strategy based on effective replenishment of certain selected aquifers. The geometric and hydraulic properties of effluent stream-aquifer systems can be used to select aquifers suitable for baseflow augmentation. A comprehensive management strategy for baseflow augmentation should also take into account the vegetative aspects. Vegetation aids in aquifer replenishment and in raising stream base levels, thereby helping to create an environment conducive to baseflow augmentation. The experiences of Camp Creek, Sheep Creek, Alkali Creek, and Trout Creek have shown that it is possible to accomplish baseflow augmentation with a broad range of management strategies. At Camp Creek, baseflow augmentation was primarily the result of livestock grazing exclusion, to allow for the recovery of riparian vegetation. This led to channel stabilization, raised stream base levels, and raised watertables. At Sheep Creek, the construction of Sheep Creek Barrier Dam caused sediment deposition behind the dam. In time, the sediment deposit acted as an artificial aquifer, capturing and storing water during the high flow season for its eventual release during the low flow season. The Alkali Creek and Trout Creek watershed rehabilitation projects showed that baseflow augmentation is usually the byproduct of watershed treatments for the control of gully erosion. Management strategies for baseflow augmentation fall under one or a combination of the following five categories:
Range management aims to reduce the hydrologic impact of livestock grazing and related activities on the land, to maintain and enhance soil infiltrability, to encourage aquifer replenishment, and to ensure the preservation of range resources. Upland vegetation management refers to the various upland practices to manage vegetation type, density, and growth for the purpose of maintaining and/or increasing soil infiltrability and encouraging aquifer replenishment. Riparian vegetation management refers to the control of types and growth of riparian vegetation for the purpose of increasing soil infiltrability, raising stream base levels, and replenishing the aquifer. Upland runoff detention and retention refers to the agricultural and related practices in uplands to control and slow down runoff, increase ponding time, and, consequently, increase total infiltration, thus replenishing the aquifer. The use of instream structures is a feasible management tool in cases where it is desired to accelerate recovery and produce a relatively fast change in stream character, with baseflow augmentation being either a primary or secondary objective of the treatment.
Outlook for Future Research The review of the literature has shown the many instances in which baseflow augmentation has been accomplished by streambank storage. The concept is not new, but merely one that has merited little attention in times where the emphasis was on water exploitation rather than on water conservation. It now appears that the managed conversion of increasing quantities of surface runoff into subsurface runoff is a viable management strategy, and one worthy of further research and development. The benefits of streambank storage are not necessarily limited to baseflow augmentation to enhance downstream uses such as water supply and hydropower generation. There are many other benefits, among which the most important are: (1) improved erosion control, (2) improved water quality, (3) cooler water temperatures, (4) enhanced fish and wildlife habitat, (5) improved aesthetics, and (6) better recreation opportunities. Therefore, baseflow augmentation has a strong interdisciplinary flavor. A comprehensive study would have to assess costs and benefits under this multiple-resource-use framework. Selected case studies have shown that it is possible to accomplish baseflow augmentation with the proper choice of management practices. For the most part, these practices have sought to maintain or increase soil infiltrability and/or ponding time to encourage subsurface water storage and groundwater accretion. Some of these practices may be readily applicable to a given site. Yet other sites may defy their application. Site-specific studies will be necessary to determine a combination of management practices best suited to a given site. Institutional incentives will also have to be developed for the technology of baseflow augmentation to be implemented on a broad scale. This literature review has shown that the physical mechanisms and related processes governing baseflow augmentation by streambank storage are reasonably well understood. Moreover, limited field experience as documented by case studies has clearly shown the potential for the wide-ranging benefits to accrue from baseflow augmentation and related watershed management practices. The need for further research on baseflow augmentation by streambank storage is recognized. Theoretical and field research should materially contribute to advance the concept of baseflow augmentation to developmental stage. More research is needed on how to successfully integrate baseflow augmentation within comprehensive resource management strategies, given the economic, political, and institutional constraints. With the current trend from natural resource exploitation to conservation, it appears that baseflow augmentation is a concept whose time has come. |