Purpose and Objective

Baseflow augmentation by streambank storage refers to the temporary storage of subsurface water, either by natural or artificial means, in floodplain aquifers, stream sides, streambanks, and/or stream bottom during the wet season to increase the magnitude and duration of dry season flows (Ponce and Lindquist 1990). This type of storage can also be used to effect a change in the hydrologic character of a stream—from one that flows intermittently (seasonally) to one that flows perennially (year-round). The temporary storage of precipitation in soils and groundwater adjacent to streams can directly benefit the landowner and downstream water users by establishing year-round flow, which reduces site erosion and runoff caused by floods and promotes healthy stands of riparian and meadow vegetation (Elmore and Beschta 1987). Vegetation cover is critical to site stability and maintenance of meadow function, which are necessary prerequisites for establishing baseflow augmentation.

The objective of this study was to build a knowledge base for baseflow augmentation and use this information to develop a study plan for quantitative assessment of this phenomenon in a mountain meadow setting. Increasing stream baseflows, especially in the dry summer months, is one of the goals listed in the charter of the Feather River CRM.

Methods

A literature review of baseflow augmentation by streambank storage, including theory and case studies, was initiated by PG&E in 1988 to provide the CRM Group with more background on stream recovery and the role of streambank moisture in maintaining a healthy stream system (Ponce 1989). Computer searches and other conventional means were used to identify literature documenting cases of baseflow augmentation. This effort identified 138 journal papers, research and technical reports, and other published and unpublished articles and reports. Of those, four case studies of baseflow augmentation were selected and reviewed in detail (Winegar 1977; Hooper, Van Haversen, and Jackson 1987; Heide 1977; Jauch 1957). All were located in western states other than California. Additional information about the case studies was obtained in a series of interviews conducted in California, Oregon, and Wyoming with professionals and practitioners who had extensive field knowledge. This information was used to develop a study plan to test baseflow augmentation in a small watershed (Ponce and Lindquist 1990).

Key Findings

• Physical mechanisms and related processes governing baseflow augmentation by streambank storage are reasonably well understood, although quantitative supporting data are not available. In the past, riparian vegetation was considered a nuisance and was thought to consume large amounts of valuable water and restrict channel capacity to pass flood flows. Removal of riparian vegetation left streambanks and adjacent meadows vulnerable to erosion and caused a shift from more moderated to "flashy" runoff patterns.
• Riparian vegetation reduces stream flow velocity and enhances aquifer storage of water. The reduced velocity increases the rate of subsurface moisture replenishment during high flows by increasing the infiltration capacity and may permanently raise the water table in the shallow floodplain aquifer. The raised water table and increased aquifer storage in meadows can change the character of the stream from intermittent to perennial.
• Four case studies show that it is possible to accomplish baseflow augmentation with a broad range of management strategies. In all cases, recovery of riparian vegetation led to channel stabilization, increased streambed levels, raised water tables, and enhanced meadow storage function. In time, adequate aquifer replenishment led to release of stored water during low-flow periods, converting the intermittent streams to perennial. Many of the gradual changes that occurred in these cases were similar to changes observed in the Red Clover Creek study area. Unfortunately, very few Sierra Nevada examples were discovered, and none of the case studies reviewed took place in California.
• That these same results were achieved using a variety of methods—from exclusion of grazing to revegetation and construction of instream barriers (check dams)—suggests that sound management practices designed to initiate and maintain vegetation are the key to a successful project. Appropriate management strategies will vary based on stream characteristics and condition, so careful planning is necessary.
• Case studies and anecdotal information indicate that successful baseflow augmentation may not yield more water volume in any given year, but it will change the pattern and timing of runoff, thereby reducing damaging peak flows and increasing beneficial summer flows. Maintaining vegetation requires water but is critical to increase meadow infiltration and aquifer storage. Research is needed to quantify these qualitative findings and determine their validity and potential application in California, specifically in the Sierra Nevada.
• A study plan was developed for the Big Flat/Cottonwood Creek Project in Plumas National Forest. The plan was partially implemented by FRCRM and the USDA Forest Service in 1994, when a series of wells and stream gages were installed to collect baseline flow and water table data. Construction of the project began in 1995, and monitoring continues. Preliminary results will be available in spring 1998.

The Red Clover Creek Erosion Control Demonstration Project successfully achieved its objectives by testing and evaluating several low-cost structural and nonstructural methods for controlling erosion and by developing an organizational process to coordinate the contributions of various stakeholders in carrying out future projects. The project benefited the CRM organization in several ways.

• It provided an opportunity to test and fine-tune the organizational process.
• It increased the visibility and credibility of CRM objectives to agencies and other potential stakeholders.
• It demonstrated to local landowners the potential benefits of restoration.
• It provided a "living laboratory" for evaluating the effectiveness of several techniques that could be implemented more broadly in the watershed.

Institutional Process

• Establishment of an institutional framework using Coordinated Resource Management was a critical factor for the success of the demonstration project. The CRM organization has successfully implemented thirty-three restoration projects in the Feather River watershed using the framework developed in the demonstration project.
• To minimize variability in results and to maintain control of the study area during a research project, long-term agreements or contracts with the landowner need to be completed prior to initiation of the study. In this case, most of the resulting variability could have been avoided if more care had been taken in selecting a site where such agreements could be negotiated.
• The restoration design of natural streams must be tempered with practical experience and empirical observations of natural stream behavior. The design process will not be perfect or consistent because every stream and location provides unique design challenges. It is to be expected that portions of reconstructed streams may suffer damage when tested by high flows. These damages should not be viewed as project failures, but rather as learning experiences and opportunities to test new solutions. It is only through trial and error that teams of specialists will learn the intricacies of natural stream design. When problems occur, they are systematically analyzed by the interdisciplinary team, and repairs incorporate better understanding of how nature works. It is expected that the success of subsequent projects will increase on their first test with a significant flood flow.

• The success of revegetation efforts depends on such factors as a realistic plan, healthy planting materials, feasible planting techniques, location, and timing. The demonstration project had best results with unrooted local willow cuttings (stakes) planted during the fall in high moisture areas. Distance to available water, not soil substrate, was the most critical factor for unrooted stakes. Rooted liners showed less success and were more vulnerable to drying out. Commercial seed mixes established more quickly than native mixes, but both mixes were doing well after three years.
• A significant proportion of the plantings was lost to severe drought conditions and streambank sloughing as the channel began to stabilize. Special consideration should be taken to postpone revegetation efforts until channel improvements have time to stabilize and moisture regimes are established. It is believed that vegetation is accelerating the healing process in the channel and is critical to reversing the degradation process.

• The dams created an artificial recharge system that raised the shallow water table over a short period of time. We found a noticeable difference in the shallow aquifer gradient attributable to the effect of the ponds on the local groundwater regime compared to control areas. The effect is specific to the area adjacent to the ponds and occurs on a seasonal basis. Variations in the gradient appear to be due to seasonal changes in the pond level, streamflow fluctuations, and infiltration from local runoff. During summer, when groundwater levels were lowest, the dams were able to recharge the water table. Over time, depth-to-groundwater levels decreased in the demonstration area and remained stable in the control area. Decreased depth to groundwater in the test area made more water available for establishment of riparian species.
• Current data indicate that water flows from the ponds into the aquifer in summer. Whether this trend is later reversed (once vegetation and soil stabilization occurs) and augments baseflows (as demonstrated in Oregon) will require further study.
• Springs and near-surface ponds in the floodplain (more than 200 feet from the channel) also recharged, creating more forage for wildlife and livestock and more cover to reduce erosion.

• Installation of the check dams clearly reduced further downcutting of the stream channel and slowed bank sloughing in the study area. Stream profile monitoring and photographic documentation showed that the channel was narrower and more sinuous and that overall sediment deposition was increasing, especially in point bars and along the toe of the bank.
• Unanticipated changes in grazing management in the downstream control area tended to confound research results because channel stability was promoted. The channel narrowed dramatically, and streambank vegetation was reestablished without the aid of check dams, only at a much slower pace. Grazing management may provide more benefits at a lower cost than check dams and intensive restoration efforts if time is not an issue and if changes in land management are feasible over the long term.
• Through qualitative observations in 1993, we conclude that the raw vertical streambanks within the demonstration site are revegetating and starting to heal. Some sloughing of vertical streambanks continues, but it is believed the recession rate is much slower than pre-project conditions. Sediment is depositing along the bank toe in the ponds, providing a substrate for the establishment of vegetation. Continued improvement is expected as long as an appropriate grazing management strategy is adopted.
• Watershed-level cross-sectional data indicate that Red Clover Creek is in transition, and channel instability is therefore likely to continue in the short term. The upper site is expected to reach equilibrium and stabilize as long as grazing pressure is reduced. The lower end is more problematic due to convergence with Chase Creek and continued grazing pressure.

• Natural vegetation recovery was a success largely due to the elevated water table and exclusion fencing. We observed a direct correlation between vegetation cover and water table levels both on streambanks and in the adjacent floodplain. Increased plant cover protects soils from erosion, increases infiltration and meadow storage, and enhances habitat for wildlife.
• Check dams reduced average depth to groundwater in areas adjacent to the ponds. This set the stage for rapid growth of vegetation—especially evident at Ponds #3 and #4, the two upstream impoundments (Figure 6-1; Sagraves 1991). In contrast, the depth to groundwater in control areas remained relatively stable but had large depth-to-groundwater values (Figure 6-5; Sagraves 1994), which continued to support sparse dry-site vegetation.
• Overall, trends indicate that the check dams had a positive effect on vegetation growth and cover, especially at stations closer to the stream.
• In spite of prolonged drought conditions, comparisons of the test and control sites over time indicate that the areas influenced by the check dams increased in diversity (more species), increased in vegetation cover, and were generally more productive than control areas. Within the test area, vegetation in high water table zones (wet sites) was more productive and had higher cover and more wet site species than areas where the water table elevation was lower (dryer sites). See Figure 8-5.
• A transition period or lag time for vegetation stabilization may occur after a restoration project is completed. Rapid improvement of the soil moisture regime may cause the dryer site vegetation to die more rapidly than the wet site vegetation can colonize, resulting in an overall reduction in cover for the first two years.
• Annual variation in vegetation data was relatively high during the study period, preventing statistical evidence of restoration success in some cases. Variation was due in part to the amount and periodicity of precipitation (different rainfall patterns tend to encourage different plant types and species); annual temperature variation; changes in grazing management outside the exclosure; the use of three different botanists to identify plants in the field; phenological differences from year to year, which in some years made it more difficult to identify certain species; and beaver activity, which caused the water table to rise above Check Dam #4 during two study years.

• Vegetation community changes over time were further analyzed from aerial infrared photographs and a GIS, which showed a visible trend toward conversion to a willow/sedge community from a sagebrush-type community in areas influenced by the dams. The expected conversion to a dominant wet-site vegetation community can be readily documented by both aerial and surface photographs. Vegetation in downstream control areas remained dominated by sparse, drought-tolerant vegetation.

• Adult fish populations in the pools created by the check dams increased dramatically during the first years of the demonstration and continued to be higher than the downstream control during this study. Due to unanticipated grazing management changes that promoted vegetation recovery and channel narrowing, fish habitat in the control area improved during the latter half of the study. Increases in trout numbers in the test area were attributable to recruitment from adjacent areas, probably in response to cooler water and better cover retained in the deep pools. The low occurrence of young-of-the-year trout indicates that little or no reproduction occurred in the control or test areas. Although the check dams provided locally improved adult trout habitat within the study area, a much more extensive effort would be required to significantly improve Red Clover Creek for all life stages.

• Water quality parameters appeared to be generally within the range for good trout habitat, but in some cases, dissolved oxygen was found to be lower and pH higher than optimal in the test and control areas. High summer water temperatures in both the control stream section and test ponds may limit the potential for trout habitat improvement.

• Wildlife species density and richness showed a significant increase in the demonstration area. Waterfowl usage and reproduction in the project area increased 700% over the control site. Deer frequented the site, whereas sampling did not detect deer use in the control area. Stream restoration methods that rely on vegetation management alone could not produce these waterfowl benefits in the same time interval (Bogener 1993). Waterfowl populations quickly benefited from the raised water table and grazing exclusion and were good indicators of short-term habitat enhancement.

• Photographic monitoring was valuable in providing a visual means to qualitatively document and assess trends occurring in the demonstration area, especially in the absence of statistically significant data. Before and after comparisons at the same station along the creek in the test area have been used effectively in presentations and posters to emphasize the benefits of water table recharge.

• The literature review indicated that baseflow augmentation by streambank storage was successful at restoration projects similar to the demonstration project. A similar process was documented in all cases, which included:

• A fully functioning meadow aquifer modifies runoff patterns and protects soils and vegetation from potentially damaging flood flows. At the same time, it provides more flow in the dry summer months for fish and downstream water users.

Though research findings were largely qualitative, much value can be derived from monitoring a natural system over a long period of time in a qualitative context. The Red Clover Creek "living laboratory" provides a rare opportunity to view these effects over a ten-year period and look at the response of natural systems to changes catalyzed by the demonstration project. Establishing and documenting long-term trends helps to explain natural variation and better understand the response of similar systems to improvement measures. Permanent streamside photo points and periodic aerial photographs provide a cost-effective way to monitor changes over large geographic areas and over long periods of time; they should be routinely used in watershed improvement projects.

Overall, the combined results from all studies indicate that check dams were effective in stabilizing the Red Clover Creek demonstration site, providing more diverse and desirable habitat for fish and wildlife, improving hydrologic characteristics of the channel, raising the shallow water table and water storage potential, enhancing vegetation cover, and initiating vegetation conversion from xeric to mesic site species.

Some issues need to be considered in conjunction with the benefits:

• Grazing management upsteam and downstream of the demonstration site may impact the sustainability and stability of the project. 
• Check dams are relatively costly, they must be maintained to avoid failure, and they do not provide good habitat for spawning and rearing of trout. Check dams are not a panacea for enhancing degraded stream systems, but they can be used effectively in the right situation to quickly recharge shallow water tables and stimulate the growth of vegetation.

Based on research results, observations, successes and failures, field experience, unanticipated constraints, interaction with the landowner, and advice from CRM supporters, the following is a list of the most significant lessons learned from ten years of research at Red Clover Creek. This information may be useful to others planning and executing similar stream improvement projects in the future.

CRM Process

• Establishment of the CRM process was an unqualified success. All stakeholders must be involved in the process up front to ensure that all views are represented, to nurture a sense of ownership and camaraderie, and to ensure that issues are resolved in the preliminary stages of planning.
• Big-picture regional approaches to watershed management can be facilitated by the CRM process and are recommended if long-term sustainable improvement is the goal.
• There is no substitute for good planning and careful consideration of actions to be taken in securing a successful project. Since each project is different—including objectives, site characteristics, and ownership issues—planning is an essential ingredient for successful implementation and long-term stakeholder support.
• Multidisciplinary teams representing a broad array of scientific expertise are necessary to ensure that technical oversight is thorough and technical objectives are met. Working together to achieve project goals is essential for success.

• Check dams are expensive and require regular maintenance. Although they are effective in providing immediate benefits to vegetation and waterfowl, they are one of many options and should be used only when appropriate. Dams must be designed and sited with care to maintain site stability and to achieve overall project goals. A long-term maintenance plan, including grazing management, should be specified up front in the planning process.
• Proper grazing management, revegetation, and exclusion fencing alone can dramatically improve degraded channels at some sites with minimal site disturbance. This was observed at the control site for the fish study, where channel narrowing, increased riparian vegetation, and reduced channel erosion resulted from changes in grazing management. These low-intensity treatments should be considered in the design process prior to selecting more expensive, higher-risk fixes, or they should be used in combination with structural measures.

• Study goals and objectives should be well thought out and focused to address realistic questions. Projects conducted outside the laboratory setting are difficult to control; maintaining consistency and minimizing variation is therefore a challenge. Clear, defined, and feasible objectives will help, along with contingency planning for unanticipated changes that affect the study design.
• To maximize the value of results, baseline data collection is essential. It is difficult to draw conclusions without pre-project data. Comparison between test and control sites provides good trend data, but the data are less informative and it requires long-term investigation to reduce variation and obtain usable results. Moreover, identifying uniform sites for comparison is not easily accomplished.
• In some cases, annual monitoring is not warranted once an annual pattern is established and if funding is limited. Channel geomorphology, aerial mapping, and fish population monitoring could be limited to high runoff years to document change and minimize resource disturbance. More frequent sampling is necessary to monitor water table levels—sampling at least bimonthly provides enough data points to analyze annual and seasonal variations. Vegetation recovery and fish and wildlife populations should be monitored at the same time every year to expedite understanding of yearly variation.
• Permanent photo points provide valuable monitoring information and can be used effectively to qualitatively document trends over time.
• Low-altitude aerial photography (especially infrared) is an effective, low-cost way to monitor over time the response of vegetation, water table, and channel shape to erosion control projects such as the Red Clover Creek demonstration. Aerial photographs taken at three- to five-year intervals are sufficient to monitor trends. GIS (to store, compare, and analyze data) and high-resolution satellite imagery could be used effectively to monitor long-term trends.
• More comprehensive geotechnical studies are needed to fully understand groundwater recharge dynamics in response to check dam installations.
• Long-term monitoring is required to detect changes from the variation in the data from year to year and to provide an opportunity for vegetation changes to stabilize. Changes in vegetation most likely continue for a number of years after project implementation. To achieve meaningful long-term trend analysis, more data points spanning a longer period of time are needed to show statistically significant results.
• Events that occurred during the study—beaver activity (submerging control wells and altering fish habitat), changes in grazing management, climatic variability, drought, and the duration of monitoring—made it difficult to draw quantitative conclusions in some cases. These factors should be taken into account when planning watershed restoration research studies. Additional sampling data points are needed to overcome variation introduced by these uncontrollable factors and to produce conclusive results.
• Channel cross-sectional monitoring is an important element for establishing geomorphological trends. Laser-based techniques, if available, are recommended over traditional survey techniques, especially if the objective is to determine the magnitude of aggradation and/or degradation. If traditional survey techniques are used, ensure that adequate points are taken and that monitoring includes data points covering the channel, the streambanks, and partially out into the floodplain.

• Locally collected unrooted willows harvested and planted in the fall (when plants are dormant) is recommended. Stakes should be planted within 3 feet of available water for best results.
• Pilot holes should be dug prior to planting unrooted stakes, especially in rocky soils, to avoid bark damage.
• Willow stakes should be soaked for 48 hours before planting to dissolve rooting inhibitor hormones.
• Unrooted stakes should be cut 24 inches or longer to facilitate reaching available water.
• Monitoring should include at least three growing seasons to determine planting success of unrooted willow stakes.
• Planting stock should be collected from the project site if possible, or it should be acclimated before planting to produce the best results.
• Planting in the channel or along the banks should be avoided for two to three years so that lateral bank sloughing stabilizes where check dams are installed.
• Although native seed mixes tend to germinate slower than commercial seed mixes, they do provide protection from erosion once established and serve to reintroduce native species into the area. Locally collected, good condition seed is desirable for best results. Planting native species is encouraged and should be routinely considered with other options when planning improvement projects.

• The raised water table created very dramatic increases in nesting waterfowl in a short period of time—a good indicator of site recovery.
• Ponds created by check dams create good adult trout habitat, but spawning and rearing habitat are generally lacking. To enhance the habitat for all life stages, it is important to apply a much broader approach that addresses problems and provides solutions for the whole stream system. Land disturbance that is allowed to continue downstream and upstream of the project area will continue to inhibit improvements for fisheries and establishment of a fully functional riparian system.
• Beaver activity can have a detrimental effect on vegetation and soils in a degraded stream system if hardwoods such as willow are in short supply. Relocating beaver until vegetation and soils stabilize is recommended to expedite site recovery.
Table of Contents