1The authors of ‘Standards for ecologically successful restoration’ (Palmer et al. 2005) are commended for clearly articulating and discussing the nuances of establishing ecological standards in river restoration practice. We agree that there is a need for better and more thoughtful standards to elevate the practice of river restoration to a science of restoration. As practitioners of fluvial restoration, and students of the sciences of geomorphology, hydrology, ecology and plant sciences, we offer our experiences and observations, noting that the desire to achieve ecologically effective project outcomes is not new. We agree that there are valid questions regarding the application of the term ‘restoration’ to projects with minimal ecological benefits.
2We also concur that guiding image development is critical to project success. However, we note that well-intentioned projects often drift from having initially sound restoration objectives to ultimately providing reduced ecological benefit. Our observations suggest that this phenomenon can be attributed to risk aversion and the progressive incorporation of rigid engineered elements. We term this project hardening, where natural channel boundary evolution and change (deformability) is progressively sacrificed. This artificially constrains natural fluvial dynamics and associated ecological function.
3We agree that meaningful pre- and post-monitoring programmes are essential to understanding project success. We encourage the academic community to make relevant data sets available to facilitate the evolution of restoration science. It is appropriate for sponsoring agencies to require the collection of these data.
4Synthesis and applications. As practitioners we suggest that more interaction with ecologists and the larger academic community is necessary so that practical experience is communicated and integrated into the emerging science of restoration. While this interaction will advance the common goal of implementing more ecologically effective projects, we also note that project participants outside the scientific community must also appreciate the challenges a project faces in meeting higher standards. These challenges are not insignificant and include convincing project sponsors, practitioners and regulators of the need for standards in project generation, implementation and monitoring on a project-by-project basis.
As restoration practitioners we note that the concept of ecologically effective restoration is not a new project goal, but one that has been actively pursued for at least 30 years (Burgess & Bides 1980; Gore 1985). Yet real questions remain concerning our abilities to achieve this goal. In this paper we respond to Palmer et al.'s (2005) call for the adoption of ecological restoration standards and how they relate to projects that involve alteration of existing channel geomorphology and natural channel design. We offer our perspective on cultural and institutional factors that impede ecologically effective restoration, including: (i) misuse of the term restoration; (ii) failure to create sound, ecologically based guiding images; (iii) a lack of risk tolerance required to achieve restoration; (iv) practitioner/sponsor inexperience and inflexibility; and (v) lack of commitment to monitoring. Our hope is that this discussion of typical project processes and failings will underscore the need for widely accepted and practised use of standards such as are proposed by Palmer et al. (2005) and expanded upon by Jansson et al. (2005).
Why the need for ecological standards?
Palmer et al. (2005) address the problem of the inappropriate application of the word restoration to projects with little ecological value. We agree that true restoration projects are too often lumped together with the plethora of stream and river manipulation projects that have co-opted the word restoration without implementing the principles (Sear 1994). Changing this will require not only an academic understanding of restoration criteria and standards, but also a fuller understanding among practitioners and sponsors.
Figure 1 illustrates the continuum of project types within a subset of stream and river restoration projects that emphasize the alteration of geomorphic variables by active manipulation of the channel plan form, bed and banks. The continuum moves from projects targeting true ecological goals and opportunities in pristine settings to erosion control and containment efforts, such as bank stabilization projects in highly constrained urban channels, where channel migration is not tolerated. While some projects will have ecologically effective outcomes, many others will result in little improvement for habitat quality or ecosystem function.
We suggest that by increasing factors of safety and incorporating large immobile structures, a project is driven away from ecological restoration towards the category of erosion control and containment. We feel that ecological restoration techniques must recognize and anticipate the fact that natural channels have largely deformable boundaries where the movement and position of channel features, such as banks, riffles and pools, is expected to change through space and time (Miller & Skidmore 1998; Cramer et al. 2003). Projects relying on fixed, in-place structures that resist or impede natural channel evolution contradict natural channel behaviour, which is a component of ecological integrity. Restoration projects should utilize natural processes to create and maintain habitat, and should favour the use of native materials. A project that introduces materials uncharacteristic of the natural environment to create habitat (such as adding boulders to a fine grained stream to cause pool scour) is therefore considered an enhancement project rather than a restoration project. A principal example of enhancement work is the placement of boulders and logs in various combinations to prevent bank scour and lateral channel adjustments. We argue that these techniques have little long-term resilience and plasticity in a dynamic environment, do not work with natural channel processes to affect recovery of natural habitat and channel form, and are therefore not sustainable. This is not to say that projects utilizing rigid or non-deformable components are not worthy, but rather that they may not be considered accurately as restoration projects.
While we recognize that project types can overlap considerably, we maintain that if sponsors and practitioners can objectively place their project along this continuum as part of the guiding image process, many benefits will be realized. Once the project is characterized, acceptable levels of risk can be determined, which provides the basis for the selection of appropriate techniques (Gillilan 1996). Project categorization early in the design phase will help sponsors understand the relative ecological value of their project. Project type description will also promote the development of project databases and the application of appropriate ecological standards to evaluate success.
Proposed standard 1: a guiding image exists
The concept of launching a river restoration project by first establishing a guiding image with a dynamic ecological end state is the most critical aspect of a restoration project. In the real world of project development, guiding image development is far too often abbreviated, restricted by preconceived ideas and self-interest, hampered by lack of experience, or stripped of original intent by either bureaucratic process or sponsor/stakeholder reluctance to accept uncertainty and risk, leading to a number of poorly conceived projects.
The creation of a sound, ecologically based guiding image requires a challenging level of critical thought and collaboration between practitioner, sponsor and stakeholders. It also requires a relatively solid understanding of the fluvial sciences in addition to a working knowledge of a broad range of restoration techniques (Sear 1994). There are several common pitfalls in developing a guiding image.
When sponsors do not have the scientific and practical knowledge to develop a guiding image, they frequently turn to the practitioner community for this work. While the practitioner community has had some success, it has also failed by repeatedly allowing projects to devolve into ones with little ecological value. The most common practitioner-based failings in guiding image development include: (i) minimizing project complexities because of a lack of practical experience, which is commonly exacerbated by a lack of familiarity with the existing body of relevant research; (ii) self-interest in maximizing the role of the practitioner in the project for economic gain; and (iii) a limited knowledge of available restoration techniques (Malakoff 2004).
As practitioners, we have all been guilty of selecting specific bank/channel stabilization techniques before the guiding image has been fully developed. For example, a project to restore salmonid habitat may fail to include an investigation of the causes of degradation in the context of historic and current fish population data. The practitioner moves directly to a vision of a channel with root wads or boulders to ‘provide cover’. If cover indeed limited the population under discussion, a more appropriate guiding image discussion would consider why the cover is absent, how analogous healthy natural systems should look (undercut banks, overhanging vegetation, aquatic macrophytes, pool depth, etc.) and finally how to restore the natural processes that will then sustain the development and maintenance of these attributes. Perhaps this is where Jansson et al.'s (2005) proposed sixth criterion would be useful, by requiring that practitioners provide an explicit explanation and testable hypotheses for how the proposed restoration techniques will achieve the stated guiding image.
Practitioner self-interest also regularly subverts the guiding image development process. In a truly open and critical image development process, all participants would be open to outcomes that may not be expected, and that may result in reduced involvement for an individual practitioner. Too often we have observed preconceived image development, with practitioners either consciously or unconsciously promoting approaches, personnel, methods and techniques in which they will have the most active role. This may result in the exclusion of relevant, more cost-effective approaches, and preferential inclusion of techniques in which practitioners hold proprietary or intellectual interest.
While practitioners and academics typically appreciate the process of melding theory and practice into a successful image and then project, the project sponsor community and stakeholders often do not have the training or the resources to commit to this process. One of the more common difficulties in the guiding image development process is the need to secure funding for a project before practitioners can be hired to help define the image. In this case, the sponsor develops a guiding image without critical input and this becomes the de facto guiding image even when subsequent scientific investigation may suggest a different image. An example of this may be a sponsor's persistent intent to ‘restore’ a river after a large flood by trying to re-establish pre-flood conditions, even following a qualified professional evaluation that identifies natural recovery and changes in water management as the most effective restoration approaches.
guiding image drift
The adoption of a poorly articulated guiding image commonly results in ‘guiding image drift’. Given a poorly conceived guiding image, goals of ecological restoration lose their original priority in the project design and implementation process, resulting in projects that fall short of ecological objectives. Such a departure from the original ecological intent is commonly manifested as ‘unintentional hardening’ of a project. This shift is shown in Fig. 1 with risk aversion moving projects away from fluvial restoration techniques with deformable components, towards containment-type projects with increasing factors of safety, reduced risk and reduced deformability.
The primary reasons for unforeseen project hardening include the following.
1A tendency to reduce project risk continually by increasing factors of safety during the design process. In a natural channel environment with unpredictable disturbance regimes, sponsors are commonly averse to adopting the level of risk necessary for true ecological restoration. A low level of risk tolerance, such as the commonly adopted 100-year flood, is such a high threshold that it almost always forces projects away from restoration techniques that anticipate channel changes of shape through time.
2Tight deadlines and/or budget constraints that preclude the development and adoption of natural process-driven alternatives.
3Bureaucratic requirements that, because of established and dogmatic patterns of practice, reduce opportunities for true restoration design, resulting in hard engineering solutions through precedent.
4Over-reaction to even moderate channel adjustments in a restored channel often leads to overly aggressive ‘site repair’. This hardening can be the result of the initial adoption of an unrealistic guiding image that anticipated full ecological benefits without acceptance of any risk of project distortion, adjustment or failure during a channel-changing flow event.
Proposed standard 2: ecosystems are improved
The standard of measurable ecosystem improvement is a laudatory one that should be embraced by the restoration community, but with the caution that we perhaps do not yet fully understand how even stable fluvial systems function geomorphically (Gurnell & Petts 1995), much less from the aquatic ecology perspective. The measurement of ecosystem improvement inherently relies on the comparison of pre- and post-project conditions. However, the interpretation of baseline data can be difficult, given that it is usually a snapshot of conditions.
The clear need for a systematic quantification of the effects of fluvial manipulation projects on ecosystem function points to the compelling and often overlooked need to gather physical and biological data where projects are proposed. As practitioners we share the discouraging reality that project sponsors seldom have the baseline conditions identified prior to project inception and are often either budgetarily limited, overly impatient or generally disinterested in collecting these data. Without adequate baseline data, a project will always become an anecdotal experiment that contributes little to scientific understanding.
Proposed standard 3: resiliency is increased
Resiliency in fluvial environments reflects the ability of a stream to recover from natural disturbances, such as flood events. In certain conditions, increased resiliency and sustainability can be achieved with a passive approach to restoration; that is, the promotion of natural recovery through the implementation of land-use planning tools and best management practices. These passive approaches, which are often overlooked, have tremendous potential to allow for long-term system function. Specific approaches to these concepts include zoning ordinances based on long-term river processes, agricultural best management practices, and land acquisition and/or conservation strategies (FISRWG 1998; Skidmore, Cooper & Boyd 1999; Washington Department of Ecology 2003). A primary benefit of these passive approaches is the ability to improve extensive portions of drainage systems in a cost-effective manner. Our experience indicates that public support for land-use planning approaches is variable. With time, however, a growing suite of successful examples will probably increase the feasibility of passive restoration approaches.
Regarding the measurement of resilience and self-sustainability (Jansson et al. 2005), it is suggested that the rates of change in a reference channel reach can be compared with the performance of these attributes in the restored channel. For example, channel bank erosion, avulsion, thalweg migration and bank accretion rates in a restored channel can be compared with those found in a reference channel experiencing a similar disturbance regime. An ecologically effective restoration project should be expected to be as resilient and sustainable as a reference channel responding to a similar hydrologic event. In general, we need to understand the natural variability in channels over time to understand how to establish envelopes for acceptable project performance.
Proposed standard 4: no lasting harm is done
Again we stress the importance (and frequent lack) of pre- and post-project monitoring both on-site and downstream as appropriate. Without pre-project monitoring, the impacts of project construction cannot be measured meaningfully, and questions about impacts cannot be addressed scientifically (Frissell & Ralph 1998). For example, claims of increased downstream sedimentation impacts because of short-term turbidity from project construction cannot be addressed without a monitoring programme. While regulatory agencies have concerns about impacts, the lack of data and monitoring of impacts resulting from various construction techniques hampers constructive dialogue and the development of appropriate project controls.
Proposed standard 5 : ecological assessment is completed
While the planning process for restoration projects typically calls for post-construction monitoring, usually only physical attributes are monitored. This may be the result of a sponsor's lack of appreciation of the need to assess ecological impacts, a lack of funds or indifference. All too often, project resources are so drained by the end of construction that prescribed monitoring efforts are abandoned.
The practice of fluvial restoration design commonly accesses empirical data sets related to channel form, hydraulic geometry and system hydrology. Unfortunately, the restoration community has so far failed to generate similar data sets for project impacts. We applaud the efforts of Palmer et al. (2005) and others to establish regional, national and international standards and restoration databases, and the regular evaluation of project success. To that end, we encourage academic institutions to provide leadership in the compilation of objective, unbiased monitoring data, and stratification of that data in terms of project type (Fig. 1) and regional setting. Additionally, government agency sponsors may initially be the most effective in promoting the collection of relevant monitoring data. From a practitioner's standpoint, the usability of the monitoring data and consequent application of that data to restoration objectives requires that it be easily accessible and consistently organized to facilitate data query and extraction.
We do not believe that the sciences relating to fluvial systems have yet coalesced into a science of river restoration, although the ideas raised by Palmer et al. (2005) are important steps in this direction. We recognize that the achievement of sound ecological standards of performance will be extremely difficult. To help improve our standards of practice, we encourage the academic community to continue their efforts towards the compilation and interpretation of a vast array of projects in many different geographical settings. An interdisciplinary study of the large body of work in place will shed light on appropriate river restoration practices and procedures, and will generate trained restoration scientists. The practice of river restoration will also benefit from a heightened awareness within the project sponsor community of the complexity of restoration principles, appropriate monitoring needs, timeframes and the costs in meeting scientific standards.
We strongly encourage stakeholders to evaluate their project and place it in the continuum of project types (Fig. 1). By objectively categorizing their project, stakeholders will consider the relative ecological benefit of their project, more clearly separating restoration projects from enhancement and containment projects. Hopefully this distinction will aid in directing funding towards projects with ecological restoration outcomes. We also strongly recommend that funding agencies take a more aggressive stance with respect to the requirement of meaningful monitoring on fluvial manipulation projects. If not required by regulation, few project sponsors will commit to effective monitoring, and the restoration community will continue to be awash in projects with anecdotal outcomes.