4. Prioritizing rehabilitation activities

Our review of the effectiveness of different watershed rehabilitation techniques was undertaken in part to help determine which techniques might be most useful. Clearly many of the techniques we reviewed succeed or fail for similar reasons (Table 17). Most notably the actions that fail either ignore large scale-watershed processes such as sediment transport or basic fish needs such as water quality and adequate instream flow. Below we provide some guidance on planning and prioritizing rehabilitation actions based on our review. We emphasizes the need to restore basic processes and address factors limiting fish production and identify other factors to consider when more detailed information is available for prioritizing rehabilitation.

TABLE 16
Average cost and range of cost of selected habitat rehabilitation activities implemented under the Pacific Coastal Salmon Recovery Fund between 2000 and 2003 in the northwest United States and throughout the United States between 1990 and 2003 as reported by the National River Restoration Science Synthesis Project (www.nrrss.umd.edu); Berhardt et al., 2005). The variation in costs regionally, nationally and among project types highlights the problems with accurate cost reporting and the range of costs among projects. Reported costs per project ranged from a few thousand dollars to several million and some projects covered more than one project type so costs may be inflated.

NA = not available

4.1 Determining what to rehabilitate

Two types of assessments or questions must be answered to identify necessary habitat rehabilitation actions and assist in planning and prioritization (Figure 12). The first group of assessments focuses on identifying disruptions to ecosystem function and the types of habitat rehabilitation necessary for ecosystem recovery. The second set of assessments concentrates on how humans have altered habitats and how those habitat changes have affected biota. Addressing these questions requires analysis of natural functioning of aquatic ecosystems (often including assessment of historical habitat types and abundance), as well as assessments of relationships between habitat and biota (Habersack, 2000; Beechie et al., 2003a). These questions motivate assessments that identify where the biological integrity of ecosystems has been degraded and where specific ecosystem processes or functions are disrupted. In combination, these assessments provide a broad understanding of actions that are likely to improve the functioning of aquatic ecosystems, and form the basis of both regional and site-specific plans for ecosystem restoration.

Assessments of ecosystem functions and biological integrity can be separated into screening assessments that identify areas where ecosystem processes and functions are most impaired, and specific field inventories to diagnose causes of ecosystem impairment and opportunities for rehabilitation. Assessments that correlate landscape and land use characteristics with population attributes can indicate which habitat changes are most likely responsible for declines in specific organisms, and therefore which broad categories of rehabilitation actions are most likely to result in improved ecosystem functioning. Direct assessments of ecosystem processes that form aquatic habitats (e.g. barrier inventories, riparian condition inventories) identify causes of degradation, as well as rehabilitation actions that are required to recover ecosystem functions and biological integrity (Table 18).

TABLE 17
Summary of number of evaluations located on effectiveness of specific techniques, common parameters monitored and common limiting factors for each rehabilitation technique reviewed. Some studies examined more than one type of rehabilitation (i.e., placement of cover logs and spawning gravel).

TABLE 18
Examples of landscape processes and functions that should be addressed in planning watershed and river restoration.

Analysis of habitat change and influences on biota help predict fish and biota response to habitat changes. Common methods include assessment of habitat change to estimate or model changes in fish populations or other biota (e.g. Beechie et al.,1994, Kemp et al.,1999, Muhar et al.,2000), and correlation analyses that relate landscape and land use characteristics to fish populations and communities without directly quantifying changes to habitat (e.g. Feist et al., 2003, Steel et al., 2004, Filipe et al., 2004). It should be noted that these type of analyses do not directly identify causes of habitat degradation or specific rehabilitation actions. Rather they provide a means of predicting ecosystem responses to rehabilitation actions (e.g. habitat-based estimates of potential population size for specific organisms), insights into potential changes in species diversity or life history diversity, and a means of identifying which species or populations are most constrained by habitat loss and therefore may be most difficult to recover (Beechie et al., 2003b, Filipe et al., 2004, McHugh et al., 2004). Thus, when combined with impaired processes and potential rehabilitation actions developed from the assessments of ecosystem functions and biological integrity they assist in the planning and prioritization of rehabilitation.

4.2 Prioritizing actions

Once the previously described assessments have been completed and a list of potential actions has been established there are several approaches for prioritizing rehabilitation actions depending on the level of information available. Based on our understanding of watershed processes, basic needs of aquatic biota, and our review of effectives of techniques, we recommend an interim approach for setting rehabilitation priorities. This begins first with protecting high quality habitats and providing adequate water quality and quantity and is then followed either sequentially or concurrently by restoring connectivity of habitats, restoring processes that create and maintain habitat, and finally, if necessary, instream habitat improvements (Figure 13). We begin this section by discussing this interim approach and then outlining other approaches for prioritizing rehabilitation.

Habitat protection is typically not seen as rehabilitation. However, given the loss of habitat and the continued degradation from human activities, protecting high quality functioning habitats should be part of the rehabilitation planning process (NRC, 1992; Roni et al., 2002; Lucchetti et al., 2005). As discussed in previous sections, it is also considered more cost-effective and reliable than trying to rehabilitate habitat once it has been degraded.

Adequate water quality and quantity are fundamental to the success of any rehabilitation action or suite of actions. Our review of published data and anecdotal evidence suggests that inadequate water quality and quantity have prevented many expensive rehabilitation projects from achieving successful increases in fish production. In some cases adequate water quality can be achieved through riparian protection or replanting, such as in the case of agricultural runoff (Sovell et al., 2000; Meals and Hopkins, 2002; Parkyn et al., 2003). However, pollutants often originate from both point and nonpoint sources including runoff from roads, treated or untreated sewage effluent, or runoff from agricultural lands and will require much more complicated measures to improve water quality. Water quantity is typically addressed by limiting the timing and volume of water withdrawals or water use. Lengthy political, legal, or legislative acts may be needed to restore water quantity and quality. While both water quality and quantity can be difficult to address, they are critical to the success of other rehabilitation actions.

Assuming problems with water quality and quantity have been or are being addressed; we suggest that restoring connectivity of habitats is the next factor to consider. Reconnection of isolated habitats relies on existing habitat and often produces both rapid and long-term results. This includes both restoring longitudinal connectivity such as removal or providing passage at manmade barriers as well lateral connectivity such as reconnecting isolated floodplain habitats and floodplains. The techniques discussed in the section on floodplain connectivity and the culvert/fish passage discussion in the roads section have all shown rapid and promising results, assuming water quality and instream flow issues have been addressed.

Setting up a watershed for long-term recovery will require the restoration of natural processes such as delivery of sediment, organic material, nutrients, and other processes. Thus the next logical step and the factors that most often limit the success of structural manipulations of instream habitat are the restoration of basic hydrologic, geologic, and riparian processes.

While considerable effort and focus is placed on structural manipulations of habitat such as placement of boulders and wood, and many have shown success at increasing local fish numbers, their success is often determined by the previously discussed factors (water quality and quantity, habitat connectivity, and watershed processes). Thus we recommend this category of techniques be implemented following or in conjunction with improvement of those factors.

The methods discussed above and in Figure 13 represent a starting point for restoring a watershed, given fairly broad goals. In many cases, rehabilitation actions may be focused on one or two species. In watersheds where information exists on factors limiting specific species, a more detailed prioritization of actions is possible. Roni et al. (2002) provided a general strategy for prioritizing actions for Pacific salmon (Figure 14). Beechie and Bolton (1999) using a relatively coarse level of resolution provided an example of how priorities for habitat rehabilitation might differ between two salmon species with different life histories and habitat requirements (Figure 15).

It is critical to bear in mind that the prioritization of actions such as outlined in Figure 15 does not alter the types of actions that are needed to restore ecosystems. Ecosystems restoration will include a wide range of recovery actions affecting the entire life cycles of multiple species. Where a single species is a primary concern, altering the sequence of those actions for rapid recovery of that species might be prudent. The limitation of a single species approach is demonstrated in western North America, where efforts initially focused on rehabilitating habitat for one species of Pacific salmon, but quickly had to change focus when additional species became threatened.

Alternative strategies for prioritizing rehabilitation that incorporate economic, ecologic, and biologic factors have been proposed or used, especially where there are multiple species of concern and prioritizing actions based on the needs of individual species will lead to conflicting priorities. For example, Sedell et al. (1990), Wasserman et al. (1995), Beechie et al. (1996), Frissell (1993), Frissell and Bayles (1996), and others have outlined rehabilitation strategies that focus on providing refugia and protecting high quality habitats. Beechie et al. (1996) outlined a prioritization strategy that focused on providing refugia for a depressed steelhead trout population in Deer Creek, Washington (USA). Other strategies might prioritize actions on potential increase in fish numbers, cost, cost per fish, aquatic diversity, assuring for metapopulations structure or diversity, or scoring based on a suite of these and other factors (e.g. Beechie et al., 1996, Frissell and Bayles, 1996; Doyle, 1997). The "Leitbild" (target vision) concept developed in Austria, which is similar to process based restoration, but includes setting numerous goals and priorities with public input, is widely accepted and consistent with European Water Framework Directive (Muhar et al., 2000; Jungwirth et al., 2002). These various strategies incorporate management goals beyond simply restoring watershed or ecosystem processes and habitat. Where at least one species appears to be at high risk of extinction, the refugia approach may be most appropriate to make sure that individual populations are preserved first. By contrast, watersheds with relatively stable populations might embark on a longer term, process-based approach to ecosystem recovery previously outlined (Figure 13 and 14).

The sequencing of rehabilitation actions under different prioritization strategies will vary. We demonstrate how priorities might differ based on different restoration prioritization schemes by running alternative scenarios. First, we developed a hypothetical list of potential rehabilitation actions along with detailed information on their cost, length and area restored, whether they provide refugia for endangered species, potential increase in fish numbers, and cost per fish (Table 19). Second, we ranked rehabilitation actions using different prioritization schemes mentioned above (Table 20). This analysis demonstrated that if rehabilitation actions were prioritized based on Figure 13 or 14, impassible culverts and reconnection of habitats would occur first, followed by road, riparian, and LWD placement. If actions were prioritized by whether they were refugia for an endangered species, instream flow and LWD placement would be first, simply because they are in a high priority area. Similarly, different cost, cost/fish, and total fish production all produced slightly different prioritization scenarios. This simple example illustrates how priorities might differ based on the method, information used, and management objectives.

4.3 Conclusions for prioritization

The appropriate method for prioritizing rehabilitation activities within a watershed will depend on numerous factors. However, our review of rehabilitation techniques suggests that an initial approach such as Figure 13 or the Lietbilt approach described by Jungwirth et al. (2002) may help lead to more successful rehabilitation projects. This approach can then be modified as more detailed information watershed conditions, potential projects and other factors become available.

TABLE 19
Hypothetical example of list of potential rehabilitation actions within a watershed. Refugia are based on Beechie et al. (1996) and include 1 = refugia (areas where recovery is relatively predictable), 2 = key habitat areas that provide for the largest long-term recovery of species of interest, but are sensitive to disturbance and more difficult to restore, and 3 = habitat areas expected to provide the smallest gain for species of interest. Coho and Chinook smolts represent expected annual increase in smolt production. All numbers are fictitious but reasonable and for demonstration purposes only. Modified from Beechie et al. (2003b).

TABLE 20
Example of different order of priorities based on use of different prioritization schemes using information presented in Table 19. Roni et al., (2002) and refugia methods of prioritization do not distinguish between projects of the same type. Hence, there are only four levels and three levels, for the two methods, respectively. Modified from Beechie et al. (2003b).