As requested by the Council on October 22, 2004, the ISRP reviewed Washington Department of Fish and Wildlife's (WDFW) step submittal for the project, Re-introduction of Lower Columbia River Chum Salmon into Duncan Creek.
Duncan Creek chum were extirpated from Duncan Creek in the late 1960s following construction of a dam and culvert complex at the creek's confluence with the Columbia main channel. Moreover, natural maintenance of the creek's side channels and seeps appears to have been stressed by upland land use activities. Most of the good spawning habitat in Duncan Creek is located just above Duncan Lake. This area is most important for chum and coho, although it is also used by fall Chinook and winter steelhead. The watershed was blocked to anadromous fish but has been re-opened by both physical facilities and agreement with landowners. Specifically, access to spawning areas in Duncan Creek has recently been improved by 1) the construction of a dam that lowers lake levels during salmonid migration periods, 2) installation of a fish ladder to provide adult passage at the mouth of Duncan Creek, and 3) creation of artificial spawning side-channels.
Now that access and rehabilitated habitat have been provided, WDFW plans on capturing up to 10% of the adult chum salmon from nearby population segments in the Lower Columbia Gorge to reintroduce adults and artificially produce juveniles (as well as provide access to strays from other sources) with the goal of establishing a self-sustaining in situ breeding population in Duncan Creek. Supplementation is expected to be sustained for three generations and the artificial side channels plan to be used until more "normative" conditions can be re-established.
WDFW also intends to evaluate adult and juvenile supplementation strategies via the project. Fish would be reared at an existing facility and no new facilities would need to be built. Consequently, this is a Combined Step Review and all relevant review elements and conditions are to be met in one submittal. Those conditions centered on the ISRP concerns from the Lower Columbia and Estuary province review that chum salmon should not be stocked until a plan for establishment of a wild chum population in the context of a watershed assessment occurs and monitoring protocols are defined.
The proposal was submitted as part of the Lower Columbia provincial review (2002) where CBFWA rated it as High Priority. The ISRP provided a fundable recommendation for the benefits to chum, coho and sea-run cutthroat, but raised the concerns described above about re-establishing chum that are addressed by this WDFW submittal.
NOAA Fisheries commented that Duncan Creek was an important project to move chum salmon spawning from the mainstem to the tributaries and believed it addressed RPAs 156 and 157 from the 2000 BiOp, which called for strengthening the threatened Lower Columbia chum ESU, especially tributary sources of production in the Lower Columbia Gorge area. Bonneville supported funding the project and agreed with NOAA Fisheries that the restoration efforts would implement RPAs 156 and 157.
In the October 22, 2004 cover letter from Council requesting further ISRP review, Council Staff noted that the Duncan Creek project potentially has substantial benefits for chum salmon restoration and would recommend funding the project. Staff also noted that the taking of captive broodstock and the proposed artificial production actions would trigger a Step Review under the Fish and Wildlife Program. That Step Review would then allow the Council to focus on addressing the ISRP concerns of conducting a watershed assessment and defining monitoring protocols prior to the stocking of a chum population. Therefore, funds associated with artificial production contained in the Construction and Implementation (objectives 3, 4 and 5) and Monitoring and Evaluation (objectives 3, 4, 5, 6, 7, and 8) budgets for the project are conditioned upon a favorable Step Review.
The ISRP response below represents the ISRP's review of the Combined Step Review and all relevant review elements and conditions
ISRP Combined Step Review
The ISRP recommends approval and funding for the Duncan Creek Combined Three-Step project. The package of proposal and supporting documents (M&E Plan, HGMP) provide a thorough and persuasive argument for the proposed set of actions, although the proposal itself could have been much better organized and should have provided a clearer background on the project and better documentation of actions that proceed from the proposed project. The ISRP comments below raise points that project sponsors, Council, and Council staff may wish to consider in bringing the project to its greatest potential for contributing to chum salmon recovery in Duncan Creek and the Lower Columbia, as well as helping provide insights into critical uncertainties associated with supplementation and habitat restoration activities.
Scientific Principles of the Fish and Wildlife Program
This project greatly benefits from the passage and in situ habitat rehabilitation steps as a precursor. These steps directly addressed the findings regarding system limitations from the EDT analysis and potential for productivity and diversity. These steps directly address Principle 1 of the 8 scientific principles from the 2000 Fish and Wildlife Program's Basinwide Provisions, as called for in the Step Review Elements. Reintroduction of this extirpated population into the watershed is clearly predicated on addressing the original stressors that caused extirpation (which has not routinely or historically been the case with reintroduction attempts - as such this is an important advancement). Moreover, the project sponsors have completed HGMP and M&E plans a priori to give consideration to evolutionary risks and a fundamental understanding of learning from and adapting management actions.
Project Success and Duration
While there are some issues that can be addressed in a revision of this document, the sponsors provide a project that benefits from several conditions that suggest a likelihood of attaining (and sufficiently monitoring and evaluating) the broader goals for the resource and narrow objectives for the project in particular. Ultimately, this project has promise because there is not a legacy of long-term artificial production effects (stock transfers, domestication, swamping, inbreeding, etc.).
While the project might conceivably fail to restore a functional and self-sustaining population (if for example blocked passage and degraded spawning habitat were not responsible for extirpation), four alternative scenarios regarding the ultimate source of population founding would lead to a project success:
1) recolonization through volitional strays;
2) recolonization from release of naturally-produced adults into the spawning channels;
3) recolonization from released of artificially produced juveniles;
4) recolonization from a combination of these sources.
While we might expect that all will be derived from common nearby sources, timeframes for recolonization, influences of breeding effects, and founder effects will differ across the scenarios. It is an important component of this project that three potential sources for recolonization can be discriminated through the presence or absence of various thermal and strontium marks. On page 9, "planners recommend that a combination of natural and hatchery production would be the most likely way to produce the most rapid sustainable improvement in chum salmon runs." Are the sponsors convinced that the combination approach is the most effective, OR is are they uncertain which of the three strategies (adults v. juveniles v. strays) will show the strongest gains (and are therefore designing the test)?
The project also relies on brood from a local source (within the power of current genetic analyses to discriminate different stocks) that are presently above critically endangered levels and will take steps to limit acknowledged risks of brood mining (stated as less than 10% of the area's population, but without describing how this will be measured and therefore practiced). The ISRP therefore recommends that WDFW better describe the methods for preventing such brood mining.
The project indicates that the side channels are "an interim measure to sustain chum salmon in Duncan Creek until more normative ecosystem function ... restored" (p.4). Some explanation of the timeframe for this expectation relative to the life expectancy of the artificial side channels is appropriate. Will normative conditions return in 10, 50, 500 years? The side channel was created by blockage of the original main channel, which caused the establishment of a new main channel. Re-establishment of normative conditions might lead to destruction of the existing side channel within the 12 years proposed as the lifespan of the project. If such an event occurred, what is the likelihood that other appropriate spawning areas would be coincidentally created? Is it possible that natural events could eliminate appropriate spawning habitat within lower Duncan Creek?
The project description indicates that three generations (12 years) of artificial production and release may be sufficient to produce a robust, stable, and self-sufficient population. Numerical targets are provided, but potential annual variation associated with external variables should be provided. Moreover, the three-generation expectation is presumably based on some projection or predictive model (including robust assumptions). Articulating these will help reviewers evaluate appropriateness. Related to this, the project indicates that the supplementation will cease if sustainability is achieved before three generations. How "sustainability" is defined is important so that future decision makers will know if they have gotten there? Lastly, if no real progress is realized in two or three generations, is there a decision point at which the program will be terminated?
The project description (page 18) notes that the project duration is set for 12 years, or three chum salmon generations, unless a self-sustaining population is established earlier (emphasis added). Given the natural annual variation observed in most salmon populations over periods much longer than 12 years, it seems extremely unlikely that a population could rebuild and be judged self-sustaining in less than 12 years (or three generations). Also given the life history of chum salmon, there may be less genetic and demographic exchange between brood years than for chinook or steelhead, something that could lead to strong and weak year classes. This would also affect any judgements about the "self-sustaining" nature of the Duncan Creek chum salmon population.
Regarding "jump starting" the population, ISRP is concerned whether the perspective that "supplementation methods to speed up recolonization" (p8) is justified. It is not clear that restoration is explicitly time-sensitive (at critical risk of demographic collapse). Indeed, the concept of "jump-starting" populations is not generally well-received in the scientific community — it remains in most instances, an untested hypothesis! Providing some justification will be important.
Monitoring and Evaluation Protocols:
The monitoring plan and protocols are excellent. The hatchery HGMP and subbasin plan are also included for context. The collection of documents seems persuasive as a study plan and watershed assessment. The monitoring and evaluation study plan is an important regional test to detect any differences in performance between wild-spawned and hatchery-planted fish. Usually, we have adequately marked hatchery fish available but no way to mark the wild-spawned juveniles. In this instance, proposers intend to use both thermal and chemical marking of juvenile fish. The hatchery fish would be marked thermally (using creation of "bar codes" on otoliths by differential growth rates at different temperatures). The well-defined area for spawning makes marking the wild-produced fish feasible (they would be netted, marked by strontium dip, and released before the hatchery fish were added). The strontium mark is perfect for this use and should be no trouble to detect above a background, even with some marine-derived elements in the system from salmon carcasses.
Related points for Monitoring and Evaluation
The monitoring and evaluation program is thorough and likely to provide the information needed to evaluate the proposed re-establishment program and to compare the various supplementation strategies. However, the monitoring and evaluation program could be expanded into two areas that would contribute significantly to issues that are among the region's most critical uncertainties: 1) measuring habitat restoration actions and fish responses; and
2) evaluating the fitness effects of various production strategies. Neither of these goals is explicitly described in the proposed project; however, including them would add substantially to what can likely be learned from this project.
Much is made in the project proposal of the habitat restoration activities in Duncan Creek and the goal of reestablishing normative conditions implied as linked to project success and to the establishment of a self-sustaining chum population in Duncan Creek. The M&E plan should explicitly define goals for the habitat conditions needed to sustain chum salmon in Duncan Creek and describe a monitoring program to identify and evaluate whether those conditions have been achieved.
Expanding the M&E program to include a fitness evaluation of the various production strategies would significantly increase what might be learned from this project. Genetic theory strongly suggests that artificial production activities can introduce risk and reduce population fitness for salmonids (see extended discussion in the ISAB's 2003 review of supplementation). The broodstock collection methods and the release of fry (rather than extended juvenile rearing) should help reduce the risks and effects of the artificial production intervention. Proposers are clearly aware of these issues (described in the HGMP and the M&E plan), but should strongly consider expanding the M&E protocols to directly measure fitness effects of the three colonization (supplementation) strategies.
A completed HGMP is provided that addresses evolutionary risks associated with components of the breeding, culture, and release of chum for this project. The HGMP appears well conceived and straight-forward with one potential exception that would benefit from some justification or reference regarding the matrix or factorial design including the use of "back-up" males. If such information is not readily available, this might be a needed area for research. Also of specific interest: are the Duncan Creek chum contributing to the viability of the Lower Columbia River Gorge ESU -- in any way detracting or interfering? Which strategies of release or reliance on straying are most effective? Do these strategies interfere with each other? What percentage of the population at any given time is artificial v. natural production? Is this changing?
While difficult to predict and describe the full range of specific conditions (including duration) necessary for the salvage collection, it will be a key consideration not to undo or counteract the progressive elements of the Duncan Creek restoration project by introducing/transporting large or unsupportable numbers of adults into the system on top of each other. We recommend some description of limits.
 ISAB 2003-3: Review of Salmon and Steelhead Supplementation.