The ISRP requested the following information:

Additional information is needed on the EcoAnalysts study of reservoir productivity. Given current stocking and net pen escape levels, what is the evidence that the fishery goals can be met in this reservoir given that the limnological study revealed “normal” invertebrate abundance? What additional research is needed to show that harvest targets are realistic? An ecosystem model approach such as EcoSim (www.ecopath.org/) would possibly assist management and evaluations. See the programmatic comments on fish stocking.

The response to this question addressed the history of trout escapes and annual variation in flows, but it did not provide the required information. Specifically, the ISRP wants evidence that triploid trout could grow to large size based on the food resources in the reservoir. Unless trout are attaining very large size simply by remaining close to commercial net pens and eating food that passes through the nets, the ISRP has not yet been presented evidence that supports the hypothesis that significant numbers of >5kg fish can be produced in Rufus Woods. Unless the trophic pathways leading to trophy trout are reasonably well understood, the assumption that releasing large numbers of smaller fish will ultimately lead to harvestable trophy trout remains untested.

More information is needed on trout survival. What are the alternative working hypotheses that could explain the apparently high loss rate of stocked fish, and how will these be tested?

The response set forth some useful hypotheses; however, it would have been helpful to provide more information on how each of the hypotheses would (or could) be tested. Information was given on what is known about the disappearance of newly-released trout, but it was difficult to understand from the response how the hypotheses regarding fish loss would be approached experimentally.Understanding mortality is important because it will bear on the size, timing, and numbers of deliberate releases. How the target release of 100,000 trout of 500g size was derived was not clear.

Additional information is needed about how food habits will be investigated, including the frequency and location of sampling, the size of trout to be examined, and the analytical methods.

The ISRP is very interested in the stable isotope graph provided in the response.Based on the clustering of trout of Lake Roosevelt origin (group A) and Rufus Woods origin (group B), it appears that Rufus Woods-origin trout still maintain an isotopic signature that suggests a continued reliance on food pellets from the net pens. If this is the case, continued leakage of food from the net pens may be needed to support large-bodied trout, unless there is another food source, for example entrained kokanee from Lake Roosevelt, which is underappreciated.

What will be the procedure for developing the long-term management plan? How will success or failure thresholds be established, and what are the contingency plans if some assumptions do not hold?

The ISRP strongly recommends that the plan include objective criteria for testing hypotheses and adjusting release strategies according to new information. It may be instructive to gradually ramp up hatchery production, beginning with carefully monitored releases of smaller numbers of fish, as well as different sizes and release times, and working up to program goals. The long-term management plan should be reviewed by the ISRP when completed and should include the steps for a structured decision making process.

During the site visit, mats of blue-green algae appeared below the fish pens while a continuous flow of food pellets flowed down pipes and into food hoppers. Does water quality influence high rates of mortality of released triploid trout? Empty stomachs from surveys may also suggest depletion of the reservoir’s food supply or simply poor adaptation to the reservoir environment by these fish. How will food depletion or poor adaptation be assessed?

The ecological footprint of net-pen operations could be very significant, and the response to our concerns on this issue was not adequate. A management plan that considers the ecological footprint of fish culture operations needs to derive from the EcoAnalysts report and additional investigations, including further eco-simulations and review.

ISRP Retrospective Evaluation of Results

The ISRP is concerned that this project is headed toward full implementation (releasing large numbers of ~1 lb trout) without sufficient testing of hypotheses regarding growth, survival, and angler harvest. As with some other projects involving fish stocking for harvest, it would appear that pilot-scale testing with limited releases of smaller fish to supplement escaped fish from commercial trout farming operations would facilitate evaluation of the assumptions behind the proposal, and in the long term could save money by reducing management strategies, such as sizes and times of release, that prove to be ineffective.

===========QUALIFICATIONS FOLLOW================

The ISRP appreciates the response of the Colville Tribe to our questions, and we acknowledge that development and management of a trophy trout fishery in Lake Rufus Woods will benefit the Tribe and recreational fishers. Given the long history of unintentional escapes of triploid rainbow trout from the commercial trout farms operating net pens in the reservoir, it is possible that planned releases of smaller hatchery trout will not do much additional harm to species of concern such as native resident rainbow ("redband") trout or kokanee. However, we remain concerned about the ecological footprint of continued fish culture operations on the reservoir, and project sponsors should take all reasonable steps to minimize unwanted effects on the reservoir's ecosystem.

We also believe that important questions remain unanswered and should be addressed at a pilot scale before proceeding to a full-scale hatchery rearing and release program for Lake Rufus Woods. These questions can largely be answered in a 1-2 year time frame, and having the answers could potentially be cost-effective over the longer term by testing key assumptions about the success of the project before attempting full-scale implementation. We believe development of the management plan should assume top priority and therefore recommend that it be completed as soon as possible. The plan should identify key assumptions regarding trout growth, survival, and harvest rates, followed by procedures for testing and monitoring each assumption, and finally by an adaptive management plan that lays out standards for project success or failure before target trout releases are undertaken.

Therefore, the ISRP finds the project meets scientific criteria with some qualifications. The following project elements are scientifically justified:

1. Completion of the management plan and initiation of a pilot-scale release of smaller (500-1,000g) trout with the objective of evaluating the assumptions about survival and growth of the smaller fish at different release timing.
2. Investigations of the food web in Lake Rufus Woods, including food habits of triploid rainbow trout.
3. Investigations of the survival of released fish, including entrainment rates at Chief Joseph Dam.
4. Studies of the growth of smaller fish, to determine if they attain the target size that will make them attractive to anglers seeking trophy trout.
5. Continued creel sampling, including stomach analysis of caught fish and angler satisfaction with the fishery.

The ISRP suggests that moving directly to target releases of 2,000 fish per release is premature until some basic questions have been addressed, and we recommend that pilot-scale release experiments be conducted first. Our qualifications thus include:

Continuation of the limnological investigations of Lake Rufus Woods, coupled with appropriate physiological models of growth and survival (e.g., EASy and EcoSim) to identify testable hypotheses. The EcoAnalysts report and its conclusions should be reviewed by the ISRP.

Learn more about early reservoir mortality or other sources of loss such as downstream entrainment, particularly in the smaller sizes of trout that are being considered for the project, before proceeding to full-scale production.

Continuation of the creel sampling program, which is based on a sound design, as well as stomach analysis of trout to learn more about what they are eating as they grow larger. The sponsors state "In 2012, stomachs from gill net and angler captured fish will be used. Each month a single 3m x 30m gill net with 4 mesh sizes will be set in three randomly selected locations in Rufus Woods. The nets will be set in the afternoon and hauled the following morning." Digestion of stomach contents, especially if fish are eating food pellets, could occur overnight and the method should be reconsidered or tested to make sure results are not biased.

Production of a report that summarizes overall evidence that releasing relatively large, stable numbers of ~1 pound trout will result in a fishery that yields trophy-size fish, at the anticipated fishing intensity. This report should be completed in 2014 and reviewed by the ISRP and should include a structured decision making process that sets forth standards for measuring project success or failure.

An analysis of the ecological impact of the hatchery rearing facility for producing the trout on the ecology of Lake Rufus Woods should be completed, and adjustments made to the fish culture facility if reducing unwanted impacts is possible.

Although the project is titled Rufus Woods Habitat/Passage Improvement, Creel and Triploid Supplementation, the proposal pertains only to the triploid stocking and sampling effort. The project sponsors have adopted a scientifically sound creel sampling strategy, but the ISRP needs additional information in order to determine the project's overall soundness.

1. Additional information is needed on the EcoAnalysts study of reservoir productivity. Given current stocking and net pen escape levels, what is the evidence that the fishery goals can be met in this reservoir given that the limnological study revealed “normal” invertebrate abundance? What additional research is needed to show that harvest targets are realistic? An ecosystem model approach such as EcoSim (www.ecopath.org/) would possibly assist management and evaluations. See the programmatic comments on fish stocking.
   
   
2. More information is needed on trout survival. What are the alternative working hypotheses that could explain the apparently high loss rate of stocked fish, and how will these be tested?
   
   
3. Additional information is needed about how food habits will be investigated, including the frequency and location of sampling, the size of trout to be examined, and the analytical methods.
   
   
4. What will be the procedure for developing the long-term management plan? How will success or failure thresholds be established, and what are the contingency plans if some assumptions do not hold?
   
   
5. During the site visit, mats of blue-green algae appeared below the fish pens while a continuous flow of food pellets flowed down pipes and into food hoppers. Does water quality influence high rates of mortality of released triploid trout? Empty stomachs from surveys may also suggest depletion of the reservoir’s food supply or simply poor adaptation to the reservoir environment by these fish. How will food depletion or poor adaptation be assessed?

1. Purpose: Significance to Regional Programs, Technical Background, and Objectives

This project had its genesis in the mid-1990s when triploid rainbow trout escaped from a commercial farming operation in Lake Rufus Woods, and subsequently turned up at very large size in the recreational trout fishery. Continued escapes of triploid rainbows, as well as some deliberate stocking, have supported a sizeable fishery that the Colville Tribe wishes to sustain to provide tribal income. The goal is to sustain an angler harvest rate of 0.5 fish per hour with an average fish size of 1.5 kg, with some fish reaching 10 kg or more, at a fishing intensity of 50,000 angler days per year. Based on evidence from a reservoir productivity review, the project sponsors feel these goals are feasible, and trophy trout can be produced in Lake Rufus Woods without the need to raise very large, expensive fish for release.

The project has been justified based on mitigation for lost anadromous salmonid production caused by Chief Joseph Dam. The technical background was explained in general terms, but without going into many details about previous findings related to rainbow trout food habits, losses to entrainment or predators, and changes in the productive capacity of the reservoir. Although the objectives were explained, they addressed only production goals and did not include conservation objectives for native fishes as reflected in the habitat/passage language in the project title.

2. History: Accomplishments, Results, and Adaptive Management (ISRP Review of Results)

The history of the Lake Rufus Woods trophy trout fishery was given and the graphs of angler days, catches per unit effort, and fish counts were useful. More information is needed concerning the EcoAnalysts report on reservoir productivity. If "densities of benthic invertebrates were within the normal range for many trout fisheries, but tending toward the “low end" and if "More than 25 % of the rainbow trout stomachs examined were empty" (p. 5 of the proposal), and further if "nearly 50% of the released fish disappeared each month" (following paragraph), it is difficult to follow the logic that releases of smaller, less expensive trout would achieve the project's trophy trout and catch rate goals. Perhaps there is more information in the findings to date that were not given in the proposal, but based on the data presented it does not appear that the suggested release strategy is likely to produce the anticipated benefits. What is the evidence to support the belief that smaller fish will achieve the growth rates needed to reach the target sizes at harvest?

Additionally, the importance of entrainment losses and possible predation losses to non-native fishes such as walleye need a more complete description of what has been learned to date. Because smaller trout are likely to be more vulnerable to predation and to downstream displacement through the dam, an earlier, smaller release strategy might be counterproductive, especially if smaller fish are released closer to Chief Joseph Dam.

The section of the proposal dealing with adaptive management also needs clarification. It is mentioned elsewhere in the narrative that stocking schedules might be modified based on incidences of escape from the net pens, but it was not clear if numerical criteria for such modifications have been developed or if these will be part of the long-term management plan. How is the carrying capacity of the reservoir figured so that overstocking can be prevented? How does the movement of fish from Lake Roosevelt into Lake Rufus Woods affect trout behavior or growth rates?

ISRP Retrospective Evaluation of Results

The proposal has clearly-stated goals but has not incorporated testable hypotheses into changes in management strategies. For example, it is apparently assumed that prolonged rearing of triploid trout to large size and high cost can be largely replaced with a stocking strategy involving smaller, less expensive fish that will result in abundant large trout for the recreational harvest. However, the ISRP is not convinced that the evidence is strong enough at this point to support such an assumption; rather, it should be treated as a testable hypothesis, and a management experiment should be put in place to evaluate it. Assumptions about how many trout the reservoir can support, and at what size distribution, should also be treated as testable hypotheses before full-scale implementation of intentional releases of triploid trout.

3. Project Relationships, Emerging Limiting Factors, and Tailored Questions for Type of Work (hatchery, RME, tagging)

The sponsors have adequately described the work's relationship with other salmonid enhancement projects in the Colville Reservation area. However, more information is needed on how this project affects conservation efforts for native interior rainbow trout (redbands). Will the program result in incidental catch of adfluvial lacustrine redband trout and how will this be monitored, if it occurs? The proposal notes that a risk assessment for native species has not been completed (p. 11), but it would seem prudent to do so in view of the considerable effort to protect the native redband trout in Lake Roosevelt. Walleye are listed as a secondary focal species, but walleye abundance, predation by walleye on young rainbow trout, or possible changes in walleye regulations are not discussed. This information should be included in the proposal.

It is mentioned that some early research showed that redband trout did poorly in a hatchery environment. It does not appear that there will be any hatchery supplementation of redband trout, but if any releases occur how will they affect the triploid rainbow fishery program in Lake Rufus Woods, or vice-versa?

The question of how the released triploid trout are going to obtain enough food to achieve the growth rates need to reach the 1.5 kg target rate for the fishery remains to be answered. Given the other information in the proposal about average benthic invertebrate densities in Rufus Woods, the relative scarcity of fish or other large-bodied prey in rainbow trout diets, and the general absence of evidence that other Columbia River reservoirs can sustainably support a trophy trout fishery similar to the one envisioned in the management plan, it would seem that understanding the food web of Rufus Woods is a key to project success. The ISRP strongly recommends that more effort be devoted to food web research in this ecosystem. It seems possible that post-release/escape growth rates of triploid trout might be fueled by residual food from the net pen operations.

4. Deliverables, Work Elements, Metrics, and Methods

The deliverables were clearly stated but more information was needed on some of the metrics and methods. The creel census was adequately described and is based on sound scientific sampling principles. Additional details are needed on stomach sample analysis including sample size; timing and location of sampling; stomach content determinations and statistical methods; and on the procedures to be used in developing the long-term management plan including specific fishery objectives, interim targets, tracking metrics, and adaptive management guidelines.

The proposal would be improved by an explanation of the rationale for the stomach sample and cobble basket study. Presumably this is an attempt to determine if food supplies are limiting growth and survival of the trout. The sponsors are referred to the ISAB food web report for information on how bioenergetics models could help.

4a. Specific comments on protocols and methods described in MonitoringMethods.org

The creel survey developed by Dr. Skalsky was well explained. No details specific to the Rufus Woods triploid rainbow trout release protocols or to the stomach sampling program were uploaded to MonitoringMethods.org.

Modified by Dal Marsters on 4/17/2012 1:08:22 PM.