This is essential work that is fundamental to stock management and monitoring. The ISRP comments are intended to improve on this already promising approach. In summary: • Consider alternative smolt and adult capture and sampling techniques following a thorough review of the history • Use the recruitment results that will be generated to evaluate previous habitat work and also test some hypotheses related to the biology—several options are suggested. • Eventually include AHA/EDT and recruitment simulation model testing and ground-truthing and begin planning for this early in the project. 1. Technical Justification, Program Significance and Consistency, and Project Relationships The proponents have done a good job at establishing clear justification for this project. The linkage with RPAs was well documented. The project will significantly improve the basic understanding of this easternmost run of winter steelhead, now on the endangered species list. ODFW is also conducting many other studies in the vicinity, and knowledge sharing will be taking place. The proposed data acquisition is essential for comprehensive population monitoring and management of this unique winter-run steelhead population, and, eventually, for evaluation of extensive watershed improvements already completed and planned. The program is entirely consistent with a well-designed population and watershed monitoring program, and will serve as an intensively monitored watershed and index site for Oregon and the Columbia Basin. The results will serve several watershed and Basin projects, as listed, and assist in stock status determinations over the broad area, as well as ground truthing of habitat and population dynamics models, in the future. Some suggestions, recommendations, and considerations as the project moves forward are included below. 2. Project History and Results This is a new project, but earlier work has been conducted in the area (including 15 years of habitat restoration work). It will be very useful to compare results of adult returns over the weir with redd counts and assess how accurately past (pre-project) redd counts may have actually indexed adult abundance. A summary of current knowledge of stock status and abundance (smolts and adults) would have benefited reviewers and planners, and should prove useful to future plans and comparisons. 3. Objectives, Work Elements, and Methods There were two basic objectives: (1) determine life cycle survival and performance metric (smolt abundance, smolts/redd, SAR), and (2) determine spawner abundance and spawner distribution (now have 60 miles of known spawning habitat). The work elements and methods seemed fairly complete, and protocols were adequately outlined. It was clear to the reviewers what major activities will be undertaken. In gaining the most benefit of this information for management purposes, it would be useful to specifically consider the following issues: 1. In most cases, it will be adequate to estimate screw trap efficiencies once per week. However, during times of high flows (depending on general watershed runoff characteristics and interannual flow variations), it may not be adequate. Peak smolt emigrations often coincide with high water periods, and inappropriate estimates of trap efficiency applied to catches at those critical times may generate much greater overall errors in some abundance estimates than at other times. It is safe to say that poor efficiency estimates during two days at the peak of the outmigration may be worse than two weeks during a slow time. For that reason, it would be worth developing a general protocol for estimating trap efficiencies that is at a minimum once per week but that is at shorter intervals for high flow/peak migration times. This should not be too difficult to do. It may require a bit more personnel at the peak times. 2. Although the proponents did not discuss the value of any hypothesis testing and the smolt run, they may have thought about the desirability of evaluating some specific aspects of the run itself and effects on SARs. For example, they might hypothesize that larger or older smolts might return at higher rates than smaller smolts or that smolts in the middle portion of the run might perform better than those near the beginning and end. It would necessitate measuring and weighing PIT tagged smolts. Smolt condition might also play a role, not just Fulton's condition index but fat reserves. It might be useful to evaluate if the non-invasive "fat meter" (from work in Ireland) now available for use can be sensitive enough to detect the fat reserves of smolts, which might be linked to their survival. 3. Similarly, it is possible to develop several testable hypotheses about factors affecting the timing of the smolt emigration, including discharge, turbidity, lunar cycle, water temperature, and such. USGS gauging stations are in the basin and this data (water discharge, turbidity, etc.) could be used to relate to smolt emigrations, adult counts at the weir, and redd counts. Several other suggestions or alternatives should be considered. Researchers have considered nearly all aspects of parr, smolt, and adult steelhead monitoring but hopefully have also considered alternative techniques for several aspects, and should be encouraged to review/visit sites with similar experiences (e.g., Keogh River, B.C., Snow Ck. WA). Adult monitoring by video alone will not permit sufficient biological sampling (e.g., scales, weight, condition, fecundity, etc. would not be available). An alternative may be partial live capture (traps) supplemented by resistivity or video counter(s). Radio tagging may be preferred to PIT tagging of adults to determine watershed distribution, proportionate composition within tributaries (given the overall estimate or count), and migratory behavior, which would also elucidate more on discreet habitat utilization and importance. Note that marked adults may be recaptured as kelts in their downstream migration, further confirming population estimates and/or counts. Likewise, full smolt enumeration and sampling (~15%, checked regularly) is preferably to RSTs. The latter provide unreliable M-R estimates with broad confidence levels. Steelhead trout, like Atlantic salmon smolts, are known to exhibit trap avoidance behavior as well as residualism, confounding M-R estimates. If using RSTs, separate mark and recapture sites are recommended, with minimal handling at the marking site, biological sampling at the recovery site. Sampling for age structure must be stratified by age and migration timing. There was little or no discussion of steelhead recruitment analyses, but the reference list indicated some awareness, although some key references on winter steelhead recruitment were missing. EDT and AHA models (which utilize EDT) may eventually be testable at this site as data accumulates. Preparations for these and related analyses (e.g., recruitment models) should begin as operation and maintenance of field work and trapping evolves. 4. M&E The M & E component was paramount and well defined; however, an attempt to evaluate habitat improvement activities conducted over the last 15 years should be considered (e.g., EDT model predictions and smolt yield, or smolts per spawner), and in relation to locations of redds, or habitat parameters.