1. Objectives

The Ocean Survival of Salmonids Project is a long-term (20-year) effort. The project monitors salmon and oceanographic conditions in the Columbia River plume and nearshore marine areas. The primary research hypothesis is that “variation in salmon survival during the first few months of ocean residency has the largest impact on cohort strength of all life stages for Pacific salmon.” The primary project objective is “to determine the physical, biological and ecological mechanisms that control survival of salmon during their early marine life.” Hypotheses were grouped into three areas: (a) growth and survival; (b) hatchery-wild interactions and density dependence; and (c) modeling, scenario planning, and recovery.

The primary hypothesis is that marine growth and survival are determined by the combined effects of physical and biological processes in the coastal ocean; the abundance, species composition, and spatial distribution of predators; and the size and availability of salmon and alternative prey.

Because hatchery-wild interactions and density dependence cannot be observed directly, three hypotheses designed to infer the effects of hatchery/wild competition were addressed: (a) there is no difference between the diet and growth of hatchery and wild fish, (b) there is no difference between the distribution and migration timing of hatchery and wild fish, and (c) there is no difference between the synchrony of responses to marine drivers for hatchery and wild fish.

Modeling, scenario planning, and recovery hypotheses are based on assumptions that early ocean growth and survival of juvenile salmon and steelhead are determined through both bottom-up and top-down processes associated with the productivity of coastal ocean and plume habitats during spring and summer. It is also assumed that survival during this period is highly variable and has an exceptionally large impact on cohort strength and, therefore, future adult returns.

While these hypotheses are highly relevant to the Council’s Fish and Wildlife Program, they are difficult to quantify, achieve, and test because of the complexity and unpredictability of the marine ecosystem. A long time-series is needed to tease out the relative impact of many factors.

The project is now in its 20th year, and there is no indication of a specific future end date. The ISRP concludes that this project will continue to provide critical information into the future and that the value of this project to the Fish and Wildlife Program will grow each year. Nevertheless, the program should not remain static. The proponents indicate that they are making efforts to place their results into an ecosystem framework and are modifying the sampling strategy (e.g., via a smaller mesh liner) to provide more information on salmon and their prey simultaneously. These future activities should be reviewed to ensure that the program continues to collect data relevant to management questions, to examine changes in survey methods, and evaluate if the new types of data collection will fill important gaps in understanding of ocean survival (e.g., impacts of forage fish abundance).

One objective is to estimate early ocean survival, but this is not directly addressed. A major question is whether the program can be modified to estimate survival in the early and subsequent ocean life stages (e.g., by the reintroduction of acoustic survey lines).

2. Methods

At present the project is contracted by BPA on an annual basis, with clear milestones for specific work elements such as research vessel cruises as well as sample and data processing.

Juvenile Salmon and Ocean Ecosystem Surveys (JSOES) relies on estuary purse seines to capture salmonids along with measurements of physical and biological characteristics of the Columbia River plume and nearshore ocean. Two additional surveys compare the attributes of hatchery and natural origin juvenile fish (late May), and they characterize the spatial distribution of juvenile salmonids, the pelagic fish community, the salmon prey field, and characteristics of the ocean environment (June). In general, methods are scientifically sound, but some survey methods could be improved. For example, methods currently used are sufficient for collecting biological samples of salmon and associated species in the catch, and for providing indices of relative abundance (i.e., catch per unit effort) but not for directly estimating salmon survival or the abundance/biomass of salmon predators and prey.

A key assumption is that catch per unit effort (CPUE) is proportional to abundance. It is not clear if this been empirically tested using estimates of smolt outmigration from CSS vs. the CPUE from the trawls.

Varying levels of salmon predators add variability to survival proportions. This makes it harder to detect the effects of ocean conditions and also means that models based on earlier years, where the predators were not prevalent, are less and less reliable.

The original trawl survey design was developed primarily to determine the early ocean distribution of juvenile Columbia River salmon. Considering the current hypotheses being tested and the low CPUE of juvenile salmon in surface trawl catches, especially wild salmon, new fisheries-oceanographic ecosystem survey designs (e.g., a gridded survey, larger trawl) and methods (e.g., combined trawl and hydroacoustic surveys) might be considered.

3. Results

The project has made significant progress in addressing a progressive series of questions and hypotheses, ones that have evolved over time as more data are gathered.

Project results show that ocean productivity controls feeding and growth of juvenile salmon (i.e., bottom-up control), leading to multiple, nonexclusive hypotheses about how ocean productivity influences salmon survival. For example, when ocean productivity is high (a) rapid growth allows juveniles to escape size-selective predation and (b) abundant alternative prey for salmon predators reduces predation pressure on salmon juveniles. The investigators conclude, “Importantly, these ocean effects can be somewhat dependent on freshwater effects (carryover effects, Gosselin et al. 2018⁴ ). Therefore, accurately evaluating freshwater management actions requires an accurate ocean context.” An economic analysis may be needed to compare the cost and effectiveness of management actions taken in freshwater (e.g., hydro-system actions) vs. actions occurring in the ocean (e.g., harvest management).

Project results addressing hatchery/wild interactions and density dependence CUs (Critical Uncertainties) were hampered by the low numbers of wild juvenile salmon in ocean research vessel catches because of low relative abundance and/or low sampling rates. Results indicate that hatchery salmon may be used as a proxy for wild salmon in some cases (e.g., migration timing, diets, and spatial distribution), but not others (e.g., size and condition factor). However, data are insufficient to determine “whether hatchery salmon have an advantage over wild salmon during different ocean conditions.”

Project results addressing modeling, recovery, and climate impacts CUs show that “stockspecific distribution, abundance, and survival of juvenile Columbia River salmon in the NCC [Northern California Current] vary synchronously with variable ocean conditions.” Investigators have developed a suite of physical, biological, and ecological indicators of ocean conditions that are useful qualitative predictors of salmon survival that are now being used by managers as early-warning indicators of recent ocean conditions or as covariates in statistical models. Initial results of marine survival modeling “confirm the importance of the marine stage in the salmon life cycle and suggest dramatic declines in population trajectories with a warming ocean.”

An important consideration is whether the project's primary objective (“to determine the physical, biological and ecological mechanisms that control survival of salmon during their early marine life") is sufficient to guide policy or actions that benefit fish and wildlife. Future changes in climate and ocean conditions are not predictable. Thus, as discussed by the investigators in their project summary, a greater emphasis on long term, high quality monitoring of ocean conditions and assessment of changes in the ocean ecosystem that affect juvenile salmon survival is essential.

The project has an excellent publication record with over 121 publications listed. Project findings are shared with others within and outside the basin via numerous journal publications, project reports, conference presentations, an excellent web site, presentations to the Council, and via the Ocean Forum. However, the ISRP notes that communication of information within the Basin will likely decline because there are limitations to project funds to attend conferences and to support other communication portals, such as the Columbia Basin Bulletin.

4. 2017 Research Plan uncertainties validation

The project narrative provided a comprehensive table and text that links its research to 39 CUs, organized into four categories: (a) survival and growth; (b) hatchery wild interactions and density dependence; (c) modeling, scenario planning, and recovery, masking effect of the ocean; and (d) other topics. The project directly addresses six CUs and indirectly addresses 33 additional CUs. Linkage to these CUs is generally appropriate, though Item F3 (What factors within and outside of the Columbia River Basin influence trends in recruitment, mortality, and abundance of Columbia River Basin fish and wildlife populations?) should have been identified as a direct link (current CUs on the web failed to make the link between this CU and the Ocean project).

Table 1 in the narrative identifies several new CUs associated with the project, including 19 new CUs. Information in the project narrative is not always sufficient to understand the new linkages, and additional justification may be needed. For example, some linkages were revised from direct to indirect and vice versa. The indirect relationships identified in Table 1 tend to be more inclusive than exclusive. Some of the linkages may be questionable (G1.1, G2.1, C1.5, F3.7, F3.8, B1, M3.1, F1.4). For example, CU 1.5 states “What are the range, magnitude, and rates of change of natural spawning fitness of integrated (supplemented) populations, and how are these related to management rules, including the proportion of hatchery fish permitted on the spawning grounds, the broodstock mining rate, and the proportion of natural origin adults in the hatchery broodstock?” and it is questionable how surveying juvenile fish in the early ocean period helps answer this CU.

Similarly, it is unclear why CU E2.5: "What additional information is needed to assess the importance of tidal freshwater, estuary, plume, and ocean habitats and their biota on focal species (anadromous salmonids, White Sturgeon, Pacific Lamprey, Eulachon)?" was not included in Table 1, given that CU E2.5 was considered by the Council to be directly addressed by this project in the Council's 2017 database.

 

⁴ Gosseline, J.L., R.W. Zabel, J.J. Anderson, J.R. Faulkner, A.M. Baptista, and B.P. Sandford. 2018. Conservation planning for freshwater-marine carryover effects on Chinook salmon survival. Ecology and Evolution 8:319-332

 

Modified by Dal Marsters on 11/8/2018 9:17:34 AM.

Comment:

In 2010, the Bonneville Power Administration addressed the need to understand salmon survival in the ocean, stating "Salmon spend most of their lives in salt water. Most don't come back to the river to spawn. If just 1 to 2 percent more juvenile salmon survived through adulthood in the ocean, the number of adult salmon that spawn would more than double" (see BPA document). Since then, BPA's need to understand ocean survival of salmon has been reaffirmed by dramatic fluctuations in ocean conditions (favorable and unfavorable) that were correlated with adult Chinook salmon and steelhead returns to the Basin. Nevertheless, funding for the Fish and Wildlife Program's research program to understand salmon survival in the ocean has been reduced by about 75%.

In this "change-of-scope" proposal, the NOAA investigators responded to the ISRP's recent (ISRP 2018-8) scientific review by providing an innovative research plan to advance quantitative understanding of the physical, biological, ecological, and ecosystem processes that impact the early ocean survival of Columbia River salmon and steelhead. The proposal focuses on the practical needs to improve forecasting of adult salmon returns and to advance decision-making about management and mitigation options in the face of future (unpredictable) changes in climate and ocean conditions. Cognizant of budgetary constraints, the NOAA investigators provided three alternative funding scenarios for project implementation. The first scenario (Option A) indicates that present funding levels cannot maintain all of the project's current field and laboratory investigations, ones that are cornerstones of this project. For example, the May survey that provides the only empirical data on juvenile steelhead and associated ocean conditions would be eliminated. Considering the current poor ocean survival of Columbia River steelhead, this loss of information is untenable to the ISRP. The second scenario (Option B) continues the May survey but does not allow implementation of the full suite of proposed objectives that advance adaptive management and mitigation practices. Thus, the ISRP recommends full implementation of the proposed project (Option C), which would include testing of hypotheses critical to understanding the top-down mechanisms (predation, predator-prey interactions) that control early ocean survival of juvenile salmonids (see ISRP 2018-8).

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

This proposal is the only remaining Fish and Wildlife Program project, as well as the only project in US coastal waters, that directly addresses the effects of ocean conditions on growth and survival of Columbia River juvenile salmon and steelhead. During the past 21 years, the project has revealed several important relationships among ocean conditions, the Columbia River plume, and the distribution, abundance, and survival of juvenile Columbia River salmonids. The biological/physical objectives of the project are clearly defined. Three alternative scenarios for project implementation and scientific objectives for each option are provided. This "change of scope" proposal includes past objectives that were reviewed by the ISRP in 2018 and provides new objectives (depending on funding level). The new objectives will address the direct causes of early ocean mortality of juvenile salmonids (predation by marine birds and piscivorous fish, and reduction in abundance of forage fishes as a buffer to predation), enable quantification of the current qualitative forecasts of adult salmon returns, and lead to an ecosystem-based model to help decouple the effects of various mitigation efforts in fresh water from the effects of a changing ocean environment.

The significance of this ongoing project to the region and to mitigation and management of Columbia River salmon and steelhead is widely recognized and cannot be overstated by the ISRP. Extending the 21-year dataset and addressing the new proposed objectives are critical to the understanding of factors affecting the growth and survival of Columbia River salmonids and how management actions in the Basin may increase salmonid survival at sea. The project has continued to adapt and change in response to scientific reviews by the ISRP and to management and evaluation needs in the Basin.

The description of the technical background is outstanding and provides a review of relevant past results and anticipated quantitative results, including strong supporting information from the primary literature.

2. Results and Adaptive Management

The ISRP reviewed the results and outcomes of this project in 2018. The ISRP views this "change of scope" proposal as an adaptive response to both the ISRP's review and lessons learned from past results. The project has an outstanding record of publication in the primary scientific literature (~130 publications), participation in scientific and management meetings, presentations to the Council including the Ocean Forum that provides outreach and interaction between scientists and managers in the Basin, and public outreach through electronic and print media. Underscoring the importance of the project, in February 2019, the Seattle Times published a multi-page article that described the ongoing effort by this project to unravel factors affecting salmon survival and abundance. Another recent (March 2019) Seattle Times article discussed the project's June 2018 survey findings, indicating improved ocean survival of juvenile Chinook salmon.

3. Methods: Project Relationships, Work Types, and Deliverables

The proposal, including new objectives, is based on sound scientific principles and methods, and includes provisions for monitoring and evaluation of results. The relationships to projects both inside and outside of the Fish and Wildlife Program are clearly described. The project is well coordinated with similar projects that focus on the marine ecology and survival of salmonid populations from other regions of the USA and Canada, including collaborative data sharing, development of new and improved methods, and scientific publications. Work types and deliverables are clearly described and achievable based on past performance.

The proponents describe objectives, methods, and effort that are specific to three funding scenarios. They also provide a detailed description of how each of four objectives is dependent on each level of funding. Although NOAA Fisheries provides matching funds for this effort, project costs have increased while the overall operating budget has declined significantly since 2012. Additionally, BPA funding for two other ocean research projects (i.e., Canadian Department of Fisheries and Oceans; Pacific Ocean Shelf Tracking project) were eliminated in 2012. The current level of reduced funding for the NOAA Fisheries Program (Option A) would lead to reduced effort and scope (e.g., no May survey, thus missing steelhead and early migrating Chinook). Option B represents partial restoration of the budget. Option C represents full budget restoration that would include testing of hypotheses critical to understanding the top-down mechanisms (predation, predator-prey interactions) that control early ocean survival of juvenile salmonids (see ISRP 2018-8).

Qualification: A synthesis of this project, as proposed by the proponents, should be completed and reviewed by the ISRP in 2011.

This is a productive and worthwhile project that has made significant contributions to understanding relationships between Chinook and coho salmon survival and ocean and plume conditions. The ISRP commends the proposed new research on abundance, distribution, timing and migration of smolts through the estuary. Another important new feature of the project is the proposed analysis of factors affecting sockeye, chum, and steelhead. However, the proposal was not clear on the extent to which data on these species were collected in previous years but not analyzed or reported. Rapid gains in knowledge could be accomplished if previous data on these species were collected, although the ISRP recognizes that these species may not be abundant in coastal research trawl samples. An important outcome of the project has been a qualitative method for forecasting salmon runs that appears to be an improvement over past methods. It is refreshing to see a project that directly addresses management concerns. The ISRP strongly concurs with the proponents that a major synthesis of this work should be completed in 2011.

Some important issues to be considered during the contracting process and in the synthesis are listed below:

1. Strategic plan. The ISRP recommends the use of synthesis results to develop a strategic plan that prioritizes project hypotheses and management objectives. The current approach is exploratory and observational, including numerous hypotheses and investigations of trails of evidence dealing with limiting factors ranging from lipids to parasites to bird predation. When arguing for an observational rather than experimental approach, the proponents state that each year/sampling season can be considered an “independent observation.” It seems unlikely that the quantitative values of physical and biological variables are independent between years, that is, there is no between-year autocorrelation. The proponents need to justify this assertion or adjust for it in their statistical analyses, as described in the synthesis objective. The strategic plan should explain in greater detail how interaction issues arising from studying four elements (bottom-up, top-down, food-web, and plume structure) at the same time will be addressed.

2. Achievable objectives. Consider whether stated objectives are achievable. For example, can the objective (discussed in proposal’s introduction) to determine decadal-scale cycles in ocean productivity be achieved? If so, when will the periodic wave length in cycles be known? If changes are periodic events without a fixed wave length or chaotic events, then how will this objective be achieved?

3. Fishing operation effects. Consider important sources of variation in research trawl and other fishing operations and fishing efficiency with respect to what is known about diel, horizontal, vertical, and seasonal distribution of juvenile salmon that might affect time-series observational data on species composition, abundance, distribution, growth, etc., of juvenile salmon in the survey area.

4. Cruise planning and coordination. The ISRP recommends that the proponents provide annual cruise plans to other related projects. The plans should include sufficient detail on how cruises in the plume, estuary, and ocean will be organized and coordinated with these other projects. For example, the current proposal lacks details on how far upstream the estuary sampling will occur. It seems the sampling will occur only in the lower reaches, and this may not be sufficient to tie in with other work, e.g., POST tagging at Sand Island, LCREP work in the marshes, etc.

5. Monitoring ocean conditions. Consider greater use of ocean monitoring data collected by other (non-BPA funded) projects for developing indices of ocean conditions, such as hydroacoustics, remote sensing, oceanographic buoys and floats, and robotic vehicles. The ISRP recommends improved coordination and collaboration with other projects and programs collecting these data.

6. Hatchery vs. wild salmon. Consider a detailed comparison of differences in condition, growth, and survival between hatchery and wild salmon of each species. The Endangered Species Act protects many salmon and steelhead ESUs in the Columbia Basin, yet this study does not address hatchery versus wild salmon issues. Hatchery salmon are released at a large size and have high lipid content, therefore hatchery fish may respond differently to environmental factors compared with wild salmon. In earlier years, many hatchery salmon were not marked and could not be readily identified. However, in recent years, including 2010, nearly all hatchery Chinook and coho salmon and steelhead, with the exception of some tribal and conservation hatchery fish, will receive an adipose fin clip. Relatively small numbers of hatchery Chinook raised in conservation hatcheries will not be marked. The ISRP recommends a detailed comparison of hatchery versus wild salmon of each species.

7. Genetic stock identification. The ISRP recommends standardization of genetic stock identification methods used by BPA-funded estuary and ocean survival projects so that results are directly comparable among projects. Different projects may currently be using different methods, but this was not clearly explained in the proposal.

8. Otolith microchemistry. The ISRP considers the value of otolith microchemistry research uncertain. The proponents need to consider specifically how this method can provide new information without extensive baseline data collection. The validity of the proposed use of genetic methods to identify stock origin of individual fish sampled for otoliths needs to be demonstrated. Use of daily otolith increments to estimate estuary and plume residence times is also uncertain. For example, project results to date have estimated that yearling Chinook salmon spend several months in the estuary/plume, which is contrary to evidence from trawl survey and tagging research. Hatchery fish are known to have high Sr/Ca ratios because of their feed. Is this another factor that will confound the proposed microchemistry work? Also the Sr/Ca transition cannot distinguish between estuary and plume habitats, an issue that was not clearly described in the proposal. A useful reference is: Elsdon, T.S. and 9 others. 2008. Otolith chemistry to describe movements and life-history parameters of fishes: hypotheses, assumptions, limitations and inferences. Oceanography and Marine Biology: An Annual Review, 2008, 46, 297-330.

9. Avian predation and alternative prey. The ISRP recommends that the effects of Caspian terns be considered in proposed research on avian predation and alternative prey (anchovy). In the estuary, Caspian tern predation is known to be related to river flows and the Pacific Decadal Oscillation. Proponents need to demonstrate collaboration with other avian predation studies.

10. Tag recovery. In addition to collection of coded-wire tags (CWT), all salmon and steelhead sampled during fishing and tagging operations should be examined for recovery of PIT tags and acoustic tags, if this is not already being done. The ISRP recommends using a handheld wand detector, V-Detector, or tunnel detector onboard the survey vessels to examine all salmon and steelhead in survey catches for CWTs, as some Columbia River hatcheries release coded-wire tagged fish that do not have an adipose fin clip.

11. Tagging effects. New proposed research involves acoustic tagging of juvenile Chinook salmon smolts in the Columbia River estuary with VEMCO and JSATS tags and tracking them as they cross several acoustic listening-lines and with mobile units in the estuary to estimate site-specific survival during outmigration. An evaluation of the effects of tagging stress on fish that are smolting is needed, as stress may be considerable and could affect behavior and survival of tagged fish. Although the proponents think survival will be high because of positive test results in 2010, up-estuary release above the receivers at Astoria and Sand Island may be an added stress to smolts that could be evaluated.

12. Collaboration. This project is collaborating with the CDFO Salmon Shelf Survival Study (#200300900) and the Pacific Ocean Survey Tracking (#200311400, POST, re-named COAST) studies. The ISRP appreciates recent improvements in coordination with these projects. Linkages between these and others studies (e.g., JSATS tagging research) in the estuary, plume, and ocean are established, but the degree of coordination needs further explanation and development. For example, the approaches by NOAA and CDFO are somewhat similar, and integration of data collection and analyses to a greater extent would strengthen results. Likewise, the proponents should consider how data from the NOAA, COAST, and JSATS tagging projects can be integrated to provide a more comprehensive analysis of factors affecting salmon survival.

13. Scientific workshop. The ISRP recommends a scientific workshop in 2011 focused on estimation of estuarine and ocean survival, forecasting of adult returns, and adaptive estuary, plume, and ocean environmental assessment for Columbia River Basin salmon and steelhead. Perhaps the proposal should include this workshop. A workshop would help to improve coordination and collaboration, standardization of methods (e.g., genetic stock identification), development of simulation and predictive models, and integration of results among Columbia River Basin estuary and ocean projects. One aspect of all projects that needs work is how to include more detail on sub-stock structure, including hatchery versus wild fish, hatchery release time, area comparisons, in-river migration and associated ocean migration, and more in the models. CDFO and NOAA seem to be taking somewhat different approaches to salmon forecasting, i.e., stoplight charts (red, yellow, and green) with a Bayesian belief network approach by CDFO versus ecosystem indicators by NOAA. Can this be reconciled?

14. Adaptive management. Consider how to better implement adaptive management to forecasted changes in ocean survival in the Columbia River system. Consider experiments designed in concert with hatchery, hydrosystem, and harvest managers to test specific hypotheses related to estuarine and early ocean survival. Proponents have indicated that management could respond to release timing and barging vs. in-river releases based on predictions from their 16 indicators and timing of upwelling, but what do managers say about the feasibility? How can managers respond to pathogen problems identified during this project? Or is this strictly an explanatory variable?

15. Sources of variation in forecasts. Consider whether ocean survival forecasts could be improved by integration of additional sources of variation in freshwater and ocean survival (e.g., ocean harvests of immatures, jacks, and adults in Alaska and Canada; bycatch in commercial groundfish fisheries; and climate and ocean conditions in offshore rearing areas)?

16. Quantitative forecasts. Qualitative methods of forecasting are helpful, but difficult for managers to apply and rely upon. That being said, proponents need to exercise caution in promoting the idea that their monitoring data will eventually lead to reliable, quantitative forecasts of ocean survival of salmon. Clearly, it is a goal of their agency to provide scientific forecasting tools to improve fishery management, but to date all quantitative ocean forecasting tools for salmon have failed, and thus expensive, long-term research vessel monitoring surveys are necessary.

17. Communicating results. Consider developing more effective approaches for communicating project results and forecasts of ocean survival of salmon directly to hatchery, hydrosystem, and harvest managers. The websites, scientific meetings, and peer-reviewed scientific publication are excellent methods for communicating with other scientists, government agencies, educational institutions, and conservation organizations, but are likely not effective tools for communicating directly with hydro, harvest, and hatchery managers.

18. Online proposal. Consider improvements to the online proposal form. Descriptions of methods in the online proposal were overly brief for some reviewers. Methods should provide sufficient stand-alone detail in the online form to enable evaluation of scientific and technical merit. The proposal could be improved if methods and metrics were explicitly stated for each objective. This is a complex proposal with six general objectives, both broad and narrow hypotheses, and “Studies” that provide metrics and methods that are intended to address multiple objectives, but the association between each specific objective and the metrics and methods that are intended to address it are unclear. For example, Study One provides methods and metrics that the proponents indicate address objectives one through six, but it is not entirely clear what methods and metrics presented in Study One address which of the six general objectives. The discussion of results in the online form would benefit from an ecosystem diagram depicting important physical and biological variables and their known or hypothesized interactions (perhaps indicated by arrows between variables). Such a diagram would provide a synopsis of the proponent’s current view of the system and how it might work, and would be beneficial in understanding the proposal. More complete details are needed on sampling methodology and analyses, along with a format that reduces the redundancies. Information on the percent of salaries for the PIs and what outside support they have would also help.

Qualification: A synthesis of this project, as proposed by the proponents, should be completed and reviewed by the ISRP in 2011. This is a productive and worthwhile project that has made significant contributions to understanding relationships between Chinook and coho salmon survival and ocean and plume conditions. The ISRP commends the proposed new research on abundance, distribution, timing and migration of smolts through the estuary. Another important new feature of the project is the proposed analysis of factors affecting sockeye, chum, and steelhead. However, the proposal was not clear on the extent to which data on these species were collected in previous years but not analyzed or reported. Rapid gains in knowledge could be accomplished if previous data on these species were collected, although the ISRP recognizes that these species may not be abundant in coastal research trawl samples. An important outcome of the project has been a qualitative method for forecasting salmon runs that appears to be an improvement over past methods. It is refreshing to see a project that directly addresses management concerns. The ISRP strongly concurs with the proponents that a major synthesis of this work should be completed in 2011. Some important issues to be considered during the contracting process and in the synthesis are listed below: 1. Strategic plan. The ISRP recommends the use of synthesis results to develop a strategic plan that prioritizes project hypotheses and management objectives. The current approach is exploratory and observational, including numerous hypotheses and investigations of trails of evidence dealing with limiting factors ranging from lipids to parasites to bird predation. When arguing for an observational rather than experimental approach, the proponents state that each year/sampling season can be considered an “independent observation.” It seems unlikely that the quantitative values of physical and biological variables are independent between years, that is, there is no between-year autocorrelation. The proponents need to justify this assertion or adjust for it in their statistical analyses, as described in the synthesis objective. The strategic plan should explain in greater detail how interaction issues arising from studying four elements (bottom-up, top-down, food-web, and plume structure) at the same time will be addressed. 2. Achievable objectives. Consider whether stated objectives are achievable. For example, can the objective (discussed in proposal’s introduction) to determine decadal-scale cycles in ocean productivity be achieved? If so, when will the periodic wave length in cycles be known? If changes are periodic events without a fixed wave length or chaotic events, then how will this objective be achieved? 3. Fishing operation effects. Consider important sources of variation in research trawl and other fishing operations and fishing efficiency with respect to what is known about diel, horizontal, vertical, and seasonal distribution of juvenile salmon that might affect time-series observational data on species composition, abundance, distribution, growth, etc., of juvenile salmon in the survey area. 4. Cruise planning and coordination. The ISRP recommends that the proponents provide annual cruise plans to other related projects. The plans should include sufficient detail on how cruises in the plume, estuary, and ocean will be organized and coordinated with these other projects. For example, the current proposal lacks details on how far upstream the estuary sampling will occur. It seems the sampling will occur only in the lower reaches, and this may not be sufficient to tie in with other work, e.g., POST tagging at Sand Island, LCREP work in the marshes, etc. 5. Monitoring ocean conditions. Consider greater use of ocean monitoring data collected by other (non-BPA funded) projects for developing indices of ocean conditions, such as hydroacoustics, remote sensing, oceanographic buoys and floats, and robotic vehicles. The ISRP recommends improved coordination and collaboration with other projects and programs collecting these data. 6. Hatchery vs. wild salmon. Consider a detailed comparison of differences in condition, growth, and survival between hatchery and wild salmon of each species. The Endangered Species Act protects many salmon and steelhead ESUs in the Columbia Basin, yet this study does not address hatchery versus wild salmon issues. Hatchery salmon are released at a large size and have high lipid content, therefore hatchery fish may respond differently to environmental factors compared with wild salmon. In earlier years, many hatchery salmon were not marked and could not be readily identified. However, in recent years, including 2010, nearly all hatchery Chinook and coho salmon and steelhead, with the exception of some tribal and conservation hatchery fish, will receive an adipose fin clip. Relatively small numbers of hatchery Chinook raised in conservation hatcheries will not be marked. The ISRP recommends a detailed comparison of hatchery versus wild salmon of each species. 7. Genetic stock identification. The ISRP recommends standardization of genetic stock identification methods used by BPA-funded estuary and ocean survival projects so that results are directly comparable among projects. Different projects may currently be using different methods, but this was not clearly explained in the proposal. 8. Otolith microchemistry. The ISRP considers the value of otolith microchemistry research uncertain. The proponents need to consider specifically how this method can provide new information without extensive baseline data collection. The validity of the proposed use of genetic methods to identify stock origin of individual fish sampled for otoliths needs to be demonstrated. Use of daily otolith increments to estimate estuary and plume residence times is also uncertain. For example, project results to date have estimated that yearling Chinook salmon spend several months in the estuary/plume, which is contrary to evidence from trawl survey and tagging research. Hatchery fish are known to have high Sr/Ca ratios because of their feed. Is this another factor that will confound the proposed microchemistry work? Also the Sr/Ca transition cannot distinguish between estuary and plume habitats, an issue that was not clearly described in the proposal. A useful reference is: Elsdon, T.S. and 9 others. 2008. Otolith chemistry to describe movements and life-history parameters of fishes: hypotheses, assumptions, limitations and inferences. Oceanography and Marine Biology: An Annual Review, 2008, 46, 297-330. 9. Avian predation and alternative prey. The ISRP recommends that the effects of Caspian terns be considered in proposed research on avian predation and alternative prey (anchovy). In the estuary, Caspian tern predation is known to be related to river flows and the Pacific Decadal Oscillation. Proponents need to demonstrate collaboration with other avian predation studies. 10. Tag recovery. In addition to collection of coded-wire tags (CWT), all salmon and steelhead sampled during fishing and tagging operations should be examined for recovery of PIT tags and acoustic tags, if this is not already being done. The ISRP recommends using a handheld wand detector, V-Detector, or tunnel detector onboard the survey vessels to examine all salmon and steelhead in survey catches for CWTs, as some Columbia River hatcheries release coded-wire tagged fish that do not have an adipose fin clip. 11. Tagging effects. New proposed research involves acoustic tagging of juvenile Chinook salmon smolts in the Columbia River estuary with VEMCO and JSATS tags and tracking them as they cross several acoustic listening-lines and with mobile units in the estuary to estimate site-specific survival during outmigration. An evaluation of the effects of tagging stress on fish that are smolting is needed, as stress may be considerable and could affect behavior and survival of tagged fish. Although the proponents think survival will be high because of positive test results in 2010, up-estuary release above the receivers at Astoria and Sand Island may be an added stress to smolts that could be evaluated. 12. Collaboration. This project is collaborating with the CDFO Salmon Shelf Survival Study (#200300900) and the Pacific Ocean Survey Tracking (#200311400, POST, re-named COAST) studies. The ISRP appreciates recent improvements in coordination with these projects. Linkages between these and others studies (e.g., JSATS tagging research) in the estuary, plume, and ocean are established, but the degree of coordination needs further explanation and development. For example, the approaches by NOAA and CDFO are somewhat similar, and integration of data collection and analyses to a greater extent would strengthen results. Likewise, the proponents should consider how data from the NOAA, COAST, and JSATS tagging projects can be integrated to provide a more comprehensive analysis of factors affecting salmon survival. 13. Scientific workshop. The ISRP recommends a scientific workshop in 2011 focused on estimation of estuarine and ocean survival, forecasting of adult returns, and adaptive estuary, plume, and ocean environmental assessment for Columbia River Basin salmon and steelhead. Perhaps the proposal should include this workshop. A workshop would help to improve coordination and collaboration, standardization of methods (e.g., genetic stock identification), development of simulation and predictive models, and integration of results among Columbia River Basin estuary and ocean projects. One aspect of all projects that needs work is how to include more detail on sub-stock structure, including hatchery versus wild fish, hatchery release time, area comparisons, in-river migration and associated ocean migration, and more in the models. CDFO and NOAA seem to be taking somewhat different approaches to salmon forecasting, i.e., stoplight charts (red, yellow, and green) with a Bayesian belief network approach by CDFO versus ecosystem indicators by NOAA. Can this be reconciled? 14. Adaptive management. Consider how to better implement adaptive management to forecasted changes in ocean survival in the Columbia River system. Consider experiments designed in concert with hatchery, hydrosystem, and harvest managers to test specific hypotheses related to estuarine and early ocean survival. Proponents have indicated that management could respond to release timing and barging vs. in-river releases based on predictions from their 16 indicators and timing of upwelling, but what do managers say about the feasibility? How can managers respond to pathogen problems identified during this project? Or is this strictly an explanatory variable? 15. Sources of variation in forecasts. Consider whether ocean survival forecasts could be improved by integration of additional sources of variation in freshwater and ocean survival (e.g., ocean harvests of immatures, jacks, and adults in Alaska and Canada; bycatch in commercial groundfish fisheries; and climate and ocean conditions in offshore rearing areas)? 16. Quantitative forecasts. Qualitative methods of forecasting are helpful, but difficult for managers to apply and rely upon. That being said, proponents need to exercise caution in promoting the idea that their monitoring data will eventually lead to reliable, quantitative forecasts of ocean survival of salmon. Clearly, it is a goal of their agency to provide scientific forecasting tools to improve fishery management, but to date all quantitative ocean forecasting tools for salmon have failed, and thus expensive, long-term research vessel monitoring surveys are necessary. 17. Communicating results. Consider developing more effective approaches for communicating project results and forecasts of ocean survival of salmon directly to hatchery, hydrosystem, and harvest managers. The websites, scientific meetings, and peer-reviewed scientific publication are excellent methods for communicating with other scientists, government agencies, educational institutions, and conservation organizations, but are likely not effective tools for communicating directly with hydro, harvest, and hatchery managers. 18. Online proposal. Consider improvements to the online proposal form. Descriptions of methods in the online proposal were overly brief for some reviewers. Methods should provide sufficient stand-alone detail in the online form to enable evaluation of scientific and technical merit. The proposal could be improved if methods and metrics were explicitly stated for each objective. This is a complex proposal with six general objectives, both broad and narrow hypotheses, and “Studies” that provide metrics and methods that are intended to address multiple objectives, but the association between each specific objective and the metrics and methods that are intended to address it are unclear. For example, Study One provides methods and metrics that the proponents indicate address objectives one through six, but it is not entirely clear what methods and metrics presented in Study One address which of the six general objectives. The discussion of results in the online form would benefit from an ecosystem diagram depicting important physical and biological variables and their known or hypothesized interactions (perhaps indicated by arrows between variables). Such a diagram would provide a synopsis of the proponent’s current view of the system and how it might work, and would be beneficial in understanding the proposal. More complete details are needed on sampling methodology and analyses, along with a format that reduces the redundancies. Information on the percent of salaries for the PIs and what outside support they have would also help.

This is an innovative project that has yielded new and critically needed information on how conditions in the ocean and plume affect salmon survival. A unique aspect of this work is the ecosystem approach that is taken to understand salmon survival. This approach is highly consistent with science principles in the Fish and Wildlife Program.

Proposals for the project have expanded to encompass new objectives well beyond the scope of those previously reviewed by the ISRP ("research in the Columbia River plume to investigate juvenile salmon growth and survival, and modeling studies to investigate management of Columbia River flows to improve habitat opportunity in the plume").

Therefore, the ISRP qualifies this "fundable" recommendation with a number of questions to be considered (although the ISRP is not requesting a response):

Could the proponents provide a strategic overview that prioritizes their proposed objectives, tasks, and subtasks, including specific information for each task on the PIs and staff, FTEs committed to that task, critical assumptions, experimental design, justification for degrees of freedom (number of years)/statistical significance, specific timelines, and costs supported by BPA? Could proponents provide an effectiveness analysis of the various results sooner than 2009, as well as a specific plan for involvement of hydro managers?

Technical and Scientific Background: The proponents have provided an excellent summary of the technical and scientific background, and the logical need to address the problem to benefit salmon is clearly defined.

Rationale and Significance to Subbasin Plans and Regional Programs: The proposal addresses objectives in the 2000 Fish and Wildlife Program Plan.

Could the proponents relate the proposal to the 2005 research plan and provide some explicit descriptions of how the research will help with Biological Opinions? The sponsors do not indicate whether the proposed work is called for in the Estuary Subbasin Plan.

Relationships to other project: There is evidence in this proposal of good integration within the large group of proponents (n=26 scientists). The modeling work is integrated with only one of the proposed or ongoing estuary projects (20030100). The proponents also relate their research to US Army Corps of Engineers and National Science Foundation funded projects. Some of the proposed work seems to be dependent upon the continuation of projects funded primarily by these other sources, which could be a problem. Will Peterson's Newport time series be funded by this proposal or from some other source? Only passing reference is made to other related and similar projects such as "Acoustic Tracking for Survival" (200311400) and the "inner estuary" (20030100) researchers. Given that the proposed ocean array studies are focused on the plume area, could the proponents enable coordination between these two projects?

At present, one of the PIs plans to participate in the 2006-2009 research vessel cruises of project #200300900 (Canada-USA Shelf Salmon Survival Study). There is duplication between these two projects on some of the proposed research, e.g., bioenergetics modeling. The proponents also plan to work closely with project #200723600 ("Strategic Adaptation of the Federal Columbia River Power System to Climate Variability and Change"), that is, use remote sensing products and habitat metrics. An integrated approach is required to move the products of research in all key habitats to management agencies. Can proponents demonstrate links to specific BPA-funded restoration or salmon management projects that might be potential users of their proposed ecological indicator/run forecast products?

Project history: This innovative project has contributed significantly to understanding how plume and near shore ocean conditions influence salmon survival. Excellent background and history material are provided. The proponents have demonstrated good monitoring for results, a strong publication record, and all data are archived and/or made available for others to use.

Objectives: A more strategic approach is required to select the most important topics to improve understanding of ocean survival. Can the proponents provide a discussion of what they see as the most important subprojects?

The desired outcome of this project (last 2 paragraphs, section F, p. 30) is that products (ecological indicators; forecasts of the effect of climate and ocean conditions on salmon survival) provided each year by the proponents will help BPA managers evaluate the success or failure of various mitigation programs. For example, if return rates of adult salmon from a particular mitigation program are lower than expected, then changes in ocean conditions "would provide a least one reason why." At the end of the next funding cycle (2007-2009), the proponents promise to provide and "in-depth analysis of the efficacy" of their monitoring and to design a smaller-scale, longer-term, cost-effective monitoring program that will provide these products for as long as managers find them useful. Could the proponents conduct this "in-depth" analysis each year? If "in-depth" analysis is postponed until the end of the next funding cycle, the proponents might discover that they have insufficient samples sizes, variables, etc., to produce the desired outcome (run forecasting products). Key to this is whether or not they have sufficient stock-specific data on Columbia River Chinook and coho salmon ESUs.

Tasks (work elements): The comprehensive ecosystem/mechanistic approach is the major strength of this proposal. Most of the scientific methods are based on sound scientific principles. Cutting edge techniques will be used to accomplish many of the objectives. On the other hand, methods for specific tasks (work elements) are often not of sufficient detail to evaluate by the narrative alone. The experimental design is very complex with multiple variables. Throughout the proposal, there is seldom if any explanation of experimental or field sampling design, how sample sizes were determined, or whether sample sizes are sufficient for the proposed statistical tests. Critical assumptions or consideration of alternative methods for specific tasks are usually not presented or discussed. There is some coordination with other projects conducting similar research. However, are the times and areas of proposed surveys complementary or redundant with other projects? The proponents are counting on models to do the integration of results; however, plans for verifying the models are not specified. Mathematical algorithms for computer models are seldom if ever described in sufficient detail to permit evaluation by reviewers from the narrative alone. The benefits of the proposed computer simulation modeling (other than to generate new hypotheses) is questionable given the lack of sufficient time series of field data from objective 1 to validate results. Methods for bringing results to managers are not well described.

Questions and comments by the reviewers on specific tasks are as follows:

Task 1.1a: The proponents imply that individual fish can be identified to stock of origin or ESU of origin. Can the proponents provide details on genetic baselines and data analysis methods?

Task 1.1b: Ocean growth and bioenergetic tasks, as well as most other tasks in this proposal, would be improved if they were genetic stock or ESU specific. Differences in ocean growth and bioenergetics between hatchery and wild fish might be significant, e.g., hatchery fish might start their ocean life with a larger reserve of lipids than wild fish, but did the proponents consider these factors?

Task 1.1c: Can the proponents describe potential problems with otolith techniques? It is not clear if catch location vs. residence time in the Columbia River plume can be resolved by this technique. Sulfur is mentioned as an isotope to be measured. Is this in addition to carbon and nitrogen? It should be.

Task 1.2a: How useful are the avian predator data without direct feeding studies?

Task 1.2b: Pathogen studies would be more useful if they were stock or ESU specific. How were sample sizes established?

Task 1.2C: Would the results be more useful if they were stock-specific?

Task 2.1a: Chinook smolts and fry likely continue to trickle out of the estuary into the autumn as per six life history types described so far. The planned sampling scheme might miss them. Will salmon in the catch be identified to stock or ESU? Will results from purse seine sampling be comparable to trawl sampling used for other tasks? Can the proponents provide detailed descriptions of sampling gear/methods, fishing stations, statistical or analytical procedures?

Task 2.1c: Fine scale studies of salmon and prey in relation to the plume are to be completed in one year (2007); does this assume that data on interannual variation at this fine scale are not necessary? This task is contingent on availability of a large NOAA vessel, as well as analyses performed as a part of studies funded by other grants (NSF, etc.). Can the proponents provide information on the experimental design, sample size/statistical power, etc., to evaluate whether the results would be statistically valid?

Task 2.1d: Can the proponents provide information on permits, methods, analytical details, etc.?

Task 3.1: This seems to be a very complex series of models - as per comments above, have they been chosen strategically?

Task 3.1a: This physical circulation model has already been developed. Can the proponents provide information as to algorithms used, how the model was validated, or how it is integrated with other models?

Task 3.1b: Can the proponents provide details on how the existing model of plankton and nutrient dynamics will be adapted for use in the Columbia River estuary and plume and coupled with the physical circulation model? The proposed computer simulations will be used to fill data gaps, but it is not clear how these will be validated.

Task 3.1c: Can the models be developed so they are stock/ESU specific and related to timing of ocean entry? The SBMs (spatially explicit) would focus on horizontal and vertical variation in salmon prey densities with respect to oceanographic features in and near the Columbia R. plume. Temporal variation is likely to be important but the proposed seasonal scale is likely too broad to capture the critical ocean entry period. The GOA/GLOBEC bioenergetic studies (Beauchamp, UW) mentioned focus on Prince William Sound pink salmon, which have a very different ocean life history than Columbia River coho and Chinook salmon. How would close coordination with this project be beneficial?

Task 3.1d: IBM models of salmon growth and migration might be more useful if they were stock/ESU specific. No mathematical algorithms are provided for modeling movements. Are existing data of fine enough scale to develop a model that can be validated?

Task 3.1e: Can the proponents provide examples of how Ecopath with ecosystem models have proven to be useful for salmon forecasting and management? Salmon are a very minor part of the California Current ecosystem. Could potential problems with this broad-scale snapshot approach be provided?

Task 3.2: A number of predictors (or forecasters? Note: the terms seem to be used interchangeably but in reality are very different, they should use forecasters) are rejected here because they need more degrees of freedom (df). How do the proponents know that the predictors they have chosen have enough degrees of freedom? Forecasts of return rates are dependent on individual genetic assignments, and it is not clear when these will be available. The proponents have some promising ecological indices but need more degrees of freedom. GAMs will be used to estimate return rates. Can methodological details be provided? A key question is whether or not stock/ESU-specific data series and sample sizes are sufficient.

Task 3.3: How do the proponents plan to engage managers? It is not clear how the managers can directly use the products provided. Can the proponents demonstrate direct coordination and input from BPA managers, as well as state and tribal fishery managers?

Monitoring and evaluation: Monitoring and evaluation of results is an integral part of the whole program, and data are used in scientific publications.

Can plans for long term M&E assessment of ocean survival, or conditions that affect ocean survival of Columbia River Basin salmonids be provided? Ultimately, the success or failure of this project will be measured by the utility of the products (ecological indices, run forecasts) to BPA managers. One concern that would benefit from further discussion in the proposal is whether the spatial, temporal, and biological scales/sample sizes are sufficient to provide useful products. In the face of increasing climate variation, it's not likely that remote sensing or computer modeling will ever be a useful substitute for direct sampling and monitoring of juvenile salmon in the Columbia R. plume. An annual "in-depth" evaluation of the efficacy of monitoring would be useful, rather than delaying this to the end of the next funding cycle.

Facilities, equipment, and personnel are better than adequate. Vessels are a key facility for the program and seem to be available. Staff proposed for the work have very good scientific credentials and are exceptionally well qualified. Can information on FTEs/hours of time commitment by the 16 PIs and 10 Associate Investigators, as well as information on which PIs and AIs who will carry out specific tasks be provided?

Information Transfer: Data will be made available in the scientific literature through peer reviewed papers and reports and through talks at scientific meetings and coastal forums. Can the proponents provide a strategy to provide for better transfer of information to people concerned with management of the river (e.g., USCE, hydro groups) since flow dynamics clearly affect the oceanography?

Benefit to focal and non-focal species: Increased knowledge of how oceanographic factors affect salmon survival will provide significant benefits to anadromous salmonids.
It should lead directly to measures that can be undertaken to improve salmon survival in the ocean and forecast return rates of salmon. This ongoing project has demonstrated significant benefits that are likely to persist over the long-term.

There are ample benefits to non-focal species such as non-salmonids and forage species through increased understanding of oceanographic processes. The proposed fieldwork may affect non-focal species, however, in general "reasonable" precautions seem to have been taken. Can information on the catch and bycatch of all non-focal species during trawl and purse seine fishing operations be provided?