1. Objectives

This project has been a successful collaboration between universities, federal researchers, and environmental consultants. The 2018 research narrative includes general goals for the management of avian predation rather than specific, measurable objectives with time-based milestones for the research to be conducted. Hypotheses associated with the research were not stated in the summary. The numerous peer-reviewed publications do state clear objectives and hypotheses. The unstated objectives are probably quantifiable and testable based on the mortality attributable to avian predators.

Project results have been very relevant to the Program and have benefited fish (i.e., decreased mortality) but have probably had a negative effect on wildlife (i.e., reduced nesting of avian predators). The proponents indicate that funding is being reduced, and they believe that additional time is needed to address ongoing questions associated with the movement of avian predators and the emergence of additional species of avian predators. They do not, however, indicate specific actions, time frames, or end-dates for the additional work.

2. Methods

Based on the results and the numerous peer-reviewed publications, the methods are scientifically sound. It is relatively easy to measure changes in bird numbers (e.g., colony size) or redistribution of birds, but it was unclear how this was linked to changes in total predation. A linkage between the two should be addressed. The proponents discuss the need to determine whether the avian-based mortality of emigrating salmonids is additive or compensatory, but they do not describe how they would do such a study.

3. Results

Although the project summary does not include quantifiable objectives, the proponents have been successful in providing information to address critical uncertainties (CUs) in the Predation theme.

One concern with results is the plots of predation-rate versus colony-size (Figure 1, Research Narrative). It is difficult to discern or compare the functional relationships between predation rate and colony size in these plots because the x-axes do not start at zero pairs of breeding birds and the y-axes are scaled differently in each plot. The implicit assumption that the relationship is linear (i.e., shown by the dashed lines fitted to the data) is likely unrealistic, at least for Middle Columbia steelhead where the y-intercept appears to be positive, implying avian predation occurs in the absence of birds. It seems more reasonable to fit non-linear curves that pass through the origin. The shape of that curve—called the “functional response to predator density”—and particularly whether the curve is concave up or down has important consequences for modeling the dynamics of predation. The authors should consider presenting estimates of biomass or number of fish eaten by birds rather than showing predation as a fraction of the run.

As a result of this project, three avian predation management plans have been developed and managers have taken steps which have initially reduced mortality of emigrating salmonids to avian predation in the estuary. Reporting is generally excellent and includes many peerreviewed publications. The latest annual report available through PISCES describes detailed results for 2016 and updates long-term time series and provides retrospective evaluations of the whole project. Collaboration and sharing of information among partners appear to be excellent. Despite these successes, many CUs remain, and recent developments suggest that predation impacts from piscivorous colonial waterbirds in the Columbia River Basin may be increasing at a time when funding for avian predation RM&E in the Basin is either being eliminated (USACE) or sharply reduced (BPA).

Rates of predation determined in this study are very site specific, but the study area is broad, including the Columbia River from the mouth (i.e., river km 0) to the head of the impoundment created by McNary Dam (i.e., river km 553). The project also monitors a variety of other piscivorous birds. Recent findings suggest that smolt predation rates by gulls nesting at some colonies in the Columbia Plateau region are comparable to, if not higher than, those of Caspian terns and double-crested cormorants nesting at colonies in the Columbia Plateau region. Summary project data are available at the project website (www.birdresearchnw.org/), and it appears that the raw data are available—perhaps on request.

4. 2017 Research Plan uncertainties validation

Agree. This project addresses CUs in the Predation theme (The list of uncertainties on the 2018 Review database (https://research.nwcouncil.org/2017/Projects) does not include the questions posed in the Research Plan “Critical uncertainties by theme: Predation”).

Specific quantifiable objectives with time-based milestones should be developed to examine additive and compensatory density dependence predation effects. As part of this, the proponents should consider how their work fits in with ecosystem effects of predation. For example, these questions below among others should be addressed: • What is the timing of predation and who are the key bird, fish and mammal predators and prey? • How significant is predation in the lower river, estuary, and ocean plume? • What is the role of prey fish abundance in buffering juvenile salmonids from fish predators? • Where in the salmonid or predator life cycle is management intervention warranted?

Comment:

This is a well-developed and well-designed proposal with obvious benefits for salmonids in the Basin. Avian predation continues to be one of greatest sources of smolt mortality for some ESA-listed salmonid populations, particularly those that must migrate through the foraging ranges of multiple breeding colonies of piscivorous water birds. However, it is important to recognize that the continuing problem of avian predation appears to be a response to widespread human-driven changes in the Basin. These changes include hatcheries that release juveniles within a consistent and limited temporal period, artificially created habitat that is ideal for bird nesting, and dam operations and configuration that slow water flows at a critical time during smolt emigration. Acknowledging that causality, further research, monitoring, and evaluation studies are warranted to more fully understand the ecosystem level effects of avian predation, to quantify changes in predator abundance and predation impacts, and to continue to evaluate, manage, and develop management plans to improve survival of ESA-listed salmonids by reducing avian predation.

The program has been successful in estimating salmonid mortality from local avian predation (e.g., from a single colony) but less successful in estimating basinwide impacts. This is especially true when actions may succeed in reducing bird numbers at a single colony but fail to reduce the total number of birds eating salmonids in the Basin. For example, perhaps birds simply move elsewhere in the Basin and total predation remains unchanged. Perhaps moving birds from the estuary (where non-salmonid prey are also abundant) to upstream sites (dominated by salmonids) leads to worse outcomes. This project has made the first steps in addressing the additive or compensatory nature of bird predation. This question, however, needs much more work for all fish predators throughout the Columbia River Basin.

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

The project has two primary objectives, with several sub-objectives. These are not stated in quantitative terms and with timelines and milestones, but many are inherently quantitative. The proponents have demonstrated through their results from past efforts that the information collected is empirical and well-analyzed. This project has largely achieved past objectives by quantifying the abundance and distribution of piscivorous colonial water birds in the Basin and estimating their predation rate on juvenile salmonids migrating downstream. This knowledge is documented in over 50 peer-reviewed publications and has been used to develop multiple management plans to cost-effectively reduce localized avian predation on salmon smolts.

2. Results and Adaptive Management

Though the project has no explicitly stated adaptive management process, it has a history of rapid adaptation to unanticipated changes in avian predator abundance, behavior, distribution, or responses to management. The project has responded rapidly to cooperating agencies such as USACE, BPA, USFWS, NMFS, state agencies, and PUDs. Several of the project sub-objectives are RPAs in the current Biological Opinion. The proponents have responded thoughtfully to several questions raised by the ISRP in our last review (ISRP 2018-8) and plan to further address these issues through new research described in the current proposal.

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

The proposed research is based on sound scientific principles and rigorous methods. Most of these were developed by researchers who have been world leaders in methods for monitoring bird colonies and measuring avian predation. The proposal describes the measurements used for each sub-objective and the products that will be produced. Exact timelines for each sub-objective are not reported, and it is unclear whether all research elements will be conducted simultaneously or phased through the funding period. The latest Annual Report (for work done in 2017) provides an excellent synthesis of sophisticated modeling to predict Caspian tern population status under alternative management scenarios.

Efforts to collect the data necessary to evaluate avian predation have been achieved through funding from multiple agencies. The proponents are concerned that USACE is no longer funding avian predation studies upstream of Bonneville Dam and that other funding for avian predation studies in the Columbia River estuary is being drastically reduced. Sufficient funding must be provided if this project is required to adequately address the issues and concerns raised by the Council and the ISRP.

This is a well developed, well designed and important program for the Fish and Wildlife Program that supports a clear need that will benefit salmonids in the Columbia River Basin. The investigators have demonstrated that avian predation concentrated in certain specific areas has a large effect on salmonid outmigrant survival. They developed the necessary data to show this need and to support the management plans to move nesting birds and reduce the predation. The work proposed will continue these efforts, support efforts to move cormorants to appropriate nesting locations, and continue to determine the importance of predation by other nesting waterbirds (including the relatively recent arrival of pelicans in the estuary). This study is important to understanding the predation rate of fish-eating birds on various salmon stocks. This rate is being evaluated in considerable detail; however, the predator influence on the overall survival rate of the various stocks seems unknown (is it mostly compensated for or is it additive)? For a true cost-benefit analysis, this question needs to be answered. Perhaps avian biologists working with salmon biologists can address this critical issue by working together on salmon life stage models for various stocks, especially since predation rates seem to vary among species and stocks.

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

Evaluate efficacy and management initiative implemented to reduce predation on juvenile salmonids by Caspian terns and double-crested cormorants on East Sand Island and develop plans for long-term management of avian predation, as warranted. Data indicates that most significant impacts to smolt survival occur in Columbia River estuary, although populations at other sites may be a concern to local stocks. The two avian species now take 15-20 million smolts annually, i.e., ~15% of all smolts. Stocks affected include every ESA-listed stock from throughout the Basin.

The project’s plan is to (1) evaluate efficacy of management initiative implemented to reduce tern predation on East Sand Island, (2) collect compile and analyze data needed for NEPA analyses required to manage cormorants on East Sand Island, and once implemented evaluate efficacy of management actions, and (3) investigate impacts on juvenile salmonids of other piscivorous birds (Brown Pelicans, White Pelicans, Brandt’s Cormorants and gull species), including interactions with smolt losses from Caspian Terns and Double-crested Cormorants. Specifically, assess changes in tern habitat use, colony size, productivity, diet composition, smolt consumption and stock-specific predation rates (associated with reducing the acreage for nesting on East Sand Island from 5 to 1.5 acres. Basically the same approach seems to be followed with cormorants, but it is still in early stages and has not progressed as far.

Nice background information is given, with many detailed publications. Tern population is relatively stable since 1998, but cormorant populations more than doubled. More salmonids are now eaten by cormorants than terns. However, similar to other predator control projects, there is the lingering concern of the importance of predation losses via birds relative to overarching factors such as ocean survival.

2. History: Accomplishments, Results, and Adaptive Management

Systemwide losses of juvenile salmonids to Caspian Terns in estuary amount to 3.6-6.7 million smolts per year, even after management to date. The colony on Rice Island was moved to E. Sand Island (closer to ocean) where diet would hopefully include fewer salmonids. A Caspian Tern dietary change indeed took place (from 90% salmonids to 47% salmonids) with a 62% reduction in consumption of smolts. Further management was needed with the goal in 2006 to redistribute half to two-thirds of E. Sand Island tern colony to alternative sites in Oregon and California, with goal to reduce smolt loss another 50% while still maintaining a viable tern population. Eight artificial islands were constructed in Oregon and California as alternative tern nesting habitat with more nesting islands planned as the size of the nesting area on E. Sand Island is reduced from 5 to 1.5 acres. Double-crested Cormorants on E. Sand Island in 2009 consumed 11.1 million smolts and the colony now represents 41% of the population in western North America. As with the tern, any management of the cormorants will likely require an analysis under NEPA which includes (1) assessment of population status in Pacific states, (2)available suitable alternative nesting habitat outside Columbia Basin, and (3) potential enhancement of salmonid recovery rates in Columbia River should management of cormorants be implemented. The project shows a history of solving problems with fish-eating birds and seems to be planning far ahead to obtain the information needed to assess responses from current or planned management activities. Many results have been published in peer reviewed journals. Raw data have been archived and are available to others. Adaptive Management is clearly demonstrated.

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

Research, monitoring and evaluation of piscivorous waterbird colonies are paramount to success of several regional fish and wildlife recovery plans. Numbers and percentages of smolts consumed by avian predators are needed to assess the success of management activities. Furthermore, the investigators plan to evaluate whether reduction in smolt consumption associated with management of birds in the estuary is compensated by commensurate increases in predation by other avian populations. The diet data will provide information on impacts of avian predation to salmonid and non-salmonid species alike. The study is designed to broaden knowledge of fish mortality through understanding predator-prey interactions (on a species and stock basis). The study is dependent upon PIT tags applied by many salmon researchers. This indicates that avian biologists are working with salmon biologists.

A big question that does not seem to be fully discussed is, “What influence has the smolt loss from avian predation had on the adult return rate (survival) of various salmon stocks?” Some scenarios were discussed, but from what was presented, these seemed to be “what if” type presentations. The time is now to ask the question, “What type of data are necessary to fully understand what percentage of this avian predation loss is additive vs. compensatory?” It would seem like a real cost-benefit analysis needs the answer to this question. Game bird and waterfowl management has been concerned about compensatory and additive mortality for decades when determining harvest rates by hunting, so perhaps some approaches can be found in that literature. Emerging factors, especially climate change as it affects bird distribution, are acknowledged and factored into study designs. The investigators are apparently working with various management agencies across a broad geographic scope.

4. Deliverables, Work Elements, Metrics, and Methods

Fourteen deliverables were described and were very specific with respect to types of data being collected and types of reports that will be prepared (usually journal articles). The techniques were described, and most were standard techniques or techniques they developed and described in earlier years of this study. A few methods were slightly modified when changes needed to be made for a different species. The deliverables were listed for both Caspian terns and double-crested cormorants. Certainly the investigators planned far ahead, e.g., collecting pre-management activity data so that it will be available for assessing responses to the management activities. Seems that many organizations and agencies are involved with this large project and cost sharing and expertise sharing is occurring.

This is a well developed, well designed and important program for the Fish and Wildlife Program that supports a clear need that will benefit salmonids in the Columbia River Basin. The investigators have demonstrated that avian predation concentrated in certain specific areas has a large effect on salmonid outmigrant survival. They developed the necessary data to show this need and to support the management plans to move nesting birds and reduce the predation. The work proposed will continue these efforts, support efforts to move cormorants to appropriate nesting locations, and continue to determine the importance of predation by other nesting waterbirds (including the relatively recent arrival of pelicans in the estuary). This study is important to understanding the predation rate of fish-eating birds on various salmon stocks. This rate is being evaluated in considerable detail; however, the predator influence on the overall survival rate of the various stocks seems unknown (is it mostly compensated for or is it additive)? For a true cost-benefit analysis, this question needs to be answered. Perhaps avian biologists working with salmon biologists can address this critical issue by working together on salmon life stage models for various stocks, especially since predation rates seem to vary among species and stocks. 1. Purpose, Significance to Regional Programs, Technical Background, and Objectives Evaluate efficacy and management initiative implemented to reduce predation on juvenile salmonids by Caspian terns and double-crested cormorants on East Sand Island and develop plans for long-term management of avian predation, as warranted. Data indicates that most significant impacts to smolt survival occur in Columbia River estuary, although populations at other sites may be a concern to local stocks. The two avian species now take 15-20 million smolts annually, i.e., ~15% of all smolts. Stocks affected include every ESA-listed stock from throughout the Basin. The project’s plan is to (1) evaluate efficacy of management initiative implemented to reduce tern predation on East Sand Island, (2) collect compile and analyze data needed for NEPA analyses required to manage cormorants on East Sand Island, and once implemented evaluate efficacy of management actions, and (3) investigate impacts on juvenile salmonids of other piscivorous birds (Brown Pelicans, White Pelicans, Brandt’s Cormorants and gull species), including interactions with smolt losses from Caspian Terns and Double-crested Cormorants. Specifically, assess changes in tern habitat use, colony size, productivity, diet composition, smolt consumption and stock-specific predation rates (associated with reducing the acreage for nesting on East Sand Island from 5 to 1.5 acres. Basically the same approach seems to be followed with cormorants, but it is still in early stages and has not progressed as far. Nice background information is given, with many detailed publications. Tern population is relatively stable since 1998, but cormorant populations more than doubled. More salmonids are now eaten by cormorants than terns. However, similar to other predator control projects, there is the lingering concern of the importance of predation losses via birds relative to overarching factors such as ocean survival. 2. History: Accomplishments, Results, and Adaptive Management Systemwide losses of juvenile salmonids to Caspian Terns in estuary amount to 3.6-6.7 million smolts per year, even after management to date. The colony on Rice Island was moved to E. Sand Island (closer to ocean) where diet would hopefully include fewer salmonids. A Caspian Tern dietary change indeed took place (from 90% salmonids to 47% salmonids) with a 62% reduction in consumption of smolts. Further management was needed with the goal in 2006 to redistribute half to two-thirds of E. Sand Island tern colony to alternative sites in Oregon and California, with goal to reduce smolt loss another 50% while still maintaining a viable tern population. Eight artificial islands were constructed in Oregon and California as alternative tern nesting habitat with more nesting islands planned as the size of the nesting area on E. Sand Island is reduced from 5 to 1.5 acres. Double-crested Cormorants on E. Sand Island in 2009 consumed 11.1 million smolts and the colony now represents 41% of the population in western North America. As with the tern, any management of the cormorants will likely require an analysis under NEPA which includes (1) assessment of population status in Pacific states, (2)available suitable alternative nesting habitat outside Columbia Basin, and (3) potential enhancement of salmonid recovery rates in Columbia River should management of cormorants be implemented. The project shows a history of solving problems with fish-eating birds and seems to be planning far ahead to obtain the information needed to assess responses from current or planned management activities. Many results have been published in peer reviewed journals. Raw data have been archived and are available to others. Adaptive Management is clearly demonstrated. 3. Project Relationships, Emerging Limiting Factors, and Tailored Questions for Type of Work (Hatchery, RME, Tagging) Research, monitoring and evaluation of piscivorous waterbird colonies are paramount to success of several regional fish and wildlife recovery plans. Numbers and percentages of smolts consumed by avian predators are needed to assess the success of management activities. Furthermore, the investigators plan to evaluate whether reduction in smolt consumption associated with management of birds in the estuary is compensated by commensurate increases in predation by other avian populations. The diet data will provide information on impacts of avian predation to salmonid and non-salmonid species alike. The study is designed to broaden knowledge of fish mortality through understanding predator-prey interactions (on a species and stock basis). The study is dependent upon PIT tags applied by many salmon researchers. This indicates that avian biologists are working with salmon biologists. A big question that does not seem to be fully discussed is, “What influence has the smolt loss from avian predation had on the adult return rate (survival) of various salmon stocks?” Some scenarios were discussed, but from what was presented, these seemed to be “what if” type presentations. The time is now to ask the question, “What type of data are necessary to fully understand what percentage of this avian predation loss is additive vs. compensatory?” It would seem like a real cost-benefit analysis needs the answer to this question. Game bird and waterfowl management has been concerned about compensatory and additive mortality for decades when determining harvest rates by hunting, so perhaps some approaches can be found in that literature. Emerging factors, especially climate change as it affects bird distribution, are acknowledged and factored into study designs. The investigators are apparently working with various management agencies across a broad geographic scope. 4. Deliverables, Work Elements, Metrics, and Methods Fourteen deliverables were described and were very specific with respect to types of data being collected and types of reports that will be prepared (usually journal articles). The techniques were described, and most were standard techniques or techniques they developed and described in earlier years of this study. A few methods were slightly modified when changes needed to be made for a different species. The deliverables were listed for both Caspian terns and double-crested cormorants. Certainly the investigators planned far ahead, e.g., collecting pre-management activity data so that it will be available for assessing responses to the management activities. Seems that many organizations and agencies are involved with this large project and cost sharing and expertise sharing is occurring.

This is a strong proposal, and avian predation is definitely a problem that has been documented in a useful series of studies. This project is being funded by a number of entities, the Corps and BPA. The Council/BPA/Corps and the sponsor should clearly delineate who is funding which tasks.

This recommendation is qualified, because the ISRP questions whether it is necessary to condition new sites for the terns (this pertains only to those sites more than 200 miles away), or even evaluate potential new habitat at great distances from the present colonies. These birds are adept at finding suitable habitat when the present habitat is no longer rendered suitable and will likely redistribute to their more historical range, instead of the recent concentration in the Columbia River estuary. More suitable alternative sites need to be provided within 200 miles of the present colonies, because birds need alternative sites or they will not readily move.

Fisheries investigators should consider a similar approach to this project's in sampling for PIT tags in dredge material at Burbank Slough (at mouth of Snake and Columbia).