Kalama River Fall Chinook Program HGMP page 9

Klickitat Spring Chinook Production Program HGMP page 35

Lewis River (Merwin) Winter Steelhead Program HGMP page 59

Skamania Summer Steelhead (Washougal River) Program HGMP page 82

References: page 104

Attachments: page 105

Comments on WDFW Chinook and Steelhead

Hatchery and Genetic Management Plans

for Columbia River Tributaries

Submitted to Washington Department of Fish and Wildlife

by Washington Trout,

October 28, 2004

INTRODUCTION

Washington Trout has reviewed chinook and steelhead Hatchery and Genetic Management Plans currently made available for public review by the Washington Department of Fish and. We are submitting the following sets of comments to WDFW for its consideration and response under this established public-input process.

We have not reviewed HGMPs for explicitly experimental chinook, steelhead, and chum programs. Washington Trout is still reviewing NOAA Fisheries’ proposed listing decision for Lower Columbia River coho. Without completing that review, we did not consider ourselves adequately prepared to comment on coho operations that could impact the LCR coho ESU. The decision to not review or submit comments on any HGMP currently available for public review should not be interpreted as an endorsement or a rejection of that program. However, where responses in any HGMP parallel or reflect concerns expressed in these comments, we encourage that those concerns be considered.

This review in most part focuses on some general concerns Washington Trout has identified that run throughout all or many of the chinook and steelhead HGMPs. These include our assessment that:

· In general, the HGMPs fail to adequately describe clear performance standards and indicators, or adequately detailed monitoring and evaluation protocols or timetables;

· A number of erroneous and/or unsupported assumptions run throughout the HGMPs;

· There is a consistent failure to quantify, as required, the estimated take of listed Puget Sound chinook;

· The HGMPs fail to describe or reconcile inconsistencies between individual programs and WDFW’s Wild Salmonid Policy;

· The HGMPs fail to consider or discuss credible, meaningful alternatives to current programs.

The HGMPs provide little reason to believe that levels of take of listed species attributable to the proposed programs are being or will be effectively contained, or that hatchery benefits justify those impacts. The HGMPs commit to NO readily identifiable, measurable, or appropriate performance standards or indicators. They fail to identify alternative management actions that will or might be undertaken in light of the evaluation of the results of a clear quantitative monitoring program.

The HGMP process offers the opportunity to evaluate several broad factors including: the justification for a particular hatchery program; the social, cultural, and economic benefits of the program; the current state of the affected listed population; the potential for the program to take listed species, including a credible quantitative estimate of the level of the potential take, and the measures proposed by the program proponents to minimize that take (including a credible quantitative estimate of the expected reduction in potential take and the ongoing monitoring and evaluation of those measures) -- and to weigh these factors against each other in order to determine if take authorization is warranted.

Justifications for programs are at best inadequately described. Risks of several types of adverse impacts to listed populations from the described programs appear to be high. Measures to minimize take are either inadequately described or based on assertions left unsupported. Likewise, the description of proposed methods for monitoring and evaluating those measures are unacceptably vague, at best.

Many of the HGMPs contain essentially the same answers to critical sections of the HGMP Template that deal with program justifications, performance standards, alternatives to the proposed actions, and the monitoring and evaluation of the proposed action, often utilizing the same vague language, consistently failing to adequately address these particular queries.

The following comments can, and should where appropriate, be applied to most if not all of the HGMPS currently available for public and federal agency review. Several sets of detailed comments on specific, representative HGMPs are included. These specific-HGMP comments address in detail and demonstrate the repeated, often redundant failure of the HGMPs to adequately address critical factors that must be evaluated in order to objectively judge whether the proposed hatchery programs should qualify for take authorization. With some slight variation to account for the minor idiosyncrasies of each HGMP, the individual reviews submitted demonstrate how virtually the same set of comments applies to the same evaluated sections of each HGMP. This pattern is repeated throughout the HGMP package.

Rather than repeat essentially the same set of comments for every HGMP, we believe the critical, central issues addressed in our extant comments can in most instances be applied to all or most of the HGMPs. Therefore, where in any other individual HGMP the responses to the specific sections cited in these reviews are substantively similar to those evaluated in these reviews, or fail to adequately provide the types of required information identified in those reviews, then those elements of our comments that can be reasonably applied to those responses should be considered as submitted comments to that individual HGMP, and should be responded to in the context of every applicable HGMP.

GENERAL COMMENTS AND CONCERNS

Failure to Adequately Describe Performance Standards and Indicators, or Monitoring and Evaluation Protocols or Timetables

Sections 1.9 and 1.10 require descriptions of Performance Standards, Performance Indicators and related Monitoring and Evaluation programs and procedures. NOAA Fisheries is very clear that identifiable quantitative measures that are clearly related to program goals and objectives and that can serve as monitoring variables be identified. All HGMPs fail to provide or identify such quantitative measures, and in general seem to confuse assertions of goals and objectives with descriptions of standards and indicators. This issue has arisen early in the process of developing the HGMP Template in connection with the Northwest Power Planning Council’s Artificial Production Review in the Columbia River Basin, and it is germane to quote the Independent Science Advisory Board (ISAB) review of the Draft Performance Standards and Indicators for Artificial Production in the Northwest Power Planning Council’s Artificial Production Review (ISAB 2002-2, February 23, 2000) which NOAA places on its HGMP website as an accompanying document for completing individual HGMPs:

· A standard is a quantifiable state or condition described in such a way that it is easy to determine whether or not it is being met (emphasis added);

· Indicators are a list of measurable metrics that bear directly on the quantitative determination as to whether or not the standard is being met (emphasis added) (p. 6).

No HGMP reviewed rises to this elementary standard of articulating performance standards and indicators relevant to program goals and objectives, especially those objectives concerned with quantifying potential levels of take and minimizing adverse impacts on listed fish.

Genuine and appropriate standards and indicators are essential to the establishment and the implementation of hatchery program monitoring and evaluation plans. They establish the list of measurable metrics that could be employed to monitor program performance. In the absence of such metrics, the HGMPs’ efforts to describe monitoring and evaluation are inevitably vague, and unacceptably so. No timelines are provided for gathering needed information or meeting performance standards.

This vagueness has the result that neither NMFS nor the public can or will be able to determine whether or not any particular program is achieving its stated or, more often, implied goals and objectives, particularly where take of listed salmon and steelhead is concerned. Washington Trout believes that quantitative standards that provide clear threshold levels of potential adverse impact to be avoided need to be stated and then clearly linked to quantitative monitoring variables and monitoring plans containing detailed timelines for achieving biologically appropriate performance standards.

All the HGMPs fail to meet these criteria, despite WDFW responses to public comments on Puget Sound HGMPs (2003) that indicated WDFW understanding of the need to provide the type of specific, detailed information missing from the HGMPs currently available for public review.

Erroneous Assumptions

The HGMPs rely upon tenuous, uncertain, and even false assumptions concerning the rearing and migratory behavior of listed juveniles and the conditions under which competition and predation by hatchery juveniles may occur. The HGMPs simply assume that there is a unique and narrow period of time during which an overwhelming majority of wild juveniles migrate downstream and out of the river basin. Hatchery release protocols that are asserted to be aimed at minimizing adverse impacts on migrating wild juveniles are not reconciled with the planned releases of hatchery juveniles several weeks prior or during periods of wild juvenile swim-up and/or migration.

Wild fish will commonly be available as rearing fish and as downstream migrants in freshwater habitats for a three to four month period. Considerable information in support of this comes from WDFW reports of downstream trapping from the Skagit, Cedar and Green rivers and Bear and Issaquah creeks in the Lake Washington system (Seiler et al. 2002a, 2002b, 2003). These reports provide substantial evidence that wild juvenile chinook downstream migration generally occurs over a protracted period of time ranging from February to July. The majority of this data is noteworthy in displaying no pronounced mode in the timing of wild chinook outmigration. Rather, outmigration appears to be more or less continuous with several small modes scattered from mid-March to mid-June.

These same fish will be available (as documented in other state and tribal reports) in estuaries and nearshore marine habitats for several additional months (e.g., Beamer et al 2000). Thus, there is really an extended period of up to six months in which problems with competition and predation will occur.

Failure to Quantify Take of Listed Salmon and Steelhead

NMFS in its January 5, 2002 guidance document titled Hatchery and Genetic Management Plan Template: Purpose, Applications, and Instructions, clearly directs HGMP applicants to supply a “numerical estimate” of expected take from hatchery operations “as best as possible”. Paragraph G is particularly explicit:

“Under the broad definition of ESA, “take” of listed species will include hatchery activities that lead to harassment, behavioral modification, capture, handling, tagging, bio-sampling, rearing, release, competition, predation, disease transfer, adverse genetic effects, injury, or mortality of listed fish. When “take” of a listed species is expected in the hatchery operation, the ESA requires that a numerical estimate be quantified as best as possible. To meet this objective, a “take table” is appended to the HGMP that, when completed for each hatchery program, will provide a uniform means to report estimated take (see Table 1) (emphasis added).”

The majority of HGMPs consistently report “unknown” for predation losses involving yearling salmonids and otherwise make no effort to estimate the likely numeric range of actual or potential take, nor do they attempt to explain why such estimates were not or could not be made. It is difficult to see how NMFS could credibly approve any HGMP that does not comply with Paragraph G of the Guidance without forcing the listed resource to bear the entire burden of the risk that arises due to the uncertainty of the level of take that is likely to occur from the various acknowledged risk factors related to hatchery practices.

Violations of the Wild Salmonid Policy

Our use of the term “Wild Salmonid Policy” in this review refers to the combination of (1) Policy of Washington Department of Fish and Wildlife and Western Washington Treaty Indian Tribes Concerning Wild Salmonids plus (2) Additional Policy Guidance on Deferred Issues Concerning Wild Salmonid Policy. Both were adopted by the Washington Fish and Wildlife Commission on December 5, 1997, and together became the existing WDFW Wild Salmonid Policy. The first document is commonly but mistakenly referred to as the “joint policy” since it was never adopted by the Northwest Treaty Indian Tribes.

A second obvious repeated violation of the Wild Salmonid Policy is with respect to the allowable percentages of hatchery fish on the spawning grounds (see Table 2, page 16 of the Additional Guidance). For a high level of similarity of hatchery fish, the maximum percent of the wild spawning population that is of hatchery origin should be 5 to 10%. For an intermediate level of similarity, the limit is 1 to 5%, while the limit for low similarity is 0 to 1%. The violations of this policy element are described in detailed responses to individual HGMPs.

Failure to Examine Credible Alternatives

An obvious course of action in view of stated program goals, the alleged and the largely unquantified (or apparently unrealized) benefits resulting from the programs and the potential risks to listed populations, is to reduce or eliminate programs altogether. It appears that consideration of such an alternative would be mandatory.

The Independent Scientific Advisory Board of the NW Power Planning Council released in 2003 a Review of Salmon and Steelhead Supplementation for the Columbia Basin. One of the ISAB’s key findings was an urgent need to develop “robust experiments with unsupplemented reference streams” (emphasis added) in order to adequately quantify the benefits and/or impacts of hatchery supplementation of native salmonid stocks throughout the basin.

Given many programs’ apparently marginal (at best) contribution to catch rates and associated mitigation obligations, it appears that several programs might make a good candidate for the types of experiments surrounding hatchery closures being recommended by the ISAB and NOAA Fisheries’ Salmon Recovery Science Review Panel. The RSRP issued a report of Panel meetings held July 2003, to discuss “how modification or closure of hatcheries provides NOAA Fisheries with opportunities to investigate the experimental effects of hatcheries on wild populations” and to “urge that NOAA Fisheries take the lead in multi-agency efforts to utilize planned hatchery closures and modifications as experimental units.”

These alternatives, it seems to us, would be consistent with overall social and legal mandates listed as “justifications” for many programs, particularly given the value such options would have for other ongoing supplementation experiments. Serious evaluation of these potential alternatives should be undertaken.

Both the ISAB Review and the RSRP report provide exceptionally valuable blueprints for significant and positive reform of Columbia Basin hatchery programs in general, particularly in the context of qualifying those programs for take authorization. The ISAB noted explicitly that, “most (Columbia Basin) hatchery programs are not integrated with natural production because they rely extensively on fish of hatchery-origin for their broodstock. Nevertheless, the hatchery productions from these programs are present in large numbers on the breeding grounds of many natural spawning stocks. In some cases this is deliberate; in others it is inadvertent. Either way, this constitutes a supplementation action.” The RSRP found “an obvious need for additional experiments testing the efficacy of conservation hatcheries and of modified systems that have the joint objective of production and conservation.” In other words, the findings and recommendations of the two reports should be applicable to most if not all programs throughout the basin. We recommend that WDFW utilize the Review of Salmon and Steelhead Supplementation (ISAB; 2003), and the Report for the RSRP Meeting Held July 21-23, 2003 (RSRP; 2003) as guides to developing reasonable alternatives to major characteristics of this and all other HGMPs currently available for public review. We have attached both documents to these comments, to be considered in WDFW’s response to public comments, as well as NOAA Fisheries’ review of submitted HGMPs.

SPECIFIC HGMPs

WE have included four individual sets of comments to the following HGMPs:

· Kalama River Fall Chinook

· Klickitat River Spring Chinook

· NF Lewis River (Merwin) Winter Steelhead

· Washougal River Skamania Summer Steelhead

The submitted reviews present detailed comments addressing and demonstrating the repeated, often redundant failure of the HGMPs to adequately address critical factors or provide sufficient information to meet the requirements of the HGMP Template.

These comments may appear to address directly only a small fraction of the HGMPs currently under review, and they do not directly address every section of the HGMP Template. However, the election of this review to not specifically comment on any individual HGMP or any HGMP-section describing specific hatchery practices should not be interpreted as approval of that HGMP or practice, or a disinclination to review it. The individual comments demonstrate how virtually the same set of comments applies to the same evaluated sections of each HGMP. This pattern is repeated throughout the HGMPs.

Where in any other individual HGMP the responses to the specific sections cited in the reviews are substantively similar to those evaluated in those reviews, or fail to adequately provide the types of required information identified in those reviews, then those elements of our comments that can be reasonably applied to those responses should be considered as submitted comments to that individual HGMP, and should be responded to in the context of every applicable HGMP.

CONCLUSION

The HGMPs by and large share a central flaw, a general unwillingness or inability to adequately quantify estimates of the harm the individual programs may be doing to listed poulations of salmon and steelhead, or estimates of the benefits the HGMPs claim derive from the programs. Levels of take are consistently characterized as unknown. Levels of benefit are described in cursory, vague language; important information about how many fish, caught where, by whom, and at what total value, is rarely if at all provided in sufficient detail. Few adequate standards, targets, or thresholds are offered for either harm or benefit. Efforts to determine and/or monitor levels of harm and benefit, standards for each, or any measures or timetables to meet those standards are not adequately described.

Several fundamental issues must be reconciled in any application for take authorization. What level of take would be authorized? What types of benefits can the level of take be weighed against? What is the difference between current levels and any appropriate standard for those levels? How were those standards determined? How will they be met, when, and how will those efforts be monitored? Failing to provide this important information creates uncertainty that could provide the potentially inappropriate implication that the level of harm is lower than it actually is, and the level of benefit greater. The overall scope and scale of the Columbia River hatchery program is simply too large to responsibly accommodate the level of uncertainty presented in the HGMPs.

Given these significant shortcomings, this review finds the applications apparently inadequate to justify take authorization under the criteria enumerated in the 4d Rule.

WDFW Hatchery and Genetic Management Plan:

Kalama River Fall Chinook Program

Comments prepared and submitted by Washington Trout

PO Box 402 Duvall, WA 98019; 425/788-1167; wildfish@washingtontrout.org

October 28, 2004

Section 1.6

The response states: “The proposed integrated strategy for this program is based on WDFW’s assessment of the genetic characteristics of the hatchery and local natural populations, the current and anticipated productivity of the habitat used by the populations, the potential for successfully implementing as isolated program, and NOAA’s proposed listing determination (69 FR 33102; 6/14/2004). Details on the methods WDFW employed to make these important assessments, and the actual findings of those assessments, would be helpful in this section.

The response expresses confidence that, mass marking would make “possible” an “accurate level of integration,” and reports that funding has been requested to begin implementation of an effort to mark all program releases. No information is provided about the funding request, the amount of the request, how long it has been pending, or what its current status is. How will the proposed timeline be affected if successful receipt of the funding is not secured?

Section 1.8

Guidance from the HGMP Template clearly instructs that responses to this section should indicate how integrated harvest programs will be operated to provide fish for harvest “while minimizing adverse effects on listed fish.” This would appear to be the primary purpose for the entire HGMP process for take authorization. Insofar as WDFW attempts to designate conservation as a program purpose, it should acknowledge and respond to the Template’s instruction to provide an indication of how the program will “enhance or benefit” the target natural population.

Instead, the response begins with a list of legal and administrative mandates for the purposes of the program (the list might have been more appropriately included in a response to Section 1.7). WT believes that some of the acts, agreements, and decisions listed are at least ambivalent in regards to mandating this or any particular hatchery program. It is entirely plausible that several of these mandates could be satisfied by any number of other approaches. Why is it socially, economically, biologically, or even legally necessary or advisable to satisfy these mandates using this program at this facility? There may be several and varied justifications for meeting these mandates through the Kalama River Fall Chinook Program, but they should be listed and described in sufficient detail to be evaluated and weighed objectively against all direct and indirect take of listed species likely to occur as a result of the program. The response should describe why it is necessary to produce, rear, and release fall chinook under the specific protocols proposed – again, in order that such justification can be weighed against the risk of potential take that may occur, relative to other options, including modifying, scaling back, replacing, or even discontinuing the program.

At any rate, this list of purported mandates does nothing to describe how the program will minimize adverse impacts on listed populations, or how it will benefit the target natural population (or indeed, how it will satisfy the listed mandates!). At least, the level of performance relative to the listed mandates should be described in detail in order to justify any biological risks of the program to listed Mid Columbia River steelhead, Lower Columbia River chinook, LCR steelhead, LCR coho (proposed for Threatened listing) and Columbia River Chum. (While the Kalama River is technically outside the MCR steelhead ESU, the likely geographical and temporal scope of various hatchery-impacts should require examination of potential and likely impacts to those listed populations and an assessment of the appropriateness of take authorization relative to those impacts.) While the response includes skeletal information about the legal, social, and political obligations of the program, and provides a cursory identification of some affected stakeholders, it fails to explain the program’s success at providing any expected mitigation or other benefits, and does not supply any quantitative or even qualitative estimates of the economic or social activity that can be directly attributed to the program.

WDFW appears to assume one or both of two things: that because some or all of the listed mandates predate the listing of the relevant ESUs, the “benefit” of raising fish for harvest in the program has already been established, and should not require detailed explication; or that the mere assertion that the program “protects listed fish and provides harvest opportunity ” is adequate to justify the program. Washington Trout considers both of these assumptions counterintuitive, and a misreading of both the spirit and the specific requirements of relevant 4d Rules and the HGMP template.

At any rate, the response lacks detail sufficient to assure that the program will indeed protect, mitigate, and enhance anadromous fish populations in the Kalama subbasin. In order to meet the HGMP requirement to adequately describe how the program will accomplish these goals, measurable, quantitative standards that provide clear threshold levels of benefits to be achieved and potential adverse impact to be avoided need to be stated, and then clearly linked to quantitative monitoring variables.

Table 1 describes aspects of program operations that are intended to reduce risks of potential adverse impacts on listed fish. However, the table only addresses a few selected risk factors; it does not address potential genetic impacts on LCR (or other) chinook populations, potential migration or other behavioral impacts on LCR (or other) chinook populations, or potential harvest impacts on relevant listed populations associated with harvest activities targeting chinook produced by the program, to name just a few. It does not actually describe in any meaningful way actual risk aversion measures. At best, it describes authorities and/or documents that could govern potential risk aversion measures. In some cases it appears only to describe what risk aversion measures are necessary.

References to other sections in the HGMP provide little clarification. Neither the table nor the references include or refer to appropriate measurable quantitative standards, and/or rely on dubious or unjustified assumptions and assertions about the sources of adverse impact and how they may best be minimized. The table itself and the information provided in the relevant HGMP references are repeated (often verbatim) without extensive case-specific qualifying information in nearly every HGMP. In some cases, program and/or species names are inappropriate to the specific HGMP, clearly indicating a cut-and-paste approach, and a sloppy one at that. It is frankly hard to escape the impression that WDFW does not take this particular ESA-related responsibility as seriously as it should. Boilerplate assertions are not an adequate substitute for honest, thoughtful, specific analyses of individual WDFW hatchery programs and their potential harmful impacts on listed salmon and steelhead populations.

Table 1 references Section 7.9 of the HGMP to provide explication of “Risk Aversion Measures” relative to the “Potential Hazard” associated with Broodstock Collection and Adult Passage. In Section 7.9, NOAA queries applicants: “Indicate risk aversion measures that will be applied to minimize the likelihood for adverse genetic or ecological effects to listed natural fish resulting from the broodstock collection program.” WDFW’s response includes the following broad assertion: “Every effort shall be made to promote local adaptation of this spring chinook population and out of basin hatchery transfers of eggs or fish for use as broodstock will only be considered in extreme cases.” While this response may indicate good management-intent (an impression weakened by the same boilerplate response included verbatim in HGMP after HGMP), it is entirely too vague to provide any explication of the specific “Risk Aversion Measures” the program will undertake to avoid relevant potential hazards. A detailed description of any of the proposed “efforts” is entirely lacking. How does WDFW define “local adaptation,” and how will it be promoted? What exactly will qualify as “extreme circumstances” that would prompt WDFW to consider out of basin hatchery transfers of eggs or fish?

This unacceptable lack of detail continues throughout the response. The statement that no known genotypic, phenotypic, or behavioral differences separate the hatchery and native populations requires some support. How effectively will the program limit take by identifying and immediately releasing non-target listed fish? How exactly will non-target fish be sorted from the holding pond, handled, and released? Can WDFW provide any estimate of expected mortality or decreased spawning success attributable to those encounters, as brief as they may be? The response to Section 7.9 includes no information regarding adult fish passage. The response does acknowledge that mass marking is necessary. WT concurs, but notes that mass marking is not truly a risk aversion measure. Mass marking is a necessary mechanism for monitoring the performance of various potential risk aversion measures. For instance, mass marking program releases would allow managers to monitor the percentage of HORs present in natural spawning areas to determine the risk of genetic interactions with listed chinook. But the mass marking itself only becomes a mechanism for averting risk when it monitors the impacts of a direct management action (i.e. reducing scale of releases) designed to contain or reduce those risks. No information is provided describing how mass marking will be employed to monitor or contain risks associated with the program.

To provide explication of measures WDFW will employ to avoid potential risk associated with competition and predation, Table 1 references Sections 2.2.3 and 10.11 of the HGMP. Section 2.2.3 presents a relatively thorough attempt to support WDFW’s assertion that competition and predation impacts can be minimized by juvenile-release strategies currently employed at most WDFW hatchery programs, including those proposed in this HGMP. This is a major improvement over previous HGMPs submitted by WDFW for public review (see Comments on WDFW Chinook, Coho, and Steelhead Hatchery and Genetic Management Plans for Puget Sound; Washington Trout, August 1, 2003).

However, we believe the response to Section 2.2.3 still fails to address several relevant issues in sufficient detail, relies on some unsupported and sometimes contradictory assumptions, ignores and/or dismisses impacts to neighboring ESUs, and is inappropriately sanguine about the risks inherent in the scientific uncertainty that the response acknowledges surrounds these issues. (See comments on Section 2.2.3.)

The response to Section 10.11 suffers from many of the same weaknesses repeated throughout the HGMP, insufficient detail and support for the broad assertions presented, information contradicted in other sections of the HGMP, a lack of case-specific information and analyses, and no acknowledgement or evaluation of likely risks to neighboring ESUs.

Guidance from NMFS on completing the HGMP Template directs applicants to “cite relevant reports… or other analysis (sic) or plans that provide pertinent background information to facilitate evaluation of the HGMP,” and to “provide additional support of critical information” submitted in the HGMPs. WDFW provides no citations or documentation to support many of the assertions made in the response.

The HGMP should at least attempt to quantify the level of take managers would expect when program operations are configured to “minimize impacts on listed fish.” What is the target level of “minimized” impact? In its January 5, 2000 guidance document (Hatchery and Genetic Management Plan Template: Purpose, Applications, and Instructions), NOAA Fisheries clearly directs HGMP applicants to supply a “numerical estimate” of expected take from hatchery operations “as best as possible” (paragragh G).

We are aware that in its response to public comments regarding Puget Sound HGMPs, WDFW asserted it was under no obligation to provide take estimates in this context. WT frankly disagrees. We find WDFW’s citation in support of this assertion at best inapplicable (WDFW clearly does have the capability to provide some reasonable numerical estimate of potential impact from at least several aspects of its hatchery operations). At any rate WDFW’s response does not reconcile with the unambiguous guidance provided in the January 2000 document, still referenced by NOAA Fisheries on its HGMP Template web page as providing “important instructions for the completion of an HGMP.” (http://www.nwr.noaa.gov/1hgmp/hgmptmpl.htm)

In sum, it appears that the response to Section 1.8 is inadequate to meet the requirements of the HGMP Template. As noted above, without an adequately described justification for the program, there is virtually no way for federal regulators or the public to evaluate or weigh the potential risks of the program against any supposed benefits, regardless of the scope or probability of those risks. The significant shortcomings in this response alone would appear to render this HGMP application inadequate for federal approval.

Sections 1.9 and 1.10

The Tables on pages 5 - 6 listing Performance Standards, Indicators, and Monitoring and Evaluation measures in general either fail to be standards or indicators, or are stated at an inappropriate level of generality. The exact same tables are presented in nearly every HGMP currently available for public review. Taken by themselves the tables do not comply with the guidance provided in the HGMP Template to provide standards that are “measurable, realistic, and time specific” (emphasis added), and to provide indicators that identify “specific parameters to be monitored and evaluated.” (Emphasis added.) The repetition of this inadequate response throughout the HGMP package suggests an attempt to avoid thorough, case-specific analyses of the risks to listed salmon and steelhead populations presented by WDFW hatchery programs, and/or the specific measures that could potentially eliminate, minimize, or even evaluate those risks.

The tables presented in Section 1.10 are each designated as applicable to either “benefits” or “risks” per Template instructions. In general, the table listing performance standards and indicators for “benefits” comes closer to meeting the Template requirements, insofar as the standards listed correspond to samples provided in the NPPC Artificial Production Review. However, in most cases the Indicators listed do not reach the level of measurable specificity directed by the APR.

For instance, the table identifies the following standard: “Assure that hatchery operations support Columbia River fish Mgt. Plan (US v Oregon), production and harvest objectives.” This appears to adequately correspond to the following sample provided in the APR: “Fish produced for harvest are produced and released in a manner enabling effective harvest, as described in all applicable fisheries management plans, while avoiding overharvest of non-target species.” (The caveat of “avoiding overharvest of non-target species” is addressed in the following table; see below.) However, the corresponding Performance Indicators do not reach the level of specificity provided in the APR samples:

Indicator: Annual number of fish produced by this program caught in all fisheries, including estimates of fish released and associated incidental mortalities, by fishery.

Indicator: Annual numbers of each non-target species caught (including fish retained and fish released/discarded) in fisheries targeting this population.
Indicator: Recreational angler days, by fishery.
Indicator: Annual escapements of natural populations that are affected by fisheries targeting program fish.
Indicator: Catch per unit effort, by fishery.

This higher level of specificity is required in order to adequately judge how the program performs relative to objectives and applicable mandates, and to weigh the relative value of that performance against negative impacts to listed stocks that may occur. It is not credible that WDFW would not have this information available. (In the context of the HGMP Template, the actual numerical estimates corresponding to these sample indicators should be provided: i.e., “Indicator: 848 Kalama-Program fall chinook harvested; 424 tribal; 222 non-tribal commercial; 222 non-tribal recreational; etc.” This is implicit in the guidance provided in the APR.)

The performance indicator provided is derived from a table provided in response to Section 1.12, reporting “current” program performance. It does not seem appropriate to merely tally up past program performance – not apparently related to any particular goal at all – and simply declare it an acceptable performance indicator for future program operations, particularly when that performance appears to be low both objectively and relative to stated program goals. Expressing the indicator as a total of “harvest plus escapement” masks an apparently low contribution to total catch rates, an average of 848 chinook. There is no accompanying discussion of how adequately these numbers “support” harvest or mitigation objectives, or how such a contribution to sport, tribal, and commercial fisheries can be considered at all “meaningful.” Indeed, the way the numbers are presented in the performance table creates the impression that the program’s contribution to harvest objectives is higher than it actually is.

The listed Performance Standard, “Program contributes to fulfilling tribal trust responsibility mandates and treaty rights” corresponds adequately to a relevant sample provided in the APR. However, the corresponding Performance Indicator is not an indicator (it really does nothing more than restate the standard), and does not even approach the measurable specificity of the samples provided in the APR:

Indicator: Total number of fish harvested in tribal fisheries targeting this program.
Indicator: Total fisher days or proportion of harvestable return taken in tribal resident fisheries, by fishery.
Indicator: Tribal acknowledgment regarding fulfillment of tribal treaty rights.

(Again, terms like “Total number” and “Total fisher days” should be replaced by actual numerical estimates.)

The listed Performance Standard, “Region-wide, groups are marked in a manner consistent with information needs and protocols to estimate impacts to natural and hatchery origin fish” corresponds adequately to a relevant sample provided in the APR. (However, some description of the relevant “information needs and protocols” would be helpful.) But again, the corresponding Performance Indicators are little more than a restatement of the standard, and do little to identify any “specific parameters to be monitored” (HGMP Template), or provide “measurable metrics that bear directly on the quantitative determination as to whether or not the standard is being met” (Artificial Production Review; p 3). The listed indicators do not acknowledge or reconcile the admission in previous and subsequent sections that program fish are not marked. We include the following examples from the APR for comparison:

Indicator: Marking rate by mark type for each release group.
Indicator: Sampling rate by mark type for each fishery.
Indicator: Number of marks of this program observed in fishery samples, and estimated total contribution of this population to fisheries, by fishery.

(Again, terms like “Marking rate,” “Sampling rate,” “Number of marks,” and “estimated total contribution” should be replaced by actual numerical estimates.)

The table does not include relevant performance standards/indicators for potential benefits associated with every purpose, goal, and/or justification of the program presented in sections 1.6 – 1.8. The APR provides sample performance standards and corresponding performance indicators for mitigation requirements, conservation of wild/naturally spawning populations, and socio-economic effectiveness (see Artificial Production Review; pp 6-19). None of these are included in the table. If nothing else, this illustrates an inherent weakness and inefficiency in WDFW’s “one size fits all” cut-and-paste approach to the development of the HGMPs currently available for review.

The “risks” table is even less adequate for the purposes of take authorization. “Minimize impacts and/or interactions to ESA listed fish” is not a bone fide standard, but at best is a program goal. The APR defines a performance standard as, “a quantifiable state or condition described in such a way that it is easy to determine whether or not it is being met.” (Emphasis added.) There is nothing quantifiable in this purported “standard,” and it is described in a way that actually makes it impossible to determine whether or not it is being met. What is meant by “minimize?” Will impacts be minimized relative to current levels, or some otherwise determined unacceptable level? What are those levels? How were they identified and/or determined? Exactly which impacts and/or interactions will be minimized, and how will that minimization be achieved? How will compliance with this “standard” be judged?

The APR provides a range of performance standards/indicators that would be pertinent to the goal of minimizing impacts and/or interactions to ESA listed fish:

Releases are sufficiently marked to allow statistically significant evaluation of program contribution to natural production, and to evaluate effects of the program on the local natural population.
Fish collected for broodstock are taken throughout the return or spawning period in proportions approximating the timing and age distribution of the population from which broodstock is taken.
Broodstock collection does not significantly reduce potential juvenile production in natural rearing areas.
Life history characteristics of the natural population do not change as a result of this artificial production program.
Annual release numbers do not exceed estimated basin-wide and local habitat capacity, including spawning, freshwater rearing, migration corridor, and estuarine and nearshore rearing.
Patterns of genetic variation within and among natural populations do not change significantly as a result of artificial production.
Collection of broodstock does not adversely impact the genetic diversity of the naturally spawning population.
Artificially produced origin adults in natural production areas do not exceed appropriate proportion of the total natural spawning population.
Juveniles are released on-station, or after sufficient acclimation to maximize homing ability to intended return locations.
Adult broodstock collection operation does not significantly alter spatial and temporal distribution of any naturally produced population.
Weir/trap operations do not result in significant stress, injury, or mortality in natural populations.
Predation by artificially produced fish on naturally produced fish does not significantly reduce numbers of natural fish.

(Note: We acknowledge that not all of these standards would be specifically applicable to this particular program in this particular ESU. However, we do believe that WDFW has been inappropriately dismissive of the potential genetic and ecological impacts of this program to other listed populations in bordering ESUs [see comments on Sections 2.2 – 2.2.2], and given the verbatim repetition of this table in nearly all the HGMPs currently available for public review, we wish to make these comments as broadly applicable to every HGMP as possible.)

The performance indicators provided in the table fail to meet a level of specificity to qualify as actual indicators. As in the Benefits table, they are generally no more than restatements of the purported standards, or at best lists of possible parameters that could serve as indicators, if they were adequately monitored (and described in measurable terms). Few of the items listed under the heading Performance Indicator are clearly stated as a measurable indicator.

The Monitoring and Evaluation Plans corresponding to the standards and indicators presented in the two tables contain no measurable criteria and no specific descriptions of actions associated with attempts to measure either fishery benefits, or impacts of hatchery releases on listed populations.

For instance, the Monitoring and Evaluation Plan corresponding to the purported standard of minimizing impacts to listed fish includes the statement: “instream evaluations of… NOR/HOR ratio on the spawning grounds.” A mere statement that NOR/HOR ratios of adults will be “evaluated” fails to specify a number for an acceptable target-ratio, much less how such a ratio is to be estimated and where and when it will be measured. An adequate Monitoring and Evaluation Plan would include a specification of index areas and frequency of spawner counts during the course of the spawning season together with a description of sample methods and associated sample sizes for estimating ages, sex ratios, and percentage of hatchery-origin fish. (An actual performance indicator associated with such monitoring plan might be "the minimum number of natural origin spawners observed in index reaches A,B, and C, are at least X,Y, and Z with a percentage of females age 4 and older of 90%.")

For the purported standard, “Harvest of hatchery-produced fish minimizes impact to wild populations,” no adequate indicators are provided, except in the most general terms (see comments on “benefits” table, above, for samples of more appropriate performance indicators). Likewise, the description of the corresponding monitoring and evaluation measures, “harvests are monitored by agencies and tribes to provide up to date information” provides little useful information. An adequate monitoring plan for these standards and indicators would at very least describe the methods by which catch will be monitored and survival rates estimated.

The failure to adequately respond to this section is acutely disappointing, given WDFW’s response to comments regarding this exact failure in the Puget Sound HGMPs made available for public review in summer 2003. In that response (WDFW; 2003), WDFW offered to revise those HGMPs to “provide specific, numeric performance measures for key program characteristics.” (Emphasis added.) As far as we know, the public has not been given opportunity to evaluate whether or how thoroughly those revisions have been completed, but the failure in this round of HGMPs to provide anything approaching quantifiable performance standards or include any measurable metrics as performance indicators is far from encouraging in this regard.

Section 1.11.2.

The response identifies the total number of fingerlings to be released annually. While this complies with the letter of the HGMP template, it fails to provide either NOAA Fisheries or the public with enough information to properly judge the scale of the hatchery releases and their potential direct and cumulative impact on listed fish in the river basin in which the releases occur, in the Columbia mainstem, and in the associated estuary and nearshore environments. Some scale of hatchery releases relative to the number of wild listed juveniles likely to be present in these environments during and shortly after the time of hatchery releases is required to adequately judge the size of the program and assess the potential contribution of the releases from specific programs and facilities to the cumulative impact of Columbia-wide hatchery releases on listed fish.

We recommend that in addition to listing specific hatchery-program releases, an estimate also be made of the total numbers (by species) of hatchery salmonid juveniles that will be released in the basin, as well as the total numbers (by species) of wild salmonid juveniles (listed and unlisted) that are expected to be rearing in and migrating through and out of the river basin in which the releases are planned to occur. We further recommend that the HGMP list estimates of the numbers of hatchery juveniles of each species of salmon that are expected to be migrating through and rearing in the Mainstem Columbia between the mouth of the river on which the hatchery in question is located (or in which the hatchery releases occur) and the Columbia estuary and associated nearshore environments, and that these numbers be compared to estimates of cumulative numbers (by species) of wild juveniles. Only this kind of comparative data in addition to the numbers of juveniles proposed to be released by the facility for which the HGMP is written can provide NOAA Fisheries and the public with the appropriate sense of the expected size of the program, and its potential contribution to cumulative impacts on all relevant listed populations from WDFW hatchery programs overall.

Section 1.12.

The response provides data on smolt-to-adult survival rates, catch- levels, though no numbers appear to be provided for total adult production levels. No discussion or analysis accompanies the data reported. The average smolt-to-adult survival rate for the brood years 1995 – 1998 appears to be .0542%, varying from a low of .0299% (1998) to .0815% (1995). No discussion accompanies this to indicate whether these survival rates are acceptable, expected, or a matter of concern, despite clear guidance in the HGMP Template to provide program goals for these parameters.

Insofar as the HGMP proposes to use the average of past program SAR and harvest rates as acceptable performance indicator for future program operations (see Section 1.10), it has an obligation to discuss the relationship of past performance to past goals. Mean performance appears to be low both objectively and relative to stated program goals. The program’s annual contribution to total-catch rates averages 848 chinook. Some discussion is due of how these numbers have and will meet stated program objectives to support meaningful harvest or meet mitigation requirements. The HGMPs proposes to integrate the hatchery and natural populations in the subbasin. It seems that some discussion is warranted regarding whether it is appropriate to integrate a hatchery population with such a low smolt-to-adult survival rate into an already depressed, listed, natural-spawning population. While an SAR rate of .0542% may be adequate for a hatchery population that experiences egg-to-emergence survival above 80%, it would certainly not sustain a naturally reproducing population that must tolerate egg-to-emergence survival as low as 10%.

(Further, the table is confusing to this review. No column provides total adult production levels, so we assume that the sum of “Total Catch” and “Hatchery Escapement” equals total adult production for each of the listed Return Years. However, this is not entirely clear because The HGMP instructs applicants to include estimates for hatchery escapement to “natural areas,” but the information is not provided and apparently unavailable. It is not identified which return year corresponds to which brood year for determining smolt to adult survival. At any rate, summing up “Total Catch” and “Hatchery Escapement” for any particular return year, and dividing by the annual release goal of 5,000,000 [or alternatively 2,500,000] will not consistently render a survival-percentage estimate that matches any presented in the table. It is possible that release goals were not met for some or all of the brood years, but if that is true it should warrant some discussion in this context. A clearer indication of total adult production is needed, with a clearer indication of each return year’s corresponding brood year.)

We believe that more is required in addressing this subsection of the HGMP than has been provided, including a description of a monitoring and evaluation plan that has been (or will be) employed in measuring program performance. Such a monitoring and evaluation plan should include features that monitor program impacts on listed fish. This will require clear statements of measurable performance standards and performance indicators. It will also require statements of appropriate management responses when specific threshold levels of indicators are attained (or fail to be attained, depending upon the manner in which the indicator is stated).

We suggest that the following be included in assessing program performance.

· Stray rates (% hatchery spawners present on spawning grounds with listed fish in specific subbasins): clear upper bounds that are in compliance with the Wild Salmonid Policy guidelines.

· The proportion that the annual number of released hatchery juveniles bears to the estimated annual number of listed conspecific juveniles within the river basin or subbasin where the hatchery releases occur: a clear upper bound combined with a scaling of the absolute number of hatchery juveniles released to the estimated juvenile freshwater carrying capacity of the basin.

· Hatchery smolt-to-adult survival rates, and wild smolt-to-adult survival rates: A lower limit to smolt-to-adult survival rates for hatchery fish should be established. Determination of an appropriate limit should include fitness considerations. Fitness considerations should include considerations of the long-term viability and productivity of the hatchery stock and considerations of the impacts on listed fish of interbreeding with hatchery strays at the upper acceptable level (specified under #1 above). A minimal, biologically acceptable lower limit on hatchery smolt-to-adult survival, however, cannot be purchased at the cost of significant size/condition differentials at the time of release between hatchery and listed juveniles. Limits (performance standards) need to be set on both the maximum size/condition differential between hatchery and listed juveniles and the minimum smolt-to-adult survival rate of hatchery juveniles. Both are required to assure that the program goal of minimizing adverse impacts on listed fish can be attained.

· In addition, a minimum wild smolt-to-adult survival rate should be established that would be sufficient to insure the recovery and long-term persistence of local in-basin populations. Estimation of this rate should take into account the modal value of age-specific female fecundity, the adult population age-structure and sex ratio, the expected range and distribution of variation in survival rates between egg deposition and adult return, and expected harvest impacts. While the role hatchery releases may have in depressing wild smolt-to-adult rates may be unknown or controversial, it is certainly unexamined and un-monitored. Knowing whether and to what extent this may be occurring would appear to be essential to providing an acceptable evaluation of the performance of a hatchery program. This cannot occur without establishing a performance standard for wild smolt-to-adult survival.

Section 1.16.

The “key issues” summarized in the response are straightforward and unsettling. An unknown proportion of hatchery and wild fish in the broodstock combined with “recent estimates” that HORs comprise 80% of natural spawners is an alarming circumstance. The “alternatives” and “reforms/investments” briefly summarized in the response are not all obviously relevant to the identified key issues. Some only restate aspects of the program proposal described in other sections of the HGMP, and/or obligations WDFW must meet to bring the program into compliance with various standards and policies. Some of this information appears inconsistent with descriptions of activities at the “Modrow Trap” that appear in other sections of the HGMP.

The table ending the response and its accompanying text appears to be a broad statement of hoped-for project outcomes for the target stocks. It is vague, confusing, offers no definition of terms or how the listed determinations were made, and at any rate does not address in any way alternatives to the program except to imply that “other strategies” will augment the hatchery program.

Some potential alternatives/reforms with more relevance to an examination of actual program alternatives are identified but not discussed. WT concurs with HGMP suggestions to evaluate potential program-alternatives including: reducing the size of the program; placing “greatest concern” on the ecological risks of competition between wild and hatchery chinook; accurately monitoring the integration of broodstock and naturally spawning populations; and monitoring hatchery and wild chinook-juvenile migration to fill important data gaps regarding competition, predation, and disease-transfer risks to listed steelhead, coho, chum, and fall chinook juveniles.

We suggest that these and other alternatives be discussed at length and in detail. The HGMP Template clearly requires that if clear alternatives to the proposed program were considered that could potentially meet stated program goals or reduce program impacts, they be described and “reasons why those actions are not being proposed” provided.

One of the program goals is to conduct hatchery operations so as to minimize potential adverse impacts on listed fish. Significant thought should be given to ways in which facility operations might be altered or other program goals modified so as to achieve the goal of minimizing potential adverse impacts. These should be enumerated and discussed here together with a statement of reasons for not adopting such changes. At a minimum considerable detail should be provided to support a claim that current operations and goals are sufficiently protective of ESA concerns.

We suggest that the following be considered among the kinds of changes that would better satisfy the goal of minimizing potential adverse impacts on listed fish. 1) reducing the proposed number of juveniles released until stray rates attributable to the program within the basin and neighboring basins are determined to be within the Wild Salmonid Policy guidelines; 2) changing rearing practices so as to produce juveniles that are similar in size and condition to wild conspecifics likely to be rearing in and migrating from the basin during the time of release; 3) within the limits of the facility, releasing juveniles over a more protracted period of time to more closely approximate the temporal distribution of wild juvenile migration, in order to avoid overwhelming wild juveniles with one large pulse of hatchery juveniles.

An obvious course of action in view of stated program goals, the alleged and the largely unquantified (or apparently unrealized) benefits resulting from the program and the potential risks to listed populations, is to reduce or eliminate the program altogether. It appears that consideration of such an alternative in this section of the HGMP would be mandatory.

Given the program’s apparently marginal (at best) contribution to catch rates and associated mitigation obligations, it appears that the Basin might make a good candidate for the types of experiments surrounding hatchery closures being recommended by the ISAB and NOAA Fisheries’ Salmon Recovery Science Review Panel. The RSRP issued a report of Panel meetings held July 2003, to discuss “how modification or closure of hatcheries provides NOAA Fisheries with opportunities to investigate the experimental effects of hatcheries on wild populations” and to “urge that NOAA Fisheries take the lead in multi-agency efforts to utilize planned hatchery closures and modifications as experimental units.”

The RSRP specifically endorsed the findings and recommendations of the ISAB Review, and made several findings and recommendations of its own. The RSRP found, among other things, that “questions on the negative impact of hatchery fish on wild stocks abound …, while scant progress has been made toward investigation and resolution of this major topic.” The report noted that “In all examples that the RSRP has been able to locate, when experiments were conducted to test claims for the success of hatcheries in promoting the conservation of naturally spawning fish, the initial claims have been proven false.” The report found that large-scale experiments involving hatchery closures could test “possible consequences” of current hatchery programs, including:

Do hatchery releases cause extreme ecological stress to natural fish in streams?
Will supplementation hatchery programs increase the number of natural-origin adults on the spawning grounds?
Are there only minor negative consequences of taking wild fish for broodstock?
Is the increased predation on natural-origin fish in a mixed-species fishery significant?
Do hatchery releases seriously influence the marine growth and survival of natural fish?
What is the effect of spawners that are strays from production hatcheries on the genetics of wild stock?

The RSRP explicitly emphasized that “critical data on the demographic and genetic effects of hatchery fish on the wild population can be obtained only by completely eliminating gene flow from the hatchery to the wild population, and by observing demographic and evolutionary changes in both populations as they (re)adapt to their own environments” (emphasis added), and that “a full assessment of interacting demographic and genetic effects of hatchery fish on the wild population can only be obtained from additional experiments in which some hatchery programs (if not the entire hatchery) are completely terminated to remove competition between hatchery and wild fish in freshwater and estuarine habitats.” The report recommends an experimental approach that stops or otherwise modifies hatchery production in a watershed and compares various aspects of subsequent performance of the affected wild salmon population with those in a population whose hatchery operations have not been altered.

These alternatives, it seems to us, would be consistent with overall social and legal mandates listed as “justifications” for the current program, particularly given the value such options would have for other ongoing supplementation experiments. Serious evaluation of these potential alternatives should be undertaken.

Related to the alternatives of program reduction or elimination would be a consideration of how and whether habitat-management efforts could replace or augment hatchery production to meet some program goals at a lower level of biological risk. Efforts on the Skagit River in Puget Sound provide an example for consideration. In 1980 and again in 1990, Seattle City Light (SCL) radically changed the operation of the Upper Skagit dams with increased commitments of flow to better accommodate salmon spawning and rearing. It is apparent there has been a shift of wild Skagit chinook production increasingly into that section upstream of Rockport.

Between 1974-1984 the percentage of the overall wild Skagit chinook population that spawned upstream of Rockport was 62%, between 1985-1993 it was 73%, and between 1094-2001 it was 78% (Connor and Pflug 2003). This sub-stock of chinook is the only one in the watershed that has remained in stable numbers in the period of spawning survey record between 1974-2001. For comparison, these same data indicate that the percentage of change in mean escapement between the 1974-1984 time period and the 1985-2001 time period was +3% for the Upper Skagit while it was -41% for the Lower Skagit and -52% for the Lower Sauk River, the major wild chinook spawning tributary to the Skagit. While the Upper Skagit wild chinook have remained stable, or increased slightly, the remaining basin has been in significant downward decline. From 1974-2001, the overall average wild Skagit chinook population escapement remained relatively stable: 1974-84 - 12,112; 1985-93 - 10,279; 1994-2001 - 11,526. Wild-chinook productivity for the population is being increasingly carried by the Upper Skagit.

Since 1980, SCL mitigation investments became increasingly focused on habitat acquisitions with related habitat protection, habitat restoration, or habitat re-creation projects (personal communication Dave Pflug 2000, 2001, 2002, per Bill McMillan, 2003). This contrasts with hydro electric dam mitigation for fish losses more commonly realized in the form of hatchery programs elsewhere. While Upper Skagit wild chinook have remained stable, the rest of the Skagit basin has remained in wild chinook decline at the same levels as other Puget Sound areas where habitat investments have most often been lower and hatchery domination commonly higher in those other river basins.

The Skagit system is the only place in the Puget Sound region where wild fish have a clear production advantage. Seiler et al. (2002a) show that the 12-year (1989-2000) annual production of wild fry and fingerlings averaged 2.8 million fish. This compares favorably with a relatively modest hatchery program planned for 672,000 fingerlings and 150,000 yearlings.

Evidence suggests that on the Skagit, where emphasis has been on moderation of hatchery chinook production, coupled with relatively intensive habitat protection and recovery, the result has been comparatively high wild fry and fingerling production. This credible alternative, with others, should be discussed and contrasted with the proposed alternative in this section, with a rationale for rejecting any. We do not believe that this HGMP can credibly qualify for take authorization without significant revision to this response.

Section 2.2.1

This response does not adequately address the guidance provided by NMFS in the HGMP Template. The response fails to describe: “adult age class structure, sex ratio, size range, migrational timing, spawning range, and spawn timing; and juvenile life history strategy, including smolt emigration timing.” The response does not address, let alone emphasize, “spatial and temporal distribution relative to hatchery fish release locations and weir sites.”

If these data are unavailable, or inadequate for inclusion in the application, then serious questions arise about the appropriateness of the program at the proposed scope.

In addition to Lower Columbia River chinook, LCR steelhead, and LCR coho (proposed for Threatened listing), the HGMP should examine the potential for impact from this proposed program to listed populations of salmon and steelhead in the Columbia River Chum, Mid Columbia River steelhead, Upper Willamette River chinook, and UWR steelhead ESUs. The HGMP Template instructs applicants to identify all listed populations that may be incidentally affected by the program, including all listed populations in the proposed program’s juvenile-release and adult-return areas.

Hatchery-release areas would include at least all freshwater, estuarine, and nearshore marine environments where hatchery juveniles would be expected to pass through and potentially commingle, interact with, and impose ecological impacts on listed fish. In this case, The HGMP appropriately acknowledges those areas would include the mainstem of the Kalama and the “Columbia River corridor,” which would include the mainstem Columbia, its estuary, and associated nearshore environments. Adult-return areas would include at least all estuarine and freshwater areas where HORs of this proposed program would be expected to pass through and potentially impose ecological impacts on listed fish, or where they might be expected to stray to and interact with listed fish on natural spawning areas, potentially imposing ecological and genetic impacts on listed fish. In this case those areas would include the mainstem of the Columbia and its estuary down stream of the Kalama, and tributaries of the Columbia between its estuary and the mouth of the Kalama. Listed fish from the above-named ESUs will be present in all these areas and “may be incidentally affected by the program.”

Section 2.2.2

The response should include appropriate descriptions of the Mid Columbia River steelhead, Columbia River Chum, Upper Willamette River chinook, and UWR steelhead ESUs. The status descriptions provided do not identify whether the information requested in this section is available for those ESUs. Without estimates of annual juvenile production and basin capacity, the HGMP cannot provide an accurate description of the scale of proposed hatchery-releases relative to the production and capacity of listed fish in the basin or basins. The HGMP should provide the relevant estimates for affected populations in the Lower Columbia River chinook, LCR steelhead, LCR coho (proposed for Threatened listing), Columbia River Chum, Mid Columbia River steelhead, Upper Willamette River chinook, and UWR steelhead ESUs. Estimates should also be provided for total hatchery releases likely to be present in the relevant areas of the Columbia mainstem, so the proposed program’s contribution to the overall scale hatchery releases in the mid and lower Columbia basins can be evaluated.

Section 2.2.3.

The response presents a relatively thorough attempt to describe program activities that may impact listed populations, and the risks of those impacts. This is a major improvement over previous HGMPs submitted by WDFW for public review (see Comments on WDFW Chinook, Coho, and Steelhead Hatchery and Genetic Management Plans for Puget Sound; Washington Trout, August 1, 2003).

However, we believe the response still fails to address several relevant issues in sufficient detail, relies on some unsupported and sometimes contradictory assumptions, ignores and/or dismisses impacts to neighboring ESUs, and is inappropriately sanguine about the risks inherent in the scientific uncertainty that the response acknowledges surrounds some of these issues.

Broodstock Program

In describing broodstock collection at Modrow, the response does not acknowledge information presented in Section 1.16 identifying problems with “unsafe handling of adult listed fish.”

The response acknowledges that program hatchery chinook cannot currently be identified from listed adults, and that indirect take from genetic introgression is “unknown.” While the level of indirect take in this instance may not be specifically “known,” it can be presumed to be quite high, given admissions in other responses that straying of HORs on natural spawning areas is as high as 80%, and that the proportions of natural/hatchery recruits in the broodstock is again, “unknown.” This high potential impact should be discussed at more length in this section, accompanied by proposals to reduce the level of straying to some more acceptable target-level.

Rearing Program

The response acknowledges that the water-intake facility at Fallert Creek is out of compliance with NOAA fish screening standards. The HGMP states that “solutions” have been “identified” to bring the facility into compliance, but no information is provided indicating how long “solutions” have been pending, or what their current status is. No time estimate is provided for bringing the facility into compliance with NOAA standards.

Information provided about fish-passage is confusing, but seems to imply that passage is a concern at Fallert Creek as well, though the response does not include discussion of remedies or timetables other than a cryptic reference to “forwarded” funding requests for “needed improvements.”

The response includes information about water quality permitting and monitoring, but none about applicable water-quality standards or program performance relative to those standards or any other conditions of the relevant permits. The types or amounts of effluent discharged from the hatchery facility are not listed.

Information about disease transfer is vague, if not to say unsettling. The “significant” consequences of outbreaks are described, but little useful information is provided regarding the history of or potential for disease transfer at this facility, other than cryptic indications of “greatly improved” health conditions and success at “reducing” disease outbreaks. What are the pathogens involved? How often have disease outbreaks occurred? What were the consequences? If performance in this regard has improved, by what measure, by how much, and for how long?

Release Program

The response provides relatively detailed information intended to support WDFW’s assertion that competition and predation impacts can be effectively minimized by juvenile-release strategies currently employed at most WDFW hatchery programs, including those proposed in this HGMP. However, the response is weakened by a reliance on some unsupported and sometimes contradictory assumptions, its failure to consider impacts to neighboring ESUs, and its inappropriate reaction to uncertainty.

The response states that 5,000,000 Kalama fall chinook are released “at a time and size that indicates fish are smolted and can emigrate quickly.” The response acknowledges that “Salmon and steelhead feed actively during their downstream migration (Becker 1973; Muir and Emmelt 1988; Sager and Glova 1988) and if they do not migrate they can compete with wild fish,” but then presents information and references to support the assertions that wild and hatchery juveniles are effectively segregated by some behavioral and habitat-preference differences between hatchery and wild juveniles, and “rapid” downstream migration of hatchery juveniles released as smolts.

The response states that, “WDFW is unaware of any studies that have empirically estimated the

competition risks to listed species posed by the program described in this HGMP.” This statement is repeated or paraphrased several times throughout the response and in nearly every HGMP. In fact the bulk and substance of the response is repeated essentially verbatim in nearly every HGMP. In the absence of such a narrow study, the HGMP invests a high degree of confidence in “studies conducted in other areas (that) indicate that this program is likely to pose a minimal risk of competition.” At best the referenced studies “suggest” support for the relevant assertion; they do not “indicate” anything in this context. But never mind semantics; without due consideration of studies in “other areas” that may suggest a potential for more than “minimal” competition-impacts from the proposed program, WDFW’s value-standard risks a bias toward particular assumptions about program performance.

The response acknowledges that releases will occur during periods of low to moderate flow, in the middle of the outmigration period for listed juvenile chinook. The language of the response appears intended to imply that the June release was chosen after consideration of the LCFRB data regarding outmigration timing of listed chinook juveniles, based on the “belief” that a June release “allows listed fish time to grow.” The assertion downplays the higher likelihood that the long-established June-release target was rationalized in the face of the 2004 data, and the belief underpinning the rationalization relies on a gross over-simplification of the temporal distribution of the migration of wild listed juveniles from freshwater to saltwater habitats

Recent data on the timing of wild juvenile chinook outmigration in mid-Puget Sound rivers gathered by the Department's own Wild Salmon Production Evaluation Unit (WSPE) (Seiler et al. 2001(a), 2001(b), 2001(c), 2002, and 2003) provides substantial evidence that wild juvenile chinook downstream migration generally occurs over a protracted period of time ranging from February to July. The majority of this data is noteworthy in displaying no pronounced mode in the timing of wild chinook outmigration. Rather, outmigration appears to be more or less continuous with several small modes scattered from mid-March to mid-June. This suggests it would be extremely unlikely in this case that hatchery smolt releases can be scheduled to temporally segregate them from all wild emigration unless hatchery releases occur in late August or early September. The response fails to acknowledge or address the likelihood of hatchery juveniles interacting with rearing juveniles of other listed species/populations, including populations that employ life-history strategies that include multi-year freshwater rearing.

The response relies on findings in Flagg et al that “it is unclear whether or not hatchery and wild chinook salmon utilize similar or different resources in the estuarine environment.” But it does not discuss current work in Puget Sound that strongly suggests hatchery and wild juvenile salmonids are commingling in near-shore habitats in Puget Sound for significant periods of time before migrating to the open ocean, any attempt at temporal segregation during emigration from upstream, freshwater habitats notwithstanding. Early data from beach-seine and surface-trawl sampling in Skagit Bay in 2002 demonstrate that hatchery-marked and unmarked chinook juveniles of various age and size classes are present together in significant ratios throughout the spring, summer, and fall, in several types of estuarine and near shore habitats. Sampled hatchery-marked juveniles are mixed with unmarked juveniles in mean percentages ranging from 10% to nearly 60% from May through November (personal comm., Casey Rice, NMFS; 2003). Both hatchery-marked and unmarked fish-presence is consistent throughout these periods, but attempts to identify exact ratios of hatchery to wild juveniles are confounded by the fact that some hatchery juveniles released outside but nearby the study area are not visibly marked, and may be entering the study area during certain sampling periods, creating a possibility of undercounting hatchery juveniles during sampling. During the periods that hatchery and wild juveniles are present together in these near shore environments, the hatchery juveniles may enjoy several competitive advantages over their wild counterparts, including most significantly size, which may contribute to create a significant risk of adverse interactions and impacts to listed chinook, including competition, displacement, and predation. WDFW is aware of these preliminary findings. These data should warrant some discussion and analysis in this context, insofar as WDFW is relying on studies in “other areas” to support the assertion that it can successfully minimize adverse impacts to listed populations by effectively segregating wild and hatchery juveniles during freshwater out-migration and rearing life stages.

The response makes a reasonably strong case that temporally and/or spatially segregating hatchery juveniles from listed juveniles can reduce or minimize competition-impacts to listed juveniles in freshwater habitats. However, the response rather overstates the conclusiveness of the evidence presented to support the assertion that wild and hatchery juveniles are being effectively segregated by the release strategies described in this and other HGMPs currently available for public review. The response acknowledges uncertainty about the level of interactions between hatchery and wild juveniles in the Columbia estuary, and about the consequences of those interactions.

The response acknowledges a high likelihood of interactions between hatchery smolts and listed juvenile steelhead and coho. Table 6 shows that steelhead swim-up takes place directly before planned hatchery releases and will be present in sizes at the border of prey-size for hatchery smolts by WDFW’s standard, and well within likely prey-size by other potentially more appropriate standards. No information is provided on the in-basin production of wild listed steelhead fry, relative to the 5,000,000 hatchery chinook-smolts (and the sum of all other hatchery releases in the basin), in order to evaluate the likely scope of potential competitive or predation interactions. The response ignores the potential for encounters with listed juveniles of other ESUs that may be rearing and/or migrating in the mainstem Columbia or its estuary, even though it acknowledges that encounters with LCR chinook, LCR coho, and LCR steelhead could occur in the Columbia mainstem. The admission that “it is unknown to what extent listed fish are available both behaviorally or spatially on the migration corridor” underscores the necessity of considering ecological impacts to other ESUs in the migration corridor, given the admission early in the response that “salmon and steelhead feed actively during their downstream migration.”

The response acknowledges a high level of scientific uncertainty regarding the frequency of competitive interactions between hatchery and wild juveniles, and in determining the consequences of those interactions. But the reaction to that uncertainty is inappropriate.

When faced with genuine uncertainty as to a potential harmful effect of a hatchery practice -- resulting either from data gaps or uncertainty regarding biological mechanisms involved in potentially harmful inter- and intra-specific interactions -- assumptions should be