Monitoring effects of harvest management decisions: mark-selective by O0Xhrcc5

VIEWS: 4 PAGES: 30

									Applying EPA’s Data Quality Objectives Process by CSMEP Harvest
Work Group to Columbia Basin Anadromous Fisheries, working draft
                            6/13/2005.
This process is presently applied at first to the Snake River Basin, but is also intended to illustrate steps
required Columbia Basinwide (including ocean fisheries). It is not the intent of the CSMEP harvest
workgroup to recommend specific analytical approaches for certain decisions, but rather to consider the
M&E requirements for a number of analytical approaches which might be used (CSMEP’s Annotated
DQOtemplatefinal112904.doc). The harvest management and monitoring programs are substantially
different for mainstem Columbia net fisheries (tribal and non-tribal commercial, subsistence, and
ceremonial), ocean fisheries, sport fisheries, and terminal tribal fisheries and different methods may to
used to monitor the targeted catch vs. the incidental catch. This document addresses the monitoring needs
of the mainstem net fisheries. Other similar reports need to be developed for the other types of fisheries.

Step 1. State the Problem

   What is the particular problem to be evaluated?
            o Targeted fisheries on Chinook, steelhead, coho, and (in some years) sockeye are managed
                by setting allowable catch and catch allocation limits and open periods for each fishery
                prior to opening a fishery (considering escapement goals and preseason and updated run
                predictions) and then adjusting those openings and limits as runs develop and catches are
                totaled.
            o Both mark-selective and non-selective fisheries exert mortality on non-targeted stocks of
                anadromous, adfluvial, and resident species that are incidentally intercepted. Removal of
                fish in fisheries can potentially affect the number of mature adults that spawn in natural
                and artificial production areas on a seasonal basis and potentially affect diversity and
                spatial structure of population components on a longer term basis if removals are
                selective of phenotypes (e.g. size, sex or age). Fishing opportunity in areas with mixed
                stocks and species inevitably results in bycatch of non-targeted species or stocks.
                Because such bycatch counts towards the harvest of the bycaught species, it must be
                accounted for. If the bycatch in a particular non-targeting fishery exceeds allowable
                catch or impacts set for that fishery or some other pre-specified limit, then management
                actions will come into play. The type of action will depend on the fishery, on the
                bycaught species and on management agreements in place.
            o Take includes direct harvest, indirect harvest (released fish that die or non-target landed
                fish). It may also be worth considering the impact on fitness of catch and released fish.
   What are the conditions or circumstances that are of concern and necessitate undertaking a
    monitoring program?
            o Escapement goals, harvest rate limits and preseason fishing regulations depend primarily
                on stock-recruitment and cohort analyses and the past history of run entry patterns. Data
                needed for these analyses include total abundance and age composition for each
                population. These data are collected from the fish removed in fisheries and those
                returning to spawn in hatcheries and in natural habitats.
            o In-season adjustments to fishing regulations are made based upon observations of the
                catch and escapement (usually mainstem dam counts) for each management group during
                passage through the mainstem Columbia and Snake rivers. Key parameters are the
                numbers of each age class observed for each management group in the catch and
                escapement estimates.


                                               Page 1 of 30
       Working DRAFT – do not cite or distribute 12/13/2004
            o  Fish removed in fisheries is a component of each brood’s progeny and needs to be
               accounted for in estimates of recruitment and parent/progeny ratios. Stock-recruitment
               relationships allow estimates of adult escapements, recruitment, and productivity at
               different life stages. Bycatch is part of the harvest and therefore is a component of run-
               size calculations. Tracking the components of run size separately over time will provide
               indications of how management actions may impact spawning escapement and provide
               guidance on altering impact changes recommended by bodies such as the TRT.
   Who are the stakeholders (i.e. agencies) ultimately responsible for making decisions related to this
    problem?
           o State agencies and tribes that co-manage fisheries impacting anadromous fish
               populations:

                        Confederated Tribes of the Umatilla Indian Reservation
                        Confederated Tribes of the Warm Springs Reservation
                        Yakama Nation
                        Idaho Department of Fish & Game
                        National Oceanic and Atmospheric Administration Fisheries
                        Nez Perce Tribe of Idaho
                        Oregon Department of Fish & Wildlife
                        Shoshone-Bannock Tribes of Fort Hall
                        U.S. Fish & Wildlife Service
                        Washington Department of Fish and Wildlife

   What other stakeholders are most affected by these decisions?
            o Commercial, recreational, tribal fishers, and associated interests; hatchery programs are
                affected both by too little or excess numbers of broodstock.
   Which entities are charged with developing the technical design of monitoring programs that address
    this problem?
            o Compact Agencies of Oregon and Washington; Technical Advisory Committee to Parties
                of U.S. v Oregon, Tribes..
   What non-technical issues (e.g., political, jurisdictional) could impact development of such
    monitoring designs?
            o Impacts to fish from other “H’s” and changes to fish marking programs are both technical
                and policy issues; other non-technical issues are changes to artificial production
                schedules, consumer market demands and health concerns (toxins).
   Design an overall conceptual model to describe the problem, which could provide the basis for all
    subsequent inputs and decisions for problem analyses
            o Figure 1 describes the general lifecycle of Snake River ESU spring/summer Chinook
                salmon. Figure 2 illustrates a generalized decision making framework. Figure 3 focuses
                on the Tier 3 actions, including harvest actions. One possible model is to plot the
                components of run size over time by stock of interest. For example, natural origin Snake
                River spring Chinook salmon can be intercepted in mark-selective commercial and sport
                fisheries downstream of Bonneville Dam, selective sport fisheries between Bonneville
                and McNary dams, and in the lower Snake River; in traditional Treaty fisheries between
                Bonneville and McNary dams; and in terminal selective sport and Treaty fisheries in
                Snake Basin tributaries. Snake River natural spring /summer chinook are assumed by
                managers to have very low impact rates in ocean fisheries. Few CWT’s from Snake
                River hatchery spring/summer chinook are recovered in ocean salmon fisheries. It is not
                known with great certainty if some ocean fisheries are sampled at high enough rates to
                reliably estimate impacts to Snake River spring/summer chinook or if there are

                                              Page 2 of 30
       Working DRAFT – do not cite or distribute 12/13/2004
                significant impacts to sublegal fish. Canadian fisheries are often not electronically
                sampled and Alaskan fisheries are not electronically sampled for CWT’s. It has also
                been recently suggested by NMFS (by Bob Lohn at a US v. Oregon meeting on May 10,
                2005) that there is some preliminary evidence that Snake River spring/summer chinook
                may be impacted to some degree by certain ocean trawl fisheries such as the Alaskan
                Pollack fishery. (INPUT CR “MAP” WITH FISHERIES AND TIMING LISTED [Tom
                Rein and Eric Tinus]


Step 2. Identify the Decision

   What is the principal CSMEP monitoring question(s) to be addressed (stated as specifically as
    possible)?
    - this focuses the search for information needed to address the problem
             o What are the inseason estimates of run size and escapement for each management group
                 (target and non-target) and how do they compare to preseason estimates? Note: for Zone
                 1-6, the management unit is all upriver spring chinook and Snake River Spring/summer
                 chinook. For Snake Basin tributary fisheries, the management units vary but are
                 generally basin level units.
             o What is the target and nontarget harvest and when is it projected to reach allowable
                 levels?
   What alternative actions could be triggered dependent on results of this monitoring?
             o Open or close various fisheries;
             o Increased or decreased harvest opportunities for fishers.
   Combine the principal monitoring question and alternative actions into a decision statement.
             o Open Fishery X during periods a, b, and c subject to the catch not exceeding Y for target
                 species M and Z for non-target species M.
             o Once the bycatch is projected to reach to quota, then the fishery would be halted,
                 postponed, or reshaped.

"Harvest information is an important component of other decisions. For example, harvest data are used in
mainstem run reconstructions to calculate inter dam survival or conversion rates and harvest rates (Yuen
et al. 2002, Appendix A). Run reconstructions have three basic components: harvest estimates, dam
counts and tributary escapements. For spring-summer chinook, mainstem run reconstructions start at the
tributaries and work downstream to the mouth of the Columbia River. Harvest below Bonneville dam is
estimated for the non-Indian commercial, sport, and miscellaneous fisheries, including total harvest, adult
harvest. Total adult returns to the mouth of the Columbia are then this harvest plus Bonneville dam count.
Information is also required on the harvest between Bonneville and McNary dam and tributary
escapements (for spring chinook only below McNary dam). Tributary escapement methods vary between
tributaries but can include additional tributary dam counts and harvest information. This information is
used to calculate the inter dam conversion rate (Appendix A), which is the inter dam passage survival
after accounting for harvest and tributary turnoff. More precise estimates of harvest will improve run
reconstruction estimates, which will improve status and trend monitoring." Dam counts amount to
“indexes” of dam passage. Dam counts are not begun at each dam at the same time. Dam counts may or
may not include some video counts. Dam counts are typically reported as daytime counts expanded for
non-counting periods (staff breaks) during the day. Managers typically have not adjusted dam counts for
fall back and normally do not try to make passage adjustments for nighttime or lock passage. More
precise dam counts would also increase accuracy of run reconstruction.



                                              Page 3 of 30
       Working DRAFT – do not cite or distribute 12/13/2004
The parts of the decision process specific to the lower Columbia River fisheries depicted in Figure 4 are
summarized as:
   1. Summary Staff Report includes
            A. Examples:
                      i. Staff Reports found at http://www.dfw.state.or.us/OSCRP/CRM/reports.html;
                     ii. Staff Report for Management of the 2003 Columbia River Non-Indian
                          Commercial Spring Chinook Fishery found at
                          http://www.dfw.state.or.us/ODFWhtml/InfoCntrFish/InterFish/03Managementof
                          CommSpringChinookFishery.pdf
            B. Reviews of past run sizes
            C. Estimates of current abundance
            D. Estimates of harvestable fish numbers
            E. Acknowledges management guidelines, e.g. E.g. ESA, US v Oregon Management
                 Agreements
            F. Describes strategy to implement fishery including
                      i. Allocation issues
                     ii. Dates and Times
                    iii. Boundaries of fishable waters
                    iv. Specific description of allowed gears
                     v. Allowable Impacts
   2. Compact Hearing
            A. In-season update of stock status and impacts, largely a review of the summary staff report
                 and any new information since the report.
            B. Describe proposed fishery
            C. Hearing - Stakeholders (Public Meeting w/ testimony) including commercial fishers,
                 sport fishers, treaty fishers, TAC staff (OR, WA, ID, USFWS, NOAAF, CTUIR,
                 CTWSR, YIN, NPT), Joint Staff ODFW WDFW, and interested parties from the public.
            D. Decision - Consensus of WA and OR designees informed by staff report and testimony.
            E. Schedule future Compact hearing to update in-season estimate of catch and escapement.

    3. Joint State Sport Hearing (similar but separate from Compact Hearing –sometimes held just
       before or after)
           A. In-season update of stock status and impacts, largely a review of the summary staff report
                and any new information since the report.
           B. Describe proposed fishery
           C. Hearing - Stakeholders (Public Meeting w/ testimony) including commercial fishers,
                sport fishers, treaty fishers, TAC staff (OR, WA, ID, USFWS, NOAAF, CTUIR,
                CTWSR, YIN, NPT), Joint Staff ODFW WDFW, and interested parties from the public.
           D. Decision - Consensus of WA and OR designees informed by staff report and testimony.
           E. Schedule future Joint State Sport hearing to update in-season estimate of catch and
                escapement.
    4. Each state also has its own process of setting sport fishing seasons and limits for fisheries entirely
       in one state’s waters.
    5. The Nez Perce, Umatilla, and Shoshone Bannock Tribes all set tributary fisheries in the Snake
       River Basin.
    6. Action Notice
           A. Examples:
                     i. Action Notices found at
                         http://www.dfw.state.or.us/OSCRP/CRM/action_notes.html
                    ii. Action Notice for Spring Chinook Commercial Demonstration Fishery found at
                         http://www.dfw.state.or.us/OSCRP/CRM/CAN/05/05 Jan 28 notice.pdf
                                               Page 4 of 30
       Working DRAFT – do not cite or distribute 12/13/2004
          B. FAX (to a list of constituents), News Release, and Internet
    7. Rule filing (simultaneous with Action Notice)
          A. To Attorney General
          B. Creation of an enforceable law for fishery implementation
    8. Conduct and Monitor Fishery
          A. Monitoring
          B. In-season reporting and fishery proposal as stipulated in compact hearing
          C. Examples:
                     i. Fact Sheets found at http://www.dfw.state.or.us/OSCRP/CRM/fact_sheets.html
                    ii. Winter Fact Sheet No. 2 - Jan. 2005 found at
                        http://www.dfw.state.or.us/OSCRP/CRM/FS/05/05 Jan 28 fact.pdf



Step 3. Identify the Inputs to the Decision

   What are the Action Levels relating to this decision (i.e., what are the monitoring outputs that will
    trigger agency response at each node in the decision flow chart)?
             o A specific level of take is allowed based on run forecasts and updated in season run sizes
                where available. Monitoring harvest for that level of take triggers management decisions
                in a fishery.
   What kinds of information are needed to make the decision?
             o Catch, Effort, CPUE, stock identification (for mainstem spring season fisheries, the only
                stock identification used in season is for below Bonneville fisheries where separation
                between Willamette and Upriver stocks are made.
             o Age-specific estimates of the numbers of each management unit (stock) in the
                escapement. Age specific data is only used in forecasting, not in in-season fishery
                management.
   What/where are the sources of this information?
            o Mainstem commercial, subsistence, ceremonial, and sport fisheries, Hatcheries,
              dams, previous fisheries, natural spawning estimates, and mark samples (see
              Table 1).
   What are the main sources of uncertainty of these data and how do the decision processes address
    those uncertainties?
            o The main source of uncertainty is statistical sampling error and perhaps bias due to
                assumption violations, such as error in assumptions regarding release mortality rates.
                Decision making is typically based on point estimates of take and addresses the
                uncertainties by adopting conservative actions.
   Do we need new data or new sampling approaches to make this decision? If so, how will it improve
    on existing data and how feasible is it to obtain?
            o More data, more research on mortality rates resulting from interceptions with varying
                gears (e.g. hook and release, tangle net release, and others). There is considerable
                uncertainty in tangle net release mortality rates. There have also been wide ranging
                estimates made for hook and release mortality rates. There are debates about whether
                hook location, barbed vs. barbless or environmental (i.e. temperature) are the most
                important determinant of release mortality. There are no estimates currently in use for
                net dropout rates for Columbia River net fisheries.
            o Expansion of Genetic Stock Identification (GSI) baseline sampling and in-season
                sampling of catch and escapement. There has been only limited GSI sampling of fish
                from harvest.
                                             Page 5 of 30
       Working DRAFT – do not cite or distribute 12/13/2004
   What are the analytical methods that are appropriate when using this data to make decisions (existing
    and/or new data)
            o Modeling exercise coupled with in-season information to update pre-season expectations
                to make in-season management decisions. Run reconstructions provide a basis for
                developing pre-season expectations (see performance measures in CSMEP C1 tables).
   How will these analyses be used to evaluate Action Levels?
            o Stock-recruitment relationships have been used to set escapement goals for some
                Columbia Basin salmon populations. Cohort analyses have been used to develop pre-
                season expectations. There are no agreed to escapement goals for the entire Snake River
                spring/summer chinook ESU. There are some tributary return goals, but these have not
                been useful for mainstem fishery management.


Step 4. Define the Boundaries of the Study

   What is the target population(s), subpopulations? (e.g. spring summer Chinook in the Snake River
    ESU)
             o ESA listed salmonids including Snake Basin Chinook, sockeye, and steelhead ESU’s. It
                 should be noted that the Snake River spring/summer ESU includes all naturally spawning
                 spring/summer chinook populations in the Snake Basin except from the Clearwater. This
                 fact complicates management below the Mouth of the Clearwater, because estimates of
                 natural origin Snake River spring/summer chinook below the Clearwater are a mix of
                 listed and non-listed fish.
             o All anadromous populations impacted by mainstem and tributary fisheries during the
                 time that Snake River spring Chinook are migrating upstream
   Where are the spatial boundaries of the study (i.e., where are the areas for which data will be
    collected/evaluated; which order streams are of interest)?
             o The Columbia River below Priest Rapids dam, Snake River to the WA, ID border, as well
                 as terminal fisheries in Snake River tributaries.
   What are the temporal boundaries of the study (i.e., what are the appropriate time frames for
    collecting this data)?
             o Annual
             o March through June for all mainstem fisheries
             o May through July for Snake Basin tributary fisheries.
   Are there practical constraints to collecting data that could impact the spatial and temporal boundaries
    of the study?
             o Budget; time required to analyze sample data in-season.
   What is the spatial scale of decision making (i.e., what is the geographic area(s) to which the
    decision will apply)?
             o The Columbia River below the mouth of the Snake River, Snake River to the WA, ID
                 border, as well as terminal fisheries in Snake River tributaries.
   What is the temporal scale of decision making (i.e., what is the time frame(s) for decision making -
    both for the overall target population and for any subpopulations of interest)?
             o Annual and in-season when data are available.


Step 5. Develop a Decision Rule



                                              Page 6 of 30
       Working DRAFT – do not cite or distribute 12/13/2004
A conceptual model of a decision tree is shown in Figure 1 (copied from CSMEP draft hydro
DQO exercise. 12/8/2004). Specific examples could be developed from recent or current harvest
management guidelines (e.g., U.S. v. Oregon Management Agreements, Section 7 or Section 10
permits. 4d rule Plans, State or tribe Fishery Management and Evaluation Plans; etc.

One potentially confounding factor in assessing Tier 3 harvest impacts on a stocks status and
trends is the pre-season forecast. Pre-season forecasts are extensively used in setting allowable
non-target impacts and therefore play a key role in the setting of fishing seasons. Errors and
uncertainties in the pre-season forecast information may mask or exaggerate the Tier 3 harvest
level impacts. Therefore, monitoring efforts should track the measurable uncertainties in those
forecasts to aid in the interpretation of Tier 3 results. Mainstem fisheries impacting Snake River
spring/summer chinook are managed on updated run sizes for all upriver spring and Snake River
spring/summer chinook. This generally allows actual harvest to be kept with in the harvest rate
limits and only occasionally has resulted in overages.

The following five decision rules apply to specific types of fisheries. The technical specifics
have been kept general because they may vary from year to year, among species of concern and
from fishery to fishery (within a grouping) based on the management agreements in place. More
specific information may be found in the appendices.


   1. If the catch of upriver spring chinook and Snake River spring or summer Chinook
      approaches X% of the total upriver spring chinook and Snake River spring summer
      chinook run size at the Columbia River Mouth in the mainstem Columbia River tribal
      spring management period Zone 6 fishery, then the fishery will be closed. X% depends
      on the allowed harvest rate in the management agreement that is based on the updated
      river mouth run size. There is a stepped harvest rate schedule in the current mainstem
      management agreement.

   2. If the catch and/or handling mortality of wild upriver spring chinook and Snake River
      spring/summer Chinook approaches X% of the wild run size in the mainstem Columbia
      River non-tribal commercial or select area fishery, then the fishery will {decision type
      here – in-season adjustments to effort level, gear type, duration, etc.}. The decision will
      depend on if the sport/commercial allocation limit is being approached or if the overall
      wild impact limit is being approached.

   3. If the catch and/or handling mortality of wild upriver spring chinook and Snake River
      spring /summer Chinook approaches X% of the cumulative run in the mainstem
      Columbia River recreational fishery, select area sport fishery, or Washington Lower
      Snake River sport fishery, then the fishery will {decision type here – in-season
      adjustments to effort level, gear type, duration, etc.}. The decision will depend on if the
      sport/commercial allocation limit is being approached or if the overall wild impact limit
      is being approached.

   4. If the catch and/or handle of the Snake River spring or summer Chinook approaches X%
      of the cumulative run in the terminal area tribal fishery in any part of the Snake River

                                           Page 7 of 30
       Working DRAFT – do not cite or distribute 12/13/2004
       Basin, then the fishery will {decision type here – in-season adjustments to effort level,
       gear type, duration, etc.}. The actual harvest limits in any terminal fishery depend both
       on the allowed ESA take if any and the state tribal allocation agreements and escapement
       objectives that may be in place in any year.

   5. If the catch and/or handle of the Snake River spring or summer Chinook approaches X%
      of the cumulative run in the terminal area non-tribal fishery of the Columbia River, then
      the fishery will {decision type here – in-season adjustments to effort level, gear type,
      duration, etc.}. The actual harvest limits in any terminal fishery depend both on the
      allowed ESA take if any and the state tribal allocation agreements and escapement
      objectives that may be in place in any year.



Step 6. Specify Tolerable Limits on Decision Errors

Fish management typically relies on point estimates rather than hypothesis tests so that decision
errors are best characterized by confidence intervals rather than Type I and Type II errors.

Three metrics important to monitor harvest impacts are total catch, stock composition and age
composition. These three sources of information can be used to estimate the total harvest
mortality on any particular stock of interest and assign that mortality to a brood year. Catch
accounting is a primary objective of management agencies and the methods used are varied and
reflect the range of management regulations. It is unlikely that one monitoring plan would be
able to cover catch accounting sufficiently for management needs. Therefore, this exercise is
focused on stock and age composition monitoring.


Stock Composition

Stock composition requires that proportions of stocks be measured, which can then be multiplied
by catch to obtain stock mortalities. Stock composition can be determined from assumptions (as
in run reconstruction), from tagging programs (CWT or PIT tag) or from scale analysis.

Using scale analysis to determine stock/age composition, scales are pressed onto acetate and
these impressions viewed using a microfiche reader to estimate age. At Bonneville Dam, known
age fish are rarely available to provide validation of age estimates. A sample of scales, including
many of the most difficult scales to read, is reviewed by John Sneva of WDFW for corroboration
of age estimates. Spring Chinook are relatively easy to age due to good scale condition with
very minimal scale resorption and over 99% of the fish being of only three age classes (yearling
outmigrants returning after one, two, or three years in saltwater). Summer Chinook scales are
more difficult to age due to a multiplicity of age classes (both yearling and subyearling
outmigrants that return after one, two, three, four, or occasionally even five years in saltwater)
and varying hatchery program rearing strategies. Fall Chinook, steelhead, and sockeye all have a
multiplicity of age classes and fewer different hatchery program rearing strategies. This makes


                                           Page 8 of 30
       Working DRAFT – do not cite or distribute 12/13/2004
scales of these species easier to age than summer Chinook salmon, though still more difficult to
age than scale samples from spring Chinook salmon.

To estimate the total run age composition, weekly age composition estimates are weighted by the
weekly run size. Fish are sampled at the Bonneville Dam Adult Fish Facility, which diverts fish
from the Washington shore ladder. Estimating age from scales collected at Bonneville Dam is
easier than from sites further upstream. The longer salmon spend in freshwater, the more fish
scales erode (resorb) from the outside. This resorption can obliterate one or occasionally even
two saltwater annuli. This problem is particularly pronounced on scales collected off salmon on
spawning grounds, though it can also be a problem at upstream dams (e.g. Lower Granite and
Wells dams) and occasionally at Bonneville Dam. Sampling is typically conducted one to three
days per week for approximately six hours per day, normally between about 9 AM and 3 PM The
assumption is made that fish diverted to this ladder represent the entire run. At this facility, fish
are anesthetized, measured for length, checked for fin marks, and sampled for scales before
being allowed to recover and released.Our goal has been to have a 90% probability of estimating
the percentage of fish belonging to a given age class within 5 percentage points (α=0.10,
d=0.05). Simulations conducted using sockeye salmon suggested that a total sample size of 575
was sufficient to meet total run age composition accuracy and precision goals. Similar results
would be expected for spring, summer, and fall Chinook salmon.

Scale pattern analysis techniques are used to estimate the stock composition at Bonneville Dam.
A classification function is created using linear discriminant analysis, and this function applied to
the Bonneville mixed-stock sample. Mixed stock composition estimates are calculated for each
week and weighted by run size to compute the overall stock composition. Classification has
been successful for spring Chinook, but only for classifying hatchery and natural origin fish
while failing to classify fish by subbasin or hatchery (Schwartzberg and Fryer 1993).

PIT tag readers for CRITFC samplers at the Bonneville Dam Adult Fish Facility would assist in
providing needed corroboration of scale age data with known age fish.For this DQO, the relevant
measure is the proportion of each harvest component made up of the Snake River Spring and
Summer Chinook ESU:

   1. What is the plausible range of values for the population parameter used to evaluate the
      decision?

       Historical percent of catch composed of Snake River spring and summer Chinook ESU.
                                                       N
                          Fishery                   (years)   Mean       Min       Max     STDev
        1. Mainstem Commercial                        25      1.08%     0.00%     4.87%    1.30%
        2. Mainstem Sport                             25      0.70%     0.00%     2.96%    0.77%
        3. Select Area Commercial                     12      0.06%     0.00%     0.18%    0.07%
        4. Select Area Test                           13      0.01%     0.00%     0.02%    0.01%
        5. Select Area Sport                           4      0.01%     0.00%     0.01%    0.00%
        8. Commercial Shad                            19      0.10%     0.00%     0.29%    0.09%
        9. Zone 6 & Snake Sport                        3      0.15%     0.13%     0.18%    0.02%
        10. Zone 6 Winter Gillnet                     25      0.38%     0.00%     4.30%    0.95%
        11. Zone 6 Commercial Gillnet                 25      0.90%     0.00%    10.47%    2.61%
        12. Ceremonial and Subsistence Platform       25      4.86%     2.60%     7.42%    1.49%

                                            Page 9 of 30
       Working DRAFT – do not cite or distribute 12/13/2004
2. What are the sources of total study error in the datasets available for making the
   decision?

   Table 1 describes the sources of study error.

3. What component of this total study error relates to sampling design error?

   Table 1 describes the sources of study error in terms of bias and precision.

4. Are there ways to manage/minimize either of these sources of error?

   Bias error tends to originate from assumption violations. Minimizing bias error typically
   requires focused studies to estimate a bias correction factor or in redesign of the sampling
   program. Without direct evidence of stock composition, e.g. from CWT or PIT tags
   implanted in natal rivers or from stocks where scale pattern analysis has been proven
   accurate, assumptions such as those used in run reconstruction must be substituted. In
   this case, either correction factors could be constructed using a short term but intensive
   tagging program or the problem could be corrected by creating a long-term tagging
   program. For example, in the recreational fishery there is CWT sampling, visual stock
   ID (some stocks are dark, e.g. URB have dark faces, while others are light, e.g.
   Willamettes). The visual stock ID is confirmed via CWTs to separate out Willamette
   from lower CR tributary stocks (Cowlitz, Kalama). So three groups for sampling are
   Willamette, URB, and lower CR tribs – all are represented with CWT’s.

   On the other hand, precision error tends to be linked to sampling effort. Typical methods
   for minimizing precision error is to increase the sampling efforts.

5. What is the apparent current baseline condition of this population parameter? How was
   this established?

   For stock composition of catch see #1. Relative to total catch, the harvest rate
   management objectives are provided in the following table.

   Spring Management Period Chinook Harvest Rate Schedule

     Harvest Rate Schedule for Chinook in Spring Management Period

     Total Upriver
      Spring and   Snake River   Treaty                            Non-Treaty
     Snake River     Natural     Zone 6     Non-Treaty    Total     Natural
       Summer Spring/Summer       Total      Natural     Natural    Limited
     Chinook Run Chinook Run     Harvest     Harvest     Harvest    Harvest
                           1          2,5          3           4          4
         Size         Size       Rate         Rate        Rate       Rate
         <27,000       <2,700        5.0%      <0.5%       <5.5%        0.5%
          27,000        2,700        5.0%       0.5%        5.5%        0.5%
          33,000        3,300        5.0%       1.0%        6.0%        0.5%
          44,000        4,400        6.0%       1.0%        7.0%        0.5%
          55,000        5,500        7.0%       1.5%        8.5%        1.0%

                                        Page 10 of 30
   Working DRAFT – do not cite or distribute 12/13/2004
           82,000              8,200          7.0%           2.0%         9.0%           1.5%
          109,000             10,900          8.0%           2.0%        10.0%
          141,000             14,100          9.0%           2.0%        11.0%
          217,000             21,700         10.0%           2.0%        12.0%
          271,000             27,100         11.0%           2.0%        13.0%
          326,000             32,600         12.0%           2.0%        14.0%
          380,000             38,000         13.0%           2.0%        15.0%
          434,000             43,400         14.0%           2.0%        16.0%
          488,000             48,800         15.0%           2.0%        17.0%



    Footnotes
    1. If the Snake River natural spring/summer forecast is less than 10% of the total upriver run size,
    the allowable mortality rate will be based on the Snake River natural spring/summer Chinook run
    size. In the event the total forecast is less than 27,000 or the Snake River natural spring/summer
    forecast is less than 2,700, Oregon and Washington would keep their mortality rate below 0.5% and
    attempt to keep actual mortalities as close to zero as possible while maintaining minimal fisheries
    targeting other harvestable runs.

    2. Treaty Fisheries include: Zone 6 Ceremonial, subsistence, and commercial fisheries from
    January 1-June 15. Harvest impacts in the Bonneville Pool tributary fisheries may be included if
    TAC analysis shows the impacts have increased from the background levels.

    3. Non-Treaty Fisheries include: Commercial and recreational fisheries in Zones 1-5 and
    mainstem recreational fisheries from Bonneville Dam upstream to the Hwy 395 Bridge in the Tri-
    Cities and commercial and recreation SAFE (Selective Areas Fisheries Evaluation) fisheries from
    January 1-June 15; Wanapum tribal fisheries, and Snake River mainstem recreational fisheries
    upstream to the Washington-Idaho border from April through June. Harvest impacts in the
    Bonneville Pool tributary fisheries may be included if TAC analysis shows the impacts have
    increased from the background levels.
    4. If the Upper Columbia River natural spring Chinook forecast is less than 1,000, then the total
    allowable mortality for treaty and non-treaty fisheries combined would be restricted to 9% or less.
    Whenever Upper Columbia River natural fish restrict the total allowable mortality rate to 9% or less,
    then non-treaty fisheries would transfer 0.5% harvest rate to treaty fisheries. In no event would
    non-treaty fisheries go below 0.5% harvest rate.


    5. The Treaty Tribes and the States of Oregon and Washington may agree to a fishery for the
    Treaty Tribes below Bonneville Dam not to exceed the harvest rates provided for in this Agreement.




6. What level of Type I vs. Type II error will be acceptable for this population parameter?

   Management is dependent on point estimates rather than on hypothesis testing so a
   discussion of precision is relevant as opposed to a discussion of Type I and Type II error.
   There is no defined precision criteria except that a 20% sample is the goal for species
   composition.

   What is the best way to evaluate/represent the quality of the decision process that uses
   data relating to the population parameter?

   The Pacific salmon fishery exists along an extensive area from central California to
   Alaska. The tribal harvest of Pacific salmon stocks has been occurring for thousands of


                                                  Page 11 of 30
   Working DRAFT – do not cite or distribute 12/13/2004
years while the commercial harvest dates back to the Civil War. The Columbia River
Basin provides spawning areas for many salmon runs.

Harvest management within the Columbia Basin is tightly regulated by state and federal
laws, an interstate Compact and an international treaty. Management is focused on two
general goals, allocation of the resources and conservation of the resources. Both goals
require extensive, dedicated analytical systems.

Allocation consists, at first, of political negotiations, or legal determinations, to apportion
the total take among user groups. Briefly, Treaties between the Columbia River tribes and
the United States and the U.S. v. Oregon court case govern issues among Columbia River
harvesters. The Pacific Fishery Management Council makes recommendations to NMFS
regarding allocation, harvest limits, and conservation for ocean fisheries south of the
US/Canada border. The Pacific Salmon Treaty governs international fishery
management issues between southern US fisheries, Canadian fisheries, and Alaskan
fisheries. Each forum has technical committees responsible for analytical work. Although
each management entity has similar roles and precepts, their geographical scopes overlap
in different degrees, such that a collaborative relationship exists between the various
management forums.

The management goals of allocation and conservation are inter-related. Because different
stocks have different productivities, and therefore, different sustainable exploitation rates,
fishery management is largely a matter of shaping fisheries in time and space so that each
stock is not over-harvested.

A variety of information is used in establishing fishing seasons. Run forecasts
(predictions of run size for a given stock) are generated primarily through run
reconstruction (aka cohort analysis.) This involves separating the catch into individual
stocks and ages. Then, earlier returning members of a cohort are used as predictors of
stock strength for the later returning ages of the cohort. Three-year-olds from a cohort are
used to predict four-year-old returns the following year, and four-year-olds are used to
predict five-year-olds the year after that.

Thus run size forecasting involves stock identification and aging technology. Stock
identification has several components. Virtually all hatchery spring chinook, coho, and
steelhead have had their adipose fins clipped to distinguish them from wild fish. Many
hatchery summer and fall chinook are also adipose fin clipped. Many fish, particularly
hatchery chinook and coho, have been marked with coded wire tags (CWTs) that identify
the stock and age when scanned. Increasingly, genetic stock identification (GSI) is
supplementing CWTs for stock identification.

Scale reading is the primary means of aging the catch. Scale samples are taken from some
of the catch and from fish sampled at dams and the lengths of those and other fish are
measured. Ages are determined by reading the scales with a microfiche reader and the
age data, and the length data, are used to create an age – length relationship. This
relationship can then be used to age the larger sample of measured fish. Note that fish

                                    Page 12 of 30
Working DRAFT – do not cite or distribute 12/13/2004
with CWTs can help ground truth the aging process because their ages are known. Scales
are also collected from a sample of Chinook and sockeye salmon and steelhead passing
the Adult Fish Facility on the Washington shore ladder of Bonneville Dam. These data
are used in run reconstruction for run forecasts in future years. In addition, the data are
used in modifying in-season forecasts.


Several management issues are arising within west coast salmon fisheries. First, the
Canadian government is reducing the number of index stocks it monitors. As a
consequence, there will fewer stocks on which to base run size forecasts. There have also
been shifts in harvest times and areas in an effort to lessen effort on weak stocks within
Canadian fisheries. Unfortunately, exploitation rates are based on historical fishing
patterns so that these shifts increase the uncertainty surrounding the exploitation rates of
all stocks in the fishery.

Clipping the adipose fins of all hatchery fish (mass marking) has made it easy to
distinguish between wild and hatchery fish in fisheries. This has enabled the
implementation of selective fisheries. Selective fisheries have in many cases allowed
increased opportunity to harvest hatchery origin fish than would have been feasible
without them. They have not generally been implemented in a manner that has reduced
harvest rates on wild fish. It has also made it difficult to identify fish with CWTs,
formerly the only ones with adipose fin clips. It now requires more sampling effort to
detect the tags because all fish need to be scanned electronically for CWTs. With larger
selective fisheries, there is a decrease in the precision in estimates of harvest impacts on
wild fish largely because of differential harvest rates between marked and unmarked fish
and the difficulty of using CWT’s in marked fish to estimate harvest impacts on
unmarked fish.

GSI technology has the potential of allowing any returning adult to be a sample for stock
identification purposes. However, GSI does not provide the ages of the fish caught.
Because GSI is an emerging technology, there is not full agreement on GSI protocols.
However, coordination is improving.

There is also little coordination among the few individuals who are given the task of scale
reading by their respective agencies. As a result, there have been no studies comparing
the results of stock aging among different scale readers. There have also been no blind
studies comparing estimated ages with the known ages of tagged fish.

Because both stock identification and aging are time consuming, determining stock
specific run sizes does not occur on a real time basis. Increasingly, passive integrative
tags (PIT tags) are being used in survival studies. These tags identify individual fish and
can provide stock identification and age on a real time basis. However, several factors
could impede their use. The tags are more expensive and the tagging procedure is more
costly and difficult. And although the infrastructure (a system of adult detectors) is
becoming well developed at many Columbia River Dams and in Columbia basin
hatcheries, it is virtually nonexistent elsewhere. Given the inexplicable poor returns this

                                   Page 13 of 30
Working DRAFT – do not cite or distribute 12/13/2004
         (2005) year, the ability for real time run assessment may be desirable. In order to
         determine if PIT tags can be useful for real time fishery management and stock
         assessment purposes, several questions need to be addressed. Appropriate indicator
         stocks and tag rates need to be determined. Fishery sampling questions need to be
         resolved (i.e.: sample rates, costs, and the applicability to fisheries where fish may be
         cleaned prior to sampling which would result in tag loss).



Step 7. Optimize the Design for Obtaining data


                                                                           Assigned To:
Existing CSMEP Tables C1 and B2 pilot inventory data do not include a      Topic for discussion at
    pathway for harvest data.                                              workshop.
Discuss promising areas of research:
    Genetic information for stock composition                              Genetics generic language -
                                                                           Annette
   Enhanced data on timing to help shape fisheries                         Stewart
   Studies on hook and release mortality rates as a function of location   Stewart
   of hooking on the fish and/or in the river.
Adapt Clair’s spreadsheet for harvest for Performance Measures: catch,     Tom/Eric (catch, HR, stock
   stock composition, harvest rate, and age composition.                   comp)
                                                                           Jeff (age)
Include sample size calculations for different objectives.                 Annette
Describe the pros and cons of the different monitoring programs in         Everyone at workshop,
     terms of spatial and temporal sampling variability and analytical     including sample size
     measurement techniques.                                               considerations.




REFERENCES



Fryer, J.K. 2005. Identification of Columbia basin sockeye salmon stocks in 2004. Columbia
River Inter-Tribal Fish Commission Technical Report 05-02. Portland, Oregon.

Miranda, D.P., J. Whiteaker, and J.K. Fryer. 2005. Age and length composition of Columbia
Basin Chinook and sockeye salmon and steelhead at Bonneville dam in 2004. Columbia River
Inter-Tribal Fish Commission Technical Report. Portland, Oregon.

Schwartzberg, M. and J.K. Fryer. 1993. Identification of hatchery and naturally spawning
stocks of Columbia Basin spring Chinook salmon by scale pattern analyses. North American
Journal of Fisheries Management 13:263-271.




                                                   Page 14 of 30
        Working DRAFT – do not cite or distribute 12/13/2004
Figure 1. Basic life cycle model for status and trend monitoring by life stage for Spring/Summer
Chinook salmon in the Snake River Basin (courtesy CSMEP Status and Trends workgroup).




                                         Page 15 of 30
       Working DRAFT – do not cite or distribute 12/13/2004
                                        Design &
                                       Implement
                                         M&E

    Revised
     M&E                                                                    Revise
                                                                              d
                                          M&E                               M&E
                                          Results                                                                 Y
          Y                              Has Goal
       Δ               Δ                # Been #                             M&                                   M&
                                         Reached S
N     M&       N      Mg        Y       S
                                          Sm
                                                         m         N          E               N                   E Δ?
                                              olt
      E?              mt                p
                                               /        S
                                                         olt                 Adeq
                     Y                      G
                                              Sp                             uate                                 N
                       ?                    D           A                       Y
                                                 S      R
                                                                         Implement
                      New                        D      s
                                                                           mgmt
                     Mgmt                                                 action
                     Action                                             plus M&E
                                                                                                   Revised         New
                                              Harvest          Hydro    Hatcheries       Habitat
                                                                                                    M&E           Mgmt
                                                                                                                  Action




                                                 Tier 3        Tier 3      Tier 3        Tier 3
                                                                                                       Y
                                                 M&E           M&E         M&E           M&E
                                                                                                    M&E       N
                                                                                                     Δ?



                                                                                                        N

                                                                              Mgmt
                                                                              Goal        N         M&E
                                                                                                   Adequate   Y
                                                                              Met?                    ?




                                                                                     Y



Figure 2. An example of a rule for making decisions about recovery. (Source:IDFG).




                                              Page 16 of 30
       Working DRAFT – do not cite or distribute 12/13/2004
                       Revised
                        M&E

                                                                          Y


                        M&E
                       Adequate                                     M&E Δ?
                                       N


                            Y                                             N

                 Implement mgmt
                 action plus M&E


                                                                         Y          N

                                                      Revised        New
  Harvest     Hydro      Hatcheries   Habitat                                  N
                                                       M&E           Mgmt          Revised
                                                                    Action ?       Criteria ?

                                                          Y
   Tier 3     Tier 3       Tier 3     Tier 3                                        Y
   M&E        M&E          M&E        M&E
                                                                N
                                                     M&E Δ ?


                                                           N


             Y      Mgmt              N               M&E
                                                                Y
                   Goal Met?                        Adequate?




Figure 3. (Revised Tier 3 Figure from S. Sharr, Feb 24, 2005)




                                                Page 17 of 30
        Working DRAFT – do not cite or distribute 12/13/2004
              Staff Report




                Compact
                Hearing



                  Designees
                  approve                N        Terminate
                  fishery                o        Fishery


                             Ye
                             s


             Action Notice          Rule Filing




                 Fishery
                 Monitoring


                 Fact Sheet




Figure 4. Decision process specific to the lower Columbia River mainstem fisheries.




                                             Page 18 of 30
       Working DRAFT – do not cite or distribute 12/13/2004
Figure 5. Map showing mainstem and tributary fisheries affecting the Snake River spring and summer Chinook
ESU.
                                           Page 19 of 30
          Working DRAFT – do not cite or distribute 12/13/2004
Table 1. Summary of inputs to the decision by fishery.
Fishery          Typical        Main Data                 How                 Alternative   Analytical                  Evaluation
                 Fishery        Sources and               Addressed in        Options       Methods                     Method
                 Timing         Uncertainties             Decisions
     TARGET
OR/WA       Jan - May           1) Effort counts via      Fisheries                         TARGET:                     Compare sample based
Columbia                        aerial flights, creel     managed                           1) Statistical estimation   point-estimate against
                                surveys, boat counts-     conservatively to                 by week for catch.          Action Level
River                           count accuracy,           ensure that                       2) Apply CWT
Mainstem                        reflective of non-count   impacts do not                    recovery for stock
Sport                           days,                     exceed                            composition
Selective.                      2) Effort expansion       sport/commercial                  (upriver/Willamette
                                algorithms include        allocation                        split).
                                environmental factors     agreements and
                                and historic effort       NOAA Fisheries                    BYCATCH:
                                profile                   guidelines.                       1) Release rates
                                3) Catch rates via        Fisheries closed                  applied to total effort
                                angler interviews-        prior to reaching                 estimates on a monthly
                                sampling error and        impacts. Final                    basis to estimate total
                                bias                      impact levels                     released stratified by
                                4) CWT recovery-          typically less                    weekday and weekend
                                sampling error(in-        than upper limit                  day.
                                season - used for         set forth by                      2) Combine monthly
                                Willamette/upriver        NOAA Fisheries.                   catch estimates for
                                stock separation only,    May not utilize                   weekdays and weekend
                                , ability to maintain     impacts fully due                 days to estimate total
                                random sample             to conservative                   monthly released catch
                                5) CWT expansion          management                        and apply a release
                                factors-accuracy,         strategy.                         mortality rate.
                                expanded for sampling
                                and mark rates                                              IMPACT RATE:
                                6) VSI stock                                                1) Compare total catch
                                composition estimates,                                      by stock to total return
                                error rate of VSI                                           by stock to estimate
                                estimates.                                                  impact rate.
                                7) Supplied hook and
                                release mortality,
                                uncertainty over

                                                                 Page 20 of 30
                           Working DRAFT – do not cite or distribute 12/13/2004
Fishery        Typical         Main Data                 How              Alternative   Analytical                  Evaluation
               Fishery         Sources and               Addressed in     Options       Methods                     Method
               Timing          Uncertainties             Decisions
                               release mortality
                               estimates.
OR/WA          Mid Feb - May   1) Average weight         1) Same as for                 1) Statistical estimation
Columbia                       data-sampling errors,     WA CR Sport                    of weight conversion
                               sample at processing      Non-Selective                  to determine total catch
River                          plants, ability to        Fishery                        by in numbers on a
Mainstem                       maintain random                                          weekly basis.
Non-Treaty                     sample                                                   2) Weekly catch
Commercial                     2) Fish ticket                                           estimates combined to
Selective                      landings-minimal                                         develop season total
                               error, accuracy in                                       catch estimates
                               filling out fish ticket                                  3) Apply CWT
                               3) CWT recovery-                                         recovery for stock
                               sampling error, ,                                        composition
                               ability to maintain                                      (upriver/Willamette
                               random sample                                            Split).
                               4) CWT expansion                                         4) Chinook marked to
                               factors-accuracy,                                        unmarked ratio applied
                               expanded for sampling                                    to weekly catch
                               and mark rates                                           estimates to determine
                               5) Chinook marked to                                     total weekly chinook
                               unmarked ratio-                                          catch
                               sampling errors,
                               possible bias, on-
                               board monitoring
                               program, ability to
                               maintain random
                               sample, sample size
ID Steelhead                   Creel w/ CWT                                                                         ± 10% , 90% of time
Recreational                   recovery + phone
                               surveys twice annually
                               Release mortality rate
                               assumed to be 5% of
                               handle.
ID Chinook                     Creel w/ CWT                                                                         ± 10% , 90% of time
Recreational                   recovery + phone                                                                     for ceeel and ±1% for

                                                                Page 21 of 30
                         Working DRAFT – do not cite or distribute 12/13/2004
Fishery         Typical       Main Data                  How                 Alternative                Analytical                  Evaluation
                Fishery       Sources and                Addressed in        Options                    Methods                     Method
                Timing        Uncertainties              Decisions
Fisheries                     surveys twice annually                                                                                mandatory check
Idaho                         Roving Creel and                                                                                      stations
                              Mandatory check
                              stations
                              Encounter rate
                              estimate for listed fish
                              likely biased high
                              because of missed
                              clips among hatchery
                              releases.
                              Release mortality rate
                              assumed to be 10% of
                              handle.
Tribal Treaty                 1. Effort counts via       Fisheries           Use PIT tag data of        1) Statistical estimation   Compare sample based
Fisheries                     aerial flights for         managed             fish tagged as juveniles   by week for catch.          point-estimate against
                              commercial fisheries-      conservatively to   and adults tagged at       2) Apply CWT                Action Level
                              count accuracy,            ensure that         Bonneville Dam to          recovery for stock
                              reflective of non-count    impacts do not      determine timing of        composition (not used
                              days.                      exceed NOAA         different stocks and       for in-season
                              2. Total catch             Fisheries           their vulnerability to     management).
                              estimation from catch      guidelines.         fisheries. This would
                              monitoring for             Fisheries closed    require recovery of PIT
                              platform, ceremonial       prior to reaching   tag data from harvested
                              permit and or              impacts. Final      fish and possibly more
                              commercial fisheries,      impact levels       extensive PIT tag
                              potential bias or          typically less      monitoring in
                              uncertainty from           than upper limit    tributaries.
                              sampling protocol or       set forth by
                              sample rates.              NOAA Fisheries.
                              3.                         May not utilize
                                                         impacts fully due
                                                         to conservative
                                                         management
                                                         strategy.



                                                                Page 22 of 30
                          Working DRAFT – do not cite or distribute 12/13/2004
Appendix A – Use of Harvest Data in Mainstem Run Reconstruction

This is one of the work products that was prepared for Northwest Power and Conservation Council to
help validate the Ecosystem Diagnosis and Treatment tool. It was presented to Council staff on a CD
September 2002. Henry Yuen (USFWS) was the lead author, working with Earl Weber, Charlie
Petrosky, and Eric Tinus. TAC does not use EDT for any aspect of fishery management.

                                    Methods for mainstem run reconstruction

In order to calculate interdam survival or conversion rates and harvest rates, the entire spring chinook run
must be reconstructed. The methods used in the run reconstruction are similar to those used in PATH
(Marmorek et al 1998) by the US. v. Oregon Technical Advisory Committee (TAC) except TAC’s run
reconstruction is for adults only whereas PATH and EDT’s product are adults and jacks combined.
TAC’s reconstruction is for the entire Snake River spring/summer chinook population, both the listed
ESU and the Clearwater. TAC also now reconstructs Snake River spring/summer chinook separately
from Upper Columbia River summer chinook. (note: TAC may end up using PIT tag data to estimate
conversion rates in the future). Considerable effort was spent on reconciling differences between this and
TAC’s databases. Any outstanding differences are noted below. There are three basic components in the
run reconstruction: dam counts, harvest estimates, and tributary escapements (or turn offs). We will
describe the data sources and calculations starting the mouth of the Columbia River and work upstream
(which will correspond to reading the tables from left to right).

Harvest below Bonneville Dam. - Begin the run reconstruction at the mouth of the Columbia River by
estimating the harvest in the non-Indian commercial, sport, and miscellaneous1 fisheries below Bonneville
dam.

The sources for spring chinook were

       Total harvest estimates were obtained from Table 27 in ODFW and WDFW 2000. Note (1) the
        numbers in this source are rounded to the nearest hundred, (2) the 1949-59 total sport harvest
        were calculated from the average 1960-74 and 1977 sport fishery exploitation rate, and (3) the
        1960-68 total sport catch were the sum of the adult and jack sport harvest. See the note below on
        how the 1960-68 jack sport harvest was calculated.
     Adult harvest estimates were obtained from Table 1 in TAC 2002 and from Table 28 in ODFW
        and WDFW 2000. Note after 1977, TAC’s numbers are not rounded to the nearest hundred.
Jack harvest = total harvest – adult harvest, except for following: (1) assume 0 jacks harvested in the
commercial fishery between 1981-2001 which were conducted early in the run and did not intercept jacks
(C. Melcher, ODFW, personal communication); (2) the number of jacks in the sport fishery between
1960 and 1968 were estimated as average 1969-78 jack sport harvest rate X Bonneville jack count, and
(3) assume 0 jacks in the miscellaneous fisheries which were conducted early in the run and did not
intercept jacks (C. Melcher, ODFW, personal communication).
The sources for summer chinook were (Note: TAC now manages Snake River summer chinook with
upriver spring chinook):

     




1
  Select Area fisheries, test fisheries, and tangle net experiments plus incidental non-retention mortality in commercial shad and
sockeye fisheries.



                                                                 23
Dam counts. - Proceed upriver to the Bonneville dam counts. The source for the adult and jack dam
counts were pages 247-248 and 251-253 from ODFW and WDFW 2000 including the 1974, 1976, and
1977 counts which were corrected for fallback. Note (1) the adult dam counts in TAC 2002 were not
corrected for fallback and (2) there were no jack counts prior to 1960.

The sum of the harvest below Bonneville dam and the Bonneville dam counts is the total return of upriver
spring chinook and Snake River spring/summer chinook at the mouth of the Columbia River. Upriver
spring chinook do not turn off into tributaries until they pass above Bonneville dam.

Harvest between Bonneville and McNary Dam. - There are both fisheries and tributary escapements
(turnoffs) between Bonneville and McNary dams. The data sources for the spring chinook fishery were:

       Total harvest estimates were obtained from Table 27 in ODFW and WDFW 2000. Note the
        numbers in this source are rounded to the nearest hundred.
       Adult harvest estimates were obtained from Table 1 in TAC 2002 and from Table 28 in ODFW
        and WDFW 2000. Note (1) after 1978, TAC’s numbers are not rounded to the nearest hundred
        and (2) there were no C&S data prior to 1977.
       Jack harvest = total harvest – adult harvest. Note (1) the tribes do not record jacks in the C&S
        fishery harvest and (2) there not enough jacks in the commercial catch for the total to round up to
        the next hundred.


Tributary escapements. – Snake River Summer chinook do not turn off the mainstem Columbia River
below McNary dam. On the other hand, there are numerous tributaries with spring chinook escapements
between Bonneville and McNary dams. The data sources for the tributary escapements are described
below.
     Wind River: The 1960-1965 total run was obtained from Beamesderfer et al 1977. The 1966-
       1969 data was calculated from Zone 6 escapement (Bonneville Dam counts – harvest) X 1970-77
       average % of Zone 6 escapement that is Wind R escapement. The 1970 to present data were
       obtained from Table 1 in Petit 2002.
     Little White Salmon River: The 1970 to present data were obtained from Table 10 in Petit 2002.
     Big White Salmon River: The 1986 to present data were obtained from Table 17 (escapement)
       and Table 18 (sport harvest) in Petit 2002.
     Hood River: The 1963-1971 total run was calculated from the sum of the Powerdale Dam
       passage counts (ODFW and CTWSRO 1990) and the sport harvest of adult salmon (unpublished
       sport harvested estimates from ODFW). The 1989-1991 adult data was calculated by multiplying
       the average 1992-94 ratio of sport catch below Powerdale Dam to Powerdale Dam with the 1989-
       91 sport catch. The 1992-1998 jack and adult data was calculated from the sum of Powerdale
       dam counts and sport harvest estimates. The 1999-2001 data is preliminary.
     Klickitat River: The 1977 to present data obtained from Table 20 in Petit 2002. Note jacks = age
       3 only. Do not include age 2 in the totals.
     Deschutes River: The 1954-68 total run was calculated from the average of the Deschutes River
       run (Beamesderfer et al. 1997), divided by Bonneville Dam count (less Zone 6 harvest) for years
       1969-1977 (i.e. 0.023), multiplied by each year's Bonneville Dam count for 1954-1968. The
       1969-76 total run was calculated from the run to the Warm Springs River (Beamesderfer et al.
       1997), fish entering Pelton Trap (Lindsay et al. 1989), and Deschutes sport and tribal inriver
       harvest estimates (Lindsay et al. 1989). The jack and adult estimates between 1977-85 were
       obtained from Lindsay et al. 1989 and between 1986-2000 from ODFW 2001. The 2001 data
       was assembled from preliminary hatchery returns, Pelton trap counts, and harvest data.
     John Day River: The 1954-58 total run was calculated from the average of 1959-67 John Day


                                                    24
            River run, divided by Bonneville Dam count (less Zone 6 harvest) for years 1959-1967 (i.e.
            0.026), multiplied by each year's Bonneville Dam. The adult and jack estimates between 1959
            and 2001 was calculated from index redd counts in three key spawning areas expanded by area
            and fish per redd. The estimate includes inriver harvest and prespawning mortality.
           Umatilla River: The 1988-89 data was obtained from the count of hatchery jacks and adults
            trapped at Threemile Dam (CTUIR and ODFW, 1990a). The 1990-99 data is the same except
            harvest downstream of the dam is included (Saul et al 2001). The 2000-2001 data is preliminary.
           Walla Walla River: The 2000 data was obtained from CTUIR and ODFW 1990b. The 2001 data
            is preliminary.
       
       
           Tucannon River: The 1969-84 total return data was calculated from index redd count expansion
            (6.7), average of 2.6 fish per redd, punch card harvest, or 0.25 sport harvest rate from Gallinat et
            al. 2001, Howell et al. 1985, and Schwartzberg and Roger, 1986. The 1985 to 2001 jack and
            adult estimates were obtained from M. Galliant, WDFW.


Interdam conversion rate. – Conversion rate is defined as the interdam passage survival after accounting
for harvest and tributary turnoff2. The conversion rate between two dams would be:

                                                                          upstreamdam yr
            interdamSurvivalRate yr 
                                                  downstreamDamyr  harvest yr  tributaryEscapement yr

For example, the Bonneville to McNary conversion rate according to this report would be

                                                             MCN yr
BONtoMCNconvRate yr 
                                         BON yr  harvest yr   tributaryEscapement yr
TAC has a slightly different method (Table 1) where an optimization routine is used to find the
conversion rate that minimized the difference between the downstream dam count and the reconstruction
upstream run, i.e.

                                                     Deschutes JohnDay  Umatilla     MCN  
min  BON yr  harvest yr  BonnevillePoolTurnoff                                          
              
                                                                               2
                                                                                      convRate  
                                                   3
                                                       convRate     3
                                                                      convRate                 


Notice that TAC uses a single project rate, which is

                                                                   upstreamdam yr
interdamSurvivalRate yr             n
                                         downstreamDamyr  harvest yr  tributaryEscapement yr


where n = number of dams between the upstream and downstream dam + 1. For example, n = 3 when the
downstream dam is Bonneville and the upstream dam is McNary. Thus all of the tributary turnoffs
between Bonneville and The Dalles Dam are assume to have 100% interdam survival rates because there
are no dams between Bonneville and the tributaries. The Deschutes escapement is assigned the single
2
    Tributary turnoff includes both tributary and hatchery escapements.



                                                                  25
project conversion rate because The Dalles dam is between Bonneville and the Deschutes River.
Likewise, escapements to the John day and Umatilla Rivers have to pass The Dalles and the John Day
dams and therefore are assigned the single project rate squared twice (i.e. Bonneville-TheDalles X
TheDalles-JohnDay) Finally, all escapements past NcNary dam encounter 3 dams above Bonneville and
the conversion between Bonneville and McNary would be the single project rate cubed, i.e. Bonneville-
TheDalles X TheDalles-JohnDay X JohnDay–McNary. The two methods are not interchangeable. See
Table 2 for examples on how to use the conversion rates properly.

Frequently, an upstream dam may count more fish than the downstream dam. This occurs with spring
and summer chinook between Bonneville and McNary Dams and with summer chinook between Ice
Harbor and Lower Granite Dams. To get around this problem, TAC, PATH and this report calculate
conversion rates between four pairs of reference points. (1) Bonneville and McNary3, (2) McNary, Priest
Rapids and Ice Harbor dams4, (3) between Ice Harbor and Lower Granite dams5, and between Priest
Rapids and Wells dams6. The biggest difference among TAC, PATH, and this report is how situations
when more fish counted upstream than downstream7 are handled. PATH and this report defined a
maximum conversion rate of 1.0 whereas TAC does not for the Bonneville to McNary dam conversion
rates. For example in 1999 (ignoring the fact that TAC’s conversion rate is for adults only), this report’s
spring chinook conversion rate through McNary reservoir was capped at 1.000 whereas TAC reported a
conversion rate of 1.32618. On the other hand, TAC has a maximum conversion rate of 1.0 for summer
chinook between Ice Harbor and Lower Granite Dam similar to PATH and this report.


Harvest rate. - Calculate the historical harvest rates as catch/total return at mouth of Columbia River.


                                                          Data Sources


Beamesderfer, R. C., H. A. Schaller, M. P. Zimmerman, C. E. Petrosky, O. P. Langness, and L. LaVoy.
      1997. Spawner-recruit data for spring and summer chinook salmon populations in Idaho, Oregon,
      and Washington in D. R. Marmorek and C. Peters (eds.) Plan for analyzing and testing
      hypotheses (PATH): report of retrospective analysis for fiscal year 1997. ESSA Technologies
      Ltd, Vancouver, B.C., Canada.

Bosch, B. 2001. Run size forecast for Yakima River adult spring chinook, 2002. Yakama Nation
       Fisheries Resource Management.




3
  TAC describes this area as Zone 6 or the mid-Columbia.
4
  TAC describes this area as McNary Reservoir. This area may be defined by other agencies as part of mid-Columbia.
5
  Also referred to as the Snake River.
6
  TAC described the area above Priest Rapids Dams as the upper Columbia while other agencies may define the area between
Priest Rapids and Wells dams as the mid-Columbia and the area above Grand Coulee as the upper Columbia.
7
  The reasons for counting more fish upstream are many: the cutoff date which define a chinook as spring or summer or fall may
not be correct because of annual variation in run timing, some dams quit counting earlier in the year than others, some dams count
fewer hours in the day than others, some dams are suspected of having chinook fall back (either over the spillway or through the
boat locks) and reascending the fish ladder, some dams may have fish migrating upstream through the boat locks and bypassing
the fish ladders and counting station altogether, etc.
8
  This example was provided only to illustrate how conversion rates greater than 1.0 are handled differently by TAC. There other
differences, namely TAC’s conversion rate is for adults only and this report is for adults and jacks combined, that was ignored in this
example.



                                                                  26
CTUIR and ODFW (Confederated Tribes of the Umatilla Indian Reservation and Oregon Department of
      Fish and Wildlife). 1990a. Umatilla River subbasin Salmon and Steelhead Production Plan,
      Columbia Basin System Planning. Northwest Power Planning Council and Columbia Basin Fish
      and Wildlife Authority.

CTUIR and ODFW (Confederated Tribes of the Umatilla Indian Reservation and Oregon Department of
      Fish and Wildlife). 1990b. Walla Walla River subbasin Salmon and Steelhead Production Plan,
      Columbia Basin System Planning. Northwest Power Planning Council and Columbia Basin Fish
      and Wildlife Authority.

Fast, D.E., Hubble, J.D. and B.D. Watson. Year of publication unknown. Yakima River spring chinook -
        The decline and recovery of a Mid-Columbia natural spawning stock, page 18 in unknown
        document. Fisheries Resource Management, Yakima Indian Nation. Toppenish, Washington.


Gallinat, M.P., J. Bumgarner, L. Ross, and M. Varney. 2001. Tucannon River spring chinook salmon
        hatchery evaluation program 2000 annual report. Washington Department of Fish and Wildlife,
        Fish Program, Science Division, Olympia, WA.

Howell, P., K. Jones, D. Scarnecchia, L LaVoy, W. Kendra, and D. Ortman. 1985. Stock assessment of
       Columbia River anadromous salmonids, volume 1. Report of the Oregon Department of Fish and
       Wildlife to Bonneville Power Administration, Portland.

Lindsay, R. B., B. C. Jonasson, R. K. Schroeder, and B. C. Cates. 1989. Spring chinook salmon in the
       Deschutes River, Oregon. Report of Oregon Department of Fish and Wildlife to Bonneville
       Power Administration, Portland.

Marmorek, D.R. and C.N. Peters (eds.), J. Anderson, R. Beamesderfer, L. Botsford, J. Collie, B. Dennis,
      R. Deriso, C. Ebbesmeyer, T. Fisher, R. Hinrichsen, M. Jones, O. Langness, L. LaVoy, G.
      Mathews, C. Paulsen, C. Petrosky, S. Salia, H. Schakker, C. Toole, C. Waltres, E. Weber, P.
      Wilson, M.P. Zimmerman. 1998. Plan for Analyzing and Testing Hypothesis (PATH):
      Retrospective and prospective analyses of spring/summer chinook reviewed in FY 1997.
      Compiled and edited by ESSA Technologies Ltd., Vancouver, B.C.

Matylewich, M. Year of publication unknown. Appendix C, adult dam conversion data in Schaller et al.
       1992. Columbia River Intertribal Fish Commission. Portland, Oregon.

ODFW. 2001. Mid Columbia District annual Report. Oregon Department of Fish and Wildlife, The
     Dalles, Oregon.

ODFW and CTWSRO. 1990. Hood River Subbasin Salmon and Steelhead Production Plan. Oregon
     Department of Fish and Wildlife and the Confederated Tribes of the Warm Springs Reservation.
     September 1990. Columbia Basin system planning. Northwest Power Planning Council and
     Columbia Basin Fish and Wildlife Authority.

ODFW and WDFW 2000 Status Report. Columbia River fish runs and fisheries. 1938-1999.

Petit, R. 2002. Escapement database for spring chinook in Washington tributaries between Bonneville
         and McNary dams, 1970-2001. Washington Department of Fish and Wildlife, Region 5.
         Vancouver, Washington.



                                                   27
Schwartzberg, M., and P. B. Roger. 1986. An annotated compendium of spawning ground surveys in the
       Columbia River basin above Bonneville Dam, 1960-1984. Columbia River Inter-Tribal Fish
       Commission, Technical Report 86-1, Portland.

Saul, D., C. Rabe, A. Davidson, and D. Rollins. 2001. Draft Walla Walla River subbasin summary
        prepared for the Northwest Power Planning Council through the facilities of the Columbia Basin
        Fish and Wildlife Authority. Portland, Oregon.
This should probably use the current BA which is dated March 7, 2005.




                                                  28
Table 2. Spring chinook BON to MCN conversion rates (revised in 2004 when the database combined Snake spring and summer Chinook) are
higher with the TAC method because passage losses are prorated between the mainstem migrants and the tributary turnoffs. The EDT method
assigns all of the passage losses to the mainstem migrants.

                                                                                                                                                TAC method9            EDT method10
    Year BON Count Zone 6      Wind      Little    White     Klickitat  Hood       Total     Deschutes John Day Umatilla Total Zone     MCN        Conv. Conv. per       Conv.     Conv.
                   Catch       River     White    Salmon      River     River Bonneville       River     River   River     6Tributary   Count     BON to Project        BON to       per
                                        Salmon     River                        Pool turnoff                                Turnoff                MCN                   MCN       Project
                                         River
     1979    48,600    2,090    3,467       1,009         --        851       3        5,330      1,870    2,198        --      9,398    19,100    0.5372     0.8129      0.5147    0.8014
     1980    53,100    1,855    3,427         974         --      1,685       3        6,089      1,441      968        --      8,498    17,947    0.4319     0.7559      0.4198    0.7488
     1981    62,827    3,398    2,700       2,601         --      2,528       4        7,833      1,900    1,408        --     11,141    29,533    0.6225     0.8539      0.6116    0.8488
     1982    70,011    5,308    1,714       3,070         --      3,238       4        8,026      3,870    1,395        --     13,291    24,922    0.5024     0.7950      0.4848    0.7855
     1983    54,898    2,531    2,751       2,705         --      2,417       3        7,876      3,302    1,262        --     12,440    29,874    0.7591     0.9122      0.7482    0.9078
     1984    46,866    3,475    2,287         533         --      1,323       3        4,146      2,253    1,036        --      7,435    23,718    0.6715     0.8757      0.6596    0.8705
     1985    83,182    3,112    4,872       1,502         --        848       5        7,227      5,163    1,665        --     14,055    55,957    0.8537     0.9486      0.8476    0.9464
     1986   118,082    7,433    7,516       1,098         0       1,112       7        9,733      7,937    2,793        --     20,463    73,246    0.8207     0.9362      0.8122    0.9330
     1987    98,573    6,679    5,246       5,831        51       1,682       6       12,816      5,237    5,083        --     23,136    58,787    0.8648     0.9527      0.8550    0.9491
     1988    90,532    7,004    2,685       5,732      147        3,929       5       12,498      5,262    2,884       13      20,657    48,885    0.7892     0.9241      0.7775    0.9195
     1989    81,267    6,726    3,379       3,983      117        5,254      99       12,832      7,799    2,612       66      23,309    32,502    0.6603     0.8708      0.6344    0.8593
     1990    94,158    6,928   14,674       3,950    1,310        2,583   1,835       24,352      7,747    2,199   2,158       36,456    43,555    0.8713     0.9551      0.8578    0.9502
     1991    57,339    3,869    9,791       2,537    1,069        1,477     662       15,536      4,900    1,142   1,291       22,869    20,302    0.6931     0.8850      0.6634    0.8722
     1992    88,425    5,748    8,538       4,085      383        1,540     882       15,428      6,297    3,138     460       25,323    48,629    0.8586     0.9504      0.8479    0.9465
     1993   110,820    7,255   12,739       7,098      590        3,702     756       24,885      3,337    4,014   1,205       33,441    58,734    0.8472     0.9462      0.8376    0.9426
     1994    20,169    1,125    1,572         817      370          958     589        4,306      1,060    2,362     263        7,991     8,768    0.8253     0.9380      0.7933    0.9257
     1995    10,194      619      487         710      132          696      92        2,117         996     369     388        3,870     4,612    0.8320     0.9405      0.8084    0.9316
     1996    51,493    2,791    8,090       6,973      980        1,156     131       17,330      3,225    3,124   2,152       25,831    14,373    0.6865     0.8822      0.6284    0.8566
     1997   114,071    8,278    7,209       4,202      162        1,861     384       13,818      2,821    2,466   2,194       21,299    56,832    0.6865     0.8821      0.6726    0.8762
     1998    38,342    2,189    1,770         776        94         702     101        3,443      1,006    1,647     409        6,505    19,415    0.6722     0.8760      0.6549    0.8684
     1999    38,669    1,962    5,029       4,321        55         728     121       10,254      3,332    1,554   1,764       16,904     9,260    0.5349     0.8117      0.4676    0.7762
     2000   178,302   11,256   21,522     11,888       211        2,582     188       36,391     12,950    5,930   4,644       59,915    64,647    0.6379     0.8608      0.6034    0.8450
     2001   391,367   54,562   25,610     14,309       880        1,098     651       42,548     10,214    6,100   4,760       63,622   258,689    0.9489     0.9827      0.9469    0.9820




9
  TAC solves for conversion rate with an optimizing routine. Objective function was to minimize the difference between BON count and the following: [Zone 6 catch +Bonneville pool
turnoff + Deschutes turnoff/cube root of conv rate + John Day & Umatilla turnoff / (cube root of conv. rate) squared + MCN count / conv rate].
10
   EDT calculates conversion rate as upstream dam count/(downstream dam count - harvest - tributary turnoffs)




                                                                                          29
Table 2. The TAC conversion rates are not interchangeable with the PATH and EDT conversion rates for 2 reasons: (1) the TAC rates are for
adults only while the PATH and EDT rates are for jacks and adults and (2) the starting and ending values for Bonneville dam will differ. A simple
test to match the correct method with the conversion rates is to back calculate the Bonneville dam counts as in the examples.

step 1: calculate conversion rates
                            Little                                Total                                  Total                    TAC method11    PATH, EDT method12
                            White           White                 Bonneville              John         Zone 6                      Conv. Conv.      Conv.   Conv.
     BON      Zone 6 Wind Salmon            Salmon Klickitat Hood Pool         Deschutes Day Umatilla Tributary MCN              BON to per BON to per
Year Count    Catch River River             River River River turnoff        River      River River Turnoff Count               MCN Project         MCN     Project

2001 391,367 54,562 25,610 14,309           880       1,098   651 42,548      10,214     6,100 4,760            63,622 258,689 0.9489 0.9827         0.9469    0.9820


step 2: back calculate BON dam count
                                Example 1                        Example 2             Example 3

                                TAC reconstruction method        PATH reconstruction method        TAC reconstruction method
                                with TAC conversion rates        with PATH conversion rates        with PATH conversion rates

MCN                                         258,689                          258,689                         258,689
MCN pool conversion rate                    0.9489                           0.9469                          0.9469
recontructed BON count                      272,617                          273,183                         273,183

John Day & Umatilla turnoff                 10,860                           10,860                          10,860
JDA pool conversion rate                    0.9656                           1                               0.9643
recontructed BON count                      11,246                           10,860                          11,262

Deschutes turnoff                           10,214                           10,214                          10,214
TDA pool conversion rate                    0.9827                           1                               0.9820
recontructed BON count                      10,394                           10,214                          10,401

BON pool turnoff                            42,548                           42,548                          42,548
BON pool conversion rate                    1                                1                               1
recontructed BON count                      42,548                           42,548                          42,548

Zone 6 harvest                              54562                            54562                           54562
total reconstructed BON count               391,367                          391,367                         391,956


11
   BON = Zone 6 harvest + BON pool tributary turn offs + Deschutes R/single project conv rate + (Umatilla + John Day R)/single project conv rate2 + MCN dam counts/single project
conv rate3
12
   BON = Zone 6 harvest + all tributary turn offs + MCN dam counts/single project conv rate3



                                                                                                                                                    (H\W\RME-CSMEP\Harvest DQO 6:16.doc)




                                                                                              30

								
To top