BPA Fish and Wildlife FY 1997 Proposal

Section 1. Administrative
Section 2. Narrative
Section 3. Budget

see CBFWA and BPA funding recommendations

Section 1. Administrative

Title of project
Run Timing Predictions for the Columbia River Basin Including Individual ESA Demes

BPA project number   9105100

Business name of agency, institution or organization requesting funding
Univ/WA

Sponsor type   WA-Consultant

Proposal contact person or principal investigator

BPA technical contact   Pat Poe, EWI 503/230-4043

Biological opinion ID   NMFS BO RPA Sec. 13f

NWPPC Program number   5.0F.5

Short description
Project focuses on developing statistical methods to predict the timing of the smolt outmigration at Columbia and Snake River Dams in real-time and conveying results interactively to the fisheries community via the World Wide Web. PROJECT START YEAR: 1991

Project start year   1991    End year   

Start of operation and/or maintenance   1992

Project development phase   Implementation

Section 2. Narrative

Related projects
Related projects include: Fish Passage Center (9403300), Smolt Monitoring by Non-Federal Entities (8712700); Smolt Monitioring by Federal Entities (8401400), PIT Tagging Wild Chinook (9102800), Columbia Basin PIT-Tag Information System (9008000), Networked-Based WWW Information Infrastructure (9207104), In-season Operations Technical Management Team (TMT) Support (9601600), and Second-Tier Database Support for Technical Management Team (TMT) (9601900).
This BPA project focuses on developing statistical methods and software tools for the systematic prediction of salmonid smolt outmigration status and trends, as well as forecasts of the cumulative passage of sensitive ESA stocks and other populations. The Smolt Monitoring Program projects, the Fish Passage Center and the PIT-Tag Information System (PTAGIS) primary-tier database centers, and the second-tier database support center DART project 9601900 provide the data used to make the real-time in-season predictions. The predictions from these tools are provided interactively to the fisheries community, including NMFS TMT via the World Wide Web to assist management of flow and spill augmentation to maximize benefits to smolt outmigration. The statistical tools are made available to all parties wishing to enhance monitoring and evaluation capabilities of their agency and of the fisheries community.

Project history
Project was initiated in response to ESA listings and the need for synthesis of biological information and development of new methodologies for application in river management to improve the protection of T&E stocks. Focuses on analysis and synthesis of existing mainstem passage data collected by the Smolt Monitoring Program and other mainstem research on wild and hatchery-reared salmonid smolts, with emphasis on wild threatened or endangered stocks and development of new approaches to assess their status. Gained knowledge and understanding will be applied to improve the effectiveness of hydrosystem operations to improve the protection of these sensitive populations through the FCRPS.
This project has evaluated the information content of historical tagging studies on Snake River salmonids. Information is being used to help design and direct future research efforts and determine reliability of existing information. Historical data has and continues to be explored to reveal possible relationships between river condition
s and salmonid survival, outmigration timing, speed, and outmigration success. Specific accomplishments by year include:
1991: Report on historical brand release data for the Snake River.
1992: Report on a "strawman" smolt monitoring design for Snake/Columbia River systems.
1993: Report on adult PIT-tag returns.
1994: Development and testing of REALTIME PIT Forecaster statistical software using pattern recognition and neuronets to predict the outmigration timing of spring runs of wild Snake River spring/summer chinook at Lower Granite Dam.
1995: Refinement of REALTIME PIT Forecaster statistical software and initial testing of REALTIME Passage Index Forecaster statistical software to predict outmigration timing of summer/fall runs of juvenile fall chinook at Lower Granite Dam. Report on transportation benefit analysis methods. Report on Priest Rapids hatchery returns versus river conditions.
1996: The project is narrowing its focus to prepare for and participate in inseason predictions of smolt outmigration timing at Snake and Columbia River Dams. This effort is expanding predictive capabilities to 4 Snake River dam sites in 1996 and to more sites in 1997, including Mid-Columbia and mainstem Columbia River Dams.

Biological results achieved
Tens of millions of coded-wire-tagged (CWT) and hundreds of thousands of passive-integrated-transponder-tagged (PIT-tagged) salmonid smolt are released into the Snake/Columbia River system annually. The data derived from these tagged fish provide needed information for postseason evaluation of outmigration success and inseason management of flow and spill. This project has provided needed analyses of smolt travel time, survival, and outmigration timing used in mainstem management and protection of endangered stocks of salmon.
Specific biological results of this project include:
1. Evaluation of travel time information contained in freeze-brand data.
2. Evaluation of the annual pattern of adult salmon returns to the upper Snake River from PIT-tag data.
3. Recommendations for complete life cycle survival studies using combined PIT/CWT releases.
4. Evaluation of the relationship between Priest Rapids hatchery adults returns and river conditions.
5. Evaluation of effects of river pulsing on smolt travel times.
6. Evaluation of the benefits of smolt transportation at Priest Rapids and McNary Dams.
Together, these biological results are helping to evaluate past management decisions and improve future efforts to enhance wild salmonid stocks. Finally, since 1994, the project has been providing run timing predictions of endangered wild salmonid stocks in the Snake River and making these predictions publicly available to the fisheries community.

Annual reports and technical papers
Technical Reports (BPA Series)
DOE/BP-16570-1, December 1990: Giorgi, A.E. 1990. Mortality of Yearling Chinook Salmon Prior to Arrival at Lower Granite Dam on the Snake River. Technical Report prepared for Bonneville Power Administration, Portland, Oregon.
DOE/BP-35885-3, October 1993: Skalski, J. R. and A.E. Giorgi. 1993. Juvenile Passage Program: A Plan for Estimating Smolt Travel Time and Survival in the Snake and Columbia Rivers. Technical Report prepared for Bonneville Power Administration.
DOE/BP-35884-4, October 1993: Smith, S.G., J.R. Skalski, and A.E. Giorgi. 1993. Statistical Evaluation of Travel Time Estimation Based on Data from Freeze-Branded Chinook Salmon on the Snake River, 1982-1990. Technical Report prepared for Bonneville Power Administration.
DOE/BP-35885-6, July 1995: Newman, K. 1995. Adult Salmonid PIT-Tag Returns to Columbia River's Lower Granite Dam. Technical Report prepared for Bonneville Power Administration.
DOE/BP-35885-7 (In Press March 1996): Skalski, J. R., G. Tartakovsky, S. G. Smith, P. Westhagen, and A.E. Giorgi. 1996. Pre-1994 Season Projection of Run-Timing Capabilities Using PIT-tag Databases. Technical Report prepared for Bonneville Power Administration.
DOE/BP-35885-8 (In Press March 1996): Townsend, R.L., P. Westhagen, D. Yasuda, and J.R. Skalski. 1996. Evaluation of the 1994 Predictions of the Run-Timing of Wild Migrant Yearling Chinook in the Snake River Basin. Trechnical Report prepared for the Bonneville Power Administration.
Other Scientific Reports Produced for Publication:
Newman, K. 1995. Experimental Designs and Statistical Models to Estimate the Effect of Transportation on Columbia River System Salmonid Survival (draft report under peer review).
Newman, K. 1995. Survival Information Provided by Combined PIT-CWT Tagging (draft report under peer review).
Perez-Comas, J.A. and J.R. Skalski. 1995. Preliminary Assessment of the Effects of Pulsed Flows on Smolt Migratory Behavior (draft report).
Townsend, R. L., J. R. Skalski, P. Westhagen, D. Yasuda, and K. Ryding. 1995. Evaluation of Real-Time Prediction of the Run-Timing of Wild Migrant Yearling Chinook in the Snake River Basin (draft report under peer review).
Skalski, J. R., and R.L. Townsend. 1994. A Comparison of the Statistical Methods of Estimating Transportation Benefit Ratios (TBR) Based on Spring Chinook Salmon on the Columbia River, 1986-1988 (draft report under peer review).

Management implications
This BPA project is providing the statistical, computer, and telecommunications needs to analyze and disseminate information on the status and timing of spring and summer outmigrations in support of the adaptive management framework.. To this end, Program REALTIME uses passage indices and PIT-tag detections to make daily predictions of the "percent run to date" and "date to specified percentiles" of wild endangered salmonid stocks in the Snake River and posts results daily on the World Wide Web for the fisheries community. These predictions help focus spill and flow augmentation to maximally benefit smolt outmigration.

Specific measureable objectives
1996: Expand development of real-time outmigration prediction to include all Lower Snake River hydroprojects as well as facilitate incorporation of Mid-Columbia predictions. Expand internet access to cover historical run timing and flow data at all major hydroprojects.
1997: Investigate extending outmigrating timing predictions to Lower Columbia hydroprojects.
1998: Adapt REALTIME data analyses of outmigration timing to the changing needs of river managers.

Testable hypothesis
The project's analyses of historical tagging data routinely tests hypotheses and looks for correlations between tag returns and environmental and human-related effects. Specific hypotheses include:
1. Flow-travel time relationships of outmigrating smolt
2. Flow-survival relationships of outmigrating smolt
3. Travel time-river pulsing relationships of outmigrating smolt
4. Temperature, turbidity, hatchery release numbers on in-river smolt survival
5. Low-adult return rate relationships.
With the increased focus on outmigration predictions, the project is testing basic hypotheses concerning outmigration timing and dynamics, that are tested annually when actual run timing is compared to daily predictions at the end of the season.

Underlying assumptions or critical constraints
1. Smolt Monitoring Program continues to operate and collect in-season and time-series data on the migrational characteristics of Columbia and Snake River populations of salmon and steelhead.
2. FPC and PTAGIS database centers continue to provide passage index and PIT-tag detection information in-season.
3. Marking of juvenile fish continues.
4. Support for adaptive management framework continues.

Methods
Historical databases on tag detections (brands, PIT-tag), tag recoveries (CWT), and ambient river conditions (flows, temperatures, spill volumes, etc.) are analyzed using SUN workstations and a variety of statistical methods (e.g., generalized linear models, nonlinear regression, mark-recapture theory). Predictions of outmigration timing
are based on advanced mathematical/statistical techniques including fuzzy logic, generalized models, and bootstrap/jackknife procedures. Output of outmigration timing predictions posted on the World Wide Web.

Brief schedule of activities
1997: Investigate extending outmigrating timing predictions to Lower Columbia hydroprojects.
1998: Adapt REALTIME data analyses of outmigration timing to include other site locations such as Mid-Columbia stocks and dams and to continue to meet the changing needs of river managers.

Biological need
Prediction of where the wild stocks of salmonid smolt are in their outmigration facilitates management decisions on the timing and placement of water budget releases and spill to protect endangered stocks.

Critical uncertainties
Project strives to use statistical methods to unravel the uncertainties of tagging data and outmigration dynamics.

Summary of expected outcome
The project will provide daily predictions of outmigration timing at Lower Granite, Little Goose, Lower Monumental, Ice Harbor, and Wanapum Dams and associated confidence intervals during spring and summer 1996 and subsequent years. Subsequent years will proceed with predictions in the Mid-Columbia and Lower Columbia.

Dependencies/opportunities for cooperation
Project has the following opportunities for interaction:
1. Results of outmigration predictions supplied to technical management team (TMT)
2. Results of outmigration predictions supplied to National Marine Fisheries Service (NMFS).

Risks
Project presents no risk to fish stocks.

Monitoring activity
The project’s outcomes are measured against the deliverables in the contract. BPA and the UW will monitor the project closely to assure that it is providing what was contracted for. Furthermore, feedback from users, and monitoring of the TMT and REALTIME Web home pages usage will be used to monitor the use of the project’s tools. The on-line comment forms on the Web pages will also be used to provide further feedback on work products. We will discuss the value of the statistical tools produced by this project with interested regional parties and make them available to all parties wishing to enhance monitoring and evaluation capabilities of their agency and of the fisheries community. Finally, we will submit technical reports for peer review prior to publication.

Section 3. Budget

* For most projects, Authorized is the amount recommended by CBFWA and the Council. Planned is amount currently allocated. Contracted is the amount obligated to date of printout.

Funding recommendations

CBFWA funding review group   Mainstem

Recommendation    Tier 2 - fund when funds available

Recommended funding level   $146,000