This project was initiated in 1991 in response to the Endangered Species Act (ESA) listings in the Snake River Basin of the Columbia River Basin. Primary objectives and management implications of this project include: (1)to address the need for further synthesis of historical tagging and other biological information to improve understanding and identify future research and analysis needs; (2)to assist in the development of improved monitoring capabilities, statistical methodologies and software tools to aid management in optimizing operational and fish passage strategies to maximize the protection and survival of listed threatened and endangered Snake River salmon populations and other listed and nonlisted stocks in the Columbia River Basin; (3)to design better analysis tools for evaluation programs; and (4)to provide statistical support to the Bonneville Power Administration and the Northwest fisheries community.

In 2005, the University of Washington developed a new statistical model to analyze the combined juvenile and adult detection histories of PIT-tagged salmon migrating through the Federal Columbia River Power System (FCRPS). This model, implemented by software Program ROSTER (River-Ocean Survival and Transportation Effects Routine), has been used to estimate survival and transportation effects on large temporal and spatial scales for PIT-tagged hatchery spring and summer Chinook salmon and steelhead released in the Snake River Basin from 1996 to 2003. Those results are reported here. Annual estimates of the smolt-to-adult return ratio (SAR), juvenile inriver survival from Lower Granite to Bonneville, the ocean return probability from Bonneville to Bonneville, and adult upriver survival from Bonneville to Lower Granite are reported. Annual estimates of transport-inriver (T/I) ratios and differential post-Bonneville mortality (D) are reported on both a systemwide basis, incorporating all transport dams analyzed, and a dam-specific basis. Transportation effects are estimated only for dams where at least 5,000 tagged smolts were transported from a given upstream release group. Because few tagged hatchery steelhead were transported in these years, no transportation effects are estimated for steelhead. Performance measures include age-1-ocean adult returns for steelhead, but not for Chinook salmon. Annual estimates of SAR from Lower Granite back to Lower Granite averaged 0.71% with a standard error (SE) of 0.18% for spring Chinook salmon from the Snake River Basin for tagged groups released from 1996 through 2003, omitting age-1-ocean (jack) returns. For summer Chinook salmon from the Snake River Basin, the estimates of annual SAR averaged 1.15% (SE=0.31%). Only for the release years 1999 and 2000 did the Chinook SAR approach the target value of 2%, identified by the NPCC as the minimum SAR necessary for recovery. Annual estimates of SAR for hatchery steelhead from the Snake River Basin averaged 0.45% (SE=0.11%), including age-1-ocean returns, for release years 1996 through 2003. For release years when the ocean return probability from Bonneville back to Bonneville could be estimated (i.e., 1999 through 2003), it was estimated that on average approximately 86% of the total integrated mortality for nontransported, tagged hatchery spring and summer Chinook, and 74% for steelhead, occurred during the ocean life stage (i.e., from Bonneville to Bonneville). This suggests that additional monitoring and research efforts should include the ocean and estuary environment. Annual estimates of the systemwide T/I are weighted averages of the dam-specific T/I ratios for each transport dam (with e"5,000 tagged fish transported), weighted by the probabilities of being transported at each dam. The systemwide T/I compares the observed SAR under the existing transportation system with the expected SAR if the transportation system had not been operated. Estimates of 1.0 indicate that the systemwide transportation program has no effect on SAR, while estimates> 1.0 indicate that the transportation program increases SAR. Excluding the 2001 release group, the geometric mean of the systemwide T/I estimates for hatchery spring Chinook salmon from the Snake River Basin was 1.15 (SE=0.03) for release years 1997 through 2003. The geometric mean of the systemwide T/I estimates for hatchery summer Chinook salmon from the Snake River Basin was 1.28 (SE=0.13) for release years 1997 through 2000 and 2003. Estimates were much higher for the 2001 release groups. These estimates reflect transportation from Lower Granite and/or Little Goose for most release years, depending on the number of tagged smolts actually transported at each dam during each release year. Differential post-Bonneville mortality (D) is the ratio of post-Bonneville survival to Lower Granite Dam of transported fish to that of nontransported ('inriver') fish. Excluding the 2001 release year, the geometric mean of the D estimates for hatchery spring Chinook salmon from the Snake River Basin was 1.00 (SE=0.09) for release years 1997 through 2003. For hatchery summer Chinook salmon from the Snake River Basin, the geometric mean of the D estimates was 1.32 (SE=0.27) for release years 1997 through 2000 and 2003. These estimates reflect transportation from Lower Granite and/or Little Goose, depending on the number of tagged smolts actually transported at each dam during each release year. Approximately half the point estimates of D for both spring and summer Chinook salmon were 1.0 or greater, indicating that for those release groups, transported fish did not have lower ocean and adult survival than nontransported fish. For those years with estimates of D

Program RealTime provided monitoring and forecasting of the 2007 inseason outmigrations via the internet for 26 PIT-tagged stocks of wild ESU Chinook salmon and steelhead to Lower Granite and/or McNary dams, one PIT-tagged hatchery-reared ESU of sockeye salmon to Lower Granite Dam, one PIT-tagged wild stock of sockeye salmon to McNary Dam, and 20 passage-indexed runs-at-large, five each to Rock Island, McNary, John Day, and Bonneville dams. Nineteen stocks are of wild yearling Chinook salmon which were captured, PIT-tagged, and released at sites above Lower Granite Dam in 2007 and have at least one year's historical migration data previous to the 2007 migration. These stocks originate in 19 tributaries of the Salmon, Grande Ronde and Clearwater Rivers, all tributaries to the Snake River, and are subsequently detected through tag identification and monitored at Lower Granite Dam. Seven wild PIT-tagged runs-at-large of Snake or Upper Columbia River ESU salmon and steelhead were monitored at McNary Dam. Three wild PIT-tagged runs-at-large were monitored at Lower Granite Dam, consisting of the yearling and subyearling Chinook salmon and the steelhead runs. The hatchery-reared PIT-tagged sockeye salmon stock from Redfish Lake was monitored outmigrating through Lower Granite Dam. Passage-indexed stocks (stocks monitored by FPC passage indices) included combined wild and hatchery runs-at-large of subyearling and yearling Chinook, coho, and sockeye salmon, and steelhead forecasted to Rock Island, McNary, John Day, and Bonneville dams.

The combined juvenile and adult detection histories of PIT-tagged wild salmonids migrating through the Federal Columbia River Power System (FCRPS) were analyzed using the ROSTER (River-Ocean Survival and Transportation Effects Routine) statistical release-recapture model. This model, implemented by software Program ROSTER, was used to estimate survival on large temporal and spatial scales for PIT-tagged wild spring and summer Chinook salmon and steelhead released in the Snake River Basin upstream of Lower Granite Dam from 1996 to 2004. In addition, annual results from wild salmonids were compared with results from hatchery salmonids, which were presented in a previous report in this series (Buchanan, R.A., Skalski, J.R., Lady, J.L., Westhagen, P., Griswold, J., and Smith, S. 2007, 'Survival and Transportation Effects for Migrating Snake River Hatchery Chinook Salmon and Steelhead: Historical Estimates from 1996-2003', Technical report, Bonneville Power Administration, Project 1991-051-00). These results are reported here. Annual estimates of the smolt-to-adult return ratio (SAR), juvenile inriver survival from Lower Granite to Bonneville, the ocean return probability from Bonneville to Bonneville, and adult upriver survival from Bonneville to Lower Granite are reported. Annual estimates of transport-inriver (T/I) ratios and differential post-Bonneville mortality (D) are reported on a dam-specific basis for release years with sufficient numbers of wild PIT-tagged smolts transported. Transportation effects are estimated only for dams where at least 1,000 tagged wild smolts were transported from a given upstream release group. Because few wild Chinook salmon and steelhead tagged upstream of Lower Granite Dam were transported before the 2003 release year, T/I and D were estimated only for the 2003 and 2004 release years. Performance measures include age-1-ocean adult returns for steelhead, but not for Chinook salmon. Spring and summer Chinook salmon release groups were pooled across the entire Snake River Basin upstream of Lower Granite Dam for this report. Annual estimates of SAR from Lower Granite back to Lower Granite averaged 0.92% with an estimated standard error (dSE) of 0.25% for wild spring and summer Chinook salmon for tagged groups released from 1996 through 2004, omitting age-1-ocean (jack) returns. Only for the 1999 and 2000 release years did the wild Chinook SAR approach the target value of 2%, identified by the NPCC as the minimum SAR necessary for recovery. Annual estimates of SAR for wild steelhead from the Snake River Basin averaged 0.63% (dSE = 0.15%), including age-1-ocean returns, for release years 1996 through 2004. For release years when the ocean return probability from Bonneville back to Bonneville could be estimated (i.e., 1999 through 2004), it was estimated that on average approximately 83% of the total integrated mortality for nontransported, tagged wild spring and summer Chinook, and 78% for steelhead (omitting the 2001 release year), occurred during the ocean life stage (i.e., from Bonneville to Bonneville). This suggests that additional monitoring and research efforts should include the ocean and estuary environment. Annual estimates of the dam-specific T/I for Lower Granite Dam were available for the 2003 and 2004 release years for both wild Chinook salmon and wild steelhead. The estimated T/I for Lower Granite was significantly> 1.0 for Chinook in 2004 (P 0.0001) and for steelhead in both 2003 (P

Since the 1994 outmigration, program RealTime has been applied to provide in-season predictions of smolt outmigration timing for individual and aggregates of listed threatened and endangered Snake River salmon stocks. Results from the 1997 smolt outmigrations of wild Snake River yearling and subyearling chinook show prediction of run-timing can be accurately forecasted. The number of release sites meeting previous years criteria for RealTime forecasts dropped to five for the wild spring/summer chinook parr PIT-tagged in 1996: Catherine Creek, Imnaha, Lostine, Minam and South Fork Salmon Rivers. An experiment in lessening previous RealTime requirements for forecasting a outmigration in progress added three release sites of chinook: Lake Creek, Secesh and South Fork Wenaha Rivers; and one release of age 1+ sockeye at Redfish Lake. Passage indices provided by the Fish Passage Center for Lower Granite Dam were monitored for the wild subyearling chinook outmigration. Investigation continued into basing predictions on historical years with similar flows as a way to improve forecasting performance for the wild subyearling outmigration. Program RealTime's output is a series of estimated percentages of the status of the smolt outmigration throughout the season. To compare the performance the program from year to year, or to compare various assumptions used set up the forecasting, the mean absolute deviance (MAD) of the daily predicted outmigration-proportion from the actual outmigration-proportion is calculated post-season. Furthermore, these MAD's are considered for three periods of the season: the first 50% of the season, the second 50%, and the entire season.

Using the pre-2000 reach survival probabilities reported in the 2000 FCRPS Biological Opinion (BO) for three selected stocks: yearling and sub-yearling chinook and steelhead, power curves were constructed for each of the two statistical hypothesis tests suggested in the BO. These power calculation results were interpreted in terms of the ability of the statistical tests to correctly identify the true states of recovery (i.e., fail or succeed in fulfilling RPA expectations). The proposed one-sided tests have a moderate to low probability of correctly assessing the true status of the recovery by the years 2005 and 2008. The relatively poor odds of making the correct decision with the BO proposed Tests 1 and 2 suggest alternative decision rules need to be investigated and developed for assessing RPA compliance. Therefore, we propose to immediately examine alternative decision rules that might maximize the likelihood of correct decisions while minimizing the prospect of incorrect decisions. The Bayesian analysis will incorporate scientific/biological knowledge/expertise.

Program RealTime provided tracking and forecasting of the 2001 inseason outmigration via the internet for eighteen PIT-tagged stocks of wild salmon and steelhead to Lower Granite and/or McNary dams and eleven passage-indexed stocks to Rock Island, McNary, or John Day dams. Nine of the PIT-tagged stocks tracked this year were new to the project. Thirteen ESUs of wild subyearling and yearling chinook salmon and steelhead, and one ESU of hatchery-reared sockeye salmon were tracked and forecasted to Lower Granite Dam. Eight wild ESUs of subyearling and yearling chinook salmon, sockeye salmon and steelhead were tracked to McNary Dam for the first time this year. Wild PIT-tagged ESUs tracked to Lower Granite Dam included yearling spring/summer chinook salmon release-recovery stocks (from Bear Valley Creek, Catherine Creek, Herd Creek, Imnaha River, Johnson Creek, Lostine River, Minam River, South Fork Salmon River, Secesh River, and Valley Creek), PIT-tagged wild runs-at-large of yearling chinook salmon and steelhead, and a PIT-tagged stock of subyearling fall chinook salmon. The stock of hatchery-reared PIT-tagged summer-run sockeye salmon smolts outmigrating to Lower Granite Dam, consisted this year of a new stock of fish from Alturas Lake Creek, Redfish Lake Creek Trap and Sawtooth Trap. The passage-indexed stocks, counted using FPC passage indices, included combined wild- and hatchery-reared runs-at-large of subyearling and yearling chinook, coho, and sockeye salmon, and steelhead migrating to Rock Island and McNary dams, and, new this year, combined wild and hatchery subyearling chinook salmon to John Day Dam. Unusual run-timing and fish passage characteristics were observed in this low-flow, negligible-spill migration year. The period for the middle 80% of fish passage (i.e., progress from the 10th to the 90th percentiles) was unusually short for nine out of ten PIT-tagged yearling spring/summer chinook salmon stocks tracked to Lower Granite Dam. It was the shortest on record for seven of these ten stocks. The nine stocks recording unusually short middle 80% periods also recorded higher-than-average recovery percentages. However the opposite trend was observed for the PIT-tagged wild subyearling chinook salmon and hatchery sockeye salmon stocks whose middle 80% period of passage to Lower Granite Dam was average to above average. Recovery percentages for these two stocks were average, compared to historical recoveries. The performance results of Program RealTime to make accurate predictions of percentiles of fish passage at an index site were mixed this year. The release-recovery stocks of wild PIT-tagged spring/summer chinook salmon tracked to Lower Granite Dam were predicted less accurately than usual, on average, with two exceptions. One of these exceptions was a stock that had its best prediction (first-half, last-half, and season-wide) ever to occur. On average, however, performance was down for predicting these stocks. The RealTime Select composite season-wide MAD was 4.3%, larger than the historical average of 2.1%. Passage percentiles for PIT-tagged runs-at-large of wild Snake River yearling and subyearling chinook salmon and of wild steelhead outmigrating to Lower Granite Dam were predicted very well this year, their second year of inclusion in the project, with season-wide MADs of 3.6%, 4.7%, and 1.8% respectively. These results, too, were mixed with respect to comparison with last year's performance. The yearling chinook stock was predicted somewhat better last year (up from 1.7% last year to 3.6% this year) but the subyearling chinook salmon and steelhead stocks were predicted better this year than last, season-wide. The steelhead stock, in particular, was predicted much better this year than last year, down to 1.8% this year from 4.8% last year. The PIT-tagged runs-at-large of wild salmon and steelhead tracked to McNary Dam in 2001 for the first time, were also well-predicted. In particular, the Snake River stocks were well-predicted, with season-wide MADs of 4.7% for subyearling chinook salmon, 3.3% for yearling chinook salmon, and 1.4% for steelhead. All three Snake River stocks were better predicted at McNary Dam than they were at Lower Granite Dam. The Upper Columbia River PIT-tagged runs-at-large of wild subyearling chinook salmon and wild steelhead were not predicted with the remarkable accuracy of the Snake River stocks, but RealTime performance for these stocks was still good, with season-wide MADs of 7.9% and 4.9%, respectively. The results of RealTime predictions of FPC passage-indexed percentiles of combined wild and hatchery-reared salmonids to Rock Island and McNary dams were comparable to last year with respect to the large variability in performance. Like last year some runs were predicted very well while others were predicted very poorly. The stocks predicted best and worst last year were not necessarily the stocks predicted best and worst this year.