This report describes studies of juvenile-salmon dam passage and apparent survival at Cougar Dam, Oregon, during two operating conditions in 2012. Cougar Dam is a 158-meter tall rock-fill dam used primarily for flood control, and passes water through a temperature control tower to either a powerhouse penstock or to a regulating outlet (RO). The temperature control tower has moveable weir gates to enable water of different elevations and temperatures to be drawn through the dam to control water temperatures downstream. A series of studies of downstream dam passage of juvenile salmonids were begun after the National Oceanic and Atmospheric Administration determined that Cougar Dam was impacting the viability of anadromous fish stocks. The primary objectives of the studies described in this report were to estimate the route-specific fish passage probabilities at the dam and to estimate the survival probabilities of fish passing through the RO. The first set of dam operating conditions, studied in November, consisted of (1) a mean reservoir elevation of 1,589 feet, (2) water entering the temperature control tower through the weir gates, (3) most water routed through the turbines during the day and through the RO during the night, and (4) mean RO gate openings of 1.2 feet during the day and 3.2 feet during the night. The second set of dam operating conditions, studied in December, consisted of (1) a mean reservoir elevation of 1,507 ft, (2) water entering the temperature control tower through the RO bypass, (3) all water passing through the RO, and (4) mean RO gate openings of 7.3 feet during the day and 7.5 feet during the night. The studies were based on juvenile Chinook salmon (Oncorhynchus tshawytscha) surgically implanted with radio transmitters and passive integrated transponder (PIT) tags. Inferences about general dam passage percentage and timing of volitional migrants were based on surface-acclimated fish released in the reservoir. Dam passage and apparent survival probabilities were estimated using the Route-Specific-Survival Model with data from surface-acclimated fish released near the water surface directly upstream of the temperature control tower (treatment group) and slightly downstream of the dam (control group). In this study, apparent survival is the joint probability of surviving and migrating through the study area during the life of the transmitters.

Two rearing groups were used to enable sufficient sample sizes for the studies. The groups differed in feed type, and for the December study only, the rearing location. Fish from each group were divided nearly equally among all combinations of release sites, release times, and surgeons. The sizes, travel times, and survivals of the two rearing groups were similar. There were statistical differences in fish lengths and travel times of the two groups, but they were small and likely were not biologically meaningful. There also was evidence of a difference in single-release estimates of survival between the rearing groups during the December study, but the differences had little effect on the relative survival estimates so the analyses of passage and survival were based on data from the rearing groups pooled.

Conditions during the December study were more conducive to passing volitionally migrating fish than conditions during the November study. The passage percentage of the fish released in the reservoir was similar between studies (about 70 percent), but the passage occurred in a median of 1.0 day during the December study and a median of 9.3 days during the November study. More than 93 percent of the dam passage of volitionally migrating fish occurred at night during each study. This finding corroborates results of previous studies at Cougar Dam and suggests that the operating conditions at night are most important to volitionally migrating fish, given the current configuration of the dam.

Most fish released near the temperature control tower passed through the RO. A total of 92.2 percent of the treatment group passed through the RO during the November study and the RO was the only route open during the December study.

The assumptions of the survival model were either met or adjusted for during each study. There was little evidence that tagger skill or premature failure of radio transmitters had an effect on survival estimates. There were statistically significant differences in travel times between treatment and control groups through several of the river reaches they had in common, but the differences were typically only a few hours, and the two groups likely experienced the same in-river conditions. There was direct evidence of bias due to detection of euthanized fish with live transmitters released as part of the study design. The bias was ameliorated by adjusting the survival estimates for the probability of detecting dead fish with live transmitters, which reduced the estimated survival probabilities by about 0.02.

The data and models indicated that the treatment effect was not fully expressed until the study reach terminating with Marshall Island Park on the Willamette River, a distance of 105.8 kilometers downstream of Cougar Dam. This was the first reach in which the 95-percent confidence interval of the estimated reach-specific relative survival overlapped 1.0, indicating similar survival of treatment and control groups. The median travel time of the treatment group from release to Marshall Island Park was 1.64 days during the November study and 1.36 days during the December study.

The survival probability of fish that passed into the RO was greater during the December study than during the November study. The relative survival probability of fish passing through the RO was 0.4594 (standard error [SE] 0.0543) during the November study and 0.7389 (SE 0.1160) during the December study. These estimates represent relative survival probabilities from release near Cougar Dam to the Marshall Island site.

The estimated survival probability of RO passage was lower than previous studies based on balloon and PIT tags, but higher than a similar study based on radio transmitters. We suggest that, apart from dam operations, the differences in survival primarily are due to the release location. We hypothesize that the balloon- and PIT-tagged fish released through a hose at a point near the RO gate opening experienced more benign conditions than the radio-tagged fish passing the RO volitionally. This hypothesis could be tested with further study. An alternative hypothesis is that some live fish remained within the study area beyond the life of their radio transmitter.

The results from these and previous studies indicate that entrainment and survival of juvenile salmonids passing Cougar Dam varies with dam operating conditions. The condition most conducive to dam passage has been the discharge and low pool elevation condition tested during December 2012. That condition included large RO gate openings and was the condition with the highest dam passage survival.