Survival predictions of steelhead in the Snake River were evaluated against survival estimated in the 1994 PIT tag study (Muir et al. 1995) (Fig. 78). The results of the validation are given in Table 77. Travel time parameters were set according to the calibration given in Table 11. Travel time was not adjusted to specifically fit the steelhead PIT tag data. In all reaches of the river the CRiSP predicted estimates of survival were lower than the PIT tag based survival estimates.
Sims and Ossiander steelhead study
Sims and Ossiander's (1981) compilation of yearly survival studies from the 1970's was used to compared CRiSP-estimated juvenile steelhead survivals (Table 78). In general, observed survivals vary considerably from year to year, but model estimates were less variable. Modeled and observed survivals agreed, although results for 1977 appear to be unusual. Note that the two low-flow years, 1973 and 1977, produce extremely high travel times which are not matched by the model, although CRiSP provides an excellent fit to travel time observations in other years. These characteristics were also observed in the spring chinook. Note that in 1973 and 1974 reports, actual numbers of fish collected are reported, but in 1975-1977, only estimated numbers of fish passing dams are reported.
Mid-Columbia steelhead survival
Steelhead survivals estimates in the mid-Columbia were available from mark-recapture experiments conducted in the 1980s. Steelhead smolts released in the Methow River and below Priest Rapids Dam were recaptured at McNary Dam (Fig. 79). This gave survival estimates from the Methow River to below Priest Rapids Dam. The data are given in McConnaha and Basham (1985) and Fish Passage Center (1986, 1987). CRiSP as calibrated provided a good fit to observed travel times, but underestimated survival by 30 to 40%. It is worth noting, however, that survival estimates depend critically on the assumption of "equal risk" for test and control releases; in the same study, similar test/control experiments were performed for steelhead in the lower Snake River, and recoveries were such that survival for the test group was estimated to be over 100% for two years running. Obviously, survival estimates using this protocol were variable. In 1986, fish were also collected at John Day Dam, providing a second estimate of survival (39%), and this estimate differs considerably from that obtained at McNary Dam (73%). Considering that the CRiSP-derived estimate falls between these two estimates (43.4%), we believe that while steelhead survival demands further investigation. In 1985 and 1986, observations were made at Rock Island Dam as well; Table 79 gives CRiSP fits to those data also.
| Fig. 79 Mid-Columbia release and recovery sites for steelhead survival studies |
|---|
 |