MEMO: August 12, 1996
TO: William Stelle, NMFS
FROM: Joshua Hayes, Columbia Basin Research, UW
RE: CRiSP validation against preliminary 1996 NMFS survival results
CC: Michael Schiewe (NMFS), James Anderson (UW)

We have examined the preliminary survival estimates based on 1996 PIT tag detections, outlined in the memo of 12 July from Michael Schiewe to William Stelle. We want to bring to your attention comparable predictions made using CRiSP 1.5.3. We used the same model parameters as were used in this season's CRiSP/RT in-season analysis (see Table 3 below), and used the most recent BPA flow/spill forecast (as of 7/29/96). The fish "release" used in the model was constructed by extracting all yearling chinook PIT tag detections at Lower Granite Dam in 1996 through May 31 and using those counts to provide a seasonal shape to the release. Results for 1996 are given below in Table 1; analyses for all four years are shown in Table 2. Parameters used in 1996 analysis are given in Table 3.

Using actual 1996 conditions, model estimates are in excellent agreement with NMFS survival estimates: overall system survival estimates, from Lower Granite to John Day, are within 2.5% of each other, and reach by reach fits are all within 5%. Of particular interest is the survival estimate from Lower Monumental to McNary, because this segment of the system probably produced the highest levels of gas-related mortality.

The model predicted total system survival to below Bonneville Dam, including transported fish (assuming 11% direct transport mortality), of 64.8%.

For all four years of study, the CRiSP model produces survival estimates that are in excellent agreement with those obtained in the NMFS PIT-tag studies. Last year's analysis suggested that temperature was a primary factor in determining mortality rate, and gas bubble disease was of secondary importance. Results for this year, however, suggest that dissolved gas may have been more important. Should NMFS wish to pursue an investigation of the importance of dissolved gas in the 1996 outmigration, we stand ready to assist in that analysis.