738 



Fishery Bulletin 93(4), 1995 



positively correlated for marked transported fish 

 CR=0.77, P<0.075) (Table 1). Too few unmarked 

 untransported groups were observed to examine the 

 correlation. Further, we did not use the limited data 

 because the <1.0% delayed mortality measured in 

 one of the 1973 tests is inexplicable (the NMFS an- 

 nual report 6 for the year stated that the results were 

 "somewhat surprising," considering that migrants 

 passing the dam suffered, in actuality, a 50% mor- 

 tality). Although not evident from the summary table 

 (Table 1), in the 1970s most of the descaled fish were 

 missing considerably more than 10% of their scales 

 as compared with present conditions where highly 

 descaled fish are the exception rather than the rule. 



Annual survival estimates were lowest for 1973 

 and 1977 (Fig. 2), the two years with the lowest river 

 flows and the highest levels of descaling at the up- 

 per dam. The relatively low survivals may well have 

 been greatly influenced by debris-related problems 

 at the upper dams and were compounded by the low 

 river flows in these two years. 



In 1980, the COE began removing debris from the 

 Lower Granite Dam forebay, and in 1981 a perma- 

 nent debris rake was installed and used for the first 

 time to remove debris from the trashracks (effective 

 rakes to remove debris from trashracks were also 

 built during the 1980's for other dams). In 1983, a 

 temporary boom was placed upstream from Lower 

 Granite Dam to divert new debris away from the 

 powerhouse. The temporary boom was replaced by a 

 permanent structure in 1984. Debris is now diverted 

 away from the powerhouse and removed from the 

 river, and trashracks are systematically cleaned. 

 Additionally, the bypass systems at Lower Granite 

 and Little Goose Dams have been substantially modi- 

 fied and improved. For example, pipe and orifice sizes 

 have been increased so that what little debris enters 

 the systems does not cause problems. Fish and de- 

 bris separators have been modified so that fish are 

 separated under water and they exit after separa- 

 tion via large flumes rather than small pipes. 

 Descaling and 48-hour delayed mortality in recent 

 years have been much less than those observed in 

 the 1970's (Table 1), even under "relatively" low-flow 

 conditions such as occurred in 1987. 



To verify improved migratory conditions for 

 chinook salmon smolts after debris problems had 

 been eliminated or greatly controlled in the Snake 

 River, we compared historic with recent Snake River 



survival estimates. Survival estimates in the 1970's 

 over a 2- or 3-dam stretch under moderate to high 

 river flows ranged from 33 to 50% (Raymond, 1979). 

 Over a comparable 2-and 3-dam stretch in 1993 and 

 1994, survival estimates were 77% 7 and 66% 8 , re- 

 spectively. For low-flow conditions, we estimated 

 survival of PIT-tagged (passive-integrated-transpon- 

 der-tagged) (Prentice et al., 1990) chinook salmon 

 smolts from Little Goose Dam to McNary Dam in 

 1992 by using Raymond's ( 1979) techniques for com- 

 paring populations of fish that passed both dams. 

 We used 50 and 75% collection efficiency estimates 

 for the two dams, respectively. Flows were similar in 

 1973 and 1992 (Fig. 4); however, our 1992 survival 

 estimate (which covered three dams and reservoirs, 

 but not the most upstream dam) was 81% compared 

 with 12% for two dams and reservoirs in 1973 

 (Raymond, 1979). 



Discussion and conclusions 



The argument for a flow-survival relationship for 

 juvenile salmonid migrants, based on the 1973-79 

 NMFS yearly point estimates of inriver survival (Fig. 

 2), is heavily influenced by the low survivals esti- 

 mated for 1973 and 1977 under low-flow conditions. 

 Low survival of river migrants (both above and 

 through the hydropower system) certainly occurred 

 during the 1973 and 1977 low-flow years. However, 

 the estimated low fish survivals within the hydro- 

 power system resulted more likely from fish encoun- 

 ters with debris at dams (encounters which were in- 

 creased because of low flows and exacerbated by spo- 

 radic turbine operations) than from river discharge. 

 Data collected in the past few years on PIT-tagged 

 fish that migrated under low to moderately-low flow 

 conditions, comparable to those in the 1970's, indi- 

 cated a substantially higher survival of juvenile 

 smolts. 



Under present conditions, low flows during the 

 spring migration may not lead to direct losses of mi- 

 grant fish as high as those in the 1970's within the 



6 Ebel, W. J., R. W. Krcma, and H. L. Raymond. 1973. Evaluation 

 of fish protection facilities at Little Goose Dam and review of 

 other studies relating to protection of juvenile salmonids in the 

 Columbia and Snake Rivers, 1973. Report to U.S. Army Corps 

 of Engineering, Portland, 52 p. Northwest Fish. Sci. Cent., 

 NMFS. 



7 Iwamoto, R. N„ W. D. Muir, B. P. Sandford, K. W. Mclntyre, D. A. 

 Frost, J. G. Williams, S. G. Smith, and J. R. Skalski. 

 1994. Survival estimates for the passage of juvenile chinook 

 salmon through Snake River dams and reservoirs. Report to 

 the Bonneville Power Admin., Portland, OR, 140 p. Northwest 

 Fish. Sci. Cent., NMFS, and Univ. Washington. 



B Muir, W D., S. G. Smith, R. N. Iwamoto, D. J. Kamikawa, K. W. 

 Mclntyre, E. P. Hockersmith, B. P. Sandford, P. A. Ocker, T. E. 

 Ruehle.J. G.Williams, and J. R. Skalski. 1994. Survival es- 

 timates for the passage of juvenile chinook salmon through 

 Snake River dams and reservoirs, 1994. Report to the 

 Bonneville Power Admin., Portland, OR, 174 p. Northwest 

 Fish. Sci. Cent., NMFS, and Univ. Washington. 



