EBEL: TRANSPORTATION OF CHINOOK SALMON AND STEELHEAD SMOLTS 



ily on river conditions. During years when survi- 

 val of natural migrants (hatchery and wild stocks 

 migrating naturally) was low, there was corres- 

 pondingly low survival of control releases and 

 greatest benefit from transportation. For exam- 

 ple, in 1973, natural migrant survival estimates 

 (Raymond 1979, see footnote 2) indicated an all- 

 time low survival rate for both juvenile chinook 

 salmon and steelhead migrants; in contrast, 

 transport-control ratios were highest — 15.4:1 for 

 chinook salmon and 13.4:1 for steelhead (Figure 

 2). Raymond (1979) compared survival estimates 

 of control releases from this study and other ear- 

 lier studies (Ebel et al. 1973; Slatick et al. 1975) 

 with survival estimates of naturally migrating 

 smolts and found a high correlation (r = 0.95 for 

 chinook salmon; 0.92 for steelhead) between sur- 

 vival of controls from transportation studies and 

 his estimates of survival of natural migrants for 

 the years 1968 to 1975. These data indicated there 

 were close relationships between survival of re- 

 leases of control fish marked for the transportation 

 studies and hatchery and wild stocks migrating 

 naturally. Raymond also correlated low survival 

 with adverse river conditions; thus benefits from 

 transportation would be highest when river condi- 

 tions are the most adverse. 



Statistical analysis of the return percentages 

 (Table 4) was done by analysis of variance. A test 



of normality (Shapiro and Wilk 1965) of the per- 

 centage return data showed that the data were 

 normally distributed (P 0.05); thus, transforma- 

 tion of percentage figures was not necessary. 

 Analysis of variance of the return percentages in- 

 dicated that the differences between "treatments" 

 (test and control releases) were significant at the 

 IVc level (Table 5). Interactions of the treatment x 

 species were significant at the d^c level, indicating 

 that the effects of treatment varied between 

 chinook salmon and steelhead. For example, the 

 mean transport/control ratio for returning 

 chinook salmon in 1972 was 1.1:1, whereas the 

 mean ratio for steelhead was 3.25: 1 . An analysis of 

 the test of treatment effects — to compare the two 

 downstream releases (both transported) and the 

 control vs. the transported groups (Table 5) — 

 clearly showed that there were no differences be- 

 tween recoveries from the Dalton Point and Bon- 

 neville Dam release sites and that the differences 

 shown between test and control groups (Figure 2) 

 were highly significant (P<0.01). Since interac- 

 tions of the treatment x species were significant 

 (P<0.05), I also analyzed the chinook salmon and 

 steelhead percentages separately (Table 6). These 

 analyses confirmed that differences shown be- 

 tween test and control groups were significant 

 (P<0.05) for both steelhead and chinook salmon 

 and that there were no differences between re- 



TabLE 4. — Releases and recaptures of experimental fish. 



'Age designation follows the formulas of Koo (1962). The number of years at sea is shown by an Arabic numeral preceded by a dot. 

 ^Return percentage adjusted according to tag detector and trap efficiency. 



^Transport,' control ratios determined by dividing estimated percentage return of controls into estimated return of transported fish. 

 •• Adult returns of spring and summer chinook salmon combined. 



497 



