NOTE Giorgi et al.: Migratory characteristics of juvenile Oncorhynchus tshawytscha 877 



temperatures ranged from 13 to 16°C, then increased 

 steadily during the summer and peaked near 21°C 

 by the end of August. Overall, 1982 was character- 

 ized by slightly higher water temperatures than the 

 other two years. 



Fish travel time from McNary Dam to John 

 Day Dam 



For the three study years, a total of 49 freeze-branded 

 groups were released to estimate fish travel time 

 through the reservoir. The number of fish released 

 in each group ranged from 1,007 to 12,310 (Tables 

 1-3). The estimated median travel time through John 

 Day Reservoir for freeze-branded groups ranged from 

 6 to 26 days in 1981, 9 to 46 days in 1982, and 7 to 29 

 days in 1983 (Tables 1-3). Overall, the estimated 

 median travel times were longest in 1982. 



All individual groups exhibited protracted passage 

 distributions at John Day Dam. The elapsed time 

 between the 10th and 90th percentile of the recap- 

 ture distributions typically exceeded several weeks 

 (Tables 1-3). The fastest moving fish, those repre- 

 sented by the 10th percentile, traversed the reser- 

 voir in 2 to 16 days. The slowest moving fish, those 

 represented by the 90th percentile, took 17 to 126 

 days to migrate through the reservoir. 



The linear regression analyses, treating each year 

 separately and pooling all years, did not identify a 

 single model that was applicable to all years. Trans- 

 formation of predictor variables did not improve the 

 model. In fact, for each year, different sets of vari- 

 ables were included in the model constructed by the 

 stepwise procedure. In 1981, the variability in travel 

 time could not be explained by any predictor 

 (Table 4), and none of the predictor variables entered 

 the model. In 1982, only one predictor, release date, 

 was entered into the model. In 1983, two variables, 

 release date and water temperature, were entered 



Table 4 



Regression models derived from stepwise multiple 

 regression. The modelling procedure was applied to 

 median zero-age chinook salmon travel times pre- 

 sented in Tables 1-3. Average water temperature, 

 inverse average flow, and Julian release date were 

 used in the model selection process. 



Year 



Model 



R 2 



1981 No variables were entered into the model 0.00 



1982 Travel time = -53.02 + 0.37 ( release date ) 0.47 



1983 Travel time = -1.16 + 5.20 (temperature) 



-0.34 (release date) 0.46 

 Combined Travel time = -22.83 + 2.36 



(temperature) 0.24 



into the model. For the three years combined, only 

 water temperature entered into the model. In all years, 

 strong correlations were observed among the three pre- 

 dictor variables, with r-values ranging from 0.64 to 0.98. 



Intrareservoir movement 



Upstream movement offish after branding was regu- 

 larly observed in the reservoir (Table 5). Detailed 

 recapture histories for individual fish were reported 

 in Sims and Miller, 8 and Miller and Sims. 910 In 1981, 

 1982, and 1983, the percentages of marked fish that 

 were recaptured at or upstream from the transect of 

 release were 67, 63, and 60%, respectively (Table 5). 

 In each year, upstream movement was observed more 

 frequently than stationary or downstream move- 

 ment. Upstream movements were often pronounced, 

 ranging from 9 to 82 km. Over the three years of 

 study, the duration of the observed upstream move- 

 ments ranged from 6 to 104 days. These observations 

 indicated that the population at large was not con- 

 sistently displaced downstream: rather, a large seg- 

 ment was engaged in pronounced upstream move- 

 ment, or was stationary for extended periods. 



8 Sims, C. W., and D. R. Miller. 1982. Effects of flow on the mi- 

 gratory behavior and survival of juvenile fall and summer 

 chinook salmon in John Day Reservoir. Northwest Fish. Sci. 

 Cent., Natl. Mar. Fish. Serv., Seattle, WA 98112-2097. Report 

 to Bonneville Power Administration, 22 p. 



9 Miller, D. R., and C. W. Sims. 1983. Effects of flow on the mi- 

 gratory behavior and survival of juvenile fall and summer 

 chinook salmon in John Day Reservoir. Northwest Fish. Sci. 

 Cent., Natl. Mar. Fish. Serv, Seattle, WA 98112-2097. Report 

 to Bonneville Power Administration, 25 p. 



10 Miller, D. R., and C. W. Sims. 1984. Effects of flow on the mi- 

 gratory behavior and survival of juvenile fall and summer 

 chinook salmon in John Day Reservoir. Northwest Fish. Sci. 

 Cent., Natl. Mar. Fish. Serv, Seattle, WA 98112-2097. Report 

 to Bonneville Power Administration, 23 p. 



