ming depth (Gray and Haynes 1977). Transmit- 

 ters were individually identifiable and operated 

 on a carrier frequency of 53 MHz. Transmitter 

 range varied with depth and transmitter orienta- 

 tion to a receiving antenna. Transmitter life was 

 2-3 wk. Receivers were capable of distinguishing 

 100 discrete crystal-tuned transmitters. 



Transmitters used in spring 1976 weighed 

 about 68 g in water and were 11.5 cm long and 2.7 

 cm in diameter. Transmitters used in 1977 were 

 about one-half the weight and two-thirds the vol- 

 ume of those used initially, weighed about 34 g in 

 water, and were 7.9 cm long and 1.9 cm in diameter. 

 In spring 1976 and 1977, chinook salmon 

 were trapped, anesthetized (tricaine methanesul- 

 fonate-quinaldine), and tagged externally with 

 radio and metal-core anchor tags (experimental) 

 or metal-core anchor tags only (controls). Tagging 

 was accomplished in cooperation with the Na- 

 tional Marine Fisheries Service (NMFS) at Little 

 Goose Dam. Methods of external tag attachment 

 were reported by Gray and Haynes (1977). After 

 tagging, salmon were transported 6.4 km down- 

 stream and released at Texas Rapids (Figure 1). 



Total lengths and weights of tagged salmon 

 ranged from 66 to 100 cm and 3.4 to 11.4 kg and 

 were consistent with the sizes offish used in other 

 Columbia River studies (Monan and Liscom see 

 footnotes 2-6). Fish movements between Lower 

 Monumental and Lower Granite Dams (Figure 1) 

 were monitored day and night for the duration of 

 transmitter life. Tagged fish passing through fish 



ladders at Little Goose and Lower Granite Dams 

 were automatically diverted into fish traps by a 

 magnetometer-triggered device (Durkin et al. 

 1969), or observed and recorded at fish-counting 

 windows. This allowed comparison of travel time 

 data for control and experimental fish. Total dis- 

 tance traveled was calculated for each fish by 

 summing movements between successive loca- 

 tions. 



Results and Discussion 



Extensive timing variability among individual 

 salmon was common throughout the study. How- 

 ever, travel times of salmon carrying external 

 radio transmitters and control fish were not sig- 

 nificantly different (Gray and Haynes 1979). Av- 

 erage passage delays at Little Goose Dam for 

 radio-tagged and control salmon (combined) were 

 216±210 h («=45) in 1976 and 90±57 h{n= 48) in 

 1977 (Table 1). Passage delays for the same fish at 

 Lower Granite Dam were <50 ±19 h (n = 3) in 

 1976 and 58 ±45 h (n = 18) in 1977. 



While our observations of delay at Lower Gran- 

 ite Dam were consistent with previous research at 

 other Columbia and Snake River Dams (Table 1), 

 it appears that excessive delays occurred at Little 

 Goose Dam, especially in 1976. Differences in pas- 

 sage times at Little Goose Dam compared with 

 other dams (Table 1) were significant (P<0.05, 

 Mann- Whitney test). 



Several observations indicate extensive salmon 



Table l.— Delay of tagged adult chinook salmon below Columbia and Snake River Dams. 



'R = radio transmitter; NT = nontelemetering fish tag. 

 ^Values averaged over all fish used in a study. 

 ^Time spent in fish ladders only. 

 ^Time from release 6.4 km downstream to dam passage. 



187 



