In a mass release, a large group (300-800) 

 was released from the collection pool b-y open- 

 ing a large gate between it and the introductory 

 pool. The fish were identified upstream from 

 the exit pool. 



Timing of Fish 



A time -event recorder noted pas sage through 

 the test area. Observers at the release, entr-y, 

 and exit points activated push button switches 

 to transmit information to the recorder, which 

 transcribed the data to an operations sheet. 



As the fish entering the 0.3- and 0.9-m.- 

 diameter pipes could not be seen during the 

 1963 tests, the timing zone was extended from 

 the downstream weir of the introductor-y pool 

 to the upstream weir of the exit pool (A to B, 

 fig. 2). Timing zones were somewhat more 

 precise in 1964, when passage was timed 

 through the introductory pool (release box to 

 point A, fig. 5) and the pipe (A to D, fig. 5). 



Arbitrar-y linnits were established in both 

 ■years so that excessive time would not be 

 spent on fish that failed to pass through the 

 test facility. In 1963, fish were allowed 45 

 minutes to pass through the approach pool 



and 35 minutes to pass through the timing 

 zone. In the 1964 experiments the fish were 

 allowed only 45 minutes to pass through the 

 introductory pool and pipe. If passage was 

 not completed within these limits, timing was 

 stopped, the fish was removed, and another fish 

 was introduced into the system. 



Comparison of Fish Passage 



Median passage times were used to compare 

 the performance of salmon and trout under the 

 various test conditions. A table of confidence 

 intervals (Dixon and Massey, 1957) was used 

 to test the significance of observed differences 

 between these median passage times. The 

 median passage time for a test condition was 

 determined by arranging the passage times 

 of individual fish in an array (table 1) and then 

 selecting the middle value. 



Terminated fish were included and assigned 

 values of 3 5+ minutes in 1963 and 45+ minutes 

 in 1964. This procedure had no effect upon the 

 median as long as 50 percent of the fish re- 

 quired less than 35 and 45 minutes respectively 

 to pass through the pipes. We were unable to 

 compute median passage times in tests where 

 most of the fish were terminated. 



Table 1. — Distribution of passage times of individual Chinook salmon, sockeye salmon, and steelhead trout through 

 30.5 m. of 0.3-m. -diameter pipe at water velocities of 0.3, 0.6, and 1.2 m.p.s., April-July 1963 



-^ 95 percent confidence intervals about "the median. 



