MAJOR and PAULIK: ENCROACHMENT OF WANAPUM DAM RESERVOIR 



for the right-bank releases. In period IV when 

 both releases were from the left bank, the travel 

 time of fish released on August 3 was signifi- 

 cantly higher than that of fish released on July 

 31. Differences were not significant in period V 

 when fish were released from both left and right 

 banks. 



There is little doubt then that in periods I to 

 III (when the spill pattern recommended by the 

 study group was in operation) , fish released from 

 the left bank were finding the ladders and trav- 

 eling over the dam faster than those released 

 on the right bank. During the last two periods 

 (V and VI), the flows were intentionally 

 switched from the center to the right and back 

 again. The effects of this change on fish passage 

 are discussed in greater detail in a later section. 



Further examination of Table 5 reveals two 

 additional characteristics about the movement 

 of tagged sockeye salmon over Rock Island Dam 

 in 1964. First is the consistency of the relative 

 passage times between ladders for fish from a 

 given tag release. This means that regardless 

 of how rapidly or slowly fish from a particular 

 release moved, they did so more or less uniformly 

 at all three ladders. Spring chinook salmon 

 (Table 2) varied much more than did the sock- 

 eye in this respect. Second, in five of six com- 

 parisons fish released on the left bank negotiated 

 the right-bank fish ladder faster than did fish 

 released on the right bank. Considered jointly, 



these two features suggest that sockeye were 

 capable of rapid lateral movement in the area 

 downstream from the dam and that the passage 

 of fish released on the right bank was somehow 

 delayed whether they chose the left, center, or 

 even the adjacent right-bank ladder. 



Table 7 depicts the travel times by period and 

 ladder, ignoring release sites. Corresponding 

 statistical tests of the hypothesis of no difference 

 in travel times between the right and center and 

 between right and left fish ladders are included. 

 For these tests, a pooled estimate of the error 

 variance with 880 degrees of freedom was com- 

 puted from within-ladder variances for the 15 

 groups. Because we tested left versus right and 

 center versus right ladders simultaneously, using 

 the same within-period data, we modified the 

 ^-test to control the type I error according to 

 the method suggested by Dunn (1961). 



Only one difference is significant. In period 

 III, the mean passage time through the center 

 ladder was less than through the right ladder. 

 In general, however, passage time does not ap- 

 pear to be influenced by the ladder chosen. 



Analysis of variance tests of the hypothesis of 

 no difference in travel time between fish released 

 on the right and left banks (all ladders com- 

 bined) in 1965 are presented in Table 8. The 

 conclusions from these tests are mixed. Fish 

 traveled over the dam faster from the left-bank 

 release site than from the right-bank release 



Table 7. — Analysis of variance tests of the hypothesis that sockeye salmon, tagged and 

 relased below Rock Island Dam in 1964, traveled over the right fish ladder at Rock Island 

 Dam equally as fast as those traveling over the left and center ladders. 



' N.S. = Not significant at the 0.05 level, accept hypothesis of equal travel times. 



* = Significant at the 0.05 level, reject hypothesis of equal travel times, and conclude that travel time through 

 the center ladder was significantly less than through the right ladder. 



135 



