TABLE 2. — Passage times of 18 chinook salmon that ascended a l-on-16 slope endless 

 fishvra.y in plunging ( circuits 1, 3, and 5) and streaming ( circuits 2, A, and 6) 

 flows ( flows were alternated on each successive circuit) 



Passage time per circuiti. 



1/ 



2/ Time in pool 2 excluded 



markedly faster in the plunging flow. When all 

 data were combined, however, the difference 

 between performance under the two conditons 

 was negligible. 



Passage times during plunging and stream- 

 ing flows usually decreased on each successive 

 circuit in the respective flows (table 2). This 

 tendency for improved performance during 

 repeated ascents was attributed to learning by 

 Collins, Gauley, and Elling (1962). 



Effect on Behavior 



The behavior of the salmon was studied in 

 plunging and streaming flows to determine 

 (1) areas of the pool used and (2) orientation 

 to flow. 



Areas of pool used . --Time in pool 13 was 

 recorded according to the quadrant in which 



the fish was observed. Time spent (resting 

 and moving) in each of the quadrants was ex- 

 pressed as a percentage of total time in the 

 pool (fig. 7). All fish spent most of the total 

 time in the lower half of the pool regardless of 

 the condition of flow. During plunging flow, 

 76 percent of the time was spent in the lower 

 downstream quadrant (quadrant 2), whereas 

 under streanning flow, the greater portion of 

 time (57 percent) was spent in the lower up- 

 stream quadrant (quadrant 3). The average 

 time that fish remained in the viewing pool 

 during each observation was slightly more 

 than 1 minute vinder both conditions of flow. 



Orientation to flow . - - A numbe r of be havio r al 

 features were evident in the study, but all were 

 classified into three general groups: (A) direct 

 in-line passage through the pool, (B) circuitous 

 passage involving continuous movement, and 



