condition on the basis of numbers of fish of 

 each species leaving the respective chan- 

 nels (fig. 7). 



Figure 7. — Recording passage of fish at upstream end of 

 transportation channels. Counts of fish leaving each 

 channel provided a record of the response of these fish 

 to the position and depth of the orifices at the entrance 

 to the channels. 



day. Accepting this assumption, we computed 

 the basic response measurements as follows: 



1. The percentage of fish entering each 

 orifice was calculated by day. 



2. The daily percentages of fish entering 

 a given orifice in each 2-day period (1 day 

 in the south and the other in the north chan- 

 nel) were averaged to yield the measurement 

 of basic response for that orifice. 



We evaluated the response to orifices sep- 

 arately for each test condition. Averages of 

 the percentage of fish entering each of two 

 orifices being compared during a 2-day test 

 period totaled 100 percent. Therefore, unequal 

 use of the orifices was shown by the de- 

 parture of the averages from 50 percent. 

 Only average percentages from one of two 

 orifices being compared were used in the 

 evaluation. Since average percentages were 

 replicated for each test condition, we applied 

 t-tests to examine departure of their means 

 from 50 percent. Arcsin transformations 

 (Snedecor, 1957) of the average percentages 

 were used to compute t_- values. 



RESPONSE OF SALMOOTDS TO ORIFICE 

 CONDITIONS 



Experimental Design 



The experiment was divided into the fol- 

 lowing four test conditions to examine the 

 response of salnnonids to orifice alignment 

 and depth: (1) a horizontal and a vertical 

 orifice, both 9 feet below the water surface; 

 (2) a horizontal and a vertical orifice, both 

 3 feet below the surface; (3) two horizontal 

 orifices, one 3 feet and one 9 feet below the 

 surface; and (4) two vertical orifices, one 3 

 feet and one 9 feet below the surface. Test 

 conditions 1 and 2 constituted one block, and 

 test conditions 3 and 4 constituted another 

 block. We randomized test conditions within 

 blocks and ran blocks alternately for three 

 or four replications. 



Because of the possibility of more fish 

 entering one channel than the other, regard- 

 less of the position or depth of the orifice, 

 the orifices were changed from one channel 

 to the other halfway through each test period. 

 Each test period lasted 2 days because it was 

 impractical, for both biological and mechani- 

 cal reasons, to modify a test condition at 

 intervals shorter than 1 day. We assigned 

 orifices to the north and south channels 

 randomly on the first day of each 2-day 

 period. 



This design requires the assumption that 

 approach patterns remain similar on both 

 days for a 2-day test period. In other words, 

 if a majority of fish approach the test panel 

 from the right side on the first day, a simi- 

 lar majority is assumed to do so the next 



The analyses of our experiments, including 

 four test conditions of orifice settings and 

 observations on the lateral orientation of fish 

 in the experimental system, are discussed 

 in the following passages: 



Test Condition 1--Deep horizontal orifice 

 vs. deep vertical orifice 



Significantly more than 50 percent of the 

 Chinook salnnon and steel-head trout entered 

 the vertical orifice (table 1). The percentage 

 of coho salmon passing through the vertical 

 orifice did not differ significantly from 50 

 percent. 



Table 1. — Average percentages^ of salnonids entering orifices in test 

 condition l--deep horizontal orifice vs. deep vertical orifice 



[ Asterisk, significant at 95-percent level; NS, not significant; mean 

 percentages with neither designation not tested] 



^ Direct observations during a basic time unit of 2 days were combined to 

 form a response measurement called "average percentage" (see text for 

 details). 



