evaluated the fish screening capability in 2000, and then again in 2002 following 

 correction of observed construction flaws. 



In 2000, we captured 48 westslope cutthroat trout using a backpack mounted, 

 battery-powered DC electrofishing unit (Smith-Root). Fish were anesthetized with 

 tricaine methanesulfonate, counted and measured for total length. After fish recovered 

 from the anesthetic, we released individual fish through the fountain intake. As fish 

 exited the intake riser, we counted and visually estimated total length to the nearest 25 

 mm and timed the duration of all fish impinged on the screen for > 2 seconds. After all 

 fish passed, we walked up-and downstream of the bypass exit to visibly detect signs of 

 related mortality or signs of injury. 



During this impingement evaluation of our initial fish screen design, we identified 

 two construction flaws that appeared to contribute to unnecessary impingement: 1) the 

 close proximity of inner chamber to a portion of outer wall of the structure, and 2) a 

 lower screen angle than specified in our prototype design (photo 1). Due to the first 

 construction flaw, the fountain was unable to completely wash debris from the edge of the 

 screen. At this location, fish were unable to wash free of the screen and were impinged 

 on the screen against the debris. Based on this observation, we modified our original 

 screen by adding a flow deflector shield to the fountain riser in order to direct water, 

 debris and fish away fi-om this area of screen. We also modified the shape of the screen 

 fi-om a low-angle flat screen to a rounded cone-shaped screen with a mean 1% slope 

 (Figure 54). 



Following these screen modifications, in 2002 we repeated the impingement trial 

 with 66 westslope cutthroat trout entrained through the fountain intake. The capture, 

 handling and observation of these fish were similar to the previous trial. During both 

 experiments, the fountain intake was operaUng at fiiU (0.14 m^s) capacity. 



We used a Mann- Whitney nonparametric tests to compare fish lengths in the 

 initial trial to fish lengths in the second trial. We also used Mann- Whitney to compare 

 the duration of impingement between the prototype and modified design. A chi-square 

 analysis was used to test whether the number of impinged fish varied by size class (50- 

 110mm, 111 -150mm, and >151mm) between trials, where the number of impinged fish 

 fi-om the original design trial was used as the expected values of impingement in the 

 modified design trial. In all cases, differences were considered significant at /^-values < 

 0.05. 



Results 



Prior to screen modification, 3 1 westslope cutthroat trout (65%) passed through 

 the fountain with no impingement (<2 seconds) on the screen. Seventeen fish (35%) 

 were impinged for >2 seconds, of which 14 managed to work free of the screen (median 

 impingement time, 30 (range, 2-1560 seconds)). Three (6%) of the sampled fish 

 remained on the fish screen after 26 minutes when we ended the experiment (Table 15). 



Following screen modificafions, all but four (6%) of the 66 fish immediately 

 passed through the fountain and screen with no impingement (<2 seconds). Of the four 

 impinged fish, all washed over the screen within four seconds (median, 2 (range = 2-3 

 seconds)). For these four fish, all impingement occurred in a localized boundary area 

 between the main flow and the shielded portion of the screen. 



88 



