There was no difference between the total length of fish in the first test and the 

 second test (Mann- Whitney, P=0.081, Table 1). nor did the proportion of impinged fish 

 vary significantly among the three size classes between the first and second trial (x^ =3.0. 

 df =2. P=0.223). The number of impinged fish declined from 37% in the first test to six 

 percent in the second test (Table 12). The duration of impingement between the first and 

 second test also declined significantly (Mann- Whitney, P=0.006). 



Upon completion of both experiments, we walked up-and downstream of the 

 bypass and found no evidence of injury or mortality resulting from impingement from 

 either test. 



Table 12, Numbers and sizes of impinged fish before (2000) and after (2002) 

 modification of the original screen. 



Total lengths (mm) Number fish Size class of impinged fish (mm) 

 Year N mean (SD), range impinged (50-110) (111-150) (>151) 

 2000 48 134(45). 61-241 17 9 3 5 



2002 66 121(38), 61-216 4 2 \ 1 



Discussion 



Our modifications and evaluations of turbulent fountain screens suggest that this 

 device can provide an effective, low-cost, low-maintenance fish screening system. Our 

 field trials further outline the importance of constructing screens to exact design 

 specifications. 



Based on our design - evaluate - modify approach, the following observations 

 will help ensure effective application of this screen in the fijture. Fabricators and 

 installers should ensure there is sufficient distance between the inner and outer chamber 

 to facilitate the movement of fish and entrained debris off the screen, plus include a 

 sloped (or crowned) screen with a minimum 1% slope. Not only do lower angle screens 

 increase impingement, but also several fish, once on the original more horizontal screen, 

 attempted to swim towards the main flow (center) of the fountain. These fish remained 

 on the screen for an extended time before escaping. A smaller diameter inner chamber 

 with minimal screen surface would also clean the screen more efficiently and reduce fish 

 contact with the screen on the outer portion of the screen. A larger intake with excessive 

 volume would serve a similar purpose by washing fish more quickly fi-om the screen. 

 Similarly, a smaller diameter out-flow pipe relative to intake pipe diameter forces 

 upwelling on the outer portion of a sloped screen and assists in washing fish from the 

 screen with less screen contact. Another possibility that was not tested might be to 

 elevate the bypass pipe or otherwise submerge the screen in order to minimize fish 

 contact with the screen and enhance fish passage over the screen. 



89 



