230 



FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE 



box below the electrical array, counted, and ex- 

 amined for mortalities or any sign of visible injury. 

 The average time required to conduct each test, 

 from the time the fish were released until they 

 were recaptured, was 25 minutes. 



After all of the fish had been examined and the 

 data recorded, they were transferred to an outdoor 

 holding tank where they were held 30 days for 

 observation. The dead fish were picked out daily 

 and carefully examined for marks or bruises. At 

 the termination of the experiment, the live fish 

 were also examined for any signs of visible injury 

 and the results were tabulated. 



After every fourth power-on test, but before the 

 water resistivity was changed for the next series 

 of tests, approximately 100 control fish were 

 tattooed and released into the experimental tank. 

 The control fish were handled in exactly the same 

 manner as the test fish, except that the array was 

 not energized while they were in the experimental 

 tank. 



The statistical analysis that follows is based on 

 a comparison of the differences in mortality in the 

 fish tested under the various electrical conditions 



and also on a comparison of the mortality of the 

 test fisli with that of the control fish. These com- 

 parisons were made for the differences that occur- 

 red during the first 24 hours following the testing 

 and also for the differences that occurred during 

 the second to thirtieth day after testing. 



RESULTS AND DISCUSSION 



The experimental results are summarized in 

 table 5, and include the number and percentage 

 of fish dead after one day and the cumulative 

 mortalities, by number and percentage, from the 

 second to the thirtieth day. The data from table 

 5 were transformed from percentages to arc 

 sin Vpercentage and the entire analysis was made 

 with the transformed data. 



Immediate effect 



Analysis of variance (table 6) to determine the 

 immediate effect (during the first day) of the 

 variables tested reveals that at the 5-percent sig- 

 nificance level a higher mortality resulted from the 

 square-wave pulses than from the half sine-wave 

 pulses, and that a higher mortality resulted from 



Table 5. — Immediate {1st day) and delayed {2d to SOth day) mortalities, by number and percentage, of the test fish 



and control fish 



111 



w 



I Explanation of test condition symbols: 

 I— Wiring pattern I. 

 II— Wiring pattern II. 

 SW— Square-wave pulses. 

 HSW— Half sine-wave pulses, 

 p.p.s.— Pulses per second. 



