232 



FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE 



Table 7. — Immediate, (first day) mean percentage mor- 

 tality resulting from each of the test conditions 



(Table of means] 



Table 8. — Analysis of variance of the immediate effect (1st 

 day) of low (1,000 and 5,000 ohm cm.) and high (10,000 

 and 15,000 ohm cm.) levels of water resistivity on the 

 mortality of fingerling silver salmon 



' See table 6. 



•Significant at 5-percent level. 



does not show whether the least-harmful varia- 

 tions of these test conditions caused a significant 

 number of mortalities in the test fish compared 

 with that in the control fish. Therefore, a chi- 

 square analysis was conducted to determine 

 whether a difference in mortality existed between 

 the test fish and the control fish. 



In this analysis, the mortalities resulting from 

 the low and high levels of water resistivity and 

 each variation of pulse shape and frequency were 

 grouped into eight combinations. Each combi- 

 nation included the mortalities resulting from one 

 variation of each of these three variables. The 

 total percentage mortality of each combination 

 was then compared with the total percentage 

 mortality of the controls. The results of this 

 analysis are shown in table 9. 



Table 9 reveals that the least-harmful test 

 variations (high water resistivity, half sine-wave 

 pulses, and low frequency) did not cause a sig- 

 nificantly higher mortality of test fish tlian of the 

 control fish. The table also shows that the test 

 variations caused a liigh mortality only when two 

 or more of the harmful variations were combined. 



Table 9. — Chi-square comparison of the immediate (1st day) 

 mortality of the test fish with that of the control fish 



Comparison ' 



Control 



LR, LF, HSW. 



Control 



LR, HF, HSW. 



Control 



LR, LF, SW... 



Control 



LR, HF, SW.. 



Control 



HR, LF, HSW. 



Control 



HR, HF, HSW 



Control _ 



HR, LF, SW.. 



Control 



HR, HF, SW.. 



AT! 



o.oo 



•7.62 

 •11.85 

 •45.60 

 0.205 

 2.437 

 0.958 

 •3.97 



1 Explanation of symbols: 



LR — Low levels of water resistivity (1,000 and 5,000 ohm cm.). 



HR— High levels of water resistivity (10,000 and 15,000 ohm cm.). 



LF — Low frequency (15 p.p.s.). 



HF— High frequency (30 p.p.s.). 



HSW — Half sine-wave pulses. 



SW — Square-wave pulses. 



• Significant at 5-percent level. 



This analysis is based on the assumption that 

 the variations of wiring pattern and voltage did 

 not cause a significant number of deaths among 

 the test fish when compared with the control fish. 

 However, since voltage and wiring pattern were 

 integral parameters of the experiment, they were 

 factors m the results of the analysis. Therefore, 

 they would have caused the entire analysis to 

 show significance if the variations of either had 

 caused a significant number of mortalities in the 

 test fish as compared with that in the control fish. 



The analysis is also based on the assumption 

 that no difTerences in mortality existed among the 

 four groups of control fish. To test this assump- 

 tion, another chi-square analysis was made, and it 

 showed that the differences in mortality among 

 control groups were not significant at the 5-percent 

 level. The chi-square value for this test was 

 1. 273 with 3 degrees of freedom. 



Delayed effect 



Analysis of variance to determine the delayed 

 effect (second to thirtieth day) reveals that at the 

 5-percent significance level there were no dif 

 ferences in mortality that could be attributed to 

 the differences between the test variations (table 

 10). Again, however, certain indications are ap 

 parent from a table of means (table 11) and are 

 summarized as follows: 



General. — With the exception of frequency, the 

 variations of the test conditions resulted in 

 approximately the same differences in mortality 



