EFFECT OF ELECTRICAL PARAMETERS OX YOUNG SALMON 



231 



the high frequency (30 pulses per second total or 

 6 p.p.s. per field) than from the low frequency 

 (15 p.p.s. total or 3 p.p.s. per field). 



Table 6. — Analysis of variance of (he immediate effect (1st 

 day) of water resistimty (rows), wiring pattern and pulse 

 shape (columns), and voltage and frequency (treatments), 

 on mortality of fingerling silver salmon 



Source of variation 



Water resistivity 



Wiring pattern and pulse shape.. 



Wiring pattern _ 



Pulse shape 



Interaction _._ _. 



Voltage and frequency 



Voltage.. _. 



Frequency 



Interaction 



Residual 



Total 



Sum of 

 squares 



129. 9289 



123.9093 



1. 5813 



122. 1578 



0. 1702 



175.6544 



43.9238 



127. 8596 



3. 8710 



80.4669 



509. 9595 



Degrees 



of 

 freedom 



Mean 

 square 



43.3096 



1. 5813 



122. 1578 



0. 1702 



43. 9238 



127. 8596 



3.8710 



13.4112 



3.23 



0.118 

 •9.109 

 0.013 



3.275 

 •9. 534 

 0.289 



"Signiflcant at 5-percent level. 



As previously mentioned, polarity of the elec- 

 trodes was not alternated when the array was 

 wired according to wiring pattern I and energized 

 with half sine-wave pulses. The polarity of the 

 electrodes was alternated, however, when the 

 array was wired according to wiring pattern I and 

 energized with square-wave pulses. Therefore, 

 the difference in mortality which in table 6 is 

 attributed to the difference between pulse shapes 

 could, in wiring pattern I, be due to the effect of 

 alternating polarity, or to a combination of alter- 

 nating polarity and the difference between pulse 

 shapas, ratlier than to the difference between 

 pulse shapes alone. 



If the difference in mortality rates was due to 

 alternating polarity, either alone or in combina- 

 tion with pulse shape, the analj-sis would be 

 expected to show a significantly higher mortality 

 resulting from wiring pattern II than from wiring 

 pattern I, since polarit}* of the electrodes was 

 alternated for both pulse shapes when the array 

 was wired according to pattern II. Also, the 

 analysis would be expected to show significance for 

 the interaction of wiring patterns and pulse 

 shapes. In both of these analyses, however, the 

 differences are not significant. Therefore, we 

 have concluded that the difference in mortality 

 was due to the difference between pulse siiapes 

 and not to the effect of alternating polarity. Fig- 

 ure 13 is a comparison of the two pulse shapes and 

 substantiates this conclusion. The figure shows 

 that the total electrical energy of a square-wave 



Figure 13. — Comparison of square-wave and half sine- 

 wave pulses. Shaded area indicates additional amount 

 of electrical energy available to the fish with each 

 square-wave pul.se. 



pulse is approximately 30 percent greater than 

 that of a half sine-wave pulse at any constant 

 value of voltage, frequency, and duration. 



Total electrical energ}' is directly proportional 

 to frequency, providing pulse shape, voltage, and 

 pulse duration remain constant. Therefore, the 

 fish subjected to the high frequency' experienced 

 twice as much electrical energy as the fish sub- 

 jected to the low frequency. This difference in 

 total electrical energ}' explains the higher mor- 

 tality that resulted from the high frequency. 



Although no significant differences in mortality 

 resulted from the differences between the vari- 

 ations of water resistivitj', waring pattern, and 

 voltage, certain indications are apparent from the 

 table of means (table 7) and are summarized as 

 follows : 



Water resistivity. — Mortalities that resulted from 

 the four levels of water resistivity were com- 

 bined into two groups: those resulting from the 

 low levels (1,000 and 5,000 ohm cm.) form one 

 group and those resulting from the high levels 

 (10,000 and 15,000 ohm cm.) form the other. 

 Under the conditions tested, it is apparent that the 

 low levels of water resistivity resulted in a higher 

 percentage of losses than the high levels. The 

 analysis of variance shown in table 8 reveals that 

 this difference is significant at the 5-percent level. 



Wirini/ patten}. — Tlie difference in mortality 

 that can be attributed to the difference between 

 wiring patterns I and II is only 0.5 percent. 



Voltage. — High voltage (250 volts) appears to 

 have resulted in a higher percentage of losses than 

 low voltage (165 volts). 



Analysis of variance shows that significant dif- 

 ferences in mortality resulted from the differences 

 between the variations of pulse shape, frequency, 

 and water resistivity (when the mortalities result- 

 ing from the four levels of water resistivity were 

 combined into a high and a low group and the 

 variance test was between groups). However, it 



