Table 2 

 Chi-square test of 1950 data from Fishing Bridge Dock 

 Total QJdd Even Expected 



1/ 



As neither of the Chi-square values were significant^ the hypo^ 

 thesis was accepted» It was concluded that half as many days would 

 sample one=half of the censused fishing parties and fishermen. ThuSj 

 if the mean catch~per°unit=effort for each of the halves were not signif- 

 icantly different^ the estimate made from either half of the census days 

 would produce the same seasonal estim^ate as that from the other half. 

 Catch-per^unit^ef forts were calculated for each of the census days of 

 1950j and it was found that the means for the odd and even census days 

 were exactly equal, to three decimal places, O»739o With this informationj 

 it was felt that with confidence we could reduce the n'omber of census 

 days during the 1951 season to four days for each two-week period without 

 decreasing the accuracy of the final estimate. 



A summary of the data collected from the rowboat rentals at Fishing 

 Bridge dock during the 1950 and 1951 seasons is shown in table 3o 



Table 3 shows a noticeable similarity in the effort^ catch-per=>unit~ 

 effort, and the estimate of the total nunber of fish taken by rowboats 

 from Fishing Bridge dock for the two years. As the catch-per=>unit-=effort 

 is an index of the relative success of fishing^, it would be essential to 

 know if the slight difference in average catch-per-unit-effort between 

 the two years is significant. A graph of the catch-per-iinit=effort 

 values for each 2=week period (fig. 2) for 1950 and 1951 shows them to 

 follow rather closely throughout the season, A graph of the rates of 

 effort shows more distant relationships. 



A test of the difference between the variance of biweekly means for 

 each year for catch-per-unit=-effort and for effort was not significant at 

 the 0o05 level for rowboat or guideboat fishing at Fishing Bridge or West 

 Thumb dock. Thus^ the "t" test was used to i:est the hypothesis that 

 there was no difference between the grand means of the biweekly means for 

 each unit. The results of the test for the rowboat fishery at Fishing 

 Bridge dock are shown in table Ii. 



1/ P — Oe05 is considered as significant in all tests in this paper o 



