in the species composition of the catch no doubt 

 influenced the fishing success, and its reaction 

 to the time of season and other physical factors. 



Various authors have recorded different 

 times of best fishing. Hiner (1943) found the 

 best fishing of the summer to be had in June 

 (many Minnesota lalces and several species of 

 fish involved) . On the other hand, Pelton (1950) 

 reported for Lake Alma, Ohio, that the late 

 summer and fall fishing was better than that of 

 the spring and early summer, except for bass, 

 which was best in June. Hart (1940) found that 

 in one year the fishing for bass in Cache Lake, 

 Ontario, improved after mid-July, but in the fol- 

 lowing year it fell off. Eschmeyer (1939) stated 

 that in Fife Lake, Michigan, the poorest fishing 

 occurred in mid-season (late July and early Aug- 

 ust). As Hansen (1942) has said, the peaks of 

 biting may occur at different times for different 

 species, and at different times in different years. 

 Altogether, these records indicate that so many 

 factors are involved, such as water temperatures, 

 species composition, etc., that there can be no 

 general rule as to when is the best time of the 

 year to go fishing. 



In table 18 are listed the range and mean 

 of the weekly averages for the summer of 1945, 

 for several of the pools, and a few of the key 

 localities. The fluctuations from the mean were 

 considerable in nearly all instances . A part of 

 this effect may have been caused by the imperfect 

 sampling technique, but a substantial part of it 

 reflects actual fluctuations in fishmg success 

 from week to week. These fluctuations are diffi- 

 cult to account for with the information at hand 

 regarding changes in physical factors. 



Figure 3 shows the day-to-day changes in 

 fishing success for the total (all pools) for the 

 winter 1944-45. The curve is smoothed by mov- 

 ing averages of fives. The calculated day-to- 

 day averages for the entire winter show a range 

 of from . 16 to 1 . 19 fish per hour, with a mean 

 of 0.61, a standard deviation of 0.24, and a co- 

 efficient of variation of V = 39 . 



February. There was little tendency for the in- 

 dividual pools or localities to follow the total 

 averages in trends throughout the seeson. There- 

 fore, the total averages are composites. Etespite 

 this fact the average does exhibit a seasonal pat- 

 tern which prooably is caused by seasonal changes 

 in physical factors . 



Hiner (1943) found that for the Minnesota 

 lakes there was no significant difference between 

 the fishing success in December and that in Janu- 

 ary. For the sauger fishery in western Lake 

 Erie, Doan (1944) reported a distinctly rising 

 success through the winter season. As with sum- 

 mer fishing, there are many factors which may 

 influence the time of best winter fishing in any 

 given body of water . 



Comparison of two years. Figure 4 shows the 

 week-to-week variations in the average (total all 

 pools) for the two winter seasons. Although 

 visual inspection of the two graphs of this figure 

 shows a superficial resemblance, there is only a 

 very weak mathematical correlation (r = 0.57; 

 P = 0.1). For a sample locality, Gremore Lake 

 (figure 5), the weekly averages for the two sea- 

 sons followed vaguely similar patterns, but a 

 mathematical correlationship cannot be demon- 

 strated. 



The grand total averages for the two 

 winter seasons are 0.54 fish per hour for 1944- 

 45, and 0.46 for 1945-46. Since for each of 

 these seasons the day-to-day or week-to-week 

 deviations from the mean were large, it cannot 

 be proved statistically that there is any significant 

 difference between the two means, or that fishing 

 was any better mathematically in one winter than 

 in the other . However, if from the computation 

 we exclude the small scattered localities with a 

 low number of hours of fishing and use only the 

 20 and 13 localities, respectively, which are used 

 in table 17, it turns out that there is a mathemat- 

 ically significant difference in the two seasonal 

 means. The means are 0.52 for 1944-45, and 

 0.37 for 1945-46. Calculations yield a value for 

 d/ ^d equal to 3 . (for which P is less than . 005) . 



Some general trends are evident; particu- 

 larly a fairly steady decline from the beginning 

 of the winter fishing to about the end of January, 

 and a secondary peak in the first few days of 



A comparison of the two all-over means 

 is perhaps an inaccurate way of comparing fish- 

 ing successes for the two winters, since the 

 territories involved were of different extent . 



31 



