on the grounds fished for other species than on 

 the actual abundance of fish. Neither would 

 abundance percentages of the whitefish taken in 

 the fall be reliable, not only because of the un- 

 known number of small -mesh nets fished with 

 the whitefish nets, but also because of the vari- 

 ations in weather conditions which in some years, 

 for example 1938, would delay the spawning 

 season to such an extent that it would not reach 

 its climax before commercial fishing operations 

 were suspended for the year. 



Catch per lift of northern pike was well 

 above average in 1930, improved in 1931, and 

 reached a maximum in 1932. A sharp decline 

 in 1933 was followed by a less sharp improve- 

 ment in 1934. The abundance of northern pike 

 appears to have been consistently below average 

 throughout the period, 1935-1938. 



Catches of freshwater drum were above 

 average in 1931, 1932, and 1935-1937 and below 

 average in 1930, 1933, 1934, and 1938. The best 

 catches were made in 1936 and the poorest in 1934. 



The preceding discussion was based on 

 the data of all fishermen in each summer's fish- 

 ing. In view of the considerable annual fluctua- 

 tions in the number of fishermen (tcible 6), the 

 possibility was considered that some of the ob- 

 served variations in the catch per lift may have 

 originated in variations from year to year in the 

 skill and ability of the fishermen engaged. To 

 test this possibility an analysis was made of the 

 annual fluctuations in the catch per lift of 19 

 selected fishermen (the best fishermen as se- 

 lected by the manager) who fished each year 

 throughout the period 1930-1938. Althou^ the 

 average catch per lift of all species made by the 

 19 fishermen was 1.5 times the average catch 

 per lift obtained by all fishermen during the 

 years 1S80-1S68, the conclusions, with respect to 

 relative abundance, that would have been drawn 

 from the data of these selected fishermen were 

 practically identical with those based on the data 

 of all fishermen. The description of fluctuations 

 in abundance from annual variations in the catch 

 per lift of all fishermen may be considered valid. 



THE WALLEYES OF LOWER RED LAKE 



The walleye, as demonstrated during the 

 preceding discussion of the statistics of the com- 

 mercial fisheries of the Red Lakes, was by far 



the most important species produced. During 

 1927-1938, the walleye averaged 62.3 percent 

 of the annual catch . 



Scale samples, accompanied by records 

 of standard, fork, and total lengths, weight, sex, 

 and stage of maturity, were collected from 127 

 individuals which represented the unselected 

 catch obtained by 3 fishermen on August 24, 1938, 

 in 18 nets. Similar data were obtained on August 

 25 from 185 fish selected to secure a good repre- 

 sentation of fish of all sizes and ages. Length 

 measurements only were obtained from 470 wall- 

 eyes, which, together with the 185 from which 

 scale samples were taken, constituted the total 

 number taken by 6 fishermen from 34 nets lifted 

 on August 25. 



The scales, cleaned and mounted on slides 

 in a glycerine -gelatine medium, were studied on 

 a microprojection machine. Measurements of the 

 projected images of the scales from the focus 

 (center) to each annulus (yearring) and to the edge 

 of the scale were used to calculate lengths of the 

 fish at the end of each year of life. Computations 

 were made by the direct -proportion method. Ages 

 are designated by roman numerals which denote 

 the num ber of annuli (year rings) on the scales or 

 completed years of life of the fish. Since the col- 

 lections were made during the growing season 

 each fish was actually a fraction of a year older 

 than the age indicates . 



Length -frequency distributions 



The length -frequency distributions of the 

 various age groups of walleyes (table 9) have been 

 based on a combination of 127 fish of an unassorted 

 sample and 185 individuals selected to obtain a 

 better representation of fish over certain length 

 intervals. Comparisons of the frequency distribu- 

 tions of the age groups in table 9 with the distribu- 

 tions based on the unassorted sample of 127 fish 

 (not shown), revealed that the inclusion of the 

 selected individuals did not affect the data signif- 

 icantly. Inclusion of the selected fish did, however, 

 affect the age composition of the samples as may 

 be seen by the comjiarison of the percentage age 

 compositions of the unassorted sample (bottom of 

 table) and the sample containing both unassorted 

 and selected fish. In the unassorted sample agp- 

 groups rV and V made up 81 .9 percent of the total; 

 in the combined samples, 64.3 percent. Age-group 

 IV dominated both samples . At the extreme right 

 of the table is the distribution of fish in the 



25 



