120 



FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE 



collections is demonstrated by the following 

 tabulation of percentages of females in samples 

 of four different year classes at various ages (where 

 represented by 25 fish or more). 



Table 35. — iSex composition of lake herring taken in gill 

 nets, 1948-52 



[Number of flsh In parentheses; males at left, females at right] 



The corresponding tabulation for collections 

 made in months other than January and February 

 demonstrates a similar trend but does contain one 

 exception, the V-group of the 1944 year class. 



In gill-net collections the percentage of females 

 varied widely from sample to sample (table 35). 

 Although the available data are insufficient for a 

 study of annual and seasonal trends, they offer 

 no evidence of disagreement with the trends 

 established in pound-net data. The change in 

 sex composition with age of gill-net caught fish, 

 however, is the reverse of that of fish taken by 

 pound nets. The gill-net samples show a clear 

 tendency toward a higher percentage of females 

 with increasing age. This progressive destruction 

 by gill nets of females in the older age groups 

 should tend to counteract the effect of pound nets 

 in cropping younger females at a faster rate than 

 males. 



The combined effects of the selective destruction 

 of the two fishing gears in determining differences 

 in sex composition with season and age cannot be 

 evaluated with data at hand. It is clear, however, 

 that females are cropped more heavily than males 

 during the winter (January-February) fishery by 

 both pound nets and gill nets. The effects of this 

 destruction of females are counteracted in part by 

 the greater destruction of males in the remaining 

 months of the year. 



Records of the sex composition of samples of 

 lake herring taken at various levels between the 

 surface and bottom (see description of oblique 



1 Collection from commercial gUl nets— 1948, 1960, 1951; collection from 

 experimental gill nets— 1962. 



gUl-net sets in Vertical Distribution in Green Bay, 

 p. 128), yielded no evidence of segregation of the 

 sexes according to depth in June or July, but they 

 indicated a strong tendency in October toward a 

 higher percentage of females in the deeper strata 

 than in the shallower (table 36). This trend was 

 much stronger in samples from nets fished in 60 

 feet of water than in 90 feet (fig. 11). 



30 ^S 60 



DEPTH (FEET) 



Figure 11. — Sex composition of lake herring taken at vari- 

 ous depths in October 1952 in 2-inch-mesh experimental 

 gill nets set obliquely from surface to the bottom. Open 

 circles indicate the data for 60-foot stations and solid 

 circles the data for 90-foot stations; the regression lines 

 were fitted by least squares. 



