LAKE HERRING OF GREEN BAY, LAKE MICHIGAN 99 



Table 10. — Weight distribution of lake herring, by age group, taken from pound nets in January and February, 1949-St 



Table 1 1 . — Relation between magnified (X41) scale diameter 

 and total length of Green Bay lake herring 



' Means for fish within a H-inch interval of totallength. 



* Means of the body-scale ratio computed for individual fish. 



direct-proportion calculations are valid for all 

 Green Bay herring. Thus, the method established 

 for Saginaw Bay lake herring is applicable to the 

 Green Bay stock as well. 



The graphical representation of the relation 

 between body length and scale diameter (fig. 3; 

 see also table 11) is based on data for the sexes 

 combined. Regression lines fitted to the data for 

 the sexes separately exhibited no statistically 

 significant differences. This body-scale relation 

 is obviously linear. The fitting of a least-squares 

 line to the means ^ of scale diameters and body 

 lengths yielded the following equation : 



Z = 0.01615 + 0.05486 S, 



where L is total length in inches and S is the 

 magnified scale diameter in millimeters. The 

 intercept of less than 0.02 inch on the axis of fish 

 length is so small that calculations of growth from 

 the least-squares equation are nearly identical 

 with those that would be obtained if the intercept 

 were assumed arbitrarily to be 0. In other words. 



' Weighted means are used to reduce error in slope of the least-squares 

 regression due to scatter. Tests of statistical significance were based on 

 individual measurements. 



388748 O — 57 3 



50 100 ISO 200 2S0 300 



SCALE DIAMETER (mjLLIMETERS X 4l) 



Figure 3. — Relation between total length of fish and mag- 

 nified (X41) scale diameter in the Green Bay lake her- 

 ring. The dots show empirical data; the slope of the 

 line is the mean body-scale ratio. 



