104 



FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE 



Hile concluded that — 



* * * if these selected groups are eliminated the re- 

 maining growth  data can be considered accurate and 

 trustworthy within very narrow limits. 



If Hile's assumptions are correct we should find 

 close agreement between growth of best-repre- 

 sented age groups of herring talten in gill nets and 

 herring of the same ages taken in the less-selective 

 pound nets. This expectation is fulfilled by the 

 data of table 17. Growth of the best represented 

 age groups (III, IV, and V) was almost identical 

 in pound-net and gill-net samples. In shorter 

 age groups, I and II, the greater calculated lengths 

 of herring from gill-net samples indicate that the 

 larger, faster-growing individuals are selected by 

 gill nets. This tendency for herring caught in gill 

 nets to be larger than those taken in pound nets 

 is still present though somewhat reduced in age 

 group III. Because the effects of gill-net selection 

 extends to ages as high as the Ill-group (which is 

 frequently dominant), most detailed analyses of 

 growth in later sections have been based on pound- 

 net samples alone. 



Table 17. — Comparison of growth of lake herring taken in 

 pound nets and in gill nets, by age groups 



' Collections from pound nets in 1948-52, and from 2 to 2)4-inch-mesh 

 experimental and commercial gill nets in 1948 and 1950-52. 



Seasonal differences 



The apparently slow growth indicated by small 

 differences between lengths of lake herring of 

 different age groups at capture, brought out in a 

 previous discussion of the length frequencies of 

 age groups, again suggests the possibility of se- 

 lective destruction of fish of more rapid growth 

 by the commercial fishery. If such a selective 



destruction is taking place and is strong, it should 

 result in growth differences detectable in samples 

 taken in the same year but several months apart. 

 That selective destruction was sufficiently great to 

 influence estimates of growth is indicated by the 

 data of table 18. In every comparison, except 

 the third year of life, in the Ill-group taken in 

 1949, fish caught earlier in the year had higher 

 calculated lengths than did those taken later. In 

 14 of 18 comparisons the advantage of the early- 

 season over the late-season fish amounted to 0.4 

 inch or more. Because of the seasonal differences 

 in growth patterns in fish of the same age group it 

 is necessary to stratify samples according to sea- 

 sons when making discriminating comparisons. 



Table 18. — Comparison of growth of lake herring, by age 

 group, taken in pound nets at different seasons, 1949-51 



Geographic differences 



That environmental conditions must differ in 

 the various parts of Green Bay is obvious (see 

 General Features of Green Bay, p. 88) . If environ- 

 mental conditions influence growth and if the pop- 

 ulation is not regularly mixed by active migration 

 or passive transport with currents, differences in 

 the growth of lake herring captured in various 

 sections of Green Bay should be detectable. 



Differences between growth in northern and 

 southern waters of the bay are indicated by com- 

 parisons of lake herring taken in pound nets at 

 the same time of year at locations separated by 

 considerable distances (table 19). In 10 compari- 

 sons of size at capture for fish of the same age, 

 northern fish were shorter in six, and longer in 

 two; lengths of the remaining two groups were 



