LAKE HERRING OF GREEN BAY, LAKE MICHIGAN 



109 



2. Segregation as to maturity during the 

 spawning run. 



3. Segregation as to size, independent of ma- 

 turity. 



4. Higher mortality rate among fast-growing 

 fish than among slow-growing. 



Hile (1936) found that discrepancies in the 

 calculated lengths of ciscoes in three of four 

 Wisconsin lakes were the result of faulty sampling 

 traceable to selective action of gill nets. Carlander 

 (1945) attributed Lee's phenomenon in ciscoes 

 of Lake of the Woods, Minnesota, to the selectivity 

 of large-mesh gill nets, as well as to differential 

 mortality of fast- and slow-growing fish as pro- 

 posed by Hile (1936). Eddy and Carlander (1942) 

 also found the phenomenon in ciscoes of Gull Lake, 

 Minnesota. 



Van Oosten (1929) and Cooper (1937), whose 

 samples came from spawning-run lake herring in 

 Saginaw Bay and Blind Lake, respectively, 

 offered similar explanations of Lee's phenomenon. 

 Their views are expressed adequately in the 

 following quotation (p. 570) from Cooper's 

 paper: 



* * * the lake herring first reaches maturity during its 

 third, fourth, or fifth year of life, depending upon individual 

 rate of growth; the more rapidly growing individuals of 

 any one year class attain maturity first. It follows that 

 the youngest year groups were represented in the catch 

 (from the spawning grounds) only by their biggest indi- 

 viduals and, as older age groups were considered, more 

 and more of those fish that had been smaller individuals 

 in their earlier years appeared in the older groups. There- 

 fore the younger age groups contained a larger proportion 

 of fast-growing fish than did the older groups and, con- 

 sequently, the computed lengths for the early years of 

 life would be greater in the younger age groups than in the 

 older. The persistence of the phenomenon in the older 

 age groups (in groups in which all individuals are mature) 

 may be explained on the basis of differential mortality, 

 that is, on the assumption that the more rapidly growing 

 fish die off earlier in life than the more slowly growing fish. 



In Green Bay, as has been pointed out, segregation 

 by size (^and hence by rate of growth within a year 

 class) appears to take place at all seasons. 



Evidence was presented by Hile (1936) that a 

 high natural mortality rate was correlated with 

 rapid growth in the cisco population of Silver 

 Lake. Cooper has suggested differential mortality 

 as a possible factor in Lee's phenomenon. Hile 

 also advanced the hypothesis that, if there was 

 scgriigation of fast- and slow-growing fish with 

 depth, the gill nets which were always fished on 



the bottom could not take equal samples of both. 

 Fry (1937) demonstrated that faster-growing 

 young fish were found in deeper waters of Lake 

 Nipissing during the summer and were joined 

 in successive years by more and more of the 

 slower-growing members of the same year class. 

 Behavior of this type explains why Lee's phe- 

 nomenon might be found in samples taken in a 

 certain location at a particular period of the year. 

 Although a difference in seasonal distribution of 

 fast- and slow-growing lake herring may exist 

 in Green Ba}' and may be contributing to Lee's 

 phenomenon there, it cannot be the main causative 

 agent, because the phenomenon exists in samples 

 collected at different depths and at different 

 locations in the same and different seasons. 



Growth compensation 



Growth compensation — the tendency for the 

 smaller fish at a particular age to have the more 

 rapid subsequent growth — seems to be common 

 among fish (V&n Oosten 1929 ; Eddy and Carlander, 

 1942). The existence of growth compensation 

 was mentioned in 4 of 14 publications on growth 

 of lake herring (Carlander 1945, in Lake of the 

 Woods tullibee; Eddy and Carlander, 1942, in the 

 tullibee of 17 Minnesota lakes; Hile 1936, in the 

 cisco of four Wisconsin lakes; and Van Oosten 

 1929, in the Saginaw Bay lake herring). Growth 

 compensation seems to be a general occurrence in 

 North American coregonids. It has been shown 

 in the following stocks: Lake Michigan kiyi 

 (Leucichthys kiyi) by Deason and Hile (1947); 

 Reighard's chub (Z. reighaidi), longjaw cisco (Z. 

 alpenae) and bloater (L. hoyi) of Lake Michigan 

 by Jobes (1943, 1949a, and 1949b); Lake Huron 

 whitefish by Van Oosten (1939) ; and Lake Superior 

 longjaw {L. zenithicus) by Van Oosten (1937). 

 McHugh (1941) did not find growth compensation 

 in several populations of Rocky Mountain white- 

 fish (Prosopium vnlliamsoni) . 



Of the authors who mentioned growth com- 

 pensation in studies of lake herring only Hile 

 (1936) and Van Oosten (1929) discussed its char- 

 acteristics in any detail. Carlander (1945, p. 

 129) stated that— 



As was demonstrated for the ciscoes by Van (losten 

 (1929) and Hile (1936), growth compensation occurs in the 

 Lake of the Woods tullibee but the compensation is not 

 great enough to overcome any advantage in length which 

 large individuals may hold at the end of the first year. 



