YELLOW PERCH IN SAGINAW BAY 



389 



between members of the same age group. Al- 

 though the same observation does not hold well 

 for the females the data offer evidence that much 

 of the disagreement among calculated lengths of 

 different age groups is established between the 

 spring and fall during the time of active growth 

 in length and when fishing intensity is high. 



Jobes (1952) discussed the possible causes of 

 these discrepancies among calculated lengths in 

 full detail. The present comments' will be lim- 

 ited, therefore, to those factors believed to be 

 most important in Saginaw Bay perch. 



Much of the disagreement in calculated lengths 

 of Saginaw Bay perch can be attributed to biased 

 samples and to the progressive elimination of the 

 more rapidly growing individuals from the popu- 

 lation. These two sources of discrepancy are 

 interrelated. 



Selection according to legal size: Although 

 this factor was mentioned in the preceding para- 

 graph, it is given a special listing because of its 

 great importance. The mortality of legal-sized 

 fish caught in commercial nets is 100 percent. 

 Inasmuch as the commercial fishery is supported 

 principally by age groups whose length distribu- 

 tions cross the legal minimum of 8V2 inches this 

 selective destruction, particularly if the fishery 

 is intensive, has a profound effect on the growth 

 characteristics of the survivors. 



Selective action of gears: The mesh sizes of 

 the commercial trap nets by which the samples 

 were taken (about 2V 2 inches, extension measure) 

 were large enough to permit the escape of the 

 smaller individuals of the younger age groups 

 (the older the fish the higher the percentage held 

 by the nets). As a result, the size at capture 

 and calculated lengths of the younger age groups 

 were overestimates of their growth. 5 Further- 

 more, these larger members of the younger age 

 groups, as a result of being caught, were sub- 

 jected to certain mortality hazards. The few that 

 had reached legal size (8y 2 inches) all were 

 killed. Those shorter than. 8y 2 inches were re- 

 turned to the water but they experienced some 

 mortality from handling and sorting (nearly all 

 fish are dead in gill nets). The extent of the 

 mortality of small fish is not known, but Jobes 

 (1952) estimated that in Lake Erie 14 percent 



6 The selective action of such other gears as pound nets, fyke 

 nets, and seines is similar to that of trap nets. Gill nets are 

 even more selective as both the smallest and largest fish escape 

 capture. 



of the perch were dead when the trap nets were 

 lifted. This selective destruction of the more 

 rapidly growing fish surely affects the growth 

 characteristics of samples from the same year 

 classes taken in subsequent years at older ages. 



Selection due to segregation by sexual maturity 

 and size: Segregation according to maturity can 

 be significant in the present study since the prin- 

 cipal samples were taken in the spawning season 

 during which time small, immature fish tend to 

 avoid the spawning grounds. Data from our 

 samples (see Size at Maturity, p. 408) gave no 

 evidence of this type of segregation, although 

 other investigators have reported it frequently. 

 A scarcity of immature fish can mean a segrega- 

 tion according to size because the fast-growing 

 individuals usually mature earlier. Selective de- 

 struction of the mature fish during the spawning 

 season is of limited consequence since most or all 

 of the spawning is covered by the closed season 

 during which all perch must be returned to the 

 water. 



Segregation according to sex and size can lead 

 to biased sampling and selective mortality at all 

 times of year. Evidence is strong for a segrega- 

 tion of sexes in various months outside the spawn- 

 ing season (Eschmeyer 1937, Weller 1938, Jobes 

 1952) . It is well known also that perch of differ- 

 ent size inhabit different regions of a lake, but 

 the nature of this segregation apparently differs 

 among populations ( Hile and Juday, 1941 ) . Data 

 are lacking for a description of segregation by 

 sex and by size in Saginaw Bay, but undoubtedly 

 it occurs. This segregation can lead both to 

 biased samples and selective mortality since fish- 

 ermen can be expected to concentrate their efforts 

 on grounds occupied by the larger fish. 



Greater natural mortality rate for fish with 

 rapid growth: Hile (19.36) found a higher natu- 

 ral mortality rate among the rapidly growing 

 ciscoes of Silver Lake (northern Wisconsin) than 

 among the slow-growing ones. Whether a simi- 

 lar differential mortality occurs in the Saginaw 

 Bay yellow perch is not known. Indeed, it would 

 be most difficult to obtain information on the 

 question in view of the known selective destruc- 

 tion of rapidly growing individuals in the fishery. 



Evidence that the discrepancies of calculated 

 lengths among the age groups of Saginaw Bay 

 yellow perch are of the type that would result 

 from the selective mortality of fish with the more 

 rapid growth is given in table 30 which contains 



