may have been held beyond their spawning time 

 in the trap nets, which were lifted only every 

 other day. 



It is also probable that some gizzard shad are 

 not able to find a suitable spawning site by the 

 time the eggs are ready to be expelled. This 

 possibility is supported by the observation that 

 some of the female shad caught in the lake by 

 commercial fishermen were releasing eggs. To 

 explore this possibility further, I determined the 

 percentage of spawning fish for the II-group and 

 older female shad caught in Fishery Bay and for 

 those captured in the lake during May, June, and 

 July 1954-55 (table 20). The state of the ovaries 

 was determined by gross examinations. The 

 much smaller percentages of spawning shad cap- 

 tured in the lake than in the bay suggest that most 

 of the ripe females migrated from deeper to shal- 

 lower water during the spawning season. This 

 finding is not in disagreement with the earlier 

 observation (tables 16 and 17) that the percentage 

 of females in June was less in the shoal water than 

 in the deeper water. The high percentage of males 

 on the spawning reef can be explained on one 

 or more of the logical assumptions that: greater 

 numbers of males than females entered the shoal 

 area from the lake; that activity was greater 

 among the males than among the females in the 

 shoal waters; that males individually stayed longer 

 on the site. 



When fish begin to spawn, the rise in the 

 "spawning" category should coincide with the 

 drop in the "nonspawning" category if the fish 

 do not migrate from the area. Lack of this rela- 



tion in the lake provides further evidence that 

 most of the fish do not remain in the deeper water 

 when they are ready to spawn. 



The rapid rise of the "spent" category in the 

 lake from mid-June through July, which is com- 

 pletely out of proportion with the drop in the 

 "spawning" group there, indicates a return of the 

 spent fish from shallow water to the open lake. 



The data of table 20 on the percentages of non- 

 spawning, spawning, and spent females captured 

 in Fishery Bay and in the lake during May, 

 June, and July 1954-55 also give a basis for the 

 estimation of the beginning, peak, and end of the 

 spawning season of II-group and older shad. 



The earliest samples containing spawning fish 

 were captured in the first quarter of June on the 

 spawning bar in Fishery Bay. The percentage 

 of spawners caught here rose from 33.3 percent 

 in that quarter to a high of 58.3 percent in the 

 next quarter, then dropped to 45.0 percent in the 

 third quarter and abruptly to 1.2 percent in the 

 last quarter of June. Spawning shad were 

 captured from the first quarter of June to the 

 first quarter of July inclusive. 



Although I lacked samples from the lake in the 

 first quarter of June (the commercial fishermen 

 of South Bass Island — the source of the lake- 

 caught gizzard shad — did not fish during the last 

 quarter of May and the first quarter of June) 

 the Fishery Bay sample for that quarter and sub- 

 sequent samples from the lake lead me to assume 

 that spawning fish, though never numerous, were 

 present here at that time also. 



Spent fish were first observed in the first quarter 



Table 20. — State of ovaries of mature II-group and older gizzard shad captured in Fishery Bay and in Lake Erie in May, 



June, and July 1954-55 



[Given as percentages of fish in the categories "not spawning," "spawning," and "spent"] 



' No females captured May 24-31. 

 GIZZARD SHAD IN WESTERN LAKE ERIE 



415 



