(H)N()VKR: I'ATrFRNS IN FFCrNDITY ( iK All.AN'I'K ' SILVKKSIDK 



(nearest millimeter) and gonad-free body weight 

 (nearest 0.1 g). The gonadosomatic index (GSI) was 

 calculated by expressing gonad weight as a percent- 

 age of total weight (Snyder 1983). 



Preliminary microscopic examination of ripe 

 ovaries from collections during the breeding season 

 revealed three general egg types that corresponded 

 with modes found in frequency distributions of egg 

 diameters from ripe females (see Figure 1 below). 

 The three egg classifications were defined by both 

 size and appearance of the ova as follows: 



Immature ova: spherical, 0.05-0.60 mm in diameter. 

 The smaller ova in this group (0.05-0.15 mm) were 

 primary oocytes with a clear cytoplasm and large 

 nucleus. The larger ova (0.15-0.60 mm) were opa- 

 que and white. These ova formed one continuous 

 mode in the frequency distribution so they were 

 considered as one group (Fig. lA, B). 



Maturing ova: spherical, 0.6-1.0 mm in diameter, 

 yellowish, and translucent. 



Ripe ova: spherical, hydrated, 1.0-1.2 mm in dia- 

 meter, hyaline and golden, with visible gelatinous 

 threads coiled around the egg. 



To confirm that these classifications represented 

 distinct groups of synchronously maturing ova, I 

 measured the diameter (random axis) of about 150 

 eggs, randomly subsampled from each of several 

 females. In females captured during the breeding 

 season, there were two distinct modes: the most ad- 

 vanced mode represented maturing eggs, and the 

 other mode represented immature eggs (Fig. 1 A). In 

 females with ripe eggs, three modes in egg fre- 

 quency were apparent: the most advanced group 

 represented ripe eggs, the intermediate mode 

 represented the next batch of maturing eggs, and 

 the remaining mode consisted of immature eggs 

 (Fig. IB). 



"Batch fecundity" was defined as the number of 

 mature eggs in the most advanced size class, and 

 presumably represented the number of eggs spawn- 

 ed at one time. As illustrated in Figure 1, the most 

 advanced size class of maturing or ripe eggs was 

 clearly distinguishable from, and did not overlap in 

 size with, the immature ova. In ripe females, batch 

 fecundity represented the number of hydrated eggs 

 (e.g., the most advanced mode in Figure IB). In 

 nonripe females, batch fecundity represented the 

 number of maturing eggs (e.g., the most advanced 

 mode in Figure lA). "Recruitment fecundity" was 

 defined as the number of eggs smaller than the most 

 advanced egg batch. Recruitment fecundity presum- 

 ably represented the number of ova from which addi- 



Immature 



Maturing 



n=l45 



.3 .5 .7 .9 1.1 



EGG DIAMETER(mm) 



Figure 1.- Frequency vs. egg size (diameter in 0.1 mm interN-als) 

 for ova randomly sampled from two female Menidia menidia col- 

 lected 6 June in Essex Bay, MA. A) A nonripe female in which 

 there is a large number of immature eggs and a clearly definable 

 mode of maturing eggs. B) A ripe female in which there is a large 

 mode representing immature eggs, an intermediate mode of eggs 

 beginning to mature, and an advanced mode of ripe eggs. 



tional batches of mature eggs could potentially arise. 

 In ripe fish containing three modes (e.g.. Fig. IB), 

 the intermediate mode of eggs in early maturation 

 was included with the immature eggs as part of the 

 recruitment egg group because these overlapped in 

 size with, and were difficult to separate from, im- 

 mature ova. The minimum size of recruitment eggs 

 was 0.05 mm in diameter. "Total fecundity" was de- 

 fined as the sum of recruitment and batch fecun- 

 dity. 



Batch, recruitment, and total fecundity of indivi- 

 dual females were estimated gravimetrically in the 

 following manner. A cross section comprising 

 10-20% of the total ovarian weight was cut from a 

 randomly chosen portion of one ovary. Both the sub- 



333 



