22 
Fishery Bulletin 99(1 ) 
Figure 4 
Final oocyte maturation (FOM) in late-developing cobia ovaries. (A) Early 
stage of FOM showing initial lipid coalescence (L). Scale bar = 0.1 mm. 
(B) More advanced stage of FOM. Lipids have coalesced to form a single 
large droplet (L) and the nucleus (N) is beginning to migrate to the 
periphery of the oocyte. Scale bar = 0.1 mm. (C) Oocytes undergoing syn- 
chronous FOM. Scale bar = 0.2 mm. 
>500 g. The observed relationship was linear, 
and the best fit equation as judged by the 
least squares criterion was mL = -8.54 + 
0.96g (r 2 =0.978). The relationship indicated 
that cobia ovaries from 500 to 1600 g were 
less dense than seawater and that ovary den- 
sity remained constant over that range. 
Batch fecundity 
Batch fecundity estimates were compared for 
samples from 11 females with ovaries fixed in 
GF, for 40 females with ovaries fixed in NBF, 
and for 26 ovarian samples examined by the 
histological method and where oocytes were 
undergoing FOM from April through Septem- 
ber. Samples of this type were limited; there- 
fore we combined observations from SEUS, 
EGOM, and NCGOM. Fecundity estimates 
for individual fish varied widely; however, 
there was no significant difference in mean 
estimates among the three methods (Mann- 
Whitney [/-test, P>0. 05; Table 7). Mean batch 
fecundity ranged from 377,000 ±64,500 eggs 
(CV=2.677) with the histological method to 
1,980,500 ±1,598,500 eggs (CV=0.875) with 
the GF method. 
Batch fecundity estimates for all three 
methods showed substantial variation. A Kol- 
mogorov-Smirnov test of normality showed 
that batch fecundity was normally distribut- 
ed for NBF samples (40 df, P>0.05) and FOM 
samples (26 df, P>0.05) and could be ana- 
lyzed by using parametric statistics. Batch 
fecundity determined with the GF method 
was not used for further analyses owing to 
the small sample size. Regression analysis 
showed a significant, positive relationship be- 
tween batch fecundity (BF) and FL (P=0.021, 
r 2 =0.132) and BF and ovary-free body weight 
(OFBW; P=0.016, r 2 =0.143) for NBF samples. 
The relationship between BF and OFBW for 
NBF samples (Fig. 6A) was described by 
BF = OFBW 1 717 -36.813. 
There was no significant relationship be- 
tween BF and FL (P=0.105) or OFBW 
(P=0.097) for FOM samples. The relationship 
between BF and OFBW for FOM samples 
(Fig. 6B) was described by 
BF = 19.29 x OFBW + 1,113,713 
[r 2 =0.11, P=0.097], 
Mean batch fecundity, as determined from 
the NBF method, averaged 854,100 ±166,200 
eggs and ranged from 247,100 ±204,400 eggs 
in August (n= 2) to 923,000 ±237,700 eggs 
in May (n= 26). Mean batch fecundity, as 
