4 ! 8 
Fishery Bulletin 96(3), 1998 
Table 1 
Effect of location of gag tissue samples on hydrated oocyte counts per unit of weight (g). Locations are anterior (1), mid (2), and 
posterior (3) of ovarian lobes. Analysis of variance indicates significance of location within a lobe for number of hydrated oocytes 
per gram of tissue. SS = sum of squares; MS = mean square. 
Positions 
of sample 
in ovary 
Lobe 1 
Lobe 2 
Both lobes 
X 
SD 
n 
X 
SD 
n 
X 
SD 
n 
1 
918 
92 
6 
951 
107 
6 
934 
97 
12 
2 
914 
81 
6 
932 
180 
6 
923 
133 
12 
3 
875 
140 
6 
931 
144 
6 
903 
138 
12 
Total 
902 
103 
18 
938 
138 
18 
920 
121 
(36) 
Source 
df 
SS 
MS 
F 
P>F 
Lobe 
1 
11,460.23 
11,460.23 
0.74 
0.40 
Region 
1 
6018.21 
6018.21 
0.39 
0.54 
Interaction 
1 
803.81 
803.81 
0.05 
0.82 
Error 
32 
496,556.28 
15,517.38 
Total 
35 
514,838.52 
n= 5,450 oocytes; range=0.04 to 1.20 mm) nor hydrated 
oocyte counts per gram (Table 1) differed significantly 
among ovarian regions, we randomly sampled one re- 
gion on each ovary to determine maturation stage, 
maximum oocyte diameter, and fecundity. Gag hydrated 
ovaries contained a mean of 920 hydrated oocytes per 
gram of ovarian tissue (range=875 to 951, SD=121, n = 6 
fish) (Table 1). The distinct clusters of oocytes of vari- 
ous diameters, including stage-1 and stage-2 oocytes, 
in gag ovaries throughout the spawning season indi- 
cated that gag is a multiple and indeterminate spawner. 
Gag ovaries were usually much larger than testes 
(mean ovary weight=46.5 g, mean testes weight= 
26.7 g). The largest ovary and testes weighed 1.60 kg and 
120.0 g, respectively. Wormlike parasites were common 
in ovaries. Histological appearance of gag ovaries (non- 
transitional) was Typical for marine teleosts (Wallace 
and Selman, 1981), except that some functional females 
appeared to contain crypts of spermatogonia. 
An unusual condition observed in three females 
(stage 1 or 2) was the presence of a 5-mm thick lin- 
ing (around the inside perimeter of the abdominal 
cavity) of compressed, hydrated oocytes apparently 
caused by rupture of the ovarian wall. These three 
fish were caught after the peak spawning season 
(May 1994, June 1993, and July 1994). We found no 
such lining in any of the other specimens examined. 
We found that sex determinations and maturation- 
stage determinations were made with variable accu- 
racy in fresh and histologically prepared samples. 
The two examinations gave the same results 89% of 
the time for sex determination but only 70% of the 
time for stage determination. The main reason for 
this poor comparison of the two staging techniques 
was that we could not accurately identify spent fe- 
males (with POFs) or transitional males from fresh 
samples, because POFs and sperm could only be posi- 
tively identified histologically. 
Sex ratio 
Sex ratio from all random samples (pooling data for 
1991-93) was approximately 49 females: 1 male. Our 
emphasis on collecting gag with dark ventral pig- 
mentation in 1994 clearly increased the overall per- 
centage of males: 52.5% of all males for 1991-94 (n= 59) 
were collected in 1994. Percentage of males was 1.9% 
in 1991-93 and 4.7% in 1994. Females were collected 
year round and males were collected in all months but 
January. Possible early transitional male gag (n= 3) 
were found only during February, April, and May. 
Ventral pigmentation 
Gag with heavy ventral pigmentation (n-6 2) were 
collected in all months but January and March in 
waters >41 m (Table 2). Sex determination (h= 60) 
showed that 5% were females (Fig. 2A), 3.3% were pos- 
sible transitional males (Fig. 2, B and C), and 91.7% 
were males (Fig. 2D). Sex and stage were homogeneous 
among six gonad regions in all heavily pigmented fish. 
Spawning period 
Winter-spring spawning, indicated by the presence 
of hydrated oocytes, was corroborated by GSI (Fig. 
