Farley and Davis: Reproductive dynamics of Thunnus maccoyii 
225 
In the laboratory, a core subsample was taken from 
each ovary while frozen. Subsamples were fixed in 
10% buffered formalin, embedded in paraffin, and 
standard sections were prepared for histological ex- 
amination (cut to 6 pm and stained with Harris’s 
haematoxylin and eosin). All ovaries collected on the 
spawning grounds and 53 of the ovaries collected 
from the Southern Ocean were processed in this way. 
Ovaries were thawed, trimmed of fat, blotted dry, 
and weighed to the nearest g. The mean diameter 
(random axis to the nearest pm) of five oocytes from 
the most advanced group of oocytes (MAGO) was 
determined for each ovary with a video coordinate 
digitizer connected to an Ikegami video camera 
mounted on a stereomicroscope at 50x magnification. 
Gonad index was calculated as 
GI = W/L 3 x IQ 4 , 
where W = gonad weight in g; and 
L = length to caudal fork in cm (Kikawa, 
1964a; Shingu, 1970). 
All ovaries were examined for residual hydrated oo- 
cytes as evidence of spawning activity. 
Histological classification 
and spawning frequency 
Because many of the ovaries were frozen before a 
subsample could be removed, it was necessary to as- 
sess whether histological sections prepared from fro- 
zen ovarian material could be used to determine go- 
nad stage and spawning activity. A comparison was 
made between histological sections prepared from 
200 ovaries which were subsampled both before and 
after freezing. Although considerable cell destruction 
was observed in the histological sections from frozen 
tissue, oocytes, atretic oocytes, and postovulatory fol- 
licles were still distinguishable and could be classified. 
Histological sections were classified with criteria 
similar to those developed for northern anchovy, 
Engraulis mordax (Hunter and Goldberg, 1980; 
Hunter and Macewicz, 1980, 1985a, 1985b), skipjack 
tuna, Katsuwonus pelamis (Hunter et al., 1986) and 
yellowfin tuna, Thunnus albacares (Schaefer, 1996). 
Each ovary was staged by the most advanced group 
of oocytes present into one of five classes: unyolked, 
early yolked, advanced yolked (Fig. 2A), migratory 
nucleus (Fig. 2B), or hydrated (Fig. 2C). In order to 
determine the relation between atresia (resorption 
of oocytes) and spawning, ovaries were also classi- 
fied by the level of a and (3 stage of atresia in ad- 
vanced yolked oocytes. During the a stage of atre- 
sia, yolk resorption takes place (Fig. 2D). During the 
P stage of atresia, the remaining granulosa and the- 
cal cells are reorganized and resorbed leaving a com- 
pact structure containing several intercellular vacu- 
oles. 
Ovaries were classified into one of the following 
five atretic states: 
0 no a atresia present, but advanced yolked oocytes 
are; 
1 <10% of advanced yolked oocytes are in the a stage 
of atresia; 
2 10-50% of advanced yolked oocytes are in the a 
stage of atresia; 
3 >50% of advanced yolked oocytes are in the a stage 
of atresia; 
4 100% of advanced yolked oocytes are in the a stage 
of atresia, or no advanced yolked oocytes are present 
but oocytes in the (3 stage of atresia are present. 
Spawning frequency was determined by the post- 
ovulatory follicle method of Hunter and Macewicz 
( 1985a). This method uses the incidence of females 
with postovulatory follicles less than 24 hours old to 
define the fraction of the population spawning. Be- 
cause the time of capture was not available, we could 
not assign ages to postovulatory follicles based on 
the estimated time of death relative to the estimated 
time of spawning. Postovulatory follicles were, there- 
fore, aged according to their state of degeneration 
with criteria developed for skipjack tuna, yellowfin 
tuna, and bigeye tuna, Thunnus obesus (Hunter et 
al., 1986; McPherson, 1988; Nikaido et al., 1991; 
Schaefer, 1996), all of which spawn in water tempera- 
tures above 24°C and resorb their postovulatory fol- 
licles within 24 hours of spawning. We assumed that 
southern bluefin tuna resorb postovulatory follicles 
at the same rate as other tropical spawning tuna 
because water temperature appears to be the domi- 
nant factor governing resorption rates ( Fitzhugh and 
Hettler, 1995). We recorded postovulatory follicles in 
histological sections according to methods of Hunter 
and Macewicz (1985a), Hunter et al. (1986) and 
Schaefer ( 1996). Postovulatory follicles were classified 
as either 0 (absent), 1 (new) (Fig. 2E), 2 (less than 12 
hours old), 3 (12 to 24 hours old), or 4 (indistinguish- 
able owing to tissue decay). The incidence of females 
with postovulatory follicles of any age was used to de- 
termine spawning frequency. 
Females were classified into one of four spawning 
states depending on the oocytes, atretic state, and 
postovulatory follicle class present. 
1 Immature: Ovaries contain no advanced yolked 
oocytes or advanced yolked oocytes in the a stage 
of atresia. No residual hydrated oocytes present. 
