Livingston et al.: Estimating the annual proportion of nonspawning adult Macruronus novaezelandiae 
105 
Table 1 
Percentages of each histological stage observed in female hoki sampled from each survey. 
Histological stage 
Dec 1991 
May 1992 
Dec 1992 
May 1993 
Chromatin nucleolar 
0 
0 
0 
0.1 
Perinucleolar 
100 
45.4 
100 
39.7 
Yolk vesicle 
0 
27.1 
0 
13.6 
Vitellogenic 
0 
27.5 
0 
43.3 
Ripe 
0 
0 
0 
3.3 
Number in sample 
452 
541 
1,039 
1,136 
Table 2 
Numbers of hoki monthly samples showing ovarian 
development on 
the Chatham Rise and west coast spawning grounds. 
Region and 
Histological stage 
month 
Perinucleolar 
Yolk vesicle Vitellogenic 
Ripe 
Chatham Rise 
January 
22 
— 
— 
— 
F ebruary 
10 
— 
— 
— 
March 
no sample 
April 
6 
9 
1 
- 
May 
10 
9 
1 
— 
West coast 
June 
4 
3 
13 
— 
July 
— 
— 
10 
5 
August 
— 
— 
9 
5 
Histology 
Monthly samples of hoki from the Cha- 
tham Rise and west coast of the South Is- 
land showed little change in oocyte stage 
in January and February, all being classi- 
fied as perinucleolar (Table 2). In April, 
May, and June, larger oocytes of the yolk 
vesicle and vitellogenic yolk stages were 
evident. By July and August, females 
sampled on the west coast of the South Is- 
land were vitellogenic or ripe and had hya- 
line oocytes (Table 2). Oocytes clearly devel- 
oped as a synchronous group, evidenced by 
the separation in size of the developing 
clutch from the reserve fund of chromatin 
nucleolar and perinucleolar oocytes ( Fig. 3 ). 
During the December trawl surveys, 
most ovaries contained oocytes that could 
be classified as late perinucleolar (Fig. 4). 
By May, however, a significant change in oocyte stage 
had occurred; many ovaries contained cortical alveoli 
organized into a ring structure and showed increased 
oocyte size and oil droplets forming around the 
nucleus (Fig. 5). Because the oocyte stage observed 
in summer appeared to be a natural holding point in 
development, we classified a fish with such a stage 
as perinucleolar. When we saw the same develop- 
ment in fish in the autumn surveys they were classi- 
fied as nonspawners. Only those fish with a prolif- 
eration of cortical alveoli and oil droplets that had 
begun to form around the nucleus were classified as 
being at or beyond the yolk vesicle stage and there- 
fore counted as spawners for the coming season. 
Table 1 shows the proportion of fish at each stage 
in each of the surveys. In both December samples, most 
or all fish were classified as perinucleolar. In contrast, 
in the May surveys, only 45.4% and 39.7% were in the 
perinucleolar stages in 1992 and 1993 respectively. 
For fish caught during the trawl survey in May 
1993, stage of development of the ovary of each fish 
in the histological sample was also evaluated mac- 
roscopically . The number of fish at each histological 
stage and at each stage of macroscopic gonad devel- 
opment are presented in Table 3. In total, of the 237 
ovaries classified as maturing, only two were classi- 
fied histologically as nonspawners. However, of the 
899 ovaries classified as resting, 450 were classified 
histologically as nonspawners and 449 as pre- 
spawners. This finding confirms that physical devel- 
opment for spawning begins before it is apparent 
macroscopically in the ovaries and reinforces the re- 
quirement for histological methods of analysis. 
The proportion of prespawners in each stratum was 
estimated from the aged histological samples from 
the May surveys (Table 4). If there were not at least 
two fish in a stratum of a particular age, the propor- 
tion was not estimated. Although there were many 
age-stratum combinations where the proportion of 
prespawners in a particular age class could not be 
estimated (mainly for the young fish and for the 1986 
cohort), these were not generally found in strata that 
supported the greatest numbers of hoki of that age 
class (Table 4). 
