104 
Fishery Bulletin 95( I ), 1997 
The numbers of hoki observed in the May surveys 
(26.8 million [1992], 24.4 million [1993]) were con- 
siderably less than in the December surveys (38.6 
million [1991], 34.8 million [1992]) (Fig. 2). In addi- 
tion, the length-frequency histograms show a decline 
in bimodality in the adult part of the distribution 
between May and December (Fig. 2). In December 
1991, 56% of females over 50 cm were in strata west 
of 170°E. In May 1992, 59% were west of this line. In 
December 1992, 49% were in the west and in May 
1993, 68% were in the west. 
There were 541 fish in the histological sample in 
May 1992 and 1,136 fish in the sample in May 1993 
(Table 1). In 1992, stratum 1 (300-600 m depth at 
Puysegur) was not sampled and female fish from 
every second station in the other strata were 
sampled. In 1993 female fish from every station in 
all 15 strata were sampled. 
Ageing 
There were very few young fish in most strata, and 
limited numbers of fish in the 1986 cohort, which 
appeared as age-6 fish in 1992 and as age-7 fish in 
1993. Although fish were aged to a maximum age of 
19 years, we combined them into a group of age 10 + 
and above. The ageing data were also used to de- 
velop age-length keys. 
