246 



Fishery Bulletin 92(2). 1994 



Central flat area 



Males 



MIm 



i { 



Females 



\ H H i * 



1984 1965 1986 1967 1986 1989 1990 



Pinnacles 



1 



I 



1 



1 1 1 1 1 1 1 



1984 1985 1986 1987 1988 1989 1990 



it** 



i r - 



— I 1 1 1 



1987 1988 1989 1990 



Year 



Figure 5 



Mean length (±2 standard errors) of orange roughy (Hoplostethus atlanticus) by year 

 in the three main spawning areas: Central Flat, Pinnacles, and Westpac Bank. 



Flat and Pinnacles. The increase in ripe fish towards 

 the end of the 1990 survey was due largely to sam- 

 pling the Westpac Bank at this time. 



The onset of spawning, defined as the first date 

 on which 20% of fish sampled were spent (after 

 Pankhurst, 1988), has been relatively consistent 

 over the years (Table 11). Actual dates based on fe- 

 males have ranged from 9 July to 16 July. 



Size at maturity 



Mean lengths at maturity for males and females by 

 year are given in Table 12. There is a significant 

 trend of decreasing mean size for males (linear re- 

 gression F-test, P<0.05), but there is no consistent 

 trend for females. 



Feeding 



Data on frequency of occurrence of broad taxonomic 

 prey groups from 1984 to 1990 showed the most 



common prey were natant decapod crustaceans and 

 fish (Table 13). 



The main groups that could be identified were 

 macrourids (small species of Coelorinchus and 

 Coryphaenoides) and myctophids (species of 

 Lampanyctus and Lampanyctodes). Natant decapod 

 crustacean prey were mainly species in the genera 

 Pasiphaea, Sergestes, Oplophorus, and Acan- 

 thephyra. Squids and amphipods were also frequent 

 prey. 



Discussion 



Orange roughy on the Challenger Plateau were over- 

 exploited in the late 1980s. Research trawl survey 

 and commercial catch and effort data show similar 

 changes in distribution and abundance. There was 

 a marked contraction in the area of high catch rates, 



