Abstract. — To enable accurate 

 ageing of orange roughy, Hoplo- 

 stethus atlanticus, eggs caught in 

 egg production surveys for biomass 

 estimation, eggs were cultured at 

 four experimental temperatures, 

 and the results were fitted to a re- 

 gression model of age-at-stage for 

 the culture temperatures used. 

 Water column ascent and descent 

 rates of early stage eggs were esti- 

 mated by using a combination of 

 experimental observation and 

 theory. Young eggs ascended rap- 

 idly (>300 m-day" 1 ), middle-aged 

 eggs approached neutral buoyancy, 

 and the oldest eggs sank rapidly. 

 The ascent and development rate 

 results were combined with data on 

 water column thermal structure to 

 produce a "thermal history model" 

 of orange roughy egg development, 

 enabling the thermal history of 

 eggs to be considered in ageing 

 them as they rise through the 

 stratified layers into the mixed 

 layer. The ascent and descent rates 

 and modelled depth-at-stage were 

 compared with field data on depth 

 distributions of stages from MOC- 

 NESS tows over the North Chatham 

 Rise, New Zealand. For the domi- 

 nant and subdominant egg stages, 

 predictions closely matched the 

 field results. The model therefore 

 provided a robust method for age- 

 ing eggs caught in an egg produc- 

 tion survey for orange roughy. The 

 ontogenetic pattern of buoyancy in 

 the eggs may partially explain the 

 distributions of early juvenile or- 

 ange roughy on the North Chatham 

 Rise. The bias in predicted egg 

 ages, caused by the assumption 

 that temperature of development 

 was constant for eggs below the 

 mixed layer, was shown to be impor- 

 tant in egg production estimation. 



Ascent rates, vertical distribution, 

 and a thermal history model of 

 development of orange roughy, 

 Hoplostethus atlanticus, 

 eggs in the water column 



John R. Zeldis 

 Paul. J. Grimes 

 Jonathan K. V. Ingerson 



MAF Fisheries, Greta Point 



Box 297, Wellington, New Zealand 



Manuscript accepted 9 September 1994. 

 Fishery Bulletin 93:373-385 1 1995). 



Orange roughy, Hoplostethus atlan- 

 ticus (family: Trachichthyidae), are 

 slow-growing, deep water (700- 

 1,500 m) fish that have been found 

 in large quantities mainly in New 

 Zealand 1 and Australia (Kailola et 

 al., 1993). In winter, orange roughy 

 aggregate to spawn near banks, pin- 

 nacles, and canyons, and these ag- 

 gregations attract large amounts of 

 fishing effort. Fisheries for orange 

 roughy began in 1978—79 in New 

 Zealand. The largest fishery is on 

 the Chatham Rise (Fig. 1 ) where re- 

 ported catches reached 13,800 met- 

 ric tons (t) in 1992-93. l The stock 

 assessment for the Chatham Rise 

 fishery is based on stock reduction 

 modelling of biomass in which com- 

 mercial catches, biological param- 

 eters, and time series of relative bio- 

 mass estimates from research trawl 

 surveys are used. The next largest 

 fishery (with reported catches of 

 9,128 1 in 1992-93 [Field et al., 1994] ) 

 is off the east coast of New Zealand, 

 and a large proportion of the catch 

 is taken from the winter spawning 

 aggregations on the Ritchie Bank 

 (Fig. 1). The stock reduction analy- 

 sis for this fishery prior to 1993-94 

 was based primarily on catch-per- 

 unit-of-effort data which were 

 highly uncertain and made the bio- 

 mass estimate very uncertain 

 (Field et al., 1994). Consequently, 



there was a need for a reliable, ab- 

 solute biomass estimate. 



After reviewing aspects of the bi- 

 ology of orange roughy planktonic 

 eggs and adult fecundity, Zeldis 

 (1993) concluded that both the an- 

 nual egg production method ( AEPM; 

 Saville, 1964 ) and the daily fecundity 

 reduction method (DFRM; Lo et al., 

 1992) would be feasible for the esti- 

 mation of absolute spawning bio- 

 mass of orange roughy on the 

 Ritchie Bank. Both the AEPM and 

 the DFRM were applied during an 

 egg production survey of Ritchie 

 Bank orange roughy biomass in 

 June-July 1993. 2 



Because egg production methods 

 rely on calculating egg production 

 rates from estimates of abundance- 

 at-age of planktonic eggs, a means 

 of ageing the eggs is required. This 

 is done by determining ages of iden- 

 tifiable morphological stages of egg 



1 Annala, J. (Compiler) 1994. Report from 

 the Special Fishery Assessment Plenary, 

 17 August 1994: stock assessments and 

 yield estimates for ORH 3B, 24 p. Unpubl. 

 rept. held in MAF Fisheries, Greta Point 

 library, Wellington, N.Z. 



2 Zeldis, J., R. I. C. C. Francis, J. K. V. 

 Ingerson, M. Clark, and P. J. Grimes. 1994. 

 A daily fecundity reduction method esti- 

 mate of Ritchie Bank and east coast North 

 Island orange roughy biomass. Draft New 

 Zealand Fisheries Assessment Research 

 Document. MAF Fisheries, Greta Point, 

 Wellington, N.Z. 



373 



