578 
Fishery Bulletin 95(3), 1997 
survey area was divided by the daily fecundity/kg of 
the females: 
B spf - Af 0 /( 1,000.0), 
where B ^ = biomass of spawning females (tons); 
N 0 = daily egg production, (eggs/day); 
D = mean daily fecundity (eggs/(kg x day)) 
for mature fish; and the factor 1,000 
converts kg to tons. 
The recruited biomass, B rec (defined as the biomass 
of fish of length >32 cm) was calculated from the bio- 
mass of Ritchie Bank spawning females, B gp ^ as 
B rec = B sp fS, 
where the scalar (S) was an estimate of the ratio B / 
B sp ^ This ratio allows for recruited females that did 
not spawn, females that did spawn but were <32 cm, 
as well as the sex ratio. 
To estimate the parameters of the above biomass 
model, the data analysis deals with three distinct 
data sets: 
• daily planktonic egg production; 
• daily fecundity per female weight; and 
• proportions spawning, recruited, and female. 
Daily planktonic egg production 
Survey design The timing of plankton sampling 
coincided with the period when orange roughy fe- 
males on the Ritchie Bank were in late maturation 
or spawning stages (mid-June to the end of the first 
week of July; Pankhurst, 1988). The location of plank- 
ton sampling was determined from research trawl 
catch rates (Fincham et al. 4 ), which showed adult 
biomass to be highly aggregated on Ritchie Hill 5 (Fig. 
2, A and B) at the northern end of Ritchie Bank. 
Ritchie Hill catch rates accounted for 84% of the rela- 
tive orange roughy biomass over the Ritchie Bank 
survey area in July 1986 (Fincham et al. 4 ). In plank- 
ton sampling during the spawning season on Ritchie 
Bank in early July 1986, 6 orange roughy eggs were 
4 Fincham, D. J., D. A. Banks, and P. J. McMillan. 1987. 
Orange roughy trawl survey, Tolaga Bay to Cape Turnagain, 14 
June to 11 July 1976: cruise report. Fisheries Research Divi- 
sion Internal Report 60. NIWA, Greta Point, Wellington, New 
Zealand, 38 p. 
5 The names “Ritchie Bank,” “Ritchie Hill,” and “North Hill” used 
in this paper refer to the features called “Ritchie Ridge,” 
“Calyptogena Bank,” and “Pantin Bank,” respectively, in the 
following: Arron, E. S., and K. B. Lewis. 1992. Mahia, 2nd ed. 
N.Z. Oceanogr. Inst. Chart, Coastal Series, 1:200,000. NIWA 
Greta Point, Wellington, New Zealand. 
caught only at stations near Ritchie Hill, and samples 
taken >20 km away contained no eggs, indicating that 
eggs were aggregated near the spawners (Zeldis, 
1993) and that plankton sampling would need to be 
highly concentrated near Ritchie Hill. 
During their first 36 hours of development, orange 
roughy eggs ascend the water column at 300-350 m/ 
day from a spawning depth of about 850 m (Zeldis et 
al., 1995). Geostrophic currents over Ritchie Bank 
during July 1986 6 were to the south and averaged 
about 12 cm/sec between 700 m and 250 m; these 
currents would displace these eggs at least 15.5 km 
to the south of Ritchie Hill by the time the eggs had 
reached 36 h of development (this is a minimum es- 
timate because there was probably some residual flow 
at the postulated level of no motion at 700 m). Con- 
sidering that drift would probably vary in direction 
and speed but would lie predominantly along 
isobaths, we designed the survey area with eight 
strata, elongated alongshore and arranged symmetri- 
cally around a central stratum centred over the top 
of Ritchie Hill (Fig. 2B). This central stratum (10.0 
x 7.6 km) was about the size of the area of high fish 
density observed during a trawl survey done in the 
area in June— July 1987 (Grimes 7 ). The middle layer 
of four strata surrounding the central stratum had 
outer boundaries of 18.5 xl3.7 km. These dimensions 
were chosen to approximate the distance at which 
catch rates of 1-day-old orange roughy eggs on the 
St. Helen’s seamount in eastern Tasmania were re- 
duced to half (9.3 km from the spawners [Koslow 8 ]). 
The outer layer of four strata had an outer boundary 
40.0 x 30.0 km long, to allow for maximal drift of 36- 
h-old eggs. 
The St. Helen’s data were used to estimate opti- 
mal allocation of stations to Ritchie Hill strata. The 
St. Helen’s data were collected over an entire spawn- 
ing season, in six separate subsurveys, each of which 
had two spatial strata. The counts in each St. Helen’s 
subsurvey and stratum were standardized such that 
their means equalled the mean across all of the 
subsurveys for each stratum. This procedure removed 
the within-season variation in mean egg density in 
each stratum. These standardized data were then 
laid over the Ritchie Bank stratum layers (central, 
middle, and outer), and mean egg densities (M) were 
estimated for each stratum layer, j. To allocate sta- 
6 Zeldis, J. R. 1986. Cruise report J08/86 (second half). MAF 
Fisheries unpublished cruise report held in NIWA Library, Greta 
Point, Wellington, New Zealand, 7 p. 
7 Grimes, P. 1987. NIWA, P.O. Box, 14-901, Kilbirnie, 
Wellington, New Zealand. Unpubl. data. 
8 Koslow, T. 1992. CSIRO Division of Fisheries, GPO Box 1538, 
Hobart, Tasmania 7001, Australia. Personal commun. 
