Zeldis et al. : An estimate of biomass of Hoplostethus atlanticus 
589 
20 June, in agreement with the reduc- 
tion actually observed by that date 
(Fig. 8). 
It was likely, however, that early 
departures caused spent fish to be un- 
der-represented in the trawl samples 
toward the end of the R t time series. 
The abundance of young planktonic 
eggs in the central strata was reduced 
by more than half between subsurvey 
3 (mid-date 30 June) and subsurvey 4 
(mid-date 4 July; Table 5). Because 
subsurvey 3 was done when the spent 
proportion was virtually zero (Fig. 7), 
this reduction of planktonic egg abun- 
dance implied that about half of the 
fish had ceased spawning by 4 July 
(assuming that the spawning rate of 
remaining active fish was constant). 
However, only 0.11 of fish were re- 
corded as spent on 4 July (Table 2; Fig. 
7); thus this proportion appeared to be 
underestimated by 0.50 - 0.11 = 0.39 
on 4 July. Commercial catch rates on 
Ritchie Hill (Fig. 9) also decreased con- 
siderably during the last week of June 
and first week of July, suggesting that 
fish abundance in the trawling area 
had declined. 
To account for the error in D that the 
underestimation of spent fish would cause, 
the proportion of active fish on 4 July was 
adjusted downward to 0.89 - 0.39 = 0.50 
to estimate a new D value of 1,106 eggs/ 
(kg x day) (assuming a linear decline be- 
tween 20 June and 4 July; Fig. 8; Table 4). 
This adjustment resulted in a period of 25 
days for an average fish to spawn com- 
pletely, which is not greatly different from 
the 19 days estimated from the time lag 
between ovulated and spent fish propor- 
tions (shown above). This re-estimation of 
D, to account for turnover, had propor- 
tional effects on B sp ^ and B rec (Table 4). 
Figure 6 
The daily production of eggs as a function of age, calculated by dividing the 
curved trajectory egg-abundance estimates (Table 3) by the duration of each 
age group and by plotting against the mean age of eggs in each age group 
( Appendix 3 ). The line was fitted by using N 0 and Z ( Appendix 3 ), calculated 
with only egg stages < 7 to 10 (crosses). Egg stages 11 to 15 (triangles), were 
not used in the production estimate because they were subject to advection 
out of the survey area (see text). 
5 Jun 10 Jun 15 Jun 20 Jun 25 Jun 30 Jun 5 Jul 10 Jul 15 Ju I 
Date 
Figure 7 
Discussion 
Time series of macroscopic ovarian stages over the survey period. Each 
point is the proportion of females at that stage, as a proportion of all 
mature females, averaged for all trawls on that date. Only trawls on 
Ritchie and North Hills were used (Table 2). 
This study has used the DFRM to estimate 
the biomass of recruited individuals in the 
mid-east coast orange roughy stock, by 
combining estimates of daily egg production rate, the 
rate of fecundity reduction, and the proportion of the 
stock that was spawning females. In this section vari- 
ous sources of error are discussed, and the methods 
and results of this study are compared with those of 
other studies. Also, a brief description is made of how 
the results from this survey have been used in as- 
sessing the stock. 
