70 
Fishery Bulletin 107(1 ) 
1990 
Year class Year class 
Figure 9 
Observed numbers of individuals of each year class of western blue groper ( Achoerodus gouldii) in commercial gillnet 
samples obtained from the south coast of Western Australia between August and July in 2004-5, 2005-6, and 2006-7 
(histograms). Dotted lines represent the expected numbers of fish of each year class, as determined from a relative 
abundance analysis and were fitted to the observed frequency of abundance of fish in each year class in each of the three 
12-month periods. Variable recruitment was assumed for creating the dotted lines; 1972, 1980, 1983, and 1990 (denoted 
with arrows) were years of particularly strong recruitment. 
1.0 
r k 
0.8 
0.6 
oc 
Q_ 
c n 
0.4 
\ — — Female 
0.2 
- \ 
0 
— _ Male 
0 0.2 0.4 0.6 0.8 
Fishing mortality (per year) 
Figure 10 
The effect of fishing mortality ( F ) on the spawning 
potential ratio (SPR) for females and males of the 
western blue groper (Achoerodus gouldii), calculated 
by using the age of 15 years (the age currently cal- 
culated to represent full recruitment into the com- 
mercial gillnet fishery). 
of A. gouldii, irrespective of their sex, the variation 
in the lengths at age became very marked after ~25 
years (Fig. 4A). This variation reflects a marked upward 
divergence in the growth of males from that of females 
among the larger and older fish (Fig. 4B). The fact that 
the von Bertalanffy growth parameters for green and 
blue fish were very similar to those derived for females 
and males, respectively, in data for the same subset of 
individuals, demonstrates that growth curves based on 
color provide a good proxy for those based on sex alone 
(Fig. 4, B and C). 
A substantial number of the larger and older fish 
obtained during the study had been eviscerated be- 
fore we collected them and therefore sex could not be 
determined and used to provide length-at-age data for 
constructing growth curves for the two sexes. Conse- 
quently, some caution must be exercised in drawing 
conclusions for the upper end of the growth curves con- 
structed using fish of known sex, or for the same fish 
after separation by color. The AIC test used on the 
above subsample demonstrated that a combination of 
length and color provided a better indication of the sex 
