68 
Fishery Bulletin 107(1) 
0 200 400 600 800 1000 
°'jj / o. "tr A "vr Tr % J ,V 
A '<£? H- A> Ar ’‘o v y ' v o 4/ 4s> 
Total length (mm) Age category (years) 
Figure 8 
Percent frequencies of occurrence, in sequential 50-mm length classes and 5-year 
age categories, of the western blue groper (Achoerodus gouldii) with (A) testes, i.e., 
males (gray histograms), and ( B ) blue coloration (gray histograms). The logistic 
curves (solid line) and their 95% confidence limits (dotted lines) were derived from 
the probability that a fish at a given length is male in (A) and blue in (B). Sample 
sizes are shown above each histogram. 
The values for the AIC showed that the dichotomous 
factor, color, is not as good a predictor as the continuous 
variable, length, that an individual is male (Table 3). 
However, the likelihood-ratio test demonstrated that the 
combination of color and length provided a significantly 
better fit than length alone (P=0.012). 
Recruitment variability, mortality, 
and spawning potential ratio 
From the data shown in Figure 3, it is assumed that 
A. gouldii becomes fully recruited to the commercial 
gillnet fishery by 15 years of age. Because a substantial 
