Coulson et al.: Biological features of Achoerodus gouldii 
71 
of a fish than either of those variables on their own. Al- 
though sex was determined for only a proportion of the 
large fish, the length and color of all fish were recorded. 
This enabled us to construct growth curves for females 
and males that incorporated the length at age for each 
individual of unknown sex, allowing for the probability 
that, from its length and color, it would be of a particu- 
lar sex. Through increasing substantially the numbers 
of large fish from those whose sex was just determined 
from their gonads, the von Bertalanffy growth curves 
thereby derived for females and males are considered to 
provide better representations of the growth throughout 
life of the two sexes of A. gouldii than those derived 
solely from fish whose sex was unknown. 
The greater length of males than females of A. goul- 
dii at corresponding ages parallels the situation re- 
corded for a number of other protogynous fishes, e.g., 
the cylindrical sandperch ( Parapercis cylindrica), the 
heavybeaked parrotfish ( Chlorurus gibbus ), the greenfin 
parrotfish ( Chlorurus sordidus), the swarthy parrot- 
fish ( Scarus niger), and the sixband parrotfish ( Sea - 
rus frenatus) (Choat et al., 2006; Munday et al., 2004; 
Walker and McCormick, 2004), as well as its congener 
Achoerodus viridis (Gillanders, 1995a). Comparisons of 
the distributions of the lengths at age of the two sexes 
(Fig. 4B) strongly indicate that the fastest growing 
A. gouldii have the greatest tendency to change from 
female to male, a conclusion which parallels that drawn 
by Munday et al. (2004) for the labrid S. frenatus. How- 
ever, some of the oldest fish were females (Fig. 4, B and 
D) and therefore some female A. gouldii were apparently 
not destined to become males, as is also the case with 
Cheilinus undulatus (Choat et al., 2006). 
As hypothesized, the greater TL attained by A. gouldii 
(1162 mm) than by A. viridis (773 mm, after conversion 
from the SL given by Gillanders, 1995a), corresponded 
to a greater maximum age of 70 years for A. gouldii in 
comparison to 35 years for A. viridis. The maximum 
age of the large temperate A. gouldii is thus twice that 
estimated by Gillanders (1995a) for A. viridis, and by 
far the greatest yet recorded for any species of labrid, 
including the large tropical species C. undulatus, which 
attains a maximum age of 32 years (Choat and Bell- 
wood, 1994; Choat et al., 2006). Although Hostetter and 
Munroe (1993) concluded that other large temperate 
labrids have long life spans, Choat et al. (2006) consid- 
ered that the life spans of the large labrids that they 
used to make this generalization, typically <30 years of 
age, are relatively short. In the context of A. gouldii, the 
exceptional age attained by males was accompanied by 
a marked tendency for growth to an asymptote, which 
is not typically the case with the males of large tropical 
labrids (Choat et al., 2006). 
Reproductive biology 
The vast majority of the A. gouldii caught by spear fish- 
ing at numerous sites in nearshore waters were obtained 
from over the granite reefs that dominate the hard- 
structured substrate found in those waters along the 
south coast of Western Australia, and around the islands 
off the coast (Kendrick, 1999; Sanderson et al., 2000). 
Because the sizes of these fish, which were representa- 
tive of the A. gouldii observed in those clear waters, were 
predominantly <600 mm and <12 years old, and there- 
fore below the TL 50 at maturity, the reef areas in near- 
shore waters constitute an important nursery habitat 
for juvenile A. gouldii. Large numbers of A. gouldii >500 
mm, many of which were mature during the spawning 
season, were caught in deeper waters and this finding 
indicates that the individuals of this species move off- 
shore as they increase in body size and become mature. 
This movement and the spawning of A. gouldii in off- 
shore, deeper waters are consistent with the conclusions 
drawn by Shepherd and Brook (2007) for this species. 
Our data indicate that, on the south coast of Western 
Australia, A. gouldii spawns in offshore waters between 
early winter and mid-spring, when water temperatures 
are declining to their minima (Fig. 5B); the preference 
for offshore waters and an early winter to mid-spring 
season for spawning parallels the situation with the 
congeneric A. viridis in eastern Australia (Gillanders, 
1995a). Although there are no data on the larval phase 
of A. gouldii, Leis and Hay (2004) have shown that the 
larvae of A. viridis settle at 7-8 mm, metamorphose 
into juveniles by about 10 mm, and have hypothesized 
that the larvae of this species have some behavioral 
control over their movement from the shelf to their 
nearshore nursery habitats. Consequently, it is proposed 
that the spawning of A. gouldii in winter enables the 
larvae of this species to recruit into protected, near- 
shore coastal waters, at a time when the temperature 
and productivity in those waters have already started 
to increase (Fig. 5A) and winds are at their weakest 
and thus collectively provide an optimum environment 
for early juvenile growth. 
The hypothesis that the greater length and age at- 
tained by A. gouldii than by A. viridis is accompanied 
by a greater length and age at maturity of females was 
confirmed by our results. With A. gouldii, the smallest 
mature female was 391 mm and the TL 50 of females at 
maturity was as high as 653 mm. Furthermore, only 
one of the 164 females <5 years old was mature, and 
maturity was not attained by -50% of females until 
they had reached 15-19 years in age. Although Gil- 
landers (1995a) did not estimate the TL 50 at maturity 
for female A. viridis, she recorded that females first 
matured at 200-220 mm SL ( = 246-270 mm TL) and 
that the majority had become mature by 240-280 mm 
SL (=294-341 mm TL). Gillanders (1995a) also found 
that, although a few female A. viridis became mature 
at the end of their second year of life, the majority ma- 
tured between their third and fifth years of life. As with 
maturity, sex change occurred at a greater length and 
age in A. gouldii than in the smaller A. viridis. Thus, 
whereas sex change commenced in A. gouldii at -650 
mm and 15 years of age, and the TL 50 at sex change 
was as high as 821 mm (TL), it began in A. viridis at 
lengths of -600 mm (500 mm SL) and as early as 10 
years of age (Gillanders 1995a). 
