Environmental Biology of Fishes 70: 257–272, 2004.
© 2004 Kluwer Academic Publishers. Printed in the Netherlands.
Reproductive biology and protandrous hermaphroditism in
Acanthopagrus latus
S. Alex Hesp, Ian C. Potter & Norman G. Hall
Centre for Fish and Fisheries Research, School of Biological Sciences and Biotechnology, Murdoch University,
Murdoch, Western Australia 6150, Australia (e-mail: i.potter@murdoch.edu.au)
Received 7 March 2003 Accepted 18 October 2003
Key words: gonadal changes, spawning period, fecundity, spawning potential ratio
Synopsis
Detailed macroscopic and histological studies of the gonads of a full size and age range of Acanthopagrus latus from
each season in Shark Bay, Western Australia, demonstrate that this species is a protandrous hermaphrodite in this
large subtropical embayment. Although our scheme for the changes that occur in the ovotestes of A. latus during life
is not consistent with some of the conclusions drawn for this species elsewhere, it is similar to that of Pollock (1985
J. Fish. Biol. 26: 301–311) for the congeneric Acanthopagrus australis. The ovotestes of males develop from gonads
which contain substantial amounts of both testicular and ovarian tissue. The testicular component of the ovotestes of
all males regresses markedly after spawning. During the next spawning season, the ovotestes either become gonads
in which the testicular zone again predominates and contains spermatids and spermatozoa (functional males), or
gonads in which the ovarian zone now predominates and contains mature oocytes (functional females). Once a fish
has become a functional female, it remains a female throughout the rest of its life. In Shark Bay, A. latus typically
spawns on a limited number of occasions during a short period in late winter and early spring and has determinate
fecundity. The mean potential annual fecundity was ca. 2 000 000. The total length of 245 mm, at which, during
the spawning period, 50% of A. latus become identifiable as males, is very similar to the current minimum legal
length (MLL) of 250 mm, which corresponds to an age of ca. 2.5 years less than that at which 50% of males become
females. Thus, although the spawning potential ratio suggests that the present fishing pressure is sustainable, the
current MLL should be reviewed if recreational fishing pressure continues to increase.
Introduction
The western yellowfin bream Acanthopagrus latus
lives in coastal waters throughout the Indo-Pacific,
where it is fished both recreationally and commercially
(Hussain & Abdullah 1977, Lee & Al-baz 1989, Allen
1997). The most important fishery for this sparid in
Australia is located in Shark Bay, a large subtropical
marine embayment on the west coast of this conti-
nent. The indications from reduced catches that A. latus
may have been subjected to overfishing in Shark Bay
led the Western Australian Department of Fisheries
to reduce the number of commercial licenses for this
species (Shaw 2000). However, until the present study,
there were no biological data for any populations of
A. latus in Australia that could be used to develop sound
management plans for this sparid in this continent.
Previous studies on the reproductive biology of
A. latus, which have all been undertaken on popula-
tions in the northern hemisphere, have shown that the
spawning period varies markedly in different regions
(Hussain & Abdullah 1977, Abu-Hakima 1984, Abol-
Munafi & Umeda 1994, Chang et al. 2002). This
species, like other sparids, possesses ovotestes, which
comprise paired bisexual gonads consisting of a medio-
dorsal ovarian zone and a latero-ventral testicular zone,
separated by a wall of connective tissue (D’Ancona
1949, Besseau & Brusle´-Sicard 1995). However, there
is a divergence of opinion as to whether all of the
individuals of A. latus are protandrous hermaphrodites

258
or whether all or some individuals are essentially gono-
chorists (compare Kinoshita 1939, Abu-Hakima 1984,
Abol-Munafi & Umeda 1994, Abou-Seedo et al. 2003).
Abol-Munafi & Umeda (1994) based their conclusion
that A. latus was a protandrous hermaphrodite on the
fact that the individuals in the population they studied
shifted from exclusively males to very largely females
as they increased in size. However, as sex-related
bimodal size–frequency distributions can be produced
by factors other than sequential hermaphroditism
(Sadovy & Shapiro 1987), this finding does not consti-
tute definitive evidence of protandry. Indeed, Buxton &
Garrett (1990) considered many of the reports that cer-
tain sparid species are protandrous to be questionable
and that at least some of those species are rudimentary
hermaphrodites, i.e. an individual progresses from a
juvenile possessing gonads with both immature testicu-
lar and ovarian tissue into either a functional male with
an ovotestis containing ovarian rudiments or a func-
tional female with an ovotestis containing testicular
rudiments.
Successful fisheries management requires that the
reproductive capacity of a stock is sustained above
a threshold level that is sufficient to ensure that
a high level of egg production is maintained and
that recruitment is not jeopardised (Goodyear 1993,
Mace & Sissenwine 1993, Mace 2001). Since exploita-
tion is often size selective, species that undergo an
irreversible sex change, i.e. protandrous and protog-
ynous hermaphrodites, may be particularly susceptible
to recruitment overfishing unless specific measures,
such as the introduction of appropriate minimum legal
lengths (MLLs) for capture, are undertaken to ensure
that sufficient numbers of both sexes are maintained
(Buxton 1992, Milton et al. 1998). The development
of reliable estimates for the optimal levels of exploita-
tion for hermaphroditic species, such as those derived
from per recruit models, requires a thorough under-
standing of the pattern of sex change exhibited by
these species as individuals increase in size and/or
age. Data on changes in sex were thus required for
producing the type of per recruit analyses that were
undertaken by Bannerot et al. (1987), Buxton (1992)
and Punt et al. (1993) for protogynous species and
by Milton et al. (1998), in essence, for a protandrous
species.
The first aim of the present study was to examine,
both macroscopically and histologically, the gonads of
a wide size and age range of A. latus from Shark Bay
in order to determine the changes undergone by the
ovotestes during the life of this hermaphroditic sparid.
Particular emphasis has been placed on ascertaining
quantitatively how the prevalence of fish with different
types of ovotestes in the different size classes changes
throughout the year and in both sequential size and
age classes during the spawning period. The latter data
have then been used to ascertain the lengths and ages
over which this species changes sex and to provide
reliable data on sex ratios for use in stock assessment.
The second aim was to determine the duration of the
spawning period and whether A. latus has determinate
or indeterminate fecundity sensu Hunter et al. (1985),
i.e. whether or not the number of eggs released by indi-
vidual females within a spawning period is determined
prior to that period. The third aim was to determine the
relationships between potential annual fecundity and
both the total length (TL) and somatic mass of A. latus.
The fourth aim was to determine the spawning biomass
per recruit of male and female A. latus and egg produc-
tion per recruit for the females by using an estimate
derived for total mortality during another study (Hall
et al. 2004) and a range of natural mortalities. Finally,
the management implications of our results for A. latus
are discussed.
Materials and methods
We sampled A. latus in Shark Bay by line fishing and
seine netting at monthly or bimonthly intervals between
August 1999 and December 2001. We also purchased
samples from commercial fishing sources. We recorded
the TL and wet mass of each fish to the nearest 1 mm
and 0.1 g, respectively, and weighed the gonads of each
fish to the nearest 0.01 g. We aged each fish using the
number of annuli on their otoliths, which had been
validated as being formed each year (Hall et al. 2004).
We assigned the gonads of a large number of A. latus
that covered a wide size range, on the basis of a macro-
scopic investigation, to one of the following categories:
(1) Very thin, strand-like and of indeterminate sex.
These are found only in juveniles <ca. 160 mm in TL.
(2) Ovotestes containing substantial amounts of both
immature testicular and ovarian material. (3) Ovotestes
in which the testicular zone clearly predominates. (4)
Ovotestes in which the ovarian zone clearly predom-
inates. We considered fish with gonads at categories
3 and 4 during the spawning period to be functional
males and functional females, respectively.
We allocated each gonad that contained almost
exclusively either ovarian or testicular tissue, on
the basis of the macroscopic characteristics of