The reproductive cycle of female Ballan wrasse Labrus bergylta in high latitude, temperate waters.

Abstract

This 2 year study examined the reproductive cycle of wild female Ballan wrasse Labrus bergylta in western Norway as a precursor to captive breeding trials. Light microscopy of ovarian histology was used to stage gonad maturity and enzyme-linked immuno-absorbent assay (ELISA) to measure plasma concentrations of the sex steroids testosterone (T) and 17beta-oestradiol (E(2)). Ovarian recrudescence began in late autumn to early winter with the growth of previtellogenic oocytes and the formation of cortical alveoli. Vitellogenic oocytes developed from January to June and ovaries containing postovulatory follicles (POF) were present between May and June. These POF occurred simultaneously among other late maturity stage oocytes. Plasma steroid concentration and organo-somatic indices increased over winter and spring. Maximal (mean +/-s.e.) values of plasma T (0.95 +/- 0.26 ng ml(-1)), E(2) (1.75 +/- 0.43 ng ml(-1)) and gonado-somatic index (I(G); 10.71 +/- 0.81) occurred in April and May and decreased greatly in July when only postspawned fish with atretic ovaries occurred. Evidence indicates that L. bergylta are group-synchronous multiple spawners with spawning occurring in spring and peaking in May. A short resting period may occur between late summer and autumn when previtellogenic oocytes predominate and steroid levels are minimal.

Full text

Page 1

Journal of Fish Biology (2010)
doi:10.1111/j.1095-8649.2010.02691.x, available online at www.interscience.wiley.com
The reproductive cycle of female Ballan wrasse Labrus
bergylta in high latitude, temperate waters
S. Muncaster*†‡§, E. Andersson*, O. S. Kjesbu*, G. L. Taranger*,
A. B. Skiftesvik† and B. Norberg†
*Institute of Marine Research, P.O. Box 187, Nordnes, N-5817 Bergen, Norway, †Institute of
Marine Research, Austevoll Research Station, N-5392 Storebø, Norway and ‡Department of
Biology, University of Bergen, N-5020 Bergen, Norway
(Received 13 August 2009, Accepted 11 April 2010)
This 2 year study examined the reproductive cycle of wild female Ballan wrasse Labrus bergylta
in western Norway as a precursor to captive breeding trials. Light microscopy of ovarian histol-
ogy was used to stage gonad maturity and enzyme-linked immuno-absorbent assay (ELISA) to
measure plasma concentrations of the sex steroids testosterone (T) and 17β-oestradiol (E2). Ovar-
ian recrudescence began in late autumn to early winter with the growth of previtellogenic oocytes
and the formation of cortical alveoli. Vitellogenic oocytes developed from January to June and
ovaries containing postovulatory follicles (POF) were present between May and June. These POF
occurred simultaneously among other late maturity stage oocytes. Plasma steroid concentration
and organo-somatic indices increased over winter and spring. Maximal (mean ± s.e.) values of
plasma T (0·95 ± 0·26 ng ml−1), E2(1·75 ± 0·43 ngml−1) and gonado-somatic index (IG; 10·71
± 0·81) occurred in April and May and decreased greatly in July when only postspawned fish
with atretic ovaries occurred. Evidence indicates that L. bergylta are group-synchronous multiple
spawners with spawning occurring in spring and peaking in May. A short resting period may occur
between late summer and autumn when previtellogenic oocytes predominate and steroid levels are
minimal.
Journal compilation © 2010 The Fisheries Society of the British Isles
© 2010 The Authors
Key words: oestradiol; oogenesis; protogynous; reproduction.
INTRODUCTION
Ballan wrasse Labrus bergylta Ascanius are common inhabitants on rocky reefs along
the Atlantic coasts of Morocco and Europe, ranging as far north as Trondheim in
Norway (Quignard & Pras, 1986). Female fish exist in harems within the reproductive
territory of a dominant male with which they will mate, spawning their benthic eggs
over temporary nests (Hilld´ en, 1984). Like many other temperate wrasses, L. bergylta
have proven effective cleaners of sea lice Lepeophtheirus salmonis from farmed
Atlantic salmon Salmo salar L. These ectoparasites represent a serious economic
and animal welfare liability to the European S. salar farming industry. Ulcerative
§Author to whom correspondence should be addressed at present address: Bay of Plenty Polytechnic,
Private Bag 12001, Tauranga 3143, New Zealand. Tel.: +64 75440920; fax: +64 75442386; email: simon.
muncaster@boppoly.ac.nz
1
© 2010 The Authors
Journal compilation © 2010 The Fisheries Society of the British Isles

Page 2

2
S. MUNCASTER ET AL.
wounds inflicted by sea lice often cause osmoregulatory stress and can lead to mass
loss and, in severe cases, even death. In addition, infected farmed salmonids have
been implicated as a contamination vector of wild populations (Bjørn & Finstad,
2002; Frazer, 2009). Chemical delousing treatments often used within the S. salar
industry have received negative attention and vary in efficacy due to resistance devel-
opment (Pike & Wadsworth, 1999; Tully & McFadden, 2000; Fallang et al., 2004).
Alternatively, biological control of sea lice is considered environmentally safe and its
application is not subject to resistance problems. Labrus bergylta are the largest of
the European wrasses attaining a maximum size of c. 60 cm (Quignard & Pras, 1986)
and are therefore of sufficient size to keep in sea cages with larger 3–5 kg S. salar. At
present, there is a growing interest in farming L. bergylta to reduce fishing pressure
on local wild stocks and to provide a consistent supply to S. salar farmers. Such an
approach should decrease the use and environmental impact of chemical delousing
agents.
It is advantageous to understand the reproductive cycle of a species in an effort to
co-ordinate effective captive breeding trials. Only limited information exists on the
reproductive physiology of L. bergylta and, in particular, its seasonality in Norway.
Teleost reproduction is often a highly seasonal event, especially at high latitudes due
to the reproductive strategy of external fertilization. The resulting larvae are generally
poorly developed at hatch and are highly susceptible to environmental fluctuations.
Mature fish, therefore, perceive seasonal cues to help direct the timing of gonad
production and breeding (Crim, 1982; Lam, 1983; Bye, 1989; Taranger et al., 1998).
These environmental signals are modulated via the brain-pituitary-gonad (BPG) axis
to effect a co-ordinated endocrine cascade that results in the development of mature
gametes (Bromage et al., 2001; Okuzawa, 2002). Oocyte development can be clas-
sified into discrete categories according to morphological developments (Wallace &
Selman, 1981; Tyler & Sumpter, 1996; Pati˜ no & Sullivan, 2002). These changes are
clearly visible using histological analysis and involve development of the protective
chorion as well as incorporation of hepatic egg yolk proteins into the ooplasm. The
penultimate development before ovulation involves the final maturation of the oocyte
(FOM) when yolk proteins coalesce and the nucleus migrates to the micropyle and
its membrane disintegrates. The events before FOM are coupled with a great increase
in oocyte diameter and are primarily directed by the sex steroid 17β-oestradiol (E2)
(Nagahama, 1994). This hormone is usually produced from the bioconversion of
testosterone (T) in the granulosa layer of the steroidogenic follicle cells surrounding
the oocytes (Nagahama, 1994). Final oocyte maturation tends to be associated with
a decrease in E2and is under the influence of progestins called maturation-inducing
steroids (MIT). Although these MIT differ between species, the most common is 17α,
20β-dihydroxy-4-pregnen-3-one (17, 20βP; Pati˜ no & Sullivan, 2002). Therefore, a
detailed overview of the female reproductive cycle of a species may be determined
through the seasonal analysis of these histological changes and the plasma borne
steroids, which drives them.
In this study, the reproductive cycle of female L. bergylta caught from a wild
population in western Norway was characterized. Histological data, gonad indices
and plasma sex steroid values collected over two complete reproductive cycles were
used to delineate ovarian recrudescence, final oocyte maturation and putative spawn-
ing. From this, the first data linking circulating steroid levels with defined stages of
oogenesis in this temperate wrasse are provided.
© 2010 The Authors
Journal compilation © 2010 The Fisheries Society of the British Isles, Journal of Fish Biology 2010, doi:10.1111/j.1095-8649.2010.02691.x

Page 3

REPRODUCTIVE CYCLE OF FEMALE LABRUS BERGYLTA
3
MATERIALS AND METHODS
Gillnets and traps were used to capture 125 female L. bergylta from 5 to 20 m depth in
western Norway (60◦06?N; 5◦10?E) between February 2005 and February 2007. Fish were
obtained monthly and stored for up to 2 days in a sea cage before sampling. The fish were
then rapidly removed from the storage cage by a dip-net, killed with a blow to the head and
measured for total length (LT) and body mass (MB) to the nearest cm and g. Blood (2 ml) was
then immediately obtained from the duct of Cuvier using cold heparinized syringes, stored on
ice until centrifugation (12 100 g, 3 min; Eppendorf mini-spin; www.eppendorf.com) after
which the plasma was frozen on dry ice for storage at −80◦C until further analysis. Ovaries
(MG) and liver (ML) were excised and weighed to the nearest 0·01 g for calculation of gonado-
somatic (IG) and hepato-somatic (IH) indices according to IG= 100 MG(MB− MG)−1and
IH= 100 ML(MB− ML)−1.
HISTOLOGY
An introductory analysis of anterior, mid and posterior portions from both ovarian lobes
indicated that oocyte development was uniform throughout the gonad. Future tissue samples
were, therefore, excised from the anterior region of the ovary, fixed in neutral phosphate-
buffered formalin and 3 μm paraffin sections were cut and stained with haematoxylin, eosin
and saffron, using conventional protocols. Ovaries were classified according to the develop-
mental stage of the leading cohort oocytes (West, 1990), using a light microscope with a
mounted digital camera. Images were processed with Adobe Photoshop 7.0. The diameter of
leading cohort oocytes that were sectioned through the nucleus was measured digitally (Image
J, National Institute of Health; http://rsbweb.nih.gov/ij). The number of oocytes (n) required
to accurately represent the diameter for each stage was determined after modelling the run-
ning normalized (moving) mean of oocyte diameter as a function of n from 1 to 30 (Howard
& Reed, 1998; Kjesbu et al., 1998) showing quickly stabilized values. Consequently, five
oocytes were measured per stage reflecting an accuracy of ±2%.
STEROID ANALYSES
Steroids were extracted from the plasma samples using ether:heptane (4:1) solvents as
described by Hyllner et al. (1994) and re-dissolved to a relevant concentration in 0·5–1 ml
of enzyme immunoassay buffer (EIA) (0·1 M KHPO4, pH 7·4, 0·4 M NaCl, 1 mM ethylene
diaminetetra acetic acid, EDTA) buffer solution (60◦C, 10 min) and stored at −20◦C until
analysed. Circulating plasma levels of T and E2were measured by enzyme-linked immunosor-
bent assay (ELISA) based on the methods described by Pradelles et al. (1985), Maclouf
et al. (1987) and Cuisset et al. (1994). Antisera, acetylcholine esterase-labelled tracers and
microplates precoated with monoclonal mouse antirabbit IgG were supplied by Cayman
Chemicals (www.caymanchem.com). Standard steroids were purchased from Sigma (Sigma
reference standards; www.sigma-aldrich.com). Logit-log binding curves of serial dilutions of
standards and plasma samples were parallel showing that the plasma extracted samples were
suited to the assay conditions. The ED80 and ED20 were 0·004 and 0·08 ngml−1for T and
0·006 and 0·6 ngml−1for E2. The detection limits of the assays were 0·008 and 0·015 ngml−1
for T and E2, respectively. Extraction efficiencies were 95% for T and 94% for E2. These
were determined by spiking replicate plasma samples with tritiated steroid and confirming
the quantity of steroid recovered. The interassay variation was 7·7% (T) and 7·6% (E2), while
the intra-assay variation was 7·0% (T) and 6·8% (E2). Antibody cross-reactivity is described
by the supplier.
STATISTICS
Data were analysed using GraphPad Prism 5.0 (GraphPad Software Inc.; http://www.graph
pad.com) and presented as mean ± s.e. Differences between groups were analysed using
one-way ANOVA and Bonferroni’s multiple comparison test for post hoc testing (P < 0·05).
Linear regression was used to test relationships between reproductive variables, while Pearson
© 2010 The Authors
Journal compilation © 2010 The Fisheries Society of the British Isles, Journal of Fish Biology 2010, doi:10.1111/j.1095-8649.2010.02691.x

Page 4

4
S. MUNCASTER ET AL.
product-moment correlation was used to determine associations between abiotic and biotic
variables. Assumptions of parametric testing were checked using Bartlett’s test for equal
variance and Kolmogorov–Smirnov’s test for normality. In situations where data varied from
these assumptions, In or inverse transformations were performed.
RESULTS
The ovaries of L. bergylta are arranged as a paired lobular structure, above the
gut in the body cavity. The ovary is of the cystovarian type with oogonial germ cells
arranged in nests inside the epithelial margin of the ovarian lamellae. During the
course of the reproductive cycle, the ovaries varied in shape from a thin clear thread
to a distended grey-coloured sac surrounded by a thin semi-transparent membrane
(tunica albuginea). Ovarian development was classified on a six-stage scale (Table I)
according to the histology of the leading cohort oocytes (Figs 1 and 2).
RELATIONSHIPS BETWEEN SEX STEROIDS,
GONADO-SOMATIC INDEX AND OOCYTE DIAMETER
Linear regression showed that both plasma T (r2= 0·3589, d.f. = 122, P <
0·001) and plasma E2 (r2= 0·4746, d.f. = 122, P < 0·001) concentrations were
positively related to leading oocyte diameters [Fig. 3(a), (b)]. Low concentrations of
both steroids, however, could be found in fish with different sized oocytes, indicating
that plasma sex steroid concentration and oocyte size were poor predictors of each
other. Linear regression analysis also indicated that IG was dependent on oocyte
Table I. Stages of ovarian maturity in Labrus bergylta, according to histological condition
and leading oocyte diameter
Stage of maturity Histological conditionOocyte diameter (μm)
n
I Immature, recovering Previtellogenic oocytes predominate,
postovulatory follicles not present
II Early maturingLeading cohort oocytes contain
cortical alveoli
III Early yolkLeading cohort oocytes contain
small vitellogenic yolk globules
and the inner zona radiata is
generally <6 μm thick
IV Late yolkLeading cohort oocytes contain
many large yolk globules, the
inner zona radiata is >6 μm thick
V Mature Leading cohort oocytes contain
hydrated oocytes with coalescent
yolk and the germinal vesicle has
started to migrate towards the
animal pole
VI Resorbing Atretic oocytes evident
106–174 25
219–265 10
307–44423
462–65838
651–7205
99–314
Total
23
124
n, number of individuals
© 2010 The Authors
Journal compilation © 2010 The Fisheries Society of the British Isles, Journal of Fish Biology 2010, doi:10.1111/j.1095-8649.2010.02691.x

Page 5

REPRODUCTIVE CYCLE OF FEMALE LABRUS BERGYLTA
5
(a) (b)
CT
Pvo
Pvo
CA
EV
(c)
Pvo
Pvo
Pvo
(d)
Pof
LV
(e)
Hy
(f)
At
Fig. 1. Stages of ovarian development in Labrus bergylta: (a) stage I, recovering, (b) stage II, early maturing,
(c) stage III, early yolk, (d) stage IV, late yolk, (e) stage V, mature and (f) stage VI, resorbing (see
Table I). Scale bars = 100 μm (a), (b), (c), (d) and (e) and 250 μm (a) and (f). At, atretic oocyte; CA,
cortical alveolus oocyte; CT, connective tissue; EV, early vitellogenic oocyte; Hy, mature hydrated
oocyte; LV, late-vitellogenic oocyte; Pof, postovulatory follicles; Pvo, previtellogenic oocyte.
diameter [r2= 0·7850, d.f. = 121, P < 0·001, Fig. 3(c)]. The maximum IG, T and
E2levels recorded were 15·6, 0·87 and 3·83 ngml−1, respectively.
SEX STEROIDS, GONADO-SOMATIC AND HEPATO-SOMATIC
INDICES IN RELATION TO OVARIAN MATURITY
Stages of ovarian maturity differed significantly (one-way ANOVA, n = 125–127,
P < 0·001) in IH, IGand concentrations of E2and T (Fig. 4). Both T and E2con-
centrations reflected a similar trend in relation to ovarian stage. There were no
© 2010 The Authors
Journal compilation © 2010 The Fisheries Society of the British Isles, Journal of Fish Biology 2010, doi:10.1111/j.1095-8649.2010.02691.x

Page 6

6
S. MUNCASTER ET AL.
(a)
Pvo
(c)
FL
OZR
IZR
Vg
Cag
FL
(d)
OZR
IZR
Vg
(b)
FL
Cag
ZR
(f)
GC
Oop
FL
Oop
IZR
OZR
(e)
Oog
Fig. 2. Enlargement of ovarian developmental stages in Labrus bergylta: (a) stage I, recovering, (b) stage II,
early maturing, (c) stage III, early yolk, (d) stage IV, late yolk, (e) stage V, mature and (f) stage VI,
resorbing (see Table I). Scale bars = 25 μm (a), (b), (c), (d), (e) and (f). Cag, cortical alveoli granules;
FL, follicular layer; GC, granulosa cells; IZR, inner zona radiata; Oog, oogonia; Oop, ooplasm; OZR,
outer zona radiata; Pvo, previtellogenic oocyte; Vg, vitellogenin globules; ZR, zona radiata.
significant differences between stages I (immature) and II (early maturing) and min-
imum concentrations of 0·04 ± 0·01 ngml−1for T and 0·06 ± 0·01 ngml−1for E2
were recorded. Steroid levels increased in fish with stage III (early vitellogenic)
ovaries and were maximal in stage IV (late vitellogenic) (0·24 ± 0·03 ngml−1
and 1·08 ± 0·14 ngml−1for T and E2, respectively). Concentrations of T and E2
decreased in stage V (mature), although these were not significantly different from
stage IV values. There was a further decrease of T (0·05 ± 0·01 ngml−1) and E2
(0·06 ± 0·02 ngml−1) in fish with stage VI (resorbing) ovaries to near minimal
levels.
The IGincreased significantly from a minimum value of 0·86 ± 0·05 in fish with
stage I ovaries to a maximum value of 11·66 ± 1·23 in fish with stage V gonads.
There was no significant difference, however, between fish with stages IV (6·92 ±
0·46) and V ovaries. Fish with stage VI ovaries had near-minimal IG(1·16 ± 0·05).
The IHdid not vary greatly although differences were evident among the six stages
© 2010 The Authors
Journal compilation © 2010 The Fisheries Society of the British Isles, Journal of Fish Biology 2010, doi:10.1111/j.1095-8649.2010.02691.x

Page 7

REPRODUCTIVE CYCLE OF FEMALE LABRUS BERGYLTA
7
1·0
0·8
0·6
0·4
0·2
0·0
(a)
(c)
20
15
10
5
0
0
200
400600
800
Oocyte diameter (μm)
T (ng ml−1 plasma)
(b)
5
4
3
2
1
0
E2 (ng ml−1 plasma)
IG
Fig. 3. Leading oocyte diameter in relation to (a) levels of plasma testosterone (T) and (b) plasma
17β-oestradiol (E2) and to (c) the gonado-somatic index (IG) in female Labrus bergylta. The curves
were fitted by: (a) y = 0·0004x, (b) y = 0·0016x and (c) y = 0·1220x.
of ovarian development. Fish with stages I and II ovaries had a similar IHof c. 1·5.
This increased to 1·86 ± 0·08 in stage III and attained maximal value of c. 2·2 in
fish with stages IV and V ovaries. There was a slight decrease of IH in fish with
stage VI ovaries to 2·00 ± 0·13.
© 2010 The Authors
Journal compilation © 2010 The Fisheries Society of the British Isles, Journal of Fish Biology 2010, doi:10.1111/j.1095-8649.2010.02691.x

Page 8

8
S. MUNCASTER ET AL.
I IIIIIIVVVI
1·6
1·4
1·2
1.0
0·8
0·5
0·4
0·3
0·2
0·1
0·0
(a)
24 251212232338 38562323
a
A
a
A
b
B
c
abc
C
BC
A
ab
Steroid (ng ml−1 plasma)
IIIIIIIVVVI
252512122323 3838552322
14
12
10
8
6
5
4
2
1
0
3
(b)
a
b
A
AB
c
ABC
C
ABC
a
BC
d
d
Stage
IG or IH
Fig. 4. Mean ± s.e. (a) plasma testosterone (T,
(IG,) and hepato-somatic index (IH,
in Labrus bergylta. Numbers beneath bars indicate sample sizes. Letters indicate significant differences
(P < 0·05) between stages of each variable (lower cases = T or IG, upper cases = E2or IH) as determined
by one-way ANOVA and Bonferroni’s multiple comparisons test.
) and 17β-oestradiol (E2,
) between different stages (see Table I) of ovarian development
) and (b) gonado-somatic index
© 2010 The Authors
Journal compilation © 2010 The Fisheries Society of the British Isles, Journal of Fish Biology 2010, doi:10.1111/j.1095-8649.2010.02691.x

Page 9

REPRODUCTIVE CYCLE OF FEMALE LABRUS BERGYLTA
9
18
24
12
6
0
Day length (h)
15
20
12
6
0
Temperature (° C)
2005
2007
Month
Feb
Mar
Apr
May
Jun
Aug
Sep
Oct
Nov
Dec
Jan
Aug
Sep
Oct
Nov
Dec
Jan
Feb
Mar
Apr
May
Jun
Jul
Jul
Fig. 5. Mean monthly values of seawater temperature (
January 2007 at 60◦06?0??N; 5◦10?0??E.
) and day length () between February 2005 and
MEAN MONTHLY VALUES OF PLASMA SEX STEROIDS
AND GONADO-SOMATIC INDEX IN RELATION TO
ENVIRONMENTAL FACTORS AND OVARIAN MATURITY
Values of sex steroids and IGshowed cyclic changes during the 2 years of study.
Sea temperature also followed an annual cycle in a similar but slightly delayed
trend to that observed in day length. The mean minimum and maximum seawater
temperatures for this period were 3·5 ± 0·1◦C and 18·8 ± 0·1◦C, respectively
(Fig. 5). Negative correlations existed between sea temperature and plasma T (r =
−0·4867, d.f. = 22, P < 0·05) and E2 (r = −0·5921, d.f. = 22, P < 0·01) levels
and IG (r = −0·6791, d.f. = 22, P < 0·01). There were no correlations between
day length and plasma T, E2or IG.
Minimum values of T, E2and IGwere recorded as day length and seawater tem-
peratures decreased, this was c. 2 months after maximal seawater temperatures and
c. 2 months before the winter solstice (Figs 5 and 6). In contrast, these physiological
variables were maximal as sea temperature and day length increased, with peak val-
ues recorded in the 2 months before the summer solstice and c. 2 to 4 months before
sea temperature maxima. There was often an overlap of 1 or 2 months during which
individual fish were recovered containing ovaries of either one or another stage of
maturity. Fish with stage I ovaries (n = 25) were found between August and Decem-
ber, although a total of only four fish in this stage were found between September
and December. Fish with stage II (n = 10) ovaries occurred between November
and February, with only one fish found in this stage in February. Stage III ovaries
(n = 23) were evident from January to April, one fish in this stage was also found
in June. Stage IV ovaries (n = 38) existed between March and June, the majority of
fish caught in May and June had ovaries in this stage. Five fish with stage V ovaries
were obtained during this study, one was in April and the remaining four were found
in May and June, these latter 2 months also coincided with the occurrence of pos-
tovulatory follicles (POF). Only fish with stage VI ovaries (n = 23) occurred in July
© 2010 The Authors
Journal compilation © 2010 The Fisheries Society of the British Isles, Journal of Fish Biology 2010, doi:10.1111/j.1095-8649.2010.02691.x

Page 10

10
S. MUNCASTER ET AL.
55557774316544 11 1198 1051220
(a)2·4
2·2
2·0
1·8
1·2
1·6
1·0
0·8
0·6
0·4
0·0
15
10
5
0
0·2
Plasma steroid concentration (ng ml −1)
∗∗
∗
∗
∗∗
∗∗∗
∗∗
∗∗
∗∗∗
∗∗∗
(b)
2005
2007
Month
Feb
Mar
Apr
May
Jun
Aug
Sep
Oct
Nov
Dec
Jan
Aug
Sep
Oct
Nov
Dec
Jan
Feb
Mar
Apr
May
Jun
Jul
Jul
IG
555577743165441111981051220
Fig. 6. Mean ± s.e. monthly values of (a) plasma testosterone ( ) and 17β-oestradiol ( ) and (b) gonado-
somatic index (IG) and stage of ovarian development in female Labrus bergylta between 2005 and
2007.
, stage I;, stage II; , stage III;
Numbers beneath the bars indicate sample size while cross symbols denote months dropped from analysis
due to insufficient samples size. The level of significance (*P < 0·05; **P < 0·01; ***P < 0·001)
between consecutive months as determined by one-way ANOVA and Bonferroni’s multiple comparisons
test is indicated.
, stage IV; , stage V;, stage VI (see Table I).
(n = 15), five fish of this stage were also present in August and one in January,
February and March.
Values of T, E2 and IG varied significantly (one-way ANOVA, n = 127, P <
0·001) in relation to month of the year. Minimum T concentrations ranging from 0·01
© 2010 The Authors
Journal compilation © 2010 The Fisheries Society of the British Isles, Journal of Fish Biology 2010, doi:10.1111/j.1095-8649.2010.02691.x

Page 11

REPRODUCTIVE CYCLE OF FEMALE LABRUS BERGYLTA
11
to 0·07 ± 0·01 ngml−1occurred between July and January. Although not signifi-
cantly different from previous months, a higher mean value of 0·19 ± 0·05 ngml−1
was recorded in February of the secondyear. The mean value of T in Febru-
ary of the first year was even higher (0·27 ± 0·12 ngml−1). Lower values were
recorded in March of both the first (0·15 ± 0·05 ngml−1) and second years (0·04 ±
0·01 ngml−1). Values in April were similar to February in the same years (0·25 ±
0·14 and 0·20 ± 0·04 ngml−1for years 1 and 2, respectively). The highest values
were recorded in May of years 1 (0·36 ± 0·14 ngml−1) and 2 (0·30 ± 0·05 ngml−1).
These values decreased significantly between May and June (0·07 ± 0·01 ngml−1)
of the second year. The following month of June was characterized by lower T val-
ues in both years (0·17 ± 0·04 and 0·07 ± 0·01 ngml−1, respectively); however,
there was only a significant decrease in the second year.
The annual profile of E2 concentrations followed a similar trend to T although
most values were not significantly different between months. Maximal E2 values
were approximately five-fold greater than T. Minimum E2values of 0·01 to 0·14 ±
0·01 ngml−1occurred between July and January. The E2concentration in the second
year doubled from January (0·14 ± 0·01 ngml−1) to February (0·28 ± 0·09 ngml−1).
Values of E2 in March (0·38 ± 0·14 and 0·12 ± 0·02 ngml−1, years 1 and 2,
respectively) were approximately half of those recorded in the previous February.
This contributed to the formation of an annual bimodal peaked trend similar to
that observed for T concentrations. Higher values were recorded in April of both
years (0·95 ± 0·26 ngml−1in year 1 and 0·78 ± 0·22 ngml−1in year 2), while
maximal values of E2were recorded in May (1·75 ± 0·43 and 1·51 ± 0·35 ngml−1,
respectively). Lower E2 values were recorded in June 0·62 ± 0·19 and 0·47 ±
0·13 ngml−1, respectively. There was a significant decrease in E2 concentrations
between June and July of both years.
Changes in IGapproximated a similar trend to that of T and E2over the 2 year
period. There was no evidence, however, of an annual bimodal peak between Febru-
ary and May. In contrast, IGvalues were greater in March than in February, although
there was only a significant increase in the first year. Values were low between July
and December (0·44 to 1·35 ± 0·11). Higher values were recorded in January of each
year (2·25 ± 0·31 and 1·87 ± 0·34 for years 1 and 2, respectively). The highest IG
values were recorded between April and June in both years (6·61 ± 0·88 to 7·44 ±
1·27 in year 1 and 4·57 ± 0·35 and 10·71 ± 0·81 in year 2). Peak values, however,
were recorded in April in the first year and in May in the second with lower values
recorded in June (6·61 ± 0·88 in year 1 and 5·41 ± 0·97 in year 2). Peak values in
May of the second year were significantly higher than both April and June of the
same year. There was a significant decrease in IGbetween June and July.
DISCUSSION
This study characterizes the reproductive cycle of female L. bergylta in Norway
using plasma T and E2 and IG, oocyte diameter and ovarian histology, collected
monthly over a 2 year period. Ovarian development of L. bergylta followed a clear
annual cycle. Such defined reproductive cycles are typical of temperate and high
latitude teleosts (Crim, 1982). Seasonal oscillations often permit only limited peri-
ods where environmental conditions are conducive to the survival of offspring. The
© 2010 The Authors
Journal compilation © 2010 The Fisheries Society of the British Isles, Journal of Fish Biology 2010, doi:10.1111/j.1095-8649.2010.02691.x

Page 12

12
S. MUNCASTER ET AL.
L. bergylta in this study appeared to enter a short quiescent period between August
and October, when IGwas at minimal levels and only previtellogenic oocytes and
nests of oogonia were present in the ovary (stage I). A few fish with postspawned,
resorbing ovaries (stage VI) containing atretic postvitellogenic oocytes, however,
were still present in August of the first year. Atresia of unspawned oocytes was
complete by September. This was also when oocyte diameter was lowest. Plasma
concentrations of T and E2were minimal in September and October, suggesting the
existence of a brief quiescent period in late summer, autumn as light period and sea
temperature started to decrease.
Ovarian recrudescence started in November and December with the appearance of
oocytes containing cortical alveoli (stage II) and a trend of greater E2and IGvalues.
A continued trend of greater T, E2and IGvalues occurred 2 months later in January
with the appearance of early vitellogenic oocytes (stage III ovaries). Two steroid
peaks were indicated in February and April with a trend for lower steroid concen-
trations in March. Although these apparent bimodal steroid peaks were consistent
in both years, the concentrations did not differ significantly between consecutive
months. Nevertheless the pattern is interesting as bimodal steroid profiles have been
observed in the ovarian recrudescence of other teleosts (Bonnin, 1979; Burke et al.,
1984; Pankhurst & Conroy, 1987, Prat et al., 1990). A significant increase in IGdur-
ing these months was only evident in the first year of sampling, however, further dis-
torting any apparent connection between ovarian development and steroid bimodality.
While the significance of bimodality may be unclear, the observed increase of IGin
the first year may be explained by E2stimulation of vitellogenesis (Nagahama, 1994).
Gonadal steroid production was greatest during vitellogenic stages of oocyte devel-
opment. The lipophospoglycoprotein, vitellogenin is produced in the liver under
receptor-mediated stimulation of E2 in teleosts (Menuet et al., 2001). Therefore,
oogenesis proceeded rapidly from February to May as E2concentrations increased,
and by March approximately half of the fish contained stage IV ovaries with late-
vitellogenic oocytes. This was reflected by significant differences in IGamong fish
with stages II, III and IV ovaries. Vitellogenesis often causes a temporary increase in
liver mass (Korsgaard-Emmersen & Emmersen, 1976; Bohemen et al., 1981; Haux
& Norberg, 1985). This is likely to explain the significant increase in IHbetween
fish with stages II and IV ovaries. Indeed, the increase in IHobserved in stages III
and IV ovaries is supported histologically through the vitellogenin incorporation evi-
dent in the oocytes of these fish. In addition, the thickness of the inner zona radiata
(IZR) increased in fish with stage IV oocytes (Table I). These choriogenins are often
derived from the liver, although not exclusively in all teleosts (Begovac & Wallace,
1989; Chang et al., 1996; Hyllner et al., 2001), under the influence of E2and may
have, in part, contributed to the increase in IHin these stages. Therefore, much of
the observed increase in IGand IHis associated with oocyte development resulting
from increased sex steroid production.
The plasma steroid levels decreased as vitellogenesis ended. Although not signifi-
cant, the mean plasma concentrations of T and E2were slightly less in fish containing
stage V ovaries with mature, hyaline oocytes. Aromatizable androgens and E2have
been shown to have a negative feedback on the pituitary release of follicle-stimulating
hormone (FSH) and a potentiating effect on luteinizing hormone (LH) release, in
many teleosts (Yaron et al., 2003). This results in a shift towards the follicular pro-
duction of progestogens that stimulate final oocyte maturation and may explain the
© 2010 The Authors
Journal compilation © 2010 The Fisheries Society of the British Isles, Journal of Fish Biology 2010, doi:10.1111/j.1095-8649.2010.02691.x

Page 13

REPRODUCTIVE CYCLE OF FEMALE LABRUS BERGYLTA
13
decreased E2concentrations observed in fish with stage V ovaries. The patterns of
FSH and LH release can be complicated depending on the spawning strategy of the
fish (Yaron et al., 2003). The first incidence of a fish with stage V ovaries was in late
April, indicating that spawning was imminent. Maximal concentrations of T and E2,
however, were not until May, after which, there were marked decreases in T, E2and
IGin June. This indicates that peak spawning occurred during these 2 months. Rapid
decreases in E2and IGfrom peak values are typical patterns of gonad maturation
and spawning in other temperate marine teleosts (Shimizu, 1997; Dahle et al., 2003;
Clark et al., 2005; Garcia-L´ opez et al., 2006). The ovaries of fish in May and June
contained an abundance of POF and were in either stage IV or V, indicating that
the spawning season started before May. Spawning seems to be complete by July as
indicated by ovaries in a postspawned state with atretic mature oocytes (stage VI).
Moreover, this notion was also supported by a significant decrease in T (year 2), E2
and IG, as well as the deflated appearance of the oocytes that had been invaded by
phagocytotic granulosa cells (Lang, 1981). Previous work in British waters classifies
L. bergylta as a group-synchronous single spawner (Dipper & Pullin, 1979). The
coincidence of POF together with oocytes in several developmental stages and the
existence of a large pool of previtellogenic oocytes in the present study indicates
that L. bergylta should be classified as a group-synchronous multiple spawner. This
was further supported by the observation that individual fish can spawn at least three
times in the same breeding season (unpubl. obs.).
Seasonal shifts of environmental variables are likely to modulate the reproductive
cycle of L. bergylta. Although there was no clear association between changes in
day length and the production of T, E2or IG, it remains likely that shifts in pho-
toperiod are involved in the timing of reproduction in this species. Photoperiod is
a primary cue affecting the timing of reproduction in most teleosts (Hoover, 1937;
Girin & Devauchelle, 1978; Whitehead et al., 1978; Smith et al., 1991; Taranger &
Hansen, 1993; Watanabe et al., 1998; Morehead et al., 2000). In addition, changes
in day length are a dominant environmental phenomenon at the latitude of this study.
Correlations between sea temperature and T, E2and IGindicate that temperature has
a significant effect on gonadal processes and reproduction. Temperature has been
suggested to fulfil a permissive role in teleost reproduction (Taranger et al., 2010).
A decrease in temperature appears to induce the reproductive cycle in percids and
moronids (Dabrowski et al., 1996; Prat et al., 1999; Migaud et al., 2002; Clark et al.,
2005). This may also be consistent with the inverse correlations observed between sea
temperature and steroid levels as well as IG, in L. bergylta.The decrease in T, E2and
IGin June, as well as the minimal levels of these variables associated with postspawn-
ing in July, contrasts with the rapidly warming sea temperatures at this time of year
and could also help explain the inverse association between these biotic and abiotic
factors. Although temperature may influence enzymatic processes in the gonads, the
data from this study do not support greater speculation. Further research into the
relationship between the photothermal cues and the reproductive cycle of L. bergylta
will have important implications for controlling reproduction in this species.
Plasma sex steroid concentrations in L. bergylta were found to be relatively
low (<5 ngml−1) like several other marine species such as the saddleback wrasse
Thalassoma duperrey (Quoy & Gaimard) (Nakamura et al., 1989), the blackeye
goby Coryphopterus nicholsii (Bean) (Kroon & Liley, 2000), honeycomb grouper
Epinephelus merra Bloch (Bhandari et al., 2003) and red porgy Pagrus pagrus L.
© 2010 The Authors
Journal compilation © 2010 The Fisheries Society of the British Isles, Journal of Fish Biology 2010, doi:10.1111/j.1095-8649.2010.02691.x

Page 14

14
S. MUNCASTER ET AL.
(Kokokiris et al., 2006), which have sex steroid levels between 1 and 10 ngml−1,
By contrast, other fish species such as salmonids may have concentrations of >50
ngml−1(Taranger et al., 1998). Low sex steroid production may be typical of fishes
that spawn multiple times within a spawning season. Pankhurst & Carragher (1991)
suggest that fishes with low steroid levels tend to follow a semi-lunar spawning
cycle while group-synchronous spawners such as plaice Pleuronectes platessa L. and
winter flounder Pseudopleuronectes americanus (Walbaum) often have higher sex
steroid levels. It would seem that L. bergylta also fits this trend as an apparent group-
synchronous multiple spawner and low steroid levels. This would be better supported
through captive manipulation of fish and the measurement of sex steroids across
a finer time scale as E2concentrations can fluctuate between the spawning of egg
batches (Zohar et al., 1998; Methven et al., 1992; Kjesbu et al., 1996). Further atten-
tion also needs to be applied to identifying final maturation-inducing steroids in these
fishes. For example, the Japanese kyusen Halichoeres poecilopterus (Richardson)
and bambooleaf Pseudolabrus japonicus (Houttuyn) wrasses both produce two pro-
gestins, 17, 20β-dihydroxy-4-pregnen-3-one (17, 20βP) and 17α,20β,21-trihydroxy-
4-pregnen-3-one (20β-S) responsible for inducing germinal vessicle breakdown in
oocytes (Matsuyama et al., 1998a, 2002). Quantification studies of these progestins
in combination with analysis of ovarian histology have determined that these fishes
follow a diurnal spawning cycle with peak spawning occurring between 0600 and
0900 hours or presumably shortly after dawn (Matsuyama et al., 1998b, 2002).
Therefore, identification and measurement of MIS progestins would provide valuable
information regarding the frequency and timing of ovulation in L. bergylta.
In summary, female L. bergylta at 60◦N experiencing water temperatures between
4 and 19◦C and a day length oscillation of 6 to 18 h follow an annual reproductive
cycle that is characterized by increasing values of T, E2and IGand culminates in
spawning in April to June. Early ovarian recrudescence begins in late autumn to
early winter when day length is near minimal and water temperature is decreasing.
Oocytes rapidly progress into vitellogenesis by late winter. Spawning appears to take
place in late spring to early summer, when both day length and water temperature
are increasing, and is complete shortly after the summer solstice as sea temperature
continues to increase. Postspawned fish undergo ovarian atresia between July and
August before possibly entering a short resting period.
The authors wish to thank the contribution of the technical staff at the Institute of Marine
Research, Austevoll Research Station and the Institute of Marine Research, Bergen. In par-
ticular: S. Gokstad, S. Kalvenes, S. Olausson, R. Karlsen for assistance with steroid work
and A. Torsvik, I. Uglenes Fiksdal and L. Kleppe for assistance with the histology samples.
The study was supported by the Research Council of Norway, grant number 153261 and the
Institute of Marine Research.
References
Begovac, P. C. & Wallace, R. A. (1989). Major vitelline envelope proteins in pipefish oocytes
originate within the follicle and are associated with the Z3 layer. Journal of Experi-
mental Zoology 251, 56–73.
Bhandari, R. K., Komuro, H., Nakamura, S., Higa, M. & Nakamura, M. (2003). Gonadal
restructuring and correlative steroid hormone profiles during natural sex change in pro-
togynous honeycomb grouper (Epinephelusmerra). ZoologicalScience 20, 1399–1404.
© 2010 The Authors
Journal compilation © 2010 The Fisheries Society of the British Isles, Journal of Fish Biology 2010, doi:10.1111/j.1095-8649.2010.02691.x

Page 15

REPRODUCTIVE CYCLE OF FEMALE LABRUS BERGYLTA
15
Bjørn, P. A. & Finstad, B. (2002). Salmon lice, Lepeophtheirus salmonis (Krøyer), infestation
in sympatric populations of Arctic charr, Salvelinus alpinus (L.), sea trout, Salmo trutta
(L.), in areas near and distant from salmon farms. ICES Journal of Marine Science 59,
131–139.
Bohemen, C. G., van Lambert, J. G. D. & Peute, J. (1981). Annual changes in plasma and
liver in relation to vitellogenesis in the female rainbow trout, Salmo gairdneri. General
and Comparative Endocrinology 44, 81–92.
Bonnin, J. P. (1979). Seasonal-variations of plasma testosterone in a teleost fish. Comptes
Rendus Hebdomadaires des Seances de l Academie des Sciences Serie D 288, 627–630.
Bromage, N., Porter, M. & Randall, C. (2001). The environmental regulation of maturation
in farmed finfish with special reference to the role of photoperiod and melatonin.
Aquaculture 197, 63–68.
Burke, M. G., Leatherland, J. F. & Sumpter, J. P. (1984). Seasonal changes in serum testos-
terone, 11-ketotestosterone and 17β-estradiol levels in brown bullhead, Ictalurus neb-
ulosus Lesueur. Canadian Journal of Zoology 62, 1195–1199.
Bye, V. J. (1989). The role of environmental factors in the timing of reproductive cycles.
In Fish Reproduction: Strategies and Tactics (Potts, G. W. & Wootton, R. J., eds),
pp. 187–205. London: Academic Press.
Chang, Y. S., Wang, S. C., Tsao, C. C. & Huang, F. L. (1996). Molecular cloning, structural
analysis, and expression of carp ZP3 gene. Molecular Reproduction and Development
44, 295–304.
Clark, R. W., Henderson-Arzapalo, A. & Sullivan, C. V. (2005). Disparate effects of constant
and annually-cycling day length and water temperature on reproductive maturation of
striped bass (Morone saxatilis). Aquaculture 249, 497–513.
Crim, L. W. (1982). Environmental modulation of annual and daily rhythms associated with
reproduction in teleost fishes. Canadian Journal of Fisheries and Aquatic Sciences 39,
17–21.
Cuisset, B., Pradelles, P., Kime, D. E., K¨ uhn, E. R., Babin, P., Davail, S. & Le Menn, F.
(1994). Enzyme immunoassay for 11-ketotestosterone using acetylcholinesterase as
label: application to the measurement of 11-ketotestosterone in plasma of Siberian
sturgeon. Comparative Biochemistry and Physiology C 108, 229–241.
Dabrowski, K., Ciereszko, R. E., Ciereszko, A., Toth, G. P., Christ, S. A., El-Saidy, D. &
Ottobre, J. S. (1996). Reproductive physiology of yellow perch (Perca flavescens):
environmental and endocrinological cues. Journal of Applied Ichthyology 12, 139–148.
Dahle, R., Taranger, G. L., Karlsen, Ø., Kjesbu, O. S. & Norberg, B. (2003). Gonadal devel-
opment and associated changes in liver size and sexual steroids during the reproductive
cycle of captive male and female Atlantic cod (Gadus morhua L.). Comparative Bio-
chemistry and Physiology A 136, 641–653.
Dipper, F. A. & Pullin, R. S. V. (1979). Gonochorism and sex-inversion in British Labridae
(Pisces). Journal of the Zoological Society of London 187, 97–112.
Fallang, A., Ramsay, J. M., Sigmund Sevatdal, S., Burka, J. F., Jewess, P., Hammell, K. L.
& Horsberg, T. E. (2004). Evidence for occurrence of an organophosphate-resistant
type of acetylcholinesterase in strains of sea lice (Lepeophtheirus salmonis Krøyer).
Pest Management Science 60, 1163–1170.
Frazer, L. N. (2009). Sea-cage aquaculture, sea lice, and declines of wild fish. Conservation
Biology 23, 599–607.
Garcia-L´ opez, A., Anguis, V., Couto, E., Canario, A. V. M., Canavate, J. P., Sarasquete, C.
& Martinez-Rodriguez, G. (2006). Non-invasive assessment of reproductive status and
cycle of sex steroid levels in a captive wild broodstock of Senegalese sole, Solea
senegalensis (Kaup). Aquaculture 254, 583–593.
Girin, M. & Devauchelle, N. (1978). Decalage de la periode de reproduction par raccour-
cissement des cycles photoperiodique et thermique chez des poissons marin. Annales
de Biologie Animale Biochimie Biophysique 18, 1059–1066.
Haux, C. & Norberg, B. (1985). The influence of estradiol-17β on the liver content of protein,
glycogen and nucleic acids in juvenile rainbow trout, Salmo gairdnerii. Comparative
Biochemistry and Physiology B 81, 275–279.
© 2010 The Authors
Journal compilation © 2010 The Fisheries Society of the British Isles, Journal of Fish Biology 2010, doi:10.1111/j.1095-8649.2010.02691.x