Professional Documents
Culture Documents
Georg Ako Poul Ou 2007
Georg Ako Poul Ou 2007
Georg Ako Poul Ou 2007
To exam whether the temperature experienced by fishes at early developmental stages can
influence their phenotype at subsequent stages, the model species used, European sea bass
Dicentrarchus labrax was subjected to water temperatures of 15 or 20° C during the half-epiboly
stage until the metamorphosis. Meristic and morphometric characters at three different stages,
well after the end of the thermal treatments, were explored. Body shape and most of the meristic
characters were significantly affected by the environmental temperature during their early life
stages. Fish body shape at 15° C tended to be more slender than at 20° C. The dorsal spines and
soft rays, the pectoral lepidotrichia and caudal dermatotrichia were significantly affected.
Phenotypic differences due to the two thermal regimes are discussed in terms of their functional
meaning during the transition from the planktonic to the littoral niche. # 2007 The Authors
Key words: Dicentrarchus labrax; early life history; meristics; morphometrics; phenotypic
plasticity; temperature.
INTRODUCTION
Environmental factors can affect fish performance and survival directly
through their action (e.g. water temperature, food and predator abundance),
or indirectly via environmentally induced phenotypic variation (Blaxter, 1992;
Chambers & Leggett, 1996; Hare & Cowen, 1997; Fuiman et al., 1998; Brander
et al., 2001; McCormick, 2003). Phenotypic plasticity is one of the most impor-
tant mechanisms of phenotypic adaptation and is defined here as the ability of
a single genotype to produce more than one alternative phenotype in response
to environmental conditions (West-Eberhard, 1989; Schlichting & Pigliucci,
1998; Robinson & Parsons, 2002; Relyea & Hoverman, 2003).
In teleosts, high levels of environmentally induced phenotypic plasticity have
been demonstrated during the early ontogenetic stages. Water temperature has
§Author to whom correspondence should be addressed. Tel.: þ30 2610 969217; fax: þ30 2610 969268;
email: koumound@upatras.gr
278
# 2007 The Authors
Journal compilation # 2007 The Fisheries Society of the British Isles
PHENOTYPIC PLASTICITY IN SEA BASS JUVENILES 279
FISH POPULATIONS
Sea bass eggs were obtained from a wild broodstock that was matured and spawned
as described in Koumoundouros et al. (2002). Egg incubation and larval rearing was
performed in four 500 l tanks, which were connected to closed re-circulation systems
equipped with a biological filter. Two experimental groups were formed (duplicate
tanks per group), according to the different temperature conditions, 15 or 20° C,
which were selected because they fall inside the natural thermal range of this species.
These temperatures have also been shown to affect the phenotypes of this species
(Pavlidis et al., 2000; Koumoundouros et al., 2001, 2002). Temperature was kept con-
stant during the different treatments by the use of a thermo-regulating system (chillers
and heaters).
At the stage of half epiboly (30 h after fertilization at 15° C), 50 000 eggs were trans-
ferred into each tank, after an acclimation period of 2–3 h. Egg incubation, as well as
the maintenance of the yolk-sac larval, larval and juvenile stages were carried out as
described in Pavlidis et al. (2000) and Koumoundouros et al. (2001). Salinity was 30
and water temperature was 200 04 and 152 03° C (mean S.D.) for the two
different treatments. At 16–17 mm total length (LT) fish were transferred to the juvenile
tanks (500 l volume), where they were kept under common rearing conditions and
ambient water temperature (191 11° C mean S.D. daily temperature) for 8 weeks.
Survival during the larval stage was 28 and 64% at 15 and 20° C, respectively, while
during the juvenile stage, survival was >95% for both populations. The differences
between the mortality rates of the different temperature treatments were expressed at
the end of yolk-sac larval stage (4–7 days post-hatching), due to the differential nega-
tive mechanic effect of a surface skimmer, used to maintain the water surface free from
oily films and thus to allow the normal inflation of the swimbladder (Koumoundouros
et al., 2001).
SAMPLING
In order to study the effect of temperature on the body shape of D. labrax, specimens
were sampled and analysed at three different times: A) at the end of the different ther-
mal treatments (141 14 mm standard length, LS; metamorphosing larvae), B) at the
end of feeding with plankton (210 19 mm LS; 3 weeks after the end of the different
thermal treatments) and C) at the beginning of the juvenile stage (381 40 mm LS, 8
weeks after the end of the different thermal treatments). In total, 89 specimens (55 from
15° C and 34 from 20° C) were examined in sample A, 505 (255 from 15° C and 250
from 20° C) in sample B, and 93 (73 from 15° C and 20 from 20° C) in sample C.
Specimens from sample A and B were anaesthetized with ethylenglycol-mono-
phenylether (02–05 ml l1), photographed, and preserved individually in phosphate-
buffered 5% formalin (Markle, 1984). Each specimen was photographed with a digital
camera (Olympus Camedia C-3030 Zoom), attached to a stereoscopic microscope
(Olympus SZX 9). Due to the opaqueness of the juveniles’ body, specimens from
sample C after anaesthetization were radiographed (Koumoundouros et al., 2000)
and examined for the existence of vertebral malformations. All specimens with deform-
ities in the cephalic region, fins or vertebral column (Koumoundouros et al., 1997, 2001,
2002, 2004; Sfakianakis et al., 2004, 2006), as well as those with damaged body parts
due to sampling manipulations, were excluded from shape analysis.
Vertebra number and fin rays were counted after the double staining of the individ-
uals for bone and cartilage (Park & Kim, 1984). A total of 100 juveniles were stained
per temperature regime.
M O R P H O M E T R I C S A N D ST A T I S T I C A L A N A LY S I S
For the study of body shape plasticity, the method of geometric morphometrics was
used, which is a relatively new technique capable of visualizing shape differences, and
decomposing them into a uniform (describing stretching, compression or shearing of
the entire landmark configuration) and a non-uniform, localized (describing differences
of individual landmarks) component. Co-variation among parts or entire regions of the
body is taken into account by the model (Bookstein, 1991; Rohlf & Marcus, 1993; Loy
et al., 1998; Adams et al., 2004).
The data set of the analysis is composed of landmark measurements (in the form of
x, y and eventually, z co-ordinates) that are acquired at homologous anatomical points
of each specimen. Then, a series of algorithms is applied to the dataset of the individual
landmark configurations, to align specimens to a uniform body orientation and Cartesian
co-ordinates (superimposition), to adjust body size to a uniform scale, and finally to
discriminate size effects on shape variability. Geometric size is estimated in terms of
centroid size (the square root of the sum of squared distances of each landmark from
the specimen’s centroid). A consensus is estimated from the new landmark configura-
tions and used as reference to obtain residuals at each landmark location. Thin-plate
spline algorithm is then applied on the consensus and residuals to estimate a new set
of variables, which is referred to as the ‘weight matrix’. The ‘weight matrix’ contains
a) the shape parameters (‘partial warps’) needed to deform (‘warp’) the form of the con-
sensus configuration to each specimen (localized shape differences), and b) that part of
shape variability that can be modelled by an affine transformation (transformation after
which parallel lines remain parallel) (uniform shape differences). Shape differences
among specimens can then be described in terms of differences in the deformation grids
or landmark-based vector plots depicting the objects. The parameters describing these
deformations (partial warp scores) can be used as shape variables for statistical compar-
isons of variation in shape within and between populations. The theory of the applied
geometric morphometric methodology is described in detail elsewhere (Bookstein, 1991;
Rohlf and Marcus, 1993; Loy et al., 1998, 1999, 2000; Adams et al., 2004).
In the present study, on each fish image, a total of 13 (stage A, B) to 14 (stage C)
landmarks were collected by means of tpsDig software (v. 1.37) (Fig. 1). A generalized
least square method was applied to adjust for centroid size (CS) and to superimpose all
landmark configurations (Rohlf & Slice, 1990). Shape differences between populations
were tested by canonical variate analysis on the weight matrix. The weight matrix was
estimated with the thin-plate spline algorithm using the landmark data set (Loy et al.,
2000, tpsRelw software, v. 1.31). To obtain the deformation grids that represent the
inter-population shape differences, shape components were regressed on the canonical
variate scores by the use of tpsRegr software (v. 1.26). The relative contribution of
the non-uniform and uniform components to shape differentiation was estimated using
the respective percentages of variance explained by the regression model, when the lat-
ter is computed including both shape components or the non-uniform component only
(Loy et al., 1998, 1999). Software cited is available at http://Life.Bio.SUNYSB.edu/
morph/morph.html. Superimposition and size adjustment was performed by means of
IMP CoordGen6d software (www.canisius.edu/~sheets/morphsoft.html).
MANOVA was conducted to test the differences of the meristic traits between the
two thermal treatments. To individually compare the meristic characters between the
two thermal treatments applied, the Mann–Whitney U-test was used, after the appro-
priate correction for tied measurements. Kolmogorof–Smirnof statistic was used to
FIG. 1. (a). Landmarks collected on the in vivo photographed larvae. 1, Anterior tip of upper jaw; 2,
middle of the skull, on the vertical axis passing through the eye centre; 3, anterior base of the 1st
dorsal fin; 4, anterior base of the 2nd dorsal fin; 5, posterior base of the 2nd dorsal fin; 6, base of the
dorsal caudal lepidotrichium; 7, base of the ventral caudal lepidotrichium; 8, posterior base of the
anal fin; 9, anterior base of the anal fin; 10, anterior base of the left pelvic fin; 11, base of the central
caudal lepidotrichium; 12, anterior margin of the eye; 13, posterior margin of the eye. (b) Landmarks
collected on the x-rayed juveniles. 1, Anterior tip of pre-maxilla; 2, middle of the skull, at the distal
exterior margin of the eyes; 3, anterior supra-occipital base; 4, anterior base of the 1st dorsal fin; 5,
anterior base of the 2nd dorsal fin; 6, dorsal margin of the caudal peduncle, perpendicular to the
22nd vertebra; 7, dorsal base of the caudal fin; 8, distal tip of the hypural bones; 9, ventral base of the
caudal fin; 10, ventral margin of the caudal peduncle, perpendicular to the 22nd vertebra; 11,
anterior base of the anal fin; 12, base of the left pelvic fin; 13, ventral tip of cleithrum; 14, posterior
proximal tip of the angular.
overall compare the frequency distribution of the meristic characters between the two
thermal treatments (Zar, 1999).
RESULTS
B O D Y S H A P E P L A ST I C I T Y
At the end of the thermal treatments (sample A), morphometric analysis re-
vealed a significant effect of water temperature on the shape of sea bass larvae
(Wilks’ l ¼ 0141, P < 0001). The distribution of the larvae along the canon-
ical axis demonstrated a complete discrimination of the individuals of the two
populations (Fig. 2). This effect of ‘developmental’ temperature on sea bass
shape was also evident in the subsequent stages (i.e. at 3 and 8 weeks after
the end of the different thermal treatments). In both B and C samples, the sig-
nificance of the differences between the two populations was high: P < 0001
(Wilks’ l ¼ 0264) and P < 0001 (Wilks’ l ¼ 0517), respectively. In these
samples, the frequency distribution along the canonical axis clearly discrimi-
nated the two groups, with a slight overlapping of class (Figs 3 and 4). The
squared Mahalanobis distances between the mean canonical values of the
two populations decreased from sample A to sample B and C (3043 for sample
FIG. 2. Histogram of the canonical variate scores for the specimens ( , 20 or , 15° C early water
temperature) from sample A. Splines show the uniform (shearing or stretching, on the top) and the
non-uniform (localized shape differences, at the bottom) components of shape change relative to
extreme values of the canonical axis.
FIG. 3. Histogram of the canonical variate scores for the specimens from sample B (see Fig. 2).
A 1111 for sample B and 554 for sample C), indicating a decrease of shape
differences between fish reared at 15 or 20° C at advancing age.
Shape differences between the two populations were mainly due to local de-
formations of the body, since the uniform shape components represented only
38% (sample A), 19% (sample B) and 11% (sample C) of the between-group
shape variation. This result is visualized by the high overlap between the two
populations in the scatterplots of the uniform components (Fig. 5), as well
as by the relative decrease of uniform shape changes from the sample A to
B to C (Figs 2, 3 and 4).
Shape difference between the two populations at the first sample was mainly
attributed to a comparatively more compressed lateral profile of the 15° C fish
which in turn was determined by the proximal transposition of the landmarks
related to the bases of the dorsal and pelvic fins (landmarks 1, 3, 4, 5, 10;
Fig. 2). This dorso-ventral compression of the 15° C fish is also demonstrated
by the spline diagrams of the uniform shape components (Fig. 2). In sample B,
shape differences were mainly due to the posterior shift of the anterior dorsal
fin base, and the ventral shift of the pelvic and of the anterior anal fin base (in
20° C fish, landmarks 3, 10, 9; Fig. 3). In sample C, shape differences were
mainly due to the dorsal transposition of the pelvic fin base and the bases of
the dorsal fin were dorsally transposed (in 20° C fish, landmarks 4, 5, 12; Fig. 4).
Finally, in sample A and C, the 20° C population was characterized by a com-
paratively anterior-ventral shift of the upper jaw (landmark 1; Figs 2 and 4).
FIG. 4. Histogram of the canonical axis scores for the specimens from sample C (see Fig. 2).
MERISTIC CHARACTERS
The temperature regimes significantly affected the fin meristic characters
(Wilks’ l ¼ 0599, approximation F6,184, P < 0001, MANOVA). The number
of dorsal spines and caudal dermatotrichia was higher at 20° C, while fish
which had experienced 15° C had more pectoral and dorsal lepidotrichia. There
were no significant differences between the two temperature groups with
respect to the caudal lepidotrichia and the meristics of the anal fin (Table I
and Fig. 6; Mann–Whitney U-test). Similarly, pelvic fins and vertebral number
were unaffected by the temperature in early stages. All fish examined had one
pelvic spine, five pelvic lepidotrichia and 25 vertebrae (including urostyle).
DISCUSSION
The environment experienced by a fish has a significant influence on the
fish’s phenotype during the entire life cycle. It is well known that factors like
water temperature, salinity, swimming exercise, food availability and food con-
tent during various developmental periods can modify a fish’s external mor-
phology (Meyer, 1987; Blaxter, 1988; Corti et al., 1996; Loy et al., 1996;
Koumoundouros et al., 1999, 2001; Peres-Neto & Magnan, 2004), rate of ontog-
eny (Kinne & Kinne, 1962; Blaxter, 1969; Polo et al., 1991; Fuiman et al., 1998;
Klimogianni et al., 2004), gender (Pavlidis et al., 2000; Koumoundouros
et al., 2002; Piferrer et al., 2005), muscle anatomy and cellularity (Stickland
et al., 1988; Johnston, 1993; Galloway et al., 1999). The understanding of the
integrated character of phenotypic plasticity, however, requires the study
FIG. 5. Distribution of fish developed at 20 ( ) or 15° C ( ) early water temperature along the uniform
shape components, shearing (Uni X) and stretching (Uni Y) for samples (a) A, (b) B and (c) C. Left
and right splines represent the extreme values of Uni X. Top and bottom splines represent the
extreme values of Uni Y.
of the extent to which environment experienced in early life stages influences the
phenotype at subsequent life stages.
The effect of water temperature during embryonic and larval stages on the
morphology of sea bass juveniles was examined, and that early thermal history
significantly affected meristic characters and shape of the fish was clearly dem-
onstrated. The environmentally induced changes in body shape are not new for
fishes (Corti et al., 1996; Day & McPhail, 1996; Koumoundouros et al., 2001;
Robinson & Parsons, 2002; Peres-Neto & Magnan, 2004). This is the first time,
as far as is known, however, that shape plasticity is shown to be induced by the
temperature conditions applied during the egg and larval stages.
Previous studies demonstrate that the impact of early developmental condi-
tions decreases substantially as development proceeds, almost leading to morpho-
functional uniformity as soon as the fish enter the juvenile stage, e.g. when
environmental variability can be dealt with through behavioural choices of
the fish (Johnston et al., 2001; Johnston & Temple, 2002). In the present study,
shape differences between the two groups exposed to 15 and 20° C were not
stable from the end of the thermal treatments to the juvenile stage. Even
though in the entire study period these differences were mainly attributed to
non-uniform components (localized shape differences), from the end of the
thermal treatments up to the juvenile stage there was a clear and gradual
decrease of the weight of the uniform component (shearing or stretching).
FIG. 6. Frequency distribution of meristic characters in fish developed at 20 ( ) or 15° C ( ) for (a)
number of dorsal spines, (b) number of dorsal lepidotrichia, (c) number of lower caudal dermato-
trichia, (d) number of upper caudal dermatotrichia, (e) number of anal lepidotrichia and (f) number
of pectoral lepidotrichia. P-values from Kolmogorof–Smirnof tests.
of the pelvic fins. The development of a slender body at cold conditions could
be the mechanism to counteract the effects of low temperature in physiological
rates of the individuals and in kinematic viscosity of the medium in which they
move in (Fuiman & Batty, 1997; Johnson et al., 1998; Hunt von Herbing,
2002). Furthermore, a streamline fish form is related to reduced drag during
prolonged swimming which is potentially useful in searching for prey in the
planktonic habitat (Webb, 1984; Robinson & Parsons, 2002; Peres-Neto & Ma-
gnan, 2004). On the other hand, the lower orientation of the paired fins serves
to improve hovering and manoeuvrability, which may be useful in the complex
littoral habitats (Webb, 1984; Robinson & Parsons, 2002; Peres-Neto & Magnan,
2004).
Reproduction in sea bass takes place in winter, when water temperature is
between 11 and 15° C (Barnabe, 1976; Mendez et al., 1995; Mananos et al.,
1997), while embryonic and larval development can occur in the wider range
of 8–20° C (Marangos et al., 1986; Jennings & Pawson, 1992). Therefore, the
different temperature conditions that were tested in the present study fall inside
the natural spectrum of the species. Additionally, those conditions were applied
during the period (egg to the metamorphosis stage) where sea bass exhibits
maximum sensitivity to temperature sex determination (Blázquez et al., 1998;
Koumoundouros et al., 2002) and a large plasticity of the developmental pat-
tern (Koumoundouros et al., 2001). Based on the above, it could be suggested
that the phenotypic response of sea bass to developmental temperature, com-
bined with its extended spawning season, is the species’ strategy for regulating
population structure amidst annual and seasonal environmental fluctuations.
Further study of thermally driven phenotypic plasticity in other eurythermal
fish species, as well as in different populations of sea bass would contribute
to a better understanding of the role of water temperature on fish populations.
We wish to express our thanks to C. Chambers, A. Jordaan and the other two
anonymous reviewers for their valuable comments and suggestions. We also thank
S. Sfenthourakis for his comments on the manuscript, I. Papadakis for the maintenance
of the fish populations, and to C. Doxa for her participation in fish sampling and
staining. This study was supported by the European Commission, RTD programme,
QLRT-2000-01233, ‘Optimisation of rearing conditions in sea bass for eliminated
lordosis and improved musculoskeletal growth’ but it does not necessarily reflect the
Commission’s views nor anticipates its future policy in this area.
References
Adams, D. C., Rohlf, F. J. & Slice, D. E. (2004). Geometric morphometrics: ten years of
progress following the ‘revolution’. Italian Journal of Zoology 71, 5–16.
Barnabe, G. (1976). Contribution a la connaissance de la biologie du loup Dicentrarchus
labrax (Poisson Serranidae) de la region de Sète. These Doct Etat, Université de
Scienci et Technologie, Languedoc, Montpellier.
Blaxter, J. H. S. (1969). Development: eggs and larvae. In Fish Physiology, Vol. III (Hoar,
W. S. & Randall, D. J., eds), pp. 177–252. New York: Academic Press.
Blaxter, J. H. S. (1988). Pattern and variety in development. In Fish Physiology, Vol. XIA
(Hoar, W. S. & Randall, D. J., eds), pp 1–58. London: Academic Press.
Blaxter, J. H. S. (1992). The effect of temperature on larval fishes. Netherlands Journal of
Zoology 42, 336–357.
Blàzquez, M., Zanuy, S., Carillo, M. & Piferrer, F. (1998). Effects of rearing temperature
on sex differentiation in the European sea bass (Dicentrarchus labrax L.). Journal of
Experimental Zoology 281, 207–216.
Blàzquez, M., Carrillo, M., Zanuy, S. & Piferrer, F. (1999). Sex ratios in offspring of
sex-reversed sea bass and the relationship between growth and phenotypic
sex differentiation. Journal of Fish Biology 55, 916–930.
Bookstein, F. L. (1991). Morphometric Tools for Landmark Data: Geometry and Biology.
Cambridge: Cambridge University Press.
Brander, K. M., Dickson, R. R. & Shepherd, J. G. (2001). Modelling the timing of
plankton production and its effect on recruitment of cod (Gadus morhua). ICES
Journal of Marine Science 58, 962–966.
Campinho, M. A., Moutou, K. A. & Power, D. M. (2004). Temperature sensitivity of
skeletal ontogeny in Oreochromis mossambicus. Journal of Fish Biology 65, 1003–1025.
Chambers, R. C. & Leggett, W. C. (1996). Maternal influences on variation in egg sizes in
temperature marine fishes. American Zoologist 36, 180–196.
Corti, M., Loy, A. & Cataudella, S. (1996). Form changes in the sea bass, Dicentrarchus
labrax (Moronidae: Teleostei), after acclimation to freshwater: an analysis using
shape coordinates. Environmental Biology of Fishes 47, 165–175.
Day, T. & McPhail, J. D. (1996). The effect of behavioural and morphological plasticity
on foraging efficiency in the threespine stickleback (Gasterosteus sp.). Oecologia
108, 380–388.
Fowler, J. A. (1970). Control of vertebral number in teleosts-an embryological problem.
Quarterly Review of Biology 45, 148–167.
Fuiman, L. A. & Batty, R. S. (1997). What a drag it is getting cold: partitioning the
physical and physiological effects of temperature on fish swimming. Journal of
Experimental Biology 200, 1745–1755.
Fuiman, L. A., Poling, K. R. & Higgs, D. M. (1998). Quantifying developmental progress
for comparative studies of larval fishes. Copeia 1998, 602–611.
Galloway, T. F., Kjorsvik, E. & Kryvi, H. (1999). Muscle growth in yolk-sac larvae of the
Atlantic halibut as influenced by temperature in the egg and yolk-sac stage. Journal
of Fish Biology 55, 26–43.
Hare, A. H. & Cowen, R. K. (1997). Size, growth, development, and survival of the
planktonic larvae of Pomatomus saltatrix (Pisces: Pomatomidae). Ecology 78,
2415–2431.
Hunt von Herbing, I. (2002). Effects of temperature on larval fish swimming performance:
the importance of physics to physiology. Journal of Fish Biology 61, 865–876.
Imre, I., McLaughlin, R. L. & Noakes, D. L. G. (2002). Phenotypic plasticity in brook
charr: changes in caudal fin induced by water flow. Journal of Fish Biology 61,
1171–1181.
Jennings, S. & Pawson, M. G. (1992). The origin and recruitment of bass, Dicentrarchus
labrax, larvae to nursery areas. Journal of the Marine Biological Association of the
United Kingdom 72, 199–212.
Johnson, T. P., Cullum, A. J. & Bennett, A. F. (1998). Partitioning the effects of
temperature and kinematic viscosity on the C-start performance of adult fishes.
Journal of Experimental Biology 201, 2045–2051.
Johnston, I. A. (1993). Temperature influences muscle differentiation and the relative
timing of organogenesis in herring (Clupea harengus) larvae. Marine Biology 116,
363–379.
Johnston, I. A. & Temple, G. K. (2002). Thermal plasticity of skeletal muscle phenotype
in ectothermic vertebrates and its significance for locomotory behaviour. Journal of
Experimental Biology 205, 2305–2322.
Johnston, I. A., Vieira, V. L. A. & Temple, G. K. (2001). Functional consequences and
population differences in the developmental plasticity of muscle to temperature in
Atlantic herring Clupea harengus. Marine Ecology Progress Series 213, 285–300.
Jordaan, A., Hayhurst, S. E. & Kling, L. J. (2006). The influence of temperature on the
stage at hatch of laboratory reared Gadus morhua and implications for compar-
isons of length and morphology. Journal of Fish Biology 68, 7–24.