Aquaculture Research, 2013, 44, 289–299
Heritabilities and correlations of deformities and
growth-related traits in the European sea bass
(Dicentrarchus labrax, L) in four different sites
Bilge Karahan1, Béatrice Chatain2, Hervé Chavanne3, Alain Vergnet2, Agnès Bardon2,
Pierrick Haffray4, Mathilde Dupont-Nivet5 & Marc Vandeputte2,5
1
Aquaculture Department, Fisheries Faculty, Ege University, Izmir, Turkey
2
Ifremer UMR110 INTREPID, Chemin de Maguelone, Palavas-Les-Flots, France
3
Panittica Pugliese, Torre Canne di Fasano, Italy
4
Sysaaf, Section aquacole, Station SCRIBE, Rennes, France
5
INRA, UMR1313 Génétique Animale et Biologie Intégrative, Jouy-en-Josas, France
Correspondence: M Vandeputte, GDR Inra-Ifremer “Amélioration génétique des Poissons”, Chemin de Maguelone, F-34250 Pala-
vas les Flots, France. E-mail: marc.vandeputte@jouy.inra.fr
Abstract
Skeletal deformities are important traits for aqua-
culture as they induce slow growing and low
market value. We studied their genetic determin-
ism and their interactions with the environment
at the ongrowing stage in 5839 European sea
bass from a partial factorial mating of 33 sires
and 23 dams, reared in four sites. All families
were mixed, and fish were first reared in one site
(site B) until 35 g mean weight, then distributed
to the four sites. A posteriori reconstruction of
pedigree with microsatellites was used. Deformi-
ties were scored internally at slaughtering, and
externally from photographs. Site B, where all
fish were initially stocked until 35 g and exposed
to forced swimming because of fast water current
showed the highest rate of deformities with 83%
and 65% from internal and external scoring
respectively. Heritability on the underlying scale
was h2 = 0.25 ± 0.03 across all sites, and varied
little between sites, while genetic correlations of
deformities between sites were always high
(>0.85). Genetic correlations between deformities
and daily growth coefficient were variable
between sites(rA = 0.50 ± 0.09, 0.43 ± 0.10,
0.32 ± 0.10, 0.18 ± 0.10 for sites A, B, C, D
respectively) and were positively linked with the
average growth rate in each site. These results
pointed out that there could be a relation
between growth rate and the evolution of defor-
mities at the grow-out stage.
© 2012 Blackwell Publishing Ltd
doi:10.1111/j.1365-2109.2011.03082.
Keywords: European sea bass, Dicentrarchus lab-
rax, skeletal deformities, heritability, genetic cor-
relations, G 9 E interactions
Introduction
Selective breeding is increasingly being considered
as a major way to improve aquaculture practices
through the use of better performing populations.
When planning breeding strategies, it is of major
importance to point out the traits that are com-
mercially important. One of the most important
problems occurring with intensive aquaculture is a
significant proportion of fish with deformities: skel-
etal deformities (fins, spinal column and head;
Boglione, Marino, Gigati, Longobardi & Marzi
2009), pigmentation problems, scale malforma-
tions (Moretti, Criado, Cittolin & Guidastri 1999)
and swim bladder non-inflation (Divanach, Papan-
droulakis, Anastasiadis, Koumoundouros & Kento-
uri 1997) are all significantly decreasing the
commercial value of fish.
As deformed fish are less acceptable to consum-
ers, they are either discarded or must be further
processed (if possible) to save any of the meat.
Moreover, even fish with lesser deformities, that
may externally look normal, may be difficult to fil-
let, resulting in lower fillet yields. Fish with verte-
bral deformities therefore translate into financial
losses to fish farmers and throughout the value
chain (Gjerde, Pante Ma & Baeverfjord, 2005).
Besides the economic consequences due to the
289
Genetics of deformities in the sea bass B Karahan et al.
appearance of fish, deformities also reduce the
physiological capacities of cultured fish to perform
well: low growth rate, high mortality during any
stressing operation (anaesthesia, reducing oxygen
rate or another type of manipulation) can result of
skeletal deformities (Barahona-Fernandes 1982;
Chatain 1994). These anomalies also have impor-
tant implications from an animal welfare point of
view. It is therefore important to identify the
causes of vertebral deformities in order to reduce
their occurrence (Gjerde et al., 2005).
Several factors may cause skeleton deformities
in fish in natural and aquaculture conditions.
Environmental factors, such as embryo density
during incubation, broodstock stress, contaminants
in the water, fast water current in the larval rear-
ing tanks (thus muscle pressure on the spine, par-
ticularly during active swimming), salinity
changes, and inappropriate oxygen levels have
effects on deformities (Chatain 1994; Divanach
et al. 1997). Temperature, which is another envi-
ronmental factor affecting organogenesis, also has
indirect effect on deformities (Abdel, Abellan,
Lopez-Albors, Valdes, Nortes & Garcia-Alkazar
2004). Nutrition also affects development and in
particular skeletal formation. Phospholipids, unsat-
urated fatty acids, proteins and vitamins have
effect on both growth and skeletal development.
For example, a moderate level of protein hydroly-
sate improves skeletal development and in con-
trast, some other components like vitamin A have
toxic effect when overdosed (Cahu, Zambonino &
Takeuchi 2003). All the previously listed effects
occur mainly during the larval and post-larval
stages. Very little attention has been given to the
later stages of growth, where some studies still
showed that proportions of deformed fish may
evolve over time (Bardon, Vandeputte, Dupont-
Nivet, Chavanne, Haffray, Vergnet & Chatain
2009; Boglione & Costa 2011).
Genetics has also been shown to have an impact
on skeletal deformities. Genetic variation studies
have been carried out in several species; salmon
(McKay & Gjerde 1986; Gjerde et al., 2005; Kause,
Ritola & Paananen 2007), trout (Kause, Ritola,
Paananen, Wahlroos & Mäntysaari 2005), carp
(Kocour, Linhart & Vandeputte 2006), sea bream
(Castro, Pino-Querido, Hermida, Chavarrias,
Romero, Garcia-Cortes, Toro & Martinez 2008),
sea bass (Bardon et al. 2009), Atlantic cod (Kols-
tad, Thorland, Refstie & Gjerde 2006) and com-
mon sole (Sae-Lim 2008). The magnitude of the
290
Aquaculture Research, 2013, 44, 289–299
heritabilities estimated leads to think that vertebral
deformity is determined by a substantial additive
genetic component (Gjerde et al., 2005). When the
effects of environment on phenotype are reduced
to a minimum, estimating genetic variation is a
very important issue for aquaculture. It may also
be important if all environmental effects cannot be
controlled and differential genetic sensitivity to
these environmental stressors exists (Kause et al.
2007), which may then be expressed as genotype
by environment (G9E) interactions.
For estimating genetic parameters in fish, the
necessity to rear families separately until the tag-
ging size may bias or reduce the precision of the
estimated parameters. In this study, a posteriori
reconstruction of pedigrees with microsatellites
was used as an alternative to separate rearing of
families. The major benefits of this methodology
are the absence of between families environmental
effects, the possibility to work in industry condi-
tions with good experimental power, and to use
highly informative factorial designs (Vandeputte,
Dupont-Nivet, Chatain & Chevassus 2001).
According to Bardon et al. (2009) spine deformi-
ties in the European sea bass have a heritable
component that may indirectly describe a sensitiv-
ity to the particular pre-growing environment
endured by fish. In this study, we took advantage
of the high incidence of inadvertent spine deformi-
ties in a large set of mixed families of sea bass to
examine the genetic component of spine deformi-
ties, comparing four different ongrowing sites
under different environmental conditions. This
allowed studying the interactions between genetic
components and environmental effects on spine
deformities and late growth.
Materials and methods
Crossing, rearing and tagging
Collection of breeders, reproduction process and
larval rearing were previously described in detail
in Dupont-Nivet, Vandeputte, Vergnet, Merdy, Haf-
fray, Chavanne and Chatain (2008). The parents
of the studied animals were wild fish of Atlantic
origin. Reproduction and first nursing of fish took
place at Panittica Pugliese (Italy). Two hundred
and fifty-three full-sib families from 33 males and
23 females were produced according to a partly
factorial mating design. Crosses were conducted
with three sets of different parents: 11 males 9 9
© 2012 Blackwell Publishing Ltd, Aquaculture Research, 44, 289–299
Aquaculture Research, 2013, 44, 289–299
females, 11 males 9 7 females and 11 males 9 7
females. Within each set a full factorial crossing
was accomplished – thus 253 individual fertiliza-
tions were performed on the same day. Eggs were
incubated (one female per incubator) for 48 h and
then mixed by equal volumes to constitute a single
batch which was kept in the same tank for larval
rearing until day 64. Fish were transferred to a
concrete raceway from day 65 to day 130. At
134 days posthatch (about 4 g), a random sample
of 16 000 fish was sent to the Ifremer station in
Palavas (France) and pre-grown in a 5-m3 tank in
a semi-recirculation system (10–30% renewal/day,
18°C, 34& average salinity). At 156 days, the
batch of fish was split at random into four 5 m3
tanks to lower the density. These four tanks will
be referred as ‘initial tank’ in the statistical analy-
sis. At 370 days, fish had reached a mean weight
of 35 g and 7000 of those were randomly chosen,
individually PIT-tagged, individually measured
(body weight and length) and fin-clipped (kept in
90% ethanol for further DNA analyses). Four
batches of 1750 fish each were constituted and
each of the batches was kept in a 5-m3 tank prior
to distribution to four different farms.
The four farms were chosen for their varying
environmental conditions. The main rearing con-
ditions are reported in Table 1. One batch of fish
was reared from day 513 in a 216-m3 sea-cage in
tropical conditions at the Ardag fish farm (Israel,
site A). This batch had been kept in farm C from
day 423 to day 510, due to transportation prob-
lems. The second batch was kept in a recirculated
system at the Ifremer station in Palavas (France,
site B), in tanks where fish density was maintained
below 30 kg m
3. In this site, the temperature
was maintained at 20–22°C. A third batch arrived
at 423 days in Panittica Pugliese (Italy, site C)
where it was reared in a 12-m3 concrete raceway
supplied with 19°C (constant temperature) bore-
hole water. The last one arrived in Viveiro
Table 1 Growing conditions in the four rearing sites
Rearing
Sites period (days) Rearing system
A* 513–734 Floating cage in tropical waters
B 420–714 Semi-closed recirculation system
C 423–795 Concrete tank with borehole water
D 420–873 Semi-intensive estuarine
*Fish of farm A were reared in farm C during the period 423–510 days posthatching.
© 2012 Blackwell Publishing Ltd, Aquaculture Research, 44, 289–299
Genetics of deformities in the sea bass B Karahan et al.
Villanova (Portugal, site D) at 420 days and was
first reared in a 8-m3 tank and then transferred to
a semi-intensive estuarine pond at 588 days.
Data collection
In each farm, fish were individually measured at
commercial size, varying from 338 (site C) to 487 g
(site A) on average, and photographed on their left
side, on a white background, for later external deter-
mination of deformities. On harvest day, prior to
measurement, all fish were euthanized with an
excess dose of 2-phenoxyethanol (0.6 mL L
1) or
eugenol (0.1 mL L
1, site C). In farm B, the fish
were not euthanized but only anaesthetized
(0.3 mL L
1 2-phenoxyethanol). Each fish was
identified through tag reading, then weighed to the
nearest 0.1 g, and its length was measured to the
nearest mm. In sites A, C and D, internal deformities
were scored by observing the spine after removing
the left fillet of each fish. In site B, fish were further
reared to 800 g mean weight after the picture was
taken at 400 g. They were then slaughtered and
scored for internal deformities. In total, 5885 fish
were measured and scored for deformities from all
farms. Regarding deformities, incidence of lordosis
(saddleback), scoliosis (lateral curvature with the
rotated vetebrae), cyphosis (humpback) and fusion
(compressed vertebrae) were determined from pho-
tographs (external deformities) as well as direct
observation of the spine (internal deformities). The
scoring from photographs was done by one person,
meanwhile scoring of internal deformities was done
by a single person in sites B, C and D, and by the
same person plus two others in site A. While some
fish had only one of these deformities, others had
several at the same time. Whether they had one or
more kinds of deformities, they were assigned as
deformed and scored 1, while undeformed fish were
scored 0. All kinds of deformities were also assigned
and scored separately for each fish.
Temperature Volume Rearing density
(°C) (m3) (kg m
3)
22–27 216 <4
20–22 5(x4) <30
19–20 12 <46
9–25 400 <2
291
Genetics of deformities in the sea bass B Karahan et al.
Parentage assignment
DNA was extracted from tissue using Qiagen
Dneasy 96 well plates following a 12-h proteinase
K digestion. Parentage assignment was performed
using Landcatch Natural Selection (Scotland)
using six microsatellite markers organized in a sin-
gle PCR multiplex. Assignments were redrawn
using the assignment software VITASSIGN (Vande-
putte, Mauger & Dupont-Nivet 2006), allowing for
one mismatch to account for genotyping errors.
The assignment rate was very high (99.2%).
Statistical analysis
Deformity scores (0 or 1) for each individual and
each type of deformity were analysed by logistic
regression, using proc-LOGISTIC of SAS® to deter-
mine the potential significant fixed effects.
The individual growth traits studied were log
weight (log transformation was used to correct
scale effects between sites due to different slaugh-
tering ages as well as a skewed distribution of
body weights within sites), body length and DGC
(daily growth coefficient):
1=3 1=3
W2
W1
DGC1
2 ¼  100;
day2
day1
where W2 is the weight at commercial size, W1 is
the weight at tagging, and day2 and day1 are age
of fish (in days after hatching) where those
weights were recorded. DGC corrects for unequal
rearing times and slaughter weights between sites,
thus allowing better comparisons between sites.
Heritabilites and standard errors (h2 ± SE) of
internal and external deformities, as well as of
growth traits were obtained with a univariate ani-
mal model using VCE 6 (Groeneveld, Kovac & Mie-
lenz 2008):
Y ¼ Xb þ Za þ Wm þ e;
where Y is the vector of observations, b is the vec-
tor of fixed effects, a is the vector of random addi-
tive genetic effects, m is a vector of random effects
common to maternal half-sibs and e is the vector
of random residual effects. X, Z and W are known
incidence matrices. Fixed effects for deformities
were overall mean, site and initial tank (in site B,
prior to tagging) as the fish from each tank were
equally distributed to all sites. Sex had no signifi-
cant effect on deformities and was not further con-
292
Aquaculture Research, 2013, 44, 289–299
sidered. For growth traits, fixed effects included
overall mean, site, sex, initial tank, and a fixed
effect of deformities, in order to take into account
the possible phenotypic impact of deformities on
growth. Non-genetic maternal effects were not sig-
nificant for deformities, so they were removed from
the model when analysing deformity traits. Defor-
mities are threshold traits which are determined
by an underlying liability trait. In this case, herita-
bility calculated on the observed scale with the
animal model were adjusted by the correction of
Dempster and Lerner (1950), which is valid here
as the incidences were between 0.25 and 0.75 in
almost all cases (Van Vleck 1972).
Genetic correlations between internal and exter-
nal deformities, deformity types and growth-related
traits were estimated with multi-trait animal mod-
els, with the same effects as described for the uni-
variate models. The observed scale was used to
estimate the genetic correlations as genetic corre-
lations between binary traits have been shown to
be equal on the observed and on the underlying
scale (Gianola, Goffinet & Bulmer 1982), and as
genetic correlations between one binary trait on
the observed scale and a normally distributed trait
are very similar to the true correlations as long as
the threshold trait does not have both low herita-
bility and low incidence (Olausson & Ronningen
1975; Mercer & Hill 1984).
In order to estimate G9E interactions, traits in
the different sites were considered as different traits
and analysed in multi-traits animal models. Genetic
correlations between sites were used as a measure-
ment of G9E interaction, as the closer to unity the
genetic correlation, the lower the GxE interaction.
Results
Mean performances
There were no differences in weight at tagging
between the groups sent in the four sites. Large dif-
ferences in slaughter weight were recorded (487 g
in site A, 824 g in site B, 338 g in site C and 358 g
in site D), but they were related both to growth rate
and to differences in age. It was expected to slaugh-
ter all fish at 400 g (commercial size) but differences
occurred from one site to another because it is not
easy to precisely estimate growth in commercial
farms. However, DGC, which is corrected for the
gross effects of age, was also very different between
sites (site A: 1.25 ± 0.02, site B: 1.18 ± 0.01, site
© 2012 Blackwell Publishing Ltd, Aquaculture Research, 44, 289–299
Aquaculture Research, 2013, 44, 289–299 Genetics of deformities in the sea bass B Karahan et al.

C: 0.86 ± 0.01, site D: 0.76 ± 0.01). Survival was liosis and 50% of the fish had lordosis. Site A was
very high in sites A and C (95.7% and 94.8%, characterized by the largest difference between the
respectively), high in site B (84.2%) and lower in incidence of internal lordosis and of scoliosis (46%
site D (67.3%). and 6% respectively).
Proportion of internal and external deformities When external skeletal deformities were scored
as well as DGC of deformed and non-deformed fish on photographs (Table 2), the figures were compa-
for each site are given in Table 2. In total 62% of rable (although generally lower) to internal defor-
all fish examined had at least one or more kinds of mity scores in most cases, but in some cases
internal skeletal deformities, while the incidence differed significantly (e.g. 6% internal scoliosis and
was 54% when deformities were scored externally 25% external ones in site A), showing some inac-
from the photographs. The incidence of fish scored curacy of external evaluations. External measures
as ‘deformed’ was lower than the sum of the inci- were matching internal ones depending on the
dences of specific deformities (lordosis, scoliosis, cy- site: the phenotypic correlation between internal
phosis and fusion) as 11% of the fish suffered from and external deformity proportions was 0.62 for
several deformities. all sites. When both traits were compared site by
The incidence of internal deformities varied from site, they were the most correlated in site A (0.70)
53% (site D) to 82% (site B), while the incidence and the least correlated (0.46) in site B.
of external deformities varied from 45% (site C) to Daily growth coefficients of deformed and unde-
65% (site B). formed fish were similar (P > 0.05), except in site
Lordosis was the most common deformity (43% A where non-deformed fish grew significantly fas-
average incidence), followed by scoliosis with 30% ter than deformed fish (Table 2).
incidence. Fusion and cyphosis were also observed in
the sample but their incidence was very low (<3%).
Heritabilities
Considering internal deformities, three sites (A,
C and D) showed a proportion of lordosis higher The heritability (liability scale) of both internal and
than that of the other deformities (46%, 35% and external deformities was 0.25 ± 0.03 when all sites
40% respectively). In site B, with the highest inci- were considered (Table 3). The range of heritabili-
dence of total deformities, 74% of the fish had sco- ties among sites was narrow for internal deformi-
ties (0.22–0.30) and wider for external deformities
(0.16–0.28). Interestingly, site D had both the
Table 2 Incidence of deformities site by site (% affected
fish) and DGCs highest heritability for internal deformities (0.30)
and the lowest one for external deformities (0.16).
A B C D All sites The heritability of DGC was low (0.17 ± 0.03)
when all sites were considered; however, when
Internal
estimated site by site, the heritability of DGC was
Lordosis 46 50 35 40 43
Scoliosis 5.7 74 22 19 30
0.23 ± 0.05 in site A, 0.25 ± 0.06 in site B,
Fusion 0.9 1.2 1.6 2.2 1.4 0.32 ± 0.06 in site C and 0.45 ± 0.08 in site D.
Cyphosis 0.9 0.7 0.4 0.5 0.7 DGC was the main growth trait considered as it is
Deformed* 53 82 57 54 62 the one that best represents differences in growth
DGC+ 1.23a 1.18a 0.87a 0.76a 1.01a
rate between sites differing in slaughtering ages
DGC
1.27b 1.19a 0.86a 0.76a 1.02a
External
and environmental conditions.
Lordosis 48 45 36 45 43
Scoliosis 25 44 25 17 28
Fusion 0.8 0.5 1.0 2.0 1.0
Interactions between traits
Cyphosis 0 0.4 0.8 1.8 0.7
Genetic correlations between internal deformities
Deformed* 54 65 45 52 54
DGC+ 1.23a 1.17a 0.87a 0.76a 1.01a
(Table 4) and growth traits were either not signifi-
DGC
1.27b 1.20a 0.86a 0.76a 1.02a cant or positive, meaning that the occurrence of
deformities is expected to increase as a conse-
*Deformed fish are fish with at least one kind of spine defor-
quence of selection for higher growth. When all
mity.
+, Deformed fish;
, non-deformed fish. Different superscript
sites were considered, the genetic correlation
letters within the same column indicate significant differences between DGC and internal deformities was signifi-
between DGC of deformed and non-deformed fish (P < 0.05). cant (0.34 ± 0.09 – Table 4), although quite

© 2012 Blackwell Publishing Ltd, Aquaculture Research, 44, 289–299 293

Genetics of deformities in the sea bass B Karahan et al.
Table 3 Heritabilities and standard errors (h2 ± SE) of internal and external deformities on the observed and on the un-
derliying liability scale (transformed according to Dempster & Lerner 1950)
Sites
A B
Deformities
Idef observed 0.17 ± 0.03 0.10
Idef* liability 0.27 ± 0.05 0.22
Edef observed 0.16 ± 0.01 0.17
Edef* liability 0.24 ± 0.05 0.28
*Corrected by Dempster and Lerner (1950).
Idef observed, observed internal deformities; Idef liability, internal deformities on liability scale; Edef observed, observed external
deformities; Edef liability, external deformities on liability scale.
different between sites (0.50 ± 0.09, 0.43 ± 0.10,
0.32 ± 0.10 and 0.18 ± 0.10 in sites A, B, C, D
respectively).
Log weight also showed some level of genetic
correlation with the internal deformities. It was not
significant at the global level (0.15 ± 0.09), but
was positively correlated with deformities in sites A
and B. The pattern was similar for body length.
When external deformities were considered,
again only genetic correlations with DGC were sig-
nificant at the global level. The main difference
with internal deformities was that genetic correla-
tions with growth traits was not significant in site
B, due to the possible errors of the external scor-
ing. Correlations between external deformities and
log weight and length were not significant in most
cases (Table 4), except for a negative correlation
with tagging weight in site D (
0.27 ± 0.10).
GxE interactions
Additive genetic correlations of deformities between
different environments are presented in Table 5.
Genetic correlations between sites were all high for
internal deformities, varying between 0.87 ± 0.01
(e.g. B–C) and 0.99 ± 0.07 (sites A–C) for internal,
and varied from 0.96 ± 0.07 (e.g. A–C) to 1 ±
0.0003 (e.g. A–D) for external deformities.
Discussion
Incidences and heritabilities
In hatchery conditions the main objective is to
increase survival rates and growth. Even though
the occurrence of high number of deformities is
not unexpected (Andrades, Becerra & Fernandez-
Llebrez 1996), our experimental population was
294
Aquaculture Research, 2013, 44, 289–299
C D All Sites
± 0.02 0.16 ± 0.03 0.19 ± 0.03 0.15 ± 0.02
± 0.04 0.26 ± 0.05 0.30 ± 0.05 0.25 ± 0.03
± 0.01 0.18 ± 0.02 0.11 ± 0.01 0.13 ± 0.01
± 0.05 0.28 ± 0.05 0.16 ± 0.03 0.25 ± 0.03
unusual with high deformity rate. Osteological
development in larval fish is a detailed process that
begins with the formation of cartilage prior to ossi-
fication (Faustino & Power, 1998) and skeletal
deformities are known to occur during larval
development and osteogenesis (Chatain 1994; Div-
anach, Boglione, Menu, Koumoundouros, Kentouri
& Cataudella 1996; Afonso, Montero, Robania,
Astora, Izquierdo & Gines 2000).
Here, the cause of the deformities was appar-
ently not to be sought during larval rearing, as
the fish that were kept by farm C for its breeding
programme, which were from the same batch, did
not suffer (at least externally) from such deformi-
ties. The probable cause is the rearing conditions
in farm B, prior to tagging. Indeed, the small fish
that arrived from farm C (134 days, 3.6 g mean
weight) were reared in 5 m3 circular tanks, where
a strong circular water current was induced for
tank self-cleaning (Bardon et al. 2009), and it is
known that the intensive swimming provoked by
such water current is not suitable for this size of
fish that have not completed their bone calcifica-
tion (Chatain 1994; Divanach et al. 1997). In the
present study, the evolution of the deformity ratio
in different ongrowing sites was investigated, as it
is known that proportions of deformed fish may
evolve over time (Bardon et al.2009; Boglione &
Costa 2011). We could show that the final propor-
tions of deformed fish were different among sites,
although they all came from the same source pop-
ulation.
Although differences between sites were
reduced when deformities were scored based on
the photographs, the proportion of deformities in
the experimental populations varied depending on
the site for both scoring procedures (internal or
external). As all fish were in the same site when
© 2012 Blackwell Publishing Ltd, Aquaculture Research, 44, 289–299
Aquaculture Research, 2013, 44, 289–299 Genetics of deformities in the sea bass B Karahan et al.

the deforming conditions (strong water current)

0.10
0.09
0.10
0.10
were applied, this difference is probably due to a

±
±
±
±
0.1
0.28

0.05

0.06
differential evolution of the initial deformities

All
which may be induced by the different rearing
conditions in each site (Table 1). Therefore, the
G9E examined here is only that of the evolution

0.10
0.10
0.03
0.10
of deformities over time. Genetic by environment

±
±
±
±
0.18

0.27

0.15

0.14
interactions in early life stage of fish, where defor-

D
mities are initiated, should also be examined in
future studies. As mortalities varied among sites, it
cannot be excluded a priori that some of the differ-
0.10
0.10
0.10
0.10
±
ences observed were caused by differential mortali-
±
±
±
0.40

0.11

0.03

0.03
ties of deformed fish. However, this cannot be a
major factor explaining the results as sites A and
C

C, where survival was close to 100%, have the

same deformity rate as site D which was the most
0.10
0.10
0.10
0.09

selective (67% survival), and would have been

±
±
±
±

expected to display a lower deformity rate under a

0.10
0.08

0.08

0.07

simple model where mortality removes the

B

deformed fish. This simple model also does not fit

results in site B, where the deformity rate is the
0.10
0.02
0.10
0.10

highest although there is a significant mortality

External

±
±
±
±

rate (16%).
0.46

0.02
0.15
0.14
Table 4 Genetic correlations between growth traits and internal and external deformities, in all sites

One important bias to deformities incidence

A

between sites is the person scoring the deformities.

There is no possible bias for external deformities as
0.09
0.09
0.09
0.09

the same person scored all the photographs, while

±
±
±
±

for external deformities there could be differences

0.34
0.15
0.15
0.07
All

between site A (scored by three persons) and sites

B, C, D (all scored by a single person). However,
0.10
0.10
0.10
0.10

this should not be a major bias as again site A is

not one of the extremes in terms of incidence and
±
±
±
±
0.18

0.12
0.04
0.04

figures of internal and external deformities are

D

quite coherent.
The heritability for skeletal deformities is
LogWt, log weight at tagging; LogW, log weight at slaughtering.

0.25 ± 0.03 for all sites on the underlying scale, a

0.10
0.10
0.10
0.10

figure comparable with a previous estimate made

±
±
±
±

in site B only (0.21 ± 0.04) by Bardon et al.

0.32

0.05
0.1
0.1
C

(2009). This result is in the range of heritability

estimates for deformities in other species such as
0.36 ± 0.14 for Atlantic salmon (Gjerde et al.,
0.10
0.10
0.10
0.10

2005), 0.64 for sire and 0.36 for dam components

±
±
±
±
0.43
0.20
0.21
0.21

of variance on the underlying liability scale in

B

another salmon study (McKay & Gjerde 1986) and

0.27 in cod (Kolstad et al. 2006). Simulation stud-
0.09
0.10
0.10
0.10

ies (Van Vleck 1972) and experimental data

Internal

(Kause et al. 2007) demonstrated that heritabilities

±
±
±
±
0.50

0.08
0.24
0.23

for threshold characters could be downbiased

A

when the proportion of abnormal fish is below 5%

which is not the case in this study. However sea
Length
LogWt

bream showed non-significant heritability obtained

LogW
DGC

adjusting an animal model for lordosis

© 2012 Blackwell Publishing Ltd, Aquaculture Research, 44, 289–299 295

Genetics of deformities in the sea bass B Karahan et al.
Table 5 Genetic correlations (RA) of deformities between sites (%)
Internal deformities
B C D B
A 0.89 ± 0.09 1.00 ± 0.07 0.91 ± 0.06
B 0.87 ± 0.01 0.87 ± 0.09
C 0.89 ± 0.07
(0.021 ± 0.019) in an experimental study, even
though the proportion of abnormal fish was over
5% (Castro et al. 2008). This difference could be
explained by the fact that the environmental fac-
tors inducing deformities can be variable, and the
sensitivity of fish to these factors can also be vari-
able, so that even in the same species, heritability
estimates between years can be highly variable
(Kause et al. 2007).
Interactions between internal and external
deformities
For external examination from the photographs,
there was a standardization because only one per-
son made the observations, while for internal
determination site by site, there were three differ-
ent people in site A and one person for each of the
other sites. In such cases to prevent human error,
it is important to have one examinator for all sam-
ples. Internal and external deformity rates were
very close to each other in sites A and C, and dif-
fered more in sites B and D. Thus for such a
research, if it is not possible that only one person
can observe internal deformities for all fish, then it
may be more reliable to observe fish externally
from the photographs by only one person. Other
advantages of external scoring is that it does not
require to kill the fish, and that it represents the
final customer’s point of view. However, its effi-
ciency might be highly dependent on the growth
stage and on the person doing the observation.
Additionally, some deformed fish remain unde-
tected by external observation as proven by the
sometimes low phenotypic correlation between
internal and external deformities (0.46 in site B),
thus potentially reducing the efficiency of individ-
ual selection against deformities, as pointed out by
Bardon et al. (2009).
As the fish were chosen at random to constitute
the different farm batches, we can make the
hypothesis that the rate of deformities was initially
the same in all batches. The differences observed
296 © 2012 Blackwell Publishing Ltd, Aquaculture Research, 44, 289–299
Aquaculture Research, 2013, 44, 289–299
External deformities
C D
1.00 ± 0.00 0.98 ± 0.04 1.00 ± 0.00
0.99 ± 0.05 0.96 ± 0.07
1.00 ± 0.07
at slaughter then should come from environmental
effects of the rearing systems, allowing the fish to
recover or not, or worsening or not the initial
deformities (Dupont-Nivet, Nomm-Karahan,
Vergnet, Merdy, Haffray, Chavanne, Chatain &
Vandeputte 2010). The much higher proportion of
internally deformed fish in farm B is probably due
to the larger size (800 g) at scoring.
When two groups (A and C) of fish were kept at
the same place for longer time, the incidences of
total deformities in both sites were very close (53%
in site A and 57% in site C). The result of the
highest incidence in site B may indicate that
occurrence of deformities at later stages depends
on different environmental conditions, although
fish are from same source.
Estimated heritability for deformities in site B
was the lowest among all sites. Due to high inci-
dence at this site, more families gave deformed off-
spring and thus heritability estimation at the site
was less precise, and in this case the Dempster
and Lerner correction is less efficient due to an
incidence over 0.75 (Van Vleck 1972).
Growth traits and interactions with deformities
Growth rate in our study was very variable
between sites, and strongly correlated with tem-
perature, which is one of the main drivers of
growth rate in fish (e.g. Imsland & Jonsdottir
2003 on cod).
Daily growth coefficient, which indicates growth
rate, seemed to have an important impact on
deformities both at the site level (e.g. more defor-
mities in sites A and B where growth was faster)
and at the genetic level (positive genetic correla-
tions – as already showed in cod by Kolstad et al.
2006) but not at the individual level as the DGC
of deformed and non-deformed fish was similar in
a given environment (showing the absence of phe-
notypic correlation between late growth and defor-
mities), except in site A, where it could however
be linked to a deformity scoring problem (different
Aquaculture Research, 2013, 44, 289–299
scorers). This could nevertheless be the result of
an increased aggravation of deformities in fast-
growing individuals and families (especially in sites
with fast growth), compensated by a lower growth
rate of the deformed individuals (i.e. lower than it
would have been without deformities) resulting in
equal late growth rate of both deformed and unde-
formed fish. In this study, males showed different
sexual maturation levels, and maturing males –
which were also the fastest growing males – had
higher deformity rates than non-maturing males
(+10% on average). However, this seems likely
linked to their faster growth rate, as it was shown
before in cod that there was no genetic correlation
between sex maturity and growth (Kolstad et al.
2006).
Interestingly, the heritability of growth rate
(measured as DGC) seemed to be inversely corre-
lated with temperature, with the highest estimate
being obtained in the coldest site (site D). At the
same time, the genetic correlations between defor-
mities and DGC seemed to be the lowest also in
site D (Fig. 1). Although standard errors were
large, genetic correlations between DGC and defor-
mities were significant (P  0.01).
Most previous studies showed that deformed fish
grow slower (Al-Harbi 2001; Fjelldal, Hansen &
Berg 2007), but at the same time high temperature
during early stages favours the development of
abnormalities due to the acceleration in develop-
ment rate it induces (Abdel et al. 2004). In this
study, we observed that high growth rate (probably
mediated by high temperatures) also affects skeletal
deformities, in later stages. The simplest explana-
0.6
Mean DGC h2 int
h2 DGC rA
0.5
0.4
0.3
0.2
0.1
0 common carp Cyprinus carpio L. Asian Fisheries Science
A B C D
Figure 1 Variations among sites of mean daily growth
co-efficiency (DGC), heritability (h²) of DGC and internal
deformities and genetic correlation (rA) between inter-
nal deformities and DGC.
© 2012 Blackwell Publishing Ltd, Aquaculture Research, 44, 289–299
Genetics of deformities in the sea bass B Karahan et al.
tion is that fast growth leads to an aggravation of
the deformities present. Figure 1 shows that when
mean DGC decreases the heritability of DGC tends
to increase, and its genetic correlation with defor-
mities on the contrary tends to decrease, while the
heritability of deformities seems little affected.
When growth rate is very high, less genetic varia-
tion for growth is present, but still the fastest grow-
ers will have a higher propensity to suffer from
aggravated or late-appearance deformities. On the
contrary, when growth rate is low due to low
water temperature, there is a higher genetic varia-
tion for growth, but the fastest growers probably
do not reach a growth rate which can either
aggravate or create late deformities. Then, the pos-
sible unfavourable impact of selection for growth
on deformities will have to be specially monitored
in sites with high growth rates, where the risk is
higher to have an increase in deformities as a cor-
related response to selection for fast growth.
Acknowledgments
This work was carried out in the framework of a
CRAFT project no. Q5CR-2002-71720 (Heritabo-
lum) funded by the EC and the private farms
Panittica Pugliese (Italy), Viveiro Villanova (Portu-
gal) and Ardag (Israel), which are also thanked for
their active participation. The present work was
also part of the programme of the Research Group
‘Fish Genetic Improvement’ between INRA and Ifr-
emer.
References
Abdel I., Abellan E., Lopez-Albors O., Valdes P., Nortes
M.J. & Garcia-Alkazar A. (2004) Abnormalities in the
juvenile stage of sea bass (Dicentrarchus labrax L.)
reared at different temperatures: types, prevalence and
effect on growth. Aquaculture International 12, 523–
538.
Afonso J.M., Montero D., Robania L., Astora N., Izquierdo
M.S. & Gines R. (2000) Association of a lordosis-scolio-
sis-cyphosis deformity in gilthead seabream (Sparus au-
rata) with family structure. Fish Physiology and
Biochemistry 22, 159–163.
Al-Harbi A.H. (2001) Skeletal deformities in cultured
14, 247–254.
Andrades J.A., Becerra J. & Fernandez-Llebrez P. (1996)
Skeletal deformities in larval, juvenile and adult stages
of cultured gilthead sea bream (Sparus aurata L.). Aqua-
culture 141, 1–11.
297
Genetics of deformities in the sea bass B Karahan et al.
Barahona-Fernandes M.H. (1982) Body deformity in
hatchery reared European seabass Dicentrarchus labrax
(L.). Types, prevalence and effect on fish survival. Jour-
nal of Fish Biology 21, 239–249.
Bardon A., Vandeputte M., Dupont-Nivet M., Chavanne
H., Haffray P., Vergnet A. & Chatain B. (2009) What
is the heritable component of spinal deformities in the
European sea bass (Dicentrarchus labrax)? Aquaculture
294, 194–201.
Boglione C. & Costa C. (2011) Skeletal deformities and
juvenile quality. In: Sparidae: Biology and Aquaculture of
Gilthead Sea Bream and Other Species (ed. by M. Pavlidis
& C. Mylonas), pp. 233–294. Blackwell, Oxford, UK.
ISBN: 978-1-4051-9772-4
Boglione C., Marino G., Gigati M., Longobardi A. & Marzi
P. (2009) Skeletal anomalies in dusky grouper Ephin-
ephelus marginatus (Lowe, 1834) juveniles reared with
different methodologies and larval densities. Aquacul-
ture 291, 48–60.
Cahu C., Zambonino J.I. & Takeuchi T. (2003) Nutri-
tional components affecting skeletal development in
fish larvae. Aquaculture 227, 245–258.
Castro J., Pino-Querido A., Hermida M., Chavarrias D.,
Romero R., Garcia-Cortes L.A., Toro M.A. & Martinez
P. (2008) Heritability of skeleton abnormalities (lordo-
sis, lack of operculum) in gilthead seabream (Sparus
aurata) supported by microsatellite family data. Aqua-
culture 279, 18–22.
Chatain B. (1994) Abnormal swimbladder development
and lordosis in sea bass (Dicentrarchus labrax) and sea
bream (Sparus auratus). Aquaculture 119, 371–379.
Dempster R.E. & Lerner I.M. (1950) Heritability of
threshold characters. Genetics 35, 212.
Divanach P., Boglione C., Menu B., Koumoundouros G.,
Kentouri M. & Cataudella S. (1996) Abnormalities in
finfish mariculture: an overview of the problem, causes
and solutions. In: Seabass and Seabream Culture: Prob-
lems and Prospects (ed. by B. Chatain, M. Saroglia, J.
Sweetman & P. Lavens), pp. 45–66. European Aqua-
culture Society, Oostende.
Divanach P., Papandroulakis N., Anastasiadis P., Koumo-
undouros G. & Kentouri M. (1997) Effect of water cur-
rents on the development of skeletal deformities in sea
bass (Dicentrarchus labrax L.) with functional swimblad-
der during postlarval and nursery phase. Aquaculture
156, 145–155.
Dupont-Nivet M., Vandeputte M., Vergnet A., Merdy O.,
Haffray P., Chavanne H. & Chatain B. (2008)
Heritabilities and GxE interactions for growth in the
European sea bass (Dicentrarchus labrax L.) using a
marker-based pedigree. Aquaculture 275, 81–87.
Dupont-Nivet M., Nomm-Karahan B., Vergnet A., Merdy
O., Haffray P., Chavanne H., Chatain B. & Vandeputte
M. (2010) G9E interactions for growth in European
seabass (Dicentrarchus labrax) are large when growth
298 © 2012 Blackwell Publishing Ltd, Aquaculture Research, 44, 289–299
Aquaculture Research, 2013, 44, 289–299
rate rather than weight is considered. Aquaculture
306, 365–368.
Faustino M. & Power D.M. (1998) Development of osteo-
logical structures in the sea bream: vertebral column
and caudal fin complex. Journal of Fish Biology 52, 11
–24.
Fjelldal P.G., Hansen T.J. & Berg A.E. (2007) A radiologi-
cal study on the development of vertebral deformities
in cultured Atlantic salmon (Salmo salar L). Aquaculture
273, 721–728.
Gianola D., Goffinet B. & Bulmer M.G. (1982) Query: Sire
evaluation with best linear unbiased predictors. Bio-
metrics 38, 1085–1088.
Gjerde B., Pante Ma J.R. & Baeverfjord G. (2005) Genetic
variation for a vertebral deformity in Atlantic salmon
(Salmo salar). Aquaculture 244, 77–87.
Groeneveld E., Kovac M. & Mielenz N. (2008) VCE User’s
Guide and Reference Manual Version 6.0. Friedrich Loef-
fler Institute, Neustadt, Germany, 125pp.
Imsland A.K. & Jonsdottir O.D.B. (2003) Linking popula-
tion genetics and growth properties of Atlantic cod.
Reviews in Fish Biology and Fisheries 13, 1–26.
Kause A., Ritola O., Paananen T., Wahlroos H. &
Mäntysaari E.A. (2005) Genetic trends in growth, sex-
ual maturity and skeletal deformations, and rate of
inbreeding in a breeding programme for rainbow trout.
Aquaculture 247, 177–187.
Kause A., Ritola O. & Paananen T. (2007) Changes in
the expression of genetic characteristics across cohorts
in skeletal deformations of farmed salmonids. Genetics
Selection Evoution 39, 529–543.
Kocour M., Linhart O. & Vandeputte M. (2006) Mouth
and fin deformities in common carp: is there a genetic
basis? Aquaculture Research 37, 419–422.
Kolstad K., Thorland I., Refstie T. & Gjerde B. (2006)
Body weight, sexual maturity and spinal deformity in
strains and families of Atlantic cod (Gadus morhua) at
two years of age at different locations along the Nor-
wegian coast. ICES Journal of Marine Sciences 63, 246–
252.
McKay L.R. & Gjerde B. (1986) Genetic variation for a
spinal deformity in Atlantic salmon, Salmo salar. Aqua-
culture 52, 263–272.
Mercer J.T. & Hill W.G. (1984) Estimation of genetic
parameters for skeletal defects in broiler chickens.
Heredity 53, 193–203.
Moretti A., Criado M.P.F., Cittolin G. & Guidastri R.
(1999) Manual on Hatchery Production of Sea Bass and
Gilthead Sea Bream-Volume 1. FAO Corporate Document
Repository.
Olausson A. & Ronningen K. (1975) Estimation of
genetic parameters for threshold characters. Acta Agri-
culture Scandinavica 25, 200–208.
Sae-Lim P. (2008) Estimation of genetic parameters of
common sole Solea solea (Linnaeus, 1758) using
Aquaculture Research, 2013, 44, 289–299 Genetics of deformities in the sea bass B Karahan et al.

pedigree reconstruction. MSc thesis, Wageningen Uni-
versity, Wageningen, the Netherlands.
Van Vleck L.D. (1972) Estimation of heritability of thresh-
old characters. Journal of Dairy Science 55, 218–225.
Vandeputte M., Dupont-Nivet M., Chatain B. & Chevassus
B. (2001) Setting up a strain testing design for the
seabass, Dicentrarchus labrax: a simulation study. Aqua-
culture 202, 329–342.
Vandeputte M., Mauger S. & Dupont-Nivet M. (2006) An
evaluation of allowing for mismatches as a way to
manage genotyping errors in parentage assignment by
exclusion. Molecular Ecology, Notes 6, 265–267.

© 2012 Blackwell Publishing Ltd, Aquaculture Research, 44, 289–299 299