AQUATIC BIOLOGY

Aquat Biol

Vol. 18: 6980, 2013

doi: 10.3354/ab00490

Published online March 5

Population subdivision of saddled seabream

Oblada melanura in the Aegean Sea revealed
by genetic and morphometric analyses
G. A. Gkafas1, C. Tsigenopoulos2, A. Magoulas2, P. Panagiotaki1, D. Vafidis1,
Z. Mamuris3, A. Exadactylos1,*
1

Department of Ichthyology and Aquatic Environment, School of Agricultural Sciences, University of Thessaly, Fytokou str.,
38446, Volos, Greece
2

Institute of Marine Biology and Genetics, Hellenic Centre of Marine Research, Heraklion, Crete 71003, Greece

Department of Biochemistry and Biotechnology, School of Health Sciences, University of Thessaly, 26 Ploutonos & Aiolou str.,
41221, Larissa, Greece

ABSTRACT: The population structure of marine fishes is often cryptic, due to limited knowledge
about life history and ecological habitats. Understanding environmental stability is a challenge to
retrieve possible structuring patterns responsible for dispersal and demographic habits. In the
present study, a combination of genetic and phenotypic assessments was applied in order to investigate population structure of the saddled seabream Oblada melanura in the Aegean Sea, eastern
Mediterranean basin. Analyses of 6 microsatellite loci and 15 morphometric characters revealed
that saddled seabream individuals differ among northern, southern and central Aegean populations. These observed boundaries may be related to the differentiated seascape of the Aegean
Sea, suggesting that oceanographic factors are a significant stressor for population subdivision of
the saddled seabream. Individual-based landscape genetic approaches and multivariate analysis
of the morphometric characters suggest the presence of habitat-related limitations of saddled
seabream dispersal potential. Molecular genetics and phenotypic analyses along with life-history
traits provide useful informative data for the management and conservation schemes applied for
this species in the Aegean Sea.
KEY WORDS: Oblada melanura Microsatellite Morphometrics Aegean Sea Population structure
Resale or republication not permitted without written consent of the publisher

The marine environment forms a continuous water

body that does not exhibit clear barriers. Although,
due to a lack of barriers, homogeneity is expected in
marine species that have a large geographic range
and high dispersal, a fine-scale population structure
is often reported for these species (Bekkevold et
al. 2005, Jrgensen et al. 2005, Hoelzel et al. 2007,
Hoelzel 2009, Kelly & Palumbi 2010, Pilot et al. 2010).
Significant genetic differentiation has been reported
for marine fishes, the extent of which varies among

species (Arculeo et al. 2003, Pinera et al. 2007, Zulliger et al. 2009, Gonzlez-Wangemert et al. 2011).
Structural patterns are not always correlated with geographic components, but are the result of complex
demographic stochasticity, such as local resources,
social coherence and environmental stability (Galarza
et al. 2009, McCusker & Bentzen 2010, Salmenkova
2011). Despite the fact that oceanographic features,
such as water temperature, salinity and oxygen level,
are known to play a key role in species habitats,
the mechanisms driving the dispersal and population
structure of marine fishes remain unclear.

*Corresponding author. Email: [email protected]

Inter-Research 2013 www.int-res.com

INTRODUCTION

70

Aquat Biol 18: 6980, 2013

The Mediterranean Sea is a closed basin ecosystem, with a relative high temperature and different
seasonal rainfall patterns and flow dynamics underlying a highly oligotrophic area (Siokou-Frangou et
al. 2002). Spatial and temporal genetic population
structure studies in the Mediterranean Sea report
differentiation between the eastern and western part
of the basin (Vias et al. 2004, Rolland et al. 2007,
Coll et al. 2010), possibly due to historical demographic processes but also due to hydrological and ecological traits (e.g. Magoulas et al. 2006). The majority
of these studies have focused on fish stock exploitation through the population structure of the species
in question, revealing patterns via life-history traits
and biogeographic barriers. Variations in oceanographic features play a key role in the life history of
such species, most obviously on larval dispersal due
to food availability. Species with no pelagic larval
dispersal tend to evolve as highly subdivided populations, suggesting low gene flow between them
(Palumbi 1994), whereas species with longer pelagic
larval duration are expected to present higher gene
flow values and lower population differentiation
(Bohonak 1999).
The saddled seabream Oblada melanura L. belongs to the family Sparidae and occurs in the Mediterranean Sea, eastern Atlantic, Biscay Bay and from
the Strait of Gibraltar to Angola. It is also recorded
in Madeira, Cape Verde and the Canary Islands
(Bauchot & Hureau 1986, Bauchot et al. 1981). It is a
gregarious species, living in coastal regions, on rocky
and seaweed (Posidonia oceanica) beds up to 30 m
deep, creating small schools near the coastline. During the day it hides in small crevices and cavities of
rocky beds (Bauchot & Hureau 1990). It is an omnivorous species, preying on small invertebrates and
phytoplankton (Klaoudatos et al. 1997, Lenfant &
Olive 1998). In the Mediterranean Sea spawning is
from April to June (Bauchot & Hureau 1986) and,
according to Jardas (1996), in the Adriatic Sea in late
summer. The saddled seabream is of high economic
importance in the Mediterranean Sea, as it is a species of interest not only for fisheries but also for
tourists (i.e. angling). It is one of the species that is
targeted to be set in aquaculture enterprises (Klaoudatos et al. 1997), as are many others members of
the Sparidae family.
To evaluate the effectiveness of biodiversity management programs detailed knowledge of the population ecology of the managed species is necessary.
Recent information on the stock definition of sparids,
in particular saddled seabream, is scarce, although
they are commercially and ecologically important fish

in the Mediterranean Sea. Genetic and abundance

data, in conjunction with morphometric assessments,
may provide important information on saddled seabream stock composition, the evolutionary mechanisms related to specific demographic patterns in
the Mediterranean Sea and the effect of selective
environmental pressures.
It is hypothesized that variation in food availability
and abundance affect the density and dispersal of
saddled seabream in the Mediterranean, and therefore, its population structure should correlate with
oceanographic characteristics that cause such variation. To test such a hypothesis, the genetic structure
of saddled seabream was analyzed and correlated
to seascape characteristics. The Aegean Sea (eastern
Mediterranean basin) is a complex ecosystem with
a highly irregular coastline and semi-isolated deep
basins (Olson et al. 2007), and presents a novel challenge as a transition between the eutrophic Black
Sea and the oligotrophic Mediterranean basin. Six
microsatellite loci were analysed to evaluate genetic
polymorphism in the Aegean Sea using a combination of recent individual-based landscape genetic
approaches.
Additionally, 15 morphological assessments were
used to determine the geographical variation of the
given stocks. Multivariate statistical analysis of morphometric characters has provided useful results in
the past for assessing stock structure of several marine fish species (Schaefer 1989), however, the use of
morphological characteristics does have some limitations they are polygenically inherited, have low
heritability and are subject to considerable environmental plasticity (Karakousis et al. 1991). Multivariate
analysis of a set of phenotypic characters is regarded
as a more appropriate method than the use of a single
character for determining relationships between populations of a given species (Thorpe 1987, Palma &
Andrade 2004).

MATERIALS AND METHODS

Sample collection and study area
Adult saddled seabream individuals were collected
as bycatch from 8 locations in the Aegean Sea. Sampling sites and sizes along with main oceanographic
features are shown in Fig. 1. The total sample size
was 514 individuals; sampling was carried out by
local fishermen using gillnets. A small amount of
muscle tissue was obtained and stored in 70%
ethanol for further analysis.

Gkafas et al.: Oblada melanura population genetics

71

Fig. 1. Oblada melanura. Oceanographic features (current patterns are illustrated with arrows), sample sites and number of
saddled seabream specimens (n) in the Aegean Sea (after Olson et al. 2007)

DNA extraction and PCR amplification

DNA was extracted from 0.01 g of muscle tissue
following the standard phenol/chloroform protocol
(Sambrook & Russel 2001). DNA was preserved in
10 mM TE and stored in 20C. A panel of 6 microsatellites was tested and optimized for the genetic
analysis: SAI10, SAI12, SAI19 (Brown et al. 2005) and
Pma1, Pma2 and GA2A (Adcock et al. 2000). Roques
et al. (2007) characterized an additional panel of species-specific microsatellites which ought to be under
future consideration for screening. The total volume
of the PCR reaction was 10 l; consisting of 40 ng of
template DNA, 10X Buffer, 1.5 mM MgCl2, 0.2 mM
dNTPs, 1 U Taq polymerase (all Invitrogen) and 0.5 M
of each primer (Operon-Invitrogen). The florescence
primers used were labelled by 4 different dyes: FAM,

HEX, ROX and TAMRA (Invitrogen). The PCR conditions were as follows: 95C for 3 min, followed by
30 cycles of 1 min at 95C, 50 s at TC annealing
of the primer set, and 30 s at 72C, with a final step
of 15 min at 72C. PCR products were verified by
1% agarose gel electrophoresis (Invitrogen). Fluorescently labelled PCR products were run on an ABI
3700 automated sequencer (Applied Biosystems).
Each specimens alleles were scored by STRand
software v.2.0 (Toonen & Hughes 2001) and 10% of
genotypes were rerun for error checking.

Genetic analysis
All loci were tested for the presence of null alleles
or allelic dropout using the software MICRO-

72

Aquat Biol 18: 6980, 2013

CHECKER v.2.2.3 (van Oosterhout et al. 2004). Exact

tests for Hardy-Weinberg equilibrium (HWE), linkage disequilibrium (LD, using Fishers exact tests),
expected heterozygosity (HEXP) and observed heterozygosity (HOBS) were carried out using the software
Genepop v.3.4 (Raymond & Rousset 1995). FST
and FIS indices (using the formulation described by
Weir & Cockerham 1984) and number of alleles per
locus were calculated using FSTAT v.2.9.3.2 software
(Goudet 2001). This software was also used to calculate the allelic richness (rarefaction) per locus and
per population. A Bonferroni correction was carried
out for each pairwise analysis (Rice 1989). A recalculation was made in order to estimate null and visible
allele frequencies for each locus and population; for
such estimations, the ENA method was used as
described in Chapuis & Estoup (2007). The ENA correction method was found to efficiently correct the
positive bias induced by the presence of null alleles
on FST estimation and provide a more accurate estimation of FST. To evaluate hypothesized patterns of
spatial genetic structure, a hierarchical analysis of
molecular variance (AMOVA) (Excoffier & Lischer
2010) was performed using Arlequin 3.5 software in
order to partition variance components attributable
to (1) variance between areas; (2) variance between
populations within areas; and (3) variance among
samples within populations. The significance of the
resulting F-indices and variance components were
permutated 10 000 times using the Bonferroni correction (Rice 1989).
Population structure was assessed using the software STRUCTURE v.2.3 (Pritchard et al. 2000a),
where identified migrants and admixed individuals
were assigned. The Correlated Allele Frequency
Model (Falush et al. 2003b) records the allele frequencies in a hypothetical ancestral population without specifying geographic area as a prior. To test the
convergence of the priors and the appropriateness of
the chosen burn-in length and simulation length, 3
independent repeats were run for each value of K,
the number of clusters (2 K 10). Burn-in length
and length of simulation were set at 500 000 and
1 000 000 repetitions, respectively. Isolation by distance software (IBD) was used to assess whether the
association between genetic similarity (FST /[1FST];
Rousset 1997) and geographic distance is statistically
significant using a Mantel test based on 1000 randomizations (Bohonak 2002). The software uses partial correlation coefficients between genetic and geographical distances. We tested for evidence of recent
bottleneck events using the software BOTTLENECK
v.1.2.02 (Piry et al. 1999) and the Stepwise Mutation

Model (SMM). A one-tailed Wilcoxon singed rank

test was used to determine numbers of loci in heterozygosity excess. The mode shift indicator tested
the allele frequency distribution that discriminates
bottlenecked populations from stable populations
(Luikart et al. 1998).
The software Barrier v.2.2 (Manni et al. 2004) was
utilized in order to identify locations and the directions of barriers using a computational geometry
approach. The Monmonier (1973) maximum difference algorithm provided a more realistic representation of the barriers in a genetic landscape and a
significance test was implemented by means of bootstrap matrix analysis. In order to obtain a geometric
satisfactory map from a list of geographic x,y coordinates, a Vorono tessellation (Vorono 1908) calculator was used. Out of this tessellation a Delaunay
triangulation (Brassel & Reif 1979) was obtained.
Additionally, Migrate v.3.3.2 (Beerli & Palczewski
2010) was implemented in order to reveal any directionality of the migration patterns.

Morphometric analysis
Fifteen morphometric characters were measured
on each specimen according to Hubbs & Lagler
(1967). These morphometric characters were: standard length, maximum body depth, minimum body
depth, caudal peduncle length, head length, diameter of eye, preorbital distance, postorbital distance,
predorsal fin distance, dorsal fin height, dorsal fin
base length, anal fin height, anal fin base length,
pectoral fin length and wet weight (Fig. 2). All measurements were taken to the nearest 0.1 mm or mg.
Parameters such as sampling timing, sexual dimorphism, allometric growth, sampling more than one
population in each water body, different phenotypic
groups within the sample, and maturation stage of
the fish could impose some limitations on a study
(Roughgarden 1972, 1974). The effect of allometry
and sexual dimorphism was minimized after the
transformation of the original measurements. In
order to overcome the fourth parameter (sampling
more than one population in each water body) sampling was restricted to as small a geographic area as
possible. As for maturation stage, every effort was
made to choose only mature fish for the analysis. To
minimize any variation resulting from allometric
growth, all morphometric measurements were standardized according to:
e = logY b (logX logX1)

Gkafas et al.: Oblada melanura population genetics

Fig. 2. Oblada melanura. Morphometric measurements: 12:

Preorbital distance; 23: diameter of eye; 34: postorbital
distance; 14: head length; 45: pectoral fin length; 16:
predorsal fin distance; 68: dorsal fin height; 69: dorsal fin
base length; 67: maximum body depth; 1011: anal fin base
length; 1012: anal fin height; 1314: minimum body depth;
1116: caudal peduncle length; 115: standard length

where e is a standardized measurement, Y is character length, b is the slope of log Y against log X plot of
each population, X is the standardized length of the
specimen and X1 is the mean standard length of the
specimen examined. According to Reist (1985), this
transformation best reflects shape variation among
groups independently of size factor. Therefore, total
length and standard lengths of each specimen were
excluded from the final analysis.
Principal component analysis (PCA) was used to
test for the contribution of the remaining 13 morphological characters in the configuration of variance.
Forward stepwise discriminant analysis (DA) based
on the generalized Mahalanobis distance was used to
determine the similarity between populations and
the ability of these characters to identify the specimens correctly. The above analysis was performed
using SPSS v.15.0 statistical software.

RESULTS
A moderately high level of polymorphism was
detected in all tested loci, ranging from 18 alleles at
the Pma1 locus to 93 alleles at the SAI19 locus. However, all loci showed evidence of the presence of
null alleles. These 6 loci also showed departures
from HWE (p < 0.01). Observed heterozygosity values were quite low across all loci for all populations
(mean HOBS = 0.597) and were comparatively lower
than overall expected heterozygosity values (mean
HEXP = 0.898). Values of HOBS, HEXP, FIS and allelic
richness, as well as the numbers of alleles are shown
in Table 1.
FST values between the 8 sampled geographical
areas of the saddled seabream ranged from 0.007

73

between Volos and Karystos to 0.1162 between Paros

and Karystos (Table 2). Most of the values were
highly significant after Bonferroni correction (Rice
1989). Pairwise FST values and their significances did
not differ from the original pattern, even when recalculated using the ENA method (see Table 2). A hierarchical analysis (AMOVA) of the genetic structure
showed that the proportion of total genetic variation
that can be ascribed to differentiation between the
southern, central and northern Aegean areas (FCT) is
close to zero (Table 3). Furthermore, genetic structuring appears to take place among populations within
each area (FSC = 0.042) and within populations (FST =
0.322). Approximately 82% of the total genetic variation was due to variation within populations and 11%
was due to variation among populations within the
3 areas. A Bayesian individual assignment implemented in STRUCTURE is shown in Fig. 3, without
using geographical area as a prior. The highest
posterior probability was for K = 4 and ln (PD) =
7866.6 (Fig. 4). The Wilcoxon sign-ranked test
under the SMM model, implemented in BOTTLENECK, revealed no deviance from the mutation-drift
equilibrium overall (data not shown). Furthermore,
the implemented IBD model of geographical and
genetic distances was insignificant.
The geometric map, using Barrier software, illustrated 4 barriers (Fig. 5). The first barrier separates
the Crete and Paros populations from the Karystos
population, the second separates the Karystos population from the Volos and Trikeri populations, the
third separates the Volos and Trikeri populations
from the Katerini and Nikiti populations, and the
fourth determines a barrier to gene flow between
the Katerini, Nikiti and Kavala populations. Analysis
using Migrate software did not reveal any migration
directionality. The connection type matrix pinpointed
that migration rates were free to vary, without any
clear direction of the migrants (data not shown).
Moreover, univariate ANOVA carried out on the
transformed data (Zar 1984) indicated significant differences between the 8 populations for all morphometric characters except maximum body depth and
anal fin base length (F212, 0.05 = 3.23, p << 0.05). The
PCA extracted 3 factors with eigenvalues > 0.75,
explaining 70.6% of the variance (Table 4).
The characters of primary importance in distinguishing between the groups were head length for
the first canonical variable and dorsal fin height
for the second. Using these morphometric characters,
each specimen could be classified (using DA) correctly to the 8 populations with an accuracy of 66.5%
(Table 5, Fig. 6).

Aquat Biol 18: 6980, 2013

74

Table 1. Oblada melanura. Genetic variation at each locus for each population of saddled seabream in the southern, central
and northern Aegean Sea. HOBS: observed heterozygosity; HEXP: expected heterozygosity; FIS: inbreeding index. **p < 0.01
Locus

Parameter

Volos

SAI10

HOBS
HEXP
Allelic richness
FIS
No. of alleles
HOBS
HEXP
Allelic richness
FIS
No. of alleles
HOBS
HEXP
Allelic richness
FIS
No. of alleles
HOBS
HEXP
Allelic richness
FIS
No. of alleles
HOBS
HEXP
Allelic richness
FIS
No of alleles
HOBS
HEXP
Allelic richness
FIS
No of alleles
HOBS
HEXP
Allelic richness
FIS

0.697
0.963
7.053
0.277**
45
0.598
0.957
6.912
0.376**
41
0.708
0.973
7.303
0.273**
58
0.467
0.788
4.397
0.408**
13
0.614
0.935
6.48
0.344**
35
0.516
0.917
6.089
0.438**
22
0.600
0.922
6.372
0.353**

SAI12

SAI19

Pma1

Pma2

GA2A

Total

Trikeri
0.5
0.938
6.430
0.474**
13
0.6
0.954
6.795
0.377**
16
0.6
0.933
6.345
0.363**
13
0.556
0.883
5.415
0.377**
9
0.9
0.964
7.039
0.068**
18
1
0.921
6.069
0.089
11
0.693
0.932
6.349
0.262**

Kavala
0.5
0.936
6.436
0.471**
19
0.688
0.901
6.024
0.243**
20
0.5
0.929
6.369
0.47**
23
0.438
0.704
3.542
0.386**
5
0.563
0.944
6.579
0.412**
20
0.5
0.911
5.927
0.459**
9
0.532
0.888
5.813
0.407**

Katerini

Karystos

0.611
0.96
6.965
0.37**
10
0.5
0.954
6.849
0.482**
13
0.842
0.959
6.982
0.125**
15
0.222
0.654
3.511
0.667**
4
0.684
0.963
7.029
0.295**
13
0.579
0.861
5.182
0.333**
9
0.573
0.892
6.086
0.379**

0.5
0.868
5.600
0.498**
15
0.6
0.953
6.772
0.383**
17
0.7
0.974
7.263
0.292**
21
0.4
0.7
3.273
0.442**
8
0.8
0.942
6.61
0.158**
20
0.6
0.921
5.986
0.361**
13
0.600
0.893
5.917
0.356**

Nikiti

Paros

Crete

0.529
0.927
6.270
0.436**
8
0.611
0.951
6.749
0.364**
8
0.765
0.963
7.042
0.211**
8
0.412
0.797
4.471
0.491**
3
0.833
0.96
6.985
0.135**
6
0.556
0.886
5.608
0.38**
5
0.618
0.914
6.188
0.336**

1
1
8
0
11
0.8
0.933
6.6
0.158**
14
0.8
0.956
6.756
0.179**
13
0.2
0.511
2.778
0.636**
5
0.4
0.844
5.178
0.556**
12
0.2
0.867
4.733
0.789**
5
0.567
0.852
5.674
0.386**

0.778
0.928
6.279
0.17
15
0.889
0.967
7.077
0.086
15
0.667
0.954
6.923
0.314**
16
0.125
0.792
3.664
0.851**
6
0.778
0.948
8
0.188**
15
0.333
0.791
3.505
0.593**
12
0.595
0.897
5.990
0.367**

Table 2. Oblada melanura. FST values between the 8 sampled geographical areas of saddled seabream (ns: non-significant,
**p < 0.01, ***p < 0.001, after Bonferroni correction). Values in parentheses are the recalculated FST values after application of
the ENA method
Volos

Trikeri

Kavala

Katerini

Karystos

Nikiti

Paros

Volos

Trikeri

0.02527***
(0.021***)

Kavala

0.01270ns
(0.012ns)

0.04246***
(0.032***)

Katerini

0.07151***
(0.06***)

0.04333***
(0.0321***)

0.09333***
(0.088***)

Karystos

0.00704ns
(0.003ns)

0.04205***
(0.031***)

0.00875ns
(0.005ns)

0.09069***
(0.089***)

Nikiti

0.06357***
(0.051***)

0.03426***
(0.023***)

0.08249***
(0.074***)

0.01131ns
(0.008ns)

0.08088***
(0.73***)

Paros

0.09476***
(0.082***)

0.04993ns
(0.031ns)

0.11406***
(0.095***)

0.01539ns
(0.009ns)

0.11622***
(0.109***)

0.03062ns
(0.016ns)

Crete

0.07369***
(0.065***)

0.05079**
(0.045**)

0.08885***
(0.081***)

0.02221ns
(0.015ns)

0.08365***
(0.0712***)

0.01958ns
(0.011ns)

0.03425ns
(0.02ns)

Crete

Gkafas et al.: Oblada melanura population genetics

75

Fig. 3. Oblada melanura. Bayesian individual assignment implemented in STRUCTURE for K = 4 clusters without using
geographical area as a prior. The y-axis represents the probability of assignment of an individual to each cluster and each color
corresponds to the suggested cluster
Table 3. Oblada melanura. AMOVA of grouped into 3 areas: northern, central and
southern Aegean Sea. ns: non-significant, ***p < 0.001
df

Between areas
2
Among populations 5
within areas
Within populations 217

Variance
component

SS

19.054 0.00623
36.388
0.11322
735.800

0.82539

The genetic individual-based approaches show that most of the

saddled seabream in the central
and south Aegean Sea form a
FCT = 0.00233ns
relatively continuous population,
FSC = 0.04227***
despite a complex coastline in
the area. However, the significant
FST = 0.32174***
and relatively high FST values
between central Aegean populations (Volos, Trikeri, Karystos) and
south Aegean populations (Paros, Crete) revealed a
subdivision of the species. The striking result of
the relatively low but significant level, in some cases,
of genetic differentiation between the northern
Aegean group of populations and the group of central
and south Aegean populations illustrates a potential
small-scale population structure in the targeted area
for the species in question. The Bayesian individual
assignment proposes a differentiation between the
northern grouping of specimens and the rest of the
study area. However, the geometric map analysis
revealed barriers to gene flow between the south and
central Aegean and between the central and north
Aegean Sea.

Variance
Fixation index
explained (%)
11.2
25.5
63.3

Fig. 4. Oblada melanura. Determination of the number of

clusters (K) including all 3 repetitions for each K without
geographical area as a prior. The highest peak denotes the
most likely number of clusters according to the Pritchard
Bayes formula. PD: probability of data

DISCUSSION

Aquat Biol 18: 6980, 2013

76

Fig. 5. Oblada melanura. Vorono tessellation (in grey) of the

points (populations) according to geographical locations
(black circles) and the corresponding Delaunay triangulation (in dark grey). The bold black lines with arrows indicate the 4 barriers. Dotted lines represent biogeographic subareas of the Aegean Sea (after Voultsiadou 2005b)
Table 4. Oblada melanura. Results of principal component analysis (PCA) and factor loadings for each morphometric variable
on the 3 extracted PCA factors after varimax normalized rotation
Factor
1
2
3
Character
Maximum body depth
Minimum body depth
Caudal peduncle length
Head length
Diameter of eye
Preorbital distance
Postorbital distance
Predorsal fin distance
Dorsa fin height
Dorsal fin base
Ana fin height
Ana fin base length
Pectoral fin base length

Eigenvalue

% of
Cumulative
variance
%

7.237
1.194
0.746

55.671
9.187
5.738

55.671
64.857
70.595

Factor 1

Factor 2

Factor 3

0.868
0.767
0.724
0.923
0.725
0.615
0.789
0.906
0.466
0.763
0.468
0.845
0.662

0.145
0.138
0.024
0.146
0.192
0.096
0.074
0.197
0.704
0.089
0.604
0.129
0.396

0.019
0.260
0.318
0.003
0.197
0.450
0.231
0.038
0.119
0.040
0.292
0.098
0.412

Moderately high levels of polymorphism were displayed at all 6 microsatellite loci, which is not an
unexpected result given the proposed higher rates of
microsatellite mutation previously reported in marine fishes (OReilly et al. 2004). The reported significant departures from HWE support the presence of
null alleles found in this study, which likely reveal
issues regarding biased patterns of the disequilibrium. Moreover, the observed deviation from HWE
and the relative high inbreeding values may be
related to small sample sizes and/or individuals from
the same progeny. The observed heterozygosity
deficit related to the high FIS values suggests that
saddled seabream do not mate randomly in the study
area and that gene flow may be spatially restricted.
However, the relatively limited gene flow and the
high FST values among samples, along with the differentiation from HWE, might indicate a subdivision
of the populations, suggesting sub-structuring of the
species in the Aegean Sea.
This differentiated pattern in the Aegean Sea was
previously reported in benthopelagic species; these
species in the Aegean Sea seem to exhibit spatial differences with respect to their bentho-geographical
distribution (Vafidis 2002, Kefalas et al. 2003, Voultsiadou 2005a). The authors argue that this dissimilarity
of the north, central and south Aegean Sea conforms
with the prolonged differences in species genetic
and morphological data in relation with geographical
and physicochemical features of the area. The results
of the present study further support such a hypothesis. The phenotypic univariate analysis indicated
significant differences between the 8 populations for
all morphometric characters. DA of morphometric
characters further revealed a clear pattern of population structure segregating the saddled seabream
individuals between the 3 major sub-areas of the
Aegean Sea. The high significance of the morphometric analysis compared with the genetic analysis suggests that oceanographic factors (i.e. currents, cyclonic circulation, bathymetry) are a significant stressor
for environmental plasticity.
According to the literature, barriers in the Aegean
Sea coincide with strong oceanographic fronts of different nature as a result of a near-surface thermohaline circulation involving the low salinity outflow
from the Black Sea in conjunction with the freshwater
input of the continental shelf and the high salinity
input from the south Aegean Sea (Poulos et al. 1997,
Kourafalou et al. 2004). In addition, a basin-wide
cyclonic circulation is taking place, resulting in subbasin scale gyres connected to the complex topography of the area (Nittis & Perivoliotis 2002). Taking

Gkafas et al.: Oblada melanura population genetics

Biogeographically speaking, this documented cyclonic circulation is not

only a food specialization habitat,
and therefore boundary for saddled
7
8
seabream dispersal potential, but also
a transition zone with respect to eco0
0
10
0
logical niches for many marine species
12.5
12.5
(e.g. Voultsiadou 2005b). The saddled
20
0
seabream is considered to have mod14.3
0
0
0
erate pelagic larval duration with
42.4
2.2
pelagic eggs and offshore larvae,
0
33.4
suggesting a shorter pelagic duration
than other marine species and an
ultimately restricted gene flow due to
tidal currents (Macpherson & Raventos
2006, Galarza et al. 2009).
Despite the reported extensive phenotypic variability within and among populations, the fact is that
specimens were collected on a differential seasonal
basis during sampling. Therefore, it is probable that
a large proportion of the intra-sample variation (i.e.
various age groups) is the result of such heterogeneous phenotypic groups. To what degree these differences are the result of environmental conditions
or genetic divergence remains to be determined,
although the present data on microsatellite variability do confirm such differentiation. Within the limits
of accuracy of the biometric experimental procedure,
the characters of primary importance in distinguishing between the studied populations were those
related to head and fin height dimensions; such characters explained approximately 70% of the total variance between the groups. Nonetheless, the DA classification accuracy for saddled seabream populations
was somewhat lower and/or similar to values reported in the literature for other marine species (e.g.
Alexander & Breden 2004, Cabral et al. 2003, Mamuris et al. 1998, Karakousis et al. 1993, Claytor & MacCrimmon 1988). However, one ought to bear in mind
that such morphometric characteristics have a lower
ability to discriminate populations than meristic
(Karakousis et al. 1991) or even evolutionarily influenced characteristics. In any case, pairwise FST significances reflect the relative described differentiation: AMOVA was highly significant on 2 of the 3
hierarchical levels. Regarding the phenotypic analysis, results from both PCA and DA were highly
significant.
Morphological variability among different geographical populations could be attributed to the variable genetic structure of populations and/or to different environmental conditions prevailing in each
geographic area. Multivariate analyses can provide

Table 5. Oblada melanura. Discriminant analysis output, showing the percentage of specimen classification in each sampled group (1 to 8). Overall
classification = 66.5%
Population
Volos
Karystos
Kavala
Katerini
Crete
Trikeri
Nikiti
Paros

75.1
20
0
10
0
10
11.1
0

4.2
70
0
5
14.3
0
5.5
0

0
0
31.3
15
0
0
33.3
0

5.7
0
31.2
35
0
10
5.5
0

0
0
0
0
71.4
0
0
66.6

15
0
12.5
15
0
80
0
0

Wilks lambda
0.82288

df

5.60709

0.00186

them in turn, the conditions in the north Aegean contain extended shelf areas, 2 cyclonic and 1 anticyclonic
system and a deep trench of 1500 m, which present
a physical geographical boundary to the central
Aegean. The central Aegean forms extended deep areas, a cyclonic system and a boundary current along
the continental shelf of an elongated island with
Karystos at the southern end. The south Aegean contains many island complexes, a main cyclonic system
and different water masses, and communicates with
the western Mediterranean basin through Crete (for a
review, see Olson et al. 2007). The different cyclone
systems in the Aegean Sea contribute to different
upwelling events and thus differentiated nutrientenriched water masses, suggesting differences in food
resources and availability for biota (Theodorou 2004).
2

Canonical variable II

3
Crete

Volos

Paros
Karystos

Kavala

6
Katerini
7

Trikeri

Nikiti

8
9
24

23

22

21

20

19

18

17

16

15

Canonical variable I
Fig. 6. Oblada melanura. Discriminant analysis plot of populations, where the 13 morphometric characters were used.
Circles represent the percentage of the total of specimens
correctly classified per population. Arrows indicate corresponding population

77

78

Aquat Biol 18: 6980, 2013

an efficient tool, along with genetic data interpretation, for stock definition, and thus would be useful for
a conservation management project plan for Hellenic
coastal resources. However, one cannot exclude the
existence of possible undetected genetic structure
of such populations, which could account for any
detected morphological variation. Such findings are
not unique, as some phenotypic adaptations are not
dependent on genetic mutations. The novelties of
some phenotypic adaptations are not immediately
expressed in the species gene pool. The lack of differences in a study does not preclude the existence of
stock differentiation, even when including the input
of genetic forces, since this is a direct consequence of
environmental adaptation (Schweigert 1991).
Historical processes producing isolation, or low
connectivity in combination with potential bottlenecks in one or more subpopulations (although not
detected in the present data set), could also lead
to morphological and genetic differentiation. Moreover, the observed absence of HWE is maybe due to
genetic drift (e.g. small sample size), and inbreeding
is such a crucial process and is remarkably quite common for species favoring reproduction in the sea. As
inbreeding is caused by mating of genetically related
individuals (common in fishes), it results in increased
homozygosity. Therefore, observed heterozygosities
are likely to be lower than expected. Although this is
not always the case, in the present study inbreeding
within populations was high enough. Furthermore, intense fishing pressure, mainly overfishing, can also
lead to potential bottlenecks and population divergence (see Perez-Ruzafa et al. 2006, Walsh et al. 2006,
Baibai et al. 2012, Limborg et al. 2012), and thus apparent small populations often suffer from the loss of
genetic diversity due to genetic drift and inbreeding
effects. However, the present study assumed a structure between the 3 areas, which affects local effective
population size and thus the level of inbreeding.
Therefore, the considerable population structure revealed by the morphometric characteristics has the
potential to influence the impact on fitness-related
traits to a greater extent than may be expected from
the genetic analysis. The Aegean Sea, being a semienclosed marginal basin of relatively small volume
compared with the open sea, shows an amplified and
very rapid response to climate change (Anagnostou
et al. 2005). A large number of faunal studies in the
area have succeeded in reconstructing the climatic
conditions and documenting the general trend from
the cold, glacial climate conditions of the late Pleistocene to the warm, interglacial conditions of the
Holocene (see Cramp et al. 1988, Geraga et al. 2000).

Cyclonic circulation, current flows and life history

seem to be the factors that best represent the nature
of barriers to gene flow encountered across the
Aegean Sea in saddled seabream populations. The
revealed population structure and the species geographic dispersal might be closely linked to the
oceanographic features encountered in the studied
area, and could act as a useful tool for the study of
other physical processes, such as the abundance of
plankton and nutrients, which regulate the ecological niche of bio-society in small scaled areas. Further
analysis is required to address the demographic patterns of the saddled seabream, as well as the ecological aspects of plasticity in this species.
LITERATURE CITED

Adcock GJ, Bernal-Ramirez JH, Hauser L, Smith P, Car-

valho GR (2000) Screening of DNA polymorphism in

samples of archived scales from New Zealand snapper.
J Fish Biol 56:12831287
Alexander HJ, Breden F (2004) Sexual isolation and extreme
morphological divergence in the Cuman guppy: a
possible case of incipient speciation. J Evol Biol 17:
12381254
Anagnostou C, Chronis G, Gogou A, Karageorgis AP and
others (2005) Paleoceanography and paleocirculation of
the Aegean Sea from the late glacial period to the present (20 000 years). In: Papathanasiou E, Zenetos A (eds)
SoHelME, State of the Hellenic Marine Environment.
HCMR Publ., Athens, p 2123
Arculeo M, Lo Brutto S, Sirna-Terranova M, Maggioa T,
Cannizzaro L, Parrinello N (2003) The stock genetic
structure of two Sparidae species, Diplodus vulgaris and
Lithognathus mormyrus, in the Mediterranean Sea. Fish
Res 63:339347
Baibai T, Oukhattar L, Vasquez Q, Javier V, Mesfioui A,
Rey-Mendez M, Soukri A (2012) First global approach:
morphological and biological variability in a genetically
homogeneous population of the European pilchard, Sardina pilchardus (Walbaum, 1792) in the North Atlantic
coast. Rev Fish Biol Fish 22:6380
Bauchot ML, Hureau CJ (1986) Sparidae. In: Whitehead PJP,
Bauchot ML, Hureau JC, Nielsen J, Tortonese E (eds)
Fishes of the North-eastern Atlantic and the Mediterranean. Vol II. UNESCO, Bungay, UK, pp. 883907
Bauchot ML, Hureau JC (1990) Sparidae. In: Quero JC,
Hureau JC, Karrer C, Post A, Saldanha L (eds) Check-list
of the Fishes of the Eastern Tropical Atlantic, Clofeta II.
UNESCO, Paris, p 790812
Bauchot ML, Hureau JC, Miquel JC (1981) Sparidae. In:
Fischer WG, Bianchi G, Scott WB (eds) FAO species
identification sheets for fishery purposes: Eastern
Central Atlantic; fishing areas 34, 47 (in part). Canada
funds-in-Trust, Ottawa
Beerli P, Palczewski M (2010) Unified framework to evaluate
panmixia and migration direction among multiple sampling locations. Genetics 185:313326
Bekkevold D, Andre C, Dahlgren TG, Clausen LA and
others (2005) Environmental correlates of population differentiation in Atlantic herring. Evolution 59:26562668
Bohonak AJ (1999) Dispersal, gene flow, and population
structure. Q Rev Biol 74:2145

Gkafas et al.: Oblada melanura population genetics

Bohonak AJ (2002) IBD (isolation by distance): a program for

analyses of isolation by distance. J Hered 93:153154

Brassel KE, Reif D (1979) A procedure to generate Thiessen
polygons. Geogr Anal 325:3136
Brown RC, Tsalavouta M, Terzoglou V, Magoulas A, McAndrew BJ (2005) Additional microsatellites for Sparus
aurata and cross-species amplification within the Sparidae family. Mol Ecol Notes 5:605607
Cabral HN, Marques JF, Rego AL, Catarino AI, Figueiredo
J, Garcia J (2003) Genetic and morphological variation of
Synaptura lusitanica (Capello, 1868) along the Portuguese coast. J Sea Res 50:167175
Chapuis MP, Estoup A (2007) Microsatellite null alleles and
estimation of population differentiation. Mol Biol Evol
24:621631
Claytor RR, MacCrimmon HR (1988) Morphometric and
meristic variability among North American Atlantic
salmon (Salmo salar). Can J Zool 66:310317
Coll M, Piroddi C, Steenbeek J, Kaschner K and others
(2010) The biodiversity of the Mediterranean Sea: estimates, patterns, and threats. PLoS ONE 5:e11842
Cramp A, Collins M, West R (1988) Late Pleistocene
Holocene sedimentation in the NW Aegean Sea: a palaeoclimatic palaeoceanographic reconstruction. Palaeogeogr Palaeoclimatol Palaeoecol 68:6177
Excoffier L, Lischer HEL (2010) Arlequin suite ver 3.5: A
new series of programs to perform population genetics
analyses under Linux and Windows. Mol Ecol Res 10:
564567
Falush D, Stephens M, Pritchard JK (2003b) Inference of
population structure using multilocus genotype data:
linked loci and correlated allele frequencies. Genetics
164:15671587
Galarza JA, Carreras-Carbonell J, Macpherson E, Pascual
M, Roques S, Turnere GF, Ricod C (2009) The influence
of oceanographic fronts and early-life-history traits on
connectivity among littoral fish species. Proc Natl Acad
Sci USA 106:14731478
Geraga M, Tsaila-Monopolis S, Ioakeim C, Papatheodorou
G, Ferentinos G (2000) Evaluation of palaeoenvironmental changes during the last 18 000 years in the Myrtoon basin, SW Aegean Sea. Palaeogeogr Palaeoclimatol
Palaeoecol 156:117
Gonzlez-Wangemert M, Froufe E, Prez-Ruzafa A, Alexandrino P (2011) Phylogeographical history of the white
seabream Diplodus sargus (Sparidae): implications for
insularity. Mar Biol Res 7:250260
Goudet J (2001) FSTAT, a program to estimate and test gene
diversities and fixation indices (version 2.9.3); http://www.
unil.ch/izea/softwares/Fstat.html
Hoelzel AR (2009) Evolution of population structure in marine mammals. In: Bertorelle G, Bruford MW, Hauffe HC,
Rizzoli A, Vernesi C (eds) Population genetics for animal conservation. Cambridge University Press, New York,
NY, p 294318
Hoelzel AR, Hey J, Dahlheim ME, Nicholson C, Burkanov V,
Black N (2007) Evolution of population structure in a
highly social top predator, the killer whale. Mol Biol Evol
24:14071415
Hubbs C, Lagler K (1967) Fishes of the Great Lakes Region.
University of Michigan Press, Ann Arbor, MI
Jardas I (1996) Jadransha ihtiofanna. Skolsha Knjiga, Zagreb
Jrgensen HBH, Hansen MM, Bekkevold D, Ruzzante DE,
Loeschcke V (2005) Marine landscapes and population
genetic structure of herring (Clupea harengus L.) in the
Baltic Sea. Mol Ecol 14:32193234
Karakousis Y, Triantaphyllidis C, Economidis P (1991) Mor-

79

phological variability among seven Greek populations of

brown trout (Salmo trutta). J Fish Biol 38:807817
Karakousis Y, Peios C, Economidis P, Triantaphyllidis C
(1993) Multivariate analysis of the morphological variability among Barbus peloponnesius (Cyprinidae) populations from Greece and two populations of B. meridionalis meridionalis and B. meridionalis petenyi. Cybium
17:229240
Kefalas E, Tsirtsis G, Castritsi-Catharios I (2003) Distribution
and ecology of Demospongiae from the circalittoral of
the islands of the Aegean Sea (Eastern Mediterranean).
Hydrobiologia 499:125134
Kelly RP, Palumbi SR (2010) Genetic structure among 50
species of the Northeastern Pacific rocky intertidal community. PLoS ONE 5:e8594
Klaoudatos S, Pouli A, Katsoris T, Conides A (1997) Study of
hermaphroditism of new euryhaline finfish species candidate for aquaculture. Technical report 1469 PENED 94,
NVMR/University of Athens
Kourafalou VH, Savvidis YG, Koutitas CG, Krestenitis YN
(2004) Modelling studies on the processes that influence
matter transfer on the Gulf of Thermaikos (NW Aegean
Sea). Cont Shelf Res 24:203222
Lenfant P, Olive C (1998) Changements gradueles du
rgime alimentarie des juvniles dOblade (Oblada melanura, Sparidae) lors du recrutement. Cybium 22:203210
Limborg MT, Hanel R, Debes PV, Ring AK, Andr C,
Tsigenopoulos CS, Bekkevold D (2012) Imprints from
genetic drift and mutation imply relative divergence
times across marine transition zones in a pan-European
small pelagic fish (Sprattus sprattus). Heredity 109:
96107
Luikart G, Allendorf FW, Cornuet JM, Sherwin WB (1998)
Distortion of allele frequency distributions provides a test
for recent population bottlenecks. J Hered 89:238247
Macpherson E, Raventos N (2006) Relationship between
pelagic larval duration and geographic distribution of
Mediterranean littoral fishes. Mar Ecol Prog Ser 327:
257265
Magoulas A, Castilho R, Caetano S, Marcato S, Patarnello T
(2006) Mitochondrial DNA reveals a mosaic pattern of
phylogeographical structure in Atlantic and Mediterranean populations of anchovy (Engraulis encrasicolus). Mol
Phylogenet Evol 39:734746
Mamuris Z, Apostolidis AP, Panagiotaki P, Theodorou AJ,
Triantaphyllidis C (1998) Morphological variation between red mullet populations in Greece. J Fish Biol 52:
107117
Manni F, Gurard E, Heyer E (2004) Geographic patterns of
(genetic, morphologic, linguistic) variation: how barriers
can be detected by Monmoniers algorithm. Hum Biol
76:173190
McCusker MR, Bentzen P (2010) Historical influences dominate the population genetic structure of a sedentary marine fish, Atlantic wolffish (Anarhichas lupus), across the
North Atlantic Ocean. Mol Ecol 19:42284241
Monmonier M (1973) Maximum-difference barriers: an alternative numerical regionalization method. Geogr Anal 3:
245261
Nittis K, Perivoliotis L (2002) Circulation and characteristics
of the North Aegean Sea: a contribution from real-time
buoy measurements. Medit Mar Sci 3:2131
Olson DB, Kourafalou VH, Johns WE, Samuels G, Veneziani
M (2007) Aegean surface circulation from a satellitetracked drifter array. J Phys Oceanogr 37:18981917
OReilly PT, Canino MF, Bailey KM, Bentzen P (2004)
Inverse relationship between FST and microsatellite poly-

80

Aquat Biol 18: 6980, 2013

morphism in the marine fish, walleye pollock (Theragra

chalcogramma): implications for resolving weak population structure. Mol Ecol 13:17991814
Palma J, Andrade JP (2004) Morphological study of Pagrus
pagrus, Pagellus bogaraveo, and Dentex dentex (Sparidae) in the Eastern Mediterranean Sea. J Mar Biol Assoc
UK 84:449454
Palumbi S (1994) Genetic divergence, reproductive isolation,
and marine speciation. Annu Rev Ecol Syst 25:547572
Perez-Ruzafa A, Gonzalez-Wanguemert M, Lenfant P, Marcos C, Garcia-Charton JA (2006) Effects of fishing protection on the genetic structure of fish populations. Biol
Conserv 129:244255
Pilot M, Dahlheim ME, Hoelzel AR (2010) Social cohesion
among kin, gene flow without dispersal and the evolution of population genetic structure in the killer whale
(Orcinus orca). J Evol Biol 23:2031
Pinera JA, Blanco G, Vazquez E, Sanchez JA (2007) Genetic
diversity of blackspot seabream (Pagellus bogaraveo)
populations of Spanish Coasts: a preliminary study. Mar
Biol 151:21532158
Piry S, Luikart G, Cornuet JM (1999) Bottleneck: a computer
program for detecting recent reductions in the effective
size using allele frequency data. J Hered 90:502503
Poulos SE, Drakopoulos PG, Collins MB (1997) Seasonal
variability in sea surface oceanographic conditions in the
Aegean Sea (eastern Mediterranean): an overview.
J Mar Syst 13:225244
Pritchard JK, Stephens M, Donnelly P (2000a) Inference
of population structure using multilocus genotype data.
Genetics 155:945959
Raymond M, Rousset F (1995) GENEPOP (version 1.2) Population genetics software for exact tests and ecumenicism. J Hered 86:248249
Reist J (1985) An empirical evaluation of several Univariate
methods that adjust for size variation in morphometric
data. Can J Zool 63:14291439
Rice WR (1989) Analyzing tables of statistical tests. Evolution 43:223225
Rolland JL, Bonhomme F, Lagardre F, Hassan M, Guinand
B (2007) Population structure of the common sole (Solea
solea) in the Northeastern Atlantic and the Mediterranean Sea: revisiting the divide with EPIC markers. Mar
Biol 151:327341
Roques S, Galarza JA, McPherson E, Turner GF, Rico C
(2007) Isolation and characterization of nine polymorphic
microsatellite markers in the two-banded sea bream
(Diplodus vulgaris) and cross-species amplification in the
white sea bream (Diplodus sargus) and the saddled bream
(Oblada melanura). Molec Ecol Notes 7(4):661663
Roughgarden J (1972) Evolution of niche width. Am Nat
106:683718
Roughgarden J (1974) Niche width: biogeographical patterns
among Anolis lizard populations. Am Nat 108:429442
Rousset F (1997) Genetic differentiation and estimation of
gene flow from F-statistics under isolation by distance.
Genetics 145:12191228
Salmenkova EA (2011) New view on the population genetic
Editorial responsibility: Francesco Patti,
Naples, Italy

structure of marine fish. Genetika 47:14401450

Sambrook J, Russel DW (2001) Molecular cloning: a laboratory manual, 3rd edn. Cold Spring Harbor Laboratory
Press, Cold Spring Harbor, NY
Schaefer KM (1989) Morphometric analysis of yellowfin
tuna Thunnus albacares from the Pacific Ocean. InterAm Trop Tuna Comm Bull 19:389427
Schweigert JF (1991) Multivariate description of Pacific herring (Clupea harengus pallasi) stocks from size and age
information. Can J Fish Aquat Sci 48:23652376
Siokou-Frangou I, Bianchi M, Christaki U, Christou ED and
others (2002) Carbon flow in the planktonic food web
along a gradient of oligotrophy in the Aegean Sea (Mediterranean sea). J Mar Syst 33-34:335353
Theodorou A (2004) Oceanography: an introduction to marine environment. Stamoulis, Athens
Thorpe RS (1987) Geographic variation: a synthesis of cause,
data pattern and congruence in relation to subspecies,
multivariate analysis and phylogenies. Boll Zool 54:311
Toonen RJ, Hughes S (2001) Increased throughput for fragment analysis on ABI Prism 377 Automated Sequencer
using a membrane comb and STRand Software. Biotechniques 31:13201324
Vafidis D (2002) Biogeography of the Anthozoa coelenterate
fauna of the Mediterranean Sea. 9th Int Congr Zoogeography and Ecology of Greece and adjacent regions
(ICZEGAR), Thessaloniki, 2225 May, Abstracts, p 159
van Oosterhout C, Hutchinson WF, Wills DPM, Shipley P
(2004) Microchecker: software for identifying and correcting genotyping errors in microsatellite data. Mol Ecol
Notes 4:535538
Vias J, Bremer JA, Pla C (2004) Phylogeography of the
Atlantic bonito (Sarda sarda) in the northern Mediterranean: the combined effects of historical vicariance, population expansion, secondary invasion, and isolation by
distance. Mol Phylogenet Evol 33:3242
Vorono MG (1908) Nouvelles application des paramtres
continus la thorie des formes quadratiques, deuxime
mmoire, recherche sur le parallloedres primitifs. J Reine
Angew Math 134:198207
Voultsiadou E (2005a) Sponge diversity in the Aegean Sea:
checklist and new information. Ital J Zool 72:5364
Voultsiadou E (2005b) Demosponge distribution in the
eastern Mediterranean: a NW-SE gradient. Helgol Mar
Res 59:237251
Walsh MR, Munch SB, Chiba S, Conover DO (2006) Maladaptive changes in multiple traits caused by fishing:
impediments to population recovery. Ecol Lett 9:142148
Weir BS, Cockerham CC (1984) Estimating F-statistics
for the analysis of population structure. Evolution 38:
13581370
Zar JH (1984) Biostatistical analysis. Prentice Hall, Upper
Saddle River, NJ
Zulliger DE, Tanner S, Ruch M, Ribi G (2009) Genetic structure of the high dispersal Atlanto-Mediterreanean sea
star Astropecten aranciacus revealed by mitochondrial
DNA sequences and microsatellite loci. Mar Biol 156:
597610
Submitted: July 24, 2012; Accepted: December 14, 2012
Proofs received from author(s): February 21, 2013