MARGEN-00322; No of Pages 10 Marine Genomics xxx (2015) xxx–xxx

Contents lists available at ScienceDirect

Marine Genomics journal homepage: www.elsevier.com/locate/margen

F

5 6 7

a

8

a r t i c l e

9 10 11 12 13

Article history: Received 16 February 2015 Received in revised form 28 April 2015 Accepted 28 April 2015 Available online xxxx

14 15 16 17 18

Keywords: Adaptation Atlantic Ocean Fish Population genomics

Departamento de Genética, Universidad de Santiago de Compostela, Facultad de Biología, Santiago de Compostela E-15706, Spain University of Leuven, Laboratory of Biodiversity and Evolutionary Genomics, Charles Deberiotstraat 32, B-3000 Leuven, Belgium Departamento de Genética, Universidad de Santiago de Compostela, Facultad de Veterinaria, Lugo E-27002, Spain

i n f o

a b s t r a c t

Partitioning phenotypic variance in genotypic and environmental variance may benefit from the population genomic assignment of genes putatively involved in adaptation. We analyzed a total of 256 markers (120 microsatellites and 136 Single Nucleotide Polymorphisms — SNPs), several of them associated to Quantitative Trait Loci (QTL) for growth and resistance to pathologies, with the aim to identify potential adaptive variation in turbot Scophthalmus maximus L. The study area in the Northeastern Atlantic Ocean, from Iberian Peninsula to the Baltic Sea, involves a gradual change in temperature and an abrupt change in salinity conditions. We detected 27 candidate loci putatively under selection. At least four of the five SNPs identified as outliers are located within genes coding for ribosomal proteins or directly related with the production of cellular proteins. One of the detected outliers, previously identified as part of a QTL for growth, is a microsatellite linked to a gene coding for a growth factor receptor. A similar set of outliers was detected when natural populations were compared with a sample subjected to strong artificial selection for growth along four generations. The observed association between FST outliers and growth-related QTL supports the hypothesis of changes in growth as an adaptation to differences in temperature and salinity conditions. However, further work is needed to confirm this hypothesis. © 2015 Published by Elsevier B.V.

E

E

C

c

T

b

O

4

Román Vilas a,⁎, Sara G. Vandamme b,1, Manuel Vera c,2, Carmen Bouza c, Gregory E. Maes b,3, Filip A.M. Volckaert b, Paulino Martínez c

R O

3Q4

P

2

A genome scan for candidate genes involved in the adaptation of turbot (Scophthalmus maximus)

D

1Q3

R

37 35 34 36

1. Introduction

39 40

Partitioning the phenotypic variance into genetic and environmental components can be challenging due to statistical correlations between genetic and environmental variation and the constitutive interaction between both effects. However, phenotypic differences may reflect differences in the underlying genetic variation caused by natural selection acting over generations in a process of adaptation to particular environmental conditions. This effect has been demonstrated through genetic variation of candidate genes within

45 46

N C O

43 44

U

41 42

R

38

⁎ Corresponding author at: Departamento de Genética, Universidad de Santiago de Compostela, Facultad de Biología, Santiago de Compostela E-15706, Spain. Tel/fax: +34 982 822428. E-mail addresses: [email protected] (R. Vilas), [email protected] (S.G. Vandamme), [email protected] (M. Vera), [email protected] (C. Bouza), [email protected] (G.E. Maes), [email protected] (F.A.M. Volckaert), [email protected] (P. Martínez). 1 Present address: Institute for Agricultural and Fisheries Research (ILVO), Animal Sciences Unit—Fisheries, Ankerstraat 1, B 8400 Ostend, Belgium. 2 Present address: Laboratori d'Ictiologia Genètica, Departamento de Biología, Facultad de Ciencias, Universidad de Girona, Campus de Montilivi s/n, E-17071 Girona, Spain. 3 Present address: Centre for Sustainable Tropical Fisheries and Aquaculture, School of Marine and Tropical Biology, James Cook University, Townsville QLD 4811, Australia.

populations along an environmental gradient (Colosimo et al., 2005; Orsini et al., 2012; Alberto et al., 2013; Hemmer-Hansen et al., 2014). The choice of the genes putatively subjected to natural selection can be made on the basis of functional knowledge (Hoffman and Willi, 2008) or by identifying Quantitative Trait Loci (QTL) involved in the development of phenotypes directly related to survival and reproduction (Slate, 2005; Storz, 2005). However, these methods require a detailed characterization of gene function or performing experimental crosses, which is not always feasible. The population genomics approach can also be used for detecting candidate genes. This involves the screening of a large set of markers scattered across the genome in order to distinguish the effects of natural selection influencing specific loci from the effects of processes that act on the whole genome, such as gene flow, genetic drift and inbreeding (Luikart et al., 2003). In contrast to the latter, loci targeted by selection or closely linked to a locus under selection (known as selective sweeps) should display three characteristics which can be statistically tested: (1) a skewed allele frequency distribution, (2) low variation within populations and (3) a higher differentiation between locally adapted populations (e.g. Lewontin and Krakauer, 1973; Schlötterer, 2002). Therefore, scanning the patterns of variation at the genomic level in order to identify loci

http://dx.doi.org/10.1016/j.margen.2015.04.011 1874-7787/© 2015 Published by Elsevier B.V.

Please cite this article as: Vilas, R., et al., A genome scan for candidate genes involved in the adaptation of turbot (Scophthalmus maximus), Mar. Genomics (2015), http://dx.doi.org/10.1016/j.margen.2015.04.011

19 20 21 22 23 24 25 26 27 28 29 30 31 32 33

47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68

92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 Q5 123 124 125 126 127 128 129 130 131 132 133 134

Fin clip samples were obtained from 286 specimens of turbot (S. maximus) captured at five locations across the Northeastern Atlantic Ocean (Fig. 1): the English Channel (EC, 50.85 N; 1.1 E); the Baltic Sea in the vicinity of the Island of Bornholm, Denmark (BS, 55.07 N; 14.55 E); the German Bight (NS, 55.60 N; 8.0 E); the Cantabric Sea (CS, 43.41 N; 7.30 E); and the Atlantic Galician coast of the Iberian Peninsula (AG, 42.14 N; 8.43 W). An additional farmed population (FAR) native to the Northeastern Atlantic Ocean was analyzed. This population is the result of four generations of strong artificial selection for growth carried out in aquaculture facilities in Galicia, Northwestern Spain. A total of 48 individuals per site (except 46 at CS) were analyzed. No specific permission for sampling was required for this study since all wild individuals sampled were obtained from commercial fishing and those from the farm were broods previously stored from an ongoing genetic breeding program. No protected species was sampled.

137 138

2.2. Genetic markers

152

Whole genomic DNA was extracted by using standard phenol– chloroform procedures. Samples were genotyped for a total of 256 codominant markers: 120 microsatellites and 136 single nucleotide polymorphisms (SNPs). All microsatellite loci had been characterized (Pardo et al., 2006, 2007; Bouza et al., 2008) and mapped (Bouza et al., 2007, 2012) previously. Because their linkage relationships could be used as a selection criterion, the 24 linkage groups that constitute the turbot genetic map were included. Furthermore, linkage equilibrium between markers could be assumed as loci assigned to the same linkage group were generally separated by large genetic distances. A second criterion for selection of microsatellites was their significant association with previously reported growth-and-resistance related QTL ( Sánchez-Molano et al., 2011; Rodríguez-Ramilo et al., 2011, 2013, 2014). Among the 120 microsatellites analyzed, 92 were linked to expressed sequence tags (ESTs) (Bouza et al., 2008, 2012) and 28 were anonymous (Pardo et al., 2007). Thirty of these 92 EST-linked microsatellite markers and the 28 anonymous ones had been previously studied in a population genomics screen using some of the samples of the present study (Vilas et al., 2010). Therefore, we extend that study by analyzing new samples, and the variation at 62 additional ESTlinked microsatellites and 136 SNPs. Fourteen of the EST-linked microsatellites described in Bouza et al. (2008) were also used by Vandamme et al. (2014). The anonymous microsatellites Sma149, 2/5TG14 and Sma22 and the EST-linked (E code) SmaE7, SmaE14 and SmaE40, showed significant association with growth-related traits (Sánchez-Molano et al., 2011). One of them, Sma149, was also associated with resistance and survival time to the infection with the parasite Philasterides dicentrarchi (Rodríguez-Ramilo et al., 2013). In addition, we used another seven microsatellites associated with QTL for resistance to pathologies: Sma168, Sma147, Sma144, Sma77, SmaE41, SmaE30 and SmaE36 (Rodríguez-Ramilo et al., 2011, 2013, 2014). Microsatellites were genotyped on an ABI PRISM® 3730 automatic sequencer (Applied Biosystems) and allele scoring was performed with GENEMAPPER 4.0 software (Applied Biosystems). All the 29 SNPs from Vera et al. (2011) included in the turbot map as framework markers (Bouza et al., 2012) were incorporated to the study. In addition, another 107 recently developed SNPs from the last turbot EST database enriched with two 454 runs (Vera et al., 2013; Ribas et al., 2013; Pereiro et al., 2012) were analyzed. The mapping position of these 107 SNPs was unknown. However, we thought it was appropriate to check either for the proximity of the outlier SNPs to other outliers based on their estimated position or for their location within the confidence interval of previously

153 154

O

F

136

R O

90 91

2.1. Sampling

P

88 89

135

D

86 87

2. Materials and methods

T

84 85

C

82 83

E

80 81

R

78 79

R

76 77

O

75

C

73 74

N

71 72

that behave differently is useful for detecting adaptive polymorphisms, particularly if combined with complementary strategies such as the use of functional information and appropriate QTL (Vasemägi and Primmer, 2005; Stinchcombe and Hoekstra, 2007; Hansen et al., 2012; Schoville et al., 2012). However, it is important to emphasize the limited status of candidates for outlier loci, which begs for additional experimental confirmation because of important limitations of the population genomics approach (Bierne et al., 2013). Correlations across populations and variation in effective population size can substantially increase the FST variance among loci and therefore make it difficult to separate historical or demographic effects from natural selection in highly subdivided species (Nei and Maruyama, 1975; Robertson, 1975). Furthermore, the effects of selection and genetic drift are frequently intertwined. For example, when selection is involved in the reinforcement of reproductive barriers, this may facilitate genome-wide neutral divergence via genetic drift (Nosil et al., 2009). In combination with population subdivision, the hitchhiking effect and the presence of intrinsic genetic incompatibilities in hybrid zones, may cause a correlation between genetic and environmental variation (Bierne, 2011; Bierne et al., 2011). Furthermore, other factors besides a complex demography and cryptic genetic structure, can also determine the appearance of selective effects on particular loci (Barrett and Hoekstra, 2011; Roesti et al., 2012). The turbot (Scophthalmus maximus L.; Scopththalmidae) is a marine fish with high fecundity and vagility, which probably explains high levels of gene flow between large panmictic populations. This hypothesis is concordant with the evidence of low genetic structure along its geographic range (Blanquer et al., 1992; Bouza et al., 1997, 2002; Coughlan et al., 1998; Vandamme et al., 2014). The observed spatial genetic homogeneity of neutral markers is an advantage in relating FST-outlier loci with signatures of divergent selection because it suggests a low genetic drift scenario (Beaumont, 2005; Nielsen et al., 2009). Vandamme et al. (2014) found evidence for putative divergent selection at both sides of a hydrodynamic front in the North Sea. However, the very wide distribution range across several environmental gradients implies the capacity of the fish to survive and reproduce in very different conditions. The Atlantic area involves a gradient in light, temperature and salinity. Temperature differences range between 16 °C in the Iberian Peninsula to 8 °C in the southern part of the Baltic Sea and differences in salinity range from 35‰ in the Iberian Peninsula to 8‰ in the Baltic (HELCOM, 2003; Álvarez et al., 2005). Whereas temperature changes gradually with latitude, an abrupt salinity drop is observed between the Baltic and the North Sea. The transition zone between the Baltic and the North Sea is recognized as a biogeographical barrier for many marine species (Johanneson and André, 2006). Differences in temperature and photoperiod are expected to particularly affect the survival and reproduction of a fish such as turbot that lives in relatively shallow waters near the coast. Although populations living across these gradients have usually shown low genetic structure for presumably neutral markers, differentiation between them is statistically significant (Nielsen et al., 2004; Suzuki et al., 2004; Florin and Höglund, 2007; Vilas et al., 2010; Vandamme et al., 2014). An explanation for this result is that turbot consist of several demes locally adapted to environmental differences in a background of relatively high levels of gene flow (Nielsen et al., 2004; Vilas et al., 2010). With the aim to identify genes with an adaptive genetic pattern we perform a genome scan analysis of five natural populations living in different environmental conditions by using a total of 256 codominant markers, most of them closely linked to functionally annotated genes (Bouza et al., 2012). The analysis included 13 microsatellites located at QTL related to growth and resistance to pathologies and the genetic variation in natural populations was also compared with an additional sample subjected to strong growth selection during several generations.

U

69 70

R. Vilas et al. / Marine Genomics xxx (2015) xxx–xxx

E

2

Please cite this article as: Vilas, R., et al., A genome scan for candidate genes involved in the adaptation of turbot (Scophthalmus maximus), Mar. Genomics (2015), http://dx.doi.org/10.1016/j.margen.2015.04.011

139 140 141 142 143 144 145 146 147 148 149 150 151

155 156 157 158 159 160 161 162 163 164 165 Q6 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196

3

D

P

R O

O

F

R. Vilas et al. / Marine Genomics xxx (2015) xxx–xxx

197 198

T

E

Fig. 1. Map of Europe showing the location of collecting sites: Atlantic Galician coast (AG), Cantabric Sea (CS), English Channel (EC), North Sea (NS) and Baltic Sea (BS).

209

2.3. Genetic variation within and among populations

210 211

216 217

Allele frequencies, average number of alleles per locus (A) and gene diversity (He ) were calculated using FSTAT v2.9.3 (Goudet, 2001). This program was also used to estimate FST values among populations (Weir and Cockerham, 1984). Conformance to Hardy– Weinberg proportions was tested using exact tests as implemented in GENEPOP v3.4 (Raymond and Rousset, 1995). Critical significance levels were adjusted for multiple tests using the Bonferroni correction.

218

2.4. Outlier tests for selection

219

To search for putative signatures of selection we applied a test based on estimates of FST (Beaumont and Balding, 2004). The basic idea is that the influence of divergent selection on a locus will increase F ST compared to that expected at neutral loci. Thus, evidence for divergent selection is obtained by looking for outliers with higher FST values than expected under neutrality. Conversely, unexpectedly low values would be indicative of balancing selection. The method is based on the work by Lewontin and Krakauer (1973),

212 213 214 215

220 221 222 223 224 225 226

E

R

205 206

R

203 204

N C O

201 202

U

199 200

C

207 208

reported QTL. For this purpose, we used the recently assembled turbot genome (A. Figueras, Spanish National Research Council, Vigo, unpublished) anchored to the linkage map of turbot using mapped markers. Scaffolds anchored to the turbot map with a significant homology of marker sequences (mostly full identity but always N 98%) and including more than two markers (frequently tens of markers) were considered for positioning the SNPs. In this way, non-mapped SNPs were located between the two closest adjacent loci of turbot linkage groups anchored to specific scaffolds (Hermida et al., 2013). SNPs were genotyped by using the MassARRAY technology from SEQUENOM, Inc. San Diego, CA, USA, in the University of Santiago de Compostela Genomics Platform.

improved by using stochastic simulations to obtain the expected neutral distribution of the statistic (Beaumont and Nichols, 1996) and inserted in a Bayesian framework (Beaumont and Balding, 2004). A logistic-regression model is used to estimate population and locus specific F ST values in combination with a hierarchical Bayesian approach. The posterior probability of including a locus specific effect (α), presumably caused by selection, is estimated by calculating a Bayes factor (BF), which is the ratio of the posterior probabilities of selective and neutral models given the data (Foll and Gaggiotti, 2008). A log10BF = 1.5–2 is considered “very strong evidence” of different statistical support for both models and corresponds to a posterior probability (P) between 0.97 and 0.99. For values above 2 evidence is interpreted as “decisive” (P = 0.99– 1). The Beaumont and Balding (2004) method implemented in BayeScan v2.01 (Foll and Gaggiotti, 2008) has some advantages over the approach implemented in Fdist (Beaumont and Nichols, 1996). For example, it does not assume a symmetrical model of gene flow and therefore allows particularly distinct populations. It also avoids the problems associated with the simulation of the neutral distribution of FST by using the empirical estimate, a value calculated with a sample of loci which also includes the outlier ones. We analyzed markers in all populations to reveal loci with a major overall effect. Because we included one population (BS) living in a different natural environment and one sample (FAR) originating from an artificially reared stock, we performed a triple approach: (1) analysis of the five natural populations including BS; (2) analysis of FAR and all natural populations except BS; and (3) analysis of all natural populations except BS. In addition, we conducted pairwise comparisons, as recommended by Tsakas and Krimbas (1976) and Vitalis et al. (2001), in order to detect local selective effects driven by the different conditions represented by the BS and FAR samples. Because microsatellites and SNPs differ in the mutation process, which affects their levels of variability within and among populations, data corresponding to both types of markers were analyzed separately.

Please cite this article as: Vilas, R., et al., A genome scan for candidate genes involved in the adaptation of turbot (Scophthalmus maximus), Mar. Genomics (2015), http://dx.doi.org/10.1016/j.margen.2015.04.011

227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 Q7 256 257 258 259 260

4 t1:1 t1:2 t1:3 t1:4

R. Vilas et al. / Marine Genomics xxx (2015) xxx–xxx

Table 1 Gene diversity (He), Hardy–Weinberg equilibrium departures (FIS) and test for conformance to expected values by locus and population for 58 microsatellite loci of turbot, analyzed by Vilas et al. (2010), of which 28 are anonymous (left) and 30 EST-linked (right) in two new populations: the English Channel (EC) and the farmed population (FAR). Significance level *: P b 0.0008.

t1:5

EC

FAR

EC

FAR

He

FIS

He

FIS

Locus

He

FIS

He

FIS

Sma284 Sma42 Sma168 Sma247 2/5TG14 Sma135 Sma22 Sma147 Sma14 Sma137 Sma142 Sma149 F8I11/8/17 Sma113 Sma117 Sma34 Sma18 Sma184 Sma185 Sma38 Sma100 Sma144 Sma175 Sma205 Sma19 Sma146 Sma282 Sma77

0.862 0.938 0.785 0.759 0.867 0.910 0.867 0.804 0.903 0.320 0.936 0.889 0.865 0.808 0.859 0.868 0.882 0.666 0.723 0.833 0.594 0.962 0.418 0.510 0.920 0.829 0.691 0.831

0.084 0.007 0.197* −0.033 0.061 0.053 0.067 0.053* 0.081 0.037 0.304* 0.043 0.153 0.193 0.095 0.053 0.019 0.217* 0.047* 0.119 0.214* 0.026 0.021 0.183 0.005 0.079 −0.076 0.064

0.867 0.906 0.631 0.646 0.825 0.821 0.878 0.691 0.765 0.377 0.873 0.801 0.497 0.699 0.731 0.802 0.868 0.257 0.525 0.807 0.408 0.876 0.341 0.459 0.895 0.770 0.551 0.759

0.183 −0.058 −0.057 −0.054 0.041 0.015 0.075 −0.085 −0.007 −0.051 0.269 0.037 −0.112 −0.043 −0.083 −0.142 0.069 0.028 0.213 0.123 0.081 0.001 0.024 0.053 0.022 −0.101 0.112 0.018

SmaE1 SmaE2 SmaE3 SmaE7 SmaE4 SmaE29 SmaE8 SmaE12 SmaE14 SmaE16 SmaE25 SmaE28 SmaE38 SmaE20 SmaE35 SmaE40 SmaE41 SmaE42 SmaE19 SmaE22 SmaE26 SmaE32 SmaE30 SmaE43 SmaE10 SmaE13 SmaE31 SmaE33 SmaE36 SmaE39

0.869 0.284 0.586 0.617 0.510 0.367 0.610 0.879 0.862 0.379 0.789 0.430 0.887 0.666 0.347 0.854 0.693 0.432 0.703 0.799 0.645 0.706 0.527 0.116 0.383 0.858 0.846 0.814 0.766 0.856

0.250 0.617* −0.066 0.139* −0.184 0.074 −0.080 0.052 0.226 0.598* 0.076 −0.135 0.026 0.086 0.624* −0.029 0.066 0.169 0.005 0.040 0.360* −0.085 0.115 −0.053 −0.104 −0.027 0.133* 0.189 0.051 0.111

0.770 0.172 0.505 0.516 0.469 0.308 0.662 0.797 0.870 0.043 0.656 0.245 0.727 0.427 0.000 0.793 0.660 0.538 0.508 0.262 0.461 0.574 0.576 0.354 0.164 0.781 0.745 0.680 0.583 0.647

0.080 −0.093 0.010 0.111 0.066 −0.084 0.119 −0.046 −0.101 −0.011 −0.027 −0.154 −0.024 0.287 0.000 −0.051 0.177 0.071 −0.024 −0.032 0.123 −0.160 −0.121 0.057 −0.063 −0.048 0.225 −0.073 −0.014 0.130

261

3. Results

262

3.1. Genetic variation within and among populations

263 264

In general, both microsatellites and SNP loci showed genotypic frequencies in accordance with the expected values under Hardy– Weinberg equilibrium (Tables 1 and 2). Four loci showed positive and significant FIS values across several populations suggesting the presence of null alleles (SmaE139, SmaE120, SmaE61 and SmaE96). Hardy–Weinberg deviations were particularly frequent in the sample from the English Channel (EC; Tables 1 and 2), which is consistent with the population genetic structure observed in that geographic area (Vandamme et al., 2014). Microsatellite analysis revealed similar levels of genetic variability in natural populations (He ranged from 0.577 in BS to 0.614 in EC; A ranged from 6.6 in the Baltic Sea (BS) to 7.4 in Atlantic Galician coast (AG). The farmed population (FAR) showed the lowest variability (He = 0.555 and A = 5.2), which is consistent with the presumed loss of diversity associated with the founder effect and the successive generations of selection. This effect was not detected with SNPs, which is consistent with the loss of rare alleles due to genetic drift. Natural population structure quantified as FST revealed that 3.7% of the variation at microsatellites and 2.9% of the variation at SNPs were due to differences between populations. When analyzed pairwise, the highest differentiation between natural populations was detected between BS and AG (FST = 0.031 and 0.043 for microsatellites and SNPs, respectively; P = 0.003) and between BS and EC (FST = 0.034 for microsatellites and SNPs; P = 0.003), while the lowest one between AG and Cantabric Sea (CS) for microsatellites (FST = 0.004; P = 0.003) and between German Bight (NS) and CS for SNPs (FST = 0.000; P = 0.173). The farmed population showed, as expected due to drift, the highest divergence (the pairwise FST ranged from 0.047 to 0.075 for microsatellites and from 0.037 to 0.085 for SNPs; P b 0.003). FST for SNPs was statistically significant only when natural populations were compared with BS or FAR samples (Table 3).

273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291

E

O

R O P

T C E

R

R

O

271 272

C

269 270

N

267 268

U

265 266

F

Locus

t1:7 t1:8 t1:9 t1:10 t1:11 t1:12 t1:13 t1:14 t1:15 t1:16 t1:17 t1:18 t1:19 t1:20 t1:21 t1:22 t1:23 t1:24 t1:25 t1:26 t1:27 t1:28 t1:29 t1:30 t1:31 t1:32 t1:33 t1:34 t1:35 t1:36

D

t1:6

3.2. Tests for outliers in natural populations

292

The global genome scan analysis using microsatellites in all natural populations identified 15 outliers with a degree of evidence interpreted as “decisive” (log10BF N 2; P N 0.99; Table 4). Three other markers were identified as outliers with “very strong evidence” (log10BF = 1.5–2; P = 0.97–0.99; Table 4). This means that genomic regions closely linked to 15% of the microsatellite loci analyzed are likely under selection. For microsatellites Sma42, Sma22, Sma144 and SmaE112 populations were more similar than expected under neutrality, thus suggesting stabilizing selection. The remaining 14 outliers revealed a pattern of variation consistent with divergent selection. The outlier status of two markers (SmaE7 and Sma146) confirmed previous results by Vilas et al. (2010). The other two previously reported (SmaE4 and SmaE12) showed some statistical support (log10BF = 1; P = 0.91–0.97). The markers SmaE4 and SmaE7 were also identified as outliers by Vandamme et al. (2014). The only marker subjected to balancing selection (Sma144), which was consistently identified by Vilas et al. (2010), was confirmed in the present study (Table 4). When the BS population was excluded from the analysis, an additional locus showed a “decisive” significance (Sma149); 12 loci did not change their significance status while eight changed, particularly SmaE167, now being non-significant. Seven out of 13 microsatellites at QTL were outliers in any of the two global comparisons performed representing more than 50% tested. Global analysis of SNPs in natural populations identified five outlier loci (6.8%), three of them (SmaSNP247, SmaSNP181 and SmaSNP314) highly significant (Table 5, Fig. 2). These three markers are located within genes coding for ribosomal proteins: L18a, L13 and L1, respectively (Vera et al., 2011, 2013). The posterior probability of the other two SNPs (SmaSNP298 and SmaSNP281) was between 0.91 and 0.97, which corresponds with a log10BF of 1 (“strong evidence”). SmaSNP298 and SmaSNP281 are located at the coding region of the queuine tRNAribosyltransferase locus and the 5′UTR of the gene coding for a

293 294

Please cite this article as: Vilas, R., et al., A genome scan for candidate genes involved in the adaptation of turbot (Scophthalmus maximus), Mar. Genomics (2015), http://dx.doi.org/10.1016/j.margen.2015.04.011

295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323

R. Vilas et al. / Marine Genomics xxx (2015) xxx–xxx

5

t2:1 t2:2 t2:3

Table 2 Gene diversity (He), Hardy–Weinberg equilibrium departures (FIS) and test for conformance to expected values by locus and population of 62 new EST-linked microsatellites of turbot. EC, BS, NS, CS, AG, and FAR are samples from the English Channel, Baltic Sea, North Sea, Cantabric Sea, Atlantic Galician coast, and the farmed population, respectively. Significance level *: P b

t2:4

0.0008.

t2:5

EC

BS

NS

CS

AG

FAR

He

FIS

He

FIS

He

FIS

He

FIS

He

FIS

He

FIS

t2:7 t2:8 t2:9 t2:10 t2:11 t2:12 t2:13 t2:14 t2:15 t2:16 t2:17 t2:18 t2:19 t2:20 t2:21 t2:22 t2:23 t2:24 t2:25 t2:26 t2:27 t2:28 t2:29 t2:30 t2:31 t2:32 t2:33 t2:34 t2:35 t2:36 t2:37 t2:38 t2:39 t2:40 t2:41 t2:42 t2:43 t2:44 t2:45 t2:46 t2:47 t2:48 t2:49 t2:50 t2:51 t2:52 t2:53 t2:54 t2:55 t2:56 t2:57 t2:58 t2:59 t2:60 t2:61 t2:62 t2:63 t2:64 t2:65 t2:66 t2:67 t2:68

Smax144 Smax272 SmaE139 Smax187 Smax218 Smax261 SmaE84 SmaE86 Smax194 Smax315 SmaE120 SmaE96 Smax154 Smax159 SmaE127 SmaE61 SmaE99 Smax244 SmaE100 Smax205 Smax283 Smax290 Smax170 Smax220 Smax254 Smax310 Smax158 Smax255 SmaE118 SmaE128 SmaE52 Smax164 Smax189 Smax215 Smax225 SmaE97 SmaE316 Smax231 SmaE117 SmaE82 Smax270 Smax302 SmaE137 Smax183 Smax184 Smax277 SmaE79 Smax180 Smax248 Smax284 SmaE78 Smax191 Smax197 Smax227 SmaE105 SmaE112 SmaE50 SmaE71 Smax276 SmaE72 SmaE167 Smax168

0.823 0.149 0.776 0.069 0.450 0.439 0.781 0.902 0.693 0.368 0.531 0.855 0.942 0.640 0.716 0.707 0.499 0.710 0.549 0.691 0.752 0.725 0.739 0.190 0.779 0.747 0.553 0.506 0.578 0.908 0.045 0.477 0.586 0.333 0.591 0.605 0.736 0.429 0.148 0.638 0.454 0.554 0.790 0.142 0.839 0.734 0.881 0.595 0.853 0.748 0.343 0.506 0.210 0.342 0.632 0.819 0.542 0.695 0.568 0.823 0.290 0.504

0.132 −0.040 0.520* −0.016 0.260* −0.111 −0.072 0.081 0.050 0.038 0.559* 0.644* 0.0590 0.201* −0.130 0.186 0.043 0.405* 0.240 0.035 0.058* −0.096 −0.065 −0.094 0.117 0.026 0.162 −0.038 −0.046 −0.028 −0.012 0.048 0.001 −0.085 0.257 0.008 0.046 0.108 0.250 −0.103 −0.032 −0.279 0.240 −0.071 0.153 −0.022 0.069 0.030 0.053 0.318* 0.444 0.227 0.290 −0.058 −0.031 −0.065 −0.088 −0.166 0.043 −0.077 −0.200 0.009

0.758 0.155 0.802 0.021 0.409 0.475 0.595 0.841 0.724 0.118 0.411 0.762 0.914 0.441 0.676 0.757 0.327 0.587 0.501 0.691 0.610 0.687 0.650 0.061 0.701 0.680 0.575 0.505 0.707 0.870 0.022 0.264 0.615 0.410 0.564 0.626 0.754 0.298 0.021 0.593 0.494 0.648 0.784 0.245 0.685 0.788 0.843 0.533 0.781 0.731 0.340 0.697 0.121 0.230 0.586 0.746 0.585 0.664 0.504 0.832 0.000 0.538

0.032 0.192 0.376 0.000 −0.068 −0.096 −0.001 −0.007 0.177 −0.056 0.482* 0.497* 0.015 0.009 −0.140 0.295* 0.053 0.218 −0.040 0.045 −0.116 −0.114 −0.025 −0.021 −0.129 0.049 0.047 0.032 0.182 0.193 0.000 0.195 −0.175 0.116 0.245 −0.054 0.181 0.428 0.000 −0.076 −0.033 −0.410 0.279 0.391 0.365* 0.054 0.074 −0.059 0.081 −0.064 0.248 0.405 −0.057 0.243 0.076 −0.060 0.109 −0.098 0.132 0.028 0.000 0.225

0.847 0.041 0.712 0.062 0.406 0.473 0.631 0.897 0.709 0.345 0.505 0.705 0.935 0.448 0.724 0.716 0.328 0.401 0.463 0.626 0.621 0.724 0.663 0.155 0.677 0.576 0.514 0.491 0.635 0.922 0.118 0.558 0.589 0.287 0.591 0.611 0.769 0.462 0.138 0.595 0.345 0.581 0.781 0.154 0.825 0.691 0.859 0.584 0.829 0.752 0.339 0.623 0.271 0.342 0.606 0.789 0.606 0.665 0.621 0.771 0.349 0.504

−0.008 −0.010 0.648* −0.014 0.126 −0.056 −0.022 0.094 −0.029 −0.025 0.440* 0.471* 0.048 −0.066 0.059 0.284 0.200 0.099 0.099 −0.098 −0.039 −0.035 0.101 0.192 0.133 −0.093 −0.012 −0.103 0.147 0.119 −0.056 0.028 0.115 0.258 −0.154 −0.174 0.011 0.217 −0.057 0.124 −0.025 −0.398 0.136 −0.080 0.267 −0.054 −0.040 −0.212 0.020 0.002 −0.005 0.388* −0.020 −0.181 0.138 −0.019 0.037 −0.064 0.211 0.134 0.512* 0.048

0.813 0.160 0.754 0.085 0.442 0.482 0.751 0.911 0.713 0.427 0.293 0.778 0.913 0.425 0.794 0.758 0.229 0.630 0.493 0.629 0.643 0.678 0.747 0.160 0.597 0.562 0.616 0.505 0.488 0.899 0.065 0.287 0.583 0.520 0.628 0.553 0.787 0.441 0.321 0.614 0.428 0.588 0.683 0.022 0.815 0.730 0.847 0.573 0.838 0.766 0.368 0.677 0.196 0.334 0.639 0.755 0.609 0.662 0.574 0.814 0.374 0.443

0.117 −0.084 0.557* −0.025 0.065 0.323 0.044 0.171 0.055 −0.325 0.667* 0.608* −0.070 −0.023 0.014 0.283* 0.052 0.275 0.179 −0.002 0.019 −0.049 −0.041 −0.084 −0.020 0.033 −0.144 0.100 −0.070 0.085 −0.023 −0.060 −0.029 0.060 0.074 0.115 0.181 0.013 0.053 0.114 −0.118 −0.331 −0.051 0.000 0.146 −0.071 0.152 −0.289 0.118 0.092 −0.064 0.261 −0.111 −0.042 −0.191 0.021 0.124 0.093 0.109 −0.014 0.048 −0.129

0.817 0.100 0.776 0.082 0.260 0.499 0.876 0.876 0.698 0.426 0.332 0.767 0.921 0.475 0.762 0.711 0.281 0.611 0.451 0.675 0.672 0.705 0.731 0.172 0.669 0.596 0.593 0.499 0.671 0.920 0.082 0.361 0.605 0.524 0.603 0.675 0.780 0.439 0.195 0.621 0.334 0.598 0.716 0.137 0.819 0.742 0.846 0.530 0.871 0.745 0.320 0.660 0.154 0.192 0.623 0.792 0.609 0.674 0.554 0.807 0.339 0.461

0.081 −0.044 0.592* −0.021 0.117 −0.127 −0.023 −0.023 −0.194 −0.027 0.579* 0.429* 0.029 −0.051 −0.012 0.296 −0.039 0.522* −0.108 −0.018 −0.023 −0.086 −0.053 −0.093 0.034 0.091 0.175 −0.194 0.287* −0.017 −0.033 −0.180 0.173 0.046 −0.070 0.043 −0.041 0.288 −0.069 0.060 0.252 −0.359 −0.093 −0.068 0.124 −0.032 0.040 −0.099 −0.051 0.049 −0.105 0.084 −0.080 −0.085 0.063 0.033 0.145 −0.050 0.210 −0.098 0.121 0.031

0.725 0.395 0.711 0.000 0.448 0.496 0.737 0.729 0.592 0.313 0.510 0.641 0.893 0.490 0.722 0.769 0.607 0.605 0.192 0.657 0.724 0.726 0.563 0.333 0.700 0.267 0.313 0.498 0.460 0.806 0.082 0.259 0.525 0.157 0.408 0.491 0.658 0.631 0.126 0.591 0.504 0.410 0.677 0.157 0.765 0.702 0.771 0.505 0.788 0.729 0.282 0.556 0.061 0.543 0.500 0.772 0.397 0.558 0.444 0.657 0.193 0.299

0.190 0.084 0.330 0.000 −0.138 −0.139 −0.086 0.169 −0.162 −0.088 0.433 0.701* 0.182 −0.172 0.116 0.190* −0.111 0.402 −0.108 0.029 −0.035 0.053 −0.171 0.000 −0.012 −0.034 0.011 0.102 0.138 0.181 −0.033 0.035 −0.070 −0.060 −0.084 0.151 0.113 0.207 0.296 −0.127 −0.115 −0.220 0.690* −0.082 0.101 −0.098 0.034 0.010 0.022 0.028 0.408 0.250 −0.021 −0.036 0.167 0.136 0.055 −0.082 0.329 −0.059 0.338 −0.066

324 325 326 327 328 329 330 331

E

T

C

E

R

R

N C O

U

phosphatidyl inositol transfer protein, respectively (Vera et al., 2013). The five outliers displayed patterns of differentiation consistent with a model of divergent selection. Pairwise population analysis of SNPs confirmed the apparent association of the outlier pattern with the BS sample (Table 6). Analysis of the relationship between the recently assembled turbot genome with the genetic map (Bouza et al., 2012) located the highly significant outliers SmaSNP247 and SmaSNP181 in QTLs for resistance to Aeromonas salmonicida and P. dicentrarchi, respectively (Table 7).

O

R O

P

F

Locus

D

t2:6

3.3. Comparison between natural and farmed populations

332

The global outlier microsatellite analysis including FAR instead of BS revealed 14 highly consistent outliers (log10BF N 2; P N 0.99). Two of them (Sma42 and Sma22) were significant for balancing selection. The remaining 12 were significant for divergent selection (Table 4). Other four loci (SmaE112, Sma168, SmaE12 and Sma142) were also significant (P = 0.97–0.99) for divergent selection. In total, only five loci (Sma42,

333

Please cite this article as: Vilas, R., et al., A genome scan for candidate genes involved in the adaptation of turbot (Scophthalmus maximus), Mar. Genomics (2015), http://dx.doi.org/10.1016/j.margen.2015.04.011

334 Q8 335 336 337 338

6 t3:1 t3:2 t3:3 t3:4

R. Vilas et al. / Marine Genomics xxx (2015) xxx–xxx

Table 3 Pairwise FST values from SNPs (above diagonal) and microsatellites (below diagonal). EC, BS, NS, CS, AG and FAR are samples for the English Channel, Baltic Sea, North Sea, Cantabric Sea and Atlantic Galician coast, respectively. P b 0.0033.

t3:5 t3:6 Q1 t3:7 t3:8 t3:9 t3:10 t3:11

339

EC BS NS CS AG FAR

Table 5 Outlier analysis based on BayeScan of SNPs in natural populations of turbot. α is the locus specific effect; P is the posterior probability that the locus is under selection and FST is the fixation index.

EC

BS

NS

CS

AG

FAR

Outlier

α

P

FST

t5:5

– 0.034* 0.020* 0.022* 0.025* 0.047*

0.034 – 0.026* 0.031* 0.030* 0.075*

0.001 0.032* – 0.009* 0.015* 0.064*

0.003 0.041* 0.000 – 0.004 0.067*

0.008 0.043* 0.002 0.003 – 0.071*

0.046* 0.085* 0.037* 0.041* 0.043* –

SmaSNP247 SmaSNP281 SmaSNP314 SmaSNP298 SmaSNP181

2.114 1.432 1.485 1.346 1.316

1.000 0.967 0.985 0.914 0.993

0.0765 0.0469 0.0491 0.0441 0.0412

t5:6 t5:7 t5:8 t5:9 t5:10

generalized pattern consistent with balancing selection. Two of these loci, Sma22 and Sma144, are located at QTL for growth and resistance, respectively (Rodríguez-Ramilo et al., 2013, 2014). The observation that more than half of the outlier microsatellites are located in QTL seems to confirm the adaptive significance of the associated traits. A statistical association between outlier loci identified by the population genomics approach and QTL for putatively adaptive traits has previously been observed in fish (Rogers and Bernatchez, 2005, 2007).

365 366

4.1. Microsatellite based outliers along a temperate and salinity gradient

373

Because the outlier analysis was performed with samples from populations living along a temperature cline, it is likely that the presumed adaptive process is related to differences for this variable or any other correlated with it. Many environmental variables are dependent on temperature, including differences in metabolism, the distribution of food, predators, parasites and competitors. Differences in temperature as an explanation for the atypical behavior of several markers is consistent with the result that most outliers remained significant when the analysis did not include the two atypical samples (BS because of the low salinity and FAR because of farming), but only those populations living along a temperature gradient (AG, CS, EC, and NS). Although temperature is gradually changing within the Atlantic Ocean, population genetic structure of turbot appears to be mostly associated with a relatively strong break within the North Sea (Vandamme et al., 2014). Four of these outliers (Sma22, SmaE7, 2/5TG14 and Sma149) are located at a growth-related QTL (Sánchez-Molano et al., 2011; Rodríguez-Ramilo et al., 2014). Locus SmaE7 is linked to a gene coding for the receptor for a fibroblast growth factor, which is likely involved in adaptation

374 375

355

4. Discussion

356

364

This study identified several candidate genes for local adaptation in natural populations of turbot living across a temperature and salinity gradient in the North Atlantic. Candidate loci were associated with 20 different linkage groups (Table 7). Fifteen microsatellites (12.5%), five of them located in QTL related to growth and pathogen resistance, showed very high statistical support as outliers. Another five loci were putative outliers (Table 4). Most outliers showed evidence of divergent selection suggesting that they are involved in local adaptation. However, four markers (Sma42, Sma22, Sma144 and SmaE112) revealed a

t4:1 t4:2 t4:3 Q2 t4:4

Table 4 Outlier analysis based on BayeScan of the microsatellite loci in all five turbot populations including the Baltic Sea and the farmed population (5p/BS and 5p/FAR, respectively) and in four populations excluding the Baltic and the farmed population (4p/noBS-FAR). α is the locus specific effect; P is the posterior probability that the locus is under selection and FST is the fixation index. Outliers identified by Vilas et al. (2010) and Vandamme et al. (2014) are indicated as V10 and V14, respectively. Significance level ***: P N 0.99; **: P = 0.97–0.99; *: P = 0.91–0.97.

357 358 359 360 361 362 363

O

R O

P

D

E

T

352 353

C

350 351

E

348 349

R

346 347

R

344 345

t4:5

O

342 343

F

354

Sma22, Sma144, SmaE112 and SmaE12) displayed patterns of variation consistent with a model of balancing selection (Table 4). However, SmaE12 was identified to be subjected to divergent selection when FAR was replaced by BS. In addition, this locus was identified as outlier only when the analysis involved BS or FAR. Nonetheless, most outliers were common to the three global tests (Table 4). Only one SNP (SmaSNP273) was identified as outlier (P = 0.995) after replacing BS by FAR, showing signals consistent with divergent selection. Furthermore, this marker was only significantly identified when FAR was compared with either of the Iberian samples (P = 0.991 and 0.946 for comparisons with CS and AG, respectively). This marker is located at the 3′UTR of the ubiquitin-protein ligase locus. The global analysis of SNPs excluding both BS and FAR did not reveal any outliers. Pairwise comparisons between natural populations and FAR using both microsatellites and SNPs basically confirmed the results of the global analysis (results not shown).

340 341

5p/BS

Sma42 Sma117 Sma22 Sma144 Sma147 Sma146 SmaE7 SmaE22 F8I11/8/17 SmaE33 2/5TG14 SmaE84 SmaE137 SmaE112 SmaE167 Sma168 SmaE99 SmaE117 SmaE12 SmaE4 Sma149 Sma142

C

t4:7 t4:8 t4:9 t4:10 t4:11 t4:12 t4:13 t4:14 t4:15 t4:16 t4:17 t4:18 t4:19 t4:20 t4:21 t4:22 t4:23 t4:24 t4:25 t4:26 t4:27 t4:28

α

V10

N

Outlier

U

t4:6

t5:1 t5:2 t5:3 t5:4

V10 V10V14

V10 V10V14

−1.198 1.240 −1.541 −1.113 1.552 1.333 2.487 0.974 1.071 1.099 1.257 3.016 1.751 −1.842 1.911 1.021 1.328 1.636 0.751 1.096 0.546 −0.055

5p/FAR

4p/noBS-FAR

P

FST

α

P

FST

α

P

FST

*** *** *** *** *** *** *** *** *** *** *** *** *** *** *** ** ** ** * * ns ns

0.0064 0.0649 0.0048 0.0068 0.0887 0.0700 0.1883 0.0520 0.0553 0.0582 0.0656 0.2737 0.1051 0.0038 0.1313 0.0555 0.0774 0.1061 0.0415 0.0584 0.0335 0.0188

−1.097 1.508 −1.192 −0.528 1.854 1.317 2.694 1.384 1.328 1.502 1.401 2.961 1.855 −1.140 −0.033 1.120 1.773 1.250 −1.228 −0.127 0.971 0.616

*** *** *** * *** *** *** *** *** *** *** *** *** ** ns ** *** ns ** ns *** **

0.0122 0.1184 0.0114 0.0203 0.1539 0.1022 0.2611 0.1091 0.1032 0.1187 0.1091 0.3023 0.1552 0.0123 0.0407 0.0904 0.1504 0.1084 0.0115 0.0415 0.0779 0.0573

−1.170 1.659 −1.354 −1.075 2.060 1.552 2.752 1.380 1.410 1.656 1.605 3.227 2.073 −1.494 −0.047 1.400 1.540 1.459 −0.946 0.245 0.994 0.500

* *** *** *** *** *** *** *** *** *** *** *** *** ** ns *** *** ns ns ns *** ns

0.0044 0.0631 0.0037 0.0046 0.0940 0.0566 0.1680 0.0500 0.0494 0.0644 0.0594 0.2355 0.0965 0.0034 0.0188 0.0531 0.0627 0.0667 0.0057 0.0210 0.0340 0.0201

Please cite this article as: Vilas, R., et al., A genome scan for candidate genes involved in the adaptation of turbot (Scophthalmus maximus), Mar. Genomics (2015), http://dx.doi.org/10.1016/j.margen.2015.04.011

367 368 369 370 371 372

376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391

7

R O

O

F

R. Vilas et al. / Marine Genomics xxx (2015) xxx–xxx

Fig. 2. FST outliers identified by BayeScan analysis of the global dataset with five natural samples (136 SNPs). A log10 Bayes factor (BF) of 1.0, 1.5 and 2.0 corresponds to posterior probabilities that a locus is under selection of 0.91, 0.97 and 0.99, respectively. For a log10BF N 2 the evidence is commonly interpreted as “decisive”.

409 410 411 412 413 414 415 416 417 418 419 420 421

C

E

407 408

R

405 406

R

403 404

N C O

401 402

U

399 400

t6:7

Outlier

BS/EC

BS/NS

BS/CS

BS/AG

t6:8 t6:9 t6:10 t6:11 t6:12

SmaSNP247 SmaSNP281 SmaSNP314 SmaSNP298 SmaSNP181

1.000 0.846 0.830 0.963 0.986

0.998 0.711 0.860 0.779 0.995

1.000 0.579 0.967 0.679 0.949

1.000 0.488 0.972 0.859 0.963

423

Table 7 Markers linked to loci likely under selection in turbot. LG is the linkage group in the genetic map of Bouza et al. (2012). Outliers within the confidence interval of QTL for growth (G) and resistance (R) are indicated.

t7:1 t7:2 t7:3 t7:4

P

Table 6 Outlier analysis of SNPs between pairs of natural populations of turbot. EC, BS, NS, CS, AG, and FAR are samples from the English Channel, Baltic Sea, North Sea, Cantabric Sea, Atlantic Galician coast, and the farmed population, respectively. Posterior probability of being under selection of five SNPs identified as outliers in the global analysis is shown. The remaining six pairwise comparisons (which do not involve BS) revealed no outliers.

397 398

also explain differences between populations in the availability of resources used for growth. Both the identification of several growthrelated loci in natural populations living along a gradient of temperature and the comparison with a sample subjected to strong artificial selection for growth, support the hypothesis of changes in body size as a by-product of the adaptation to differences in temperature or any other variable correlated with it. Four microsatellites linked to genes (SmaE167, SmaE117, SmaE12 and SmaE4) were identified as outliers only when BS was included in the analysis, which suggests that this sample affects most the outlier pattern. Two of them (SmaE12 and SmaE4) were previously identified as outliers; they are most likely related to the salinity cline at the entrance of the Baltic Sea (Vilas et al., 2010; Vandamme et al., 2014). In contrast to SmaE4, very high statistical support is found for

D

t6:1 t6:2 t6:3 t6:4 t6:5 t6:6

395 396

T

422

(Vilas et al., 2010; Vandamme et al., 2014). Most markers identified as outliers remained highly significant when the Baltic sample (BS) was replaced for a farmed population which had been strongly selected for growth during four generations (FAR). One marker (Sma149), located at a QTL for both growth and resistance to the infection with P. dicentrarchi (Rodríguez-Ramilo et al., 2013), was identified as an outlier only when the analysis included the FAR sample (Table 4). This could be related to the opportunistic infection of farmed populations originating from the study area by this parasite (Iglesias et al., 2001). However, Sma149 also seems to be an outlier after excluding FAR and BS samples. Taken together these observations suggest (1) that the anomalous behavior of most markers does not solely depend on the peculiar characteristics of the Baltic Sea (e.g. low salinity) and (2) an adaptive relationship between temperature and growth rate. The latter hypothesis is supported by the detection of several outliers along a cline of temperature which are located in QTLs for growth. However, when the growth selected farm sample is excluded from the outlier analysis, such markers are still identified as outliers. This result suggests a strong effect on wild populations. The presumed relationship between temperature and growth rate can be direct or mediated by other traits. For example, the optimal temperature for the growth of turbot declines with increasing size (Imsland et al., 1996), an observation consistent with the hypothesis of differences in body size as an adaptation to cold. Another possibility is that adaptation to differences in temperature entails changes in allele frequency by selection of genes involved in growth due to the effect of temperature on reproduction and the need of the species to find a balance between the amount of resources used in reproduction and somatic growth. In fact, as commonly observed in fish, the highest growth rate occurs before sexual maturity. Thus, differences in fecundity may be due to local adaptation, which could

393 394

E

392

Outlier

LG QTL Annotation

Sma42 Sma117 Sma22 Sma144 Sma147 Sma146 SmaE7 SmaE22 F8I11/8/17 SmaE33 2/5TG14 SmaE84 SmaE137 SmaE112 SmaE167 Sma168 SmaE99 SmaE117

1 21 11 3 6 14 6 12 24 23 19 7 16 – 22 2 9 9

– – G R R – G – – – G – – – – R – –

SmaE12 SmaE4 Sma149 Sma142 SmaSNP247 SmaSNP281 SmaSNP314 SmaSNP298 SmaSNP181

17 8 15 17 – – 13 21 4

– – G/R – R – – – R

– – – – – – Fibroblast growth factor receptor – – Trifunctional enzyme subunit beta, mitochondrial-like – Cathepsin O Fibrinogen alpha chain-like Villin-1 T-complex protein 1 subunit gamma-like – Cytosolic malate dehydrogenase thermostable form Microtubule-associated tumor suppressor 1 homolog A-like Trap alpha-translocon protein Beta-microglobuline – – Ribosomal protein L18a Phosphatidyl inositol transfer protein Ribosomal protein L1 Queuine t-RNA ribosyltransferase Ribosomal protein L13

Please cite this article as: Vilas, R., et al., A genome scan for candidate genes involved in the adaptation of turbot (Scophthalmus maximus), Mar. Genomics (2015), http://dx.doi.org/10.1016/j.margen.2015.04.011

424 425 426 427 428 429 430 431 432 433 434 435 436

t7:5 t7:6 t7:7 t7:8 t7:9 t7:10 t7:11 t7:12 t7:13 t7:14 t7:15 t7:16 t7:17 t7:18 t7:19 t7:20 t7:21 t7:22 t7:23 t7:24 t7:25 t7:26 t7:27 t7:28 t7:29 t7:30 t7:31 t7:32

444 445 446 447 448 449 450 451 Q9 452 453 454 455

458

Unlike the microsatellite markers, the outlier analysis based on SNPs revealed a much lower number of significant markers. This might be attributed to different mutational dynamics on the functional constraints 461 experienced (DeFaveri et al., 2013), but also that these markers had not 462 been a priori selected by their association to previously detected QTLs. 463 Five gene-associated SNPs showed evidence for divergent selection, 464 but only when BS was included in the pairwise analyses. The outlier pat465 tern was always consistent with a model assuming divergent selection 466 between BS and the remaining populations. These results suggest local 467 adaptation of turbot populations to salinity (although there are other 468 potential drivers than salinity such as differences in environmental pol469 lution or in any other selective pressure). However, it is striking that at 470 least four of the five outliers are either linked to genes coding for ribo471 somal proteins or directly related to the production of cellular proteins. 472 In addition, the two most significant SNPs are within ribosomal genes 473 located in two QTL for resistance to pathologies, which could be related 474 to the known influence of salinity on the immune system of fish 475 (Bowden, 2008). It is interesting that SNP outliers seem to be related 476 to the BS sample while microsatellite outliers are mainly related to the 477 Atlantic samples. This difference could be related to the close functional 478 relationship of the SNP outliers. The relatively low number of SNP out479 liers detected in Atlantic populations suggests an effect on SNPs specif480 ically linked to the Baltic Sea and it is likely that the functional 481 relationship is not a coincidence. 482 The Baltic Sea, almost fully surrounded by land and heavily influ483 enced by freshwater runoff, has a distinct salinity gradient from south 484 to north (HELCOM, 2003). The limited exchange rate explains the very 485 high level of pollution (HELCOM, 2011; Hong et al., 2012). Therefore, 486 at least two hypotheses could explain putative signatures of divergent 487 selection on ribosomal genes associated to the comparisons between 488 the Baltic with any other population: (1) they are the outcome of local 489 adaptation to salinity or (2) they are related to environmental pollution. 490 It is known that the expression of ribosomal proteins is induced by ex491 Q10 posure of marine fish to pollutants (Campbell and Devlin, 1997; 492 Handley-Goldstone et al., 2005; Oh et al., 2009). Although differences 493 in gene expression do not necessarily imply changes in allele frequen494 cies, eventually ribosomal genes may record the effects of natural selec495 tion leading to adaptation to a more polluted environment; particularly, 496 as in this case, when markers are located within the coding region or the 497 5′UTR (Hoffman and Willi, 2008; Roelofs et al., 2010). On the other 498 hand, if turbot populations from the Baltic Sea have adapted to low sa499 linity conditions by increased fecundity, a trait correlated with differ500 ences in female size, this adaptation may involve changes in genes

Acknowledgments

538

We thank Lucía Insua for the technical assistance and Carlos Fernández for the help with the turbot database. SNP genotyping services were provided by the University of Santiago de Compostela node of the Spanish National Centre of Genotyping (CeGen-ISCIII). S.V. acknowledges a PhD scholarship of ILVO-Vlaanderen and G.E.M. acknowledges a fellowship of the Research Foundation-Flanders. This study was supported by the Consolider Ingenio Aquagenomics (CSD200700002) and the Science and Education Spanish Ministry (AGL2009-11782).

539

References

548

Alberto, F.J., Derory, J., Boury, C., Frigerio, J.M., Zimmermann, N.E., Kremer, A., 2013. Imprints of natural selection along environmental gradients in phenology-related genes of Quercus petraea. Genetics 195, 495–512. Álvarez, I., de Castro, M., Gómez-Gesteira, M., Prego, R., 2005. Inter- and intra-annual analysis of the salinity and temperature evolution in the Galician Rías Baixas-ocean boundary (northwest Spain). J. Geophys. Res. 110 (C04008). Barrett, R.D.H., Hoekstra, H.E., 2011. Molecular spandrels: test of adaptation at the genetic level. Nat. Rev. Genet. 12, 767–780. Beaumont, M.A., 2005. Adaptation and speciation: what can FST tell us? Trends Ecol. Evol. 20, 435–440. Beaumont, M.A., Balding, D.J., 2004. Identifying adaptive genetic divergence among populations from genome scans. Mol. Ecol. 13, 969–980. Beaumont, M.A., Nichols, R.A., 1996. Evaluating loci for the use in the genetic analysis of population structure. Proc. Biol. Sci. 263, 1619–1626. Bierne, N., 2011. The distinctive footprints of local hitchhiking in a varied environment and global hitchhiking in a subdivided population. Evolution 64, 3254–3272. Bierne, N., Welch, J., Loire, E., Bonhomme, F., David, P., 2011. The coupling hypothesis: why genome scans may fail to map local adaptation genes. Mol. Ecol. 20, 2044–2072.

549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566

U

N

C

O

R

R

E

C

T

459 460

F

443

O

4.2. SNP based outliers and the unique position of the Baltic Sea

441 442

501 502

R O

457

439 440

related to somatic growth and reproduction (Nissling et al., 2013). Genes encoding ribosomal proteins may show the effects of natural selection for growth because the biosynthesis of these proteins is related to the rate of ribosomal formation, an essential process in actively dividing tissues. However, both hypotheses are not mutually exclusive because ribosomal proteins also show extra-ribosomal functions probably relevant in response to environmental stress. Hence, the unexpected behavior of these SNPs could be simply the result of adaptation to harsher environmental conditions. All explanations considered so far assumed that selection plays a causal role in the atypical behavior of these loci. However, outliers could be evidencing other phenomena such as differences associated with the hybridization of distinct populations at the study area, not necessarily related to adaptation to different environmental conditions (Bierne et al., 2013). The possibility of hybrid zones of turbot from the transition zone between the Baltic Sea and the North Sea has been suggested by Nielsen et al. (2004). The effect of different selective pressures acting on both sides of the transition zone reduces gene flow in this area. In such case, outlier analyses would identify genomic regions with reduced gene flow due to its putative contribution to local adaptation. However, outliers in cryptic hybrid zones could also be revealing genetic incompatibilities or differences in mate choice between diverging populations (Bierne et al., 2011). Actually, there are many other caveats that make it advisable to consider FST outliers just as candidate loci, requiring further investigation (Bierne et al., 2013). Here, we identified several loci that deserve the status of candidates to be involved in the adaptation of turbot populations most likely related to differences in temperature and salinity (and maybe pollution). Furthermore, results suggest that the presumed adaptive process entails changes in growth rate as a side effect, which is consistent with studies suggesting that turbot in the Baltic Sea grows slower and reaches a smaller size in comparison with other populations from the North Atlantic Ocean and the Mediterranean Sea (Stankus, 2003; Nissling et al., 2013). Finding genetic markers associated with growth whose variation in natural populations may be showing the effects of adaptation to different environmental conditions has implications for aquaculture of this species because they may serve to improve conditions for cultivation.

P

456

SmaE117 and SmaE167. Locus SmaE12 was an outlier only when the analysis included BS or FAR. Nonetheless, this locus showed higher FST than expected under neutrality when the analysis includes BS and the opposite pattern when this sample was replaced by FAR. This suggests that the variation at SmaE12 is driven in opposite directions by these two populations under different evolutionary pressures, including those potentially imposed by differences in salinity and growth rate. This marker is closely linked to the gene coding for the trap alphatranslocon protein, which is related to membrane processes (Vilas et al., 2010). An explanation for these contrasting patterns is the adaptation to varying conditions of salinity, negatively affecting the reproductive success. Consistent with this hypothesis, it has been observed that eggs of turbot from the Baltic populations have high survival rates in salinities between 10‰ and 15‰, conditions under which no or only low survival occurs among the turbot from the North Sea (Karas and Klingsheim, 1997; Nissling et al., 2006). Recent observations suggest that selection for a high fecundity result in lower somatic growth in the Baltic Sea (Nissling et al., 2013). However, as in the case of temperature, it is likely that other environmental factors correlated with differences in salinity may be operating (Vandamme et al., 2014).

D

437 438

R. Vilas et al. / Marine Genomics xxx (2015) xxx–xxx

E

8

Please cite this article as: Vilas, R., et al., A genome scan for candidate genes involved in the adaptation of turbot (Scophthalmus maximus), Mar. Genomics (2015), http://dx.doi.org/10.1016/j.margen.2015.04.011

503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537

540 541 542 543 544 545 546 547

R. Vilas et al. / Marine Genomics xxx (2015) xxx–xxx

N C O

R

R

E

C

D

P

R O

O

F

Nielsen, E.E., Hemmer-Hansen, J., Larsen, P.F., Bekkevold, D., 2009. Population genomics of marine fishes: identifying adaptive variation in space and time. Mol. Ecol. 18, 3128–3150. Nissling, A., Johansson, U., Jacobsson, M., 2006. Effects of salinity and temperature conditions on the reproductive success of turbot (Scophthalmus maximus) in the Baltic Sea. Fish. Res. 80, 230–238. Nissling, A., Florin, A.B., Thorsen, A., Bergström, U., 2013. Egg production of turbot, Scophthalmus maximus, in the Baltic Sea. J. Sea Res. 84, 77–86. Nosil, P., Funk, D.J., Ortiz-Barrientos, D., 2009. Divergent selection and heterogeneous genomic divergence. Mol. Ecol. 18, 375–402. Oh, J.H., Moon, H.B., Choe, E.S., 2009. Changes in differentially expressed genes in the liver of Oryzias latipes by binary exposure to carcinogenic polycyclic aromatic hydrocarbons. Kor. J. Environ. Biol. 27, 391–396. Orsini, L., Spanier, K.I., De Meester, L., 2012. Genomic signature of natural and anthropogenic stress in wild populations of the waterflea Daphnia magna: validation in space, time and experimental evolution. Mol. Ecol. 21, 2160–2175. Pardo, B.G., Hermida, M., Fernández, C., Bouza, C., Pérez, M., Llavona, A., Sánchez, L., Martínez, P., 2006. A set of highly polymorphic microsatellites useful for kinship and population analysis in turbot (Scophthalmus maximus L.). Aquac. Res. 37, 1578–1582. Pardo, B.G., Fernández, C., Hermida, M., Vázquez-López, A., Pérez, M., Presa, P., Calaza, M., Álvarez-Dios, J.A., Comesaña, A.S., Raposo-Guillán, J., Bouza, C., Martínez, P., 2007. Development and characterization of 248 novel microsatellites markers in turbot (Scophthalmus maximus). Genome 50, 329–332. Pereiro, P., Balseiro, P., Romero, A., Dios, S., Forn-Cuni, G., Fuste, B., Planas, J.V., Beltrán, S., Novoa, B., Figueras, A., 2012. High throughput sequence analysis of turbot (Scophthalmus maximus) transcriptome using 454-pyrosequencing for the discovery of antiviral immune genes. Plos One 7 (e35369). Raymond, M., Rousset, F., 1995. GENEPOP (version 1.2): population genetics software for exact test and ecumenicism. J. Hered. 86, 248–249. Ribas, L., Pardo, B.G., Fernández, C., Álvarez-Dios, J.A., Gómez-Tato, A., Quiroga, M.I., Planas, J.V., Sitja-Bobadilla, A., Martínez, P., Piferrer, F., 2013. A combined strategy involving Sanger and 454 pyrosequencing increases genomic resources to aid in the management of reproduction, disease control and genetic selection in the turbot (Scophthalmus maximus). BMC Genomics 14, 180. Robertson, A., 1975. Remarks on the Lewontin–Krakauer test. Genetics 80, 396. Rodríguez-Ramilo, S.T., Toro, M.A., Bouza, C., Hermida, M., Pardo, B.G., Cabaleiro, S., Martínez, P., Fernández, C., 2011. QTL detection for Aeromonas salmonicida resistance related traits in turbot (Scophthalmus maximus). BMC Genomics 12, 541. Rodríguez-Ramilo, S.T., Fernández, J., Toro, M.A., Bouza, C., Hermida, M., Fernández, C., Pardo, B.G., Cabaleiro, S., Martínez, P., 2013. Uncovering QTL for resistance and survival time to Philasterides dicentrarchi in turbot (Scophthalmus maximus). Anim. Genet. 44, 149–157. Rodríguez-Ramilo, S.T., De la Herrán, R., Ruiz-Rejón, C., Hermida, M., Fernández, C., Pereiro, P., Figueras, A., Bouza, C., Toro, M.A., Martínez, P., Fernández, J., 2014. Identification of quantitative trait loci associated to resistance to viral haemorrhagic septicaemia (VHS) in turbot (Scophthalmus maximus): a comparison between bacterium, parasite and virus diseases. Mar. Biotechnol. http://dx.doi.org/10.1007/s10126013-9544-x. Roelofs, D., Morgan, J., Sturzenbaun, S., 2010. The significance of genome-wide transcriptional regulation in the evolution of stress tolerance. Evol. Ecol. 24, 527–539. Roesti, M., Hendry, A.P., Salzburger, W., Berner, D., 2012. Genome divergence during evolutionary diversification as revealed in replicate lake-stream stickleback population pairs. Mol. Ecol. 21, 2852–2862. Rogers, S.M., Bernatchez, L., 2005. Integrating QTL mapping and genome scans towards the characterization of candidate loci under parallel selection in the lake whitefish (Coregonus clupeaformis). Mol. Ecol. 14, 351–361. Rogers, S.M., Bernatchez, L., 2007. The genetic architecture of ecological speciation and the association with signatures of selection in natural lake whitefish (Coregonus sp., Salmonidae) species pairs. Mol. Biol. Evol. 24, 1423–1438. Sánchez-Molano, E., Cerna, A., Toro, M.A., Bouza, C., Hermida, M., Pardo, B.G., Cabaleiro, S., Fernández, C., Martínez, P., 2011. Detection of growth-related QTL in turbot (Scophthalmus maximus). BMC Genomics 12, 473. Schlötterer, C., 2002. A microsatellite-based multilocus screen for the identification of local selective sweeps. Genetics 160, 753–763. Schoville, S.D., Bonin, A., François, O., Lobreaux, S., Melodelima, C., Manel, S., 2012. Adaptive genetic variation on the landscape: methods and cases. Annu. Rev. Ecol. Evol. Syst. 43, 23–43. Slate, J., 2005. Quantitative trait locus mapping in natural populations: progress, caveats and future directions. Mol. Ecol. 14, 363–379. Stankus, S., 2003. The peculiarities of turbot (Psetta maxima L.) biology and their role in the ecosystem of the Baltic Sea coastal zone of Lithuania. Acta Zool. Lit. 13, 217–238. Stinchcombe, J.R., Hoekstra, H.E., 2007. Combining population genomics and quantitative genetics: finding the genes underlying ecologically important traits. Heredity 100, 158–170. Storz, J.F., 2005. Using genome scans of DNA polymorphism to infer adaptive population divergence. Mol. Ecol. 14, 671–688. Suzuki, N., Nishida, M., Yoseda, K., ÜstÜndağ, C., Şahin, T., Amaoka, K., 2004. Phylogeographic relationships within the Mediterranean turbot inferred by mitochondrial DNA haplotype variation. J. Fish Biol. 65, 580–585. Tsakas, S., Krimbas, C.B., 1976. Testing heterogeneity of F-values: suggestion and a correction. Genetics 84, 399–401. Vandamme, S.G., Maes, G., Raeymaekers, J., Cottenie, K., Imsland, A.K., Hellemans, B., Lacroix, G., Mac Aoidh, E., Martinsohn, J., Martínez, P., Robbens, J., Vilas, R., Volckaert, F., 2014. Environmental selective pressure reinforces historical differentiation in turbot (Scophthalmus maximus). Mol. Ecol. 23, 618–636.

E

T

Bierne, N., Roze, D., Welch, J., 2013. Pervasive selection or is it…? Why are FST outliers sometimes so frequent? Mol. Ecol. 22, 261–264. Blanquer, A., Alayse, J.P., Berrada-Rkhami, O., Berrebi, P., 1992. Allozyme variation in turbot (Psetta maxima) and brill (Scophthalmus rhombus) (Osteichtyes, Pleuronectiformes, Scophthalmidae) throughout their range in Europe. J. Fish Biol. 41, 725–736. Bouza, C., Sánchez, L., Martínez, P., 1997. Gene diversity analysis in natural populations and cultured stocks of turbot (Scophthalmus maximus L.). Anim. Genet. 28, 28–36. Bouza, C., Presa, P., Castro, J., Sánchez, L., Martínez, P., 2002. Allozyme and microsatellite diversity in natural and domestic populations of turbot (Scophthalmus maximus) in comparison with other Pleuronectiformes. Can. J. Fish. Aquat. Sci. 59, 1460–1473. Bouza, C., Hermida, M., Pardo, B.G., Fernández, C., Fortes, G.G., Castro, J., Sánchez, L., Presa, P., Pérez, M., Sanjuán, A., de Carlos, A., Álvarez-Dios, J.A., Ezcurra, S., Cal, R.M., Piferrer, F., Martínez, P., 2007. A microsatellite genetic map of turbot (Scophthalmus maximus). Genetics 177, 2457–2467. Bouza, C., Hermida, M., Millán, A., Vilas, R., Vera, M., Fernández, C., Calaza, M., Pardo, B.G., Martínez, P., 2008. Characterization of EST-derived microsatellites for gene mapping and evolutionary genomics in turbot. Anim. Genet. 39, 666–670. Bouza, C., Hermida, M., Pardo, B.G., Vera, M., Fernández, C., de la Herrán, R., Navajas-Pérez, R., Álvarez-Dios, J.A., Gómez-Tato, A., Martínez, P., 2012. An Expressed Sequence Tag (EST)-enriched genetic map of turbot (Scophthalmus maximus): a useful framework for comparative genomics across model and farmed teleosts. BMC Genet. 13, 54. Bowden, T.J., 2008. Modulation of the immune system of fish by their environment. Fish Shellfish Immunol. 25, 373–383. Campbell, P.M., Devlin, R.H., 1997. Increased CYP1A1 and ribosomal protein L5 gene expression in a teleost: the response of juvenile Chinook salmon to coal dust exposure. Aquat. Toxicol. 38, 1–15. Colosimo, P.F., Hosemann, K.E., Balabhadra, S., Villareal, G., Dickson, M., Grimwood, J., Schmutz, J., Myers, R.M., Schluter, D., Kingsley, D.M., 2005. Widespread parallel evolution in sticklebacks by repeat fixation of ectodysplasin alleles. Science 307, 1928–1933. Coughlan, J.P., Imsland, A.K., Galvin, P.T., Fitzgerald, R.D., Naevdal, G., Cross, T.F., 1998. Microsatellite DNA variation in wild populations and farmed strains of turbot from Ireland and Norway: a preliminary study. J. Fish Biol. 52, 916–922. DeFaveri, J., Viitaniemi, H., Leder, E., Merilä, J., 2013. Characterizing genic and nongenic molecular markers: comparison of microsatellites and SNPs. Mol. Ecol. Resour. 13, 377–392. Florin, A.B., Höglund, J., 2007. Absence of population structure of turbot (Psetta maxima) in the Baltic Sea. Mol. Ecol. 16, 115–126. Foll, M., Gaggiotti, O., 2008. Estimating selection with different markers and varying demographic scenarios: a Bayesian perspective. Genetics 180, 977–993. Goudet, J., 2001. FSTAT, a program to estimate and test gene diversities and fixation indices (version 2.9.3). Available from. http://www.unil.ch/izea/softwares/fstat.html. Handley-Goldstone, H.M., Grow, M.W., Stegeman, J.J., 2005. Cardiovascular gene expression profiles of dioxin exposure in zebrafish embryos. Toxicol. Sci. 85, 683–693. Hansen, M.M., Olivieri, I., Waller, D.M., Nielsen, E.E., GeM Working Group, 2012. Monitoring adaptive genetic responses to environmental change. Mol. Ecol. 21, 1311–1329. HELCOM, 2003. The Baltic marine environment 1999–2002. Balt. Sea Environ. Proc. N° 87. HELCOM, 2011. The fifth Baltic Sea pollution load compilation (PLC-5). Balt. Sea Environ. Proc. No. 128. Hemmer-Hansen, J., Therkildsen, N.O., Meldrup, D., Nielsen, E.E., 2014. Conserving marine biodiversity: insights from life-history trait candidate genes in Atlantic cod (Gadus morhua). Conserv. Genet. 15, 213–228. Hermida, M., Bouza, C., Fernández, C., Sciara, A.A., Rodríguez-Ramilo, S.T., Fernández, J., Martínez, P., 2013. Compilation of mapping resources in turbot (Scophthalmus maximus): a new integrated consensus genetic map. Aquaculture 414, 19–25. Hoffman, A.A., Willi, Y., 2008. Detecting genetic responses to environmental change. Nat. Rev. Genet. 9, 421–432. Hong, B., Swaney, D.P., Mörth, C.M., Smedberg, E., Hägg, H.E., Humborg, C., Howarth, R.W., Bouraoui, F., 2012. Evaluating regional variation of net anthropogenic nitrogen and phosphorus inputs (NANI/NAPI), major drivers, nutrient retention pattern and management implications in the multinational areas of Baltic Sea basin. Ecol. Model. 227, 17–135. Iglesias, R., Paramá, A., Álvarez, M.F., Leiro, J., Fernández, J., Sanmartín, M.L., 2001. Philasterides dicentrarchi (Ciliophora: Scuticociliatida) as the causative agent of the scuticociliatosis in farmed turbot Scophthalmus maximus in Galicia (NW Spain). Dis. Aquat. Org. 46, 47–55. Imsland, A.K., Sunde, L.M., Folkvord, A., Stefansson, S.O., 1996. The interaction of temperature and fish size on growth of juvenile turbot. J. Fish Biol. 49, 926–940. Johanneson, K., Andre, C., 2006. Life on the margin: genetic isolation and diversity loss in a peripheral marine ecosystem, the Baltic Sea. Mol. Ecol. 15, 2013–2029. Karas, P., Klingsheim, V., 1997. Effects of temperature and salinity on embryonic development of turbot (Scophthalmus maximus L.) from the North Sea, and comparisons with Baltic populations. Helgol. Mar. Res. 51, 241–247. Lewontin, R., Krakauer, J., 1973. Distribution of gene frequency as a test of the theory of the selective neutrality of polymorphisms. Genetics 74, 175–195. Luikart, G., England, P.R., Tallmon, D., Jordon, S., Taberlet, P., 2003. The power and promise of population genomics: from genotyping to genome typing. Nat. Rev. Genet. 4, 981–994. Nei, M., Maruyama, T., 1975. Lewontin–Krakauer test for neutral genes. Genetics 80, 395. Nielsen, E.E., Nielsen, P.H., Meldrup, D., Hansen, M.M., 2004. Genetic population structure of turbot (Scophthalmus maximus L.) supports the presence of multiple hybrid zones for marine fishes in the transition zone between the Baltic Sea and the North Sea. Mol. Ecol. 13, 585–595.

U

567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652

9

Please cite this article as: Vilas, R., et al., A genome scan for candidate genes involved in the adaptation of turbot (Scophthalmus maximus), Mar. Genomics (2015), http://dx.doi.org/10.1016/j.margen.2015.04.011

653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738

10

739 740 741 742 743 744 745 746 747 748 749

R. Vilas et al. / Marine Genomics xxx (2015) xxx–xxx

Vasemägi, A., Primmer, C.R., 2005. Challenges for identifying functionally important genetic variation: the promise of combining complementary research strategies. Mol. Ecol. 14, 3623–3642. Vera, M., Álvarez-Dios, J.A., Millán, A., Pardo, B.G., Bouza, C., Hermida, M., Fernández, C., de la Herrán, R., Molina-Luzón, M.J., Martínez, P., 2011. Validation of single nucleotide polymorphism (SNP) markers from an immune Expressed Sequenced Tag (EST) turbot (Scophthalmus maximus) data base. Aquaculture 313, 31–41. Vera, M., Álvarez-Dios, J.A., Fernández, C., Bouza, C., Vilas, R., Martínez, P., 2013. Development and validation of single nucleotide polymorphisms (SNPs) markers from two transcriptomic 454-run of turbot (Scophthalmus maximus) using high throughput genotyping. Int. J. Mol. Sci. 14, 5694–5711.

Vilas, R., Bouza, C., Vera, M., Millán, A., Martínez, P., 2010. Variation in anonymous and EST-microsatellites suggests adaptive population divergence in turbot. Mar. Ecol. Prog. Ser. 420, 231–239. Vitalis, R., Dawson, K., Boursot, P., 2001. Interpretation of variation across marker loci as evidence of selection. Genetics 158, 1811–1823. Weir, B.S., Cockerham, C.C., 1984. Estimating F-statistics for the analysis of population structure. Evolution 38, 1358–1370.

U

N

C

O

R

R

E

C

T

E

D

P

R O

O

F

757

Please cite this article as: Vilas, R., et al., A genome scan for candidate genes involved in the adaptation of turbot (Scophthalmus maximus), Mar. Genomics (2015), http://dx.doi.org/10.1016/j.margen.2015.04.011

750 751 752 753 754 755 756