Fistularia commersonii (Teleostea: Fistulariidae): walking through the Lessepsian paradox of mitochondrial DNA

Italian Journal of Zoology, 2015, 1–14 http://dx.doi.org/10.1080/11250003.2015.1046958

Fistularia commersonii (Teleostea: Fistulariidae): walking through the Lessepsian paradox of mitochondrial DNA

D. SANNA1*, F. SCARPA1, T. LAI1, P. COSSU1, M. FALAUTANO2, L. CASTRIOTA2, F. ANDALORO2, M. C. FOLLESA3, P. FRANCALACCI1, M. CURINI-GALLETTI1, & M. CASU1 1

Dipartimento di Scienze della Natura e del Territorio, Sezione di Zoologia, Archeozoologia e Genetica, Università di Sassari, Italy, 2ISPRA (Istituto Superiore per la Protezione e la Ricerca Ambientale), STS Palermo, Italy, and 3Dipartimento di Scienze della Vita e dell’Ambiente, Università di Cagliari, Italy (Received 2 October 2014; accepted 21 April 2015)

Abstract The Mediterranean Lessepsian migrations excite the interest of biologists who are devoted to inferring the effects of selection on the genetic structure of immigrants. The bluespotted cornetfish Fistularia commersonii is an Indo-Pacific species that was first recorded in the Levantine cost of Mediterranean, and within a few years, it rapidly expanded throughout the entire basin. Studies on its genetic variability, performed via mitochondrial sequencing of the Mediterranean specimens, suggest that a limited number of mitochondrial lineages passed through the Suez Canal. However, nuclear markers provide a scenario, with a high genetic variability among the Mediterranean F. commersonii migrants, along with the occurrence of haplotype sharing between the Mediterranean and the Red Sea. The aim of this study was to enlarge the number of Mediterranean sites in order to evaluate if the rapid expansion and different patterns of spread of F. commersonii in the basin could have led to a genetic structuring. The analysis was carried out by sequencing mitochondrial D-loop I in individuals from Sardinia, Sicily, Tunisia, Lampedusa, Libya and Lebanon. Sequences available from previous studies were included in the data set, allowing us to obtain a data set that likely represents the entire distribution range of the species. Results suggest the possible occurrence of two mitochondrial lineages involved in the Mediterranean invasion of F. commersonii, a bottleneck may have caused a loss in the genetic variation, leading to the fixation of specific lineages as an adaptive response to the new environmental conditions. Keywords: Bluespotted cornetfish, bioinvasion, mitochondrial DNA, Mediterranean Sea

Introduction Biological invasions have always been a focus of researchers due to their potential for exploring the evolutionary forces that cause relevant changes in native biodiversity and ecosystem functioning (Simberloff 2006). Some of the most dramatic impacts have been reported in aquatic habitats (Kolar & Lodge 2002), where a great deal of competition occurs between native and alien species with similar ecological niches. Falling into this category, the spread of Indo-Pacific species from the Red Sea into the Mediterranean following the opening of the Suez Canal in 1869 (the so-called “Lessepsian migration”; Por 1978) represents the “most important biogeographic phenomenon

witnessed in the contemporary oceans” and offers a unique opportunity to study the ecological and evolutionary processes that occur during a veritable biological invasion. With regards to fish, the first alien fish species was reported in the Mediterranean in 1902 (Tillier 1902; from Mavruk & Avsar 2008). Since then, 85 new species of fish have entered the Mediterranean during the last century from the Suez Canal (Fricke et al. 2012), representing approximately a quarter to a half of the world’s marine fish invaders (Lockwood et al. 2007). A twofold factor may account for the high rate of migration of alien fish: the occurrence of several vacant ecological niches in the Mediterranean (EEA 1999; see Hierro et al. 2005;

*Correspondence: D. Sanna, Dipartimento di Scienze della Natura e del Territorio, Sezione di Zoologia, Archeozoologia e Genetica, Università di Sassari, Via Francesco Muroni 25, 07100 Sassari, Italy. Tel: +39 079 228630. Fax: +39 079 228665. Email: [email protected] This article was originally published with errors. This version has been corrected. Please see Corrigendum (http://dx.doi.org/10.1080/11250003.2015.1055164). © 2015 Unione Zoologica Italiana

2

D. Sanna et al.

Paavola et al. 2005); and, in the case of the competition between the native and invader species, the higher adaptation ability of the Lessepsian species compared to the indigenous Mediterranean species (Ben-Tuvia 1964; Por 1978). Because adaptation may also be related to the amount of genetic variability of a given species, several molecular studies have been carried out on Lessepsian invader fish in order to understand how the passage through the canal affects their genetic patterns (e.g. Hassan & Bonhomme 2005; Azzurro et al. 2006; Bariche & Bernardi 2009). Genetic analyses carried out using allozymic, nuclear and mitochondrial markers on Lessepsian migrants that successfully entered and established in the Mediterranean (see Bernardi et al. 2010 for details) provide evidence of an overall lack of genetic differentiation between the Mediterranean and Red Sea fish populations, without the genetic signature of bottlenecks or loss of diversity. Even when considering that the results obtained could have been biased by either the sampling design, which was primarily focused in the area near the Suez Canal, or the large amount of time that passed between the occurrence of the invasion and the assessment of its genetic variation, this phenomenon is slightly anomalous considering that genetic bottlenecks often represent a diffused feature in biological invasions (see Allendorf & Lundquist 2003; but see also Estoup & Guillemaud 2010 and references therein). Conversely, the sudden spreading of the IndoPacific bluespotted cornetfish, Fistularia commersonii Rüppell, 1838 (Teleostea: Fistulariidae) across the Mediterranean basin represents an engaging paradox, as it is the only Mediterranean Lessepsian fish that shows, at least in the mtDNA genome, a rough signature of a genetic bottleneck, with a strong genetic differentiation from its Red Sea counterpart (Golani et al. 2007; Bernardi et al. 2010; Sanna et al. 2011; Tenggardjaja et al. 2014). Despite its low level of mitochondrial genetic variation in the Mediterranean, F. commersonii is now considered one of the 100 “worst” invasive species in Europe (DAISIE 2008; http://www.europe-aliens.org). Only about 7 or 8 years after its discovery in the eastern Mediterranean (Golani 2000; Wirtz & Debelius 2003), the species has been able to colonise both the eastern and western coasts of the basin (see Azzurro et al. 2013 for details), becoming the farthest recorded Lessepsian fish migrant (CIESM 2009). Its recent rapid invasion is thought to be mediated by an active adult migration (Merella et al. 2010; Azzurro et al. 2013), occurring in parallel along both the southern and northern coasts of the Mediterranean (Azzurro et al. 2013).

The first molecular studies devoted to disentangling the genetic pattern of the F. commersonii Mediterranean populations used several mitochondrial DNA molecular markers (see Golani et al. 2007; Bernardi et al. 2010; Sanna et al. 2011; Tenggardjaja et al. 2014) and pointed out the occurrence of few haplotypes and low levels of genetic variance, consistent with the signature of a genetic bottleneck. Conversely, Tenggardjaja et al. (2014), using a nuclear marker that was supposed to be under selection (Rhodopsin), did not find evidence of decreasing genetic variability among the Mediterranean F. commersonii migrants but instead found that they shared haplotypes with the Red Sea populations. These authors suggested that selection, rather than differences between the male and female population dynamics, might explain the contrasting results observed between the mitochondrial and nuclear markers, likely playing an important role in shaping the genetic pattern of the species (Tenggardjaja et al. 2014 and references therein). Because the use of mitochondrial markers serves many functions for studying biological invasions, including the identification of source populations, testing models of population growth and stochastic forces, distinguishing between single and multiple invasions, and estimating gene flow and dispersal among invasive populations (see Avise et al. 1987 for a review), further analysis on a geographic area encompassing the entire distribution range of the species could be helpful for shedding some light on the role played by the evolutionary forces in shaping the mitochondrial pattern of F. commersonii in the Mediterranean. Therefore, in this study, we aimed to investigate the evolutionary forces that likely affected the Mediterranean invasion of F. commersonii, from a mitochondrial perspective. We used the first segment of the mitochondrial control region (D-loop I) to analyse samples from the western and eastern Mediterranean basins and to perform genetic analyses inclusive of the samples from the entire distribution range of F. commersonii. D-loop I is, at present, the sole marker that allows the merging of newly obtained sequences with the sequences that are currently available for the species from other geographic regions (eastern Mediterranean and the Red Sea: Golani et al. 2007; Indian Ocean, South Pacific Ocean and North Pacific Ocean: Bernardi et al. 2010). Even though D-loop I is not directly under selective pressure, being linked with the coding region of mtDNA, it could be affected by the selective processes which act on the entire mitochondrial genome (e.g. Ballard & Whitlock 2004; Bazin et al. 2006; Galtier et al. 2009).

MtDNA genetic structure of Fistularia commersonii

3

Materials and methods

DNA extraction

Sampling

DNA was isolated from the muscular tissue using the Qiagen DNeasy tissue kit, and the DNA concentration was estimated via fluorometry (average value per sample, 15 ng/μl).

Forty-nine Fistularia commersonii specimens from different western (Sardinia) and eastern (Tunisia, Lampedusa, Libya, Lebanon) Mediterranean sites were analysed (see Table I and Figure 1 for details). The sequences provided by Golani et al. (2007) and Bernardi et al. (2010) were used to obtain a 281-bp alignment D-loop I data set consisting of 162 individuals that were representative of the Mediterranean Sea (Italy, Tunisia, Libya, Lebanon, Israel, Greece, Turkey: MED), the northern Red Sea (Egypt, Israel: RED), the Indian Ocean (Seychelles: IND), the southern Pacific Ocean (French Polynesia: SPAC) and the northern Pacific Ocean (Mexico: NPAC) regions.

Poylmerase chain reaction (PCR) The mitochondrial regions were amplified using the primers CR-A and CR-E (Lee et al. 1995), which amplify a 281-bp fragment of D-loop I. Each 25 μl PCR mixture contained approximately 75 ng of total genomic DNA, 0.2 μM of each primer, 2.5 U of EuroTaq DNA Polymerase (Euroclone), 1× reaction buffer and 200 µM of dNTP mix. The MgCl2 concentration was set at 2.5 mM. The

Table I. Data collection: marine ecoregion (ME), sampling locality, label (LB), sampling period (SP), and sample sizes (N) for individuals of Fistularia commersonii belonging to the geographic areas considered. Marine ecoregions: Western Mediterranean (WES); Tunisian Plateau/Gulf of Sidra (TUN); Levantine Sea (LEV); Aegean Sea (AEG); Northern and Central Red Sea (RED). Area

State

Mediterranean Italy

ME WES

TUN

Tunisia Libya Israel

Red Sea Indian Ocean Pacific Ocean

Lebanon Turkey Greece Egypt Israel Seychelles French Polynesia Mexico

LEV

AEG RED

Locality

LB

N

SP

# GenBank

Sardinia Arbatax

ARB

3

Calamosca Capo Malfatano Capo Comino Golfo di Cagliari

CAL CML COM CAG

1 1 1 6

10/2005, 10/2007, 01/2008 09/2007 12/2007 09/2007 09–11/2007

Golfo dell’Asinara Oristano Sant’Antioco Su Pallosu Torre delle Stelle Sicily Sciacca*1 Cavallo Bianco Lampedusa island Lampedusa island*1 Sfax Teboulba Tripoli Ashdod*1 Haifa*1 Jaffa*1 Ouza – Beirut Karaburnu*1 Rhodes island*1 Marsa Alam*1 Eilat*1 Seychelles*2 Moorea *2 Rangiroa*2 Baja California*2

ASN ORI SAN PAL TDS

1 1 2 1 1

08/2011 11/2007 10/2007 11/2007 01/2008

KM521157 KM521166 KM521151 KM521155–6, KM521159–61, KM521165 KM521168 KM521154 KM521162–3 KM521164 KM521167

SCI LMP

1 23

12/2005 01/2005, 11/2007

– KM521175–97

LAM SFX TEB TRI ASH HAI JAF LIB TUR RHO MAL EIL SEY MOO RAN MUE

12 1 3 2 3 12 2 1 1 21 15 29 2 3 5 7

01/2005 12/2006 12/2006 12/2006 06/2003 09/2005 10/2003 07/2013 09/2006 09–12/2005 12/2005 12/2004, 06/2005 01/2002 11/2006 – –

– KM521173 KM521170–1, KM521174 KM521169, KM521172 – – – KM521198 – – – – – – – –

KM521150, KM521152–3

Asterisks (*) and superscript numbers identify samples whose sequences were taken from 1Golani et al. (2007) and 2Bernardi et al. (2010).

4

D. Sanna et al.

Figure 1. Sampling plan: map indicating the localisation of the collecting sites. In the Mediterranean: (1) ASN; (2) ORI, TDS, PAL; (3) SAN; (4) CAL, CML, CAG; (5) ARB; (6) COM; (7) SCI; (8) TEB; (9) SFX; (10) TRI; (11) LMP, LAM; (12) TUR; (13) RHO; (14) LIB; (15) HAI; (16) JAF; (17) ASH. The localities are labelled as reported in Table I.

PCR were performed according to the following steps: 1 cycle of 2 min at 94°C, 35 cycles of 1 min at 94°C, 1 min at 48°C and 1 min 30 s at 72°C. At the end of the amplification, a post-treatment of 5 min at 72°C and a final cooling at 4°C was applied. The electrophoretic runs were carried out at 4 V/cm for 20 min on 2% agarose gels, which were made using 0.5× TBE buffer containing ethidium bromide (10 mg/ml) to stain the DNA fragments. The PCR products did not show any occurrence of aspecificity, therefore excluding the possibility of multiple nuclear mtDNA-like sequences. The PCR products were purified using ExoSAP-IT (USB Corporation) and sequenced using an external sequencing core service (Macrogen Inc., Europe). Each mitochondrial Dloop I fragment was sequenced in both the forward and reverse direction, and the corresponding sequencing runs were repeated twice in order to verify the reliability of the results. Dual peaks of similar height, which could be interpreted as evidence of mitochondrial pseudogenes in the nucleus (Numts) or heteroplasmy, were not observed in any of the electropherograms. Data analysis The sequences were aligned using the program Clustal W (Thompson et al. 1994), implemented in the BioEdit 7.1.3.0 software package (Hall 1999)

and then deposited in the GenBank database (accession numbers: KM521150–KM521198, see Table I for details). The level of genetic polymorphism for each biogeographical sector was assessed to estimate the number of polymorphic sites (s), the number of haplotypes (H), the nucleotide diversity (π), the haplotype diversity (h) and the mean number of pairwise differences (d), using the software package DnaSP 5.10 (Librado & Rozas 2009). A median-joining network (Bandelt et al. 1999) was constructed using the software package Network 4.6.1.2 (www.fluxus-engineering.com) to infer the genetic relationships among the haplotypes and to detect the occurrence (if any) of discrete genetic clusters. The transitions and transversions were equally weighted. Due to the lack of knowledge regarding the possible occurrence of retromutation events, the same weight (10) was assigned to all of the observed polymorphisms. The presence of population genetic structure was assayed by the Bayesian model-based clustering algorithm implemented in BAPS 5.3 (Corander & Tang 2007), which uses genetic information to assign individuals probabilistically to groups without presuming predefined populations (Pearse & Crandall 2004). Clustering was performed using the module for linked molecular data and applying the codon linkage model, which is appropriate for sequence data. Each analysis was run 10 times with a vector of K values = 1 to 22, each with 6

MtDNA genetic structure of Fistularia commersonii replicates. Haplotypes were organised into haplogroups following the partition of sequences into the distinct genetic clusters evidenced by Bayesian clustering. The historical population dynamics were assessed by comparing the observed mismatch distribution of the DNA substitution pairwise differences to a model of sudden population expansion (Rogers & Harpending 1992; Schneider & Excoffier 1999) using the software Arlequin 3.5.1.3 (Excoffier & Lischer 2010). Such distributions are unimodal if the populations have experienced recent expansion and multimodal if the populations are at demographic equilibrium or when the populations are significantly subdivided. The sum of square differences (SSD) between the observed and expected mismatch distributions was used to test the probability of obtaining a simulated SSD that was larger than or equal to that observed. Tajima’s D (1989) and Fu’s Fs (1997) neutrality tests were used to infer any departures from the population equilibrium models using the software Arlequin. Significant negative Tajima’s D values are expected to occur in cases of recent population expansion, severe bottlenecks and/or after a selective sweep. Positive values are expected in cases of balancing selection, population subdivision or recent bottlenecks. Similarly, significant negative Fu’s Fs values indicate an excess of rare haplotypes, which could be caused by recent population expansion, whereas positive values indicate balancing selection, population structure or moderate bottlenecks (Soriano et al. 2008 and references therein). Combining different neutrality tests can help to distinguish between the different evolutionary processes responsible for departures from the equilibrium; Fu’s Fs is superior for detecting demographic expansions, whereas Tajima’s D can better detect bottlenecks and population contractions (Soriano et al. 2008).

5

The best probabilistic model of sequence evolution was determined using the program jModelTest 2.1.1 (Posada 2008) and included an optimised maximum likelihood search using the Akaike Information Criterion (AIC) to select the best-fitting model. jModelTest selected the GTR+I+G model of nucleotide substitution as the model that best fit our data. The occurrence of significant levels of genetic structuring among biogeographical sectors was tested using a hierarchical analysis of molecular variance (AMOVA). We used the approach and the settings described in Excoffier et al. (1992), implemented in the software Arlequin. The Tamura and Nei (1993) genetic distances with a gamma distribution (α = 0.331) were used, according to the most likely model of sequence evolution estimated with jModelTest using the default options.

Results Lessepsian migration Individuals were grouped according to the location of the collecting sites in some of the marine ecoregions proposed by Spalding et al. (2007) for Mediterranean and Red Sea: (1) Western Mediterranean: WES (Sardinia); (2) Tunisian Plateau/Gulf of Sidra: TUN (Tunisia, Lampedusa, Libya, Sicily); (3) Aegean Sea: AEG (Turkey, Greece); (4) Levantine Sea: LEV (Israel, Lebanon); (5) Northern and Central Red Sea: RED (Egypt, Israel). In the Mediterranean, 8 different haplotypes (Table II), defined by 14 polymorphic sites, were identified. The total mean haplotype and nucleotide diversity values were h = 0.671 and π = 0.0154, respectively. The lowest haplotype and nucleotide diversity values were found among individuals from the eastern Mediterranean basin (AEG, LEV). In particular, the entire sample from

Table II. Sample sizes and levels of genetic diversity obtained for each Mediterranean marine ecoregions (Spalding et al. 2007) considered. N, sample sizes; S, number of polymorphic sites; H, number of haplotypes; %: relative frequency of different haplotypes per group; h, haplotype diversity; π, nucleotide diversity; d, mean of pairwise nucleotide differences. Mediterranean marine ecoregions: Western Mediterranean (WES); Tunisian Plateau/Gulf of Sidra (TUN); Levantine Sea (LEV); Aegean Sea (AEG); Northern and Central Red Sea (RED). Marine ecoregions are labelled as in Table I. Sites with gaps were not considered. Sample WES TUN LEV AEG RED Tot

N

S

H

%

h

π

d

19 42 18 22 44 145

12 13 0 12 64 66

3 7 1 2 41 49

15.79 16.67 0 9.09 93.18 33.79

0.485 0.719 0 0.173 0.997 0.841

0.0164 0.0198 0 0.0074 0.0496 0.0346

4.608 5.560 0 2.078 13.884 9.693

6

D. Sanna et al.

Figure 2. Lessepsian migration data set: median-joining network with the haplotypes grouped according to the sampling locations of the individuals in the major marine ecoregions proposed by Spalding et al. (2007) for Mediterranean and Red Sea: Western Mediterranean (Sardinia); Tunisian Plateau/Gulf of Sidra (Tunisia, Libya, Lampedusa, Sicily); Aegean Sea (Turkey, Rhodes); Levantine Sea (Israel, Lebanon); Northern and Central Red Sea: RED (Egypt, Israel).The small black plots on the nodes show median vectors representing the hypothetical connecting sequences that were calculated using the maximum parsimony method. The numbers of mutations between haplotypes that are greater than one are reported on the network branches. The numbers inside the spots indicate the number of individuals sharing the sequence. The spots without numbers are unique to a single individual.

the Levantine Sea exhibited a single haplotype. Higher values of genetic diversity were found for the westernmost collecting sites (WES, TUN), and the samples from the Tunisian Plateau/Gulf of Sidra region showed the highest levels of genetic variation (h = 0.719 and π = 0.0198), although these values may be slightly biased by a discrepancy in the sample size, as this area included twice the number of individuals (42) compared to the other areas considered in this study. Overall, individuals from the westernmost Mediterranean regions (WES, TUN) exhibit rates of genetic variation that are nearly two- to three-times higher than those reported for the easternmost sites (LEV, AEG). In Red Sea, 41 different haplotypes (Table II), defined by 64 polymorphic sites, were identified. The total mean haplotype and nucleotide diversity values were h = 0.997 and π = 0.0050, respectively. A list of all of the haplotypes that were identified is reported in supplementary Table S1. No evidence of haplotype sharing was found among Mediterranean and Red Sea. Furthermore, in Mediterranean, three

haplotypes encompassed the 95% of individuals, while in the Red Sea all haplotypes were privately owned with the only exception of three haplotypes, each of them occurring in two individuals. The median-joining network analysis, carried out on the Mediterranean and Red Sea samples (Figure 2), identified two main clusters of haplotypes (clusters A and B in the Figure 2). Four individuals occupied an intermediate position between the two clusters. Cluster A exclusively included individuals from the Red Sea, while cluster B encompasses individuals from the Red Sea along with the whole sample from the Mediterranean. Within the Mediterranean, the three most represented haplotypes were widespread in the entire basin without a geographical distribution restricted to only one of the marine ecoregions proposed by Spalding et al. (2007). The entire set of samples from the Levantine Sea, along with 91% of the individuals from the Aegean Sea, seven samples from Lampedusa, and one sample from Sicily, share the most diffused haplotype. The second most common haplotype, which diverged for a single point

MtDNA genetic structure of Fistularia commersonii mutation, spread throughout the Western Mediterranean and Tunisian Plateau/Gulf of Sidra regions. The third most common Mediterranean haplotype diverged from the former two for 12 point mutations. This haplotype was found in individuals from all of the Mediterranean areas that were considered, with the exception of the Levantine Sea. It is mainly widespread in the Tunisian Plateau/Gulf of Sidra region (65% of the individuals). Ten of the 17 Red Sea individuals included in cluster B laid on a separated branch sharing a private common ancestor. The Bayesian assignment analysis (see boxes in Figure 2) identified five distinct haplotype groups, hereafter denoted as groups I, II, III, IV and V (see Supplementary Tables S1 and S2 for details on the frequency distribution of the groups among the

7

individuals). Group I was exclusive to Mediterranean samples, corresponding to the two most represented haplotypes in the basin along with two derived sequences from the Western Mediterranean. Group II was less common and encompassed individuals from the Red Sea, which were placed on a separate branch of the cluster B of network. Group III was spread across the remaining Mediterranean individuals belonging to cluster B, along with some samples from the Red Sea. Group IV was the least widespread, being exclusively present in four individuals from the Red Sea. And finally, group V was the second most widespread, being only represented among individuals from Red Sea. The mismatch distribution analysis carried out on Mediterranean samples (Figure 3A) showed a

Figure 3. Lessepsian migration data set: graphs of the mismatch distributions of the Mediterranean samples. The x-axis reports the observed distribution of the pairwise nucleotide differences, and the y-axis reports the frequencies. SSD, sum of squared deviations. The significant values are denoted in bold.

8

D. Sanna et al.

bimodal distribution of pairwise DNA differences with the occurrence of a peak close to the origin (starting shape) along with a further unimodal right-handed peak. Nonetheless, the observed mismatch distribution still fits the expected distribution under a sudden population expansion model (SSD = 0.082, P = 0.232). The mismatch distribution of the samples from the Red Sea (Figure 3B) resembled the unimodal mismatch distribution expected under the Rogers and Harpending (1992) model of sudden demographic expansion (SSD = 0.003, P = 0.545). The values of Tajima’s D (D = 1.647, P = 0.954) and Fu’s Fs tests (Fs = 4.694, P = 0.928), carried out on the Mediterranean sample, did not provide evidence of a significant departure from the equilibrium models that were expected under neutrality. Nonetheless, the positive value of Tajima’s D could provide some evidence of bottlenecks occurred in the Mediterranean. With regard to the Red Sea samples, Fu’s Fs showed a significant departure from the equilibrium (Fs = −24.197, P ˂ 0.001), whereas Tajima’s D did not (D = −0.232, P = 0.462), possibly reflecting the reduced power of D compared to Fs at detecting demographic expansions (Soriano et al. 2008). Therefore, the discordant result of the two tests, whose statistical power differs according to the underlying population scenario, may be consistent with a population expansion. The AMOVA (Table S3) was carried out for the Mediterranean sample in order to evidence, if any, the occurrence of genetic structuring within the basin. Results showed a significant genetic differentiation (ΦCT = 0.325, P = 0.004) between the four groups of individuals defined according to the biogeographic criteria (Spalding et al. 2007): (1) Western Mediterranean; (2) Tunisian Plateau/Gulf of Sidra; (3) Aegean Sea; (4) Levantine Sea. However, the genetic differentiation was maximised (ΦCT = 0.408, P ˂ 0.001) when the samples from

Sicily and Lampedusa were considered as a separate fifth group. When alternative groupings of the samples were tested, the AMOVA showed a decrease in the proportion of ΦCT variance (data not shown).

Phylogeographical analysis at global level Sixty-three different haplotypes (Table III), defined by 73 polymorphic sites, were identified. The total mean haplotype and nucleotide diversity values were h = 0.766 and π = 0.0374, respectively. The lowest haplotype and nucleotide diversity values were identified among individuals from the Mediterranean sites (h = 0.671 and π = 0.0154), which showed eight haplotypes. Higher values of genetic diversity (h = 0.997 and π = 0.0497) were identified at the sites outside of the Mediterranean, which were characterised by a high rate of different haplotypes (92%). The median-joining network analysis (Figure 4) showed a high level of variability, with 89% of sequences consisting of geographically private haplotypes. The only exception was observed for two individuals from the southern Pacific Ocean (SPAC), which share a haplotype with one individual from the Red Sea (RED); an overall lack of haplotype sharing was evidenced among all of the analysed regions (MED, RED, IND, SPAC, NPAC). Samples from the Indian Ocean (IND) and the southern Pacific Ocean (SPAC) spread throughout both clusters A and B identified in the previous reported median-joining network analysis (see Figure 2), while the northern Pacific Ocean (NPAC) sample is placed on a diverging private branch included in cluster A. There is no evidence of spatial structuring at either the small or large geographic scale. However, two exceptions were identified in individuals from the northern Pacific Ocean (Mexico, NPAC) and individuals belonging

Table III. Sample sizes and levels of genetic diversity obtained for each geographical macro-area considered. N, sample sizes; S, number of polymorphic sites; H, number of haplotypes; %, relative frequency of different haplotypes per group; h, haplotype diversity; π, nucleotide diversity; d, mean of pairwise nucleotide differences. Mediterranean Sea (MED), Red Sea (RED), Indian Ocean (IND), southern Pacific Ocean (SPAC), northern Pacific Ocean (NPAC). Sites with gaps were not considered. Sample

N

S

H

%

h

π

d

MED RED IND SPAC NPAC Tot

101 44 2 8 7 162

14 64 16 34 9 73

8 41 2 7 7 63

7.92 93.18 100 87.5 100 38.89

0.671 0.997 1.000 0.964 1.000 0.766

0.0154 0.0496 0.0569 0.0492 0.0095 0.0374

4.327 13.884 16.000 13.821 2.571 10.062

MtDNA genetic structure of Fistularia commersonii

9

Figure 4. Phylogeographical survey data set: median-joining network with the haplotypes coloured according to the geographic distribution of the individuals in the five macro-areas. Boxes I–V report the results of the Bayesian assignment. The small black plots on the nodes show the median vectors representing the hypothetical connecting sequences that were calculated using the maximum parsimony method. The number of mutations between the haplotypes that are greater than one are reported on the network branches. The numbers inside the spots indicate the number of individuals sharing the sequence. The spots without numbers are unique to a single individual.

to Mediterranean cluster B, which lay in two different peripheral positions in the network. The Bayesian assignment analysis (see boxes in Figure 4) identified five distinct haplotype groups, corresponding to groups I, II, III, IV and V already reported (see boxes in Figure 2; see Supplementary Tables S1 and S2 for details on the frequency distribution of the groups among the individuals). Group I, widespread in 48% of the individuals, is exclusive to the Mediterranean samples. Group II encompassed individuals from the Red Sea (RED) and the southern Pacific Ocean (SPAC). Group III spread across Mediterranean individuals along with some samples from the Red Sea (RED) and the Indian Ocean (IND). Group IV was widespread in 3% of the individuals only, being exclusively present among five individuals from the Red Sea (RED). Group V was widespread in 20% of the individuals, being represented throughout all regions considered, with the exception of the Mediterranean. The mismatch distribution of the pairwise DNA differences carried out on the entire data set was not significantly different from that expected under the population expansion model (SSD = 0.034,

P = 0.330), although it showed more than one peak (Figure 5). Neutrality tests carried on the entire data set showed negative but nonsignificant values for Tajima’s D (D = −0.678, P = 0.278). Significant negative Fu’s Fs neutrality values (Fs = −23.865, P = 0.001) were a possible consequence of an excess of rare alleles, which might be consistent with a recent population expansion or the occurrence of selective sweep (Fu 1997). AMOVA (Table S4) was used to detect the largest differences between the groups (MED, RED, SPAC, NPAC). Samples with less than five individuals (Seychelles, IND) were excluded from the population-level analysis due to a lack of statistical power. The analysis maximised the highest genetic differentiation (ΦCT = 0.456, P ˂ 0.001) when three groups were considered: (1) NPAC, (2) MED and (3) RED, SPAC. However, a similar value of molecular variance (ΦCT = 0.448, P ˂ 0.01) was obtained when treating the Red Sea (RED) and the southern Pacific Ocean (SPAC) as separate groups [(1) NPAC, (2) MED, (3) RED and (4) SPAC]. When alternative groupings

10

D. Sanna et al.

Figure 5. Phylogeographical survey data set: graphs of the mismatch distributions carried out for the entire data set (a) and the Red Sea samples (b). The x-axis reports the observed distribution of the pairwise nucleotide differences and the y-axis reports the frequencies. SSD, sum of squared deviations. The significant values are denoted in bold.

of the samples were tested [e.g. (1) NPAC+SPAC, RED, MED; (2) NPAC+SPAC+RED, MED; and (3) NPAC, SPAC+RED, MED], the AMOVA showed a decrease in the proportion of ΦCT variance (data not shown). Furthermore, when the possible occurrence of a genetic relationship between the Mediterranean and Red Sea was tested (MED+RED, SPAC, NPAC) the proportion of ΦCT variance approached zero (ΦCT = 0.028, P = 0.830). Discussion Fistularia commersonii is considered to be one of the most invasive species in the Mediterranean. Its rapid invasion occurred in parallel along both the southern and northern coasts of the basin, but different rates of environmental suitability for the species were reported by Azzurro et al. (2013). The authors suggested that environmental parameters may represent a crucial hindrance to the progress of F. commersonii, which tends to invade coastal zones with an average biological productivity and high salinity, while avoiding both highly productive and oligotrophic areas. The single individual captured along the coast of Lebanon approximately 25 years before the Mediterranean invasion (Bariche et al. 2014) may represent an early introduction of F. commersonii that was not followed by a successful establishment of the population, suggesting that the environmental parameters occurring in the basin during that time period prevented any successful invasions. Subsequently, significant alterations in the biotic parameters of the Mediterranean, most likely due to the climate

change (see Lejeusne et al. 2010 references therein), may have favoured the spread of the species during the twenty-first century. To date, the three molecular mitochondrial surveys that have been conducted on F. commersonii (Golani et al. 2007; Sanna et al. 2011; Tenggardjaja et al. 2014) retrieved low variability at D-loop region in the Mediterranean, suggesting that high levels of genetic diversity may not bet a crucial factor for the successful invasion of this species. In contrast, Tenggardjaja et al. (2014) also showed a high level of genetic variability using the nuclear molecular marker Rhodopsin. This finding gives rise to a new question: are there environmental factors that exclusively affected the mitochondrial molecule during the Mediterranean invasion of F. commersonii? Tenggardjaja et al. (2014) hypothesised that adaptive evolution (see Willis & Orr 1993 for details) or selective sweep (see Maruyama & Birky 1991; Fu 1997; Filatov et al. 2000 for details) were the most likely factors to play (or have played) an important role in shaping the mitochondrial genetic patterns of the species, while they excluded the occurrence of a sex bias across the Suez Canal and the Mediterranean basin. Our study showed that the Mediterranean is characterised by the presence of few mitochondrial haplotypes, and remarkably none of which have ever been reported in the Red Sea. Although results may be biased by limits of the current sampling plan (e.g. 44 specimens from only 2 localities from the Red Sea could not be representative of the whole genetic variability of this area), at present such a finding may be explained as a possible consequence of the selective pressures

MtDNA genetic structure of Fistularia commersonii promoted by the new environmental condition of the Mediterranean. In such a context, specific mitochondrial genotypes may have undergone a selective sweep, which promoted the fixation and spread of alleles that were adapted to specific environmental condition, different from Red Sea. When adaptive evolution takes place, these genotypes can be rare or not expressed in the source population, but became common and distinctive in the introduced population as a consequence of the founder event (Willis & Orr 1993). In these cases, because only a small and rare portion of the genetic variability of the source population is transferred to the colonising population, it produces a reduction in the mitochondrial variation of migrants and a significant genetic structuring between the new habitat and the origin site. However, at the current state of knowledge, other alternative hypothesis cannot be ruled out a priori. For instance, the high level of pollutants in the Suez Canal (Gladstone et al. 1999) may have prompted repeated demographic bottlenecks that led to the lowering of the mitochondrial lineages of the F. commersonii migrants. Indeed, Por (1978) evidenced that mineral oil negatively affects the migration throughout the canal of some fish. Interestingly, the same genetic bottlenecks may have had positive impacts on the introduced individuals by purging deleterious mitochondrial alleles (for more details, see Bouzat 2010). Furthermore, as suggested by Azzurro et al. (2013), the low level of mitochondrial variability in F. commersonii invaders could be related to the high initial population growth rates. Indeed, during rapid range expansion rare alleles may reach high frequencies in the populations at the front of the expansion wave (Excoffier et al. 2009). The lower level of genetic variability of the mtDNA compared to the nuclear markers (see Tenggardjaja et al. 2014) could also be due to the occurrence of a stochastic lineage sorting, which was based on the lower effective population size of the mtDNA among the Lessepsian migrants. Finally, considering that “Mediterranean” genotypes of F. commersonii may be present in unsampled areas (e.g. the Southern Red Sea and the Gulf of Aden), a more extensive geographic sampling plan (at present limited to only two northern Red Sea sites), especially from other marine ecoregions of the source area of invasion, combined with a better molecular sampling (e.g. sequencing of fragments from the mitochondrial coding region, surveys of multilocus nuclear markers), are needed to clarify

11

the genetic structuring of the Mediterranean populations of F. commersonii. As well, a genomic scan coupled with outlier tests will allow the identification of loci that are candidate for divergent or balancing selection (Luikart et al. 2003). A suggestive result is that individuals from the central–west Mediterranean exhibited higher levels of genetic variation than those from the eastern Mediterranean, particularly from the Aegean Sea. We may explain this fact with the higher environmental suitability of the central Mediterranean region for the species, as hypothesised by Azzurro et al. (2013). It is also noteworthy that a slight but significant genetic substructuring was observed among the Mediterranean marine ecoregions surveyed. Likely, the observed genetic pattern may reflect (1) the different ecological conditions characterising the marine ecoregions, (2) the result of stochastic events (Hudson & Turelli 2003), and/or (3) the occurrence of multiple events of invasion from genetically distinct sources (Estoup & Guillemaud 2010) that have adapted later to well-defined environmental conditions. The occurrence of two groups of haplotypes, homogeneously distributed in the Mediterranean, but resulting in a slight genetic substructuring among the Mediterranean marine ecoregions, is also consistent with the rapid invasion of F. commersonii, which likely occurred along two pathways (the southern and northern coasts of the basin), as proposed by Azzurro et al. (2013). The environmental plasticity of the species, along with the lack of autochthonous competitors, may have allowed a rapid colonisation of ecologically different Mediterranean habitats, though the importance of these mechanisms still needs to be directly assessed in the future. However, because F. commersonii invaded the Mediterranean recently, being unlikely to have attained migration-drift equilibrium, we should pay caution in the interpretation of our results. Indeed, non-equilibrium models, in which few individuals colonise a new habitat and subsequently display rapid population growth, may create large differences among populations through founder effects (Allen et al. 2010). The high initial genetic divergence will persist until the time needed to reach equilibrium, which depends on the migration rate (Allen et al. 2010). We cannot exclude that observed patterns of population differentiation are related to different propagule pressures (e.g. Fields & Taylor 2014 and references therein). For instance, such a mechanism would be important if few individuals contribute to the gene pool of subsequent generation because of high reproductive variance, leading to temporal

12

D. Sanna et al.

chaotic genetic patchiness (Johnson & Black 1984). When considering the entire data set, the occurrence of two distinct groups within the species was observed. Although the low sample size of the IndoPacific F. commersonii individuals requires caution in data interpretation, this finding is suggestive of the existence of two divergent taxonomic units currently living in sympatry. Our results also showed a significant genetic structuring of the most peripheral populations (from the Mediterranean and Mexico) that potentially diverged in allopatry at the opposite edges of the distribution range of the species. On the contrary, the samples from the Red Sea and southern Pacific Ocean showed a moderate level of genetic affinity. It is worth noting that historical demographic analyses suggested an overall scenario of sudden demographic expansion, even with a background of subdivided populations throughout the entire distribution range (Mediterranean, Red Sea, North Pacific Ocean, South Pacific Ocean) of the species. Although the mitochondrial DNA of F. commersonii most likely underestimates the effective genetic variability of the Mediterranean Lessepsian migrants (Tenggardjaja et al. 2014), such a marker could represent a useful tool to infer the possible occurrence of discrete genetic clades throughout the entire distribution range of the species, thus allowing its use in reconstructing a species’ population history at different levels of the temporal scale (Casu et al. 2011; Sanna et al. 2013a, 2013b).

Acknowledgements We thank Dr Paolo Merella and Dr Antonio Pais for their help in collecting of specimens.

Funding This work was supported by MURST (Ministero per l’Università e la Ricerca Scientifica e Tecnologica) ex-60% (grant to PF), and by Fondazione Banco di Sardegna, 2013 (grant to MC).

Supplemental data Supplemental data for this article can be accessed here.

References Allen MR, Thum RA, Cáceres CE. 2010. Does local adaptation to resources explain genetic differentiation among Daphnia populations? Molecular Ecology 19:3076–3087. doi:10.1111/ mec.2010.19.issue-15. Allendorf FW, Lundquist LL. 2003. Introduction: Population biology, evolution, and control of invasive species. Conservation Biology 17:24–30. doi:10.1046/j.15231739.2003.02365.x. Avise JC, Arnold J, Ball RM, Bermingham E, Lamb T, Neigel JE, Reeb CA, Saunders NC. 1987. Intraspecific phylogeography: The mitochondrial DNA bridge between population genetics and systematics. Annual Review of Ecology and Systematics 18:489–522. Azzurro E, Golani D, Bucciarelli G, Bernardi G. 2006. Genetics of the early stages of invasion of the Lessepsian rabbitfish Siganus luridus. Journal of Experimental Marine Biology and Ecology 333:190–201. doi:10.1016/j. jembe.2005.12.002. Azzurro E, Soto S, Garofalo G, Maynou F. 2013. Fistularia commersonii in the Mediterranean Sea: Invasion history and distribution modeling based on presence-only records. Biological Invasions 15:977–990. doi:10.1007/s10530-012-0344-4. Ballard JWO, Whitlock MC. 2004. The incomplete natural history of mitochondria. Molecular Ecology 13:729–744. doi:10.1046/ j.1365-294X.2003.02063.x. Bandelt HJ, Forster P, Rohl A. 1999. Median-joining networks for inferring intraspecific phylogenies. Molecular Biology and Evolution 16:37–48. doi:10.1093/oxfordjournals.molbev. a026036. Bariche M, Bernardi G. 2009. Lack of a genetic bottleneck in a recent Lessepsian bioinvader, the blue-barred parrotfish, Scarus ghobban. Molecular Phylogenetics and Evolution 53:592–595. doi:10.1016/j.ympev.2009.06.017. Bariche M, Kazanjian G, Azzurro E. 2014. A lag of 25 years: Evidence from an old capture of Fistularia commersonii Rüppell, 1838 from Lebanon (Mediterranean Sea). Journal of Applied Ichthyology 30:535–536. doi:10.1111/jai.12394. Bazin E, Glémin S, Galtier N. 2006. Population size does not influence mitochondrial genetic diversity in animals. Science 312:570–571. doi:10.1126/science.1122033. Ben-Tuvia A. 1964. Two siganids fishes of Red Sea origin in the eastern Mediterranean. Bulletin of the Sea Fisheries Research Station, Haifa 37:3–10. Bernardi G, Golani D, Azzurro E. 2010. The genetics of Lessepsian bioinvasions. In: Golani D, Appelbaum-Golani B, editors. Fish Invasions of the Mediterranean Sea: Change and Renewal. Sofia-Moscow: Pensoft Publishers. pp. 71–84. Bouzat JL. 2010. Conservation genetics of population bottlenecks: The role of chance, selection, and history. Conservation Genetics 11:463–478. doi:10.1007/s10592-010-0049-0. Casu M, Sanna D, Cossu P, Lai T, Francalacci P, Curini-Galletti M. 2011. Molecular phylogeography of the microturbellarian Monocelis lineata (Platyhelminthes: Proseriata) in the NorthEast Atlantic. Biological Journal of the Linnean Society 103:117–135. doi:10.1111/bij.2011.103.issue-1. CIESM 2009. CIESM Atlas of exotic species in the Mediterranean Sea. Available: http://www.ciesm.org/online/ atlas/index.htm. Accessed July 2012 21. Corander J, Tang J. 2007. Bayesian analysis of population structure based on linked molecular information. Mathematical Biosciences 205:19–31. doi:10.1016/j.mbs.2006.09.015.

MtDNA genetic structure of Fistularia commersonii DAISIE 2008. European invasive Alien species gateway. Avaliable: http://www.europe-aliens.org. Accessed June 2012 15. EEA 1999. Environment in the European Union at the turn of the century. Environmental Assessment Report No. 2. Copenhagen, Denmark: European Environment Agency. Estoup A, Guillemaud T. 2010. Reconstructing routes of invasion using genetic data: Why, how and so what? Molecular Ecology 19:4113–4130. doi:10.1111/mec.2010.19.issue-19. Excoffier L, Foll M, Petit RJ. 2009. Genetic consequences of range expansions. Annual Review of Ecology, Evolution, and Systematics 40:481–501. doi:10.1146/annurev. ecolsys.39.110707.173414. Excoffier L, Lischer HEL. 2010. Arlequin suite ver 3.5: A new series of programs to perform population genetics analyses under Linux and Windows. Molecular Ecology Resources 10:564–567. doi:10.1111/j.1755-0998.2010.02847.x. Excoffier L, Smouse PE, Quattro JM. 1992. Analysis of molecular variance inferred from metric distances among DNA haplotypes: Application to human mitochondrial DNA data. Genetics 131:479–491. Fields PD, Taylor DR. 2014. Determinants of genetic structure in a non equilibrium metapopulation of the plant Silene latifolia. PLoS ONE 9:e104575. doi:10.1371/journal. pone.0104575. Filatov DA, Monéger F, Negrutiu I, Charlesworth D. 2000. Low variability in a Y-linked plant gene and its implications for Y-chromosome evolution. Nature 404:388–390. doi:10.1038/ 35006057. Fricke R, Golani D, Appelbaum-Golani B. 2012. First record of the Indian Ocean anchovy Stolephorus insularis Hardenberg, 1933 in the Mediterranean (Clupeiformes: Engraulidae). BioInvasions Records 1:303–306. doi:10.3391/ bir.2012.1.4.11. Fu YX. 1997. Statistical tests of neutrality of mutations against population growth, hitchhiking and background selection. Genetics 147:915–925. Galtier N, Nabholz B, Glémin S, Hurst GDD. 2009. Mitochondrial DNA as a marker of molecular diversity: A reappraisal. Molecular Ecology 18:4541–4550. doi:10.1111/ mec.2009.18.issue-22. Gladstone W, Tawfiq N, Nasr D, Andersen I, Cheung C, Drammeh H, Krupp F, Lintner S. 1999. Sustainable use of renewable resources and conservation in the Red Sea and Gulf of Aden: Issues, needs and strategic actions. Ocean & Coastal Management 42:671–697. doi:10.1016/S0964-5691(99) 00040-X. Golani D. 2000. First record of the Bluespotted cornetfish from the Mediterranean. Journal of Fish Biology 56:1545–1547. doi:10.1111/j.0022-1112.2005.00643.x. Golani D, Azzurro E, Corsini-Foka M, Falautano M, Andaloro F, Bernardi G. 2007. Genetic bottlenecks and successful biological invasions: The case of a recent lessepsian migrant. Biology Letters 3:541–545. doi:10.1098/rsbl.2007.0308. Hall TA. 1999. BioEdit: A user-friendly biological sequence alignment editor and analysis program for Windows 95/98/NT. Nucleic Acids Symposium Series 41:95–98. Hassan M, Bonhomme F. 2005. No reduction in neutral variability of mitochondrial and nuclear genes for a Lessepsian migrant, Upeneus moluccensis. Journal of Fish Biology 66:865– 870. doi:10.1111/jfb.2005.66.issue-3. Hierro JL, Maron JL, Callaway RM. 2005. A biogeographical approach to plant invasions: The importance of studying exotics in their introduced and native range. Journal of Ecology 93:5–15. doi:10.1111/jec.2005.93.issue-1.

13

Hudson RR, Turelli M. 2003. Stochasticity overrules the ‘‘threetimes rule’’: Genetic drift, genetic draft, and coalescence times for nuclear loci versus mitochondrial DNA. Evolution 57:182–190. Johnson MS, Black R. 1984. Pattern beneath the chaos: The effect of recruitment on genetic patchiness in an intertidal limpet. Evolution 38:1371–1383. Kolar CS:, Lodge DM. 2002. Ecological predictions and risk assessment for alien fishes in North America. Science 298:1233–1236. doi:10.1126/science.1075753. Lee WJ, Conroy J, Howell WH, Kocher TD. 1995. Structure and evolution of teleost mitochondrial control regions. Journal of Molecular Evolution 41:54–66. doi:10.1007/BF00174041. Lejeusne C, Chevaldonné P, Pergent-Martini C, Boudouresque CF, Pérez T. 2010. Climate change effects on a miniature ocean: The highly diverse, highly impacted Mediterranean Sea. Trends in Ecology and Evolution 25:250–260. doi:10.1016/j.tree.2009.10.009. Librado P, Rozas J. 2009. DnaSP v5: A software for comprehensive analysis of DNA polymorphism data. Bioinformatics 25:1451–1452. doi:10.1093/bioinformatics/btp187. Lockwood JL, Hoopes MF, Marchetti MP. 2007. Invasion ecology. Malden: Blackwell Publishing. Luikart G, England PR, Tallmon D, Jordan S, Taberlet P. 2003. The power and promise of population genomics: From genotyping to genome typing. Nature Reviews Genetics 4:981–994. doi:10.1038/nrg1226. Maruyama T, Birky Jr CW. 1991. Effects of periodic selection on gene diversity in organelle genomes and other systems without recombination. Genetics 127:449–451. Mavruk S, Avsar D. 2008. Non-native fishes in the Mediterranean from the Red Sea, by way of the Suez Canal. Reviews in Fish Biology and Fisheries 18:251–262. doi:10.1007/s11160-0079073-7. Merella P, Casu M, Garippa G, Pais A. 2010. Lessepsian fish migration: Genetic bottlenecks and parasitological evidence. Journal of Biogeography 37:978–980. doi:10.1111/ jbi.2010.37.issue-5. Paavola M, Olenin S, Leppäkoski E. 2005. Are invasive species most successful in habitats of low native species richness across European brackish water seas?. Estuarine Coastal and Shelf Science 64:738–750. doi:10.1016/j.ecss.2005.03.021. Pearse DE, Crandall KA. 2004. Beyond FST: Analysis of population genetic data for conservation. Conservation Genetics 5:585–602. doi:10.1007/s10592-003-1863-4. Por FD. 1978. Lessepsian migration – the influx of Red Sea biota into the Mediterranean by way of the Suez Canal. Berlin: Springer-Verlag. Posada D. 2008. jModelTest: Phylogenetic model averaging. Molecular Biology and Evolution 25:1253–1256. doi:10.1093/molbev/msn083. Rogers AR, Harpending H. 1992. Population growth makes waves in the distribution of pairwise genetic differences. Molecular Biology and Evolution 9:552–569. Sanna D, Biagi F, Ben Alaya H, Maltagliati F, Addis A, Romero A, De Juan J, Quignard J-P, Castelli A, Franzoi P, Torricelli P, Casu M, Carcupino M, Francalacci P. 2013a. Mitochondrial DNA variability of the pipefish Syngnathus abaster. Journal of Fish Biology 82:856–876. doi:10.1111/ jfb.2013.82.issue-3. Sanna D, Cossu P, Dedola GL, Scarpa F, Maltagliati F, Castelli A, Franzoi P, Lai T, Cristo B, Curini-Galletti M, Francalacci P, Casu M. 2013b. Mitochondrial DNA reveals genetic structuring of Pinna nobilis across the Mediterranean Sea. PlosOne. doi:10.1371/journal.pone.0067372.

14

D. Sanna et al.

Sanna D, Merella P, Lai T, Farjallah S, Francalacci P, CuriniGalletti M, Pais A, Casu M. 2011. Combined analysis of four mitochondrial regions allowed the detection of several matrilineal lineages of the Lessepsian fish Fistularia commersonii in the Mediterranean Sea. Journal of the Marine Biological Association of the United Kingdom 91:1289–1293. doi:10.1017/S0025315410001451. Schneider S, Excoffier L. 1999. Estimation of past demographic parameters from the distribution of pairwise differences when the mutation rates vary among sites: Application to human mitochondrial DNA. Genetics 152:1079–1089. Simberloff D. 2006. Invasional meltdown 6 years later: Important phenomenon, unfortunate metaphor, or both. Ecology Letters 9:912–919. doi:10.1111/j.1461-0248.2006.00939.x. Soriano AR, Ramos-Onsins SE, Rozas J, Calafell F, Navarro A. 2008. Statistical power analysis of neutrality tests under demographic expansions, contractions and bottlenecks with recombination. Genetics 179:555–567. doi:10.1534/genetics.107. 083006. Spalding MD, Fox HE, Allen GR, Davidson N, Ferdaña ZA, Finlayson M, Halpern BS, Jorge MA, Lombana A, Lourie SA, Martin KD, McManus E, Molnar J, Recchia CA, Robertson J. 2007. Marine ecoregions of the world: A bioregionalization of coastal and shelf areas. BioScience 57:573–583.

Tajima F. 1989. Statistical method for testing the neutral mutation hypothesis by DNA Polymorphism. Genetics 123:585–589. Tamura K, Nei M. 1993. Estimation of the number of nucleotide substitutions in the control region of mitochondrial DNA in humans and chimpanzees. Molecular Biology and Evolution 10:512–526. Tenggardjaja K, Jackson A, Leon F, Azzurro E, Golani D, Bernardi G. 2014. Genetics of a Lessepsian sprinter: The bluespotted cornetfish, Fistularia commersonii. Israel Journal of Ecology & Evolution. doi:10.1080/15659801.2013. 898402. Thompson JD, Higgins DG, Gibson TJ. 1994. CLUSTAL W: Improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Research 22:4673–4680. doi:10.1093/nar/22.22.4673. Tillier JB. 1902. Le Canal de Suez et sa faune ichthyologique. Memories de la Societe Zoologique de France 15:279–318. Willis JH, Orr HA. 1993. Increased heritable variation following population bottlenecks – the role of dominance. Evolution 47:949–957. Wirtz P, Debelius H. 2003. Mediterranean and Atlantic invertebrate guide. Hackenheim: Conchbooks.