**4.1 Levels of population variation**

Low levels of intraspecific variation reported by many authors (Palumbi, 1996; Stepien et al., 1999; Saavedra & Pena, 2004; Pujolar et al., 2010) indicate that the mt 16S rDNA gene is highly conserved in many taxa, apparently due its functional role in protein assembly. However, it was not the case in the present study. Values of nucleotide diversity found for *F. glaber* (π=0.004) and *F. proteus* (π=0.005) are higher than the values reported for Adriatic populations of the same taxa (π=0.001 and π=0.000 respectively) (Pujolar et al., 2010). Also our values are higher than those estimated for other Pectinidae taxa as *Chlamys nobilis* (π =0.0012; Yuan et al., 2009), *C*. *farreri* (π=0.0035; Kong et al., 2003), *Pecten jacobaeus* (π=0.0015), *P*. *novaezelandiae* (π=0.0020), *P*. *fumatus* (π=0.000) and *P*. *maximus* (π=0.0039) (Saavedra & Pena, 2004).

The levels of intrapopulation variation for *F. glaber* and *F. proteus*, revealed by sequences of the mt DNA 16S rDNA region in the present study are also relatively high. Values of haplotype diversity ranged from 0.7793 to 0.9744 for populations of both taxa, are much higher than the values of 0.000 and 0.400 reported for the Adriatic populations (Pujolar et al., 2010) and higher than the values of 0.5356 and 0.7328 reported for two wild populations of *C*. *nobilis*, using sequences of the same region (Yuan et al., 2009). Also the highest value of π revealed in our study (0.0078) exceeds much the value of π=0.0012 observed in *C*. *nobilis* populations (Yuan et al., 2009). Kong et al. (2003) reported lower to similar values of π (0.00175 to 0.00591) and similar values of h (0.750 to 1.000) for different populations of Zhikong scallop *C. farreri*, based on the sequencing analysis of 16S ribosomal gene. Also, values of nucleotide and haplotype diversity observed in our samples are considerably higher than the values observed for different populations of the Asian moon scallop *Amusium pleuronectes* (0.000 to 0.0017 for π; 0.000 to 0.0511 for h) (Mahidol et al., 2007b).

The demographic analysis revealed that all the studied samples of both taxa were probably stable over time. Pujolar et al. (2010) who have studied the demographic history of the Adriatic *Flexopecten* complex using three mitochondrial and one nuclear gene proposed that the lack of large demogarphic expansion since the last glaciations, suggests that colonization must have been conducted by a substantial number of individuals that occupied the new habitat relatively fast. Demographic expansion does not appear either as a correct explanation of the observed genetic variation in our samples. A possible explanation for high values of intrapopulation indices for *F. glaber* and *F. proteus* could be the large effective population sizes as genetic variability is abundant. It is common knowledge that large *Ne* results to large amounts of neutral variability in populations.

An alternative hypothesis to explain this finding is that there is more than one species in the data set, but the character used to identify it is not valid. In other words, there may be more than one evolutionary entity, so we are artificially lumping more than one species in the analysis. One way to sort this out was to include other taxa in the analysis and make a phylogenetic analysis of the haplotypes, to see if they form unique phylogenetic species. There may simply be cryptic species to be considered or perhaps the morphotype doesn't correspond to the phylotype, which would certainly not be the first time in molluscs (Knowlton, 2000). The phylogenetic analysis of haplotypes revealed no clustering corresponding to each putative species (*F. glaber* and *F. proteus*). Also, the taxa *A. irradians* and *C. multistriata* formed a separate group in the Neighbour – Joining tree.

The decreased intrapopopulation genetic diversity can have serious consequences on the survival and reproduction of the species (Soule, 1980). Nevertheless, in the marine invertebrates reduction of genetic variation is expressed mainly with the loss of rare haplotypes despite with the total reduction of heterozygosity in a population (Gosling, 2003). Our samples are characterized from high DNA variability and the existence of some rare haplotypes, a result which indicates the good population status for both taxa. As the fishing of the Aegean *F. glaber* and *F. proteus* stocks is forbidden since 2002 because of heavy metals (Cd) presence in high concentrations (Koutsoubas et al., 2007), our data are fully justified.

Despite the high degree of genetic variation found within populations, the results of the statistical analysis of the 16S rDNA haplotypes indicated that little geographic structure was present among populations. The genetics of marine species with pelagic larval development has often been characterized by low genetic variation among populations, a pattern driven by high dispersal capabilities and large scale oceanic mixing (Reichow & Smith 2001; Rivera et al., 2004). After spawning and external fertilization, developing larvae spend a variable period of time as part of the plankton, which can be passively drifted by water currents (Seed, 1969; Cho et al., 2007). Similarly, gene flow among the populations of the present study seems not to be significantly blocked in this relatively small area of the Gulf of Thessaloniki. Consequently, our results support the general observation that marine invertebrates possessing a planktonic larval stage have a high genetic relatedness due to increased potential for larval dispersal (Crisp, 1978).

#### **4.2 Systematic status of the two taxa**

The presence of *F. proteus* in the Aegean Sea does not support the restricted gene flow or the recent origin of this form, reported by Pulojar et al. (2010). It seems that the high dispersal of the larval stage has provoked the expansion of this morph outside of the Adriatic Sea or *F. proteus* has appeared for the first time in other Mediterranean region, apart from Adriatic Sea.

The phylogenetic position of *Chlamys glabra* (now *F. glaber*) is under discussion. Previous phylogenetic data (Canappa et al., 2000a, 2000b; Barucca et al., 2004; Liu et al., 2007) support the idea that *C. glabra* seems to belong to a completely distinct genus from *Chlamys*, in agreement with some malacologists who have included it in the genus *Flexopecten* (Vaught, 1989; Rombouts, 1991). According to these studies *C. glabra* is clustered together with *Αequipecten opercularis* and not with other species of the genus *Chlamys* i. e. *C. islandica*, *C. varia* and *C. farreri*.

Later on Saavedra & Pena (2006) used the updated latin name *F. glaber* instead of *C. glabra* in a phylogenetic analysis for American scallops, and they also found that taxa *F. glaber* and *Α. opercularis* clustered together. There has been one effort for identification of *C. glabra* larvae from other bivalve species larvae, with a method based on PCR - SSCP combined with sequencing of partial 18S rDNA region, but it was not successful (Livi et al., 2006). In the latest study of the two Adriatic *Flexopecten* forms using three mitochondrial and one nuclear gene, Pujolar et al. (2010) suggest that *F. glaber* and *F. proteus* are the same species as evidenced by both putative species appearing mixed in all genetic trees, with no clustering according to species.

The decreased intrapopopulation genetic diversity can have serious consequences on the survival and reproduction of the species (Soule, 1980). Nevertheless, in the marine invertebrates reduction of genetic variation is expressed mainly with the loss of rare haplotypes despite with the total reduction of heterozygosity in a population (Gosling, 2003). Our samples are characterized from high DNA variability and the existence of some rare haplotypes, a result which indicates the good population status for both taxa. As the fishing of the Aegean *F. glaber* and *F. proteus* stocks is forbidden since 2002 because of heavy metals (Cd) presence in high concentrations (Koutsoubas et al., 2007), our data are fully

Despite the high degree of genetic variation found within populations, the results of the statistical analysis of the 16S rDNA haplotypes indicated that little geographic structure was present among populations. The genetics of marine species with pelagic larval development has often been characterized by low genetic variation among populations, a pattern driven by high dispersal capabilities and large scale oceanic mixing (Reichow & Smith 2001; Rivera et al., 2004). After spawning and external fertilization, developing larvae spend a variable period of time as part of the plankton, which can be passively drifted by water currents (Seed, 1969; Cho et al., 2007). Similarly, gene flow among the populations of the present study seems not to be significantly blocked in this relatively small area of the Gulf of Thessaloniki. Consequently, our results support the general observation that marine invertebrates possessing a planktonic larval stage have a high genetic relatedness due to

The presence of *F. proteus* in the Aegean Sea does not support the restricted gene flow or the recent origin of this form, reported by Pulojar et al. (2010). It seems that the high dispersal of the larval stage has provoked the expansion of this morph outside of the Adriatic Sea or *F. proteus* has appeared for the first time in other Mediterranean region, apart from Adriatic

The phylogenetic position of *Chlamys glabra* (now *F. glaber*) is under discussion. Previous phylogenetic data (Canappa et al., 2000a, 2000b; Barucca et al., 2004; Liu et al., 2007) support the idea that *C. glabra* seems to belong to a completely distinct genus from *Chlamys*, in agreement with some malacologists who have included it in the genus *Flexopecten* (Vaught, 1989; Rombouts, 1991). According to these studies *C. glabra* is clustered together with *Αequipecten opercularis* and not with other species of the genus *Chlamys* i. e. *C. islandica*, *C.* 

Later on Saavedra & Pena (2006) used the updated latin name *F. glaber* instead of *C. glabra* in a phylogenetic analysis for American scallops, and they also found that taxa *F. glaber* and *Α. opercularis* clustered together. There has been one effort for identification of *C. glabra* larvae from other bivalve species larvae, with a method based on PCR - SSCP combined with sequencing of partial 18S rDNA region, but it was not successful (Livi et al., 2006). In the latest study of the two Adriatic *Flexopecten* forms using three mitochondrial and one nuclear gene, Pujolar et al. (2010) suggest that *F. glaber* and *F. proteus* are the same species as evidenced by both putative species appearing mixed in all genetic trees, with no clustering

increased potential for larval dispersal (Crisp, 1978).

**4.2 Systematic status of the two taxa** 

justified.

Sea.

*varia* and *C. farreri*.

according to species.

The present study tried to resolve the two Aegean taxa relationship using a mitochondrial gene. Sequencing analysis of 16S rDNA revealed no taxon specific positions that could discriminate between different species. Additionally, two highly frequent and six less frequent haplotypes are shared between the two taxa, while no clustering in different groups was detected. This is an evidence of interbreeding between *F. glaber* and *F. proteus*, suggesting that they are conspecific taxa. Similarly, Wilding et al. (1999) found no high frequency private haplotypes within *P. maximus* and *P*. *jacobaeus* after a mtDNA PCR - RFLP analysis, and this fact together with the low genetic divergence demonstrated that the two species are also capable of hybridization. Later on, Saavedra & Pena (2004), in a phylogenetic analysis of *P. maximus* and *P. jacobaeus* noticed that the two most common haplotypes appeared in both taxa and this supports the held view of conspecificity between the two European taxa.

Another evidence of conspecific taxa was the levels of genetic distance. The genetic distance found between the two taxa (D=0.005) does not suggest genetic differences between two species, and shown to be of a similar magnitude to intraspecific than interspecific values. This value is considerably lower than expected for congeneric species (Liu et al., 1998; Stepien et al., 1999; Saavedra & Pena 2004; Yu et al., 2004; Mahidol et al., 2007a). Our data cannot even justify the classification of *F. proteus* as a subspecies of *F. glaber*. According to Therriault et al. (2004), the previous classification of *Dreissena bugensis* as a subspecies of *D*. *rostriformis* should be revised because genetic distance between them was relatively low (0.003–0.004; a value similar to our results), suggesting that these two taxa constitute a single species.

The existence of a single species in our study is also reinforced by the exact test of population differentiation. When pooling together the three sampling sites of *F. glaber* and *F. proteus* respectively, no statistically significant population differentiation was detected. So, our findings support the aspect that *F. proteus* is only a form or simply a synonyme of *F. glaber*.

In Bivalve, differences in shell morphology are often the result of phenotypic plasticity, so that many past descriptions of species were unjustified (reviewed by Knowlton, 2000). Consequently, the number of genetic studies that have resulted in the synonimization of species is probably higher for bivalves than most other marine invertebrate phyla. Our study is one example of numerous cases where distinctions between sympatric taxa based on shell morphology were not supported by genetic data (Wilding et al., 1999; reviewed by Knowlton, 2000). Present data reinforce also the validity of the mtDNA markers for clarifying uncertain systematic relationships among taxa.

A global DNA-based barcode identification system that is applicable to all animal species provides a universal tool for the identification of different species. The barcode system is based on sequence diversity in a single gene region (a section of the mitochondrial DNA cytochrome c oxidase I gene, COI). When the reference sequence library is in place, new specimens can be identified by comparing their DNA barcode sequences against this barcode reference library. Herbert et al. (2004a, 2004b) have demonstrated that the COI region is appropriate for discriminating between closely related species and these results have prompted international efforts to accelerate the process of cryptic species identification (Herbert et al., 2003). A similar survey based on the barcode identification system, could be done in the future for the two *Flexopecten* forms, in order to confirm the existence of a single species.



Appendix 1. Polymorphic sites, haplotype frequencies and genetic diversity indices, in all three studied populations of both taxa. The nucleotides in each position are given in comparison to haplotype H1. For all haplotypes variable sites are indicated, while identity is given by dots. In brackets it is the sample size. gA: *F. glaber* from Airport; gP: *F. glaber* from Paliomana; gN: *F. glaber* from Naziki; pA: *F. proteus* from Airport; pP: *F. proteus* from Paliomana; pN: *F. proteus* from Naziki.

120 Aquaculture

**5. Appendix** 
