Genetic Diversity of Three Spotted Seahorse , Hippocampus trimaculatus ( Leach , 1814 ) in India Using Four Microsatellite Loci

Seahorse populations are declining year by year not only in India but also throughout the world, because of over-fishing and increasing demand in Chinese market. The three spotted seahorse, Hippocampus trimaculatus is one of the dominant species and distributed all along the Indian coast. To study the genetic structure is very essential to conserve these species effectively. Hippocampus trimaculatus samples (n = 60/population) were collected from Mullimunai in Palk Bay, Tuticorin in Gulf of Mannar and Vizhinjam in south Malabar in India as by-catch in small trawlnets. Microsatellites are being widely applied in animal genome mapping and phylogenetic analysis because of their co-dominant inheritance and high degree of polymorphism. The molecular polymorphism of microsatellite DNA has proved to be a potent tool in the analysis of several aspects of population genetics. In the present study, four microsatellite primers were used to investigate the genetic difference and structure of three selected populations of H. trimaculatus. The result showed the overall FST value (0.0989) of the microsatellite loci between Mullimunai and Vizhinjam was significantly different (p<0.0001). The genetic distance between Mullimunai and Tuticorin was 0.183; between Tuticorin and Vizhinjam was 0.461; and Mullimunai and Vizhinjam was 0.837. There was no statistical evidence of recent severe bottlenecks in any of the three populations. Continuous monitoring of microsatellite variations within the populations of all the three locations was suggested to determine whether genetic variation within the populations is stabilized between year classes.


Introduction
A genetically inherited variant in which the phenotype can be inferred from the genotype during genetic screening is known as genetic marker.The utility of DNAbased marker is generally determined by the technology that is used to reveal DNA-based polymorphism (Gupta and Gupta, 2011).They are also used to identify different species in the event of taxonomic disputes and also to detect genetic diversity in a species.(Gopalakrishnan and Mohindra, 2001).Microsatellites are short tandemly arrayed di-, tri-, or tetra-nucleotide repeat sequences with repeat size of 1-6 bp repeated several times flanked by regions of non-repetitive unique DNA sequences (Tautz, 1989).The development of polymorphic microsatellite markers to determine the population structure of the Patagonian toothfish, Dissostichus eleginoides, has been reported by Reilly and Ward (1998).Microsatellites that have been largely utilized for population studies are single locus ones in which both the alleles in a heterozygote show co-dominant pattern of inheritance (Gopalakrish-nan and Mohindra, 2001).Many authors have reported microsatellite polymorphism and sequences in some marine and freshwater fish species using population genetic analysis (Appleyard et al., 2002;Han et al., 2000).Neff et al. (1999) described 10 microsatellite loci from the bluegill (Lepomis macrochirus) and discussed their evolution within the family Centarchidae.Kellog et al. (1995) applied microsatellite-fingerprinting approach to address questions about paternity in cichlids.
A number of attempts have been made to study the cross-species amplification of microsatellite loci in fishes.May et al. (1997) reported the microsatellite genetic variation through cross species amplification in sturgeons Acipenser and Scaphirhynchus.Only two studies were made for population genetics in seahorses using microsatellite marker in worldwide so far.Four polymorphic microsatellite loci were used to asses biological parentage from a natural population of the western Australian seahorse Hippocampus angustus ( Jones et al., 1998).Later, Jones et al. (2003) used three microsatellite loci for genetic parentage in Hippocampus subelongatus.There is no such molecu-DNA extraction and microsatellite DNA analysis Total DNA was extracted by standard phenol-chloroform method (Taggart et al., 1992).The DNA samples were screened for DNA variation at four microsatellite loci: Han 05, Han 06, Han 15 ( Jones et al., 1998), Han 16 ( Jones et al., 2003).Polymerase chain reactions (PCR) were carried out in a thermal cycler (M.J. Research, USA) employing the microsatellite primers.Amplifications were performed in 25 µl reaction mixture containing 1x reaction buffer (10 mM Tris, 50 mM KCl, 0.01% gelatin, pH 9.0) with 1.5 mM MgCl 2 , 5 pM of each primer, 200 mM dNTPs, 2 U Taq DNA polymerase (Genei, India) and 20 ng of template DNA.The temperature profile of the PCR was; 94°C for 3 min of an initial denaturing cycle followed by 35 cycles of 94°C denaturation for 30 s, a 50-54°C annealing for 30 s (varied according to primers), a 72°C extension cycle for 1 min; followed by a final extension step at 72°C for 5 min.Then each PCR product (2 μl) was electrophoresed in 10% acrylamide gel, which was fixed and stained with silver nitrate.

Data analysis
To analyze the variation in microsatellite loci, allele frequencies at each locus were calculated with Genetix software (Belkhir et al., 1997).A locus was considered to be polymorphic when the frequency of the most common allele was equal to or less than 0.99 (Nei, 1987).The mean number of alleles per locus, observed and expected heterozygosities (H obs and H exp ) and percentage of polymorphic loci for overall and each population were calculated with Genetix software.
Exact P-tests for conformity to Hardy-Weinberg Equilibrium (probability and score test) were performed by the Markov Chain method using Genepop (Raymond and Rousset, 1998) with parameters, dememorization = 1000; batches = 10 and iterations = 100; and based upon a null hypothesis of random union of gametes.The significant criteria were adjusted for the number of simultaneous tests using sequential Bonferroni technique (Rice, 1989).
The coefficient of genetic differentiation (F ST ) and the inbreeding coefficient (F IS ) were estimated through estimator of Weir and Cockerham (1984).Estimation of average F ST and determining whether the values are significantly lar study was reported in H. trimaculatus in India as well as in other countries.
Seahorse populations are declining year by year not only in India but throughout the world, because of overfishing and increasing demand in Chinese market (Lipton and Thangaraj, 2002).Among the six Indian seahorse species viz., H. kuda, H. fuscus, H. histix, H. kelloggi, H. trimaculatus and H. spinossisimus, the three spotted seahorse, H. trimaculatus is the dominant species and distributed throughout entire coast of the nation (Murugan et al., 2008).The body colour of this species is golden orange, sandy coloured to totally black, some have striped in brown and white colour.The distinguish characters are: very sharp hook like cheek and eye spine, quite flat in appearance, narrow head, no nose spine, large dark spots on the dorso-lateral surface of the first, fourth and seventh trunk rings more common in male than female.They inhabited more than 20 m deep waters with gravel, sandy or clay bottom.To conserve these species effectively, it is essential to study the genetic structure of it.In the present study, the genetic structure of three populations was assessed using microsatellite markers.

Collection of samples
The three spotted seahorse, Hippocampus trimaculatus (Fig. 1) samples (n = 60/population) were collected from Mullimunai (9° 44' N; 79° 10' E) in Palk Bay, Tuticorin (8° 47' N; 78° 80' E) in Gulf of Mannar and Vizhinjam (8° 21' N; 76° 58' E) in south Malabar (Fig. 2) as by-catch in small trawlnets.The tissue samples (finclips) for DNA extraction were stored in sterile eppendorf tubes containing 95% ethanol.different from zero; and calculation of pair-wise population F ST values (θ) and their significance levels, were carried out using Genepop.This program performs numerical resampling by bootstrapping (1000 times) and jack-knife procedures in order to estimate confidence intervals and the significance of values.The F ST value is used to estimate the amount of gene flow (Nm) using the corrected private -allele-based model of Barton and Slatkin (1986) in Genepop, between populations.The expected frequency of null alleles was calculated according to Van Oosterhoust et al. (2004, 2006) using Microchecker.
The three populations were tested for genetic bottleneck using the software Bottleneck to determine whether the populations exhibited the observed gene diversity higher than the expected equilibrium gene diversity by Wilcoxon sign-rank test (Luikart, 1996).Genetic similarity and distance between pairs of populations of H. trimaculatus were estimated using Popgene (Yeh et al., 1999).Nei and Li's (1979) pair-wise genetic similarity (SI) among the three populations were computed and converted by Popgene into genetic distance (GD) according to Hillis and Moritz's (1990) formula, GD = 1-SI.

Results
All the four amplified microsatellite loci were polymorphic in all the three populations.All the four primers indicated positive F IS values in different populations (Tab.1).The estimated null allele frequency was not significant (p < 0.05) at all the four tested loci using different algorithms, indicating the absence of null alleles and false homozygotes.
There was no significant association indicative of linkage disequilibrium (p > 0.05) between any pair of microsatellite loci for any population.It was therefore assumed that allelic variation at microsatellite loci could be considered independent.The F Statistics and gene flow (Nm) for each locus in the overall populations are given in Tab. 2. In adittion genetic variations such as observed heterozygosity (H obs ) and expected heterozygosity (H exp ) in the overall populations for each locus are given in Tab. 2. The summary of genetic variations including inbreeding coefficient (F IS ), probability value of significant deviation from Hardy Weinberg Equilibrium (P HW ), probability value of significant heterozygosity deficiency (P score ) are given in Tab. 1.

Discussion
Generally, the development of new species-specific microsatellite primers is expensive and time consuming, but cross-species amplification is cheap and fast.Primers developed for a particular species have been successfully tested for cross-species amplification in several fish species (Gopalakrishnan et al., 2004;Lal et al., 2003;Mohindra et al., 2000).In the present study, successful cross-priming was obtained with four primer pairs and all the loci were polymorphic and ideal to be used as markers in stock identification studies.However, the optimum annealing temperature to get scorable band in H. trimaculatus slightly differed from the resource species.The present study demonstrated successful cross-priming of microsatellite loci, between the fish species of closely related.The use of heterologous PCR primers would significantly reduce the cost of developing similar set of markers for other Hippocampus species found in India.
The number of alleles at different microsatellite loci in H. trimaculatus varied from 6 to 26 with an average value of 15.Primers Han 05 and Han 16 exhibited maximum The probability test revealed that the observed allele frequencies in most of the loci showed significant deviation (p < 0.05) from Hardy Weinberg Equilibrium (Tab.1).The Infinite Allele Model (IAM), Two Phase Model (TPM) and the significance test for each locus in the three populations are given in Tab. 3. The test did not exhibit significant excess of observed gene diversity (p > 0.05) on average loci for all the locations under both mutation models.There was no indication of reduction of rare alleles, indicating the absence of evidence of recent genetic bottleneck in any of the populations.
The coefficient of genetic differentiation (pair-wise F ST ) differed between populations significantly (p < 0.0001) and the result is given in Tab. 4. Nei's (1978) unbiased genetic distances estimated between pairs of three populations of H. trimaculatus are given in Tab. 5.The genetic distance between Mullimunai and Tuticorin was 0.183; between Tuticorin and Vizhinjam was 0.461; and Mullimunai and Vizhinjam was 0.837.exploitation (Gopalakrishnan et al., 2004;Rico et al., 1997, Yue et al., 2000).There were twelve stock specific alleles found in Mullimunai, thirteen in Tuticorin and nine alleles in Vizhinjam population.The detection of these private alleles in specific populations might be the absence of mixing of the gene pool between populations.In Clarias macrocephalus, Na-Nakorn et al. (1999) reported twenty stock-specific markers in three loci in four populations in Thailand.Scribner et al. (1996) reported 22 stock specific alleles in three populations of Chinook salmon (Oncorhynchus tshawytscha) in Canada.
The overall F ST value (0.0989) of the microsatellite loci between Mullimunai and Vizhinjam was significantly different (p < 0.0001).Levels of genetic differentiation demonstrated here for H. trimaculatus between the Mullimunai and Vizhinjam population (F ST = 0.0989) are comparable to those significant values seen in the Pacific herring (F ST = 0.023), Atlantic herring (F ST = 0.035) and widespread anadromous fish like Atlantic salmon (F ST = 0.054) (McConnell et al., 1995).In the present study, the Nm value indicated chances of restricted migration between populations and that genetic differentiation over time might have occurred due to genetic drift.However, low values of Nm have been reported in species exhibiting significant genetic differences and geographic isolation among populations as in the present study (Angers et al., 1995;Coughlan et al., 1998).
A significant reduction in microsatellite heterozygosity, but non-significant loss of microsatellite alleles was observed in all the three populations.Comparison of observed heterozygocities with that expected, under a two phase model (TPM) did not support the hypothesis of a severe bottleneck in the seahorse populations.Therefore, there was no statistical evidence of recent severe bottlenecks in any of the three populations.Continuous monitoring of microsatellite variations within the populations of all the three locations is suggested to determine whether genetic variation within the populations is stabilized between year classes.

Conclusions
The analysis using novel hypervariable microsatellite loci in H. trimaculatus revealed several significant results.First, the potential use of heterologous PCR primers was explored and they appeared to be conserved in this species.Second, all the four microsatellite loci were polymorphic and showed heterogeneity in allele frequency in three populations.Third, the study suggested that these three natural populations of this species, viz., Mullimunai, Tuticorin and Vizhinjam are divergent in their genetic characteristics can be clearly identified through microsatellite loci.
number of allele, meaning 22 and 26 respectively compared to other primers (six alleles).High microsatellite allele variation was recorded in Thai silver barb (Puntius gonionotus) in four microsatellite loci with average of 13.8 alleles per locus (Kamonrat, 1996), and in a number of marine fishes such as the whiting (14-23 alleles/locus) (Rico et al., 1997) and red seabream (16-32 alleles/locus) (Takagi et al., 1999).Jones et al. (1998) reported 9 to 22 alleles in the Australian seahorse, H. angustus by using four microsatellite loci primers.Neff and Gross (2001) reported that marine species had greater microsatellite allele variation as compared to freshwater species and that this was consistent with the increased effective population sizes.
Variations of allele sizes were quite low for Han-15 locus which might be due to their small number of repeat units, and similar level of allele size variation is reported in selected freshwater teleosts (Carvalho and Hauser, 1994).The same kind of relationship was observed in H. angustus ( Jones et al., 1998) where the allelic size of locus Han-15 was somewhat less with less number of repeat ([GT] 9 ).However, the locus, Han-06 has 22 alleles in the present study and this is comparable with the earlier reports of 22 alleles in the Australian seahorse ( Jones and Avise, 2001).The mean observed number of alleles at each locus (na = 15.00) in the present study was similar as in H. angustus (na = 15.75) ( Jones et al., 1998).
In H. trimaculatus, the mean observed heterozygosity (H obs ) for locus Han-05 per population was 0.2333 and the mean expected heterozygosity (H exp ) per locus per population was 0.7492.In the case of H. angustus, Jones et al. (1998), reported a higher value of mean observed heterozygosity (H obs = 0.9490) for the same locus.However, the mean expected heterozygosity (H exp ) was much higher for Han-06 in the present study (H exp = 0.9301), but in H. angustus it was less (H exp = 0.08770) ( Jones et al., 1998).In H. trimaculatus, a significant overall deficiency of heterozygosity was revealed in all the populations for all the loci whereas in the H. angustus all the loci showed significant by high value of heterozygotes ( Jones et al., 1998).
A significant deviation from HWE were found within samples across loci rather than within loci and across most of the samples.There was also no instance of non-amplifying samples in repeated trials with any of the primer pairs in H. trimaculatus.Van Oosterhout et al. (2004) suggested that, the overall homozygosity can be due to deviations from panmixia, inbreeding, short allele dominance, stuttering or large allele dropouts.
Seahorses are generally characterized by sparse distribution, low mobility, small home ranges, low fecundity, mate and habitat fidelity.Due to this, inbreeding can happen, which will result in the deficiency of heterozygotes and deviation from HWE.Similar situation was reported in other fishes that showed decline in catches due to over

Fig 1 .
Fig 1.The male three spotted seahorse, H. trimaculatus Fig 2. Map showing the collection site Tab. 1. Summary of genetic variation statistics of four microsatellite loci in three populations H obs = Observed heterozygosity; H nb = Heterozygosity (Non-biased) for small populations; F IS =Inbreeding coefficient; P HW = Probability value of significant deviation from HWE; P score = Probability value of significant heterozygosity deficiency; P (0.95) = Polymorphism at 0.95 criteria; P (0.99) = Polymorphism at 0.99 criteria; A n = Mean number of alleles per locus Tab. 2. F statistics and Observed (H obs ) and Expected (H exp ) heterozygosity of overall populations