Phenotype and genotype differentiation between flathead grey mullet [ Mugil cephalus ] and thinlip grey mullet [ Liza ramada ( Pisces : Mugilidae ) ]

This study aimed to study the phenotype and genotype differentiation and to compare the amount of differences in phenotype based on morphometric character indices and meristic counts with the amount of differences in genotype based on random amplified polymorphic DNA (RAPD) fingerprinting between two Mugilidae, Flathead grey mullet (Mugil cephalus) and Thinlip grey mullet (Liza ramada). The results showed that there were significant differences in most of the morphometric character indices but no significant differences were detected in most of the meristic counts between M. cephalus and L. ramada. In addition, the euclidean distance between M. cephalus and L. ramada using hierarchical cluster analysis of quantitative phenotype based on morphometric character indices and meristic counts and condition factor was (0.421). Genotype analysis based on RAPD fingerprint showed highly genetic dissimilarity (0.437) between M. cephalus and L. ramada. These results confirmed that the amount of differences in genotype reflected the same amount of differences in phenotype between M. cephalus and L. ramada. Therefore, either phenotype analysis can be use based on a large number of morphometric character indices and meristic counts or genotype analysis based on RAPD fingerprinting to discriminate M. cephalus and L. ramada with the same results.


INTRODUCTION
Mullets are euryhaline fish widely distributed in tropical and subtrobical and estuaries.Mullets are catadromous spawning migrating fish, the young life before maturity remains predominantly in the system of rivers and lakes (Lee and Tamaru, 1988;El-Deeb et al., 1996).Five mullet species occur in Egyptian water.According to the different regions, the most commonly-cultured of mullets are Mugil cephalus and Liza ramada (Oren, 1981;Lee and Tamaru, 1988;El-Sayed, 1991).It is considered an important source of animal protein for the peoples of the Pacific Basin, Southeast Asia, India, the Mediterranean, Eastern Europe and, many parts of central and South America (Lee and Tamaru, 1988).The identification of the fry to be used for stocking purposes is of practical interest and often causes problems for farmers who want to stock the ponds with specific species (Menezes et al., 1992).Different methods are used for identification but phenotype based on morphometric and meristic is considered as earliest and authentic methods for the identification of fish species in fish biology to measure discreteness and relationships among various taxonomic categories.There are many well documented morphometric studies which provide evidence for stock discreteness (Avsar, 1994;Haddon and Willis, 1995;Bembo et al., 1996;Anene, 1999;Turan, 1999;Zafar et al., 2002;Barriga-Sosa et al., 2004;Doherty and McCarthy, 2004;Naesje et al., 2004).Morphometric is the external measurements of an organism, while meristic counts means serial counts of body elements (Talwar and Jhingran, 1992).Morphological characters including meristic counts and body proportions often vary clinically (that is along a geographic gradient) (Lindsey, 1988).Ihssen et al. (1981), Allendorf (1988), Swaine et al. (1991) and Turan (1999) reported that phenotypic adaptations do not necessarily result in genetic changes in the population and thus, the detection of such phenotypic differences among populations cannot usually be taken as evidence of genetic differentiation.Studies of morphological character variation are, therefore, vital in order to elucidate patterns observed in phenotypic and genetic character variation among fish populations (Beheregaray and Levy, 2000).
Therefore, the aim of this study was to study the phenotype and genotype differentiation and to compare the amount of differences in phenotype based on morphometric character indices and meristic counts with the amount of differences in genotype based on RAPD fingerprinting between two Mugilidae, Flathead grey mullet, M. cephalus and Thinlip grey mullet, L. ramada.

MATERIALS AND METHODS
This study was carried out at fish breeding and production laboratory, Animal and Fish Production Department, Faculty of Agriculture (Saba-Bacha), Alexandria University and Nucleic Acids Research Department, Genetic Engineering and Biotechnology Research Institute (GEBRI), City for Scientific Research and Technology Applications, Alexandria, Egypt.

Specimen collection
Larvae of Flathead grey mullet, M. cephalus and Thinlip grey mullet, L. ramada were collected from Mediterranean sea and transferred to culture into earthen ponds at Faculty of agriculture (Saba-Bacha) fish farm for 12 month.From both sex, a total of 216 specimens of Flathead grey mullet, M. cephalus and 234 of Thinlip grey mullet, L. ramada were randomly collected.Their weight and length ranged from (32.10 to 50.00 g and 16.50 to 19.80 cm) and (28.00 to 45.50 g and 15.60 to 18.60 cm), respectively.

Quantitative phenotype analysis
A total of 34 morphometric characters and 9 meristic counts were recorded within each species as descriped by Bagenal (1978), Ezzat et al. (1979), Taniguchi et al. (1996), Sahu et al. (2000), Costa et al. (2003) and Doherty and McCarty (2004).Condition factor of each species were measured using the following equation: K= (W/L 3 ) 100 (Lagler, 1956), Where, W is the body weight (g) and L is the total length (cm).All morphometric characters were transformed by dividing the measurement by either total or standard length or head length of each fish to minimize the effect of fish size (Tables 1 and 3).
Phenotype differentiation between Flathead grey mullet, M. cephalus and Thinlip grey mullet, L. ramada based on morphometric character indices and meristic counts was analysed by means of the hierarchical cluster analysis of the STATISTICA package, version 5.00 (StatSoft, 1995).The cluster analysis using unweighted pair group average method (UPGMA: Sneath and Sokal, 1973) was performed on the matrix of euclidean distance in order to depict hierarchically the shape differences between Flathead grey mullet, M. cephalus and Thinlip grey mullet, L. ramada.

Statistical analysis
Data significance of the morphometric character indices and meristic counts were analyzed using unpaired Student ' s t-test (P<0.05) according to Snedecor and Cochran (1980).

Genotype analysis
Genotype analysis was performed based on RAPD fingerprinting.DNA was extracted from liver tissues of each Flathead grey mullet, M. cephalus and Thinlip grey mullet, L. ramada according to the method described by Bardakci and Skibinski (1994).In this study ten and twenty base long oligonucleiotide primers were used to initiate polymerase chain reaction (PCR) amplifications.Primers were randomly selected on the basis of GC content and annealing temperature for RAPD-PCR amplification (Table 2).
PCR amplifications were performed according to the procedure of Williams et al. (1990Williams et al. ( , 1993)).The reaction (25 µl) was carried out in a mixture consisting of 0.8 U of Taq DNA polymerase (Fanzyme), 25 pmol dNTPs, and 25 pmol of random primer, 2.5 µl 10X Taq DNA polymerase buffer and 40 ng of genomic DNA.The final reaction mixture was placed in a DNA thermal cycler (Eppendorf).The PCR programme included an initial denaturation step at 94°C for 2 min followed by 45 cycles with 94°C for 30 s for DNA denaturation, annealing as mentioned with each primer (Table 2), extension at 72°C for 30 s and final extension at 72°C for 10 min were carried out.Samples were cooled at 4°C.
The amplified DNA fragments were separated on 2.5% agarose gel and stained with ethidium bromide.ΦX174 DNA Ladder marker (bp 1335, 1078, 872, ………72) was used in this study.The amplified patterns were visualized on an UV transilluminator and photographed by gel documentation system.
RAPD patterns were analyzed and scored from photographs.For the analysis and comparison of the patterns, a set of distinct, wellseparated bands were selected.The genotypes were determined by recording the presence (1) or absence (0) in the RAPD profiles.Genetic similarity (GS) between Flathead grey mullet, M. cephalus and Thinlip grey mullet, L. ramada was calculated according to the formula given by Nei and Li (1979): Bij=2 Nij/(Ni + Nj), where Nij is the number of common bands observed in individuals i and j, and Ni and Nj are the total number of bands scored in individuals i and j respectively, with regard to all assay units.Thus, GS reflects the proportion of bands shared between two individuals and ranges from zero (no common bands) to one (all bands identical).Genetic dissimilarity (GD) was calculated as: GD = 1-GS (Bartfai et al., 2003).

Quantitative phenotype analysis
Mean values of morphometric character indices were compared between M. cephalus and L. r amada.The   3).Moreover, the results of meristic counts showed that the highest and lowest mean value of Pec FRC and AFRC were obtained by M. cephalus and differ significantly from that of L. ramada.While, the other meristic counts records of FDFSC, SDFRC, Pel FSC, Pel FRC, AFSC, CFRC and TNN, showed no significant differences between M. cephalus and L. ramada.In addition, M. cephalus had significant superiority of condition factor (K) compared with L. ramada (Table 3).The euclidean distance between M. cephalus and L. ramada using hierarchical cluster analysis of quantitative phenotype based on morphometric character indices, meristic counts and condition factor was 0.421.

Genotype analysis
All the ten different primers used in this study produced different RAPD band patterns (Table 4).The number of amplified bands detected varied, depending on the primers and species; in addition to ensure that the amplified DNA bands originated from genomic DNA and not from primer artifacts.Also, negative control was done for each primer/species combination.No amplification was detected in the control reactions.All amplification products were found to be reproducible when reactions were repeated using the same reaction conditions (Table 4 and Figure 1).The RAPD fingerprint was used for the detection of the genetic diversity between M. cephalus and L. ramada.The results showed highly genetic dissimilarity range (0.00 to 1.00) with an average of 0.437 using different random primers (Figure 1 and Table 4).

DISCUSSION
Morphometric characters and meristic counts have been the most widely used tool in racial studies in fish taxonomy (Anene, 1999).Meristic counts were much easier to evaluate and seem to be advantageous because most counts can be collected from live fish.However, meristic data alone may not provide the detail necessary to discern dissimilarities between different populations at the same species and sometimes at the same genus (North et al., 2002).According to Katselis et al. (2006), discriminate analysis in eight morphometric characters of the fry of four Grey mullets species in Western Greece showed that M. cephalus and Chelon labrosus were quite similar in morphology, while L. aurata and Liza saliens were rather different.The study also stated that the analysis of the morphometric variation of mullet fry could be used for their discrimination.
Furthermore, Akyol and Kinacigil (2001) found that the discriminate analysis in seven morphometric characters in adult specimens of Grey mullets showed that L. saliens and L. aurata were similar in forms, while M. cephalus and C. labrosus were rather different.Also, the results of morphological measurements on the L. saliens and L. ramada, obtained by Minos et al. (1994Minos et al. ( , 1995) ) supported these findings.The results of this study based on 34 morphometric character indices and 9 meristic counts are consistent with these findings as most of the morphometric character indices showed significant differences between M. cephalus and L. ramada, while most of the meristic counts showed insignificant differences between M. cephalus and L. ramada.
The main advantages of RAPD markers are the

Liza ramada
Mean±S.E.possibility of working with anonymous DNA and the relatively low expense, also fast and simple to produce RAPD marker (Hadrys et al., 1992;Elo et al., 1997;Ali et al., 2004).Moreover, RAPD analysis might be useful for systematic investigation at the level of species and subspecies (Bardakci and Skibinski, 1994), and more sensitive and technically easier to perform and produced results with low statistical error, whereas DNA fingerprinting detected greater genetic differentiation between Nile tilapia stains than other molecular techniques such as multilocus minisatellite marker (Naish et al., 1995).

Morphometric analysis
The results of this study are consistent with the findings reported by (Papasotiropoulos et al., 2001(Papasotiropoulos et al., , 2002(Papasotiropoulos et al., , 2007;;Blel et al., 2008;Semina et al., 2007), they applied several techniques for genetic analysis among Mugilidae family and reported that M. cephalus is the most distinct species compared with Liza genus.Also, the results confirmed that the amount of differences in genotype reflected the same amount of differences in phenotype between M. cephalus and L. ramada.In addition, these results stated that genetic analysis based on RAPD fingerprinting confirm the existing taxonomic system based on morphometric character indices and meristic counts and confirmed the results reported by Li et al. (1993) and Louis and Barlow (1987).
The major limitation of morphological characters at the intra-specific level is that phenotypic variation is not directly under genetic control but subjected to environmental modification (Clayton, 1981).Therefore, it can be use either phenotype analysis based on a large number of morphometric character indices and meristic counts or genotype analysis based on RAPD fingerprinting to discriminate M. cephalus and L. ramada with the same results.

Figure 1 .
Figure 1.Example of RAPD amplification products.Lane M: ΦX174 DNA marker, the lanes 1 and 2 of each primer are M. cephalus and L. ramada, respectively.

Table 1 :
Quantitative phenotype traits based on morphometric characters and meristic counts used for differentiation analysis between Mugil cephalus and Liza ramada.

Table 2 .
The sequences, GC% and the annealing temperatures of the primers used.

Table ( 3
): Means and standard error of quantitative phenotype traits based on morphometric character indices and meristic counts used for differentiation analysis between Mugil cephalus and Liza ramada.

Table 4 .
Total number of band, polymorphic bands and genetic dissimilarity between Mugil cephalus and Liza ramada using different random sequence of primers.