Abstract: Haruan, the local name for snakehead Channa striatus, indigenous to Malaysia, is a tropical and air breathing carnivorous fresh water fish in the family Channidae, which have important economics value as consumption’s fish and medicine value. In this study, a total of 120 specimens of Haruan Channa striatus were collected from the sampling stations in Johor, Kedah, Kelantan, Pahang, Perak and Terengganu. The result of physical-chemical parameters of the natural habitat were mainly Dissolved Oxygen (DO) within range from 2.62-33.70 mg L^-1; pH between 5.16-6.90; temperature within 29.43-31.86°C; turbidity of the water between -10.00 to 14.57 NTU; conductivity of 0.06-0.97 mS cm^-1 and salinity of 0-20 ppt. In the mixed populations, length weight relationships indicated the pattern of negative allometric type of growth (b<3). The result shows that the length of frequency distribution by populations from Johor, Kedah, Kelantan, Pahang, Perak and Terengganu were 25.97±2.88, 29.97±4.75, 31.52±3.32, 27.74±3.16, 30.34±2.87 and 29.53±6.67 cm TL, respectively. Variation in twenty four morphometric characteristics of Haruan C. striatus were examined by multivariate analysis. The discriminant analysis and the Unweighted Pair Group Method with Arithmeric averaging (UPGMA) cluster analysis based on Mahalanobis distance between group centroids showed that populations at Kelantan, Perak and Terengganu had quite similar morphological distant, while Johor, Kedah and Pahang were rather different. Such measurements are useful in the management of different cultures of Haruan C. striatus for Good Agriculture Practices (GAP) in farming technology, as well as the basic for Good Manufacturing Practices (GMP) production.

INTRODUCTION

Haruan, the local name for snakehead Channa striatus, indigenous to Malaysia, is a tropical and air breathing carnivorous fresh water fish in the family Channidae. There are altogether 30 species in the family reported in the world and eight of them were found in Malaysia including Toman C. micropeltes, Bujuk C. lucius and Haruan C. striatus. Members of the family are also found in all ASEAN countries namely Myanmar, Thailand, Laos, Cambodia, Vietnam, Brunei, Philippines, Indonesia and Singapore (Mat Jais, 1997).

Haruan in particular has the diploid chromosome number of C. striatus was 2n = 44 (Donsakul and Magtoon, 1991). The Haruan is a wild species found in small rivers, lakes, pools and shallow water body and in their natural habitats can survive in harsh environments with low dissolved oxygen and high ammonia (Marimuthu and Haniffa, 2007). Such Haruan C. striatus is known to exhibit high levels of cannibalism as juveniles, where the fish was associated with heterogeneous size variation, limited food availability, high population densities, limited refuge areas and light conditions (Qin and Fast, 1996). The fish for many decades had been known traditionally as remedy for wound healing. People in Malaysia and most other countries consumed Haruan C. striatus as source of protein to induce healing after giving birth and now becoming supplementary among caesarean mothers and illnesses such as cancer and diabetic gangrene (Mat Jais et al., 1994; Mat Jeis, 1997). Haruan C. striatus are in great demand in the local domestic markets and detailed knowledge on the population structure of Haruan C. striatus is needed for sound management and successful commercial fishing of this species.

Morphometric analysis has been very useful for separated rating species, populations and races in the past have been widely used for the identification of different fish stocks. Such morphometric study of fish populations are very important for understanding the interactive effect of environment, selection and heredity on the body shapes and sizes within a species (Saini et al., 2008). Meristic characters are the number of discrete, serially repeated, countable structures that are fixed in embryos or larvae. Studies of morphologic variation between populations continue to have an important role to play in stock identification, while stable difference in shape between groups of fish may reveal different growth, mortality or reproductive rates that are relevant for the definition of stocks (Turan et al., 2006). Length-Weight Relationships (LWR) provide basic information in fisheries biology, being useful to determine the weight of an individual fish of known length or total weight from length-frequency distribution and to compare specific growth among different regions. This relationship was initially used to obtain information on the growth condition of fish and to find out whether the somatic growth was isometric or allometric (Heydranejad, 2009).

Given the growing polar importance of Haruan C. striatus, detailed morphometric studies on this species are required and provide an important first step in what well surely become as specification and standard in a Good Agriculture Practices (GAP) for farming technology. Therefore, the present study aimed to determine the morphometric differences in six populations selected in Peninsular Malaysia. The objective of the present study was to find out the data and information of aspects, which include the relationship of length-weight relationships; size-frequency distribution, measurement morphometric and meristic study and water quality of different geographical populations in several populations in Peninsular Malaysia elucidate morphometric variations in this species to support and standard in Good Agriculture Practices for farming technology. Morphometric variations between stocks can provide a basis for stock structure and might be useful for studying short-term, environmentally induced variation, for example, in fisheries management (Begg et al., 1999).

MATERIALS AND METHODS

Sample collections: In this study, a total of 120 specimens of Haruan C. striatus (20.60-44.10 cm total body length) were collected from the sampling stations namely Sungai Petani, Kedah (N 5°38'33.33" and E 100°28'09.27"); Bagan Datoh, Perak (N 3°57'51.31" and E 100°45'13.51"); Kuantan, Pahang (N 3°49'48.45" and E 103°19'15.53"); Pontian, Johor (N 1°29'14.93" and E 103°23'57.85"); Kuala Terengganu, Terengganu (N 5°19'29.40" and E 103°08'27.29") and Tanah Merah, Kelantan (N 5°48'21.80" and E 102°09'17.17") in part of Peninsular Malaysia. These stations were selected base on geographical distances and previous studied (Ambak et al., 2006). The fish were transported to the laboratory for further analysis, excepted for Kedah and Terengganu were done on the spot at the stations.

In the laboratory, the fish were placed in an aquarium and the next day the fish were killed on the spot with ice (Nuove Technologie Del Freddo Ice Makers). After morphometric and meristic were done, the heart, liver, eggs (if available) and fillet were preserved in 95% ethanol (HmbG Chemicals, 2.5L) for further genetic marker analysis. The specimens were deposited at Faculty of Medicine and Health Sciences, University Putra Malaysia from July 2007 to January 2008. Distribution of fish into six groups in accordance with the total length of the body was done namely Johor 21.90-31.80 cm (20 specimens); Kedah 22.00-40.40 cm (20 specimens); Kelantan 24.50-37.70 cm (20 specimens); Pahang 23.20-35.50 cm (20 specimens); Perak 24.20-35.40 cm (20 specimens) and Terengganu 20.60-44.10 cm (20 specimens).

Water parameters: Physical-chemical measurements were analyzed at the lower source of the water body (between 0.3-0.5 m depth of below water surface) at stations including pH, conductivity (mS cm^-1), turbidity (NTU), dissolved oxygen (mg L^-1), temperature (°C) and salinity (ppt) using a Horiba U-10 Water Quality Checker. Physical-chemical variables measured of these stations are shown in Fig. 1.

Morphological identification: In total 5 meristic characters and 24 morphometric measurements were selected for comparisons. These are standard length, dorsal fin rays (D), Pectoral fin rays (P₁), Pelvic fin rays (P₂), anal fin rays (A), Caudal fin rays (C), total number in percentage of Standard Length (SL), Pre-Anal Length (PAL), Pre-Pelvic Length (PPelL), Pre-Pectoral Length (PPecL), Caudal Peduncle Length (CPL), Caudal Peduncle Depth (CPD), Body Depth (BD), Head Length (HL), Head Width (HW), Head Depth (HD), Dorsal-Fin Base Length (DFBL), Anal-Fin Length (AFL), Pelvic-Fin Length (PelFL), Pectoral-Fin Length (PecFL), Caudal-Fin Length (CFL), Caudal-Fin Height (CFH), Fork Length (FL), Total Length (TL), Pre-Dorsal Length (PDL), Dorsal Spine Length (DSL), Snout Length (SnL), Eye Diameter (ED), Pre-Orbital Length (POL) and Upper Jaw Length (UPL). All the characters were measured to the nearest centimeters using vernier caliper (American Scientific LLC, 150x0.02 mm) and weighted (Ohaus Ts400 Precision Standard Toploading Balance) to nearest gram. Standard length was expressed as percentages of the standard. Head measurements were expressed as percentages of the head length. All method of meristic counts and morphometric measurements are given in Fig. 2.

Fig. 1:	Map showing distribution of Peninsular Malaysia and the sampling stations (●) of Haruan Channa striatus used in this study namely Johor, Kedah, Perak, Kelantan, Pahang and Terengganu

Data and statistical analysis: In the present study, there were significant linear correlations between all morphometric characters and standard length of fish. All individual morphometric data were thus corrected, using the following equation (Elliott et al., 1995):

Fig. 2:

An outline drawing of Haruan Channa striatus showing the various body measurements studies. 1. Standard Length (SL). 2. Pre-anal Length (PAL). 3. Pre-pelvic Length (PPelL). 4. Pre-pectoral Length (PPecL). 5. Caudal Peduncle Length (CPL). 6. Caudal Peduncle Depth (CPD). 7. Body Depth (BD). 8. Head Length (HL). 9. Head Width (HW). 10. Head Depth (HD). 11. Dorsal-fin Base Length (DFBL). 12. Anal-fin Length (AFL). 13. Pelvic-fin Length (PelFL). 14. Pectoral-fin Length (PecFL). 15. Caudal-fin Length (CFL). 16. Caudal-fin Height (CFH). 17. Fork Length (FL). 18. Total Length (TL). 19. Pre-dorsal Length (PDL). 20. Dorsal Spine Length (DSL). 21. Snout Length (SnL). 22. Eye Diameter (ED). 23. Pre-orbital Length (POL). 24. Upper Jaw Length (UPL)

Working formula for each specimen for given morphometric character is:

Where:

The standardized data were rechecked by correlation against size to see whether the size dependence had been removed. The standardized data were used in all statistical analysis.

Standardized data were analyzed by univariate and multivariate methods. Differences among geographic sample were tested by one-way Analysis of Variance (ANOVA) followed by post hoc multiple comparison test of specific means using Tukey’s Honestly Significant Difference (HSD) multiple comparison test for unequal sample sizes (Sokal and Rohlf, 1994; Zar, 1998). The tests were considered significant at 0.05 probability level. Raw meristic counts with observable variation within samples were compared using Kruskal-Wallis nonparametric test considering the difference in the median as well as the distribution.

The length weight relationships were calculated from Total Length (TL) and Body Weights (BW) of 120 individuals were measured. To establish the length-weight relationship, the commonly used relationship was applied by Ricker (1975) and Quinn and Deriso (1999) was applied:

Where:

The parameters a and b were estimated by least squares linear regression on log-log data transformed data:

The coefficient of determination (r²) was used as an indicator of the quality of the linear regression (Scherrer, 1984). In addition, 95% confidence limits of the parameter b and the statistical significance level of r² were estimated. The value b = 3 indicates that the fish grows symmetrically or isometrically (provided its specific gravity remains constants). Values other than 3 indicate allometric growth: if b>3, the growth is called positive allometric and if b<3, it is called negative allometric.

Discriminant Function Analysis (DFA) were performed to identify the characters that were important in distinguishing population groups and to formulate classification functions for each location. Based on these functions, DFA reclassifies individual fish into groups, which are then compared with a prior group of that individual to get the percentage of classification success. Mahalanobis Distances (D²), as an indication of morphologic distance, were calculated between samples from DFA (Hair et al., 1995) to investigate the phenotypic relationship between populations a dendrogram was constructed based on Mahalanobis Distances (D²) using UPGMA cluster analysis of arithmetic averages. All analyses were conducted with the statistics packages NTSYS v.2.1 (Rohlf, 2000), XLSTAT (Addinsoft, XLSTAT Version 2009.4.05) and SPSS v.16.0.

RESULTS

Table 1 presents data on the water physical-chemical parameters of Haruan C. striatus habitats on six populations in Peninsular Malaysia. An analysis of variance (one-way ANOVA) showed that all physical-chemical parameters pH, conductivity (mS cm^-1), turbidity (NTU), temperature (°C), dissolved oxygen (mg L^-1) and salinity (ppt) were highly significant (p<0.05) between distributions. Water temperatures through the Johor ranged from about 29-30°C (Fig. 1). The average temperatures of Kedah and Kelantan were about 30-32.9°C with similar ranges. The averages temperatures of Pahang, Perak and Terengganu were about 28 and 33.9°C (Table 1), respectively. Differences were observed in the relative locations of Johor (M = 29.4, SD = 0.53), Kedah (M = 30.9, SD = 0.69), Kelantan (M = 31.9, SD = 0.50), Pahang (M = 30.9, SD = 0.69), Perak (M = 33.1, SD = 0.70) and Terengganu (M = 30.6, SD = 1.61); the differences were significant (F_5.36 = 14.336, p = 0.000). Post-hoc Tukey HSD tests was showed in Table 1. The average of conductivity ranged from 0.02-0.80 mS cm^-1; while there was a significant effect of conductivity by distribution at the p<0.05 level for the six stations. Turbidity readings, measured in Nephelometric Turbidity Units. The turbidity of Haruan C. striatus resources at Kelantan were significantly higher than those in other stations and averaged bout six times higher than other stations. The differences were significant (F_5.36 = 15.138, p = 0.000). In contrast, dissolved oxygen (mg L^-1), pH and salinity (ppt) range were 2.00-33.70 mg L^-1, 4.10-7.24, 0.00-0.20 ppt, respectively.

A total of 120 specimens of Haruan C. striatus, made up of 20 each from six station sites, were examined and analysed. The range and mean values of different morphometric characters during the study are given in Table 2. Mean standard length of Haruan C. striatus was 24.76±3.64 cm and the details weighted of Haruan C. striatus collected during the study are given in Table 2.

Table 1:	Physical-chemical parameters six stations of water resources in selected part of Peninsular Malaysia
*Highly significant at (p<0.05); NS = Not Significant at (p>0.05); Mean±SD Subjected to Tukey HSD Post-hoc test; Mean±SD in column with dissimilar superscript letter is significantly different (p = 0.05); Cond = Conductivity (mS cm-1); Turb = Turbidity (Nephelometric Turbidity Units); Temp = Temperature (°C); DO = Dissolved Oxygen (mg L-1); Sal = Salinity (ppt)

Table 2:	Analysis of Variance (One-way ANOVA) of twenty four morphometric characteristics (cm) showed four characteristics were significantly different (p<0.05) among distribution namely Johor, Kedah, Kelantan, Pahang, Perak and Terengganu in part of Peninsular Malaysia
*Highly significant at (p<0.05); NS = Not Significant at (p>0.05); Mean±SD Subjected to Tukey HSD Post-hoc test; Mean±SD in column with dissimilar superscript letter is significantly different (p = 0.05); PDL = Pre-Dorsal Length; DSL = Dorsal Spine Length; SnL = Snout Length; ED = Eye Diameter; POL = Pre-orbital Length; UPL = Upper Jaw Length; W = Weight

Table 3:	Meristic characteristics of Haruan C. striatus among six populations namely Johor,
*Highly significant at (p<0.05); NS = Not Significant at (p>0.05); D = Dorsal Fin Rays; P1 = Pectoral Fin Rays; P2 = Pelvic Fin Rays; A = Anal Fin Rays; C = Caudal Fin Rays

Analysis of Variance (one-way ANOVA) showed that out of twenty four morphometric, twenty one characteristics (SL = Standard Length, PAL = Pre-Anal Length, PpelL = Pre-pelvic Length, PpecL = Pre-pectoral Length, CPL = Caudal Peduncle Length, CPD = Caudal Peduncle Depth, HL = Head Length, HW = Head Width, HD = Head Depth, DFBL = Dorsal-Fin Base Length, AFL = Anal-Fin Length, PecFL = Pectoral-Fin Length, CFL = Caudal-Fin Length, CFH = Caudal-Fin Height, FL = Fork Length, TL=Total Length, DSL = Dorsal Spine Length, SnL = Snout Length, ED = Eye Diameter, POL = Pre-Orbital Length and UPL = Upper Jaw Length) were highly significant (p<0.05) between Johor, Kedah, Kelantan, Pahang, Perak and Terengganu (Table 2). The graph of the means boxplots from twenty four morphometric characteristic also show twenty one (A, B, C, D, E, F, H, I, J, K, L, N, O, P, Q, R, T, U, V, W, X) morphometric characteristic that were significant different for mean value of the morphometric characters between Johor, Kedah, Kelantan, Pahang, Perak and Terengganu (Table 2). The physical morphology, meristic counts were relatively homogenous among the specimens although slight and there is no significant variation in the distribution in the number of Dorsal fin rays (D), Pectoral fin rays (P₁), Pelvic fin rays (P₂), anal fin rays (A) and Caudal fin rays (C) (Table 3) (Kruskal-Wallis test, p<0.05).

Result from the analysis length-weight relationship of Haruan C. striatus, without differentiating sexes among six stations is to be known that the growth of Haruan C. striatus were negative allometric with value b = 0.047 whereas b<3 and statistically significant, p<0.05 (Fig. 3). The calculated equations for the length-weight relationship was W = 19.920L^0.047 or Log₁₀W = 1.285+0.047 Log₁₀ L (r² = 0.789).

If the analysis length-weight relationship were differentiated by stations the value b is between 0.039-0.065. Haruan C. striatus were catch in all 6 stations have growth polar negative allometric (b<3) and were shown in Table 4. The length-weight equations were calculated as for Johor, W = 19.309L^0.044 or Log₁₀W = 1.286+0.044 Log₁₀ L (r² = 0.490); Kedah, W = 17.355L^0.064 or Log₁₀W = 1.239+0.064 Log₁₀ L (r² = 0.785); Kelantan, W = 12.059L^0.043 or Log₁₀W = 1.323+0.043 Log₁₀ L (r² = 0.526); Pahang, W = 1.222L^0.065 or Log₁₀W = 1.222+0.065 Log₁₀ L (r² = 0.772); Perak, W = 1.237L^0.060 or Log₁₀W = 1.237+0.060 Log₁₀ L (r² = 0.953) and Terengganu W = 21.024L^0.039 or Log₁₀W = 1.323+0.039 Log₁₀ L (r² = 0.926) (Fig. 3), respectively. Figure 4 shows the length-frequency histogram of Haruan C. striatus. The result showed that the length of frequency distribution by populations from Johor, Kedah, Kelantan, Pahang, Perak and Terengganu were 25.97±2.88 cm TL, 29.97±4.75 cm TL, 31.52±3.32 cm TL, 27.74 ±3.16 cm TL, 30.34±2.87 cm TL and 29.53±6.67 cm TL, respectively.

Fig. 3:	Relationships between total length (cm) and weight (g) Haruan C. striatus, where a) is total of among six populations, b) Johor, c) Kedah, d) Kelantan, e) Pahang, f) Perak and g) Terengganu

Fig. 4:	Total length frequency distribution, where is a) Johor, b) Kedah, c) Kelantan, d) Pahang, e) Perak and f) Terengganu

Derived discriminant function in DFA (Addinsoft, XLSTAT Version 2009.4.05) using twenty four of the size-correction morphometric characters identified ED as significant contributors, which were found to be useful in population differentiation in order of importance (Wilk’s Lambda = 0.653, approx. F (5,114) = 12.121, p<0.0001), while BD had the least contribution (p = 0.215) (Table 5). As revealed by the respective eigenvectors, first function is weighted most positive heavily by CPD, HD, CFH, TL, PDL, DSL and ED, while second function showed a contrast between positive weighted by SL, PAL, PPelL, AFL and SnL. About 39.49% of all discriminatory power is explained by the first function and therefore it is clearly the most important discrimination function. As detected by the respective eigenvalues, the first two Canonical Variables (CV) were able to explain 69% of the total variance in the data. The first Canonical Variable (CV1) was accounted responsible 39.49% of the variation among-groups, the second (CV2) accounted for 29% (Table 6) and had eigenvalues of 2.132 and 1.568, respectively.

Table 4:	Relationship between total length (cm) and weight (g) six stations selected in the part of
N: Sample size, a and b: Parameters of the length-weight relationship: r2: Coefficient of determination

While 3-5 Canonical Variable accounted for 15, 13 and 4% of the residual variation, respectively and therefore were excluded from the remaining analysis. Morphological differences were greatest in function 1, where Johor, Kedah, Kelantan, Pahang, Perak and Terengganu, group centroid values were -2.698, 1.972, 0.816, -0.487, 0.190 and 0.206, respectively (Fig. 5). In function 2, group centroid values for where Johor, Kedah, Kelantan, Pahang, Perak and Terengganu were -0.875, -2.035, 1.300, -0.404, 1.280 and 0.787, respectively. The plot of first two canonical variables failed to show significant separation of samples with group means located close to each other especially Perak, Kelantan and Terengganu with 95% confidence ellipses for the six populations overlapping greatly (Fig. 6). However, Johor, Pahang and Kedah were clearly separated from others group, particularly from Perak, Kelantan and Terengganu.

Table 5:	Summary results of the discriminant function analysis of standardized morphological data

Table 6:	The eigenvectors for the first five canonical discriminant funtions, eigenvalue and cumulative proportion percent of the total variance accounted for and variability percentage from a DFA on measurement of 24 morphometric characters from six stations Haruan Channa striatus in Peninsular Malaysia

Fig. 5:	Scatterplot of the centroid values (mean of discriminate functions) separating into their groups

Fig. 6:	Ninety five percent confidence ellipses of CV scores of morphometric characters, the first Canonical Variables (CV1) accounts for 39.49% and second (CV2) accounts for 29.06% of the between group variability

When comparing Mahalanobis Distance (D²), between each pair of samples on the basis of twenty four morphometric variables (Table 7), Terengganu was fairly similar to those from Perak and Kelantan.

In contrast, Pahang was also clearly distant from Kelantan, Perak and Terengganu. Johor and Kedah were also separated from the others populations. These relationships among the populations were more clearly shown by cluster analysis (Fig. 7).

Fig. 7:	Cluster diagrams using the UPGMA algorithm on the matrix of Mahalanobis distance between populations

Table 7:	Squared Mahalanobis distance between pairs of populations (below diagonal) and the probability value (above diagonal) (asterisk indicate tests that are not significant at the level α = 0.003 with sequential Bonferroni adjustment)

Table 8:	Proportion of Haruan C. striatus correctly classified into groups by classification functions based on a prior knowledge of their original locations

The derived classification functions could correctly classify only an average of 83.33% into their respective a priori groups (Table 8). While the best classification success was observed for Haruan C. striatus from Kedah (95%), the lowest success was Haruan C. striatus from Pahang (65%) whereas those for Johor, Kelantan, Perak and Terengganu were 90, 85, 80 and 85%, respectively.

DISCUSSION

In this study, physical-chemical measurements were obtained during the catch of Haruan C. striatus on the site among six stations. However, Haruan C. striatus was caught first before water measurements were taken. These were done to confirm Haruan C. striatus populations and one of typical habitats of Haruan C. striatus found in Malaysia. Base on the result, there were significant different physical-chemical parameters in environmental Haruan C. striatus. The ranges of physical-chemical parameters sample showed pH of 5.10-7.24, while pH levels in Perak were higher than other distribution. According to Lee and Ng (1994) and Ali (1990), Haruan C. striatus are commonly found in highly adaptability water pH >5, ranged at pH values of 4.25-9.40 and 5.0-9.4 in year 1986, respectively. However, conductivity of 0.02-0.80 mS cm^-1 and also have significant differences p<0.05 (Tukey’s HSD post-hoc) among the six distributions. According to the previous study (Mat Jais, 1997) and (Lee and Ng, 1994), found and recorded, conductivity range 0.07-0.30 mS cm^-1 and 20.8-20.9 μS cm^-1, respectively. The turbidity measured on all the samples were <1 Nephelometric Turbidity Unit (NTU) except for distribution of Perak, -10.00 NTU. Turbidity range from -10.00-82.00 NTU, Dissolved Oxygen of 2.00-33.70 mg/L^-1, there was significant differences HSD post-hoc p<0.05 both turbidity and Dissolved Oxygen among the six distributions. However, this study was recorded at temperature ranges between 28-33.90°C, while previous study done by Mat Jais (1997) was recorded at 25-30°C. According to Lee and Ng (1994) and Haruan C. striatus able to tolerate temperatures of 11-40°C, while (Ali, 1990) reported 22.8-40.9°C and able to survive in slightly brackish water. There was significant differences <0.05 (Tukey’s HSD post-hoc) among the six distributions. Salinity was 0.0% in all of distributions.

According to Table 1, Haruan C. striatus were one of the fish that can withstand living in low water condition, low oxygen level and with pH level were slightly acidic. A variation in the environmental such as temperature-salinity were probability effect on survival, growth, morphometry beside other environmental factors and were important source in manipulated to determine relationship physiological in fish. Temperature usually has a greater effect on fish growth than salinity (Rombough, 1996) and can affect virtually all aspects of fish reproduction (Van der Kraak and Pankhurst, 1996) and larval development including hatching size, efficiency of yolk utilization, growth, feeding rate, time to metamorphosis, digestion rates and metabolic demand (Blaxter, 1969; Rombough, 1996). Historically, the morphology of fishes has been the primary source of information for taxonomic and evolutionary studies. There are numerous characters available for morphological study. These characters were most commonly divided into categories; morphometric and meristic, where it is also have long been used to stocks identifications (Turan et al., 2004). The differences in the morphological and meristic characters of specimen were supposed to be in association with environmental ecosystems from which they originated (Cakic et al., 2002; Franicevic et al., 2005).

There were revealed significant differences in morphometrics between six populations of Haruan C. striatus in Peninsular Malaysia. However, all six populations had similar average lengths and this will be affecting allometric polar growth of Haruan C. striatus in Peninsular Malaysia. The average size of both the populations was very similar and probably the average physical-chemical parameters were significantly similar in Peninsular Malaysia beside other environmental factors.

There was a clear morphological distinction between certain characters in both populations. It is often difficult to explain the causes of morphological differences between populations (Cadrin, 2000). These differences may be genetically related differences or they might be associated with phenotypic plasticity in response to different environmental factors in each area (Murta, 2000) and also could be expected for distant sampling localities. Significant differences between six populations showed that out of twenty four morphometric, twenty one characteristics (SL = Standard Length, PAL = Pre-Anal Length, PpelL = Pre-Pelvic Length, PpecL = Pre-Pectoral Length, CPL = Caudal Peduncle Length, CPD = Caudal Peduncle Depth, HL = Head Length, HW = Head Width, HD = Head Depth, DFBL = Dorsal-Fin Base Length, AFL = Anal-Fin Length, PecFL = Pectoral-Fin Length, CFL = Caudal-Fin Length, CFH = Caudal-Fin Height, FL = Fork Length, TL = Total Length, DSL = Dorsal Spine Length, SnL = Snout Length, ED = Eye Diameter, POL = Pre-orbital Length and UPL = Upper Jaw Length) were highly significant (p<0.05) between Johor, Kedah, Kelantan, Pahang, Perak and Terengganu (Table 2). Thus, morphological variation can reflect genetic differences between stock or environmental differences between localities.

The growth polar among Haruan C. striatus populations selected in Peninsular Malaysia was negative allometric type of growth where the value of b values is 0.047 and there were no significant between populations. These was against (Ali, 1990) where Haruan C. striatus at Perak was the value of b>3.092 in order of positive allometric type of growth in year 1985 and 1987. The value of b-value must be equal to 3 if fishes have to maintain their shape as they grow, but there is no theory that says in which case the b-value can be expected to be negatively or positively allometric. The biological interpretation of the numerical values of the parameters a and b is not straight forward, except that when growth is isometric, a can be interpreted as a condition factor. When growth is allometric, the role of a as the condition factor is questionable (Khaironizam and Rashid, 2002). The declines in growth differences may be condition indicate low availability or poorer quality of food and pressure stress. Analysis of Length Weight relationships of fish species is often to determine the effect of different treatments some population, where biomass often estimated measurement can be time-consuming in the fish population (Martin-Smith, 1996). In addition, the data on length and weight can also provide important clues to climatic and environmental changes and change in human subsistence practices (Samat et al., 2008). Length and weight measurement in conjunction with age data can give information on the stock composition, age at maturity, life span, mortality, growth and production (Fafioye and Oluajo, 2005).

The size structure among populations Haruan C. striatus in the investigated areas consists of relatively higher percentage in Perak than other populations selected. Perak attained a greater size indicating a size dimorphism. According to size structure obtained, there was a possibility based on measurement of size showed that Haruan C. striatus can still grow. These have been proved with (Froose and Pauly, 2009) where the measurement of Haruan C. striatus that have been recorded was 100 cm. Size structure (frequency) of any fish is important to know the status of the population structure of that fish in nature including traits as the river health, stock condition and breeding period of the fish (Nurul Amin et al., 2005; Ranjan et al., 2005).

This study was to investigate morphological variation among six stations of Haruan C. striatus populations from various geographic in Peninsular Malaysia. Although, the majority of the morphometric variations in Haruan C. striatus were between-individuals-within-regions, at least some of the variation related to size had a geographic basis (Eastham et al., 2002). Results from both univariate and multivariate analysis provide congruent evidence for the existence of significant morphological variation in Haruan C. striatus among selected population’s localities of Peninsular Malaysia. Derived classification functions provided moderate percentage of classification success (ranging from 65-95%), indicating that morphological characters used in these analysis provide discriminatory power for the Haruan C. striatus populations studied. Therefore, morphological characters measurement is a helpful tool for discrimination of fish population. Morphometrics measurements combining with imagine analysis is steps ahead to produce a better understanding fish stock structures (Pollar et al., 2007). Among five eigenvalues in discriminating the populations, the first canonical variables is clearly most important variation among region groups, where the other function is comparatively much smaller. The other functions were excluded from the remaining analysis except for first two functions. The first two canonical variables accounted 68.5% of total observed variance among the twenty four morphometrics variables; only positive weighted were selected. Discriminant analysis was successful in grouping population (Fig. 6). The standardized coefficients (loadings) for the functions (Table 6) reflect the correlation between the original variable and the function can be thought of as the proportion of the variance that is shared (Hair et al., 1995). The first Canonical Variables (CV1) were dominated by positive weighted (large loading) for CPD, HD, CFH, TL, PDL, DSL and ED, the SL, PAL, PPelL, AFL and SnL for canonical variables. The rest of the morphometrics also contributed significant to variation but were not good enough identified of individual populations. These findings showed ED were significant identified contributor that were found mostly to be use useful in population differentiation from other morphometric characters, in order of important (Wilk’s Lambda = 0.653, approx. F(12.121) = 3.115, p<0.0001). Such differences between the populations may be related to different habitats characteristics, such as temperature, turbidity, food availability, water depth and flow. These findings also showed there are none of the meristic characters of significant variation in Haruan C. striatus.

The canonical variates analysis and cluster among six populations, although there are the occurrence of some overlapped between individuals of some populations, it is clear that these are separated from each other and that populations located in the same geographical regions are grouped together. Populations from Kelantan, Pahang and Perak are close together to each other and are grouped together.

Although, there are overlapping among individuals Johor, Kedah and Terengganu but it is clearly separated groups from Kelantan, Pahang and Perak. The detected differentiations for these groupings are probably the geographical proximities and also similar climatic conditions of locations. The results of cluster analysis based on Mahalanobis Distance (D²) again confirms this separation between geographic populations which are located a cline from Johor, Kedah and Terengganu were separated from other populations in Peninsular Malaysia. The previous study genetic variation using Random Amplified Polymorphic DNA also confirms that several populations on Haruan C. striatus in Peninsular Malaysia (Ambak et al., 2006). However, these phenotype morphometric will be compared and carries on to the further studies using genetic, Polymerase Chain Reaction-Restriction Fragment Length Polymorphism (PCR-RFLP) analysis mitochondrial DNA cyto b gene.

CONCLUSION

The results indicate that significant heterogeneity in morphology exist among some of the studied populations of Haruan C. striatus. Observed morphological variation probability occurred due to factors such as partial isolation of the populations, environmental and local adaptations. The length-weight relationships, size-frequency distribution compositions, morphometric and meristic result distinguishing among populations in Peninsular Malaysia that probability are useful in fisheries management, as well as standardization for a Good Agriculture Practices (GAP) in farming technology, as well as the basis for Good Manufacturing Practices (GMP) production (Begg et al., 1999). This result will be verified and associated in our future genetic study.

ACKNOWLEDGEMENTS

This research was supported by Biosatellite Program directly under National Biotechnology Centre, MOSTI in collaboration with UPEN Pahang BSP©/BTK/006 and University Putra Malaysia for facilities.