Morphological characterization of the progenies of pure and reciprocal crosses of Pangasianodon hypophthalmus (Sauvage, 1878) and Clarias gariepinus (Burchell, 1822)

Twenty-five traditional and thirty-four geometric morphometric comparisons were carried out on pure and reciprocal crosses of Pangasianodon hypophthalmus (Sauvage, 1878) and Clarias gariepinus (Burchell, 1822). Thirty fish samples each of the C. gariepinus (CH), P. hypophthalmus (PH), Pangapinus (♀PH × ♂CG) and the two distinct morphotypes of the Clariothalmus (♀CG × ♂PH) (Clarias-like and Panga-like) between the ages of four and six months were used for this study. Phenotypically, the Clarias-like Clariothalmus and the Pangapinus progenies were indistinguishable from their maternal parents while the Panga-like Clariothalmus was a phenotypic intermediary of the putative parents but looks more closely to the paternal parent. Hence, both univariate proportion and multivariate analysis of the collected data successfully separated the different fishes into three multivariate spaces. The analysis of the dendrogram with complete linkage and Euclidean distance further showed the close relationship of the isolated Panga-like Clariothalmus progenies to the paternal parent, however, Clarias-like Clariothalmus and the Pangapinus were completely intermingled with their maternal parents. The most important index of discrimination of these fishes into different multivariate spaces was the fin characteristic which showed 100% exclusive ranges for the individual groups in many cases.

, dorsal attachment of the caudal fin to the tail 4 , ventral attachment of the caudal fin to the tail 5 , posterior end of the anal fin 6 , origin of the anal fin 7 , origin of the pelvic fin 8 , origin of the pectoral fin 9  and ambiguous [6][7][8] . Geometric morphometric on the other hand is a landmark-based technique and considered the most rigorous morphometric technique ever [9][10][11] . It is capable of processing morphometric data from digital images with landmark points quickly and with high precision 9,12,13 . The integration of geometric morphometric data with other analytic tools such as biochemical, geographical, molecular and morphological parameters could better describe phylogenetic relationships among fishes and shed light to many ambiguous taxonomic ranks [14][15][16] .
Recently, we produced novel hybrid progenies from intergeneric crosses of Asian catfish Pangasianodon hypophthalmus (S.) and African catfish Clarias gariepinus (B.) 17,18 . Beyond the production of a novel aquaculture candidate, we also anticipated providing solutions to some of the breeding problems associated with the production of the pure crosses by the hybridization between the two species. For instance, the killing of male C. gariepinus to obtain testis is perfectly complimented and could be avoided due to the ease of sperm stripping from the male P. hypophthalmus. Also, the early maturity (9months) and high fecundity of the female C. gariepinus eliminate seasonal production characteristics associated with P. hypophthalmus brood fish due to the difficulty in obtaining gravid female (late maturity period of about 3 years). More so, early report of the hybrid ♀C. gariepinus × ♂P. hypophthalmus has demonstrated superior growth performance and heterosis over both pure parents 17 . In view of the performance characteristics of the hybrids and popularity of the pure crosses, it is important to urgently provide a quick and rapid identification tools for the hybrid since data on genetic discrimination is not available. In this study, morphological data were collected using traditional and geometric measurement hence, overcoming the drawbacks inherent in the use of traditional multivariate techniques alone. It is believed that the combination of both approaches would allow for accurate characterization of the novel hybrids between the African and Asian catfishes.

SN
Landmark points Character description Origin of the pelvic fin to the origin of the pectoral fin.

Materials and Methods
Progenies of C. gariepinus (CH), P. hypophthalmus (PH), and the reciprocal crosses Pangapinus (♀PH × ♂CG) and Clariothalmus (♀CG × ♂PH) were obtained from similar breeding history using the method described by Okomoda et al. 17,18 . In brief, six sexually mature P. hypophthalmus and C. gariepinus (between 1-2.5 kg) were injected with Ovaprim ® hormone at a dosage of 0.5 ml/kg. After stripping of the females, the pooled eggs of the different species were divided separately into two portions. One portion was used for the production of pure progenies (♀CG × ♂CG and ♀PH × ♂PH); while the other portion was used for the reciprocal crosses Clariothalmus (♀CG × ♂PH) and Pangapinus (♀PH × ♂CG). The progenies obtained were cultured for at least four months (4-6 months) at the School of Fisheries and Aquaculture Sciences hatchery of the Universiti Malaysia Terengganu, Malaysia before morphological analysis was done. Thirty 19 fish samples each of the progenies of the pure C. gariepinus, P. hypophthalmus, Pangapinus (Panga-like) and the two observed morphotypes of the Clariothalmus (Clarias-like and Panga-like) were used for morphological characterization in this study. The experimental protocols for this study were approved by the Universiti Malaysia Terengganu committee on research. All methods used in this study involving the care and use of animals were in accordance with international, national, and institutional guidelines. Twenty-five conventional traditional morphometric data were collected from each fish (some of which are described in Fig. 1).
These includes total length (TL), standard length (SL), dorsal fin length (DFL), dorsal fin height (DFH), distances between dorsal fin end and adipose fin origin (DBDAF), predorsal distance (PDD), pelvic fin length (PFL), pelvic      Body related measurements were expressed as percentages of standard length, while head related parameters were expressed as percentages of head length. Descriptive statistics of data were done and further subjected to analysis of variance using Minitab 14 software. The percentages and exclusive ranges of the data from the fins characters were also determined. This was done by first sorting data in ascending order using the Microsoft Excel software and the exclusive ranges of a paired combination of the different groups were determined. Trust network method was also used to measure morphological traits. Ten landmark point (Fig. 2) were identified namely Snout 1 , origin of dorsal fin 2 , posterior end of the dorsal fin 3 , dorsal attachment of the caudal fin to the tail 4 , ventral attachment of the caudal fin to the tail 5 , posterior end of the anal fin 6 , origin of the anal fin 7 , origin of the pelvic fin 8 , origin of the pectoral fin 9 and the posterior point of the eye 10 . Thirty-six distances between the different landmark points were recorded as shown in Table 1. Values from the landmark distance measured were expressed as percentages of standard length.  To ensure that variations in this study were only attributed to body shape differences, and not to the relative sizes of the fish, size effects from the data set were eliminated, by standardizing the morphometric parameters (from traditional and trust network measurement) using the allometric formula given by Elliott et al. 20 : where M is the original measurement, M adj is the size-adjusted measurement, Lo is the TL of the fish, and Ls is the overall mean of the TL for all fish from all samples. Parameter b was estimated for each character from the observed data as the slope of the regression of log M on log Lo, using all fish in all groups. The data collected were subjected to Principal component analysis (PCA) using PAST free software to obtain sample centroids graph and then determine the morphological character that contributes most to the separation of the fishes into distinct groups. Dendrograms with complete linkage and Euclidean distances of the fishes were also determined.

Result and Discussion
Despite the ease of phenotypic discrimination between two species involved in interspecific or intergeneric crosses, the use of morphological characterization could pose a number of challenges because of inappropriate techniques and the occurrence of growth allometry 16,21 . Hence, discrimination depending solely on size influenced morphometric traits becomes difficult or largely inaccurate. It is, however, important to eliminate size effect or growth-related shape changes and then elucidate shape differences among the different fish groups. The morphometric parameters expressed as a proportion of standard/head length revealed three distinct multivariate spaces for the five fish groups in this study. The Clarias-like Clariothalmus and the Pangapinus progenies        the shape of the mouth was more like those of L. rohita. Also, the hybrid ♀C. gariepinus × ♂Clarias batrachus (L.) was phenotypically similar to C. batrachus 25 . The reciprocal hybrid between Pangasius djambal (B.) and P. hypophthalmus showed intermediate phenotypic characteristics but had a strong similarity with the latter than the former 26 . Legendre et al. 27 and Akinwande et al. 2 , had earlier concluded that the display of intermediate phenotypic characteristics in the ♀C. gariepinus × ♂Heterobranchus longifilis (V.) hybrid was an indication of true hybridization. Hence, it is thought that they resulted from the fusion of the genetic material of both parents. They further hypothesized that phenotypical inheritance in true hybrid has paternal dominance. In the study, the Panga-like Clariothalmus progeny had an adipose fin and a forked tail just like the paternal parent. However, it has an elongated dorsal fin just like those in the maternal parent (but shorter than), hence may be the only true hybrid of the reciprocal crosses. According to Chevassus 28 , fertilization of a fish egg by a heterospecific sperm may lead to the production of haploid, gynogenesis/androgenesis, hybrid diploid, hybrid triploid, hybrid tetraploid, a combination of some or all the mentioned or death. Progenies of many distance hybridization are often composed of different phenotypic characters 29 which is suggestive of significant differences in genetic composition and inheritance pattern 17 . Therefore, the largely indistinguishable morphology of Purebred C. gariepinus and P. hypophthalmus with the Clarias-like Clariothalmus and Pangapinus progenies respectively may be a pointer to the presence of gynogenetic individuals. Environmental, geographical, and genetic adaptation has been implicated as possible causes of phenotypic variations between strains in many previous studies 19,30 , observations of the present study, however, is largely linked to differences in species and pattern of genetic inheritance in the different morphotypes of the reciprocal hybrids.      The morphometric variability among the three groups in this study was mainly due to the variation of characters related to fins, and body characteristic. This is because the effect of size was successfully eliminated by the allometric transformation and demonstrated by univariate proportion and multivariate analysis. However, with 100% exclusive ranges observed for the different groups for most fin characters (Tables 7-11), this is likely the easiest and most rapid index of discrimination of the three groups applicable in the field. Haddon and Willis 31 stated that morphometrics of the head and body depth have been regarded as the most important characters for discrimination of Devil anglerfish Lophius vomerinus (V.), Pacific herring Clupea pallasi pallasi (V.) and Orange roughy Hoplostethus atlanticus (C.) 31 . Solomon et al. 19 , however, reported head length, body depth at anus and the eye diameter as the most influential morphometric parameters used to discriminate fish strains from cultured and wild Clarias gariepinus. However, the suitability of fin and body characteristics in the study as opposed to reports from the previous study may be due to the clear and unambiguous intergeneric phenotypic differences of the pure crosses used. Based on the results of the present study, the fin characteristic appears to be the most promising index of morphological discrimination. Despite the fact that morphological approach alone is insufficient to investigate hybridization status of fishes, this research has provided useful assumption on the nature of the hybrid progenies and a quick/cheap identification tools for field application. However, a combination of both morphological, genetic and cytogenetic data in future studies could provide a clearer understanding of the nature of the progenies gotten. Table 11. Comparison between the progenies of Clariothalmus and Pangapinus for the exclusive ranges of fin measurement expressed as a percentage of standard length (n = 30). Minimum-maximum exclusive range observed (percentage). Keys: DFL = Dorsal fin length; DFH = Dorsal fin height; PFL = Pelvic fin length; PFH = Pelvic fin height; PeFL = Pectoral fin length; PeFH = Pectoral fin height; AFL = Anal fin length; AFH = Anal fin height; CFL = Caudal fin length; CFH = Caudal fin height.