Combining ability and heterosis in egusi-melon Citrullus lanatus (Thumb.) Matsum and Nakai

This study was carried out to evaluate the genetic potential of seven egusi-melon genotypes for seed weight and its components following the diallel mating scheme. Seven parents were selected from among 28 genotypes obtained from indigenous farmers in five ecological zones (forest, derived savanna, southern, northern and Sudan savanna) of Nigeria. The experiment was laid out in a completely randomized block design with four replications. Specific combining ability variance was of greater importance than general combining ability variance in the inheritance of all the traits under study. Okwe-2-IM x KTGO-GMB was the best combiner for number of pods and pod weight per plant while KOGI-NITN x MOKWA-NG was the best combiner seed weight per plant. Positive and highly significant (P<0.01) heterosis was recorded among crosses for number of primary vines, number of pods and seed production efficiency per plant indicating their importance in hybrid development for improved yield. Running Title: Combining ability and heterosis in egusi-melon.


INTRODUCTION
Egusi-melon is a member of the Cucurbitaceae family (Schippers, 2000). It is a variety of melon seed which is popularly called "egusi" in West Africa (Akpambang et al., 2008). Egusi-melon is a multipurpose crop that is mainly cultivated for its seeds. The seed of egusi-melon is nutritious, and contains up to 41.6% crude protein and 57.3% total lipid which are exploitable (Fakou et al., 2004). The crop is an important income earner in the food economics of most farmers especially women (DSC, 2006). Egusi-melon seeds are obtained either in shelled or unshelled forms in West African markets and used greatly in West African cookery (Akpambang et al., 2008). The crop is useful as live mulch for weed control and soil moisture conservation at high populations (Ikeorgu et al., 1989). It provides excellent ground cover (Lal, 1995), reduces surface runoff and controls erosion (Willy and Chinaka, 1985). In Nigeria limited attempt has been made for the genetic improvement of the crop. This may be due to the scanty information on the nature of gene action controlling the expression of yield and yield contributing traits in the crop. The heterozygous nature consequent to the outcrossing breeding system of egusi-melon may open the scope of development of openpollinated as well as hybrid varieties with better seed yield and quality. In a breeding program, it is important to know the combining abilities of parents to be constituted into hybrids. Combing ability and heterosis are the most important genetic parameters for breeding improved varieties. The variance of general combining ability (GCA) and specific combining ability (SCA) are related to the type of gene action. Information on the relative importance of the additive (GCA) and non-additive (SCA) gene action within a breeding population is important because it is used to determine as to which breeding procedure will effectively improve the performance of the trait of interest (Dudley and Moll, 1969). It is also necessary to have detailed information about the desirable parental combinations in any breeding program that can involve a high degree of heterotic response. Phenotypic selection of parents for hybridization based on their performance per se alone may not always be a viable procedure, since phenotypically superior genotypes may yield inferior hybrids and or poor recombinants in the segregating generations. It is essential that parents are selected on the basis of their genetic worth. Therefore, selection of potential parents based on genetic information and knowledge of their combining ability is important. Combining ability predicts performance of parents in cross combinations and thus helps in identifying the desirable parents as well as the crosses for crop improvement program. It also helps in getting idea about the nature of gene action for a particular trait (Nazzin Uddin et al., 2009). Hybrids due to their high yield potential are used for increasing the productivity of crops. Shankara et al., (2005) reported that hybrid tomato usually produces higher yield. Hybrid plants are usually heavy producers, and combine the character of the parent plants. Heterosis can be expressed when the parents of a hybrid have different allele at a locus and there is some level of dominance or epistasis among the alleles (Falconer and Mackay, 1996). Heterotic studies can provide the basis for the exploitation of valuable hybrid combinations in future breeding programs. Heterosis has been utilized in some crops, including cucurbits, to exploit dominance variance through the production of hybrids (Cramer and Wehner, 1999). Heterosis in a trait varies from population to population of the same crop species. Ghanderi and Lower (1979c) reported high-parent heterosis for fruit yield in cucumber while Cramer and Wehner (1999) did not observe heterosis and inbreeding depression for fruit yield in inbreeds parents of the same crop.
The objectives of this study were to use diallel analysis according to Griffing's (1956) Method 1, Model 11 to determine general and specific combining abilities of seven egusi-melon genotypes as well as gene action for six yield components and the heterotic performance of the F 1 hybrid progeny.

MATERIALS AND METHODS
Seven genotypes of egusi-melon taken from breeding program at the Federal College of Agriculture, Ishiagu, Ebonyi State, Nigeria, namely, AKARA-AB., OKWE-2-IM., KOGI-NITN., MOKWA-NG., MADLS-NG., KTGO-GMB and ABTK-ADP. were used to make a full diallel cross to obtain forty-nine treatments (7 parents, 21 F1s and 21reciprocals). The experimental design was Randomized Complete Block Design with four replications. Three seeds of each genotype were planted at a spacing of 1.0 x 1.0 m which was later thinned down to one seedling at two weeks after planting. Each plot contained twenty plants per genotype. Data were collected from all plants for number of primary vines, leading vine length, number of pods, seed production efficiency and seed weight. Seed production efficiency is the proportion of dry seed weight to fresh fruit weight expressed as percentage and has been found to range between 2.0 -2.75 percent in Nigeria (Philips, 1964). Analysis of combining ability was carried out according to Griffing's (1956) Model 1 and Method 2 following Sing and Chaudhary (1976). The magnitude of heterosis was estimated in relation to mid-parent (Mp) and better-parent (Bp) values using the method of Hayes et al., (1956). Test of significance was done as described by Hayes et al., (1956).

Analysis of variance
The analysis of variance showed that mean square due to diallel progenies was highly significantly different (P<0.01) for all the traits under study except for leading vine length per plant (P<0.05) indicating the existence of genetic variability among the egusi-melon diallel progenies for most of the traits under study ( Table 1). The decomposition of the mean square due to diallel progenies for the traits under study into its components (Table 1) showed that parents were highly significantly different (P<0.01) for number of pod per plant, pod weight per plant and seed weight per plant. Significant differences (P<0.01) were observed among F 1 s progenies for all the traits under study except for seed production efficiency per plant (P>0.05). Reciprocal progenies (rF 1 s) were highly significantly different (P<0.01) from each other for all the traits (Table  1). Furthermore, the difference between F 1 progenies and their reciprocals (F 1 vs rF 1 ) were highly significantly different (P<0.01) for two traits, pod weight per plant and seed weight per plant which indicate considerable reciprocal differences and suggesting the presence of maternal effects in the control of the expression of these traits. The difference between parents and crosses (Parents vs crosses) was highly significant (P<0.01) for number of primary vines per plant, number of pods per plants, pod weight per plant and seed production efficiency indicating heterosis in resultant hybrid progenies ( Table 1).  Analysis of variance for combining ability for the traits under study is presented in Table 2. The mean squares for general combining ability (GCA) for all the traits under consideration were non-significant (P>0.05) indicating that variability was not present among the parents for GCA for the six traits under study. On the other hand, mean squares for SCA were highly significant (P<0.01) for number of pods plant -1 , pod weight plant -1 and seed weight plant -1 . This suggests the importance of non-additive gene effects in the expression of these traits. Mean squares for reciprocal effects were significant (P<0.01) for number of primary vines plant -1 , pod weight plant -1 and seed weight plant -1 (Table 2) indicating the importance of reciprocal effects in the expression of these traits.

Analysis of variance for combining ability
The variance components given in Table 2 were estimated to determine the importance of additive and dominance components in the control of characters under study. The variance (δ 2 s ) due to SCA was higher in magnitude than variance (δ 2 g ) due to GCA for all the traits under study. Further, the values of GCA variance (δ 2 g ) to SCA variance (δ 2 s ) ratio which is less than 1.0 for all the traits support the predominance of non-additive gene effects in governing the expression of these traits. These results are encouraging from the point of view of heterosis exploitation. Ghaderi and Lower (1979) obtained similar results for fruit weight and Dogra and Kanwar (2011) for number of fruits and fruit weight in cucumber. The general combining ability effect represents the additive nature of gene action and it is used to designate the performance of the parents in general combination. In the present study, GCA mean squares was non-significant (P>0.05) for all the characters under study indicating lack of variability in the general combining ability effects of the parents. As a result it was not necessary to estimate the general combining ability effect of the parents in the diallel crosses.

Specific combining ability effect
Specific combining ability effect signifies the predominance of non-additive gene action in the expression of traits. It denotes the highly specific combining ability leading to good performance of some specific cross combinations. The usefulness of a particular cross in exploiting heterosis is judged by the SCA effect of component parents. Table 3 revealed that OKWE -2 -IM x KTGO-GMB was the best cross combination for two traits: number of pods plant -1 and pod weight plant -1 while KOGI-NITN x MOKWA-NG was the best cross combination for seed weight plant -1 . Similar results showed that, MOKWA-NG x MADLS-NG was the best cross combination to increase seed production efficiency per plant -1 on account of being associated with positive and significant SCA effect.

Reciprocal effects
Reciprocal effects are important because they could detect a desirable female seed parent in hybridization programme, particularly for producing commercial F 1 hybrids (Sharma, 1988). Table 3 indicated that OKWE-2-IM will produce progenies with better performance for pod weight plant -1 and seed weight plant -1 when deployed as a female parent than as a pollen source. Bahari et al., (2012) obtained similar results for fruit yield and fruit number in cucumber. Pavasia et al. (1999) recommended that crosses portraying reciprocal effects should not be mixed with direct crosses.

Heterosis
Significant difference between parents and crosses which is an indication of heterosis was observed only for number of primary vines plant -1 , number of pods plant -1 and percentage seed production efficiency plant -1 . Mid-parent and better-parent heterosis estimates for all the characters under study are presented in Table 4 and Table 5 for direct f 1 and reciprocal cross combinations, respectively. Generally, mid-parent heterosis estimates were higher than their corresponding better parent heterosis estimates. Both positive and negative mid-parent and better-parent heterosis were also recorded for these characters. The cross MOKWA-NG x MADLS-NG obtained a relatively high and positive mid-parent (>30%) and better-parent (>30%) heterosis for number of primary vines plant -1 and seed production efficiency. Similar results in reciprocal cross showed that MOKWA-NG x KOGI-NITN obtained a relatively high and positive mid-parent (>40%) and better-parent (>30%) heterosis for number of primary vines plant-1 while KOGI-NITN x OKWE-2-IM cross combination was associated with a relatively high and positive mid-parent (>23%) and better-parent (>11%) heterosis for seed production efficiency. Even though the differences between parents and crosses were non-significant (P>0.05) for pod weight and seed weight plant -1 respectively, the cross KOGI-NITN x MOKWA-NG was associated with a relatively high midparent (>20%) and better-parent (19%) heterosis for pod weight and an equally high mid-parent (>40%) and betterparent (>40%) heterosis for seed weight plant -1 .
The results indicated preponderance of non-additive genetic variance in the expression of the six characters under study which offers good scope for the exploitation of heterosis or hybrid vigor in enhancing the productivity of egusi-melon. Furthermore, consideration should be given to specific combining ability and reciprocal effects in selecting sources of inbreds with potential for improving seed production efficiency and seed yield in egusi-melon. *, ** significant at 0.05 and 0.01 level of probability respectively

19.44
*, ** significant at 0.05 and 0.01 level of probability respectively Table 5: Mid-parent (Mp) and Better-parent (Bp) heterosis of six characters for reciprocal cross combinations of egusi-melon Genotypes in a 7x7 diallel cross.