https://orcid.org/0000-0003-4902-7530
Figures
Abstract
Cucurbita moschata varieties grown in Africa have very low yield. They have been neglected, and totally ignored in agricultural research programs. However, interest in their fruits, seeds, flowers and leaves is growing nowadays due to their nutritional and medicinal potentials. That growing interest has prompted plant breeders and agronomists to develop research programs for their improvement. A complete diallel cross analysis of four parental lines, Long, Zouan-H, Oval, and Soubre and their twelve F1 hybrids, was carried out in a farming environment at the University Nangui Abrogoua, Abidjan, Côte d’Ivoire. The four parental lines and the F1 hybrids were evaluated for their general performances, combining abilities, potency ratio and heterosis effects. The investigated traits included plant height, and eleven fruit- and seed-related characters. The analysis of variance showed significant differences for all traits studied. In addition, the diallel model yielded highly significant gca effects of the female parents. The gca effects of the male parents were significant for all traits except plant height, length of the fruit, width of the fruit and length of the seed. Highly significant sca effects were observed in the crosses for all the traits. Strong maternal effects were observed for the weight and diameter of the fruit, weight of the pulp, number of seeds per fruit, weight of the fresh seeds and 100-seed weight. The general predictive ratio approached the value 1 for all the traits except weight of the fresh seed and width of the dry seed. Most of the characters under this study are predominantly determined by the effects of additive genes. But, weight of the fresh seed and width of the dry seed may be controlled by non-additive genes. Mid-parent heterosis was significant for all measured traits in the crosses, except the length of the fruit. And better-parent heterosis was significant for all traits except plant height, number of fruits per plant and length of the fruit. Gene expression is described by a super-dominance for many traits, and partial dominance for some other traits in all twelve F1 hybrids. Classification of the parental lines based on the effects of their general combining ability grouped the Soubre lines as promising contributors to fruit yield. The parental lines Long and Oval formed another group likely on the basis of the small size of their fruits, the small pulps, the smaller number of fruits per plant and the large number of seeds per fruit. However, Long would be a candidate parent for the development of cultivars with longer vegetative growth. The parental line Zouan-H formed the third group and it was mostly characterized by its large number of seeds per fruit and relatively large fruits.
Citation: KOUAGO BA, SEKA D, BROU KF, BONNY BS, KOFFI KHJ, ADJOUMANI K, et al. (2024) Combining ability analysis of Cucurbita moschata D. in Côte d’Ivoire and classification of promising lines based on their gca effects. PLoS ONE 19(8): e0305798. https://doi.org/10.1371/journal.pone.0305798
Editor: Sumit Jangra, University of Florida Institute of Food and Agricultural Sciences, UNITED STATES OF AMERICA
Received: June 4, 2024; Accepted: August 3, 2024; Published: August 22, 2024
Copyright: © 2024 KOUAGO et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the manuscript and its Supporting Information files.
Funding: The author(s) received no specific funding for this work.
Competing interests: The authors have declared that no competing interests exist.
1. Introduction
Cucurbita moschata is an annual crop that is primarily grown for its young shoots, fruits, seeds, flowers and leaves. Its fruits highly contribute balanced nutrition for humans [1]. It is used in various food items. For example, Cucurbita moschata powder is used to enhance the β-carotene content of Asian noodles [2]. The flour of the peeled and unpeeled fruit of Cucurbita moschata is found comparable to wheat flour [3]. The fruit of Cucurbita moschata is often used as an ingredient in baked goods, salami, and sausage [1] for its property as a natural food preservative [4]. It is consumed as pie in North America, and as a stew or a soup in Africa. The unripe fruit is eaten as a boiled vegetable in many regions of the world. The seeds are used with honey to prepare desserts in Central America [5]. In many other countries, the seeds are eaten as snack, after salting and roasting [6]. Almost all organs of Cucurbita moschata, except the stem and the roots, are used in human diet. And, for that dietary diversity, Cucurbita moschata may be considered as part of the solution to the food challenges associated with the relentless increase of world population. Besides, it may also be considered as one of the most important vegetable crop for humans [7] due to its richness in nutrients and its generous fruit size. Cucurbita moschata is relatively high in energy and carbohydrates. It is a good source of vitamins, and its content of carotenoid pigments and minerals are particularly high [8]. It has a very high potential to cover the nutritional needs of the population, particularly vulnerable groups with regard to the need for vitamin A [9]. It is very rich in many essential compounds for the human body. It contains eight amino acids, vitamins A, and C, various minerals, carotene, and trace elements of phosphorus, potassium, calcium, magnesium, and zinc [10]. It contains antioxidants that help to protect the body against free radicals and, to lower the risk of severe diseases. Its nutritional benefits for humans are considerably high. Oil extracted from the seeds of Cucurbita moschata is very appreciated due to its high nutritional quality and medicinal properties [5, 6, 11, 12]. The seeds and the fruits of Cucurbita moschata have a wide range of bioactive compounds used in the treatment or the prevention of diabetes, cancer, fungal and microbial infections [5, 6, 13]. They are used in traditional medicine in developing countries as well as in advanced countries.
Despite its many nutritional and medicinal benefits, Cucurbita moschata remains a marginalized crop. In Côte d’Ivoire, the yield of Cucurbita moschata is very insignificant, and the crop has totally been ignored in agricultural research programs. Given all the nutritional potential of this species, consideration should be given to this crop in agronomic research in order to improve its productivity, its fruit and seed yields and its other characteristics of nutritional values to humans such as its leaves. The productivity of this species can be increased by improving the genetic architecture across different cultivars or through the selection of high yielding cultivars [14]. For this, knowledge of certain genetic parameters that govern the important agronomic traits of the crop is necessary, and can be obtained through diallel cross design [15] and implementation. Diallel cross design is used in plant and animal genetic research to estimate general combining ability (gca), specific combining ability (sca), potency ratio and heterosis for a population from randomly chosen parental lines [16]. Combining ability analysis helps to identify superior parents to be used in breeding programs or to identify promising cross combinations for cultivar development [17]. Crop breeders typically utilize combining ability analysis to choose parents with high general combining ability and hybrids with high specific combining ability effects. General combining ability is a measure of additive gene activity that relates to the average performance of a genotype in a series of hybrid combinations. Specific combining ability evaluates the average performance of certain hybrid combinations compared to the parental lines and is the result of dominance, epistatic deviation, and genotype by environment interactions. Therefore, both gca and sca effects are important in the selection or development of breeding populations [18]. They are referred to in the selection of parents with the potential to produce hybrids exhibiting greater heterosis. Heterosis is a quantifiable, trait-dependent and environment-specific phenotype, which is the performance of the F1 hybrid exceeding that of the parents [19].
The objectives of this study are to: 1) assess the combining abilities of the parental lines and the progenies from the crosses on the fruit, seed traits and plant height; and 2) to classify parental lines based on their gca effects in the Cucurbita moschata germplasm development and the improvement of cultivars.
2. Materials and methods
2.1. Plant material and field experiment
The plant material is composed of the seeds of four inbred lines of Cucurbita moschata and their F1 hybrids. They are the parental lines Oval (O), Long (L), Zouan-H (Z), and Soubre (S), and their F1 hybrids resulting from the complete diallel crosses of these four lines. The seeds of these parental lines come from accessions originating from Korhogo, Ferkessédougou, Zouan-Hounien and Soubre, Côte d’Ivoire, respectively, (Table 1). Accessions of the cultivars Oval, Long, Zouan-H and Soubre first underwent four successive cycles of self-fertilization to create the respective inbred lines, O, L, Z, S, used in the complete diallel crosses to obtain the seeds of the different families of F1 hybrids. A total of twelve hybrids and the four inbred lines were used in this study.
The experiment was carried out from June to November 2022 on a field of 5673 m2 (93 m x 61 m) using a randomized complete block design with three replications, in a farming environment at the University Nangui Abrogoua, Abidjan, Côte d’Ivoire. Each block measured 93 m by 18 m (1674 m2), and included 16 plots, randomly made of the twelve hybrids and the four parental lines. Each plot had a surface area of 54 m2 (9 m x 6 m) with 3 sowing lines and 4 sowing points per sowing line. Only one F1 hybrid or one parent is randomly assigned to a plot. The experimental field had a total of 576 plants. All the other agronomic management practices, such as weeding, hoeing were applied as recommended, and as needed.
2.2. Data collection and analysis
Data were collected at harvest on a sample of 10 plants randomly selected from each of the four parental lines and twelve F1 hybrids in each block. Twelve quantitative characters listed in Table 2 were used to evaluate the hybrids and the parents. The data collected served to compute the descriptive statistics for those characters. An analysis of variance was performed with the measured traits serving as the response variables. Blocks and the 16 genotypic strains were the factors. When significant differences were observed, the method of Tukey was used to separate means due to the genotypic effect of the strains at the 5% level of probability. In addition, the effects of the general combining ability (gca) of parent i and parent j and the specific combining ability (sca) of the cross between two parents i and j were determined according to the following expressions [15], ;
; and
where gcai and gcaj are the gca effects of parent i and parent j, respectively, and scaij is the sca effect of the cross,
is the mean of all crosses with the same parent i;
is the mean of all crosses with the same parent j; Yij is the mean of the crosses between parent i and parent j; and
is the mean of all crosses.
The genetic variances of the gca and sca effects were assessed with the linear diallel model [15] given as follows, , where Yijk is the observed response value of the cross between parents i and j; μ is the overall mean; gi is the general combining ability effect of parent i; gj is the general combining ability of the parent j; sij is the specific combining ability of the cross between parents i and j; rij is the effect of the reciprocal cross between parents i and j; bk is the effect of the kth block; and εijk is the experimental error.
The relative importance of the gca and sca effects were determined with the general predictive ratio (gpr) [20, 21] given as , where
is the variance of the gca effect and
is the variance of the sca effect. A gpr closer to unity indicates the hybrid’s performance is a function of the gca alone [20]. The percentage of heterosis was estimated with two methods, the mid-parent heterosis (mph) and the better-parent heterosis (bph). They are
and
, where F1 is the mean performance of the F1 hybrid,
is the average value of the two parents, P1 and P2, and bp is the value of the better parent. A test of hypothesis on the significance of the two statistics mph and bph uses the Student’s t test with respective standard error,
and
where MSE is the error variance [22, 23]. The potency ratio (P) determined the degree of gene dominance and was computed as
[24, 25], with P2 representing the average performance of the better parent, and P1 the average performance of the weaker parent. The space of P is the set of real numbers. However, P = ±1% indicates complete dominance, −1%<P<1% means partial dominance, P = 0% means absence of dominance and P>1% or P<−1% is the indication of super-dominance. The positive or negative sign of P defines the direction of dominance with respect to the parents [24]. Randomly selected individuals of each parental line were used to create classes of promising parental lines based on their gca effects on plant height and fruit- and seed-yield traits. The statistical software R [26], the packages lmdiallel [27] and ape [28] and the online application iTOL [29] were used for data analysis and graphics.
3. Results
3.1. Mean performances of parents and hybrids
The means of the plant height, fruit- and seed-yield traits are reported in Table 3. The analysis of variance showed highly significant differences among the parental lines and the hybrid combinations for all the traits studied. Fig 1 shows the diversity of the fruit traits among the parental lines. The method of Tukey for the multiple comparison of means helped to identify significantly different means. It could be seen that the line identified as Long has the highest plant height among the four parental lines. The hybrids involving Long as a parent have the highest plant height. The Soubre parent is characterized by higher number of fruits that have larger dimensions (length, width, and diameter of the fruit) and are the heaviest. Hybrids produced with Soubre as female parent have the largest number of fruits, the longest, widest and heaviest fruits. In addition, the fruits of the Soubre parents have the heaviest pulp. That trait is also inherited in the hybrids with Soubre as a female parent.
(A) Fruit of the parental line Soubre. (B) Fruit of the parental line Long. (C) Fruit of the parental line Oval. (D) Fruit of the parental line Zouan-H.
The parental lines Zouan-H, Long and Oval have significantly higher number of seeds per fruit. But the seeds are smaller and lighter. The hybrids produced from these three parents also have the highest number of seeds. In particular, Z♀ x L♂, L♀ x Z♂, Z♀ x O♂ and O♀ x Z♂ have doubled the number of seeds per fruit compared to their parents. Hybrids produced from those three parents have smaller seeds that are lighter. The Soubre parent has a smaller number of seeds that are comparatively bigger and heavier. Hybrids involving Soubre as a parent have smaller number of seeds per fruit, except O♀ x S♂ and S♀ x Z♂. But, they have heavier and bigger seeds. In most cases, the hybrids outperformed the parents. The hybrids S♀ x O♂ and O♀ x S♂ more than doubled their 100-seed weight compared to either parent.
3.2. Heterosis and values of the potency ratio for the different traits
3.2.1. Estimates of heterosis in the F1 hybrids.
The estimates of heterosis expressed as a percent increase (or decrease) with respect to the average of the two parents (mph) and the better parent (bph) are presented in Tables 4 and 5, respectively.
Mid-parent heterosis was significant for all measured traits in the crosses, except the length of the fruit. For the character plant height, significant mid-parent heterosis was observed with hybrids involving the parents Long and Zouan-H, and the parents Long and Oval. Regarding the number of fruits per plant, mid-parent heterosis was significant only in hybrids where Soubre is the female parent. None of the estimated better-parent heterosis was significant for the number of fruits per plant. For the length of the fruit, none of the estimated heterosis was significant, whether computed on the basis of the average of the two parents or the better parent. Significant mid-parent heterosis was observed in the hybrids from the crosses between the parental lines Soubre and Oval, Soubre and Zouan-H, and in the hybrid Z♀ x O♂ for the character width of the fruit. For that character, the significant estimates of better-parent heterosis were negative. They were observed in the hybrids involving the parents Zouan-H and Long, and Soubre and Long, whether used as male or female parent. For the diameter of the fruit, significant estimates of mid-parent heterosis were observed in the crosses Z♀ x L♂, S♀ x L♂, and S♀ x O♂. The reciprocal crosses yielded significant negative estimates of mid-parent heterosis. Also, significant estimates of mid-parent heterosis were observed with the hybrids from the cross Z♀ x O♂ and its reciprocal. All significant estimates of better-parent heterosis were negative. They were observed in the hybrids L♀ x Z♂, L♀ x O♂, L♀ x S♂, and O♀ x S♂. For the weight of the pulp, all estimates of mid-parent heterosis were significant except in the following hybrids O♀ x Z♂, S♀ x Z♂, L♀ x O♂, and O♀ x L♂. However, the estimated better-parent heterosis was not significant for the same trait in all hybrids except Z♀ x S♂, L♀ x S♂, and O♀ x S♂. For the character number of seeds per fruit, significant estimates of heterosis computed with either method were seen in the crosses involving Zouan-H and Long and Zouan-H and Oval and their reciprocals. Mid-parent heterosis was also significant in the hybrids S♀ x Z♂ and S♀ x L♂ and better-parent heterosis was significant in the latter hybrid. With the weight of fresh seeds per fruit, heterosis whether estimated with the average of the two parents or with the better parent, was significant in all crosses except Z♀ x L♂, L♀ x Z♂ and O♀ x L♂. And significant heterosis was found for the 100-seed weight in all hybrids except L♀ x O♂, O♀ x L♂ and S♀ x L♂. That observation applied to the two methods of estimation. Regarding the character length of the dry seed, mid-parent heterosis was significant in all hybrids except those from crosses involving Zouan-H and Long, and in the hybrid O♀ x L♂. For the same character, estimate of better-parent heterosis was significant only in crosses where the parents were Zouan-H and Oval, Long and Soubre, and Oval and Soubre. In hybrids from the crosses involving Zouan-H and Oval, better-parent heterosis was negative. For the width of the dry seed, heterosis computed with either method was significant in all hybrids. These results implied that the different hybrids exhibited varied performances according to the traits observed.
3.2.2. Effects of gene dominance.
The effects of gene dominance are determined by the values of the potency ratio. They are reported in Table 6. In the absence of a significance test, values were interpreted as they were. We did not assume a value -0.99 to equate -1.00 or a value 0.02 to equate 0.00, without statistical support. Doing so would give totally different interpretations of gene expression in many cases of the quantitative analysis.
The gene(s) controlling plant height exhibited a super-dominance in all the hybrids examined because the potency ratio was greater than one (P > 1) in all the crosses of the inbred lines of Cucurbita moschata used in this study. For the number of fruits per plant, a partial dominance was observed in the following F1 hybrids L♀ x Z♂, O♀ x Z♂, L♀ x S♂, and O♀ x S♂ because -1 < P < 1 and P ≠ 0, and a super-dominance in the expression of the character in all the other hybrids. The negative value of the potency ratio observed in the F1 hybrid L♀ x Z♂ indicates a reduction in the number of fruits per plant for that hybrid. Partial dominance was observed in the expression of the length of the fruits in all hybrids. For F1 (Z♀ x O♂) and F1 (L♀ x O♂) and their respective reciprocals, the values of the potency ratio were negative, indicating a relative reduction of the length of the fruits in those hybrids. The F1 hybrids issued from the cross Z♀ x S♂ and its reciprocal showed a super-dominance of the gene(s) determining the width of the fruits in those hybrids. In all the other hybrid combinations, partial dominance was observed in the expression of the width of the fruit, and the values of the potency ratio were negative for the following F1 hybrids Z♀ x L♂, L♀ x Z♂, L♀ x O♂, and S♀ x L♂. For the weight of the fruit, a super-dominance in the expression of the trait was observed in the F1 hybrids Z♀ x L♂, L♀ x Z♂, Z♀ x O♂, O♀ x Z♂, and S♀ x O♂. All the other hybrids exhibited partial dominance, and the values of the potency ratio were negative for the following F1 hybrids Z♀ x S♂, O♀ x L♂, L♀ x S♂, and O♀ x S♂. Expression of the diameter of the fruit varied too, according to the hybrids. A super-dominance in the expression of the diameter of the fruit was observed in the F1 (Z♀ x L♂), F1 (Z♀ x O♂), F1 (O♀ x Z♂), and F1 (S♀ x O♂), and partial dominance for the character in all the other F1 hybrids. Four F1 hybrids had a negative value of the potency ratio for the diameter of the fruit. For the weight of the pulp, the expression of the trait varied according to the hybrids, but in a different pattern. In the F1 hybrids from the crosses involving Zouan-H and Long, Zouan-H and Oval, their respective reciprocals and the F1 (S♀ x O♂), the gene(s) that determined the weight of the pulp were super-dominant. In all the other hybrids, partial dominance was observed in the expression of the character with some cases of negative value of the potency ratio. The expressions of the examined seed-related traits were mostly super-dominant. For the number of seeds per fruit, super-dominance was observed in the determination of the character in all F1 hybrids except F1 (Z♀ x S♂), F1 (L♀ x S♂), and F1 (S♀ x O♂) where -1 < P < 0. The gene(s) governing the weight of fresh seeds were super-dominant in all F1 hybrids except F1 (Z♀ x L♂), F1 (L♀ x Z♂), and F1 (O♀ x L♂) where partial dominance was observed. A super-dominance was observed in all hybrids for the 100-seed weight except F1 (S♀ x L♂) where we observed partial dominance. For the length of the dry seed, partial dominance was observed in the F1 (L♀ x Z♂), F1 (Z♀ x O♂), F1 (O♀ x Z♂), and F1 (O♀ x L♂). In all the other hybrid combinations, the expression of the length of the dry seed was characterized by a super-dominance of the long seed. And the width of the seed was characterized by a super-dominance of the wide seed in all hybrids studied.
3.2.3. Analysis of variance of the gca and sca of the traits.
The examined traits significantly differed according to blocks (Table 7). Across all the blocks, we observed highly significant gca effects of the female parents for all traits studied, indicating a very large variation of the gca effects of the inbred lines when used as females. When used as male parents, the gca effects were significant for all the traits except plant height, length of the fruit, width of the fruit and length of the seed. For all the traits, the variation in gca of the female parent was far greater than that of the male parent. Highly significant sca effects were observed for all the traits studied. The reciprocal effects were highly significant for weight and diameter of the fruit, weight of the pulp, number of seeds per fruit, weight of the fresh seeds and 100-seed weight, indicating strong maternal effects on those traits. The general predictive ratio is given as . A value of gpr = 2/3 indicates equal contribution of additive and non-additive gene effects in the determination of a trait. A value of gpr < 2/3 means the predominance of non-additive gene effects over the additive ones in the expression of the trait. And a value of gpr > 2/3 means that additive gene effects predominantly determine the trait. Especially, a gpr closer to 1 means only additive gene effects control the expression of the trait. In this study, the general predictive ratio was greater than two-third for all the traits except weight of the fresh seed and width of the dry seed. We may infer from those observations that most of the characters under this study are predominantly determined by the effects of additive gene(s). But, weight of the fresh seed and width of the dry seed are mostly controlled by non-additive gene(s). In particular, traits related to the fruit have a general predictive ratio greater than 0.9 and very close to 1, meaning that they are almost solely determined by additive gene effects.
The estimates of general combining ability varied across parental lines as well as the traits, with negative and positive values (Table 8). In general, negative gca effects would be desirable for plant height. We would want the plant to complete its vegetative growth early enough in order to allocate larger proportion of dry matter to the development of the fruits and the seeds. Positive gca effects would be preferred for fruit- and seed-related traits because they are the commonly harvestable organs that determine yield. The parental line Soubre had negative gca effects for plant height and positive gca effects for all the fruit- and seed-related traits except the number of seeds per fruit. The parent Soubre would be suitable for developing a shorter C. moschata hybrid with higher number of large fruits, and large seeds. The parent Zouan-H has negative gca effects for plant height, and positive gca effects for fruit- and seed-related traits except the number of fruits per plant and fruit length. Both Soubre and Zouan-H have very high estimates of gca effects on the weight of the fruit and pulp mass. If interest lies in a longer vegetative growth with reduced fruit- and seed-yield, the inbred line Long would be preferred as it has very high gca effects for plant height and negative or non-significant gca effects for fruit- and seed-related traits, except number of seeds per fruit. The parental line Oval would be suitable for reducing plant height and increasing the number of seeds per fruit. It has significant negative gca effects for plant height, and positive gca effects for number of seeds per fruit. Except the conspicuous difference in plant height, Long and Oval have many similarities, especially their negative gca effects on the fruit and seed traits.
Based on the sca effects, we may affirm that crosses involving the parental line Long result in hybrids with increased plant height. Crosses between Zouan-H and Long increase fruit weight, pulp weight, and number of seeds per fruit, along with an increased plant height. And crosses between Zouan-H and Oval produce hybrids with significantly reduced plant height, increased fruit weight and increased number of seeds per fruit. Combination of crosses involving the parent Soubre results in hybrids with increased fruit yield. And hybrids from crosses involving Oval and Soubre have reduced plant height, increased fruit weight and increased seed weight.
Random samples from each inbred line in each of the three blocks were used to develop a classification model, using the gca effects of the parental lines. Based on the within-group sum-of-squares, a three-group model was found appropriate because it had the minimum error variance. Fig 2 is the classification tree that assembled the parental lines in three groups based on their gca effects, each group with a distinctive color. The first group is identified with the green color and it assembled the parental line Soubre. The Soubre line used as a parent has a very high general combining ability effect on fruit traits such as weight of the fruit, diameter of the fruit, length of the fruit, weight of the pulp, and number of fruits per plant. It also has a high positive gca effect on the 100-seed weight. The second group is identified with the orange color and includes the parental line Zouan-H. It is characterized by a significantly high gca effect on the weight of the fruit, the diameter of the fruit, and the number of seeds per fruit but the seeds have a reduced size. The third group, identified with the red color, includes the parental lines Long and Oval. They both have significant positive gca effects on the number of seeds per fruit, and significant negative gca effects on all the fruits traits and many of the seed traits. The parental line Long is mostly characterized by an extended vegetative growth. Hence, Soubre would be a promising line in a breeding program where improvement of fruit yield is the objective. The parent Long would be promising in a program to develop hybrids for extended vegetative development. For improvement of the seed traits, the parental lines Oval, Long and Zouan-H could serve as a female parent if an increase in the number of seeds per fruit is the objective. The parent Soubre could also serve as a female parent if the objective is to increase the size of the seeds.
4. Discussion
The evaluation of the parental lines with respect to the fruits, the seed traits and plant height showed that the lines Soubre, Oval, Long, and Zouan-H have very contrasting characters, and pointed out the existence of a large genetic diversity in the Cucurbita moschata germplasm. In fact, the genetic diversity of this species is considerable in terms of the shapes, the forms and the sizes of the fruits and the seeds and their growth cycle [30]. The phenotypic variability in these parental lines has already been described [31]. The phenotypic variability observed in the four lines can be used in selection program for the genetic improvement of the traits of interest in Cucurbita moschata [32, 33]. The study of the performances of the parents and their F1 hybrid combinations revealed significant differences between the genotypes for each of the characters evaluated in this work. Some F1 hybrids obtained much better performances than others for each trait studied. The hybrids could be proposed as candidates in breeding programs aimed at improving traits where theses hybrids have recorded their best performances [34].
Significant positive heterosis effects were observed in different hybrids for several traits. This hybrid performance could be a consequence of the large genetic distance separating the different parental lines from each other [35]. Hybrids resulting from genetically distant parents express a heterosis effect [36]. And hybrid vigor is based on the complementation of the gametic contributions of the parents by favorable dominant genes [37] and therefore, the expression of the heterosis effect could be linked in part to the non-additive action of genes, dominance and epistasis. The detection of heterosis in a species is an important step for the improvement of the species. Indeed, the expression of heterosis by hybrids for a given trait indicates their potential to produce superior cultivars through the selection of transgressive segregants in segregating populations [38, 39]. Heterosis breeding is a potential tool to improve the quantity, quality and productivity of bitter gourd [40], which cannot be done by traditional methods.
Analysis of variance of the gca and sca revealed significant general and specific combining ability effects for the evaluated traits. These results suggested the involvement of additive and non-additive gene actions in the expression of the characters. Estimates of the general predictive ratio, , for each trait varied between 0.506 and 0.989 with most estimates being greater than 0.900. A gpr estimate of 2/3 is the indication of equal variance of the gca and the sca in the distribution of the variances among the components. For the traits studied in this experiment, the gpr estimates were greater than 2/3 for all traits except the weight of the fresh seeds and the width of the dry seed. This indicated the preponderance of additive gene effects in the genetic control of most of the traits studied. The average performance of a genotype in a series of hybrid combinations will be most relevant in the design of a breeding procedure in cultivar development. If a particular cultivar has a high gca for a trait, it means that the cultivar would be a valuable parent in the breeding program to improve that trait [41]. In our study, Soubre is the best general combiner for a large number of characters studied with the exception of the number of seeds per fruit and plant height. The lines Oval, Long and Zouan-H stand out as the best overall combiners for number of seeds per fruit. Long is the best general combiner for plant height. Thus, each of the four parental lines could be used in a selection program to improve the traits of interest in Cucurbita moschata. Also, the significant gca values for a given trait revealed that the selection and hybridization methods would lead to interesting genetic improvement for the trait, due to the accumulation of desirable and favorable alleles from both parents in the targeted genotype [34].
The best combination in terms of gca effects consistently involved one or two best overall combiners as parents. In other words, most crosses resulting in high gca involved at least one parent with a high gca (high × high, high × low) for the trait in question. The current results are in line with those reported elsewhere [42] based on watermelon crosses with high gca effects from parents with either high × high or high × low gca effects. The crossover having a high significant value of sca involving two best general combiners, would be due to the additive gene action and the additive × non-additive gene interaction which are fixable [36, 43]. For a given character, the best combination in terms of gca obtained with a cross involving a single parent having a high gca is the result of the additive gene interaction × dominance in the expression of the trait [44] and therefore, the diversity of values of parental gca would play an important role in the production of hybrids with significant gca values [36, 45]. However, for a given trait, parents with poor general combining skills sometimes give good combinations when crossed with each other, as is the case here in the cross between Zouan-H × Long for the weight of the fruit and the weight of the pulp. Significant sca effects resulting from crosses of parents with low gca indicate the presence of epistasis (non-allelic interaction) at heterozygous loci that are not fixable [44]. It is therefore suggested to use these crosses for the selection of a plant in later generations of breeding [44]. It should also be noted that non-significant sca can be obtained after a crossover involving two best general combiners [14]. Therefore, it cannot be generalized that parents with high gca effects could only produce good hybrids [46]. Furthermore, combinations having recorded high sca values can be improved through conventional genetic improvement methods such as biparental crosses and selective diallel crosses, subsequently, followed by the pedigree selection method, to break any epistatic link that may occur, in order to isolate transgressive segregants [45, 47, 48].
5. Conclusion
Our study helped to test four parental lines of Cucurbita moschata and their twelve F1 hybrids from a complete diallel cross for the evaluation of fruit and seed traits and plant height. It appears from our work that there is very high genetic diversity among the parental lines Soubre, Long, Oval, and Zouan-H and their descendants. Heterosis effects relative to the average of the two parents were observed in the hybrids for all traits except fruit length. With respect to the better parent, heterosis for different combinations of parents was significant for all traits except plant height, number of fruit per plant and fruit length. Significant gca effects were observed for all traits for the different parents. The traits studied were predominantly under the influence of genes with additive effects. The Soubre parental line proved to be the best combiner for several traits. However, Zouan-H was the best combiner for the number of seeds while Long was the best combiner for longer vegetative development. Oval was the best candidate for number of seeds per plant and for shorter plants. Significant sca effects were recorded at some crosses for each trait, suggesting that these crosses can be used for the development of new and more productive hybrids.
References
- 1. Hosen M, Rafii M.Y, Mazlan N, Jusoh M, Chowdhury MFN, Yusuff O, et al. (2022) Estimation of Heterosis and Combining Ability for Improving Yield, Sweetness, Carotenoid and Antioxidant Qualities in Pumpkin Hybrids (Cucurbita moschata Duch. Ex Poir.). Horticulturae, 8, 863. https://doi.org/10.3390/horticulturae8100863.
- 2. Lee CH, Cho JK, Lee SJ, Koh W, Park W, Kim CH (2002) Enhancing b-carotene content in Asian noodles by adding pumpkin powder. Cereal Chem. 79: 593–595.
- 3. Aziah AAN and Komathi CA (2009) Physicochemical and functional properties of peeled and unpeeled pumpkin flour. J. Food Sci. 74 (7): S328–33. pmid:19895499
- 4. Mokhtar M, Bouamar S, Di Lorenzo A, Temporini C, Daglia M, and Riazi A (2021) The Influence of Ripeness on the Phenolic Content, Antioxidant and Antimicrobial Activities of Pumpkins (Cucurbita moschata Duchesne). Molecules 26, 3623. pmid:34199320
- 5. Yadav M, Jain S, Tomar R, Prasad GBKS H (2010) Medicinal and biological potential of pumpkin: an updated review. Nutr. Res. Rev. 23, 184–190. pmid:21110905
- 6.
Marie-Magdeleine C, Mahieu M, and Archimède H (2011) Pumpkin (Cucurbita moschata Duchesne ex Poir.) Seeds as an Anthelmintic Agent?. Nuts and Seeds in Health and Disease Prevention, Chapitre 110 (1st Ed.), Academic Press—Elsevier, 978-0123-756-886. ffhal-02810473.
- 7. Abdein MAE, Hassan HMF and Dalia HM (2017) General performance, combining abilities and yield component traits in Pumpkin (Cucurbita moschata Poir.) at different conditions. Curr. Appl. Sci. technol. 17 (1): 121–129.
- 8. Jahan TA, Islam A, Rasul MG, Mian MAK, and Haque MM (2012) Heterosis of Qualitative and Quantitative Characters in Sweet Gourd (Cucurbita Moschata Duch. Ex Poir). Afr. J. Food Agric. Nutr. Dev. 12(3): 6186–6199.
- 9. Jahan M, Koocheki A, Tahami MK, Amiri MB, and Nassiri MM (2012) The effects of simultaneous application of different organic and biological fertilizers on quantitative and qualitive characteristics of Cucurbita pepo L. J. Life Sci. 12 (1): 1145–1149.
- 10. Men X, Choi SI, Han X, Kwon HY, Jang GW, Choi YE, et al. (2020) Physicochemical, nutritional and functional properties of Cucurbita moschata. Food Sci. Biotechnol. 30, 171–183.
- 11. Darrudi R, Nazeri V, Soltani F, Shokrpour M, and Ercolano MR (2018) Evaluation of combining ability in Cucurbita pepo L. and Cucurbita moschata Duchesne accessions for fruit and seed quantitative traits. J. Appl. Res. Med. Aromat Plants. Volume 9, Pages 70–77.
- 12. Caili F, Huan S, and Quanhong L (2006) A review on pharmacological activities and utilization technologies of pumpkin. Plant Foods Hum. Nutr. 61, 73e80. pmid:16758316
- 13. Enneb S, Drine S, Bagues M, Triki T, Boussora F, Guasmi F, et al. (2020). Phytochemical profiles and nutritional composition of squash (Cucurbita moschata D.) from Tunisia. S. Afr. J. Bot. 130: 165–71.
- 14. Brou KF, Sie RS, Adjoumani K, Saraka YDM, Kouago BA and Adahi CD. (2018) Diallel analysis in Citrullus mucosospermus (Fursa) for fruit traits. Afr. J. Agric. Res. 13 (38): 2006–2015.
- 15. Griffing B (1956) Concept of general and specific combining ability in relation to diallel crossing systems. Australian Journal of Biological Sciences 9(4): 463–493.
- 16. Feyzian E, Dehghani H, Rezai AM and Javaran MJ (2009) Diallel cross analysis for maturity and yield-related traits in melon (Cucumis melo L.). Euphytica 168:215–223.
- 17.
Acquaah G (2007) Principles of Plant Genetics and Breeding. Oxford:Wiley-Blackwell.
- 18. Viana JMS and Matta FP (2003) Analysis of general and specific combining abilities of popcorn populations, including selfed parents. Genet. Mol. Biol. 26:465–471.
- 19. Yu K, Wang H, Liu X, Xu C, Li Z, Xu X, et al. (2020). Large-Scale Analysis of Combining Ability and Heterosis for Development of Hybrid Maize Breeding Strategies Using Diverse Germplasm Resources. Front. Plant Sci. 11:660. pmid:32547580
- 20. Backer RJ (1978). Issues in diallel analysis. Crop Sci. 18: 533–536.
- 21. Hung HY and Holland JB (2012). Diallel analysis of resistance to Fusarium ear rot and fumonisin contamination in maize. Crop Sci. 52:2173–2181.
- 22. Wynne JC, Emery DA and Rice PW (1970). Combining ability estimates in Arachis hypogaea L. II. Field performance of F1 hybrids. Crop Sci. 10 (1): 713–715.
- 23. Heydari F, Ramezanpour SS, Soltanloo H, Mehdi KA and Shaban K (2012). Estimation of genetic parameters and effective factors in resistance to septoria leaf blotch of wheat under field condition. Int. J. Agric. and Crop Sci. 4(18): 1376–1384.
- 24.
Smith HH (1952). Fixing transgressive vigor in Nicotiana rustica. In: Goodwin TW ed. Heterosis. Ames, IA, Iowa State College Press. Pp. 161–174.
- 25. Zeinanloo A, Shahsavari A, Mohammadi A and Naghavi MR (2009). Variance Component and heritability of some fruit characters in olive (Olea europaea L.). Scientia Horticulturae 123:68–72.
- 26.
R Core Team (2023). _R: A Language and Environment for Statistical Computing_. R Foundation for Statistical Computing, Vienna, Austria. <https://www.R-project.org/>.
- 27. Onofri A, Terzaroli N and Russi L (2020). Linear models for diallel crosses: a review with R functions. Theor. Appl. Genet. pmid:33156356
- 28. Paradis E and Schliep K (2019). “ape 5.0: an environment for modern phylogenetics and evolutionary analyses in R.” _Bioinformatics_, *35*, 526–528. <https://doi.org/10.1093/bioinformatics/bty633>. pmid:30016406
- 29. Letunic I and Bork P (2021). Interactive Tree Of Life (iTOL) v5: an online tool for phylogenetic tree display and annotation. Nucleic Acids Res. 2:49(W1):W293–W296. pmid:33885785; PMCID: PMC8265157.
- 30. Ananda G (2015). Courges et Courgettes. Légumes et Fruits, Kokopelli 1(1): 1–17.
- 31. Seka D, Kouago BA, Bonny BS (2023). Assessment of the variability of the morphological traits and differentiation of Cucurbita moschata in Cote d’Ivoire. Sci Rep. 2023 Mar 6;13(1):3689. pmid:36878922; PMCID: PMC9988981.
- 32. Ahsan M.Z., Majidano M.S., Bhutto H., Soomro A.W., Panhwar F.H., Channa A.R., et al. (2015). Genetic variability, coefficient of variance, heritability and genetic advance of some Gossypium hirsutum L. accessions. J. Ag. Sci. 7 (2):147–151.
- 33. Bello OB, Aminu D, Gambo A, Azeez AH, Lawal M, Ali I, et al. (2015) Genetic diversity, heritability and genetic advance in okra [Abelmoschus esculentus (L.) Moench]. Bangladesh J. Plant Breed. Gen., 28 (2):25–38.
- 34. Golabadi M, Golkar P and Eghtedary A (2015) Combining ability analysis of fruit yield and morphological traits in greenhouse cucumber (Cucumis sativus L.). Canadian J. Plant Sci. 95 (2):377–385.
- 35. Adjoumani K, Kouonon LC, Koffi GK, Bony BS, Brou KF, Akaffou DS, et al. (2016) Analysis on genetic variability and heritability of fruit characters in Citrullus lanatus (Thunb.) Matsumura and Nakai (Cucurbitaceae) cultivars. J. Anim. Plant Sci. 28 (1): 4340–4355.
- 36. Banerjee PP, and Kole PC (2009) Analysis of genetic architecture physiological characters in sesame (Sesamum indicum L.). Euphytica, 168:11–22.
- 37. Yousfi K (2011) Etude agronomique et analyse diallèle de six variétés de blé dur (Triticum durum Desf.). Mémoire soutenu le 17/11/2011 à l’Ecole Nationale Agronomique El-Harrach-Alger en vue de l’obtention du diplôme magistère. 65 pages.
- 38. Mohammadi AA, Saeidi G and Arzani A (2010) Genetic analysis of some agronomic traits in flax (Linum usitatissimum L) Australian J. Crop Sci. 4(5):343–352.
- 39. Wannows AA, Sabbouh MY and Al-Ahmad SA (2015) Generation mean analysis technique for determining genetic parameters for some quantitative traits in two maize hybrids (Zea mays L.). Jordan J. Agric. Sci. 11 (1): 59–73.
- 40. Thingamajig C and Pugalendhi L (2013) Heterosis studies in Bitter Gourd for yield and related. Int. J. Vegetable Sci. 19 (2): 109–125.
- 41. Masny A, Madry W and Żurawicz E. (2005) Combining ability analysis of fruit yield and fruit quality in ever-bearing strawberry cultivars using an incomplete diallel cross design. J. Fruit Ornam Plant Res. 13(1): 5–17.
- 42. Hatem AK (2009). Heterosis and combining ability for yield and some characters in watermelon (Citrullus lanatus, Thunb). Minfya J. Agric. Res., 34 (1): 147–162.
- 43. Adarsh MN and Kumari P (2015). Combining ability and gene action studies for important horticultural traits in chilli (Capsicum annuum L.). Int. J. farm Sci. 5(1): 251–262.
- 44. Nataša L, Sofija P, Miodrag D, Hristov N, Vukosavljev M and Srećkov Z (2014). Diallel analysis for spike length in winter wheat. Turkish J. Agric. Natural Sci. 2 (1): 1455–1459.
- 45. Venel C (2022). Analyse du déterminisme génétique de l’aptitude à la combinaison et de l’effet des hétérosis des caractères liés au rendement chez labetterave sucrière en utilisant les designs North Carolina II et North Carolina III. Sciences du Vivant (q-bio), dumass-03855426: 1–100.
- 46. El-Tahawey MAFA Kandeel AM., Youssef SMS and Abd El-Salam MMM (2015). Heterosis, potence ratio, combining ability and correlation of some economic traits in diallel crosses of pumpkins. Egypt. J. Plant Breed. 19(2):419–439 (2015).
- 47. Iqbal AM, Nehvi FA, Wani SA, Dar ZA, Lone AA and Qadri H (2015). Combining ability study over environments in dry beans (Phaseolus vulgaris L.) SAARC J. Agric. 10(2): 61–69.
- 48. Fellahi Z, Hannachi A, Oulmi A and Bouzerzour H (2018). Analyse des aptitudes générale et spécifique à la combinaison chez le blé tendre (Triticum aestivum L.). Revue Agriculture. 9(1):60–70.