Phenotypic Variability of Local Squash ( Cucurbita maxima Duch.) Grown in the Passoré Province of Burkina Faso

of


Introduction
The term squash refers to several species of the Cucurbitaceae family, usually grown for their edible fruits or sometimes for their oilseeds. Squash has an essential role in the diet of rural communities as well as in traditional medicine [1]. It is used either fresh as a dessert [2] or cooked as a main dish or ingredient in pies, soups, stews and in bakeries [3]. It is also used in the preparation of marinades and juices [2]. In addition to its nutritional role, squash is a vegetable-fruit rich in nutrients essential to the proper functioning of the body. Biochemical analyses have shown that squash is rich in water, bioactive components (with 80% of β-carotenes), as well as polysaccharides, dietary fibres, proteins and minerals. It is also a source of vitamins B6, K, E, A and C [4]. Seeds are rich in fatty acids whose linoleic acid is the main component [5,6].
Squash (Cucurbita maxima Duch) is a more widely grown species [7]. Its growing system requires at least a temperature of 25°C during the growing period and low humidity [8]. In Burkina Faso, its growing conditions are very poorly documented and the existing diversity within the plant is still unknown. The main objective of this study entitled phenotypic variability of the local squash (Cucurbita maxima Duch.) grown in the Passoré province of Burkina Faso is to know the genetic diversity of squash grown in the Passoré province. It is specially meant for (i) identifying the discriminating characters, (ii) establishing the level of Passoré Province of Burkina Faso variability of the species in the Passoré province, and finally (iii) establishing the organization of this variability. The plant material is composed of twenty (20) accessions collected in two districts (Yako and Arbollé) in the Passoré province. Among these 20 accessions, 10 are from the districts of Yako and 10 from the districts of Arbollé.

Experimental Site, Experimental Setup and Growing Practices
A trial was conducted during the 2017-2018 crop year at the Institut du Développement Rural (IDR) experimental station located in Gampéla following a randomised complete block design with three replications. Gampéla is located about twenty kilometres from Ouagadougou on the Ouagadougou-Niamey stretch. In each repeat, each accession has been sowed in a row of two seeds/hole. Row-space and hole-space are of 2 metres. The spaces between repeats are of 2 metres too. Before sowing on July 12 th 2018, the plot was ploughed with a tractor, fertilized with organic manure at a rate of 50 t/ha and then levelled with a hoe. The first weeding took place on 01 August, 21 days after sowing. During this hoeing, a weeding was carried out with one plant per hole. At the same time, a second fertilization with manure was made with 30 g of organic manure from small ruminants. Two other weeding operations were carried out upon request.

Data Collection
Thirty-two (32) variables, eleven (11) of which are qualitative and twenty-one (21) quantitative, selected mostly from the descriptor of Cucurbita mochata Duch proposed in 2007 by UPOV, were used to characterize the accessions.
The observations on the eleven (11) qualitative variables were spread throughout the plant development cycle. These variables were about leaf blade margin, fruit color (CFT), mesocarp color (CDM), seed color (CDG) and peduncle color (CDP), fruit longitudinal section shape (FFT), position of fruit maximum diameter (PDM), fruit longitudinal axis curvature (CLF), fruit base profile (PSB), fruit apex profile (PSF) and fruit stria (CAF). For the quantitative variables, the measurement techniques are reported in Table 1.

Data Analysis
The qualitative data underwent a descriptive analysis through the calculation of frequencies. For the quantitative data, an analysis of variance (ANOVA) was carried out to identify the discriminating characteristics in order to assess the level of variability. Principal component analysis (PCA) was performed to determine the main associations between the variables and to extract the most representative ones for the grouping of accessions by hierarchical ascending classification (HAC) using Ward's aggregation method. The groups resulting from this organization were characterized through discriminating factorial analysis (DFA). The Newman-Keuls test of separation of means at the 5% threshold was performed in order to identify the characteristics that discriminate the groups resulting from the HAC. On the basis of the leaf blade margin, two morphotypes were observed ( Figure 2). These are the morphotype with a full blade margin ( Figure 2a) which represents 64% of the collection and the morphotype with a weakly incised blade margin, representing 36% of the collection (Figure 2b).

Description of Variability Using Qualitative Characteristics
Observations made on the fruits show a variability in color and form. Thus, on the basis of fruit color, 4 modalities were observed ( Figure 3). The fruits were either orange in color with a frequency of 48% of the accessions analyzed, either yellow (28%), or light yellow (18%) or green, which represents 6% of the collection.
Observations made on the fruits show a variability in color and form. Thus, on the basis of fruit color, 4 modalities were observed ( Figure 3). The fruits were either orange in color with a frequency of 48% of the accessions analyzed, either yellow (28%), or light yellow (18%) or green, which represents 6% of the collection.  The cylindrical to spongy stalks were either yellow in color corresponding to 70% of the collection (Figure 5a) or green representing 30% of the collection (Figure 5b).   The results of the analysis of variance (Table 2) show that nine characters out of the twenty-two discriminate significantly between accessions at the 5% threshold, thus indicating the existence of agro-morphological variability within the collection. On the other hand, they are low (CV < 30%) for the other sixteen (16) traits.  (Table 3). Thus, axis 1 with 24% of the total inertia positively associates the characters leaf width (r = 0.941), leaf length (r = 0.901), petiole length (r = 0.678), petiole diameter (r = 0.767) and stem diameter (r = 0.778). It is negatively correlated with the number of the holes grown (r= -0.589). Axis 1 can therefore be defined as the axis of vegetative development. Axis 2, which accounts for 19.34% of the total variance, positively associates the characters internode length (r = 0.587), fruit weight (r = 0.917), mesocarp thickness (r = 0.724) and number of primary branches (r = 0.871). It was negatively correlated with the seed thickness parameter (r= -0.456). Axis 3, with 9.42% of the total inertia, is positively correlated with maximum fruit diameter (r = 0.560) and negatively correlated with seed weight (r= -0.503). Axes 2 and 3 can be considered as production axes. Axis 4, which accounts for 8.63% of the variability, positively associates the characters date at maturity (r= 0.659), date 50% female flowering (r= 0.586) and seed width (r= 0.558) and negatively the parameter fruit length (r= -0.493). Axis 4 can be considered as the cycle axis.

Organization of Squash Diversity in the Passoré
Province The dendrogram resulting from the hierarchical ascending classification (HAC) carried out on the weighted averages of the Euclidean distances gives a structuring of the variability of the twenty (20) accessions in three (3) morphological groups G1, G2 and G3 (Figure 7). These three groups are made up of 5, 5 and 10 accessions respectively. Figure 8 shows that axis 1 with 73.92% inertia and axis 2 with 26.08% inertia explain 100% of the total variability. The relationship of the groups with the axes shows that group 1 is negatively correlated with axis 1, which opposes it to group 2. Group 3 is positively correlated with axis 2. Furthermore, the Mahalanobis Mc Lachlan, (1992) distance squares (Table 4) from this analysis show that the groups are significantly different from one another at the rate of 5%. They also show that groups 2 and 3 are the most distant (27.03%). The results of the analysis of variance combined with the Newman-Keuls test (Table 5) indicate that the characters 50% male flowering, fruit length, maximum fruit diameter and mesocarp thickness significantly discriminate the three groups. Thus, Group 1 is characterized by plants with small leaves, small cylindrical fruits, early flowering and late maturity. This group is made up of low-performance accessions. Group 2 is characterized by plants with medium performance. It includes accessions with medium diameter, round fruits containing large seeds. Group 3 includes the high performance accessions, characterized by large leaves with long fruits of high maximum diameter. Thus, these accessions have large cylindrical fruits with a high mesocarp (flesh) diameter.

Discussion
The round form of the peduncle observed for all accessions analyzed suggests that the local squash grown in the Passoré province belongs to the same species. According to Gagnon et al., the form of the peduncle is a distinguishing parameter between the different species of the Cucurbita genus most commonly grown [9]. According to them, the species C. maxima is characterized by a rounded peduncle, with a spongy appearance without marked ribs. Therefore, the local squash collection of this study would belong to the species Cucurbita maxima. Moreover, the characteristics of the leaves, namely the entire or slightly incised margin of the blades of the whole collection, confirm the belonging of the accessions to the species Cucurbita maxima Duch. Indeed, the leaf margin is one of the important criteria in the classification of species of the genus Cucurbita. The leaves of Cucurbita maxima are generally whole or slightly incised. However hand, those of Cucurbita pepo are deeply incised and those of Cucurbita moschata are weakly incised and spotted with white [10].
The existence of several discriminating characters and the structuring of the accessions into three phenotypic groups indicate the existence of morphological variability in the accessions. Thus, the variation in fruit form, color, length, diameter and size could be attributed to genetic factors within the accessions. According to Nee et al., and Abdullah et al., Cucurbita produce fruits of different sizes depending on the genetic makeup [11,12]. According to Okombe et al., reported that quality traits are under the control of additive effect gene with about 30 genes coding for quality traits in cultivated Cucurbita species [13]. Earlier studies by Chung et al., Du et al., and Tarchoun et al., also found variability in fruit form and color [14][15][16]. The greater variability observed in the characteristics of the fruit could be justified by the fact that the fruit is the organ of interest, and therefore morphological variability is maintained by the local population, which is managed according to preferences.
Taking into consideration the plant's reproductive system, the existence of intermediate forms would result from natural crosses between different morphotypes of the same species.
Thus, the variability observed within the accessions analyzed could be partly due to the plant's mode of reproduction. Indeed, squash has a mixed mode of reproduction dominated by allogamy [17]. This mode of reproduction favours trait recombination and intra-population heterogeneity. According to Sagnard et al., reproductive biology, pollen dispersal capacity, breeding practices and seed exchange also play a role in structuring diversity [18].
Furthermore, the structuring of the morphological diversity observed offers possibilities for genetic improvement of the species, either by direct selection or by hybridization between groups. Thus, the organization of diversity indicates that accessions in morphological group 3 perform well compared to groups 1 and 3 and would be potential progenitors of many leaf biomass and fruit yield traits in a selection process. The characters that most discriminate the morphological groups formed are male flowering date, fruit length, maximum fruit diameter and mesocarp thickness. Thus, hybridizations between group 2 plants that produce large fruits and group 3 plants that produce long fruits would result in long-fruited varieties with large sizes. Furthermore, the very strong and positive association between the number of primary branches, fruit weight, mesocarp thickness and axis 2 is an indication of a linear relationship between these parameters. This means that the number of branches would increase with the number of fruits. This relationship is very important because the same plant provides leaves and fruits, all of which are used in the diet. Another important relationship is that between fruit weight and mesocarp diameter. This indicates that both parameters grow in the same direction. Thus the heavier the fruit, the more mesocarp (flesh) it contains, which is used for food preparation. These positive correlations observed between these traits can facilitate genetic improvement because improving one trait also improves the other.

Conclusion and Prospects
This study highlighted the existence of significant agromorphological diversity within the collection of squash accessions grown in the Passoré province. This observed variability is more noticeable with the fruit quality traits. Interesting correlations between the traits of interest were also noted. The study also showed a structuring of squash accessions into three (3) agro-morphological groups whose characters that most discriminate the morphological groups formed are flowering date, fruit diameter, fruit length.
The study also showed great morphological heterogeneity among the squash accessions of Passoré. The information from this study can be used as a springboard to extend the study on samples of the whole country and to further studies.