Genetic Differentiation of ARC Soybean [ Glycine Max ( L . ) Merrill ] Accessions Based on Agronomic and Nutritional Quality Traits

Soybean is one of the most important leguminous crops grown globally for food and feed. The study of genetic diversity is invaluable for efficient utilization, conservation and management of germplasm collections. The study aims at assessing genetic diversity present among the soybean genotypes using phenotypic markers. The restriction maximum likelihood revealed highly significant differences among the genotypes for eight quantitative traits. The principal component analysis revealed three most important PCs contributing 63.19%, 25.43% and 8.88% to the total variation of 97.5%, respectively. Seed yield was highly significant and highly correlated with seed number per plant, pod weight per plant, pod number per plant, and hundred seed weight but negatively correlated with seed number per pod. The hierarchical clustering revealed three major clusters with further sub-clusters. The accessions 2015/06/12, 69 S 10, PR 154-14, R 5-4-2 M, Hawkeye (USSR), and PR 145-2 were the most diverse. There were significant differences among the accessions based on nutritional quality traits such as oil, protein and stearic acid across the locations. The protein content varied from 29.1% to 35.6%, oil content varied from 10.6% to 20.7% whereas oleic acid and ash varied between 6.8% and 30.8%, and 4.3% and 8.2%, respectively. There was vast genetic diversity among the soybean genotypes. The presence of genetic diversity will aid breeders in selections and hybridization programmes for crop improvement.


Introduction
Soybean [Glycine max (L.) Merrill] (2n=2x=40) is one of the most important legumes produced worldwide.According to Food and Agriculture Organisation of the United Nations [1] the three major world-producing countries are U.S.A (90.6 million metric tonnes), Brazil (68.5 million metric tonnes) and Argentina (52.6 million metric tonnes).The total production in Africa was 1.5 million tonnes with West Africa producing 437,115 metric tonnes.Nigeria is the leading producer in West Africa with 393,860 metric tonnes [2].In South Africa soybean is produced in almost all provinces with the Free State being the major producer.In 2014, South Africa produced 8851 metric tons in an area of 502900 hectares.Soybean is grown primarily for the production of seed and has several uses in the food and industrial sectors.It is the most important crop provider of proteins and oil used in animal nutrition and for human consumption.It contains 40 to 42% good quality protein and 18 to 22% oil comprising 85% unsaturated fatty acids and is free from cholesterol, it is highly desirable in the human diet [3].Besides fixing the atmospheric nitrogen, this crop has the ability to grow in a wide range of environments, to reduce soil erosion, to suppress weeds and to suit inter and sequential cropping patterns.
It is valued finding to understand genetic diversity and relationship for facilitating the transfer of useful genes among cultivated species and maximizing the use of available germplasm resources.The extent of genetic diversity in germplasm can be assessed through morphological characterization.The characterized material then helps the plant breeders to select the accessions to be utilized in hybridization programme [4].An investigation of genetic relatedness at a broad level may provide important information about the historical relationship among different genotypes.It reveals genetic backgrounds and relationships of germplasm and also provides strategies to establish unitize and manage crop core collections [5].Therefore the knowledge of the genetic variation within accessions from germplasm collections is essential to the choice of strategies to incorporate useful diversity into the program to facilitate the introgression of genes of interest into commercial cultivars, to understand the evolutionary relations among accessions, to better sample germplasm diversity and to increase conservation efficiency [6].
Morphological characters, both quantitative and qualitative have long been used to identify species, genera, to evaluate systematic relationships, and to discriminate between varieties [7].In breeding practice and seed production, the role of morphological descriptor is very important, since the distinguishing between varieties can be done quickly and precisely.
Qualitative traits are usually controlled by a few genes, thus easily observable and suitable for cultivar differentiation and identification.On the other hand, quantitative traits have more limitations in cultivar description, since they are affected by environmental effects, developmental stage of the plant and the generation of selfing of breeding material.According to Khalid, et al. [8] scientific classification of the plant still relies on morphological traits.They are easier to work with, cost effective and easy to score and requires less time and finally they do not need any technical knowledge.Kumar, et al. [9] evaluated genetic diversity and interrelationship of agro-morphological traits in soybean genotypes.Vesna, et al. [10] determined genetic relatedness of soybean genotypes based on agro-morphological traits and DNA markers and they found genetic differences among the genotypes.Malek, et al. [11] also assessed genetic variability and association of characters among the soybean mutants and reported a vast genetic variability.Khatab, et al. [12] reported the presence of genetic diversity among soybean genotypes assessed through agro-morphological descriptors.
Soybean has nutritional quality attributes such as protein and oil that makes it an important food crop.It has 40-42% high quality protein and 18-22% oil comprising up of 85% unsaturated fatty acid as well as 12% carbohydrates [13].Soybean not only contains high quality protein, but the protein content is also much higher than that of other plant foods.Soy protein is valued as a healthy protein due to containing a balanced proportion of all of the important and essential amino acids required by the human body [14].It can provide two fold more proteins as compared to any other vegetable crop or grain [15].Soy oil can serve as a good source of oleic and linoleic acid, even the partially hydrogenated soy oil contains 25% linoleic and 3% linolenic acid [16].Soybean oil is also a good source of vitamin E [17].Some recent studies revealed variations among the soybean accessions based on quality traits such as oil [18].However, the nutritional quality traits vary depending on the varieties grown.There are limited studies on the analysis of genetic diversity among the soybean accessions based on nutritional quality traits in South Africa.There is still need to understand and record variations due to other nutritional quality traits that soybean cultivars possess.This will aid breeders to improve the quality of soybean cultivars and for selection of best accessions for both quality agronomic attributes.The objective of the study was to assess genetic diversity using agro-morphological and nutritional quality traits among soybean genotypes grown in South Africa.

Plant Material. Experimental Layout and Management
Ninety-eight soybean genotypes maintained at the Agricultural Research Council-Grain Crops Institute were planted in Potchefstroom (26.7145° S. 27.0970° E) and Brits (25.6100° S. 27.7960° E) in alpha lattice design replicated two times.Each plot consisted of two 4 m length rows with a spacing of 75 cm between the rows and 10 cm between the plants.Fertilizer 2:3:4 was applied before planting.The plants were irrigated using sprinkler irrigation system.The pre-emergent weeds were controlled by herbicide Bateleur Gold and post emergent weeds were controlled by both Basagran EC and manually.Lime Ammonium Nitrate was top dressed 45 days after germination i.e. before the plants flower.The cultural practices were applied as per soybean planting recommendations.

Data Collection and Analysis
The agro-morphological data recorded is indicated in Table 2.At harvest, five plants were randomly taken from each plot to measure days to 50% flowering; number of branches per plant; hundred seed weight (g); pod weight (g); seed number per pod; seed number per plant; pod number per plant; and seed yield (g). Analysis of variance was performed for all traits in order to test the significance of variation among the genotypes.The data were further subjected to principal component analysis and correlations.The dendrogram was constructed to study the genetic relatedness among the tested accessions using hierarchical clustering in GenStat 18 th version.

Agro-Morphological Diversity 3.1.1. Analysis of Variance of Nine Agro-Morphological Traits
There were significant differences (P ≤ 0.05) observed among the accessions grown in Potchefstroom based on seed number per plant and seed yield; and highly significant differences based on days to 50% flowering, pod weight per plant, and seed number per pod (P ≤ 0.001) (Table 3).Seed number per plant ranged between 46 and 47, and seed yield ranged between 5.58 and 156.3 g.Days to 50% flowering ranged between 55 and 121,5.Pod weight per plant ranged from 13.33 to 4.21 g.Seed number per pod ranged between 1 and 7 pods.In Brits, significant differences (P ≤ 0.05) were observed among the accessions based on number of branches per plant and pod number per plant (Table 4).Number of branches ranged between 3.8 and 8.3, and pod number per plant also ranged between 31.7 and 47.0.The genotypic effects were highly significantly different across the two locations based on number of branches per plant and seed number per plant; and there were significant differences on pod length, seed number per pod, and seed yield (Table 5).Hundred seed weight, days to flowering, pod number per plant and pod weight per plant were non-significant.
Number of branches per plant ranged between 3.6 and 8.0.Seed number per plant ranged between 71.9 and 418.2, whereas pod length, seed number per pod, and seed yield ranged from 27.8 to 44.3 mm, 1 to 3, and 11.6 to 96.8 g, respectively.
The two sites showed highly significant differences (P ≤ 0.001) in quantitative traits such as number of branches per plant, pod weight per plant, seed number per pod, seed number per plant, and seed yield and significant differences on days to 50% flowering.On the other hand, highly significant differences were observed on the genotype x site interaction for days to 50% flowering, seed number per plant and seed yield.The significant differences on genotype x site interaction could be attributed to the different reactions of the accessions to sites or due to differences between the sites.

Principal Component Analysis
The agronomic data were subjected to principal component analysis (PCA), which revealed that the three most important PCs contributed 63.2%, 25.4% and 8.9% of the total variation, respectively (Table 6).Seed number per plant, pod number per plant, and pod weight were the traits that contributed the most variation in the first PC.Seed number per plant and pod number per plant were the traits that contributed the most variation in the second PC, whereas pod number per plant, and seed number per plant were the largest contributors to the variation observed in the third PC.The principal component biplot (Figure 1) grouped the tested soybean accessions into two major groups.The accessions exhibiting early flowering, high seed number per plant, high pod weight and high seed yield were grouped together.

Correlation Analysis among Phenotypic Traits
The phenotypic traits were analysed using pair-wise rank correlations coefficients.The results and association of the traits are reported based on the significance levels of 5% (p < 0.05) and 1% (p < 0.001).Seed yield was highly significant and highly correlated with seed number per plant, pod weight per plant, pod number per plant, and hundred seed weight but negatively correlated with seed number per pod.However, was also positively and significantly associated with pod length and days to 50% flowering.Seed number per plant was highly significant and positively correlated with pod weight per plant, pod number per plant, days to 50% flowering but negatively correlated with seed number per pod.It was also significant and positively correlated with pod length.Seed number per pod was highly significant and negatively associated with pod weight per plant, days to 50% flowering, but positively correlated with number of branches.Pod weight per plant was highly and positively associated with pod length, pod number per plant, days to 50% flowering.Pod length was significant and positively correlated with pod number per plant, and number of branches.Pod number per plant was highly and significantly associated with branch number and days to 50% flowering, respectively.

Nutritional Quality Diversity
The nutritional quality traits were analysed using analysis of variance.There were significant differences (P< 0.05) among the soybean accessions planted in Potchefstroom based on ash, moisture, and oil contents (Table 8).The ash content varied between 4.3 and 10.9%.Moisture content varied between 1.7 and 12.8% whereas oil content ranged between 7.1 and 23.9%.In Brits, highly significant differences were observed in fiber, moisture, oil, palmitic acid, and protein contents, and significant differences were observed for linolenic acid (Table 9).Fiber content varied between 4.5 and 5.5%.Moisture varied between 3.1 and 4.6%.Oil content varied between 12.0 and 21.8%.Palmitic acid ranged from 55.9 to 64.2% and protein content varied between 29.6 and 35.8% whereas linolenic acid ranged from 3.1 to 4.7%.When assessing the reaction of the soybean accessions across the two locations, highly significant differences were observed for ash, moisture, oil, oleic acid and stearic acid and significant differences were observed for protein content (Table 10).Ash content varied between 4.3% and 8.2%.Moisture content ranged from 2.8% to 12.8%.Oil content varied between 10.6% and 20.7%.Oleic acid varied between 6.8% and 30.8% and stearic acid ranged between 31.6% and 263.5% whereas protein content was ranging between 29.1% and 35.6%.The two sites showed highly significant differences among the genotypes based on linolenic acid, moisture content, oleic acid, stearic acid and significant differences were observed for linoleic acid and oil content.The genotype x site interaction was significant for ash and non-significant for all other quality traits.

Discussion
The principal component analysis (PCA) is a statistical method commonly used in population genetics to identify structure in the distribution of genetic variation across geographical location and ethnic background [19].The goal is to evaluate the importance of each variable in relation to the total available variation among genotypes.The method provides an opportunity to exclude less important traits in the group studied [20] and simultaneously determine which traits are the most important.In this study, the traits that contributed to the most variation were seed number per pod, pod number per plant, seed number per plant and seed yield.
The correlation information of the tested traits can be provided to the breeders in the direct and indirect selection programmes.The genotypes with significant correlations and desirable traits can be selected concurrently.Moreover, understanding the relationship between yield and its component traits is of great importance to a breeder for making the best use of these relationships in selecting desirable genotypes for yield improvement programs [21,22].Significant positive correlations of days to flowering, number of branches and pods per plant, pod length, seed number per pod with seed yield were reported by Malek, et al. [11] which concur with the results in this study.This means that in selecting high yielding genotypes these characters should be given more emphasis as the best selection criteria.Seed yield always showed a positive correlations with other desirable yield traits [23,24] which indicates that the increase in one trait would result in the increase of the other; that is, simultaneous increase or decrease of both traits would be easy.The strong positive correlation of seed yield with other yield traits indicated that it would be very easy to identify a soybean genotype having higher seed yield simultaneously with higher number of pods per plant and but difficult with number of seeds per pod.Similar results were reported in other studies [25][26][27][28].
Traditional cluster analysis could provide an easy and effective way in determining the genetic diversity of germplasm collections [29].It is commonly used to study genetic diversity and for forming core subset for grouping accessions with similar characteristics into one homogenous category [30].It is also used to summarize information on relationships between accessions by grouping similar units so that the relationship is easily understood.In this study, the accessions were clustered into three major groups of which two of them were further sub clustered into three groups based on the agro-morphological traits measured.The most diverse lines can be used as parents for hybridization for improvement of genes of novelty.
Genetic diversity evaluation among germplasm is a necessity and a prerequisite in any hybridization program and would promote the efficient use of genetic variations [31,32].To improve an efficient crop, it is essential to obtain the information on genetic diversity and relationships among breeding materials for a plant breeder.This is because the assessment of genetic diversity is important not only for crop improvement but also for efficient management and conservation of germplasm resources.In this study, the nutritional quality traits showed a vast diversity among the accessions.The oil and protein content were within the range recorded in other studies [33].The oil content ranged from 13.8% to 22.5% whereas protein ranged between 37.0% and 50.1%) with a mean of 19% and 42.9%, respectively.The vast divergence of the accessions will assist breeders in selecting accessions with good quality for hybridization and conservation.
The analysis of genetic diversity plays a fundamental role in identification of parents [34] and it can help to achieve long-term selection gain [35].As a traditional method, morphological traits used to assess genetic divergence and classify existing germplasm materials.However, this technique is a low level but powerful taxonomic tool and has been utilized for the preliminary grouping of germplasm prior to their characterization using more precise marker technologies.Although the genetic base of soybean cultivars is considered to be extremely narrow [36] studies of genetic diversity in soybean have been conducted using morphological characteristics [37,38].Recently, De Chavez, et al. [39] reported a vast phenotypic diversity among the soybean accessions in Phillipines.Khatab, et al. [12] reported the presence of genetic diversity among soybean genotypes assessed through agro-morphological descriptors.Hamzekhanlu, et al. [40] studied 34 mutant lines including one control cultivar and detected variability for ten quantitative traits in soybean.Iqbal, et al. [41] reported significant differences among all the assessed phenotypic traits.

Conclusion
Most of the traits showed positive correlations between each other, which will assist in the combined improvement of these traits by selecting only highly and positively correlated and easily measurable phenotypic traits although most were highly significant and positively correlated with seed yield per plant.The accessions were clustered into three major groups with subgroups, which showed existence of a vast genetic diversity among the accessions.The most divergent accessions were 2015/06/12, 69 S 10, PR 154-14, R 5-4-2 M, Hawkeye (USSR), and PR 145-2.The nutritional quality traits also varied significantly among the accessions.The presence of genetic diversity can be useful for breeding and selection of parents for transgressive segregation.

Table - 2
. List of abbreviations

Table - 3
. Means of agro-morphological traits of 96 soybean accessions planted in Potchefstroom.

Table - 4
. Means of agro-morphological traits of 98 soybean accessions planted in Brits.

Table - 5
. Means and mean squares of combined analysis of variance of agro-morphological traits of 96 soybean accessions planted across two sites.Potchefstroom and Brits.2016/17

Table - 6
. Factor loadings of the three PCs based on agronomic traits

Table - 8
. Means of nutritional quality traits of 96 soybean accessions planted in Potchefstroom.