https://orcid.org/0009-0006-7033-2758
Figures
Abstract
Soybean (Glycine max) is one of the world’s most important oilseed crops and has adapted to various environmental conditions. Yields of soybeans in Nigeria are notably low due to different production constraints, including the limited availability of improved varieties and the slow replacement rate of old varieties with new and high-yielding ones. Ensuring high genetic diversity in the working germplasm is among the primary factors for the success of breeding programs in identifying high-yielding and well-adapted improved varieties. This study aimed to assess the genetic diversity and population structure of 45 soybean breeding lines of the International Institute of Tropical Agriculture soybean breeding program at the advanced evaluation stage based on phenotypic traits and SNP markers to support breeding strategies. Field trials were conducted in 2022 across three International Institute of Tropical Agriculture stations in Nigeria using a 5 × 9 alpha-lattice design with three replications. The collected yield and yield component data were subjected to analysis of variance, mean comparison, principal component analyses, and cluster analyses using R software. The genotypes were further assessed using 10,630 SNP markers obtained from DArTseq genotyping. The combined analysis of variance revealed a significant genotype × location interaction for grain yield and a highly significant difference in days to 50% flowering and days to 95% maturity. The genotypes G02, G10, G11, G01, and G24 were significantly superior in grain yield. Principal component analysis showed that the first three components explained 64.8% of total variation, with major contributions from traits such as lodging score, hundred seed weight, plant height, nodulation, and days to 50% flowering. Hierarchical clustering grouped the genotypes into five clusters, highlighting desirable traits such as high yield, early maturity, and lodging tolerance. SNP-based population structure grouped the genotypes into three distinct subpopulations. The SNP markers showed average observed heterozygosity, expected heterozygosity, minor allele frequency, and polymorphic information content of 0.08, 0.27, 0.20, and 0.22, respectively, which showed the existence of considerable genetic variation among the studied genotypes.
Citation: Assfaw AT, Bunmi O, Paterne A, Chigeza G, Mushoriwa H, Fowobaje K, et al. (2025) Genetic diversity and population structure analysis of soybean [Glycine max (L.) Merrill] genotypes based on agro-morphological traits and SNP markers. PLoS One 20(10): e0332895. https://doi.org/10.1371/journal.pone.0332895
Editor: Yêyinou Laura Estelle Loko, UNSTIM: Universite Nationale des Sciences Technologies Ingenierie et Mathematiques, BENIN
Received: February 20, 2025; Accepted: September 5, 2025; Published: October 10, 2025
Copyright: © 2025 Assfaw 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: The data underlying the results presented in the study are available from IITA Data Bank, https://data.iita.org/ with the associated DOI, https://doi.org/10.25502/mja8-g609/d.
Funding: This research was funded by the Pan African University Life and Earth Science Institute (including Health and Agriculture) (PAULESI) as part of the MSc funding awarded to ATA with grant number PAU020110MB (https://www.pau-au.africa/institues/paulesi). The study was also supported by the International Institute of Tropical Agriculture (IITA) in connection with the United States Agency for International Development (IITA/USAID) Genetic Improvement in Soy Project with number PJ-2315 (https://www.iita.org/iita-countries/nigeria/). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Soybean [Glycine max (L.) Merrill] is one of the leading oilseed crops globally, accounting for approximately 57% of global vegetable oil production [1,2]. It is cultivated in numerous countries and serves as a major source of vegetable oil and protein, essential in food, feed, and various industrial applications [3]. With a protein level of 40–42% and an oil content of 18–22%, soybeans have twice as much protein as meat or chicken and all eight essential amino acids required for a child’s healthy development [4]. Soybeans were domesticated around the 11th century BC in Northeast China and subsequently spread across Asia, the USA, Brazil, and Argentina [5]. Nigeria ranks as the second-largest producer of soybeans in Africa, following South Africa [6]. Soybeans were first introduced to Nigeria in 1908, with successful commercial cultivation beginning in 1937, using the Malayan variety in Benue State [7]. The crop is adapted to diverse environmental conditions and is predominantly grown under rain-fed conditions [8]. Despite Nigeria being the second highest producer of soybeans in Africa, the national average yield is < 1 ton/ha, which is far below the potential yield of the crop (over 3 tons/ha) [6]. However, the limited availability of high-yielding and disease-resistant varieties and the slow rate of replacement of old varieties with high-yielding, climate, and stress-resilient varieties are among the major factors contributing to the low yields of soybeans. Hence, enhancing the genetic improvement of the crop to make high-yielding, climate-resilient varieties available for production can have paramount importance in increasing soybean production, which will greatly benefit humanity, primarily by reducing malnutrition. Genetic improvement of soybeans plays a pivotal role in addressing malnutrition in Nigeria through enhancing the crop’s nutritional composition [9]. Currently, efforts are underway by the IITA soybean breeding program to improve protein content, fatty acid profiles, and reduce anti-nutritional factors such as phytic acid and protease inhibitors, thereby increasing the crop’s nutritional value and its role in combating malnutrition. Recent studies have shown that local soybean varieties differ significantly in their proximate composition, including protein and fat levels, highlighting the potential for selecting superior nutritional lines [10,11]. Additionally, integrating biofortification strategies into soybean breeding can help enhance the supply of essential micronutrients like iron and zinc, which are commonly deficient in low-income populations [12,13]. Thus, soybean improvement strengthens food security and serves as a tool for mitigating micronutrient deficiencies and protein-energy malnutrition [14,15].
A fundamental step for success in any breeding program is evaluating and understanding the extent of genetic variability in the crop of interest [16]. The genetic diversity of a crop species can be assessed using phenotypic traits and molecular markers [17]. A phenotypic diversity study is the standard method of evaluating the extent of genetic diversity and determining the agronomic value and grouping of crop germplasm [18]. Understanding the phenotypic variation and trait relationships helps crop breeders to develop more adaptable and productive varieties [17]. Key traits, like number of seeds per plant/pod, number of pods per plant, 100 seeds weight, leaflet shape, flower color, stem architecture, number of days till flowering or maturity, plant height, pubescence type, and density, grain yield and other related factors are among the phenotypic qualities that are typically evaluated for genetic diversity of soybean [1 9–25]. Several studies on phenotypic traits have found a high genetic diversity in soybean germplasms. Liu et al. [26] reported high phenotypic variation in characterizing 138 soybean accessions based on yield and yield-related agro-morphological traits. Similarly, Bairagi et al. [27] reported high genetic variation among 32 soybean genotypes based on ten morphological traits, while Kuswantoro et al. [28] studied the phenotypic diversity of 100 soybean genotypes and reported significant variations for all the agronomic traits. Based on nine agro-morphological traits, Marconato et al. [29] also reported the existence of high genetic diversity among the 93 soybean accessions maintained by the Brazilian Agricultural Research Corporation (EMBRAPA) gene bank. All the aforementioned findings indicate that the agro-morphological traits were helpful in assessing genetic diversity that facilitates the utilization of the genetic resources for the genetic improvement of the crop. In crop species and their relatives, selection based on phenotypic features is still frequently used and will likely continue to be an important approach in determining the extent of diversity [25].
Molecular markers are the preferred approach for assessing genetic diversity due to their excellent repeatability, superior genome coverage, automation potential, great variability, neutrality, and lack of sensitivity to environmental variations [30]. There are reports on using different types of molecular markers for diversity and population structure studies in soybeans. However, SNP markers are the most commonly used molecular markers in genetic diversity in most of the recent studies, given that they are extensively spread across the plant’s genome [31]. This is due to its affordability, target accuracy, and codominant character [32,33]. Genetic diversity studies based on SNP markers were conducted on various crops including soybeans [34–37] cassava [33]; maize [38–40]; and yam [41–43]. Despite, the great significance of assessing the phenotypic and genotypic diversity, the study materials that were developed by the IITA soybean breeding program and were at an advanced stage of evaluation have not been assessed for their genetic diversity based on both the phenotypic traits and molecular markers. Therefore, the objectives of this study were to evaluate the genetic diversity and population structure of the advanced soybean breeding lines of the IITA soybean breeding program for yield and yield-related traits using agro-morphological traits and SNP markers to recommend the best-performing varieties for direct production or use as parental lines for future genetic improvement of the crop.
Materials and methods
Description of the study area
The field experiments were conducted across three stations, i.e., Ibadan, Zaria, and Ikenne of the International Institute of Tropical Agriculture (IITA), Nigeria in the 2022 cropping season. The study locations represent Nigeria’s different soybean production environments and are characterized by different agroclimatic conditions presented in Table 1.
Experimental materials
A total of 45 soybean breeding lines, which are part of the working germplasm of the IITA soybean breeding program, along with the IITA check (TGX-1951-3F a variety developed by IITA and released in Nigeria) and a commercial check (SC-Signa, a variety released by a private company called SeedCo in Nigeria) were used in the study. Among the entries, 38 genotypes were developed by and sourced from the soybean breeding program of IITA that were at an advanced stage of yield trials in Nigeria; four from the USDA soybean genetic resource center, one from Ghana, and one from Uganda. The study genotypes with their corresponding sources are presented in Table 2.
Experimental design and management
The field trial was laid out in a 5 × 9 alpha lattice design with three replications. Each entry was planted in a plot of 4 rows of 4 m length. The spacings between rows and plants were 50 cm and 5 cm, respectively. The two middle rows were harvested to measure plot yield and other related traits, and two border rows were left to exclude the border effect. A mixture of NPK and TSP fertilizers was applied at 25g/row at a 1:2 ratio at planting. The seeds were inoculated with Bradrhizobium japonicum inoculant called Nodumax, manufactured by the IITA Business Incubation Platform (BIP). All the rest of the management practices were applied as per the recommendation for the crop [46].
Phenotypic data collection
The agro-morphological traits including plant height and root nodule score were determined from the average values of five randomly selected plants of each genotype, whereas days to 50% flowering, days to 95% maturity, lodging score, shattering score, hundred seed weight and grain yield were collected on a plot basis from the trials following the soybean descriptor of IBPGR [47] as shown in Table 3.
Phenotypic data analysis
Analysis of variance and mean comparisons.
The quantitative data were subjected to a combined analysis of variance to test for significant differences among genotypes using the linear mixed model (LMM) procedure of the R software package (version 4.3.1, 2023). Locations and replications within locations were considered random effects, whereas the genotypes were considered as fixed effects and used to determine the significance level of genotypes (G), environments (E), and their interaction (GEI). The combined ANOVA model used in this study is provided in the following equation.
where Yijkl is the response of the ith genotype in jth environment and kth replication within the jth environment and lth block within replication; μ is the grand mean, Gi is the effect of ith genotype; Ej is the effect of jth environment; Rk(j) is the effect of kth replication within the jth environment; Bl(jk) is the effect of lth block in the jth environment and kth replication; GEij is the interaction effect of ith genotype and jth environment; and eijkl is the random error effect.
The mean comparisons were done using the least significant differences (LSD) at a 5% level of significance.
Cluster analysis.
Cluster analyses were used to group the genotypes into homogeneous forms based on quantitative characters. A dissimilarity matrix was first computed using Euclidean distance, which is appropriate for continuous quantitative traits. Hierarchical clustering was then performed using Ward’s D² method (implemented in R as ward.D2), based on the dissimilarity coefficients among the 45 soybean genotypes. The analysis was performed using the base R function, and the dendextend package was used only to visualize the dendrogram (version 4.3.1, R Core Team, 2023).
Principal components analysis (PCA).
Principal component analysis (PCA) was computed to determine the traits that accounted for much of the total variation and to assess the extent of genetic diversity in the studied genotypes. The analysis was performed using the ‘FactoMiner’ package for PCA and the ‘factoextra’ package for visualization in R software [48]. In this analysis, only principal components (PCs) with eigenvalues greater than one were considered important for the total variations. PCA was not used to select traits for clustering.
Genotyping
Leaf sampling, DNA extraction, and genotyping data processing.
Seeds of the 45 genotypes were sown in the screen house at IITA, Ibadan, for sampling. For the analysis, fully expanded but young leaves from three-week-old seedlings were collected from four to five plants of each of the 45 genotypes (Table 2) using a leaf puncher and kept in a zip-lock bag on ice and later stored at −80 ◦C in a deep freezer dryer. Before genomic deoxyribonucleic acid (DNA) extraction, each sample leaf was bulked and lyophilized for 72 hours in a Labconco Freezone 2.5 L System lyophilizer (Marshall Scientific, LABCONCO, Kansas, MO, USA) and reduced to a fine powder in the SpexTM Sample Prep 2010 Geno/Grinder (Thomas Scientific, Metuchen, NJ, USA). The deoxyribonucleic acid (DNA) was extracted using a technique developed by Intertek-AgriTech (http://www.intertech.com/agriculture/agritech/), accessed 16 January 2024, and based on the LGC oKtopureTM automated high-throughput ‘sbeadexTM’ DNA extraction and purification system (https://www.biosearchtech.com/), accessed 16 January 2024.
Magnetic separation was used in the ‘sbeadexTM’ technique to prepare nucleic acids. The first stage in this process was to homogenize leaf tissue samples in 96 deep-well plates using steel bead grinding. The ground tissue was treated with a DNA extraction buffer using LGC’s ‘sbeadexTM’ kit for plant DNA preparation (https://www.biosearchtech.com/, accessed 16 January 2024). Finally, super-paramagnetic particles coated with ‘sbeadexTM’ surface chemistry that catches nucleic acids from a sample were used to purify extracted DNA. Purified DNA was eluted and used in downstream procedures. Medium-throughput genotyping was conducted in a 96 plex DArTseq protocol, and SNPs were called using the DArT’s proprietary software, DArTSoft, as described by Kilian et al. [49]. Each sequencing result’s reads and tags were mapped to the G. max reference genome, which was used to convert the raw HapMap file to a Variant Call Format (VCF).
Genotypic data analysis.
A total of 59,126 SNP markers were identified from the raw DArTseq SNP-derived dataset before quality assessment. VCFtools [50] was used to perform the initial filtering, which involved removing SNP markers with a minor allele frequency (MAF) < 0.01, markers with >20% missing data (i.e., SNP call rate < 80%), unmapped markers to any chromosome, and duplicated markers. Subsequently, PLINK v1.9 was employed for additional quality control, specifically for excluding SNPs with high heterozygosity using Hardy–Weinberg Equilibrium (HWE)-based filtering. In the end, 10,630 informative SNP markers were retained and used for the subsequent analysis. Diversity indices statistics, such as observed heterozygosity (Ho), expected heterozygosity (He), minor allele frequency (MAF), and the polymorphic information content (PIC) were estimated using PLINK 1.9 [51]. Ho was calculated with the method suggested by Chesnokov and Artemyeva [52]:
He was determined following the equations given by Liu [53]
Where the summing is over all possible alleles, and pi is the frequency of the ith allele
The MAF values were calculated using Xue et al. [54] equation as follows:
where Xi is the number of minor alleles detected at a point, and X is the total number of genotypes detected at a point.
In the same way, the PIC values were calculated by the following formula: Amiryousefi et al. [55]
Where pi and pj represent the population frequencies of the i-th and j-th alleles, respectively. The first summation includes all alleles, while the double summation covers all combinations, where i ≠ j.
Bayesian information criterion (BIC) was used to define the optimum sub-populations (K) using discriminant analysis of principal components (DAPC) and which was implemented in R using the ‘adegenet’ package [56]. Using the ancestry probability, the level of admixture was estimated, and individuals were assigned to a specific population when their membership coefficient in that group was ≥ 0.70. Genotypes with membership coefficients less than 0.70 at each assigned K were considered admixed. Coefficients of similarity showing genetic distances among the soybean lines were calculated in R software following Gower’s Distance model [57].
Results
Combined analysis of variance for agronomic traits over locations
The results from the combined analysis of variance (ANOVA) across three locations are presented in Table 4. The combined ANOVA revealed a significant (p ≤ 0.05) genotype × location interaction for grain yield and a highly significant (P < 0.001) difference for days to 95% maturity and days to 50% flowering. However, a non-significant genotype x location interaction was observed for shattering, root nodules, lodging, plant height, and hundred seed weight. The genotype effect was highly significant (p ≤ 0.001) for all the traits except for hundred seed weight. The location effect also showed a significant effect (P < 0.05) for most traits, except for lodging, days to 50% flowering, and hundred seed weight.
Location-specific performance of superior genotypes
Given the significant genotype × environment (G × E) interaction, the best-performing genotypes for each location were identified based on grain yield, days to 50% flowering, and days to 95% maturity. At Ibadan, genotype G24 recorded the highest grain yield (3,910 kg/ha), while G10 and G11 exhibited the earliest flowering (47 days) and maturity (109 days), respectively. At Zaria, genotype G05 achieved the highest yield (3,070 kg/ha), whereas genotypes G23 and G32 were the earliest to flower (37.9 days) and mature (102 days), respectively. At Ikenne, genotype G36 had the highest yield (3,850 kg/ha), while G32 and G03 recorded the earliest flowering (47.2 days) and maturity (124 days), respectively. Detailed values are provided in S5 Table.
Mean performances of the soybean genotypes for yield and yield-related traits
The mean comparisons of agronomic and yield-related traits that were determined during the study are presented in S3 Table. High mean yields of 3310, 3210, 3060, 3050, and 2990 kg/ha were recorded for genotypes G02, G10, G11, G01, and G24, respectively. The lowest-yielding genotypes were G35 (1700 kg/ha), G15 (2020 kg/ha), G18 (2040 kg/ha), and G28 (2140 kg/ha). Days to 95% maturity varied from 112 to 121 days with G32 and G33 maturing earlier (112 days) than the rest, while G25 was relatively late maturing (121 days). Mean days to maturity value of 117 days were recorded for genotypes G02, G26, G36, G39, G40 and G45. Days to 50% flowering varied from 44 to 55 days with the earliest genotype being G23 (44 days) followed by G33 with 45 days and G24 and G39 with 46 days. The genotypes that took longer to flower were G04, G05, G08, and G13 all with 55 days. Shattering ranged from 0.96 to 3.5 with an average of 1.3. Genotype G45 had a higher shattering score of 3.5. Lower shattering scores were recorded by G43 (0.96), G27, G40, and G07 (0.97). The maximum lodging score was recorded by G44 (2.43), G01 (2.4), and G08 (2.3), while the minimum was observed in G22 (1), G20 (1.02), and G17 (1.06). The root nodule score ranged from 2.17 to 3.61 with an average of 3.13. Among the 45 soybean genotypes, 27 genotypes nodulated more efficiently (above average) and the best were genotypes G02, G16 (3.61), G24, G29, G45 (3.56), and G01, G07, G15 (3.50). Genotypes G17 (102 cm), G20 (98.6 cm), G16 (96.4 cm), G19 (95.2 cm) and G26 (94.3 cm) were found to be the tallest. Genotypes G02 (63.6 cm), G03 (64.8 cm), G31 (67.4 cm), and G12 (68.3 cm) were the shortest genotypes in this study.
Principal component analysis (PCA)
The first three principal components with eigenvalues greater than one contributed to 64.8% of the total variation among the genotypes (Table 5, S1 Table). The first principal component (PC1) accounted for 27.2% of the total variation. The lodging score showed a high and positive association, while days to 95% maturity had a negative association with PC1. The considerable variation observed in the second principal component (21.3%) was positively and highly correlated with hundred seed weight and plant height, and negatively correlated with days to flowering. The third principal component accounted for 16.3% of the total variation and had a high positive correlation with nodulation and days to 50% flowering. The biplot based on PC1 and PC2 that captured 27.2 and 21.3% of the total variation, respectively, with a cumulative contribution of 48.7% (Fig 1) displayed that genotypes G15, G24, G45, G13, and G02 were far from the origin, whereas G23, G43, G11, G07, G01, and G09 were located close to the origin.
Where SHS = shattering score, RtNd = root nodule score, Lodg = lodging score, PH = plant height (cm), D50F = days to 50% flowering, D95M = days to 95% maturity, HSW = hundred seed weight (gm), GY = grain yield (kg/ha).
Cluster analysis
The cluster analysis resulting from the hierarchical ascending classification (HAC) grouped the soybean genotypes into five cluster groups, as shown in Table 6 and Fig 2.
Cluster I had the highest number of genotypes (12); while cluster IV had the lowest number of four genotypes. F-test revealed significant differences among the five clusters for grain yield, plant height, and lodging score (p < 0.001), and for days to 95% maturity (p < 0.01). Cluster IV exhibited the highest mean grain yield (2786.85 kg/ha), while Cluster III recorded the lowest (2319.34 kg/ha), indicating meaningful yield differences across clusters. For days to 95% maturity, Cluster V was the latest maturing group (116 days), significantly differing from Cluster II, the earliest group (109 days), suggesting potential suitability for diverse growing seasons. Lodging score also varied significantly, with Cluster V showing the lowest lodging score (1.23), making it desirable for environments prone to lodging. Similarly, plant height differed significantly across clusters, with Cluster V being the tallest (92.10 cm) and Cluster II the shortest (70.92 cm). In contrast, differences among clusters for days to 50% flowering, hundred seed weight, root nodule score, and shattering score were not statistically significant based on F-test.
Genetic diversity of soybean displayed by SNP markers
The results of the SNP markers’ statistics are summarized in Table 7 and S1 Fig. The observed heterozygosity ranged from 0.00 to 0.16, with a mean of 0.08, while the expected heterozygosity (He) ranged between 0.00 and 0.50, with a mean of 0.27. The minor allele frequency (MAF) ranged between 0.01 and 0.50 with a mean value of 0.20 and the polymorphic information content (PIC) ranged from 0.00 to 0.38 with an average value of 0.22.
The population structure analysis in this study identified three sub-populations among the 45 soybean genotypes based on the optimal K = 3 determined based on the BIC method (Fig 3).
Using the 70% cut-off criterion of the membership probability threshold, 27 genotypes were successfully assigned to the three different sub-populations. The remaining 18 genotypes with a probability of associations of less than 70% were considered as an admixed population (Fig 4, S2 Table).
Red, blue, and green color denotes sub-populations I, II, and III, respectively.
Sub-population I (red) was comprised of 13 genotypes that were sourced from IITA and Uganda, while the seven genotypes were allocated in sub-population II (blue) that were sourced from Ghana and IITA. Sub-population III (green) consisted of seven genotypes obtained from USDA and IITA. The phylogenetic tree also showed three sub-populations with higher degrees of admixture similar to the BIC (Fig 5).
Dots indicating individual genotypes.
Similarly, the DAPC assigned the genotypes to three cluster groups and clearly showed a higher degree of admixture among the genotypes. The first and second PCs accounted for 35.8 and 13.8% of the total variation, respectively. The members of Cluster III were the most compact in distribution, while those of Cluster I were the most widely distributed along with the axes of the first two PCs (Fig 6).
Note Each color corresponds to population structuring and grouping.
Genetic distance between genotypes based on 10,630 SNP markers
The genetic distances across the genotypes are summarized in S4 Table. In cluster one, the highest genetic distance (0.335) was observed between genotypes SY007 and SY024, while the lowest genetic distance (0.14) was observed between genotypes SY013 and SY012. In cluster two, the maximum genetic distance (0.335) was observed between genotypes SY026 and SY030, while the minimum genetic distance (0.012) was observed between genotypes SY037 and SY029. In cluster three, the widest genetic distance (0.335) was observed between genotypes SY030 and SY026, while the shortest genetic distance (0.014) was observed between genotypes SY026 and SY020.
Discussion
Genetic diversity analysis in crops is required for the success of any plant breeding effort [58]. Hence, evaluating crops’ population structure and genetic diversity is vital for adopting effective genetic resource management and conservation measures [59]. The genotype × location interactions across the three locations were significant for grain yield, and highly significant for days to 95% maturity, and days to 50% flowering, implying genotypes exhibited varying relative performances across locations for these traits. The results are in line with the findings of Tesfaye et al. [60], Krisnawati and Adie [61], and Pedro [62], who reported a highly significant genotype x environment interaction for days to 50% flowering, days to 95% maturity, and grain yield. Njoroge et al. [63], Nachilima [64], and Abebe et al. [65] also reported significant genotype x environment interaction for grain yield. Five genotypes, i.e., G02, G10, G11, G01, and G24 showed significantly superior performance for grain yield (Table 5). These genotypes can be released as varieties for direct production by farmers after further evaluation or can be used as parents in the soybean hybridization programs. Getnet [66], Beyene et al. [67], and Thio et al. [68] also obtained different mean yield performances among soybean genotypes in their studies.
Days to 95% maturity varied from 112 to 121 days, with G32 and G33 maturing earlier than the rest, while G25 maturing relatively late. Beyene et al. [67], Thio et al. [68], Sileshi [69], Goonde and Ayana [70] and Yirga et al. [71] reported comparable trends in variability in days to maturity for different soybean genotypes. Days to 50% flowering varied from 44 to 55 days with the earliest genotype being G23, followed by G33. The genotypes that took longer to flower were G04, G05, G08, and G13 all with 55 days. The results for days to 50% flowering agree with the findings of Goonde and Ayana [70] and Jandong et al. [72], and Akter [73]. Shattering ranged from 0.96 to 3.5 with the highest tolerant genotypes being G43, G07, and G27, whereas G45 was highly susceptible to pod-shattering. This result aligns with Fatima et al. [45] and Aondover et al. [74], who reported different mean performances in pod shattering. G44 recorded the maximum lodging score, while the minimum was observed in G22. This aligns with the work of Antwi-Boasiako [75], who found significant variation in tolerance to lodging and shattering across the 34 soybean genotypes evaluated. The root nodule score ranged from 2.17 to 3.61 with an average of 3.13. Among the 45 soybean genotypes, 27 nodulated more efficiently, and the best genotypes were G02, G16, G24, G29, G45, G01, G07, and G15. These results agree with those of Thio et al. [68] and Bello et al. [76], who reported different mean performances in root nodule scores. The plant height among the genotypes varied from 63.6 to 102 cm. The tallest plant height was recorded for G17, whereas the shortest was recorded on G02, which was also the top-performing genotype for grain yield. The results of this study is in line with the findings of Thio et al. [68] and Njoroge and Njeru [77], who reported plant height ranged from 51.1 cm to 102.8 cm and 60 cm to 109 cm, respectively. In contrast, Hizli et al. [78] reported up to 151.8 cm plant height, which is taller than the heights found in this study.
Principal component analysis (PCA) aids in simplifying multidimensional data by breaking down its complexity into simpler principal components (PCs) by retaining important information [79,80]. The first three PCs with eigenvalues greater than one, which accounted for about 64.8% of the total variation were considered important (Table 5). In a similar study on soybeans, Aondover et al. [74] reported that the first three principal components accounted for 76.4% of the total variation. Similarly, Verma et al. [81] reported that the first three PCs contributed 91.7% of the total variation. In their study, Akter [73] and Vijayakumar et al. [80] also reported that the first three PCs accounted for 77.7% and 73.7% of the total variation, respectively. Each PC is independently associated with the different yield and yield-related variables examined. The first PC that showed the highest contributions (27.2%) to the total variation was highly and positively associated with lodging and highly and negatively associated with days to 95% maturity, indicating the importance of these traits in the genetic improvement of the studied germplasm. The average contributions of RtNd, HSW, PH, D50F, and Lodg were high in the principal axes. This finding implies that these traits account for the majority of the variations in soybeans and further influence the yield of soybeans. This finding aligns with the results of Aondover et al. [74], Adetiloye et al. [82], Sivabharathi et al. [83], Vianna et al. [84], and Leite et al. [85], who reported the mean contributions of RtNd, HSW, PH, D50F, and Lodg were high in the principal axes. Thus, these traits must be considered when selecting grain yield. Soybean genotypes were spread over the biplot based on the first two PCs, also reflecting high variability in the studied germplasm. Genotypes G15, G24, G45, G13, and G02 were far from the origin and identified as highly interactive and diverse genotypes; whereas G23, G43, G11, G07, G01, and G09 were located close to the origin and considered stable genotypes or less diverse genotypes (Fig 1). Comparable results were reported by Sivabharathi et al. [83], who performed principal component analysis on 135 soybean genotypes and reported genotypes MACS 1460, EC 18736, and PK 1038 as highly divergent genotypes, while JS 89–24, NRC 25, NRC 2007-G-1–13, NRC 43 and PK 7247 genotypes were less diverse and stable.
The studied genotypes were grouped into five distinct clusters, which also indicates the divergent nature of these germplasm. This result was comparable to the findings of Getnet [66] and Ghiday and Sentayehu [86], who reported the clustering of 49 soybean genotypes into five and three distinct clusters, respectively. Similarly, Yirga et al. [71] performed cluster analysis of 100 soybean genotypes based on morphological traits that were grouped into five different clusters, which implied the presence of genetic variability among the tested genotypes. Adetiloye et al. [82] also reported the clustering of 43 accessions into nine distinct clusters. These findings indicate the studied soybean genotypes exhibited wide genetic diversity. Utilizing genetically diverse parents in hybridization programs is critical in crop improvement activities [73] in ensuring high genetic recombinations in the progeny populations. Considering the cluster mean values presented in Table 6 and the significant differences revealed by F-test, the observed significant differences among clusters for key agronomic traits highlight the potential of cluster-based selection in soybean breeding. Notably, Cluster IV, which exhibited the highest grain yield, and Cluster V, which combined the latest maturity duration and tolerance to lodging, represent valuable genetic sources for breeding programs. For instance, Cluster II, with its early maturity and shorter plant height, may be suitable for regions with shorter growing periods. The present outcome supports prior findings [87–90].
In this study, the observed heterozygosity (Ho) of 0.08 was lower than the expected heterozygosity of the genotypes, possibly due to soybeans’ high self-pollinating nature [91]. This implies high possibilities of inbreeding and fixation at most loci [92,93]. The mean observed heterozygosity of 0.08 obtained in this study is higher compared to the 0.02 and 0.06 reported by Tsindi et al. [35] and Chiemeke et al. [94], respectively, in soybeans using SNP markers. However, it is lower compared to the 0.22 stated by Yohane et al. [95] when assessing 81 pigeon pea genotypes based on 4122 SNP markers, but similar to the results of Gumede et al. [59], when assessing 90 cowpea accessions using 5864 SNP markers. The average expected heterozygosity (He = 0.27) found in the current study was lower than the He of 0.31 reported by Abebe et al. [34] and Tsindi et al. [35] evaluating 65 and 210 soybean genotypes using 1223 and 403 SNP markers, respectively. The minor allele frequency (MAF) values indicate the prevalence of the less common allele at a locus, helping to distinguish common from rare variants. Minor allele frequency (MAF) varied from 0.01 to 0.50 with an average of 0.20. The average MAF value found in this study was greater than the 0.24 reported by Tsindi et al. [35] based on the 403 SNPs in 210 soybean genotypes, 0.25 reported by Naflath et al. [96] based on the 29,955 SNPs in 96 soybean genotypes and 0.32 reported by Chander et al. [97] based on 186 SNPs in 155 soybean lines. Dube et al. [40] found similar results in maize with the MAF ranging from 0.01 to 0.5. A mean PIC value of 0.22 indicates that the markers had moderate informativeness for distinguishing among genotypes. A mean PIC value of 0.22 in the current study was lower than the PIC values reported in various crops such as 0.25 in soybean [34], 0.31 in maize [40], 0.36 in wing yam [41], 0.29 in wheat [98], 0.27 in cowpea [59], 0.26 in pea [99]. However, it is comparable to the results in other legumes, i.e., 0.22 in common bean [100] and 0.23 in cowpea [101].
The observed (Ho = 0.08) and expected (He = 0.27) heterozygosity values in this study reflect soybean’s self-pollinating nature, which naturally results in lower heterozygosity [91,102]. However, the level of expected heterozygosity still indicates the presence of useful genetic variation across the population [103]. This moderate diversity is essential for breeders because it enables effective selection of genotypes with favorable traits [59,104]. The disparity between Ho and He also suggests some degree of inbreeding and allele fixation, highlighting the need for incorporating genetically diverse parents in the crossing program to maintain and further increase the genetic base of IITA soybean [92,97,104]. Besides, by comparing Ho and He to other soybean germplasm such as Abebe et al. [34], Tsindi et al. [35], and Chiemeke et al. [94], breeders can pinpoint the most diverse parents for hybridization and thus maximize recombination of useful alleles [105,106]. Additionally, moderate average values PIC (0.22) confirm that the SNP markers used are informative enough to distinguish genotypes and support population structure analysis [54,97]. These diversity indices together provide valuable guidance for parental selection, germplasm conservation, and marker deployment in soybean improvement programs [54,105].
Population structure analysis is the process of determining a breeding line’s ancestry using genotypic data [38,107]. The SNP analysis identified three sub-populations (K = 3) among the 45 soybean genotypes based on the optimal K = 3 determined according to the BIC method, which was not consistent with the results of phenotypic clustering. The inconsistency might be that most of the phenotypic traits are controlled by polygenes, and environmental conditions highly influence these traits [108,109]. In the same way, the phylogenetic tree and DAPC also showed three sub-populations with higher degrees of admixture. The high degree of admixture found suggests that there was either gene flow or that these subpopulations had a common ancestor [91]. Chander et al. [97] found comparable amounts of admixture in their study of 155 soybean genotypes, predominantly IITA-bred soybean varieties. The overall results of the population structure analysis are in agreement with the results of Abebe et al. [34] and Fatokun et al. [101], who identified three sub-populations among the 65 soybean genotypes and 298 cowpea accessions, respectively based on BIC and DAPC methods.
The analysis of genetic distances within clusters revealed considerable variation, reflecting the underlying genetic diversity among the genotypes. In cluster I, the greatest genetic distance was recorded between genotypes SY007 and SY024, suggesting substantial genetic differentiation between these two individuals. This could imply limited recent shared ancestry or contrasting selection pressures. On the other hand, the lowest genetic distance was observed between SY013 and SY012, indicating a closer genetic relationship, possibly due to recent common ancestry or shared breeding lineage [107,110]. In cluster II, the widest genetic distance was found between genotypes SY026 and SY030, mirroring the divergence seen in Cluster 1. Such high differentiation might suggest that these genotypes belong to distinct subpopulations or have undergone different evolutionary pressures. Conversely, the minimal genetic distance in this cluster was surprisingly low between SY037 and SY029, indicating a near-identical genetic makeup. This could reflect either clonal propagation, recent hybridization, or strong genetic conservation between the two genotypes. Cluster III also exhibited a maximum genetic distance, again between SY030 and SY026, reinforcing the notion of pronounced genetic divergence between these two genotypes across clusters. The lowest genetic distance in this cluster was between SY026 and SY020, suggesting a close relationship and potential kinship or shared origin. This pattern of high and low pairwise distances within clusters indicates that while clustering helps group genetically similar individuals, considerable diversity still exists within each group, which is critical for maintaining adaptive potential and guiding breeding decisions. Various genetic distances within soybean germplasm have been observed in a similar study by [35,37,102,111]
Conclusion
The combined analysis of variance revealed significant differences among genotypes, locations, and the genotype × location interaction for days to 95% maturity and grain yield, indicating that genotypes exhibited different relative performances across locations for these traits. The genotypes G02, G10, G11, G01, and G24 were found to be the most high-yielding (3050−2990 kg/ha). The first three principal components explained 64.8% of the variation, with traits like HSW, RtNd, PH, Lodg, and D50F contributing the most in the principal axes. Cluster analysis grouped the 45 genotypes into five clusters, suggesting moderate variation. Based on cluster mean values and the significant differences, genotypes selected from distant clusters such as clusters II and IV, were desirable for use as parents in future hybridization programs to develop high-yielding soybean varieties.
From the molecular analysis, 10,630 SNP markers were used, showing moderate informativeness with a mean polymorphic information content (PIC) of 0.22. The mean observed heterozygosity (Ho = 0.08), expected heterozygosity (He = 0.27), and minor allele frequency (MAF = 0.20) values indicate moderate genetic variability within the genotypes. Population structure analysis using DAPC and BIC identified three subpopulations, confirming considerable genetic diversity among the studied soybean genotypes.
Supporting information
S1 Table. Eigenvalues, proportion of variance, and cumulative proportion of the studied soybean genotypes for eight agronomic and yield traits.
D50F = days to 50% flowering, SHS = shattering score, RtNd = root nodule score, Lodg = lodging score, PH = plant height (cm), D95M = days to 95% maturity, HSW = hundred seed weight (gm), GY = grain yield (kg/ha).
https://doi.org/10.1371/journal.pone.0332895.s001
(CSV)
S2 Table. List of genotypes and their grouping based on 10,630 SNP markers using the 70% cut-off criterion of the membership probability threshold.
https://doi.org/10.1371/journal.pone.0332895.s002
(CSV)
S3 Table. Mean performances of 45 soybean genotypes evaluated at three locations in Nigeria for agronomic and yield-related traits.
RtNd = root nodule score, Lodg = lodging score, D50F = days to 50% flowering, PH = plant height in centimeter, D95M = days to 95% maturity, SHS = shattering score, GY = grain yield in kilogram per hectare, SE = standard error, LSD = least significant difference.
https://doi.org/10.1371/journal.pone.0332895.s003
(DOCX)
S4 Table. Genetic distance matrix of the 45 soybean genotypes based on SNP markers.
https://doi.org/10.1371/journal.pone.0332895.s004
(CSV)
S5 Table. Location-wise mean performances for grain yield, D50F, and D95M of the study genotypes.
https://doi.org/10.1371/journal.pone.0332895.s005
(XLSX)
S1 Fig. Summary statistics of 10,630 single nucleotide polymorphism (SNP) markers used for genotyping 45 soybean genotypes.
https://doi.org/10.1371/journal.pone.0332895.s006
(PNG)
Acknowledgments
The first author is grateful for the scholarship granted by the African Union Commission for her MSc studies at the Pan African University Life and Earth Sciences Institute (including Health and Agriculture), University of Ibadan, Nigeria. We would like to acknowledge the staff and technicians of the soybean breeding and biometric unit, at the International Institute of Tropical Agriculture for all their invaluable guidance.
References
- 1. Berk Z. Technology of production of edible flours and protein products from soybean. FAO Agric Serv Bull. 1992:1–10.
- 2. Yadava DK, Vasudev S, Singh N, Mohapatra T, Prabhu KV. Breeding major oil crops: present status and future research needs. Technol Innov Major World Oil Crop. 2012;1:17–51.
- 3. Wilson RF. Soybean: market driven research needs. In: Plant genetics and genomics: crops and models. Springer New York; 3–15.
- 4. Malik MFA, Qureshi AS, Ashraf M, Ghafoor A. Genetic variability of the main yield related characters in soybean. Int J Agric Biol. 2006;8:815–9.
- 5. Hymowitz T, Shurtleff WR. Debunking soybean myths and legends in the historical and popular literature. Crop Sci. 2005;45(2):473–6.
- 6.
FAOSTAT. A list of countries by soybean production from 2016 to 2020. https://en.wikipedia.org/wiki/List_of_countries_by_soybean_production
- 7. Akande SR, Oyekan PO, Adesoye AI. Soybean in Nigeria: introduction, production and utilization. Legume. 2013;16:38.
- 8. Chandrawat KS. Study on genetic variability, heritability and genetic advance in soybean. Int J Pure App Biosci. 2017;5(1):57–63.
- 9. Singer WM, Lee Y-C, Shea Z, Vieira CC, Lee D, Li X, et al. Soybean genetics, genomics, and breeding for improving nutritional value and reducing antinutritional traits in food and feed. Plant Genome. 2023;16(4):e20415. pmid:38084377
- 10. Getaneh Zewudie K, Gemede HF. Assessment of nutritional, antinutritional, antioxidant and functional properties of different soybean varieties: implications for soy milk development. Cogent Food Agricul. 2024;10(1).
- 11. Joseph OO, Uzoma AC, Juliana AN, Precious S. Proximate and anti-nutritional analyses of three soybeans (Glycine max (L.) Merrill). IMCC J Sci. 2024;4(1):10–6.
- 12. Kumar A, V RK, Singh C, V SK, Agarwal DK, Pal G, et al. Bioprospecting nutraceuticals from soybean (Glycine max) seed coats and cotyledons. Indian J Agri Sci. 2019;89(12):2064–8.
- 13. Tamangwa MW, Djikeng FT, Feumba RD, Sylvia VT, Loungaing VD, Womeni HM. Nutritional composition, phytochemical, and functional properties of six soybean varieties cultivated in Cameroon. Legum Sci. 2023;5(4):e210.
- 14. Nwosu DJ, Olubiyi MR, Aladele SE, Apuyor B, Okere AU, Lawal AI, et al. Proximate and mineral composition of selected soybean genotypes in Nigeria. J Plant Develop. 2019;26:67–76.
- 15. Fasusi SA, Kim J-M, Kang S. Current status of soybean production in Nigeria: constraint and prospect. Korean J Int Agric. 2022;34(2):149–56.
- 16. Meena BL, Das SP, Meena SK, Kumari R, Devi AG, Devi HL. Assessment of GCV, PCV, heritability and genetic advance for yield and its components in field pea (Pisum sativum L.). Int J Curr Microbiol Appl Sci. 2017;6(5):1025–33.
- 17. Nkhoma N, Shimelis H, Laing MD, Shayanowako A, Mathew I. Assessing the genetic diversity of cowpea [Vigna unguiculata (L.) Walp.] germplasm collections using phenotypic traits and SNP markers. BMC Genet. 2020;21(1):110. pmid:32948123
- 18. Singh S, Trivedi PKA, Koshal AK. Seed protein profiling through electrophoresis in mungbean Vigna radiata (L.) Wilczek. J Plant Dev Sci. 2015;7(4):337–40.
- 19. Chen Y, Nelson RL. Genetic variation and relationships among cultivated, wild, and semiwild soybean. Crop Sci. 2004;44(1):316–25.
- 20. Dayaman V, Senthil N, Raveendran M, Nadarajan N, Shanmugasundaram P, Balasubramanian P. Diversity analysis in selected Indian soybean Glycine max (L.) Merrill using morphological and SSR markers. Int J Integr Biol. 2009;5:125–9.
- 21. Liu M, Zhang M, Jiang W, Sun G, Zhao H, Hu S. Genetic diversity of shaanxi soybean landraces based on agronomic traits and SSR markers. African J Biotechnol. 2011;10(24):4823–37.
- 22. Malik MFA, Ashraf M, Qureshi AS, Khan MR. Investigation and comparison of some morphological traits of the soybean populations using cluster analysis. Pakistan J Bot. 2011;43(2):1249–55.
- 23. Matsuo É, Sediyama T, Cruz CD, Oliveira RD de L, Oliveira R de CT, Nogueira APO. Genetic diversity in soybean genotypes with resistance to Heterodera glycines. Crop Breed Appl Biotechnol. 2011;11(4):304–12.
- 24. Salimi S, Lahiji HS, Abadi GM, Salimi S, Moradi S. Genetic diversity in soybean genotypes under drought stress condition using factor analysis and cluster analysis. World Appl Sci J. 2012;16(4):474–8.
- 25.
Mushoriwa H. Breeding gains, diversity analysis and inheritance studies on soybean Glycine max (L.) Merrill germplasm in Zimbabwe. 2013:210.
- 26. Liu Z, Li H, Fan X, Huang W, Yang J, Wen Z, et al. Phenotypic characterization and genetic dissection of nine agronomic traits in Tokachi nagaha and its derived cultivars in soybean (Glycine max (L.) Merr.). Plant Sci. 2017;256:72–86. pmid:28167041
- 27. Bairagi V, Mishra S, Sen R, Dixit S, Tyagi DB. Assessment of genetic variability in soybean (Glycine max L., Merrill). Biol Forum–An Int J. 2023;15:258–63.
- 28. Kuswantoro H, Adie MM, Putri PH. Genetic variability, heritability, and genotypic correlation of soybean agronomic characters. Bul Palawija. 2021;19(2):117.
- 29. Marconato MB, Pereira EM, Silva FM, Bizari EH, Pinheiro JB, Mauro AO, et al. Genetic divergence in a soybean (Glycine max) diversity panel based on agro-morphological traits. Genet Mol Res. 2016;15(4). pmid:27886340
- 30. Bhanu AN. Assessment of genetic diversity in crop plants - an overview. Adv Plants Agric Res. 2017;7(3).
- 31. Ganal MW, Altmann T, Röder MS. SNP identification in crop plants. Curr Opin Plant Biol. 2009;12(2):211–7. pmid:19186095
- 32. Ren J, Sun D, Chen L, You FM, Wang J, Peng Y, et al. Genetic diversity revealed by single nucleotide polymorphism markers in a worldwide germplasm collection of durum wheat. Int J Mol Sci. 2013;14(4):7061–88. pmid:23538839
- 33. Prempeh RNA, Manu-aduening JA, Quain MD, Asante IK, Offei SK, Danquah EY. Assessment of genetic diversity among cassava landraces using single nucleotide polymorphic markers. Afr J Biotechnol. 2020;19(6):383–91.
- 34. Abebe AT, Kolawole AO, Unachukwu N, Chigeza G, Tefera H, Gedil M. Assessment of diversity in tropical soybean (Glycine max (L.) Merr.) varieties and elite breeding lines using single nucleotide polymorphism markers. Plant Genet Resour. 2021;19(1):20–8.
- 35. Tsindi A, Eleblu JSY, Gasura E, Mushoriwa H, Tongoona P, Danquah EY, et al. Analysis of population structure and genetic diversity in a Southern African soybean collection based on single nucleotide polymorphism markers. CABI Agric Biosci. 2023;4(1).
- 36. Potapova NA, Zlobin AS, Perfil’ev RN, Vasiliev GV, Salina EA, Tsepilov YA. Population structure and genetic diversity of the 175 soybean breeding lines and varieties cultivated in West Siberia and other regions of Russia. Plants (Basel). 2023;12(19):3490. pmid:37836230
- 37. Andrijanić Z, Nazzicari N, Šarčević H, Sudarić A, Annicchiarico P, Pejić I. Genetic diversity and population structure of european soybean germplasm revealed by single nucleotide polymorphism. Plants (Basel). 2023;12(9):1837. pmid:37176892
- 38. Boakyewaa Adu G, Badu-Apraku B, Akromah R, Garcia-Oliveira AL, Awuku FJ, Gedil M. Genetic diversity and population structure of early-maturing tropical maize inbred lines using SNP markers. PLoS One. 2019;14(4):e0214810. pmid:30964890
- 39. Badu-Apraku B, Garcia-Oliveira AL, Petroli CD, Hearne S, Adewale SA, Gedil M. Genetic diversity and population structure of early and extra-early maturing maize germplasm adapted to sub-Saharan Africa. BMC Plant Biol. 2021;21(1):96. pmid:33596835
- 40. Dube SP, Sibiya J, Kutu F. Genetic diversity and population structure of maize inbred lines using phenotypic traits and single nucleotide polymorphism (SNP) markers. Sci Rep. 2023;13(1):17851. pmid:37857752
- 41. Agre P, Asibe F, Darkwa K, Edemodu A, Bauchet G, Asiedu R, et al. Phenotypic and molecular assessment of genetic structure and diversity in a panel of winged yam (Dioscorea alata) clones and cultivars. Sci Rep. 2019;9(1):18221. pmid:31796820
- 42. Bhattacharjee R, Agre P, Bauchet G, De Koeyer D, Lopez-Montes A, Kumar PL, et al. Genotyping-by-sequencing to unlock genetic diversity and population structure in white yam (Dioscorea rotundata Poir.). Agronomy. 2020;10(9):1437.
- 43. Agre PA, Dassou AG, Loko LEY, Idossou R, Dadonougbo E, Gbaguidi A, et al. Diversity of white guinea yam (Dioscorea rotundata Poir.) cultivars from Benin as revealed by agro-morphological traits and SNP markers. Plant Genet Resour. 2021;19(5):437–46.
- 44. Adewale BD, Ojo DR, Abberton M. GGE biplot application for adaptability of African yam bean grain yield to four agro-ecologies in Nigeria. Afr Crop Sci J. 2017;25(3):333.
- 45. Fatima UA, Mohammed MS, Oyekunle M, Abdulmalik MM, Usman A. Screening soybean (Glycine max (L.) Merrill) genotypes for resistance to pod shattering in Zaria, Nigeria. FUDMA J Sci. 2020;4(1):727–31.
- 46.
Omoigui LO, Kamara AY, Kamai N, Dugje IY, Ekeleme F, Kumar P. Guide to soybean production in northern Nigeria. Nigeria: International Institute of Tropical Agriculture; 2020.
- 47.
International Board for Plant Genetic Resources IBPGR. Descriptors for soybean. 1984.
- 48.
R Development Core Team. R: a language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing; 2007.
- 49.
Kilian A, Sanewski G, Ko L. The application of DArTseq technology to pineapple. In: XXIX International Horticultural Congress on Horticulture: Sustaining Lives, Livelihoods and Landscapes (IHC2014), 2014. 181–188.
- 50. Danecek P, Auton A, Abecasis G, Albers CA, Banks E, DePristo MA, et al. The variant call format and VCFtools. Bioinformatics. 2011;27(15):2156–8. pmid:21653522
- 51. Purcell S, Neale B, Todd-Brown K, Thomas L, Ferreira MAR, Bender D, et al. PLINK: a tool set for whole-genome association and population-based linkage analyses. Am J Hum Genet. 2007;81(3):559–75. pmid:17701901
- 52. Chesnokov YV, Artemyeva AM. Evaluation of the measure of polymorphism information of genetic diversity. Agric Biol. 2015;50(5):571–8.
- 53.
Liu BH. Statistical genomics: linkage, mapping and QTL analysis. Boca Raton, Florida, USA: CRC Press; 1998.
- 54. Xue H, Yu X, Fu P, Liu B, Zhang S, Li J, et al. Construction of the core collection of Catalpa fargesii f. duclouxii (Huangxinzimu) based on molecular markers and phenotypic traits. Forests. 2021;12(11):1518.
- 55. Amiryousefi A, Hyvönen J, Poczai P. iMEC: online marker efficiency calculator. Appl Plant Sci. 2018;6(6):e01159. pmid:30131901
- 56. Jombart T. adegenet: a R package for the multivariate analysis of genetic markers. Bioinformatics. 2008;24(11):1403–5. pmid:18397895
- 57. Gower JC. A general coefficient of similarity and some of its properties. Biometrics. 1971;27(4):857.
- 58. Ebert AW, Engels JMM. Plant biodiversity and genetic resources matter!. Plants. 2020;9:1706.
- 59. Gumede MT, Gerrano AS, Amelework AB, Modi AT. Analysis of genetic diversity and population structure of cowpea (Vigna unguiculata (L.) Walp) genotypes using single nucleotide polymorphism markers. Plants (Basel). 2022;11(24):3480. pmid:36559592
- 60. Tesfaye A, Githiri M, Derera J, Debele T. Genetic variability in soybean (Glycine max L.) for low soil phosphorus tolerance. Ethiop J Agric Sci. 2017;27(2):1–15.
- 61. Krisnawati A, Adie MM. Yield stability of soybean promising lines across environments. IOP Conf Ser: Earth Environ Sci. 2018;102:012044.
- 62.
Pedro JA. Evaluation of soybean Glycine max (L.) Merrill] genotypes for grain yield and associated agronomic traits under low and high phosphorus environments. Pietermaritzburg, South Africa: University of KwaZulu-Natal; 2018.
- 63. James NN, James OO, Maurice EO. Evaluation of soybean [Glycine max (L.)Merr.] genotypes for agronomic and quality traits in Kenya. Afr J Agric Res. 2015;10(12):1474–9.
- 64.
Nachilima C. Genotype plus genotype-by-environment interaction analysis of soybean Glycine max (L. Merrill) across production environments in Southern Africa. The University of Zambia; 2021.
- 65. Abebe AT, Adewumi AS, Adebayo MA, Shaahu A, Mushoriwa H, Alabi T, et al. Genotype x environment interaction and yield stability of soybean (Glycine max l.) genotypes in multi-environment trials (METs) in Nigeria. Heliyon. 2024;10(19):e38097. pmid:39398076
- 66. Enideg Getnet B. Genetic variability, heritability and expected genetic advance as indices for selection in soybean [Glycine max (L.) Merrill] varieties. Am J Life Sci. 2018;6(4):52.
- 67. Beyene D, Jalata Z, Geleta N. Exploring the genetic variability, heritability and genetic gain in soybean Glycine max (L.) Merrill genotypes. Int J Agric Res. 2022;17(2):43–9.
- 68. Ibié GT, Nofou O, Inoussa D, Frank E, Fidèle BN, Pierre AEDS, et al. Evaluation of medium maturity group of soybean (Glycine max L. Merr) for agronomic performance and adaptation in Sudanian zone of Burkina Faso. Afr J Agric Res. 2022;18(4):264–75.
- 69. Sileshi Y. Estimation of variability, correlation and path analysis in soybean Glycine max (L.) Merr. genotypes at Jimma, South Western Ethiopia. J Nat Sci Res. 2019;9(7):22–9.
- 70. Goonde DB, Ayana NG. Genetic diversity and character association for yield and yield related traits in soybean (Glycine max L.) genotypes. J Agric Sci Food Res. 2021;2(1):1–5.
- 71. Yirga M, Sileshi Y, Tesfaye A, Hailemariam M. Genetic variability and association of traits in soybean (Glycine max (L.)) genotypes in Ethiopia. Ethiop J Crop Sci. 2022;9(2):49–74.
- 72. Jandong EA, Uguru MI, Okechukwu EC. Estimates of genetic variability, heritability and genetic advance for agronomic and yield traits in soybean (Glycine max L.). African J Biotechnol. 2020;19(4):201–6.
- 73.
Akter S. Character association and diversity analysis of different genotypes of soybean Glycine max (L.) Merr. Dhaka-1207, Bangladesh: Sher-e-Bangla Agricultural University; 2021.
- 74. Aondover S, Lekan BL, Terkimbi V. Correlation, path coefficient and principal component analysis of seed yield in soybean genotypes. Int J Adv Res. 2013;1(7):1–5.
- 75. Antwi-boasiako A. Screening of soybean Glycine max (L.) Merrill genotypes for resistance to lodging and pod shattering. Int J Agron Agric Res. 2017;10(5):1–8.
- 76. Bello LL, Shaahu A, Vange T. Studies on relationship between seed yield and yield components in soybean Glycine max (L.) Merrill. Electron J Plant Breed. 2012;3(4):1012–7.
- 77.
Njoroge JN, Njeru PNM. Performance of soybean genotypes evaluated for yield and protein content in Nakuru County, Kenya. In: Joint Proceedings of the 27th Soil Science Society of East Africa and the 6th African Soil Science Society Conference, 2013. 255–61.
- 78. Hizli H, Çubukçu P, Şahar AK. Path analysis, genetic variability and correlation studies of related characters for forage soybean (Glycine max (L.) Merill). OKÜ Fen Bil Ens Dergisi ((OKU J Nat App Sci). 2023;6(2):1513–28.
- 79. Jolliffe IT, Cadima J. Principal component analysis: a review and recent developments. Philos Trans A Math Phys Eng Sci. 2016;374(2065):20150202. pmid:26953178
- 80. Vijayakumar E, Sudhagar R, Vanniarajan C, Ramalingam J, Allan V, Senthil N. Estimating the breeding potency of a soybean core set. Int J Agric Biol. 2022;27(3):184–96.
- 81. Verma V, Shrivastava MK, Mehra S, Amrate PK, Yadav RB. Estimation of genetic parameters for yield associated traits and principal component in advance breeding lines of soybean Glycine max. (L.) Merrill. Int J Curr Microbiol Appl Sci. 2021;10(1):2704–10.
- 82. Adetiloye IS, Ariyo OJ, Alamu O, Osewa SO. Agronomic potential and genetic diversity of 43 accession of tropical soybean Glycine max (L) Merrill. Int J Plant Res. 2020;10:33–9.
- 83. Sivabharathi RC, Muthuswamy A, Anandhi K, Karthiba L. Genetic diversity studies of soybean Glycine max (L.) Merrill germplasm accessions using cluster and principal component analysis. Legum Res - an Int J. 2023;1:1–6.
- 84. Vianna VF, Unêda-Trevisoli SH, Desidério JA, Santiago SD, Charnai K, Júnior JAF. The multivariate approach and influence of characters in selecting superior soybean genotypes. African J Agric Res. 2013;8(30):4162–9.
- 85. Leite WS, Unêda-Trevisoli SH, Silva FM, Silva AJ, Di Mauro AO. Identification of superior genotypes and soybean traits by multivariate analysis and selection index. Rev Cienc Agron. 2018;49(3):491–500.
- 86. Ghiday T, Sentayehu A. Genetic divergence analysis on some soybean Glycine max (L.) Merrill genotypes grown in Pawe, Ethiopia. Am J Agric Environ Sci. 2015;15(10):1927–33.
- 87. Khan M, Karim M, Haque M, Karim A, Mian M. Variations in agronomic traits of soybean genotypes. SAARC J Agric. 2015;12(2):90–100.
- 88. Alina MM, Kingstone M. Genetic differentiation of ARC soybean [Glycine Max (L.) Merrill] accessions based on agronomic and nutritional quality traits. Agricul Food Sci Res. 2018;5(1):6–22.
- 89. El-Hashash EF. Genetic diversity of soybean yield based on cluster and principal component analyses. J Adv Biol Biotechnol. 2016;10(3):1–9.
- 90. Mahbub MM, Rahman MM, Hossain MS, Nahar L, Shirazy BJ. Morphophysiological variation in soybean Glycine max (L.) Merrill. Am J Agric Environ Sci. 2016;16(2):234–8.
- 91. Liu Y, Li Y, Zhou G, Uzokwe N, Chang R, Chen S, et al. Development of soybean EST-SSR markers and their use to assess genetic diversity in the subgenus soja. Agricul Sci China. 2010;9(10):1423–9.
- 92. Gupta S, Manjaya J. Genetic diversity and population structure of Indian soybean [Glycine max (L.) Merr.] revealed by simple sequence repeat markers. J Crop Sci Biotechnol. 2017;20(3):221–31.
- 93. Nawaz MA, Lin X, Chan T-F, Lam H-M, Baloch FS, Ali MA, et al. Genetic architecture of wild soybean (Glycine soja Sieb. and Zucc.) populations originating from different East Asian regions. Genet Resour Crop Evol. 2021;68(4):1577–88.
- 94. Chiemeke FK, Olasanmi B, Agre PA, Mushoriwa H, Chigeza G, Abebe AT. Genetic diversity and population structure analysis of soybean [Glycine max (L.) Merrill] genotypes using agro-morphological traits and SNP markers. Genes (Basel). 2024;15(11):1373. pmid:39596572
- 95. Yohane EN, Shimelis H, Laing M, Shayanowako A. Genetic diversity and grouping of pigeonpea [Cajanus cajan Millspaugh] germplasm using SNP markers and agronomic traits. PLoS One. 2022;17(11):e0275060. pmid:36327283
- 96. Naflath TV, Rajendra PS, Ravikumar RS. Population structure and genetic diversity characterization of soybean for seed longevity. PLoS One. 2022;17(12):e0278631. pmid:36472991
- 97. Chander S, Garcia-Oliveira AL, Gedil M, Shah T, Otusanya GO, Asiedu R, et al. Genetic diversity and population structure of soybean lines adapted to Sub-Saharan Africa using single nucleotide polymorphism (SNP) markers. Agronomy. 2021;11(3):604.
- 98. Kumar D, Chhokar V, Sheoran S, Singh R, Sharma P, Jaiswal S, et al. Characterization of genetic diversity and population structure in wheat using array based SNP markers. Mol Biol Rep. 2020;47(1):293–306. pmid:31630318
- 99. Brhane H, Hammenhag C. Genetic diversity and population structure analysis of a diverse panel of pea (Pisum sativum). Front Genet. 2024;15:1396888. pmid:38873115
- 100. Nkhata W, Shimelis H, Melis R, Chirwa R, Mzengeza T, Mathew I, et al. Population structure and genetic diversity analyses of common bean germplasm collections of East and Southern Africa using morphological traits and high-density SNP markers. PLoS One. 2020;15(12):e0243238. pmid:33338076
- 101. Fatokun C, Girma G, Abberton M, Gedil M, Unachukwu N, Oyatomi O, et al. Genetic diversity and population structure of a mini-core subset from the world cowpea (Vigna unguiculata (L.) Walp.) germplasm collection. Sci Rep. 2018;8(1):16035. pmid:30375510
- 102. Rani R, Raza G, Tung MH, Rizwan M, Ashfaq H, Shimelis H, et al. Genetic diversity and population structure analysis in cultivated soybean (Glycine max [L.] Merr.) using SSR and EST-SSR markers. PLoS One. 2023;18(5):e0286099. pmid:37256876
- 103. Yirgu M, Kebede M, Feyissa T, Lakew B, Woldeyohannes AB, Fikere M. Single nucleotide polymorphism (SNP) markers for genetic diversity and population structure study in Ethiopian barley (Hordeum vulgare L.) germplasm. BMC Genom Data. 2023;24(1):7. pmid:36788500
- 104. Sodedji FAK, Agbahoungba S, Agoyi EE, Kafoutchoni MK, Choi J, Nguetta S-PA, et al. Diversity, population structure, and linkage disequilibrium among cowpea accessions. Plant Genome. 2021;14(3):e20113. pmid:34275189
- 105. Carpentieri-Pípolo V, Pípolo AE, Silva FAM da, Petek MR. Soybean parent selection based on genetic diversity. Braz arch biol technol. 2000;43(3):295–300.
- 106. Miller MJ, Song Q, Fallen B, Li Z. Genomic prediction of optimal cross combinations to accelerate genetic improvement of soybean (Glycine max). Front Plant Sci. 2023;14:1171135. pmid:37235007
- 107. Lawson DJ, Falush D. Population identification using genetic data. Annu Rev Genomics Hum Genet. 2012;13:337–61. pmid:22703172
- 108.
Pham T. Analyses of genetic diversity and desirable traits in sesame (Sesamum indicum L., Pedaliaceae) implication for breeding and conservation. Swedish University of Agricultural Sciences; 2011.
- 109. Dey SS, Singh AK, Chandel D, Behera TK. Genetic diversity of bitter gourd (Momordica charantia L.) genotypes revealed by RAPD markers and agronomic traits. Sci Hortic. 2006;109:21–8.
- 110. Uba CU, Oselebe HO, Tesfaye AA, Abtew WG. Genetic diversity and population structure analysis of bambara groundnut (Vigna subterrenea L) landraces using DArT SNP markers. PLoS One. 2021;16(7):e0253600. pmid:34197522
- 111. Fu Y-B, Cober ER, Morrison MJ, Marsolais F, Peterson GW, Horbach C. Patterns of genetic variation in a soybean germplasm collection as characterized with genotyping-by-sequencing. Plants (Basel). 2021;10(8):1611. pmid:34451656