Citation: Obua, T.; Sserumaga, J.P.; Tukamuhabwa, P.; Namara, M.; Awio, B.; Mugarra, J.; Tusiime, G.; Chigeza, G. Unravelling Yield and Yield-Related Traits in Soybean Using GGE Biplot and Path Analysis. Agronomy 2024, 14, 2826. https://doi.org/10.3390/ agronomy14122826 Academic Editor: Aiming Qi Received: 23 September 2024 Revised: 24 November 2024 Accepted: 25 November 2024 Published: 27 November 2024 Copyright: © 2024 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/). Article Unravelling Yield and Yield-Related Traits in Soybean Using GGE Biplot and Path Analysis Tonny Obua 1,* , Julius Pyton Sserumaga 2 , Phinehas Tukamuhabwa 1, Mercy Namara 1, Bruno Awio 1, Johnson Mugarra 1 , Geoffrey Tusiime 1 and Godfree Chigeza 3 1 School of Agricultural Sciences, College of Agricultural and Environmental Sciences, Makerere University, Kampala P.O. Box 7062, Uganda; phinehas.tukamuhabwa@mak.ac.ug (P.T.); mercynamara@gmail.com (M.N.); awiobruno14@gmail.com (B.A.); mugarrajohnson4@gmail.com (J.M.); gwtusiime@gmail.com (G.T.) 2 National Agricultural Research Organization, National Livestock Resources Research Institute, Kampala P.O. Box 5704, Uganda; j.serumaga@gmail.com 3 IITA Southern Africa Research and Administrative Hub (SARAH) Campus, Chelston Lusaka P.O. Box 310142, Zambia; g.chigeza@cgiar.org * Correspondence: obuatonny@gmail.com; Tel.: +256-779615960 Abstract: Soybean (Glycine max) is a vital crop for food, animal feed, and industrial products. However, its yield performance is significantly affected by genotype-by-environment interaction (GEI), which complicates the selection of high-yielding, stable varieties. This study aimed to evaluate the yield performance and stability of 12 elite soybean varieties across five major production areas in Uganda using GGE biplot and path analysis. The varieties were planted in a randomized complete block design with three replications over two consecutive seasons. Results revealed significant differences in grain yield among the varieties, locations, and their interactions (p < 0.001). The highest-yielding varieties were Maksoy 5N (979 kg ha−1), Maksoy 4N (978 kg ha−1), Maksoy 3N (930 kg ha−1), and Signal (930 kg ha−1). GGE biplot analysis grouped the locations into two mega- environments, with the Maksoy varieties exhibiting greater yield stability compared to Seed Co. varieties. Path analysis showed that traits such as the number of lower internodes, central internode length, and filled pods had the highest positive direct effects on grain yield. This study provides insights into soybean breeding in tropical environments, highlighting traits that can be targeted to improve yield and stability. The findings offer a framework for breeding programs in Uganda and similar agro-ecological regions, promoting more resilient and productive soybean varieties. This study also illustrated the potential advantages of employing more complex mathematical techniques like path analysis to uncover yield and yield-related traits in soybean breeding programs. Keywords: biplot; correlation; GEI; GGE biplots; path analysis; yield stability 1. Introduction Soybean (Glycine max) is one of the most widely grown and used oilseeds [1]. Soybean uses range from human food (19.2%) to animal feed (77%) and industrial products (3.8%) [2]. Soybean production has expanded significantly over the last two decades [3]. The greatest soybean production increase occurred in the latter half of the 20th century, when it increased tenfold, from 27 million tons (MT) in 1962 to 349 MT in 2022 [4]. The rapid increase in production is because of the ever-increasing demand for soybeans as a fundamental source of protein in animal feed, especially in developing and emerging countries [5,6]. Furthermore, it is anticipated that 2030 global soybean production will reach 371 MT, an annual growth of 2.2% [7]. Nutritionally, soybean grain has 40% protein, 20% oil, high-calorie content, and several vital amino acids [8]. Soybeans form root symbioses with rhizobia that fix atmospheric nitrogen in the soil, thereby improving soil fertility [9]. In tropical Africa, characterized by infertile soils and limited fertilizer use, soybean is a suitable crop for sustainable agricultural production systems due to their capacity to fix Agronomy 2024, 14, 2826. https://doi.org/10.3390/agronomy14122826 https://www.mdpi.com/journal/agronomy https://doi.org/10.3390/agronomy14122826 https://doi.org/10.3390/agronomy14122826 https://creativecommons.org/ https://creativecommons.org/licenses/by/4.0/ https://creativecommons.org/licenses/by/4.0/ https://www.mdpi.com/journal/agronomy https://www.mdpi.com https://orcid.org/0000-0003-1134-3255 https://orcid.org/0000-0001-5152-2835 https://orcid.org/0000-0001-5809-1177 https://doi.org/10.3390/agronomy14122826 https://www.mdpi.com/journal/agronomy https://www.mdpi.com/article/10.3390/agronomy14122826?type=check_update&version=2 Agronomy 2024, 14, 2826 2 of 14 atmospheric nitrogen into the soil [10]. Therefore, there is minimum need for nitrogen fertilizers in such production systems [9]. Soybean is grown as both a food and a cash crop [11]. Therefore, in tropical Africa, it is necessary to promote soybean as a cash crop. Several outstanding soybean varieties that are widely cultivated throughout Africa have been developed for more than 20 years by Seed Co. (Sandton, South Africa) and Makerere University (Kampala, Uganda). However, little is known about the agronomic traits, yield components, and relationships between the yield components of these soybean varieties. Additionally, contrasting information has been reported on the most important yield component that contributes to the total grain yield [12–17]. The complex trait of soybean grain yield is largely affected by genotype-by-environment interaction (GEI) [14]. A crossover form of GEI for grain yield was found at four locations in Ethiopia during a multi- environment study comprising 32 genotypes. At four locations in Ethiopia, a crossover form of GEI for grain yield was found during a multi-environment study comprising 32 genotypes. The same study found three genotypes that showed excellent grain yield and stability in performance across the test settings [18]. Similarly, W9837 x Cikuray-66 was the most stable and high-yielding genotype in another multi-environment study that tested 10 black-seeded soybean genotypes at sixteen locations [19]. Another study in Zambia identified TGX 1988-22F as the best genotype for general adaptability because of its great adaptability across all locations [20]. Additionally, yield results from the association of different yield-related traits may directly or indirectly affect yield [14,21]. Designing an effective plant breeding program requires an accurate estimation of the relationship between yield-related traits and their effects on the total yield. Several studies have reported a correlation between the yield of soybeans and phenotypic characteristics, including the number of branches, length of internode, days to 50% flowering, days to physiological maturity, and plant height [22,23]. Additionally, several studies have shown a correlation between yield-related characteristics, including the quantity of seeds per plant, the number of pods per plant, the number of seeds per pod, and the weight of 100 seeds [24,25]. A simple correlation analysis between two traits is commonly performed using a linear correlation coefficient. However, a simple correlation analysis might not accurately depict the cause-and-effect relationship between yield and associated traits, which could lead to inefficiencies in the selection process [26]. To assess whether there is a cause-and-effect relationship between two variables, path analysis partitions the correlation coefficients between the variables, making it more accurate than correlation analysis [22,27]. As far as we are aware, no comprehensive study has been conducted to estimate GEI for yield and relationships among different yield-related traits on the two sets of soybean varieties developed by Makerere University and Seed Co. This study seeks to establish and quantify genotype × environment interactions and relationships to identify indirect criteria for grain yield selection in soybeans. Therefore, the purpose of this study was to evaluate the yield performance and stability of 12 elite soybean varieties in Uganda’s major soybean-growing areas and to determine the relationship between related traits and grain yield. 2. Materials and Methods 2.1. Experimental Materials There were 12 soybean genotypes in all; five were developed by the Makerere Univer- sity Centre for Soybean Improvement and Development (MAKCSID), while the remaining seven were developed by Seed Co. (Table 1). Every experimental material used in this study is a released soybean variety that is widely cultivated in several African countries. All the experimental materials were evaluated in Uganda in a multi-environment trial. Agronomy 2024, 14, 2826 3 of 14 Table 1. Characteristics of the soybean genotypes used in the study. Genotype Origin 1. Maksoy 1N Makerere University 2. Maksoy 2N Makerere University 3. Maksoy 3N Makerere University 4. Maksoy 4N Makerere University 5. Maksoy 5N Makerere University 6. Saga Seed Co. 7. Saxon Seed Co. 8. Sentinel Seed Co. 9. Sequel Seed Co. 10. Signal Seed Co. 11. Squire Seed Co. 12. Status Seed Co. 2.2. Description of the Study Locations Five locations that represent the main soybean-growing regions in Uganda were used for the soybean grain yield trial (Table 2). Three locations are found in Lake Victoria Cres- cent: Namulonge, Kabanyolo, and Nakabango; Ngetta is in the north-western savannah grasslands; and Mubuku in the western medium-high farmland. The varying climatic conditions in these locations have a significant effect on the grain yield of soybean (Table 2). In this study, the Mubuku irrigation scheme was selected to evaluate the adaptability of soybean genotypes to irrigated field conditions. Table 2. Description of the five locations used in this study. Location Region Coordinates Altitude (masl) Mean Annual Temperature (◦C) Mean Annual Rainfall (mm) Namulonge Central 0o32′ N/32o37′ E 1160 22.6 1400 Kabanyolo Central 0o28′ N/32o36′ E 1180 21.4 1234 Nakabango Eastern 0o29′ N/33◦14′ E 1210 22.8 1400 Ngetta Northern 2◦17′ N/32◦56′ E 1103 24.7 1200 Mubuku Western 0o13′ N/30o08′ E 1007 27.8 750 Source: Meteorological station data obtained from each location; masl = meters above sea level. 2.3. Experimental Design The soybean genotypes were planted using a randomized complete block design (RCBD) with three replications. Three 5 m long rows, with 60 cm between rows and 5 cm between plants within a row, were used to represent each soybean genotype. Two seasons were used in this trial: the first rainfall season of 2016 (2016 A) and the second rainfall season of 2016 (2016 B). Three weedings were performed each season to keep the trials free of weeds. In these trials, no agrochemicals were used to manage the pests. 2.4. Data Collection Several yield-related traits were collected from the soybean varieties during the grow- ing season. These included the number of branches, plant height, pod clearance, internode length, number of internodes, days to 50% flowering, and days to physiological maturity. Additional yield-related traits were recorded at harvest, such as yield, 100-seed weight, and the number of full pods per plant. 2.4.1. Days to Flowering (DTF) and Days to Maturity (DTM) Days to flowering was determined by counting the number of days from the date of planting until 50% of the plants in a plot flowered. Days to maturity was determined by counting the number of days between the planting date and the day that the plants in each plot reached physiological maturity. Agronomy 2024, 14, 2826 4 of 14 2.4.2. Plant Height, Pod Clearance, and Number of Branches At 85 days after planting, the number of primary (NPB) and secondary branches (NSB), plant height (PH), and pod clearance were recorded. Using a ruler, the height of the plant was measured from the ground to the top of the main stem axis, and the pod clearance was measured from the ground to the first pods. The number of branches per plant was counted. Ten randomly chosen plants in each plot were examined for these characteristics, and the mean was determined. 2.4.3. Length and Number of Internodes Each sampled individual plant was divided into three sections: lower, center, and upper. A meter ruler was used to measure the length of each section, and the number of internodes was counted for all ten randomly selected plants. 2.4.4. Yield and Yield-Related Traits During harvesting, the number of mature-filled pods per plant was counted to de- termine the number of pods per plant. The number of seeds per pod was counted from the ten randomly selected soybean plants, and a mean was calculated for every plot. The number of seeds per pod, 100-seed weight, and plot yield were determined for each plot. To calculate the pod yield per plot, the harvested soybean plants were threshed, and the seeds were sun-dried to about 10% moisture content and weighed. The plot yields were then converted to yield in Kg Ha−1. 2.5. Data Analysis 2.5.1. Multi-Locational Yield Data The agricolae R package was used to conduct analysis of variance (ANOVA) for grain yield independently for each location and combined across locations [28]. Genotypes were treated as fixed effects, and replications and blocks within them as random effects. To test genotype variations and the presence of GEI, variances were separated into pertinent sources of variation for the combined analysis. To dissect the GEI of each variable and evaluate genotype stability in performance in different conditions, GGE biplot analysis was applied to the adjusted grain yield obtained via ANOVA [29]. Soybean grain yield in each test location was assessed using the ggplot function in tidyverse package in R software, Version 4.4.1 [30]. Mean separation was performed using the Least Significant Difference (LSD) test using the LSD.test function from agricolae package in R software, Version 4.4.1 [30]. 2.5.2. Correlation Analysis Correlation analysis was carried out for the different yield-related traits among both the Seed Co. and Maksoy varieties. The Pearson technique and the corrplot package version 0.88 of the R software were used to conduct the correlation analysis [31]. The correlation coefficients were categorized as follows: weak (0.0–0.3), moderate (0.3–0.6), strong (0.6–0.9), and very strong (0.9–1) [14]. 2.5.3. Path Analysis The direct and indirect effects of yield-related traits on the grain yield of the 12 soybean genotypes were determined by path analysis. The path analysis considered grain yield as a response/dependent variable and all the studied traits as predictive/independent variables. Estimation and fractionation of the effects of yield-related traits were performed with agricolae package version 1.3-5 in R software [28]. The direct and indirect effects of the yield-related traits were classified as negligible (0.00–0.09), low (0.1–0.19), moderate (0.2–0.29), and high (0.3–0.99) as described by [32]. Agronomy 2024, 14, 2826 5 of 14 3. Results 3.1. Grain Yield Performance of Genotypes Across Locations The grain yield performance varied significantly across the 12 soybean genotypes, locations, and their interactions (p < 0.001) (Table 3). Among the tested varieties, the top performers were the Maksoy lines, with Maksoy 5N achieving the highest grain yield of 979 kg ha−1, followed by Maksoy 4N (978 kg ha−1) and Maksoy 3N (930 kg ha−1). The best Seed Co. variety, Signal, produced 930 kg ha−1, but other Seed Co. varieties like Sentinel had considerably lower yields (656 kg ha−1) (Table 4). Location-wise, Ngetta showed the highest mean grain yield of 1256 kg ha−1, followed by Mubuku (1107 kg ha−1). Namulonge recorded the lowest grain yield among the five locations, averaging 406 kg ha−1 (Table 4). Table 3. Analysis of variance of 12 soybean genotypes tested for grain yield in Uganda across two seasons and five locations. Source of Variation DF Sum Sq Mean Sq F Value F Pr. Season 1 5,262,177 5,262,177 84.68 *** Genotype 11 3,408,556 309,869 4.99 *** Location 4 32,565,992 8,141,498 131.01 *** Genotype × Season 11 1,275,954 115,996 1.87 ns Genotype × Location 44 9,570,005 217,500 3.5 *** Location × Season 4 15,471,863 3,867,966 62.24 *** Genotype × Location × Season 44 2,585,386 58,759 0.95 ns Rep 2 51,681 25,840 0.42 Residual 232 14,417,735 62,145 Total 353 83,847,837 *** indicate significance at α ≤ 0.001 respectively. ns = not significant. Table 4. Grain yield (kg ha−1) of 12 soybean genotypes tested in five locations and two seasons in Uganda. Genotype Kabanyolo Mubuku Nakabango Namulonge Ngetta Mean Maksoy 1N 666 1067 859 286 1421 860 abc Maksoy 2N 686 947 689 570 1234 825 abc Maksoy 3N 729 1242 808 523 1349 930 bc Maksoy 4N 774 1187 699 459 1771 978 c Maksoy 5N 739 1094 1088 406 1569 979 c Saga 809 1342 765 379 1142 887 abc Saxon 763 1237 726 378 1031 827 abc Sentinel 646 844 648 362 906 681 a Sequel 674 1262 507 212 804 692 a Signal 675 989 787 526 1672 930 bc Squire 816 1195 770 343 599 745 ab Status 636 873 844 428 1571 870 abc Mean 718 1107 766 406 1256 850 Genotype means followed by different letter(s) are significantly different using LSD (p < 0.05). 3.2. Genotype Yield Stability Using GGE Biplot Analysis The GGE biplot analysis (Figure 1) explained 74.01% of the total GEI variation, identi- fying Maksoy 5N as the highest-yielding genotype across all environments and showing some instability. Maksoy 3N was the most stable variety, consistently performing well across environments, making it ideal for broad adaptation. On the other hand, Sequel and Sentinel had low yields and were unstable across the test environments. Agronomy 2024, 14, 2826 6 of 14 Agronomy 2024, 14, 2826 6 of 15 3.2. Genotype Yield Stability Using GGE Biplot Analysis The GGE biplot analysis (Figure 1) explained 74.01% of the total GEI variation, identifying Maksoy 5N as the highest-yielding genotype across all environments and showing some instability. Maksoy 3N was the most stable variety, consistently performing well across environments, making it ideal for broad adaptation. On the other hand, Sequel and Sentinel had low yields and were unstable across the test environments. Figure 1. GGE Biplot for combined analysis of 12 soybean genotypes tested for grain yield over two seasons at five locations. For broad adaptation in the target locations, the ideal genotype should have high mean performance and stability. In the biplot, an ideal genotype is closer to the origin. Thus, in the current study, the GGE biplot illustrated that Maksoy 5N was the most ideal genotype, followed by Maksoy 4N and Maksoy 3N (Figure 2). Conversely, less stable genotypes and low yield are undesirable since they are located distant from the ideal genotypes. The most undesirable variety was Sequel, followed by Sentinel (Figure 2). Figure 1. GGE Biplot for combined analysis of 12 soybean genotypes tested for grain yield over two seasons at five locations. For broad adaptation in the target locations, the ideal genotype should have high mean performance and stability. In the biplot, an ideal genotype is closer to the origin. Thus, in the current study, the GGE biplot illustrated that Maksoy 5N was the most ideal genotype, followed by Maksoy 4N and Maksoy 3N (Figure 2). Conversely, less stable genotypes and low yield are undesirable since they are located distant from the ideal genotypes. The most undesirable variety was Sequel, followed by Sentinel (Figure 2). Agronomy 2024, 14, 2826 7 of 15 Figure 2. GGE Biplot for combined analysis of 12 soybean genotypes tested for grain yield over two seasons at five locations. The biplot analysis also grouped the locations into two mega-environments. Saga performed best in Kabanyolo and Mubuku, while Maksoy 5N was superior in Nge a and Nakabango and Namulonge (Figure 3). Figure 3. GGE Biplot for combined analysis of 12 soybean genotypes tested for grain yield over two seasons at five locations. Figure 2. GGE Biplot for combined analysis of 12 soybean genotypes tested for grain yield over two seasons at five locations. Agronomy 2024, 14, 2826 7 of 14 The biplot analysis also grouped the locations into two mega-environments. Saga performed best in Kabanyolo and Mubuku, while Maksoy 5N was superior in Ngetta and Nakabango and Namulonge (Figure 3). Agronomy 2024, 14, 2826 7 of 15 Figure 2. GGE Biplot for combined analysis of 12 soybean genotypes tested for grain yield over two seasons at five locations. The biplot analysis also grouped the locations into two mega-environments. Saga performed best in Kabanyolo and Mubuku, while Maksoy 5N was superior in Nge a and Nakabango and Namulonge (Figure 3). Figure 3. GGE Biplot for combined analysis of 12 soybean genotypes tested for grain yield over two seasons at five locations. Figure 3. GGE Biplot for combined analysis of 12 soybean genotypes tested for grain yield over two seasons at five locations. The use of GGE biplot analysis to determine the ideal locations for multi-environmental trials (METs) is an intriguing application. An ideal testing environment should be represen- tative and discriminating. Since Nakabango was closest to the center, it was identified as the ideal test location (Figure 4). Mubuku was not an ideal location because it was far away from the center (Figure 4). Agronomy 2024, 14, 2826 8 of 15 The use of GGE biplot analysis to determine the ideal locations for multi- environmental trials (METs) is an intriguing application. An ideal testing environment should be representative and discriminating. Since Nakabango was closest to the center, it was identified as the ideal test location (Figure 4). Mubuku was not an ideal location because it was far away from the center (Figure 4). Figure 4. A GGE Comparison biplot displaying performance and stability of 12 soybean genotypes tested in five locations and two seasons concerning grain yield. 3.3. Correlations Between Grain Yield and Yield-Related Traits Correlation analysis revealed significant positive associations between grain yield and several yield-related traits (Figure 5). The number of filled pods per plant (r = 0.84), the number of lower internodes (r = 0.59), and the hundred-seed weight (r = 0.61) had the strongest positive correlations with yield (p < 0.001). Traits like plant height (r = 0.28) and the number of branches (r = 0.28) also showed positive associations with grain yield, though to a lesser degree. Conversely, grain yield had a negative correlation with pod clearance (r = −0.44), indicating that higher pod clearance reduced yield (p < 0.01). Figure 4. A GGE Comparison biplot displaying performance and stability of 12 soybean genotypes tested in five locations and two seasons concerning grain yield. Agronomy 2024, 14, 2826 8 of 14 3.3. Correlations Between Grain Yield and Yield-Related Traits Correlation analysis revealed significant positive associations between grain yield and several yield-related traits (Figure 5). The number of filled pods per plant (r = 0.84), the number of lower internodes (r = 0.59), and the hundred-seed weight (r = 0.61) had the strongest positive correlations with yield (p < 0.001). Traits like plant height (r = 0.28) and the number of branches (r = 0.28) also showed positive associations with grain yield, though to a lesser degree. Conversely, grain yield had a negative correlation with pod clearance (r = −0.44), indicating that higher pod clearance reduced yield (p < 0.01). Agronomy 2024, 14, 2826 9 of 15 Figure 5. A corrplot displaying the distribution and association of the studied traits. Flowering = number of days to 50% flowering, maturity = number of days to physiolgical maturity, PH = plant height, PC = pod clearance, FP = number of filled pods per plant, SPP = number of pods per plant, branches = number of branches per plant, LLI = length of lower internodes, LCI = length of central internodes, LUI = length of upper internodes, NLI = number of lower internodes, NCI = number of central internodes, NUI = number of upper internodes, HSW = 100 seed weight (g), yield = grain yield in Kg per hectare. *, **, and *** indicate significance at p ≤ 0.05, p ≤ 0.01, and p ≤ 0.001, respectively; values without * are not significant at p < 0.05. 3.4. Path Analysis of Yield-Related Traits Path analysis (Table 5 and Figure 6) identified the number of lower internodes as having the highest positive direct effect on grain yield (373.08), followed by the length of the central internodes (54.08) and the number of filled pods per plant (48.65). These traits represent the most critical factors for direct selection in breeding programs aimed at increasing soybean grain yields. Negative direct effects on yield were observed for the number of central internodes (−200.74) and the length of lower internodes (−91.71), suggesting that excessive internode growth may reduce yield. The number of days to physiological maturity also had a small negative direct effect on yield (−10.88), indicating that earlier maturing varieties may offer yield advantages. Table 5. Estimates of direct and indirect effects of the different traits on grain yield of soybean varieties tested. Response Trait Estimate Std.Err Z-Value p (>|z|) Std.Lv Std.All Flowering 1.206 8.280 0.146 0.884 1.206 0.012 Maturity −10.884 3.736 −2.913 0.004 −10.884 −0.250 PH 7.159 3.400 2.105 0.035 7.159 0.291 Branches 49.671 25.396 1.956 0.050 49.671 0.114 FP 48.654 6.944 7.007 0.000 48.654 0.608 SPP −31.087 29.998 −1.036 0.300 −31.087 −0.054 LLI −91.707 15.286 −5.999 0.000 −91.707 −0.548 LCI 54.076 14.225 3.801 0.000 54.076 0.425 LUI −8.724 5.676 −1.537 0.124 −8.724 −0.220 NLI 373.075 80.081 4.659 0.000 373.075 0.423 NCI −200.742 90.325 −2.222 0.026 −200.742 −0.232 NUI −18.975 26.594 −0.714 0.476 −18.975 −0.074 HSW 13.042 6.744 1.934 0.053 13.042 0.163 Figure 5. A corrplot displaying the distribution and association of the studied traits. Flowering = number of days to 50% flowering, maturity = number of days to physiolgical maturity, PH = plant height, PC = pod clearance, FP = number of filled pods per plant, SPP = number of pods per plant, branches = number of branches per plant, LLI = length of lower internodes, LCI = length of central internodes, LUI = length of upper internodes, NLI = number of lower internodes, NCI = number of central internodes, NUI = number of upper internodes, HSW = 100 seed weight (g), yield = grain yield in Kg per hectare. *, **, and *** indicate significance at p ≤ 0.05, p ≤ 0.01, and p ≤ 0.001, respectively; values without * are not significant at p < 0.05. 3.4. Path Analysis of Yield-Related Traits Path analysis (Table 5 and Figure 6) identified the number of lower internodes as having the highest positive direct effect on grain yield (373.08), followed by the length of the central internodes (54.08) and the number of filled pods per plant (48.65). These traits represent the most critical factors for direct selection in breeding programs aimed at increasing soybean grain yields. Negative direct effects on yield were observed for the number of central internodes (−200.74) and the length of lower internodes (−91.71), suggesting that excessive internode growth may reduce yield. The number of days to physiological maturity also had a small negative direct effect on yield (−10.88), indicating that earlier maturing varieties may offer yield advantages. Agronomy 2024, 14, 2826 9 of 14 Table 5. Estimates of direct and indirect effects of the different traits on grain yield of soybean varieties tested. Response Trait Estimate Std.Err Z-Value p (>|z|) Std.Lv Std.All Flowering 1.206 8.280 0.146 0.884 1.206 0.012 Maturity −10.884 3.736 −2.913 0.004 −10.884 −0.250 PH 7.159 3.400 2.105 0.035 7.159 0.291 Branches 49.671 25.396 1.956 0.050 49.671 0.114 FP 48.654 6.944 7.007 0.000 48.654 0.608 SPP −31.087 29.998 −1.036 0.300 −31.087 −0.054 LLI −91.707 15.286 −5.999 0.000 −91.707 −0.548 LCI 54.076 14.225 3.801 0.000 54.076 0.425 LUI −8.724 5.676 −1.537 0.124 −8.724 −0.220 NLI 373.075 80.081 4.659 0.000 373.075 0.423 NCI −200.742 90.325 −2.222 0.026 −200.742 −0.232 NUI −18.975 26.594 −0.714 0.476 −18.975 −0.074 HSW 13.042 6.744 1.934 0.053 13.042 0.163 Agronomy 2024, 14, 2826 10 of 15 Figure 6. Cause and effect relationships among the predictor traits associated with grain yield. Flw = number of days to 50% flowering, Mtr = number of days to physiological maturity, PH = plant height, PC = pod clearance, FP = number of filled pods per plant, SPP = number of pods per plant, Brn = number of branches per plant, LLI = length of lower internodes, LCI = length of central internodes, LUI = length of upper internodes, NLI = number of lower internodes, NCI = number of central internodes, NUI = number of upper internodes, HSW = 100 seed weight (g), Yld = grain yield in Kg per hectare. Green arrows = positive effects, Red arrows = negative effects. 4. Discussion 4.1. Grain Yield Performance of Genotypes Across Locations This study demonstrated significant genotype-by-environment interactions (GEIs) among the 12 soybean genotypes tested across five locations in Uganda, underscoring the importance of selecting varieties that are well-adapted to specific environments. The top- performing genotypes, Maksoy 5N, 4N, and 3N, consistently outyielded their counterparts, with Maksoy 5N exhibiting the highest grain yield (902 kg ha−1) across all environments. This superior performance reflects their local adaptation to the Ugandan climate, which includes periods of erratic rainfall, high temperatures, and varying soil fertility. In contrast, Seed Co varieties, which are bred primarily for Southern African climates, showed less stability and lower yields in most Ugandan environments. Sentinel, for example, had the lowest yield among all genotypes (656 kg ha−1), likely due to its reduced adaptability to warmer, more variable conditions. These findings support previous research showing that genotypes bred for specific regions tend to perform be er under local environmental stresses [18,19]. As climate change intensifies, resulting in greater temperature fluctuations and unpredictable rainfall, locally bred varieties like the Maksoy lines may become even more critical for sustaining soybean production in Uganda. The differences in performance across locations can also be a ributed to local environmental conditions. For instance, Nge a and Mubuku, which showed the highest mean yields, benefi ed from irrigation and relatively stable water availability, minimizing the impact of drought stress. Similar results have been observed in other multi- environment trials, where stable water supply or irrigation significantly improved yields under otherwise challenging conditions [20,33]. As climate change alters rainfall pa erns, the integration of irrigation or water management practices may be crucial for maintaining high soybean yields in Uganda. Figure 6. Cause and effect relationships among the predictor traits associated with grain yield. Flw = number of days to 50% flowering, Mtr = number of days to physiological maturity, PH = plant height, PC = pod clearance, FP = number of filled pods per plant, SPP = number of pods per plant, Brn = number of branches per plant, LLI = length of lower internodes, LCI = length of central internodes, LUI = length of upper internodes, NLI = number of lower internodes, NCI = number of central internodes, NUI = number of upper internodes, HSW = 100 seed weight (g), Yld = grain yield in Kg per hectare. Green arrows = positive effects, Red arrows = negative effects. 4. Discussion 4.1. Grain Yield Performance of Genotypes Across Locations This study demonstrated significant genotype-by-environment interactions (GEIs) among the 12 soybean genotypes tested across five locations in Uganda, underscoring the importance of selecting varieties that are well-adapted to specific environments. The top- performing genotypes, Maksoy 5N, 4N, and 3N, consistently outyielded their counterparts, with Maksoy 5N exhibiting the highest grain yield (902 kg ha−1) across all environments. This superior performance reflects their local adaptation to the Ugandan climate, which includes periods of erratic rainfall, high temperatures, and varying soil fertility. In contrast, Seed Co varieties, which are bred primarily for Southern African climates, showed less stability and lower yields in most Ugandan environments. Sentinel, for ex- ample, had the lowest yield among all genotypes (656 kg ha−1), likely due to its reduced adaptability to warmer, more variable conditions. These findings support previous research Agronomy 2024, 14, 2826 10 of 14 showing that genotypes bred for specific regions tend to perform better under local envi- ronmental stresses [18,19]. As climate change intensifies, resulting in greater temperature fluctuations and unpredictable rainfall, locally bred varieties like the Maksoy lines may become even more critical for sustaining soybean production in Uganda. The differences in performance across locations can also be attributed to local envi- ronmental conditions. For instance, Ngetta and Mubuku, which showed the highest mean yields, benefitted from irrigation and relatively stable water availability, minimizing the impact of drought stress. Similar results have been observed in other multi-environment tri- als, where stable water supply or irrigation significantly improved yields under otherwise challenging conditions [20,33]. As climate change alters rainfall patterns, the integration of irrigation or water management practices may be crucial for maintaining high soybean yields in Uganda. 4.2. Genotype Yield Stability Using GGE Biplot Analysis GGE biplot analysis provided valuable insights into both the yield performance and stability of the tested genotypes. While Maksoy 5N exhibited the highest overall yield, Maksoy 3N was the most stable genotype across all environments, making it a prime candidate for broad adaptability under varying environmental conditions. This stability is particularly important in the face of climate change, which is expected to increase environmental variability in tropical regions like Uganda [34]. The identification of three distinct mega-environments further emphasizes the need for targeted breeding strategies. For example, Saga, which performed best in Kabanyolo and Mubuku, may be specifically recommended for regions with similar environmental conditions, while Maksoy 4N, which excelled in Ngetta and Nakabango, should be pri- oritized in areas with similar agro-climatic profiles. These findings align with previous studies in Ethiopia and Zambia, where multiple mega-environments were identified due to the increasing unpredictability of rainfall and temperature [18,20]. The clustering of locations into different mega-environments suggests that Uganda’s agricultural landscape is becoming more segmented due to climate-driven changes. Such segmentation could guide breeding programs to develop varieties tailored to specific regions, particularly those most vulnerable to the adverse effects of climate change, such as drought-prone areas or regions experiencing increased temperature extremes. The three mega-environments identified in this study indicate that at least one of the mega-environments chosen must be used for successful soybean breeding and selection. This agrees with a study by [33], who identified three mega-environments in a MET that was conducted in eight locations in Uganda. This result, however, is contrary to findings by [35], who demonstrated that Uganda possessed two mega-environments for the yields of soybean grain. The observations from METs in Uganda in the last 10 years illustrate that initially one of two mega-environments were identified for soybean grain yield. However, this has significantly changed in the five years where more mega-environments have been identified in recent studies. This could be evidence of climate change that has affected the grain yield of soybean in Uganda. A study by [35] had Maksoy 3N as the winning cultivar in the major mega-environments; however, this observation was not the same in the current study. This could also be attributed to the effects of climate change, like different rainfall patterns during the two studies. Additionally, in the current study, the soybean varieties studied have been recently released and are superior to Maksoy 3N in terms of grain yield. In the current study, Saga is the winning cultivar in Mubuku and Kabanyolo. Therefore, this variety can be recommended for these two locations and the surrounding areas. 4.3. Correlations Between Grain Yield and Yield-Related Traits Understanding the primary factors affecting soybean grain yield can offer important insights for choosing breeding and management strategies to increase productivity. The elements of the correlation analysis results show if there is a positive or negative relation- ship among studied traits. If this produces a positive correlation value, increasing the Agronomy 2024, 14, 2826 11 of 14 first component will also increase the second component. Nevertheless, raising the first component will result in a drop in the second component if the correlation coefficient is negative [16,36]. This study revealed several important correlations between grain yield and yield- related traits, which can serve as indirect selection criteria for improving soybean produc- tivity. The strong positive correlations between grain yield and traits such as the number of filled pods per plant (r = 0.84) and hundred-seed weight (r = 0.61) suggest that breeding programs focusing on these traits are likely to enhance yield stability and performance. These results are consistent with previous studies, which have similarly found strong correlations between these traits and soybean yield [12,17]. In addition, several studies have found a significant relationship between soybean grain yield and the number of pods per plant [16,23]. Hundred-seed weight has also been reported in previous studies to be positively correlated with grain yield [16]. The following traits have also been re- ported to be positively correlated to grain yield in soybean: number of branches [16] and plant height [13]. Notably, the negative correlation between grain yield and pod clearance (r = −0.44) highlights a key factor for breeding programs in tropical regions. Lower pod clearance may help protect soybean plants from the adverse effects of drought, heavy rainfall, or high winds, which are expected to become more frequent due to climate change. In environments where moisture stress is common, selecting for traits that minimize pod exposure to environmental extremes could help mitigate yield losses. These findings em- phasize the need for breeding strategies that prioritize stress tolerance, particularly in regions increasingly affected by climate instability. 4.4. Path Analysis of Yield-Related Traits To make sound breeding and management decisions to improve grain yield, plant scientists must clearly understand the different yield-related traits in soybeans. Path analysis was created to examine and interpret relationships between a variety of traits [37,38]. Path analysis has been employed by numerous researchers to predict soybean grain production using various traits [14]. In the current study, the path analysis identified several yield- related traits with significant direct effects on grain yield, providing further insight into the genetic factors driving soybean productivity. The number of lower internodes had the highest positive direct effect on grain yield, followed by the length of the central internodes and the number of filled pods per plant. These traits should be prioritized in breeding programs aiming to enhance yield potential, as they directly contribute to improved productivity [38]. Internode number and length are affected by both genotypic and environmental factors and have been reported to be identical [39]. According to reports, higher CO2 levels and temperatures during the growing season have two effects on plant weight: they increase the number of internodes or internode length [39]. However, internode number and length are also influenced by genotypic factors [40]. According to the current study, grain yield was positively impacted by the number of branches per plant. This agrees with previous studies that have reported that the number of branches per plant has a significant direct contribution to grain yield in soybeans [38]. Grain yield was also directly and significantly impacted by the number of pods per plant. The crop will yield more if it produces more photosynthate (source) and pods (sink) [41]. The availability of enough photosynthate to meet the needs of the developing pod and seeds determines whether the flower is aborted or a pod is set [41]. The amount of photosynthate produced per plant is influenced by both environmental and genetic factors. For instance, more photosynthate is produced and more pods are established during a growing season when the conditions are favorable. The observed results suggest that soybean genotypes with a high number of pods per plant tend to allocate more photosynthate to the development of pods, hence the resultant high grain yields. These findings concur with earlier research that showed that the number of pods per plant is a significant factor in soybean crop yield. Agronomy 2024, 14, 2826 12 of 14 Hundred-seed weight had a direct and significant effect on grain yield. Seed weight is influenced by the length of the seed fill phase and the rate of seed growth [42]. Seeds with a high 100-seed weight typically have faster seed growth rates and more photosynthesis requirements [41]. The weight of the seeds is genetically controlled and largely constant during the seed-filling stage. While favorable late-season rainfall and temperatures do not impact seed number, they can enhance seed weight by extending the seed fill period. Additionally, a severe late-season drought may end the seed-filling cycle early, lower seed size, and result in a low grain yield [41]. Therefore, soybean genotypes with a high 100-seed weight generally have a high seed growth rate or longer seed-filling period or both. The results agree with earlier reports by [25] that 100-seed weight has a positive contribution to grain yield. Interestingly, the negative direct effects of the number of central internodes and the length of lower internodes suggest that excessive vegetative growth may reduce grain yield. This may be due to the trade-off between vegetative and reproductive growth, where excessive allocation of resources to stem elongation limits the energy available for pod development. These findings align with previous research that highlights the importance of balancing vegetative and reproductive growth for optimizing yield. Given the anticipated impact of climate change on temperature and moisture avail- ability, these insights are particularly relevant. Selecting genotypes with balanced growth traits that can maintain productivity under both favorable and stressful conditions will be critical for ensuring food security in the coming decades. This reinforces the need for integrating climate resilience into breeding programs, as soybean plants with optimized architecture will be better able to cope with the increasing environmental pressures [41]. 5. Conclusions The results of this study demonstrate that the Maksoy soybean varieties, specifically Maksoy 5N, 4N, and 3N, exhibited superior yield performance and stability across multiple environments in Uganda compared to Seed Co. varieties. Maksoy 5N achieved the highest mean grain yield of 902 kg ha−1, followed closely by Maksoy 4N (896 kg ha−1) and Maksoy 3N (866 kg ha−1), highlighting their adaptability to the diverse agro-ecological conditions of Uganda. These varieties also demonstrated stable performance across environments, as revealed by GGE biplot analysis, making them ideal candidates for low-input agricultural systems in sub-Saharan Africa. In contrast, Seed Co. varieties, while exhibiting higher yields under favorable condi- tions, were less stable across environments, indicating that they may be more suited for high-input commercial farming systems where optimal growing conditions can be main- tained. Path analysis revealed that traits such as the number of lower internodes, length of central internodes, and filled pods per plant had the highest direct positive effects on grain yield, providing valuable insights for future breeding efforts targeting yield improvement in soybeans. These findings emphasize the importance of selecting genotypes with both high yield potential and stability to address the challenges of genotype-by-environment interactions. The identified yield-related traits can serve as indirect selection criteria for enhancing soybean productivity in tropical environments. Moving forward, the promotion of Maksoy varieties, particularly for regions with variable climatic conditions, could significantly improve soybean yields and contribute to sustainable agricultural practices in Uganda and similar agro-ecological regions. Future developments in soybean breeding should focus on enhancing traits like the number of lower internodes, filled pods per plant, and hundred-seed weight, which were identified as key contributors to yield under variable conditions. Additionally, incorporat- ing climate resilience traits such as drought tolerance, disease resistance, and pest tolerance into breeding programs will be essential to sustain high yields as environmental stresses intensify. There is also a need for further research on how climate change will impact different agro-ecological zones in Uganda and beyond, allowing for more precise breeding Agronomy 2024, 14, 2826 13 of 14 and management recommendations. Overall, this research provides valuable insights into breeding strategies for soybeans in tropical environments, particularly in the face of climate change. By focusing on stable, high-yielding varieties like Maksoy 5N, 4N, and 3N, breeding programs can contribute to more resilient agricultural systems across sub-Saharan Africa. Author Contributions: Conceptualization, T.O., M.N. and P.T.; study design, T.O., M.N. and P.T.; data collection, T.O., M.N. and J.M.; data analysis, J.P.S., J.M. and B.A.; data interpretation, J.P.S., J.M. and B.A.; writing, T.O., J.P.S., P.T., M.N., B.A., J.M. and G.C.; literature search, G.T. and G.C.; figures, G.T. and G.C. All authors have read and agreed to the published version of the manuscript. Funding: This research was funded by the International Institute of Tropical Agriculture (IITA) and Makerere University Centre for Soybean Improvement and Development (MAKCSID). Data Availability Statement: The original contributions presented in the study are included in the article, further inquiries can be directed to the corresponding author. Acknowledgments: The authors thank Makerere University for providing the experimental materials used in this research. Conflicts of Interest: The authors declare no conflicts of interest. References 1. Gupta, K.S. Breeding Oilseed Crops for Sustainable Production: Opportunities and Constraints; Academic Press: Cambridge, MA, USA, 2015; ISBN 978-0-12-801309-0. 2. Our World in Data. Available online: https://ourworldindata.org/drivers-of-deforestation#is-our-appetite-for-soy-driving- deforestation-in-the-amazon (accessed on 6 August 2024). 3. Agralytica. Agralytica United Soybean Board Connections 2012: Soybean Market Scan; Agralytica: Washington, DC, USA, 2012; p. 97. 4. FAOSTAT. 2022. Available online: http://www.fao.org/faostat (accessed on 17 November 2024). 5. WWF. The Growth of Soy: Impacts and Solutions; WWF International: Gland, Switzerland, 2014. ISBN 978-2-940443-79-6. Available online: http://issuu.com/wwfsoyreport/docs/wwf_soy_report_final_jan_19/1?e=10667775/6569194 (accessed on 26 October 2024). 6. Joala, R. The Rise of Soya in Zambia and the Integration of smallholder farmers. Inst. Poverty Land Agrar. Stud. 2018, 1–4. 7. Masuda, T.; Goldsmith, P.D. World soybean production: Area harvested, yield, and long-term projections. Int. Food Agribus. Manag. Rev. 2009, 12, 143–162. [CrossRef] 8. Singh, P.; Kumar, R.; Sabapathy, S.N.; Bawa, A.S. Functional and edible uses of soy protein products. Compr. Rev. Food Sci. Food Saf. 2008, 7, 14–28. [CrossRef] 9. Hardarson, G.; Atkins, C. Optimising biological N2 fixation by legumes in farming systems. Plant Soil 2003, 252, 41–54. [CrossRef] 10. Sanginga, N.; Woomer, P.L. Integrated Soil Fertility Management in Africa: Principles Practices and Development Process; Tropical Soil Biology and Fertility (TSBF); International Centre for Tropical Agriculture (CIAT): Nairobi, Kenya, 2009; ISBN 978-92-9059-261-7. 11. Sanginga, P.C.; Adesina, A.A.; Manyong, V.M.; Otite, O.; Dashiell, K.E. Social impact of soybean in Nigeria’s southern Guinea savanna. Trop. Agric. 1999, 34. 12. Sarutayophat, T. Correlation and path coefficient analysis for yield and its components in vegetable soybean. Songklanakarin J. Sci. Technol. 2012, 34, 273–277. 13. Balla, M.Y.; Ibrahim, S.E. Genotypic Correlation and Path Coefficient Analysis of Soybean [Glycine max (L.) Merr.] for Yield and Its Components. Agric. Res. Technol. 2017, 7, 3–7. 14. Ferrari, M.; Carvalho, I.R.; de Pelegrin, A.J.; Nardino, M.; Szareski, V.J.; Olivoto, T.; da Rosa, T.C.; Follmann, D.N.; Pegoraro, C.; da Maia, L.C.; et al. Path analysis and phenotypic correlation among yield compnents of soybean using environmental stratification methods. Aust. J. Crop Sci. 2018, 12, 193–202. [CrossRef] 15. Wei, M.C.F.; Molin, J.P. Soybean yield estimation and its components: A linear regression approach. Agric. Switz. 2020, 10, 348. [CrossRef] 16. Karyawati, A.S.; Puspitaningrum, E.S.V. Correlation and path analysis for agronomic traits contributing to yield in 30 genotypes of soybean. Biodiversitas 2021, 22, 1146–1151. [CrossRef] 17. Lekota, M.P.; Morojele, M.E.; Motanyane, M.S. Assessment of Yield and Yield Components of Soya-Bean (Glycine max (L.) Merril) Grown Under Conventional Agronomic Practices of Lesotho. SSRN Electron. J. 2021, 9, 1–7. [CrossRef] 18. Mulugeta, A.; Kidane, S.; Abadi, S.; Fisha, Z. GGE biplots to analyze soybean multi-environment yield trial data in north Western Ethiopia. J. Plant Breed. Crop Sci. 2013, 5, 245–254. [CrossRef] 19. Adie, M.M.; Krisnawati, A.; Gatut-Wahyu, A.S. Assessment of Genotype × Environment Interactions for Black Soybean Yield using Ammi and GGE Biplot. Int. J. Agric. Innov. Res. 2014, 2, 673–678. https://ourworldindata.org/drivers-of-deforestation#is-our-appetite-for-soy-driving-deforestation-in-the-amazon https://ourworldindata.org/drivers-of-deforestation#is-our-appetite-for-soy-driving-deforestation-in-the-amazon http://www.fao.org/faostat http://issuu.com/wwfsoyreport/docs/wwf_soy_report_final_jan_19/1?e=10667775/6569194 https://doi.org/10.22004/ag.econ.92573 https://doi.org/10.1111/j.1541-4337.2007.00025.x https://doi.org/10.1023/A:1024103818971 https://doi.org/10.21475/ajcs.18.12.02.pne488 https://doi.org/10.3390/agriculture10080348 https://doi.org/10.13057/biodiv/d220309 https://doi.org/10.2139/ssrn.3794627 https://doi.org/10.5897/JPBCS13.0403 Agronomy 2024, 14, 2826 14 of 14 20. Cheelo, P.; Lungu, D.; Mwala, M. GGE Biplot Analysis for Identification of Ideal Soybean [Glycine max L. Merrill] Test and Production Locations in Zambia. J. Exp. Agric. Int. 2017, 15, 1–15. [CrossRef] 21. El-Mohsen, A. Agronomical evaluation of six soybean cultivars using correlation and regression analysis under different irrigation regime conditions. J. Plant Breed. Crop Sci. 2013, 5, 91–102. [CrossRef] 22. de Carvalho, C.G.P.; Arias, C.A.A.; de Toledo, J.F.F.; de Oliveira, M.F.; Vello, N.A. Correlações e análise de trilha em linhagens de soja semeadas em diferentes épocas. Pesqui. Agropecuária Bras. 2002, 37, 311–320. [CrossRef] 23. Akram, R.M.; Fares, W.M.; Fateh, H.S.A.; Rizk1, A.M.A. Genetic variability, correlation and path analysis in soybean. Legume Res. 2011, 34, 36–40. 24. Iqbal, S.; Mahmood, T.; Tahira Muhammad Ali, M.A.; Sarwar, M. Path Coefficient Analysis in Different Genotypes of Soybean (Glycine max (L.) Merril). Pak. J. Biol. Sci. 2003, 6, 1085–1087. [CrossRef] 25. Vu, T.T.H.; Le, T.T.C.; Vu, D.H.; Nguyen, T.T.; Ngoc, T. Correlations and Path Coefficients for Yield Related Traits in Soybean Progenies. Asian J. Crop Sci. 2019, 11, 32–39. [CrossRef] 26. Rodrigues, G.B.; Marim, B.G.; da Silva, D.J.H.; Mattedi, A.P.; de Souza, V. Path analysis of primary and secondary yield components in tomato plants of the Salad group. Pesqui. Agropecu. Bras. 2010, 45, 155–162. [CrossRef] 27. Dvorjak, D.S.; Unêda-Trevisoli, S.H.; De Sousa Leite, W.; Da Silva, A.J.; Da Silva, F.M.; Di Mauro, A.O. Correlations and path analysis in soybean progenies with resistance source to cyst nematode (race 3). Comun. Sci. 2019, 10, 168–175. [CrossRef] 28. De Mendiburu, F. Agricolae: Statistical Procedures for Agricultural Research. R Package Version 1.3-5. 2021. Available online: https://cran.r-project.org/web/packages/agricolae/index.html (accessed on 26 October 2024). 29. Yan, W.; Rajcan, I. Biplot analysis of test sites and trait relations of soybean in Ontario. Crop Sci. 2002, 42, 11–20. [CrossRef] [PubMed] 30. Wickham, H.; François, R.; Henry, L.; Müller, K. Dplyr: A Grammar of Data Manipulation. R Package Version 0.8.4. Available online: https://CRAN.R-project.org/package=dplyr (accessed on 26 October 2024). 31. Taiyun, W.; Viliam, S. R package “corrplot”: Visualization of a Correlation Matrix. (Version 0.88). Available online: https: //github.com/taiyun/corrplot (accessed on 26 October 2024). 32. Lenka, D.; Misra, B. Path Coefficient Analysis of Yield in Rice Varieties. Indian J. Agric. Sci. 1973, 43, 376–379. 33. Obua, T.; Sserumaga, J.P.; Awio, B.; Nganga, F.; Odong, T.L.; Tukamuhabwa, P.; Tusiime, G.; Mukasa, S.B.; Nabasirye, M. Multi-environmental evaluation of protein content and yield stability among tropical soybean genotypes using gge biplot analysis. Agronomy 2021, 11, 1265. [CrossRef] 34. Yan, W.; Kang, M.S.; Ma, B.; Woods, S.; Cornelius, P.L. GGE biplot vs. AMMI analysis of genotype-by-environment data. Crop Sci. 2007, 47, 643–655. [CrossRef] 35. Tukamuhabwa, P.; Oloka, H.K.; Sengooba, T.; Kabayi, P. Yield stability of rust-resistant soybean lines at four mid-altitude tropical locations. Euphytica 2012, 183, 1–10. [CrossRef] 36. Valencia-Ramírez, R.A.; Ligarreto-Moreno, G.A. Phenotypic correlation and path analysis for yield in soybean (Glycine max (L.) Merril). Acta Agronómica 2012, 61, 353–363. 37. Kozak, M.; Azevedo, R.A. Does using stepwise variable selection to build sequential path analysis models make sense? Physiol. Plant. 2011, 141, 197–200. [CrossRef] 38. Xu, C.; Li, R.; Song, W.; Wu, T.; Sun, S.; Hu, S.; Han, T.; Wu, C. Responses of Branch Number and Yield Component of Soybean Cultivars Tested in Different Planting Densities. Agriculture 2021, 11, 69. [CrossRef] 39. Allen, L.H.; Zhang, L.; Boote, K.J.; Hauser, B.A. Elevated temperature intensity, timing, and duration of exposure affect soybean internode elongation, mainstem node number, and pod number per plant. Crop J. 2018, 6, 148–161. [CrossRef] 40. Pereira-Flores, M.E.; Justino, F.B. Yield Components and Biomass Partition in Soybean: Climate Change Vision; IntechOpen: London, UK, 2019. 41. Van Roekel, R.J.; Purcell, L.C. Understanding and Increasing Soybean Yields. 2016, 1–6. Available online: https://dr.lib.iastate. edu/entities/publication/864f0deb-5cb3-437d-a0da-ea6b049a2e14 (accessed on 26 October 2024). [CrossRef] 42. Van Roekel, R.J.; Purcell, L.C.; Salmerón, M. Physiological and management factors contributing to soybean potential yield. Field Crops Res. 2015, 182, 86–97. [CrossRef] Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. https://doi.org/10.9734/JEAI/2017/30154 https://doi.org/10.5897/JPBCS2013.0389 https://doi.org/10.1590/S0100-204X2002000300012 https://doi.org/10.3923/pjbs.2003.1085.1087 https://doi.org/10.3923/ajcs.2019.32.39 https://doi.org/10.1590/S0100-204X2010000200006 https://doi.org/10.14295/cs.v10i1.1697 https://cran.r-project.org/web/packages/agricolae/index.html https://doi.org/10.2135/cropsci2002.1100 https://www.ncbi.nlm.nih.gov/pubmed/11756248 https://CRAN.R-project.org/package=dplyr https://github.com/taiyun/corrplot https://github.com/taiyun/corrplot https://doi.org/10.3390/agronomy11071265 https://doi.org/10.2135/cropsci2006.06.0374 https://doi.org/10.1007/s10681-011-0404-3 https://doi.org/10.1111/j.1399-3054.2010.01431.x https://doi.org/10.3390/agriculture11010069 https://doi.org/10.1016/j.cj.2017.10.005 https://dr.lib.iastate.edu/entities/publication/864f0deb-5cb3-437d-a0da-ea6b049a2e14 https://dr.lib.iastate.edu/entities/publication/864f0deb-5cb3-437d-a0da-ea6b049a2e14 https://doi.org/10.31274/icm-180809-195 https://doi.org/10.1016/j.fcr.2015.05.018 Introduction Materials and Methods Experimental Materials Description of the Study Locations Experimental Design Data Collection Days to Flowering (DTF) and Days to Maturity (DTM) Plant Height, Pod Clearance, and Number of Branches Length and Number of Internodes Yield and Yield-Related Traits Data Analysis Multi-Locational Yield Data Correlation Analysis Path Analysis Results Grain Yield Performance of Genotypes Across Locations Genotype Yield Stability Using GGE Biplot Analysis Correlations Between Grain Yield and Yield-Related Traits Path Analysis of Yield-Related Traits Discussion Grain Yield Performance of Genotypes Across Locations Genotype Yield Stability Using GGE Biplot Analysis Correlations Between Grain Yield and Yield-Related Traits Path Analysis of Yield-Related Traits Conclusions References