Analysis of Bambara groundnut (Vigna subterranea (L.) Verdc.) diversity towards improved yield

PhD (Biology), North-West University, Mahikeng Campus

A considerable set of underutilized crops perform better than their more accepted counterparts when cultivated in less nutritive soils and grossly unfavorable environmental conditions. This advantage necessitates the development of these special crops for the good of sustainable agriculture. Based on the effect of a changing climate and the expected population increase, it is necessary to find a means of improving crop diversities and yield. Less valued crops need to be accepted and incorporated into the food system. Bambara groundnut (Vigna subterranea (L.) Verdc.) is one such crop with so much promise of help in enhancing food security. There is a need to find approaches to improve crop yield and stop this decline in food availability, especially in the less-developed world. Bambara groundnut (BGN) and other legumes with the basic compositions of required nutrients such as proteins, oils, and carbohydrates have been major sources of food for humans and animals alike. One of the ways used in selecting crops and improving yields is the application of multi-environment trials (MET) which have been employed in various crops to select the best cultivars that adapt well to various environments. This has aided in the development of many adaptive and stable cultivars across environments. More importantly, advances in next-generation sequencing (NGS) technologies coupled with improvements in bioinformatics tools have strengthened research in plant breeding to tackle food security. The use and accuracy of molecular breeding has been improved through marker-assisted selection (MAS), genotyping, and gene editing, among others. Improvements in genotyping by sequencing (GBS) have been widely successful in crops with reference genomes. Most less-studied crops do not have a reference genome yet, but other approaches can be employed, such as genome-wide association studies (GWAS), quantitative trait loci (QTL) analysis, and comparative genomics, among others. In this study, a set of 95 accessions of Bambara groundnut that have not been DArT-characterized were selected from the germplasm collection in the IITA Gene bank. These accessions were evaluated for morphological traits in a MET in 2018 and 2019 in Ibadan and Ikenne, South-West Nigeria. Ibadan is in the derived savanna and Ikenne is in the tropical rain forest. To validate their ability to enhance food and nutrition security, their nutrient, antinutrient, mineral components, and stress responses were accessed. The objectives of the field trials were to evaluate the diversities in the phenotypic and agro-morphological traits in the selected accessions, to examine the effect of the environment on the individual traits and accessions, to discover the most stable and adaptable accession in terms of yield among the selected accessions, and to select the best environment for the crop. Experiments were laid out in a randomized complete block design (RCBD), replicated three times. The plot area was 3m2 with 10 plants per plot. Spacing between each plant was 0.3m and inter-plot spacing was 1m. An alley of 1m separates each replicate. The plants were rainfed and irrigated as appropriate and all standard agronomic practices were observed. After planting, young leaves from 2-week-old plants were collected and DNA was extracted for DArT sequencing. Data were collected from the fields at the appropriate time using the field book. For the nutrient and antinutrient components, good-looking seeds were selected after harvest and analyzed in the Food and Nutrition laboratory. Drought assessment was carried out in the screen house in IITA, Ibadan. Wooden boxes were used and arranged using RCBD in three replicates. Five accessions were planted in each box with 6 plants per accession which were later thinned to 3 after 2 weeks. The boxes were irrigated to field capacity for 24hrs before planting and the moisture content at field capacity was recorded. After planting, watering was done regularly for 4 weeks when plants were fully established and the watering was stopped. Individual plants were scored for wilting, stem greenness, chlorophyll content, and leaf senescence. Scoring was done on days 7, 10, and 13 before watering was resumed. Boxes were watered to field capacity on the day of resumption of irrigation, thereafter once every 2 days for 2 weeks until the experiment was concluded. The collected data were subjected to ANOVA, and the means were separated using the Fischer LSD test. Principal component analysis (PCA), correlation, and cluster analysis were also evaluated for the different traits. Furthermore, a genome-wide analysis study was conducted on the stress-treated plants to identify genomic regions and candidate genes regulating the stress-response traits studied. The Eberhart and Russell method and GGE biplot were used to analyze the stability analysis and predict the best genotypes and best environment. Results showed that location was highly significant for all the traits (p <0.0001) except for plant height and leaf length. The accessions varied significantly in plant height, leaf length and width, chlorophyll content, number of petioles, germination count, petiole length, number of pods per plant, number of seeds, hundred seed weight, seed length, seed width, seed thickness, and yield (p <0.0001) while days to flowering (p<0.001) and days to 50% germination and total seed weight (p <0.01) were also significant but their responses to the trait days to emergence was not significant. The interaction effect of location and accession was highly significant (p <0.0001) on leaf width, chlorophyll content, number of petioles, germination count, number of pods, number of seeds, and yield, while plant height was also significant at p <0.001, leaf length and seed length were significant at p <0.01, and seed width was significant at p <0.05. However, the interaction between accession and year was highly significant for plant height, leaf width, number of pods, and number of seeds (p <0.0001) and leaf length (p <0.001). There was a highly significant effect of location, accession, and year interaction on leaf length, leaf width, petiole length, number of pods, and number of seeds (p <0.0001), plant height and days to flowering (p <0.01), and hundred seed weight (p <0.05). This implies that high levels of variability and heterogeneity exist among accessions, locations, and years in response to the traits scored. Principal components 1 (24.67%) and 2 (17.63%) account for 42.3% of the total variance observed. Among the variables, seed width (19.53%), seed thickness (19.58%), hundred seed weight (16.98%), seed length (15.93%) and yield (9.76%) were the major contributing traits in PC1, while number of seeds (21.78%), number of pods (18.48%), total seed weight (13.96%), plant height (9.12%), and petiole length (8.93%) were the major contributing traits in PC2. From the biplot, accessions loading on PC1 are high yielding with thick seeds and long seeds while at the same time having high hundred seed weight. Accessions loaded on PC2 have a high number of seeds, number of pods, and total seed weight. The cluster analysis grouped the accessions into 4 clusters (red, green, blue, and purple) based on the agro-morphological traits with the clusters in red having the highest number of accessions (37 accessions) followed by the ones in green (30), blue having 11, and the purple cluster with 17 accessions. There are lots of significant correlations among the traits scored. In the analysis of the genetic parameters, the phenotypic variance is higher than the genotypic variance in all the traits. Yield (kg ha−1) reported higher phenotypic (19,476.39) and genotypic (5,159.09) variances, while the lower phenotypic (0.68) and genotypic (0.23) variances were observed in leaf width. The traits such as LLE (GCV 7.18, PCV 19.95), GCT (GCV 6.50, PCV 19.61), DTF (GCV 6.31, PCV 10.81), SEEDL (GCV 14.41, PCV 18.19), SEEDW (GCV 11.84, PCV 16.13), and SEEDT (GCV 13.48, PCV 17.66) showed below 20% of phenotypic coefficient of variation (PCV) and genotypic coefficient of variation (GCV). Yield stability analysis showed that seed yield was significantly affected by genotype, environment, and GEI. The mean squares of the accessions were highly significant, also the effect of the environment over the years. Out of the 95 accessions studied, 22 were found to be good-performing and stable, 6 were found to be adaptable, while the remaining accessions were not affected by the environmental factors. The biplot explained 80% of the total variation observed. The first principal component (axis1) explained 48.59% and the second principal component (axis2) explained 31.41%. Accessions TVSu-1866, TVSu-2022, TVSu-2017, TVSu-1943, TVSu-1892, TVSu-2060, and TVSu-1557 were all located at the corners of the polygon in the "which won where" view of the relationship between accessions and environments, indicating that these accessions were outstanding in those environments. TVSu-1706, TVSu-2018, TVSu-1785, TVSu- 1895, and TVSu-1951 accessions performed consistently across all environments. IB2019 was the closest to being an ideal environment, while TVSu-2020 and TVSu-1649 were the most ideal accessions, followed by accessions TVSu-2021, TVSu-1664, TVSu-1866, and TVSu-2025. However, accessions TVSu-1557, TVSu-2060, TVSu-2056, and TVSu-2042 were the worst accessions in terms of yield performance as they are located far from the center of the concentric circle. The result of the nutrient and antinutrient composition shows a highly significant difference for the traits. The two components account for 41.2% of the total variations observed. The clustering based on the traits depicts four main groups. According to the correlation matrix, protein was significantly correlated with ash, fat, and phytate. Fat correlated with moisture content and tannin; tryptophan correlated slightly with protein content and correlated highly with tannin; moisture content and tannin were also highly correlated. Correlation between drought response traits showed a significant positive correlation between chlorophyll content and recovery. PCA of the traits showed variation in response levels on the three different days that data was taken. Further clustering analysis grouped the accessions based on the response traits. A total of twenty significant SNPs (considering thresholds of log (p) ≤ 0.001 with R2 ≥ 9%) from both the GLM (15) and MLM (5) models were identified by GWAS analysis of the BGN accessions in response to water stress using the Mungbean reference genome. In the study, twelve SNPs associated with drought stress response were identified. These SNPs are co-localized with the Vradi07g31020, Vradi05g01630, Vradi06g04840, Vradi06g03310, and Vradi04g08510 genes which encode for a transaldolase, pectin esterase, proline transporter 1, GDSL esterase, and outer plastidial membrane protein porin respectively. As well as the Vradi03g05310, Vradi10g10930, Vradi03g08520, and Vradi02g06260 genes encoding for cell division cycle 20.2- cofactor of APC complex, putative tRNA (cytidine(32)/guanosine(34)-2'-O)-methyltransferase, UDP- glycosyltransferase 76F1, and SHUGOSHIN 2 respectively. Finally, the genes Vradi04g01720, Vradi02g02440, and Vradi07g10090 were discovered, which encode the origin of replication complex subunit 1A, alanine--tRNA ligase, and salicylic acid-binding protein 2. This study showed that BGN can help improve food and nutritional security, and the accessions used can serve as a source of parent lines for improved varieties.

Doctoral