Abstract:
Mung bean is one of the most important legume crops. It serves as a source of protein for poor farmers in drought-prone areas of Ethiopia. It is used as a source of food and insurance crops during the drought season. Mung bean is adaptable to a wide range of climatic conditions. Despite adaptation and importance in moisture-stressed areas, its productivity is generally low due to a lack of suitable and drought-tolerant varieties. Moreover, the crop has got limited research attention compared to other lowland pulses. Comprehensive information on genetic diversity, drought tolerance, and performance of mung bean is lacking. The present study included field, laboratory, and greenhouse experiments. The objective of the study was to assess the diversity of 60 mung bean genotypes using morphological and molecular (SNP) markers for developing high-yielding, drought tolerance, and adaptable varieties. Morphological diversity study was conducted in 2018 main cropping season, while drought tolerance experiment was executed in 2018 offseason at Jinka Agricultural Research Center (JARC) using a 6 x 10 alpha lattice design, while 15 genotypes were evaluated across six locations in 2019 main cropping season for genotype by environment interaction and stability using a randomized complete block design. All field data were collected on plant and plot bases. The in vitro and greenhouse experiments were conducted at the tissue culture laboratory of Areka Agricultural Research Center during 2020 using a completely randomized design, while the molecular diversity study was conducted at Integrated Genotyping Service and Support (IGSS) platform located at Biosciences Eastern and Central Africa (BecA-ILRI) Hub in Nairobi during 2020 using SNPs. Sixty mung bean genotypes were characterized in a 6 x 10 alpha lattice design for morphological markers and drought screening experiments. Forty-four and seventeen data were collected for morphological diversity and drought screening experiments, respectively. Phenotypic diversity was observed among mung bean genotypes based on qualitative traits, while all individual phenotypic classes were not evenly distributed. The first seven principal components explained 94 and 80.1% of the total variation for qualitative and quantitative traits, respectively. Almost all traits were important contributors to morphological divergence. The analysis of variance among the 60 mung bean genotypes revealed highly significant differences for most of the traits. In this study, high estimates of GCV and PCV were noted for harvest index, while moderate estimates of GCV and PCV were observed for, peduncle length, pod length, the number of seeds per pod, seed yield per plant, and hundred seed weight. The lowest GCV and PCV estimates were obtained for days to flowering, days to maturity, seed yield per hectare, and biomass yield per hectare. High heritability coupled with high GAM estimates was recorded for plant height, the number of primary branches per plant, a hundred seed weight, and harvest index, while most traits had high heritability coupled with moderate GAM, indicated that the traits were least influenced by the environmental factors. The magnitude of genotypic correlations was higher than phenotypic correlations for most of the studied traits. Seed yield was positively and highly significantly correlated with most of the traits at the phenotypic and genotypic levels. The values for the genotypic direct and indirect effects were higher than the phenotypic direct xxi and indirect effects for most of the studied traits, indicating that the traits had a strong genetically inherited relationship with seed yield. The analysis of variance under contrasting water regimes revealed that there were significant variations among the genotypes. The mean seed yield under the moisture stress condition was 663.5 kg ha-1 compared to 1062.55 kg ha-1 under the non-stressed condition, with an average yield reduction of 37.53%. The genotypes G22 (NLLP-MGC-22), G24 (NLLP-MGC-24), G28 (VC1973A), and G34 (VC6368 (46-40-4)) were the first four promising and high yielders under moisture stress conditions. Stress tolerance index (STI), yield index (YI), harmonic mean (HM), geometric mean productivity (GMP), and mean relative performance (MRP) were found to be more suitable indices since these indices had the highest correlation with seed yield under both stress and non-stress conditions. Increasing polyethylene glycol concentration from 0% to 2% in the medium causes a gradual decrease in most of the studied parameters except root length. In general, most of the regenerants obtained from G34 showed the best performance under the greenhouse were drought-tolerant under the in vitro condition, suggesting that the accumulated performance of the tested regenerants under in vitro condition was realized under greenhouse condition. It also indicated that in vitro culture is an important tool to develop drought-tolerant genotypes and to improve desirable agronomical traits. As to GEI, the combined analysis of variance revealed that there were significant variations among genotype, environments, and GEI for yield and yield-related traits. Analysis of variance from the AMMI model indicated the contribution of environment, genotype, and GEI was 59.6%, 16.8%, and 14.8% of the total variation in seed yield, respectively. The magnitude of the GEI sum squares was 4.4 times the genotypes sum squares for seed yield. The sum squares of the IPCA1 and IPCA2 explained 47.4% and 7.4% of the GEI variation, respectively. The results from AMMI, Pi, GGE biplot, YSI, and ASV analyses depicted that the NLLP-MGC-24 (G6), Acc006 (G13), and NLLP MGC-15 (G3) were identified as stable and high yielder genotypes across the environments. However, G1 (NLLP-MGC-01), G4 (NLLP-MGC-20), G5 (NLLP-MGC-22), G9 (NM94 (VC6371-94)), G10 (VC6368 (46-40-4)), G11 (NLLP-MGC-06), G12 (Acc002), and G15 (NVL-1) were identified as least stable, and can be recommended for specific environments. AMMI1 biplot showed that Kako was the potential and favorable environment while Humbo was an unfavorable environment for mung bean production. Molecular diversity analysis indicated that the number of polymorphic SNP markers per chromosome varied from 153 to 381 with a mean of 302. The analysis of molecular variance showed that the differentiations accounted for among the population, among individuals, and within individuals were 3%, 74%, and 23% of the total variations, respectively. The mean polymorphic information content (PIC) value was 0.235 and ranged from 0.197 on chromosome 10 to 0.309 on chromosome 3. The average values of expected heterozygosity and fixation index were 0.237 and 0.029, respectively. The mean gene flow among populations was 8.37, indicating a high level of genetic diversity within individuals. The cluster analysis was grouped into three distinct clusters based on genetic similarity. The first three principal coordinate axes explained 48.7% of the total variation. The scatter plot differentiated the population based on their genetic similarity. Generally, the present study confirmed the presence of genetic diversity among genotypes for drought tolerance, yield, and yield-related traits. Therefore, future studies should use the results of the present findings as a benchmark for future collection, and characterization, and on-farm and ex-situ conservations of mung bean landraces.