YONAS MOGES GELAW

GENETIC DIVERSITY, ASSOCIATION MAPPING ANALYSIS AND INHERITANCE OF SEED IRON AND ZINC CONCENTRATION OF COMMON BEAN (PHASEOLUS VULGARIS L.) IN ETHIOPIA

ABSTRACT 

Common bean is the main staple food for more than 300 million people in east Africa, Latin America and southern Europe. In Ethiopia common bean is widely grown as nutrition crop as well as source of income and export market. It is one of the main crops which can serve as a good source of minerals and protein. Despite common bean is considered to be a rich source of Fe and Zn, genetic variability existed among the genotypes. The presence of genetic variation is helpful to alleviate current and future global food and nutrition scarcity. In Ethiopia, studies related to seed Fe and Zn concentration and large scale characterization of the genotypes and association of iron and zinc with high density SNP markers are very limited. Therefore, the main objective of this study was to develop common bean varieties with high Fe and Zn with reasonable yield and consumer acceptance, and thereby contribute to the improvement of food and nutrition security for smallholder farmers in Ethiopia. The specific objectives of the study were to: (i) assess genetic variability among common bean germplasm for micronutrient concentration, morpho-agronomic traits; (ii) determine the genetic diversity and population structure of common bean germplasm using high density SNP markers; (iii) identify SNP markers associated with Fe and Zn concentration in common bean and (iv) determine the inheritance of Fe and Zn concentration and other agro-morphological traits in common bean. Three hundred common bean genotypes were planted at MARC and HU in 2018 using alpha lattice design and replicated thrice to evaluate micronutrient concentration and agro-morphological traits. The genotypes showed significant variation for Fe, Zn and agro-morphological traits. In the combined analysis, seed Fe concentration ranged from 61.7 to 103.1 ppm, while Zn concentration ranged from 25.5 to 42.2 ppm. Five genotypes (NUA227, NUA517, NUA514, NUA515 and NC-08) had ≥90ppm Fe concentration and five genotypes (NUA514, NUA517, Montcalm, NC-08 and Awash Melka) had ≥38ppm Zn concentration. Based on the whole genome sequencing of the genotypes generated 11480 high quality DArTSeq SNP markers revealed that the overall mean Nei’s gene diversity, PIC and inbreeding coefficient (FIS) of the entire common bean genotypes were 0.38, 0.30 and 0.85, respectively. Based on geographic region, the highest Nei’s gene diversity was observed in the landrace collected from Oromia region. Pre-grouping of the genotypes into the different regions of collection and improved varieties revealed that the highest genetic distance was scored between SNNPR and CIAT (0.492). Clustering of the common bean genotypes based on geographical area revealed that the genotypes were not clustered based on their geographical location which could be due to duplication of the landraces. Based on the Bayesian clustering algorism implemented in STRUCTURE software using the admixture model, the optimal subpopulations were six (K=6). Genome wide association analysis revealed that 53 SNP markers with a cut-off point of  log(p)≥5.36 were significantly associated with seed Fe and Zn concentration. Forty-one of the significant SNPs were found at Melkassa location of which 40 were associated with seed Fe concentration and one was for Zn concentration. At Haramaya, six SNPs were detected for Fe and one for Zn concentration. In the combined analysis of the two locations, 23 SNPs were associated with Fe concentration and one is for Zn concentration. Five of the significant SNPs, SNP9735, SNP6123, SNP995, SNP5916, SNP11261 were found to be stable across the testing locations and in the combined analysis. However, any of the markers for Zn concentration were not stable. The effects of the significant markers ranged from 8 to 13% for Fe concentration and 8 to 9% for Zn concentration. Generation mean analysis of the three sets of crosses revealed that the scaling and χ2 test for goodness of fit were significant in all the three sets of crosses revealed that the additivedominance model was inadequate to elucidate the inheritance of Fe and Zn concentration and all other agronomic traits. This was an indicative of the involvement of allelic and non- allelic interaction effects in the control Fe, Zn and other agronomic traits. In conclusion, common bean genotypes showed a wide range of phenotypic variation for Fe and Zn and other morphoagronomic traits. The SNP based diversity showed that the landraces were diverse and provide new genetic variants. The wide genetic diversity in the studied genotypes would help to develop common bean genotypes with desirable traits. The genome wide association analysis indicated that five of the SNP markers associated with Fe concentration were stable across the environments which is a genotypic characteristic with little or no environmental effect. The stable markers can be used in molecular breeding program to select genotypes with increased Fe concentration in the seed using marker assisted selection after validation of the markers in diverse set of genotypes. This would help to fast track the biofortification strategy to increase and stabilize Fe concentration in the seed of common bean and develop varieties which could reduce micronutrient deficiency. The generation mean analysis showed that both fixable and non-fixable gene effects coupled with duplicate epistasis contributed to the inheritance of Fe and Zn concentration and the majority of the traits.