Genetic Diversity, Population Structure, and Marker-Trait Association Analysis in Sugarcane (Saccharum spp. L.) Genotypes of Ethiopia

Abstract

Sugarcane (Saccharum spp. hybrid) is the primary sugar and bioenergy crop in tropical and subtropical regions, including Ethiopia. However, its production and productivity have been severely hampered mainly due to a lack of improved varieties. Analyzing sugarcane germplasm genetic diversity and population structure provides valuable genetic information that breeders can use in improvement programs. In Ethiopian sugarcane germplasm collections, many genotypes remain phenotypically and molecularly uncharacterized. Additionally, no studies on marker-trait associations for Ethiopian sugarcane germplasm have been reported. Therefore, the present study aimed to investigate the genetic diversity, population structure, and association mapping of major agronomic traits of sugarcane genotypes of Ethiopia. Field experiments were conducted at Metehara and Kessem Sugar Estates between 2021 and 2022, involving 144 sugarcane genotypes. The experiment was laid out as a partially balanced lattice design with two replications. The data collected on 16 qualitative traits, 16 quantitative traits and 20 SSR markers was subjected to different statistical analysis. With regard to qualitative morphological diversity, the results indicated that the highest Shannon diversity index (H') was recorded for stalk corky patches while the lowest has been recorded for stalk growth crack. The H' pooled across qualitative traits by country of collection ranged from 0.00 for Mauritius and Brazil to 0.83 for India, with a mean of 0.62, implying there was high phenotypic variability within the Indian population and among the overall studied populations. The genotypes were grouped into four distinct clusters based on qualitative traits. The analysis of variance revealed significant differences (P < 0.01) among genotypes for all quantitative traits. PCA result showed the first four principal components (PCs) contributed to 81.44% of the overall variation. Multivariate cluster analysis grouped the genotypes into five distinct clusters based on genetic distance. Highly significant positive genotypic and phenotypic association (P < 0.01) was revealed between cane yield and traits, including sprouting percentage, tiller number, millable cane number, stalk height, single cane weight, and internode number. GenotypeCP6023, CP70/321, Mex57/197, D41/46, CP971944, B657/150, and B49-244 were selected based on Smith-Hazel selection index. The genotypic and phenotypic path coefficients showed a significant positive direct effect of the number of millable canes (1.174 and 0.75936) and single cane weight (1.067 and 0.49977) on cane yield. The study found high genotypic variability, high heritability, and high genetic advance for the number of tillers, millable canes, and cane yield, indicating that these traits can serve as selection criteria. Sugarcane genotypes CP6023, CP70/321, Mex57/197, D41/46, CP971944, B657/150, and B49-244 were selected for their higher mean performance of sugar yield than standard check and commercial varieties. Molecular analysis revealed a high level of genetic diversity in the markers, with an average of 1.9 numbers of alleles, effective alleles of 1.59, Shannon's information index of 0.35, Nei's gene diversity of 0.48, and polymorphic percentage loci of 88.59%, polymorphic information content of 0.64, and genetic differentiation of 0.37. Among the study populations, India had the most observed alleles (1.86), effective alleles (1.60), Shannon's information index (0.49), Nei's gene diversity (0.34), and unexpected heterozygosity (0.34). Cluster analysis using the UPGMA method showed four main clusters. The population structure analysis also identified six subpopulations within the genotype sets. AMOVA analysis elucidated that 96% and 4% of the total variability were within and among populations, respectively. The principal coordinate analysis grouped the genotypes into four clusters in conformity with cluster analysis. Similarly, ten of the SSR markers UGSM665, UGSM667, UGSM681, SMC1852LA, SMC286CS, SMC477CG, SMC36BUQ, SOMS168, SOGL41, and P89, showed high discriminatory power, with polymorphic information content values above average. General linear model and mixed linear model analysis showed that marker alleles UGSM665-8, SMC36BUQ-21, and P-142-7 were significantly associated (p < 0.001) with the number of millable canes, single cane weight, and leaf width traits, respectively. In this study, considerable phenotypic and genotypic genetic diversity was observed among the studied genotypes. In general, the results from this diversity and marker-trait association study based on qualitative and quantitative traits and SSR markers will be useful for breeders in selecting the best parental combinations for starting effective breeding populations and in marker-assisted breeding. Therefore, the current study will contribute to paving reliable ways for future sugarcane breeding programs in the country.