FIELD AND MOLECULAR SCREENING FOR STRIGA RESISTANCE IN SELECTED FINGER MILLET (Eleusine coracana, L. Gaertn) GERMPLASM IN WESTERN KENYA

Abstract

Finger millet (Eleusine coracana) is an importance food crop in Africa and Asia. Its grain is richer in protein, fat and minerals than other major cereals. The parasitic weed Striga hermonthica (Del.) Benth seriously limits finger millet production. The damage of Striga to cereal crop is more severe under drought and low soil fertility. The main objectives of this study were to: (i) assess the effect of Striga infestation on finger millet based on agro-morphological traits, (ii) determine genetic basis of resistance of finger millet to S. hermonthica using genome-wide selection with single nucleotide polymorphism (SNP) markers through Genotyping by sequencing (GBS) and (iii) determine genetic diversity among the selected finger millet genotypes against S hermonthica. One hundred finger millet genotypes were evaluated for resistance against S hermonthica (Del) Benth under field conditions at Alupe and Kibos sites in Western Kenya. The genotypes were planted in control and experimental plots inoculated with Striga and plant growth monitored to maturity. All accessions were genotyped–by-sequencing (GBS) and data analyzed using the non-reference based Universal Network Enabled Analysis Kit (UNEAK) pipeline. Genome wide association studies (GWAS) were done to establish the association of detected SNPs with Striga resistance based on field results. Statistical analysis of phenotypic data using Statistical Analysis System (SAS) PROC ANOVA revealed highly significant differences among genotypes for morphological traits at P<0.05. Six genotypes showed high resistance to Striga with a mean Striga count of 0 while the most susceptible genotype had Striga count mean of 69.17 at maturity. In molecular analysis 117542 SNPs from raw GBS data used in GWAS revealed that markers TP 85424 and TP 88244 were associated with Striga resistance in the 95 genotypes. Principal Component Analysis revealed that the first and third component axes accounted for 2.5% and 8% of total variance respectively and the genotypes were distributed according to their reaction to Striga weed. Genetic diversity analysis grouped the 95 accessions into three major clusters containing 32 (A), 56 (B), and 7 (C) genotypes each. All finger millet genotypes that showed resistance to Striga in the field were from cluster B while the most susceptible genotypes were from cluster A. Results revealed genetic variation for Striga resistance in cultivated finger millet genotypes and hence the possibility of marker – assisted breeding for the trait. It is suggested that more studies including more genotypes and wild relatives be carried out to understand further the resistance to Striga in Eleusine genera