Utility of indirect and direct selection traits for improving Striga resistance in two sorghum recombinant inbred populations

Abstract

Breeding of sorghum ( Sorghum bicolor L. Moench) for resistance to the parasitic weed Striga hermonthica (Del.) Benth. has been hampered by the difficulty of evaluating host resistance in the field and lack of reliable screening techniques. Therefore, we investigated the value of various indirect and direct measures of Striga resistance as selection traits. Two sorghum recombinant inbred populations of 226 F 3:5 lines each were developed from the crosses (1) IS 9830×E 36-1 and (2) N 13×E 36-1. Striga-resistant line IS 9830 is characterized by low stimulation of Striga seed germination, whereas Striga-susceptible line E 36-1 produces germination stimulants in abundance. Line N 13 possesses “mechanical” resistance and probably also an antibiosis mechanism. Resistance was assessed in 1997 and 1998 using in vitro agar-gel assays with Striga seeds from Kenya, Mali, and Niger, pot trials in the respective three countries, and field experiments in Kenya and Mali. The agar-gel assay proved to be a useful, precise and fast indirect selection method to screen for sorghum entries with the low-stimulant character. However, correlation analysis showed that this resistance mechanism was ineffective in some environments, especially in Kenya, pointing to the necessity of field evaluation. Because of low heritability estimates and moderate to low correlations to Striga resistance under field conditions, pot screening appeared to be of limited use in breeding programs. The field trials confirmed the effectiveness of several direct measures of Striga resistance in sorghum: emerged Striga counts, Striga severity index, and area under the Striga number or severity progress curves. A two-row plot field layout with an empty row between plots, coupled with artificial infestation of test rows, lattice design and six replications offered an improved screening procedure that achieved high heritability. Significant genotype×environment interactions in the field experiments stress the importance of multi-locational trials to achieve stable Striga resistance.