BREEDING SORGHUM (Sorghum bicolor L. Moench) FOR HIGH GRAIN YIELD AND RESISTANCE TO SORGHUM MIDGE (Stenodiplosis sorghicola Coquillett) IN NIGER
According to the results of rural appraisal, farming in the study sites was mainly smallholder family system. Farmers were predominantly young sorghum producers with low levels of education and relatively little farming experience. The use of local varieties was dominant in farming system. The seed supply system was mainly through family networks. The phenotypic characteristics preferred by farmers were white seed color, intermediate stem height, big panicle size and large hard seed. Sorghum midge is one of the major limiting factors. The insect was well known by farmers since it has been experienced by most of the farmers in their fields where it affects all their sorghum plants almost every year. However, no control measures are being used to overcome sorghum midge infestation.
Combining ability studies revealed significant differences among lines, testers and the interaction between lines and testers in both planting date for most of the characters under study. Additive gene action was predominant for all traits in both planting dates as well as across planting date. Non additive gene action was also predominant for plant height, days to 50% flowering, grain yield and resistance to midge. Based on General Combining Ability effects (GCA) and Specific Combining Ability (SCA) effects, some lines and hybrids have been identified with good yield potential and resistance to sorghum midge
High Phenotypic Coefficient of Variation (PCV) versus genotypic coefficient of variation (GCV) was observed in both sites and planting dates. Across planting dates at both Konni and Maradi, grain yield, plant height, panicle weight and midge damage had high heritability coupled with high estimates of genetic advance. At Konni, high genetic advance coupled with high heritability was detected for grain yield, plant height, panicle weight and resistance to midge.
Similarly, results were at Maradi, for grain yield, plant height and panicle weight. The variation observed among genotypes was represented by the first three Principal Components (PC). The 22.73% variation in the PCI was due to plant height, 1000 seeds weight, grain yield and days to flowering. PCII accounted for 17.99% variation associated with days to flowering, panicle compactness, midge damage and 1000 seeds weight. The PCIII presented 15.50% of the total variation mainly due to awned glumes and midge damage. Based on the similarity within and between classes, genotypes were divided into 18 cluster groups. Correlation analysis suggested that 1000 seeds weight, days to flowering, plant height and awned glumes can be good selection criteria for yield and resistance to midge in sorghum for early planting at Konni. Low 1000 seed weight, presence of awned glumes and compact panicles would be indices for midge resistance for late planting at Konni. Reduction of awned glumes on panicles, reduction in panicle compactness and increase in days to flowering could be useful as selection criteria for yield in early planting at Maradi. However, selection for resistance to midge and high grain yield for late planting sorghum at Maradi, should be focused on early maturing sorghum with compact panicles having awned glumes.
Two distinct groups of environments for evaluation of grain yield were observed. Genotype L232 was found to be the best genotype for grain yield in the environments 1, 3, 4 and genotype L17 was the best for grain yield in the environment 2. Second planting date at Konni was found to be the most discerning environment while the first planting date at Konni was found to be suitable for selecting widely adapted genotypes for grain yield. Genotypes L232, L17, L182, L202 L168 and L64, were found to be the ideal and most desirable genotypes for yield stability.