Tapping the Sudanese sorghum germplasm for grain yield stability and tolerance to drought and spotted stem borer (Chilo partellus, Swinehoe) damage

Two hundred and ten, seventy four, twenty four and eleven genotypes of sorghum were evaluated under full irrigation and partial irrigation in the Gezira Research Station Farm, Wad Medani, Sudan, in season 2001, 2002, 2003 and 2004, respectively in a randomized complete block design with three replicates, to estimate the variability, genotype-environment interaction, stability, character association and path analysis. The data were collected on 11 characters namely plant height, days to 50% flowering, number of leaves, flag leaf area, number of tillers per plant, lodging, senescence, yield, 100 grain weight, number of grains per panicle and stem borer damage. Significant differences among genotypes for the studied characters were found in almost all seasons and under both types of irrigation, indicating that sorghum genotypes were highly variable for these characters, therefore, will respond to selection. The plant height, flowering time, flag leaf area, yield, grain weight, number of grains per panicle and stem borer damage under water stress were less than that under full irrigation by 11 %, 3%, 14%, 19%, 8%, 12% and 39%, respectively. However, number of tillers, senescence and lodging were higher than under full irrigation by 15%, 16% and 9%, respectively, whereas number of leaves per plant was not affected by water stress. The largest decreases induced by drought were observed for stem borer damage and yield. The 19% yield loss reflects the relatively mild moisture stress in comparison with the last season where a yield loss of 36% was experienced. Flowering time, number of leaves per plant, senescence, lodging and stem borer damage under full and partial irrigation over seasons were increased by (14, 15%), (8, 18%), (16, 24%), (25, 35%) and (115, 141 %), respectively. However, flag leaf area, yield and grain weight under both watering regimes were reduced by (30, .34%), (27, 30%) and (9, 31 %), respectively. Plant height and number of tillers were reduced under full irrigation but increased under partial irrigation whereas number of grains per panicle increased under full irrigation and reversed under partial irrigation. Irrigation treatments significantly affected most of the characters except plant height, number of tillers and number of grains per panicle. The interaction of genotypes x season indicated the relative performance of genotypes over seasons. Using the overall mean of the environments, it is seen that genotype PI 568377 and PI 568383 were the most stable. Based on the genotypic variance and the coefficient of variation genotype PI 568377 would be judged as stable and PI 568383 as fairly stable. Further, these genotypes contributed the least to the genotype by environment interaction as measured by ecovalence (wi square m ) and interaction variance( oi square m ). In addition this genotype have small deviation from regression on site index as measured by the deviation mean square, 8 oi square m and having slope closest to 1.00 among the seven genotypes. On the other hand, genotype PI 568383 showed good stability on the basis of low PC I scores compared to genotype PI 568377. Ranking of correlation revealed that number of seeds per panicle is first, grain weight is the second under stress but third under full irrigation and the third under stress is number of leaves while plant height ranked as a second under irrigation. The path coefficient analysis indicated that number of grains per panicle and 100 grain weight were the main components determining grain yield in sorghum. The greatest direct effect on grain yield was exerted by number of grains per panicle. The contribution of these components depended on other characters such as plant height which may be considered as secondary yield component