Evaluation and selection of 20 sorghum [Sorghum bicolor (L.) Moench] genotypes for drought tolerance

A field study was conducted at ARC - Grain Crops Institute, Potchefstroom Experimental Farm and at Taung Crop Production Center during 2006/07 summer growing season. Twenty sorghum genotypes were planted in two separate blocks, under full irrigation and rain fed conditions with plot size of 4 rows x 5 m x 0.9 m giving a density of 55 555 plants ha-1. The experiment was laid out in a randomised complete block design replicated three times. The irrigated block received scheduled irrigation from planting until maturity, while the rain fed block received irrigation prior to germination only. Agronomic traits measured were plant height, stem diameter, biomass, flowering date, panicle exertion, panicle length, leaf area, grain yield, and thousand seed mass at Potchefstroom and Taung experiments. Drought susceptibility index (DSI) was quantified using the formula: DSI = [1 - (Ydi/ Ypi)]/ [1 - (YD/YP)] and %yield reduction (%YR) was calculated using the formula: %YR = (Ypi –Ydi)/ Ypi x 100. Significant variations among genotypes with regard to grain yield (GYLD), plant height (PH), panicle length (PL), biomass (BM), stem diameter (SD), panicle exertion (PEx), and 1000 seed mass (TSM) were observed at Potchefstroom under rain fed conditions, while under irrigated conditions significant variations were only observed for GYLD, PH, PL, BM, days to 50% flowering (DF), SD, leaf area (LA), PEx and TSM. At Potchefstroom genotypes varied significantly with regard to PH, PL, BM, DF, SD, LA, PEx and TSM under rain fed conditions, while under irrigated conditions genotypes varied with all traits measured with exception of harvest index (HI) and relative water content (RWC). At Potchefstroom, soil water deficits significantly affected GYLD, PH, HI, PEx and RWC, while at Potchefstroom soil water deficits significantly affected GYLD, PH, SD, LA, PEx, TSM and RWC. Genotypes varied with their level of resistance/ susceptibility to soil water deficits at both Potchefstroom and Taung. Some remarkable correlations among traits measured were observed under rain fed and irrigated conditions and across the treatments at both Potchefstroom and Taung. At Potchefstroom, significant correlation was only observed between drought susceptibility index (DSI) and PH under rain fed conditions, while at Potchefstroom negative and significant correlation was only observed between DSI and GYLD under rain fed conditions. At Potchefstroom, GYLD under rain fed significantly related to GYLD under irrigated conditions and across the treatments. However, GYLD under irrigated conditions significantly related to GYLD across the treatments. At Potchefstroom, GYLD under rain fed conditions significantly correlated with GYLD across soil water regimes, while GYLD under irrigated conditions significantly correlated with GYLD across the treatments. Genotypes that exhibited the combination of high yield potential and resistant traits were recommended. Genotypes that exhibited high resistant traits with low yield potential were recommended for breeders to incorporate those traits into susceptible genotypes with high yield potential.