Evaluation of Partially Converted Lines from the Sorghum Conversion Program to Determine Combining Ability and Heterosis for Early Testing

Hybrid breeding technology has been very successful in increasing yield of many cereal crops including sorghum (Sorghum bicolor (L.) Moench). The process of developing and evaluating the performance of hybrids is the most expensive and time-consuming activity. In sorghum, developing parental inbred lines and evaluating their potential hybrid performance is very expensive and time-consuming. Predicting hybrid performance in any way possible might help reduce the number of crosses needed and evaluated. The study was comprised of 40 lines, selected by the percentage of exotic genome recovered and the generation (BCF2:3 high, BCF2:3 low, F2:3 high, and F2:3 low), and four elite testers. The 160 F1 hybrids and parents were evaluated in a randomized complete block design with two replications for combining ability and heterosis during the 2015-2016 cropping season. The results indicated that sufficient genetic variability was observed for all characters evaluated. Analysis of general combining ability (GCA) revealed that among the parents, lines RSC117-4 BCF2:3 high, RSC83-1 BCF2:3 low, RSC83-14 BCF2:3 high, and RSC112-19 BCF2:3 high, and the testers A.301 and A.319 were promising general combiners for increasing grain yield and most other important traits. Therefore, the parents can be used in sorghum breeding programs aimed toward developing high-yielding hybrids. A.301RSC83-14 BCF2:3 high produced the third largest overall mean grain yield for hybrids, yielding slightly more than commercial check 301/41, with 67.3 days to anthesis and 127.5 cm total plant height. The hybrid had significant midparent heterosis (MPH) with a positive specific combining ability (SCA) effect and GCA combining effect combination of high x high which provides complementary gene action. The hybrid had good combination for earliness, dwarfing genes, and large yield. A.319RSC83-1 F2:3 high had the largest overall mean for grain yield, out yielding four of the best six commercial check hybrids, with 72.2 days to anthesis and 153.0 cm total plant height, with a positive MPH and SCA effect with high x low GCA effects combination indicating additive x dominance type of gene interaction. The cross could produce desirable transgressive segregates because of the additive genetic system in one general combiner of the parent and complementary epistatic effects in the other. The hybrid was taller and yielded more, with heavier panicle weights and 1000-kernel weight, and significantly greater MPH per se parents. Line x tester analysis revealed that the contribution of the lines to the total sum of squares was greater than testers in all studied traits except the number of days to anthesis. The variance due to GCA (σ²gca) was less than for SCA (σ²sca) for all traits except plant exsertion and concentration of protein, starch, fiber, and fat in grain, suggesting preponderance of non-additive gene action controlling the characters. Dominance variance (σ²D) was greater than additive variance (σ²A) for all traits except plant exsertion and concentration of protein, starch, fiber, and fat in the grain. The results are supported by the ratio of variance of general to specific combining ability (σ²gca/ σ²sca) which was smaller than unity and by the degree of dominance (σ²D/σ²A) that uses values greater than unity for all traits except plant exsertion where σ²A was larger than σ²D, with the degree of dominance being less than unity. Selections based on the predominance of dominance variance suggested the exploitation of heterosis.