|dc.description.abstract||The increase in energy demand, volatile oil prices and climate change has led South Africa to reduce its dependency on fossil fuels and promote biofuels. Sweet sorghum [Sorghum bicolor (L.) Moench] has been considered as one of the promising crops due to its sugar-rich stalk to supplement sugarcane which is the major feedstock for bioethanol. Establishing genotypic variability for biomass yield and sugar-related traits in sweet sorghum is therefore essential for developing superior cultivars. The objectives of the study were: (i) to assess sweet sorghum lines for agronomic performance and genetic diversity using quantitative morphological traits and (ii) to assess sweet sorghum lines for genetic diversity and interrelationships using simple sequence repeat (SSR) markers.
Twenty-five sweet sorghum lines collected from International Crops Research Institute for the Semi-Arid Tropics (ICRISAT-Kenya) and the African Centre for Crop Improvement (ACCI-South Africa) were evaluated during the 2015/2016 season in KwaZulu-Natal (KZN) province at Ukulinga Research Farm in Pietermaritzburg and Makhathini Research Station in Jozini. Seven agronomic traits; fresh biomass yield, fresh stalk yield, grain yield, plant height, stalk diameter, panicle length and days to 50% flowering, and six quality traits; fibre, dry matter, °brix, °total brix, total fermentable sugars and ethanol were recorded. The sweet sorghum lines revealed highly significant variations for the 13 quantitative characters assessed in this study. The extent of variation was highly influenced by environment and genotype by environment interaction. Genotypes designated as IS 2331, IESV 92008 DL, ICSV 700, AS 244, URJA and SS 27 were identified as suitable genotypes with high plant height, dry matter, fibre, °brix, °total brix, total fermentable sugars and ethanol.
The specified genotypes also exhibited medium to late maturity with relatively high fresh biomass and fresh stalk yield. Genotype 91018 LT showed the highest fresh biomass yield, fresh stalk yield, stalk diameter and relatively high grain yield. High levels of trait heritability were observed for fresh stalk yield (98%), stalk diameter (93%), fresh biomass yield (81%), panicle length (76%), fibre (73%) and plant height (66%). Heritability estimates were influenced by the environment and genotype by environment interaction. Principal component analysis resulted in the first three principal components showing 83% of the total variability among the genotypes. Ethanol, total fermentable sugars, °total brix, fresh stalk yield and °brix contributed mainly to PC 1, whereas fresh biomass yield and stalk diameter contributed mainly to PC 2. The dendrogram generated from cluster analysis divided the genotypes into two main clusters
and three singletons (ICSB 324, ICSB 654 and ICSV 700). Cluster I comprised 54% of the total germplasm and included only one ACCI genotype (SS 27), while cluster II comprised of 33% of the total variation.
The morphological variability analysis of the genotypes was also complimented with the use of molecular markers. The 24 sweet sorghum lines were genotyped with 10 simple sequence repeat (SSR) markers and distance-based method was used to analyze the data. Variation was observed for all the markers with allelic size ranging from 1 to 36 bp. A total of 61 alleles were generated with an average of 6.1 alleles per locus. The polymorphism information content (PIC) values ranged from 0.32 to 0.86 with an overall mean value of 0.62, showing a high discriminating ability of the markers used. The largest genetic distance was observed for AS 244 (GD = 1.9), while IESV 92001 DL and IESV 92008 had the smallest genetic distance (GD = 0.50). The dendrogram generated from cluster analysis using SSR markers classified the 24 sweet sorghum lines into two major clusters. Cluster I comprised of 12.5% of the total genotypes which included URJA, SS 27 and ICSB 654. It was observed that all ACCI genotypes apart from AS 244 were grouped in Cluster I, with URJA and SS 27 being very closely related. Cluster II was observed to be the largest (87.5%) with 21 genotypes, which further formed 3 sub-clusters (A, B and C) and a singleton (AS 244). The results from molecular marker characterization were similar to those obtained using PCA analysis of morphological traits which grouped genotypes into four clusters, with the same type of genotypes in each group. The information obtained in this study coupled with phenotypic characterization can be used by plant breeders to select parents or pure lines that can be used in developing improved cultivars. This will therefore contribute to the production of sweet sorghum and promotion of its use for bioethanol in South Africa.||en_US