GENETIC AND PHENOTYPIC DIVERSITY IN SORGHUM FOR IMPROVEMENT AS A BIOFUEL FEEDSTOCK

Abstract

Sorghum [Sorghum bicolor (L.) Moench] is the sixth most important grain crop worldwide. In addition to grain varieties, diversifying selection in sorghum breeding has produced elite types for forage, silage, sugar, syrup, fiber, and broom production. Sorghum, however, has yet to undergo selection for whole plant energy production. Recent initiatives and improvements in grain, stem sugar, and lignocellulosic ethanol suggest that all plant parts may be valuable for energy production in the future. The primary focus of this dissertation was to understand the genetics of tradeoffs in yield and composition of grain, stem sugar and lignocellulosic biomass, and identify genetic diversity in grain and sweet sorghum types. A biparental recombinant inbred line population, derived from a cross between a sweet and a grain sorghum, was used to identify QTL for non-structural carbohydrates in the first study and structural carbohydrates in the second. Including both studies, a total of 303 QTL were identified across 69 reported traits. From these studies, a major finding was that tradeoffs are minimal between carbohydrates under favorable conditions. This supports a sink limited model of whole plant energy production. Identified tradeoffs co-localized with QTL for height, flowering time and stand density suggesting some pleiotropic effects. Another important finding in the second study was that the genetic control of structural composition was found to be different between stem and leaf tissue. This was also true for protein, with QTL for grain, stem and leaf crude protein failing to co-localize. These results suggest independent selection on tissues would be most effective. Additional products of these first two studies were a genetic map and NIRS equations for grain, stem and leaf tissue. The third study used a diverse panel of 125 sweet, grain and landrace sorghums to examine genetic relationships within the sweet sorghums and between sweet and grain sorghums. Using principal coordinate analysis, three main populations within sweet sorghum were identified, syrup types, sugar and energy types, and amber types. These had some correspondence with grain races kafir, bicolor and caudatum. Association mapping detected four major height QTL, and one QTL for brix.