Until recently, systematic breeding of sweet sorghum was sporadic with the USDA-ARS Meridian, Mississippi sweet sorghum breeding program being the only project devoted exclusively to this crop. Though it closed in 1988, many of the existing sweet sorghum varieties in the United States were derived from that program [86-88]. These varieties represent an array of different maturities, heights and agronomic packages adapted to regions as best identified by producers. They have become standards throughout the world and form the basis for sweet sorghum breeding programs that are being established in many regions of the world. With increasing interest in bioenergy, newer sweet sorghum varieties have been released in India and China [11].

Prior breeding efforts in energy and sweet sorghum focused on pure-line cultivar develop­ment. Sweet sorghum cultivars were selected from intentional crosses and advanced through several generations via self-pollination and selection to uniformity. Traits of importance included maturity, height, sugar yield, sugar concentration, sugar quality and agronomic adaptation to stalk rot, drought, and stem borers. Initial breeding efforts did not emphasize bagasse quality because it was of little importance to small-scale sweet sorghum producers. In an industrial setting, however, bagasse is substantially more important, as it represents an energy source to produce electricity or to be converted to ethanol itself once lignocellulosic conversion processes become profitable [43].

For reasons previously mentioned, hybrid sweet sorghum cultivars are crucial for indus­trial production systems. Because sugar concentration is a primarily additive trait, a true sweet sorghum hybrid is often difficult to produce, since most parental seed lines have low sugar concentrations in the juice and the resulting hybrids have intermediate sugar concentrations [89]. Consequently, even though juice volumes were heterotic, overall sugar yields were reduced relative to sweet sorghum varieties because of lower sugar concentra­tions in the parental seed lines. Therefore, it was necessary to develop sweet sorghum seed parents to produce true sweet sorghum hybrids. Development of pollinator parents was also important, but most sweet sorghum varieties can serve as pollinator parents as they typically restore fertility to their hybrids made using either A1 or A2 cytoplasm.

Sweet sorghum parental seed lines are being developed in several programs around the world [11]. Studies of these hybrids are superior to seed parents and numerically equal if not superior to pollen parents (i. e., sweet sorghum varieties) [45,46]. While first generation hybrids did not always outperform their respective pollinator parents, hybrid seed production capacity was four to six times greater than for a pure-line sweet variety and it was much easier to harvest. Furthermore, it is logical to expect that yield and quality of subsequent hybrids will be better than those currently available. In India, excellent progress has been made in developing hybrid sweet sorghum lines from sweet and grain parents. From the ICRISAT sorghum improvement program, Reddy et al. [90] described six hybrids significantly lower in brix than the control genotype that were nonetheless significantly higher in sugar yield than the control. Additionally, and not surprisingly, these six hybrids also produced significantly greater grain yields than the sweet sorghum control.

Hybrid sweet sorghum breeding methodologies follow traditional sorghum breeding approaches with modifications in traits and selection protocol. For example, because of the additive effect of sugar concentration, it is critical to select for sugar concentration in both seed and pollinator parents. However, juice volume is more of a dominant trait; consequently, it may be selected in either parent and will be expressed in the hybrid. Most breeding programs will use a pedigree approach followed by sterilization of the seed parent and test-crossing of both types on standard testers [65]. In addition, and just like grain crops, a range of hybrid maturities will be necessary to ensure an optimal distribution of hybrids for a continual harvest season [14].

For improvement purposes it is best to define total sugar yield using the individual components that contribute to it, which are juice yield and soluble sugar concentration. Juice yield is related to biomass yield, and thus total biomass yield is critical in sweet sorghum breeding [34]. Often, reports in the literature estimate sugar yield using total dry biomass yield and a coefficient for juice content and soluble sugar concentration reported in the brix units. However, juice extractability and the proportion of sugar to total soluble materials must be considered in selection, which if based on brix and/or moisture content alone could be misleading [45]. Lodging and stress tolerance are also traits of interest for sweet sorghum breeders insofar as they affect harvestability, stability and fermentable sugar yield. Finally, an important trait unique to sweet sorghum (as compared to grain sorghum) is the duration of optimal sugar yield in the hybrids. It is generally agreed that sugar yields peak in sorghum prior to physiological maturity and, therefore, if that yield can be maintained for a longer period it could extend the economic harvest season, which is crucial for development of a sustainable biofuel industry.