Opportunities for Applied Marker-Assisted Breeding in Sorghum: Moving towards the Next Green Revolution.

Development of sorghum (Sorghum bicolor) cytoplasmic-genetic male-sterility systems in the 1950s first permitted commercial cultivation of F1 hybrids. Subsequent diversification of elite genepools with tropically-adapted zera-zera germplasm from Eastern Africa allowed development of food and feed cultivars (both hybrids and OPVs) with higher grain yield potential. However, since then most applied sorghum improvement globally has focused on enhancing resistance to biotic production constraints (e.g. diseases, insect pests, and Striga), or tolerance to abiotic stresses (e.g. drought and acid soils), with limited further enhancement of grain or stover yield potential. Over the past two decades, development of molecular genetic tools has consumed a large portion of global research resources available for sorghum improvement. Development of many different types of molecular markers and establishment of marker-trait associations in sorghum has resulted in many publications, but few visible applied products. Biofuel potential of sweet sorghum recently led to development of an aligned sorghum genome sequence – perhaps the ultimate genetic tool if we can use it efficiently. Public research programs in the USA and around the world are now striving to use genetic resources and these biotechnological tools to improve the efficiency of sorghum breeding programs, and emerging opportunities to exploit next generation sequencing approaches for "genotyping-by-sequencing" are expected to make such research even more attractive. Recent successes include SSR-based genetic diversity assessment of representative sets of national, regional and global sorghum germplasm. This information can now be used to sample germplasm for association mapping, identify parents for more conventional mapping studies, or hunt for improved heterosis. QTL mapping, marker-assisted backcrossing, and phenotypic assessment of MABC products to validate QTLs have become almost routine. There are opportunities now to exploit genetic male-sterility combined with marker-assisted backcrossing to quickly complete single- or double-QTL introgressions (by harvest of selected selfed BC2F2 plants), and use high-density genotyping to improve the effectiveness of population improvement while “mapping-as-you-go”. As marker datapoint costs continue to decline we will be able to efficiently improve our most elite materials for any heritable trait we can reliably phenotype, and should finally start reaping the over-due rewards for our long-term investments in sorghum molecular genetic tools.