CAB International 2013. Realizing Africa's Rice Promise (eds M.C.S. Wopereis et al.) 311than white rice in Kumasi (the second largest city in Ghana), but this is not true for other markets in the country (Sakurai et al., 2006). Nigerians generally prefer cooked rice of a 'harder' texture than consumers in neighbouring countries. Watanabe et al. (2002aWatanabe et al. ( ,b, 2006 showed that many grain quality characteristics such as grain breakage, grain whiteness after milling, and protein and amylose content were significantly different among 47 rice cultivars cultivated under wet-and dry-season growing conditions in Côte d'Ivoire. In a different trial, where effects of different harvesting dates on milling and related traits were examined in several varieties, interactive effects of harvesting dates and varieties were significant for husking recovery and head-rice ratio, while no significant interactive effects were observed in grain dimensions, milling recovery and chalkiness (Futakuchi et al., 2001 -see definitions of grain quality below). These results illustrate that grain quality is a function of the variety (intrinsic factors) and the production and processing environment (extrinsic factors). It also underlines the importance of conducting grain quality assessments under standardized cultivation and postharvest practices.Enhancing grain quality of locally produced rice in Africa needs to consider: (i) clear grain

Background: Reproductive-stage drought stress is a major impediment to rice production in rainfed areas. Conventional and marker-assisted breeding strategies for developing drought-tolerant rice varieties are being optimized by mining and exploiting adaptive traits, genetic diversity; identifying the alleles, and understanding their interactions with genetic backgrounds for their increased contribution to drought tolerance. Field experiments were conducted in this study to identify marker-trait associations (MTAs) involved in response to yield under reproductive-stage (RS) drought. A diverse set of 280 indica-aus accessions was phenotyped for ten agronomic traits including yield and yield-related traits under normal irrigated condition and under two managed reproductive-stage drought environments. The accessions were genotyped with 215,250 single nucleotide polymorphism markers. Results: The study identified a total of 219 significant MTAs for 10 traits and candidate gene analysis within a 200 kb window centred from GWAS identified SNP peaks detected these MTAs within/ in close proximity to 38 genes, 4 earlier reported major grain yield QTLs and 6 novel QTLs for 7 traits out of the 10. The significant MTAs were mainly located on chromosomes 1, 2, 5, 6, 9, 11 and 12 and the percent phenotypic variance captured for these traits ranged from 5 to 88%. The significant positive correlation of grain yield with yield-related and other agronomic traits except for flowering time, observed under different environments point towards their contribution in improving rice yield under drought. Seven promising accessions were identified for use in future genomics-assisted breeding programs targeting grain yield improvement under drought. Conclusion: These results provide a promising insight into the complex genetic architecture of grain yield under reproductive-stage drought in different environments. Validation of major genomic regions reported in the study will enable their effectiveness to develop drought-tolerant varieties following marker-assisted selection as well as to identify genes and understanding the associated physiological mechanisms.