PREDICTING QUANTITATIVE VARIATION WITHIN RICE GERMPLASM USING MOLECULAR MARKERS

Diverse Asian rice (Oryza sativa) germplasm has been used to identify associations between various quantitative traits and RAPD molecular ma rkers using multiple regression analysis. This has allowed us to predi ct for other samples of germplasm their performance for traits such as culm length and number, days to flowering, grain width, and panicle a nd leaf length using only RAPD marker data. Such predictive capability is possible because of the availability of extensive diversity held i n genebanks, and can be used in the future to facilitate the exploitat ion of that biodiversity. More specifically the methodology could faci litate crop improvement by rapid ideotype prediction. For the mapping and isolation of QTLs (genes controlling quantitative traits) the meth od would provide information to guide the selection of parental materi al for hybridization and markers expected to show linkage to QTLs. It may also be possible that these associations could lead the way toward s marker-assisted selection during breeding programs. In the future, t his demonstration of association between markers and easily measured t raits could also be extended to the study of important adaptive traits , such as stress tolerance, found either within germplasm collections or in natural populations.