Abstract only

Presented at International Conference on Integrated Approaches to Improve Crop production Under Drougt-Prone Environments (INTERDROUGHT), 3; Shangai, China, 11-16 Oct, 2009

Screening for drought tolerance (DT) means evaluating genotypes, alleles, or transgenes for yield under conditions analogous to natural stress in the target population of environments (TPE). By definition, droughts are unpredictable, causing practical problems in developing screens. Breeders and geneticists must (1) ensure that values of genotypes or alleles estimated in screens are predictive of those under natural stress, (2) ensure that stress means are estimated with adequate precision for selection or QTL mapping, and (3) for cultivar development, decide how to weight stress versus non-stress (and secondary trait) means in selection decisions. Large commercial breeding programs have relied on extensive multi-location testing, identifying DT cultivars at locations with natural drought stress. This ensures that over many years, relevant stresses are appropriately represented, but is too expensive for most breeding programs, genetic analysis, and transgene evaluation. The CIMMYT maize and IRRI rice breeding and drought genetics programs rely mainly on managed-stress screening (MSS) for DT in dry seasons. MSS improves reliability of stress application but introduces other problems, including screening in out-of-season conditions that are unrepresentative for temperature and daylength, generation of data out of sequence with breeding program advancement needs, and inappropriate weighting of means from stress and non-stress trials. Addressing these problems is becoming urgent in maize with the advent of genomic selection, adopted in CIMMYT?s DT breeding effort due to the highly polygenic nature of the trait and the failure to detect many useful large-effect QTL. Generally, MSS in maize and rice has proven adequately repeatable on a single-site basis, and CIMMYT and IRRI breeding programs have demonstrated that gains in DT from MSS are expressed in the TPE. However, empirical evidence on genetic correlations between traits in MSS at particular sites and in the TPE is urgently needed. Large-effect QTLs influencing yield under stress have been detected in MSS in rice, usually with greater precision and at lower cost than that for QTLs affecting other drought-related traits. The IRRI drought genetics program has been successful in identifying and coarse-mapping such QTLs, and in physiological analysis and elucidating the basis of allele effects on yield. Fine-mapping and cloning of these loci must be accelerated. For this, facilities permitting more reliable and precise phenotyping are needed.

Conservation Agriculture Program|Research and Partnership Program

Text in English

INT1888|INT3064

CIMMYT Staff Publications Collection