Risk and uncertainty as factors in crop improvement research.

The organization and productivity of a crop improvement system is examined in the light of the environmental variability in which the crop is grown. The trade-offs between wide adaptability and genotype tailoring are discussed. Adaptability is defined as the performance of a genotype with respect to environmental factors that change across locations, and stability as genotype performance with respect to environmental factors that change across time within a given location. Research is considered as a systematic search for improved genotypes within a distribution of potentially discoverable genotypes. Genotype-environment interactions reduce productivity of research outside its primary target environment, while spill-over effects work in the opposite direction. An optimal degree of clustering of environments will maximize research productivity. Analysis of International Rice Yield Nursery (IRYN) and All India Coordinated Rice Improvement Program (AICRIP) data provide conflicting evidence on the correlation of adaptability and stability, perhaps because. the AICRIP data are clustered while the IRYN data are not