Improvement of crop yield in dry environments: benchmarks, levels of organisation and the role of nitrogen.

Crop yield in dry environments can be improved with complementary approaches including selecting for yield in the target environments, selecting for yield potential, and using indirect, trait- or genomic-based methods. This paper (i) outlines the achievements of direct selection for yield in improving drought adaptation, (ii) discusses the limitations of indirect approaches in the context of levels of organization, and (iii) emphasizes trade-offs and synergies between nitrogen nutrition and drought adaptation. Selection for yield in the water- and nitrogen-scarce environments of Australia improved wheat yield per unit transpiration at a rate of 0.12kg ha(-1) mm(-1) yr(-1); for indirect methods to be justified, they must return superior rates of improvement, achieve the same rate at lower cost or provide other cost-effective benefits, such as expanding the genetic basis for selection. Slow improvement of crop adaptation to water stress using indirect methods is partially related to issues of scale. Traits are thus classified into three broad groups: those that generally scale up from low levels of organization to the crop level (e.g. herbicide resistance), those that do not (e.g. grain yield), and traits that might scale up provided they are considered in a integrated manner with scientifically sound scaling assumptions, appropriate growing conditions, and screening techniques (e.g. stay green). Predicting the scalability of traits may help to set priorities in the investment of research efforts. Primary productivity in arid and semi-arid environments is simultaneously limited by water and nitrogen, but few attempts are made to target adaptation to water and nitrogen stress simultaneously. Case studies in wheat and soybean highlight biological links between improved nitrogen nutrition and drought adaptation.