Crop improvement in the 21st century.

Crop yields increased dramatically in the 20th century as recorded on Broadbalk or in world averages. The vast majority of that increase has occurred since the last world war and has been powered by changes in the genetic potential of the crop and in the way in which it has been managed. Nevertheless, the challenge to feed a world population that is likely to rise to 8 billion is formidable, particularly since recent analyses suggest that the rate of increase in yields of several crops may have dropped over the last decade. What are the opportunities to meet this challenge and to continue to improve the yields of our crops? Improvements in agronomy are likely to be more concerned with efficiency and elegance rather than in major breakthroughs. More sophisticated crop protection chemicals designed on the basis of vastly increased screening potentials and (at last?) possibilities of rational design will be supplemented by a battery of decision support systems to aid management choices which can be precisely implemented. Genetic improvement is the area in which to-look for the major breakthroughs. The broad potential of recombinant DNA technology will provide the possibility of both molecular analyses of crop productivity and ways in which it may be possible to improve that productivity. The goal of analysis may be approached in three ways: starting at the beginning by generating complete sequences of the plant genome; starting at the end by genetic analysis of phenotypes using genetic marker technology; or, starting in the middle, by metabolic analysis. Improvements may be obtained by re-assorting what has been achieved through enhanced breeding technologies, by randomly induced change, and by generation of totally new possibilities through biochemical engineering. Examples of all approaches will be given. The onset of genomics will provide massive amounts of information, but the success will depend on using that to improve crop phenotypes. The ability to meet the challenges of the 21st century will depend on the ability to close that 'phenotype gap'.