ABSTRACT
To avoid a net expansion of harvested cropland, it is proposed crop yields must grow by at least a third more in the next 45 years than they did in the previous half-century while land sparing strategies will be needed to minimize the negative impacts of food production on biodiversity. To achieve this, plant breeding programmes should use the power of selection to develop phenotypically and genetically diverse high-yielding populations. High crop yields can be achieved either by increasing ‘yield potential’ through plant breeding, or by reducing ‘yield gaps’ by developing cultivars with improved adaptation to stressful environments. Breeding gains are believed to be feasible due to the existence of genetic variation for most agronomic traits, including variation in crop wild relatives, landraces, or other available bred-germplasm. Improved knowledge of how genetic diversity of crops is distributed within resources of these kinds will facilitate the selection of suitably diverse parents in breeding crosses while advances in automated plant phenotyping and phenomics will allow rapid screening of traits in the resulting offspring. The use of high-throughput DNA marker tools, and associated bioinformatic analyses, is also facilitating the search for useful sources of variation for cross-breeding. This chapter provides an overview of the use of genetic diversity in plant breeding, including what can be learned from studies of crop domestication and evolution. The chapter concludes that it will be necessary to embrace a knowledge-led plant breeding approach to leverage agricultural biodiversity as effectively as possible.
