Lecture 5 Classical and Molecular Approaches to Plant Improvement

The genetic aspect of plant improvement involves the identification and generation of superior genotypes. This might involve selecting the best single line from a series of pure lines, or increasing the frequencies of favorable alleles in a segregating population, or transferring desirable alleles from a line or wild stain into existing elite germplasm, or selecting the best lines to cross to produce superior hybrids. While there is a rich history of quantitative-genetic methods for all of these avenues of improvement using solely phenotypic information, many (but not all) of these can potentially be enhanced by using molecular information. We start with a brief review of some of the purely phenotypic approaches (using trait measurements on collections of known relatives), and then examine where molecular methods can augment our efficiency in creating superior genotypes. A key point to stress is that phenotypic approaches have a long history of remarkable success, and one should not simply chose a molecular-base approach without reason. Often far more improvement can occur by devoting the funds for a molecular study into a larger phenotypic selection design (such as using the extra resources to test more lines and/or more environments). Our next generation of plant breeders (or improvement engineers) must not only have a mastery of molecular biology and genomics/metabolimics/proteomics, but also a deep understanding (and appreciation) of this rich history of phenotype-based approaches. Fortunately, many of the statistical tools of the modern plant breeder are applicable to many -omics problems in addition to strictly phenotypic-based approaches. Indeed, many of the statistical tools used for the analysis of complex -omics problems have their roots in historical tools used by generations of breeders.