Breeding Techniques

Most of the traits we manage are economically important and very complex in their genetic nature. Multi-stage and multi-trait selection across environments is essential for corn genetic improvement at NDSU. The NDSU corn breeding program utilizes a pedigree selection program for inbred line development with early (S1 and S2) and late generation testing (S3 and so forth) for combining ability. Early selections are evaluated per se across locations for early vigor, emergence percentage, disease resistance (through a disease nursery), and for several female and male traits. Germplasm adaptation is carried out with stratified mass selection and different type of backross systems (not used extensively) on exotic introductions. Intra and inter-population recurrent selection methods are utilized for germplasm improvement of genetically broad-based populations. The step number one in germplasm improvement is the utilization of several mating systems (e.g. diallel, design II) for progeny seed production (e.g. half-sibs, full-sibs, etc). As a second step, progenies are evaluated across environments (e.g. usually in partially balanced lattice designs or augmented designs). Since we are selecting several traits at the same time we use index selection (e.g. rank-summation index, heritability index). The third step involves recombining selected progenies for the next cycle of selection (e.g. bulk-entry method). Top progenies are included in our modified pedigree selection program for testcross (tc) and single-cross (sc) trials as shown below.

In cooperation with industry, we utilize to a lesser degree double haploid and molecular breeding methodologies.

Plant Breeding and Quantitative Genetics

Plant breeding works toward the identification of superior genotypes or combination of genotypes. The principles used to identify superior genotypes are similar across species.

Plant breeding was initiated as soon as we discovered the importance of plants as source of food, fiber, and fuel. Civilizations were established when plants were adapted and improved to specific environments and old improvement methods were very effective based upon the germplasm available today. Improvement was based on traits controlled by a few genes and was effective for increasing the frequency of alleles controlling their expression. Visual selection for traits that were desirable (e.g. cultural reasons) was emphasized until the end of the 19^th century and plant breeding methods were improved as information became available on the inheritance of traits and mode of plant development and reproduction. However, the scientific basis of plant breeding started in the 1900s. The rediscovery of Mendelian genetics and the development of the statistical concepts of randomization and replication had considerable impact on plant breeding methods. They provided a genetic basis for the variation observed among individuals and valid experimental techniques for measuring those differences.

Breeders work with traits and environments and their task is to increase the frequency of favorable alleles of traits that are controlled by a large, unknown number of genes. Breeders decide which combination of traits and environments are needed to breed for. A trait controlled by a few genes that are not affected by the environment can be improved very effectively. However, most traits that concern the breeder are economically very important and are quantitative in nature. Quantitative traits are controlled by a large, but unknown, number of genes, each having a small effect on the total expression of the trait. The environment in which they are measured determines their effects. They are characterized by degree differences among phenotypes that do not fall into distinct categories. These traits have a complex inheritance including dominance, epistasis, linkage, and the interaction of genetic and environmental effects. Quantitative traits are controlled by the joint action of many genes and genetic improvement of plants has been successful even though the knowledge of genes controlling these traits was minimal. Knowledge of the number of genes underlying quantitative traits is assumed to be large and usually estimates for the number of loci controlling a trait do not provide information on the presence of favorable alleles.