The improvement of leaf rust resistance in selected bread wheat lines

English: The objective of this study was not only to improve leaf rust (caused by Puccinia triticina) resistance in selected wheats (Triticum aestivum L.), but to focus on durability as well as agronomic acceptability of resistant lines. This was achieved by traditional breeding techniques as well as with the use of AFLP analysis. Seven bread wheat lines were obtained from ARC-Small Grain Institute (SGI) where they were developed. Six leaf rust resistance sources were obtained from the University of the Free State. The breeding strategy focused on creating lines that contained both seedling and adult plant resistance genes. Lr34 was chosen as the adult plant resistance source as it is an important gene due to its durability and interaction with other leaf rust resistance genes. It is also an easy gene to follow as it is associated with leaf tip necrosis. In an attempt to create genotypes that will remain durable, four seedling leaf rust resistance genes (Lr21, Lr32, Lr36 and Lr41) were each combined with Lr34 in the seven SGI backgrounds. By combining seedling and adult plant genes, and selecting only the lowest infection types throughout all phases of testing, a significant shift in the leaf rust resistance of the population occurred. Greenhouse evaluations of plant architecture and agronomic performance of lines containing both a seedling gene and Lr34, indicated that individual plants, similar or better than the original SGI parent, had been selected. Several of these selections can therefore be incorporated in larger, mainstream breeding programmes. It is, however, imperative that the agronomic and quality value of selections be determined under field conditions. More efficient manipulation of an undesignated Lr gene was also attempted. This gene, which originates from T. monococcum, was incorporated in certain bread wheat lines and cultivars. Since the gene is characterized by an immune response to South African pathotypes of leaf rust, it is considered valuable in terms of rust resistance. However, its single gene nature implies it will not remain durable and that it has to be protected in complex resistance gene combinations. To achieve this, linked molecular markers are needed. In this study a putative AFLP marker for leaf rust resistance was generated by S12/M14 and S12/M44 using pooled DNA. Analysis of individual plants from which the DNA bulks were constructed indicated that the marker might not be closely linked to the Lr gene of interest. Validation in three additional backgrounds revealed several instances of recombination between the marker and the gene, emphasising the need to do a proper linkage study. Should the linkage distance be acceptable, the marker might still be useful as it proved to be polymorphic in five different wheat backgrounds. The linkage distance is also needed before the intensive effort to clone and convert the AFLP band to a more user-friendly STS marker will be considered.