Genetic variation, breeding potential and cytogenetic profile of Ethiopian tetraploid wheat (Triticum turgidum L.) landraces

The over-all aim of this thesis is to provide som base-line knowledge for efficient utilisation of the indigenous tetraploid wheat landraces of Ethiopia for breeding purposes. Diversity of 37 populations, collected from the four major production regions, was estimated by the Shannon-Weaver Diversity Index (H') using five qualitative traits. For populations, H' was ranged from completely monomorphic to highly polymorphic. For the entire collection, H' was high and reached maximum under relatively favourable climatic conditions. Most traits were not conspicuously unique to any single region or altitude group, but glume colour and spike density may be considered varietal features for specific localities. Sixty randomly selected pure-lines were evaluated for 12 quantitative traits. Significant genotype mean differences were observed. However, not all traits showed high coefficients of variation and genetic advance. Contrasting merits were found between landraces and the exotic variety, 'Boohai'. Generally, it may be possible to improve the landraces by direct selection for grain yield or indirect selection for kernel weight, but this approach is limited and temporary. Therefore, crossing landraces with exotic germplasm, particularly to improve their fertility, should be given more emphasis. However, a method to come up with varieties that possess compromised characteristics of the two germplasm groups has to be envisaged. There is a considerable chromosomal divergence in the landraces due to reciprocal translocations. Most of them might have noncentrometric breakpoints. C-band polymorphism was observed for all chromosomes except 1A and 2B. New C-bands and banding patterns were also observed. Two of these could have resulted from translocations between homoeologous chromosomes of the B and D genomes. This is conditional evidence for the introgression of the Q-factor complex from hexaploid wheat into the tetraploids that, consequently, posed taxonomic difficulties. From breeding perspective, hybridisation schemes involving the landraces need to be followed by bulk-selection until meiotic stability is more prevalent. Simple crosses alone may not be enough so that backcrossing is necessary.