DEVELOPMENT OF GENOMIC TOOLS FOR ACCELERATED BREEDING OF CRESTED WHEATGRASS [Agropyron cristatum (L.) Gaertn.]

Crested wheatgrass [Agropyron cristatum (L.) Gaertn.], particularly valued as forage crop for its early spring vigor providing high quality, highly palatable forage before the native forage grasses. However, the nutritive value of the forage starts declining after heading and becomes less palatable to livestock. The objectives of this thesis research were: 1) To assess the genetic diversity and population structure of crested wheatgrass using genome-wide Single Nucleotide Polymorphism (SNP) markers generated using Genotyping-by-Sequencing (GBS); 2) To develop a high-resolution linkage map in an intraspecific F1 mapping population of crested wheatgrass; and 3) To study the prediction ability of genomic selection models in crested wheatgrass. Breeding initiatives for the development of high yield, high palatable and late maturing cultivars begins with uncovering of the extent of genetic diversity available and its utilization in the breeding program. Molecular characterization of un-sequenced plant species with complex genomes is now possible by GBS using recent next generation sequencing technologies (NGS). SNP markers were used to assess the genetic diversity present in 192 genotypes from 12 tetraploid lines of crested wheatgrass. The model-based Bayesian analysis revealed four major clusters of the samples assayed. The diversity analysis revealed 15.8% of SNP variation residing among the 12 lines. The principal coordinates analysis and dendrogram were able to distinguish four lines of Asian origin from Canadian cultivars and breeding lines. With an attempt to utilize SNP markers towards molecular breeding, a study to develop genetic linkage maps in Agropyron cristatum utilizing segregating F1 mapping population developed from intraspecific cross of two diploid elite cultivars grouped 678 SNP markers into seven linkage groups. The linkage map generated here could be saturated for quantitative trait loci (QTL) mapping and marker assisted selection (MAS). However, MAS is less effective for improvement of traits controlled by a large number of small effect QTLs. Genomic selection (GS) is gaining attention towards improving genetic gain of quantitative traits. Genomic selection study showed moderate prediction accuracies (in range of 0.20–0.41) for traits such as ADF, TPP, CD, PH, LW and FRG suggesting that it is possible to implement GS. The additive GS models in this study were similar in prediction suggesting prediction ability is less influenced by the model preference. Increasing SNP densities did not improved prediction ability except for few traits suggesting the traits could have been influenced by large number of small effect QTLs such that GBS approach adopted was unable to capture the effect across the genome. However, GS in crested wheatgrass can be improved with refining structure and size of training population, field experiments to accurately measure phenotypic records and utilize GS models to incorporate genotype-by-environment and gene interactions/ epistasis, for this comparison of parametric models with non-parametric models in GS could provide insight