Abstract
The loss of genetic diversity in bread wheat (Triticum aestivum L.) due to bottlenecks from polyploidy, domestication, and modern plant breeding can be compensated by introgressing exotic germplasm. A successful approach to capture genetic diversity is the production of primary synthetic bread wheats, which are contemporary reconstitutions of the ancestral genomes of wheat from diverse wild relatives. However, this diverse germplasm has many undesirable characters, making direct use in breeding programs difficult. To increase the speed of introgression of exotic germplasm, genomic selection approaches could be applied to enable rapid cycles of selection. To test this, selected lines from double haploid and recombinant inbred line populations between six different primary synthetics and the elite cultivar 'Opata M85' were evaluated for grain yield and other important agronomic traits. Field trials were conducted at CIMMYT (International Center for Maize and Wheat Improvement) over 2 yr in irrigated, heat, and drought-stressed environments. Several synthetic-derived lines outperformed the elite parent Opata M85 in all environments, indicating that the primary synthetics contribute alleles increasing yield. Whole-genome profiles were generated using genotyping-by-sequencing. Five different whole-genome prediction models that can be applied for genomic selection were evaluated for prediction accuracy using cross validation. Overall, the prediction models had moderate predictive ability, slightly lower than expected based on the heritability of the traits. While a viable approach, rapid cycle genomic selection for introgression of exotic alleles might be restricted with synthetic-derived wheat due to complex and confounding physiological effects.