Location: Hard Winter Wheat Genetics ResearchTitle: Genomic selection in preliminary yield trials in a winter wheat breeding program Author
|Belamkar, Vikas - UNIVERSITY OF NEBRASKA|
|Hussain, Waseem - UNIVERSITY OF NEBRASKA|
|Jarquin, Diego - UNIVERSITY OF NEBRASKA|
|El-basyoni, Ibrahim - NATIONAL CENTER FOR AGRICULTURE AND FORESTRY TECHNOLOGIES (CENTA)|
|Poland, Jesse - KANSAS STATE UNIVERSITY|
|Lorenz, Aaron - UNIVERSITY OF MINNESOTA|
|Baenziger, P. Stephen - UNIVERSITY OF NEBRASKA|
Submitted to: G3, Genes/Genomes/Genetics
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 6/19/2018
Publication Date: N/A
Interpretive Summary: Yield testing in small plots is the principal bottleneck of wheat breeding programs, requiring a substantial investment of program resources. The requirement for timely harvest and the limited throughput of research combines restricts the number of selections a program can successfully test. Differences in performance of selections in different locations compounds the challenge to breeders. A high quantity of DNA sequence information can be gathered on selected lines at a fraction of the cost of executing an individual yield trial plot. Wheat breeders, therefore, are turning to DNA sequence information, combined with yield data on a subset of selections, to predict performance of untested materials, an approach known alternatively as Genomic Selection or Genomic Prediction. This study evaluated the application of this genomic-assisted approach to the working breeding program at the University of Nebraska. The results of this evaluation indicate that the breeding program could operate efficiently by testing 50% of its selections in any given environment. This genomics-assisted approach also can accelerate the return of high performing selections to the crossing program. And the genomics-assisted approach can ensure that valuable selections are not lost due to adverse weather, such as hail storms, or unusual disease epidemics.
Technical Abstract: Genomic selection (GS) is now routinely performed in plants to test the effectiveness of GS for predicting phenotype using internal cross-validation experiments. However, few studies put the results into practice. Here, we examine evaluate the performance of GS for predicting grain yield and determine prospects for cultivar development. This study comprised examined four nurseries of F3:6 and F3:7 evaluated trialed at 6 to 10 locations each year. Grain yield was analyzed using mixed models that accounted for experimental design and spatial variations. Genotype-by-sequencing provided nearly 27,000 high-quality SNPs. Average genomic prediction ability (PA), estimated for each year, by randomly masking lines as missing in steps of 10% from 10% to 90%, and using the remaining lines as a training dataset, ranged from 0.229 to 0.552. The PA estimated for a new year using the other years ranged from 0.167 to 0.282. Further, we tracked lines advanced based on phenotype from each of the four F3:6 nurseries. Lines with both above average genomic estimated breeding value and phenotypic value were retained for longer times in the breeding program. The number of lines selected for advancement was substantially greater when predictions were made for 50% of the lines in each year. Hence, evaluation of only 50% of the lines yearly seems possible. Correlation between F3:6 predictions and phenotypes with their corresponding F3:7 phenotypes showed GS outperformed phenotypic selection in 2012 and 2015, emerged ~50% successful in 2014, and had lower performance in 2013. This study provides insights to assess and integrate GS in breeding programs of autogamous crops.