Utilizing genomic prediction to boost hybrid performance in a sweet corn breeding program.

Area of Science:

• Plant breeding
• Quantitative genetics
• Genomics

Background:

• Sweet corn breeding relies on developing elite inbred lines for commercial hybrids.
• Genomic selection (GS) offers in silico prediction of hybrid performance, potentially enhancing breeding programs.

Purpose of the Study:

• To evaluate the effectiveness of genomic selection for hybrid prediction in sweet corn.
• To compare different genomic prediction models and cross-validation schemes.

Main Methods:

• Evaluated 506 hybrids across six environments over two years.
• Assessed 20 plant, ear, and flavor traits.
• Compared GBLUP and RKHS models with additive and additive+dominance kernels in single- and multi-trait frameworks using three cross-validation schemes.

Main Results:

• Genomic prediction models demonstrated good trait prediction accuracies.
• GBLUP consistently outperformed RKHS across all tested schemes.
• Additive plus dominance kernels showed marginal improvements over additive kernels in some models.
• Within-site across-year and across-site models performed better with the CV0 scheme than CV00.

Conclusions:

• Genomic selection, particularly using the GBLUP model, is a reliable tool for sweet corn hybrid prediction.
• Implementing GS can optimize the testcross stage, identifying superior hybrids for advanced testing and increasing genetic gain.