2018 ASHS Annual Conference

Genomic selection (GS) has been demonstrated as an effective method for increasing genetic gains for traits with complex genetics in animals and plants. However, its application in horticultural crops has been limited. The application of GS in the University of Florida strawberry breeding program was evaluated in a series of annual trials established from the 2013-2014 season until the 2017-2018 season at the Gulf Coast Research and Education Center in Balm, Florida. These trials comprised more than 2200 field-tested seedlings and selections, which were genotyped using Affymetrix SNP Axiom^® arrays. Differences among GS methods were found, and on average the predictive ability (PA) of genomic breeding values (GBV) for true validations, across methods and across years, was moderate to high ranging from 0.3 for total marketable yield to 0.5 for average fruit weight. As data were aggregated from multiple breeding cycles in the training population, the PA increased for all traits in test populations. Between 500 and 1000 randomly selected markers resulted in more than 85% of the PA obtained with 8000 markers. In addition, within-family selection of seedlings generated gains from 4% to 16% above the family mean by selecting the top 10% of predicted individuals within families. Genomic selection has been effectively integrated into the UF strawberry breeding program for parental selection, decreasing the length of the breeding cycle, with strong potential for additional gains via seedling selection if cost-effective, low-density, genotyping can be obtained.