2019 ASHS Annual Conference

Genome-wide selection (GWS) is being increasingly utilized in animal and plant breeding to improve genetic gains for polygenic traits, but implementation has generally lagged behind in fruit breeding programs. The strawberry breeding program at the University of Florida began exploring this methodology five years ago as part of the RosBREED project, using trials of advanced selections to train statistical models that were validated in future years trials. In 2015, GWS for parent selection was integrated into the breeding program. Since that time, various improvements in methodology have been implement, and practical applications have been made. First, prediction of genomic expected breeding values of first-year seedlings, prior to their inclusion in clonally replicated trials in the following year, allowed early parent selection for a subset of crosses. This is accomplished in practice by genotyping advanced selections several months prior to their inclusion in clonally replicated trials, generating performance predictions for multiple traits, and selecting a few predicted top-performers for immediate inclusion in crosses. This has increased genetic gains for traits such as fruit size and soluble solids content by reducing the average length of the breeding cycle. Second, once phenotypic data is available, the prediction of breeding values is now done with marker-based relationship matrices instead of pedigree data. This translates into estimates of genetic values with higher precision. Third, phenotypic and marker data from each year has been added to the training population to predict genetic values for the subsequent cycle. This increases the size of the training population each cycle, which has resulted in higher predictive abilities leading to higher genetic gains. Fourth, recent research on GWS to choose the best seedlings from biparental crosses has been successful. Not only are substantial genetic gains above family means possible, but only 500 markers are needed. Thus, within-family seedling GWS can be implemented to increase genetic gains for polygenic traits if cost-effective, low-density genotyping can be developed for strawberry.