Genomic Selection for F Hybrid Breeding in Strawberry ( × ).

Cultivated strawberry is the most widely consumed fruit crop in the world, and therefore, many breeding programs are underway to improve its agronomic traits such as fruit quality. Strawberry cultivars were vegetatively propagated through runners and carried a high risk of infection with viruses and insects. To solve this problem, the development of F hybrid seeds has been proposed as an alternative breeding strategy in strawberry. In this study, we conducted a potential assessment of genomic selection (GS) in strawberry F hybrid breeding. A total of 105 inbred lines were developed as candidate parents of strawberry F hybrids. In addition, 275 parental combinations were randomly selected from the 105 inbred lines and crossed to develop test F hybrids for GS model training. These populations were phenotyped for petiole length, leaf area, Brix, fruit hardness, and pericarp color. Whole-genome shotgun sequencing of the 105 inbred lines detected 20,811 single nucleotide polymorphism sites that were provided for subsequent GS analyses. In a GS model construction, inclusion of dominant effects showed a slight advantage in GS accuracy. In the across population prediction analysis, GS models using the inbred lines showed predictability for the test F hybrids and vice versa, except for Brix. Finally, the GS models were used for phenotype prediction of 5,460 possible F hybrids from 105 inbred lines to select F hybrids with high fruit hardness or high pericarp color. These F hybrids were developed and phenotyped to evaluate the efficacy of the GS. As expected, F hybrids that were predicted to have high fruit hardness or high pericarp color expressed higher observed phenotypic values than the F hybrids that were selected for other objectives. Through the analyses in this study, we demonstrated that GS can be applied for strawberry F hybrid breeding.