Genome-wide association studies and genomic prediction for key ornamental traits in tetraploid pot carnation

Abstract Background Carnation ( Dianthus caryophyllus L.) is one of the most commercially important ornamental crops worldwide. Similar to many ornamental species, genomic selection has yet to be widely adopted as a breeding tool in carnation. Moreover, the widespread development and commercialization of polyploid carnation cultivars makes their genomic study particularly relevant. In this study, we investigate the genetic architecture of key ornamental traits in a tetraploid pot carnation diversity panel using genome-wide association studies (GWAS), and evaluate the potential of genomic prediction for ornamental trait improvement in this panel. For genomic prediction, we assessed three SNP marker encodings capturing different allele dosage information, a pseudo-diploid encoding, a tetraploid additive dosage encoding, and an encoding accounting for both additive and non-additive genetic effects. In addition, we carried out a GWAS-assisted genomic prediction by fitting top-scoring GWAS markers as fixed effects to the genomic prediction model. Results Through GWAS, we identified potential genomic regions of interest associated with seven of the eight traits studied. Genomic prediction achieved moderate to high prediction accuracy across traits and encodings, ranging on average from 0.54 to 0.82 for flowering-related traits (earliness, flower filling, flower count, and flower size) and from 0.39 to 0.48 for plant architecture traits (branching, habit, plant height, and vigour). Fitting top-scoring GWAS markers as fixed effects provided only modest gains in prediction accuracy (up to 0.03) for three of the eight traits (earliness, flower filling, and branching), with no improvement for the remaining traits under any of the three encodings. Conclusions These findings extend the knowledge on the genetic basis of key ornamental traits and attest to the potential of genomic selection for breeding tetraploid pot carnation.