Abstract
BackgroundSweet cherry is consumed widely across the world and provides substantial economic benefits in regions where it is grown. While cherry breeding has been conducted in the Pacific Northwest for over half a century, little is known about the genetic architecture of important traits. We used a genome-enabled mixed model to predict the genetic performance of 505 individuals for 32 phenological, disease response and fruit quality traits evaluated in the RosBREED sweet cherry crop data set. Genome-wide predictions were estimated using a repeated measures model for phenotypic data across 3 years, incorporating additive, dominance and epistatic variance components. Genomic relationship matrices were constructed with high-density SNP data and were used to estimate relatedness and account for incomplete replication across years.ResultsHigh broad-sense heritabilities of 0.83, 0.77, and 0.76 were observed for days to maturity, firmness, and fruit weight, respectively. Epistatic variance exceeded 40% of the total genetic variance for maturing timing, firmness and powdery mildew response. Dominance variance was the largest for fruit weight and fruit size at 34% and 27%, respectively. Omission of non-additive sources of genetic variance from the genetic model resulted in inflation of narrow-sense heritability but minimally influenced prediction accuracy of genetic values in validation. Predicted genetic rankings of individuals from single-year models were inconsistent across years, likely due to incomplete sampling of the population genetic variance.ConclusionsPredicted breeding values and genetic values revealed many high-performing individuals for use as parents and the most promising selections to advance for cultivar release consideration, respectively. This study highlights the importance of using the appropriate genetic model for calculating breeding values to avoid inflation of expected parental contribution to genetic gain. The genomic predictions obtained will enable breeders to efficiently leverage the genetic potential of North American sweet cherry germplasm by identifying high quality individuals more rapidly than with phenotypic data alone.
Highlights
Sweet cherry is consumed widely across the world and provides substantial economic benefits in regions where it is grown
This study addresses a deficiency of published information on genetic parameters for sweet cherry breedingrelevant traits beyond those influenced primarily by large-effect QTLs by obtaining robust estimates of genetic variance components
The genetic values and the improved understanding of the genetic architecture of important traits in sweet cherry obtained from this multi-year data set of a large pedigree-connected population represent a clear opportunity for genetic improvement
Summary
Sweet cherry is consumed widely across the world and provides substantial economic benefits in regions where it is grown. We used a genome-enabled mixed model to predict the genetic performance of 505 individuals for 32 phenological, disease response and fruit quality traits evaluated in the RosBREED sweet cherry crop data set. Genome-wide predictions were estimated using a repeated measures model for phenotypic data across 3 years, incorporating additive, dominance and epistatic variance components. More than 2.8 million tons of sweet cherry fruit were produced in 2014 [1]. The U.S sweet cherry industry and consumers have previously prioritized which fruit trait thresholds are essential for a successful cultivar. Sweet cherry producers identified fruit size, flavor, firmness, and powdery mildew resistance as trait priorities in a survey conducted in 2011 [4].
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have