Prediction of Genetic Potential of Candidate Apple Cultivars for Fruit Quality from Unreplicated Multi-location Field Trials

Evaluation of the phenotypic expression of traits under field conditions is a fundamental process for predicting the genetic potential of new candidate cultivars. However, the accuracy of genetic potential predicted from phenotypic observations is compromised by non-genetic effects. Replication of individuals through grafting and randomization may be employed to estimate the average effect of a candidate, from which genetic potential may be inferred. In this study, we describe the use of pedigree information in the analysis of unreplicated genetic trials as another method to improve the accuracy of predicting genetic potential from phenotypic records. The data in this analysis is from the pedigree-linked RosBREED apple reference germplasm set including trees planted at three locations: Wenatchee, WA (n = 284), Victoria, MN (n = 300), and Geneva, NY (n = 170). There were less than five individuals that were in common among the trials. However, the deep pedigree relationships among parents and ancestors that spanned seven generations were used to establish genetic connections among trials and hence enabled genetic effects to be separated from non-genetic effects. This approach was also used to quantify the magnitude of genotype-by-environment interactions for several traits.