Comparative analysis of methods for estimating genetic parameters of fruit-quality traits in apple breeding program

Apples, originating from Central Asia, are widely cultivated worldwide, with China producing 47.6 million tonnes annually. Despite their global importance, breeding programs often focus on a limited number of high-quality cultivars, potentially threatening genetic diversity. To mitigate this risk, advanced statistical methods could be employed to improve selection strategies. For example, the multi-trait BLUP method, which accounts for genetic correlations among traits, can reduce selection bias but increases computational complexity due to the model’s intricacy and slow convergence of the REML process. Bayesian methods, like Markov Chain Monte Carlo (MCMC), offer a solution to these challenges. This study compares single-trait and multi-trait approaches using REML/BLUP and MCMC/BLUP to estimate variance components and predict genetic values. Over two seasons, phenotypic data from 304 seedlings and 16 parents were analyzed using an animal model to predict their genetic values related to fruit quality traits. Significant progress was made in estimating genetic parameters and selecting parents for traits such as fruit weight, flesh firmness, and soluble solids content in the University of Udine’s apple breeding program. The multi-trait BLUP method improved the accuracy of predicted breeding values, particularly for fruit weight, polar diameter, flesh firmness, and soluble solids content. The multivariate repeated measures model, despite low correlations between some trait pairs, was recommended for integrating multi-season results and accounting for trait correlations. The Bayesian MCMC approach proved superior in genetic evaluations, offering higher heritabilities and genetic gains compared to Fisherian methods (REML). It effectively handles small sample sizes, captures more genetic variance, and enhances breeding value predictions, recommending traits like fruit size, firmness, and sweetness for selecting superior parents.