Apple (Malus spp.) is a widely cultivated economic crop. Leaf-related traits, such as leaf size, pigment content, and photosynthetic rate, serve as important indicators of light use efficiency and are associated with apple fruit quality and potential yields. Identifying quantitative trait loci (QTLs) and genes related to these traits is essential for plant breeding. Genomic selection (GS) presents a promising approach, particularly for improving leaf-related traits in apples. This study focused on identifying QTLs and associated genes, and evaluating the use of GS for leaf-related traits in apples. By genotyping an F1 population (‘Luli’ × ‘Red No. 1’) and constructing a high-density linkage map with 3,759 bin markers, 69 QTLs related to leaf traits were identified, explaining 10.4∼17.8% of the phenotypic variance. Forty-six candidate genes were predicted from these QTLs, resulting in 14, 14, 13, and 5 genes for leaf size, pigment content, photosynthetic rate, and fast chlorophyll fluorescence parameters, respectively. The study also assessed the accuracy of GS prediction for leaf traits using genome-wide markers and QTL interval markers through a 10-fold cross-validation. Results indicated that ridge regression BLUP (RR-BLUP) and gradient boosting decision tree (GBDT) methods using QTL interval SNP markers demonstrated higher prediction accuracies compared to whole genome markers for most traits, suggesting the feasibility of employing QTL interval markers for GS prediction. Notably, all models provided valuable genome prediction results with 4K markers or when the training population represented 80% of the total population. These findings significantly contribute to gene discovery and genetic improvement efforts related to leaf traits while highlighting the potential utility of GS in accelerating apple breeding programs.
Read full abstract