The first application of genomic selection for tree morphological traits in <i>Pinus koraiensis</i> using markers identified through a genome-wide association study

Abstract We present an inaugural comprehensive genomic selection study on Pinus koraiensis, a tree species of significant economic and ecological value in Korea and China. Utilizing genotyping-by-sequencing technology, we analyzed a diverse panel of 191 individuals representing 23 half-sib families. Research on P. koraiensis has so far been limited due to the lack of a reference genome. In this study, we performed a genome-wide association study on five traits, including diameter at breast height, tree height, crown area, crown diameter, and crown volume. Subsequently, we conducted genomic prediction to select the optimal prediction model by comparing 12 genomic statistical models across five genetic populations. We then evaluated the accuracy of the genomic prediction by combining markers selected from the genome-wide association study with the prediction model and assessed the feasibility of genomic selection in genetically distinct populations. We found that conducting separate genomic prediction based on genetic subpopulations did not significantly influence the predictive accuracy for P. koraiensis. However, restricting the number of markers to a −log10 (P-value) threshold of 1.0 (indicating statistical significance of marker–trait associations) improved the accuracy of all eight examined statistical models for all five traits, with the highest accuracy reaching 0.962 for crown area using genomic best linear unbiased prediction. Moreover, our genomically estimated breeding values showed a high correlation with the actual measured phenotypic values, indicating the practical applicability of our marker filtering, and the optimized statistical models in breeding selection. Remarkably, this approach was even applicable to genetically distinct subpopulations. In conclusion, this study demonstrates that marker filtering can lead to an improvement in genomic selection accuracy for Korean pine, ultimately allowing for the shortening of breeding cycles, stabilizing forests, and increasing productivity.