Accuracy of Genomic Prediction Based on the Plus-Tree Populations in Cryptomeria japonica

Cryptomeria japonica is one of the most important forestry species in Japan. In order to shorten improvement period and to enhance prediction accuracy in this species, we performed genomic prediction for various traits, such as growth, wood properties, and male strobilus production based on three different populations of the first-generation plus tree clones. For genotype data, we carried out the genotyping of sample clones for over 70,000 single nucleotide polymorphisms (SNPs) using a 70K SNP array of Axiom genotyping system designed for C. japonica. Phenotype data of the sample clones were obtained from clonal test sites with ramet replications. More than 31,000 SNPs were informative for the following analyses. The accuracies of genomic prediction were different among traits and/or prediction models. Predictions for male strobilus production showed higher accuracy than ones of the other traits. The accuracies also differed among populations. We discussed the availability of genomic selection for C. japonica improvement in the near future.