Abstract:

Molecular breeding refers to selection practices based on DNA markers associated with phenotypic traits, called also as marker assisted selection or marker assisted breeding, in which genetic engineering breeding and geno?mics aided breeding were included in a broad term. It provided a potentially efficient option for early selection and accelerated breeding in forest trees. This review presents the advancement and prospects of genomics information resources for tree molecular breeding studies. In recent three decades, molecular marker techniques have been developed from earlier low-throughput assay to currently high-throughput microarray- and next-generation-sequencing-based platforms, such as genotyping by sequencing, transcriptome sequencing, genome re-sequencing, target amplicon sequencing and exome sequencing, and these high-throughput techniques have been widely used in the three main approaches for identifying trait-related DNA variations in forest trees, including linkage mapping, association genetics and genomic selection studies. More than 50 tree species have been genome sequenced since the first release of Populustrichocarpa whole genome sequence. Linkage mapping and association genetics studies have resulted in many genomic loci related with growth, wood properties and stress responses as well as non-wood product quality traits in more than 10 tree genera, highlighting the trends below: ① a multitude of phenotypic traits investigated, covering economic characteristics, physiological indices and metabolic composition; ② hundreds of thousands of markers across the whole genome scanned; ③ integration of multi-omics data; ④ large population size for high-resolution fine mapping; ⑤ multiple site trials for dissecting the interactions between genotype and environment and between genotype and age; ⑥ mining candidate genes from reference genome sequence and/or differentially expressed transcriptomic genes. Genomic selection models have been formulated for a number of major cultivation tree species in genera Eucalyptus, Pinus and Picea. Other genomic information resour ces such as pan genomes, computational tools and software packages, functional genes, genome editing techniques and online databases have also been available. Challenges that forest tree molecular breeding is facing include:① how to obtain environmentally stable trait-related genomic loci and genomic selection models; ②lacking of autonomous, non-destructive and high-throughput phenotyping methods; ③ difficulty in high-quality genome assembly for genome-large conifers and polyploidy tree species; ④ application uncertainty and balance of additional investment for marker-assisted selection in a breeding program; ⑤ technical barrier in accelerated breeding for most tree species. In the post-genome area, molecular breeding will be integrated effectively into tree breeding programs and can be expected to contribute largely to capturing higher genetic gains as compared to the traditional breeding.

Key words: forest tree molecular breeding, genomics, linkage mapping, association studies, genomic selection（GS), accelerated breeding