Abstract: Subtropical evergreen broad-leaved trees are usually vulnerable to freezing stress, while hexaploid wild Camellia oleifera shows strong freezing tolerance. As a valuable genetic resource of woody oil crop C. oleifera, wild C. oleifera can serve as a case for studying the molecular bases of adaptive evolution to freezing stress. Here, 47 wild C. oleifera from 11 natural distribution sites in China and 4 relative species of C. oleifera were selected for genome sequencing. “Min Temperature of Coldest Month” (BIO6) had the highest comprehensive contribution to wild C. oleifera distribution. The population genetic structure of wild C. oleifera could be divided into two groups: in cold winter (BIO6 ≤ 0 °C) and warm winter (BIO6 > 0 °C) areas. Wild C. oleifera in cold winter areas might have experienced stronger selection pressures and population bottlenecks with lower N_e than those in warm winter areas. 155 single-nucleotide polymorphisms (SNPs) were significantly correlated with the key bioclimatic variables (106 SNPs significantly correlated with BIO6). Twenty key SNPs and 15 key copy number variation regions (CNVRs) were found with genotype differentiation > 50% between the two groups of wild C. oleifera. Key SNPs in cis-regulatory elements might affect the expression of key genes associated with freezing tolerance, and they were also found within a CNVR suggesting interactions between them. Some key CNVRs in the exon regions were closely related to the differentially expressed genes under freezing stress. The findings suggest that rich SNPs and CNVRs in polyploid trees may contribute to the adaptive evolution to freezing stress.

Key words: Adaptive evolution, Camellia oleifera, Copy number variations, Freezing stress, Polyploid, Single-nucleotide polymorphisms