Abstract
Dendrobium nobile Lindl. is one of the most important Orchid plants worldwide. The genotype-by-sequencing (GBS) method has now been widely used to access genetic diversity because of its high-throughput and cost-effective in molecular markers. The goal of this study was to employ the GBS technique for diversity evaluation of D. nobile and determine genetic differences between populations. A total of 129 accessions of D. nobile collected originally between 2019 and 2020 from 10 imitation-wild cultivated populations growing in Sichuan, Guizhou and Yunnan of southwestern China were sequenced, a total of 135G clean reads and a total of 836,786 SNPs of high quality data was yielded and were used for final analysis of genetic diversity and population structure. The quality value 20(Q20) ≥ 92.61%, the quality value 30(Q30) ≥ 82.38%. The GC contents distributed between 37.58% and 38.82%. It was also found that more transitions than transversions, and the ratio of transition/transversion varied from 1.804 to 1.911. By the methods of STRUCTURE, the most appropriate number was found to be k=3, all accessions of D. nobile were classified into three groups, excepts for 14 accessions belonging to admixed group. Phylogenetic tree and principal component analysis (PCA) were consistent with the result. The first two principal components explained a total of 23.25% of the variation by PCA. The genetic diversity of ML population showed the lower genetic diversity as indicated by the effective number of alleles (Ne) = 1.287, polymorphism information content (PIC) = 0.141, and Shannon's information index (I) = 0.205, while WT population showed slightly higher genetic diversity by the Ne =1.512, PIC =0.256, and I =0.360. ML population and other nine populations (FB, FM, FX, LJ, SJ, SP, WL, WT and XM) were the most divergent between them respectively owing to all pairwise Fst values above 0.25, while FM population and FX population were considered identical because the pairwise Fst value was 0.0 between the two populations. Correlation analysis showed that highly significant correlation was observed between genetic distance and actual geographical distance (r = 0.854, P < 0.0001), indicating that the genetic differentiation of the 10 D.nobile populations conformed to the geographical isolation model. Analysis of molecular variance (AMOVA) revealed that the genetic variation was greater within populations (87.8%) than among populations (12.2%). This confirmed that intra-population variation was the main source of genetic variation in 10 D. nobile populations. The results also showed that Nm = 1.799 > 1, indicating that there was gene exchange between different populations. Analysis of unweighted pair-group method with arithmetic mean (UPGMA) suggested that the 10 populations were classified into three groups (Group I, Group II and Group III), Group III could be further divided into two subgroups (Group IIIa and Group IIIb). The results will not only provide valuable information for the level of genetic diversity of D.nobile growing in southwestern of China but also help for formulation of strategies for resource protection and utilization. Moreover, GBS appears as an efficient tool to detect intra-population variation.