Kyu Tae Park, Young-Ho Kim, Sang-Jun Kim, KyoungSu Choi
The wild population (GJ) showed moderate genetic diversity compared with other Korean endemic species, and positive inbreeding coefficients were observed in all populations. Population structure analyses based on the clone-corrected and LD-pruned 5,539-SNP dataset revealed genetic differentiation among populations. Principal component analysis showed that GJ and HM clustered closely, whereas GC, HT, and YM showed varying degrees of differentiation. ADMIXTURE analysis identified K = 3 as the best-supported value based on the lowest cross-validation error, with GJ and HM sharing a common ancestry component, YM showing a distinct ancestry pattern, and GC and HT exhibiting mixed ancestry. Pairwise fixation index (FST ) values ranged from -0.069 to 0.126. Finite effective population size estimates were obtained only for GJ and HM, whereas finite estimates could not be obtained for GC, HT, and YM, likely reflecting the small number of independent individuals remaining after clone correction.
INTRODUCTION: Veronica pusanensis is a critically endangered plant endemic to Korea and is restricted to a single wild population in Busan. Understanding its genetic diversity, population structure, and effective population size is essential for conservation genomics and for evaluating the representativeness of ex situ collections.
METHODS: We applied triple-digest restriction site-associated DNA sequencing (3RAD-seq) to compare genetic diversity and population structure between the wild population and four ex situ populations maintained in botanical gardens, identifying 43,885 single-nucleotide polymorphisms (SNPs) across 95 individuals. Clone correction based on the full SNP dataset reduced the dataset to 59 representative multilocus genotypes.
RESULTS: The wild population (GJ) showed moderate genetic diversity compared with other Korean endemic species, and positive inbreeding coefficients were observed in all populations. Population structure analyses based on the clone-corrected and LD-pruned 5,539-SNP dataset revealed genetic differentiation among populations. Principal component analysis showed that GJ and HM clustered closely, whereas GC, HT, and YM showed varying degrees of differentiation. ADMIXTURE analysis identified K = 3 as the best-supported value based on the lowest cross-validation error, with GJ and HM sharing a common ancestry component, YM showing a distinct ancestry pattern, and GC and HT exhibiting mixed ancestry. Pairwise fixation index (FST ) values ranged from -0.069 to 0.126. Finite effective population size estimates were obtained only for GJ and HM, whereas finite estimates could not be obtained for GC, HT, and YM, likely reflecting the small number of independent individuals remaining after clone correction.
DISCUSSION: Genetic differences among ex situ populations may reflect differences in source material, propagation history, and clonal redundancy. These findings suggest that ex situ groups differ in their ability to represent the genetic composition of the wild population. Conservation of V. pusanensis should prioritize in situ habitat protection, the maintenance of genetically distinct lineages in ex situ collections, and improved documentation of source materials and propagation history.