Genetic variation of coreoleuciscus splendidus populations (Teleostei; Cypriniformes) from four major river drainage systems in South Korea as assessed by AFLP markers

Abstract

Amplified fragment length polymorphism (AFLP) technique was used to assess genetic variation and population structuring of Coreoleuciscus splendidus (Teleostei; Cypriniformes) from four major river drainage systems in South Korea (i.e. the Han and Geum Rivers belonging to the West Korea Subdistrict and the Seomjin and Nakdong Rivers belonging to the South Korea Subdistrict). Four combinations of selective primers generated 152–186 DNA fragments, of which 45–59 were polymorphic. The Geum River population had the lowest genetic variation in terms of average heterozygosity. The pairwise genetic differentiation (F _ST) estimates showed that the four populations of C. splendidus were genetically differentiated (P < 0.05). The F _ST values were higher between population groups of the West and South Subdistricts than between populations from the two river systems within each subdistrict. A dendrogram based on the distance matrix revealed that all C. splendidus individuals bifurcated into two subdistrict population groups and were further divided into two river populations within each subdistrict. Individuals from each population consistently clustered together with high statistical support. The analysis of molecular variance after partitioning of C. splendidus populations into two groups (i.e. the West and South Korea Subdistricts) revealed very strong geographic structuring between them with 74.01% of total variance (P < 0.001). Taken together, the AFLP markers clearly distinguished not only between two population groups of the West and South Subdistricts but also between populations from two river systems within each subdistrict. Morphological differences between the two subdistrict population groups in body color and the presence of black array(s) on dorsal, anal, and caudal fin rays agreed with their high degree of genetic variation. Our findings strongly indicate that restricted dispersal and gene flow are apparent among four geographically separate populations of C. splendidus, and each population should be treated as an evolutionarily significant unit; this should be an a priori consideration for developing their conservation strategies.