SPATIAL STRUCTURE OF GENETIC VARIATION OF SSR MARKERS FOR WILD POPULATIONS OF ACTINIDIA CHINENSIS AND A. DELICIOSA

Genetic variations in two wild populations of A. chinensis and A. deliciosa were studied using spatial autocorrelation analysis. This paper considers the distribution characteristics of genetic variation within natural populations of the genus Actinidia, and the mechanism of speciation in evolution. A total of 104 alleles was screened from 9 pairs of SSR primers for two populations, and 63 SSR alleles were selected, according to the phenotype frequency between 20% and 80%, to calculate the Moran’s I spatial autocorrelation coefficient for the two populations of A. chinensis and A. deliciosa by using equal gene frequency correlograms with ten distance classes. There was a weak spatial structure for A. chinensis and A. deliciosa in the two populations, in which some alleles showed correlation of significance although most alleles were randomly distributed in each population. Individuals within 100 m distance had positive and significant correlation of alleles, and showed negative correlation with increasing geographical distance. The preliminary results showed that the genetic variation within populations of Actinidia were distributed in a lump-like structure at shorter distances, and cline structures were found with increasing separation. The spatial structure of genetic variation for Actinidia populations was the result of the combination of the biological characteristics for pollination, seed dispersal and natural habitat, and the adjacency distribution of seed and the effective distance of pollen spread were concluded to be the main factors which caused the patch pattern. A conservation strategy for the genus Actinidia was also put forward based on this study.