Abstract: 【Objective】 By comparison analysis of the results of different sampling strategies and genetic distances, the method for constructing core collection of Xinjiang wild apricot based on molecular markers data was studied in order to establish the optimum core collection resources which will be beneficial to the protection and use of resources. 【Method】 Taking 135 wild apricots as materials that were come from Huocheng Daxigou, Xinyuan Boersai and Gongliu Yiligedai in Xinjiang, using UPGMA stepwise clustering method according to SM, Jaccard and Nei&Li genetic distances for initial collection, until any one sampling location was no collection into the core collection, then an allele preferred sampling strategy was used to construct wild apricot core collection and compared with the random sampling strategy. The number of lost allele and t-test of number of polymorphic loci, percentage of polymorphic loci, observed number of alleles, effective number of alleles, Nei’s genetic diversity and Shannon information index for gene diversity were used to determine the optimal building methods. T-tests of core collection, initial collection and reserve collection were conducted, and the genetic diversity of them was compared to evaluate the representative of core collections. The principal coordinate analysis method and the phenotypic traits of initial and core collections were used to confirm the core germplasm.【Result】Compared with the random sampling strategy, allele preferred sampling strategy could construct more representative core collections that with higher values of genetic diversity indexes and little polymorphic loci. According to Nei&Li genetic distance, the constructed core collection of Xinjiang wild apricot was better than by SM and Jaccard genetic distance for these core collections had high values of genetic diversity indexes. The analysis of principal coordinate analysis and phenotypic traits showed that the core collection of Xinjiang wild apricot constructed by allele preferred sampling strategy and Nei&Li genetic distance could more comprehensively represent at the genetic diversity of wild apricot initial collection. The 31 core collection resources of wild apricot includes 22.96% germplasm samples of the initial collection, the retention ratio of number of polymorphic loci, percentage of polymorphic loci, observed number of alleles, effective number of alleles, Nei's genetic diversity and Shannon information index were 92.69%, 98.83%, 99.42%, 103.26%, 109.24% and 108.31%, respectively. 【Conclusion】 The method of allele preferred sampling strategy and Nei&Li genetic distance by stepwise clustering is a suitable method for constructing Xinjiang wild apricot core collection. These results demonstrated that the 31 core collection could stand for original collection excellently, at the same time this research method of the construction of core collection has important reference values for other crops.

Key words: Xinjiang wild apricot, core collection, ISSR markers, allele preferred sampling strategy, genetic diversity