Objective  The method of constructing microcore germplasm was discussed to provide scientific basis for precise selection of hybrid parents or research materials of P. tomentosa, and to provide reference for other species to construct microcore germplasm.

  Method  In this study, molecular genotypes of 272 P. tomentosa samples were detected by fluorescent SSR molecular markers. The contribution value of each sample to the overall genetic diversity was calculated and sorted from large to small. Then, the representativity of microcore germplasm was analyzed and the appropriate sampling proportion was determined by comparing the average effective number of alleles Ne, the average Shannon’s information index I, the average expected heterogenicity He and the average polymorphic information content PIC values obtained by sampling proportion of top 25%, 20%, 15%, 10%, 5%, 2.5%, 2% and 1.5%. Finally, t test was used for statistical verification. According to the band-shaped results of all SSR primers, fingerprints of each microcore germplasm were formed. According to the difference of fingerprint, the specific alleles were analyzed and excavated.

  Result  When the sampling proportion gradually decreased from 25% to 2%, three important genetic diversity parameters of Ne, He and PIC gradually reached the maximum values, respectively, while I reached the maximum value when the sampling proportion was 10%. When the sampling ratio was 2%, the values of Ne, I, He and PIC were 3.513, 1.254, 0.643 and 0.597, respectively, which were all greater than 2.075, 0.825, 0.432 and 0.364 of the original germplasm, indicating that we had effectively removed the genetic redundancy, and the constructed microcore germplasm had abundant genetic diversity. The t test results showed that there was no significant difference in genetic diversity between the constructed microcore germplasm and the original germplasm, which was reliable and representative and could be used as the microcore germplasm. The proportion of hybrid germplasm was more than 50% regardless of the sampling proportion of 10%, 5% or 2%. Sample 218 had a specific allele at locus 11, sample 261 had a specific allele at locus 16, and sample 263 had a specific allele at locus 7.

  Conclusion  When the sampling ratio of microcore germplasm of poplar is 2%−10%, it is the most scientific, reliable, simple and efficient, among which the most accurate sampling ratio is 2%. If the number of germplasm resources is large, it can be appropriately reduced according to the genetic diversity, and on the contrary, it can be increased appropriately. It is suggested that when constructing core germplasm or microcore germplasm in the future, rather than grouping, each sample is firstly ranked from largest to smallest according to its contribution to overall genetic diversity. It is proved from the molecular level that hybrid germplasm plays an important role in the conservation of germplasm resources and the genetic breeding of trees. At the same time, this study also provides reference for the construction of microcore germplasm of similar species.