Construction of Core Collection of Schima superba Based on SSR Molecular Markers

Objective: The most suitable sampling strategy and proportion for core collection of Schima superba were defined through a comparative analysis and evaluation, and a core collection of S. superba was constructed. Molecular identity of the core collection was further studied, providing theoretical basis and core materials for further research and utilization of S. superba germplasms. It also provides a basis for the core collection construction of other woody plants. Method:Taking 754 accessions of S. superba as materials that were come from seven provinces of China. The methods of M strategy (maximization strategy, M), random sampling (R), simulated annealing algorithm maximizing the genetic diversity (SAGD) and simulated annealing algorithm maximizing the number of alleles (SANA) were used to construct the core collections using 13 SSR primer pairs. The parameters of genetic diversity, such as number of alleles (Na), effective number of alleles (Ne) and Shannon's information index (I), were used to determine the optimal method. Result: 128 alleles (Na) were detected in 13 SSR primer pairs, the average of Ne was 3.47. High genetic diversity was revealed in the germplasm resources of S. superba (I=1.39). Comparative analysis showed that the core collection constructed by SANA, SAGD and M strategies were all better than R strategy. The retained of Na in the core collection was low, although the core collection with SANA and SAGD are better representation of the original collection. The retained of Na in the core collection that was constructed by M strategy was apparently higher than other core collections that were constructed by other three strategies. According to the parameters of genetic diversity, and the number of sampling, the core collection constructed by M strategy could preserve the greatest level of genetic diversity of germplasm resources with the minimum sample size, indicating that M strategy was optimal in terms of establishing the most representative core collection, although SANA and SAGD methods all have good representation. The principal coordinate analysis (PCoA) showed that the core collection constructed by M strategy could more comprehensively represent at the genetic diversity level of S. superba. The 115 accessions of S. superba includes 15.3% samples of the germplasms, the retention ratio of Na, Ne and I were 93.8%, 115.6% and 109.9%, and the specific molecular identity for 115 core collections were established with 99.99% probability of confidence. These results demonstrated that the 115 accessions could stand for the initial collection, at the same time a set of unique molecular identity was established for 115 accessions based on 13 SSR loci. Conclusion: M strategy is a suitable method for constructing S. superba core collection. These results demonstrated that the 115 accessions as core collection could represent the original germplasms, at the same time this research method of the construction of core collection would provide an example for other tree species. © 2017, Editorial Department of Scientia Silvae Sinicae. All right reserved.