Establishment of a core collection of cultivated apricot based on phenotypic characters in Xinjiang
The genetic diversity in 35 phenotypic characters based on plant organs, such as leaf, fruit, nutlet, and kernel, of 310 cultivated apricot plants was used for the establishment of a core collection. A total of 24 core collections were constructed at 25% sampling proportion by using three sampling strategies, including preferred sampling, deviation sampling, and random sampling, in combination with two genetic distances, Euclidean and Mahalanobis, and four clustering methods including the UPGMA, Ward's method, single linkage, and complete linkage. Using the best construction method, the optimum sampling proportion was determined under eight different sampling proportions. We evaluated the representation of the core collection based on the genetic diversity, variation, and t-test of different quantitative characters of the initial, core, and reserve collections. The representation of the core collection was confirmed by using principal component analysis. The Euclidean distance was much better than the Mahalanobis distance. Among the four clustering methods, we found that the single-linkage method was better than the UPGMA, Ward's method, and complete linkage for constructing the core collection. The preferred sampling strategy was suitable for establishing the core collection of cultivated apricot, but the deviation and random sampling methods were not. The optimum sampling proportion was 25%. The 75 core collection resources for cultivated apricot were considered good representatives of the initial collection and retained the genetic diversity and variation well. At 25% sampling proportion, use of the stepwise clustering method with Euclidean distance, and single linkage combined with preferred sampling is the most suitable method for establishing the most representative core collection of cultivated apricot.
Liu, J., Liao, K., Cao, Q., Sun, Q., Liu, H. and Jia, Y. (2018). Establishment of a core collection of cultivated apricot based on phenotypic characters in Xinjiang. Acta Hortic. 1214, 257-268
genetic diversities, genetic distance, clustering method, sampling strategy, phenotypic character