Abstract: In order to rapid and precisely identify the excellent gene resource for the breeding of tea plant, the construction on core collection of tea plant is necessary. Based on the passport and evaluated datum of 1048 tea accessions in “Evaluated database of tea germplasms”, the optimum sampling proportion, grouping methods, intra-group sampling proportion and methods, in order to obtain the optimum sampling strategy to establish the core collection of China tea germplasm. Results showed that, (1) 20% was the optimum sampling proportion, at this proportion, the genetic diversity index of core collection was higher than other proportions, and the reserved ratio of phenotypic charater of core collection was higher than 95%. (2) At 20% sampling proportion, 25 primary core collections using 3 grouping methods, 4 intra-group sampling proportions, 2 intra-group individual selecting methods were evaluated using 5 parameters to determine the optimum sampling methods. The five parameters were a) ratio of phenotype reserved, b) phenotypic of variance value, c) variance of phenotypic frequency, d) Shannon-weave genetic diversity index, e) coefficient of variation quantitative characters. Results showed that there were different ranks when different evaluated parameter was used. It implied that there was significant interaction between grouping methods, sampling proportion and sampling methods within group. Combined the ranking results of 5 parameters, the optimum sampling strategy to establish primary core collection of tea plant is cultivated region grouping-logarithm sampling proportion-cluster individual selecting.

Key words: tea germplasm (Camellia spp.), core collection, sampling strategy