Abstract

Abstract Genetic diversity of 5029 accessions in Chinese candidate core collections was analyzed at 78 SSR loci by fluorescence microsatellite markers. Based on the four hundred thousands of SSR datapoints, core collection was established for Chinese 23090 wheat accessions. Genetic representation and diversity of the core were calculated in this paper. Through selective sampling (adjust sampling number and entries), we only use 23.1% (i.e. 1160 accessions) of candidate core collections including 762 landraces, 348 modern varieties and 50 introduced varieties, to keep 94.9% alleles in the candidate collections. The core only took 5% of the base collections, but could capture 91.5% of genetic variation in the base collections. This is much higher than the accepted standard i.e.10% of the base accessions capture more than 70% of the genetic diversity. However, if core entries were chosen completely randomly, the entries took more than 10.4% (~350 accessions) and 18.9% (~300 accessions) of candidates in the landrace and modern variety respectively to reach more than 70% representation. Genetic representation of selective sampling (91.5%) is much higher than that by random sampling (79.8%). Total 1557 alleles were amplified at the 78 loci in the 1160 core collections. The landraces have more allelic variation than the modern varieties (1408 vs 1158). Average genetic richness and genetic diversity indexes of the landraces were also higher than that of the latter. Structure and principal coordinate analysis showed that the landraces and the modern varieties were two relatively independent populations.