Wheat genetic resources: A total of 40 accessions, which include 9 Tibetan weedrace (TW), 9 Xingjiang rice wheat (XR), 14 Yunnan hulled wheat (YH) and 8 Sichuan White Wheat (SWW) were employed in this study (Table 1).
DNA extraction: Genomic DNA was extracted from a bulk sampling of a minimum of ten individual plants for each accession following the procedure described by Sharp et al. (1988).
STS analysis: Fourteen STS markers, located on all 7 homoeologous chromosome groups of wheat, were used in this study (Table 2). All these STS-PCR markers were, from RFLP clones of the wheat or barely mapping projects (Chen et al. 1994; Talbert et al. 1994). STS-PCR was conducted following the procedure described by Talbert et al. (1994). Amplified products were subsequently digested with approximately 1 U of HinfI and HhaI (MBI) per reaction mixture for 1 h at 37^oC, and then analyzed by electrophoresis in 2.0% agarose gels and stained with ethidium bromide.
SSR analysis: Twenty-four SSR markers, which located on 21 chromosomes (Devos et al. 1995; Roder et al. 1998), were used in this study (Table 3). PCR amplifications were carried out under the conditions described by Roder et al. (1998). The products were separated on 3.0% agarose gel and visualized by ethidium bromide staining.
Data scoring and analysis: Images for STS and SSR analysis were photographed, captured using ImageMaster VDS (Amersham Pharmacia Biotech). The size of each DNA band was automatically estimated using ImageMaster 1D Elite Software (Amersham Pharmacia Biotech). For each accession x marker combination, the presence (1) or absence (0) of an amplified band or allele was treated as an independent character without consideration of the quantitative aspects of the results, i.e. band intensity. The data matrix was then used to calculate genetic similarity index (GS) (Nei and Li 1979):
GS = 2Nij/(Ni + Nj)
where Nij if; the number of bands (alleles) in common between genotypes i and j, and Ni and Nj are the total number of bands (alleles) observed for genotypes i and j, respectively. Based on the (1-GS) matrix, a dendrogram showing the genetic relationships between genotypes was constructed using the unweighted pair-group method with arithmetic average (UPGMA) (Sneath and Sokal 1973) through the computer software NTSYS-pc Version 1.80 (RohIf 1993).

STS variations: The 14 STS markers (Table 2) were used to reveal genetic difference among 40 Chinese endemic wheat accessions. Upon digestion with Hinfl and HhaI, 11 out of 14 markers (78.6%) and 16 out of 28 marker/enzyme combinations (57.1%) revealed polymorphisms. The G36/HhaI combination generated more bands than any other marker/enzyme combination, while the B5/HinfI, D18/HhaI and E16/ HinfI combinations could only generate one band. Among 40 Chinese endemic wheats, a total of 121 bands were observed in 28 marker/enzyme combinations, with 4.3 bands per marker/enzyme combination. Thirty-nine out of 121 bands (32.2%) were polymorphic, among which 1 to 8 polymorphic bands were generated by each informative marker/enzyme combination.

All the 121 bands, generated from 28 STS marker/enzyme combinations, were used to calculated genetic similarity index (GS) among 40 landraces (Table 4). Among 40 Chinese endemic wheat accessions, the GS value ranged from 0.645 to 0.989, with the mean of 0.822. The mean GS values within SWW, YH, TW and XR groups were 0.921, 0.907, 0.849 and 0.842, respectively. It indicated that the mean GS values of SWW and YH wheat groups were relatively higher than those of the TW and XR groups.

The genetic relationships within and between groups were estimated by a UPGMA cluster analysis of (1-GS) matrix (Fig. 1). The results showed that all 40 landraces could be distinguished by STS-PCR markers. Two distinct groups were evident, with the second group (II) including all 9 accessions from the XR wheat group. Three subgroups (Ia, Ib and Ic) were evident for the first group (I), with the subgroup Ib and Ic more closely related. The subgroup lc was a cluster for the accessions from the SWW group. The subgroup Ib included all 14 YH wheat accessions and 4 TW wheat accessions. It indicated that the XR wheat group was genetically distinct from other three Chinese endemic wheat groups, while the SWW group was genetically related to the YH group. The TW wheat group is more diverse than the SWW and YH groups, with some accessions more related to the YH group.
SSR polymorphisms: Twenty-four SSR markers, located on 21 wheat chromosomes, were used for PCR amplification of the genomic DNA of 40 Chinese endemic wheat accessions, among which PCR products of 21 SSR makers (87.5%) showed polymorphism (Fig.2). A total of 83 alleles were detected among the 40 accessions. The number of alleles ranged from 1 to 7, with an average of 3.5 alleles per SSR locus. On 3 SSR loci (i.e. gamma-gliadin, LMW-Glu and GWM111) more than 6 alleles could be detected.