The samples for this study were collected from fifty-one cultivars of C. miniata, eight wild individuals of C. nobilis and two hybrids between the two species. The microsatellite-enriched libraries were constructed following the procedure of Fast Isolation by AFLP of Sequences Containing repeats (FIASCO) [13]. In brief, total genomic DNA was extracted from a single individual each of C. nobilis and C. miniata. About 300 ng genomic DNA for each species was completely digested with MseI separately (New England BioLabs, Beverly, MA, USA). Then, these digested products were ligated to the MseI adaptor pair (5′-TACTCAGGACTCAT-3′/5′-GACGATGAGTCCTGAG-3′) [14]. The diluted digestion-ligation mixture (1:10) was then amplified with MseI-N primer (5′-GATGAGTCCTGAGTAAN-3′) using the following cycle conditions: 95 °C for 5 min, 20 cycles of 94 °C for 30 s, 53 °C for 1 min, 72 °C for 1 min, and a finally extension at 72 °C for 5 min.

For enrichment, these PCR products were denatured and hybridized with biotin-labeled (AC)₁₅ probes, and then these DNA fragments were captured by streptavidin-coated magnetic beads (Promega, Madison, WI, USA). These enrichment products were purified with a Gel Extraction Kit (Takara, Liaoning, China) and ligated into the pMD18 vector (Takara) and transformed into E. coli strain JM109 (Takara). The positive clones were identified by PCR amplification using (AC)₁₀ and M₁₃₊/M₁₃₋ primers, respectively. If one of the primer combinations was successfully amplified, it suggested that the amplicons contained microsatellite motifs. The positive clones for C. miniata and C. nobilis were sequenced on an ABI 3730 DNA analyzer (Applied Biosystems, Foster City, CA, USA) using M13F as sequencing primer. Primers were designed according to Li et al. [15]: 1. the length of primer was 18–25 bp; 2. annealing temperature was 55–65 °C; 3. GC content of primer was 40–60%.

Primer pairs were used to amplify DNA of 61 individuals. PCR amplifications were performed in a total volume of 20 μL containing the following components: 50 ng of DNA template, 0.2 μM of each dNTP (10 mM), 0.5 pmol of each primer, 1× PCR buffer (containing 2.5 mM Mg²⁺) and 1 U of high-fidelity Taq polymerase (Takara). The PCR reactions were performed on an ABI 2720 thermocycler: an initial denaturation of 5 min at 95 °C, 35 cycles of 30 s at 94 °C, annealing temperature (Ta) for 30 s (Ta for each primer pairs were listed in Table 2), 30 s at 72 °C, and a final extension of 72 °C for 8 min. These amplification products were resolved on 6% polyacrylamide denaturing gel electrophoresis and visualized using silver staining. To discriminate the real alleles from Taq polymerase errors and nonspecific amplification products, only clear bands with expected size were considered and the rare alleles (only occured once) were removed from the dataset.

The preliminary population genetic analyses for these microsatellite loci were performed using the software GENEPOP with the default settings and assumptions [16], including the number of alleles per locus (N_A), observed heterozygosity (H_O), expected heterozygosity (H_E) and linkage disequilibrium (LD).