Barthea Hook. f. (Melastomataceae) is a monotypic genus endemic to southern China. The only species, B. barthei (Hance ex Benth.) Krasser, is an evergreen shrub and has a disjunct distribution in southern mainland China (Guangdong Province, Guangxi Province, Fujian Province, Hunan Province, and Hong Kong) and Taiwan (Chen, 1984; Chen and Renner, 2007). There are two varieties for this species, B. barthei var. barthei and B. barthei var. valdealata C. Hansen. While B. barthei var. barthei is found throughout the range of the species, the variety B. barthei var. valdealata is confined to Shangsi County, Guangxi Province. The two varieties differ mainly in the width of the capsule wings (1 mm wide for B. barthei var. barthei vs. 2 mm for B. barthei var. valdealata [Chen, 1984; Chen and Renner, 2007]). Both varieties occur on forested mountain slopes at 500–1500 m elevation.

Patterns of disjunct distributions and their formative mechanisms have long been an important topic in the field of phytogeography. Disjunct distributions of plants can be used to reveal the relationships between floras of two or more regions. There are 35 genera of seed plants that have a southern mainland China–Taiwan disjunct distribution (Chen et al., 2012); however, when and how the disjunct distribution of these plants formed remains elusive (Chen et al., 2012; Ye et al., 2012). As a typical species with a disjunct distribution in southern mainland China and Taiwan, B. barthei can be used to address this phytogeographic question. Moreover, because the trait used to distinguish the two varieties of B. barthei shows substantial variation, the taxonomic treatment of them as two varieties is doubtful. Molecular data may help resolve these evolutionary or taxonomic questions. However, to our knowledge, there have been no molecular markers developed for B. barthei so far. In this study, we developed and characterized 27 nuclear simple sequence repeat (SSR) markers for B. barthei using paired-end reads (250 bp) generated using an Illumina HiSeq 2500 system (Illumina, San Diego, California, USA).

METHODS AND RESULTS

We sampled 20 individuals from each of three natural populations, namely, Nanling in Ruyuan, Guangdong Province (NA); Sanzhoutian in Shenzhen, Guangdong Province (SA); and Shiwandashan in Shangsi, Guangxi Province (SH) (Appendix 1). The SH population represents B. barthei var. valdealata, while the other two populations represent B. barthei var. barthei. Genomic DNA was isolated from silica-dried leaves using the cetyltrimethylammonium bromide (CTAB) method (Doyle and Doyle, 1987). We first constructed a genomic DNA library with 400-bp inserts from an individual of B. barthei collected in SA. The genomic DNA library was sequenced using an Illumina HiSeq 2500 system (approximately 1/10 Illumina lane) at Berry Genomics (Beijing, China). Initial quality filtering was performed with the sequencing company's in-house script fastqc_adapter_pe. Reads with (1) adapters, (2) >10% ambiguous base calls (Ns), or (3) >50% bases ≤5 in the Phred quality score were removed. A total of 17.58 million 250-bp clean paired-end reads were obtained. The reads were then deposited in the National Center for Biotechnology Information (NCBI) Sequence Read Archive (accession no. SRR5130485). Each pair of reads was assembled into contigs using mothur version 1.37.6 (Schloss et al., 2009) with the parameters minlength = 350, maxlength = 500, minoverlap = 50, and mismatches = 2. We then clustered these contigs using CD-HIT version 4.6 (Li and Godzik, 2006) with the minimum identity of 98%. A total of 4,572,771 clusters were generated. When a cluster included more than three members, the longest one was used to seek SSR motifs. Using MISA software (Thiel et al., 2003), a total of 2535 SSR loci were detected from the 124,768 clusters with more than three members. Primer pairs for the 90 SSR loci with the longest dinucleotide repeats were designed with Primer3 (Rozen and Skaletsky, 1999).

Table 1.

Characteristics of 27 SSR markers developed for Barthea barthei.

To screen the SSR primers, we conducted PCR amplification using one individual each from the SA and SH populations in a final reaction volume of 20 µL containing ∼20 ng DNA, 10 µL 10× PCR buffer with Mg²⁺, 0.25 mM dNTPs, 1 µM each of forward and reverse primer, and 1 unit EasyTaq DNA polymerase (TransGen Biotech, Beijing, China). PCR was conducted for all primers using the following cycling program: 4 min of denaturation at 94°C; followed by 30 cycles of 40 s at 94°C, 40 s at the annealing temperature of each primer pair, and 60 s at 72°C; with a final extension of 8 min at 72°C. The PCR products were run in a 1.2% agarose gel to see if the expected size was obtained for each primer pair. Twenty-seven primer pairs designed for B. barthei produced single-band PCR products with the expected size (Table 1).

Table 2.

Genetic diversity at 18 SSR loci in three natural populations of Barthea barthei.^a

To test the polymorphism level of the 27 primer pairs, PCR was conducted for all of the samples of B. barthei using the conditions mentioned above. We labeled the forward primers of the 27 primer pairs with the fluorescent dye FAM or HEX. Using GeneScan 500 ROX (Applied Biosystems, Waltham, Massachusetts, USA) as an internal size standard, we determined the fragment sizes of the PCR products on an ABI 3100 DNA Sequencer with Genotyper 4.0 (Applied Biosystems). Parameters of genetic diversity in each population and genetic differentiation among populations of B. barthei were calculated using GenAlEx version 6.5 (Peakall and Smouse, 2012).

Our results showed that 18 SSR markers were polymorphic in B. barthei (Table 2), and two to nine alleles were detected for these SSR loci at the species level. The observed heterozygosity of these polymorphic loci varied from 0 to 0.850, and the expected heterozygosity varied from 0 to 0.809. Of the 18 polymorphic SSR loci, three, five, and six in the SH, NA, and SA populations, respectively, exhibited significant deviations from Hardy–Weinberg equilibrium (Table 2). A measure of pairwise genetic differentiation between populations (F_ST) indicated that genetic differentiation between NA and SA populations was the highest (F_ST = 0.474), while lower genetic differentiation was observed between SH and NA populations (F_ST = 0.418) and between SH and SA populations (F_ST = 0.387). Therefore, our SSR data showed that genetic differentiation between the two varieties is lower than that between the two populations of B. barthei var. barthei.

CONCLUSIONS

This is the first set of molecular markers developed for B. barthei, an evergreen shrub with a disjunct distribution in southern mainland China and Taiwan. The 18 polymorphic markers may be useful for phytogeographic studies of B. barthei to reveal the formative mechanisms of the southern mainland China–Taiwan disjunct distribution. Lower differentiation between the two varieties than between allopatric populations of the variety B. barthei var. barthei suggests that the taxonomic division of B. barthei as two varieties may not hold.