OBJECTIVETo evaluate the ability of several hotspot candidate sequence of DNA barcodes for identifying medicinal plant species in Verbenaceae.

METHODUsing universal primers, three chloroplast sequences, psbA-trnH, rbcL, matK, two nuclear ribosomal DNA ITS2 and ITS were amplified and sequenced. PCR amplification and sequencing efficiency, intra- and inter-specific variation, barcoding gap and identification efficiency (with BLAST 1 and Nearest Distance methods) were used to evaluate these loci.

RESULTThe rate of successful amplification using matK was too low to further analyze, and the rate of successful amplification and sequencing using psbA-trnH, ITS, and ITS2 and rbcL sequence was 83.6%, 83.6%, 96.4%, 98.2%, respectively. The rate of successful identification using psbA-trnH, ITS, and ITS2 was 100% at the species level except that rbcL was 77.8%, 75.9% for 55 samples belonging to 32 species, but ITS2 did better in intra- and inter-specific variation, barcoding gap than the other loci. The rate of successful identification of ITS2 was 89.5%, 87.6% even when joining the date of 165 samples from GenBank.

CONCLUSIONIt proposes that the combination of ITS2 and psbA-trnH senquence is promising for the identification of the species in Verbenaceae.