Oysters are distributed around the world along the shores of all continents (except Antarctica) and some oceanic islands. They are important species in the shellfish industry; the world production of oysters was 4.4 million tons in 2011, accounting for more than 30% of the total marine molluscan yield (FAO, 2013). Due to their economic value and significant role in nearshore ecosystems (Ruesink, Lenihan & Trimble, 2005), oysters have been extensively studied. All commercially important oysters that are common along the coast of China belong to the genus Crassostrea, including C. gigas (Thunberg, 1793), C. angulata (Lamarck, 1819), C. ariakensis (Fujita, 1913), C. hongkongensis (Lam & Morton, 2003) and C. sikamea (Amemiya, 1928) (Liu et al., 2011). They are commonly found on rocks, concrete and other hard substrates in almost all intertidal zones and harbours. Because of their phenotypically plastic shell morphology, oysters cannot be reliably identified using morphological characteristics. The use of morphological characteristics has led to numerous errors and confusion in oyster classification (Harry, 1985; Li & Qi, 1994;Wang, Zhang & Zhang, 2004), with negative effects on culture efficiency and conservation. Hence, alternative identification techniques that are efficient and accurate are needed for oyster research and aquaculture. In some previous studies, DNA markers have been used for the identification of oyster species (Klinbunga et al., 2003, 2005; Cordes, Stubbs & Reece, 2005; Wang & Guo, 2008b). In particular, mtDNA data have been useful to identify oyster species by means of restriction-fragment length polymorphisms (Pie et al., 2006), denaturing-gradient gel electrophoresis (Yu & Li, 2008), nucleotide sequencing (Reece et al., 2008; Yu et al., 2003) or specific multiplex PCR (Wang & Guo, 2008a). However, these methods are mostly complicated and labour intensive, hindering their high-throughput applicability. High-resolution melting (HRM) analysis is a highly sensitive molecular method, which distinguishes specimens on the basis of single-nucleotide polymorphisms (SNPs) or small deletions in amplified DNA sequences (Reed, Kent & Wittwer, 2007; Vandersteen et al., 2007). All the processes of PCR amplification during HRM take place in the same tube during 2 h, the product is denatured and the changes in sample fluorescence with temperature are monitored. This technique has already been used to screen for genetic variation in human disease (Vossen et al., 2009), mammals (Berry & Sarre, 2007), insects (Malewski et al., 2010; David & Cheolho, 2013) and fish (Haynes et al., 2009; McGlauflin et al., 2010). Thus, it has been considered to be a rapid, sensitive, reliable and cost-effective method (Camila & Mauricio, 2011). In oysters, HRM has been successfully used for identification of Crassostrea species and for detection of hybrids between C. sikamea and C. angulata, based on mitochondrial cytochrome c oxidase subunit I (COI) sequences (Wang et al., 2014; Xu et al., 2014). As the rate of molecular evolution of COI is about three times greater than that of the mitochondrial large ribosomal subunit (16S rDNA) (Knowlton & Weigt, 1998), there is usually some difficulty in amplification using universal primers. Therefore, in this study we tested the utility of relatively conserved 16S rRNA sequences in HRM analysis, and show that this provides a fast and effective method to differentiate five commercially important species of Crassostrea oysters. In total, 95 oysters were collected from typical areas where they occur. Crassostrea gigas (n 1⁄4 20) was obtained from the rocks of coastal waters around Rushan, Shandong Province; C. angulata (n 1⁄4 19) was collected from Putian, Fujian Province; C. ariakensis (n 1⁄4 19) was sampled from the Sea of Ariake, Japan; C. hongkongensis (n 1⁄4 19) was collected from Beihai, Guangxi Province; and Crassostrea sikamea (n 1⁄4 18) was collected from Haimen, Jiangshu Province. Genomic DNA was extracted from adductor muscle by standard proteinase-K digestion and phenolchloroform extraction. The purity of the samples was measured on a NanoDrop-2000 spectrophotometer (NanoDrop Technologies) at 260/280 nm, adjusted to the same level. The DNA extracts were stored at 220 8C until used. Mitochondrial 16S rRNA sequences of all known Crassostrea species were downloaded from GenBank (accession numbers: AF280611, KC170322, FJ743507, HQ660978, JF808181). A 300-bp 16S rRNA sequence alignment that contained variation unique to each of the five species was examined for variation between species. Primers were designed using the Primer Premier v. 5.0 program (PREMIER Biosoft International). Design criteria included a similar annealing temperature at 60+ 2 8C and target amplicon size of 60–150 bp. A primer set (F: 50-AGATTTTTAGGTGGGGCG-30, R: 50-GCGTAAC TTCTCCTATGATCG-30) was created based on consensus regions between the different Crassostrea species and a 83 bp amplicon with unique form-specific melting curves, to ensure maximum resolution in HRM analysis. All DNA extractions were diluted to 10 ng/ml. PCR amplification and HRM analysis were performed on a LightCycler480 real-time PCR instrument (Roche Diagnostics). The reaction