DNA barcoding analysis
After photographing, the samples preliminarily identified to be different species were preceded with DNA barcoding. Total DNA was isolated from a small piece of muscle tissue of fish larvae or a single fish egg using standard Wizard® Genomic DNA Purification Kit and following the manufacturer’s instructions (Promega Inc, U.S.A.), rehydrated in ddH2O and examined through OD260/OD280 spectrometry and agarose gel electrophoresis, and then kept at 4 °C for subsequent PCR amplification.
Amplification of the COI DNA fragment via polymerase chain reaction (PCR) used the universal primers of FishF1 and FishR1 (Ward et al. 2005) and referred PCR conditions as described by Hebert and Gregory (2005), Becker et al. (2015), and Ivanova et al. (2007). PCR products were verified via gel electrophoresis using 1.5% agarose gel stained with SybrGreen fluorescent dye. Quality amplicons were sequenced at Sangon Biotech Cooperation Ltd (Shanghai, China).
DNA sequences were confirmed, edited and assembled using Sequencher v5.4 (Gene Codes) and then aligned in Clustal X2 (Chenna et al. 2003). Additional sequences were retrieved from the GenBank database (http://www. ncbi.nlm.nih.gov/) via the Basic Local Alignment Search Tool (BLAST) for identifying specimens when the resulting sequences were at a minimum of 98% similarity. K-2-P distances (Kimura 1980) were used to calculate pairwise differences of the sequences in MEGA v.7.0 (Kumar et al. 2016). In addition, a clustering analysis of neighbor-joining (NJ) based on K-2-P distances and 10,000 bootstrap replications was performed to provide a graphic representation of relationship of taxa using MEGA v.7.0.