Distance- based approaches
The Kimura two-parameter (K2P) and p-distance model are used to calculate the pairwise genetic distance and construct the neighbor-joining tree (NJ) on the MEGA7.0. The haplotype diversity and nucleotide diversity of COI sequences were calculated using DnaSP. Then ML tree analysis was implemented using RaxmlGUI. All analysis used the default parameters and 1,000 bootstraps. In all trees, bootstrap values below 70% are not shown (Shen et al., 2016 ).
According to the phylogenetic trees of the original sequencing sequences and morphological characteristics, no less than 3 DNA sequences of each category were selected for further species confirmation by the IDENTIFICATION of BOLD (Barcode of Life Data System) and the BLAST of NCBI (National Biotechnology Information Center), to evaluate the accuracy of the morphological identification results and obtain reference sequences with high relative similarity. In the selection of similarity sequences, we have defined 97% as a relatively loose standard to indicate potential species identification (E. H. K. Wong & Hanner, 2008 ).
In this study, a total 42 COI sequence with high similarity is obtained by aligning from GenBank. Gammarus pisinnus (GenBank accession number: KF824592) selected as the outer group. All DNA sequences obtained in this study were submitted to GenBank and BOLD; their accession numbers are provided in the electronic appendix (Table S2, Table S3), which can be found on the website of the National Center for Biotechnology Information (NCBI).
In addition to, Automatic Barcode Gap Discovery (ABGD) analysis was implemented on the website (http://wwwabi.snv.jussieu.fr/public/abgd/abgdweb.html), using K80, relative gap width (X = 1.0) and set the rest of the parameters as default values.