Individual samples were recovered by identifying individual barcodes and adapters. Quality score trimming was performed with default settings using MiSeq Reporter Software (Illumina, San Diego, CA, USA). Paired-end reads were assembled and mapped to their respective GenBank reference sequences in CLC Genomics v. 7.5
https://www.qiagenbioinformatics.com/ Nuclear ambiguities were called when two alleles were present at a 0.3 frequency threshold. Loci were aligned via MUSCLE v. 7 (Edgar, 2004). NCBI GenBank references used for mapping included NC021751, NC028013, HM106320, AP017412, HM106321, NC024942, KT901457, and HM106318 for mitogenomes, AB564081, EF987774, AF498194, AB564085, AB564089, EF987776, EF987777, EF987778, and DQ205808 for GHR, and AB564080, EF987758, AF498170, AB564084, AB564088, EF987761, AB564084, and EF987763 for FES.
Proposed species can be most easily diagnosed using genetics as their mitochondrial DNA is distinct and reciprocally monophyletic. For mitochondrial gene cytochrome oxidase I (COX1, 1,542 bp), the barcoding gene, a minimum 47 nucleotide changes distinguish M. erminea and M. haidarum haplotypes from M. richardsonii and 29 nucleotide changes diagnose M. erminea from M. haidarum. Further, for mitochondrial gene cytochrome b (cytb, 1,137 bp), a minimum of 15 nucleotide changes separate M. erminea and M. haidarum cytb haplotypes. 30 nucleotide differences distinguish M. erminea from richardsonii and 31 distinguish M. haidarum from M. richardsonii.