Mitochondrial Sequencing and haplotype network
Mitochondrial DNA haplotypes were obtained from a subset of 67 individuals from the five genetic clusters. PCR reactions were performed using a Taq-Pro COMPLETE kit (Denville Scientific, CB4065-4) targeting a partial region of the COI gene with the Lep50 primer set from Folmer et al. (1994) and the thermocycler profile from Herbert et al. (2003). PCR products were cleaned using an EXOSAP-IT kit (ThermoFisher, 78201.1.ML) and prepared for sequencing using a BigDye Terminator v.3.1 Cycle Sequencer Kit (Applied Biosystems, 4337454). Sanger sequencing was done using an Applied Biosystems 3500 Genetic Analyzer. Chromatograms were cleaned and aligned using the program Geneious v.9.1 (Kearse et al., 2012).
A haplotype network analysis was conducted using the 67 COI sequences obtained in this study combined with 218 C. variipennis complex sequences previously collected (M. Hopken unpublished data). Sequences were aligned in MEGA v.10.1.8 (Kumar, Stecher, Li, Knyaz, & Tamura, 2018) and trimmed to 546 bp to ensure all sequences contained identical lengths. A median-joining analysis was performed using NETWORK v.5.0.1.0 (Bandelt, Forster, & Rohl, 1999). Specimens collected in this study were assigned a color based on the results from the SNP clustering analyses while the remaining samples were left unassigned. All individuals were used to calculate the mean uncorrectedp- divergence between and within the different groupings inferred from the haplotype network using MEGA.