Estimation of the origin of the S morph lineage
The relaxed clock method was used to estimate the approximate divergence date of the S gyne from the L queen. Three sets of monophyletic lineages, thought to have diverged approximately 1 MYA (million years ago), 2 MYA, and 3 MYA, were used, i.e., Myrmica excelsa andM. taediosa , M. sulcinodis and M. xavieri , andM. tobiasi and M. georgica , respectively (GenBank Acc. No: FJ824432, GQ255131, GQ255141, GQ255197, GQ255192, and GQ255145) (Jansen & Savolainen, 2010). Jansen & Savolainen (2010) estimated the divergence time of holarctic Myrmica ants using mitochondrial and nuclear genes; the COI data and their estimates were extracted.
The HKY+G+I model (gamma distribution shape value: 1.26247; proportion of invariant sites: 0.61287) was selected as the best fit evolutionary substitution model based on the Bayesian information criterion, as determined using MEGA (Kumar et al., 2008; Kumar et al., 1994). For the clock method, Bayesian Markov chain Monte Carlo was run for 100 million generations. Trees were sampled every 1000 generations using BEAST (ver. 1.8.0) (Drummond & Rambaut, 2007). Posterior distributions for parameter estimates and likelihood scores were visualized using Tracer (ver. 1.5) to examine tree appropriateness. The trees were consolidated to a maximum clade credibility tree with median heights after discarding the first 15000 trees as burn-in. The resultant tree was visualized, with 95% HPD (highest posterior density), using FigTree (ver. 1.40). It was further edited with additional data using Adobe Illustrator CS6 (Adobe Inc.).
Association between wing morphology and Wolbachia infection status
The chi-square independence test in SPSS (Release 17.0; Chicago, IL, USA) was used to examine whether there is a relationship between wing morph and Wolbachia infection. For statistical analysis, the three USA individuals were excluded owing to uncertainty with respect to their wing morphology.