Bayesian analyses of the nucleotide and amino acid analyses were each carried out using MrBayes, v.3.1.2 [75]. Following the recommendations of Rota-Stabelli et al. [76], for the amino acid data set the mtZoa model was applied (settings were rates = gamma, ngammacat = 5, aamodel = fixed (rtrev) ); two chains (temp = 0.2) were run for 5,000,000 generations and sampled every 1,000 generations. Using the unsaturated gene partitions only, Modeltest 3.7maxX [77] was used to estimate a suitable model for nucleotide substitution; this was equivalent to GTR+I+G and settings were nst = 6, rates = invgamma, ngammacat = 4. Four chains (temp = 0.2) were run for 5,000,000 generations and sampled every 1,000 generations. For each analysis convergence was assessed using Tracer v 1.4 [78], with a discarded burn-in period of 5,000 trees. Posterior probabilities provided evidence of nodal support. All trees were rooted against Hymenolepis diminuta.