Maximum parsimony
and maximum likelihood trees were generated with heuristic
searches. Starting tree(s) were obtained via stepwise addition.
Addition of sequences was at random with 100 replicates. The
branch-swapping algorithm was tree-bisection reconnection
(TBR). We ran two Bayesian analyses with four chains each, one
million generations, sampling frequency of 100, and a burn-in of
250,000 generations. Support for nodes was assessed with bootstrap
(1000 replicates for neighbour joining and maximum parsimony
and 100 replicates for maximum likelihood) and posterior
probability for Bayesian analyses.