Hypogymnia phylogeny, including Cavernularia, reveals biogeographic structure

DNA sequence evolution through nucleotide substitution may be assimilated to a stationary Markov process. The fundamental equations of the general model, with 12 independent substitution parameters, are used to obtain a formula which corrects the effect of multiple and parallel substitutions on the measure of evolutionary divergence between two homologous sequences. We show that only reversible models, with six independent parameters, allow the calculation of the substitution rates. Simulation experiments on DNA sequence evolution through nucleotide substitution call into question the effectiveness of the general model (and of any other more detailed description); nevertheless, the general model results are slightly superior to any of its particular cases.

A key element to a successful Markov chain Monte Carlo (MCMC) inference is the programming and run performance of the Markov chain. However, the explicit use of quality assessments of the MCMC simulations-convergence diagnostics-in phylogenetics is still uncommon. Here, we present a simple tool that uses the output from MCMC simulations and visualizes a number of properties of primary interest in a Bayesian phylogenetic analysis, such as convergence rates of posterior split probabilities and branch lengths. Graphical exploration of the output from phylogenetic MCMC simulations gives intuitive and often crucial information on the success and reliability of the analysis. The tool presented here complements convergence diagnostics already available in other software packages primarily designed for other applications of MCMC. Importantly, the common practice of using trace-plots of a single parameter or summary statistic, such as the likelihood score of sampled trees, can be misleading for assessing the success of a phylogenetic MCMC simulation. The program is available as source under the GNU General Public License and as a web application at http://ceb.scs.fsu.edu/awty.

Despite the recent progress in molecular phylogenetics, many of the deepest relationships among the main lineages of the largest fungal phylum, Ascomycota, remain unresolved. To increase both resolution and support on a large-scale phylogeny of lichenized and non-lichenized ascomycetes, we combined the protein coding-gene RPB2 with the traditionally used nuclear ribosomal genes SSU and LSU. Our analyses resulted in the naming of the new subclasses Acarosporomycetidae and Ostropomycetidae, and the new class Lichinomycetes, as well as the establishment of the phylogenetic placement and novel circumscription of the lichen-forming fungi family Acarosporaceae. The delimitation of this family has been problematic over the past century, because its main diagnostic feature, true polyspory (numerous spores issued from multiple post-meiosis mitoses) with over 100 spores per ascus, is probably not restricted to the Acarosporaceae. This observation was confirmed by our reconstruction of the origin and evolution of this form of true polyspory using maximum likelihood as the optimality criterion. The various phylogenetic analyses carried out on our data sets allowed us to conclude that: (1) the inclusion of phylogenetic signal from ambiguously aligned regions into the maximum parsimony analyses proved advantageous in reconstructing phylogeny; however, when more data become available, Bayesian analysis using different models of evolution is likely to be more efficient; (2) neighbor-joining bootstrap proportions seem to be more appropriate in detecting topological conflict between data partitions of large-scale phylogenies than posterior probabilities; and (3) Bayesian bootstrap proportion provides a compromise between posterior probability outcomes (i.e., higher accuracy, but with a higher number of significantly supported wrong internodes) vs. maximum likelihood bootstrap proportion outcomes (i.e., lower accuracy, with a lower number of significantly supported wrong internodes).