Higher level phylogenetic relationships of euascomycetes (Pezizomycotina) inferred from a combined analysis of nuclear and mitochondrial sequence data

Abstract

A combined data set of nuclear SSU rDNA, LSU rDNA, and mitochondrial SSU rDNA sequences was analyzed in order to examine the relationships of the major clades of euascomycetes. Partial sequences of 14 ascomycetes were determined and aligned with the corresponding sequences of 16 other ascomycetes retrieved from Genbank. The alignment was analyzed using maximum parsimony (MP) and a Bayesian analysis with Markov chain Monte Carlo (B/MCMC). The classification based on single-gene studies is supported, but the confidence is enhanced in the concatenated analysis. The monophyly of the superclass Leotiomyceta, which includes all euascomycetes with inoperculate asci, is strongly supported. The polyphyly of ascolocularous fungi is supported. The group is divided into two groups: the Dothideomycetes basal to all other Leotiomyceta and the Chaetothyriomycetes as sister-group to Eurotiomycetes. The Lecanoromycetes appear as a monophyletic group with strong support and form a sister-group to the Chaetothyriomycetes/Eurotiomycetes clade, but this lacks support. The Leotiomycetes and Sordariomycetes form a strongly supported sister-group. Alternative topologies are tested using parametric bootstrapping; a basal position of the Eurotiomycetes and Leotiomycetes in the Leotiomyceta cannot be rejected, while such a position can be rejected for Chaetothyriomycetes, Lecanoromycetes and Sordariomycetes. The character evolution with regard to ascoma type, ascus type and ascoma-ontogeny is examined using MP and maximum likelihood (ML). While it appears most likely that the ancestor of the inoperculate ascomycetes had apothecia and an ascohymenial ascoma-ontogeny using MP methods, the ML approach shows that there is some uncertainty at the current state of knowledge. The improvement of confidence of the combined data set in comparison with single-gene studies makes us confident that analyses with additional data sets will further improve the confidence and eventually uncover the branching order of euascomycetes.