Supervisor: teadur Kadri Põldmaa
Opponent: prof. Donald H. Pfister (Harvardi Ülikool, USA)
Summary:
Systematics of many fungal groups at different taxonomic levels has remained obscure. This can be ascribed to scarceness of morphological characters that are used for recognition of taxa. Pezizales, Hypocreales and Helotiales are three of the largest orders among ascomycetes with debatable taxonomy. In this study, genetic variability and taxon boundaries of different ranks were assessed in selected genera: at species level in a small subgroup of Hypomyces (Hypocreales), at species and genus level in Geopora (Pezizales), and at family level in available species of Encoelia (Helotiales, incertae sedis). rDNA as well as several protein-coding genes were used to assess phylogenetic relationships. Hypomyces rosellus turned out to be paraphyletic, comprising several cryptic as well as previously described species that have different patterns of geographical distribution and host specificity. The majority of isolates causing cobweb disease in mushroom farms were found to belong to Hypomyces odoratus. The genus Geopora is paraphyletic. Ascocarp type provides morphological distinction of two well-supported monophyletic groups which we suggest to recognise as separate genera. Apothecial G. arenicola, G. tenuis and G. sepulta and ptychothecial G. tolucana are recognised as monophyletic species. The genus Encoelia is highly polyphyletic. The type species of the genus, E. furfuracea, forms a strongly supported sister group to Hemiphacidiaceae. Only two species can be unequivocally placed to monophyletic families, Sclerotiniaceae and Rutstroemiaceae, respectively. Three species were found to belong to the monophyletic group 'Encoelioideae' sensu Peterson & Pfister. In Geopora and Encoelia, the resolving power of rDNA ITS enables to use this region as barcoding gene. In studied species of Hypomyces, however, the ITS does not provide sufficient distinction. Incorporation of protein-coding gene sequences in phylogenetic analyses greatly improves the resolution and support values for most branches in phylogenetic trees. Reference sequences from reliably identified fruit-bodies obtained in this study can help with identification of numerous fungal species from environmental and agricultural samples.