The phylogeny of the family Bacidiaceae (Lecanorales, Ascomycota) was investigated using 65 nuclear ITS1-5·8S-ITS2 ribosomal DNA sequences, 63 of which were newly determined. After exclusion of ambiguous alignment, the data set contained 285 variable characters, 212 of which were parsimony-informative. Phylogenetic estimations were performed with maximum parsimony (unweighted and weighted) and maximum likelihood optimality criteria. Four different phylogenetic hypotheses were tested using a parametric bootstrap approach to simulate the expected null distribution of the difference between the globally optimal tree and the best (constrained) tree agreeing with the null hypotheses under unweighted and weighted parsimony, and maximum likelihood: (1) the genus Bacidia is monophyletic; (2) the genus Bacidina is monophyletic; (3) the genus Toninia is monophyletic; and (4) the family Ramalinaceae is monophyletic and distinct from a monophyletic Bacidiaceae. The monophyly of Bacidia, Toninia, and the Ramalinaceae was rejected under all circumstances. Hence, Bacidiaceae is likely to be a younger synonym of Ramalinaceae. The monophyly of Bacidina was not rejected under any optimality criterion. Furthermore, the data set suggests that the Bacidia beckhausii and B. sabuletorum groups are unrelated to Bacidia s. str., that Megalaria is monophyletic, and that Lecania auct. is polyphyletic.

Fifty-three isolates of Pisolithus were obtained from various locations in central and eastern Australia. These, along with two isolates from south-east Asia and one from USA, were compared using ITS–RFLP and ITS sequencing analyses. Results of the RFLP analysis initially grouped the isolates into eight RFLP types. ITS sequences were obtained for at least one representative isolate from each RFLP type and compared by a neighbour-joining analysis with Pisolithus ITS sequences available in databases. The majority of isolates clustered into four groups within two major clades, each clade comprising isolates of similar basidiospore characteristics. Most Australian isolates corresponded with recent provisional descriptions of P. albus or P. marmoratus. One isolate (LJ30) had low sequence identity (61·6–78%) to the other isolates and probably represents a separate undescribed Australian species. Significant intraspecific variation was observed in ITS–RFLP profiles for the putative P. albus isolates, suggesting that the sole use of RFLP analysis in diversity assessment may overestimate Pisolithus species richness.

The phylogeny of the Dothideomycetes was investigated using nucleotide sequences of SSU and LSU rDNA. 32 new sequences of these regions from 18 species were aligned with five sequences obtained from GenBank, including two representatives of Pezizomycetes used as outgroup. A combined data set of SSU and LSU rDNA was analysed phylogenetically using neighbour-joining and maximum likelihood. The pseudoparaphysate taxa (Pleosporales incl. Melanommatales) form a monophyletic group. A separation of Melanommatales can be rejected using parametric bootstrapping, and this supports previous results obtained from SSU rDNA studies. The aparaphysate Dothideomycetes form a sister-group to the Pleosporales, but with low bootstrap support. Within these species, two well-supported groups can be distinguished, the Dothideales s. str. and the Capnodiales with Myriangiales as sister-groups. The Myriangiales appear paraphyletic, but this has only poor bootstrap support. A monophyly of Dothideales comprising Capnodiales can be rejected with parametric bootstrapping as well as a subdivision of the loculoascomycetes based on form and opening of the ascomata and ascus. The phylogeny of the aparaphysate taxa requires further studies.

Phylogenetic trees of 16 Claviceps species were constructed based on alignments of 5.8S rDNA and the adjacent ITS1 and ITS2 spacers. Two highly supported clades were found: (1) C. paspali, C. zizaniae, C. grohii, C. sulcata, C. fusiformis, and C. purpurea; and (2) C. citrina, C. phalaridis, two unidentified Claviceps spp. (isolates PM and SG), C. sorghicola, C. gigantea, C. sorghi, C. africana, C. viridis, and C. pusilla. No relationship was found between the species placement and its morphological markers. The probe from C. purpurea gene cpd1 for dimethylallyl tryptophan synthase, the first enzyme of alkaloid biosynthesis, was hybridized to Pst I digested genomes of the above species under non-stringent conditions. Hybridizing DNA was present in all species of clade 1, although the signal of the C. paspali gene was weaker. In clade 2, only C. africana, C. gigantea, and C. pusilla gave weak positive signals. Colorimetric detection found small amounts of alkaloids in cultures of Claviceps sp. SG and PM but despite that, no cpd1 hybridizing bands were found. The occurrence of two major clades of Claviceps and their biogeography suggests, that the genus originates from South America and that the evolution of its species was influenced by comigration with their hosts and with the global climatic changes that influenced spreading of grass subfamilies.

Arbuscular mycorrhizal (AM) fungi, classified in the order Glomales, cannot easily be identified by hyphal morphology. PCR can be used successfully to identify the fungi within the host or soil. Some of the primers that have been used are based on a few sequences of a variable region in the SSU rRNA gene. To verify the specificity of primers of this type which have been used in earlier publications, and to provide data for future primer design, we analysed 51 partial Glomales SSU rRNA gene sequences from 33 glomalean isolates.

To distinguish contaminants from the AM fungi (AMF) all sequences were verified by phylogenetic analyses. Together with sequences from the EMBL database a set of 88 sequences from 58 isolates was analysed, comprising 39 species. The phylogenetic trees reveal at least three distinct well supported Glomus clades and other interesting relationships within the Glomales. Sequence data show that most of the primers published in earlier studies are not useful at the proposed taxonomic level. For example, the VANS1 primer site is not specific for the Glomales, and is homologous to fewer than half of the AMF sequences investigated. Primer regions that have been considered to be family-specific within the Glomales do not show strict specificity. Because clades in the phylogenetic trees fit well to several sequence patterns, priming sites for AMF are discussed and sequence data for future analyses provided.

The genetic relationships among 132 isolates of Sclerotium rolfsii (teleomorph Athelia rolfsii) collected during 1967–97 from 36 different host species over a wide geographic range representing 13 countries were investigated using mycelial compatibility groupings and RAPD analysis. A smaller group of 15 Sclerotium delphinii isolates from five host species and a limited geographic distribution was also studied. The development of aversion reactions following mycelial pairings of isolates in all possible combinations on potato dextrose agar was used to differentiate 71 mycelial compatibility groups (MCG) in S. rolfsii and five MCG in S. delphinii. Many MCG were unique single-member groups and these generally were found in widely separated geographic regions or countries. There was no clear relationship between host of origin and MCG, except for a majority of isolates of S. rolfsii from turfgrass that belonged to MCG 1. Within a specific geographic region, e.g. California, there usually were several different MCG present, some of which were recovered from the same host species. In addition, specific MCG of S. rolfsii were recovered from widely separated geographic regions as well as different host species, e.g. MCG 1 was recovered from turfgrass, carrot, tobacco, and tomato in California, Georgia, North Carolina, and Mexico, respectively. The extent of genetic diversity within and among MCG of S. rolfsii and S. delphinii was studied using RAPD analysis. Isolates from different MCG could be differentiated by their unique banding patterns using six primers. There were no discernible relationships among the various MCG using UPGMA analysis. Isolates within a particular MCG were also genetically diverse, but shared greater numbers of common bands and clustered together. Only a few members of some MCG in S. rolfsii and S. delphinii that had identical RAPD patterns were considered to be clonally derived. The extent of genetic diversity among isolates of S. delphinii was lower than that observed in S. rolfsii.

Cercospora caricis is being considered as a bioherbicide agent for use against purple nutsedge (Cyperus rotundus), one of the world's worst weeds. However, first its efficiency must be improved, for example by genetic transformation. By optimizing physical and biological parameters, the particle gun acceleration method (biolistics) has been adapted to transform mycelial cells of C. caricis. Two genes were expressed in C. caricis by biolistic transformation. The β-glucuronidase gene (GUS) fused to the GDP1 promoter of Cochliobolus hetrostrophus and the hygromycin B resistance gene under control of the PtrpC promoter of Aspergillus nidulans. Although the transformation frequency was not high, all transformants were stable when they were propagated on a selective medium after eight subsequent transfers.

Seventy-three isolates of Phytophthora cinnamomi were collected from diseased Eucalyptus marginata (jarrah) and Corymbia calophylla (marri) trees in two forest communities in the southwest of Western Australia. Both populations of P. cinnamomi were examined for phenotypic and genotypic variation. Microsatellite DNA analysis showed that all isolates were of the same clonal lineage. We show, for the first time for P. cinnamomi, that morphological and pathogenic variation between populations of the clonal lineage are very broad and continuous. The phenotypes examined included growth rates and colony morphology on potato dextrose agar at different temperatures, sporangial and gametangial morphology, ability to form lesions in detached jarrah and marri stems, and ability to cause deaths of clonal jarrah plants in a glasshouse trial. Phenotype variation was derived asexually. All phenotypes investigated varied independently from one another. Cluster analysis of 24 morphological and pathogenicity phenotypes identified two main clusters of isolates corresponding to each population. The ability to cause deaths in both populations ranged from killing all plants within 59 d to plants being symptomless 182 d after inoculation.

Based on ITS and β-tubulin sequence data of 33 isolates, the newly introduced genus, Phaeomoniella was confirmed as being distinct from Phaeoacremonium (Pm.). Phylogeny inferred from DNA sequences and cultural characteristics also confirmed the species status of Pm. aleophilum and Pm. angustius, which were recently reduced to synonymy. Pm. aleophilum has an optimum growth rate at 30 °C and the ability to grow at 35°, whereas Pm. angustius has an optimum growth rate at 25 °C and does not grow at 35°. Furthermore, ITS and β-tubulin sequence data showed Pm. viticola to be indistinguishable from Pm. angustius, while a new species, Pm. mortoniae, could be distinguished from this complex.

The loculoascomycetes are defined by the initiation of ascoma development prior to dikaryotisation and functionally bitunicate asci. Previous molecular studies suggested non-monophyly of loculoascomycetes and consequently, Chaetothyriomycetes and Dothideomycetes were distinguished. We have used sequences of nu SSU rDNA, nu LSU rDNA, and mt SSU rDNA to re-evaluate the monophyly of loculoascomycetes. 13 new sequences of these regions from 10 species, including two representatives of Pezizomycetes used as outgroup, were aligned with sequences obtained from GenBank. A combined data set was analysed phylogenetically using maximum parsimony. The Chaetothyriomycetes and Eurotiomycetes form a sister-group, supported by a bootstrap value of 97%, suggesting that loculoascomycetes are not monophyletic. A topology constrained to loculoascomycete monophyly can be rejected as being significantly worse using parametric bootstrapping. Our results also indicate that mt SSU rDNA sequence data are useful as additional characters to elucidate the phylogeny of ascomycetes at the rank of different classes. Additional data from more representatives of these groups and other ascomycetes are required to clarify whether the loculoascomycetes are a paraphyletic or polyphyletic assemblage.

To obtain an understanding of the phylogenetic relationships among the Georgefischeriales, septation, cellular interactions, teliospores, basidia, cultures and nucleotide sequences from the 5′ terminal domain of the nuclear large subunit rRNA gene were studied. Analyses of both morphological and molecular characters yield similar phylogenetic conclusions. The Georgefischeriales are divided into three groups, corresponding to the Eballistraceae, Georgefischeriaceae, and Tilletiariaceae. The basal dichotomy is between the Eballistraceae and the branch uniting the Georgefischeriaceae and Tilletiariaceae. The Tilletiariaceae are phragmobasidiate, whereas the Eballistraceae and the Georgefischeriaceae are holobasidiate. The Eballistraceae differ from the Georgefischeriaceae and Tilletiariaceae in the lack of the ballistospore mechanism. The systematic position of Tilletiopsis minor is unclear. The Eballistraceae, Eballistra and Phragmotaenium are proposed as new taxa. The descriptions of the Tilletiariaceae and Jamesdicksonia are emended. Except for Entyloma majus, E. parvum, Georgefischeria, Jamesdicksonia brunkii, J. obesa, Tilletiaria anomala, and Tolyposporella chrysopognis, the teleomorphic species of the Georgefischeriales are presented as new combinations.

The number of known species of fungi is estimated as at least 74 K, but could be as much as 120 K with allowances for ‘orphaned’ species. Yet in 1990 the magnitude of fungal diversity was estimated ‘conservatively’ at 1·5 M species. This figure has been widely accepted as a working hypothesis, but subsequent estimates have ranged from 500 K to 9·9 M and the bases of these suggestions are analyzed. Additional data pertinent to the estimation of the number of fungal species on Earth that has become available since 1990 is discussed. Site inventories demonstrate the need for long-term (20 yr plus) intensive studies to determine the number of species in a site. Fresh data sets on fungus[ratio ]plant ratios and degrees of host specificity, especially from well-studied hosts in the tropics, are consistent with earlier estimates. The extent of novelty discovered in recent monographic generic revisions and studies of species in particular habitats varies from 0–96%. Allowances for cryptic species, now known to be widespread by incompatibility and molecular studies, could on their own justify an upward revision by a factor of at least five. To enable confidence in any overall estimate to be increased, more detailed studies, especially on particular sites in the tropics, are needed. The consensus of tropical and molecular mycologists in particular is that an increased estimate could be justified. However, it is prudent to retain 1·5 M as the current working hypothesis for the number of fungi on Earth while additional data to test it further accumulates.

The Laboulbeniales have had a clouded taxonomic history. Initially interpreted as abnormal cuticular outgrowths of arthropods, they have since been considered to be acanthocephalans, basidiomycetes, and zygomycetes as well as ascomycetes. Phylogenetic analysis of partial sequences of SSU rDNA with species of Chytridiomycota as outgroup taxa, and including trichomycete and entomophthoralean Zygomycota, provides evidence of the placement of a Laboulbeniales-Pyxidiophora clade among the Ascomycota. Support for parsimony analyses came from maximum likelihood ratio tests applied to trees with ten different topological constraints. The placement among Ascomycota, excluded from Sordariomycetes, is well supported, and indicative of a distinct lineage, class Laboulbeniomycetes. At this time, however, we are unable to discover close relatives of the Laboulbeniomycetes clade among other filamentous ascomycetes.

Mycosphaerella contains some of the most damaging foliar pathogens of Eucalyptus. Identification of the species has relied mainly on characters such as ascospore morphology and the mode of ascospore germination. Wide variation and overlap in morphological characters, especially ascospore shape and size, has lead to confusion in species concepts. This study examines the potential of RAPD–PCR to differentiate Mycosphaerella species that cause leaf diseases of Eucalyptus in Australia. RAPD–PCR was conducted on 39 isolates representing 10 species of Mycosphaerella collected from southern Australia. Primers were chosen mainly to distinguish between the two most common and damaging species in Australia, M. cryptica and M. nubilosa. The RAPD analysis revealed five groups within the 39 isolates, four of which represent distinct species: M. cryptica, M. gregaria, M. nubilosa and M. marksii. For these four species, the cluster analysis and RAPD patterns gave results that fitted well with classical taxonomic criteria. Variation within M. cryptica, M. nubilosa and M. marksii was low, even though the isolates within these species originated from diverse hosts and locations. The remaining species in this study grouped together and could not be easily separated using the RAPD data here.

Three fungal species responsible for anthracnose canker, perennial canker, and bull's-eye rot of apple have been considered members of the genus Pezicula for a number of years. Recent studies, however, have provided evidence to (re-)classify these species as Neofabraea. There has been a long historical debate regarding the taxonomy of two of these fungi. In Europe, both Neofabraea malicorticis and N. perennans have generally been considered N. malicorticis, while in North America a species distinction has been maintained. Phylogenetic analyses of Neofabraea isolates were based on DNA sequences of the internal transcribed spacer region of nuclear rDNA (38 isolates), the mitochondrial rDNA small subunit (partial; 48 isolates), the β-tubulin gene (partial; 25 isolates), and a combined data set (21 isolates). Our work provides evidence for the existence of four distinct Neofabraea apple pathogens including N. malicorticis, N. perennans, N. alba, and a putative new Neofabraea species that was isolated in both Europe and eastern North America. Our results indicate that the primary Neofabraea species causing tree cankers and bull's-eye rot in North America are N. malicorticis and N. perennans in the west, and N. alba in eastern Canada. N. perennans, N. alba, and the undescribed Neofabraea species were found in Europe but the presence of N. malicorticis was not confirmed by our limited sampling. Inclusion of Rosa spp. in the host range of N. malicorticis is merited.

The nucleotide sequences of the nuclear ribosomal DNA, including ITS1, ITS2 and the 5·8S rDNA, were determined for 54 specimens representing 13 Leveillula species on 50 different host plant species. The maximum nucleotide sequence diversity among taxa in ITS1-5·8S-ITS2 was 7·1%. Phylogenetic analyses revealed that Leveillula species formed six clades and three basal taxa. The taxonomic positions of several species that are well characterized by morphology of conidia, especially primary conidia, are supported by the present molecular analyses. This result shows that the morphology of primary conidia mostly provides a good criterion to identify Leveillula species. Twenty-six collections of Leveillula taurica recovered from 14 different plant families formed a distinct clade with L. chrozophorae, L. duriaei and L. elaeagni, and showed high homology in the ITS regions. Eight isolates of L. taurica recovered from the Asteraceae, Balsaminaceae, Fabaceae and Campanulaceae showed high sequence diversity, in contrast with other L. taurica specimens, and clustered separately or with other taxa. The result of Kishino-Hasegawa and Templeton tests demonstrated that the possibility of monophyly of L. taurica s. lat. could be significantly rejected, indicating that L. taurica s. lat. is a species complex composed of several biological species.

The two subpopulations of the Dutch elm disease pathogen Ophiostoma novo-ulmi, previously known as the Eurasian (EAN) and North American (NAN) races, are redesignated as subspecies novo-ulmi and americana. In addition to their partial reproductive isolation, wide range of physiological and molecular differences and different geographic ranges, the two subspecies can be discriminated by their perithecial form and dimensions. These perithecial differences are described, using perithecia produced in multiple intra-subspecies crosses. Perithecia of subsp. novo-ulmi have an average neck length of ca 450 μm, base width of ca 103 μm and neck length:base width ratio ca 4.4. Perithecia of subsp. americana have an average neck length of ca 295 μm, base width ca 116 μm and neck length:base width ratio ca 2.6. The average shape and dimensions of subsp. americana perithecia is similar to that of O. ulmi. The average perithecial form of subsp. novo-ulmi, as well as of O. himal-ulmi, is rather distinctive. The current known geographical distribution of subspecies novo-ulmi and americana, based on > 6500 samples, is presented.

SSU rRNA gene sequences of Anamylopsora pulcherrima (Anamylopsoraceae), Placopsis gelida, Trapelia involuta and T. placodioides (Agyriaceae) were determined and aligned with the corresponding sequences of 39 other ascomycetes. Phylogenetic analysis (maximum parsimony, spectral analysis) of these sequences placed the Agyriineae outside the Lecanorales and suggested a sister group relationship to Ostropales/Pertusariales. The resurrection of the order Agyriales is proposed based on molecular and morphological data, such as faintly amyloid asci opening by dehiscence. The sequence data support the two families Agyriaceae and Anamylopsoraceae to be accommodated within the order Agyriales. The Agyriales, Ostropales and Pertusariales can be placed in the class Lecanoromycetes.

Species of Mycosphaerella and their anamorphs are commonly found on the leaves of Myrtaceae, many of which are defoliated by these pathogens. The taxonomy of these fungi has been based on minute morphological differences, and virtually nothing is known regarding their relatedness to each other. In this study, we present a phylogeny of 30 species of Mycosphaerella or their anamorphs from myrtaceous hosts, based on sequence data from the ITS regions of the ribosomal RNA operon. Fifteen of the species were also analysed for the 5′ end of the large subunit (28S), which produced a phylogeny similar to that obtained for the ITS data set. The Mycosphaerella species included in this analysis are all regarded as representatives of section Plaga, and appear to represent a monophyletic assemblage. Mycosphaerella lateralis was the only species shown to have a wide host range. In general, species clustered together based on their anamorph genera. Species with Colletogloeopsis and Stenella anamorphs always grouped in two respective clusters. However, species with Mycovellosiella, Phaeophleospora, Pseudocercospora and Uwebraunia anamorphs occurred separately, suggesting that they have evolved more than once within Mycosphaerella. Based on the ITS data set, all morphospecies were also shown to be phylogenetic species, although too few isolates were available to address questions relating to intraspecies variation. Nevertheless, ITS sequence data proved sufficient to distinguish morphologically similar taxa that have hitherto only been distinguished based on ascospore germination patterns and anamorph characteristics. Sequence data presented in this study should facilitate the identification of Mycosphaerella species occurring on Myrtaceae in the future.

Simple sequence repeat (SSR)–PCR amplification using a microsatellite primer (GACA)4 and ribosomal RNA gene sequencing were used to examine the intraspecific diversity in the mycoparasite Coniothyrium minitans based on 48 strains, representing eight colony types, from 17 countries world-wide. Coniothyrium cerealis, C. fuckelii and C. sporulosum were used for interspecific comparison. The SSR–PCR technique revealed a relatively low level of polymorphism within C. minitans but did allow some differentiation between strains. While there was no relationship between SSR–PCR profiles and colony type, there was some limited correlation between these profiles and country of origin. Sequences of the ITS 1 and ITS 2 regions and the 5.8S gene of rRNA genes were identical in all twenty-four strains of C. minitans examined irrespective of colony type and origin. These results indicate that C. minitans is genetically not very variable despite phenotypic differences. ITS and 5.8S rRNA gene sequence analyses showed that C. minitans had similarities of 94% with C. fuckelii and C. sporulosum (which were identical to each other) and only 64% with C. cerealis. Database searches failed to show any similarity with the ITS 1 sequence for C. minitans although the 5.8S rRNA gene and ITS 2 sequences revealed an 87% similarity with Aporospora terricola. The ITS sequence including the 5.8S rRNA gene sequence of Coniothyrium cerealis showed 91% similarity to Phaeosphaeria microscopica. Phylogenetic analyses using database information suggest that C. minitans, C. sporulosum, C. fuckelii and A. terricola cluster in one clade, grouping with Helminthosporium species and ‘Leptosphaeria’ bicolor. Coniothyrium cerealis grouped with Ampelomyces quisqualis and formed a major cluster with members of the Phaeosphaeriacae and Phaeosphaeria microscopica.