In the continual search by both pharmaceutical and agricultural industries for new products, natural selection has been found to be superior to combinatorial chemistry for discovering novel substances that have the potential to be developed into new industrial products. Since natural products are adapted to a specific function in nature, the search for novel secondary metabolites should concentrate on organisms that inhabit novel biotopes. Endophytic fungi inhabit such a biotope. In the course of the last 12 years, we have isolated ∼ 6500 endophytic fungi from herbaceous plants and trees, screened them for biological activities, and have isolated and determined the structures of the biologically active compounds. Correlations were found between biological activity and biotope, e.g. a higher proportion of the fungal endophytes, in contrast to the soil isolates, inhibited at least one of the test organisms for antialgal and herbicidal activities. The substances isolated originated from different biosynthetic pathways: isoprenoid, polyketide, amino acid derivatives, and belonged to diverse structural groups: terpenoids, steroids, xanthones, chinones, phenols, isocumarines, benzopyranones, tetralones, cytochalasines, and enniatines. The potential role of the endophyte and its biologically active metabolites in its association with its host has been investigated. The fungal endophytes possess the exoenzymes necessary to colonize their hosts and they grow well in the apoplastic washing fluid of the host. When the roots of larch are colonized, the association with the host may be mutualistic, improving growth of the host and supplying the mycobiont with enough nourishment to extensively colonize the host's roots. The concentrations of some plant defence metabolites are lower than in the control when the host is infected with a pathogen than with an endophyte. We hypothesize that the interaction fungal endophyte–plant host is characterized by a finely tuned equilibrium between fungal virulence and plant defence. If this balance is disturbed by either a decrease in plant defence or an increase in fungal virulence, disease develops. Not only must the endophyte synthesize metabolites to compete first with epiphytes and then with pathogens in order to colonize the host, but presumably also to regulate metabolism of the host in their delicately balanced association. The utilization of a biotope such as that of the fungal endophyte is one aspect of intelligent screening, another very important one is the taxonomy of the fungus in order to avoid redundant structural isolations. It is not a random walk through a random forest. Many groups of fungi in different biotopes are waiting to be exploited.

Verticillium fungicola, a severe mycopathogen of the cultivated mushroom Agaricus bisporus, was successfully transformed using both PEG-mediated and Agrobacterium-mediated techniques. PEG-mediated co-transformation was successful with hygromycin B resistance (hph), uidA (β-glucuronidase GUS), and green fluorescent protein (GFP) genes. Agrobacterium-mediated transformation was successful with the hph gene. Transformation frequencies of up to 102 transformants per μg DNA and 4068 transformants per 105 conidia were obtained for PEG-mediated and Agrobacterium-mediated transformation respectively. Expression of integrated genes in co-transformants was stable after 18 months of successive sub-culturing on non-selective medium, and following storage at −80 °C in glycerol. Molecular analysis of PEG-mediated transformants showed integration of the transforming genes into the target genome. Molecular analysis of Agrobacterium-mediated transformants showed integration of transforming DNA as single copies within the target genome. Co-transformants exhibited symptoms of disease in inoculation experiments and were at least as virulent as the wild-type fungus. GFP and GUS expression were observed in-vivo with the GFP-tagged strain showing great potential as a tool in epidemiological and host-pathogen interaction studies. The development of transformation systems for V. fungicola will allow in-depth molecular studies of the interaction of this organism with A. bisporus.

The phylogeny of Trichoderma and the phylogenetic relationships of its species was investigated by maximum parsimony analysis and distance analysis of DNA sequences from multiple genetic loci. 18S rDNA sequence analysis suggests that the genus Trichoderma evolved at the same time as Hypomyces and Fusarium and thus about 110 Myr ago. 28S rDNA sequence analysis shows that the genus Trichoderma is part of a monophyletic branch within the Hypocreaceae, which also includes Arachnocrea and Aphysiostroma in basal positions. Gene trees inferred from a combined analysis of the nuclear ribosomal internal transcribed spacer (ITS1 and 2), the D1 and D2 region of the 28S rDNA, the small subunit of the mitochondrial rDNA (mitSSU), the fifth and part of the sixth exon of translation elongation factor 2 (tef1), and a fragment of ech42 provide strong statistical support for a phylogeny consistent with the existence of four clades: clade A comprises species of Bissett's (1991) sect. Trichoderma but also T. hamatum, T. pubescens, T. asperellum, and T. strigosum; clade C comprises all the species contained in section Longibrachiatum as revised by Samuels et al. (1998), and clade D contains only T. aureoviride which is genetically most distant to all other species. Clade B, on the other hand, contains a large and taxonomically heterogeneous mixture of species, among which several subclades could be identified: subclade B1 containing H. lactea, H. citrina, H. citrina var. americana, H. lutea, and an unnamed T. sp. 1; subclade B2 containing T. stromaticum, and an unnamed T. sp. PPRI3559; subclade B3 containing T. fertile, T. oblongisporum, and H. hunua; subclade B5 containing T. polysporum, T. croceum, Hypocrea pilulifera, and T. minutisporum; and a larger subclade (B4), in which three strain clusters could be distinguished: one comprising T. harzianum, T. inhamatum, H. vinosa, and T. aggressivum, another one containing T. fasciculatum, T. longipile, and T. strictipile; and a third containing T. virens, T. flavofuscum, and T. crassum. The position of the remaining species of subclade B4 (T. spirale, T. cfr aureoviride, H. tawa, and T. tomentosum) was not resolved. A comparison of the topologies of the individual gene trees was concordant with the topology of the combined tree in most cases, but also revealed incongruent positions for a few species (T. oblongisporum, T. longipile, T. fasciculatum) which was most pronounced in the ITS1 and 2 tree. The results confirm the recent concept for sects Longibrachiatum and Trichoderma, indicate that the sects Hypocreanum and Pachybasium cannot be distinguished phylogenetically, and provide a first phylogenetic basis for dissection of the latter two sections.

Rhynchosporium secalis causes an economically important foliar disease of barley, rye, and other grasses known as leaf blotch or scald. This species has been difficult to classify due to a paucity of morphological features; the genus Rhynchosporium produces conidia from vegetative hyphae directly, without conidiophores or other structures. Furthermore, no teleomorph has been associated with R. secalis, so essentially nothing is known about its phylogenetic relationships. To identify other fungi that might be related to R. secalis, the 18S ribosomal RNA gene and the internal transcribed spacer (ITS) region (ITS1, 5.8S rRNA gene, and ITS2) were sequenced and compared to those in databases. Among 31 18S sequences downloaded from GenBank, the closest relatives to R. secalis were two species of Graphium (hyphomycetes) and two other accessions that were not identified to genus or species. Therefore, 18S sequences were not useful for elucidating the phylogenetic relationships of R. secalis. However, analyses of 76 ITS sequences revealed very close relationships among R. secalis and species of the discomycete genera Tapesia and Pyrenopeziza, as well as several anamorphic fungi including soybean and Adzuki-bean isolates of Phialophora gregata. These species all clustered together with 100% bootstrap support. On the basis of these results, the teleomorph of R. secalis, if it exists, most likely will be a small apothecium produced directly on dead, infected host tissue. The ITS analysis also indicated that higher-level classifications within the discomycetes need to be revised, and that Tapesia and Pyrenopeziza probably do not belong in the Dermateaceae.

Species of Melanospora are parasitic on or closely associated with other fungi. The morphological characters traditionally used to place Melanospora and the Ceratostomataceae in Sordariales do not agree with their placement in our phylogenetic study. The nuclear encoded small subunit ribosomal DNA (SSU nrDNA) grouped four species of Melanospora and several allied genera and species in Hypocreales, an order rich in mycoparasites. An analysis based on characters derived from deduced amino acids from the second largest subunit of the RNA polymerase II (RPB2) gene further supports the close relationship of Melanospora and Hypocreales. The nuclear encoded large subunit ribosomal DNA (LSU nrDNA) phylogeny inferred six groups for 40 species in Hypocreales and allied genera (Bionectria, Claviceps, Hypocrea, Nectria, and Niesslia). Four species of Melanospora, including the type, and two Sphaerodes species formed a basal clade. Scopinella was placed in a distinct clade. Two other allies of Melanospora, Syspastospora and Persiciospora, were closely related to or within the Hypocrea and Nectria groups, respectively. The study argues for the reclassification of taxa that previously were placed in the Sordariales.

The species of Psilocybe sect. Psilocybe, formerly classified in the genus Deconica, were investigated using morphology, mating behaviour and RAPD analysis. Psilocybe inquilinus and P. crobula do not seem to be closely related. Based on the morphology, two varieties could be accommodated in P. subviscida, namely as vars. subviscida and velata. The mating group of P. montana is characterized by rather thick-walled, dark spores with a fairly large germ-pore. Putative representatives of P. muscorum and P. physaloides freely interbreed with typical P. montana, and, consequently, these taxa are considered to represent one variable species. The ex-type strain of P. chionophila did not mate with isolates of P. montana. One collection of P. chionophila from a lowland habitat, morphologically resembling P. montana, was found to be interfertile with the ex-type strain of P. chionophila, but not with P. montana. We identified several collections as P. magica, which is morphologically similar to P. schoeneti. Mating studies showed that these specimen belong to the same biological species, but failed to mate with the ex-type of P. schoeneti.

The symbiotic phenotype of a lichen arises through the interaction and cooperation of two or more genetically unrelated partners. Ultrastructural and molecular methods were used to investigate the changes that take place during early stages of the lichenization process. The resynthesis of prethallus stages of Baeomyces rufus was studied by co-culturing under sterile conditions the isolated, axenically grown mycobiont and its green algal photobiont Elliptochloris bilobata. The lichenization process was monitored by SEM. One day after co-culture, symbionts were bound together by a newly secreted mucilage. By day 12, photobiont induced, morphological changes in the mycobiont were visible. Aerial hyphae grew around photobiont cells, showed a high frequency of branching and formed appressoria on the algal wall surface. By day 28, many photobiont cells were completely engulfed by hyphae and soredia-like clusters were observed. These morphological developments resemble lichenized structures formed in the natural lichen. cDNA-AFLP was used to investigate gene expression profiles on day 12 of co-culture. Differential gene expression patterns revealed that few genes were induced, and many fungal and algal genes seemed to be suppressed in the early stages of lichenization.

All isolates of Rhizoctonia solani (teleomorph Thanatephorus cucumeris) anastomosis group (AG) 3 from potato tubers in commercial fields in Scotland were found to contain more than one double-stranded RNA (dsRNA) element ranging in size from 0·3 to 15 kb. Strains of other AGs (2-1, 3, 4, 5) from culture collections also contained dsRNA elements. Isolates of AG 3 from different black scurf lesions on single tubers showed similar dsRNA profiles and, when opposed on agar plates, underwent compatible hyphal fusion where the colonies met. Isolates from different tubers in the same commercial field often showed somatic incompatibility (mean 44%) leading to cytoplasmic death when colonies fused. Isolates from different fields always showed somatic incompatibility after hyphal fusion, suggesting that the complex population structure of AG 3 could significantly limit the transmission of dsRNA between strains. The incidence of different dsRNA elements in AG 3 isolates could not be related to pathogenicity in seedling tests on tomato, carrot, lettuce, radish, cress or onion. Double-stranded RNA elements could not be eliminated from AG 3 isolates by repeated hyphal-tip subculture or extended (8 week) incubation on agar containing cycloheximide. Some isolates lost the smaller dsRNA elements (1·0 to 1·6 kb) during these procedures, and one isolate (of AG 4) lost a 6·5 kb dsRNA element, but the larger elements (up to 9·5 or 12 kb) were always retained. When assessed by low stringency hybridisation, the dsRNA elements of R. solani AG 3 and of strains from AGs 2-1, 4 and 5 showed no homology with a plasmid containing cDNA of the complete hypovirus dsRNA genome of Cryphonectria parasitica.

DNA was extracted from soil collected from the upper 15 cm of the profile at three sclerophyll forest sites in New South Wales three days before, and 14 d after prescribed burning events. rDNA internal transcribed spacer (ITS) regions were amplified by PCR using fungus-specific primers, and ITS products cloned. Restriction fragment length polymorphism (RFLP) analysis was conducted on the ITS clones using the endonucleases HaeIII, HinfI, MboI and AluI. Cloned ITS products from the three sites were thus separated into 120 RFLP types and the ITS product for each RFLP type was sequenced. Comparison of sequences with those available in the GenBank nucleotide database allowed putative assignment of ITS clones to the following functional groups: ectomycorrhizal (ECM), arbuscular mycorrhizal (AM), ericoid mycorrhizal (ERM), decomposer basidiomycetes (DEC), other soil fungi (OSF). While abundant ECM-like clones were amplified from pre-fire soils at the North Rocks and South Maroota sites, they were largely absent in the post-fire soils at these sites, being replaced mainly by microfungi. In contrast, no decrease in ECM-like ITS clones was observed at the Ridgecrop site following fire. The data indicate that, while the abundance of extraradical ECM fungal propagules in the upper soil profile may be reduced by prescribed burning, the effect is site-specific.

Many species formerly listed in Entyloma have been removed and are now placed in several orders. This study aims to clarify the framework of the genus Entyloma and the phylogeny of these plant parasitic smut fungi with molecular data. Analyses of LSU and ITS sequences are presented and support the monophyly of the order Entylomatales. The sequences show higher similarities within the examined species of Entyloma than within other smut families and genera, suggesting a recent radiation. Within the Entylomatales a cluster of anamorphic Tilletiopsis washingtonensis, T. lilacina and T. cremea collections is the sister group to Entyloma. The phylogenetic relationships in the genus Entyloma are a result of joint evolution with their hosts. The analyses of the sequence data show unresolved groups on Ranunculales and a well-supported group on Asteridae.

Restrictions patterns of the ITS rDNA and a partial fragment of the β-tubulin gene were shown to help identify Phaeoacremonium species associated with diseased grapevines. PCR-RFLP markers revealed a high frequency of P. parasiticum, together with P. aleophilum, in fungi isolated from stems and trunks of mature diseased vines showing ‘hoja de malvón’ disease, and from young declining plants in Argentina. This distribution is different from that in vineyards in other parts of the world, where P. aleophilum is dominant. Indeed, this is the first report associating P. parasiticum with diseased vines.

Phylogenetic analyses of nucleotide data from the ITS and 5.8S rDNA gene region were performed on 37 strains of Onygenales representing all the accepted species of Auxarthron (9 strains), several species of Amauroascus (16 strains), and morphologically similar genera (3 strains of Amaurascopsis, 1 of Apinisia, 1 of Shanorella, 1 of Mallochia, 2 of Nannizziopsis, 1 of Pectinotrichum, and 3 of Uncinocarpus). The inferred phylogeny provided strong support for the monophyly of Auxarthron. We propose the new combinations Auxarthron kuehnii and A. pseudoreticulatum combs nov. and provide a key to the accepted species of Auxarthron. Although this study demonstrated that Amauroascus is polyphyletic, relationships among the members of that genus could not be resolved in phylogenies inferred from the analyses of ITS sequences; the new combination A. queenslandicus comb. nov. is also proposed. Molecular data suggested that Amaurascopsis reticulatus, characterised by the absence of discrete ascomata and prominently echinulate ascospores, can be accomodated in Mallochia.

Mycotoxins are secondary metabolites produced by many phytopathogenic and food spoilage fungi including Penicillium, Fusarium and Aspergillus. The toxicity and carcinogenicity of many of these mycotoxins, and their potential to contaminate foods and animal feedstuffs is a cause of serious concern globally, both from a food safety and food trade standpoint. Thus the rapid identification of mycotoxigenic fungi would be desirable, such that early intervention steps could be applied to help limit the amounts of contaminated materials, particularly cereals and cereal-based products, gaining access to the human food chain. With this in mind a number of PCR-based methodologies have been developed for the identification of mycotoxin biosynthetic genes in different fungal genera, together with assays developed using other genes or random amplification of polymorphic DNA (RAPD) methodologies for the identification of specific toxigenic fungi. In addition, reverse transcription (RT)-PCR, competitive PCR and Real Time quantitative PCR methodologies have also been developed for this purpose. The development of each of these techniques, their usefulness, limitations and adaptability will be discussed together with descriptions of specific examples where these techniques have been utilised in different experimental settings.

Differences in morphology, ITS, IGS, mtSSU and β-tubulin sequences and UP–PCR hybridization were compared between morphologically identified F. avenaceum, F. arthrosporioides, F. anguioides, F. tricinctum, F. graminum and F. acuminatum strains. According to the combined β-tubulin, IGS and ITS tree, the strains of the Fusarium avenaceum/F. arthrosporioides/F. tricinctum species complex species can be divided into seven clusters supported by bootstrap values higher than 50%. The two main groups of European F. avenaceum, which cannot be distinguished by morphology, were separated in the tree based on β-tubulin sequences and less clearly in trees based on IGS and ITS sequences. MtSSU sequences were identical in all F. avenaceum and F. tricinctum strains studied. The European F. avenaceum strains of main group II had identical β-tubulin sequences with one American F. avenaceum strain and four European F. arthrosporioides strains, while F. avenaceum strains of main group I were closely related to two European F. arthrosporioides strains and to one Japanese F. anguioides strain. According to the combined β-tubulin/IGS/ITS sequence tree, European F. arthrosporioides strains were divided into four groups; F. tricinctum strains formed a well-supported cluster, in which two European clusters were separated from one African isolate. In the IGS sequence tree two European F. acuminatum strains together with one American F. acuminatum strain formed a cluster, which was separate from another American F. acuminatum strain. The F. acuminatum cluster was nested within the large F. tricinctum cluster together with one F. reticulatum strain in the combined IGS/β-tubulin tree. Several strains may be intermediates between the F. avenaceum/F. arthrosporioides/F. anguioides and F. tricinctum clusters and represent their own species. These results are partially supported by the results of UP–PCR hybridization analysis. Thus the molecular results may be helpful in future revision in the taxonomy of these species.

Genetic relationships among the violet root rot fungi were studied by hyphal anastomosis and sequencing of the ITS-5·8S rDNA regions. Twelve isolates of Helicobasidium mompa, one isolate each of H. purpreum and H. compactum, two isolates of Helicobasidium sp., eight isolates of Rhizoctonia crocorum, and three isolates of R. violacea anastomosed in most combinations, resulting in the death of fused cells and adjacent cells. Phylogenetic analyses of the ITS-5·8S rDNA regions separated 31 violet root rot fungal isolates into five lineages: H. mompa from Japan and Korea; R. crocorum from Europe and New Zealand; H. purpureum, R. crocorum, and R. violacea from Europe; Helicobasidium sp. from Japan; and H. compactum from the USA. Within each lineage there was no more than one nucleotide substitution, and similarity ranged from 81·8% to 98·2% between lineages. The discrepancy between the results from the two methods are discussed in terms of the biological species concept in the complex.

Fluorescent stains offer an effective means of visualising bacteria and fungi in soil or litter samples. Fluorescent brightener (FB) 28 (also known as Calcofluor White M2R) is commonly used to stain such microorganisms. However, during production of soil thin-sections we observed erratic crystallisation of this stain, particularly in soils colonised by Rhizoctonia solani. We report on the evaluation of alternative stains to FB 28 for their propensity to crystallize in interaction with fungi, staining efficiency, and suitability for application in soil thin-section production. All of the additional candidate stains namely Fluorescent Brightener Agent (FBA) 15/25, FBA 71 and SCRI Renaissance 2200 (SR 2200) were highly effective in staining agar-cultured hyphae, but differed in the degree to which they stained hyphae cultured in soil. All stains tested, except SR 2200, stained hyphae of R. solani insufficiently when growing on or through soil. These stains also showed extensive crystallisation in solutions that had been in contact with R. solani colonised soil. However, SR 2200 stained hyphae of R solani growing over soil as effectively as hyphae growing on agar and showed no evidence of crystallisation; the intensity of staining exceeded that of the ‘benchmark’ FB 28 for hyphae grown in two soil types. These excellent fluorescent properties of FBA 220 persisted in soil thin-sections, resulting in bright hyphae that could be readily visualised in situ in undisturbed soils.

The phylogenetic position of members of the entirely asexually reproducing lichen-forming genera Lepraria and Leproloma was investigated using sequence data from the ITS1-5.8S-ITS2 and small subunit (SSU) nuclear ribosomal DNA. Phylogenetic reconstructions were carried out using a likelihood-based Bayesian Markov chain Monte Carlo (MCMC) tree sampling technique and the unweighted least squares optimality criterion, the latter based on maximum likelihood distances obtained via an alignment-free distance estimation technique. The results indicate that most species currently referred to Lepraria and Leproloma form a single monophyletic group. This monophyletic group is the sister group to Stereocaulon and Muhria and belongs in the Stereocaulaceae (Lecanorales, Lecanoromycetes, Ascomycota). This indicates that an ancestor of Lepraria switched from a sexual to an asexual mode of dispersal. Subsequent speciation must have taken place in the absence of sexual processes, which contradicts the view of asexual taxa as ‘evolutionary dead ends’. Leproloma is polyphyletic and nested within Lepraria. Lepraria flavescens is a Lecanora, probably belonging in subgenus Glaucomaria. Lepraria lesdainii and L. obtusatica are unrelated to each other and to other species currently referred to Lepraria or Leproloma. Leprocaulon and Crocynia are distantly related to the core group of Lepraria and Leproloma.

Two methods are currently being used to describe ECM fungal communities associated with root tips: molecular techniques and morphological classification. Previous studies have found that these two approaches give conflicting results, with several fungal genotypes being identified from different ectomycorrhizas within the same morphotype. This has led researchers to question the usefulness of the morphological approach. The objective of this study was to compare the two approaches on ectomycorrhizas collected from three plant species growing in two different environments. Specifically, mycorrhizas were classified using a detailed morphological approach and then were subjected to PCR-RFLP analysis of the ITS region of the rRNA gene repeat. Ectomycorrhizas of Douglas-fir (Pseudotsuga menziesii) and paper birch (Betula papyrifera) were sampled from three widely dispersed sites with different soil types in the southern interior of British Columbia. Ectomycorrhizas of Douglas-fir and arbutoid mycorrhizas from Arctostaphylos uva-ursi were sampled from a fourth site in a different biogeoclimatic zone. For eight of eleven dominant morphotypes, one main RFLP banding pattern was observed. Ninety-three % of the mycorrhizas analyzed in these eight morphotypes would have been classified in the same way by either method. Five of the eight morphotypes were positively identified as Russula nigricans, Lactarius sp., Leccinum scabrum, Rhizopogon sect. Villosuli, and Thelephora terrestris by matching the RFLPs to those of fungal fruit bodies in our database or by sequencing the ITS region. The other morphotypes producing one dominant RFLP pattern were designated as Cenococcum, E-strain and Mycelium radicis atrovirens (MRA) based on their morphology. Morphotyping did not distinguish amongst major RFLP types for mycorrhizas classified as Amphinema-like, Piloderma-like and Rhizopogon-like A. We conclude that detailed morphological classification can be very useful as the primary method of ectomycorrhizal classification, when used in conjunction with molecular techniques. This approach will allow for an efficient use of research funds.