• The mycorrhizal basidiomycetes are known to have multiple, independent evolutionary origins from saprotrophic ancestors. To date, a number of studies have revealed functional resemblance of mycorrhizal fungi to free-living saprotrophs, but information on the ability of saprotrophic fungi to perform as mycorrhizal symbionts is scarce.
• Here, the objective was to investigate the ability of three wood-decay fungi, Phlebiopsis gigantea, Phlebia centrifuga and Hypholoma fasciculare, to colonize fine roots of conifer seedlings.
• For each fungus, mycorrhizal syntheses were attempted with Picea abies and Pinus sylvestris. After 24 wk, isolation of fungi and direct sequencing of fungal internal transcribed spacer (ITS) rDNA were carried out from healthy-looking surface-sterilized root tips that yielded both pure cultures and ITS sequences of each inoculated strain. Mycelial mantle of P. gigantea was frequently formed on root tips of P. abies, and microscopical examination has shown the presence of intercellular hyphae inside the roots.
• The results provide evidence of the ability of certain wood-decay fungi to colonise fine roots of tree seedlings.
In mycorrhizal symbiosis, the fungal symbiont is located in the root and its hyphal connections with the soil ensure the absorption of soil-derived nutrients, while obtaining organic carbon from the plant (Smith & Read, 1997). By contrast, saprotrophic fungi utilize dead organic matter and tend to be physically discrete and functionally independent (Cooke & Rayner, 1984). It is known, however, that biotrophy and saprotrophy are not necessarily mutually exclusive ways of life, and there is no reason why some fungi should not be capable of both under different circumstances (Cooke & Whipps, 1993). This idea was supported by phylogenetic analyses of ectomycorrhizal and free-living saprotrophic basidiomycetes, which indicated that ectomycorrhizal fungi have evolved independently from multiple lineages of saprotrophic fungi (Hibbett et al., 2000; Bruns & Shefferson, 2004; James et al., 2006).
Although mycorrhizal fungi have been studied in relation to functional resemblance to free-living saprotrophs, information on the ability of saprotrophic fungi to behave as mycorrhizal symbionts is scarce. In the past, several fungi fruiting on dead wood (e.g. Tomentella spp., Clavulicium delectabile (Jacks.) Hjortst.) were suspected, but not demonstrated, to be saprobes (Larsen, 1968, 1974; Eriksson & Ryvarden, 1973; Hansen & Knudsen, 1997), and were recently reclassified as ectomycorrhizal (Koljalg et al., 2000; Tedersoo et al., 2003). Yet, Armillaria spp. can form an orchid type of mycorrhiza in addition to being saprotrophic or even pathogenic in woody substrates (Smith & Read, 1997).
In our other recent work, a total of 33 distinct morphotypes was observed among 30 166 root tips from 660 healthy-looking conifer seedlings in forest nurseries, and one to 22 root tips per morphotype (130 in total) were subjected to direct ITS sequencing of the fungal rDNA (Menkis et al., 2005). Direct sequencing of the material revealed the presence of 60 fungal taxa, only 27 of which (45%) were previously known as mycorrhizal. Surprisingly, four rare morphotypes (each observed just once) yielded ITS sequences which gave an almost exact match, both in our database and in the NCBI GenBank (Table 1), with the sequences of three well-known wood-decay fungi –Phlebiopsis gigantea (Fr.) Jülich, Phlebia centrifuga P.Karst., and Hypholoma fasciculare (Huds.Fr.) Kummer, which are widespread decomposers of dead conifer wood in temperate and boreal forests (Jahn, 1979; Eriksson et al., 1981; Hansen & Knudsen, 1992, 1997; Ryman & Holmasen, 1998). The objectives of the present study were to confirm the ability of the mentioned wood-decay fungi to colonize fine roots of conifer seedlings and to characterize the morphology of their mycelial structures.
Table 1. Previously unpublished data on mycorrhizal morphotypes and fungal internal transcribed spacer (ITS) rDNA sequences from four root tips in which the presence of wood-decay fungi (in bold) was revealed by the study of Menkis et al. (2005)
The database of fungal ITS rDNA sequences at the Department of Forest Mycology and Pathology, Swedish University of Agricultural Sciences, currently comprises over 3000 sequences of over 1000 species of wood-inhabiting and mycorrhizal fungi.
Isolated either from the spore print or the fruitbody of a given fungus, sequenced, and deposited in culture collection of the Department of Forest Mycology and Pathology, Swedish University of Agricultural Sciences. The identification of fungi outside our reference is based on their sequence similarity to sequences available in the GenBank database (Altschul et al., 1997).
Over 99% match also with the GenBank sequences of respective species.
The following fungal strains were used for colonization of root systems of Picea abies and Pinus sylvestris: Hypholoma fasciculare (strain RV181), Phlebia centrifuga (strain NK3), and Phlebiopsis gigantea (Rotstop®, Kemira OY, Finland). The first two strains were isolated directly from the fruitbodies of the respective species, while the third strain originated from an in vitro germinated spore suspension of P. gigantea, the Rotstop biological control agent, used to control root rot of conifers in Fennoscandia (Thor, 2003). All three strains were stored in the culture collection at the Department of Forest Mycology and Pathology, of the Swedish University of Agricultural Sciences. Before the experiment, the cultures were maintained in darkness at 20°C on modified Melin-Norkrans (MMN) medium (Marx, 1969).
Experimental microcosm systems were constructed using the mycorrhizal synthesis systems described in previous studies (Finlay, 1989; Rosling et al., 2004). Two-wk-old seedlings were aseptically inoculated with agar plugs from the fungal stock cultures in Petri dishes with growth substrate of sterilized fine sphagnum peat : vermiculate : 1/10 strength liquid MMN mixture in the ratio 1 : 4 : 2. Five replicate microcosms were constructed for each tree species/fungus combination (two tree species × three fungi × five replicates, or 30 microcosms in total). Inoculated microcosms were incubated in a growth chamber at 20°C, with a 16 h light : 8 h dark photoperiod.
After 24 wk, individual root tips and lateral roots were examined for fungal colonization. Isolation of fungi into pure culture was performed from root tips following the procedures of Menkis et al. (2005), and included one root tip from two replicates of each microcosm type, or a total of 12 root tips. Vegetative compatibility tests between inoculated and reisolated strains were carried out as described by Vasiliauskas & Stenlid (1998). The direct sequencing of fungal ITS rDNA included one root tip from all five replicates of each microcosm type, or a total of 30. The tips were surface-sterilized in 33% hydrogen peroxide for c. 30 s and rinsed three times in sterile deionized water (Danielson, 1984; Sieber, 2002). Extraction of fungal DNA from the roots, amplification and sequencing were performed exactly as by Menkis et al. (2005) and Rosling et al. (2003).
Furthermore, roots representing each microcosm system (one root per each fungus/plant system, or six in total) were sectioned to 5 µm thick on an ultratome and stained as described by Egger & Paden (1986). Sections were brought down to 70% ethanol, stained with 1% safranin O in 50% ethanol for 20 min, rinsed in water, dehydrated through two changes of 100% ethanol, stained with 1% fast green in 100% ethanol for 10 s, rinsed in 100% ethanol, then transferred through a series of four jars of a saturated solution of orange G in clove oil (2 min per jar). This procedure allowed lignified tissues to stain red and hyphae to stain green. Microscopic examination and photography were carried out at ×400 magnification using a Carl Zeiss Axioplan microscope equipped with differential interference contrast (Nomarski) optics.
All plants in all microcosms remained healthy-looking and vigorous after 24 wk, and the substrate containing their root systems was extensively colonized by the fungal mycelia (Fig. 1a). In each of the five replicates of the P. abies–P. gigantea treatment, about two-thirds of all root tips showed a fungal mantle, which could be classed as a ‘pale reddish’ or ‘yellow reddish’ morphotype (Fig. 1b), and was similar to that observed on root tips colonized by the fungus under field conditions (Table 1). By contrast, the external fungal mantle was completely absent from all five of the remaining treatments, namely P. abies with H. fasciculare and P. centrifuga, and all microcosms of P. sylvestris with all three species of fungi. Despite this, each of two attempted isolations from each of the six microcosm types in all 12 cases yielded the same fungus that had been inoculated, as their mycelia had similar morphology and were vegetatively compatible (anastomosed into a single entity when paired on a Petri dish, implying genetic similarity).
Direct ITS rDNA sequencing from five root tips (replicates) of each microcosm type yielded the sequence of the inoculated fungus in one to four occasions. In each intraspecific comparison, the sequences were 100% similar among themselves and to the sequence of the inoculated strain. Thus, out of five root tips per microcosm tested for each fungus with each of the two tree species, H. fasciculare was detected from one root tip of P. abies, and one of P. sylvestris (GenBank accession number DQ320134), P. centrifuga was detected from two root tips of P. abies, and one of P. sylvestris (GenBank accession number DQ320135), and P. gigantea was detected from two root tips of P. abies, and four of P. sylvestris (GenBank accession number DQ320133). No other fungus was detected either by sequencing or by isolation in any of the 30 microcosms checked.
Sectioning, staining and microscopic examination of roots have demonstrated the presence of P. gigantea hyphae inside P. abies roots, exhibiting a pattern of intercellular colonization of epidermal root cells (Fig. 1c). Although microscopy has also revealed certain hypha-like structures in roots taken from other five fungus–plant microcosm systems, in those cases fungal penetration inside the roots could not be clearly documented by staining.
In this study, as in numerous preceding works, pure culture synthesis techniques (axenic conditions) were used to establish if known fungal species could colonize fine roots of trees (Riffle, 1973; Danielson, 1984; Egger & Paden, 1986, Buscot & Kottke, 1990; Yamada et al., 2001). In the cited works, the observed associations ranged from mycorrhizal symbiosis to pathogenicity, while many species did not show any affinity to roots. Yet, as axenic experimental design provides a method of testing plant–fungus interactions without interference from other soil organisms and under ecological conditions that differ from those in the field, care must be taken in extrapolating these results to natural conditions. Thus, the ability of a fungus to form mycorrhizas with a host in pure culture does not prove mycorrhizal formation in nature and vice versa (Riffle, 1973; Egger & Paden, 1986; Buscot & Kottke, 1990).
Yet, in the present experiment, we have supporting evidence that the tested fungi are also able to colonize fine conifer seedling roots under xenic conditions in the field. In fact, the current results complement the findings of our earlier study, in which the colonization of healthy-looking conifer root tips by H. fasciculare, P. centrifuga and P. gigantea was observed in the bare root nursery (Menkis et al., 2005).
Consequently, the results of this work provide more evidence for the ability of wood-decay fungi to live in the soil and to colonize healthy fine roots of conifer seedlings. The most evident pattern was exhibited by P. gigantea in P. abies roots, on the tips of which the fungus consistently formed mycelial mantle (Fig. 1b) and spread inside intercellularly between epidermal cells (Fig. 1c). Despite the fact that this was not explicitly demonstrated in the remaining microcosm systems, both the isolations and direct sequencing from root tips following their intensive external sterilization indicated clearly and consistently the presence of active mycelium of each tested fungus in the internal root tissue of healthy-looking seedlings of both tree species.
In the present study, no visible negative impacts on seedling vitality were observed in our microcosms during a half-year period (Fig. 1a). By contrast, a few similar attempts at microcosm syntheses of conifer seedling roots with saprotrophic soil-inhabiting basidiomycetes Leucopaxillus giganteus (Sow.Fr.) Sing. and Collybia dryophila (Bull.Fr.) Kummer have resulted in retarded growth (Yamada et al., 2001) or death of a plant (Riffle, 1973).
The rhizosphere environment has not been previously considered as a possible ecological niche for wood-decomposing fungi (Rayner & Boddy, 1988). Therefore, to our knowledge these are the first reports on the presence of wood-decaying saprotrophs in fine, apparently healthy roots of trees and in the soil both in vitro and in vivo. This provides new insights into the ecology of wood-decay fungi, since, in the past, studies of their community ecology have been based on the occurrence of their fruitbodies on a given woody substrate, which was consequently regarded as the ultimate habitat of those fungi in nature (Jahn, 1979; Eriksson et al., 1981; Breitenbach & Kränzlin, 1986; Hansen & Knudsen, 1992, 1997; Ryvarden & Gilbertson, 1993, 1994; Ryman & Holmasen, 1998). It is known, however, that such data can provide misleading information, since the fruitbody distribution does not necessarily fully reflect mycelial distribution and activity (Rayner & Boddy, 1988). For example, our recent work has revealed a poor correspondence between fruitbody occurrence on woody debris and the relative abundance of fungal mycelia within the given substrate (Allmer et al., 2006).
The currently reported results expand our knowledge of the habitat of wood-decay saprotrophs, demonstrating the ability of some of these fungi to live in the rhizosphere and colonize fine roots of trees, resembling closely the patterns characteristic for mycorrhizal fungi. Moreover, when exposed both to dead wood and to soil, mycelia of P. centrifuga exhibited more extensive growth in the soil (Fig. 1d). Consequently, two questions arise: do the observed colonization patterns result in a functional mycorrhizal relationship; and are other species of wood-decay fungi also able to colonize fine roots of trees in a similar manner? Furthermore, P. gigantea is a large-scale biological control agent, its spore suspension being sprayed on cut stump surfaces in over 100 000 ha of forest in Europe annually (Thor, 2003). As it was demonstrated that the fungus persist in sprayed stumps for up to 7 yr, and produces abundant sporocarps there which release viable basidiospores to the environment (Vasiliauskas et al., 2005), it would be of interest to check if such an increased load of inoculum has any impact on mycorrhizal communities in treated forest stands.
The functional relationships of H. fasciculare, P. centrifuga and P. gigantea with P. abies and P. sylvestris are therefore being investigated in ongoing studies. Moreover, we are presently monitoring microcosms containing 200 species of wood-decay basidiomycetes with three species of trees (P. abies, P. sylvestris and Betula pendula Roth). Analysis of these systems will be carried out by investigating the potential for bidirectional transfer of nutrients and carbohydrates (Finlay & Read, 1986; Lindahl et al., 1999, 2001), and should provide new information on how widespread the ability to colonize healthy roots is in fungal saprotrophs. For the future, it would be of utmost importance to study the functional relations of saprobic fungi and trees in situ, open to competition between fungi. The investigations of fungi on seedling roots that are established on Rotstop-treated stumps would be of particular interest.
This work was supported by a grant from the Swedish Research Council for Environment, Agricultural Sciences and Spatial Planning (FORMAS). We thank Mrs Anette Axen for technical assistance and anonymous referee for the constructive comments and suggestions.