Interactions between mycelia of the ectomycorrhizal fungus Suillus variegatus and the wood decomposing fungus Hypholoma fasciculare were studied in soil microcosms. The ectomycorrhizal mycelium extended from Pinus sylvestris seedlings and the saprotrophic fungus grew out into the soil from birch wood blocks of two alternative sizes. Transfer of 32P between the interacting mycelia was measured non-destructively using electronic autoradiography. The outcome of the interactions was clearly affected by the size of the wood blocks. In systems with large wood blocks (1.6 cm3) the wood decomposer fungus overgrew the mycelium of the mycorrhizal fungus whereas in systems with small wood blocks (0.44 cm3) the mycorrhizal mycelium overgrew the wood decomposer fungus. The saprotrophic fungus was able to capture significantly more 32P from labelled mycorrhizal mycelium when growing out from larger wood blocks. The mycorrhizal fungus was able to capture significantly more 32P from labelled saprotrophic mycelium when the wood blocks were smaller.
In many ecosystems, especially those dominated by coniferous trees and ericoid shrubs, little easily available nitrogen and phosphorus is found in the soil. Organisms that forage for nutrients in the soil are likely to compete with each other for these scarce resources [1–2]. In the boreal forest, basidiomycetous fungi play important roles, both as degraders of organic material and as symbionts forming ectomycorrhizal associations with trees. Although saprotrophic fungi and ectomycorrhizal fungi have traditionally been treated in separate functional contexts, there is no evolutionary support for this strict separation , and fungi from the two groups may be more similar than earlier thought. Saprotrophic soil fungi usually import nutrients into cellulose rich resource units [4–6], and ectomycorrhizal fungi supply both themselves and their host trees with nutrients from the soil . This suggests that saprotrophic and mycorrhizal fungi compete with each other for soil nutrients, as first suggested by Gadgil and Gadgil . In the typical boreal forest, inorganic forms of nutrients occur only in low concentrations . However, saprotrophic fungi as well as many ectomycorrhizal fungi  can degrade organic macromolecules to assimilable forms. Thus, ectomycorrhizal fungi can compete with saprotrophic fungi, not only for inorganic nutrients, but also for organic nutrient sources. Fungal mycelium in itself constitutes a significant pool of organic nutrients. Up to 20% of the total amounts of nitrogen and phosphorus in a boreal forest soil has been estimated to be incorporated into dead and active fungal mycelium . Nutrients in fungal mycelium are usually not complexed with recalcitrant polyphenolic compounds or protected from degradation by thick lignocellulose cell walls and are therefore likely to be an attractive nutrient source for both saprotrophic and mycorrhizal fungi.
Competition between a saprotrophic and a mycorrhizal fungus for nutrients in fungal mycelium has been studied in soil microcosms by Lindahl et al. . The cord forming saprotroph Hypholoma fasciculare, extending into soil from wood blocks, was confronted with an ectomycorrhizal fungus (Suillus variegatus or Paxillus involutus), growing from pine seedling roots. The mycelium of one of the interacting fungi was labelled with 32P and the ability of the other fungus to capture phosphorus from the mycelium was tested. In that experiment, transfer of 32P mainly occurred from labelled saprotrophic mycelium to the mycorrhizal plants. The transfer of phosphorus from labelled mycorrhizal mycelium to the wood block was much slower, indicating that the mycorrhizal fungi, under the conditions of this particular experiment, were superior competitors for phosphorus. In another microcosm study of interacting fungi , where 14C fluxes from pulse-labelled, ectomycorrhizal seedlings to their associated mycorrhizal fungus (Suillus bovinus) were monitored, interaction with a wood decomposing fungus (Phanerochaete velutina) inhibited carbon allocation to the mycorrhizal mycelia. As yet, there are few other studies which have attempted to investigate competitive interactions between saprotrophic and mycorrhizal fungi and the factors determining the outcome of mycelial interactions are still largely unexamined.
In a study of combative interactions between wood rotting fungi, wood blocks were inoculated with different fungi and combined pair-wise . In these experiments, the size of the wood blocks was a crucial factor in determining the outcome of the interactions. Here we describe radioactive labelling experiments using electronic autoradiography and interaction microcosms similar to those used by Lindahl et al. . The aim was to test whether the outcome of the fungal interactions and the net fluxes of 32P between the fungi change in favour of the saprotrophic fungus, when the resources available to the saprotrophic fungus are increased.
2. Materials and methods
2.1 Design of microcosms
Elongated soil microcosms  were prepared: forest floor material was collected from a mixed coniferous forest (Lunsen, Uppsala, Sweden) and stored at 4°C until used. In plastic trays (2×20×1 cm), 1-mm-thick strips of Perspex were covered with 2.5 g dry weight of sieved (2-mm mesh) forest floor material (the fresh/dry weight ratio was determined on separate samples), which was slightly compressed to a thickness of 3 mm. De-ionised water was added to a moisture level of 280%.
Wood blocks of two different sizes were prepared from dry birch trunk wood. The large wood blocks were rectangular with 40×16-mm sides and the small blocks were circular with a diameter of 15 mm. The blocks were 2.5 mm thick regardless of size. Thus, the volume of a large wood block was 3.6 times larger than that of a small wood block (1.6 cm3 compared to 0.44 cm3). The wood blocks were autoclaved twice (121°C, 45 min) and placed on Petri dishes with Hagem agar  which were pre-inoculated with the cord forming white rot fungus H. fasciculare (Fr.) Kumm. (strain LH16 isolated in Lund, Sweden). The wood blocks were organised with four large blocks or 12 small blocks in each dish. After 20 days the wood blocks were fully covered by mycelium and transferred to the soil microcosms. At one end of the microcosms, soil was removed and replaced with the inoculated blocks. The large blocks were placed at the very end of the trays and the small blocks were placed about 3 cm from the end of the trays so that the distance between the edge of the wood blocks and the opposite end of the trays was the same in all systems.
Seven days after the wood blocks were introduced into the microcosms, pine seedlings (Pinus sylvestris L.), inoculated with the ectomycorrhizal fungus S. variegatus (Fr.) O. Kuntze (strain UP78 isolated in Uppsala, Sweden), were transferred to the systems. The seedlings were inoculated using the Petri dish method of Duddridge  with some modifications; the ratio of peat:vermiculite:modified Melin-Norkrans (MMN) in the medium was 1:4:2 and MMN with 1.25 g l−1 glucose and the concentrations of mineral nutrients reduced to 1/10 was used. When the seedlings were 13 weeks old they were transferred to the microcosms. The inoculated roots were carefully washed free of vermiculite in de-ionised water, non-mycorrhizal long roots were cut off and the root systems were folded to occupy 5 cm at the end of the soil trays opposite to the wood blocks, leaving the shoots protruding through notches in the short walls. To facilitate 32P labelling of the mycorrhizal mycelium, small plastic cups were placed under the roots of some of the plants. The systems were incubated horizontally in propagators at 17°C with a 16 h photoperiod and photon fluence rate of 300 μmol m−2 s−1. Water loss through desiccation was compensated for gravimetrically by the addition of de-ionised water twice every week with a Pasteur pipette. Direct wetting of the hydrophobic mycelial structures was avoided to prevent damage to the mycelium.
2.2 Labelling with 32P
At the time of first physical contact between the mycelia of the two fungi, one of the fungi was supplied with 32P. This occurred at different times due to differences in growth rate between different experimental units. In half of the systems the mycorrhizal fungus was supplied with 19 kBq carrier-free 32P as orthophosphate by placing droplets of water (10 μl) containing the radioactive tracer in the mycelium-covered cups under the roots. In the rest of the systems the saprotrophic fungus was supplied with the same amount of radioactive tracer in 10-μl water droplets that were placed on top of the mycelium-covered wood blocks.
The distribution of radioactivity in the microcosms was measured non-destructively at 5–6-day intervals using a Packard Instant Imager electronic autoradiography system (Packard Instrument Co., Meriden, CT, USA, see  for details). To acquire an autoradiographic image, radioactive disintegration events were counted for 4–7 min with a spatial resolution of 0.5 mm. The experiments were interrupted before any mycelium of the non-labelled fungus had reached the sites of 32P addition (the wood blocks or the plastic cups).
2.4 Calibration of counting efficiency for 32P in wood blocks
Microcosms were prepared with small or large wood blocks but no plants. After 5 weeks of incubation, 0.5 ml of 0.1 M KH2PO4 solution containing 0.5 μCi of 32P was supplied to the growing mycelial fronts. After an uptake period of 1 week, the radioactivity in the wood blocks was measured using electronic autoradiography. Immediately after autoradiographic scanning, the wood blocks were removed from the soil and digested in 5 ml of a 1:1 mixture of nitric acid (68%) and perchloric acid (70%). Large wood blocks were divided into three separate samples before they were digested. The samples were bleached by addition of 5 ml hydrogen peroxide (35%) and the radioactivity was counted in a Packard TriCarb 2100 TR scintillation counter (Packard Instrument Co., Meriden, CT, USA) using Cerenkov counting, correcting for colour quenching and decay.
2.5 Analysis of autoradiographic images
In each of the acquired autoradiographic images, three areas were defined over which the counts per minute (cpm) were integrated. Data were recorded separately for plants including mycorrhizal roots, wood blocks and the mycelial interaction zone between the plant roots and the wood blocks. The measured cpm data were converted to disintegrations per minute (dpm) using counting efficiencies of 0.25 for activity in the soil and 0.24 for activity in the plants . For the wood blocks, counting efficiencies were obtained from the separate calibration experiment. All radioactivity measurements were corrected for background activity and isotope decay.
The dynamics of isotope transfer between the fungi were analysed using a mathematical regression model described in detail by Lindahl et al. . First the negative exponential functions were found that best described the translocation of activity out from the microcosm component (wood block or root zone) to which the radioisotope was added. For each individual microcosm non-linear regression was used to find values for the constants a and b in the equation
where t is time and AO is the activity (dpm) outside the microcosm component to which the radioisotope was added (activity in the interaction zone+plant in systems where 32P was added to the wood block and activity in the interaction zone+wood block in systems where 32P was added to the plant roots). The constant a represents the level of activity that AO asymptotically approaches. The constant b represents the rate with which AO approaches a. The constants a and b were used in the equation
where AR is the activity in the ‘receiving’ microcosm component (the ‘receiving’ microcosm component is the plant including mycorrhizal root tips, in systems where 32P was added to the wood block, and the wood block in systems where 32P was added to the plant roots). Non-linear regression was used to find values for c, which represents the fraction of the activity in the interaction zone that is captured by the receiving fungus per unit time. This value was assumed to be constant during the experiment. Possible differences in transfer rates between microcosms with large wood blocks and microcosms with small wood blocks were statistically tested using analysis of variance (ANOVA). Two ANOVA tests were performed; one of the 32P transfer rate from labelled saprotrophic mycelium to plants and one of the transfer rate from labelled mycorrhizal mycelium to wood blocks. The null hypothesis was that there was no difference in average transfer rate (the constant c in Eq. 2) between systems with small and large wood blocks.
3.1 Morphological responses
The first systems were labelled 22 days after the plants were introduced into the systems and the last systems were labelled after another 17 days. Generally, Hypholoma mycelia growing from large wood blocks had higher extension rates than mycelia growing from small wood blocks. Thus, the meeting of the fungi and the addition of 32P occurred earlier in systems with large wood blocks (at an average of 24 days after the plants were introduced) than in systems with small wood blocks (at an average of 32 days after the plants were introduced). In addition to the higher extension rate, mycelia extending from the larger wood blocks grew more vigorously than mycelia extending from small wood blocks.
The first morphological response to physical contact between the fungi was the production of bands of very dense mycelium by H. fasciculare in the area of hyphal interaction between the fungi (Fig. 1b). This band was most pronounced in systems with large wood blocks and in some of these systems the mycelial growth associated with confrontation was so dense that the mycelium almost formed a continuous tissue. H. fasciculare also showed this kind of confrontation response when extending from smaller wood blocks, but not as intensely and not in all of the replicates.
In 14 of the 16 systems with large wood blocks, the bands of dense mycelium expanded and advanced through soil occupied by Suillus mycelium towards the plant roots to form patches of dense Hypholoma mycelium (Fig. 1c). In half of the systems where Hypholoma formed patches, the development of mycorrhizal mycelium was very poor after meeting the saprotrophic mycelium and there was little visible mycorrhizal mycelium in the interaction zone at the end of the experiment. In the rest of the systems, growing mycorrhizal mycelium could be observed within the Hypholoma patches. In two of the systems with large wood blocks, no dense growth of Hypholoma mycelium could be observed and in one of those systems, Suillus formed a dense patch over the saprotrophic mycelium.
In seven of the 16 systems with small wood blocks the Suillus mycelium overgrew the saprotrophic mycelium and formed dense patches over the Hypholoma mycelial fronts (Fig. 2b). The Hypholoma patches characteristic for the systems with large wood blocks were found in five of the systems with small wood blocks. The growth of saprotrophic mycelium in these patches was less vigorous than in most of the systems with large wood blocks and growth of mycorrhizal mycelium could always be observed within the patches. In two of the systems with Hypholoma patches, the Suillus mycelium managed to grow through the Hypholoma patch and form a patch behind the mycelial front of the saprotroph (Fig. 2c). In four of the systems with small wood blocks, neither of the fungi produced dense mycelial patches (Fig. 2d).
3.2 Counting efficiency
The counting efficiency of the electronic autoradiography system was estimated to 0.095 for 32P in large wood blocks and 0.083 for 32P in small wood blocks.
In the systems with labelled Suillus mycelium, the amounts of activity in the extraradical mycorrhizal mycelium were highly variable. No significant differences in translocation rate of 32P from the labelling cup to the interaction zone could be found between systems with small or large wood blocks (Fig. 3a). Clear transfer of 32P from labelled mycorrhizal mycelium to the wood blocks was found in 14 of the 16 experimental units. The transfer rate of 32P to the wood blocks was higher in systems with large wood blocks than in systems with small wood blocks (Fig. 3b). Expressed in relation to the amounts of 32P outside the root zone (the 32P available to the Hypholoma), on average, four times as much 32P was found in large wood blocks as in small wood blocks at the end of the experiment (Fig. 3c). Values of the model parameters a and b were found that produced good approximations of radioisotope translocation to the extraradical mycelium (Fig. 3a). The model approximations of the transfer of 32P to the wood blocks tended to overestimate transfer rates early in the experiment and underestimate transfer rates at the end of the experiment (Fig. 3b). The model parameter c (the fraction of 32P in the interaction zone that was transferred to the wood blocks per unit time) was, on average, more than twice as high in systems with large wood blocks compared to systems with small wood blocks (Table 1). When tested with ANOVA the estimated differences in the model parameter c were found to be significant (Table 2a).
Table 1. Frequency of mycelial patches and 32P transfer rates between interacting mycelia of Suillus variegatus and Hypholoma fasciculare in soil microcosms
Number of systems with mycelial patches (out of 16)
Average 32P transfer (percent of activity in interaction zone per day)
From labelledSuillus mycelium to wood blocks
From labelled Hypholoma mycelium to plants
All values are means±SE
a In two of these systems, the Suillus patches were developed only after the mycorrhizal fungus had grown trough the front of the Hypholoma, which also formed patches (Fig. 2c).
Large wood blocks
Small wood blocks
Table 2a. Frequency of mycelial patches and 32P transfer rates between interacting mycelia of Suillus variegatus and Hypholoma fasciculare in soil microcosms
Wood block size
In systems with labelled Hypholoma mycelium, the initial translocation rate of 32P from the labelled mycelium in the wood blocks to the interaction zone was significantly higher in systems with large wood blocks than in systems with small wood blocks (Fig. 4a). Clear transfer of 32P to the mycorrhizal root tips was found in 15 of the 16 experimental units. The transfer rate of 32P to the mycorrhizal root tips was similar in all systems (Fig. 4b). However, when expressed in relation to the amounts of 32P outside the wood blocks (the 32P available to the Suillus), on average almost twice as much 32P was found in plants of systems with small wood blocks compared to plants of systems with large wood blocks at the end of the experiment (Fig. 4c). Values of the model parameters a and b were found that produced good approximations of radioisotope translocation out of the wood blocks (Fig. 4a). As in the case of 32P transfer to the wood blocks, the model approximations of 32P transfer to the plants tended to overestimate transfer rates early in the experiment and underestimate transfer rates at the end of the experiment (Fig. 4b). The model parameter c (the fraction of 32P in the interaction zone that was transferred to the plants per unit time) was, on average, more than twice as high in systems with small wood blocks compared to systems with large wood blocks (Table 1). When tested with ANOVA the estimated differences in the model parameter c were found to be significant (Table 2b).
Table 2b. ANOVA of 32P transfer rate from labelled Hypholoma mycelium to plants
Wood block size
The interacting soil mycelia displayed a range of different morphological interaction responses. The responses ranged from dominance of H. fasciculare, where the growth of the soil mycelium of the saprotroph was stimulated by the interaction, through intermediate forms, to dominance of S. variegatus, where the mycelium of the mycorrhizal fungus formed a dense patch over the saprotrophic mycelium. When growing on agar, H. fasciculare responds to interaction with other fungi by the formation of dense mycelial structures similar to those observed in the above described experiments . The ability of Suillus species to form dense mycelial patches in response to enrichment of the substrate with organic nutrients has been described by Bending and Read . In the present study there was a clear correlation between the size of the wood blocks and the morphological outcome of the interactions. Mycelium of H. fasciculare usually dominated in microcosms with large wood blocks, whereas in systems with small wood blocks S. variegatus often seemed to outcompete the saprotroph. This observation agrees with the finding of Holmer and Stenlid  that the combative strength of wood decaying fungi is dependent on the volume of wood that they occupy.
The outcome of competitive events involving fungi is determined by the relative inoculum potential of the competitors. Inoculum potential ‘sensu Garrett’ is defined as ‘the energy of growth of a fungus available for colonisation of a substrate at the surface of the substrate to be colonised’. This concept includes both the biomass of the fungi, the concentration of resources in the biomass and the availability of resources at the interaction site. In the case of rhizomorphic fungi with high capacity for resource translocation, the term inoculum potential must also include the resource content and resource availability of the whole mycelium as well as the resistance to translocation. The wood block size could potentially influence all of these parameters. In the here described experiments, differences in inoculum potential were not necessarily a direct effect of the wood block size, but could also have been influenced by the inoculation method. A significant fraction of the resources in the saprotrophic mycelium probably originated from the agar on which the wood was inoculated. Resources, taken up from the agar, were distributed among four units in the case of large wood blocks, but among 12 units in the case of small wood blocks. The differences in initial resource content could influence all the parameters determining the inoculum potential at the interaction site. It is not possible to conclude which particular resources caused the higher competitivity of the mycelia growing from the larger wood blocks. Growth rate and decay activity of wood decomposing fungi are, however, suggested to be tightly correlated with the amount of nitrogen available to the fungi [20,21].
Transfer of phosphorus between the interacting mycelia occurred in both directions. The size of the 32P fluxes was dependent on the size of the wood blocks; the saprotrophic fungus captured more and lost less 32P when supplied with a larger resource unit. The average transfer rate from Hypholoma mycelium to the plants in systems with small wood blocks, estimated to 0.3% of the activity in the interaction zone per day, was lower than the average transfer rate of 0.8% estimated in similar systems in previous experiments . The transfer rate of 32P from labelled Suillus mycelium to small wood blocks was higher in the present experiments than in these earlier experiments. As the same strain of H. fasciculare was used in all of the experiments, suggesting that the Suillus strain used by Lindahl et al.  was more competitive than the strain used in the experiments described in this paper.
There are several problems associated with the quantification of the phosphorus transfer between the mycelia. Since the total amount of phosphorus in the mycelium is unknown, the specific activity (the ratio of 32P to non-radioactive P) is also unknown. Thus, no conclusions can be drawn about the transfer rates of non-radioactive phosphorus and the direction of net phosphorus translocation between the mycelia. The data on 32P transfer to the receiving component (Figs. 3b and 4b) therefore provide little information, unless expressed in relation to the amounts of 32P available to the receiving fungus (Figs. 3c and 4c). If all phosphorus pools in the mycelium are uniformly labelled with 32P, the fraction of 32P in the interaction zone that is transferred is the same as the fraction of non-radioactive P. However, if the tracer isotope is preferentially incorporated into labile compounds, then the transferred fraction of non-radioactive P will be overestimated.
The fraction of the available tracer isotope that is found in the receiving component at a specific point in time (Figs. 3c and 4c) is a simple estimate of the phosphorus transfer, but not good for comparing different systems, since the amount of available 32P changes during the experiment (Figs. 3a and 4a). In systems where all the 32P translocation into the interaction zone occurs early in the experiment, the 32P found in the interaction zone at a specific point in time has been available to the receiving fungus for a longer time than in systems where the amounts of 32P in the interaction zone increase gradually. The regression model used in this paper compensates for these differences between systems and combines all measurements of a microcosm into one transfer rate value. The regression model assumes that the transfer rate between the fungi is constant during the experiment. This assumption is not fully correct; in most microcosms the 32P transfer seems to be an accelerating process. The model thus tends to overestimate transfer rates early in the experiment and underestimate transfer rates at the end of the experiment (Figs. 3b and 4b).
Since it was impossible to assign 32P activity in the interaction zone to either of the interacting fungi, it was necessary to consider activity to be transferred from the labelled to the receiving fungus, only when the activity could be measured in the plant or the wood block. The transfer rates are thus underestimated, as a fraction of the captured phosphorus is immobilised in the soil mycelium. It is hard to speculate about the partitioning of phosphorus between soil mycelium and plant/wood block since this probably differs between different experimental designs. In an experiment with pine seedlings colonised with S. luteus, 10–20% of the acquired phosphorus was immobilised in the extraradical mycelium . H. fasciculare, growing out into soil from wood blocks, immobilised about 10% of the acquired 32P in the soil mycelium regardless of wood block size .
The different counting efficiency of the electronic autoradiography system when measuring radioactivity in small or large wood blocks is surprising, as the wood blocks had the same thickness. The lower absorption of radiation by the large wood blocks could however be explained by a lower water content in the large wood blocks compared to the small wood blocks, which had a larger surface in contact with the moist soil.
The experiments suggest that combative fungal interactions, earlier observed in wood or on agar , also occur between mycelia growing in soil. Antagonistic interactions between ectomycorrhizal fungi have been observed by Wu et al. . The experiments described in this paper confirm previous findings  that ectomycorrhizal fungi can also be antagonistic against a saprotrophic fungus, but in addition, that this occurs only when the competitive strength of the saprotroph is restricted by a limited resource supply. When the saprotrophic fungus had access to a larger resource, the growth of the mycorrhizal mycelium seemed to be reduced by the interaction, in a manner similar to that described by Leake et al. . Competition for nutrients between soil fungi thus seems to be size-asymmetric, which means that larger individuals suppress the growth of their smaller neighbours and thereby acquire more nutrients in relation to their size than smaller individuals. This would be in contrast to competition for nutrient between plants, which is thought to be size-symmetric .
Many experiments have investigated the ability of the soil mycelium of wood decaying fungi to take up inorganic phosphate from the soil and translocate it to high quality resource units . The present study shows that the soil mycelium of a saprotrophic, cord forming fungus can compete with a mycorrhizal fungus and its host plant, not only for inorganic nutrients in the soil solution but also for nutrients already taken up by the mycorrhizal fungus. However, to compete successfully with the mycorrhizal fungus, the saprotrophic fungus is dependent on a sufficient supply of energy and nutrients.
In old forest stands, mycorrhizal genets are few and large compared to young stands, where they are many and small . Field and laboratory experiments have shown that the production of mycorrhizal fruitbodies and the respiration of mycorrhizal mycelium decline sharply, as the flux of plant assimilates to the mycelium is disrupted, showing the dependence of these fungi on current assimilates from the trees [29,30]. It is thus likely that when a forest area is clear cut, the old and well developed mycorrhizal mycelia will not persist and that newly planted seedlings will have to form symbiotic associations with young and small mycorrhizal individuals. Clear cutting sites should provide optimal conditions for saprotrophic fungi with large amounts of high quality sources of energy and nutrients present in the form of stumps, roots and slash. Highly competitive saprotrophic fungi could interfere with the development of new mycorrhizal mycelia on seedlings, reducing the chances of seedling survival. This could explain why seedling survival increases when slash is removed after clear cutting compared to when the slash is left at the clear cutting site .
In naturally regenerating boreal forests, young seedlings are often found on very decomposed stumps or logs –‘nurse logs’. Rhizomorphs and sporocarps of mycorrhizal fungi are commonly found in highly decomposed wood  and the stumps and logs provide the mycorrhizal fungi with a large volume of substrate where all high quality carbohydrates have been used up by the former activities of wood decaying fungi. Competition from saprotrophic fungi should thus be minimal, and this may contribute to the suitability of ‘nurse logs’ as substrates for young seedlings. During decomposition of logs, nutrients from the wood or imported from the surrounding soil accumulate in the mycelium of the wood rotting fungi. At advanced stages of decomposition, mycelium of dead or weak wood rotting fungi could provide the mycorrhizal fungi and their associated seedlings with a high quality nutrient substrate.
Financial support from NFR (The Swedish Natural Science Research Council), SJFR (The Swedish Council for Forestry and Agricultural Research) and the Knut and Alice Wallenberg Foundation is gratefully acknowledged.