Effects of Fire on Ectomycorrhizal Fungi in Fennoscandian Boreal Forests

Fire, the primary natural disturbance factor in Fennoscandian boreal forests, is considered to have exerted major selection pressure on most boreal forest organisms. However, recent studies show that few ectomycorrhizal (EM) fungi appear to have evolved post-fi re adaptations, no succession of EM fungi following fi re is apparent after low intensity fi res, and only two EM fungal taxa exclusively fruit at post-fi re conditions. In this paper I review the present knowledge of effects of forest fi re on EM fungal communities in Fennscandian boreal forests, put into perspective by a comparison from other parts of the world. Characteristically, these boreal forests consist of less than a handful of tree species, e.g. Scots pine and Norway spruce, while the below ground communities of EM fungi is impressively species rich with presently more than 700 known taxa. Commonly, forest fi res in Fennoscandia have been of low intensity, with a considerable portion of the trees surviving and the organic humus layer partly escaping combustion. Hence, EM fungi appear to largely have evolved under conditions characterised by a more or less continuous presence of their hosts. In fact, the composition of EM fungi within a forest appear be more variable due to spatial variation than due to wildfi re. However, in areas with high intensity burns and high tree mortality, most EM fungi may locally be killed. Thus, the legacy of EM fungi following wildfi re depends on the survival of trees, which determine the potential for mycorrhizal growth, and the combustion and heating of the organic soil, which directly correlate to mortality of mycorrhizas. The questions if and to what degree fi res may be of signifi cance for yet unidentifi ed spatiotemporal dynamics of EM fungal populations and communities are discussed. Recent experiments indicate a few EM fungi are favoured by high intensity burn conditions whereas others disappear. The consequences of wildfi res in temperate conifer forests differ considerably from those in boreal forests. Wildfi res in temperate conifer forests are typically high intensity stand-replacing fi res that cause a total combustion of organic layers. Subsequently, prefi re EM fungal communities are largely eradicated and a succession of post-fi re EM fungi is initiated.


Introduction
Boreal forests constitute the largest terrestrial biome, covering 11% of the earth's terrestrial surface and forming a continuous northern circumpolar vegetation zone (Bonan and Shugart 1989). Environmental factors that determine the vegetation patterns are largely similar throughout the region. Coniferous trees prevail at all sites, typically with single or few tree species of Picea, Pinus, Larix and Abies. Boreal forests of Fennoscandia (e.g. in Sweden, Norway and Finland) are relatively homogeneous. Two conifers dominate, Scots pine (Pinus sylvestris) and Norway spruce (Picea abies), with minor contributions of birch (Betula spp), aspen (Populus tremula), alder (Alnus spp), and goat willow (Salix caprea) (Ahti et al. 1968, Essen et al. 1992. Before active fi re suppression started at the end of 1900 century, recurring fi re was a common feature in the boreal forests of Fennoscandia (Zackrisson 1977). Forest fi re is considered to be one of the major disturbance factors in this ecosystem and most forest organisms are presumed to have evolved under a strong selection pressure from wildfi res (Essen et al. 1992, Johnson 1992, Goldammer and Furyaev 1996, Engelmark 1999. Forest fi res have since long been the focal point of interest for boreal forest ecologists. However, due to earlier methodological diffi culties to monitor belowground ectomycorrhizal (EM) fungal communities, knowledge of fi re effects on EM fungi is yet scarce. Furthermore, most such studies have examined effects of prescribed burning after clear-cutting (i.e. Mikola et al. 1964, Harvey et al. 1980aand 1980b, Pilz and Perry 1984, Stendell et al. 1999, Mah et al. 2001. The common feature of these studies is a reduction in diversity and abundance of EM fungi caused by cutting rather than by fi re. However, recently effects of natural wildfi res have been studied (Visser 1995, Jonsson et al. 1999a, Baar et al. 1999, Grogan et al. 2000. Fennoscandian studies indicate that fi re effects on the EM fungal community are low, probably because most fi res are of low intensity and a considerable portion of the trees survives. (Mikola et al. 1964, Jonsson et al. 1999a. In contrast, fi res strongly affect EM fungal com-munities in temperate forests where fi res commonly are of high intensity and result in stand replacement, e.g. in Mediterranean and in temperate North America (Danielson 1984, Visser 1995, Torres and Honrubia 1997, Horton et al. 1998, Baar et al. 1999, Grogan et al. 2000 2  Dahlberg et al. 2000). The total number of fungal taxa in this ecosystem is considerably greater as about 65% of fungal taxa in Sweden are microfungi, i.e. either lacking known sexual structures or having sporocarps smaller than 1 mm (Gärdenfors 2000).

Mycorrhizal Fungi in Boreal Forest
With few exceptions, all plants in boreal forests form mycorrhizal associations and rely upon them for their fi tness and acquisition of nutrients. Grasses, ferns and most herbs form arbuscular mycorrhiza (AM), the most ancient and widespread mycorrhizal form of fungi lacking reproductive structures. The actual number of AM-forming fungal taxa in boreal Fennoscandia is not known, but low, as not more than 150 taxa are recognised worldwide. Forest fi re effects on AM fungi have not been studied in Scandinavia, but likely resemble the decrease or total loss reported from grassland ecosytems in the United States (e.g. Wicklow-Howard 1989, Amaranthus andTrappe 1993). Ericaceous plants, that characteristically monopolise and constitute the fi eld vegetation layer in boreal forests, form ericoid mycorrhiza. Worldwide, about 10 fungal taxa, predominantely ascomycetes, are known to form this association, which colonise the hair-roots. Because rhizomes and hair-roots of ericaceous plants are more superfi cially located in the organic layer of the forest fl oor than are roots of trees (cf. Persson 1983), fi re will have a more detrimental effect on communities of ericoid fungi than on those of EM fungi. However, the survival of ericoid fungi is probably considerable, as rhizomes of ericaceous plants commonly survive after fi res of low severity or spatially varying severity (Uggla 1958, Schimmel andGranström 1996). The boreal forest trees, both conifers and broadleaved, form ectomycorrhiza (EM) and the majority of these trees' root tips (>95%) are colonised by symbiotic EM fungi (Melin 1927, Smith andRead 1997). The root tips and the fungi form composite organs termed mycorrhizas, the sites of nutrient and carbon transfer between the two symbionts (Smith and Read 1997). The approximately 700 species of EM macrofungi recorded in Swedish boreal forests constitute 10-15% of all worldwide known EM fungi (Dahlberg et al. 2000, Molina et al. 1992. The most species rich genera are Cortinarius with slightly more than 200 taxa, Russula with about 80 taxa and Lacatrius with about 60 taxa.

EM Fungal Communities
Sporocarp surveys of Fennoscandian boreal forest stands reveal more than 100 fruiting EM fungal species as present (Dahlberg 1991, Ohenoja 1993. The great majority of the EM fungal biomass and fungal activity, however, is located in the forest fl oor as mycorrhizas and from the extending mycelia (Mikola el at. 1966, Smith andRead 1997). Studies of composition and dynamics of EM fungal communities have been severely hampered by the typically invisible growth of forest fungi (Dahlberg 2001, Horton and. However, the last decade's progress in identifi cation of mycobionts on single mycorrhizas has repeatedly demonstrated that fungi that do not form obvious fruiting structures, form the major mycorrhizal abundance (Gardes andBruns 1996, Horton and. Boreal EM fungal communities are impressively diverse, have high species richness, and fruiting species typically constitute less than 30% of the community of mycorrhizas (Dahlberg et al. 1997, Jonsson et al. 1999a, Jonsson et al. 2000, cf. Horton and Bruns 2001. Typically, a few EM fungal taxa account for the greatest proportion of the mycorrhizal biomass while the contribution of most taxa is small (Horton and Bruns 2001). Apparently, resource allocation to production of sporocarps versus mycorrhizas and mycelial growth varies among species (Gardes and Bruns 1996). Certain species that abundantly form mycorrhizas are poorly or non-represented Additional fungi mainly occurring 0 3 2 5 at post fi re conditions No. of red listed fungi 6 2 2 10 associated with fi re 2 1 of which six taxa fruit of wood pieces on the ground. 2 all favoured but not exclusively associated by post-fi re conditions. In total 200 of the 609 fungi red listed in Sweden, occur in boreal forests.
in the aboveground fruiting record and commonly fruiting species may conversely have only a limited representation in below-ground mycorrhizal networks. Some EM fungal taxa may be abundant both as sporocarps and mycorrhizas, e.g species of Suillus (Gardes and Bruns 1996). Other EM fungal taxa, e.g. corticoid or tomentelloid fungi, may either produce small amounts of inconspicuous sporocarps or completely lack sporocarps, despite being abundant as mycorrhizas (Erland and Taylor 1999, Kõjlag et al. 2001. The mechanisms that create this impressive diversity are yet poorly understood. Two main hypotheses are; 1) those based on niche differentiation between coexisting EM fungal species, and, paradoxically, 2) those based on similarities in competitive ability, i.e. non-equilibrium models. An example of the former is the intermediate disturbance hypothesis stating that a disturbance leads to a predictable sequence of species replacing each other. An intermediate range of disturbance would allow a range of species traits, early to late successional species, to coexist and lead to high species richness. The latter assumes stochastic elements which enhance coexistence: so called lottery models, state that due to poor dispersal ability, low local abundance and chance events, species may be absent in a neighbourhood and their abundance recruitment may thus be limited. The best competitor that happens to colonise a particular local site may well be inferior on a larger scale. Potentially, this can lead to unlimited species richness.

EM Fungal Populations
Spatiotemporal patterns and demographic processes of fungal populations are not well known, as effi cient monitoring of fungal genotypes is precluded by their growth within opaque substrates. Mycelial growth ensures the persistence and local spread of established genotypes while sexual spores secure both short and long distance dispersal and the rise of new genotypes. EM fungi do not form asexual spores, conidia, important for many saprotrophic and pathogenic fungi (Hutchison 1989).
Potentially, mycelia may have unlimited growth and age. It is merely the resource a mycelium occupies that establishes the limit of space and age of a fungal individual. Fungi that occupy discrete resource units such as wood require regular dispersal for their long-term survival. However, soil-dwelling fungi such as EM fungi, exploiting more or less indefi nite resources have the capacity for extensive mycelial growth and long survival without sexual reproduction and dispersal of spores. Hypothetically, the longevity of an individual EM fungal mycelium could be as long as their host trees are continuously present. However, the relative fi tness of an individual EM fungal mycelium is strongly infl uenced by changes in its local environment as well as by competition with other EM fungi. Circumstantial evidence suggests that the ability of established mycelia and mycorrhizas to colonise new short roots is probably much higher than that of spores to successfully establish in forest conditions, i.e. clonal traits have a higher fi tness than sexual traits in normal forest conditions (cf. Deacon and Fleming 1992). The few EM fungi that have been studied in boreal forests, (Suillus bovinus, S. variegatus and Hydnellum ferrugineum) reveal that genotypes may be both large (>300 m 2 ) and old (>150 years). Studies demonstrated the ability of these fungi to spread by mycelial growth in soil in excess of 30 m (Hintikka and Näykki 1967, Dahlberg and Stenlid 1995, Dahlberg 1997. Large genets indicate old mycelial structures that have grown from a point source over decades (Dahlberg and Stenlid 1995). The potential annual mycelial extension is probably a few decimetres in forest conditions (cf Fioro-Donno and Martin 2001). It has been demonstrated that naturally regenerated seedlings are colonised by the same set of EM fungi that are present on the surrounding trees, and probably largely from mycelia (Jonsson et al. 1999b).
Recent investigations of EM fungi from disturbed temperate forests with Hebeloma cylindrosporum and from undisturbed forest settings with Laccaria spp Amanita frenchati, Russula cremoricolor, and Lactarius xanthogalus demonstrate numerous small sized and short-lived genotypes (Gryta et al. 1997, 2000, Gherbi et al. 1999, Redecker et al. 2001, Fioro-Donno and Martin 2001. These results suggest, that spore establishment may be more important, at least for certain species in undisturbed forests, than mycelial spread. In saprotrophic and pathogenic fungi, population structures have been examined for a large number of species. Genotypes of different species are found to range in size from millimetres to kilometres with considerable heterogeneity in the size range for a single species (Anderson and Kohn 1998).

Frequency and Behaviour of Fennoscandian Forest Fires
Fennoscandian wildfi res in boreal forests are commonly of low intensity, large Scots pines usually survive and a signifi cant portion of the organic layer is left intact below the uppermost charcoal layer (Schimmel and Granström 1997). The occurrence of high temperatures within and beneath the mor-humus layer during a fi re is coupled to the depth of burn, and directly infl uences the survival of biological structures in the ground (Schimmel and Granström 1996). Detrimental temperatures normally extend less than a 2-3 cm into the mineral soil (Schimmel and Granström 1996). Among and within different fi res, the variation in depth of burn can range from most of the organic soil left intact, to combustion of almost all organic material (J. Schimmel, personal communication), affecting the local composition of EM fungi and infl uencing their spatial patchiness. Spatiotemporal variations at both the regional and local scale are important to consider when evaluating biological effects of fi res (cf Engelmark 1999). Historically, forest landscapes consisted of forests with different historic fi re frequency and varying probability of ignition. Hörnberg et al. (1998) suggested three post-fi re patterns in different Scots pine-Norway spruce forest types; a) frequently disturbed dry-mesic forests where fi res occurred every 50-100 years (Zackrisson 1977, Engelmark 1984, b) non-frequently disturbed swamp forests, with an average interval between fi res of over 400 years, and c) nonpyrogenic, mainly mesic, forests. The latter Scots pine forest type is known close to the Fennoscandian mountain range (Zackrisson et al. 1995).
At the local scale, the effects of fi re on living organisms vary both between fi re events and between different sections of the same fi re, largely owing to differences in fi re behaviour (Van Wagner 1983). The energy output in the spreading front of a forest fi re, together with properties of the tree species involved, determine the level of tree mortality caused by the fi re (Van Wagner 1983). Fires in Scandinavian boreal forests have commonly been of low intensity (Östlund et al. 1997). A substantial number of trees usually survived through several fi re events (Zackrisson 1977, Engelmark 1984, Niklasson 1998. Still, high intensity fi res can also occur and tree survival may vary considerably within the same fi re. Burn intensity is infl uenced by fuel loading, wind speed, and the moisture content of fi ne fuel particles (Schimmel and Granström 1996). Irrespective of the fi re intensity, fi res in boreal forests typically combust part of the organic soil layer. The degree of consumption of the organic soil layers, i.e. the depth of burn, depends primarily on the depth of the dry soil layer which in turn is determined by the length of time without rain before the fi re (Schimmel and Granström 1997).

Fire Dependent Organisms in Boreal Forest
Slightly more than 80 species in Swedish boreal forests are considered fi re-dependent, i.e. dependent upon fi re and the immediate post-fi re conditions to complete their life cycles and secure their long-term survival. This set of species includes more than 38 for insects, 40 for macrofungi (i.e. taxa fruiting exclusively at immediate postfi re conditions), fewer than fi ve for vascular plants, a few lichens, but no mosses (Engelmark 1999, Gärdenfors 2000, Johannesson and Dahlberg 2001. The major impact of forest fi re on organisms in boreal forests is attributed to the conditions created by fi re and the subsequent interactions of organisms with those environmental conditions. Accordingly, an estimated several hundred boreal organisms (pyrophilic or post-fi re adapted) are favoured by forest fi re rather than being strictly fi re-dependent. It is apparent that the spatial and temporal heterogeneity created by fi re among and within forests at the landscape level, combined with varying biological legacy, will affect the presence and abundance of the succeeding forest organisms. Fire-dependent macrofungi in Sweden are predominantly saprotrophs; 32 occur on soil or litter and two occur on wood. Five additional taxa fruit almost exclusively at post-fi re conditions, of these three are saprotrophs, one is the EM basidiomycete Hebeloma antracpohilum and one is the putative EM ascomycete fungus Geopyxis carbonaria (Vrålstad et al. 1998). Hebeloma antracophilum has a more southern distribution and just reaches the boreal zone, whereas G. carbonaria is common throughout the boreal zone (Hallingbäck and Aronsson 1998). Ten of the 296 red-listed macrofungi present in the boreal parts of Sweden are associated with post-fi re conditions (Table 1). None is strictly associated with burnings.

Impacts of Fire on Fungi
Forest fi res cause radical changes in the community of fruiting saprotrophic fungi (Moser 1949, Petersen 1970and 1971, Holm 1995, Rahko 1997. The destructive nature of fi re immediately effects wood inhabiting fungi by decreasing species richness and altering species composition (Penttilä and Kotiranta 1996). Several studies report a succession of ground-dwelling post-fi re fungi, in particular the early successional and abundant fruiting of discomycetes within the order Pezizales (Moser 1949, Petersen 1970and 1971, Holm 1995, Rahko 1997. EM fungi are affected by factors infl uencing the photosynthesis of the trees. For example, changes in stand biomass, successional age, or disturbances such as forest fi re will quantitatively effect the abundance of ectomycorrhizas and the production of sporocarps. Such quantitative and qualitative fi re effects on EM fungi are summarised and discussed below.

Effects on Sporocarp Production of EM Fungi
Severe fi res in coniferous forests with high tree mortality cause a drastic or total disappearance of fruiting of EM fungi (Petersen 1971, Holm 1995. When fi re intensity is low, only the upper portion of humus is consumed, trees survive, and the number, biomass and species composition of fruiting EM fungi remains largely unchanged (Rahko 1997, Jonsson et al. 1999a, Dahlberg et al. 2001. In a 1-50-year-old chronosequence of four burned stands paired with adjacent old forests in Northern Sweden, differences in the occurrence of fruiting species were attributed more to natural spatial variation than to fi re (Jonsson et al. 1999a, personal observations). In one of the stands, a low intensity underburn occurred beneath the mature forest canopy in an old-growth forest with low tree mortality. In this stand, the number of fruiting EM fungal taxa the following two years was 26, compared to 41 in the nearby old stand (Linder et al. 1999, Jonsson et al. 1999a. This difference in fruiting may be due to spatial variation as well as to: a) reduced allocation of photosynthate to the forest soil resulting from fi re damaged trees, b) reduced abundance of mycorrhizas, but not necessarily extermination of EM fungal taxa, and c) hampered fruiting due to altered environmental conditions in the organic layer, i.e. presence of charcoal and absence of bottom and fi eld vegetation. In another comparison between underburned and nonburned forests, sporocarp production was strongly affected, yet species richness of fruiting EM taxa was similar and minor differences in composition could not be attributed to the fi re . Recurrent fi re reduces the organic layer and fi eld observations suggest a thin humus layer as a prerequisite for fruiting of certain Cortinarius taxa (Phlegmacium) (Tor Erik Brandrud, personal communition). Sporocarp studies after fi res in boreal Canada suggest low survival of EM fungi (Danielson 1984, Visser 1995, while a study after low-intensity fi re in Virginia, USA, showed fruiting of most EM fungi was unaffected (Palmer et al. 1994).

Effects on Mycorrhizal Abundance
Fire causes a signifi cant loss of EM biomass in the upper litter and organic soil horizons. The mycorrhizal abundance in the organic layer is generally strongly reduced (Jonsson et al. 1999a). However, Mikola et al. (1964) reported that EM fungi protected by the soil-layer from the detrimental penetration of heat, survive and may thus link the EM fungal community between fi re affected forest generations. The mortality of mycorrhizas correlates with severity of the burn and tree survivial (Fig. 1, Dahlberg et al. 2001). However, irrespective of fi re severity, live mycorrhizas were present in the mineral layer . A vertical stratifi cation of different EM fungal taxa has been briefl y described, but is not yet well studied, e.g. that some taxa are more prolifi c in the organic layer and some taxa are more prolifi c in the mineral soil (Visser 1995, Stendell et al. 1999, Taylor and Bruns 1999. In dry forests, mycorrhizas are located deeper in the mineral soil than in mesic forests. Mycorrhizas are probably present as deep as tree roots are, i.e. several meters (Kalela 1949). Whereas mycorhizal biomass was reduced eight-fold in the litter-and organic layer in a prescribed burning of a ponderosa pine forest in USA, it was not signifi cantly reduced in the mineral layer (Stendell et al. 1999). Even after a stand replacing wildfi re on the Californian coast, USA, the fi ve most abundant EM fungal pioneers colonising the emerging coniferous seedlings appeared to have survived the fi re primarily as resistant propagules (spores, sclerotia or mycorrhizal root tips) in the mineral soil rather than as re-colonising spores (Baar et al. 1999).

Effects on Composition of EM Fungal Species
General successional patterns of EM fungal communities in relation to successions of fi re history are not apparent in Fennoscandian boreal forests. Certain EM fungal species occur exclusively in conditions associated with old forests, others are predominant after specifi c disturbance events, but most EM fungal taxa appear at almost any forest succesional age (cf. Hallingbäck and Aronsson 1998, Gärdenfors 2000, Dahlberg 2001). Spatial variation in species composition seems to be more prominent than the effects of fi re, supporting the hypothesis of signifi cant survival of EM fungi in boreal forests disturbed by wildfi re (Mikola et al. 1964, Jonsson et al. 1999a). The natural shift in composition of trees during forest successions will, however, affect the EM fungal community as some EM fungi differ in the preference and specifi city for host trees (Molina et al. 1992, Dahlberg et al. 2000. Soil composition and litter quality subsequently infl uence the properties of the organic layer and may also affect EM fungi (cf Wardle et al. 1998). For example, after a forest fi re no immediate effects on the frequency or distribution of EM fungal taxa were detected, despite reduction in mycorrhizal abundance . These fi ndings are likely due to high spatial variation, as only a few taxa were widespread and the majority were recorded only once ).
Island size is frequently a major determinant of the species composition of biological communities. A study of fi fty islands of varying size in an archipelago in the northern Swedish boreal forest zone, in which larger islands naturally burn more frequently than smaller ones through wildfi re arising from lightning strike, revealed a signifi cant relationship between island area and -Lower EM fungal legacy -Higher potential for colonisation from spores -High EM fungal legacy -Lower potential for colonisation from spores Fire severity -consumption of the organic layer Fire intensity -tree mortality Fig. 1. Schematic relationships between fi re intensity, fi re severity and survival of the mycorrhizal community.
We have conducted an analysis of the EM fungal communities on these islands, and our preliminary results demonstrate high species richness, high heterogeneity within and among islands, but no apparent taxa or set of EM fungal taxa occurring exclusively in early or late forest succession (Anders Dahlberg, personal observations). Wildfi re produces a charcoal layer, which in forests in northern Sweden may range from 1000-2000 kg ha -1 (Zackrisson et al. 1996). These amounts of charcoal appear to be suffi cient to have an important ecological effect through adsorption (Zackrisson et al. 1996, Pietikainen et al. 2000. For example, charcoal effi ciently adsorbs phenolic compounds produced by the late successional dwarf shrub Empetrum hermaphroditum. Phytotoxic effects, produced by this shrub, negatively infl uences the composition of boreal forest plants and may also affect EM fungal composition (Wardle et al. 1998). More subtle indirect effects from ash leaching have also been suggested to affect mycorrhizal fungi via pH and nutritional shifts (Stendell et al. 1999), but a controlled study in United States suggested composition and distribution EM fungal inoccula to be the primary determinants of EM community structure (Grogan et al. 2000).
In contrast, Danielson (1984) and Visser (1995) showed wildfi re in jack pine (Pinus banksiana) stands in Canada to have large impacts on EM fungal communities by both reducing the mycorrhizal abundance and changing species composition. The high intensity, stand-replacing fi res in these studies strongly altered the ground vegetation and humus conditions. The foliage of jack pine is highly combustible, and in dense forests intense crown fi res often result in tree mortality (Engelmark 1999).

Effects on Fungal Population Dynamics
Presently, we know little about necessary conditions for colonisation by EM fungi or of the frequency of new colonisation in secondary forest conditions and even less to what extent fi re alters these processes in boreal forests. It has been suggested that fungal genotypes of S. variegatus have the potential to survive fi res and to persist between forest generations, although a severe fi re with high tree mortality may be fatal to most genets (Dahlberg 1997). It has also been speculated that certain post-fi re conditions, particularly following deep-burning fi res that result in high tree mortality as well as mortality of much of the existing EM community, may provide recruitment for colonisation of new genets . However, a recent study fails to support that post-fi re conditions affect the establishment or spatial population structure of S. variegatus (Dahlberg, personal observations). Furthermore, a recent experiment demonstrates both mycorrhizal abundance and fruiting of Lactarius mammosum to be strongly enhanced after severe burns. Such conditions are clearly disadvantageous for Piloderma reticulatum, the EM fungus dominating mycorrhizas in Swedish Scots pine forests (Anders Dahlberg, personal observations).
Spore germination of certain saprophytic and pathogenic post-fi re fungi is stimulated by heat (El-Abyad andWebster 1968, Jalaluddin 1967). This response is not known from EM fungi, but addition of charcoal has been demonstrated to improve germination by removing inhibitory compounds for some EM fungi (i.e. Fries and Sun 1992). However, spores from the post-fi re discomycete G. carbonaria germinate without any external stimuli. The life cycle of G. carbonaria has been suggested to include a biotrophic association, possibly EM with Norway spruce, and where sexual reproduction and dispersal occur only at immediate post-fi re conditions (Vrålstad et al. 1998). Similarly, the wood inhabiting ascomycete Daldinia loculata exclusively fruits in post-fi re conditions on dead trees, but latent mycelia are apparently present in living birches between fi re events .
from the World Wildlife Foundation of Sweden (WWF) and the Swedish Research council for Forestry and Agricultural Research.