Ectomycorrhizal fungi: exploring the mycelial frontier


  • Editor: Ramon Diaz Orejas

Correspondence: Ian C. Anderson, The Macaulay Institute, Craigiebuckler, Aberdeen AB15 8QH, UK. Tel.: +44 1224 498200; ext. 2357; fax: +44 1224 498207; e-mail:


Ectomycorrhizal (ECM) fungi form mutualistic symbioses with many tree species and are regarded as key organisms in nutrient and carbon cycles in forest ecosystems. Our appreciation of their roles in these processes is hampered by a lack of understanding of their soil-borne mycelial systems. These mycelia represent the vegetative thalli of ECM fungi that link carbon-yielding tree roots with soil nutrients, yet we remain largely ignorant of their distribution, dynamics and activities in forest soils. In this review we consider information derived from investigations of fruiting bodies, ECM root tips and laboratory-based microcosm studies, and conclude that these provide only limited insights into soil-borne ECM mycelial communities. Recent advances in understanding soil-borne mycelia of ECM fungi have arisen from the combined use of molecular technologies and novel field experimentation. These approaches have the potential to provide unprecedented insights into the functioning of ECM mycelia at the ecosystem level, particularly in the context of land-use changes and global climate change.


Many tree species in forest habitats worldwide rely upon mutualistic ectomycorrhizal (ECM) fungi to fulfil their nutrient requirements (Smith & Read, 1997). These fungi contribute to tree nutrition by means of mineral weathering (Landeweert et al., 2001) and mobilization of nutrients from organic complexes (Read & Perez-Moreno, 2003). They are also an important avenue for the delivery of carbon to soil and are responsible for a substantial component of forest-soil carbon fluxes (Söderström, 1992; Högberg et al., 2001; Högberg & Högberg, 2002; Godbold et al., 2006; Hobbie, 2006). ECM fungi are thus regarded as key elements of forest nutrient cycles and as strong drivers of forest ecosystem processes (Read et al., 2004).

The fungi that form ECM associations comprise a taxonomically broad suite of basidiomycetes and, to a lesser extent, ascomycetes (Smith & Read, 1997). Typically, the fungi form a mycelial mantle around short lateral roots of their hosts and penetrate between epidermal and cortical cells, surrounding them with a highly branched structure, the Hartig net (Peterson et al., 2004). Significant variation exists in the morphology of ECM root tips that are infected by different fungal taxa, and analysis of macroscopic and microscopic characteristics of ECM roots is used widely for identification of the ECM fungi (Agerer, 1987–2002). The fungus–plant interface formed by the Hartig net in the ECM root is functionally critical to the mutualism, as it represents the interface across which nutrients and carbon are transferred between the partners (Smith & Read, 1997).

In addition to the structures formed at the host root, ECM fungi produce mycelia that extend from the mantle into the surrounding soil (Fig. 1), although the extent and structure of this extramatrical mycelium is thought to differ between ECM fungal taxa (Agerer, 2001). The soil-borne mycelia of ECM fungi are regarded as functionally important to the mutualism in foraging for, and translocation of, nutrients and water. They also infect short lateral roots (Smith & Read, 1997) and potentially form a common mycelial network that might facilitate carbon and/or nutrient movement between individual tree hosts (Simard & Durall, 2004; Selosse et al., 2006). At the ecosystem level these mycelia are of further significance in the mobilization of nutrients from organic and recalcitrant inorganic sources, both in competition and in cooperation with soil-dwelling saprotrophs, in the relocation of nutrients within the soil–plant continuum, and in the delivery and distribution of carbon belowground (Leake et al., 2002; Wallander, 2006). Despite the importance of soil-borne mycelia of ECM fungi in forest ecosystems, along with their likely importance in the global carbon budget (Hobbie, 2006) and sensitivity to global change factors such as rising atmospheric CO2 (Rillig et al., 2002), we know relatively little regarding their extent, diversity and spatial location in forest soils. In particular, the current lack of fundamental information regarding the physical and physiological continuity of belowground mycelia limits our understanding of the importance of ECM mycelial networks in nutrient cycling and in the distribution of carbon and nutrients within forest-soil systems (Cairney, 2005). Furthermore, although many detailed investigations of communities of ECM fungi colonizing root tips have been undertaken (see below), the relative importance of particular taxa in root-tip communities gives no insight into the extent of their mycelial systems in soil (Genney et al., 2006), nor into their contributions to nutrient and carbon cycling.

Figure 1.

 Mycelial systems of ectomycorrhizal fungi. (a) A typical microcosm system demonstrating the mycelium of Suillus variegatus growing from a Scots pine seedling (photo taken by P.M.A. Fransson). (b) Russula sp. fruiting body and associated mycelium. (c) A close-up of (b) showing the base of the fruiting-body stipe, the associated white mycelium, and ECM root tips (arrows) colonized by this fungus. (d) A dense mycelial mat that has been lifted from a rock surface in a native Scots pine forest. Patches of brown, pink, white and yellow mycelium of different fungal species are visible (photo taken by C.D. Campbell).

To date, investigation of ECM mycelia in soil has been constrained by the difficulty in identifying ECM mycelia and in separating these from mycelia of non-ECM soil fungi. Most of our current understanding has thus been derived via inference from the distribution of fruiting bodies and ECM root tips in the field and/or from observations of mycelia produced in microcosms in the laboratory (Fig. 1). As we will outline in this review, many of the assumptions upon which these inferences are based are somewhat tenuous, and current understanding of complex ECM mycelial systems in forest soils remains limited. Recent years have, however, witnessed the development and application of a suite of molecular methods that are yielding unprecedented insights into ECM mycelia in forest soils. In this review we highlight some of the limitations of our current knowledge, along with the most significant recent advances in our appreciation of the distribution, dynamics and activities of soil-borne ECM mycelial systems along with their responses to elevated atmospheric CO2 concentrations.

Fruiting bodies as indicators of community structure

Although frequently used in surveys of higher fungal diversity, fruiting bodies provide little useful information regarding the diversity of ECM fungal mycelia in soil. Clearly, the presence of an epigeous fruiting body indicates that mycelium of the fungus must be present in one form or another in the underlying soil. Because many ECM taxa produce no epigeous fruiting bodies, or may not have fruited during the duration of the survey, it is impossible to infer the importance of a fungus in the belowground community on the basis of fruiting-body observations. In fact, where fruiting-body data have been compared with ECM root tips (see below), there is generally a very poor correlation between the two, with the latter typically indicating greater species richness (Erland & Taylor, 2002). Furthermore, there is evidence that, for Laccaria spp. at least, physiological and/or genetic differences may influence fruiting at the intraspecific level (Selosse et al., 2001).

Estimation of mycelial distribution using fruiting bodies

For those ECM fungi that produce epigeous fruiting bodies, their presence has been used to infer the distribution of mycelia of the fungi in forest soils. This method centres on analysis of the mycelial genotype in the vegetative tissue of the fruiting body and mapping the distribution of the genotypes in a forest stand. Fruiting bodies that have the same vegetative genotype must have developed from a genetically identical mycelium that has grown through soil, thus allowing estimation of the distribution of soil-borne mycelia (Dahlberg & Stenlid, 1995). Initially this was achieved by isolation of vegetative fruiting-body stipe mycelium into axenic culture, followed by somatic incompatibility testing to identify which isolates are genetically identical (e.g. Dahlberg & Stenlid, 1990, 1994; Baar et al., 1994). While useful for fungi that are readily cultured, many ECM fungi are difficult or impossible to isolate into axenic culture by conventional methods (Brundrett et al., 1996), limiting the usefulness of somatic compatibility testing for ECM fungi. Concerns have also been raised about the resolution of the method for identifying genotypes of some ECM taxa (Sen, 1990; Jacobson et al., 1993).

More recently, molecular methods for the identification of genotypes using DNA extracted from vegetative fruiting-body tissue have been used in this context. A range of dominant molecular markers, including random amplified polymorphic DNA (RAPD), simple sequence repeat (SSR), amplified fragment length polymorphism (AFLP), single-strand conformational polymorphism (SSCP) and inter-retrotransposon amplified polymorphism (IRAP) markers (e.g. De la Bastide et al., 1994; Anderson et al., 1998; Bonello et al., 1998; Sawyer et al., 1999; Redecker et al., 2001; Murata et al., 2005), along with codominant repeat SSR markers (e.g. Zhou et al., 2001a; Dunham et al., 2003; Kretzer et al., 2004; Wu et al., 2005; Bergemann et al., 2006), has now been used to infer the genotype distributions of certain ECM fungi in a variety of forest habitats. While some of these molecular methods may not resolve all genotypes in a population (Redecker et al., 2001; Dunham et al., 2003; Bagley & Orlovich, 2004), they have been widely used to estimate the spatial distribution of ECM fungal genotypes.

It is thus evident that genotypes of some ECM fungi are spread over tens of metres in some forests, implying that mycelia can extend for considerable distances, and persist for several years, in the field (e.g. Dahlberg & Stenlid, 1990; Baar et al., 1994; Dahlberg, 1997; Anderson et al., 1998; Bonello et al., 1998; Sawyer et al., 1999). Estimates of belowground mycelial growth rates (e.g. Dahlberg & Stenlid, 1990; Bonello et al., 1998; Sawyer et al., 1999; Lian et al., 2006) and investigations of the survival of genotypes introduced as ECM inoculum in forest plantations (e.g. De la Bastide et al., 1994; Selosse et al., 1998; Sawyer et al., 2001) provide strong supplementary evidence that mycelia of certain ECM fungi can persist in forest soils for at least tens of years. Whether these genotypes persist in soil as continuous mycelia [=genets (Dahlberg & Stenlid, 1995)], or have fragmented into smaller mycelia [=ramets (Dahlberg & Stenlid, 1995)] as a result of, for example, feeding activities of mycophagous soil arthropods (Schneider et al., 2005) or other forms of disturbance, remains unclear.

Fruiting-body collections indicate that multiple genotypes of some species, such as Laccaria spp., frequently occur in an area of a few square metres, suggesting the presence of many spatially restricted and ephemeral mycelia in the underlying soil (e.g. Gryta et al., 1997; Gherbi et al., 1999; Fiore-Donno & Martin, 2001; Dunham et al., 2003; Murata et al., 2005). Within genera such as Amanita (Redecker et al., 2001; Sawyer et al., 2001, 2003; Bagley & Orlovich, 2004; Liang et al., 2005), Russula (Redecker et al., 2001; Bergemann & Miller, 2002; Liang et al., 2004; Riviere et al., 2006) or Tricholoma (Murata et al., 2005; Gryta et al., 2006; Lian et al., 2006) there are varying reports of widespread and spatially restricted genotypes. Similar observations have been made in population studies of individual species such as Suillus bovinus (Dahlberg & Stenlid, 1990, 1994), Laccaria bicolor (Baar et al., 1994; Selosse et al., 1998) and Pisolithus albus (Anderson et al., 1998, 2001), making it difficult to predict genotype distribution, and so infer the dimensions of belowground mycelia, strictly on the basis of taxonomic affiliation.

The occurrence of multiple spatially restricted genotypes of an ECM fungus is thought to reflect frequent mycelial establishment from meiospores, while a local population dominated by widespread genotypes is characteristic of the sustained growth of persistent belowground mycelia (Dahlberg & Stenlid, 1995). There is some evidence that, for certain ECM fungi, the former may prevail in recently disturbed or younger forests, while the latter predominate in mature undisturbed forests (Dahlberg & Stenlid, 1990, 1995). This is not always the case, however, because the distribution of host tree species in a mixed forest may influence the distribution of ECM genotypes (Zhou et al., 2000). Furthermore, multiple spatially restricted genotypes have been identified in populations of some ECM fungi in mature forests (e.g. Gherbi et al., 1999; Fiore-Donno & Martin, 2001; Redecker et al., 2001), and widespread genotypes in relatively young plantations (Selosse, 2003). The meaning of these observations in terms of the distribution of soil-borne mycelia is questionable. At best, such information based on fruiting-body distribution allows estimation of possible mycelial dimensions in the underlying soil. It provides no information on mycelial continuity, density or vertical distribution in the soil profile.

Community structure and spatial distribution based on ectomycorrhizal root tips

Root tip-based assessment of belowground ECM fungal diversity

With the realization that analysis of ECM fungal communities on the basis of fruiting-body data is a poor indicator of belowground communities of the fungi (Horton & Bruns, 2001; Erland & Taylor, 2002) came acceptance that communities of ECM fungi are best analysed on the basis of identification and enumeration of short lateral roots infected by ECM fungal taxa (=ECM roots). Initially achieved by morphological identification of ECM roots (e.g. Taylor & Alexander, 1989; Massicotte et al., 1999), PCR-based molecular approaches have become de rigueur for these investigations in recent years (Horton & Bruns, 2001; Anderson, 2006), and a great deal of thought has been invested in designing appropriate sampling strategies (e.g. Horton & Bruns, 2001; Taylor, 2002). Although some specific habitats appear to be characterized by low species diversity of ECM roots (Chambers et al., 2005), numerous such investigations have revealed that communities of ECM roots in multifarious forest habitats are generally extremely species-rich (reviewed Horton & Bruns, 2001). It is also evident that communities of ECM roots vary with vegetation/environmental gradients and chronosequences (e.g. Kernaghan & Harper, 2001; Wurzenburger et al., 2004; Dickie & Reich, 2005; Palfner et al., 2005; Robertson et al., 2006).

ECM root species richness and/or community structure are influenced by a range of factors, including soil type (Gehring et al., 1998; Moser et al., 2005), soil moisture (e.g. Taylor & Alexander, 1989; Kårén et al., 1996; Shi et al., 2002; Walker et al., 2005), season (Giachini et al., 2004), natural nutrient gradients (Toljander et al., 2006), fire (e.g. Visser, 1995; Stendell et al., 1999; Grogan et al., 2000; Smith et al., 2005; and see the review by Cairney & Bastias, 2006), activities of herbivores and plant parasites (Brown et al., 2001; Cullings et al., 2005; Mueller & Gehring, 2006), and the quality and quantity of organic matter (e.g. Conn & Dighton, 2000; Cullings et al., 2003). ECM root-tip communities can also be strongly influenced by a range of forest management practices (reviewed Jones et al., 2003) and by gradients of nitrogen deposition (Taylor et al., 2000; Lilleskov et al., 2002; Dighton et al., 2004). Further anthropogenic factors that have been shown to influence ECM root communities include localized nitrogen addition (e.g. Kårén & Nylund, 1997; Fransson et al., 2000; Peter et al., 2001; Avis et al., 2003; Berch et al., 2006; Carfrae et al., 2006), acid precipitation (e.g. Rapp & Jentschke, 1994; Qian et al., 1998; Roth & Fahey, 1998), elevated atmospheric CO2 concentrations (Rey & Jarvis, 1997; Fransson et al., 2001; Kasurinen et al., 2005), toxic metal contamination (e.g. Markkola et al., 2002), and other forms of anthropogenic pollution (reviewed Cairney & Meharg, 1999).

Diversity and distribution of ECM root tips at different scales

Although belowground root-tip analyses have been instrumental in furthering our understanding of ECM fungal communities, it is without doubt that such studies are confounded both by sampling intensity and by the scale at which the study was conducted (Horton & Bruns, 2001; Taylor, 2002). By constructing species−area curves for data published in four previous studies, Horton & Bruns (2001) demonstrated that, in all but one, insufficient samples were analysed to have covered the diversity of ECM taxa present. This demonstrates that a potentially inaccurate or partial picture of diversity and community composition can be generated if the sampling intensity is insufficient. While this alone has consequences for data interpretation, sampling intensity cannot be considered independently from scale (Lilleskov et al., 2004). This is particularly important because in many field studies the desirable scale for the analysis is the stand or forest level, but, in reality, pattern in most ECM communities occurs at much finer scales. For example, analysis of ECM root-tip community data from eight studies suggested that it is necessary to take cores at least 3 m apart in order to achieve the greatest sampling efficiency for stand-level community analysis (Lilleskov et al., 2004). However, ECM root-tip abundance and similarity of community composition have been shown to be highly variable at much finer scales (5–20 cm) (Tedersoo et al., 2003; Izzo et al., 2005), with a complete change in ECM species composition being observed at a scale of 50 cm in a mixed forest (Tedersoo et al., 2003).

There is an obvious trade-off between scale and sampling intensity in root tip-based investigations of ECM communities, and alternative sampling approaches, including the selection of random root tips from soil cores (Horton & Bruns, 2001; Tedersoo et al., 2003), down to as few as three ECM root-tips per sample (Koide et al., 2005b), have been suggested as a possible way of tackling this problem. In most ECM communities, a small number of ECM species are highly abundant and dominant (Taylor, 2002). In these situations it is difficult to see how this would not confound the ecological interpretation of the data when so few root tips are analysed.

Vertical stratification in ECM root-tip communities

A major advance in our understanding of the belowground ecology of ECM fungi has been the observation that ECM root-tip communities are vertically stratified in the soil profile, with different fungi typically occupying different horizons (Goodman & Trofymow, 1998; Rosling et al., 2003; Tedersoo et al., 2003; Izzo et al., 2005; Baier et al., 2006; Genney et al., 2006). While this might signal a degree of niche differentiation between ECM taxa, caution is required in interpretation because the vertical distribution of ECM roots does not necessarily reflect the distribution of mycelia of the respective taxa (Genney et al., 2006; and see below).

Despite the extensive and often painstaking effort that has been devoted to investigating communities of ECM roots, our understanding of the belowground ecology of ECM fungi in forest ecosystems remains limited. The ECM root-based studies have provided unprecedented insights into the diversity of fungal taxa that form ECM associations, along with aspects of their dynamics under a range of conditions, yet they have failed to consider the components of ECM fungi that are functionally most important in forest soil nutrient and carbon cycling processes − the soil-borne mycelial systems.

Similar to aboveground and belowground comparisons of ECM communities, numerous studies have reported a lack of congruence between ECM species colonizing root tips and those detected using molecular methods to be present as mycelia in the same soil volume (Koide et al., 2005a; Kjøller, 2006). Therefore, the true diversity of ECM species may remain uncovered if those species that are only present as mycelia in a given soil volume, and which are likely to be of some functional importance to their host partners, are not considered. In addition, even where intense sampling protocols have been adopted, the spatial distribution patterns of root-tip communities may not necessarily reflect the patterns of the mycelial distribution in soil of the constituent species (Genney et al., 2006).

Because ECM fungi are ‘nonresource-unit-restricted’ fungi (sensuBoddy, 1999), their mycelia forage through soil to greater or lesser extents for nutrients, water and new short-lateral roots to colonize (Read, 1992). In doing so, they form indeterminate, structurally and physiologically heterogeneous interconnected networks of which the colonized roots of the plant host form a part (Rayner, 1991; Cairney & Burke, 1996). There is a growing awareness not only that investigation of ECM mycelia in forest soils is imperative to a functional understanding of ECM communities, but also that communities of ECM fungi when considered from the perspective of their mycelial systems in soil may be quite different from those identified on the basis of colonized root tips (e.g. Horton & Bruns, 2001; Cairney, 2005; Anderson, 2006; Genney et al., 2006; Wallander, 2006).

Estimation of mycelial distribution in ectomycorrhizal populations using root tips

In addition to community analysis, ECM roots have been used to estimate the distribution of ECM mycelial genotypes belowground. This work has confirmed the observations based on fruiting-body distribution (see above) that genotypes of some taxa are spatially restricted belowground and that the distribution of others can be widespread (Guidot et al., 2001; Hirose et al., 2004; Kretzer et al., 2004; Lian et al., 2006). It is also clear that genotypes of some ECM fungal taxa can infect multiple hosts in the field simultaneously (Lian et al., 2006), providing circumstantial support for the notion of a common mycelial network in forest soil (Peter, 2006). Importantly, it has also become evident that while estimates of genotype distribution based on fruiting-body distribution sometimes correlate roughly with estimates based on root tips (Guidot et al., 2001; Zhou et al., 2001b; Lian et al., 2006), in other instances root tip-based estimates indicate more widespread genotype distribution (Hirose et al., 2004). Furthermore, there is strong evidence that belowground genotypes of at least some taxa, can differ in temporal fruiting patterns (Guidot et al., 2001; Zhou et al., 2001b; Hirose et al., 2004), meaning that, unless repeated sampling is conducted, fruiting-body analysis may only unveil part of the population that is present. Although, with an appropriate sampling strategy, root-tip studies may provide some information on the vertical distribution of mycelial genotypes in ECM roots (Zhou et al., 2001b), they do not necessarily provide an accurate reflection of the vertical distribution of the equivalent soil-borne mycelia (Genney et al., 2006).

Direct observation of ectomycorrhizal mycelia in the field

The indeterminate filamentous nature of fungal hyphae, coupled with the complex nature of soil fungal communities (Fig. 1) and difficulties in identifying fungal mycelia in the absence of fruiting bodies, makes direct observation and identification of fungal mycelia in soil virtually impossible, and ECM fungi are no exception in this regard (Bridge & Spooner, 2001; Cairney, 2005). Some higher fungi produce macroscopic linear mycelial aggregates (=rhizomorphs sensuCairney et al., 1991) during growth through soil. In the case of certain saprotrophic fungi, robust rhizomorphs are produced within the soil−litter interface, and this has facilitated excavation, direct measurement and mapping of more-or-less intact mycelial systems (see Thompson, 1984). Although many ECM fungi form rhizomorphs, these are typically diminutive, and thus direct observations of ECM mycelia in soil have been largely confined to those produced within a few centimetres of ECM root tips or fruiting bodies. From investigations of this nature, it is evident that different ECM fungi produce different amounts of soil-borne mycelium, and that the propensity for mycelia to differentiate to form rhizomorphs also varies considerably (reviewed by Agerer, 2001). Mycelia that remain nonrhizomorphic are thought to reflect a limited ability to explore surrounding soil, while mycelia that comprise highly differentiated rhizomorphs are regarded as more adapted to long-distance exploration (Agerer, 2001). Between the two extremes are a range of more-or-less differentiated mycelial types that facilitate medium-distance exploration in soil (Agerer, 2001). According to Agerer (2001), the most highly differentiated rhizomorphs can explore ‘up to several decimetres’ from the root surface in soil. Difficulties in excavating even the most robust ECM rhizomorph systems in soil make it difficult to determine the extent to which this underestimates the exploration potential of ECM mycelia in soil, but nonetheless this classification scheme represents a useful way to categorize ECM fungi based on the potential ecological relevance of their mycelial systems.

Some ECM fungi produce dense mycelial mats or shiro in forest soil. Mat-forming fungi such as Gaultheria and Hysterangium spp. produce persistent dense mycelia that occupy the upper few centimetres in soil and occupy areas that are typically <1 m2 (Cromack et al., 1979; Griffiths et al., 1991). While discrete mats produced by these fungi can be identified and measured in soil, it is not clear whether they comprise single or multiple mycelia of the fungi in question. The shiro produced by Tricholoma matsutake persist for many years and develop by outward growth in a ‘fairy ring’ fashion from a central point for up to several metres (Ogawa, 1977). Recent analysis of vegetative mycelia from fruiting bodies indicate that a single shiro of T. matsutake can comprise multiple mycelia (Murata et al., 2005); however, the distribution of these mycelia within the shiro remains unclear.

Observations of mycelia in microcosms

Our understanding of the physiological activities of ECM mycelia in soil has been built largely on microcosm studies. The most widely used microcosm system comprises a thin layer of milled peat or soil between two Perspex sheets (Duddridge et al., 1980), or a similar construction (Fig. 1). Such microcosms have been used to elegantly demonstrate the roles of ECM mycelia in the absorption and translocation of water, phosphorus and nitrogen from soil to the plant host (Duddridge et al., 1980; Finlay & Read, 1986a, b; Finlay et al., 1988, 1989; Ek et al., 1994; Andersson et al., 1996; Timonen et al., 1996; Ek, 1997), along with establishing the mycelia as important conduits for carbon movement from tree hosts into soil (Finlay & Read, 1986a, b; Ek, 1997; Leake et al., 2001; Mahmood et al., 2001; Wu et al., 2002; Heinonsalo et al., 2004; Rosling et al., 2004). The microcosms have also been pivotal in identifying potential patterns and processes involved in the growth, development and differentiation of ECM mycelial systems (Read, 1992; Donnelly et al., 2004), along with the likely influence of edaphic changes such as pH (Erland et al., 1990; Ek et al., 1994; Mahmood et al., 2001). An important observation here has been the foraging behaviour of ECM mycelia in soil and, in particular, their abilities to densely colonize discrete patches of various organic substrates or necromass from which they can effect nutrient mobilization (Finlay & Read, 1986a, b; Carleton & Read, 1991; Bending & Read, 1995a, b; Pérez-Moreno & Read, 2000, 2001a, b; Leake et al., 2001; Lilleskov & Bruns, 2003; Donnelly et al., 2004; Wallander & Pallon, 2005). Microcosms have also been used to investigate factors such as ECM mycelial longevity and temporal changes in elemental composition (Downes et al., 1992; Wallander & Pallon, 2005), interactions with soil minerals (Wallander et al., 2002), responses of mycelia to elevated atmospheric CO2 concentrations (e.g. Rouhier & Read, 1998, 1999; Fransson et al., 2005), spatial heterogeneity in enzyme and gene expression in mycelial systems (Timonen & Sen, 1998; Wright et al., 2005), and to quantify ECM myelial area, density, biomass and elemental content in soil (Ek, 1997; Schubert et al., 2003; Agerer & Raidl, 2004; Donnelly et al., 2004; Hagerberg et al., 2005).

Although microcosm-based studies have facilitated significant advances in our appreciation of the activities of ECM mycelia in soil, there is a danger that they may have painted a somewhat unrealistic picture of the extent of ECM mycelial system development in the field. Images of ECM plants in microcosms generally portray a seedling with a small root system and an ECM mycelial system that radiates throughout much of the microcosm soil, exploring an area of soil that is several times that occupied by the root system (Fig. 1). The ‘extensive soil-colonizing vegetative mycelium’ produced in microcosms is viewed as representing an ideal model of ECM systems in forest soil (Sen, 2000); however, such extrapolation needs to be undertaken with caution for several reasons. First, the extent of mycelial development in more-or-less homogenous milled or sieved microcosm substrates is unlikely to reflect that through the heterogeneous matrix of natural forest soil. Indeed, as highlighted by Smith & Read (1997), the intensive hyphal colonization observed in patches of introduced organic matter (e.g. Finlay & Read, 1986a, b; Bending & Read, 1995b; Pérez-Moreno & Read, 2000; Leake et al., 2001) may be more representative of mycelial growth in forest soil than the extensive mycelial exploration that occurs through the homogeneous microcosm substrates. The largely two-dimensional nature of the microcosms means that exploration of the substrate is effectively on a single plane, and thus the distance that mycelium grows from the root system is likely to be greater than would be the case for the equivalent mycelial biomass in a three-dimensional soil mass. Even where three-dimensional microcosms have been used (e.g. Coutts & Nicholl, 1990), however, observations of mycelia rely on growth at the interface between the soil and the transparent microcosm window, and, as such, are unlikely to reflect growth of the mycelia within the soil column. Add to this the lack of vertical stratification of soil (see below) in the microcosms (although the rudimentary soil profile reconstruction of Heinonsalo et al. (2004) partially addressed this), along with the small size of seedlings, for which the carbon balance and patterns of carbon allocation are likely to be dissimilar to that of mature trees in the field (Ericsson et al., 1996), and the relative growth of mycelia in microcosms may not be much of a guide to mycelia in the field.

Despite these limitations, rates of growth of ECM mycelia in microcosms (Read, 1992) have been used to estimate rates of mycelial expansion in the field (e.g. Bonello et al., 1998; Sawyer et al., 1999). With the above in mind, and the fact that the nature of the substrate and edaphic conditions can strongly affect ECM mycelial growth (Erland et al., 1990; Smith & Read, 1997), such estimation must be viewed with a degree of caution. Attention must also be drawn to the fact that, in most cases, microcosms reflect mycelial growth of a single ECM fungus in the absence of competing ECM or other macro fungi. Where competing ECM fungi have been coinoculated in microcosms, it is clear that mycelial development can be altered significantly (Wu et al., 1999). While the outcomes of interactions appear to vary, largely according to relative carbon availability to the individual fungi (Lindahl, 2000; Lindahl et al., 2001), it is further evident that the presence of saprotrophic basidiomycete mycelia can influence the growth and development of ECM mycelia in microcosms (Lindahl et al., 1999, 2001; Leake et al., 2001, 2002). Because mycelia of different taxa may preferentially grow at different depths in the soil (Lindahl et al., 1999), the largely two-dimensional nature of the microcosms will reduce the potential for different mycelia to avoid competition by vertical stratification of growth (Cairney, 2005). Care must therefore be taken when attempting to extrapolate outcomes of competitive interactions between mycelia in microcosms to the field.

Estimation of mycelial biomass in the field

Given the apparent importance of ECM mycelia in forest carbon and nutrient cycles, an understanding of their belowground biomass is pivotal to fully appreciating their contributions to biogeochemical processes. While estimates of mycelial biomass in simple nonsoil substrates have been obtained in the laboratory (e.g. Colpaert et al., 1992; Rousseau et al., 1994), determining ECM mycelial biomass in field soil has, until recently, remained a vexed issue. Högberg & Högberg (2002) adopted an indirect method that involved tree girdling, soil fumigation and measurements of dissolved organic carbon to estimate that ECM mycelium accounts for >32% of the total microbial biomass in a boreal forest, while estimates have also been derived by long-term incubation of soil in the laboratory. Incubation in the absence of a host plant means that ECM mycelia should degrade during the incubation period, and thus the difference between fungal biomass, estimated by ergosterol or phospholipid fatty acid (PLFA) analysis, before and after incubation, is used as an estimate of ECM fungal biomass (Wallander et al., 2004). Such studies indicate that ECM mycelial biomass decreases with soil depth (Bååth et al., 2004; Wallander et al., 2004).

Hyphal ingrowth bags offer a more direct method for analysis of ECM mycelial growth and biomass in soil. These are nylon mesh (50 μm) bags containing acid-washed sand that are buried in soil and later retrieved for analysis of total fungal content by ergosterol or PLFA analysis (Wallander et al., 2001). The absence of organic matter in the sand substrate is thought to select for mycelia of ECM fungi over saprotrophs, because ECM fungi have a carbon source from their tree hosts (Wallander et al., 2001). Comparison of mycelial biomass in hyphal ingrowth bags from trenched plots (designed to sever ECM roots from their tree hosts, and so exclude ECM mycelia from the plots) with that from untrenched plots suggests that some 85–90% of the mycelium that colonizes the sand-filled bags in Swedish forest soils is ECM (Wallander et al., 2001; Hagerberg & Wallander, 2002; Nilsson & Wallander, 2003). This is supported by δ13C values derived from the bags (Wallander et al., 2001), along with molecular analysis of the fungal taxa present as mycelia in hyphal ingrowth bags from a range of forest habitats (Wallander et al., 2003; Bastias et al., 2006; Kjøller, 2006 and see below). The method relies upon the assumptions that the use of acid-washed sand provides a reasonable approximation of mycelial growth in native soil, and that there is no turnover of mycelium during the period in which bags are buried in soil (Wallander et al., 2001; Hendricks et al., 2006). Recent data suggest that ECM mycelial colonization of bags containing acid-washed sand can be considerably reduced compared with that of native soil, indicating that the former may underestimate ECM biomass in the surrounding soil, a fact that may be exacerbated if mycelial turnover occurs during the burial period (Hendricks et al., 2006). The use of soil as a substrate in ingrowth bags, however, requires that there is relatively low fungal biomass in the soil (Wallander, 2006), and further work on the influence of substrate is clearly required. It has also been suggested that ingrowth bags may select for certain ECM fungal taxa that are able to colonize the substrate rapidly at the expense of slower-growing taxa (Wallander et al., 2003).

Despite the various questions regarding the efficacy of ingrowth bags for investigation of ECM mycelia, this method remains the most direct for estimation of ECM mycelial biomass in the field. Work of this nature indicates that ECM fungal biomass in coniferous forests can be of the order of 100–600 kg ha−1 and that it comprises a substantial part of the soil fungal biomass (Wallander et al., 2001, 2004; Hagerberg et al., 2003; Hendricks et al., 2006). Growth of ECM mycelium appears to be seasonal, with greatest activity in autumn and little or no growth during winter (Wallander et al., 2001; Hagerberg & Wallander, 2002). ECM mycelial biomass has also been shown to vary with soil depth and host tree species, being generally greater in the upper part of the soil profile than in the underlying soil and correlated with the distribution of tree roots in the profile (Wallander et al., 2004; Göransson et al., 2006). It is also evident that ECM mycelial biomass in forest soil can be negatively influenced by soil nutrient, particularly nitrogen, status (Nilsson & Wallander, 2003; Nilsson, 2004; Nilsson et al., 2005; Hendricks et al., 2006). In contrast, fertilization of arctic tundra has recently been shown to significantly increase ECM mycelial biomass (Clemmensen et al., 2006). The extent to which these observations reflect altered mycelium production per se, or a shift in the ECM fungal community towards taxa that produce more/less soil-borne mycelium, however, remains unclear. It is also evident that ECM mycelial production can be influenced by temperature (Clemmensen et al., 2006), and that mycelia can grow in response to certain soil minerals (Hagerberg & Wallander, 2002; Hagerberg et al., 2003) and can contribute to the mobilization of minerals such as apatite (Wallander et al., 2002, 2003).

Analysis of ECM communities by direct DNA extraction

In recent years methods have been developed to allow analysis of microbial communities by direct nucleic acid extraction from soil. These techniques include cloning, denaturing gradient gel electrophoresis (DGGE), temperature gradient gel electrophoresis (TGGE) and terminal restriction fragment length polymorphism (T-RFLP). The technical considerations and limitations associated with these approaches in the context of both general soil (Anderson & Cairney, 2004; Bidartondo & Gardes, 2005) and ECM (Anderson, 2006) fungal communities have been reviewed recently and thus will not be considered here.

Direct DNA extraction from soil

Recent investigations based on direct DNA extraction from soil have yielded information on the diversity and distribution of ECM mycelia in soil (Table 1). Although they did not target ECM fungi specifically, several studies identified mycelia of ECM fungi as being present in the general fungal assemblages of forest soils (Chen & Cairney, 2002; Anderson et al., 2003; Landeweert et al., 2003a). Given that the majority of ECM fungi are basidiomycetes (Smith & Read, 1997), basidiomycete-specific PCR primers have been used to increase the detection of ECM mycelia in soil DNA extracts. Using this method, Landeweert et al. (2003a) reported vertical stratification of certain ECM fungi in the soil profile, while Smit et al. (2003) demonstrated that removal of litter and humus layers from a pine plantation increased the diversity of ECM mycelia. However, the use of basidiomycete-specific primers for the analysis of ECM mycelial communities has been cautioned (Anderson, 2006), because the diversity (e.g. Vrålstad et al., 2002; Tedersoo et al., 2006) and abundance (e.g. Koide et al., 2005a) of ascomycetes in ECM root-tip communities is often high.

Table 1.   Investigations of soil ECM mycelial communities using direct DNA extraction and molecular analysis methods
Forest typeDNA extracted fromAnalysis methodTarget fungal groupReference
Larix kaempferi forestSoil and ECM rootsSimple sequence repeat analysisSuillus grevilleiZhou et al. (2001b)
Sclerophyll forestSoilCloning/sequencingSoil fungiChen & Cairney (2002)
Pinus resinosa plantationSoilT-RFLPECMDickie et al. (2002)
Pinus pinaster forestSoilCompetitive PCRHebeloma cylindrosporumGuidot et al. (2002)
Pinus sylvestris forestSoilDGGE/sequencingSoil fungiAnderson et al. (2003)
Coniferous forestSoilCloning/sequencingBasidiomycetesLandeweert et al. (2003a)
Pinus sylvestris plantationSoilDGGE and cloning/sequencingBasidiomycetesSmit et al. (2003)
Pinus taeda plantationSoilT-RFLPBasidiomycetesEdwards et al. (2004)
Pinus pinaster forestSoilCompetitive PCRHebeloma cylindrosporumGuidot et al. (2004)
Pinus resinosa plantationSoil and ECM rootsT-RFLPECMKoide et al. (2005a)
Pinus resinosa plantationSoil and ECM rootsT-RFLPECMKoide et al. (2005b)
Pinus sylvestris forestSoil and ECM rootsDGGE/sequencingBasidiomycetesLandeweert et al. (2005)
Wet sclerophyll forestHyphal ingrowth bagsDGGE and cloning/sequencingECMBastias et al. (2006)
Pinus sylvestris plantationSoil and ECM rootsT-RFLPECMGenney et al. (2006)
Fagus sylvatica forestHyphal ingrowth bagsCloning/sequencingECMKjøller (2006)
Picea abies plantationHyphal ingrowth bagsDGGE/sequencingECMKorkama et al. (2007)
Pinus sylvestris forestSoilT-RFLP and cloning/sequencingECM and saprotrophic fungiLindahl et al. (2007)

An alternative means of targeting ECM fungi in DNA extracted directly from soil is T-RFLP analysis and interrogation of a database containing reference terminal fragments of known ECM fungi from the site. This has facilitated confirmation of the vertical stratification of ECM mycelial communities in the soil profile (Dickie et al., 2002; Genney et al., 2006). More importantly, Koide et al. (2005b) demonstrated that positive or negative interspecific interactions can influence the distribution of ECM mycelia in soil and that such interactions can vary according to soil conditions.

T-RFLP analysis of DNA extracted from multiple adjacent soil cores allowed Edwards et al. (2004) to estimate the minimum width of mycelial patches of several ECM taxa in a horizontal plane, the widest patch-size recorded being 160 cm for Thelephora terrestris. Moreover, Genney et al. (2006) coupled T-RFLP analysis with a soil-slicing approach to investigate the distribution of ECM mycelia in a volume of soil (800 cm3) at a fine scale. Mycelial patches of individual ECM taxa were variously shown to occupy different volumes up to a maximum of 312 cm3 for Cadophora finlandica. A combination of T-RFLP analysis and cloning has also been used to demonstrate vertical separation of ECM and saprotrophic mycelia in a forest-soil profile (Lindahl et al., 2007). While each of these studies demonstrated the potential distribution of mycelia of various ECM species in forest soil, none considered the distribution of individual mycelia (=genet) of the species. Zhou et al. (2001a, b), however, have demonstrated that simple sequence repeat (SSR) markers can be applied to investigate the distribution of individual mycelia of Suillus grevillei following direct DNA extraction from soil. With the increasing development of SSR markers for ECM fungi (e.g. Bergemann & Miller, 2002; Kretzer et al., 2004; Hitchcock et al., 2006), information regarding the distribution of individual mycelia within the soil volume should increase markedly in the near future.

Even where individual mycelia of ECM fungal taxa can be identified in soil, observations are restricted to determining only the presence or absence of the mycelium in a sampled soil volume, and the methods provide no information on mycelial density. Competitive PCR, however, offers a means of achieving this goal. Indeed, Guidot et al. (2002, 2004) have used this approach successfully to show that soil-borne mycelia of Hebeloma cylindrosporum are ephemeral, extend for <50 cm, and decrease in density with increasing distance from the base of basidiomes.

A fundamental assumption in the interpretation of data generated by direct DNA extraction from soil is that the taxa identified are present as mycelia rather than as spores or other resting propagules. Although approaches such as the collection of soil samples after basidiome production has ceased (Dickie et al., 2002) have been used to minimize the likely detection of spore DNA, the fact that DNA from a small number of H. cylindrosporum spores can be detected in soil DNA extracts (Guidot et al., 2004) demonstrates the potential for spore DNA to be amplified. Given that dormant propagules are likely to contain little RNA relative to active mycelia, RNA-based approaches (e.g. Anderson & Parkin, 2006) may provide a more robust means to detect active ECM mycelia in soil in the future.

The importance of analysing ECM fungal communities by direct DNA extraction from soil is emphasized by the lack of congruence between community structures inferred from root-tip and soil-based analyses (Koide et al., 2005a; Landeweert et al., 2005). Furthermore, it is evident that spatial segregation can exist between root tips and mycelia of individual ECM taxa in the soil profile, with abundant mycelium of some species occurring in the absence of infected roots tips (Genney et al., 2006). Indeed, the same appears to be true for individual mycelia of S. grevillei (Zhou et al., 2001b).

DNA extraction from hyphal ingrowth bags

Hyphal ingrowth bags, in conjunction with ergosterol and/or PLFA analysis, has been the main method used to measure mycelial growth and biomass in forest soil (see above). More recently, however, the approach has been combined with direct DNA extraction from the ingrowth-bag substrate and molecular methods to identify the mycelia of ECM fungal taxa colonizing the bags. Regardless of the molecular method used, these studies indicated that 83–91% of taxa colonizing hyphal ingrowth bags in a wet sclerophyll forest (Bastias et al., 2006), a Fagus sylvatica forest (Kjøller, 2006) or a Picea abies plantation (Korkama et al., 2007) were ECM fungi, which is broadly similar to hyphal ingrowth bag-based estimates of ECM fungal biomass in Swedish forest soils using δ13C measurements and trenching (Wallander et al., 2001; Hagerberg & Wallander, 2002; Nilsson & Wallander, 2003).

The hyphal ingrowth-bag approach has been successfully used in conjunction with DGGE to investigate the effects of plant host and forest management practices on ECM mycelial communities. Thus, Korkama et al. (2007) demonstrated that fast-growing Picea abies clones support different ECM fungal mycelial communities from slower-growing clones, with Atheliaceae taxa more abundant in the former. Bastias et al. (2006) used this method to establish that long-term repeated prescribed burning of wet sclerophyll forest can significantly alter the composition of mycelial communities of ECM fungi in the upper 10 cm of the soil profile. Cloning and sequencing of DNA extracted from the ingrowth bags further suggested that frequent burning had a strong negative effect on Cortinariaceae and a positive effect on Thelephoraceae taxa (Bastias et al., 2006). Comparison of ECM fungi colonizing hyphal ingrowth bags with those species colonizing root tips adjacent to the ingrowth bags in a Fagus sylvatica forest showed that Boletoid species occurred more frequently as mycelia than in root tips, in contrast to Russuliod and Cortinarius spp., for which the opposite was true (Kjøller, 2006). It should be noted that, in each of these investigations, only sand was used as substrate in the ingrowth bags. Because the nature of the substrate can influence mycelial colonization of hyphal ingrowth bags (Hendricks et al., 2006), it may also bias community composition. To date there is no published comparison between mycelia of ECM fungi colonizing sand-filled hyphal ingrowth bags with mycelia in the surrounding soil.

A combination of PLFA analysis and molecular approaches allows determination of both mycelial biomass and ECM species identity in hyphal ingrowth bags (Korkama et al., 2007), but it is difficult to correlate these two independent measurements directly. Correlation of both relative abundance (community evenness) and species identity (species richness) is important for determining the mycelial biomass of individual ECM species and the response of individual ECM taxa in manipulative field experiments. This approach is common in root tip-based community studies of ECM fungi, although the limitations of root tip-based measurements of relative abundance have been previously discussed (Taylor, 2002). While determination of the relative abundance of mycelia of individual ECM species is yet to be achieved, this should be possible in the future by combining quantitative molecular methods such as competitive PCR (e.g. Guidot et al., 2002, 2004) or real-time quantitative PCR (Landeweert et al., 2003b; Schubert et al., 2003; Raidl et al., 2005) with the hyphal ingrowth-bag approach. Given the known variation in the growth form of mycelial systems of different ECM species (Agerer, 2001), this will be an important future development in the use of hyphal ingrowth bags.

ECM mycelia and climate change

The importance of understanding the diversity, extent and dynamics of ECM mycelial systems in forest soils is emphasized by current interest in belowground responses to elevated atmospheric CO2 in the context of global climate change (Pendall et al., 2004) and the need to understand both plant and ECM fungal responses (Alberton et al., 2005). ECM fungal mycelia can comprise 80% of the total fungal biomass and 30% of the microbial biomass in some forest soils (Wallander et al., 2001, 2003; Högberg & Högberg, 2002), with carbon allocation to ECM fungi estimated to be as much as 22% of net primary production (Hobbie, 2006). ECM fungi are thus an important component of forest carbon cycles, and the effects of elevated atmospheric CO2 on these fungi have deservedly received recent attention. Much of this work has focused on interactions between CO2 concentration and root colonization by ECM fungi. With one exception (Rouhier & Read, 1999), such investigations have reported increased percentage root colonization by ECM fungi under elevated atmospheric CO2 conditions (Norby et al., 1987; Ineichen et al., 1995; Berntson et al., 1997; Godbold et al., 1997; Tingey et al., 1997; Rouhier & Read, 1998; Walker et al., 1998; Kasurinen et al., 2005).

Elevated atmospheric CO2 has also been shown to alter ECM fungal root-tip community structure in growth-chamber experiments (Godbold & Berntson, 1997; Godbold et al., 1997; Rey & Jarvis, 1997; Rygiewicz et al., 2000) and in the field (Fransson et al., 2001; Kasurinen et al., 2005). In one such experiment, a change in ECM root-tip community composition in favour of morphotypes that appeared to produce emanating hyphae and/or rhizomorphs was noted under elevated atmospheric CO2 conditions (Godbold & Bernston, 1997; Godbold et al., 1997). Because no attempt was made to quantify soil-borne mycelia, however, this provides no direct information on the mycelial response of ECM fungi under these conditions. Several laboratory-based microcosm experiments have been undertaken with a view to determining the responses of ECM mycelia to elevated atmospheric CO2 (Table 2). While Gorissen & Kuyper (2000) observed no significant increase in mycelial production, other studies have reported increases in mycelial biomass or spread of up to 400% (Ineichen et al., 1995; Rouhier & Read, 1998, 1999; Fransson et al., 2005). It should be emphasized that these investigations were conducted in simplified microcosm systems (see above) using seedlings inoculated with a single ECM fungus. As such, the observed responses may not reflect the responses of ECM mycelia to elevated atmospheric CO2 in the field, highlighting the need for field-based investigations. Although elevated atmospheric CO2 was found to alter the structure of a basidiomycete mycelial community (that included ECM fungi) in a scrub oak forest (Klamer et al., 2002), to date only one such study has focused on ECM fungi. By analysing PLFAs from hyphal ingrowth bags buried under Betula pendula trees in open-top chambers, Kasurinen et al. (2005) established that elevated atmospheric CO2 had no significant effect on mycelial production by ECM fungi.

Table 2.   Responses of ECM fungal mycelia to elevated atmospheric CO2 concentrations
FungusHost treeExperimental systemMycelial responseReference
Hebeloma crustuliniformePinus sylvestrisPot experimentIncreased mycelial biomass (300%)Fransson et al. (2005)
Laccaria bicolorPinus sylvestrisPetri-dish microcosmNo effect on hyphal lengthGorissen & Kuyper (2000)
Paxillus involutusPinus sylvestrisPerspex microcosmIncreased mycelial spread (>400%)Rouhier & Read (1998)
Paxillus involutusBetula pendulaPerspex microcosmIncreased mycelial spread (30%)Rouhier & Read (1999)
Pisolithus tinctoriusPinus sylvestrisPetri-dish microcosmIncreased mycelial biomass (200%)Ineichen et al. (1995)
Suillus bovinusPinus sylvestrisPerspex microcosmIncreased mycelial spread (200%)Rouhier & Read (1998)
Suillus bovinusPinus sylvestrisPetri-dish microcosmNo effect on hyphal lengthGorissen & Kuyper (2000)
ECM mycelial communityBetula pendulaField experiment (hyphal ingrowth bags)No effect on mycelial biomassKasurinen et al. (2005)

The inability to generalize based on this single field experiment notwithstanding, these data relate only to the response of the ECM mycelial community and therefore provide no information on responses of individual ECM species to elevated atmospheric CO2 in the field. Further studies on how mycelial systems of ECM fungi respond to elevated atmospheric CO2 concentrations will be crucial to modelling and understanding carbon dynamics in complex forest ecosystems under future climate-change scenarios.


Soil-borne mycelia of ECM fungi are a dominant and important component of soil microbial communities in forest ecosystems, where they play crucial roles in nutrient and carbon cycling processes. Our appreciation of the roles of ECM fungi in these processes is hampered by a fundamental lack of understanding of their mycelial systems in soil. Much of the currently available information on the diversity, distribution and dynamics of ECM fungi is derived from investigations of fruiting bodies and ECM root tips, and provides only limited insights into soil-borne ECM mycelial communities. Information derived from microcosm studies, while unarguably enhancing our understanding of ECM mycelial functioning, cannot necessarily be extrapolated directly to mycelia in the field, emphasizing the necessity for direct analysis of soil-borne mycelia of ECM fungi in situ. Molecular approaches, combined with new experimental approaches, such as hyphal ingrowth bags, facilitate analysis of ECM fungal communities from the perspective of mycelia in soil. Furthermore, they have the potential to increase our understanding of the distribution of individual mycelia and so enhance our appreciation of the physical and physiological continuity of mycelial systems in soil. Such information will give unprecedented insights into the functioning of ECM mycelia at the ecosystem level and will assume particular importance when considering the responses of ECM fungi to future perturbations, particularly in the context of land-use changes and global climate change.


This work was supported by an Australian Research Council Linkage International Awards grant (LX0455012) and a Macaulay Development Trust collaboration grant. I.C.A. receives funding from the Scottish Executive Environment and Rural Affairs Department.