Introduction

Soil microbes are the primary regulators of terrestrial carbon and nutrient cycling. A powerful cocktail of lingocellulolytic enzymes, capable of the initial deconstruction of complex organic compounds, make saprotrophic fungi the dominant agents of plant litter decomposition in many terrestrial (primarily woodland) ecosystems (Baldrian and Valášková, 2008). During litter decomposition, the C:N ratio gradually decreases and inorganic nutrients are released into the surrounding environment (Lindahl et al., 2002). This ‘slow cycling of nutrients’ maintains ecosystem stability, ensuring the retention of nutrients within late-succession soils (Fierer et al., 2009). Saprotrophic fungi also contribute up to 90% of total heterotrophic respiration in woodland ecosystems (Cooke and Rayner, 1984) and are considered key regulators of soil carbon fluxes between the biosphere and atmosphere.

Saprotrophic fungi are continuously engaged in the search for fresh nutrients. Whereas some (termed resource-unit-restricted) disseminate to encounter new resources solely by production of propagules (spores or hyphal fragments), others (termed non-unit-restricted) extend as hyphae, and interconnect discrete organic resource units. Aggregations of hyphae (forming linear organs known as cords) enable some fungal species to extend vast distances (covering many m2) at the soil–litter interface (Boddy, 2000). Their ability to translocate carbon, nitrogen and phosphorus throughout these cords means that saprotrophic fungi are important agents of nutrient redistribution in soil (Cairney, 2005).

Mycelial growth inevitably leads to encounters with antagonistic soil organisms. Invertebrates contribute extensively to the species and functional diversity within soils. Their classification is usually based on size (micro-, meso- or macrofauna) or functional group (Rusek, 1998). In terms of nutrient cycling, the most important are the arthropods, oligochaetes, molluscs and nematodes, the majority of which are primarily mycophagous (Pollierer et al., 2009). Damage to hyphal compartments and faecal pellet production directly influence the release of labile nutrients into the surrounding soil (Boddy and Watkinson, 1995); an important process in the recycling of nutrients to plants (Bardgett and Chan, 1999). Invertebrate ‘grazing’ can also influence mycelial growth and physiology (Crowther et al., 2011a, 2011c). Given the prominent roles of saprotrophic fungi in nutrient decomposition and redistribution (Hättenschwiler et al., 2005), grazing may represent the primary role of soil fauna in woodland ecosystem functioning.

The various mechanisms by which invertebrates and saprotrophic fungi interact (trophic and non-trophic) have been reviewed extensively (Lussenhop, 1992; Maraun et al., 2003). This review focuses on the consequences of grazing interactions on fungal functioning and community ecology (Figure 1). Specifically, it explores the effects of soil invertebrates on: (i) fungal-mediated nutrient distribution (via mycelia and spores); (ii) mycelial physiology and nutrient cycling (extracellular enzyme production and respiration); and (iii) fungal community composition and diversity.

Figure 1
figure 1

Breakdown of the key functional and ecological aspects of fungus–grazer interactions.

Full size image

Mycelial growth and propagule dissemination: redistribution of nutrients

Propagule dissemination

Grazing invertebrates can damage or destroy fungal propagules, but can also act as vectors, assisting in their dispersal. This can be passive— resulting from external adherence of propagules to invertebrate bodies or passage in guts —or active, when fungal propagules are carried in specialised sacs (for example, mycangia) on, or in, the invertebrate’s body. Active dissemination often occurs where mutualistic symbiosis has evolved (for example, between the higher termites (Macrotermitinae) and basidiomycetes in the genus Termitomyces (Aanen and Boomsma, 2006), attine ants and basidiomycete Attamyces, Leucoagaricus and Lepiota (North et al., 1997), and woodwasps (Siricidae) and Amylostereum (Basidiomycota) (Slippers et al., 2003)). Passive dispersal of fungal propagules by oribatid mites (Renker et al., 2005), earthworms (Moody et al., 1996), collembola (Visser et al., 1987) and enchytraeids (Hedlund and Augustsson, 1995) can also result in the vertical and horizontal redistribution of propagules within litter layers.

Passive transport via passage through invertebrate guts is not always successful; germination and subsequent development can be inhibited. Earthworm (Lumbricus terrestris and Aporrectodea longa) grazing, for example, reduced or prevented germination of several basidiomycete and ascomycete spores, but propagule survival varied between species; germination of ascomycete, Chaetomium globosum, spores was stimulated following ingestion by A. longa (Moody et al., 1996). This specialised association conferred a selective advantage to the stimulated fungal species following grazing. Similar symbiotic interactions with termites (Mueller and Gerardo, 2002) and dipteran larvae (Nuss, 1982) have also been found to stimulate spore germination. For some fungi (for example, basidiomycete species of Ganoderma), passage of spores through invertebrate guts is essential to allow successful germination (Nuss, 1982).

Mycelial distribution

The extensive mycelial systems of non-unit-restricted fungi represent highly dynamic channels through which nutrients are translocated. Network growth and development are key processes in the redistribution of nutrients within woodland soils (Cairney, 2005). By severing mycelia and ingesting growing hyphal tips, mycophagous collembola can restrict mycelial extension (Tordoff et al., 2006) and high-intensity enchytraeid and woodlouse grazing can completely remove foraging Hirsutella rhossiliensis and Resinicium bicolor systems, respectively (Jaffee et al., 1997; Crowther et al., 2011a). As well as limiting the ability of foraging basidiomycetes to encounter new resources, grazing also disrupts the translocation of carbon (Butenschoen et al., 2007) and nitrogen (Tordoff et al., 2011) through mycelial systems. These changes in nutrient partitioning affect the dynamics and spatial heterogeneity of forest floor nutrients.

Mycelial systems are highly dynamic and most species show distinct growth responses during grazing. Compensatory growth, analogous to that seen in plants during herbivory (McNaughton, 1983), is characterised by increased mycelial extension and branching of hyphae around thick basidiomycete cords. This may facilitate increased nutrient uptake by fine hyphae to counteract the negative effects of grazing (Bengtsson et al., 1993). It has also been interpreted as an ‘escape response’, increasing mycelial extension into uncolonised regions of soil (Hedlund et al., 1991). These effects depend on grazing intensity. Compensatory growth is commonly associated with low-density collembola grazing (Figure 2), whereas growth is inhibited ‘beyond a certain density threshold’ (Crowther and A’Bear, 2012). By controlling grazer populations, predatory invertebrates can exert top–down control of fungal communities. Predation of mycophagous collembola (Folsomia fimetaria) by predatory mites (Hypoaspis aculeifer), for example, limited their grazing potential, indirectly stimulating growth and respiration rates of three soil ascomycetes (Hedlund and Ohrn, 2000). In contrast, factors (abiotic and biotic) contributing to increased collembola abundance often lead to reduced fungal biomass (Lenoir et al., 2007).

Figure 2
figure 2

Digital images showing mycelia of Hypholoma fasciculare (a, b, c), and Phanerochaete velutina (d, e, f) during ungrazed (a and d), low intensity (b and e) and high intensity (c and f) F. candida grazing treatments following 10 days of growth from 2 × 2 × 1 cm wood blocks on 24 × 24 cm soil trays. Low-intensity grazing (783 collembola m−2) stimulated mycelial extension, whereas high intensity (1566 collembola m−2) restricted growth of both fungi.

Full size image

Interspecific variation in invertebrate size, metabolic activity and feeding preference also leads to contrasting grazing pressures. This is apparent in the stark interspecific differences between collembola grazing effects (Tordoff et al., 2008). Moreover, variation between invertebrate orders often outweighs that within orders. Macrofauna (isopods and millipedes) have a greater influence on basidiomycete growth than meso- and microfauna (Crowther et al., 2011a). The capacity of larger invertebrates to sever thick cords and consume entire mycelia enables them to limit mycelial foraging (Crowther and A’Bear, 2012), whereas lower intensity grazing by smaller invertebrates often stimulated fungal growth (Crowther et al., 2011b). The potential for mycelial networks to acquire new resources and redistribute nutrients, therefore, depends on both invertebrate density and community composition.

Physiological responses: decomposition and nutrient cycling

Enzyme production

Hydrolytic enzymes produced by saprotrophic fungi are responsible for the initial steps in the deconstruction of plant cell walls and the mineralisation of complex compounds into simple inorganic molecules (sugars, amino acids, NH4+, PO4−3, H2O and CO2) that can be assimilated by plants and microbes. Lignocellulolytic enzyme production by saprotrophic basidiomycetes colonising leaf litter increases during macrofauna (Scheu, 1993) and collembola (Parkinson et al., 1979) activity. This may be attributed, in part, to litter comminution, but fungal physiological responses to grazing will also contribute to the enhanced nutrient mineralisation (Osono, 2007). Few studies have separated these indirect (litter comminution) and direct (mycelial physiological responses to grazing) effects, but in one such study, collembola (Protaphorura armata) grazing increased protease and α-amylase production by the zygomycete, Umbelopsis isabellina, growing in agar (Hedlund et al., 1991). The nematode, Panagrellus redivivus, also stimulated protease production by the basidiomycete P. velutina, but induced the opposite effect while grazing on Stereum hirsutum (Dyer et al., 1992). Species-specific fungal enzymatic responses have also been recorded in soil (Crowther et al., 2011c). Hypholoma fasciculare and P. velutina (exhibiting fast, extensive growth), increased production of cellulolytic and phosphorolytic enzymes during macro-invertebrate grazing, whereas the slow-growing species, R. bicolor, reduced enzyme production, probably to conserve energy during mycelial disturbance (Figure 3). Contrasting enzymatic responses of fungal species suggest that the impacts of soil fauna on fungal-mediated nutrient mineralisation are dependent on fungal community composition.

Figure 3
figure 3

Activities of 1,4-β-glucosidase (BG), 1,4-α-glucosidase (AG), cellobiohydrolase (CBH), 1,4-β-xylosidase (BX), arylsulfatase (AS), N-acetylglucosaminidase (NAG), a acid phosphatase (AP) and phosphodiesterase (PDE) enzymes in soil colonised by Phanerochaete velutina, Hypholoma fasciculare and Resinicium bicolor. Mean and s.e. are shown for each enzyme activity during fungus-only control (), F. candida (), O. asellus (), Blaniulus guttulatus () and Panagrellus redivivus () grazing treatments as well as uncolonised soil (). Figure was modified from Crowther et al. (2011c).

Full size image

As with mycelial distribution, grazer effects on fungal enzyme production are dependent on grazing intensity. Macrofauna had a greater influence than meso- or microfauna on the production of eight hydrolytic enzymes (associated with C-, N- and P-cycling) by basidiomycete networks in soil (Crowther et al., 2011c). Although the nature of enzymatic responses were fungal species-specific, it was generally the high-intensity grazing associated with larger invertebrates that initiated the changes in fungal physiology.

Decomposition and respiration

Enzymatic responses of saprotrophic fungi will have direct consequences for wood and leaf-litter decomposition. Generally, invertebrate activity is thought to stimulate litter decomposition (Bardgett, 2005) although, as with enzyme production, few studies have distinguished between direct and indirect grazer effects. When restricted to feeding on extra-resource mycelia growing in soil, collembola affect the abilities of fungi to decompose wood resources (Tordoff et al., 2008). These effects vary between fungal species; during grazing, H. fasciculare showed increased rates of wood decay, whereas the opposite was true of R. bicolor (Crowther et al., 2011b). This difference is directly related to opposing enzymatic responses of these two fungal species during grazing.

While the importance of invertebrate density on extracellular enzyme production remains unexplored, effects can be seen on fungal-mediated litter decomposition. By reducing fungal biomass and activity, high-density collembola populations reduce wood decay rates (Tordoff et al., 2006, 2008), but lower intensity grazing can stimulate fungal nutrient uptake from wood (Crowther et al., 2011a). This reflects the contrasting effects of micro- and macrofauna on fungal-mediated wood decay. Heavy grazing on extra-resource mycelia by the woodlouse, O. asellus, reduced the decomposition of R. bicolor-colonised wood, whereas less intense nematode, P. redivivus, grazing stimulated decay rates, although this effect was not statistically significant (Crowther et al., 2011b).

During decomposition, CO2 is released from litter resources via microbial respiration, which is directly related to biomass and enzyme activity (Table 1); at high densities collembola reduce fungal respiration rates (Bardgett et al., 1993), whereas low-intensity grazing stimulates CO2 production (Hanlon and Anderson, 1979). This density-dependent trend is also found during enchytraeid (Hedlund and Augustsson, 1995) and oribatid mite (Kaneko et al., 1998) grazing. Although the importance of invertebrate species identity is as yet unknown, the clear parallels between mycelial enzyme activity and respiration suggest that macrofauna are likely to be particularly important in influencing the fungal-mediated flux of carbon between the terrestrial and atmospheric carbon pools.

Table 1 Effects of invertebrate grazing on fungal biomass and activity
Full size table

Fungal community structure

The majority of soil fauna are generalists, capable of exploiting a variety of microbial resources within highly diverse and heterogeneous environments (Setälä et al., 2005). Indiscriminate feeding can have major consequences for fungal species richness. Wicklow and Yocom (1982) found that the number of coprophilous fungal species on rabbit faeces was reduced during grazing by sciarid fly, Lycoriella mali, larvae. By ingesting entire microfungi (1–100 μm), indiscriminate grazers reduced the amount of competition between fungi and remaining fungi were ‘released’ from competitive stress. This led to increased rates of organic matter decomposition. Via this process, indiscriminate grazers can alter the cycling of nutrients and efflux of CO2 from soil.

Despite their polyphagous nature most mycophagous fauna display distinct feeding preferences for nutritious or palatable fungi (Maraun et al., 2003). Preferential grazing can exert selective pressures on fungal communities, favouring the growth of less palatable species. This process is particularly important during early stages of fungal succession as late-succession fungi are often less susceptible to grazers (Lussenhop, 1992). Selective grazing by F. candida on primary saprotrophs led to faster replacement by secondary saprotrophs on spruce and fir needles (Klironomos et al., 1992). By restricting the less competitive species, collembola facilitated fungal succession in decaying litter. Grazers also regulate this process by an opposing mechanism; stimulation of the dominant competitor. In soil microcosms grazing by F. candida stimulated growth of the dominant species, P. velutina, over its opponent, H. fasciculare (Rotheray et al., 2011). This supported previous studies where collembola reinforced the outcomes of competitive mycelial interactions, stimulating fungal species turnover but not eventual species composition (Parkinson et al., 1979; McLean et al., 1996).

Although numerous studies show selective collembola grazing, none have reported the complete reversal of mycelial interaction outcomes. Newell (1984a, 1984b) provided some evidence that collembola grazing can alter the relative abundances of competing basidiomycetes in Picea sitchensis needle litter, but this did not lead to the complete replacement of a dominant species by a less competitive opponent. In most cases, the grazing pressures exerted by mesofauna are not strong enough, even at twice the recorded field densities, to alter fungal competitive abilities (McLean et al., 1996). In contrast, macrofauna, can exert stronger top–down pressures on fungal community composition (Crowther et al., 2011d). By ingesting entire R. bicolor mycelial systems the isopod, O. asellus, prevented the competitive exclusion of H. fasciculare and P. velutina. Nematodes (P. redivivus), also influenced the progression of fungal interactions, stimulating growth of the less competitive H. fasciculare, enabling it to overcome its opponent, R. bicolor. Thus, two opposing mechanisms (suppression of the dominant competitor and stimulation of the less competitive species) allow soil fauna to determine the outcomes of competitive mycelial interactions and regulate fungal community compositions in soil.

Perhaps the most extreme example of invertebrates influencing fungal community composition is found with the mutualistic symbiotic relationships of higher termites and ants with their respective basidiomycete symbionts. In both systems, the invertebrates maintain monocultures of the fungi by antibiotic secretions and physical grooming (North et al., 1997; Aanen and Boomsma, 2006). Invertebrates also influence fungal community composition through differential survival of ingested spores. The stimulated germination of C. globosum spores following earthworm, A. longa, grazing exerted selective pressures on the resulting microbial community (Moody et al., 1996). Although grazing will not influence the competitive abilities of germinating fungi, the short-term selective advantage and rapid establishment of stimulated species is a key process influencing fungal species composition and activity.

Comparisons with mycorrhizal fungi

Soil faunal grazing can influence the development of plant–mycorrhizal associations, as well as the transfer of nutrients between symbionts (Fitter and Sanders, 1992; Gange and Brown, 2002). A number of parallels can be drawn between the influence of grazers on saprotrophic and mycorrhizal systems: (i) spore ingestion can have positive and negative effects on spore distribution, depending on the grazer–fungus combination (Harinikumar and Bagyraj, 1994); (ii) severing of hyphae can restrict growth and nutrient translocation throughout hyphal networks (Klironomos and Kendrick, 1995); and (iii) grazing can have positive or negative effects on fungal activity (Gormsen et al., 2004) with consequences for nutrient mineralisation and primary productivity. As with saprotrophic fungi, grazing effects are density dependent. High-intensity grazing generally reduces mycorrhizal activity, whereas low-intensity grazing can stimulate growth, directly increasing plant productivity (Klironomos and Ursic, 1998). Contrasting effects of earthworm and collembola populations suggest that effects are also taxon specific (Gormsen et al., 2004).

Invertebrate preference for saprotrophic over mycorrhizal fungi represents a clear difference between the effects of grazers on these two major fungal groups. When provided with a choice, invertebrates consistently prefer, and reproduce more successfully on, saprotrophic than mycorrhizal fungi (Gange, 2000). As a result, in complex, multi-species environments soil fauna commonly stimulate mycorrhizal growth by suppression of competing or inhibitory fungi (Klironomos and Kendrick, 1995). Plant pathogenic fungi are also generally preferred to mycorrhizal species, highlighting further the positive effects of grazers on plant productivity and mycorrhizal development (Lussenhop, 1992). Variation in the susceptibility of major fungal functional groups can lead to shifts in fungal community composition. Collembola grazing, for example, destabilised the saprotrophic fungal community, making it more susceptible to the negative effects of the competing arbuscular mycorrhizal species, Glomus mosseae (Tiunov and Scheu, 2005). Whereas changes in invertebrate communities are likely to have direct implications for saprotrophic fungi, the effects on mycorrhizal systems are likely to be more indirect.

Conclusions

There has been a growing interest in the role of soil biodiversity on terrestrial ecosystem processes (for example, soil nutrient availability and decomposition) and the related ecosystem services (primary productivity and soil carbon storage) they provide. Two trends pervading this review may help predict how invertebrate community structure and diversity affect saprotrophic fungal activity and functioning. These are underpinned by the fact that the biological and functional responses of saprotrophic fungi to grazing are intrinsically linked; increased mycelial growth is generally coupled with increased enzyme production and respiration, and vice versa. First, the density-dependent impacts of grazing suggest that factors contributing to an increase in soil invertebrate abundance are likely to reduce the biomass of saprotrophic mycelial systems. This would decrease fungal-mediated nutrient translocation, mineralisation and CO2 efflux into the atmosphere. In contrast, reducing soil faunal abundance is likely to stimulate mycelial respiration and soil nutrient availability. Second, taxon-specific effects suggest that macrofauna exert the strongest grazing pressures on mycelial systems. Predicted changes in the relative abundance of earthworms, isopods and millipedes as a result of climate and land use change (David and Handa, 2010) will be of particular importance because, even at low densities, these grazers have a dominant role in regulating mycelial activity. It may, as in aboveground systems, be functional group representation, rather than invertebrate species diversity, per se, that determines mycelial functioning in soil (Gessner et al., 2010). The capacity of macrofauna to prevent the exclusion of less competitive fungi further highlights their importance in regulating microbial community composition and diversity which, in turn, affect woodland nutrient cycling.

In isolation, these functional consequences of grazing interactions are predictable, but these effects may be complicated by a myriad of other belowground interactions. As well as decomposing organic matter and increasing inorganic nutrient availability for plants, saprotrophic fungi also compete with plants and their mycorrhizal symbionts for these nutrients. Low-intensity grazing can stimulate fungal-mediated nutrient mineralisation with positive effects on plant growth (Bardgett and Chan, 1999), but it could also increase saprotrophic mycelial growth and nutrient uptake, eventually limiting primary productivity. Furthermore, understanding the top–down consequences of grazing on fungal growth and activity requires an appreciation of the concurrent bottom-up effects; invertebrate population responses to fungal resources will affect grazing intensities. Although detailed, microcosm-based manipulations have provided a good mechanistic understanding of these belowground interactions, this acquired knowledge must now be applied to more natural, multi-species environments to allow accurate predictions regarding the role of soil biodiversity on terrestrial ecosystem functioning.