The biology of mycorrhizas is a function partially of the wide diversity of fungi, partially of the wide diversity of plants involved, and partially the wide diversity of soils where plant and fungus interact, but also by the various morphologies these associations take. Simplistically, these associations between plant and fungus are symbiotic and almost entirely mutualistic in that both fungus and plant partners benefit [1]. The benefits seem predominately connected to improved nutrition of the host and the infecting agent, but extend to attenuation of hormonal balance, physical protection, chemical protection and modification of other rhizosphere organisms that impact competition for substrates (Table 1). The association, however, is tenuously balanced. It is easy to say that these mycorrhizal associations are mutualistic and not parasitic, but words do not capture easily the concept that there is a gradient from the mutualistic to the parasitic. Environmental conditions fluctuate enough that any given association moves back and forth along the mutualism-parasitism continuum. So it is likely that the benefits to the partners are not only quantitatively unequal but also qualitatively unbalanced: one partner gaining more than the other. In other words, one partner leans more towards parasitism than the other. One of the ironies of these associations is that it is not always clear, even in a general sense, which partner leans towards the parasitic and which partner is therefore forced towards being parasitized. Our human experience with pathogenic fungi would lead us to accept the notion that the fungi would tend towards the parasitic habit—the plant being parasitized. There is evidence, however, that one of the largest groups of fungi, those that form the arbuscular mycorrhizas, are heavily controlled by their plant partners. These plants lean towards parasitism of the fungi [2].

One of the most significant events in the successful colonization of land by plants was the evolution of biotrophic, root-inhabiting symbioses [3-5]. Mycorrhizas (“fungus-roots”) are symbiotic associations between specialized soil fungi and plants, and represent one of the many specialized members of micro-organisms that inhabit the rhizosphere. These include N-fixing symbioses including rhizobia and Actinomycetes and other specialized bacteria such as mycorrhiza helper bacteria [6-9]. Functionally, this group of fungi has evolved along with their plant-hosts (Table 2) and display a variety of ways in which they interact with both readily and poorly available nutrient resources. Such interaction often results in enhanced plant growth [1]. Mycorrhizal fungi have been shown to play a role in the dissolution of parent rock in more established soil, and the mycorrhizal association with plant roots provides a physical bridge—involved in the absorption and delivery of nutrients from the soil matrix—to the plant-host via the mycelial network [1]. This classical, symbiotic description of the mycorrhizal association is completed with the obligatory return of carbon from the plant-host to the fungus—in a completely balanced manner. However, each type of mycorrhiza is likely to have its own characteristic function, e.g., for achlorphyllous plant-hosts, this delivery mechanism can also involve the acquisition of carbon-resources via mycorrhizal-haustorial connections from adjacent autotrophic plants.

For the purposes of this review: mycorrhizal fungi are dual soil-plant inhabitants; a mycorrhizal association is the relationship between the fungus and the plant-host; and mycorrhizal diversity encompasses the entire range of form and functions—ranging along the functional continuum from symbiotic, commensalistic, mutualistic all the way through to exploitive/parasitism [34]. Herein the discussion of parasitism will be limited to those circumstances where the normal mycorrhizal functionality has turned pathogenic. Propagules for this review refer to any material, sexual spores, asexual spores, hyphal fragments, infected root fragments, etc., that are capable of providing infection or colonization of an uninfected or uncolonized root resulting in a mycorrhizal association. Inoculum is nearly a synonym for propagules, but implies an anthropogenic overlay—propagules cultivated, enhanced or used by humans.

Currently, structural mycorrhizal diversity is categorized into seven types of equal taxonomic rank: arbuscular (AM), orchid mycorrhizas (OM), ericoid mycorrhizas (EM), ecto- (ECM), ectendo-(ECTENDO), arbutoid (ARBM), and monotropoid (MM) [13,35] (Table 3). Arbuscular mycorrhizas have turned out to be much more diverse in structural features than previously thought (e.g., [36-38]). There is much structural homology exhibited among ecto-, ectendo-, arbutoid and monotropoid mycorrhizas [2] and together they comprise a distinct ECM lineage. Imhof [38] proposes OM as a third, distinct lineage from AM and ECM. Smith and Read [35] provide an in-depth treatment of the mycorrhizal symbiosis.

In brief, Smith and Read [35] indicate that AM describe the association formed by members of the Glomeromycota. This is the most ancient of mycorrhizal types, with fossils from the Devonian containing both arbuscules and vesicles [23]. Fungi in this group have not been shown to be culturable on defined media and it is assumed that they are wholly dependent on the photosynthetic plant. In ECM, the fungus forms a structure called a mantle (or sheath) which encloses the rootlet. From it hyphae or rhizomorphs radiate out into the substrate. Hyphae also penetrate inwards of the root to form a complex intercellular system, which appears microscopically in cross section as a network of hyphae, termed the Hartig net [35]. The ECTENDO are similar in structural nature to ECM, except that the sheath may be reduced or absent, and hyphae penetrate into the cells of the plant. The distinction between these types is confounded in that the same species of fungus (fungi include members of all orders of Basidiomycetes and Ascomycetes) may form ECM on one species of plant and ECTENDO on others [35].

Heath plants (family: Ericaceae) are uniquely hosts to ericoid mycorrhizas [40]. In many autotrophic members of the Ericaceae and related families, the hair-like roots are enmeshed in an extensive weft of hyphae, which also penetrate the cells of the root—normally, no sheath is formed (cf. ECM) [35]. The fungi currently identified as forming EM are Ascomycetes [35] and Basidiomycetes [41]. Arbutoid mycorrhizas possess sheath, external hyphae and usually well-developed Hartig net. Arbutoid mycorrhizas are formed, in the main, by autotrophic trees and shrubs, although some of the plants, such as Pyrola, are herbs and are partially achlorophyllous [35]. The closely related Monotropoideae are all achlorophyllous (herbaceous sub-family of the Ericaceae). The first accurate descriptions of the fungal partners of monotropes were provided by Martin [42]— again, septate-fungi belonging to orders of Basidiomycetes and Ascomycetes. Monotropoid mycorrhizas have a well-developed fungal sheath, and Hartig net. They also possess a highly specialized haustorium-like structure (the fungal peg) which penetrates the epidermal cells and goes through a developmental cycle of its own as the plant grows and achieves anthesis. Arbutoid and monotropoid mycorrhizas are variations of the ECM-type and the fungi associated can also form ECM on neighbouring, autotrophic plants, and hemi-parasitism of organic carbon is assumed [35].

Orchids can be either wholly or partially achlorophyllous for some part of their life cycle. They form mycorrhizas with Basidiomycetes of various affinities [35]. The division between orchids that are green for part of their lives and those that are wholly achlorophyllous is mirrored by the identities of their fungal associates—the fungal symbionts of green orchids are saprophytes (e.g., Rhizoctonia spp.) although known perfect stages are all Basidiomycetes [43]. The mycorrhizal fungi of achlorophyllous orchids are likely to form ECM on autotrophic hosts constituting a biological bridge between them and the carbon-seeking orchids. There is considerable detail known about the relationship of orchids to their fungal symbionts; however, relationships have not been wholly established such that general concepts can be articulated [35]. The impressive diversity of the single family Orchidaceae is matched only by the Asteraceae and the Poaceae [44]. Although largely tropical and subtropical in their habitat, of the estimated 700 genera and more than 25,000 species worldwide, there are members found in all terrestrial habitats except at the poles [45]. Most orchids are autotrophic, rarely saprotropic, but are not known to be parasitic on other plants. Almost half of all genera that contain achlorophyllous mycotrophic plants are in the Orchidaceae. Orchids are commonly terrestrial, lithophytic and epiphytic; they are rarely semi-aquatic or subterranean [43].

Continued consideration of OM and mycorrhizal associations will provide a useful construct for the interpretation of newly discovered mycorrhizal associations in both autotrophic plant-hosts (e.g., Monotropa) and achlorophyllous (e.g., Corallorhiza) plant-hosts—where the function of the association shifts from the mutualistic end of the symbiotic spectrum (via fungal nutrient acquisition for the host—C for the fungus), through to the exploitative/epiparasitic (via sequestration of C—from neighbouring, autotrophic plants). Although orchids represent a vibrant and diverse group of plants, there are species, especially at/near the limits of their range, which are subject to perturbations. In North America, the western prairie fringed orchid is a legally defined threatened and endangered species [46].

There are ongoing discussions around the definitions of mycorrhizas, mycorrhizal associations and hence, mycorrhizal diversity—especially with the emergence of the application of molecular-based techniques (see [38,47,48]). For the purposes of this review, we use mycorrhizal diversity in the broadest sense—encompassing the structural, form and functional diversity of plant-fungal root symbioses types or categories. Within each of the mycorrhizal types (taxonomic categories), there exists further levels of diversity. This takes into account multiple lineages of fungi that encompass diversity of nutrient uptake (e.g., phosphorus, zinc, sulphur, and nitrogen), other essentials (e.g., water), production of sequestering agents, production of antibiotics, enhancement of soil structure, etc.

Population and community regulation can result from either promotion or reduction in the growth, fitness or reproductive potential of an organism. If the fitness of one organism in the community is altered to a greater or lesser extent than another, the result is a changed dominance of the favoured species in the community that occurs over successive generations [1]. Mycorrhiza influence the growth and fitness of the plant-host, through the extramatrical hyphal network providing access to a larger pool of nutrients—either mineral (e.g., phosphorus) or from organic sources (e.g., humus) or autotrophic plants. As such, the actual effect of the mycorrhizal association above the cost of the maintenance of the association is dependent on the rate of growth of the extraradical hyphae of the fungal species [49]. Mycorrhizas also improve plant-water relations [50,51], and reduce pathogenic infections. Mycorrhizas also contribute to plant C-allocation [52]. However, there is evidence that mycorrhizas mediate uptake of other non-nutrients and even toxic materials [53]. Alternatively there is evidence that AM fungi increase plant tolerance to heavy metals [54].

Because mycorrhizae influence plant fitness, the resulting enhanced plant nutrition may result in increased biomass production—which could be translated into improved reproductive success. Thus, the association can influence interspecific plant competition dynamics by preferentially improving recruitment. There is also evidence that the association may result in the uptake of metabolically active organics [55] or provide mechanical and chemical defenses against plant pathogens [56,57]. This in turn can influence plant demographic responses, community structure, and ultimately, the successional dynamics of some plant communities [58].

The major ecosystem function of mycorrhizas is to assist plant-hosts in the acquisition of resources (nutrients, water, C) from soil. Read [59,60] put forward a hypothesis that the dominant type of mycorrhiza in an ecosystem was related to the soil conditions and the nature of the major form of nutrient from which the plant community derived its nutrition. Read showed the geographical distribution of the main mycorrhizal types in the world as follows:

AM habit was dominant in the temperate and tropical grasslands, tropical forests, and desert communities

In ecosystems dominated by graminoids, AM fungi are much more prevalent than other categories of mycorrhizal fungi. Physical disturbance of these grasslands has been shown to depress activity of these fungi [61]. Alpine and tundra grasslands also have a substantial community that is dominated by AM plants, although there are shrubs or stunted trees in the alpine grasslands that have ECM associations. And while the dark septate root endophytes (DSE, Table 3) in a range of plant taxa from Polar regions are not as yet considered to be true mycorrhizal symbionts [62], recent studies suggest some benefits to plant growth (under certain conditions).

The importance of mycorrhizal contribution in forested ecosystems was shown by Vogt et al. [63], who demonstrated that the percentage of net primary production represented by mycorrhizal fungi was 14–15% (with up to 45% in young forest stands and 75% in mature forest stands). ECM have been shown to produce N-degrading protease enzymes and P-solubilizing acid phosphatase enzymes—enabling them access to forest-floor nutrient pools [64] (see also [65]).

Weather conditions are major drivers of bacterial and fungal infectious diseases of emerging plants [66]. Environmental stresses such as drought or nutrient deficiency can predispose plants to diseases. The ecological impacts of non-indigenous invasive fungi as forest pathogens are increasing in global frequency [67]. Efficacious ECM associations are considered to convey some level of host resistance to tree hosts [56]. Therefore continuity in the ECM association will continue to provide ecosystem-resilience from pathogenic as well as other potentially invasive mycorrhizal fungi [68].

Numerous species of forest-dwelling small mammals rely on the fruiting bodies of ECM as a primary source of food [69]. Others too rely on hypogeous fungi, including Australian mammals— Marsupialia and Eutheria alike [70], while primates have also been shown to be mycophagous (e.g., [71]). Malajczuk et al. [72] considered that the morphology of some ECM increased their dependence on animal mycophagy for spore dispersal—as hypogenous fungi remain below ground and do not release spores into the air [73]. The lack of mammal-dispersed hypogenous ECM is considered to hinder Pinaceae invasions into previously unsuitable habitats [74]. Insects and other invertebrates are known to ingest spore materials of mycorrhizal fungi [75] and there is evidence these are important in the dispersal of mycorrhizal fungi [76,77].

The mycorrhizal fungi forming associations with ericaceous plant communities are capable of producing enzymes enabling the plant-host to access organic forms of nutrients. In addition to the direct nutritional benefits of EM colonization, the ability of the fungus to sequester, and in some cases metabolize, metal ions that are otherwise toxic appears to be of importance [35].

There are several scenarios that emerge when considering mycorrhizal association in the context of climate change. The central theme of this paper is to present those scenarios that are the most likely in the context of proposed climate change models. Central to examination of these possibilities is addressing the following fundamental questions: (1) What are the effects of environmental drivers on mycorrhizal plant-hosts? (2) What are the effects of environmental drivers on mycorrhizal fungi? How do these changes to mycorrhizal fungi affect mycorrhizal association with their plant-partner? What will happen to mycorrhizal diversity as a consequence of any de-coupling of mycorrhizal associations? (5) Will changes to mycorrhizal fungi in terrestrial soil-vegetation systems influence plant, community, ecosystem, and biospheric/global processes?

To address these questions, we pose four hypotheses. Addressing these is fundamental to addressing the resiliency questions in the last section.

The response of mycorrhizal associations to perturbations in their environment in the past is a predictor of how they will react to perturbations in the future. Global climate change (GCC) is driven by several fundamental ecologically important drivers [79]:

It is a central premise of biogeography that the broad-scale distribution of terrestrial ecosystem complexes is determined largely by climate and, as such, can be altered by climatic changes due to natural and human-mediated activities [164]. Climate is also a major determinant for the phenology, physiology, distribution and interactions of plants (determined largely by temperature range, quality and quantity of light, seasonal water availability, humidity, barometric pressure and the generation of wind, and extent and intensity of erosion of parent material). The strong interrelationship between climate, soils and plants is a basis for anticipating substantial changes in natural terrestrial biomes in response to GCC (Figure 2).

Most soil taxonomies rely on measurement of abiotic properties for categorization. In Jenny's [306] state factors, the organismic factor is generally ignored or minimized perhaps because it is hard to measure and not well understood. Soils represent a continuous tension between abiotic factors (e.g., topography, geology and some of climate) and biotic factors (organisms, biochemicals and the rest of climate). This tension is integrated across time such that at any instant soils are manifest with characteristics which in themselves are complex. A conceptual simplification (and probably over-simplification) is to place soil in a spectrum that ranges from the completely abiotic at one end to the completely biotic at the other. The completely abiotic end could be little more than physically weathered rock (e.g., by freezing and thawing, wetting drying, etc.) but with no significant biological components. The completely biotic end for all practical purposes does not exist, but imagine a site with a diversity of plants, plant roots, arthropods, fungi, etc., and with deep, rapidly decomposing litter and high levels of organic carbon in the upper soil horizons (rich, black, earthy). With characteristics of abiotic and biotic parameters, any given soil can be placed appropriately along this continuum: some where abiotic parameters dominate; others where biotic factors dominate.

The focus here is to place soils into a GCC context as ameliorators of GCC and as habitat for mycorrhizal associations. A soil taxonomy has been chosen to explain how mycorrhizal forms and functions are attenuated by overarching soil characteristics and how these soil characteristics stratify mycorrhizal forms and function in a maxtrix of GCC.

The taxonomy chosen for this is the US Taxonomy. The US Soil Taxonomy [307] is translatable into some other taxonomies (e.g., the New Zealand Taxonomy [308]; The FAO system [309], the World Reference Base of Soil Resources for soil correlation [310], the French Taxonomy [311] andThe Russian Taxonomy [312]). Fundamental soil taxa of the US Soil Taxonomy [307] (especially at the Order level) have characteristics that suggest that they fit naturally and easily in nature. However, the separation between one order and another is often artificial. For example, the US Soil Taxonomy separates grassland soils (Mollisols) from all other soils by the amount of organic matter in the surface soil and that these soils are fertile. That this is defined precisely as “a soil having at least 1% soil organic matter (or 0.58% organic carbon) in the surface soil when mixed to 18 cm, and having a base saturated of at least 50%” illustrates the artificial demarcation.

In the weathering of soils are several trends among mycorrhizal associations. First is that slightly weathered or embryonic soils (Entisols and Inceptisols) can be generally occupied by a full spectrum of mycorrhizal associations. Where Entisols and Inceptisols occur in deserts—a common occurrence— these systems are dominated by AM associations to the probably complete exclusion of ECM associations. Staffeldt and Vogt [314] found desert plants in New Mexico (USA) to be infected with AM fungi. These included grasses, shrubs, forbs, cacti and yuccas. In Entisols from central Wyoming (USA) Singleton and Williams ([315]—Torriorthents) found exclusively AM associations on clayey and sandy substrates. Others have also observed AM associations in desert environments on a diversity of plants growing in Entisols and have not observed ECM mycorrhizas ([300,316]— Torrifluvents); ([53]—Torriorthents, Ustifluvents and Torrifluvents). Klopatek et al. [317] showed that AM fungi dominated in an Inceptisol under Pinus edulis and Juniperus osteosperma woodland although they recognized these systems had ECM components too (on Pinus).

On soils in alpine environments and derived from crystalline and calcareous parent materials (probably Entisols and Inceptisols, respectively) Lesica and Antibus [318] found plant species (grasses, forbs and shrubs) to be associated with AM fungi, but they found ECM fungal associates as well (e.g., with Dryas and Polygonium spp.). Haselwandter [319] confirms this mixture of AM and ECM plants in alpine areas of Europe and points out that alpine areas may also have abundant members of the Ericaceae and thus ericoid mycorrhizas (EM). Williams et al. [320] confirm mixtures of AM and ECM plants in alpine environments on alpine turf soils (Inceptisols-[Cryumbrepts]).

A highly studied soil type is that derived from dune sand. Often these have little or no horizon development but still classify as sandy Entisols [Psamments]. Active dunes appear to be colonized, sometimes heavily, by AM fungi only [111,266,321-324]. Jehne and Thompson [325] made the same observation but noted the presence of ECM associations in stabilized dune sites. Kurtböke et al. [111] described that hind-dune vegetation was characterized and dominated by the presence of a Casuarina species with the bipartite symbiotic complex comprised of AM and actinomycte-facilitated N-fixation [326].

In the Alfisols (Figure 4) the order divides into those soils dominated by grasses thus principally occupied by AM associations and those dominated by forest trees thus principally occupied by ECM associations. The AM occupy soils that on further development become Mollisols and Vertisols (both having near-neutral to alkaline pH) where these fungal associations generally dominate. The vegetation on these soil-types is mostly grass and forbs. Alfisols that have been grasslands and then converted to intensive agriculture continue to be occupied by AM associations [327]. Forested Alfisols, however, have mostly occupancy by ECM [328-332].

In undisturbed grasslands (Mollisols), AM mycorrhizal associations dominate to the near exclusion of other mycorrhizal associations (e.g., [333-336]). In farmed Mollisols, soils remain dominated by AM associations even on non-native plants [337]. Plants normally associated with ECM fungi do so only when appropriately inoculated [338].

Presumably Vertisols would be occupied by AM mycorrhizas. Because of the shrinking and swelling driven pedoturbation, woody plants would not survive on these soils, but only herbaceous and graminoids. Dry Vertisols that do not shrink and swell regularly would be too dry and likely too alkaline to support plants occupied by ECM fungi.

The ECM fungal associations develop and become dominant in those soils that are more leached and become more acid: Ultisols, Spodosols. The vegetation on these soils is dominated by conifers and deciduous forests, at least historically. Ultisols are characteristically dominated by ECM associations on conifers and other hosts (e.g., Quercus spp.) under undisturbed conditions [339]. Ultisols are often farmed, however, and undergo transformation to systems where AM fungi are present [340,341].

Spodosols are noted by several authors as hosting primarily ectomycorrhizal systems ([98]—Abies balsamea, [342]—Pinus radiata and Pinus elliottii, [343]). Hayman and Mosse [344] show that Spodosols in the field could support AM associations when original vegetation was removed, replaced with Trifolium repens, fertilized and inoculated with an AM fungus. Haselwandter [345] has shown that ericaceous plants growing on subalpine and alpine Spodosols are EM.

Aridisols generally evolve from Entisols but have characteristics of other soil orders too, but with the caveat that they are dry. Almost universally they have high accumulation of salts and considerable levels of high-exchange-capacity clays. Plants in this soil order have almost all AM associations (e.g., [315,292]—Haplargids, [316]—Haplargids, Calciorthids & Comborthids]).

Oxisols are highly weathered soils not known outside of the tropics or subtropics. They may evolve from the Mollisol-Vertisolic soil branch (Figure 4), but are more likely to evolve from the Ultisol-Spodosolic branch. These soils may be occupied by a diversity of mycorrhizal associations including AM, ECM and others. When forested, these soils are occupied by associations that are AM, ECM, ECTENDO and presumably OM [346]. Thomazina [346] looked at plants from more than 23 families including the Orchidaceae. When disturbed and planted to crops, these soils eventually become inhabited mostly by AM fungi in part because these soils are characteristically phosphorus deficient and crop plants do not survive well without AM associations [347-349].

Andisols, as noted, develop in a diversity of climates and can have a full component of mycorrhizal associations. Waters et al. [350] report ECM associations on mature Abies stands in Northern California. James et al. [351] report similar from Andisols in Scotland.

Histisols as noted are mostly too wet for abundant occupancy by any mycorrhizal associations although there are exceptions. Indeed, all of the other orders (except Aridisols) have wet or aquic phases (designated at the suborder or great group levels) where mycorrhizal associations are depauperate because of excessive wetness. Even the Gelisols that have histic properties are also characterized by a general lack of mycorrhizal fungi. Theorodou [352] showed five ectomycorrhizal fungi and their host, Pinus radiata, were intolerant of high water contents. In his 1974 review, Slankis [99] reported that excessive water prevented formation of ECM on hosts. The Gelisols that have inceptic properties are likely much more diverse in their mycorrhizal associations. Linkins and Antibus [353] report ECM associations of Salix rotundifolia on Gelisols (polygonal soils with 2 cm of surface organic matter) in the Alaskan tundra. Christie and Nicolson [354] report AM fungi on two angiosperms (one a grass) were infected with AM fungi on Gelisols on the Faulkland Islands (51°S) and the South Georgia Islands (54°S) but were uninfected on similar soils on the South Orkney Islands (60°S), the South Shetland Islands (64°S) and the Antarctic Peninsula (64–68°S).

For AM fungi, in their review, Sylvia and Williams [355] indicated these fungi were intolerant of high soil-water contents where oxygen was restricted and CO₂ elevated (i.e., anaerobic conditions). AM fungi, however, have been reported on aquatic plants/vegetation complexes where aqueous conditions were aerobic (review by Sylvia and Williams [355]). At the dry end of the spectrum, numerous studies show AM fungi infect plants at very low water potentials and may substantially aid plants in uptake of water under drought conditions (reviewed [355]).

Lodge [356] showed that Salix and Populus species, genera that can be both ECM and AM infected, were both dominated by AM infection under dry soils. For Populus plants, they were equally infected by AM and ECM under very wet conditions. For Salix, however, in very wet soils plants were dominated by AM associations and in moist soils dominated by ECM associations. Haselwandter [319] points out that plants in alpine environments in nivation hollows are not mycorrhizal presumably due to the excessively wet conditions due to high snow accumulation and persistent snow melt wetness. In summary, wetness and acidity controls the type of mycorrhizal association that will develop on a given soil. Where soil orders are not distinctly acid or wetness is variable or dry, these drive the type of mycorrhizal association that can colonize and result in occupancy of wider mycorrhiza diversity.

A soil feature that also determines mycorrhizal development is buffering capacity. In this review, buffering capacity is the capacity to resist changes in pH and for this definition captures important characteristics of climate change. Atmospheric pollutants are often manifest as deposition of NOx and SOx, both of which in the presence of water in soils will result in a tendency towards lowering pH. Increased partial pressure of CO₂ in the atmosphere results in generation of carbonates as carbonic acid. Increased average temperatures will increase decomposition, which in turn releases more CO₂, but also generates more organic acids in soils. Buffering capacity is germane to issues of GCC since so many processes in GCC result in generation of acids. The summarial result is that many GCC processes contribute to the decrease of soil pH—and in turn, low pH may be attenuated by buffering capacity. In systems where buffering capacity is high (Aridisols, Mollisols, Vertisols) the input of atmospheric acids will be neutralized for often a very long time. Soils with low buffering capacity (Entisols, Inceptisols, Ultisols, Spodosols, Oxisols) will be much more impacted by acid inputs. Acid inputs will accentuate the depression pH in these soils and thus depress biological activity. It follows that those soil taxa characterized typically with moderate buffering capacity (Alfisols, Histisols, and some Andisols) will be moderately impacted by atmospheric inputs of acid formers.

Several authors indicate that ECM associations are particularly sensitive to inputs of nitrogen (e.g., [165,259,357]). When then N comes from the atmosphere, it is mostly as the NO₃ anion. When it is as ammonium (NH₄⁺cation) then this is often converted by prokaryotes to NO₃⁻ with additional release of acidic protons (H⁺). Whether the overall negative impact of increased N inputs on the ECM system is due to imbalance of nutrients (ECM fungi dislike high available N) or to increased acidity is not well worked out, but it is clear this is detrimental to ECM associations. Some ECM associations are highly tolerant of or even thrive in very low pH systems. The ECM fungus Pisolithus tinctorius is among those fungi most tolerant of acidity, with three New Zealand species only co-occurring in geothermal areas [358]. However, this species appears to be tolerant mostly of acidity derived from sulphur (e.g., sulfate) and not from nitrogen (nitrate).

A fundamental observation is that ectomycorrhizas generally occupy soils of fairly low pH and high accumulations of organic surface litter. Some of these fungi are adapted even to very low pH (e.g., P. tinctorius), but many are intolerant of further decreases in pH and almost universally they are intolerant of high pH and high salts. In a 1974 review, Slankis [99] showed that most ECM species in pure culture had maximum growth between pH of 4-6, some were capable of growth at pH as low as 2.7, but none grew at pH above 7.

Arbuscular mycorrhiza fungi were shown to exist in soils at a wide range of pH [359] from 4.5 up to 8.5, but there was considerable species-specificity for particular pH optima. Green and Graham [360] showed very similar results in their 1976 report on the germination of AM spores. Sylvia and Williams [355] in their review cite numerous reports that support the contention that “AM fungi do not readily adapt to soils with a pH different from their soil of origin… However, there are AM fungi that are adapted to a wide range of soil pHs”.

The typic soil taxa at the order level of the US Soil Taxonomy [307] can be grouped according to their buffering capacity. Those with low buffering capacity are the Entisols, Inceptisols, Spodosols, Ultisols, Oxisols and Gelisols. Those with intermediate buffering capacity are the forested Alfisols, Andisols and Histosols. Those with high buffering capacity are the Aridisols, grassy Alfisols, Mollisols, and Vertisols. Further, those taxa with low buffering capacity tend to have low pHs especially those that are more weathered. Those taxa with high buffering capacity tend towards higher pHs. Those with intermediate buffering capacity have near neutral or variable pH.

As each of the various soil orders of the US Soil Taxonomy are examined in detail, there are many examples that violate the above principles of buffering capacity and pH. For that matter, there are many that violate the principles articulated in the last section regarding mycorrhizal occupancy. These chemical principles as well as the mycorrhizal principles reported herein are those that are most likely, according to sweeps made of the literature that correlated with the central tendencies of the 12 order-level taxa of US Soil Taxonomy. Clearly these principles must be viewed as hypotheses but the literature we have examined fails to support the contrary view (i.e., the null hypothesis).

Several other principles can be drawn from the above. In the forested and related soil-taxa, or those that are unweathered or only slightly weathered, ECM associations are common if not dominant. These are all of the taxa with low pH and low buffering capacity. Certainly other associations can be found too. OM are sometimes common, MM can be present, in the understorey of many forests EM can be found and of course the elusive ECTENDO may be present. However, ECM dominate and are associated with the dominant and deeper rooted species. When these soils are disturbed—e.g., forest removed and soils ploughed or forests removed via fire that deeply heats the soil—often AM associations will become established on such sites. Where such soils are converted to intensive crop agriculture, often these systems are dominated by AM associations to the exclusion of ECM and other mycorrhizas. Those soils of low buffering capacity and moderate to low pH are also susceptible to changes in pH. Clearly in such situations where soil pH drops, those ECM that are tolerant or adapted to a low pH environment will survive and perhaps proliferate. Those that are not so adapted or tolerant will be selected against. Hosts of fungal associations may also have intolerance or tolerance to depressed pH and may or may not provide an appropriate host for ECM fungi. Further, increases in average temperature may enhance decomposition of soil organics. This increased decomposition can be further enhanced by the input of inorganic nitrogen. Often nitrogen is limiting to decomposition in high organic carbon systems. Input of nitrate, even though an acid former, will accelerate organic matter decomposition which, among other things, produces organic acids that have the potential to further exacerbate decomposition. Recent work by Eisenlord and Zak [102] present an argument that in part supports this view, but they also make the point that increased soil organic matter accumulation could result.

In those soil taxa that have high buffering capacity and which are often neutral to relatively high pH, AM associations dominate. There are situations where other mycorrhizas can exist and even persist in these soils. OM may exist as well as EM and perhaps a few MM. However, ECM and ECTENDO are for all intents absent. Certainly plants can be inoculated and introduced into such systems [338], but this is rare and artificial. Further, when these high buffering capacity, high pH systems are disturbed, ECM do not invade these soils [316]. They remain as AM-dominated systems. Also, these systems are fairly immune to changes in pH. To summarize then, ECM occupy soils that are susceptible to increasingly lowering of pH. Further, these mycorrhizas do not invade soils of high pH and high buffering capacity. Conversely, AM fungi dominate in systems having much higher pH and buffering capacity. These systems are often immune or nearly so from changes in pH derived from atmospheric deposition. Further, as low-buffering-capacity soils are disturbed, perhaps even disturbed by radical changes in soil pH, the AM associations have the capacity to move into these systems and form associations with the exotic/invasive opportunistic plant-hosts that are either resistant to the change, or have the capacity to quickly invade available niches. Many invasive species seem, however, to be non-mycorrhizal or facultatively mycorrhizal—but, there is emerging evidence to suggest that the situation is both species and context dependent. Sanon et al. [361], in their study of the invasive, non-mycorrhizal tree Gmelina arborea, observed that AM colonization was significantly and positively correlated with plant diversity (i.e., in the absence of the invasive tree, diversity was higher). While, the work of Greipsson and Di Tommaso [362] demonstrated that exotic weeds had the ability to form mycorrhizal associations with the native suite of AM fungi—their results suggest that invasion into new areas by non-native plant species may alter the occurrence of AM fungi in the resident soil. Furthermore, the exotic plants were shown to be able to selectively recruit a particular AM fungal partner from the native AM fungal suite, causing a shift in mycorrhizal diversity by preferentially allowing the chosen AM fungal partner to rapidly proliferate. The outcome of this selective inoculum's amplification resulted in improved performance of the weed-species compared to the non-mycorrhizal counterpart—so the weed actually gained a benefit from the association with the indigenous AM fungi. Pringle et al. [363] summarize the suite of interactions succinctly: several criteria can determine the potential for mycorrhizal symbioses to constrain or facilitate the invasion process, particularly whether an introduced plant requires the mycorrhizal association, the invasive plant's degree of flexibility in associating with a range of the indigenous species, and whether or not suitable mycorrhizas “co-invade” with the invasive plant (or arrive independently). Following their successful establishment, invasive plants are influenced by (and can influence) their own (and neighbouring) mycorrhizal symbioses, and these can further influence the trajectory of the invasion, with flow-on effects which can impact community integrity and resilience to future incursions.

One of the consistently challenging conundrums that have emerged through this review process is that mycorrhizal diversity ranges across many spatial scales—and within the rhizosphere of a given plant-host there exist opportunities for further levels of diversity (e.g., [364]). This variation in scales works its way up from the microscopic-level through to influencing biospheric process—and furthermore, there is the real potential for reverse feed-back loops, from the level of biosphere back down, to influence the physiology of the individual mycorrhizal hypha (see Figures 1 & 2). A further layer of complexity is delivered when we consider these changes across a temporal scale—in keeping with the seasonal rhythms that influence plant and fungal phenology (especially leaf and root flush of plants and the hyphal extension and sporulation of fungi). Swarts and Dixon [124], in their paper on integrated approaches to orchid conservation, provide a temporal continuum from immediate impacts (1–2 years) to long-term impacts (20–50 years) in relation to the sequence of abiotic and biotic factors that have an impact upon terrestrial orchid populations. Figure 5a reproduces their derived timescale, and while this eloquently articulates the impacts on orchids and orchid mycorrhizas, we consider that a putative timescale for GCC impacts could be generated from this basis—to incorporate all mycorrhizal types, and hence, mycorrhizal diversity.

We also consider there to be a compelling nexus between the timescale articulated by Swarts and Dixon [124] regarding orchid conservation and the identified ecological drivers that are associated with GCC (as per [79]) listed below:

Increasing atmospheric levels and associated climatic changes

Global temperature rise

We have pictorially amalgamated the timescale of Swarts and Dixon [124] with the ecological drivers of Tylianakis et al. [79]. In Figure 5b we have retained the immediate and long-term time references, but have also added, short-term (3–10 years) and intermediate-term (11–20 years) time references, based on our interpolation of the time frames for the ecological drivers of GCC on mycorrhizal diversity (Figure 5b). The individual ecological drivers have been partitioned into four putative timescales as follows:

Immediate (1–2 years) impacts associated with ecosystem fragmentation and habitat loss will be immediately realized through the removal of plant-hosts and significant disturbance of the soil/rhizosphere habitat of mycorrhizas (Figure 5c)

The above- and below-ground impacts of chronic/catastrophic ecosystem modification that results in permanent land-use change will potentially result in:

The loss of geographically restricted plant taxa

Because of their dependence upon their host for C-supply, orchid mycorrhizas and all heterotrophic mycorrhizal groups will be immediately impacted through ecosystem fragmentation leading to loss of habitat and, more importantly, their plant-hosts (Figure 5d). Orchid mycorrhizas are described as “delicately balanced mutual antagonisms” [43,44]. Further, plant propagules (spore-like seeds) are often difficult to germinate and undergo a saprotropic phase prior to development of mycorrhizal forms. Although orchids and their mycorrhizal symbionts have evolved to occupy a remarkably diverse set of habitats [43,44], most of these habitats are of low to moderate pH and low buffering capacity and as a consequence are susceptible to further acidification.

Transient changes in the availability of some classes of mycorrhizal inoculum have not been seen to permanently alter the infective ability of AM mycorrhizas (e.g., [301]). This is because some of the propagules (such as infected root pieces) are quite robust and can remain for some time as a viable source of inoculum. Reinvasions by soil mesofauna also provide a mechanism for AM inoculum to be returned to a site. Klironomos and Moutouglis [365] demonstrated that soil arthropods such as collembola can act as effective dispersal agents. More recently, Perez-Morena and Read [366] have demonstrated a pathway for the transfer of nutrients from dead nematodes to mycorrhizal plants—thus demonstrating the capacity for mycorrhizal fungi to recycle nutrients, even more effectively than from saprotrophic fungi alone.

However, habitat modification leading to the displacement of larger, mycophagous mammals has been implicated in the loss in the dispersal pathway for ECM fungi [72]. Indeed, Pyare and Longland [367] concluded that the integrity of entire small-mammal communities may be important to the maintenance of ECM diversity in coniferous forests. However, like AM fungi described above, the loss of infective propagules may only be transient. In his study on the effects of fire on Tasmanian bettongs (Bettongia gaimardi), Johnson [128] demonstrated that all measured variables returned to control values about four months after fire. This follows the findings of Bellgard et al. [295], whereby the ability of the flora to resprout from underground organs provides for a stable habitat and carbohydrate supply for the mycorrhizal partners to survive the effects of wildfire, and continue their below-ground functions.

The invasion of native ecosystems by adventive incursions of weeds, pests and plant diseases will potentially result in:

Invasive mycorrhizal fungi displacing native symbionts to the detriment of native/indigenous plant-hosts (e.g., [368])

We consider that ECM fungi will be the principal group of mycorrhizas that will be subject to short-term impacts (Figure 5d). Dickie et al. [370] consider that the process is facilitated by coinvasion of plants and their mycorrhizal counterparts. Their example of Pinus contorta invading endemic communities of Nothofagus solandri var. cliffortioides also demonstrates the potential for exotic mycorrhizal fungi to displace native fungal species and drive the mycorrhizal community to a more simplified version of the native complement. The preceding example of course raises the role of human-mediated introductions that result from the global trade in horticultural and agricultural plant species. These effectively remove plants from their native context and place them into non-indigenous regions. The new arrivals are placed into situations where their usual associates are not present—this release from the usual suite of herbivores and pathogens may explain their success in the introduced range.

Hansen [285] considers that many Phytophthora species are benign in their native, coevolved plant communities, but given the opportunity of introduction to new hosts in new environments, new opportunities for dispersal, or unexpected sexual recombination, can result in dramatic epidemics in forests around the world. Phytophthora ramorum (cause of sudden oak death in western North America and also damaging in Europe) provides a current example, dramatically illustrating the potential of these pathogens for rapid ecological damage. The consequence of removal of native mycorrhizal plant-hosts will be a concomitant decrease in their abundance. While it is unlikely that these invasive plant pathogens will push any of their major susceptible hosts to final range-wide extinctions, death of the tree hosts creates opportunities for invasions by exotic plants and/or exotic mycorrhizal fungi. The additive effects of disease and biotic invasions may be enough to realize local endemic plant extinctions, and the concomitant loss of their ecosystem services, including hosts of ECM fungi.

Alteration to the nutritional balance of soils through the enrichment of N, decrease in soil pH (e.g., through enrichment with NOx and SOx) and addition of environmental pollutants will potentially result in:

Changes in insect herbivory patterns

Empirical evidence suggests that there are several mechanisms by which N deposition may affect interactions between plants and insect herbivores. The most likely mechanisms are deposition-induced shifts in the quality and availability of plant-host tissues [371]. In New Mexico and Arizona, there have been massive die-offs of the Piñon pine (Pinus edulis) due proximally to bark beetles (Coleoptera: Curculionidae, Scolytinae). The empirical study of Jones et al. [372] demonstrated that tree mortality increased 8% and beetle activity increased 20% under the highest rates of nitrogen. The other factors involved in this complex problem such as ozone (as an atmospheric pollutant) and drought will be discussed in the following sections, but for our current purposes, the summarial removal of keystone plant species from forest ecosystems will have a concomitant impact upon the distribution and diversity of ECM fungi.

Associated with increases in grassland N-economies was a reduction in the below-ground allocation of C to arbuscules, coils and extraradical hyphae [136]. Concomitant impacts to the species composition of the AM fungal community was also observed—probably associated with the allocation plasticity within individual fungal taxa [136]. From her earlier study, Johnson [373] had indicated that 8 years of fertilization altered the species composition of AM fungal communities—there was an observed reduction in species diversity and a disproportionate increase of a single Glomus species. From her controlled greenhouse study, she demonstrated that fertilization of nutrient-depauperate soils favours the proliferation of inferior AM fungal mutualists. Additionally, the community-level shifts in AM species abundance may be an indirect response to increasing fungal competition for a more limited supply of host photosynthate.

Conversely, however, in a long-time, conventionally tilled agricultural soil in Czechia, a synergistic effect of organic and mineral fertilization on the development of AM was demonstrated by Gryndler et al. [374]. The length of AM mycelium and sporulation both increased in proportion with increasing mineral fertilization. However, correlation analysis showed the dependence of length of AM hyphae and sporulation on soil available P. The observed differences could therefore be related to the original fertility status of the experimental field-soil studied in Czechia. If, as suggested by Johnson [373], the suite of AM fungi had already been altered through the long fertilization history, selection of inferior mutualists may have already occurred. Thus, the observed extraradical stimulation may not have had any biological significance with respect to plant performance.

Arnebrant [260] demonstrated that all five of the different ECM fungal isolates she examined were negatively affected by nitrogen amendments, although sensitivity to amendments varied between the isolates. The extramatrical mycelial growth was inhibited in all species examined—although the percentage inhibition varied between 3 and 80%. This suggests that some ECM species may be more tolerant to N amendments and thus could have some selective advantage in an N-enriched soil environment. This selection could in turn shift the ECM community composition, if the N-sensitive species were not able to produce a competitive quantity of extramatrical hyphal growth to compete with the more N-resistant strains.

Roth and Fahey [375] demonstrated a shift in the composition of the ECM community that occurred in response to acid precipitation treatments. Among the seven ECM types studied, the percentage composition of one morphotype increased and another decreased in response to higher rain acidity. Nitrogen and S depositions are usually connected with each other in the field, and it is often impossible to distinguish their effects. Large concentrations of ammonium in rainwater [376] or gaseous ammonia [377] have been found to cause a decline in the amount of mycorrhiza in coniferous seedlings. Markkola et al. [258] demonstrated a reduction in sporophore production of ECM fungi at high N sites. As stated by Sårstad and Jenssen [378] the high niche differentiation and high species diversity of fungi in forest ecosystems means that marginal differences in environmental factors can result in substantial shifts in species composition within an area. However, it has been observed that “generalist” ECM species that form a symbiosis with a wide range of species seem to less affected by increased N availability than “specialist” ECM species, especially those living in symbioses with conifers [379]. Frey et al. [380] identified that the active fungal biomass was 27–61% and 42–69% lower in the N-fertilized compared with control plots in hardwood and pine stands, respectively. In the pine stand, ECM fungal community diversity was lower in the N-treated plot than in the control plot. Finally, differences were also observed in ECM community structure between control and N-fertilized plots—with a reduction in those species with the highest relative frequencies in the control community.

Edaphic inputs of SOx that effectively reduce soil pH can liberate toxic metals such as aluminium, manganese and magnesium [165], and hence make them more plant available. This increased toxicity leads to reduced root growth, root dieback and reduced mycorrhizal fungal growth and colonization. Dighton and Skeffington [357] found that simulated acid rain caused a change in the ECM community structure by reducing the occurrence of mycorrhizal morphotypes that were multibranched and that were associated with large amounts of extraradical hyphae. In Finland, Markkola and Ohtonen [381] demonstrated that Piloderma, Dermocybe and Hebeloma were significantly reduced in the presence of acidifying pollutants, whereas Cenococcum was seen to increase in frequency. Once again, there was the identification of differential susceptibility to soil acidification—akin to the differences in responses to amendments of N. This demonstrates the potential for soil acidification to select ECM that are more tolerant to decreases in soil pH. If this change in frequency is great enough to affect the individual's reproductive fitness, then we will see a permanent shift in the community diversity of ECM.

Pollutant-induced reductions in the photosynthetic capacity of tree canopies potentially reduces the allocation of C to roots and their mycorrhizas [1]. Reduced energy supply both reduces the overall mycorrhizal colonization of roots and favours those fungal species that can survive on low carbohydrate supplies [165]. Stankevičienė and Pečiulytė [259] demonstrated that the lowest diversity of ECM morphotypes was associated with forests characterized by the highest concentrations of heavy metals and mineral nutrients (N and P). Interestingly, microfungal abundance was negatively correlated with the distance from the pollution source; meanwhile the number of genera (i.e., diversity) increased with distance from the pollution source. The favoured microfungi included the phytopathogenic species of the genera Cladosporium spp. and the entomopathogenic fungus Paecilomyces farinosus. Other pollution-resistant microfungal taxa included Mucor hiemalis and Trichoderma viridies—both fungi known to improve the mobility of heavy metals in polluted soil [382]. This indicates that the shift in microfungal diversity favours the proliferation of pollution-resistant species, some which can further exacerbate the toxic nature of the soil—potentially excluding the return of metal-sensitive species (including ECM fungi).

Thus, vulnerable mycorrhizal types subject to intermediate-term GCC changes include “specialist” ECM species associated with forest ecosystems and AM associated with grassland ecosystems (Figure 5d). AM have repeatedly been demonstrated to alleviate heavy metal stress of plants. Associated with plants under heavy stress of either Cd, Cu, Zn was the response of specific AM fungal genes in the intra- and extra-radical mycelium [383]. This “level” of responsiveness was interesting, as “heavy metal-derived oxidative stress may be the primary concern of the fungal partner in the symbiosis”. The level of pre-adaptation to tolerance to heavy metal stress would vary—thus presenting a selective pressure on the AM fungal community.

Alteration of rainfall distributions will increase the frequency of droughts, which will cause plant stress. Coupled with this will be temperature increase in response to increased CO₂ concentration. The combined influence of these factors will result in:

Increased frequency of drought facilitating pathogen invasions in forests (e.g., [68])

Drought-induced predisposition, rendering plants and trees susceptible to pathogenic infection, has long been recognized (e.g., [228]). However, drought-induced tree mortality is difficult to predict, because it is a non-linear, threshold process [388]. The severe drought and heat wave experienced in Europe during the summer of 2003 raised concern, especially if warmer temperatures raised the effects of drought on tree mortality—both for background rates of mortality and for regional die-off events [68]. Considerable circumstantial evidence is available linking drought and disease in forest ecosystems in Europe, North America and Australia—with the principal agents of disease being canker/dieback pathogens (e.g., Botryosphaeria) and root disease (e.g., Phytophthora). The relationship between the severity of the environmental stress, prevalence of the pathogen, and the severity of disease is articulated via the “disease triangle”—with the well-being/susceptibility of the host being a critical factor in understanding the extent of damage to the host by the disease [389].

Direct drought effects on pathogens are generally negative, and potentially also affect mycorrhizal efficacy. The results of Valdés et al. [390] demonstrated that both total fine-root biomass and ECM-root biomass are strongly affected by severe drought in high elevation, tropical plantation forests. The comprehensive review of water relations, drought and AM symbiosis of Augé [180] highlights the following facts on the influence of AM fungi and drought physiology:

AM effects on plant water relations are not as dramatic and consistent as those on P acquisition and host growth

Seasonal and drought-related declines in the extraradical hyphae of AM fungi and percent root colonized have been observed in many, but not all studies [263]. Commonly, the decline in exploitation of soil has been associated with a decline in soil water content. Increased frequency and duration of drought episodes will impact negatively on AM plants through the concomitant reduction in access to essential plant nutrients (in the water phase). However, compensatory effects associated with increased CO₂ enrichment may offset any short-term losses; however, sustained invasion pressures from non-mycorrhizal, exotic plants may be at some advantage if the efficacy of AM association is reduced through reduced soil exploitation.

Interactions between Phytophthora infections, flooding and droughts are well known and generally accepted as triggers of epidemics (e.g., [391]). However, conjecture remains on the interaction between water availability and disease expression. This is especially relevant for oomycete root-rot pathogens (such as Phytophthora), whereby sufficient soil water is essential to facilitate the germination and sporulation of dormant propagules (e.g., oospores). Permanent changes to rainfall distributions could potentially increase the frequency of episodic rainfall events leading to local waterlogging. For example, Phytophthora-mediated oak decline in Central Europe and parts of Western and Northern Europe has been attributed to long-term changes in climatic conditions in the second half of the 20th century (i.e., a continuous increase in mean winter temperatures, a seasonal shift of precipitation from summer into winter-time, and a tendency towards episodic, heavy rainfall events) [392]. Indeed, waterlogging or impeded drainage has been shown to be sufficient to kill trees in the absence of root pathogens. Davison [393] investigated the interaction between waterlogging alone and in combination with infection by P. cinnamomi. She articulated that there were indeed a complex of disorders: deaths of mid- and understorey species caused by P. cinnamomi, deaths of groups of Eucalyptus marginata (jarrah) caused by waterlogging, a background mortality of isolated jarrah deaths, and a crown decline of jarrah—on some sites all problems occurred simultaneously (see also [394]). Thus through either a predisposition and/or a multiple stress hypothesis, decline of keystone species in forest ecosystems will impact on the viability and ultimately the diversity of mycorrhizas.

In our previous discussion, bark beetle impacts on forest health were linked to N-amendments. Another particular concern is the potential for warmer temperatures to compound the effects of increasingly severe droughts by triggering widespread vegetation shifts via woody plant mortality [385]. Adams et al. [385] provide compelling evidence that warmer temperatures shorten the time to drought-induced mortality in Piñon pine trees by nearly a third, with temperature-dependent differences in cumulative respiration costs implicating carbon starvation (whereby trees keep stomata closed to maintain safe levels of xylem pressure—but stopping most photosynthesis), as the primary mechanisms for mortality. Furthermore, projected increase in drought frequency due to changes in precipitation and increase in stress from biotic agents (e.g., bark beetles) would further exacerbate tree mortality. The death of tree hosts would impact upon the fitness of the associated ECM—with broad-scale removal of a keystone species, effectively disrupting the host-mycorrhiza relationship: thereby reducing mycorrhizal diversity.

Cullings et al. [138] showed that by experimentally defoliating Pinus contorta (lodgepole pine) while maintaining the canopy of Engelemann spruce (Picea engelmannii) the ectomycorrhizal species composition responded accordingly. The premature defoliation of Pinus contorta changed the relative abundance of ECM of the two tree species from a ratio of approximately 6:1 without treatment to a near 1:1 ratio post-treatment. In addition, ECM species composition changed significantly post-defoliation; the system dominant, an Inocybe species, was rare in defoliation plots, while Agaricoid and Suilloid species that were rare in controls were dominant in treatments. Furthermore, species of ECM fungi associating with both lodgepole pine and Engelmann spruce were affected, which indicates that changing the photosynthetic capacity of one species can affect mycorrhizal associations of neighbouring non-defoliated trees.

Indeed, vegetation shifts have profound ecological impacts as they provide important climate-ecosystem feedback through their alteration of carbon, water and energy exchanges of the land surface [385]. Mycorrhizal diversity will respond to temperature changes. Invasion by warm-season plant-hosts will change mycorrhizal functioning. This will likely be accompanied by invasions of warm-season-adapted ECM fungi [395]. Wolf et al. [299] showed that CO₂ enrichment and plant species richness impact AM fungal spore communities. Under plant monocultures, only Glomus clarum responded significantly to CO₂ elevation out of 11 species present. In contrast, under polyculture, this response was not detectable. In a similar way, Chung et al. [386] showed that higher plant species richness fostered greater microbial biomass as well as the abundance of saprophytic and AM fungi. Moreover, the effect of plant species richness on microbial communities was significantly modified by elevated CO₂ and N deposition. For instance, microbial biomass and fungal diversity increased with greater species richness, but only under combinations of elevated CO₂ and ambient N, or ambient CO₂ and N deposition.

Further to this, Johnson [137] found that plant species responded individualistically to the presence of AM fungi and the availability of CO₂. Starting with mesocosms of 14 plant species, ambient or elevated CO₂ and low or enriched N factorial treatments were investigated. After 1 year, significantly, different plant communities arose under different treatment conditions. Plant species richness was lowest at plus N mesocosms, and highest in plus AM and plus CO₂ treatments. At ambient CO₂, AM fungi reduced plant richness but at elevated CO₂ they increased it. This was caused by preferential mortality of several C3 forbs and may suggest that CO₂ enrichment ameliorates the carbon cost of some AM symbioses.

Many studies have demonstrated plant responses to warming temperatures, but as advancement in the timing of phenological events and in range shifts. Mountain gradients have provided a very powerful opportunity for studying species range changes (e.g., [396])—and furthermore, these particular contextual settings are appropriate, as GCC-induced changes to vegetation are expected to be accentuated at high latitudes and altitudes (e.g., [87]). Dullinger et al. [397] considers that invasion of abandoned subalpine pastures and tundra grasslands will be facilitated by invasive, aggressive conifers (in this case, Pinus mugo Turra). It is well accepted that the degree of resistance to invasion of grasslands by tree and shrub species is strongly affected by competition with vigorous grass and herb species (e.g., [398]). However, the concomitant supply of propagules (i.e., seed from P. mugo) and the depressed resistance to invasibility (due to invasion pressure from a highly competitive Carex sp.) of the “host communities” will see shifts into these grassland communities by trees.

A curious, complex interaction is emerging in another high altitude/latitudinal complex associated with the forest-steppe ecotone of Mongolia. In this contextual setting, drought and high temperatures at the soil-air interface are crucial for inhibiting tree encroachment. For example, the southern distribution of taiga forest dominated by the keystone species Larix sibirica, although modified by human activities (e.g., logging, livestock breeding and wildfire), is still largely determined by seasonal drought (e.g., [399]). Treelines towards the steppe are generally not static—but underlie spatial variation with temporarily varying influences of climate, fire activity, herbivore densities, and human utilization [400]. However, interannual variation in insect herbivore pressure is actuating another layer of selective pressure to restrict tree establishment. The sowing and planting experiments of Dulamsuren et al. [399] demonstrated that drought exerts a strong “post-zygotic” pressure on seedling mortality. However, 2-year-old seedlings of L. sibirica suffered from drought as well as from feeding pressure by Gypsy moth larvae (Lymantria dispar), early in the growing season. Follow-up herbivory pressure from grasshoppers and rodents persisted throughout the growing season, resulting in two-thirds of seedlings dying due to insect and small mammal herbivory and one-third due to drought-related damage. Hauck et al. [400] identified that herbivory pressure was greatly exaggerated at the forest edge, and that while defoliation of trees growing at the forest edge was severe in the first year of their study, the gypsy moth invasion responsible for this defoliation did not cause lethal damage in the trees. However, sub-lethal damage in seedlings can be exacerbated through the added impact of waves of herbivory pressure that can cause a spatial shift in the treeline—with a concomitant impact upon the mycorrhizal complement.

Olsrud et al. [401] examined the how EM, fine edophytes and DSE in roots responded to elevated atmospheric CO₂ concentrations and warming in the dwarf understorey of a birch forest in the subarctic region of northern Sweden. The EM colonization in ericaceous dwarf shrubs increased under elevated CO₂ concentrations, but did not respond to warming following 6 years of treatment. It was suggested that the higher EM colonization under elevated CO₂ might be due to increased transport of carbon below-ground to acquire limiting resources such as N, which was found to be diluted in leaves of ericaceous plants under enhanced CO₂. Conversely, the elevated CO₂ did not affect total plant cover but the plant cover was increased under warming, which may reflect increased N availability in the soil. Fine endophyte colonization decreased in grass roots under enhanced CO₂ and under warming—possibly due to an increase in root growth. DSE hyphal colonization in grass roots significantly increased under warmer conditions, but did not respond to elevated CO₂. For this reason we have placed EM* in the 3–10 year, short-term temporal scenario depicted in Figure 5d. This may only be relevant at high latitudes and altitudes, but it seems that the three-way interaction identified by Olsrud et al. [401] could be a powerful predictor of the possible outcomes of competitive interactions between mycorrhizal fungi and other root endophytes.

Conjecture remains as to the plausible function of DSE—clearly DSE form symbiotic associations with roots of their plant-host; the fungal symbionts inhabit the absorbing organs of the host forming the dual organ as described by Trappe [402]. Questions remain as to the function of this dual organ; whether it is an organ of absorption and whether DSE associations can be considered facultatively mutualistic, with occasional instances or parasitism [403-405]. Regardless of function, it is obvious, because of their location in the grass-root rhizosphere and their ability to respond to GCC drivers, that they may be capable of competitive exclusion—to the detriment of symbiotic fungi (e.g., AM). The picture of total “environmental DNA” generated by Vandenkoornhuyse et al. [406] demonstrated that the roots of the grass Arrhenatherum elatius contained 49 different sequences (so called “phylotypes”). Of the 40 different phylotypes, all were found to be fungal! From this analysis, it was clear that among the phylotypes were relatives of mutualistic AM fungi. However, the authors could only speculate on the possible roles of the other 94% of the root fungal diversity that they documented. Is it possible one or more of these phylotypes could usurp the root-space to the competitive exclusion of mutualistc partners?

This type of analysis sheds light on the unknown spectrum of micro-fungal diversity of fungi that are potentially in competition with mycorrhizal endophytes for host-root tissue. The GCC-driven changes affecting soil microhabitats could be seen to favour certain non-mycorrhizal components of the soil microflora. As described in our above example, Olsrud et al. [401] found that soil warming favoured DSE but did not enhance EM colonization. What are the consequences of the competitive exclusion of EM fungi in this circumstance? A reduction in the efficacy of the EM association could in turn alter the plant community structure (potentially from shrubs to grasses). The fertilization effect of enhanced N (thought to be a consequence of GCC along with soil warming) may change the level of host dependence upon mycorrhizal endophytes. This in turn may continue to favour rapidly colonizing, aggressive AM-dominated monocots. With the concomitant shift in mycorrhizal diversity from EM-dominated ericaceous shrubs to AM-dominated grasslands will be changes to soil biology, which in turn will potentially alter the biological component involved in the adhesion of soil particles to form peds (i.e., structure) and, ultimately, influence the biogeochemistry of subarctic soil ecosystems [401].

The increase in CO₂ has been associated with an increase in root biomass and an increase in the C:N ratio, possibly due to a change in the balance between allocation of C to root growth and carbon storage [1]. Mycorrhizal tissue has been estimated to comprise a significant fraction of soil organic matter and below-ground biomass in a range of systems. The current literature suggests that in an enriched CO₂ environment, mycorrhizal fungi might sequester increased amounts of C in living, dead and residual biomass in the soil [247,407]. The concomitant stressors being applied from increased N additions, drought frequency and biotic invasions of non-mycorrhizal plants could see dis-function and a “resource-based” dislocation of the C-fixing machinery delivered through mycorrhizal diversity. Important to recognize is that below ground, mycorrhizal plants shifted allocation of carbon to pools that are rapidly turned over, primarily to fine roots and fungal hyphae [407], and host root and fungal respiration. Mycorrhizae alter the size of below-ground carbon pools, the quality and, therefore, the retention time of carbon below ground. The data of Nottingham et al. [387] indicate that if elevated atmospheric CO₂ and altered climate stressors alter mycorrhizal colonization in forests, the role of forests in sequestering carbon could be altered. Further to this, changes in ECM assemblages have been shown to occur in response to increased CO₂ (e.g., [408,409]). Arguments of Eisenlord and Zak [102] suggest that increase in exogenous N will shut off the certain carbon degradations of Basidiomycetes. This function is taken over by other lignin-degrading organisms that do not degrade lignin to CO₂, but rather to polyphenols.

Both CO₂ enrichment and mycorrhizae can indirectly influence soil moisture via effects on stomatal conductance (e.g., [410]). Soil moisture can increase under elevated atmospheric CO₂ because of the commonly observed reduction in stomatal conductance, particularly in C3 plant species [411]. The increased soil moisture generated by elevated CO₂ has been shown to increase rates of microbial metabolism and nutrient cycling. Stomatal conductance and transpiration rates are generally higher in AM compared with non-AM plants in both mesic and drought conditions [180]. Increased N availability typically stimulates above-ground plant productivity, stomatal conductance and whole-plant water use. Because elevated CO₂, AM fungi and N availability have contrasting effects on plant growth, water use, and soil moisture, the interactive effects may have unexpected effects on plant community structure [137] (see also [412]). Superimposed upon this are the potentiality of biotic incursions, and other stressors threatening biodiversity (including abiotic factors).

Once there are major perturbations to the global hydrological cycle that permanently change rainfall patterns and distributions, no aspect of the global mycorrhizal diversity will remain unaffected (Figure 5d). Why do we consider that these modified interactions affecting mycorrhizal diversity are so important for plant performance?

Because presence and/or absence of mycorrhizal symbionts can affect distribution and abundance of plant species, which in turn influences evolution of the host and plant community response.

GCC is not a recent phenomenon, but has played a principal role in historical evolution of the diversity of life on planet Earth at least over the last 500 m.y. The episodic punctuation of the continuity of life on Earth has, in the main, been associated with significant GCC events (e.g., expansions of ice-sheets). Fungi emerged in the Proterozoic and so have witnessed the suite of six major extinction events and other GCC drivers. Historically, fungi (including ancestral mycorrihzal fungi) were mining nutrients from primitive, undeveloped soils. All primitive plants that successfully colonized land required some form of fungal partnership, because their roots were not capable of accessing the necessary essential plant nutrients and water. This continued until such time as some root systems became sophisticated enough (and/or soils became sufficiently developed) for plants to “go it alone” without a fungal partner.

The source and extent of mycorrhizal diversity was due, in part, to GCC. There have been five episodes when two-thirds cumulative of all living species were removed from the planet. And while this represented the end for some, these events represented opportunities for mycorrhizal fungi to diversify their host range—repeatedly. Thus mycorrhizal diversity is the result of an accumulation of [plant-host] “associative-memory”, reinforced by “new encounters” with plant species that relied upon a successful mycorrhizal association in order to become established in the post-apocalyptic edaphic substrates.

Pankhow et al. [413] suggested that the main role of the mycorrhizal symbiosis is not during the early productive stages of plant succession in ecosystems, but rather in the protective stages during which most resources are entrained in plant biomass. As such, perturbations to this association that influence the efficacy/benefits to plants will affect plant health. Poor plant health can predispose them to diseases as mycorrhizae play an important role in protecting plant-hosts against pathogens [35,414]. Thus mycorrhizas are essential components contributing to plant health and, hence, contribute to the resilience of ecosystems to perturbations such as climate change, invasive species incursions and pollution. But is this significant in relation to the trajectory of plant communities through the influence of autecological patterns?

Kohut [415] propounds that genetic selection occurs when an entity (in this case mycorrhizal fungi) is subject to an agent of selection (in this case GCC) and three conditions for a property (in this mycorrhizal diversity) responsive to the agent are satisfied at the population level. Firstly, there must be variation in the population. Secondly, part of the variation is genetically controlled, and finally, variation in the property must affect reproductive fitness. If these conditions are satisfied, the frequency distribution of the property, and gene frequency associated with it, will change over time in response to selection.

It is obvious that variation exists at a range of scales within mycorrhizal fungi within populations— across all types or categories of mycorrhizas. This variation is governed by a complex array of genetic-based processes—and even though the balance of the genetic systems varies across the mycorrhizal types (i.e., they do not all follow the classical, panmictic, Mendelian-based construct), they are all genetically-based. And, it is obvious that there are direct physiological influences of GCC, and indirect influences of GCC-mediated changes to plant-hosts that affect the reproductive fitness of mycorrhizal fungi—their physiology, biology and ecology. These factors dictate that there will be changes in relative gene frequencies that will be realized as changes in the species composition at the community-level.

And because of the above, the GCC-induced changes which influence mycorrhizal fungal fitness will principally influence mycorrhizal diversity through:

Ecosystem fragmentation/habitat loss, which can spatially de-couple the mycorrhizal fungi from their plant-host. Habitat fragmentation will lead to loss of rare taxa; this will influence mycorrhizal diversity by selection against highly-specialized, co-evolved “clades” of fungi which form intimately-linked, mutually-dependent partnerships.

The schema extracted from Staddon et al. [166] (Figure 2) usefully articulates the two-way arrows introduced in Figure 1 between the various hierarchical levels that exist in complex ecosystems. This schema articulates some of the levels of interactions and demonstrates the various feedback mechanisms at a simple scale of a three-species system. We have demonstrated that changes in mycorrhizal diversity will occur at the level of:

species

The relationship between mycorrhizas and plant productivity will continue to influence the Earth's atmospheric composition through the following escalation of scales commencing from the individual fungal hypha upwards to the global biomic-level via the following set of interrelationships:

Mycorrhizal fungi form absorptive accessories to roots, linking them to the individual fitness of their plant-host.

Thus changes in mycorrhizal diversity will have consequence at the biospheric-level and thus can either contribute to ameliorating the long-term effects of GCC (e.g., through C-storage, or facilitating the revegetation and hence restoration of ecosystem processes) or facilitate the GCC drivers (e.g., through release of C). GCC was also instrumental in achieving a habitable planet, 100 billion years ago. Fungi (including the ancestral mycorrhizas) contributed to making the planet habitable for terrestrial plants, which, in turn, facilitated the change in oxygen concentration necessary to allow more sophisticated levels of macro-evolution [33]. This ultimately led to the evolution of humans. The irony is that there is little support of the null hypothesis that humans over the last 10,000 years have NOT facilitated the commencement of the sixth major extinction and further, there is little support of the null that through their contributions to GCC now do not threaten the diversity of mycorrhizas.

Changes in geographical distribution of plant species along with their associated mycorrhizal fungi has been a long-term and slow change as plants have colonized new areas since the early Devonian. Plants species have continued to change in distribution as suitable climatic and edaphic conditions that allow them to successfully compete with established plant species become available. Climatic, atmospheric and edaphic changes associated with the environmental consequences of GCC will allow plants to invade marginal habitats as conditions become suitable). The newly established plants will facilitate the establishment of their associated mycorrhizal partners to access key plant nutrients. As a consequence, the plants and their associates that are eased out of formerly optimal habitats (either through GCC, habitat fragmentation or human-mediated biotic invasions) will be those that are dependent on highly evolved associations with mycorrhizal fungi (e.g., Orchidaceae and other achlorphyllous plants). This includes low-buffering-capacity ecosystems on highly weathered or recalcitrant substrates (e.g., alpine ecosystems on crystalline substrates) most highly leached forested systems, many highly weather systems in the humid tropics and others (see section 4). Additionally, biodiversity hotspots, which once covered 12% of the land surface but today are restricted to 1.4%, seem especially conducive to local or even general botanic extirpation because the majority of this loss is recent, and many of the species in these areas are already geographically restricted and threatened with extinction (see [416]).

It is a legitimate criticism that the hypotheses that initiate this review are simplistic and indeed they are. However, we have used them in part to ground this review to the point that we can make some general projections. This is fraught with problems. The laboratory in which this review attends is the entire terrestrial World. The ecosystems specifically are not specific but are all ecosystems. The mycorrhizal fungi reflect, and likely more than we know, this massive diversity of soils and plants as they interact across landscapes and are impacted by a climate that gives evidence of changing physically and chemically. Still, there are the hypotheses and here we attempt to address them directly and with as little ambiguity as we can muster. There is caution here also to be exercised. The authors of this document are humbled by the vast diversity that interplays on this planet, and of course especially by the diversity of the symbioses we review in this document. To be sure, we are aware that there is much we do not know. The danger lies in that readers will take what we have found here as absolute truth. This attitude could not be further from our intentions and violates the methodology of science [417]. In our field, there are no absolutes. We deal with probability, risk and complexity. Still, basically, what do we claim to have uncovered? Here we rely directly on examination of null hypotheses. Succinctly, these are as follows:

Null Hypothesis: Mycorrhizal diversity did not evolve in response to changes in environmental drivers. We find lack of evidence to support the null hypothesis (see especially Table 2).

Although we state our hypotheses or predictions above, there are other well-known and some recent reviews that to some degree confirm these, to some degree refute and likewise do neither. We have worked the findings of these reviewers into the discussion throughout our above analyses and discussion. However, readers will find these more comprehensive summaries enlightening and perhaps justification for continued research: Dighton and Jansen [165], Molina et al. [161], Bledsoe [418], Miller and Jastrow [419], Thomas et al. [420], Blankinship and Hungate [421], and Rillig [422].