Global changes can impact both plants and fungi directly, or indirectly via modifications to the symbiosis (Rillig et al. 2002; Singh et al. 2010). For example, it is generally assumed that atmospheric CO2 enrichment directly influences plants and indirectly influences fungi though increased photosynthate (Drigo et al. 2008; Fig. 1). In contrast, changes to biota and N eutrophication directly impact both plants and fungi. Understanding both individual and interactive effects of environmental changes on plants and fungi are necessary to predict the net effects of global changes on mycorrhizal symbioses.
Microbial feedbacks have been insinuated in community and ecosystem responses to global change drivers, but we currently lack sufficient knowledge to predict the relative importance, or even the direction, of these feedbacks (Singh et al. 2010). The principles of optimal allocation, biotic context and adaptability may provide a useful platform to begin to construct and test hypotheses about the mechanisms of mycorrhizal responses to global change factors and the outcomes of these responses for communities and ecosystems.
Enrichment of atmospheric CO2 and N
In the past century, human activities have increased the concentration of CO2 in the atmosphere from ~250 ppm to 385 ppm. Although there are exceptions, particularly in N-limited systems (e.g. Parrent & Vilgalys 2007), CO2 enrichment generally increases the biomass of roots and mycorrhizal fungi (Rillig et al. 2002; Drigo et al. 2008). This finding is in accord with Principle 1: if enriched atmospheric CO2 increases the production of photosynthate such that plant demand for soil nutrients increases, then allocation belowground to mycorrhizas is expected to increase.
Concurrently with CO2 enrichment, humans have more than doubled the rate of N input into the biosphere (Vitousek et al. 1997). This has been linked to changes in the species composition of communities and accelerated loss of plant diversity (Clark & Tilman 2008), and mycorrhizal fungal diversity (Lilleskov et al. 2002; Liu et al. 2012). Mycorrhizas exhibit variable responses to N enrichment, and fungal biomass has been reported to decrease, increase or remain unchanged (reviewed in Rillig et al. 2002). Nevertheless, these variable responses to N enrichment generally conform to the predictions of optimal allocation in Principle 1. A comparison of five long-term field experiments demonstrated that N enrichment caused AM colonization of roots, spore biovolume and density of extraradial hyphae (outside the root) to decrease at sites with ample soil P, and increase at a site with limited levels of soil P (Johnson et al. 2003a). We expect this pattern if plants allocate energy to structures needed to acquire the most limiting resource: adding N to P-rich soil will diminish the relative value of roots and mycorrhizas, while N enrichment of a P-limited soil will exacerbate P limitation and enhance the value of the symbioses.
There is increasing recognition of the importance of synergistic effects when multiple anthropogenic changes occur. A free air CO2 enrichment (FACE) experiment in Switzerland showed that after 10 years of treatment, the CO2 × N interaction had a significant effect on AM fungi, even though there was not a significant main effect of CO2 (Staddon et al. 2004). Interestingly, Staddon and colleagues found that when N was in limited supply (but not enriched), the formation of AM colonisation inside plant roots decreased in response to enrichment of CO2, while the density of extraradical hyphae increased. This conforms with the expectation of optimal allocation because when N is more limited than C, there should be increased growth of extraradical hyphae to forage for N. In contrast, the BioCON FACE experiment in Minnesota USA showed that CO2 enrichment significantly increased the density of extraradical hyphae, but there was no interaction with N supply (Antoninka et al. 2011). There are many possible reasons for the different findings at the two FACE experiments, including differences in the soil properties, climate and the composition of the plant and fungal communities. So while the optimal allocation principle can help predict the general direction of future shifts, differences among communities and ecosystems introduce serious variation.
The potential for mitigating rising CO2 levels though increased allocation and storage in mycorrhizas and other belowground C pools is currently debated (Kowalchuk 2012; Phillips et al. 2012; Verbruggen et al. 2013). Recent studies suggest that when atmospheric CO2 levels are elevated, mycorrhizas may actually speed up the decomposition of soil organic matter, and thus generate net C-sources rather than C-sinks (Cheng et al. 2012; Phillips et al. 2012). Even though Glomeromycota are not saprotrophic themselves, under elevated CO2, AM fungi appear to have a priming effect on bacteria and other organisms in the mycorrhizosphere such that decomposition of organic matter is accelerated. Priming appears to stimulate saprotrophic bacteria to release N which is rapidly assimilated by both fungi and plants in AM symbioses (Cheng et al. 2012). Notably, the degree to which AM fungi enhanced decomposition can vary among taxa. When grown under elevated CO2, Gigaspora margarita and Glomus clarum enhanced decomposition much more strongly than Acaulospora morrowiae (Cheng et al. 2012). This pattern is intriguing because a different research team showed that CO2 enrichment generated a very rapid shift in symbiotic partners inside plant roots from Acaulospora to Glomus (Drigo et al. 2010). This study highlights the need for future research to identify functional groups of AM fungi that may be distinguished according to their enzymatic capabilities and competitive dominance under different resource environments.
There is some evidence that altering the abiotic environment (i.e. nutrient availability) can also lead to genetic changes in AM fungi (Ehinger et al. 2009), and these changes may be linked to differences in symbiotic functioning. Adaptation of AM fungi in response to abiotic selection pressures, Principle 3, has been demonstrated previously, with researchers finding that a gradual shift in CO2 levels over 21 generations provided time for adaptation to occur and generated a much different mycorrhizal response compared to a sudden CO2 increase during a single generation (Klironomos et al. 2005). Similarly, a study at the Swiss FACE site that compared the nutritional response of white clover to colonisation by Glomus isolates derived from elevated or ambient CO2 showed that within just 8 years of CO2 treatment, the functioning of Glomus isolates diverged such that isolates from plots treated with elevated CO2 improved the N nutrition of their host plants significantly more than those in plots treated with ambient CO2 (Gamper et al. 2005; Fig. 7a). Furthermore, a larger proportion of N was derived from soil pools rather than through N-fixation (Fig. 7b). These results are in accord with the recent discoveries that CO2 enrichment generates a strong selection pressure for enhanced N uptake capacity (Cheng et al. 2012); and, that geographical isolates of AM fungi differ in their abilities to acquire N from the soil (Johnson et al. 2010; Ji et al. 2013). Does the heterokaryotic characteristic of AM fungal clones allow for rapid genetic and functional changes when fungi are exposed to changing environmental conditions (Fig. 6c, d)? More research is needed to understand these types of rapid functional changes in AM fungi.
Figure 7. Rapid selection for efficient nutrient uptake and transfer by AM fungi can influence host plant performance in changing environments. (a) Concentration of foliar N and (b) proportion of total foliar N acquired from the soil in Trifolium repens grown with AM fungi isolated from field plots treated for 8 years with either ambient (open bars) or elevated (black bars) CO2. Data courtesy of Gamper et al. (2005), (n = 7), mean ± SE * significant difference at P ≤ 0.05.
Download figure to PowerPoint
Research suggests that ectomycorrhizal and AM communities respond similarly to CO2 enrichment. In both type of mycorrhizas, CO2 enrichment has been shown to accelerate decomposition of soil organic matter through priming (Carney et al. 2007; Phillips et al. 2012). In ectomycorrhizal systems, the fungi, in addition to associated saprotrophic bacteria, have the enzymatic capability to break down complex organic compounds (Buée et al. 2007; Drake et al. 2011). Fourteen years of CO2 enrichment of loblolly pine at the Duke FACE site in North Carolina USA enhanced decomposition rates because of more rapid turnover of roots and EM structures in addition to a priming effect on old soil organic matter (Phillips et al. 2012). Likewise, 6 years of CO2 enrichment of a scrub-oak ecosystem in Florida USA increased production of phenol oxidase (Carney et al. 2007; Fig. 8a), a key enzyme in the decomposition of lignin and other recalcitrant organic materials; and it also increased the abundance of fungi relative to bacteria (Fig. 8b).
Figure 8. Enzymatic activity and composition of soil microbes is influenced by 6 years of CO2 enrichment of a scrub-oak ecosystem. (a) Phenol oxidase and (b) fungus:bacteria ratios were significantly higher under elevated CO2 compared to ambient. Means ± SE (n = 8 for ambient, and n = 6 for elevated CO2). Data courtesy of Carney et al. 2007.
Download figure to PowerPoint
The principles of optimal allocation, biotic context and adaptability can help explain the apparent lack of an enhanced mycorrhizal C-sink under elevated CO2. Principle 1, optimal allocation, accounts for the higher N demand in response to increased availability of labile C belowground (Hu et al. 2001). Principle 2, biotic context, predicts that fungi may become competitively dominant in CO2 enriched systems because they can better translocate nutrients and have a lower C : N ratio than bacteria. Consequently, N limitation induced by CO2 enrichment generally favours fungi over bacteria (Hu et al. 2001), with fungi utilising lignolytic enzymes to acquire recalcitrant forms of organic N under elevated CO2. Finally, Principle 3, adaptability, accounts for the rapid change in the ability of AM fungi to supply additional N to their host plants. These principles allow us to make predictions that can be tested in future experiments. For example, the optimal allocation principle predicts that CO2 enrichment of N-rich soil will not stimulate production of lignolytic enzymes because there is already ample N in the system. In this situation, the increased production of mycorrhizas under elevated CO2 may indeed increase the accrual of C in soil organic matter. Studies of mycorrhizas across more habitat types are clearly necessary before the symbiosis can be effectively managed for its role in C and N cycling.
Another challenge in managing the symbiosis for N and C cycling is that complex interactions among anthropogenic enrichment of resources and mycorrhizal feedbacks may generate large-scale state changes in communities and ecosystems. The critical threshold levels at which state changes occur are difficult to predict because of the complexity of the interactions among factors. One of the most famous examples of a state-shift related to mycorrhizas is the relationship between anthropogenic N deposition and the loss of European heathlands. These ecosystems are characterised by acidic, organic rich soils and thick stands of heather (Calluna vulgaris) with ericoid mycorrhizas that have enzyme systems capable of releasing tightly bound N from polyphenolic compounds (Fig. 3). Airborne N inputs diminish N limitation and increase biomass production, which makes P more limited. At some point during this eutrophication process a threshold is reached, and the ericoid heather loses its competitive advantage and becomes replaced by Molinia caerulea an AM grass with a lower R* for P and a higher R* for N (Aerts 2002). This example illustrates how eutrophication of a limiting resource can restructure above- and belowground feedbacks and facilitate the conversion of heathland to grassland.
The structure of grasslands may also be influenced by mycorrhizal responses to N and CO2 enrichment. In community-scale mesocosm experiments, grasses with C3 photosynthesis generally benefit less from mycorrhizas and more from CO2 enrichment compared to C4 grasses (Johnson et al. 2003b; Antoninka et al. 2009). These findings support the idea that mycorrhizas can mediate competitive interactions among different functional groups of plants and that resource availability will affect the outcome of these interactions on community diversity. Principle 2, biotic context, is relevant here because most of the problematic exotic grasses invading native North America grasslands are C3 species that are not strongly dependent on mycorrhizas (Wilson & Hartnett 1998). Invasive species and land-use change, discussed below, are among the two most important anthropogentic drivers altering mycorrhizal relationships.
Changes to biota: invasive species
Studies suggest that mycorrhizal fungi can play an important role in determining patterns of abundance and invasiveness of introduced species (Bever et al. 2010). Introduced host plants have been shown to alter closely interlinked ecological relationships, many of which have co-evolved within native systems (Inderjit & van der Putten 2010). Although subtle, and rarely studied, these re-organisation processes are likely very common responses to global change that can decrease the efficacy of symbiotic partnerships and potentially lead to long-term, system-wide changes.
When invasive plants are introduced, at least three microbially mediated outcomes have been identified (Bever et al. 2010; Inderjit & van der Putten 2010; Fig. 9). All three outcomes can be linked to biotic context (Principle 2), and adaptability (Principle 3). One possibility is an ‘enhanced mutualism response’ in which native mutualisms facilitate the success of invading plant species (Reinhart & Callaway 2006). Mycorrhizas have been shown to facilitate the spread of non-native plants from low to super high abundance if the invading plant species benefits more from the symbioses than the indigenous plants. For example, spotted knapweed, Centaurea maculosa (Fig. 9a), is a noxious perennial plant that appears to generate an enhanced mutualism with native AM fungi. Many invading plant species exhibit higher growth rates than endemic species (e.g. van Kleunen et al. 2010a), and associating with native AM symbionts may facilitate this process by allowing them to gain access to more nutrients to support this growth. This highly efficient mycorrhizal partnership has likely facilitated the spread of knapweed throughout much of the native prairie in the northwestern USA (Harner et al. 2010). More generally, mutualistic facilitation of plant invasion arises when AM fungi alter nutrient uptake, competitive dynamics, successional changes and/or plant–herbivore interactions to the advantage of the exotic species and detriment of the native species (Shah et al. 2009). In furthering Principle 2, ecologists are designing new tools for assessing the determinants of invasiveness (van Kleunen et al. 2010b), which will eventually facilitate a more systematic approach to predicting mycorrhizal response to invasive species.
Figure 9. Examples of three microbially mediated outcomes of plant species invasion. (a) Enhanced mutualistic response – Knapweed (Centaurea maculosa) relies on AM fungi to aid its invasion of the Bitterroot Valley grasslands in Western mountain ranges of the USA (photo: Dan Mummey). (b) Degraded mutualism response – Garlic mustard (Alliaria petiolata), an invasive species in North America, has been shown to facilitate its spread by disrupting mycorrhizal associations (photo: Ben Wolfe). (c) Mutualistic barrier – Invasive pines (i.e. Pinus contorta) only became a problem when appropriate fungal symbionts were introduced to the environment (photo: Jon Sullivan).
Download figure to PowerPoint
It has been predicted that enhanced mutualistic responses are also most likely to emerge when the invasive hosts associate with widespread, generalist AM fungal taxa, rather than specialised taxa (Moora et al. 2011). This is in accord with Principle 3, mycorrhizal fungal adaptability. One idea is that the wide host range observed in AM fungi may be maintained by co-existence of genetically different nuclei and that the fungus can maximise its fitness in a given host by an alteration in the relative frequency of genetically different nuclei. There is some evidence for this in the study by Ehinger et al. 2009; in which small genetic changes were observed in clonal AM fungal lines when grown on different host plants. If true, then this would mean that AM fungi could have the capacity to rapidly adapt to new invasive plant species. In these cases, the native plant species may disappear, but the native, generalist fungi, which are now better adapted to the invasive host, remains in symbiosis with the invasive species. A similar response may occur in host plants as well: one recent case documented a generally non-mycorrhizal invasive plant species hybridising with a generally mycorrhizal native species to produce invasive offspring that benefited from the mycorrhizal association (Eberl 2011).
A second microbially mediated outcome occurs when a non-native plant species reduces the densities of native fungal symbionts and causes the subsequent loss of native host plants (Wilson et al. 2012). Entitled the ‘degraded mutualism hypothesis’ this outcome is possible when native plants are more dependent on mycorrhizal symbioses than invasive species (Shah et al. 2009) and/or invasive species directly degrade microbial targets (Cipollini et al. 2012). Invasion by non-native plants with low mycorrhizal dependencies can reduce mycorrhizal fungal densities in the soil, further facilitating invasion by non-mycorrhizal plants (Vogelsang & Bever 2009). This has been shown repeatedly across different mycorrhizal types (Shah et al. 2009; Castellano & Gorchov 2012), and can lead to long-term legacy effects, even after the invasive species has been removed (Meinhardt & Gehring 2012).
In some extreme cases, an introduced plant species can be lethal for the native symbiosis (Cipollini et al. 2012). A well-known example involves Alliaria petiolata (Fig. 9b), a European invader of North American forests, which suppresses native plant growth by actively disrupting mutualistic associations between native canopy tree seedlings and both AM (Stinson et al. 2006) and ectomycorrhizal (Wolfe et al. 2008) fungi. Suppression of North American mycorrhizal fungi by A. petiolata corresponds to decreased growth of North American plant species that rely heavily on fungal symbionts (Callaway et al. 2008) and changes in AM fungal community composition (Barto et al. 2011).
A third outcome is that an introduced plant species is unable to establish because the new habitat is void of the appropriate symbiont. In accordance with Principle 2, host fitness will depend strongly on the biotic context: in these cases, the lack of specific mycorrhizal propagules can be a critical factor in slowing invasion of fragile habitats. For example, initial plantings of pine (Pinus sp.; Fig. 9c) in the southern hemisphere often failed because of the absence of the necessary ectomycorrhizal fungi (reviewed in Pringle et al. 2009). Ironically, after the appropriate fungal symbionts are introduced to the environment, pine has become a problematic invasive species in many areas of the world (Nuñez et al. 2009). Exotic pine species have been shown to host exotic ectomycorrhizal fungi, and their spread into new habitats is essentially a co-invasion of exotic fungi and its host (Dickie et al. 2010). Although some attention has been paid to the potential dangers of the application of commercial products to purposely introduce mycorrhizal fungal inoculum into novel habitats (e.g. Schwartz et al. 2006), this area requires more research, especially in situations where fungi introduced as inoculum can become invasive (Jairus et al. 2011), or can facilitate invasion of exotic plant species.
Changes to biota: land-use changes
Humans have fragmented or removed over half of the global and temperate broadleaf and mixed forests, and roughly one quarter of the tropical rainforests (Wade et al. 2003). This is worrying because forests store ~45% of terrestrial C and contribute ~50% of terrestrial net primary production (Bonan 2008). Similar to invasions by introduced species, land-use change can dramatically alter native mycorrhizal symbioses, forcing changes in the aboveground community, and revealing feedbacks in the belowground community. At a global scale, the loss of forests and heathlands and the gain of deserts and croplands will reduce the abundance of ecto and ericoid mycorrhizas and increase the abundance of arbuscular mycorrhizas (Fig. 3). These changes may have major impacts on soil properties, heterotrophic respiration and the potential for belowground C sequestration.
A review of the literature suggests that if clear-cut forests are replanted with tree seedlings, mycorrhizal colonisation of the seedlings is generally not reduced; however, there are often dramatic shifts in the species composition of ectomycorrhizal fungi (Jones et al. 2003; Curlevski et al. 2010). These changes in fungal communities are likely driven by physical, chemical and biological changes in the reforested environment. It has been suggested that post-disturbance communities of ectomycorrhizal fungi may better aid their hosts under the altered environmental conditions (Jones et al. 2003). This seems likely because the deforestation and reforestation process is likely to have a dramatic impact on the organic horizon of the forest floor and species of ectomycorrhizal fungi differ in their enzymatic capabilities to acquire nutrients from detritus (Buée et al. 2007; Pritsch & Garbaye 2011). Future studies are needed to determine whether symbiotic function and fungal fitness are optimised in the mycorrhizas that reassemble following disturbance.
Conversion of diverse natural forests to plantations of exotic tree species may reduce species richness as well as change the composition of the native community of mycorrhizal fungi. Introduced Pinus contorta in New Zealand was found to host only 14 fungal species (93% were exotic), while co-occurring endemic Nothofagus solandri hosted 98 species of native ectomycorrhizal fungi (Dickie et al. 2010). A comprehensive study suggested that even if species richness of ectomycorrhizal fungal communities is maintained, the species composition is generally altered when native forests are converted to plantations (O'Hanlon & Harrington 2012). This result was confirmed in a study investigating the effects of conversion of native forest to avocado plantations; differences in mycorrhizal composition, but not richness, were found (Gonzalez-Cortes et al. 2012). In accordance with Principle 2, changes in species composition can have strong functional consequences for biotic interactions, raising the risk that future conversion back to native forest will be problematic. For example, changes in the community composition of mycorrhizal fungi have been shown to constrain the re-establishment of desired flora (William et al. 2011).
Tropical forest fragmentation and conversion to agricultural systems has been shown to negatively affect AM fungal interactions, with root colonisation, spore diversity and spore abundance positively correlated to fragment size, which is also negatively correlated with soil N and P availability (Grilli et al. 2012). In accordance with Principle 1, optimal allocation, we would expect higher soil fertility in fragmented plots to be correlated with a reduction in mycorrhizal fungi, as was found (Fig. 10).
Figure 10. Relationship between fragment size of Chaco forest in Co′rdoba, Argentina and the abundance of mycorrhizas, and the availability soil P, and N. Top left, total mycorrhizal colonisation of roots from Euphorbia acerensis (filled diamond) and E. dentate (open square). Top right, abundance of AM fungal spore abundance per 100 g of soil. Bottom left, concentration of soil P (ppm). Bottom right, concentration of soil NH4+ (ppm). Means ± SE (n = 8). Data courtesy Grilli et al. 2012.
Download figure to PowerPoint
With a growing interest in the functioning of mycorrhizal communities in agricultural systems, significant progress has been made in studying how land management regimes affects AM fungal communities. For instance, differences in the composition of fungal communities are often noted in high vs. low tillage regimes (Alguacil et al. 2008) and when grasslands, organic and conventional agricultural systems are compared (Verbruggen et al. 2010). These studies indicate that disturbances such as tillage – which disrupts AM hyphal networks – can negatively affect crop-mycorrhizal interactions compared to when hyphal networks are left intact (Verbruggen & Kiers 2010). Principle 3 suggests that intact networks of AM fungi have the ability to adapt rapidly to environmental changes by alterations in nucleotype frequencies within hyphal networks. This could explain the surprising resilience of AM fungal communities when, for example, fires drove the conversion of a Costa Rican dry tropical forest to a monoculture of African grass (Johnson & Wedin 1997). Whether all mycorrhizal communities can recover when land is restored after disturbance is far from clear: long-term effects of disturbance appear to be very site specific (e.g. Tedersoo et al. 2011). Understanding the patterns of recovery after major and minor land-use changes is critical for predicting community level responses. This knowledge hinges on first understanding the composition of mycorrhizal fungal communities in natural undisturbed ecosystems. Very few studies have assessed mycorrhizal fungi in natural habitats and there is an urgent need to recognise the importance of protecting natural areas for the preservation of these unseen, but important organisms so that future studies can use these as reference sites to guide ecosystem restoration efforts (Turrini & Giovannetti 2012).
Although loss of biodiversity is a major concern, loss of global carbon stores may be an even more pressing issue under current global change trajectories. The density of mycorrhizal hyphae in grasslands is often strongly correlated with soil organic C content (Wilson et al. 2009). Over the past century, natural grasslands have been increasingly converted to cropland, accelerating soil erosion, land degradation and releasing soil organic C stores (Qiu et al. 2012). Although careful management of ectomycorrhizas may help increase soil C sequestration within forest ecosystems (Hoeksema & Classen 2012), conversion of natural grasslands into tree plantations for C mitigation projects may be misguided. This is particularly true if ectomycorrhizal fungi that are introduced with the exotic trees may access C pools not available to the native AM fungi and actually increase C lost through respiration (Chapela et al. 2001). Does the short-term gain of an aboveground C-sink (wood) outweigh the long-term loss of a belowground C storage in recalcitrant organic compounds (Drake et al. 2011)? Given the current trajectory of land conversion as demands for food and agricultural commodities rise, more research is needed to provide reliable estimates of long-term C storage in above and belowground reservoirs across different types of ecosystems (Reich 2011). Mycorrhizas are an important component of belowground C-budgets and projections of future C-dynamics (Drake et al. 2011).