Can common mycorrhizal fungal networks be managed to enhance ecosystem functionality?

Several studies conducted in controlled environments show that key functional at tributes of common mycorrhizal networks (CMNs), which inter-link different plants, are influenced by management practices. Here, we highlight the need to consider how land management affects the ubiquity and function of CMNs in nature to maximize the role of mycorrhizal fungi in enhancing ecosystem services. We emphasize that CMNs can sometimes negatively affect aspects of plant performance, but there remain major gaps in understanding before explicit consideration of CMN management can be delivered. a mycorrhizal influence establishment, nutrition, productivity and defense, nutrient distribution and storage, and multitrophic interactions. However, many of these studies have focused on the importance of common mycorrhizal networks in ecological contexts and there has been less emphasis in managed systems, including croplands, grassland, agrofor estry and forestry, on which humankind relies. Here we review the evidence of the potential importance of common mycorrhizal networks in managed systems, and pro vide insight into how these networks could be managed effectively to maximize the functions and outputs from managed systems. We also emphasize possible negative effects of common mycorrhizal networks on plant performance and question popu lar views that mycorrhizal networks may offer a panacea for enhancing ecosystem services. We highlight the need to gain greater insight into the ubiquity, functioning, and response to management interventions of common mycorrhizal networks and, critically, the need to determine the extent to which these networks can add value to the promotion of mycorrhizal colonization.


| INTRODUC TI ON
Mycorrhizal fungi are largely obligate biotrophs forming associations with the majority of terrestrial plant species (Brundrett & Tedersoo, 2018), and often form complex and (usually) dense networks of extraradical mycelium (ERM) in soil (Miller et al., 1995). The ERM forages soil that is not accessible to the root system thus granting increased access to elements that are needed in large amounts by host plants, notably phosphorus (P) and nitrogen (N; Wipf et al., 2019). A fascinating feature of the ERM is its capacity to interconnect numerous plants belonging to the same or different species (Rhodes, 2017;Simard, 2018;Simard et al., 1997), forming common mycorrhizal networks (CMNs), often coined the 'wood-wide web' (Rhodes, 2017).
The potential functional consequences of such underground physical connections linking individual plants challenge our thinking of resource capture and multitrophic interactions, and as a result has spawned numerous popular views, sometimes exaggerated, extolling the virtues of CMNs for both natural and managed systems (Simard, 2018). Such enthusiasm for CMNs is not surprising given their functional capabilities. Indeed, CMNs formed by arbuscular mycorrhizal (AM) fungi have been shown to facilitate movement of water (H 2 O; Egerton-Warburton et al., 2007), and significant amounts of P (Mikkelsen et al., 2008) as demonstrated through the use of radioisotopes. CMNs formed by ectomycorrhizal (ECM) fungi have been shown to facilitate transfer of carbon (C; Selosse et al., 2006;Rog et al., 2020), N (He et al., 2005) and F I G U R E 1 Impacts of mycorrhizal colonization and integration of plants into common mycorrhizal networks. The arrows indicate a hypothetical virtuous circle in which promotion of colonization may lead to more abundance CMNs, which leads to more effective colonization of seedlings. Green shading highlights those interventions and services that have greatest potential to be implemented in specific ecosystems | 3 ALAUX et AL. . Furthermore, involvement of CMNs is not limited to nutrient exchange but can also involve allelochemicals (Barto et al., 2011) for AM fungi, and defence information (Gilbert & Johnson, 2017;Oelmüller, 2019) for both ECM and AM fungi, which can impact multitrophic interactions . Therefore, CMNs can actively participate in the well-documented beneficial effect of mycorrhizal fungi on plant nutrition (Smith & Read, 2008) as well as improving plant resistance and tolerance to abiotic (e.g. drought; (Plouznikoff et al., 2016)) and biotic (Whipps, 2004) stresses. Thus, there is growing evidence of the multifunctional effects of CMNs across ecosystems involving different types of mycorrhizal fungi. A burgeoning literature and plethora of commercial products and organizations have focused on promoting mycorrhizal fungal colonization of plants of interest (i.e. horticultural or crop species), especially those forming arbuscular mycorrhizas (Berruti et al., 2016;Chen et al., 2018). Yet, there remain few attempts to explicitly manage CMNs to enhance valuable ecosystem services, and tackle globally important grand challenges related to sustainable agriculture, production of fuel and fiber, water use and biodiversity conservation. Therefore, new information to inform guidance on how best to manage ecosystems is required.
This need is particularly important because many of Earth's terrestrial ecosystems have been managed to some extent; a recent consensus view estimates that 51% of land area has been modified for human purposes (Hooke et al., 2012). This estimate includes the areas used for croplands (13%) and pasture grassland (23%), and planted (2.1%) and logged (1.8%) forest, as well as urbanization.
Additionally, humans may influence unmanaged 'natural' systems indirectly due to anthropogenic activities (e.g. foraging, hunting, pollutant deposition). Thus, a vast range of specific land-uses are potentially affected by how humans influence mycorrhizal fungi  and the CMNs that they develop. Nevertheless, agroecosystems, including crops and grassland, and forests, productive forestry, agroforestry and forest restoration and creation, are the primary systems of interest from the perspective of CMNs.
Here, we provide an overview of the evidence of the potential importance of CMNs in managed systems and provide insight into how CMNs could be managed effectively to maximize the functions and outputs from these systems with reduced farming C cost. Indeed, conventional farming often has a high C cost notably by over fertilization or pest management and tillage (C footprint, Figure 1; Hillier et al., 2009). Crucially, we also emphasize the need to consider potential negative effects of CMNs on plant performance, and to ensure popular views of CMN function are evidence-based, so that effective management strategies can be carefully developed. CMNs are intrinsically linked to the need for plants both to form mycorrhizas and produce ERM (Rhodes, 2017).
Therefore, this synthesis inevitably considers hyphal production and density alongside our focus on the added impact (see

| HOW WIDE S PRE AD IS THE ' WOOD -WIDE WEB '?
In order to effectively manage CMNs, we need to have reliable data both on their ubiquity and extent, but also their functionality over distance. Alas, such information remains scarce, especially in AM dominated systems, despite previous calls to collect such data (Babikova et al., 2014), no doubt because of the difficulty in reliably estimating the presence and extent of CMNs. Notable exceptions include information on the extent of CMNs formed by particular species of ECM fungi in temperate forests in the US (Simard & Durall, 2004) and Canada (Beiler et al., 2010). For example, CMNs formed by Rhizopogon spp. Fr. extended across a 20 m × 20 m plot and often interconnected several trees (Beiler et al., 2010).
Nevertheless, we need far more studies of this kind before broad statements about CMN ubiquity in forests and woodland can be relied upon. Heathland vegetation dominated by ericaceous species also supports extensive ERM, but the fine roots of the host plants which are adept at foraging through soil mean that the fungi are generally confined close to their surface. Some of the fungi forming ericoid mycorrhizas can also form CMNs but these generally occur at localized scales (Kjøller et al., 2010). Remarkably, some fungi can simultaneously form classic ectomycorrhizas on trees and ericoid mycorrhizas on ericaceous shrubs, although these CMNs are also restricted to distances <4 cm (Grelet et al., 2010;Kjøller et al., 2010).
It is often assumed (especially in the popular science press) that CMNs formed by both AM and ECM fungi can cover large areas, potentially at field scales. Such enormous networks are unlikely to form in nature, but more information is needed to objectively inter- Increasing numbers of studies have quantified hyphal density (Kim et al., 2013) and even hyphal turnover (Hagenbo et al., 2017) by both ECM and AM fungi, including in managed ecosystems (Table 1). The often large densities of hyphae in soil in itself provides strong evidence for the potential to form CMNs, and perhaps this information is adequate to infer the presence of CMNs from a management perspective, without the need to undertake expensive, technically challenging, and time-consuming molecular or functional analyses. Nevertheless, we argue that greater focus is needed to quantify how land-management practices affect the ability of mycorrhizal fungi to form functional and stable networks, and provide more rigorous test of the hypothesis that ERM density is an adequate proxy for CMN formation. Moreover, relatively simple experiments could also be undertaken to confirm the ability of different mycorrhizal fungi to form CMNs under controlled conditions. The assumption is that all mycorrhizal fungi have this capacity, but a systematic test has not been done. The capacity to form CMNs may indeed be a functionally important trait that adds to the growing number of studies using traits to shed new light on the functional significance of microbial diversity (de Bello et al., 2010;Fry et al., 2019;Wang et al., 2019).
In soil, the ERM can represent between 20% and 30% of the total soil microbial biomass (Leake et al., 2004) and frequently accounts for tens, and sometimes up to a hundred m of hyphae per cm 3 of soil (Miller et al., 1995). Growth rates of mycorrhizal fungi (calculated by dividing total hyphal length by the number of days growth) can be fast: in vivo model systems using AM fungi estimated growth rates of 738-1,067 mm day −1 (Giovannetti et al., 2001), thus indicating great potential of fungi to interconnect many individual roots. Anastomosis between ERM of different AM host plants has been observed (Novais et al., 2017) and the frequency of fusions between hyphae coming in contact was high (67%-77%), as was the number of fusions along hyphae (0.46-0.51 mm −1 ; Novais et al., 2017).
The speed of CMN formation varies depending on the environmental conditions, notably the growing medium and distance between plants. Indeed, the latter stages of colonization of the potato variety Bintje has been obtained after 12 days using in vitro CMN (Gallou et al., 2010), whereas it was observed after TA B L E 1 Typical hyphal densities in soil from ecosystems with contrasting land use and management
Generalist fungi have broad host ranges potentially working as interaction network hubs in plant-fungus networks and thereby connecting otherwise isolated groups of plants (Olesen et al., 2007).
In forests, plants and fungi were associated with a narrower range of partners than expected under models that assumed random associations between hosts and symbionts, and the networks were compartmentalized into modules of closely associated plants and fungi (Toju et al., 2015). In those networks some trees were linked with up to 48 other trees, forming distinct hubs (Simard, 2018). The ability of some ECM fungi to form such inter-connected networks clearly offers opportunities for exploitation to enhance key ecosystem services related to forestry and forest restoration.

| FUN C TI ONAL AT TRIBUTE S OF CMN S RELE VANT TO MANAG EMENT INTERVENTI ON S
Here, we outline some of the key functional attributes of CMNs that offer the most potential to be influenced by management practices or, conversely, which may detrimentally affect plant performance to be done to test the significance of these findings under more realistic field-based conditions. Nevertheless, we illustrate potential guidance for land managers ( Figure 1); notably, we highlight that farming systems with light to no-till, cover crops and use of simultaneously diverse crops are expected to make the best use of CMNs.

| Colonization of seedling recruits
The ability of plant seedlings to become established is critical for their survival and becoming rapidly colonized by beneficial mycorrhizal fungi facilitates this process. From a management perspective (e.g. when creating new woodland, promoting establishment of species of conservation concern, growing food crops) ensuring that focal plants become rapidly colonized by beneficial fungi is therefore important.
Demonstrating such effects in the field is more challenging, and experiments often rely on mesh cores (e.g. Johnson et al., 2001) to manipulate seedling access to ERM supported by surrounding adult plants. This approach is not perfect because, first, it is rarely possible to unequivocally determine whether a seedling is connected to a wider CMN and, second, the mechanism of how a seedling benefits (or not) from integration into a CMN is difficult to quantify (e.g.

| Distribution and retention of mineral nutrients
A key question from a management perspective is: do plants integrated into common mycorrhizal networks gain more nutrients than plants that are solely associated with an isolated community of mycorrhizal fungi? Several studies have tackled this question but the findings are inconsistent with some reporting that integration into CMNs leads to positive effects on nutrient uptake and biomass compared to non-mycorrhizal controls (Francis & Read, 1995) or isolated mycorrhizal host , and some others reporting negative effects (Francis & Read, 1995;Janoušková et al., 2011;Merrild et al., 2013). While in some cases experimental designs pre-

| Pest and pathogen resistance
Below-ground signaling can occur between plants infected with pests to non-infested neighbors via CMNs formed by AM fungi (Babikova, Gilbert, et al., 2013;Song et al., 2010Song et al., , 2014, endophyte fungi (Vahabi et al., 2018), and ECM fungi . CMNs plant defence: genes or compounds) within 24 hr or even 6 hr after donor plant being exposed to the stressor (Alaux et al., 2020;Song et al., 2019), with peak response on the receiver plant between 48 hr and 100 hr . The duration over which signals are effective remains unknown (Gilbert & Johnson, 2017) but in recent studies appeared transitory, suggesting a priming phase (Alaux et al., 2020). Plant responses have been related to production of jasmonic acid (JA), ethylene (Eth; Alaux et al., 2020;Song et al., 2014Song et al., , 2019, and salicylic acid (SA; Song et al., 2010;. Interestingly, the plant response (i.e. priming of plant defence) seems to activate only particular aspects of JA response (Song et al., 2019), which may limit the potential cost of induced defence as a result of CMN signal transfer. To effectively use this approach and reduce pest-related crop losses additional elements need to be investigated in field situations, notably the putative relay of the signals among plants and potential links between other CMNs (Wipf et al., 2019).   (Liang et al., 2020). Possible mechanisms behind these observations relate to the increased protection from pathogens conferred by ECM fungi compared to AM fungi, and also the ability of ECM plants to access a greater range of P forms, including complex organic molecules such as phytate (Liu et al., 2018).

| IMPAC T OF MANAG EMENT PR AC TICE S ON ERM DEN S IT Y AND CMN FUN C TI ON IN S ITU
Few studies have explicitly focused on understanding how management practices may affect CMN functioning, even though fungal networks have potential to be affected (mostly) negatively by many conventional land-management practices. The evidence suggests that AM fungi have a key role in sustainable agriculture  but we still lack clarity on how CMNs can further benefit these practices. In addition, other management scenarios should not be ignored, especially forestry and woodland creation and restoration (Figure 1), which is high on the political and societal agenda (Bastin et al., 2019). Furthermore high ERM density could increase C sequestration above the effects already observed in no-till management (around 115kg C /ha/year in United Kingdom; Hillier et al., 2009), and thus contribute to several possible ecosystem services provided by CMNs.
Intercropping is already used widely and often consists of combination of plant species in order to achieve over yielding due to the complementary strategies of the partner species. For example, mixtures typically comprise combinations of cereals and legumes (Inal et al., 2007), but also plants with either C4 or C3 photosynthetic pathways, such as maize and chili pepper (Hu et al., 2019). In addition to reducing loss of N from legumes (Hamel et al., 1991), CMNs can simultaneously improve the efficiency of the maize root system for the recovery of N excreted by co-occurring and interconnected soybeans (Hamel et al., 1991). Furthermore, AM fungal colonization increased when maize and chili pepper were intercropped and formed a CMN, with maize supplying part of the C for increasing AM fungi propagules in pepper roots. CMNs increased fruit yield via improving P distribution to chili pepper. Amongst a given CMN, the P competitive strategy differed for each plant host with notably higher ability of pepper against maize to access P (Hu et al., 2019), thus highlighting the need of careful selection of plants associated during the same cropping season.
Cover cropping has been reported to contribute to the restoration and maintenance of AM fungi (Verbruggen et al., 2010).
Cover cropping during winter periods allows the maintenance of a potential AM fungal inoculum increasing colonization in the subsequent cultivated crops (Verzeaux et al., 2017). By providing sus- mother plants, Figure 1) via CMNs for newly planted saplings, which has been shown to improve sapling growth (Teste & Simard, 2008).
Increasing focus has been put on the interactions between crops and weeds; AM fungi can suppress weeds if they are non-hosts (Daisog et al., 2012) that could be exploited through co-cultivated in intercropping systems. The effect of biomass removal from grazing on CMN functioning remains to be tested. The effects of grazing on AM colonization of roots is ambiguous (Barto & Rillig, 2010) partly through complex interactions with biological (e.g. plant and fungal species), edaphic and climatic variables (Faghihinia et al., 2020 Hetrick et al., 1992), as well as in other ecologically important traits.
Therefore, we expect ERM production to be also dependent on host genotype identity, and creating mixtures within a field or between growing seasons offers promise to enhance ecosystem services and

| SYNTHE S IS AND FUTURE D IREC TI ON
Humans need to make a step-change in how land is managed if we are to tackle society's grand challenges and the sustainable development goals set out by the United Nations. Capitalizing on the functional attributes of mycorrhizal fungi offers promise ( Figure 1), but we highlight that greater explicit focus on CMN ubiquity, functioning, and response to management interventions is needed to gain a much-needed 'holistic' view of how mycorrhizal fungi can be managed to enhance ecosystem services. A popular view of CMNs is that they offer many benefits for the function- It is vital that key knowledge gaps are filled, notably the need to characterize under field conditions the AM species involved in the formation of CMNs. Furthermore, a consistent approach to experimentation, data collection, and synthesis among mycorrhizal type would also be beneficial; consideration of P uptake remains a focus in AM systems, but the role of P in ECM systems is relatively poorly studied.
We highlight the need for further basic and applied research on the ubiquity and function of CMNs in nature in order to inform land-management policy and guidance. Notably, we need to know under which circumstances do CMNs really have value for land-managers? And can CMNs ever be considered as separate from mycorrhizal colonization in nature, and therefore explicitly influenced by management? Answering such questions is vital not just in managed systems but also from a fundamental ecological context.

ACK N OWLED G M ENTS
This work was supported by funding from the Natural Environment Research Council (NE/P001823/1). DJ receives partial support from the N8 AgriFood program.

AUTH O R CO NTR I B UTI O N S
P-LA, YZ, and DJ wrote the manuscript. YZ collected data used in table and made the summary figure. LG contributed manuscript editing and provided central feedback. DJ and LG secured the funding.