Introduction

An appreciation for the insights of our scientific ancestors is quite valuable. It can reduce the waste of time and other resources in attempts to “reinvent the wheel”. It can provide context for current research. It may even help us overcome contemporary research problems and clear up contemporary controversies by providing points of view that are uncolored by current dogma. We illustrate the value of understanding the older literature by highlighting a few of the insights of mycorrhiza function by John (Jack) Laker Harley (1911–1990) as they relate to some of the most important research of today.

Harley completed his undergraduate (post-secondary) education in Botany at the University of Oxford in the 1930s. Although Harley’s interests were primarily physiological, interactions with Professors Tansley and Clapham undoubtedly helped him to appreciate the ecological context. For his D. Phil., also at Oxford, Harley studied the mycorrhizas of European beech (Fagus sylvatica L.), but he apparently considered the work to be of relative unimportance and switched temporarily to working on the nitrogen metabolism of a non-mycorrhizal fungus. As with so many from that generation, Harley’s career was interrupted by “the war” and, from 1940 to 1945, he served as an officer in the Royal Army in India, Burma, the USA, and Ceylon (now Sri Lanka). Upon leaving Ceylon, Harley returned to staff at Oxford and soon the beech mycorrhiza was, once again, the subject of his enthusiasm (Smith and Lewis 1994).

Harley was a critical thinker. He could be critical of his own research and he could be critical of the research of others. For example, he gave the concluding address to the second European symposium on mycorrhiza in Prague in 1988 (Harley 1989). In that address he said: “Before I discuss the important differences between this and previous mycorrhizal conferences, I would like to make a few criticisms, but as we say in Britain, no names, no pack drill. In other words, I won’t mention names in order to save embarrassment.” But, of course, at such a small conference the identification of individuals might only have required the mention of research topics. There is usually little to be gained by being critical only to embarrass someone. But there can be much to be gained by using criticism to raise the level of research.

Harley was a pioneer “mycorrhizast,” a term he lightheartedly coined to describe those who enthusiastically study mycorrhizal symbioses. Much of his research concerned various aspects of the physiology of “ectotrophic” mycorrhizas, now known as ectomycorrhizas. His lasting contributions to our field, however, transcend physiology and beeches and ectomycorrhizas and continue to influence research in many fields. Academic administrators have created perverse incentives by relying on various quantitative measures of research impact when considering a scientist for promotion. But it may be instructive, particularly for them, to consider the fact that Harley’s ability to contribute very meaningfully to our field did not occur as a consequence of a high h-index, or because he published in journals with high impact factors, or because he was awarded a tremendous sum in grants. It stemmed from simply being a talented researcher and communicator. In particular, Harley had (1) a penchant for viewing the mycorrhiza in a broad biological context, (2) the ability to synthesize seemingly disparate bits of information into important insights, and (3) a desire to share ideas in order to stimulate further research, mainly through his books and reviews. Although Harley’s research was published decades ago, much of it continues to be quite relevant today.

It is the continued relevance of “older literature” that we wish to highlight here. Indeed, our purpose is to convince our readers that there is real value in reading older materials, and not just for historical interest. We illustrate this with just a few of many examples of Harley’s insights that laid the groundwork for some very important lines of research of today. In particular, we address six insights of Harley and his students concerning mycotrophy, the new niche, the sheath, C cycling, N cycling, and mutualism.

Mycotrophy

Harley used the term “mycotrophy” to describe the feeding of mycorrhizal fungi. In 1975 when he wrote a chapter entitled “Problems of Mycotrophy” (Harley 1975), the extent to which mycorrhizal fungi are dependent on their hosts for reduced carbon in the form of photosynthate was not entirely clear. It is hardly any clearer today (Koide et al. 2008).

On one hand, the old term “ectotrophic” and its alternative “endotrophic” suggest that mycorrhizal fungi obtain nutrition at least partly biotrophically, while either remaining outside root cells or by forming structures within them. Indeed, Harley recognized a classic ectomycorrhiza as one in which the fungus derives the majority of its reduced carbon in the form of soluble carbohydrate directly from the living plant (Harley 1975). Moreover, summarizing the conclusions of Gray and Williams (1971), Harley wrote “the estimated yearly accretion of carbon, from the obvious sources of leaf fall and so on, to the soil is often inadequate to explain both the rate of CO2 production from the soil and the maintenance of the estimated biomass of the soil organisms...” (Harley 1975). Harley recognized photosynthate as the hidden C source: “[mycorrhizal fungi] are adapted to receiving a supply of carbon direct from the photosynthetic products of the autotrophic partners rather than primarily or solely from the humus or dead tissues indirectly derived from photosynthesis after death of the autotroph” (Harley 1975).

Nevertheless, Harley was aware that among ectomycorrhizal fungi, considerable variation occurred in their capacity to acquire C from litter (Harley 1969). He argued that it would be extremely risky to define the mycotrophic status of mycorrhizas too narrowly, writing “[mycorrhizas] grade from clearly mutualistic symbioses to … casual root-surface and rhizosphere associations on the other” (Harley 1975). Harley further explained “as more and more cases [of mycorrhiza] were described, the contrast between the forms called mycorrhizas and the more casual association between fungi and roots became less and less clear” (Harley 1969). Harley appreciated the large degree of variability existing among the various kinds of plant-fungal associations better than some do, perhaps, and that knowledge predisposed him to view mycorrhizas as occupying a considerable space along trophic and anatomical continua with no clear boundaries (Fig. 1). In his view, mycorrhizal fungi are as their name indicates, associated with roots in one way or another but, as a group, unspecifiable in terms of exact anatomical relationship with the root or their degree of biotrophy.

Fig. 1
figure 1

Harley appreciated the variability existing among the various kinds of plant-fungal associations, and that predisposed him to view mycorrhizas as occupying a considerable space along trophic and anatomical continua with no clear boundaries. Narrower concepts of “mycorrhiza” exclude some associations that Harley may have included as a mycorrhiza

Full size image

The variability in the C nutrition of mycorrhizal fungi continues to be of great interest today. For example, the research of Rineau et al. (2013) and Lindahl and Tunlid (2014) suggests that some ectomycorrhizal fungi, while participating in litter decomposition are, nonetheless, not saprotrophic. On the other hand, Vasiliauskas et al. (2007) suggest that some saprotrophic fungi colonize roots in a manner consistent with ectomycorrhizal fungi. This was further supported by Smith et al. (2017), who surveyed 201 saprotrophic basidiomycetes for their ability to colonize root systems of Pinus sylvestris and Picea abies seedlings. They found that a surprisingly high proportion of those fungi (16.9%) were capable of colonizing the surface, epidermis, or cortex of roots, and some of them even formed mantles or Hartig nets. Taken together, these modern contributions suggest that the trophic status of ectomycorrhizal fungi, as a group, may be quite broad. In fact, the breadth of the trophic status of mycorrhizal fungi has not been all that remarkable since we realized that ectomycorrhizal fungi evolved within saprotrophic lineages on multiple occasions (Hibbett et al. 2000). What is remarkable is that Harley recognized this in 1975.

The new niche

Obviously, the retention of even rudimentary powers of saprotrophy by ectomycorrhizal fungi (Harley 1969) would give them some primacy by preemption over pure saprotrophs attempting to utilize dead roots as a carbon source (Koide et al. 2008). However, mycorrhizal fungi are, as a rule, not good saprotrophs and, as indicated earlier, several fungal lineages have basically given up saprotrophy as they evolved to exploit the mainly biotrophic, mycorrhizal niche (Hibbett et al. 2000).

In Harley’s words, with mycorrhizal fungi “there is a short-circuiting of carbon direct from photosynthesis into the heterotroph” (Harley 1975). Harley attributed quite a bit of significance to the novelty of the biotrophic status of mycorrhizal fungi and for good reason. First, consider the C flow paths to saprotrophic fungi (Fig. 2a.i) and biotrophic, ectomycorrhizal fungi (Fig. 2a.ii). Compared to saprotrophic fungi, ectomycorrhizal fungi have moved upstream, so to speak, as close to the C source (photosynthate) as possible, hence the “short circuit.” The transfer of energy from the host is more efficient to the biotroph than to the saprotroph because the utilization of sugars is more efficient in producing biomass than is the utilization of complex polymers such as cellulose or lignin (Manzoni et al. 2012). Thus, for a given amount of photosynthesis, the total supportable biomass of biotrophic, mycorrhizal fungi should be greater than the total supportable biomass of saprotrophic fungi. This fact has obvious relevance to the C economy of individual mycorrhizal fungi and to the relative abundance of mycorrhizal to saprotrophic fungi, but it also has relevance to overall ecosystem C cycling (see the “Ectomycorrhizal fungi and ecosystem C cycling” section, below).

Fig. 2
figure 2

a Compared to saprotrophic fungi, ectomycorrhizal fungi are closer to the C source (photosynthate). b If fungal growth is limited by carbohydrate, and if host photosynthesis is limited by P, then when the fungus shares P with the host, its photosynthesis rate increases, potentially allowing the fungus to grow more rapidly. c Harley reasoned that nutrient-absorbing surfaces consisting of hyphae are less costly in terms of photosynthate than roots. Therefore, for a given amount of P uptake, less photosynthate need be allocated belowground when plants are mycorrhizal

Full size image

Second, Harley recognized that biotrophy need not be a zero-sum game. In other words, the gain of C by the fungus need not represent a net loss of C to the plant. He wrote “[in a reciprocal mutualism]...a factor or factors limiting growth of each partner must be supplied by the other”. Thus, if fungal growth is limited by carbohydrate, and if host photosynthesis is limited by P, then sharing P with the host would increase host photosynthesis (Reid et al. 1983) and, consequently, lead to a net gain in C for both the plant and fungus (Fig. 2b). Thus, by virtue of transporting phosphate to their hosts, more C enters the biosphere than would be possible without the fungi. Again, this has relevance to the C economy of individual mycorrhizal fungi, to the balance between mycorrhizal and saprotrophic fungi and to ecosystem C cycling (see the “Ectomycorrhizal fungi and ecosystem C cycling” section, below).

Third, based on the work of Sanders and Tinker (1973), Harley (1975) reasoned that nutrient-absorbing surfaces consisting of hyphae are less costly in photosynthate than when they consist of roots. Therefore, for a given rate of P uptake, less photosynthate needs to be allocated belowground when plants are mycorrhizal than when they are not, leading to a C savings for the plant, additional C available to the fungus (Fig. 2c), and an impact on ecosystem C cycling (see the “Ectomycorrhizal fungi and ecosystem C cycling” section, below).

When Harley pointed out that mycorrhizal fungi occupied a biotrophic niche distinct from the one occupied by saprotrophic fungi, he seemed to be aware of the ecological consequences that are only now being more fully explored. For example, the biotrophic nature of mycorrhizal fungi, linked with reciprocal P exchange, suggests that the balance between saprotrophic and mycorrhizal fungi should be affected by soil P fertility, and this has only recently been shown (Kyaschenko et al. 2017).

The sheath and maintenance of the new niche

If hyphae are less costly than roots as nutrient-absorbing organs, one can easily envision selection for plants to become highly dependent on mycorrhizal fungi for phosphate uptake, particularly plants with thick roots that are poorly adapted to nutrient capture. Indeed, tree species that do not proliferate roots in nutrient hot spots either proliferate mycorrhizal fungal hyphae (Cheng et al. 2016) or preferentially associate with mycorrhizal fungi with greater exploration distances than tree species that can proliferate roots in nutrient hotspots (Chen et al. 2016, 2018). While Harley was wary of constructing evolutionary hypotheses that could not be directly tested, he nonetheless offered some choice insights into important evolutionary processes. For example, if mycorrhizal fungi had some means of promoting dependence of the host on the fungus for nutrient uptake, the fungus would be all the more evolutionarily stable in its biotrophic niche. The fungal mantle or sheath may serve this purpose.

Harley was fascinated by the fungal sheath (Harley 1969, 1975). Describing the sheaths produced by mycorrhizal fungi of the Basidiomycota, he wrote “the consistent presence of a sheath in so many genera of gymnosperms and angiosperms cannot be ignored in hypotheses of function for it is an expensive structure in upkeep. It must, you would think, have a selective value …” (Harley 1975). At some point, Harley realized that dependence of the host on the fungus was enforced by the sheath for, after all, the sheath prevents contact between root and soil, thereby preventing the root from taking up nutrients directly from the soil. Although ectomycorrhizal short roots may resume growth and become largely non-mycorrhizal long roots (Harley 1969), on most occasions, fungal sheaths appear to shield the vast majority of root surface from the soil. In describing the ectomycorrhizal short roots of European beech, Harley wrote “the [root] branch initials develop from the pericycle immediately opposite the xylem poles and each young branch, as it passes through the cortex, forms a bulge in the sheath which is visible on the surface of the mother-root. The fungal sheath above the bulge does not break under tension but extends so that throughout its growth the young branch remains totally encased in fungal tissue” (Harley 1969).

The sheath, therefore, has the potential to act as the nutrient gatekeeper. It could only do so, however, if it not only intercepted nutrients on their way to the root, but also controlled whether the nutrients were, in fact, transferred. Harley was well aware that ectomycorrhizal fungi are not merely pawns of the host when it comes to phosphate transfer. He wrote “when excised beech mycorrhizas were allowed to absorb labelled phosphate from a solution of low concentration, about 90 per cent was found to remain in the fungal tissue during the first few hours. Only when phosphate concentrations far above those expected in the soil were applied, did the sheath allow a greater proportion than 10 per cent to pass to the host” (Harley 1969). More recently Wu (2014) demonstrated that photosynthate is transferred from shoot to fungal sheath within 1 day of synthesis, while soil phosphate is accumulated in the fungal sheath and then transferred to the plant only after a few days (Wu 2014). These observations suggest that despite receiving carbohydrate from the plant, at least some ectomycorrhizal fungi need not immediately respond by transferring phosphate to the host, nor need they transfer all of the phosphate they have collected (Fig. 3). Indeed, Harley (1975) recognized that because the sheath prevents direct root uptake of phosphate and acts as a storage organ rather than a mere pass-through for phosphate, the fungus may be in nearly complete control of plant phosphate acquisition. Ectomycorrhizal fungi may act similarly with respect to N, as they are capable of immobilizing N, reducing its absorption by roots (Näsholm et al. 2013). Thus, by assuming the roles of nutrient gatekeepers, mycorrhizal fungi have made themselves indispensable to the host, thus safeguarding their biotrophic niche.

Fig. 3
figure 3

Harley recognized that because the sheath prevents direct root uptake of phosphate, and acts as a storage organ rather than a mere pass-through for phosphate, the fungus is in nearly complete control of phosphate acquisition. By assuming the role of P gatekeeper, mycorrhizal fungi have made themselves indispensable, thus safeguarding their biotrophic niche

Full size image

Without some way of exerting control over plant nutrient uptake, the fungus may have no means of maintaining its indispensability. Thus, while some ectomycorrhizal associations may not involve a well-developed sheath, one cannot envision a strong natural selection maintaining those associations. Hibbett et al. (2000) suggested that evolutionary instability of ectomycorrhizal symbiosis may stem, in part, from the lack of mechanisms that protect the fungi from exploitation by the plants. As recognized by Harley, however, a well-developed sheath could serve this purpose: The sheath “must, you would think, have a selective value …” (Harley 1975). It would be interested to see if the unstable lineages as highlighted by Hibbett et al. (2000) were primarily those in which sheaths were not well-developed.

Ectomycorrhizal fungi and ecosystem C cycling

The capacity of mycorrhizal fungi to influence important ecosystem processes was largely unappreciated in the decades of Harley’s research activities. Few individuals of that period asked whether mycorrhizal fungi had a significant effect on C cycling probably because the issue of global climate change and the role of C cycling in climate was not the large issue it is today.

We already indicated that mycorrhizal fungi may increase the total quantity of C flowing through an ecosystem when they supply nutrients to the host and when those nutrients limit photosynthesis. Obviously, this may have large consequences by itself. We point out here, however, that mycorrhizal fungi may have additional impacts on ecosystem C cycling; a large flux of C passes directly through mycorrhizal fungi so there are additional impacts when mycorrhizal fungi alter the quality of available C in the ecosystem.

It has recently become quite clear that the total quantity of C passing through ectomycorrhizal fungi is significant. But it was actually decades ago when Harley first suggested that the mycorrhiza effect on C cycling was underappreciated. He wrote, “From both an ecological and a physiological point of view it is important to gain some estimate of the quantities of photosynthetic products which are directly diverted to the mycorrhizal and other symbiotic and associated heterotrophs in an ecosystem…a conservative value equivalent to about 10% of that used in timber production, 500 kg per hectare per year, of a temperate forest is estimated. This does not include carbohydrate used by mycelium in the soil....The importance of symbiosis in ecosystems has been much underestimated...” (Harley 1975). Harley realized that the estimate of 10% of timber production was probably a severe underestimate because it was based on the standing biomass of sheaths and sporocarps, ignoring the biomass of the mycelium in the soil and the turnover of fungal biomass. Indeed, he estimated that the mycelium represented a significant avenue for carbon transfer from host to soil. The “estimates of carbon dioxide released from the soil (so-called soil respiration) are usually much greater than can be explained by the breakdown of the amount of detritus falling upon them. Quantities between one-third and two-thirds of the CO2 released cannot be accounted for....It is extremely probable that much of the CO2 originates fairly directly from the consumption of the photosynthetic products by mycorrhizal hyphae…” (Harley 1975). Indeed, Ekblad et al. (2013) showed that turnover of the ectomycorrhizal fungal mycelium accounts for up to several hundred kilograms per hectare per year, doubling Harley’s estimate of the contribution by mycorrhizal sporocarps and sheaths.

Of course, the amount of C passing through the mycorrhizal fungal pool would make little difference to C cycling if it did not influence the rate at which organic matter is decomposed and, thus, the extent to which it remains sequestered in the soil. But, in fact, it may. Consider a world in which there are no mycorrhizal fungi. The C flow path would look like this:

  1. 1)

    Photosynthate→ plant biomass→ saprotrophic microorganisms

However, in a world with mycorrhizal fungi, there is an additional, parallel C flow path:

2) Photosynthate→ mycorrhizal fungal biomass→ saprotrophic microorganisms.

Harley noted that “The importance of symbiosis in ecosystems has been much underestimated” because the presence of mycorrhizal fungi in ecosystems can substantially alter both the quantity (see above) and quality of the C available to decomposer organisms.

For example, ectomycorrhizal fungi, under special circumstances, may actually retard litter decomposition, possibly via competition with saprotrophs (Fig. 4.i). This is known as the “Gadgil effect” (Gadgil and Gadgil 1971). Summarizing the findings of Gadgil and Gadgil, Harley wrote, “At present it is possible to imagine that the host root, after infection, acts as a center from which the mycorrhizal fungi exploit the humus layers in competition with free-living saprophytes.” “If it is true that all or a large part of the carbon for mycorrhizal fungi comes from the host…the fungi are able, by use of this carbon source, to remove inorganic nutrients selectively from the F and H soil layers. The effect of this will inevitably be to slow up the rate of humus breakdown which is in most soils limited by nutrients such as nitrogen and phosphate” (Harley 1975). More recently, the extraction of organic nutrients from litter by ectomycorrhizal fungi is also thought to contribute to the Gadgil effect (Orwin et al. 2011). The use of the word “inevitably” is, perhaps, hyperbolic because, as Harley indicated, the Gadgil effect caused by nutrient competition is only expected when decomposition is limited by nutrients, and he indicated that this was only true “in most soils”. It is no longer clear that decomposition is limited by nutrients such as N or P even “in most soils.” In fact, the Gadgil effect does not always occur, as indicated by Fernandez and Kennedy (2015), and this variability may have something to do with the relative strengths of suppression and priming (see below) of saprotrophs by ectomycorrhizal fungi. Ultimately, the strengths of these effects may hinge on numerous environmental factors, such as soil moisture (Koide and Wu 2003) and soil fertility (Fernandez and Kennedy 2015), which was recently shown to be negatively correlated with both the abundance of ectomycorrhizal fungi and soil organic matter accumulation in boreal forests (Kyaschenko et al. 2017).

Fig. 4
figure 4

Schematic diagram of the potential effects of ectomycorrhizal fungi may have on carbon cycling. Arrows represent either suppression (blue) or acceleration (red) of soil C cycling. (i) The suppression of saprotrophic activity via competitive interactions with ectomycorrhizas, or the ”Gadgil effect,” is hypothesized to decelerate soil organic matter decomposition rates and increase C held in SOM. (ii) Conversely, ectomycorrhizal fungi may accelerate soil organic matter decomposition rates by “priming” saprotrophic fungi with labile C exudates. (iii) The turnover of labile ectomycorrhizal necromass (dead biomass) represents a very fast cycling input; however, this C may be efficiently used by bacteria and subsequently stabilized by soil matrix interactions (iv). (v) Some ectomycorrhizal fungi produce very recalcitrant necromass that is very resistant to decomposition and is thought to contribute to C storage in SOM

Full size image

Mycorrhizal fungi may also contribute significantly to rapidly cycling C pools, and Harley wondered about the ecological consequences of this. He asked, “to what extent do mycorrhizal fungi secrete into the soil organic acids - hydroxyacids - such as malic or citric…?” Rygiewicz and Andersen (1994) indicated that mycorrhizal trees released far more C through rapidly cycling (labile) pools than non-mycorrhizal trees. Högberg and Högberg (2002) indicated that ectomycorrhizal fungal mycelium contributes about half of the dissolved organic carbon, which can serve as a source of carbon for other microbes. Rapidly cycling C compounds produced by ectomycorrhizal fungi may increase decomposition of soil organic matter by stimulating the activity of soil saprotrophs via a process known as priming (Kuzyakov et al. 2000, see also Fig. 4.ii). Chitin, a polymer produced in the cell walls of ectomycorrhizal fungi, is highly labile and a potentially important source of C and N for saprotrophic microorganisms (Fernandez and Koide 2012; Koide et al. 2011). Therefore, colonization of roots by ectomycorrhizal fungi may significantly prime fine root decomposition (Koide et al. 2011). On the other hand, the production of labile organic compounds by mycorrhizal fungi may contribute significantly to stable soil organic matter when they are processed by saprotrophic microbes that produce novel, recalcitrant materials that become sorbed to mineral surfaces (Cotrufo et al. 2013, see also Fig. 4.iv).

Some mycorrhizal fungal C pools are highly recalcitrant (Fig. 4.v), and these can result in significant soil organic matter retention (Wang et al. 2017). Melanin, for example, is highly resistant to decomposition (Fernandez and Koide 2014; Siletti et al. 2017; Fernandez and Kennedy 2018; Lenaers et al. 2018) and thus may contribute significantly to stable soil organic matter. Indeed, Clemmensen et al. (2013) demonstrated that slow decomposition rates of mycorrhizal fungal necromass may contribute significantly to organic matter accumulation.

Harley’s interest in the effect of mycorrhizal fungi on ecosystem C cycling was prescient. Although the topic was of minor interest in the 1970s, it has become an extremely important topic today because of our concern for the role of C cycling in climate stability.

Ectomycorrhizal fungi and ecosystem N cycling

The uptake of mineral N by ectomycorrhizal fungi has multiple consequences. First, N may be passed to the host. When photosynthesis is N-limited, this may increase the amount of reduced C entering the ecosystem. However, Harley recognized that in order for the fungus to pass materials to the host, “…the fungus must have absorbing organs in the soil through which materials are absorbed in excess of its own requirements” (Harley 1952). In other words, unless the fungus can acquire more of a resource than it needs, it is illogical to assume it would pass this limiting resource on to its host. Indeed, ectomycorrhizal fungi can immobilize N from the host plant (Näsholm et al. 2013; Hasselquist et al. 2016). Koide and Kabir (2001) showed that the likelihood of nutrient immobilization by ectomycorrhizal fungi is much greater for N than for P. Harley himself showed that even when an increase in P concentration of the host occurs as a consequence of ectomycorrhizal colonization, there may be no increase in N concentration (Harley 1969). Nevertheless, net transfer of N from fungus to host may occur and have a significant impact on the N economy of the plant (Abuzinadah and Read 1986), and thereby influence all processes limited by N, including photosynthesis.

Second, the transfer of N from fungus to host occurs concomitantly with transfer of C in the same direction. Even when ectomycorrhizal fungi acquire mineral N, it is rapidly converted into amino acid before it is transferred to a host. Lewis (1975), one of Harley’s PhD students, wrote “Harley (1964) and Carrodus (1966 1967) showed that, in beech mycorrhizas, absorbed ammonia is rapidly converted, principally in the sheath, to the amino compounds, glutamic and aspartic acids and, especially, glutamine”. Insofar as reduced C is acquired biotrophically, most of the C in the newly produced amino acid is derived from the plant. Harley (1969) wrote “…the carbon skeleton of glutamate which is transferred from fungus to host is probably derived from host photosynthate to the extent of three-quarters of the carbon involved, the remaining quarter coming from CO2”. One consequence of this situation was obvious to Harley. “There is thus undoubtedly some recycling of carbon derived from the host through the fungus and back to the root” (Harley 1969). “Lewis pointed out that carbon which moved from host to fungus as carbohydrate could return as amino-compounds so that bilateral transport of carbon should not be equated with that of carbohydrate…” (Lewis 1975). Harley, therefore, appreciated that it is possible for C to be transferred from one plant to another via the hyphae of an ectomycorrhizal fungus in the following way: transfer of carbohydrate from host 1 to fungus, use of that carbohydrate by fungus to produce amino acid from mineral N, transfer of the amino acid from fungus to host 2 (Fig. 5). This could explain, as Harley (1975) put it, the “reverse flows of 14C as by Reid and Woods (1969) in ectotrophic mycorrhiza from fungus to host…”, and may explain results that suggest carbohydrate can be transferred from one host to another via mycorrhizal fungi.

Fig. 5
figure 5

Harley appreciated that it is possible for C to be transferred from one plant to another via the hyphae of an ectomycorrhizal fungus in the following way: transfer of carbohydrate from host 1 to fungus, use of that carbohydrate by fungus to produce amino acid from mineral N, transfer of the amino acid from fungus to host 2

Full size image

While transfer of C as amino acid from host to host is highly probable, it does not seem likely that carbohydrate would be transferred via ectomycorrhizal fungi. Indeed, Harley and Lewis clearly recognized the evolution of specific physiological mechanisms to promote unidirectional transfer of carbohydrate from plant to fungus. Surely one was the conversion of carbohydrate from glucose to trehalose, which cannot diffuse passively across membranes and even if it were, is poorly utilized by plants. Harley (1969) wrote “…fungus and higher plant…have very different internal carbohydrate substances. It is of special significance that mannitol, trehalose, and glycogen, which are the important storage carbohydrates of the fungal layers, are not found in the host-plant and are not readily absorbed by it if made available. By contrast the host carbohydrates, sucrose, glucose, and fructose, are readily absorbed and utilized by the fungus. This allows the hypothesis to be formulated that the fungus can absorb carbohydrate from the host and convert it into forms that are unavailable for reciprocal flow”. It seems unlikely, then, that photosynthate itself moves from host to host via mycorrhizal fungi as has been implied (Song et al. 2015).

There is considerable interest in the movement of materials from plant to plant via mycorrhizal fungal hyphae. In particular, the idea that canopy trees are capable of sharing carbohydrate with light-limited trees in the understory has captured the imagination of some (Beiler et al. 2010; Wiemken and Boller 2002). However, while Harley worked out the means by which some C can be transferred from plant to plant in the form of amino acid, he did not find it likely that this could have a major impact on the C economy of the recipient plant or that carbohydrate could be involved in the exchange of C. Indeed, Fitter et al. (1998) have shown that when labeled carbohydrate moves from one plant into an arbuscular mycorrhiza fungus and then to the roots of another plant, the C stays within the fungus and does not actually transfer to the other plant.

Mutualism

Some have assumed mycorrhizas of all kinds are, by definition, mutualistic because of reciprocal exchange of limiting resources (see Jones and Smith 2004 for a discussion of this topic). But such a definition is obviously much narrower than Harley’s. He wrote, “Indeed, the general question ‘Is mycorrhizal infection beneficial to the hosts that possess it?’ is frequently posed, but…is not really susceptible of a satisfactory answer” (Harley 1969). He explained that “[mycorrhizas] grade from clearly mutualistic symbioses to biotrophic parasites” (Harley 1975) and “mycorrhizal associations are separable only with difficulty from many kinds of pathological and non-pathological associations of roots and other plants organs with soil microorganisms” (Harley 1949). Consistent with Harley’s point of view, a few have attempted to broaden the narrow, mutualistic definition of mycorrhizas and to recast them as associations that occupy rather liberal ranges along evolutionary and functional continua (Johnson et al. 1997; Jones and Smith 2004; Koide et al. 2008), but these attempts have not convinced everyone. It strikes us that the controversy concerning the definition of mycorrhizal fungi based on their net effects on their hosts is similar to that which has occurred for what are currently called endophyticFootnote 1 fungi. The term “endophytic” simply indicates that the organisms live inside plant tissues; it is neutral with respect to function. Some have defined endophytic fungi as mutualists or commensals, but it is clear that their net impact on a host is context-dependent and may be negative (Delaye et al. 2013).

Some have convincingly argued that many factors frequently collude to select for mutualisms, especially under stable environmental circumstances (Garrett 1950; Schwartz and Hoeksema 1998). However, a mycorrhiza need not always be mutualistic to be selected for. Selection can occur for associations that are only mutualistic on balance, which is to say, more often than not. Surely ectomycorrhizas exist that are, on balance, mutualistic, especially the set of associations involving certain basidiomycetes and their hosts in the Pinaceae or Cupuliferae (Fagaceae and Betulaceae) (Harley 1969). In these cases of what might be termed the “classic” ectomycorrhiza, the fungi benefit from the carbohydrate obtained from the living host, and the host may benefit from the phosphate obtained from the fungi. While that couplet is frequently recited as a mantra as if it were always true, it need not be true in all instances. Mutualism depends on this reciprocal exchange of limiting resources, and there may be times in natural systems when P does not limit plant growth and carbohydrate does not limit fungal growth. For example, water may limit the growth of both. Thus, ectomycorrhizal associations that are sometimes less than beneficial for the host do occur (Karst et al. 2008).

Conclusion

In this contribution, we have shown that one of the great mycorrhiza researchers of the twentieth century, Jack Harley, and his students anticipated much of our current research. We hope our limited discussion has provided some motivation for reading the older literature. It may reduce the waste of resources “reinventing the wheel”, it may provide context for current research, and it may even help overcome contemporary controversies by providing a point of view that is uncolored by current dogma. Therefore, to our older readers, we suggest that you read the old literature. To our younger readers, we suggest that you read the ancient literature. We are certain that we will all benefit by doing so.