This study has established, for the first time, that the mycorrhizal associations of a green-leaved terrestrial orchid can function mutualistically, with photosynthate passing from the plant to the fungus, in return for mineral nutrients (N) passing from the fungus to the plant. It is surprising that so fundamental an aspect of orchid mycorrhizal functioning has not hitherto been established, despite the fact that orchids were amongst the first group of plants in which benefits of mycorrhizal associations were recognized (Bernard, 1899). More than a century later, despite the enormous interest in the cultivation, conservation and biodiversity of members of the family Orchidaceae, the crucial question as to whether the C invested by orchid mycorrhizal fungi in the germination and establishment of seedlings can potentially be later repaid in green-leaved species by photosynthate is, only now, at least partially answered. Our studies therefore have far-reaching implications for understanding the co-evolution of orchids and their mycorrhizal fungi and the functional significance of these associations.
C and N movement from fungus to plant
Both C and N, when supplied as double-labelled glycine to distal parts of the extraradical mycelium of G. repens, were transferred through the hyphae to the roots, the rhizomes and the green shoots of the plant. These findings are novel in the case of both elements. Alexander & Hadley (1985) investigated the movement of 14C, supplied to mycorrhizal mycelium associated with G. repens in the form of labelled insoluble carbohydrate, to heterotrophic protocorms and into the rhizomes and green shoots of plantlets of G. repens. The autotrophic plantlets, in contrast to the protocorms, failed to obtain C from the applied substrate, even after a prolonged period of exposure to darkness designed to enhance their C demand. On the basis of these results, Alexander & Hadley (1985) concluded that the orchid mycorrhiza underwent a physiological change which stopped the movement of C from the fungus to the plant once the orchid became photosynthetic. Clearly, the results obtained in the current study do not support such a view.
As the size and stage of development of the plantlets used in this study were similar to those used by Alexander & Hadley (1985), it is unlikely that phenological differences can account for the apparent anomalies. Most probably the differences are attributable to the nature of the substrates and environmental conditions. In their experiments, Alexander & Hadley (1985) employed the N-free substrate cellulose as the C source as this was considered to simulate field conditions. Cellulose is indeed a potentially predominant source of C for heterotrophs growing in soil, but there are many other possible sources of the element. Amongst these, C sources linked to N in amino acid and peptide forms must be considered. It is now recognized that amino acids, whether in soil pools, rhizosphere exudates or plant tissues, provide an excellent source of both C and N for fungi (Näsholm & Persson, 2001; Taylor et al., 2004) and there is good evidence from other types of mycorrhizal symbiosis, particularly the ecto-types predominating in forest soils of the kind supporting G. repens, that glycine-derived C is absorbed by mycorrhizal fungi and transferred to the host plant (Taylor et al., 2004). Benefits to the autotroph in terms of improved N nutrition have been documented when this and other soil-borne amino compounds have been provided to mycorrhizal fungi (Abuzinadah & Read, 1988; Näsholm & Persson, 2001; Taylor et al., 2004). The possibility of augmentation of autotroph C supplies has also been considered (Abuzinadah & Read, 1989). Our results suggest that C and N derived from amino acid is assimilated by the orchid mycorrhizal fungus and transferred to the autotrophic plant.
The much lower ratio of excess 13C:15N in the external mycelium compartment compared with glycine source (Fig. 3) indicates that the glycine is used primarily as a C source by the fungus. The change in enrichment ratio must reflect loss of 13CO2 through fungal respiration. A low 13C:15N ratio was also found in the roots and rhizomes, and fungal biomass is a significant component of these tissues. However, the plant shoots had an excess 13C:15N ratio that was more than an order of magnitude higher than in the external mycelium and root components and was also significantly higher than that of the glycine source (P < 0.05). There is no established pathway for inorganic N transfer from fungus to plant in mycorrhizal associations, but the transfer of amino acid is well established (Martin & Botton, 1993). Two distinct pathways of C and N transfer from fungus to plant are suggested by the shoot isotopic enrichment ratios. First, cotransport of glycine-derived 15N and 13C to the plant is likely to occur as amino acid (Smith & Read, 1997). If the main amino acid transferred from the fungus to the plant is not glycine, but, for example, glutamine, the transamination could account for the change in 13C:15N ratio. Alternatively, if the glycine is transferred intact, an additional C transfer pathway must exist to explain the greater relative enrichment of shoots in 13C than the source. The established pathway of C flux from the fungus to the plant in nonautotrophic orchid plantlets is via the fungal sugar trehalose (Smith, 1967). Fungal deamination of glycine and its conversion into sugars, a small amount of which is passed onto the plant, is consistent with the suggestion that the fungus is primarily using the glycine as a C source in our experiment.
We can exclude leakage of 13CO2 from the root and fungal compartment as the cause of the higher 13C:15N ratio of shoots compared with the glycine source. The control plants, with their shoots intermingling with those of the experimental plants, were used to derive the background 13C and 15N values so the calculations of excess 13C account for any theoretical leakage and fixation.
It has long been assumed that green orchids are released from their dependence on fungal C as adults (Smith & Read, 1997); however, recent evidence, including that of the current study, suggests that some C flow to the orchid is maintained into the autotrophic adult life stage (Gebauer & Meyer, 2003; Bidartondo et al., 2004). We found, using both double-labelled 13C-15N glycine (Fig. 2) and 14C-labelled glycine (Fig. 4) significant fungus-to-plant transfer of labelled C. Of the total labelled C found in the biomass, in each case 30% and 9% were found in the shoots after 36 h (13C-15N glycine) and 72 h (14C glycine), respectively. It is important to recognize that both of these measurements are based on single harvests, and that the rate of turnover of the glycine-derived labelled C is likely to be much higher in fungal than in plant tissues, so that the isotopic enrichment and pool size of the isotopes in the plants and fungal components will be temporally dynamic. A comprehensive budget for C allocation from fungus to plant would therefore need to study both the rates of turnover and transfer over a time course. Nonetheless, the data presented here are consistent with the suggestion that many green orchids may have the capacity for mixotrophy (Julou et al., 2005), whereby they can obtain C through photosynthesis (autotrophically) and heterotrophically from their fungal association. In nature, this may enable green orchids to grow at or below the compensation point (Julou et al., 2005).
The transfer of N from fungus to plant appears not to have been previously reported in orchids. Excess glycine-derived 15N appeared in the external mycelium, roots and shoots (Fig. 2). Over the 14-d incubation period, of the total 15N transferred across the Petri dish divider, only 2% was passed to shoots, whereas 20% was found in the roots. The rate of fungal-to-plant transfer of N is likely to be dependent on the activity and nature of the plant–fungal interfaces in the roots (e.g. whether young active peloton cells are abundant, or if the fungus is mainly being digested by the plant).
Transfer of C from plant to fungus
Our demonstration that C, fed to the shoots as 14CO2, was readily assimilated and transferred to the rhizome and onwards to the extending external mycelial system of the mycorrhizal fungus (Figs 5, 6), is of importance both in confirming an autotrophic capability in the plant and in showing that the products of photosynthesis can readily be used to support nutrient foraging by the fungus. It is clear that the 14C present in the fungus is not derived from scavenging of root exudate C (Supplementary material, Fig. S1a,b) as only 0.3% of the 14C fixed by the plant was detected in the agar in contact with roots of mycorrhizal plants before external mycelium grew out of the root. Root exudation is approximately an order of magnitude lower than the 2.6% of the 14C fixed by plants that was detected in external mycorrhizal mycelia. Transfer of C from plant to fungus must therefore occur through the plant cell–fungal interface of the mycorrhiza.
These results are in marked contrast with those reported in earlier studies. Hadley & Purves (1974) detected very little transfer of C to rhizomes, even up to 25 d after 14CO2 labelling of shoots of the orchid, and no radioactivity was detected in emerging mycelium. Similarly, Alexander & Hadley (1985) found that < 0.5% of the 14C fixed by shoots was found in the rooting substrate and mycelia, even after a period of postlabelling incubation in darkness designed to increase sink strength. This is an order of magnitude lower than the proportion of photosynthate C allocated into external mycelium we now report, but similar to the allocation of C to root exudation (Supplementary material Fig. S1). The reasons for the differences in the results obtained from what appear to be experiments of similar design, are unclear, but may relate to the levels of physiological activity, and hence sink sizes, of the respective tissues in the experiments. We took particular care to minimize heating of our plants and their mycorrhizal networks using evaporative cooling, whilst providing moderate rates of irradiance and high humidity, as would be found in the field. In the earlier studies, photorespiration is likely to have been greater. Certainly, under the conditions described in the current experiments, the mycorrhizal mycelium developed extensively and proved to constitute a vigorous sink.
Whatever the reasons for the disparity in the outcomes between the earlier and the current experiments, the implications of the present results for the functioning of the orchid mycorrhizal system in nature are profound. The earlier studies concluded that orchids differ fundamentally from other types of mycorrhizal plants with respect to the movement of C from plant to fungus. However, the present study shows that orchid mycorrhizal fungal growth and activity can be supported by current photosynthate supplied by the green plant, as in arbuscular, ecto- and ericoid mycorrhizas (Smith & Read, 1997). Under these conditions, it is reasonable to assume that, as in the other types of mycorrhiza, allocation of photosynthate to the mycelium facilitates foraging for key nutrients. Indeed, Alexander et al. (1984) showed that phosphorus inflow to mycorrhizal G. repens was 100 times higher than that in the same plants treated with fungicide to inhibit mycorrhizal functioning. The present study also reveals that the mycorrhiza has a highly significant role in the orchid's uptake of N, the other major plant growth-limiting nutrient.
Clearly the C demands of the fungus, and to a lesser extent of the green plantlet, can be partially met by exploitation of organic substrates, as in the case reported here. The extent to which C acquired heterotrophically from the substrate can subsidise the demands of the mycelium and of the mycorrhiza remains to be quantitatively determined, but it is important to bear in mind that the pathways for such acquisition, and hence for bidirectional flow, of C are present. Thus, such bidirectional C flow complicates the interpretation and quantification of the net direction of C flow, regardless of whether the polarity is in favour of the plant or the fungus. The relative importance of the two pathways is therefore likely to vary both temporally and spatially, particularly in orchids such as G. repens which characteristically inhabit forest floor habitats with dappled shade. The demonstration (Gebauer & Meyer, 2003; Bidartondo et al., 2004) that the C and N stable isotope signatures of orchids growing in such environments are enriched relative to those of nonorchid species is indicative of the presence of distinctive pathways for C and N acquisition and transfer.
In the current study, we attempted to estimate the net direction of C flow by combining results from Expts 2 and 3, in the first of which 14C-labelled glycine was supplied to the external mycelium and in the other 14CO2 was supplied to the plant shoots (Fig. 7). In these experiments, the plants were of uniform size and developmental stage, based on fixed selection criteria (see the section entitled Plant and fungal material in the Materials and Methods). The 14C fixed by the orchid shoots and allocated to the external mycorrhizal mycelium was 2.6% of net 14C fixation. This figure does not include any 14C respired by the mycelium and excludes C allocation to fungal structures within the root and rhizome. The proportion of net fixation allocated by the orchid to the external mycelium is almost identical to values recently reported from a laboratory study of C allocation to external mycelium of arbuscular mycorrhizal fungi in a grassland turf (Johnson et al., 2002), where it was found that 3.4% of net assimilation was allocated to the external mycorrhizal mycelium 72 h after 14CO2 pulse labelling of the shoots, comprising 2.7% in mycelia, the remainder (0.7%) being respiration (Johnson et al., 2002). The transfer of C to the fungus is likely to be further underestimated as the plant will have begun to be CO2 starved by the end of the labelling period. We monitored the 14CO2 concentrations at intervals during the period of shoot exposure to the gas, and the plants were harvested within hours of this being almost depleted, to minimize C starvation.
Figure 7. Comparison of fungus-to-orchid and orchid-to-fungus carbon transfer in Goodyera repens mycorrhiza. Transfer of 14C from fungal mycelium supplied with 14C glycine is shown for roots (░) and shoots (▪). Note that the values for roots will include fungus in the roots. The orchid-to-fungus transfer of 14C fixed from 14CO2 by the shoots and allocated to roots (░) and external mycorrhizal mycelium (□) is shown. The values for roots comprise 14C in both plant and fungal tissues. The data are from separate experiments carried out under the same conditions. Bars sharing the same letters are not significantly different [P > 0.05, one-way analysis of variance (anova), degrees of freedom (df) = 3,20; F = 30.14; P < 0.001]. n = 3–6.
Download figure to PowerPoint
In the present study, our reciprocal estimates of fungus-to-plant 14C allocation into roots using [14C]glycine will overestimate C flux into root tissues as much of the 14C may be retained in the fungus rather than passed onto the plant cells (Fig. 7). Despite the underestimate of the plant-to-fungus C transfer and overestimate of the fungus-to-plant C transfer, we found that C allocation from the plant to the external mycelium is significantly greater (P < 0.05) than C allocation from the external mycelium to the roots (Fig. 7). Comparison of the plant-to-fungal and fungal-to-plant C allocation to shoots and external mycelium indicates that the net flux of C is from the plant to the fungus (Fig. 7). Combined with the evidence of fungus-to-plant N transfer in this orchid mycorrhiza, we conclude that the association is mutualistic in our experiments. Simultaneous labelling of the mycelial and shoots systems with different C isotopes and under different conditions of irradiance will be required to elucidate formally the eventual net polarity of these C fluxes under varying light conditions. Experiments to determine net fluxes of C under these varying conditions are currently underway.