Mycorrhizal Acquisition of Inorganic Phosphorus by the Green-leaved Terrestrial Orchid Goodyera repens

Abstract

Background and Aims

Mycorrhizal fungi play a vital role in providing a carbon subsidy to support the germination and establishment of orchids from tiny seeds, but their roles in adult orchids have not been adequately characterized. Recent evidence that carbon is supplied by Goodyera repens to its fungal partner in return for nitrogen has established the mutualistic nature of the symbiosis in this orchid. In this paper the role of the fungus in the capture and transfer of inorganic phosphorus (P) to the orchid is unequivocally demonstrated for the first time.

Methods

Mycorrhiza-mediated uptake of phosphorus in G. repens was investigated using spatially separated, two-dimensional agar-based microcosms.

Results

External mycelium growing from this green orchid is shown to be effective in assimilating and transporting the radiotracer ³³P orthophosphate into the plant. After 7 d of exposure, over 10 % of the P supplied was transported over a diffusion barrier by the fungus and to the plants, more than half of this to the shoots.

Conclusions Goodyera repens

can obtain significant amounts of P from its mycorrhizal partner. These results provide further support for the view that mycorrhizal associations in some adult green orchids are mutualistic.

INTRODUCTION

With an estimated 20 000 to 35 000 species, the Orchidaceae is the largest family in the plant kingdom (Cribb et al., 2003), its rapid diversification being facilitated by flowers producing prodigious quantities of typically wind-dispersed ‘dust’ seeds (Arditti and Ghani, 2000). Crucial to the success of this strategy is reliance upon symbiotic mycorrhizal fungi that provide the essential organic carbon (C) (Smith, 1966) and nutrients required for seedling establishment. Whilst a minority (approx. 10 %) of orchids never photosynthesize, retaining their myco-heterotrophic mode of nutrition and remaining parasitic upon fungi throughout their lives (Leake, 1994, 2004), most develop green leaves and may cease to depend on their mycorrhiza for carbon. However, as most adult orchids have coarse roots they are likely to remain very dependent upon mycorrhiza for mineral nutrient [particularly phosphorus (P)] uptake as is the pattern in other coarse-rooted plants (Baylis, 1975).

It has been shown recently, for the first time, that in an adult green-leaved orchid Goodyera repens, net carbon flow through orchid mycorrhiza can be from plant to fungus (Cameron et al., 2006). This suggests that after the initially myco-heterotrophic phase of development, the symbiosis in green-leaved adults follows the conventional pattern of mycorrhizal mutualisms in which carbon from the plant is exchanged for mineral nutrients gathered by the fungus. Whilst adult G. repens plants incur a C cost in maintaining their mycorrhizal symbionts, the extent to which this orchid benefits from the provision of mineral nutrient supply from the fungus has not been fully established.

The mycorrhizal fungus supplies the orchid with organic nitrogen (N) (Cameron et al., 2006) and a further study has demonstrated P transfer to juvenile protocorms (Smith, 1966). Alexander et al. (1984) demonstrated that the inflow of [³²P]orthophosphate to adult G. repens plants was less by two orders of magnitude when the roots were treated with fungicide, relative to untreated plants. While this provided circumstantial evidence that the mycorrhizal fungus was involved in P uptake and transfer, direct evidence of such transfer is lacking.

In this study, adult G. repens plants were grown in microcosms composed of separate root and shoot compartments (Cameron et al., 2006) enabling mycelial uptake and transfer of P to the plant to be traced and quantified. By supplying ³³P-labelled orthophosphate to mycorrhizal mycelium growing in symbiosis with the orchid, the hypothesis that the fungus takes up P and facilitates its transfer to the partner plant was tested.

MATERIALS AND METHODS

Clonal mats of Goodyera repens Br. (creeping lady's tresses) were collected from beneath a stand of Pinus sylvestris at Tentsmuir Forest, Fife, Scotland (National Grid Reference: NO503263) and grown under controlled conditions: temperature 15 °C, photoperiod 12 h, photon flux density 150 µmol m⁻² s⁻¹ (at canopy height), relative humidity 85 %. Five similar plants were divided from the clonal mat of stock plants and were surface sterilized in 5 % (v/v) sodium hypochlorite for 1 min, washed four times with sterile distilled water. Each plant was placed in a microcosm with their roots in the upper compartment of a 9-cm-diameter two-compartment Petri dish (Sterilin, Stone, Staffordshire, UK) forming a root and hyphae compartment (RHC). Both the upper compartment (the RHC) and the lower hyphal compartment (HC) of the split-plate Petri dish contained 1 % (w/v) non-nutrient Plantgar agar (Duchefa, Haarlem, The Netherlands), with the antibacterial Novobiocin (50 mg L⁻¹ from Sigma-Aldrich, Gillingham, Dorset, UK), to a depth of 7 mm (Cameron et al., 2006). Shoots protruded from the dishes through a slot cut in the side, and were held in place with a plug of non-absorbent cotton wool. A 1·5-cm-diameter well was cut out of the agar in the HC (Fig. 1). The dishes were incubated under controlled conditions, as previously described. Using the methods described by Bidartondo et al. (2004), molecular analysis of extracts of genomic DNA from axenic cultures of fungi isolated from the plant roots confirmed that the mycorrhizal fungus was Ceratobasidium cornigerum (Cameron et al., 2006).

Fig. 1.

(A) Microcosm used to supply [³³P]orthophosphoric acid to the fungal symbiont of G. repens by means of a hyphal compartment (HC) separated by a diffusion barrier (a, surface-sterilized orchid; b, cotton wool plug; c, split-plate Petri dish with lid and base sealed with sterile anhydrous lanolin; d, fungal mycelium; e, ³³P source). (B) Digital autoradiograph 7 d after [³³P]orthophosphoric acid was added to a well cut into the agar in the hyphal compartment (HC) of the microcosm, showing transport over a diffusion barrier into the root and hyphal compartment (RHC). A photograph showing the exact position of the orchid plant in the microcosm is superimposed. The colour scale indicates the number of counts detected in 1 h in each of the 0·25 mm² pixels.

After 2 weeks, mycelium of C. cornigerum grew out of the plants and covered the agar surface having crossed the plastic barrier that prevents diffusion of elements between the RHC and HC of the microcosm (Cameron et al., 2006). Once the mycelium had reached the well in the HC, 100 µL of [³³P]orthophosphoric acid, containing 4·7 MBq (specific activity 148 TBq mmol⁻¹) equating to 1·05 ng was added to the well and this was filled with cooled, molten agar. These systems have been tested extensively by Cameron et al. (2006) for the potential risk that the compounds introduced into the labelling well could pass into the RHC of the microcosm by diffusion or by tracking the outside of hyphae; no such diffusion or tracking were recorded (Cameron et al., 2006).

After 7 d the orchids were removed. The RHC and HC were lifted intact from each Petri dish and the distribution of ³³P in the fungal mycelium and agar was then imaged using digital autoradiography (Packard Instant Imager; Isotech, Chesterfield, UK). The radioisotope imaging could not be applied to the plants due to the strong quenching effects of their tissues.

The roots and shoots of the orchid and the RHC were oven-dried at 80 °C for 48 h, weighed and digested in 1 mL of concentrated sulphuric acid and heated to 350 °C in a heating block (Grant instruments, Shepreth, Cambridgeshire, UK ) for 5 min. Oxidation to a colourless solution was achieved by addition of 200 µL of 100 volumes hydrogen peroxide (modified from Bowman, 1988). Digests were made up to 10 mL with distilled water, and a 1-mL aliquot was mixed with 10 ml of emulsifying scintillant (Ecosafe, Fisher Biosciences, UK) for liquid scintillation counting of ³³P (Packard Tri-carb 3100TR; Isotech). The ³³P content (mg) of samples was calculated using equation 1 after correcting counts for radioactive decay.

1

where M_33P = mass of ³³P (mg), C_DPM = counts as disintegrations per minute, S_Act = specific activity of the source (Bq mmol⁻¹), D_f = dilution factor (in this case 10) and M_wt = molecular mass (of P).

Differences between treatment means were analysed by ANOVA and Fisher's multiple comparison test using Minitab 13. Where necessary to satisfy the test assumptions, data were arcsine square root transformed. Untransformed means and associated standard errors are presented. Counts of radioactivity in each pixel obtained from the radioisotope imager were exported to Stanford Graphics image analysis software, and converted to colour scale using the spectral plotter function. The final figure was assembled using Adobe Photoshop (version 4.0.1) to allow the photograph of the plant to be superimposed on the colour digital autoradiograph image to show the exact position of the plant and mycorrhizal mycelium.

RESULTS

Autoradiography of the RHC and HC showed that the mycelial networks of C. cornigerum growing out of the roots of the orchid had taken up ³³P (Fig. 1). The ³³P was transferred from the source in the HC through the mycelial network and across hyphal bridges formed over the diffusion barrier (Fig. 1). The ³³P was transported throughout the mycelium, both towards and away from the plants.

The mean total amount of ³³P recovered in the fungus (in the RHC), roots and shoots (0·1 ng) (Fig. 2) accounted for 8·2 % of the ³³P supplied (Table 1). The mean ³³P-content decreased in the step from fungus-to-root (from 6·6 to 0·4 % of the total supplied). However, ³³P-content increased from root-to-shoot (from 0·4 to 0·7 % of the total supplied), the differences between the amounts in external fungal mycelium and in both roots and shoots were significant (P < 0·05). However, the differences between roots and shoots were not significant (P > 0·05; Fig. 2).

Fig. 2.

(A) The amount of ³³P supplied and (B) mean total ³³P transferred across the diffusion barrier by the mycorrhizal mycelial network (lightly shaded column) and the distribution of this between fungus, roots and shoots (deeply shaded columns). Error bars represent + 1 s.e., n = 5 microcosms. Bars sharing the same letter are not significantly different (P > 0·05, one-way ANOVA: d.f. = 3,15; F = 30·86; P < 0·001).

Table 1.

The mean percentage of the ³³P supplied assimilated by the fungus and transferred into the upper compartments in fungal mycelium, orchid roots and shoots and their total, and the percentage of the total ³³P in the upper compartment found in the fungal and plant components

	% of 33P supplied*	% of 33P uptake†
Shoots	0·7 ± 0·5a	6·3 ± 4·6a
Roots	0·4 ± 0·2a	5·4 ± 4·0a
Fungus	6·6 ± 0·6b	83·5 ± 8·6b
Total	8·2 ± 1·1b	n.a.

Rows sharing the same letter codes do not differ significantly.

n.a., Not available.

* ANOVA: d.f. = 3,15; F = 34·18; P < 0·001.

^† ANOVA: d.f. = 2,11; F = 33·29; P < 0·001.

Of the total ³³P transferred across the diffusion barrier, >10 % was found in the orchid, 6·3 % (± 0·5) being found in the orchid shoots, confirming that P had moved from fungus to plant tissues (Table 1). The mean rate of fungus-to-shoot transfer being 1·1 fmol h⁻¹ averaged over the 7-d incubation. The differences between the percentage of the ³³P taken up by the plant–fungal system and present in the external fungal mycelium and in the plant tissues were significant (P < 0·05), but there was no significant difference between roots and shoots (P > 0·05; Table 1).

DISCUSSION

This study is the first to combine spatially explicit tracing of radiolabelled P through orchid mycorrhizal mycelium with quantification of the net amounts of this P distributed between fungus and plant tissues. The earlier pioneering study by Alexander et al. (1984) provided circumstantial evidence that most of the P uptake by G. repens depends upon its mycorrhiza, but in their experiment it was not possible to control for any non-target effects of the fungicide used to provide ‘non-mycorrhizal’ plants. Uptake of ³³P by the mycorrhizal fungus and its subsequent transfer to the orchid by the mycorrhizal mycelium pathway has now been conclusively demonstrated.

The physiological pathways responsible for P flow though orchid mycorrhizal networks and to the partner orchid are unclear. Ashford et al. (1994) demonstrated fluxes of short-chain polyphosphates through the motile tubule and vacuole system present in ectomycorrhizal fungi; although likely, this transport mechanism remains to be proven in orchid mycorrhizal networks. The mechanisms of P transfer to the plant again remain to be fully elucidated. Lysis of fungal pelotons (intra-cellular hyphal coils) may play a role in facilitating nutrient transfer but direct cross-membrane transfer is also likely to be implicated (Smith and Read, 1997).

Transfer of P from fungus-to-host can occur rapidly in mycorrhizal symbioses. Finlay and Read (1986) showed extensive labelling of ecto-mycorrhizal networks and the co-associated Pinus sylvestris plants after as little as 82 h after the fungus had been supplied with ³²P, with the fungus transporting the P over distances of several decimeters. In the present study, extensive labelling of the fungus and co-associated fungus was observed over 7 d. Such rapid P uptake represents further evidence for dynamic fungus-to-plant resource transfer in addition to fungal lysis.

These findings are of considerable importance for understanding the role of mycorrhiza in adult green orchids where there can be a net carbon cost to the plant in maintaining its symbiont (Cameron et al., 2006). The benefits to the orchid in terms of improved access to the two main plant growth-limiting nutrients nitrogen (Cameron et al., 2006) and P (Alexander et al., 1984; the present study) are likely to offset these carbon costs. Establishing whether this represents a mutualistic symbiosis critically relies on investigating the reciprocal benefit to the orchid in return for its carbon outlay. A direct fungal pathway for N flux to the orchid has also been established (Cameron et al., 2006). The present study demonstrates an additional fungal pathway facilitating P transfer to the plant. The hypothesis that orchids and their mycorrhiza can indeed operate in a mutualistic fashion is further supported.