Wide geographical and ecological distribution of nitrogen and carbon gains from fungi in pyroloids and monotropoids (Ericaceae) and in orchids


Author for correspondence:Gerhard Gebauer Tel: +49 921552060 Fax: +49 921552564 Email: gerhard.gebauer@uni-bayreuth.de


  • •  Stable isotope abundance analyses recently revealed that some European green orchids and pyroloids (Ericaceae) are partially myco-heterotrophic, exploiting mycorrhizal fungi for organic carbon and nitrogen. Here we investigate related species to assess their nutritional mode across various forest and climate types in Germany and California.
  • •  C- and N-isotope signatures of five green pyroloids, three green orchids and several obligate myco-heterotrophic species (including the putatively fully myco-heterotrophic Pyrola aphylla) were analysed to quantify the green plants’ nutrient gain from their fungal partners and to investigate the constancy of enrichment in 13C and 15N of fully myco-heterotrophic plants from diverse taxa and locations relative to neighbouring autotrophic plants.
  • •  All green pyroloid and one orchid species showed significant 15N enrichment, confirming incorporation of fungi-derived N compounds while heterotrophic C gain was detected only under low irradiance in Orthilia secunda. Pyrola aphylla had an isotope signature equivalent to those of fully myco-heterotrophic plants.
  • •  It is demonstrated that primarily N gain from mycorrhizal fungi occurred in all taxonomic groups investigated across a wide range of geographical and ecological contexts. The 13C and 15N enrichment of obligate myco-heterotrophic plants relative to accompanying autotrophic plants turned out as a fairly constant parameter.


More than 400 species of vascular plants have only traces of chlorophylls, and are almost completely dependent on their mycorrhizal fungal symbionts for supplies of C and N. Such ‘myco-heterotrophic’ (MH; Leake, 1994) plants occur frequently in the Orchidaceae, and one of their ecological attributes is that they are independent of irradiance and able to colonize deeply shaded forest habitats. In contrast to the majority of autotrophic orchids that form mycorrhizas with fungi of the polyphyletic rhizoctonia group (Bernard, 1909), many MH orchids form mycorrhizas with fungi that are ectomycorrhizal (Bidartondo et al., 2004; Selosse et al., 2004).

Fungal tissues and the nutrients derived from heterotrophic sources are enriched in the heavy stable isotopes of N (Gebauer & Dietrich, 1993) and C (Gleixner et al., 1993) relative to those of accompanying autotrophs. Trudell et al. (2003) observed that the δ13C and δ15N values of a diverse range of MH species were similar to those of their ectomycorrhizal fungal associates. Similarly, Gebauer & Meyer (2003) concluded on the basis of natural stable isotope abundance that some green and hence putatively autotrophic orchids gained C and N from their fungal symbionts in addition to the C obtained by photosynthesis and the N from the soil. This mixed mode of nutrition, which involves the acquisition of C and N through association with fungi as well as through autotrophic processes, has been referred to as partial myco-heterotrophy (Gebauer & Meyer, 2003).

While partial myco-heterotrophy has been well documented in the Orchidaceae (Gebauer & Meyer, 2003; Bidartondo et al., 2004; Leake, 2004; Selosse et al., 2004; Bidartondo, 2005; Gebauer, 2005; Julou et al., 2005; Abadie et al., 2006), its occurrence beyond monocotyledonous plants was found only recently within the Ericaceae (Tedersoo et al., 2007). The dicotyledonous family Ericaceae contains lineages of very closely related tribes which, as in the case of Orchidaceae, contain some species that are fully MH and others that are autotrophic. The tribes Monotropeae and Pterosporeae (Kron et al., 2002) (subsequently referred to here as ‘monotropoids’) consist entirely of ectomycorrhizal basidiomycete-dependent MH species that form distinctive monotropoid mycorrhizas (Duddridge & Read, 1982), while the tribe Pyroleae (subsequently referred to here as ‘pyroloids’) consists, with the exception of one putatively MH species (Pyrola aphylla), of green and hence putatively autotrophic species, forming ectendo or arbutoid mycorrhizas (Smith & Read, 1997). All other tribes of the Ericaceae are, as far as is known, fully autotrophic. Tedersoo et al. (2007) identified several endophytic and ectomycorrhizal fungal associates on pyroloid roots, and found that the C and N nutrition of four green pyroloid species (Chimaphila umbellata, Orthilia secunda, Pyrola chlorantha, Pyrola rotundifolia) from two Estonian boreal coniferous forests relies partly on fungi.

This study on nutrient gain from fungi extends the spectrum of species investigated to a wide range of geographical and ecological circumstances, and is the first work to include isotope data of P. aphylla– the sole putatively obligate MH pyroloid. Material of five green pyroloid species (C. umbellata, O. secunda, P. chlorantha, Pyrola minor, Pyrola picta) and three green orchids (Epipactis gigantea, Goodyera oblongifolia, Platanthera leucostachys), and of neighbouring autotrophic and MH plants, was collected from 12 sites in Germany, northern California and southern California varying from temperate to Mediterranean-type climate, and from dense and very dark deciduous forest sites at low elevations to open mixed evergreen forest sites at elevations between 1500 and 1900 m asl. Mycorrhizal fungi of the green pyroloids investigated were identified using molecular analysis, and leaf isotope signatures of pyroloid and orchid species were compared to autotrophic and MH reference plants. For each pyroloid or orchid that showed evidence of nutrient gain from fungi, the percentage of fungi-derived N and C was calculated. To standardize one parameter of these calculations, the stability of enrichment in 15N and 13C of obligate MH species from various taxonomic groups and locations (Corallorhiza maculata, Neottia nidus-avis, albino Cephalanthera damasonium, albino Cephalanthera longifolia, Monotropa hypopitys, Pterospora andromedea, Sarcodes sanguinea) relative to accompanying obligate autotrophic plants was investigated, yielding a mean enrichment factor. This uniform enrichment factor represents a suitable reference point for such calculations if fully myco-heterotrophic plants are missing at a study site.

Materials and Methods

Study sites

Samples were collected from six forest sites in Germany (G1–G6) and six in California (C1–C6). The German sites are located in north-east Bavaria (49°40′ to 49°55′ N and 10°49′ to 11°32′ E) at 300–520 m elevation with mean annual precipitation of 650–1000 mm, mean annual temperature of 6–9°C (German weather service, http://www.dwd.de) and a large gradient in relative light availability to the forest understorey. Site G1 is a dense and very dark broadleaf forest dominated by Fagus sylvatica with a sparse and patchy cover of understorey vegetation. Site G2 is a clearing in the same forest. Sites G3–G5 are located in an open Pinus sylvestris stand with a species-rich understorey. The overstorey is interspersed with some Quercus rubra, Quercus robur and Fagus sylvatica trees. Site G6, an ancient sandpit, is an open stand mainly composed of P. sylvestris, Picea abies, F. sylvatica, Q. robur and Corylus avellana with an understorey dominated by mosses. Five of the Californian sites (C1–C5) are located in the San Bernardino and San Jacinto Mountains in southern California (34°04′ to 34°14′ N and 116°47′ to 117°14′ W) at elevations between 1500 and 1900 m asl, and they are open mixed pine forests dominated by Pinus ponderosa and Quercus kelloggii with a sparse understorey of herbs and shrubs that vary in composition between sites (Allen et al., 2007). Mean annual precipitation accounts for 600–1100 mm and mean annual temperature is 9–15°C (http://www.cimis.water.ca.gov, http://www.jamesreserve.edu). The six German sites (Harrison et al., 2000), and the sites in southern California (Allen et al., 2007) are characterized by considerably high atmospheric N-deposition rates. Site C6 is a Sierran mixed conifer forest situated at the University of California's Blodgett Experimental Forest in northern California (38°54′ N, 120°39′ W) at an elevation of approx. 1340 m, a mean annual precipitation of 1600 mm and mean annual temperature of 11.2°C (http://ecology.cnr.berkeley.edu/properties/blodgett.html). The forest is dominated by Abies concolor and Pseudotsuga menziesii with a sparse understorey and a midstorey composed mainly of Q. kelloggii and A. concolor saplings.

Sampling scheme and species investigated

Sampling was performed in 2001 (southern California, June/July) and in 2005 (Germany, June/July; northern California, August). Five 1-m2 plots were selected for all 12 sites, with the exception of C6 where three plots were selected. Each plot included a green pyroloid and/or orchid species and at least two autotrophic reference plant species (excepting pyroloids and orchids) of different functional groups and growth forms. Most of the plots (except for sites G6 and C5) also contained fully myco-heterotrophic plants. Leaf material of all plants was taken following the criteria described by Gebauer & Meyer (2003). In total, 48 plant species (n = 386), including five green pyroloids, three green orchids, one MH pyroloid species, two MH orchids and three MH monotropoids, were sampled (Table S1 in Supplementary material). The reference species were grouped into five functional types according to their mycorrhiza (Wang & Qiu, 2006) or their ability to utilize atmospheric N2: ectomycorrhizal plants (ECM, n = 99), ericoid mycorrhizal plants (ERI, n = 20), arbutoid mycorrhizal plants (ARB, n = 23), plants forming arbuscular mycorrhizas or nonmycorrhizal plants (AM/NON, n = 109), and plants potentially living in symbiosis with N2-fixing microbes (FIX, n = 15). Sampled pyroloids, orchids and monotropoids were categorized as either green (GP/GO, n = 64) or MH (MHP/MHO/MHM, respectively, n = 56). At each site, soil samples from the organic layer were taken (n = 100). Root samples were collected from all green pyroloid individuals at sites G1–G6 and C6 for identification of mycorrhizal fungi.

Light measurements

At the beginning of July, relative light availability under the forest canopy (%) at sites G1–G6 was calculated by comparing PAR measurements (Quantum Sensor, Li-Cor, Lincoln, NE, USA) at the forest ground and outside the forest.

Stable isotope abundance analysis

Leaf and soil samples were oven-dried and ground to a fine powder. Relative N and C isotope abundances of all samples from sites G1–G6 and C1–C5 were measured using a dual element analysis mode with an elemental analyser coupled to a continuous flow isotope ratio mass spectrometer as described by Bidartondo et al. (2004). Samples from northern California (C6) were analysed in a dual element analysis mode with a continuous flow mass spectrometer (20/20 mass spectrometer, PDZ Europa Ltd, Northwich, UK). Measured abundances are denoted as δ values, calculated according to the following equation:

δ15N or δ13C = (Rsample/Rstandard – 1) × 1000 (‰)

where Rsample and Rstandard are the ratios of heavy isotope to light isotope of the samples and the respective standard. Standard gases were calibrated with respect to international standards using the reference substances N1 and N2 for the nitrogen isotopes and ANU sucrose and NBS19 for the carbon isotopes. The samples from site C6 were analysed using standards N2 and NIST 1577 bovine liver, or NIST 1547 peach leaf and corn flour, respectively.


Data sets from the three major regions (Germany, southern California and northern California) were tested for differences in δ15N and δ13C using the Kruskal–Wallis nonparametric test and Bonferroni-corrected Mann–Whitney U-tests for post hoc comparisons. Afterwards, δ values within resulting groups were tested for normal distribution and homogeneity of variances, and compared by anova and Tukey's honestly significant difference.

Calculation of C and N gains from fungi

For all green pyroloids and orchids that were significantly different from reference plants in δ15N and/or δ13C, a linear two-source isotopic mixing model, as described by Gebauer & Meyer (2003), was used to calculate the relative contribution of N or C derived from fungal material to the N or C content of the green pyroloids and orchids (%xdf where x= N or C). The model is based on individual δ values of the green pyroloids and orchids (δxGP or δxGO), mean δ values of co-occurring reference plants (δxREF), and the mean enrichment factor (ɛ) of fully MH plants (ɛMH-REF = δxMH – δxREF; Table 1): %xdf = (δxGO/GP – δxREF)/ɛMH-REF × 100. Mean %xdf values were tested for significant difference from zero (no N or C gain from fungi) using Student's t-test. Data are given as means ±1 SD.

Table 1.  Relative enrichment for 15N and 13C in leaves of fully myco-heterotrophic orchids (MHO), pyroloids (MHP) and monotropoids (MHM)
SiteMyco-heterotrophic species (functional group)15N ɛMH-REF (‰)13C ɛMH-REF (‰)
  1. The table includes samples from this and other studies: *, Gebauer & Meyer (2003);†, Bidartondo et al. (2004);‡, Julou et al. (2005);§, Abadie et al. (2006). Enrichment factors (ɛMH-REF) represent the difference between mean δ values in leaves of reference plants and mean δ values of myco-heterotrophic plants collected at the respective sites. Mean ɛMH-REF are used for further calculations (Table 3). For numbers of replicates and site codes see Table S1. Values from the literature are based on n = 5.

Fagus sylvatica forest (G1)Neottia nidus-avis (MHO) 7.67.9
Fagus sylvatica forest clearing (G2)Neottia nidus-avis (MHO)10.57.9
Pinus sylvestris/Quercus robur forest (G3–G5)Monotropa hypopitys (MHM)12.17.9
Mixed pine forest (C1)Corallorhiza maculata (MHO)11.96.7
Sarcodes sanguinea (MHM)15.74.8
Mixed pine forest (C2)Corallorhiza maculata (MHO)13.67.8
Sarcodes sanguinea (MHM)14.36.6
Mixed pine forest (C3)Pterospora andromedea (MHM) 9.55.3
Mixed pine forest (C4)Pterospora andromedea (MHM) 8.94.7
Sarcodes sanguinea (MHM)12.04.3
Mixed conifer forest (C6)Pterospora andromedea (MHM) 8.06.5
Pyrola aphylla (MHP)13.87.8
Fagus sylvatica forest (S.E. Germany)Neottia nidus-avis (MHO)*12.28.4
Neottia nidus-avis (MHO)11.68.1
Quercus robur/Corylus avellana forest (N. France)Cephalanthera damasonium albino (MHO)13.99.2
Juniperus communis/Pinus sylvestris shrubland (Estonia)Cephalanthera longifolia albino (MHO)§11.86.6
Mean ± 1 SD 11.7 ± 2.36.9 ± 1.5

Molecular identification of mycorrhizal fungi

Individual fine root tips from single Pyrola, Orthilia and Chimaphila plants that appeared to be colonized by fungi during microscopical examination were separated from the rhizome and placed in lysis buffer. Samples were frozen and thawed three times before grinding the softened tissue with a micropestle. Genomic DNA of samples from German sites (G1–G6) was extracted following methods described elsewhere (Gardes & Bruns, 1993) but using GeneClean (Q-BioGene, Carlsbad, CA, USA) for DNA binding and purification. For DNA extraction from P. picta roots (C6), Qiagen DNeasy kit (Qiagen, Valencia, CA, USA) was used following the manufacturer's instructions, with the extracted DNA eluted into 50 µl AE buffer. Using PCR, the nuclear ribosomal internal transcribed spacer (ITS) region was amplified with the fungal-specific primers ITS1F and ITS4 and PCR conditions described by Gardes & Bruns (1993). Positive PCR products were purified using QIAquick 96 kits (Qiagen) (samples from G1–G6), or 0.5 µl Exosap-IT (GE Healthcare, Piscataway, NJ, USA) combined with 3.5 µl PCR product and 1 µl dH2O (samples from C6). DNA sequencing was performed on an ABI3100 Genetic Analyzer using BigDye ver. 3.1 chemistry (Applied Biosystems, Foster City, CA, USA) and absolute ethanol/EDTA precipitation. Electrophoretograms were checked using sequence navigator ver. 1.0.1 (Applied Biosystems) or sequencher ver. 4.2.2 (Gene Codes Corporation, Ann Arbor, MI, USA). All samples with strong PCR amplification of single templates were blasted in GenBank to ascertain taxonomic affinity. For C6 samples, all the sequences that appeared to contain more than one template were cloned using the TOPO TA Kit for Sequencing (Invitrogen, Carlsbad, CA, USA) following the manufacturer's instructions. Positive clones were picked, amplified with plasmid primers T3 and T7, sequenced, and visually aligned at 95% similarity using sequencher ver. 4.2.2. From the clone library, the longest fragment of each dominant haplotype was then blasted in GenBank to ascertain taxonomic affinity to the genus level. The GenBank accession numbers are EF372399–EF372410 (samples from G1–G6) and EF101760–EF101777 (C6 samples).


Comparison of δ15N and δ13C between functional groups

Within most of the sites, leaves of plants belonging to the functional groups ECM, AM/NON, ERI and ARB showed similar 15N and 13C abundances. Consequently, data of these groups were pooled for each site, forming one group of reference plants (REF) and describing one endpoint of the linear two-source isotopic mixing model. Because of their association with N-fixing bacteria, FIX plants can use atmospheric N215N = 0‰), in addition to soil mineral N, which leads to δ15N values generally closer to zero and makes them unsuitable as N references. Data from fully MH plants (contributing to the other endpoint of the mixing model) were pooled as well for each site (MH).

Except for G6 and C5, which did not include MH plants, all sites showed significant differences in N and C isotopic composition between functional groups (Table S2). Compared to the respective autotrophic references, MH plants were significantly and highly enriched in 15N and 13C at all sites. All green pyroloids from Germany and California were consistently and significantly enriched in 15N compared to the respective REF plants (Figs 1, 2). The δ15N values in leaves of some green pyroloids –O. secunda at a dark beech forest (Fig. 1a), P. chlorantha (Fig. 1d), P. picta (Fig. 2c,d,f) – were even not statistically different from those of MH plants at the same sites. All other green pyroloids had an N isotopic composition intermediate between the REF and the MH group. The green orchid P. leucostachys (C2, C5) was also significantly enriched in 15N compared to the respective autotrophic references, and depleted in 15N compared to the MH plants, whereas the δ15N values of E. gigantea (C4) and G. oblongifolia (C6) were not significantly different from those of REF plants. Concerning 13C abundances, the effects on isotopic composition within green pyroloids and green orchids were weak. Except for O. secunda from a very dark beech forest (G1), which had an isotopic position intermediate between REF and MH plants (Fig. 1a), no significant differences in δ13C were detected between green pyroloids, green orchids and their references.

Figure 1.

Mean (± 1 SD) values of δ15N and δ13C in leaves of four green pyroloid species, two fully myco-heterotrophic species, 12 fully autotrophic reference species and in soil samples collected from six forest sites in north-eastern Bavaria, Germany: (a) Fagus sylvatica forest G1; (b) F. sylvatica forest clearance G2; (c–e) mixed Pinus sylvestris forest G3–G5; (f) mixed conifer forest G6. Symbols represent classification according to functional groups: ectomycorrhizal reference plants (ECM, □), arbuscular mycorrhizal or nonmycorrhizal reference plants (AM/NON,○), ericoid mycorrhizal reference plants (ERI, ▿), green pyroloids (GP, •), myco-heterotrophic orchids (MHO, ▪), myco-heterotrophic monotropoids (MHM, ▴). For numbers of replicates see Table S1. Error bars are missing if n < 3. Plant species: Ap, Acer pseudoplatanus; Ca, Corylus avellana; Cm, Convallaria majalis; Ea, Epilobium angustifolium; Fs, Fagus sylvatica; Fv, Fragaria vesca; Hs, Hieracium sylvaticum; Pa, Picea abies; Qr, Quercus robur; Rs, Rubus saxatilis; Sa, Sorbus aucuparia; Vm, Vaccinium myrtillus.

Figure 2.

Mean (± 1 SD) values of δ15N and δ13C in leaves of two green pyroloid species, three green orchids, four fully myco-heterotrophic species, 22 fully autotrophic reference species and in soil samples collected from five forest sites in southern California: (a–e) mixed pine–oak forest C1–C5; and from one site in northern California: (f) mixed conifer forest C6. Symbols represent classification according to functional groups: ectomycorrhizal reference plants (ECM, □), arbuscular mycorrhizal or nonmycorrhizal reference plants (AM/NON, ○), arbutoid mycorrhizal references (ARB, ▵), ericoid mycorrhizal references (ERI, ▿), reference plants potentially living in symbiosis with N2-fixing microbes (FIX, ◊), green orchids (GO, ◆), green pyroloids (GP, •), myco-heterotrophic orchids (MHO, ▪), myco-heterotrophic pyroloids/monotropoids (MHP/MHM, ▴). For numbers of replicates see Table S1. Error bars are missing if n < 3. Plant species: Ac, Abies concolor; Am, Acer macrophyllum; Ap, Arctostaphylos patula; Ar, Alnus rhombifolius; Bc, Bromus carinatus; Cd, Calocedrus decurrens; Ee, Elymus elymoides; Eg, Elymus glaucus; Ec (part c), Epilobium canum; Ec (part b), Eriophyllum confertifolium; Ge, Glyceria elata; Lb, Linnea borealis; Lc, Lotus crassifolius; Pa, Pteridium aquilinum; Pl, Pinus lambertiana; Pm, Pseudotsuga menziesii; Pp, Pinus ponderosa; Qk, Quercus kellogii; Rs, Ribes sp.; Ro, Rhododendron occidentale; Ss, Smilacina sp.; Sl, Stipa lettermanii.

Comparison of δ15N and δ13C between sites

According to the Kruskal–Wallis test (P < 0.001 for δ15N and δ13C) and post hoc comparisons, leaf δ15N values of reference plants from the three major regions (Germany, southern California, northern California) differed significantly from each other and were positively correlated with the mean δ15N values of the respective soil samples. Because of the Mediterranean climate, δ13C signatures of reference plants from southern California were significantly less negative than those of northern California and Germany. Differences were also found between samples from southern and northern California (Mann–Whitney U-test, P < 0.001 after Bonferroni correction) but not between samples from Germany and northern California (Mann–Whitney U-test, P = 1 after Bonferroni correction). Subsequent anovas within regions were conducted with respect to the results of the Mann–Whitney U-tests.

As shown by the site anovas (Tables S3 & S4), 15N and 13C abundances in the soil and in leaves of REF plants varied in a highly significant manner between sites, even within major regions (P < 0.001). These variations demonstrate that data from different sites cannot be pooled and must be analysed separately. However, despite varying δ values between sites in reference plants and soil samples, the enrichment factor of obligate MH plants relative to accompanying autotrophic plants (ɛMH-REF) remained considerably stable, irrespective of taxonomic groups or climatic conditions.

Mycorrhizal fungi of pyroloid roots

Light microscope observations revealed that roots of all pyroloid species investigated were colonized by hyphae forming a Hartig net and intracellular coils restricted to the para-epidermal cells. Whereas no fungal sheath was found for species collected at German sites, colonization in P. picta (C6) resembled ectendomycorrhizae, with a sparse, netted fungal mantle. As described by Robertson & Robertson (1985), we observed individual cells with varying degrees of fungal colonization. Accordingly, ‘pyroloid’ mycorrhizas have been described as arbutoid (Robertson & Robertson, 1985; Molina et al., 1992) with an intracellular infection similar to that seen in ericoid mycorrhizas (Read, 1983). Molecular identification of fungal partners revealed that both basidiomycetes and ascomycetes associate with pyroloid species (Table 2). The majority of the fungi detected are obligate mycorrhizal basidiomycetes that form ectomycorrhizas with trees. In contrast to root samples of P. picta, only a few positive amplifications were obtained from samples of pyroloid species collected in Germany. This may be caused by degradation during the longer storage of root samples from sites G1–G6 before molecular analyses.

Table 2.  Mycorrhizal fungi detected in roots of pyroloid species from six German sites and one northern Californian site
Pyroloid speciesSitenMycorrhizal fungi
  1. Obligate ectomycorrhizal lineages are shown in bold. *, Lineages that contain some ectomycorrhizal taxa. Site codes are given in Table S1. n, Number of individuals of which root samples were taken. Numbers in parentheses after each fungus indicate the number of plants it was associated with.

Orthilia secundaG15Piloderma sp. (1), Ascomycete sp.* (1)
Orthilia secundaG25Sebacina sp. (1), Pezizales sp.* (2)
Orthilia secundaG35Piloderma spp. (2), Wilcoxina sp. (1)
Pyrola chloranthaG45Russula sp. (1)
Chimaphila umbellataG55Tomentella sp. (1), Phialophora sp.* (1)
Pyrola minorG65Laccaria sp. (1)
Pyrola pictaC65Cortinarius sp. (1), Hebeloma sp. (1), Piloderma spp. (2), Rhizopogon spp. (2), Russula spp. (3), Thelephoraceae spp. (2), Tomentella spp. (2), Helotiales sp.* (1), Phialocephala spp. (1), Wilcoxina sp. (1), Cadophora sp. (1), Leptodontidium sp. (1)

N and C gain of GP and GO from fungal association

The relative contribution of N or C derived from fungi to the N or C content of P. leucostachys and the green pyroloid species was calculated via the linear two-source mixing model, using the mean enrichment factor of fully MH plants (ɛMH-REF = 11.7 ± 2.3‰ for 15N; ɛMH-REF = 6.9 ± 1.5‰ for 13C), which was determined using the present data set and all suitable isotope data for MH plants described in the literature (Table 1).

The results show that P. leucostachys and all green pyroloid species gained a remarkable amount of N from their fungal associations – varying between 25 ± 7% and 32 ± 14% for P. leucostachys, and ranging from 30 ± 9% (P. minor) to 83 ± 26% (P. picta) for the pyroloid species (Table 3). By contrast, a significant C gain (28 ± 12%) could be detected only for O. secunda growing under extremely low light conditions at site G1.

Table 3.  Percentage of nitrogen (Ndf ± 1 SD) and carbon (Cdf ± 1 SD) derived from mycorrhizal fungi in the leaves of green orchids and pyroloids, calculated from δ values and mean enrichment factors (Table 1) using a linear two-source isotopic mixing model
Pyroloid/orchid speciesSiteLight availability (%)Ndf (%) ± 1 SDCdf (%) ± 1 SD
  1. Significance levels for deviations from zero based on a Student's t-test: **, P < 0.01; ***, P < 0.001. In accordance with the model assumptions no calculations were made for species that were not statistically different in their δ values from the respective references, or if replication was insufficient. For numbers of replicates and site codes see Table S1. Relative light availability under the forest canopy (%) was measured at sites G1–G6.

Orthilia secundaG1  156 ± 21**28 ± 12**
Orthilia secundaG210062 ± 15**
Orthilia secundaG3 1559 ± 27**
Pyrola chloranthaG4 1574 ± 29**
Chimaphila umbellataG5 1540 ± 5***
Pyrola minorG6 2530 ± 9**
Platanthera leucostachysC2 32 ± 14**
Pyrola pictaC3 83 ± 26**
Pyrola pictaC4 70 ± 11***
Platanthera leucostachysC5 25 ± 7***


Mycorrhizal fungi of pyroloid roots

Most of the fungi identified as mycorrhizal associates of pyroloid roots are known also to be ectomycorrhizal associates of forest trees. The pattern of symbiotic associations in this group of plants is thus seen to be analogous to that in fully MH (Taylor & Bruns, 1997; McKendrick et al., 2002; Selosse et al., 2002) and partially MH orchids (Bidartondo et al., 2004; Selosse et al., 2004), and in monotropoids (Bidartondo & Bruns, 2001, 2002; Trudell et al., 2003), which are specialized on particular fungi of a broad spectrum of ectomycorrhizal fungal taxa. It is also consistent with that recorded for members of the genera Arbutus and Arctostaphylos, the arbutoid mycorrhizas of which have been known for some time to be produced by a diverse range of ectomycorrhizal fungi (Molina & Trappe, 1982; Mühlmann & Göbl, 2006). Specialization of pyroloid species on ectomycorrhizal fungi is consistent with the situation reported by Tedersoo et al. (in press), who observed the presence of ectomycorrhizal fungi in the pyroloid species O. secunda, P. chlorantha and C. umbellata. However, while the present study found a diverse range of ectomycorrhizal associates in these and related pyroloid species, Tedersoo et al. (in press) suggest a preference for fungi of the genus Tricholoma. It will be important to determine the extent of specificity occurring in the tribe Pyroleae.

Some aspects of the ecology of pyroloid plants provide an indirect indication that there may be specificity in their symbiotic relationships. They are known to be particularly susceptible to disturbance. Thus, reports from the Pacific north-west (Halpern & Spies, 1995), Norway (Haugset et al., 1996), Canada (Timoney et al., 1997; A. Schmiedinger and co-workers, unpublished data) and central Alaska (Rees & Juday, 2002) indicate that species such as P. picta, P. chlorantha, Pyrola asarifolia, C. umbellata and Chimaphila menziesii are lost from sites that experience disturbance from logging or burning. In addition to the loss of any competitive advantage that these species might possess from tolerance to the shade typical of primary forests, logging is known to cause changes in the ectomycorrhizal key fungal species (Jones et al., 2003). Among other environmental factors to which both pyroloid and monotropoid plants are known to be sensitive is anthropogenic N deposition. Nitrogen inputs of up to 50 kg N ha−1 yr−1 in the San Bernardino Mountains have been shown to be associated with loss of P. picta and of P. andromedea since 1973 (Allen et al., 2007). Again, the loss of fungal diversity associated with N deposition is well documented (Wallenda & Kottke, 1998; Taylor et al., 2000), and the possibility that elimination of key fungal symbionts contributes to the loss of these species is worthy of investigation.

Isotopic evidence for N and C gain from mycorrhizal fungi

All the pyroloid species examined were significantly enriched in 15N relative to co-associated reference plants, indicating the incorporation of fungi-derived N. In the species with the highest level of such enrichment, P. picta, as much as 83% (±26%) of N was obtained by this route. No correlation could be found between N gain from mycorrhizal fungi and irradiance at the site of origin. The observations of N gain in the green pyroloids are consistent with those seen in the four pyroloid species examined by Tedersoo et al. (2007). Hence the N demand of all hitherto investigated pyroloid species is supplied by fungi-derived N.

A significant MH gain of C (28 ± 12%) could be detected in only one species (O. secunda), and then only at one of its sites (G1), where ground-level irradiance was particularly low. As no C gain was detected in O. secunda when growing at higher irradiances, it is unlikely that the single instance of C gain is species-dependent. A lower C than N gain may be attributed to greater complexity of pathways for C gain and loss than for N (Bidartondo et al., 2004). The study suggests that the negative relationship between the quantity of C gain from fungi and light availability, previously described by Gebauer (2005) in partially MH orchids, may also be found in pyroloids, but further analyses are required to confirm this. Our failure, with the exception of the one species and site, to detect C gain from fungal sources stands in marked contrast to the findings of the recent study by Tedersoo et al. (2007), in which significant C gains of up to 68% were recorded in four pyroloid species. It is not clear whether the discrepancy between the two studies arises from differences in the light climates to which the plants were exposed, but it seems unlikely in view of the broad inter- and intrasite range of irradiances found in the German and Californian locations examined in the present study. Hashimoto et al. (2005) have provided preliminary evidence of C transfer from the overstorey tree Larix kaempferi to Pyrola incarnata through mycelia of co-associated thelephoroid fungi. Rigorous quantification of such C flows under controlled conditions of irradiance, preferably with parallel characterization of foliar isotopic signatures, is urgently needed.

Among the putatively autotrophic orchids, no significant C gain from fungi could be detected. However, in the case of 15N, P. leucostachys, in contrast to E. gigantea and G. oblongifolia, was significantly enriched. Hence it can be hypothesized that P. leucostachys has switched its mycorrhizal partnership from a rhizoctonia-type fungus to an ectomycorrhizal associate.

All the fully MH orchids and monotropoids showed strong enrichment in 15N and 13C relative to neighbouring autotrophs. This reflects their dependence on N and C sources that have been metabolized by ectomycorrhizal fungi. The mean enrichment factor (ɛMH-REF) calculated from the broad spectrum of fully MH species for 15N (11.7 ± 2.3‰) and 13C (6.9 ± 1.5‰) can be useful for linear mixing model calculations, especially to obtain unbiased values for N and C gains of partially MH species if obligate MH plants are missing at a study site (as at sites G6 and C5). Standard deviations of ɛMH-REF may result to some degree from the MH plants’ preference towards certain fungi, or may be a legacy of specificity during underground development from seed.

The isotopic analyses in this work indicate that P. aphylla has isotope signatures typical of a fully MH plant. If more extensive analysis of this species confirms this finding it will be the only fully MH species, besides the monotropoid lineage, in the Ericaceae. Furthermore, this study demonstrates that, irrespective of site elevation, climate or forest type, (ecto)mycorrhizal fungi supply significant proportions of the required N in two more pyroloid and one more orchid species. With the exception of O. secunda at the notably shaded site, no evidence was obtained for fungal augmentation of plant C supply. Further experimental investigation, particularly of the roles of irradiance in determining the polarity of C movement, are required before the nature and full extent of the dependence of pyroloid upon myco-heterotrophy can be appreciated.


The authors would like to thank Dr Martin I. Bidartondo (Imperial College London and Royal Botanic Gardens, Kew, UK) for his valuable support and guidance in fungal DNA analysis of the root samples collected in Germany. We thank Margarete Wartinger (Universität Bayreuth, Germany) and Stefania S. Mambelli (University of California, Berkeley, CA, USA) for skilful technical assistance in sample preparation and mass spectrometry, Andrew Sanders (Herbarium of UC Riverside) for competent determination of plant material collected in southern California, and numerous members of the Arbeitskreis Flora Nordostbayerns (in particular Peter Ille and Hermann Bösche) for their helpful hints on German Pyrola and Chimaphila sites. Kind permission for the plant sampling by the respective regional authorities is gratefully acknowledged. The present study contributes to the DFG project GE 565/7-1.