The ectomycorrhizal specialist orchid Corallorhiza trifida is a partial myco-heterotroph


  • Katja Zimmer,

    1. Laboratory of Isotope Biogeochemistry, Bayreuth Center of Ecology and Environmental Research (BayCEER), University of Bayreuth, D–95440 Bayreuth, Germany
    Search for more papers by this author
  • Cornelia Meyer,

    1. Laboratory of Isotope Biogeochemistry, Bayreuth Center of Ecology and Environmental Research (BayCEER), University of Bayreuth, D–95440 Bayreuth, Germany
    Search for more papers by this author
  • Gerhard Gebauer

    1. Laboratory of Isotope Biogeochemistry, Bayreuth Center of Ecology and Environmental Research (BayCEER), University of Bayreuth, D–95440 Bayreuth, Germany
    Search for more papers by this author

Author for correspondence:
Gerhard Gebauer
Tel:+49 921552060
Fax:+49 921552564


  • • The leafless, circumboreal orchid Corallorhiza trifida is often assumed to be fully myco-heterotrophic despite contrary evidence concerning its ability to photosynthesize. Here, its level of myco-heterotrophy is assessed by analysing the natural abundance of the stable nitrogen and carbon isotopes 15N and 13C, respectively.
  • • The mycorrhizal associates and chlorophyll contents of C. trifida were investigated and the C and N isotope signatures of nine C. trifida individuals from Central Europe were compared with those of neighbouring obligate autotrophic and myco-heterotrophic reference plants.
  • • The results show that C. trifida only gains c. 52 ± 5% of its total nitrogen and 77 ± 10% of the carbon derived from fungi even though it has been shown to specialize on one specific complex of ectomycorrhizal fungi similar to fully myco-heterotrophic orchids. Concurrently, compared with other Corallorhiza species, C. trifida contains a remarkable amount of chlorophyll.
  • • Since C. trifida is able to supply significant proportions of its nitrogen and carbon demands through the same processes as autotrophic plants, this species should be referred to as partially myco-heterotrophic.


Within the Orchidaceae, > 100 species are nonphotosynthetic and depend on carbon (C) and nitrogen (N) supplies from associated fungi. The majority of these ‘myco-heterotrophic’ orchids are subterranean for most of their life cycle and their organs are well adapted to this habit in function and morphology (Leake, 1994). The orchid genus Corallorhiza Gagnebin (Epidendroideae) comprises 10 leafless species of temperate-boreal terrestrial orchids. All of them are distributed in North and Central America, with exception of the circumboreal pale coral root orchid Corallorhiza trifida (Freudenstein, 1992). Species of the genus Corallorhiza are, in general, known to be fully myco-heterotrophic. Again, C. trifida is an exception. Although the ‘mycotrophic’ nature of this species was already recognized in 1898 (Jennings & Hanna), many questions concerning the nutritional mode of C. trifida have been raised, by pigment analyses and assimilation experiments (Montfort & Küsters, 1940) and by comparative studies of the plastid DNA (Freudenstein & Doyle, 1994). However, despite the contrary evidence C. trifida remains known as an obligate myco-heterotrophic plant (Downie, 1943; Zelmer & Currah, 1995; McKendrick et al., 2000a,b).

Natural stable isotope abundances can be used to assess the nutritional mode of orchid species (Gebauer & Meyer, 2003). This method is based on two findings: fungal tissues are enriched in the heavy stable isotopes of N (Gebauer & Dietrich, 1993) and C (Gleixner et al., 1993) relative to accompanying autotrophic plants; and obligate myco-heterotrophic plants show isotopic signatures similar to those of their fungal associates (Trudell et al., 2003). Furthermore, Gebauer & Meyer (2003) found that some green and hence putatively autotrophic orchid species are enriched in 15N and 13C compared with neighbouring autotrophic plants but depleted in the heavy isotopes relative to obligate myco-heterotrophic plants. They concluded that these species use a mixed nutritional mode, where the acquisition of C and N through mycorrhizal fungi subsidizes the nutrient supply through autotrophic processes and referred to these plants as partially myco-heterotrophic.

Furthermore, there is evidence that the nutritional mode of orchid species is linked to association with certain functional groups of fungi. Dearnaley (2007) recently summarized the literature on orchid mycorrhiza. While roots of fully autotrophic orchids are generally associated with diverse saprotrophic, rhizoctonia-forming basidiomycete fungi (Bernard, 1909), several obligate myco-heterotrophic orchids are known to be highly specialized on a phylogenetically narrow range of fungi that simultaneously form ectomycorrhizas with roots of neighbouring trees (Taylor & Bruns, 1997; Taylor et al., 2003; McKendrick et al., 2002; Selosse et al., 2002). Consistent with intermediate isotopic signatures of partially myco-heterotrophic orchids, Bidartondo et al. (2004) could provide a ‘missing link in the evolution’ and show that these putatively green orchids are also connected to ectomycorrhizal fungi of a diverse range and that some species may simultaneously associate with rhizoctonia-forming fungi. The high specialization towards certain ectomycorrhizal fungi in species belonging to the genus Corallorhiza (Zelmer & Currah, 1995; Taylor & Bruns, 1997, 1999; McKendrick et al., 2000b) supports the assumption that C. trifida is obligate myco-heterotrophic. However, if C. trifida is not fully myco-heterotrophic this would be an example of a partially myco-heterotrophic species highly specialized on ectomycorrhizal fungi.

This study is the first to investigate the nutritional mode of the obviously greenish C. trifida by using natural stable isotope abundance analysis. The application of a linear mixing model provides not only a qualitative conclusion but is also suited to quantitatively estimate the level of myco-heterotrophy of a plant (Gebauer & Meyer, 2003). Isotopic signatures together with data of chlorophyll contents and molecular identification of fungal associates will help to answer the question of whether C. trifida is fully myco-heterotrophic.

Materials and Methods

Study site

Samples were collected from a forest site located in NE Bavaria, Germany (49°40′N and 11°23′E) at 522 m elevation with mean annual precipitation of 820 mm, mean annual temperature of 8°C (German weather service, The site is a dense broadleaf forest dominated by Fagus sylvatica with a sparse and patchy cover of understory vegetation (Cephalanthero-Fagion). Soil is lithic leptosol originating from Jurassic dolomite with a shallow organic layer and a pH of 7.2 (0–5 cm) measured in H2O.

Sampling scheme and investigated species

Sampling for isotope ratio analysis was performed in 2004 and 2005 (June). Scale-like leaves and parts of the stem including flowers or seed capsules, respectively, were collected of 17 obligate myco-heterotrophic plants (Leake, 1994; Bidartondo, 2005) (Neottia nidus-avis, Monotropa hypopitys) and of nine individuals of Corallorhiza trifida (Fig. 1). As reference, leaf material of three to four obligate autotrophic plant species (n = 68) was taken in close spatial proximity (within 1 m2) to each of the Neottia, Monotropa and Corallorhiza individuals, following the criteria described by Gebauer & Meyer (2003). In total, 10 plant species (n = 94) were sampled, including Neottia nidus-avis (L.) Rich. (n(2004) = 4, n(2005) = 9), Monotropa hypopitys L. (n(2004) = 4), Corallorhiza trifida Châtel. (n(2004) = 4, n(2005) = 5), Fagus sylvatica L. (n(2004) = 8, n(2005) = 10), Convallaria majalis L. (n(2004) = 8, n(2005) = 10), Acer pseudoplatanus L. (n(2004) = 8, n(2005) = 5), Fragaria vesca L. (n(2004) = 4), Sorbus aucuparia L. (n(2005) = 5), Hieracium sylvaticum (L.) Grufb. (n(2005) = 5) and Rubus saxatilis L. (n(2005) = 5). For identification of mycorrhizal fungi, two root samples were collected from each of the four C. trifida individuals in 2004. Furthermore, the whole aboveground biomass of 10 C. trifida individuals was taken for quantitative chlorophyll analysis (23 May 2006).

Figure 1.

Corallorhiza trifida growing in a dense Fagus sylvatica forest in eastern Hesse, Germany. (a) Several flowering individuals on 5 May 2005; (b) seed capsules on 27 May 2005. Courtesy of Marco Klüber.

Stable isotope abundance analysis

Leaf and stem samples were oven-dried and ground to a fine powder. Relative N and C isotope abundances were measured using a dual element analysis mode with an elemental analyser coupled to a continuous flow isotope ratio mass spectrometer as described in Bidartondo et al. (2004). Measured abundances are denoted as δ values, which were 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 by using the reference substances N1 and N2 for N isotopes and ANU sucrose and NBS 19 for C isotopes, provided by the International Atomic Energy Agency (Vienna, Austria).


Variation in δ15N and δ13C between the two sampling years was assessed using Student's t-tests. Afterwards, pooled data were tested for differences in δ values using the Kruskal–Wallis nonparametric test and Bonferroni-corrected (Holm, 1979) Mann–Whitney U-tests for post hoc comparisons.

Calculation of C and N gains from fungi

As described by Gebauer & Meyer (2003), a linear two-source isotopic mixing model was used to calculate the approximate relative contribution of N or C derived from fungal material to the N or C content of C. trifida (%xdf with x as N or C, respectively). The model is based on individual δ values of C. trifidaxCT), mean δ values of co-occurring autotrophic reference plants (δxREF) and on the mean enrichment factor ɛ of sampled neighbouring fully myco-heterotrophic (MH) plants (ɛMH-REF = δxMH − δxREF):

%xdf = (δxCT − δxREF)/ɛMH-REF × 100

Data are given as means ± 1 SD.

Molecular identification of mycorrhizal fungi

From each of the four C. trifida individuals (2004), two root sections colonized by fungi were taken and placed in lysis buffer. Samples were frozen and thawed three times before grinding the softened tissue with a micropestle. Genomic DNA was extracted following methods described elsewhere (Gardes & Bruns, 1993) but using GeneClean (Q-BioGene, Carlsbad, CA, USA) for DNA binding and purification. Using polymerase chain reaction (PCR), the nuclear ribosomal internal transcribed spacer (ITS) region was amplified with the fungal-specific primers ITS1F and ITS4 and the PCR conditions described in Gardes & Bruns (1993). Positive PCR products were purified using QIAquick 96 kits (Qiagen, Valencia, CA, USA). DNA sequencing was performed on an ABI3100 Genetic Analyzer using BigDye v.3.1 chemistry (Applied Biosystems, Foster City, CA, USA) and absolute ethanol/ethylenediaminetetraacetic acid (EDTA) precipitation. Electrophoretograms were checked using Sequence Navigator v.1.0.1 (Applied Biosystems). All samples with strong PCR amplification of single templates were blasted in GenBank to ascertain taxonomic affinity. The GenBank accession number is EF471313.

Chlorophyll content

Before extraction of chlorophylls a and b (Chla and b), stem height and mass of the 10 C. trifida individuals was determined. N,N′ethylformamide was added to the fresh, ground material (whole aboveground biomass) and samples were kept in the dark at −23°C for 8 d. After centrifugation, absorbance of the supernatants was spectrophotometrically measured at 646.8, 663.8 and 750 nm. Chlorophyll concentrations (µg ml−1) were calculated according to the following equations (Porra et al., 1989): Chla = 12.00 (A663.8 − A750) – 3.11 (A646.8 − A750); Chlb = 20.78 (A646.8 − A750) – 4.88 (A663.8 − A750). The Chla + b contents are given in µg gFW−1.

Results and Discussion

Mycorrhizal association

Molecular identification of fungal partners revealed that all of the eight roots were exclusively colonized by fungi belonging to the obligate ectomycorrhizal basidiomycete genus Tomentella (Thelephoraceae). No other mycorrhizal associations were detected.

These results are consistent with previous findings. Zelmer & Currah (1995) first isolated a clamp-bearing basidiomycete from pelotons of C. trifida roots and demonstrated that the same fungus is ectomycorrhizal on Pinus contorta seedlings. Taylor (1998) observed associations between thelephoroid fungi and C. trifida in North America, and studies comparing DNA sequences of fungi on C. maculata with those of ectomycorrhizal fungi on adjacent tree roots have confirmed that these coral root orchids are indeed associated with ectomycorrhizal fungi (Taylor & Bruns, 1997). McKendrick et al. (2000b) found that the specificity of C. trifida towards fungi exclusively belonging to the Thelephora–Tomentella complex applies from the earliest stages of seed germination through adulthood and flowering.

Both C. maculata and C. mertensiana are known to form mycorrhizas exclusively with fungi in the Russulaceae (Taylor & Bruns, 1999; Taylor et al., 2004) and the desert orchid Hexalectris spicata is specialized on ectomycorrhizal Sebacinaceae (Taylor et al., 2003). The relationship of these fungi to their ectomycorrhizal partners, however, is unlikely to be as specific (Zelmer & Currah, 1995; Gardes & Bruns, 1996).

Since it has been estimated that c. 15% of net C fixation by ectomycorrhizal trees is allocated to their ectomycorrhizal fungal partners (Finlay & Söderström, 1992), McKendrick et al. (2000a) concluded that the switch in achlorophyllous orchids from soil-inhabiting rhizoctonia-type mycorrhizal association, to one involving ectomycorrhizal fungi might be based upon better access of the latter to C supplies.

Chlorophyll content

Corallorhiza trifida individuals contained on average 26 ± 11 µg gFW−1 of Chla + b (Table 1) which accounts for only 1% of the chlorophyll amount detected in green leaves of Cephalanthera damasonium (Julou et al., 2005). However, the area-to-mass ratio of stems and leaves is highly different and relating to the same mass, stems contain considerably more nonassimilating tissues than leaves. Thus, the chlorophyll content in stems of the leafless orchid C. trifida must be lower than the chlorophyll content in green leaves of plants. This also becomes apparent from the close (P < 0.001), negative correlations between Chla + b content and fresh weight of each C. trifida individual (r2 = 0.896, Fig. 2), or between Chla + b content and stem height (r2 = 0.893, not shown). In 1940, Montfort & Küsters found that inflorescences of C. innata R.Br. (= C. trifida) contained 33% (and stems with young fruits even 60%) of the chlorophyll content detected in inflorescences of the fully autotrophic orchid Listera ovata. Cummings & Welschmeyer (1998) determined the pigment composition for 10 species of putatively achlorophyllous angiosperms including two orchids (Cephalanthera austinae and Corallorhiza maculata) by high-pressure liquid chromatography (HPLC). They detected chlorophyll a in all taxa, but chlorophyll b was only detected in Corallorhiza. Compared with the total content of chlorophyll and chlorophyll-related pigments in C. maculata (26.75 ng gFW−1) (Cummings & Welschmeyer, 1998), C. trifida contains a three orders of magnitude higher amount of chlorophyll (Table 1). Furthermore, the mean chlorophyll a : b ratio of C. trifida (Table 1) is similar to that found in other C3 plants (Larcher, 2003).

Table 1.  Fresh weight, stem height, chlorophyll contents (Chla, Chlb, Chla + b) and chlorophyll ratios (Chla/b) of the aboveground biomass of 10 Corallorhiza trifida individuals collected from a dense Fagus sylvatica forest in northeastern Bavaria, Germany
Fresh weight (g)Stem height (cm)Chla (µg gFW−1)Chlb (µg gFW−1)Chla+b (µg gFW−1)Chla/b
  1. Means ± 1 SD are shown in bold type.

0.411 ± 0.12812.6 ± 2.418.2 ± 8.47.9 ± 2.926.1 ± 11.22.2 ± 0.3
Figure 2.

Correlation of chlorophyll a + b content and fresh weight of the aboveground biomass of 10 Corallorhiza trifida individuals collected from a dense Fagus sylvatica forest in northeastern Bavaria, Germany: y = −83.12x + 60.267; r2 = 0.896; P < 0.001.

If photosynthetic function of a plant is altered or absent, concomitant changes in the plastome may have occurred (Palmer et al., 1988). Following this assumption, Freudenstein & Doyle (1994) examined the variation in plastid DNA among species of Corallorhiza. They found deletions of genes relevant for photosynthesis in all Corallorhiza species investigated with the exception of C. trifida and concluded that Corallorhiza may be a genus ‘on its way’ to a heterotrophic existence. However, more-recent studies have shown that the deletion of photosynthesis genes in the plastid genome is not necessarily linked to the trophic strategy of the plants (Randle & Wolfe, 2005; Young & dePamphilis, 2005) and C. trifida frequently is still described as a nonphotosynthetic plant.

Isotopic signature and nutrient gain from the fungal partner

Based on Bonferroni-corrected (Holm, 1979) Mann–Whitney U-tests, highly significant differences existed between the isotopic signatures of all groups tested (obligate autotrophic plants, REF; C. trifida individuals, CT; obligate myco-heterotrophic plants, MH). Compared with the REF group (mean δ15N = −6.1 ± 1.2‰, mean δ13C = −32.0 ± 1.3‰), myco-heterotrophic species were significantly enriched by 11.2 ± 3.1‰ in 15N and by 8.3 ± 1.0‰ in 13C (Fig. 3), which is in accordance with the enrichment factors observed in other obligate myco-heterotrophic plants relative to accompanying autotrophic plants (Zimmer et al., 2007). Individuals of C. trifida were also highly enriched in both, 15N and 13C, compared with the autotrophic reference plants showing the incorporation of fungal-derived organic compounds.

Figure 3.

Mean δ15N and δ13C values in the aboveground biomass of Corallorhiza trifida (CT, circles), two obligate myco-heterotrophic plant species (MH, upward-pointing triangle) and in leaves of seven fully autotrophic species (REF, downward-pointing triangle) collected from a dense Fagus sylvatica forest in northeastern Bavaria, Germany. Boxes are defined by standard deviations of the mean δ values from each group (REF, CT, MH). For numbers of replicates see the Materials and Methods section. Plant species: Ap, Acer pseudoplatanus; Cm, Convallaria majalis; Ct, Corallorhiza trifida; Fs, Fagus sylvatica; Fv, Fragaria vesca; Hs, Hieracium sylvaticum; Mh, Monotropa hypopitys; Nn, Neottia nidus-avis; Rs, Rubus saxatilis; Sa, Sorbus aucuparia.

However, relative to obligate myco-heterotrophic plants, C. trifida was significantly depleted in the heavy stable isotopes of N and C (Fig. 3). The latter is different from findings in C. maculata, which has isotope signatures typical of a fully myco-heterotrophic plant (Zimmer et al., 2007). Hence, in contrast to other species of the genus, aboveground organs of C. trifida are photosynthetically active. Therefore, flowering adults of this species are not fully but only partially myco-heterotrophic. This does not exclude full myco-heterotrophy during belowground phases of this orchid's life cycle. The individuals investigated in this study also use high amounts of soil borne N. Based on estimations from the two-source linear mixing model, C. trifida gains c. 52 ± 5% of its N from fungal association, and approx. 77 ± 10% of its total C demand is derived from fungi. Although these estimated values are not validated through direct measurements of photosynthetic C gain in the individuals sampled, our results confirm previous findings by Montfort & Küsters (1940) who observed photosynthesis in C. innata (= C. trifida) by measuring the plant's CO2 exchange. They found that inflorescences and especially stems with maturing fruits were able to compensate for respiratory losses of CO2 through assimilation and calculated a quotient Q (Q = assimilation/respiration) of 2.2 for the plants including young fruits, indicating a positive CO2 balance. The detection of a higher C than N gain from fungi (Student's t-test, P < 0.001) is unusual for partially myco-heterotrophic plants (see Gebauer & Meyer, 2003; Bidartondo et al., 2004; Zimmer et al., 2007) and might be caused by a N-limited fungal partner. Since several Tomentella species are wood inhabiting (Küffer & Senn-Irlet, 2005) and may therefore have lower N concentrations than those living on humus (Gebauer & Taylor, 1999) C. trifida might be forced to supply its N demand through higher uptake of soil-borne inorganic N.

As it is has now been shown that C. trifida is able to supply its N and C demand through autotrophic processes, this leafless orchid should no longer be referred to as fully or obligate myco-heterotrophic. A possibly similar situation has been shown for the Mediterranean orchid Limodorum abortivum (Girlanda et al., 2006). This also has reduced leaves, associates predominantly with narrow clades of ectomycorrhizal fungi (Russula spp.) and contains photosynthetic pigments. However, its photosynthetic activity was found to be insufficient to compensate for respiration in adult plants (Girlanda et al., 2006).


We would like to thank Dr Martin I. Bidartondo (Imperial College London and Royal Botanic Gardens, Kew, UK) for valuable support in fungal DNA analysis and Nicole A. Hynson (University of California, Berkeley, USA) for critical comments on an earlier version of the manuscript. We also thank Marco Klüber for courtesy of professional pictures of Corallorhiza trifida. This work was supported by the German Research Foundation and contributes to the DFG project GE 565/7-1. Kind permission for orchid sampling by the Regierung von Oberfranken is gratefully acknowledged.