The chlorophyll-containing orchid Corallorhiza trifida derives little carbon through photosynthesis


Author for correspondence:
Duncan D. Cameron
Tel:+44 (0) 114 2220066


  • • While measurements of tissue stable isotope signatures and isotope mixing models have suggested that the green orchid Corallorhiza trifida is photosynthetically active and hence only partially mycoheterotrophic, these assumptions have not been validated by direct analysis of carbon assimilation.
  • • The photosynthetic capabilities of three orchid species assumed on the basis of the indirect methods or chlorophyll content to have differing trophic strategies: Neottia nidus-avis (fully mycoheterotrophic), Cephalanthera damasonium (partially autotrophic), C. trifida (partially autotrophic), as well as saplings of an autotrophic tree, Fagus sylvatica, were investigated by combining the determination of chlorophyll content and fluorescence, with direct measurement of the potential for CO2 assimilation using 13C isotope tracers in the field.
  • • Chlorophyll content and fluorescence values were indicative of ineffective photochemical processes in Neottia and reduced efficiency of photochemical processes in Corallorhiza. These differences are reflected in the mean assimilation rates of 13CO2 of 594 ± 129, 331 ± 72, 12.4 ± 2.4 and 7.3 ± 0.9 µg g−1 h−1 for Fagus, Cephalanthera, Corallorhiza and Neottia, respectively.
  • • Our study, while confirming the fully mycoheterotrophic status of Neottia and the partially autotrophic condition in Cephalanthera, also demonstrates under field conditions that Corallorhiza is physiologically closer to the fully mycoheterotrophic condition than has previously been recognized.


The overwhelming majority of plants form mutualistic symbioses with soil fungi, termed mycorrhizas, in which the plant supplies fixed carbon (C) to the fungal symbionts in return for the provision of mineral nutrients by the fungal partner (Smith & Read, 2008). The functional status of these symbioses in orchids has, however, been controversial. It is accepted that all orchids begin their life cycle with a mycoheterotrophic (sensu Leake, 1994) growth phase in which the fungal symbionts provide C and mineral nutrients to the orchid seedling (Smith, 1966; McKendrick et al., 2000a) without obvious benefit to themselves. It is also thought that, with the exception of a small proportion of species (c. 1%) that retain the fully mycoheterotrophic (achlorophyllous) condition in adulthood, the green shoots of orchids emerging aboveground have the potential for autotrophy. Cameron et al. (2006, 2008a) showed that the green forest orchid Goodyera repens was able to engage in a mutualistic symbiosis with its fungus partner, the partnership enabling the plant to repay the C invested in it during its early achlorophyllous stage. However, the generality of this observation in other green orchids remains to be elucidated. The extent to which the green shoots are actually photosynthetic is less clear in those orchid species, such as Corallorhiza trifida Chatel. in which leaves have been reduced to scales but green stems and capsules appear to retain some potential for autotrophic activity.

Progress towards identification of the sources of C and nitrogen (N) acquired by orchids has been provided by measurements of the natural abundance of these elements in their tissues (Gebauer & Meyer, 2003). These reveal that fully mycoheterotrophic orchids have distinctively enriched δ13C (and δ15N) signatures relative to those seen in species that are autotrophic at maturity. Zimmer et al. (2008) showed that in Corallorhiza the natural abundance of tissue 13C was slightly depleted (δ13C = −25.6) relative to that seen in the neighbouring chlorophyll-free orchid Neottia nidus-avis13C = −24) and the ericaceous herb Monotropa hypopitys13C = −22), both of which are generally accepted to be fully mycoheterotrophic. By contrast, Corallorhiza, Neottia and Monotropa were all shown to be significantly enriched in 13C relative to co-occurring green autotrophic reference plants (δ13C = −32). These findings, by themselves, are indicative of a rather low photosynthetic capability in Corallorhiza. However, using data obtained by the application of a two-source mixing model proposed by Gebauer & Meyer (2003), Zimmer et al. (2008) went on to calculate that up to 33% of the carbon gained by Corallorhiza may be derived from photosynthetic activity. These authors recognized the need to validate such estimated values through direct measurements of photosynthetic C gain. The need for such validation has been further highlighted by Barrett & Freudenstein (2008) who confirmed the presence of the plastid-encoded Rubisco large subunit gene, rbcL, in C. trifida as well as in its closest relatives in the genus, and called for direct measurements of the potential of these plants to express photosynthetic activity.

Here, by in situ field measurements in the same population that was used by Zimmer et al. (2008), we determined the ability of C. trifida and of co-occurring species known to be either fully (N. nidus-avis) or partially mycoheterotrophic orchid (Cephalanthera damasonium) or autotrophic (small saplings of Fagus sylvatica), to fix atmospheric CO2. By coupling the measurement of chlorophyll fluorescence and content, we re-evaluate the nutritional status of this orchid.

Materials and Methods

Field site

All experiments were undertaken at a forest site located in northeast Bavaria, Germany (49°40′ N and 11°23′ E) at 522 m elevation with mean annual precipitation of 820 mm and mean annual temperature of 8°C (German weather service, The site is a dense broadleaf forest dominated by F. sylvatica L. with a sparse and patchy cover of understory vegetation. 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 (Zimmer et al., 2008). Total chlorophyll extractions, measurements of chlorophyll fluorescence parameters and 13CO2 pulse chase experiments were performed in May 2008.

Chlorophyll content

The youngest fully expanded leaf was removed from four individuals of N. nidus-avis L. Rich. and C. damasonium (Miller) Druce, four saplings of Fagus and the whole stem of eight individuals of Corallorhiza. Shoots were harvested and kept on ice in the dark until extraction of chlorophyll (within 1 h). Dry weights were estimated using the FW : DW ratio of additional harvested leaves/stems and the surface area estimated using the FW : area ratio of the same leaves/stems (data not shown). Leaves or stems were ground in a mortar and pestle with a small amount of acid-washed sand (as an abrasive) and 5 ml of 80% ice-cold acetone. The mortar and pestle was washed out twice with a further 2 ml of acetone and transferred to a centrifuge tube. The samples were centrifuged at 8000 g for 5 min and the supernatant diluted to 10 ml total volume with 80% ice-cold acetone. The optical density of the supernatant was measured at 645 and 663 nm using a Hitachi U-2001 spectrophotometer.

Chla (mg l−1) = (12.7 × OD663) – (2.69 × OD645) (Eqn 1)
Chlb (mg l−1) = (22.9 × OD645) – (4.68 × OD663) (Eqn 2)

The chlorophyll concentration (mg l−1 of extract) was calculated according to Arnon (1949) using Eqns 1 and 2 and re-expressed as mg of chlorophyll per cm2 (surface area) to facilitate comparison of stem data collected from Corallorhiza with leaf data from the other species. Surface area of the Corallorhiza stem was calculated as a truncated cone.

Chlorophyll fluorescence

The maximum and quantum yields (Fv/Fm) and steady-state quantum yield of photosystem II (ΦPSII) of the youngest fully expanded leaf of four individuals of Cephalanthera and Fagus or the top of the stem below the first flower of intact Corallorhiza and Neottia shoots were measured in the field using a pulse-modulated fluorimeter (FMS2; Hansatech Ltd, King's Lynn, UK). The Fv/Fm is defined by Eqn 3 and ΦPSII by Eqn 4 after Maxwell and Johnson (2000).

image(Eqn 3)
image(Eqn 4)

F0 is the minimal level of fluorescence; Fm is the maximum fluorescence (after the application of the saturating flash); inline image is the maximum fluorescence in the light; and Ft is the steady state fluorescence immediately before the flash.

Samples were dark-adapted for 15 min before measurements of Fv/Fm, and the intensity of the 0.7-s light pulse of 8000 µmol photons m−2 s−1. Leaves were adapted to an actinic beam until F0 stabilized to obtain ΦPSII. In the ΦPSII measurements, the light pulse had an intensity of 8000 µmol photons m−2 s−1 for 0.7 s. All fluorescence parameters were estimated as per the manufacturer's instructions.

In situ 13CO2 pulse labelling

Four individuals of Neottia and Cephalanthera, four saplings of Fagus and eight individuals of Corallorhiza were identified in the field and sealed into a plastic bag that transmitted, on average, 95% photosynthetically active radiation (PAR). A PTFE vial containing 50 mg of 99 atom % Ca13CO3 was attached to the inside wall of the bag (4000 cm3) before labelling. A gas-tight seal was made around the stems using anhydrous lanolin. HCl (1% v : v) was injected through the wall of the bag and into the vial to liberate 13CO2 and the resulting hole sealed with PTFE tape (Fig. 1b). Plants were maintained in the labelling bags for 4 h and the PAR was recorded at canopy height every hour throughout the labelling period. Four control plants of each species studied were harvested and dried in order to establish the natural abundance 13C signature of the plants.

Figure 1.

(a) Corallorhiza trifida, (b) Neottia nidus-avis and (c) Cephalanthera damasonium growing beneath a stand of Fagus near Bayreuth, Bavaria, Germany. (d) Labelling chamber for the introduction of the 13CO2 label.

At harvest, plants were dried at 80°C for 48 h and weighed (there was c. 1 h between harvest and samples entering the drying oven, labelled and unlabeled samples were dried in separate ovens). The samples were homogenized separately and a 5 µg subset of each constituent part was analysed for 13C by continuous flow mass spectrometry (PDZ Europa 2020 Isotope Ratio Mass Spectrometer – IRMS coupled to a PDZ ANCA GSL preparation unit, SerCon, Crewe, UK). Data were collected as δ13C relative to the Pee Dee Belemnite international standard for 13C and re-expressed as atom %. The excess (above background) mass of 13C was calculated using Eqn 5.

image(Eqn 5)

MEx = Mass (excess) of the isotope in µg; Atlab = atom % of the isotope in labelled plant; Atcont = atom % of the isotope in paired control plant; M = biomass of sample (µg); and %C = percentage of carbon.

Statistical analysis

Differences between treatment means were analysed by anova followed by Fisher's multiple comparison test using Minitab 13 (Minitab Inc., State College, PA, USA). Data were transformed either using Log10 or Box-Cox (Minitab 13) transformations when they failed to meet the assumptions of anova. Untransformed means and associated standard errors are presented.


Chlorophyll content

The total amount of chla + chlb (µg cm−2) was significantly different between all means (anova (Log10): df = 3,17; F = 368; P < 0.001) with Neottia containing the lowest amount of chlorophyll and Cephalanthera the highest (Fig. 2a). By contrast, there was no significant difference in the chla : chlb ratio of Fagus, Cephalanthera and Corallorhiza. The chla : chlb ratio of Neottia was however significantly lower than that of all other species (anova (Log10): df = 3,17; F = 60.3; P < 0.001) (Fig. 2b).

Figure 2.

(a) Total chlorophyll content and (b) chla : chlb ratio in Neottia nidus-avis (mycoheterotroph), Corallorhiza trifida (partial mycoheterotroph sensu Zimmer et al., 2008), Cephalanthera damasonium (partial mycoheterotroph sensu Gebauer & Meyer, 2003) and Fagus sylvatica (autotroph). Chlorophyll content is expressed as a function of surface area of the sample leaves in all cases except Corallorhiza, which is leafless. In this last case stems were analysed and the surface area calculated as a truncated cone. Bars with differing letters are significantly different (anova: P < 0.05). Error bars represent + 1 SE; n = 4–8.

Chlorophyll fluorescence parameters (Fv/Fm and ΦPSII)

Maximum quantum yield (Fv/Fm) was measured for Fagus, Cephalanthera and Corallorhiza but could not be measured for Neottia as steady-state F0 was never detected (Fig. 3a). The Fv/Fm was highest for Fagus (0.85 ± 0.001) and was significantly different from Cephalanthera (0.81 ± 0.003) (anova: df = 2,12; F = 43.4; P < 0.001), although both values are considered to be within the range for healthy plants (Maxwell & Johnson, 2000). The Fv/Fm for Corallorhiza (0.71 ± 0.015) was significantly lower than that of both Fagus and Cephalanthera (anova: df = 2,12; F = 43.4; P < 0.001) (Fig. 3a). The ΦPSII was determined for Fagus (0.85 ± 0.005), Cephalanthera (0.80 ± 0.012) and Corallorhiza (0.71 ± 0.020) but again could not be measured for Neottia (Fig. 3b). The ΦPSII values recorded for all species were significantly different (anova: df = 2,12; F = 22.2; P < 0.001) (Fig. 3b).

Figure 3.

(a) Maximum quantum yield (Fv/Fm) and (b) steady-state quantum yield (ΦPSII) of photosystem II for the stems of Neottia nidus-avis (mycoheterotroph) and Corallorhiza trifida (partial mycoheterotroph) and the leaves of Cephalanthera damasonium (partial mycoheterotroph) and Fagus sylvatica (autotroph). Bars with differing letters are significantly different (anova: P < 0.05). Error bars represent + 1 SE; n = 4–8. Note: no values for Fv/Fm or ΦPSII could be obtained for Neottia as steady-state F0 was not detectable following the application of the actinic beam.

In situ 13CO2 pulse labelling

The shoots of all target species –Fagus, Cephalanthera, Corallorhiza and Neottia –contained the 13C label (atom % excess and thus above background) after 4 h of exposure (Fig. 4). There was significantly more of the 13C label (µg g−1 DW) in the shoots of Cephalanthera and Fagus than was detected in either Corallorhiza or Neottia (anova (Box-Cox): df = 3,19; F = 91.8; P < 0.001) (Fig. 4a). However, there were no significant differences in the amount of the 13C label present in the tissues of Cephalanthera compared with Fagus or those of Corallorhiza compared with Neottia (anova: P > 0.05) (Fig. 4a). In terms of proportion of 13CO2 label supplied that was fixed by the plant there was no significant difference in the percentage of the label fixed by Fagus compared with Cephalanthera (anova (Box-Cox): P > 0.05; Fig. 4b). Both Fagus and Cephalanthera fixed a significantly greater percentage of the label supplied than either Corallorhiza or Neottia while, somewhat surprisingly, Neottia contained a greater percentage of the label than Corallorhiza (anova (Box-Cox): df = 3,19; F = 95.1; P < 0.001) (Fig. 4b).

Figure 4.

(a) Total amount of 13C and (b) the percentage of the supplied 13C present in plant shoots of Neottia nidus-avis (mycoheterotroph), Corallorhiza trifida (partial mycoheterotroph), Cephalanthera damasonium (partial mycoheterotroph) and Fagus sylvatica (autotroph) after 4 h exposure to a 13CO2 source. Mean photosynthetically active radiation (PAR) is given above each bar. Bars with differing letters are significantly different (anova: P < 0.05). Error bars represent + 1 SE. n = 4–8.

The amount of 13C present in the plant shoots and the percentage of the label fixed are not functions of light availability as there is no relationship between PAR and 13C content (Fig. 4).


A recent analysis (Zimmer et al., 2008) of the enrichment of the stable isotopes 13C and 15N in tissues of C. trifida indicated a small but statistically significant depletion in the natural abundance of these two elements (δ13C = −25.6 and δ15N = −0.3) in this orchid relative to that seen in co-occurring plants of the fully mycoheterotrophic orchid N. nidus-avis13C = −24.0; δ15N = 4.9). These values are similar to those recorded in the present study (δ13C = −24.2 and −23.15 for Corallorhiza and Neottia respectively; see the Supporting Information, Fig. S1). The observed difference between the two orchids was interpreted to indicate that Corallorhiza obtained a proportion of its C from photosynthesis. Indeed, Zimmer et al. (2008), on the basis of a stable isotope mixing model, concluded that the Corallorhiza plants they analysed had gained c. 33% of their carbon through autotrophic C fixation. Since the levels of C fixation by Corallorhiza observed in the present study were only c. 2% of those seen in co-occurring Fagus, it would seem that its photosynthetic capacity is an order of magnitude lower than that of normal autotrophs.

Since, during their short period of development above ground, the green stems of this orchid will normally be exposed to diffuse irradiance, it is logical to expect that some autotrophic activity could occur in their tissues. The presence of genes encoding for chlorophyll synthesis (Barrett & Freudenstein, 2008), the demonstration of the occurrence of chlorophylls a and b by Zimmer et al. (2008) and in the present study, as well the presence of chlorophyll fluorescence, are also supportive of the notion that some potential for autotrophy can be expected in this orchid. However, all of these approaches to the question of the extent of photosynthetic activity in Corallorhiza are essentially indirect, there being only one previous report of direct analysis of its ability to assimilate C (Montfort & Küsters, 1940). This indicated that some autotrophic C fixation did occur in the orchid.

The direct measurements of 13CO2 assimilation reported in the present study indicate that under similar conditions of irradiance, the quantities of C fixed by photosynthesis in Corallorhiza are negligible relative to those seen in a co-occurring partially mycoheterotrophic (sensu Julou et al., 2005) leafy green orchid C. damasonium or in saplings of F. sylvatica. Moreover, the amount of 13CO2 assimilated by Corallorhiza was not significantly different from that detected in the tissues of Neottia. In this latter case, the 13CO2 detected must be a result of diffusion, incorporation into organic acids via the phosphoenolpyruvate (PEP) carboxylase pathway or through nonphotochemical processes as, by general consent, Neottia is unable to photosynthesize owing to a lack of critical light-harvesting pigments (Drude, 1873; Montfort & Küsters, 1940; Reznik, 1958; Reznik et al., 1969; Menke & Schmidt, 1976).

Since our measurements were made under conditions of direct diffused solar irradiance and at the stage of maximum shoot extension in plants with fully developed green seed capsules it seems unlikely that there are other environmental or developmental conditions that would be more favourable for the expression of photosynthetic activity. Indeed, since all species examined co-occurred on identical substrates within a few metres of each other it is reasonable to assume that they were all experiencing the same soil conditions. Further, the phenology of this orchid is such that the opportunity for significant autotrophic accumulation of C is inevitably restricted as the flowering spikes are exposed above ground for only a very restricted period of time, normally not more than 2 months. Additional direct measurements of the kind described here are desirable and preferably these should be carried out over a longer period than the 4-h duration employed in this study. Nonetheless, it is apparent from the results obtained in the parallel analyses of Cephalanthera that this period of exposure is sufficient to reveal C fixation when and where it is taking place. Moreover, such exposure times are as long as or longer than those routinely employed for assessment of photosynthetic activity using infrared gas analysis techniques.

Chlorophyll fluorescence parameters for Corallorhiza indicate, on the one hand, the presence of active photosystem II reaction centres but on the other that the quantum efficiency of PSII (ΦPSII) is reduced. In this orchid, the mean value of 0.7 for Fv/Fm was substantially lower than the multispecies average of 0.83 recorded for healthy autotrophs by Maxwell & Johnson (2000). Such a value is potentially indicative of photo-inhibition (Cameron et al., 2008b), although the analyses of Ritchie (2006) suggest that the values of Fv/Fm and ΦPSII recorded here for Corallorhiza are not necessarily so depressed as to predict the absence of photosynthesis. Our failure to detect more than minimal C fixation suggests, therefore, that most of the excitation energy of the light harvesting complex of PSII (LHCII) is being transferred to an alternative electron acceptor, it then being dissipated as heat through nonphotosynthetic metabolism, as described by Krause & Weis (1991). Such nonphotochemical quenching (NPQ) processes are known to be facilitated by xanthophylls (Johnson et al., 1993). Neither NPQ nor carotenoid composition of Corallorhiza could be measured in the present study, but in view of the observation (Bungard et al., 1999) that the nonphotosynthetic holoparasite Cuscuta reflexa possesses a novel type of NPQ-related xanthophyll cycle linked with the transition from autotrophy to heterotrophy, analyses of these pathways in the orchid are called for.

Clearly, C. trifida, while retaining the genes encoding for chlorophyll synthesis (Barrett & Freudenstein, 2008), represents a late stage in the evolutionary development towards complete mycoheterotrophy. However, it appears from the present study that in a physiological context this orchid has moved more closely towards the fully mycoheterotrophic condition than has previously been recognized. These observations are consistent with those indicating that, in nature, C. trifida is routinely involved in tripartite symbiotic associations between ectomycorrhizal fungi and autotrophic overstory trees (Zelmer & Currah, 1995; McKendrick et al., 2000b). The mycelia of the fungal partners have been shown to provide pathways through which C is transferred from the trees to the large coralloid root systems that constitute the slowly developing belowground storage structures characteristic of this genus (McKendrick et al., 2000a). Having formed such an effective mechanism for assimilate acquisition it is perhaps not surprising that the contribution of photosynthesis to the C economy of the orchid, has, as indicated here by direct measurements, been so greatly reduced.


We thank Dr Janice Lake and Dr Jonathan Leake (University of Sheffield, UK) for critical comments on the manuscript, Prof. Julie Scholes (University of Sheffield) for assistance in the interpretation of the chlorophyll fluorescence data and Prof. David Beerling (University of Sheffield) for supplying the portable fluorimeter. We also thank Irene Johnson for expert technical support and Heather Walker (University of Sheffield) for analysing the samples for 13C content. This project was funded by the Natural Environment Research Council, UK (Award Number NE/E014070/1 to D.D.C.) and Deutsche Forschungsgemeinschaft, Germany (Award Number GE 565/7-1 to G.G.).