The degree of mycoheterotrophic carbon gain in green, variegated and vegetative albino individuals of Cephalanthera damasonium is related to leaf chlorophyll concentrations

Authors

  • Marcus Stöckel,

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

Author for correspondence:
Gerhard Gebauer
Tel: +49 (0)921 552060
Email: gerhard.gebauer@uni-bayreuth.de

Summary

  • Achlorophyllous variants of some forest orchids are known to reach almost the same size as their green forms. These vegetative albino forms cover their entire carbon (C) demand through fungi that simultaneously form ectomycorrhizae with trees, while green variants partially draw on C from photosynthesis and C from fungal hosts. Here, we investigate whether the amount of C derived from either source is proportional to leaf chlorophyll concentration. The discovery of two Cephalanthera damasonium populations with variegated leaves enabled a continuous bridging of leaf chlorophyll concentrations between green and albino forms.
  • Leaves of 27 green, variegated and albino individuals of C. damasonium were compared for chlorophyll concentrations, C sources (as characterized by 13C abundances) and total C and nitrogen (N) concentrations.
  • We found a linear relationship between leaf chlorophyll concentrations and the proportional reliance on fungi as a C source. Furthermore, we show that the shift in C gain through mycoheterotrophic means significantly changes leaf total C and N concentrations.
  • Our results document that partial mycoheterotrophy in C. damasonium is not a static nutritional mode but a flexible mechanism related inter alia to leaf chlorophyll concentrations. The change in proportional reliance on fungi as a C source affects leaf chemical composition.

Introduction

Because of their minute and almost endospermless seeds, all of the c. 25 000 orchid species (Dressler, 2004) are assumed to be initially mycoheterotrophic, that is, they rely on a fungal host in their very early seedling stage for their carbon and mineral nutrient gain (Smith & Read, 2008). Later, most orchids develop green leaves and become autotrophic. However, c. 200 orchid species remain achlorophyllous. These achlorophyllous orchids rely for their carbon and presumably mineral nutrient supply on the fungal host throughout their life cycle, and therefore their nutrition is called full mycoheterotrophy (Leake, 1994; for definitions see also Merckx et al., 2009). Recent investigations based on stable carbon (C) and nitrogen (N) isotope abundance analysis revealed for some green orchids belonging to various tribes of the subfamily Epidendroideae, and mostly growing at shaded forest sites, a tapping on two different C sources: C from photosynthesis and C from fungal hosts simultaneously sharing ectomycorrhizae with trees (Gebauer & Meyer, 2003; Bidartondo et al., 2004; Zimmer et al., 2007, 2008; Selosse & Roy, 2009; Liebel et al., 2010). Among this mixotrophic, or, more specifically, partially mycoheterotrophic, group of orchids are some species that have long been known to produce fully achlorophyllous but still leaf-bearing variants (Renner, 1938; Mairold & Weber, 1950). These vegetative albino variants reach almost the same size as the green forms (Julou et al., 2005; Tranchida-Lombardo et al., 2010). Their isotope signatures are similar to fully mycoheterotrophic plants associated with fungi forming ectomycorrhizae, indicating a 100% C and a considerable N gain from the fungal source (Julou et al., 2005; Abadie et al., 2006; Preiss & Gebauer, 2008).

In addition to the frequent green and the rare albino forms of some orchid species, individuals with a continuous range between these extremes have episodically been found in nature. For Epipactis helleborine, a variegated individual was found with white and green striped leaves (Salmia, 1989). Epipactis microphylla had yellowish-white striped leaves and Gymnadenia conopsea had white bordered leaves (Renner, 1938). Such phenotypes are called variegation forms (Kirk & Tilney-Bassett, 1978). Species with variegated leaves are also well known for many nonorchids from various taxonomic groups. Simultaneous occurrence of individuals with fully green and variegated leaves of the same species is common in dense forests as well as in open habitats. Although their photosynthetic capacity is reduced, individuals with variegated leaves are considered to be less attacked by leaf-mining herbivores (Smith, 1986). Until now, evidence for such intermediate phenotypes of Cephalanthera damasonium has been lacking (Abadie et al., 2006; Tranchida-Lombardo et al., 2010).

In this study we report for the first time on two C. damasonium populations that represent a continuum, including fully green, variegated and albino individuals. This continuum provides the unique opportunity to test whether the degree of mycoheterotrophic nutrient gain by this orchid is related to leaf chlorophyll concentrations and to investigate consequences of the mycoheterotrophic nutrient gain on leaf C and N concentrations. This study complements the recent finding of a direct relationship between light availability and dependence on the fungal C source in green C. damasonium individuals (Preiss et al., 2010).

Materials and Methods

Study site and sampling scheme

Leaves of Cephalanthera damasonium (Mill.) Druce and surrounding reference plants were sampled at two different sites in northeast Bavaria, Germany (49°41′N, 11°02′E and 50°01′N, 11°20′E). One site is a mixed deciduous forest dominated by Acer campestre and Fagus sylvatica. The other sampling site is a deciduous forest dominated by F. sylvatica. The mean annual temperature at both sites is 6–9°C, with a mean annual precipitation of 650–1000 mm (German weather service, http://www.dwd.de). Both of these forest sites have a homogenous dense canopy. Thus, the light intensity available to forests’ ground vegetation is low. Based on simultaneously performed measurements of photosynthetic active radiation (Quantum Sensor; Li-Cor, Lincoln, NE, USA) close to the orchid leaves and outside of the forests, we calculated for the investigated areas within these forests relative light availabilities between 1 and 4%. Green, variegated and albino individuals of this study are growing in close spatial proximity and thus receive similar light intensities.

In July 2004 and June 2008, leaves of 27 C. damasonium individuals were sampled, with a special focus on differences in leaf colour. The second highest leaf of normal green, albino and variegated individuals (Fig. 1) was cut and prepared for further analysis. From regular inspection, we confirmed that leaf colour of the investigated individuals did not visually change from the beginning of leaf appearance until collection, and that all of the investigated individuals were growing only vegetatively during the years of this study. Additionally, leaves of accompanying autotrophic reference plants were sampled (Table 1). To minimize spatial effects of microclimate and δ13C of CO2 in air on leaf δ13C, a sampling scheme as described in detail by Gebauer & Meyer (2003) was applied; that is, leaves of green, variegated or albino C. damasonium individuals and leaves of the respective reference plants were collected within a distance of 1 m2 and at the same height above ground.

Figure 1.

 Variegated individuals of Cephalanthera damasonium collected at the mixed deciduous forest site in northeast Bavaria, Germany. The photographs show a homogenous distribution of variegation over all leaves of the investigated individuals.

Table 1.   Sampling sites and years of investigated Cephalanthera damasonium individuals and autotrophic reference plants
SiteYearNumber of investigated individuals
C. damasoniumReference plants
GreenVariegatedAlbino
  1. See Table 2 for species of the collected reference plants.

Mixed forest200447415
Mixed forest200844012
Beech forest2008220 6

In addition, seven further leaves of variegated C. damasonium individuals were sampled in 2004 in the mixed forest stand. These leaves were used to compare the C isotope signatures and C and N concentrations in green and white parts within individual leaves with the respective data of fully green or albino individuals. Both parts of these leaves were cut off and separately prepared for stable isotope abundance analysis without determination of their chlorophyll concentrations.

Chlorophyll concentration analysis

To minimize the destructive harvest of leaf material from the protected orchids, and specifically from the very rare variegated and albino individuals, and in order to analyse leaf δ13C and chlorophyll concentrations from identical material, leaf chlorophyll concentrations were measured in a nondestructive way using a chlorophyll meter (SPAD-502; Konica Minolta Sensing Inc., Osaka, Japan). The chlorophyll concentration of the uniform white and green leaves was measured five times at different positions of the sampled leaves. The chlorophyll concentrations of reference plants collected in 2008 were determined in the same way. For the variegated leaves, the spot measurements were carried out five times on green and five times on white parts. After this the leaves were photographed and the proportion of green and white parts was determined by setting a 1 × 1 mm grid over the pictures. The chlorophyll concentration of the whole leaves was calculated from the Soil Plant Analysis Development (SPAD) values of white and green parts weighted by their respective areas. For leaf sections intermediate between green and white, a mean chlorophyll concentration of the green and white parts was assumed. These intermediate sections always covered very minor proportions of the leaf areas. The SPAD values, which are correlated to the chlorophyll content, were converted into chlorophyll concentrations using Eqn 1 (Monje & Bugbee, 1992):

image(Eqn 1)

Stable isotope abundance analysis

Leaf samples were dried at 105°C, ground and stored in a desiccator until further analysis. Relative C stable isotope abundances and C and N concentrations were measured using a dual-element analysis mode with an elemental analyser coupled to a continuous-flow isotope ratio mass spectrometer (Bidartondo et al., 2004). The δ notation was used for measured isotope abundances according to Eqn 2:

image(Eqn 2)

where R is the ratio of heavy to light isotopes in the sample or the respective standard. CO2 standard gas calibration, control of reproducibility and accuracy of the isotope abundance measurements and calculation of total C and N concentrations in the leaf samples followed the protocol described by Gebauer & Schulze (1991).

Statistics

Because of the small numbers and rather inhomogeneous distributions of replicates, all statistical comparisons between groups were based on nonparametric tests. δ13C values and total C and N concentrations of reference plants of different years and sites were tested for differences with the Kruskal–Wallis test. Chlorophyll concentrations of reference plants were only measured in 2008 at two different sites. For this reason the Mann–Whitney U-test was used to test chlorophyll concentrations of reference plants for differences. The Mann–Whitney U-test was also used to compare mean values from the green and white parts of variegated leaves, to compare mean values of green or albino leaves with green or white parts of variegated leaves and to compare mean N concentrations of reference plant leaves with those of C. damasonium leaves. Values of α < 0.05 were assumed to be significant. Regression and correlation analyses between chlorophyll concentration and δ13C or total C or N concentration of C. damasonium leaves were performed using SigmaPlot v. 11.0 (Systat Software, Inc., San José, CA, USA). Data used for the regression equations were visually checked for normal distributions of the residuals and homoscedasticity of the residuals variance. To test two linear regression equations obtained from data of different years or sites for differences, a method equivalent to an analysis of covariance (ANCOVA) was used and performed with GraphPad Prism v. 5 (GraphPad Software, Inc., San Diego, CA, USA). All other statistical tests were carried out using SPSS v. 16.0 (SPSS, Inc., Chicago, IL, USA). Data are mean values ± 1 SD if not stated otherwise.

Results

Chlorophyll concentrations in the leaves of the 27 C. damasonium individuals ranged from 1.5 to 276 mg m−2 (Fig. 2). The mean chlorophyll concentration found for albino leaves was 1.6 ± 0.1 mg m−2 (= 4) and for leaves of green individuals was 214 ± 43 mg m−2 (= 10). Thus, leaf chlorophyll concentration of green individuals was c. 130 times higher than that of albinos. Chlorophyll concentrations in the leaves of variegated individuals (= 13) ranged between the two extremes. The reference plants had mean chlorophyll concentrations in the range 86 ± 22 to 196 mg m−2 (Table 2).

Figure 2.

 δ13C values (a), total carbon (C) concentrations (b), and total nitrogen (N) concentrations (c) as a function of chlorophyll concentrations in leaves of green (squares), variegated (circles) and albino (triangles) Cephalanthera damasonium individuals (= 27). Regression curves (solid lines) are given with 95% confidence intervals (dotted lines): (a) = −0.016− 23.47, F1;25 = 56.4, r²adj = 0.68, < 0.001; (b) = 0.016x + 34.58, F1;25 = 38.8, r²adj = 0.72, < 0.001; (c) = −0.005x + 3.93, F1;25 = 10.0, r²adj = 0.26, = 0.004.

Table 2.   Mean δ13C values, total carbon (C), total nitrogen (N) and chlorophyll concentrations [Chl] of autotrophic reference species of the different sampling sites and years
YearSitenSampleδ13C (‰) ± SDTotal C (mmol g−1 DW) ± SDTotal N (mmol g−1 DW) ± SD[Chl] (mg m−2) ± SD
  1. Data show no statistical differences between years and sites for reference δ13C values (= 1.885, df = 2, = 0.39), total C (= 0.495, df = 2, = 0.781) and total N (= 1.202, df = 2, = 0.548) concentrations (Kruskal–Wallis test). No significant differences (= 33.00, = 0.779) in chlorophyll concentrations between different sites in 2008 were obvious (Mann–Whitney U-test).

2004Mixed forest4Acer pseudoplatanus−32.1 ± 0.438.3 ± 0.41.7 ± 0.1 
4Crataegus monogyna−31.8 ± 0.636.7 ± 0.81.1 ± 0.0 
4Fagus sylvatica−32.0 ± 0.840.1 ± 0.41.3 ± 0.1 
4Fragaria vesca−30.6 ± 0.637.7 ± 0.31.6 ± 0.2 
4Galium odoratum−32.2 ± 0.733.9 ± 0.51.7 ± 0.2 
2008Mixed forest4Fagus sylvatica−31.4 ± 1.340.0 ± 0.41.5 ± 0.186 ± 22
4Galium odoratum−31.9 ± 0.735.3 ± 0.51.7 ± 0.2140 ± 30
4Crataegus monogyna−31.4 ± 0.236.8 ± 2.41.3 ± 0.1115 ± 26
2008Broadleaf forest2Fagus sylvatica−33.140.51.696
2Fragaria vesca−31.837.11.4114
1Convallaria majalis−30.438.01.7121
1Plantago lanceolata−33.535.32.1196

There were no significant differences between δ13C values or total C or N concentrations in the leaves of the reference plants collected in different years and at different sites (Table 2). There were also no significant distinctions in the regression equations for chlorophyll concentrations of C. damasonium leaves from different years or sites vs δ13C values (slopes, = 3.37, dfn = 2, dfd = 21, = 0.053; intercepts, = 2.08, dfn = 2, dfd = 23, = 0.15) or total C (slopes, = 0.51, dfn = 2, dfd = 21, = 0.61; intercepts, = 0.74, dfn = 2, dfd = 23, = 0.49) or total N concentrations (slopes, = 0.02, dfn = 2, dfd = 21, = 0.98; intercepts, = 1.57, dfn = 2, dfd = 23, = 0.23). This justifies pooling of C. damasonium leaves of different years and sites for all further regression analyses.

While mean δ13C values of the reference plants varied in narrow ranges from −30.4 to −33.5‰ (Table 2), δ13C values of C. damasonium leaves were in general less negative and the variation was considerably greater, ranging from −22.0 to −28.8‰ (Fig. 2a). A comparison of the C. damasonium leaves revealed that green leaves were the ones most depleted in 13C, while albino leaves were the least depleted. Variegated leaves ranged in between C. damasonium leaf δ13C values were negatively correlated with leaf chlorophyll concentrations; that is, leaves with the highest chlorophyll concentrations were the ones most depleted in 13C and vice versa (Fig. 2a).

Mean leaf total C concentrations of the reference plants ranged between 35.3 and 40.5 mmol C g–1 DW (Table 2), while leaves of C. damasonium had a tendency towards lower C concentrations (33.1–38.8 mmol C g–1 DW), with albino and variegated leaves having lower total C concentrations than leaves of green individuals. The mean difference in leaf total C concentrations between green (38.0 ± 0.7 mmol C g–1 DW) and significantly different (< 0.001, = 0.005) albino individuals (33.5 ± 0.3 mmol C g–1 DW) was 4.5 mmol C g–1 DW. Thus, C. damasonium leaf total C concentrations were positively correlated with chlorophyll concentrations (Fig. 2b).

Leaf total N concentrations of the investigated C. damasonium individuals ranged from 1.9 to 5.4 mmol N g–1 DW (Fig. 2c) with a mean value of 3.4 ± 0.9 mmol N g–1 DW and thus were significantly higher (= 5.00, <0.001) than mean leaf total N concentrations of the reference plants (1.5 ± 0.3 mmol N g–1 DW; Table 2). C. damasonium leaf total N concentrations were negatively correlated with the chlorophyll concentrations; that is, leaves of albino and variegated individuals had higher total N concentrations than leaves of green individuals (Fig. 2c).

A comparison of white and green parts of seven leaves of variegated C. damasonium individuals indicated a significant (= 2.00, = 0.002) difference in δ13C values even within these leaves. White parts (δ13C = −23.8 ± 0.8‰) were more enriched in 13C than green parts (−25.6 ± 0.7‰), and, analogous to albino and green individuals, the total C concentrations were significantly (= 3.00, = 0.004) lower in white parts (35.4 ± 1.0 mmol C g–1 DW) than in green parts (37.0 ± 0.8 mmol C g–1 DW). However, no significant differences (= 20.00, = 0.62) in total N concentrations were found between white (2.7 ± 1.1 mmol N g–1 DW) and green parts (3.1 ± 0.7 mmol N g–1 DW) of variegated leaves. A further comparison of δ13C values and total C and N concentrations in white or green parts of variegated C. damasonium leaves with the leaves of albino or fully green individuals also revealed mostly significant differences. White parts of variegated leaves were more depleted in 13C (= 1.0, = 0.014) and had higher total C (= 1.0, = 0.014) and lower total N concentrations (< 0.001, = 0.008) than leaves of albino individuals, while green parts of variegated leaves were less depleted in 13C (= 12.0, = 0.025) and had lower total C concentrations (= 12.0, = 0.025) than leaves of fully green individuals. Only total N concentrations of green parts in variegated leaves and of leaves of fully green C. damasonium individuals were not significantly distinguished (= 29.0, = 0.558).

Discussion

Chlorophyll is an essential prerequisite for plant photosynthesis. Therefore, it is clear that achlorophyllous plants require a C source alternative to atmospheric CO2 assimilated by chlorophyllous plants in photosynthesis. They either penetrate with haustoria into the xylem, and in most cases also into the phloem, of their host plants and tap on C compounds and mineral nutrients transported by their hosts (holoparasites), or they utilize C from various types of fungi associations in their roots (mycoheterotrophs). In both cases, the C isotope signature of the hosts determines the C isotope composition of their own tissue. While holoparasites have C isotope signatures almost identical to their host plants (Ziegler, 1996), mycoheterotrophs are enriched in 13C compared with accompanying autotrophic plants (Gebauer & Meyer, 2003; Trudell et al., 2003; Preiss & Gebauer, 2008; Ogura-Tsujita et al., 2009) because of a 13C enrichment of fungal tissue in comparison to accompanying plants (Högberg et al., 1999; Kohzu et al., 1999). As already known from previous studies (Julou et al., 2005; Abadie et al., 2006) and confirmed in this investigation, Cephalanthera albino forms are enriched in 13C, similar to fully mycoheterotrophic plants. As also confirmed in this investigation and already known from other studies, green individuals of C. damasonium are enriched in 13C compared with reference plants (Gebauer & Meyer, 2003; Bidartondo et al., 2004; Julou et al., 2005; Liebel et al., 2010), despite having leaf chlorophyll concentrations in the typical range of autotrophic forest ground plants (Cameron et al., 2009). However, this 13C enrichment is lower than that for fully mycoheterotrophic plants or albinos and is not constant. For this reason they are considered to be partial mycoheterotrophs. Albino phenotypes of C. damasonium were found to associate with a greater diversity of basidiomycetae than fully green individuals (Julou et al., 2005). Thus, a greater diversity of fungal partners might help achlorophyllous C. damasonium individuals to compensate for the lack of C gain from photosynthesis. This investigation of C. damasonium individuals with variegated leaves provided us, for the first time, with the opportunity to document that the proportional reliance of this orchid species on the fungal C source is, under similar light conditions, linearly correlated with leaf chlorophyll concentration. Even within variegated leaves, a difference in the C isotope signature between green and white leaf sections was found. Differences in the C isotope signature of green and white parts of variegated C. damasonium leaves were, however, smaller than differences in the leaf C isotope signature of fully green and albino individuals, indicating a certain mixing of C from photosynthesis and the fungal source between green and white parts of the variegated leaves. Further investigations on variegated leaves of fully autotrophic plants are required to elucidate whether, and by how much, a general difference in δ13C of photosynthetic and nonphotosynthetic plant tissues (Gebauer & Schulze, 1991; Cernusak et al., 2009) contributes to the differences in C isotope signature found between green and white parts of C. damasonium leaves.

A recent investigation revealed that, for two Cephalanthera species, the local light climate at the forest ground was a further factor modulating the proportional degree of mycoheterotrophic C gain. The reliance on the fungal C source by green Cephalanthera plants capable of partially mycoheterotrophic C gain obviously increases continuously with decreasing light availability for photosynthesis (Preiss et al., 2010). Also, for a much larger set of orchid species growing in a wide range of habitat types and receiving very different light intensities, a relationship between light climate and proportional C gain from the fungal source was found (Liebel et al., 2010).

The variability in reliance on C from photosynthesis or from tapping on a fungal source also affects inter alia the chemical composition of the plant tissue. Fully and partially mycoheterotrophic orchids are already known to have unusually high total N concentrations (Gebauer & Meyer, 2003; Julou et al., 2005; Liebel & Gebauer, 2010; Liebel et al., 2010). The high total N concentrations in mycoheterotrophic orchids appear to stem from the fact that fungi have considerably higher total N concentrations in their tissue than most autotrophic plants (Gebauer & Dietrich, 1993; Gebauer & Taylor, 1999). C gain from the fungal source is obviously associated with a considerable N gain, leading to N concentrations in mycoheterotrophic orchids considerably above those known for the majority of autotrophic plants (Gebauer et al., 1988). Laboratory 13C and 15N labelling studies with the mycoheterotrophic orchid Rhizanthella gardneri (Bougoure et al., 2010) and the green orchid Goodyera repens (Cameron et al., 2006) also indicate a simultaneous fungus-to-plant transport of organic C and N compounds. Here, we document for green, variegated and albino individuals of C. damasonium that the continuous decrease in leaf chlorophyll concentration and, as a consequence, the proportional increase in reliance on C from host fungi are accompanied by a linear increase in leaf N concentration. This observation is in contrast to what is commonly understood about the relationships between leaf N concentrations and photosynthesis. According to textbook knowledge, the capacity for leaf photosynthesis should increase with increasing leaf total N concentration (Schulze et al., 2005), because usually 20–30% (Feller et al., 2008) of the leaf N in C3 plants is bound in the enzyme Rubisco, which is directly involved in photosynthetic C fixation. Thus, our best prediction is that most N in the tissue of fully and partially mycoheterotrophic orchids is a waste product and does not serve any physiological function. An analysis of the chemical nature of the N compounds responsible for the high total N concentrations in the tissue of mycoheterotrophic orchids is urgently required to elucidate this puzzling phenomenon.

Carbon gain through the fungal source affects not only leaf total N concentrations of green, variegated and albino individuals of C. damasonium, but also leaf total C concentrations. C concentrations in plant tissue are generally considered to vary within narrow ranges (Öztürk et al., 1981). Here, we found that leaf total C concentration decreased with decreasing chlorophyll concentration and thus increasing reliance on C supply from the fungal source. The decrease in leaf total C concentration (in mmol g–1 DW) from green to albino individuals was even higher than the increase in leaf total N concentration. Thus, a potential C dilution effect caused by the extraordinarily high leaf total N concentrations of the albino individuals can only partially explain the changes in leaf total C concentration. Again, further analysis of peculiarities in the chemical composition of fully and partially mycoheterotrophic orchid tissue is required to fully understand the changes in leaf total C concentration.

In summary, in this investigation of the orchid species C. damasonium we show a close linear relationship between leaf chlorophyll concentration and proportional C gain through mycoheterotrophic means (represented by the leaf δ13C values) and consequences for the leaf total C and N concentrations. Although rather rare in nature, variegated orchid forms turned out as ideally suited tools to bridge between the green and albino endpoints of a continuum in proportional reliance by plants on fungi as a C source.

Acknowledgements

The authors thank Isolde Baumann, Iris Schmiedinger, Christine Tiroch (BayCEER – Laboratory of Isotope Biogeochemistry, University of Bayreuth) and Marga Wartinger (Department of Plant Ecology, University of Bayreuth) for skilful technical assistance with stable isotope abundance measurements, and the Regierung von Oberfranken for permission to collect leaf material of the protected orchids. Valuable comments by three anonymous reviewers on an earlier version of this manuscript are highly appreciated. This work was supported by the German Research Foundation (DFG, project GE 565/7-1).

Ancillary