Characterization and spatial distribution of ectomycorrhizas colonizing aspen clones released in an experimental field
- First Online:
- Cite this article as:
- Kaldorf, M., Renker, C., Fladung, M. et al. Mycorrhiza (2004) 14: 295. doi:10.1007/s00572-003-0266-1
- 169 Views
Ectomycorrhizas (EM) from aspen clones released on an experimental field were characterized by morphotyping, restriction analysis and internal transcribed spacer (ITS) sequencing. In addition, their community structure and spatial distribution was analyzed. Among the 23 observed morphotypes, six mycobionts dominated, forming roughly 90% of all ectomycorrhizas: Cenococcum geophilum, Laccaria sp., Phialocephala fortinii, two different Thelephoraceae, and one member of the Pezizales. The three most common morphotypes had an even spatial distribution, reflecting the high degree of homogeneity of the experimental field. The distribution of three other morphotypes was correlated with the distances to the spruce forest and deciduous trees bordering the experimental field. These two patterns allowed two invasion strategies of ectomycorrhizal fungi (EMF) to be recognized, the success of which depends on adaptation of the EMF to local ecological conditions.
For most trees in temperate and boreal forests, establishment, growth and survival are dependent on colonization by ectomycorrhizal fungi (EMF) (Smith and Read 1997). EMF belong to the Basidiomycota, Ascomycota and also Zygomycota, and their estimated diversity is above 5,000 species (Molina et al. 1992). Below ground, ectomycorrhiza (EM) communities are often species-rich, with anything from 20 to over 50 EM species colonizing single tree species at a given stand (Horton and Bruns 2001). In this context, the starting point of each investigation in the field is characterization of EM at the species level. This goal can be achieved by combining detailed light-microscope-based morphological and anatomical description—morphotyping (Agerer 1991)—with molecular approaches. Molecular characterization is based on PCR amplification of the internal transcribed spacer (ITS) region within the rDNA, which has been shown to be a suitable species marker for EMF (Buscot et al. 2000 and references therein). This method has, therefore, been used in many studies on EM community structure since the pioneering work of Gardes et al. (1991). To date, the majority of studies have focused on EM of coniferous trees, mainly of the genera Picea and Pinus. In the Colour Atlas of Ectomycorrhizae (Agerer 1987–1998), the largest available systematic collection of EM morphotype descriptions, the keys for Picea and Pinus EM contain 91 and 67 morphotypes, respectively, while for fast-growing deciduous trees like Eucalyptus, Populus or Salix only 3–4 morphotypes are included.
In the last decade, Populus has become one of the most interesting trees for biotechnology. Besides being of commercial importance, e.g., for the paper industry, poplars combine many biotechnological advantages, such as rapid growth, simple in vitro propagation and the existence of genetic transformation systems (Fladung and Ahuja 1996). In the near future, genetically modified (transgenic) poplars could be cultivated in large plantations. Therefore, it becomes necessary to characterize their mycobionts, especially with a view to assessing the mycorrhization pattern of transgenic trees under field conditions in comparison to that of wild type trees. In a recent study, we initiated such a comparative study on Populus (Kaldorf et al. 2002) within the frame of the first release experiment involving transgenic trees in Germany (Fladung and Muhs 2000).
Several mechanisms, such as host plant specificity (Cullings et al. 2000), clonal versus sexual spread (Fiore-Donno and Martin 2001), competition (Wu et al. 1999) and ecophysiological preferences (Bruns 1995) of EMF, have been discussed to explain the formation of spatial patterns in EM communities. In the present work, given that in the experiment mentioned above defined aspen lines were released in a homogeneous field, we take the opportunity to complete the detailed description of aspen EM morphotypes by microscopic and molecular methods and to evaluate the impact of ecological factors not related to host plant effects on the formation of the local EM community.
Materials and methods
Description of the field site
In this context, the northern part of the field clearly received more sunlight than the southern one. This was confirmed by measurements of the light intensity with a LI-COR Quantum/Radiometer/Photometer, Model LI-185B (LI-COR Biosciences, Lincoln, Neb.), taken with five replicates 1 m above ground at three points (close to the spruce forest, central and close to the deciduous trees) within each sector of the experimental field on a sunny summer day (see Results). This may explain why, during the field work, we often observed a high soil moisture content in the shady southern part of the field.
The plant material investigated consisted of seven lines of hybrid aspen (Populus tremula L. × P. tremuloides Michx.) planted out in 1996. One line was the parental wild type clone Esch5, from which six transgenic lines were derived and multiplied by micropropagation before planting out (Fladung et al. 1997). Four of these transgenic lines carried the rolC gene from Agrobacterium rhizogenes under the control of the 35S promoter of cauliflower mosaic virus on the plasmid pPCV002-35S-rolC (Spena et al. 1987) as a morphologically selectable marker to analyze transgene stability in the field. These 35S-rolC transgenic aspen with constitutive rolC expression were characterized by altered levels of several endogenous growth regulators, resulting in reduced plant height, shortened internodes and smaller leaves (Fladung et al. 1997). Two further lines were obtained by transformation with the plasmid pPCV002-rbcS-rolC (Schmülling et al. 1993), in which rolC is under the control of the light-inducible rbcS promoter from potato. Under field conditions, rbcS-rolC transgenic trees, expressing the rolC gene mainly in the leaves, were not different phenotypically from the parental aspen line Esch5 (Fladung and Muhs 2000).
Each aspen line was represented in the field by four randomized blocks with 8 plants in each (in total 32 plants per line). The experimental field was divided into four sectors (Fig. 1), each of them containing one block of each aspen line. By this planting design, an even spatial distribution of wild type, 35S-rolC transgenic and rbcS-rolC transgenic aspen was obtained.
Each EM sample consisted of one root fragment of approximately 5–8 mm diameter and 100–120 cm length taken together with its derived fine rootlets. Most samples contained between 250 and 400 root tips with fully developed EM, all of which were characterized by morphotyping as described below. The root fragments were dug out, starting from trunks, in order to unambiguously assign each of them to individual trees. Most trees were sampled only once during the field experiment in order to reduce disturbance of the root systems, especially of the small 35S-rolC transgenic plants. However, for a few trees double sampling was performed to check whether the distribution of different morphotypes observed within the single samples was representative for a given tree. The root samples were stored at 4°C in plastic bags filled with water for up to 4 weeks until analysis (Kaldorf et al. 2002). Samples were collected 12 times between September 1998 and October 2001 to include all morphotypes independent of seasonal or successional effects. At each sampling date, between 9 and 15 root samples were taken, covering at least three of the four sectors of the experimental field with samples from at least one wild type, one 35S-rolC transgenic and one rbcS-rolC transgenic tree per sector.
In 1998/1999, the work was focused on the morphotype characterization, while the data collected in 2000/2001 concerned the spatial distribution of the different morphotypes. Overlapping of spatial by temporal effects was reduced by excluding the data collected in 1998/1999 from spatial pattern analysis, as the EM community changed considerably between 1998/1999 and 2000/2001 (data not shown). Additionally, statistical analyses of spatial patterns were performed for the eight morphotypes that were found in at least five of the six samplings in 2000/2001. Linear and multiple regression analyses were performed to test correlations between spatial patterns and environmental factors for statistical significance.
Morphotyping of EM
Morphotypes were described on the basis of fresh EM root tips as recommended by Agerer (1991). The first step was to examine morphological characters such as color, shape, branching patterns, and surface structure under a dissecting microscope. In addition, the anatomy of the hyphal mantle and Hartig net was described based on tangential sections through the mantle observed with a Zeiss Axioplan light microscope at 400×–1,000× magnification. Photographs of the mycorrhizal habitus and of cross-sections were taken with an MC100 microscope camera (Zeiss, Oberkochen) using Kodak EPY64T film.
Characterization of morphotypes by PCR-RFLP
Extraction of genomic DNA for PCR was performed as described by Doyle and Doyle (1990) from single EM roots tips (fresh weight 0.1–1 mg) homogenized with micropestles in 100 μl CTAB DNA extraction buffer. Purified DNA was dissolved in 100 μl sterile water and stored at 4°C. The PCR assays contained 5 μl 10× Taq polymerase reaction buffer (Promega, Heidelberg, Germany), 4 μl 25 mM MgCl2, 10 nmol of each deoxynucleotide, 50 pmol of each of the primers ITS1 and ITS4 (White et al. 1990) and 1 μl template DNA in a total volume of 50 μl. After 10 min of denaturation at 95°C, PCR was started by adding 2 U Taq DNA polymerase (Promega). The PCR program comprised 32 cycles (40 s at 92°C, 40 s at 52°C, 40 s at 72°C) using an OmniGene HB-TR3 thermocycler (MWG-Biotech, Ebersberg, Germany). PCR products were cleaved, in single enzyme digests, with AluI, EcoRI, BsuRI, HinfI or MspI (all from MBI Fermentas, St. Leon-Rot, Germany). The lengths of amplification products and restriction fragments were determined by electrophoresis on 2% agarose gels run at 10 V/cm.
Cloning and sequencing
PCR products were cloned into the pCR4-Topo Vector (Invitrogen, Karlsruhe, Germany) and transformed into TOP10 Chemically Competent Escherichia coli following the manufacturer’s protocol provided with the TOPO TA Cloning Kit. Sequencing was performed on a LI-COR DNA Sequencer Long Reader 4200 using the Thermo Sequenase fluorescent labeled primer cycle sequencing kit with 7-deaza-dGTP (Amersham Pharmacia, Little Chalfont, UK). Sequence data were compared with GenBank entries using the BLASTN program (Altschul et al. 1997).
Morphotypes with few or no emanating hyphae; cystidia and rhizomorphs missing
EM 6.2: dominating morphotype, abundance 31.4% of all EM. Branched EM; white to beige; mantle with 5–7 layers of hyphae, plectenchymatic (similar to mantle type B-H, according to Agerer 1991); emanating hyphae rare, with clamps. (Tomentella sp., see Figs. 2e, 3e).
EM 6.1: very common, abundance 10.5%. Unbranched EM; white to beige; mantle 5–7 layers, plectenchymatic (similar to type H), similar to EM 6.2; emanating hyphae rare, with clamps. (Laccaria sp., see Figs. 2h, 3h).
EM 15: rare, abundance 0.6%. Branched EM; gray; mantle 5–7 layers, plectenchymatic (similar to type B); emanating hyphae rare, with clamps. (Laccaria sp.).
EM 1: very rare, abundance 0.1%. Unbranched EM; beige; mantle 3–5 layers; outer hyphal layers pseudoparenchymatic (similar to type L), inner layers plectenchymatic (type H-E); emanating hyphae missing.
EM 21: very rare, abundance 0.01%. Branched EM; dark brown; mantle 2–4 layers, plectenchymatic (similar to type C-E); emanating hyphae missing.
Morphotypes with well developed emanating hyphae; cystidia and rhizomorphs missing
EM 5: very common, abundance 14.0%. Unbranched EM; black; mantle 3–5 layers, plectenchymatic (similar to type B-E); emanating hyphae septate without clamps. (Phialocephala fortinii, see Figs. 2c, 3c).
EM 18: common, abundance 4.9%. Unbranched EM, often spherical; black, shining; mantle 2–3 layers, plectenchymatic (similar to type G); emanating hyphae much longer than mycorrhizal roots, septate without clamps. (Cenococcum geophilum, see Figs. 2k, 3k).
EM 3: scattered, abundance 1.3%. Branched EM with short lateral roots; brown; mantle 3–5 layers, pseudoparenchymatic (similar to type M); emanating hyphae septate with clamps. (Lactarius sp., see Figs. 2g, 3g. As Lactarius EM usually do not possess clamps, the “emanating hyphae” were possibly foreign hyphae growing on the mantle).
EM 20: rare, abundance 0.9%. Branched EM with long lateral roots; blackish brown; mantle 3–5 layers, pseudoparenchymatic (similar to type L-M); emanating hyphae septate with clamps. (Thelephoraceae, see Figs. 2f, 3f).
EM 7: rare, abundance 0.5%. Branched EM; light brown; mantle 2–4 layers, plectenchymatic (similar to type B-E); emanating hyphae woolly, septate with clamps.
EM 11: rare, abundance 0.4%. Unbranched EM; beige; mantle 2–3 layers, plectenchymatic (similar to type A); emanating hyphae not septate.
EM 12: very rare, abundance 0.1%. Unbranched EM; brown; mantle 4–6 layers, pseudoparenchymatic (similar to type Q); emanating hyphae septate without clamps.
EM 13: very rare, abundance 0.1%. Branched EM; brown; mantle 2–3 layers, plectenchymatic (similar to type B); emanating hyphae septate with clamps.
EM 9: very rare, abundance 0.1%. Unbranched EM; brown; mantle 1–2 layers, plectenchymatic (similar to type H-M); emanating hyphae septate without clamps.
EM 4: very rare, abundance 0.04%. Branched EM; brown; mantle 3–5 layers, pseudoparenchymatic (similar to type L); emanating hyphae septate without clamps.
Morphotypes with cystidia
EM 17: scattered, abundance 1.9%. Branched EM, often globular; orange-brown; mantle 4–6 layers, pseudoparenchymatic (similar to type L-M); few needle-shaped cystidia, other emanating hyphae absent. (Tuber sp., see Figs. 2j, 3j).
EM 19: very rare, abundance 0.1%. Unbranched EM; brown; mantle 2–4 layers, plectenchymatic (similar to type D); many needle-shaped cystidia, other emanating hyphae absent. (Tuber sp.).
Morphotypes with rhizomorphs
EM 22: scattered, abundance 1.3%. Branched EM; silver-white; mantle 5–6 layers, plectenchymatic (similar to type A-B); emanating hyphae septate with clamps; well developed rhizomorphs. (Agaricales, see Figs. 2l, 3l).
EM 16: scattered, abundance 1.2%. Branched EM; yellow; mantle 8–10 layers, plectenchymatic (similar to type A); emanating hyphae rare, septate without clamps; well developed, strong rhizomorphs. (Boletaceae, see Figs. 2i, 3i).
EM 23: Rare; abundance 0.6%. Branched EM; silvery; mantle 4–6 layers, outer layers plectenchymatic (similar to type H), inner layers nearly pseudoparenchymatic; emanating hyphae septate with clamps; fine rhizomorphs formed by only few hyphae. (Hebeloma helodes).
EM formed by two mycobionts
All mixed EM belonged to the white/beige morphotypes EM 6.1 (Laccaria sp.) or EM 6.2 (Tomentella sp.), secondarily colonized by the mycobiont of either EM 5 (P. fortinii) or EM 18 (C. geophilum). The mixed type EM 6.1/5 (Laccaria/Phialocephala) was very common (abundance 15.9%), with developmental levels of Phialocephala ranging from single hyphae colonizing the EM surface to nearly closed layers of hyphae. In contrast, the mixed type EM 6.2/5 (Tomentella/Phialocephala), characterized by a loose colonization with Phialocephala hyphae, occurred scattered (abundance 2.7%). In the case of the rare mixed types between EM 6.1/18 (Laccaria/Cenococcum, abundance 0.2%) and EM 6.2/18 (Tomentella/Cenococcum, abundance 0.1%), secondary colonization was mostly restricted to the tip of the mycorrhizas.
Assessment and identification of morphotypes by PCR-RFLP and ITS sequencing
Restriction fragment length (in bp) of PCR-amplified internal transcribed spacer (ITS) regions of the 14 most common ectomycorrhiza (EM) morphotypes colonizing different hybrid aspen lines grown on an experimental field in Großhansdorf (Germany). Fragment length was determined from the ITS sequences in all cases except EM 3, EM 10 and EM 15, for which it was estimated after PCR-RFLP analysis (Kaldorf et al. 2002)
Identification of the mycobionts of EM morphotypes colonizing different hybrid aspen lines
Best BLAST hit
Pezizales sp. d334, AF266709
Pezizales EM 2
Phialocephala fortinii, AY078141
Laccaria laccata, AF204814
Tomentella ellisii, AF272913
Pezizales EM 10
Thelephoraceae sp. C.t.-3, AF184742
Thelephoraceae EM 14
Xerocomus pruinatus, AF402140
Boletaceae EM 16
ITS sequence, anatomy
Tuber maculatum, AF106889
ITS sequence, anatomy
Cenococcum geophilum, AY112935.1
Thelephoraceae sp. C.t.-3, AF184742
Thelephoraceae EM 20
Entoloma nitidum, AF335449
Hebeloma helodes, AF124710
Spatial patterns of EM distribution
Average light exposure of 12 spots at the experimental field. Measurements of light intensity were taken over a sunny summer day with five replications between 9:00 a.m. and 3:00 p.m. 1 m above ground. For the organization of the field and sector delimitation see Fig. 1
Light intensity (μE m−2 sec−1)
Sector 1, close to spruce forest
Sector 2, close to spruce forest
Sector 3, close to spruce forest
Sector 4, close to spruce forest
Sector 1, central
Sector 2, central
Sector 3, central
Sector 4, central
Sector 1, close to deciduous trees
Sector 2, close to deciduous trees
Sector 3, close to deciduous trees
Sector 4, close to deciduous trees
In contrast, an uneven spatial distribution related to the factors mentioned above was observed for three morphotypes. The occurrence of C. geophilum (EM 18, Fig. 6f) showed a highly significant positive correlation with the distance from the deciduous trees (r2=0.285, P<0.001) as well as a highly significant negative correlation with the light exposure (r2=0.210, P<0.001) and the distance from the spruce forest (r2=0.266, P<0.001). An opposite distribution pattern was confirmed for the only two morphotypes with differentiated rhizomorphs, namely Agaricales EM 22 (Fig. 6h) and Boletaceae EM 16 (Fig. 6i). Their occurrence was significantly positively correlated with the distance from the spruce forest (r2=0.164, P=0.003 for EM 22 and r2=0.132, P=0.007 for EM 16) and significantly negatively correlated with the distance from the deciduous trees (r2=0.152, P=0.004 for EM 22 and r2=0.098, P=0.021 for EM 16). The influence of light exposure on the distribution of Agaricales EM 22 and Boletaceae EM 16 was not significant.
Among the rare morphotypes, EM 1, EM 9 and EM 21 occurred only once in a single root sample. Other rare types (e.g. EM 4, EM 12, EM 13, EM 19) were found in less than five samples (abundance per sample below 10%), located in two or three different sectors in all four cases. For these rare morphotypes, no spatial preferences within the field were apparent. Even Lactarius sp. (EM 3), EM 11 and Thelephoraceae EM 20 were restricted to three or less samples, but in contrast to the other rare morphotypes mentioned above, these morphotypes were then dominant within samples (maximal abundance of 44.4% for EM 11, 65.2% for EM 20 and 75.1% for EM 3).
Community structure of aspen EM
The combination of morphotyping and PCR-based analyses of parts of the rDNA is a well-accepted method to describe EM and analyze their community structure (Horton and Bruns 2001). The community observed at our experimental field seems to be typical for a young monoculture of trees. The number of EM morphotypes detected—23— is in accordance with the 16 to 24 morphotypes reported in young stands of alder, birch, pine and spruce (Hashimoto and Hyakumachi 2000; Jonsson et al. 1999; Kårén and Nylund 1997; Pritsch et al. 1997).
Species within the Thelephoraceae and Russulaceae are among the most frequent and abundant EMF in Europe and North America (Horton and Bruns 2001). Forming 38% of all EM, the Thelephoraceae (Tomentella sp. EM 6.2, Thelephoraceae EM 14 and Thelephoraceae EM 20) were dominant on our field, while the Russulaceae represented only 1.3% of the EM with one morphotype (Lactarius sp. EM 3). The two black morphotypes (EM 5 and EM 18) were formed by two widespread fungi, P. fortinii and C. geophilum, respectively. P. fortinii (EM 5) occurs in the entire temperate zone of the northern hemisphere without apparent host specificity (Addy et al. 2000, and references therein). It belongs to the artificial group of dark septate endophytes, for which interactions with plant roots ranging from parasitic to mycorrhizal have been described (Jumpponen et al. 1998; Wang and Wilcox 1985). C. geophilum (EM 18) is considered as a nearly ubiquitous EMF with worldwide distribution and low, or no, host specificity (Kovács et al. 2000; Wurzburger and Bledsoe 2001). These two fungi were involved in the formation of about 40% of all EM at our field site, either alone, or in mixed EM with Tomentella sp. (EM 6.2) or Laccaria sp. (EM 6.1). Laccaria, which formed 10.5% of all EM, has also been described as one of the genera colonizing deciduous trees like birch at early stages of EM succession (Mason et al. 1983). The last common morphotype found, Pezizales EM 2, was identified only at the order level. Consequently, its ecological abilities cannot be discussed. Together, these 6 EM fungi formed nearly 90% of all mycorrhizas, while none of the other 17 morphotypes had an abundance above 2%.
EM morphotypes often occur as clusters (Horton and Bruns 2001, and references therein), making it difficult to establish correlations between their distribution and local variations of factors such as pH, temperature, moisture or nutrient availability in the soil. Using in vitro cultures, Sanchez et al. (2001) demonstrated that changes in pH, water supply and temperature had different effects on the growth of eight EMF species. In a field study, increasing atmospheric nitrogen deposition was found to correlate with a drastic change in EM community structure (Lilleskov et al. 2002). Both observations demonstrate that EMF species with ecophysiological specificity exist; such specificity has been suggested to be an important factor for EMF biodiversity and community structure (Bruns 1995). In normal field situations however, attempts to analyze such relationships is traditionally hampered by overlapping heterogeneities in the soil and the host plants community with small-scale microclimatic variations.
A central goal of the field release experiment in Großhansdorf was the assessment of mycorrhizal colonization of transgenic aspen. This aspect was considered in a former article showing that in the first 2 years of the experiment (1998/1999), the expression of the rolC gene from A. rhizogenes in transgenic aspen had no influence on their mycorrhization rates. Additionally, other than a reduced compatibility between P. fortinii (EM 5) and one of the four 35S-rolC transgenic aspen lines, no differences in EM community composition were observed (Kaldorf et al. 2002). In this context, the release of genetically closely related aspen clones on a homogeneous field offered a unique opportunity to follow the establishment of an EM community and to analyze the effect of a few varying ecological factors such as the light and drought regime or the influence of neighboring forest stands.
Due to the context of the experiment, the sampling technique used had to allow for correct assignment of each root tip to an individual tree and should not endanger the survival of the saplings. Analyses of EM were therefore based on a 1-m-long sector of a single lateral root per tree. This procedure is uncommon, as randomly taken soil cores are used in most studies on EM community structure (Horton and Bruns 2001). When two samples from the same tree were compared, very similar EM compositions were found, as indicated by high Sørensen index values. This demonstrates that our single sample procedure correctly reflected the EM community of each tree analyzed. Therefore, analysis of the spatial EM distribution was possible, and two patterns in the establishment of EMF were recognized.
The first pattern results from an invasion starting from mycorrhizal trees adjacent to the experimental field and can be termed “vicinal invasion”. This concerns on the one hand the morphotypes Boletaceae EM 16 and Agaricales EM 22, whose occurrence, based on the r2 values, appeared to be linked to the proximity of deciduous trees. On the other hand, it applies to the area colonized by C. geophilum (EM 18), which correlated with the distance to the bordering spruce forest. C. geophilum has been described as one of the most drought-resistant EMF (Pigott 1982). Most likely, the presence of C. geophilum (EM 18) mainly on the wettest part of our experimental field was not related to this ecological trait. However, this does not exclude that the extent of a vicinal invasion can be influenced by local ecological factors. Even if not proven by the correlation analysis, the extension of Boletaceae EM 16 and Agaricales EM 22 at the more sunny and dry part of the field fits well with the fact that these morphotypes were the only ones with a hydrophobic hyphal mantle surface and well differentiated rhizomorphs, two characters typical of drought-resistant EMF with medium- to long-distance exploration strategies (Agerer 2001; Unestam and Sun 1995). As indicated by r2 values below 0.3, additional environmental factors (e.g., temperature or soil moisture) as well as spread of EM fungi by chance may have additionally contributed to the formation of spatial patterns.
The second pattern of EM establishment observed in the field can be termed “random invasion”. This pattern is well documented by the extension of the three most common morphotypes, EM 5 (P. fortinii), EM 6.1 (Laccaria sp.) and EM 6.2 (Tomentella sp.), which were found randomly distributed on over 90% of the root samples. The origin and the time of the invasion cannot be determined, but several sources can be hypothesized: (1) an invasion with the plants themselves if they were already mycorrhizal in the greenhouse prior to their outplanting; (2) a massive invasion via immediately germinating spores or via any propagules resting in the soil prior to planting. For the three morphotypes mentioned above, the invasion can be considered as massive and the mycobionts as well-adapted to the ecological conditions of a young plantation. A random invasion can also result in a more discrete extension of EMF. This was the case for rare morphotypes detected only a few times on different independent trees (EM 4, EM 12, EM 13, EM 19) or even only once on a single tree (EM 1, EM 9 and EM 21). These morphotypes did not spread after their punctual random invasions in 1998–2000 and therefore seemed not to be competitive compared to the dominating morphotypes. Scarce random invasion of more competitive EMF is illustrated by the distribution patterns of Lactarius sp. EM 3, EM 11 and Thelephoraceae EM 20, as these morphotypes were very abundant on the few trees on which they were detected.
The observation of two establishment strategies of EM on our field site (vicinal invasion and random invasion) is a first documentation that the filter theory proposed to explain assembly rules in the regeneration of plant communities (Fattorini and Halle, in press) could also be suitable for microorganisms such as EMF. According to this theory, the two factors determining the successful establishment of an organism would be its arrival at the plot and its ability to pass a “filter” constituted by the local ecological constraints. Verifying this kind of theory in the field is especially difficult in the case of EMF. However, our study indicates that it becomes possible when working under simplified conditions with clonal plants on roughly homogenized fields.
This work was kindly supported by a grant from the Federal Ministry of Education and Research, Germany (Grant number: 0311387). We wish to thank Dr. J. Perner for his help with statistical analysis and T. Spribille for critical reading of the manuscript.