Introduction

Bog ecosystems belong to the oldest vegetation forms with more or less constant conditions for thousands of years. The bryophyte genus Sphagnum, consisting of approximately 300 different species, is worldwide distributed and a dominant component of the mire and bog vegetation (Daniels and Eddy, 1985). During bog formation, the lower parts of the Sphagnum plants die off and build deep layers of dead plant material, the so-called peat. Sphagnum bogs are unique habitats for numerous plant and animal species, and their environmental and ecological significance is immense because of the well-recognized role in the global carbon budget and emission of the greenhouse gas methane (Succow and Joosten, 2001; Raghoebarsing et al., 2005). High acidity (pH 3.5–5.0), low temperature, water saturation together with oxygen deficiency and extremely low concentration of mineral nutrients are characteristic for these extreme habitats. Furthermore, Sphagnum leafs are highly specialized: they form a special tissue of living, chlorophyll-containing chlorocytes and dead, cell content-free hyalocytes, which are responsible for the huge potential to store water. Due to their morphology and ecology, Sphagnum plants are unique host plants for microorganisms. So far, only the microbial populations involved in CH4 cycling, that is methanotrophic bacteria (Dedysh et al., 1998, 2001; Dedysh, 2002; Morris et al., 2002; Raghoebarsing et al., 2005; Merila et al., 2006) and methanogenic archaea (Horn et al., 2003; Sizova et al., 2003; Kotsyurbenko et al., 2004) have attracted research interest. Studies of Mitchell et al. (2003) analyzed the bacterial abundance and activity in water samples collected from aquatic Sphagnum habitats, while in other studies the bacterial community at the oxic and anoxic regions of Sphagnum peat bogs was investigated (Morales et al., 2006). However, the composition of bacterial communities associated with living Sphagnum plants remains largely unknown.

Several studies have identified different biotic and abiotic factors influencing the structural and functional diversity of bacterial communities associated with different crop plants (Berg et al., 2002, 2006; Kowalchuk et al., 2002; Garbeva et al., 2004; Costa et al., 2006). Evidence has also been presented that the structure of rhizosphere microbial communities is influenced by the plant species, because of differences in root exudation and rhizodeposition (Brimecombe et al., 2001; Smalla et al., 2001; De Boer et al., 2006). Host specificity is also known for the microbial communities living in the phyllosphere (Yang et al., 2000). However, Costa et al. (2006) showed that the impact of soil type is higher than those of the plant species, as analyzed for different crops at different sites in Germany. Bezemer et al. (2006) found interaction between the vegetation and terrestrial microbial communities. Additionally, biogeographical aspects and endemicity of microorganisms (see review in Ramette and Tiedje, 2007) can influence biodiversity of plant-associated bacteria. The extent to which plants are associated with specific bacterial communities in their natural habitats in different geographical regions, especially in the case of bryophytes, is still unclear.

The objective of this work was to analyze the associated bacteria of two different Sphagnum species, Sphagnum magellanicum and Sphagnum fallax (class Sphagnopsida, family Sphagnaceae), which were isolated from six bog ecosystems at different sites in the Northeast of Germany and Southwest of Norway, and to determine the extent to which the bryophyte species, the diversity of the surrounding plant community and the geographical region influence biodiversity. S. magellanicum BRID. (section Sphagnum) is typical for strong acidic, oligotrophic and ombrotrophic habitats, whereas S. fallax H. KLINGGR. (section Cuspidata) grows in weakly acid, more mesotrophic situations influenced by minerotrophic groundwater. Both bryophytes colonize extreme and specific habitats, and they belong to the typical and very important vegetation in peat bogs (Daniels and Eddy, 1985; Nordbakken et al., 2003). The selection of species and sampling locations were designed to ensure (i) bryophyte species difference, (ii) ecological difference (oligotrophic and mesotrophic mires) and (iii) geographical difference (Germany/Norway). Molecular tools were used for investigating the biodiversity of the bacterial populations associated with the two Sphagnum species. Particular interest was layed on the analysis of members of the genus Burkholderia and Serratia, which belong to the dominant bacteria associated with Sphagnum (Opelt and Berg, 2004; Belova et al., 2006).

Materials and methods

Sampling and isolation of the bacterial fraction

Adult gametophytes of both bryophytes S. magellanicum and S. fallax were sampled in three different natural habitats in the Northeast of Germany and Southwest of Norway (Table 1). The bryophyte samples were collected in August and September in 2004. We have chosen two common European Sphagnum species growing in light and wet mires, but in different ecological situations (Table 2). The ecological characters of the different habitats were characterized by different abiotic conditions, especially with regard to moisture, soil pH and nutrient content.

Table 1 Sampling locations
Full size table
Table 2 Some ecological characters of the habitats
Full size table

From each of the six investigated Sphagnum bogs, six (three from S. magellanicum and three from S. fallax), different composite samples (consisting of five single samples of approximately 20 Sphagnum plantlets) were taken and analyzed separately. The living green parts of the gametophytes were placed into sterile petri dishes and transported to the laboratory, and then 5 g were transferred to a sterile stomacher bag. Extraction of the moss-associated bacteria from the gametophytes was carried out as described by Opelt and Berg (2004).

Scanning electron microscopy

Gametophytes were fixed in glutaraldehydephosphate buffer (2%; 0.1 M; pH 7.2) and post fixed in 2% glutaraldehyde buffer. After removal of the fixative by a wash in phosphate buffer, samples were dehydrated, critical point dried and coated with gold before undergoing scanning electron microscopy (Carl Zeiss, Oberkochen, Germany).

Total-community DNA isolation

The total-community DNA of bacterial cell consortia was extracted as described by Martin-Laurent et al. (2001). For mechanical lysis, samples were homogenized in a FastPrep Instrument (Qbiogene, BIO101, Carlsbad, CA, USA) for 30 s at speed 5.0 m/sec. The extracted DNA was purified by using the Geneclean Spin kit (Qbiogene) according to the manufacturer's protocol.

Molecular analysis by SSCP

Fingerprinting of the moss-associated bacterial communities by single-strand conformation polymorphism (SSCP) was carried out as described by Schwieger and Tebbe (1998). Briefly, 16S rRNA fragments were amplified by PCR using bacterial DNA isolated from moss tissue as the template, with the eubacterial primer pair Unibac-II-515F/Unibac-II-927rP along with specific primers for Serratia (Lieber et al., 2003). For the analysis of the Burkholderia sp community a nested PCR was performed. In a first PCR, the eubacterial primer pair Eub1/Eub2 (Lane, 1991) was used in a 20 μl reaction mixture as follows: 95°C for 5 min, followed by 10 cycles of 95°C for 30 s, 52°C for 30 s and 72°C for 1 min 40 s, followed by 20 cycles of 95°C for 30 s, and 72°C for 1 min 30 s +10 s/cycle with a 72°C final extension step for 5 min before holding at 15°C when the reaction was terminated. The second PCR was performed in a 20 μl reaction mixture using 1 μl of the product from the first PCR and the Burkholderia-specific primer pair BKH143Fw (5′-TGGGGGATAGCYCGGCG-3′) and BKH1434Rw (5′-TGCGGTTAGRCTASCYACT-3′) (Schönmann et al., unpublished). The Burkholderia-specific primers were designed using the phylogenetic software package ARB with probe design and probe match tools (Ludwig et al., 2004; www.arb-home.de) under Linux Suse 8.2 with the ARB 16S rRNA database (ssu_jan04_full.arb), which was updated with the available sequences found in the hierarchy browser of the ribosomal database project (rdp) (http://rdp.cme.msu.edu/index.jsp) (Release 9.27, 25. April 2005, 136355 16S rRNA sequences). Only sequences over 1200 bp length belonging to the path: class Proteobacteria/order Burkholderiales/family Burkholderiaceae/genus Burkholderia (May 2004: 0/525/0) out of Bergey's Taxonomy were chosen. Parameters for the Burkholderia-specific PCR were as follows: 94°C for 4 min, followed by 40 cycles of 94°C for 40 s, 52°C for 40 s and 72°C for 1 min 30 s, with a 72°C final extension step for 10 min. After the second PCR, we obtained DNA fragments of 1300 bp. The third PCR was performed to obtain smaller fragments for the analysis with the SSCP. For this PCR, the eubacterial primer pair Unibac-II-515F/Unibac-II-927rP (Lieber et al., 2003) was used. For the 60 μl reaction mixture, 6 μl from a 1:50 dilution of the second PCR product were used as the template and we obtained DNA fragments of 412 bp. The amplicons were separated by using the TGGE Maxi system (Biometra, Göttingen, Germany) at 400 V and 26°C for 26 h using 8% (wt/vol) acryl amide gels. Silver staining was used to detect DNA in SSCP gels (Bassam et al., 1991).

Construction of a Burkholderia clone library

16S rRNA gene fragments of Burkholderia sp were amplified by using the Burkholderia-specific forward primer BKH143Fw and the reverse primer BKH1434Rw (Schönmann et al., unpublished). A nested PCR was performed using EX Taq polymerase (Takara Biomedicals, Otsu, Shiga, Japan) and the eubacterial primers 616 V and 630R (Brosius et al., 1981) in a first amplification round, followed by a second PCR using the Burkholderia-specific primers BKH143Fw and BKH1434Rw. Amplification reactions were set up in a total volume of 50 μl reaction mix containing community DNA in concentrations ranging from 25 to 130 ng/μl. Reaction mixtures contained 25 μ M primers, 2 mM MgCl2 (Promega, Madison, WI, USA) and 1.5 U of Ex Taq/Taq polymerase. Thermal cycling was carried out as follows: initial denaturation at 94°C for 4 min, followed by 40 cycles of denaturation at 94°C for 40 s, annealing at 52°C for 40 s and elongation at 72°C for 1 min 30 s. Cycling was completed by a final elongation at 72°C for 10 min. The PCR amplification products were separated on 1.0% agarose gels and the bands with the expected sizes were excised and the DNA was purified using the QIAquick Gel Extraction kit (Qiagen, Hilden, Germany) according to the manufacturer's instructions. The amplicons were ligated into the cloning vector pCR 2.1 (Invitrogen Corp., Carlsbad, Canada) and transformed into Escherichia coli. Cloning, restriction enzyme analysis, and transformation of E. coli XL1 blue were performed essentially as described previously (Sambrook et al., 1989). For phylogenetic analyses, 16S rRNA sequences of closely related organisms were collected at the public database rdp (http://rdp.cme.msu.edu). The 16S rRNA moss clone sequences were added to the alignment of about 200 closely related small-subunit rRNA sequences by using the alignment tool of the ARB program package (O Strunk and W Ludwig, http://www.arb-home.de). Alignments were refined by visual inspection. Phylogenetic analyses were performed using neighbor joining.

Computer-assisted cluster analysis and calculation of diversity indices

Computer-assisted evaluation of bacterial community profiles obtained by SSCP was performed by using the GelCompar program (Applied Maths, Kortrijk, Belgium). To compare the community fingerprints, the SSCP gels were scanned transmissively (Epson perfection 4990 Photo, Nagano, Japan) to obtain digitized gel images. The gels were normalized, which made it possible to compare band pattern of different gels, and the background was subtracted (details at the manufacturer manual). The Pearson's correlation index (r) for each pair of lanes within a gel was calculated as a measure of similarity between the community fingerprints. Finally, cluster analysis was performed by applying the unweighted pair group method using average linkages (UPGMA) to the matrix of similarities obtained. To assess differences in the diversity of the various bacterial communities, the diversity index H′ was calculated according to Shannon and Weaver (1999). H′ was calculated according to the equation H′=−ΣSi=1 [(ni/N)ln(ni/N)] where ni is the relative abundance of a single taxon (a single SSCP band), N is the total abundance of all taxa, and S is the total number of abundant species. The relative abundance of the SSCP bands was calculated using the GelCompar program. For this, the different peaks of a densitometric curve of each SSCP lane were analyzed. For each SSCP pattern, the maximum peak value was 255 (100%) and only peaks (bands) that showed a value more than 5% were included in the calculation. Statistically significant differences were determined by the Mann–Whitney U-test (P⩽0.05).

Habitat characterization by the analysis of the composition of the plant vegetation

In each of the 12 bryophyte collecting points, we recorded all bryophyte and vascular plant species (lichens did not occur, fungi were ignored as well as microscopic plants) in a square of 2 × 2 meter around the sampling point. Plant abundance was estimated by Braun–Blanquet approach with a conventional scale first published by Reichelt and Wilmans (1973). The collected data were analyzed by the unweighted TWINSPAN ordination analysis (Hill, 1979). TWINSPAN is a common numerical classification method by hierarchical division specifically developed for vegetation data. The program is based on the concept that a group of ecological similar sample plots will have preferential species. Species and relevés are sorted depending on the maximum likelihood based on a reciprocal averaging ordination, refined by a discriminant function analysis (Hill, 1979). For ecological characterization of the Sphagnum habitats, we used indicator values by Ellenberg et al. (1991). These are empirical developed ordinal values of environmental variables along ecological gradients of light, temperature, continentality, soil moisture, soil pH, fertility and salinity. They have been widely used and have been tested, improved and verified by measurements for central European vascular plants and bryophytes during a period of many decades (Diekmann, 2003).

For multivariate analysis of the vegetation data, we used the indirect unimodal correspondence analysis (CA) of the software package Canoco 4.5 (Lepš and Šmilauer, 2003). Indicator values are given as supplementary environmental data. Species nomenclature was followed according to Ellenberg et al. (1991).

Results

Microscopic observation of surfaces of gametophytes

Impressions of the specific microenvironment were obtained by analyzing gametophytes by electron microscopy (Figure 1). Figures 1a and b show that the surfaces of leaves from S. fallax and S. magellanicum, respectively, are colonized by bacteria. The Sphagnum tissue forms a regular pattern of alternating green (chlorocytes) containing chloroplasts and hyaline cells, which appear to shape a very specialized microenvironment for the bacteria. The hyaline cells with large pores, normally filled with water, are dead at maturity and now serve to provide water. In general, both Sphagnum species showed similar surface structures and they appear to form similar microhabitat structures for bacteria.

Figure 1
figure 1

Scanning electron micrographs of the surface of gametophytes of S. fallax (a) and S. magellanicum (b). Triangles indicate the microorganisms on the bryophyte surface.

Full size image

Molecular fingerprinting and diversity of Sphagnum-associated bacterial communities

SSCP gels showed that the diversity of 16S rRNA fragments amplified from Sphagnum community DNA is high. For both bryophyte species we found between 30 and 50 bands for the total bacterial and the Burkholderia-specific community patterns. About 20–30 bands were found for the Serratia-specific SSCP patterns. For eubacterial as well as group-specific community patterns, dominant bands were detectable.

To analyze the bryophyte specificity of the bacterial community composition, Pearson's correlation coefficients were compared by UPGMA. The eubacterial community patterns of S. magellanicum from different geographical sites formed one cluster and the patterns of S. fallax formed another cluster at similarity of 15% (Figure 2). These two clusters indicate that the Sphagnum-associated eubacterial communities are specific for the moss species, irrespective of the sampling site. The similarity between the associated bacterial communities of S. fallax from German and Norwegian sites was up to 70%, whereas the similarity between the associated bacterial communities of S. magellanicum from both sites was up to 65%. In contrast to the high plant specificity, no influence of the geographical site could be detected. With regard to the Serratia- and Burkholderia-specific SSCP patterns also a high degree of similarity between S. magellanicum-associated bacterial communities from different geographical regions was observed (Figures 3 and 4). Likewise, the similarities of S. fallax-associated communities of the genera Serratia and Burkholderia from different sites was very high. The patterns obtained from S. magellanicum and S. fallax, respectively, formed one cluster at a similarity level of 20% for Serratia and 24% for Burkholderia (Figures 3 and 4). For both bryophytes we found that the Serratia- or Burkholderia-specific SSCP patterns of one bryophyte and one sampling site is not so often clustered together than for the eubacterial communities. However, for Serratia-specific communities of S. fallax from different geographical sites, the similarity was up to 95% (Figure 3). The geographical site had a greater effect on the Burkholderia-specific SSCP patterns of associated bacterial communities in case of S. fallax than S. magellanicum. In fact, at a similarity level of 52%, the patterns of associated Burkholderia-specific communities of S. fallax from the German and Norwegian sites, respectively, formed only one cluster (Figure 4).

Figure 2
figure 2

Dendrogram based on amplified 16S rRNA fragments of S. magellanicum-associated bacterial communities from different sites in Germany (G1-3SM) and Norway (N1-3SM) and from S. fallax from different sites in Germany (G1-3SF) and Norway (N1-3SF) obtained by using eubacterial primers and separated by SSCP. The patterns obtained were grouped by UPGMA. Double-headed vertical arrows indicate the similarity for the groupings. SM=Sphagnum magellanicum; SF=Sphagnum fallax; SSCP=single-strand conformation polymorphism.

Full size image
Figure 3
figure 3

Dendrogram based on amplified 16S rRNA fragments of S. magellanicum-associated bacterial communities from different sites in Germany (G1-3SM) and Norway (N1-3SM) and from S. fallax from different sites in Germany (G1-3SF) and Norway (N1-3SF) obtained by using Serratia-specific primers and separated by SSCP. The patterns obtained were grouped by UPGMA. Double-headed vertical arrows indicate the similarity for the groupings. SM=Sphagnum magellanicum; SF=Sphagnum fallax; SSCP=single-strand conformation polymorphism.

Full size image
Figure 4
figure 4

Dendrogram based on amplified 16S rRNA fragments of S. magellanicum-associated bacterial communities from different sites in Germany (G1-3SM) and Norway (N1-3SM), and from S. fallax from different sites in Germany (G1-3SF) and Norway (N1-3SF) obtained by using Burkholderia-specific primers and separated by SSCP. The patterns obtained were grouped by UPGMA. Double-headed vertical arrows indicate the similarity for the groupings. SM=Sphagnum magellanicum; SF=Sphagnum fallax; SSCP=single-strand conformation polymorphism.

Full size image

The Shannon diversity index H′ was calculated for the bacterial communities from all sampling locations (Table 3). Diversity indices of H′=3.24 and H′=3.25 on average were calculated for the eubacterial communities associated with S. fallax in Germany and Norway, whereas the indices for S. magellanicum were significantly lower (H′=2.69 and H′=2.91). In contrast to the specificity for both Sphagnum species, there was no statistically significant difference between the different geographical sites.

Table 3 Biodiversity indices for eubacterial SSCP patterns and vascular plant and bryophyte community calculated using the Shannon diversity index
Full size table

Analysis of a 16S rRNA gene library of moss-associated Burkholderia species

The 16S rRNA genes of members of the genus Burkholderia associated with the mosses were PCR amplified and cloned as described in Materials and methods. Altogether, 38 clones were sequenced and compared to sequences deposited in GenBank. Sequences were designated as follows: M1 for S. fallax (Dänschenburger Moor, Germany), M2 for S. magellanicum (Sørdalen peat mire, Norway) and M3 for S. magellanicum (Seljestad peat mire, Norway). Analysis of the sequence polymorphism of the 16S rRNA genes was used to infer the phylogenetic positions of the sequences (Figure 5). Examination of the phylogram revealed that the sequences formed distinct phylogenetic clusters.

Figure 5
figure 5

16S rRNA gene-based neighbor joining tree showing the phylogenetic relationship of 38 peat bog clones (M1, M2 and M3) to the nearest phylogenetic neighbors within the genus Burkholderia. Pandorea apista typestrain (AF139173) was used as outgroup in this analysis. Bootstrap values for 1000 replications superior to 50% are indicated. Type strains are depicted as T. The scale bar represents 0.1 substitution per nucleotide position.

Full size image

A large group of clones formed a cluster with the novel species Burkholderia bryophila (strains 1S18, A5, 1A11 and S512), which was isolated from Sphagnum rubellum (Vandamme et al., in press) and Burkholderia sp strain Y, which was recently isolated from a Sphagnum peat bog in Estonia (Belova et al., 2006). Furthermore, two sequences from Burkholderia sp isolates from woodland (WR19X and RW36) and from Burkholderia sp SB1, F4W and F4, all of which were isolated in the course of the analysis of the microbial community associated with Sphagnum peat bogs by Belova et al. (2006), were located within the same cluster. Four sequences (M2-K4, M2-K14, M3-K2 and M3-K6) clustered with the heavy metal resistant Burkholderia sp strain DM-Ni1, which was isolated from a moderate acid mine drainage (DQ419958), Burkholderia phenazinium and Burkholderia megapolitana, which was recently isolated from Aulacomnium palustre, grown in the natural reserve ‘Ribnitzer Großes Moor’ (Mecklenburg-Pommerania, Germany). In conclusion, this entire branch contains mainly strains associated with the peat of acidic Sphagnum bogs and acid mine drainage. Another large cluster forms a separate branch that is most closely related to B. cepacia (92–98% similarity). Among these sequences, M1-K13 and M2-K81 showed high degrees of similarities (98%) to the 2,4-dichlorophenoxyacetic acid degrading strain Burkholderia sp Ff54 and B. cepacia 383, an isolate from forest soil in Trinidad (Stanier et al., 1966). A single sequence, M1-K7, showed 92% similarity to Burkholderia sp strain SB6, which was isolated from Belova et al. (2006) in a tundra bog (Vorkuta) and clustered together with Burkholderia sp strain DM5 isolated from a moderate acid mine drainage.

Habitat characterization and composition of the different plant communities

The different Sphagnum locations in Germany and Norway were selected with respect to abiotic conditions including moisture, soil pH and nutrient content. The occurrence and abundance of bryophyte and vascular plant species in the 12 sampling squares is shown in Table 4. Ten species correlated to S. magellanicum, and four species correlated to S. fallax were detected. Several bryophyte or vascular plant species were found only once. Our data provide strong evidence that different groups of species prefer the S. magellanicum and the S. fallax habitats. The occurrences of the different plant species were evaluated by the TWINSPAN analysis. The ordination of the species list by TWINSPAN analysis is provided in Figure 6. Regardless of the composition of the plant communities, all S. magellanicum sampling sites and all locations of S. fallax showed a high level of similarity, respectively. It is evident that the habitat, as characterized by the specific abiotic conditions, including pH and nutrient availability, has a higher influence for the bryophyte and vascular plant species composition than the geographical distance.

Table 4 Vegetation relevés (bryophyte and vascular plant species) of the 12 sampling squares
Full size table
Figure 6
figure 6

Dendrogram based on the composition of different plant communities analyzed by recording all bryophyte and vascular plant species around the different S. magellanicum sampling points in Germany (G1-3SM) and Norway (N1-3SM), and the different S. fallax sampling points in Germany (G1-3SF) and Norway (N1-3SF) by TWINSPAN analysis. SM=Sphagnum magellanicum; SF=Sphagnum fallax.

Full size image

The ecological characteristics of the macrohabitats were determined by the aid of ecological indicator values (Table 2). The results were in good agreement with common geobotanical knowledge about the habitat preferences of both Sphagnum species. For better evidence, we further checked synoptic data from 465 vegetation relevés of the typical S. fallax community from Northeast Germany (Timmermann, 2001) and 104 relevés from Scandinavia (Dierßen, 1996) as well as 382 vegetation relevés of the S. magellanicum community from Northeast Germany and 83 relevés from Scandinavia by the same authors (Table 2). With regard to the different abiotic conditions, the habitats of both Sphagnum mosses are highly specific. The habitats of S. magellanicum were characterized by an extremely acidic pH value (indicator value 1.3–1.7), and an extremely low nutrient content (indicator nutrient value 1.0–1.6). However, the locations of S. fallax showed a higher nutrient content (indicator nutrient value 1.8–2.8) and pH value (indicator value 2.3–3.0) than those of S. magellanicum (Table 2). A CA of the sampling sites and selected species revealed statistically significant differences (Figure 7). Indicator values are given as supplementary environmental data. The first and second axes explain 39.2% and 14.8% variability of the data set, respectively. The correlation coefficient of the pH value is 0.9364, and of the nutrient value 0.8492. Interestingly, the same differences between both Sphagnum species were found in the Western Alps by measurements in the field too (Miserere et al., 2003).

Figure 7
figure 7

Correspondence analysis biplot (indirect unimodal gradient analysis) of sampling sites and selected species of Table 4. The first and second axes explain 39.2% and 14.8% variability of the data set, respectively. Correlation coefficient of pH value is 0.9364, of nutrient value 0.8492 regarding the first axis. Dashed line indicates the variability of the two communities, and full line the high correlated species. G=Germany, N=Norway, SF=Sphagnum fallax, SM=Sphagnum magellanicum; Arabic numerals represent the sampling site (1–3). Species names are given in a 3+4 abbreviation manner.

Full size image

The average species richness was much higher in the S. magellanicum habitats (9–19 species, average 11.7) than in S. fallax habitats (4–17 species, average 8.0) (Table 4). This was confirmed by Shannon diversity index, which was based on the plant communities of all sampling sites (Table 3). For statistical significance (α=0.01) of these results, 291 relevés of S. fallax community and 122 relevés of the S. magellanicum community from Northeast Germany using data by Timmermann (2001) were checked. For an average relevé area of 4 m2, we calculated an average species number of 8.6 for the S. fallax community and 12.9 for the S. magellanicum community.

Discussion

In this study we were able to demonstrate that the bacterial communities associated with bryophytes of the genus Sphagnum show a high degree of host specificity independent of the geographical region, where they occurred: in temperate bogs in Germany or in boreal bogs in Norway. The same was found to be true for the surrounding plant communities, which were highly specific for each Sphagnum species independent of the site. On the other hand, biodiversity, which was analyzed for the bacterial as well as plant communities, was different: the S. fallax-associated bacterial communities displayed a statistically significant higher degree of diversity than those associated with S. magellanicum. In contrast, analyses of plant communities of Sphagnum-specific habitats resulted in a higher diversity of S. magellanicum-specific habitats for all six sites.

Although an impact of the plant species on their associated bacteria could be clearly shown, other factors such as soil quality, climate and geographical zones may also influence the bacterial community structure (see review in Garbeva et al., 2004; De Boer et al., 2006). Over the past few years, the role of biogeography for prokaryotic diversity was subject to debates (see review in Ramette and Tiedje, 2007). For example, strong endemicity of fluorescent Pseudomonas genotypes, which were isolated from soil samples from 10 sites on four continents, was observed, suggesting that these heterotrophic soil bacteria are not globally mixed (Cho and Tiedje, 2000). In this light, the high similarity of Sphagnum-associated bacteria independent of the location was unexpected. In fact, we were unable to detect a statistically significant difference between bacterial communities analyzed from Germany and Norway. Furthermore, some of the 16S rRNA gene sequences in our Burkholderia clone library showed a very high similarity with sequences of strains obtained from a Sphagnum bog in Estonia (Belova et al., 2006). Taken into consideration that, at the genus and species resolution levels, the prokaryotes tend to be cosmopolitan (Ramette and Tiedje, 2007), this result is especially interesting. Intriguingly, Burkholderia strains isolated from German and Norwegian Sphagnum bogs showed very similar BOX-PCR patterns (Opelt et al., 2007). The long history of the vegetation type can be one reason for specificity. Bog ecosystems belong to the oldest vegetation type of Europe, for example the Dänschenburger Moor investigated in this study (G2) is about 8000 years old (Precker and Krbetschek, 1996). In contrast to most vegetation types, the Sphagnum bog vegetation belongs to the azonal vegetation with vast distribution areas. This means that the composition of plant species is similar over large distances, and under the condition of water surplus, the bog vegetation can establish itself in different humid climates. Moreover, there might be a very close interaction between Sphagnum mosses and their associated bacteria. Bacteria on and in Sphagnum may be important to survive in a harsh environment; they fix nitrogen, solubilize phosphate and show plant growth promoting abilities (unpublished results).

A high microbial diversity and complexity was determined for the bacterial communities in the phyllosphere of both moss species. In a previous study it was also shown that these microbial consortia are highly active (Opelt et al., 2007). This is similar to the so-called rhizosphere effect, which describes the phenomenon that due to the roots exudates, the activity of the microorganisms in the rhizosphere is greatly enhanced relative to those of communities inhabiting other plant microenvironments or bulk soil (Brimecombe et al., 2001). However, Sphagnum mosses have no root, no rhizosphere and to transportation system through the plants. They live on peat, which is a substrate made up of partially rotted remains of themselves. However, in their phyllosphere, they are heavily colonized by microorganisms. Like the rhizosphere of higher plants, which is responsible for pathogens defence against soil-borne pathogens (Cook et al., 1995), the phyllosphere of Sphagnum plants shows a high antagonistic potential against pathogenic fungi (Opelt et al., 2007).

Interestingly, Sphagnum mosses are ectohydric and depend on high atmospheric humidity and wet conditions for growth and survival. They have the ability to release hydrogen (H+) ions in exchange for dissolved cations (Andrus, 1986), and have a high capacity for cation exchange, with more cation-binding sites per unit area of cell wall than any other plant (Gignac, 1987). This process alters the pH of Sphagnum surroundings and acidifies the Sphagnum habitat. The plant also modifies its environment considerably by accumulating peat and hence raising the vegetation surface. Sphagnum has furthermore the ability to absorb large amounts of water, usually 10–20 times its dry weight (Andrus, 1986). In conclusion, Sphagnum gametophytes can absorb water and dissolved minerals over their surfaces and this may be one reason for bacterial colonization. Also, the specialized leaf structure of content cavernous hyalocytes, as shown in Figure 1, may present an extraordinary habitat for microorganisms. For the ecosystem Sphagnum bog high bacterial abundances were also shown in earlier studies (Dedysh, 2002; Mitchell et al., 2003; Morales et al., 2006). Despite of high acidity and an extremely low nutrient content, Sphagnum bogs are colonized by large numbers of diverse microorganisms. Previous work suggested that dissolved organic carbon rather than pH affects bacterial abundance (Fisher et al., 1997).

Biodiversity in bacterial as well as plant communities were calculated using Shannon's diversity index (Shannon and Weaver, 1949), an index to measure diversity or entropy in ecosystems. For plants in Germany as well as Norway, the S. magellanicum-associated communities show a significant higher diversity than S. fallax-associated plant communities. In contrast, for the eubacterial communities associated with S. fallax, a higher diversity was observed than for S. magellanicum. Again, this could be detected for the German as well as Norwegian sites. In contrast to S. magellanicum, S. fallax grows in weakly acid, some more mesotrophic situations. Higher content of nutrients and less acidity could explain the higher diversity of bacteria in this microenvironment. Natural habitats below pH 5 are considered extreme acidic environments, which are inhabited by acidophilic microorganisms. Acidophiles have evolved a variety of specialized mechanisms to maintain their internal cellular pH at a constant level (usually 7.2), and therefore they are highly specialized bacteria.

In conclusion, due to their phylogenetical age and their interesting morphology and ecology, Sphagnum plantlets are interesting objects to study microbial ecology. Here, we could demonstrate a high specificity and biodiversity for Sphagnum-associated communities.