Impact of biotic and a-biotic parameters on structure and function of microbial communities living on sclerotia of the soil-borne pathogenic fungus Rhizoctonia solani.

The plant pathogen Rhizoctonia solani is very difficult to control due to its persistent, long-living sclerotial structures in soil. Sclerotia are the main source of infection for Rhizoctonia diseases, which cause high yield losses on a broad host range world-wide. Little is known about micro-organisms associated with sclerotia in soil. Therefore, microbial communities of greenhouse and field incubated Rhizoctonia sclerotia were analysed by a multiphasic approach. Using microbial fingerprints performed by PCR-SSCP, sclerotia-associated bacterial communities showed a high diversity, whereas only a few fungi could be detected. Statistical analysis of fingerprints revealed the influence of soil types, incubation conditions (greenhouse, field), and incubation time (5 and 12 weeks) on the bacterial as well as fungal community. No significant differences were found for the microbial community associated with different Rhizoctonia anastomosis sub-groups (AG 1-IB and AG 1-IC). Rhizoctonia sclerotia are an interesting bio-resource: high proportions of fungal cell-wall degrading isolates as well as those with antagonistic activity towards R. solani were found. While a fraction of 28.4% of sclerotia-associated bacteria (=40 isolates) with antagonistic properties was determined, only 4.4% (=6 isolates) of the fungal isolates were antagonistic. We identified strong antagonists of the genera Bacillus, Enterobacter, Pseudomonas, and Stenotrophomonas, which can be used as biological control agents incorporated in soil or applied to Rhizoctonia host plants.

Introduction

Rhizoctonia solani Kühn (teleomorph Thanatephorus cucumeris [Frank] Donk) is a widely distributed soil-borne plant pathogen responsible for economic yield losses on a high diversity of crops (Sneh et al., 1996; Kiewnick et al., 2001; Büttner et al., 2003). The life cycle of Rhizoctonia includes parasitic and saprophytic periods as well as sexual and asexual reproduction. In soil, R. solani exists as vegetative mycelium or sclerotia, which are asexually produced. Sclerotia are known as the major source of infection for Rhizoctonia diseases (Anderson, 1982). The small (1–3 mm diameter), irregular-shaped brown structures consist of dense masses of monilioid cells. Due to this compact structure and the presence of melanin in all cell walls, sclerotia of R. solani are protected against various environmental conditions (Sneh et al., 1996). In addition, sclerotia of R. solani exude liquid brown droplets that are complex mixtures composed of phenolics, carboxylic acids, carbohydrates, fatty acids, and amino acids, which contributed to their antifungal and phytotoxic activities (Aliferis and Jabaji, 2010). R. solani is characterised by a high intra-specific diversity and comprises both multinucleate and binucleate species, which are further taxonomically divided according to their hyphal anastomosis into anastomosis groups (AGs; Anderson, 1982; Carling et al., 2002; Grosch et al., 2007). Both factors – persistent sclerotia and high diversity – contribute to the fact that Rhizoctonia diseases are difficult to control.

In the past, Rhizoctonia diseases were managed by soil fumigation using the ozone-depleting chemical methyl bromide, which was completely banned in 2006. Today, available fungicides to control Rhizoctonia diseases are limited, and fungicide spraying may prevent symptoms but does not kill the survival structures. Furthermore, other Rhizoctonia diseases like black scurf on potato are managed by crop rotation to reduce the pathogen inoculums in soil. However, Rhizoctonia sp. can use volunteers or weeds to build-up inoculums that can serve as a “green bridge” to the next crop (Keijer et al., 1997). These backgrounds together with increased public concerns about environmental problems and human health in industrial agriculture (Horrigan et al., 2002), emphasize the need for biological control strategies against Rhizoctonia.

The use of naturally occurring antagonistic microorganisms to control soil-borne pathogens offers sustainable alternatives to chemical pesticides (Lugtenberg and Kamilova, 2009; Berg, 2009). In biocontrol studies, introduced fungal antagonists like Trichoderma harzianum and Gliocladium virens successfully suppressed R. solani on several hosts (Elad et al., 1980; Grosch et al., 2006; Lahlali and Hijri, 2010). In addition, bacterial antagonists like Bacillus subtilis, Pseudomonas fluorescens, Pseudomonas jessenii, Serratia plymuthica and Lysobacter capsici (Chernin et al., 1995; Scherwinski et al., 2008; Postma et al., 2010) have shown potential to control R. solani. However, the majority of reported strains were only able to antagonise hyphae of the pathogen and not the survival structures (Marcello et al., 2010). Only two specific mycoparasites of Rhizoctonia sclerotia were reported: Verticillium biguttatum (Jager et al., 1991) and Trichoderma virens (Liu et al., 2010). Therefore, analysis of sclerotia-associated microorganisms is not only interesting to understand fungal ecology; results can be used to develop new biocontrol strategies. During their growth, sclerotia accumulate relatively high concentrations of carbohydrates, fats, and proteins, which provide good nutritional niches for microorganisms in soil (Willetts, 1971). In the mycospheres of different basidiomycetous fungi, species-specific bacterial ‘fungiphiles’ were detected (Warmink et al., 2009), and many fungal structures are colonised by bacteria (rev. in Grube and Berg, 2009). We hypothesize that sclerotia can be a source for adapted antagonistic microorganisms with negative effect on longevity of sclerotia in soil.

Therefore, the objective of our study was to analyse sclerotia-associated microbial communities in detail. In a cultivation-independent approach, we characterised the diversity of bacterial and fungal communities on sclerotia depending on: (i) different soil types, (ii) different incubation conditions in greenhouse and field trials, (iii) under different incubation times, and (iv) on different anastomosis groups of the fungus. Furthermore, by a cultivation-dependent approach, we analysed: (i) the antagonistic potential of bacterial and fungal isolates towards R. solani and (ii) the antagonistic mechanisms of the sclerotia-associated bacteria.

Materials and methods

Incubation of R. solani sclerotia in soil

For formation of sclerotia the R. solani isolates of different anastomosis sub-groups, AG 1-IB; and, AG 1-IC were cultivated for three weeks on potato dextrose agar (PDA, VWR International GmbH, Berlin, Germany) at 20 °C. Fifteen sclerotia of similar size were wrapped into nylon fabric parcels and 5 parcels were buried in 500 ml pots (10 cm × 10 cm × 10 cm) at 5 cm depth. The pots were filled with different soil types obtained from bulk soil (10 cm) of various locations (Table 1). In these pots, the sclerotia were incubated for 5 weeks (experiment 1) and for 12 weeks (experiment 2) in the dark at 15 °C/10 °C (16 h/8 h, day/night). A third experiment was carried out by incubation of sclerotia in the various soil types under field conditions from end of November 2006 to beginning of April 2007. In this case the pots were buried in top soil in Großbeeren (latitude: 52°N, 13°E). The average temperature during the incubation time was 3.9 °C (minimum −3.1 °C, maximum 14.8 °C) in 5 cm bulk soil. In all experiments each treatment included five replicates with 15 sclerotia each. After incubation, sclerotia were unwrapped under semi-sterile conditions and either used for cultivation of associated fungi and bacteria or for DNA-extraction. Samples were characterised based on soil type (S1–S4), cultivation conditions (F, field trial; GH, green house trial), anastomosis sub-group of R. solani AG 1-IB or AG 1-IC (B or C), sample repetition (1–5), and incubation time (5 weeks or 22 weeks = *); e.g. S1_F_B_1.

Table 1

Soil characteristics.

Soil number	Location	Soil type	pH	Nta (mg/100 g)	Ctb (%)	P lact.c (mg/100 g)	Po lact.c (mg/100 g)
Soil 1 (S1)	Golzow	Alluvial loam	7.6	151.0	1.81	46.0	24.4
Soil 2 (S2)	Plattling	Sandy loam	7.1	151.0	1.18	30.0	38.0
Soil 3 (S3)	Großbeeren	Diluvial sand	6.6	70.5	0.79	22.5	17.3
Soil 4 (S4)	Großbeeren	Diluvial sand	3.6	229.0	2.83	8.8	8.9

Nt – total nitrogen.

Ct – total carbon content.

Phosphorous (P) and potassium (Po) were measured using the lactat method according to VDLUFA-Methodenbuch (1991).

Structural analysis of sclerotia-associated microorganisms by Single Strand Conformational Polymorphism Analysis (SSCP)

DNA of R. solani sclerotia-associated bacterial and fungal cell consortia was extracted using the FastDNA Spin Kit for Soil (BIO 101, Carlsbad, USA) according to the manufacturer's protocol. Fingerprints of bacterial and fungal communities were obtained by SSCP analysis carried out as described by Schwieger and Tebbe (1998). Briefly, bacterial and fungal gene sequences were amplified by using the primers Unibac-II-515f (5′-GTG CCA GCA GCC GC-3′) and Unibac-II-927rP (5′-CCC GTC AAT TYM TTT GAG TT-3′) primers or ITS1F (5′-CTT GGT CAT TTA GAG GAA GTA A-3′) and ITS2R (5′-GCT GCG TTC TTC ATC GAT GC-3′) primers. The amplicons were separated by using the TGGE Maxi system (Biometra, Göttingen, Germany) at 400 V and 26 °C. Silver staining was used to detect DNA in SSCP gels (Bassam et al., 1991). The band patterns representing the microbial community composition of the different treatments were hierarchically clustered to produce dendrograms using the unweighted pair group method with arithmetic averages (UPGMA). The actual distances calculated and the distances shown on the dendrogram were compared by Pearson product-moment correlation (cophenic correlation) to ensure that the dendrogram gave an accurate representation of the data.

Isolation of sclerotia-associated microorganisms

For isolation of bacterial strains, sclerotia of each replicate were cultivated in nutrient broth (NB, VWR International GmbH, Germany) for 24 h at 28 °C. Three aliquots (500 μl) were taken from each generated culture suspension and further plated onto nutrient agar (NB amended with 15 g l−1 agar to) to isolate associated bacteria. Plates were incubated for five days at 20 °C. Bacteria visually distinguishable based on colony morphology were selected from the plates and transferred to nutrient agar. Colonisation of sclerotia with fungi were investigated by incubation of sclerotia on water agar (1.2%) amended with 50 μg ml−1 streptomycin and 50 μg ml−1 penicillin. The agar plates were incubated in the dark at 20 °C. After 1–3 days, hyphal tips from colonies were isolated and transferred to PDA. Cultures were stored at 10 °C until further use.

In vitro screening for antifungal and hydrolytic activity and production of secondary metabolites

Bacterial isolates were tested in a dual-culture assay with R. solani AG 1-IB on Waksman agar containing per liter 5 g of peptone (Carl-Roth, Karlsruhe, Germany), 10 g of glucose (Carl-Roth), 5 g of NaCl (Carl-Roth), 3 g of yeast extract (Carl-Roth) and 20 g of agar (Carl-Roth). Zones of inhibition were measured and evaluated according to Berg et al. (2002).

Chitinase activity (beta-1,4-glucosamine polymer degradation) was tested in chitin minimal medium by the method of Chernin et al. (1995). Clearing zones were detected five days after incubation at 20 °C. Beta-glucanase activity was tested by using chromogenic (azurine-dyed, cross-linked [AZCL]) substrates (Megazyme, Bray, Ireland). Formation of blue halos was recorded until 5 days after incubation. Protease activity (casein degradation) was determined from clearing zones in skim milk agar (50 ml of sterilized skim milk mixed at 55 °C with 50 ml of 1/5 tryptic soy agar and 4% agar) after 5 days of incubation at 20 °C.

Antibiosis against R. solani by the bacterial strains was assayed on Waksman agar plates (15 ml, pH 7–8, 9 cm diameter) containing 5 ml of sterile culture filtrate (24-h culture, nutrient broth II (Sifin), sterile filtration using 0.2 μm filters). An actively growing R. solani mycelial disk from the margin of a 7 day old PDA-culture (0.8 cm diameter) was placed in the centre of a Waksman agar plate. The growth of R. solani inoculant was compared to its growth on a control plate using sterile water instead of culture filtrate.

The ability to produce indole-3-acetic acid (IAA) was tested for selected bacterial strains in a plate test according to Sawar and Kremer (1995). The test was modified by using an Immobilion™-P Transfer Membrane (Millipore, Billerica, USA) instead of a Nitrocellulose membrane.

Identification of antagonistic isolates

Initial genotypic characterisation of selected bacterial antagonists ARDRA (amplified rDNA restriction analysis) was performed. For amplification of bacterial 16S RNA genes Eub 1F primer (5′-GAG TTT GAT CCT GGC TCA G-3′) and 1492R (5′-TAC GGY TAC CTT GTT ACG ACT T-3′) were used. Thereby obtained, 1400 bp long PCR fragments were digested at 37 °C over a period of at least 3 h, using 0.25 μl restriction endonuclease HhaI (New England Biolabs), 2 μl of NEB-4 puffer (New England Biolabs), 0.2 μl of BSA (10 mg ml−1) (New England Biolabs), 2.55 μl of water and 15 μl of clean PCR product. A 10 μl aliquot of restriction product was separated by gel electrophoresis on 2.0% agarose (Carl-Roth) gels in 0.5× Tris–borate–EDTA buffer for 3 h. Representative antagonists of selected and characterised antagonistic bacteria were identified by partial 16S rRNA gene sequencing and subsequent alignment using the BLAST algorithm (Altschul et al., 1997).

Nucleotide sequence accession numbers

Sequence accession numbers for sequences of the partial 16S rRNA gene (FR752798 to FR752814 and FR752817 to FR752818) of the bacterial strains were submitted to the NCBI nucleotide sequence database.

Statistical analysis

To compare SSCP fingerprints of the microbial communities, a computer-assisted cluster analysis was carried out using the GelCompar II® software (Applied Math, Kortrijk, Belgium). Computer-assisted pattern analyses of molecular fingerprints of the communities resulted in correlation matrices, which were subjected to significance test by applying permutation tests with 104 random permutations of sample elements (Kropf et al., 2004). The relationships between samples taken from different soil types, at different incubation times and conditions as well as from anastomosis groups were investigated.

Results

Molecular fingerprinting of sclerotia-associated bacterial and fungal communities

The cultivation-independent analysis of sclerotia-associated microbial populations was performed to get insights into the impact of environmental factors on their structure. In general, SSCP (Single Strand Conformational Polymorphism Analysis) fingerprints of bacterial and fungal sclerotia-associated communities revealed differences in abundances and diversity. The bacterial community was much more diverse than the fungal community (data not shown). In fungal community fingerprints with less than 10 different bands were observed; especially very few bands were found for the fungal community obtained from sclerotia incubated in the field. Differences between field and greenhouse incubated communities were not found in this extent for the bacterial communities; the bacterial population was always highly diverse and consisted of numerous populations.

On the basis of microbial fingerprints performed by PCR-SSCP, dendrograms and statistical analyses were performed to analyse the impact of biotic and a-biotic parameters on the structure of microbial communities living on sclerotia. According to these analyses, bacterial as well as fungal fingerprints revealed the influence of soil type, incubation conditions (greenhouse and field) and incubation time (5 and 12 weeks). All factors affected significantly the microbial community (P ≤ 0.05). In contrast, no significant differences were found for the structure of microbial communities associated with different anastomosis AG 1 sub-groups (IB and IC) of R. solani (Figs. 1 and 2 and Table 2).

Table 2

Statistical analysis of investigated parameters. Microbial fingerprints obtained by PCR-SSCP were analysed using permutation test (Kropf et al., 2004) P ≤ 0.05.

Parameter	Compared data set	P value for
		Bacterial communities	Fungal communities
Soil type	S1 + S2 + S3 + S4a	0.0000	0.0001
Incubation condition	GH + Fb	0.0000	0.0001
Incubation time	5 + 12 weeks	0.0000	0.0000
Anastomosis group	IB + ICc	0.6235	0.1277

S1 Golzow, S2 Plattling, S3 Großbeeren, S4 Großbeeren (see Table 1).

Greenhouse (GH) trial or field (F) trial.

AG 1-IB or AG 1-IC.

Antagonistic potential of sclerotia-associated microorganisms

Bacterial as well as fungal strains were isolated from the incubated Rhizoctonia sclerotia. While the bacterial isolates showed a high diversity based on colony morphology, a limited morphologically visual diversity was observed for fungal isolates associated with sclerotia (<7 genera, like Mucor spp., Fusarium spp., Trichoderma spp., Cylindrocarpon spp., and Penicillium spp. isolates with sterile mycelium). Altogether, 141 bacterial and 270 fungal isolates were obtained and analysed for in vitro activity towards R. solani. A proportion of 28.4% (40 isolates) from the bacterial isolates and of 4.4% (6 isolates) of the fungal isolates was able to antagonise R. solani. In addition, to estimate lytic properties on the fungal cell wall, hydrolytic enzymes were investigated: 82.4% of selected antagonistic isolates showed protease activity, 57.6% had β-1,4-glucanase activity and 35.3% showed chitinases. Furthermore, a high proportion of isolates was able to produce antifungal metabolites; 87.5% of all tested strains were able to produce such compounds in vitro. Using these data to compare the activity of sclerotia-associated bacteria at the two sampling times, we found that the antagonistic population of sclerotia-associated bacteria increased after a prolonged period of incubation (Fig. 3). The proportion of bacteria producing all three hydrolytic enzymes was more than doubled (increased from 14.3% to 31.6%).

Fig. 3

Summary of antagonistic mechanisms of the sclerotia-associated bacteria to produce proteases, chitinases, beta-1,4-glucanases or the plant growth hormone indole-3-acetic acid (IAA) after different durations of incubation (grey 5 weeks, black 12 weeks).

For further analysis of the antagonistic microorganisms, only bacteria were considered: the antagonistic fungal isolates were characterised by a slow growth, weak antagonism and none mycoparasitic activity, and are therefore not appropriate as biocontrol agents. To select bacterial antagonists, the antifungal as well as fungal cell wall-degrading properties were considered (Table 3).

Table 3

Results obtained from physiological and molecular characterisation of selected antagonistic sclerotia-associated bacteria. ARDRA, amplified rDNA restriction analysis; n.d., not determined.

Isolate	ARDRA group	Closest database match (16S rRNA gene similarity)	Accession no.	SIb	Antagonistic activitya	IAAc	Protease	Chitinase	Glucanase	Antibiosisa
B104	1				+	+	+	+	–	+++
B163	1	Pseudomonas fluorescens	GU198122.1	99	++	n.d.	n.d.	n.d.	n.d.	++
B44	2	Pseudomonas fluorescens	HQ317182.1	100	+	n.d.	n.d.	n.d.	n.d.	n.d.
B106	3	Bacillus cereus	HM771661.1	100	++	–	+	+	+	++
B156	3	Bacillus thuringiensis	HQ256544.1	99	+++	–	+	+	–	+++
B32	4	Pseudomonas chlororaphis	HM241942.1	97	+++	+	+	–	–	++
B13	4				+	+	+	–	–	+
B50	4				+	+	+	–	–	++
B111	4				++	+	+	–	–	++
B102	5	Enterobacter aerogenes	GQ165811.1	99	+	+	–	–	+	+
B2	5	Enterobacter ludwigii	AM184285.1	98	++	n.d.	n.d.	n.d.	n.d.	++
B3	5				++	n.d.	n.d.	n.d.	n.d.	++
B31a	6	Stenotrophomonas maltophilia	HQ166115.1	98	++	–	+	+	–	+++
B31b	6	Stenotrophomonas maltophilia	HQ407233.1	99	+	+	+	+	–	++
B168	6				+	+	+	+	–	+++
B47	7	Bacillus subtilis	HQ123476.1	100	+++	+	+	–	+	+++
B48	7	Bacillus subtilis	EU627169.1	100	+++	+	+	–	+	+++
B158	8				+	+	+	–	–	++
B142	8	Pseudomonas sp.	FJ225200.1	99	++	–	+	+	+	++
B167	8				+++	–	–	+	–	+
B119	8				+	+	+	+	+	+
B164	9				+	–	+	–	–	++
B136	9	Pseudomonas putida	HQ658765.1	100	+	+	–	–	–	++
B103	9				+	+	–	–	–	++
B134	9				+	+	–	–	–	n.d.
B21	10	Bacillus cereus	HM771661.1	99	+	–	+	+	–	++
B23	10	Bacillus cereus	GQ462533.1	98	++	–	+	+	+	++
B40	10				++	–	+	+	+	++
B182	10	Bacillus cereus	HQ407233.1	100	+++	–	+	+	+	+++
B107	11				+	+	–	–	–	+
B121	11				+	+	+	+	+	++
B147	11	Bacillus thuringiensis	HQ256544.1	99	+	+	+	+	–	++
B169	11				+	n.d.	+	n.d.	–	++
B178	11	Bacillus thuringiensis	DQ286343.1	98	++	–	+	+	+	+++

+ 0–5 mm, ++ 5–10 mm, +++ >10 mm radius of zone of inhibition in dual culture assay or mycelial growth area of R. solani in antibiosis tests, –no suppression.

Plant growth hormone indole-3-acetic acid (IAA).

Similarity index.

Identification and characterisation of Rhizoctonia antagonists

For the identification of bacterial antagonists, the isolates were firstly grouped by ARDRA (Table 3). From each ARDRA group, 34 isolates were randomly selected and identified by partial 16S rRNA gene sequencing. Sclerotia-associated bacteria belong to four phylogenetic groups: Enterobactericeae, Pseudomonadaceae, Xanthomonadaceae, and Firmicutes. Altogether, 18 different strains were identified, wich resulted into nine species: Bacillus cereus, Bacillus subtilis, Bacillus thuringiensis, Enterobacter aerogenes, Enterobacter ludwigii, Pseudomonas chlororaphis, Pseudomonas fluorescens, Pseudomonas putida, and Stenotrophomonas maltophilia. Furthermore, selected isolates were characterised regarding their antagonistic and fungal cell wall-degrading activity, IAA production and antibiosis capacity (Table 3). All strains, which were selected according to their antagonistic activity, produced antifungal antibiotics. In addition, a high proportion was characterised by proteolytic, chitinolytic and glucanolytic properties. Interestingly many strains produced the phytohormone indole-3-acetic acid (IAA), which is not only known to influence growth of plants; it is also a signal molecule and growth factor for fungi (Prusty et al., 2004).

Discussion

While sclerotia formation and structure of diverse plant pathogenic fungi are well-investigated, only little is known about sclerotia-inhabiting microorganisms. In this study, we found that sclerotia of the plant pathogenic fungus R. solani were colonised by various specific bacterial and fungal isolates. Moreover, especially between the bacterial inhabitants, we found a high proportion, approx. 1/3 of the bacteria, which expressed antagonistic traits against R. solani. In addition, a high proportion of sclerotia-associated micro-organisms posses fungal cell wall-degrading abilities. These facts show that sclerotia-associated bacteria are a potential bio-resource to select novel biocontrol agents against sclerotia-forming pathogens.

Our data showed that in all experiments sclerotia were colonised by similar microbial populations. However, several biotic and a-biotic factors were identified, which influenced the structure of the microbial communities. Interestingly, the structure of bacterial communities was influenced by the same environmental factors than the fungal community. The latter includes soil types, incubation conditions, and incubation time. Soil type, especially pH, is an important driver of the microbial communities in soil (Rousk et al., 2010). We investigated three different soil types (alluvial and sandy loam, diluvial sand), which were characterised by extremely different pH values (3.6–7.6). They can explain the high impact of this factor on the structure of sclerotia-associated microorganisms. Furthermore, incubation conditions in greenhouse and field differed strongly in temperature; the difference was more than 20 °C. This could be a key factor here. The time of incubation was also an important impact factor. The microbial fingerprints reflect the impact of these factors. We found that the proportion of associated bacteria producing hydrolytic enzymes was more than doubled after 12 in comparison to 5 weeks. No impact was found on the structure of microbial communities associated with different R. solani AG 1 sub-groups (IB and IC). However, the similarity in the ITS region of the isolates from each AG 1 sub-group is, although they can be distinguished (Grosch et al., 2007), very high, whereas isolates of R. solani AG 1 were clearly separated from R. solani AG 2-1, AG 4, and binucleate Rhizoctonia AG-Bb and AG-K (Toda et al., 2011). While the impact of plant species and cultivars on structure and function of plant-associated microbial communities is well-known (rev. in Berg and Smalla, 2009), specificity of fungi–bacteria interaction is only partly understood (Grube and Berg, 2009).

The composition of the fungal community was characterised by low diversity, which was found in cultivation-dependent and -independent experiments. Sclerotia-associated fungi could be influenced by antifungal metabolites produced by sclerotia themselves. Aliferis and Jabaji (2010) identified a high number of antifungal and phytotoxic compounds in the exudates of R. solani. Antibiotic properties are widespread amongst Rhizoctonia spp.; antibiotic compounds were also found in the long-living sclerotia of Rhizoctonia cerealis and Rhizoctonia oryzea (Burton and Coley-Smith, 1993). Furthermore, beside the presence of antibiotic compounds, melanin contributes to longevity and resistance to microbial lysis (Sneh et al., 1996).

In contrast to the results obtained by cultivation-independent methods, cultivation captures only a very low proportion of microorganisms (Schwieger and Tebbe, 1998). But the combination of both assays has advantages: functions can be assessed and isolates can be used for biotechnological applications. Till now, only cultivation-dependent studies about sclerotia-colonising bacteria were already published. For example, Gladders and Coley-Smith (1980) showed a high diversity of Bacillus on sclerotia of Rhizoctonia tuliparum. They also identified antibacterial pyrone antibiotics produced by the sclerotia and showed that isolates belonging to the Bacillus group and Serratia sp. were less sensitive against these antibiotics. In our study, it was not possible to calculate the number of bacterial and fungal microorganisms. We selected abundant and culturable microorganisms by incubation of sclerotia in NB or water agar. However, during the incubation of sclerotia in NB media the bacteria propagate already, so the counting of the CFU on agar media reflect not the number of bacteria on the sclerotia surface only the diversity. To assess bacterial abundances on sclerotia, FISH (fluorescence in situ hybridization) in combination with confocal laser scanning microscopy would be a useful method (Grube and Berg, 2009). We used the cultivation approach to isolate strains of Bacillus, Enterobacter, Pseudomonas and Stenotrophomonas with antagonistic properties against R. solani. All genera are well-known that they harbour potent antagonists against plant pathogens (Haas and Défago, 2005; Ryan et al., 2009). However, strain-specific antagonistic activity was reported (Berg et al., 2002; Zachow et al., 2010), and several of the isolated bacteria were highly active in vitro antagonists. Fungal isolates from sclerotia belonging mainly to the genera Mycor, Fusarium, Cylindrocarpon and Trichoderma. Fungi of these genera were also observed on R. solani sclerotia from potato tubers (Chand and Logan, 1984), and on sclerotia of R. tuliparum (Gladders and Coley-Smith, 1980). Our results support previous findings that the outer layers of sclerotia may be invaded by a number of soil fungi but we could not find specific mycoparasites of R. solani.

Interestingly, we found a high proportion of bacterial antagonists on sclerotia. In contrast to the rhizosphere of plants, which is a hotspot for antagonists (Berg et al., 2002) the number of antagonistic bacteria was much higher on sclerotia. In our study we could show, that fungal sclerotia are also an interesting reservoir for naturally occurring antagonists. The knowledge about sclerotia-associated microorganisms and molecular mechanisms involved in sclerotial lysis can support biological control strategies against fungi producing long-term survival structures.