OUP user menu

Humus bacteria of Norway spruce stands: plant growth promoting properties and birch, red fescue and alder colonizing capacity

S. Elo , L. Maunuksela , M. Salkinoja-Salonen , A. Smolander , K. Haahtela
DOI: http://dx.doi.org/10.1111/j.1574-6941.2000.tb00679.x 143-152 First published online: 1 February 2000

Abstract

We studied the potential of the humus layer of the Norway spruce stands to supply beneficial rhizobacteria to birch (Betula pendula), alder (Alnus incana) and fescue grass (Festuca rubra), representatives of pioneer vegetation after clear-cutting of the coniferous forest. Axenically grown seedlings of these species were inoculated with the acid spruce humus, pH 3.7–5.3. Actinorhizal propagules, capable of nodulating alder, were present in high density (103 g−1) in humus of long-term limed plots, whereas plots with nitrogen fertilization contained almost none (≤10 g−1). The genera most frequently found in the humus were Bacillus, Paenibacillus, Arthrobacter, Nocardia, Rhodococcus and Pseudomonas, independently of prior liming or fertilization of the plots. The taxa found in the seedling roots differed from that in humus by the prevalence of the Gram-negative genera Pseudomonas, Alcaligenes and Comamonas. Enrichment cultures of the roots on nitrogen-free media yielded Paenibacillus and Rhodococcus species. Nitrogen-fixing R. erythropolis and a novel Paenibacillus, closest by full sequence of 16S rDNA to P. durus, represented new classes of nitrogen-fixing rhizosphere bacteria. In addition, nitrogen-fixing R. fascians was found in the humus. The rhizoflora and humus contained high proportions of bacteria antagonistic towards plant pathogenic Rhizoctonia sp., Botrytis cinerea and Fusarium culmorum. The antagonistic isolates also commonly produced siderophores and/or cell wall degrading enzymes.

Keywords
  • Spruce humus rhizobacterium
  • Nitrogen fixation
  • Fungal antagonist
  • Alder nodulation

1 Introduction

Conifers dominate in the mature Finnish economically managed forests. The raw material demand of the pulp and paper industry has recently shifted towards hardwoods because the present environmental policy requires elemental chlorine-free or totally chlorine-free bleaching of wood pulp. Birch also grows more rapidly than spruce or pine in the boreal climate. For these reasons former conifer stands are often planted with birch after harvesting. Soils under birch (Betula pendula, B. pubescens) possess a more active and rich microflora than those under soft woods, Scots pine (Pinus sylvestris) and especially Norway spruce (Picea abies) [1].

Bacteria in the rhizosphere and rhizoplane of grasses and crop plants (e.g. [2, 3]) have been investigated but few studies have focused on root surface bacteria of forest trees. Interactions with the host by rhizosphere bacteria of Scots pine (Pinus sylvestris) were studied by Timonen et al. [4] and of Pinus radiata by Garbaye and Bowen [5], with a hybrid spruce host (Picea glauca×engelmannii) by Chanway et al. [6], and with Douglas fir (Pseudotsuga menziesii) by Axelrood et al. [7]. Soils under birch had higher Frankia nodulation capacities than soils under alder [8]. Root surface bacteria of the herbaceous plants growing on the forest floor may serve as a microbe pool in early phases of tree development. Plant growth promoting properties of boreal coniferous forest humus bacteria may be of importance when replanting the forest but have not been studied.

We studied the uptake of bacteria by axenic seedlings of silver birch (Betula pendula), fescue grass (Festuca rubra) and alder (Alnus incana) from the humus of Norway spruce stands without broad-leaved trees. We found that silver birch and fescue grass selected to their rhizoplanes bacteria capable of fixing nitrogen and bacteria with antagonistic properties to plant pathogenic fungi. The spruce humus also contained a high number of alder infective Frankia.

2 Materials and methods

2.1 Study sites and humus sampling

The Norway spruce (Picea abies L.) stands at Heinola (61°9′N/26°3′E), aged 42 years, and at Kerimäki (61°51′N/29°22′E), aged 60 years, in southern Finland, grew on mineral soils and were of the OMT type (Oxalis-Myrtillus) [9]. Fertilization histories of the four plots at each site are recorded in Table 1. There was no actinorhizal vegetation. The soil type was podzol and the humus type mor. Twenty-eight soil cores (diameter 5 cm) of the humus layer were collected in September 1992 and combined to give one composite sample per plot. This number of soil cores was sampled to even out the microbial diversity known to occur in natural soils. The samples were pooled, so no statistics were possible between the individual cores. Green plant material was removed, the samples were sieved through a 2.8-mm sieve and stored in polyethene bags overnight at 6°C. Selected humus characteristics are given in Table 1.

View this table:
1

Humus properties from the Norway spruce stands used as inoculum for seedlings of birch and fescue grass

Stand site and age (years)Treatment of the plotD.m. (%)Bulk density (g cm−3)Organic matter (% d.m.)pH (CaCl2)
Kerimäki (60)O540.3513.7
Ca530.3514.7
N500.2673.6
CaN500.3555.3
Heinola (42)O630.5204.2
Ca660.6165.3
N640.5213.7
CaN620.5224.7
The plots had a history of being limed (Ca), nitrogen fertilized (N), limed and fertilized (CaN) or untreated (O), described in [10]. The latest application of lime was 12–14 years, and that of N addition 3–5 years before the humus layer for this study was sampled. D.m.=dry matter.

2.2 Plant material and inoculation

Seeds of red fescue (Festuca rubra) were surface sterilized by soaking in 94% ethanol for 1 min, in 5% (w/v) sodium hypochlorite for 15 min and rinsed five times in sterile distilled water [11]. Seeds of silver birch (Betula pendula) were soaked in distilled water overnight, surface sterilized with 0.01% of Tween 80 in 30% H2O2 for 25 min, rinsed as above and allowed to germinate [12]. The seedlings were grown in tubes with glass beads as described in [12] for 2–4 days and then inoculated with 0.5 ml of humus suspension (20% w/v in phosphate buffered saline (PBS)).

2.3 Isolation and enumeration of bacteria from humus and plant roots

Nodulation of Alnus incana seedlings by Frankia was determined as described in [13]. Serial dilutions of humus suspension (5 ml, 10% w/v in half-strength Nutrient solution [14]) were pipetted into four replicate bottles (120 ml) with two 7-week-old alder seedlings growing in Nutrient solution in liquid cultures. Seven days later the seedlings were transferred to N-free half-strength Nutrient solution. The Nutrient solution was replaced once a week and root nodules were counted after 6 weeks. A nodulation unit was defined as a unit which can induce formation of a single nodule.

Humus suspension (20% w/v in PBS) was plated as serial dilutions on three replicate plates on Luria-Bertani [15], King's B [15], Malate [2], and Actinomycete isolation agar (Difco) with cycloheximide (75 mg l−1). Colony forming units were counted after 5 days incubation at 28°C. Roots, homogenized in PBS [16], were similarly plated on three replicate plates 7 weeks after inoculation. Pure cultures were prepared of all colony types. Nitrogen-fixing activity was scored from 0.2 ml of the humus and root suspensions (two replicates) in 22-ml serum bottles with 5 ml of N-free semisolid Malate medium [16] supplemented with 0.5% (w/v) glucose after 7 days incubation at 28°C. The acetylene reduction assay was performed as described in [2]. Serial dilutions from the positive bottles were plated on Luria agar and on Malate agar, prepared to pure cultures and tested for acetylene reduction. Pure cultures were stored at −70°C in N-free medium with 20% glycerol.

2.4 Identification of bacterial isolates

The isolates were Gram-stained using Hucker's modification [15], and analyzed by API 20 E, API 20 NE and API 50 CHB strip tests (BioMérieux SA, Marcy l'Etoile, France), and by whole cell fatty acid composition using the MIDI Microbial Identification System, library version 3.9 (Microbial ID Inc., Newark, DE, USA), as described in [17].

For rep-PCR the cells were suspended in TE buffer (10 mM Tris, pH 8.0, 1 mM EDTA), heated in a boiling water bath for 5 min and centrifuged. The supernatant was used as template and REP1R-I, REP2-I, ERIC1R, ERIC2 [18] and BOXA1R [19] as primers. Reaction conditions for the REP and ERIC primer pairs were as described in [18]. Annealing temperature was 52°C for the BOXA1R primer. Dynazyme DNA polymerase (Finnzymes, Espoo, Finland) was used. PCR products were analyzed by gel electrophoresis on 1.2% agarose gels, stained with ethidium bromide and photographed.

Nucleic acids were extracted as described in [20]. The 16S rRNA gene was amplified in two steps using primer pairs pA-pE′ and pD-pH′[21]. The sequencing was done by the solid phase method [22] using an automated ALF DNA sequencer (Pharmacia, Uppsala, Sweden), and a Dye Terminator sequencing kit and an ABI Prism 377XL DNA sequencer (Perkin Elmer, Foster City, CA, USA). The sequences were assembled using the xgap program in the Staden Package. The search for homologous sequences was done with the FASTA program of the Wisconsin GCG package. The 16S rRNA gene sequences were deposited in the EMBL database under the accession numbers AJ011321AJ011333.

2.5 Antagonistic activity

Rhizoctonia solani HK1, Botrytis cinerea HK2, Fusarium culmorum HK3 (Department of Plant Biology, University of Helsinki) and uninucleate Rhizoctonia sp. 263 and 264 isolated from nursery-grown conifer seedlings [23] were used as indicator strains for antagonistic bacteria. Fluorescent pseudomonads and streptomycete-like isolates were tested on dual-growth plates on potato dextrose agar (PDA) (Biokar Diagnostics, Beauvais, France) (pH 5.2) and bacilli on Malate-yeast extract-glucose-sucrose agar (MYGS). MYGS was modified from solid Malate medium [2] with 0.17% (w/v) each of malic acid, glucose and sucrose, 0.1% (w/v) of yeast extract, 5 μg of biotin l−1 and 10 μg of p-aminobenzoic acid l−1, pH 6.8. A piece of PDA culture of the fungus or streptomycete-like culture was placed on the test plate. Pseudomonads and bacilli were streaked on the test plates perpendicular to the fungus. Plates were incubated at room temperature until the fungi had grown over the control plates without bacteria. Antifungal activity was recorded as the width of the zone of growth inhibition between the fungus and the organism tested. A biocontrol agent, Streptomyces griseoviridis SE1 [24], was used as a reference strain.

2.6 Other substances beneficial for plant growth

Production of hydrogen cyanide (HCN) from glycine (4.4 g l−1) was read after 72 h at 28°C with King's B or MYGS medium as a base [25] using picric acid/Na2CO3 filter paper fixed to the underside of the Petri dish lids [26]. Chitinolytic activity was tested by a modification of the method of Chernin et al. [27] using synthetic agar medium (SM) with 0.2% (w/v) colloidal chitin. The plates were incubated at 28°C and zones of chitin clearing recorded after 7–14 days. Hydrolysis of pectin was detected by the method of Chatterjee and Starr [28]. Carboxymethylcellulase (CMCase) activity was detected according to [29]. Siderophore production was measured with the chrome azurol S (CAS) agar and the microtiter method using CAS assay solution described in [30] except that 5-sulfosalicylic acid was not used. Indoleacetic acid (IAA) production was detected by a rapid in situ assay [31]. Fluorescent pseudomonads were grown on King's B with 5 mM l-tryptophan and the other isolates on MYGS medium with 5 mM l-tryptophan. For simplicity, we use the term ‘plant growth promoting substances’ in the present paper also to indicate these activities (HCN, chitinase, CMCase, siderophore).

3 Results

3.1 Colonization of axenic seedlings of birch and fescue grass by humus bacteria from spruce stands

To answer the question whether coniferous forest humus contained bacteria capable of colonizing birch or grass roots, we exposed axenic seedlings of silver birch (Betula pendula) and fescue grass (Festuca rubra) to humus extracts obtained from spruce stands at Heinola and at Kerimäki.

The resident bacterial communities in the humus and that of the birch and fescue grass roots after 7 weeks of exposure to the humus were analyzed. Identifications of the aerobic heterotrophic isolates from the combined spruce humus samples and from the exposed seedlings are compiled in Fig. 1. Four differently fertilized plots were analyzed from two different stands (Table 1). The same species were obtained from the humus layers of the eight plots. The results were therefore combined in Fig. 1. In the humus 81% of the isolates (n=92) were Gram-positive, the genera most frequently found were identified as species of Bacillus, Paenibacillus, Brevibacillus and those of the actinobacteria Arthrobacter, Nocardia, Rhodococcus and Streptomyces. The abundance displayed in the column height reflects the prevalence of the different taxa among the isolates. Streptomyces represented less than 5% of the colonies but all colonies (n=23 in humus, n=1 in birch roots) were selected in the study because of their expected antagonistic properties. Streptomyces included the species S. albus, S. californicus, S. exfoliatus, S. griseoflavus, S. halstedii, S. lavendulae, S. rochei and S. violaceusniger. Of the Gram-negative taxa in the humus, Pseudomonas (11%) was the most prevalent. Distribution of the taxa found on the roots of birch and fescue grass differed from that found in the humus used to expose the seedlings of these plants. Gram-negatives constituted 73 and 58%, respectively, of the isolates from birch (n=64) and fescue grass (n=59). Pseudomonas species were the prevalent taxon in both birch and fescue grass (49 and 35%, respectively), together with Alcaligenes and Comamonas (17 and 8 isolates). None of these taxa was obtained from the humus. The plant root actinobacterial genera differed from those found in the humus; Arthrobacter was less prevalent and Nocardia more prevalent in the roots than in the humus. Bacillus and Paenibacillus species (15% of the isolates) were found in the fescue grass seedlings but not in the birch. In general, bacilli (Bacillus, Brevibacillus or Paenibacillus) were less prevalent in the roots than in the humus. The roots of birch and fescue grass seedlings did not select Streptomyces from the humus. It is possible that the exposure time (49 days) of the seedlings was too short for these slow-growing bacteria to colonize the roots.

1

Heterotrophic aerobic bacteria found in the humus layer of Norway spruce stands and in the roots of birch (Betula pendula) or fescue grass (Festuca rubra) after exposing the axenic seedlings for 7 weeks to the humus. The spruce humus contained 106–107 cfu of aerobic heterotrophs g−1 of organic matter and the exposed seedlings 104–106 cfu/seedling. The figure shows the distribution of 325 isolates representing 49% of all isolates, identified using FAME analysis (SIM>0.1) and API strip tests (probability≥80%). From the humus, 241 isolates were obtained, 131 of which were tentatively identified. Of the 209 birch seedling isolates 90 were identified and of the 217 fescue grass isolates 104. Arthrobacter included the species A. agilis, A. aurescens, A. citreus, A. ilicis and A. viscosus. Nocardia included N. asteroides and N. globerula. Rhodococcus included R. erythropolis, R. fascians and R. globerulus. Bacillus included B. circulans, B. megaterium, B. mycoides, and B. sphaericus. Paenibacillus included P. macerans, P. pabuli and P. polymyxa. Brevibacillus included B. brevis and B. laterosporus. ‘Other Gram-positive’ included Cellulomonas turbata, Clavibacter michiganensis, Kocuria kristinae, Kurthia gibsonii, Microbacterium imperiale, Micrococcus luteus and Oerskovia xanthineolytica. ‘Enterobacteria’ included Citrobacter sp., Kluyvera ascorbata, K. cryocrescens, Serratia proteamaculans, Yersinia fredereksenii and Y. pseudotuberculosis ssp. pseudotuberculosis. Pseudomonas included Ps. chlororaphis, Ps. fluorescens and Ps. putida. Comamonas included C. acidovorans. Alcaligenes included Alc. piechaudii, Alc. xylosoxidans ssp. xylosoxidans and Alc. xylosoxidans ssp. denitrificans. ‘Other Gram-negative’ included Acidovorax avenae, Burkholderia cepacia, Chryseobacterium indologenes, Janthinobacterium lividum, Ochrobactrum anthropi, Sphingomonas paucimobilis, Stenotrophomonas maltophilia, Variovorax paradoxus and Xanthobacter agilis.

In summary, our data show that the seedlings positively selected certain genera from the humus inoculum, notably the Gram-negatives Pseudomonas, Comamonas and Alcaligenes and the actinobacteria Nocardia and Rhodococcus. The roots of fescue grass attracted spore-forming bacilli.

3.2 Nitrogen-fixing taxa in the humus of spruce stands and in seedlings exposed to it

When no nitrogen-fixing isolates were found from humus by direct plating, the effort was extended by preparing 12 enrichment cultures resulting in 813 isolates, 263 from the humus of the nonfertilized control plots (O), 294 from birch and 256 from fescue grass seedlings inoculated with the control plot humus. Forty-one isolates (5%) were active in the acetylene reduction assay, 32 from humus, six from birch and three from fescue grass. To see how many parallel isolates the enrichment procedure had generated, all 41 isolates were analyzed by REP-PCR (17–27 fragments, 200–4000 bp). Nine unique fingerprint patterns were found and a further four patterns were shared by 32 isolates. These 32 were analyzed with ERIC and BOX primer sets. It was found that isolates producing similar patterns in REP-PCR were similar also with BOX (14–19 fragments, 200–4000 bp) or with ERIC (12–15 fragments, 200–4000 bp) PCR. This result sets the total number of different pattern groups at 13. Isolates sharing patterns also shared the origin (humus, birch or fescue grass) in all cases.

Representatives of each of the 13 different pattern groups were subjected to sequence analysis of 16S rRNA genes. The sequences of 16S rDNA were compared to those in the EMBL database. The same isolates were also analyzed by FAME. Table 2 shows the identities as shown by 16S rDNA sequencing. Similar species assignments were obtained only for Rhodococcus fascians and R. erythropolis by the two methods. One explanation for the poor match between the two methods may lie in the fact that most of the isolates grew only slowly in TSA, the standard medium used for identification by the commercial library.

View this table:
2

Identification of nitrogen-fixing isolates by 16S rDNA sequencing

IsolateSite of isolationSequenced 16S rDNA size (bp)Accession numberClosest EMBL library strainsSimilarity (%)
KM6Humus/Kerimäki840AJ011329Rhodococcus fascians D188100.0
KM8Humus/Kerimäki1455AJ011321Paenibacillus durus DSM 1735T95.0
KM16Humus/Kerimäki1506AJ011333Yersinia ruckeri ATCC 29473T96.2
HM26Humus/Heinola1530AJ011326Paenibacillus durus DSM 1735T95.9
HM29Humus/Heinola794AJ011331Pseudomonas azotoformans IAM 160398.0
Pseudomonas synxantha IAM 1235698.0
Pseudomonas fluorescens IAM 1202297.7
HM31Humus/Heinola1531AJ011327Paenibacillus durus DSM 1735T96.0
HM35Humus/Heinola1462AJ011330Phyllobacterium myrsinacearum IAM 13584T99.2
Phyllobacterium rubiacearum IAM 13587T98.9
KK19Birch/Kerimäki1540AJ011322Paenibacillus durus DSM 1735T96.0
KK20Birch/Kerimäki1531AJ011323Paenibacillus durus DSM 1735T96.0
HK40Birch/Heinola1518AJ011332Stenotrophomonas maltophilia LMG 1108797.9
Xanthomonas vasicola LMG 73697.6
KN24Fescue/Kerimäki1450AJ011324Paenibacillus durus DSM 1735T95.9
KN25Fescue/Kerimäki1450AJ011325Paenibacillus durus DSM 1735T96.0
HN41Fescue/Heinola832AJ011328Rhodococcus erythropolis ATCC 19566100.0
T=Type strain.

In summary, the nitrogen-fixing isolates were identified to the genera Rhodococcus, Paenibacillus, Pseudomonas and Phyllobacterium. In addition, isolates of two unknown γ-proteobacterial genera were recognized on the basis of their 16S rDNA sequences. It is remarkable that out of the 13 independent genotypes tested, seven were identified as Paenibacillus sp., with similarities of 95.0–96.0% to P. durus of the almost full 16S rDNA sequence. This taxon was found in the humus and in the roots of both the fescue and the birch, showing that the seedlings picked up nitrogen-fixing Paenibacillus species from the humus. The similarities between the seven Paenibacillus isolates were 97.3–100%. Therefore the isolates may represent one or several species not yet sequenced. Two species of nitrogen-fixing Rhodococcus, R. fascians and R. erythropolis, were found, one in humus and the other in fescue grass. The isolate KM6 was identified as R. fascians with similarity of 100% (840 nucleotides) and the isolate HN41 as R. erythropolis with similarity 99.8% (832 nucleotides) to the respective type strains. The isolates HM29, HM35 and HK40 were identified as species of Pseudomonas, Phyllobacterium and Stenotrophomonas, respectively, with high similarities (Table 2). Isolate KM16 likely represents a new species.

Alder (Alnus incana) seedlings were used as the tool to enumerate nodulation-positive Frankia propagules in the humus. Alder nodulating capacity of the spruce humus soils taken from differently fertilized plots (Table 1) varied greatly, whereas other bacterial species were similarly distributed among the eight plots (Fig. 1). The limed plots of both stands (Table 1) contained 870±358 nodulation units g−1 organic matter, whereas there was almost no nodulation capacity (8±3 g−1 organic matter) in the humus of the nitrogen-fertilized plots of the same stands. The limed and nitrogen-treated plots had nodulation capacities similar to those of the plots with no treatment (244±45 and 265±139 g−1 organic matter, respectively). Plate counts of aerobic heterotrophic bacteria in the same humus soils differed little between the plots, being 106–107 cfu g−1 organic matter in all cases (legend to Fig. 1).

3.3 Antagonistic activities of spruce humus bacteria towards selected plant pathogenic fungi

The plant growth promoting potential of spruce humus bacteria was analyzed by testing their capability to inhibit plant pathogenic fungi. Pseudomonads, bacilli and streptomycetes were selected for testing as they are known to be potential antagonists. The Gram-negative isolates and Streptomyces sp. were tested at pH 5.2, i.e. thus near the pH of the humus (Table 1). Spore-forming bacilli were tested at neutral pH. Two target fungi, uninucleate Rhizoctonia sp. 263 and 264, originated from a conifer tree nursery, represented conifer pathogens. In addition, three species of commonly occurring plant pathogenic fungi were used as target organisms. Thirty-one (48% of the number tested) Streptomyces isolates (Fig. 2), 18 (40%) aerobic spore-forming bacilli (Fig. 3) and 17 (17%) fluorescent Gram-negative isolates (Fig. 3) formed an inhibition zone of >1 mm towards one or more of the tested fungi. Thirty-seven bacterial isolates (18%) were strongly (≥5 mm) antagonistic. All antagonistic Streptomyces isolates were more effective against the uninucleate Rhizoctonia strains than towards the other plant pathogenic fungi (Fig. 2). Two isolates (KMOA5, KMCA1), tentatively identified as Streptomyces albus, produced the widest (>5 mm) inhibition zones, otherwise the antagonistic activity did not correlate with Streptomyces species. Fig. 3 shows the inhibition zones of spruce humus and seedling isolates other than Streptomyces. None of the isolates produced inhibition zones >10 mm to any of the tested fungi. The majority of the isolates with moderately antagonistic effects (>4 mm inhibition zone) were Gram-positives (B. mycoides and B. sphaericus) or Ps. fluorescens.

2

Inhibition of fungal growth by Streptomyces isolates from spruce humus and roots of birch and red fescue. Sixty-five isolates were tested against Rhizoctonia solani HK1 on dual plates. The 31 inhibitory isolates, shown in A were further tested with four other fungi (B–D). Rhizoctonia sp. 263 and Rhizoctonia sp. 264 gave identical responses except to two isolates (differences in the inhibition zones 6 and 1 mm). S. griseoviridis SE1 (Mycostop®) was used as a positive reference strain giving inhibition zones of 7 mm (R. solani HK1), 3 mm (Rhizoctonia sp. 263), 6 mm (Rhizoctonia sp. 264), 17 mm (B. cinerea HK2) and 11 mm (F. culmorum HK3). Streptomyces species names are tentative and based on FAME analysis and MIDI library version 3.9.

3

Inhibition of fungal growth by aerobic spore-forming bacilli and fluorescent Gram-negative isolates from spruce humus and roots of birch and red fescue. Forty-five bacilli (Bacillus, Brevibacillus and Paenibacillus isolates) and 100 Gram-negative (Pseudomonas and Comamonas) isolates were tested as described in Fig. 2. Eighteen and 15 positive strains, respectively, shown in A, were further tested (B–D). Rhizoctonia sp. 263 and Rhizoctonia sp. 264 gave identical responses except to one isolate with a difference in the inhibition zone of 4 mm, and to nine isolates ≤2 mm.

All antagonistic Streptomyces isolates (n=31) except one originated from spruce humus samples. The antagonists found from birch and fescue grass seedlings were identified as S. halstedii (1 isolate), P. pabuli (2), Bacillus sp. (2), Ps. fluorescens (10) and C. acidovorans (3).

Ninety-five Gram-negative isolates, mostly fluorescent pseudomonads, were tested for emission of HCN. Ten fluorescent isolates appeared positive, six of these also inhibited fungal growth. So the growth inhibition in vitro (Figs. 2 and 3) was not explainable by HCN production alone. Sixty-six presumptive plant growth promoting isolates with antagonistic, nitrogen-fixing or HCN producing capacity were tested for excretion of cell wall degrading enzymes. The results are shown in Fig. 4. Antagonistic spore-forming bacilli and Streptomyces, and nitrogen-fixing isolates appeared to commonly produce the tested enzymes and many of them hydrolyzed all tested substrates.

4

Cell wall degrading enzyme productivity by potentially plant growth promoting isolates from spruce humus. The pHs of the plating media were 7.0 (chitinase), 7.3 (pectinase) and 7.5 (CMCase). A clearing zone or halo of >1 mm was regarded as a positive result. Streptomyces included the isolates as in Fig. 2. ‘Spore-formers’ and ‘Gram-negative rods’ include the isolates as in Fig. 3. ‘Nitrogen fixers’ are the isolates identified in Table 2 as species of Paenibacillus (7 isolates), Pseudomonas (1), Stenotrophomonas (1) and unknown, closest related by 16S rDNA to Yersinia (1).

3.4 The production of siderophores or IAA by spruce humus bacteria

The production of siderophores and IAA was investigated to further elucidate the plant growth promotion potential of the spruce humus bacteria (Table 3). Pseudomonas and Comamonas isolates produced siderophores and IAA and most Streptomyces isolates produced siderophores. Table 3 shows that 41 out of the 52 (81%) tested isolates had potential for siderophore production, while 34% had potential for IAA production. If this potential is expressed under environmental conditions (pH 3.7–5.3 in humus, vs. pH 6.8 in test medium), siderophore production may be of importance for plants in forest soils.

View this table:
3

Potential for siderophore and/or IAA production by antagonistic or nitrogen-fixing bacteria from humus layer of spruce stand and humus inoculated birch or red fescue roots

Bacterial isolatesNumber of positive/tested isolates
SiderophoresdIAAe
Antagonistic
Aerobic spore formersa7/131/17
Gram-negative rodsb16/1615/16
Streptomyces14/17ND
Nonantagonistic
Nitrogen fixersc4/61/7
ND=not determined.
  • bInclude genera Comamonas, Pseudomonas and Sphingomonas.

    cInclude genera Paenibacillus, Pseudomonas, Stenotrophomonas and Yersinia, two isolates weakly antagonistic.

    aInclude genera Bacillus, Brevibacillus and Paenibacillus.

    dPositive reaction on CAS agar (a color change around colonies) or by the microtiter method (absorbance <75% of the uninoculated control).

    eA color change on the membrane was regarded as a positive reaction.

4 Discussion

The humus layer of the boreal conifer forest is a hostile environment to bacteria because of the low pH (Table 1) [10]. Our study shows that aerobic spore-forming bacilli as well as nocardioform actinomycetes were prevalent in the humus layer of the Norway spruce stands and the number and variety of Gram-negative isolates were small. Our findings supplement those of Timonen et al. [4] who found low numbers of fluorescent Gram-negatives in Finnish pine forest humus and domination by spore-forming bacteria. Sørheim et al. [32] investigated the acid (pH 4.8, in H2O) floor of Norwegian deciduous forest (beech) and found that the colony-forming isolates were dominantly Gram-negative, Pseudomonas and Alcaligenes, whereas the frequency of spore-forming bacteria was low. The compositions of bacteria in forest humus may therefore not only be determined by pH.

It is known that soil devoid of actinorhizal plants may have capacity of infecting alder seedlings with Frankia [13] although free-living Frankia have not been shown to fix atmospheric nitrogen in soil. Frankia has been shown to colonize and grow on the birch root surfaces [12] but its function for birch is not known. We found a high nodulation capacity for alder in spruce humus soil, devoid of any deciduous trees. Planting with birch has been shown to increase the nodulation capacity of soil for alder of the soils also under experimental conditions in a greenhouse [33].

Nitrogen fixation by pure cultures of Frankia is known to be inhibited by combined nitrogen (e.g. [34]). Our results indicate that long-term nitrogen fertilization debilitated the alder nodulation capacity of the acid conifer humus although there was no observable effect on the aerobic heterotrophic flora measurable by plate counting. Sensitivity towards fertilization has also been reported earlier for the endophytic nitrogen fixers of sugarcane [35]. Frankia nodulation capacity in the limed humus soil was higher than in the unlimed plots with lower pH. A positive correlation between the pH and the nodulation capacity of soil was noted by Smolander and Sundman [13].

We found two nitrogen-fixing Rhodococcus isolates, from the spruce humus and from red fescue roots inoculated with the humus (Table 2). To our knowledge, nitrogen-fixing isolates of Rhodococcus have not been reported before. Our findings of nitrogen-fixing Pseudomonas in spruce humus soil expand on the previous report for red fescue roots [2] collected from meadows.

Nitrogen-fixing Paenibacillus strains have been found widely present in the rhizospheres of different grasses and in soil (e.g. [36]) but the present report is the first description of nitrogen-fixing Paenibacillus from birch roots and from conifer forest humus. 16S rDNA sequences of the seven independent (judged by rep-PCR) nitrogen-fixing Paenibacilli and the two Gram-negative isolates were less than 98% (Table 2) similar to any sequence in the databases, indicating that these belonged to species not yet sequenced. Birch seedlings have been shown to respond to the presence of nitrogen-fixing Enterobacter, Klebsiella and Pseudomonas by a change of root morphology and increased growth of plant biomass [12].

Our study shows that axenic roots of red fescue and birch seedlings selectively attracted bacteria from the surrounding humus soil. Pseudomonas and Alcaligenes groups were scarce in soil but frequent in roots infected by that same soil (Fig. 1). The results of Rosencrantz et al. [37] indicated a similar situation in agricultural soil where rice roots were mainly colonized by α- and β-proteobacteria while Gram-positive bacteria dominated in the surrounding bulk soil. The composition of the root exudates differs with plant species [38]. There have been reports indicating that plant species select the uptake of root bacteria from soil [3942]. In our study the roots favored Gram-negatives. Red fescue also attracted spore-forming bacilli. Latour et al. [39] suggested that in the combination of soil and plant, soil was responsible for the diversity of the bacterial populations associated with the roots. Our results indicate that plant species select bacteria associated with the roots from the bacterial pool in the soil.

About one-third of the Streptomyces, of the spore-forming bacilli and of the fluorescent Gram-negative isolates from the humus layer of the Norway spruce stands and from the roots possessed antifungal properties on the tested fungi pathogenic to conifer seedlings and crop plants. Birch pathogens were not tested because they are poorly known. Many of the antagonistic bacteria had potential for producing fungal cell wall hydrolyzing enzymes, siderophores or IAA in vitro. The antifungal activity of rhizobacteria has been ascribed to these bacterial groups by many authors (e.g. [7, 43, 44]). Of the Streptomyces isolates from spruce humus 46% suppressed one or more of the indicator fungi, similar to the frequency reported previously for rhizosphere and non-rhizosphere isolates [7, 44]. It was interesting in our study that the fungal root pathogens originating from a conifer nursery were highly susceptible to several of the Streptomyces isolates from spruce humus. The humus bacteria may thus have natural potential for attenuating autochthonous fungi as biological control agents in spruce nurseries. Several reports indicate a correlation between in vitro growth inhibition and biocontrol [45], but many other studies found no advantage in screening the strains in vitro [4345]. Broadbent et al. [46] noted that strains nonantagonistic in plate assays were generally inactive in soil also. Although there are limitations in using in vitro assays, they might aid in the first screen of strains for further study.

Acknowledgements

The authors thank Erja Silván for bacterial isolations, Riitta Boeck for gas chromatography analyses, Nina Verta for siderophore assays, and Lars Paulin and Sini Suomalainen for 16S rDNA sequencing. This work was financially supported by the Academy of Finland (K.H., M.S.-S.) and Helsinki University Fund for Excellence (M.S.-S.).

References

  1. [1].
  2. [2].
  3. [3].
  4. [4].
  5. [5].
  6. [6].
  7. [7].
  8. [8].
  9. [9].
  10. [10].
  11. [11].
  12. [12].
  13. [13].
  14. [14].
  15. [15].
  16. [16].
  17. [17].
  18. [18].
  19. [19].
  20. [20].
  21. [21].
  22. [22].
  23. [23].
  24. [24].
  25. [25].
  26. [26].
  27. [27].
  28. [28].
  29. [29].
  30. [30].
  31. [31].
  32. [32].
  33. [33].
  34. [34].
  35. [35].
  36. [36].
  37. [37].
  38. [38].
  39. [39].
  40. [40].
  41. [41].
  42. [42].
  43. [43].
  44. [44].
  45. [45].
  46. [46].
View Abstract