Microbial Community Changes in a Perturbed Agricultural Soil Investigated by Molecular and Physiological Approaches

Abstract

Changes in soil microbial activity and diversity after incubation either with nitrogen or with a mixture of methane and air were examined. The perturbation by methane and air were characterized in detail and led to reduced diversity and enrichment of methanotrophs which were identified by denaturing gradient gel electrophoresis and 16S rRNA sequencing.

Biologically mediated processes in soils are central to the ecological function of soils. Soil methane oxidation is caused by methanotrophs and a heterogeneous collection of cooxidizing bacteria (3–6, 9, 10, 13, 25). Environmental perturbation such as soil management introduces poorly known changes in microbial communities (7, 19, 26). The response of bacterial communities to environmental changes can be assessed by analyzing in situ activity in combination with molecular analysis of the community DNA. In this study the response of the soil bacterial community as well as methanotrophic and methylotrophic activities to anaerobic and methane-enriched atmosphere was investigated. The aim was to investigate whether shifts in microbial activity were reflected in shifts in the community structure. Bacterial community DNA was analyzed to assess the total diversity and generate a profile of the community. Changed profiles indicated altered community structure. Broad resolution approaches such as DNA reannealing kinetics (28), base composition profiles (24, 28), and denaturing gradient gel electrophoresis (DGGE) (20) were applied.

Soil collection and perturbation.

Samples (depth, 0 to 5 cm) were taken from an organic agricultural soil (Krohnestykket, Stend, south of Bergen, Norway). Methods of sampling, characterization, and storage are described elsewhere (15). In laboratory experiments the soil was incubated in two different types of atmosphere at 15°C for 3 weeks. One set was incubated with N₂ gas (N₂ perturbation) and the other was incubated with an atmosphere of air with 17% methane (CH₄ perturbation). Twenty grams of soil (wet weight) was incubated for 3 weeks in 120-ml sterile serum bottles capped with butyl rubber stoppers. The headspace atmosphere was created by flushing to obtain a final concentration of 99.8% nitrogen (flushed once) or 17% methane in air (flushed twice a week).

Physiological measurements.

Methanol and methane were measured by gas chromatography as described by Lindahl et al. (18) and Jensen and Olsen (14), respectively. Methanol oxidation measurements were modified as follows: bottles containing 5 g (wet weight) of soil, 20 ml of autoclaved distilled water, and 5.0 μl of methanol (100% high-performance liquid chromatography grade) were incubated in a shaking water bath at 15°C at 150 rpm. Rate constants of atmospheric methane oxidation and methanol oxidation were calculated (2, 13).

Genetic analyses.

Soil bacteria were extracted from nine 20-g (wet weight) soil samples, and DNA from each soil bacterial fraction was isolated and purified as described by Torsvik et al. (27). Thermal denaturation and reassociation were determined spectrophotometrically in a Cary 4E, UV-visible light spectrophotometer with temperature holder (Varian Instruments) as previously described (27). The DNA complexity was calculated as described by Torsvik et al. (27) with a data acquisition program developed by Svein Norland (University of Bergen, Norway).

The V3 regions of 16S rRNA genes from community DNA were PCR amplified and analyzed by DGGE as described by Øvreås et al. (23). DGGE bands selected for sequencing were reamplified and sequenced as previously described (1, 23).

Physiological changes of soil bacterial communities.

Both perturbations changed the activity of the methane oxidizers and the methanol oxidizers (Table 1). Methane caused the most dramatic changes with highly increased oxidation rates. Nitrogen under practically anaerobic conditions had the opposite effect, reducing both oxidation rates. The control soil revealed methanol oxidation to be biologic and perturbation to increase methane production (Table 1).

TABLE 1.

Physiological changes in the community of methane and methanol oxidizers in agricultural soil after 3 weeks of perturbation at 15°C^a

Perturbation	Rate of methane oxidation (pmol g−1 [dry wt] of soil h−1 at 1 ppmv methane)b	Rate of methanol oxidation (nmol g−1 [dry wt] of soil h−1 at 5.3 mM methanol)c
None (control)	0.6 ± 0.1 (0.0 ± 0.3)	646.6 ± 5.5 (31.7 ± 32.4)
99.8% nitrogen	−0.8 ± 0.2 (−1.2 ± 0.6)	297.2 ± 57.0 (−56.7 ± 40.3)
17% methane in air	97.4 ± 3.9 (−2.1 ± 0.4)	4,242.7 ± 77.9 (−62.6 ± 14.5)

^aNegative values indicate production. Values are means ± standard errors (n = 3). 

^bValues in parentheses indicate methane production when oxidation was blocked by dimethyl ether (21). ppmv, part per million per volume. 

^cValues in parentheses indicate methanol oxidation in soil which was autoclaved twice for 60 min each time with a 10-min intermediate cooling (29). 

Genetic changes of soil microbial communities.

Community DNA extracted from the bacterial fraction from the control soil had a steeper melting profile than that of perturbed soils (Fig. 1A). The difference in mean melting temperatures (T_m) between that for DNA from N₂-perturbed soil (78.4°C) and that for DNA from CH₄-perturbed soil (79.2°C) was only 0.8°C. The T_m of the control soil DNA was 79.0°C. This could idicate minor differences between the communities. The DNA melting range (25 to 75% melting), however, revealed greater community differences. The control soil DNA melted over a smaller temperature range (4.3°C) than those from the perturbed soils (6.3 and 7.6°C for N₂- and CH₄-perturbed soils, respectively). The base composition of DNA from the control soil and the N₂-perturbed soil ranged from 45 to 70 and 40 to 70 mol% G+C, respectively. The DNA from CH₄-perturbed soil ranged from 35 to 77 mol% G+C. The base composition profile had two peaks, indicating that the community consisted of two fractions; one ranging from 35 to 60 and the other from 62 to 77 mol% G+C (Fig. 1B).

FIG. 1.

(A) Thermal denaturation for DNA from control soil (▵), N₂-perturbed soil (◊), CH₄-perturbed soil (○), and E. coli (□). (B) DNA melting profiles (1st derivative of the melting curves). Melting profiles were converted to moles percent G+C (mol % G+C) profiles by using the equation mol% G+C = [(T_m/50.2)−0.99] × 100 (18a). The moles percent G+C of DNA from control soil (▵), N₂-perturbed soil (◊), CH₄-perturbed soil (○), and E. coli (□) are shown.

Reassociation showed that perturbation caused a significant change in community diversity (Fig. 2). The reassociation rate of DNA from the control soil was low and showed the highest C₀t_1/2 value (6,300 mol s⁻¹ liter⁻¹), corresponding to approximately 8,000 different Escherichia coli genomes. N₂ perturbation decreased the C₀t_1/2 value to 2,500 mol s⁻¹ liter⁻¹, which corresponds to 3,200 different E. coli genomes. The lowest C₀t_1/2 value (300 mol s⁻¹ liter⁻¹) was found in DNA from the CH₄-perturbed soil, corresponding to 380 different genomes. The results show that perturbations caused a significant reduction in bacterial diversity, but the data provided no information on species composition.

FIG. 2.

Reassociation (C₀t plots) of DNA from bacterial fractions of control soil (▵), N₂-perturbed soil (◊), CH₄-perturbed soil (○), and E. coli (□). The DNA was sheared to about 420,000 Da and reassociated at 49°C.

Separation of PCR-amplified 16S ribosomal DNA by DGGE showed a complex community structure with more than 100 different bands covering the entire gradient (Fig. 3). N₂ perturbation resulted in minor differences in the community profile with two bands becoming slightly more intense. In the CH₄-perturbed soil some intensified bands were seen on top of the community profile, and we were able to purify and reamplify three of these. Sequences of the purified fragments revealed that they showed phylogenetic affiliation to Methylomicrobium album (band 1), Methylobacter sp. strain BB5.1 (band 2), and Methylomonas rubra (band 3) with sequence homology of >90%. These are all type I methanotrophs belonging to the γ subclass of the class Proteobacteria.

FIG. 3.

DGGE analysis of PCR-amplified 16S ribosomal DNA fragments from soil bacterial communities. DNA was derived from the control soil, a high-organic pastureland (lane A), the same soil incubated for 3 weeks at room temperature with methane (lane B), and the same soil incubated for 3 weeks with nitrogen (lane C). The positions and numbering of bands discussed in the text are indicated with arrows.

We observed gas perturbation to change both the structure and the physiology of microbial communities in an agricultural soil. All changes were found after only 3 weeks and incubation at 15°C. Substrate addition (CH₄) produced greater changes than oxygen removal (N₂). The atmospheric methane oxidation increased after methane and decreased after nitrogen perturbation. Methanol oxidation changed in parallel, indicating that methanol and atmospheric methane were oxidized by methanotrophs. Fast-growing methanotrophs are likely to have leaked exudates such as methanol or formate (11, 12), which can be substrates for, e.g., Methanosarcina spp. or Methylobacterium spp. (2, 8). Hence, perturbation with a substrate (CH₄) specific for the methanotrophs may cause secondary effects which change other microbial populations in the soil as well.

Shifts in the bacterial community structure were observed after both perturbations, as reflected in the broader DNA melting range, skewed base distribution towards lower moles percent G+C, and reduced DNA complexity. The most profound changes were found in CH₄-perturbed soil. The molecular analyses strongly indicated growth of methanotrophic bacteria during the methane perturbation. During the N₂ perturbation the moles percent G+C was lowered by 5%, indicating growth of bacterial types not abundant under aerobic conditions.

DNA from the control soil had an extremely slow reassociation rate, indicating high bacterial diversity prior to perturbation. In this soil the genetic heterogeneity was more than twice the diversity in the N₂-perturbed soil. In the CH₄-perturbed soil the diversity was 1/20 of that in the control soil. The diversity reflects the number of genetically different bacterial types responding to the added substrate and indicates an outgrowth of a few dominant bacterial species. Complex microbial communities like soil are difficult to analyze using the DGGE technique (22). More than 100 amplified fragments were seen in all the samples. Use of this method indicated that anaerobic conditions (N₂) caused growth of bacteria which were suppressed under aerobic conditions or were stimulated by substrates becoming available, e.g., from cells lysed due to oxygen deficiency. In the CH₄-perturbed soil, growth of methane oxidizers was demonstrated by the DGGE analysis. Sequence analysis suggested that the dominant community members in this soil were related to cultivable methanotrophs, all of which are found to have low affinity to methane (16, 17).

The parallel increase in methanol and methane oxidation indicated that methanotrophs consume both methane and methanol in the soil. In the methane-perturbed soil, DGGE analyses revealed an increase in numbers of known methanotrophs, and an increased oxidation rate at 1 part per million per volume (ppmv) methane was found. Our data support the view that in the presence of atmospheric methane concentrations the known methanotrophs oxidize atmospheric methane by the concomitant consumption of methanol (6, 13).

We conclude that apparent changes in microbial activity and community structure were observed after perturbation. The bacterial communities from perturbed soils showed a significant decrease in diversity compared to that from the control soil. The methane-perturbed soil was dominated by a few community members showing phylogenetic affiliation to known methanotrophic bacteria. Our investigation indicated that these methanotrophs, although they were related to cultured, low-affinity methanotrophs, were responsible for oxidation of atmospheric concentrations of methane. Our investigation demonstrates that the polyphasic approach can elucidate the relationship between microbial activity and community structure and identify the bacteria which have a key role in a process.

Nucleotide sequence accession numbers. The sequences obtained in this study are available from GenBank (band 1, AF067423; band 2, AF067424; band 3, AF067425).