OUP user menu

Molecular analysis of ammonia-oxidising bacteria in soil of successional grasslands of the Drentsche A (The Netherlands)

George A. Kowalchuk , Atie W. Stienstra , G. Hans J. Heilig , John R. Stephen , Jan W. Woldendorp
DOI: http://dx.doi.org/10.1111/j.1574-6941.2000.tb00685.x 207-215 First published online: 1 March 2000

Abstract

Changes in the community structure of chemolitho-autotrophic ammonia-oxidising bacteria of the β-subgroup Proteobacteria were monitored during nutrient-impoverishment management of slightly acidic, peaty grassland soils, which decreased in pH with succession. Specific PCR, cloning and sequence analysis, denaturing gradient gel electrophoresis (DGGE) and probe hybridisation were used to analyse rDNA sequences directly recovered from successional soils. Four previously characterised ammonia oxidiser sequence clusters were recovered from each soil, three associated with the genus Nitrosospira and one with the genus Nitrosomonas. All samples were dominated by Nitrosospira-like sequences. Nitrosospira cluster 3 was the most commonly recovered ammonia oxidiser group in all fields, but a greater representation of Nitrosospira clusters 2 and 4 was observed in older fields. Most probable number (MPN) counts were conducted using neutral and slightly acid conditions. Neutral pH (7.5) MPNs suggested a decrease in ammonia oxidiser numbers in later successional fields, but this trend was not observed using slightly acid (pH 5.8) conditions. Analysis of terminal MPN dilutions revealed a distribution of sequence clusters similar to direct soil DNA extractions. However, an increased relative recovery of Nitrosospira cluster 2 was observed for acid pH MPNs compared to neutral pH MPNs from the most acidic soil tested, in agreement with current hypotheses on the relative acid tolerance of this group.

Keywords
  • β-Subgroup Proteobacteria
  • Soil pH
  • Nitrification
  • DGGE
  • Most probable number (MPN)
  • Oligonucleotide hybridisation
  • Nitrosospira
  • Nitrosomonas

1 Introduction

In an attempt to counteract the decrease in floral diversity resulting from the use of nitrogen-rich fertilisers, significant efforts have been made in recent decades to convert intensely used fields into more natural species-rich grasslands via the use of impoverishment management strategies. Such strategies involve a halt in fertiliser application, mowing once a year with hay removal, and irregular grazing. Secondary plant succession begins with the implementation of the impoverishment management. In the Drentsche A region in the northeast of the Netherlands, pastures have been taken out of fertilisation periodically over the last 50 years, thus creating a natural experiment for the study of nitrogen cycle processes related to the induced environmental changes associated with impoverishment field management. This peaty, stream valley region is lime poor and therefore slightly acid (pH 4.2–5.5), and secondary succession is accompanied by a decrease in pH following the onset of impoverishment management [1]. Secondary succession has also been shown to correlate with changes in the community structure of dominant soil bacteria in these fields [2].

Chemolithotrophic ammonia oxidation is a key process in the global nitrogen cycle and is the principle route of ammonia conversion in many soil ecosystems [35]. The secondary plant succession that occurs during impoverishment management is accompanied by decreases in nitrification, culturable nitrifier numbers and net N mineralisation [1,68]. Although environmental factors such as ammonia availability and pH have been shown to affect ammonia oxidiser growth and survival in pure culture experiments, our understanding of how these factors impact natural ammonia oxidiser populations in terrestrial ecosystems remains limited. Ammonia oxidiser populations have been shown to up-regulate certain metabolic activities that promote survival during times of prolonged substrate limitation, but such experiments have been conducted with pure cultures of more easily isolated ammonia oxidiser strains (e.g. Nitrosomonas europaea) [911,12]. Similarly, environmental pH can affect nitrification activity both in pure culture and in soil. Ammonia oxidiser pure cultures typically fail to nitrify below pH 6 [4], but some isolates can adapt to acid conditions after exposure to pH fluctuations or immobilisation in alginate [13]. Furthermore, autotrophic nitrification has been shown to continue in low pH soils [14], and a previous study in the Drentsche A revealed that the optimal pH for nitrification in soil suspensions decreased with decreasing pH of the soil [1]. Thus, culture-based studies have failed to resolve whether ammonia oxidation in acid environments is due to the adaptation of generalist ammonia oxidiser populations, perhaps attached to surfaces or in pH neutral microsites [4], or the presence of particular ammonia oxidiser ecotypes that are well suited to acid environments.

Ecological studies concerning β-subgroup ammonia-oxidising bacteria have been hampered by difficulties encountered with their isolation and manipulation in pure culture. However, the monophyletic nature of the β-subgroup ammonia oxidisers has allowed the development of a variety of PCR-based strategies designed to detect these bacteria in the environment by either targeting the 16S rRNA gene [1518] or the amoA gene [19], which encodes the alpha subunit of ammonia monooxygenase [20]. All known ammonia-oxidising strains isolated from terrestrial ecosystems fall within the Nitrosospira/Nitrosomonas radiation of the β-subgroup Proteobacteria based upon 16S rDNA sequences [21,22]. These two genera can be further subdivided into at least four 16S rDNA sequence clusters each (Fig. 1) [16,23], based upon sequences recovered from the environment, and the relative abundance of certain sequence clusters has been correlated with specific environmental conditions, including soil pH and agricultural use [18,2427].

1

Neighbour joining tree of partial 16S rDNA sequences focusing on the β-subgroup ammonia oxidiser clade. Sequences recovered from the clone libraries generated in this study are prefixed with ‘pDA’ and codes indicate the sample site and DGGE migration category (see Table 2). For instance, clone pDA-A(h).3 would indicate the DGGE class 3 clone sequenced from sample A(h). Sequences whose names begin with ‘Env’ and ‘pH’ were recovered via PCR from marine sediment and arable soil samples, respectively, and the A1bM3 and Enr-ZD5 sequences were recovered from enrichment cultures of marine sediment and arable soil, respectively [15]. Identical sequences have been placed on the same branch.

This study sought to investigate the response of ammonia oxidiser populations to changes in ammonia availability and pH encountered in successional grassland soils. To this end, detailed ammonia oxidiser community fingerprints were generated by coupling PCR amplification of specific 16S rDNA targets, from soil DNA extracts, with the use of denaturing gradient gel electrophoresis (DGGE) and oligonucleotide probe hybridisation [16,17]. A second PCR primer set [15] was also used for the direct amplification, cloning and partial sequencing of ammonia oxidiser-like 16S rDNA fragments. In addition, the role of pH in selecting specific ammonia oxidiser populations was tested using most probable number (MPN) methods with either neutral or slightly acid culture media. Enrichment cultures, derived from the highest positive MPN dilutions, were characterised by DGGE and hybridisation to allow for comparison between direct and culture-mediated analyses. A culture-independent method of evaluating population size, a competitive PCR assay targeting amoA [28], was compared to the MPN dilution results.

2 Materials and methods

2.1 Sample sites and sampling procedure

The sampling area, the Drentsche A, is located along a brook in a glacial sand plain in the northeastern part of the Netherlands (53°05′N, 6°40′E). The field locations, labelled A–D, have been taken out of fertilisation for a number of years (Table 1). The groundwater table for these fields is usually low in summer (>120 cm) and high in winter (<40 cm). Certain areas of the region can also become inundated with floodwaters that can persist for as long as several months per annum. To account for differences in exposure to flooded conditions, field A was sampled in two locations, A(1) and A(h), at locations with mean ground water tables of 45 cm and 80 cm respectively. The wetter section of field A was dominated by Agrostis stolonifera, Holcus lanatus and Poa trivialis, the drier section by Lolium perenne, Dactylus glomerata and Elymus repens. The dominant plant species for field B was H. lanatus. The most abundant plant species in field C were H. lanatus, Anthoxanthum odoratum and Deschampsia cespitosa, whereas the most dominant species in field D were Juncus acutiflorus and Agrostis capillaris [6,8]. All samples were taken in March of 1996. The sampling procedure and determination of soil parameters and potential nitrification activity (PNA), were as previously described [1,26]. This involved collection of three soil cores (30 mm diameter; 100 mm depth) from each of 15 equidistant (5 m apart) points within each field. Each group of three cores was sieved, analysed for moisture content and the 15 samples per site pooled using an equal amount of each three-core subsample. Triplicate subsamples from these pooled samples were used for enumeration of ammonia-oxidising bacteria using MPN and cPCR methods.

View this table:
1

Soil characteristics, including potential nitrifying activity (PNA)

FieldYears since fertilisationOrganic matter (%)NO3-N (μmol g−1 dry soil)NH4+-N (μmol g−1 dry soil)pH (KCl)Potential nitrifying activity (nmol NO3+NO2 h−1 g−1 dry soil)
A(l)79.00.2240.1815.19.2
A(h)711.00.4250.1135.212.2
B1111.00.1120.8204.812.3
C248.40.2340.3554.411.4
D>507.10.0900.4294.23.7
Field age is given in reference to the date of sampling.

2.2 MPN cultures and preparation of template DNA for PCR

Ammonia-oxidising bacteria were enumerated using the most probable number (MPN) procedure described by Stienstra et al. [1], with the following modifications: 5 mM (NH4)2SO4 as ammonia source, 50 mg l−1 cycloheximide, and an initial pH of either 7.5 or 5.8, buffered by 5 mM MOPS (4-[N-morpholino]butanesulfonic acid) or MES (2-[N-morpholino]ethanesulfonic acid), respectively. Counts were carried out in triplicate for each medium. MPN microtitre plates were wrapped in aluminium foil to prevent evaporation and incubated in the dark for 12 weeks at 27°C without shaking. Positive nitrification activity was determined as the presence of nitrate and/or nitrite (>0.1 mM NO3+NO2), as detected using a Technicon Traacs 800 autoanalyser (Technicon Instruments Corp. Tarrytown, NY, USA). The highest MPN dilution to show positive nitrification for each dilution series was used as inoculum (0.5 ml) in 9.5 ml of the same culture medium as used for MPN enumeration with the addition of 0.04% w/v bromothymol blue as pH indicator. Enrichment cultures were grown in the dark at 27°C for 60 days without shaking and bacterial cells harvested by centrifugation at 14 000×g for 15 min. All but 0.4 ml of the supernatant was decanted, and cell pellets were resuspended in the remaining medium and transferred to 0.5 ml microcentrifuge tubes. Concentrated cell suspensions were incubated for 10 min at 100°C, frozen for 30 min at −20°C, and heated once more to lyse cells. Lysates were stored at −20°C until further use. PCR-DGGE and hybridisation (see below) recovered single sequence types from all but four enrichment cultures, which each contained two distinct ribosomal sequence types. Each ribotype was treated separately for the ribotype frequency analysis.

2.3 DNA extraction from soil, 16S rDNA analyses and amoA target number estimation

DNA was directly extracted from soil by mechanical disruption [16], followed by composite agarose gel-mediated DNA purification [17]. PCR, using either DNA extracted from soil or enrichment cultures, was performed using the CTO189f-GC and CTO637r primers, for the specific amplification of β-subgroup ammonia oxidiser 16S rDNA [17], with the incorporation of a GC-clamp. Reactions used 1 μl template (approximately 30 ng of DNA for soil extractions) in a 50 μl reaction volume using the PCR conditions described by Kowalchuk et al. [17]. Partial 16S rDNA fragments were also generated by PCR from soil DNA using the βAMOf and βAMOr primers and PCR conditions described in [15]. PCR product concentrations were estimated by comparison to known standards after agarose gel electrophoresis (1.5% w/v agarose, 0.5×TBE; 1×TBE=90 mM Tris–Borate, 2 mM EDTA, pH 8.3) and ethidium bromide staining. Competitive PCR for the estimation of amoA target numbers was as described previously [28], using the same DNA extractions as template and was performed in triplicate.

2.4 DGGE and hybridisation analyses

DGGE analyses were performed in triplicate as described previously [17] using a D-Gene apparatus (BioRad, Hercules, CA, USA). Approximately 200 ng PCR product was loaded in each lane. DNA was transferred to nylon membranes using a semi-dry electroblotting apparatus (model SD, BioRad). Hybridisation conditions were according to Stephen et al. [18], and signal intensities determined densitometrically from autoradiographs or by direct phosphoimager counts [18]. Hybridisation intensities for control lanes, containing known quantities of 16S rDNA previously defined with respect to sequence cluster affinity [16,17], were used to calibrate the relative hybridisation efficiencies of the sequence cluster-specific probes and to standardise counts between membranes. The combined signals from the four sequence clusters represented between 96% and 103% of the total β-subgroup ammonia oxidiser hybridisation signal per lane of the DGGE gel analysed. The summed signals per lane were standardised to 100% for graphical representation.

2.5 Generation, screening and analysis of 16S rDNA clones

Cloning of PCR products generated with the primers βAMOf and βAMOr in Escherichia coli was performed using the T-Vector kit according to manufacturer's instructions (Promega Corp., Madison, WI, USA). Colony PCR, using the vector primers T-7 and SP6 was performed to identify transformants containing inserts of the appropriate size (approximately 1.26 kb including vector sequences). Colonies containing such inserts (30 per library) were further screened by PCR for the presence of β-subgroup ammonia oxidiser-like 16S rDNA fragments using the CTO189f-GC and CTO654r primers [17,24]. DGGE was subsequently performed for clones that yielded PCR product and one clone of each DGGE migration category per sample (see Table 2) was selected for sequence analysis. Double-stranded cycle sequencing was performed using 1 μg plasmid DNA as template and the primers CTO189f and CTO654r as well as primers internal to the vector, as described previously [17]. Phylogenetic analysis was performed as previously described [16,17], for 287 positions that could be unambiguously aligned for all sequences used. Bootstrapping was conducted with 100 replicates. Novel partial 16S rDNA sequences have been deposited in the EMBL databank (Accession numbers AJ131791AJ131801).

View this table:
2

PCR and DGGE screening of 16S rDNA clones

FieldNumber of clones giving positive PCR with CTO primersaNumber of clones per DGGE migration category
1 (43.5%)2 (44.5%)3 (45%)4 (45.5%)5 (46%)
A(l)11 (37%)30611
A(h)12 (40%)10902
B10 (33%)00721
C9 (30%)11421
D6 (20%)01212

2.6 Statistical analyses

Linear correlations (Pearson's R) for the comparison of the sequence cluster distributions detected by the different methods (direct extraction vs. the two enrichment strategies) were performed using Statistica, version 5.1 for Windows (Statsoft Inc., Tulsa, OK, USA). Student's t-test and Chi-square test were used to evaluate the significance of differences within the DGGE-hybridisation, MPN count and amoA competitive PCR data [1,28] and calculations were carried out within an Excel spreadsheet (Microsoft Office 97, Microsoft Corp.).

3 Results and discussion

3.1 PCR, DGGE and hybridisation analysis of ammonia oxidiser-like 16S rDNA sequences from directly extracted soil DNA

A positive PCR signal was generated using the β-subgroup ammonia oxidiser-specific primers, CTO189f-GC and CTO654r, for all fields of the secondary successional series examined. For all samples, ammonia oxidiser-specific PCR-DGGE produced six detectable bands in three doublets, within the range of 43–46.5% denaturant chemicals (Fig. 2A). The appearance of doublet bands is consistent with other DGGE results using these primers and was due to an ambiguous position in the reverse primer [17].

2

A: DGGE analysis of PCR-recovered 16S rDNA fragments from Drentsche A grassland fields. PCR used the CTO189fGC and CTO637r primers [17], and field designations are as given in the text, with numbers in parentheses representing the number of years under impoverishment management. 16S rDNA fragments of known ammonia oxidiser sequence cluster affiliation were also included as controls (first five lanes [17]). B: Results of hybridisation analysis of environmental DGGE patterns. Three gels similar to that shown in panel A were transferred to a nylon membrane and hybridised with the oligonucleotide probes specific for the following β-subgroup ammonia oxidiser sequence clusters: Nitrosospira cluster 2 (NspC12_458) ▧, Nitrosospira cluster 3 (NspC13_454) ▩, Nitrosospira cluster 4 (NspC14_446). Nitrosospira cluster 6a (NmoC16a_205) ▤ [18]. Values represent the average of three analyses, and the intervals shown above the bars refer to standard deviation.

Hybridisation analysis identified the uppermost doublet as Nitrosomonas cluster 6 [18], previously detected in a variety of environments [15,16]. Sequences belonging to a subgroup of this cluster, currently termed cluster 6a, have been detected in arable soil and freshwater sediment environments [16,18,24]. The specific oligonucleotide probe for this subgroup (NmoCL6a_205) [18] hybridised with this uppermost DGGE doublet, identifying it as Nitrosomonas cluster 6a. All other detectable bands on the DGGE hybridised with probes specific for Nitrosospira-like sequences. The lowest doublet corresponded to Nitrosospira cluster 4 and the middle doublet contained a mixture of Nitrosospira clusters 2 and 3. The co-migration of these latter two sequence clusters during DGGE has been observed previously [17,18].

Quantification of cluster-specific hybridisation signals revealed the relative amount of PCR product recovered for each of the detected sequence clusters across the successional fields sampled (Fig. 2B). All samples contained a predominance of Nitrosospira-like sequences, ranging from 88–95% of the total β-subgroup ammonia oxidiser signal. Nitrosospira cluster 3 was the most commonly detected signal in all fields, although its frequency decreased from 77% in the A(h) sample to 36% in sample D. The relative proportions of both Nitrosospira clusters 2 and 4 increased with field age from 10% and 4% respectively in sample A(h) to 31% and 23% in field D. No consistent patterns were found with respect to the distribution of Nitrosomonas cluster 6a. Differences between the A(l) and A(h) samples were minor.

3.2 Phylogenetic analysis of cloned 16S rDNA fragments

PCR using the β-AMOf and β-AMOr primers [15], yielded positive results for all fields tested. This primer set has previously been found to be semi-specific for the amplification of β-subgroup ammonia oxidiser-like 16S rDNA fragments [16]. For this reason, transformants containing inserts were screened by PCR using the CTO189fGC/CTO654r primer set, previously shown to be diagnostic for the presence of ammonia oxidiser-like inserts [17,24]. The proportion of inserts which showed positive amplification with the CTO189fGC/CTO654r primers varied from 20% in sample D to 40% in the A(h) sample (Table 2). The relative recovery of ammonia oxidiser-like clones with the βAMO primers may be influenced by the ratio of β-subgroup ammonia oxidiser 16S rDNA to 16S rDNA recovered from closely related β-subgroup Proteobacteria compatible with these primers.

Five DGGE mobility classes were observed among the clones examined, and phylogenetic analysis of clones selected for partial sequencing detected the same four ammonia oxidiser sequence clusters as detected by PCR-DGGE using the CTO189f-GC/CTO654r primers (Fig. 1). DGGE classes 1, 2, and 5 produced sequences showing affinity with Nitrosomonas cluster 6a, Nitrosospira cluster 2 and Nitrosospira cluster 4, respectively. DGGE migration category 3 included sequences from Nitrosospira clusters 2, 3 and 4, and DGGE migration category 4 produced sequences affiliated with Nitrosospira clusters 3 and 4. It should be noted that some sequence variation was probably missed given that only a single clone of each DGGE migration category per sample was chosen for sequence analysis. DGGE migration alone has previously been shown to be a poor indicator of sequence identity or phylogenetic affiliation [17]. Sequence analysis was not, however, intended to be an exhaustive cloning-based study, but was used to confirm the specificity and comprehensiveness of the PCR/DGGE/hybridisation strategy. Although the sequence data from cloned 16S rDNA fragments agreed well with the sequence cluster distributions presented in Fig. 2B, the number of clones examined was too small to allow a statistical comparison. Nevertheless, that the same ammonia oxidiser sequence clusters were detected using different primer sets in these two strategies strengthens the suggestion that the detected groups represent the dominant β-subgroup ammonia oxidisers in these samples.

3.3 MPN counts and 16S rDNA analysis of enrichment cultures

MPN counts using neutral pH culture medium suggested a decrease in the culturable ammonia oxidiser cell count with respect to time since fertiliser application (Fig. 3). However, with both MPN and cPCR counts (see below), the use of pooled samples rather than multiple subsamples from each site precluded estimation of within-field variation. Thus, while the consistency in population structure described within and between experimental methods suggests that differences in ammonia oxidiser population structure varied with age, the lack of knowledge of variation within each field prevented strong statistical support for these conclusions. The above trend was not observed for the MPN counts conducted with slightly acid pH culture medium, where no significant differences were detected (Fig. 3).

3

Summary of hybridisation results from PCR-DGGE and MPN counts. Black bars, ■, represent the number of amoA target sequences per gram dry soil, as estimated by competitive PCR [36]. Spotted bars, Embedded Image, and gray shaded bars, ▩, give the MPN counts determined using culture media of pH 7.5 and pH 5.8, respectively. Bars marked with the same letter are not significantly different from each other at the level of 5% (Tukey's text). Intervals shown above the bars refer to standard deviation.

In order to assess the effect of culturing and culture medium pH on the recovery of ammonia oxidiser sequence clusters, enrichment cultures derived from terminal positive MPN dilutions were screened by PCR-DGGE and hybridisation analysis. A total of 300 enrichments were tested, divided evenly between the two media tested and five field sites examined (i.e. 30 cultures field−1 medium−1). All but four of the MPN cultures used in community structure analysis appeared to contain a single ammonia oxidiser ribotype, though neither hybridisation analysis nor sequencing of the excised DGGE bands are adequate methods to prove that other ribotypes were not co-cultured (e.g. [29]). The same four ammonia oxidiser sequence clusters were recovered as were detected by the direct PCR-DGGE analysis, and all four sequence clusters were detected in each of the soil samples (Table 3). The distribution of clusters within a treatment was compared to those recovered by direct analysis of soil DNA and between the two types of medium used for each field. High levels of correlation were found between the direct analyses and the terminal dilutions grown at pH 7.5 (r=0.98, P<0.01 for all fields). Terminal MPN cultures grown at pH 5.8 from all fields were also distributed between the four sequence clusters significantly in accordance with the results gained by the direct analyses, though less closely than those grown at pH 7.5 (r=0.89, P<0.05). Thus, the ammonia oxidiser community structures, as assayed by these three methods, did not differ significantly. The only significant differences in population structure were between those suggested by MPN terminal dilutions from field D grown at pH 7.5 and pH 5.8 (r=0.85, P=0.07). The lack of a significant correlation between these distributions was mainly due to the relative recovery of Nitrosospira clusters 2 and 3. As with the direct PCR-DGGE analysis, enrichment cultures using pH 7.5 medium showed an increase in Nitrosospira cluster 2 recovery with increased field maturity, but Nitrosospira cluster 3 still dominated 40% of the cultures examined for field D (Table 3). In contrast, the pH 5.8 medium only showed a 10% recovery of Nitrosospira cluster 3 cultures from field D, with the majority of cultures (53%) being dominated by Nitrosospira cluster 2.

View this table:
3

Distribution of ammonia oxidiser clusters in MPN-derived enrichment cultures

FieldpH 7.5 mediumpH 5.8 medium
Nitrosospira cluster 2Nitrosospira cluster 3Nitrosospira cluster 4Nitrosomonas cluster 6aNitrosospira cluster 2Nitrosospira cluster 3Nitrosospira cluster 4Nitrosomonas cluster 6a
A(1)1213551285
A(h)1215341394
B3223261284
C6146412864
D8127316383

PCR, DGGE and hybridisation were used to determine the cluster designations of ammonia oxidiser populations enriched in cultures derived from terminal MPN dilutions (30 field−1 medium−1). Correlation analyses did not suggest that the population structure of any field differed from that suggested by the direct hybridisation analysis, nor did comparison of population structures suggested by the two media differ significantly, with the exception of field D (r=0.85, P=0.07).

Acidic MPN conditions produced lower MPN values for younger fields than neutral MPN cultures, but gave higher MPN counts for the more acidic, late successional fields (Fig. 3). The pH of the MPN culture medium significantly influenced both the number of ammonia oxidiser cells detected and, for field D, the relative distribution of the ammonia oxidiser sequence clusters present in the highest positive dilutions (Table 3). The slightly acid conditions enhanced recovery of Nitrosospira cluster 2 at the expense of Nitrosospira cluster 3, and the MPN counts may have been affected by differences in growth rates of different ammonia oxidisers under neutral and acid pH conditions. These results are in agreement with previous soil suspension experiments that showed a decrease in the optimal pH of nitrification with increased field age and decreased soil pH [1].

The estimation of amoA gene targets in the soil by competitive PCR did not reveal any significant differences between the successional fields examined. The number of amoA targets detected was between 5 and 500 times higher than the number of cells estimated by MPN, depending on the sample and the pH of the culture medium used (Fig. 3). The relatively low MPN counts for late successional fields using the neutral pH culture medium suggests that this medium may not be efficient for recovery of some ammonia-oxidising bacteria from these acidic soils. Notably, the proportion of Nitrosospira cluster 2 MPNs recovered from all soils by neutral pH medium was reduced when compared to the acidic medium (Table 3).

The detection of a dominance of Nitrosospira-like bacteria by all the methods used is consistent with a number of studies suggesting nitrosospiras to be the dominant β-subgroup ammonia oxidisers in a number of soil environments [1618,30,31]. The relative recovery of the detected ammonia oxidiser groups differed little between the analyses used, but was apparently influenced by the pH of the culture medium (see above). These differences were tentatively attributed to varying degrees of culture bias imparted by the two culture media used [4] but differences due to random or systematic biases in the PCR [32] or DNA extraction protocols [33,34] cannot be ruled out. Although ammonia availability and soil pH were implicated as major factors influencing the distribution of ammonia oxidiser populations in these successional fields, potential affects of other factors such as the production of allelochemicals by late successional plants or differences in moisture content cannot be excluded [3537].

3.4 Conclusions

A shift in β-subgroup ammonia oxidiser community structure was apparent between early and late successional fields. Two of the three methods used to assay the population size did not indicate any significant change with succession. The dominance of Nitrosospira cluster 3, previously detected as dominant in neutral pH agricultural soils [16,18], decreased with field age. In contrast, there was a relative increase in the recovery of Nitrosospira cluster 2 and Nitrosospira cluster 4-affiliated sequences. Nitrosospira cluster 2-affiliated sequences have previously been associated with acidic soil environments [16,18]. The increased detection of Nitrosospira cluster 2 sequences in older, more acidic fields is consistent with the hypothesis that this phylogenetic grouping carries a selective advantage over other Nitrosospira lineages in low pH soil environments. The increased detection of Nitrosospira cluster 2-affiliated sequences in acid pH MPN cultures, as compared to neutral pH MPNs, further supports the hypothesis that acid conditions preferentially select for this ammonia oxidiser group. Nitrosospira cluster 4 has been shown to increase in abundance relative to Nitrosospira cluster 3 in neutral pH soils that have never been tilled or have been taken out of agricultural production for extended periods of time [26,27]. These findings support the hypothesis that distinct ammonia oxidiser populations are associated with the specific environmental conditions of either early or late successional pastures.

Acknowledgments

G.A.K. was funded by an NWO grant to the Netherlands Graduate School of Functional Ecology. The authors thank Wietse de Boer and Lijbert Brussaard for helpful comments on the manuscript and Sarah J. Macnaughton (CEB) for invaluable assistance with the statistical analyses.

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].
View Abstract