Growth of ammonia-oxidizing archaea in soil microcosms is inhibited by acetylene


  • Editor: Christoph Tebbe

Correspondence: Graeme W. Nicol, Institute of Biological and Environmental Sciences, University of Aberdeen, Cruickshank Building, St. Machar Drive, Aberdeen AB24 3UU, UK. Tel.: +44 12 24 272 258; fax: +44 12 24 272 703; e-mail:


Autotrophic ammonia-oxidizing bacteria were considered to be responsible for the majority of ammonia oxidation in soil until the recent discovery of the autotrophic ammonia-oxidizing archaea. To assess the relative contributions of bacterial and archaeal ammonia oxidizers to soil ammonia oxidation, their growth was analysed during active nitrification in soil microcosms incubated for 30 days at 30 °C, and the effect of an inhibitor of ammonia oxidation (acetylene) on their growth and soil nitrification kinetics was determined. Denaturing gradient gel electrophoresis (DGGE) analysis of bacterial ammonia oxidizer 16S rRNA genes did not detect any change in their community composition during incubation, and quantitative PCR (qPCR) analysis of bacterial amoA genes indicated a small decrease in abundance in control and acetylene-containing microcosms. DGGE fingerprints of archaeal amoA and 16S rRNA genes demonstrated changes in the relative abundance of specific crenarchaeal phylotypes during active nitrification. Growth was also indicated by increases in crenarchaeal amoA gene copy number, determined by qPCR. In microcosms containing acetylene, nitrification and growth of the crenarchaeal phylotypes were suppressed, suggesting that these crenarchaea are ammonia oxidizers. Growth of only archaeal but not bacterial ammonia oxidizers occurred in microcosms with active nitrification, indicating that ammonia oxidation was mostly due to archaea in the conditions of the present study.


Ammonia oxidation by aerobic autotrophic bacteria is a two-step process: ammonia is oxidized to hydroxylamine by ammonia monooxygenase (AMO) and hydroxylamine is subsequently oxidized to nitrite by hydroxylamine oxidoreductase (Prosser, 1989). Acetylene is a potent inhibitor of this process (Hynes & Knowles, 1978, 1982; Bédard & Knowles, 1989) and Hyman & Wood (1985) have shown that acetylene is a suicide substrate for the AMO, i.e. the oxidation of acetylene by AMO results in permanent inhibition of the enzyme. Recovery of bacterial ammonia oxidizers exposed to acetylene takes at least 10 days and requires protein synthesis (Hynes & Knowles, 1978, 1982; Hyman & Wood, 1985). Acetylene does not impair the conversion of hydroxylamine into nitrite (Hynes & Knowles, 1978). Complete inhibition of ammonia oxidation in pure culture of the autotrophic ammonia oxidizer Nitrosomonas europaea occurs at ≥1 Pa partial pressure (0.001% v/v) of acetylene, but no effect is observed at 0.1 Pa (Hynes & Knowles, 1978). In contrast, acetylene does not efficiently inhibit ammonia oxidation by heterotrophic nitrifiers (Hynes & Knowles, 1982; Moir et al., 1996; Daum et al., 1998).

Inhibition of soil ammonia oxidation by acetylene has been investigated extensively (Walter et al., 1979; Mosier, 1980; Berg et al., 1982; McCarty & Bremner, 1986, 1991; Garrido et al., 2000). Low acetylene concentrations (c. 10 Pa) are generally sufficient for 100% inhibition of autotrophic ammonia oxidation under standard incubation conditions (Berg et al., 1982; Klemedtsson et al., 1988; Bateman & Baggs, 2005), with some exceptions (Garrido et al., 2000), and acetylene inhibition is being developed to manage groundwater pollution by nitrate and emission of the greenhouse gas nitrous oxide, associated with the microbial oxidation of ammonia-based fertilizers (Subbarao et al., 2006; Singh & Verma, 2007). Inhibition of soil ammonia oxidation by acetylene is also used in routine methods for assessment of nitrous oxide emission (Klemedtsson et al., 1988; Kester et al., 1996) and to distinguish its production by nitrifiers and denitrifiers (Ryden et al., 1979) in soil.

Until recently, autotrophic ammonia-oxidizing bacteria were considered to be responsible for the majority of ammonia oxidation in soil. Recent studies, however, have provided evidence for the existence of novel lineages of autotrophic ammonia-oxidizing prokaryotes within the archaeal phylum Crenarchaeota (Könneke et al., 2005; de la Torréet al., 2008; Hatzenpichler et al., 2008; Prosser & Nicol, 2008). These crenarchaea possess genes homologous to those encoding subunits A, B and C of the AMO (amoA, amoB and amoC homologues) of autotrophic ammonia-oxidizing bacteria. Cultivated representatives of this group have been showed to grow with ammonia as the sole energy source, with near stoichiometric conversion of ammonia to nitrite (Könneke et al., 2005; de la Torréet al., 2008) and incorporation of inorganic carbon (Hatzenpichler et al., 2008). Archaeal amoA genes are widely distributed in soils, and quantification of amoA genes and transcripts in several soils indicates greater abundance (Leininger et al., 2006; He et al., 2007; Nicol et al., 2008) and transcriptional activity (Leininger et al., 2006; Nicol et al., 2008) of archaeal than bacterial ammonia oxidizers, suggesting an important role for crenarchaea in soil ammonia oxidation.

The discovery of abundant archaeal ammonia oxidizers in soil emphasizes the need to characterize conditions favourable for their growth and to evaluate the use of acetylene to block all assumed autotrophic ammonia oxidation. The aim of this study was therefore to assess the growth of archaeal and bacterial ammonia oxidizers in soil during active ammonia oxidation and to characterize the effect of acetylene on their abundances and soil nitrification kinetics using DNA-based analyses of 16S rRNA and amoA gene abundances and community structures. The study was performed in microcosms containing soil for which a previous study (Tourna et al., 2008) showed changes in the relative abundance of archaeal amoA gene copies and transcripts during nitrification when soil was incubated at 30 °C. This may therefore represent a system whereby the growth of ammonia-oxidizing archaea can be reproducibly demonstrated and monitored under different conditions.

Materials and methods

Soil microcosms and analysis

Soil was collected from the upper 10 cm of an agricultural plot (Scottish Agricultural College, Craibstone, Aberdeen, Scotland, Grid reference NJ872104), which has been maintained since 1961 at a pH of c. 7. Detailed soil characteristics are provided by Kemp et al. (1992). Ten soil samples of 200 g, collected at random from the plot, were pooled, sieved (3.35-mm mesh size) and stored at 4 °C before construction of microcosms. Microcosms consisted of 10 g of fresh soil in 144-mL sterile serum bottles. Four sets of nine replicate microcosms were prepared and sealed with rubber stoppers and aluminium caps. Acetylene was then added to the headspace of 27 microcosms (three of the four sets) at partial pressures of 10 Pa (0.01%), 100 Pa (0.1%), and 1000 Pa (1%), respectively. Microcosms (9) without acetylene were used as controls. Microcosms were incubated at 30 °C in the dark and destructive sampling was carried out after 10, 20 and 30 days of incubation (three replicate microcosms per treatment and incubation time). During the incubation, microcosms were opened and vented every 2 or 3 days to maintain aerobic conditions and acetylene partial pressures were re-established each time.

Soil was collected from the microcosms and immediately frozen and stored at −80 °C before chemical analysis and nucleic acid extractions. Three additional soil samples (10 g each) were stored at −80 °C at the beginning of the experiment, to characterize nitrogen and microbial characteristics. Ammonia and combined nitrite+nitrate concentrations were determined colorimetrically by flow injection analysis (FIA star 5010 Analyser, Tecator) (Allen, 1989) after extraction from 5 g of soil in 40 mL of 1 M KCl. Nucleic acids were extracted from 0.5 g of soil as described by Griffiths et al. (2000) and as modified by Nicol et al. (2005).

Denaturing gradient gel electrophoresis (DGGE) analysis

DGGE analysis of total crenarchaeal communities and of bacterial and archaeal ammonia oxidizers was performed as described previously (Nicol et al., 2008). Total crenarchaeal communities were analysed by amplification of 16S rRNA genes from extracted DNA using a nested PCR approach. Primary amplification was performed using primers A109f (Groβkopf et al., 1998) and a modified version (Nicol et al., 2008) of 1492r (Lane, 1991). Amplification products were used as template for secondary amplification using non-thermophilic crenarchaeal-specific primers 771f and 957r (Ochsenreiter et al., 2003) with primer 957r containing additionally the GC clamp of primer P3 (Muyzer et al., 1993). Communities of ammonia-oxidizing bacteria were characterized by DGGE analysis of 16S rRNA genes as described by Freitag et al. (2006). Bacterial ammonia oxidizer 16S rRNA genes were amplified with CTO189f and CTO654r PCR primers (Kowalchuk et al., 1997) and amplicons were nested in a second round of PCR amplification with P3 (357f-GC) and P2 (518r) (Muyzer et al., 1993). DGGE analysis of ammonia-oxidizing archaea was performed by amplification of amoA genes using primers CrenamoA23f/CrenamoA616r (Tourna et al., 2008) and do not require a GC clamp. Cycling conditions for all PCRs were 95 °C for 5 min; followed by 10 cycles of 94 °C for 30 s, 55 °C for 30 s, 72 °C for 1 min; followed by 25 cycles of 92 °C for 30 s, 55 °C for 30 s, 72 °C for 1 min; followed by 10 min at 72 °C, except for primer set A109f/1492r, which used an extension time of 2 min.

DGGE was performed using a DCode Universal Mutation Detection System (Bio-Rad, Hertfordshire, UK) as described previously (Nicol et al., 2005). Gels contained 8% (w/v) polyacrylamide and a linear gradient of 35–70% denaturant for crenarchaeal 16S rRNA gene assays, 35–65% for bacterial ammonia oxidizer 16S rRNA gene assays and 15–55% for crenarchaeal amoA assays. Gels were electrophoresed in 6.5 L of 1 × TAE buffer at a constant temperature of 60 °C for 900 min at 100 V and silver stained as described previously (Nicol et al., 2005) before scanning using an Epson GT9600 scanner with transparency unit (Epson, Hemel Hempstead, UK).

Cloning and sequence analysis of archaeal 16S rRNA and amoA gene sequences

First-round archaeal 16S rRNA and amoA gene PCR products were cloned into pGEM-T Easy vector (Promega, Southampton, UK) and screened by DGGE as described previously (Nicol et al., 2005) to select those clones representative of bands of interest in community DGGE profiles. Sequenced clones of each band position were obtained from at least three libraries derived from different microcosms. Phylogenetic distance analysis of derived amoA amino acid sequences was performed using the Jones, Taylor and Thornton substitution model with site variation (invariable sites and eight variable γ rates) using phylip (Felsenstein, 2007). LogDet distance analysis of 16S rRNA gene sequences was performed using paup (Swofford, 1998) using invariable sites estimated from a maximum likelihood analysis. Additionally, maximum likelihood and parsimony analyses were performed for both data sets using phyml (Guindon & Gascuel, 2003) and phylip, respectively.

Quantitative PCR (qPCR)

Quantification of total bacterial amoA genes and crenarchaeal amoA gene sequences corresponding to DGGE band positions 1 and 2 was based on a SybrGreen I approach. Bacterial amoA genes were amplified with primers amoA-1F and amoA-2R (Rotthauwe et al., 1997) and a dilution series (102–107amoA gene copies) of Nitrosospira multiformis ATCC25196 genomic DNA was used as a standard with efficiency of 92% and r2 value of 0.989. Specific crenarchaeal amoA gene sequences were amplified with newly designed primers, CrenamoA1-165f (5′-AATATCGCAAACGTTGATGCTTGCA-3′) and CrenamoA1-390r (5′-TAAAGACATTCCTACCAGTAC-3′) for sequences corresponding to DGGE band position 1 and CrenamoA2-165f (5′-AATATCACAAACACTGATGTTAGTA-3′) and CrenamoA2-390r (5′-CAATGACATTCCACACATCAC-3′) for band position 2. The specificity of these two primer sets was assessed by testing the ability of the primers to amplify 23 crenarchaeal amoA gene fragments obtained from our clone libraries. Three of these crenarchaeal amoA gene fragments were representative of DGGE band position 1, three were representative of DGGE band position 2 and the remaining 17 were nontarget DGGE band positions representative of the diversity of the nontarget sequences in the clone libraries. Successful and efficient PCR amplification, as assessed by visualization of a PCR product of the expected size by agarose gel electrophoresis, was obtained only for the crenarchaeal amoA gene fragments targeted by the primers, i.e. those related to DGGE band position 1 were amplified by the primers CrenamoA1-165f/390r and those related to DGGE band position 2 by the primers CrenamoA2-165f/390r. Two clones containing the crenarchaeal amoA gene fragment corresponding to DGGE band positions 1 and 2 were amplified with vector primers SP6 and T7 and dilution series (102–107amoA gene copies) of these PCR products were used as standards. For both crenarchaeal amoA assays, a fragment of 226 bp was amplified with efficiencies of 82% and 100% and r2 value of 0.976 and 0.99, respectively. Each reaction was performed in a 25-μL volume containing 12.5 μL of 2 × QuantiFast SYBR® Green PCR Master Mix (Qiagen), 0.2 mg mL−1 BSA and 0.8 μM (bacterial amoA assay) or 0.9 μM (crenarchaeal amoA assays) of each primer. For analysis of soil DNA extracts, 10 ng of DNA was added per reaction. Amplification was performed in a DNA Engine Opticon 2 System (MJ Research) and cycling conditions were as described in the QuantiFast SYBR® Green PCR handbook. Melting curve analysis was performed at the end of each qPCR run to indicate amplification of specific products only, before confirmation by standard agarose gel electrophoresis. All qPCR data presented were from independent extractions from triplicate microcosms.

Statistical analysis

Nitrification process and qPCR data from triplicate microcosms were analysed using a general linear model of anova using minitab 15 (Minitab, PA).

Accession numbers

All sequences have been deposited in the GenBank database with accession numbers FJ971889FJ971897.


Soil nitrification kinetics

Nitrification was assessed during incubation of soil microcosms as changes in the concentrations of ammonium (Fig. 1a) and nitrite+nitrate (Fig. 1b). In control microcosms, ammonium concentrations (Fig. 1a) remained at low levels (<0.6 μg NH4+–N g−1 soil), without significant change (P>0.05), and the concentrations of nitrite+nitrate (up to 47 μg NO2–N/NO3–N g−1 soil) increased significantly (P<0.05) during incubation for 30 days. In contrast, concentrations of ammonium (Fig. 1a) in all microcosms containing acetylene increased significantly (P<0.05) and continuously (up to 27 μg NH4+–N g−1 soil), while concentrations of nitrite+nitrate (Fig. 1b) after incubation for 30 days were not significantly different (P>0.05) from the initial value (13 μg NO2–N/NO3–N g−1 soil). Concentrations of ammonium (Fig. 1a) and nitrite+nitrate (Fig. 1b) did not differ (P>0.05) significantly between microcosms with different acetylene concentrations at any of the sample times.

Figure 1.

 Ammonia (a) and nitrite+nitrate (b) concentrations in soil incubated at 30°C for 30 days at various acetylene partial pressures. Triplicate microcosms were destructively sampled at each time point and mean values plotted. Error bars represent SEM values.

Changes in ammonia oxidizer community structure

The structure of crenarchaeal and bacterial ammonia oxidizer communities was assessed by DGGE analysis of 16S rRNA and amoA gene fragments amplified from DNA extracted from soil collected at the start of the experiment (day 0) and after incubation of microcosms for 30 days (Fig. 2). Bacterial and archaeal DGGE profiles from microcosms with different acetylene concentrations were similar and therefore only the profiles obtained from microcosms treated with the lowest applied concentration of 10 Pa (0.01%) acetylene are presented. DGGE profiles from replicate microcosms were very similar for all DGGE assays, reflecting high reproducibility between different microcosms. DGGE fingerprints of bacterial ammonia oxidizer 16S rRNA genes did not differ significantly across all treatments (Fig. 2a). Successful and reproducible amplification of bacterial amoA genes could not be obtained with GC-clamped primers, as reported previously (Nicol et al., 2008), and therefore DGGE analysis of amoA-defined bacterial ammonia oxidizer communities was not performed. In contrast, DGGE profiles of crenarchaeal 16S rRNA and amoA genes in control microcosms changed during the 30-day incubation period (Fig. 2b and c). The main differences between the profiles recorded at days 0 and 30 were increases in the relative intensity of one band for the 16S rRNA gene assay (Fig. 2b) and of two bands for the amoA assay (Fig. 2c, bands 1 and 2) that were minor components in the initial profiles (day 0), but which increased in relative intensity at day 30. In contrast, DGGE profiles of the crenarchaeal communities (Fig. 2b) and those of the archaeal ammonia oxidizers (Fig. 2c) in microcosms incubated for 30 days with 0.01% acetylene were similar to those obtained at day 0.

Figure 2.

 DGGE analysis of bacterial ammonia oxidizer 16S rRNA genes (a), crenarchaeal 16S rRNA genes (b) and archaeal ammonia oxidizer amoA genes (c) in soil incubated at 30°C for 30 days with and without 10 Pa partial pressure of acetylene. Each lane represents a profile derived from an individual microcosm. Bands highlighted by black arrows are considered to increase in relative intensity during the 30-day incubation period.

Phylogenetic analysis

DGGE analysis established that the relative intensity of one crenarchaeal 16S rRNA gene band (Fig. 2b) as well as two archaeal amoA bands (Fig. 2c) increased during incubation of control microcosms, but not in microcosms containing acetylene. For both genes, phylogenetic analysis (Fig. 3) indicated that these sequences were derived from organisms placed within lineages associated with marine and subsurface environments (Group 1.1a lineage) and were distinct from phylogenetic lineages dominated by sequences from soil environments (Group 1.1b lineage). This would indicate that the 16S rRNA gene sequence group was probably derived from one of the same populations with a corresponding amoA gene that was shown to increase in relative intensity after incubation. For both genes, the sequences were shown to be placed in the same clades of sequences derived from the same soil demonstrated previously to increase in relative intensity after incubation at 30 °C (Tourna et al., 2008).

Figure 3.

 Phylogenetic analysis of cloned amoA and 16S rRNA gene sequences (highlighted in blue) representative of band positions shown to increase in relative intensity in DGGE profiles after incubation at 30°C. Selected reference sequences are representative of those from two lineages dominated by sequences from marine (Group 1.1a) and soil (Group 1.1b) environments plus a thermophilic ammonia-oxidizing archaeal lineage (ThAOA). These include sequences derived from organisms from the same soil in which DGGE profiles have previously been shown to change during identical incubation conditions (Tourna et al., 2008) (highlighted in red) and sequences from organisms for which both 16S rRNA and amoA gene sequences are known (in grey). (a) Jones, Taylor and Thornton distance analysis (invariable sites plus eight variable rates of substitution) of derived amoA amino acids sequences. The target group of two specific qPCR assays developed are also highlighted. (b) LogDet distance analysis of 16S rRNA genes (variable sites only). Phylogenetic analyses used unambiguously aligned homologous positions only and bootstrap support for both trees represent the lowest percentage of 1000, 1000 and 100 bootstrap replicates from distance, parsimony and maximum likelihood analyses, respectively.

Growth of ammonia oxidizer communities

Growth of ammonia oxidizers was assessed during incubation of soil microcosms by quantification of amoA genes (Fig. 4), using qPCR assays specific for betaproteobacterial amoA genes and for specific crenarchaeal amoA genes related to DGGE band positions 1 and 2 (Fig. 2c). qPCR amplification of total crenarchaeal amoA genes could not be obtained with efficiencies routinely >80% and specific qPCR assays with greater efficiencies were therefore developed for specific sequence groups. Abundance of betaproteobacterial amoA genes in control microcosms decreased significantly (anova, log-transformed data, P<0.05) from c. 4.9 × 105 to 3.4 × 105 copies g−1 soil during the first 20 days of incubation without further change at 30 days (Fig. 4a). Similarly, bacterial amoA gene abundance in microcosms containing acetylene decreased significantly from 4.9 × 105 to 2.9 × 105 copies g−1 soil after 10 days and did not change during the final 20 days of incubation. In contrast, abundance of crenarchaeal amoA genes, corresponding to DGGE bands 1 and 2, increased significantly (anova, log-transformed data, P<0.05) in control microcosms (up to 1.1 × 106 and 2.3 × 106 copies g−1 soil, respectively) during the 30-day incubation period but did not change in microcosms containing acetylene (Fig. 4b and c).

Figure 4.

 The effect of a 10 Pa acetylene (C2H2) partial pressure on the abundance of bacterial ammonia oxidizers (a) and of two populations of archaeal ammonia oxidizers (b and c) in soil microcosms incubated at 30° for 30 days. Crenarchaeal populations are related to DGGE band positions 1 (b) and 2 (c) in Fig. 2c. Quantification of amoA gene copies was performed with a SybrGreen I quantitative PCR approach. Three replicate microcosms were analysed for each time point. Error bars represent SEs.


The aims of this study were to determine whether bacterial and archaeal ammonia oxidizers grow during active nitrification and whether growth kinetics of ammonia oxidizers and soil nitrification are affected by acetylene. The experimental strategy consisted of monitoring, over 30 days, the composition and abundance of bacterial and archaeal ammonia oxidizer communities in soil microcosms using DGGE and qPCR analyses of 16S rRNA and amoA genes. Active nitrification occurred in control microcosms throughout the incubation period as shown by the steady production of nitrite+nitrate (Fig. 1b) and consistently low ammonium concentrations (Fig. 1a). As no ammonia was added to the microcosms, the nitrification process was presumed to be fuelled by ammonia generated by continuous nitrogen mineralization. DGGE analysis of archaeal amoA genes in control microcosms showed that the relative abundance of two crenarchaeal phylotypes increased during the experiment (Fig. 2c), suggesting that the populations belonging to these groups were growing. Growth of these populations was demonstrated by qPCR analysis of their amoA gene copy number (Fig. 4b and c), which increased significantly during incubation. In contrast, the composition (Fig. 2a) of the bacterial ammonia oxidizer community remained unchanged throughout the 30-day period and their abundance decreased slightly (Fig. 4a). Thus, growth of archaeal, but not bacterial, ammonia oxidizers occurred only during the nitrification activity, suggesting that ammonia oxidation was mainly due to archaea rather than bacteria, under the conditions of this study. Growth of both groups of amoA-containing archaea through ammonia oxidation is supported by Tourna et al. (2008) who showed, under identical conditions, that amoA genes belonging to these archaeal phylotypes (Fig. 3) are expressed during nitrification. Moreover, Tourna et al. (2008) did not detect bacterial amoA gene transcripts, which is consistent with low rates of bacterial ammonia oxidation.

Communities of bacterial ammonia oxidizers in microcosms containing acetylene did not vary (Fig. 2a) during incubation and their abundance decreased slightly (Fig. 4a). Partial pressures of acetylene ≥10 Pa inhibited the growth of archaeal populations that grew in control microcosms, as shown by qPCR analysis of their amoA genes (Fig. 4b and c). These data show that low concentrations of acetylene inhibit the growth in soil of amoA-containing archaea. Acetylene might suppress the growth of amoA-containing archaea by inactivating the archaeal AMO protein, as demonstrated in ammonia-oxidizing bacteria (Hyman & Wood, 1985; Hyman & Arp, 1992). However, as the enzymes and metabolic pathways of ammonia-oxidizing archaea have not yet been characterized (Hallam et al., 2006b; Prosser & Nicol, 2008), and might differ significantly from those of ammonia-oxidizing bacteria, the possibility that acetylene interferes with components other than the AMO protein in the archaea cannot be excluded. Future studies based on pure cultures are therefore necessary to elucidate the mechanism by which acetylene blocks the growth of archaeal ammonia oxidizers.

Acetylene at partial pressure ≥10 Pa completely blocked the production of nitrite+nitrate recorded in control microcosms (Fig. 1b), as reported previously for other soils (Berg et al., 1982; Bateman & Baggs, 2005), and led to an increase in ammonium concentration, through mineralization of organic matter (Fig. 1a). Communities of bacterial ammonia oxidizers exposed to acetylene and those in control microcosms had similar compositions (Fig. 2a) and abundances (Fig. 4a) showing that ammonia-oxidizing bacteria in control microcosms were not actively growing during nitrification and therefore probably contributed little to the nitrification process. In contrast, growth of amoA-containing archaea was observed during active nitrification, but not when nitrification was blocked by acetylene, providing further evidence that most of the ammonia oxidation in this soil was carried out by archaea.

These results contrast with those of Jia & Conrad (2009), who determined the influence of acetylene on amoA gene abundance and ammonia and nitrate production in N-fertilized agricultural soils. Changes in abundance of bacterial, rather than archaeal amoA genes correlated best with nitrification rate, with bacterial growth occurring only in actively nitrifying microcosms. Intriguingly, in microcosms containing acetylene at 100 Pa (a concentration one order of magnitude greater than that used in this study), archaeal populations still appeared to grow, even when there was no nitrifying activity. The Group 1.1a crenarchaea showing nitrification-related growth in the Craibstone soil belong to a lineage quite distinct from the Group 1.1b archaea growing in the Jia & Conrad (2009) study. It is possible that Groups 1.1a and 1.1b soil crenarchaea have distinct physiologies, with Group 1.1b perhaps displaying the capability for mixotrophic growth or with a different function for AMO. Growth of Group 1.1b soil crenarchaea in the absence of nitrification might otherwise result from the ability to use energy and carbon storage compounds similarly to that of the ammonia-oxidizing bacteria N. multiformis (Norton et al., 2008) and as suggested by Hatzenpichler et al. (2008) for the thermophilic Group 1.1b crenarchaeote Nitrososphaera gargensis. In addition, the concentrations of ammonia in the two studies were different. Archaeal growth was demonstrated in soil microcosms with <1 μg N g−1 soil whereas bacterial growth dominated in microcosms supplemented at least once with ammonium to a final concentration of 100 μg N g−1 soil. Ammonia concentration is known to select for different groups of bacterial ammonia oxidizers, through variation in tolerance to high ammonia levels (Webster et al., 2005), and Valentine (2007) proposed that archaeal ammonia oxidizers might be better adapted to low-energy flux resulting from reduced concentrations of ammonia and/or oxygen (Hoehler, 2004), potentially outcompeting bacterial ammonia oxidizer in conditions of low-energy supply. Indeed, the highest level of ammonia oxidation achieved by the archaeon N. gargensis (Hatzenpichler et al., 2008) was recently shown to occur below 3.08 mM ammonium concentration, while the lowest ammonium concentration known to inhibit the growth of a bacterial ammonia oxidizer is 21.4 mM (Suwa et al., 1994; Koops & Pommerening-Röser, 2001). Nevertheless, it is not currently known whether archaeal ammonia oxidizers in soils are sensitive to high concentrations of ammonium, and an alternative explanation may be differences in interaction with nitrogen mineralizers (Schimel & Bennett, 2004), as suggested by the increased abundance of archaeal but not bacterial ammonia oxidizers in the rice rhizosphere compared with the bulk soil (Chen et al., 2008).

Phylogenetic analysis of amoA gene sequences (Fig. 3a) placed the archaeal ammonia oxidizers growing during nitrification within well-supported monophyletic groups also containing sequences retrieved from soils. Nevertheless, these monophyletic branches were placed within a lineage dominated by amoA gene sequences retrieved from the marine environment (Fig. 3a), including sequences of the cultivated ammonia oxidizer Nitrosopumilus maritimus (Könneke et al., 2005) and the sponge symbiont Cenarchaeum symbiosum (Hallam et al., 2006a), both belonging to the crenarchaeal Group 1.1a (Prosser & Nicol, 2008). A similar result was obtained for the phylogenetic analyses (Fig. 3b) of clone sequences representative of the crenarchaeal 16S rRNA gene DGGE band that increased in relative abundance during active nitrification (Fig. 2b), suggesting that both populations of archaeal ammonia oxidizers, characterized by different amoA genes, might have closely related 16S rRNA gene sequences. Previous studies of several soils (Buckley et al., 1998; Ochsenreiter et al., 2003), including that of Craibstone (Nicol et al., 2008; Tourna et al., 2008), have shown that most crenarchaeal 16S rRNA and amoA gene sequences in soils are placed within the Group 1.1b lineage. This suggests that ammonia oxidation in the conditions of the present study resulted from the activity of archaeal groups that are less common in soil and unfortunately provides little insight into the functional activity of the numerically dominant AMO-possessing lineage.

In conclusion, this study demonstrated growth of archaeal ammonia oxidizers during active nitrification in soil microcosms. Inhibition of nitrification by acetylene resulted in suppression of their growth, indicating that these archaeal groups are ammonia oxidizers. This is further supported by a previous study of the same soil (Tourna et al., 2008), showing, under identical conditions, transcription of amoA genes belonging to these archaeal groups during nitrification. Bacterial ammonia oxidizers did not grow during nitrification, providing compelling evidence that ammonia oxidation was mostly due to archaea. Phylogenetic analysis of archaeal 16S rRNA and amoA gene sequences placed groups of archaeal ammonia oxidizers that grew during nitrification within a monophyletic branch, which does not contain the most abundant lineage of soil archaeal ammonia oxidizers (Prosser & Nicol, 2008) and which accounts for a small proportion of their diversity (Nicol et al., 2008; Tourna et al., 2008). Together, these results raise the question of the ecological function of the most abundant amoA-containing soil archaea in soil and of the functional diversity of archaeal ammonia oxidizers. They also highlight the need to characterize the environmental factors determining growth and activity of different ammonia-oxidizing archaeal lineages and to assess whether archaeal and bacterial ammonia oxidizers occupy distinctive niches or compete for ammonia and, if so, to determine the trade-offs that allow their coexistence in soils.


The authors would like to thank Mr Lawrence Maurice and the SAC Craibstone Estate (Aberdeen) for access to the Woodlands Field pH plots. P.O. is funded by a Marie-Curie Intra-European Fellowship (PIEF-GA-2008-220639) under the 7th Framework Program of the European Union and G.W.N. by an Advanced Fellowship (NE/D010195/1) from the UK Natural Environment Research Council (NERC).