Ammonia-oxidizing archaea (AOA) outnumber ammonia-oxidizing bacteria (AOB) in many terrestrial and aquatic environments. Although nitrification is the primary function of aquarium biofilters, very few studies have investigated the microorganisms responsible for this process in aquaria. This study used quantitative real-time PCR (qPCR) to quantify the ammonia monooxygenase (amoA) and 16S rRNA genes of Bacteria and Thaumarchaeota in freshwater aquarium biofilters, in addition to assessing the diversity of AOA amoA genes by denaturing gradient gel electrophoresis (DGGE) and clone libraries. AOA were numerically dominant in 23 of 27 freshwater biofilters, and in 12 of these biofilters AOA contributed all detectable amoA genes. Eight saltwater aquaria and two commercial aquarium nitrifier supplements were included for comparison. Both thaumarchaeal and bacterial amoA genes were detected in all saltwater samples, with AOA genes outnumbering AOB genes in five of eight biofilters. Bacterial amoA genes were abundant in both supplements, but thaumarchaeal amoA and 16S rRNA genes could not be detected. For freshwater aquaria, the proportion of amoA genes from AOA relative to AOB was inversely correlated with ammonium concentration. DGGE of AOA amoA genes revealed variable diversity across samples, with nonmetric multidimensional scaling (NMDS) indicating separation of freshwater and saltwater fingerprints. Composite clone libraries of AOA amoA genes revealed distinct freshwater and saltwater clusters, as well as mixed clusters containing both freshwater and saltwater amoA gene sequences. These results reveal insight into commonplace residential biofilters and suggest that aquarium biofilters may represent valuable biofilm microcosms for future studies of AOA ecology.
Citation: Sauder LA, Engel K, Stearns JC, Masella AP, Pawliszyn R, Neufeld JD (2011) Aquarium Nitrification Revisited: Thaumarchaeota Are the Dominant Ammonia Oxidizers in Freshwater Aquarium Biofilters. PLoS ONE 6(8): e23281. https://doi.org/10.1371/journal.pone.0023281
Editor: Jack Anthony Gilbert, Argonne National Laboratory, United States of America
Received: June 30, 2011; Accepted: July 11, 2011; Published: August 16, 2011
Copyright: © 2011 Sauder et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This research was funded by the Natural Sciences and Engineering Research Council of Canada (NSERC). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Ammonia is a toxic metabolic waste product excreted by fish and other aquatic organisms. Ammonia toxicity can threaten aquatic ecosystem health and is a particular concern for relatively closed ecosystems, such as aquaculture operations and home aquaria, in which ammonia can quickly accumulate to lethal concentrations in the absence of active nitrification. The un-ionized form of ammonia (NH3) is particularly toxic to fish; stress, disease, and death may be associated with concentrations that exceed 0.1 mg L−1 in aquarium and aquaculture systems , . In order to convert ammonia to nitrate, aquarium biofilters are designed to promote the growth and activity of nitrifying populations due to the high surface area of filter support material (e.g. sponge, ceramic or polymer) and rapid flow rates of aerated water. Despite their importance to fish health and identical function within many industrial biofilters, including aquaculture and wastewater treatment, little is known of the microorganisms catalyzing nitrification in association with aquarium biofilter support material.
Before the discovery of ammonia-oxidizing archaea (AOA), belonging to the newly proposed phylum Thaumarchaeota , , molecular approaches were used to investigate ammonia-oxidizing bacteria (AOB) and nitrite-oxidizing bacteria in freshwater and marine aquaria , . In particular, Hovanec and DeLong used oligonucleotide probes to target bacterial nitrifiers in freshwater and saltwater aquarium biofilter DNA extracts. Although Nitrosomonas-like bacteria from the Betaproteobacteria were associated with the saltwater aquaria in their study, they did not detect these bacteria in most of the freshwater aquarium biofilter extracts tested. They concluded that “the bacterial species responsible for nitrification in simple freshwater systems remain unknown” . Subsequent studies determined that Nitrosomonas spp. could indeed be enriched from freshwater aquarium biofilters , suggesting their potential involvement in ammonia oxidation under in situ conditions. Since the discovery of AOA, two studies have investigated the presence and diversity of both AOA and AOB in marine biofiltration systems , . Foesel and colleagues found that Nitrosomonas-like AOB were numerically dominant in a marine aquaculture biofilm. Urakawa and colleagues identified the presence of amoA genes from AOB and AOA in selected marine aquarium biofilters from a public aquarium in Japan (i.e. sunfish tank, cold water tank, and a coastal fish tank), and suggested that the diversity of AOA and AOB was decreased in low temperature marine aquaria. Candidatus Nitrosopumilus maritimus SCM1, the first AOA representative isolated in pure culture, was obtained from saltwater aquarium gravel . Despite these initial studies, no research has yet investigated the abundance of AOB and AOA in freshwater aquarium biofilters.
Based on the ubiquity and high abundance of AOA in natural environments, the inability of Hovanec and DeLong (1996) to detect AOB in freshwater aquaria, and the isolation of the first ammonia-oxidizing archaeon from aquarium substrate , we hypothesized that AOA dominate freshwater aquarium biofilters and play an important role in aquarium nitrification. In addition to determining the abundances of AOA and AOB in aquaria, the objectives of this study were to assess the diversity of AOA amoA genes in aquaria, and to determine how these genes clustered with sequences derived from environmental sources and cultured AOA representatives. The results of this study revealed that, based on amoA gene abundances, AOA were the dominant putative ammonia oxidizers in the majority of freshwater and saltwater aquaria. These results provide first evidence for the important role of AOA in freshwater aquarium filtration and suggest possible niche adaptation of AOA to conditions associated with freshwater aquarium biofilters.
Twenty-seven freshwater and eight saltwater aquarium filter samples were collected from retail and residential locations in three cities (Table S1). All biofilters sampled in this study were derived from standalone aquaria in homes (or offices), or from display tanks in retail outlets, reflecting conditions common to most residential or retail aquaria, respectively. In addition to aquarium biofilters, we included two aquarium supplements in the analysis. The aquaria selected for sampling ranged in pH from 7.6 to 9.2 and varied in their fish and live plant composition. The aquaria contained a variety of fish including mixed tropical, goldfish, South American cichlids and African cichlids. Three aquaria had received antibiotic treatment in the previous six months (SW4, SW5, FW8) and several were known to have received doses of bacterial filter supplement when first established (e.g. FW12, FW13, FW19, FW25). Ammonium concentrations of aquaria ranged from below detection to approximately 0.5 mg L−1, with the majority of aquaria below 100 µg L−1. In 28 of the 32 aquaria studied, nitrite (NO2−) was below detection. As expected, significant positive correlations were observed between ammonium and nitrite concentrations (r = 0.48, p<0.05; Table S2) and nitrite and nitrate concentrations (r = 0.52, p<0.05; Table S2). Aquaria ranged in size from 5 gallons to greater than 400 gallons (for large retail show tanks) and approximate fish numbers ranged from zero (in a plant tank; FW3) up to 300 (FW11). Although not a perfect measure of fish biomass, the approximate number of fish per gallon was positively correlated with ammonium concentration (r = 0.60, p<0.001; Table S2). Water hardness of freshwater aquaria was as low as 25 ppm (in FW12, a breeding tank using softened water), however, the water hardness of 25 of 27 aquaria was >150 ppm. Hardness of saltwater samples could not be determined using the kit utilized. Alkalinity (i.e. carbonate alkalinity/hardness) ranged from below detection (e.g. FW24, FW17) to 300 ppm (e.g. SW1, FW3). Neither hardness nor alkalinity correlated significantly with any other water chemistry parameters (Tables S2 and S3).
AOA and AOB abundances
Real-time PCR results demonstrated that thaumarchaeal amoA genes were dominant in 23 of the 27 sampled freshwater filters (Fig. 1; Table S1). For 12 of the freshwater biofilters, thaumarchaeal amoA genes represented the entire detected amoA gene signal, including FW13, FW15, FW16, and FW25, aquaria that received bacterial aquarium supplements when first established. For saltwater aquaria, amoA genes from both AOA and AOB were detected in all samples, with AOA dominating five of eight samples. For both commercially available aquarium supplements, AOB amoA genes were abundant, and both AOA amoA and 16S rRNA genes were below detection limits.
Relative 16S rRNA (A) and amoA (B) gene abundances for bacteria and thaumarchaea in saltwater (SW1–SW8) and freshwater (FW1–F27) aquaria and in aquarium supplements (SP1, SP2). These data were calculated from duplicate qPCR amplifications.
Both bacterial and thaumarchaeal 16S rRNA genes were detected in all aquarium DNA extracts; however, for the majority of samples, bacterial 16S rRNA genes greatly outnumbered thaumarchaeal 16S rRNA genes (Fig. 1; Table S1). In one aquarium biofilter (FW11), thaumarchaeal and bacterial 16S rRNA gene copy numbers were approximately equal. For all aquaria, bacterial amoA gene copy numbers were at least three orders of magnitude less than bacterial 16S rRNA genes (Table S1). In some aquaria (e.g. FW2, FW27, SW4), thaumarchaeal amoA and 16S rRNA gene copy numbers were approximately equal. In other cases (e.g. FW11, FW12, SW8), thaumarchaeal amoA gene copies were orders of magnitude less abundant than thaumarchaeal 16S rRNA genes.
We also examined aspects of water chemistry (Table S1) to assess correlations that might provide an explanation for differential amoA gene abundances. Regression analyses focused on freshwater aquarium samples because these were the primary focus of this study, and because too few saltwater samples were available to yield statistically significant correlations (data not shown). Correlations were also calculated that included both fresh and saltwater samples, which yielded similar results to freshwater correlations (Tables S2 and S3). For freshwater aquaria, we observed a significant negative correlation between ammonium concentrations and the proportion of amoA genes belonging to AOA rather than AOB (r = −0.85, p<0.001, R2 = 0.72; Fig. 2 and Table S3). Low ammonia concentrations were typically associated with high AOA amoA gene abundances, although one sample (FW20) had high proportions of AOB amoA genes, despite an ammonia concentration of less than 20 µg L−1. In all cases, higher concentrations of ammonium were associated with higher relative abundances of AOB amoA genes. No other factors related to water chemistry or aquarium setup (e.g. pH, hardness, alkalinity) yielded significant correlations with amoA gene abundances (Table S2). Ammonia concentrations in aquarium FW27 fluctuated very little, both on daily and monthly scales (Fig. S1A). The high proportion (>85%) of AOA in this filter was also consistent over a two-year sampling period (Fig. S1B). Together, these results suggest temporal consistency in individual freshwater aquarium filter environmental conditions and AOA communities.
AOA amoA gene copies are expressed as a percentage of the total amoA gene copies (per ng DNA). The coefficient of determination (R2) for the linear regression is 0.7249. The associated Pearson correlation coefficient (r) is −0.8518, with an associated p-value of <0.001. See Table S1 for all sample data.
AOA gene diversity
In order to assess the sample-to-sample variability of thaumarchaeal populations possessing amoA genes, all biofilter DNA extracts were subjected to DGGE analysis. Based on the physical separation of PCR products by sequence heterogeneity and G+C content , the patterns generated by amoA gene amplicons varied between samples from very simple with few bands (e.g. FW8, FW20, SW4) to relatively complex with many bands (e.g. FW10, FW15, SW1; Fig. 3). DGGE patterns revealed shared bands between aquarium biofilters from the same location (e.g. FW8 and FW9, FW25 and FW26, and SW1 and SW2), indicating that location-specific factors likely influenced the specific composition of AOA. Despite these location-specific similarities, distinct clustering of fingerprints was not observed based on location, and many bands were shared between multiple samples from various locations (Fig. 3A and B, Table S1). Based on visual inspection of aligned fingerprints, a possible shift in G+C content was observed between salt- and freshwater fingerprints, with saltwater bands typically melting at lower denaturant concentrations than the majority of freshwater sequences (Fig. 3A and B). This apparent difference in G+C content was supported by sequences from both clone libraries and sequenced DGGE bands. Freshwater amoA gene sequences from clone libraries had an average G+C content of 45.4%, versus 43.8% in saltwater clones. Sequences of DGGE bands showed a similar pattern, with average G+C contents of 45.5% and 44.0% for freshwater and saltwater bands, respectively. Although relatively small (∼1.5%), this difference in average G+C content between freshwater and saltwater sequences was statistically significant (p<0.0001) for both clones and DGGE sequences, as determined by unpaired t tests. Nonmetric multidimensional scaling (NMDS) using densitometric curves generated from DGGE fingerprints revealed that freshwater and saltwater fingerprints were largely separated in two-dimensional space. However, some freshwater samples (e.g. FW1, FW12, FW15) were intermediate to freshwater and saltwater clusters (Fig. 3C).
Saltwater (A) and freshwater (B) fingerprints have been normalized and aligned. Bands chosen for sequencing are indicated with triangles: white triangles correspond to bands appearing in Figure 4. Black triangles represent failed sequencing reactions. Clustering of freshwater and saltwater fingerprints (C) is based on nonmetric multidimensional scaling (NMDS) using Pearson correlations of background-subtracted densitometric curves.
In addition to DGGE, we generated clone libraries of 261 and 84 sequences for freshwater and saltwater composite thaumarchaeal amoA gene PCR amplicons, respectively. Multidimensional scaling of translated and aligned amoA gene sequences (from DGGE bands, clone libraries, and reference sequences) revealed four distinct clusters, each of which contained both clones and DGGE band sequences (Fig. 4). The first cluster (cluster 1; Fig. 4) contained saltwater and freshwater clone library sequences in approximately equal proportions, as well as both saltwater and freshwater DGGE band sequences. The freshwater DGGE band sequences that fell into this cluster were generated from samples which had fingerprints that fell near the periphery of the freshwater cluster, close to saltwater sequences (Fig. 3C).
Sequences were obtained from clone libraries and DGGE bands derived from freshwater and saltwater aquarium filter samples. The inset panel provides a summary of the number of freshwater and saltwater clone library sequences contained within each cluster. Sequences obtained from DGGE bands correspond to the white triangles in Figure 3, and are labelled with sample and band numbers. Selected Genbank sequences from uncultivated clones and reference strains are included for comparison. The sampling environment and Genbank accession numbers for all reference sequences are included within parentheses.
Both the second and third cluster (clusters 2 and 3; Fig. 4) were dominated by saltwater clone sequences. Cluster 2 was a distinctly saline cluster that contained amoA gene sequences from AOA originating from saline environments, including Ca. N. maritimus and Candidatus Cenarchaeum symbiosum A [YP_875342; 12], as well as an environmental clone obtained from a saltwater aquarium biofilter [AB373235; 9]. The majority of saltwater DGGE band sequences also fell into this cluster, including SW2-3, SW1-2, SW8-2 and SW6-2, sequences that represent a band shared across the majority of saltwater fingerprints (Fig. 3A). Cluster 3 was variable in composition, and contained both freshwater and saltwater DGGE band sequences. Reference amoA gene sequences in this cluster were from soil fosmid 54d9 [AJ627422; 13] and hotspring organisms Candidatus Nitrososphaera gargensis [ABY77595; 14] and Candidatus Nitrosocaldus yellowstonii [ABY83788; 15].
Cluster 4 was a distinct freshwater cluster; it was the largest of the clusters and contained the vast majority of the freshwater clone library and DGGE band sequences generated in this study. This cluster contained all sequences generated from a particular DGGE band that was shared over approximately half of the freshwater fingerprints (i.e. FW2-4, 6-11, 13, 14, 17, 19, 27; Fig. 3B). In addition, reference sequences from a variety of low-salinity environments fall into this cluster, including rivers, lakes, sediments, and wastewater, as well as the recently enriched Candidatus Nitrosoarchaeum limnia .
The present study has generated data that challenge decades of common knowledge regarding nitrogen cycling within aquarium filtration systems and solves outstanding questions remaining since Nitrosomonas spp. were undetected in freshwater aquaria . From the sampling of 27 standalone freshwater aquaria from homes and retail outlets, our qPCR results have revealed a dominant AOA population in most of the freshwater filters sampled (Fig. 1). Indeed, 12 of the aquarium filters were associated only with an AOA amoA gene signal, despite using 40 cycles for the PCR. These results are important because they provide the first qualitative evidence that AOA may act alone in catalyzing ammonia oxidation. Though also generally dominated by AOA amoA, saltwater aquaria were more variable in the relative abundances of AOA and AOB, with both groups detected in each biofilter. Based on amoA gene abundances, the results of this study suggest that both AOA and AOB contribute to nitrification in saltwater aquaria and that AOA are the dominant contributors to nitrification in most established freshwater aquaria.
The results of this study also answer longstanding questions related to nitrification in aquarium biofilter environments. For example, the discovery that abundance of bacterial amoA genes may be orders of magnitude less than bacterial 16S rRNA genes (Table S1) provides an explanation for previous studies that were unable to detect Nitrosomonas spp. with gene probes , despite the ability to detect Nitrosomonas spp. with PCR primers for most samples . Probe hybridization methods have limited detection for genes that represent less than 1% of the total community , whereas the high sensitivity of PCR can allow even a single copy of a gene to be detected if the amplification is not inhibited. Relative proportions of AOB in saltwater aquaria were generally greater than in freshwater biofilters (Fig. 1), which may explain the ability of both probe hybridization and PCR amplification to detect Nitrosomonas spp. in saltwater aquarium biofilters , , .
For a variety of genes in Bacteria, including both 16S rRNA and amoA genes, copy numbers within a cell are variable and often greater than one, and can therefore not be taken as a direct indication of population sizes. Conversely, available genomes from AOA representatives, including Ca. N. maritimus  and Ca. C. symbiosum , suggest that AOA cells contain one gene copy of each amoA and 16S rRNA. Whether all Thaumarchaeota possess ammonia monooxygenase genes is unknown. However, thaumarchaeal 16S rRNA genes have been detected that are up to 100 times more abundant than archaeal amoA genes , suggesting that some thaumarchaeal lineages do not gain energy by oxidizing ammonia. Some aquarium biofilters examined in this study (e.g. FW2, FW27, SW4; Table S1) yielded amoA gene copy numbers approximately equal to thaumarchaeal 16S rRNA copy numbers, implying that all Thaumarchaeota present in these aquaria possess amoA genes and presumably oxidize ammonia. Interestingly, other aquaria contained 16S rRNA genes that were orders of magnitude higher than AOA amoA genes (e.g. FW11, FW12, SW8; Table S1), which may imply to existence of thaumarchaeal lineages that do not oxidize ammonia.
AOA amoA gene diversity was variable across the filters collected in this study (Fig. 3A and B), with DGGE patterns ranging from relatively simple with few bands to complex with greater than ten discernible bands. In addition, many bands were shared across multiple aquarium filters from a variety of locations. Multidimensional scaling of DGGE fingerprints revealed that freshwater and saltwater fingerprints were largely separated in two-dimensional space. However, fingerprints of some freshwater samples were intermediate to major saltwater and freshwater clusters (Fig. 3C). DGGE band sequences derived from these intermediate fingerprints fell into a cluster of sequences containing approximately equal proportions of freshwater and saltwater clones (cluster 1, Fig. 4). Visual inspection of DGGE profiles (Fig. 3A and B) suggested a possible shift in amoA gene G+C content between freshwater and saltwater samples, and this was likely a factor in the separation observed. This trend was supported by clone libraries, which indicated that freshwater AOA amoA sequences had a higher average G+C content than their marine counterparts. The ecological implications of this finding are yet to be determined, but this poses an interesting hypothesis for future research. Multidimensional scaling of thaumarchaeal amoA clone library sequences (Fig. 4) revealed that freshwater and saltwater AOA amoA gene sequences largely cluster in distinct groups, which supports the separation observed in ordination analysis of DGGE profiles. Clusters were observed that predominantly contained saltwater sequences (i.e. clusters 2 and 3; Fig. 4), while one cluster (cluster 1; Fig. 4) contained approximately equal proportions of freshwater and saltwater sequences. Based on multidimensional scaling of both AOA amoA DGGE profiles and gene sequences, salinity appears to be a major factor in AOA differentiation. Nonetheless, these results suggest that at least some sequences are similar between environments, an observation that may indicate halotolerance of some phylotypes. Despite incomplete separation of saltwater and freshwater AOA amoA gene sequences, the majority (>80%) of freshwater sequences derived from clone libraries and DGGE bands clustered with sequences derived from a variety of freshwater environments, including lake sediments, wastewater, paddy soil, lakes and rivers (Fig. 4). This clustering confirms previous evidence for niche adaptation of aquatic freshwater AOA [e.g. 20],  and suggests that aquaria might provide valuable microcosms for investigating the ecology of aquatic AOA.
Although thaumarchaeal amoA gene copies may be several thousand fold more abundant than betaproteobacterial amoA genes in some marine  and terrestrial  environments, the relative contributions of AOA and AOB to environmental ammonia oxidation and the factors that affect their activity have been difficult to confirm. The concentration of ammonia may be a major factor affecting bacterial and thaumarchaeal ammonia oxidation. Recent studies have provided evidence that AOB are the dominant ammonia-oxidizing organisms in ammonia-rich soil and aquatic environments. For example, a recent study used stable-isotope labelled carbon dioxide and supplemented ammonia in soil to demonstrate that the labelled carbon was assimilated primarily into the nucleic acid of bacterial nitrifiers . Foesel and colleagues (2008) found that Nitrosomonas-like AOB were numerically dominant in a marine aquaculture biofiltration system receiving high-ammonia influent (ranging from 340–1700 µg L−1). Further, the metabolic and numerical dominance of bacterial ammonia oxidizers in the presence of high ammonium concentrations is consistent with a previous study  investigating the eventual establishment of bacterial ammonia oxidizer colonization of ammonia-supplemented aquarium biofilters (5–60 mg NH3 L−1). In addition, most tested activated sludge bioreactors with high influent ammonia concentrations have been dominated by AOB populations , .
Predominance of AOA in low ammonia conditions has now been relatively well established in soil environments , , . However, the role of ammonia in regulating ammonia-oxidizing populations in aquatic environments has not yet been well studied. That AOA are better adapted to low ammonia environments is supported by the present study with aquarium biofilters, where ammonium concentrations are maintained at consistently low concentrations, and certainly well below 1 mg N L−1 in all sampled aquaria (Table S1; Fig. 3). The majority of freshwater aquaria with very low ammonium concentrations (<100 µg L−1) were associated with increased proportions of AOA, and a significant inverse correlation was identified (Fig. 2 and Table S3). For statistical analyses, inclusion of additional samples with intermediate to high ammonia concentrations would have been preferable; however, established and maintained aquaria are typically kept at low ammonia concentrations to ensure fish health, and as a result we were unable to locate additional ammonia-rich aquaria within the timeframe of this study. The results of this study suggest that ammonia concentration in freshwater environments is an important parameter for determining the relative abundance of AOA and AOB. These results are consistent with other studies that have retrieved AOA amoA genes from low ammonia aquatic environments and demonstrated corresponding activity , , . In addition, growth kinetics of Ca. N. maritimus str. SCM1 demonstrated a half saturation constant (Km) for ammonia that is substantially lower than for cultured AOB representatives  and similar to the measured ammonium concentrations associated with most of the aquaria sampled in this study (i.e. low µg L−1 ranges). However, because Ca. N. maritimus is the only AOA representative for which kinetic studies have been reported, it remains unclear whether all AOA are similarly adapted to oligotrophic conditions and demonstrate high substrate affinities for ammonia. For example, the recently isolated Nitrososphaera viennensis tolerates ammonium concentrations of up to 20 mM , which is considerably higher than the inhibitory concentrations of 2 to 3 mM that have been reported for Ca. N. maritimus and Ca. N. gargensis , .
Although the detectable ammonia concentrations in established freshwater aquaria are typically low as a result of biological ammonia oxidation, a preference for high ammonia concentrations by AOB suggests a possible role for their involvement in first establishing an aquarium when ammonia concentrations may approach levels associated with fish toxicity. In addition, ammonium concentration was positively and significantly correlated with the number of fish per gallon of aquarium water (Tables S2 and S3), suggesting that AOB may also be important for heavily stocked tanks that experience chronic high ammonia concentrations.
This study has identified that AOA are the dominant ammonia oxidizing microorganisms in freshwater aquarium biofilters. Aquarium ammonium concentrations were significantly and inversely correlated with AOA∶AOB ratios. Freshwater aquarium AOA amoA gene sequences largely clustered with other freshwater-associated sequences. This work provides a foundation for future studies of aquarium nitrification and AOA ecology. Aquaria may serve as valuable microcosms to investigate the factors affecting AOA and AOB dynamics in both natural and engineered aquatic communities, including wastewater treatment systems, aquaculture, lakes, rivers and oceans.
Materials and Methods
Freshwater and saltwater aquarium biofilters were sampled from retail aquarium outlets and homes in Waterloo, Kitchener and Cambridge (Ontario, Canada) between June and December 2009 (Table S1). A total of 27 freshwater and 8 saltwater filters were analyzed in this study; filter types included sponge, floss, baffle, and live rock, and all filters collected were composed of cotton or synthetic polymeric material (e.g. nylon). Filter samples were collected using flamed forceps and scissors to cut small slices (e.g. 1 cm×1 cm×3 cm) from sponge material in external aquarium filtration systems. All filter samples were placed into 50-ml sterile tubes and stored on ice until returned to the laboratory within a few hours. Aquarium water samples were collected in 50-ml sterile tubes and stored on ice before being frozen at −80°C. Note that additional samples were collected for this study but were not included either due to poor yields of nucleic acid from the filter material or lack of amplification for both thaumarchaeal and bacterial amoA genes. A filter from one aquarium (FW27) was sampled four times over the course of two years to assess temporal stability in AOA/AOB ratios, and water was sampled several times (over six months and hourly over one day) to assess stability in ammonium concentrations within a given aquarium. The pH was assessed for all water samples with a DELTA 320 pH meter (Mettler Toledo, Columbus, OH). Ammonium concentrations were assessed fluorometrically, according to a previously published method  using a TD 700 fluorometer (Turner Designs, Sunnyvale, CA) and calculated from linear standard curves. Other water chemistry parameters were assessed with a simple water test kit available commercially (Quick Dip Aquarium Multi-Test Kit, Jungle Laboratories Corporation, Cibolo, TX). All sample information is contained within Table S1. We also sampled from eight saltwater aquarium filters and two aquarium supplements for comparison (Table S1). Bacterial supplements (typically bottled liquid suspensions) are intended to aid in populating newly established aquaria with active nitrifying bacteria, to help ensure that ammonia and nitrite concentrations remain below toxic levels during the initial 1–2 months of aquarium filter colonization. The aquarium supplements included were Cycle (SP1; Rolf C. Hagen Inc., Montreal, Canada), and Bio-Support (SP2; Big Al's Distribution Centre, Niagara Falls, NY).
A harsh nucleic acid extraction technique  was adapted to extract nucleic acids from thawed sponge filter material that had been cut into small fragments with flame-sterilized scissors. For supplements, aliquots (15 mL) of liquid aquarium supplements were pelleted by centrifugation at 7,000×g for 30 min, then suspended in lysis buffer for extraction. A beadbeating extraction was performed according to the published protocol with minor modifications. Nucleic acids were extracted from an equal volume of filter material rather than equivalent weight due to the variation in filter media porosity. The porous nature of filter material also required the phenol-chloroform-CTAB extraction buffer to be decanted away from the sponge material after cell lysis, prior to centrifugation and separation of aqueous and organic phases. To precipitate purified nucleic acids, two volumes of polyethylene glycol (PEG) solution (30% PEG 6000 and 1.6 M NaCl) were used in combination with linear polyacrylamide (AppliChem, Darmstadt, Germany) as a co-precipitant to avoid introducing exogenous DNA detected in commercial supplies of glycogen . All extracts were separated on a 1% agarose gel with Gel Red nucleic acid stain (Biotium, Hayward, CA), visualized with an AlphaImager HP (Alpha Innotech Corporation, Santa Clara, CA) and quantified densitometrically by comparison to dilutions of known quantities of lamba DNA (New England Biolabs, Pickering, Canada) using AlphaView software (Alpha Innotech Corporation).
Quantitative real-time PCR
Quantification of AOA and AOB amoA genes was performed using primers Arch-amoAF and Arch-amoAR  and amoA-1F and amoA-2R , respectively. Thaumarchaeal and bacterial 16S rRNA genes were quantified using primers 771F and 957R  and 341F and 518R , respectively. All real-time PCR amplifications were performed in duplicate with a reaction volume of 12.5 µl, which contained 2× iQ SYBR Green Supermix (Bio-Rad, Mississauga, Canada), 5 pmol of each primer, 5 µg of bovine serum albumin and 1 µl of template. Real-time PCR was performed on a CFX96 system (Bio-Rad). For both 16S rRNA genes, PCR conditions were 95°C for 3 min followed by 40 cycles of 95°C for 20 s, 55°C for 30 s and 72°C for 30 s (with fluorescence values recorded after the extension step). For amoA genes, the PCR conditions were the same as above, except with an extension time of 1 minute and annealing temperatures of 60°C and 58.5°C for bacterial and thaumarchaeal amoA genes, respectively. For all amplification reactions, melting curves from 65°C to 95°C were performed after each run with an incremental increase in temperature of 0.5°C.
PCR amplicons were used as standard template DNA, and were generated using the primers indicated above for their respective genes. We used genomic DNA from aquarium FW27 to generate standards for thaumarchaeal and bacterial amoA and thaumarchaeal 16S rRNA genes. Genomic DNA from Escherichia coli genomic DNA was used to generate bacterial 16S rRNA gene standards. Standard curves were constructed using serial dilutions of standard template DNA plotted against the cycle threshold (Ct) values for each dilution. Amplification efficiencies ranged from 90.6–98.2%, and coefficients of determinations (R2) ranged from 0.988 to 0.999. Melting curves calculated for each target sequence showed single peaks and all PCR products were verified on a 1% agarose gel. Starting DNA copy numbers for each sample were calculated from the linear regression equation of each standard curve.
Denaturing gradient gel electrophoresis
DGGE analysis of AOA amoA genes was performed as described previously  with minor modifications. Samples were run on 6% acrylamide gels, which provided better resolution than 8% acrylamide gels (data not shown). AOA amoA genes were amplified using primers CrenamoA23f and CrenamoA616r with thermal cycling as described elsewhere . The DGGE system used was a DGGEK-2401 (C.B.S. Scientific Company, Del Mar, CA) using previously described technical modifications . Gels were run for 15 h at 85 V and subsequently stained with SYBR green (Invitrogen) for 1 h. Gels were scanned using the Typhoon 9400 Variable Mode Imager (GE Healthcare, Piscataway, NJ). Individual DGGE bands were excised, amplified (using the above primers and conditions) and sequenced. From the original gel images, fingerprints were normalized for multi-gel alignment with GelCompar II (Applied Maths, Austin, TX) and a nonmetric multidimensional scaling (NMDS) plot was generated based on Pearson product moment correlations of background-subtracted densitometric curves.
Clone libraries and ordination analysis
PCR amplicons for sequencing were generated using Arch-amoAF and Arch-amoAR, with thermal cycling as described previously . Composites of freshwater and saltwater samples were produced by pooling equal nanogram amounts of PCR products from all freshwater and saltwater samples, respectively. Composite PCR products were ligated into the pGEM®-T Easy Vector (Promega, Madison, WI) according to the manufacturer's protocol. Single colonies were picked randomly and grown up in Luria Burtani broth containing ampicillin (100 µg ml−1), followed by a plasmid extraction and sequencing of inserts with the M13f primer. A total of 288 and 96 clones were sequenced for the composite freshwater and saltwater libraries, respectively.
All ordination analyses were performed using translated amino acid sequences. DNA sequences derived from both DGGE bands and clone libraries were translated using dna2pep . After discarding sequences containing stop codons in the amino acid translation, a total of 261 freshwater clones and 84 saltwater clones remained. Reference sequences were obtained from GenBank for environmental clones as well as isolated or enriched AOA representatives. The collection of sequences was aligned using MUSCLE  and the resulting alignment was cropped so that all sequences spanned the same 160 amino acid region. All positions in the alignment were used for ordination analyses. A distance matrix was produced using protdist  and scaled by Kruskal's nonmetric multidimensional scaling using the MASS package . All DNA sequences generated in this study were submitted to Genbank with accession numbers JN183456–JN183849.
Pearson product-moment correlation coefficients and coefficients of determination were calculated in Excel 2010 (Microsoft, Redmond, WA), and associated p-values were calculated using InStat 3 (GraphPad Inc. San Diego, CA). Unpaired t tests were utilized to compare means in G+C content between freshwater and saltwater sequences and were conducted in InStat 3.
Aquarium FW27 temporal patterns. Ammonium concentrations (A) are shown over several months (during 2010), and hourly over a 12 hour period. Proportions of AOA/AOB in the FW27 sponge filter DNA extract are shown from four time points over the course of 2 years.
Details of aquaria and associated quantitaive real-time PCR data.
Pearson correlation coefficients for aquarium chemistry parameters and AOA/AOB abundances for all aquaria.
This research is dedicated to Roger Knowles (1929–2009). We thank local aquarium retail outlets and members of the Kitchener Waterloo Aquarium Society (KWAS) for their enthusiastic participation in this study by providing aquarium filter samples.
Conceived and designed the experiments: JDN LAS. Performed the experiments: LAS KE JCS. Analyzed the data: LAS APM JDN. Wrote the paper: LAS JDN. Collected and processed samples: RP.
- 1. Parker R (2002) Aquaculture Science. Albany: Delmar. 621 p.
- 2. Andrews C, Exell A, Carrington N (1988) The Manual of Fish Health. Morris Plains: Tetra Press.
- 3. Brochier-Armanet C, Boussau B, Gribaldo S, Forterre P (2008) Mesophilic Crenarchaeota: proposal for a third archaeal phylum, the Thaumarchaeota. Nat Rev Microbiol 6: 245–252.
- 4. Spang A, Hatzenpichler R, Brochier-Armanet C, Rattei T, Tischler P, et al. (2010) Distinct gene set in two different lineages of ammonia-oxidizing archaea supports the phylum Thaumarchaeota. Trends Microbiol 18: 331–340.
- 5. Hovanec T, DeLong E (1996) Comparative analysis of nitrifying bacteria associated with freshwater and marine aquaria. Appl Environ Microbiol 62: 2888–2896.
- 6. Hovanec TA, Taylor LT, Blakis A, DeLong EF (1998) Nitrospira-like bacteria associated with nitrite oxidation in freshwater aquaria. Appl Environ Microbiol 64: 258–264.
- 7. Burrell P, Phalen C, Hovanec T (2001) Identification of bacteria responsible for ammonia oxidation in freshwater aquaria. Appl Environ Microbiol 67: 5791–5800.
- 8. Foesel BU, Gieseke A, Schwermer C, Stief P, Koch L, et al. (2008) Nitrosomonas Nm143-like ammonia oxidizers and Nitrospira marina-like nitrite oxidizers dominate the nitrifier community in a marine aquaculture biofilm. FEMS Microbiol Ecol 63: 192–204.
- 9. Urakawa H, Tajima Y, Numata Y, Tsuneda S (2008) Low temperature decreases the phylogenetic diversity of ammonia-oxidizing archaea and bacteria in aquarium biofiltration systems. Appl Environ Microbiol 74: 894–900.
- 10. Könneke M, Bernhard AE, de la Torre JR, Walker CB, Waterbury JB, et al. (2005) Isolation of an autotrophic ammonia-oxidizing marine archaeon. Nature 437: 543–546.
- 11. Green SJ, Leigh MB, Neufeld JD (2010) Denaturing gradient gel electrophoresis (DGGE) for microbial community analysis. In: Timmis KN, editor. Microbiology of Hydrocarbon and Lipid Microbiology. Berlin: Springer-Verlag Berlin Heidelberg. pp. 4137–4158.
- 12. Hallam SJ, Mincer TJ, Schleper C, Preston CM, Roberts K, et al. (2006) Pathways of carbon assimilation and ammonia oxidation suggested by environmental genomic analyses of marine Crenarchaeota. PLoS Biol 4: e95.
- 13. Treusch AH, Leininger S, Kletzin A, Schuster SC, Klenk H-P, et al. (2005) Novel genes for nitrite reductase and Amo-related proteins indicate a role of uncultivated mesophilic crenarchaeota in nitrogen cycling. Environ Microbiol 7: 1985–1995.
- 14. Hatzenpichler R, Lebedeva EV, Spieck E, Stoecker K, Richter A, et al. (2008) A moderately thermophilic ammonia-oxidizing crenarchaeote from a hot spring. Proc Natl Acad Sci USA 105: 2134–2139.
- 15. de la Torre JR, Walker CB, Ingalls AE, Könneke M, Stahl DA (2008) Cultivation of a thermophilic ammonia oxidizing archaeon synthesizing crenarchaeol. Environ Microbiol 10: 810–818.
- 16. Blainey PC, Mosier AC, Potanina A, Francis CA, Quake SR (2011) Genome of a low-salinity ammonia-oxidizing archaeon determined by single-cell and metagenomic analysis. PLoS ONE 6: e16626.
- 17. Neufeld JD, Mohn WW, de Lorenzo V (2006) Composition of microbial communities in hexachlorocyclohexane (HCH) contaminated soils from Spain revealed with a habitat-specific microarray. Environ Microbiol 8: 126–140.
- 18. Walker CB, de la Torre JR, Klotz MG, Urakawa H, Pinel N, et al. (2010) Nitrosopumilus maritimus genome reveals unique mechanisms for nitrification and autotrophy in globally distributed marine crenarchaea. Proc Natl Acad Sci U S A 107: 8818–8823.
- 19. Agogué H, Brink M, Dinasquet J, Herndl GJ (2008) Major gradients in putatively nitrifying and non-nitrifying Archaea in the deep North Atlantic. Nature 456: 788–791.
- 20. Herrmann M, Saunders A, Schramm A (2008) Archaea dominate the ammonia-oxidizing community in the rhizosphere of the freshwater macrophyte Littorella uniflora. Appl Environ Microbiol 74: 3279–3283.
- 21. Herrmann M, Saunders A, Schramm A (2009) Effect of lake trophic status and rooted macrophytes on community composition and abundance of ammonia-oxidizing prokaryotes in freshwater sediments. Appl Environ Microbiol 75: 3127–3136.
- 22. Wuchter C, Abbas B, Coolen MJL, Herfort L, van Bleijswijk J, et al. (2006) Archaeal nitrification in the ocean. Proc Natl Acad Sci USA 103: 12317–12322.
- 23. Leininger S, Urich T, Schloter M, Schwark L, Qi J, et al. (2006) Archaea predominate among ammonia-oxidizing prokaryotes in soils. Nature 442: 806–809.
- 24. Jia Z, Conrad R (2009) Bacteria rather than Archaea dominate microbial ammonia oxidation in an agricultural soil. Environ Microbiol 11: 1658–1671.
- 25. Wells G, Park H, Yeung C, Eggleston B, Francis C, et al. (2009) Ammonia-oxidizing communities in a highly aerated full-scale activated sludge bioreactor: betaproteobacterial dynamics and low relative abundance of Crenarchaea. Environ Microbiol 11: 2310–2328.
- 26. Limpiyakorn T, Sonthiphand P, Rongsayamanont C, Polprasert C (2011) Abundance of amoA genes of ammonia-oxidizing archaea and bacteria in activated sludge of full-scale wastewater treatment plants. Bioresource Technol 102(4): 3694–3701.
- 27. Di HJ, Cameron KC, Shen J-P, Winefield CS, O'Callaghan M, et al. (2010) Ammonia-oxidizing bacteria and archaea grow under contrasting soil nitrogen conditions. FEMS Microbiol Ecol 72: 386–394.
- 28. Verhamme DT, Prosser JI, Nicol GW (2011) Ammonia concentration determines differential growth of ammonia-oxidising archaea and bacteria in soil microcosms. ISME J 5(6): 1067–1071.
- 29. Taylor AE, Zeglin LH, Dooley S, Myrold DD, Bottomley PJ (2010) Evidence for different contributions of Archaea and Bacteria to the ammonia-oxidizing potential of diverse Oregon soils. Appl Environ Microbiol 76: 7691–7698.
- 30. Herrmann M, Scheibe A, Avrahami S, Kusel K (2011) Ammonium availability affects the ratio of ammonia-oxidizing bacteria to ammonia-oxidizing archaea in simulated creek ecosystems. Appl Environ Microbiol 77: 1896–1899.
- 31. Reigstad L, Richter A, Daims H, Urich T, Schwark L, et al. (2008) Nitrification in terrestrial hot springs of Iceland and Kamchatka. FEMS Microbiol Ecol 64: 167–174.
- 32. Beman J, Popp B, Francis C (2008) Molecular and biogeochemical evidence for ammonia oxidation by marine Crenarchaeota in the Gulf of California. ISME J 2: 429–441.
- 33. Martens-Habbena W, Berube PM, Urakawa H, de la Torre JR, Stahl DA (2009) Ammonia oxidation kinetics determine niche separation of nitrifying Archaea and Bacteria. Nature 461: 976–979.
- 34. Tourna M, Stieglmeiera M, Spanga A, Könneke M, Schintlmeisterc A, et al. (2011) Nitrososphaera viennensis, an ammonia oxidizing archaeon from soil. Proc Natl Acad Sci U S A 108(20): 8420–8425.
- 35. Holmes R, Aminot A, Kérouel R, Hooker B, Peterson B (1999) A simple and precise method for measuring ammonium in marine and freshwater ecosystems. Can J Fish Aquat Sci 56: 1801–1808.
- 36. Griffiths RI, Whiteley AS, O'Donnell AG, Bailey MJ (2000) Rapid method for coextraction of DNA and RNA from natural environments for analysis of ribosomal DNA- and rRNA-based microbial community composition. Appl Environ Microbiol 66: 5488–5491.
- 37. Bartram AK, Poon C, Neufeld JD (2009) Nucleic acid contamination of glycogen used in nucleic acid precipitation and assessment of linear polyacrylamide as an alternative co-precipitant. Biotechniques 47: 1019–1022.
- 38. Francis C, Roberts K, Beman J, Santoro A, Oakley B (2005) Ubiquity and diversity of ammonia-oxidizing archaea in water columns and sediments of the ocean. Proc Natl Acad Sci USA 102: 14683–14688.
- 39. Rotthauwe J, Witzel K, Liesack W (1997) The ammonia monooxygenase structural gene amoA as a functional marker: molecular fine-scale analysis of natural ammonia-oxidizing populations. Appl Environ Microbiol 63: 4704–4712.
- 40. Ochsenreiter T, Selezi D, Quaiser A, Bonch-Osmolovskaya L, Schleper C (2003) Diversity and abundance of Crenarchaeota in terrestrial habitats studied by 16S RNA surveys and real time PCR. Environ Microbiol 5: 787–797.
- 41. Muyzer G, de Waal EC, Uitterlinden AG (1993) Profiling of complex microbial populations by denaturing gradient gel electrophoresis analysis of polymerase chain reaction-amplified genes coding for 16S rRNA. Appl Environ Microbiol 59: 695–700.
- 42. Nicol G, Leininger S, Schleper C, Prosser J (2008) The influence of soil pH on the diversity, abundance and transcriptional activity of ammonia oxidizing archaea and bacteria. Environ Microbiol 10: 2966–2978.
- 43. Tourna M, Freitag T, Nicol G, Prosser J (2008) Growth, activity and temperature responses of ammonia-oxidizing archaea and bacteria in soil microcosms. Environ Microbiol 10: 1357–1364.
- 44. Wernersson R (2006) Virtual Ribosome - a comprehensive DNA translation tool with support for integration of sequence feature annotation. Nucl Acids Res 34: W385–W388.
- 45. Edgar RC (2004) MUSCLE: multiple sequence alignment with high accuracy and high throughput. Nucl Acids Res 32: 1792–1797.
- 46. Felsenstein J (1989) PHYLIP - phylogeny inference package (version 3.2). Cladistics 5:
- 47. Venables WN, Ripley BD (2002) Modern Applied Statistics with S. New York: Springer.