Fmicb 08 00101

ORIGINAL RESEARCH
published: 30 January 2017

doi: 10.3389/fmicb.2017.00101
Freshwater Recirculating
Aquaculture System Operations
Drive Biofilter Bacterial Community
Shifts around a Stable Nitrifying
Consortium of Ammonia-Oxidizing
Archaea and Comammox Nitrospira
Ryan P. Bartelme, Sandra L. McLellan and Ryan J. Newton *
School of Freshwater Sciences, University of Wisconsin-Milwaukee, Milwaukee, WI, USA
Recirculating aquaculture systems (RAS) are unique engineered ecosystems that

Edited by: minimize environmental perturbation by reducing nutrient pollution discharge. RAS
Hongyue Dang, typically employ a biofilter to control ammonia levels produced as a byproduct of
Xiamen University, China
fish protein catabolism. Nitrosomonas (ammonia-oxidizing), Nitrospira, and Nitrobacter
Reviewed by:
Uwe Strotmann, (nitrite-oxidizing) species are thought to be the primary nitrifiers present in RAS
Westfälische Hochschule, Germany biofilters. We explored this assertion by characterizing the biofilter bacterial and archaeal
Sebastian Luecker,
community of a commercial scale freshwater RAS that has been in operation for >15
Radboud University Nijmegen,
Netherlands years. We found the biofilter community harbored a diverse array of bacterial taxa
Hidetoshi Urakawa, (>1000 genus-level taxon assignments) dominated by Chitinophagaceae (∼12%) and
Florida Gulf Coast University, USA
Acidobacteria (∼9%). The bacterial community exhibited significant composition shifts
*Correspondence:
Ryan J. Newton with changes in biofilter depth and in conjunction with operational changes across a fish
newtonr@uwm.edu rearing cycle. Archaea also were abundant, and were comprised solely of a low diversity
assemblage of Thaumarchaeota (>95%), thought to be ammonia-oxidizing archaea
Specialty section:
This article was submitted to
(AOA) from the presence of AOA ammonia monooxygenase genes. Nitrosomonas
Aquatic Microbiology, were present at all depths and time points. However, their abundance was >3
a section of the journal
orders of magnitude less than AOA and exhibited significant depth-time variability not
Frontiers in Microbiology
observed for AOA. Phylogenetic analysis of the nitrite oxidoreductase beta subunit
Received: 07 November 2016
Accepted: 13 January 2017 (nxrB) gene indicated two distinct Nitrospira populations were present, while Nitrobacter
Published: 30 January 2017 were not detected. Subsequent identification of Nitrospira ammonia monooxygenase
Citation: alpha subunit genes in conjunction with the phylogenetic placement and quantification
Bartelme RP, McLellan SL and
Newton RJ (2017) Freshwater
of the nxrB genotypes suggests complete ammonia-oxidizing (comammox) and
Recirculating Aquaculture System nitrite-oxidizing Nitrospira populations co-exist with relatively equivalent and stable
Operations Drive Biofilter Bacterial
abundances in this system. It appears RAS biofilters harbor complex microbial
Community Shifts around a Stable
Nitrifying Consortium of communities whose composition can be affected directly by typical system operations
Ammonia-Oxidizing Archaea and while supporting multiple ammonia oxidation lifestyles within the nitrifying consortium.
Comammox Nitrospira.
Front. Microbiol. 8:101. Keywords: recirculating aquaculture system, biofilter, nitrifiers, ammonia-oxidizing archaea, comammox,
doi: 10.3389/fmicb.2017.00101 microbial communities, Nitrospira
Frontiers in Microbiology | www.frontiersin.org 1 January 2017 | Volume 8 | Article 101

Bartelme et al. Recirculating Aquaculture Biofilter Microorganisms
INTRODUCTION The non-nitrifying component of RAS biofilter communities

also impact biofilter function. Heterotrophic biofilm overgrowth
The development of aquacultural technology allows societies to can limit oxygen availability to the autotrophic nitrifying
reduce dependency on capture fisheries and offset the effects community resulting in reduced ammonia-oxidation rates
of declining fish numbers (Barange et al., 2014). Aquaculture (Okabe et al., 1995). Conversely, optimal heterotrophic biofilm
production now accounts for nearly 50% of fish produced formation protects the slower-growing autotrophs from biofilm
for consumption, and estimates indicate a five-fold increase shear stress and recycles autotrophic biomass (Kindaichi
in production will be required in the next two decades et al., 2004). Previous studies have suggested the diversity
to meet societal protein demands (FAO, 2014). However, of non-nitrifying microorganisms in RAS biofilters could be
expanding production will increase the environmental impact of large and sometimes contain opportunistic pathogens and
aquaculture facilities and raises important concerns regarding the other commercially detrimental organisms (Schreier et al.,
sustainability of aquaculture practices. Recirculating aquaculture 2010). However, most of these studies used low-coverage
systems (RAS) have been developed to overcome pollution characterization methods (e.g., DGGE, clone libraries) to
concerns and stocking capacity limits of conventional terrestrial describe the taxa present, so the extent of this diversity
aquaculture facilities (Chen et al., 2006; Martins et al., 2010). RAS and similarity among systems is relatively unknown. Recently,
offer several advantages over traditional flow-through systems the bacterial community of a set of seawater RAS biofilters
including: 90–99% reduced water consumption (Verdegem et al., run with different salinity and temperature combinations was
2006; Badiola et al., 2012), more efficient waste management characterized with massively parallel sequencing technology (Lee
(Piedrahita, 2003), and potential for implementation at locations et al., 2016). This study provided the first deeper examination of a
that decrease distance to market (Martins et al., 2010). RAS RAS biofilter microbial community, and revealed a highly diverse
components are similar to those used in wastewater treatment, bacterial community that shifted in response to environmental
including solids capture and removal of nitrogenous waste from conditions but more consistent nitrifying assemblage typically
excess animal waste and undigested feed. The advancement of dominated by Nitrospira-classified microorganisms.
RAS technology and advantages over flow-through systems has In this study, we aimed to deeply characterize the bacterial and
led to increasing RAS use, especially among countries that place archaeal community structure of a commercial-scale freshwater
high value on minimizing environmental impacts (Badiola et al., RAS raising Perca flavescens (Yellow perch) employing a fluidized
2012) and in urban areas where space is limiting (Klinger and sand biofilter that has been in operation for more than 15
Naylor, 2012). years. We hypothesized that the biofilter sand biofilm community
Nitrifying biofilters are a critical component of most RAS and would exhibit temporal variability linked to environmental
an important determinant of operational success. These biofilters changes associated with the animal rearing process and a diverse
are also cited as the biggest hurdle for RAS start-up and the nitrifying assemblage. To address these questions, we used
most difficult component to manage once the RAS is in operation massively parallel sequencing to characterize the bacterial and
(Badiola et al., 2012). RAS biofilters act to remove nitrogenous archaeal biofilter community across depth and time gradients.
waste byproducts generated by fish protein catabolism and We also identified and phylogenetically classified nitrification
oxidation processes. Ammonia and nitrite are of most concern to marker genes for the ammonia monooxygenase alpha subunit
freshwater aquaculturalists, with the toxic dose of both nitrogen (amoA; Rotthauwe et al., 1997; Pester et al., 2012; van Kessel et al.,
species depending on pH and the aquatic organism being reared 2015) and nitrite oxidoreductase alpha (nxrA; Poly et al., 2008;
(Lewis and Morris, 1986; Randall and Tsui, 2002). In RAS process Wertz et al., 2008) and beta (nxrB; Pester et al., 2014) subunits
engineering, designers typically cite the principle nitrifying taxa present in the biofilter, and then tracked their abundance with
as Nitrosomonas spp. (ammonia-oxidizers) and Nitrobacter spp. biofilter depth and over the course of a fish rearing cycle.
(nitrite-oxidizers) (Kuhn et al., 2010) and model system capacity
from these organisms’ physiologies (Timmons and Ebeling,
MATERIALS AND METHODS
2013). It is now clear Nitrosomonas and Nitrobacter are typically
absent or in low abundance in freshwater nitrifying biofilters UWM Biofilter Description
(Hovanec and DeLong, 1996) while Nitrospira spp. are common All samples were collected from the University of Wisconsin-
(Hovanec et al., 1998). More recent studies of freshwater Milwaukee Great Lakes Aquaculture Facility RAS biofilter
aquaculture biofilters have expanded the nitrifying taxa present (UWM biofilter). Measured from the base, the biofilter stands
in these systems to include ammonia-oxidizing archaea (AOA), ∼2.74 m tall, with a diameter of ∼1.83 m. The water level within
a variety of Nitrospira spp., and Nitrotoga (Sauder et al., 2011; the biofilter is ∼2.64 m from the base, with the fluidized sand
Bagchi et al., 2014; Hüpeden et al., 2016). Further studies are filter matrix extending to a height of ∼1.73 m from the base. The
needed to understand whether other nitrifying consortia co- biofilter is filled with Wedron 510 silica sand, which is fluidized
inhabit RAS biofilters with Nitrosomonas and Nitrobacter spp., or to ∼200% starting sand volume by the use of 19 schedule 40
if diverse assemblages of nitrifying organisms are characteristic of PVC probes, each with a diameter of 3.175 cm. The probes
high-functioning systems. A more refined understanding of RAS receive influent from the solid waste clarifier, which upwells
biofilter nitrifying consortia physiology would inform system through the filter matrix. Samples for this study were taken at
design optimization and could alter parameters that are now three depths within the fluidized sand biofilter, defined as surface
considered design constraints. (∼1.32–1.42 m from biofilter base), middle (∼0.81–0.91 m from

biofilter base), and bottom (∼0.15–0.30 m, from biofilter base). EMD Millipore, Darmstadt, Germany), frozen at −80◦ C, and
Depictions of the UWM biofilter and sample sites are shown macerated with a sterilized spatula prior to DNA extraction.
in Figure 1. The maximum flow rate of the biofilter influent To separately address the spatial distribution of bacterial taxa,
is 757 L per minute, which gives a hydraulic residence time depth samples were taken from the filter matrix by using 50
of ∼9.52 min. Typical system water quality parameters are as mL syringes with attached weighted Tygon tubing (3.2 mm
follows (mean ± standard deviation): pH 7.01 ± 0.09, oxidation- ID, 6.4 mm OD; Saint-Gobain S.A., La Défense, Courbevoie,
reduction potential 540 ± 50 (mV), water temperature 21.7 ± France). Samples were binned into categories by approximate
0.9 (◦ C), and biofilter effluent dissolved oxygen (DO) 8.20 ± 0.18 distance from the filter base as surface, middle and bottom.
mg/L. The biofilter is designed to operate maximally at 10 kg feed Tubing was sterilized with 10% bleach and rinsed 3X with sterile
per day, which is based on the predicted ammonia production deionized water between sample collections. DNA was extracted
by fish protein catabolism at this feeding rate (Timmons and separately from biofilter sand and water samples (∼1 g wet
Ebeling, 2013). weight and 100 mL, respectively) using the MP Bio FastDNA R
SPIN Kit for Soil (MP Bio, Solon, OH, USA) according to the
Sample Collection, Processing, and DNA manufacturer’s instructions except that each sample underwent
Extraction 2 min of bead beating with the MP Bio FastDNA R
SPIN kit’s
Samples from the top of the biofilter matrix were collected in included beads at the Mini-BeadBeater-16’s only operational
autoclaved 500 mL polypropylene bottles. Two samples from the speed (Biospec Products, Inc., Bartlesville, OK, USA). DNA
surface of the biofilter were collected during the final 2 months quality and concentration was checked using a NanoDrop R
Lite
of one Yellow perch rearing cycle and then immediately before (Thermo Fisher Scientific Inc., Waltham, MA, USA). Sample
the initiation of a new rearing cycle in the system. After stocking details and associated environmental data and molecular analyses
the system with fish, samples were collected approximately every are listed in Table S1.
week through the first half of the new rearing cycle (the strains
of Yellow perch present during this study need ∼9 months Ammonia and Nitrite Measurements
to grow to market size). Following collection, water from the For both the time series and depth profiles, a Seal Analytical
biofilter matrix samples was decanted into a second sterile 500 AA3 Autoanalyzer (Seal Analytical Inc., Mequon, WI, USA) was
mL bottle for further processing. Then, approximately 1 g wet used to quantify ammonia and nitrite, using the manufacturer’s
weight sand was removed from the sample bottle and frozen supplied phenol and sulfanilamide protocols on two separate
at −80◦ C for storage prior to DNA extraction. Water samples channels. To quantify only nitrite, the cadmium reduction
were filtered onto 0.22 µm filters (47 mm mixed cellulose esters, column was not incorporated into the Auto Analyzer. RAS
operators recorded all other chemical parameters from
submerged probes measuring temperature, pH, and oxidation-
reduction potential. Per the laboratory standard operating
procedure, RAS operators used Hach colorimetric kits to
measure rearing tank concentrations of ammonia and nitrite.
16S rRNA Gene Sequencing

To maximize read depth for a temporal study of the biofilter
surface communities, we used the illumina HiSeq platform and
targeted the V6 region of the 16S rRNA gene for Archaea and
Bacteria separately. In total, we obtained community data from
15 dates for the temporal analysis. To interrogate changes in the
spatial distribution of taxa across depth in the biofilter and obtain
increased taxonomic resolution, we used 16S rRNA gene V4-V5
region sequencing on an illumina MiSeq. We obtained samples
from three depths n = 5 for the surface, n = 5 for the middle,
and n = 4 for the bottom. Sample metadata are listed in Table
S1. Extracted DNA samples were sent to the Josephine Bay Paul
Center at the Marine Biological Laboratory (V6 Archaea and V6
Bacteria; V4-V5 samples from 12/8/2014 to 2/18/2015) and the
Great Lakes Genomic Center (V4-V5 samples from 11/18/2014,
12/2/2014, 12/18/2014) for massively parallel 16S rRNA gene
sequencing using previously published bacterial (Eren et al.,
FIGURE 1 | llustration of the UW-Milwaukee recirculating aquaculture 2013) and archaeal (Meyer et al., 2013) V6 illumina HiSeq and
system (RAS) fluidized sand biofilter. For illustration purposes only a single bacterial V4-V5 illumina MiSeq chemistries (Huse et al., 2014b;
inflow pipe is shown. Nineteen of these pipes are present in the system. Water Nelson et al., 2014). Reaction conditions and primers for all
flow is depicted with directional arrows, sample locations are indicated by
circles, and the biofilter height is listed.
illumina runs are detailed in the aforementioned citations, and
may be accessed at: https://vamps.mbl.edu/resources/primers.

php#illumina. Sequence run processing and quality control for of archaeal amoA and Nitrospira sp. nxrB. One sample from
the V6 dataset are described in Fisher et al. (2015), while the center of the sand biofilter was used to construct clone
CutAdapt was used to trim the V4-V5 data of low quality libraries for betaproteobacterial amoA and comammox amoA.
nucleotides (phred score <20) and primers (Martin, 2011; Fisher The center biofilter sample was chosen as it produced well-
et al., 2015). Trimmed reads were merged using Illumina-Utils defined amplicons suitable for cloning target amoA genes. All
as described previously (Newton et al., 2015). Minimum entropy PCR reactions for clone libraries were constructed using a TOPO
decomposition (MED) was implemented on each dataset to PCR 2.1 TA cloning kit plasmid (Invitrogen, Life Technologies,
group sequences (MED nodes = operational taxonomic units, Carlsbad, CA). Libraries were sequenced on an ABI 3730
OTUs) for among sample community composition and diversity Sanger-Sequencer with M13 Forward primers. Vector plasmid
analysis (Eren et al., 2015). MED uses information uncertainty sequence contamination was removed using DNAStar (Lasergene
calculated via Shannon entropy at all nucleotide positions of Software, Madison, WI).
an alignment to split sequences into sequence-similar groups Cloned sequences of Betaproteobacteria amoA, Archaea
(Eren et al., 2015). The sequence datasets were decomposed amoA, and Nitrospira nxrB from this study were added to
with the following minimum substantive abundance settings: ARB alignment databases from previous studies (Abell et al.,
bacterial V6, 377; archaeal V6, 123; bacterial V4-V5, 21. The 2012; Pester et al., 2012, 2014). Comammox amoA sequences
minimum substantive threshold sets the abundance threshold for from this study were aligned with those from van Kessel
MED node (i.e., OTU) inclusion in the final dataset. Minimum et al. (2015), Pinto et al. (2015), and Daims et al. (2015)
substantive abundances were calculated by dividing the sum total using MUSCLE and imported into a new ARB database where
number of 16S rRNA gene sequences per dataset by 50,000 as the alignment was heuristically corrected before phylogenetic
suggested in the MED best practices (sequence counts are listed tree reconstruction. For the AOA, AOB, and Nitrospira amoA
in Table S2). The algorithm Global Alignment for Sequence phylogenies, relationships were calculated using Maximum-
Taxonomy (GAST) was used to assign taxonomy to sequence Likelihood (ML) with RAxML on the Cipres Science Gateway
reads (Huse et al., 2008), and the website Visualization and (Miller et al., 2010; Stamatakis, 2014) and Bayesian inference (BI)
Analysis of Microbial Population Structures (VAMPS; Huse et al., using MrBayes with a significant posterior probability of <0.01
2014a), was used for data visualization. and the associated consensus tree (Abell et al., 2012; Pester et al.,
2012, 2014) from ARB incorporated into a tree block within
Comammox amoA PCR the input nexus file to reduce calculation time (Miller et al.,
To target comammox Nitrospira amoA for PCR and subsequent 2010; Ronquist et al., 2012). Consensus trees were then calculated
cloning and sequencing, amoA nucleotide sequences from van from the ML and BI reconstructions using ARB’s consensus tree
Kessel et al. (2015) and Daims et al. (2015) were aligned using algorithm (Ludwig et al., 2004).
MUSCLE (Edgar, 2004). The alignment was imported into The Nitrospira nxrB sequences generated in this study were
EMBOSS to generate an amoA consensus sequence (Rice et al., significantly shorter than those used for nxrB phylogenetic
2000). Primer sequences were identified from the consensus reconstruction in Pester et al. (2014), so we did not perform
using Primer3Plus (Untergasser et al., 2012), and the candidates phylogenetic reconstructions as with the other marker genes.
along with the methane monooxygenase subunit A (pmoA) Instead, the UWM Biofilter and Candidatus Nitrospira nitrificans
primers suggested by van Kessel et al. (2015), were evaluated sequences were added to the majority consensus tree from Pester
against the consensus sequence in SeqMan Pro (DNAStar), using et al. (2014) using the Quick-Add Parsimony tool of the ARB
MUSCLE (Edgar, 2004). The pmoA forward primer (Luesken package (Ludwig et al., 2004). This tool uses sequence similarity
et al., 2011) and candidate primer COM_amoA_1R (this study; to add sequences to pre-existing trees without changing the tree
Table 1) offered the best combination of read length and topology.
specificity, and subsequently were used to amplify amoA genes
from our samples. qPCR Assays for Target Marker Genes
Quantitative PCR assays were designed to differentiate two
Clone Library Construction and Nitrospira nxrB genotypes and two Nitrosomonas amoA
Phylogenetic Analysis genotypes in our system. Potential qPCR primer sequences
Multiple endpoint PCR approaches were used to investigate the were identified using Primer3Plus (Untergasser et al., 2012)
nitrifying community composition of the RAS fluidized sand on MUSCLE (Edgar, 2004) generated alignments in DNAStar
biofilter for amoA (Gammaproteobacteria, Betaproteobacteria, (Lasergene Software, Madison, WI). Primer concentrations and
Archaea, and comammox Nitrospira), nxrA (Nitrobacter spp.), annealing temperatures were optimized for specificity to each
and nxrB (non-Nitrobacter NOB). The primer sets and reaction reaction target. Primers were checked using Primer-BLAST
conditions used are listed in Table 1. All endpoint PCR reactions on NCBI to ensure the assays matched their target genes. The
were carried out at a volume of 25 µl: 12.5 µl 2x Qiagen newly designed primers were tested for between genotype
PCR master mix (Qiagen, Hilden, Germany), 1.5 µl appropriate cross-reactivity using the non-target genotype sequence in both
primer mix (F&R), 0.5 µl bovine serum albumin (BSA), 0.75 µl endpoint and real time PCR dilution series. After optimization,
50 mM MgCl2 , and 1 µl DNA extract. all assays amplified only the target genotype. Due to high
DNA samples of biofilter water and sand from four different sequence similarity between the two archaeal amoA genotypes
rearing cycle time-points were used to construct clone libraries (>90% identity) in our system, a single qPCR assay to target

TABLE 1 | Primer sets used for endpoint PCR and qPCR.
Target organisms Gene Assay Forward primer Reverse primer Primer Approximate Thermocycler temperature Citation
target type conc. (nM) product size programs
Bartelme et al.
(BP)
Betaproteobacteria amoA Endpoint 1F 5′ -GGG GHT TYT ACT 2R 5′ -CCC CTC KGS AAA GCC 300 490 1 × 95◦ C 5:00 min; 30 × 95◦ C 0:30 Rotthauwe et al., 1997;
AOB PCR GGT GGT-3′ TTC TTC-3′ min, 53◦ C 0:30 min, 72◦ C 0:30 min; Christman et al., 2011
1 × 72◦ C 7:00 min
Gammaproteobacteria amoA Endpoint 3F 5′ -GGT GAG TGG GYT 4R 5′ -GCT AGC CAC TTT 300 560 1 × 95◦ C 5:00 min; 30 × 95◦ C 0:30 Christman et al., 2011
AOB PCR AAC MG-3′ CTG-3′ min, 48◦ C 0:30 min, 72◦ C 0:30 min;
1 × 72◦ C 7:00 min
Ammonia-oxidizing amoA Endpoint 19F 5′ -ATG GTC TGG YTW 616R 5′ -GCC ATC CAB CKR 300 637 1× 95◦ C 5:00 min; 30 × 95◦ C 0:30 Tourna et al., 2008; Pester
Frontiers in Microbiology | www.frontiersin.org

Archaea PCR AGA CG-3′ TAN GTC CA-3′ min, 50◦ C 0:30 min, 72◦ C 0:30 min; et al., 2012
1 × 72◦ C 7:00 min
Comammox amoA amoA Endpoint pmoA-189b-F 5′ -GGN GAC Com_amoA_1_R 5′ -CGA GAT 300 520 1 × 95◦ C 10:00 min; 35 × 95◦ C 0:40 Fwd (Luesken et al., 2011) &
PCR TGG GAC TTY TGG-3′ CAT GGT GCT GTG AC-3′ min, 56◦ C 0:40 min, 72◦ C 0:15 min; Rev This Study
1 × 72◦ C 7:00 min
Nitrobacter spp. nxrA Endpoint F1nxrA 5′ -CAG ACC GAC R2nxrA 5′ -TCC ACA AGG AAC 300 322 1 × 94◦ C 5:00 min; 35 × 94◦ C 0:30 Fwd (Poly et al., 2008) and
PCR GTG TGC GAA AG-3′ GGA AGG TC-3′ min, 55◦ C 0:45 min, 72◦ C 1:00 min; Rev (Wertz et al., 2008)
1 × 72◦ C 10:00 min
Non-Nitrobacter nxrB Endpoint nxrB169f 5′ -TAC ATG TGG nxrB638r 5′ -CGG TTC TGG 300 485 1 × 95◦ C, 5:00 min; 35 × 95◦ C 0:40 Pester et al., 2014
NOB PCR TGG AAC A-3′ TCR ATC A-3′ min, 50◦ C 0:40 min, 72◦ C 1:30 min;
1 × 72◦ C 10:00 min
UWM AOA - Total amoA qPCR Arch-amoAF 5′ -CTG ACT Arch-amoAR 5′ -CCC AAT GCA 200 170 1 × 95◦ C 2:00 min; 40 × 95◦ C 0:05 This Study
5
GGG CGT GGA CAT CA-3′ AAC CAT GCA CC-3′ min, 62◦ C 0:45 min
UWM Nitroso - 1 amoA qPCR Beta-amoA-m1-F 5′ -TCG Beta-amoA-m2-R 5′ -ACA AAC 200 70 1 × 95◦ C 2:00 min; 40 × 95◦ C 0:05 This Study
AAC AAG GTT CAC TCC GCT GAG AAG AAC GC-3′ min, 61◦ C 0:45 min
GTA C-3′
UWM Nitroso - 2 amoA qPCR Beta-amoA-O2-F 5′ -ATT Beta-amoA-O2-R 5′ -TAT GAC 200 145 1 × 95◦ C 2:00 min; 40 × 95◦ C 0:05 This Study
TGG ACC GAC CCA CTT CAC CAA ACG TAC GC-3′ min, 60◦ C 0:45 min
ACC-3′
Nitrospira nxrB nxrB qPCR NitrospiraG1-a-F 5′ -TAT NitrospiraG1-a-R 5′ -ATG TTC 200 104 1 × 95◦ C 2:00 min; 40 × 95◦ C 0:05 This Study
uwm-1 GGG GTG TTC GAA GGG ACG AAG CGC CAT TC-3′ min, 67◦ C 0:45 min
ATG-3′
Nitrospira nxrB nxrB qPCR NitrospiraG2-a-F 5′ -ACG NitrospiraG2-a-R 5′ -CGG CAT 200 123 1 × 95◦ C 2:00 min; 40 × 95◦ C 0:05 This Study
uwm-2 TCA AAA TCA CGC AGC CGA AAA TGG TCA TCC-3′ min, 65◦ C 0:45 min
TG-3′
Comammox UWM amoA qPCR UWM_comammox_amoA_F1 UWM_comammox_amoA_R1 200 70 1 × 95◦ C 2:00 min; 40 × 95◦ C 0:05 This Study
amoA 5′ -CGG ACT ACA TGG 5′ -GAG CCC ACT TCG ATC min, 59◦ C 0:45 min
GCT TTG C-3′ ATC C-3′
Recirculating Aquaculture Biofilter Microorganisms
January 2017 | Volume 8 | Article 101

both genotypes was developed using the steps described above. biomass concentration, we used the mean cell diameter (0.96
The two closely related sequence types were pooled in equimolar µm) for Candidatus Nitrosocosmicus franklandus (Lehtovirta-
amounts for reaction standards. A comammox amoA qPCR Morley et al., 2016) to calculate the biovolume of a single cell,
primer set was developed using the same methods as the other and used the conversion factor of 310 fg∗ C/µm3 (Mußmann
assays presented in this study. All assay conditions are listed in et al., 2011) to relate biovolume to endogenous respiration.
Table 1. All qPCR assays were run on an Applied Biosystems The modeled biomass concentration was plotted vs. a range of
StepOne Plus thermocycler (Applied Biosystems, Foster City, potential MCRT for a RAS fluidized sand filter (Summerfelt,
CA). Cloned target genes were used to generate standard Personal communication). The results of all amoA qPCR
curves from 1.5 × 106 to 15 copies per reaction. All reactions assays were combined to estimate total ammonia-oxidizing
were carried out in triplicate, with melt curve and endpoint microorganism biomass in copy numbers per gram wet weight
confirmation of assays (qPCR standard curve parameters and sand. Modeled biomass was then compared to our AOM qPCR
efficiency are listed in Table S3). assay results. A commented R-script for the model is available on
GitHub (https://github.com/rbartelme/BFprojectCode.git).
Statistics and Data Analysis
Taxonomy-based data were visualized with heatmaps constructed NCBI Sequence Accession Numbers
in the R statistical language (R Core Team, 2014), by Bacterial V6, V4-V5, and Archaeal V6 16S rRNA gene sequences
implementing functions from the libraries gplots, Heatplus from generated in this study are available from the NCBI SRA
Bioconductor Lite, VEGAN, and RColorBrewer. MED nodes (SRP076497; SRP076495; SRP076492). Partial gene sequences for
were used in all sample diversity metrics. The EnvFit function amoA and nxrB are available through NCBI Genbank and have
in the VEGAN (Oksanen et al., 2015) R package was used to accession numbers KX024777–KX024822.
test the relationship between RAS observational data and changes
in the biofilter bacterial community composition. Pearson’s RESULTS
correlations were calculated using the Hmisc package in R
(Harrell, 2016) to test whether 16S rRNA, amoA, and nxrB gene Biofilter Chemistry Results
copies correlated over time. Kruskal–Wallis rank sum tests were RAS operations data was examined from the beginning of a
performed in the R base statistics package (R Core Team, 2014) Yellow perch rearing cycle until ∼6 months afterward. The
to test whether the populations of the aforementioned genes mean biofilter influent concentrations of ammonia and nitrite
were stratified by depth. The ADONIS function from VEGAN were, respectively, 9.02 ± 4.76 and 1.69 ± 1.46 µM. Biofilter
was used on the V4-V5 depth dataset to test the significance effluent ammonia concentrations (3.84 ± 7.32 µM) remained
of the observed Bray-Curtis dissimilarity as a function of depth within the toxicological constraints (<60 µM) of P. flavescens
categorical factors, with strata = NULL since the same biofilter reared in the system. On occasion, nitrite accumulated above the
was sampled multiple times. recommended threshold of 0.2 µM in both the rearing tank (0.43
± 0.43 µM) and biofilter effluent (0.73 ± 0.49 µM). No major
Biomass Model fish illnesses were reported during the RAS operational period.
To determine whether the observed ammonia removal could Environment and operations data are listed in Table S1.
provide the energy needed to support the number of potential
ammonia-oxidizing microorganisms (AOM) in the biofilter Bacterial and Archaeal Assemblages
as quantified via qPCR, we modeled steady-state biomass within the Biofilter
concentration from measured ammonia oxidation with the The characterization of the RAS biofilter bacterial community
following equation: revealed that both the sand-associated and water communities
∗ were diverse at a broad taxonomic level; 17 phyla averaged >0.1%
θx Yao in each of the biofilter sand and water bacterial communities
XAO = 1SNH3
θ 1 + bAO ∗ θx (See Table S2 for sample taxonomic characterization to
genus). Proteobacteria (on average, 40% of biofilter sand
XAO is defined as the biomass concentration of ammonia community sequences and 40% of water sequences) and
oxidizers in milligrams per liter in previous models (Mußmann Bacteroidetes (18% in sand, 33% in water) dominated both
et al., 2011), however, in this study we converted to cells per water and sand bacterial communities. At family-level taxonomic
wet gram of sand by identifying the mean grams of sand per classification, the biofilter sand-associated community was
liter water in the biofilter. Θx is the mean cell residence time distinct from the water community. The greatest proportion
(MCRT) in days and was unknown for the system. Θ is the of sequences in the sand samples were classified to the
hydraulic retention time in days, which, is ∼9.52 min, or 0.0066 bacterial groups, Chitinophagaceae (mean relative abundance,
days in this system. YAO is the growth yield of ammonia oxidizers, 12%), Acidobacteria family unknown (9%), Rhizobiales family
and bAO is the endogenous respiration constant of ammonia unknown (6%), Nocardioidaceae (4%), Spartobacteria family
oxidizers, which were estimated as 0.34 kg volatile suspended unknown (4%), and Xanthomonadales family unknown (4%),
solids (VSS)/kg NH4+ −N and 0.15 d−1 from Mußmann et al. while the water samples were dominated by sequences classified
(2011). ∆SNH3 is the change in substrate ammonia concentration to Chitinophagaceae (14%), Cytophagaceae (8%), Neisseriaceae
between influent and effluent in mg/L. To calculate XAO , or (8%), and Flavobacteriaceae (7%). At the genus-level Kribbella,

Chthoniobacter, Niabella, and Chitinophaga were the most

numerous classified taxa, each with on average >3% relative
abundance in the biofilter samples.
Using Minimum Entropy Decomposition (MED) to obtain
highly discriminatory sequence binning, we identified 1261
nodes (OTUs) across the bacterial dataset. A MED-based
bacterial community composition comparison (Figure 1)
supported the patterns observed using broader taxonomic
classification indicating that the biofilter sand-associated
community was distinct from the assemblage present in the
biofilter water.
In contrast to the large diversity in the bacterial community,
we found the archaeal community to be dominated by a single
taxonomic group, affiliated with the genus Nitrososphaera
This taxon made up >99.9% of the Archaea-classified
sequences identified in the biofilter samples (Table S2). This
taxon also was represented almost completely by a single
sequence (>95% of Archaea-classified sequences) that was
identical to a number of database deposited Thaumarchaeota
sequences, including the complete genome of Candidatus
Nitrosocosmicus oleophilus (CP012850), along with clones from
activated sludge, wastewater treatment, and freshwater aquaria
(KR233006, KP027212, KJ810532–KJ810533).
The initial biofilter community composition characterization
revealed distinct communities between the biofilter sand and
decanted biofilter water (Figure 2). Based on this data and
that fluidized-bed biofilter nitrification occurs primarily in
particle-attached biofilms (Schreier et al., 2010), we focused
our further analyses on the biofilter sand matrix. In the
sand samples, we observed a significant change in bacterial
community composition (MED nodes) over time (Table 2).
The early portion of the study, which included a period while
market sized Yellow perch were present in the system (sample
−69 and −26), a fallow period following fish removal (sample FIGURE 2 | Dendrogram illustrating the bacterial community
0), and time following re-stocking of mixed-age juvenile fish composition relationships among biofilter sand and biofilter water
(sample 7 and 14), had a more variable bacterial community samples. A complete-linkage dendrogram is depicted from Bray–Curtis
sample dissimilarity relationships based on Minimum Entropy Decomposition
composition (Bray-Curtis mean similarity 65.2 ± 6.5%) than the
node distributions among samples (V6 dataset). The leaves of the dendrogram
remaining samples (n = 9) collected at time points after an adult are labeled with the day count, where 0 represents the beginning of a fish
feed source had been started (20.0 ± 6.4%, Figure 3). Several rearing cycle. Negative numbers are days prior to a new rearing cycle. The day
operational and measured physical and chemical parameters, count is followed by the date sampled (mm.dd.yy). See Table S1 for sample
including oxidation-reduction potential, feed size, conductivity, metadata.
and biofilter effluent nitrite were correlated (p < 0.05)
with the time-dependent changes in bacterial community
composition (see Table 2 for environmental correlation Nitrifying Community Composition and
results). Phylogeny
Using a second sequence dataset (V4-V5 16S rRNA gene The massively parallel 16S rRNA gene sequencing data indicated
sequences), we examined the bacterial community composition bacterial taxa not associated with nitrification comprised the
associated with sand across a depth gradient (surface, middle, majority (∼92%) of the sand biofilter bacterial community. In
bottom). We found the bacterial communities in the top sand contrast, >99.9% of the archaeal 16S rRNA gene sequences were
samples were distinct from those in the middle and bottom classified to a single taxon associated with known AOA. Among
(ADONIS R2 = 0.74, p = 0.001; Figure 4). The Planctomycetes the bacterial taxa, Nitrosomonas represented <1% of the total
were a larger portion of the community in the surface sand community across all samples and no Nitrobacter sequences
(on average 15.6% of surface sand vs. 9.6% of middle/bottom were obtained. We also were unable to amplify Nitrobacter
sand), whereas the middle and bottom layers harbored a greater nxrA genes (Figure S1) with a commonly used primer set (Poly
proportion of Chitinophagaceae (7.4% in surface vs. 16.8% in et al., 2008; Wertz et al., 2008). In contrast, Nitrospira was fairly
middle/bottom) and Sphingomonadaceae (2.4% in surface vs. abundant, comprising 2–5% of the total bacterial community
7.9% in middle/bottom; Figure 4). (Table S2).

TABLE 2 | Environmental variable to bacterial community composition

correlations.
Variablea,b Dim1 Dim2 R2 Pr(>r)
Days From Start 0.836 0.548 0.94 0.002

Number of Fish −0.839 −0.544 0.77 0.024
Fish Mortalities 0 0 0 1
Culled Fish 0 0 0 1
System pH −0.454 0.891 0.03 0.911
Air Temperature 0.844 0.537 0.39 0.326
Water Temperature 0.752 0.659 0.69 0.05
Conductivity 0.970 −0.242 0.82 0.042
System Ammonia 0.651 0.759 0.50 0.19
System Nitrite 0.823 −0.568 0.87 0.011
Biofilter PSI 0.473 0.881 0.70 0.081
Biofilter Influent Ammonia 0.297 0.955 0.63 0.097
Biofilter Effluent Ammonia −0.582 0.813 0.03 0.949 FIGURE 3 | Non-metric multidimensional scaling plot of Bray–Curtis
bacterial community composition dissimilarity between sample time
Biofilter Influent Nitrite 0.687 0.727 0.69 0.057
points. nMDS Stress = 0.07 and dimensions (k) = 2. Arrows indicate the
Biofilter Effluent Nitrite 0.782 0.623 0.81 0.01 sample progression through time from the end of one rearing cycle
ORP 0.928 −0.374 0.82 0.021 (daynumber −69 and −26), to a period with no fish (0), and into the
Feed Size 0.991 −0.133 0.88 0.042 subsequent rearing cycle (7–126). The circle indicates samples taken after fish
had grown to a size where feed type and amount were stabilized (3 mm
kg feed 0.798 0.603 0.47 0.19
pelleted feed diet and 3–7 kg of feed per day).
Percent Variance Explainedc 23.8 11.0 – –
a The V6 16S rRNA gene biofilter sand bacterial community composition data were related
to the system metadata in Table S1 using environmental vector fitting of a principal Daims et al., 2015; van Kessel et al., 2015; Figure 7A). Because of
coordinates analysis (Oksanen et al., 2015; VEGAN EnvFit function).
b Days the association of Nitrospira nxrB uwm–2 with comammox nxrB
From Start, Days following the start of a rearing cycle; Culled fish, the number of
fish removed from the system up to the point of sampling; System pH, pH in the rearing sequences, we further examined the biofilter for the presence
tank; ORP, oxidation reduction potential; Biofilter PSI is the pressure within the biofilter of Nitrospira-like amoA genes. We subsequently amplified a
manifold, in pounds per square inch.
c Percent variance explained by the first and second axes in the bacterial community
single Nitrospira-like amoA out of the biofilter samples, and
composition ordination.
phylogenetic inference placed this amoA on a monophyletic
branch with currently known Nitrospira amoA sequences, but in
a distinct cluster (Figure 7B) with a drinking water metagenome
In addition to the 16S rRNA gene community data, we contig (Pinto et al., 2015) and a “Crenothrix pmoA/amoA” Paddy
amplified, cloned, and sequenced nitrifying marker genes Soil Clone (KP218998; van Kessel et al., 2016). A link to ARB
representing the dominant nitrifying taxa in the UWM biofilter. databases containing these data may be found at https://github.
The archaeal amoA sequences (KX024777–KX024795) clustered com/rbartelme/ARB_dbs.
into two distinct genotypes, with an average nucleotide identity
ranging from 97 to 99%. Both genotypes placed phylogenetically Temporal and Spatial Quantification of
in the Nitrososphaera sister cluster (Figure 5), which includes Nitrification Marker Genes
the candidate genus, Nitrosocosmicus (Lehtovirta-Morley et al., We investigated the temporal and spatial stability of the nitrifying
2016), but the sequences were most closely related to the amoA organisms in the UWM biofilter by developing qPCR assays
genes from Archaeon G61 (97% nucleotide identity; KR233005). specific to identified amoA and nxrB genes. Within the ammonia-
Sequenced amplicons for betaproteobacterial amoA (KX024803– oxidizing community, the AOA and comammox-Nitrospira
KX024810) also revealed the presence of two AOB genotypes (amoA assay) had space-time abundance patterns distinct from
affiliated with Nitrosomonas. These Nitrosomonas genotypes that of the Nitrosomonas genotypes. For example, the AOA
were most closely related (99% identity) to environmental and comammox-Nitrospira were numerically dominant (range
sequences obtained from freshwater aquaria and activated sludge = 450–6500:1) to Nitrosomonas (combined UWM nitroso-1
(Figure 6). and nitroso-2 genotypes) across all samples (Figure 8; Table 3).
The UWM biofilter sand also harbored two phylogenetically The AOA and comammox-Nitrospira also had more stable
distinct and divergent clades of nxrB sequences (85–86% abundances over time [Coefficient of variation (CV) = 0.38 and
nucleotide identity between genotypes; KX024811–KX024822) 0.55 vs. 1.33 and 1.32 for nitroso-1 and nitroso-2; Figure 8], copy
affiliated with the genus Nitrospira. Nitrospira nxrB uwm-1 number concentrations that were less impacted by biofilter depth
formed a clade distinct from cultivated Nitrospira spp. (∼92% (Table 3), and comammox-Nitrospira were approximately 1.9x
nucleotide identity to Nitrospira bockiana). Nitrospira nxrB more abundant than AOA throughout the biofilter. Lastly, the
uwm-2 clustered phylogenetically with Nitrospira spp., which two Nitrosomonas amoA genotypes exhibited a strong temporal
have been implicated in complete nitrification (i.e., comammox; abundance correlation (Pearson’s R = 0.90, pseudo p = 0.0002)

FIGURE 4 | Depth comparison of bacterial biofilter community composition. A heatmap is depicted for all bacterial families with ≥1% relative abundance in
any sample. Taxon relative abundance was generated from V4–V5 16S rRNA gene sequencing and is indicated with a scale from 0 to 25%. The dendrogram
represents Bray-Curtis dissimilarity between sample community composition. Sample IDs are listed and sample depth is indicated by on the plot next to the
dendrogram. Sample names correspond to sample metadata in Table S1.
that was not shared with AOA or the comammox-Nitrospira For example, the model indicates ammonia oxidizer biomass
(Pearson’s R = 0.65 and 0.69, and pseudo p = 0.031 and 0.019, reaches near maximum by a mean cell residence time (MCRT)
respectively). of 20 days (Figure 10). At this 20-day MCRT, the model indicates
Within the nitrite-oxidizing community, the abundance of the ammonia removal rate measured could support ∼6.2X more
both Nitrospira genotypes (nxrB uwm-1 and uwm-2) was in cells than we observed (Figure 10).
the range of 108 CN/g sand, and each exhibited temporal
and spatial (depth) abundance stability (Table 3; Figure 8). The DISCUSSION
two genotypes also exhibited abundance co-variance across all
samples (Pearson’s R = 0.71, pseudo p = 0.0002). Despite these Biofilter Microbial Community Composition
abundance pattern similarities, the two genotypes had differential In this study, we generated data that deeply explored the
associations with other nitrifying taxa marker genes. Genotype microbial community composition for a production-scale
uwm-1, which is phylogenetically associated with strict nitrite- freshwater RAS nitrifying biofilter, expanding our understanding
oxidizers, had strong abundance co-variation with the AOA of the complexity of these systems beyond previous reports
amoA (Pearson’s R = 0.90, pseudo p ≤ 0.0001), while genotype (Sugita et al., 2005; Sauder et al., 2011; Blancheton et al.,
uwm-2 (phylogenetically associated with comammox-Nitrospira) 2013). This deeper coverage gave us the power to examine
had a stronger relationship to the Nitrospira amoA (Pearson’s R temporal and depth distributions for both total bacterial and
= 0.82, pseudo p ≤ 0.0001; Figure 9). archaeal communities and the potential nitrifying member
consortia therein. In previous studies of freshwater RAS
biofilters, Actinobacteria, Gammaproteobacteria, Plantomycetes,
Ammonia-Oxidizing Microorganism and Sphingobacteria were identified as dominant taxa,
Biomass Model while at more refined taxonomic levels Acinetobacteria,
The estimated cell densities for ammonia oxidizers in the biofilter Cetobacterium, Comamonas, Flectobacillus, Flavobacterium,
were modeled as a function of mean cell residence time (MCRT). and Hyphomicrobium were common (Sugita et al., 2005). All of
Since the biofilter MCRT was unknown, a range of values (1–30 these genera were present and relatively abundant (>0.5% total
days) was used in the model. The model suggests the combined community; genus level taxonomic breakdown in Table S2) in
estimated ammonia oxidizer cell densities (Nitrosomonas + AOA our biofilter sand samples, suggesting there may be selection
+ commamox-Nitrospira) could be supported by the ammonia pressures for heterotrophs that act universally across systems.
oxidation observed, and in fact over-estimated these densities. Some researchers have hypothesized that each RAS biofilter

FIGURE 5 | Ammonia-oxidizing Archaea consensus tree. A consensus phylogenetic tree was generated from maximum likelihood and Bayesian inference
phylogenetic reconstructions. Consensus tree support is indicated by colored circles at tree nodes. Collapsed nodes and assigned names are based off of Pester et al.
(2012). Clone and taxonomic names are followed by NCBI accession numbers. Ammonia-oxidizing archaea amoA sequences generated in this study are highlighted.
should have a unique microbial community composition shaped making robust comparisons across systems and identifying
by operational controls and components implemented in the underlying community composition trends that relate to system
RAS (Sugita et al., 2005; Blancheton et al., 2013). In support operations.
of this idea, many of the most abundant bacterial genera in Different components of RAS are expected to have unique
our system (e.g., Kribbella, Niabella, Chitinophaga, Byssovorax, environmental selective pressures, and thus multiple distinct
Hyphomicrobium) had not been reported as abundant in other microbial communities should be present within a single
systems. While it is likely true that each microbial community RAS. Our community data indicates there are consistent
assemblage will be unique among RAS biofilters, i.e., each and significant differences in the biofilter sand and water
biofilter has a unique “microbial fingerprint,” the low number of communities. These differences included community members
RAS biofilters with community composition information to date that were ubiquitous in, but nearly exclusive to the water samples.
and the low sequencing depth within existing studies, prohibits These taxa could be remnant members derived from previous

FIGURE 6 | Ammonia-oxidizing Bacteria consensus tree. A consensus phylogenetic tree was generated from maximum likelihood and Bayesian inference
phylogenetic reconstructions. Consensus tree support is indicated by colored circles at tree nodes. Collapsed nodes and assigned names are based off of Abell et al.
(2012). Clone and taxonomic names are followed by NCBI accession numbers. The clade containing Nitrosomonas amoA genotype, UWM nitroso-1 amoA is
highlighted in green, and UWM nitroso-2 amoA is highlighted in yellow.
components in the system (e.g., rearing tank, clarifier), but the an independent community develops on the biofilter media
high shear force in a fluidized sand bed may make for inconsistent (Blancheton et al., 2013).
passage of these inflow microorganisms. The water samples also Our time series indicates RAS biofilter bacterial community
had decreased representation of prominent sand-associated taxa, composition change correlates with environmental parameter
including most known nitrifiers, so studies sampling biofilter shifts related to fish growth (i.e., number of fish, water
outflow water would not represent accurately the microbial temperature, conductivity, oxidation-reduction potential, and
assemblages associated with nitrification. These observations feed size). This result is consistent with the hypothesis
support previous observations to the same effect, further lending that biofilter bacterial community variation follows feed
support to the idea that a transient planktonic microbial and fish growth driven shifts in the C/N ratio (Michaud
assemblage is constantly moving through RAS components while et al., 2006, 2014). The community variability is seemingly

FIGURE 7 | Consensus phylogenetic trees for Nitrospira-like (A) nxrB and (B) amoA genes. For the nxrB phylogeny, the consensus tree from Pester et al. (2014)
is illustrated. The UWM Biofilter and Candidatus Nitrospira nitrificans sequences were added to this phylogenetic reconstruction with the Quick-Add Parsimony tool of
the ARB package (Ludwig et al., 2004), so as not to change the tree topology. For the amoA phylogeny, a consensus phylogenetic tree was generated from maximum
likelihood and Bayesian inference phylogenetic reconstructions. Consensus tree support is indicated by colored circles at tree nodes. Clone names are followed by
NCBI accession numbers or a manuscript citation. In both trees, sequences generated in this study are highlighted with colored boxes.
confined to the non-nitrifying members of the biofilter, as the biofilter revealed distinct microbial communities in each sand
dominant nitrifying organisms changed little in composition stratum, suggesting a potential partitioning across physical
or abundance over time. Sampling different depths in the and chemical gradients within the biofilter. In contrast to the

FIGURE 8 | Nitrification marker gene concentration over time. Plot (A) illustrates amoA copy number (CN) per gram of biofilter sand and plot (B) nxrB CN per
gram of biofilter sand for all identified genotypes. Standard deviation of triplicate qPCR reactions is indicated for each sample. The x-axis indicates time, with timepoint
0 representing the beginning of one fish rearing cycle. Samples collected in the previous rearing cycle are labeled with negative values. See Table S1 for sample
metadata.
TABLE 3 | Nitrification marker gene concentrations in biofilter sand.
qPCR Assaya Bottom (CN/g)b Middle (CN/g) Surface (CN/g) Significanced
UWM AOA-Total (amoA)c 2.1 × 108 ± 0.2 × 108 2.6 × 108 ± 0.8 × 108 1.0 × 108 ± 0.06 × 108 χ 2 = 5.4 and p = 0.07
UWM Nitroso–1 (amoA) 4.6 × 105 ± 0.3 × 105 3.6 × 104 ± 1.3 × 104 4.5 × 104 ± 2.9 × 104 χ 2 = 5.6 and p = 0.06
UWM Nitroso–2 (amoA) 2.0 × 104 ± 0.4 × 104 4.0 × 103 ± 1.7 × 103 3.5 × 103 ± 1.9 × 103 χ 2 = 5.4 and p = 0.07
Nitrospira nxrB uwm-1 5.8 × 108 ± 1.0 × 108 7.4 × 108 ± 3.9 × 108 4.6 × 108 ± 1.3 × 108 χ 2 = 2.3 and p = 0.32
Nitrospira nxrB uwm-2 4.9 × 108 ± 1.8 × 108 4.6 × 108 ± 2.1 × 108 4.2 × 108 ± 1.4 × 108 χ 2 = 0.35 and p = 0.84
Comammox (amoA) 3.5 × 108 ± 0.7 × 108 3.9 × 108 ± 1.0 × 108 2.5 × 108 ± 0.9 × 108 χ 2 = 1.7 and p = 0.43
a Mean and standard deviation are listed.

b Bottom, middle, and surface depth categories are defined as: surface (∼1.32–1.42 m from biofilter base), middle (∼0.81–0.91 m from biofilter base), and bottom (∼0.15–0.30 m, from
biofilter base).
c For nxrB, n = 4, and for amoA n = 3. Corresponding samples are listed in Table S1.
d χ 2 and P-values from Kruskal–Wallis Rank Sum assessment of depth as a significant factor in nitrification marker gene distribution.
observed temporal variation, these differences were present both in Nitrosomonas and Nitrobacter bioreactors (Sedlacek et al.,
in the heterotrophic assemblages, and in the abundance of 2016). It is unknown whether these interactions extend to
nitrifiers. It appears this biofilter maintains a stable, but depth other ammonia and nitrite-oxidizing taxa or other systems, but
partitioned nitrifying community in the midst of a shifting the interplay between heterotrophs and nitrifiers as a means
bacterial community, whose composition is linked to variation to enhance nitrification rates in RAS should be investigated.
in nutrient inputs, ultimately stemming from the output of fish Further data across systems and over longer periods in a
growth. single system are also needed to bound “normal” vs. stochastic
Generally, the RAS biofilter heterotrophic microbial system variability and identify key taxa or community assembly
community is viewed only as competing with nitrifiers for principles governing RAS.
resources, and system design guidelines recommend operations
based on this premise (Okabe et al., 1995). However, this view Nitrifying Consortia
may confine further development of biofilter technology, as Prior to metagenomic studies, members of a few bacterial clades
it is becoming apparent that the heterotrophic community were believed to be responsible for ammonia oxidation.
context can play a broader role in nitrification. Our data clearly The isolation of the first ammonia-oxidizing archaeon,
indicates the heterotroph community varies substantially during Nitrosopumilus maritimus, altered global nitrification models
“typical” fish rearing cycles. It is possible under some scenarios (Könneke et al., 2005). AOA are ubiquitous in both natural
that these changes could impact nitrification. For example, and engineered environments and are seemingly differentiated
certain heterotrophs are known to enhance nitrification rates by niche from ammonia-oxidizing bacteria (AOB) based

to be competitive in systems with limited substrate influx, and

comammox Nitrospira have proven to be common in drinking
water systems (Pinto et al., 2015). Part of the initial discovery
of comammox included a comammox Nitrospira from a RAS
(van Kessel et al., 2015), but in the anoxic portion of a trickling
biofilter. Thus, RAS biofilters, which often have a municipal
water source and relatively low nutrient influx may be a common
reservoir of comammox Nitrospira colonization.
The physiology of the UWM RAS biofilter AOA cannot
be interpreted from our dataset, but both the AOA genotypes
cluster phylogenetically within the Nitrososphaera sister cluster,
which is represented mainly by cloned amoA sequences from
soil, sediment, and some AOA associated with freshwater
aquaria. Recently an organism given the name Candidatus
Nitrosocosmicus franklandus (Lehtovirta-Morley et al.,
2016) was isolated from the Nitrososphaera sister cluster.
Ca. Nitrosocosmicus spp. appear to be suited to tolerate higher
concentrations of ammonia and nitrite than other AOA, and
are capable of ureolytic growth (Lehtovirta-Morley et al., 2016),
both of which could be beneficial traits in RAS environments.
AOA, now have been detected in freshwater, brackish, and saline
RAS that also span a variety of cultured species, ranging from
FIGURE 9 | Heatmap of abundance pattern correlations for nitrifier finfish to crustaceans (Urakawa et al., 2008; Sauder et al., 2011;
genotypes. Pearson’s correlation coefficient values (r) are listed and colored
Sakami et al., 2012). Given the common AOA dominance over
according to the strength of the abundance correlation between marker genes
for each genotype. Purple colors indicate stronger correlations and green Nitrosomonas in RAS nitrifying biofilters, including in our study
colors indicate weaker correlations. system, a greater understanding of AOA ecophysiology is needed
to understand how system designs could be used to maximize
AOA capabilities.
Although AOA appear widespread in RAS biofilters, the
on ammonia concentration, where AOA outcompete AOB presence of AOA with comammox Nitrospira in our system
at relatively low concentrations (Hatzenpichler, 2012). This suggests understanding AOA physiology may be only a part
relationship appears to extend to freshwater biofilters, as it of understanding RAS biofilter nitrification. It is clear this
was shown recently that AOA dominate in freshwater aquaria environment generally favors the proliferation of organisms
biofilters when ammonia concentrations are low (<30 µM; thought to be high affinity, low substrate specialists and can
Pester et al., 2011). Our data support these previous findings, as support a complex nitrifying consortium. However, further work
AOA were 6 × 105 times more abundant than both Nitrosomonas is needed to understand how ammonia-oxidation partitions
genotypes in the UWM biofilter, which maintains similarly low between the various ammonia-oxidizers competing for substrate
influent ammonia concentrations (mean = 9 µM). AOA showed and how system operations can take advantage of potentially
little abundance variation with depth or over time (<3X change) flexible ammonia-oxidizer physiologies.
while Nitrosomonas exhibited an order of magnitude greater In our system, we did not detect Nitrobacter, whose
abundance during later periods in the fish rearing cycle and physiological constraints are often used when calculating
deeper in the biofilter (Table 3). System ammonia is highest late RAS biofiltration capacity. Instead we identified Nitrospira as
in the rearing cycle (Table S1) and presumably deeper in the the dominant nitrite-oxidizing bacteria (NOB). Nitrospira are
biofilter, which is nearest to the influent ports. generally considered K-strategist NOB favoring oligotrophic
Although, AOA were numerically dominant over AOB, a environments, while Nitrobacter are r-strategist copiotrophs
presumed third ammonia-oxidizer was also present in the (Nowka et al., 2015). Nitrospira uwm-1 exhibited a strong
biofilter sand matrix. Identification of Nitrospira-like amoA abundance pattern correlation with AOA, had abundances
(Figure 7B) in the biofilter and the strong correlation between roughly equal (∼108 nxrB CN/g sand) to that of the AOA,
the abundance of the Nitrospira nxrB uwm-2 gene and and clustered phylogenetically with known nitrite-oxidizing
this Nitrospira amoA, suggests a complete ammonia-oxidizing Nitrospira. Together, this suggests Nitrospira uwm-1 is the
Nitrospira spp. resides in the UWM biofilter. In fact, we found primary strict nitrite-oxidizing bacterium in this biofilter. The
that the comammox amoA was the most abundant ammonia- dominance of Nitrospira in this system and several other RAS
oxidizing gene in the biofilter (on average 1.9X that of AOA (Schreier et al., 2010; van Kessel et al., 2010; Auffret et al., 2013;
amoA). Similar to the AOA, the comammox Nitrospira exhibited Brown et al., 2013; Kruse et al., 2013) indicates there is a versatile
little abundance variation with depth or over time, which metabolic network driving RAS biofilter nitrification. For
suggests the AOA and comammox Nitrospira stably co-exist example, nitrite-oxidizing Nitrospira spp. possess a diverse array
throughout this system. The comammox reaction is predicted of metabolic pathways, and have been shown experimentally

FIGURE 10 | Model output of ammonia-oxidizer cell concentration as a function of biofilter mean cell residence time (MCRT). The red line indicates
ammonia-oxidizer cell abundance estimates from the mean change in ammonia concentration across the filter matrix as a function of mean cell residence time. The
shaded gray region represents the range of cell abundance estimates from the minimum and maximum observed ammonia removal rates. The horizontal dashed line
indicates qPCR estimated total ammonia-oxidizer abundance (ammonia-oxidizing Archaea + ammonia-oxidizing Bacteria + comammox Nitrospira) in the system.
to hydrolyze urea and cyanate to ammonia, thereby initiating and include an updated understanding of cross-feeding between
nitrification through cross-feeding with AOA/AOB. This process AOM and NOB (De Schryver and Vadstein, 2014; Daims et al.,
is counter to the supposed role of nitrite oxidizers solely as 2016).
converters of nitrite to nitrate (Daims et al., 2016). Whether This study builds upon the accumulating body of evidence
or not Nitrospira in RAS move nitrogen pools through these that biofilter microbial communities in freshwater RAS are
alternate pathways is not yet known. dynamic, diverse, and more distributed by resource availability
Given the diversity of nitrifiers and burgeoning understanding than is often considered in the design process. Our results
of nitrifier metabolic flexibility, it is possible that some of the along with others (Sakami et al., 2012; Brown et al., 2013)
identified ammonia-oxidizing organisms in our system were indicate the microorganisms carrying out nitrification in RAS are
not carrying out ammonia oxidation, as this scenario has been different than those used traditionally to model RAS nitrifying
observed in municipal wastewater treatment systems (Mußmann capacity. This disconnect suggests there is potential to further
et al., 2011). Our model indicates the measured ammonia fine-tune biofilter design to take advantage of these newly
removal could support the predicted ammonia-oxidizer biomass, discovered physiologies and alter start-up procedures so that
and in fact overestimated the number of ammonia oxidizing animal production objectives are matched to the nitrifying
cells present. This overestimation could be the result of the microorganisms most capable of meeting those demands.
model’s reliance on biomass production from traditional AOM Incorporating this knowledge would provide opportunities to
metabolisms, which many not represent accurately biomass develop new system operations, such as operating at a lower
production from ammonia oxidation for metabolically flexible pH (Hüpeden et al., 2016), and could move system optimization
ammonia-oxidizers or comammox Nitrospira (Costa et al., beyond that bound by current nitrification models. Yet,
2006). Also, the cell volume used in the model is based on many unknowns remain, including how differences in system
measurements of Candidatus Nitrosocosmicus franklandus, a scale, water properties, and system initiation with subsequent
relatively small microorganism; thus differences in cell size founder effects influence biofilter community composition,
across ammonia-oxidizing taxa also may be contributing to stability, and ultimately performance. Further use of microbial
the overestimation of biomass. In order to accurately predict ecological theory in aquaculture has the potential to extend
ammonia consumption to biomass production ratios, which are RAS capabilities, identify currently unrecognized interactions
used to constrain biofilter design, future models will need to between microorganisms and system design, and facilitate
account for the substrate kinetic differences between ammonia replicable zero discharge systems (De Schryver and Vadstein,
oxidizer metabolic pathways, differences in cell size among taxa, 2014).

AUTHOR CONTRIBUTIONS provided access to their RAS system and operator data, Katherine
Halmo who aided in DNA extraction, Jenny Fisher who provided
RB contributed to the development of research project goals, R script proofreading and code suggestions, Melinda Bootsma
carried out the lab work and most of the data analysis, and was the and Patricia Bower who were consulted during qPCR assay
primary author in writing and revising the manuscript. SM was development, Ameet Pinto and Brett Mellbye who provided
involved in writing and editing the manuscript and provided the Comammox amoA sequences and Nitrobacter spp. gDNA
primary source of funding. RN contributed to the development respectively, Steve Summerfelt who provided mean cell residence
of research project goals, provided data analysis, was involved in times for typical commercial-scale aquaculture biofilters, and
all of the writing and editing of the manuscript, and contributed Christopher E Lawson who provided valuable comments on
a source of project funding. earlier versions of the manuscript. We also appreciate the
technical expertise in massively parallel sequencing provided by
FUNDING The Marine Biological Laboratory at Woods Hole and the Great
Lakes Genomic Center at UW-Milwaukee. Finally, we would like
Funding for this work was provided by a University of Wisconsin to acknowledge the insightful discussions from colleagues and
System Incentive grant to the School of Freshwater Sciences and attendees of ICoN4 and ISME16.
through start-up laboratory funds to RN.
SUPPLEMENTARY MATERIAL
ACKNOWLEDGMENTS
The Supplementary Material for this article can be found
We appreciate the insight and technical assistance provided online at: http://journal.frontiersin.org/article/10.3389/fmicb.
by a number of colleagues, including: the Binkowski Lab, who 2017.00101/full#supplementary-material
REFERENCES Daims, H., Lebedeva, E. V., Pjevac, P., Han, P., Herbold, C., Albertsen, M., et al.
(2015). Complete nitrification by Nitrospira bacteria. Nature 528, 504–509.
Abell, G. C. J., Robert, S. S., Frampton, D. M. F., Volkman, J. K., Rizwi, F., doi: 10.1038/nature16461
Csontos, J., et al. (2012). High-throughput analysis of ammonia oxidiser Daims, H., Lücker, S., and Wagner, M. (2016). A new perspective on microbes
community composition via a novel, amoA-based functional gene array. PLoS formerly known as nitrite-oxidizing bacteria. Trends Microbiol. 2016, 1–14.
ONE 7:e51542. doi: 10.1371/journal.pone.0051542 doi: 10.1016/j.tim.2016.05.004
Auffret, M., Yergeau, É., Pilote, A., Proulx, É., Proulx, D., Greer, C. W., et al. De Schryver, P., and Vadstein, O. (2014). Ecological theory as a foundation
(2013). Impact of water quality on the bacterial populations and off-flavours to control pathogenic invasion in aquaculture. ISME J. 8, 2360–2368.
in recirculating aquaculture systems. FEMS Microbiol. Ecol. 84, 235–247. doi: 10.1038/ismej.2014.84
doi: 10.1111/1574-6941.12053 Edgar, R. C. (2004). MUSCLE: multiple sequence alignment with high
Badiola, M., Mendiola, D., and Bostock, J. (2012). Recirculating Aquaculture accuracy and high throughput. Nucleic Acids Res. 32, 1792–1797.
Systems (RAS) analysis: main issues on management and future challenges. doi: 10.1093/nar/gkh340
Aquac. Eng. 51, 26–35. doi: 10.1016/j.aquaeng.2012.07.004 Eren, A. M., Morrison, H. G., Lescault, P. J., Reveillaud, J., Vineis, J. H., and Sogin,
Bagchi, S., Vlaeminck, S. E., Sauder, L. A., Mosquera, M., Neufeld, J. M. L. (2015). Minimum entropy decomposition: unsupervised oligotyping for
D., and Boon, N. (2014). Temporal and spatial stability of ammonia- sensitive partitioning of high-throughput marker gene sequences. ISME J. 9,
oxidizing archaea and bacteria in aquarium biofilters. PLoS ONE 9:e113515. 968–979. doi: 10.1038/ismej.2014.195
doi: 10.1371/journal.pone.0113515 Eren, A. M., Vineis, J. H., Morrison, H. G., and Sogin, M. L. (2013). A
Barange, M., Merino, G., Blanchard, J. L., Scholtens, J., Harle, J., Allison, filtering method to generate high quality short reads using illumina paired-
E. H., et al. (2014). Impacts of climate change on marine ecosystem end technology. PLoS ONE 8:e66643. doi: 10.1371/annotation/afa5c40d-c6
production in societies dependent on fisheries. Nat. Clim. Chang. 4, 211–216. 04-46ae-84c4-82cb92193a5e
doi: 10.1038/nclimate2119 FAO (2014). The State of World Fisheries and Aquaculture.
Blancheton, J. P., Attramadal, K. J. K., Michaud, L., d’Orbcastel, E. R., and Fisher, J. C., Newton, R. J., Dila, D. K., and McLellan, S. L. (2015). Urban microbial
Vadstein, O. (2013). Insight into bacterial population in aquaculture systems ecology of a freshwater estuary of Lake Michigan. Elem. Sci. Anthr. 3:64.
and its implication. Aquac. Eng. 53, 30–39. doi: 10.1016/j.aquaeng.2012. doi: 10.12952/journal.elementa.000064
11.009 Harrell, F. E. Jr., (2016). Hmisc: harrell miscellaneous. R package version 4.0-2.
Brown, M. N., Briones, A., Diana, J., and Raskin, L. (2013). Ammonia- Available online at: https://CRAN.R-project.org/package=Hmisc
oxidizing archaea and nitrite-oxidizing nitrospiras in the biofilter of a Hatzenpichler, R. (2012). Diversity, physiology, and niche differentiation
shrimp recirculating aquaculture system. FEMS Microbiol. Ecol. 83, 17–25. of ammonia-oxidizing archaea. Appl. Environ. Microbiol. 78, 7501–7510.
doi: 10.1111/j.1574-6941.2012.01448.x doi: 10.1128/AEM.01960-12
Chen, S., Ling, J., and Blancheton, J.-P. (2006). Nitrification kinetics of Hovanec, T. A., and DeLong, E. F. (1996). Comparative analysis of nitrifying
biofilm as affected by water quality factors. Aquac. Eng. 34, 179–197. bacteria associated with freshwater and marine aquaria. Appl. Environ.
doi: 10.1016/j.aquaeng.2005.09.004 Microbiol. 62, 2888–2896.
Christman, G. D., Cottrell, M. T., Popp, B. N., Gier, E., and Kirchman, D. L. Hovanec, T. A., Taylor, L. T., Blakis, A., and Delong, E. F. (1998). Nitrospira-like
(2011). Abundance, diversity, and activity of ammonia-oxidizing prokaryotes bacteria associated with nitrite oxidation in freshwater aquaria. Appl. Environ.
in the coastal Arctic ocean in summer and winter. Appl. Environ. Microbiol. 77, Microbiol. 64, 258–264.
2026–2034. doi: 10.1128/AEM.01907-10 Hüpeden, J., Wegen, S., Off, S., Lücker, S., Bedarf, Y., Daims, H., et al. (2016).
Costa, E., Pérez, J., and Kreft, J. U. (2006). Why is metabolic labour divided Relative abundance of Nitrotoga in a biofilter of a cold freshwater aquaculture
in nitrification? Trends Microbiol. 14, 213–219. doi: 10.1016/j.tim.2006. plant appears to be stimulated by a slightly acidic pH-value. Appl. Environ.
03.006 Microbiol. 82, 1838–1845. doi: 10.1128/AEM.03163-15

Huse, S. M., Dethlefsen, L., Huber, J. A., Welch, D. M., Relman, D. A., Miller, M. A., Pfeiffer, W., and Schwartz, T. (2010). “Creating the CIPRES Science
and Sogin, M. L. (2008). Exploring microbial diversity and taxonomy Gateway for inference of large phylogenetic trees,” in 2010 Gateway Computing
using SSU rRNA hypervariable tag sequencing. PLoS Genet. 4:e1000255. Environments Workshop, GCE 2010 (New Orleans, LA), 1–8.
doi: 10.1371/journal.pgen.1000255 Mußmann, M., Brito, I., Pitcher, A., Sinninghe Damsté, J. S., Hatzenpichler, R.,
Huse, S. M., Mark Welch, D. B., Voorhis, A., Shipunova, A., Morrison, Richter, A., et al. (2011). Thaumarchaeotes abundant in refinery nitrifying
H. G., Eren, A. M. et al. (2014a). VAMPS: a website for visualization sludges express amoA but are not obligate autotrophic ammonia oxidizers.
and analysis of microbial population structures. BMC Bioinformatics 15:41. Proc. Natl. Acad. Sci. U.S.A. 108, 16771–16776. doi: 10.1073/pnas.1106427108
doi: 10.1186/1471-2105-15-41 Nelson, M. C., Morrison, H. G., Benjamino, J., Grim, S. L., and Graf, J. (2014).
Huse, S. M., Young, V. B., Morrison, H. G., Antonopoulos, D. A., Kwon, J., Dalal, Analysis, optimization and verification of Illumina-generated 16S rRNA gene
S., et al. (2014b). Comparison of brush and biopsy sampling methods of the amplicon surveys. PLoS ONE 9:e94249. doi: 10.1371/journal.pone.0094249
ileal pouch for assessment of mucosa-associated microbiota of human subjects. Newton, R. J., McLellan, S. L., Dila, D. K., Vineis, J. H., Morrison, H. G., Murat
Microbiome 2, 1–5. doi: 10.1186/2049-2618-2-5 Eren, A., et al. (2015). Sewage reflects the microbiomes of human populations.
Kindaichi, T., Ito, T., and Okabe, S. (2004). Ecophysiological interaction between MBio 6, e02574–e02514. doi: 10.1128/mBio.02574-14
nitrifying bacteria and heterotrophic bacteria in autotrophic nitrifying biofilms Nowka, B., Daims, H., and Spieck, E. (2015). Comparative oxidation
as determined by microautoradiography-fluorescence in situ hybridization kinetics of nitrite-oxidizing bacteria: nitrite availability as key factor
ecophysiological interaction between nitrifying Bacteria and He. Appl. Environ. for niche differentiation. Appl. Environ. Microbiol. 81, 745–753.
Microbiol. 70, 1641–1650. doi: 10.1128/AEM.70.3.1641-1650.2004 doi: 10.1128/AEM.02734-14
Klinger, D., and Naylor, R. (2012). Searching for solutions in aquaculture: Okabe, S., Hirata, K., and Watanabe, Y. (1995). Dynamic changes in spatial
charting a sustainable course. Annu. Rev. Environ. Resour. 37, 247–276. microbial distribution in mixed-population biofilms: experimental
doi: 10.1146/annurev-environ-021111-161531 results and model simulation. Water Sci. Technol. 32, 67–74.
Könneke, M., Bernhard, A. E., de la Torre, J. R., Walker, C. B., Waterbury, J. B., doi: 10.1016/0273-1223(96)00009-1
and Stahl, D. A. (2005). Isolation of an autotrophic ammonia-oxidizing marine Oksanen, J., Blanchet, F. G., Kindt, R., Legendre, P., Minchin, P. R., O’Hara, R.
archaeon. Nature 437, 543–546. doi: 10.1038/nature03911 B., et al. (2015). vegan: Community Ecology Package. Available online at: http://
Kruse, M., Keuter, S., Bakker, E., Spieck, E., Eggers, T., and Lipski, A. (2013). cran.r-project.org/package=vegan
Relevance and diversity of Nitrospira populations in biofilters of brackish RAS. Pester, M., Maixner, F., Berry, D., Rattei, T., Koch, H., Lücker, S., et al. (2014).
PLoS ONE 8:e64737. doi: 10.1371/journal.pone.0064737 NxrB encoding the beta subunit of nitrite oxidoreductase as functional and
Kuhn, D. D., Drahos, D. D., Marsh, L., and Flick, G. J. (2010). phylogenetic marker for nitrite-oxidizing Nitrospira. Environ. Microbiol. 16,
Evaluation of nitrifying bacteria product to improve nitrification 3055–3071. doi: 10.1111/1462-2920.12300
efficacy in recirculating aquaculture systems. Aquac. Eng. 43, 78–82. Pester, M., Rattei, T., Flechl, S., Gröngröft, A., Richter, A., Overmann, J., et al.
doi: 10.1016/j.aquaeng.2010.07.001 (2012). amoA-based consensus phylogeny of ammonia-oxidizing archaea and
Lee, D. E., Lee, J., Kim, Y. M., Myeong, J. I., and Kim, K. H. (2016). deep sequencing of amoA genes from soils of four different geographic regions.
Uncultured bacterial diversity in a seawater recirculating aquaculture system Environ. Microbiol. 14, 525–539. doi: 10.1111/j.1462-2920.2011.02666.x
revealed by 16S rRNA gene amplicon sequencing. J. Microbiol. 54, 296–304. Pester, M., Schleper, C., and Wagner, M. (2011). The Thaumarchaeota: an emerging
doi: 10.1007/s12275-016-5571-4 view of their phylogeny and ecophysiology. Curr. Opin. Microbiol. 14, 300–306.
Lehtovirta-Morley, L. E., Ross, J., Hink, L., Weber, E. B., Gubry-Rangin, C., doi: 10.1016/j.mib.2011.04.007
Thion, C., et al. (2016). Isolation of “Candidatus Nitrosocosmicus franklandus”, Piedrahita, R. H. (2003). Reducing the potential environmental impact of tank
a novel ureolytic soil archaeal ammonia oxidiser with tolerance to high aquaculture effluents through intensification and recirculation. Aquaculture
ammonia concentration. FEMS Microbiol. Ecol. 95, 1–10. doi: 10.1093/femsec/ 226, 35–44. doi: 10.1016/S0044-8486(03)00465-4
fiw057 Pinto, A. J., Marcus, D. N., Ijaz, Z., Bautista-de los Santos, Q. M., Dick, G. J.,
Lewis, W. M., and Morris, D. P. (1986). Toxicity of nitrite to fish: a reivew. Trans. and Raskin, L. (2015). Metagenomic evidence for the presence of comammox
Am. Fish. Soc. 115, 183–195. doi: 10.1577/1548-8659(1986)115<183:TONTF> nitrospira-like bacteria in a drinking water system. mSphere 1, e00054–e00015.
2.0.CO;2 doi: 10.1128/mSphere.00054-15
Ludwig, W., Strunk, O., Westram, R., Richter, L., Meier, H., Yadhukumar, et al. Poly, F., Wertz, S., Brothier, E., and Degrange, V. (2008). First exploration
(2004). ARB: a software environment for sequence data. Nucleic Acids Res. 32, of Nitrobacter diversity in soils by a PCR cloning-sequencing approach
1363–1371. doi: 10.1093/nar/gkh293 targeting functional gene nxrA. FEMS Microbiol. Ecol. 63, 132–140.
Luesken, F. A., van Alen, T. A., van der Biezen, E., Frijters, C., Toonen, G., doi: 10.1111/j.1574-6941.2007.00404.x,
Kampman, C., et al. (2011). Diversity and enrichment of nitrite-dependent Randall, D., and Tsui, T. K. (2002). Ammonia toxicity in fish. Mar. Pollut. Bull. 45,
anaerobic methane oxidizing bacteria from wastewater sludge. Appl. Microbiol. 17–23. doi: 10.1016/S0025-326X(02)00227-8
Biotechnol. 92, 845–854. doi: 10.1007/s00253-011-3361-9 R Core Team (2014). R: A Language and Environment for Statistical Computing.
Martin, M. (2011). Cutadapt removes adapter sequences from high-throughput Available online at: http://www.r-project.org/
sequencing reads. EMBnet J. 17, 10–12. doi: 10.14806/ej.17.1.200 Rice, P., Longden, I., and Bleasby, A. (2000). EMBOSS: the European
Martins, C. I. M., Eding, E. H., Verdegem, M. C. J., Heinsbroek, L. T. N., molecular biology open software suite. Trends Genet. 16, 276–277.
Schneider, O., Blancheton, J. P., et al. (2010). New developments in recirculating doi: 10.1016/S0168-9525(00)02024-2
aquaculture systems in Europe: a perspective on environmental sustainability. Ronquist, F., Teslenko, M., Van Der Mark, P., Ayres, D. L., Darling, A.,
Aquac. Eng. 43, 83–93. doi: 10.1016/j.aquaeng.2010.09.002 Höhna, S., et al. (2012). Mrbayes 3.2: efficient bayesian phylogenetic inference
Meyer, J. L., Akerman, N. H., Proskurowski, G., and Huber, J. A. (2013). and model choice across a large model space. Syst. Biol. 61, 539–542.
Microbiological characterization of post-eruption “snowblower” vents at axial doi: 10.1093/sysbio/sys029
seamount, juan de fuca ridge. Front. Microbiol. 4:153. doi: 10.3389/fmicb.2013. Rotthauwe, J. H., Witzel, K. P., and Liesack, W. (1997). The ammonia
00153 monooxygenase structural gene amoA as a functional marker: molecular
Michaud, L., Blancheton, J. P., Bruni, V., and Piedrahita, R. (2006). Effect fine-scale analysis of natural ammonia-oxidizing populations. Appl. Environ.
of particulate organic carbon on heterotrophic bacterial populations and Microbiol. 63, 4704–4712.
nitrification efficiency in biological filters. Aquac. Eng. 34, 224–233. Sakami, T., Andoh, T., Morita, T., and Yamamoto, Y. (2012). Phylogenetic
doi: 10.1016/j.aquaeng.2005.07.005 diversity of ammonia-oxidizing archaea and bacteria in biofilters
Michaud, L., Lo Giudice, A., Interdonato, F., Triplet, S., Ying, L., and of recirculating aquaculture systems. Mar. Genomics 7, 27–31.
Blancheton, J. P. (2014). C/N ratio-induced structural shift of bacterial doi: 10.1016/j.margen.2012.04.006
communities inside lab-scale aquaculture biofilters. Aquac. Eng. 58, 77–87. Sauder, L. A., Engel, K., Stearns, J. C., Masella, A. P., Pawliszyn, R., and Neufeld, J.
doi: 10.1016/j.aquaeng.2013.11.002 D. (2011). Aquarium nitrification revisited: thaumarchaeota are the dominant

ammonia oxidizers in freshwater aquarium biofilters. PLoS ONE 6:e23281. van Kessel, M. A. H. J., Mesman, R. J., Arshad, A., Metz, J. R., Tom Spanings,
doi: 10.1371/journal.pone.0023281 F. A., van Dalen, S. C. M., et al. (2016). Branchial nitrogen cycle symbionts
Schreier, H. J., Mirzoyan, N., and Saito, K. (2010). Microbial diversity of biological can remove ammonia in fish gills. Environ. Microbiol. Rep. 8, 590–594.
filters in recirculating aquaculture systems. Curr. Opin. Biotechnol. 21, 318–325. doi: 10.1111/1758-2229.12407
doi: 10.1016/j.copbio.2010.03.011 van Kessel, M. A. H. J., Speth, D. R., Albertsen, M., Nielsen, P. H., Op den Camp, H.
Sedlacek, C. J., Nielsen, S., Greis, K. D., Haffey, W. D., Revsbech, N. P., J. M., Kartal, B., et al. (2015). Complete nitrification by a single microorganism.
Ticak, T., et al. (2016). The effect of bacterial community members Nature 528, 555–559. doi: 10.1038/nature16459
on the proteome of the ammonia-oxidizing bacterium Nitrosomonas van Kessel, M. A. H. J., Harhangi, H. R., van de Pas-Schoonen, K., van
sp. Is79. Appl. Environ. Microbiol. 82, 4776–4788. doi: 10.1128/aem. de Vossenberg, J., Flik, G., Jetten, M. S. M., et al. (2010). Biodiversity
01171-16 of N-cycle bacteria in nitrogen removing moving bed biofilters for
Stamatakis, A. (2014). RAxML version 8: a tool for phylogenetic analysis freshwater recirculating aquaculture systems. Aquaculture 306, 177–184.
and post-analysis of large phylogenies. Bioinformatics 30, 1312–1313. doi: 10.1016/j.aquaculture.2010.05.019
doi: 10.1093/bioinformatics/btu033 Verdegem, M. C. J., Bosma, R. H., and Verreth, J. A. J. (2006). Reducing water
Sugita, H., Nakamura, H., and Shimada, T. (2005). Microbial communities use for animal production through aquaculture. Int. J. Water Resour. Dev. 22,
associated with filter materials in recirculating aquaculture systems of 101–113. doi: 10.1080/07900620500405544
freshwater fish. Aquaculture 243, 403–409. doi: 10.1016/j.aquaculture.2004. Wertz, S., Poly, F., Le Roux, X., and Degrange, V. (2008). Development and
09.028 application of a PCR-denaturing gradient gel electrophoresis tool to study the
Timmons, M. B., and Ebeling, J. M. (2013). Recirculating Aquaculture, 3rd Edn. diversity of Nitrobacter-like nxrA sequences in soil. FEMS Microbiol. Ecol. 63,
Ithaca, NY: Ithaca Publishing Company, LLC. 261–271. doi: 10.1111/j.1574-6941.2007.00416.x
Tourna, M., Freitag, T. E., Nicol, G. W., and Prosser, J. I. (2008). Growth,
activity and temperature responses of ammonia-oxidizing archaea Conflict of Interest Statement: The authors declare that the research was
and bacteria in soil microcosms. Environ. Microbiol. 10, 1357–1364. conducted in the absence of any commercial or financial relationships that could
doi: 10.1111/j.1462-2920.2007.01563.x be construed as a potential conflict of interest.
Untergasser, A., Cutcutache, I., Koressaar, T., Ye, J., Faircloth, B. C., Remm, M.,
et al. (2012). Primer3-new capabilities and interfaces. Nucleic Acids Res. 40, Copyright © 2017 Bartelme, McLellan and Newton. This is an open-access article
e115. doi: 10.1093/nar/gks596 distributed under the terms of the Creative Commons Attribution License (CC BY).
Urakawa, H., Tajima, Y., Numata, Y., and Tsuneda, S. (2008). Low temperature The use, distribution or reproduction in other forums is permitted, provided the
decreases the phylogenetic diversity of ammonia-oxidizing archaea and original author(s) or licensor are credited and that the original publication in this
bacteria in aquarium biofiltration systems. Appl. Environ. Microbiol. 74, journal is cited, in accordance with accepted academic practice. No use, distribution
894–900. doi: 10.1128/AEM.01529-07 or reproduction is permitted which does not comply with these terms.

Fmicb 08 00101

Uploaded by

Document Informationclick to expand document informationm

Document Informationclick to expand document information

Copyright:

Available Formats

Fmicb 08 00101

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Fmicb 08 00101

Uploaded by

Copyright:

Available Formats

ORIGINAL RESEARCH

published: 30 January 2017

School of Freshwater Sciences, University of Wisconsin-Milwaukee, Milwaukee, WI, USA

Recirculating aquaculture systems (RAS) are unique engineered ecosystems that

Frontiers in Microbiology | www.frontiersin.org 1 January 2017 | Volume 8 | Article 101

INTRODUCTION The non-nitrifying component of RAS biofilter communities

Frontiers in Microbiology | www.frontiersin.org 2 January 2017 | Volume 8 | Article 101

16S rRNA Gene Sequencing

Frontiers in Microbiology | www.frontiersin.org 3 January 2017 | Volume 8 | Article 101

Frontiers in Microbiology | www.frontiersin.org 4 January 2017 | Volume 8 | Article 101

Frontiers in Microbiology | www.frontiersin.org

January 2017 | Volume 8 | Article 101

Frontiers in Microbiology | www.frontiersin.org 6 January 2017 | Volume 8 | Article 101

Chthoniobacter, Niabella, and Chitinophaga were the most

Frontiers in Microbiology | www.frontiersin.org 7 January 2017 | Volume 8 | Article 101

TABLE 2 | Environmental variable to bacterial community composition

Variablea,b Dim1 Dim2 R2 Pr(>r)

Days From Start 0.836 0.548 0.94 0.002

Frontiers in Microbiology | www.frontiersin.org 8 January 2017 | Volume 8 | Article 101

Frontiers in Microbiology | www.frontiersin.org 9 January 2017 | Volume 8 | Article 101

Frontiers in Microbiology | www.frontiersin.org 10 January 2017 | Volume 8 | Article 101

Frontiers in Microbiology | www.frontiersin.org 11 January 2017 | Volume 8 | Article 101

Frontiers in Microbiology | www.frontiersin.org 12 January 2017 | Volume 8 | Article 101

TABLE 3 | Nitrification marker gene concentrations in biofilter sand.

qPCR Assaya Bottom (CN/g)b Middle (CN/g) Surface (CN/g) Significanced

a Mean and standard deviation are listed.

Frontiers in Microbiology | www.frontiersin.org 13 January 2017 | Volume 8 | Article 101

to be competitive in systems with limited substrate influx, and

Frontiers in Microbiology | www.frontiersin.org 14 January 2017 | Volume 8 | Article 101

Frontiers in Microbiology | www.frontiersin.org 15 January 2017 | Volume 8 | Article 101

Frontiers in Microbiology | www.frontiersin.org 16 January 2017 | Volume 8 | Article 101

Frontiers in Microbiology | www.frontiersin.org 17 January 2017 | Volume 8 | Article 101

Frontiers in Microbiology | www.frontiersin.org 18 January 2017 | Volume 8 | Article 101

You might also like