relationships between nitrogen transformation rates and gene abundance in a riparian buffer soil
TRANSCRIPT
Relationships Between Nitrogen Transformation Rates and GeneAbundance in a Riparian Buffer Soil
Lin Wu • Deanna L. Osmond • Alexandria K. Graves •
Michael R. Burchell • Owen W. Duckworth
Received: 7 February 2012 / Accepted: 20 July 2012 / Published online: 22 August 2012
� Springer Science+Business Media, LLC 2012
Abstract Denitrification is a critical biogeochemical
process that results in the conversion of nitrate to volatile
products, and thus is a major route of nitrogen loss from
terrestrial environments. Riparian buffers are an important
management tool that is widely utilized to protect water
from non-point source pollution. However, riparian buffers
vary in their nitrate removal effectiveness, and thus there is
a need for mechanistic studies to explore nitrate dynamics
in buffer soils. The objectives of this study were to
examine the influence of specific types of soluble organic
matter on nitrate loss and nitrous oxide production rates,
and to elucidate the relationships between these rates and
the abundances of functional genes in a riparian buffer soil.
Continuous-flow soil column experiments were performed
to investigate the effect of three types of soluble organic
matter (citric acid, alginic acid, and Suwannee River dis-
solved organic carbon) on rates of nitrate loss and nitrous
oxide production. We found that nitrate loss rates increased
as citric acid concentrations increased; however, rates of
nitrate loss were weakly affected or not affected by the
addition of the other types of organic matter. In all
experiments, rates of nitrous oxide production mirrored
nitrate loss rates. In addition, quantitative polymerase chain
reaction (qPCR) was utilized to quantify the number of
genes known to encode enzymes that catalyze nitrite
reduction (i.e., nirS and nirK) in soil that was collected at
the conclusion of column experiments. Nitrate loss and
nitrous oxide production rates trended with copy numbers
of both nir and 16s rDNA genes. The results suggest that
low-molecular mass organic species are more effective at
promoting nitrogen transformations than large biopolymers
or humic substances, and also help to link genetic potential
to chemical reactivity.
Keywords Riparian buffer � Nitrogen � Denitrification �Quantitative PCR � Nitrous oxide � Nitrate
Introduction
Riparian buffers are an important management tool for
reducing the impact of agricultural nonpoint source nitrate
(NO3-) pollution on water quality. These buffer strips of
uncultivated land bordering drainages have proven effec-
tive at reducing NO3- fluxes to surface waters; however,
buffers vary widely in their NO3- removal efficacy (Hill
1996; Smith and others 2006; Lowrance and others 1997;
Osborne and Kovacic 1993; Mayer and others 2007). At
the Center for Environmental Farming (CEFS; Goldsboro,
NC), which contains a long term monitoring site associated
with this study (Dukes and others 2002; King 2006; Knies
2009), the type of vegetation grown on riparian buffers has
been implicated as a key factor in controlling buffer
effectiveness, although the mechanism of action has not
been conclusively demonstrated (King 2006). Riparian
vegetation also has been widely reported to be a key factor
in controlling pollutant dynamics, partially by providing
L. Wu � D. L. Osmond � A. K. Graves � O. W. Duckworth (&)
Department of Soil Science, North Carolina State University,
Raleigh, NC 27695-7619, USA
e-mail: [email protected]
Present Address:L. Wu
Department of Statistics, University of North Carolina,
Chapel Hill, NC, USA
M. R. Burchell
Department of Biological and Agricultural Engineering, North
Carolina State University, Raleigh, NC 27695-7625, USA
123
Environmental Management (2012) 50:861–874
DOI 10.1007/s00267-012-9929-z
different types of organic carbon (C) that may affect soil
biogeochemistry in the riparian zone (Dosskey and others
2010).
The riparian zone has been identified as an area that may
have particularly high rates of denitrification (Hedin and
others 1998). Although denitrification is notoriously diffi-
cult to assess because of the nature of the products and the
complicated web of environmental variables that partially
control the process (Firestone and Davidson 1989; Knowles
1982; Myrold and Tiedje 1985; Nommik 1956), it is gen-
erally accepted that availability of NO3- and organic C are
the two major limiting factors for denitrification under
anaerobic conditions (Greenan and others 2006; Rivett and
others 2008; Starr and Gillham 1993; Hill and others 2000;
Bradley and others 1992). Significant relationships have
been observed between denitrification rates and various
operationally defined C fractions, including soluble,
extractable, and mineralizable C (Beauchamp and others
1980; Bijay-Singh and Whitehead 1988; Burford and
Bremner 1975; Davidson and others 1987; Hill and Cardaci
2004; Myrold and Tiedje 1985; Stanford and others 1975;
McCarty and Bremner 1992; Pintar and Lobnik 2005;
Hernandez and Mitsch 2007). However, which fractions of
soil organic matter correlate with reactivity differs between
studies, emphasizing the variability in the bioavailability of
soluble C and the possible influence of other edaphic and
environmental factors on the process. Because many con-
voluting factors control denitrification, continuous-flow soil
column experiments, which have been successfully used to
quantify rates of NO3- loss and nitrous oxide (N2O) pro-
duction at specifically defined chemical and hydrological
conditions, are a powerful tool to study the effects of C
concentration and type on denitrification (Knies 2009;
Pavel and others 1996; Willems and others 1997). Although
it is difficult to truly decouple N transformations, these
columns allow for the isolation of environmental variables
and control of the type of organic C present.
Molecular biology techniques provide insights into
denitrification that may be complementary to chemical
measurements of N transformations (Wallenstein and others
2006). Biological denitrification involves four enzyme
catalyzed reductions: NO3- ? NO2
- ? NO ? N2O ?N2. A key step in the denitrification process is the con-
version of NO2- to NO (nitric oxide), which is the first step
in the reaction sequence that produces a gaseous product
that can be volatized and exported from the buffer system.
This reaction is catalyzed by enzymes encoded by nitrite
(NO2-) reductase genes (i.e., nirK and nirS) that can be
exploited as molecular biomarkers to study biological
denitrification (Wallenstein and others 2006; Throback and
others 2004; Philippot 2002). Although studies have sug-
gested that denitrification rates trend with bacterial counts
(Miller and others 2012; Ardakani and others 1975;
Bowman and Focht 1974; Myrold and Tiedje 1985), others
have not seen a clear relationship between denitrification
rate and microbial community composition (Henry and
others 2008; Mounier and others 2004; Groffman and others
2006). Thus, the relationship between chemical denitrifi-
cation rate and microbial activity is poorly understood.
In the present study, column experiments were utilized
to evaluate NO3- loss and N2O production rates in a
riparian buffer soil as a function of dissolved organic C
type and concentration. In addition, we utilized quantitative
PCR to measure the copy number of nirS and nirK func-
tional genes to explore the relationship between N trans-
formation rates and gene abundance. The specific
objectives of this study were to: (1) determine the effects of
specific organic molecules on rates of N reduction (NO3-
loss and N2O oxide production rates) and (2) quantitatively
link this chemical reactivity to genetic potential (gene copy
numbers) in a riparian buffer soil. The results provide
specific information about the effects of different classes of
organic molecules on the biology and chemistry of N
transformations in anaerobic soils.
Materials and Methods
Soil Materials
A soil was collected in May 2009 from a fescue-vegetated
riparian buffer located adjacent to a drainage ditch at the
Center for Environmental Farming Systems (CEFS) in
Goldsboro, North Carolina. The buffer vegetation was
established in 1998, with the adjacent field historically used
as a cattle pasture and spray field for swine waste (Dukes
and others 2002). This site has been the subject of a long-
term nutrient dynamics monitoring project (Dukes and
others 2002; King 2006; Knies 2009). Soil was collected
from the riparian buffer by using a hand auger from a depth
of 3 m, which is the location of the aquifer that was the
subject of monitoring studies. Soil samples were kept on
ice during transportation and stored at 4 �C in the dark until
packed into soil columns. Based on the county soil map,
buffer soil was classified as Lumbee sandy loam (Barnhill
and others 1974). Our analysis of texture via the hydrom-
eter method and sieving (Gee and Or 2002) revealed the
soil to be a sand (96.7 % sand, 1.0 % silt, 2.3 % clay).
Total N and organic C concentrations were 0.02 % and
0.07 %, respectively, as determined by Perkin Elmer 2400
CHNS elemental analyzer.
Experimental Design
All solutions used in the study were made with type I
deionized water (DI) with a resistivity of 18.3 MX cm. All
862 Environmental Management (2012) 50:861–874
123
influent solutions were boiled and then continuously
purged with humidified argon (Ar) gas to exclude oxygen.
A NO3- concentration of 5.0 mg N L-1 (as KNO3) was
used in all experiments. This value was chosen because it is
an environmentally relevant concentration that is repre-
sentative of NO3- concentrations typically found in aqui-
fers at CEFS (King 2006; Knies 2009). Field measurements
have suggested that organic C availability limits denitrifi-
cation in these buffers (Knies 2009). Thus, influent solu-
tions also contained different concentrations of three
different organic C sources, as described below.
Three types of organic matter were utilized as model
compounds to represent broader families of organic mol-
ecules. Citric acid (C6H8O7, Acros Organics), a low
molecular mass organic acid (LMMOA), was chosen to
represent metabolites that are commonly exuded by plants
(Jones 1998; Curl and Truelove 1986; Jones and others
2003). Alginic acid (approximately [C6H8O6]n, Acros
Organics), a large macromolecule composed of uronic
sugar acids, was selected to represent environmentally
common microbial polysaccharides (Perry and others 2006;
Perry and others 2004). To represent soluble humic sub-
stances, Suwannee River dissolved organic carbon
(SRDOC) was obtained from International Humic Acid
Society (Suwannee River natural organic matter sample
1R101 N) and used without further purification. This
reference material has been extensively characterized,
with details available on the supplying society’s website
(http://www.ihss.gatech.edu/). Most notably, the SRDOC
has been found to be composed of 52.47 % C and 1.10 %
N by mass. For each type of organic molecule, experiments
were conducted with 4.0 mg C L-1, 8.0 mg C L-1, and
16.0 mg C L-1. A control experiment was also conducted
with no C added to the solution. Initial solution pH was 5.7
and 5.5 for alginic acid and SRDOC. To keep pH constant
between trials and avoid confounding effects from low pH
(Groffman and others 1991), inlet solutions containing
citric acid were adjusted to pH 6.0 by addition of NaOH.
Sampling and Analysis
Columns were constructed from polyvinyl chloride (PVC)
pipe that was 13.2 cm long and 3.0 cm in diameter. Soil
was packed into columns to a bulk density of 1.6 g cm-3, a
representative value for soil at CEFS (Dukes 2000). Col-
umns were sealed with rubber stoppers lined with fiberglass
fabric and equipped with plastic tubing outlets. Stoppers
were attached with waterproof polyurethane-based glue to
columns to ensure a tight fit.
Anaerobic continuous flow column experiments were
conducted to measure the rates of NO3- loss and N2O
production in the presence of different forms and concen-
trations of organic C (Pavel and others 1996; Knies 2009).
Boiled and Ar-purged solutions were pumped through soil
columns via fluorinated ethylene propylene (FEP) tubing
wrapped in aluminum foil to exclude atmospheric gases,
except for a short stretch of tygon tubing that was within a
peristaltic pump (Manostat). These precautions minimized
contamination from atmospheric gases, thus keeping col-
umns anaerobic. All experiments were conducted at
ambient laboratory temperature (25 ± 2 �C).
For each C type and concentration, columns experi-
ments were conducted for six days, which was sufficient
for the effluent to reach a steady-state concentration of
dissolved NO3- (cf. days 3-6 in Fig. 1a); all experiments
were conducted in triplicate. Each day, approximately
30 ml of effluent was collected in 50 ml disposable plastic
beakers over a timed interval (ca. 30 min). The mass of
effluent was measured to determine specific flow rate,
0
1
2
3
4
5
6
[NO
3- ] (m
g N
L-1)
Time (day)
0
1
2
3
4
5
6
Time (day)
0
1
2
3
4
5
6
1 2 3 4 5 6
1 2 3 4 5 6
1 2 3 4 5 6
Time (day)
a
b
c
[NO
3- ] (m
g N
L-1)
[NO
3- ] (m
g N
L-1)
influent
effluent of triplicate columns
influent
effluent of triplicate columns
influent
effluent of triplicate columns
Fig. 1 NO3- concentrations in the influent (9) and effluent (open
markers) in experiments containing 8 mg C L-1 a citric acid, b alginic
acid, and c SRDOC. Lines (solid = influent and dashed = effluent)
are presented to guide the eye and do not represent a model fit
Environmental Management (2012) 50:861–874 863
123
which was 19.0–32.3 ml h-1 for all experiments. The flow
rate was chosen to optimize residence time such that the
system would have a reasonable limit of detection for
NO3- loss rate, but would keep the steady-state NO3
-
concentration from becoming depleted ([NO3-] [
2 mg N L-1). The pH of effluent samples was measured
using an Accumet Excel pH/conductivity meter (XL20).
During experiments containing alginic acid and SRDOC,
pH change from the influent to effluent was pH \ 0.5; for
citric acid experiments, the change in pH = 0.3–1.0. In all
cases, pH drift was upward. Effluent samples were filtered
by syringe through 0.22 lm nylon filters (Millipore), fro-
zen, and stored at -20 �C for further analysis.
Effluent from each column (ca. 5 ml) was collected via a
hypodermic needle into a sealed vacutainer tube (Becton,
Dickinson and Company; 22.4 ml capacity) for determi-
nation of N2O. Attempts to utilize evacuated vacutainers
were unsuccessful because the vacuum produced suction
on the column outlet that increased flow rates and fre-
quently caused column clogging, presumably due to the
dislodging of soil particles. Therefore, solution was col-
lected into air-filled vacutainers and the background con-
tribution from air was subtracted, as described below.
Samples were analyzed for NO3-, NO2
-, and N2O to
determine rates of NO3- loss and N2O production. For
determination of NO3- loss, the NO3
- concentrations of
the influent and effluent solutions were determined color-
metrically with a sulfanilamide colorimetric reagent using
flow injection analysis (QuikChem Method 10107-01-A,
Lachat Instruments QuikChem 8000) with detection limit
of 0.05 mg N L-1. In experiments with measureable NO3-
loss, NO2- concentration in the effluent was also measured
colormetrically via the same method as NO3-. The NO2
-
concentrations were less than 1.00 mg N L-1 in all cases,
with most below the detection limit of 0.05 mg N L-1
(data not shown).
For analysis of N2O production, 10 ml of gas was man-
ually extracted from the headspace of vacutainer tube with a
gastight syringe and injected into a gas chromatograph
(Hewlett Packard 5890 GC-ECD, injector tempera-
ture = 60 �C). A Henry’s Law calculation (KH = 0.0257
M atm-1; Stumm and Morgan 1996) indicated that ca. 85 %
of N2O will be present in the headspace gas phase at equi-
librium. It should be noted that this calculation was specific
to volume of both the container and water sample, and cannot
necessarily be generalized to others systems. The concen-
tration of dissolved N2O resulting from partitioning from the
effluent in the gas phase was determined by difference from
the ambient background concentration (320 ppbV), thus
yielding an effective detection limit of 50 ppbV. The con-
centration of N2O that was dissolved in the effluent was then
calculated by assuming the headspace gas concentration was
originally dissolved in the solution. We acknowledge that
this approach contains an approximation (i.e., that gas phase
N2O accounts for all N2O in the system) that may result in a
small systematic underreporting of N2O production.
The total dissolved C concentration in effluent samples
from citric acid treatments was measured with TOC/N
Analyzer (Shimadzu Scientific Instruments, Japan). The
average effluent dissolved C concentration was
2.6 mg L-1, 4.0 mg L-1, and 4.7 mg L-1 for experiments
initially containing 4.0 mg C L-1, 8.0 mg C L-1, and
16.0 mg C L-1 citric acid, respectively. These values
suggest that the C concentration was not completely
depleted in the column by microbial processes.
Calculation of NO3- Loss and N2O Production Rates
The overall mass-normalized NO3- loss rate (RNO�
3;
mg N h-1 g-1) was calculated using the following equa-
tion (Knies 2009; Pavel and others 1996):
RNO�3¼
q� D NO�3� �
mð1Þ
where q was flow rate (L h-1), D[NO3-] was the change of
NO3- concentration (mg N L-1) between inflow and
effluent, and m was the mass of soil packed into the column
(g). The rate calculation was based on the steady-state
concentration, which typically was the final 4-5 days of the
experiment. The limit of detection (LOD) of
RNO�3
= 4 9 10-5 mg N h-1 g-1 was calculated via Eq. 1
by determining the minimum reliable difference between
influent and effluent (0.1 mg N L-1).
Similarly, the N2O production rate (RN2O;
mg N h-1 g-1) was calculated as follows:
RN2O ¼q� D N2O½ �
mð2Þ
where q was the flow rate (L h-1), DC (N2O) was the
concentration of N2O dissolved in column effluent
(mg N L-1), and m was the initial mass of soil in the
column (g). The LOD for N2O production was estimated to
be RN2O = 10-8 mg N h-1 g-1 loss rate, as calculated via
Eq. 2 from the detection limit against background. For both
NO3- loss and N2O production rates, uncertainty was
reported as twice the standard deviation (±1 standard
deviation) of three replicates of experiments.
Quantitative PCR
At the conclusion of experiments, soil samples were col-
lected in beakers, homogenized by mechanical mixing with
a sterile spatula, and stored at 4 �C for no more than
2 weeks, at which point DNA was extracted from soil in
each column. The PowerSoilTM
DNA Isolation Kit (Mo BIO
Laboratories, USA) was used to extract and purify genomic
864 Environmental Management (2012) 50:861–874
123
DNA from 1.0 g homogenized soil (wet weight) from each
soil column; this procedure was completed twice for each
column. Extracted DNA was quantified by spectropho-
tometry at 260 nm using a Nanodrop spectrophotometer
(Thermo scientific, USA). DNA was stored at -20 �C until
amplified by quantitative PCR (qPCR).
Previously published primers were used in this study to
amplify 16 rDNA and nir genes (Kandeler and others 2006;
Throback and others 2004; Henry and others 2004). The
primer sequence and annealing temperature for corre-
sponding targets are shown in Table 1. Amplification of
qPCR products conducted by using a MasterCycler� EP
Realplex cycler (Eppendorf, USA).
Amplifications were performed in 25 lL reaction mix-
tures using QuantiFastTM
SYBR� Green PCR Kit (QIA-
GEN, Valencia, CA). The reaction mixture contained
2.5 lL of each primer for gene (1 lM final concentration
for each), 12.5 lL of 2 9 master mix (including HotStar
Taq Plus DNA polymerase, QuantiFast SYBR Green PCR
buffer, dNTP mix with dUTP, SYBR Green I, ROX dye
and 5 mM MgCl2), 2 lL of DNA template (ca. 10 ng total
DNA), and RNase-free water to complete the 25 lL
volume.
Thermocycler conditions for nirS qPCR were 300 s at
95 �C for enzyme activation as recommended by the
manufacturer (QuantiFastTM
SYBR� Green PCR Kit,
QIAGEN), followed by 35 cycles of 45 s at 94 �C for
denaturation, 45 s at 57 �C for annealing, and 60 s at
72 �C for extension and data acquisition. The reaction
was completed after 300 s at 72 �C for final extension and
a final temperature gradient step from 60 �C to 95 �C
with an increase of 0.2 deg s-1 used to obtain a specific
denaturation curve. The conditions for nirK and 16s
rDNA qPCR were similar as those for nirS, except that
the annealing temperatures were 58 �C and 53 �C,
respectively (Henry and others 2004; Nogales and others
2002). Quantitative PCR efficiencies for were 91 %,
101 %, and 94 % for 16s rDNA, nirK, and nirS genes,
respectively.
The purity of amplified products was verified by the
observation of a single melting curve or the presence of a
unique band of the expected size in a 2 % (weight base)
agarose gel stained with ethidium bromide. Standard
curves for absolute quantitation were obtained from serial
dilutions of a known amount of genomic DNA (obtained
from the American Type Culture Collection; ATCC) con-
taining a fragment of the 16s rDNA (ATCC8750, 3285 kb),
nirS (ATCC 17699D-5, 4052 kb), or nirK (ATCC8750,
3285 kb) gene. The critical threshold (Ct) values were
plotted as a log function of the target DNA copy numbers.
Tenfold serial dilutions of standard genomic DNA ranging
from 102 to 106 gene copies were used in triplicate as a
template to determine the standard curve. Three points of
the standards were used as positive controls in each reac-
tion. A reaction mixture with no template was used as a
negative control. Uncertainty was determined as the stan-
dard deviation of nine replicate amplifications for each
trial.
Statistical Analysis
All statistical calculations were performed in SAS 9.2
(Cary, NC). Two-way analyses of variance (ANOVA) were
performed to compare the effects of C type and concen-
tration on NO3- loss rates, N2O production rates, and nirK,
nirS, and 16S rDNA gene abundances. Means were com-
pared using the Fisher’s least significant difference (LSD)
test at significant level P \ 0.05. For NO3- loss and N2O
production rates and all gene abundances, the C concen-
tration, C type, and a two-way interaction were significant.
The two-way interactions are evident in the data, which
show that all rates and gene abundances trend positively
with C concentration but increase at differing rates (cf.
Tables 2, 3; Figs. 2, 3). The physical rationale for the
different responses to C type is explored in the Discussion
Section. Additionally, one-way analyses of variance
(ANOVA) were performed to compare the gene abun-
dances and N transformation rates both as a function of C
Table 1 Primer sets used for amplification of nir and 16s rDNA (Kandeler and others 2006; Throback and others 2004; Henry and others 2004)
Gene Amplicon size (bp) Primer Annealing
temperature (�C)Name Sequence (50-30)
nirS 425 nirSCd3aF AACGYSAAGGARACSGGa 57
nirSR3cd GASTTCGGRTGSGTCTTSAYGAA
nirK 165 nirK876 ATYGGCGGVAYGGCGA 60
nirK1040 GCCTCGATCAGRTTRTGGTT
16s 200 Eub338 ACTCCTACGGGAGGCAGCAG 53
Eub518 ATTACCGCGGCTGCTGG
a Y was a mixture of C and T. S was a mixture of C and G. R was a mixture of A and G
Environmental Management (2012) 50:861–874 865
123
concentration and type. Again, means were compared
using the Fisher’s least significant difference (LSD) test at
significant level P \ 0.05.
Results
Rates of Nitrate Loss
Examples of data from column experiments, with con-
centrations of NO3- in effluent from triplicate column
experiments plotted as a function of time, are shown in
Fig. 1. Large differences in the rates of NO3- loss were
observed in experiments that utilize different C types. In
Fig. 1A, decreases in NO3- concentration (up to
2.1 mg N L-1) were observed in citric acid treatments at
the concentration of 8 mg C L-1. However, alginic acid
(Fig. 1B) and SRDOC (Fig. 1C) showed no detectable
decrease of NO3- concentrations in the effluent.
Similar results were also evident in experiments at other
concentrations. The average NO3- loss rates (as calculated
from Eq. 1) for all combinations of C type and concen-
tration are shown in Fig. 2a and Table 2. Rates of NO3-
loss increased with increasing citric acid concentration
(R2 = 0.98). Alginic acid had a statistically significant
effect on NO3- loss rates in experiments containing 4 and
16 mg C L-1; however, all rates were significantly less
than those promoted by citric acid at corresponding con-
centrations. In control experiments with no C added, as
well as in all SRDOC treatments, only small changes in
Table 2 The NO3- loss rate and N2O production rate (mg N h-1 g-1) for given combination of carbon type and concentration
Rate Concentration
(mg C L-1)
Rate (mg N h-1 g-1)
Citric Acid Alginic Acid SRDOC
NO3- loss 0 5 ± 6 9 10-5A,a 5 ± 6 9 10-5A,a 5 ± 6 9 10-5A,a
4 1.0 ± 0.8 9 10-4A,b 4 ± 3 9 10-5B,b LOD
8 4 ± 1 9 10-4A,c LOD LOD
16 8 ± 2 9 10-4A,d 4 ± 2 9 10-5B,b LOD
N2O production 0 4 ± 1 9 10-7A,a 4 ± 1 9 10-7A,a 4 ± 1 9 10-7A,a
4 4 ± 3 9 10-7A,a 3.3 ± 0.9 9 10-7A,a 4 ± 2 9 10-7A,a
8 2 ± 1 9 10-6A,b 4 ± 1 9 10-7B,a 3 ± 1 9 10-7B,a
16 3.8 ± 0.9 9 10-6A,c 2.3 ± 0.9 9 10-7B,b 5 ± 2 9 10-7B,b
For NO3- loss rate, limit of detection (LOD)\4 9 10-5 mg N h-1 g-1, based on the reliable detection limit of NO3
- loss (0.1 mg N g-1). For
N2O production rate, LOD\10-8 mg N h-1 g-1, based on detection limit of N2O detection against background (50 ppbV). For both NO3- loss
and N2O production rates, uncertainty is reported as twice the standard deviation (±1 standard deviation) of three replicates of experiments. For
NO3- loss and N2O production rates, values followed by the same capital letter in a row or a lowercase letter within a column are not
significantly different (P [ 0.05) based on a LSD test
Table 3 Gene copy numbers for different carbon types as a function of concentration
Gene Concentration
(mg C L-1)
Gene copy numbers (copy g-1 soil)
Citric acid Alginic acid SRDOC
16s rDNA 0 6 ± 3 9 106A,a 6 ± 3 9 106A,a 6 ± 3 9 106A,a
4 8 ± 3 9 107A,b 2 ± 3 9 107B,b 1.4 ± 0.4 9 107B,b
8 1.2 ± 0.4 9 108A,b 2.7 ± 0.7 9 107B,b 1.4 ± 0.3 9 107B,b
16 1.8 ± 0.3 9 108A,c 3 ± 1 9 107B,b 1.3 ± 0.3 9 107B,b
nirK 0 5 ± 4 9 106A,a 5 ± 4 9 106A,a 5 ± 4 9 106A,a
4 7 ± 3 9 107A,b 1.7 ± 0.9 9 107B,a 1.2 ± 0.5 9 107B,a
8 7 ± 2 9 107A,b 2.3 ± 0.8 9 107B,a 1.2 ± 0.3 9 107B,a
16 9 ± 4 9 107A,b 2 ± 1 9 107B,a 1.0 ± 0.3 9 107B,a
nirS 0 2 ± 2 9 106A,a 2 ± 2 9 106A,a 2 ± 2 9 106B,a
4 3 ± 1 9 107A,b 7 ± 4 9 106B,b 6 ± 3 9 106B,a
8 4 ± 1 9 107A,b 1.3 ± 0.7 9 107B,b 9 ± 6 9 106B,a
16 4 ± 2 9 107A,b 1.2 ± 0.7 9 107B,b 6 ± 3 9 106B,a
Uncertainty is reported as twice the standard deviation (± 1 standard deviation) of nine amplifications at each condition. For a gene, values
followed by the same capital letter in a row or a lowercase letter within a column are not significantly different (P [ 0.05) based on a LSD test
866 Environmental Management (2012) 50:861–874
123
concentration, which were within experimental uncer-
tainty, were observed between the influent and effluent
concentrations. The lack of appreciable NO3- loss in
control experiments was consistent with the assertion that
C limited NO3- loss in our soil.
Rates of Nitrous Oxide Production
Similar to NO3- loss, N2O production rates (Eq. 2) were
greatest for citric acid experiments at all concentrations,
followed by the alginic acid and SRDOC experiments
(Table 2). In experiments with citric acid, mean N2O
production rates increased linearly with increased citric
acid concentration (Fig. 2b; R2 = 0.92). In contrast, N2O
production rates in alginic acid and SRDOC treatments
were approximately an order of magnitude lower than the
rate at the corresponding citric acid concentration for
experiments with 8 and 16 mg C L-1. Although all N2O
production rates were above the LOD, production rates for
alginic acid and SRDOC experiments were statistically
different from controls only in experiments with
16 mg C L-1. These results suggest that alginic acid and
SRDOC stimulate to N2O production in our system only
when present at high concentration, consistent with low
NO3- loss rates. However, rates of NO3 loss were between
100- to 250-fold greater that N2O production rates, indi-
cating that only a small portion of NO3- lost was converted
to N2O.
Quantification of nir and 16s rDNA Genes
The 16s rDNA, nirK, and nirS abundances in the unit of
copy number g-1 soil (Table 3) were grouped by C type
and plotted as a function of concentration in Fig. 3. There
was a clear trend of increasing 16s rDNA gene copy
numbers (8 ± 3 9 107 to 1.8 ± 0.3 9 108 copy g-1 soil)
0
0.2
0.4
0.6
0.8
1.0N
O3-
loss
rat
e
Carbon concentration (mg C L-1)
Citric acid
Alginic acid
SRDOC
0
2.0
4.0
6.0
N2O
pro
duct
ion
rate
(mg
N h
r-1 g
-1)
× 10
-6 Citric acid
Alginic acid
SRDOC
0 4 8 12 16
a
Carbon concentration (mg C L-1)0 4 8 12 16
b
(mg
N h
r-1 g
-1)
× 10
-3
R2 = 0.98
R2 = 0.92
Fig. 2 Average rates of a NO3- loss and b NO2 production rates as a
function of carbon type and concentration. Lines represent a least-
squares fit of a linear model to the citric acid data. For both NO3- loss
and N2O production rates, uncertainty is reported as twice the
standard deviation (±1 standard deviation) of three replicates of
experiments
nirK
copy
num
ber
(×10
8 )
0 mg C L-1
4 mg C L-1
8 mg C L-1
16 mg C L-1
nirS
copy
num
ber
(×10
8)
0 mg C L-1
4 mg C L-1
8 mg C L-1
16 mg C L-1
0
0.5
1.0
1.5
2.0
2.5
Citric acid
16 r
DN
A c
opy
num
ber
(×10
8)
Carbon type
0 mg C L-1
4 mg C L-1
8 mg C L-1
16 mg C L-1
a
SWDOCAlginic acid
b
c
Citric acid
Carbon type
SWDOCAlginic acid
Citric acid
Carbon type
SWDOCAlginic acid
0
0.5
1.0
1.5
2.0
2.5
0
0.5
1.0
1.5
2.0
2.5
a
b
b
c
ab b b
a b b b
a
bb b
a b b b aaaa
aa a a
aaaaa
bb
b
Fig. 3 The a nirK b nirS, and c 16s rDNA copy numbers as a
function of carbon concentration (mg C L-1) for different carbon
types. Uncertainty is reported as twice the standard deviation
(±1 standard deviation) of nine amplifications at each condition.
Within the same treatment, bars with the same letter above them are
not significantly different (P [ 0.05) based on a LSD test
Environmental Management (2012) 50:861–874 867
123
with increasing C concentration in the citric acid treatment.
Experiments containing alginic acid and SRDOC contained
higher 16s rDNA gene copy numbers than control experi-
ments; however, gene copy numbers were lower than in the
corresponding experiments with citric acid, and did not
vary with increasing C concentrations.
Soil samples in the citric acid treatment contained more
nir gene copies than in control experiments or corre-
sponding experiments with the other C sources (Fig. 3a, b).
However, although all citric acid experiments contained
higher nir copy numbers than the control, there was no
trend in the average nirK or nirS gene copy numbers with
increasing C concentrations. Experiments with alginic acid
had higher nirS numbers than control, but nirK gene copy
numbers were not significantly different from the control.
Similarly, experiments with SRDOC did not produce an
increase in nir genes copy number from the control. Soil
samples with alginic acid and SRDOC contained fewer nir
gene copies per gram than corresponding experiments with
citric acid for all C concentrations.
Discussion
Effect of Organic Carbon on Rates of Nitrate Loss
and Nitrous Oxide Production
Organic C availability often limits denitrification in
groundwater, riparian buffers, and other types of wetlands
(Greenan and others 2006; Rivett and others 2008), and has
been suggested to limit activity at in soils at CEFS, the site
from which these samples were obtained (Knies 2009).
However, C type, as well as concentration, may regulate
organic C availability, and thus the denitrification rate.
Denitrification has been shown to correlate more strongly
with readily decomposable soil organic matter than with
total organic C (Beauchamp and others 1980; Bijay-Singh
and Whitehead 1988; Burford and Bremner 1975; David-
son and others 1987; Hill and Cardaci 2004; Myrold
and Tiedje 1985; Stanford and others 1975; McCarty and
Bremner 1992; Pintar and Lobnik 2005; Hernandez and
Mitsch 2007). The nature of organic C may also be
responsible for ecosystem-scale differences in denitrifica-
tion (Ward and others 2009).
In our experiments, citric acid was most effective in
promoting NO3- loss and N2O production. Citric acid, a
LMMOA and intermediate in common metabolic cycles, is
likely to be readily available to organisms, and thus pro-
moted NO3- reduction in our experiments. The release of
LMMOAs, such as citric acid, malic acid, and acetic acid,
from root exudation and cell lysis has been well docu-
mented in surface soil and the rhizosphere (Jones 1998;
Jones and others 2003). These molecules are rapidly
mineralized in soils (van Hees and others 2005), and thus
are a good substrate for microbial growth and the con-
sumption of electron acceptors. Our results suggested that
LMMOAs in soil organic C may be tightly coupled to
NO3- loss, in agreement with results from previous labo-
ratory (Henry and others 2008) and field studies (Stow and
others 2005; Sirivedhin and Gray 2006; Bernhardt and
Likens 2002).
Several similar column studies provide a basis of com-
parison for the rates of NO3- loss derived from our study.
Pavel et al. (1996) utilized surface soils with high DOC
concentrations (9.3 % and 7.8 %) without additional C
addition, but used a similar continuous-flow column
experimental approach. In situations where transport pro-
cesses did not control the reaction, their rates were
approximately 10-4 to 10-5 mg N h-1 g-1 soil. Knies
(2009) employed a similar methodology and concentration
range as the current study, but used D-glucose as C source;
NO3- loss rates also varied between 10-4 to 10-5
mg N h-1 g-1 soil, and generally increased with increas-
ing C concentration. These results thus broadly agree with
our measured rates of NO3- loss (Table 2).
Alginic acid had a small effect NO3- loss rates and only
promoted N2O production at high concentration in our
experiments. Many authors have suggested that soluble C
sources are key drivers of microbial activity, although not
all soluble C is equally bioavailable (Boyer and Groffman
1996; Cook and Allan 1992). The generally accepted
dogma is that high molecular mass organic molecules are
more recalcitrant than low molecular mass organic mole-
cules (Tate 1987), which could be a possible explanation
for the low NO3- loss and N2O production rates in our
alginic acid column experiments. This observation is sup-
ported by previous work that demonstrated that glucose
enhanced the denitrification potential more effectively than
the biopolymer cellulose or grass litter (Dendooven and
others 1996).
Our results indicated that SRDOC did not affect NO3-
loss rates and only promoted N2O production at high
concentration in our experiments. This result was consis-
tent with studies that have shown denitrification to be
limited even in the presence of abundant DOC (Starr and
Gillham 1993), perhaps because of the nature of the C type.
Previous results have also demonstrated that denitrification
in riparian buffer soils is highest in surface horizons and
depends on organic matter content (Groffman and others
1992; Lowrance 1992; Groffman and others 1991; Willems
and others 1997); however, surface horizons have high C
concentrations that may contain more labile types of C in
addition to humic substances (Thurman 1985; Tate 1987).
One possible reason for the low reactivity of the SRDOC is
that humic substances tend to be recalcitrant, and thus have
long environmental residence times and degrade slowly via
868 Environmental Management (2012) 50:861–874
123
microbial processes (Thurman 1985). Additionally,
SRDOC sample has a high C/N ratio (48:1), which may not
be optimal for supporting microbial activity. Van Mooy
and others (2002) found that denitrifying microbes pref-
erentially consumed N-rich organic C when given suffi-
cient C supply. Also, Groffman and others (1991) found
higher denitrification potential in grassy vegetated filter
strips as compared to forest strips, which was attributed to
lower C/N ratio (\10) in the grassy buffer.
Although our results indicated low rates of NO3- loss
and N2O production for alginic acid and SRDOC, the
molecules in fact may promote N transformations in the
field. The residence time in our columns (ca. 2.8 h) was
short when compared to the typical time required for
groundwater to pass through a riparian buffer, wetlands, or
the hyporheic zone. Thus, if these molecules are abundant
in a specific environment, they may appreciably promote
denitrification, albeit at a slower rate, as long as other
conditions are favorable and the residence time of the
groundwater is sufficient (Ocampo and others 2006). This
idea is also supported by the observed increase in nirS gene
copy number upon addition of alginic acid, which suggests
some stimulation of N cycle bacteria in our soil.
It is also important to note the limitations of the
chemical measurements described in the present study.
Measurements of N transformations, even in carefully
controlled experimental conditions, are complicated
because myriad processes are possible. In our anaerobic
columns, where the major electron donor and acceptor
were organic C and NO3-, respectively, it is reasonable to
assume that denitrification, a process that produces gaseous
nitrogen oxides and N2, was occurring in our system.
However, several different factors convoluted the inter-
pretation of our kinetic data. First, it should be noted that
the mass of NO3- lost in our system greatly exceeded the
mass of N2O produced for all measurements, consistent
with previous conceptual models of denitrification (Fire-
stone and Davidson 1989). However, N2O may have been
produced by other processes [e.g., nitrification (Bremner
and Blackmer 1978)], and thus other N transformation
processes may have affected our N2O production rates.
Second, N2, the major production of denitrification, was
not measured due to the difficulties in its quantification
(Groffman and others 2006). We were thus unable to
complete a full mass balance for N in our system. This
limited our ability to relate influent NO3-concentration to
the concentration of N compounds in the effluent. Finally,
although the N content of our soil is low (0.02 %), it was
not possible to conclusively exclude the possibility that N
mineralization and nitrification may result in internal pro-
duction of NO3- in our columns, which could convolute
mass balance approaches and may have affected our
measured rates. The possible contributions of other N
transformation processes thus limit the strict interpretation
of the kinetic measurements in these columns to net rates of
NO3- loss and N2O production.
Effect of Organic Carbon Gene Abundances
Microbial communities may be altered by C amendments
that select for populations that are most competitive in
terms of growth rates and their ability to utilize nutrients
(Drenovsky and others 2004). Our study found an increase
in nirK gene abundance in citric acid experiments as
compared to untreated controls, but no change for systems
with alginic acid or SRDOC. Henry and others (2004)
observed a four-fold increase in nirK gene copy number in
soil amended with a unspecified organic C source. Simi-
larly, the nirK abundance correlated with quantity of soil
organic C in a receding glacier foreland, although little
information was provided regarding the type of C (Kan-
deler and others 2006). Taken together, these observations
suggest that the type and concentration of C are key factors
in controlling the denitrifying bacteria community.
16s rDNA copy numbers have been commonly used to
estimate the bacterial population in soils. Compared to
previous studies, which found the 16s rDNA copy numbers
in agricultural soils to be ca. 6 9 107 to 1 9 108 copy g-1
soil (Dandie and others 2008; Miller and others 2009), our
16s rDNA gene abundances were somewhat low (Table 1),
consistent with the previous assertion that low concentra-
tions of organic C may be limiting biological activity in
CEFS soils (Knies 2009). In our soil, 16s rDNA gene
abundances increased upon addition of all types of C,
suggesting stimulation of the overall microbial community
by C addition.
Copy numbers of nir genes, which potentially have a
functional relationship with N reduction rates, may also be
compared to prior results in observed in soils. Our results for
nirK in trials without C addition (5 9 106 copy g-1 soil)
were in rough agreement with previous studies of environ-
mental samples that have reported 105 to 106 copy g-1 soil
of nirK in sandy soils (Henry and others 2004), 5 9 104
to 107 copy g-1 soil in a subsoil of a fir plantation (Levy-
Booth and Winder 2010), 4.1-7.3 9 107 copy g-1 soil in a
agricultural field, 107 copy g-1 soil in rice paddy soil
(Yoshida and others 2009), and 107 copy g-1 soil of glacial
soils (Kandeler and others 2006). Our nirS gene abundance
for soil without C addition (2 9 106 copy g-1 soil) agreed
well with what has been reported for other unamended
environmental samples, including estuarine sediments (105
to 106 copy g-1 soil) (Smith and others 2006), a subsoil of a
fir plantation (3 9 105 copy g-1 soil) (Levy-Booth and
Winder 2010), rice paddy soil (106 copy g-1 soil) (Yoshida
and others 2009) and soil of glacial soils (5 9 106 to 1 9 108
copy g-1) (Kandeler and others 2006).
Environmental Management (2012) 50:861–874 869
123
Bacterial Community Composition
Table 4 shows the ratios of the genes in soils from our
column experiments. Changes in the ratios provide infor-
mation on how community composition changes with C
addition. The ratio of nirK to 16s rDNA varies from 0.50-
0.88, a somewhat surprising result that reflects the low 16s
rDNA copy number measured in our soils. Although some
studies have found a significantly smaller ratio (Kandeler
and others 2006; Henry and others 2008), Daidie and others
(2008) found similar ratios of nirK:16s rDNA in an agri-
cultural field soil. nirS genes were generally less abundant
than nirK in our soil, with ratios of nirK to nirS ranging
from 1.3 to 2.4 (Table 4). The ratio of these genes in lit-
erature studies of soils is highly variable (Bothe and others
2000; Yoshida and others 2009, 2010; Kandeler and others
2006; Levy-Booth and Winder 2010; Henry and others
2008; Dandie and others 2011), suggesting that local fac-
tors may control which gene is more abundant. However,
the relative consistency of ratios of nirK:nirS in our
experiments indicated that the denitrifier community
composition is relatively stable regardless of the C type and
concentration, in accordance with a previous study that
noted a correlation between nirK and nirS gene abundance
(Levy-Booth and Winder 2010). Although multiple gene
copies may be present in a single cell [Philipott (2002)
reported that only one copy of nirK is present in a bacte-
rium] and the primer bias cannot be excluded, our qPCR
results in control experiments suggest that the nirK-har-
boring denitrifiers might be more abundant than nirS-har-
boring denitrifiers in our riparian buffer soil.
Gene Abundances and Rates of Nitrate Loss
and Nitrous Oxide Production
Although nir genes are responsible for NO2-, not NO3
- or
NO, reduction, it is reasonable to postulate that the rates of
these processes are strongly correlated, especially given the
previous observation that NO is an unstable intermediate in
denitrification (Firestone and Davidson 1989) and our low
observed concentrations of NO2- in the effluent. In our study,
the nirK and nirS gene copy numbers trended with increasing
rates of NO3- loss or N2O production (Fig. 4a–d); however,
there was not a linear correlation between either nir gene
abundance and these rates. In contrast, the 16s rDNA gene
copy number (Fig. 4e, f) correlated with both NO3- loss rate
(R2 = 0.85) and N2O production rate (R2 = 0.72). Overall,
our chemical and biological measurements were consistent
with previous work that noted an increase in both the deni-
trification rate, and the abundance of denitrifying and total
bacteria (as determined by most probable number analysis) in
soils amended with high concentrations of glucose and NO3-
(Ardakani and others 1975; Bowman and Focht 1974). It is
important to remember other biological factors may affect the
relationship between copy gene abundance and denitrifica-
tion rate. Most notably, elevated metabolic activity, which
may be measured by nir gene expression (mRNA quantifi-
cation) instead of gene abundance, could also result in
increased NO3- loss and N2O production rates.
Henry and others (2008) and Mounier and others (2004)
observed increases in denitrification rates but only small
changes in denitrifying bacterial community composition in
agricultural soils amended with root exudates (a mixture of
sugars, organic acids, and amino acids, which are somewhat
analogous to citric acid) and corn mucilage (a complicated
mixture of polysaccharides, which are structurally related to
alginic acid). In the context of our findings, their results are
somewhat surprising in that their experiments are somewhat
analogous to our experiments with citrate, which resulted in
an increase in NO3- loss rate, N2O production rate, and in nir
gene copy number, and with alginate, which resulted in small
changes to NO3- loss rates and nirS gene copy numbers.
Although some of these differences may result from differ-
ences in experimental conditions (e.g., silty clay loam vs.
sand, macrocosm vs. continuous flow column experiment,
concentration of NO3- and organic C, etc.), the results also
underscore the potential importance of specific molecules in
preferentially promoting specific microbial processes.
It should be noted that, as in any PCR-based study, there
are artifacts associated with amplification process. The use of
specific primers targeting specific genetic sequences may not
allow the quantification of all the target functional genes in a
population. The possible presence of multiple gene copy
numbers in an organism may distort the view of community
composition. In addition, PCR may amplify genes that are in
dead or dormant organisms. However, because we measured
changes with C addition from an unamended control system,
the significance of the last point is minimized. It also is
possible that storage before amplification may have affected
the microbial community.
Table 4 Ratios of the NO2- reduction genes nirK and nirS copy
numbers to the 16s rDNA copy number
Type Concentration
(mg C L-1)
nirK/16S nirS/16S nirK/nirS
Citric Acid 0 0.83 0.33 2.5
4 0.88 0.38 2.3
8 0.58 0.33 1.8
16 0.50 0.22 2.3
Alginic Acid 4 0.85 0.35 2.4
8 0.85 0.48 1.8
16 0.67 0.40 1.7
SRDOC 4 0.86 0.43 2.0
8 0.86 0.64 1.3
16 0.77 0.47 1.7
870 Environmental Management (2012) 50:861–874
123
The results of this study indicate that structure of soluble
C may be critical in regulation of soil N transformations,
thus highlighting the potential importance of vegetation in
promoting denitrification in riparian buffers and other ter-
restrial landscapes. Although a recent meta-analysis did not
identify vegetation as a key factor for controlling subsur-
face N removal in riparian buffers, the effect may have
been masked in such an analysis by other soil biogeo-
chemical factors (Mayer and others 2007). Metabolites,
such as citric acid and other LMMOAs, are often leaked
from roots, accounting for up to one-fifth of the C fixed by
plants (Nguyen 2003), and thus are often present in high
concentrations in the rhizosphere (Jones 1998; Curl and
Truelove 1986; Jones and others 2003) where they may
affect the microbial ecology of denitrifying bacteria
(Philippot and others 2007). In some environments,
LMMOA and other small molecules may account for up to
30 % of organic C (van Hees and others 2005). These
molecules, which have short environmental residence
times, are critical drivers of many environmental processes,
and may also be important connections between the C and
N cycles.
Conclusions
The rates of N transformations in soil may vary with
organic C concentrations and type. When sufficient NO3-
is present in system, availability of liable organic C may
limit denitrification under anaerobic conditions. Results
from molecular biology approaches indicated that specific
components of soil organic matter may have differing
effects on the rates of N transformations by influencing
gene abundances in riparian buffer soil. Citric acid, a labile
organic metabolite, exhibited significant enhancement of N
transformation rates and gene abundances; however,
increased concentrations of alginic acid and SRDOC either
did not affect or had a small positive effect on NO3- loss
0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
nirK
copy
num
ber
(×10
8 )
0
0.2
0.4
0.6
0.8
nirS
copy
num
ber
(×10
8 )
a b
c
0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
nirK
copy
num
ber
(×10
8)
0 1 2 3 4 50 0.2 0.60.4 0.8
0 0.2 0.60.4 0.80
0.2
0.4
0.6
0.8
nirS
copy
num
ber
(×10
8 )
0 1 2 3 4 5
d
0
0.5
1.0
1.5
2.0
2.5
16s
rDN
Aco
pynu
mbe
r ( ×
10 )
0 1 2 3 4 50 0.2 0.60.4 0.8
16s
rDN
Aco
pynu
mbe
r ( ×
10 )
0
0.5
1.0
1.5
2.0
2.5
N2O production rate (mg N hr-1 g-1) × 10-6
e fN2O production rate (mg N hr-1 g-1) × 10-6
N2O production rate (mg N hr-1 g-1) × 10-6
NO3- loss rate (mg N hr-1 g-1) × 10-3
NO3- loss rate (mg N hr-1 g-1) × 10-3
NO3- loss rate (mg N hr-1 g-1) × 10-3
R2 = 0.85 R2 = 0.72
8 8
Fig. 4 NO3- loss rates versus
a nirK, c nirS, and e 16s rDNA
gene copy numbers;
N2Oproduction rates vs. b nirK,
d nirS, and f 16s rDNA gene
copy numbers. Lines represent a
least-squares fit of a linear
model to the 16s rDNA data.
Legend: squares alginic acid,
triangle SWDOC, diamondscitric acid
Environmental Management (2012) 50:861–874 871
123
rate, N2O production, and the copy number of functional
genes. The ratio of nirK:nirS genes in amended soils was
relatively constant, suggesting reasonable stability in the
denitrifying community structure. Longer residence time
experiments with complex or recalcitrant C sources may
help to further refine and quantify their effect on N trans-
formation rates and microbial community dynamics.
Additionally, more field studies that relate differing types
of C to functional gene copy numbers or microbial com-
munity dynamics may help to further refine the relation-
ships between N transformations, biological activity, and C
availability.
Acknowledgments We thank Lauren Saal, Emily Dell, Guillemo
Ramirez, Sara Knies, Brad Robinson, Chris Ashwell, Greg Dick,
Suzanna Brauer, Jeff White, and John Walker. This work was sup-
ported by NC Department of Environment & Natural Resources,
Conservation Reserve Enhancement Program. Lin Wu thanks the
Society of Wetland Scientists for a student research grant.
References
Ardakani MS, Belser LW, McLaren AD (1975) Reduction of nitrate
in a soil column during continuous flow. Soil Science Society of
America Journal 39:290–294
Barnhill WL, Goodwin J, Bostian MR, McLoda NA, Leishman GW,
Scanu RJ (1974) Soil survey of Wayne County, North Carolina.
Soil Conservation Service, Washington, DC
Beauchamp EG, Gale C, Yeomans JC (1980) Organic matter
availability for denitrification in soils of different textures and
drainage classes. Communications in Soil Science and Plant
Analysis 11:1221–1233
Bernhardt ES, Likens GE (2002) Dissolved organic carbon enrich-
ment alters nitrogen dynamics in a forest stream. Ecology
83:1689–1700
Bijay-Singh RydenJC, Whitehead DC (1988) Some relationships
between denitrification potential and fractions of organic carbon
in air-dried and field-moist soils. Soil Biology & Biochemistry
20:737–741
Bothe H, Jost G, Schloter M, Ward BB, Witzel KP (2000) Molecular
analysis of ammonia oxidation and denitrification in natural
environments. FEMS Microbiology Reviews 24:673–690
Bowman RA, Focht DD (1974) Influence of glucose and nitrate
concentrations upon denitrification rates in sandy soils. Soil
Biology & Biochemistry 6:297–301
Boyer JN, Groffman PM (1996) Bioavailability of water extractable
organic carbon fractions in forest and agricultural soil profiles.
Soil Biology & Biochemistry 28:783–790
Bradley PM, Fernandez M, Chapelle FH (1992) Carbon limitation of
denitrification rates in an anaerobic groundwater system. Envi-
ronmental Science and Technology 26:2377–2381
Bremner JM, Blackmer AM (1978) Nitrous oxide: emission from
soils during nitrification of fertilizer nitrogen. Science 199:
295–296
Burford JR, Bremner JM (1975) Relationships between denitrification
capacities of soils and total, water soluble and readily decom-
posable soil organic-matter. Soil Biology & Biochemistry 7:
389–394
Cook BD, Allan DL (1992) Dissolved organic carbon in old field
soils- compositional changes during the biodegredation of soil
organic-matter. Soil Biology & Biochemistry 24:595–600
Curl EA, Truelove B (1986) The rhizosphere. Springer, New York
Dandie CE, Burton DL, Zebarth BJ, Henderson SL, Trevors JT,
Goyer C (2008) Changes in bacterial denitrifier community
abundance over time in an agricultural field and their relation-
ship with denitrification activity. Applied and Environmental
Microbiology 74:5997–6005
Dandie CE, Wertz S, Leclair CL, Goyer C, Burton DL, patten CL,
Zebarth BJ, Trevors JT (2011) Abundance, diversity and
functional gene expression of dentrifier comminties in adjacent
riparian and agricultural zones. FEMS Microbiology Ecology
77:69–82
Davidson EA, Galloway LF, Strand MK (1987) Assessing available
carbon: comparison of techniques across selected forest soils.
Communications in Soil Science and Plant Analysis 18:45–64
Dendooven L, Splatt P, Anderson JM (1996) Denitrification in
permanent pasture soil as affected by different forms of C
substrate. Soil Biology & Biochemistry 28:141–149
Dosskey MG, Vidon P, Gurwick NP, Allan CJ, Duval TP, Lowrance
R (2010) The role of riparian vegetation in protecting and
improving chemical water quality in streams. Journal of the
American Water Resources Association 46:261–277
Drenovsky RE, Vo D, Graham KJ, Scow KM (2004) Soil water
content and organic carbon availability are major determinants
of soil microbial community composition. Microbial Ecology
48:424–430
Dukes MD (2000) Effect of riparian buffers and controlled drainage
on shallow groundwater quality in the North Carolina Middle
Coastal Plain. Doctoral Dissertation, North Carolina State
University, Raleigh
Dukes MD, Evans RO, Gilliam JW, Kunickis SH (2002) Effect of
riparian buffer width and vegetation type on shallow groundwa-
ter quality in the Middle Coastal Plain of North Carolina.
Transactions of the ASAE 45:327–336
Firestone MK, Davidson EA (1989) Microbiological basis of NO and
N2O production and consumption in soil. In: Andreae MO,
Schimel DS (eds) Exchange of trace gases between terrestrial
ecosystems and the atmosphere, vol 47. Life Sciences Research
Report. Wiley, Hoboken, NJ, pp 7–21
Gee GW, Or D (2002) Particle-size analysis. In: Dane JH, Topp GC
(eds) Methods of soil analysis. Part 4. Physical methods. Soil
Science Society of America, Madison, WI
Greenan CM, Moorman TB, Kaspar TC, Parkin TB, Jaynes DB
(2006) Comparing carbon substrates for denitrification of
subsurface drainage water. Journal of Environmental Quality
35:824–829
Groffman PM, Axelrod EA, Lemunyon JL, Sullivan WM (1991)
Denitrification in grass and forest vegitated filter strips. Journal
of Environmental Quality 20:671–674
Groffman PM, Gold AJ, Simmons RC (1992) Nitrate dynamics in
riparian forests-microbial studies. Journal of Environmental
Quality 21:666–671
Groffman PM, Altabet MA, Bohlke JK, Butterbach-Bahl K, David
MB, Firestone MK, Giblin AE, Kana TM, Nielsen LP, Voytek
MA (2006) Methods for measuring denitrification: diverse
approaches to a difficult problem. Ecological Applications
16:2091–2122
Hedin LO, von Fischer JC, Ostrom NE, Kennedy BP, Brown MG,
Robertson GP (1998) Thermodynamic constraints on nitrogen
transformations and other biogeochemical processes at soil-
stream interfaces. Ecology 79:684–703
Henry S, Baudoin E, Lopez-Gutierrez JC, Martin-Laurent F,
Baumann A, Philippot L (2004) Quantification of denitrifying
bacteria in soils by nirK gene targeted real-time PCR. Journal of
Microbiological Methods 59:327–335
Henry S, Texier S, Hallet S, Bru D, Dambreville C, Cheneby D, Bizouard
F, Germon JC, Philippot L (2008) Disentangling the rhizosphere
872 Environmental Management (2012) 50:861–874
123
effect on nitrate reducers and denitrifiers: insight into the role of root
exudates. Environmental Microbiology 10:3082–3092
Hernandez ME, Mitsch WJ (2007) Denitrification potential and
organic matter as affected by vegetation community, wetland
age, and plant introduction in created wetlands. Journal of
Environmental Quality 36:333–342
Hill AR (1996) Nitrate removal in stream riparian zones. Journal of
Environmental Quality 25:743–755
Hill AR, Cardaci M (2004) Denitrification and organic carbon
availability in riparian wetland soils and subsurface sediments.
Soil Science Society of America Journal 68:320–325
Hill AR, Devito KJ, Campagnolo S, Sanmugadas K (2000) Subsur-
face denitrification in a forest riparian zone: interactions between
hydrology and supplies of nitrate and organic carbon. Biogeo-
chemistry 51:193–223
Jones DL (1998) Organic acids in the rhizosphere: a critical review.
Plant and Soil 205:25–44
Jones DL, Dennis PG, Owen AG, van Hees PAW (2003) Organic acid
behavior in soils: misconceptions and knowledge gaps. Plant and
Soil 248:31–41
Kandeler E, Deiglmayr K, Tscherko D, Bru D, Philippot L (2006)
Abundance of narG, nirS, nirK, and nosZ genes of denitrifying
bacteria during primary successions of a glacier foreland.
Applied and Environmental Microbiology 72:5957–5962
King SE (2006) Riparian buffer effectiveness in removing ground-
water nitrate as influenced by vegetative type. Master’s Thesis,
North Carolina State University, Raleigh
Knies SV (2009) Riparian buffer effectiveness at removal of NO3-N
from groundwater in the middle Coastal Plain of North Carolina.
Master’s Thesis, North Carolina State University, Raleigh
Knowles R (1982) Denitrification. Microbiological Reviews 46:43–70
Levy-Booth DJ, Winder RS (2010) Quantification of nitrogen
reductase and nitrite reductase genes in soil of thinned and
clear-cut Douglas-Fir stands by using real-time PCR. Applied
and Environmental Microbiology 76:7116–7125
Lowrance R (1992) Groundwater nitrate and denitrification in a
coastal plain riparian forest. Journal of Environmental Quality
21:401–405
Lowrance R, Altier LS, Newbold JD, Schnabel RR, Groffman PM,
Denver JM, Correll DL, Gilliam JW, Robinson JL, Brinsfield
RB, Staver KW, Lucas W, Todd AH (1997) Water quality
functions of riparian forest buffers in Chesapeake Bay water-
sheds. Environmental Management 21:687–712
Mayer PM, Reynolds SK, McCutchen MD, Canfield TJ (2007) Meta-
analysis of nitrogen removal in riparian buffers. Journal of
Environmental Quality 36:1172–1180
McCarty GW, Bremner JM (1992) Availability of organic carbon for
denitrification of nitrate in subsoils. Biology and Fertility of
Soils 14:219–222
Miller MN, Zebarth BJ, Dandie CE, Burton DL, Goyer C, Trevors JT
(2009) Denitrifier community dynamics in soil aggregates under
permanent grassland and arable cropping systems. Soil Science
Society of America Journal 73:1843–1851
Miller MN, Dandie CE, Zebarth BJ, Burton DL, Goyer C, Trevors JT
(2012) Influence of carbon amendments on soil denitrifier
abundance in soil microcosms. Geoderma 170:48–55
Mounier E, Hallet S, Cheneby D, Benizri E, Gruet Y, Nguyen C,
Piutti S, Robin C, Slezack-Deschaumes S, Martin-Laurent F,
Germon JC, Philippot L (2004) Influence of maize mucilage on
the diversity and activity of the denitrifying community.
Environmental Microbiology 6:301–312
Myrold DD, Tiedje JM (1985) Establishment of denitrification
capacity in soil: effects of carbon, nitrate and moisture. Soil
Biology & Biochemistry 17:819–822
Nguyen C (2003) Rhizodeposition of organic C by plants: mecha-
nisms and controls. Agronomie 23:375–396
Nogales B, Timmis KN, Nedwell DB, Osborn AM (2002) Detection and
diversity of expressed denitrification genes in estuarine sediments
after reverse transcription-PCR amplification from mRNA.
Applied and Environmental Microbiology 68:5017–5025
Nommik H (1956) Denitrification in soil. Acta Agriculturæ Scandi-
navica 6:195–228
Ocampo CJ, Oldham CE, Sivapalan M (2006) Nitrate attenuation in
agricultural catchments: shifting balances between transport and
reaction. Water Resources Research 42:W01408
Osborne LL, Kovacic DA (1993) Riparian vegetated buffer strips in
water-quality restoration and stream management. Freshwater
Biology 29:243–258
Pavel EW, Reneau RB, Berry DF, Smith EP, Mostaghimi S (1996)
Denitrification potential of nontidal riparian wetland soils in the
Virginia coastal plain. Water Research 30:2798–2804
Perry TD, Duckworth OW, McNamara CJ, Martin ST, Mitchell R
(2004) The effects of the biologically produced polymer alginic
acid on macroscopic and microscopic calcite dissolution rates.
Environmental Science and Technology 38:3040–3046
Perry TD, Cygan RT, Mitchell R (2006) Molecular models of alginic
acid: interactions with calcium ions and calcite surfaces.
Geochimica et Cosmochimica Acta 70:3508–3532
Philippot L (2002) Denitrifying genes in bacterial and archaeal
genomes. Biochimica et Biophysica Acta (BBA) - Gene Structure
and Expression 1577:355–376
Philippot L, Hallin S, Schloter M (2007) Ecology of denitrifying
prokaryotes in agricultural soil. Advances in Agronomy 96:
249–305
Pintar M, Lobnik F (2005) The impact of nitrate and glucose availability
on the denitrification at different soil depths. Fresenius Environ-
mental Bulletin 14:514–519
Rivett MO, Buss SR, Morgan P, Smith JWN, Bemment CD (2008)
Nitrate attenuation in groundwater: a review of biogeochemical
controlling processes. Water Research 42:4215–4232
Sirivedhin T, Gray KA (2006) Factors affecting denitrification rates in
experimental wetlands: field and laboratory studies. Ecological
Engineering 26:167–181
Smith TA, Osmond DL, Gilliam JW (2006) Riparian buffer width and
nitrate removal in a lagoon-effluent irrigated agricuttural area.
Journal of Soil and Water Conservation 61:273–281
Stanford G, Vanderpol RA, Dzienia S (1975) Denitrification rates in
relation to total and extractable soil carbon. Soil Science Society
of America Journal 39:284–289
Starr RC, Gillham RW (1993) Denitrification and organic carbon
availability in two aquifers. Ground Water 31:934–947
Stow CA, Walker JT, Cardoch L, Spence P, Stow CA (2005) N2O
emissions from streams in the Neuse River watershed, North
Carolina. Environmental Science and Technology 39:6999–7004
Stumm W, Morgan JJ (1996) Aquatic chemistry, 3rd edn. Wiley, New
York
Tate RL (1987) Soil organic matter: biological and ecological effects.
Wiley, New York
Throback IN, Enwall K, Jarvis A, Hallin S (2004) Reassessing PCR
primers targeting nirS, nirK and nosZ genes for community
surveys of denitrifying bacteria with DGGE. FEMS Microbiol-
ogy Ecology 49:401–417
Thurman EM (1985) Organic geochemistry of natural waters. Kluwer,
Boston
van Hees PAW, Jones DL, Finlay R, Godbold DL, Lundstom US
(2005) The carbon we do not see: the impact of low molecular
weight compounds on carbon dynamics and respiration in forest
soils: a review. Soil Biology & Biochemistry 37:1–13
Van Mooy BAS, Keil RG, Devol AH (2002) Impact of suboxia on
sinking particulate organic carbon: enhanced carbon flux and
preferential degradation of amino acids via denitrification.
Geochimica et Cosmochimica Acta 66:457–465
Environmental Management (2012) 50:861–874 873
123
Wallenstein MD, Myrold DD, Firestone M, Voytek M (2006)
Environmental controls on denitrifying communities and deni-
trification rates: insights from molecular methods. Ecological
Applications 16:2143–2152
Ward BB, Devol AH, Rich JJ, Chang BX, Bulow SE, Naik H,
Pratihary A, Jayakumar A (2009) Denitrification as the dominant
nitrogen loss process in the Arabian Sea. Nature 461:78–81
Willems HPL, Rotelli MD, Berry DF, Smith EP, Reneau RB,
Mostaghimi S (1997) Nitrate removal in riparian wetland soils:
effects of flow rate, temperature, nitrate concentration and soil
depth. Water Research 31:841–849
Yoshida M, Ishii S, Otsuka S, Senoo K (2009) Temporal shifts in
diversity and quantity of nirS and nirK in a rice paddy field soil.
Soil Biology & Biochemistry 41:2044–2051
Yoshida M, Ishii S, Otsuka S, Senoo K (2010) nirK-harboring
denitrifiers are more responsive to denitrification inducing
conditions in rice paddy soil than nirS-harboring bacteria.
Microbes and Environments 25:45–48
874 Environmental Management (2012) 50:861–874
123