Content uploaded by Jay Zarnetske
Author content
All content in this area was uploaded by Jay Zarnetske on Jul 07, 2014
Content may be subject to copyright.
Dynamics of nitrate production and removal as a function
of residence time in the hyporheic zone
Jay P. Zarnetske,
1
Roy Haggerty,
1
Steven M. Wondzell,
2
and Michelle A. Baker
3
Received 18 March 2010; revised 7 December 2010; accepted 21 December 2010; published 10 March 2011.
[1] Biogeochemical reactions associated with stream nitrogen cycling, such as
nitrification and denitrification, can be strongly controlled by water and solute residence
times in the hyporheic zone (HZ). We used a whole‐stream steady state
15
N‐labeled
nitrate (
15
NO
3
−
) and conservative tracer (Cl
−
) addition to investigate the spatial and
temporal physiochemical conditions controlling the denitrification dynamics in the HZ of
an upland agricultural stream. We measured solute concentrations (
15
NO
3
−
,
15
N
2
(g), as
well as NO
3
−
,NH
3
, DOC, DO, Cl
−
), and hydraulic transport parameters (head, flow rates,
flow paths, and residence time distributions) of the reach and along HZ flow paths of
an instrumented gravel bar. HZ exchange was observed across the entire gravel bar (i.e., in
all wells) with flow path lengths up to 4.2 m and corresponding median residence
times greater than 28.5 h. The HZ transitioned from a net nitrification environment at
its head (short residence times) to a net denitrification environment at its tail (long
residence times). NO
3
−
increased at short residence times from 0.32 to 0.54 mg‐NL
−1
until
a threshold of 6.9 h and then consistently decreased from 0.54 to 0.03 mg‐NL
−1
.
Along these same flow paths, declines were seen in DO (from 8.31 to 0.59 mg‐O
2
L
−1
) and
DOC (from 3.0 to 1.7 mg‐CL
−1
). The rates of the DO and DOC removal and net
nitrification were greatest during short residence times, while the rate of denitrification was
greatest at long residence times.
15
NO
3
−
tracing confirmed that a fraction of the NO
3
−
removal was via denitrification as
15
N
2
was produced across the entire gravel bar HZ.
Production of
15
N
2
across all observed flow paths and residence times indicated that
denitrification microsites are present even where nitrification was the net outcome. These
findings demonstrate that the HZ is an active nitrogen sink in this system and that the
distinction between net nitrification and denitrification in the HZ is a function of residence
time and exhibits threshold behavior. Consequently, incorporation of HZ exchange and
water residence time characterizations will improve mechanistic predictions of nitrogen
cycling in streams.
Citation: Zarnetske, J. P., R. Haggerty, S. M. Wondzell, and M. A. Baker (2011), Dynamics of nitrate production and removal
as a function of residence time in the hyporheic zone, J. Geophys. Res., 116, G01025, doi:10.1029/2010JG001356.
1. Introduction
[2] Surplus nitrogen adversely affects aquatic systems,
contributing to extensive surface and groundwater degrada-
tion, which is a persistent and growing global problem
[Schlesinger et al., 2006; Diaz and Rosenberg, 2008].
Stream ecosystems can be important locations of N retention
along the continuum between terrestrial and ocean environ-
ments. Research has established that headwater and mid-
network streams are most effective at regulating downstream
nitrogen exports [Peterson et al., 2001; Alexander et al.,
2000; Mulholland et al., 2008]. These same small streams
are also where stream‐groundwater (hyporheic, HZ) flux is
greatest relative to surface flux [Anderson et al., 2005].
Previous work clearly shows that HZ exchange can regulate
nitrogen [Holmes et al. , 1996; Wondzell and Swanson, 1996;
Hill and Lymburner, 1998]. However, the linkages between
HZ hydrology and stream nitrogen export are poorly
understood and there is no clear mechanistic representation
of HZ controls on nitrogen flux through streams [Duff and
Triska, 2000; Böhlke et al., 2009]. Hence, there is a need
to quantify the coupling of HZ hydrology and biogeo-
chemical conditions and their role in creating stream nitrogen
sources and sinks. Until the linkages between HZ hydrology
and nitrogen biogeochemistry are established, it will be
unclear how the HZ influences nitrogen dynamics at reach
and catchment scales. In this study we move toward this
mechanistic understanding of nitrogen fate and transport by
1
Department of Geosciences and Water Resources Graduate Program,
Oregon State University, Corvallis, Oregon, USA.
2
Olympia Forestry Sciences Laboratory, Pacific Northwest Research
Station, Olympia, Washington, USA.
3
Department of Biology and the Ecology Center, Utah State University,
Logan, Utah, USA.
Copyright 2011 by the American Geophysical Union.
0148‐0227/11/2010JG001356
JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 116, G01025, doi:10.1029/2010JG001356, 2011
G01025 1of12
observing and relating the HZ residence timescales to N
transformations.
[
3] There are many processes that temporarily remove or
relocate inorganic N from stream water (e.g., sorption onto
substrate, assimilation into plants or microbes), but denitri-
fication is the primary mechanism by which inorganic N is
permanently removed from the stream system. Conse-
quently, denitrification has been the source of significant
research because of its potential role in regulating the
downstream transport of inorganic N. Denitrification in
streams is primarily regulated by redox conditions, NO
3
−
concentrations, and labile dissolved organic carbon (DOC)
availability [Holmes et al., 1996; Duff et al., 1996; Baker
et al., 1999]. The HZ is known to create strong gradients
in each of these conditions regulating nitrogen cycling
[Triska et al., 1993; Jones and Holmes, 1996; Hill and
Lymburner, 1998; Hedin et al., 1998]. Consequently, the
HZ has been identified as a potential hot spot for denitrifi-
cation in aquatic systems [McClain et al., 2003]. However,
the HZ is not simply a net sink (via denitrification), it can
also be a source (via nitrification) of nitrate where ever
nitrification exceeds denitrification [Jones et al., 1995]. The
nitrate produced in the HZ can fuel primary production in
surface waters [Valett et al., 1994; Henry and Fisher, 2003].
[
4] The role each HZ plays in regulating downstream
nitrogen export is variable in space and time [e.g., Wondzell
and Swanson, 1996]. This is due to the temporal and spatial
variation in substrate and transport limitations on nitrogen
transformations. For example substrate limitations acting on
denitrification, such as the type and quantity of nitrogen and
DOC entering the HZ, can vary significantly in time (e.g.,
seasonally [ Kaplan and Newbold, 2000]). On the other
hand, physical transport conditions of the HZ, which are a
function of energy gradients and hydraulic conductivity, will
regulate the rate at which nitrogen, dissolved oxygen (DO),
and DOC are supplied to the sediment [Baker et al., 2000].
For example, DO exerts a strong control on nitrogen
dynamics in the HZ, and research has shown that DO
dynamics are related to water residence time in the HZ
[Triska et al., 1993; Findlay , 1995; Valett et al., 1996;
Morrice et al., 2000]. Recently, research has started to
integrate physical transport and biogeochemical approaches
to assess HZ denitrification as a function of HZ residence
time [e.g., Gu et al., 2007; Clilverd et al., 2008; Pinay et al.,
2009]. The use of
15
N tracers has also advanced our
understanding of aquatic N cycling. Böhlke et al. [2004,
2009] and Mulholland et al. [2004, 2008] demonstrated the
usefulness of the field
15
N tracer approach for determining
denitrification rates of streams at the reach scale. Böhlke
et al. [2004, 2009], in particular, showed that the key
sources of uncertainty in measuring reach denitrification
with traditional mass balance approaches, nitrification and
nitrate uptake, can be accounted for with the use of
15
N
tracer approach. Although many advances about nitrogen
cycling in streams have resulted from this
15
N tracer
work, these studies were unable to account for the entire
nitrogen budget in streams. Böhlke et al. [2004] and
Mulholland et al. [2004, 2008] suggest that a portion of
the unaccounted for nitrogen may be due to benthic and
hyporheic nitrogen retention and removal processes.
Recently, Böhlke et al. [2009] demonstrated with whole
stream
15
NO
3
−
addition experiments that between 14 and
97% of whole stream denitrification was attributed to HZ
denitrification.
[
5] The objective of this study was to assess the substrate
and transport conditions controlling net HZ denitrification.
We hypothesize that biogeochemical reactions associated
with stream nitrogen cycling, such as nitrification and
denitrification are strongly controlled by water residence
times in the HZ. To test the hypothesis, we conducted a
whole‐stream steady state
15
NO
3
−
and a conservative tracer
experiment in an upland agricultural stream to measure the
in situ spatial and temporal hydraulic and biogeochemical
conditions controlling HZ nitrification and denitrification.
We show here that
15
NO
3
−
tracing techniques characterize
HZ denitrification and that the conditions conducive to net
denitrification vary with subsurface residence times. Ulti-
mately, relating HZ denitrification controls to residence
times will help to upscale denitrification measurements to
reach and network scales in a way that is linked quantita-
tively to transient storage.
2. Methods
2.1. Study Site
[
6] The study site consists of a 303 m reach containing an
instrumented gravel bar hyporheic zone on Drift Creek
(Figure 1a), a third‐order stream within the Willamette
River basin in western Oregon, USA (44.9753°N,
122.8259°W). The drainage area above the study reach is
6517 ha, and has mixed land use dominated by agriculture
(lower catchment) and forestry (upper catchment). The
catchment population is predominantly rural with septic
systems, another potential source of N in the study stream.
Annual precipitation is 1190 mm and comes primarily
during the winter as rain. Base flow discharge gradually
decreases to an annual minimum (<50 L s
−1
) in early Sep-
tember. The study reach was modified by channelization in
the past, as were many of the streams in this agricultural
region. The channelized stream is incised into competent
bedrock consisting of andesite flow breccias and is now
separated from an active floodplain. The incised active
channel is 10–20 m wide and is bounded by steep banks 3–
5 m high. The alluvial thickness above bedrock (as depth to
refusal) varies from 0 to ≥ 1.5 m. Consequently, the reach
has a limited and constrained hyporheic zone. The study
reach has a slope of 0.007 m m
−1
and the morphology is
primarily a planebed channel with occasional pool‐riffle
sequences (see Montgomery and Buffington [1997] for
definitions of channel types). The streambed consists of
poorly sorted sand, gravel, cobbles, and boulders.
[
7] The hyporheic zone study site is a lateral gravel bar
approximately 6.1 m by 4 m (Figure 1b). This gravel bar is
adjacent to a riffle on one side and connected to the bedrock
channel bank on the other side. The gravel bar separates two
pools and spans a head loss across the riffle of 0.13 m. The
alluvium comprising this gravel bar was uniformly 1.2 m
thick. The observed and modeled subsurface exchange
across this gravel bar primarily occurs along lateral flow
paths from the head to the tail of the bar (Figure 1b). This
gravel bar was instrumented with a well network (n = 11) of
3.8 cm I.D. schedule 40 PVC wells screened 0.2–0.4 m
below ground surface. Chloride (Cl
−
) tracer tests conducted
prior to the experiment demonstrated that all wells were
ZARNETSKE ET AL.: RESIDENCE TIME CONTROLS ON HYPORHEIC NITROGEN G01025G01025
2of12
connected to stream water and that well water originated
from the stream and not from the groundwater aquifer. This
lack of groundwater inflows reduces the uncertainty of
15
N
tracing interpretation in this HZ system; groundwater inflow
of nitrogen is a common source of uncertainty seen in other
systems [e.g., Böhlke et al., 2004]. Background stream and
hyporheic biogeochemistry are presented in Table 1.
2.2. Field Procedures
[
8] We performed a whole‐stream steady state d
15
NO
3
−
and conservative tracer (Cl
−
) injection on 23–24 August
2007 when discharge was relatively stable. Following
methods from Mulholland et al. [2004], an injection solu-
tion of d
15
NO
3
−
(as 99% enriched K
15
NO
3
−
) and Cl
−
(as
NaCl
−
) was released at a constant rate (154 mL min
−1
) using
a peristaltic pump at the head of the reach for 27.5 h starting
at 1428 h (Geopump Series I, Geotech Environmental
Equipment, Denver, Colorado, USA, note that the use of
trade names in this publication is for reader information and
does not imply endorsement by the U.S. Department of
Agriculture of any product or service). The amount of
K
15
NO
3
−
introduced to the reach was calculated to produce a
target d
15
N enrichment of 10 000‰ in the stream water
NO
3
−
. The Cl
−
mass addition target was to elevate the
background stream Cl
−
400% and generate an electrical
conductivity increase of 50%. The solution was injected into
a turbulent riffle sufficiently upstream of the first sampling
location to guarantee compete lateral and vertical channel
mixing at all downstream sampling locations. The K
15
NO
3
−
addition produced a 3% increase in ambient stream N‐NO
3
−
.
[
9] Electrical conductivity was used to measure the real‐
time Cl
−
transport through the stream and hyporheic zone.
The electrical conductivity measurements were taken every
60 s in all 11 wells and in the stream water at the head and
tail of the gravel bar. These electrical conductivity mea-
surements were made with 13 multiplexed, in situ, CS547A
conductivity and temperature probes connected to a CR1000
(Campbell Scientific, Logan, Utah, USA). The Cl
−
transport
at the end of the experimental reach was captured via auto-
mated sampling every 10 min until plateau and then every
2 h during plateau (ISCO model 3700, Lincoln, Nebraska,
USA) and subsequently field measured with an electrical
conductivity meter (YSI model 63, Yellow Springs, Ohio,
USA). These electrical conductivity measurements were
used to characterize the solute transport dynamics including
flow rates, flow paths, and residence times as well as to
inform the timing of the sampling regime described below.
[
10] The water sampling regime consisted of collecting
multiple rounds of stream and hyporheic samples during the
two phases of the experiment: preinjection and plateau
(steady state). For each location (11 wells plus stream water
at the gravel bar head), repeated sampling occurred during
the preinjection (n = 3) and plateau (n = 5) periods (Figure 2).
The plateau sampling period was initiated at 22.5 h after
injection when all hyporheic wells demonstrated near steady
state electrical conductivity. Repeated hyporheic samples
were collected approximately every 1 h during the plateau
period.
[
11] Water samples were collected for key solute con-
centrations and d
15
N enrichments relevant to the respiratory
denitrification process (d
15
NO
3
−
, d
15
N
2
(g), as well as NO
3
−
,
NH
3
, DOC, DO, Cl
−
, and specific ultraviolet absorption
(SUVA
254
). Hyporheic well samples were collected with a
field peristaltic pump (Masterflex L/S, Vernon Hills, Illi-
nois, USA) [Woessner, 2007]. All water samples were
immediately filtered through ashed Whatman GF/F glass
fiber filters (0.7 mm pore size) into acid washed HDPE
bottles (60 mL for nutrient chemistry and 1 L for
15
N iso-
tope samples). Following filtering, nutrient chemistry sam-
ples and isotope samples were stored on ice in the field and
later refrigerated at 4°C or frozen in the laboratory until
processed and analyzed. DO concentrations were measured
in situ with a calibrated YSI DO Meter (Model 52) at all
locations prior to collecting each round of samples. Samples
were also collected for d
15
N
2
O (g), but were unable to be
Figure 1. (a) Map of the Drift Creek study site showing
tracer injection site and the gravel bar hyporheic site. (b) Map
of the hyporheic study site showing locations of wells (dots
with cross hairs) and water potentiometic surface during the
injection experiment. Stream briefly bifurcates near gravel
bar (i.e., not a tributary confluence), and water chemistry
is the same across channel. Dashed arrow indicates a single
representative simulated flow path between the head and tail
of the gravel bar.
ZARNETSKE ET AL.: RESIDENCE TIME CONTROLS ON HYPORHEIC NITROGEN G01025G01025
3of12
analyzed due to technical problems at the stable isotope
laboratory. However, denitrification in freshwater and
nearshore marine system sediments consists almost entirely
of N
2
production with N
2
O/N
2
production ratios generally
between < 0.001 and < 0.05 [Seitzinger, 1988; Mulholland
et al., 2004], so
15
N
2
by itself is capable of characterizing
the majority of the denitrification dynamics.
[
12] The
15
N gas collection for each sample occurred in
the field and followed procedures adapted from Hamilton
and Ostrom [2007]. A peristaltic pump was used to collect
80 mL water samples into a 140 mL plastic syringe (Becton‐
Dickinson, Franklin Lakes, NJ, USA) fitted with stopcocks.
All visible bubbles were expelled so that there was no
headspace. To avoid any atmospheric N contamination,
sample syringes were submerged under water in a proces-
sing tub kept at stream temperature. An underwater transfer
of 40 mL high purity He was added to each sample syringe.
Sample syringes were then gently shaken for 10 min to
permit equilibration of the N
2
(g) into the He headspace.
Following equilibration, approximately 14 mL of headspace
gas was then injected into preevacuated 12 mL exetainers
(Labco Limited, Wycombe, UK). Exetainers were pre-
evacuated by pumping them down to a pressure of <50 mTorr
using a Welch vacuum pump (Model DirectTorr 8905,
Skokie, Illinois, USA) and then stored underwater in He
purged DI water‐filled centrifuge tubes until sample collec-
tion. Sample‐filled exetainers were then returned to their
zero headspace He purged DI water‐filled centrifuge tubes
for storage until analysis.
[
13] Following the experiment (5–6 Sept. 2007) and dur-
ing similar stable low flow conditions, we collected detailed
thalweg surface water and channel surface topography data
for the reach using a Topcon total station (Model GTS‐226,
Livermore, California, USA) and standard surveying meth-
ods with spatial resolution of x ≤ 1m,y ≤ 1m,z ≤ 0.01 m
for the greater reach and x ≤ 0.1 m, y ≤ 0.1 m, z ≤ 0.005 m
for the instrumented gravel bar site.
2.3. Laboratory Procedures
[
14] Stream and hyporheic samples were analyzed for
NO
3
−
,NH
3
, DOC, and Cl
−
at the Cooperative Chemical
Analytical Laboratory (Corvallis, Oregon, USA). The NO
3
−
and NH
3
measurements were made by a Technicon Auto‐
Analyzer II. The NO
3
−
and NH
3
nutrient analyses were
performed following standard colorimetric methodology and
had detection limits of 0.001 mg L
−1
and 0.01 mg L
−1
,
respectively. The concentration of total DOC was deter-
mined with a Shimadzu TOC‐VCSH Combustion Analyzer
(Tokyo, Japan; detection limit = 0.05 mg L
−1
). The Cl
−
was
determined by ion chromatography (Dionex 1500, Sunny-
vale, California, USA; detection limit = 0.01 mg L
−1
).
SUVA values were determined by dividing the UV absor-
bance measured at l = 254 nm by the DOC concentration
and are reported in the units of liter per milligram carbon per
meter [Weishaar et al., 2003].
[
15] The d
15
N content of the stream and hyporheic water
NO
3
−
was determined by methods adapted from Sigman
et al. [1997] and Mulholland et al. [2004], which are
briefly summarized below. Prior to d
15
N analysis,
15
NO
3
−
samples with blanks and standards were processed in the
following manner. First, a volume of each sample (0.25–1L;
processing volume is dependent on N content of each sample)
was stripped of its dissolved NH
4
+
and had its NO
3
−
con-
centrated. Second, the concentrated sample NO
3
−
was cap-
tured on a prepared filter via a reduction/diffusion/sorption
Figure 2. The stream at gravel bar and representative distal
HZ well (K2) electrical conductivity (EC) breakthrough
curves showing the timing of the repeated preinjection and
plateau sampling events. Gaps in well EC data represent
times when EC probes were disturbed via sampling.
Table 1. Background Stream and Hyporheic Biogeochemistry
a
Site DO (mg‐O
2
L
−1
)NO
3
−
(mg‐NL
−1
)NH
3
(mg‐NL
−1
) DOC (mg‐CL
−1
)
d
15
N‐NO
3
−
(‰ Versus AIR)
d
15
N‐N
2
(‰ Versus AIR)
Stream 8.31 ± 0.43 0.32 ± 0.01 0.02 ± 0.02 3.01 ± 0.15 2.19 ± 0.61 3.00 ± 0.37
G1 2.07 ± 0.05 0.54 ± 0.06 0.11 ± 0.02 2.07 ± 0.15 5.48 ± 3.33 1.20 ± 0.61
H1 3.27 ± 0.09 0.43 ± 0.05 0.05 ± 0.01 2.18 ± 0.14 5.35 ± 2.27 0.80 ± 0.61
H2 1.30 ± 0.03 0.27 ± 0.03 0.06 ± 0.01 2.01 ± 0.17 5.00 ± 3.26 0.50 ± 1.01
H3 0.72 ± 0.06 0.33 ± 0.03 0.05 ± 0.01 2.01 ± 0.20 14.92 ± 2.25 0.60 ± 1.10
I1 1.09 ± 0.05 0.25 ± 0.04 0.08 ± 0.01 1.94 ± 0.15 18.10 ± 2.88 1.50 ± 0.79
I3 0.93 ± 0.09 0.13 ± 0.00 0.07 ± 0.01 1.98 ± 0.10 14.18 ± 1.99 2.80 ± 0.61
J1 0.70 ± 0.06 0.07 ± 0.02 0.01 ± 0.01 1.66 ± 0.12 4.03 ± 2.18 0.30 ± 3.18
J2 0.61 ± 0.09 0.13 ± 0.01 0.02 ± 0.01 1.79 ± 0.10 8.85 ± 2.30 2.30 ± 0.87
J3 0.51 ± 0.05 0.11 ± 0.01 0.04 ± 0.01 1.76 ± 0.08 11.35 ± 2.13 3.30 ± 3.46
K2 0.65 ± 0.05 0.08 ± 0.01 0.04 ± 0.01 1.71 ± 0.06 8.13 ± 2.25 0.00 ± 2.82
K3 0.59 ± 0.10 0.09 ± 0.01 0.02 ± 0.01 1.70 ± 0.11 10.25 ± 2.96 0.00 ± 0.37
a
Mean of three observations before injection ± 1 standard error.
ZARNETSKE ET AL.: RESIDENCE TIME CONTROLS ON HYPORHEIC NITROGEN G01025G01025
4of12
procedure (full reduction of NO
3
−
to NH
4
+
, which is then
converted to NH
3
that diffuses into the headspace and ulti-
mately gets captured on the acidified sorption filter). After
complete transfer of NO
3
−
to the sample filter, the samples
were sealed and sent for
15
NO
3
−
analysis. All d
15
NO
3
−
and
d
15
N‐gas samples were analyzed by the Marine Biological
Laboratory Stable Isotope Facility (MBL, Woods Hole,
Massachusetts, USA). Replicate analyses of the water and gas
samples show the precision of d
15
NO
3
−
and d
15
N
2
isotope
measurements is ±80.0‰ and ±0.2‰, respectively.
2.4. Parameter Calculations
[
16] Electrical conductivity breakthrough curves (as a
measure of Cl
−
transport) at the head of the gravel bar and in
each well were used to measure the median residence time
of the study reach and the HZ water in each well. The
median residence time was calculated as the time required to
raise the EC in the well to one half the plateau concentra-
tion. In the case of the wells, median residence times were
calculated based on the observed Cl
−
arrival at the gravel
bar, not the start of the injection.
[
17] To estimate reactive solute transport through and
removal by the HZ, we compared the observed NO
3
−
,NH
3
,
DO, and DOC concentrations to the conservative tracer
concentrations at steady state conditi ons [after Morrice
et al., 2000]. In the absence of biological or chemical
removal, the reactive compounds and Cl
−
transport should
be identical. Based upon this assumption, we calculated the
NO
3
−
,NH
3
, DO, and DOC concentrations according to the
measured Cl
−
concentrations observed at each well:
R
pred;x;t
¼ R
inj;x
C
x;t
C
x;t¼0
C
inj
C
x;t¼0
þ R
x;t¼0
1
C
x;t
C
x;t¼0
C
inj
C
x;t¼0
ð1Þ
where, R is the solute concentration of interest (NO
3
−
,NH
3
,
DO, or DOC), C is the conservative tracer concentration
(Cl
−
), and subscripts pred, inj, x, t represent predicted well
concentration at plateau, injection concentration in the
stream, well location, and sample time period, respectively.
We then calculated the difference between the measured and
predicted reactive solute concentrations for each well during
the plateau conditions. Reactive solute removal occurs when
the observed concentration is less than the predicted con-
centration and production occurs when the observed con-
centration is greater than the predicted concentration.
3. Results
3.1. Stream Hydrology and Chemical Conditions
During Experiment
[
18] Streamflow conditions were relatively stable over the
experiment with a mean flow of 22 L s
−1
and a variance of ±
2.2 L s
−1
. This variance in flow did not create any mea-
surable change in the stage of the stream along the reach,
near the gravel bar, or in the heads of the gravel bar wells.
Surveying of channel topography and geometry yielded a
reach mean depth, d, of 0.23 m and a mean wetted width, w,
of 5.21 m. Repeated measurements of the head at wells
(before plateau sampling disturbances) reflected stable sur-
face water elevations as there was no detectible variation
during the experiment. Stream and HZ water temperature
ranged between 14.1 and 16.5°C during the injection period.
Measured surface water nutrient and chemistry conditions
were stable across the experiment and did not show diel
patterns with NO
3
−
(0.318–0.325 mg‐NL
−1
), NH
3
(0.021–
0.024 mg‐NL
−1
), DOC (2.95–3.45 mg‐CL
−1
), DO (8.10–
8.51 mg‐O
2
L
−1
), and pH (6.65–6.85).
3.2. Spatial Dynamics of Hyporheic Transport and N
Transformation Conditions
[
19] Chloride plateau concentration conditions were
achieved in all 11 hyporheic wells, demonstrating good
connectivity with surface water. Nominal flow path lengths
(Figure 3) ranged from 0.5 m (H1) to 4.2 m (K3). The mean
DO decreased from 8.31 to 0.59 mg‐O
2
L
−1
along HZ flow
paths (Figure 4a) and the mean DOC decreased from 3.01 to
1.7 mg‐CL
−1
(Figure 4d). The DOC SUVA
254
concentra-
tions were more spatially variable than DOC, indicating that
different locations within the HZ had different quantities of
aromatic DOC (Figure 4d, contour map), but did generally
decrease along flow paths (3.22 to 0.94 L mg‐C
−1
m
−1
).
Along flow paths, DO and DOC removal rates were largest
in the first 2 m of the flow paths but continued across
the entire gravel bar. In contrast, the N species did not
consistently decrease along flow paths. Nitrate increased at
the proximal end of the flow paths (<0.55 m) from 0.34 to
0.54 mg‐NL
−1
and then decreased along the remainder of
the flow paths from 0.54 to 0.02 mg‐NL
−1
(Figure 4b).
Similarly, NH
3
increased from 0.02 to 0.11 mg‐NL
−1
at
the proximal end of the flow paths and then decreased from
0.11 to 0.01 mg‐NL
−1
along the remainder of the flow paths
(Figure 4c).
[
20] Tracing of
15
NO
3
−
confirmed that a fraction of the
NO
3
−
removal was via denitrification as
15
N
2
was produced
across the entire gravel bar HZ (Figure 4f). Production of
15
N
2
occurred along all portions of the flow paths, even
portions characterized by net nitrification (elevated NO
3
−
and
NH
3
). Importantly, there was no consistent spatial gradient
in the
15
NO
3
enrichment from proximal to distal ends of the
HZ flow paths (Figure 4e) which, if present, would indicate
nonsteady state dynamics. Therefore the range of
15
NO
3
−
Figure 3. Nominal flow path length and median residence
time for each hyporheic zone well.
ZARNETSKE ET AL.: RESIDENCE TIME CONTROLS ON HYPORHEIC NITROGEN G01025G01025
5of12
Figure 4. Spatial steady state hyporheic zone biogeochemical and
15
N conditions: (a) dissolved oxygen
(DO, mg‐O
2
L
−1
), (b) nitrate (NO
3
−
,mg‐NL
−1
), (c) ammonia (NH
3
,mg‐NL
−1
), (d) total dissolved
organic carbon (DOC, mg‐CL
−1
) with SUVA
254
contours (interval equals 0.1 L mg
−1
m
−1
), (e) d
15
N‐
nitrate (d
15
NO
3
−
, ‰ versus AIR), and (f) d
15
N‐dinitrogen (d
15
N
2
, ‰ versus AIR). Maps present spatially
interpolated mean values generated from repeated samples (n = 5) collected during tracer plateau condi-
tions. Stream water values did not vary between the head and tail of the gravel bar.
ZARNETSKE ET AL.: RESIDENCE TIME CONTROLS ON HYPORHEIC NITROGEN G01025G01025
6of12
enrichment is a function of steady state hydrologic and
biological conditions. The gravel bar plateau
15
NO
3
−
enrichment ranged from 4 260‰ (well I3) to 6 805 (well
J2), while the stream water ranged from 9 935 to 10 092‰.
3.3. Temporal Dynamics of Hyporheic Transport
and N Transformation
[
21] Tracing Cl
−
transport through the HZ generated
median residence times ranging from 3.8 to 28.5 h (Figure 3).
The shortest residence times were generally associated with
the head of the gravel bar while the longest were located at
the tail of the gravel bar. However, the longest median res-
idence time (28.5 h) was observed at a midbar well, J1.
Comparison of residence times to measured biogeochemical
conditions (Figure 5) indicates that residence times less than
6.9 h were associated with a net dominance of oxic con-
ditions and aerobic microbial processes (O
2
respiration and
nitrification) while residence times beyond 6.9 h were
associated with a net dominance of hypoxic‐anoxic condi-
tions and anaerobic microbial processes (denitrification).
More specifically, the greatest rates of DO, DOC, and
SUVA
254
reduction corresponded with the greatest rates of
NH
3
and NO
3
−
production (Figure 5), all of which cooc-
curred during the first 6.9 h of observed transport. Beyond
6.9 h of residence time, DO, DOC, and SUVA
254
continue
to decrease gradually to a minimum of 0.51 mg‐O
2
L
−1
,
1.66 mg‐CL
−1
, and 0.94 L mg‐C
−1
m
−1
, respectively.
Further, SUVA
254
indicates that the highest fraction of
labile DOC occurred when residence times were smallest,
and that labile DOC was largely depleted by the time water
attained larger residence times (Figure 5d).
[
22] The concentration of
15
N
2
increased along proximal
portions of flow paths with residence times < 3.8 h, reaching
a peak at the most distal portions of the flow paths where
residence times exceeded 22 h (Figure 5f). The concentra-
tion of NO
3
−
(mg‐NL
−1
) also increased along the proximal
portions of flow paths where residence times were less than
6.9 h, indicating concurrent nitrification and denitrification
throughout the proximal portion of the gravel bar (Figure 5).
Increases in NH
3
were concurrent with consumption of DO
and DOC (Figure 6). Comparison of DO and DOC con-
centrations with the conservative transport of Cl
−
demon-
strates that DO and DOC show net loss over all residence
times (Figure 6). Conversely, NH
3
shows net production
until the very longest residence time of 28.5 h when it shows
net loss, while NO
3
−
shows net production until 18.2 h fol-
lowed by net loss.
4. Discussion
[23] We utilized a whole‐stream steady state
15
NO
3
−
and
conservative tracer (Cl
−
) addition to observe spatial and
temporal hydraulic and physiochemical conditions control-
ling NO
3
−
dynamics in a HZ. Our results illustrate that
nitrification and denitrification: (1) create nonlinear NO
3
−
dynamic along HZ transport, (2) are not exclusively segre-
gated processes in space and time, and (3) are strongly
controlled by water and solute residence times in the HZ.
From these findings we are able to confirm and build upon
earlier conceptual frameworks [Jones and Holmes, 1996;
Valett et al., 1996; Hedin et al., 1998] that relate HZ nitri-
fication and denitrification dynamics along flow paths and
with residence times.
4.1. Spatial and Residence Time Dynamics
of Hyporheic N
[
24] Nitrate production versus removal can be site‐ and
scale‐dependent, and the hyporheic biogeochemistry in our
study shows general spatial patterning in net N transfor-
mation processes consistent with earlier studies [e.g.,
Holmes et al., 1994; Pinay et al., 1994; Holmes et al., 1996].
The upgradient end of the HZ flow paths is dominated by
oxic conditions and is a net NO
3
−
production hot spot, while
the middle and downgradient parts of the flow paths are
anoxic and are a net NO
3
−
removal hot spot (Figure 4). At the
scale of the entire gravel bar, however, this HZ is a net NO
3
−
removal hot spot for the stream. In contrast, other, smaller
HZ units in the stream may be production hot spots because
of their shorter flow paths and residence times. The resi-
dence time can be used to mark where this HZ turns from
one redox condition to another: from net NO
3
−
production to
net NO
3
−
removal (Figure 7). In this HZ, a water residence
time of 6.9 h marks the threshold that separates these con-
ditions. This observed threshold supports the HZ N trans-
formation conceptual framework put forth by Jones and
Holmes [1996] and Hedin et al. [1998]. The creation of
this residence time threshold is complex and is a function of:
(1) HZ water temperature (as it controls microbial activity
and DO in water), (2) concentration of DO across the HZ
(controlled by biological oxygen demand and advected
supply), (3) HZ DOC supply and quality, (4) amount of
NO
3
−
in the HZ system, and (5) the physical hydraulics
subsumed in the physical residence time of water (e.g., head
gradient, hydraulic conductivity, advection, and dispersion).
[
25] As shown in Figure 7, at times shorter than the
threshold, transport and substrate conditions promote min-
eralization of stream sourced DON or particulate organic
matter and subsequent nitrification with denitrification (N
2
production) limited to microsite react ions [Sheibley et al.,
2003]. As residence times increase, the extent of anaerobic
water in the HZ grows and the effective nitrification rate
decreases until its contribution to net NO
3
−
production is
negligible. At the threshold residence time, both processes
are cooccurring and this is likely the location where the
greatest rates of denitrification will be observed. Beyond the
threshold, the HZ is dominated by net denitrification, and
concentrations of both NH
3
and NO
3
−
decrease rapidly as
denitrification is not substrate (labile DOC) limited and NO
3
−
production rapidly decreases. Finally, at much longer resi-
dence times, the rate of denitrification will decrease due to
an increasing DOC substrate limitation and lack of NO
3
−
,
even though redox conditions remain appropriate to carry
out denitrification.
[
26] Similar nitrification‐denitrification coupling has been
observed in stream HZs where DON ‐rich anoxic ground-
water flows into oxic stream sediments [e.g., Hedin et al.,
1998; Sheibley et al., 2003]. The thermodynamic frame-
work of Hedin et al. [1998] clearly illustrates that microbial
redox processes, including nitrification‐denitrification, will
be tightly coupled in the riparian and hyporheic environ-
ments where solutes exchange across oxic‐anoxic bound-
aries. Our study illustrates this tight coupling occurs in space
and time as shown by the spatial distribution of N species
ZARNETSKE ET AL.: RESIDENCE TIME CONTROLS ON HYPORHEIC NITROGEN G01025G01025
7of12
Figure 5. Steady state hyporheic zone biogeochemical and
15
N conditions relative to median residence
time: (a) dissolved oxygen (DO), (b) nitrate (NO
3
−
), (c) ammonia (NH
3
), (d) total disolved organic carbon
(DOC) with SUVA
254
, (e) d
15
N‐nitrate (d
15
NO
3
−
), and (f) d
15
N‐dinitrogen (d
15
N
2
). Each data point repre-
sents the mean values generated from repeated samples (n = 5, error bars = ± 2 standard error) collected
during tracer plateau conditions. Stream water values are shown as a median residence time = 0 h.
ZARNETSKE ET AL.: RESIDENCE TIME CONTROLS ON HYPORHEIC NITROGEN G01025G01025
8of12
(Figure 4) and the residence time threshold between net
nitrification and denitrification (Figure 7). This suggests that
a thermodynamic approach is useful and should be coupled
to physical transport dynamics in order to better understand
the spatial and temporal distribution of different hyporheic
redox environments.
[
27] Earlier studies [e.g., Valett et al., 1996; Pinay et al.,
2009] indicate that biologically mediated N fluxes through
the HZ are explicitly a function of HZ flow path length and
implicitly a function of residence time (i.e., space for time
[see Jones and Holmes, 1996, and references therein]). In
this study,
15
N tracing explicitly shows that there is a rela-
tionship between N fluxes and HZ residence time. Earlier
studies by Valett et al. [1996] and Pinay et al. [2009] col-
lected measures of HZ solute transit times during their N
addition experiments and found that NO
3
−
uptake and deni-
trification were the dominant N transformations processes.
However, they did not observe the coupling of nitrification
and denitrification (i.e., nitrification preceding denitrifica-
tion along increasing residence times). In their studies, NO
3
−
uptake and denitrification processes occurred across all
measured transport times achieving maximums during the
first 1.7 h of travel time. Conversely, our study shows a
different N removal pattern, one that is more complex, with
nitrification occurring at short residence times and denitri-
fication at late residence times (Figures 6 and 7). A main
difference between our study and these earlier studies is the
use of
15
NO
3
−
to trace N dynamics under near ambient NO
3
−
conditions. Valett et al. [1996] and Pinay et al. [2009]
elevated NO
3
−
concentrations in their HZ studies (6.4‐ to
24‐fold NO
3
−
increase, respectively). NO
3
−
transformations
are concentration‐dependent, and elevated NO
3
−
conditions
can increase rates of HZ uptake and denitrification [Jones
and Holmes, 1996]. Consequently, the use of
15
NO
3
−
allowed us to observe different NO
3
−
transformations under
approximately ambient concentrations.
[
28] In addition to NO
3
−
removal via denitrification, bac-
terial assimilation is an important N retention process and
may account for a significant fraction of the observed DOC
and NO
3
−
removal across the HZ. Following Sobczak et al.
[2003], using the observed DOC loss across the HZ =
1.35 mg‐CL
−1
, and assuming a microbial growth efficiency
of 50% and a microbial C:N = 7:1, a potential 0.68 mg‐CL
−1
and 0.1 mg‐NL
−1
could be assimilated into microbial bio-
mass. Under these assumptions, microbial assimilation can
Figure 6. Apparent net production and removal based upon conservative transport relative to median
residence time for: (a) dissolved oxygen (DO), (b) nitrate (NO
3
−
), (c) ammonia (NH
3
), and (d) total
disolved organic carbon (DOC). Positive values represent production, negative values represent removal,
and the zero line represents con servative transport of compound. Each data point represents the mean
values generated from repeated samples (n = 5, error bars = ± 2 standard error) collected during tracer
plateau conditions.
ZARNETSKE ET AL.: RESIDENCE TIME CONTROLS ON HYPORHEIC NITROGEN G01025G01025
9of12
account for 21% of the NO
3
−
loss across the gravel bar HZ.
This microbial assimilation of NO
3
−
is supported by the
lower
15
N enrichment values observed in the HZ
15
NO
3
−
compared to the stream surface waters (Figure 4). Both
microbial assimilation and denitrification will act to lower
15
N enrichment of the HZ
15
NO
3
−
pool. After accounting
for assimilation of NO
3
−
, the respiratory denitrification may
account for as much as 79% of the total NO
3
−
removed from
stream water flowing through the HZ of this gravel bar.
[
29] DOC quantity and quality also clearly depend on
residence time. This is expected given the role DOC plays as
a substrate for the observed nitrification and denitrification.
Previous work has demonstrated a strong positive relation-
ship between DOC loss and bacterial productivity along
hyporheic flow paths [Sobczak and Findlay, 2002]. The
concurrent declines in DOC and DO concentration with
increasing flow path length and residence times indicates
strong aerobic metabolism in this gravel bar HZ. Similar
observations of DOC declines were also seen along flow
path length in the controlled mesocosm experiments of
Sobczak et al. [2003]; they found that the DOC dynamics
were a function of rapid microbial utilization of the bio-
available fractions of the DOC followed by conservative
transport of the unavailable DOC fraction. In this study, we
are able to relate these DOC spatial dynamics to transport
times. In doing so, we see the same conservative transport of
the less labile fraction of the DOC at longer flow paths and
residence times (Figure 5d); however, some labile DOC
must be present at later residence times to provide necessary
substrate for denitrification.
4.2. The
15
N Tracing Shows Overlapping N Process
Domains
[
30] At the organismal level, microbial denitrification
requires anoxic conditions. However, it is well known that
denitrification can occur in anoxic biofilms within bulk
conditions that are oxic [e.g., Holmes et al., 1996]. Hence, it
is possible for denitrification and nitrification to proceed
concurrently in a small volume of the HZ. Our observation
of
15
N
2
production at wells exhibiting net NO
3
−
production
(Figure 7) indicates that over short residence times the water
has encountered denitrification microsites in less mobile
pore water. When low‐flow water samples were collected
from wells, water and solutes were accessed from both
mobile and immobile domains, as seen in other groundwater
well sampling regimes [e.g., Harvey and Gorelick, 2000]. In
the case of our study, at short residence times, the advec-
tion‐dominated mobile domain likely supports the aerobic
processes and the diffusion‐dominated immobile domain
supports anaerobic processes. In contrast, at long residence
times both mobile and immobile domains become anoxic as
the DO is utilized along the flow path. The use of
15
N
tracing permitted us to see that a HZ water sample contains
signals from multiple distinct N transformation domains.
This overlapping domain complexity can be accounted for
with a residence time distribution perspective, because the
residence time distribution of sample volume integrates the
Figure 7. Conceptual model showing a continuum between net hyporheic nitrification and denitrifica-
tion conditions (labeled lines) as a function residence time. Note that the conceptual model is overlaid on
the observed steady state nitrification‐denitrification species, dissolved organic carbon, and dissolved
oxygen conditions: dissolved oxygen (DO, green circles), nitrate (NO
3
−
, blue diamonds), ammonia
(NH
3
, gray triangles), total d issolved organic carbon (DOC, orange squares), and d
15
N‐dinitrogen
(d
15
N
2
, red circles). Each data point represents the normalized mean values generated from repeated pla-
teau samples (n = 5, error bars = ± 2 standard error, values were normalized by maximum observed con-
centration or
15
N enrichment).
ZARNETSKE ET AL.: RESIDENCE TIME CONTROLS ON HYPORHEIC NITROGEN G01025G01025
10 of 12
effect of different process domains encountered during
hyporheic transport (e.g., large subsurface heterogeneity and
gradients created by advective‐dispersive transport, mixing
flow paths, or substrate patchiness).
[
31] Future work should explore the role of these over-
lapping process domains in light of other possible explana-
tions that we are unable to address with the current study. For
example, labile DOC subsurface heterogeneity can also
generate spatial and temporal N transformation heterogene-
ity. Another issue is that rapid localized redox condition
changes could be created while drawing the low‐flow water
sample. During the sampling, an otherwise anoxic region in
the HZ around the well may have more oxic water drawn into
the well via the pumping of upgradient preferential flow
paths. In this case, a denitrification signal from the anoxic
pore waters surrounding the well will be combined with the
upgradient waters properties such as higher DO concentra-
tions and nitrification.
5. Conclusions
[32] Results from the coupled conservative and
15
N
reactive tracer experiment provide definitive evidence of in
situ denitrification occurring in the HZ. Further, the com-
parison between conservative tracer and reactive
15
N tracers
enable us to relate the fate and transport of DO, DOC, and
NO
3
−
to water residence times. In this hyporheic zone, short
residence times were dominated by aerobic metabolic pro-
cesses such as the rapid utilization of DO and DOC and the
production of ammonia and nitrate (ammonification and
nitrification, respectively). However, a clear denitrification
signal, concurrent in both space and time, was also observed
during short residence times indicating anaerobic microsites.
Beyond a residence time threshold of 6.9 h in this HZ, the
anaerobic metabolic process of denitrification dominated
the system, and resulted in a net removal of nitrate from the
stream. Thus, this HZ was a hot spot for nitrogen transfor-
mation, where hot spots of nitrate production and removal
were distinguished by residence time.
[
33] In this gravel bar HZ, the combination of
15
N tracing
with a relatively elegant, highly instrumented hydrologic
system (Figure 1), we see that residence time helps make N
transformation relationships more clear (Figure 7). The
actual spatial and temporal location of the threshold between
net N transformation process domains is expected to vary
between representative HZ units given their exact combi-
nation of hydrologic and upgradient biogeochemical and
substrate characteristics. Further, this threshold is likely to
vary in time with daily and seasonal changes in hydraulic,
temperature, and water chemistry conditions. Ultimately,
relating hyporheic denitrification controls to residence times
will help to upscale denitrification measurements to stream
reach and network scales in a way that is linked quantita-
tively to transient storage.
[
34] Acknowledgments. Support for this project was pr ovided via
NSF grant EAR‐041240 to R.H., S.M.W., and M.A.B. and NSF grant
DGE‐0333257 and OSU Institute for Water and Watersheds (IWW) grant
to J.P.Z. Further support was provided by the Hollis M. Dole Environmen-
tal Geology Fund at OSU. Any opinions, findi ngs, and conc lusions or
recommendations exp ressed in this m aterial are those of the auth ors
and do not necessarily reflect the views of the NSF. We thank the asso-
ciate editor and two anonymous reviewers whose comments impro ved
this manuscript. Special thanks to: V. Adams, S. Baxter, P. Zarnetske,
A. Argerich, and B. Burkholde r for field/lab assistance; L. Ashkenas
and S. Thomas for advising J.P.Z. on stable
15
N handling; M. Otter
of MBL’s Stable Isotope Laboratory for analyzing
15
N samples; and
C. Jones and K. Motter of CCAL and OSU IWW Collaboratory for help
with analyzing general water chemistry.
References
Alexander, R. B., et al. (2000), Effect of stream channel size on the delivery
of nitrogen to the Gulf of Mexico, Nature, 403,758–761, doi:10.1038/
35001562.
Anderson, J. K., et al. (2005), Patterns in stream longitudinal profiles and
implications for hyporheic exchange flow at the H. J. And rews Experi-
mental Forest, Oregon, USA, Hydrol. Process., 19, 2931–2949,
doi:10.1002/hyp.5791.
Baker, M. A., et al. (1999), Acetate retention and metabolism in the hypor-
heic zone of a mountain stream, Limnol. Oceanogr., 44, 1530–1539,
doi:10.4319/lo.1999.44.6.1530.
Baker, M. A., et al. (2000), Organic carbon supply and metabolism in a
shallow groundwater ecosystem, Ecology, 81, 3133–3148, doi:10.1890/
0012-9658(2000)081[3133:OCSAMI]2.0.CO;2.
Böhlke, J. K., et al. (2004), Reach‐scale isotope tracer experiment to quan-
tify denitrification and related processes in a nitrate‐rich stream, midcon-
tinent United States, Limnol. Oceanogr., 49, 821–838, doi:10.4319/
lo.2004.49.3.0821.
Böhlke, J. K., et al. (2009), Multi‐scale measurements and modeling of
denitrification in streams with varying flow and nitrate concentration in
the upper Mississippi River basin, USA, Biogeochemistry, 93, 117–
141, doi:10.1007/s10533-008-9282-8.
Clilverd, H. M., et al. (2008), Nitrogen retention in the hyporheic zone of a
glacial river in interior Alaska, Biogeochemistry, 88,31–46, doi:10.1007/
s10533-008-9192-9.
Diaz, R . J., and R. Rosenberg (2008), Spreading dead zones and conse-
quences for marine ecosystems, Science, 321, 926–929, doi:10.1126/
science.1156401.
Duff, J. H., and F. J. Triska (2000), Nitrogen biogeochemistry and surface‐
subsurface exchange in streams, in Streams and Ground Waters, edited
by J. A. Jones and P. J. Mulholland, pp. 197–217, Academic, San Diego,
Calif.
Duff, J. H., et al. (1996), Nitrate reduction in sediments of lowland tropical
streams draining swamp forest in Costa Rica: An ecosystem perspective,
Biogeochemistry, 33, 179–196, doi:10.1007/BF02181071.
Findlay, S. (1995), Importanc e of surface‐subsurface exchange in stream
ecosystems: The hyporheic zone,
Limnol. Oceanogr., 40,159–164,
doi:10.4319/lo.1995.40.1.0159.
Gu, C. H., G. M. Hornberger, A. L. Mills, J. S. Herman, and S. A. Flewelling
(2007), N itrate reduction in streambed sediments: Effects of flow and
biogeochemical kinetics, Water Resour. Res., 43, W12413, doi:10.1029/
2007WR006027.
Hamilton, S. K., and N. E. Ostrom (2007), Measurement of the stable iso-
tope ratio of dissolved N‐2inN‐15 tracer experiments, Limnol. Ocea-
nogr. Methods, 5, 233–240.
Harvey, C. F., and S. M. Gorelick (2000), Rate‐limited mass transfer or
macrodispersion: Which dominates plume evolution at the Macrodisper-
sion Experiment (MADE) site?, Water Resour. Res., 36,637–650,
doi:10.1029/1999WR900247.
Hedin, L. O., et al. (1998), Thermodynamic constraints on nitrogen trans-
formations and other biogeochemical processes at soil‐stream interfaces,
Ecology, 79, 684–703.
Henry, J., and S. Fisher (2003), Spatial segregation of periphyton commu-
nities in a desert stream: Causes and consequences for N cycling, J. N.
Am. Benthol. Soc., 22, 511–527, doi:10.2307/1468349.
Hill, A. R., and D. J. Lymburner (1998), Hypor heic zone chemi stry and
stream‐subsurface exchange in two groundwater‐fed streams, Can. J.
Fish. Aquat. Sci., 55, 495–506, doi:10.1139/cjfas-55-2-495.
Holmes, R. M., et al. (1994), Parafluvial nitrogen dynamics in a desert
stream ecosystem, J. N. Am. Benthol. Soc., 13,468–478, doi:10.2307/
1467844.
Holmes, R. M., et al. (1996), Denitrification in a nitrogen‐limited stream
ecosystem, Biogeochemistry, 33, 125–146, doi:10.1007/BF02181035.
Jones, J. B., and R. M. Holmes (1996), Surface‐subsurface interactions in
stream ecosystems, Trends Ecol. Evol., 11, 239–242, doi:10.1016/0169-
5347(96)10013-6.
Jones,J.B.,S.G.Fisher,andN.B.Grimm(1995),Nitrificationinthe
hyporheic zone of a desert stream ecosystem, J. N. Am. Benthol. Soc.,
14, 249–258, doi:10.2307/1467777.
Kaplan, L. A., and J. D. Newbold (2000), Surface and subsurface dissolved
organic carbon, in Streams and Ground Waters, edited by J. A. Jones and
P. J. Mulholland, pp. 237 –253, Academic, San Diego, Calif.
ZARNETSKE ET AL.: RESIDENCE TIME CONTROLS ON HYPORHEIC NITROGEN G01025G01025
11 of 12
McClain, M. E., et al. (2003), Biogeochemical hot spots and hot moments
at the interface of terrestrial and aquatic ecosystems, Ecosystems, 6, 301–
312, doi:10.1007/s10021-003-0161-9.
Montgomery, D. R., and J. M. Buffington (1997), Channel‐reach morphol-
ogy in mountain drainage b asins, Geol. Soc. Am. Bull. , 109,596–611,
doi:10.1130/0016-7606(1997)109<0596:CRMIMD>2.3.CO;2.
Morrice, J. A., et al. (2000), Terminal electron accepting processes in the
alluvial sediments of a headwater stream, J. N. Am. Benthol. Soc., 19,
593–608, doi:10.2307/1468119.
Mulholland, P. J., et al. (2004), Stream denitri fication and total nitrate
uptake rates measured using a field N‐15 tracer addition approach,
Limnol. Oceanogr., 49, 809–820, doi:10.4319/lo.2004.49.3.0809.
Mulholland, P. J., et al. (2008), Stream denitrification across biomes and its
response to anthr opogenic nitrate loading, Nature, 452, 202–205,
doi:10.1038/nature06686.
Peterson, B. J., et al. (2001), Control of nitrogen export from watersheds by
headwater streams, Science, 292,86–90, doi:10.1126/science.1056874.
Pinay, G., et al. (1994), The role of denitrification in nitrogen removal
in river corridors, in Global Wetlands: Old World and New,editedby
W. J. Mitsch, pp. 107–116, Elsevier, Amsterdam.
Pinay, G., et al. (2009), Nitrate removal in the hyporheic zone of a salmon
river in Alaska, River Res. Appl., 25, 367–375, doi:10.1002/rra.1164.
Schlesinger, W. H., et al. (2 006), Global change: T he nitrogen cycle and
rivers, Water Resour. Res., 42, W03S06, doi:10.1029/2005WR004300.
Seitzinger, S. P. (19 88), Denitrification in freshwater and coa stal marine
ecosystems: Ecological and geochemical significance, Limnol. Ocea-
nogr., 33, 702–724, doi:10.4319/lo.1988.33.4_part_2.0702.
Sheibley, R. W., et al. (2003), Numerical modeling of coupled nitrification‐
denitrification in sediment perfusion cores from the hyporheic zone of the
Shingobee River, MN, Adv. Water Resour., 26,977–987, doi:10.101 6/
S0309-1708(03)00088-5.
Sigman, D. M., et al. (1997), Natural abundance‐level measurement of the
nitrogen isotopic composition of oceanic nitrate: An adaptation of the
ammonia diffusion method,
Mar. Chem., 57, 227–242, doi:10.1016/
S0304-4203(97)00009-1.
Sobczak, W. V., and S. Findlay (2002), Variation in bioavailability of dis-
solved organic carbon among stream hyporheic flowpaths, Ecology, 83,
3194–3209, doi:10.1890/0012-9658(2002)083[3194:VIBODO]2.0.
CO;2.
Sobczak, W. V., et al. (2003), Relationships between DOC bioavailability
and nitrate re moval i n an upland stream: An experimental approach,
Biogeochemistry, 62, 309–327, doi:10.1023/A:1021192631423.
Trisk a, F. J., et al. (1993), The role of water exchange between a stream
channel and its hyporheic zone in nitrogen cycling at the terrestrial‐
aquatic interface, Hydrobiologia, 251,167–184, doi:10.1007/
BF00007177.
Valett, H. M., S. G. Fisher, N. B. Grimm, and P. Camill (1994), Vertical
hydrologic exchange and ecological stability of a desert stream ecosys-
tem, Ecology, 75, 548–560, doi:10.2307/1939557.
Valett, H. M., et al. (1996), Parent lithology, surface‐groundwater
exchange, and nitrate retention in headwater streams, Limnol. Oceanogr.,
41, 333–345, doi:10.4319/lo.1996.41.2.0333.
Weishaar, J. L., et al. (2003), Evaluation of specific ultraviolet absorbance
as an indicator of the c hemical co mposition and reactivity of dissolved
organic carbon, Environ. Sci. Tec hnol., 37, 4702– 4708, doi:10.1021/
es030360x.
Woessner, W. W. (2007), Building a compact, low‐cost, and portable peri-
staltic sampling pump, Ground Water, 45, 795–797, doi:10.1111/j.1745-
6584.2007.00346.x.
Wondzell, S. M., and F. J. Swanson (1996), Seasonal and storm dynamics
of the hyporheic zone of a 4th‐order mountain stream. 2. Nitrogen
cycling, J. N. Am. Benthol. Soc., 15,20–34, doi:10.2307/1467430.
M. A. Baker, Department of Biology, Utah State University, Logan, UT
84322, USA.
R. Haggerty and J. P. Zarnetske, Depa rtment of Geosciences, Oregon
State University, Corvallis, OR 97331, USA. (zarnetsj@geo.oregonstate.edu)
S. M. Wondzell, Olympia Forestry Sciences Laboratory, Pacific
Northwest Research Station, Olympia, WA 98512, USA.
ZARNETSKE ET AL.: RESIDENCE TIME CONTROLS ON HYPORHEIC NITROGEN G01025G01025
12 of 12
