Differences in Dissolved Organic Carbon and Nitrogen - SUNY-ESF

JOURNAL OF THE AMERICAN WATER RESOURCES ASSOCIATION
Vol. 44, No. 6
AMERICAN WATER RESOURCES ASSOCIATION
December 2008
DIFFERENCES IN DISSOLVED ORGANIC CARBON AND NITROGEN RESPONSES
TO STORM-EVENT AND GROUND-WATER CONDITIONS IN A FORESTED,
GLACIATED WATERSHED IN WESTERN NEW YORK1
Shreeram Inamdar, Julia Rupp, and Myron Mitchell2
ABSTRACT: Differences in the storm-event responses of dissolved organic carbon (DOC) and nitrogen (DON) in
streamflow and ground water were evaluated for a glaciated forested watershed in western New York. Eight to
ten storm events with varying rainfall amounts, intensities, and antecedent moisture conditions were studied for
three catchments (1.6, 3.4, and 696 ha) over a three-year period (2003-2005). Concentrations of DOC in streamflow exiting the catchments were significantly higher for storm events following a dry period, whereas no similar
response was observed for DON. Highest DON concentrations in streamflow were typically associated with storm
events following wet antecedent moisture conditions. In addition to antecedent moisture conditions, DOC concentrations were also positively correlated with precipitation amounts, while DON did not reveal a consistent pattern. Streamwater and ground-water concentrations of DOC during storm events were also strongly correlated
with riparian ground-water depths but a similar relationship was not observed for DON. Ground-water DON
concentrations were also more variable than DOC. We hypothesized that the differences in DOC and DON
responses stemmed from the differences in catchment sources of these solutes. This study suggests that while
DOC and DON are intrinsically linked as components of dissolved organic matter, their dynamics and exports
from watersheds may be regulated by a different set of mechanisms and factors. Identifying these differences is
critical for developing more reliable and robust models for transport of dissolved organic matter.
(KEY TERMS: dissolved organic matter; nitrogen; solute flushing; storm events; ground water; antecedent moisture conditions; drying and rewetting.)
Inamdar, Shreeram, Julia Rupp, and Myron Mitchell, 2008. Differences in Dissolved Organic Carbon and Nitrogen Responses to Storm-Event and Ground-Water Conditions in a Forested, Glaciated Watershed in Western
New York. Journal of the American Water Resources Association (JAWRA) 44(6):1-16. DOI: 10.1111 ⁄ j.17521688.2008.00251.x
(DOC) and nitrogen (DON) from watersheds. DOC
plays an important role in the acid-base chemistry of
acid sensitive freshwater systems (Herczeg et al.,
1985; Driscoll et al., 1988); affects the complexation,
solubility, and mobility of metals such as aluminum
and mercury (Driscoll et al., 1988); influences the
INTRODUCTION
Recently there has been considerable interest in
understanding the sources, flow paths, and factors
influencing the exports of dissolved organic carbon
1
Paper No. JAWRA-07-0130-P of the Journal of the American Water Resources Association (JAWRA). Received September 14, 2007;
accepted April 10, 2008. ª 2008 American Water Resources Association. Discussions are open until June 1, 2009.
2
Respectively, Assistant Professor, Bioresources Engineering, 264 Townsend Hall, University of Delaware, Newark, Delaware 19716;
Former graduate student, Great Lakes Center, Buffalo State College, Buffalo, New York 14222; and Professor, Environment and Forest Biology, SUNY-ESF, Syracuse, New York 13210 (E-Mail ⁄ Inamdar: [email protected]).
JOURNAL
OF THE
AMERICAN WATER RESOURCES ASSOCIATION
1
JAWRA
INAMDAR, RUPP,
MITCHELL
pronounced than DOC values for events following a
dry period. Goller et al. (2006) attributed the higher
DON concentrations to preferential release of hydrophilic DOM which is richer in N (Qualls and Haines,
1991). Bernal et al. (2005) reported lower DOC:DON
ratios in streamflow for wet conditions and storm
events as opposed to dry conditions and base flow in
a forested Mediterranean catchment in Spain. These
differences in DOC and DON relationships clearly
indicate the need for additional studies, especially
across a broad range of storm-event conditions.
The rise in ground-water elevations into DOC-rich
surficial soil horizons and the subsequent flushing of
DOC have been hypothesized as an important mechanism for export of DOC in catchment streamflow
(Hornberger et al., 1994; Boyer et al., 1997; Creed
and Band, 1998). While a few studies have provided
support for this hypothesis by finding a match in the
temporal pattern of ground-water elevations and
DOC concentrations in streamflow (e.g., Inamdar and
Mitchell, 2006), we know of no study that has explicitly evaluated the relationship between changes in
ground-water elevations and the DOC concentrations.
Similarly, no study has explored the relationship
between ground-water elevations and DON. Do DON
concentrations follow the same flushing pattern as
that hypothesized for DOC? If yes, is the DONground-water relationship as strong as that for DOC
and ground water? These are important questions
and need to be addressed, especially, if we are to
identify the differences in DOC and DON dynamics.
In previous studies, we (Inamdar and Mitchell,
2006, 2007a) identified the sources and flow paths of
DOC and DON for the Point Peter Brook watershed
(PPBW), a glaciated, forested watershed in western
New York. Investigations were performed across multiple catchments and storm events over a three-year
period. These studies indicated important differences
in the responses of DOC and DON. Concentrations of
DON in streamflow during storm events consistently
increased and peaked before the corresponding DOC
values. The peak in DOC concentrations typically
occurred at or after the discharge peak and on the
recession limb of the hydrograph. Furthermore, the
rise and fall of DON concentrations were much
quicker than the corresponding DOC values. We
attributed the differences in storm-event patterns to
different sources of DOC and DON in the catchment
– DON being derived from throughfall and forest
floor while DOC being derived from throughfall,
forest floor as well as the soil (Inamdar and Mitchell,
2007a). However, these studies did not investigate
how storm-event conditions (e.g., rainfall amount
and antecedent moisture conditions) and groundwater elevations influenced the concentrations and
fluxes of DOC and DON in catchment streamflow.
adsorption of pesticides in soils (Worrall et al., 1997);
and has been linked to the formation of potentially
carcinogenic trihalomethanes when surface water is
chlorinated for drinking (Siddiqui et al., 1997; Nokes
et al., 1999). DON constitutes a significant portion of
the total nitrogen (N) flux for some ecosystems
(Hedin et al., 1995; Campbell et al., 2000; Perakis
and Hedin, 2002; Neff et al., 2003; Pellerin et al.,
2006) and a large portion of DON can become bioavailable for estuarine plankton (Seitzinger et al.,
2002). While DOC and DON are intrinsically linked
as components of dissolved organic matter (DOM)
recent studies suggest that there may be important
differences in the responses of DOC and DON to
varying hydrologic conditions (Kalbitz et al., 2000;
Prechtel et al., 2000; Michalzik et al., 2001; Aitkenhead-Peterson et al., 2003; Willett et al., 2004; Inamdar and Mitchell, 2007a). These differences may
especially be apparent during storm events when
these solutes are mobilized and the exports are at
their maximum (Hagedorn et al., 2000, 2001; Buffam
et al., 2001; Inamdar et al., 2006; Turgeon and Courchesne, 2008).
Elevated exports of DOC and DON during storm
events have been attributed to throughfall input, litter leachate, ‘‘flushing’’ of organic and DOM-rich soil
horizons by near-surface runoff, and release of
instream and hyporheic sources of the solutes with
elevated streamflow (Boyer et al., 1997; Creed and
Band, 1998; Kaushal and Lewis, 2003, 2005; Goller
et al., 2006; Inamdar and Mitchell, 2006, 2007a).
Evaluating the sources and mechanisms responsible
for the exports of DOC and DON has typically been
done using a small number of storm events (Hagedorn et al., 2000; Buffam et al., 2001; Inamdar et al.,
2004). While many studies have investigated the role
of precipitation, water fluxes, and moisture at the
laboratory, plot, and field scales for DOC (see Kalbitz
et al., 2000 for an exhaustive review), fewer studies
have simultaneously compared the controls of stormevent conditions on DOC and DON at the catchment
scale (Bernal et al., 2005; Goller et al., 2006; Cooper
et al., 2007; Turgeon and Courchesne, 2008). Cooper
et al. (2007) working in an 80 ha glaciated, forested
watershed in Scotland evaluated the influence of four
different factors on DOC and DON exports and found
that event magnitude (i.e., rainfall amount) combined
with the flow through the Bs horizon had the greatest
influence on the exports. Turgeon and Courchesne’s
(2008) research in a 5 ha Canadian Shield catchment
in Quebec, Canada revealed considerably higher DOC
concentrations on rewetting after dry periods while
the response for DON was muted. In contrast, Goller
et al. (2006) working in a montane forested
watershed in Ecuador found that the increase in
DON concentrations in litter leachate was more
JAWRA
AND
2
JOURNAL
OF THE
AMERICAN WATER RESOURCES ASSOCIATION
DIFFERENCES
IN
DOC
AND
DON RESPONSES
TO
STORM-EVENT
AND
GROUND-WATER CONDITIONS
30¢ W). Mean annual winter temperature is )3C
and the mean summer temperature is 21C. Annual
precipitation averages 1,006 mm of which 200250 mm occurs as snow (20 year average based on
the National Atmospheric Deposition Program
Weather Station at Chautauqua, New York; 35 km
southwest of the PPBW; NADP, 2006). Bedrock geology in the region consists of stratified limestone, dolomite with gypsum and shale of the Upper Devonian
period (Olcott, 1995). The parent material was
derived from glacial till (Kent Drift of Woodfordian
formed 19,000 years B.P.) (Phillips, 1988). Vegetation
on ridgetops and hillslopes is dominated by deciduous
trees including sugar maple (Acer saccharum), black
maple (Acer nigrum), American beech (Fagus grandiflora), yellow birch (Betula alleghaniensis) with larger
proportions of conifers including hemlock (Tsuga
canadensis), and white pine (Pinus strobus) in valleybottoms. Topography of the entire watershed is fairly
distinct with wide ridgetops, steep hillslopes, and
narrow valley bottoms. Slope gradients in the
watershed range from 0 to 69%, with a mean gradient of 14%. Elevation ranges from 254 to 430 m above
mean sea level. A low-permeability clay layer, that is
a part of the till layer, generates perched water
tables and forces water to move as shallow subsurface
flow on the steep hillslopes. The depth to the clay ⁄ till
measured using soil cores varies from 1.2 to 1.7 m in
the valley-bottom locations, 0.3-0.5 m along the side
slopes, and 0.6 m at the ridgetops.
Furthermore, the relationship of ground water with
DOC vis-à-vis DON was also not evaluated. Here, we
build on our prior knowledge of the mechanisms of
DOC and DON in the PPBW to address specific questions that include the following:
1. What is the influence of storm rainfall amounts,
intensity, and antecedent moisture conditions on
DOC and DON concentrations and fluxes? Do
the influences differ for DOC and DON?
2. What is the relationship between ground-water
elevations and DOC and DON concentrations
and does the relationship differ for DOC and
DON?
SITE DESCRIPTION AND METHODS
Site Description
This study was conducted in the PPBW (Figure 1),
located in Cattaraugus County and 55 km southeast
of Buffalo in New York State (42 26¢ 30¢ N; )78 55¢
Watershed Instrumentation and Sampling
Sampling was performed for catchments S1
(696 ha), S2 (3.4 ha), and S3 (1.9 ha) (Figure 1).
Catchment S3 is nested within S2, and S2 within S1.
A detailed description of the catchments and instrumentation and sampling plan is provided in Inamdar
and Mitchell (2006, 2007a). Briefly, precipitation was
recorded using a tipping-bucket rain gage while
stream stage was recorded using a pressure transducer. At S1, a stage-discharge relationship was
developed for the stream channel while Parshall
flumes were installed on streams at S2 and S3 (all
perennial streams). Ground-water elevations were
recorded using pressure transducers nested within
logging wells that were constructed of 5 cm (ID) PVC
tubing. Two logging wells R1 and R2 were located in
the riparian area of catchment S2 (Figure 1). One
hillslope well (H2) was positioned in a saturated area
on a hillslope bench. Multiple other wells were also
located in other parts of the catchment including a
valley-bottom ground-water logging well (H7) and an
adjoining ground-water sampling well (HS7) (Figure 1). Ground-water logging and sampling wells
FIGURE 1. Location of Point Peter Brook Watershed (PPBW) in
New York State (A); Sampled Catchments S1, S2, and S3 (B); and
Sampling Locations and Devices Within the Catchments (C).
JOURNAL
OF THE
AMERICAN WATER RESOURCES ASSOCIATION
3
JAWRA
INAMDAR, RUPP,
MITCHELL
discharge peak except for the event of August 30,
2005 which was a long event (intermittent rainfall
for 18 h) associated with the remnants of Hurricane
Katrina (NOAA, 2007). It is important to note here
that not all catchments were sampled simultaneously
(stream chemistry) and across all the selected events.
Antecedent moisture conditions for each storm
were computed using multiple indices: (1) summation
of the precipitation amounts for seven days prior to
the event (antecedent precipitation index – API7), (2)
summation of streamflow at outlet of S1 for 24 h
prior to the event, and (3) average of ground-water
depths from the soil surface (antecedent groundwater index – AGI7) for seven days prior to the event
for hillslope-bench well H2 and riparian well R2.
Ground-water depths at the hillslope benches were
much more dynamic than those at valley-bottom
riparian positions in PPB (Inamdar and Mitchell,
2007a). Inamdar and Mitchell (2007b) also found that
ground-water depths were a more reliable indicator
of antecedent moisture conditions in the watershed,
including the watershed response to storms, compared to the seven-day antecedent precipitation
(API7) values. One factor that likely influenced the
relationship between the API7 values was related to
the variable temporal influences of evapotranspiration. Such influences would most likely be most
evident during periods of high (late summer) or
extremely low (early spring) evapotranspiration.
Therefore, we used the 24-h antecedent streamflow
for S1 and antecedent moisture indices for ground
water (AGI7 for H2 and R2) associated with the
selected events to classify the antecedent periods into
‘‘dry,’’ ‘‘moderately wet,’’ and ‘‘wet’’ moisture conditions. In addition to antecedent moisture conditions,
rainfall amount, intensity, catchment runoff ratio,
and the change in ground-water depths at H2 and R2
for the events were also used in our analyses. The
change in ground-water depth was the difference
between the depth prior to the event and the depth
at peak saturation during the event.
To investigate if the DOC and DON concentrations
in stream discharge differed significantly among the
antecedent moisture groups (dry, moderately wet,
and wet), the event concentrations for each group
were compared using the Mann-Whitney test. In
addition, DOC and DON responses for individual
events were evaluated including – event fluxes, flowweighted concentrations (FWC), percent increases in
concentration (%V), increases in concentrations per
unit discharge (dV ⁄ dQ), and rates of change of DOC
and DON concentrations on the rising limb. Event
flux was computed by the summation of the product
of discharge and event concentrations. FWC was the
ratio of the event flux and the event discharge. Percent increase in concentration was the percent
were screened for the full length from 30 cm below
the soil surface.
Water chemistry was monitored by grab sampling
and automatic sampling using ISCO samplers. Grab
sampling was performed twice a month for: valleybottom and hillslope ground-water wells, surface
seeps, and lysimeters located in valley-bottom and
hillslope-bench saturated areas. Storm event sampling of streamflow for the catchments was performed
using a limited number of ISCO samplers over 20032005 and thus not all catchments were sampled
simultaneously for all events. In the summer of 2004,
an ISCO sampler was also installed for ground-water
well HS7 and ground-water samples were collected
for selected storm events. The ISCO was programmed
to collect samples on the rising and falling limbs of
the ground-water hydrograph. All samples were collected within 24 h of an event in 250 ml Nalgene bottles, and samples were filtered with 0.5 lm filters
prior to analysis.
Analyses performed on the samples included: DOC
on a Tekmar-Dohrmann Phoenix 8000 TOC analyzer
and anions (NO
3 ) on a Dionex IC. Total dissolved N
(TDN) was determined using persulfate oxidation
(Ameel et al., 1993) followed by colorimetric analysis
on an autoanalyzer. Ammonia was also determined
on an autoanalyzer using the Berthelot Reaction followed by colorimetric analysis. Concentrations of
DON were estimated by subtracting NHþ
4 and NO3
þ
from TDN. If the sum of NH4 and NO3 equaled or
exceeded the TDN value, DON concentration was
recorded as zero (McHale et al., 2000). The overall
estimate of analytical uncertainty for the DON analyses was ±5-10%. The laboratory is a participant in
the U.S. Geological Survey (USGS) performance evaluation program to ensure data quality. A system of
calibration QC, detection QC, analytical blanks, and
replicates is used with every set of samples (Mitchell
et al., 2001).
Selection of Storm Events and Hydrologic and Solute
Response Parameters for the Events
Multiple storm events were monitored for catchments S1, S2, and S3 from May 2003 through August
2005. Events that are evaluated in this study include:
June 8, July 27, and August 9, 2003 (summer
events); September 1 and 22, October 14, and November 19, 2003 (fall events); April 12, May 9, 20, 22b
(second event on that date), and May 27, 2004 (spring
events); and August 30, 2005 (summer event). These
events were selected as they were representative of
the events for those seasons and had the most
complete dataset across the catchments. Most of
the selected events produced a unimodal (single)
JAWRA
AND
4
JOURNAL
OF THE
AMERICAN WATER RESOURCES ASSOCIATION
DIFFERENCES
IN
DOC
AND
DON RESPONSES
TO
STORM-EVENT
RESULTS
Hydrologic Conditions for the Selected Storm Events
Of the selected events, events of September 22,
2003, and August 30, 2005, had the driest antecedent
moisture conditions while the events of May 27, May
22b, April 12, and April 19, 2004, had the wettest
(Table 1). Moisture conditions in the PPBW were typically at their wettest during the spring months
(March-May) and driest in late summer (August).
The shallow ground-water depth values associated
with well H2 (AGI7 in Table 1) during the spring
events indicate surface saturation on hillslope
benches (e.g., May 27, April 12 and 19). Events also
differed considerably in the amounts and intensities
of rainfall input. The largest event (66 mm) of August
30, 2005, was the remnant of Hurricane Katrina that
also occurred when the watershed was experiencing
the lowest antecedent moisture conditions during the
three-year study (2003-2005). However, the rainfall
input for this event was sufficiently large to cause
surface-saturation at hillslope benches and valleybottom riparian areas (e.g., compare AGI7 for H2 and
R2 and the resultant decrease in ground-water
depths in Table 1 associated with this event). The
second-largest rainfall event (32 mm, October 14,
2003) also occurred following a dry period; however,
the rainfall input was not sufficient to saturate valley-bottom riparian locations (as indicated by depth
of ground water in riparian well R1, Table 1). Events
during the spring (April, May) were moderate to
small in size (<17 mm) but typically yielded greater
catchment runoff ratios because of wetter antecedent
soil moisture conditions. The most intense rainfall
event occurred on July 27, 2003 under moderately
wet soil moisture conditions.
Relationship Between Ground-Water Depths and
DOC and DON Concentrations in Streamflow and
Ground Water
Our intent here was to determine if DOC and
DON displayed similar flushing responses in the
PPBW. For this, we compared the ground-water
depths in riparian wells R1 and R2 (Figure 1, within
catchment S2) against the concentrations of DOC and
DON in streamflow at S2. The correlations between
the two parameters were quantified using Spearman
rank correlation (SPSS Inc.). Wells R1 and R2 were
selected as they were representative of ground-water
elevations in the valley-bottom riparian area in catchment S2 and thus were well suited to characterize
the flushing response of DOM for catchment S2. Well
R2 was located 4.2 m from S2, while well R1 was
located at a distance of 24.1 m. In addition to correlations, we also investigated if there was any relationship between percent change in ground-water depth
in the wells and corresponding percent changes in
concentrations of DOC and DON recorded in streamflow at S2. The strength of the relationship between
percent change in DOC and ground-water depth was
quantified using the Pearson statistic. Ideally, it
would also have been nice to compare the groundwater depths in R1 and R2 against concentrations of
DOC and DON in the ground water at R1 and R2
itself. That, we assume, would have been a stronger
test of DOC and DON flushing. However, while we
did not have this data for R1 and R2, we did have
ground-water concentrations of DOC and DON for
well HS7 and corresponding ground-water elevations
OF THE
AMERICAN WATER RESOURCES ASSOCIATION
GROUND-WATER CONDITIONS
for well H7 for seven storm events in 2004. These
events included – June 14; July 7, 15, 15, 18, 19; and
September 16 (all 2004). For these seven events, we
determined the correlations (Spearman rank)
between ground-water depths recorded at logging
well H7 and the corresponding concentrations of
DOC and DON at sampling well HS7. Well HS7 was
about 3 m from well H7.
change between concentration at start of event and
the peak concentration; and the dV ⁄ dQ was the ratio
of increase in concentration (peak minus the start
value) and the increase in discharge (peak discharge
minus the discharge at the start of the event). The
rate of change of DOC and DON concentrations on
the rising limb was the ratio of the increase in the
solute concentration (difference in peak and starting
concentration) and the time required for the increase.
Furthermore, to allow direct comparison between the
rates of change of DOC and DON, the rate parameter
(dV ⁄ dt) was normalized by the corresponding peak
solute concentration. Events that did not reveal a
clear increase in DOC and DON concentrations on
the hydrograph rising limb were not included for this
analysis.
Correlations between flow-weighted DOC and DON
concentrations and event parameters (total rain, rainfall intensity, antecedent moisture indices, total discharge, and peak discharge) were evaluated using the
Spearman rank tests (SPSS Inc., Chicago, Illinois).
JOURNAL
AND
DOC and DON Responses for the Storm Events
DOC concentrations were significantly greater for
events following drier antecedent periods (Table 2).
In comparison, concentrations for DON did not differ
5
JAWRA
INAMDAR, RUPP,
AND
MITCHELL
TABLE 1. Hydrologic Conditions Associated With the Selected Storm Events.
Antecedent Wetness
Indices
Rainfall
Events
Total
(mm)
Duration
(hrs)
Increase in
GroundWater
Depth (m)
Catchment Runoff
Ratio*
Peak 10-min
Intensity (mm)
API7
mm
24 h S1
Streamflow
(mm)
AGI7
R2 (m)
AGI7
H2 (m)
Well
H2
Well
R2
S1
696 ha
S2
3.4 ha
S3
1.6 ha
0.7
6.1
2.3
2.5
17.5
2.8
0.0
1.8
0.40
0.31
0.44
0.18
0.49
0.50
0.50
0.35
0.61
0.68
0.30
1.34
0.36
0.54
0.41
1.3
0.08
0.17
0.04
0.34
0.19
0.10
0.19
0.13
0.17
na
na
0.35
na
0.13
0.15
na
4.3
8.4
4.1
6.9
12.7
47.5
31.5
23.1
0.91
1.06
0.62
0.94
na
0.42
0.44
0.32
0.03
0.12
0.02
0.12
0.17
0.21
0.10
0.22
na
0.29
0.17
0.12
0.23
0.27
0.16
0.10
0.25
0.29
0.27
0.18
na
na
na
0.28
1.0
2.3
3.3
5.6
2.0
3.8
27.4
0.2
39.6
25.4
33.8
83.0
1.36
1.42
1.71
1.78
2.72
1.99
0.39
0.34
0.29
0.31
0.30
0.23
0.04
0.03
)0.02
0.10
0.08
)0.04
0.17
0.15
0.05
0.12
0.05
0.07
0.17
0.08
0.06
0.14
0.08
0.02
0.61
0.19
0.45
0.35
na
0.26
na
0.37
0.26
0.40
0.27
0.32
0.34
0.49
0.27
0.36
0.39
0.33
Dry antecedent conditions
Sep 1, 2003
15.7
9.2
Sep 22, 2003
16.5
5.3
Oct 14, 2003
32.0
14.1
Aug 30, 2005
66.0
18.2
Moderately wet antecedent conditions
Jun 8, 2003
12.1
2.7
Jul 27, 2003
24.1
4.7
Aug 9, 2003
11.2
1.5
May 20, 2004
11.2
0.5
Wet antecedent conditions
Nov 19, 2003
12.7
6.7
Apr 12, 2004
16.0
4.6
Apr 19, 2004
4.3
0.2
May 9, 2004
16.0
4.0
May 22b, 2004
5.6
0.3
May 27, 2004
10.7
1.5
Notes: na, data not available.
*Ratio of event discharge to event rainfall.
TABLE 2. Mean DOC and DON Concentrations and Mann-Whitney Test Results for the Three
Groups of Antecedent Moisture Conditions – Dry, Moderately Wet (Mod), and Wet Conditions.
Mean DOC Conc. (lmol C ⁄ l)
Catchment
S1
S2
S3
Mean DON Conc. (lmol N ⁄ l)
Dry
Mod
Wet
Dry
Mod
Wet
479 (33) a [50]
341 (39) a [30]
337 (32) a [18]
426 (26) a [34]
317 (44) a [41]
207 (22) b [22]
304 (31) b [30]
209 (33) b [39]
166 (31) c [42]
29 (98) d [50]
15 (83) d [30]
9 (99) d [18]
22 (95) d [34]
16 (97) d [41]
16 (100) d [22]
22 (88) d [30]
30 (110) d [39]
13 (146) d [40]
Note: The coefficient of variation is provided within curved brackets. The number of data points within each group is provided within square
brackets. For a given catchment, different letters among the groups indicate significant differences while groups having the same letters indicate nonsignificant differences.
intense event of July 27, 2003 (47 lmol N ⁄ l). Meanwhile, the event of April 12, 2003, which had one of
the lowest FWC of DOC for S1, produced the thirdhighest DON concentration (39 lmol N ⁄ l). A comparison of the DOC:DON ratios for S1 show that except
for the event of August 30, the events with dry antecedent moisture conditions had higher ratios than
events with wetter antecedent conditions.
Overall, a similar pattern in FWC was observed
for the smaller catchments S2 and S3 (Tables 4 and
5). For catchment S2, highest DOC concentrations
were recorded for the events of July 27 and August
30, while the lowest values were observed for the wet
events of April 12 and May 27, 2004. In contrast, the
highest DON concentration was recorded for the
significantly among any of the antecedent moisture
groups (Table 2). When FWC for individual events
are compared, highest flow-weighted DOC concentrations from catchment S1 were recorded for the dry
events of October 14, 2003, and August 30, 2005,
(both 584 lmol C ⁄ l) (Table 3). The next highest DOC
concentration (540 lmol C ⁄ l) for catchment S1 was
observed for the intense event of July 27, 2003. In
contrast, the wet events of April 12 and 19 resulted
in the lowest flow-weighted DOC concentrations
(Table 3) for S1. DON concentrations, on the other
hand, did not follow the same pattern as that for
DOC. The highest flow-weighted DON concentration
(71 lmol N ⁄ l) from catchment S1 was recorded for
the largest event of August 30 followed by the most
JAWRA
6
JOURNAL
OF THE
AMERICAN WATER RESOURCES ASSOCIATION
DIFFERENCES
IN
DOC
AND
DON RESPONSES
TO
STORM-EVENT
AND
GROUND-WATER CONDITIONS
TABLE 3. DOC and DON Responses for Catchment S1 (696 ha).
Dry Antecedent
Conditions
Parameters
DOC
Flux
FWC
%V
dV ⁄ dQ
dV ⁄ dt
DON
Flux
FWC
%V
dV ⁄ dQ
dV ⁄ dt
DOC:DON
Moderately Wet
Antecedent Conditions
Wet Antecedent
Conditions
Sep 1,
2003
Sep 22,
2003
Oct 14,
2003
Aug 30,
2005
Jun 8,
2003
Jul 27,
2003
Aug 9,
2003
Nov 19,
2003
Apr 12,
2004
Apr 19,
2004
14
489
113
1680
1.1
7.5
450
162
1075
2.1
35
584
222
924
1.2
50
584
201
394
1.4
10
353
79
991
1.2
35
540
131
254
2.5
8.2
450
99
696
3.4
30
392
167
720
2.2
9.1
297
112
662
2.0
5.6
289
68
830
1.5
0.3
12
272
72
1.8
41
0.2
9
193
57
15
50
0.7
12
2323
161
1.7
49
6.1
71
1212
90
3.0
8
0.3
11
348
68
8.4
32
3.1
47
1826
57
1.2
11
0.6
31
1904
130
44
15
1.5
19
408
78
4.1
21
1.2
39
906
245
21
8
0.2
10
1783
72
5.0
29
Note: Responses include: DOC and DON flux (mol ⁄ ha), flow-weighted concentration (FWC; lmol ⁄ l), % increase in concentration (%V), ratio of
change in concentration to change in discharge (dV ⁄ dQ; lmol ⁄ L ⁄ mm), rate of change of concentrations (dV ⁄ dt) on the rising limb of hydrograph and the ratio of flow-weighted DOC and DON concentrations (DOC:DON) during storm events.
TABLE 4. DOC and DON Responses for Catchment S2 (3.4 ha).
Dry Antecedent
Conditions
Parameters
DOC
Flux
FWC
%V
dV ⁄ dQ
dV ⁄ dt
DON
Flux
FWC
%V
dV ⁄ dQ
dV ⁄ dt
DOC:DON
Moderately Wet
Antecedent Conditions
Wet Antecedent
Conditions
Sep 1,
2003
Aug 30,
2005
Jun 8,
2003
Jul 27,
2003
Aug 9,
2003
May 20,
2004
Apr 12,
2004
7.1
264
74
189
3.4
102
449
137
209
1.8
6.3
205
46
170
15
36
517
246
146
9.6
7.6
250
104
244
10
5.8
293
83
60
24
9.6
163
43
59
9.5
0.1
5
264
44
42
46
5.7
25
298
25
18
18
0.2
5
35
5
12
41
1.2
18
634
10
22
29
0.2
8
246
17
42
31
0.8
40
116
14
65
7
0.4
6
940
21
55
27
May 9,
2004
May 22b,
2004
May 27,
2004
21
336
201
79
29
3.3
215
52
58
32
5.7
166
31
40
12
1.0
16
132
4
138
21
0.3
17
761
25
102
13
1.8
52
142
67
na
3
Note: Responses include: DOC and DON flux (mol ⁄ ha), flow-weighted concentration (FWC; lmol ⁄ l), % increase in concentration (%V), ratio of
change in concentration to change in discharge (dV ⁄ dQ; lmol ⁄ l ⁄ mm), rate of change of concentrations (dV ⁄ dt) on the rising limb of hydrograph and the ratio of flow-weighted DOC and DON concentrations (DOC:DON) during storm events.
for another dry event – September 22 was not very
different than ratios observed for the wetter events.
The differences in DOC and DON concentration
relationships among the dry and wet events are also
apparent from the concentration-discharge (C-Q)
plots for the catchments (Figures 2-4). The dry events
produced higher DOC concentrations while high
DON values were associated with events following
wet antecedent moisture conditions. The largest
event of August 30 which started out with the driest
antecedent moisture conditions yielded both high
event of May 27. However, a similarly high value was
not recorded for the event of April 12. The highest
DOC:DON ratio for S2 (46) was associated with the
dry event of September 1, while the lowest ratio (3)
was observed for the May 27 event. Similarly for S3,
the maximum flow-weighted DOC concentration (484)
was associated with the dry event of September 22
while the highest flow-weighted DON concentration
occurred for the wet event of May 27. Although the
highest DOC:DON ratio for S3 (65) was observed for
the dry event of October 14, the corresponding value
JOURNAL
OF THE
AMERICAN WATER RESOURCES ASSOCIATION
7
JAWRA
INAMDAR, RUPP,
AND
MITCHELL
TABLE 5. DOC and DON Responses for Catchment S3 (1.6 ha).
Dry Antecedent
Conditions
Parameters
DOC
Flux
FWC
%V
dV ⁄ dQ
dV ⁄ dt
DON
Flux
FWC
%V
dV ⁄ dQ
dV ⁄ dt
DOC:DON
Sep 22, 2003
Oct 14, 2003
Moderately Wet
Antecedent Conditions
Wet Antecedent
Conditions
Nov 19, 2003
May 20, 2004
Apr 12, 2004
May 9, 2004
May 22b, 2004
May 27, 2004
9.1
424
133
364
20
13
259
15
52
11
10
234
61
210
4.0
5.4
172
101
146
45
10
137
51
65
4.3
13
226
175
94
29
3.3
156
36
37
47
5.6
162
32
91
20
0.3
15
700
33
80
28
0.2
4
325
16
78
65
0.9
21
597
96
19
11
0.4
13
299
35
136
13
0.2
3
1357
18
27
46
0.6
10
1087
9
13
23
0.2
8
3921
23
199
19
1.2
36
161
41
110
4
Note: Responses include: DOC and DON flux (mol ⁄ ha), flow-weighted concentration (FWC; lmol ⁄ l), % increase in concentration (%V), ratio of
change in concentration to change in discharge (dV ⁄ dQ; lmol ⁄ l ⁄ mm), rate of change of concentrations (dV ⁄ dt) on the rising limb of hydrograph and the ratio of flow-weighted DOC and DON concentrations (DOC:DON) during storm events.
FIGURE 2. Streamflow Discharge (mm ⁄ h) vs. Concentrations of DOC (lmol C ⁄ l) and DON (lmol N ⁄ l), and
DOC:DON Ratios for Selected Events for Catchment S1. Events with dry antecedent moisture conditions (AMC),
September 1 and 22, October 14, August 30; moderately wet AMC, July 27; and wet AMC, April 12.
FIGURE 3. Streamflow Discharge (mm ⁄ h) vs. Concentrations of DOC (lmol C ⁄ l) and DON (lmol N ⁄ l), and
DOC:DON Ratios for Selected Events for Catchment S2. Events with dry antecedent moisture conditions (AMC),
August 30; moderately wet AMC, July 27, May 20; and wet AMC, April 12, May 22b, May 27.
JAWRA
8
JOURNAL
OF THE
AMERICAN WATER RESOURCES ASSOCIATION
DIFFERENCES
IN
DOC
AND
DON RESPONSES
TO
STORM-EVENT
AND
GROUND-WATER CONDITIONS
FIGURE 4. Streamflow Discharge (mm ⁄ h) vs. Concentrations of DOC (lmol C ⁄ l) and DON (lmol N ⁄ l), and
DOC:DON Ratios for Selected Events for Catchment S3. Events with dry antecedent moisture conditions (AMC),
September 22, October 14; moderately wet AMC, May 20; and wet AMC, April 12, May 22b, May 27.
DOC as well as high DON concentrations. The C-Q
plots for DOC followed an anticlockwise hysteresis
because of the higher DOC concentrations on the
hydrograph falling limb whereas DON values showed
a clockwise hysteresis pattern. The C-Q plots also
revealed another interesting result – that for the
large events (e.g., July 27 and August 30 for S1,
August 30 for S2, and September 22 for S3) the DOCQ relationship displayed a monotonic increase
followed by a plateau in concentrations whereas a
similar pattern was seen in the DON-Q relationships.
Clearly, this suggests that for peak discharge conditions DOC exports from the catchments is supply-limited while no such limitations occur for DON.
The percent increase in the solute concentrations
(%V in Tables 3-5) during the events was more for
DON than DOC, but a systematic trend with antecedent moisture conditions was not observed (Tables 3-5).
Similarly the unit increase in solute concentrations
with discharge (dV ⁄ dQ) did not reveal a systematic
trend with antecedent moisture conditions other than
the observation that the values were high for dry
events in catchment S1 (Tables 3-5). The rates of
change of solute concentrations (dV ⁄ dt in Tables 3-5)
were again much greater for DON than DOC.
Although data have not been included here, DON
concentrations also displayed a sharper decrease than
DOC concentrations on the recession limb of the
hydrograph (note Figure 5). This indicates that DON
concentrations increased and receded much more
quickly (were more flashy) during the events than
the corresponding DOC values.
For catchment S1, flow-weighted DOC concentrations were strongly correlated with total rain, 24-h
antecedent streamflow, and AGI7H2 (Table 6). DOC
concentrations were inversely correlated with 24-h
antecedent streamflow and positively correlated with
JOURNAL
OF THE
AMERICAN WATER RESOURCES ASSOCIATION
AGI7H2 indicating that higher DOC concentrations
were associated with lower 24-h antecedent streamflow and greater ground-water depths (drier antecedent conditions). In comparison, catchments S2 and S3
showed significant correlations for 24-h antecedent
streamflow, and AGI7H2. While there was some correlation between DOC concentrations and total rain
for catchments S2 and S3, the relationships were not
significant. In contrast to DOC, no significant relationships were observed between flow-weighted DON
concentrations and total rain and antecedent wetness
indices. The only significant relationship that was
observed for DON was with total Q (catchment S1)
and peak Q (catchments S1 and S2).
Ground-Water Depths During Storm Events and DOC
and DON Concentrations
The temporal pattern of discharge and solute concentrations at S2 and ground-water depths for wells
R1 and R2 are presented in Figure 5 for selected
events of August 9, 2003, and May 20, 2004 (similar
patterns were observed for the other events). DON
concentrations increased and peaked earlier than
DOC (Figure 5). DOC concentrations peaked later in
the event and had a closer temporal match with the
riparian ground-water depths for R1 and R2. DOC
concentrations, however, dropped much more rapidly
during recession while ground-water depths receded
steadily. Among the riparian wells, the response for
well R2 which was located closer to the stream (4.2 m
from S2) was more pronounced and rapid than well
R1 (24.1 m from S2).
Not surprisingly, the correlations of DOC with
ground-water depths were much stronger than those
for DON (Table 7). For six of the seven events, DOC
9
JAWRA
INAMDAR, RUPP,
AND
MITCHELL
FIGURE 5. Temporal Patterns of Precipitation (top panel); S2 Discharge and DOC and DON Concentrations (middle panel); and
Ground-Water Depths (below soil surface) for Riparian Wells R1 and R2 (bottom panel) for Events of August 9, 2003, and May 20, 2004.
TABLE 6. Spearman Rank Correlations Between Flow-Weighted Event Concentrations
of DOC and DON and Selected Hydrologic Attributes for Catchments S1, S2, and S3.
Parameter
S1 [n = 10]
DOC
DON
S2 [n = 10]
DOC
DON
S3 [n = 8]
DOC
DON
Peak 10-min
Rain Intensity
API7
24-h Antecedent
S1 Discharge
0.79**
0.44 (0.19)
)0.07 (0.83)
)0.16 (0.65)
)0.25 (0.48)
0.00
)0.75* (0.01)
)0.03 (0.92)
0.76* (0.01)
0.12 (0.73)
0.46 (0.18)
0.71* (0.02)
0.63* (0.04)
0.68* (0.03)
0.56 (0.09)
)0.13 (0.71)
0.43 (0.21)
0.37 (0.29)
0.11 (0.75)
0.50 (0.13)
)0.37 (0.29)
0.28 (0.42)
0.78**
0.08 (0.82)
0.38 (0.27)
0.15 (0.67)
0.62 (0.05)
0.64* (0.04)
0.65 (0.08)
)0.35 (0.39)
0.21 (0.60)
0.18 (0.67)
)0.33 (0.42)
0.64 (0.08)
)0.76* (0.03)
0.00
0.74* (0.03)
)0.16 (0.69)
Total Rain
AGI7H2
Total Q
)0.19 (0.65)
)0.44 (0.27)
Peak Q
0.28 (0.49)
)0.43 (0.28)
Notes: p-values are provided within brackets and number of events (n) for each catchment are provided within square brackets.
*p < 0.05; **p < 0.01.
percent change in ground-water depth at R2
(R = 0.88; p < 0.01) (Figure 6). Again, in contrast to
the DOC result, no relationship was observed
between %DON change and ground-water depth at
R2 (R = 0.09; p = 0.81). The strong relationship
between DOC and ground-water depth observed for
well R2 however did not extend over for well R1
(R = 0.30; p = 0.31). The relationship between percent
change in DON and ground-water depth at R1 was
nonexistent (R = )0.31; p = 0.45).
Ground-water DOC concentrations at HS7 were
also strongly (R values ‡0.6, Table 8) correlated with
ground-water depths for five of the seven events
with significant correlation for four of those events.
Again, in contrast to DOC, no correlation was
concentrations in streamflow at S2 were significantly
and strongly (negatively) correlated with the groundwater depths recorded for riparian well R2. The negative correlation indicates that as ground water
approached the soil surface (ground-water depths
decreased), DOC concentrations increased. Similarly,
five of the 10 events showed significant negative correlations between DOC concentrations at S2 and the
ground-water depth for well R1. In striking contrast,
no significant correlations were observed between
DON concentrations and the corresponding groundwater depths for either of the two wells. In addition
to the strong correlation for DOC and ground-water
depth, the percent change in DOC concentrations at
S2 were also strongly related to corresponding
JAWRA
10
JOURNAL
OF THE
AMERICAN WATER RESOURCES ASSOCIATION
DIFFERENCES
IN
DOC
AND
DON RESPONSES
TO
STORM-EVENT
AND
GROUND-WATER CONDITIONS
TABLE 7. Spearman Rank Correlation Between Riparian Ground-Water Depths
and Streamwater DOC and DON Concentrations Exiting Catchment S2.
Correlation With Well R2
Events
n
DOC
July 27, 2003
Aug 9, 2003
Oct 14, 2003
Nov 19, 2003
Apr 12, 2004
May 9, 2004
May 20, 2004
May 22b, 2004
May 27, 2004
Aug 30, 2005
12
12
12
12
12
8
9
7
12
16
)0.98**
)0.24 (0.44)
0.30 (0.33)
)0.83**
0.08 (0.78)
)0.85**
)0.85**
)0.75* (0.05)
0.51 (0.08)
)0.80**
DON
)0.39
0.42
0.38
0.15
0.30
)0.07
0.03
)0.53
0.65
0.10
(0.21)
(0.17)
(0.21)
(0.63)
(0.31)
(0.87)
(0.93)
(0.21)
(0.02)*
(0.69)
Correlation With Well R1
DOC
DON
)0.95**
0.44 (0.15)
0.27 (0.40)
)0.81**
)0.09 (0.76)
)0.74* (0.03)
)0.83**
)0.72 (0.06)
0.59* (0.04)
)0.89**
0.25 (0.42)
0.71* (0.01)
0.37 (0.23)
0.10 (0.74)
0.41 (0.18)
0.08 (0.84)
0.04 (0.91)
)0.56 (0.19)
0.68* (0.01)
0.06 (0.80)
Notes: n, number of data points per event.
p-values are provided within brackets.
*p < 0.05; **p < 0.01.
FIGURE 6. Pearson Correlations (R) Between Percent Change in Riparian Ground-Water Depths for Wells R1 and R2 and Percent Change in
Streamflow (S2) Concentrations of DOC and DON. Wells R1 and R2 were located 24.1 and 4.2 m, respectively, from the stream outlet at S2.
observed between ground-water DON concentrations
and depths. The temporal patterns for four (July 7, 14,
and 15, and September 16) of the seven events are preJOURNAL
OF THE
AMERICAN WATER RESOURCES ASSOCIATION
sented in Figure 7. For the event of July 7 (top panel),
concentrations of DOC displayed a decreasing trend
prior to the rise in ground-water levels. However, a
11
JAWRA
INAMDAR, RUPP,
AND
MITCHELL
TABLE 8. Spearman Rank Correlation Between Storm-Event Ground-Water Depths at H7
and Corresponding Ground-Water DOC and DON Concentrations Recorded at Well HS7.
Events (all 2004)
Rainfall Amount (mm)
n
DOC
11.7
27.4
10.7, 7.3
37.0
3.8
8.9
7.9
10
10
10
12
11
12
11
)0.13 (0.72)
)0.60 (0.07)
)0.66* (0.04)
)0.05 (0.88)
)0.78** (<0.01)
)0.68* (0.01)
)0.75** (<0.01)
June 14
July 7
July 14
July 15
July 18
July 19
September 16
DON
0.18
)0.11
)0.32
0.19
)0.19
)0.39
)0.32
(0.60)
(0.75)
(0.36)
(0.54)
(0.57)
(0.19)
(0.34)
Notes: n, number of data points per event.
p-values are provided within brackets. The negative correlations indicate that as ground-water depth decreases (water approaches the
surface) the concentrations increase.
*p < 0.05; **p < 0.01.
and 4). While there were some differences in DOC
and DON responses among the catchments for individual storm events, the overall response of the
catchments to antecedent moisture conditions was
similar. We believe the differences for the individual
events likely stem from the difference in catchment
scales, topography, and the extent of saturated areas
(Inamdar and Mitchell, 2006).
While there have been numerous studies that
have investigated the influence of drying and rewetting on DOC concentrations (Kalbitz et al., 2000)
only a few have explicitly compared the DOC and
DON concentrations for these conditions. The few
studies that have compared DOC and DON have
been performed across varying scales and settings
(laboratory vs. field) and looked at DOM concentrations in different watershed compartments. Not surprisingly then, the results have been conflicting.
Small-scale and laboratory-scale studies have typically found greater increases in DON concentrations
(vs. DOC) in response to drying and rewetting episodes. Laboratory studies on forest floor of a Norway
spruce (Prechtel et al., 2000) showed dramatic
increase in DON concentrations on rewetting after
10 and 20 days of drying but a similar response was
not observed for DOC. Similarly, litter leachate concentrations measured by Goller et al. (2006) in a
tropical, montane forested watershed in Ecuador
registered a greater increase in DON concentrations
vs. DOC for events following a dry period. In general, the increase in DOM concentrations following
drying and rewetting episodes have been attributed
to (Lundquist et al., 1999): (1) disruption of soil
aggregates and release of previously sequestered
organic matter, (2) cell lysis and release of organic
matter from microbial biomass, and (3) reduced
microbial utilization during dry periods (Lundquist
et al., 1999; Miller et al., 2005). Christ and David
(1994, 1996) reported greater amounts of hydrophilic
neutrals and bases following dry soil conditions and
rise in ground-water level resulted in a sharp increase
in DOC followed by a peak prior to maximum groundwater rise. In contrast, DON concentrations did not
decrease prior to the event and also did not show an
increase as distinct as DOC. For the sequence of
events during July 14-15 (middle panel), concentrations of DOC showed a distinct peak for the first event
but a muted response for the following events. DON
concentrations also followed a pattern similar to DOC
but with much greater variability in the concentrations. Surprisingly, DOC and DON concentrations followed opposite trends for the first-half of the event of
September 16, 2004. While DOC concentrations
increased and peaked with rise in ground water, DON
values displayed a dilution pattern.
DISCUSSION
DOC and DON Responses to Storm-Event Conditions
Our results highlight some important differences
in the responses of DOC and DON to varying stormevent conditions. Concentrations of DOC were higher
for events following dry antecedent moisture conditions while DON concentrations were highest for
events following wet antecedent moisture conditions
(e.g., May 27) or large events that resulted in surface
saturation (e.g., August 30, 2005). Although the event
of August 30 was classified as an event with dry
antecedent moisture conditions, this event differed
from the other dry events in that the large rainfall
input (66 mm) resulted in surface saturation in hillslope and valley-bottom riparian positions (Table 1).
The dry conditions preceding the event likely contributed to the high DOC concentrations whereas the
wetting-up of the catchment during the event likely
increased the potential for export of DON (Table 3
JAWRA
12
JOURNAL
OF THE
AMERICAN WATER RESOURCES ASSOCIATION
DIFFERENCES
IN
DOC
AND
DON RESPONSES
TO
STORM-EVENT
suggested that this was a consequence of disruption
and turnover of the microbial biomass. Following
the observations of Christ and David (1994, 1996),
Prechtel et al. (2000) hypothesized that the greater
DON concentrations in their laboratory study were
likely due to preferential release of N-rich hydrophilic DOM associated with cell lysis. Similarly, Goller et al. (2006) also attributed the higher DON
OF THE
AMERICAN WATER RESOURCES ASSOCIATION
GROUND-WATER CONDITIONS
concentrations in litter leachate to preferential
release of N-rich hydrophilic DOM.
In contrast to small-scale and laboratory studies,
catchment-scale studies comparing concentrations of
DOC and DON in streamflow reveal opposite results.
Catchment-scale studies have found more pronounced
release of DOC following drying and rewetting episodes compared to DON. Turgeon and Courchesne
(2008) working in a 5 ha forested watershed in the
Canadian Shield, Quebec, Canada reported considerably higher DOC concentrations in streamflow on
rewetting after dry periods whereas the response for
DON was muted. Turgeon and Courchesne (2008)
attributed the higher DOC to preferential flushing of
C-rich hydrophobic organic matter under water-repellent soil conditions resulting from the dry periods.
Similarly, Bernal et al. (2005) reported lower DOC:
DON ratios in streamflow for wet conditions and
storm events as opposed to dry conditions and base
flow in a forested Mediterranean catchment in Spain.
The observations of Turgeon and Courchesne (2008)
and Bernal et al. (2005) are clearly more in agreement with our observations for PPB. The contrasting
results from the catchment- vs. small ⁄ laboratory-scale
studies suggest that the response of DOM to drying
and rewetting episodes at the catchment scale are
likely regulated by a more complex set of controls
than those influencing DOM release at the small
scale. Clearly, the expression of DOM in streamflow
is influenced by additional factors beyond the accumulation of DOM in source pools (e.g., litter and soil)
via cell lysis or disruption of soil aggregates. The
types of hydrologic flow paths, their intersection with
DOM pools, and the mobility of DOC and DON along
these flow paths are factors that likely play an important role in determining the DOM concentrations in
streamflow.
Our previous work based on EMMA (Inamdar and
Mitchell, 2007a) suggested that DON was primarily
leached from throughfall and the forest floor while
DOC contributions occurred from throughfall, forest
floor, as well as the soil profile. The strong correlation
between ground-water depths and DOC and the
simultaneous absence of such a relationship for DON
as reported in this study also suggests that DON was
derived from a surficial source. We hypothesize that
wet catchment conditions likely provided a greater
opportunity for mobilization of surficial DON via
saturation excess runoff and near-surface flow and
thus contributed to the greater expression of DON in
streamflow during these periods. This hypothesis
would also support the low DON concentrations
under dry antecedent moisture conditions as surface
runoff would be minimal, unless a large event led to
surface saturation (as in case of the August 30, 2005
event). In contrast, the soil DOC pool was likely an
FIGURE 7. Storm-Event Patterns of Ground-Water Depth
(m from surface) and Ground-Water Concentrations
of DOC (lmol C ⁄ l) and DON (lmol N ⁄ l) for Well HS7. DOC is
indicated by open circles (o) and DON is shown by crosses (+).
JOURNAL
AND
13
JAWRA
INAMDAR, RUPP,
AND
exceed the DON amounts in the catchment. The
ground-water-DOC correlations (Table 8) certainly
suggest a greater amount of DOC in the soil profile
or ground water than DON.
important contributor to elevated DOC concentrations associated with events following dry conditions.
While the sources of DON and DOC explain their
release during wet and dry events, we did not have
any data on hydrophobic or hydrophilic fractions to
evaluate the role of these DOM constituents in influencing DOC and DON concentrations. The higher
DOC concentrations following dry periods at PPB
would suggest that a greater proportion of C-rich
hydrophobic DOM was released in runoff following
dry antecedent periods while the N-rich hydrophilic
DOM was mobilized during wet storm-event conditions. The greater mobility of the hydrophilic fraction
could also explain the quick rise and fall of DON concentrations (rate of change and percent change;
Tables 3-5) as opposed to the slower response for
DOC. However, in the absence of data on DOM quality, we can only speculate about these hypotheses.
Flow-weighted event DOC concentrations were positively correlated with total rain across all three catchments but the correlation was significant only for
catchment S1 (Table 6). In comparison, DON indicated
a nonsignificant positive correlation with total rain for
catchment S1 and weak negative correlations for
catchments S2 and S3 (Table 6). Again, contrasting
observations have been reported in literature on the
influence of rainfall inputs on DOC and DON concentrations (Dosskey and Bertsch, 1997; Kalbitz et al.,
2000; Cooper et al., 2007). Cooper et al. (2007) found
that a combination of event magnitude and flux
through the Bs horizon were the most influential factors for both DOC and DON in an upland headwater
catchment in the United Kingdom. They hypothesized
that the large rain input resulted in an elevated water
flux through the soil and contributed to the high DOC
exports. In contrast, Michalzik and Matzner (1999) did
not find any correlation between DOC or DON and
water inputs for the forest floor and Guggenberger and
Zech (1994) reported no such relationship for the mineral soil. The size of the available pool of DOC or DON
vis-à-vis the volume of rainfall input may be an important factor that dictates whether concentrations
increase or decrease with rainfall input (Kalbitz et al.,
2000). A large pool of the solute may result in increase
in concentrations with rainfall input while a small pool
may be exhausted quickly and exhibit a dilution of the
solute in runoff. Catchments may have multiple pools
of DOM which get mobilized at different rainfall
amounts, flow paths, or moisture thresholds.
The differential behavior of DOC and DON to rainfall input at PPB could possibly be associated with
the pool sizes of these solutes in the catchment and
the mobilization of these pools with size of the events.
The stronger positive correlation of DOC with rainfall
input (vs. DON) suggests that the DOC pools in PPB
(throughfall, forest floor, and the soil profile) likely
JAWRA
MITCHELL
Ground-Water Depths and DOC and DON
Concentrations
Correlations of ground-water depths (wells R1 and
R2) with streamflow DOC concentrations at S2 were
much stronger and significant than the corresponding
DON values. A strong relationship was observed
between percent change in ground-water depth and
percent change in DOC for riparian well R2 located
4.2 m from S2 (Figure 6). Similarly, ground-water
DOC concentrations for well HS7 were also strongly
coupled with ground-water levels for a majority of
the seven storm events, while DON did not reveal a
similar response. While the role of ground water in
DOC flushing has long been recognized (Hornberger
et al., 1994; Boyer et al., 1997; Creed and Band,
1998) we believe this is the first study that explicitly
shows the strong correlation of ground water with
DOC and simultaneously a much weaker relationship
for DON. The strong DOC relationship supports the
flushing response for DOC with a rising water table,
whereas the weaker relationship for DON suggests
that DON is not flushed as efficiently as DOC.
The size of the DOC and DON pools in the soil, the
adsorption-desorption kinetics, and the mobility of
these solutes may be some of the factors that are likely
determining the relationship of these solutes with
ground water. Concentrations of DOC in ground water
in PPB clearly indicate that there is considerably much
more DOC in the soils as opposed to DON. Furthermore, N-rich hydrophilic fractions of DOM are more
mobile (Qualls and Haines, 1991; Ussiri and Johnson,
2004) and thus are likely to be exhausted more rapidly
than the C-rich hydrophobic fractions. In addition to
DOC, C-rich hydrophobic fractions that are preferentially adsorbed to soil surfaces (Kalbitz et al., 2000),
may continue to slowly desorb and contribute to DOC
as ground-water rise and come in contact with fresh
soil surfaces (Qualls et al., 2002). These hypotheses
however can only be confirmed by a simultaneous evaluation of DOC and DON concentrations and the hydrophobic and hydrophilic fractions of DOM.
CONCLUSIONS
This study highlighted critical differences in
responses of DOC and DON to storm event and
14
JOURNAL
OF THE
AMERICAN WATER RESOURCES ASSOCIATION
DIFFERENCES
IN
DOC
AND
DON RESPONSES
TO
STORM-EVENT
ground-water conditions in watersheds. Antecedent
moisture conditions had a greater influence on DOC
dynamics than DON. DOC concentrations were
higher for events following dry antecedent moisture
conditions while highest DON concentrations were
observed for events following wet conditions. DOC
concentrations in streamflow and ground water were
strongly coupled with the ground-water depths while
a similar pattern was not observed for DON. These
observations suggest that while DOC and DON are
intrinsically linked as components of DOM their
dynamics and exports from catchments may be regulated by different sets of mechanisms and factors.
Determining these differences and identifying the
reasons behind these responses are critical for developing more reliable and robust models for DOM fate
and transport. This study also underscored the value
of high-frequency storm-event data that allowed us to
identify the differences in DOC and DON and thus
facilitated an improved understanding of DOM.
This study was funded through a USDA NRI grant (#200200847) awarded to S. Inamdar. We are very grateful to the Gowanda Water Department for providing access to the Point Peter Brook
watershed. Joanna Tuk Riley, Christopher Riley, and Aneal Padmanabha are thanked for watershed instrumentation and data collection. Pat McHale and David Lyons provided the water chemistry
results expeditiously. The Great Lakes Center at Buffalo State College provided tuition support for Joanna and Julie.
LITERATURE CITED
Aitkenhead-Peterson, J.A., W.H. McDowell, and J.C. Neff, 2003.
Sources, Production, and Regulation of Allochthonous Dissolved
Organic Matter Inputs to Surface Waters. In: Aquatic Ecosystems: Interactivity of Dissolved Organic Matter, S. Findlay and
R. Sinsabaugh (Editors). Academic Press, Amsterdam, Netherlands, pp. 26-70.
Ameel, J.J., R.P. Axler, and C.J. Owen, 1993. Persulfate Digestion
for Determination of Total Nitrogen and Phosphorus in LowNutrient Waters. American Environmental Laboratory 10:1-111.
Bernal, S., A. Butturni, and F. Sabater, 2005. Seasonal Variation
of Dissolved Nitrogen and DOC:DON Ratios in an Intermittent
Mediterranean Stream. Biogeochemistry 75:351-372.
Boyer, E.W., G.M. Hornberger, K.E. Bencala, and D.M. McKnight,
1997. Response Characteristics of DOC Flushing in an Alpine
Catchment. Hydrological Processes 11:1635-1647.
Buffam, I., J.N. Galloway, L.K. Blum, and K.J. McGlathery, 2001.
A Stormflow ⁄ Baseflow Comparison of Dissolved Organic Matter
Concentrations and Bioavailability in an Appalachian Stream.
Biogeochemistry 53(3):269-306.
Campbell, J.L., J.W. Hornbeck, W.H. McDowell, D.C. Buso, J.B.
Shanley, and G.E. Likens, 2000. Dissolved Organic Nitrogen
Budgets for Upland Forested Ecosystems in New England. Biogeochemistry 49:123-142.
Christ, M.J. and M.B. David, 1994. Fractionation of Dissolved
Organic Carbon in Soil Water: Effects of Extraction and Storage
OF THE
AMERICAN WATER RESOURCES ASSOCIATION
GROUND-WATER CONDITIONS
Methods. Communications in Soil Science Plant Annals
25:3305-3319.
Christ, M.J. and M.B. David, 1996. Temperature and Moisture
Effects on the Production of Dissolved Organic Carbon in Spodosol. Soil Biology and Biochemistry 28:1191-1199.
Cooper, R., V. Thoss, and H. Watson, 2007. Factors Influencing the
Release of Dissolved Organic Carbon and Dissolved Forms of
Nitrogen From a Small Upland Headwater During Autumn
Runoff Events. Hydrological Processes 21(5):622-633.
Creed, I.F. and L.E. Band, 1998. Export of Nitrogen From Catchments Within a Temperate Forest: Evidence for a Unifying
Mechanism Regulated by Variable Source Area Dynamics.
Water Resources Research 34(11):3105-3120.
Dosskey, M.G. and P.M. Bertsch, 1997. Transport of Dissolved
Organic Matter Through a Sandy Forest Soil. Soil Science Society of America Journal 61:920-927.
Driscoll, C.T., R.D. Fuller, and D.M. Simone, 1988. Longitudinal
Variations in Trace Metal Concentrations in a Northern
Forested Ecosystem. Journal of Environmental Quality 17:101107.
Goller, R., W. Wilcke, K. Fleischbein, C. Valarezo, and W. Zech,
2006. Dissolved Nitrogen, Phosphorus, and Sulfur Forms in the
Ecosystem Fluxes of a Montane Forest in Ecuador. Biogeochemistry 77:57-89.
Guggenberger, G. and W. Zech, 1994. Composition and Dynamics
of Dissolved Organic Carbohydrates and Lignin-Degradation
Products in Two Coniferous Forests, N.E. Bavaria, Germany.
Soil Biology and Biochemistry 26:19-27.
Hagedorn, F., J. Bucher, and P. Schleppi, 2001. Contrasting
Dynamics of Dissolved Inorganic and Organic Nitrogen in Soil
and Surface Waters of Forested Catchments With Gleysols. Geoderma 100:173-192.
Hagedorn, F., P. Schleppi, P. Waldner, and H. Fluhler, 2000.
Export of Dissolved Organic Carbon and Nitrogen From Gleysol
Dominated Catchments – The Significance of Water Flow Paths.
Biogeochemistry 50:137-161.
Hedin, L.O., J.J. Armesto, and A.H. Johnson, 1995. Patterns of
Nutrient Loss From Unpolluted, Old-Growth Temperate Forests
– Evaluation of Biogeochemical Theory. Ecology 76:493-509.
Herczeg, A.L., W.S. Broecker, R.F. Anderson, S.L. Schiff, and D.W.
Schindler, 1985. A New Method for Monitoring Temporal
Trends in Acidity of Freshwaters. Nature 315:133.
Hornberger, G.M., K.E. Bencala, and D.M. McKnight, 1994. Hydrological Controls on Dissolved Organic Carbon During Snowmelt
in the Snake River Near Montezuma, Colorado. Biogeochemistry 25:147-165.
Inamdar, S.P., S. Christopher, and M.J. Mitchell, 2004. Export
Mechanisms for Dissolved Organic Carbon and Nitrate During
Summer Storm Events in a Glaciated Forested Catchment in
New York, USA. Hydrological Processes 18: 2651-2661.
Inamdar, S.P., N.O. Leary, M.J. Mitchell, and J.T. Riley, 2006. The
Impact of Storm Events on Solute Exports From a Glaciated
Forested Watershed in Western New York, USA. Hydrological
Processes 20:3423-3439.
Inamdar, S.P. and M.J. Mitchell, 2006. Hydrologic and Topographic
Controls on Storm-Event Exports of Dissolved Organic Carbon
(DOC) and Nitrate Across Catchment Scales. Water Resources
Research 42:W03421, doi:10.1029/2005WR004212.
Inamdar, S.P. and M.J. Mitchell, 2007a. Storm Event Exports of
Dissolved Organic Nitrogen (DON) Across Multiple Catchments
in a Glaciated Forested Watershed. Journal of Geophysical
Research, 112:G02014, doi:10.1029/2006JG000309.
Inamdar, S.P. and M.J. Mitchell, 2007b. Contributions of Riparian
and Hillslope Waters to Storm Runoff Across Multiple
Catchments and Storm Events in a Glaciated Forested
Watershed. Journal of Hydrology 341:116-130.
ACKNOWLEDGMENTS
JOURNAL
AND
15
JAWRA
INAMDAR, RUPP,
AND
Qualls, R.G. and B.L. Haines, 1991. Geochemistry of Dissolved
Organic Nutrients in Water Percolating Through a Forest Ecosystem. Soil Science Society of America Journal 55:1112-1123.
Qualls, R.G., B.L. Haines, W.T. Swank, and S.W. Tyler, 2002.
Retention of Soluble Organic Nutrients by a Forested Ecosystem. Biogeochemistry 61:135-171.
Seitzinger, S.P., R.W. Sanders, and R. Styles, 2002. Bioavailability
of DON From Natural and Anthropogenic Sources to Estuarine
Phytoplankton. Limnology and Oceanography 47:353-366.
Siddiqui, M.S., G.L. Amy, and B.D. Murphy, 1997. Ozone
Enhanced Removal of Natural Organic Matter From Drinking
Water Sources. Water Research 31:3098-3106.
Turgeon, J.M.L. and F. Courchesne, 2008. Hydrochemical Behavior
of Dissolved Nitrogen and Carbon in a Headwater Stream of the
Canadian Shield: Relevance of Antecedent Soil Moisture Conditions. Hydrological Processes 22(3):327-339.
Ussiri, D.A.N. and C.E. Johnson, 2004. Sorption of Organic Carbon
Fractions by Spodosol Mineral Horizons. Soil Science Society of
America Journal 68:253-262.
Willett, V.B., B.A. Reynolds, P.A. Stevens, S.J. Ormerod, and D.L.
Jones, 2004. Dissolved Organic Nitrogen Regulation in Freshwaters. Journal of Environmental Quality 33:201-209.
Worrall, F., A. Parker, J.E. Rae, and A.C. Johnson, 1997. A Study
of Sorption Kinetics of Isoproturon on Soil and Subsoil, the Role
of Dissolved Organic Carbon. Chemosphere 34:87-97.
Kalbitz, K., S. Solinger, J.H. Park, B. Michalzik, and E. Matzner,
2000. Controls on the Dynamics of Dissolved Organic Matter in
Soils: A Review. Soil Science 165:277-304.
Kaushal, S.S. and W.M. Lewis, Jr., 2003. Patterns in the Chemical
Fractionation of Organic Nitrogen in Rocky Mountain Streams.
Ecosystems 6:483-492.
Kaushal, S.S. and W.M., Lewis, Jr., 2005. Fate and Transport of
Organic Nitrogen in Minimally Disturbed Montane Streams of
Colorado, USA. Biogeochemistry 74:303-321.
Lundquist, E.J., L.E. Jackson, and K.M. Scow, 1999. Wet-Dry
Cycles Affect Dissolved Organic Carbon in Two California Agricultural Soils. Soil Biology and Biochemistry 31:1031-1038.
McHale, M.R., M.J. Mitchell, J.J. McDonnell, and C.P. Cirmo,
2000. Nitrogen Solutes in an Adirondack Watershed: Importance of Dissolved Organic Nitrogen. Biogeochemistry 48:165184.
Michalzik, B., K. Kalbitz, J.H. Park, S. Solinger, and E. Matzner,
2001. Fluxes and Concentrations of Dissolved Organic Carbon
and Nitrogen – A Synthesis of Temperate Forests. Biogeochemistry 52:173-205.
Michalzik, B. and E. Matzner, 1999. Fluxes and Dynamics of Dissolved Organic Nitrogen and Carbon in a Spruce (Picea Abies
Karst.) Forest Ecosystem. European Journal of Soil Science
50:579-590.
Miller, A., J.P. Schimel, T. Meixner, J.O. Sickman, and J.M.
Melack, 2005. Episodic Rewetting Enhances Carbon and
Nitrogen Release From Chapparal Soils. Soil Biology and
Biochemistry 37:2195-2204.
Mitchell, M.J., G.G. McGee, P. McHale, and K.C. Weathers, 2001.
Experimental Design and Instrumentation for Analyzing Solute
Concentrations and Fluxes for Quantifying Biogeochemical Processes in Watersheds, Methodology Paper Series of the 4th
International Conference on ILTER in East Asia and Pacific
Region. ILTER Network, Ulaanbaatar-Hatgal, Mongolia, pp. 1521.
NADP (National Atmospheric Deposition Program), 2006. National
Atmospheric Deposition Program Weather site: http://nadp.sws.
uiuc.edu/sites/siteinfo.asp?net=NTN&id=NY10, accessed December, 2006.
Neff, J.C., F.S. Chapin, and P.M. Vitousek, 2003. Breaks in Cycle:
Dissolved Organic Nitrogen in Terrestrial Ecosystems. Frontiers
in Ecology and Environment 1(14):205-211.
NOAA (National Oceanic and Atmospheric Administration), 2007.
National Oceanic and Atmospheric Administration. Hurricane
Katrina web site: http://www.katrina.noaa.gov/, accessed August
2007.
Nokes, C.J., E. Fenton, and C.J. Randall, 1999. Modeling the Formation of Brominated Trihalomethanes in Chlorinated Drinking
Waters. Water Research, 33:3557-3568.
Olcott, P.G., 1995. Groundwater Atlas of the United States: Connecticut, Maine, Massachusetts, New Hampshire, New York,
Rhode Island, Vermont. USGS Publication No. HA 730-M.
USGS, Reston, Virginia.
Pellerin, B.A., S.S. Kaushal, and W.H. McDowell, 2006. Does
Anthropogenic Nitrogen Enrichment Increase Organic Nitrogen
Concentrations in Runoff From Forested and Human-Dominated Watersheds? Ecosystems 9(5):852-864.
Perakis, S.S. and L.O. Hedin, 2002. Nitrogen Loss From Unpolluted South American Forests Mainly Via Dissolved Organic
Compounds. Nature (London) 415:416-419.
Phillips, R.A., 1988. Relationship Between Glacial Geology and
Streamwater Chemistry in an Area Receiving Acid Deposition.
Journal of Hydrology, 101:267-273.
Prechtel, A., C. Alewell, B. Michalzik, and E. Matzner, 2000. Different Effect of Drying on the Fluxes of Dissolved Organic Carbon
and Nitrogen From a Norway Spruce Forest Floor. Journal of
Plant Nutrition Soil Science 163:517-521.
JAWRA
MITCHELL
16
JOURNAL
OF THE
AMERICAN WATER RESOURCES ASSOCIATION