JEQ Journal of Natural Resources and Life Sciences Education
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Related articles in JEQ
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (30)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Royer, T. V.
Right arrow Articles by David, M. B.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Royer, T. V.
Right arrow Articles by David, M. B.
Agricola
Right arrow Articles by Royer, T. V.
Right arrow Articles by David, M. B.
Related Collections
Right arrow Biogeochemical Processes
Right arrow Watershed and Landscape Processes
Right arrow Nutrient Cycling
Right arrow Nitrogen
Published in J. Environ. Qual. 33:1296-1304 (2004).
© ASA, CSSA, SSSA
677 S. Segoe Rd., Madison, WI 53711 USA

TECHNICAL REPORTS

Landscape and Watershed Processes

Transport and Fate of Nitrate in Headwater Agricultural Streams in Illinois

Todd V. Royera,c,*, Jennifer L. Tankb and Mark B. Davida

a Department of Natural Resources and Environmental Sciences, University of Illinois, 1102 South Goodwin Avenue, Urbana, IL 61801
b Department of Biological Sciences, University of Notre Dame, Notre Dame, IN 46556
c Department of Biological Sciences, Kent State University, 256 Cunningham Hall, Kent, OH 44242

* Corresponding author (troyer{at}kent.edu).

Received for publication June 13, 2003.

   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Nitrogen inputs to the Gulf of Mexico have increased during recent decades and agricultural regions in the upper Midwest, such as those in Illinois, are a major source of N to the Mississippi River. How strongly denitrification affects the transport of nitrate (NO3–N) in Illinois streams has not been directly assessed. We used the nutrient spiraling model to assess the role of in-stream denitrification in affecting the concentration and downstream transport of NO3–N in five headwater streams in agricultural areas of east-central Illinois. Denitrification in stream sediments was measured approximately monthly from April 2001 through January 2002. Denitrification rates tended to be high (up to 15 mg N m–2 h–1), but the concentration of NO3–N in the streams was also high (>7 mg N L–1). Uptake velocities for NO3–N (uptake rate/concentration) were lower than reported for undisturbed streams, indicating that denitrification was not an efficient N sink relative to the concentration of NO3–N in the water column. Denitrification uptake lengths (the average distance NO3–N travels before being denitrified) were long and indicated that denitrification in the streambed did not affect the transport of NO3–N. Loss rates for NO3–N in the streams were <5% d–1 except during periods of low discharge and low NO3–N concentration, which occurred only in late summer and early autumn. Annually, most NO3–N in these headwater sites appeared to be exported to downstream water bodies rather than denitrified, suggesting previous estimates of N losses through in-stream denitrification may have been overestimated.

Abbreviations: DOC, dissolved organic carbon


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
THE POLLUTING OF North American coastal waters by river-borne nutrients, primarily N and P, during recent decades has resulted in coastal eutrophication and increased occurrences of harmful algal blooms, hypoxic zones, and fish kills (National Research Council, 2000; Boesch et al., 2001; Diaz, 2001). For example, the N load carried by the Mississippi River to the Gulf of Mexico has increased since the middle of the 20th century (Turner and Rabalais, 1991) and is now causing the development of seasonal hypoxia in an increasingly large area in the Gulf of Mexico (e.g., Rabalais et al., 2002). Nonpoint sources, particularly agricultural runoff, are often responsible for the increased nutrient loads carried by rivers to coastal waters (National Research Council, 2000; Boesch et al., 2001).

Streams in the agricultural regions of the upper Midwest are major contributors to N loads in the Mississippi River (Alexander et al., 2000; Goolsby et al., 1999). In the intensively agricultural areas of Illinois, watershed yields of total N are typically >15 kg N ha–1 yr–1 (Goolsby et al., 1999) and can exceed 40 kg N ha–1 yr–1 in wet years (David and Gentry, 2000). In contrast, yields from the western regions of the Mississippi River basin are typically <1 kg N ha–1 yr–1 (Goolsby et al., 1999; Alexander et al., 2000). The dissolved N pool in agricultural streams of the Midwest is dominated by NO3–N, which can often exceed 10 mg NO3–N L–1 in headwater streams (David et al., 1997; Goolsby et al., 1999).

As NO3–N is transported downstream it is subject to retention via biotic and abiotic processes. The nutrient spiraling concept (Newbold, 1992) describes the downstream movement of N as it cycles between organic N retained in biomass and dissolved inorganic N in the water column. Of the processes affecting NO3–N retention in streams, only denitrification results in a loss of N from the riverine system. Assimilatory uptake by vascular plants, algae, and microbes generally represents only short-term retention of NO3–N because the organic N is eventually remineralized. Quantifying the role of in-stream denitrification and the fate of NO3–N during stream transport will improve our understanding of the links between the agricultural areas of the Midwest and N loading to the Gulf of Mexico.

As stream size increases the likelihood of NO3–N being denitrified declines sharply (Howarth et al., 1996; Alexander et al., 2000), suggesting that headwater streams are critical locations for processing N (Peterson et al., 2001). However, this generality appears not to hold in all cases. In-stream denitrification had no substantial effect on the annual export of NO3–N from agricultural streams in Ontario (Hill, 1979) and Sweden (Jansson et al., 1994), or forested streams in the southern Appalachian Mountains (Martin et al., 2001). Conversely, mass-balance studies, although not measuring denitrification directly, often identify it as a major pathway for N loss (e.g., Howarth et al., 1996; Alexander et al., 2000; David and Gentry, 2000). For example, previous work in Illinois suggested that, state-wide, in-stream denitrification could be a substantial sink for N, potentially accounting for 132000 Mg N yr–1 (Mg = 106 g) (David and Gentry, 2000). Similarly, Alexander et al. (2000) determined that for headwater streams in the Mississippi River basin, 45.5% of the N load was retained per day of travel time, presumably through in-stream denitrification.

We examined the role of in-stream denitrification as a mechanism for N retention in five headwater, agricultural streams in east-central Illinois. We hypothesized that rates of in-stream denitrification would be high and have a significant effect on the flux of NO3–N from the streams. Our approach was to use the nutrient spiraling concept (Newbold, 1992) to assess how denitrification affected the transport of NO3–N through the streams. Nutrient spiraling metrics are commonly used to describe the transport and uptake of N in streams with low available N (e.g., Peterson et al., 2001; Hall and Tank, 2003), but an application of nutrient spiraling to examine N processing in N-rich agricultural streams has not been conducted. Our specific objectives in this paper are to (i) estimate the distance traveled by NO3–N in the streams before being denitrified; (ii) calculate the loss rates for NO3–N due to denitrification; and (iii) describe environmental factors, such as nutrient limitation and hydrology, responsible for temporal patterns in how denitrification influenced the dynamics of NO3–N in these streams.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Study Sites
The study was conducted from January 2001 through January 2002 in five headwater sites within three of the major river basins in east-central Illinois: the Sangamon, Embarras, and Kaskaskia Rivers. Each site was located in an area of extensive row-crop agriculture (Table 1). All but the fourth-order site are located in incised, channelized streams with grass as the only riparian vegetation; a description of channel alterations to the streams of east-central Illinois is given in Rhoads and Herricks (1996). The fourth-order site is located in a non-incised section of the Embarras River with a narrow corridor of hardwood forest, although all upstream tributaries are channelized and land use in the basin as a whole is dominated by agriculture. The study streams are low gradient and characterized by the flashy hydrology typical of streams in the area (David et al., 1997; Mitchell et al., 2000). However, floods rarely overtop the stream banks because of the incised channels. As is common in the Midwest, the fertilized cropland within these watersheds is underlain by extensive networks of tiles that drain excess water and associated solutes directly to headwater streams (David et al., 1997). Because of this, stream water concentrations of NO3–N in east-central Illinois routinely exceed 10 mg L–1 and can approach 20 mg NO3–N L–1 following heavy precipitation (David et al., 1997; Mitchell et al., 2000).


View this table:
[in this window]
[in a new window]
  		Table 1. Site abbreviations, coordinates, and channel and watershed characteristics for the study sites on agricultural streams in east-central Illinois.

 
Monitoring Physiochemical Variables
To further characterize the streams and relate the denitrification measurements to annual patterns in water chemistry and hydrology, we measured several environmental variables at each of the sites during the study. Water chemistry was sampled intensively throughout the study with a combination of routine grab samples (approximately weekly) supplemented by automated samplers that collected samples more frequently during floods. Water samples were analyzed for major nutrients including NO3–N, NH4–N, organic N, soluble reactive phosphorus (SRP), total P, and dissolved organic carbon (DOC) using standard methods (American Public Health Administration, 1998) and a Dionex (Sunnyvale, CA) DX-120 ion chromatograph, Technicon (Tarrytown, NY) Autoanalyzer, or Dohrmann (Mason, OH) DC-80 carbon analyzer. Organic N was determined as the difference between total N and the dissolved inorganic fractions.

Water temperature in the streams was recorded hourly with HOBO temperature loggers (Onset Computer, Bourne, MA). Discharge was monitored by either installing a stilling well and stage recorder or locating the sites in proximity to existing gaging stations operated by the USGS or the Illinois State Water Survey. Size composition of the benthic sediments was determined twice during spring 2001 at each site by collecting sediment samples from the upper 5 cm of the streambed, drying the material, and sieving it into particle classes. Four classes were used: coarse particulate organic matter (>1 mm), gravel (>2 mm), sand (53 µm–2 mm), and fine sediment (<53-µm organic and inorganic particles). The relative abundance of each class was expressed as a percentage of the dry mass of the total sample.

Denitrification Rates
Rates of benthic denitrification were measured at the sites from April 2001 through January 2002. Sediment sampling was at times prevented by floods or ice cover, and occasionally during late summer the small streams would dry up. The sampling regime varied among the sites, but no site was sampled fewer than five times. We used the C2H2 inhibition method to estimate rates of benthic denitrification in sediment slurries (Knowles, 1990; Martin et al., 2001). Chloramphenicol, an antibiotic that suppresses de novo enzyme production but does not inhibit the action of existing enzymes, was added to the slurries at a concentration of 5 mM. The use of chloramphenicol in sediment slurries reduces bottle effects and improves estimates of in situ denitrification rates (Smith and Tiedje, 1979). The assays were limited in duration to 3 h and conducted at stream temperature without the addition of NO3–N or DOC beyond that in the stream water (except for enrichment experiments described below). Based on our results and those of another study (Rudolph et al., 1991), we believe the C2H2 inhibition method was appropriate for these streams and did not bias the conclusions of the study (see Discussion).

For all assays, benthic sediments were collected from the upper 5 cm of the stream bed at several locations within a 10-m reach of the stream. The sediments were combined into a composite sample and taken immediately to the laboratory where 25 to 30 cm3 of sediment was placed in 150-mL media bottles (n = 4 per date and site). Unfiltered stream water was added to bring the total volume of the sediment slurry to 75 mL. Oxygen in the headspace and slurry was removed by purging the media bottles with ultrapure helium; bottles were shaken periodically during the purging. We conducted enrichment experiments in August 2001 at two of the sites to test for limitation of denitrification by N or DOC. Three treatments were used: ambient, +NO3–N, and +DOC. For the +NO3–N treatment, 1 mL of stock solution (0.35 mg NO3–N mL–1) was added to the media bottles so that the final concentration in the bottles was approximately 5 mg NO3–N L–1, depending on the ambient concentration. Glucose was used for the DOC amendment and the final concentration in the bottles was approximately 30 mg L–1.

During all assays, the bottles were kept in the dark in an incubator set at ambient stream temperature, and were not shaken except before sampling the headspace to equilibrate N2O in the sediment and aqueous phases. Gas samples were collected from the headspace of each media bottle at the beginning of the assay and hourly thereafter. Samples were analyzed for N2O on a Varian (Palo Alto, CA) 3600 gas chromatograph equipped with a Porapak Q column and a 63Ni electron-capture detector (oven temperature = 70°C, flow rate = 30 mL min–1). Following removal of the final gas sample, the sediment in each bottle was collected and the dry mass and ash-free dry mass (AFDM) measured by drying the sediment at 60°C, combusting the organics at 550°C, rewetting the sediment, drying at 60°C, and obtaining the final mass (difference between pre- and post-combustion mass = AFDM).

For each site and date, five quantitative sediment samples were collected, and the AFDM was measured in each to estimate the standing stock of AFDM in the upper 5 cm of the stream bed. To express the denitrification rates on an areal basis, the rates were multiplied by the mean standing stock (g AFDM m–2) at each site.

Nutrient Spiraling Metrics
To determine the importance of in-stream denitrification as a mechanism for NO3–N retention, we calculated the denitrification uptake length (Sw,dn), defined as the average distance a NO3–N molecule travels before being denitrified in the benthic sediments of a stream (Fig. 1). This term is similar to traditional uptake length (Sw) (Newbold, 1992) but differs in that it is a spiraling metric specific for denitrification. The term Sw,dn was calculated as:

[1]
where v is stream velocity (m s–1), h is stream depth (m), and Vf,dn is the uptake velocity of NO3–N due to denitrification in the benthic sediments. The term Vf,dn was calculated as:

[2]
where U is the areal denitrification rate (mg N m–2 s–1) in the sediments and C is NO3–N concentration in the water column (mg m–3). Short uptake lengths (approximately ≤1 km) indicate that NO3–N is likely to be denitrified in the stream before traveling long distances downstream. Conversely, long uptake lengths indicate that denitrification has little influence on the export of NO3–N from the watershed and that NO3–N entering headwater streams is likely to reach larger rivers and downstream water bodies.



View larger version (23K):
[in this window]
[in a new window]
  		Fig. 1. Conceptual diagram of nutrient spiraling (modified from Newbold, 1992). The stream consists of two compartments, the water column (W) and the benthic sediments (B). In this model, uptake (U) of NO3–N is due to denitrification, making the uptake length denitrification-specific.

 
To compare our assessment of the role of denitrification with that presented by Alexander et al. (2000), we also calculated the loss rate, –k, for NO3–N in the streams. Values for –k were calculated as:

[3]
and then scaled from a fraction s–1 to % d–1. This metric is an instantaneous measure of the percentage of the NO3–N load lost to denitrification per day and indicates the effect that denitrification has on the concentration and load of NO3–N in the stream water.

For the past 4 to 8 yr, intensive monitoring of NO3–N concentrations has been conducted with a combination of flow-triggered automatic samplers and routine manual sampling at three of the sites (EMC, BDO, and LFK; see Table 1). This has allowed us to measure NO3–N concentrations during short-term periods of high discharge when much of the annual export of NO3–N occurs in these agricultural watersheds (David et al., 1997; Mitchell et al., 2000). Historic values of Sw,dn were estimated as described above (Eq. [1] and [2]) using mean depth and velocity determined from the gaging records, NO3–N concentrations from the intensive monitoring, and a denitrification rate of 15 mg N m–2 h–1, the highest value measured during the study (see Results). By using a constant and high rate of denitrification, our long-term assessment describes the maximum role of denitrification as a N sink and gives conservative estimates of the distance NO3–N traveled in these streams. We recognize that stream depths and velocities estimated from gaging records represent conditions at the gaging station, but not necessarily conditions throughout a stream reach. However, we believe this approach is robust enough to allow general conclusions regarding long-term temporal patterns in the effect of denitrification on NO3–N dynamics in these streams.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Physiochemical Variables
Stream NO3–N concentrations were high (5–15 mg NO3–N L–1) for the first 6 mo of 2001 and then declined dramatically during July to <1 mg NO3–N L–1 before increasing again in October (Fig. 2, Table 2). Other forms of N had lower concentrations and a smaller range of values than did NO3–N (Table 2). This pattern was consistent among all the study streams and is typical of the annual pattern for agricultural streams in east-central Illinois (David et al., 1997; Mitchell et al., 2000). During 2001, mean values of DOC among the sites ranged from approximately 2 to 7 mg L–1 with the lowest values occurring during autumn and early winter (Table 2). Stream concentrations of total P and soluble reactive P are affected by agricultural drainage and tended to be greatest during the first 6 mo of 2001 (Table 2). During floods, soluble reactive P concentrations exceeded 1 mg P L–1 in 2001, but during baseflow mean values were generally <75 µg P L–1 (Table 2).



View larger version (39K):
[in this window]
[in a new window]
  		Fig. 2. Mean daily water temperature and stream NO3–N concentrations at selected sites from January 2001 through January 2002. All sites showed similar patterns in temperature and NO3–N, and therefore only selected sites are shown. (Site characteristics are given in Table 1.)

 

View this table:
[in this window]
[in a new window]
  		Table 2. Mean nutrient concentrations in the study sites in east-central Illinois from January 2001 through January 2002.

 
Mean daily water temperatures during 2001 were <8°C until approximately April. From April on, temperatures increased steadily, reaching peaks of 25 to 30°C in late summer (Fig. 2). Benthic sediments in the streams were predominately sand and gravel with these two classes representing 95% or more of the sediment dry mass at all sites. Fine sediments (particles <53 µm), although visually abundant at each site, represented only 1 to 3% of the dry mass. Only EMC, the site with a narrow forest corridor, had measurable amounts (4%) of coarse particulate organic matter. Based on visual observations, the composition and texture of the sediments at each site did not change noticeably during the study period.

Denitrification Rates
Rates of in-stream denitrification ranged during the study from <0.1 to 15 mg N m–2 h–1 (>1000 µmol N m–2 h–1) (Table 3). No consistent pattern was observed between stream size and rates of denitrification. Overall, denitrification rates were highly variable during the course of the study both within and among sites. There was no consistent relationship between concentrations of NO3–N in the stream water and rates of denitrification in the benthic sediments. However, concentrations dropped below 5 mg NO3–N L–1 only during July through September.


View this table:
[in this window]
[in a new window]
  		Table 3. Discharge, stream water concentration of NO3–N, and benthic denitrification rates in the study sites from April 2001 through January 2002.

 
The two sites used for the enrichment study, BLS and BDO, had ambient nitrate concentrations of 1.05 and 0.36 mg NO3–N L–1, respectively, on the days the experiments were performed. Nitrate amendments to sediments from each site significantly increased denitrification rates above the rates measured under ambient conditions in August (Fig. 3; ANOVA, P < 0.001, Tukey's pairwise comparisons, {alpha} = 0.05). Amendments with DOC did not affect denitrification rates in sediments from either site.



View larger version (21K):
[in this window]
[in a new window]
  		Fig. 3. Mean (plus one standard deviation) denitrification rates in August 2001 under ambient and amended conditions at two sites, BLS and BDO. (Site characteristics are given in Table 1.) Within a site, treatments with different letters are significantly different from each other (p < 0.05, Tukey's pairwise comparisons). AFDM, ash-free dry mass; DOC, dissolved organic carbon.

 
Nutrient Spiraling Metrics
Denitrification uptake velocity, Vf,dn, is a measure of the efficiency of denitrification in the stream bed relative to the availability of NO3–N in the water column. Values for Vf,dn were low, with all but two measurements being less than 0.04 mm min–1 (Table 4). The denitrification uptake lengths (Sw,dn) were generally long and often exceeded 200 km (Table 4). The only values less than 100 km occurred during July and September and corresponded with the highest values of Vf,dn. Overall, the long uptake lengths indicated that denitrification did not affect the movement of NO3–N through these headwater streams during the study period.


View this table:
[in this window]
[in a new window]
  		Table 4. Denitrification uptake velocity (Vf,dn), uptake length (Sw,dn), and nitrate N loss rate (–k) calculated for the study sites from April 2001 through January 2002.{dagger}

 
Using historical data on stream flow, values of Sw,dn were calculated for three sites for the past 4 to 8 yr. All sites showed a similar, cyclic pattern in which Sw,dn was longest seasonally from winter through mid-summer, corresponding to the time of highest NO3–N concentrations (Fig. 4). During this period, values of Sw,dn were routinely >200 km and occasionally exceeded 3000 km. Extremely high values of Sw,dn should not be interpreted as predictive of the actual transport distance for NO3–N because of the changes in stream morphology, hydrology, and uptake rates that occur over such distances. Rather, Sw,dn serves as an index of the importance of in-stream denitrification. In this regard, the past values of Sw,dn indicate that when NO3–N concentrations increased annually with the onset of agricultural drainage, in-stream denitrification did not influence the load of NO3–N moving through the streams. For example, when NO3–N concentrations were 5 mg NO3–N L–1 or greater, the median values of Sw,dn ranged from 280 to 532 km among the three streams and the 25th percentiles ranged from 162 to 262 km.



View larger version (36K):
[in this window]
[in a new window]
  		Fig. 4. Calculated denitrification uptake length (Sw,dn) and measured stream water concentrations of NO3–N at sites BDO, LFK, and EMC, 1993–2002. (Site characteristics are given in Table 1.) The term Sw,dn was calculated using stream gaging records and an in-stream denitrification rate of 15 mg N m–2 h–1.

 
Nitrate N loss rates, –k, represent the fraction of the NO3–N load lost to denitrification per day. The values of –k were variable spatially and temporally, ranging from <0.1 to 273% d–1 (Table 4). Across all sites, the mean value for April through June was 2.4% d–1 (standard deviation = 3.5% d–1). The greatest values of –k occurred during late summer and autumn, particularly when high denitrification rates coincided with relatively low stream NO3–N concentrations. On two occasions in July and September, –k exceeded 100% d–1, indicating that demand by in-stream denitrification could potentially remove 100% of the NO3–N load, in the absence of nitrification or other sources of NO3–N.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
CONCLUSIONS
 REFERENCES
 
Obtaining accurate rates of in-stream denitrification can be difficult and expensive, and so the role of denitrification as a N sink is often estimated by difference using a mass balance approach (e.g., Howarth et al., 1996; Alexander et al., 2000; David and Gentry, 2000). We used the commonly accepted C2H2 inhibition technique to measure denitrification in sediment slurries (Knowles, 1990). We recognize that the C2H2 inhibition technique can, in some situations, significantly underestimate denitrification rates (Seitzinger et al., 1993) and that it is particularly inappropriate when used in sediment cores with low available NO3–N or coupled nitrification–denitrification (Rudolph et al., 1991; Seitzinger et al., 1993). Conversely, Rudolph et al. (1991) found that if NO3–N was ≥10 µM, the C2H2 inhibition technique was acceptable for estimating in situ denitrification rates when used in sediment slurries, such as in our study. We also added chloramphenicol to the slurries (see above), which has been shown to improve estimates of in situ denitrification rates (Smith and Tiedje, 1979). We believe the C2H2 inhibition technique was appropriate for the streams we examined. The production of N2O in the bottles was linear for all but three assays, and given the generally high denitrification rates we measured (see below) it is unlikely that the rates we report are underestimated to any great extent.

Among studies that have expressed in-stream denitrification rates in areal units, peak values are often <5 mg N m–2 h–1 (e.g., Seitzinger, 1988; Thompson et al., 2000; Kemp and Dodds, 2002), although rates up to 60 mg N m–2 h–1 have been reported (Howarth et al., 1996). Using an innovative whole-stream technique, Laursen and Seitzinger (2002) made three estimates of denitrification in the Iroquois River system in Illinois and reported rates of 3.8, 47.6, and 118.6 mg N m–2 h–1 and suggested that rates in this range may be typical of low-gradient, agricultural streams. Likewise, denitrification rates tended to be high in the sites we examined, with nearly one-third of the measurements greater than 5 mg N m–2 h–1 and the highest value reaching 15 mg N m–2 h–1. Our rates are in the range reported by Laursen and Seitzinger (2002) and demonstrate that benthic sediments in the agricultural streams of east-central Illinois are capable of supporting high rates of denitrification for much of the year. The rates we measured were highly variable spatially and temporally, as has also occurred in similar studies (Garcia-Ruiz et al., 1998; Thompson et al., 2000; Kemp and Dodds, 2002). A large degree of spatial and/or temporal variability may be a characteristic of denitrification in agricultural streams, particularly those in which discharge and NO3–N loads are closely tied to precipitation and tile drainage.

The enrichment experiments showed that nitrate additions to sediment slurries increased denitrification rates in late summer, whereas rates did not respond to additions of labile DOC. Other workers have reported the same result for streams with low available N (Holmes et al., 1996; Martin et al., 2001). This pattern suggests that the availability of NO3–N can at times limit denitrification in stream sediments, even in streams that are nitrate-rich for much of the year. For our sites, the period of nitrate limitation appeared to be late summer and early autumn when discharge and NO3–N concentrations were low. An alternative explanation is that denitrification rates in late summer and autumn are driven by coupled nitrification–denitrification, rather than by NO3–N in the water column. Our methods do not allow us to conclude whether denitrification in late summer was limited by NO3–N concentrations in the water column or simply coupled at that time to nitrification. Regardless, late summer to early autumn is a period of low discharge and low NO3–N concentrations for streams in east-central Illinois, making the time period relatively unimportant in terms of annual N export (David et al., 1997; Mitchell et al., 2000).

Uptake velocity is a measure of demand relative to concentration and reflects the efficiency of processes within the stream at removing a nutrient from the water column (Davis and Minshall, 1999). Other studies have reported uptake velocities for NO3–N in streams that are two to three orders of magnitude greater than those we measured (Davis and Minshall, 1999; Peterson et al., 2001; Hall and Tank, 2003). This difference could have resulted from the fact that we examined only one of the processes (denitrification) that affects the uptake velocity of NO3–N, although it is assumed to be the primary mechanism of N retention in streams of the Mississippi River basin (Alexander et al., 2000). Our Vf,dn results indicate that, although the rates were generally high, denitrification in the stream beds was not an efficient N sink relative to the concentration of NO3–N in the water column. Because benthic denitrification did not affect water column concentrations of NO3–N, the distance that NO3–N was estimated to travel before being denitrified (Sw,dn) tended to be long. The fate of NO3–N in the headwater sites we studied appeared to be export to downstream water bodies, rather than denitrification.

In their assessment of N transport and retention in the Mississippi River basin, Alexander et al. (2000) used a mean annual loss rate (–k) of 45.5% d–1 for headwater streams. In the headwater streams we examined, –k was <5% d–1 except during July through September. During July through September, in-stream denitrification did have a strong influence on NO3–N movement through some of the streams (see Table 4), and this is probably the typical condition for late summer and autumn when agricultural drainage has ceased and NO3–N concentrations often drop to <0.5 mg L–1. In relation to annual load, however, high rates of NO3–N loss during late summer and autumn are misleading, because by that time of the year most of the NO3–N export has already occurred (David et al., 1997). On the basis of our results, we cannot rule out substantial in-stream retention of N at the scale of the Mississippi River basin, but in the headwater sites we examined denitrification in the benthic sediments did not appear to be a significant N sink relative to the annual load of N.

Using the site with the most complete data set as an example (BLS), we show in Fig. 5 and 6 the relationship between hydrology (discharge), NO3–N concentrations, and in-stream retention of NO3–N by denitrification, as occurred in an agricultural stream in east-central Illinois. There is a strong relationship between increasing discharge and stream NO3–N concentrations (Fig. 5) because most of the flow originates from agricultural drainage (David et al., 1997; Mitchell et al., 2000). Figure 6 shows the relationships between the concentration, –k, and mean daily load of NO3–N from the watershed. Increases in discharge lead to corresponding increases in both NO3–N concentrations and water depth, thereby reducing the ability of denitrification in the stream bed to affect the NO3–N load. Alternatively, during late summer, when discharge and NO3–N are both low, the stream bed is effective at removing NO3–N from the water column and controlling the transport of NO3–N through the stream.



View larger version (21K):
[in this window]
[in a new window]
  		Fig. 5. Relationship between mean daily discharge and stream NO3–N concentration for January 2001 through January 2002 in a headwater agricultural stream in Illinois. (Site characteristics are given in Table 1.)

 


View larger version (32K):
[in this window]
[in a new window]
  		Fig. 6. (A) Nitrate concentrations at site BLS, January 2001 through January. (B) Mean daily nitrate load and measured rates of nitrate N loss, –k (% d–1), for site BLS. (Site characteristics are given in Table 1.)

 
Hydrology is a major factor influencing the retention of NO3–N (and other nutrients) in streams. Meyer and Likens (1979) identified two distinct conditions for streams regarding nutrients: a processing-retention mode and a through-put mode. During periods of high discharge, nutrient inputs to the stream are quickly exported with little retention or processing (i.e., through-put mode). We suggest that hydrology is the primary control on the export of NO3–N from headwater agricultural streams and that the streams are in a through-put mode (Meyer and Likens, 1979) for all but a few weeks in late summer and autumn. The switching between through-put and processing modes is shown clearly by the estimates of past Sw,dn values (Fig. 4). Hill (1979) also concluded that during critical times of the year (i.e., high discharge, high NO3–N concentrations) the export of NO3–N from agricultural streams in Ontario was not affected by denitrification.

We suggest that the through-put mode that occurs in the streams we examined is due largely to the drainage and channelization that has occurred in the watersheds (see Rhoads and Herricks, 1996). With the extensive network of subsurface tile drains now in place throughout the Midwest, precipitation and solutes are quickly drained into channelized headwater streams, bypassing riparian buffer zones (Hill, 1976). The dynamics of N cycling and the role of denitrification in Illinois streams before agricultural development are unknown. However, an undisturbed prairie stream in Kansas showed substantial retention of inorganic N (Dodds et al., 2000), suggesting the headwater streams in east-central Illinois are less retentive of N now than they were before European settlement and conversion of the native prairie and wetlands to agriculture.


   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
REFERENCES
 
The nutrient spiraling model (Newbold, 1992) was used to assess the role of in-stream denitrification in affecting the concentration and transport of NO3–N in headwater agricultural streams in east-central Illinois. Nutrient spiraling has been used successfully to examine nutrient cycling in relatively undisturbed, low-nutrient streams (e.g., Hall and Tank, 2003), but we believe this study is the first to apply the concept to nitrogen-rich, agricultural streams. The results indicate that denitrification in the stream sediments did not affect the transport of NO3–N or represent a substantial N sink in the sites we examined. Although the areal denitrification rates tended to be high, the concentrations of NO3–N in the streams also were high and resulted in an overall low efficiency of NO3–N removal.

There are more than 20 million ha of drained cropland in the Mississippi River basin, much of it concentrated in the Corn Belt (Goolsby et al., 1999). Our study was limited to five headwater streams in east-central Illinois, but we believe the sites are typical of headwater streams in the Corn Belt that receive tile drainage. If the results obtained from our sites are representative of headwater agricultural streams throughout the Corn Belt, previous studies of N transport in Illinois (David and Gentry, 2000) and the Mississippi River basin (Alexander et al., 2000) may have overestimated the loss of N through denitrification in headwater streams. In the northeastern United States, headwater streams appear to be the major habitat for N removal (Seitzinger et al., 2002). In the agricultural Midwest, we suggest habitats such as reservoirs or floodplain wetlands may support greater N removal than do headwater streams, particularly those streams in tile-drained watersheds.


   ACKNOWLEDGMENTS
 
We thank Jamie Schaller, Kyle Herrman, Karen Starks, and Christy Davidson for assistance in the field and laboratory. Discussions with Steven Thomas helped develop the ideas presented in this paper. Robert Hall provided helpful comments on a draft of the paper, as did three anonymous reviewers. Discharge data for three of the sites was provided by the Illinois State Water Survey or the USGS. The Illinois State Water Survey also provided long-term data on water chemistry. This research was funded by the State of Illinois through the Illinois Council on Food and Agricultural Research (C-FAR), the Illinois Water Resources Center, and the USDA-CSREES 406 Water Quality Program.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 


Related articles in JEQ:

This Issue in Journal of Environmental Quality

JEQ 2004 33: 1177-1182. [Full Text]  



This article has been cited by other articles:


Home page
 J. Environ. Qual.Home page
L. J. Puckett, C. Zamora, H. Essaid, J. T. Wilson, H. M. Johnson, M. J. Brayton, and J. R. Vogel
Transport and Fate of Nitrate at the Ground-Water/Surface-Water Interface
J. Environ. Qual., May 1, 2008; 37(3): 1034 - 1050.
[Abstract] [Full Text] [PDF]

Home page
 J. Environ. Qual.Home page
J. H. Duff, A. J. Tesoriero, W. B. Richardson, E. A. Strauss, and M. D. Munn
Whole-Stream Response to Nitrate Loading in Three Streams Draining Agricultural Landscapes
J. Environ. Qual., May 1, 2008; 37(3): 1133 - 1144.
[Abstract] [Full Text] [PDF]

Home page
 J. Environ. Qual.Home page
J. L. Domagalski, S. Ator, R. Coupe, K. McCarthy, D. Lampe, M. Sandstrom, and N. Baker
Comparative Study of Transport Processes of Nitrogen, Phosphorus, and Herbicides to Streams in Five Agricultural Basins, USA
J. Environ. Qual., May 1, 2008; 37(3): 1158 - 1169.
[Abstract] [Full Text] [PDF]

Home page
 J. Environ. Qual.Home page
T. V. Royer, M. B. David, L. E. Gentry, C. A. Mitchell, K. M. Starks, T. Heatherly II, and M. R. Whiles
Assessment of Chlorophyll-a as a Criterion for Establishing Nutrient Standards in the Streams and Rivers of Illinois
J. Environ. Qual., March 1, 2008; 37(2): 437 - 447.
[Abstract] [Full Text] [PDF]

Home page
 J. Environ. Qual.Home page
E. Mehnert, H.-H. Hwang, T. M. Johnson, R. A. Sanford, W. C. Beaumont, and T. R. Holm
Denitrification in the Shallow Ground Water of a Tile-Drained, Agricultural Watershed
J. Environ. Qual., January 9, 2007; 36(1): 80 - 90.
[Abstract] [Full Text] [PDF]

Home page
 J. Environ. Qual.Home page
M. R. Opdyke, M. B. David, and B. L. Rhoads
Influence of Geomorphological Variability in Channel Characteristics on Sediment Denitrification in Agricultural Streams
J. Environ. Qual., October 27, 2006; 35(6): 2103 - 2112.
[Abstract] [Full Text] [PDF]

Home page
 J. Environ. Qual.Home page
J. M. Sogbedji and G. F. McIsaac
Evaluation of the ADAPT Model for Simulating Nitrogen Dynamics in a Tile-Drained Agricultural Watershed in Central Illinois.
J. Environ. Qual., September 1, 2006; 35(5): 1914 - 1923.
[Abstract] [Full Text] [PDF]

Home page
 J. Environ. Qual.Home page
S. H. Ensign, S. K. McMillan, S. P. Thompson, and M. F. Piehler
Nitrogen and Phosphorus Attenuation within the Stream Network of a Coastal, Agricultural Watershed.
J. Environ. Qual., July 1, 2006; 35(4): 1237 - 1247.
[Abstract] [Full Text] [PDF]

Home page
 J. Environ. Qual.Home page
A. M. Morgan, T. V. Royer, M. B. David, and L. E. Gentry
Relationships among Nutrients, Chlorophyll-a, and Dissolved Oxygen in Agricultural Streams in Illinois
J. Environ. Qual., May 31, 2006; 35(4): 1110 - 1117.
[Abstract] [Full Text] [PDF]

Home page
 J. Environ. Qual.Home page
P. W. van Driel, W. D. Robertson, and L. C. Merkley
Upflow Reactors for Riparian Zone Denitrification
J. Environ. Qual., February 2, 2006; 35(2): 412 - 420.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Related articles in JEQ
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (30)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Royer, T. V.
Right arrow Articles by David, M. B.
Right arrow Search for Related Content
PubMed
Right arrow PubMed C