HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Figures Only Full Text (PDF) Free Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via ISI Web of Science (2) Citing Articles via Google Scholar Google Scholar Articles by Panno, S. V. Articles by Hwang, H.-H. Search for Related Content PubMed PubMed Citation Articles by Panno, S. V. Articles by Hwang, H.-H. Agricola Articles by Panno, S. V. Articles by Hwang, H.-H. Related Collections Surface Water Quality Nutrients Nutrient Cycling Water Pollution Nitrogen

TECHNICAL REPORTS

Surface Water Quality

Isotopic Evidence of Nitrate Sources and Denitrification in the Mississippi River, Illinois

^a Illinois State Geological Survey, Natural Resources Building, 615 E. Peabody Street, Champaign, IL 61820
^b Illinois State Water Survey, 2204 Griffith Drive, Champaign, IL 61820-7495

^* Corresponding author (panno{at}isgs.uiuc.edu)

Received for publication January 14, 2005.

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
Anthropogenic nitrate (NO₃^–) within the Mississippi–Atchafalaya River basin and discharge to the Gulf of Mexico has been linked to serious environmental problems. The sources of this NO₃^– have been estimated by others using mass balance methods; however, there is considerable uncertainty in these estimates. Part of the uncertainty is the degree of denitrification that the NO₃^– has undergone. The isotopic composition of NO₃^– in the Mississippi River adjacent to Illinois and tile drain (subsurface drain) discharge in agricultural areas of east-central Illinois was examined using N and O isotopes to help identify the major sources of NO₃^– and assess the degree of denitrification in the samples. The isotopic evidence suggests that most of the NO₃^– in the river is primarily derived from synthetic fertilizers and soil organic N, which is consistent with published estimates of N inputs to the Mississippi River. The 1:2 relationship between ¹⁸O and ¹⁵N also indicate that, depending on sample location and season, NO₃^– in the river and tile drains has undergone significant denitrification, ranging from about 0 to 55%. The majority of the denitrification appears to have occurred before discharge into the Mississippi River.

Abbreviations: ¹⁵N, isotopic composition of nitrogen • ¹⁸O, isotopic composition of oxygen • , isotope enrichment factor • MARB, Mississippi–Atchafalaya River Basin • OM, organic matter

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
EXCESS N, primarily as NO₃^–, is a major pollutant of rivers, lakes, and estuaries and can adversely affect aquatic life (Parry, 1998). Nitrogen has been linked to the hypoxic zone that develops annually in the bottom waters of the Gulf of Mexico, threatening aquatic life and commercial fisheries (Goolsby et al., 1999; Rabalais et al., 2001, 2002). The N load in the Mississippi–Atchafalaya River Basin (MARB) doubled from the 1960s to the 1980s, and the mean NO₃–N concentration in the Mississippi River also doubled during that period (Goolsby et al., 1999). This increase has been attributed primarily to the use of synthetic fertilizers in the Corn Belt states of the midwestern USA (Rabalais et al., 2001). Goolsby et al. (1999) estimated that 60% of the N inputs into the MARB are from synthetic fertilizer and mineralized soil N. The remaining 40% comes from legumes (N₂ fixation), manure, atmospheric deposition, and point sources such as effluent from sewage treatment facilities. Policymakers are using estimated contributions of these N sources, typically calculated by mass balance approaches, to develop strategies to reduce NO₃^– inputs to the Mississippi River in the hope of decreasing the hypoxic zone. There is, however, a great deal of uncertainty in these estimates and the link between fertilizer use and hypoxia in the Gulf of Mexico is poorly understood. When comparing the inputs and outputs of the N mass-balance calculations for various river systems, there is typically a significant deficit in the output portion, most of which is generally attributed to denitrification (David and Gentry, 2000).

The N isotopic ratio has been used to evaluate NO₃^– sources and geochemical processes affecting the concentration of NO₃^– (Létolle, 1980, Hübner, 1986, Heaton, 1986, Komor and Anderson, 1993). For example, NO₃^– originating from sewage and livestock effluent is typically more enriched in the heavier N isotope (¹⁵N) compared with N and NO₃^– derived from atmosphere deposition, fertilizer N, and soil organic N (Kendall, 1998). There is significant overlap, however, of ¹⁵N values for NO₃^– derived from the latter three sources. When attempts have been made to distinguish NO₃^– derived from such sources using only ¹⁵N values, many uncertainties arise because of the initial overlap in ¹⁵N values and the isotopic changes that occur due to fractionation effects caused by geochemical processes such as denitrification (Kohl et al., 1971; Hauck et al., 1972). In recent years, researchers have included the measurement of O isotope ratios of NO₃^– as well as N isotopes, producing more definitive results. A number of researchers have used this dual-isotope technique to identify dominant sources of NO₃^– in natural waters and to determine whether or not denitrification has occurred (Heaton, 1986; Böttcher et al., 1990; Aravena et al., 1993; Wassenaar, 1995; Aravena and Roberston, 1998; Panno et al., 2001; Burns and Kendall, 2002; Chang et al., 2002).

Illinois is a major source of N to the Mississippi River. Most of the N from Illinois is agricultural in origin, but sewage is also significant, primarily from the Chicago area. David and Gentry (2000) estimated that sewage contributes about 16% of the total N to rivers in Illinois. Battaglin et al. (2001) used the dual-isotope approach to, in part, assess the degree of denitrification in the Mississippi River at several different sites from Illinois to Louisiana, but drew no conclusions as to the dominant sources of NO₃^–. Chang et al. (2002) were more successful by comparing land use with isotopic results; they observed distinct, although overlapping, isotopic signatures. The objectives of our investigation were to identify the major sources contributing to the total dissolved NO₃^– in the upper Mississippi River using N and O isotopes from the NO₃^– ion together with published mass balance data and to assess the occurrence and degree of denitrification.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
Sample Collection and Analysis
Water samples from the Mississippi River were collected quarterly from August 2001 through May 2002 near the intakes for local water utilities at Davenport, IA, Quincy, IL, and Chester, IL (Fig. 1). These utilities pump their water through pipes that either extend 10 m from the shore or from a concrete sea wall that extends to a depth of 10 m. No major tributaries enter the Mississippi River between the Illinois–Wisconsin border and Davenport. Between Davenport and Quincy, the Rock, Iowa, and Des Moines rivers discharge into the Mississippi. Between Quincy and Chester, two of the largest tributaries in the Midwest, the Illinois and Missouri rivers, enter the Mississippi River. All of these tributaries drain land in the Corn Belt region of the midwestern USA. Because there were no major tributaries <60 km upstream of any of the sampling sites, we assumed that the concentrations of dissolved constituents were fairly well mixed and uniform across the river at each sampling site. This is supported by Meade et al. (1995), who compared single-point pumped samples collected from the center of the river with composite sampling across the width and depth of the Mississippi River and reported that the dissolved composition of the single pumped samples was comparable to the composite river water samples. To help test this assumption, during one sampling event we took samples at three locations across the river at Quincy and compared results with a sample taken at the municipal utility plant intake.

View larger version (31K): [in this window] [in a new window]   Fig. 1. Sampling locations. The light gray area on the U.S. map shows the extent of the Mississippi River watershed.

Measurements of temperature, pH, Eh (redox potential), and specific conductance (SpC) were taken in the field during sample collection. Water samples were collected using a peristaltic pump, and were filtered through in-line 0.45-µm filter capsules and stored in polyethylene bottles. The tile water samples were collected directly from the stream of water exiting the tile drains or in a clean 19-L bucket, and then pumped through an in-line filter. Filtered samples were analyzed for anions, dissolved organic carbon (DOC), NH₄–N, dissolved Kjeldahl N (DKN), and NO₃^– isotopes. Samples collected for the analysis of ¹⁵N and ¹⁸O of the NO₃^– ion were acidified with HCl to a pH <2. All samples were transported in ice-filled coolers to the laboratory, and kept refrigerated at approximately 4°C until analysis.

Anion concentrations, including NO₃–N, were determined using ion chromatography. The DOC was determined by difference (total carbon – inorganic carbon) using persulfate–oxidation and infrared detection (Greenburg et al., 1987). The DKN was determined by converting organic N to NH₄⁺ by digestion with H₂SO₄; the resultant solution was analyzed for NH₄–N using a titrimetric procedure (Bremner and Mulvaney, 1982). Ammonium was determined using semi-automated colorimetry (R.A. Cahill, unpublished data, 1985). Analytical precision for the above methods is as follows: ion chromatography (±0.02 mg/L for Cl; ±0.02 mg/L for NO₃–N; ±0.02 mg/L for SO₄; ±0.23 mg/L for alkalinity), DOC (±0.02 mg/L), DKN (±0.25 mg/L), NH₄–N (±0.10 mg/L). These data are based on 2 of the repeated analysis of a standard (n = 10).

Isotopic analyses of N and O were conducted using published methods (Silva et al., 1994; 2000; Wassenaar, 1995) with some modifications (Hwang et al., 1999; Panno et al., 2001). The samples were first boiled under acidification to remove HCO₃^– and dissolved CO₂. The dissolved organic matter (DOM) and SO₄^2– were also removed to minimize contamination of the ¹⁸O due to the O in the SO₄^2– ions and the DOM, and to help eliminate anion interference during the ion-exchange step for NO₃^– extraction. The DOM was removed using a silicalite molecular sieve, and the SO₄^2– was removed by precipitation as BaSO₄. After removal of HCO₃^–, SO₄^2–, and DOM, the NO₃^– was extracted using an anion-exchange column packed with BioRad AG 1-X8 resin (BioRad, Hercules, CA). Nitrate collected on the anion-exchange column was eluted with HBr solution (Hwang et al., 1999) and converted to AgNO₃ by adding AgO. The AgNO₃ was precipitated by freeze-drying the sample in a vacuum system. The dried AgNO₃ was converted to N₂ and CO₂ for ¹⁵N and ¹⁸O analysis, respectively, by combustion techniques (Hwang et al., 1999; Silva et al., 2000). Both gases were analyzed on an isotope ratio mass spectrometer. International isotope standards IAEA-N1, IAEA-N2, USGS25, and USGS26 were used for ¹⁵N calibration. International standard IAEA-N3 was used for ¹⁸O calibration. Reproducibility of duplicate analyses was 0.9 for ¹⁸O (averaging ±0.3), and 0.3 for ¹⁵N (averaging ±0.1).

Recent work by Révész and Böhlke (2002) showed that the methods used to determine ¹⁸O values from the combustion of NO₃^– in sealed quartz glass tubes resulted in the interaction of the quartz O with the resultant CO₂ from the NO₃^– combustion, altering the ¹⁸O measured value of the NO₃^–. The new high-temperature conversion technique discussed by Révész and Böhlke (2002) does not contain any quartz in the reaction chamber and results in a more accurate ¹⁸O value for the NO₃^–. Because most of the published ¹⁸O data for NO₃^– up to this point were produced using quartz glass sealed-tube combustion techniques, however, we have used the sealed-tube combustion ¹⁸O values for this investigation. We have analyzed three National Institute of Standards and Technology (NIST) NO₃^– standards recently prepared by the USGS for ¹⁸O analyses and calculated a correction for the sealed-tube ¹⁸O values to adjust for the quartz O influence so that the values can be compared with the new high-temperature conversion technique in the future (Table 1).

View this table: [in this window] [in a new window]   Table 1. Results from comparison of National Institute of Standards and Technology (NIST) standard 18O values with our replicate values and adjusted values based on a least squares equation.

[1]

where X is the isotope of interest (e.g., ¹⁸O or ¹⁵N), and R is the ratio of ¹⁸O/¹⁶O or ¹⁵N/¹⁴N, respectively. Reference standards used in Eq. [1] are the Vienna-Standard Mean Ocean Water for O and air for N. The isotopic enrichment factor () for a chemical reaction can be defined as:

[2]

where = R_p/R_s, R_p is the isotopic ratio of the reaction product, and R_s is the isotopic ratio of the reaction substrate (reactant). The Rayleigh equation describes the evolution of the isotopic composition of the residual substrate during certain reactions where the products are removed from chemical or isotopic equilibrium with the reactant. The following is a simplified version of the Rayleigh equation:

[3]

where _o is the initial composition of the substrate and f is the remaining fraction of the substrate (Clark and Fritz, 1997).

Typical ranges of ¹⁵N and ¹⁸O for various natural and anthropogenic sources of NO₃^– are defined by boxes in Fig. 2 (Kendall et al., 1995; Clark and Fritz, 1997; Mengis et al., 2001). Soil OM has a large range of ¹⁵N values, although most soils have values between 2 and 10 (Fogg et al., 1998; Kendall, 1998).

View larger version (25K): [in this window] [in a new window]   Fig. 2. Ranges of 18O and 15N values for potential NO3– sources and values measured in Mississippi River and tile drain samples. The domain of soil organic matter is shaded to show the range of greatest (darkest) to lowest (lightest) frequency. The domain of soil organic matter partially overlaps that of reduced N fertilizer, manure, and sewage with respect to 15N. The 18O of NO3– from these reduced N sources was calculated as described in the discussion. The large arrow represents a denitrification vector; as denitrification proceeds, the 15N and 18O values of the remaining NO3– progressively increase in the direction of the vector. The size of circles represents NO3–N concentration; shaded circles indicate tile drain samples. In all cases, analytical errors are smaller than the circle size. Synthetic NO3– Fertilizer refers to synthetic fertilizer applied as NO3– (e.g., the NO3– in NH4NO3) and Reduced N Fertilizer refers to NO3– produced by nitrification of applied NH3 (as NH3 or the NH4+ in NH4NO3) and urea.

Most of the reduced N applied to fields is quickly oxidized to NO₃^–, which is taken up by plants, leached from the soil zone, or denitrified to N₂. A portion of the N from fertilizers can be lost via volatilization or be immobilized in the soil by uptake, storage, and recycling in the microbial biomass within the soil zone (Burkart and James, 1999; Mengis et al., 2001). Some of these processes can greatly affect the isotopic composition of the NO₃^– formed in the soil zone. For example, fertilizer N added as NO₃^– and taken up by plants or microbes, converted to organic N, and then returned to the soil and reoxidized to NO₃^– would no longer have the ¹⁸O value of the initial applied NO₃^–, but instead would have a ¹⁸O value like that of other reduced N sources that were nitrified in the soil zone (Mengis et al., 2001). A sample so affected would be expected to plot in the Soil OM domain in Fig. 2. Thus, NO₃^– in ground or surface water whose original source was synthetic fertilizer may have different isotopic signatures depending on how long it remained in the soil–plant environment; i.e., fertilizer NO₃^– leached soon after application is isotopically different than NO₃^– from fertilizer taken up by plants and later returned to the soil.

Biologically mediated reactions such as denitrification cause the N and O isotopes to fractionate, increasing the ¹⁵N and ¹⁸O values of the remaining (undenitrified) NO₃^–. The fractionation factors for ¹⁵N and ¹⁸O vary depending on the local conditions and rates of reaction; however, the ratio of the change in ¹⁸O and ¹⁵N during denitrification is typically close to 1:2 (Böttcher et al., 1990; Aravena and Robertson, 1998; Kendall, 1998; Mengis et al., 1999), as shown by the vector in Fig. 2.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
The NO₃–N concentrations in the 2001–2002 Mississippi River samples ranged from 0.2 to 2.9 mg/L, with mean concentrations for Davenport, Quincy, and Chester being 1.2, 1.7, and 2.1 mg/L, respectively (Table 2). These concentrations are similar to mean values measured between 1980 and 1996 at stations in the Mississippi River near our sites (Goolsby et al., 1999). Nitrate-N concentrations in the tile drain samples ranged from 6.5 to 15.3 mg/L, with a mean value of 11.5 mg/L (Table 2).

The ¹⁵N values of the NO₃^– in the water samples from the Mississippi River and the tile drains range was 4.8 to 16.4 and the ¹⁸O values range was 6.6 to 13.9 (Table 2); the data plot within a narrow band on a ¹⁵N–¹⁸O plot (Fig. 2). The samples having the greatest NO₃–N concentrations tended to have the lowest ¹⁵N and ¹⁸O values. Plots of NO₃–N concentrations vs. ¹⁵N and ¹⁸O values for the river samples show a significant negative correlation (r² = 0.64 and 0.61, respectively), and the isotopic values tended to decrease downstream (Fig. 3). The isotopic values of NO₃^– were lowest and showed less variability among the samples collected during the winter and spring, when NO₃–N concentrations were greatest. The tile drain samples had a similar isotopic range to the river samples, and negative correlations between NO₃–N and the isotopic values, although the correlations were not as strong as that observed for the Mississippi River samples (Fig. 3). Nitrate-N concentrations also tended to be greater in the spring and winter for the tile drains.

View larger version (16K): [in this window] [in a new window]   Fig. 3. Nitrate-N concentrations vs. (a) 15N and (b) 18O of NO3– for Mississippi River and tile drain samples. Letters indicate sample locations (D = Davenport, Q = Quincy, C = Chester). The gray hexagon represents the average composition of the four river samples collected in April 2004 at Quincy. The month of sample collection is shown for tile samples. Lines are linear regressions for river samples (r2 = 0.64 for 15N and 0.61 18O).

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
Sources of Nitrate to the Mississippi River
The isotopic data plot outside of the estimated domains representing the range of isotopic values expected for the major NO₃^– sources. The distribution of data could be explained by denitrification, mixing of different NO₃^– sources, or both. Plots of ¹⁵N and ¹⁸O vs. (NO₃–N)^–1 and ln(NO₃–N), which have been used by some researchers to distinguish between mixing and fractionation processes, were inconclusive. Linear relationships (r² > 0.5) were found for the river samples in both types of plots.

If mixing between atmospheric NO₃^– and NO₃^– from reduced sources was the dominant process affecting the isotopic values, the fact that the ¹⁸O values of the samples are significantly greater than the values estimated for the reduced N sources would require a large amount of the NO₃^– to reach the river as precipitation, without passing through the soil zone. Using a ¹⁸O value of 45 for NO₃^– in precipitation (Kendall, 1998), we calculate this amount to be at least 10% of the total NO₃^–, clearly an unrealistic value. Similarly, samples with relatively elevated ¹⁵N values would require a large input of manure and sewage if mixing was the only process affecting isotopic values. More than 50% of the NO₃^– would have to come from manure and sewage sources for the ¹⁵N value to exceed 12. If we expanded the ¹⁸O range of the source domains to reach 14 or 16, as observed in some samples of NO₃^– in forest soils (Mayer et al., 2001; Burns and Kendall, 2002), and then used mixing to explain the isotopic results for the Mississippi River and tile drain samples, it would suggest that the samples with the most positive isotopic values were generated from nearly 100% sewage or manure NO₃^– inputs. These scenarios are also obviously unrealistic; for comparison, the Illinois River, which receives discharge from some of the largest wastewater treatment plants in the world, has been estimated to receive 21% of its N load from sewage based on mass balance approaches (David and Gentry, 2000). It is clear, however, that some of the N in the Mississippi River comes from sewage sources, especially downstream of the mouth of the Illinois River.

Because mixing of sources alone cannot account for the range of isotopic values for the samples, it appears that denitrification has played an important role. If the NO₃^– is primarily from one source, then following the denitrification vector back toward its intersection with a source domain would indicate the dominant source. Doing this with the Mississippi River and tile data suggests that the NO₃^– in the samples comes primarily from fertilizer N and soil OM (see dotted line in Fig. 2). The vector shown in Fig. 2 was passed through the center of the observed data. This vector suggests a ¹⁸O value for the fertilizer and soil OM sources that is in the heavy part of these source domains for which the ¹⁸O composition was calculated. These slightly greater ¹⁸O values may be due to the contribution of NO₃^– from atmospheric and synthetic NO₃^– fertilizer sources. Approximately 15% of the synthetic N fertilizer applied to Illinois crop land is synthetic NO₃^– (USDA, 1999). To account for the possible input of higher ¹⁸O values that are associated with synthetic NO₃^– fertilizers, typically 19 to 25 (Mayer et al., 2001), we added a 15% contribution of NO₃^– with a ¹⁸O value of 22 (measured from a sample of NH₄NO₃ fertilizer obtained in southwestern Illinois) to estimate the range of values that would be expected from a combination of NO₃^– from nitrification of reduced N fertilizer and synthetic NO₃^– (15% NO₃^– + 85% Reduced Fertilizer in Fig. 2). However, we would normally not expect to observe the impact of the greater ¹⁸O values from the synthetic NO₃^– because of the rapid uptake and transformations of NO₃^– to organic N within the soil zone followed by the eventual oxidation back to NO₃^–, resetting the ¹⁸O value (Mengis et al., 2001). We could expect that the isotopically heavier O of synthetic NO₃^– may impact the ¹⁸O of the overall NO₃^– during periods of slow biological activity or rapid runoff of the NO₃^– from the fields to the streams, as observed by Mengis et al. (2001). Additionally, in some forested areas ¹⁸O-enriched atmospheric NO₃^– was observed to impact the ¹⁸O of NO₃^– in the shallow soil water and stream runoff (Durka et al., 1994; Burns and Kendall, 2002), indicating that NO₃^– with isotopically heavy ¹⁸O can, under certain conditions, affect the ¹⁸O value of the NO₃^– in shallow soil water and runoff.

Because NO₃^– is not initially present in the reduced N sources (reduced fertilizer, soil OM, manure, and sewage), the ¹⁸O values for NO₃^– that are produced by the oxidation of reduced N must be estimated. This was done by assuming that the biologically mediated nitrification process for most of the soils in the midwestern USA derives one-third of the O for the NO₃^– from atmospheric O and two-thirds from the surrounding water (Amberger and Schmidt, 1987; Böttcher et al., 1990; Kendall, 1998). Recent studies suggest that this is probably an oversimplification for predicting the ¹⁸O of NO₃^– during nitrification (Mayer et al., 2001; Burns and Kendall, 2002). These studies found that, in some cases, the ¹⁸O is greater than expected based on incubation experiments and on measurements of NO₃^– from forest soils, suggesting that perhaps the contribution of atmospheric O is sometimes greater than one-third. Mayer et al. (2001) suggested that in low-pH environments, which they observed in forest soils, a different bacterial process dominates the nitrification reaction and utilizes a greater amount of atmospheric O. Soil pHs in the forest studied by Mayer et al. (2001) were 3.2 to 3.3, whereas in Illinois, agricultural soil pHs are closer to 6.5 and range from 5.2 to 7.5 for the state (Illinois State Water Survey, 2005). There is also the possibility that the ¹⁸O of the soil water may be enriched in ¹⁸O due to respiratory isotope fractionation or evaporative effects that may occur within the soil zone (Kendall, 1998; Burns and Kendall, 2002).

In agricultural settings, ¹⁸O values for NO₃^– in shallow ground water samples are commonly quite low, with many of the larger values attributed to denitrification processes or fast throughput of NO₃^– fertilizers (Mengis et al., 2001; Aravena et al., 1993; Beaumont, 2003). The lower ¹⁸O values in these studies are consistent with the two-thirds water, one-third atmospheric O contribution for nitrification. There is always a range of values, however, which probably reflects the variable ¹⁸O of the soil water as well as the influence of some of the processes discussed above. It appears the environmental setting and climatic conditions affect the initial ¹⁸O value of NO₃^– during nitrification of reduced N sources.

Because our samples are primarily from agricultural settings, we decided to use the two-thirds water and one third atmospheric contribution of O to estimate the values for the initial ¹⁸O of NO₃^– for nitrification of reduced N sources. The range of ¹⁸O of NO₃^– from the nitrification reaction will vary depending on the ¹⁸O of the shallow ground water in contact with the nitrifying bacteria. Because the isotopic composition of shallow ground water reflects the average isotopic composition of precipitation for a geographic region, we used the average range of ¹⁸O_H2O for precipitation in Illinois and bordering states, which generally ranges from about –4 to –11 (Coplen and Kendall, 2000; summarized in Clark and Fritz, 1997). Using these ¹⁸O_H2O values gives an expected range of ¹⁸O_NO3 of 5.2 to 0.5 for the reduced N sources of NO₃^– (Fig. 2).

Denitrification
While the dominant source(s) of the NO₃^– for the tile drains and the Mississippi River appears to be the same, the NO₃–N concentrations were much lower for the river samples than those from the tile drains. The fact that the isotopic values for the tile and river samples cover a similar range (Fig. 2) is indirect evidence that in-stream denitrification was probably minimal in the Mississippi River. Battaglin et al. (2001) also concluded that there was little or no in-stream denitrification in the Mississippi River in their study of the MARB. The lack of evidence of denitrification in the Mississippi River was not unexpected, as N loss in streams declines rapidly with increasing channel size (Alexander et al., 2000). Dilution and biological uptake of NO₃^– are probably the dominant mechanisms for the lower NO₃–N concentrations in the river samples, as Battaglin et al. (2001) suggested. The majority of the NO₃^– entering streams and rivers in the Corn Belt passes through the soil zone and into the ground water before discharging to surface water. Therefore, it seems likely that most of the denitrification, indicated by the isotopic data, probably occurs within soils and ground water rather than in surface water. This conclusion is supported by Beaumont (2003), who obtained a similar range of isotopic values for shallow ground water samples collected in an agricultural watershed in central Illinois; the majority clustered along the same vector as the tile and Mississippi River samples reported here.

We approximated the degree of denitrification that has probably occurred in the Upper Mississippi River basin and the tile samples collected for this study by applying typical isotopic enrichment factors () to the Rayleigh equation. Although Battaglin et al. (2001) and Chang et al. (2002) were not able to distinguish denitrification in their MARB samples, our seasonally collected, more localized water samples followed a well-defined denitrification pattern. Further, we were primarily interested in the overall dentrification exhibited by water in the Mississippi River in comparison with that of tile drain water. To calculate the degree of denitrification for these data, an initial isotopic composition is needed. We chose to use two initial ¹⁵N and ¹⁸O values. The first estimate uses the average isotopic composition of four samples from the tile drains collected in the spring, during April and May (¹⁵N = 6.1 and ¹⁸O = 7.2), as the starting isotopic composition. About one-half of the N fertilizer applied to cropland in the midwestern USA is applied in the spring (typically April–May). Precipitation is usually greatest in the spring in Illinois, resulting in more rapid runoff and probably a better representation of initial isotopic composition of recently formed NO₃^– in the soil zone; rainfall amounts in April and May were slightly below average in 2000 and above average in 2002 and 2004. However, this estimated initial isotopic composition may be conservatively high because these two tiles maintained discharge all year, indicating the influence of local ground water with a longer flow path than most tile drains. Thus, there could be some denitrification already recorded in these spring samples. Lower ¹⁵N and ¹⁸O values for NO₃^– samples from tile drains, lysimeters, and shallow wells have been observed in other agricultural settings (Beaumont, 2003; Mengis et al., 2001; Aravena et al., 1993). For example, the average ¹⁵N and ¹⁸O values for spring samples of tile drains that discharged seasonally in a nearby agricultural field in central Illinois were 4.7 and 5.7, respectively (Beaumont, 2003).

The second set of initial ¹⁵N and ¹⁸O values for calculating the degree of denitrification was determined using theoretical contributions from the different major sources of NO₃^– as determined by mass-balance calculations for large river systems with emphasis on the Upper Mississippi River basin. Using published studies, we estimated that approximately 14% of the N load is from sewage and manure discharge, 5% from atmospheric deposition, 30% from synthetic fertilizers, and 51% from soil OM (including legumes and pasture) (Howarth et al., 1996; Burkart and James, 1999; Goolsby et al., 1999; David and Gentry, 2000). Using these percentages and the average isotopic compositions of the different major sources of NO₃^– (Table 3), the theoretical ¹⁵N would be 4.4 and the ¹⁸O 6.4 for the initial NO₃^– before denitrification effects for the Mississippi River data.

Using these published values and the "initial" isotopic compositions of NO₃^– estimated from the spring tile drain samples and the theoretical contributions from major NO₃^– sources, as discussed above, the amount of denitrification for the NO₃^– in the Mississippi River and the tile drain samples was calculated (Fig. 4). Because the drainage basin of the Upper Mississippi River is large and contains many different soil types and land uses, we feel that applying average or typical isotopic values for the various end-member NO₃^– sources and using typical values is a reasonable approach for modeling the data. Assuming that the spread of the isotopic values of the NO₃^– in water samples collected from the Mississippi River is due primarily to denitrification, these calculated vectors indicate that 0 to 55% of the original NO₃^– has been denitrified, with most of the samples falling between 10 and 40%. The isotopic values of samples from the tile drains suggest that 0 to 45% denitrification may have occurred in the water from the two cultivated fields sampled in central Illinois (Fig. 4). The sample from Tile 1 collected in August 2002 (with the largest ¹⁵N and ¹⁸O values) had the lowest flow and probably represented base-flow conditions (i.e., shallow ground water), suggesting a longer residence time for the NO₃^– in this sample compared with the other tile samples. We should keep in mind that if smaller values were used for these calculations, greater degrees of denitrification would be estimated, while larger values would result in smaller degrees of denitrification. For example, if values of –10 for ¹⁵N and –5 for ¹⁸O were used, then the sample with the largest ¹⁵N and ¹⁸O values for the Mississippi River samples would be estimated to have gone through approximately 65% denitrification; if values of –20 were used, then the largest degree of denitrification would be approximately 40%. Even with all the uncertainties involved with these calculations, the results indicate that denitrification is an important mechanism for NO₃^– loss, and help to validate what others have postulated using mass-balance calculations (Howarth et al., 1996; Burkart and James, 1999; Goolsby et al., 1999; David and Gentry, 2000).

View larger version (23K): [in this window] [in a new window]   Fig. 4. Detailed portion from Fig. 2, showing the two denitrification lines calculated from the two estimated initial 15N and 18O values (gray hexagons). Open symbols represent river samples, closed symbols represent tile drain samples. Letters indicate river sample locations (D = Davenport, Q = Quincy, C = Chester). Tick marks on the vectors indicate an estimation of the extent of denitrification (percentage of original NO3– lost, using values of –15.9 for 15N and –8 for 18O in Eq. [3] for NO3– having initial values of 4.4 for 15N and 6.4 for 18O [solid lines] and for NO3– having initial values of 6.1 for 15N and 7.2 for 18O [dashed lines]).

The isotopic data from the Mississippi River also suggest there was greater denitrification of the NO₃^– in upstream samples during the summer and fall seasons. The greater degrees of denitrification in the upstream samples may indicate relatively longer travel times in the more northern parts of the Upper Mississippi River basin than the southern parts during the summer and fall seasons. A greater percentage of the soils in Illinois and Iowa are drained with tiles compared with the soils in Wisconsin and Minnesota (Zucker and Brown, 1998). Tile drains provide a shorter and faster route for water to pass through the subsurface, limiting denitrification.

	CONCLUSIONS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
The greatest concentrations of NO₃–N for our seasonally collected water samples in both the Mississippi River and tile drain samples were observed in the samples collected during the winter and spring. Although the number of our samples is limited, the timing of the greatest NO₃^– concentrations agrees with previous studies of the Mississippi and Illinois rivers and follows or coincides with the timing of the major fertilizer application periods for the midwestern USA. The isotopic values of the NO₃^– ions from the Mississippi River and tile drain samples were lowest during the winter and spring collection periods, when the NO₃–N concentrations were greatest.

The isotopic composition of the NO₃^– from the Mississippi River and the tile drains overlap one another, suggesting that the dominant source of the NO₃^– for both data sets is similar. The ¹⁵N and ¹⁸O values for NO₃^– from the Mississippi River and tile drains follow a linear trend that suggests the isotopic compositions were significantly affected by denitrification. Mixing of different sources of NO₃^– did not sufficiently explain the range of isotopic data observed for the Mississippi River and tile drain samples. The trend and seasonal variation of the isotopic composition in both data sets indicated that the NO₃^– is primarily derived from agricultural fertilizers and soil organic N. Although the Mississippi River NO₃^– should have a greater variety of sources, such as manure and sewage inputs, the isotopic data are not consistent with a large fraction of the NO₃^– coming from either source, but small contributions from these two sources are likely. Thus, our isotope data support the findings of other investigators who used mass-balance approaches and concluded that elevated NO₃–N concentrations in the Mississippi River are primarily due to crop-related agricultural N sources (Turner and Rabalais, 1994; Howarth et al., 1996; Burkart and James, 1999; Goolsby et al., 1999).

Using published values and calculated starting ¹⁵N and ¹⁸O compositions for NO₃^–, the estimated degree of denitrification varied from about 0 to nearly 55% for the Mississippi River samples and from about 0 to 45% for the tile drain samples, depending on sample location and season. The majority of the denitrification appears to have occurred before discharge into the Mississippi River, presumably in the soil zone and shallow ground water environments. To establish better estimates of the degree of denitrification that actually take place, additional research is needed to determine the best isotope enrichment factors and help establish the initial isotopic composition of the NO₃^– for the various sources in the agriculturally dominated midwestern USA.

	ACKNOWLEDGMENTS

 
This research was supported through a grant from the Illinois Groundwater Consortium at Southern Illinois University in Carbondale and by the Illinois Department of Natural Resources. We thank the Iowa-American Water Co., the Quincy Water Co., and the Chester Water Co. for their assistance in collecting river-water samples. We also thank E. Mehnert, B. Herzog, and J. Goodwin of the ISGS, and H. Wehrmann, E. Krug, T. Holm, and D. Winstanley of the ISWS for their critical reviews and insightful comments. Publication of this article has been authorized by the Chiefs of the Illinois State Geological and Water Surveys.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 

• Alexander, R.B., R.A. Smith, and G.E. Schwarz. 2000. Effect of stream channel size on the delivery of nitrogen to the Gulf of Mexico. Nature (London) 403:758–761.[CrossRef][Medline]
• Amberger, A., and H.-L. Schmidt. 1987. Naturliche isotopengehalte von nitrate als indikatoren fur dessen hAntwerkunft. Geochim. Cosmochim. Acta 51:2699–2705.
• Aravena, R., M.L. Evans, and J.A. Cherry. 1993. Stable isotopes of oxygen and nitrogen in source identification of nitrate from septic systems. Ground Water 31:180–186.[CrossRef][ISI]
• Aravena, R., and W.D. Robertson. 1998. Use of multiple isotope tracers to evaluate denitrification in ground water: Study of nitrate from a large-flux septic system plume. Ground Water 36:975–982.[CrossRef]
• Bates, H.K., G.E. Martin, and R.F. Spalding. 1998. Kinetic isotope effects in production of nitrite-nitrogen and dinitrogen gas during in situ denitrification. J. Environ. Qual. 27:183–192.[ISI]
• Bates, H.K., and R.F. Spalding. 1998. Aquifer denitrification as interpreted from in situ microcosm experiments. J. Environ. Qual. 27:174–182.[ISI]
• Battaglin, W.A., C. Kendall, C.C.Y. Chang, S.R. Silva, and D.H. Campbell. 2001. Chemical and isotopic evidence of nitrogen transformation in the Mississippi River, 1997–98. Hydrol. Process. 15:1285–1300.[CrossRef]
• Beaumont, W.C. 2003. Denitrification in a subsurfaced drained, agricultural watershed in Central Illinois. MS thesis. Dep. of Geology, Univ. of Illinois, Urbana-Champaign.
• Böhlke, J.K., S.J. Mroczkowski, and T.B. Coplen. 2003. Oxygen isotopes in nitrate: New reference materials for 18O: 17O: 16O measurements and observations on nitrate-water equilibrium. Rapid Commun. Mass Spectrom. 17:1835–1846.[CrossRef][ISI][Medline]
• Böttcher, J., O. Strebel, S. Voerkelius, and H.-L. Schmidt. 1990. Using isotope fractionation of nitrate nitrogen and nitrate oxygen for evaluation of denitrification in a sandy aquifer. J. Hydrol. (Amsterdam) 114:413–424.
• Bremner, J.M., and C.S. Mulvaney. 1982. Nitrogen-total. p. 595–624. In A.L. Page et al. (ed.) Methods of soil analysis. Part 2. ASA and SSSA, Madison, WI.
• Burkart, M.R., and D.E. James. 1999. Agricultural-nitrogen contributions to hypoxia in the Gulf of Mexico. J. Environ. Qual. 28:850–859.[Abstract/Free Full Text]
• Burns, D.A., and C. Kendall. 2002. Analysis of ¹⁵N and ¹⁸O to differentiate NO₃^– sources in runoff at two watersheds in the Catskill Mountains of New York. Water Resour. Res. 38:9-1–9-11.
• Clark, I.D., and P. Fritz. 1997. Environmental isotopes in hydrogeology. Lewis Publ., New York.
• Chang, C.C.Y., C. Kendall, S.R. Silva, W.A. Battaglin, and D.H. Campbell. 2002. Nitrate stable isotopes: Tools for determining nitrate sources among different land uses in the Mississippi River Basin. Can. J. Fish. Aquat. Sci. 59:1874–1885.[CrossRef]
• Coplen, T.B., and C. Kendall. 2000. Stable hydrogen and oxygen isotope ratios for selected sites of the U.S. Geological Survey's NASQAN and benchmark surface-water networks. U.S. Geol. Surv. Open-File Rep. 00-160. USGS, Reston, VA.
• David, M.B., and L.E. Gentry. 2000. Anthropogenic inputs of nitrogen and phosphorus and riverine export for Illinois, USA. J. Environ. Qual. 29:494–508.[Abstract/Free Full Text]
• Durka, W., E.-D. Shulze, G. Gebauer, and S. Voerkelius. 1994. Effects of forest decline on uptake and leaching of deposited nitrate determined from ¹⁵N and ¹⁸O measurements. Nature (London) 172:765–767.[CrossRef]
• Fausey, N.R., L.C. Brown, H.W. Belcher, and R.S. Kanwar. 1995. Drainage and water quality in Great Lakes and Cornbelt states. J. Irrig. Drain. Eng. 121:283–288.[CrossRef]
• Fogg, G.E., D.E. Rolston, D.L. Decker, D.T. Louie, and M.E. Grismer. 1998. Spatial variation in nitrogen isotope values beneath nitrate contamination sources. Ground Water 36:418–426.[CrossRef][ISI]
• Goolsby, D.A., W.A. Battaglin, G.B. Lawrence, R.S. Artz, B.T. Aulenbach, R.P. Hooper, D.R. Keeney, and G.J. Stensland. 1999. Flux and sources of nutrients in the Mississippi–Atchafalaya River Basin: Topic 3 report for the integrated assessment on hypoxia in the Gulf of Mexico. NOAA Coastal Ocean Prog. Anal. Ser. 17. NOAA Coastal Ocean Office, Silver Spring, MD.
• Greenburg, A.E., J.J. Connor, and D. Jenkins. 1987. Standard methods for the examination of water and wastewater. 15th ed. Am. Public Health Assoc., Washington, DC.
• Hauck, R.D., W.V. Bartholomew, J.M. Bremner, F.E. Broadbent, H.H. Cheng, A.P. Edwards, D.R. Keeney, J.O. Legg, S.R. Olsen, and L.K. Porter. 1972. Use of variations in natural nitrogen isotope abundance for environmental studies: A questionable approach. Science (Washington, DC) 177:453–454.[Free Full Text]
• Heaton, T.H.E. 1986. Isotopic studies of nitrogen pollution in the hydrosphere and atmosphere: A review. Chem. Geol. 59:87–102.
• Hoefs, J. 1987. Stable isotope geochemistry. 3rd ed. Springer-Verlag, New York.
• Howarth, R.W., G. Billen, D. Swaney, A. Townsend, N. Jaworski, K. Lajtha, J.A. Downing, R. Elmgren, N. Caraco, T. Jordan, F. Berendse, J. Freney, B. Kudeyarov, P. Murdoch, and Z. Zhao-Liang. 1996. Regional nitrogen budgets and riverine N and P fluxes for the drainages to the North Atlantic Ocean: Natural and human influences. Biogeochemistry 35:75–139.[CrossRef][ISI]
• Hübner, H. 1986. Isotope effects on nitrogen in the soil and biosphere. p. 361–425. In P. Fritz and J. Fontes (ed.) Handbook of environmental isotope geochemistry 2. Elsevier, Amsterdam.
• Hwang, H.-H., J. Liu, and K.C. Hackley. 1999. Method improvement for oxygen isotope analysis in nitrates. p. A-23. In Geological Society of America, Abstracts with Programs, North-Central Section, 31. Geol. Soc. Am., Boulder, CO.
• Illinois State Water Survey. 2005. Alternative crop suitability maps-soils. Illinois State Water Survey, Champaign, IL. Available at www.sws.uiuc.edu/data/altcrops/gisoils.asp (accessed 15 Sept. 2005; verified 27 Oct. 2005).
• Kendall, C. 1998. Tracing nitrogen sources and cycling in catchments. p. 517–576. In C. Kendall and J.J. McDonnell (ed.) Isotope tracers in catchment hydrology. Elsevier Science, Amsterdam.
• Kendall, C., M. Sklash, and T.D. Bullen. 1995. Isotope tracers of water and solute sources in catchments. p. 261–303. In S.T. Trudgill (ed.) Solute modeling in catchment systems. John Wiley & Sons, New York.
• Kohl, D.H., G.B. Shearer, and B. Commoner. 1971. Fertilizer nitrogen: Contribution to nitrate in surface water in a Corn Belt watershed. Science (Washington, DC) 174:1331–1334.[Abstract/Free Full Text]
• Komor, S.C., and H.W. Anderson, Jr. 1993. Nitrogen isotopes as indicators of nitrate sources in Minnesota Sand-Plains aquifers: Natural sources versus human pollution. J. Hydrol. (Amsterdam) 61:285–301.
• Létolle, R. 1980. Nitrogen-15 in the natural environment. p. 407–433. In P. Fritz and J. Fontes (ed.) Handbook of environmental isotope geochemistry 1. Elsevier, Amsterdam.
• Mariotti, A., A. Landriau, and B. Simon. 1988. ¹⁵N isotope biogeochemistry and natural denitrification process in groundwater: Application to the chalk aquifer of northern France. Geochim. Cosmochim. Acta 52:1869–1878.
• Mayer, B., S.M. Bollwerk, T. Mansfeldt, B. Hutter, and J. Veizer. 2001. The oxygen isotope composition of nitrate generated by nitrification in acid forest floors. Geochim. Cosmochim. Acta 65:2743–2756.[CrossRef]
• Meade, R.H., J.A. Moody, and H.H. Stevens, Jr. 1995. Sampling the big rivers. p. 40–51. In R.H. Meade (ed.) Contaminants in the Mississippi River, 1987–92. U.S. Geol. Surv. Circ. 1133. USGS, Reston, VA.
• Mengis, M., S.L. Schiff, M. Harris, M.C. English, R. Aravena, R.J. Elgood, and A. MacLean. 1999. Multiple geochemical and isotopic approaches for assessing ground water NO₃^– elimination in a riparian zone. Ground Water 37:448–457.[CrossRef][ISI]
• Mengis, M., U. Walther, S.M. Bernasconi, and B. Wehrli. 2001. Limitations of using ¹⁸O for the source identification of nitrate in agricultural soils. Environ. Sci. Technol. 35:1840–1844.[Medline]
• Panno, S.V., K.C. Hackley, H.H. Hwang, and W.R. Kelly. 2001. Determination of the sources of nitrate contamination in karst springs using isotopic and chemical indicators. Chem. Geol. 179:113–128.
• Parry, R. 1998. Agricultural phosphate and water quality: A U.S. Environmental Protection Agency perspective. J. Environ. Qual. 27:258–261.[ISI]
• Rabalais, N.N., R.E. Turner, and D. Scavia. 2002. Beyond science into policy: Gulf of Mexico hypoxia and the Mississippi River. BioScience 52:129–142.[CrossRef][ISI]
• Rabalais, N.N., R.E. Turner, and W.J. Wiseman, Jr. 2001. Hypoxia in the Gulf of Mexico. J. Environ. Qual. 30:320–329.[Abstract/Free Full Text]
• Révész, K., and J.-K. Böhlke. 2002. Comparison of ¹⁸O measurements in nitrate by different combustion techniques. Anal. Chem. 74:5410–5413.[Medline]
• Silva, S.R., C. Kendall, C.C. Chang, J.C. Radyk, and D.H. Wilkinson. 1994. A new method of preparing dissolved nitrate for nitrogen and oxygen isotopic analysis. EOS Trans. 75:280 (abstr.).
• Silva, S.R., C. Kendall, D.H. Wilkinson, C.C.Y. Chang, A.C. Zeigler, and R. Avanzino. 2000. A new method for collection of nitrate from fresh water and the analysis of nitrogen and oxygen ratios. J. Hydrol. (Amsterdam) 228:22–36.
• Turner, R.E., and N.N. Rabalais. 1994. Coastal eutrophication near the Mississippi River delta. Nature (London) 368:619–621.[CrossRef][ISI]
• USDA. 1999. Illinois Agricultural Annual Summary. Fertilizer. NASS-USDA, Washington, DC. Available at www.nass.usda.gov/il/1999/99110.htm (verified 16 Oct. 2005).
• Vogel, J.C., A.S. Talma, and T.H.E. Heaton. 1981. Gaseous nitrogen as evidence for denitrification in groundwater. J. Hydrol. (Amsterdam) 50:191–200.
• Wassenaar, L.I. 1995. Evaluation of the origin and fate of nitrate in the Abbotsford Aquifer using the isotopes of ¹⁵N and ¹⁸O in NO₃^–. Appl. Geochem. 10:391–405.[CrossRef]
• Zucker, L.A., and L.C. Brown (ed.). 1998. Agricultural drainage: Water quality impacts and subsurface drainage studies in the Midwest. Ohio State Univ. Ext. Bull. 871. Ohio State Univ., Columbus, OH. Available at http://ohioline.osu.edu/b871/index.html (verified 16 Oct. 2005).

This Article Abstract <