- Email: [email protected]
Quantification of diffuse nitrate inputs into a small river system using stable isotopes of oxygen and nitrogen in nitrate
Organic Geochemistry Organic Geochemistry 37 (2006) 1333–1342 www.elsevier.com/locate/orggeochem
Quantification of diffuse nitrate inputs into a small river system using stable isotopes of oxygen and nitrogen in nitrate Barbara Deutsch
a,*
, Melanie Mewes b, Iris Liskow a, Maren Voss
a
a
b
Baltic Sea Research Institute, Seestr. 15, 18119 Rostock, Germany Chair of Landscape Economics, University of Greifswald, Grimmer Str. 88, 17487 Greifswald, Germany Received 31 August 2005; accepted 14 April 2006 Available online 9 August 2006
Abstract To identify and quantify diffuse nitrate inputs into a river sub basin in Mecklenburg-Vorpommern (Germany) a dualisotope approach with d15N and d18O in nitrate was carried out from October 2002 to March 2003. Three nitrate sources (water from artificially drained agricultural soils, groundwater, atmospheric deposition) and the river were sampled monthly to bimonthly. Nitrate of the drainage water had a concentration weighted mean (cwm) d15N value of 10.4& and d18O of 4.7&, and was significantly different to the groundwater nitrate (cwm d15N = 0.6&; d18O = 1.4&). The low d18O values indicated that most of the nitrate from these sources was formed during the nitrification process of soil organic N. Nitrate of atmospheric deposition had a cwm d15N value of 0.1& and a d18O value of 51.7&. River nitrate showed cwm values of 9.0& in d15N and 6.0& in d18O close to the isotope values of the drainage water nitrate. The isotope values were used in a three source mixing-model, to determine the contribution of each sampled nitrate source to the total river nitrate. The mixing-model revealed that the nitrate from the drainage water contributed 86% of the river nitrate. Contribution of nitrate from groundwater and atmospheric deposition was 11% and 3%, respectively. These results agree with estimations of nitrate input data for this sub basin given by a nutrient emissions model. 2006 Elsevier Ltd. All rights reserved.
1. Introduction Although inputs of nutrients have been reduced in the recent years, the excessive load of nitrogen and phosphorous is still one of the major ecological problems of the Baltic Sea (Sta˚lnacke et al., 1999). In 2000 an amount of 814 · 103 t nitrogen entered the Baltic Sea, and more than 84% of the total N *
Corresponding author. Tel.: +49 381 5197 417; fax: +49 381 5197 440. E-mail address: [email protected] (B. Deutsch).
load derived from rivers (HELCOM, 2003). Worldwide, almost 70% of the riverine N load consists of dissolved organic nitrogen (DON; Meybeck, 1982), but experiments showed that its bioavailability may be as low as 2–16% (Stepanauskas and Leonardson, 1999), while nitrate is the dominant inorganic N-species lost to the aquatic environment (Addiscott et al., 1992) and is rapidly consumed. Diffuse nitrate inputs such as fertilizer runoff from farmland, atmospheric deposition and groundwater input are especially hard to identify, because they are emitted over large areas. Nowadays, a powerful tool to distinguish different nitrate sources is the
0146-6380/$ - see front matter 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.orggeochem.2006.04.012
1334
B. Deutsch et al. / Organic Geochemistry 37 (2006) 1333–1342
determination of stable isotope ratios of nitrogen and oxygen (Wassenaar, 1995; Spoelstra et al., 2001; Chang et al., 2002; Mayer et al., 2002; Piatek et al., 2005). The d15N values of nitrate from different sources often show overlapping ranges, but the additional measurement of the d18O values allows a more precise classification (Mayer et al., 2002). Nitrate, derived from sewage and manure, is isotopically distinct from atmospheric nitrate in d15N (7–20&, 10& to 8&, respectively), as well as in d18O (<15& compared to 25–75&; Wassenaar, 1995; Kendall, 1998). Nitrate originating from mineral fertilizers shows d15N values of 0 ± 4& (Kendall, 1998), and d18O values of 22 ± 3& (Amberger and Schmidt, 1987) because of their production from atmospheric nitrogen (d15N = 0&) and oxygen (d18O = 23.5&). However, the isotopic composition of nitrate collected in drainage tiles and runoff ditches does not reflect exactly the isotope values of the fertilizer applied, but is altered due to isotope fractionation processes (Kendall and Aravena, 1999). Flipse and Bonner (1985) demonstrated that groundwater nitrate produced under fertilized fields showed d15N values up to 12.4& higher than the fertilizers applied, and explained this difference as the volatile loss of ammonia from the fertilizer, containing reduced nitrogen forms. Another fractionation process is denitrification, which increases the d15N and d18O values of the residual nitrate with an enrichment of d18O:d15N close to 1:2 (Bo¨ttcher et al., 1990). Nitrate uptake by plants (Ho¨gberg, 1997) and soil N-mineralization, including ammonification and nitrification may also modify the isotope signature of nitrate (Iqbal et al., 1997; Mayer et al., 2001). The objective of this study was to test whether the isotopic composition of nitrate from three diffuse nitrate sources can be used to quantify the diffuse nitrate inputs into a sub basin of the Warnow River (444 km2), which also supplies the city of Rostock (200,000 inhabitants) with drinking water. The three nitrate sources, drainage water from fertilized fields, groundwater, and atmospheric deposition, as well as the river itself were sampled regularly from October 2002 to March 2003. The samples were analysed for nitrate concentration, d15N and d18O values. The data were used in a three source mixing-model (Phillips and Koch, 2002) to determine the percentage of every sampled nitrate source to the river nitrate. For the successful application of a conservative isotope mixing-model, it is necessary that the isotope values of the river nitrate
are not altered due to fractionation processes. For this purpose, sampling was carried out during late fall and winter. At low water temperature microbial activity is reduced (Pfenning and McMahon, 1996) and therefore alteration of the river nitrate isotope values due to fractionation processes is minimal. The usefulness of this approach as an additional method for nitrate source quantification beside model estimations is discussed. 2. Material and methods 2.1. Study area The Warnow River is located in MecklenburgVorpommern (north eastern Germany) and flows into the Southern Baltic Sea at the City of Rostock (Fig. 1a and b). With a length of 149 km and a drainage area of 3270 km2, mainly dominated by agricultural and forested areas (63% and 24% respectively; Pagenkopf, 2001) it is the second largest river system of Mecklenburg-Vorpommern (Thiele and Mehl, 1995). The Warnow River is characterized as polytrophic (Bo¨rner et al., 1994) due to a high amount of nutrient inputs (total nitrogen: 4140 t yr1, phosphorous: 200 t yr1), and a production rate of 650 g C m2 yr1. The riversystem is divided into seven sub basins, and the investigated sub basin ‘Middle Warnow’ covers an area of 444 km2. Land use in the sub basin is dominated by agriculture, with arable land (49%) and pasture (23%), followed by forests (17%), urban areas (8%), and waters (3%; Pagenkopf, 2001). According to the Mesoscale Agricultural Mapping Programme (MMK), which provides information about soil textures, 75% of the agricultural land in the sub basin ‘Middle Warnow’ are waterlogged soils. Estimations showed that up to 50% of the agricultural area might be artificially drained (Bockholt and Kappes, 1994). 2.2. Sampling From November 2002 to April 2003 two tile drain outlets, located on the area of a farmers’ cooperative near Rostock (Fig. 1c) were sampled monthly to bimonthly (five samples). The distance between the tile drains and the Warnow River is approx. 2 km. The drainage system is located at a soil depth of 60–100 cm. The soil textures are mainly sandy loams (46%) and sands (24%) from moraine substrates. The dominant soil types are
B. Deutsch et al. / Organic Geochemistry 37 (2006) 1333–1342
1335
Fig. 1. Site map of the Warnow River system (a) located at the south western coast of the Baltic sea (b). The dark grey area shows the sub basin ‘Middle Warnow’. Numbers in circles indicate the sampling sites (1: tile drain outlets, ditches, and river sampling; 2: groundwater sampling; 3: sampling of atmospheric deposition in Rostock-Warnemuende). Map (c) shows the tile drain outlets, ditches, and the river sampling site in more detail.
cambisols, luvisols and gleysols. The drained areas are approximately 0.025 km2 (outlet 1) and 0.15 km2 (outlet 2), respectively. Outlet 2 was also supplied with surface water of two small ponds located in the field. Cultivated crops were winter wheat/winter barley in 2002 and sugar beets/corn in 2003 (outlet 1) and for tile drain outlet 2 winter wheat in 2002 and winter barley in 2003. Fertilizer application and the composition of fertilizer used are shown in Table 1. Application of manure on both areas takes place every fifth to sixth year, and the last application on the soil above outlet 2 was in autumn 1998 with a similar amount as applied above tile 1 in 2002. Two adjacent ditches (ditch 1 and 2; Fig. 1c), connecting the field runoff
with the Warnow River were sampled additionally, to examine the possible alteration of the isotopic composition of nitrate due to isotope fractionation processes during the passage of the drainage water to the river. Three samples were taken from the Warnow River from January 2003 to March 2003 close to the mouth of the adjacent ditches (Fig. 1c). Sampling took place from a landing stage at a maximum depth of 0.3 m. Water temperatures measured close to the sampling location decreased from 8.6 C (October 2002) to 0.1 C (December 2002) and then increased to 6.5 C (March 2003; Table 2; unpublished data, Office for the Environment and Nature, Rostock). Groundwater samples were taken from a
1336
B. Deutsch et al. / Organic Geochemistry 37 (2006) 1333–1342
Table 1 Application date, amount, and composition of the fertilizers used on the soils above the sampled drainage tiles Date
Amount applied (kg N ha1)
Fertilizer
Nitrogen compounds
Tile drain outlet 1 03-25-2002 04-23-2002 05-29-2002 August 2002
85 70 79 232
AHL Urea KAS Manure
28% N ð7% NHþ 4 ; 7% NO3 ; 14% amideÞ 46% N (46% amide) 27% N ð13:5% NHþ 4 ; 13:5% NO3 Þ
Tile drain outlet 2 03-27-2002 04-22-2002 05-28-2002 03-18-2003 04-28-2003
66 61 148 79 90
AHL NTS NTS AHL Urea
28% 27% 27% 28% 46%
Table 2 Water temperatures of the Warnow River measured close to the sampling location during the sampling period Date
Temperature (C)
10-28-02 11-12-02 11-27-02 12-10-02 01-09-02 01-22-03 02-04-03 02-19-03 03-05-03 03-18-03
8.6 5.0 5.7 0.1 1.3 3.7 1.5 1.8 3.5 6.5
sampling site near the village of Reez close to a wood and arable land (Fig. 1a) from November 2002 to March 2003 (five samples). The well had a depth of 8.4 m and sampling was carried out with a submerged pump at a maximum discharge flow of 5 l min1. The aquifer had a thickness of approx. 5 m and there was a direct discharge into the river (K. Hennig, Eurawasser GmbH, pers. com.). Eight samples of precipitation were collected from October 2002 to April 2003 at the German Meterological Service in Rostock-Warnemuende, 15 km north of the study area (Fig. 1a). Samplers were funnel shaped (diameter 24 cm) with a vial below. After every rainfall they were immediately emptied and subsampled for nitrate analysis. Because of low nitrate concentrations samples of consecutive rainfall events were combined for isotope analysis. 2.3. Laboratory work After sampling, the water was filtered through a 0.45 lm membrane filter. The determination of
N N N N N
ð7% NHþ 4 ; 7% ð6% NHþ 4 ; 8% ð6% NHþ 4 ; 8% ð7% NHþ 4 ; 7% (46% amide)
NO 3 ; 14% NO 3 ; 13% NO 3 ; 13% NO 3 ; 14%
amideÞ amideÞ amideÞ amideÞ
nitrate concentration was carried out after Grasshoff et al. (1999). Subsequently, samples were prepared for isotopic analysis of d15N and d18O in nitrate using the method of Silva et al. (2000). In brief, samples were passed through a cation exchange resin (5 ml AG 50W-X4, H+-form; Biorad), followed by an anion exchange resin (2 ml AG1-X8, Cl-form; Biorad). At least 60 lMol nitrate was finally collected on the anion exchange resin. The resins with the absorbed nitrate were stored in a refrigerator for several weeks until further preparation. For further processing samples were eluted from the anion exchange resin with 15 ml of 3 M HCl, and neutralized with 6 g Ag2O to obtain a pH of 5.5–6. The precipitated AgCl and remaining Ag2O was removed by filtration (0.45 lm membrane filter). Two milliliters of 1 M BaCl solution was added to the filtrate to remove SO2 and PO3 4 4 . Precipitated BaSO4 and Ba3(PO4)2 were removed by filtration (0.45 lm membrane filter). The sample was passed through a cation exchange resin (5 ml AG 50W-X4, H+-form; Biorad) to eliminate the excess Ba2+. Then a second neutralization with Ag2O (1–2 g) was carried out and the resulting AgCl and excess Ag2O were removed by filtration (0.45 lm membrane filter). The solution, now containing Ag+ and NO 3 , was freezedried and the remaining solid AgNO3 was weighed into silvercaps for determination of d15N and d18O values of nitrate. The d18O values were determined with a Thermo Finnigan Delta Plus isotope ratio mass spectrometer (IRMS) after pyrolysis in a Thermo Finnigan TC/EA. Temperature of pyrolysis was 1350 C. The isotope ratios are given as d values in per mil (&) relative to a standard and were calculated after the following equation:
B. Deutsch et al. / Organic Geochemistry 37 (2006) 1333–1342
1337
d18 O ¼ ð18 O=16 Osample =18 O=16 Ostandard Þ 1 1000:
3. Results
The same equation is used for calculation of the d15N values. For d18O the international standard is Vienna standard mean ocean water (VSMOW) defined as 0&. Calibration of the reference gas CO was done with IAEA-KNO3 (d18O = 25.1 ± 0.6&; n = 51) and IAEA-C3-Cellulose (d18O = 32.2 ± 0.2&; n = 38). Precision of the measurement was verified by repeated analysis of an internal laboratory standard. The standard was KNO3 from Merck with the lot No. A286463 (d18O = 24.6 ± 0.7&; n = 52). Samples for d15N were also analysed in a Thermo Finnigan Delta Plus IRMS after combustion in a Flash EA at a temperature of 1020 C vs. air nitrogen. International reference materials used were IAEA-N1 (d15N = 0.4 ± 0.07&) and IAEA-N2 (d15N = 20.3 ± 0.09&), the internal laboratory standards were acetanilide (d15N = 1.7 ± 0.2&; n = 244), peptone (d15N = 5.7 ± 0.2&; n = 245) and Merck-KNO3 (d15N = 0.4 ± 0.16%; n = 20). Reproducibility of the preparation procedure was tested with AgNO3, which was prepared from an internal lab standard along with the samples. The standard was Merck-KNO3 ð1000 lMol 18 NO 3 =lÞ with SO4 (620 lMol). d O values of the resulting AgNO3 were 23.4 ± 0.7& (n = 16) and d15N = 0.4 ± 0.2& (n = 16). Since the d15N and d18O values of the produced AgNO3 were in the range of the standard deviations of the d values of the original Merck-KNO3, the isotope values of the prepared samples were not corrected.
The nitrate concentrations, d15N, and d18 O–NO 3 values of all samples are summarized in Table 3. The highest nitrate concentrations were found in the two tile drain outlets and in the groundwater, with a maximum value of 1462 lMol in outlet 2 at the end of December 2002. Outlet 2 and the groundwater showed high variability in nitrate concentrations, with no visible upward or downward trend. Table 3 Nitrate concentration, d15N, and d18 O–NO 3 values of the tile drain outlets, ditches, groundwater, atmospheric deposition, and the river
To estimate the contribution of the nitrate sources to the Warnow River a mixing-model based on mass balance equations was used (Phillips and Koch, 2002). The equations are d15 NW ¼ fD d15 ND þ fG d15 NG þ fA d15 NA ; 18
18
18
18
Tile drain outlet 1
10-29-02 12-04-02 12-20-02 01-17-03 02-13-03 03-14-03
663 844 813 837 848 966
12.1 10.8 11.9 11.7 12.6 9.6
6.6 5.8 5.8 5.3 4.4 4.1
Tile drain outlet 2
10-29-02 12-04-02 12-20-02 01-17-03 02-13-03 03-14-03
1105 533 1462 151 518 574
9.2 10.9 8.5 4.3 9.8 10.1
3.4 4.4 2.7 14.6 5.5 3.9
Ditch 1
12-04-02 12-20-02 01-17-03 02-13-03 03-14-03
680 666 447 403 613
9.1 9.8 11.9 9.8 8.6
4.6 4.2 5.2 4.4 4.3
Ditch 2
12-04-02 12-20-02 01-17-03 02-13-03 03-14-03
522 335 287 453 423
9.3 7.9 9.3 7.4 8.2
3.6 4.5 6.5 3.6 3.9
Groundwater
11-22-02 12-17-02 01-17-03 02-13-03 03-14-03
137 880 915 369 827
1.6 0.5 0.0 3.5 1.0
1.2 0.9 2.4 2.4 0.5
Atmospheric deposition
10-04-02 10-18-02 10-23-02 10-28-02 10-30-02 01-15-03 01-29-03 03-10-03
46 36 44 17 82 89 49 139
3.7 0.9 0.9 0.9 0.9 2.4 2.4 0.4
38.0 60.7 60.7 60.7 60.7 49.1 49.1 55.6
River
01-17-03 02-13-03 03-14-03
259 135 238
8.2 9.3 9.6
6.5 6.5 5.3
ð1Þ
d OW ¼ fD d OD þ fG d OG þ fA d OA ;
ð2Þ
1 ¼ fD þ fG þ fA :
ð3Þ
The subscripts D, G and A represent the three sampled sources: drainage water (D), groundwater (G), atmospheric deposition (A), and W represents the Warnow River; f is defined as the fraction of the respective source. The nitrate isotope values (d15N and d18O) are the concentration weighted mean values.
d18 O–NO 3 ð‰Þ
Date
2.4. Mixing-model
Nitrate (lMol)
d15 N–NO 3 ð‰Þ
Sample
1338
B. Deutsch et al. / Organic Geochemistry 37 (2006) 1333–1342
In outlet 1 nitrate increased from 663 to 966 lMol during the sampling period. The nitrate drinking water limit of 806 lMol (50 mg NO3 l1) was exceeded in 59% of the drainage and groundwater samples. Mean concentrations of outlets 1, 2, and the groundwater were 829, 724, and 625 lMol, respectively. The ditches had mean concentrations of 562 (ditch 1) and 404 lMol (ditch 2). Atmospheric deposition showed the lowest concentrations with 17–139 lMol (mean 63 lMol), and there was no correlation between nitrate concentration and amount of precipitation. The Warnow River had concentrations between 135 and 259 lMol. The two tile drain outlets differed in their isotopic ratios. While the d15 N–NO 3 values in outlet 1 were in a range between 9.6& and 12.6& and the d18 O–NO 3 values slightly decreased from 6.6& to 4.1&, nitrate in tile drain outlet 2 showed higher variability in d15N (4.3–10.9&) and d18O (2.7– 14.6&). A strong increase in the d18 O–NO 3 values in combination with a strong decrease in d15N was observed on January 17. This increase in d18O was also visible in nitrate of ditch 2, but co-occurred with increasing d15 N–NO 3 values. The concentration weighted mean d15 N–NO 3 values of the tile drain outlets 1 and 2 were 11.4& and 9.2&, respectively. The concentration weighted mean d18 O–NO 3 values were 5.3& and 4.0& (Fig. 2). The ditches had concentration weighted mean d15 N–NO 3 values 60
Tile Drain 1 Tile Drain 2 Runoff Ditch 1 Runoff Ditch 2 Groundwater Atm. Deposition River
δ18O-NO3- [‰]
50
40
Table 4 Isotopic composition of the total nitrogen and oxygen as well as of the nitrate compounds of the mineral fertilizers applied on the catchment area of the tile drain outlets in 2002 and 2003 Fertilizer
d15Ntotal (&)
d15 N–NO 3 ð‰Þ
d18 N–NO 3 ð‰Þ
AHL NTS KAS (+CaO) Urea
3.8 0.1 1.7 0.7
0.8 4.4
25.7 19.4 20.5a
a
Data from Amberger and Schmidt (1987).
of 9.7& and 8.4&, and the d18 O–NO 3 values were 4.5& and 4.2&, respectively (Fig. 2). To investigate whether the application of mineral fertilizers affects the isotopic composition of the drainage water nitrate, the isotopic composition of the fertilizers used was additionally determined. d15N values of the mineral fertilizers varied between 3.8& and 4.4& and d18 O–NO 3 values between 19.4& and 25.7& (Table 4). The isotopic composition of the river nitrate was close to that of the two tile drain outlets and the ditches, with a concentration weighted mean d15N of 9.0&, and d18O of 6.0& (Fig. 2). A single-factor analysis of variance (ANOVA) of these five groups of samples indicated no significant difference (p > 0.05) for both d15N and d18O values. Groundwater nitrate had a concentration weighted mean d15N value of 0.6& and d18O of 1.4&. Nitrate in atmospheric deposition showed d15N values between 0.4& and 3.7& (cwm 0.1&) and d18O values of 49.1–60.7& (cwm 51.7&; Fig. 2). A single-factor ANOVA confirmed that three groups can be significantly distinguished from each other by means of their d15N and d18O values in nitrate (p < 0.05). These groups are both tile drain outlets, ditches and the river (group 1), groundwater (group 2), and atmospheric deposition (group 3). 4. Discussion
8
4
0 -2
0
2
4
6
8
10
12
δ15N-NO3- [‰] Fig. 2. Concentration weighted mean d18O–NO3 values vs. concentration weighted mean d15N–NO3 values including standard deviation for all samples; notice the break of the y-axis.
The tile drain outlets showed the highest nitrate concentrations of all sampling sites. This was expected because of the long-term application of mineral and organic fertilizer. The d15 N–NO 3 values in the drainage water are in a range reported for nitrate in agricultural soils with both application of mineral fertilizers as well as manure (Amberger and Schmidt, 1987; Iqbal et al., 1997). The mineral fertilizers used in our area showed typical d15N values between 3.8& and 0.7& for total N, and the
B. Deutsch et al. / Organic Geochemistry 37 (2006) 1333–1342
nitrate compound of the AHL and NTS fertilizers had 0.8& and 4.4&, respectively. Application of manure and mineral fertilizers, which contain urea and/or ammonium compounds leads to volatile loss of 15N-depleted ammonia. The remaining 15Nenriched ammonia is converted to 15N-enriched nitrate during the nitrification process (Heaton, 1986; Kendall, 1998). The d18 O–NO 3 values for both outlets were in a range given for nitrate generated during the nitrification process, which shows d18O values between 2& and 14& (Kendall, 1998; Mayer et al., 2001). d18 O–NO 3 values from 5.7& to 10.5& were reported from groundwater beneath agricultural land only treated with mineral fertilizers (Amberger and Schmidt, 1987). Mineral fertilizers are made of atmospheric O2, which has a d18O value of around 23.5& (Kroopnick and Craig, 1972), and the mineral fertilizers applied on our fields had d18 O–NO 3 values between 19.4& and 25.7&. The decrease in d18 O–NO 3 in outlet 1 and the simultaneous increase in the nitrate concentration might be the result of an increased nitrification rate during the sampling period. Nitrate that leaches from beneath agricultural fields during fall and winter mainly derives from the mineralization of soil organic matter (Lammel, 1990). The decreasing 15 d18 O–NO 3 and d N–NO3 values in outlet 1 indicate that no denitrification occurred in the soil during the sampling period. Bo¨ttcher et al. (1990) reported during the denitrification process a simul 15 taneous increase in the d18 O–NO 3 and d N–NO3 values in the residual nitrate with a ratio close to 1:2. In our study only nitrate of outlet 2 showed such an increase in the d18O and d15N values (ratio of 1:1.7) between the first and second sampling date, associated with a strong decrease in nitrate concentration. This might be an indicator for denitrification in the soil. Because of the similar soil textures and similar fertilizing practice, the greater variability in the nitrate isotopic composition and in the nitrate concentrations of outlet 2 compared to outlet 1 can only be explained by differences in the drained area. While outlet 1 reflected the drained agricultural area, outlet 2 was further supplied with surface water from two small ponds. Dilution and mixing during heavy rainfall or snowmelt events could have altered the concentrations and the isotopic composition of the nitrate. Rainwater nitrate, as well as nitrate in snow is characterized by high d18O values >30& and low d15N-values (Durka et al., 1994; Campbell et al., 2002; Piatek et al., 2005). This is probably the reason for the d18 O–NO 3 peak and
1339
the d15 N–NO 3 decrease in outlet 2 on January 17, 2003. The meteorological data reported a snowmelt event a few days before sampling, so that the decrease in the nitrate concentration and the increase in the d18 O–NO 3 might be the result of a mixture of nitrate-poor snowmelt water with low 18 d15 N–NO 3 and high d O–NO3 values with the soil water. The adjacent ditches connecting the tile drain outlets with the river showed similar nitrate isotope values to the outlets suggesting that the nitrate discharged from the drainage tiles enters the river without fractionation. The lower nitrate concentrations of the ditches compared to the tile drain outlets can be the result of dilution with nitrate poor water. Also NO 3 uptake processes that may occur without isotope fractionation cannot be excluded. These could be denitrification in sediments or assimilation by macroalgae, (Mariotti et al., 1988; Ho¨gberg, 1997). However, they seem unlikely because of the low water temperatures (Table 2). Pfenning and McMahon (1996) reported a 77% decrease in the denitrification rate during their laboratory experiments when the incubation temperature was lowered from 22 to 4 C. The low d18 O–NO 3 values between 0.5& and 2.4& indicate that the groundwater nitrate is likewise generated during the mineralization of soil organic N (Amberger and Schmidt, 1987; Kendall, 1998; Mayer et al., 2001). However, the very low d15 N–NO 3 values (0.5& to 3.5&) and the partially high nitrate concentrations (>50% of all groundwater samples exceeded the nitrate drinking water limit) could also be a result of a long-term input of mineral fertilizer nitrogen (Freyer and Aly, 1974). The samples from atmospheric deposition had the lowest nitrate concentrations of all sampled sources (17–139 lmol) and are in the range reported for other regions in Germany with similar moderate industrial pollution (UBA, 2002). Also, their isotope values (d18O = 39.0–60.7; d15N = 0.4–3.7) are typical for nitrate from atmospheric deposition (Durka et al., 1994; Campbell et al., 2002). d15 N–NO 3 values between 4.3& and 2.4& were found in atmospheric deposition collected in Ju¨lich (Germany), and a seasonal trend with higher values during fall and winter was observed (Freyer, 1991). The elevated d15N values of the river nitrate well reflect the agricultural land use of the sub basin. A positive correlation between percentage of land used for agriculture and urban purposes and mean d15N
1340
B. Deutsch et al. / Organic Geochemistry 37 (2006) 1333–1342
values of nitrate for 16 river-systems in the northeastern US is reported from Mayer et al. (2002). They determined a mean d15N value of 8.4& for the Schuylkill River with 38.4% of land used for agriculture. Although fractionation processes like denitrification and assimilation in the river cannot absolutely be excluded, the d15N and d18O values in the river nitrate do not indicate such a process. Whereas the increase in the d15 N–NO 3 value might be an indicator for assimilation or denitrification, the simultaneous decrease in the d18 O–NO 3 values argue against fractionation. We therefore assume, that fractionation of the river nitrate was not significant during the sampling period. The mixing-model indicated that the nitrate in the sub basin ‘Middle Warnow’ mainly consists of drainage water (86%) and groundwater (11%) (Table 5), although both sources showed similar mean nitrate concentrations. The atmospheric deposition had a share of 3% in the river nitrate. These data are compared with good agreement to estimates of a modified version of the MONERIS model (Modelling Nutrient Emissions in River Systems; Behrendt et al., 1999) for the period 1995–1999 (Table 6; Pagenkopf, 2001). MONERIS estimates N inputs from point sources and six diffuse pathways, which are drainage water, ground-
Table 5 and Concentration weighted mean values of d15 N–NO 3 d18 O–NO 3 used in the mixing-model and the resulting fractions of each source with standard error Source
d15 N–NO 3 ð‰Þ
d18 N–NO 3 ð‰Þ
Fraction [f]
Drainage water Groundwater Rain Mixture River
10.4 0.6 0.1
4.7 1.4 51.7
0.86 ± 0.07 0.11 ± 0.07 0.03 ± 0.02
9.0
6.0
Table 6 Comparison of the estimated proportion of each source on the Warnow River nitrate in the sub basin ‘Middle Warnow’, given from the modified MONERIS model (Pagenkopf, 2001) and this study Pathways
Percentage of input MONERIS
This study
Tile drain outlets Groundwater Atmospheric deposition
80 15 5
86 ± 7 11 ± 7 3±2
water, atmospheric deposition, erosion, urban areas, and surface runoff. To allow a comparison of our results with the model data, the estimated inputs from drainage water, groundwater, and atmospheric deposition given by the modified MONERIS model were added together and set as 100%. Although the MONERIS model estimates inputs of total dissolved inorganic nitrogen, we think that a comparison with our estimation of nitrate inputs is maintainable, since most of the inorganic nitrogen is emitted to the aquatic ecosystems in the form of nitrate (Addiscott et al., 1992). Furthermore, our estimate only represents the winter period, whereas the model calculates annual mean values. Several studies have shown that the main portion of nitrate leaches from agricultural soils during fall and winter, when most nitrate derives from mineralization of soil organic N (Kirchmann et al., 2002). Combined with the high amount of water discharge due to low evapotranspiration and high rainfalls during winter and fall, more nitrate leaches to the drainage system (Lammel, 1990). According to the MMK data our investigated agricultural soils (water saturated sandy loams and sands) account for 11% of the total agricultural area in the sub basin ‘Middle Warnow’, which is dominated by water saturated loams (50%) and peat soils (20%). It cannot be excluded that the drainage water from loam and peat soils show elevated d15 N–NO 3 and d18 O–NO 3 values because of a higher influence of denitrification, but results from a second study carried out on agricultural fields in the same sub basin showed similar isotope values in drainage water nitrate (Deutsch et al., unpublished data). Uncertainty in source attribution mainly exists for the share of groundwater nitrate, since only one aquifer was sampled. Only a 2& enrichment 18 in the d15 N–NO 3 and d O–NO3 values as reported by Bo¨ttcher et al. (1990) during denitrification, would increase the contribution of the groundwater nitrate from 11% to 15%. 5. Conclusions Our study shows that the stable isotope ratios of nitrogen and oxygen in nitrate can be used to quantify the sources of nitrate to a river. Given that the various sources differ substantially in their isotopic composition the mixing-model approach provides a trustworthy estimate for the contribution of the different sources. The choice of the winter season
B. Deutsch et al. / Organic Geochemistry 37 (2006) 1333–1342
for sampling may have supported the quality of our results – that are based on only few samples – since low temperatures reduce microbial fractionation processes. Acknowledgements This work was part of the project ‘‘The southern Baltic Sea and its coast in change’’, supported by the German Federal Environmental Foundation. The authors thank K. Hennig from the Eurawasser GmbH, R. Bo¨rner from the StAUN Rostock and R. Kolbe from the LUNG Gu¨strow for their cooperativeness and support. Special thanks to Bernhard Mayer, Katrin Schmidt and Angus Atkinson for their helpful and critical comments. References Addiscott, T.M., Whitmore, A.P., Powlson, D.S., 1992. Farming, Fertilizers and the Nitrate Problem. CAB International, Wallingford. Amberger, A., Schmidt, H.-L., 1987. Natu¨rliche Isotopengehalte von Nitrat als Indikatoren fu¨r dessen Herkunft. Geochimica et Cosmochimica Acta 51, 2699–2705. Behrendt, H., Huber, P., Kornmilch, M., Opitz, D., Schmoll, O., Scholz, G., Uebe, R., 1999. Na¨hrstoffbilanzierung der Flussgebiete Deutschlands. Umweltbundesamt, Berlin. Bockholt, R., Kappes, K., 1994. Stoffaustrag u¨ber Entwa¨sserungssysteme - ein aktueller Niedermoorgru¨nland ¨ kologische Untersuchungen im Ackerland - Vergleich, O Warnow-Einzugsgebiet II, 2/1994, LAUN M-V, pp. 136–146. Bo¨rner, R., Bo¨nsch, R., Fadschild, K., Gosselck, F., Hu¨bener, T., Klinkenberg, G., Kolbow, D., Lill, D., Merkel, G., Neumann, C., Randow, F.F.E., Schlungbaum, G., Selig, U., Winkler, H., 1994. Ein Beitrag zur Biologie der Warnow, eines ¨ kologische norddeutschen nacheiszeitlichen Tieflandflusses, O Untersuchungen im Warnow-Einzugsgebiet II, Heft 2, Schriftenreihe des Landesamtes fu¨r Umwelt und Natur Mecklenburg-Vorpommern, Landesamt fu¨r Umwelt und Natur Mecklenburg-Vorpommern, pp. 56–92. Bo¨ttcher, J., Strebel, O., Voerkelius, S., Schmidt, H.-L., 1990. Using isotope fractionation of nitrate–nitrogen and nitrate– oxygen for evaluation of microbial denitrification in a sandy aquifer. Journal of Hydrology 114, 413–424. Campbell, D.H., Kendall, C., Chang, C.C.Y., Silva, S.R., Tonnessen, K.A., 2002. Pathways for nitrate release from an alpine watershed: determination using d15N and d18O. Water Resources Research 38, 10-1–10-9. Chang, C.C.Y., Kendall, C., Silva, S.R., Battaglin, W.A., Campbell, D.H., 2002. Nitrate stable isotopes: tools for determining nitrate sources among different land uses in the Mississippi River Basin. Canadian Journal of Fisheries and Aquatic Sciences 59, 1874–1885. Durka, W., Schulze, E.-D., Gebauer, G., Voerkelius, S., 1994. Effects of forest decline on uptake and leaching of deposited nitrate determined from 15N and 18O measurements. Nature 372, 765–767.
1341
Flipse, W.J., Bonner, F.T., 1985. Nitrogen-isotope ratios of nitrate in ground water under fertilized fields, Long Island, New York. Groundwater 23, 59–67. Freyer, H.D., Aly, A.I.M., 1974. Nitrogen-15 variations in fertilizer nitrogen. Journal Environmental Quality 3, 405–406. Freyer, H.D., 1991. Seasonal variation of 15N/14N ratios in atmospheric nitrate species. Tellus 43, 30–44. Grasshoff, K., Kremling, K., Ehrhardt, M., 1999. Methods of Seawater Analysis, third ed. Wiley VCH, Weinheim. Heaton, T.H.E., 1986. Isotopic studies of nitrogen pollution in the hydrosphere and atmosphere: a review. Chemical Geology 59, 87–102. HELCOM, 2003. Executive summary of the Fourth Baltic Sea Pollution Load Compilation PLC-4. HELCOM, Helsinki. Ho¨gberg, P., 1997. 15N natural abundance in soil–plant systems. New Phytologist 137, 179–203. Iqbal, M.Z., Krothe, N.C., Spalding, R.F., 1997. Nitrogen isotope indicators of seasonal source variability to groundwater. Environmental Geology 32, 210–218. Kendall, C., 1998. Tracing nitrogen sources and cycles in catchments. In: Kendall, C., McDonnell, J.J. (Eds.), Isotope Tracers in Catchment Hydrology. Elsevier, Amsterdam, pp. 519–576. Kendall, C., Aravena, R., 1999. Nitrate isotopes in groundwater systems. In: Cook, P.G., Herozeg, A.L. (Eds.), Environmental Tracers in Subsurface Hydrology. Kluwer Academic, Boston, pp. 261–297. Kirchmann, H., Johnston, A.E.J., Bergstro¨m, L.F., 2002. Possibilities for reducing nitrate leaching from agricultural land. Ambio 31, 404–408. Kroopnick, P.M., Craig, H., 1972. Atmospheric oxygen: isotopic composition and solubility fractionation. Science 175, 54–55. Lammel, J., 1990. Na¨hrstoffaustra¨ge durch Vorfluter und Dra¨ne. Landwirtschaftsverlag GmbH, Mu¨nster-Hiltrup. Mariotti, A., Landreau, A., Simon, B., 1988. 15N isotope biogeochemistry and natural denitrification process in groundwater: application to the chalk aquifer of Northern France. Geochimica et Cosmochimica Acta 52, 1869–1878. Mayer, B., Bollwerk, S.M., Mansfeldt, T., Hu¨tter, B., Veizer, J., 2001. The oxygen isotope composition of nitrate generated by nitrification in acid forest floors. Geochimica et Cosmochimica Acta 65, 2743–2756. Mayer, B., Boyer, E.W., Goodale, C., Jaworski, N.A., van Breemen, N., Howarth, R.W., Seitzinger, S., Billen, G., Lajtha, K., Nadelhoffer, K., van Dam, D., Hetling, L.J., Nosal, M., Paustian, K., 2002. Sources of nitrate in rivers draining sixteen watersheds in the northeastern US: isotopic constraints. Biogeochemistry 57/58, 171–197. Meybeck, M., 1982. Carbon, nitrogen, and phosphorous transport by world rivers. The American Journal of Science 282, 401–450. Pagenkopf, W., 2001. Aktuelle Na¨hrstoffbilanzierung fu¨r Teilgebiete des Einzugsgebiets der Warnow. Geodaten Integration & Analyse, Berlin. Pfenning, K.S., McMahon, P.B., 1996. Effect of nitrate, organic carbon, and temperature on potential denitrification rates in nitrate-rich riverbed sediments. Journal of Hydrology 187, 283–295. Phillips, D.L., Koch, P.L., 2002. Incorporating concentration dependence in stable isotope mixing models. Oecologica 130, 114–125.
1342
B. Deutsch et al. / Organic Geochemistry 37 (2006) 1333–1342
Piatek, K.B., Mitchell, M.J., Silva, S.R., Kendall, C., 2005. Sources of nitrate in snowmelt discharge: evidence from water chemistry and stable isotopes of nitrate. Water, Air, and Soil Pollution 165, 13–35. Silva, S.R., Kendall, C., Wilkison, D.H., Ziegler, A.C., Chang, C.C.Y., Avanzino, R.J., 2000. A new method for collection of nitrate from fresh water and the analysis of nitrogen and oxygen isotope ratios. Journal of Hydrology 228, 22–36. Spoelstra, J., Schiff, S.L., Elgood, R.J., Semkin, R.G., Jeffries, D.S., 2001. Tracing the sources of exported nitrate in the Turkey Lakes Watershed using 15N/14N and 18O/16O isotopic ratios. Ecosystems 4, 536–544. Sta˚lnacke, P., Grimvall, A., Sundblad, K., Tonderski, A., 1999. Estimation of riverine loads of nitrogen and phosphorous to
the Baltic Sea, 1970–1993. Environmental Monitoring and Assessment 58, 173–200. Stepanauskas, R., Leonardson, L., 1999. Bioavailability of wetland-derived DON to freshwater and marine bacterioplankton. Limnology and Oceanography 44, 1477–1485. Thiele, V., Mehl, D., 1995. Ein o¨kologisch begru¨ndetes Sanierungskonzept fu¨r das Gewa¨ssereinzugsgebiet der Warnow (Mecklenburg-Vorpommern). Landesamt fu¨r Umwelt und Natur Mecklenburg-Vorpommern, Gu¨strow. UBA, 2002. Jahresbericht 2001 aus dem Messnetz des Umweltbundesamtes. Umweltbundesamt, Berlin. Wassenaar, L.I., 1995. Evaluation of the origin and fate of nitrate in the Abbotsford Aquifer using the isotopes of 15N and 18O in NO 3 . Applied Geochemistry 10, 391–405.
Quantification of diffuse nitrate inputs into a small river system using stable isotopes of oxygen and nitrogen in nitrate Barbara Deutsch
a,*
, Melanie Mewes b, Iris Liskow a, Maren Voss
a
a
b
Baltic Sea Research Institute, Seestr. 15, 18119 Rostock, Germany Chair of Landscape Economics, University of Greifswald, Grimmer Str. 88, 17487 Greifswald, Germany Received 31 August 2005; accepted 14 April 2006 Available online 9 August 2006
Abstract To identify and quantify diffuse nitrate inputs into a river sub basin in Mecklenburg-Vorpommern (Germany) a dualisotope approach with d15N and d18O in nitrate was carried out from October 2002 to March 2003. Three nitrate sources (water from artificially drained agricultural soils, groundwater, atmospheric deposition) and the river were sampled monthly to bimonthly. Nitrate of the drainage water had a concentration weighted mean (cwm) d15N value of 10.4& and d18O of 4.7&, and was significantly different to the groundwater nitrate (cwm d15N = 0.6&; d18O = 1.4&). The low d18O values indicated that most of the nitrate from these sources was formed during the nitrification process of soil organic N. Nitrate of atmospheric deposition had a cwm d15N value of 0.1& and a d18O value of 51.7&. River nitrate showed cwm values of 9.0& in d15N and 6.0& in d18O close to the isotope values of the drainage water nitrate. The isotope values were used in a three source mixing-model, to determine the contribution of each sampled nitrate source to the total river nitrate. The mixing-model revealed that the nitrate from the drainage water contributed 86% of the river nitrate. Contribution of nitrate from groundwater and atmospheric deposition was 11% and 3%, respectively. These results agree with estimations of nitrate input data for this sub basin given by a nutrient emissions model. 2006 Elsevier Ltd. All rights reserved.
1. Introduction Although inputs of nutrients have been reduced in the recent years, the excessive load of nitrogen and phosphorous is still one of the major ecological problems of the Baltic Sea (Sta˚lnacke et al., 1999). In 2000 an amount of 814 · 103 t nitrogen entered the Baltic Sea, and more than 84% of the total N *
Corresponding author. Tel.: +49 381 5197 417; fax: +49 381 5197 440. E-mail address: [email protected] (B. Deutsch).
load derived from rivers (HELCOM, 2003). Worldwide, almost 70% of the riverine N load consists of dissolved organic nitrogen (DON; Meybeck, 1982), but experiments showed that its bioavailability may be as low as 2–16% (Stepanauskas and Leonardson, 1999), while nitrate is the dominant inorganic N-species lost to the aquatic environment (Addiscott et al., 1992) and is rapidly consumed. Diffuse nitrate inputs such as fertilizer runoff from farmland, atmospheric deposition and groundwater input are especially hard to identify, because they are emitted over large areas. Nowadays, a powerful tool to distinguish different nitrate sources is the
0146-6380/$ - see front matter 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.orggeochem.2006.04.012
1334
B. Deutsch et al. / Organic Geochemistry 37 (2006) 1333–1342
determination of stable isotope ratios of nitrogen and oxygen (Wassenaar, 1995; Spoelstra et al., 2001; Chang et al., 2002; Mayer et al., 2002; Piatek et al., 2005). The d15N values of nitrate from different sources often show overlapping ranges, but the additional measurement of the d18O values allows a more precise classification (Mayer et al., 2002). Nitrate, derived from sewage and manure, is isotopically distinct from atmospheric nitrate in d15N (7–20&, 10& to 8&, respectively), as well as in d18O (<15& compared to 25–75&; Wassenaar, 1995; Kendall, 1998). Nitrate originating from mineral fertilizers shows d15N values of 0 ± 4& (Kendall, 1998), and d18O values of 22 ± 3& (Amberger and Schmidt, 1987) because of their production from atmospheric nitrogen (d15N = 0&) and oxygen (d18O = 23.5&). However, the isotopic composition of nitrate collected in drainage tiles and runoff ditches does not reflect exactly the isotope values of the fertilizer applied, but is altered due to isotope fractionation processes (Kendall and Aravena, 1999). Flipse and Bonner (1985) demonstrated that groundwater nitrate produced under fertilized fields showed d15N values up to 12.4& higher than the fertilizers applied, and explained this difference as the volatile loss of ammonia from the fertilizer, containing reduced nitrogen forms. Another fractionation process is denitrification, which increases the d15N and d18O values of the residual nitrate with an enrichment of d18O:d15N close to 1:2 (Bo¨ttcher et al., 1990). Nitrate uptake by plants (Ho¨gberg, 1997) and soil N-mineralization, including ammonification and nitrification may also modify the isotope signature of nitrate (Iqbal et al., 1997; Mayer et al., 2001). The objective of this study was to test whether the isotopic composition of nitrate from three diffuse nitrate sources can be used to quantify the diffuse nitrate inputs into a sub basin of the Warnow River (444 km2), which also supplies the city of Rostock (200,000 inhabitants) with drinking water. The three nitrate sources, drainage water from fertilized fields, groundwater, and atmospheric deposition, as well as the river itself were sampled regularly from October 2002 to March 2003. The samples were analysed for nitrate concentration, d15N and d18O values. The data were used in a three source mixing-model (Phillips and Koch, 2002) to determine the percentage of every sampled nitrate source to the river nitrate. For the successful application of a conservative isotope mixing-model, it is necessary that the isotope values of the river nitrate
are not altered due to fractionation processes. For this purpose, sampling was carried out during late fall and winter. At low water temperature microbial activity is reduced (Pfenning and McMahon, 1996) and therefore alteration of the river nitrate isotope values due to fractionation processes is minimal. The usefulness of this approach as an additional method for nitrate source quantification beside model estimations is discussed. 2. Material and methods 2.1. Study area The Warnow River is located in MecklenburgVorpommern (north eastern Germany) and flows into the Southern Baltic Sea at the City of Rostock (Fig. 1a and b). With a length of 149 km and a drainage area of 3270 km2, mainly dominated by agricultural and forested areas (63% and 24% respectively; Pagenkopf, 2001) it is the second largest river system of Mecklenburg-Vorpommern (Thiele and Mehl, 1995). The Warnow River is characterized as polytrophic (Bo¨rner et al., 1994) due to a high amount of nutrient inputs (total nitrogen: 4140 t yr1, phosphorous: 200 t yr1), and a production rate of 650 g C m2 yr1. The riversystem is divided into seven sub basins, and the investigated sub basin ‘Middle Warnow’ covers an area of 444 km2. Land use in the sub basin is dominated by agriculture, with arable land (49%) and pasture (23%), followed by forests (17%), urban areas (8%), and waters (3%; Pagenkopf, 2001). According to the Mesoscale Agricultural Mapping Programme (MMK), which provides information about soil textures, 75% of the agricultural land in the sub basin ‘Middle Warnow’ are waterlogged soils. Estimations showed that up to 50% of the agricultural area might be artificially drained (Bockholt and Kappes, 1994). 2.2. Sampling From November 2002 to April 2003 two tile drain outlets, located on the area of a farmers’ cooperative near Rostock (Fig. 1c) were sampled monthly to bimonthly (five samples). The distance between the tile drains and the Warnow River is approx. 2 km. The drainage system is located at a soil depth of 60–100 cm. The soil textures are mainly sandy loams (46%) and sands (24%) from moraine substrates. The dominant soil types are
B. Deutsch et al. / Organic Geochemistry 37 (2006) 1333–1342
1335
Fig. 1. Site map of the Warnow River system (a) located at the south western coast of the Baltic sea (b). The dark grey area shows the sub basin ‘Middle Warnow’. Numbers in circles indicate the sampling sites (1: tile drain outlets, ditches, and river sampling; 2: groundwater sampling; 3: sampling of atmospheric deposition in Rostock-Warnemuende). Map (c) shows the tile drain outlets, ditches, and the river sampling site in more detail.
cambisols, luvisols and gleysols. The drained areas are approximately 0.025 km2 (outlet 1) and 0.15 km2 (outlet 2), respectively. Outlet 2 was also supplied with surface water of two small ponds located in the field. Cultivated crops were winter wheat/winter barley in 2002 and sugar beets/corn in 2003 (outlet 1) and for tile drain outlet 2 winter wheat in 2002 and winter barley in 2003. Fertilizer application and the composition of fertilizer used are shown in Table 1. Application of manure on both areas takes place every fifth to sixth year, and the last application on the soil above outlet 2 was in autumn 1998 with a similar amount as applied above tile 1 in 2002. Two adjacent ditches (ditch 1 and 2; Fig. 1c), connecting the field runoff
with the Warnow River were sampled additionally, to examine the possible alteration of the isotopic composition of nitrate due to isotope fractionation processes during the passage of the drainage water to the river. Three samples were taken from the Warnow River from January 2003 to March 2003 close to the mouth of the adjacent ditches (Fig. 1c). Sampling took place from a landing stage at a maximum depth of 0.3 m. Water temperatures measured close to the sampling location decreased from 8.6 C (October 2002) to 0.1 C (December 2002) and then increased to 6.5 C (March 2003; Table 2; unpublished data, Office for the Environment and Nature, Rostock). Groundwater samples were taken from a
1336
B. Deutsch et al. / Organic Geochemistry 37 (2006) 1333–1342
Table 1 Application date, amount, and composition of the fertilizers used on the soils above the sampled drainage tiles Date
Amount applied (kg N ha1)
Fertilizer
Nitrogen compounds
Tile drain outlet 1 03-25-2002 04-23-2002 05-29-2002 August 2002
85 70 79 232
AHL Urea KAS Manure
28% N ð7% NHþ 4 ; 7% NO3 ; 14% amideÞ 46% N (46% amide) 27% N ð13:5% NHþ 4 ; 13:5% NO3 Þ
Tile drain outlet 2 03-27-2002 04-22-2002 05-28-2002 03-18-2003 04-28-2003
66 61 148 79 90
AHL NTS NTS AHL Urea
28% 27% 27% 28% 46%
Table 2 Water temperatures of the Warnow River measured close to the sampling location during the sampling period Date
Temperature (C)
10-28-02 11-12-02 11-27-02 12-10-02 01-09-02 01-22-03 02-04-03 02-19-03 03-05-03 03-18-03
8.6 5.0 5.7 0.1 1.3 3.7 1.5 1.8 3.5 6.5
sampling site near the village of Reez close to a wood and arable land (Fig. 1a) from November 2002 to March 2003 (five samples). The well had a depth of 8.4 m and sampling was carried out with a submerged pump at a maximum discharge flow of 5 l min1. The aquifer had a thickness of approx. 5 m and there was a direct discharge into the river (K. Hennig, Eurawasser GmbH, pers. com.). Eight samples of precipitation were collected from October 2002 to April 2003 at the German Meterological Service in Rostock-Warnemuende, 15 km north of the study area (Fig. 1a). Samplers were funnel shaped (diameter 24 cm) with a vial below. After every rainfall they were immediately emptied and subsampled for nitrate analysis. Because of low nitrate concentrations samples of consecutive rainfall events were combined for isotope analysis. 2.3. Laboratory work After sampling, the water was filtered through a 0.45 lm membrane filter. The determination of
N N N N N
ð7% NHþ 4 ; 7% ð6% NHþ 4 ; 8% ð6% NHþ 4 ; 8% ð7% NHþ 4 ; 7% (46% amide)
NO 3 ; 14% NO 3 ; 13% NO 3 ; 13% NO 3 ; 14%
amideÞ amideÞ amideÞ amideÞ
nitrate concentration was carried out after Grasshoff et al. (1999). Subsequently, samples were prepared for isotopic analysis of d15N and d18O in nitrate using the method of Silva et al. (2000). In brief, samples were passed through a cation exchange resin (5 ml AG 50W-X4, H+-form; Biorad), followed by an anion exchange resin (2 ml AG1-X8, Cl-form; Biorad). At least 60 lMol nitrate was finally collected on the anion exchange resin. The resins with the absorbed nitrate were stored in a refrigerator for several weeks until further preparation. For further processing samples were eluted from the anion exchange resin with 15 ml of 3 M HCl, and neutralized with 6 g Ag2O to obtain a pH of 5.5–6. The precipitated AgCl and remaining Ag2O was removed by filtration (0.45 lm membrane filter). Two milliliters of 1 M BaCl solution was added to the filtrate to remove SO2 and PO3 4 4 . Precipitated BaSO4 and Ba3(PO4)2 were removed by filtration (0.45 lm membrane filter). The sample was passed through a cation exchange resin (5 ml AG 50W-X4, H+-form; Biorad) to eliminate the excess Ba2+. Then a second neutralization with Ag2O (1–2 g) was carried out and the resulting AgCl and excess Ag2O were removed by filtration (0.45 lm membrane filter). The solution, now containing Ag+ and NO 3 , was freezedried and the remaining solid AgNO3 was weighed into silvercaps for determination of d15N and d18O values of nitrate. The d18O values were determined with a Thermo Finnigan Delta Plus isotope ratio mass spectrometer (IRMS) after pyrolysis in a Thermo Finnigan TC/EA. Temperature of pyrolysis was 1350 C. The isotope ratios are given as d values in per mil (&) relative to a standard and were calculated after the following equation:
B. Deutsch et al. / Organic Geochemistry 37 (2006) 1333–1342
1337
d18 O ¼ ð18 O=16 Osample =18 O=16 Ostandard Þ 1 1000:
3. Results
The same equation is used for calculation of the d15N values. For d18O the international standard is Vienna standard mean ocean water (VSMOW) defined as 0&. Calibration of the reference gas CO was done with IAEA-KNO3 (d18O = 25.1 ± 0.6&; n = 51) and IAEA-C3-Cellulose (d18O = 32.2 ± 0.2&; n = 38). Precision of the measurement was verified by repeated analysis of an internal laboratory standard. The standard was KNO3 from Merck with the lot No. A286463 (d18O = 24.6 ± 0.7&; n = 52). Samples for d15N were also analysed in a Thermo Finnigan Delta Plus IRMS after combustion in a Flash EA at a temperature of 1020 C vs. air nitrogen. International reference materials used were IAEA-N1 (d15N = 0.4 ± 0.07&) and IAEA-N2 (d15N = 20.3 ± 0.09&), the internal laboratory standards were acetanilide (d15N = 1.7 ± 0.2&; n = 244), peptone (d15N = 5.7 ± 0.2&; n = 245) and Merck-KNO3 (d15N = 0.4 ± 0.16%; n = 20). Reproducibility of the preparation procedure was tested with AgNO3, which was prepared from an internal lab standard along with the samples. The standard was Merck-KNO3 ð1000 lMol 18 NO 3 =lÞ with SO4 (620 lMol). d O values of the resulting AgNO3 were 23.4 ± 0.7& (n = 16) and d15N = 0.4 ± 0.2& (n = 16). Since the d15N and d18O values of the produced AgNO3 were in the range of the standard deviations of the d values of the original Merck-KNO3, the isotope values of the prepared samples were not corrected.
The nitrate concentrations, d15N, and d18 O–NO 3 values of all samples are summarized in Table 3. The highest nitrate concentrations were found in the two tile drain outlets and in the groundwater, with a maximum value of 1462 lMol in outlet 2 at the end of December 2002. Outlet 2 and the groundwater showed high variability in nitrate concentrations, with no visible upward or downward trend. Table 3 Nitrate concentration, d15N, and d18 O–NO 3 values of the tile drain outlets, ditches, groundwater, atmospheric deposition, and the river
To estimate the contribution of the nitrate sources to the Warnow River a mixing-model based on mass balance equations was used (Phillips and Koch, 2002). The equations are d15 NW ¼ fD d15 ND þ fG d15 NG þ fA d15 NA ; 18
18
18
18
Tile drain outlet 1
10-29-02 12-04-02 12-20-02 01-17-03 02-13-03 03-14-03
663 844 813 837 848 966
12.1 10.8 11.9 11.7 12.6 9.6
6.6 5.8 5.8 5.3 4.4 4.1
Tile drain outlet 2
10-29-02 12-04-02 12-20-02 01-17-03 02-13-03 03-14-03
1105 533 1462 151 518 574
9.2 10.9 8.5 4.3 9.8 10.1
3.4 4.4 2.7 14.6 5.5 3.9
Ditch 1
12-04-02 12-20-02 01-17-03 02-13-03 03-14-03
680 666 447 403 613
9.1 9.8 11.9 9.8 8.6
4.6 4.2 5.2 4.4 4.3
Ditch 2
12-04-02 12-20-02 01-17-03 02-13-03 03-14-03
522 335 287 453 423
9.3 7.9 9.3 7.4 8.2
3.6 4.5 6.5 3.6 3.9
Groundwater
11-22-02 12-17-02 01-17-03 02-13-03 03-14-03
137 880 915 369 827
1.6 0.5 0.0 3.5 1.0
1.2 0.9 2.4 2.4 0.5
Atmospheric deposition
10-04-02 10-18-02 10-23-02 10-28-02 10-30-02 01-15-03 01-29-03 03-10-03
46 36 44 17 82 89 49 139
3.7 0.9 0.9 0.9 0.9 2.4 2.4 0.4
38.0 60.7 60.7 60.7 60.7 49.1 49.1 55.6
River
01-17-03 02-13-03 03-14-03
259 135 238
8.2 9.3 9.6
6.5 6.5 5.3
ð1Þ
d OW ¼ fD d OD þ fG d OG þ fA d OA ;
ð2Þ
1 ¼ fD þ fG þ fA :
ð3Þ
The subscripts D, G and A represent the three sampled sources: drainage water (D), groundwater (G), atmospheric deposition (A), and W represents the Warnow River; f is defined as the fraction of the respective source. The nitrate isotope values (d15N and d18O) are the concentration weighted mean values.
d18 O–NO 3 ð‰Þ
Date
2.4. Mixing-model
Nitrate (lMol)
d15 N–NO 3 ð‰Þ
Sample
1338
B. Deutsch et al. / Organic Geochemistry 37 (2006) 1333–1342
In outlet 1 nitrate increased from 663 to 966 lMol during the sampling period. The nitrate drinking water limit of 806 lMol (50 mg NO3 l1) was exceeded in 59% of the drainage and groundwater samples. Mean concentrations of outlets 1, 2, and the groundwater were 829, 724, and 625 lMol, respectively. The ditches had mean concentrations of 562 (ditch 1) and 404 lMol (ditch 2). Atmospheric deposition showed the lowest concentrations with 17–139 lMol (mean 63 lMol), and there was no correlation between nitrate concentration and amount of precipitation. The Warnow River had concentrations between 135 and 259 lMol. The two tile drain outlets differed in their isotopic ratios. While the d15 N–NO 3 values in outlet 1 were in a range between 9.6& and 12.6& and the d18 O–NO 3 values slightly decreased from 6.6& to 4.1&, nitrate in tile drain outlet 2 showed higher variability in d15N (4.3–10.9&) and d18O (2.7– 14.6&). A strong increase in the d18 O–NO 3 values in combination with a strong decrease in d15N was observed on January 17. This increase in d18O was also visible in nitrate of ditch 2, but co-occurred with increasing d15 N–NO 3 values. The concentration weighted mean d15 N–NO 3 values of the tile drain outlets 1 and 2 were 11.4& and 9.2&, respectively. The concentration weighted mean d18 O–NO 3 values were 5.3& and 4.0& (Fig. 2). The ditches had concentration weighted mean d15 N–NO 3 values 60
Tile Drain 1 Tile Drain 2 Runoff Ditch 1 Runoff Ditch 2 Groundwater Atm. Deposition River
δ18O-NO3- [‰]
50
40
Table 4 Isotopic composition of the total nitrogen and oxygen as well as of the nitrate compounds of the mineral fertilizers applied on the catchment area of the tile drain outlets in 2002 and 2003 Fertilizer
d15Ntotal (&)
d15 N–NO 3 ð‰Þ
d18 N–NO 3 ð‰Þ
AHL NTS KAS (+CaO) Urea
3.8 0.1 1.7 0.7
0.8 4.4
25.7 19.4 20.5a
a
Data from Amberger and Schmidt (1987).
of 9.7& and 8.4&, and the d18 O–NO 3 values were 4.5& and 4.2&, respectively (Fig. 2). To investigate whether the application of mineral fertilizers affects the isotopic composition of the drainage water nitrate, the isotopic composition of the fertilizers used was additionally determined. d15N values of the mineral fertilizers varied between 3.8& and 4.4& and d18 O–NO 3 values between 19.4& and 25.7& (Table 4). The isotopic composition of the river nitrate was close to that of the two tile drain outlets and the ditches, with a concentration weighted mean d15N of 9.0&, and d18O of 6.0& (Fig. 2). A single-factor analysis of variance (ANOVA) of these five groups of samples indicated no significant difference (p > 0.05) for both d15N and d18O values. Groundwater nitrate had a concentration weighted mean d15N value of 0.6& and d18O of 1.4&. Nitrate in atmospheric deposition showed d15N values between 0.4& and 3.7& (cwm 0.1&) and d18O values of 49.1–60.7& (cwm 51.7&; Fig. 2). A single-factor ANOVA confirmed that three groups can be significantly distinguished from each other by means of their d15N and d18O values in nitrate (p < 0.05). These groups are both tile drain outlets, ditches and the river (group 1), groundwater (group 2), and atmospheric deposition (group 3). 4. Discussion
8
4
0 -2
0
2
4
6
8
10
12
δ15N-NO3- [‰] Fig. 2. Concentration weighted mean d18O–NO3 values vs. concentration weighted mean d15N–NO3 values including standard deviation for all samples; notice the break of the y-axis.
The tile drain outlets showed the highest nitrate concentrations of all sampling sites. This was expected because of the long-term application of mineral and organic fertilizer. The d15 N–NO 3 values in the drainage water are in a range reported for nitrate in agricultural soils with both application of mineral fertilizers as well as manure (Amberger and Schmidt, 1987; Iqbal et al., 1997). The mineral fertilizers used in our area showed typical d15N values between 3.8& and 0.7& for total N, and the
B. Deutsch et al. / Organic Geochemistry 37 (2006) 1333–1342
nitrate compound of the AHL and NTS fertilizers had 0.8& and 4.4&, respectively. Application of manure and mineral fertilizers, which contain urea and/or ammonium compounds leads to volatile loss of 15N-depleted ammonia. The remaining 15Nenriched ammonia is converted to 15N-enriched nitrate during the nitrification process (Heaton, 1986; Kendall, 1998). The d18 O–NO 3 values for both outlets were in a range given for nitrate generated during the nitrification process, which shows d18O values between 2& and 14& (Kendall, 1998; Mayer et al., 2001). d18 O–NO 3 values from 5.7& to 10.5& were reported from groundwater beneath agricultural land only treated with mineral fertilizers (Amberger and Schmidt, 1987). Mineral fertilizers are made of atmospheric O2, which has a d18O value of around 23.5& (Kroopnick and Craig, 1972), and the mineral fertilizers applied on our fields had d18 O–NO 3 values between 19.4& and 25.7&. The decrease in d18 O–NO 3 in outlet 1 and the simultaneous increase in the nitrate concentration might be the result of an increased nitrification rate during the sampling period. Nitrate that leaches from beneath agricultural fields during fall and winter mainly derives from the mineralization of soil organic matter (Lammel, 1990). The decreasing 15 d18 O–NO 3 and d N–NO3 values in outlet 1 indicate that no denitrification occurred in the soil during the sampling period. Bo¨ttcher et al. (1990) reported during the denitrification process a simul 15 taneous increase in the d18 O–NO 3 and d N–NO3 values in the residual nitrate with a ratio close to 1:2. In our study only nitrate of outlet 2 showed such an increase in the d18O and d15N values (ratio of 1:1.7) between the first and second sampling date, associated with a strong decrease in nitrate concentration. This might be an indicator for denitrification in the soil. Because of the similar soil textures and similar fertilizing practice, the greater variability in the nitrate isotopic composition and in the nitrate concentrations of outlet 2 compared to outlet 1 can only be explained by differences in the drained area. While outlet 1 reflected the drained agricultural area, outlet 2 was further supplied with surface water from two small ponds. Dilution and mixing during heavy rainfall or snowmelt events could have altered the concentrations and the isotopic composition of the nitrate. Rainwater nitrate, as well as nitrate in snow is characterized by high d18O values >30& and low d15N-values (Durka et al., 1994; Campbell et al., 2002; Piatek et al., 2005). This is probably the reason for the d18 O–NO 3 peak and
1339
the d15 N–NO 3 decrease in outlet 2 on January 17, 2003. The meteorological data reported a snowmelt event a few days before sampling, so that the decrease in the nitrate concentration and the increase in the d18 O–NO 3 might be the result of a mixture of nitrate-poor snowmelt water with low 18 d15 N–NO 3 and high d O–NO3 values with the soil water. The adjacent ditches connecting the tile drain outlets with the river showed similar nitrate isotope values to the outlets suggesting that the nitrate discharged from the drainage tiles enters the river without fractionation. The lower nitrate concentrations of the ditches compared to the tile drain outlets can be the result of dilution with nitrate poor water. Also NO 3 uptake processes that may occur without isotope fractionation cannot be excluded. These could be denitrification in sediments or assimilation by macroalgae, (Mariotti et al., 1988; Ho¨gberg, 1997). However, they seem unlikely because of the low water temperatures (Table 2). Pfenning and McMahon (1996) reported a 77% decrease in the denitrification rate during their laboratory experiments when the incubation temperature was lowered from 22 to 4 C. The low d18 O–NO 3 values between 0.5& and 2.4& indicate that the groundwater nitrate is likewise generated during the mineralization of soil organic N (Amberger and Schmidt, 1987; Kendall, 1998; Mayer et al., 2001). However, the very low d15 N–NO 3 values (0.5& to 3.5&) and the partially high nitrate concentrations (>50% of all groundwater samples exceeded the nitrate drinking water limit) could also be a result of a long-term input of mineral fertilizer nitrogen (Freyer and Aly, 1974). The samples from atmospheric deposition had the lowest nitrate concentrations of all sampled sources (17–139 lmol) and are in the range reported for other regions in Germany with similar moderate industrial pollution (UBA, 2002). Also, their isotope values (d18O = 39.0–60.7; d15N = 0.4–3.7) are typical for nitrate from atmospheric deposition (Durka et al., 1994; Campbell et al., 2002). d15 N–NO 3 values between 4.3& and 2.4& were found in atmospheric deposition collected in Ju¨lich (Germany), and a seasonal trend with higher values during fall and winter was observed (Freyer, 1991). The elevated d15N values of the river nitrate well reflect the agricultural land use of the sub basin. A positive correlation between percentage of land used for agriculture and urban purposes and mean d15N
1340
B. Deutsch et al. / Organic Geochemistry 37 (2006) 1333–1342
values of nitrate for 16 river-systems in the northeastern US is reported from Mayer et al. (2002). They determined a mean d15N value of 8.4& for the Schuylkill River with 38.4% of land used for agriculture. Although fractionation processes like denitrification and assimilation in the river cannot absolutely be excluded, the d15N and d18O values in the river nitrate do not indicate such a process. Whereas the increase in the d15 N–NO 3 value might be an indicator for assimilation or denitrification, the simultaneous decrease in the d18 O–NO 3 values argue against fractionation. We therefore assume, that fractionation of the river nitrate was not significant during the sampling period. The mixing-model indicated that the nitrate in the sub basin ‘Middle Warnow’ mainly consists of drainage water (86%) and groundwater (11%) (Table 5), although both sources showed similar mean nitrate concentrations. The atmospheric deposition had a share of 3% in the river nitrate. These data are compared with good agreement to estimates of a modified version of the MONERIS model (Modelling Nutrient Emissions in River Systems; Behrendt et al., 1999) for the period 1995–1999 (Table 6; Pagenkopf, 2001). MONERIS estimates N inputs from point sources and six diffuse pathways, which are drainage water, ground-
Table 5 and Concentration weighted mean values of d15 N–NO 3 d18 O–NO 3 used in the mixing-model and the resulting fractions of each source with standard error Source
d15 N–NO 3 ð‰Þ
d18 N–NO 3 ð‰Þ
Fraction [f]
Drainage water Groundwater Rain Mixture River
10.4 0.6 0.1
4.7 1.4 51.7
0.86 ± 0.07 0.11 ± 0.07 0.03 ± 0.02
9.0
6.0
Table 6 Comparison of the estimated proportion of each source on the Warnow River nitrate in the sub basin ‘Middle Warnow’, given from the modified MONERIS model (Pagenkopf, 2001) and this study Pathways
Percentage of input MONERIS
This study
Tile drain outlets Groundwater Atmospheric deposition
80 15 5
86 ± 7 11 ± 7 3±2
water, atmospheric deposition, erosion, urban areas, and surface runoff. To allow a comparison of our results with the model data, the estimated inputs from drainage water, groundwater, and atmospheric deposition given by the modified MONERIS model were added together and set as 100%. Although the MONERIS model estimates inputs of total dissolved inorganic nitrogen, we think that a comparison with our estimation of nitrate inputs is maintainable, since most of the inorganic nitrogen is emitted to the aquatic ecosystems in the form of nitrate (Addiscott et al., 1992). Furthermore, our estimate only represents the winter period, whereas the model calculates annual mean values. Several studies have shown that the main portion of nitrate leaches from agricultural soils during fall and winter, when most nitrate derives from mineralization of soil organic N (Kirchmann et al., 2002). Combined with the high amount of water discharge due to low evapotranspiration and high rainfalls during winter and fall, more nitrate leaches to the drainage system (Lammel, 1990). According to the MMK data our investigated agricultural soils (water saturated sandy loams and sands) account for 11% of the total agricultural area in the sub basin ‘Middle Warnow’, which is dominated by water saturated loams (50%) and peat soils (20%). It cannot be excluded that the drainage water from loam and peat soils show elevated d15 N–NO 3 and d18 O–NO 3 values because of a higher influence of denitrification, but results from a second study carried out on agricultural fields in the same sub basin showed similar isotope values in drainage water nitrate (Deutsch et al., unpublished data). Uncertainty in source attribution mainly exists for the share of groundwater nitrate, since only one aquifer was sampled. Only a 2& enrichment 18 in the d15 N–NO 3 and d O–NO3 values as reported by Bo¨ttcher et al. (1990) during denitrification, would increase the contribution of the groundwater nitrate from 11% to 15%. 5. Conclusions Our study shows that the stable isotope ratios of nitrogen and oxygen in nitrate can be used to quantify the sources of nitrate to a river. Given that the various sources differ substantially in their isotopic composition the mixing-model approach provides a trustworthy estimate for the contribution of the different sources. The choice of the winter season
B. Deutsch et al. / Organic Geochemistry 37 (2006) 1333–1342
for sampling may have supported the quality of our results – that are based on only few samples – since low temperatures reduce microbial fractionation processes. Acknowledgements This work was part of the project ‘‘The southern Baltic Sea and its coast in change’’, supported by the German Federal Environmental Foundation. The authors thank K. Hennig from the Eurawasser GmbH, R. Bo¨rner from the StAUN Rostock and R. Kolbe from the LUNG Gu¨strow for their cooperativeness and support. Special thanks to Bernhard Mayer, Katrin Schmidt and Angus Atkinson for their helpful and critical comments. References Addiscott, T.M., Whitmore, A.P., Powlson, D.S., 1992. Farming, Fertilizers and the Nitrate Problem. CAB International, Wallingford. Amberger, A., Schmidt, H.-L., 1987. Natu¨rliche Isotopengehalte von Nitrat als Indikatoren fu¨r dessen Herkunft. Geochimica et Cosmochimica Acta 51, 2699–2705. Behrendt, H., Huber, P., Kornmilch, M., Opitz, D., Schmoll, O., Scholz, G., Uebe, R., 1999. Na¨hrstoffbilanzierung der Flussgebiete Deutschlands. Umweltbundesamt, Berlin. Bockholt, R., Kappes, K., 1994. Stoffaustrag u¨ber Entwa¨sserungssysteme - ein aktueller Niedermoorgru¨nland ¨ kologische Untersuchungen im Ackerland - Vergleich, O Warnow-Einzugsgebiet II, 2/1994, LAUN M-V, pp. 136–146. Bo¨rner, R., Bo¨nsch, R., Fadschild, K., Gosselck, F., Hu¨bener, T., Klinkenberg, G., Kolbow, D., Lill, D., Merkel, G., Neumann, C., Randow, F.F.E., Schlungbaum, G., Selig, U., Winkler, H., 1994. Ein Beitrag zur Biologie der Warnow, eines ¨ kologische norddeutschen nacheiszeitlichen Tieflandflusses, O Untersuchungen im Warnow-Einzugsgebiet II, Heft 2, Schriftenreihe des Landesamtes fu¨r Umwelt und Natur Mecklenburg-Vorpommern, Landesamt fu¨r Umwelt und Natur Mecklenburg-Vorpommern, pp. 56–92. Bo¨ttcher, J., Strebel, O., Voerkelius, S., Schmidt, H.-L., 1990. Using isotope fractionation of nitrate–nitrogen and nitrate– oxygen for evaluation of microbial denitrification in a sandy aquifer. Journal of Hydrology 114, 413–424. Campbell, D.H., Kendall, C., Chang, C.C.Y., Silva, S.R., Tonnessen, K.A., 2002. Pathways for nitrate release from an alpine watershed: determination using d15N and d18O. Water Resources Research 38, 10-1–10-9. Chang, C.C.Y., Kendall, C., Silva, S.R., Battaglin, W.A., Campbell, D.H., 2002. Nitrate stable isotopes: tools for determining nitrate sources among different land uses in the Mississippi River Basin. Canadian Journal of Fisheries and Aquatic Sciences 59, 1874–1885. Durka, W., Schulze, E.-D., Gebauer, G., Voerkelius, S., 1994. Effects of forest decline on uptake and leaching of deposited nitrate determined from 15N and 18O measurements. Nature 372, 765–767.
1341
Flipse, W.J., Bonner, F.T., 1985. Nitrogen-isotope ratios of nitrate in ground water under fertilized fields, Long Island, New York. Groundwater 23, 59–67. Freyer, H.D., Aly, A.I.M., 1974. Nitrogen-15 variations in fertilizer nitrogen. Journal Environmental Quality 3, 405–406. Freyer, H.D., 1991. Seasonal variation of 15N/14N ratios in atmospheric nitrate species. Tellus 43, 30–44. Grasshoff, K., Kremling, K., Ehrhardt, M., 1999. Methods of Seawater Analysis, third ed. Wiley VCH, Weinheim. Heaton, T.H.E., 1986. Isotopic studies of nitrogen pollution in the hydrosphere and atmosphere: a review. Chemical Geology 59, 87–102. HELCOM, 2003. Executive summary of the Fourth Baltic Sea Pollution Load Compilation PLC-4. HELCOM, Helsinki. Ho¨gberg, P., 1997. 15N natural abundance in soil–plant systems. New Phytologist 137, 179–203. Iqbal, M.Z., Krothe, N.C., Spalding, R.F., 1997. Nitrogen isotope indicators of seasonal source variability to groundwater. Environmental Geology 32, 210–218. Kendall, C., 1998. Tracing nitrogen sources and cycles in catchments. In: Kendall, C., McDonnell, J.J. (Eds.), Isotope Tracers in Catchment Hydrology. Elsevier, Amsterdam, pp. 519–576. Kendall, C., Aravena, R., 1999. Nitrate isotopes in groundwater systems. In: Cook, P.G., Herozeg, A.L. (Eds.), Environmental Tracers in Subsurface Hydrology. Kluwer Academic, Boston, pp. 261–297. Kirchmann, H., Johnston, A.E.J., Bergstro¨m, L.F., 2002. Possibilities for reducing nitrate leaching from agricultural land. Ambio 31, 404–408. Kroopnick, P.M., Craig, H., 1972. Atmospheric oxygen: isotopic composition and solubility fractionation. Science 175, 54–55. Lammel, J., 1990. Na¨hrstoffaustra¨ge durch Vorfluter und Dra¨ne. Landwirtschaftsverlag GmbH, Mu¨nster-Hiltrup. Mariotti, A., Landreau, A., Simon, B., 1988. 15N isotope biogeochemistry and natural denitrification process in groundwater: application to the chalk aquifer of Northern France. Geochimica et Cosmochimica Acta 52, 1869–1878. Mayer, B., Bollwerk, S.M., Mansfeldt, T., Hu¨tter, B., Veizer, J., 2001. The oxygen isotope composition of nitrate generated by nitrification in acid forest floors. Geochimica et Cosmochimica Acta 65, 2743–2756. Mayer, B., Boyer, E.W., Goodale, C., Jaworski, N.A., van Breemen, N., Howarth, R.W., Seitzinger, S., Billen, G., Lajtha, K., Nadelhoffer, K., van Dam, D., Hetling, L.J., Nosal, M., Paustian, K., 2002. Sources of nitrate in rivers draining sixteen watersheds in the northeastern US: isotopic constraints. Biogeochemistry 57/58, 171–197. Meybeck, M., 1982. Carbon, nitrogen, and phosphorous transport by world rivers. The American Journal of Science 282, 401–450. Pagenkopf, W., 2001. Aktuelle Na¨hrstoffbilanzierung fu¨r Teilgebiete des Einzugsgebiets der Warnow. Geodaten Integration & Analyse, Berlin. Pfenning, K.S., McMahon, P.B., 1996. Effect of nitrate, organic carbon, and temperature on potential denitrification rates in nitrate-rich riverbed sediments. Journal of Hydrology 187, 283–295. Phillips, D.L., Koch, P.L., 2002. Incorporating concentration dependence in stable isotope mixing models. Oecologica 130, 114–125.
1342
B. Deutsch et al. / Organic Geochemistry 37 (2006) 1333–1342
Piatek, K.B., Mitchell, M.J., Silva, S.R., Kendall, C., 2005. Sources of nitrate in snowmelt discharge: evidence from water chemistry and stable isotopes of nitrate. Water, Air, and Soil Pollution 165, 13–35. Silva, S.R., Kendall, C., Wilkison, D.H., Ziegler, A.C., Chang, C.C.Y., Avanzino, R.J., 2000. A new method for collection of nitrate from fresh water and the analysis of nitrogen and oxygen isotope ratios. Journal of Hydrology 228, 22–36. Spoelstra, J., Schiff, S.L., Elgood, R.J., Semkin, R.G., Jeffries, D.S., 2001. Tracing the sources of exported nitrate in the Turkey Lakes Watershed using 15N/14N and 18O/16O isotopic ratios. Ecosystems 4, 536–544. Sta˚lnacke, P., Grimvall, A., Sundblad, K., Tonderski, A., 1999. Estimation of riverine loads of nitrogen and phosphorous to
the Baltic Sea, 1970–1993. Environmental Monitoring and Assessment 58, 173–200. Stepanauskas, R., Leonardson, L., 1999. Bioavailability of wetland-derived DON to freshwater and marine bacterioplankton. Limnology and Oceanography 44, 1477–1485. Thiele, V., Mehl, D., 1995. Ein o¨kologisch begru¨ndetes Sanierungskonzept fu¨r das Gewa¨ssereinzugsgebiet der Warnow (Mecklenburg-Vorpommern). Landesamt fu¨r Umwelt und Natur Mecklenburg-Vorpommern, Gu¨strow. UBA, 2002. Jahresbericht 2001 aus dem Messnetz des Umweltbundesamtes. Umweltbundesamt, Berlin. Wassenaar, L.I., 1995. Evaluation of the origin and fate of nitrate in the Abbotsford Aquifer using the isotopes of 15N and 18O in NO 3 . Applied Geochemistry 10, 391–405.