Limitations of Using δ 18 O for the Source Identification of Nitrate in Agricultural Soils MARTIN MENGIS, † ULRICH WALTHER, ‡ STEFANO M. BERNASCONI, § AND BERNHARD WEHRLI* ,† Swiss Federal Institute for Environm ental Science and Technology (EAWAG), Lim nological Research Center, CH-6047 Kastanienbaum, Switzerland, Swiss Federal Research Station for Agroecology and Agriculture, Postfach, CH-8046 Zurich, Switzerland, and Swiss Federal Institute of Technology (ETH), Geological Institute, ETH Zentrum, CH-8092 Zurich, Switzerland The stable isotopic composition (δ 15 N and δ 18 O) of nitrate was analyzed in two lysimeter field experiments in order to identify the conditions under which the dual isotope approach can be applied to identify the main source of nitrate in agricultural soils. The first field experiment involved six lysimeters beneath fields that had been fertilized for 10 yr with the same type of fertilizer (NH 4 NO 3 ; δ 15 N )+1.2‰, δ 18 O )+18.6‰). The isotope ratios of NO 3 - in the leachate (δ 15 N ≈ 0‰; δ 18 O ≈ +2‰) could not be interpreted in a conventional way with either fertilizer or soil organic nitrogen as main sources. These results provided clear evidence for the microbial immobilization and subsequent mineralization and nitrification to NO 3 - (mineralization-immobilization turnover concept). This process masked the original oxygen isotope ratio of the fertilizer source during the summer when microbial activity was high. A second experiment involving the application of Ca(NO 3 ) 2 to three lysimeters during the winter confirmed that the dual isotope approach remains valid for the source identification of nitrate under conditions of low microbial activity. The study reveals the limitation of the dual isotope approach to characterize nitrate sources under biologically active conditions and the ability to quantify microbial processes when the main sources can be controlled. Introduction Leaching of nitrate from farmland is a common problem affecting groundwater quality in Europe and North America (1-3).Nitrate in soils originates from different sources,which include mineralization of soil organic nitrogen, application of organic and inorganic N fertilizers, and atmospheric deposition. These different sources produce nitrate with distinct isotopic composition (Figure 1). Nitrate originating from manure is characterized by high δ 15 N compositions (>10‰ ) (1, 2).Inorganic nitrate fertilizers (4)are significantly more positive in δ 18 O(>15‰ ) as compared to nitrate from most other sources (δ 18 O <10‰ ). The δ 15 N composition has been used for almost three decades to trace nitrate sources (2, 5, 6).More recently,the δ 18 Oofnitratehasbeen combined with δ 15 N values in a dualisotope approach (3, 4, 7)to identify the sources of soil nitrate. However, it has also been shown that microbial trans- formations introduce complications in the interpretation of the isotopic composition of nitrate (8). Processes such as denitrification lead to isotopic fractionation, masking the effect ofthe originalnitrate source on the isotope fingerprint ofsoilnitrate (9-12).Thismaybeoneofthereasons,together with analytical difficulties, why only a limited number of field studies in agricultural environments used both δ 15 N and δ 18 Ofor the source identification ofsoilnitrate.Contrary to the expectations based on Figure 1, some of the studies (7) based on the dual isotope approach in agricultural areas reported no significant differences in δ 18 O values and therefore could not distinguish between the various possible nitrate sources.On the other hand,severalstudies in forested (13, 14) and alpine catchments (15) proved that δ 18 O-nitrate can be a powerful tracer of different nitrate sources. In addition to denitrification, other microbial nitrate transformation processes such as nitrate uptake and trans- formation into biomassmayalso lead to fractionation effects. However, manypublished field studies applying the isotope approach to identifythe source ofnitrate in agriculturalsoils neglected the possible fractionation effects of microbial processes other than denitrification. The goal of this study was to determine (i) under which circumstances the analysis of δ 18 O-nitrate can be used as a direct tracer of the source of nitrate in agricultural soils and (ii)under which conditionsfractionation effectsbymicrobial immobilization mask the isotopic signature of the nitrate source.These questions need to be answered before the dual isotope approach can be used appropriately in field studies. Experimental Section Lysimeter Experiments. Lysimeters were used for the field experiments because they allow sampling of leachate that percolatesthrough the upper horizon ofsoilsfrom a specified area.The lysimeters are located at the Swiss FederalResearch Station for Agroecology and Agriculture in Zurich and have a diameter and depth of 80 and 84 cm, respectively. They were installed in 1986 and filled from bottom to top as follows: 6 cm of gravel (diameter 3-5 cm), 4 cm of fine *Corresponding author e-mail: bernhard.wehrli@eawag.ch; phone: +41 41 349 21 17. † EAWAG, Limnological Research Center. ‡ Swiss FederalResearch Station for Agroecologyand Agriculture. § ETH, Geological Institute. FIGURE 1. Schematic diagram of the isotopic composition of soil nitrate from various sources (modified after refs 3, 4, and 13). The arrow s indicate the change in the isotopic composition of nitrate due to fractionation by partial denitrification (9, 10, 12). Environ. Sci. Technol. 2001, 35, 1840-1844 1840 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 9, 2001 10.1021/es0001815 CCC: $20.00  2001 American Chemical Society Published on Web 03/28/2001