We report a novel method for measurement of the oxygen isotopic composition (18O/16O) of nitrate (NO3-) from both seawater and freshwater. The denitrifier method, based on the isotope ratio analysis of nitrous oxide generated from sample nitrate by cultured denitrifying bacteria, has been described elsewhere for its use in nitrogen isotope ratio (15N/14N) analysis of nitrate. (1) Here, we address the additional issues associated with 18O/16O analysis of nitrate by this approach, which include (1) the oxygen isotopic difference between the nitrate sample and the N20 analyte due to isotopic fractionation associated with the loss of oxygen atoms from nitrate and (2) the exchange of oxygen atoms with water during the conversion of nitrate to N2O. Experiments with 18O-labeled water indicate that water exchange contributes less than 10%, and frequently less than 3%, of the oxygen atoms in the N20 product for Pseudomonas aureofaciens. In addition, both oxygen isotope fractionation and oxygen atom exchange are consistent within a given batch of analyses. The analysis of appropriate isotopic reference materials can thus be used to correct the measured 18O/16O ratios of samples for both effects. This is the first method tested for 18O/16O analysis of nitrate in seawater. Benefits of this method, relative to published freshwater methods, include higher sensitivity (tested down to 10 nmol and 1 microM NO3-), lack of interference by other solutes, and ease of sample preparation.

We report a new method for measurement of the isotopic composition of nitrate (NO3-) at the natural-abundance level in both seawater and freshwater. The method is based on the isotopic analysis of nitrous oxide (N20) generated from nitrate by denitrifying bacteria that lack N2O-reductase activity. The isotopic composition of both nitrogen and oxygen from nitrate are accessible in this way. In this first of two companion manuscripts, we describe the basic protocol and results for the nitrogen isotopes. The precision of the method is better than 0.2/1000 (1 SD) at concentrations of nitrate down to 1 microM, and the nitrogen isotopic differences among various standards and samples are accurately reproduced. For samples with 1 microM nitrate or more, the blank of the method is less than 10% of the signal size, and various approaches may reduce it further.