ABSTRACT: We developed a simple and reliable method which allows simultaneous isotope-ratio analysis of inorganic (DIN) and organic (DON) forms of nitrogen extracted from seawater. All forms
of nitrogen under analysis are converted to ammonium, by diffusionwith magnesium oxide, prior to collection on glass-fiber filters appropriate for mass spectrometric assay of ¹⁵N. Oxidized DIN forms (nitrate, nitrite) are reduced to ammonium in the presence of Devarda alloy. Conversion of DON to ammonium isperformed by wet oxidation using potassium persulfate and subsequent reduction of the nitrate formed. Recovery tests, both for total nitrogen and ¹⁵N content, showed that this procedure is suitable for application in DI¹⁵N-isotopedilution experiments and DON-release studies. Recovery of total nitrogen from DIN and DON was nearly complete (94 to 97%). The variability in the experimental determination of ¹⁵N abundance was &LT2% and &LT4% for DIN and DON, respectively. Weused the method to balance the ¹⁵N budget in nitrate and ammonium uptake experiments conducted in an oligotrophic area (tropical North Atlantic) by including, in addition to the substrate (DIN) and biomass (PON) pool, the DON pool. However, theuse of glass-fiber filters (GF/F) for the collection of particulate matter produced a significant artifact, i.e. a large amount of small particles (&LT0.7 um, PON_&LTGF/F; prochlorophytes and/or bacteria) passed through these filters and wererecovered together with the DON in a combined pool. While inclusion of this combined pool led virtually to a complete accounting for the ¹⁵N label (99%) in all samples for nitrate uptake and in those for ammonium uptake incubated for &LT8 h, nomass balance was achieved during ammonium uptake lasting 10 to 24 h. We suggest that the ¹⁵N that was still missing (13%) resulted mainly from bottle containment effects such as ammonium-ion adsorption and/or PON adherence onto incubationbottle walls. Transfer of ¹⁵N label to the combined pool (nitrate experiment) and to the DON and PON_&LTGF/F pools (ammonium experiment) represented up to 41, 38 and 20% of the total ¹⁵N taken up as DIN, respectively, anddepended strongly upon the length of incubation. Failure to take these pathways of the missing ¹⁵N into account during traditional ¹⁵N uptake experiments involves risk of substantially underestimating new and regenerated production,at least in oligotrophic areas. The latter fact has considerable significance in the design of future ¹⁵N tracer methodologies.