Isotopic constraints on the pre-industrial oceanic nitrogen budget

The size of the bioavailable (i.e., fixed) nitrogen inventory in the ocean influences global marine productivity and the biological carbon pump. Despite its importance, the pre-industrial rates for the major source and sink terms of the oceanic fixed nitrogen budget, N-2 fixation and denitrification, respectively, are not well known. These processes leave distinguishable imprints on the ratio of stable nitrogen isotopes, delta N-15, which can therefore help to infer their patterns and rates. Here we use delta N-15 observations from the water column and a new database of seafloor measurements to constrain rates of N-2 fixation and denitrification predicted by a global three-dimensional Model of Ocean Biogeochemistry and Isotopes (MOBI). Sensitivity experiments were performed to quantify uncertainties associated with the isotope effect of denitrification in the water column and sediments. They show that the level of nitrate utilization in suboxic zones, that is the balance between nitrate consumption by denitrification and nitrate replenishment by circulation and mixing (dilution effect), significantly affects the isotope effect of water column denitrification and thus global mean delta(NO3-)-N-15. Experiments with lower levels of nitrate utilization within the suboxic zone (i.e., higher residual water column nitrate concentrations, ranging from 20 to 32 mu M) require higher ratios of benthic to water column denitrification, BD: WCD = 0.75-1.4, to satisfy the global mean NO3- and delta(NO3-)-N-15 constraints in the modern ocean. This suggests that nitrate utilization in suboxic zones plays an important role in global nitrogen isotope cycling. Increasing the net fractionation factor epsilon(BD) for benthic denitrification (epsilon(BD) = 0-4 %) requires even higher ratios, BD: WCD = 1.4-3.5. The model experiments that best reproduce observed seafloor delta N-15 support the middle to high-end estimates for the net fractionation factor of benthic denitrification (epsilon(BD) = 2-4 %). Assuming a balanced fixed nitrogen budget, we estimate that pre-industrial rates of N-2 fixation, water column denitrification, and benthic denitrification were between 195-350 (225), 65-80 (76), and 130-270 (149) Tg N yr(-1), respectively, with our best model estimate (epsilon(BD) = 2 %) in parentheses. Although uncertainties still exist, these results suggest that marine N-2 fixation is occurring at much greater rates than previously estimated and the residence time for oceanic fixed nitrogen is between similar to 1500 and 3000 yr.