The distinct nitrogen isotopic compositions of low and high molecular weight marine DON

To constrain the sources and cycling of bulk and size-fractionated marine dissolved organic nitrogen (DON), samples were collected for concentration and isotopic (δ15N) analysis from the western tropical North Atlantic Ocean as well as over the North Australian shelf. Bulk DON concentrations are typically between 4 and 6 μM, with low molecular weight (LMW, i.e., < 1000 Da) DON accounting for 60 to 70% of total DON. The δ15N of both bulk and high molecular weight (HMW, i.e., > 1000 Da to < 0.2 μm) DON from the western tropical North Atlantic is similar to previous measurements, as well as to bulk and HMW DON collected off the North Australian Shelf, ~ 4‰. Here we report the first measurement of marine LMW DON δ15N, and a coherent pattern emerges where LMW DON δ15N < HMW DON δ15N. An analytical concern is that the bulk DON δ15N as calculated from our coupled HMW and LMW measurements is often lower than that measured directly for the bulk DON. Despite this discrepancy, the self-consistency of the data in other regards argues for the robustness of the basic observation of lower δ15N in LMW relative to HMW DON. One explanation for this isotopic difference is considered most likely, based on the model that DON δ15N elevation relative to PON is fundamentally due to DON production from PON without isotopic fractionation, coupled with DON destruction occurring with fractionation. In this model, HMW DON loss must occur predominantly through the breakage of N-containing bonds, for which isotope fractionation should be substantial, whereas LMW DON loss is partly due to direct assimilation by phytoplankton and other microbes, which appears to occur with only minor isotopic fractionation. However, an alternative hypothesis exists: LMW and HMW DON may have distinct sources, for example, prokaryotic and eukaryotic plankton, respectively. In either case, it appears that most LMW DON has not passed through the HMW DON pool, as this would require an unrealistically large isotope fractionation at the HMW to LMW conversion. Finally, neither the concentration nor the δ15N of the bulk or either size fraction of DON varies with in situ N2 fixation rates measured at the time of sample collection. However, the δ15N of both bulk and LMW DON from the western tropical North Atlantic shows significant differences between samples collected six months apart, which may be due to seasonally variable stimulation of primary production by the Amazon River.