Nitrogen isotopic evidence for a shift from nitrate- to diazotroph-fueled export production in the VAHINE mesocosm experiments

In a coastal lagoon with a shallow, 25 m water column off the southwest coast of New Caledonia, large-volume ( ∼  50 m<sup>3</sup>) mesocosm experiments were undertaken to track the fate of newly fixed nitrogen (N). The mesocosms were intentionally fertilized with 0.8 µM dissolved inorganic phosphorus to stimulate diazotrophy. N isotopic evidence indicates that the dominant source of N fueling export production shifted from subsurface nitrate (NO<sub>3</sub><sup>−</sup>) assimilated prior to the start of the 23-day experiments to N<sub>2</sub> fixation by the end of the experiments. While the <i>δ</i><sup>15</sup>N of the sinking particulate N (PN<sub>sink</sub>) flux changed during the experiments, the <i>δ</i><sup>15</sup>N of the suspended PN (PN<sub>susp</sub>) and dissolved organic N (DON) pools did not. This is consistent with previous observations that the <i>δ</i><sup>15</sup>N of surface ocean N pools is less responsive than that of PN<sub>sink</sub> to changes in the dominant source of new N to surface waters. In spite of the absence of detectable NO<sub>3</sub><sup>−</sup> in the mesocosms, the <i>δ</i><sup>15</sup>N of PN<sub>sink</sub> indicated that NO<sub>3</sub><sup>−</sup> continued to fuel a significant fraction of export production (20 to 60 %) throughout the 23-day experiments, with N<sub>2</sub> fixation dominating export after about 2 weeks. The low rates of organic N export during the first 14 days were largely supported by NO<sub>3</sub><sup>−</sup>, and phytoplankton abundance data suggest that sinking material primarily comprised large diatoms. Concurrent molecular and taxonomic studies indicate that the diazotroph community was dominated by diatom–diazotroph assemblages (DDAs) at this time. However, these DDAs represented a minor fraction (< 5 %) of the total diatom community and contributed very little new N via N<sub>2</sub> fixation; they were thus not important for driving export production, either directly or indirectly. The unicellular cyanobacterial diazotroph, a <i>Cyanothece</i>-like UCYN-C, proliferated during the last phase of the experiments when N<sub>2</sub> fixation, primary production, and the flux of PN<sub>sink</sub> increased significantly, and <i>δ</i><sup>15</sup>N budgets reflected a predominantly diazotrophic source of N fueling export. At this time, the export flux itself was likely dominated by the non-diazotrophic diatom, <i>Cylindrotheca closterium</i>, along with lesser contributions from other eukaryotic phytoplankton and aggregated UCYN-C cells, as well as fecal pellets from zooplankton. Despite comprising a small fraction of the total biomass, UCYN-C was largely responsible for driving export production during the last  ∼  10 days of the experiments both directly ( ∼  5 to 22 % of PN<sub>sink</sub>) and through the rapid transfer of its newly fixed N to other phytoplankton; we infer that this newly fixed N was transferred rapidly through the dissolved N (including DON) and PN<sub>susp</sub> pools. This inference reconciles previous observations of invariant oligotrophic surface ocean DON concentrations and <i>δ</i><sup>15</sup>N with incubation studies showing that diazotrophs can release a significant fraction of their newly fixed N as some form of DON.