Description: Diazotrophic dinitrogen (N2) fixation contributes ~76% to "new" nitrogen inputs to the sunlit open ocean, but environmental factors determining N2 fixation rates are not well constrained. Excess phosphate (phosphate-nitrate/16 〉 0) and iron availability control N2 fixation rates in the eastern tropical North Atlantic (ETNA), but it remains an open question how excess phosphate is generated within or supplied to the phosphate-depleted sunlit layer. Our observations in the ETNA region (8°N-15°N, 19°W-23°W) suggest that Prochlorococcus and Synechococcus, the two ubiquitous non-diazotrophic cyanobacteria with cellular N:P ratios higher than the Redfield ratio, create an environment of excess phosphate, which cannot be explained by diapycnal mixing, atmospheric, and riverine inputs. Thus, our results unveil a new biogeochemical niche construction mechanism by non-diazotrophic cyanobacteria for their diazotrophic phylum group members (N2 fixers). Our observations may help to understand the prevalence of diazotrophy in low-phosphate, oligotrophic regions.

Description: Marine nitrogen (N2) fixation supports significant primary productivity in the global ocean. However, in one of the most productive regions of the world ocean, the northern Humboldt Upwelling System (HUS), the magnitude and spatial distribution of this process remains poorly characterized. This study presents a spatially resolved dataset of N2 fixation rates across six coastal transects of the northern HUS off Peru (8°S – 16°S) during austral summer. N2 fixation rates were detected throughout the waters column including within the OMZ between 12°S and 16°S. N2 fixation rates were highest where the subsurface Oxygen Minimum Zone (OMZ, O2 〈20 µmol L-1) was most intense and estimated nitrogen (N) loss was highest. There, rates were measured throughout the water column. Hence the vertical and spatial distribution of rates indicates colocation of N2 fixation with N loss in the coastal productive waters of the northern HUS. Despite high phosphate and total dissolvable iron (TdFe) concentrations throughout the study area, N2 fixation was still generally low (1.19 ± 3.81 nmol L-1 d-1) and its distribution could not be directly explained by these two factors. Our results suggest that the distribution was likely influenced by a complex interplay of environmental factors including phytoplankton biomass and organic matter availability, and potentially iron, or other trace metal (co)-limitation of both N2 fixation and primary production. In general, our results support previous conclusions that N2 fixation in the northern HUS plays a minor role as a source of new N and to replenish the regional N loss. Key Points: A north-to-south pattern in N2 fixation rates was observed implying increased N turnover between 12°S and 16°S where N loss was pronounced Highest N2 fixation rates were measured in coastal productive waters above and within the OMZ, showing no clear relationship with Fe or P The magnitude of N2 fixation was low compared to predictions, estimated to account for ∼0.3% of primary production and 〈2% of local N loss