Cell-specific measurements show nitrogen fixation by particle-attached putative non-cyanobacterial diazotrophs in the North Pacific Subtropical Gyre

Biological nitrogen fixation is a major important source of nitrogen for low-nutrient surface oceanic waters. Nitrogen-fixing (diazotrophic) cyanobacteria are believed to be the primary contributors to this process, but the contribution of non-cyanobacterial diazotrophic organisms in oxygenated surface water, while hypothesized to be important, has yet to be demonstrated. In this study, we used simultaneous 15N-dinitrogen and 13C-bicarbonate incubations combined with nanoscale secondary ion mass spectrometry analysis to screen tens of thousands of mostly particle-associated, cell-like regions of interest collected from the North Pacific Subtropical Gyre. These dual isotope incubations allow us to distinguish between non-cyanobacterial and cyanobacterial nitrogen-fixing microorganisms and to measure putative cell-specific nitrogen fixation rates. With this approach, we detect nitrogen fixation by putative non-cyanobacterial diazotrophs in the oxygenated surface ocean, which are associated with organic-rich particles (<210 µm size fraction) at two out of seven locations sampled. When present, up to 4.1% of the analyzed particles contain at least one active putative non-cyanobacterial diazotroph. The putative non-cyanobacterial diazotroph nitrogen fixation rates (0.76 ± 1.60 fmol N cell−1 d−1) suggest that these organisms are capable of fixing dinitrogen in oxygenated surface water, at least when attached to particles, and may contribute to oceanic nitrogen fixation.


nature research | reporting summary
April 2020 Field-specific reporting Please select the one below that is the best fit for your research. If you are not sure, read the appropriate sections before making your selection. Did the study involve field work?
Yes No N2 fixation rates were calculated from 7 locations across the North Pacific Subtropical Gyre at 15 m depth using dual-isotope incubations of 15N2 gas and 13C-bicarbonate. Each location had a natural-light treatment (natural light/dark cycle) and an all-dark treatment (24-hours of darkness) as a factorial design structure. Each treatment consisted of biological triplicates. A single subsample of the natural-light treatment from each location was used for cell-specific N2 fixation rate measurements by nanoSIMS. The one sample was analyzed with nanoSIMS at over 150 (~30 x 30 µm) analysis frames from each location. Replicate DNA samples from the seven locations were amplified for the nifH gene and sequenced. Sequences were pooled for average relative abundance.
Samples were chosen from the across the North Pacific Subtropical Gyre as it has low nutrients and is know to have nitrogen fixation. The samples represent populations of plankton microorganisms between 0.2 and 210 micrometers. Samples were collected periodically across the Pacific to cover a broad geography and environmental conditions and increase the probability of encountering non-cyanobacterial N2 fixing organisms. Isotopically labeled (15N2, 13C-bicarbonate) planktonic community incubations from the 7 locations across the North Pacific Subtropical Gyre. Isotope incubations were filtered onto~0.7 µm (GF/F, community N2 fixation) and 0.2 µm (silver, cell-specific N2 fixation) pore size filters for analysis. In addition, corresponding planktonic samples without isotope additions were collected and filtered (0.2 µm) for nifH gene amplification and sequencing.
A sample size of 4L for isotope incubations was chosen to ensure enough biomass (>10 µg N) was collected on the filter to accurately quantify 15N enrichment from particulate matter. Previous studies in similar nutrient-poor systems determined the minimum biomass necessary for reliable quantification (White et al., 2020). In addition, Triplicate incubations were chosen for a more robust average value and associated error. Subsamples for nanoSIMS analysis ranged from 0.1 to 0.5 L of incubation volume from each location. Based on the maximum abundances of the cells of interest (non-cyanobacterial diazotroph) in similar habitats (103 cells L-1), this volume would be sufficient while minimizing fixative waste used in sample preparation.
While at sea, data was collected in recorded in the cruise-specific laboratory notebook by K.H., E.W.K.M., and K.A.T.K. Data were compiled into Excel spreadsheets by K.A.T.K. NanoSIMS data was automatically saved to internal servers by X.M., P.K.W., and K.H. at Lawrence Livermore National Laboratory. All analyzed nanoSIMS data was collected in excel spreadsheets by K.H. IRMS data was collected at the Isotope Biochemistry Laboratory in Hawaii and analyzed in excel by K.H. MIMS data was collected the University of California, Santa Cruz and analyzed in excel by K.H. Raw nifH sequence data was collected at the University of Illinois Core Facility and transferred via BaseSpace to K.A.T.K who analyzed and processed sequences in a remote server.
Samples were collected aboard the R/V Sally Ride in a relocation transect across the North Pacific from Guam to San Francisco. Samples were collected from November 5 to November 26, 2019. Seven samples were chosen across the Pacific in total covering approximately 5000 nautical miles, with 400-700 nautical miles between each sample. The large spatial scale was chosen to investigate as much area and different systems as possible provided the limitations set by nanoSIMS instrument time.
Raw sequencing data was excluded during sequence analysis if it did not pass quality controls, which included removing: chimeric sequences, singletons, OTUs with less than 10 read counts, any non-diazotrophic sequences, and sequences with stop codons in the middle of the reading frame. No data was excluded from community N2 fixation rate measurements. NanoSIMS data were also removed if they did not pass quality controls. For nanoSIMS data, regions of interest with 15N enrichment values above the threshold (3x standard deviation of the standard cells) were discarded if the total rare isotopic counts were too low (< 100), the associated error was too high (> 30% to the 15N enrichment value), the ROI size was too small (< 0.4 µm diameter), lack of a clear cell-like outline in either 12C2-or 14N12C-or 15N enrichment was not measured in the majority of the defined ROI. Additionally, all ROIs, enriched and unenriched, were excluded if they did not meet minimum count threshold in 12C2-or 14N12C-.
Reproducibility in a dynamic environment such as the surface ocean is nearly impossible, and our results represent a snapshot of an ever-changing system. Furthermore, Diazotroph distribution and the N2 fixation activity are dependent on factors specific to each location. However, isotope incubations and subsequent nanoSIMS analysis were conducted at 7 distinct locations, and N2-fixing cells were found at 6 out of 7 stations which validates the reproducibility of our results within that system.
Randomization did not apply to this study. Samples were not grouped.
Blinding was not necessary for this study. Cells were considered enriched and fixing N2 if the 15N enrichment value was higher than 3x the standard deviation of regularly used standards. This binary definition of enrichment does not allow for bias results.