Table of Contents

• Coverage
• Dataset Description
  • Methods & Sampling
  • Data Processing Description
  • BCO-DMO Curation Notes
• Data Files
• Supplemental Files
• Related Publications
• Parameters
• Instruments
• Project Information
• Funding

Sample collection and solid phase extraction:
Surface water samples (500 milliliters (mL)) were collected at all river stations via a submersible pump lowered ~1.0 meter (m) below the water's surface. For the collection of submarine groundwater samples, U.S. Geological Survey small boat operations along the West Florida Shelf (WFS) were conducted. Surface water, bottom water, and submarine groundwater samples (2 liters (L)) were collected via peristaltic pump along three coastal transects.

Samples were filtered through 0.2 micrometer (um) filters (polyethersulfone) into acid-cleaned polypropylene bottles to 2L. The sample was then acidified to a pH of 2-3 with hydrochloric acid (trace metal grade) and pumped at 25 milliliters per minute (mL/min) through Bond-Elut PPL solid phase extraction (SPE) columns (1 gram (g), 6 mL, Agilent Technologies) that had been previously activated by passing 6 mL each of methanol (MeOH, Optima LCMS grade, Fisher Scientific) and ultrapure water (qH₂O, 18.2 MΩ) through the column. SPE columns were rinsed with 10 mL acidified qH₂O (pH = 2) to remove salts and frozen at -20 degrees Celsius (°C) immediately after sample collection and returned to the laboratory for further processing. Sample columns were gravity eluted with 10 mL LCMS Optima grade methanol. The eluent was collected in 15 mL polypropylene centrifuge tube and evaporated in a SpeedVac to concentrate to a volume of ~ 0.2 mL. The eluent was transferred to a 2 mL centrifuge tube and brought up to a final volume of 2 mL with qH₂O. Each sample was spiked with 20 microliters (µL) of 100 micromolar (µM) B12 stock for a concentration of 1 µM cyanocobalamin. Solid phase extraction process blank samples were prepared with the same procedure without sample loading.

Analysis by LCMS:
Tubes were centrifuged at max speed for 5 minutes to sediment any particles, and then the spiked samples were transferred to plastic microcentrifuge tubes for storage in freezer, 1 mL was used for analysis on Orbitrap in 2 mL plastic autosampler vials. A pooled sample from each batch of samples for each Orbitrap run was prepared by combining 15 µL from each sample. Chromatographic separation was performed using a Phenyl-Hexyl column (Waters ACQUITY Premier CSH) with dimensions of 2.1 x 100 millimeters (mm) and a 1.7 µm particle size. The column temperature was maintained at 30.0 °C throughout the analytical run. Mobile phase A was composed of water supplemented with 0.1% formic acid, while mobile phase B consisted of methanol with 0.1% formic acid. The system operated at a constant flow rate of 200 µL/min with a 30-minute gradient starting from 95% solvent A and 5% solvent B to 5% solvent A and 95% solvent B and a 5-minute hold at 95% solvent B. Solvents were then switched back to the initial condition of 95% solvent A and held for 6 minutes to re-equilibrate the column. The total injection volume for each sample was 20 µL. Mass spectral data were acquired using an Orbitrap IQ-X Tribrid Mass Spectrometer (Thermo Fisher Scientific) operating in positive electrospray ionization (ESI) mode. The ESI source parameters were optimized with a spray voltage of 3500 volts (V), a sheath gas flow of 50 (Arb), an auxiliary gas flow of 10 (Arb), and a sweep gas flow of 1 (Arb). The ion transfer tube and vaporizer temperatures were set to 325 °C and 350 °C, respectively. Full MS scans were recorded in the Orbitrap mass analyzer across a mass range of 100 to 1000 m/z. The resolution was set to 500,000 at m/z 200 to ensure high mass accuracy for the analytes. Data acquisition settings included the collection of 1 microscan per data point with a maximum injection time of 1014 milliseconds (ms).

All data were quality controlled and processed using CoreMS with inhouse python based scripts (10.5281/zenodo.19744141). Open source data files were generated using MS Convert. These data files are publicly available in MassIVE under accession #MSV000101377. Orbitrap analyses included sample analyses that are tabulated in the sample list as well as solid phase extraction process blanks analyses (denoted with 'blank' in the file name), and pooled sample quality control analyses (denoted with 'pooled' in the file name).

Quality control:
First, quality control was conducted by checking the extracted ion chromatography (EIC) of internal cyanocobalamin standard (m/z = 678.2915 for the doubly charged form of cyanocobalamin, with a m/z tolerance of 15 parts per million (ppm)) for each sample. Peak areas of internal standard across a time range of 10.5 to 12.0 minutes were calculated. Samples with internal standard peak area differing by more than two standard deviations from the mean across the data set were excluded from further analysis.

Molecular formula assignment:
To assign molecular formula, mass spectra was averaged in 2-minute intervals over a range of 0-30 minutes and a mass range between 100-580 m/z. Each spectra was internally calibrated based on a polynomial correction applied to peaks within a CHON series reference list that included calibrant mass peaks across the entire m/z and retention time range. Search criteria for molecular formula assignments included: C 1-40, H 4-80, O 0-16, N 0-8, S 0-1, and Na 0-1, with a maximum allowed double bond equivalent (DBE = C – (H/2) + (N/2) + 1) of 16 and a maximum mass error (the difference between the average of measured m/z and the theoretical ion mass) of ± 1.5 ppm. When multiple formulas were assigned to a peak within the mass error range, the one with the highest confidence score was selected.

Feature list generation:
With assignments for all samples, a combined list of features (molecular formula assigned m/z peak at a specific retention time interval) was constructed to compare molecular abundance across samples. Features that represent the same mass peak but were assigned with different formulas in different samples were consolidated by choosing the assignment with the highest confidence score. Elemental ratios (O/C, N/C, and H/C) of each feature were calculated based on the molecular formula. Features were classified into different molecular classes based on their stoichiometric ratios. After removing isotopologues, features that satisfied the following criteria were noted as clean features and exported as a clean feature list for downstream analysis: (1) is detected in more than 4 samples, (2) is within the mass range of 100-580 m/z, (3) has a max intensity greater than 20,000, (4) had m/z error within 4 times the standard error of the expected error (expected error is a rolling average of features across the dataset), (5) max intensity in the sample is 5 times greater than in blank samples.

- Imported original file "Clean_Featurelist.csv" into the BCO-DMO data processing system.
- Renamed fields to comply with BCO-DMO naming conventions.
- Saved the final file as "1004302_v1_sting_sgd_river_lcms_dom.csv".

NSF Award Abstract:
This project will investigate how groundwater discharge delivers important nutrients to the coastal ecosystems of the West Florida Shelf. Preliminary studies indicate that groundwater may supply both dissolved organic nitrogen (DON) and iron in this region. In coastal ecosystems like the West Florida Shelf that have very low nitrate and ammonium concentrations, DON is the main form of nitrogen available to organisms. Nitrogen cycling is strongly affected by iron availability because iron is essential for both photosynthesis and for nitrogen fixation. This study will investigate the sources and composition of DON and iron, and their influence on the coastal ecosystem. The team will sample offshore groundwater wells, river and estuarine waters, and conduct two expeditions across the West Florida Shelf in winter and summer. Investigators will participate in K-12 and outreach activities to increase awareness of the project and related science. The project will fund the work of six graduate and eight undergraduate students across five institutions, furthering NSF’s goals of education and training.

Motivated by preliminary observations of unexplained, tightly-correlated DON and dissolved iron concentrations across the West Florida Shelf (WFS), the proposed work will quantify the flux and isotopic signatures of submarine groundwater discharge (SGD)-derived DON and iron to the WFS, and evaluate the bioavailability of this temporally-variable source using four seasonal near-shore campaigns sampling offshore groundwater wells, estuarine, and riverine endmembers and two cross-shelf cruises. The work will evaluate whether SGD stimulates nitrogen fixation on the WFS, and the potential for the stimulated nitrogen fixation to further modify the chemistry of DON and dissolved iron in the region. The cross-shelf cruises will investigate hypothesized periods of maximum SGD and Trichodesmium abundance (June), and reduced river discharge and SGD (February), thus comparing two distinct biogeochemical regimes. The concentrations and isotopic compositions of DON and dissolved iron, molecular composition of DON, and the concentration and composition of iron-binding ligands will be characterized. Nitrogen fixation rates and Trichodesmium spp. abundance and expression of iron stress genes will be measured. Fluxes of DON and iron from SGD and rivers will be quantified with radium isotope mass balances. The impacts of SGD on nitrogen fixation and DON/ligand production will be constrained with incubations of natural phytoplankton communities with submarine groundwater amendments. Two hypotheses will be tested: 1) SGD is the dominant source of bioavailable DON and dissolved iron on the WFS, and 2) SGD-alleviation of iron stress changes the dominant Trichodesmium species on the WFS, increases nitrogen fixation rates and modifies DON and iron composition. Overall, the work will establish connections between marine nitrogen and iron cycling and evaluate the potential for coastal inputs to modify water along the WFS before export to the Atlantic Ocean. This study will thus provide a framework to consider these boundary fluxes in oligotrophic coastal systems and the relative importance of rivers and SGD as sources of nitrogen and iron in other analogous locations, such as coastal systems in Australia, India, and Africa, where nitrogen fixation and SGD have also been documented.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.