Sample collection and processing
Suspended particulate organic matter (POM) samples were collected by McLane pumps according to GEOTRACES sampling protocols (Lam et al. 2018). Seawater was pumped through a 51-micrometer (μm) polyester prefilter and a Whatman QMA quartz filter (Cytiva). The particles collected on the QMA filter represent 1-51 μm size fraction of suspended POM. Up to 1500 liters (L) of seawater is pumped through each pair of filters, and we received a subsample equivalent to 1/16 of each QMA filter, representing 50-100 L of seawater.

Filters were frozen (-20 degrees Celsius (°C)) immediately after sample collection and returned to the laboratory for processing. Filters are first extracted with 10 milliliters (mL) MQ under ultrasonication at 35% power for 5 minutes (Branson Ultrasonics, model 102C). The MQ extract was collected and the filter was extracted again with 10 mL distilled MeOH under ultrasonication at 35% power for 10 minutes. Then, the MQ extract and MeOH extract were combined, and diluted with 200 mL MQ. The sample was pumped through a 0.2 μm PES filter capsule (P/N SLGPM33RS, MilliporeSigma) and a Bond-Elut ENV SPE cartridge (1 g, 6 mL, P/N 12255012, Agilent Technologies) that had been previously activated by passing 6 mL each of distilled methanol (MeOH, Optima LCMS grade, Fisher Scientific) and ultrapure water (qH2O, 18.2 MΩ) through the column.

Columns were thawed and washed with 6 mL qH2O (to reduce salts) and the qH2O wash was discarded. Ligands were then eluted with 6 mL distilled MeOH into acid-cleaned 10 mL polypropylene tubes. A 10 µL stock solution of 2.2 micromoles (µM) Ga-Desferrioxamine-E (Ga-DFOE) was added to each sample as an internal standard. The sample was concentrated to about 500 microliters (µL) by vacuum centrifugation (SpeedVac, Thermo Scientific; 35 °C, 5 hours). A 100 µL aliquot of the sample was taken, mixed with 100 µL of qH2O, and immediately analyzed by LC-MS.

High pressure liquid chromatography-Inductively coupled plasma mass spectrometry
Chromatographic analyses were performed on a bioinert Dionex Ultimate 3000 LC system fitted with a loading pump, a nano pump, and a 10-port switching valve (Li et al. 2021). During the loading phase, 200 µL of sample were withdrawn into the sample loop, then pushed onto a C18 trap column (3.5 μm, 0.5 mm x 35 mm, PN 5064-8260, Agilent Technologies) by the loading pump at 25 μL/min for 10 minutes. The loading solvent is a mixture of 95% solvent A (5 mM aqueous ammonium formate, Optima, Fisher Scientific) and 5% solvent B (5 mM methanolic ammonium formate). During the elution phase, the solvent was delivered by the nano pump at 10 µL/min, and the trap column outflow directed onto two C18 columns (3.5 μm, 0.5 mm x 150 mm, PN 5064-8262, Agilent Technologies) connected in series. Samples were separated with an 80-minute linear gradient from 95% solvent A and 5% solvent B to 95% solvent B, followed by isocratic elution at 95% solvent B for 10 minutes. Meanwhile, the loading pump solvent was switched to 100% qH2O, increased to 35 µL/min and directed as a post column make-up flow, which was infused with the column eluant into the ICPMS. The high aqueous content of the combined flow serves to minimize the effect of changes in solvent composition (in this case increasing methanol content during the analysis) on the detector response to Fe and Ga.

The combined flow from the LC was analyzed using a Thermo Scientific iCAP Q quadrupole mass spectrometer fitted with a perfluoroalkoxy micronebulizer (PFA-ST, Elemental Scientific), and a cyclonic spray chamber cooled to 4 °C (Boiteau and Repeta, 2015). Measurements were made in kinetic energy discrimination (KED) mode, with a helium collision gas flow of 4-4.5 mL/min to minimize isobaric 40Ar16O+ interferences on 56Fe. Oxygen was introduced into the sample carrier gas at 25 mL/min to prevent the formation of reduced organic deposits onto the ICPMS skimmer and sampling cones. Isotopes monitored were 56Fe, 54Fe, 57Fe, 69Ga and 71Ga.

External and Internal Standards
The Fe detector response was calibrated using the siderophore ferrichrome which elutes at ~ 40 minutes in our chromatographic analysis. Stock solutions of 250 µM of ferrichrome were diluted to prepare standards with 2 nM, 5 nM, 10 nM, 20 nM, and 40 nM of the siderophore. Then, 5 µL of 2.2 µM Ga-DFOE was added to 995 µL of each standard. Next, a 100 µL aliquot was taken, mixed with 100 µL of qH2O, and analyzed by LC-ICPMS. A plot of the ratio of Fe-56 (ferrichrome):Ga-69 (Ga-DFOE) peak areas against ferrichrome/Ga-DFOE concentration yields a relationship (r2 ~0.999) between 0.2-4 pmole of ferrichrome. Calibrations and process blanks were made for every 10-20 samples analyzed, with only small changes (RSD~30%) in the slope of the calibration relationship observed over the course of the ~2 years of sample analysis. Concentrations of iron ligands in each sample were measured by plotting the FeL/Ga-DFOE peak area on the most appropriate calibration curve.

High pressure liquid chromatography-Electrospray ionization mass spectrometry
To verify the assignment of Fe-Ls to known siderophores, samples were analyzed by LC-ESIMS. The eluant from the LC, without qH2O infusion, was coupled to a Thermo Scientific Orbitrap Fusion mass spectrometer equipped with a heated electrospray ionization source. ESI source parameters were set to a capillary voltage of 3500 V, sheath, auxiliary and sweep gas flow rates of 5, 2, and 0 (arbitrary units), and ion transfer tube and vaporizer temperatures of 275°C and 20°C. MS1 scans for a m/z range of 150-1900 were collected in high resolution (450K) positive ion mode.

The LC-ESIMS data was converted from raw file format to mzXML (MSconvert, Chambers, Maclean, Burke et al. 2012). The mzXML is imported to Matlab, and aligned with ICPMS data using the retention time of Ga-DFOE, which was obtained by monitoring m/z of 667.26 by ESIMS and 69Ga by ICPMS. Then, the m/z and intensity from each scan are extracted, and ordered by scan number into a scan number/mass (m/z)/intensity matrix, which is then interrogated by mass search algorithms (Boiteau and Repeta, 2015; Li et al. 2021). The algorithms find pairs of co-eluting peaks with a difference of 1.995 amu in m/z and a ratio of 15.7 in intensity, which represent isotopologues of Fe containing complexes.

Instrumentation
We used a Gilson Aspec GX-271 to recover samples from solid phase extraction columns. Extracted samples were reduced in volume using a Thermo/Savant RVT 1505 vacuum centrifuge. Concentrated samples were analyzed by high pressure liquid chromatography using a Dionex Ultimate 3000 (liquid chromatograph) coupled to a Thermo iCap QC inductively coupled plasma mass spectrometer or a Thermo Orbitrap Fusion mass spectrometer fitted with a heated electrospray interface.