Incubation Setup:

Surface water was collected for shipboard incubations in June 2022 aboard the R/V Sikuliaq using a trace metal clean surface pump “towfish” system (Mellett and Buck, 2020). Filtered (<0.2 µm, Acropak) seawater from the towfish was homogenized in three acid-cleaned and seawater rinsed 50-L carboys that were filled round-robin style (Burns et al., 2023). Each carboy was then bubbled overnight with a custom CO2-air mixture to achieve the target pH levels of pH 8.1, 7.6, and 7.1, which was verified with shipboard spectrophotometric pH analyses using the Byrne MICA system (Adornato et al., 2016). Unfiltered surface seawater was then collected and homogenized in a fourth acid-cleaned and seawater-rinsed carboy using the trace metal clean towfish. Trace metal clean polycarbonate incubation bottles were then filled two-thirds with filtered seawater and one-third unfiltered seawater, amended for the nutrient and/or iron treatment, sealed with the caps/threads wrapped in parafilm and electrical tape, and delivered to deckboard flow-through seawater incubators that were covered in screening to mimic surface light levels. Once all incubation bottles were in the incubators, the time-zero sampling of each incubation began. Six incubations were conducted, two each at a coastal upwelling station (Inc 1, 2; 40.112 ºN, 125.56 ºW), in the oligotrophic central North Pacific (Inc 3, 4; 35 ºN, 145 ºW), and at Ocean Station PAPA (Inc 5, 6; 50 ºN, 145 ºW) in the subarctic North Pacific. For incubations 1-4, all incubation bottles were spiked with chelexed stocks of nitrate and phosphate, and aged (for trace metal cleanliness) silicic acid stocks, to target additions of 10 µM nitrate, silicic acid, and 0.8 µM phosphate; no macronutrients were added to Incs 5 and 6, which were already macronutrient replete. Replicates of pH treatment were additionally spiked with 1 nM ⁵⁷FeCl₃ as a dissolved iron addition. Incubation bottles were labeled according to treatment and were the same across light bottles all incubations: A = pH 8.1, B = pH 8.1 + Fe, C = pH 7.6, D = pH 7.6 + Fe, E = pH 7.1, F = pH 7.1 + Fe. Replicates of each treatment were also incubated in heavy duty black contractor bags to serve as dark controls (G = A, H = B, I = C, J = D, K = E, L = F), which were sampled on day final only.

Incubation sampling:

Triplicate bottles from each incubation were sampled daily over the course of the experiments. Incubation bottles were brought in from the incubators into a clean lab bubble in the ship, where they were washed down with Milli-Q and transferred into a clean hood. After gently inverting to mix, each was subsampled for pH, chlorophyll a, and particulate organic nutrients. The remaining contents of each bottle were filtered in the bubble clean hood on a custom acrylic filtration rig outfitted with dual stage Teflon filtration holders (Savillex) that allows the filtrate to go directly into sample bottles after passing through consecutive 5 µm and 0.4 µm acid-cleaned polycarbonate track-etched (PCTE; Whatman) filters. Samples for dissolved trace metals were collected in acid-cleaned and triple-rinsed narrow mouth low density polyethylene bottles, acidified with 0.024 M ultrapure hydrochloric acid (to pH ~1.8), and stored for shore-based analysis at the University of Nagasaki (Yoshiko Kondo). Samples for dissolved iron and nickel speciation were collected in acid-cleaned, Milli-Q-conditioned, and triple-rinsed narrow mouth fluorinated high density polyethylene bottles (Nalgene) and analyzed shipboard for dissolved iron speciation (Lise Artigue, Kristen Buck lab) before freezing at -20 ºC for shore-based dissolved nickel speciation analyses at Oregon State University (Matthew Koteskey, Kristen Buck lab). Samples for pH were analyzed shipboard (Drajed Seto, Mark Wells lab).

Sample analyses – dissolved trace metals:

The concentrations of dissolved iron, manganese, nickel, zinc, and copper were analyzed by high resolution inductively coupled plasma mass spectrometry (Thermo Scientific ELEMENT II) with a preconcentration flow injection system seaFAST-pico (Elemental Scientific Inc., ESI) at Nagasaki University (Yoshiko Kondo and Shigenobu Takeda). Acidified samples were measured without UV-oxidation, and dissolved copper concentrations should be considered ‘reactive Cu’ as total recovery may have been hindered by organic complexation in these samples. Briefly, dissolved trace metals in the samples were preconcentrated on a Nobias-chelate PA1 resin, eluted with 2 M HNO3, and quantified by calibration curve prepared with SAFe and GEOTRACES reference samples (S1, GS, and GD) (https://www.geotraces.org/standards-and-reference-materials/).

A subset of incubation samples were analyzed for total dissolved nickel concentrations by competitive ligand exchange-adsorptive cathodic stripping voltammetry following UV-oxidation of remaining volume in nickel speciation samples. Frozen seawater samples were thawed at room temperature in the lab at Oregon State University and UV-oxidized in Teflon jars (Savillex) with quartz lids for at least ninety minutes in a Jelight model 342 UVO cleaner. Following UV-oxidation, seawater sample aliquots were buffered with a borate-ammonium buffer and amended with 200 µM dimethylglyoxime to complex the dissolved nickel in the sample and form an electroactive complex, which was then measured by standard addition on a hanging mercury drop electrode (BioAnalytical Systems, Inc.).

Sample analyses – labile dissolved nickel:

The concentration of labile dissolved nickel concentrations was measured by competitive ligand exchange-adsorptive cathodic stripping voltammetry using the added ligand dimethylglyoxime (DMG; van den Berg and Nimmo 1987) and following a modification of previously described procedures (Saito et al. 2004; Boiteau et al. 2016). Briefly, seawater sample aliquots were buffered with a borate-ammonium buffer and equilibrated overnight with 200 µM DMG. Following equilibration, the amount of dissolved nickel in the samples that was bound to DMG was measured on a hanging mercury drop electrode and quantified by standard additions of dissolved nickel to the sample. All measurements, of the sample and of the standard additions, were conducted in triplicate. The concentration of labile dissolved nickel was determined from the slope of the standard curve and the triplicate measurements of the initial sample, and the results presented as averages and standard deviations of the three values.

Sample analyses – pH:

pH was measured using a USB4000 fiber optic spectrometer (Ocean Optics) with purified meta-Cresol Purple (mCP) as the pH indicator dye (Liu et al., 2011). The system comprised a open top, flow-thru cell positioned in a temperature controlled (20·°C) water bath. The cell was zeroed by manually injecting a blank or reference sample (seawater without mCP) and recording the absorbance at 434, 578, and 700 nm (reference). For sample analysis, 1 µL of purified mCP indicator solution was drawn into a clean 3 ml syringe followed by 2 ml of seawater sample. The solution was mixed gently to ensure uniform distribution of the indicator while avoiding air bubble formation. The solution then was manually injected into the flow-thru cell (using excess volumes for rinse) and allowed to thermally equilibrate. Once absorbance values had stabilized (1-3 min) the values were recorded at 434, 578, and 700 nm. Seawater pH was calculated on the total scale using the absorbance ratio (578/434) according to Liu et al. (2011). All samples were analyzed in triplicate and the results presented as the averages and standard deviations of the three values.