Chlorophyll
In the dark, 250 mL of seawater was filtered onto 25 mm GF/F filters.  Samples were stored in the dark at -20C until analyzed according to protocols adapted from Strickland and Parsons (1972).  Briefly, samples were extracted in 90% acetone in the dark (4C, 9 hr) and measured using an 10AU fluorometer (Turner).  Sample signals were calibrated using a chlorophyll-a standard (Sigma C6144) and were corrected for pheopigments by accounting for the fluorescence of extracts before and after acidification to 0.003 M HCl.

Nitrogen Species
Concentrations of nitrate + nitrite were measured by chemiluminescence after reduction in a hot acidic vanadyl sulfate solution on a NOx analyzer (Braman and Hendrix, 1989). Concentrations of nitrite were quantified by using the Griess–Ilosvay method followed by measuring absorption 540 nm (Grasshoff et al., 1999), and nitrate was quantified by difference. Concentrations of ammonium were measured by fluorescence using the OPA method (Holmes et al., 1999).

Microbial Abundance
Seawater samples were preserved for flow cytometry with 0.5% glutaraldehyde (final concentration), flash frozen in liquid nitrogen and stored at -80°C until analysis. Bacteria and group-specific phytoplankton counts were conducted on a Guava EasyCyte HT flow cytometer (Millipore). Instrument-specific beads were used to calibrate the cytometer. Samples were analyzed at a low flow rate (0.24 µL s-1) for 3 min.  To enumerate bacteria, samples were diluted (1:100) with filtered seawater (0.01 µm).  Samples and filtered seawater blanks were stained with SYBR Green I (Invitrogen) according to the manufacturer’s instructions and incubated in a 96-well plate in the dark at room temperature for 1 hr.  Bacterial cells were counted based on diagnostic forward scatter vs. green fluorescence signals.  Major phytoplankton groups were distinguished based on plots of forward scatter vs. orange (phycoerythrin-containing Synechococcus sp.), and forward scatter vs. red (eukaryotes).  Size classes of eukaryotic phytoplankton were further distinguished based on forward scatter (pico-, nano- and large eukaryotes). 

Dissolved organic carbon
Filtered water samples for total dissolved organic carbon were pipetted into acid-washed combusted glass vials, acidified to pH = 2 with 12 M hydrochloric acid, and stored at 4 °C until analysis on a Shimadzu TOC-5050A total organic carbon analyzer. The coefficient of variability between replicate injections was less than 1%.

Reactive oxygen species
Water samples were collected directly from a trace metal clean rosette into acid-washed, opaque bottles and stored in a shipboard flow-through seawater incubator. A subset of each water sample was filtered (0.2 um), amended with 50 µM diethylene-triaminepentaacetic acid (DTPA, Sigma), and aged overnight in the shipboard seawater flow through incubator (termed AFSW). Additionally, particle associated superoxide signals were determined by filtering (0.2 um) a subset of each water sample and measuring the chemiluminescent signal within 25-30 minutes after filtering (termed FFSW). Superoxide signals were measured by pumping unfiltered water (UFSW), FFSW, or AFSW from dark bottles using a high accuracy peristaltic pump directly into a flow-through FeLume Mini system (Waterville Analytical, Waterville ME) within the ship laboratory. Superoxide detection was based on the reaction between superoxide and a chemiluminescent probe, a methyl cypridina luciferin analog (MCLA, Santa Cruz Biotechnology) (Rose et al., 2008) as before (Roe et al., 2016). The travel time of the water samples in the opaque FeLume tubing was approximately 20 sec. Data was collected for several minutes (~2-4 min) once a steady-state signal was achieved. At least 12 hours following filtering, the superoxide signals within the AFSW for each depth was measured to establish the baseline. At the end of each run, 800 U L-1 superoxide dismutase (SOD, Sigma) was added to seawater samples. The total dark superoxide signal produced in the seawater was defined as the difference between the UFSW signal (subtracted by the SOD baseline) and AFSW signal (subtracted by the SOD baseline) (similar to (Roe et al., 2016)). The particle associated signal was defined as the UFSW signal (subtracted by the SOD baseline) and FFSW signal (subtracted by the SOD baseline) (similar to (Roe et al., 2016)). While the SOD baseline has an autoxidation artifact, this artifact is canceled by taking the difference between two signals. The chemiluminescent signals were converted to superoxide concentration by conducted calibrations in the same aged-filtered seawater used for the baseline at each depth. Calibrations were conducted using potassium dioxide (Sigma) as detailed previously (Zhang et al., 2016). 

Designations are: 
Total superoxide = [(UFSW) - (UFSW-SOD)] - [(AFSW)- (AFSW-SOD)]
Particle superoxide = [(UFSW) - (UFSW-SOD)] - [(FFSW)- (FFSW-SOD)]

Hydrogen peroxide concentrations. Hydrogen peroxide concentration was quantified based on the oxidation of colorless AmplifluTM Red (AR, Sigma) to pink-colored resorufin by hydrogen peroxide. For hydrogen peroxide concentration analysis, pre-mixed AR and horse radish peroxidase (HRP) stock solution was added at a final concentration of 50 µmol L-1 AR and 1 kU L-1 HRP to filtered (0.2 um) and unfiltered seawater samples in a clear 96-well microplate (Zhang et al., 2016). Light absorbance was measured at 570 nm (Abs570, maximum absorbance of resorufin) and 700 nm (Abs700, to account for background absorbance) on a SpectraMax® M3 multi-mode microplate reader. The difference between Abs570 and Abs700 (i.e., Abs570-700) was used for calculating hydrogen peroxide concentrations in seawater samples based on a calibration. The calibration factor was determined by standard addition of hydrogen peroxide into 0.2-µm filtered seawater from each station and depth as described previously (Zhang et al., 2016). To account for autoxidation of AR, 200 kU L-1 catalase (Sigma) was added to the blanks prior to the addition of AR and HRP. The hydrogen peroxide concentrations in seawater samples were determined by applying the calibration factor to the blank-corrected Abs570-700values. Net and gross hydrogen peroxide concentrations were defined as the levels measured in unfiltered and filtered seawater samples, respectively. 

Manganese
Seawater samples (1 L) were filtered through 0.2 µm membrane filters (Millipore) within one hour of collection using acid-washed Savillex vacuum-filtration rigs. The filtrate was poured into new 15 mL falcon tubes and the filter was immediately amended with a leuco-based dye for Mn oxide concentration in a separate 15 mL falcon tube.

The leucoberbelin blue (LBB) assay for Mn oxides (denoted MnOx hereafter) was previously adapted from Altman (1972) to examine coastal water column sites (Oldham et al., 2015; 2017a, 2017b, 2017c). In this assay, the filter is amended with 3 mL of 20 µM LBB dye solution (LBB, Sigma-Aldrich). The dye color forms upon oxidation of the LBB molecule by Mn oxides and can be measured spectrophotometrically. The LBB stock solution was prepared by dissolving the powder in Milli-Q water to a concentration of 4 % and adding 40 µL of 10 M sodium hydroxide (NaOH) per 10 mL of stock solution. Working solutions are subsequently prepared by diluting the stock solution into 1 % acetic acid, to 0.4 % LBB. A calibration curve was generated using KMnO4, for which equivalent absorbance for Mn(IV) is calculated based on 2.5 more Mn(IV) being required relative to Mn(VII) to oxidize the LBB. In our set-up, a 100 cm pathlength cell (Liquid Waveguide Capillary Cell) was coupled to a Flame UV-Vis (Ocean Optics), set up with SpectraSuite software. Using a 100 cm pathlength cell allows for a detection limit of 0.2 nM but also requires our re-filtration of samples (0.2 µm luer-lok syringe filter, Millipore) prior to injection into the cell to avoid particulate interference and clogging of the capillary cell. Samples reacted with the LBB dye overnight in the dark prior to analysis, then absorbance at 623 nm was recorded.  If sample absorbance was too high, samples were diluted 10-20 times in Milli-Q water.

For soluble Mn speciation analysis, an established spectrophotometric porphyrin addition method was employed (Madison et al., 2011; Oldham et al., 2017a), which uses the ligand T(4-CP)P (or α, β, γ, δ-tetrakis(4-carboxyphenyl)porphine, to 2.33 x 10-7 M final sample cocentration). In this method, cadmium chloride (CdCl2; to 2.4 x 10-7 M) is added to form a complex with the porphyrin, in the presence of an imidazole tetraborate buffer (pH = 8.2). The sample is added to the mixture (diluted 10-fold with Milli-Q water to avoid chloride interference) and any Mn(II) in the sample undergoes a metal substitution reaction with the Cd over the course of a 1 hour reaction in a 90°C water bath. The solution is cooled, then analyzed using the 100-cm UV-Vis spectrophotometric set-up described above. Total dissolved Mn is analyzed in the same way, but after the addition of 1.4 µM hydroxylamine hydrochloride to the sample (reacted overnight in a refrigerator). The difference between the total dissolved Mn and the Mn(II) gives the Mn(III)-L in the sample. We note that during the heated reaction with no reducing agent, it is likely that some Mn(III)-L complexes undergo a ligand substitution reaction with the added porphyrin, and thus our method likely underestimates Mn(III)-L, particularly for weaker complexes. For all samples, assays were run in triplicate for both Mn(II) and Mn total, and peak height for all assays was determined using a baseline subtraction performed using ECD-Soft peak correction software.