Samples were collected from 5 L Teflon-coated Niskin-X bottles hung on non-metal line, and clean techniques were followed throughout (Bruland et al., 1979). Niskin-X bottles were transferred to a clean lab immediately after retrieval. At most stations a single sample was collected from 20 m. Four-depth profiles extending to 200 m were collected at 6 stations along the transect.

For dissolved metal analysis, 300 mL of seawater was filtered through acid-washed 47 mm 0.4 μm polycarbonate filters into 250 mL LDPE bottles which were acid-cleaned following GEOTRACES protocols (Cutter et al., 2010). A vacuum filtration apparatus was used, with all parts in contact with the sample composed of acid-washed PFA. The assembly was rinsed with 3 aliquots of sample prior to sample collection. Samples were acidified with Optima HCl to 0.024 M, double bagged, and stored for analysis on shore.

Dissolved samples were analyzed using an ESI seaFAST SP2 coupled to a Perkin Elmer Nexion 350D ICP-MS. Sample aliquots were transferred to 30 mL LDPE bottles. Approximately 8 mL of sample was taken up to rinse and fill a 6 mL preconcentration loop. Samples were buffered in-line with resin-cleaned 4 M ammonium acetate buffer-adjusted to pH = 6.0 ± 0.1 and passed through an ESI pre-concentration column with a 200 μL resin bed volume. The final concentration of the buffer in the samples was 0.88 M. Metals were eluted off the column with 200 μL of 1.6 M Optima-grade HNO₃. Eluted metals passed through a PFA nebulizer, quartz spray chamber, and nickel cones on their way to an ammonia-filled dynamic reaction cell (0.55 mL/min flow).

Quantification was achieved by standard additions automated by the seaFAST system. A standard addition solution of 6.9 nM Fe, 46 nM Mn, 4.5 nM Zn, and 1.1 nM Pb was made in 0.024 M HCl. Concentrations in this solution were quantified using an external curve. Additions were introduced as the sample filled the preconcentration loop. The volume introduced by standard additions was kept constant, with 0.024 M Optima HCl making up the difference for samples with no additions. Each sample was spiked with 2 additions averaging roughly 100% and 200% of the sample concentration.

Approximately 4L of water from Niskin bottles was filtered for particulate metals. The bottles were pressurized to approximately 5 psi with 0.2 μm filtered air and seawater passed through 25 mm 0.4 μm Supor filter membranes. The membranes were housed in Swinnex polypropylene filter holders and attached to the Niskin-X bottles with Teflon tubing and polycarbonate fittings. The membrane was transferred to an acid washed petri-slide, double bagged, and kept frozen at -20°C for analysis on shore.

Labile and total particulate concentrations were analyzed through sequential digestion of the Supor membranes following protocol of Rauschenberg and Twining (2015). Membranes were first leached using an acetic acid-hydroxylamine solution (Berger et al., 2008) then transferred to a PFA vial and digested with a 4M HCl/4M HNO3/4M HF mixture (Ohnemus et al., 2014). Reference materials and process blanks – filters that had 2 L of 0.2 μm filtered seawater passed through them –were sequentially digested alongside the samples.

Digests were analyzed using a Thermo Element2 HR-ICP-MS equipped with a quartz nebulizer, cyclonic spray chamber, and nickel cones. Cd was analyzed in low resolution while all other elements were analyzed in medium resolution. In-115 was used as a drift monitor, and quantification was performed using an external curve.