To characterize the nutritional values of Southeast Pacific Ocean waters and determine the limiting nutrient(s) in each regime, uncontaminated seawater samples (n=173) were drawn from 21 deployments of nine 5-liter (L) Niskin-X bottles suspended from Kevlar line at varying depths between 30 meters (m) and 1050 m.

Sampling procedure:
Niskin-X bottles were carried from a UNOLS trace-metal clean van to the starboard squirt boom, where they were secured by hand to the 1/4-inch Aracom Miniline using stainless steel hardware. The Niskins were attached incrementally so that each Niskin was lowered to its target depth when the deepest Niskin reached its target depth (~1000 m). Niskin bottles were allowed to sit at depth for ~5 minutes before triggered to close using a series of Teflon messengers.

The Niskins were recovered and returned to the trace metal clean van, where they were secured in PVC racks with wedges to keep the caps tight. A series of air lines were connected to the air ports at the top of the Niskins, which were then overpressured (~6-8 psi) with filtered air using a Gast pump.

Unfiltered samples were drawn from each Niskin-X for salinity and major nutrient measurements (ODF). The remainder of the volume was filtered using 47-millimeter (mm), 0.4-micrometer (μm) pore-size Isopore membrane filters installed in Advantec filter cartridges. The filtrate was collected into trace metal-cleaned (acid-washed) 125-milliliter (mL) LDPE bottles for dissolved trace metal concentrations.

Special modifications to sampling operations:

1. A small Ronstan Series 60 Single Orbit block (model #7298755) was used instead of a larger metering block, which eliminated the risk of the line jumping the sheave and the need for a deck-mounted snatch block. The block was attached to the starboard squirt boom which allowed for the Niskin-X bottles to be attached to the Aracom line without having to lean too far over the side of the ship.

2. The Niskin-X bottles were deployed from the starboard side of the ship by securing the Hawboldt winch at an angle, just aft of the ResTech locker. This allowed the Aracom line to be threaded through the Ronstan block mounted to the squirt boom. Even in heavier seas, the Niskin-X bottles were deployed reliably and reproducibly due to this arrangement (which was designed by Matt Durham, UCSD ResTech).

3. In order to better estimate the depths at which the bottles were deployed, a miniature temperature and pressure sensor (centi-TD, 5-1500 m; Star Oddi) was secured to each Niskin-X bottle. These sensors recorded the temperature and pressure (i.e., depth) at preset intervals (every 60 seconds). These depths were double-checked by comparing the salinity and nutrient values determined in samples from each Niskin-X bottle against (1) the temperature and salinity values from the ship's CTD sensors and (2) discrete salinity and nutrient concentrations determined from the ship's bottles.

Analytical procedure (dissolved trace metals):
Filtered seawater samples were acidified to 0.024 M HCl using concentrated Fisher Optima HCl. The samples were allowed to sit for ~6 months to allow complete desorption of metals from the bottle walls, according to Jensen et al. 2020.

Aliquots of acidified seawater (~15 mL) were transferred to 30 mL FEP bottles (Savillex) for UV-oxidation (1.5 hours), according to Milne et al. 2010. The samples were then processed online using a SeaFAST S2 system (Elemental Scientific, Inc.), which includes online buffering with ammonium acetate buffer before passing over a 200-microliter (µL) column filled with Nobias Chelate PA-1. The extraction procedure captures trace metals from 10 mL of buffered seawater while allowing most of the major sea-salt cations (e.g., Na and Mg) to be directed to waste. The column is then eluted with 0.5 mL of 1.6 M HNO3 (Fisher Optima) to release the trace metals, and this eluate is introduced directly to the High Resolution-Inductively Coupled Plasma-Mass Spectrometer (Thermo ELEMENT 2) for analysis. By eliminating the major seasalts and concentrating the trace metals by a factor of ~20 (10 mL sample loaded, 0.5 mL eluted), the matrix effects are minimized, and the metal detection limits are increased.

Metal concentrations were quantified upon comparison against external standards prepared in low metal surface seawater, which were processed identically to the samples. Blanks (n=93) were determined by running a complete loading-extraction-elution cycle without seawater (i.e., leaving the sample probe out of any sample or rinse solution, taking up only HEPA-filtered air). These blanks were applied to all samples and reference materials to determine any background metal contributions from the SeaFAST system, ICPMS sample introduction components, or reagent blanks.

Reference materials were included in every analytical run, including the GEOTRACES GS (n=9) and GD (n=3) consensus reference materials, as well as an in-house "daily check" sample of surface seawater collected from the Gulf of Mexico (n=18).

Details of the reference materials and daily check sample are included with the air blanks in the "QA/QC" summary file "BCO-DMO_SAMW_2021_dTM_QAQC_2024_06_04" (PDF and Excel formats provided; see Supplemental Files section of metadata).