Sampling:

Seawater samples were collected via a CTD rosette deployed to a maximum depth of 500 m during the AMT-30 transect cruise. Depth profile samples from the solar noon CTD Niskin bottles were taken every three days at 11 of the 54 total stations, with 12 samples per cast (250 mL each) ranging from 5 m to 500 m depth for a total of 132 samples for iodide analysis along the transect. Sample containers were rinsed 3 times with sample seawater prior to filling. Samples were filtered to remove bacteria and other particles through 0.8 μm pre-filters followed by 0.2 μm filters (AcropakTM 1500 Supor Capsule, Pall Corporation) using a Masterflex pump and then placed into opaque 60 mL bottles and frozen at -20°C (Campos et al., 1996).

Additionally, larger volume, unfiltered samples (1-3 L) were collected from depths corresponding to light levels of 7% and 1% (the extent of the DCM) for use in the incubation experiments. The iodine isotope (mostly ¹²⁹I^-) spike was added to the larger volume first to homogenize the iodine speciation and respective isotope ratio before aliquoting into replicates. The larger volumes were spiked to a target concentration of 70 nM ¹²⁹I^- solution (Eckert and Ziegler Isotope Products^©) (Schnur et al., 2024, Hardisty et al., 2020, Ştreangă et al., 2024, Hardisty et al., 2021). Notably, the isotope spike included a NaI carrier, which also contained ¹²⁷I^-, and was added at a volume so total iodide added was approximately 140 nM (¹²⁹I/¹²⁷I of I^- ~1).

We highlight that, while the spike is mostly iodide, it has trace amounts of iodate resulting in detectable ¹²⁹I/¹²⁷I for the iodate pool in each experiment. This effect is similar to previous studies and is maintained in order to prime the iodate pool with ¹²⁹I so that increases and decreases in the ¹²⁹I/¹²⁷I isotope ratio are detectable via MC-ICP-MS, which cannot resolve natural background ¹²⁹I/¹²⁷I ratios, which are on the order of 10^-12 (Qi et al., 2024, Qi et al., 2023, Nagai et al., 2015, Wefing et al., 2019, Casacuberta et al., 2018, Zhang et al., 2023, He et al., 2013).

Each spiked carboy was split into triplicate 250 mL incubations (Schnur et al., 2024). Incubations occurred within on-deck flow-through incubators with screens replicating the light levels and chillers maintaining temperatures of the sample depths. Samples for t₀ were immediately subsampled after adding the spike. All subsamples were filtered at 0.2 μm to end interaction with biology/particles, put into amber high-density polyethylene (HDPE) Nalgene bottles, and frozen at -20°C. All subsamples (t₀, t₁, t₂) were ~60 mL. Incubation conditions were monitored for phytoplankton and microbial analysis onboard (see next section). Incubations at the beginning southernmost leg of the cruise were run for ~6 days and this was shortened based on taxonomic analysis indicating cell losses on longer timescales. All samples were stored frozen (at -20°C) until analysis at Michigan State University.

Analytical Methods:

Methods are adapted from Hardisty et al., (2020) and Schnur et al., (2024) and are summarized below.

Iodide concentration was measured in depth profiles and iodide and iodate concentration were measured in incubation time points using an established ion-exchange chromatography protocol from Hardisty et al., (2020) and Schnur et al., (2024) and summarized below.

The iodine speciation was conducted using glass columns packed with PYREX glass wool and 1 mL of AG1-X8 resin, which were pre-cleaned to eliminate residual iodine before sample processing. Iodide was eluted from the seawater matrix after iodate and a dissolved organic iodine (DOI) were released from the resin. Iodate and DOI fractions were collected independently, but DOI was not measured. The iodate fraction was then reduced to iodide using concentrated hydrochloric acid (HCl) and 0.3M sodium bisulfite (NaHSO₃). This fraction underwent a second round of chromatography on cleaned resin as performed previously for iodide.

For quality control, a 200ppb iodide solution (diluted from a 1000 ± 4 μg mL^-1 iodide standard in 1% tetraethylammonium (TEA)) or dissolved solid KIO₃ in 18.2 MΩ·cm water was processed through the columns alongside the samples to assess elution efficiency and yield of iodide and iodate, respectively. 18.2 MΩ·cm water blanks were included for each column set to check for contamination, and at least one replicate sample was processed in each column set to evaluate reproducibility. Iodide concentrations [¹²⁷I⁻] were measured in diluted samples at Michigan State University using a Thermo Scientific iCap triple-quad inductively coupled plasma mass spectrometer (ICP-MS-TQ), in both single-quad (SQ) and triple-quad (TQ) modes with O₂ reaction cell gas. Data correction was performed using internal standards (In, Rh, and Cs) from Inorganic Ventures^©. Multi-point calibration curves and column standards were serially diluted from 1000 ± 4 μg mL^-1 iodide standard in 1% TEA. Iodate recovery yields from the processed KIO₃ standard solutions were 90-95%, while iodide yields were near-complete (~100%).

Iodine isotope ratios (¹²⁹I/¹²⁷I) were analyzed at the Woods Hole Oceanographic Institution (WHOI) using a Thermo Finnegan Neptune MC-ICP-MS, following previous methods , which are summarized below. These measurements were performed on aliquots of the eluted iodine species (I^- and IO₃^-) recovered during the ion exchange chromatography used for ICP-MS concentration analysis. Each day before sample analyses were performed, the instrument was tuned to maximize beam intensity for accuracy of the instrument. Seven Faraday cups L3, L2, L1, C, H1, H2 and H3 were used to monitor the ¹²⁶Te, ¹²⁷I, ¹²⁸Te, ¹²⁹I, ¹³⁰Te, ¹³¹Xe and ¹³²Xe isotopes, respectively. Mass bias corrections were applied using a 500 ppb Te solution (Inorganic Ventures^©), and potential isobaric interferences on both Te and I were tracked by monitoring ¹³¹Xe over the course of sample analysis. Iodine samples were introduced into the instrument using a gas-based "sparge" method in which an inert Ar carrier gas containing Te originating from Aridus II desolvation system is bubbled through the sample where volatilized iodine is carried with the Ar and Te into the MC-ICP-MS plasma. The collected data were corrected for interferences and instrumental mass bias to yield a final ¹²⁹I/¹²⁷I ratio and paired standard deviation.