Sampling and analytical methodology:

Sample collection and filtering: Water column samples were collected from a trace metal rosette outfitted with Teflon-coated X-Niskin samplers (Ocean Test Equipment) and deployed on a nonmetallic line. Samples were filtered through acid-cleaned 142-mm polycarbonate filters with a 0.4 µm pore size (Geotech Environmental) in a laminar flow clean hood. Please see Noble et al. (2012) for detailed information on sample acquisition and processing.

Filtered seawater samples for iron speciation (organic complexation) were collected in 500 mL acid-cleaned Nalgene narrow-mouth fluorinated high-density polyethylene (FPE) bottles that had been filled with Milli-Q for at least two weeks after acid cleaning and rinsed three times with filtered seawater prior to sample collection (Buck et al. 2012). All filtered seawater samples for dissolved iron speciation were frozen at -20 ºC shipboard by the sampling team and shipped to the Bermuda Institute of Ocean Sciences for laboratory-based analyses post cruise.

Dissolved iron speciation (organic complexation) analyses: Dissolved Fe speciation (organic complexation) was analyzed using a competitive ligand exchange- adsorptive cathodic stripping voltammetry (CLE-ACSV) method with salicylaldoxime as the added competing ligand (Buck et al. 2007, 2012), modified from the original Rue and Bruland (1995) method. Dissolved Fe additions of 0 to 7.5 nM were used in the titrations, for a total of 12 points in each titration. Deposition times of 120 s were applied to the analyses. All analyses were performed on Bioanalytical Systems (BASi) Controlled Growth Mercury Electrodes (static mercury drop setting, size 14) with Epsilon e2 (BASi) electrochemical analyzers. There are no reference samples available for iron speciation/ organic complexation measurements of iron in seawater, though this method was shown to compare well with other labs and techniques in the GEOTRACES intercalibration exercises (Buck et al. 2012, 2016).

FLAG: The standard Ocean Data View qualifying flags were used (reference all flags at https://www.bodc.ac.uk/data/codes_and_formats/odv_format/):

1: Good Value: Good quality data value that is not part of any identified malfunction and has been verified as consistent with real phenomena during the quality control process. [Used when data were shown to be reproducible]

2: Probably Good Value: Data value that is probably consistent with real phenomena but this is unconfirmed or data value forming part of a malfunction that is considered too small to affect the overall quality of the data object of which it is a part. [Used when the reported value reflects analysis of a single or unverified replicate]

3: Probably Bad Value: Data value recognized as unusual during quality control that forms part of a feature that is probably inconsistent with real phenomena. [Not used]

4: Bad Value: An obviously erroneous data value. [Not used]

5: Changed Value: Data value adjusted during quality control. [Not used]

6: Value Below Detection Limit: The level of the measured phenomenon was too small to be quantified by the technique employed to measure it. The accompanying value is the detection limit for the technique or zero if that value is unknown. [We report “not_detected” rather than zero or a detection limit value for this case]

7: Value in Excess: The level of the measured phenomenon was too large to be quantified by the technique employed to measure it. The accompanying value is the measurement limit for the technique. [Not used]

8: Interpolated Value: This value has been derived by interpolation from other values in the data object. [Not used]

9: Missing Value: The data value is missing. Any accompanying value will be a magic number representing absent data. [Not used]