Samples were taken during cruise R/V Atlantis AT39-01. ROV Jason II dives 1024 – 1035, IODP CORK observatories U1382A, U1383C

Detailed methodology for sample collection, shipboard procedures and sample storage conditions provided in BCO-DMO dataset “Carbon Geochemistry Samples List” (see related datasets)

Briefly, DIC samples for isotopic analysis were transferred from sampling bags through Masterflex Bio-Pharm silicone tubing pre-cleaned with 10% HCl to 100 mL Wheaton wide-mouth reagent bottles with ground-glass stoppers such that turbulence during fluid transfer was minimized. Samples were preserved according to NOSAMS DIC sampling protocol: http://www.whoi.edu/fileserver.do?id=75006&pt=2&p=75096 and stored at room temperature in the dark. DIC samples were submitted to the NOSAMS AMS facility for analysis (https://www2.whoi.edu/site/nosams/). DIC was extracted from fluid samples by acidification and water stripping. d13C values are measured on CO2 splits by IRMS and reported with a conservative uncertainty of 0.1‰, higher than each individual measurement. D14C values are measured by AMS and uncertainty is reported for each individual measurement. All isotopic measurements and error estimations were made according to standard NOSAMS protocols.

For fluids sampled in 2012, DIC concentration is measured as CO2 yield after acidification and water stripping reported with an uncertainty of 5%. For fluids sampled in 2014, DIC concentrations are calculated from pH and alkalinity concentrations reported by Wheat et al. (2020) using Matlab package CO2SYS (https://github.com/jonathansharp/CO2-System-Extd). Propagated uncertainty for calculated DIC concentrations is 1%. For fluids sampled in 2017, DIC was measured on an AS-C3 analyzer (Apollo Scitech, Newark, DE, USA). After acidification with phosphoric acid, the evolved CO2 was extracted and carried by pure N2 gas to an infrared CO2 detector (Li-Cor 6262). Total alkalinity was measured by Gran titration with an open-cell titration system (AS-ALK-2, Apollo SciTech, Newark, DE, USA). Alkalinity and DIC analyses were calibrated with Certified Reference Materials (A. G. Dickson, Scripps Institution of Oceanography) and measured in duplicate with repeat concentrations within 0.1% and measurement uncertainty of 2 umol/kg for both parameters on each bottle. pH values were calculated from measured alkalinity and DIC concentrations using Matlab package CO2SYS with propagated uncertainty of pH values is ± 0.01.

DOC samples were transferred to 1000-mL amber glass bottles. A parallel set of samples was filtered through pre-combusted GF/F filters to remove particulate matter but these samples yielded more carbon than observed in previous years and it was determined that they were likely to be contaminated. Values reported are from whole, unfiltered fluids which yielded less CO2 than filtered fluids supporting our hypothesis that filtered samples were contaminated. Particles were not observed in the whole fluids during transfer. Frozen DOC samples were submitted to the NOSAMS AMS facility for analysis (https://www2.whoi.edu/site/nosams/). DOC concentrations are measured as CO2 yield after UV-oxidation of samples. d13C values are measured on CO2 splits by IRMS and original values reported with a conservative uncertainty of 0.5‰, higher than each individual measurement. D14C values are measured by AMS and uncertainty is reported for each individual measurement. All measurements and error estimations were made according to standard NOSAMS protocols. NOSAMS revised all previously reported DOC data in 2019 according to new internal estimation and correction for carbon blank in their DOC oxidation procedure. Original and corrected DOC data are both reported in the dataset for samples measured before 2019. For corrected d13C values, individual error estimates are reported as well.

Procedures for 13C and 14C analysis of DIC and DOC as well as concentrations of DOC are described in Shah Walter, S. R., et al. (2018).

Alkalinity and DIC analyses were calibrated with Certified Reference Materials (A. G. Dickson, Scripps Institution of Oceanography)